{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 32607,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.0667284838855267,
      "epoch": 9.20063714412223e-05,
      "grad_norm": 15.893442153930664,
      "learning_rate": 0.0001,
      "loss": 1.3744,
      "mean_token_accuracy": 0.6716951392591,
      "num_tokens": 32043.0,
      "step": 1
    },
    {
      "entropy": 2.8813244104385376,
      "epoch": 0.0001840127428824446,
      "grad_norm": 7.945219993591309,
      "learning_rate": 9.999693317385837e-05,
      "loss": 3.2837,
      "mean_token_accuracy": 0.45890693739056587,
      "num_tokens": 64000.0,
      "step": 2
    },
    {
      "entropy": 1.9456694647669792,
      "epoch": 0.0002760191143236669,
      "grad_norm": 1.9805350303649902,
      "learning_rate": 9.999386634771675e-05,
      "loss": 2.225,
      "mean_token_accuracy": 0.5570643804967403,
      "num_tokens": 96394.0,
      "step": 3
    },
    {
      "entropy": 1.491881251335144,
      "epoch": 0.0003680254857648892,
      "grad_norm": 0.6315288543701172,
      "learning_rate": 9.999079952157514e-05,
      "loss": 1.6004,
      "mean_token_accuracy": 0.6262646466493607,
      "num_tokens": 128843.0,
      "step": 4
    },
    {
      "entropy": 1.204836416989565,
      "epoch": 0.0004600318572061115,
      "grad_norm": 0.5164754390716553,
      "learning_rate": 9.99877326954335e-05,
      "loss": 1.4138,
      "mean_token_accuracy": 0.6705983318388462,
      "num_tokens": 161036.0,
      "step": 5
    },
    {
      "entropy": 1.1962174996733665,
      "epoch": 0.0005520382286473338,
      "grad_norm": 2.0098628997802734,
      "learning_rate": 9.998466586929187e-05,
      "loss": 1.6444,
      "mean_token_accuracy": 0.639203991740942,
      "num_tokens": 192703.0,
      "step": 6
    },
    {
      "entropy": 1.1457996368408203,
      "epoch": 0.0006440446000885561,
      "grad_norm": 0.3965838849544525,
      "learning_rate": 9.998159904315025e-05,
      "loss": 1.3046,
      "mean_token_accuracy": 0.6767523474991322,
      "num_tokens": 224101.0,
      "step": 7
    },
    {
      "entropy": 1.1448344737291336,
      "epoch": 0.0007360509715297784,
      "grad_norm": 1.4503222703933716,
      "learning_rate": 9.997853221700862e-05,
      "loss": 1.3455,
      "mean_token_accuracy": 0.6773927994072437,
      "num_tokens": 255025.0,
      "step": 8
    },
    {
      "entropy": 1.2764364778995514,
      "epoch": 0.0008280573429710008,
      "grad_norm": 0.35333123803138733,
      "learning_rate": 9.9975465390867e-05,
      "loss": 1.3044,
      "mean_token_accuracy": 0.6737880185246468,
      "num_tokens": 286786.0,
      "step": 9
    },
    {
      "entropy": 1.2712209224700928,
      "epoch": 0.000920063714412223,
      "grad_norm": 0.26453709602355957,
      "learning_rate": 9.997239856472537e-05,
      "loss": 1.2955,
      "mean_token_accuracy": 0.6700094677507877,
      "num_tokens": 318609.0,
      "step": 10
    },
    {
      "entropy": 1.1138665080070496,
      "epoch": 0.0010120700858534453,
      "grad_norm": 0.27728816866874695,
      "learning_rate": 9.996933173858375e-05,
      "loss": 1.1502,
      "mean_token_accuracy": 0.7026896476745605,
      "num_tokens": 350295.0,
      "step": 11
    },
    {
      "entropy": 1.1433785781264305,
      "epoch": 0.0011040764572946677,
      "grad_norm": 0.22404825687408447,
      "learning_rate": 9.996626491244212e-05,
      "loss": 1.2029,
      "mean_token_accuracy": 0.6917385309934616,
      "num_tokens": 382004.0,
      "step": 12
    },
    {
      "entropy": 1.1485241651535034,
      "epoch": 0.00119608282873589,
      "grad_norm": 0.21652007102966309,
      "learning_rate": 9.996319808630049e-05,
      "loss": 1.2024,
      "mean_token_accuracy": 0.6947129927575588,
      "num_tokens": 414590.0,
      "step": 13
    },
    {
      "entropy": 1.1348076686263084,
      "epoch": 0.0012880892001771122,
      "grad_norm": 0.2439424693584442,
      "learning_rate": 9.996013126015887e-05,
      "loss": 1.121,
      "mean_token_accuracy": 0.7054240368306637,
      "num_tokens": 447045.0,
      "step": 14
    },
    {
      "entropy": 1.0515992045402527,
      "epoch": 0.0013800955716183345,
      "grad_norm": 0.24095486104488373,
      "learning_rate": 9.995706443401725e-05,
      "loss": 1.0491,
      "mean_token_accuracy": 0.7255404032766819,
      "num_tokens": 478415.0,
      "step": 15
    },
    {
      "entropy": 1.1961029656231403,
      "epoch": 0.0014721019430595568,
      "grad_norm": 0.23444510996341705,
      "learning_rate": 9.995399760787562e-05,
      "loss": 1.2147,
      "mean_token_accuracy": 0.6856391355395317,
      "num_tokens": 510753.0,
      "step": 16
    },
    {
      "entropy": 1.0586777180433273,
      "epoch": 0.0015641083145007792,
      "grad_norm": 0.21574203670024872,
      "learning_rate": 9.995093078173399e-05,
      "loss": 1.0934,
      "mean_token_accuracy": 0.7118845842778683,
      "num_tokens": 542882.0,
      "step": 17
    },
    {
      "entropy": 1.2174208275973797,
      "epoch": 0.0016561146859420015,
      "grad_norm": 0.23548097908496857,
      "learning_rate": 9.994786395559235e-05,
      "loss": 1.2737,
      "mean_token_accuracy": 0.6743902489542961,
      "num_tokens": 574353.0,
      "step": 18
    },
    {
      "entropy": 1.1256860978901386,
      "epoch": 0.0017481210573832239,
      "grad_norm": 0.2055496722459793,
      "learning_rate": 9.994479712945073e-05,
      "loss": 1.1642,
      "mean_token_accuracy": 0.6958713382482529,
      "num_tokens": 606581.0,
      "step": 19
    },
    {
      "entropy": 1.0308941081166267,
      "epoch": 0.001840127428824446,
      "grad_norm": 0.20362018048763275,
      "learning_rate": 9.994173030330912e-05,
      "loss": 1.0365,
      "mean_token_accuracy": 0.7224165089428425,
      "num_tokens": 638668.0,
      "step": 20
    },
    {
      "entropy": 1.2041791826486588,
      "epoch": 0.0019321338002656683,
      "grad_norm": 0.31872403621673584,
      "learning_rate": 9.993866347716748e-05,
      "loss": 1.2492,
      "mean_token_accuracy": 0.685604389756918,
      "num_tokens": 670104.0,
      "step": 21
    },
    {
      "entropy": 1.1800084374845028,
      "epoch": 0.0020241401717068907,
      "grad_norm": 0.2158932238817215,
      "learning_rate": 9.993559665102585e-05,
      "loss": 1.1974,
      "mean_token_accuracy": 0.6903716772794724,
      "num_tokens": 702052.0,
      "step": 22
    },
    {
      "entropy": 1.1455806605517864,
      "epoch": 0.002116146543148113,
      "grad_norm": 0.2111409455537796,
      "learning_rate": 9.993252982488423e-05,
      "loss": 1.1461,
      "mean_token_accuracy": 0.6971441358327866,
      "num_tokens": 733600.0,
      "step": 23
    },
    {
      "entropy": 1.0466794520616531,
      "epoch": 0.0022081529145893354,
      "grad_norm": 0.27453601360321045,
      "learning_rate": 9.99294629987426e-05,
      "loss": 1.0335,
      "mean_token_accuracy": 0.7275227941572666,
      "num_tokens": 765408.0,
      "step": 24
    },
    {
      "entropy": 1.120854638516903,
      "epoch": 0.0023001592860305577,
      "grad_norm": 0.20742784440517426,
      "learning_rate": 9.992639617260098e-05,
      "loss": 1.096,
      "mean_token_accuracy": 0.7104541622102261,
      "num_tokens": 797954.0,
      "step": 25
    },
    {
      "entropy": 1.1934521719813347,
      "epoch": 0.00239216565747178,
      "grad_norm": 0.2353135347366333,
      "learning_rate": 9.992332934645935e-05,
      "loss": 1.1969,
      "mean_token_accuracy": 0.6894397512078285,
      "num_tokens": 830357.0,
      "step": 26
    },
    {
      "entropy": 1.0816687010228634,
      "epoch": 0.0024841720289130024,
      "grad_norm": 0.20302598178386688,
      "learning_rate": 9.992026252031773e-05,
      "loss": 1.0882,
      "mean_token_accuracy": 0.7138861753046513,
      "num_tokens": 862657.0,
      "step": 27
    },
    {
      "entropy": 1.1710936054587364,
      "epoch": 0.0025761784003542243,
      "grad_norm": 0.2950412929058075,
      "learning_rate": 9.99171956941761e-05,
      "loss": 1.2213,
      "mean_token_accuracy": 0.6872150637209415,
      "num_tokens": 894761.0,
      "step": 28
    },
    {
      "entropy": 1.0848508141934872,
      "epoch": 0.0026681847717954467,
      "grad_norm": 0.22355246543884277,
      "learning_rate": 9.991412886803447e-05,
      "loss": 1.1243,
      "mean_token_accuracy": 0.7048648484051228,
      "num_tokens": 925967.0,
      "step": 29
    },
    {
      "entropy": 1.2089909054338932,
      "epoch": 0.002760191143236669,
      "grad_norm": 0.2289072871208191,
      "learning_rate": 9.991106204189285e-05,
      "loss": 1.2091,
      "mean_token_accuracy": 0.6832928396761417,
      "num_tokens": 957524.0,
      "step": 30
    },
    {
      "entropy": 1.2757915444672108,
      "epoch": 0.0028521975146778913,
      "grad_norm": 0.2619670629501343,
      "learning_rate": 9.990799521575123e-05,
      "loss": 1.312,
      "mean_token_accuracy": 0.6680754534900188,
      "num_tokens": 988738.0,
      "step": 31
    },
    {
      "entropy": 1.1263886131346226,
      "epoch": 0.0029442038861191137,
      "grad_norm": 0.42301061749458313,
      "learning_rate": 9.99049283896096e-05,
      "loss": 1.1402,
      "mean_token_accuracy": 0.7081449925899506,
      "num_tokens": 1019851.0,
      "step": 32
    },
    {
      "entropy": 1.1219137981534004,
      "epoch": 0.003036210257560336,
      "grad_norm": 0.22893165051937103,
      "learning_rate": 9.990186156346797e-05,
      "loss": 1.1133,
      "mean_token_accuracy": 0.7073247544467449,
      "num_tokens": 1051554.0,
      "step": 33
    },
    {
      "entropy": 1.0775936767458916,
      "epoch": 0.0031282166290015584,
      "grad_norm": 0.1935524195432663,
      "learning_rate": 9.989879473732635e-05,
      "loss": 1.0592,
      "mean_token_accuracy": 0.7178852446377277,
      "num_tokens": 1084107.0,
      "step": 34
    },
    {
      "entropy": 1.088371992111206,
      "epoch": 0.0032202230004427807,
      "grad_norm": 0.21425174176692963,
      "learning_rate": 9.989572791118472e-05,
      "loss": 1.1229,
      "mean_token_accuracy": 0.7036660052835941,
      "num_tokens": 1115946.0,
      "step": 35
    },
    {
      "entropy": 1.1267016679048538,
      "epoch": 0.003312229371884003,
      "grad_norm": 0.21512165665626526,
      "learning_rate": 9.98926610850431e-05,
      "loss": 1.1506,
      "mean_token_accuracy": 0.6959264352917671,
      "num_tokens": 1147753.0,
      "step": 36
    },
    {
      "entropy": 0.9860531277954578,
      "epoch": 0.0034042357433252254,
      "grad_norm": 0.20237453281879425,
      "learning_rate": 9.988959425890146e-05,
      "loss": 1.0075,
      "mean_token_accuracy": 0.7293110303580761,
      "num_tokens": 1179873.0,
      "step": 37
    },
    {
      "entropy": 1.2115809097886086,
      "epoch": 0.0034962421147664477,
      "grad_norm": 0.21071083843708038,
      "learning_rate": 9.988652743275985e-05,
      "loss": 1.2138,
      "mean_token_accuracy": 0.687472153455019,
      "num_tokens": 1211906.0,
      "step": 38
    },
    {
      "entropy": 1.0664190016686916,
      "epoch": 0.00358824848620767,
      "grad_norm": 0.19886374473571777,
      "learning_rate": 9.988346060661821e-05,
      "loss": 1.058,
      "mean_token_accuracy": 0.71553760394454,
      "num_tokens": 1244426.0,
      "step": 39
    },
    {
      "entropy": 1.1510485298931599,
      "epoch": 0.003680254857648892,
      "grad_norm": 0.19857457280158997,
      "learning_rate": 9.988039378047658e-05,
      "loss": 1.1691,
      "mean_token_accuracy": 0.6979049108922482,
      "num_tokens": 1276002.0,
      "step": 40
    },
    {
      "entropy": 1.165607638657093,
      "epoch": 0.0037722612290901143,
      "grad_norm": 0.19593724608421326,
      "learning_rate": 9.987732695433496e-05,
      "loss": 1.1344,
      "mean_token_accuracy": 0.6992177776992321,
      "num_tokens": 1307262.0,
      "step": 41
    },
    {
      "entropy": 1.1791815012693405,
      "epoch": 0.0038642676005313367,
      "grad_norm": 0.20384101569652557,
      "learning_rate": 9.987426012819335e-05,
      "loss": 1.1838,
      "mean_token_accuracy": 0.6937372386455536,
      "num_tokens": 1339221.0,
      "step": 42
    },
    {
      "entropy": 1.172286819666624,
      "epoch": 0.0039562739719725595,
      "grad_norm": 0.19468903541564941,
      "learning_rate": 9.987119330205171e-05,
      "loss": 1.1751,
      "mean_token_accuracy": 0.6908092014491558,
      "num_tokens": 1370441.0,
      "step": 43
    },
    {
      "entropy": 1.081815343350172,
      "epoch": 0.004048280343413781,
      "grad_norm": 0.2200634777545929,
      "learning_rate": 9.986812647591008e-05,
      "loss": 1.0664,
      "mean_token_accuracy": 0.7244240567088127,
      "num_tokens": 1401754.0,
      "step": 44
    },
    {
      "entropy": 1.2078289203345776,
      "epoch": 0.004140286714855003,
      "grad_norm": 0.19776010513305664,
      "learning_rate": 9.986505964976846e-05,
      "loss": 1.2207,
      "mean_token_accuracy": 0.6822244971990585,
      "num_tokens": 1433814.0,
      "step": 45
    },
    {
      "entropy": 1.1718518324196339,
      "epoch": 0.004232293086296226,
      "grad_norm": 0.1949436366558075,
      "learning_rate": 9.986199282362684e-05,
      "loss": 1.1766,
      "mean_token_accuracy": 0.689038347452879,
      "num_tokens": 1466150.0,
      "step": 46
    },
    {
      "entropy": 1.1159547194838524,
      "epoch": 0.004324299457737448,
      "grad_norm": 0.18857932090759277,
      "learning_rate": 9.985892599748521e-05,
      "loss": 1.1349,
      "mean_token_accuracy": 0.694551058113575,
      "num_tokens": 1498097.0,
      "step": 47
    },
    {
      "entropy": 1.0545888021588326,
      "epoch": 0.004416305829178671,
      "grad_norm": 0.17756043374538422,
      "learning_rate": 9.985585917134358e-05,
      "loss": 1.0355,
      "mean_token_accuracy": 0.7207311727106571,
      "num_tokens": 1529768.0,
      "step": 48
    },
    {
      "entropy": 1.2191857025027275,
      "epoch": 0.004508312200619893,
      "grad_norm": 0.1978643387556076,
      "learning_rate": 9.985279234520195e-05,
      "loss": 1.2319,
      "mean_token_accuracy": 0.6804237402975559,
      "num_tokens": 1561641.0,
      "step": 49
    },
    {
      "entropy": 1.067577201873064,
      "epoch": 0.0046003185720611154,
      "grad_norm": 0.18102110922336578,
      "learning_rate": 9.984972551906033e-05,
      "loss": 1.0732,
      "mean_token_accuracy": 0.7154321148991585,
      "num_tokens": 1594025.0,
      "step": 50
    },
    {
      "entropy": 1.082532148808241,
      "epoch": 0.004692324943502337,
      "grad_norm": 0.1834568977355957,
      "learning_rate": 9.984665869291871e-05,
      "loss": 1.0925,
      "mean_token_accuracy": 0.7136896662414074,
      "num_tokens": 1625865.0,
      "step": 51
    },
    {
      "entropy": 1.1882275715470314,
      "epoch": 0.00478433131494356,
      "grad_norm": 0.19145984947681427,
      "learning_rate": 9.984359186677708e-05,
      "loss": 1.1896,
      "mean_token_accuracy": 0.6846486479043961,
      "num_tokens": 1657728.0,
      "step": 52
    },
    {
      "entropy": 0.9871219620108604,
      "epoch": 0.004876337686384782,
      "grad_norm": 0.18753796815872192,
      "learning_rate": 9.984052504063545e-05,
      "loss": 1.0052,
      "mean_token_accuracy": 0.7346924133598804,
      "num_tokens": 1690329.0,
      "step": 53
    },
    {
      "entropy": 1.049861989915371,
      "epoch": 0.004968344057826005,
      "grad_norm": 0.19207102060317993,
      "learning_rate": 9.983745821449383e-05,
      "loss": 1.0752,
      "mean_token_accuracy": 0.7105509489774704,
      "num_tokens": 1722832.0,
      "step": 54
    },
    {
      "entropy": 1.0746104791760445,
      "epoch": 0.005060350429267227,
      "grad_norm": 0.4220045506954193,
      "learning_rate": 9.98343913883522e-05,
      "loss": 1.1222,
      "mean_token_accuracy": 0.7099413871765137,
      "num_tokens": 1755600.0,
      "step": 55
    },
    {
      "entropy": 1.190536953508854,
      "epoch": 0.005152356800708449,
      "grad_norm": 0.19676953554153442,
      "learning_rate": 9.983132456221058e-05,
      "loss": 1.1398,
      "mean_token_accuracy": 0.6977646127343178,
      "num_tokens": 1787337.0,
      "step": 56
    },
    {
      "entropy": 1.326043676584959,
      "epoch": 0.005244363172149671,
      "grad_norm": 0.39878496527671814,
      "learning_rate": 9.982825773606894e-05,
      "loss": 1.2605,
      "mean_token_accuracy": 0.6741001829504967,
      "num_tokens": 1819095.0,
      "step": 57
    },
    {
      "entropy": 1.07636122033,
      "epoch": 0.005336369543590893,
      "grad_norm": 0.1924779862165451,
      "learning_rate": 9.982519090992733e-05,
      "loss": 1.0969,
      "mean_token_accuracy": 0.706939484924078,
      "num_tokens": 1850593.0,
      "step": 58
    },
    {
      "entropy": 1.0179176963865757,
      "epoch": 0.005428375915032116,
      "grad_norm": 0.20269745588302612,
      "learning_rate": 9.98221240837857e-05,
      "loss": 1.0597,
      "mean_token_accuracy": 0.7214597202837467,
      "num_tokens": 1882091.0,
      "step": 59
    },
    {
      "entropy": 0.987612746655941,
      "epoch": 0.005520382286473338,
      "grad_norm": 0.20211850106716156,
      "learning_rate": 9.981905725764406e-05,
      "loss": 1.0485,
      "mean_token_accuracy": 0.7169143334031105,
      "num_tokens": 1913474.0,
      "step": 60
    },
    {
      "entropy": 0.9966298453509808,
      "epoch": 0.005612388657914561,
      "grad_norm": 0.19133584201335907,
      "learning_rate": 9.981599043150244e-05,
      "loss": 1.0422,
      "mean_token_accuracy": 0.7224424257874489,
      "num_tokens": 1945522.0,
      "step": 61
    },
    {
      "entropy": 1.1048134565353394,
      "epoch": 0.005704395029355783,
      "grad_norm": 0.19354288280010223,
      "learning_rate": 9.981292360536082e-05,
      "loss": 1.1547,
      "mean_token_accuracy": 0.6927171051502228,
      "num_tokens": 1978033.0,
      "step": 62
    },
    {
      "entropy": 1.1163930743932724,
      "epoch": 0.0057964014007970055,
      "grad_norm": 0.20973046123981476,
      "learning_rate": 9.980985677921919e-05,
      "loss": 1.1142,
      "mean_token_accuracy": 0.7007377371191978,
      "num_tokens": 2008721.0,
      "step": 63
    },
    {
      "entropy": 1.11024309694767,
      "epoch": 0.005888407772238227,
      "grad_norm": 0.18586821854114532,
      "learning_rate": 9.980678995307756e-05,
      "loss": 1.1115,
      "mean_token_accuracy": 0.7096363380551338,
      "num_tokens": 2039797.0,
      "step": 64
    },
    {
      "entropy": 1.1026447750627995,
      "epoch": 0.00598041414367945,
      "grad_norm": 0.2010764479637146,
      "learning_rate": 9.980372312693594e-05,
      "loss": 1.0939,
      "mean_token_accuracy": 0.709170650690794,
      "num_tokens": 2071326.0,
      "step": 65
    },
    {
      "entropy": 1.0555282533168793,
      "epoch": 0.006072420515120672,
      "grad_norm": 0.1917371153831482,
      "learning_rate": 9.980065630079431e-05,
      "loss": 1.0527,
      "mean_token_accuracy": 0.7188236601650715,
      "num_tokens": 2104026.0,
      "step": 66
    },
    {
      "entropy": 1.1160909458994865,
      "epoch": 0.006164426886561895,
      "grad_norm": 0.21426288783550262,
      "learning_rate": 9.979758947465269e-05,
      "loss": 1.1107,
      "mean_token_accuracy": 0.7048729918897152,
      "num_tokens": 2136794.0,
      "step": 67
    },
    {
      "entropy": 1.006285447627306,
      "epoch": 0.006256433258003117,
      "grad_norm": 0.17311859130859375,
      "learning_rate": 9.979452264851106e-05,
      "loss": 0.9998,
      "mean_token_accuracy": 0.7299099713563919,
      "num_tokens": 2169462.0,
      "step": 68
    },
    {
      "entropy": 1.23557785525918,
      "epoch": 0.006348439629444339,
      "grad_norm": 0.19895704090595245,
      "learning_rate": 9.979145582236944e-05,
      "loss": 1.252,
      "mean_token_accuracy": 0.6738821007311344,
      "num_tokens": 2200877.0,
      "step": 69
    },
    {
      "entropy": 1.1019893176853657,
      "epoch": 0.0064404460008855614,
      "grad_norm": 0.18910272419452667,
      "learning_rate": 9.978838899622781e-05,
      "loss": 1.1202,
      "mean_token_accuracy": 0.6998571455478668,
      "num_tokens": 2231588.0,
      "step": 70
    },
    {
      "entropy": 1.0300692841410637,
      "epoch": 0.006532452372326783,
      "grad_norm": 0.18493330478668213,
      "learning_rate": 9.978532217008618e-05,
      "loss": 1.0702,
      "mean_token_accuracy": 0.7154638580977917,
      "num_tokens": 2263883.0,
      "step": 71
    },
    {
      "entropy": 1.1393884010612965,
      "epoch": 0.006624458743768006,
      "grad_norm": 0.19347243010997772,
      "learning_rate": 9.978225534394456e-05,
      "loss": 1.1797,
      "mean_token_accuracy": 0.6906083673238754,
      "num_tokens": 2296288.0,
      "step": 72
    },
    {
      "entropy": 1.0748408548533916,
      "epoch": 0.006716465115209228,
      "grad_norm": 0.18773241341114044,
      "learning_rate": 9.977918851780294e-05,
      "loss": 1.1091,
      "mean_token_accuracy": 0.7066237106919289,
      "num_tokens": 2327450.0,
      "step": 73
    },
    {
      "entropy": 1.0307633616030216,
      "epoch": 0.006808471486650451,
      "grad_norm": 0.1866629272699356,
      "learning_rate": 9.97761216916613e-05,
      "loss": 1.0411,
      "mean_token_accuracy": 0.7207498736679554,
      "num_tokens": 2359015.0,
      "step": 74
    },
    {
      "entropy": 1.045896802097559,
      "epoch": 0.006900477858091673,
      "grad_norm": 0.18500326573848724,
      "learning_rate": 9.977305486551967e-05,
      "loss": 1.0629,
      "mean_token_accuracy": 0.7127506658434868,
      "num_tokens": 2390830.0,
      "step": 75
    },
    {
      "entropy": 1.0593245476484299,
      "epoch": 0.0069924842295328955,
      "grad_norm": 0.17050045728683472,
      "learning_rate": 9.976998803937804e-05,
      "loss": 1.047,
      "mean_token_accuracy": 0.7208504602313042,
      "num_tokens": 2422785.0,
      "step": 76
    },
    {
      "entropy": 1.1434853821992874,
      "epoch": 0.007084490600974117,
      "grad_norm": 0.18553388118743896,
      "learning_rate": 9.976692121323644e-05,
      "loss": 1.1318,
      "mean_token_accuracy": 0.7010665126144886,
      "num_tokens": 2454468.0,
      "step": 77
    },
    {
      "entropy": 1.1408165097236633,
      "epoch": 0.00717649697241534,
      "grad_norm": 0.18438254296779633,
      "learning_rate": 9.97638543870948e-05,
      "loss": 1.118,
      "mean_token_accuracy": 0.6995579488575459,
      "num_tokens": 2485845.0,
      "step": 78
    },
    {
      "entropy": 1.0247646942734718,
      "epoch": 0.007268503343856562,
      "grad_norm": 0.17853163182735443,
      "learning_rate": 9.976078756095317e-05,
      "loss": 1.04,
      "mean_token_accuracy": 0.7206488102674484,
      "num_tokens": 2518370.0,
      "step": 79
    },
    {
      "entropy": 1.0868795104324818,
      "epoch": 0.007360509715297784,
      "grad_norm": 0.18062998354434967,
      "learning_rate": 9.975772073481154e-05,
      "loss": 1.0893,
      "mean_token_accuracy": 0.7098815627396107,
      "num_tokens": 2549367.0,
      "step": 80
    },
    {
      "entropy": 1.0856024138629436,
      "epoch": 0.007452516086739007,
      "grad_norm": 0.17523346841335297,
      "learning_rate": 9.975465390866992e-05,
      "loss": 1.087,
      "mean_token_accuracy": 0.7058769091963768,
      "num_tokens": 2581719.0,
      "step": 81
    },
    {
      "entropy": 1.0571518577635288,
      "epoch": 0.007544522458180229,
      "grad_norm": 0.18031632900238037,
      "learning_rate": 9.97515870825283e-05,
      "loss": 1.0688,
      "mean_token_accuracy": 0.71573830768466,
      "num_tokens": 2613770.0,
      "step": 82
    },
    {
      "entropy": 1.0958705246448517,
      "epoch": 0.0076365288296214515,
      "grad_norm": 0.18454986810684204,
      "learning_rate": 9.974852025638667e-05,
      "loss": 1.1103,
      "mean_token_accuracy": 0.70598429068923,
      "num_tokens": 2645771.0,
      "step": 83
    },
    {
      "entropy": 1.1435690522193909,
      "epoch": 0.007728535201062673,
      "grad_norm": 0.19003550708293915,
      "learning_rate": 9.974545343024504e-05,
      "loss": 1.1678,
      "mean_token_accuracy": 0.6923130080103874,
      "num_tokens": 2677562.0,
      "step": 84
    },
    {
      "entropy": 1.0437137372791767,
      "epoch": 0.007820541572503895,
      "grad_norm": 0.18445762991905212,
      "learning_rate": 9.974238660410342e-05,
      "loss": 1.0637,
      "mean_token_accuracy": 0.7191505506634712,
      "num_tokens": 2709465.0,
      "step": 85
    },
    {
      "entropy": 1.033041786402464,
      "epoch": 0.007912547943945119,
      "grad_norm": 0.18696832656860352,
      "learning_rate": 9.973931977796179e-05,
      "loss": 1.0749,
      "mean_token_accuracy": 0.7103376239538193,
      "num_tokens": 2741052.0,
      "step": 86
    },
    {
      "entropy": 1.1052805744111538,
      "epoch": 0.00800455431538634,
      "grad_norm": 0.18216514587402344,
      "learning_rate": 9.973625295182017e-05,
      "loss": 1.15,
      "mean_token_accuracy": 0.6966203674674034,
      "num_tokens": 2773631.0,
      "step": 87
    },
    {
      "entropy": 1.1706587970256805,
      "epoch": 0.008096560686827563,
      "grad_norm": 0.1851716786623001,
      "learning_rate": 9.973318612567854e-05,
      "loss": 1.1871,
      "mean_token_accuracy": 0.6885181404650211,
      "num_tokens": 2806328.0,
      "step": 88
    },
    {
      "entropy": 1.0726494491100311,
      "epoch": 0.008188567058268785,
      "grad_norm": 0.18485216796398163,
      "learning_rate": 9.973011929953692e-05,
      "loss": 1.0773,
      "mean_token_accuracy": 0.7048849239945412,
      "num_tokens": 2837765.0,
      "step": 89
    },
    {
      "entropy": 1.2616961225867271,
      "epoch": 0.008280573429710007,
      "grad_norm": 0.2037215232849121,
      "learning_rate": 9.972705247339529e-05,
      "loss": 1.2269,
      "mean_token_accuracy": 0.6795659698545933,
      "num_tokens": 2869965.0,
      "step": 90
    },
    {
      "entropy": 1.121584478765726,
      "epoch": 0.00837257980115123,
      "grad_norm": 0.17375269532203674,
      "learning_rate": 9.972398564725366e-05,
      "loss": 1.1206,
      "mean_token_accuracy": 0.7035504058003426,
      "num_tokens": 2902732.0,
      "step": 91
    },
    {
      "entropy": 1.0537205636501312,
      "epoch": 0.008464586172592452,
      "grad_norm": 0.18438859283924103,
      "learning_rate": 9.972091882111204e-05,
      "loss": 1.0591,
      "mean_token_accuracy": 0.7204208634793758,
      "num_tokens": 2934394.0,
      "step": 92
    },
    {
      "entropy": 1.0765618160367012,
      "epoch": 0.008556592544033674,
      "grad_norm": 0.18548394739627838,
      "learning_rate": 9.971785199497042e-05,
      "loss": 1.0946,
      "mean_token_accuracy": 0.7067177481949329,
      "num_tokens": 2966269.0,
      "step": 93
    },
    {
      "entropy": 1.1326740346848965,
      "epoch": 0.008648598915474896,
      "grad_norm": 0.18695706129074097,
      "learning_rate": 9.971478516882879e-05,
      "loss": 1.1364,
      "mean_token_accuracy": 0.7007706575095654,
      "num_tokens": 2998336.0,
      "step": 94
    },
    {
      "entropy": 1.1041999869048595,
      "epoch": 0.00874060528691612,
      "grad_norm": 0.17914101481437683,
      "learning_rate": 9.971171834268715e-05,
      "loss": 1.1128,
      "mean_token_accuracy": 0.7018284164369106,
      "num_tokens": 3030835.0,
      "step": 95
    },
    {
      "entropy": 1.0158977136015892,
      "epoch": 0.008832611658357341,
      "grad_norm": 0.1744915097951889,
      "learning_rate": 9.970865151654554e-05,
      "loss": 1.0195,
      "mean_token_accuracy": 0.7307408563792706,
      "num_tokens": 3062030.0,
      "step": 96
    },
    {
      "entropy": 1.110980585217476,
      "epoch": 0.008924618029798563,
      "grad_norm": 0.18282753229141235,
      "learning_rate": 9.97055846904039e-05,
      "loss": 1.1197,
      "mean_token_accuracy": 0.702539723366499,
      "num_tokens": 3094141.0,
      "step": 97
    },
    {
      "entropy": 1.0316043384373188,
      "epoch": 0.009016624401239785,
      "grad_norm": 0.18121623992919922,
      "learning_rate": 9.970251786426228e-05,
      "loss": 1.0654,
      "mean_token_accuracy": 0.7126892581582069,
      "num_tokens": 3126612.0,
      "step": 98
    },
    {
      "entropy": 1.0957234241068363,
      "epoch": 0.009108630772681009,
      "grad_norm": 0.18234795331954956,
      "learning_rate": 9.969945103812065e-05,
      "loss": 1.1237,
      "mean_token_accuracy": 0.698351114988327,
      "num_tokens": 3157958.0,
      "step": 99
    },
    {
      "entropy": 1.083721086382866,
      "epoch": 0.009200637144122231,
      "grad_norm": 0.17708368599414825,
      "learning_rate": 9.969638421197903e-05,
      "loss": 1.0797,
      "mean_token_accuracy": 0.7101627215743065,
      "num_tokens": 3190140.0,
      "step": 100
    },
    {
      "entropy": 1.0303197726607323,
      "epoch": 0.009292643515563453,
      "grad_norm": 0.17580947279930115,
      "learning_rate": 9.96933173858374e-05,
      "loss": 1.0418,
      "mean_token_accuracy": 0.722861647605896,
      "num_tokens": 3222104.0,
      "step": 101
    },
    {
      "entropy": 1.0852229669690132,
      "epoch": 0.009384649887004675,
      "grad_norm": 0.1876814216375351,
      "learning_rate": 9.969025055969577e-05,
      "loss": 1.0762,
      "mean_token_accuracy": 0.7111622951924801,
      "num_tokens": 3253985.0,
      "step": 102
    },
    {
      "entropy": 1.0656349509954453,
      "epoch": 0.009476656258445897,
      "grad_norm": 0.17324751615524292,
      "learning_rate": 9.968718373355415e-05,
      "loss": 1.0748,
      "mean_token_accuracy": 0.7125000432133675,
      "num_tokens": 3286359.0,
      "step": 103
    },
    {
      "entropy": 1.133026022464037,
      "epoch": 0.00956866262988712,
      "grad_norm": 0.19071149826049805,
      "learning_rate": 9.968411690741253e-05,
      "loss": 1.1509,
      "mean_token_accuracy": 0.6924744583666325,
      "num_tokens": 3318294.0,
      "step": 104
    },
    {
      "entropy": 1.0567429475486279,
      "epoch": 0.009660669001328342,
      "grad_norm": 0.184308722615242,
      "learning_rate": 9.96810500812709e-05,
      "loss": 1.071,
      "mean_token_accuracy": 0.7111519351601601,
      "num_tokens": 3349857.0,
      "step": 105
    },
    {
      "entropy": 1.1476225852966309,
      "epoch": 0.009752675372769564,
      "grad_norm": 0.17656715214252472,
      "learning_rate": 9.967798325512927e-05,
      "loss": 1.12,
      "mean_token_accuracy": 0.7016064189374447,
      "num_tokens": 3381786.0,
      "step": 106
    },
    {
      "entropy": 0.9738841466605663,
      "epoch": 0.009844681744210786,
      "grad_norm": 0.18338768184185028,
      "learning_rate": 9.967491642898764e-05,
      "loss": 0.9896,
      "mean_token_accuracy": 0.7332197390496731,
      "num_tokens": 3413128.0,
      "step": 107
    },
    {
      "entropy": 1.1190122663974762,
      "epoch": 0.00993668811565201,
      "grad_norm": 0.1801600456237793,
      "learning_rate": 9.967184960284602e-05,
      "loss": 1.1207,
      "mean_token_accuracy": 0.7012488469481468,
      "num_tokens": 3445319.0,
      "step": 108
    },
    {
      "entropy": 1.0729020908474922,
      "epoch": 0.010028694487093232,
      "grad_norm": 0.17818088829517365,
      "learning_rate": 9.96687827767044e-05,
      "loss": 1.0996,
      "mean_token_accuracy": 0.7114726230502129,
      "num_tokens": 3477870.0,
      "step": 109
    },
    {
      "entropy": 0.9894374907016754,
      "epoch": 0.010120700858534453,
      "grad_norm": 0.1673220843076706,
      "learning_rate": 9.966571595056277e-05,
      "loss": 1.0296,
      "mean_token_accuracy": 0.7245632782578468,
      "num_tokens": 3510356.0,
      "step": 110
    },
    {
      "entropy": 1.0988226309418678,
      "epoch": 0.010212707229975675,
      "grad_norm": 0.17366047203540802,
      "learning_rate": 9.966264912442113e-05,
      "loss": 1.1232,
      "mean_token_accuracy": 0.707622442394495,
      "num_tokens": 3542667.0,
      "step": 111
    },
    {
      "entropy": 1.023915782570839,
      "epoch": 0.010304713601416897,
      "grad_norm": 0.17236974835395813,
      "learning_rate": 9.965958229827952e-05,
      "loss": 1.0401,
      "mean_token_accuracy": 0.7258173190057278,
      "num_tokens": 3574433.0,
      "step": 112
    },
    {
      "entropy": 1.0756308734416962,
      "epoch": 0.010396719972858121,
      "grad_norm": 0.1818331927061081,
      "learning_rate": 9.965651547213788e-05,
      "loss": 1.0958,
      "mean_token_accuracy": 0.704191118478775,
      "num_tokens": 3606327.0,
      "step": 113
    },
    {
      "entropy": 1.0850714892148972,
      "epoch": 0.010488726344299343,
      "grad_norm": 0.17513324320316315,
      "learning_rate": 9.965344864599627e-05,
      "loss": 1.0801,
      "mean_token_accuracy": 0.7091819941997528,
      "num_tokens": 3638129.0,
      "step": 114
    },
    {
      "entropy": 1.2840271592140198,
      "epoch": 0.010580732715740565,
      "grad_norm": 0.19304576516151428,
      "learning_rate": 9.965038181985463e-05,
      "loss": 1.3104,
      "mean_token_accuracy": 0.6577373184263706,
      "num_tokens": 3670330.0,
      "step": 115
    },
    {
      "entropy": 1.070245523005724,
      "epoch": 0.010672739087181787,
      "grad_norm": 0.17702911794185638,
      "learning_rate": 9.964731499371301e-05,
      "loss": 1.0614,
      "mean_token_accuracy": 0.7105001769959927,
      "num_tokens": 3701810.0,
      "step": 116
    },
    {
      "entropy": 1.1323248483240604,
      "epoch": 0.01076474545862301,
      "grad_norm": 0.17637357115745544,
      "learning_rate": 9.964424816757138e-05,
      "loss": 1.1118,
      "mean_token_accuracy": 0.7014639340341091,
      "num_tokens": 3733961.0,
      "step": 117
    },
    {
      "entropy": 0.9435235187411308,
      "epoch": 0.010856751830064232,
      "grad_norm": 0.1700422763824463,
      "learning_rate": 9.964118134142975e-05,
      "loss": 0.912,
      "mean_token_accuracy": 0.7540346011519432,
      "num_tokens": 3766324.0,
      "step": 118
    },
    {
      "entropy": 1.0387593246996403,
      "epoch": 0.010948758201505454,
      "grad_norm": 0.17679983377456665,
      "learning_rate": 9.963811451528813e-05,
      "loss": 1.043,
      "mean_token_accuracy": 0.7226700335741043,
      "num_tokens": 3798399.0,
      "step": 119
    },
    {
      "entropy": 1.0451080091297626,
      "epoch": 0.011040764572946676,
      "grad_norm": 0.1833013892173767,
      "learning_rate": 9.963504768914651e-05,
      "loss": 1.0827,
      "mean_token_accuracy": 0.7135558277368546,
      "num_tokens": 3830949.0,
      "step": 120
    },
    {
      "entropy": 1.0882374159991741,
      "epoch": 0.0111327709443879,
      "grad_norm": 0.1769454926252365,
      "learning_rate": 9.963198086300488e-05,
      "loss": 1.1276,
      "mean_token_accuracy": 0.7030789442360401,
      "num_tokens": 3862243.0,
      "step": 121
    },
    {
      "entropy": 0.9281687214970589,
      "epoch": 0.011224777315829122,
      "grad_norm": 0.17013674974441528,
      "learning_rate": 9.962891403686325e-05,
      "loss": 0.9444,
      "mean_token_accuracy": 0.742885272949934,
      "num_tokens": 3894386.0,
      "step": 122
    },
    {
      "entropy": 1.033750969916582,
      "epoch": 0.011316783687270343,
      "grad_norm": 0.1758384257555008,
      "learning_rate": 9.962584721072162e-05,
      "loss": 1.0568,
      "mean_token_accuracy": 0.7223353795707226,
      "num_tokens": 3926782.0,
      "step": 123
    },
    {
      "entropy": 1.136779386550188,
      "epoch": 0.011408790058711565,
      "grad_norm": 0.18879708647727966,
      "learning_rate": 9.962278038458001e-05,
      "loss": 1.1352,
      "mean_token_accuracy": 0.6941037364304066,
      "num_tokens": 3958414.0,
      "step": 124
    },
    {
      "entropy": 1.105912957340479,
      "epoch": 0.011500796430152787,
      "grad_norm": 0.1762920320034027,
      "learning_rate": 9.961971355843838e-05,
      "loss": 1.1068,
      "mean_token_accuracy": 0.703526221215725,
      "num_tokens": 3990819.0,
      "step": 125
    },
    {
      "entropy": 1.0447328872978687,
      "epoch": 0.011592802801594011,
      "grad_norm": 0.19029872119426727,
      "learning_rate": 9.961664673229675e-05,
      "loss": 1.0377,
      "mean_token_accuracy": 0.721235129982233,
      "num_tokens": 4023295.0,
      "step": 126
    },
    {
      "entropy": 1.0258752293884754,
      "epoch": 0.011684809173035233,
      "grad_norm": 0.17486336827278137,
      "learning_rate": 9.961357990615513e-05,
      "loss": 1.043,
      "mean_token_accuracy": 0.7114723920822144,
      "num_tokens": 4055792.0,
      "step": 127
    },
    {
      "entropy": 1.1351664178073406,
      "epoch": 0.011776815544476455,
      "grad_norm": 0.1835290491580963,
      "learning_rate": 9.96105130800135e-05,
      "loss": 1.1606,
      "mean_token_accuracy": 0.6948552615940571,
      "num_tokens": 4087323.0,
      "step": 128
    },
    {
      "entropy": 1.0432838648557663,
      "epoch": 0.011868821915917677,
      "grad_norm": 0.18978850543498993,
      "learning_rate": 9.960744625387188e-05,
      "loss": 1.0704,
      "mean_token_accuracy": 0.7125457674264908,
      "num_tokens": 4119998.0,
      "step": 129
    },
    {
      "entropy": 1.0382169298827648,
      "epoch": 0.0119608282873589,
      "grad_norm": 0.17402221262454987,
      "learning_rate": 9.960437942773025e-05,
      "loss": 1.0428,
      "mean_token_accuracy": 0.7186609581112862,
      "num_tokens": 4151785.0,
      "step": 130
    },
    {
      "entropy": 0.9941658936440945,
      "epoch": 0.012052834658800122,
      "grad_norm": 0.17558282613754272,
      "learning_rate": 9.960131260158863e-05,
      "loss": 1.0007,
      "mean_token_accuracy": 0.7280302010476589,
      "num_tokens": 4183830.0,
      "step": 131
    },
    {
      "entropy": 1.0216947868466377,
      "epoch": 0.012144841030241344,
      "grad_norm": 0.1805516481399536,
      "learning_rate": 9.9598245775447e-05,
      "loss": 1.0407,
      "mean_token_accuracy": 0.7166652455925941,
      "num_tokens": 4215451.0,
      "step": 132
    },
    {
      "entropy": 1.0451236963272095,
      "epoch": 0.012236847401682566,
      "grad_norm": 0.18230095505714417,
      "learning_rate": 9.959517894930536e-05,
      "loss": 1.0667,
      "mean_token_accuracy": 0.7071390524506569,
      "num_tokens": 4247721.0,
      "step": 133
    },
    {
      "entropy": 0.9948875792324543,
      "epoch": 0.01232885377312379,
      "grad_norm": 0.18703404068946838,
      "learning_rate": 9.959211212316374e-05,
      "loss": 1.0102,
      "mean_token_accuracy": 0.729287151247263,
      "num_tokens": 4280102.0,
      "step": 134
    },
    {
      "entropy": 1.0634112022817135,
      "epoch": 0.012420860144565012,
      "grad_norm": 0.18192103505134583,
      "learning_rate": 9.958904529702213e-05,
      "loss": 1.0809,
      "mean_token_accuracy": 0.7082584127783775,
      "num_tokens": 4311981.0,
      "step": 135
    },
    {
      "entropy": 1.1690315753221512,
      "epoch": 0.012512866516006233,
      "grad_norm": 0.1822914481163025,
      "learning_rate": 9.95859784708805e-05,
      "loss": 1.1624,
      "mean_token_accuracy": 0.6938770450651646,
      "num_tokens": 4343337.0,
      "step": 136
    },
    {
      "entropy": 1.0696013569831848,
      "epoch": 0.012604872887447455,
      "grad_norm": 0.17731152474880219,
      "learning_rate": 9.958291164473886e-05,
      "loss": 1.0644,
      "mean_token_accuracy": 0.7144139632582664,
      "num_tokens": 4375251.0,
      "step": 137
    },
    {
      "entropy": 1.056776750832796,
      "epoch": 0.012696879258888677,
      "grad_norm": 0.17421656847000122,
      "learning_rate": 9.957984481859723e-05,
      "loss": 1.0549,
      "mean_token_accuracy": 0.7194741331040859,
      "num_tokens": 4407573.0,
      "step": 138
    },
    {
      "entropy": 0.9020832404494286,
      "epoch": 0.012788885630329901,
      "grad_norm": 0.1633674055337906,
      "learning_rate": 9.957677799245561e-05,
      "loss": 0.9044,
      "mean_token_accuracy": 0.7556024454534054,
      "num_tokens": 4439825.0,
      "step": 139
    },
    {
      "entropy": 1.0089088529348373,
      "epoch": 0.012880892001771123,
      "grad_norm": 0.18037532269954681,
      "learning_rate": 9.957371116631399e-05,
      "loss": 1.0301,
      "mean_token_accuracy": 0.7221655361354351,
      "num_tokens": 4470894.0,
      "step": 140
    },
    {
      "entropy": 1.033136684447527,
      "epoch": 0.012972898373212345,
      "grad_norm": 0.1732766330242157,
      "learning_rate": 9.957064434017236e-05,
      "loss": 1.068,
      "mean_token_accuracy": 0.7135338932275772,
      "num_tokens": 4503398.0,
      "step": 141
    },
    {
      "entropy": 0.9998885989189148,
      "epoch": 0.013064904744653567,
      "grad_norm": 0.17721126973628998,
      "learning_rate": 9.956757751403073e-05,
      "loss": 1.0133,
      "mean_token_accuracy": 0.7290425077080727,
      "num_tokens": 4534741.0,
      "step": 142
    },
    {
      "entropy": 0.9691000506281853,
      "epoch": 0.01315691111609479,
      "grad_norm": 0.18603132665157318,
      "learning_rate": 9.956451068788911e-05,
      "loss": 0.9715,
      "mean_token_accuracy": 0.7310564815998077,
      "num_tokens": 4567267.0,
      "step": 143
    },
    {
      "entropy": 1.1178338378667831,
      "epoch": 0.013248917487536012,
      "grad_norm": 0.18301568925380707,
      "learning_rate": 9.956144386174748e-05,
      "loss": 1.1393,
      "mean_token_accuracy": 0.6937997974455357,
      "num_tokens": 4599399.0,
      "step": 144
    },
    {
      "entropy": 1.012466624379158,
      "epoch": 0.013340923858977234,
      "grad_norm": 0.17691315710544586,
      "learning_rate": 9.955837703560586e-05,
      "loss": 1.0239,
      "mean_token_accuracy": 0.7204160578548908,
      "num_tokens": 4630289.0,
      "step": 145
    },
    {
      "entropy": 1.0570966117084026,
      "epoch": 0.013432930230418456,
      "grad_norm": 0.17416663467884064,
      "learning_rate": 9.955531020946423e-05,
      "loss": 1.0422,
      "mean_token_accuracy": 0.7184420041739941,
      "num_tokens": 4661510.0,
      "step": 146
    },
    {
      "entropy": 1.0585036389529705,
      "epoch": 0.013524936601859678,
      "grad_norm": 0.17640922963619232,
      "learning_rate": 9.955224338332261e-05,
      "loss": 1.0724,
      "mean_token_accuracy": 0.7158324383199215,
      "num_tokens": 4693678.0,
      "step": 147
    },
    {
      "entropy": 1.0154268220067024,
      "epoch": 0.013616942973300902,
      "grad_norm": 0.17352929711341858,
      "learning_rate": 9.954917655718098e-05,
      "loss": 1.0253,
      "mean_token_accuracy": 0.7224516868591309,
      "num_tokens": 4725627.0,
      "step": 148
    },
    {
      "entropy": 1.034794270992279,
      "epoch": 0.013708949344742124,
      "grad_norm": 0.17139284312725067,
      "learning_rate": 9.954610973103934e-05,
      "loss": 1.04,
      "mean_token_accuracy": 0.7241168990731239,
      "num_tokens": 4757512.0,
      "step": 149
    },
    {
      "entropy": 1.1209576539695263,
      "epoch": 0.013800955716183345,
      "grad_norm": 0.18446780741214752,
      "learning_rate": 9.954304290489773e-05,
      "loss": 1.1365,
      "mean_token_accuracy": 0.6923755146563053,
      "num_tokens": 4788745.0,
      "step": 150
    },
    {
      "entropy": 1.0725781880319118,
      "epoch": 0.013892962087624567,
      "grad_norm": 0.189184308052063,
      "learning_rate": 9.953997607875611e-05,
      "loss": 1.0938,
      "mean_token_accuracy": 0.7101431675255299,
      "num_tokens": 4819699.0,
      "step": 151
    },
    {
      "entropy": 1.0915080159902573,
      "epoch": 0.013984968459065791,
      "grad_norm": 0.18500500917434692,
      "learning_rate": 9.953690925261447e-05,
      "loss": 1.1154,
      "mean_token_accuracy": 0.7026957459747791,
      "num_tokens": 4850393.0,
      "step": 152
    },
    {
      "entropy": 0.937639731913805,
      "epoch": 0.014076974830507013,
      "grad_norm": 0.1730193942785263,
      "learning_rate": 9.953384242647284e-05,
      "loss": 0.9263,
      "mean_token_accuracy": 0.7394020035862923,
      "num_tokens": 4882557.0,
      "step": 153
    },
    {
      "entropy": 1.1863786354660988,
      "epoch": 0.014168981201948235,
      "grad_norm": 0.1915581077337265,
      "learning_rate": 9.953077560033122e-05,
      "loss": 1.1604,
      "mean_token_accuracy": 0.6907501555979252,
      "num_tokens": 4914020.0,
      "step": 154
    },
    {
      "entropy": 1.0118658803403378,
      "epoch": 0.014260987573389457,
      "grad_norm": 0.16420020163059235,
      "learning_rate": 9.952770877418959e-05,
      "loss": 1.0315,
      "mean_token_accuracy": 0.7209053263068199,
      "num_tokens": 4946504.0,
      "step": 155
    },
    {
      "entropy": 1.1827064417302608,
      "epoch": 0.01435299394483068,
      "grad_norm": 0.19992844760417938,
      "learning_rate": 9.952464194804797e-05,
      "loss": 1.2142,
      "mean_token_accuracy": 0.6828213706612587,
      "num_tokens": 4978513.0,
      "step": 156
    },
    {
      "entropy": 1.1684614419937134,
      "epoch": 0.014445000316271902,
      "grad_norm": 0.1951766461133957,
      "learning_rate": 9.952157512190634e-05,
      "loss": 1.2079,
      "mean_token_accuracy": 0.6831134855747223,
      "num_tokens": 5010828.0,
      "step": 157
    },
    {
      "entropy": 1.0371404737234116,
      "epoch": 0.014537006687713124,
      "grad_norm": 0.1801554560661316,
      "learning_rate": 9.951850829576472e-05,
      "loss": 1.0702,
      "mean_token_accuracy": 0.7141414321959019,
      "num_tokens": 5042103.0,
      "step": 158
    },
    {
      "entropy": 1.1546086817979813,
      "epoch": 0.014629013059154346,
      "grad_norm": 0.1825013905763626,
      "learning_rate": 9.951544146962309e-05,
      "loss": 1.167,
      "mean_token_accuracy": 0.6934010982513428,
      "num_tokens": 5073180.0,
      "step": 159
    },
    {
      "entropy": 1.0603557638823986,
      "epoch": 0.014721019430595568,
      "grad_norm": 0.17046773433685303,
      "learning_rate": 9.951237464348146e-05,
      "loss": 1.0773,
      "mean_token_accuracy": 0.7160672694444656,
      "num_tokens": 5105608.0,
      "step": 160
    },
    {
      "entropy": 0.9954036884009838,
      "epoch": 0.014813025802036792,
      "grad_norm": 0.17080140113830566,
      "learning_rate": 9.950930781733984e-05,
      "loss": 1.0012,
      "mean_token_accuracy": 0.7265287451446056,
      "num_tokens": 5137311.0,
      "step": 161
    },
    {
      "entropy": 0.9965865425765514,
      "epoch": 0.014905032173478014,
      "grad_norm": 0.182280033826828,
      "learning_rate": 9.950624099119822e-05,
      "loss": 1.0291,
      "mean_token_accuracy": 0.7281336486339569,
      "num_tokens": 5169735.0,
      "step": 162
    },
    {
      "entropy": 1.0358598940074444,
      "epoch": 0.014997038544919235,
      "grad_norm": 0.17245079576969147,
      "learning_rate": 9.950317416505659e-05,
      "loss": 1.0606,
      "mean_token_accuracy": 0.719555176794529,
      "num_tokens": 5201620.0,
      "step": 163
    },
    {
      "entropy": 1.1126755364239216,
      "epoch": 0.015089044916360457,
      "grad_norm": 0.17259301245212555,
      "learning_rate": 9.950010733891496e-05,
      "loss": 1.1028,
      "mean_token_accuracy": 0.7077676318585873,
      "num_tokens": 5233795.0,
      "step": 164
    },
    {
      "entropy": 1.1025396808981895,
      "epoch": 0.015181051287801681,
      "grad_norm": 0.16939501464366913,
      "learning_rate": 9.949704051277334e-05,
      "loss": 1.0794,
      "mean_token_accuracy": 0.7139687314629555,
      "num_tokens": 5264977.0,
      "step": 165
    },
    {
      "entropy": 1.0551231913268566,
      "epoch": 0.015273057659242903,
      "grad_norm": 0.17037580907344818,
      "learning_rate": 9.949397368663172e-05,
      "loss": 1.0413,
      "mean_token_accuracy": 0.7213192842900753,
      "num_tokens": 5296111.0,
      "step": 166
    },
    {
      "entropy": 1.123452853411436,
      "epoch": 0.015365064030684125,
      "grad_norm": 0.1779470592737198,
      "learning_rate": 9.949090686049009e-05,
      "loss": 1.1234,
      "mean_token_accuracy": 0.7065025307238102,
      "num_tokens": 5327243.0,
      "step": 167
    },
    {
      "entropy": 1.04978758841753,
      "epoch": 0.015457070402125347,
      "grad_norm": 0.1631164401769638,
      "learning_rate": 9.948784003434846e-05,
      "loss": 1.037,
      "mean_token_accuracy": 0.7182903625071049,
      "num_tokens": 5359642.0,
      "step": 168
    },
    {
      "entropy": 1.0534667447209358,
      "epoch": 0.015549076773566569,
      "grad_norm": 0.1729961782693863,
      "learning_rate": 9.948477320820682e-05,
      "loss": 1.0537,
      "mean_token_accuracy": 0.7168702408671379,
      "num_tokens": 5390436.0,
      "step": 169
    },
    {
      "entropy": 1.0613337643444538,
      "epoch": 0.01564108314500779,
      "grad_norm": 0.17368227243423462,
      "learning_rate": 9.94817063820652e-05,
      "loss": 1.1085,
      "mean_token_accuracy": 0.709616132080555,
      "num_tokens": 5423093.0,
      "step": 170
    },
    {
      "entropy": 0.9376764446496964,
      "epoch": 0.015733089516449014,
      "grad_norm": 0.15980516374111176,
      "learning_rate": 9.947863955592359e-05,
      "loss": 0.9481,
      "mean_token_accuracy": 0.7427655793726444,
      "num_tokens": 5455365.0,
      "step": 171
    },
    {
      "entropy": 1.0727378018200397,
      "epoch": 0.015825095887890238,
      "grad_norm": 0.17579393088817596,
      "learning_rate": 9.947557272978195e-05,
      "loss": 1.0873,
      "mean_token_accuracy": 0.7099951170384884,
      "num_tokens": 5487414.0,
      "step": 172
    },
    {
      "entropy": 1.042777244001627,
      "epoch": 0.015917102259331458,
      "grad_norm": 0.17277003824710846,
      "learning_rate": 9.947250590364032e-05,
      "loss": 1.0302,
      "mean_token_accuracy": 0.71774647757411,
      "num_tokens": 5519135.0,
      "step": 173
    },
    {
      "entropy": 1.0699250139296055,
      "epoch": 0.01600910863077268,
      "grad_norm": 0.18426410853862762,
      "learning_rate": 9.94694390774987e-05,
      "loss": 1.0659,
      "mean_token_accuracy": 0.7109970711171627,
      "num_tokens": 5550857.0,
      "step": 174
    },
    {
      "entropy": 1.0907160378992558,
      "epoch": 0.016101115002213902,
      "grad_norm": 0.17302244901657104,
      "learning_rate": 9.946637225135707e-05,
      "loss": 1.0969,
      "mean_token_accuracy": 0.7070281803607941,
      "num_tokens": 5582652.0,
      "step": 175
    },
    {
      "entropy": 1.0843562595546246,
      "epoch": 0.016193121373655125,
      "grad_norm": 0.1732192039489746,
      "learning_rate": 9.946330542521545e-05,
      "loss": 1.0937,
      "mean_token_accuracy": 0.7056828290224075,
      "num_tokens": 5614761.0,
      "step": 176
    },
    {
      "entropy": 1.0434652902185917,
      "epoch": 0.01628512774509635,
      "grad_norm": 0.17828090488910675,
      "learning_rate": 9.946023859907382e-05,
      "loss": 1.0766,
      "mean_token_accuracy": 0.7133230268955231,
      "num_tokens": 5647528.0,
      "step": 177
    },
    {
      "entropy": 1.0202577151358128,
      "epoch": 0.01637713411653757,
      "grad_norm": 0.18082240223884583,
      "learning_rate": 9.94571717729322e-05,
      "loss": 1.044,
      "mean_token_accuracy": 0.7185595631599426,
      "num_tokens": 5679999.0,
      "step": 178
    },
    {
      "entropy": 0.9970671012997627,
      "epoch": 0.016469140487978793,
      "grad_norm": 0.1666569709777832,
      "learning_rate": 9.945410494679057e-05,
      "loss": 1.0018,
      "mean_token_accuracy": 0.7270110733807087,
      "num_tokens": 5712313.0,
      "step": 179
    },
    {
      "entropy": 1.006702933460474,
      "epoch": 0.016561146859420013,
      "grad_norm": 0.16732476651668549,
      "learning_rate": 9.945103812064894e-05,
      "loss": 1.0299,
      "mean_token_accuracy": 0.717566329985857,
      "num_tokens": 5744643.0,
      "step": 180
    },
    {
      "entropy": 1.0571333318948746,
      "epoch": 0.016653153230861237,
      "grad_norm": 0.1745980978012085,
      "learning_rate": 9.944797129450732e-05,
      "loss": 1.061,
      "mean_token_accuracy": 0.7116348259150982,
      "num_tokens": 5776647.0,
      "step": 181
    },
    {
      "entropy": 0.9314371161162853,
      "epoch": 0.01674515960230246,
      "grad_norm": 0.16854697465896606,
      "learning_rate": 9.94449044683657e-05,
      "loss": 0.9454,
      "mean_token_accuracy": 0.7435161843895912,
      "num_tokens": 5808550.0,
      "step": 182
    },
    {
      "entropy": 1.026913408190012,
      "epoch": 0.01683716597374368,
      "grad_norm": 0.1700485348701477,
      "learning_rate": 9.944183764222407e-05,
      "loss": 1.0513,
      "mean_token_accuracy": 0.7180965468287468,
      "num_tokens": 5840629.0,
      "step": 183
    },
    {
      "entropy": 1.0966105721890926,
      "epoch": 0.016929172345184904,
      "grad_norm": 0.17706620693206787,
      "learning_rate": 9.943877081608244e-05,
      "loss": 1.126,
      "mean_token_accuracy": 0.6992902867496014,
      "num_tokens": 5872691.0,
      "step": 184
    },
    {
      "entropy": 1.0974591076374054,
      "epoch": 0.017021178716626128,
      "grad_norm": 0.18316107988357544,
      "learning_rate": 9.943570398994082e-05,
      "loss": 1.115,
      "mean_token_accuracy": 0.6996882744133472,
      "num_tokens": 5904599.0,
      "step": 185
    },
    {
      "entropy": 1.0567415244877338,
      "epoch": 0.017113185088067348,
      "grad_norm": 0.17517054080963135,
      "learning_rate": 9.943263716379919e-05,
      "loss": 1.074,
      "mean_token_accuracy": 0.7146077081561089,
      "num_tokens": 5936624.0,
      "step": 186
    },
    {
      "entropy": 1.0514247193932533,
      "epoch": 0.01720519145950857,
      "grad_norm": 0.19081106781959534,
      "learning_rate": 9.942957033765757e-05,
      "loss": 1.0542,
      "mean_token_accuracy": 0.7200583145022392,
      "num_tokens": 5968252.0,
      "step": 187
    },
    {
      "entropy": 1.1320592164993286,
      "epoch": 0.017297197830949792,
      "grad_norm": 0.17685173451900482,
      "learning_rate": 9.942650351151594e-05,
      "loss": 1.1212,
      "mean_token_accuracy": 0.698618657886982,
      "num_tokens": 6000105.0,
      "step": 188
    },
    {
      "entropy": 1.06346470490098,
      "epoch": 0.017389204202391016,
      "grad_norm": 0.17193131148815155,
      "learning_rate": 9.942343668537432e-05,
      "loss": 1.0708,
      "mean_token_accuracy": 0.7091571725904942,
      "num_tokens": 6032033.0,
      "step": 189
    },
    {
      "entropy": 1.041716579347849,
      "epoch": 0.01748121057383224,
      "grad_norm": 0.17374545335769653,
      "learning_rate": 9.942036985923268e-05,
      "loss": 1.0562,
      "mean_token_accuracy": 0.7153358533978462,
      "num_tokens": 6064049.0,
      "step": 190
    },
    {
      "entropy": 1.0139716528356075,
      "epoch": 0.01757321694527346,
      "grad_norm": 0.17219294607639313,
      "learning_rate": 9.941730303309105e-05,
      "loss": 1.0272,
      "mean_token_accuracy": 0.722566194832325,
      "num_tokens": 6095045.0,
      "step": 191
    },
    {
      "entropy": 0.9830888733267784,
      "epoch": 0.017665223316714683,
      "grad_norm": 0.17141008377075195,
      "learning_rate": 9.941423620694943e-05,
      "loss": 0.9854,
      "mean_token_accuracy": 0.7306378521025181,
      "num_tokens": 6127211.0,
      "step": 192
    },
    {
      "entropy": 1.0846739821135998,
      "epoch": 0.017757229688155903,
      "grad_norm": 0.1723085641860962,
      "learning_rate": 9.941116938080782e-05,
      "loss": 1.0821,
      "mean_token_accuracy": 0.7097069770097733,
      "num_tokens": 6159231.0,
      "step": 193
    },
    {
      "entropy": 0.937518822029233,
      "epoch": 0.017849236059597127,
      "grad_norm": 0.18954673409461975,
      "learning_rate": 9.940810255466618e-05,
      "loss": 0.936,
      "mean_token_accuracy": 0.7467454820871353,
      "num_tokens": 6191745.0,
      "step": 194
    },
    {
      "entropy": 1.0816699527204037,
      "epoch": 0.01794124243103835,
      "grad_norm": 0.18237680196762085,
      "learning_rate": 9.940503572852455e-05,
      "loss": 1.0907,
      "mean_token_accuracy": 0.7099614441394806,
      "num_tokens": 6223665.0,
      "step": 195
    },
    {
      "entropy": 1.0073212906718254,
      "epoch": 0.01803324880247957,
      "grad_norm": 0.17136633396148682,
      "learning_rate": 9.940196890238292e-05,
      "loss": 1.0353,
      "mean_token_accuracy": 0.7224341034889221,
      "num_tokens": 6255146.0,
      "step": 196
    },
    {
      "entropy": 1.0104856751859188,
      "epoch": 0.018125255173920794,
      "grad_norm": 0.1720370203256607,
      "learning_rate": 9.939890207624131e-05,
      "loss": 1.0442,
      "mean_token_accuracy": 0.7195315323770046,
      "num_tokens": 6287341.0,
      "step": 197
    },
    {
      "entropy": 0.9401248022913933,
      "epoch": 0.018217261545362018,
      "grad_norm": 0.1726427525281906,
      "learning_rate": 9.939583525009968e-05,
      "loss": 0.9578,
      "mean_token_accuracy": 0.736522663384676,
      "num_tokens": 6318967.0,
      "step": 198
    },
    {
      "entropy": 0.9967475347220898,
      "epoch": 0.018309267916803238,
      "grad_norm": 0.17046281695365906,
      "learning_rate": 9.939276842395805e-05,
      "loss": 1.0109,
      "mean_token_accuracy": 0.7263761423528194,
      "num_tokens": 6351358.0,
      "step": 199
    },
    {
      "entropy": 1.0387144275009632,
      "epoch": 0.018401274288244462,
      "grad_norm": 0.1755915731191635,
      "learning_rate": 9.938970159781642e-05,
      "loss": 1.0394,
      "mean_token_accuracy": 0.7176682725548744,
      "num_tokens": 6382909.0,
      "step": 200
    },
    {
      "entropy": 1.0051369480788708,
      "epoch": 0.018493280659685682,
      "grad_norm": 0.1623706817626953,
      "learning_rate": 9.93866347716748e-05,
      "loss": 0.9926,
      "mean_token_accuracy": 0.7339497320353985,
      "num_tokens": 6415624.0,
      "step": 201
    },
    {
      "entropy": 0.952204056084156,
      "epoch": 0.018585287031126906,
      "grad_norm": 0.17320609092712402,
      "learning_rate": 9.938356794553318e-05,
      "loss": 0.9504,
      "mean_token_accuracy": 0.7400987520813942,
      "num_tokens": 6446550.0,
      "step": 202
    },
    {
      "entropy": 1.0588614493608475,
      "epoch": 0.01867729340256813,
      "grad_norm": 0.17322760820388794,
      "learning_rate": 9.938050111939155e-05,
      "loss": 1.0573,
      "mean_token_accuracy": 0.7115290686488152,
      "num_tokens": 6479318.0,
      "step": 203
    },
    {
      "entropy": 1.0380604639649391,
      "epoch": 0.01876929977400935,
      "grad_norm": 0.17777806520462036,
      "learning_rate": 9.937743429324992e-05,
      "loss": 1.0591,
      "mean_token_accuracy": 0.7090704552829266,
      "num_tokens": 6510442.0,
      "step": 204
    },
    {
      "entropy": 0.99451944231987,
      "epoch": 0.018861306145450573,
      "grad_norm": 0.16736510396003723,
      "learning_rate": 9.93743674671083e-05,
      "loss": 0.9963,
      "mean_token_accuracy": 0.7252599634230137,
      "num_tokens": 6542415.0,
      "step": 205
    },
    {
      "entropy": 0.9844184592366219,
      "epoch": 0.018953312516891793,
      "grad_norm": 0.18517622351646423,
      "learning_rate": 9.937130064096667e-05,
      "loss": 1.0066,
      "mean_token_accuracy": 0.7265032194554806,
      "num_tokens": 6574502.0,
      "step": 206
    },
    {
      "entropy": 1.035576917231083,
      "epoch": 0.019045318888333017,
      "grad_norm": 0.1741742640733719,
      "learning_rate": 9.936823381482505e-05,
      "loss": 1.0491,
      "mean_token_accuracy": 0.7159487903118134,
      "num_tokens": 6606430.0,
      "step": 207
    },
    {
      "entropy": 0.9525736570358276,
      "epoch": 0.01913732525977424,
      "grad_norm": 0.16638284921646118,
      "learning_rate": 9.936516698868341e-05,
      "loss": 0.9748,
      "mean_token_accuracy": 0.7349927350878716,
      "num_tokens": 6638183.0,
      "step": 208
    },
    {
      "entropy": 1.1414592079818249,
      "epoch": 0.01922933163121546,
      "grad_norm": 0.17697302997112274,
      "learning_rate": 9.93621001625418e-05,
      "loss": 1.1442,
      "mean_token_accuracy": 0.6937038414180279,
      "num_tokens": 6669724.0,
      "step": 209
    },
    {
      "entropy": 1.0954368561506271,
      "epoch": 0.019321338002656684,
      "grad_norm": 0.1847495436668396,
      "learning_rate": 9.935903333640016e-05,
      "loss": 1.0898,
      "mean_token_accuracy": 0.703035693615675,
      "num_tokens": 6700961.0,
      "step": 210
    },
    {
      "entropy": 1.0242870301008224,
      "epoch": 0.019413344374097908,
      "grad_norm": 0.16485972702503204,
      "learning_rate": 9.935596651025853e-05,
      "loss": 1.0118,
      "mean_token_accuracy": 0.7284379340708256,
      "num_tokens": 6733540.0,
      "step": 211
    },
    {
      "entropy": 1.0484031476080418,
      "epoch": 0.019505350745539128,
      "grad_norm": 0.17411889135837555,
      "learning_rate": 9.935289968411691e-05,
      "loss": 1.0673,
      "mean_token_accuracy": 0.7132063210010529,
      "num_tokens": 6765886.0,
      "step": 212
    },
    {
      "entropy": 0.9708690419793129,
      "epoch": 0.019597357116980352,
      "grad_norm": 0.16723890602588654,
      "learning_rate": 9.93498328579753e-05,
      "loss": 0.9725,
      "mean_token_accuracy": 0.7321078032255173,
      "num_tokens": 6797605.0,
      "step": 213
    },
    {
      "entropy": 1.0366210862994194,
      "epoch": 0.019689363488421572,
      "grad_norm": 0.17233191430568695,
      "learning_rate": 9.934676603183366e-05,
      "loss": 1.0492,
      "mean_token_accuracy": 0.7130952589213848,
      "num_tokens": 6830167.0,
      "step": 214
    },
    {
      "entropy": 0.9746213294565678,
      "epoch": 0.019781369859862796,
      "grad_norm": 0.16717267036437988,
      "learning_rate": 9.934369920569203e-05,
      "loss": 0.9903,
      "mean_token_accuracy": 0.7313282191753387,
      "num_tokens": 6862028.0,
      "step": 215
    },
    {
      "entropy": 0.9781723245978355,
      "epoch": 0.01987337623130402,
      "grad_norm": 0.17425300180912018,
      "learning_rate": 9.934063237955041e-05,
      "loss": 1.0007,
      "mean_token_accuracy": 0.7261261530220509,
      "num_tokens": 6893433.0,
      "step": 216
    },
    {
      "entropy": 1.0848586782813072,
      "epoch": 0.01996538260274524,
      "grad_norm": 0.17312081158161163,
      "learning_rate": 9.933756555340878e-05,
      "loss": 1.0906,
      "mean_token_accuracy": 0.7051990665495396,
      "num_tokens": 6925307.0,
      "step": 217
    },
    {
      "entropy": 1.0657422728836536,
      "epoch": 0.020057388974186463,
      "grad_norm": 0.17260460555553436,
      "learning_rate": 9.933449872726716e-05,
      "loss": 1.0512,
      "mean_token_accuracy": 0.7107369191944599,
      "num_tokens": 6957480.0,
      "step": 218
    },
    {
      "entropy": 0.9927574396133423,
      "epoch": 0.020149395345627683,
      "grad_norm": 0.18328817188739777,
      "learning_rate": 9.933143190112553e-05,
      "loss": 1.0321,
      "mean_token_accuracy": 0.7282743565738201,
      "num_tokens": 6989781.0,
      "step": 219
    },
    {
      "entropy": 1.0282342210412025,
      "epoch": 0.020241401717068907,
      "grad_norm": 0.17313358187675476,
      "learning_rate": 9.932836507498391e-05,
      "loss": 1.0487,
      "mean_token_accuracy": 0.7234283275902271,
      "num_tokens": 7021039.0,
      "step": 220
    },
    {
      "entropy": 0.9458490014076233,
      "epoch": 0.02033340808851013,
      "grad_norm": 0.16703543066978455,
      "learning_rate": 9.932529824884228e-05,
      "loss": 0.9491,
      "mean_token_accuracy": 0.734752707183361,
      "num_tokens": 7053406.0,
      "step": 221
    },
    {
      "entropy": 0.8986207470297813,
      "epoch": 0.02042541445995135,
      "grad_norm": 0.15605689585208893,
      "learning_rate": 9.932223142270065e-05,
      "loss": 0.8957,
      "mean_token_accuracy": 0.751799687743187,
      "num_tokens": 7086042.0,
      "step": 222
    },
    {
      "entropy": 1.0543724969029427,
      "epoch": 0.020517420831392574,
      "grad_norm": 0.17028896510601044,
      "learning_rate": 9.931916459655903e-05,
      "loss": 1.0599,
      "mean_token_accuracy": 0.7119736634194851,
      "num_tokens": 7118196.0,
      "step": 223
    },
    {
      "entropy": 1.041070356965065,
      "epoch": 0.020609427202833795,
      "grad_norm": 0.16802090406417847,
      "learning_rate": 9.931609777041741e-05,
      "loss": 1.0373,
      "mean_token_accuracy": 0.722816102206707,
      "num_tokens": 7149702.0,
      "step": 224
    },
    {
      "entropy": 0.9995945934206247,
      "epoch": 0.020701433574275018,
      "grad_norm": 0.17345477640628815,
      "learning_rate": 9.931303094427578e-05,
      "loss": 1.0295,
      "mean_token_accuracy": 0.7267900444567204,
      "num_tokens": 7180887.0,
      "step": 225
    },
    {
      "entropy": 1.0745280124247074,
      "epoch": 0.020793439945716242,
      "grad_norm": 0.1859765499830246,
      "learning_rate": 9.930996411813414e-05,
      "loss": 1.098,
      "mean_token_accuracy": 0.704229149967432,
      "num_tokens": 7211903.0,
      "step": 226
    },
    {
      "entropy": 1.1080476678907871,
      "epoch": 0.020885446317157462,
      "grad_norm": 0.17670086026191711,
      "learning_rate": 9.930689729199251e-05,
      "loss": 1.1084,
      "mean_token_accuracy": 0.6990596875548363,
      "num_tokens": 7244198.0,
      "step": 227
    },
    {
      "entropy": 0.8533573169261217,
      "epoch": 0.020977452688598686,
      "grad_norm": 0.1631222814321518,
      "learning_rate": 9.93038304658509e-05,
      "loss": 0.8657,
      "mean_token_accuracy": 0.7622958980500698,
      "num_tokens": 7276229.0,
      "step": 228
    },
    {
      "entropy": 1.0012922808527946,
      "epoch": 0.02106945906003991,
      "grad_norm": 0.16546288132667542,
      "learning_rate": 9.930076363970928e-05,
      "loss": 1.0117,
      "mean_token_accuracy": 0.7259608060121536,
      "num_tokens": 7308048.0,
      "step": 229
    },
    {
      "entropy": 1.1597312614321709,
      "epoch": 0.02116146543148113,
      "grad_norm": 0.19198071956634521,
      "learning_rate": 9.929769681356764e-05,
      "loss": 1.2095,
      "mean_token_accuracy": 0.6846582293510437,
      "num_tokens": 7339030.0,
      "step": 230
    },
    {
      "entropy": 0.9319143630564213,
      "epoch": 0.021253471802922353,
      "grad_norm": 0.16337594389915466,
      "learning_rate": 9.929462998742601e-05,
      "loss": 0.9414,
      "mean_token_accuracy": 0.742465440183878,
      "num_tokens": 7371606.0,
      "step": 231
    },
    {
      "entropy": 1.1574942097067833,
      "epoch": 0.021345478174363573,
      "grad_norm": 0.1782786101102829,
      "learning_rate": 9.929156316128439e-05,
      "loss": 1.1667,
      "mean_token_accuracy": 0.6903795711696148,
      "num_tokens": 7404004.0,
      "step": 232
    },
    {
      "entropy": 0.9820975102484226,
      "epoch": 0.021437484545804797,
      "grad_norm": 0.16371804475784302,
      "learning_rate": 9.928849633514276e-05,
      "loss": 0.9925,
      "mean_token_accuracy": 0.7278564684092999,
      "num_tokens": 7436024.0,
      "step": 233
    },
    {
      "entropy": 1.1143397502601147,
      "epoch": 0.02152949091724602,
      "grad_norm": 0.1745738834142685,
      "learning_rate": 9.928542950900114e-05,
      "loss": 1.0993,
      "mean_token_accuracy": 0.7055794410407543,
      "num_tokens": 7467730.0,
      "step": 234
    },
    {
      "entropy": 1.0289190784096718,
      "epoch": 0.02162149728868724,
      "grad_norm": 0.1755571812391281,
      "learning_rate": 9.928236268285951e-05,
      "loss": 1.0331,
      "mean_token_accuracy": 0.7184229083359241,
      "num_tokens": 7499693.0,
      "step": 235
    },
    {
      "entropy": 1.0330001264810562,
      "epoch": 0.021713503660128464,
      "grad_norm": 0.16859924793243408,
      "learning_rate": 9.927929585671789e-05,
      "loss": 1.0286,
      "mean_token_accuracy": 0.7246302105486393,
      "num_tokens": 7531898.0,
      "step": 236
    },
    {
      "entropy": 1.0337454117834568,
      "epoch": 0.021805510031569685,
      "grad_norm": 0.16738027334213257,
      "learning_rate": 9.927622903057626e-05,
      "loss": 1.0294,
      "mean_token_accuracy": 0.7227383069694042,
      "num_tokens": 7564291.0,
      "step": 237
    },
    {
      "entropy": 1.147235494107008,
      "epoch": 0.021897516403010908,
      "grad_norm": 0.1868073046207428,
      "learning_rate": 9.927316220443463e-05,
      "loss": 1.1635,
      "mean_token_accuracy": 0.6888069398701191,
      "num_tokens": 7596172.0,
      "step": 238
    },
    {
      "entropy": 1.0232941769063473,
      "epoch": 0.021989522774452132,
      "grad_norm": 0.17438723146915436,
      "learning_rate": 9.927009537829301e-05,
      "loss": 1.0459,
      "mean_token_accuracy": 0.7175581082701683,
      "num_tokens": 7628213.0,
      "step": 239
    },
    {
      "entropy": 0.9586729891598225,
      "epoch": 0.022081529145893352,
      "grad_norm": 0.17714299261569977,
      "learning_rate": 9.926702855215139e-05,
      "loss": 0.9893,
      "mean_token_accuracy": 0.7272261045873165,
      "num_tokens": 7660119.0,
      "step": 240
    },
    {
      "entropy": 0.9716028645634651,
      "epoch": 0.022173535517334576,
      "grad_norm": 0.16684122383594513,
      "learning_rate": 9.926396172600976e-05,
      "loss": 0.9562,
      "mean_token_accuracy": 0.7407666072249413,
      "num_tokens": 7691405.0,
      "step": 241
    },
    {
      "entropy": 0.9912166967988014,
      "epoch": 0.0222655418887758,
      "grad_norm": 0.16787931323051453,
      "learning_rate": 9.926089489986813e-05,
      "loss": 1.001,
      "mean_token_accuracy": 0.7326581627130508,
      "num_tokens": 7723103.0,
      "step": 242
    },
    {
      "entropy": 1.0880173332989216,
      "epoch": 0.02235754826021702,
      "grad_norm": 0.1724301427602768,
      "learning_rate": 9.92578280737265e-05,
      "loss": 1.1175,
      "mean_token_accuracy": 0.7024666741490364,
      "num_tokens": 7754934.0,
      "step": 243
    },
    {
      "entropy": 1.103767417371273,
      "epoch": 0.022449554631658243,
      "grad_norm": 0.18482881784439087,
      "learning_rate": 9.925476124758489e-05,
      "loss": 1.1171,
      "mean_token_accuracy": 0.7053827308118343,
      "num_tokens": 7786846.0,
      "step": 244
    },
    {
      "entropy": 1.1106660179793835,
      "epoch": 0.022541561003099463,
      "grad_norm": 0.1809864044189453,
      "learning_rate": 9.925169442144326e-05,
      "loss": 1.145,
      "mean_token_accuracy": 0.6927794851362705,
      "num_tokens": 7818471.0,
      "step": 245
    },
    {
      "entropy": 1.0490600168704987,
      "epoch": 0.022633567374540687,
      "grad_norm": 0.17510442435741425,
      "learning_rate": 9.924862759530162e-05,
      "loss": 1.0377,
      "mean_token_accuracy": 0.714185319840908,
      "num_tokens": 7850985.0,
      "step": 246
    },
    {
      "entropy": 0.9827646017074585,
      "epoch": 0.02272557374598191,
      "grad_norm": 0.16269072890281677,
      "learning_rate": 9.924556076916e-05,
      "loss": 0.9903,
      "mean_token_accuracy": 0.7312794104218483,
      "num_tokens": 7883736.0,
      "step": 247
    },
    {
      "entropy": 0.950873289257288,
      "epoch": 0.02281758011742313,
      "grad_norm": 0.1698489636182785,
      "learning_rate": 9.924249394301837e-05,
      "loss": 0.9561,
      "mean_token_accuracy": 0.740793701261282,
      "num_tokens": 7915107.0,
      "step": 248
    },
    {
      "entropy": 1.0899472013115883,
      "epoch": 0.022909586488864354,
      "grad_norm": 0.17524008452892303,
      "learning_rate": 9.923942711687675e-05,
      "loss": 1.0865,
      "mean_token_accuracy": 0.707052443176508,
      "num_tokens": 7946704.0,
      "step": 249
    },
    {
      "entropy": 1.0490071326494217,
      "epoch": 0.023001592860305575,
      "grad_norm": 0.1749829202890396,
      "learning_rate": 9.923636029073512e-05,
      "loss": 1.0472,
      "mean_token_accuracy": 0.7164794690907001,
      "num_tokens": 7978206.0,
      "step": 250
    },
    {
      "entropy": 1.0940250530838966,
      "epoch": 0.023093599231746798,
      "grad_norm": 0.17316441237926483,
      "learning_rate": 9.92332934645935e-05,
      "loss": 1.1086,
      "mean_token_accuracy": 0.7035567164421082,
      "num_tokens": 8010569.0,
      "step": 251
    },
    {
      "entropy": 0.9831471145153046,
      "epoch": 0.023185605603188022,
      "grad_norm": 0.17159403860569,
      "learning_rate": 9.923022663845187e-05,
      "loss": 0.9737,
      "mean_token_accuracy": 0.7336002290248871,
      "num_tokens": 8042545.0,
      "step": 252
    },
    {
      "entropy": 1.0237773321568966,
      "epoch": 0.023277611974629242,
      "grad_norm": 0.17272478342056274,
      "learning_rate": 9.922715981231024e-05,
      "loss": 1.0399,
      "mean_token_accuracy": 0.7164369411766529,
      "num_tokens": 8074756.0,
      "step": 253
    },
    {
      "entropy": 0.9993202462792397,
      "epoch": 0.023369618346070466,
      "grad_norm": 0.16742482781410217,
      "learning_rate": 9.922409298616862e-05,
      "loss": 1.0292,
      "mean_token_accuracy": 0.7237544730305672,
      "num_tokens": 8106878.0,
      "step": 254
    },
    {
      "entropy": 1.0520861260592937,
      "epoch": 0.02346162471751169,
      "grad_norm": 0.1790105253458023,
      "learning_rate": 9.9221026160027e-05,
      "loss": 1.0857,
      "mean_token_accuracy": 0.7121801152825356,
      "num_tokens": 8138578.0,
      "step": 255
    },
    {
      "entropy": 1.0587811097502708,
      "epoch": 0.02355363108895291,
      "grad_norm": 0.1684771478176117,
      "learning_rate": 9.921795933388537e-05,
      "loss": 1.0777,
      "mean_token_accuracy": 0.7124390602111816,
      "num_tokens": 8170435.0,
      "step": 256
    },
    {
      "entropy": 0.9994714818894863,
      "epoch": 0.023645637460394133,
      "grad_norm": 0.162997767329216,
      "learning_rate": 9.921489250774374e-05,
      "loss": 1.0174,
      "mean_token_accuracy": 0.7226156555116177,
      "num_tokens": 8202719.0,
      "step": 257
    },
    {
      "entropy": 1.0775707848370075,
      "epoch": 0.023737643831835353,
      "grad_norm": 0.17524001002311707,
      "learning_rate": 9.92118256816021e-05,
      "loss": 1.0715,
      "mean_token_accuracy": 0.7084484696388245,
      "num_tokens": 8234194.0,
      "step": 258
    },
    {
      "entropy": 0.9745459966361523,
      "epoch": 0.023829650203276577,
      "grad_norm": 0.17161127924919128,
      "learning_rate": 9.920875885546049e-05,
      "loss": 0.9872,
      "mean_token_accuracy": 0.7289598882198334,
      "num_tokens": 8266441.0,
      "step": 259
    },
    {
      "entropy": 1.037320762872696,
      "epoch": 0.0239216565747178,
      "grad_norm": 0.17044520378112793,
      "learning_rate": 9.920569202931887e-05,
      "loss": 1.0413,
      "mean_token_accuracy": 0.7180324047803879,
      "num_tokens": 8298867.0,
      "step": 260
    },
    {
      "entropy": 0.9870064221322536,
      "epoch": 0.02401366294615902,
      "grad_norm": 0.16930106282234192,
      "learning_rate": 9.920262520317724e-05,
      "loss": 0.9938,
      "mean_token_accuracy": 0.7334766872227192,
      "num_tokens": 8330670.0,
      "step": 261
    },
    {
      "entropy": 1.0553007870912552,
      "epoch": 0.024105669317600244,
      "grad_norm": 0.17598919570446014,
      "learning_rate": 9.91995583770356e-05,
      "loss": 1.0727,
      "mean_token_accuracy": 0.7125541009008884,
      "num_tokens": 8362592.0,
      "step": 262
    },
    {
      "entropy": 0.9725649319589138,
      "epoch": 0.024197675689041465,
      "grad_norm": 0.1814746856689453,
      "learning_rate": 9.919649155089399e-05,
      "loss": 0.9655,
      "mean_token_accuracy": 0.734922245144844,
      "num_tokens": 8395258.0,
      "step": 263
    },
    {
      "entropy": 1.10891717299819,
      "epoch": 0.024289682060482688,
      "grad_norm": 0.1690898835659027,
      "learning_rate": 9.919342472475235e-05,
      "loss": 1.0987,
      "mean_token_accuracy": 0.7040084637701511,
      "num_tokens": 8426905.0,
      "step": 264
    },
    {
      "entropy": 1.072560764849186,
      "epoch": 0.024381688431923912,
      "grad_norm": 0.17726925015449524,
      "learning_rate": 9.919035789861074e-05,
      "loss": 1.0731,
      "mean_token_accuracy": 0.7107347324490547,
      "num_tokens": 8458525.0,
      "step": 265
    },
    {
      "entropy": 0.9354403205215931,
      "epoch": 0.024473694803365132,
      "grad_norm": 0.1571246087551117,
      "learning_rate": 9.91872910724691e-05,
      "loss": 0.949,
      "mean_token_accuracy": 0.7437668591737747,
      "num_tokens": 8490779.0,
      "step": 266
    },
    {
      "entropy": 0.9857550449669361,
      "epoch": 0.024565701174806356,
      "grad_norm": 0.17514096200466156,
      "learning_rate": 9.918422424632748e-05,
      "loss": 1.0015,
      "mean_token_accuracy": 0.7249152995646,
      "num_tokens": 8522916.0,
      "step": 267
    },
    {
      "entropy": 1.0511173941195011,
      "epoch": 0.02465770754624758,
      "grad_norm": 0.17077544331550598,
      "learning_rate": 9.918115742018585e-05,
      "loss": 1.0863,
      "mean_token_accuracy": 0.7092981413006783,
      "num_tokens": 8554639.0,
      "step": 268
    },
    {
      "entropy": 0.9296869300305843,
      "epoch": 0.0247497139176888,
      "grad_norm": 0.15971146523952484,
      "learning_rate": 9.917809059404422e-05,
      "loss": 0.9405,
      "mean_token_accuracy": 0.7404145710170269,
      "num_tokens": 8586912.0,
      "step": 269
    },
    {
      "entropy": 1.0132721588015556,
      "epoch": 0.024841720289130023,
      "grad_norm": 0.16618159413337708,
      "learning_rate": 9.91750237679026e-05,
      "loss": 1.0335,
      "mean_token_accuracy": 0.7160226330161095,
      "num_tokens": 8618305.0,
      "step": 270
    },
    {
      "entropy": 1.042600091546774,
      "epoch": 0.024933726660571243,
      "grad_norm": 0.17565205693244934,
      "learning_rate": 9.917195694176098e-05,
      "loss": 1.06,
      "mean_token_accuracy": 0.7123121619224548,
      "num_tokens": 8649405.0,
      "step": 271
    },
    {
      "entropy": 0.9981344006955624,
      "epoch": 0.025025733032012467,
      "grad_norm": 0.17246222496032715,
      "learning_rate": 9.916889011561935e-05,
      "loss": 1.0024,
      "mean_token_accuracy": 0.7237122468650341,
      "num_tokens": 8681294.0,
      "step": 272
    },
    {
      "entropy": 1.0209496840834618,
      "epoch": 0.02511773940345369,
      "grad_norm": 0.16784678399562836,
      "learning_rate": 9.916582328947772e-05,
      "loss": 1.0036,
      "mean_token_accuracy": 0.7270780764520168,
      "num_tokens": 8713661.0,
      "step": 273
    },
    {
      "entropy": 0.9887788705527782,
      "epoch": 0.02520974577489491,
      "grad_norm": 0.1617519110441208,
      "learning_rate": 9.916275646333609e-05,
      "loss": 0.99,
      "mean_token_accuracy": 0.7291125245392323,
      "num_tokens": 8746074.0,
      "step": 274
    },
    {
      "entropy": 1.0359355509281158,
      "epoch": 0.025301752146336134,
      "grad_norm": 0.16999831795692444,
      "learning_rate": 9.915968963719447e-05,
      "loss": 1.0335,
      "mean_token_accuracy": 0.7175508104264736,
      "num_tokens": 8778046.0,
      "step": 275
    },
    {
      "entropy": 1.0637561157345772,
      "epoch": 0.025393758517777355,
      "grad_norm": 0.17285063862800598,
      "learning_rate": 9.915662281105285e-05,
      "loss": 1.0669,
      "mean_token_accuracy": 0.712538868188858,
      "num_tokens": 8810719.0,
      "step": 276
    },
    {
      "entropy": 1.1193933859467506,
      "epoch": 0.02548576488921858,
      "grad_norm": 0.1822524517774582,
      "learning_rate": 9.915355598491122e-05,
      "loss": 1.1189,
      "mean_token_accuracy": 0.6996394693851471,
      "num_tokens": 8842553.0,
      "step": 277
    },
    {
      "entropy": 1.01952401176095,
      "epoch": 0.025577771260659802,
      "grad_norm": 0.1692618429660797,
      "learning_rate": 9.91504891587696e-05,
      "loss": 1.0436,
      "mean_token_accuracy": 0.7193587422370911,
      "num_tokens": 8874183.0,
      "step": 278
    },
    {
      "entropy": 1.023792751133442,
      "epoch": 0.025669777632101022,
      "grad_norm": 0.16517235338687897,
      "learning_rate": 9.914742233262797e-05,
      "loss": 1.0305,
      "mean_token_accuracy": 0.7161813750863075,
      "num_tokens": 8906386.0,
      "step": 279
    },
    {
      "entropy": 1.0364782847464085,
      "epoch": 0.025761784003542246,
      "grad_norm": 0.17772650718688965,
      "learning_rate": 9.914435550648633e-05,
      "loss": 1.0604,
      "mean_token_accuracy": 0.7158019207417965,
      "num_tokens": 8938098.0,
      "step": 280
    },
    {
      "entropy": 1.0896171107888222,
      "epoch": 0.025853790374983466,
      "grad_norm": 0.1809283047914505,
      "learning_rate": 9.914128868034472e-05,
      "loss": 1.1138,
      "mean_token_accuracy": 0.7009692788124084,
      "num_tokens": 8970207.0,
      "step": 281
    },
    {
      "entropy": 1.0545813851058483,
      "epoch": 0.02594579674642469,
      "grad_norm": 0.17223653197288513,
      "learning_rate": 9.91382218542031e-05,
      "loss": 1.0585,
      "mean_token_accuracy": 0.717927597463131,
      "num_tokens": 9002756.0,
      "step": 282
    },
    {
      "entropy": 1.0567163042724133,
      "epoch": 0.026037803117865913,
      "grad_norm": 0.1732850819826126,
      "learning_rate": 9.913515502806147e-05,
      "loss": 1.0529,
      "mean_token_accuracy": 0.7167305201292038,
      "num_tokens": 9035051.0,
      "step": 283
    },
    {
      "entropy": 1.0590565912425518,
      "epoch": 0.026129809489307133,
      "grad_norm": 0.1697113811969757,
      "learning_rate": 9.913208820191983e-05,
      "loss": 1.072,
      "mean_token_accuracy": 0.7100552208721638,
      "num_tokens": 9066738.0,
      "step": 284
    },
    {
      "entropy": 1.0653780922293663,
      "epoch": 0.026221815860748357,
      "grad_norm": 0.1827525943517685,
      "learning_rate": 9.912902137577821e-05,
      "loss": 1.1067,
      "mean_token_accuracy": 0.7032938525080681,
      "num_tokens": 9098032.0,
      "step": 285
    },
    {
      "entropy": 1.0552263334393501,
      "epoch": 0.02631382223218958,
      "grad_norm": 0.17574644088745117,
      "learning_rate": 9.91259545496366e-05,
      "loss": 1.0687,
      "mean_token_accuracy": 0.7146542407572269,
      "num_tokens": 9130477.0,
      "step": 286
    },
    {
      "entropy": 1.138304192572832,
      "epoch": 0.0264058286036308,
      "grad_norm": 0.17941315472126007,
      "learning_rate": 9.912288772349496e-05,
      "loss": 1.1307,
      "mean_token_accuracy": 0.6898724101483822,
      "num_tokens": 9162249.0,
      "step": 287
    },
    {
      "entropy": 1.0071697924286127,
      "epoch": 0.026497834975072024,
      "grad_norm": 0.1770455241203308,
      "learning_rate": 9.911982089735333e-05,
      "loss": 1.0195,
      "mean_token_accuracy": 0.7261302918195724,
      "num_tokens": 9193869.0,
      "step": 288
    },
    {
      "entropy": 1.06841379404068,
      "epoch": 0.026589841346513245,
      "grad_norm": 0.17001360654830933,
      "learning_rate": 9.91167540712117e-05,
      "loss": 1.0827,
      "mean_token_accuracy": 0.7073499746620655,
      "num_tokens": 9225603.0,
      "step": 289
    },
    {
      "entropy": 1.0507192350924015,
      "epoch": 0.02668184771795447,
      "grad_norm": 0.16357386112213135,
      "learning_rate": 9.911368724507008e-05,
      "loss": 1.0614,
      "mean_token_accuracy": 0.7118343971669674,
      "num_tokens": 9258371.0,
      "step": 290
    },
    {
      "entropy": 1.0142629072070122,
      "epoch": 0.026773854089395692,
      "grad_norm": 0.16867002844810486,
      "learning_rate": 9.911062041892846e-05,
      "loss": 1.0252,
      "mean_token_accuracy": 0.7260223962366581,
      "num_tokens": 9289788.0,
      "step": 291
    },
    {
      "entropy": 1.0012767743319273,
      "epoch": 0.026865860460836912,
      "grad_norm": 0.17086906731128693,
      "learning_rate": 9.910755359278683e-05,
      "loss": 1.0076,
      "mean_token_accuracy": 0.7254475429654121,
      "num_tokens": 9321586.0,
      "step": 292
    },
    {
      "entropy": 1.079236552119255,
      "epoch": 0.026957866832278136,
      "grad_norm": 0.17184540629386902,
      "learning_rate": 9.91044867666452e-05,
      "loss": 1.0926,
      "mean_token_accuracy": 0.7041040547192097,
      "num_tokens": 9354146.0,
      "step": 293
    },
    {
      "entropy": 0.9915536604821682,
      "epoch": 0.027049873203719356,
      "grad_norm": 0.1680537909269333,
      "learning_rate": 9.910141994050358e-05,
      "loss": 1.0015,
      "mean_token_accuracy": 0.725428719073534,
      "num_tokens": 9386302.0,
      "step": 294
    },
    {
      "entropy": 1.0315271653234959,
      "epoch": 0.02714187957516058,
      "grad_norm": 0.17213064432144165,
      "learning_rate": 9.909835311436195e-05,
      "loss": 1.0478,
      "mean_token_accuracy": 0.7170222699642181,
      "num_tokens": 9418582.0,
      "step": 295
    },
    {
      "entropy": 1.0609200932085514,
      "epoch": 0.027233885946601803,
      "grad_norm": 0.17062462866306305,
      "learning_rate": 9.909528628822033e-05,
      "loss": 1.061,
      "mean_token_accuracy": 0.7154916524887085,
      "num_tokens": 9450630.0,
      "step": 296
    },
    {
      "entropy": 1.0197396203875542,
      "epoch": 0.027325892318043023,
      "grad_norm": 0.16153572499752045,
      "learning_rate": 9.90922194620787e-05,
      "loss": 1.012,
      "mean_token_accuracy": 0.722888607531786,
      "num_tokens": 9482509.0,
      "step": 297
    },
    {
      "entropy": 1.0288560949265957,
      "epoch": 0.027417898689484247,
      "grad_norm": 0.16769398748874664,
      "learning_rate": 9.908915263593708e-05,
      "loss": 1.0286,
      "mean_token_accuracy": 0.7228196486830711,
      "num_tokens": 9514677.0,
      "step": 298
    },
    {
      "entropy": 1.0114391669631004,
      "epoch": 0.02750990506092547,
      "grad_norm": 0.1745382398366928,
      "learning_rate": 9.908608580979545e-05,
      "loss": 1.0268,
      "mean_token_accuracy": 0.7244320549070835,
      "num_tokens": 9546815.0,
      "step": 299
    },
    {
      "entropy": 1.046730425208807,
      "epoch": 0.02760191143236669,
      "grad_norm": 0.17665036022663116,
      "learning_rate": 9.908301898365381e-05,
      "loss": 1.057,
      "mean_token_accuracy": 0.7112526260316372,
      "num_tokens": 9579056.0,
      "step": 300
    },
    {
      "entropy": 1.0406011492013931,
      "epoch": 0.027693917803807915,
      "grad_norm": 0.16710369288921356,
      "learning_rate": 9.90799521575122e-05,
      "loss": 1.0296,
      "mean_token_accuracy": 0.7178775668144226,
      "num_tokens": 9611561.0,
      "step": 301
    },
    {
      "entropy": 1.0182667076587677,
      "epoch": 0.027785924175249135,
      "grad_norm": 0.172878235578537,
      "learning_rate": 9.907688533137058e-05,
      "loss": 1.0276,
      "mean_token_accuracy": 0.7248609066009521,
      "num_tokens": 9643126.0,
      "step": 302
    },
    {
      "entropy": 0.9999312683939934,
      "epoch": 0.02787793054669036,
      "grad_norm": 0.170984148979187,
      "learning_rate": 9.907381850522895e-05,
      "loss": 1.0095,
      "mean_token_accuracy": 0.7265542894601822,
      "num_tokens": 9675352.0,
      "step": 303
    },
    {
      "entropy": 1.063875325024128,
      "epoch": 0.027969936918131582,
      "grad_norm": 0.16845428943634033,
      "learning_rate": 9.907075167908731e-05,
      "loss": 1.058,
      "mean_token_accuracy": 0.7120595499873161,
      "num_tokens": 9707663.0,
      "step": 304
    },
    {
      "entropy": 1.007687259465456,
      "epoch": 0.028061943289572802,
      "grad_norm": 0.16646651923656464,
      "learning_rate": 9.906768485294568e-05,
      "loss": 1.0245,
      "mean_token_accuracy": 0.7184119075536728,
      "num_tokens": 9739133.0,
      "step": 305
    },
    {
      "entropy": 1.0903705582022667,
      "epoch": 0.028153949661014026,
      "grad_norm": 0.16926497220993042,
      "learning_rate": 9.906461802680406e-05,
      "loss": 1.1003,
      "mean_token_accuracy": 0.7001234889030457,
      "num_tokens": 9771504.0,
      "step": 306
    },
    {
      "entropy": 0.9451657794415951,
      "epoch": 0.028245956032455246,
      "grad_norm": 0.1617155224084854,
      "learning_rate": 9.906155120066244e-05,
      "loss": 0.9663,
      "mean_token_accuracy": 0.7369178421795368,
      "num_tokens": 9803686.0,
      "step": 307
    },
    {
      "entropy": 1.040549948811531,
      "epoch": 0.02833796240389647,
      "grad_norm": 0.1738441288471222,
      "learning_rate": 9.905848437452081e-05,
      "loss": 1.0805,
      "mean_token_accuracy": 0.7120287865400314,
      "num_tokens": 9835487.0,
      "step": 308
    },
    {
      "entropy": 1.0481392070651054,
      "epoch": 0.028429968775337693,
      "grad_norm": 0.16806606948375702,
      "learning_rate": 9.905541754837919e-05,
      "loss": 1.0478,
      "mean_token_accuracy": 0.714185394346714,
      "num_tokens": 9868255.0,
      "step": 309
    },
    {
      "entropy": 1.0710994862020016,
      "epoch": 0.028521975146778913,
      "grad_norm": 0.16525167226791382,
      "learning_rate": 9.905235072223756e-05,
      "loss": 1.0628,
      "mean_token_accuracy": 0.7152832001447678,
      "num_tokens": 9900598.0,
      "step": 310
    },
    {
      "entropy": 1.021608266979456,
      "epoch": 0.028613981518220137,
      "grad_norm": 0.17442676424980164,
      "learning_rate": 9.904928389609593e-05,
      "loss": 1.0374,
      "mean_token_accuracy": 0.7184325978159904,
      "num_tokens": 9933282.0,
      "step": 311
    },
    {
      "entropy": 0.9169892519712448,
      "epoch": 0.02870598788966136,
      "grad_norm": 0.1616387814283371,
      "learning_rate": 9.904621706995431e-05,
      "loss": 0.9081,
      "mean_token_accuracy": 0.7482005804777145,
      "num_tokens": 9964747.0,
      "step": 312
    },
    {
      "entropy": 1.005317285656929,
      "epoch": 0.02879799426110258,
      "grad_norm": 0.16239547729492188,
      "learning_rate": 9.904315024381269e-05,
      "loss": 1.0069,
      "mean_token_accuracy": 0.7258857823908329,
      "num_tokens": 9996951.0,
      "step": 313
    },
    {
      "entropy": 0.9987409636378288,
      "epoch": 0.028890000632543805,
      "grad_norm": 0.16818250715732574,
      "learning_rate": 9.904008341767106e-05,
      "loss": 1.0247,
      "mean_token_accuracy": 0.7210480906069279,
      "num_tokens": 10029242.0,
      "step": 314
    },
    {
      "entropy": 0.964073896408081,
      "epoch": 0.028982007003985025,
      "grad_norm": 0.16994626820087433,
      "learning_rate": 9.903701659152943e-05,
      "loss": 0.982,
      "mean_token_accuracy": 0.7298190332949162,
      "num_tokens": 10060991.0,
      "step": 315
    },
    {
      "entropy": 0.9543931856751442,
      "epoch": 0.02907401337542625,
      "grad_norm": 0.16848395764827728,
      "learning_rate": 9.90339497653878e-05,
      "loss": 0.9554,
      "mean_token_accuracy": 0.7392972037196159,
      "num_tokens": 10092197.0,
      "step": 316
    },
    {
      "entropy": 1.09754778444767,
      "epoch": 0.029166019746867472,
      "grad_norm": 0.1701204478740692,
      "learning_rate": 9.903088293924619e-05,
      "loss": 1.1101,
      "mean_token_accuracy": 0.7063821814954281,
      "num_tokens": 10124316.0,
      "step": 317
    },
    {
      "entropy": 1.04968723654747,
      "epoch": 0.029258026118308692,
      "grad_norm": 0.16718478500843048,
      "learning_rate": 9.902781611310456e-05,
      "loss": 1.0484,
      "mean_token_accuracy": 0.7153932489454746,
      "num_tokens": 10156969.0,
      "step": 318
    },
    {
      "entropy": 1.0543867610394955,
      "epoch": 0.029350032489749916,
      "grad_norm": 0.1761874258518219,
      "learning_rate": 9.902474928696293e-05,
      "loss": 1.0653,
      "mean_token_accuracy": 0.7080084159970284,
      "num_tokens": 10188671.0,
      "step": 319
    },
    {
      "entropy": 0.9748063422739506,
      "epoch": 0.029442038861191136,
      "grad_norm": 0.16548579931259155,
      "learning_rate": 9.90216824608213e-05,
      "loss": 0.9972,
      "mean_token_accuracy": 0.7286731041967869,
      "num_tokens": 10220771.0,
      "step": 320
    },
    {
      "entropy": 0.9415609240531921,
      "epoch": 0.02953404523263236,
      "grad_norm": 0.16453656554222107,
      "learning_rate": 9.901861563467968e-05,
      "loss": 0.9522,
      "mean_token_accuracy": 0.7377289608120918,
      "num_tokens": 10253379.0,
      "step": 321
    },
    {
      "entropy": 1.0831911489367485,
      "epoch": 0.029626051604073583,
      "grad_norm": 0.16672056913375854,
      "learning_rate": 9.901554880853806e-05,
      "loss": 1.1081,
      "mean_token_accuracy": 0.7069454789161682,
      "num_tokens": 10285118.0,
      "step": 322
    },
    {
      "entropy": 0.940744161605835,
      "epoch": 0.029718057975514803,
      "grad_norm": 0.16158607602119446,
      "learning_rate": 9.901248198239642e-05,
      "loss": 0.9307,
      "mean_token_accuracy": 0.7433595806360245,
      "num_tokens": 10316942.0,
      "step": 323
    },
    {
      "entropy": 1.037747297435999,
      "epoch": 0.029810064346956027,
      "grad_norm": 0.16587531566619873,
      "learning_rate": 9.900941515625479e-05,
      "loss": 1.0514,
      "mean_token_accuracy": 0.7119894027709961,
      "num_tokens": 10348879.0,
      "step": 324
    },
    {
      "entropy": 1.1704398952424526,
      "epoch": 0.029902070718397247,
      "grad_norm": 0.18025659024715424,
      "learning_rate": 9.900634833011317e-05,
      "loss": 1.1702,
      "mean_token_accuracy": 0.6833353005349636,
      "num_tokens": 10380799.0,
      "step": 325
    },
    {
      "entropy": 1.003839686512947,
      "epoch": 0.02999407708983847,
      "grad_norm": 0.16270101070404053,
      "learning_rate": 9.900328150397154e-05,
      "loss": 0.9927,
      "mean_token_accuracy": 0.72690499573946,
      "num_tokens": 10412441.0,
      "step": 326
    },
    {
      "entropy": 1.124996893107891,
      "epoch": 0.030086083461279695,
      "grad_norm": 0.17676809430122375,
      "learning_rate": 9.900021467782992e-05,
      "loss": 1.113,
      "mean_token_accuracy": 0.6959314085543156,
      "num_tokens": 10443653.0,
      "step": 327
    },
    {
      "entropy": 1.0923333875834942,
      "epoch": 0.030178089832720915,
      "grad_norm": 0.17544600367546082,
      "learning_rate": 9.899714785168829e-05,
      "loss": 1.0823,
      "mean_token_accuracy": 0.706365168094635,
      "num_tokens": 10475689.0,
      "step": 328
    },
    {
      "entropy": 1.0189785361289978,
      "epoch": 0.03027009620416214,
      "grad_norm": 0.17241306602954865,
      "learning_rate": 9.899408102554667e-05,
      "loss": 1.0328,
      "mean_token_accuracy": 0.7179580628871918,
      "num_tokens": 10507652.0,
      "step": 329
    },
    {
      "entropy": 1.0077387765049934,
      "epoch": 0.030362102575603362,
      "grad_norm": 0.16874080896377563,
      "learning_rate": 9.899101419940504e-05,
      "loss": 1.016,
      "mean_token_accuracy": 0.723788995295763,
      "num_tokens": 10539961.0,
      "step": 330
    },
    {
      "entropy": 1.0799332521855831,
      "epoch": 0.030454108947044582,
      "grad_norm": 0.17733213305473328,
      "learning_rate": 9.898794737326341e-05,
      "loss": 1.0997,
      "mean_token_accuracy": 0.7020038515329361,
      "num_tokens": 10572018.0,
      "step": 331
    },
    {
      "entropy": 0.9794340506196022,
      "epoch": 0.030546115318485806,
      "grad_norm": 0.17030216753482819,
      "learning_rate": 9.898488054712179e-05,
      "loss": 0.9891,
      "mean_token_accuracy": 0.7313051484525204,
      "num_tokens": 10603519.0,
      "step": 332
    },
    {
      "entropy": 1.0041477158665657,
      "epoch": 0.030638121689927026,
      "grad_norm": 0.16841651499271393,
      "learning_rate": 9.898181372098017e-05,
      "loss": 1.0202,
      "mean_token_accuracy": 0.7223141491413116,
      "num_tokens": 10635784.0,
      "step": 333
    },
    {
      "entropy": 0.959893062710762,
      "epoch": 0.03073012806136825,
      "grad_norm": 0.5746650695800781,
      "learning_rate": 9.897874689483854e-05,
      "loss": 0.9843,
      "mean_token_accuracy": 0.7305244393646717,
      "num_tokens": 10667797.0,
      "step": 334
    },
    {
      "entropy": 1.000692903995514,
      "epoch": 0.030822134432809473,
      "grad_norm": 0.1638183742761612,
      "learning_rate": 9.89756800686969e-05,
      "loss": 1.0132,
      "mean_token_accuracy": 0.7241126894950867,
      "num_tokens": 10700031.0,
      "step": 335
    },
    {
      "entropy": 0.9820724241435528,
      "epoch": 0.030914140804250693,
      "grad_norm": 0.16778138279914856,
      "learning_rate": 9.897261324255527e-05,
      "loss": 0.9854,
      "mean_token_accuracy": 0.731487549841404,
      "num_tokens": 10731655.0,
      "step": 336
    },
    {
      "entropy": 0.9605408236384392,
      "epoch": 0.031006147175691917,
      "grad_norm": 0.16641676425933838,
      "learning_rate": 9.896954641641366e-05,
      "loss": 0.9656,
      "mean_token_accuracy": 0.7369475848972797,
      "num_tokens": 10762923.0,
      "step": 337
    },
    {
      "entropy": 0.9578608199954033,
      "epoch": 0.031098153547133137,
      "grad_norm": 0.16755624115467072,
      "learning_rate": 9.896647959027204e-05,
      "loss": 0.9675,
      "mean_token_accuracy": 0.7369200922548771,
      "num_tokens": 10795679.0,
      "step": 338
    },
    {
      "entropy": 0.9871492795646191,
      "epoch": 0.03119015991857436,
      "grad_norm": 0.16907480359077454,
      "learning_rate": 9.89634127641304e-05,
      "loss": 1.0065,
      "mean_token_accuracy": 0.7239247746765614,
      "num_tokens": 10827842.0,
      "step": 339
    },
    {
      "entropy": 0.9234685860574245,
      "epoch": 0.03128216629001558,
      "grad_norm": 0.1675400733947754,
      "learning_rate": 9.896034593798879e-05,
      "loss": 0.9124,
      "mean_token_accuracy": 0.7448959350585938,
      "num_tokens": 10858112.0,
      "step": 340
    },
    {
      "entropy": 0.9544042199850082,
      "epoch": 0.031374172661456805,
      "grad_norm": 0.17795246839523315,
      "learning_rate": 9.895727911184715e-05,
      "loss": 0.9408,
      "mean_token_accuracy": 0.7383503653109074,
      "num_tokens": 10889947.0,
      "step": 341
    },
    {
      "entropy": 1.0911731496453285,
      "epoch": 0.03146617903289803,
      "grad_norm": 0.1762719303369522,
      "learning_rate": 9.895421228570552e-05,
      "loss": 1.0983,
      "mean_token_accuracy": 0.70502058416605,
      "num_tokens": 10921061.0,
      "step": 342
    },
    {
      "entropy": 0.996724471449852,
      "epoch": 0.03155818540433925,
      "grad_norm": 0.6094722151756287,
      "learning_rate": 9.89511454595639e-05,
      "loss": 0.9862,
      "mean_token_accuracy": 0.7317587547004223,
      "num_tokens": 10952779.0,
      "step": 343
    },
    {
      "entropy": 0.9543893337249756,
      "epoch": 0.031650191775780476,
      "grad_norm": 0.16950005292892456,
      "learning_rate": 9.894807863342229e-05,
      "loss": 0.9869,
      "mean_token_accuracy": 0.7318556904792786,
      "num_tokens": 10984672.0,
      "step": 344
    },
    {
      "entropy": 0.9094912111759186,
      "epoch": 0.03174219814722169,
      "grad_norm": 0.16138726472854614,
      "learning_rate": 9.894501180728065e-05,
      "loss": 0.9011,
      "mean_token_accuracy": 0.7496962770819664,
      "num_tokens": 11016694.0,
      "step": 345
    },
    {
      "entropy": 0.9806951321661472,
      "epoch": 0.031834204518662916,
      "grad_norm": 0.17097319662570953,
      "learning_rate": 9.894194498113902e-05,
      "loss": 0.9908,
      "mean_token_accuracy": 0.7286317944526672,
      "num_tokens": 11047913.0,
      "step": 346
    },
    {
      "entropy": 0.9940336756408215,
      "epoch": 0.03192621089010414,
      "grad_norm": 0.16413697600364685,
      "learning_rate": 9.893887815499739e-05,
      "loss": 0.9996,
      "mean_token_accuracy": 0.7252663858234882,
      "num_tokens": 11080672.0,
      "step": 347
    },
    {
      "entropy": 0.9757176823914051,
      "epoch": 0.03201821726154536,
      "grad_norm": 0.16703853011131287,
      "learning_rate": 9.893581132885577e-05,
      "loss": 1.0129,
      "mean_token_accuracy": 0.7229591645300388,
      "num_tokens": 11112994.0,
      "step": 348
    },
    {
      "entropy": 1.0398691222071648,
      "epoch": 0.03211022363298659,
      "grad_norm": 0.17201068997383118,
      "learning_rate": 9.893274450271415e-05,
      "loss": 1.0487,
      "mean_token_accuracy": 0.7168041057884693,
      "num_tokens": 11144588.0,
      "step": 349
    },
    {
      "entropy": 0.988932553678751,
      "epoch": 0.032202230004427804,
      "grad_norm": 0.18128854036331177,
      "learning_rate": 9.892967767657252e-05,
      "loss": 1.0086,
      "mean_token_accuracy": 0.7216555811464787,
      "num_tokens": 11176346.0,
      "step": 350
    },
    {
      "entropy": 0.9840069264173508,
      "epoch": 0.03229423637586903,
      "grad_norm": 0.1725808084011078,
      "learning_rate": 9.892661085043089e-05,
      "loss": 0.9899,
      "mean_token_accuracy": 0.7303399927914143,
      "num_tokens": 11208714.0,
      "step": 351
    },
    {
      "entropy": 0.9876339249312878,
      "epoch": 0.03238624274731025,
      "grad_norm": 0.16013628244400024,
      "learning_rate": 9.892354402428927e-05,
      "loss": 0.9757,
      "mean_token_accuracy": 0.7265531346201897,
      "num_tokens": 11240982.0,
      "step": 352
    },
    {
      "entropy": 0.989908829331398,
      "epoch": 0.032478249118751475,
      "grad_norm": 0.16509833931922913,
      "learning_rate": 9.892047719814764e-05,
      "loss": 0.975,
      "mean_token_accuracy": 0.7294861748814583,
      "num_tokens": 11272566.0,
      "step": 353
    },
    {
      "entropy": 0.9569560680538416,
      "epoch": 0.0325702554901927,
      "grad_norm": 0.15770207345485687,
      "learning_rate": 9.891741037200602e-05,
      "loss": 0.9456,
      "mean_token_accuracy": 0.7373073622584343,
      "num_tokens": 11304426.0,
      "step": 354
    },
    {
      "entropy": 1.0845520347356796,
      "epoch": 0.032662261861633915,
      "grad_norm": 0.16966894268989563,
      "learning_rate": 9.891434354586439e-05,
      "loss": 1.0896,
      "mean_token_accuracy": 0.7080894820392132,
      "num_tokens": 11336514.0,
      "step": 355
    },
    {
      "entropy": 0.9902093820273876,
      "epoch": 0.03275426823307514,
      "grad_norm": 0.18424855172634125,
      "learning_rate": 9.891127671972277e-05,
      "loss": 0.9918,
      "mean_token_accuracy": 0.7236075885593891,
      "num_tokens": 11368580.0,
      "step": 356
    },
    {
      "entropy": 1.1136357635259628,
      "epoch": 0.03284627460451636,
      "grad_norm": 0.17254707217216492,
      "learning_rate": 9.890820989358114e-05,
      "loss": 1.1236,
      "mean_token_accuracy": 0.7045253440737724,
      "num_tokens": 11400835.0,
      "step": 357
    },
    {
      "entropy": 0.9630032256245613,
      "epoch": 0.032938280975957586,
      "grad_norm": 0.1826625019311905,
      "learning_rate": 9.89051430674395e-05,
      "loss": 0.9559,
      "mean_token_accuracy": 0.7342922613024712,
      "num_tokens": 11431769.0,
      "step": 358
    },
    {
      "entropy": 0.9532694295048714,
      "epoch": 0.03303028734739881,
      "grad_norm": 0.1662711501121521,
      "learning_rate": 9.890207624129788e-05,
      "loss": 0.9534,
      "mean_token_accuracy": 0.7346100136637688,
      "num_tokens": 11463296.0,
      "step": 359
    },
    {
      "entropy": 0.9254776574671268,
      "epoch": 0.033122293718840026,
      "grad_norm": 0.1689961701631546,
      "learning_rate": 9.889900941515627e-05,
      "loss": 0.9396,
      "mean_token_accuracy": 0.7428248673677444,
      "num_tokens": 11496064.0,
      "step": 360
    },
    {
      "entropy": 0.9055741056799889,
      "epoch": 0.03321430009028125,
      "grad_norm": 0.1550629734992981,
      "learning_rate": 9.889594258901463e-05,
      "loss": 0.915,
      "mean_token_accuracy": 0.7497055381536484,
      "num_tokens": 11527366.0,
      "step": 361
    },
    {
      "entropy": 1.0096909366548061,
      "epoch": 0.033306306461722474,
      "grad_norm": 0.16857384145259857,
      "learning_rate": 9.8892875762873e-05,
      "loss": 1.0191,
      "mean_token_accuracy": 0.7246541567146778,
      "num_tokens": 11559845.0,
      "step": 362
    },
    {
      "entropy": 0.9251548163592815,
      "epoch": 0.0333983128331637,
      "grad_norm": 0.1722024530172348,
      "learning_rate": 9.888980893673137e-05,
      "loss": 0.9505,
      "mean_token_accuracy": 0.7423433810472488,
      "num_tokens": 11592188.0,
      "step": 363
    },
    {
      "entropy": 1.0308907218277454,
      "epoch": 0.03349031920460492,
      "grad_norm": 0.17410510778427124,
      "learning_rate": 9.888674211058976e-05,
      "loss": 1.0403,
      "mean_token_accuracy": 0.7180306762456894,
      "num_tokens": 11623350.0,
      "step": 364
    },
    {
      "entropy": 1.007260225713253,
      "epoch": 0.033582325576046144,
      "grad_norm": 0.16690702736377716,
      "learning_rate": 9.888367528444813e-05,
      "loss": 1.009,
      "mean_token_accuracy": 0.7210812196135521,
      "num_tokens": 11654689.0,
      "step": 365
    },
    {
      "entropy": 1.061899933964014,
      "epoch": 0.03367433194748736,
      "grad_norm": 0.16727423667907715,
      "learning_rate": 9.88806084583065e-05,
      "loss": 1.0664,
      "mean_token_accuracy": 0.7090633101761341,
      "num_tokens": 11686702.0,
      "step": 366
    },
    {
      "entropy": 0.9429048523306847,
      "epoch": 0.033766338318928585,
      "grad_norm": 0.17015305161476135,
      "learning_rate": 9.887754163216487e-05,
      "loss": 0.9443,
      "mean_token_accuracy": 0.7376565635204315,
      "num_tokens": 11717811.0,
      "step": 367
    },
    {
      "entropy": 0.8862669691443443,
      "epoch": 0.03385834469036981,
      "grad_norm": 0.16411320865154266,
      "learning_rate": 9.887447480602325e-05,
      "loss": 0.8929,
      "mean_token_accuracy": 0.7531328350305557,
      "num_tokens": 11749909.0,
      "step": 368
    },
    {
      "entropy": 0.913277804851532,
      "epoch": 0.03395035106181103,
      "grad_norm": 0.16492368280887604,
      "learning_rate": 9.887140797988163e-05,
      "loss": 0.9184,
      "mean_token_accuracy": 0.7473931983113289,
      "num_tokens": 11782223.0,
      "step": 369
    },
    {
      "entropy": 1.0298655815422535,
      "epoch": 0.034042357433252256,
      "grad_norm": 0.16377604007720947,
      "learning_rate": 9.886834115374e-05,
      "loss": 1.0157,
      "mean_token_accuracy": 0.7220143675804138,
      "num_tokens": 11814336.0,
      "step": 370
    },
    {
      "entropy": 1.0704091377556324,
      "epoch": 0.03413436380469347,
      "grad_norm": 0.16085398197174072,
      "learning_rate": 9.886527432759838e-05,
      "loss": 1.0914,
      "mean_token_accuracy": 0.7112949825823307,
      "num_tokens": 11846901.0,
      "step": 371
    },
    {
      "entropy": 0.9985511340200901,
      "epoch": 0.034226370176134696,
      "grad_norm": 0.1590549647808075,
      "learning_rate": 9.886220750145675e-05,
      "loss": 0.9953,
      "mean_token_accuracy": 0.7257275655865669,
      "num_tokens": 11879614.0,
      "step": 372
    },
    {
      "entropy": 1.055351547896862,
      "epoch": 0.03431837654757592,
      "grad_norm": 0.16869869828224182,
      "learning_rate": 9.885914067531512e-05,
      "loss": 1.0589,
      "mean_token_accuracy": 0.7152610644698143,
      "num_tokens": 11910902.0,
      "step": 373
    },
    {
      "entropy": 0.8921605572104454,
      "epoch": 0.03441038291901714,
      "grad_norm": 0.16095079481601715,
      "learning_rate": 9.88560738491735e-05,
      "loss": 0.9155,
      "mean_token_accuracy": 0.748100932687521,
      "num_tokens": 11942929.0,
      "step": 374
    },
    {
      "entropy": 0.9579007774591446,
      "epoch": 0.03450238929045837,
      "grad_norm": 0.16929422318935394,
      "learning_rate": 9.885300702303188e-05,
      "loss": 0.9662,
      "mean_token_accuracy": 0.7306705750524998,
      "num_tokens": 11974875.0,
      "step": 375
    },
    {
      "entropy": 0.9048180021345615,
      "epoch": 0.034594395661899584,
      "grad_norm": 0.16311734914779663,
      "learning_rate": 9.884994019689025e-05,
      "loss": 0.9266,
      "mean_token_accuracy": 0.7411008439958096,
      "num_tokens": 12006861.0,
      "step": 376
    },
    {
      "entropy": 1.0420637503266335,
      "epoch": 0.03468640203334081,
      "grad_norm": 0.1754646599292755,
      "learning_rate": 9.884687337074861e-05,
      "loss": 1.0571,
      "mean_token_accuracy": 0.7111155167222023,
      "num_tokens": 12038885.0,
      "step": 377
    },
    {
      "entropy": 0.9592690393328667,
      "epoch": 0.03477840840478203,
      "grad_norm": 0.1727132797241211,
      "learning_rate": 9.884380654460698e-05,
      "loss": 0.9812,
      "mean_token_accuracy": 0.7319749407470226,
      "num_tokens": 12071132.0,
      "step": 378
    },
    {
      "entropy": 1.054789762943983,
      "epoch": 0.034870414776223255,
      "grad_norm": 0.1689300239086151,
      "learning_rate": 9.884073971846536e-05,
      "loss": 1.073,
      "mean_token_accuracy": 0.7101123034954071,
      "num_tokens": 12103591.0,
      "step": 379
    },
    {
      "entropy": 1.0424779579043388,
      "epoch": 0.03496242114766448,
      "grad_norm": 0.16431975364685059,
      "learning_rate": 9.883767289232375e-05,
      "loss": 1.0377,
      "mean_token_accuracy": 0.718727245926857,
      "num_tokens": 12135896.0,
      "step": 380
    },
    {
      "entropy": 1.1518799401819706,
      "epoch": 0.035054427519105695,
      "grad_norm": 0.1767517626285553,
      "learning_rate": 9.883460606618211e-05,
      "loss": 1.1547,
      "mean_token_accuracy": 0.6907789967954159,
      "num_tokens": 12168424.0,
      "step": 381
    },
    {
      "entropy": 1.0508774667978287,
      "epoch": 0.03514643389054692,
      "grad_norm": 0.1719999462366104,
      "learning_rate": 9.883153924004048e-05,
      "loss": 1.0452,
      "mean_token_accuracy": 0.7188280262053013,
      "num_tokens": 12199723.0,
      "step": 382
    },
    {
      "entropy": 1.0620715208351612,
      "epoch": 0.03523844026198814,
      "grad_norm": 0.17118346691131592,
      "learning_rate": 9.882847241389886e-05,
      "loss": 1.0438,
      "mean_token_accuracy": 0.7135345675051212,
      "num_tokens": 12231269.0,
      "step": 383
    },
    {
      "entropy": 0.9520346149802208,
      "epoch": 0.035330446633429366,
      "grad_norm": 0.15975673496723175,
      "learning_rate": 9.882540558775723e-05,
      "loss": 0.9222,
      "mean_token_accuracy": 0.7415171004831791,
      "num_tokens": 12262710.0,
      "step": 384
    },
    {
      "entropy": 1.0256397128105164,
      "epoch": 0.03542245300487059,
      "grad_norm": 0.1609017252922058,
      "learning_rate": 9.882233876161561e-05,
      "loss": 1.0294,
      "mean_token_accuracy": 0.720424871891737,
      "num_tokens": 12295019.0,
      "step": 385
    },
    {
      "entropy": 1.0355511009693146,
      "epoch": 0.035514459376311806,
      "grad_norm": 0.16495536267757416,
      "learning_rate": 9.881927193547398e-05,
      "loss": 1.0571,
      "mean_token_accuracy": 0.7152581363916397,
      "num_tokens": 12327088.0,
      "step": 386
    },
    {
      "entropy": 0.9235962964594364,
      "epoch": 0.03560646574775303,
      "grad_norm": 0.16323792934417725,
      "learning_rate": 9.881620510933236e-05,
      "loss": 0.9316,
      "mean_token_accuracy": 0.7438319884240627,
      "num_tokens": 12359154.0,
      "step": 387
    },
    {
      "entropy": 0.8861152045428753,
      "epoch": 0.035698472119194254,
      "grad_norm": 0.1591137945652008,
      "learning_rate": 9.881313828319073e-05,
      "loss": 0.92,
      "mean_token_accuracy": 0.7447265721857548,
      "num_tokens": 12391851.0,
      "step": 388
    },
    {
      "entropy": 0.9790819771587849,
      "epoch": 0.03579047849063548,
      "grad_norm": 0.17388558387756348,
      "learning_rate": 9.88100714570491e-05,
      "loss": 0.9991,
      "mean_token_accuracy": 0.7259571775794029,
      "num_tokens": 12424254.0,
      "step": 389
    },
    {
      "entropy": 0.9728363789618015,
      "epoch": 0.0358824848620767,
      "grad_norm": 0.16432583332061768,
      "learning_rate": 9.880700463090748e-05,
      "loss": 0.9872,
      "mean_token_accuracy": 0.7257611304521561,
      "num_tokens": 12456704.0,
      "step": 390
    },
    {
      "entropy": 1.0038979351520538,
      "epoch": 0.035974491233517925,
      "grad_norm": 0.16394533216953278,
      "learning_rate": 9.880393780476586e-05,
      "loss": 1.0351,
      "mean_token_accuracy": 0.722713477909565,
      "num_tokens": 12488357.0,
      "step": 391
    },
    {
      "entropy": 1.0411340072751045,
      "epoch": 0.03606649760495914,
      "grad_norm": 0.17753370106220245,
      "learning_rate": 9.880087097862423e-05,
      "loss": 1.0566,
      "mean_token_accuracy": 0.7073158174753189,
      "num_tokens": 12520055.0,
      "step": 392
    },
    {
      "entropy": 1.0669416636228561,
      "epoch": 0.036158503976400365,
      "grad_norm": 0.17230726778507233,
      "learning_rate": 9.87978041524826e-05,
      "loss": 1.0845,
      "mean_token_accuracy": 0.7072788029909134,
      "num_tokens": 12551800.0,
      "step": 393
    },
    {
      "entropy": 1.1152261942625046,
      "epoch": 0.03625051034784159,
      "grad_norm": 0.17747318744659424,
      "learning_rate": 9.879473732634096e-05,
      "loss": 1.1358,
      "mean_token_accuracy": 0.6900845691561699,
      "num_tokens": 12584321.0,
      "step": 394
    },
    {
      "entropy": 0.983378030359745,
      "epoch": 0.03634251671928281,
      "grad_norm": 0.16044682264328003,
      "learning_rate": 9.879167050019934e-05,
      "loss": 0.9855,
      "mean_token_accuracy": 0.7247141003608704,
      "num_tokens": 12615438.0,
      "step": 395
    },
    {
      "entropy": 0.9914127290248871,
      "epoch": 0.036434523090724036,
      "grad_norm": 0.1740035116672516,
      "learning_rate": 9.878860367405773e-05,
      "loss": 0.9869,
      "mean_token_accuracy": 0.7332474254071712,
      "num_tokens": 12647357.0,
      "step": 396
    },
    {
      "entropy": 1.0712041072547436,
      "epoch": 0.03652652946216525,
      "grad_norm": 0.17429697513580322,
      "learning_rate": 9.87855368479161e-05,
      "loss": 1.0497,
      "mean_token_accuracy": 0.7074903659522533,
      "num_tokens": 12679607.0,
      "step": 397
    },
    {
      "entropy": 1.044678132981062,
      "epoch": 0.036618535833606476,
      "grad_norm": 0.17273195087909698,
      "learning_rate": 9.878247002177448e-05,
      "loss": 1.0424,
      "mean_token_accuracy": 0.7187317945063114,
      "num_tokens": 12711499.0,
      "step": 398
    },
    {
      "entropy": 0.9989720210433006,
      "epoch": 0.0367105422050477,
      "grad_norm": 0.16758893430233002,
      "learning_rate": 9.877940319563284e-05,
      "loss": 0.988,
      "mean_token_accuracy": 0.7266386933624744,
      "num_tokens": 12742815.0,
      "step": 399
    },
    {
      "entropy": 1.083141215145588,
      "epoch": 0.036802548576488923,
      "grad_norm": 0.17270106077194214,
      "learning_rate": 9.877633636949121e-05,
      "loss": 1.0855,
      "mean_token_accuracy": 0.7046500369906425,
      "num_tokens": 12774717.0,
      "step": 400
    },
    {
      "entropy": 0.9950155466794968,
      "epoch": 0.03689455494793015,
      "grad_norm": 0.16009636223316193,
      "learning_rate": 9.877326954334959e-05,
      "loss": 0.9901,
      "mean_token_accuracy": 0.7269642017781734,
      "num_tokens": 12806646.0,
      "step": 401
    },
    {
      "entropy": 1.0237273834645748,
      "epoch": 0.036986561319371364,
      "grad_norm": 0.16479110717773438,
      "learning_rate": 9.877020271720797e-05,
      "loss": 1.0326,
      "mean_token_accuracy": 0.7236855439841747,
      "num_tokens": 12838327.0,
      "step": 402
    },
    {
      "entropy": 1.007001630961895,
      "epoch": 0.03707856769081259,
      "grad_norm": 0.16550607979297638,
      "learning_rate": 9.876713589106634e-05,
      "loss": 1.0116,
      "mean_token_accuracy": 0.7222214266657829,
      "num_tokens": 12869628.0,
      "step": 403
    },
    {
      "entropy": 0.8934124559164047,
      "epoch": 0.03717057406225381,
      "grad_norm": 0.17071421444416046,
      "learning_rate": 9.876406906492471e-05,
      "loss": 0.8859,
      "mean_token_accuracy": 0.7562615610659122,
      "num_tokens": 12901987.0,
      "step": 404
    },
    {
      "entropy": 1.0114603377878666,
      "epoch": 0.037262580433695035,
      "grad_norm": 0.17034700512886047,
      "learning_rate": 9.876100223878309e-05,
      "loss": 1.0192,
      "mean_token_accuracy": 0.7229459024965763,
      "num_tokens": 12933987.0,
      "step": 405
    },
    {
      "entropy": 1.0978050529956818,
      "epoch": 0.03735458680513626,
      "grad_norm": 0.17636926472187042,
      "learning_rate": 9.875793541264147e-05,
      "loss": 1.112,
      "mean_token_accuracy": 0.698073074221611,
      "num_tokens": 12965520.0,
      "step": 406
    },
    {
      "entropy": 1.0788080766797066,
      "epoch": 0.037446593176577475,
      "grad_norm": 0.17143350839614868,
      "learning_rate": 9.875486858649984e-05,
      "loss": 1.0943,
      "mean_token_accuracy": 0.7021720893681049,
      "num_tokens": 12997010.0,
      "step": 407
    },
    {
      "entropy": 1.0392850413918495,
      "epoch": 0.0375385995480187,
      "grad_norm": 0.17000417411327362,
      "learning_rate": 9.875180176035821e-05,
      "loss": 1.0581,
      "mean_token_accuracy": 0.7103569321334362,
      "num_tokens": 13029326.0,
      "step": 408
    },
    {
      "entropy": 1.004322987049818,
      "epoch": 0.03763060591945992,
      "grad_norm": 0.166298970580101,
      "learning_rate": 9.874873493421658e-05,
      "loss": 1.0134,
      "mean_token_accuracy": 0.7229413129389286,
      "num_tokens": 13060953.0,
      "step": 409
    },
    {
      "entropy": 0.9710695706307888,
      "epoch": 0.037722612290901146,
      "grad_norm": 0.1652998924255371,
      "learning_rate": 9.874566810807496e-05,
      "loss": 0.989,
      "mean_token_accuracy": 0.7285130433738232,
      "num_tokens": 13092530.0,
      "step": 410
    },
    {
      "entropy": 0.9180749878287315,
      "epoch": 0.03781461866234237,
      "grad_norm": 0.17449809610843658,
      "learning_rate": 9.874260128193334e-05,
      "loss": 0.9364,
      "mean_token_accuracy": 0.7407023012638092,
      "num_tokens": 13124442.0,
      "step": 411
    },
    {
      "entropy": 1.11996004357934,
      "epoch": 0.037906625033783586,
      "grad_norm": 0.17418552935123444,
      "learning_rate": 9.873953445579171e-05,
      "loss": 1.1248,
      "mean_token_accuracy": 0.6937156431376934,
      "num_tokens": 13156173.0,
      "step": 412
    },
    {
      "entropy": 1.053414773195982,
      "epoch": 0.03799863140522481,
      "grad_norm": 0.16683727502822876,
      "learning_rate": 9.873646762965007e-05,
      "loss": 1.0614,
      "mean_token_accuracy": 0.7129155397415161,
      "num_tokens": 13188083.0,
      "step": 413
    },
    {
      "entropy": 1.0056521221995354,
      "epoch": 0.038090637776666034,
      "grad_norm": 0.16785532236099243,
      "learning_rate": 9.873340080350846e-05,
      "loss": 1.0149,
      "mean_token_accuracy": 0.7221137546002865,
      "num_tokens": 13220595.0,
      "step": 414
    },
    {
      "entropy": 1.0297675020992756,
      "epoch": 0.03818264414810726,
      "grad_norm": 0.17071309685707092,
      "learning_rate": 9.873033397736682e-05,
      "loss": 1.0386,
      "mean_token_accuracy": 0.7179465256631374,
      "num_tokens": 13252387.0,
      "step": 415
    },
    {
      "entropy": 0.9606161341071129,
      "epoch": 0.03827465051954848,
      "grad_norm": 0.16314919292926788,
      "learning_rate": 9.87272671512252e-05,
      "loss": 0.9785,
      "mean_token_accuracy": 0.7281481437385082,
      "num_tokens": 13283850.0,
      "step": 416
    },
    {
      "entropy": 0.9838493391871452,
      "epoch": 0.0383666568909897,
      "grad_norm": 0.16217157244682312,
      "learning_rate": 9.872420032508357e-05,
      "loss": 0.9807,
      "mean_token_accuracy": 0.7286479398608208,
      "num_tokens": 13315672.0,
      "step": 417
    },
    {
      "entropy": 0.9159945547580719,
      "epoch": 0.03845866326243092,
      "grad_norm": 0.15739157795906067,
      "learning_rate": 9.872113349894196e-05,
      "loss": 0.8987,
      "mean_token_accuracy": 0.7468747198581696,
      "num_tokens": 13347850.0,
      "step": 418
    },
    {
      "entropy": 1.0316320173442364,
      "epoch": 0.038550669633872145,
      "grad_norm": 0.17621657252311707,
      "learning_rate": 9.871806667280032e-05,
      "loss": 1.0524,
      "mean_token_accuracy": 0.7118074521422386,
      "num_tokens": 13379244.0,
      "step": 419
    },
    {
      "entropy": 0.9849085547029972,
      "epoch": 0.03864267600531337,
      "grad_norm": 0.16847988963127136,
      "learning_rate": 9.871499984665869e-05,
      "loss": 1.006,
      "mean_token_accuracy": 0.7315991632640362,
      "num_tokens": 13411613.0,
      "step": 420
    },
    {
      "entropy": 1.0356707982718945,
      "epoch": 0.03873468237675459,
      "grad_norm": 0.16560828685760498,
      "learning_rate": 9.871193302051707e-05,
      "loss": 1.0535,
      "mean_token_accuracy": 0.7156078405678272,
      "num_tokens": 13443656.0,
      "step": 421
    },
    {
      "entropy": 0.9872327260673046,
      "epoch": 0.038826688748195816,
      "grad_norm": 0.17459335923194885,
      "learning_rate": 9.870886619437545e-05,
      "loss": 1.0031,
      "mean_token_accuracy": 0.7265473678708076,
      "num_tokens": 13475255.0,
      "step": 422
    },
    {
      "entropy": 1.0325854793190956,
      "epoch": 0.03891869511963703,
      "grad_norm": 0.16438168287277222,
      "learning_rate": 9.870579936823382e-05,
      "loss": 1.0241,
      "mean_token_accuracy": 0.7219798676669598,
      "num_tokens": 13507733.0,
      "step": 423
    },
    {
      "entropy": 1.006734848022461,
      "epoch": 0.039010701491078256,
      "grad_norm": 0.1691765934228897,
      "learning_rate": 9.870273254209219e-05,
      "loss": 0.9863,
      "mean_token_accuracy": 0.7285861372947693,
      "num_tokens": 13539348.0,
      "step": 424
    },
    {
      "entropy": 1.002249352633953,
      "epoch": 0.03910270786251948,
      "grad_norm": 0.16368570923805237,
      "learning_rate": 9.869966571595056e-05,
      "loss": 0.9871,
      "mean_token_accuracy": 0.7271999455988407,
      "num_tokens": 13571405.0,
      "step": 425
    },
    {
      "entropy": 1.0048393718898296,
      "epoch": 0.039194714233960704,
      "grad_norm": 0.27615973353385925,
      "learning_rate": 9.869659888980894e-05,
      "loss": 1.0087,
      "mean_token_accuracy": 0.7247118502855301,
      "num_tokens": 13602797.0,
      "step": 426
    },
    {
      "entropy": 0.9325462207198143,
      "epoch": 0.03928672060540193,
      "grad_norm": 0.16752378642559052,
      "learning_rate": 9.869353206366732e-05,
      "loss": 0.9377,
      "mean_token_accuracy": 0.745978232473135,
      "num_tokens": 13634546.0,
      "step": 427
    },
    {
      "entropy": 0.9929099604487419,
      "epoch": 0.039378726976843144,
      "grad_norm": 0.17789685726165771,
      "learning_rate": 9.869046523752569e-05,
      "loss": 1.0047,
      "mean_token_accuracy": 0.7219155803322792,
      "num_tokens": 13666314.0,
      "step": 428
    },
    {
      "entropy": 0.9159045815467834,
      "epoch": 0.03947073334828437,
      "grad_norm": 0.1598764806985855,
      "learning_rate": 9.868739841138407e-05,
      "loss": 0.9291,
      "mean_token_accuracy": 0.7393090985715389,
      "num_tokens": 13698160.0,
      "step": 429
    },
    {
      "entropy": 1.0404502265155315,
      "epoch": 0.03956273971972559,
      "grad_norm": 0.1731593757867813,
      "learning_rate": 9.868433158524244e-05,
      "loss": 1.0645,
      "mean_token_accuracy": 0.7102674506604671,
      "num_tokens": 13730033.0,
      "step": 430
    },
    {
      "entropy": 1.114392176270485,
      "epoch": 0.039654746091166815,
      "grad_norm": 0.17988698184490204,
      "learning_rate": 9.86812647591008e-05,
      "loss": 1.1273,
      "mean_token_accuracy": 0.7005251944065094,
      "num_tokens": 13761711.0,
      "step": 431
    },
    {
      "entropy": 0.9822462163865566,
      "epoch": 0.03974675246260804,
      "grad_norm": 0.1688549816608429,
      "learning_rate": 9.867819793295919e-05,
      "loss": 0.9881,
      "mean_token_accuracy": 0.7253421731293201,
      "num_tokens": 13793848.0,
      "step": 432
    },
    {
      "entropy": 0.9594956189393997,
      "epoch": 0.039838758834049255,
      "grad_norm": 0.162195086479187,
      "learning_rate": 9.867513110681757e-05,
      "loss": 0.9381,
      "mean_token_accuracy": 0.7382728308439255,
      "num_tokens": 13825858.0,
      "step": 433
    },
    {
      "entropy": 0.9629478938877583,
      "epoch": 0.03993076520549048,
      "grad_norm": 0.1678634136915207,
      "learning_rate": 9.867206428067594e-05,
      "loss": 0.9364,
      "mean_token_accuracy": 0.7352337725460529,
      "num_tokens": 13856521.0,
      "step": 434
    },
    {
      "entropy": 0.9395595081150532,
      "epoch": 0.0400227715769317,
      "grad_norm": 0.1596844494342804,
      "learning_rate": 9.86689974545343e-05,
      "loss": 0.9413,
      "mean_token_accuracy": 0.7388866692781448,
      "num_tokens": 13888613.0,
      "step": 435
    },
    {
      "entropy": 1.0746241621673107,
      "epoch": 0.040114777948372926,
      "grad_norm": 0.17097659409046173,
      "learning_rate": 9.866593062839267e-05,
      "loss": 1.0859,
      "mean_token_accuracy": 0.6997541300952435,
      "num_tokens": 13920209.0,
      "step": 436
    },
    {
      "entropy": 0.9571687839925289,
      "epoch": 0.04020678431981415,
      "grad_norm": 0.16474372148513794,
      "learning_rate": 9.866286380225107e-05,
      "loss": 0.9739,
      "mean_token_accuracy": 0.7354234457015991,
      "num_tokens": 13952483.0,
      "step": 437
    },
    {
      "entropy": 0.8903435505926609,
      "epoch": 0.040298790691255366,
      "grad_norm": 0.162520632147789,
      "learning_rate": 9.865979697610943e-05,
      "loss": 0.8897,
      "mean_token_accuracy": 0.7485632933676243,
      "num_tokens": 13983339.0,
      "step": 438
    },
    {
      "entropy": 1.0988422967493534,
      "epoch": 0.04039079706269659,
      "grad_norm": 0.1772916615009308,
      "learning_rate": 9.86567301499678e-05,
      "loss": 1.1463,
      "mean_token_accuracy": 0.6962193101644516,
      "num_tokens": 14015302.0,
      "step": 439
    },
    {
      "entropy": 1.0508275460451841,
      "epoch": 0.040482803434137814,
      "grad_norm": 0.16714204847812653,
      "learning_rate": 9.865366332382617e-05,
      "loss": 1.0611,
      "mean_token_accuracy": 0.7144768461585045,
      "num_tokens": 14047693.0,
      "step": 440
    },
    {
      "entropy": 1.053979579359293,
      "epoch": 0.04057480980557904,
      "grad_norm": 0.16524246335029602,
      "learning_rate": 9.865059649768455e-05,
      "loss": 1.0647,
      "mean_token_accuracy": 0.707475584000349,
      "num_tokens": 14079607.0,
      "step": 441
    },
    {
      "entropy": 0.9862862415611744,
      "epoch": 0.04066681617702026,
      "grad_norm": 0.16560500860214233,
      "learning_rate": 9.864752967154293e-05,
      "loss": 1.0034,
      "mean_token_accuracy": 0.7281387448310852,
      "num_tokens": 14112375.0,
      "step": 442
    },
    {
      "entropy": 1.0791384130716324,
      "epoch": 0.04075882254846148,
      "grad_norm": 0.17855489253997803,
      "learning_rate": 9.86444628454013e-05,
      "loss": 1.0705,
      "mean_token_accuracy": 0.7084731496870518,
      "num_tokens": 14143777.0,
      "step": 443
    },
    {
      "entropy": 1.0406125336885452,
      "epoch": 0.0408508289199027,
      "grad_norm": 0.1689472496509552,
      "learning_rate": 9.864139601925967e-05,
      "loss": 1.0206,
      "mean_token_accuracy": 0.7244569584727287,
      "num_tokens": 14176256.0,
      "step": 444
    },
    {
      "entropy": 1.0324552282691002,
      "epoch": 0.040942835291343925,
      "grad_norm": 0.16691230237483978,
      "learning_rate": 9.863832919311805e-05,
      "loss": 0.9982,
      "mean_token_accuracy": 0.7252889536321163,
      "num_tokens": 14208118.0,
      "step": 445
    },
    {
      "entropy": 0.9507054463028908,
      "epoch": 0.04103484166278515,
      "grad_norm": 0.17725025117397308,
      "learning_rate": 9.863526236697642e-05,
      "loss": 0.9394,
      "mean_token_accuracy": 0.7385678701102734,
      "num_tokens": 14239863.0,
      "step": 446
    },
    {
      "entropy": 0.9064260274171829,
      "epoch": 0.04112684803422637,
      "grad_norm": 0.16163617372512817,
      "learning_rate": 9.86321955408348e-05,
      "loss": 0.9039,
      "mean_token_accuracy": 0.7509748563170433,
      "num_tokens": 14272369.0,
      "step": 447
    },
    {
      "entropy": 1.1047485694289207,
      "epoch": 0.04121885440566759,
      "grad_norm": 0.17077849805355072,
      "learning_rate": 9.862912871469317e-05,
      "loss": 1.1291,
      "mean_token_accuracy": 0.6987112313508987,
      "num_tokens": 14304949.0,
      "step": 448
    },
    {
      "entropy": 0.9774901755154133,
      "epoch": 0.04131086077710881,
      "grad_norm": 0.1709381490945816,
      "learning_rate": 9.862606188855155e-05,
      "loss": 1.0215,
      "mean_token_accuracy": 0.7202001512050629,
      "num_tokens": 14337132.0,
      "step": 449
    },
    {
      "entropy": 0.9640143178403378,
      "epoch": 0.041402867148550036,
      "grad_norm": 0.1706177443265915,
      "learning_rate": 9.862299506240992e-05,
      "loss": 1.0043,
      "mean_token_accuracy": 0.7218887060880661,
      "num_tokens": 14369274.0,
      "step": 450
    },
    {
      "entropy": 0.9488422982394695,
      "epoch": 0.04149487351999126,
      "grad_norm": 0.1612362116575241,
      "learning_rate": 9.861992823626828e-05,
      "loss": 0.9503,
      "mean_token_accuracy": 0.7361940965056419,
      "num_tokens": 14401497.0,
      "step": 451
    },
    {
      "entropy": 1.0639336369931698,
      "epoch": 0.041586879891432484,
      "grad_norm": 0.17198891937732697,
      "learning_rate": 9.861686141012667e-05,
      "loss": 1.0742,
      "mean_token_accuracy": 0.7064671106636524,
      "num_tokens": 14433160.0,
      "step": 452
    },
    {
      "entropy": 0.9131907224655151,
      "epoch": 0.04167888626287371,
      "grad_norm": 0.15841710567474365,
      "learning_rate": 9.861379458398505e-05,
      "loss": 0.9032,
      "mean_token_accuracy": 0.7455384209752083,
      "num_tokens": 14465155.0,
      "step": 453
    },
    {
      "entropy": 1.005853183567524,
      "epoch": 0.041770892634314924,
      "grad_norm": 0.1667892187833786,
      "learning_rate": 9.861072775784342e-05,
      "loss": 1.0081,
      "mean_token_accuracy": 0.7247069738805294,
      "num_tokens": 14496056.0,
      "step": 454
    },
    {
      "entropy": 0.9747581593692303,
      "epoch": 0.04186289900575615,
      "grad_norm": 0.1641293466091156,
      "learning_rate": 9.860766093170178e-05,
      "loss": 0.9754,
      "mean_token_accuracy": 0.7309782467782497,
      "num_tokens": 14528432.0,
      "step": 455
    },
    {
      "entropy": 0.9135883450508118,
      "epoch": 0.04195490537719737,
      "grad_norm": 0.1626873016357422,
      "learning_rate": 9.860459410556015e-05,
      "loss": 0.9142,
      "mean_token_accuracy": 0.7430468536913395,
      "num_tokens": 14560104.0,
      "step": 456
    },
    {
      "entropy": 1.0127588957548141,
      "epoch": 0.042046911748638595,
      "grad_norm": 0.17270705103874207,
      "learning_rate": 9.860152727941853e-05,
      "loss": 1.021,
      "mean_token_accuracy": 0.7224766612052917,
      "num_tokens": 14591753.0,
      "step": 457
    },
    {
      "entropy": 0.9827232696115971,
      "epoch": 0.04213891812007982,
      "grad_norm": 0.18595170974731445,
      "learning_rate": 9.859846045327691e-05,
      "loss": 0.9864,
      "mean_token_accuracy": 0.729983314871788,
      "num_tokens": 14623459.0,
      "step": 458
    },
    {
      "entropy": 1.028811253607273,
      "epoch": 0.042230924491521035,
      "grad_norm": 0.1726570576429367,
      "learning_rate": 9.859539362713528e-05,
      "loss": 1.0439,
      "mean_token_accuracy": 0.7178481444716454,
      "num_tokens": 14655035.0,
      "step": 459
    },
    {
      "entropy": 1.0910115167498589,
      "epoch": 0.04232293086296226,
      "grad_norm": 0.17036913335323334,
      "learning_rate": 9.859232680099366e-05,
      "loss": 1.101,
      "mean_token_accuracy": 0.7012305520474911,
      "num_tokens": 14686749.0,
      "step": 460
    },
    {
      "entropy": 1.0199798494577408,
      "epoch": 0.04241493723440348,
      "grad_norm": 0.16065360605716705,
      "learning_rate": 9.858925997485203e-05,
      "loss": 1.0138,
      "mean_token_accuracy": 0.7204526774585247,
      "num_tokens": 14718454.0,
      "step": 461
    },
    {
      "entropy": 0.9938953183591366,
      "epoch": 0.042506943605844706,
      "grad_norm": 0.16792386770248413,
      "learning_rate": 9.85861931487104e-05,
      "loss": 0.9841,
      "mean_token_accuracy": 0.7303761541843414,
      "num_tokens": 14749963.0,
      "step": 462
    },
    {
      "entropy": 1.0054540112614632,
      "epoch": 0.04259894997728593,
      "grad_norm": 0.16829322278499603,
      "learning_rate": 9.858312632256878e-05,
      "loss": 1.0004,
      "mean_token_accuracy": 0.7285681366920471,
      "num_tokens": 14782141.0,
      "step": 463
    },
    {
      "entropy": 1.0681659132242203,
      "epoch": 0.042690956348727147,
      "grad_norm": 0.17712776362895966,
      "learning_rate": 9.858005949642716e-05,
      "loss": 1.0851,
      "mean_token_accuracy": 0.7081322111189365,
      "num_tokens": 14813717.0,
      "step": 464
    },
    {
      "entropy": 1.0289651975035667,
      "epoch": 0.04278296272016837,
      "grad_norm": 0.16890780627727509,
      "learning_rate": 9.857699267028553e-05,
      "loss": 1.0278,
      "mean_token_accuracy": 0.7198490500450134,
      "num_tokens": 14845695.0,
      "step": 465
    },
    {
      "entropy": 0.9659386463463306,
      "epoch": 0.042874969091609594,
      "grad_norm": 0.1655670404434204,
      "learning_rate": 9.85739258441439e-05,
      "loss": 0.948,
      "mean_token_accuracy": 0.7323296628892422,
      "num_tokens": 14876542.0,
      "step": 466
    },
    {
      "entropy": 1.082155641168356,
      "epoch": 0.04296697546305082,
      "grad_norm": 0.1711169332265854,
      "learning_rate": 9.857085901800227e-05,
      "loss": 1.0844,
      "mean_token_accuracy": 0.7022744417190552,
      "num_tokens": 14908336.0,
      "step": 467
    },
    {
      "entropy": 1.0541718192398548,
      "epoch": 0.04305898183449204,
      "grad_norm": 0.16499820351600647,
      "learning_rate": 9.856779219186065e-05,
      "loss": 1.0626,
      "mean_token_accuracy": 0.7141343988478184,
      "num_tokens": 14941049.0,
      "step": 468
    },
    {
      "entropy": 0.990358866751194,
      "epoch": 0.04315098820593326,
      "grad_norm": 0.1699855923652649,
      "learning_rate": 9.856472536571903e-05,
      "loss": 0.9717,
      "mean_token_accuracy": 0.7303466461598873,
      "num_tokens": 14971931.0,
      "step": 469
    },
    {
      "entropy": 1.0079026594758034,
      "epoch": 0.04324299457737448,
      "grad_norm": 0.18114951252937317,
      "learning_rate": 9.85616585395774e-05,
      "loss": 1.0553,
      "mean_token_accuracy": 0.7136130630970001,
      "num_tokens": 15003359.0,
      "step": 470
    },
    {
      "entropy": 1.044851902872324,
      "epoch": 0.043335000948815705,
      "grad_norm": 0.179159477353096,
      "learning_rate": 9.855859171343576e-05,
      "loss": 1.0798,
      "mean_token_accuracy": 0.709804143756628,
      "num_tokens": 15035784.0,
      "step": 471
    },
    {
      "entropy": 0.8731819577515125,
      "epoch": 0.04342700732025693,
      "grad_norm": 0.1596076339483261,
      "learning_rate": 9.855552488729415e-05,
      "loss": 0.8797,
      "mean_token_accuracy": 0.7525278553366661,
      "num_tokens": 15067234.0,
      "step": 472
    },
    {
      "entropy": 1.005760069936514,
      "epoch": 0.04351901369169815,
      "grad_norm": 0.16880400478839874,
      "learning_rate": 9.855245806115251e-05,
      "loss": 1.0111,
      "mean_token_accuracy": 0.7245165891945362,
      "num_tokens": 15099527.0,
      "step": 473
    },
    {
      "entropy": 1.0048915781080723,
      "epoch": 0.04361102006313937,
      "grad_norm": 0.16019918024539948,
      "learning_rate": 9.85493912350109e-05,
      "loss": 1.0068,
      "mean_token_accuracy": 0.7270114235579967,
      "num_tokens": 15131948.0,
      "step": 474
    },
    {
      "entropy": 1.0810152143239975,
      "epoch": 0.04370302643458059,
      "grad_norm": 0.1750817745923996,
      "learning_rate": 9.854632440886926e-05,
      "loss": 1.0713,
      "mean_token_accuracy": 0.7154526896774769,
      "num_tokens": 15163943.0,
      "step": 475
    },
    {
      "entropy": 0.8974347617477179,
      "epoch": 0.043795032806021816,
      "grad_norm": 0.14788956940174103,
      "learning_rate": 9.854325758272764e-05,
      "loss": 0.8466,
      "mean_token_accuracy": 0.7642371952533722,
      "num_tokens": 15195645.0,
      "step": 476
    },
    {
      "entropy": 0.9626926407217979,
      "epoch": 0.04388703917746304,
      "grad_norm": 0.16394424438476562,
      "learning_rate": 9.854019075658601e-05,
      "loss": 0.9659,
      "mean_token_accuracy": 0.735569316893816,
      "num_tokens": 15228042.0,
      "step": 477
    },
    {
      "entropy": 1.0828361846506596,
      "epoch": 0.043979045548904264,
      "grad_norm": 0.18153053522109985,
      "learning_rate": 9.853712393044438e-05,
      "loss": 1.1257,
      "mean_token_accuracy": 0.6984194666147232,
      "num_tokens": 15259780.0,
      "step": 478
    },
    {
      "entropy": 0.990173913538456,
      "epoch": 0.04407105192034549,
      "grad_norm": 0.16425317525863647,
      "learning_rate": 9.853405710430276e-05,
      "loss": 0.9918,
      "mean_token_accuracy": 0.7264580465853214,
      "num_tokens": 15292094.0,
      "step": 479
    },
    {
      "entropy": 0.9448204785585403,
      "epoch": 0.044163058291786704,
      "grad_norm": 0.16387133300304413,
      "learning_rate": 9.853099027816114e-05,
      "loss": 0.9468,
      "mean_token_accuracy": 0.7394870258867741,
      "num_tokens": 15323733.0,
      "step": 480
    },
    {
      "entropy": 1.0291991084814072,
      "epoch": 0.04425506466322793,
      "grad_norm": 0.1707683503627777,
      "learning_rate": 9.852792345201951e-05,
      "loss": 1.0745,
      "mean_token_accuracy": 0.7095636986196041,
      "num_tokens": 15356438.0,
      "step": 481
    },
    {
      "entropy": 1.0038556605577469,
      "epoch": 0.04434707103466915,
      "grad_norm": 0.1699240803718567,
      "learning_rate": 9.852485662587788e-05,
      "loss": 1.0219,
      "mean_token_accuracy": 0.7195817418396473,
      "num_tokens": 15387811.0,
      "step": 482
    },
    {
      "entropy": 1.0137430615723133,
      "epoch": 0.044439077406110375,
      "grad_norm": 0.17154064774513245,
      "learning_rate": 9.852178979973625e-05,
      "loss": 1.0437,
      "mean_token_accuracy": 0.7150721289217472,
      "num_tokens": 15419270.0,
      "step": 483
    },
    {
      "entropy": 0.8973237853497267,
      "epoch": 0.0445310837775516,
      "grad_norm": 0.1608092188835144,
      "learning_rate": 9.851872297359464e-05,
      "loss": 0.8974,
      "mean_token_accuracy": 0.7489767372608185,
      "num_tokens": 15451102.0,
      "step": 484
    },
    {
      "entropy": 0.9676474370062351,
      "epoch": 0.044623090148992815,
      "grad_norm": 0.1591765433549881,
      "learning_rate": 9.851565614745301e-05,
      "loss": 0.9604,
      "mean_token_accuracy": 0.738324161618948,
      "num_tokens": 15483530.0,
      "step": 485
    },
    {
      "entropy": 0.9637531265616417,
      "epoch": 0.04471509652043404,
      "grad_norm": 0.15718592703342438,
      "learning_rate": 9.851258932131138e-05,
      "loss": 0.9645,
      "mean_token_accuracy": 0.7379612438380718,
      "num_tokens": 15515934.0,
      "step": 486
    },
    {
      "entropy": 1.046225119382143,
      "epoch": 0.04480710289187526,
      "grad_norm": 0.1642787605524063,
      "learning_rate": 9.850952249516974e-05,
      "loss": 1.0441,
      "mean_token_accuracy": 0.7204109765589237,
      "num_tokens": 15548279.0,
      "step": 487
    },
    {
      "entropy": 1.160289503633976,
      "epoch": 0.044899109263316486,
      "grad_norm": 0.18160724639892578,
      "learning_rate": 9.850645566902813e-05,
      "loss": 1.1412,
      "mean_token_accuracy": 0.6901385225355625,
      "num_tokens": 15580358.0,
      "step": 488
    },
    {
      "entropy": 0.9516584761440754,
      "epoch": 0.04499111563475771,
      "grad_norm": 0.16449078917503357,
      "learning_rate": 9.850338884288651e-05,
      "loss": 0.9357,
      "mean_token_accuracy": 0.7386973090469837,
      "num_tokens": 15612380.0,
      "step": 489
    },
    {
      "entropy": 1.0169189162552357,
      "epoch": 0.04508312200619893,
      "grad_norm": 2.589756965637207,
      "learning_rate": 9.850032201674488e-05,
      "loss": 1.0007,
      "mean_token_accuracy": 0.7259696684777737,
      "num_tokens": 15644546.0,
      "step": 490
    },
    {
      "entropy": 0.9180504865944386,
      "epoch": 0.04517512837764015,
      "grad_norm": 0.1614355593919754,
      "learning_rate": 9.849725519060326e-05,
      "loss": 0.9257,
      "mean_token_accuracy": 0.7417123541235924,
      "num_tokens": 15677072.0,
      "step": 491
    },
    {
      "entropy": 1.02853412181139,
      "epoch": 0.045267134749081374,
      "grad_norm": 0.32491713762283325,
      "learning_rate": 9.849418836446162e-05,
      "loss": 1.0385,
      "mean_token_accuracy": 0.7164977677166462,
      "num_tokens": 15708767.0,
      "step": 492
    },
    {
      "entropy": 1.0291013531386852,
      "epoch": 0.0453591411205226,
      "grad_norm": 0.17130427062511444,
      "learning_rate": 9.849112153831999e-05,
      "loss": 1.0284,
      "mean_token_accuracy": 0.7171813435852528,
      "num_tokens": 15740996.0,
      "step": 493
    },
    {
      "entropy": 0.9812391102313995,
      "epoch": 0.04545114749196382,
      "grad_norm": 0.16899001598358154,
      "learning_rate": 9.848805471217837e-05,
      "loss": 0.9823,
      "mean_token_accuracy": 0.7301101200282574,
      "num_tokens": 15773315.0,
      "step": 494
    },
    {
      "entropy": 0.9653379768133163,
      "epoch": 0.04554315386340504,
      "grad_norm": 0.1684325486421585,
      "learning_rate": 9.848498788603676e-05,
      "loss": 0.9779,
      "mean_token_accuracy": 0.72777896001935,
      "num_tokens": 15805366.0,
      "step": 495
    },
    {
      "entropy": 0.9256263040006161,
      "epoch": 0.04563516023484626,
      "grad_norm": 0.17400746047496796,
      "learning_rate": 9.848192105989512e-05,
      "loss": 0.9318,
      "mean_token_accuracy": 0.7422780506312847,
      "num_tokens": 15836787.0,
      "step": 496
    },
    {
      "entropy": 0.9993999637663364,
      "epoch": 0.045727166606287485,
      "grad_norm": 0.17800436913967133,
      "learning_rate": 9.847885423375349e-05,
      "loss": 1.0107,
      "mean_token_accuracy": 0.7217923253774643,
      "num_tokens": 15868379.0,
      "step": 497
    },
    {
      "entropy": 1.0723948143422604,
      "epoch": 0.04581917297772871,
      "grad_norm": 0.17987535893917084,
      "learning_rate": 9.847578740761186e-05,
      "loss": 1.1054,
      "mean_token_accuracy": 0.7003863900899887,
      "num_tokens": 15899644.0,
      "step": 498
    },
    {
      "entropy": 0.9887304380536079,
      "epoch": 0.04591117934916993,
      "grad_norm": 0.17515528202056885,
      "learning_rate": 9.847272058147024e-05,
      "loss": 1.0131,
      "mean_token_accuracy": 0.7256871573626995,
      "num_tokens": 15930404.0,
      "step": 499
    },
    {
      "entropy": 1.0227028280496597,
      "epoch": 0.04600318572061115,
      "grad_norm": 0.17213261127471924,
      "learning_rate": 9.846965375532862e-05,
      "loss": 1.0259,
      "mean_token_accuracy": 0.7217556983232498,
      "num_tokens": 15963017.0,
      "step": 500
    },
    {
      "entropy": 0.9503715112805367,
      "epoch": 0.04609519209205237,
      "grad_norm": 0.16630369424819946,
      "learning_rate": 9.846658692918699e-05,
      "loss": 0.9486,
      "mean_token_accuracy": 0.738661877810955,
      "num_tokens": 15994718.0,
      "step": 501
    },
    {
      "entropy": 0.9515565820038319,
      "epoch": 0.046187198463493596,
      "grad_norm": 0.15728501975536346,
      "learning_rate": 9.846352010304536e-05,
      "loss": 0.9327,
      "mean_token_accuracy": 0.7443478964269161,
      "num_tokens": 16026921.0,
      "step": 502
    },
    {
      "entropy": 0.9882955569773912,
      "epoch": 0.04627920483493482,
      "grad_norm": 0.16177190840244293,
      "learning_rate": 9.846045327690374e-05,
      "loss": 0.982,
      "mean_token_accuracy": 0.7290174178779125,
      "num_tokens": 16059334.0,
      "step": 503
    },
    {
      "entropy": 0.9713782593607903,
      "epoch": 0.046371211206376044,
      "grad_norm": 0.17542076110839844,
      "learning_rate": 9.845738645076211e-05,
      "loss": 0.9608,
      "mean_token_accuracy": 0.7331028394401073,
      "num_tokens": 16089982.0,
      "step": 504
    },
    {
      "entropy": 1.1129347868263721,
      "epoch": 0.04646321757781726,
      "grad_norm": 0.1756894290447235,
      "learning_rate": 9.845431962462049e-05,
      "loss": 1.1011,
      "mean_token_accuracy": 0.696821041405201,
      "num_tokens": 16121407.0,
      "step": 505
    },
    {
      "entropy": 0.9673193357884884,
      "epoch": 0.046555223949258484,
      "grad_norm": 0.1794266402721405,
      "learning_rate": 9.845125279847886e-05,
      "loss": 0.9899,
      "mean_token_accuracy": 0.7275874689221382,
      "num_tokens": 16153246.0,
      "step": 506
    },
    {
      "entropy": 0.9888970255851746,
      "epoch": 0.04664723032069971,
      "grad_norm": 0.1732143610715866,
      "learning_rate": 9.844818597233724e-05,
      "loss": 0.9943,
      "mean_token_accuracy": 0.7196294888854027,
      "num_tokens": 16185560.0,
      "step": 507
    },
    {
      "entropy": 1.0084959976375103,
      "epoch": 0.04673923669214093,
      "grad_norm": 0.16477088630199432,
      "learning_rate": 9.84451191461956e-05,
      "loss": 1.0178,
      "mean_token_accuracy": 0.7248051688075066,
      "num_tokens": 16217723.0,
      "step": 508
    },
    {
      "entropy": 1.0579893682152033,
      "epoch": 0.046831243063582155,
      "grad_norm": 0.1757245808839798,
      "learning_rate": 9.844205232005397e-05,
      "loss": 1.0792,
      "mean_token_accuracy": 0.7037666104733944,
      "num_tokens": 16248816.0,
      "step": 509
    },
    {
      "entropy": 0.9246227145195007,
      "epoch": 0.04692324943502338,
      "grad_norm": 0.16566230356693268,
      "learning_rate": 9.843898549391235e-05,
      "loss": 0.9683,
      "mean_token_accuracy": 0.7360695637762547,
      "num_tokens": 16281008.0,
      "step": 510
    },
    {
      "entropy": 1.0756047256290913,
      "epoch": 0.047015255806464595,
      "grad_norm": 0.17584773898124695,
      "learning_rate": 9.843591866777074e-05,
      "loss": 1.0733,
      "mean_token_accuracy": 0.7046601213514805,
      "num_tokens": 16312613.0,
      "step": 511
    },
    {
      "entropy": 1.0292916856706142,
      "epoch": 0.04710726217790582,
      "grad_norm": 0.16579344868659973,
      "learning_rate": 9.84328518416291e-05,
      "loss": 1.0205,
      "mean_token_accuracy": 0.7179612033069134,
      "num_tokens": 16344471.0,
      "step": 512
    },
    {
      "entropy": 0.9672829136252403,
      "epoch": 0.04719926854934704,
      "grad_norm": 0.16674520075321198,
      "learning_rate": 9.842978501548747e-05,
      "loss": 0.956,
      "mean_token_accuracy": 0.7386903539299965,
      "num_tokens": 16376025.0,
      "step": 513
    },
    {
      "entropy": 0.951029684394598,
      "epoch": 0.047291274920788266,
      "grad_norm": 0.16606657207012177,
      "learning_rate": 9.842671818934584e-05,
      "loss": 0.9282,
      "mean_token_accuracy": 0.7439240366220474,
      "num_tokens": 16408793.0,
      "step": 514
    },
    {
      "entropy": 1.0138020440936089,
      "epoch": 0.04738328129222949,
      "grad_norm": 0.16918045282363892,
      "learning_rate": 9.842365136320422e-05,
      "loss": 0.9916,
      "mean_token_accuracy": 0.7240328527987003,
      "num_tokens": 16440081.0,
      "step": 515
    },
    {
      "entropy": 0.9985101297497749,
      "epoch": 0.04747528766367071,
      "grad_norm": 0.16846980154514313,
      "learning_rate": 9.84205845370626e-05,
      "loss": 1.0057,
      "mean_token_accuracy": 0.7223841436207294,
      "num_tokens": 16472231.0,
      "step": 516
    },
    {
      "entropy": 1.029073879122734,
      "epoch": 0.04756729403511193,
      "grad_norm": 0.16698753833770752,
      "learning_rate": 9.841751771092097e-05,
      "loss": 1.0091,
      "mean_token_accuracy": 0.7174799740314484,
      "num_tokens": 16504535.0,
      "step": 517
    },
    {
      "entropy": 1.0676442086696625,
      "epoch": 0.047659300406553154,
      "grad_norm": 0.17522063851356506,
      "learning_rate": 9.841445088477934e-05,
      "loss": 1.0884,
      "mean_token_accuracy": 0.7038857564330101,
      "num_tokens": 16536450.0,
      "step": 518
    },
    {
      "entropy": 0.9670837223529816,
      "epoch": 0.04775130677799438,
      "grad_norm": 0.167304128408432,
      "learning_rate": 9.841138405863772e-05,
      "loss": 0.9748,
      "mean_token_accuracy": 0.729322899132967,
      "num_tokens": 16567837.0,
      "step": 519
    },
    {
      "entropy": 0.9472971484065056,
      "epoch": 0.0478433131494356,
      "grad_norm": 0.16083310544490814,
      "learning_rate": 9.840831723249609e-05,
      "loss": 0.9645,
      "mean_token_accuracy": 0.7310571931302547,
      "num_tokens": 16599657.0,
      "step": 520
    },
    {
      "entropy": 1.050604097545147,
      "epoch": 0.04793531952087682,
      "grad_norm": 0.17711392045021057,
      "learning_rate": 9.840525040635447e-05,
      "loss": 1.0855,
      "mean_token_accuracy": 0.7092321552336216,
      "num_tokens": 16630704.0,
      "step": 521
    },
    {
      "entropy": 1.014123309403658,
      "epoch": 0.04802732589231804,
      "grad_norm": 0.17138954997062683,
      "learning_rate": 9.840218358021285e-05,
      "loss": 1.0393,
      "mean_token_accuracy": 0.7118928618729115,
      "num_tokens": 16662594.0,
      "step": 522
    },
    {
      "entropy": 1.0701517760753632,
      "epoch": 0.048119332263759265,
      "grad_norm": 0.173891082406044,
      "learning_rate": 9.839911675407122e-05,
      "loss": 1.0983,
      "mean_token_accuracy": 0.7016746997833252,
      "num_tokens": 16694245.0,
      "step": 523
    },
    {
      "entropy": 1.0471469946205616,
      "epoch": 0.04821133863520049,
      "grad_norm": 0.1749057173728943,
      "learning_rate": 9.839604992792959e-05,
      "loss": 1.0247,
      "mean_token_accuracy": 0.7195007130503654,
      "num_tokens": 16726548.0,
      "step": 524
    },
    {
      "entropy": 0.9441933836787939,
      "epoch": 0.04830334500664171,
      "grad_norm": 0.1559460461139679,
      "learning_rate": 9.839298310178797e-05,
      "loss": 0.9394,
      "mean_token_accuracy": 0.7386441603302956,
      "num_tokens": 16758437.0,
      "step": 525
    },
    {
      "entropy": 0.9703962281346321,
      "epoch": 0.04839535137808293,
      "grad_norm": 0.16279587149620056,
      "learning_rate": 9.838991627564635e-05,
      "loss": 0.9476,
      "mean_token_accuracy": 0.7357053533196449,
      "num_tokens": 16789658.0,
      "step": 526
    },
    {
      "entropy": 1.0778263621032238,
      "epoch": 0.04848735774952415,
      "grad_norm": 0.1774022877216339,
      "learning_rate": 9.838684944950472e-05,
      "loss": 1.0648,
      "mean_token_accuracy": 0.70652761682868,
      "num_tokens": 16821415.0,
      "step": 527
    },
    {
      "entropy": 1.0894627943634987,
      "epoch": 0.048579364120965376,
      "grad_norm": 0.17032833397388458,
      "learning_rate": 9.838378262336308e-05,
      "loss": 1.0661,
      "mean_token_accuracy": 0.708966538310051,
      "num_tokens": 16853750.0,
      "step": 528
    },
    {
      "entropy": 0.9477349556982517,
      "epoch": 0.0486713704924066,
      "grad_norm": 0.1650213748216629,
      "learning_rate": 9.838071579722145e-05,
      "loss": 0.9719,
      "mean_token_accuracy": 0.7361862733960152,
      "num_tokens": 16885712.0,
      "step": 529
    },
    {
      "entropy": 1.0294384360313416,
      "epoch": 0.048763376863847824,
      "grad_norm": 0.1641075164079666,
      "learning_rate": 9.837764897107983e-05,
      "loss": 1.0118,
      "mean_token_accuracy": 0.7239221557974815,
      "num_tokens": 16917687.0,
      "step": 530
    },
    {
      "entropy": 1.008217878639698,
      "epoch": 0.04885538323528904,
      "grad_norm": 0.16842016577720642,
      "learning_rate": 9.837458214493822e-05,
      "loss": 1.0344,
      "mean_token_accuracy": 0.7140330150723457,
      "num_tokens": 16949029.0,
      "step": 531
    },
    {
      "entropy": 1.0954037606716156,
      "epoch": 0.048947389606730264,
      "grad_norm": 0.1698404848575592,
      "learning_rate": 9.837151531879658e-05,
      "loss": 1.1082,
      "mean_token_accuracy": 0.6955718025565147,
      "num_tokens": 16980918.0,
      "step": 532
    },
    {
      "entropy": 0.9094005711376667,
      "epoch": 0.04903939597817149,
      "grad_norm": 0.16477352380752563,
      "learning_rate": 9.836844849265495e-05,
      "loss": 0.931,
      "mean_token_accuracy": 0.739186592400074,
      "num_tokens": 17011627.0,
      "step": 533
    },
    {
      "entropy": 1.0502020604908466,
      "epoch": 0.04913140234961271,
      "grad_norm": 0.17290066182613373,
      "learning_rate": 9.836538166651333e-05,
      "loss": 1.064,
      "mean_token_accuracy": 0.7079633697867393,
      "num_tokens": 17044181.0,
      "step": 534
    },
    {
      "entropy": 0.9552319645881653,
      "epoch": 0.049223408721053935,
      "grad_norm": 0.1622290164232254,
      "learning_rate": 9.83623148403717e-05,
      "loss": 0.956,
      "mean_token_accuracy": 0.7367539927363396,
      "num_tokens": 17076248.0,
      "step": 535
    },
    {
      "entropy": 0.9200166612863541,
      "epoch": 0.04931541509249516,
      "grad_norm": 0.1618937999010086,
      "learning_rate": 9.835924801423008e-05,
      "loss": 0.9399,
      "mean_token_accuracy": 0.7418420985341072,
      "num_tokens": 17108734.0,
      "step": 536
    },
    {
      "entropy": 0.9201037250459194,
      "epoch": 0.049407421463936375,
      "grad_norm": 0.15908339619636536,
      "learning_rate": 9.835618118808845e-05,
      "loss": 0.9168,
      "mean_token_accuracy": 0.7472463771700859,
      "num_tokens": 17140694.0,
      "step": 537
    },
    {
      "entropy": 1.0076869651675224,
      "epoch": 0.0494994278353776,
      "grad_norm": 0.16837343573570251,
      "learning_rate": 9.835311436194683e-05,
      "loss": 1.0012,
      "mean_token_accuracy": 0.7212460935115814,
      "num_tokens": 17172544.0,
      "step": 538
    },
    {
      "entropy": 1.0293160416185856,
      "epoch": 0.04959143420681882,
      "grad_norm": 0.172407329082489,
      "learning_rate": 9.83500475358052e-05,
      "loss": 1.0324,
      "mean_token_accuracy": 0.7161697596311569,
      "num_tokens": 17204918.0,
      "step": 539
    },
    {
      "entropy": 1.011706616729498,
      "epoch": 0.049683440578260046,
      "grad_norm": 0.1620049625635147,
      "learning_rate": 9.834698070966357e-05,
      "loss": 0.9896,
      "mean_token_accuracy": 0.7315260767936707,
      "num_tokens": 17236934.0,
      "step": 540
    },
    {
      "entropy": 1.0055107064545155,
      "epoch": 0.04977544694970127,
      "grad_norm": 0.16998472809791565,
      "learning_rate": 9.834391388352195e-05,
      "loss": 1.0082,
      "mean_token_accuracy": 0.7236494198441505,
      "num_tokens": 17267766.0,
      "step": 541
    },
    {
      "entropy": 0.9763946197926998,
      "epoch": 0.04986745332114249,
      "grad_norm": 0.16132113337516785,
      "learning_rate": 9.834084705738033e-05,
      "loss": 0.9743,
      "mean_token_accuracy": 0.7367479205131531,
      "num_tokens": 17299798.0,
      "step": 542
    },
    {
      "entropy": 1.0388979576528072,
      "epoch": 0.04995945969258371,
      "grad_norm": 0.17427237331867218,
      "learning_rate": 9.83377802312387e-05,
      "loss": 1.0635,
      "mean_token_accuracy": 0.7076380364596844,
      "num_tokens": 17331596.0,
      "step": 543
    },
    {
      "entropy": 0.9870390817523003,
      "epoch": 0.050051466064024934,
      "grad_norm": 0.16441281139850616,
      "learning_rate": 9.833471340509707e-05,
      "loss": 0.9615,
      "mean_token_accuracy": 0.7325824201107025,
      "num_tokens": 17363308.0,
      "step": 544
    },
    {
      "entropy": 0.9407383874058723,
      "epoch": 0.05014347243546616,
      "grad_norm": 0.1604689359664917,
      "learning_rate": 9.833164657895543e-05,
      "loss": 0.943,
      "mean_token_accuracy": 0.7365087680518627,
      "num_tokens": 17395091.0,
      "step": 545
    },
    {
      "entropy": 1.0013816580176353,
      "epoch": 0.05023547880690738,
      "grad_norm": 0.1669689565896988,
      "learning_rate": 9.832857975281382e-05,
      "loss": 1.0139,
      "mean_token_accuracy": 0.7213457152247429,
      "num_tokens": 17426381.0,
      "step": 546
    },
    {
      "entropy": 0.9162729755043983,
      "epoch": 0.0503274851783486,
      "grad_norm": 0.15749605000019073,
      "learning_rate": 9.83255129266722e-05,
      "loss": 0.9231,
      "mean_token_accuracy": 0.7459959164261818,
      "num_tokens": 17458873.0,
      "step": 547
    },
    {
      "entropy": 1.0329615660011768,
      "epoch": 0.05041949154978982,
      "grad_norm": 0.17582499980926514,
      "learning_rate": 9.832244610053056e-05,
      "loss": 1.0477,
      "mean_token_accuracy": 0.710836473852396,
      "num_tokens": 17491280.0,
      "step": 548
    },
    {
      "entropy": 0.8994362950325012,
      "epoch": 0.050511497921231045,
      "grad_norm": 0.16417865455150604,
      "learning_rate": 9.831937927438893e-05,
      "loss": 0.923,
      "mean_token_accuracy": 0.740050058811903,
      "num_tokens": 17523000.0,
      "step": 549
    },
    {
      "entropy": 1.0479478985071182,
      "epoch": 0.05060350429267227,
      "grad_norm": 0.1678975224494934,
      "learning_rate": 9.831631244824731e-05,
      "loss": 1.0577,
      "mean_token_accuracy": 0.7132011875510216,
      "num_tokens": 17555460.0,
      "step": 550
    },
    {
      "entropy": 1.0371168218553066,
      "epoch": 0.05069551066411349,
      "grad_norm": 0.17465507984161377,
      "learning_rate": 9.831324562210568e-05,
      "loss": 1.0347,
      "mean_token_accuracy": 0.715814758092165,
      "num_tokens": 17587010.0,
      "step": 551
    },
    {
      "entropy": 0.9364320486783981,
      "epoch": 0.05078751703555471,
      "grad_norm": 0.1658712774515152,
      "learning_rate": 9.831017879596406e-05,
      "loss": 0.9437,
      "mean_token_accuracy": 0.735924705862999,
      "num_tokens": 17619179.0,
      "step": 552
    },
    {
      "entropy": 0.912044133991003,
      "epoch": 0.05087952340699593,
      "grad_norm": 0.15669046342372894,
      "learning_rate": 9.830711196982244e-05,
      "loss": 0.914,
      "mean_token_accuracy": 0.74590864777565,
      "num_tokens": 17651947.0,
      "step": 553
    },
    {
      "entropy": 1.0051476322114468,
      "epoch": 0.05097152977843716,
      "grad_norm": 0.1639261096715927,
      "learning_rate": 9.830404514368081e-05,
      "loss": 1.0048,
      "mean_token_accuracy": 0.7183269262313843,
      "num_tokens": 17684073.0,
      "step": 554
    },
    {
      "entropy": 1.007434420287609,
      "epoch": 0.05106353614987838,
      "grad_norm": 0.16628429293632507,
      "learning_rate": 9.830097831753918e-05,
      "loss": 1.0178,
      "mean_token_accuracy": 0.7237194627523422,
      "num_tokens": 17716329.0,
      "step": 555
    },
    {
      "entropy": 0.9027214385569096,
      "epoch": 0.051155542521319604,
      "grad_norm": 0.16451740264892578,
      "learning_rate": 9.829791149139755e-05,
      "loss": 0.9169,
      "mean_token_accuracy": 0.7412130460143089,
      "num_tokens": 17748237.0,
      "step": 556
    },
    {
      "entropy": 0.9756428450345993,
      "epoch": 0.05124754889276082,
      "grad_norm": 0.16507889330387115,
      "learning_rate": 9.829484466525594e-05,
      "loss": 0.9721,
      "mean_token_accuracy": 0.7340170815587044,
      "num_tokens": 17780645.0,
      "step": 557
    },
    {
      "entropy": 0.970120832324028,
      "epoch": 0.051339555264202044,
      "grad_norm": 0.16375243663787842,
      "learning_rate": 9.829177783911431e-05,
      "loss": 0.9585,
      "mean_token_accuracy": 0.7333049513399601,
      "num_tokens": 17812773.0,
      "step": 558
    },
    {
      "entropy": 0.9633282534778118,
      "epoch": 0.05143156163564327,
      "grad_norm": 0.1719096601009369,
      "learning_rate": 9.828871101297268e-05,
      "loss": 0.9813,
      "mean_token_accuracy": 0.7322728931903839,
      "num_tokens": 17844376.0,
      "step": 559
    },
    {
      "entropy": 1.0544274300336838,
      "epoch": 0.05152356800708449,
      "grad_norm": 0.172214537858963,
      "learning_rate": 9.828564418683105e-05,
      "loss": 1.0719,
      "mean_token_accuracy": 0.708241768181324,
      "num_tokens": 17875893.0,
      "step": 560
    },
    {
      "entropy": 0.9789530634880066,
      "epoch": 0.051615574378525715,
      "grad_norm": 0.16846075654029846,
      "learning_rate": 9.828257736068943e-05,
      "loss": 0.9937,
      "mean_token_accuracy": 0.7256319187581539,
      "num_tokens": 17907521.0,
      "step": 561
    },
    {
      "entropy": 0.9505788125097752,
      "epoch": 0.05170758074996693,
      "grad_norm": 0.16983327269554138,
      "learning_rate": 9.827951053454781e-05,
      "loss": 0.9656,
      "mean_token_accuracy": 0.7319183200597763,
      "num_tokens": 17939316.0,
      "step": 562
    },
    {
      "entropy": 0.9364796727895737,
      "epoch": 0.051799587121408155,
      "grad_norm": 0.16691994667053223,
      "learning_rate": 9.827644370840618e-05,
      "loss": 0.954,
      "mean_token_accuracy": 0.733334057033062,
      "num_tokens": 17971562.0,
      "step": 563
    },
    {
      "entropy": 1.093628104776144,
      "epoch": 0.05189159349284938,
      "grad_norm": 0.17215657234191895,
      "learning_rate": 9.827337688226455e-05,
      "loss": 1.1,
      "mean_token_accuracy": 0.700460460036993,
      "num_tokens": 18003552.0,
      "step": 564
    },
    {
      "entropy": 1.0223537348210812,
      "epoch": 0.0519835998642906,
      "grad_norm": 0.16822141408920288,
      "learning_rate": 9.827031005612293e-05,
      "loss": 1.0358,
      "mean_token_accuracy": 0.7176885083317757,
      "num_tokens": 18035588.0,
      "step": 565
    },
    {
      "entropy": 0.9041565731167793,
      "epoch": 0.052075606235731826,
      "grad_norm": 0.16017886996269226,
      "learning_rate": 9.82672432299813e-05,
      "loss": 0.8889,
      "mean_token_accuracy": 0.7488632276654243,
      "num_tokens": 18067353.0,
      "step": 566
    },
    {
      "entropy": 1.0327695086598396,
      "epoch": 0.05216761260717305,
      "grad_norm": 0.16594529151916504,
      "learning_rate": 9.826417640383968e-05,
      "loss": 1.025,
      "mean_token_accuracy": 0.7211896292865276,
      "num_tokens": 18099113.0,
      "step": 567
    },
    {
      "entropy": 0.9946831949055195,
      "epoch": 0.05225961897861427,
      "grad_norm": 0.1627434939146042,
      "learning_rate": 9.826110957769804e-05,
      "loss": 1.0086,
      "mean_token_accuracy": 0.7295725010335445,
      "num_tokens": 18131160.0,
      "step": 568
    },
    {
      "entropy": 0.9707537852227688,
      "epoch": 0.05235162535005549,
      "grad_norm": 0.16237430274486542,
      "learning_rate": 9.825804275155643e-05,
      "loss": 0.9552,
      "mean_token_accuracy": 0.7393050342798233,
      "num_tokens": 18162855.0,
      "step": 569
    },
    {
      "entropy": 0.9826064929366112,
      "epoch": 0.052443631721496714,
      "grad_norm": 0.16453254222869873,
      "learning_rate": 9.825497592541479e-05,
      "loss": 0.9786,
      "mean_token_accuracy": 0.7294545508921146,
      "num_tokens": 18194300.0,
      "step": 570
    },
    {
      "entropy": 0.9876124411821365,
      "epoch": 0.05253563809293794,
      "grad_norm": 0.16435474157333374,
      "learning_rate": 9.825190909927316e-05,
      "loss": 0.9753,
      "mean_token_accuracy": 0.7324376478791237,
      "num_tokens": 18225521.0,
      "step": 571
    },
    {
      "entropy": 0.8610127381980419,
      "epoch": 0.05262764446437916,
      "grad_norm": 0.15473127365112305,
      "learning_rate": 9.824884227313154e-05,
      "loss": 0.8549,
      "mean_token_accuracy": 0.7599730007350445,
      "num_tokens": 18257791.0,
      "step": 572
    },
    {
      "entropy": 0.8465031944215298,
      "epoch": 0.05271965083582038,
      "grad_norm": 0.15734876692295074,
      "learning_rate": 9.824577544698992e-05,
      "loss": 0.8582,
      "mean_token_accuracy": 0.7577155008912086,
      "num_tokens": 18289719.0,
      "step": 573
    },
    {
      "entropy": 0.9973167181015015,
      "epoch": 0.0528116572072616,
      "grad_norm": 0.16978318989276886,
      "learning_rate": 9.824270862084829e-05,
      "loss": 1.0121,
      "mean_token_accuracy": 0.7228571400046349,
      "num_tokens": 18321434.0,
      "step": 574
    },
    {
      "entropy": 0.87562489323318,
      "epoch": 0.052903663578702825,
      "grad_norm": 0.158506378531456,
      "learning_rate": 9.823964179470666e-05,
      "loss": 0.8895,
      "mean_token_accuracy": 0.7497245036065578,
      "num_tokens": 18353794.0,
      "step": 575
    },
    {
      "entropy": 0.9533016718924046,
      "epoch": 0.05299566995014405,
      "grad_norm": 0.16541872918605804,
      "learning_rate": 9.823657496856503e-05,
      "loss": 0.9454,
      "mean_token_accuracy": 0.7388780787587166,
      "num_tokens": 18385555.0,
      "step": 576
    },
    {
      "entropy": 0.9944054484367371,
      "epoch": 0.05308767632158527,
      "grad_norm": 0.16287018358707428,
      "learning_rate": 9.823350814242341e-05,
      "loss": 0.9957,
      "mean_token_accuracy": 0.7251812964677811,
      "num_tokens": 18418031.0,
      "step": 577
    },
    {
      "entropy": 0.9897592887282372,
      "epoch": 0.05317968269302649,
      "grad_norm": 0.1654137820005417,
      "learning_rate": 9.823044131628179e-05,
      "loss": 0.972,
      "mean_token_accuracy": 0.7269420549273491,
      "num_tokens": 18449867.0,
      "step": 578
    },
    {
      "entropy": 1.0996099598705769,
      "epoch": 0.05327168906446771,
      "grad_norm": 0.17410515248775482,
      "learning_rate": 9.822737449014016e-05,
      "loss": 1.112,
      "mean_token_accuracy": 0.7021746337413788,
      "num_tokens": 18481309.0,
      "step": 579
    },
    {
      "entropy": 0.9526960216462612,
      "epoch": 0.05336369543590894,
      "grad_norm": 0.1659986823797226,
      "learning_rate": 9.822430766399853e-05,
      "loss": 0.966,
      "mean_token_accuracy": 0.7339910306036472,
      "num_tokens": 18512907.0,
      "step": 580
    },
    {
      "entropy": 1.0467368625104427,
      "epoch": 0.05345570180735016,
      "grad_norm": 0.17283470928668976,
      "learning_rate": 9.822124083785691e-05,
      "loss": 1.0499,
      "mean_token_accuracy": 0.7104004435241222,
      "num_tokens": 18543923.0,
      "step": 581
    },
    {
      "entropy": 1.0006827861070633,
      "epoch": 0.053547708178791384,
      "grad_norm": 0.16461893916130066,
      "learning_rate": 9.821817401171528e-05,
      "loss": 1.0159,
      "mean_token_accuracy": 0.7240800522267818,
      "num_tokens": 18576062.0,
      "step": 582
    },
    {
      "entropy": 0.9019467569887638,
      "epoch": 0.0536397145502326,
      "grad_norm": 0.16039246320724487,
      "learning_rate": 9.821510718557366e-05,
      "loss": 0.9083,
      "mean_token_accuracy": 0.743061650544405,
      "num_tokens": 18607134.0,
      "step": 583
    },
    {
      "entropy": 0.9684820659458637,
      "epoch": 0.053731720921673824,
      "grad_norm": 0.16630397737026215,
      "learning_rate": 9.821204035943204e-05,
      "loss": 0.9933,
      "mean_token_accuracy": 0.7307913824915886,
      "num_tokens": 18639079.0,
      "step": 584
    },
    {
      "entropy": 0.962744913995266,
      "epoch": 0.05382372729311505,
      "grad_norm": 0.16911311447620392,
      "learning_rate": 9.82089735332904e-05,
      "loss": 0.9789,
      "mean_token_accuracy": 0.7290943525731564,
      "num_tokens": 18670621.0,
      "step": 585
    },
    {
      "entropy": 0.9364055283367634,
      "epoch": 0.05391573366455627,
      "grad_norm": 0.16078977286815643,
      "learning_rate": 9.820590670714877e-05,
      "loss": 0.9252,
      "mean_token_accuracy": 0.7425668202340603,
      "num_tokens": 18702112.0,
      "step": 586
    },
    {
      "entropy": 0.9418363124132156,
      "epoch": 0.054007740035997495,
      "grad_norm": 0.16166391968727112,
      "learning_rate": 9.820283988100714e-05,
      "loss": 0.9429,
      "mean_token_accuracy": 0.7364592291414738,
      "num_tokens": 18734141.0,
      "step": 587
    },
    {
      "entropy": 1.0757093168795109,
      "epoch": 0.05409974640743871,
      "grad_norm": 0.16913855075836182,
      "learning_rate": 9.819977305486552e-05,
      "loss": 1.0768,
      "mean_token_accuracy": 0.7042913101613522,
      "num_tokens": 18766542.0,
      "step": 588
    },
    {
      "entropy": 0.9823492988944054,
      "epoch": 0.054191752778879936,
      "grad_norm": 0.16311554610729218,
      "learning_rate": 9.81967062287239e-05,
      "loss": 0.975,
      "mean_token_accuracy": 0.731862660497427,
      "num_tokens": 18798699.0,
      "step": 589
    },
    {
      "entropy": 0.9856898225843906,
      "epoch": 0.05428375915032116,
      "grad_norm": 0.162748321890831,
      "learning_rate": 9.819363940258227e-05,
      "loss": 0.9882,
      "mean_token_accuracy": 0.7264587134122849,
      "num_tokens": 18831094.0,
      "step": 590
    },
    {
      "entropy": 0.9645604155957699,
      "epoch": 0.05437576552176238,
      "grad_norm": 0.16255250573158264,
      "learning_rate": 9.819057257644064e-05,
      "loss": 0.9594,
      "mean_token_accuracy": 0.7331061474978924,
      "num_tokens": 18863564.0,
      "step": 591
    },
    {
      "entropy": 0.9556578919291496,
      "epoch": 0.054467771893203606,
      "grad_norm": 0.16019552946090698,
      "learning_rate": 9.818750575029902e-05,
      "loss": 0.9554,
      "mean_token_accuracy": 0.7303911596536636,
      "num_tokens": 18895284.0,
      "step": 592
    },
    {
      "entropy": 1.029592514038086,
      "epoch": 0.05455977826464483,
      "grad_norm": 0.16847193241119385,
      "learning_rate": 9.818443892415739e-05,
      "loss": 1.0226,
      "mean_token_accuracy": 0.7190484926104546,
      "num_tokens": 18927640.0,
      "step": 593
    },
    {
      "entropy": 0.9549121893942356,
      "epoch": 0.05465178463608605,
      "grad_norm": 0.1589156687259674,
      "learning_rate": 9.818137209801577e-05,
      "loss": 0.9563,
      "mean_token_accuracy": 0.7374194748699665,
      "num_tokens": 18959648.0,
      "step": 594
    },
    {
      "entropy": 0.9448903910815716,
      "epoch": 0.05474379100752727,
      "grad_norm": 0.16188326478004456,
      "learning_rate": 9.817830527187414e-05,
      "loss": 0.9391,
      "mean_token_accuracy": 0.7418914958834648,
      "num_tokens": 18990902.0,
      "step": 595
    },
    {
      "entropy": 0.9302772954106331,
      "epoch": 0.054835797378968494,
      "grad_norm": 0.17251761257648468,
      "learning_rate": 9.817523844573252e-05,
      "loss": 0.9393,
      "mean_token_accuracy": 0.7366911098361015,
      "num_tokens": 19022741.0,
      "step": 596
    },
    {
      "entropy": 0.8710775151848793,
      "epoch": 0.05492780375040972,
      "grad_norm": 0.16116918623447418,
      "learning_rate": 9.817217161959089e-05,
      "loss": 0.8936,
      "mean_token_accuracy": 0.7474441565573215,
      "num_tokens": 19054346.0,
      "step": 597
    },
    {
      "entropy": 0.9943357296288013,
      "epoch": 0.05501981012185094,
      "grad_norm": 0.16670559346675873,
      "learning_rate": 9.816910479344926e-05,
      "loss": 1.0106,
      "mean_token_accuracy": 0.7208894491195679,
      "num_tokens": 19086890.0,
      "step": 598
    },
    {
      "entropy": 0.9530030973255634,
      "epoch": 0.05511181649329216,
      "grad_norm": 0.16489286720752716,
      "learning_rate": 9.816603796730764e-05,
      "loss": 0.9588,
      "mean_token_accuracy": 0.7296338677406311,
      "num_tokens": 19118142.0,
      "step": 599
    },
    {
      "entropy": 0.8646176792681217,
      "epoch": 0.05520382286473338,
      "grad_norm": 0.16042181849479675,
      "learning_rate": 9.816297114116602e-05,
      "loss": 0.8834,
      "mean_token_accuracy": 0.7538492642343044,
      "num_tokens": 19150555.0,
      "step": 600
    },
    {
      "entropy": 0.9691968113183975,
      "epoch": 0.055295829236174605,
      "grad_norm": 0.16191726922988892,
      "learning_rate": 9.815990431502439e-05,
      "loss": 0.9762,
      "mean_token_accuracy": 0.730664562433958,
      "num_tokens": 19182543.0,
      "step": 601
    },
    {
      "entropy": 0.9071401357650757,
      "epoch": 0.05538783560761583,
      "grad_norm": 0.1626460999250412,
      "learning_rate": 9.815683748888275e-05,
      "loss": 0.9264,
      "mean_token_accuracy": 0.7449319437146187,
      "num_tokens": 19215060.0,
      "step": 602
    },
    {
      "entropy": 0.968286719173193,
      "epoch": 0.05547984197905705,
      "grad_norm": 0.1695675104856491,
      "learning_rate": 9.815377066274112e-05,
      "loss": 0.9747,
      "mean_token_accuracy": 0.7253856211900711,
      "num_tokens": 19246779.0,
      "step": 603
    },
    {
      "entropy": 0.9944621175527573,
      "epoch": 0.05557184835049827,
      "grad_norm": 0.1652134507894516,
      "learning_rate": 9.815070383659952e-05,
      "loss": 0.9995,
      "mean_token_accuracy": 0.7228061333298683,
      "num_tokens": 19279225.0,
      "step": 604
    },
    {
      "entropy": 0.9213341325521469,
      "epoch": 0.05566385472193949,
      "grad_norm": 0.17412346601486206,
      "learning_rate": 9.814763701045789e-05,
      "loss": 0.932,
      "mean_token_accuracy": 0.7381015047430992,
      "num_tokens": 19310793.0,
      "step": 605
    },
    {
      "entropy": 0.9655103795230389,
      "epoch": 0.05575586109338072,
      "grad_norm": 0.169269859790802,
      "learning_rate": 9.814457018431625e-05,
      "loss": 0.9626,
      "mean_token_accuracy": 0.7287351712584496,
      "num_tokens": 19341538.0,
      "step": 606
    },
    {
      "entropy": 0.9082975052297115,
      "epoch": 0.05584786746482194,
      "grad_norm": 0.16562685370445251,
      "learning_rate": 9.814150335817462e-05,
      "loss": 0.8921,
      "mean_token_accuracy": 0.7557516805827618,
      "num_tokens": 19373131.0,
      "step": 607
    },
    {
      "entropy": 0.9801355004310608,
      "epoch": 0.055939873836263164,
      "grad_norm": 0.17090274393558502,
      "learning_rate": 9.8138436532033e-05,
      "loss": 0.9943,
      "mean_token_accuracy": 0.7248963490128517,
      "num_tokens": 19404428.0,
      "step": 608
    },
    {
      "entropy": 1.0464669652283192,
      "epoch": 0.05603188020770438,
      "grad_norm": 0.16184228658676147,
      "learning_rate": 9.813536970589138e-05,
      "loss": 1.0378,
      "mean_token_accuracy": 0.7186429165303707,
      "num_tokens": 19436419.0,
      "step": 609
    },
    {
      "entropy": 1.0447305738925934,
      "epoch": 0.056123886579145604,
      "grad_norm": 0.16772888600826263,
      "learning_rate": 9.813230287974975e-05,
      "loss": 1.051,
      "mean_token_accuracy": 0.7093856781721115,
      "num_tokens": 19468448.0,
      "step": 610
    },
    {
      "entropy": 0.9712334387004375,
      "epoch": 0.05621589295058683,
      "grad_norm": 0.16332221031188965,
      "learning_rate": 9.812923605360812e-05,
      "loss": 0.9666,
      "mean_token_accuracy": 0.7356022410094738,
      "num_tokens": 19500444.0,
      "step": 611
    },
    {
      "entropy": 0.9937260150909424,
      "epoch": 0.05630789932202805,
      "grad_norm": 0.17043672502040863,
      "learning_rate": 9.81261692274665e-05,
      "loss": 0.9982,
      "mean_token_accuracy": 0.7271057181060314,
      "num_tokens": 19532443.0,
      "step": 612
    },
    {
      "entropy": 1.074421912431717,
      "epoch": 0.056399905693469275,
      "grad_norm": 0.17714686691761017,
      "learning_rate": 9.812310240132487e-05,
      "loss": 1.0956,
      "mean_token_accuracy": 0.7010564096271992,
      "num_tokens": 19563968.0,
      "step": 613
    },
    {
      "entropy": 0.8630592413246632,
      "epoch": 0.05649191206491049,
      "grad_norm": 0.16226260364055634,
      "learning_rate": 9.812003557518325e-05,
      "loss": 0.8873,
      "mean_token_accuracy": 0.7529776021838188,
      "num_tokens": 19596197.0,
      "step": 614
    },
    {
      "entropy": 0.9494337886571884,
      "epoch": 0.056583918436351716,
      "grad_norm": 0.1683243066072464,
      "learning_rate": 9.811696874904163e-05,
      "loss": 0.9583,
      "mean_token_accuracy": 0.7314475849270821,
      "num_tokens": 19628453.0,
      "step": 615
    },
    {
      "entropy": 0.9498128183186054,
      "epoch": 0.05667592480779294,
      "grad_norm": 0.1673831194639206,
      "learning_rate": 9.81139019229e-05,
      "loss": 0.9615,
      "mean_token_accuracy": 0.7353732399642467,
      "num_tokens": 19661123.0,
      "step": 616
    },
    {
      "entropy": 0.9863654561340809,
      "epoch": 0.05676793117923416,
      "grad_norm": 0.16342179477214813,
      "learning_rate": 9.811083509675837e-05,
      "loss": 0.9677,
      "mean_token_accuracy": 0.7291566766798496,
      "num_tokens": 19693161.0,
      "step": 617
    },
    {
      "entropy": 1.048695120960474,
      "epoch": 0.056859937550675387,
      "grad_norm": 0.16500575840473175,
      "learning_rate": 9.810776827061674e-05,
      "loss": 1.0384,
      "mean_token_accuracy": 0.7136950083076954,
      "num_tokens": 19725846.0,
      "step": 618
    },
    {
      "entropy": 0.8852150216698647,
      "epoch": 0.0569519439221166,
      "grad_norm": 0.1556621938943863,
      "learning_rate": 9.810470144447512e-05,
      "loss": 0.8488,
      "mean_token_accuracy": 0.760561365634203,
      "num_tokens": 19757634.0,
      "step": 619
    },
    {
      "entropy": 1.0031042732298374,
      "epoch": 0.05704395029355783,
      "grad_norm": 0.17492416501045227,
      "learning_rate": 9.81016346183335e-05,
      "loss": 1.0059,
      "mean_token_accuracy": 0.7231873422861099,
      "num_tokens": 19789873.0,
      "step": 620
    },
    {
      "entropy": 0.9483481906354427,
      "epoch": 0.05713595666499905,
      "grad_norm": 0.1657971441745758,
      "learning_rate": 9.809856779219187e-05,
      "loss": 0.9279,
      "mean_token_accuracy": 0.7393903471529484,
      "num_tokens": 19821539.0,
      "step": 621
    },
    {
      "entropy": 0.9609033651649952,
      "epoch": 0.057227963036440274,
      "grad_norm": 0.16380846500396729,
      "learning_rate": 9.809550096605023e-05,
      "loss": 0.9426,
      "mean_token_accuracy": 0.7385537661612034,
      "num_tokens": 19852712.0,
      "step": 622
    },
    {
      "entropy": 0.9985687658190727,
      "epoch": 0.0573199694078815,
      "grad_norm": 0.1723949909210205,
      "learning_rate": 9.809243413990862e-05,
      "loss": 1.0237,
      "mean_token_accuracy": 0.7171773463487625,
      "num_tokens": 19884992.0,
      "step": 623
    },
    {
      "entropy": 0.9404069483280182,
      "epoch": 0.05741197577932272,
      "grad_norm": 0.16850896179676056,
      "learning_rate": 9.808936731376698e-05,
      "loss": 0.9524,
      "mean_token_accuracy": 0.7279058285057545,
      "num_tokens": 19917236.0,
      "step": 624
    },
    {
      "entropy": 0.9506558645516634,
      "epoch": 0.05750398215076394,
      "grad_norm": 0.16849637031555176,
      "learning_rate": 9.808630048762536e-05,
      "loss": 0.9653,
      "mean_token_accuracy": 0.7319151535630226,
      "num_tokens": 19949583.0,
      "step": 625
    },
    {
      "entropy": 0.9219024889171124,
      "epoch": 0.05759598852220516,
      "grad_norm": 0.16453535854816437,
      "learning_rate": 9.808323366148373e-05,
      "loss": 0.9215,
      "mean_token_accuracy": 0.7456205748021603,
      "num_tokens": 19981623.0,
      "step": 626
    },
    {
      "entropy": 1.013407040387392,
      "epoch": 0.057687994893646385,
      "grad_norm": 0.17392785847187042,
      "learning_rate": 9.808016683534211e-05,
      "loss": 1.0485,
      "mean_token_accuracy": 0.7131614647805691,
      "num_tokens": 20013224.0,
      "step": 627
    },
    {
      "entropy": 0.9801394529640675,
      "epoch": 0.05778000126508761,
      "grad_norm": 0.1665116846561432,
      "learning_rate": 9.807710000920048e-05,
      "loss": 0.9839,
      "mean_token_accuracy": 0.7241210155189037,
      "num_tokens": 20045761.0,
      "step": 628
    },
    {
      "entropy": 0.9025226980447769,
      "epoch": 0.05787200763652883,
      "grad_norm": 0.1585804522037506,
      "learning_rate": 9.807403318305885e-05,
      "loss": 0.9077,
      "mean_token_accuracy": 0.7458856888115406,
      "num_tokens": 20078143.0,
      "step": 629
    },
    {
      "entropy": 0.9686283729970455,
      "epoch": 0.05796401400797005,
      "grad_norm": 0.16333197057247162,
      "learning_rate": 9.807096635691723e-05,
      "loss": 0.9665,
      "mean_token_accuracy": 0.7290001772344112,
      "num_tokens": 20110479.0,
      "step": 630
    },
    {
      "entropy": 0.9154176339507103,
      "epoch": 0.05805602037941127,
      "grad_norm": 0.16539502143859863,
      "learning_rate": 9.806789953077561e-05,
      "loss": 0.9424,
      "mean_token_accuracy": 0.7402860037982464,
      "num_tokens": 20142214.0,
      "step": 631
    },
    {
      "entropy": 0.9714864902198315,
      "epoch": 0.0581480267508525,
      "grad_norm": 0.16718418896198273,
      "learning_rate": 9.806483270463398e-05,
      "loss": 0.9718,
      "mean_token_accuracy": 0.7289915159344673,
      "num_tokens": 20173331.0,
      "step": 632
    },
    {
      "entropy": 0.8834399022161961,
      "epoch": 0.05824003312229372,
      "grad_norm": 0.16289383172988892,
      "learning_rate": 9.806176587849235e-05,
      "loss": 0.8866,
      "mean_token_accuracy": 0.7500693127512932,
      "num_tokens": 20205091.0,
      "step": 633
    },
    {
      "entropy": 0.9796353541314602,
      "epoch": 0.058332039493734944,
      "grad_norm": 0.16527166962623596,
      "learning_rate": 9.805869905235072e-05,
      "loss": 1.0043,
      "mean_token_accuracy": 0.7237598747014999,
      "num_tokens": 20236994.0,
      "step": 634
    },
    {
      "entropy": 0.909358948469162,
      "epoch": 0.05842404586517616,
      "grad_norm": 0.15741148591041565,
      "learning_rate": 9.80556322262091e-05,
      "loss": 0.9055,
      "mean_token_accuracy": 0.7489256225526333,
      "num_tokens": 20269097.0,
      "step": 635
    },
    {
      "entropy": 1.0318338200449944,
      "epoch": 0.058516052236617384,
      "grad_norm": 0.16267406940460205,
      "learning_rate": 9.805256540006748e-05,
      "loss": 1.0238,
      "mean_token_accuracy": 0.7141961343586445,
      "num_tokens": 20301196.0,
      "step": 636
    },
    {
      "entropy": 1.040244996547699,
      "epoch": 0.05860805860805861,
      "grad_norm": 0.17041735351085663,
      "learning_rate": 9.804949857392585e-05,
      "loss": 1.0522,
      "mean_token_accuracy": 0.7122577428817749,
      "num_tokens": 20333326.0,
      "step": 637
    },
    {
      "entropy": 0.8750684708356857,
      "epoch": 0.05870006497949983,
      "grad_norm": 0.1510256975889206,
      "learning_rate": 9.804643174778421e-05,
      "loss": 0.8608,
      "mean_token_accuracy": 0.7590311542153358,
      "num_tokens": 20365251.0,
      "step": 638
    },
    {
      "entropy": 0.9018204025924206,
      "epoch": 0.058792071350941055,
      "grad_norm": 0.16271653771400452,
      "learning_rate": 9.80433649216426e-05,
      "loss": 0.9062,
      "mean_token_accuracy": 0.7503054365515709,
      "num_tokens": 20396798.0,
      "step": 639
    },
    {
      "entropy": 1.0158135928213596,
      "epoch": 0.05888407772238227,
      "grad_norm": 0.1715613305568695,
      "learning_rate": 9.804029809550096e-05,
      "loss": 1.0386,
      "mean_token_accuracy": 0.7216349169611931,
      "num_tokens": 20428488.0,
      "step": 640
    },
    {
      "entropy": 1.0293894298374653,
      "epoch": 0.058976084093823496,
      "grad_norm": 0.16205903887748718,
      "learning_rate": 9.803723126935935e-05,
      "loss": 1.0273,
      "mean_token_accuracy": 0.7176331728696823,
      "num_tokens": 20460364.0,
      "step": 641
    },
    {
      "entropy": 0.9762745574116707,
      "epoch": 0.05906809046526472,
      "grad_norm": 0.16408756375312805,
      "learning_rate": 9.803416444321773e-05,
      "loss": 0.9721,
      "mean_token_accuracy": 0.7313985861837864,
      "num_tokens": 20493105.0,
      "step": 642
    },
    {
      "entropy": 1.039294358342886,
      "epoch": 0.05916009683670594,
      "grad_norm": 0.1625339537858963,
      "learning_rate": 9.80310976170761e-05,
      "loss": 1.0087,
      "mean_token_accuracy": 0.7228131368756294,
      "num_tokens": 20524742.0,
      "step": 643
    },
    {
      "entropy": 0.9685094505548477,
      "epoch": 0.05925210320814717,
      "grad_norm": 0.1643614023923874,
      "learning_rate": 9.802803079093446e-05,
      "loss": 0.9516,
      "mean_token_accuracy": 0.7309996075928211,
      "num_tokens": 20556870.0,
      "step": 644
    },
    {
      "entropy": 0.9781741611659527,
      "epoch": 0.05934410957958838,
      "grad_norm": 0.17166036367416382,
      "learning_rate": 9.802496396479284e-05,
      "loss": 0.9816,
      "mean_token_accuracy": 0.7277286499738693,
      "num_tokens": 20587590.0,
      "step": 645
    },
    {
      "entropy": 0.9321715421974659,
      "epoch": 0.05943611595102961,
      "grad_norm": 0.15926851332187653,
      "learning_rate": 9.802189713865123e-05,
      "loss": 0.9296,
      "mean_token_accuracy": 0.739145640283823,
      "num_tokens": 20619987.0,
      "step": 646
    },
    {
      "entropy": 0.8368671890348196,
      "epoch": 0.05952812232247083,
      "grad_norm": 0.15730631351470947,
      "learning_rate": 9.80188303125096e-05,
      "loss": 0.8331,
      "mean_token_accuracy": 0.761964064091444,
      "num_tokens": 20651859.0,
      "step": 647
    },
    {
      "entropy": 0.8395576402544975,
      "epoch": 0.059620128693912054,
      "grad_norm": 0.15268109738826752,
      "learning_rate": 9.801576348636796e-05,
      "loss": 0.8089,
      "mean_token_accuracy": 0.768634881824255,
      "num_tokens": 20683616.0,
      "step": 648
    },
    {
      "entropy": 1.0215886235237122,
      "epoch": 0.05971213506535328,
      "grad_norm": 0.17276978492736816,
      "learning_rate": 9.801269666022633e-05,
      "loss": 1.0533,
      "mean_token_accuracy": 0.7140671759843826,
      "num_tokens": 20714915.0,
      "step": 649
    },
    {
      "entropy": 0.8608022443950176,
      "epoch": 0.059804141436794495,
      "grad_norm": 0.16115765273571014,
      "learning_rate": 9.800962983408471e-05,
      "loss": 0.9137,
      "mean_token_accuracy": 0.7497973553836346,
      "num_tokens": 20746818.0,
      "step": 650
    },
    {
      "entropy": 0.9612746220082045,
      "epoch": 0.05989614780823572,
      "grad_norm": 0.16959808766841888,
      "learning_rate": 9.800656300794309e-05,
      "loss": 0.9726,
      "mean_token_accuracy": 0.7311719208955765,
      "num_tokens": 20778602.0,
      "step": 651
    },
    {
      "entropy": 0.9875885434448719,
      "epoch": 0.05998815417967694,
      "grad_norm": 0.16284504532814026,
      "learning_rate": 9.800349618180146e-05,
      "loss": 0.9966,
      "mean_token_accuracy": 0.7249477878212929,
      "num_tokens": 20810961.0,
      "step": 652
    },
    {
      "entropy": 0.9503909014165401,
      "epoch": 0.060080160551118166,
      "grad_norm": 0.16453111171722412,
      "learning_rate": 9.800042935565983e-05,
      "loss": 0.957,
      "mean_token_accuracy": 0.7324551977217197,
      "num_tokens": 20843389.0,
      "step": 653
    },
    {
      "entropy": 1.0258820429444313,
      "epoch": 0.06017216692255939,
      "grad_norm": 0.1692570596933365,
      "learning_rate": 9.799736252951821e-05,
      "loss": 1.0428,
      "mean_token_accuracy": 0.7197996079921722,
      "num_tokens": 20875637.0,
      "step": 654
    },
    {
      "entropy": 0.9980999454855919,
      "epoch": 0.06026417329400061,
      "grad_norm": 0.16594214737415314,
      "learning_rate": 9.799429570337658e-05,
      "loss": 1.0001,
      "mean_token_accuracy": 0.7217110693454742,
      "num_tokens": 20907585.0,
      "step": 655
    },
    {
      "entropy": 1.05484539270401,
      "epoch": 0.06035617966544183,
      "grad_norm": 0.17190194129943848,
      "learning_rate": 9.799122887723496e-05,
      "loss": 1.0513,
      "mean_token_accuracy": 0.7128286510705948,
      "num_tokens": 20938690.0,
      "step": 656
    },
    {
      "entropy": 0.9426363259553909,
      "epoch": 0.06044818603688305,
      "grad_norm": 0.168714702129364,
      "learning_rate": 9.798816205109333e-05,
      "loss": 0.9587,
      "mean_token_accuracy": 0.7298561073839664,
      "num_tokens": 20970885.0,
      "step": 657
    },
    {
      "entropy": 0.8432633951306343,
      "epoch": 0.06054019240832428,
      "grad_norm": 0.15879929065704346,
      "learning_rate": 9.798509522495171e-05,
      "loss": 0.8637,
      "mean_token_accuracy": 0.758915726095438,
      "num_tokens": 21002897.0,
      "step": 658
    },
    {
      "entropy": 0.9566614702343941,
      "epoch": 0.0606321987797655,
      "grad_norm": 0.16445676982402802,
      "learning_rate": 9.798202839881008e-05,
      "loss": 0.9573,
      "mean_token_accuracy": 0.7340426258742809,
      "num_tokens": 21034574.0,
      "step": 659
    },
    {
      "entropy": 1.0204659141600132,
      "epoch": 0.060724205151206724,
      "grad_norm": 0.16645212471485138,
      "learning_rate": 9.797896157266844e-05,
      "loss": 1.0242,
      "mean_token_accuracy": 0.7199134491384029,
      "num_tokens": 21066688.0,
      "step": 660
    },
    {
      "entropy": 1.0958493538200855,
      "epoch": 0.06081621152264794,
      "grad_norm": 0.17646637558937073,
      "learning_rate": 9.797589474652682e-05,
      "loss": 1.1094,
      "mean_token_accuracy": 0.6997278966009617,
      "num_tokens": 21099014.0,
      "step": 661
    },
    {
      "entropy": 1.0627868585288525,
      "epoch": 0.060908217894089164,
      "grad_norm": 0.16993241012096405,
      "learning_rate": 9.79728279203852e-05,
      "loss": 1.0459,
      "mean_token_accuracy": 0.7173599191009998,
      "num_tokens": 21131271.0,
      "step": 662
    },
    {
      "entropy": 1.0007079876959324,
      "epoch": 0.06100022426553039,
      "grad_norm": 0.1617305874824524,
      "learning_rate": 9.796976109424357e-05,
      "loss": 0.9693,
      "mean_token_accuracy": 0.7304843105375767,
      "num_tokens": 21163033.0,
      "step": 663
    },
    {
      "entropy": 0.9579825848340988,
      "epoch": 0.06109223063697161,
      "grad_norm": 0.16417573392391205,
      "learning_rate": 9.796669426810194e-05,
      "loss": 0.9545,
      "mean_token_accuracy": 0.735863633453846,
      "num_tokens": 21194433.0,
      "step": 664
    },
    {
      "entropy": 0.9619919620454311,
      "epoch": 0.061184237008412835,
      "grad_norm": 0.163755863904953,
      "learning_rate": 9.796362744196031e-05,
      "loss": 0.9563,
      "mean_token_accuracy": 0.7320655733346939,
      "num_tokens": 21225835.0,
      "step": 665
    },
    {
      "entropy": 0.929864663630724,
      "epoch": 0.06127624337985405,
      "grad_norm": 0.16609299182891846,
      "learning_rate": 9.796056061581869e-05,
      "loss": 0.9473,
      "mean_token_accuracy": 0.7374901138246059,
      "num_tokens": 21257647.0,
      "step": 666
    },
    {
      "entropy": 0.8059710673987865,
      "epoch": 0.061368249751295276,
      "grad_norm": 0.15594281256198883,
      "learning_rate": 9.795749378967707e-05,
      "loss": 0.8209,
      "mean_token_accuracy": 0.7718138620257378,
      "num_tokens": 21290287.0,
      "step": 667
    },
    {
      "entropy": 0.9865346737205982,
      "epoch": 0.0614602561227365,
      "grad_norm": 0.16683848202228546,
      "learning_rate": 9.795442696353544e-05,
      "loss": 0.9886,
      "mean_token_accuracy": 0.7253305949270725,
      "num_tokens": 21322298.0,
      "step": 668
    },
    {
      "entropy": 0.9109359867870808,
      "epoch": 0.06155226249417772,
      "grad_norm": 0.16543343663215637,
      "learning_rate": 9.795136013739381e-05,
      "loss": 0.9057,
      "mean_token_accuracy": 0.7455473244190216,
      "num_tokens": 21354325.0,
      "step": 669
    },
    {
      "entropy": 0.9279623962938786,
      "epoch": 0.06164426886561895,
      "grad_norm": 0.1550312638282776,
      "learning_rate": 9.794829331125219e-05,
      "loss": 0.9208,
      "mean_token_accuracy": 0.7456959709525108,
      "num_tokens": 21386876.0,
      "step": 670
    },
    {
      "entropy": 0.9169107973575592,
      "epoch": 0.06173627523706016,
      "grad_norm": 0.15861962735652924,
      "learning_rate": 9.794522648511056e-05,
      "loss": 0.9222,
      "mean_token_accuracy": 0.7412844784557819,
      "num_tokens": 21419111.0,
      "step": 671
    },
    {
      "entropy": 0.9637071453034878,
      "epoch": 0.06182828160850139,
      "grad_norm": 0.16707174479961395,
      "learning_rate": 9.794215965896894e-05,
      "loss": 0.9628,
      "mean_token_accuracy": 0.7339217066764832,
      "num_tokens": 21451138.0,
      "step": 672
    },
    {
      "entropy": 1.060596663504839,
      "epoch": 0.06192028797994261,
      "grad_norm": 0.1684867888689041,
      "learning_rate": 9.793909283282732e-05,
      "loss": 1.0696,
      "mean_token_accuracy": 0.7061467282474041,
      "num_tokens": 21482995.0,
      "step": 673
    },
    {
      "entropy": 0.9379373602569103,
      "epoch": 0.062012294351383834,
      "grad_norm": 0.16227133572101593,
      "learning_rate": 9.793602600668569e-05,
      "loss": 0.9503,
      "mean_token_accuracy": 0.7331300526857376,
      "num_tokens": 21514573.0,
      "step": 674
    },
    {
      "entropy": 0.9449528157711029,
      "epoch": 0.06210430072282506,
      "grad_norm": 0.16113756597042084,
      "learning_rate": 9.793295918054406e-05,
      "loss": 0.9573,
      "mean_token_accuracy": 0.7376067005097866,
      "num_tokens": 21547028.0,
      "step": 675
    },
    {
      "entropy": 0.8376788571476936,
      "epoch": 0.062196307094266275,
      "grad_norm": 0.1625738888978958,
      "learning_rate": 9.792989235440242e-05,
      "loss": 0.8365,
      "mean_token_accuracy": 0.7637463100254536,
      "num_tokens": 21578925.0,
      "step": 676
    },
    {
      "entropy": 0.9880265146493912,
      "epoch": 0.0622883134657075,
      "grad_norm": 0.16655480861663818,
      "learning_rate": 9.792682552826082e-05,
      "loss": 0.9893,
      "mean_token_accuracy": 0.7207247093319893,
      "num_tokens": 21610214.0,
      "step": 677
    },
    {
      "entropy": 0.9505514353513718,
      "epoch": 0.06238031983714872,
      "grad_norm": 0.1738533228635788,
      "learning_rate": 9.792375870211919e-05,
      "loss": 0.9372,
      "mean_token_accuracy": 0.7403922118246555,
      "num_tokens": 21641181.0,
      "step": 678
    },
    {
      "entropy": 1.0301849208772182,
      "epoch": 0.062472326208589946,
      "grad_norm": 0.1642848551273346,
      "learning_rate": 9.792069187597756e-05,
      "loss": 1.037,
      "mean_token_accuracy": 0.7126371003687382,
      "num_tokens": 21673222.0,
      "step": 679
    },
    {
      "entropy": 1.006374005228281,
      "epoch": 0.06256433258003116,
      "grad_norm": 0.16483119130134583,
      "learning_rate": 9.791762504983592e-05,
      "loss": 1.0097,
      "mean_token_accuracy": 0.7190174385905266,
      "num_tokens": 21705297.0,
      "step": 680
    },
    {
      "entropy": 0.9991269782185555,
      "epoch": 0.06265633895147239,
      "grad_norm": 0.16770169138908386,
      "learning_rate": 9.79145582236943e-05,
      "loss": 1.0231,
      "mean_token_accuracy": 0.7178670018911362,
      "num_tokens": 21737521.0,
      "step": 681
    },
    {
      "entropy": 0.9773536287248135,
      "epoch": 0.06274834532291361,
      "grad_norm": 0.16942104697227478,
      "learning_rate": 9.791149139755269e-05,
      "loss": 0.9747,
      "mean_token_accuracy": 0.7256374433636665,
      "num_tokens": 21769768.0,
      "step": 682
    },
    {
      "entropy": 1.002865333110094,
      "epoch": 0.06284035169435484,
      "grad_norm": 0.1735992431640625,
      "learning_rate": 9.790842457141105e-05,
      "loss": 1.0143,
      "mean_token_accuracy": 0.7216566205024719,
      "num_tokens": 21801639.0,
      "step": 683
    },
    {
      "entropy": 0.9517209865152836,
      "epoch": 0.06293235806579606,
      "grad_norm": 0.15748152136802673,
      "learning_rate": 9.790535774526942e-05,
      "loss": 0.9502,
      "mean_token_accuracy": 0.7377259433269501,
      "num_tokens": 21834407.0,
      "step": 684
    },
    {
      "entropy": 0.9664385803043842,
      "epoch": 0.06302436443723727,
      "grad_norm": 0.1655987650156021,
      "learning_rate": 9.79022909191278e-05,
      "loss": 0.9764,
      "mean_token_accuracy": 0.7306914515793324,
      "num_tokens": 21866476.0,
      "step": 685
    },
    {
      "entropy": 0.9957532845437527,
      "epoch": 0.0631163708086785,
      "grad_norm": 0.16823437809944153,
      "learning_rate": 9.789922409298617e-05,
      "loss": 0.9942,
      "mean_token_accuracy": 0.725135650485754,
      "num_tokens": 21897458.0,
      "step": 686
    },
    {
      "entropy": 0.972180750221014,
      "epoch": 0.06320837718011972,
      "grad_norm": 0.16990965604782104,
      "learning_rate": 9.789615726684455e-05,
      "loss": 1.0008,
      "mean_token_accuracy": 0.7290299013257027,
      "num_tokens": 21928749.0,
      "step": 687
    },
    {
      "entropy": 1.0729428641498089,
      "epoch": 0.06330038355156095,
      "grad_norm": 0.16951923072338104,
      "learning_rate": 9.789309044070292e-05,
      "loss": 1.0653,
      "mean_token_accuracy": 0.7079884931445122,
      "num_tokens": 21960656.0,
      "step": 688
    },
    {
      "entropy": 0.9752461276948452,
      "epoch": 0.06339238992300217,
      "grad_norm": 0.16529016196727753,
      "learning_rate": 9.78900236145613e-05,
      "loss": 0.9413,
      "mean_token_accuracy": 0.735041294246912,
      "num_tokens": 21992222.0,
      "step": 689
    },
    {
      "entropy": 0.9720484800636768,
      "epoch": 0.06348439629444338,
      "grad_norm": 0.1628667712211609,
      "learning_rate": 9.788695678841967e-05,
      "loss": 0.9595,
      "mean_token_accuracy": 0.7347554862499237,
      "num_tokens": 22024570.0,
      "step": 690
    },
    {
      "entropy": 0.9574784226715565,
      "epoch": 0.06357640266588462,
      "grad_norm": 0.16064247488975525,
      "learning_rate": 9.788388996227804e-05,
      "loss": 0.9627,
      "mean_token_accuracy": 0.7343390956521034,
      "num_tokens": 22057154.0,
      "step": 691
    },
    {
      "entropy": 0.9737029746174812,
      "epoch": 0.06366840903732583,
      "grad_norm": 0.16917623579502106,
      "learning_rate": 9.788082313613642e-05,
      "loss": 0.982,
      "mean_token_accuracy": 0.7244335189461708,
      "num_tokens": 22088396.0,
      "step": 692
    },
    {
      "entropy": 0.8450344726443291,
      "epoch": 0.06376041540876706,
      "grad_norm": 0.1518240123987198,
      "learning_rate": 9.78777563099948e-05,
      "loss": 0.8658,
      "mean_token_accuracy": 0.7551683560013771,
      "num_tokens": 22120920.0,
      "step": 693
    },
    {
      "entropy": 1.0285722315311432,
      "epoch": 0.06385242178020828,
      "grad_norm": 0.17033198475837708,
      "learning_rate": 9.787468948385317e-05,
      "loss": 1.0254,
      "mean_token_accuracy": 0.7211484871804714,
      "num_tokens": 22152721.0,
      "step": 694
    },
    {
      "entropy": 1.0109255872666836,
      "epoch": 0.0639444281516495,
      "grad_norm": 0.16604195535182953,
      "learning_rate": 9.787162265771154e-05,
      "loss": 1.0217,
      "mean_token_accuracy": 0.7168786190450191,
      "num_tokens": 22185159.0,
      "step": 695
    },
    {
      "entropy": 0.9595106393098831,
      "epoch": 0.06403643452309073,
      "grad_norm": 0.16640381515026093,
      "learning_rate": 9.78685558315699e-05,
      "loss": 0.9702,
      "mean_token_accuracy": 0.7288764454424381,
      "num_tokens": 22217298.0,
      "step": 696
    },
    {
      "entropy": 1.016012068837881,
      "epoch": 0.06412844089453194,
      "grad_norm": 0.16934356093406677,
      "learning_rate": 9.786548900542829e-05,
      "loss": 1.0243,
      "mean_token_accuracy": 0.712494395673275,
      "num_tokens": 22249149.0,
      "step": 697
    },
    {
      "entropy": 0.9197441339492798,
      "epoch": 0.06422044726597317,
      "grad_norm": 0.1601523607969284,
      "learning_rate": 9.786242217928667e-05,
      "loss": 0.9343,
      "mean_token_accuracy": 0.740419827401638,
      "num_tokens": 22281293.0,
      "step": 698
    },
    {
      "entropy": 0.9875156804919243,
      "epoch": 0.06431245363741439,
      "grad_norm": 0.17086534202098846,
      "learning_rate": 9.785935535314503e-05,
      "loss": 0.9692,
      "mean_token_accuracy": 0.7260996103286743,
      "num_tokens": 22312358.0,
      "step": 699
    },
    {
      "entropy": 0.9348448477685452,
      "epoch": 0.06440446000885561,
      "grad_norm": 0.16300342977046967,
      "learning_rate": 9.78562885270034e-05,
      "loss": 0.9197,
      "mean_token_accuracy": 0.7400068640708923,
      "num_tokens": 22344790.0,
      "step": 700
    },
    {
      "entropy": 0.9701763093471527,
      "epoch": 0.06449646638029684,
      "grad_norm": 0.16712148487567902,
      "learning_rate": 9.785322170086178e-05,
      "loss": 0.9917,
      "mean_token_accuracy": 0.7253463603556156,
      "num_tokens": 22376996.0,
      "step": 701
    },
    {
      "entropy": 0.8692387416958809,
      "epoch": 0.06458847275173805,
      "grad_norm": 0.1571662575006485,
      "learning_rate": 9.785015487472015e-05,
      "loss": 0.8697,
      "mean_token_accuracy": 0.7569755800068378,
      "num_tokens": 22409346.0,
      "step": 702
    },
    {
      "entropy": 0.9658809565007687,
      "epoch": 0.06468047912317929,
      "grad_norm": 0.16465206444263458,
      "learning_rate": 9.784708804857853e-05,
      "loss": 0.9717,
      "mean_token_accuracy": 0.7279982939362526,
      "num_tokens": 22441219.0,
      "step": 703
    },
    {
      "entropy": 0.9933779127895832,
      "epoch": 0.0647724854946205,
      "grad_norm": 0.16633573174476624,
      "learning_rate": 9.784402122243691e-05,
      "loss": 1.0104,
      "mean_token_accuracy": 0.7267393246293068,
      "num_tokens": 22473434.0,
      "step": 704
    },
    {
      "entropy": 0.9351341016590595,
      "epoch": 0.06486449186606172,
      "grad_norm": 0.16399052739143372,
      "learning_rate": 9.784095439629528e-05,
      "loss": 0.932,
      "mean_token_accuracy": 0.7443361170589924,
      "num_tokens": 22504731.0,
      "step": 705
    },
    {
      "entropy": 1.0217381305992603,
      "epoch": 0.06495649823750295,
      "grad_norm": 0.1639397293329239,
      "learning_rate": 9.783788757015365e-05,
      "loss": 1.0123,
      "mean_token_accuracy": 0.7212101444602013,
      "num_tokens": 22536556.0,
      "step": 706
    },
    {
      "entropy": 1.0639294534921646,
      "epoch": 0.06504850460894417,
      "grad_norm": 0.17265541851520538,
      "learning_rate": 9.783482074401202e-05,
      "loss": 1.0568,
      "mean_token_accuracy": 0.7043555565178394,
      "num_tokens": 22568306.0,
      "step": 707
    },
    {
      "entropy": 0.8266307599842548,
      "epoch": 0.0651405109803854,
      "grad_norm": 0.15222366154193878,
      "learning_rate": 9.78317539178704e-05,
      "loss": 0.8199,
      "mean_token_accuracy": 0.7694737762212753,
      "num_tokens": 22600890.0,
      "step": 708
    },
    {
      "entropy": 0.8785643726587296,
      "epoch": 0.06523251735182661,
      "grad_norm": 0.1603255420923233,
      "learning_rate": 9.782868709172878e-05,
      "loss": 0.8674,
      "mean_token_accuracy": 0.7533312663435936,
      "num_tokens": 22632668.0,
      "step": 709
    },
    {
      "entropy": 0.9772775769233704,
      "epoch": 0.06532452372326783,
      "grad_norm": 0.16208812594413757,
      "learning_rate": 9.782562026558715e-05,
      "loss": 0.9682,
      "mean_token_accuracy": 0.7358207292854786,
      "num_tokens": 22663928.0,
      "step": 710
    },
    {
      "entropy": 0.9169563576579094,
      "epoch": 0.06541653009470906,
      "grad_norm": 0.16006122529506683,
      "learning_rate": 9.782255343944552e-05,
      "loss": 0.9112,
      "mean_token_accuracy": 0.7430835999548435,
      "num_tokens": 22696082.0,
      "step": 711
    },
    {
      "entropy": 0.9908410981297493,
      "epoch": 0.06550853646615028,
      "grad_norm": 0.1732882261276245,
      "learning_rate": 9.78194866133039e-05,
      "loss": 1.0063,
      "mean_token_accuracy": 0.7236398309469223,
      "num_tokens": 22727854.0,
      "step": 712
    },
    {
      "entropy": 1.027011003345251,
      "epoch": 0.06560054283759151,
      "grad_norm": 0.16827496886253357,
      "learning_rate": 9.781641978716227e-05,
      "loss": 1.0199,
      "mean_token_accuracy": 0.7127940803766251,
      "num_tokens": 22759919.0,
      "step": 713
    },
    {
      "entropy": 0.9439230598509312,
      "epoch": 0.06569254920903272,
      "grad_norm": 0.1674676239490509,
      "learning_rate": 9.781335296102065e-05,
      "loss": 0.9612,
      "mean_token_accuracy": 0.7319885641336441,
      "num_tokens": 22791554.0,
      "step": 714
    },
    {
      "entropy": 0.952055312693119,
      "epoch": 0.06578455558047394,
      "grad_norm": 0.1640789806842804,
      "learning_rate": 9.781028613487902e-05,
      "loss": 0.9714,
      "mean_token_accuracy": 0.7325408682227135,
      "num_tokens": 22823959.0,
      "step": 715
    },
    {
      "entropy": 0.9567240849137306,
      "epoch": 0.06587656195191517,
      "grad_norm": 0.16402341425418854,
      "learning_rate": 9.78072193087374e-05,
      "loss": 0.9638,
      "mean_token_accuracy": 0.732719574123621,
      "num_tokens": 22855853.0,
      "step": 716
    },
    {
      "entropy": 0.9778101556003094,
      "epoch": 0.06596856832335639,
      "grad_norm": 0.16831116378307343,
      "learning_rate": 9.780415248259576e-05,
      "loss": 0.9808,
      "mean_token_accuracy": 0.7261964827775955,
      "num_tokens": 22886943.0,
      "step": 717
    },
    {
      "entropy": 1.0320146083831787,
      "epoch": 0.06606057469479762,
      "grad_norm": 0.16433413326740265,
      "learning_rate": 9.780108565645413e-05,
      "loss": 1.0498,
      "mean_token_accuracy": 0.7127593494951725,
      "num_tokens": 22918589.0,
      "step": 718
    },
    {
      "entropy": 1.0451893992722034,
      "epoch": 0.06615258106623884,
      "grad_norm": 0.16586624085903168,
      "learning_rate": 9.779801883031251e-05,
      "loss": 1.0403,
      "mean_token_accuracy": 0.7101616859436035,
      "num_tokens": 22950549.0,
      "step": 719
    },
    {
      "entropy": 1.0089669600129128,
      "epoch": 0.06624458743768005,
      "grad_norm": 0.16412770748138428,
      "learning_rate": 9.77949520041709e-05,
      "loss": 1.0175,
      "mean_token_accuracy": 0.7171782068908215,
      "num_tokens": 22982525.0,
      "step": 720
    },
    {
      "entropy": 0.9115235470235348,
      "epoch": 0.06633659380912128,
      "grad_norm": 0.1757228672504425,
      "learning_rate": 9.779188517802926e-05,
      "loss": 0.9025,
      "mean_token_accuracy": 0.7461497895419598,
      "num_tokens": 23015078.0,
      "step": 721
    },
    {
      "entropy": 0.956471286714077,
      "epoch": 0.0664286001805625,
      "grad_norm": 0.15770560503005981,
      "learning_rate": 9.778881835188763e-05,
      "loss": 0.9436,
      "mean_token_accuracy": 0.7347383312880993,
      "num_tokens": 23046758.0,
      "step": 722
    },
    {
      "entropy": 0.8515592478215694,
      "epoch": 0.06652060655200373,
      "grad_norm": 0.15324951708316803,
      "learning_rate": 9.7785751525746e-05,
      "loss": 0.8435,
      "mean_token_accuracy": 0.7604524195194244,
      "num_tokens": 23078838.0,
      "step": 723
    },
    {
      "entropy": 1.0155523493885994,
      "epoch": 0.06661261292344495,
      "grad_norm": 0.16911432147026062,
      "learning_rate": 9.77826846996044e-05,
      "loss": 1.0325,
      "mean_token_accuracy": 0.7149862349033356,
      "num_tokens": 23110895.0,
      "step": 724
    },
    {
      "entropy": 1.0014025121927261,
      "epoch": 0.06670461929488618,
      "grad_norm": 0.16532562673091888,
      "learning_rate": 9.777961787346276e-05,
      "loss": 1.0086,
      "mean_token_accuracy": 0.7236623056232929,
      "num_tokens": 23142033.0,
      "step": 725
    },
    {
      "entropy": 0.9258328080177307,
      "epoch": 0.0667966256663274,
      "grad_norm": 0.1567230522632599,
      "learning_rate": 9.777655104732113e-05,
      "loss": 0.9368,
      "mean_token_accuracy": 0.739589300006628,
      "num_tokens": 23174203.0,
      "step": 726
    },
    {
      "entropy": 1.056148249655962,
      "epoch": 0.06688863203776861,
      "grad_norm": 0.17096057534217834,
      "learning_rate": 9.77734842211795e-05,
      "loss": 1.0844,
      "mean_token_accuracy": 0.7052101269364357,
      "num_tokens": 23206749.0,
      "step": 727
    },
    {
      "entropy": 0.9294457361102104,
      "epoch": 0.06698063840920984,
      "grad_norm": 0.15669047832489014,
      "learning_rate": 9.777041739503788e-05,
      "loss": 0.9125,
      "mean_token_accuracy": 0.7409942634403706,
      "num_tokens": 23238825.0,
      "step": 728
    },
    {
      "entropy": 0.9112744592130184,
      "epoch": 0.06707264478065106,
      "grad_norm": 0.16031840443611145,
      "learning_rate": 9.776735056889626e-05,
      "loss": 0.9211,
      "mean_token_accuracy": 0.7456766106188297,
      "num_tokens": 23270934.0,
      "step": 729
    },
    {
      "entropy": 0.8610086254775524,
      "epoch": 0.06716465115209229,
      "grad_norm": 0.16114585101604462,
      "learning_rate": 9.776428374275463e-05,
      "loss": 0.8759,
      "mean_token_accuracy": 0.7515043914318085,
      "num_tokens": 23303164.0,
      "step": 730
    },
    {
      "entropy": 0.9923807885497808,
      "epoch": 0.0672566575235335,
      "grad_norm": 0.17049486935138702,
      "learning_rate": 9.7761216916613e-05,
      "loss": 1.0091,
      "mean_token_accuracy": 0.7261724770069122,
      "num_tokens": 23335126.0,
      "step": 731
    },
    {
      "entropy": 1.0289140231907368,
      "epoch": 0.06734866389497472,
      "grad_norm": 0.17657940089702606,
      "learning_rate": 9.775815009047138e-05,
      "loss": 1.0175,
      "mean_token_accuracy": 0.7188629806041718,
      "num_tokens": 23365955.0,
      "step": 732
    },
    {
      "entropy": 0.9260178916156292,
      "epoch": 0.06744067026641595,
      "grad_norm": 0.16398519277572632,
      "learning_rate": 9.775508326432975e-05,
      "loss": 0.9394,
      "mean_token_accuracy": 0.740141972899437,
      "num_tokens": 23397892.0,
      "step": 733
    },
    {
      "entropy": 0.9722410663962364,
      "epoch": 0.06753267663785717,
      "grad_norm": 0.1668296903371811,
      "learning_rate": 9.775201643818813e-05,
      "loss": 0.9908,
      "mean_token_accuracy": 0.731209821999073,
      "num_tokens": 23429804.0,
      "step": 734
    },
    {
      "entropy": 0.9129518233239651,
      "epoch": 0.0676246830092984,
      "grad_norm": 0.16129791736602783,
      "learning_rate": 9.774894961204651e-05,
      "loss": 0.9065,
      "mean_token_accuracy": 0.7429094351828098,
      "num_tokens": 23461042.0,
      "step": 735
    },
    {
      "entropy": 1.0367199257016182,
      "epoch": 0.06771668938073962,
      "grad_norm": 0.15939047932624817,
      "learning_rate": 9.774588278590488e-05,
      "loss": 1.0261,
      "mean_token_accuracy": 0.7220759354531765,
      "num_tokens": 23492865.0,
      "step": 736
    },
    {
      "entropy": 0.9758146665990353,
      "epoch": 0.06780869575218083,
      "grad_norm": 0.16275274753570557,
      "learning_rate": 9.774281595976324e-05,
      "loss": 0.9657,
      "mean_token_accuracy": 0.730737242847681,
      "num_tokens": 23524670.0,
      "step": 737
    },
    {
      "entropy": 0.9374345373362303,
      "epoch": 0.06790070212362206,
      "grad_norm": 0.16051124036312103,
      "learning_rate": 9.773974913362161e-05,
      "loss": 0.9341,
      "mean_token_accuracy": 0.7444140948355198,
      "num_tokens": 23557259.0,
      "step": 738
    },
    {
      "entropy": 1.0072547420859337,
      "epoch": 0.06799270849506328,
      "grad_norm": 0.16542603075504303,
      "learning_rate": 9.773668230748e-05,
      "loss": 1.0278,
      "mean_token_accuracy": 0.7148850113153458,
      "num_tokens": 23587987.0,
      "step": 739
    },
    {
      "entropy": 1.0053394995629787,
      "epoch": 0.06808471486650451,
      "grad_norm": 0.17000436782836914,
      "learning_rate": 9.773361548133837e-05,
      "loss": 1.0117,
      "mean_token_accuracy": 0.7174012921750546,
      "num_tokens": 23619095.0,
      "step": 740
    },
    {
      "entropy": 0.9260540809482336,
      "epoch": 0.06817672123794573,
      "grad_norm": 0.15828430652618408,
      "learning_rate": 9.773054865519674e-05,
      "loss": 0.9349,
      "mean_token_accuracy": 0.7401717789471149,
      "num_tokens": 23650640.0,
      "step": 741
    },
    {
      "entropy": 0.9547841921448708,
      "epoch": 0.06826872760938694,
      "grad_norm": 0.1632172018289566,
      "learning_rate": 9.772748182905511e-05,
      "loss": 0.9439,
      "mean_token_accuracy": 0.7397718019783497,
      "num_tokens": 23683159.0,
      "step": 742
    },
    {
      "entropy": 0.9822579361498356,
      "epoch": 0.06836073398082818,
      "grad_norm": 0.1617399901151657,
      "learning_rate": 9.772441500291349e-05,
      "loss": 0.9715,
      "mean_token_accuracy": 0.7286654934287071,
      "num_tokens": 23715371.0,
      "step": 743
    },
    {
      "entropy": 0.9397387616336346,
      "epoch": 0.06845274035226939,
      "grad_norm": 0.16105400025844574,
      "learning_rate": 9.772134817677186e-05,
      "loss": 0.9313,
      "mean_token_accuracy": 0.7390888780355453,
      "num_tokens": 23746734.0,
      "step": 744
    },
    {
      "entropy": 1.0103630274534225,
      "epoch": 0.06854474672371062,
      "grad_norm": 0.16748052835464478,
      "learning_rate": 9.771828135063024e-05,
      "loss": 1.0123,
      "mean_token_accuracy": 0.720540527254343,
      "num_tokens": 23778568.0,
      "step": 745
    },
    {
      "entropy": 0.8300569839775562,
      "epoch": 0.06863675309515184,
      "grad_norm": 0.1525866687297821,
      "learning_rate": 9.771521452448861e-05,
      "loss": 0.8154,
      "mean_token_accuracy": 0.7673349007964134,
      "num_tokens": 23810087.0,
      "step": 746
    },
    {
      "entropy": 0.9432720765471458,
      "epoch": 0.06872875946659306,
      "grad_norm": 0.15817712247371674,
      "learning_rate": 9.771214769834699e-05,
      "loss": 0.9445,
      "mean_token_accuracy": 0.7349588461220264,
      "num_tokens": 23842181.0,
      "step": 747
    },
    {
      "entropy": 0.9867074005305767,
      "epoch": 0.06882076583803429,
      "grad_norm": 0.17190751433372498,
      "learning_rate": 9.770908087220536e-05,
      "loss": 1.0277,
      "mean_token_accuracy": 0.7208070978522301,
      "num_tokens": 23873738.0,
      "step": 748
    },
    {
      "entropy": 0.9327075630426407,
      "epoch": 0.0689127722094755,
      "grad_norm": 0.17223809659481049,
      "learning_rate": 9.770601404606373e-05,
      "loss": 0.9525,
      "mean_token_accuracy": 0.7346966341137886,
      "num_tokens": 23905432.0,
      "step": 749
    },
    {
      "entropy": 0.9508248716592789,
      "epoch": 0.06900477858091673,
      "grad_norm": 0.18050900101661682,
      "learning_rate": 9.770294721992211e-05,
      "loss": 0.9502,
      "mean_token_accuracy": 0.7313392050564289,
      "num_tokens": 23936465.0,
      "step": 750
    },
    {
      "entropy": 1.0264729671180248,
      "epoch": 0.06909678495235795,
      "grad_norm": 0.1730281561613083,
      "learning_rate": 9.769988039378049e-05,
      "loss": 1.0622,
      "mean_token_accuracy": 0.7115480825304985,
      "num_tokens": 23968162.0,
      "step": 751
    },
    {
      "entropy": 0.9294424448162317,
      "epoch": 0.06918879132379917,
      "grad_norm": 0.159836545586586,
      "learning_rate": 9.769681356763886e-05,
      "loss": 0.9337,
      "mean_token_accuracy": 0.7404293045401573,
      "num_tokens": 24000250.0,
      "step": 752
    },
    {
      "entropy": 0.9844078049063683,
      "epoch": 0.0692807976952404,
      "grad_norm": 0.1691402941942215,
      "learning_rate": 9.769374674149722e-05,
      "loss": 0.9953,
      "mean_token_accuracy": 0.7197633534669876,
      "num_tokens": 24031949.0,
      "step": 753
    },
    {
      "entropy": 0.9914673715829849,
      "epoch": 0.06937280406668161,
      "grad_norm": 0.1598096340894699,
      "learning_rate": 9.769067991535559e-05,
      "loss": 0.9812,
      "mean_token_accuracy": 0.7298092544078827,
      "num_tokens": 24063321.0,
      "step": 754
    },
    {
      "entropy": 0.9517121911048889,
      "epoch": 0.06946481043812285,
      "grad_norm": 0.16180214285850525,
      "learning_rate": 9.768761308921397e-05,
      "loss": 0.9351,
      "mean_token_accuracy": 0.735908892005682,
      "num_tokens": 24095220.0,
      "step": 755
    },
    {
      "entropy": 1.0087159909307957,
      "epoch": 0.06955681680956406,
      "grad_norm": 0.1683640480041504,
      "learning_rate": 9.768454626307236e-05,
      "loss": 1.0,
      "mean_token_accuracy": 0.7229274809360504,
      "num_tokens": 24127139.0,
      "step": 756
    },
    {
      "entropy": 0.8741592951118946,
      "epoch": 0.06964882318100528,
      "grad_norm": 0.15390855073928833,
      "learning_rate": 9.768147943693072e-05,
      "loss": 0.8724,
      "mean_token_accuracy": 0.7545308023691177,
      "num_tokens": 24159776.0,
      "step": 757
    },
    {
      "entropy": 0.9977220296859741,
      "epoch": 0.06974082955244651,
      "grad_norm": 0.16438770294189453,
      "learning_rate": 9.767841261078909e-05,
      "loss": 1.0023,
      "mean_token_accuracy": 0.7199131026864052,
      "num_tokens": 24191352.0,
      "step": 758
    },
    {
      "entropy": 0.9235007651150227,
      "epoch": 0.06983283592388773,
      "grad_norm": 0.2680729925632477,
      "learning_rate": 9.767534578464747e-05,
      "loss": 0.922,
      "mean_token_accuracy": 0.7445998787879944,
      "num_tokens": 24222978.0,
      "step": 759
    },
    {
      "entropy": 0.9299619011580944,
      "epoch": 0.06992484229532896,
      "grad_norm": 0.16164758801460266,
      "learning_rate": 9.767227895850585e-05,
      "loss": 0.9483,
      "mean_token_accuracy": 0.7354999706149101,
      "num_tokens": 24255342.0,
      "step": 760
    },
    {
      "entropy": 0.917940890416503,
      "epoch": 0.07001684866677017,
      "grad_norm": 0.1616727113723755,
      "learning_rate": 9.766921213236422e-05,
      "loss": 0.9179,
      "mean_token_accuracy": 0.7413064204156399,
      "num_tokens": 24287900.0,
      "step": 761
    },
    {
      "entropy": 1.053685488179326,
      "epoch": 0.07010885503821139,
      "grad_norm": 0.1635410040616989,
      "learning_rate": 9.766614530622259e-05,
      "loss": 1.0446,
      "mean_token_accuracy": 0.7076585590839386,
      "num_tokens": 24320219.0,
      "step": 762
    },
    {
      "entropy": 0.8969144448637962,
      "epoch": 0.07020086140965262,
      "grad_norm": 0.15589921176433563,
      "learning_rate": 9.766307848008097e-05,
      "loss": 0.8934,
      "mean_token_accuracy": 0.7494642175734043,
      "num_tokens": 24352930.0,
      "step": 763
    },
    {
      "entropy": 0.9762221947312355,
      "epoch": 0.07029286778109384,
      "grad_norm": 0.17016766965389252,
      "learning_rate": 9.766001165393934e-05,
      "loss": 0.9911,
      "mean_token_accuracy": 0.7291914746165276,
      "num_tokens": 24384128.0,
      "step": 764
    },
    {
      "entropy": 0.9689153954386711,
      "epoch": 0.07038487415253507,
      "grad_norm": 0.16347086429595947,
      "learning_rate": 9.765694482779772e-05,
      "loss": 0.9538,
      "mean_token_accuracy": 0.7400826029479504,
      "num_tokens": 24414983.0,
      "step": 765
    },
    {
      "entropy": 0.9211628139019012,
      "epoch": 0.07047688052397628,
      "grad_norm": 0.16349484026432037,
      "learning_rate": 9.76538780016561e-05,
      "loss": 0.9221,
      "mean_token_accuracy": 0.7421093508601189,
      "num_tokens": 24447312.0,
      "step": 766
    },
    {
      "entropy": 0.9753438048064709,
      "epoch": 0.0705688868954175,
      "grad_norm": 0.16988539695739746,
      "learning_rate": 9.765081117551447e-05,
      "loss": 1.0007,
      "mean_token_accuracy": 0.7239793911576271,
      "num_tokens": 24479148.0,
      "step": 767
    },
    {
      "entropy": 1.002135954797268,
      "epoch": 0.07066089326685873,
      "grad_norm": 0.1716102957725525,
      "learning_rate": 9.764774434937284e-05,
      "loss": 1.0209,
      "mean_token_accuracy": 0.715527381747961,
      "num_tokens": 24510263.0,
      "step": 768
    },
    {
      "entropy": 0.9307844415307045,
      "epoch": 0.07075289963829995,
      "grad_norm": 0.16261951625347137,
      "learning_rate": 9.76446775232312e-05,
      "loss": 0.9392,
      "mean_token_accuracy": 0.7415165230631828,
      "num_tokens": 24542841.0,
      "step": 769
    },
    {
      "entropy": 1.0119842812418938,
      "epoch": 0.07084490600974118,
      "grad_norm": 0.16798041760921478,
      "learning_rate": 9.764161069708959e-05,
      "loss": 1.0295,
      "mean_token_accuracy": 0.7190869897603989,
      "num_tokens": 24573839.0,
      "step": 770
    },
    {
      "entropy": 0.8963761664927006,
      "epoch": 0.0709369123811824,
      "grad_norm": 0.16296271979808807,
      "learning_rate": 9.763854387094797e-05,
      "loss": 0.8813,
      "mean_token_accuracy": 0.75478770211339,
      "num_tokens": 24606174.0,
      "step": 771
    },
    {
      "entropy": 0.9868187941610813,
      "epoch": 0.07102891875262361,
      "grad_norm": 0.16351190209388733,
      "learning_rate": 9.763547704480634e-05,
      "loss": 0.9873,
      "mean_token_accuracy": 0.723734874278307,
      "num_tokens": 24637787.0,
      "step": 772
    },
    {
      "entropy": 0.973717924207449,
      "epoch": 0.07112092512406484,
      "grad_norm": 0.16125066578388214,
      "learning_rate": 9.76324102186647e-05,
      "loss": 0.9658,
      "mean_token_accuracy": 0.7296981699764729,
      "num_tokens": 24669449.0,
      "step": 773
    },
    {
      "entropy": 0.9297757185995579,
      "epoch": 0.07121293149550606,
      "grad_norm": 0.16825729608535767,
      "learning_rate": 9.762934339252309e-05,
      "loss": 0.9549,
      "mean_token_accuracy": 0.7393349185585976,
      "num_tokens": 24701453.0,
      "step": 774
    },
    {
      "entropy": 0.9122928343713284,
      "epoch": 0.07130493786694729,
      "grad_norm": 0.15279382467269897,
      "learning_rate": 9.762627656638145e-05,
      "loss": 0.9132,
      "mean_token_accuracy": 0.744269885122776,
      "num_tokens": 24733802.0,
      "step": 775
    },
    {
      "entropy": 0.9462030939757824,
      "epoch": 0.07139694423838851,
      "grad_norm": 0.16476961970329285,
      "learning_rate": 9.762320974023983e-05,
      "loss": 0.9397,
      "mean_token_accuracy": 0.7325603887438774,
      "num_tokens": 24765705.0,
      "step": 776
    },
    {
      "entropy": 0.9936528950929642,
      "epoch": 0.07148895060982972,
      "grad_norm": 0.16848473250865936,
      "learning_rate": 9.76201429140982e-05,
      "loss": 1.0133,
      "mean_token_accuracy": 0.7192382104694843,
      "num_tokens": 24796858.0,
      "step": 777
    },
    {
      "entropy": 0.9840080440044403,
      "epoch": 0.07158095698127095,
      "grad_norm": 0.16142773628234863,
      "learning_rate": 9.761707608795658e-05,
      "loss": 1.0047,
      "mean_token_accuracy": 0.7264664433896542,
      "num_tokens": 24829090.0,
      "step": 778
    },
    {
      "entropy": 0.8958589006215334,
      "epoch": 0.07167296335271217,
      "grad_norm": 0.1665801852941513,
      "learning_rate": 9.761400926181495e-05,
      "loss": 0.9098,
      "mean_token_accuracy": 0.7445541583001614,
      "num_tokens": 24860926.0,
      "step": 779
    },
    {
      "entropy": 0.891968484967947,
      "epoch": 0.0717649697241534,
      "grad_norm": 0.15826746821403503,
      "learning_rate": 9.761094243567332e-05,
      "loss": 0.8884,
      "mean_token_accuracy": 0.7500179074704647,
      "num_tokens": 24893343.0,
      "step": 780
    },
    {
      "entropy": 0.9690477028489113,
      "epoch": 0.07185697609559462,
      "grad_norm": 0.167481929063797,
      "learning_rate": 9.76078756095317e-05,
      "loss": 0.9725,
      "mean_token_accuracy": 0.7298942096531391,
      "num_tokens": 24925334.0,
      "step": 781
    },
    {
      "entropy": 0.938049141317606,
      "epoch": 0.07194898246703585,
      "grad_norm": 0.16644228994846344,
      "learning_rate": 9.760480878339008e-05,
      "loss": 0.9341,
      "mean_token_accuracy": 0.7358887791633606,
      "num_tokens": 24956938.0,
      "step": 782
    },
    {
      "entropy": 0.9837486334145069,
      "epoch": 0.07204098883847707,
      "grad_norm": 0.16411328315734863,
      "learning_rate": 9.760174195724845e-05,
      "loss": 0.97,
      "mean_token_accuracy": 0.7262499928474426,
      "num_tokens": 24988783.0,
      "step": 783
    },
    {
      "entropy": 1.0175521448254585,
      "epoch": 0.07213299520991828,
      "grad_norm": 0.1714836210012436,
      "learning_rate": 9.759867513110682e-05,
      "loss": 1.0069,
      "mean_token_accuracy": 0.7224933691322803,
      "num_tokens": 25018786.0,
      "step": 784
    },
    {
      "entropy": 1.0210258662700653,
      "epoch": 0.07222500158135951,
      "grad_norm": 0.17129401862621307,
      "learning_rate": 9.759560830496519e-05,
      "loss": 1.0408,
      "mean_token_accuracy": 0.7160746045410633,
      "num_tokens": 25051181.0,
      "step": 785
    },
    {
      "entropy": 0.8809431008994579,
      "epoch": 0.07231700795280073,
      "grad_norm": 0.15995007753372192,
      "learning_rate": 9.759254147882357e-05,
      "loss": 0.8921,
      "mean_token_accuracy": 0.7458332069218159,
      "num_tokens": 25082231.0,
      "step": 786
    },
    {
      "entropy": 0.8314573392271996,
      "epoch": 0.07240901432424196,
      "grad_norm": 0.15146297216415405,
      "learning_rate": 9.758947465268195e-05,
      "loss": 0.821,
      "mean_token_accuracy": 0.7656078785657883,
      "num_tokens": 25114469.0,
      "step": 787
    },
    {
      "entropy": 0.9689861610531807,
      "epoch": 0.07250102069568318,
      "grad_norm": 0.16891618072986603,
      "learning_rate": 9.758640782654032e-05,
      "loss": 0.9748,
      "mean_token_accuracy": 0.7321065627038479,
      "num_tokens": 25145649.0,
      "step": 788
    },
    {
      "entropy": 0.929735329002142,
      "epoch": 0.0725930270671244,
      "grad_norm": 0.15692980587482452,
      "learning_rate": 9.758334100039868e-05,
      "loss": 0.9344,
      "mean_token_accuracy": 0.7373642511665821,
      "num_tokens": 25177567.0,
      "step": 789
    },
    {
      "entropy": 0.9218393638730049,
      "epoch": 0.07268503343856562,
      "grad_norm": 0.15871797502040863,
      "learning_rate": 9.758027417425707e-05,
      "loss": 0.9307,
      "mean_token_accuracy": 0.7394563555717468,
      "num_tokens": 25209745.0,
      "step": 790
    },
    {
      "entropy": 0.9620450288057327,
      "epoch": 0.07277703981000684,
      "grad_norm": 0.15939058363437653,
      "learning_rate": 9.757720734811543e-05,
      "loss": 0.9808,
      "mean_token_accuracy": 0.7276261821389198,
      "num_tokens": 25242051.0,
      "step": 791
    },
    {
      "entropy": 0.8812330979853868,
      "epoch": 0.07286904618144807,
      "grad_norm": 0.16241149604320526,
      "learning_rate": 9.757414052197382e-05,
      "loss": 0.9099,
      "mean_token_accuracy": 0.746483813971281,
      "num_tokens": 25274482.0,
      "step": 792
    },
    {
      "entropy": 0.9856306165456772,
      "epoch": 0.07296105255288929,
      "grad_norm": 0.16723030805587769,
      "learning_rate": 9.757107369583218e-05,
      "loss": 0.9668,
      "mean_token_accuracy": 0.729243416339159,
      "num_tokens": 25306128.0,
      "step": 793
    },
    {
      "entropy": 0.9258338510990143,
      "epoch": 0.0730530589243305,
      "grad_norm": 0.1538238823413849,
      "learning_rate": 9.756800686969056e-05,
      "loss": 0.9007,
      "mean_token_accuracy": 0.7487154603004456,
      "num_tokens": 25337734.0,
      "step": 794
    },
    {
      "entropy": 0.9189766198396683,
      "epoch": 0.07314506529577174,
      "grad_norm": 0.15357853472232819,
      "learning_rate": 9.756494004354893e-05,
      "loss": 0.922,
      "mean_token_accuracy": 0.7440185956656933,
      "num_tokens": 25370143.0,
      "step": 795
    },
    {
      "entropy": 1.0439080595970154,
      "epoch": 0.07323707166721295,
      "grad_norm": 0.16735073924064636,
      "learning_rate": 9.75618732174073e-05,
      "loss": 1.0449,
      "mean_token_accuracy": 0.7094504870474339,
      "num_tokens": 25402041.0,
      "step": 796
    },
    {
      "entropy": 0.9542340748012066,
      "epoch": 0.07332907803865418,
      "grad_norm": 0.1582360565662384,
      "learning_rate": 9.75588063912657e-05,
      "loss": 0.9347,
      "mean_token_accuracy": 0.7360214106738567,
      "num_tokens": 25434049.0,
      "step": 797
    },
    {
      "entropy": 0.9874455407261848,
      "epoch": 0.0734210844100954,
      "grad_norm": 0.16485659778118134,
      "learning_rate": 9.755573956512406e-05,
      "loss": 1.0005,
      "mean_token_accuracy": 0.7200411334633827,
      "num_tokens": 25464641.0,
      "step": 798
    },
    {
      "entropy": 1.003897998481989,
      "epoch": 0.07351309078153662,
      "grad_norm": 0.17197616398334503,
      "learning_rate": 9.755267273898243e-05,
      "loss": 1.006,
      "mean_token_accuracy": 0.7257940694689751,
      "num_tokens": 25496483.0,
      "step": 799
    },
    {
      "entropy": 0.879679124802351,
      "epoch": 0.07360509715297785,
      "grad_norm": 0.15795229375362396,
      "learning_rate": 9.75496059128408e-05,
      "loss": 0.8622,
      "mean_token_accuracy": 0.7517740428447723,
      "num_tokens": 25528055.0,
      "step": 800
    },
    {
      "entropy": 1.046953234821558,
      "epoch": 0.07369710352441906,
      "grad_norm": 0.1727726310491562,
      "learning_rate": 9.754653908669918e-05,
      "loss": 1.0626,
      "mean_token_accuracy": 0.7050612010061741,
      "num_tokens": 25560089.0,
      "step": 801
    },
    {
      "entropy": 0.9426787309348583,
      "epoch": 0.0737891098958603,
      "grad_norm": 0.163871631026268,
      "learning_rate": 9.754347226055756e-05,
      "loss": 0.9619,
      "mean_token_accuracy": 0.7296958267688751,
      "num_tokens": 25592304.0,
      "step": 802
    },
    {
      "entropy": 0.8039157781749964,
      "epoch": 0.07388111626730151,
      "grad_norm": 0.15506954491138458,
      "learning_rate": 9.754040543441593e-05,
      "loss": 0.8373,
      "mean_token_accuracy": 0.7683628164231777,
      "num_tokens": 25624182.0,
      "step": 803
    },
    {
      "entropy": 1.0090614184737206,
      "epoch": 0.07397312263874273,
      "grad_norm": 0.16896405816078186,
      "learning_rate": 9.75373386082743e-05,
      "loss": 1.0282,
      "mean_token_accuracy": 0.7205565087497234,
      "num_tokens": 25655860.0,
      "step": 804
    },
    {
      "entropy": 0.9871361739933491,
      "epoch": 0.07406512901018396,
      "grad_norm": 0.1660793572664261,
      "learning_rate": 9.753427178213268e-05,
      "loss": 0.9809,
      "mean_token_accuracy": 0.719932209700346,
      "num_tokens": 25687256.0,
      "step": 805
    },
    {
      "entropy": 0.8788506984710693,
      "epoch": 0.07415713538162517,
      "grad_norm": 0.15512098371982574,
      "learning_rate": 9.753120495599105e-05,
      "loss": 0.8932,
      "mean_token_accuracy": 0.7470815107226372,
      "num_tokens": 25719872.0,
      "step": 806
    },
    {
      "entropy": 0.9695808105170727,
      "epoch": 0.0742491417530664,
      "grad_norm": 0.17574496567249298,
      "learning_rate": 9.752813812984943e-05,
      "loss": 0.9827,
      "mean_token_accuracy": 0.7296533323824406,
      "num_tokens": 25751797.0,
      "step": 807
    },
    {
      "entropy": 1.0101498495787382,
      "epoch": 0.07434114812450762,
      "grad_norm": 0.16445888578891754,
      "learning_rate": 9.75250713037078e-05,
      "loss": 1.0148,
      "mean_token_accuracy": 0.7191212624311447,
      "num_tokens": 25783625.0,
      "step": 808
    },
    {
      "entropy": 1.1196264550089836,
      "epoch": 0.07443315449594884,
      "grad_norm": 0.17525547742843628,
      "learning_rate": 9.752200447756618e-05,
      "loss": 1.103,
      "mean_token_accuracy": 0.6990806236863136,
      "num_tokens": 25814980.0,
      "step": 809
    },
    {
      "entropy": 0.9786029830574989,
      "epoch": 0.07452516086739007,
      "grad_norm": 0.16183589398860931,
      "learning_rate": 9.751893765142455e-05,
      "loss": 0.9504,
      "mean_token_accuracy": 0.731063287705183,
      "num_tokens": 25846662.0,
      "step": 810
    },
    {
      "entropy": 0.8960175663232803,
      "epoch": 0.07461716723883129,
      "grad_norm": 0.15868550539016724,
      "learning_rate": 9.751587082528291e-05,
      "loss": 0.871,
      "mean_token_accuracy": 0.754888791590929,
      "num_tokens": 25878019.0,
      "step": 811
    },
    {
      "entropy": 1.0339534543454647,
      "epoch": 0.07470917361027252,
      "grad_norm": 0.17218108475208282,
      "learning_rate": 9.75128039991413e-05,
      "loss": 1.0332,
      "mean_token_accuracy": 0.7134471423923969,
      "num_tokens": 25910136.0,
      "step": 812
    },
    {
      "entropy": 0.8484597615897655,
      "epoch": 0.07480117998171373,
      "grad_norm": 0.1623019427061081,
      "learning_rate": 9.750973717299968e-05,
      "loss": 0.8753,
      "mean_token_accuracy": 0.7547342143952847,
      "num_tokens": 25942064.0,
      "step": 813
    },
    {
      "entropy": 1.0285358019173145,
      "epoch": 0.07489318635315495,
      "grad_norm": 0.16945849359035492,
      "learning_rate": 9.750667034685804e-05,
      "loss": 1.0369,
      "mean_token_accuracy": 0.7142699807882309,
      "num_tokens": 25973691.0,
      "step": 814
    },
    {
      "entropy": 0.8631247319281101,
      "epoch": 0.07498519272459618,
      "grad_norm": 0.15859360992908478,
      "learning_rate": 9.750360352071641e-05,
      "loss": 0.8623,
      "mean_token_accuracy": 0.755243431776762,
      "num_tokens": 26005745.0,
      "step": 815
    },
    {
      "entropy": 0.897196713835001,
      "epoch": 0.0750771990960374,
      "grad_norm": 0.1596507728099823,
      "learning_rate": 9.750053669457478e-05,
      "loss": 0.9157,
      "mean_token_accuracy": 0.7448350861668587,
      "num_tokens": 26037786.0,
      "step": 816
    },
    {
      "entropy": 0.9098497070372105,
      "epoch": 0.07516920546747863,
      "grad_norm": 0.16219158470630646,
      "learning_rate": 9.749746986843316e-05,
      "loss": 0.9153,
      "mean_token_accuracy": 0.7434411533176899,
      "num_tokens": 26068966.0,
      "step": 817
    },
    {
      "entropy": 0.9070624820888042,
      "epoch": 0.07526121183891984,
      "grad_norm": 0.1559060662984848,
      "learning_rate": 9.749440304229154e-05,
      "loss": 0.9053,
      "mean_token_accuracy": 0.7430212162435055,
      "num_tokens": 26101312.0,
      "step": 818
    },
    {
      "entropy": 0.9519104026257992,
      "epoch": 0.07535321821036106,
      "grad_norm": 0.1593141108751297,
      "learning_rate": 9.749133621614991e-05,
      "loss": 0.9508,
      "mean_token_accuracy": 0.7314362488687038,
      "num_tokens": 26134080.0,
      "step": 819
    },
    {
      "entropy": 0.9597192704677582,
      "epoch": 0.07544522458180229,
      "grad_norm": 0.16427716612815857,
      "learning_rate": 9.748826939000828e-05,
      "loss": 0.9715,
      "mean_token_accuracy": 0.7315452955663204,
      "num_tokens": 26165992.0,
      "step": 820
    },
    {
      "entropy": 0.9430848769843578,
      "epoch": 0.07553723095324351,
      "grad_norm": 0.1697760671377182,
      "learning_rate": 9.748520256386666e-05,
      "loss": 0.9578,
      "mean_token_accuracy": 0.7328658252954483,
      "num_tokens": 26197742.0,
      "step": 821
    },
    {
      "entropy": 1.0156432874500751,
      "epoch": 0.07562923732468474,
      "grad_norm": 0.16890943050384521,
      "learning_rate": 9.748213573772503e-05,
      "loss": 1.035,
      "mean_token_accuracy": 0.7160553000867367,
      "num_tokens": 26229654.0,
      "step": 822
    },
    {
      "entropy": 0.9517482109367847,
      "epoch": 0.07572124369612596,
      "grad_norm": 0.15917159616947174,
      "learning_rate": 9.747906891158341e-05,
      "loss": 0.9505,
      "mean_token_accuracy": 0.7384414523839951,
      "num_tokens": 26262173.0,
      "step": 823
    },
    {
      "entropy": 0.8981260899454355,
      "epoch": 0.07581325006756717,
      "grad_norm": 0.15398278832435608,
      "learning_rate": 9.747600208544178e-05,
      "loss": 0.8786,
      "mean_token_accuracy": 0.7535396926105022,
      "num_tokens": 26294574.0,
      "step": 824
    },
    {
      "entropy": 1.0011459030210972,
      "epoch": 0.0759052564390084,
      "grad_norm": 0.1741645783185959,
      "learning_rate": 9.747293525930016e-05,
      "loss": 0.9999,
      "mean_token_accuracy": 0.7213692516088486,
      "num_tokens": 26326084.0,
      "step": 825
    },
    {
      "entropy": 0.9742907732725143,
      "epoch": 0.07599726281044962,
      "grad_norm": 0.17058828473091125,
      "learning_rate": 9.746986843315853e-05,
      "loss": 0.9806,
      "mean_token_accuracy": 0.726009164005518,
      "num_tokens": 26357870.0,
      "step": 826
    },
    {
      "entropy": 0.9268339164555073,
      "epoch": 0.07608926918189085,
      "grad_norm": 0.1637573540210724,
      "learning_rate": 9.74668016070169e-05,
      "loss": 0.9429,
      "mean_token_accuracy": 0.7373054288327694,
      "num_tokens": 26390107.0,
      "step": 827
    },
    {
      "entropy": 0.9235973954200745,
      "epoch": 0.07618127555333207,
      "grad_norm": 0.16202713549137115,
      "learning_rate": 9.746373478087528e-05,
      "loss": 0.9293,
      "mean_token_accuracy": 0.7402238510549068,
      "num_tokens": 26422607.0,
      "step": 828
    },
    {
      "entropy": 0.9813753589987755,
      "epoch": 0.07627328192477328,
      "grad_norm": 0.16816136240959167,
      "learning_rate": 9.746066795473366e-05,
      "loss": 0.9832,
      "mean_token_accuracy": 0.7284796349704266,
      "num_tokens": 26454534.0,
      "step": 829
    },
    {
      "entropy": 0.9882197715342045,
      "epoch": 0.07636528829621451,
      "grad_norm": 0.17259398102760315,
      "learning_rate": 9.745760112859203e-05,
      "loss": 0.9988,
      "mean_token_accuracy": 0.7289420776069164,
      "num_tokens": 26486028.0,
      "step": 830
    },
    {
      "entropy": 1.0355807766318321,
      "epoch": 0.07645729466765573,
      "grad_norm": 0.15970394015312195,
      "learning_rate": 9.745453430245039e-05,
      "loss": 1.012,
      "mean_token_accuracy": 0.7242952808737755,
      "num_tokens": 26518792.0,
      "step": 831
    },
    {
      "entropy": 0.925754964351654,
      "epoch": 0.07654930103909696,
      "grad_norm": 0.1539068967103958,
      "learning_rate": 9.745146747630877e-05,
      "loss": 0.916,
      "mean_token_accuracy": 0.739347230643034,
      "num_tokens": 26551167.0,
      "step": 832
    },
    {
      "entropy": 0.9974164552986622,
      "epoch": 0.07664130741053818,
      "grad_norm": 0.16784407198429108,
      "learning_rate": 9.744840065016714e-05,
      "loss": 0.9832,
      "mean_token_accuracy": 0.7245735190808773,
      "num_tokens": 26582089.0,
      "step": 833
    },
    {
      "entropy": 0.8255923576653004,
      "epoch": 0.0767333137819794,
      "grad_norm": 0.14990749955177307,
      "learning_rate": 9.744533382402552e-05,
      "loss": 0.8194,
      "mean_token_accuracy": 0.7655452415347099,
      "num_tokens": 26613878.0,
      "step": 834
    },
    {
      "entropy": 0.8874995224177837,
      "epoch": 0.07682532015342063,
      "grad_norm": 0.15795740485191345,
      "learning_rate": 9.744226699788389e-05,
      "loss": 0.8883,
      "mean_token_accuracy": 0.7470597922801971,
      "num_tokens": 26645711.0,
      "step": 835
    },
    {
      "entropy": 0.9171003624796867,
      "epoch": 0.07691732652486184,
      "grad_norm": 0.1662931889295578,
      "learning_rate": 9.743920017174227e-05,
      "loss": 0.9362,
      "mean_token_accuracy": 0.7387787289917469,
      "num_tokens": 26676664.0,
      "step": 836
    },
    {
      "entropy": 0.8538406230509281,
      "epoch": 0.07700933289630307,
      "grad_norm": 0.16301359236240387,
      "learning_rate": 9.743613334560064e-05,
      "loss": 0.8727,
      "mean_token_accuracy": 0.7524186931550503,
      "num_tokens": 26709191.0,
      "step": 837
    },
    {
      "entropy": 1.0009033530950546,
      "epoch": 0.07710133926774429,
      "grad_norm": 0.16412551701068878,
      "learning_rate": 9.743306651945901e-05,
      "loss": 1.0242,
      "mean_token_accuracy": 0.7177151963114738,
      "num_tokens": 26741069.0,
      "step": 838
    },
    {
      "entropy": 0.9472977332770824,
      "epoch": 0.07719334563918552,
      "grad_norm": 0.16523784399032593,
      "learning_rate": 9.742999969331739e-05,
      "loss": 0.9536,
      "mean_token_accuracy": 0.733122780919075,
      "num_tokens": 26773204.0,
      "step": 839
    },
    {
      "entropy": 0.9989877268671989,
      "epoch": 0.07728535201062674,
      "grad_norm": 0.1683507263660431,
      "learning_rate": 9.742693286717577e-05,
      "loss": 1.0079,
      "mean_token_accuracy": 0.7174751572310925,
      "num_tokens": 26805585.0,
      "step": 840
    },
    {
      "entropy": 0.9562830775976181,
      "epoch": 0.07737735838206795,
      "grad_norm": 0.1637813150882721,
      "learning_rate": 9.742386604103414e-05,
      "loss": 0.9494,
      "mean_token_accuracy": 0.7355855666100979,
      "num_tokens": 26837863.0,
      "step": 841
    },
    {
      "entropy": 0.9378573261201382,
      "epoch": 0.07746936475350918,
      "grad_norm": 0.16462884843349457,
      "learning_rate": 9.742079921489251e-05,
      "loss": 0.9525,
      "mean_token_accuracy": 0.7417920753359795,
      "num_tokens": 26870159.0,
      "step": 842
    },
    {
      "entropy": 0.9731441214680672,
      "epoch": 0.0775613711249504,
      "grad_norm": 0.16435545682907104,
      "learning_rate": 9.741773238875088e-05,
      "loss": 0.9729,
      "mean_token_accuracy": 0.7314001396298409,
      "num_tokens": 26901919.0,
      "step": 843
    },
    {
      "entropy": 0.9813031405210495,
      "epoch": 0.07765337749639163,
      "grad_norm": 0.16944929957389832,
      "learning_rate": 9.741466556260927e-05,
      "loss": 0.9854,
      "mean_token_accuracy": 0.7247452922165394,
      "num_tokens": 26933876.0,
      "step": 844
    },
    {
      "entropy": 0.9196974337100983,
      "epoch": 0.07774538386783285,
      "grad_norm": 0.15995725989341736,
      "learning_rate": 9.741159873646764e-05,
      "loss": 0.9142,
      "mean_token_accuracy": 0.7426532916724682,
      "num_tokens": 26964778.0,
      "step": 845
    },
    {
      "entropy": 0.8704250119626522,
      "epoch": 0.07783739023927407,
      "grad_norm": 0.15568718314170837,
      "learning_rate": 9.7408531910326e-05,
      "loss": 0.8867,
      "mean_token_accuracy": 0.7550181336700916,
      "num_tokens": 26996849.0,
      "step": 846
    },
    {
      "entropy": 0.9554555602371693,
      "epoch": 0.0779293966107153,
      "grad_norm": 0.16840651631355286,
      "learning_rate": 9.740546508418437e-05,
      "loss": 0.9552,
      "mean_token_accuracy": 0.734230637550354,
      "num_tokens": 27029158.0,
      "step": 847
    },
    {
      "entropy": 0.8672808967530727,
      "epoch": 0.07802140298215651,
      "grad_norm": 0.15096792578697205,
      "learning_rate": 9.740239825804276e-05,
      "loss": 0.8502,
      "mean_token_accuracy": 0.7600157968699932,
      "num_tokens": 27061561.0,
      "step": 848
    },
    {
      "entropy": 0.9291978813707829,
      "epoch": 0.07811340935359774,
      "grad_norm": 0.16496247053146362,
      "learning_rate": 9.739933143190114e-05,
      "loss": 0.9508,
      "mean_token_accuracy": 0.7339499704539776,
      "num_tokens": 27093223.0,
      "step": 849
    },
    {
      "entropy": 0.9408350810408592,
      "epoch": 0.07820541572503896,
      "grad_norm": 0.1602609008550644,
      "learning_rate": 9.73962646057595e-05,
      "loss": 0.9514,
      "mean_token_accuracy": 0.7350383996963501,
      "num_tokens": 27125344.0,
      "step": 850
    },
    {
      "entropy": 0.9225513786077499,
      "epoch": 0.07829742209648018,
      "grad_norm": 0.16747024655342102,
      "learning_rate": 9.739319777961787e-05,
      "loss": 0.9324,
      "mean_token_accuracy": 0.7374814115464687,
      "num_tokens": 27156802.0,
      "step": 851
    },
    {
      "entropy": 0.968829870223999,
      "epoch": 0.07838942846792141,
      "grad_norm": 0.15946544706821442,
      "learning_rate": 9.739013095347625e-05,
      "loss": 0.9685,
      "mean_token_accuracy": 0.7297152131795883,
      "num_tokens": 27188752.0,
      "step": 852
    },
    {
      "entropy": 0.971568375825882,
      "epoch": 0.07848143483936262,
      "grad_norm": 0.16191044449806213,
      "learning_rate": 9.738706412733462e-05,
      "loss": 0.9591,
      "mean_token_accuracy": 0.7315336354076862,
      "num_tokens": 27221032.0,
      "step": 853
    },
    {
      "entropy": 0.9695960693061352,
      "epoch": 0.07857344121080385,
      "grad_norm": 0.15656960010528564,
      "learning_rate": 9.7383997301193e-05,
      "loss": 0.9573,
      "mean_token_accuracy": 0.729323472827673,
      "num_tokens": 27253481.0,
      "step": 854
    },
    {
      "entropy": 0.9361900966614485,
      "epoch": 0.07866544758224507,
      "grad_norm": 0.1591096818447113,
      "learning_rate": 9.738093047505137e-05,
      "loss": 0.935,
      "mean_token_accuracy": 0.736208476126194,
      "num_tokens": 27285879.0,
      "step": 855
    },
    {
      "entropy": 1.002680692821741,
      "epoch": 0.07875745395368629,
      "grad_norm": 0.16537711024284363,
      "learning_rate": 9.737786364890975e-05,
      "loss": 0.9762,
      "mean_token_accuracy": 0.7292106449604034,
      "num_tokens": 27317240.0,
      "step": 856
    },
    {
      "entropy": 0.9484521243721247,
      "epoch": 0.07884946032512752,
      "grad_norm": 0.1627333015203476,
      "learning_rate": 9.737479682276812e-05,
      "loss": 0.966,
      "mean_token_accuracy": 0.7320380657911301,
      "num_tokens": 27349278.0,
      "step": 857
    },
    {
      "entropy": 0.9685053564608097,
      "epoch": 0.07894146669656874,
      "grad_norm": 0.15924955904483795,
      "learning_rate": 9.737172999662649e-05,
      "loss": 0.9756,
      "mean_token_accuracy": 0.7344861850142479,
      "num_tokens": 27381223.0,
      "step": 858
    },
    {
      "entropy": 0.9423946104943752,
      "epoch": 0.07903347306800997,
      "grad_norm": 0.15639211237430573,
      "learning_rate": 9.736866317048487e-05,
      "loss": 0.9296,
      "mean_token_accuracy": 0.7374790124595165,
      "num_tokens": 27413553.0,
      "step": 859
    },
    {
      "entropy": 0.9082550667226315,
      "epoch": 0.07912547943945118,
      "grad_norm": 0.16244761645793915,
      "learning_rate": 9.736559634434325e-05,
      "loss": 0.9402,
      "mean_token_accuracy": 0.7383254840970039,
      "num_tokens": 27445911.0,
      "step": 860
    },
    {
      "entropy": 0.9876478649675846,
      "epoch": 0.0792174858108924,
      "grad_norm": 0.16642430424690247,
      "learning_rate": 9.736252951820162e-05,
      "loss": 1.0224,
      "mean_token_accuracy": 0.7194645293056965,
      "num_tokens": 27478431.0,
      "step": 861
    },
    {
      "entropy": 0.9400400072336197,
      "epoch": 0.07930949218233363,
      "grad_norm": 0.1696026623249054,
      "learning_rate": 9.735946269205999e-05,
      "loss": 0.9594,
      "mean_token_accuracy": 0.7286202944815159,
      "num_tokens": 27510623.0,
      "step": 862
    },
    {
      "entropy": 0.9244946613907814,
      "epoch": 0.07940149855377485,
      "grad_norm": 0.16291528940200806,
      "learning_rate": 9.735639586591837e-05,
      "loss": 0.9584,
      "mean_token_accuracy": 0.7340520657598972,
      "num_tokens": 27543118.0,
      "step": 863
    },
    {
      "entropy": 0.8927987217903137,
      "epoch": 0.07949350492521608,
      "grad_norm": 0.1594545841217041,
      "learning_rate": 9.735332903977674e-05,
      "loss": 0.9016,
      "mean_token_accuracy": 0.7476598992943764,
      "num_tokens": 27575643.0,
      "step": 864
    },
    {
      "entropy": 1.0563612878322601,
      "epoch": 0.0795855112966573,
      "grad_norm": 0.16871725022792816,
      "learning_rate": 9.735026221363512e-05,
      "loss": 1.0397,
      "mean_token_accuracy": 0.7152613364160061,
      "num_tokens": 27607725.0,
      "step": 865
    },
    {
      "entropy": 0.9597365148365498,
      "epoch": 0.07967751766809851,
      "grad_norm": 0.16241663694381714,
      "learning_rate": 9.734719538749349e-05,
      "loss": 0.9544,
      "mean_token_accuracy": 0.735468003898859,
      "num_tokens": 27639503.0,
      "step": 866
    },
    {
      "entropy": 0.8980777077376842,
      "epoch": 0.07976952403953974,
      "grad_norm": 0.15998417139053345,
      "learning_rate": 9.734412856135187e-05,
      "loss": 0.8943,
      "mean_token_accuracy": 0.7470311857759953,
      "num_tokens": 27671208.0,
      "step": 867
    },
    {
      "entropy": 0.8800961971282959,
      "epoch": 0.07986153041098096,
      "grad_norm": 0.1601392924785614,
      "learning_rate": 9.734106173521023e-05,
      "loss": 0.8691,
      "mean_token_accuracy": 0.7504118531942368,
      "num_tokens": 27702278.0,
      "step": 868
    },
    {
      "entropy": 0.9181556589901447,
      "epoch": 0.07995353678242219,
      "grad_norm": 0.15945619344711304,
      "learning_rate": 9.73379949090686e-05,
      "loss": 0.9146,
      "mean_token_accuracy": 0.7429975792765617,
      "num_tokens": 27734046.0,
      "step": 869
    },
    {
      "entropy": 0.9215871319174767,
      "epoch": 0.0800455431538634,
      "grad_norm": 0.16096268594264984,
      "learning_rate": 9.733492808292698e-05,
      "loss": 0.9301,
      "mean_token_accuracy": 0.7395832352340221,
      "num_tokens": 27765853.0,
      "step": 870
    },
    {
      "entropy": 0.9064399190247059,
      "epoch": 0.08013754952530462,
      "grad_norm": 0.15812036395072937,
      "learning_rate": 9.733186125678537e-05,
      "loss": 0.9259,
      "mean_token_accuracy": 0.7397037670016289,
      "num_tokens": 27796898.0,
      "step": 871
    },
    {
      "entropy": 0.9026679843664169,
      "epoch": 0.08022955589674585,
      "grad_norm": 0.1564193069934845,
      "learning_rate": 9.732879443064373e-05,
      "loss": 0.904,
      "mean_token_accuracy": 0.7462964579463005,
      "num_tokens": 27828862.0,
      "step": 872
    },
    {
      "entropy": 0.9240762330591679,
      "epoch": 0.08032156226818707,
      "grad_norm": 0.15576978027820587,
      "learning_rate": 9.73257276045021e-05,
      "loss": 0.911,
      "mean_token_accuracy": 0.7446960024535656,
      "num_tokens": 27860984.0,
      "step": 873
    },
    {
      "entropy": 0.8864658661186695,
      "epoch": 0.0804135686396283,
      "grad_norm": 0.15587686002254486,
      "learning_rate": 9.732266077836047e-05,
      "loss": 0.9003,
      "mean_token_accuracy": 0.74685113504529,
      "num_tokens": 27893579.0,
      "step": 874
    },
    {
      "entropy": 0.9486214481294155,
      "epoch": 0.08050557501106952,
      "grad_norm": 0.1666061282157898,
      "learning_rate": 9.731959395221885e-05,
      "loss": 0.9427,
      "mean_token_accuracy": 0.7366765663027763,
      "num_tokens": 27924502.0,
      "step": 875
    },
    {
      "entropy": 1.0246227085590363,
      "epoch": 0.08059758138251073,
      "grad_norm": 0.1655074506998062,
      "learning_rate": 9.731652712607723e-05,
      "loss": 1.0152,
      "mean_token_accuracy": 0.722616620361805,
      "num_tokens": 27957046.0,
      "step": 876
    },
    {
      "entropy": 0.9945472665131092,
      "epoch": 0.08068958775395196,
      "grad_norm": 0.1661675125360489,
      "learning_rate": 9.73134602999356e-05,
      "loss": 0.9804,
      "mean_token_accuracy": 0.7284702956676483,
      "num_tokens": 27988917.0,
      "step": 877
    },
    {
      "entropy": 0.8925548940896988,
      "epoch": 0.08078159412539318,
      "grad_norm": 0.163802370429039,
      "learning_rate": 9.731039347379397e-05,
      "loss": 0.9195,
      "mean_token_accuracy": 0.7428703345358372,
      "num_tokens": 28020501.0,
      "step": 878
    },
    {
      "entropy": 0.9524826221168041,
      "epoch": 0.08087360049683441,
      "grad_norm": 0.16367146372795105,
      "learning_rate": 9.730732664765235e-05,
      "loss": 0.9533,
      "mean_token_accuracy": 0.7341503724455833,
      "num_tokens": 28051512.0,
      "step": 879
    },
    {
      "entropy": 0.9492873661220074,
      "epoch": 0.08096560686827563,
      "grad_norm": 0.15978164970874786,
      "learning_rate": 9.730425982151073e-05,
      "loss": 0.9426,
      "mean_token_accuracy": 0.7328274585306644,
      "num_tokens": 28083463.0,
      "step": 880
    },
    {
      "entropy": 0.9374285116791725,
      "epoch": 0.08105761323971684,
      "grad_norm": 0.1634908765554428,
      "learning_rate": 9.73011929953691e-05,
      "loss": 0.9418,
      "mean_token_accuracy": 0.73641112819314,
      "num_tokens": 28114357.0,
      "step": 881
    },
    {
      "entropy": 0.9979594834148884,
      "epoch": 0.08114961961115807,
      "grad_norm": 0.1572573184967041,
      "learning_rate": 9.729812616922747e-05,
      "loss": 0.9851,
      "mean_token_accuracy": 0.7266973666846752,
      "num_tokens": 28146784.0,
      "step": 882
    },
    {
      "entropy": 0.9088423773646355,
      "epoch": 0.08124162598259929,
      "grad_norm": 0.16149698197841644,
      "learning_rate": 9.729505934308585e-05,
      "loss": 0.9075,
      "mean_token_accuracy": 0.7481547631323338,
      "num_tokens": 28178476.0,
      "step": 883
    },
    {
      "entropy": 0.9582634121179581,
      "epoch": 0.08133363235404052,
      "grad_norm": 0.16516552865505219,
      "learning_rate": 9.729199251694422e-05,
      "loss": 0.9432,
      "mean_token_accuracy": 0.737374272197485,
      "num_tokens": 28210964.0,
      "step": 884
    },
    {
      "entropy": 0.9771178923547268,
      "epoch": 0.08142563872548174,
      "grad_norm": 0.1658152937889099,
      "learning_rate": 9.72889256908026e-05,
      "loss": 0.9563,
      "mean_token_accuracy": 0.7355898842215538,
      "num_tokens": 28243180.0,
      "step": 885
    },
    {
      "entropy": 0.9357305876910686,
      "epoch": 0.08151764509692296,
      "grad_norm": 0.15437202155590057,
      "learning_rate": 9.728585886466098e-05,
      "loss": 0.9313,
      "mean_token_accuracy": 0.7362537570297718,
      "num_tokens": 28275219.0,
      "step": 886
    },
    {
      "entropy": 1.030756052583456,
      "epoch": 0.08160965146836419,
      "grad_norm": 0.17871473729610443,
      "learning_rate": 9.728279203851935e-05,
      "loss": 1.0493,
      "mean_token_accuracy": 0.717909686267376,
      "num_tokens": 28306359.0,
      "step": 887
    },
    {
      "entropy": 1.0051693692803383,
      "epoch": 0.0817016578398054,
      "grad_norm": 0.16931192576885223,
      "learning_rate": 9.727972521237771e-05,
      "loss": 1.0198,
      "mean_token_accuracy": 0.7161805108189583,
      "num_tokens": 28338062.0,
      "step": 888
    },
    {
      "entropy": 0.8886675722897053,
      "epoch": 0.08179366421124663,
      "grad_norm": 0.1541573703289032,
      "learning_rate": 9.727665838623608e-05,
      "loss": 0.8986,
      "mean_token_accuracy": 0.747088111937046,
      "num_tokens": 28369979.0,
      "step": 889
    },
    {
      "entropy": 1.011715266853571,
      "epoch": 0.08188567058268785,
      "grad_norm": 0.16964338719844818,
      "learning_rate": 9.727359156009446e-05,
      "loss": 1.042,
      "mean_token_accuracy": 0.7169215492904186,
      "num_tokens": 28401897.0,
      "step": 890
    },
    {
      "entropy": 0.9651166312396526,
      "epoch": 0.08197767695412907,
      "grad_norm": 0.16081279516220093,
      "learning_rate": 9.727052473395284e-05,
      "loss": 0.977,
      "mean_token_accuracy": 0.7289307303726673,
      "num_tokens": 28433657.0,
      "step": 891
    },
    {
      "entropy": 1.0382549222558737,
      "epoch": 0.0820696833255703,
      "grad_norm": 0.16813810169696808,
      "learning_rate": 9.726745790781121e-05,
      "loss": 1.059,
      "mean_token_accuracy": 0.7075018435716629,
      "num_tokens": 28465203.0,
      "step": 892
    },
    {
      "entropy": 0.9477774538099766,
      "epoch": 0.08216168969701151,
      "grad_norm": 0.15902648866176605,
      "learning_rate": 9.726439108166958e-05,
      "loss": 0.9546,
      "mean_token_accuracy": 0.7347676791250706,
      "num_tokens": 28497521.0,
      "step": 893
    },
    {
      "entropy": 1.0245006084442139,
      "epoch": 0.08225369606845274,
      "grad_norm": 0.1653822660446167,
      "learning_rate": 9.726132425552796e-05,
      "loss": 1.0291,
      "mean_token_accuracy": 0.7162531726062298,
      "num_tokens": 28529556.0,
      "step": 894
    },
    {
      "entropy": 0.9168008118867874,
      "epoch": 0.08234570243989396,
      "grad_norm": 0.15906579792499542,
      "learning_rate": 9.725825742938633e-05,
      "loss": 0.9139,
      "mean_token_accuracy": 0.7480812519788742,
      "num_tokens": 28560914.0,
      "step": 895
    },
    {
      "entropy": 0.9632975794374943,
      "epoch": 0.08243770881133518,
      "grad_norm": 0.16391408443450928,
      "learning_rate": 9.725519060324471e-05,
      "loss": 0.9654,
      "mean_token_accuracy": 0.7307506985962391,
      "num_tokens": 28592992.0,
      "step": 896
    },
    {
      "entropy": 1.0200887471437454,
      "epoch": 0.08252971518277641,
      "grad_norm": 0.16870862245559692,
      "learning_rate": 9.725212377710308e-05,
      "loss": 1.029,
      "mean_token_accuracy": 0.7160102762281895,
      "num_tokens": 28624823.0,
      "step": 897
    },
    {
      "entropy": 1.033143013715744,
      "epoch": 0.08262172155421763,
      "grad_norm": 0.1692463457584381,
      "learning_rate": 9.724905695096146e-05,
      "loss": 1.0334,
      "mean_token_accuracy": 0.7121374420821667,
      "num_tokens": 28657377.0,
      "step": 898
    },
    {
      "entropy": 0.992195125669241,
      "epoch": 0.08271372792565886,
      "grad_norm": 0.17153699696063995,
      "learning_rate": 9.724599012481983e-05,
      "loss": 0.9913,
      "mean_token_accuracy": 0.7304934673011303,
      "num_tokens": 28688854.0,
      "step": 899
    },
    {
      "entropy": 0.7990805227309465,
      "epoch": 0.08280573429710007,
      "grad_norm": 0.1482461392879486,
      "learning_rate": 9.72429232986782e-05,
      "loss": 0.7767,
      "mean_token_accuracy": 0.7740639187395573,
      "num_tokens": 28720636.0,
      "step": 900
    },
    {
      "entropy": 0.9079835526645184,
      "epoch": 0.0828977406685413,
      "grad_norm": 0.16627106070518494,
      "learning_rate": 9.723985647253658e-05,
      "loss": 0.8968,
      "mean_token_accuracy": 0.753268089145422,
      "num_tokens": 28752199.0,
      "step": 901
    },
    {
      "entropy": 0.8319191001355648,
      "epoch": 0.08298974703998252,
      "grad_norm": 0.15427318215370178,
      "learning_rate": 9.723678964639496e-05,
      "loss": 0.8339,
      "mean_token_accuracy": 0.7580270878970623,
      "num_tokens": 28784431.0,
      "step": 902
    },
    {
      "entropy": 0.9139825850725174,
      "epoch": 0.08308175341142374,
      "grad_norm": 0.16638121008872986,
      "learning_rate": 9.723372282025333e-05,
      "loss": 0.9387,
      "mean_token_accuracy": 0.735026840120554,
      "num_tokens": 28815597.0,
      "step": 903
    },
    {
      "entropy": 0.8832446783781052,
      "epoch": 0.08317375978286497,
      "grad_norm": 0.16459091007709503,
      "learning_rate": 9.72306559941117e-05,
      "loss": 0.8846,
      "mean_token_accuracy": 0.7492220662534237,
      "num_tokens": 28847441.0,
      "step": 904
    },
    {
      "entropy": 0.9302517957985401,
      "epoch": 0.08326576615430618,
      "grad_norm": 0.16204756498336792,
      "learning_rate": 9.722758916797006e-05,
      "loss": 0.9304,
      "mean_token_accuracy": 0.7373491860926151,
      "num_tokens": 28879377.0,
      "step": 905
    },
    {
      "entropy": 0.9527584686875343,
      "epoch": 0.08335777252574741,
      "grad_norm": 0.1586255133152008,
      "learning_rate": 9.722452234182844e-05,
      "loss": 0.9572,
      "mean_token_accuracy": 0.7334199920296669,
      "num_tokens": 28911622.0,
      "step": 906
    },
    {
      "entropy": 0.8616615459322929,
      "epoch": 0.08344977889718863,
      "grad_norm": 0.161000058054924,
      "learning_rate": 9.722145551568683e-05,
      "loss": 0.8689,
      "mean_token_accuracy": 0.7534597851336002,
      "num_tokens": 28943577.0,
      "step": 907
    },
    {
      "entropy": 0.9634058512747288,
      "epoch": 0.08354178526862985,
      "grad_norm": 0.1671752780675888,
      "learning_rate": 9.72183886895452e-05,
      "loss": 0.9649,
      "mean_token_accuracy": 0.7340079918503761,
      "num_tokens": 28975360.0,
      "step": 908
    },
    {
      "entropy": 1.0335550345480442,
      "epoch": 0.08363379164007108,
      "grad_norm": 0.16297176480293274,
      "learning_rate": 9.721532186340356e-05,
      "loss": 1.0182,
      "mean_token_accuracy": 0.7176880799233913,
      "num_tokens": 29007510.0,
      "step": 909
    },
    {
      "entropy": 1.0312865935266018,
      "epoch": 0.0837257980115123,
      "grad_norm": 0.17152875661849976,
      "learning_rate": 9.721225503726194e-05,
      "loss": 0.9976,
      "mean_token_accuracy": 0.7250109612941742,
      "num_tokens": 29040083.0,
      "step": 910
    },
    {
      "entropy": 0.9341313913464546,
      "epoch": 0.08381780438295353,
      "grad_norm": 0.161838561296463,
      "learning_rate": 9.720918821112031e-05,
      "loss": 0.933,
      "mean_token_accuracy": 0.7396265156567097,
      "num_tokens": 29071586.0,
      "step": 911
    },
    {
      "entropy": 0.9120601788163185,
      "epoch": 0.08390981075439474,
      "grad_norm": 0.15900327265262604,
      "learning_rate": 9.720612138497869e-05,
      "loss": 0.905,
      "mean_token_accuracy": 0.742607057094574,
      "num_tokens": 29103477.0,
      "step": 912
    },
    {
      "entropy": 0.9376872777938843,
      "epoch": 0.08400181712583596,
      "grad_norm": 0.15983904898166656,
      "learning_rate": 9.720305455883706e-05,
      "loss": 0.9429,
      "mean_token_accuracy": 0.7407850995659828,
      "num_tokens": 29135221.0,
      "step": 913
    },
    {
      "entropy": 0.9235844798386097,
      "epoch": 0.08409382349727719,
      "grad_norm": 0.163807675242424,
      "learning_rate": 9.719998773269544e-05,
      "loss": 0.9457,
      "mean_token_accuracy": 0.7377686202526093,
      "num_tokens": 29167554.0,
      "step": 914
    },
    {
      "entropy": 1.0049229823052883,
      "epoch": 0.0841858298687184,
      "grad_norm": 0.16817159950733185,
      "learning_rate": 9.719692090655381e-05,
      "loss": 1.0417,
      "mean_token_accuracy": 0.7223370149731636,
      "num_tokens": 29199156.0,
      "step": 915
    },
    {
      "entropy": 0.9806271158158779,
      "epoch": 0.08427783624015964,
      "grad_norm": 0.16544672846794128,
      "learning_rate": 9.719385408041218e-05,
      "loss": 0.9976,
      "mean_token_accuracy": 0.7249796949326992,
      "num_tokens": 29231224.0,
      "step": 916
    },
    {
      "entropy": 0.9456051588058472,
      "epoch": 0.08436984261160085,
      "grad_norm": 0.16046495735645294,
      "learning_rate": 9.719078725427057e-05,
      "loss": 0.9611,
      "mean_token_accuracy": 0.7322219461202621,
      "num_tokens": 29263583.0,
      "step": 917
    },
    {
      "entropy": 0.8887595124542713,
      "epoch": 0.08446184898304207,
      "grad_norm": 0.15719042718410492,
      "learning_rate": 9.718772042812894e-05,
      "loss": 0.9045,
      "mean_token_accuracy": 0.7454048916697502,
      "num_tokens": 29295970.0,
      "step": 918
    },
    {
      "entropy": 0.9542812034487724,
      "epoch": 0.0845538553544833,
      "grad_norm": 0.15589575469493866,
      "learning_rate": 9.718465360198731e-05,
      "loss": 0.9575,
      "mean_token_accuracy": 0.73128217831254,
      "num_tokens": 29328571.0,
      "step": 919
    },
    {
      "entropy": 0.9638505727052689,
      "epoch": 0.08464586172592452,
      "grad_norm": 0.1626119166612625,
      "learning_rate": 9.718158677584568e-05,
      "loss": 0.9587,
      "mean_token_accuracy": 0.7316138707101345,
      "num_tokens": 29360282.0,
      "step": 920
    },
    {
      "entropy": 0.9536610208451748,
      "epoch": 0.08473786809736575,
      "grad_norm": 0.16201965510845184,
      "learning_rate": 9.717851994970406e-05,
      "loss": 0.9595,
      "mean_token_accuracy": 0.7355928458273411,
      "num_tokens": 29392190.0,
      "step": 921
    },
    {
      "entropy": 0.9360164199024439,
      "epoch": 0.08482987446880697,
      "grad_norm": 0.15674394369125366,
      "learning_rate": 9.717545312356244e-05,
      "loss": 0.8839,
      "mean_token_accuracy": 0.7519618459045887,
      "num_tokens": 29423861.0,
      "step": 922
    },
    {
      "entropy": 1.0307834781706333,
      "epoch": 0.08492188084024818,
      "grad_norm": 0.16153934597969055,
      "learning_rate": 9.71723862974208e-05,
      "loss": 1.0185,
      "mean_token_accuracy": 0.7177651040256023,
      "num_tokens": 29455990.0,
      "step": 923
    },
    {
      "entropy": 0.9291120283305645,
      "epoch": 0.08501388721168941,
      "grad_norm": 0.16381607949733734,
      "learning_rate": 9.716931947127917e-05,
      "loss": 0.9162,
      "mean_token_accuracy": 0.743769109249115,
      "num_tokens": 29487296.0,
      "step": 924
    },
    {
      "entropy": 0.9701007567346096,
      "epoch": 0.08510589358313063,
      "grad_norm": 0.1646110713481903,
      "learning_rate": 9.716625264513756e-05,
      "loss": 0.9608,
      "mean_token_accuracy": 0.7276989035308361,
      "num_tokens": 29519663.0,
      "step": 925
    },
    {
      "entropy": 0.9279647544026375,
      "epoch": 0.08519789995457186,
      "grad_norm": 0.15385520458221436,
      "learning_rate": 9.716318581899592e-05,
      "loss": 0.9357,
      "mean_token_accuracy": 0.7402789257466793,
      "num_tokens": 29552178.0,
      "step": 926
    },
    {
      "entropy": 0.8788296394050121,
      "epoch": 0.08528990632601308,
      "grad_norm": 0.15612773597240448,
      "learning_rate": 9.71601189928543e-05,
      "loss": 0.8831,
      "mean_token_accuracy": 0.7533745802938938,
      "num_tokens": 29583821.0,
      "step": 927
    },
    {
      "entropy": 0.9935759231448174,
      "epoch": 0.08538191269745429,
      "grad_norm": 0.1681300401687622,
      "learning_rate": 9.715705216671267e-05,
      "loss": 1.0094,
      "mean_token_accuracy": 0.7206141501665115,
      "num_tokens": 29615976.0,
      "step": 928
    },
    {
      "entropy": 0.970613107085228,
      "epoch": 0.08547391906889552,
      "grad_norm": 0.16945363581180573,
      "learning_rate": 9.715398534057105e-05,
      "loss": 1.0021,
      "mean_token_accuracy": 0.728080615401268,
      "num_tokens": 29647492.0,
      "step": 929
    },
    {
      "entropy": 0.966818779706955,
      "epoch": 0.08556592544033674,
      "grad_norm": 0.1615465134382248,
      "learning_rate": 9.715091851442942e-05,
      "loss": 0.9924,
      "mean_token_accuracy": 0.7278959788382053,
      "num_tokens": 29679928.0,
      "step": 930
    },
    {
      "entropy": 0.9005981758236885,
      "epoch": 0.08565793181177797,
      "grad_norm": 0.1578885018825531,
      "learning_rate": 9.714785168828779e-05,
      "loss": 0.9369,
      "mean_token_accuracy": 0.7381422705948353,
      "num_tokens": 29712603.0,
      "step": 931
    },
    {
      "entropy": 1.052727472037077,
      "epoch": 0.08574993818321919,
      "grad_norm": 0.1807403564453125,
      "learning_rate": 9.714478486214617e-05,
      "loss": 1.0473,
      "mean_token_accuracy": 0.7114957869052887,
      "num_tokens": 29743642.0,
      "step": 932
    },
    {
      "entropy": 0.976007804274559,
      "epoch": 0.0858419445546604,
      "grad_norm": 0.15990310907363892,
      "learning_rate": 9.714171803600455e-05,
      "loss": 0.9699,
      "mean_token_accuracy": 0.7286074087023735,
      "num_tokens": 29774794.0,
      "step": 933
    },
    {
      "entropy": 1.0614879056811333,
      "epoch": 0.08593395092610163,
      "grad_norm": 0.16741439700126648,
      "learning_rate": 9.713865120986292e-05,
      "loss": 1.0598,
      "mean_token_accuracy": 0.7112355940043926,
      "num_tokens": 29806128.0,
      "step": 934
    },
    {
      "entropy": 1.0479108653962612,
      "epoch": 0.08602595729754285,
      "grad_norm": 0.16534516215324402,
      "learning_rate": 9.713558438372129e-05,
      "loss": 1.0307,
      "mean_token_accuracy": 0.7147114090621471,
      "num_tokens": 29838752.0,
      "step": 935
    },
    {
      "entropy": 0.9310126714408398,
      "epoch": 0.08611796366898408,
      "grad_norm": 0.15779109299182892,
      "learning_rate": 9.713251755757966e-05,
      "loss": 0.9212,
      "mean_token_accuracy": 0.7420204244554043,
      "num_tokens": 29870953.0,
      "step": 936
    },
    {
      "entropy": 0.8316780142486095,
      "epoch": 0.0862099700404253,
      "grad_norm": 0.15946152806282043,
      "learning_rate": 9.712945073143804e-05,
      "loss": 0.8048,
      "mean_token_accuracy": 0.7661098130047321,
      "num_tokens": 29902218.0,
      "step": 937
    },
    {
      "entropy": 0.9058275409042835,
      "epoch": 0.08630197641186652,
      "grad_norm": 0.15455356240272522,
      "learning_rate": 9.712638390529642e-05,
      "loss": 0.8719,
      "mean_token_accuracy": 0.754690807312727,
      "num_tokens": 29934942.0,
      "step": 938
    },
    {
      "entropy": 0.9955501146614552,
      "epoch": 0.08639398278330775,
      "grad_norm": 0.16655628383159637,
      "learning_rate": 9.712331707915479e-05,
      "loss": 0.9878,
      "mean_token_accuracy": 0.724040474742651,
      "num_tokens": 29966489.0,
      "step": 939
    },
    {
      "entropy": 0.9390711151063442,
      "epoch": 0.08648598915474896,
      "grad_norm": 0.16280466318130493,
      "learning_rate": 9.712025025301316e-05,
      "loss": 0.9315,
      "mean_token_accuracy": 0.7416763491928577,
      "num_tokens": 29998770.0,
      "step": 940
    },
    {
      "entropy": 0.7573208399116993,
      "epoch": 0.0865779955261902,
      "grad_norm": 0.14976248145103455,
      "learning_rate": 9.711718342687154e-05,
      "loss": 0.764,
      "mean_token_accuracy": 0.7827834263443947,
      "num_tokens": 30030935.0,
      "step": 941
    },
    {
      "entropy": 0.9194397516548634,
      "epoch": 0.08667000189763141,
      "grad_norm": 0.16441328823566437,
      "learning_rate": 9.71141166007299e-05,
      "loss": 0.9405,
      "mean_token_accuracy": 0.7399983778595924,
      "num_tokens": 30063414.0,
      "step": 942
    },
    {
      "entropy": 0.9308536238968372,
      "epoch": 0.08676200826907263,
      "grad_norm": 0.17267094552516937,
      "learning_rate": 9.711104977458829e-05,
      "loss": 0.9488,
      "mean_token_accuracy": 0.7382223382592201,
      "num_tokens": 30095947.0,
      "step": 943
    },
    {
      "entropy": 1.0473821051418781,
      "epoch": 0.08685401464051386,
      "grad_norm": 0.18497617542743683,
      "learning_rate": 9.710798294844665e-05,
      "loss": 1.0757,
      "mean_token_accuracy": 0.7023703455924988,
      "num_tokens": 30126915.0,
      "step": 944
    },
    {
      "entropy": 0.9788549281656742,
      "epoch": 0.08694602101195507,
      "grad_norm": 0.1653570979833603,
      "learning_rate": 9.710491612230504e-05,
      "loss": 1.0098,
      "mean_token_accuracy": 0.7249509617686272,
      "num_tokens": 30158787.0,
      "step": 945
    },
    {
      "entropy": 0.8788519687950611,
      "epoch": 0.0870380273833963,
      "grad_norm": 0.15479259192943573,
      "learning_rate": 9.71018492961634e-05,
      "loss": 0.9045,
      "mean_token_accuracy": 0.7480422705411911,
      "num_tokens": 30191078.0,
      "step": 946
    },
    {
      "entropy": 0.9450247101485729,
      "epoch": 0.08713003375483752,
      "grad_norm": 0.16010959446430206,
      "learning_rate": 9.709878247002177e-05,
      "loss": 0.955,
      "mean_token_accuracy": 0.7302958145737648,
      "num_tokens": 30222685.0,
      "step": 947
    },
    {
      "entropy": 0.890707204118371,
      "epoch": 0.08722204012627874,
      "grad_norm": 0.15926292538642883,
      "learning_rate": 9.709571564388015e-05,
      "loss": 0.8939,
      "mean_token_accuracy": 0.7437068559229374,
      "num_tokens": 30254738.0,
      "step": 948
    },
    {
      "entropy": 0.9819902554154396,
      "epoch": 0.08731404649771997,
      "grad_norm": 0.16164140403270721,
      "learning_rate": 9.709264881773853e-05,
      "loss": 0.9714,
      "mean_token_accuracy": 0.7303546518087387,
      "num_tokens": 30286992.0,
      "step": 949
    },
    {
      "entropy": 1.0133906826376915,
      "epoch": 0.08740605286916119,
      "grad_norm": 0.16301406919956207,
      "learning_rate": 9.70895819915969e-05,
      "loss": 1.0064,
      "mean_token_accuracy": 0.7252141684293747,
      "num_tokens": 30317711.0,
      "step": 950
    },
    {
      "entropy": 0.9229645505547523,
      "epoch": 0.08749805924060242,
      "grad_norm": 0.15729881823062897,
      "learning_rate": 9.708651516545527e-05,
      "loss": 0.9047,
      "mean_token_accuracy": 0.7399939298629761,
      "num_tokens": 30349749.0,
      "step": 951
    },
    {
      "entropy": 0.9705100767314434,
      "epoch": 0.08759006561204363,
      "grad_norm": 0.16773590445518494,
      "learning_rate": 9.708344833931365e-05,
      "loss": 0.9779,
      "mean_token_accuracy": 0.7285280711948872,
      "num_tokens": 30382081.0,
      "step": 952
    },
    {
      "entropy": 1.0154315792024136,
      "epoch": 0.08768207198348485,
      "grad_norm": 0.1626855432987213,
      "learning_rate": 9.708038151317202e-05,
      "loss": 0.9727,
      "mean_token_accuracy": 0.7255200147628784,
      "num_tokens": 30413813.0,
      "step": 953
    },
    {
      "entropy": 0.9858567733317614,
      "epoch": 0.08777407835492608,
      "grad_norm": 0.16285663843154907,
      "learning_rate": 9.70773146870304e-05,
      "loss": 0.9911,
      "mean_token_accuracy": 0.724607028067112,
      "num_tokens": 30446059.0,
      "step": 954
    },
    {
      "entropy": 0.9113628417253494,
      "epoch": 0.0878660847263673,
      "grad_norm": 0.16188208758831024,
      "learning_rate": 9.707424786088877e-05,
      "loss": 0.9158,
      "mean_token_accuracy": 0.7436315082013607,
      "num_tokens": 30478312.0,
      "step": 955
    },
    {
      "entropy": 0.9495079331099987,
      "epoch": 0.08795809109780853,
      "grad_norm": 0.16167901456356049,
      "learning_rate": 9.707118103474715e-05,
      "loss": 0.958,
      "mean_token_accuracy": 0.7363086566329002,
      "num_tokens": 30510177.0,
      "step": 956
    },
    {
      "entropy": 0.898770909756422,
      "epoch": 0.08805009746924974,
      "grad_norm": 0.1584126353263855,
      "learning_rate": 9.706811420860552e-05,
      "loss": 0.9153,
      "mean_token_accuracy": 0.743336521089077,
      "num_tokens": 30541279.0,
      "step": 957
    },
    {
      "entropy": 0.8961479924619198,
      "epoch": 0.08814210384069097,
      "grad_norm": 0.15940997004508972,
      "learning_rate": 9.706504738246389e-05,
      "loss": 0.9344,
      "mean_token_accuracy": 0.742788091301918,
      "num_tokens": 30573349.0,
      "step": 958
    },
    {
      "entropy": 1.0329826027154922,
      "epoch": 0.08823411021213219,
      "grad_norm": 0.1674753576517105,
      "learning_rate": 9.706198055632227e-05,
      "loss": 1.0348,
      "mean_token_accuracy": 0.711038064211607,
      "num_tokens": 30605561.0,
      "step": 959
    },
    {
      "entropy": 0.9822486266493797,
      "epoch": 0.08832611658357341,
      "grad_norm": 0.16457316279411316,
      "learning_rate": 9.705891373018065e-05,
      "loss": 0.9985,
      "mean_token_accuracy": 0.7178325690329075,
      "num_tokens": 30637849.0,
      "step": 960
    },
    {
      "entropy": 0.9190222397446632,
      "epoch": 0.08841812295501464,
      "grad_norm": 0.16368675231933594,
      "learning_rate": 9.705584690403902e-05,
      "loss": 0.918,
      "mean_token_accuracy": 0.7449268810451031,
      "num_tokens": 30669136.0,
      "step": 961
    },
    {
      "entropy": 0.965740017592907,
      "epoch": 0.08851012932645586,
      "grad_norm": 0.16236372292041779,
      "learning_rate": 9.705278007789738e-05,
      "loss": 0.954,
      "mean_token_accuracy": 0.7334070317447186,
      "num_tokens": 30700943.0,
      "step": 962
    },
    {
      "entropy": 1.0067159943282604,
      "epoch": 0.08860213569789709,
      "grad_norm": 0.16343580186367035,
      "learning_rate": 9.704971325175575e-05,
      "loss": 1.0182,
      "mean_token_accuracy": 0.7169825956225395,
      "num_tokens": 30732881.0,
      "step": 963
    },
    {
      "entropy": 0.9352018758654594,
      "epoch": 0.0886941420693383,
      "grad_norm": 0.1546681970357895,
      "learning_rate": 9.704664642561415e-05,
      "loss": 0.9387,
      "mean_token_accuracy": 0.7408855333924294,
      "num_tokens": 30765418.0,
      "step": 964
    },
    {
      "entropy": 0.9172436706721783,
      "epoch": 0.08878614844077952,
      "grad_norm": 0.1559939980506897,
      "learning_rate": 9.704357959947251e-05,
      "loss": 0.9108,
      "mean_token_accuracy": 0.7435907423496246,
      "num_tokens": 30797656.0,
      "step": 965
    },
    {
      "entropy": 1.014381468296051,
      "epoch": 0.08887815481222075,
      "grad_norm": 0.16157332062721252,
      "learning_rate": 9.704051277333088e-05,
      "loss": 1.0231,
      "mean_token_accuracy": 0.7173183783888817,
      "num_tokens": 30829707.0,
      "step": 966
    },
    {
      "entropy": 1.0633132681250572,
      "epoch": 0.08897016118366197,
      "grad_norm": 0.16106529533863068,
      "learning_rate": 9.703744594718925e-05,
      "loss": 1.0288,
      "mean_token_accuracy": 0.714535303413868,
      "num_tokens": 30861270.0,
      "step": 967
    },
    {
      "entropy": 0.9028850793838501,
      "epoch": 0.0890621675551032,
      "grad_norm": 0.15666131675243378,
      "learning_rate": 9.703437912104763e-05,
      "loss": 0.891,
      "mean_token_accuracy": 0.7529587112367153,
      "num_tokens": 30893414.0,
      "step": 968
    },
    {
      "entropy": 0.901664711534977,
      "epoch": 0.08915417392654441,
      "grad_norm": 0.1575678586959839,
      "learning_rate": 9.703131229490601e-05,
      "loss": 0.887,
      "mean_token_accuracy": 0.7463204972445965,
      "num_tokens": 30924595.0,
      "step": 969
    },
    {
      "entropy": 0.8727289512753487,
      "epoch": 0.08924618029798563,
      "grad_norm": 0.15660585463047028,
      "learning_rate": 9.702824546876438e-05,
      "loss": 0.8836,
      "mean_token_accuracy": 0.7513514570891857,
      "num_tokens": 30956483.0,
      "step": 970
    },
    {
      "entropy": 0.9171620942652225,
      "epoch": 0.08933818666942686,
      "grad_norm": 0.17053432762622833,
      "learning_rate": 9.702517864262275e-05,
      "loss": 0.9262,
      "mean_token_accuracy": 0.7398470938205719,
      "num_tokens": 30988480.0,
      "step": 971
    },
    {
      "entropy": 0.9093360602855682,
      "epoch": 0.08943019304086808,
      "grad_norm": 0.16148212552070618,
      "learning_rate": 9.702211181648113e-05,
      "loss": 0.9246,
      "mean_token_accuracy": 0.7400834634900093,
      "num_tokens": 31019856.0,
      "step": 972
    },
    {
      "entropy": 0.8862450309097767,
      "epoch": 0.08952219941230931,
      "grad_norm": 0.16030707955360413,
      "learning_rate": 9.70190449903395e-05,
      "loss": 0.8896,
      "mean_token_accuracy": 0.7453945688903332,
      "num_tokens": 31052045.0,
      "step": 973
    },
    {
      "entropy": 0.9758000336587429,
      "epoch": 0.08961420578375053,
      "grad_norm": 0.15898574888706207,
      "learning_rate": 9.701597816419788e-05,
      "loss": 0.9842,
      "mean_token_accuracy": 0.7304329834878445,
      "num_tokens": 31083941.0,
      "step": 974
    },
    {
      "entropy": 0.9716283865272999,
      "epoch": 0.08970621215519174,
      "grad_norm": 0.15825632214546204,
      "learning_rate": 9.701291133805625e-05,
      "loss": 0.97,
      "mean_token_accuracy": 0.7302758395671844,
      "num_tokens": 31116068.0,
      "step": 975
    },
    {
      "entropy": 0.9637182056903839,
      "epoch": 0.08979821852663297,
      "grad_norm": 0.16271869838237762,
      "learning_rate": 9.700984451191463e-05,
      "loss": 0.9474,
      "mean_token_accuracy": 0.737514078617096,
      "num_tokens": 31147660.0,
      "step": 976
    },
    {
      "entropy": 0.9852610863745213,
      "epoch": 0.08989022489807419,
      "grad_norm": 0.16070248186588287,
      "learning_rate": 9.7006777685773e-05,
      "loss": 0.9931,
      "mean_token_accuracy": 0.7235134318470955,
      "num_tokens": 31179226.0,
      "step": 977
    },
    {
      "entropy": 0.9003812521696091,
      "epoch": 0.08998223126951542,
      "grad_norm": 0.1594938486814499,
      "learning_rate": 9.700371085963136e-05,
      "loss": 0.9194,
      "mean_token_accuracy": 0.7465169504284859,
      "num_tokens": 31211573.0,
      "step": 978
    },
    {
      "entropy": 0.9566617347300053,
      "epoch": 0.09007423764095664,
      "grad_norm": 0.16396847367286682,
      "learning_rate": 9.700064403348975e-05,
      "loss": 0.9447,
      "mean_token_accuracy": 0.7320454865694046,
      "num_tokens": 31242392.0,
      "step": 979
    },
    {
      "entropy": 0.936132188886404,
      "epoch": 0.09016624401239785,
      "grad_norm": 0.16580185294151306,
      "learning_rate": 9.699757720734813e-05,
      "loss": 0.9535,
      "mean_token_accuracy": 0.7372171021997929,
      "num_tokens": 31274115.0,
      "step": 980
    },
    {
      "entropy": 0.9733963198959827,
      "epoch": 0.09025825038383908,
      "grad_norm": 0.15742434561252594,
      "learning_rate": 9.69945103812065e-05,
      "loss": 0.9695,
      "mean_token_accuracy": 0.7304288670420647,
      "num_tokens": 31306634.0,
      "step": 981
    },
    {
      "entropy": 0.9539418555796146,
      "epoch": 0.0903502567552803,
      "grad_norm": 0.16396547853946686,
      "learning_rate": 9.699144355506486e-05,
      "loss": 0.9788,
      "mean_token_accuracy": 0.7302551195025444,
      "num_tokens": 31338210.0,
      "step": 982
    },
    {
      "entropy": 0.9622322767972946,
      "epoch": 0.09044226312672153,
      "grad_norm": 0.16082173585891724,
      "learning_rate": 9.698837672892324e-05,
      "loss": 0.9607,
      "mean_token_accuracy": 0.7372207976877689,
      "num_tokens": 31370348.0,
      "step": 983
    },
    {
      "entropy": 0.8344350997358561,
      "epoch": 0.09053426949816275,
      "grad_norm": 0.1532052904367447,
      "learning_rate": 9.698530990278161e-05,
      "loss": 0.8519,
      "mean_token_accuracy": 0.7534242309629917,
      "num_tokens": 31402803.0,
      "step": 984
    },
    {
      "entropy": 0.9055809564888477,
      "epoch": 0.09062627586960396,
      "grad_norm": 0.1572045087814331,
      "learning_rate": 9.698224307664e-05,
      "loss": 0.8988,
      "mean_token_accuracy": 0.7476417310535908,
      "num_tokens": 31435068.0,
      "step": 985
    },
    {
      "entropy": 0.912225678563118,
      "epoch": 0.0907182822410452,
      "grad_norm": 0.1611575037240982,
      "learning_rate": 9.697917625049836e-05,
      "loss": 0.9346,
      "mean_token_accuracy": 0.7391578741371632,
      "num_tokens": 31466455.0,
      "step": 986
    },
    {
      "entropy": 0.9728972613811493,
      "epoch": 0.09081028861248641,
      "grad_norm": 0.16346091032028198,
      "learning_rate": 9.697610942435674e-05,
      "loss": 0.9742,
      "mean_token_accuracy": 0.7236883230507374,
      "num_tokens": 31498083.0,
      "step": 987
    },
    {
      "entropy": 0.9775727540254593,
      "epoch": 0.09090229498392764,
      "grad_norm": 0.16748863458633423,
      "learning_rate": 9.697304259821511e-05,
      "loss": 0.983,
      "mean_token_accuracy": 0.7276958897709846,
      "num_tokens": 31529416.0,
      "step": 988
    },
    {
      "entropy": 0.9349136538803577,
      "epoch": 0.09099430135536886,
      "grad_norm": 0.15753895044326782,
      "learning_rate": 9.696997577207348e-05,
      "loss": 0.9253,
      "mean_token_accuracy": 0.7442970164120197,
      "num_tokens": 31561610.0,
      "step": 989
    },
    {
      "entropy": 0.9907169118523598,
      "epoch": 0.09108630772681008,
      "grad_norm": 0.1604354977607727,
      "learning_rate": 9.696690894593186e-05,
      "loss": 0.9783,
      "mean_token_accuracy": 0.7261944971978664,
      "num_tokens": 31593021.0,
      "step": 990
    },
    {
      "entropy": 0.919900968670845,
      "epoch": 0.0911783140982513,
      "grad_norm": 0.15658411383628845,
      "learning_rate": 9.696384211979024e-05,
      "loss": 0.9246,
      "mean_token_accuracy": 0.7386965937912464,
      "num_tokens": 31625019.0,
      "step": 991
    },
    {
      "entropy": 0.935480635613203,
      "epoch": 0.09127032046969252,
      "grad_norm": 0.16053679585456848,
      "learning_rate": 9.696077529364861e-05,
      "loss": 0.9399,
      "mean_token_accuracy": 0.734456479549408,
      "num_tokens": 31657015.0,
      "step": 992
    },
    {
      "entropy": 0.9686630330979824,
      "epoch": 0.09136232684113375,
      "grad_norm": 0.1582288295030594,
      "learning_rate": 9.695770846750698e-05,
      "loss": 0.9613,
      "mean_token_accuracy": 0.7333079725503922,
      "num_tokens": 31688965.0,
      "step": 993
    },
    {
      "entropy": 0.8596264086663723,
      "epoch": 0.09145433321257497,
      "grad_norm": 0.15340855717658997,
      "learning_rate": 9.695464164136535e-05,
      "loss": 0.8612,
      "mean_token_accuracy": 0.7520762085914612,
      "num_tokens": 31721733.0,
      "step": 994
    },
    {
      "entropy": 1.0097669586539268,
      "epoch": 0.09154633958401619,
      "grad_norm": 0.16293388605117798,
      "learning_rate": 9.695157481522373e-05,
      "loss": 1.0103,
      "mean_token_accuracy": 0.719074472784996,
      "num_tokens": 31753606.0,
      "step": 995
    },
    {
      "entropy": 1.0406215749680996,
      "epoch": 0.09163834595545742,
      "grad_norm": 0.16303348541259766,
      "learning_rate": 9.694850798908211e-05,
      "loss": 1.0257,
      "mean_token_accuracy": 0.7151807248592377,
      "num_tokens": 31785190.0,
      "step": 996
    },
    {
      "entropy": 0.9660936444997787,
      "epoch": 0.09173035232689863,
      "grad_norm": 0.16666698455810547,
      "learning_rate": 9.694544116294048e-05,
      "loss": 0.9795,
      "mean_token_accuracy": 0.7266343012452126,
      "num_tokens": 31816318.0,
      "step": 997
    },
    {
      "entropy": 0.9574544876813889,
      "epoch": 0.09182235869833986,
      "grad_norm": 0.16571584343910217,
      "learning_rate": 9.694237433679884e-05,
      "loss": 0.9616,
      "mean_token_accuracy": 0.7317477092146873,
      "num_tokens": 31848369.0,
      "step": 998
    },
    {
      "entropy": 0.9956452772021294,
      "epoch": 0.09191436506978108,
      "grad_norm": 0.16728518903255463,
      "learning_rate": 9.693930751065723e-05,
      "loss": 0.9785,
      "mean_token_accuracy": 0.7275924384593964,
      "num_tokens": 31879808.0,
      "step": 999
    },
    {
      "entropy": 0.9244445860385895,
      "epoch": 0.0920063714412223,
      "grad_norm": 0.16287016868591309,
      "learning_rate": 9.693624068451561e-05,
      "loss": 0.9218,
      "mean_token_accuracy": 0.7368968240916729,
      "num_tokens": 31912088.0,
      "step": 1000
    },
    {
      "entropy": 0.8629881627857685,
      "epoch": 0.09209837781266353,
      "grad_norm": 0.15344002842903137,
      "learning_rate": 9.693317385837397e-05,
      "loss": 0.875,
      "mean_token_accuracy": 0.7488552555441856,
      "num_tokens": 31944216.0,
      "step": 1001
    },
    {
      "entropy": 0.8888028115034103,
      "epoch": 0.09219038418410475,
      "grad_norm": 0.15599994361400604,
      "learning_rate": 9.693010703223234e-05,
      "loss": 0.8912,
      "mean_token_accuracy": 0.7515064738690853,
      "num_tokens": 31976447.0,
      "step": 1002
    },
    {
      "entropy": 0.8835399970412254,
      "epoch": 0.09228239055554598,
      "grad_norm": 0.15850293636322021,
      "learning_rate": 9.692704020609072e-05,
      "loss": 0.9004,
      "mean_token_accuracy": 0.7478994764387608,
      "num_tokens": 32008215.0,
      "step": 1003
    },
    {
      "entropy": 0.860474169254303,
      "epoch": 0.09237439692698719,
      "grad_norm": 0.16349774599075317,
      "learning_rate": 9.692397337994909e-05,
      "loss": 0.8884,
      "mean_token_accuracy": 0.7479382380843163,
      "num_tokens": 32040790.0,
      "step": 1004
    },
    {
      "entropy": 0.8956769034266472,
      "epoch": 0.09246640329842841,
      "grad_norm": 0.1614445000886917,
      "learning_rate": 9.692090655380747e-05,
      "loss": 0.8965,
      "mean_token_accuracy": 0.7496082931756973,
      "num_tokens": 32071954.0,
      "step": 1005
    },
    {
      "entropy": 0.9629922471940517,
      "epoch": 0.09255840966986964,
      "grad_norm": 0.1616009920835495,
      "learning_rate": 9.691783972766584e-05,
      "loss": 0.9762,
      "mean_token_accuracy": 0.7235319204628468,
      "num_tokens": 32104034.0,
      "step": 1006
    },
    {
      "entropy": 0.8950593881309032,
      "epoch": 0.09265041604131086,
      "grad_norm": 0.15602898597717285,
      "learning_rate": 9.691477290152422e-05,
      "loss": 0.8863,
      "mean_token_accuracy": 0.7516770213842392,
      "num_tokens": 32135981.0,
      "step": 1007
    },
    {
      "entropy": 0.9091406762599945,
      "epoch": 0.09274242241275209,
      "grad_norm": 0.1596544235944748,
      "learning_rate": 9.691170607538259e-05,
      "loss": 0.9087,
      "mean_token_accuracy": 0.7438617385923862,
      "num_tokens": 32167523.0,
      "step": 1008
    },
    {
      "entropy": 0.9169737324118614,
      "epoch": 0.0928344287841933,
      "grad_norm": 0.17722569406032562,
      "learning_rate": 9.690863924924096e-05,
      "loss": 0.9197,
      "mean_token_accuracy": 0.7474200092256069,
      "num_tokens": 32198797.0,
      "step": 1009
    },
    {
      "entropy": 1.0030145812779665,
      "epoch": 0.09292643515563452,
      "grad_norm": 0.16721847653388977,
      "learning_rate": 9.690557242309934e-05,
      "loss": 1.0277,
      "mean_token_accuracy": 0.7241777032613754,
      "num_tokens": 32231306.0,
      "step": 1010
    },
    {
      "entropy": 0.882943008095026,
      "epoch": 0.09301844152707575,
      "grad_norm": 0.15226633846759796,
      "learning_rate": 9.690250559695772e-05,
      "loss": 0.8888,
      "mean_token_accuracy": 0.7468382380902767,
      "num_tokens": 32263754.0,
      "step": 1011
    },
    {
      "entropy": 0.9571144878864288,
      "epoch": 0.09311044789851697,
      "grad_norm": 0.16107352077960968,
      "learning_rate": 9.689943877081609e-05,
      "loss": 0.976,
      "mean_token_accuracy": 0.7320382297039032,
      "num_tokens": 32295547.0,
      "step": 1012
    },
    {
      "entropy": 0.9551483504474163,
      "epoch": 0.0932024542699582,
      "grad_norm": 0.15751492977142334,
      "learning_rate": 9.689637194467446e-05,
      "loss": 0.9249,
      "mean_token_accuracy": 0.7424855157732964,
      "num_tokens": 32327522.0,
      "step": 1013
    },
    {
      "entropy": 0.9662023484706879,
      "epoch": 0.09329446064139942,
      "grad_norm": 0.16114582121372223,
      "learning_rate": 9.689330511853284e-05,
      "loss": 0.9503,
      "mean_token_accuracy": 0.7355744950473309,
      "num_tokens": 32359437.0,
      "step": 1014
    },
    {
      "entropy": 0.9885205440223217,
      "epoch": 0.09338646701284065,
      "grad_norm": 0.16707557439804077,
      "learning_rate": 9.68902382923912e-05,
      "loss": 1.0038,
      "mean_token_accuracy": 0.7238613702356815,
      "num_tokens": 32391868.0,
      "step": 1015
    },
    {
      "entropy": 0.936811089515686,
      "epoch": 0.09347847338428186,
      "grad_norm": 0.1547304093837738,
      "learning_rate": 9.688717146624959e-05,
      "loss": 0.9431,
      "mean_token_accuracy": 0.735978327691555,
      "num_tokens": 32424294.0,
      "step": 1016
    },
    {
      "entropy": 0.8488725870847702,
      "epoch": 0.09357047975572308,
      "grad_norm": 0.15775495767593384,
      "learning_rate": 9.688410464010796e-05,
      "loss": 0.8355,
      "mean_token_accuracy": 0.7626862078905106,
      "num_tokens": 32455830.0,
      "step": 1017
    },
    {
      "entropy": 0.9438190348446369,
      "epoch": 0.09366248612716431,
      "grad_norm": 0.1681807041168213,
      "learning_rate": 9.688103781396634e-05,
      "loss": 0.964,
      "mean_token_accuracy": 0.73463549092412,
      "num_tokens": 32487831.0,
      "step": 1018
    },
    {
      "entropy": 0.9804133027791977,
      "epoch": 0.09375449249860553,
      "grad_norm": 0.16500993072986603,
      "learning_rate": 9.68779709878247e-05,
      "loss": 0.9948,
      "mean_token_accuracy": 0.7251145020127296,
      "num_tokens": 32520015.0,
      "step": 1019
    },
    {
      "entropy": 0.9319795221090317,
      "epoch": 0.09384649887004676,
      "grad_norm": 0.1867542564868927,
      "learning_rate": 9.687490416168307e-05,
      "loss": 0.9181,
      "mean_token_accuracy": 0.7421147972345352,
      "num_tokens": 32551269.0,
      "step": 1020
    },
    {
      "entropy": 0.9278343692421913,
      "epoch": 0.09393850524148797,
      "grad_norm": 0.1569521576166153,
      "learning_rate": 9.687183733554145e-05,
      "loss": 0.925,
      "mean_token_accuracy": 0.7419033385813236,
      "num_tokens": 32583414.0,
      "step": 1021
    },
    {
      "entropy": 0.972417987883091,
      "epoch": 0.09403051161292919,
      "grad_norm": 0.16502360999584198,
      "learning_rate": 9.686877050939984e-05,
      "loss": 0.952,
      "mean_token_accuracy": 0.7339511848986149,
      "num_tokens": 32615651.0,
      "step": 1022
    },
    {
      "entropy": 1.0116845145821571,
      "epoch": 0.09412251798437042,
      "grad_norm": 0.1652497500181198,
      "learning_rate": 9.68657036832582e-05,
      "loss": 1.008,
      "mean_token_accuracy": 0.7223413996398449,
      "num_tokens": 32648077.0,
      "step": 1023
    },
    {
      "entropy": 0.9818507730960846,
      "epoch": 0.09421452435581164,
      "grad_norm": 0.16716963052749634,
      "learning_rate": 9.686263685711657e-05,
      "loss": 0.9874,
      "mean_token_accuracy": 0.7258833236992359,
      "num_tokens": 32680365.0,
      "step": 1024
    },
    {
      "entropy": 0.8803936950862408,
      "epoch": 0.09430653072725287,
      "grad_norm": 0.1573384404182434,
      "learning_rate": 9.685957003097494e-05,
      "loss": 0.8766,
      "mean_token_accuracy": 0.7523023635149002,
      "num_tokens": 32712321.0,
      "step": 1025
    },
    {
      "entropy": 0.9271412063390017,
      "epoch": 0.09439853709869409,
      "grad_norm": 0.16411113739013672,
      "learning_rate": 9.685650320483332e-05,
      "loss": 0.9248,
      "mean_token_accuracy": 0.7422735467553139,
      "num_tokens": 32743387.0,
      "step": 1026
    },
    {
      "entropy": 0.9524050913751125,
      "epoch": 0.0944905434701353,
      "grad_norm": 0.22662502527236938,
      "learning_rate": 9.68534363786917e-05,
      "loss": 0.953,
      "mean_token_accuracy": 0.7313084416091442,
      "num_tokens": 32775722.0,
      "step": 1027
    },
    {
      "entropy": 0.9947683997452259,
      "epoch": 0.09458254984157653,
      "grad_norm": 0.1656235307455063,
      "learning_rate": 9.685036955255007e-05,
      "loss": 0.9845,
      "mean_token_accuracy": 0.7281756065785885,
      "num_tokens": 32807527.0,
      "step": 1028
    },
    {
      "entropy": 0.9173492938280106,
      "epoch": 0.09467455621301775,
      "grad_norm": 0.16093690693378448,
      "learning_rate": 9.684730272640844e-05,
      "loss": 0.9235,
      "mean_token_accuracy": 0.7405849806964397,
      "num_tokens": 32839590.0,
      "step": 1029
    },
    {
      "entropy": 0.9861542209982872,
      "epoch": 0.09476656258445898,
      "grad_norm": 0.1708606779575348,
      "learning_rate": 9.684423590026682e-05,
      "loss": 1.0125,
      "mean_token_accuracy": 0.7231595553457737,
      "num_tokens": 32870792.0,
      "step": 1030
    },
    {
      "entropy": 1.0397276245057583,
      "epoch": 0.0948585689559002,
      "grad_norm": 0.17119857668876648,
      "learning_rate": 9.684116907412519e-05,
      "loss": 1.0524,
      "mean_token_accuracy": 0.7068664059042931,
      "num_tokens": 32902997.0,
      "step": 1031
    },
    {
      "entropy": 0.9507414177060127,
      "epoch": 0.09495057532734141,
      "grad_norm": 0.1637730598449707,
      "learning_rate": 9.683810224798357e-05,
      "loss": 0.9428,
      "mean_token_accuracy": 0.739409901201725,
      "num_tokens": 32935279.0,
      "step": 1032
    },
    {
      "entropy": 0.8370507471263409,
      "epoch": 0.09504258169878264,
      "grad_norm": 0.17573779821395874,
      "learning_rate": 9.683503542184194e-05,
      "loss": 0.8315,
      "mean_token_accuracy": 0.7615006566047668,
      "num_tokens": 32968022.0,
      "step": 1033
    },
    {
      "entropy": 0.8869328573346138,
      "epoch": 0.09513458807022386,
      "grad_norm": 0.17310485243797302,
      "learning_rate": 9.683196859570032e-05,
      "loss": 0.887,
      "mean_token_accuracy": 0.7481881231069565,
      "num_tokens": 32999935.0,
      "step": 1034
    },
    {
      "entropy": 0.9130650032311678,
      "epoch": 0.09522659444166509,
      "grad_norm": 0.15953503549098969,
      "learning_rate": 9.682890176955869e-05,
      "loss": 0.8967,
      "mean_token_accuracy": 0.7464870139956474,
      "num_tokens": 33031262.0,
      "step": 1035
    },
    {
      "entropy": 0.9816085286438465,
      "epoch": 0.09531860081310631,
      "grad_norm": 0.1664724349975586,
      "learning_rate": 9.682583494341705e-05,
      "loss": 0.9645,
      "mean_token_accuracy": 0.7292580828070641,
      "num_tokens": 33063018.0,
      "step": 1036
    },
    {
      "entropy": 0.8484234064817429,
      "epoch": 0.09541060718454752,
      "grad_norm": 0.16271883249282837,
      "learning_rate": 9.682276811727543e-05,
      "loss": 0.8514,
      "mean_token_accuracy": 0.7581972815096378,
      "num_tokens": 33094982.0,
      "step": 1037
    },
    {
      "entropy": 0.8872726447880268,
      "epoch": 0.09550261355598876,
      "grad_norm": 0.15918877720832825,
      "learning_rate": 9.681970129113382e-05,
      "loss": 0.9017,
      "mean_token_accuracy": 0.7498243227601051,
      "num_tokens": 33126769.0,
      "step": 1038
    },
    {
      "entropy": 0.9097235277295113,
      "epoch": 0.09559461992742997,
      "grad_norm": 0.1634138971567154,
      "learning_rate": 9.681663446499218e-05,
      "loss": 0.9208,
      "mean_token_accuracy": 0.7413439862430096,
      "num_tokens": 33158201.0,
      "step": 1039
    },
    {
      "entropy": 0.8990794476121664,
      "epoch": 0.0956866262988712,
      "grad_norm": 0.1598498672246933,
      "learning_rate": 9.681356763885055e-05,
      "loss": 0.8955,
      "mean_token_accuracy": 0.7434409782290459,
      "num_tokens": 33190317.0,
      "step": 1040
    },
    {
      "entropy": 0.8658686876296997,
      "epoch": 0.09577863267031242,
      "grad_norm": 0.15954849123954773,
      "learning_rate": 9.681050081270893e-05,
      "loss": 0.8835,
      "mean_token_accuracy": 0.7540551163256168,
      "num_tokens": 33222715.0,
      "step": 1041
    },
    {
      "entropy": 0.8142774850130081,
      "epoch": 0.09587063904175364,
      "grad_norm": 0.15088464319705963,
      "learning_rate": 9.680743398656731e-05,
      "loss": 0.8036,
      "mean_token_accuracy": 0.7712161093950272,
      "num_tokens": 33254947.0,
      "step": 1042
    },
    {
      "entropy": 0.9317591525614262,
      "epoch": 0.09596264541319487,
      "grad_norm": 0.16430522501468658,
      "learning_rate": 9.680436716042568e-05,
      "loss": 0.9276,
      "mean_token_accuracy": 0.74178671464324,
      "num_tokens": 33286816.0,
      "step": 1043
    },
    {
      "entropy": 0.922197887673974,
      "epoch": 0.09605465178463608,
      "grad_norm": 0.16266439855098724,
      "learning_rate": 9.680130033428405e-05,
      "loss": 0.9209,
      "mean_token_accuracy": 0.7399053052067757,
      "num_tokens": 33319191.0,
      "step": 1044
    },
    {
      "entropy": 0.8657081201672554,
      "epoch": 0.09614665815607731,
      "grad_norm": 0.16114577651023865,
      "learning_rate": 9.679823350814243e-05,
      "loss": 0.8626,
      "mean_token_accuracy": 0.755803506821394,
      "num_tokens": 33351336.0,
      "step": 1045
    },
    {
      "entropy": 1.0185527689754963,
      "epoch": 0.09623866452751853,
      "grad_norm": 0.1669594943523407,
      "learning_rate": 9.67951666820008e-05,
      "loss": 1.042,
      "mean_token_accuracy": 0.7172301560640335,
      "num_tokens": 33383101.0,
      "step": 1046
    },
    {
      "entropy": 0.8908280953764915,
      "epoch": 0.09633067089895975,
      "grad_norm": 0.15896745026111603,
      "learning_rate": 9.679209985585918e-05,
      "loss": 0.9002,
      "mean_token_accuracy": 0.7485417276620865,
      "num_tokens": 33414846.0,
      "step": 1047
    },
    {
      "entropy": 0.9098036773502827,
      "epoch": 0.09642267727040098,
      "grad_norm": 0.1624700129032135,
      "learning_rate": 9.678903302971755e-05,
      "loss": 0.9191,
      "mean_token_accuracy": 0.7450025863945484,
      "num_tokens": 33445968.0,
      "step": 1048
    },
    {
      "entropy": 0.8776053879410028,
      "epoch": 0.0965146836418422,
      "grad_norm": 0.157759428024292,
      "learning_rate": 9.678596620357593e-05,
      "loss": 0.8716,
      "mean_token_accuracy": 0.7503661215305328,
      "num_tokens": 33477663.0,
      "step": 1049
    },
    {
      "entropy": 0.9139085151255131,
      "epoch": 0.09660669001328342,
      "grad_norm": 0.1621224284172058,
      "learning_rate": 9.67828993774343e-05,
      "loss": 0.9135,
      "mean_token_accuracy": 0.7473532110452652,
      "num_tokens": 33509405.0,
      "step": 1050
    },
    {
      "entropy": 0.9733624868094921,
      "epoch": 0.09669869638472464,
      "grad_norm": 0.1676139384508133,
      "learning_rate": 9.677983255129267e-05,
      "loss": 0.9793,
      "mean_token_accuracy": 0.7292316779494286,
      "num_tokens": 33541217.0,
      "step": 1051
    },
    {
      "entropy": 0.8864538930356503,
      "epoch": 0.09679070275616586,
      "grad_norm": 0.15317882597446442,
      "learning_rate": 9.677676572515105e-05,
      "loss": 0.8746,
      "mean_token_accuracy": 0.7549472898244858,
      "num_tokens": 33573693.0,
      "step": 1052
    },
    {
      "entropy": 1.0338486805558205,
      "epoch": 0.09688270912760709,
      "grad_norm": 0.16735602915287018,
      "learning_rate": 9.677369889900943e-05,
      "loss": 1.0699,
      "mean_token_accuracy": 0.7064846158027649,
      "num_tokens": 33606121.0,
      "step": 1053
    },
    {
      "entropy": 0.8866732828319073,
      "epoch": 0.0969747154990483,
      "grad_norm": 0.1557445079088211,
      "learning_rate": 9.67706320728678e-05,
      "loss": 0.88,
      "mean_token_accuracy": 0.7527556009590626,
      "num_tokens": 33637879.0,
      "step": 1054
    },
    {
      "entropy": 0.8826326318085194,
      "epoch": 0.09706672187048954,
      "grad_norm": 0.1545235514640808,
      "learning_rate": 9.676756524672617e-05,
      "loss": 0.8658,
      "mean_token_accuracy": 0.7500460743904114,
      "num_tokens": 33669781.0,
      "step": 1055
    },
    {
      "entropy": 0.9689504578709602,
      "epoch": 0.09715872824193075,
      "grad_norm": 0.16593037545681,
      "learning_rate": 9.676449842058453e-05,
      "loss": 0.9747,
      "mean_token_accuracy": 0.7251571975648403,
      "num_tokens": 33701308.0,
      "step": 1056
    },
    {
      "entropy": 0.9733709320425987,
      "epoch": 0.09725073461337197,
      "grad_norm": 0.1710357367992401,
      "learning_rate": 9.676143159444291e-05,
      "loss": 0.9774,
      "mean_token_accuracy": 0.7302632331848145,
      "num_tokens": 33732847.0,
      "step": 1057
    },
    {
      "entropy": 0.9180822595953941,
      "epoch": 0.0973427409848132,
      "grad_norm": 0.1564447581768036,
      "learning_rate": 9.67583647683013e-05,
      "loss": 0.8982,
      "mean_token_accuracy": 0.7466649785637856,
      "num_tokens": 33764874.0,
      "step": 1058
    },
    {
      "entropy": 0.9152316823601723,
      "epoch": 0.09743474735625442,
      "grad_norm": 0.1574152261018753,
      "learning_rate": 9.675529794215966e-05,
      "loss": 0.9271,
      "mean_token_accuracy": 0.7390196546912193,
      "num_tokens": 33797307.0,
      "step": 1059
    },
    {
      "entropy": 0.9296806827187538,
      "epoch": 0.09752675372769565,
      "grad_norm": 0.16534656286239624,
      "learning_rate": 9.675223111601803e-05,
      "loss": 0.9528,
      "mean_token_accuracy": 0.7388513460755348,
      "num_tokens": 33829520.0,
      "step": 1060
    },
    {
      "entropy": 1.0560470670461655,
      "epoch": 0.09761876009913686,
      "grad_norm": 0.16798868775367737,
      "learning_rate": 9.674916428987641e-05,
      "loss": 1.0695,
      "mean_token_accuracy": 0.7047516629099846,
      "num_tokens": 33861514.0,
      "step": 1061
    },
    {
      "entropy": 0.8678808100521564,
      "epoch": 0.09771076647057808,
      "grad_norm": 0.15756385028362274,
      "learning_rate": 9.674609746373478e-05,
      "loss": 0.8599,
      "mean_token_accuracy": 0.7545629553496838,
      "num_tokens": 33893253.0,
      "step": 1062
    },
    {
      "entropy": 0.9519164338707924,
      "epoch": 0.09780277284201931,
      "grad_norm": 0.15441997349262238,
      "learning_rate": 9.674303063759316e-05,
      "loss": 0.9283,
      "mean_token_accuracy": 0.7379158958792686,
      "num_tokens": 33925567.0,
      "step": 1063
    },
    {
      "entropy": 0.9062401410192251,
      "epoch": 0.09789477921346053,
      "grad_norm": 0.15285854041576385,
      "learning_rate": 9.673996381145153e-05,
      "loss": 0.9016,
      "mean_token_accuracy": 0.7511668279767036,
      "num_tokens": 33957095.0,
      "step": 1064
    },
    {
      "entropy": 0.8826639018952847,
      "epoch": 0.09798678558490176,
      "grad_norm": 0.1611899733543396,
      "learning_rate": 9.673689698530991e-05,
      "loss": 0.8893,
      "mean_token_accuracy": 0.7482551038265228,
      "num_tokens": 33988960.0,
      "step": 1065
    },
    {
      "entropy": 1.0272873491048813,
      "epoch": 0.09807879195634298,
      "grad_norm": 0.17073515057563782,
      "learning_rate": 9.673383015916828e-05,
      "loss": 1.0306,
      "mean_token_accuracy": 0.714377149939537,
      "num_tokens": 34021185.0,
      "step": 1066
    },
    {
      "entropy": 0.9972330816090107,
      "epoch": 0.09817079832778419,
      "grad_norm": 0.1619044542312622,
      "learning_rate": 9.673076333302665e-05,
      "loss": 0.9857,
      "mean_token_accuracy": 0.7264757268130779,
      "num_tokens": 34052951.0,
      "step": 1067
    },
    {
      "entropy": 0.8821131717413664,
      "epoch": 0.09826280469922542,
      "grad_norm": 0.15990856289863586,
      "learning_rate": 9.672769650688503e-05,
      "loss": 0.8875,
      "mean_token_accuracy": 0.7503055110573769,
      "num_tokens": 34084016.0,
      "step": 1068
    },
    {
      "entropy": 0.9054156877100468,
      "epoch": 0.09835481107066664,
      "grad_norm": 0.15616628527641296,
      "learning_rate": 9.672462968074341e-05,
      "loss": 0.9124,
      "mean_token_accuracy": 0.7486260384321213,
      "num_tokens": 34116784.0,
      "step": 1069
    },
    {
      "entropy": 0.976450253278017,
      "epoch": 0.09844681744210787,
      "grad_norm": 0.15976481139659882,
      "learning_rate": 9.672156285460178e-05,
      "loss": 0.9535,
      "mean_token_accuracy": 0.7336842603981495,
      "num_tokens": 34149033.0,
      "step": 1070
    },
    {
      "entropy": 0.8957774676382542,
      "epoch": 0.09853882381354909,
      "grad_norm": 0.1565529704093933,
      "learning_rate": 9.671849602846015e-05,
      "loss": 0.887,
      "mean_token_accuracy": 0.7503524050116539,
      "num_tokens": 34180688.0,
      "step": 1071
    },
    {
      "entropy": 0.8924312628805637,
      "epoch": 0.09863083018499032,
      "grad_norm": 0.16206608712673187,
      "learning_rate": 9.671542920231853e-05,
      "loss": 0.8949,
      "mean_token_accuracy": 0.7436465509235859,
      "num_tokens": 34212407.0,
      "step": 1072
    },
    {
      "entropy": 0.8703174404799938,
      "epoch": 0.09872283655643153,
      "grad_norm": 0.15224649012088776,
      "learning_rate": 9.67123623761769e-05,
      "loss": 0.8667,
      "mean_token_accuracy": 0.7529569268226624,
      "num_tokens": 34243775.0,
      "step": 1073
    },
    {
      "entropy": 0.817875262349844,
      "epoch": 0.09881484292787275,
      "grad_norm": 0.165665864944458,
      "learning_rate": 9.670929555003528e-05,
      "loss": 0.8115,
      "mean_token_accuracy": 0.7664502784609795,
      "num_tokens": 34275426.0,
      "step": 1074
    },
    {
      "entropy": 1.0160892121493816,
      "epoch": 0.09890684929931398,
      "grad_norm": 0.16657552123069763,
      "learning_rate": 9.670622872389364e-05,
      "loss": 1.031,
      "mean_token_accuracy": 0.7158407270908356,
      "num_tokens": 34307552.0,
      "step": 1075
    },
    {
      "entropy": 0.9705043807625771,
      "epoch": 0.0989988556707552,
      "grad_norm": 0.1613999903202057,
      "learning_rate": 9.670316189775203e-05,
      "loss": 0.9713,
      "mean_token_accuracy": 0.7268971800804138,
      "num_tokens": 34339041.0,
      "step": 1076
    },
    {
      "entropy": 0.8460965864360332,
      "epoch": 0.09909086204219643,
      "grad_norm": 0.15994155406951904,
      "learning_rate": 9.67000950716104e-05,
      "loss": 0.8633,
      "mean_token_accuracy": 0.755526214838028,
      "num_tokens": 34371474.0,
      "step": 1077
    },
    {
      "entropy": 0.9997069947421551,
      "epoch": 0.09918286841363765,
      "grad_norm": 0.16577109694480896,
      "learning_rate": 9.669702824546876e-05,
      "loss": 1.0158,
      "mean_token_accuracy": 0.7151198834180832,
      "num_tokens": 34403139.0,
      "step": 1078
    },
    {
      "entropy": 0.8220634870231152,
      "epoch": 0.09927487478507886,
      "grad_norm": 0.15536828339099884,
      "learning_rate": 9.669396141932714e-05,
      "loss": 0.8613,
      "mean_token_accuracy": 0.7651693262159824,
      "num_tokens": 34435635.0,
      "step": 1079
    },
    {
      "entropy": 0.9301723409444094,
      "epoch": 0.09936688115652009,
      "grad_norm": 0.15812771022319794,
      "learning_rate": 9.669089459318552e-05,
      "loss": 0.9312,
      "mean_token_accuracy": 0.7425216287374496,
      "num_tokens": 34467666.0,
      "step": 1080
    },
    {
      "entropy": 0.8817839231342077,
      "epoch": 0.09945888752796131,
      "grad_norm": 0.15515027940273285,
      "learning_rate": 9.668782776704389e-05,
      "loss": 0.8887,
      "mean_token_accuracy": 0.7482687644660473,
      "num_tokens": 34499544.0,
      "step": 1081
    },
    {
      "entropy": 0.9201575219631195,
      "epoch": 0.09955089389940254,
      "grad_norm": 0.1571827083826065,
      "learning_rate": 9.668476094090226e-05,
      "loss": 0.9073,
      "mean_token_accuracy": 0.7487991563975811,
      "num_tokens": 34531369.0,
      "step": 1082
    },
    {
      "entropy": 0.9239728301763535,
      "epoch": 0.09964290027084376,
      "grad_norm": 0.15828335285186768,
      "learning_rate": 9.668169411476063e-05,
      "loss": 0.9118,
      "mean_token_accuracy": 0.7416358850896358,
      "num_tokens": 34563212.0,
      "step": 1083
    },
    {
      "entropy": 0.9013906978070736,
      "epoch": 0.09973490664228497,
      "grad_norm": 0.1541551798582077,
      "learning_rate": 9.667862728861902e-05,
      "loss": 0.8786,
      "mean_token_accuracy": 0.7491864114999771,
      "num_tokens": 34595265.0,
      "step": 1084
    },
    {
      "entropy": 1.0174082405865192,
      "epoch": 0.0998269130137262,
      "grad_norm": 0.16384847462177277,
      "learning_rate": 9.667556046247739e-05,
      "loss": 1.0052,
      "mean_token_accuracy": 0.7226257584989071,
      "num_tokens": 34627456.0,
      "step": 1085
    },
    {
      "entropy": 1.0035100504755974,
      "epoch": 0.09991891938516742,
      "grad_norm": 0.16659247875213623,
      "learning_rate": 9.667249363633576e-05,
      "loss": 1.0145,
      "mean_token_accuracy": 0.7195708006620407,
      "num_tokens": 34659294.0,
      "step": 1086
    },
    {
      "entropy": 0.8904071226716042,
      "epoch": 0.10001092575660865,
      "grad_norm": 0.1593467742204666,
      "learning_rate": 9.666942681019413e-05,
      "loss": 0.8984,
      "mean_token_accuracy": 0.7446462102234364,
      "num_tokens": 34690983.0,
      "step": 1087
    },
    {
      "entropy": 0.9790074042975903,
      "epoch": 0.10010293212804987,
      "grad_norm": 0.1585438996553421,
      "learning_rate": 9.666635998405251e-05,
      "loss": 0.992,
      "mean_token_accuracy": 0.7231258042156696,
      "num_tokens": 34722864.0,
      "step": 1088
    },
    {
      "entropy": 0.9666843581944704,
      "epoch": 0.10019493849949108,
      "grad_norm": 0.16131551563739777,
      "learning_rate": 9.666329315791089e-05,
      "loss": 0.955,
      "mean_token_accuracy": 0.7351276576519012,
      "num_tokens": 34754897.0,
      "step": 1089
    },
    {
      "entropy": 0.8937724698334932,
      "epoch": 0.10028694487093232,
      "grad_norm": 0.15138773620128632,
      "learning_rate": 9.666022633176926e-05,
      "loss": 0.8664,
      "mean_token_accuracy": 0.753471415489912,
      "num_tokens": 34786436.0,
      "step": 1090
    },
    {
      "entropy": 0.9003929756581783,
      "epoch": 0.10037895124237353,
      "grad_norm": 0.15860742330551147,
      "learning_rate": 9.665715950562763e-05,
      "loss": 0.8983,
      "mean_token_accuracy": 0.7465070001780987,
      "num_tokens": 34817508.0,
      "step": 1091
    },
    {
      "entropy": 0.9523097351193428,
      "epoch": 0.10047095761381476,
      "grad_norm": 0.16087223589420319,
      "learning_rate": 9.6654092679486e-05,
      "loss": 0.9588,
      "mean_token_accuracy": 0.7335548102855682,
      "num_tokens": 34849829.0,
      "step": 1092
    },
    {
      "entropy": 0.9897637516260147,
      "epoch": 0.10056296398525598,
      "grad_norm": 0.1690727174282074,
      "learning_rate": 9.665102585334437e-05,
      "loss": 1.0142,
      "mean_token_accuracy": 0.7204057015478611,
      "num_tokens": 34881166.0,
      "step": 1093
    },
    {
      "entropy": 0.959464754909277,
      "epoch": 0.1006549703566972,
      "grad_norm": 0.1661749929189682,
      "learning_rate": 9.664795902720276e-05,
      "loss": 0.9709,
      "mean_token_accuracy": 0.7302238345146179,
      "num_tokens": 34912777.0,
      "step": 1094
    },
    {
      "entropy": 0.9007907155901194,
      "epoch": 0.10074697672813843,
      "grad_norm": 0.17002630233764648,
      "learning_rate": 9.664489220106112e-05,
      "loss": 0.8841,
      "mean_token_accuracy": 0.7535837031900883,
      "num_tokens": 34945307.0,
      "step": 1095
    },
    {
      "entropy": 0.9063800685107708,
      "epoch": 0.10083898309957964,
      "grad_norm": 0.15781642496585846,
      "learning_rate": 9.66418253749195e-05,
      "loss": 0.9032,
      "mean_token_accuracy": 0.749104131013155,
      "num_tokens": 34977263.0,
      "step": 1096
    },
    {
      "entropy": 0.908812876790762,
      "epoch": 0.10093098947102087,
      "grad_norm": 0.16161276400089264,
      "learning_rate": 9.663875854877787e-05,
      "loss": 0.9303,
      "mean_token_accuracy": 0.7401559688150883,
      "num_tokens": 35009319.0,
      "step": 1097
    },
    {
      "entropy": 0.9026827737689018,
      "epoch": 0.10102299584246209,
      "grad_norm": 0.15920984745025635,
      "learning_rate": 9.663569172263624e-05,
      "loss": 0.916,
      "mean_token_accuracy": 0.7485790550708771,
      "num_tokens": 35041570.0,
      "step": 1098
    },
    {
      "entropy": 0.956053651869297,
      "epoch": 0.10111500221390331,
      "grad_norm": 0.16887539625167847,
      "learning_rate": 9.663262489649462e-05,
      "loss": 0.977,
      "mean_token_accuracy": 0.7279412932693958,
      "num_tokens": 35073557.0,
      "step": 1099
    },
    {
      "entropy": 0.8551238477230072,
      "epoch": 0.10120700858534454,
      "grad_norm": 0.15822964906692505,
      "learning_rate": 9.6629558070353e-05,
      "loss": 0.8387,
      "mean_token_accuracy": 0.7585063502192497,
      "num_tokens": 35105422.0,
      "step": 1100
    },
    {
      "entropy": 0.8338537495583296,
      "epoch": 0.10129901495678575,
      "grad_norm": 0.15613920986652374,
      "learning_rate": 9.662649124421137e-05,
      "loss": 0.8587,
      "mean_token_accuracy": 0.7573117949068546,
      "num_tokens": 35137190.0,
      "step": 1101
    },
    {
      "entropy": 0.9067161530256271,
      "epoch": 0.10139102132822699,
      "grad_norm": 0.16520637273788452,
      "learning_rate": 9.662342441806974e-05,
      "loss": 0.9306,
      "mean_token_accuracy": 0.7347854487597942,
      "num_tokens": 35169174.0,
      "step": 1102
    },
    {
      "entropy": 0.9459751918911934,
      "epoch": 0.1014830276996682,
      "grad_norm": 0.16897724568843842,
      "learning_rate": 9.662035759192812e-05,
      "loss": 0.9673,
      "mean_token_accuracy": 0.7290536165237427,
      "num_tokens": 35200896.0,
      "step": 1103
    },
    {
      "entropy": 0.8430897425860167,
      "epoch": 0.10157503407110942,
      "grad_norm": 0.1533231884241104,
      "learning_rate": 9.661729076578649e-05,
      "loss": 0.8463,
      "mean_token_accuracy": 0.7605555765330791,
      "num_tokens": 35233362.0,
      "step": 1104
    },
    {
      "entropy": 0.9403893984854221,
      "epoch": 0.10166704044255065,
      "grad_norm": 0.15612873435020447,
      "learning_rate": 9.661422393964487e-05,
      "loss": 0.9547,
      "mean_token_accuracy": 0.7400345988571644,
      "num_tokens": 35265812.0,
      "step": 1105
    },
    {
      "entropy": 1.0277419053018093,
      "epoch": 0.10175904681399187,
      "grad_norm": 0.17066191136837006,
      "learning_rate": 9.661115711350324e-05,
      "loss": 1.0418,
      "mean_token_accuracy": 0.7106143981218338,
      "num_tokens": 35297444.0,
      "step": 1106
    },
    {
      "entropy": 0.9457535222172737,
      "epoch": 0.1018510531854331,
      "grad_norm": 0.15654437243938446,
      "learning_rate": 9.660809028736162e-05,
      "loss": 0.9239,
      "mean_token_accuracy": 0.7407226487994194,
      "num_tokens": 35330015.0,
      "step": 1107
    },
    {
      "entropy": 0.9294662438333035,
      "epoch": 0.10194305955687431,
      "grad_norm": 0.1531471461057663,
      "learning_rate": 9.660502346121999e-05,
      "loss": 0.8924,
      "mean_token_accuracy": 0.7504142187535763,
      "num_tokens": 35361049.0,
      "step": 1108
    },
    {
      "entropy": 0.9033985212445259,
      "epoch": 0.10203506592831553,
      "grad_norm": 0.15580113232135773,
      "learning_rate": 9.660195663507836e-05,
      "loss": 0.8816,
      "mean_token_accuracy": 0.7505446337163448,
      "num_tokens": 35393274.0,
      "step": 1109
    },
    {
      "entropy": 0.9048357978463173,
      "epoch": 0.10212707229975676,
      "grad_norm": 0.1531599909067154,
      "learning_rate": 9.659888980893674e-05,
      "loss": 0.8848,
      "mean_token_accuracy": 0.7556491568684578,
      "num_tokens": 35424654.0,
      "step": 1110
    },
    {
      "entropy": 0.9034750461578369,
      "epoch": 0.10221907867119798,
      "grad_norm": 0.15479497611522675,
      "learning_rate": 9.659582298279512e-05,
      "loss": 0.881,
      "mean_token_accuracy": 0.7480965480208397,
      "num_tokens": 35456567.0,
      "step": 1111
    },
    {
      "entropy": 0.8956337310373783,
      "epoch": 0.10231108504263921,
      "grad_norm": 0.15655048191547394,
      "learning_rate": 9.659275615665349e-05,
      "loss": 0.9083,
      "mean_token_accuracy": 0.7485470101237297,
      "num_tokens": 35489103.0,
      "step": 1112
    },
    {
      "entropy": 0.9457507394254208,
      "epoch": 0.10240309141408042,
      "grad_norm": 0.1576923429965973,
      "learning_rate": 9.658968933051185e-05,
      "loss": 0.9557,
      "mean_token_accuracy": 0.735408890992403,
      "num_tokens": 35521030.0,
      "step": 1113
    },
    {
      "entropy": 0.9384334124624729,
      "epoch": 0.10249509778552164,
      "grad_norm": 0.16467717289924622,
      "learning_rate": 9.658662250437022e-05,
      "loss": 0.9568,
      "mean_token_accuracy": 0.7333705574274063,
      "num_tokens": 35552232.0,
      "step": 1114
    },
    {
      "entropy": 0.82656304910779,
      "epoch": 0.10258710415696287,
      "grad_norm": 0.15733087062835693,
      "learning_rate": 9.65835556782286e-05,
      "loss": 0.8247,
      "mean_token_accuracy": 0.7590740993618965,
      "num_tokens": 35583698.0,
      "step": 1115
    },
    {
      "entropy": 0.9833521023392677,
      "epoch": 0.10267911052840409,
      "grad_norm": 0.1641973853111267,
      "learning_rate": 9.658048885208698e-05,
      "loss": 1.0147,
      "mean_token_accuracy": 0.7188589200377464,
      "num_tokens": 35614878.0,
      "step": 1116
    },
    {
      "entropy": 0.8835620544850826,
      "epoch": 0.10277111689984532,
      "grad_norm": 0.15869088470935822,
      "learning_rate": 9.657742202594535e-05,
      "loss": 0.8875,
      "mean_token_accuracy": 0.7516782358288765,
      "num_tokens": 35647207.0,
      "step": 1117
    },
    {
      "entropy": 0.9517400711774826,
      "epoch": 0.10286312327128654,
      "grad_norm": 0.1626524031162262,
      "learning_rate": 9.657435519980372e-05,
      "loss": 0.9685,
      "mean_token_accuracy": 0.7301286123692989,
      "num_tokens": 35679324.0,
      "step": 1118
    },
    {
      "entropy": 0.9061287343502045,
      "epoch": 0.10295512964272775,
      "grad_norm": 0.16136522591114044,
      "learning_rate": 9.65712883736621e-05,
      "loss": 0.9052,
      "mean_token_accuracy": 0.7399870231747627,
      "num_tokens": 35709638.0,
      "step": 1119
    },
    {
      "entropy": 0.9113659039139748,
      "epoch": 0.10304713601416898,
      "grad_norm": 0.15852506458759308,
      "learning_rate": 9.656822154752048e-05,
      "loss": 0.9191,
      "mean_token_accuracy": 0.742094062268734,
      "num_tokens": 35741210.0,
      "step": 1120
    },
    {
      "entropy": 1.0113764591515064,
      "epoch": 0.1031391423856102,
      "grad_norm": 0.16707167029380798,
      "learning_rate": 9.656515472137885e-05,
      "loss": 1.0138,
      "mean_token_accuracy": 0.7160374745726585,
      "num_tokens": 35772957.0,
      "step": 1121
    },
    {
      "entropy": 0.9776623621582985,
      "epoch": 0.10323114875705143,
      "grad_norm": 0.24139249324798584,
      "learning_rate": 9.656208789523722e-05,
      "loss": 0.9722,
      "mean_token_accuracy": 0.7255499921739101,
      "num_tokens": 35804624.0,
      "step": 1122
    },
    {
      "entropy": 0.8269810415804386,
      "epoch": 0.10332315512849265,
      "grad_norm": 0.15414750576019287,
      "learning_rate": 9.65590210690956e-05,
      "loss": 0.8327,
      "mean_token_accuracy": 0.765818364918232,
      "num_tokens": 35836744.0,
      "step": 1123
    },
    {
      "entropy": 0.7696375567466021,
      "epoch": 0.10341516149993386,
      "grad_norm": 0.1514318585395813,
      "learning_rate": 9.655595424295397e-05,
      "loss": 0.7736,
      "mean_token_accuracy": 0.7792934477329254,
      "num_tokens": 35868888.0,
      "step": 1124
    },
    {
      "entropy": 0.9165477007627487,
      "epoch": 0.1035071678713751,
      "grad_norm": 0.1574585735797882,
      "learning_rate": 9.655288741681235e-05,
      "loss": 0.8817,
      "mean_token_accuracy": 0.7515552639961243,
      "num_tokens": 35900505.0,
      "step": 1125
    },
    {
      "entropy": 0.93998783826828,
      "epoch": 0.10359917424281631,
      "grad_norm": 0.16181664168834686,
      "learning_rate": 9.654982059067072e-05,
      "loss": 0.9158,
      "mean_token_accuracy": 0.7416956163942814,
      "num_tokens": 35933020.0,
      "step": 1126
    },
    {
      "entropy": 0.8888120427727699,
      "epoch": 0.10369118061425754,
      "grad_norm": 0.15146636962890625,
      "learning_rate": 9.65467537645291e-05,
      "loss": 0.8702,
      "mean_token_accuracy": 0.7530062273144722,
      "num_tokens": 35965251.0,
      "step": 1127
    },
    {
      "entropy": 0.9545511826872826,
      "epoch": 0.10378318698569876,
      "grad_norm": 0.16361840069293976,
      "learning_rate": 9.654368693838747e-05,
      "loss": 0.9426,
      "mean_token_accuracy": 0.7365666106343269,
      "num_tokens": 35996389.0,
      "step": 1128
    },
    {
      "entropy": 0.8557638861238956,
      "epoch": 0.10387519335713999,
      "grad_norm": 0.16363433003425598,
      "learning_rate": 9.654062011224583e-05,
      "loss": 0.8516,
      "mean_token_accuracy": 0.7548157721757889,
      "num_tokens": 36027510.0,
      "step": 1129
    },
    {
      "entropy": 0.8681304268538952,
      "epoch": 0.1039671997285812,
      "grad_norm": 0.16204281151294708,
      "learning_rate": 9.653755328610422e-05,
      "loss": 0.8938,
      "mean_token_accuracy": 0.7480642050504684,
      "num_tokens": 36058249.0,
      "step": 1130
    },
    {
      "entropy": 0.9061204828321934,
      "epoch": 0.10405920610002242,
      "grad_norm": 0.16870051622390747,
      "learning_rate": 9.65344864599626e-05,
      "loss": 0.9272,
      "mean_token_accuracy": 0.7406693212687969,
      "num_tokens": 36089904.0,
      "step": 1131
    },
    {
      "entropy": 0.8978628180921078,
      "epoch": 0.10415121247146365,
      "grad_norm": 0.2717898190021515,
      "learning_rate": 9.653141963382097e-05,
      "loss": 0.9002,
      "mean_token_accuracy": 0.7447415627539158,
      "num_tokens": 36121509.0,
      "step": 1132
    },
    {
      "entropy": 0.9479461200535297,
      "epoch": 0.10424321884290487,
      "grad_norm": 0.16139201819896698,
      "learning_rate": 9.652835280767933e-05,
      "loss": 0.9532,
      "mean_token_accuracy": 0.7297844812273979,
      "num_tokens": 36152996.0,
      "step": 1133
    },
    {
      "entropy": 0.8584220185875893,
      "epoch": 0.1043352252143461,
      "grad_norm": 0.15595276653766632,
      "learning_rate": 9.652528598153771e-05,
      "loss": 0.8545,
      "mean_token_accuracy": 0.757264468818903,
      "num_tokens": 36185554.0,
      "step": 1134
    },
    {
      "entropy": 0.8959417846053839,
      "epoch": 0.10442723158578732,
      "grad_norm": 0.16639848053455353,
      "learning_rate": 9.652221915539608e-05,
      "loss": 0.8955,
      "mean_token_accuracy": 0.7450969852507114,
      "num_tokens": 36217117.0,
      "step": 1135
    },
    {
      "entropy": 0.8949821796268225,
      "epoch": 0.10451923795722853,
      "grad_norm": 0.16121737658977509,
      "learning_rate": 9.651915232925446e-05,
      "loss": 0.9081,
      "mean_token_accuracy": 0.7438309378921986,
      "num_tokens": 36249429.0,
      "step": 1136
    },
    {
      "entropy": 0.8880996480584145,
      "epoch": 0.10461124432866976,
      "grad_norm": 0.1555497944355011,
      "learning_rate": 9.651608550311283e-05,
      "loss": 0.8868,
      "mean_token_accuracy": 0.7501143775880337,
      "num_tokens": 36281270.0,
      "step": 1137
    },
    {
      "entropy": 0.8846673518419266,
      "epoch": 0.10470325070011098,
      "grad_norm": 0.16152282059192657,
      "learning_rate": 9.651301867697121e-05,
      "loss": 0.8851,
      "mean_token_accuracy": 0.7479728311300278,
      "num_tokens": 36313014.0,
      "step": 1138
    },
    {
      "entropy": 0.9909635446965694,
      "epoch": 0.10479525707155221,
      "grad_norm": 0.18244989216327667,
      "learning_rate": 9.650995185082958e-05,
      "loss": 0.9985,
      "mean_token_accuracy": 0.7193295061588287,
      "num_tokens": 36344775.0,
      "step": 1139
    },
    {
      "entropy": 0.8116722833365202,
      "epoch": 0.10488726344299343,
      "grad_norm": 0.1561717987060547,
      "learning_rate": 9.650688502468795e-05,
      "loss": 0.8232,
      "mean_token_accuracy": 0.7624444738030434,
      "num_tokens": 36377368.0,
      "step": 1140
    },
    {
      "entropy": 0.962451845407486,
      "epoch": 0.10497926981443464,
      "grad_norm": 0.16757145524024963,
      "learning_rate": 9.650381819854633e-05,
      "loss": 0.9895,
      "mean_token_accuracy": 0.7306251116096973,
      "num_tokens": 36409033.0,
      "step": 1141
    },
    {
      "entropy": 0.8374790716916323,
      "epoch": 0.10507127618587588,
      "grad_norm": 0.15268345177173615,
      "learning_rate": 9.650075137240471e-05,
      "loss": 0.8403,
      "mean_token_accuracy": 0.7610041983425617,
      "num_tokens": 36441629.0,
      "step": 1142
    },
    {
      "entropy": 0.8009271528571844,
      "epoch": 0.10516328255731709,
      "grad_norm": 0.15544068813323975,
      "learning_rate": 9.649768454626308e-05,
      "loss": 0.8178,
      "mean_token_accuracy": 0.7638004496693611,
      "num_tokens": 36473992.0,
      "step": 1143
    },
    {
      "entropy": 0.918566320091486,
      "epoch": 0.10525528892875832,
      "grad_norm": 0.16062887012958527,
      "learning_rate": 9.649461772012145e-05,
      "loss": 0.9113,
      "mean_token_accuracy": 0.7453842647373676,
      "num_tokens": 36505679.0,
      "step": 1144
    },
    {
      "entropy": 0.8723691739141941,
      "epoch": 0.10534729530019954,
      "grad_norm": 0.16232752799987793,
      "learning_rate": 9.649155089397982e-05,
      "loss": 0.8595,
      "mean_token_accuracy": 0.7579716853797436,
      "num_tokens": 36537351.0,
      "step": 1145
    },
    {
      "entropy": 0.9807293154299259,
      "epoch": 0.10543930167164076,
      "grad_norm": 0.17774683237075806,
      "learning_rate": 9.64884840678382e-05,
      "loss": 0.9835,
      "mean_token_accuracy": 0.7251734808087349,
      "num_tokens": 36568764.0,
      "step": 1146
    },
    {
      "entropy": 0.9323286805301905,
      "epoch": 0.10553130804308199,
      "grad_norm": 0.1614963263273239,
      "learning_rate": 9.648541724169658e-05,
      "loss": 0.9247,
      "mean_token_accuracy": 0.7368593029677868,
      "num_tokens": 36600905.0,
      "step": 1147
    },
    {
      "entropy": 0.8306343927979469,
      "epoch": 0.1056233144145232,
      "grad_norm": 0.15686990320682526,
      "learning_rate": 9.648235041555495e-05,
      "loss": 0.8187,
      "mean_token_accuracy": 0.7701854668557644,
      "num_tokens": 36632643.0,
      "step": 1148
    },
    {
      "entropy": 0.8811053596436977,
      "epoch": 0.10571532078596443,
      "grad_norm": 0.1646854132413864,
      "learning_rate": 9.647928358941331e-05,
      "loss": 0.8994,
      "mean_token_accuracy": 0.7448853179812431,
      "num_tokens": 36664486.0,
      "step": 1149
    },
    {
      "entropy": 0.8647363968193531,
      "epoch": 0.10580732715740565,
      "grad_norm": 0.15757900476455688,
      "learning_rate": 9.64762167632717e-05,
      "loss": 0.8676,
      "mean_token_accuracy": 0.7621036171913147,
      "num_tokens": 36696515.0,
      "step": 1150
    },
    {
      "entropy": 0.992485873401165,
      "epoch": 0.10589933352884687,
      "grad_norm": 0.16723927855491638,
      "learning_rate": 9.647314993713006e-05,
      "loss": 0.9842,
      "mean_token_accuracy": 0.7219220995903015,
      "num_tokens": 36728221.0,
      "step": 1151
    },
    {
      "entropy": 0.9119985029101372,
      "epoch": 0.1059913399002881,
      "grad_norm": 0.15609526634216309,
      "learning_rate": 9.647008311098844e-05,
      "loss": 0.8892,
      "mean_token_accuracy": 0.7492692843079567,
      "num_tokens": 36760176.0,
      "step": 1152
    },
    {
      "entropy": 0.9725981503725052,
      "epoch": 0.10608334627172931,
      "grad_norm": 0.1750059425830841,
      "learning_rate": 9.646701628484681e-05,
      "loss": 0.9895,
      "mean_token_accuracy": 0.7284197695553303,
      "num_tokens": 36792372.0,
      "step": 1153
    },
    {
      "entropy": 0.8699761964380741,
      "epoch": 0.10617535264317055,
      "grad_norm": 0.16103379428386688,
      "learning_rate": 9.64639494587052e-05,
      "loss": 0.8832,
      "mean_token_accuracy": 0.7475852482020855,
      "num_tokens": 36824450.0,
      "step": 1154
    },
    {
      "entropy": 0.9101373143494129,
      "epoch": 0.10626735901461176,
      "grad_norm": 0.1653328537940979,
      "learning_rate": 9.646088263256356e-05,
      "loss": 0.9156,
      "mean_token_accuracy": 0.7380640543997288,
      "num_tokens": 36856695.0,
      "step": 1155
    },
    {
      "entropy": 0.8482017517089844,
      "epoch": 0.10635936538605298,
      "grad_norm": 0.1592244803905487,
      "learning_rate": 9.645781580642193e-05,
      "loss": 0.8425,
      "mean_token_accuracy": 0.7571935392916203,
      "num_tokens": 36888373.0,
      "step": 1156
    },
    {
      "entropy": 0.9492213726043701,
      "epoch": 0.10645137175749421,
      "grad_norm": 0.16343535482883453,
      "learning_rate": 9.645474898028031e-05,
      "loss": 0.9502,
      "mean_token_accuracy": 0.7316171824932098,
      "num_tokens": 36919868.0,
      "step": 1157
    },
    {
      "entropy": 0.9440405517816544,
      "epoch": 0.10654337812893543,
      "grad_norm": 0.16608335077762604,
      "learning_rate": 9.645168215413869e-05,
      "loss": 0.9754,
      "mean_token_accuracy": 0.7324131540954113,
      "num_tokens": 36951911.0,
      "step": 1158
    },
    {
      "entropy": 0.9583054333925247,
      "epoch": 0.10663538450037666,
      "grad_norm": 0.1622367799282074,
      "learning_rate": 9.644861532799706e-05,
      "loss": 0.9587,
      "mean_token_accuracy": 0.7319944314658642,
      "num_tokens": 36984530.0,
      "step": 1159
    },
    {
      "entropy": 0.868174958974123,
      "epoch": 0.10672739087181787,
      "grad_norm": 0.157622829079628,
      "learning_rate": 9.644554850185543e-05,
      "loss": 0.8731,
      "mean_token_accuracy": 0.7521060034632683,
      "num_tokens": 37017181.0,
      "step": 1160
    },
    {
      "entropy": 0.9422637708485126,
      "epoch": 0.10681939724325909,
      "grad_norm": 0.16432589292526245,
      "learning_rate": 9.644248167571381e-05,
      "loss": 0.9224,
      "mean_token_accuracy": 0.7373744137585163,
      "num_tokens": 37048535.0,
      "step": 1161
    },
    {
      "entropy": 0.9234974272549152,
      "epoch": 0.10691140361470032,
      "grad_norm": 0.16373319923877716,
      "learning_rate": 9.643941484957219e-05,
      "loss": 0.9346,
      "mean_token_accuracy": 0.7404068410396576,
      "num_tokens": 37079810.0,
      "step": 1162
    },
    {
      "entropy": 0.9439622424542904,
      "epoch": 0.10700340998614154,
      "grad_norm": 0.163079634308815,
      "learning_rate": 9.643634802343056e-05,
      "loss": 0.9431,
      "mean_token_accuracy": 0.7391014732420444,
      "num_tokens": 37112039.0,
      "step": 1163
    },
    {
      "entropy": 0.8304878268390894,
      "epoch": 0.10709541635758277,
      "grad_norm": 0.15716643631458282,
      "learning_rate": 9.643328119728893e-05,
      "loss": 0.8298,
      "mean_token_accuracy": 0.7615072317421436,
      "num_tokens": 37143681.0,
      "step": 1164
    },
    {
      "entropy": 0.9277613312005997,
      "epoch": 0.10718742272902398,
      "grad_norm": 0.16060766577720642,
      "learning_rate": 9.643021437114731e-05,
      "loss": 0.9065,
      "mean_token_accuracy": 0.7429410293698311,
      "num_tokens": 37174208.0,
      "step": 1165
    },
    {
      "entropy": 0.9364616386592388,
      "epoch": 0.1072794291004652,
      "grad_norm": 0.16461218893527985,
      "learning_rate": 9.642714754500568e-05,
      "loss": 0.9367,
      "mean_token_accuracy": 0.7390764616429806,
      "num_tokens": 37206827.0,
      "step": 1166
    },
    {
      "entropy": 0.9196576457470655,
      "epoch": 0.10737143547190643,
      "grad_norm": 0.16141366958618164,
      "learning_rate": 9.642408071886406e-05,
      "loss": 0.9205,
      "mean_token_accuracy": 0.7377213537693024,
      "num_tokens": 37238894.0,
      "step": 1167
    },
    {
      "entropy": 1.0235708802938461,
      "epoch": 0.10746344184334765,
      "grad_norm": 0.1670745611190796,
      "learning_rate": 9.642101389272243e-05,
      "loss": 1.0262,
      "mean_token_accuracy": 0.7184600159525871,
      "num_tokens": 37271310.0,
      "step": 1168
    },
    {
      "entropy": 0.7872585002332926,
      "epoch": 0.10755544821478888,
      "grad_norm": 0.15099723637104034,
      "learning_rate": 9.641794706658081e-05,
      "loss": 0.8028,
      "mean_token_accuracy": 0.7691438607871532,
      "num_tokens": 37304078.0,
      "step": 1169
    },
    {
      "entropy": 0.9138952381908894,
      "epoch": 0.1076474545862301,
      "grad_norm": 0.15415091812610626,
      "learning_rate": 9.641488024043917e-05,
      "loss": 0.9013,
      "mean_token_accuracy": 0.7451988831162453,
      "num_tokens": 37336432.0,
      "step": 1170
    },
    {
      "entropy": 0.9151985794305801,
      "epoch": 0.10773946095767131,
      "grad_norm": 0.15855877101421356,
      "learning_rate": 9.641181341429754e-05,
      "loss": 0.9046,
      "mean_token_accuracy": 0.7472184859216213,
      "num_tokens": 37368392.0,
      "step": 1171
    },
    {
      "entropy": 0.8853078465908766,
      "epoch": 0.10783146732911254,
      "grad_norm": 0.16170835494995117,
      "learning_rate": 9.640874658815592e-05,
      "loss": 0.9041,
      "mean_token_accuracy": 0.7445632927119732,
      "num_tokens": 37400677.0,
      "step": 1172
    },
    {
      "entropy": 0.8597267027944326,
      "epoch": 0.10792347370055376,
      "grad_norm": 0.166159525513649,
      "learning_rate": 9.64056797620143e-05,
      "loss": 0.8734,
      "mean_token_accuracy": 0.7534822970628738,
      "num_tokens": 37432620.0,
      "step": 1173
    },
    {
      "entropy": 0.9492680914700031,
      "epoch": 0.10801548007199499,
      "grad_norm": 0.16760988533496857,
      "learning_rate": 9.640261293587267e-05,
      "loss": 0.9736,
      "mean_token_accuracy": 0.7311322875320911,
      "num_tokens": 37464018.0,
      "step": 1174
    },
    {
      "entropy": 0.8578727580606937,
      "epoch": 0.10810748644343621,
      "grad_norm": 0.16266760230064392,
      "learning_rate": 9.639954610973104e-05,
      "loss": 0.8571,
      "mean_token_accuracy": 0.758011557161808,
      "num_tokens": 37495672.0,
      "step": 1175
    },
    {
      "entropy": 0.8890331462025642,
      "epoch": 0.10819949281487742,
      "grad_norm": 0.16001437604427338,
      "learning_rate": 9.639647928358941e-05,
      "loss": 0.8901,
      "mean_token_accuracy": 0.7485282197594643,
      "num_tokens": 37527105.0,
      "step": 1176
    },
    {
      "entropy": 0.9647010564804077,
      "epoch": 0.10829149918631865,
      "grad_norm": 0.16460822522640228,
      "learning_rate": 9.639341245744779e-05,
      "loss": 0.9489,
      "mean_token_accuracy": 0.7372633032500744,
      "num_tokens": 37558924.0,
      "step": 1177
    },
    {
      "entropy": 1.033889513462782,
      "epoch": 0.10838350555775987,
      "grad_norm": 0.16659210622310638,
      "learning_rate": 9.639034563130617e-05,
      "loss": 1.025,
      "mean_token_accuracy": 0.7165649831295013,
      "num_tokens": 37590682.0,
      "step": 1178
    },
    {
      "entropy": 1.024806346744299,
      "epoch": 0.1084755119292011,
      "grad_norm": 0.1633300483226776,
      "learning_rate": 9.638727880516454e-05,
      "loss": 1.0264,
      "mean_token_accuracy": 0.7159375883638859,
      "num_tokens": 37622737.0,
      "step": 1179
    },
    {
      "entropy": 0.9304421618580818,
      "epoch": 0.10856751830064232,
      "grad_norm": 0.15979604423046112,
      "learning_rate": 9.638421197902291e-05,
      "loss": 0.9211,
      "mean_token_accuracy": 0.7414129972457886,
      "num_tokens": 37654693.0,
      "step": 1180
    },
    {
      "entropy": 0.9538629502058029,
      "epoch": 0.10865952467208354,
      "grad_norm": 0.15762676298618317,
      "learning_rate": 9.638114515288129e-05,
      "loss": 0.9313,
      "mean_token_accuracy": 0.7405924797058105,
      "num_tokens": 37687044.0,
      "step": 1181
    },
    {
      "entropy": 0.8385770544409752,
      "epoch": 0.10875153104352477,
      "grad_norm": 0.15908768773078918,
      "learning_rate": 9.637807832673966e-05,
      "loss": 0.8366,
      "mean_token_accuracy": 0.7593835964798927,
      "num_tokens": 37717953.0,
      "step": 1182
    },
    {
      "entropy": 0.8700844012200832,
      "epoch": 0.10884353741496598,
      "grad_norm": 0.16258792579174042,
      "learning_rate": 9.637501150059804e-05,
      "loss": 0.8872,
      "mean_token_accuracy": 0.7474203854799271,
      "num_tokens": 37749844.0,
      "step": 1183
    },
    {
      "entropy": 0.9191421903669834,
      "epoch": 0.10893554378640721,
      "grad_norm": 0.1625436693429947,
      "learning_rate": 9.63719446744564e-05,
      "loss": 0.9557,
      "mean_token_accuracy": 0.7369426377117634,
      "num_tokens": 37781798.0,
      "step": 1184
    },
    {
      "entropy": 0.8815950453281403,
      "epoch": 0.10902755015784843,
      "grad_norm": 0.15976464748382568,
      "learning_rate": 9.636887784831479e-05,
      "loss": 0.8858,
      "mean_token_accuracy": 0.7485115937888622,
      "num_tokens": 37813935.0,
      "step": 1185
    },
    {
      "entropy": 0.9487125314772129,
      "epoch": 0.10911955652928966,
      "grad_norm": 0.16349990665912628,
      "learning_rate": 9.636581102217316e-05,
      "loss": 0.9678,
      "mean_token_accuracy": 0.7307746522128582,
      "num_tokens": 37846422.0,
      "step": 1186
    },
    {
      "entropy": 0.9389483779668808,
      "epoch": 0.10921156290073088,
      "grad_norm": 0.16426877677440643,
      "learning_rate": 9.636274419603152e-05,
      "loss": 0.9484,
      "mean_token_accuracy": 0.7360130548477173,
      "num_tokens": 37878200.0,
      "step": 1187
    },
    {
      "entropy": 0.8950897268950939,
      "epoch": 0.1093035692721721,
      "grad_norm": 0.16332906484603882,
      "learning_rate": 9.63596773698899e-05,
      "loss": 0.8917,
      "mean_token_accuracy": 0.7537476047873497,
      "num_tokens": 37910448.0,
      "step": 1188
    },
    {
      "entropy": 0.8959545940160751,
      "epoch": 0.10939557564361332,
      "grad_norm": 0.17104771733283997,
      "learning_rate": 9.635661054374829e-05,
      "loss": 0.8937,
      "mean_token_accuracy": 0.7442556731402874,
      "num_tokens": 37942282.0,
      "step": 1189
    },
    {
      "entropy": 0.9303874671459198,
      "epoch": 0.10948758201505454,
      "grad_norm": 0.16186782717704773,
      "learning_rate": 9.635354371760665e-05,
      "loss": 0.9164,
      "mean_token_accuracy": 0.7406597584486008,
      "num_tokens": 37973958.0,
      "step": 1190
    },
    {
      "entropy": 0.924135372042656,
      "epoch": 0.10957958838649577,
      "grad_norm": 0.15974308550357819,
      "learning_rate": 9.635047689146502e-05,
      "loss": 0.9065,
      "mean_token_accuracy": 0.7446037232875824,
      "num_tokens": 38005708.0,
      "step": 1191
    },
    {
      "entropy": 0.8299891129136086,
      "epoch": 0.10967159475793699,
      "grad_norm": 0.15791286528110504,
      "learning_rate": 9.63474100653234e-05,
      "loss": 0.8321,
      "mean_token_accuracy": 0.7613086141645908,
      "num_tokens": 38038396.0,
      "step": 1192
    },
    {
      "entropy": 0.8359809089452028,
      "epoch": 0.1097636011293782,
      "grad_norm": 0.15945649147033691,
      "learning_rate": 9.634434323918177e-05,
      "loss": 0.8415,
      "mean_token_accuracy": 0.7634959109127522,
      "num_tokens": 38070026.0,
      "step": 1193
    },
    {
      "entropy": 0.9454305730760098,
      "epoch": 0.10985560750081944,
      "grad_norm": 0.16911645233631134,
      "learning_rate": 9.634127641304015e-05,
      "loss": 0.9543,
      "mean_token_accuracy": 0.7351836785674095,
      "num_tokens": 38102014.0,
      "step": 1194
    },
    {
      "entropy": 0.8455300014466047,
      "epoch": 0.10994761387226065,
      "grad_norm": 0.15199436247348785,
      "learning_rate": 9.633820958689852e-05,
      "loss": 0.8417,
      "mean_token_accuracy": 0.7604790925979614,
      "num_tokens": 38133884.0,
      "step": 1195
    },
    {
      "entropy": 0.8646016959100962,
      "epoch": 0.11003962024370188,
      "grad_norm": 0.1596190184354782,
      "learning_rate": 9.63351427607569e-05,
      "loss": 0.8474,
      "mean_token_accuracy": 0.760076604783535,
      "num_tokens": 38165787.0,
      "step": 1196
    },
    {
      "entropy": 0.9365963451564312,
      "epoch": 0.1101316266151431,
      "grad_norm": 0.1643386334180832,
      "learning_rate": 9.633207593461527e-05,
      "loss": 0.934,
      "mean_token_accuracy": 0.7342923209071159,
      "num_tokens": 38197596.0,
      "step": 1197
    },
    {
      "entropy": 1.0257736630737782,
      "epoch": 0.11022363298658432,
      "grad_norm": 0.17303325235843658,
      "learning_rate": 9.632900910847364e-05,
      "loss": 1.0658,
      "mean_token_accuracy": 0.7098408564925194,
      "num_tokens": 38230098.0,
      "step": 1198
    },
    {
      "entropy": 1.0279016233980656,
      "epoch": 0.11031563935802555,
      "grad_norm": 0.1734132170677185,
      "learning_rate": 9.632594228233202e-05,
      "loss": 1.0374,
      "mean_token_accuracy": 0.710333913564682,
      "num_tokens": 38260742.0,
      "step": 1199
    },
    {
      "entropy": 0.988643130287528,
      "epoch": 0.11040764572946676,
      "grad_norm": 0.16288378834724426,
      "learning_rate": 9.63228754561904e-05,
      "loss": 0.9841,
      "mean_token_accuracy": 0.7274706810712814,
      "num_tokens": 38293322.0,
      "step": 1200
    },
    {
      "entropy": 0.9842981025576591,
      "epoch": 0.110499652100908,
      "grad_norm": 0.1583864539861679,
      "learning_rate": 9.631980863004877e-05,
      "loss": 0.9591,
      "mean_token_accuracy": 0.7328468821942806,
      "num_tokens": 38325707.0,
      "step": 1201
    },
    {
      "entropy": 0.9645696990191936,
      "epoch": 0.11059165847234921,
      "grad_norm": 0.15911807119846344,
      "learning_rate": 9.631674180390714e-05,
      "loss": 0.9437,
      "mean_token_accuracy": 0.7354324497282505,
      "num_tokens": 38357351.0,
      "step": 1202
    },
    {
      "entropy": 0.9264683276414871,
      "epoch": 0.11068366484379043,
      "grad_norm": 0.16384205222129822,
      "learning_rate": 9.63136749777655e-05,
      "loss": 0.9363,
      "mean_token_accuracy": 0.7381680496037006,
      "num_tokens": 38389933.0,
      "step": 1203
    },
    {
      "entropy": 0.9327795393764973,
      "epoch": 0.11077567121523166,
      "grad_norm": 0.1632637083530426,
      "learning_rate": 9.63106081516239e-05,
      "loss": 0.9215,
      "mean_token_accuracy": 0.7395609430968761,
      "num_tokens": 38421928.0,
      "step": 1204
    },
    {
      "entropy": 0.8881975449621677,
      "epoch": 0.11086767758667287,
      "grad_norm": 0.16038884222507477,
      "learning_rate": 9.630754132548227e-05,
      "loss": 0.8809,
      "mean_token_accuracy": 0.7521721497178078,
      "num_tokens": 38453863.0,
      "step": 1205
    },
    {
      "entropy": 0.8493070900440216,
      "epoch": 0.1109596839581141,
      "grad_norm": 0.16059279441833496,
      "learning_rate": 9.630447449934064e-05,
      "loss": 0.8581,
      "mean_token_accuracy": 0.7503038309514523,
      "num_tokens": 38485924.0,
      "step": 1206
    },
    {
      "entropy": 0.848507534712553,
      "epoch": 0.11105169032955532,
      "grad_norm": 0.15460970997810364,
      "learning_rate": 9.6301407673199e-05,
      "loss": 0.8599,
      "mean_token_accuracy": 0.7540242336690426,
      "num_tokens": 38517608.0,
      "step": 1207
    },
    {
      "entropy": 1.0172521360218525,
      "epoch": 0.11114369670099654,
      "grad_norm": 0.1650722771883011,
      "learning_rate": 9.629834084705738e-05,
      "loss": 1.0239,
      "mean_token_accuracy": 0.7180158123373985,
      "num_tokens": 38549788.0,
      "step": 1208
    },
    {
      "entropy": 0.9597753770649433,
      "epoch": 0.11123570307243777,
      "grad_norm": 0.16278965771198273,
      "learning_rate": 9.629527402091577e-05,
      "loss": 0.9551,
      "mean_token_accuracy": 0.731382705271244,
      "num_tokens": 38582101.0,
      "step": 1209
    },
    {
      "entropy": 0.9449393004179001,
      "epoch": 0.11132770944387899,
      "grad_norm": 0.16609980165958405,
      "learning_rate": 9.629220719477413e-05,
      "loss": 0.9441,
      "mean_token_accuracy": 0.7326384335756302,
      "num_tokens": 38613484.0,
      "step": 1210
    },
    {
      "entropy": 0.9704713299870491,
      "epoch": 0.11141971581532022,
      "grad_norm": 0.16518573462963104,
      "learning_rate": 9.62891403686325e-05,
      "loss": 0.9762,
      "mean_token_accuracy": 0.7278703674674034,
      "num_tokens": 38644771.0,
      "step": 1211
    },
    {
      "entropy": 0.9279946126043797,
      "epoch": 0.11151172218676143,
      "grad_norm": 0.16806304454803467,
      "learning_rate": 9.628607354249088e-05,
      "loss": 0.9362,
      "mean_token_accuracy": 0.734474454075098,
      "num_tokens": 38677020.0,
      "step": 1212
    },
    {
      "entropy": 0.8255847655236721,
      "epoch": 0.11160372855820265,
      "grad_norm": 0.15495117008686066,
      "learning_rate": 9.628300671634925e-05,
      "loss": 0.8265,
      "mean_token_accuracy": 0.7633616365492344,
      "num_tokens": 38708178.0,
      "step": 1213
    },
    {
      "entropy": 0.8723923414945602,
      "epoch": 0.11169573492964388,
      "grad_norm": 0.16134938597679138,
      "learning_rate": 9.627993989020763e-05,
      "loss": 0.851,
      "mean_token_accuracy": 0.7510189339518547,
      "num_tokens": 38739773.0,
      "step": 1214
    },
    {
      "entropy": 0.8964626677334309,
      "epoch": 0.1117877413010851,
      "grad_norm": 0.16615474224090576,
      "learning_rate": 9.6276873064066e-05,
      "loss": 0.9066,
      "mean_token_accuracy": 0.7433793321251869,
      "num_tokens": 38771738.0,
      "step": 1215
    },
    {
      "entropy": 0.9249050095677376,
      "epoch": 0.11187974767252633,
      "grad_norm": 0.16456910967826843,
      "learning_rate": 9.627380623792438e-05,
      "loss": 0.9332,
      "mean_token_accuracy": 0.7363257557153702,
      "num_tokens": 38803097.0,
      "step": 1216
    },
    {
      "entropy": 0.918994415551424,
      "epoch": 0.11197175404396754,
      "grad_norm": 0.16103795170783997,
      "learning_rate": 9.627073941178275e-05,
      "loss": 0.9282,
      "mean_token_accuracy": 0.7439801022410393,
      "num_tokens": 38835056.0,
      "step": 1217
    },
    {
      "entropy": 0.9190790541470051,
      "epoch": 0.11206376041540876,
      "grad_norm": 0.16347742080688477,
      "learning_rate": 9.626767258564112e-05,
      "loss": 0.9368,
      "mean_token_accuracy": 0.7371826469898224,
      "num_tokens": 38867074.0,
      "step": 1218
    },
    {
      "entropy": 0.8763457387685776,
      "epoch": 0.11215576678684999,
      "grad_norm": 0.15757107734680176,
      "learning_rate": 9.62646057594995e-05,
      "loss": 0.8705,
      "mean_token_accuracy": 0.752184446901083,
      "num_tokens": 38899315.0,
      "step": 1219
    },
    {
      "entropy": 0.8639146462082863,
      "epoch": 0.11224777315829121,
      "grad_norm": 0.15921445190906525,
      "learning_rate": 9.626153893335788e-05,
      "loss": 0.8377,
      "mean_token_accuracy": 0.7558577954769135,
      "num_tokens": 38930092.0,
      "step": 1220
    },
    {
      "entropy": 1.0135957710444927,
      "epoch": 0.11233977952973244,
      "grad_norm": 0.16591723263263702,
      "learning_rate": 9.625847210721625e-05,
      "loss": 1.0247,
      "mean_token_accuracy": 0.7142692767083645,
      "num_tokens": 38961755.0,
      "step": 1221
    },
    {
      "entropy": 0.9697911143302917,
      "epoch": 0.11243178590117366,
      "grad_norm": 0.15928500890731812,
      "learning_rate": 9.625540528107462e-05,
      "loss": 0.9569,
      "mean_token_accuracy": 0.7309919185936451,
      "num_tokens": 38993739.0,
      "step": 1222
    },
    {
      "entropy": 0.8771725594997406,
      "epoch": 0.11252379227261487,
      "grad_norm": 0.15610209107398987,
      "learning_rate": 9.6252338454933e-05,
      "loss": 0.8577,
      "mean_token_accuracy": 0.7529030852019787,
      "num_tokens": 39025075.0,
      "step": 1223
    },
    {
      "entropy": 0.825790349394083,
      "epoch": 0.1126157986440561,
      "grad_norm": 0.154636949300766,
      "learning_rate": 9.624927162879137e-05,
      "loss": 0.8387,
      "mean_token_accuracy": 0.7595239467918873,
      "num_tokens": 39057485.0,
      "step": 1224
    },
    {
      "entropy": 0.8548086732625961,
      "epoch": 0.11270780501549732,
      "grad_norm": 0.15643763542175293,
      "learning_rate": 9.624620480264975e-05,
      "loss": 0.8565,
      "mean_token_accuracy": 0.748670969158411,
      "num_tokens": 39089670.0,
      "step": 1225
    },
    {
      "entropy": 0.8604331947863102,
      "epoch": 0.11279981138693855,
      "grad_norm": 0.15389719605445862,
      "learning_rate": 9.624313797650811e-05,
      "loss": 0.8716,
      "mean_token_accuracy": 0.7539368383586407,
      "num_tokens": 39121159.0,
      "step": 1226
    },
    {
      "entropy": 0.8740938268601894,
      "epoch": 0.11289181775837977,
      "grad_norm": 0.16768302023410797,
      "learning_rate": 9.62400711503665e-05,
      "loss": 0.9015,
      "mean_token_accuracy": 0.7455953173339367,
      "num_tokens": 39153290.0,
      "step": 1227
    },
    {
      "entropy": 0.8753143921494484,
      "epoch": 0.11298382412982098,
      "grad_norm": 0.1602894365787506,
      "learning_rate": 9.623700432422486e-05,
      "loss": 0.8792,
      "mean_token_accuracy": 0.7519412450492382,
      "num_tokens": 39185073.0,
      "step": 1228
    },
    {
      "entropy": 0.9621951431035995,
      "epoch": 0.11307583050126221,
      "grad_norm": 0.17920830845832825,
      "learning_rate": 9.623393749808323e-05,
      "loss": 0.981,
      "mean_token_accuracy": 0.7270716913044453,
      "num_tokens": 39215803.0,
      "step": 1229
    },
    {
      "entropy": 0.95667290315032,
      "epoch": 0.11316783687270343,
      "grad_norm": 0.1612055003643036,
      "learning_rate": 9.623087067194161e-05,
      "loss": 0.9651,
      "mean_token_accuracy": 0.7321768999099731,
      "num_tokens": 39247716.0,
      "step": 1230
    },
    {
      "entropy": 0.8866273295134306,
      "epoch": 0.11325984324414466,
      "grad_norm": 0.16174910962581635,
      "learning_rate": 9.62278038458e-05,
      "loss": 0.8813,
      "mean_token_accuracy": 0.7499629780650139,
      "num_tokens": 39279558.0,
      "step": 1231
    },
    {
      "entropy": 0.9229472950100899,
      "epoch": 0.11335184961558588,
      "grad_norm": 0.1614387035369873,
      "learning_rate": 9.622473701965836e-05,
      "loss": 0.9411,
      "mean_token_accuracy": 0.7425826601684093,
      "num_tokens": 39311970.0,
      "step": 1232
    },
    {
      "entropy": 0.9434081166982651,
      "epoch": 0.1134438559870271,
      "grad_norm": 0.16200323402881622,
      "learning_rate": 9.622167019351673e-05,
      "loss": 0.9342,
      "mean_token_accuracy": 0.738484013825655,
      "num_tokens": 39343624.0,
      "step": 1233
    },
    {
      "entropy": 0.9549262374639511,
      "epoch": 0.11353586235846833,
      "grad_norm": 0.1632819026708603,
      "learning_rate": 9.62186033673751e-05,
      "loss": 0.952,
      "mean_token_accuracy": 0.7290073409676552,
      "num_tokens": 39375888.0,
      "step": 1234
    },
    {
      "entropy": 0.8881600592285395,
      "epoch": 0.11362786872990954,
      "grad_norm": 0.16566340625286102,
      "learning_rate": 9.621553654123348e-05,
      "loss": 0.8993,
      "mean_token_accuracy": 0.7481296434998512,
      "num_tokens": 39407285.0,
      "step": 1235
    },
    {
      "entropy": 0.871746014803648,
      "epoch": 0.11371987510135077,
      "grad_norm": 0.15194818377494812,
      "learning_rate": 9.621246971509186e-05,
      "loss": 0.8383,
      "mean_token_accuracy": 0.7641799114644527,
      "num_tokens": 39439167.0,
      "step": 1236
    },
    {
      "entropy": 0.9456565119326115,
      "epoch": 0.11381188147279199,
      "grad_norm": 0.16021110117435455,
      "learning_rate": 9.620940288895023e-05,
      "loss": 0.9454,
      "mean_token_accuracy": 0.7346018217504025,
      "num_tokens": 39471680.0,
      "step": 1237
    },
    {
      "entropy": 0.9114279672503471,
      "epoch": 0.1139038878442332,
      "grad_norm": 0.15624144673347473,
      "learning_rate": 9.62063360628086e-05,
      "loss": 0.9113,
      "mean_token_accuracy": 0.741263385862112,
      "num_tokens": 39503424.0,
      "step": 1238
    },
    {
      "entropy": 0.8248308431357145,
      "epoch": 0.11399589421567444,
      "grad_norm": 0.15364912152290344,
      "learning_rate": 9.620326923666698e-05,
      "loss": 0.8172,
      "mean_token_accuracy": 0.7631268091499805,
      "num_tokens": 39535084.0,
      "step": 1239
    },
    {
      "entropy": 0.9492751248180866,
      "epoch": 0.11408790058711565,
      "grad_norm": 0.16309025883674622,
      "learning_rate": 9.620020241052536e-05,
      "loss": 0.9656,
      "mean_token_accuracy": 0.7299348637461662,
      "num_tokens": 39566638.0,
      "step": 1240
    },
    {
      "entropy": 0.9135283194482327,
      "epoch": 0.11417990695855688,
      "grad_norm": 0.1669679433107376,
      "learning_rate": 9.619713558438373e-05,
      "loss": 0.9347,
      "mean_token_accuracy": 0.7380702719092369,
      "num_tokens": 39598552.0,
      "step": 1241
    },
    {
      "entropy": 0.8984729759395123,
      "epoch": 0.1142719133299981,
      "grad_norm": 0.15988226234912872,
      "learning_rate": 9.61940687582421e-05,
      "loss": 0.8949,
      "mean_token_accuracy": 0.7479907236993313,
      "num_tokens": 39629386.0,
      "step": 1242
    },
    {
      "entropy": 1.0124025158584118,
      "epoch": 0.11436391970143933,
      "grad_norm": 0.1736791580915451,
      "learning_rate": 9.619100193210048e-05,
      "loss": 1.0516,
      "mean_token_accuracy": 0.7104871943593025,
      "num_tokens": 39661583.0,
      "step": 1243
    },
    {
      "entropy": 0.8712000399827957,
      "epoch": 0.11445592607288055,
      "grad_norm": 0.15194231271743774,
      "learning_rate": 9.618793510595884e-05,
      "loss": 0.8454,
      "mean_token_accuracy": 0.760894950479269,
      "num_tokens": 39694028.0,
      "step": 1244
    },
    {
      "entropy": 0.9235191307961941,
      "epoch": 0.11454793244432177,
      "grad_norm": 0.16357599198818207,
      "learning_rate": 9.618486827981723e-05,
      "loss": 0.927,
      "mean_token_accuracy": 0.7381685860455036,
      "num_tokens": 39726391.0,
      "step": 1245
    },
    {
      "entropy": 0.987502921372652,
      "epoch": 0.114639938815763,
      "grad_norm": 0.16443833708763123,
      "learning_rate": 9.61818014536756e-05,
      "loss": 0.9856,
      "mean_token_accuracy": 0.7266838662326336,
      "num_tokens": 39758801.0,
      "step": 1246
    },
    {
      "entropy": 0.9411337189376354,
      "epoch": 0.11473194518720421,
      "grad_norm": 0.15912885963916779,
      "learning_rate": 9.617873462753398e-05,
      "loss": 0.9447,
      "mean_token_accuracy": 0.7357535995543003,
      "num_tokens": 39790931.0,
      "step": 1247
    },
    {
      "entropy": 0.8603390119969845,
      "epoch": 0.11482395155864544,
      "grad_norm": 0.159799724817276,
      "learning_rate": 9.617566780139234e-05,
      "loss": 0.8542,
      "mean_token_accuracy": 0.7572375386953354,
      "num_tokens": 39823298.0,
      "step": 1248
    },
    {
      "entropy": 1.0349065996706486,
      "epoch": 0.11491595793008666,
      "grad_norm": 0.16297271847724915,
      "learning_rate": 9.617260097525071e-05,
      "loss": 1.0412,
      "mean_token_accuracy": 0.7150174602866173,
      "num_tokens": 39855468.0,
      "step": 1249
    },
    {
      "entropy": 0.9715097993612289,
      "epoch": 0.11500796430152788,
      "grad_norm": 0.16377684473991394,
      "learning_rate": 9.616953414910909e-05,
      "loss": 0.9676,
      "mean_token_accuracy": 0.7303238995373249,
      "num_tokens": 39887506.0,
      "step": 1250
    },
    {
      "entropy": 0.9331888128072023,
      "epoch": 0.1150999706729691,
      "grad_norm": 0.15661361813545227,
      "learning_rate": 9.616646732296747e-05,
      "loss": 0.9508,
      "mean_token_accuracy": 0.7338228896260262,
      "num_tokens": 39918763.0,
      "step": 1251
    },
    {
      "entropy": 0.9242482483386993,
      "epoch": 0.11519197704441032,
      "grad_norm": 0.15961304306983948,
      "learning_rate": 9.616340049682584e-05,
      "loss": 0.9359,
      "mean_token_accuracy": 0.7375055626034737,
      "num_tokens": 39950347.0,
      "step": 1252
    },
    {
      "entropy": 0.794408243149519,
      "epoch": 0.11528398341585155,
      "grad_norm": 0.1502804160118103,
      "learning_rate": 9.616033367068421e-05,
      "loss": 0.7885,
      "mean_token_accuracy": 0.7726824283599854,
      "num_tokens": 39981713.0,
      "step": 1253
    },
    {
      "entropy": 0.8597808182239532,
      "epoch": 0.11537598978729277,
      "grad_norm": 0.1505114585161209,
      "learning_rate": 9.615726684454259e-05,
      "loss": 0.8415,
      "mean_token_accuracy": 0.7574262320995331,
      "num_tokens": 40013292.0,
      "step": 1254
    },
    {
      "entropy": 0.8962214607745409,
      "epoch": 0.11546799615873399,
      "grad_norm": 0.16228564083576202,
      "learning_rate": 9.615420001840096e-05,
      "loss": 0.9059,
      "mean_token_accuracy": 0.7461973503232002,
      "num_tokens": 40043801.0,
      "step": 1255
    },
    {
      "entropy": 0.9787055291235447,
      "epoch": 0.11556000253017522,
      "grad_norm": 0.16328176856040955,
      "learning_rate": 9.615113319225934e-05,
      "loss": 0.9716,
      "mean_token_accuracy": 0.7294313348829746,
      "num_tokens": 40075473.0,
      "step": 1256
    },
    {
      "entropy": 0.941753227263689,
      "epoch": 0.11565200890161643,
      "grad_norm": 0.1578305959701538,
      "learning_rate": 9.614806636611771e-05,
      "loss": 0.9356,
      "mean_token_accuracy": 0.7382269948720932,
      "num_tokens": 40107421.0,
      "step": 1257
    },
    {
      "entropy": 0.9922678060829639,
      "epoch": 0.11574401527305767,
      "grad_norm": 0.16627971827983856,
      "learning_rate": 9.614499953997609e-05,
      "loss": 0.9886,
      "mean_token_accuracy": 0.7248015590012074,
      "num_tokens": 40139420.0,
      "step": 1258
    },
    {
      "entropy": 0.9118014425039291,
      "epoch": 0.11583602164449888,
      "grad_norm": 0.16349591314792633,
      "learning_rate": 9.614193271383446e-05,
      "loss": 0.9499,
      "mean_token_accuracy": 0.7369318641722202,
      "num_tokens": 40171487.0,
      "step": 1259
    },
    {
      "entropy": 0.8469453938305378,
      "epoch": 0.1159280280159401,
      "grad_norm": 0.15751893818378448,
      "learning_rate": 9.613886588769283e-05,
      "loss": 0.8674,
      "mean_token_accuracy": 0.7554065398871899,
      "num_tokens": 40203466.0,
      "step": 1260
    },
    {
      "entropy": 0.9490509815514088,
      "epoch": 0.11602003438738133,
      "grad_norm": 0.1600155383348465,
      "learning_rate": 9.613579906155121e-05,
      "loss": 0.9643,
      "mean_token_accuracy": 0.734326284378767,
      "num_tokens": 40235607.0,
      "step": 1261
    },
    {
      "entropy": 0.9248876925557852,
      "epoch": 0.11611204075882255,
      "grad_norm": 0.16124874353408813,
      "learning_rate": 9.613273223540959e-05,
      "loss": 0.9275,
      "mean_token_accuracy": 0.7398980371654034,
      "num_tokens": 40267601.0,
      "step": 1262
    },
    {
      "entropy": 0.9632008746266365,
      "epoch": 0.11620404713026378,
      "grad_norm": 0.15711267292499542,
      "learning_rate": 9.612966540926796e-05,
      "loss": 0.9363,
      "mean_token_accuracy": 0.7372532747685909,
      "num_tokens": 40299802.0,
      "step": 1263
    },
    {
      "entropy": 0.9267589189112186,
      "epoch": 0.116296053501705,
      "grad_norm": 0.16260387003421783,
      "learning_rate": 9.612659858312632e-05,
      "loss": 0.9283,
      "mean_token_accuracy": 0.73813172057271,
      "num_tokens": 40331546.0,
      "step": 1264
    },
    {
      "entropy": 1.013652851805091,
      "epoch": 0.11638805987314621,
      "grad_norm": 0.1654907613992691,
      "learning_rate": 9.612353175698469e-05,
      "loss": 1.0023,
      "mean_token_accuracy": 0.7146297432482243,
      "num_tokens": 40363659.0,
      "step": 1265
    },
    {
      "entropy": 0.9678255245089531,
      "epoch": 0.11648006624458744,
      "grad_norm": 0.1616593450307846,
      "learning_rate": 9.612046493084307e-05,
      "loss": 0.9467,
      "mean_token_accuracy": 0.7359376661479473,
      "num_tokens": 40395123.0,
      "step": 1266
    },
    {
      "entropy": 1.0728706941008568,
      "epoch": 0.11657207261602866,
      "grad_norm": 0.16114768385887146,
      "learning_rate": 9.611739810470145e-05,
      "loss": 1.0705,
      "mean_token_accuracy": 0.7078097350895405,
      "num_tokens": 40427194.0,
      "step": 1267
    },
    {
      "entropy": 0.9485188946127892,
      "epoch": 0.11666407898746989,
      "grad_norm": 0.16422154009342194,
      "learning_rate": 9.611433127855982e-05,
      "loss": 0.9543,
      "mean_token_accuracy": 0.7288158871233463,
      "num_tokens": 40459108.0,
      "step": 1268
    },
    {
      "entropy": 0.9691872745752335,
      "epoch": 0.1167560853589111,
      "grad_norm": 0.1650066077709198,
      "learning_rate": 9.611126445241819e-05,
      "loss": 0.9736,
      "mean_token_accuracy": 0.7273393124341965,
      "num_tokens": 40491164.0,
      "step": 1269
    },
    {
      "entropy": 0.8388831131160259,
      "epoch": 0.11684809173035232,
      "grad_norm": 0.1474425196647644,
      "learning_rate": 9.610819762627657e-05,
      "loss": 0.8015,
      "mean_token_accuracy": 0.7715040296316147,
      "num_tokens": 40523076.0,
      "step": 1270
    },
    {
      "entropy": 0.8737820871174335,
      "epoch": 0.11694009810179355,
      "grad_norm": 0.15791308879852295,
      "learning_rate": 9.610513080013494e-05,
      "loss": 0.8688,
      "mean_token_accuracy": 0.7504031881690025,
      "num_tokens": 40554759.0,
      "step": 1271
    },
    {
      "entropy": 0.9471858963370323,
      "epoch": 0.11703210447323477,
      "grad_norm": 0.1593601405620575,
      "learning_rate": 9.610206397399332e-05,
      "loss": 0.9522,
      "mean_token_accuracy": 0.7352076917886734,
      "num_tokens": 40587290.0,
      "step": 1272
    },
    {
      "entropy": 0.8698337860405445,
      "epoch": 0.117124110844676,
      "grad_norm": 0.15851740539073944,
      "learning_rate": 9.609899714785169e-05,
      "loss": 0.8806,
      "mean_token_accuracy": 0.7459563165903091,
      "num_tokens": 40618369.0,
      "step": 1273
    },
    {
      "entropy": 0.8455539979040623,
      "epoch": 0.11721611721611722,
      "grad_norm": 0.15500161051750183,
      "learning_rate": 9.609593032171007e-05,
      "loss": 0.8518,
      "mean_token_accuracy": 0.7552744783461094,
      "num_tokens": 40650623.0,
      "step": 1274
    },
    {
      "entropy": 0.9396741762757301,
      "epoch": 0.11730812358755843,
      "grad_norm": 0.16230222582817078,
      "learning_rate": 9.609286349556844e-05,
      "loss": 0.9398,
      "mean_token_accuracy": 0.7304256968200207,
      "num_tokens": 40681877.0,
      "step": 1275
    },
    {
      "entropy": 0.9025632776319981,
      "epoch": 0.11740012995899966,
      "grad_norm": 0.16150829195976257,
      "learning_rate": 9.60897966694268e-05,
      "loss": 0.932,
      "mean_token_accuracy": 0.7390985749661922,
      "num_tokens": 40714105.0,
      "step": 1276
    },
    {
      "entropy": 0.8881730511784554,
      "epoch": 0.11749213633044088,
      "grad_norm": 0.15971910953521729,
      "learning_rate": 9.608672984328519e-05,
      "loss": 0.9214,
      "mean_token_accuracy": 0.7416803315281868,
      "num_tokens": 40745946.0,
      "step": 1277
    },
    {
      "entropy": 0.8691030815243721,
      "epoch": 0.11758414270188211,
      "grad_norm": 0.1657966822385788,
      "learning_rate": 9.608366301714357e-05,
      "loss": 0.8982,
      "mean_token_accuracy": 0.7390356659889221,
      "num_tokens": 40777794.0,
      "step": 1278
    },
    {
      "entropy": 0.9578011557459831,
      "epoch": 0.11767614907332333,
      "grad_norm": 0.16132250428199768,
      "learning_rate": 9.608059619100194e-05,
      "loss": 0.9568,
      "mean_token_accuracy": 0.731116134673357,
      "num_tokens": 40809362.0,
      "step": 1279
    },
    {
      "entropy": 0.8265602830797434,
      "epoch": 0.11776815544476454,
      "grad_norm": 0.15483814477920532,
      "learning_rate": 9.60775293648603e-05,
      "loss": 0.8246,
      "mean_token_accuracy": 0.7657246440649033,
      "num_tokens": 40841879.0,
      "step": 1280
    },
    {
      "entropy": 0.8498229719698429,
      "epoch": 0.11786016181620577,
      "grad_norm": 0.16051693260669708,
      "learning_rate": 9.607446253871867e-05,
      "loss": 0.8458,
      "mean_token_accuracy": 0.7561776153743267,
      "num_tokens": 40873793.0,
      "step": 1281
    },
    {
      "entropy": 0.8779421057552099,
      "epoch": 0.11795216818764699,
      "grad_norm": 0.15922938287258148,
      "learning_rate": 9.607139571257707e-05,
      "loss": 0.863,
      "mean_token_accuracy": 0.7531022801995277,
      "num_tokens": 40905840.0,
      "step": 1282
    },
    {
      "entropy": 0.9131678827106953,
      "epoch": 0.11804417455908822,
      "grad_norm": 0.15531784296035767,
      "learning_rate": 9.606832888643544e-05,
      "loss": 0.9046,
      "mean_token_accuracy": 0.7454161271452904,
      "num_tokens": 40938294.0,
      "step": 1283
    },
    {
      "entropy": 0.9142397008836269,
      "epoch": 0.11813618093052944,
      "grad_norm": 0.16224299371242523,
      "learning_rate": 9.60652620602938e-05,
      "loss": 0.9021,
      "mean_token_accuracy": 0.7436867207288742,
      "num_tokens": 40969805.0,
      "step": 1284
    },
    {
      "entropy": 0.8809251487255096,
      "epoch": 0.11822818730197066,
      "grad_norm": 0.1542155146598816,
      "learning_rate": 9.606219523415218e-05,
      "loss": 0.8736,
      "mean_token_accuracy": 0.7532003335654736,
      "num_tokens": 41002092.0,
      "step": 1285
    },
    {
      "entropy": 0.9415201731026173,
      "epoch": 0.11832019367341189,
      "grad_norm": 0.15238253772258759,
      "learning_rate": 9.605912840801055e-05,
      "loss": 0.921,
      "mean_token_accuracy": 0.7392208613455296,
      "num_tokens": 41034637.0,
      "step": 1286
    },
    {
      "entropy": 0.8537130504846573,
      "epoch": 0.1184122000448531,
      "grad_norm": 0.1576310694217682,
      "learning_rate": 9.605606158186893e-05,
      "loss": 0.8663,
      "mean_token_accuracy": 0.7504206262528896,
      "num_tokens": 41066236.0,
      "step": 1287
    },
    {
      "entropy": 0.8941641449928284,
      "epoch": 0.11850420641629433,
      "grad_norm": 0.15747779607772827,
      "learning_rate": 9.60529947557273e-05,
      "loss": 0.8969,
      "mean_token_accuracy": 0.7481241710484028,
      "num_tokens": 41098354.0,
      "step": 1288
    },
    {
      "entropy": 0.9433057494461536,
      "epoch": 0.11859621278773555,
      "grad_norm": 0.15869227051734924,
      "learning_rate": 9.604992792958568e-05,
      "loss": 0.9369,
      "mean_token_accuracy": 0.734813317656517,
      "num_tokens": 41130356.0,
      "step": 1289
    },
    {
      "entropy": 0.8895032368600368,
      "epoch": 0.11868821915917677,
      "grad_norm": 0.16359210014343262,
      "learning_rate": 9.604686110344405e-05,
      "loss": 0.889,
      "mean_token_accuracy": 0.7488925568759441,
      "num_tokens": 41161739.0,
      "step": 1290
    },
    {
      "entropy": 0.8882927969098091,
      "epoch": 0.118780225530618,
      "grad_norm": 0.15495671331882477,
      "learning_rate": 9.604379427730242e-05,
      "loss": 0.8726,
      "mean_token_accuracy": 0.7462094239890575,
      "num_tokens": 41193827.0,
      "step": 1291
    },
    {
      "entropy": 0.8922304287552834,
      "epoch": 0.11887223190205921,
      "grad_norm": 0.16136017441749573,
      "learning_rate": 9.60407274511608e-05,
      "loss": 0.923,
      "mean_token_accuracy": 0.7413236014544964,
      "num_tokens": 41226090.0,
      "step": 1292
    },
    {
      "entropy": 1.0411744825541973,
      "epoch": 0.11896423827350044,
      "grad_norm": 0.16743867099285126,
      "learning_rate": 9.603766062501918e-05,
      "loss": 1.0629,
      "mean_token_accuracy": 0.7125931233167648,
      "num_tokens": 41258355.0,
      "step": 1293
    },
    {
      "entropy": 0.9646545648574829,
      "epoch": 0.11905624464494166,
      "grad_norm": 0.16477781534194946,
      "learning_rate": 9.603459379887755e-05,
      "loss": 0.9878,
      "mean_token_accuracy": 0.7236351259052753,
      "num_tokens": 41290449.0,
      "step": 1294
    },
    {
      "entropy": 0.8724432066082954,
      "epoch": 0.11914825101638288,
      "grad_norm": 0.15520206093788147,
      "learning_rate": 9.603152697273592e-05,
      "loss": 0.8854,
      "mean_token_accuracy": 0.7512686215341091,
      "num_tokens": 41322472.0,
      "step": 1295
    },
    {
      "entropy": 1.0334160886704922,
      "epoch": 0.11924025738782411,
      "grad_norm": 0.17018243670463562,
      "learning_rate": 9.602846014659429e-05,
      "loss": 1.0578,
      "mean_token_accuracy": 0.7093744091689587,
      "num_tokens": 41354631.0,
      "step": 1296
    },
    {
      "entropy": 0.8692147396504879,
      "epoch": 0.11933226375926533,
      "grad_norm": 0.15482290089130402,
      "learning_rate": 9.602539332045267e-05,
      "loss": 0.8507,
      "mean_token_accuracy": 0.75397614762187,
      "num_tokens": 41386676.0,
      "step": 1297
    },
    {
      "entropy": 0.870785865932703,
      "epoch": 0.11942427013070656,
      "grad_norm": 0.15196461975574493,
      "learning_rate": 9.602232649431105e-05,
      "loss": 0.8613,
      "mean_token_accuracy": 0.755605436861515,
      "num_tokens": 41418720.0,
      "step": 1298
    },
    {
      "entropy": 1.0388873219490051,
      "epoch": 0.11951627650214777,
      "grad_norm": 0.1653686761856079,
      "learning_rate": 9.601925966816942e-05,
      "loss": 1.0266,
      "mean_token_accuracy": 0.7155215814709663,
      "num_tokens": 41449563.0,
      "step": 1299
    },
    {
      "entropy": 1.0249877907335758,
      "epoch": 0.11960828287358899,
      "grad_norm": 0.16697122156620026,
      "learning_rate": 9.601619284202778e-05,
      "loss": 1.0187,
      "mean_token_accuracy": 0.7147185653448105,
      "num_tokens": 41480552.0,
      "step": 1300
    },
    {
      "entropy": 0.8812817446887493,
      "epoch": 0.11970028924503022,
      "grad_norm": 0.1539660394191742,
      "learning_rate": 9.601312601588617e-05,
      "loss": 0.8602,
      "mean_token_accuracy": 0.7542939521372318,
      "num_tokens": 41512666.0,
      "step": 1301
    },
    {
      "entropy": 0.9086950607597828,
      "epoch": 0.11979229561647144,
      "grad_norm": 0.15985901653766632,
      "learning_rate": 9.601005918974453e-05,
      "loss": 0.9122,
      "mean_token_accuracy": 0.7423644848167896,
      "num_tokens": 41544291.0,
      "step": 1302
    },
    {
      "entropy": 0.9174596555531025,
      "epoch": 0.11988430198791267,
      "grad_norm": 0.1565527617931366,
      "learning_rate": 9.600699236360292e-05,
      "loss": 0.9132,
      "mean_token_accuracy": 0.7424374893307686,
      "num_tokens": 41576656.0,
      "step": 1303
    },
    {
      "entropy": 0.8075642287731171,
      "epoch": 0.11997630835935388,
      "grad_norm": 0.16462011635303497,
      "learning_rate": 9.600392553746128e-05,
      "loss": 0.8091,
      "mean_token_accuracy": 0.7658773101866245,
      "num_tokens": 41608802.0,
      "step": 1304
    },
    {
      "entropy": 0.8962103370577097,
      "epoch": 0.12006831473079511,
      "grad_norm": 0.15050943195819855,
      "learning_rate": 9.600085871131966e-05,
      "loss": 0.8791,
      "mean_token_accuracy": 0.7527477517724037,
      "num_tokens": 41641307.0,
      "step": 1305
    },
    {
      "entropy": 0.8839820213615894,
      "epoch": 0.12016032110223633,
      "grad_norm": 0.157679945230484,
      "learning_rate": 9.599779188517803e-05,
      "loss": 0.9193,
      "mean_token_accuracy": 0.7409001365303993,
      "num_tokens": 41673942.0,
      "step": 1306
    },
    {
      "entropy": 0.9388176463544369,
      "epoch": 0.12025232747367755,
      "grad_norm": 0.15649941563606262,
      "learning_rate": 9.59947250590364e-05,
      "loss": 0.949,
      "mean_token_accuracy": 0.7278507873415947,
      "num_tokens": 41706191.0,
      "step": 1307
    },
    {
      "entropy": 0.925638634711504,
      "epoch": 0.12034433384511878,
      "grad_norm": 0.1594221442937851,
      "learning_rate": 9.599165823289478e-05,
      "loss": 0.9365,
      "mean_token_accuracy": 0.7373437881469727,
      "num_tokens": 41738211.0,
      "step": 1308
    },
    {
      "entropy": 0.9623091034591198,
      "epoch": 0.12043634021656,
      "grad_norm": 0.1614353060722351,
      "learning_rate": 9.598859140675316e-05,
      "loss": 0.9611,
      "mean_token_accuracy": 0.7300198599696159,
      "num_tokens": 41769243.0,
      "step": 1309
    },
    {
      "entropy": 0.87416772544384,
      "epoch": 0.12052834658800123,
      "grad_norm": 0.16179439425468445,
      "learning_rate": 9.598552458061153e-05,
      "loss": 0.8911,
      "mean_token_accuracy": 0.7521978691220284,
      "num_tokens": 41801341.0,
      "step": 1310
    },
    {
      "entropy": 0.903215266764164,
      "epoch": 0.12062035295944244,
      "grad_norm": 0.15914539992809296,
      "learning_rate": 9.59824577544699e-05,
      "loss": 0.9154,
      "mean_token_accuracy": 0.7424189038574696,
      "num_tokens": 41833642.0,
      "step": 1311
    },
    {
      "entropy": 0.9014825485646725,
      "epoch": 0.12071235933088366,
      "grad_norm": 0.15694665908813477,
      "learning_rate": 9.597939092832827e-05,
      "loss": 0.9205,
      "mean_token_accuracy": 0.7389534637331963,
      "num_tokens": 41865892.0,
      "step": 1312
    },
    {
      "entropy": 0.8975206315517426,
      "epoch": 0.12080436570232489,
      "grad_norm": 0.15848146378993988,
      "learning_rate": 9.597632410218665e-05,
      "loss": 0.8642,
      "mean_token_accuracy": 0.7452213801443577,
      "num_tokens": 41897204.0,
      "step": 1313
    },
    {
      "entropy": 0.8691624291241169,
      "epoch": 0.1208963720737661,
      "grad_norm": 0.15127311646938324,
      "learning_rate": 9.597325727604503e-05,
      "loss": 0.852,
      "mean_token_accuracy": 0.7555100135505199,
      "num_tokens": 41928757.0,
      "step": 1314
    },
    {
      "entropy": 0.9531728215515614,
      "epoch": 0.12098837844520734,
      "grad_norm": 0.16214917600154877,
      "learning_rate": 9.59701904499034e-05,
      "loss": 0.9487,
      "mean_token_accuracy": 0.7345303632318974,
      "num_tokens": 41960214.0,
      "step": 1315
    },
    {
      "entropy": 0.8498926572501659,
      "epoch": 0.12108038481664855,
      "grad_norm": 0.162801593542099,
      "learning_rate": 9.596712362376178e-05,
      "loss": 0.854,
      "mean_token_accuracy": 0.7585731036961079,
      "num_tokens": 41992230.0,
      "step": 1316
    },
    {
      "entropy": 0.9361480586230755,
      "epoch": 0.12117239118808977,
      "grad_norm": 0.158907949924469,
      "learning_rate": 9.596405679762015e-05,
      "loss": 0.9382,
      "mean_token_accuracy": 0.7340890765190125,
      "num_tokens": 42024506.0,
      "step": 1317
    },
    {
      "entropy": 0.8803095314651728,
      "epoch": 0.121264397559531,
      "grad_norm": 0.15394945442676544,
      "learning_rate": 9.596098997147851e-05,
      "loss": 0.8862,
      "mean_token_accuracy": 0.7493113130331039,
      "num_tokens": 42057116.0,
      "step": 1318
    },
    {
      "entropy": 0.981334213167429,
      "epoch": 0.12135640393097222,
      "grad_norm": 0.1622510701417923,
      "learning_rate": 9.59579231453369e-05,
      "loss": 0.9775,
      "mean_token_accuracy": 0.7281354516744614,
      "num_tokens": 42088545.0,
      "step": 1319
    },
    {
      "entropy": 0.8302001766860485,
      "epoch": 0.12144841030241345,
      "grad_norm": 0.1583346724510193,
      "learning_rate": 9.595485631919528e-05,
      "loss": 0.834,
      "mean_token_accuracy": 0.7623110041022301,
      "num_tokens": 42121007.0,
      "step": 1320
    },
    {
      "entropy": 0.9328108727931976,
      "epoch": 0.12154041667385466,
      "grad_norm": 0.15617331862449646,
      "learning_rate": 9.595178949305365e-05,
      "loss": 0.9201,
      "mean_token_accuracy": 0.7406449504196644,
      "num_tokens": 42152533.0,
      "step": 1321
    },
    {
      "entropy": 0.8520091995596886,
      "epoch": 0.12163242304529588,
      "grad_norm": 0.15073151886463165,
      "learning_rate": 9.594872266691201e-05,
      "loss": 0.8365,
      "mean_token_accuracy": 0.7635049447417259,
      "num_tokens": 42184999.0,
      "step": 1322
    },
    {
      "entropy": 0.8666563853621483,
      "epoch": 0.12172442941673711,
      "grad_norm": 0.1630660742521286,
      "learning_rate": 9.594565584077038e-05,
      "loss": 0.8554,
      "mean_token_accuracy": 0.7574391067028046,
      "num_tokens": 42216563.0,
      "step": 1323
    },
    {
      "entropy": 0.9101729765534401,
      "epoch": 0.12181643578817833,
      "grad_norm": 0.1543581336736679,
      "learning_rate": 9.594258901462878e-05,
      "loss": 0.895,
      "mean_token_accuracy": 0.7461829036474228,
      "num_tokens": 42248663.0,
      "step": 1324
    },
    {
      "entropy": 0.9893680065870285,
      "epoch": 0.12190844215961956,
      "grad_norm": 0.1749403476715088,
      "learning_rate": 9.593952218848714e-05,
      "loss": 0.9708,
      "mean_token_accuracy": 0.7252227105200291,
      "num_tokens": 42280538.0,
      "step": 1325
    },
    {
      "entropy": 0.892228826880455,
      "epoch": 0.12200044853106078,
      "grad_norm": 0.1589728593826294,
      "learning_rate": 9.593645536234551e-05,
      "loss": 0.896,
      "mean_token_accuracy": 0.7447525262832642,
      "num_tokens": 42312509.0,
      "step": 1326
    },
    {
      "entropy": 0.8364868648350239,
      "epoch": 0.12209245490250199,
      "grad_norm": 0.1602228432893753,
      "learning_rate": 9.593338853620388e-05,
      "loss": 0.8508,
      "mean_token_accuracy": 0.756545003503561,
      "num_tokens": 42344396.0,
      "step": 1327
    },
    {
      "entropy": 0.8401671275496483,
      "epoch": 0.12218446127394322,
      "grad_norm": 0.1565535068511963,
      "learning_rate": 9.593032171006226e-05,
      "loss": 0.8598,
      "mean_token_accuracy": 0.7501671090722084,
      "num_tokens": 42376729.0,
      "step": 1328
    },
    {
      "entropy": 0.92406390234828,
      "epoch": 0.12227646764538444,
      "grad_norm": 0.16469357907772064,
      "learning_rate": 9.592725488392064e-05,
      "loss": 0.9626,
      "mean_token_accuracy": 0.7325879819691181,
      "num_tokens": 42408380.0,
      "step": 1329
    },
    {
      "entropy": 0.84736836142838,
      "epoch": 0.12236847401682567,
      "grad_norm": 0.15513655543327332,
      "learning_rate": 9.592418805777901e-05,
      "loss": 0.8431,
      "mean_token_accuracy": 0.7582830339670181,
      "num_tokens": 42440932.0,
      "step": 1330
    },
    {
      "entropy": 0.9204532355070114,
      "epoch": 0.12246048038826689,
      "grad_norm": 0.16738682985305786,
      "learning_rate": 9.592112123163738e-05,
      "loss": 0.9343,
      "mean_token_accuracy": 0.7366954833269119,
      "num_tokens": 42473397.0,
      "step": 1331
    },
    {
      "entropy": 0.8773025572299957,
      "epoch": 0.1225524867597081,
      "grad_norm": 0.16079069674015045,
      "learning_rate": 9.591805440549576e-05,
      "loss": 0.8766,
      "mean_token_accuracy": 0.7528524920344353,
      "num_tokens": 42505979.0,
      "step": 1332
    },
    {
      "entropy": 0.89029336348176,
      "epoch": 0.12264449313114933,
      "grad_norm": 0.1592784821987152,
      "learning_rate": 9.591498757935413e-05,
      "loss": 0.8902,
      "mean_token_accuracy": 0.744793176651001,
      "num_tokens": 42537535.0,
      "step": 1333
    },
    {
      "entropy": 0.9126544371247292,
      "epoch": 0.12273649950259055,
      "grad_norm": 0.16223393380641937,
      "learning_rate": 9.591192075321251e-05,
      "loss": 0.9135,
      "mean_token_accuracy": 0.7423229739069939,
      "num_tokens": 42569594.0,
      "step": 1334
    },
    {
      "entropy": 0.9771235771477222,
      "epoch": 0.12282850587403178,
      "grad_norm": 0.17410695552825928,
      "learning_rate": 9.590885392707088e-05,
      "loss": 0.9688,
      "mean_token_accuracy": 0.7271896004676819,
      "num_tokens": 42600994.0,
      "step": 1335
    },
    {
      "entropy": 1.003642249852419,
      "epoch": 0.122920512245473,
      "grad_norm": 0.16569291055202484,
      "learning_rate": 9.590578710092926e-05,
      "loss": 1.0222,
      "mean_token_accuracy": 0.7134799920022488,
      "num_tokens": 42633278.0,
      "step": 1336
    },
    {
      "entropy": 0.9009298980236053,
      "epoch": 0.12301251861691422,
      "grad_norm": 0.15926679968833923,
      "learning_rate": 9.590272027478763e-05,
      "loss": 0.9084,
      "mean_token_accuracy": 0.7428008280694485,
      "num_tokens": 42665214.0,
      "step": 1337
    },
    {
      "entropy": 0.9603922367095947,
      "epoch": 0.12310452498835545,
      "grad_norm": 0.16201116144657135,
      "learning_rate": 9.5899653448646e-05,
      "loss": 0.9543,
      "mean_token_accuracy": 0.7354647144675255,
      "num_tokens": 42696794.0,
      "step": 1338
    },
    {
      "entropy": 0.9607611708343029,
      "epoch": 0.12319653135979666,
      "grad_norm": 0.15695728361606598,
      "learning_rate": 9.589658662250438e-05,
      "loss": 0.9433,
      "mean_token_accuracy": 0.732390783727169,
      "num_tokens": 42729118.0,
      "step": 1339
    },
    {
      "entropy": 0.937091987580061,
      "epoch": 0.1232885377312379,
      "grad_norm": 0.16493549942970276,
      "learning_rate": 9.589351979636276e-05,
      "loss": 0.9248,
      "mean_token_accuracy": 0.7373012453317642,
      "num_tokens": 42760580.0,
      "step": 1340
    },
    {
      "entropy": 0.9200702998787165,
      "epoch": 0.12338054410267911,
      "grad_norm": 0.1624705046415329,
      "learning_rate": 9.589045297022112e-05,
      "loss": 0.9154,
      "mean_token_accuracy": 0.7398549802601337,
      "num_tokens": 42792354.0,
      "step": 1341
    },
    {
      "entropy": 0.8025525808334351,
      "epoch": 0.12347255047412033,
      "grad_norm": 0.15369336307048798,
      "learning_rate": 9.588738614407949e-05,
      "loss": 0.8256,
      "mean_token_accuracy": 0.7677078992128372,
      "num_tokens": 42824999.0,
      "step": 1342
    },
    {
      "entropy": 0.8614583425223827,
      "epoch": 0.12356455684556156,
      "grad_norm": 0.15983784198760986,
      "learning_rate": 9.588431931793786e-05,
      "loss": 0.8705,
      "mean_token_accuracy": 0.7562061212956905,
      "num_tokens": 42855960.0,
      "step": 1343
    },
    {
      "entropy": 0.881206326186657,
      "epoch": 0.12365656321700277,
      "grad_norm": 0.1611868143081665,
      "learning_rate": 9.588125249179624e-05,
      "loss": 0.8973,
      "mean_token_accuracy": 0.7474408335983753,
      "num_tokens": 42888256.0,
      "step": 1344
    },
    {
      "entropy": 0.9357227012515068,
      "epoch": 0.123748569588444,
      "grad_norm": 0.16418229043483734,
      "learning_rate": 9.587818566565462e-05,
      "loss": 0.9376,
      "mean_token_accuracy": 0.7395183816552162,
      "num_tokens": 42918977.0,
      "step": 1345
    },
    {
      "entropy": 0.8898397963494062,
      "epoch": 0.12384057595988522,
      "grad_norm": 0.15771222114562988,
      "learning_rate": 9.587511883951299e-05,
      "loss": 0.892,
      "mean_token_accuracy": 0.7475471459329128,
      "num_tokens": 42951240.0,
      "step": 1346
    },
    {
      "entropy": 0.9032508302479982,
      "epoch": 0.12393258233132644,
      "grad_norm": 0.1650296151638031,
      "learning_rate": 9.587205201337137e-05,
      "loss": 0.9051,
      "mean_token_accuracy": 0.7468995042145252,
      "num_tokens": 42982705.0,
      "step": 1347
    },
    {
      "entropy": 0.9133551232516766,
      "epoch": 0.12402458870276767,
      "grad_norm": 0.1615152209997177,
      "learning_rate": 9.586898518722974e-05,
      "loss": 0.9192,
      "mean_token_accuracy": 0.7387451305985451,
      "num_tokens": 43014513.0,
      "step": 1348
    },
    {
      "entropy": 1.0035524629056454,
      "epoch": 0.12411659507420889,
      "grad_norm": 0.18631993234157562,
      "learning_rate": 9.586591836108811e-05,
      "loss": 1.0202,
      "mean_token_accuracy": 0.7165511026978493,
      "num_tokens": 43047035.0,
      "step": 1349
    },
    {
      "entropy": 0.8903112858533859,
      "epoch": 0.12420860144565012,
      "grad_norm": 0.15951476991176605,
      "learning_rate": 9.586285153494649e-05,
      "loss": 0.8995,
      "mean_token_accuracy": 0.7450625076889992,
      "num_tokens": 43079192.0,
      "step": 1350
    },
    {
      "entropy": 0.8962217308580875,
      "epoch": 0.12430060781709133,
      "grad_norm": 0.15349577367305756,
      "learning_rate": 9.585978470880487e-05,
      "loss": 0.8913,
      "mean_token_accuracy": 0.7459709905087948,
      "num_tokens": 43111336.0,
      "step": 1351
    },
    {
      "entropy": 0.9824104551225901,
      "epoch": 0.12439261418853255,
      "grad_norm": 0.1629566103219986,
      "learning_rate": 9.585671788266324e-05,
      "loss": 0.9779,
      "mean_token_accuracy": 0.7294751591980457,
      "num_tokens": 43143210.0,
      "step": 1352
    },
    {
      "entropy": 0.8909053448587656,
      "epoch": 0.12448462055997378,
      "grad_norm": 0.1611369550228119,
      "learning_rate": 9.58536510565216e-05,
      "loss": 0.8928,
      "mean_token_accuracy": 0.7482354044914246,
      "num_tokens": 43175574.0,
      "step": 1353
    },
    {
      "entropy": 0.9078603200614452,
      "epoch": 0.124576626931415,
      "grad_norm": 0.16770216822624207,
      "learning_rate": 9.585058423037997e-05,
      "loss": 0.9219,
      "mean_token_accuracy": 0.7385167479515076,
      "num_tokens": 43207563.0,
      "step": 1354
    },
    {
      "entropy": 0.9457318540662527,
      "epoch": 0.12466863330285623,
      "grad_norm": 0.16901403665542603,
      "learning_rate": 9.584751740423836e-05,
      "loss": 0.9789,
      "mean_token_accuracy": 0.7298157438635826,
      "num_tokens": 43239727.0,
      "step": 1355
    },
    {
      "entropy": 0.7926583401858807,
      "epoch": 0.12476063967429744,
      "grad_norm": 0.1529662013053894,
      "learning_rate": 9.584445057809674e-05,
      "loss": 0.7791,
      "mean_token_accuracy": 0.772360410541296,
      "num_tokens": 43270755.0,
      "step": 1356
    },
    {
      "entropy": 0.886942557990551,
      "epoch": 0.12485264604573866,
      "grad_norm": 0.15925097465515137,
      "learning_rate": 9.58413837519551e-05,
      "loss": 0.8695,
      "mean_token_accuracy": 0.7511759772896767,
      "num_tokens": 43302637.0,
      "step": 1357
    },
    {
      "entropy": 0.9399158507585526,
      "epoch": 0.12494465241717989,
      "grad_norm": 0.15796351432800293,
      "learning_rate": 9.583831692581347e-05,
      "loss": 0.9322,
      "mean_token_accuracy": 0.7384955734014511,
      "num_tokens": 43335143.0,
      "step": 1358
    },
    {
      "entropy": 0.9325386770069599,
      "epoch": 0.1250366587886211,
      "grad_norm": 0.1641385406255722,
      "learning_rate": 9.583525009967185e-05,
      "loss": 0.9126,
      "mean_token_accuracy": 0.7446613125503063,
      "num_tokens": 43366262.0,
      "step": 1359
    },
    {
      "entropy": 0.7758619263768196,
      "epoch": 0.12512866516006232,
      "grad_norm": 0.14610634744167328,
      "learning_rate": 9.583218327353024e-05,
      "loss": 0.752,
      "mean_token_accuracy": 0.7798018008470535,
      "num_tokens": 43398136.0,
      "step": 1360
    },
    {
      "entropy": 0.9312492199242115,
      "epoch": 0.12522067153150357,
      "grad_norm": 0.15372642874717712,
      "learning_rate": 9.58291164473886e-05,
      "loss": 0.9312,
      "mean_token_accuracy": 0.7378669194877148,
      "num_tokens": 43430564.0,
      "step": 1361
    },
    {
      "entropy": 0.9555746503174305,
      "epoch": 0.12531267790294479,
      "grad_norm": 0.16121193766593933,
      "learning_rate": 9.582604962124697e-05,
      "loss": 0.955,
      "mean_token_accuracy": 0.7344971969723701,
      "num_tokens": 43462310.0,
      "step": 1362
    },
    {
      "entropy": 0.8892284110188484,
      "epoch": 0.125404684274386,
      "grad_norm": 0.15434499084949493,
      "learning_rate": 9.582298279510535e-05,
      "loss": 0.8908,
      "mean_token_accuracy": 0.7497995942831039,
      "num_tokens": 43494606.0,
      "step": 1363
    },
    {
      "entropy": 0.9992723986506462,
      "epoch": 0.12549669064582722,
      "grad_norm": 0.16130992770195007,
      "learning_rate": 9.581991596896372e-05,
      "loss": 1.0102,
      "mean_token_accuracy": 0.7208214811980724,
      "num_tokens": 43526699.0,
      "step": 1364
    },
    {
      "entropy": 0.9079805351793766,
      "epoch": 0.12558869701726844,
      "grad_norm": 0.15785521268844604,
      "learning_rate": 9.58168491428221e-05,
      "loss": 0.9193,
      "mean_token_accuracy": 0.7422300651669502,
      "num_tokens": 43559188.0,
      "step": 1365
    },
    {
      "entropy": 0.8955215811729431,
      "epoch": 0.12568070338870968,
      "grad_norm": 0.15921910107135773,
      "learning_rate": 9.581378231668047e-05,
      "loss": 0.9107,
      "mean_token_accuracy": 0.7461718767881393,
      "num_tokens": 43590383.0,
      "step": 1366
    },
    {
      "entropy": 0.9458691142499447,
      "epoch": 0.1257727097601509,
      "grad_norm": 0.16171136498451233,
      "learning_rate": 9.581071549053885e-05,
      "loss": 0.9556,
      "mean_token_accuracy": 0.7290371023118496,
      "num_tokens": 43621959.0,
      "step": 1367
    },
    {
      "entropy": 0.9895393438637257,
      "epoch": 0.1258647161315921,
      "grad_norm": 0.16478824615478516,
      "learning_rate": 9.580764866439722e-05,
      "loss": 0.9921,
      "mean_token_accuracy": 0.7248892262578011,
      "num_tokens": 43654468.0,
      "step": 1368
    },
    {
      "entropy": 0.9287047646939754,
      "epoch": 0.12595672250303333,
      "grad_norm": 0.160200297832489,
      "learning_rate": 9.580458183825559e-05,
      "loss": 0.9195,
      "mean_token_accuracy": 0.7397038154304028,
      "num_tokens": 43686735.0,
      "step": 1369
    },
    {
      "entropy": 0.8816953413188457,
      "epoch": 0.12604872887447455,
      "grad_norm": 0.16115164756774902,
      "learning_rate": 9.580151501211397e-05,
      "loss": 0.8794,
      "mean_token_accuracy": 0.7526337429881096,
      "num_tokens": 43718170.0,
      "step": 1370
    },
    {
      "entropy": 0.9544548876583576,
      "epoch": 0.1261407352459158,
      "grad_norm": 0.16148696839809418,
      "learning_rate": 9.579844818597235e-05,
      "loss": 0.9515,
      "mean_token_accuracy": 0.7347146458923817,
      "num_tokens": 43750632.0,
      "step": 1371
    },
    {
      "entropy": 1.0148214735090733,
      "epoch": 0.126232741617357,
      "grad_norm": 0.16944923996925354,
      "learning_rate": 9.579538135983072e-05,
      "loss": 1.0164,
      "mean_token_accuracy": 0.7153962999582291,
      "num_tokens": 43781681.0,
      "step": 1372
    },
    {
      "entropy": 0.8881180249154568,
      "epoch": 0.12632474798879822,
      "grad_norm": 0.15640972554683685,
      "learning_rate": 9.579231453368909e-05,
      "loss": 0.8982,
      "mean_token_accuracy": 0.7436126992106438,
      "num_tokens": 43813855.0,
      "step": 1373
    },
    {
      "entropy": 0.93473981320858,
      "epoch": 0.12641675436023944,
      "grad_norm": 0.15791121125221252,
      "learning_rate": 9.578924770754747e-05,
      "loss": 0.9356,
      "mean_token_accuracy": 0.742297176271677,
      "num_tokens": 43846517.0,
      "step": 1374
    },
    {
      "entropy": 0.8538608811795712,
      "epoch": 0.12650876073168066,
      "grad_norm": 0.15636909008026123,
      "learning_rate": 9.578618088140584e-05,
      "loss": 0.8411,
      "mean_token_accuracy": 0.7586209625005722,
      "num_tokens": 43878338.0,
      "step": 1375
    },
    {
      "entropy": 0.825945045799017,
      "epoch": 0.1266007671031219,
      "grad_norm": 0.1545090228319168,
      "learning_rate": 9.578311405526422e-05,
      "loss": 0.8191,
      "mean_token_accuracy": 0.7658201716840267,
      "num_tokens": 43910222.0,
      "step": 1376
    },
    {
      "entropy": 0.8869588673114777,
      "epoch": 0.12669277347456312,
      "grad_norm": 0.15499937534332275,
      "learning_rate": 9.578004722912258e-05,
      "loss": 0.9031,
      "mean_token_accuracy": 0.7419997341930866,
      "num_tokens": 43942758.0,
      "step": 1377
    },
    {
      "entropy": 0.91963130235672,
      "epoch": 0.12678477984600434,
      "grad_norm": 0.1623668074607849,
      "learning_rate": 9.577698040298097e-05,
      "loss": 0.9329,
      "mean_token_accuracy": 0.7402029670774937,
      "num_tokens": 43973772.0,
      "step": 1378
    },
    {
      "entropy": 0.8775487281382084,
      "epoch": 0.12687678621744555,
      "grad_norm": 0.1552736759185791,
      "learning_rate": 9.577391357683933e-05,
      "loss": 0.8879,
      "mean_token_accuracy": 0.7492555752396584,
      "num_tokens": 44005705.0,
      "step": 1379
    },
    {
      "entropy": 0.927038948982954,
      "epoch": 0.12696879258888677,
      "grad_norm": 0.15983612835407257,
      "learning_rate": 9.57708467506977e-05,
      "loss": 0.928,
      "mean_token_accuracy": 0.7426281794905663,
      "num_tokens": 44037590.0,
      "step": 1380
    },
    {
      "entropy": 0.8520211782306433,
      "epoch": 0.12706079896032801,
      "grad_norm": 0.1526368409395218,
      "learning_rate": 9.576777992455608e-05,
      "loss": 0.8339,
      "mean_token_accuracy": 0.7652990408241749,
      "num_tokens": 44069649.0,
      "step": 1381
    },
    {
      "entropy": 0.9200831651687622,
      "epoch": 0.12715280533176923,
      "grad_norm": 0.1639328896999359,
      "learning_rate": 9.576471309841446e-05,
      "loss": 0.9257,
      "mean_token_accuracy": 0.7412961311638355,
      "num_tokens": 44101416.0,
      "step": 1382
    },
    {
      "entropy": 0.8470993041992188,
      "epoch": 0.12724481170321045,
      "grad_norm": 0.1600313037633896,
      "learning_rate": 9.576164627227283e-05,
      "loss": 0.8499,
      "mean_token_accuracy": 0.7564584910869598,
      "num_tokens": 44132927.0,
      "step": 1383
    },
    {
      "entropy": 0.9344243407249451,
      "epoch": 0.12733681807465166,
      "grad_norm": 0.15922343730926514,
      "learning_rate": 9.57585794461312e-05,
      "loss": 0.9252,
      "mean_token_accuracy": 0.7399983033537865,
      "num_tokens": 44165158.0,
      "step": 1384
    },
    {
      "entropy": 0.8967868983745575,
      "epoch": 0.12742882444609288,
      "grad_norm": 0.15718422830104828,
      "learning_rate": 9.575551261998957e-05,
      "loss": 0.8862,
      "mean_token_accuracy": 0.7437204085290432,
      "num_tokens": 44196753.0,
      "step": 1385
    },
    {
      "entropy": 0.8965760283172131,
      "epoch": 0.12752083081753413,
      "grad_norm": 0.16037434339523315,
      "learning_rate": 9.575244579384795e-05,
      "loss": 0.9118,
      "mean_token_accuracy": 0.7457964047789574,
      "num_tokens": 44227949.0,
      "step": 1386
    },
    {
      "entropy": 0.9602333530783653,
      "epoch": 0.12761283718897534,
      "grad_norm": 0.16215936839580536,
      "learning_rate": 9.574937896770633e-05,
      "loss": 0.9509,
      "mean_token_accuracy": 0.7275675125420094,
      "num_tokens": 44260270.0,
      "step": 1387
    },
    {
      "entropy": 0.9736673645675182,
      "epoch": 0.12770484356041656,
      "grad_norm": 0.15929417312145233,
      "learning_rate": 9.57463121415647e-05,
      "loss": 0.9809,
      "mean_token_accuracy": 0.7328590862452984,
      "num_tokens": 44292345.0,
      "step": 1388
    },
    {
      "entropy": 0.9236449562013149,
      "epoch": 0.12779684993185778,
      "grad_norm": 0.15642336010932922,
      "learning_rate": 9.574324531542307e-05,
      "loss": 0.9332,
      "mean_token_accuracy": 0.7394499257206917,
      "num_tokens": 44324279.0,
      "step": 1389
    },
    {
      "entropy": 0.8794885501265526,
      "epoch": 0.127888856303299,
      "grad_norm": 0.150704488158226,
      "learning_rate": 9.574017848928145e-05,
      "loss": 0.8524,
      "mean_token_accuracy": 0.7605950981378555,
      "num_tokens": 44356371.0,
      "step": 1390
    },
    {
      "entropy": 1.0406970605254173,
      "epoch": 0.12798086267474024,
      "grad_norm": 0.16482026875019073,
      "learning_rate": 9.573711166313982e-05,
      "loss": 1.038,
      "mean_token_accuracy": 0.7079443000257015,
      "num_tokens": 44388133.0,
      "step": 1391
    },
    {
      "entropy": 0.751538960263133,
      "epoch": 0.12807286904618145,
      "grad_norm": 0.1450197398662567,
      "learning_rate": 9.57340448369982e-05,
      "loss": 0.7323,
      "mean_token_accuracy": 0.7856653071939945,
      "num_tokens": 44420568.0,
      "step": 1392
    },
    {
      "entropy": 0.90673753246665,
      "epoch": 0.12816487541762267,
      "grad_norm": 0.15802042186260223,
      "learning_rate": 9.573097801085657e-05,
      "loss": 0.8959,
      "mean_token_accuracy": 0.7491278238594532,
      "num_tokens": 44452478.0,
      "step": 1393
    },
    {
      "entropy": 0.9203658103942871,
      "epoch": 0.1282568817890639,
      "grad_norm": 0.16187678277492523,
      "learning_rate": 9.572791118471495e-05,
      "loss": 0.943,
      "mean_token_accuracy": 0.7378385253250599,
      "num_tokens": 44484924.0,
      "step": 1394
    },
    {
      "entropy": 0.813607320189476,
      "epoch": 0.1283488881605051,
      "grad_norm": 0.15430891513824463,
      "learning_rate": 9.572484435857331e-05,
      "loss": 0.7966,
      "mean_token_accuracy": 0.7726012878119946,
      "num_tokens": 44517112.0,
      "step": 1395
    },
    {
      "entropy": 0.993693646043539,
      "epoch": 0.12844089453194635,
      "grad_norm": 0.16756080090999603,
      "learning_rate": 9.572177753243168e-05,
      "loss": 1.0035,
      "mean_token_accuracy": 0.720903679728508,
      "num_tokens": 44549304.0,
      "step": 1396
    },
    {
      "entropy": 0.8364639040082693,
      "epoch": 0.12853290090338756,
      "grad_norm": 0.15957613289356232,
      "learning_rate": 9.571871070629006e-05,
      "loss": 0.8479,
      "mean_token_accuracy": 0.7598730958998203,
      "num_tokens": 44581454.0,
      "step": 1397
    },
    {
      "entropy": 0.8466631025075912,
      "epoch": 0.12862490727482878,
      "grad_norm": 0.15988339483737946,
      "learning_rate": 9.571564388014845e-05,
      "loss": 0.8775,
      "mean_token_accuracy": 0.748354397714138,
      "num_tokens": 44613151.0,
      "step": 1398
    },
    {
      "entropy": 0.904773224145174,
      "epoch": 0.12871691364627,
      "grad_norm": 0.16558140516281128,
      "learning_rate": 9.571257705400681e-05,
      "loss": 0.9161,
      "mean_token_accuracy": 0.7479781620204449,
      "num_tokens": 44645257.0,
      "step": 1399
    },
    {
      "entropy": 0.9671018701046705,
      "epoch": 0.12880892001771121,
      "grad_norm": 0.1613721400499344,
      "learning_rate": 9.570951022786518e-05,
      "loss": 0.9679,
      "mean_token_accuracy": 0.7287022992968559,
      "num_tokens": 44677511.0,
      "step": 1400
    },
    {
      "entropy": 0.8655254878103733,
      "epoch": 0.12890092638915246,
      "grad_norm": 0.16883884370326996,
      "learning_rate": 9.570644340172355e-05,
      "loss": 0.8948,
      "mean_token_accuracy": 0.7485690638422966,
      "num_tokens": 44708779.0,
      "step": 1401
    },
    {
      "entropy": 0.9527161046862602,
      "epoch": 0.12899293276059368,
      "grad_norm": 0.17072026431560516,
      "learning_rate": 9.570337657558194e-05,
      "loss": 0.9447,
      "mean_token_accuracy": 0.730373527854681,
      "num_tokens": 44740716.0,
      "step": 1402
    },
    {
      "entropy": 0.8556098453700542,
      "epoch": 0.1290849391320349,
      "grad_norm": 0.15850068628787994,
      "learning_rate": 9.570030974944031e-05,
      "loss": 0.8613,
      "mean_token_accuracy": 0.7535067461431026,
      "num_tokens": 44772427.0,
      "step": 1403
    },
    {
      "entropy": 0.8870825320482254,
      "epoch": 0.1291769455034761,
      "grad_norm": 0.15278637409210205,
      "learning_rate": 9.569724292329868e-05,
      "loss": 0.8709,
      "mean_token_accuracy": 0.7515799291431904,
      "num_tokens": 44805064.0,
      "step": 1404
    },
    {
      "entropy": 0.8493108432739973,
      "epoch": 0.12926895187491733,
      "grad_norm": 0.14963141083717346,
      "learning_rate": 9.569417609715706e-05,
      "loss": 0.8281,
      "mean_token_accuracy": 0.7599457427859306,
      "num_tokens": 44837654.0,
      "step": 1405
    },
    {
      "entropy": 0.8674461226910353,
      "epoch": 0.12936095824635857,
      "grad_norm": 0.15199655294418335,
      "learning_rate": 9.569110927101543e-05,
      "loss": 0.8571,
      "mean_token_accuracy": 0.7571729980409145,
      "num_tokens": 44870215.0,
      "step": 1406
    },
    {
      "entropy": 0.9104496315121651,
      "epoch": 0.1294529646177998,
      "grad_norm": 0.15655122697353363,
      "learning_rate": 9.568804244487381e-05,
      "loss": 0.8933,
      "mean_token_accuracy": 0.747913122177124,
      "num_tokens": 44902098.0,
      "step": 1407
    },
    {
      "entropy": 0.8995660915970802,
      "epoch": 0.129544970989241,
      "grad_norm": 0.16437354683876038,
      "learning_rate": 9.568497561873218e-05,
      "loss": 0.891,
      "mean_token_accuracy": 0.7484185509383678,
      "num_tokens": 44934140.0,
      "step": 1408
    },
    {
      "entropy": 0.8376863431185484,
      "epoch": 0.12963697736068222,
      "grad_norm": 0.15102721750736237,
      "learning_rate": 9.568190879259056e-05,
      "loss": 0.8232,
      "mean_token_accuracy": 0.764384750276804,
      "num_tokens": 44966364.0,
      "step": 1409
    },
    {
      "entropy": 0.8983563557267189,
      "epoch": 0.12972898373212344,
      "grad_norm": 0.1639176458120346,
      "learning_rate": 9.567884196644893e-05,
      "loss": 0.8875,
      "mean_token_accuracy": 0.7487972751259804,
      "num_tokens": 44998373.0,
      "step": 1410
    },
    {
      "entropy": 0.8669734224677086,
      "epoch": 0.12982099010356468,
      "grad_norm": 0.15847107768058777,
      "learning_rate": 9.56757751403073e-05,
      "loss": 0.8695,
      "mean_token_accuracy": 0.7527007162570953,
      "num_tokens": 45029878.0,
      "step": 1411
    },
    {
      "entropy": 0.8699670396745205,
      "epoch": 0.1299129964750059,
      "grad_norm": 0.15936405956745148,
      "learning_rate": 9.567270831416568e-05,
      "loss": 0.8754,
      "mean_token_accuracy": 0.7521078325808048,
      "num_tokens": 45061832.0,
      "step": 1412
    },
    {
      "entropy": 0.8051696121692657,
      "epoch": 0.13000500284644712,
      "grad_norm": 0.15526987612247467,
      "learning_rate": 9.566964148802406e-05,
      "loss": 0.8063,
      "mean_token_accuracy": 0.7685359306633472,
      "num_tokens": 45093749.0,
      "step": 1413
    },
    {
      "entropy": 0.8304604254662991,
      "epoch": 0.13009700921788833,
      "grad_norm": 0.1628667712211609,
      "learning_rate": 9.566657466188243e-05,
      "loss": 0.8571,
      "mean_token_accuracy": 0.7583776526153088,
      "num_tokens": 45125833.0,
      "step": 1414
    },
    {
      "entropy": 0.8962295353412628,
      "epoch": 0.13018901558932955,
      "grad_norm": 0.15798132121562958,
      "learning_rate": 9.56635078357408e-05,
      "loss": 0.9252,
      "mean_token_accuracy": 0.7397789061069489,
      "num_tokens": 45158024.0,
      "step": 1415
    },
    {
      "entropy": 0.9412603136152029,
      "epoch": 0.1302810219607708,
      "grad_norm": 0.16267749667167664,
      "learning_rate": 9.566044100959916e-05,
      "loss": 0.9491,
      "mean_token_accuracy": 0.7322576493024826,
      "num_tokens": 45189322.0,
      "step": 1416
    },
    {
      "entropy": 0.8350544311106205,
      "epoch": 0.130373028332212,
      "grad_norm": 0.15579266846179962,
      "learning_rate": 9.565737418345754e-05,
      "loss": 0.8475,
      "mean_token_accuracy": 0.7550486177206039,
      "num_tokens": 45221525.0,
      "step": 1417
    },
    {
      "entropy": 0.7900904975831509,
      "epoch": 0.13046503470365323,
      "grad_norm": 0.15703168511390686,
      "learning_rate": 9.565430735731592e-05,
      "loss": 0.808,
      "mean_token_accuracy": 0.7669675126671791,
      "num_tokens": 45253331.0,
      "step": 1418
    },
    {
      "entropy": 0.8832404986023903,
      "epoch": 0.13055704107509444,
      "grad_norm": 0.1606397181749344,
      "learning_rate": 9.565124053117429e-05,
      "loss": 0.8835,
      "mean_token_accuracy": 0.7473745383322239,
      "num_tokens": 45285653.0,
      "step": 1419
    },
    {
      "entropy": 0.884585052728653,
      "epoch": 0.13064904744653566,
      "grad_norm": 0.15162524580955505,
      "learning_rate": 9.564817370503266e-05,
      "loss": 0.8717,
      "mean_token_accuracy": 0.7572435364127159,
      "num_tokens": 45317926.0,
      "step": 1420
    },
    {
      "entropy": 0.8618516884744167,
      "epoch": 0.1307410538179769,
      "grad_norm": 0.14983543753623962,
      "learning_rate": 9.564510687889104e-05,
      "loss": 0.8602,
      "mean_token_accuracy": 0.7534883841872215,
      "num_tokens": 45350301.0,
      "step": 1421
    },
    {
      "entropy": 0.8881712406873703,
      "epoch": 0.13083306018941812,
      "grad_norm": 0.15896674990653992,
      "learning_rate": 9.564204005274941e-05,
      "loss": 0.89,
      "mean_token_accuracy": 0.7489947788417339,
      "num_tokens": 45382194.0,
      "step": 1422
    },
    {
      "entropy": 1.0374965369701385,
      "epoch": 0.13092506656085934,
      "grad_norm": 0.16714240610599518,
      "learning_rate": 9.563897322660779e-05,
      "loss": 1.0217,
      "mean_token_accuracy": 0.7151946127414703,
      "num_tokens": 45414413.0,
      "step": 1423
    },
    {
      "entropy": 0.991882685571909,
      "epoch": 0.13101707293230055,
      "grad_norm": 0.16575656831264496,
      "learning_rate": 9.563590640046616e-05,
      "loss": 0.9945,
      "mean_token_accuracy": 0.718562126159668,
      "num_tokens": 45445484.0,
      "step": 1424
    },
    {
      "entropy": 0.9294509626924992,
      "epoch": 0.13110907930374177,
      "grad_norm": 0.15846474468708038,
      "learning_rate": 9.563283957432454e-05,
      "loss": 0.9117,
      "mean_token_accuracy": 0.7415791526436806,
      "num_tokens": 45477790.0,
      "step": 1425
    },
    {
      "entropy": 0.9414418786764145,
      "epoch": 0.13120108567518302,
      "grad_norm": 0.15932688117027283,
      "learning_rate": 9.562977274818291e-05,
      "loss": 0.9425,
      "mean_token_accuracy": 0.7337660267949104,
      "num_tokens": 45509966.0,
      "step": 1426
    },
    {
      "entropy": 0.899724468588829,
      "epoch": 0.13129309204662423,
      "grad_norm": 0.15942607820034027,
      "learning_rate": 9.562670592204128e-05,
      "loss": 0.9143,
      "mean_token_accuracy": 0.7456745393574238,
      "num_tokens": 45542145.0,
      "step": 1427
    },
    {
      "entropy": 0.8454543855041265,
      "epoch": 0.13138509841806545,
      "grad_norm": 0.16145804524421692,
      "learning_rate": 9.562363909589966e-05,
      "loss": 0.8399,
      "mean_token_accuracy": 0.7555093877017498,
      "num_tokens": 45573540.0,
      "step": 1428
    },
    {
      "entropy": 0.830769807100296,
      "epoch": 0.13147710478950667,
      "grad_norm": 0.15599028766155243,
      "learning_rate": 9.562057226975804e-05,
      "loss": 0.8504,
      "mean_token_accuracy": 0.756127867847681,
      "num_tokens": 45606059.0,
      "step": 1429
    },
    {
      "entropy": 0.8958615437150002,
      "epoch": 0.13156911116094788,
      "grad_norm": 0.15909335017204285,
      "learning_rate": 9.561750544361641e-05,
      "loss": 0.9109,
      "mean_token_accuracy": 0.7405203878879547,
      "num_tokens": 45638336.0,
      "step": 1430
    },
    {
      "entropy": 0.8807841539382935,
      "epoch": 0.13166111753238913,
      "grad_norm": 0.1510554701089859,
      "learning_rate": 9.561443861747478e-05,
      "loss": 0.8624,
      "mean_token_accuracy": 0.7528376281261444,
      "num_tokens": 45670330.0,
      "step": 1431
    },
    {
      "entropy": 1.0259534195065498,
      "epoch": 0.13175312390383034,
      "grad_norm": 0.16470204293727875,
      "learning_rate": 9.561137179133314e-05,
      "loss": 1.0252,
      "mean_token_accuracy": 0.7134462296962738,
      "num_tokens": 45701936.0,
      "step": 1432
    },
    {
      "entropy": 0.9712209701538086,
      "epoch": 0.13184513027527156,
      "grad_norm": 0.15517905354499817,
      "learning_rate": 9.560830496519152e-05,
      "loss": 0.9503,
      "mean_token_accuracy": 0.7307788021862507,
      "num_tokens": 45734261.0,
      "step": 1433
    },
    {
      "entropy": 1.0422740504145622,
      "epoch": 0.13193713664671278,
      "grad_norm": 0.16536995768547058,
      "learning_rate": 9.56052381390499e-05,
      "loss": 1.0218,
      "mean_token_accuracy": 0.7202978283166885,
      "num_tokens": 45765717.0,
      "step": 1434
    },
    {
      "entropy": 0.8892856333404779,
      "epoch": 0.132029143018154,
      "grad_norm": 0.15771764516830444,
      "learning_rate": 9.560217131290827e-05,
      "loss": 0.8981,
      "mean_token_accuracy": 0.7388820014894009,
      "num_tokens": 45797884.0,
      "step": 1435
    },
    {
      "entropy": 0.8051073830574751,
      "epoch": 0.13212114938959524,
      "grad_norm": 0.15174847841262817,
      "learning_rate": 9.559910448676666e-05,
      "loss": 0.8152,
      "mean_token_accuracy": 0.763037633150816,
      "num_tokens": 45830191.0,
      "step": 1436
    },
    {
      "entropy": 0.9558395482599735,
      "epoch": 0.13221315576103645,
      "grad_norm": 0.16841810941696167,
      "learning_rate": 9.559603766062502e-05,
      "loss": 0.965,
      "mean_token_accuracy": 0.7280736938118935,
      "num_tokens": 45862439.0,
      "step": 1437
    },
    {
      "entropy": 0.9421539343893528,
      "epoch": 0.13230516213247767,
      "grad_norm": 0.16177257895469666,
      "learning_rate": 9.559297083448339e-05,
      "loss": 0.9412,
      "mean_token_accuracy": 0.7356430999934673,
      "num_tokens": 45894635.0,
      "step": 1438
    },
    {
      "entropy": 0.957996055483818,
      "epoch": 0.1323971685039189,
      "grad_norm": 0.1660410612821579,
      "learning_rate": 9.558990400834177e-05,
      "loss": 0.9481,
      "mean_token_accuracy": 0.7355706579983234,
      "num_tokens": 45926189.0,
      "step": 1439
    },
    {
      "entropy": 0.8232878874987364,
      "epoch": 0.1324891748753601,
      "grad_norm": 0.15279632806777954,
      "learning_rate": 9.558683718220015e-05,
      "loss": 0.8207,
      "mean_token_accuracy": 0.7650743238627911,
      "num_tokens": 45957619.0,
      "step": 1440
    },
    {
      "entropy": 0.8591838888823986,
      "epoch": 0.13258118124680135,
      "grad_norm": 0.153257817029953,
      "learning_rate": 9.558377035605852e-05,
      "loss": 0.8848,
      "mean_token_accuracy": 0.7529977038502693,
      "num_tokens": 45989559.0,
      "step": 1441
    },
    {
      "entropy": 0.8238626755774021,
      "epoch": 0.13267318761824257,
      "grad_norm": 0.1557403802871704,
      "learning_rate": 9.558070352991689e-05,
      "loss": 0.8382,
      "mean_token_accuracy": 0.7654239237308502,
      "num_tokens": 46021306.0,
      "step": 1442
    },
    {
      "entropy": 0.984395120292902,
      "epoch": 0.13276519398968378,
      "grad_norm": 0.1676916778087616,
      "learning_rate": 9.557763670377526e-05,
      "loss": 0.9805,
      "mean_token_accuracy": 0.7303400635719299,
      "num_tokens": 46053086.0,
      "step": 1443
    },
    {
      "entropy": 0.8886575177311897,
      "epoch": 0.132857200361125,
      "grad_norm": 0.15613077580928802,
      "learning_rate": 9.557456987763365e-05,
      "loss": 0.8919,
      "mean_token_accuracy": 0.7450644709169865,
      "num_tokens": 46085399.0,
      "step": 1444
    },
    {
      "entropy": 0.9149897582828999,
      "epoch": 0.13294920673256622,
      "grad_norm": 0.1613151729106903,
      "learning_rate": 9.557150305149202e-05,
      "loss": 0.9141,
      "mean_token_accuracy": 0.7407809644937515,
      "num_tokens": 46116288.0,
      "step": 1445
    },
    {
      "entropy": 0.8291662000119686,
      "epoch": 0.13304121310400746,
      "grad_norm": 0.1582048386335373,
      "learning_rate": 9.556843622535039e-05,
      "loss": 0.8268,
      "mean_token_accuracy": 0.7661645747721195,
      "num_tokens": 46148607.0,
      "step": 1446
    },
    {
      "entropy": 0.951840054243803,
      "epoch": 0.13313321947544868,
      "grad_norm": 0.16414111852645874,
      "learning_rate": 9.556536939920876e-05,
      "loss": 0.9642,
      "mean_token_accuracy": 0.7255589812994003,
      "num_tokens": 46180393.0,
      "step": 1447
    },
    {
      "entropy": 0.9651191756129265,
      "epoch": 0.1332252258468899,
      "grad_norm": 0.1618623286485672,
      "learning_rate": 9.556230257306714e-05,
      "loss": 0.9807,
      "mean_token_accuracy": 0.7258917205035686,
      "num_tokens": 46212271.0,
      "step": 1448
    },
    {
      "entropy": 1.0434480756521225,
      "epoch": 0.1333172322183311,
      "grad_norm": 0.16256369650363922,
      "learning_rate": 9.555923574692552e-05,
      "loss": 1.0313,
      "mean_token_accuracy": 0.7154682017862797,
      "num_tokens": 46244359.0,
      "step": 1449
    },
    {
      "entropy": 0.8898940347135067,
      "epoch": 0.13340923858977236,
      "grad_norm": 0.15305295586585999,
      "learning_rate": 9.555616892078389e-05,
      "loss": 0.8708,
      "mean_token_accuracy": 0.7529643960297108,
      "num_tokens": 46275908.0,
      "step": 1450
    },
    {
      "entropy": 0.8457957338541746,
      "epoch": 0.13350124496121357,
      "grad_norm": 0.15215720236301422,
      "learning_rate": 9.555310209464225e-05,
      "loss": 0.8461,
      "mean_token_accuracy": 0.7600013390183449,
      "num_tokens": 46308148.0,
      "step": 1451
    },
    {
      "entropy": 0.8477643057703972,
      "epoch": 0.1335932513326548,
      "grad_norm": 0.15866334736347198,
      "learning_rate": 9.555003526850064e-05,
      "loss": 0.8166,
      "mean_token_accuracy": 0.7661799043416977,
      "num_tokens": 46338730.0,
      "step": 1452
    },
    {
      "entropy": 0.8573301117867231,
      "epoch": 0.133685257704096,
      "grad_norm": 0.15068146586418152,
      "learning_rate": 9.5546968442359e-05,
      "loss": 0.8414,
      "mean_token_accuracy": 0.7616315670311451,
      "num_tokens": 46371147.0,
      "step": 1453
    },
    {
      "entropy": 0.8572186846286058,
      "epoch": 0.13377726407553722,
      "grad_norm": 0.15416494011878967,
      "learning_rate": 9.554390161621739e-05,
      "loss": 0.8502,
      "mean_token_accuracy": 0.7541092596948147,
      "num_tokens": 46403377.0,
      "step": 1454
    },
    {
      "entropy": 0.9415202960371971,
      "epoch": 0.13386927044697847,
      "grad_norm": 0.1650015413761139,
      "learning_rate": 9.554083479007575e-05,
      "loss": 0.9504,
      "mean_token_accuracy": 0.7292389497160912,
      "num_tokens": 46434624.0,
      "step": 1455
    },
    {
      "entropy": 1.0338180717080832,
      "epoch": 0.13396127681841968,
      "grad_norm": 0.1708630621433258,
      "learning_rate": 9.553776796393413e-05,
      "loss": 1.0459,
      "mean_token_accuracy": 0.710181187838316,
      "num_tokens": 46466625.0,
      "step": 1456
    },
    {
      "entropy": 0.890923760831356,
      "epoch": 0.1340532831898609,
      "grad_norm": 0.1639585942029953,
      "learning_rate": 9.55347011377925e-05,
      "loss": 0.9174,
      "mean_token_accuracy": 0.7409874685108662,
      "num_tokens": 46498859.0,
      "step": 1457
    },
    {
      "entropy": 0.7902647480368614,
      "epoch": 0.13414528956130212,
      "grad_norm": 0.15480412542819977,
      "learning_rate": 9.553163431165087e-05,
      "loss": 0.7869,
      "mean_token_accuracy": 0.7682091109454632,
      "num_tokens": 46530861.0,
      "step": 1458
    },
    {
      "entropy": 0.9016091823577881,
      "epoch": 0.13423729593274333,
      "grad_norm": 0.167546346783638,
      "learning_rate": 9.552856748550925e-05,
      "loss": 0.9042,
      "mean_token_accuracy": 0.7400625944137573,
      "num_tokens": 46561437.0,
      "step": 1459
    },
    {
      "entropy": 0.9803922958672047,
      "epoch": 0.13432930230418458,
      "grad_norm": 0.16903378069400787,
      "learning_rate": 9.552550065936763e-05,
      "loss": 1.0007,
      "mean_token_accuracy": 0.7213688008487225,
      "num_tokens": 46592797.0,
      "step": 1460
    },
    {
      "entropy": 0.9514728561043739,
      "epoch": 0.1344213086756258,
      "grad_norm": 0.16266781091690063,
      "learning_rate": 9.5522433833226e-05,
      "loss": 0.9538,
      "mean_token_accuracy": 0.7330202981829643,
      "num_tokens": 46624232.0,
      "step": 1461
    },
    {
      "entropy": 0.8207721151411533,
      "epoch": 0.134513315047067,
      "grad_norm": 0.15729418396949768,
      "learning_rate": 9.551936700708437e-05,
      "loss": 0.8232,
      "mean_token_accuracy": 0.7674405314028263,
      "num_tokens": 46656125.0,
      "step": 1462
    },
    {
      "entropy": 0.9634261578321457,
      "epoch": 0.13460532141850823,
      "grad_norm": 0.17248524725437164,
      "learning_rate": 9.551630018094274e-05,
      "loss": 0.9809,
      "mean_token_accuracy": 0.7236034795641899,
      "num_tokens": 46686896.0,
      "step": 1463
    },
    {
      "entropy": 0.9230004139244556,
      "epoch": 0.13469732778994944,
      "grad_norm": 0.15707097947597504,
      "learning_rate": 9.551323335480112e-05,
      "loss": 0.9109,
      "mean_token_accuracy": 0.7411488257348537,
      "num_tokens": 46718927.0,
      "step": 1464
    },
    {
      "entropy": 0.8120103217661381,
      "epoch": 0.1347893341613907,
      "grad_norm": 0.14958244562149048,
      "learning_rate": 9.55101665286595e-05,
      "loss": 0.7927,
      "mean_token_accuracy": 0.7734591253101826,
      "num_tokens": 46751046.0,
      "step": 1465
    },
    {
      "entropy": 0.9720970839262009,
      "epoch": 0.1348813405328319,
      "grad_norm": 0.16194313764572144,
      "learning_rate": 9.550709970251787e-05,
      "loss": 0.9734,
      "mean_token_accuracy": 0.7252835631370544,
      "num_tokens": 46783194.0,
      "step": 1466
    },
    {
      "entropy": 0.8762975186109543,
      "epoch": 0.13497334690427312,
      "grad_norm": 0.1605098992586136,
      "learning_rate": 9.550403287637625e-05,
      "loss": 0.8747,
      "mean_token_accuracy": 0.7494844533503056,
      "num_tokens": 46814558.0,
      "step": 1467
    },
    {
      "entropy": 1.0193149112164974,
      "epoch": 0.13506535327571434,
      "grad_norm": 0.16349397599697113,
      "learning_rate": 9.550096605023462e-05,
      "loss": 1.0085,
      "mean_token_accuracy": 0.7192102633416653,
      "num_tokens": 46846872.0,
      "step": 1468
    },
    {
      "entropy": 0.8644775934517384,
      "epoch": 0.13515735964715556,
      "grad_norm": 0.15632364153862,
      "learning_rate": 9.549789922409298e-05,
      "loss": 0.8657,
      "mean_token_accuracy": 0.758609127253294,
      "num_tokens": 46878288.0,
      "step": 1469
    },
    {
      "entropy": 0.9141164738684893,
      "epoch": 0.1352493660185968,
      "grad_norm": 0.16748885810375214,
      "learning_rate": 9.549483239795137e-05,
      "loss": 0.9252,
      "mean_token_accuracy": 0.737995482981205,
      "num_tokens": 46910152.0,
      "step": 1470
    },
    {
      "entropy": 0.8981011733412743,
      "epoch": 0.13534137239003802,
      "grad_norm": 0.15873192250728607,
      "learning_rate": 9.549176557180975e-05,
      "loss": 0.9334,
      "mean_token_accuracy": 0.741527620702982,
      "num_tokens": 46942594.0,
      "step": 1471
    },
    {
      "entropy": 0.8247240036725998,
      "epoch": 0.13543337876147923,
      "grad_norm": 0.15040016174316406,
      "learning_rate": 9.548869874566812e-05,
      "loss": 0.8162,
      "mean_token_accuracy": 0.7664057724177837,
      "num_tokens": 46974892.0,
      "step": 1472
    },
    {
      "entropy": 0.7996957898139954,
      "epoch": 0.13552538513292045,
      "grad_norm": 0.15322554111480713,
      "learning_rate": 9.548563191952648e-05,
      "loss": 0.7901,
      "mean_token_accuracy": 0.7710999585688114,
      "num_tokens": 47005454.0,
      "step": 1473
    },
    {
      "entropy": 0.8677499089390039,
      "epoch": 0.13561739150436167,
      "grad_norm": 0.15616901218891144,
      "learning_rate": 9.548256509338485e-05,
      "loss": 0.8612,
      "mean_token_accuracy": 0.7536361999809742,
      "num_tokens": 47037199.0,
      "step": 1474
    },
    {
      "entropy": 0.9226998798549175,
      "epoch": 0.1357093978758029,
      "grad_norm": 0.16280454397201538,
      "learning_rate": 9.547949826724325e-05,
      "loss": 0.9352,
      "mean_token_accuracy": 0.7386032938957214,
      "num_tokens": 47068972.0,
      "step": 1475
    },
    {
      "entropy": 0.8511421792209148,
      "epoch": 0.13580140424724413,
      "grad_norm": 0.15268120169639587,
      "learning_rate": 9.547643144110161e-05,
      "loss": 0.8197,
      "mean_token_accuracy": 0.7648099288344383,
      "num_tokens": 47100795.0,
      "step": 1476
    },
    {
      "entropy": 0.991644598543644,
      "epoch": 0.13589341061868535,
      "grad_norm": 0.16335169970989227,
      "learning_rate": 9.547336461495998e-05,
      "loss": 0.9976,
      "mean_token_accuracy": 0.7279123477637768,
      "num_tokens": 47132512.0,
      "step": 1477
    },
    {
      "entropy": 0.9475472010672092,
      "epoch": 0.13598541699012656,
      "grad_norm": 0.1582222878932953,
      "learning_rate": 9.547029778881835e-05,
      "loss": 0.9246,
      "mean_token_accuracy": 0.7349614687263966,
      "num_tokens": 47164785.0,
      "step": 1478
    },
    {
      "entropy": 0.8885447606444359,
      "epoch": 0.13607742336156778,
      "grad_norm": 0.1621338278055191,
      "learning_rate": 9.546723096267673e-05,
      "loss": 0.893,
      "mean_token_accuracy": 0.7510957531630993,
      "num_tokens": 47196582.0,
      "step": 1479
    },
    {
      "entropy": 0.9588793590664864,
      "epoch": 0.13616942973300902,
      "grad_norm": 0.16168516874313354,
      "learning_rate": 9.546416413653511e-05,
      "loss": 0.9617,
      "mean_token_accuracy": 0.7329396195709705,
      "num_tokens": 47228827.0,
      "step": 1480
    },
    {
      "entropy": 0.8150004111230373,
      "epoch": 0.13626143610445024,
      "grad_norm": 0.15117791295051575,
      "learning_rate": 9.546109731039348e-05,
      "loss": 0.8132,
      "mean_token_accuracy": 0.7682842388749123,
      "num_tokens": 47260496.0,
      "step": 1481
    },
    {
      "entropy": 0.9063622020184994,
      "epoch": 0.13635344247589146,
      "grad_norm": 0.16281457245349884,
      "learning_rate": 9.545803048425185e-05,
      "loss": 0.9158,
      "mean_token_accuracy": 0.7445373684167862,
      "num_tokens": 47291856.0,
      "step": 1482
    },
    {
      "entropy": 0.8509590812027454,
      "epoch": 0.13644544884733267,
      "grad_norm": 0.1577964723110199,
      "learning_rate": 9.545496365811023e-05,
      "loss": 0.8518,
      "mean_token_accuracy": 0.76293920353055,
      "num_tokens": 47323829.0,
      "step": 1483
    },
    {
      "entropy": 0.869157712906599,
      "epoch": 0.1365374552187739,
      "grad_norm": 0.15847274661064148,
      "learning_rate": 9.54518968319686e-05,
      "loss": 0.8949,
      "mean_token_accuracy": 0.7520448453724384,
      "num_tokens": 47356024.0,
      "step": 1484
    },
    {
      "entropy": 0.8899447582662106,
      "epoch": 0.13662946159021513,
      "grad_norm": 0.15856720507144928,
      "learning_rate": 9.544883000582698e-05,
      "loss": 0.8927,
      "mean_token_accuracy": 0.7470324896275997,
      "num_tokens": 47387872.0,
      "step": 1485
    },
    {
      "entropy": 0.8399642929434776,
      "epoch": 0.13672146796165635,
      "grad_norm": 0.15805940330028534,
      "learning_rate": 9.544576317968535e-05,
      "loss": 0.8435,
      "mean_token_accuracy": 0.7591837681829929,
      "num_tokens": 47419440.0,
      "step": 1486
    },
    {
      "entropy": 0.862132953479886,
      "epoch": 0.13681347433309757,
      "grad_norm": 0.15930739045143127,
      "learning_rate": 9.544269635354373e-05,
      "loss": 0.8506,
      "mean_token_accuracy": 0.7596270404756069,
      "num_tokens": 47450738.0,
      "step": 1487
    },
    {
      "entropy": 0.9156553968787193,
      "epoch": 0.13690548070453878,
      "grad_norm": 0.16053490340709686,
      "learning_rate": 9.54396295274021e-05,
      "loss": 0.9215,
      "mean_token_accuracy": 0.7399528473615646,
      "num_tokens": 47482652.0,
      "step": 1488
    },
    {
      "entropy": 0.8765908479690552,
      "epoch": 0.13699748707598,
      "grad_norm": 0.15102703869342804,
      "learning_rate": 9.543656270126046e-05,
      "loss": 0.8642,
      "mean_token_accuracy": 0.7519312389194965,
      "num_tokens": 47515203.0,
      "step": 1489
    },
    {
      "entropy": 0.929722972214222,
      "epoch": 0.13708949344742125,
      "grad_norm": 0.16479045152664185,
      "learning_rate": 9.543349587511885e-05,
      "loss": 0.934,
      "mean_token_accuracy": 0.7361554466187954,
      "num_tokens": 47545946.0,
      "step": 1490
    },
    {
      "entropy": 0.8781662285327911,
      "epoch": 0.13718149981886246,
      "grad_norm": 0.15641742944717407,
      "learning_rate": 9.543042904897723e-05,
      "loss": 0.871,
      "mean_token_accuracy": 0.7544177807867527,
      "num_tokens": 47578279.0,
      "step": 1491
    },
    {
      "entropy": 0.9010445736348629,
      "epoch": 0.13727350619030368,
      "grad_norm": 0.15929129719734192,
      "learning_rate": 9.54273622228356e-05,
      "loss": 0.8675,
      "mean_token_accuracy": 0.7536949478089809,
      "num_tokens": 47609756.0,
      "step": 1492
    },
    {
      "entropy": 0.8101601470261812,
      "epoch": 0.1373655125617449,
      "grad_norm": 0.15118253231048584,
      "learning_rate": 9.542429539669396e-05,
      "loss": 0.8021,
      "mean_token_accuracy": 0.769440121948719,
      "num_tokens": 47641689.0,
      "step": 1493
    },
    {
      "entropy": 0.9826366417109966,
      "epoch": 0.1374575189331861,
      "grad_norm": 0.17489178478717804,
      "learning_rate": 9.542122857055233e-05,
      "loss": 0.9801,
      "mean_token_accuracy": 0.7230763360857964,
      "num_tokens": 47673512.0,
      "step": 1494
    },
    {
      "entropy": 0.9948286898434162,
      "epoch": 0.13754952530462736,
      "grad_norm": 0.16823458671569824,
      "learning_rate": 9.541816174441071e-05,
      "loss": 1.0026,
      "mean_token_accuracy": 0.7178121320903301,
      "num_tokens": 47705429.0,
      "step": 1495
    },
    {
      "entropy": 0.903372161090374,
      "epoch": 0.13764153167606857,
      "grad_norm": 0.1602449119091034,
      "learning_rate": 9.54150949182691e-05,
      "loss": 0.9228,
      "mean_token_accuracy": 0.7376360595226288,
      "num_tokens": 47737783.0,
      "step": 1496
    },
    {
      "entropy": 0.8364967592060566,
      "epoch": 0.1377335380475098,
      "grad_norm": 0.15546752512454987,
      "learning_rate": 9.541202809212746e-05,
      "loss": 0.8458,
      "mean_token_accuracy": 0.7578377984464169,
      "num_tokens": 47770370.0,
      "step": 1497
    },
    {
      "entropy": 0.9463928639888763,
      "epoch": 0.137825544418951,
      "grad_norm": 0.16655172407627106,
      "learning_rate": 9.540896126598584e-05,
      "loss": 0.9558,
      "mean_token_accuracy": 0.7322683446109295,
      "num_tokens": 47802731.0,
      "step": 1498
    },
    {
      "entropy": 0.9242000319063663,
      "epoch": 0.13791755079039222,
      "grad_norm": 0.1686069816350937,
      "learning_rate": 9.540589443984421e-05,
      "loss": 0.9812,
      "mean_token_accuracy": 0.7295314557850361,
      "num_tokens": 47834965.0,
      "step": 1499
    },
    {
      "entropy": 0.8720885049551725,
      "epoch": 0.13800955716183347,
      "grad_norm": 0.15979434549808502,
      "learning_rate": 9.540282761370258e-05,
      "loss": 0.8728,
      "mean_token_accuracy": 0.7521369494497776,
      "num_tokens": 47866879.0,
      "step": 1500
    },
    {
      "entropy": 0.9144723750650883,
      "epoch": 0.13810156353327468,
      "grad_norm": 0.16163739562034607,
      "learning_rate": 9.539976078756096e-05,
      "loss": 0.9218,
      "mean_token_accuracy": 0.7431746833026409,
      "num_tokens": 47899104.0,
      "step": 1501
    },
    {
      "entropy": 0.9037791453301907,
      "epoch": 0.1381935699047159,
      "grad_norm": 0.16473473608493805,
      "learning_rate": 9.539669396141934e-05,
      "loss": 0.8989,
      "mean_token_accuracy": 0.7471611127257347,
      "num_tokens": 47929641.0,
      "step": 1502
    },
    {
      "entropy": 0.873839758336544,
      "epoch": 0.13828557627615712,
      "grad_norm": 0.15165023505687714,
      "learning_rate": 9.539362713527771e-05,
      "loss": 0.8597,
      "mean_token_accuracy": 0.7544801756739616,
      "num_tokens": 47962107.0,
      "step": 1503
    },
    {
      "entropy": 0.9012771509587765,
      "epoch": 0.13837758264759834,
      "grad_norm": 0.15976887941360474,
      "learning_rate": 9.539056030913608e-05,
      "loss": 0.8758,
      "mean_token_accuracy": 0.7531430572271347,
      "num_tokens": 47994049.0,
      "step": 1504
    },
    {
      "entropy": 0.9222031719982624,
      "epoch": 0.13846958901903958,
      "grad_norm": 0.15270550549030304,
      "learning_rate": 9.538749348299444e-05,
      "loss": 0.8901,
      "mean_token_accuracy": 0.7466108947992325,
      "num_tokens": 48026817.0,
      "step": 1505
    },
    {
      "entropy": 0.8834109716117382,
      "epoch": 0.1385615953904808,
      "grad_norm": 0.1656990349292755,
      "learning_rate": 9.538442665685283e-05,
      "loss": 0.8857,
      "mean_token_accuracy": 0.7494886107742786,
      "num_tokens": 48057884.0,
      "step": 1506
    },
    {
      "entropy": 0.8789761662483215,
      "epoch": 0.138653601761922,
      "grad_norm": 0.15314273536205292,
      "learning_rate": 9.538135983071121e-05,
      "loss": 0.8744,
      "mean_token_accuracy": 0.7524401769042015,
      "num_tokens": 48090408.0,
      "step": 1507
    },
    {
      "entropy": 0.860306516289711,
      "epoch": 0.13874560813336323,
      "grad_norm": 0.1534280776977539,
      "learning_rate": 9.537829300456958e-05,
      "loss": 0.851,
      "mean_token_accuracy": 0.7576924115419388,
      "num_tokens": 48122294.0,
      "step": 1508
    },
    {
      "entropy": 0.933868020772934,
      "epoch": 0.13883761450480445,
      "grad_norm": 0.15948344767093658,
      "learning_rate": 9.537522617842794e-05,
      "loss": 0.9322,
      "mean_token_accuracy": 0.7369281835854053,
      "num_tokens": 48154367.0,
      "step": 1509
    },
    {
      "entropy": 0.9980332925915718,
      "epoch": 0.1389296208762457,
      "grad_norm": 0.1685190349817276,
      "learning_rate": 9.537215935228632e-05,
      "loss": 1.0021,
      "mean_token_accuracy": 0.7187242805957794,
      "num_tokens": 48185814.0,
      "step": 1510
    },
    {
      "entropy": 0.8835068270564079,
      "epoch": 0.1390216272476869,
      "grad_norm": 0.1598827838897705,
      "learning_rate": 9.536909252614469e-05,
      "loss": 0.8939,
      "mean_token_accuracy": 0.746815487742424,
      "num_tokens": 48217618.0,
      "step": 1511
    },
    {
      "entropy": 0.8843892924487591,
      "epoch": 0.13911363361912812,
      "grad_norm": 0.16501133143901825,
      "learning_rate": 9.536602570000307e-05,
      "loss": 0.8891,
      "mean_token_accuracy": 0.74925871565938,
      "num_tokens": 48249118.0,
      "step": 1512
    },
    {
      "entropy": 0.9995120503008366,
      "epoch": 0.13920563999056934,
      "grad_norm": 0.17007550597190857,
      "learning_rate": 9.536295887386144e-05,
      "loss": 1.015,
      "mean_token_accuracy": 0.7165202684700489,
      "num_tokens": 48280919.0,
      "step": 1513
    },
    {
      "entropy": 0.9234212972223759,
      "epoch": 0.13929764636201056,
      "grad_norm": 0.16136333346366882,
      "learning_rate": 9.535989204771982e-05,
      "loss": 0.9346,
      "mean_token_accuracy": 0.7414928078651428,
      "num_tokens": 48312907.0,
      "step": 1514
    },
    {
      "entropy": 0.785212080925703,
      "epoch": 0.1393896527334518,
      "grad_norm": 0.1573466956615448,
      "learning_rate": 9.535682522157819e-05,
      "loss": 0.7782,
      "mean_token_accuracy": 0.7797213345766068,
      "num_tokens": 48344096.0,
      "step": 1515
    },
    {
      "entropy": 0.872599221765995,
      "epoch": 0.13948165910489302,
      "grad_norm": 0.15678797662258148,
      "learning_rate": 9.535375839543656e-05,
      "loss": 0.8743,
      "mean_token_accuracy": 0.7531127072870731,
      "num_tokens": 48375872.0,
      "step": 1516
    },
    {
      "entropy": 0.93693882599473,
      "epoch": 0.13957366547633424,
      "grad_norm": 0.1591949164867401,
      "learning_rate": 9.535069156929494e-05,
      "loss": 0.9506,
      "mean_token_accuracy": 0.7316819168627262,
      "num_tokens": 48408547.0,
      "step": 1517
    },
    {
      "entropy": 0.9289727360010147,
      "epoch": 0.13966567184777545,
      "grad_norm": 0.16631434857845306,
      "learning_rate": 9.534762474315332e-05,
      "loss": 0.9378,
      "mean_token_accuracy": 0.7328994087874889,
      "num_tokens": 48440266.0,
      "step": 1518
    },
    {
      "entropy": 0.8922602757811546,
      "epoch": 0.13975767821921667,
      "grad_norm": 0.1614350825548172,
      "learning_rate": 9.534455791701169e-05,
      "loss": 0.8898,
      "mean_token_accuracy": 0.7465514466166496,
      "num_tokens": 48470750.0,
      "step": 1519
    },
    {
      "entropy": 0.8415650241076946,
      "epoch": 0.1398496845906579,
      "grad_norm": 0.15855540335178375,
      "learning_rate": 9.534149109087006e-05,
      "loss": 0.8539,
      "mean_token_accuracy": 0.7541022598743439,
      "num_tokens": 48501446.0,
      "step": 1520
    },
    {
      "entropy": 0.8369389250874519,
      "epoch": 0.13994169096209913,
      "grad_norm": 0.15269261598587036,
      "learning_rate": 9.533842426472843e-05,
      "loss": 0.8317,
      "mean_token_accuracy": 0.761881809681654,
      "num_tokens": 48533915.0,
      "step": 1521
    },
    {
      "entropy": 0.9087646678090096,
      "epoch": 0.14003369733354035,
      "grad_norm": 0.17284798622131348,
      "learning_rate": 9.533535743858682e-05,
      "loss": 0.9193,
      "mean_token_accuracy": 0.7405981458723545,
      "num_tokens": 48565896.0,
      "step": 1522
    },
    {
      "entropy": 0.920044731348753,
      "epoch": 0.14012570370498156,
      "grad_norm": 0.1634625345468521,
      "learning_rate": 9.533229061244519e-05,
      "loss": 0.9436,
      "mean_token_accuracy": 0.7330580167472363,
      "num_tokens": 48598362.0,
      "step": 1523
    },
    {
      "entropy": 0.9196216017007828,
      "epoch": 0.14021771007642278,
      "grad_norm": 0.16235344111919403,
      "learning_rate": 9.532922378630356e-05,
      "loss": 0.8987,
      "mean_token_accuracy": 0.7421354539692402,
      "num_tokens": 48629908.0,
      "step": 1524
    },
    {
      "entropy": 0.9084957912564278,
      "epoch": 0.14030971644786402,
      "grad_norm": 0.16361317038536072,
      "learning_rate": 9.532615696016192e-05,
      "loss": 0.8956,
      "mean_token_accuracy": 0.7468879707157612,
      "num_tokens": 48661682.0,
      "step": 1525
    },
    {
      "entropy": 0.9380924291908741,
      "epoch": 0.14040172281930524,
      "grad_norm": 0.16031771898269653,
      "learning_rate": 9.53230901340203e-05,
      "loss": 0.944,
      "mean_token_accuracy": 0.7310221381485462,
      "num_tokens": 48693758.0,
      "step": 1526
    },
    {
      "entropy": 0.8924711793661118,
      "epoch": 0.14049372919074646,
      "grad_norm": 0.15422584116458893,
      "learning_rate": 9.532002330787869e-05,
      "loss": 0.8683,
      "mean_token_accuracy": 0.7507140003144741,
      "num_tokens": 48725666.0,
      "step": 1527
    },
    {
      "entropy": 0.9133307598531246,
      "epoch": 0.14058573556218767,
      "grad_norm": 0.16041408479213715,
      "learning_rate": 9.531695648173705e-05,
      "loss": 0.9084,
      "mean_token_accuracy": 0.7393610663712025,
      "num_tokens": 48757335.0,
      "step": 1528
    },
    {
      "entropy": 0.935918066650629,
      "epoch": 0.1406777419336289,
      "grad_norm": 0.162759929895401,
      "learning_rate": 9.531388965559544e-05,
      "loss": 0.9205,
      "mean_token_accuracy": 0.7380355782806873,
      "num_tokens": 48789253.0,
      "step": 1529
    },
    {
      "entropy": 0.8984809126704931,
      "epoch": 0.14076974830507014,
      "grad_norm": 0.15972734987735748,
      "learning_rate": 9.53108228294538e-05,
      "loss": 0.9031,
      "mean_token_accuracy": 0.7473898828029633,
      "num_tokens": 48821227.0,
      "step": 1530
    },
    {
      "entropy": 0.9905790127813816,
      "epoch": 0.14086175467651135,
      "grad_norm": 0.16385893523693085,
      "learning_rate": 9.530775600331217e-05,
      "loss": 0.9924,
      "mean_token_accuracy": 0.7238439582288265,
      "num_tokens": 48853699.0,
      "step": 1531
    },
    {
      "entropy": 0.8665602579712868,
      "epoch": 0.14095376104795257,
      "grad_norm": 0.1583719104528427,
      "learning_rate": 9.530468917717055e-05,
      "loss": 0.8892,
      "mean_token_accuracy": 0.7493539080023766,
      "num_tokens": 48885328.0,
      "step": 1532
    },
    {
      "entropy": 0.7871650364249945,
      "epoch": 0.14104576741939379,
      "grad_norm": 0.1514895260334015,
      "learning_rate": 9.530162235102893e-05,
      "loss": 0.7903,
      "mean_token_accuracy": 0.7727830708026886,
      "num_tokens": 48917072.0,
      "step": 1533
    },
    {
      "entropy": 0.9596593640744686,
      "epoch": 0.141137773790835,
      "grad_norm": 0.16787609457969666,
      "learning_rate": 9.52985555248873e-05,
      "loss": 0.9763,
      "mean_token_accuracy": 0.7321631051599979,
      "num_tokens": 48948940.0,
      "step": 1534
    },
    {
      "entropy": 0.7926683686673641,
      "epoch": 0.14122978016227625,
      "grad_norm": 0.1529580056667328,
      "learning_rate": 9.529548869874567e-05,
      "loss": 0.7771,
      "mean_token_accuracy": 0.7764681428670883,
      "num_tokens": 48980773.0,
      "step": 1535
    },
    {
      "entropy": 0.9480520263314247,
      "epoch": 0.14132178653371746,
      "grad_norm": 0.16902871429920197,
      "learning_rate": 9.529242187260404e-05,
      "loss": 0.9544,
      "mean_token_accuracy": 0.7349111922085285,
      "num_tokens": 49012472.0,
      "step": 1536
    },
    {
      "entropy": 0.9515025764703751,
      "epoch": 0.14141379290515868,
      "grad_norm": 0.16660866141319275,
      "learning_rate": 9.528935504646242e-05,
      "loss": 0.9646,
      "mean_token_accuracy": 0.7285832278430462,
      "num_tokens": 49044155.0,
      "step": 1537
    },
    {
      "entropy": 0.9830679893493652,
      "epoch": 0.1415057992765999,
      "grad_norm": 0.1669303923845291,
      "learning_rate": 9.52862882203208e-05,
      "loss": 0.9926,
      "mean_token_accuracy": 0.7214728817343712,
      "num_tokens": 49075698.0,
      "step": 1538
    },
    {
      "entropy": 0.9938245601952076,
      "epoch": 0.14159780564804111,
      "grad_norm": 0.1661204695701599,
      "learning_rate": 9.528322139417917e-05,
      "loss": 0.9988,
      "mean_token_accuracy": 0.7248219549655914,
      "num_tokens": 49106781.0,
      "step": 1539
    },
    {
      "entropy": 0.9125274457037449,
      "epoch": 0.14168981201948236,
      "grad_norm": 0.15893520414829254,
      "learning_rate": 9.528015456803754e-05,
      "loss": 0.889,
      "mean_token_accuracy": 0.7503414042294025,
      "num_tokens": 49138895.0,
      "step": 1540
    },
    {
      "entropy": 0.8700472079217434,
      "epoch": 0.14178181839092358,
      "grad_norm": 0.1545281857252121,
      "learning_rate": 9.527708774189592e-05,
      "loss": 0.8607,
      "mean_token_accuracy": 0.7518544606864452,
      "num_tokens": 49171080.0,
      "step": 1541
    },
    {
      "entropy": 0.9424622878432274,
      "epoch": 0.1418738247623648,
      "grad_norm": 0.16081099212169647,
      "learning_rate": 9.527402091575429e-05,
      "loss": 0.9429,
      "mean_token_accuracy": 0.738890890032053,
      "num_tokens": 49203435.0,
      "step": 1542
    },
    {
      "entropy": 0.9055228345096111,
      "epoch": 0.141965831133806,
      "grad_norm": 0.15805533528327942,
      "learning_rate": 9.527095408961267e-05,
      "loss": 0.903,
      "mean_token_accuracy": 0.7435910925269127,
      "num_tokens": 49235070.0,
      "step": 1543
    },
    {
      "entropy": 1.024103932082653,
      "epoch": 0.14205783750524723,
      "grad_norm": 0.17289242148399353,
      "learning_rate": 9.526788726347104e-05,
      "loss": 1.0322,
      "mean_token_accuracy": 0.7105959802865982,
      "num_tokens": 49266629.0,
      "step": 1544
    },
    {
      "entropy": 0.9870683625340462,
      "epoch": 0.14214984387668847,
      "grad_norm": 0.1672264039516449,
      "learning_rate": 9.526482043732942e-05,
      "loss": 0.9879,
      "mean_token_accuracy": 0.7192868515849113,
      "num_tokens": 49298827.0,
      "step": 1545
    },
    {
      "entropy": 0.8665574379265308,
      "epoch": 0.1422418502481297,
      "grad_norm": 0.16494561731815338,
      "learning_rate": 9.526175361118778e-05,
      "loss": 0.8826,
      "mean_token_accuracy": 0.7542853578925133,
      "num_tokens": 49331036.0,
      "step": 1546
    },
    {
      "entropy": 0.9165347069501877,
      "epoch": 0.1423338566195709,
      "grad_norm": 0.16177764534950256,
      "learning_rate": 9.525868678504615e-05,
      "loss": 0.9363,
      "mean_token_accuracy": 0.7352379523217678,
      "num_tokens": 49362332.0,
      "step": 1547
    },
    {
      "entropy": 1.0385689958930016,
      "epoch": 0.14242586299101212,
      "grad_norm": 0.1670178472995758,
      "learning_rate": 9.525561995890453e-05,
      "loss": 1.0408,
      "mean_token_accuracy": 0.7058207131922245,
      "num_tokens": 49394656.0,
      "step": 1548
    },
    {
      "entropy": 0.9335563387721777,
      "epoch": 0.14251786936245334,
      "grad_norm": 0.16502214968204498,
      "learning_rate": 9.525255313276292e-05,
      "loss": 0.9268,
      "mean_token_accuracy": 0.7365954034030437,
      "num_tokens": 49426241.0,
      "step": 1549
    },
    {
      "entropy": 0.8997564129531384,
      "epoch": 0.14260987573389458,
      "grad_norm": 0.15600638091564178,
      "learning_rate": 9.524948630662128e-05,
      "loss": 0.9005,
      "mean_token_accuracy": 0.7421587593853474,
      "num_tokens": 49458019.0,
      "step": 1550
    },
    {
      "entropy": 0.9630580693483353,
      "epoch": 0.1427018821053358,
      "grad_norm": 0.1570965051651001,
      "learning_rate": 9.524641948047965e-05,
      "loss": 0.9537,
      "mean_token_accuracy": 0.7331337705254555,
      "num_tokens": 49490724.0,
      "step": 1551
    },
    {
      "entropy": 0.8782177157700062,
      "epoch": 0.14279388847677701,
      "grad_norm": 0.15707643330097198,
      "learning_rate": 9.524335265433802e-05,
      "loss": 0.8689,
      "mean_token_accuracy": 0.7549863681197166,
      "num_tokens": 49523358.0,
      "step": 1552
    },
    {
      "entropy": 0.9765381813049316,
      "epoch": 0.14288589484821823,
      "grad_norm": 0.16169218719005585,
      "learning_rate": 9.52402858281964e-05,
      "loss": 0.9891,
      "mean_token_accuracy": 0.7207485735416412,
      "num_tokens": 49555207.0,
      "step": 1553
    },
    {
      "entropy": 0.8779881037771702,
      "epoch": 0.14297790121965945,
      "grad_norm": 0.15375083684921265,
      "learning_rate": 9.523721900205478e-05,
      "loss": 0.8582,
      "mean_token_accuracy": 0.7555305399000645,
      "num_tokens": 49587012.0,
      "step": 1554
    },
    {
      "entropy": 0.970871202647686,
      "epoch": 0.1430699075911007,
      "grad_norm": 0.16862569749355316,
      "learning_rate": 9.523415217591315e-05,
      "loss": 0.9559,
      "mean_token_accuracy": 0.7329207547008991,
      "num_tokens": 49618509.0,
      "step": 1555
    },
    {
      "entropy": 0.9228195361793041,
      "epoch": 0.1431619139625419,
      "grad_norm": 0.1545485556125641,
      "learning_rate": 9.523108534977152e-05,
      "loss": 0.892,
      "mean_token_accuracy": 0.747595239430666,
      "num_tokens": 49650944.0,
      "step": 1556
    },
    {
      "entropy": 1.0221968218684196,
      "epoch": 0.14325392033398313,
      "grad_norm": 0.16275210678577423,
      "learning_rate": 9.52280185236299e-05,
      "loss": 1.0181,
      "mean_token_accuracy": 0.7139876931905746,
      "num_tokens": 49682425.0,
      "step": 1557
    },
    {
      "entropy": 0.8230613321065903,
      "epoch": 0.14334592670542434,
      "grad_norm": 0.1514221876859665,
      "learning_rate": 9.522495169748827e-05,
      "loss": 0.8292,
      "mean_token_accuracy": 0.7596264705061913,
      "num_tokens": 49714801.0,
      "step": 1558
    },
    {
      "entropy": 0.830515656620264,
      "epoch": 0.14343793307686556,
      "grad_norm": 0.14815635979175568,
      "learning_rate": 9.522188487134665e-05,
      "loss": 0.8202,
      "mean_token_accuracy": 0.7649087533354759,
      "num_tokens": 49746272.0,
      "step": 1559
    },
    {
      "entropy": 0.8281339034438133,
      "epoch": 0.1435299394483068,
      "grad_norm": 0.15298017859458923,
      "learning_rate": 9.521881804520503e-05,
      "loss": 0.8104,
      "mean_token_accuracy": 0.7703306190669537,
      "num_tokens": 49777762.0,
      "step": 1560
    },
    {
      "entropy": 0.9009035006165504,
      "epoch": 0.14362194581974802,
      "grad_norm": 0.15961822867393494,
      "learning_rate": 9.52157512190634e-05,
      "loss": 0.8866,
      "mean_token_accuracy": 0.748963724821806,
      "num_tokens": 49810102.0,
      "step": 1561
    },
    {
      "entropy": 1.000514842569828,
      "epoch": 0.14371395219118924,
      "grad_norm": 0.1680361032485962,
      "learning_rate": 9.521268439292177e-05,
      "loss": 1.0284,
      "mean_token_accuracy": 0.7126848436892033,
      "num_tokens": 49842213.0,
      "step": 1562
    },
    {
      "entropy": 0.8935029655694962,
      "epoch": 0.14380595856263045,
      "grad_norm": 0.15975192189216614,
      "learning_rate": 9.520961756678015e-05,
      "loss": 0.9051,
      "mean_token_accuracy": 0.7460601069033146,
      "num_tokens": 49873433.0,
      "step": 1563
    },
    {
      "entropy": 0.7654416114091873,
      "epoch": 0.1438979649340717,
      "grad_norm": 0.15598370134830475,
      "learning_rate": 9.520655074063853e-05,
      "loss": 0.7936,
      "mean_token_accuracy": 0.7714344710111618,
      "num_tokens": 49905852.0,
      "step": 1564
    },
    {
      "entropy": 0.8499294575303793,
      "epoch": 0.14398997130551291,
      "grad_norm": 0.1578521579504013,
      "learning_rate": 9.52034839144969e-05,
      "loss": 0.8671,
      "mean_token_accuracy": 0.7553885504603386,
      "num_tokens": 49938458.0,
      "step": 1565
    },
    {
      "entropy": 0.9480381943285465,
      "epoch": 0.14408197767695413,
      "grad_norm": 0.16868247091770172,
      "learning_rate": 9.520041708835526e-05,
      "loss": 0.9801,
      "mean_token_accuracy": 0.7241949401795864,
      "num_tokens": 49970725.0,
      "step": 1566
    },
    {
      "entropy": 0.9170596860349178,
      "epoch": 0.14417398404839535,
      "grad_norm": 23.895511627197266,
      "learning_rate": 9.519735026221363e-05,
      "loss": 1.2885,
      "mean_token_accuracy": 0.7296508736908436,
      "num_tokens": 50001669.0,
      "step": 1567
    },
    {
      "entropy": 0.9274100437760353,
      "epoch": 0.14426599041983657,
      "grad_norm": 0.411363810300827,
      "learning_rate": 9.519428343607201e-05,
      "loss": 0.9537,
      "mean_token_accuracy": 0.7330594472587109,
      "num_tokens": 50033113.0,
      "step": 1568
    },
    {
      "entropy": 0.9212099947035313,
      "epoch": 0.1443579967912778,
      "grad_norm": 0.31847384572029114,
      "learning_rate": 9.51912166099304e-05,
      "loss": 0.9445,
      "mean_token_accuracy": 0.7430686056613922,
      "num_tokens": 50064945.0,
      "step": 1569
    },
    {
      "entropy": 0.9685227833688259,
      "epoch": 0.14445000316271903,
      "grad_norm": 0.16561812162399292,
      "learning_rate": 9.518814978378876e-05,
      "loss": 0.9431,
      "mean_token_accuracy": 0.7352004684507847,
      "num_tokens": 50097321.0,
      "step": 1570
    },
    {
      "entropy": 0.8279615417122841,
      "epoch": 0.14454200953416024,
      "grad_norm": 0.1530257761478424,
      "learning_rate": 9.518508295764713e-05,
      "loss": 0.8058,
      "mean_token_accuracy": 0.7686503194272518,
      "num_tokens": 50129738.0,
      "step": 1571
    },
    {
      "entropy": 0.9095079228281975,
      "epoch": 0.14463401590560146,
      "grad_norm": 0.15791893005371094,
      "learning_rate": 9.518201613150551e-05,
      "loss": 0.8872,
      "mean_token_accuracy": 0.7457893788814545,
      "num_tokens": 50161868.0,
      "step": 1572
    },
    {
      "entropy": 0.9339880868792534,
      "epoch": 0.14472602227704268,
      "grad_norm": 0.15989340841770172,
      "learning_rate": 9.517894930536388e-05,
      "loss": 0.9266,
      "mean_token_accuracy": 0.7377822324633598,
      "num_tokens": 50194036.0,
      "step": 1573
    },
    {
      "entropy": 0.9374928586184978,
      "epoch": 0.14481802864848392,
      "grad_norm": 0.157978355884552,
      "learning_rate": 9.517588247922226e-05,
      "loss": 0.9421,
      "mean_token_accuracy": 0.7380478791892529,
      "num_tokens": 50226786.0,
      "step": 1574
    },
    {
      "entropy": 0.9742381852120161,
      "epoch": 0.14491003501992514,
      "grad_norm": 0.16245165467262268,
      "learning_rate": 9.517281565308063e-05,
      "loss": 0.9753,
      "mean_token_accuracy": 0.7247503586113453,
      "num_tokens": 50258978.0,
      "step": 1575
    },
    {
      "entropy": 0.9068496227264404,
      "epoch": 0.14500204139136635,
      "grad_norm": 0.1590886265039444,
      "learning_rate": 9.516974882693901e-05,
      "loss": 0.9077,
      "mean_token_accuracy": 0.7483755275607109,
      "num_tokens": 50291218.0,
      "step": 1576
    },
    {
      "entropy": 0.8540623169392347,
      "epoch": 0.14509404776280757,
      "grad_norm": 0.15828856825828552,
      "learning_rate": 9.516668200079738e-05,
      "loss": 0.8409,
      "mean_token_accuracy": 0.7592857554554939,
      "num_tokens": 50322114.0,
      "step": 1577
    },
    {
      "entropy": 0.9355777502059937,
      "epoch": 0.1451860541342488,
      "grad_norm": 0.15918008983135223,
      "learning_rate": 9.516361517465575e-05,
      "loss": 0.928,
      "mean_token_accuracy": 0.7335886768996716,
      "num_tokens": 50353989.0,
      "step": 1578
    },
    {
      "entropy": 0.95644561201334,
      "epoch": 0.14527806050569003,
      "grad_norm": 0.16408760845661163,
      "learning_rate": 9.516054834851413e-05,
      "loss": 0.947,
      "mean_token_accuracy": 0.7324465103447437,
      "num_tokens": 50384827.0,
      "step": 1579
    },
    {
      "entropy": 0.8444613199681044,
      "epoch": 0.14537006687713125,
      "grad_norm": 0.15476804971694946,
      "learning_rate": 9.515748152237251e-05,
      "loss": 0.8316,
      "mean_token_accuracy": 0.7628643475472927,
      "num_tokens": 50417149.0,
      "step": 1580
    },
    {
      "entropy": 0.9362065382301807,
      "epoch": 0.14546207324857247,
      "grad_norm": 0.1665036529302597,
      "learning_rate": 9.515441469623088e-05,
      "loss": 0.9321,
      "mean_token_accuracy": 0.7332706153392792,
      "num_tokens": 50448440.0,
      "step": 1581
    },
    {
      "entropy": 0.9548250734806061,
      "epoch": 0.14555407962001368,
      "grad_norm": 0.16666781902313232,
      "learning_rate": 9.515134787008925e-05,
      "loss": 0.9711,
      "mean_token_accuracy": 0.7277650907635689,
      "num_tokens": 50479944.0,
      "step": 1582
    },
    {
      "entropy": 0.8269635140895844,
      "epoch": 0.1456460859914549,
      "grad_norm": 0.1511041820049286,
      "learning_rate": 9.514828104394761e-05,
      "loss": 0.8327,
      "mean_token_accuracy": 0.764365091919899,
      "num_tokens": 50512052.0,
      "step": 1583
    },
    {
      "entropy": 0.9229615107178688,
      "epoch": 0.14573809236289614,
      "grad_norm": 0.16137047111988068,
      "learning_rate": 9.5145214217806e-05,
      "loss": 0.9213,
      "mean_token_accuracy": 0.7383043356239796,
      "num_tokens": 50544253.0,
      "step": 1584
    },
    {
      "entropy": 0.8085582070052624,
      "epoch": 0.14583009873433736,
      "grad_norm": 0.1549147665500641,
      "learning_rate": 9.514214739166438e-05,
      "loss": 0.8169,
      "mean_token_accuracy": 0.7654806636273861,
      "num_tokens": 50576272.0,
      "step": 1585
    },
    {
      "entropy": 0.8763374015688896,
      "epoch": 0.14592210510577858,
      "grad_norm": 0.16401927173137665,
      "learning_rate": 9.513908056552274e-05,
      "loss": 0.8922,
      "mean_token_accuracy": 0.748044803738594,
      "num_tokens": 50607401.0,
      "step": 1586
    },
    {
      "entropy": 0.9314355067908764,
      "epoch": 0.1460141114772198,
      "grad_norm": 0.16168606281280518,
      "learning_rate": 9.513601373938111e-05,
      "loss": 0.9374,
      "mean_token_accuracy": 0.7302105724811554,
      "num_tokens": 50639428.0,
      "step": 1587
    },
    {
      "entropy": 0.9708931967616081,
      "epoch": 0.146106117848661,
      "grad_norm": 0.1647098958492279,
      "learning_rate": 9.513294691323949e-05,
      "loss": 0.9592,
      "mean_token_accuracy": 0.7255076915025711,
      "num_tokens": 50670938.0,
      "step": 1588
    },
    {
      "entropy": 1.0324012711644173,
      "epoch": 0.14619812422010225,
      "grad_norm": 0.1731657236814499,
      "learning_rate": 9.512988008709786e-05,
      "loss": 1.043,
      "mean_token_accuracy": 0.7068525590002537,
      "num_tokens": 50702194.0,
      "step": 1589
    },
    {
      "entropy": 0.9194666184484959,
      "epoch": 0.14629013059154347,
      "grad_norm": 0.15946748852729797,
      "learning_rate": 9.512681326095624e-05,
      "loss": 0.9289,
      "mean_token_accuracy": 0.7358950562775135,
      "num_tokens": 50734315.0,
      "step": 1590
    },
    {
      "entropy": 0.7473869267851114,
      "epoch": 0.1463821369629847,
      "grad_norm": 0.14720673859119415,
      "learning_rate": 9.512374643481462e-05,
      "loss": 0.7244,
      "mean_token_accuracy": 0.7865483909845352,
      "num_tokens": 50766102.0,
      "step": 1591
    },
    {
      "entropy": 0.8850968182086945,
      "epoch": 0.1464741433344259,
      "grad_norm": 0.15959887206554413,
      "learning_rate": 9.512067960867299e-05,
      "loss": 0.9036,
      "mean_token_accuracy": 0.7517639100551605,
      "num_tokens": 50798226.0,
      "step": 1592
    },
    {
      "entropy": 0.8950379714369774,
      "epoch": 0.14656614970586712,
      "grad_norm": 0.16577966511249542,
      "learning_rate": 9.511761278253136e-05,
      "loss": 0.9139,
      "mean_token_accuracy": 0.7478294894099236,
      "num_tokens": 50829370.0,
      "step": 1593
    },
    {
      "entropy": 0.9149220306426287,
      "epoch": 0.14665815607730837,
      "grad_norm": 0.15883855521678925,
      "learning_rate": 9.511454595638973e-05,
      "loss": 0.902,
      "mean_token_accuracy": 0.7472071722149849,
      "num_tokens": 50861552.0,
      "step": 1594
    },
    {
      "entropy": 0.9705828800797462,
      "epoch": 0.14675016244874958,
      "grad_norm": 0.15756767988204956,
      "learning_rate": 9.511147913024812e-05,
      "loss": 0.9596,
      "mean_token_accuracy": 0.7337660379707813,
      "num_tokens": 50893640.0,
      "step": 1595
    },
    {
      "entropy": 0.8951478563249111,
      "epoch": 0.1468421688201908,
      "grad_norm": 0.1565169245004654,
      "learning_rate": 9.510841230410649e-05,
      "loss": 0.8799,
      "mean_token_accuracy": 0.748848631978035,
      "num_tokens": 50925277.0,
      "step": 1596
    },
    {
      "entropy": 0.9125607758760452,
      "epoch": 0.14693417519163202,
      "grad_norm": 0.1660483330488205,
      "learning_rate": 9.510534547796486e-05,
      "loss": 0.9332,
      "mean_token_accuracy": 0.7422141805291176,
      "num_tokens": 50957177.0,
      "step": 1597
    },
    {
      "entropy": 0.9766602162271738,
      "epoch": 0.14702618156307323,
      "grad_norm": 0.16096630692481995,
      "learning_rate": 9.510227865182323e-05,
      "loss": 0.9665,
      "mean_token_accuracy": 0.7305490337312222,
      "num_tokens": 50988784.0,
      "step": 1598
    },
    {
      "entropy": 0.8283349201083183,
      "epoch": 0.14711818793451448,
      "grad_norm": 0.1592991203069687,
      "learning_rate": 9.509921182568161e-05,
      "loss": 0.8315,
      "mean_token_accuracy": 0.7597574442625046,
      "num_tokens": 51020490.0,
      "step": 1599
    },
    {
      "entropy": 0.9024342969059944,
      "epoch": 0.1472101943059557,
      "grad_norm": 0.15830719470977783,
      "learning_rate": 9.509614499953999e-05,
      "loss": 0.9086,
      "mean_token_accuracy": 0.7421135604381561,
      "num_tokens": 51052536.0,
      "step": 1600
    },
    {
      "entropy": 0.9151125848293304,
      "epoch": 0.1473022006773969,
      "grad_norm": 0.1568811982870102,
      "learning_rate": 9.509307817339836e-05,
      "loss": 0.8982,
      "mean_token_accuracy": 0.7489960603415966,
      "num_tokens": 51084474.0,
      "step": 1601
    },
    {
      "entropy": 0.9008803144097328,
      "epoch": 0.14739420704883813,
      "grad_norm": 0.15986542403697968,
      "learning_rate": 9.509001134725672e-05,
      "loss": 0.8719,
      "mean_token_accuracy": 0.7530812472105026,
      "num_tokens": 51116683.0,
      "step": 1602
    },
    {
      "entropy": 0.8322046250104904,
      "epoch": 0.14748621342027934,
      "grad_norm": 0.1528061181306839,
      "learning_rate": 9.50869445211151e-05,
      "loss": 0.8259,
      "mean_token_accuracy": 0.7639396339654922,
      "num_tokens": 51148448.0,
      "step": 1603
    },
    {
      "entropy": 0.9749828763306141,
      "epoch": 0.1475782197917206,
      "grad_norm": 0.1644148975610733,
      "learning_rate": 9.508387769497347e-05,
      "loss": 0.9968,
      "mean_token_accuracy": 0.7195788957178593,
      "num_tokens": 51180859.0,
      "step": 1604
    },
    {
      "entropy": 0.8427757658064365,
      "epoch": 0.1476702261631618,
      "grad_norm": 0.1584179699420929,
      "learning_rate": 9.508081086883186e-05,
      "loss": 0.8656,
      "mean_token_accuracy": 0.7566654123365879,
      "num_tokens": 51212962.0,
      "step": 1605
    },
    {
      "entropy": 0.9580745249986649,
      "epoch": 0.14776223253460302,
      "grad_norm": 0.17220841348171234,
      "learning_rate": 9.507774404269022e-05,
      "loss": 0.9966,
      "mean_token_accuracy": 0.720277339220047,
      "num_tokens": 51244644.0,
      "step": 1606
    },
    {
      "entropy": 0.7457785978913307,
      "epoch": 0.14785423890604424,
      "grad_norm": 0.14772135019302368,
      "learning_rate": 9.50746772165486e-05,
      "loss": 0.7376,
      "mean_token_accuracy": 0.7854622229933739,
      "num_tokens": 51276335.0,
      "step": 1607
    },
    {
      "entropy": 1.0144655648618937,
      "epoch": 0.14794624527748546,
      "grad_norm": 0.16768665611743927,
      "learning_rate": 9.507161039040697e-05,
      "loss": 1.0158,
      "mean_token_accuracy": 0.7216131202876568,
      "num_tokens": 51308674.0,
      "step": 1608
    },
    {
      "entropy": 0.8372501917183399,
      "epoch": 0.1480382516489267,
      "grad_norm": 0.17383073270320892,
      "learning_rate": 9.506854356426534e-05,
      "loss": 0.8438,
      "mean_token_accuracy": 0.757979366928339,
      "num_tokens": 51340677.0,
      "step": 1609
    },
    {
      "entropy": 0.9049365930259228,
      "epoch": 0.14813025802036792,
      "grad_norm": 0.16453367471694946,
      "learning_rate": 9.506547673812372e-05,
      "loss": 0.9123,
      "mean_token_accuracy": 0.744102381169796,
      "num_tokens": 51372231.0,
      "step": 1610
    },
    {
      "entropy": 0.8901003040373325,
      "epoch": 0.14822226439180913,
      "grad_norm": 0.15820017457008362,
      "learning_rate": 9.50624099119821e-05,
      "loss": 0.8929,
      "mean_token_accuracy": 0.7500216439366341,
      "num_tokens": 51403657.0,
      "step": 1611
    },
    {
      "entropy": 1.0162953436374664,
      "epoch": 0.14831427076325035,
      "grad_norm": 0.1622859686613083,
      "learning_rate": 9.505934308584047e-05,
      "loss": 1.0068,
      "mean_token_accuracy": 0.7181491032242775,
      "num_tokens": 51435273.0,
      "step": 1612
    },
    {
      "entropy": 0.9441129676997662,
      "epoch": 0.14840627713469157,
      "grad_norm": 0.15788370370864868,
      "learning_rate": 9.505627625969884e-05,
      "loss": 0.9267,
      "mean_token_accuracy": 0.7424830570816994,
      "num_tokens": 51467509.0,
      "step": 1613
    },
    {
      "entropy": 0.9139817301183939,
      "epoch": 0.1484982835061328,
      "grad_norm": 0.15706057846546173,
      "learning_rate": 9.505320943355721e-05,
      "loss": 0.907,
      "mean_token_accuracy": 0.7465889565646648,
      "num_tokens": 51499553.0,
      "step": 1614
    },
    {
      "entropy": 0.8888718262314796,
      "epoch": 0.14859028987757403,
      "grad_norm": 0.15246433019638062,
      "learning_rate": 9.505014260741559e-05,
      "loss": 0.8713,
      "mean_token_accuracy": 0.7511124052107334,
      "num_tokens": 51531831.0,
      "step": 1615
    },
    {
      "entropy": 0.9300766065716743,
      "epoch": 0.14868229624901524,
      "grad_norm": 0.16273950040340424,
      "learning_rate": 9.504707578127397e-05,
      "loss": 0.929,
      "mean_token_accuracy": 0.7376671209931374,
      "num_tokens": 51564065.0,
      "step": 1616
    },
    {
      "entropy": 0.8716898243874311,
      "epoch": 0.14877430262045646,
      "grad_norm": 0.15254119038581848,
      "learning_rate": 9.504400895513234e-05,
      "loss": 0.8454,
      "mean_token_accuracy": 0.7593253031373024,
      "num_tokens": 51596444.0,
      "step": 1617
    },
    {
      "entropy": 0.9185812976211309,
      "epoch": 0.14886630899189768,
      "grad_norm": 0.1590704470872879,
      "learning_rate": 9.504094212899072e-05,
      "loss": 0.9215,
      "mean_token_accuracy": 0.739941593259573,
      "num_tokens": 51628469.0,
      "step": 1618
    },
    {
      "entropy": 0.9598508812487125,
      "epoch": 0.14895831536333892,
      "grad_norm": 0.17196008563041687,
      "learning_rate": 9.503787530284909e-05,
      "loss": 0.9762,
      "mean_token_accuracy": 0.7267376519739628,
      "num_tokens": 51659274.0,
      "step": 1619
    },
    {
      "entropy": 0.8701883684843779,
      "epoch": 0.14905032173478014,
      "grad_norm": 0.16367904841899872,
      "learning_rate": 9.503480847670745e-05,
      "loss": 0.8835,
      "mean_token_accuracy": 0.7545709908008575,
      "num_tokens": 51691489.0,
      "step": 1620
    },
    {
      "entropy": 0.937255248427391,
      "epoch": 0.14914232810622136,
      "grad_norm": 0.16556379199028015,
      "learning_rate": 9.503174165056584e-05,
      "loss": 0.9446,
      "mean_token_accuracy": 0.7369784973561764,
      "num_tokens": 51722997.0,
      "step": 1621
    },
    {
      "entropy": 0.9459223821759224,
      "epoch": 0.14923433447766257,
      "grad_norm": 0.15829089283943176,
      "learning_rate": 9.502867482442422e-05,
      "loss": 0.9454,
      "mean_token_accuracy": 0.7368673421442509,
      "num_tokens": 51754866.0,
      "step": 1622
    },
    {
      "entropy": 0.8856813870370388,
      "epoch": 0.1493263408491038,
      "grad_norm": 0.15842947363853455,
      "learning_rate": 9.502560799828259e-05,
      "loss": 0.8979,
      "mean_token_accuracy": 0.7480881065130234,
      "num_tokens": 51786692.0,
      "step": 1623
    },
    {
      "entropy": 0.9174236468970776,
      "epoch": 0.14941834722054503,
      "grad_norm": 0.16238179802894592,
      "learning_rate": 9.502254117214095e-05,
      "loss": 0.9222,
      "mean_token_accuracy": 0.740988839417696,
      "num_tokens": 51819025.0,
      "step": 1624
    },
    {
      "entropy": 0.9222218878567219,
      "epoch": 0.14951035359198625,
      "grad_norm": 0.16102536022663116,
      "learning_rate": 9.501947434599932e-05,
      "loss": 0.9144,
      "mean_token_accuracy": 0.7343608476221561,
      "num_tokens": 51851341.0,
      "step": 1625
    },
    {
      "entropy": 0.9440353214740753,
      "epoch": 0.14960235996342747,
      "grad_norm": 0.16402681171894073,
      "learning_rate": 9.50164075198577e-05,
      "loss": 0.95,
      "mean_token_accuracy": 0.7281482107937336,
      "num_tokens": 51882700.0,
      "step": 1626
    },
    {
      "entropy": 0.9396484531462193,
      "epoch": 0.14969436633486868,
      "grad_norm": 0.1618112325668335,
      "learning_rate": 9.501334069371608e-05,
      "loss": 0.9507,
      "mean_token_accuracy": 0.7343941554427147,
      "num_tokens": 51915124.0,
      "step": 1627
    },
    {
      "entropy": 0.9482117705047131,
      "epoch": 0.1497863727063099,
      "grad_norm": 0.16965095698833466,
      "learning_rate": 9.501027386757445e-05,
      "loss": 0.9789,
      "mean_token_accuracy": 0.7258496694266796,
      "num_tokens": 51947089.0,
      "step": 1628
    },
    {
      "entropy": 0.9070003405213356,
      "epoch": 0.14987837907775114,
      "grad_norm": 0.15995164215564728,
      "learning_rate": 9.500720704143282e-05,
      "loss": 0.9187,
      "mean_token_accuracy": 0.7418252304196358,
      "num_tokens": 51979395.0,
      "step": 1629
    },
    {
      "entropy": 0.8971259109675884,
      "epoch": 0.14997038544919236,
      "grad_norm": 0.1557328850030899,
      "learning_rate": 9.50041402152912e-05,
      "loss": 0.9105,
      "mean_token_accuracy": 0.743866816163063,
      "num_tokens": 52011267.0,
      "step": 1630
    },
    {
      "entropy": 0.9155143424868584,
      "epoch": 0.15006239182063358,
      "grad_norm": 0.15526001155376434,
      "learning_rate": 9.500107338914957e-05,
      "loss": 0.9081,
      "mean_token_accuracy": 0.7454564422369003,
      "num_tokens": 52043454.0,
      "step": 1631
    },
    {
      "entropy": 0.8047334589064121,
      "epoch": 0.1501543981920748,
      "grad_norm": 0.15748849511146545,
      "learning_rate": 9.499800656300795e-05,
      "loss": 0.7867,
      "mean_token_accuracy": 0.778761051595211,
      "num_tokens": 52075313.0,
      "step": 1632
    },
    {
      "entropy": 1.039855144917965,
      "epoch": 0.150246404563516,
      "grad_norm": 0.16922444105148315,
      "learning_rate": 9.499493973686632e-05,
      "loss": 1.0292,
      "mean_token_accuracy": 0.7172029055655003,
      "num_tokens": 52107610.0,
      "step": 1633
    },
    {
      "entropy": 0.8415661714971066,
      "epoch": 0.15033841093495726,
      "grad_norm": 0.15803636610507965,
      "learning_rate": 9.49918729107247e-05,
      "loss": 0.8224,
      "mean_token_accuracy": 0.7655546069145203,
      "num_tokens": 52139711.0,
      "step": 1634
    },
    {
      "entropy": 0.8486452475190163,
      "epoch": 0.15043041730639847,
      "grad_norm": 0.15710069239139557,
      "learning_rate": 9.498880608458307e-05,
      "loss": 0.8437,
      "mean_token_accuracy": 0.7613230831921101,
      "num_tokens": 52171558.0,
      "step": 1635
    },
    {
      "entropy": 0.9922747574746609,
      "epoch": 0.1505224236778397,
      "grad_norm": 0.16343528032302856,
      "learning_rate": 9.498573925844144e-05,
      "loss": 1.0109,
      "mean_token_accuracy": 0.7208983823657036,
      "num_tokens": 52203794.0,
      "step": 1636
    },
    {
      "entropy": 0.9449852779507637,
      "epoch": 0.1506144300492809,
      "grad_norm": 0.1622820943593979,
      "learning_rate": 9.498267243229982e-05,
      "loss": 0.9575,
      "mean_token_accuracy": 0.7337242364883423,
      "num_tokens": 52235726.0,
      "step": 1637
    },
    {
      "entropy": 0.8581573590636253,
      "epoch": 0.15070643642072212,
      "grad_norm": 0.16376474499702454,
      "learning_rate": 9.49796056061582e-05,
      "loss": 0.8848,
      "mean_token_accuracy": 0.7479197978973389,
      "num_tokens": 52267204.0,
      "step": 1638
    },
    {
      "entropy": 0.9218185395002365,
      "epoch": 0.15079844279216337,
      "grad_norm": 0.158268004655838,
      "learning_rate": 9.497653878001657e-05,
      "loss": 0.9178,
      "mean_token_accuracy": 0.7391237653791904,
      "num_tokens": 52299185.0,
      "step": 1639
    },
    {
      "entropy": 0.8835579361766577,
      "epoch": 0.15089044916360458,
      "grad_norm": 0.15415672957897186,
      "learning_rate": 9.497347195387493e-05,
      "loss": 0.861,
      "mean_token_accuracy": 0.7504927590489388,
      "num_tokens": 52331194.0,
      "step": 1640
    },
    {
      "entropy": 0.8204936571419239,
      "epoch": 0.1509824555350458,
      "grad_norm": 0.1545938402414322,
      "learning_rate": 9.49704051277333e-05,
      "loss": 0.8242,
      "mean_token_accuracy": 0.7596048302948475,
      "num_tokens": 52363335.0,
      "step": 1641
    },
    {
      "entropy": 0.8901687487959862,
      "epoch": 0.15107446190648702,
      "grad_norm": 0.15599650144577026,
      "learning_rate": 9.49673383015917e-05,
      "loss": 0.8796,
      "mean_token_accuracy": 0.75063756108284,
      "num_tokens": 52395417.0,
      "step": 1642
    },
    {
      "entropy": 0.8809293136000633,
      "epoch": 0.15116646827792823,
      "grad_norm": 0.15762925148010254,
      "learning_rate": 9.496427147545006e-05,
      "loss": 0.8778,
      "mean_token_accuracy": 0.7492437995970249,
      "num_tokens": 52428119.0,
      "step": 1643
    },
    {
      "entropy": 0.8918488249182701,
      "epoch": 0.15125847464936948,
      "grad_norm": 0.1594894379377365,
      "learning_rate": 9.496120464930843e-05,
      "loss": 0.9085,
      "mean_token_accuracy": 0.7405115179717541,
      "num_tokens": 52459248.0,
      "step": 1644
    },
    {
      "entropy": 0.9354335218667984,
      "epoch": 0.1513504810208107,
      "grad_norm": 0.1632739007472992,
      "learning_rate": 9.49581378231668e-05,
      "loss": 0.9173,
      "mean_token_accuracy": 0.7434283494949341,
      "num_tokens": 52490965.0,
      "step": 1645
    },
    {
      "entropy": 0.9619724005460739,
      "epoch": 0.1514424873922519,
      "grad_norm": 0.165689155459404,
      "learning_rate": 9.495507099702518e-05,
      "loss": 0.9635,
      "mean_token_accuracy": 0.7333577275276184,
      "num_tokens": 52522463.0,
      "step": 1646
    },
    {
      "entropy": 0.8529836051166058,
      "epoch": 0.15153449376369313,
      "grad_norm": 0.1498270481824875,
      "learning_rate": 9.495200417088356e-05,
      "loss": 0.8224,
      "mean_token_accuracy": 0.7644657455384731,
      "num_tokens": 52554227.0,
      "step": 1647
    },
    {
      "entropy": 0.7959712594747543,
      "epoch": 0.15162650013513435,
      "grad_norm": 0.15403427183628082,
      "learning_rate": 9.494893734474193e-05,
      "loss": 0.7847,
      "mean_token_accuracy": 0.7762076631188393,
      "num_tokens": 52586658.0,
      "step": 1648
    },
    {
      "entropy": 0.8755505420267582,
      "epoch": 0.1517185065065756,
      "grad_norm": 0.17974773049354553,
      "learning_rate": 9.494587051860031e-05,
      "loss": 0.8671,
      "mean_token_accuracy": 0.7540891207754612,
      "num_tokens": 52618143.0,
      "step": 1649
    },
    {
      "entropy": 0.8522152528166771,
      "epoch": 0.1518105128780168,
      "grad_norm": 0.15996713936328888,
      "learning_rate": 9.494280369245868e-05,
      "loss": 0.8376,
      "mean_token_accuracy": 0.7645979337394238,
      "num_tokens": 52649143.0,
      "step": 1650
    },
    {
      "entropy": 0.9097826182842255,
      "epoch": 0.15190251924945802,
      "grad_norm": 0.16049756109714508,
      "learning_rate": 9.493973686631705e-05,
      "loss": 0.9008,
      "mean_token_accuracy": 0.7420790232717991,
      "num_tokens": 52681282.0,
      "step": 1651
    },
    {
      "entropy": 0.8929805681109428,
      "epoch": 0.15199452562089924,
      "grad_norm": 0.16685135662555695,
      "learning_rate": 9.493667004017543e-05,
      "loss": 0.9063,
      "mean_token_accuracy": 0.7408395633101463,
      "num_tokens": 52711970.0,
      "step": 1652
    },
    {
      "entropy": 0.9144176207482815,
      "epoch": 0.15208653199234046,
      "grad_norm": 0.1620255559682846,
      "learning_rate": 9.493360321403381e-05,
      "loss": 0.9437,
      "mean_token_accuracy": 0.7365058809518814,
      "num_tokens": 52744358.0,
      "step": 1653
    },
    {
      "entropy": 0.9382633678615093,
      "epoch": 0.1521785383637817,
      "grad_norm": 0.1609598696231842,
      "learning_rate": 9.493053638789218e-05,
      "loss": 0.9411,
      "mean_token_accuracy": 0.7369315847754478,
      "num_tokens": 52776211.0,
      "step": 1654
    },
    {
      "entropy": 0.8694864884018898,
      "epoch": 0.15227054473522292,
      "grad_norm": 0.15863408148288727,
      "learning_rate": 9.492746956175055e-05,
      "loss": 0.8794,
      "mean_token_accuracy": 0.7579607553780079,
      "num_tokens": 52808889.0,
      "step": 1655
    },
    {
      "entropy": 0.8612206224352121,
      "epoch": 0.15236255110666413,
      "grad_norm": 0.17168854176998138,
      "learning_rate": 9.492440273560891e-05,
      "loss": 0.881,
      "mean_token_accuracy": 0.7515956498682499,
      "num_tokens": 52840161.0,
      "step": 1656
    },
    {
      "entropy": 0.8907250948250294,
      "epoch": 0.15245455747810535,
      "grad_norm": 0.15931160748004913,
      "learning_rate": 9.49213359094673e-05,
      "loss": 0.8952,
      "mean_token_accuracy": 0.7495500147342682,
      "num_tokens": 52872333.0,
      "step": 1657
    },
    {
      "entropy": 0.8718052245676517,
      "epoch": 0.15254656384954657,
      "grad_norm": 0.15084831416606903,
      "learning_rate": 9.491826908332568e-05,
      "loss": 0.8428,
      "mean_token_accuracy": 0.7588514536619186,
      "num_tokens": 52905030.0,
      "step": 1658
    },
    {
      "entropy": 1.0191697403788567,
      "epoch": 0.1526385702209878,
      "grad_norm": 0.16582950949668884,
      "learning_rate": 9.491520225718405e-05,
      "loss": 0.9942,
      "mean_token_accuracy": 0.7198337092995644,
      "num_tokens": 52936460.0,
      "step": 1659
    },
    {
      "entropy": 0.7939396910369396,
      "epoch": 0.15273057659242903,
      "grad_norm": 0.149077907204628,
      "learning_rate": 9.491213543104241e-05,
      "loss": 0.7821,
      "mean_token_accuracy": 0.7755911909043789,
      "num_tokens": 52968539.0,
      "step": 1660
    },
    {
      "entropy": 0.8676322288811207,
      "epoch": 0.15282258296387025,
      "grad_norm": 0.15120075643062592,
      "learning_rate": 9.49090686049008e-05,
      "loss": 0.8448,
      "mean_token_accuracy": 0.7550942711532116,
      "num_tokens": 53000579.0,
      "step": 1661
    },
    {
      "entropy": 0.8399743940681219,
      "epoch": 0.15291458933531146,
      "grad_norm": 0.15518411993980408,
      "learning_rate": 9.490600177875916e-05,
      "loss": 0.8408,
      "mean_token_accuracy": 0.7580276504158974,
      "num_tokens": 53031962.0,
      "step": 1662
    },
    {
      "entropy": 0.8637074064463377,
      "epoch": 0.15300659570675268,
      "grad_norm": 0.15760450065135956,
      "learning_rate": 9.490293495261754e-05,
      "loss": 0.8502,
      "mean_token_accuracy": 0.7583245672285557,
      "num_tokens": 53062831.0,
      "step": 1663
    },
    {
      "entropy": 0.8890084810554981,
      "epoch": 0.15309860207819392,
      "grad_norm": 0.161193385720253,
      "learning_rate": 9.489986812647591e-05,
      "loss": 0.9005,
      "mean_token_accuracy": 0.7455227673053741,
      "num_tokens": 53094330.0,
      "step": 1664
    },
    {
      "entropy": 0.9590332098305225,
      "epoch": 0.15319060844963514,
      "grad_norm": 0.1679432988166809,
      "learning_rate": 9.48968013003343e-05,
      "loss": 0.975,
      "mean_token_accuracy": 0.7301226817071438,
      "num_tokens": 53126446.0,
      "step": 1665
    },
    {
      "entropy": 0.8744850568473339,
      "epoch": 0.15328261482107636,
      "grad_norm": 0.15566346049308777,
      "learning_rate": 9.489373447419266e-05,
      "loss": 0.8762,
      "mean_token_accuracy": 0.7518120110034943,
      "num_tokens": 53158252.0,
      "step": 1666
    },
    {
      "entropy": 0.8664066232740879,
      "epoch": 0.15337462119251757,
      "grad_norm": 0.15962710976600647,
      "learning_rate": 9.489066764805103e-05,
      "loss": 0.9012,
      "mean_token_accuracy": 0.7430575080215931,
      "num_tokens": 53190677.0,
      "step": 1667
    },
    {
      "entropy": 0.8496988452970982,
      "epoch": 0.1534666275639588,
      "grad_norm": 0.16696470975875854,
      "learning_rate": 9.488760082190941e-05,
      "loss": 0.8883,
      "mean_token_accuracy": 0.7505216896533966,
      "num_tokens": 53222384.0,
      "step": 1668
    },
    {
      "entropy": 0.9118777886033058,
      "epoch": 0.15355863393540004,
      "grad_norm": 0.16979368031024933,
      "learning_rate": 9.488453399576779e-05,
      "loss": 0.9126,
      "mean_token_accuracy": 0.74341681599617,
      "num_tokens": 53253638.0,
      "step": 1669
    },
    {
      "entropy": 0.9640098884701729,
      "epoch": 0.15365064030684125,
      "grad_norm": 0.16880130767822266,
      "learning_rate": 9.488146716962616e-05,
      "loss": 0.9791,
      "mean_token_accuracy": 0.7237409800291061,
      "num_tokens": 53285827.0,
      "step": 1670
    },
    {
      "entropy": 0.9875710643827915,
      "epoch": 0.15374264667828247,
      "grad_norm": 0.16397635638713837,
      "learning_rate": 9.487840034348453e-05,
      "loss": 0.9531,
      "mean_token_accuracy": 0.7307684496045113,
      "num_tokens": 53317059.0,
      "step": 1671
    },
    {
      "entropy": 0.8687214441597462,
      "epoch": 0.15383465304972369,
      "grad_norm": 0.16013942658901215,
      "learning_rate": 9.48753335173429e-05,
      "loss": 0.8639,
      "mean_token_accuracy": 0.7535696551203728,
      "num_tokens": 53349438.0,
      "step": 1672
    },
    {
      "entropy": 0.8929279148578644,
      "epoch": 0.1539266594211649,
      "grad_norm": 0.15438346564769745,
      "learning_rate": 9.487226669120128e-05,
      "loss": 0.8748,
      "mean_token_accuracy": 0.7507476694881916,
      "num_tokens": 53380945.0,
      "step": 1673
    },
    {
      "entropy": 0.9531895443797112,
      "epoch": 0.15401866579260615,
      "grad_norm": 0.16354206204414368,
      "learning_rate": 9.486919986505966e-05,
      "loss": 0.9536,
      "mean_token_accuracy": 0.7283436134457588,
      "num_tokens": 53413238.0,
      "step": 1674
    },
    {
      "entropy": 0.9441161006689072,
      "epoch": 0.15411067216404736,
      "grad_norm": 0.15975528955459595,
      "learning_rate": 9.486613303891803e-05,
      "loss": 0.9442,
      "mean_token_accuracy": 0.7342549152672291,
      "num_tokens": 53445633.0,
      "step": 1675
    },
    {
      "entropy": 0.9476718939840794,
      "epoch": 0.15420267853548858,
      "grad_norm": 0.16097865998744965,
      "learning_rate": 9.48630662127764e-05,
      "loss": 0.9558,
      "mean_token_accuracy": 0.7322823256254196,
      "num_tokens": 53477696.0,
      "step": 1676
    },
    {
      "entropy": 0.9025196582078934,
      "epoch": 0.1542946849069298,
      "grad_norm": 0.16239745914936066,
      "learning_rate": 9.485999938663478e-05,
      "loss": 0.9005,
      "mean_token_accuracy": 0.7389497086405754,
      "num_tokens": 53509230.0,
      "step": 1677
    },
    {
      "entropy": 0.8449400253593922,
      "epoch": 0.15438669127837104,
      "grad_norm": 0.15214960277080536,
      "learning_rate": 9.485693256049314e-05,
      "loss": 0.8461,
      "mean_token_accuracy": 0.7549704201519489,
      "num_tokens": 53541527.0,
      "step": 1678
    },
    {
      "entropy": 0.9128388278186321,
      "epoch": 0.15447869764981226,
      "grad_norm": 0.14962369203567505,
      "learning_rate": 9.485386573435152e-05,
      "loss": 0.8979,
      "mean_token_accuracy": 0.7470507025718689,
      "num_tokens": 53573844.0,
      "step": 1679
    },
    {
      "entropy": 0.9056573286652565,
      "epoch": 0.15457070402125347,
      "grad_norm": 0.1597958356142044,
      "learning_rate": 9.48507989082099e-05,
      "loss": 0.8967,
      "mean_token_accuracy": 0.7463571988046169,
      "num_tokens": 53605904.0,
      "step": 1680
    },
    {
      "entropy": 0.8708511292934418,
      "epoch": 0.1546627103926947,
      "grad_norm": 0.15483857691287994,
      "learning_rate": 9.484773208206827e-05,
      "loss": 0.8749,
      "mean_token_accuracy": 0.7523781917989254,
      "num_tokens": 53638185.0,
      "step": 1681
    },
    {
      "entropy": 0.917914479970932,
      "epoch": 0.1547547167641359,
      "grad_norm": 0.15941733121871948,
      "learning_rate": 9.484466525592664e-05,
      "loss": 0.9064,
      "mean_token_accuracy": 0.7437374778091908,
      "num_tokens": 53670507.0,
      "step": 1682
    },
    {
      "entropy": 0.886337973177433,
      "epoch": 0.15484672313557715,
      "grad_norm": 0.15965482592582703,
      "learning_rate": 9.484159842978502e-05,
      "loss": 0.8893,
      "mean_token_accuracy": 0.7522208616137505,
      "num_tokens": 53702779.0,
      "step": 1683
    },
    {
      "entropy": 0.8622843474149704,
      "epoch": 0.15493872950701837,
      "grad_norm": 0.15925782918930054,
      "learning_rate": 9.48385316036434e-05,
      "loss": 0.8518,
      "mean_token_accuracy": 0.7463275194168091,
      "num_tokens": 53733993.0,
      "step": 1684
    },
    {
      "entropy": 0.823455385863781,
      "epoch": 0.15503073587845959,
      "grad_norm": 0.15244053304195404,
      "learning_rate": 9.483546477750177e-05,
      "loss": 0.8186,
      "mean_token_accuracy": 0.767518799751997,
      "num_tokens": 53766076.0,
      "step": 1685
    },
    {
      "entropy": 0.8463437557220459,
      "epoch": 0.1551227422499008,
      "grad_norm": 0.15847264230251312,
      "learning_rate": 9.483239795136014e-05,
      "loss": 0.8388,
      "mean_token_accuracy": 0.7582208067178726,
      "num_tokens": 53798192.0,
      "step": 1686
    },
    {
      "entropy": 0.902086790651083,
      "epoch": 0.15521474862134202,
      "grad_norm": 0.16070100665092468,
      "learning_rate": 9.482933112521851e-05,
      "loss": 0.9254,
      "mean_token_accuracy": 0.7412216104567051,
      "num_tokens": 53830249.0,
      "step": 1687
    },
    {
      "entropy": 0.9523467421531677,
      "epoch": 0.15530675499278326,
      "grad_norm": 0.15972301363945007,
      "learning_rate": 9.482626429907689e-05,
      "loss": 0.9442,
      "mean_token_accuracy": 0.7311886921525002,
      "num_tokens": 53862059.0,
      "step": 1688
    },
    {
      "entropy": 0.8706722110509872,
      "epoch": 0.15539876136422448,
      "grad_norm": 0.16081814467906952,
      "learning_rate": 9.482319747293527e-05,
      "loss": 0.8745,
      "mean_token_accuracy": 0.7482874393463135,
      "num_tokens": 53894355.0,
      "step": 1689
    },
    {
      "entropy": 0.7717008851468563,
      "epoch": 0.1554907677356657,
      "grad_norm": 0.15584036707878113,
      "learning_rate": 9.482013064679364e-05,
      "loss": 0.7715,
      "mean_token_accuracy": 0.7790940143167973,
      "num_tokens": 53926166.0,
      "step": 1690
    },
    {
      "entropy": 0.862653486430645,
      "epoch": 0.1555827741071069,
      "grad_norm": 0.16037125885486603,
      "learning_rate": 9.481706382065201e-05,
      "loss": 0.8682,
      "mean_token_accuracy": 0.7545990012586117,
      "num_tokens": 53957634.0,
      "step": 1691
    },
    {
      "entropy": 0.8897218629717827,
      "epoch": 0.15567478047854813,
      "grad_norm": 0.1623382568359375,
      "learning_rate": 9.481399699451039e-05,
      "loss": 0.902,
      "mean_token_accuracy": 0.7461304813623428,
      "num_tokens": 53989190.0,
      "step": 1692
    },
    {
      "entropy": 0.9146887063980103,
      "epoch": 0.15576678684998937,
      "grad_norm": 0.1723165512084961,
      "learning_rate": 9.481093016836876e-05,
      "loss": 0.9181,
      "mean_token_accuracy": 0.7455864064395428,
      "num_tokens": 54021257.0,
      "step": 1693
    },
    {
      "entropy": 0.7838038373738527,
      "epoch": 0.1558587932214306,
      "grad_norm": 0.16362956166267395,
      "learning_rate": 9.480786334222714e-05,
      "loss": 0.806,
      "mean_token_accuracy": 0.7706562504172325,
      "num_tokens": 54052555.0,
      "step": 1694
    },
    {
      "entropy": 0.9310091473162174,
      "epoch": 0.1559507995928718,
      "grad_norm": 0.16300317645072937,
      "learning_rate": 9.48047965160855e-05,
      "loss": 0.9447,
      "mean_token_accuracy": 0.7334553487598896,
      "num_tokens": 54084668.0,
      "step": 1695
    },
    {
      "entropy": 0.8971166908740997,
      "epoch": 0.15604280596431302,
      "grad_norm": 0.15724237263202667,
      "learning_rate": 9.480172968994389e-05,
      "loss": 0.8885,
      "mean_token_accuracy": 0.7470222786068916,
      "num_tokens": 54115968.0,
      "step": 1696
    },
    {
      "entropy": 0.8418847545981407,
      "epoch": 0.15613481233575424,
      "grad_norm": 0.15366658568382263,
      "learning_rate": 9.479866286380226e-05,
      "loss": 0.8582,
      "mean_token_accuracy": 0.7534540891647339,
      "num_tokens": 54147725.0,
      "step": 1697
    },
    {
      "entropy": 0.811025470495224,
      "epoch": 0.1562268187071955,
      "grad_norm": 0.14820407330989838,
      "learning_rate": 9.479559603766062e-05,
      "loss": 0.7872,
      "mean_token_accuracy": 0.7764305435121059,
      "num_tokens": 54179885.0,
      "step": 1698
    },
    {
      "entropy": 0.9349726289510727,
      "epoch": 0.1563188250786367,
      "grad_norm": 0.15831083059310913,
      "learning_rate": 9.4792529211519e-05,
      "loss": 0.9219,
      "mean_token_accuracy": 0.7413631565868855,
      "num_tokens": 54212250.0,
      "step": 1699
    },
    {
      "entropy": 0.9500974379479885,
      "epoch": 0.15641083145007792,
      "grad_norm": 0.16674832999706268,
      "learning_rate": 9.478946238537739e-05,
      "loss": 0.9505,
      "mean_token_accuracy": 0.734423391520977,
      "num_tokens": 54243528.0,
      "step": 1700
    },
    {
      "entropy": 0.9455444179475307,
      "epoch": 0.15650283782151914,
      "grad_norm": 0.17008273303508759,
      "learning_rate": 9.478639555923575e-05,
      "loss": 0.9212,
      "mean_token_accuracy": 0.7416010238230228,
      "num_tokens": 54274752.0,
      "step": 1701
    },
    {
      "entropy": 0.9908579103648663,
      "epoch": 0.15659484419296035,
      "grad_norm": 0.16920419037342072,
      "learning_rate": 9.478332873309412e-05,
      "loss": 0.9836,
      "mean_token_accuracy": 0.7259894125163555,
      "num_tokens": 54306892.0,
      "step": 1702
    },
    {
      "entropy": 0.9702217988669872,
      "epoch": 0.1566868505644016,
      "grad_norm": 0.16528670489788055,
      "learning_rate": 9.478026190695249e-05,
      "loss": 0.9837,
      "mean_token_accuracy": 0.7259367853403091,
      "num_tokens": 54339107.0,
      "step": 1703
    },
    {
      "entropy": 0.9183321706950665,
      "epoch": 0.15677885693584281,
      "grad_norm": 0.16281481087207794,
      "learning_rate": 9.477719508081087e-05,
      "loss": 0.9213,
      "mean_token_accuracy": 0.7377865985035896,
      "num_tokens": 54370647.0,
      "step": 1704
    },
    {
      "entropy": 0.9342836998403072,
      "epoch": 0.15687086330728403,
      "grad_norm": 0.15916863083839417,
      "learning_rate": 9.477412825466925e-05,
      "loss": 0.9351,
      "mean_token_accuracy": 0.7422629743814468,
      "num_tokens": 54402832.0,
      "step": 1705
    },
    {
      "entropy": 0.9554618373513222,
      "epoch": 0.15696286967872525,
      "grad_norm": 0.15732289850711823,
      "learning_rate": 9.477106142852762e-05,
      "loss": 0.9386,
      "mean_token_accuracy": 0.7365233488380909,
      "num_tokens": 54435062.0,
      "step": 1706
    },
    {
      "entropy": 0.8860854264348745,
      "epoch": 0.15705487605016646,
      "grad_norm": 0.1528858095407486,
      "learning_rate": 9.476799460238599e-05,
      "loss": 0.8731,
      "mean_token_accuracy": 0.7499328218400478,
      "num_tokens": 54467253.0,
      "step": 1707
    },
    {
      "entropy": 0.8063505329191685,
      "epoch": 0.1571468824216077,
      "grad_norm": 0.15440621972084045,
      "learning_rate": 9.476492777624437e-05,
      "loss": 0.7931,
      "mean_token_accuracy": 0.7704049423336983,
      "num_tokens": 54499154.0,
      "step": 1708
    },
    {
      "entropy": 0.8728880137205124,
      "epoch": 0.15723888879304893,
      "grad_norm": 0.15466676652431488,
      "learning_rate": 9.476186095010274e-05,
      "loss": 0.8648,
      "mean_token_accuracy": 0.7545252218842506,
      "num_tokens": 54531459.0,
      "step": 1709
    },
    {
      "entropy": 0.9096110761165619,
      "epoch": 0.15733089516449014,
      "grad_norm": 0.16493366658687592,
      "learning_rate": 9.475879412396112e-05,
      "loss": 0.9189,
      "mean_token_accuracy": 0.7390969879925251,
      "num_tokens": 54563307.0,
      "step": 1710
    },
    {
      "entropy": 0.987253475934267,
      "epoch": 0.15742290153593136,
      "grad_norm": 0.16845540702342987,
      "learning_rate": 9.47557272978195e-05,
      "loss": 0.9979,
      "mean_token_accuracy": 0.7209025919437408,
      "num_tokens": 54595452.0,
      "step": 1711
    },
    {
      "entropy": 0.8169391378760338,
      "epoch": 0.15751490790737258,
      "grad_norm": 0.1545519232749939,
      "learning_rate": 9.475266047167787e-05,
      "loss": 0.8119,
      "mean_token_accuracy": 0.7663094475865364,
      "num_tokens": 54627370.0,
      "step": 1712
    },
    {
      "entropy": 0.9229889288544655,
      "epoch": 0.15760691427881382,
      "grad_norm": 0.16811498999595642,
      "learning_rate": 9.474959364553624e-05,
      "loss": 0.9486,
      "mean_token_accuracy": 0.7300241254270077,
      "num_tokens": 54658659.0,
      "step": 1713
    },
    {
      "entropy": 0.8482905868440866,
      "epoch": 0.15769892065025504,
      "grad_norm": 0.15875771641731262,
      "learning_rate": 9.47465268193946e-05,
      "loss": 0.872,
      "mean_token_accuracy": 0.7514349035918713,
      "num_tokens": 54691151.0,
      "step": 1714
    },
    {
      "entropy": 0.8890507891774178,
      "epoch": 0.15779092702169625,
      "grad_norm": 0.15801656246185303,
      "learning_rate": 9.4743459993253e-05,
      "loss": 0.8864,
      "mean_token_accuracy": 0.7439468502998352,
      "num_tokens": 54723719.0,
      "step": 1715
    },
    {
      "entropy": 0.8801919780671597,
      "epoch": 0.15788293339313747,
      "grad_norm": 0.160419762134552,
      "learning_rate": 9.474039316711137e-05,
      "loss": 0.8898,
      "mean_token_accuracy": 0.7471887245774269,
      "num_tokens": 54755920.0,
      "step": 1716
    },
    {
      "entropy": 0.9148578569293022,
      "epoch": 0.1579749397645787,
      "grad_norm": 0.16441474854946136,
      "learning_rate": 9.473732634096973e-05,
      "loss": 0.9181,
      "mean_token_accuracy": 0.7420166656374931,
      "num_tokens": 54787052.0,
      "step": 1717
    },
    {
      "entropy": 0.8919218704104424,
      "epoch": 0.15806694613601993,
      "grad_norm": 0.15505415201187134,
      "learning_rate": 9.47342595148281e-05,
      "loss": 0.9049,
      "mean_token_accuracy": 0.7442884705960751,
      "num_tokens": 54819413.0,
      "step": 1718
    },
    {
      "entropy": 0.9873096197843552,
      "epoch": 0.15815895250746115,
      "grad_norm": 0.16581912338733673,
      "learning_rate": 9.473119268868648e-05,
      "loss": 0.9712,
      "mean_token_accuracy": 0.7265113368630409,
      "num_tokens": 54851101.0,
      "step": 1719
    },
    {
      "entropy": 0.8947155028581619,
      "epoch": 0.15825095887890236,
      "grad_norm": 0.15772901475429535,
      "learning_rate": 9.472812586254487e-05,
      "loss": 0.8806,
      "mean_token_accuracy": 0.7487349845468998,
      "num_tokens": 54883466.0,
      "step": 1720
    },
    {
      "entropy": 0.7966335341334343,
      "epoch": 0.15834296525034358,
      "grad_norm": 0.15847474336624146,
      "learning_rate": 9.472505903640323e-05,
      "loss": 0.7927,
      "mean_token_accuracy": 0.772607509046793,
      "num_tokens": 54916061.0,
      "step": 1721
    },
    {
      "entropy": 0.9063409864902496,
      "epoch": 0.1584349716217848,
      "grad_norm": 0.16203106939792633,
      "learning_rate": 9.47219922102616e-05,
      "loss": 0.8889,
      "mean_token_accuracy": 0.7440407648682594,
      "num_tokens": 54948000.0,
      "step": 1722
    },
    {
      "entropy": 0.8444563671946526,
      "epoch": 0.15852697799322604,
      "grad_norm": 0.15534907579421997,
      "learning_rate": 9.471892538411998e-05,
      "loss": 0.823,
      "mean_token_accuracy": 0.7644043453037739,
      "num_tokens": 54979550.0,
      "step": 1723
    },
    {
      "entropy": 0.9037407338619232,
      "epoch": 0.15861898436466726,
      "grad_norm": 0.15733015537261963,
      "learning_rate": 9.471585855797835e-05,
      "loss": 0.8888,
      "mean_token_accuracy": 0.7466606646776199,
      "num_tokens": 55011393.0,
      "step": 1724
    },
    {
      "entropy": 0.8407981544733047,
      "epoch": 0.15871099073610848,
      "grad_norm": 0.15854699909687042,
      "learning_rate": 9.471279173183673e-05,
      "loss": 0.8648,
      "mean_token_accuracy": 0.7527385503053665,
      "num_tokens": 55042793.0,
      "step": 1725
    },
    {
      "entropy": 0.8450030721724033,
      "epoch": 0.1588029971075497,
      "grad_norm": 0.1563250720500946,
      "learning_rate": 9.47097249056951e-05,
      "loss": 0.8262,
      "mean_token_accuracy": 0.7662789635360241,
      "num_tokens": 55075019.0,
      "step": 1726
    },
    {
      "entropy": 0.8880973272025585,
      "epoch": 0.1588950034789909,
      "grad_norm": 0.1669611930847168,
      "learning_rate": 9.470665807955348e-05,
      "loss": 0.9191,
      "mean_token_accuracy": 0.7397681996226311,
      "num_tokens": 55106710.0,
      "step": 1727
    },
    {
      "entropy": 0.8997660670429468,
      "epoch": 0.15898700985043215,
      "grad_norm": 0.16458453238010406,
      "learning_rate": 9.470359125341185e-05,
      "loss": 0.9074,
      "mean_token_accuracy": 0.7468083016574383,
      "num_tokens": 55138955.0,
      "step": 1728
    },
    {
      "entropy": 0.868724636733532,
      "epoch": 0.15907901622187337,
      "grad_norm": 0.15582101047039032,
      "learning_rate": 9.470052442727022e-05,
      "loss": 0.8533,
      "mean_token_accuracy": 0.7572338879108429,
      "num_tokens": 55170623.0,
      "step": 1729
    },
    {
      "entropy": 0.8749800883233547,
      "epoch": 0.1591710225933146,
      "grad_norm": 0.1657828688621521,
      "learning_rate": 9.46974576011286e-05,
      "loss": 0.8639,
      "mean_token_accuracy": 0.752534557133913,
      "num_tokens": 55202591.0,
      "step": 1730
    },
    {
      "entropy": 0.9054227471351624,
      "epoch": 0.1592630289647558,
      "grad_norm": 0.15695258975028992,
      "learning_rate": 9.469439077498698e-05,
      "loss": 0.8784,
      "mean_token_accuracy": 0.74924461171031,
      "num_tokens": 55233868.0,
      "step": 1731
    },
    {
      "entropy": 0.9592415764927864,
      "epoch": 0.15935503533619702,
      "grad_norm": 0.16555941104888916,
      "learning_rate": 9.469132394884535e-05,
      "loss": 0.9454,
      "mean_token_accuracy": 0.7300211787223816,
      "num_tokens": 55265919.0,
      "step": 1732
    },
    {
      "entropy": 1.0271290801465511,
      "epoch": 0.15944704170763827,
      "grad_norm": 0.17038317024707794,
      "learning_rate": 9.468825712270372e-05,
      "loss": 1.0137,
      "mean_token_accuracy": 0.7169928103685379,
      "num_tokens": 55298003.0,
      "step": 1733
    },
    {
      "entropy": 0.9904250241816044,
      "epoch": 0.15953904807907948,
      "grad_norm": 0.1647581309080124,
      "learning_rate": 9.468519029656208e-05,
      "loss": 0.994,
      "mean_token_accuracy": 0.7199724428355694,
      "num_tokens": 55329836.0,
      "step": 1734
    },
    {
      "entropy": 0.8312585912644863,
      "epoch": 0.1596310544505207,
      "grad_norm": 0.15537254512310028,
      "learning_rate": 9.468212347042046e-05,
      "loss": 0.8397,
      "mean_token_accuracy": 0.7610099799931049,
      "num_tokens": 55362272.0,
      "step": 1735
    },
    {
      "entropy": 0.7786161825060844,
      "epoch": 0.15972306082196192,
      "grad_norm": 0.15685470402240753,
      "learning_rate": 9.467905664427885e-05,
      "loss": 0.7759,
      "mean_token_accuracy": 0.773269783705473,
      "num_tokens": 55393817.0,
      "step": 1736
    },
    {
      "entropy": 0.8771466538310051,
      "epoch": 0.15981506719340313,
      "grad_norm": 0.16725194454193115,
      "learning_rate": 9.467598981813721e-05,
      "loss": 0.8991,
      "mean_token_accuracy": 0.7454415708780289,
      "num_tokens": 55425499.0,
      "step": 1737
    },
    {
      "entropy": 0.7983042448759079,
      "epoch": 0.15990707356484438,
      "grad_norm": 0.159865602850914,
      "learning_rate": 9.467292299199558e-05,
      "loss": 0.8273,
      "mean_token_accuracy": 0.7678916156291962,
      "num_tokens": 55457699.0,
      "step": 1738
    },
    {
      "entropy": 0.865582350641489,
      "epoch": 0.1599990799362856,
      "grad_norm": 0.16646823287010193,
      "learning_rate": 9.466985616585396e-05,
      "loss": 0.8879,
      "mean_token_accuracy": 0.7530284151434898,
      "num_tokens": 55490225.0,
      "step": 1739
    },
    {
      "entropy": 0.9023640714585781,
      "epoch": 0.1600910863077268,
      "grad_norm": 0.16383902728557587,
      "learning_rate": 9.466678933971233e-05,
      "loss": 0.9048,
      "mean_token_accuracy": 0.7460097074508667,
      "num_tokens": 55522774.0,
      "step": 1740
    },
    {
      "entropy": 0.8834324218332767,
      "epoch": 0.16018309267916803,
      "grad_norm": 0.15716809034347534,
      "learning_rate": 9.466372251357071e-05,
      "loss": 0.8841,
      "mean_token_accuracy": 0.7473073527216911,
      "num_tokens": 55554842.0,
      "step": 1741
    },
    {
      "entropy": 0.8234524670988321,
      "epoch": 0.16027509905060924,
      "grad_norm": 0.15976832807064056,
      "learning_rate": 9.46606556874291e-05,
      "loss": 0.811,
      "mean_token_accuracy": 0.7651534415781498,
      "num_tokens": 55585496.0,
      "step": 1742
    },
    {
      "entropy": 0.9703383073210716,
      "epoch": 0.1603671054220505,
      "grad_norm": 0.16275815665721893,
      "learning_rate": 9.465758886128746e-05,
      "loss": 0.9701,
      "mean_token_accuracy": 0.7283554673194885,
      "num_tokens": 55616925.0,
      "step": 1743
    },
    {
      "entropy": 0.9022040776908398,
      "epoch": 0.1604591117934917,
      "grad_norm": 0.1629457175731659,
      "learning_rate": 9.465452203514583e-05,
      "loss": 0.9065,
      "mean_token_accuracy": 0.7455358617007732,
      "num_tokens": 55647958.0,
      "step": 1744
    },
    {
      "entropy": 0.9282336980104446,
      "epoch": 0.16055111816493292,
      "grad_norm": 0.16090373694896698,
      "learning_rate": 9.46514552090042e-05,
      "loss": 0.8978,
      "mean_token_accuracy": 0.7430650927126408,
      "num_tokens": 55679420.0,
      "step": 1745
    },
    {
      "entropy": 0.9594156555831432,
      "epoch": 0.16064312453637414,
      "grad_norm": 0.16427485644817352,
      "learning_rate": 9.464838838286258e-05,
      "loss": 0.9474,
      "mean_token_accuracy": 0.73216113448143,
      "num_tokens": 55711077.0,
      "step": 1746
    },
    {
      "entropy": 0.9214438199996948,
      "epoch": 0.16073513090781535,
      "grad_norm": 0.1624341905117035,
      "learning_rate": 9.464532155672096e-05,
      "loss": 0.9244,
      "mean_token_accuracy": 0.7433156706392765,
      "num_tokens": 55742802.0,
      "step": 1747
    },
    {
      "entropy": 0.844107486307621,
      "epoch": 0.1608271372792566,
      "grad_norm": 0.15857581794261932,
      "learning_rate": 9.464225473057933e-05,
      "loss": 0.8434,
      "mean_token_accuracy": 0.7593778781592846,
      "num_tokens": 55775387.0,
      "step": 1748
    },
    {
      "entropy": 0.8997818976640701,
      "epoch": 0.16091914365069782,
      "grad_norm": 0.15934298932552338,
      "learning_rate": 9.46391879044377e-05,
      "loss": 0.9087,
      "mean_token_accuracy": 0.7421522997319698,
      "num_tokens": 55807765.0,
      "step": 1749
    },
    {
      "entropy": 0.8983301296830177,
      "epoch": 0.16101115002213903,
      "grad_norm": 0.1656133532524109,
      "learning_rate": 9.463612107829608e-05,
      "loss": 0.9159,
      "mean_token_accuracy": 0.7403369024395943,
      "num_tokens": 55839791.0,
      "step": 1750
    },
    {
      "entropy": 0.9200187772512436,
      "epoch": 0.16110315639358025,
      "grad_norm": 0.16497880220413208,
      "learning_rate": 9.463305425215445e-05,
      "loss": 0.929,
      "mean_token_accuracy": 0.7368572689592838,
      "num_tokens": 55870160.0,
      "step": 1751
    },
    {
      "entropy": 0.8285903967916965,
      "epoch": 0.16119516276502147,
      "grad_norm": 0.16666880249977112,
      "learning_rate": 9.462998742601283e-05,
      "loss": 0.8521,
      "mean_token_accuracy": 0.7571747116744518,
      "num_tokens": 55901909.0,
      "step": 1752
    },
    {
      "entropy": 1.012224555015564,
      "epoch": 0.1612871691364627,
      "grad_norm": 0.1608617752790451,
      "learning_rate": 9.46269205998712e-05,
      "loss": 1.0228,
      "mean_token_accuracy": 0.7143578082323074,
      "num_tokens": 55933919.0,
      "step": 1753
    },
    {
      "entropy": 0.994809728115797,
      "epoch": 0.16137917550790393,
      "grad_norm": 0.1732369363307953,
      "learning_rate": 9.462385377372958e-05,
      "loss": 1.0156,
      "mean_token_accuracy": 0.7177874594926834,
      "num_tokens": 55966220.0,
      "step": 1754
    },
    {
      "entropy": 0.9800427034497261,
      "epoch": 0.16147118187934514,
      "grad_norm": 0.16336874663829803,
      "learning_rate": 9.462078694758794e-05,
      "loss": 0.9661,
      "mean_token_accuracy": 0.7283734194934368,
      "num_tokens": 55998591.0,
      "step": 1755
    },
    {
      "entropy": 0.9412521384656429,
      "epoch": 0.16156318825078636,
      "grad_norm": 0.1608155518770218,
      "learning_rate": 9.461772012144631e-05,
      "loss": 0.9115,
      "mean_token_accuracy": 0.7445107027888298,
      "num_tokens": 56030939.0,
      "step": 1756
    },
    {
      "entropy": 0.8990996517241001,
      "epoch": 0.16165519462222758,
      "grad_norm": 0.15456418693065643,
      "learning_rate": 9.46146532953047e-05,
      "loss": 0.8781,
      "mean_token_accuracy": 0.7530279569327831,
      "num_tokens": 56062781.0,
      "step": 1757
    },
    {
      "entropy": 0.9321890398859978,
      "epoch": 0.16174720099366882,
      "grad_norm": 0.1617853045463562,
      "learning_rate": 9.461158646916307e-05,
      "loss": 0.9149,
      "mean_token_accuracy": 0.7408480830490589,
      "num_tokens": 56095305.0,
      "step": 1758
    },
    {
      "entropy": 0.8277414031326771,
      "epoch": 0.16183920736511004,
      "grad_norm": 0.1555449366569519,
      "learning_rate": 9.460851964302144e-05,
      "loss": 0.8245,
      "mean_token_accuracy": 0.764634445309639,
      "num_tokens": 56127546.0,
      "step": 1759
    },
    {
      "entropy": 0.9614381939172745,
      "epoch": 0.16193121373655125,
      "grad_norm": 0.16077855229377747,
      "learning_rate": 9.460545281687981e-05,
      "loss": 0.9522,
      "mean_token_accuracy": 0.7331216596066952,
      "num_tokens": 56159622.0,
      "step": 1760
    },
    {
      "entropy": 0.9039736315608025,
      "epoch": 0.16202322010799247,
      "grad_norm": 0.160927414894104,
      "learning_rate": 9.460238599073818e-05,
      "loss": 0.9219,
      "mean_token_accuracy": 0.7412565797567368,
      "num_tokens": 56191827.0,
      "step": 1761
    },
    {
      "entropy": 0.9086353853344917,
      "epoch": 0.1621152264794337,
      "grad_norm": 0.157014399766922,
      "learning_rate": 9.459931916459657e-05,
      "loss": 0.9025,
      "mean_token_accuracy": 0.7405021451413631,
      "num_tokens": 56223630.0,
      "step": 1762
    },
    {
      "entropy": 0.8508599549531937,
      "epoch": 0.16220723285087493,
      "grad_norm": 0.15543513000011444,
      "learning_rate": 9.459625233845494e-05,
      "loss": 0.8677,
      "mean_token_accuracy": 0.7482740134000778,
      "num_tokens": 56255018.0,
      "step": 1763
    },
    {
      "entropy": 0.8361765667796135,
      "epoch": 0.16229923922231615,
      "grad_norm": 0.15574875473976135,
      "learning_rate": 9.459318551231331e-05,
      "loss": 0.8272,
      "mean_token_accuracy": 0.7634220272302628,
      "num_tokens": 56286744.0,
      "step": 1764
    },
    {
      "entropy": 0.861007260158658,
      "epoch": 0.16239124559375737,
      "grad_norm": 0.1528848260641098,
      "learning_rate": 9.459011868617168e-05,
      "loss": 0.8596,
      "mean_token_accuracy": 0.7511874288320541,
      "num_tokens": 56318645.0,
      "step": 1765
    },
    {
      "entropy": 0.8888132087886333,
      "epoch": 0.16248325196519858,
      "grad_norm": 0.15893776714801788,
      "learning_rate": 9.458705186003006e-05,
      "loss": 0.8876,
      "mean_token_accuracy": 0.747937086969614,
      "num_tokens": 56350575.0,
      "step": 1766
    },
    {
      "entropy": 0.9418562352657318,
      "epoch": 0.1625752583366398,
      "grad_norm": 0.16961203515529633,
      "learning_rate": 9.458398503388844e-05,
      "loss": 0.9567,
      "mean_token_accuracy": 0.7308631576597691,
      "num_tokens": 56382531.0,
      "step": 1767
    },
    {
      "entropy": 0.8901054635643959,
      "epoch": 0.16266726470808104,
      "grad_norm": 0.17205610871315002,
      "learning_rate": 9.458091820774681e-05,
      "loss": 0.8831,
      "mean_token_accuracy": 0.7497150525450706,
      "num_tokens": 56414998.0,
      "step": 1768
    },
    {
      "entropy": 0.855336032807827,
      "epoch": 0.16275927107952226,
      "grad_norm": 0.16010114550590515,
      "learning_rate": 9.457785138160518e-05,
      "loss": 0.8746,
      "mean_token_accuracy": 0.7516513615846634,
      "num_tokens": 56446523.0,
      "step": 1769
    },
    {
      "entropy": 0.8605861179530621,
      "epoch": 0.16285127745096348,
      "grad_norm": 0.15431128442287445,
      "learning_rate": 9.457478455546356e-05,
      "loss": 0.8343,
      "mean_token_accuracy": 0.7618341483175755,
      "num_tokens": 56478772.0,
      "step": 1770
    },
    {
      "entropy": 0.9252845570445061,
      "epoch": 0.1629432838224047,
      "grad_norm": 0.16140660643577576,
      "learning_rate": 9.457171772932192e-05,
      "loss": 0.9298,
      "mean_token_accuracy": 0.7377407178282738,
      "num_tokens": 56511530.0,
      "step": 1771
    },
    {
      "entropy": 0.9086951315402985,
      "epoch": 0.1630352901938459,
      "grad_norm": 0.15055805444717407,
      "learning_rate": 9.45686509031803e-05,
      "loss": 0.873,
      "mean_token_accuracy": 0.7459032200276852,
      "num_tokens": 56543628.0,
      "step": 1772
    },
    {
      "entropy": 0.9088000282645226,
      "epoch": 0.16312729656528716,
      "grad_norm": 0.16467337310314178,
      "learning_rate": 9.456558407703869e-05,
      "loss": 0.8933,
      "mean_token_accuracy": 0.7491044253110886,
      "num_tokens": 56575551.0,
      "step": 1773
    },
    {
      "entropy": 0.922598272562027,
      "epoch": 0.16321930293672837,
      "grad_norm": 0.16504676640033722,
      "learning_rate": 9.456251725089706e-05,
      "loss": 0.9091,
      "mean_token_accuracy": 0.7419484034180641,
      "num_tokens": 56607609.0,
      "step": 1774
    },
    {
      "entropy": 0.9443248175084591,
      "epoch": 0.1633113093081696,
      "grad_norm": 0.16118492186069489,
      "learning_rate": 9.455945042475542e-05,
      "loss": 0.9578,
      "mean_token_accuracy": 0.7299931831657887,
      "num_tokens": 56639739.0,
      "step": 1775
    },
    {
      "entropy": 0.8760691024363041,
      "epoch": 0.1634033156796108,
      "grad_norm": 0.16077694296836853,
      "learning_rate": 9.455638359861379e-05,
      "loss": 0.8901,
      "mean_token_accuracy": 0.7448839545249939,
      "num_tokens": 56671737.0,
      "step": 1776
    },
    {
      "entropy": 0.9142513684928417,
      "epoch": 0.16349532205105202,
      "grad_norm": 0.16895584762096405,
      "learning_rate": 9.455331677247217e-05,
      "loss": 0.9446,
      "mean_token_accuracy": 0.7319508232176304,
      "num_tokens": 56703203.0,
      "step": 1777
    },
    {
      "entropy": 0.927996676415205,
      "epoch": 0.16358732842249327,
      "grad_norm": 0.16760072112083435,
      "learning_rate": 9.455024994633055e-05,
      "loss": 0.9537,
      "mean_token_accuracy": 0.72987986728549,
      "num_tokens": 56734893.0,
      "step": 1778
    },
    {
      "entropy": 0.9208510965108871,
      "epoch": 0.16367933479393448,
      "grad_norm": 0.16630584001541138,
      "learning_rate": 9.454718312018892e-05,
      "loss": 0.9413,
      "mean_token_accuracy": 0.7387723363935947,
      "num_tokens": 56765574.0,
      "step": 1779
    },
    {
      "entropy": 0.932685412466526,
      "epoch": 0.1637713411653757,
      "grad_norm": 0.15871945023536682,
      "learning_rate": 9.454411629404729e-05,
      "loss": 0.9262,
      "mean_token_accuracy": 0.7364226020872593,
      "num_tokens": 56798323.0,
      "step": 1780
    },
    {
      "entropy": 1.0062441863119602,
      "epoch": 0.16386334753681692,
      "grad_norm": 0.1644701063632965,
      "learning_rate": 9.454104946790567e-05,
      "loss": 0.9913,
      "mean_token_accuracy": 0.7196023017168045,
      "num_tokens": 56830211.0,
      "step": 1781
    },
    {
      "entropy": 0.8387195058166981,
      "epoch": 0.16395535390825813,
      "grad_norm": 0.15080101788043976,
      "learning_rate": 9.453798264176404e-05,
      "loss": 0.816,
      "mean_token_accuracy": 0.7627396583557129,
      "num_tokens": 56862308.0,
      "step": 1782
    },
    {
      "entropy": 0.9452171511948109,
      "epoch": 0.16404736027969938,
      "grad_norm": 0.1582026183605194,
      "learning_rate": 9.453491581562242e-05,
      "loss": 0.9501,
      "mean_token_accuracy": 0.7398050241172314,
      "num_tokens": 56894690.0,
      "step": 1783
    },
    {
      "entropy": 0.9867070578038692,
      "epoch": 0.1641393666511406,
      "grad_norm": 0.1728513538837433,
      "learning_rate": 9.453184898948079e-05,
      "loss": 0.9744,
      "mean_token_accuracy": 0.7249099910259247,
      "num_tokens": 56925609.0,
      "step": 1784
    },
    {
      "entropy": 0.8673921562731266,
      "epoch": 0.1642313730225818,
      "grad_norm": 0.1749183088541031,
      "learning_rate": 9.452878216333917e-05,
      "loss": 0.8608,
      "mean_token_accuracy": 0.7569119147956371,
      "num_tokens": 56957595.0,
      "step": 1785
    },
    {
      "entropy": 0.8007127642631531,
      "epoch": 0.16432337939402303,
      "grad_norm": 0.15025275945663452,
      "learning_rate": 9.452571533719754e-05,
      "loss": 0.7923,
      "mean_token_accuracy": 0.7708672694861889,
      "num_tokens": 56988920.0,
      "step": 1786
    },
    {
      "entropy": 0.922088373452425,
      "epoch": 0.16441538576546424,
      "grad_norm": 0.15726657211780548,
      "learning_rate": 9.45226485110559e-05,
      "loss": 0.9123,
      "mean_token_accuracy": 0.7417571544647217,
      "num_tokens": 57021472.0,
      "step": 1787
    },
    {
      "entropy": 0.9163676761090755,
      "epoch": 0.1645073921369055,
      "grad_norm": 0.16552789509296417,
      "learning_rate": 9.451958168491429e-05,
      "loss": 0.9578,
      "mean_token_accuracy": 0.7277688384056091,
      "num_tokens": 57052634.0,
      "step": 1788
    },
    {
      "entropy": 0.8649253733456135,
      "epoch": 0.1645993985083467,
      "grad_norm": 0.15514369308948517,
      "learning_rate": 9.451651485877267e-05,
      "loss": 0.8698,
      "mean_token_accuracy": 0.7529787011444569,
      "num_tokens": 57085330.0,
      "step": 1789
    },
    {
      "entropy": 0.9113885294646025,
      "epoch": 0.16469140487978792,
      "grad_norm": 0.16266684234142303,
      "learning_rate": 9.451344803263104e-05,
      "loss": 0.9172,
      "mean_token_accuracy": 0.7455113753676414,
      "num_tokens": 57117203.0,
      "step": 1790
    },
    {
      "entropy": 0.8250732347369194,
      "epoch": 0.16478341125122914,
      "grad_norm": 0.15113677084445953,
      "learning_rate": 9.45103812064894e-05,
      "loss": 0.8247,
      "mean_token_accuracy": 0.7608569040894508,
      "num_tokens": 57149011.0,
      "step": 1791
    },
    {
      "entropy": 0.8691405653953552,
      "epoch": 0.16487541762267036,
      "grad_norm": 0.15609388053417206,
      "learning_rate": 9.450731438034777e-05,
      "loss": 0.8508,
      "mean_token_accuracy": 0.7568701282143593,
      "num_tokens": 57180787.0,
      "step": 1792
    },
    {
      "entropy": 0.8784987758845091,
      "epoch": 0.1649674239941116,
      "grad_norm": 0.15780064463615417,
      "learning_rate": 9.450424755420615e-05,
      "loss": 0.8786,
      "mean_token_accuracy": 0.749070018529892,
      "num_tokens": 57212441.0,
      "step": 1793
    },
    {
      "entropy": 0.8403750397264957,
      "epoch": 0.16505943036555282,
      "grad_norm": 0.155840203166008,
      "learning_rate": 9.450118072806453e-05,
      "loss": 0.8497,
      "mean_token_accuracy": 0.7601048983633518,
      "num_tokens": 57244365.0,
      "step": 1794
    },
    {
      "entropy": 0.838858000934124,
      "epoch": 0.16515143673699403,
      "grad_norm": 0.15987055003643036,
      "learning_rate": 9.44981139019229e-05,
      "loss": 0.8329,
      "mean_token_accuracy": 0.7612287476658821,
      "num_tokens": 57276154.0,
      "step": 1795
    },
    {
      "entropy": 0.7599440813064575,
      "epoch": 0.16524344310843525,
      "grad_norm": 0.15457147359848022,
      "learning_rate": 9.449504707578127e-05,
      "loss": 0.766,
      "mean_token_accuracy": 0.776075541973114,
      "num_tokens": 57308175.0,
      "step": 1796
    },
    {
      "entropy": 0.8075404651463032,
      "epoch": 0.1653354494798765,
      "grad_norm": 0.1509494036436081,
      "learning_rate": 9.449198024963965e-05,
      "loss": 0.7999,
      "mean_token_accuracy": 0.7686601802706718,
      "num_tokens": 57339984.0,
      "step": 1797
    },
    {
      "entropy": 0.9947471916675568,
      "epoch": 0.1654274558513177,
      "grad_norm": 0.16688406467437744,
      "learning_rate": 9.448891342349802e-05,
      "loss": 0.9795,
      "mean_token_accuracy": 0.7283100448548794,
      "num_tokens": 57371098.0,
      "step": 1798
    },
    {
      "entropy": 0.9369434304535389,
      "epoch": 0.16551946222275893,
      "grad_norm": 0.16810213029384613,
      "learning_rate": 9.44858465973564e-05,
      "loss": 0.9538,
      "mean_token_accuracy": 0.73018803820014,
      "num_tokens": 57402882.0,
      "step": 1799
    },
    {
      "entropy": 0.8812098540365696,
      "epoch": 0.16561146859420015,
      "grad_norm": 0.1688583940267563,
      "learning_rate": 9.448277977121477e-05,
      "loss": 0.9082,
      "mean_token_accuracy": 0.7456429451704025,
      "num_tokens": 57434595.0,
      "step": 1800
    },
    {
      "entropy": 0.8906464502215385,
      "epoch": 0.16570347496564136,
      "grad_norm": 0.16459906101226807,
      "learning_rate": 9.447971294507315e-05,
      "loss": 0.9028,
      "mean_token_accuracy": 0.7469335086643696,
      "num_tokens": 57466901.0,
      "step": 1801
    },
    {
      "entropy": 0.9337184261530638,
      "epoch": 0.1657954813370826,
      "grad_norm": 0.1629372090101242,
      "learning_rate": 9.447664611893152e-05,
      "loss": 0.9363,
      "mean_token_accuracy": 0.734474990516901,
      "num_tokens": 57498673.0,
      "step": 1802
    },
    {
      "entropy": 0.8730062209069729,
      "epoch": 0.16588748770852382,
      "grad_norm": 0.15843915939331055,
      "learning_rate": 9.44735792927899e-05,
      "loss": 0.8887,
      "mean_token_accuracy": 0.7526570595800877,
      "num_tokens": 57530721.0,
      "step": 1803
    },
    {
      "entropy": 0.8847571611404419,
      "epoch": 0.16597949407996504,
      "grad_norm": 0.1584598422050476,
      "learning_rate": 9.447051246664828e-05,
      "loss": 0.8606,
      "mean_token_accuracy": 0.7584680840373039,
      "num_tokens": 57563217.0,
      "step": 1804
    },
    {
      "entropy": 0.9700459931045771,
      "epoch": 0.16607150045140626,
      "grad_norm": 0.16291660070419312,
      "learning_rate": 9.446744564050665e-05,
      "loss": 0.9613,
      "mean_token_accuracy": 0.7355053387582302,
      "num_tokens": 57595478.0,
      "step": 1805
    },
    {
      "entropy": 0.9375690799206495,
      "epoch": 0.16616350682284747,
      "grad_norm": 0.1539071798324585,
      "learning_rate": 9.446437881436502e-05,
      "loss": 0.8817,
      "mean_token_accuracy": 0.7521509900689125,
      "num_tokens": 57627893.0,
      "step": 1806
    },
    {
      "entropy": 0.9058787152171135,
      "epoch": 0.16625551319428872,
      "grad_norm": 0.15624967217445374,
      "learning_rate": 9.446131198822339e-05,
      "loss": 0.8793,
      "mean_token_accuracy": 0.7508661113679409,
      "num_tokens": 57659607.0,
      "step": 1807
    },
    {
      "entropy": 0.9439679458737373,
      "epoch": 0.16634751956572993,
      "grad_norm": 0.15770089626312256,
      "learning_rate": 9.445824516208177e-05,
      "loss": 0.947,
      "mean_token_accuracy": 0.7352209985256195,
      "num_tokens": 57691702.0,
      "step": 1808
    },
    {
      "entropy": 0.9460605271160603,
      "epoch": 0.16643952593717115,
      "grad_norm": 0.15990819036960602,
      "learning_rate": 9.445517833594015e-05,
      "loss": 0.953,
      "mean_token_accuracy": 0.731384065002203,
      "num_tokens": 57723542.0,
      "step": 1809
    },
    {
      "entropy": 0.8863812424242496,
      "epoch": 0.16653153230861237,
      "grad_norm": 0.16009779274463654,
      "learning_rate": 9.445211150979852e-05,
      "loss": 0.8835,
      "mean_token_accuracy": 0.7530708834528923,
      "num_tokens": 57756057.0,
      "step": 1810
    },
    {
      "entropy": 0.9464438445866108,
      "epoch": 0.16662353868005358,
      "grad_norm": 0.16117577254772186,
      "learning_rate": 9.444904468365688e-05,
      "loss": 0.9531,
      "mean_token_accuracy": 0.7309932634234428,
      "num_tokens": 57787926.0,
      "step": 1811
    },
    {
      "entropy": 0.9742291569709778,
      "epoch": 0.16671554505149483,
      "grad_norm": 0.16853958368301392,
      "learning_rate": 9.444597785751527e-05,
      "loss": 0.9821,
      "mean_token_accuracy": 0.7299029007554054,
      "num_tokens": 57819326.0,
      "step": 1812
    },
    {
      "entropy": 1.0114005357027054,
      "epoch": 0.16680755142293605,
      "grad_norm": 0.16757190227508545,
      "learning_rate": 9.444291103137363e-05,
      "loss": 1.0376,
      "mean_token_accuracy": 0.7090245075523853,
      "num_tokens": 57850826.0,
      "step": 1813
    },
    {
      "entropy": 0.9165155664086342,
      "epoch": 0.16689955779437726,
      "grad_norm": 0.1611187905073166,
      "learning_rate": 9.443984420523201e-05,
      "loss": 0.9209,
      "mean_token_accuracy": 0.7389795556664467,
      "num_tokens": 57883274.0,
      "step": 1814
    },
    {
      "entropy": 0.8793531209230423,
      "epoch": 0.16699156416581848,
      "grad_norm": 0.15698812901973724,
      "learning_rate": 9.443677737909038e-05,
      "loss": 0.8744,
      "mean_token_accuracy": 0.7513559907674789,
      "num_tokens": 57915909.0,
      "step": 1815
    },
    {
      "entropy": 1.0087986439466476,
      "epoch": 0.1670835705372597,
      "grad_norm": 0.1689801663160324,
      "learning_rate": 9.443371055294876e-05,
      "loss": 1.0108,
      "mean_token_accuracy": 0.7207426652312279,
      "num_tokens": 57946399.0,
      "step": 1816
    },
    {
      "entropy": 0.9170082397758961,
      "epoch": 0.16717557690870094,
      "grad_norm": 0.16203944385051727,
      "learning_rate": 9.443064372680713e-05,
      "loss": 0.9345,
      "mean_token_accuracy": 0.7348842956125736,
      "num_tokens": 57978328.0,
      "step": 1817
    },
    {
      "entropy": 0.861324604600668,
      "epoch": 0.16726758328014216,
      "grad_norm": 0.15819591283798218,
      "learning_rate": 9.44275769006655e-05,
      "loss": 0.8763,
      "mean_token_accuracy": 0.7510399147868156,
      "num_tokens": 58010752.0,
      "step": 1818
    },
    {
      "entropy": 0.9799933917820454,
      "epoch": 0.16735958965158337,
      "grad_norm": 0.16706959903240204,
      "learning_rate": 9.442451007452388e-05,
      "loss": 0.9858,
      "mean_token_accuracy": 0.7285223975777626,
      "num_tokens": 58042292.0,
      "step": 1819
    },
    {
      "entropy": 0.9507187269628048,
      "epoch": 0.1674515960230246,
      "grad_norm": 0.1628035604953766,
      "learning_rate": 9.442144324838226e-05,
      "loss": 0.9357,
      "mean_token_accuracy": 0.7324147261679173,
      "num_tokens": 58074107.0,
      "step": 1820
    },
    {
      "entropy": 0.964561752974987,
      "epoch": 0.1675436023944658,
      "grad_norm": 0.16119347512722015,
      "learning_rate": 9.441837642224063e-05,
      "loss": 0.9615,
      "mean_token_accuracy": 0.7303733937442303,
      "num_tokens": 58105868.0,
      "step": 1821
    },
    {
      "entropy": 0.9015658274292946,
      "epoch": 0.16763560876590705,
      "grad_norm": 0.1629556268453598,
      "learning_rate": 9.4415309596099e-05,
      "loss": 0.9007,
      "mean_token_accuracy": 0.7462888173758984,
      "num_tokens": 58138169.0,
      "step": 1822
    },
    {
      "entropy": 0.8483394980430603,
      "epoch": 0.16772761513734827,
      "grad_norm": 0.15847481787204742,
      "learning_rate": 9.441224276995737e-05,
      "loss": 0.8383,
      "mean_token_accuracy": 0.7653291411697865,
      "num_tokens": 58170608.0,
      "step": 1823
    },
    {
      "entropy": 0.8451830819249153,
      "epoch": 0.16781962150878948,
      "grad_norm": 0.1520046591758728,
      "learning_rate": 9.440917594381575e-05,
      "loss": 0.8311,
      "mean_token_accuracy": 0.7609617039561272,
      "num_tokens": 58202210.0,
      "step": 1824
    },
    {
      "entropy": 0.9427127167582512,
      "epoch": 0.1679116278802307,
      "grad_norm": 0.16389355063438416,
      "learning_rate": 9.440610911767413e-05,
      "loss": 0.9324,
      "mean_token_accuracy": 0.7387063913047314,
      "num_tokens": 58234360.0,
      "step": 1825
    },
    {
      "entropy": 0.8891644217073917,
      "epoch": 0.16800363425167192,
      "grad_norm": 0.1667487770318985,
      "learning_rate": 9.44030422915325e-05,
      "loss": 0.9046,
      "mean_token_accuracy": 0.7443866729736328,
      "num_tokens": 58265291.0,
      "step": 1826
    },
    {
      "entropy": 0.8799921870231628,
      "epoch": 0.16809564062311316,
      "grad_norm": 0.16341419517993927,
      "learning_rate": 9.439997546539086e-05,
      "loss": 0.9067,
      "mean_token_accuracy": 0.7409439831972122,
      "num_tokens": 58297813.0,
      "step": 1827
    },
    {
      "entropy": 0.8428296744823456,
      "epoch": 0.16818764699455438,
      "grad_norm": 0.15748274326324463,
      "learning_rate": 9.439690863924925e-05,
      "loss": 0.8329,
      "mean_token_accuracy": 0.7597299888730049,
      "num_tokens": 58328535.0,
      "step": 1828
    },
    {
      "entropy": 0.8273894861340523,
      "epoch": 0.1682796533659956,
      "grad_norm": 0.15217240154743195,
      "learning_rate": 9.439384181310761e-05,
      "loss": 0.819,
      "mean_token_accuracy": 0.7627876736223698,
      "num_tokens": 58361122.0,
      "step": 1829
    },
    {
      "entropy": 0.9610063135623932,
      "epoch": 0.1683716597374368,
      "grad_norm": 0.16728657484054565,
      "learning_rate": 9.4390774986966e-05,
      "loss": 0.9591,
      "mean_token_accuracy": 0.7264038920402527,
      "num_tokens": 58392852.0,
      "step": 1830
    },
    {
      "entropy": 0.9310832601040602,
      "epoch": 0.16846366610887803,
      "grad_norm": 0.15612289309501648,
      "learning_rate": 9.438770816082436e-05,
      "loss": 0.9228,
      "mean_token_accuracy": 0.7391585856676102,
      "num_tokens": 58425184.0,
      "step": 1831
    },
    {
      "entropy": 0.7883002031594515,
      "epoch": 0.16855567248031927,
      "grad_norm": 0.15211637318134308,
      "learning_rate": 9.438464133468274e-05,
      "loss": 0.7902,
      "mean_token_accuracy": 0.7738193050026894,
      "num_tokens": 58457159.0,
      "step": 1832
    },
    {
      "entropy": 0.9498493075370789,
      "epoch": 0.1686476788517605,
      "grad_norm": 0.15690286457538605,
      "learning_rate": 9.438157450854111e-05,
      "loss": 0.9328,
      "mean_token_accuracy": 0.734191682189703,
      "num_tokens": 58489235.0,
      "step": 1833
    },
    {
      "entropy": 0.9377902150154114,
      "epoch": 0.1687396852232017,
      "grad_norm": 0.15876504778862,
      "learning_rate": 9.437850768239948e-05,
      "loss": 0.921,
      "mean_token_accuracy": 0.7422527074813843,
      "num_tokens": 58521841.0,
      "step": 1834
    },
    {
      "entropy": 0.8864666670560837,
      "epoch": 0.16883169159464292,
      "grad_norm": 0.15714214742183685,
      "learning_rate": 9.437544085625788e-05,
      "loss": 0.8724,
      "mean_token_accuracy": 0.7530332393944263,
      "num_tokens": 58554170.0,
      "step": 1835
    },
    {
      "entropy": 0.8813110329210758,
      "epoch": 0.16892369796608414,
      "grad_norm": 0.1594560146331787,
      "learning_rate": 9.437237403011624e-05,
      "loss": 0.8887,
      "mean_token_accuracy": 0.7488854713737965,
      "num_tokens": 58585353.0,
      "step": 1836
    },
    {
      "entropy": 0.9039559848606586,
      "epoch": 0.16901570433752539,
      "grad_norm": 0.1625613123178482,
      "learning_rate": 9.436930720397461e-05,
      "loss": 0.9307,
      "mean_token_accuracy": 0.7355100885033607,
      "num_tokens": 58617214.0,
      "step": 1837
    },
    {
      "entropy": 0.857202772051096,
      "epoch": 0.1691077107089666,
      "grad_norm": 0.1541692465543747,
      "learning_rate": 9.436624037783298e-05,
      "loss": 0.8648,
      "mean_token_accuracy": 0.7526212744414806,
      "num_tokens": 58648907.0,
      "step": 1838
    },
    {
      "entropy": 0.7868402227759361,
      "epoch": 0.16919971708040782,
      "grad_norm": 0.15645843744277954,
      "learning_rate": 9.436317355169136e-05,
      "loss": 0.8029,
      "mean_token_accuracy": 0.7684844247996807,
      "num_tokens": 58680533.0,
      "step": 1839
    },
    {
      "entropy": 0.7920204848051071,
      "epoch": 0.16929172345184904,
      "grad_norm": 0.1585509330034256,
      "learning_rate": 9.436010672554974e-05,
      "loss": 0.8039,
      "mean_token_accuracy": 0.7680395990610123,
      "num_tokens": 58711877.0,
      "step": 1840
    },
    {
      "entropy": 0.9269446805119514,
      "epoch": 0.16938372982329025,
      "grad_norm": 0.1614672988653183,
      "learning_rate": 9.435703989940811e-05,
      "loss": 0.9386,
      "mean_token_accuracy": 0.7387395948171616,
      "num_tokens": 58743657.0,
      "step": 1841
    },
    {
      "entropy": 0.9241079576313496,
      "epoch": 0.1694757361947315,
      "grad_norm": 0.15631060302257538,
      "learning_rate": 9.435397307326648e-05,
      "loss": 0.9131,
      "mean_token_accuracy": 0.74215392395854,
      "num_tokens": 58775336.0,
      "step": 1842
    },
    {
      "entropy": 0.9517180733382702,
      "epoch": 0.1695677425661727,
      "grad_norm": 0.15634024143218994,
      "learning_rate": 9.435090624712486e-05,
      "loss": 0.9174,
      "mean_token_accuracy": 0.7396046333014965,
      "num_tokens": 58806961.0,
      "step": 1843
    },
    {
      "entropy": 0.9258603788912296,
      "epoch": 0.16965974893761393,
      "grad_norm": 0.15717947483062744,
      "learning_rate": 9.434783942098323e-05,
      "loss": 0.9215,
      "mean_token_accuracy": 0.7375474758446217,
      "num_tokens": 58839513.0,
      "step": 1844
    },
    {
      "entropy": 0.9045845046639442,
      "epoch": 0.16975175530905515,
      "grad_norm": 0.15425685048103333,
      "learning_rate": 9.434477259484161e-05,
      "loss": 0.879,
      "mean_token_accuracy": 0.7501320019364357,
      "num_tokens": 58872164.0,
      "step": 1845
    },
    {
      "entropy": 0.9177219197154045,
      "epoch": 0.16984376168049636,
      "grad_norm": 0.15842808783054352,
      "learning_rate": 9.434170576869998e-05,
      "loss": 0.9236,
      "mean_token_accuracy": 0.7388772144913673,
      "num_tokens": 58903405.0,
      "step": 1846
    },
    {
      "entropy": 0.8883706144988537,
      "epoch": 0.1699357680519376,
      "grad_norm": 0.15935038030147552,
      "learning_rate": 9.433863894255836e-05,
      "loss": 0.9031,
      "mean_token_accuracy": 0.7455946169793606,
      "num_tokens": 58935244.0,
      "step": 1847
    },
    {
      "entropy": 0.8977056592702866,
      "epoch": 0.17002777442337882,
      "grad_norm": 0.1567680388689041,
      "learning_rate": 9.433557211641673e-05,
      "loss": 0.899,
      "mean_token_accuracy": 0.7477743178606033,
      "num_tokens": 58966904.0,
      "step": 1848
    },
    {
      "entropy": 0.8123586624860764,
      "epoch": 0.17011978079482004,
      "grad_norm": 0.15393827855587006,
      "learning_rate": 9.433250529027509e-05,
      "loss": 0.8089,
      "mean_token_accuracy": 0.7651770859956741,
      "num_tokens": 58998425.0,
      "step": 1849
    },
    {
      "entropy": 0.8685755357146263,
      "epoch": 0.17021178716626126,
      "grad_norm": 0.15230123698711395,
      "learning_rate": 9.432943846413347e-05,
      "loss": 0.8633,
      "mean_token_accuracy": 0.7513212189078331,
      "num_tokens": 59030991.0,
      "step": 1850
    },
    {
      "entropy": 0.996167179197073,
      "epoch": 0.17030379353770247,
      "grad_norm": 0.16995355486869812,
      "learning_rate": 9.432637163799186e-05,
      "loss": 0.9957,
      "mean_token_accuracy": 0.718870609998703,
      "num_tokens": 59063281.0,
      "step": 1851
    },
    {
      "entropy": 0.9604267254471779,
      "epoch": 0.17039579990914372,
      "grad_norm": 0.16216331720352173,
      "learning_rate": 9.432330481185022e-05,
      "loss": 0.9593,
      "mean_token_accuracy": 0.7290573716163635,
      "num_tokens": 59095922.0,
      "step": 1852
    },
    {
      "entropy": 0.885973397642374,
      "epoch": 0.17048780628058494,
      "grad_norm": 0.15408769249916077,
      "learning_rate": 9.432023798570859e-05,
      "loss": 0.8855,
      "mean_token_accuracy": 0.7510918751358986,
      "num_tokens": 59128246.0,
      "step": 1853
    },
    {
      "entropy": 0.8039834406226873,
      "epoch": 0.17057981265202615,
      "grad_norm": 0.15309977531433105,
      "learning_rate": 9.431717115956696e-05,
      "loss": 0.8007,
      "mean_token_accuracy": 0.7673305794596672,
      "num_tokens": 59160231.0,
      "step": 1854
    },
    {
      "entropy": 0.9160208106040955,
      "epoch": 0.17067181902346737,
      "grad_norm": 0.16488729417324066,
      "learning_rate": 9.431410433342534e-05,
      "loss": 0.941,
      "mean_token_accuracy": 0.7335659489035606,
      "num_tokens": 59192492.0,
      "step": 1855
    },
    {
      "entropy": 0.7917605899274349,
      "epoch": 0.17076382539490859,
      "grad_norm": 0.16340278089046478,
      "learning_rate": 9.431103750728372e-05,
      "loss": 0.8225,
      "mean_token_accuracy": 0.7655744291841984,
      "num_tokens": 59224016.0,
      "step": 1856
    },
    {
      "entropy": 0.8084781393408775,
      "epoch": 0.17085583176634983,
      "grad_norm": 0.15780781209468842,
      "learning_rate": 9.430797068114209e-05,
      "loss": 0.8266,
      "mean_token_accuracy": 0.7568622156977654,
      "num_tokens": 59256426.0,
      "step": 1857
    },
    {
      "entropy": 0.8871316742151976,
      "epoch": 0.17094783813779105,
      "grad_norm": 0.1542637199163437,
      "learning_rate": 9.430490385500046e-05,
      "loss": 0.8757,
      "mean_token_accuracy": 0.7519547455012798,
      "num_tokens": 59288713.0,
      "step": 1858
    },
    {
      "entropy": 0.9869510233402252,
      "epoch": 0.17103984450923226,
      "grad_norm": 0.16224883496761322,
      "learning_rate": 9.430183702885884e-05,
      "loss": 0.9614,
      "mean_token_accuracy": 0.7330340966582298,
      "num_tokens": 59321367.0,
      "step": 1859
    },
    {
      "entropy": 1.0379321537911892,
      "epoch": 0.17113185088067348,
      "grad_norm": 0.16850978136062622,
      "learning_rate": 9.429877020271721e-05,
      "loss": 1.0353,
      "mean_token_accuracy": 0.7148441299796104,
      "num_tokens": 59352644.0,
      "step": 1860
    },
    {
      "entropy": 0.8512087445706129,
      "epoch": 0.1712238572521147,
      "grad_norm": 0.15685105323791504,
      "learning_rate": 9.429570337657559e-05,
      "loss": 0.8253,
      "mean_token_accuracy": 0.7593910396099091,
      "num_tokens": 59384604.0,
      "step": 1861
    },
    {
      "entropy": 0.9197734668850899,
      "epoch": 0.17131586362355594,
      "grad_norm": 0.1591319739818573,
      "learning_rate": 9.429263655043397e-05,
      "loss": 0.8984,
      "mean_token_accuracy": 0.7394395656883717,
      "num_tokens": 59416322.0,
      "step": 1862
    },
    {
      "entropy": 0.83257388882339,
      "epoch": 0.17140786999499716,
      "grad_norm": 0.1549418717622757,
      "learning_rate": 9.428956972429234e-05,
      "loss": 0.8346,
      "mean_token_accuracy": 0.7629285864531994,
      "num_tokens": 59448943.0,
      "step": 1863
    },
    {
      "entropy": 0.9548040069639683,
      "epoch": 0.17149987636643838,
      "grad_norm": 0.1675068885087967,
      "learning_rate": 9.42865028981507e-05,
      "loss": 0.9551,
      "mean_token_accuracy": 0.7319391667842865,
      "num_tokens": 59481363.0,
      "step": 1864
    },
    {
      "entropy": 0.8938488233834505,
      "epoch": 0.1715918827378796,
      "grad_norm": 0.15862978994846344,
      "learning_rate": 9.428343607200907e-05,
      "loss": 0.9038,
      "mean_token_accuracy": 0.7440845817327499,
      "num_tokens": 59513516.0,
      "step": 1865
    },
    {
      "entropy": 0.9240580834448338,
      "epoch": 0.1716838891093208,
      "grad_norm": 0.162907212972641,
      "learning_rate": 9.428036924586746e-05,
      "loss": 0.9194,
      "mean_token_accuracy": 0.7435371279716492,
      "num_tokens": 59545301.0,
      "step": 1866
    },
    {
      "entropy": 0.8283825144171715,
      "epoch": 0.17177589548076205,
      "grad_norm": 0.1529933214187622,
      "learning_rate": 9.427730241972584e-05,
      "loss": 0.8092,
      "mean_token_accuracy": 0.7666553258895874,
      "num_tokens": 59577319.0,
      "step": 1867
    },
    {
      "entropy": 0.9682305194437504,
      "epoch": 0.17186790185220327,
      "grad_norm": 0.1727389544248581,
      "learning_rate": 9.42742355935842e-05,
      "loss": 0.9685,
      "mean_token_accuracy": 0.726842425763607,
      "num_tokens": 59608538.0,
      "step": 1868
    },
    {
      "entropy": 0.8739567399024963,
      "epoch": 0.1719599082236445,
      "grad_norm": 0.15527045726776123,
      "learning_rate": 9.427116876744257e-05,
      "loss": 0.8725,
      "mean_token_accuracy": 0.7530970685184002,
      "num_tokens": 59641016.0,
      "step": 1869
    },
    {
      "entropy": 0.9649168588221073,
      "epoch": 0.1720519145950857,
      "grad_norm": 0.16884897649288177,
      "learning_rate": 9.426810194130095e-05,
      "loss": 0.97,
      "mean_token_accuracy": 0.7300683669745922,
      "num_tokens": 59672789.0,
      "step": 1870
    },
    {
      "entropy": 0.8569833412766457,
      "epoch": 0.17214392096652692,
      "grad_norm": 0.16047517955303192,
      "learning_rate": 9.426503511515932e-05,
      "loss": 0.8698,
      "mean_token_accuracy": 0.7477751262485981,
      "num_tokens": 59704541.0,
      "step": 1871
    },
    {
      "entropy": 0.8406945317983627,
      "epoch": 0.17223592733796816,
      "grad_norm": 0.1621973067522049,
      "learning_rate": 9.42619682890177e-05,
      "loss": 0.8492,
      "mean_token_accuracy": 0.7573677487671375,
      "num_tokens": 59736305.0,
      "step": 1872
    },
    {
      "entropy": 0.8772669285535812,
      "epoch": 0.17232793370940938,
      "grad_norm": 0.15862306952476501,
      "learning_rate": 9.425890146287607e-05,
      "loss": 0.8743,
      "mean_token_accuracy": 0.751763604581356,
      "num_tokens": 59768621.0,
      "step": 1873
    },
    {
      "entropy": 0.8772459179162979,
      "epoch": 0.1724199400808506,
      "grad_norm": 0.16818562150001526,
      "learning_rate": 9.425583463673445e-05,
      "loss": 0.9206,
      "mean_token_accuracy": 0.7448192238807678,
      "num_tokens": 59800540.0,
      "step": 1874
    },
    {
      "entropy": 0.8375705778598785,
      "epoch": 0.17251194645229181,
      "grad_norm": 0.1598014086484909,
      "learning_rate": 9.425276781059282e-05,
      "loss": 0.8534,
      "mean_token_accuracy": 0.7576014995574951,
      "num_tokens": 59832585.0,
      "step": 1875
    },
    {
      "entropy": 0.9514926746487617,
      "epoch": 0.17260395282373303,
      "grad_norm": 0.15755562484264374,
      "learning_rate": 9.424970098445119e-05,
      "loss": 0.9445,
      "mean_token_accuracy": 0.7323351576924324,
      "num_tokens": 59864720.0,
      "step": 1876
    },
    {
      "entropy": 0.8084214963018894,
      "epoch": 0.17269595919517428,
      "grad_norm": 0.15232513844966888,
      "learning_rate": 9.424663415830957e-05,
      "loss": 0.8002,
      "mean_token_accuracy": 0.7686799764633179,
      "num_tokens": 59897361.0,
      "step": 1877
    },
    {
      "entropy": 0.941080778837204,
      "epoch": 0.1727879655666155,
      "grad_norm": 0.162842258810997,
      "learning_rate": 9.424356733216795e-05,
      "loss": 0.9321,
      "mean_token_accuracy": 0.7354013286530972,
      "num_tokens": 59929405.0,
      "step": 1878
    },
    {
      "entropy": 0.9692078158259392,
      "epoch": 0.1728799719380567,
      "grad_norm": 0.1627609133720398,
      "learning_rate": 9.424050050602632e-05,
      "loss": 0.9393,
      "mean_token_accuracy": 0.7363405264914036,
      "num_tokens": 59960778.0,
      "step": 1879
    },
    {
      "entropy": 0.9325489066541195,
      "epoch": 0.17297197830949793,
      "grad_norm": 0.15714102983474731,
      "learning_rate": 9.423743367988469e-05,
      "loss": 0.9233,
      "mean_token_accuracy": 0.7390247248113155,
      "num_tokens": 59993064.0,
      "step": 1880
    },
    {
      "entropy": 0.9120017848908901,
      "epoch": 0.17306398468093914,
      "grad_norm": 0.16190214455127716,
      "learning_rate": 9.423436685374305e-05,
      "loss": 0.9045,
      "mean_token_accuracy": 0.7406519912183285,
      "num_tokens": 60024629.0,
      "step": 1881
    },
    {
      "entropy": 0.9417525641620159,
      "epoch": 0.1731559910523804,
      "grad_norm": 0.15931543707847595,
      "learning_rate": 9.423130002760145e-05,
      "loss": 0.9233,
      "mean_token_accuracy": 0.7374418713152409,
      "num_tokens": 60056606.0,
      "step": 1882
    },
    {
      "entropy": 0.8955351933836937,
      "epoch": 0.1732479974238216,
      "grad_norm": 0.15810322761535645,
      "learning_rate": 9.422823320145982e-05,
      "loss": 0.9119,
      "mean_token_accuracy": 0.740511491894722,
      "num_tokens": 60089113.0,
      "step": 1883
    },
    {
      "entropy": 0.9876630119979382,
      "epoch": 0.17334000379526282,
      "grad_norm": 0.165706604719162,
      "learning_rate": 9.422516637531819e-05,
      "loss": 0.9892,
      "mean_token_accuracy": 0.7245639860630035,
      "num_tokens": 60119962.0,
      "step": 1884
    },
    {
      "entropy": 0.8956912364810705,
      "epoch": 0.17343201016670404,
      "grad_norm": 0.16372308135032654,
      "learning_rate": 9.422209954917655e-05,
      "loss": 0.9045,
      "mean_token_accuracy": 0.7458113171160221,
      "num_tokens": 60151801.0,
      "step": 1885
    },
    {
      "entropy": 0.8040297999978065,
      "epoch": 0.17352401653814525,
      "grad_norm": 0.15230725705623627,
      "learning_rate": 9.421903272303493e-05,
      "loss": 0.8001,
      "mean_token_accuracy": 0.7688378877937794,
      "num_tokens": 60182971.0,
      "step": 1886
    },
    {
      "entropy": 0.9118799418210983,
      "epoch": 0.1736160229095865,
      "grad_norm": 0.1633504182100296,
      "learning_rate": 9.421596589689332e-05,
      "loss": 0.8856,
      "mean_token_accuracy": 0.7484063655138016,
      "num_tokens": 60214750.0,
      "step": 1887
    },
    {
      "entropy": 0.8926295302808285,
      "epoch": 0.17370802928102771,
      "grad_norm": 0.15819217264652252,
      "learning_rate": 9.421289907075168e-05,
      "loss": 0.88,
      "mean_token_accuracy": 0.7453234866261482,
      "num_tokens": 60247202.0,
      "step": 1888
    },
    {
      "entropy": 0.8760696277022362,
      "epoch": 0.17380003565246893,
      "grad_norm": 0.15865299105644226,
      "learning_rate": 9.420983224461005e-05,
      "loss": 0.9058,
      "mean_token_accuracy": 0.7442635595798492,
      "num_tokens": 60279723.0,
      "step": 1889
    },
    {
      "entropy": 0.8700458966195583,
      "epoch": 0.17389204202391015,
      "grad_norm": 0.1593678891658783,
      "learning_rate": 9.420676541846843e-05,
      "loss": 0.8925,
      "mean_token_accuracy": 0.7463311813771725,
      "num_tokens": 60312216.0,
      "step": 1890
    },
    {
      "entropy": 0.9081114921718836,
      "epoch": 0.17398404839535137,
      "grad_norm": 0.1627340018749237,
      "learning_rate": 9.42036985923268e-05,
      "loss": 0.9172,
      "mean_token_accuracy": 0.7409970425069332,
      "num_tokens": 60343658.0,
      "step": 1891
    },
    {
      "entropy": 0.9439083747565746,
      "epoch": 0.1740760547667926,
      "grad_norm": 0.16297918558120728,
      "learning_rate": 9.420063176618518e-05,
      "loss": 0.9489,
      "mean_token_accuracy": 0.7335664182901382,
      "num_tokens": 60375761.0,
      "step": 1892
    },
    {
      "entropy": 0.866368655115366,
      "epoch": 0.17416806113823383,
      "grad_norm": 0.15370438992977142,
      "learning_rate": 9.419756494004356e-05,
      "loss": 0.8533,
      "mean_token_accuracy": 0.7553115338087082,
      "num_tokens": 60407299.0,
      "step": 1893
    },
    {
      "entropy": 0.8385331127792597,
      "epoch": 0.17426006750967504,
      "grad_norm": 0.1528184413909912,
      "learning_rate": 9.419449811390193e-05,
      "loss": 0.8341,
      "mean_token_accuracy": 0.7628591507673264,
      "num_tokens": 60439515.0,
      "step": 1894
    },
    {
      "entropy": 0.958952110260725,
      "epoch": 0.17435207388111626,
      "grad_norm": 0.16451141238212585,
      "learning_rate": 9.41914312877603e-05,
      "loss": 0.9643,
      "mean_token_accuracy": 0.7299320884048939,
      "num_tokens": 60471421.0,
      "step": 1895
    },
    {
      "entropy": 0.9270216189324856,
      "epoch": 0.17444408025255748,
      "grad_norm": 0.16445523500442505,
      "learning_rate": 9.418836446161867e-05,
      "loss": 0.9503,
      "mean_token_accuracy": 0.7359288483858109,
      "num_tokens": 60503548.0,
      "step": 1896
    },
    {
      "entropy": 0.846885584294796,
      "epoch": 0.17453608662399872,
      "grad_norm": 0.1647835373878479,
      "learning_rate": 9.418529763547705e-05,
      "loss": 0.8577,
      "mean_token_accuracy": 0.7575631700456142,
      "num_tokens": 60535181.0,
      "step": 1897
    },
    {
      "entropy": 0.8832699693739414,
      "epoch": 0.17462809299543994,
      "grad_norm": 0.15676485002040863,
      "learning_rate": 9.418223080933543e-05,
      "loss": 0.8922,
      "mean_token_accuracy": 0.746915016323328,
      "num_tokens": 60567499.0,
      "step": 1898
    },
    {
      "entropy": 0.8710623979568481,
      "epoch": 0.17472009936688115,
      "grad_norm": 0.15335582196712494,
      "learning_rate": 9.41791639831938e-05,
      "loss": 0.8643,
      "mean_token_accuracy": 0.7527070604264736,
      "num_tokens": 60599713.0,
      "step": 1899
    },
    {
      "entropy": 0.8578383438289165,
      "epoch": 0.17481210573832237,
      "grad_norm": 0.1570405513048172,
      "learning_rate": 9.417609715705217e-05,
      "loss": 0.8444,
      "mean_token_accuracy": 0.7536755613982677,
      "num_tokens": 60631095.0,
      "step": 1900
    },
    {
      "entropy": 0.8966676965355873,
      "epoch": 0.1749041121097636,
      "grad_norm": 0.16427390277385712,
      "learning_rate": 9.417303033091055e-05,
      "loss": 0.9043,
      "mean_token_accuracy": 0.7457891143858433,
      "num_tokens": 60662702.0,
      "step": 1901
    },
    {
      "entropy": 0.9051119722425938,
      "epoch": 0.17499611848120483,
      "grad_norm": 0.15830010175704956,
      "learning_rate": 9.416996350476892e-05,
      "loss": 0.8816,
      "mean_token_accuracy": 0.7485939674079418,
      "num_tokens": 60695143.0,
      "step": 1902
    },
    {
      "entropy": 0.8896239250898361,
      "epoch": 0.17508812485264605,
      "grad_norm": 0.15849563479423523,
      "learning_rate": 9.41668966786273e-05,
      "loss": 0.8813,
      "mean_token_accuracy": 0.7472447268664837,
      "num_tokens": 60727449.0,
      "step": 1903
    },
    {
      "entropy": 0.9845099411904812,
      "epoch": 0.17518013122408727,
      "grad_norm": 0.16663126647472382,
      "learning_rate": 9.416382985248566e-05,
      "loss": 0.9788,
      "mean_token_accuracy": 0.7253895811736584,
      "num_tokens": 60758539.0,
      "step": 1904
    },
    {
      "entropy": 0.8367153108119965,
      "epoch": 0.17527213759552848,
      "grad_norm": 0.1567063331604004,
      "learning_rate": 9.416076302634405e-05,
      "loss": 0.8276,
      "mean_token_accuracy": 0.7577414326369762,
      "num_tokens": 60790383.0,
      "step": 1905
    },
    {
      "entropy": 0.8066691048443317,
      "epoch": 0.1753641439669697,
      "grad_norm": 0.15768389403820038,
      "learning_rate": 9.415769620020241e-05,
      "loss": 0.7903,
      "mean_token_accuracy": 0.7732889391481876,
      "num_tokens": 60821777.0,
      "step": 1906
    },
    {
      "entropy": 0.8438106011599302,
      "epoch": 0.17545615033841094,
      "grad_norm": 0.154905766248703,
      "learning_rate": 9.415462937406078e-05,
      "loss": 0.8202,
      "mean_token_accuracy": 0.7615134976804256,
      "num_tokens": 60853067.0,
      "step": 1907
    },
    {
      "entropy": 0.8664483912289143,
      "epoch": 0.17554815670985216,
      "grad_norm": 0.16080358624458313,
      "learning_rate": 9.415156254791916e-05,
      "loss": 0.8759,
      "mean_token_accuracy": 0.752622690051794,
      "num_tokens": 60885306.0,
      "step": 1908
    },
    {
      "entropy": 0.8651197906583548,
      "epoch": 0.17564016308129338,
      "grad_norm": 0.15626160800457,
      "learning_rate": 9.414849572177754e-05,
      "loss": 0.8839,
      "mean_token_accuracy": 0.745987094938755,
      "num_tokens": 60917652.0,
      "step": 1909
    },
    {
      "entropy": 0.8212180305272341,
      "epoch": 0.1757321694527346,
      "grad_norm": 0.15831954777240753,
      "learning_rate": 9.414542889563591e-05,
      "loss": 0.8405,
      "mean_token_accuracy": 0.7584434375166893,
      "num_tokens": 60949746.0,
      "step": 1910
    },
    {
      "entropy": 0.8278720322996378,
      "epoch": 0.17582417582417584,
      "grad_norm": 0.16028863191604614,
      "learning_rate": 9.414236206949428e-05,
      "loss": 0.8255,
      "mean_token_accuracy": 0.7625668682157993,
      "num_tokens": 60980967.0,
      "step": 1911
    },
    {
      "entropy": 0.9243301004171371,
      "epoch": 0.17591618219561705,
      "grad_norm": 0.1606753170490265,
      "learning_rate": 9.413929524335265e-05,
      "loss": 0.9073,
      "mean_token_accuracy": 0.7430240288376808,
      "num_tokens": 61012937.0,
      "step": 1912
    },
    {
      "entropy": 0.875252265483141,
      "epoch": 0.17600818856705827,
      "grad_norm": 0.1650233119726181,
      "learning_rate": 9.413622841721103e-05,
      "loss": 0.8786,
      "mean_token_accuracy": 0.7480699643492699,
      "num_tokens": 61045226.0,
      "step": 1913
    },
    {
      "entropy": 0.8323456011712551,
      "epoch": 0.1761001949384995,
      "grad_norm": 0.15924206376075745,
      "learning_rate": 9.413316159106941e-05,
      "loss": 0.8414,
      "mean_token_accuracy": 0.7571388334035873,
      "num_tokens": 61077399.0,
      "step": 1914
    },
    {
      "entropy": 0.8103114906698465,
      "epoch": 0.1761922013099407,
      "grad_norm": 0.15570461750030518,
      "learning_rate": 9.413009476492778e-05,
      "loss": 0.7824,
      "mean_token_accuracy": 0.7727453000843525,
      "num_tokens": 61109115.0,
      "step": 1915
    },
    {
      "entropy": 0.8294900190085173,
      "epoch": 0.17628420768138195,
      "grad_norm": 0.16319166123867035,
      "learning_rate": 9.412702793878615e-05,
      "loss": 0.8302,
      "mean_token_accuracy": 0.7586394585669041,
      "num_tokens": 61140368.0,
      "step": 1916
    },
    {
      "entropy": 0.9500578865408897,
      "epoch": 0.17637621405282317,
      "grad_norm": 0.16365797817707062,
      "learning_rate": 9.412396111264453e-05,
      "loss": 0.9468,
      "mean_token_accuracy": 0.7268388345837593,
      "num_tokens": 61172180.0,
      "step": 1917
    },
    {
      "entropy": 0.9131770320236683,
      "epoch": 0.17646822042426438,
      "grad_norm": 0.15897084772586823,
      "learning_rate": 9.41208942865029e-05,
      "loss": 0.9019,
      "mean_token_accuracy": 0.7448991909623146,
      "num_tokens": 61203652.0,
      "step": 1918
    },
    {
      "entropy": 0.8933025561273098,
      "epoch": 0.1765602267957056,
      "grad_norm": 0.156026229262352,
      "learning_rate": 9.411782746036128e-05,
      "loss": 0.8772,
      "mean_token_accuracy": 0.7492431998252869,
      "num_tokens": 61234705.0,
      "step": 1919
    },
    {
      "entropy": 0.9026918932795525,
      "epoch": 0.17665223316714682,
      "grad_norm": 0.1661609262228012,
      "learning_rate": 9.411476063421965e-05,
      "loss": 0.9121,
      "mean_token_accuracy": 0.7400776334106922,
      "num_tokens": 61265932.0,
      "step": 1920
    },
    {
      "entropy": 0.8438852876424789,
      "epoch": 0.17674423953858806,
      "grad_norm": 0.15465471148490906,
      "learning_rate": 9.411169380807803e-05,
      "loss": 0.837,
      "mean_token_accuracy": 0.7550399228930473,
      "num_tokens": 61298068.0,
      "step": 1921
    },
    {
      "entropy": 0.9316592924296856,
      "epoch": 0.17683624591002928,
      "grad_norm": 0.16610406339168549,
      "learning_rate": 9.41086269819364e-05,
      "loss": 0.9497,
      "mean_token_accuracy": 0.7349800132215023,
      "num_tokens": 61329958.0,
      "step": 1922
    },
    {
      "entropy": 0.8953296281397343,
      "epoch": 0.1769282522814705,
      "grad_norm": 0.1639757603406906,
      "learning_rate": 9.410556015579478e-05,
      "loss": 0.9198,
      "mean_token_accuracy": 0.7405243739485741,
      "num_tokens": 61362276.0,
      "step": 1923
    },
    {
      "entropy": 0.8736464381217957,
      "epoch": 0.1770202586529117,
      "grad_norm": 0.16952674090862274,
      "learning_rate": 9.410249332965316e-05,
      "loss": 0.8829,
      "mean_token_accuracy": 0.7433557249605656,
      "num_tokens": 61393645.0,
      "step": 1924
    },
    {
      "entropy": 0.802812235429883,
      "epoch": 0.17711226502435293,
      "grad_norm": 0.15684747695922852,
      "learning_rate": 9.409942650351153e-05,
      "loss": 0.8132,
      "mean_token_accuracy": 0.7672380842268467,
      "num_tokens": 61425927.0,
      "step": 1925
    },
    {
      "entropy": 0.9181756488978863,
      "epoch": 0.17720427139579417,
      "grad_norm": 0.1554405242204666,
      "learning_rate": 9.40963596773699e-05,
      "loss": 0.9147,
      "mean_token_accuracy": 0.7380742691457272,
      "num_tokens": 61457762.0,
      "step": 1926
    },
    {
      "entropy": 0.9950628988444805,
      "epoch": 0.1772962777672354,
      "grad_norm": 0.16184380650520325,
      "learning_rate": 9.409329285122826e-05,
      "loss": 0.9775,
      "mean_token_accuracy": 0.7249091789126396,
      "num_tokens": 61490018.0,
      "step": 1927
    },
    {
      "entropy": 0.9058981128036976,
      "epoch": 0.1773882841386766,
      "grad_norm": 0.16267764568328857,
      "learning_rate": 9.409022602508664e-05,
      "loss": 0.9041,
      "mean_token_accuracy": 0.737959947437048,
      "num_tokens": 61521500.0,
      "step": 1928
    },
    {
      "entropy": 0.8950675036758184,
      "epoch": 0.17748029051011782,
      "grad_norm": 0.1572149693965912,
      "learning_rate": 9.408715919894502e-05,
      "loss": 0.8991,
      "mean_token_accuracy": 0.7412764392793179,
      "num_tokens": 61552535.0,
      "step": 1929
    },
    {
      "entropy": 0.8722300231456757,
      "epoch": 0.17757229688155904,
      "grad_norm": 0.15969207882881165,
      "learning_rate": 9.408409237280339e-05,
      "loss": 0.8753,
      "mean_token_accuracy": 0.7523846551775932,
      "num_tokens": 61584629.0,
      "step": 1930
    },
    {
      "entropy": 0.8401013910770416,
      "epoch": 0.17766430325300028,
      "grad_norm": 0.15525026619434357,
      "learning_rate": 9.408102554666176e-05,
      "loss": 0.83,
      "mean_token_accuracy": 0.7585150972008705,
      "num_tokens": 61616563.0,
      "step": 1931
    },
    {
      "entropy": 0.9182925298810005,
      "epoch": 0.1777563096244415,
      "grad_norm": 0.16628873348236084,
      "learning_rate": 9.407795872052014e-05,
      "loss": 0.9264,
      "mean_token_accuracy": 0.7425150536000729,
      "num_tokens": 61648681.0,
      "step": 1932
    },
    {
      "entropy": 0.9674626514315605,
      "epoch": 0.17784831599588272,
      "grad_norm": 0.1649910807609558,
      "learning_rate": 9.407489189437851e-05,
      "loss": 0.9563,
      "mean_token_accuracy": 0.7306757904589176,
      "num_tokens": 61680109.0,
      "step": 1933
    },
    {
      "entropy": 0.883151214569807,
      "epoch": 0.17794032236732393,
      "grad_norm": 0.16164125502109528,
      "learning_rate": 9.407182506823689e-05,
      "loss": 0.8791,
      "mean_token_accuracy": 0.748978815972805,
      "num_tokens": 61712739.0,
      "step": 1934
    },
    {
      "entropy": 0.8568417485803366,
      "epoch": 0.17803232873876515,
      "grad_norm": 0.15556243062019348,
      "learning_rate": 9.406875824209526e-05,
      "loss": 0.8456,
      "mean_token_accuracy": 0.7571213878691196,
      "num_tokens": 61745052.0,
      "step": 1935
    },
    {
      "entropy": 0.7681448012590408,
      "epoch": 0.1781243351102064,
      "grad_norm": 0.15426810085773468,
      "learning_rate": 9.406569141595364e-05,
      "loss": 0.749,
      "mean_token_accuracy": 0.7827073112130165,
      "num_tokens": 61777132.0,
      "step": 1936
    },
    {
      "entropy": 0.8204676918685436,
      "epoch": 0.1782163414816476,
      "grad_norm": 0.1503227800130844,
      "learning_rate": 9.406262458981201e-05,
      "loss": 0.7951,
      "mean_token_accuracy": 0.7690162435173988,
      "num_tokens": 61809578.0,
      "step": 1937
    },
    {
      "entropy": 0.8179285116493702,
      "epoch": 0.17830834785308883,
      "grad_norm": 0.15755023062229156,
      "learning_rate": 9.405955776367038e-05,
      "loss": 0.8263,
      "mean_token_accuracy": 0.7617741972208023,
      "num_tokens": 61842085.0,
      "step": 1938
    },
    {
      "entropy": 0.8645897079259157,
      "epoch": 0.17840035422453004,
      "grad_norm": 0.15597712993621826,
      "learning_rate": 9.405649093752876e-05,
      "loss": 0.8647,
      "mean_token_accuracy": 0.7521870099008083,
      "num_tokens": 61873907.0,
      "step": 1939
    },
    {
      "entropy": 0.9659742340445518,
      "epoch": 0.17849236059597126,
      "grad_norm": 0.16275624930858612,
      "learning_rate": 9.405342411138714e-05,
      "loss": 0.9879,
      "mean_token_accuracy": 0.7224089317023754,
      "num_tokens": 61906491.0,
      "step": 1940
    },
    {
      "entropy": 0.8824424482882023,
      "epoch": 0.1785843669674125,
      "grad_norm": 0.16074663400650024,
      "learning_rate": 9.40503572852455e-05,
      "loss": 0.8624,
      "mean_token_accuracy": 0.7484971359372139,
      "num_tokens": 61938772.0,
      "step": 1941
    },
    {
      "entropy": 0.8316727019846439,
      "epoch": 0.17867637333885372,
      "grad_norm": 0.15254919230937958,
      "learning_rate": 9.404729045910387e-05,
      "loss": 0.8091,
      "mean_token_accuracy": 0.7678681574761868,
      "num_tokens": 61970845.0,
      "step": 1942
    },
    {
      "entropy": 0.8521139826625586,
      "epoch": 0.17876837971029494,
      "grad_norm": 0.1611201912164688,
      "learning_rate": 9.404422363296224e-05,
      "loss": 0.8412,
      "mean_token_accuracy": 0.761069193482399,
      "num_tokens": 62002875.0,
      "step": 1943
    },
    {
      "entropy": 0.848340617492795,
      "epoch": 0.17886038608173616,
      "grad_norm": 0.15629222989082336,
      "learning_rate": 9.404115680682062e-05,
      "loss": 0.8492,
      "mean_token_accuracy": 0.7606053091585636,
      "num_tokens": 62035351.0,
      "step": 1944
    },
    {
      "entropy": 0.8376112412661314,
      "epoch": 0.17895239245317737,
      "grad_norm": 0.16578906774520874,
      "learning_rate": 9.4038089980679e-05,
      "loss": 0.8564,
      "mean_token_accuracy": 0.7559154778718948,
      "num_tokens": 62067194.0,
      "step": 1945
    },
    {
      "entropy": 0.9160560555756092,
      "epoch": 0.17904439882461862,
      "grad_norm": 0.1683865338563919,
      "learning_rate": 9.403502315453737e-05,
      "loss": 0.9205,
      "mean_token_accuracy": 0.7384261265397072,
      "num_tokens": 62098528.0,
      "step": 1946
    },
    {
      "entropy": 0.8885866664350033,
      "epoch": 0.17913640519605983,
      "grad_norm": 0.16348950564861298,
      "learning_rate": 9.403195632839574e-05,
      "loss": 0.8796,
      "mean_token_accuracy": 0.7515899911522865,
      "num_tokens": 62130030.0,
      "step": 1947
    },
    {
      "entropy": 0.9445893652737141,
      "epoch": 0.17922841156750105,
      "grad_norm": 0.16221517324447632,
      "learning_rate": 9.402888950225412e-05,
      "loss": 0.9557,
      "mean_token_accuracy": 0.7316840030252934,
      "num_tokens": 62162156.0,
      "step": 1948
    },
    {
      "entropy": 0.8256821744143963,
      "epoch": 0.17932041793894227,
      "grad_norm": 0.16154508292675018,
      "learning_rate": 9.402582267611249e-05,
      "loss": 0.8302,
      "mean_token_accuracy": 0.7654169648885727,
      "num_tokens": 62194229.0,
      "step": 1949
    },
    {
      "entropy": 0.8002823200076818,
      "epoch": 0.17941242431038348,
      "grad_norm": 0.16087652742862701,
      "learning_rate": 9.402275584997087e-05,
      "loss": 0.8115,
      "mean_token_accuracy": 0.7657873369753361,
      "num_tokens": 62225406.0,
      "step": 1950
    },
    {
      "entropy": 0.8366348538547754,
      "epoch": 0.17950443068182473,
      "grad_norm": 0.1604553908109665,
      "learning_rate": 9.401968902382924e-05,
      "loss": 0.8341,
      "mean_token_accuracy": 0.7619588673114777,
      "num_tokens": 62257195.0,
      "step": 1951
    },
    {
      "entropy": 0.9393236972391605,
      "epoch": 0.17959643705326594,
      "grad_norm": 0.16409383714199066,
      "learning_rate": 9.401662219768762e-05,
      "loss": 0.9159,
      "mean_token_accuracy": 0.7402084767818451,
      "num_tokens": 62289490.0,
      "step": 1952
    },
    {
      "entropy": 0.8731848355382681,
      "epoch": 0.17968844342470716,
      "grad_norm": 0.15401241183280945,
      "learning_rate": 9.401355537154599e-05,
      "loss": 0.8713,
      "mean_token_accuracy": 0.7510554194450378,
      "num_tokens": 62321514.0,
      "step": 1953
    },
    {
      "entropy": 0.8477256558835506,
      "epoch": 0.17978044979614838,
      "grad_norm": 0.1566876620054245,
      "learning_rate": 9.401048854540436e-05,
      "loss": 0.8381,
      "mean_token_accuracy": 0.7641479410231113,
      "num_tokens": 62353283.0,
      "step": 1954
    },
    {
      "entropy": 0.953221982344985,
      "epoch": 0.1798724561675896,
      "grad_norm": 0.16360488533973694,
      "learning_rate": 9.400742171926275e-05,
      "loss": 0.9294,
      "mean_token_accuracy": 0.7360645607113838,
      "num_tokens": 62384867.0,
      "step": 1955
    },
    {
      "entropy": 0.9862979240715504,
      "epoch": 0.17996446253903084,
      "grad_norm": 0.17085519433021545,
      "learning_rate": 9.400435489312112e-05,
      "loss": 0.9875,
      "mean_token_accuracy": 0.7191058546304703,
      "num_tokens": 62416123.0,
      "step": 1956
    },
    {
      "entropy": 0.8249899782240391,
      "epoch": 0.18005646891047206,
      "grad_norm": 0.15461786091327667,
      "learning_rate": 9.400128806697949e-05,
      "loss": 0.8243,
      "mean_token_accuracy": 0.7621666416525841,
      "num_tokens": 62448753.0,
      "step": 1957
    },
    {
      "entropy": 0.8629487045109272,
      "epoch": 0.18014847528191327,
      "grad_norm": 0.15502046048641205,
      "learning_rate": 9.399822124083786e-05,
      "loss": 0.8588,
      "mean_token_accuracy": 0.757506225258112,
      "num_tokens": 62481003.0,
      "step": 1958
    },
    {
      "entropy": 0.8199736587703228,
      "epoch": 0.1802404816533545,
      "grad_norm": 0.16673530638217926,
      "learning_rate": 9.399515441469624e-05,
      "loss": 0.8343,
      "mean_token_accuracy": 0.7647030875086784,
      "num_tokens": 62512866.0,
      "step": 1959
    },
    {
      "entropy": 0.9487345237284899,
      "epoch": 0.1803324880247957,
      "grad_norm": 0.17352715134620667,
      "learning_rate": 9.399208758855462e-05,
      "loss": 0.9911,
      "mean_token_accuracy": 0.7245938181877136,
      "num_tokens": 62544707.0,
      "step": 1960
    },
    {
      "entropy": 0.9104325380176306,
      "epoch": 0.18042449439623695,
      "grad_norm": 0.16445976495742798,
      "learning_rate": 9.398902076241299e-05,
      "loss": 0.9318,
      "mean_token_accuracy": 0.7354235388338566,
      "num_tokens": 62577100.0,
      "step": 1961
    },
    {
      "entropy": 0.8994204178452492,
      "epoch": 0.18051650076767817,
      "grad_norm": 0.16265100240707397,
      "learning_rate": 9.398595393627135e-05,
      "loss": 0.8981,
      "mean_token_accuracy": 0.7433972097933292,
      "num_tokens": 62609302.0,
      "step": 1962
    },
    {
      "entropy": 0.92415926232934,
      "epoch": 0.18060850713911938,
      "grad_norm": 0.15947270393371582,
      "learning_rate": 9.398288711012974e-05,
      "loss": 0.9281,
      "mean_token_accuracy": 0.7294612862169743,
      "num_tokens": 62641268.0,
      "step": 1963
    },
    {
      "entropy": 0.9489542972296476,
      "epoch": 0.1807005135105606,
      "grad_norm": 0.1628192514181137,
      "learning_rate": 9.39798202839881e-05,
      "loss": 0.9529,
      "mean_token_accuracy": 0.7311035767197609,
      "num_tokens": 62672737.0,
      "step": 1964
    },
    {
      "entropy": 0.9142560213804245,
      "epoch": 0.18079251988200182,
      "grad_norm": 0.16272202134132385,
      "learning_rate": 9.397675345784648e-05,
      "loss": 0.8838,
      "mean_token_accuracy": 0.7449588216841221,
      "num_tokens": 62703883.0,
      "step": 1965
    },
    {
      "entropy": 0.9141163378953934,
      "epoch": 0.18088452625344306,
      "grad_norm": 0.15744638442993164,
      "learning_rate": 9.397368663170485e-05,
      "loss": 0.8942,
      "mean_token_accuracy": 0.7445085272192955,
      "num_tokens": 62735528.0,
      "step": 1966
    },
    {
      "entropy": 0.84230612590909,
      "epoch": 0.18097653262488428,
      "grad_norm": 0.1555042713880539,
      "learning_rate": 9.397061980556323e-05,
      "loss": 0.8246,
      "mean_token_accuracy": 0.7620725929737091,
      "num_tokens": 62767243.0,
      "step": 1967
    },
    {
      "entropy": 0.8740823641419411,
      "epoch": 0.1810685389963255,
      "grad_norm": 0.15724396705627441,
      "learning_rate": 9.39675529794216e-05,
      "loss": 0.8701,
      "mean_token_accuracy": 0.747488297522068,
      "num_tokens": 62799970.0,
      "step": 1968
    },
    {
      "entropy": 0.9185877647250891,
      "epoch": 0.1811605453677667,
      "grad_norm": 0.16185419261455536,
      "learning_rate": 9.396448615327997e-05,
      "loss": 0.9282,
      "mean_token_accuracy": 0.7406797222793102,
      "num_tokens": 62831989.0,
      "step": 1969
    },
    {
      "entropy": 0.9178366959095001,
      "epoch": 0.18125255173920793,
      "grad_norm": 0.1610131561756134,
      "learning_rate": 9.396141932713835e-05,
      "loss": 0.9274,
      "mean_token_accuracy": 0.7371709644794464,
      "num_tokens": 62863623.0,
      "step": 1970
    },
    {
      "entropy": 0.8601364158093929,
      "epoch": 0.18134455811064917,
      "grad_norm": 0.16155873239040375,
      "learning_rate": 9.395835250099673e-05,
      "loss": 0.8555,
      "mean_token_accuracy": 0.7587932161986828,
      "num_tokens": 62894881.0,
      "step": 1971
    },
    {
      "entropy": 0.8727019280195236,
      "epoch": 0.1814365644820904,
      "grad_norm": 0.15873903036117554,
      "learning_rate": 9.39552856748551e-05,
      "loss": 0.8513,
      "mean_token_accuracy": 0.7576525211334229,
      "num_tokens": 62927006.0,
      "step": 1972
    },
    {
      "entropy": 0.7951199021190405,
      "epoch": 0.1815285708535316,
      "grad_norm": 0.16684651374816895,
      "learning_rate": 9.395221884871347e-05,
      "loss": 0.7748,
      "mean_token_accuracy": 0.7764547653496265,
      "num_tokens": 62959195.0,
      "step": 1973
    },
    {
      "entropy": 0.8584694042801857,
      "epoch": 0.18162057722497282,
      "grad_norm": 0.15548956394195557,
      "learning_rate": 9.394915202257184e-05,
      "loss": 0.8638,
      "mean_token_accuracy": 0.7531874440610409,
      "num_tokens": 62990617.0,
      "step": 1974
    },
    {
      "entropy": 0.7681335564702749,
      "epoch": 0.18171258359641404,
      "grad_norm": 0.15133829414844513,
      "learning_rate": 9.394608519643022e-05,
      "loss": 0.7651,
      "mean_token_accuracy": 0.7752537503838539,
      "num_tokens": 63023001.0,
      "step": 1975
    },
    {
      "entropy": 0.8975860178470612,
      "epoch": 0.18180458996785528,
      "grad_norm": 0.1556512415409088,
      "learning_rate": 9.39430183702886e-05,
      "loss": 0.8956,
      "mean_token_accuracy": 0.7440182268619537,
      "num_tokens": 63055389.0,
      "step": 1976
    },
    {
      "entropy": 0.86626011505723,
      "epoch": 0.1818965963392965,
      "grad_norm": 0.16330760717391968,
      "learning_rate": 9.393995154414697e-05,
      "loss": 0.9035,
      "mean_token_accuracy": 0.7453521937131882,
      "num_tokens": 63087662.0,
      "step": 1977
    },
    {
      "entropy": 0.8465371988713741,
      "epoch": 0.18198860271073772,
      "grad_norm": 0.15979325771331787,
      "learning_rate": 9.393688471800533e-05,
      "loss": 0.8631,
      "mean_token_accuracy": 0.7496358975768089,
      "num_tokens": 63119080.0,
      "step": 1978
    },
    {
      "entropy": 0.7927393540740013,
      "epoch": 0.18208060908217893,
      "grad_norm": 0.15314452350139618,
      "learning_rate": 9.393381789186372e-05,
      "loss": 0.7972,
      "mean_token_accuracy": 0.7687264382839203,
      "num_tokens": 63150610.0,
      "step": 1979
    },
    {
      "entropy": 0.802070289850235,
      "epoch": 0.18217261545362015,
      "grad_norm": 0.15278708934783936,
      "learning_rate": 9.393075106572208e-05,
      "loss": 0.8192,
      "mean_token_accuracy": 0.7721150778234005,
      "num_tokens": 63182989.0,
      "step": 1980
    },
    {
      "entropy": 0.9702539928257465,
      "epoch": 0.1822646218250614,
      "grad_norm": 0.16206657886505127,
      "learning_rate": 9.392768423958047e-05,
      "loss": 0.9796,
      "mean_token_accuracy": 0.7228400334715843,
      "num_tokens": 63214901.0,
      "step": 1981
    },
    {
      "entropy": 0.7839300334453583,
      "epoch": 0.1823566281965026,
      "grad_norm": 0.1466144621372223,
      "learning_rate": 9.392461741343883e-05,
      "loss": 0.763,
      "mean_token_accuracy": 0.7762608975172043,
      "num_tokens": 63246976.0,
      "step": 1982
    },
    {
      "entropy": 0.7909334190189838,
      "epoch": 0.18244863456794383,
      "grad_norm": 0.14542445540428162,
      "learning_rate": 9.392155058729721e-05,
      "loss": 0.7808,
      "mean_token_accuracy": 0.7757250480353832,
      "num_tokens": 63279619.0,
      "step": 1983
    },
    {
      "entropy": 0.9851250760257244,
      "epoch": 0.18254064093938505,
      "grad_norm": 0.1624215692281723,
      "learning_rate": 9.391848376115558e-05,
      "loss": 0.9643,
      "mean_token_accuracy": 0.7262255065143108,
      "num_tokens": 63311159.0,
      "step": 1984
    },
    {
      "entropy": 0.8264614325016737,
      "epoch": 0.18263264731082626,
      "grad_norm": 0.15101690590381622,
      "learning_rate": 9.391541693501395e-05,
      "loss": 0.7999,
      "mean_token_accuracy": 0.7653401903808117,
      "num_tokens": 63342747.0,
      "step": 1985
    },
    {
      "entropy": 0.9237053617835045,
      "epoch": 0.1827246536822675,
      "grad_norm": 0.1684730350971222,
      "learning_rate": 9.391235010887233e-05,
      "loss": 0.9342,
      "mean_token_accuracy": 0.7342002280056477,
      "num_tokens": 63374678.0,
      "step": 1986
    },
    {
      "entropy": 0.868966493755579,
      "epoch": 0.18281666005370872,
      "grad_norm": 0.15601295232772827,
      "learning_rate": 9.390928328273071e-05,
      "loss": 0.8764,
      "mean_token_accuracy": 0.7515330202877522,
      "num_tokens": 63407049.0,
      "step": 1987
    },
    {
      "entropy": 0.819223340600729,
      "epoch": 0.18290866642514994,
      "grad_norm": 0.14965568482875824,
      "learning_rate": 9.390621645658908e-05,
      "loss": 0.8112,
      "mean_token_accuracy": 0.7661146633327007,
      "num_tokens": 63439259.0,
      "step": 1988
    },
    {
      "entropy": 0.7846616264432669,
      "epoch": 0.18300067279659116,
      "grad_norm": 0.1548296958208084,
      "learning_rate": 9.390314963044745e-05,
      "loss": 0.7909,
      "mean_token_accuracy": 0.7720412723720074,
      "num_tokens": 63471431.0,
      "step": 1989
    },
    {
      "entropy": 0.8260967396199703,
      "epoch": 0.18309267916803237,
      "grad_norm": 0.16107745468616486,
      "learning_rate": 9.390008280430583e-05,
      "loss": 0.8054,
      "mean_token_accuracy": 0.7688294462859631,
      "num_tokens": 63503717.0,
      "step": 1990
    },
    {
      "entropy": 0.834751795977354,
      "epoch": 0.18318468553947362,
      "grad_norm": 0.15310858190059662,
      "learning_rate": 9.38970159781642e-05,
      "loss": 0.8287,
      "mean_token_accuracy": 0.7624571397900581,
      "num_tokens": 63535664.0,
      "step": 1991
    },
    {
      "entropy": 0.9050851464271545,
      "epoch": 0.18327669191091484,
      "grad_norm": 0.15908728539943695,
      "learning_rate": 9.389394915202258e-05,
      "loss": 0.8852,
      "mean_token_accuracy": 0.7468139603734016,
      "num_tokens": 63567527.0,
      "step": 1992
    },
    {
      "entropy": 0.9065479021519423,
      "epoch": 0.18336869828235605,
      "grad_norm": 0.15439389646053314,
      "learning_rate": 9.389088232588095e-05,
      "loss": 0.8946,
      "mean_token_accuracy": 0.7443993464112282,
      "num_tokens": 63600070.0,
      "step": 1993
    },
    {
      "entropy": 0.9310688227415085,
      "epoch": 0.18346070465379727,
      "grad_norm": 0.1685931235551834,
      "learning_rate": 9.388781549973933e-05,
      "loss": 0.9321,
      "mean_token_accuracy": 0.7354470491409302,
      "num_tokens": 63630583.0,
      "step": 1994
    },
    {
      "entropy": 0.8912995792925358,
      "epoch": 0.18355271102523849,
      "grad_norm": 0.22048255801200867,
      "learning_rate": 9.38847486735977e-05,
      "loss": 0.8999,
      "mean_token_accuracy": 0.7403712943196297,
      "num_tokens": 63662610.0,
      "step": 1995
    },
    {
      "entropy": 0.9158200062811375,
      "epoch": 0.18364471739667973,
      "grad_norm": 0.16192208230495453,
      "learning_rate": 9.388168184745606e-05,
      "loss": 0.9381,
      "mean_token_accuracy": 0.7375936694443226,
      "num_tokens": 63695028.0,
      "step": 1996
    },
    {
      "entropy": 0.8296115919947624,
      "epoch": 0.18373672376812095,
      "grad_norm": 0.1535319685935974,
      "learning_rate": 9.387861502131445e-05,
      "loss": 0.8274,
      "mean_token_accuracy": 0.7615524679422379,
      "num_tokens": 63727738.0,
      "step": 1997
    },
    {
      "entropy": 0.8706228844821453,
      "epoch": 0.18382873013956216,
      "grad_norm": 0.15712161362171173,
      "learning_rate": 9.387554819517283e-05,
      "loss": 0.8708,
      "mean_token_accuracy": 0.7500491812825203,
      "num_tokens": 63759036.0,
      "step": 1998
    },
    {
      "entropy": 0.8559218496084213,
      "epoch": 0.18392073651100338,
      "grad_norm": 0.15609563887119293,
      "learning_rate": 9.38724813690312e-05,
      "loss": 0.8553,
      "mean_token_accuracy": 0.7534160315990448,
      "num_tokens": 63791226.0,
      "step": 1999
    },
    {
      "entropy": 0.8668279834091663,
      "epoch": 0.1840127428824446,
      "grad_norm": 0.15756767988204956,
      "learning_rate": 9.386941454288956e-05,
      "loss": 0.8549,
      "mean_token_accuracy": 0.751496221870184,
      "num_tokens": 63822639.0,
      "step": 2000
    },
    {
      "entropy": 0.9212057013064623,
      "epoch": 0.18410474925388584,
      "grad_norm": 0.16219118237495422,
      "learning_rate": 9.386634771674793e-05,
      "loss": 0.92,
      "mean_token_accuracy": 0.7382799237966537,
      "num_tokens": 63853556.0,
      "step": 2001
    },
    {
      "entropy": 0.7749503999948502,
      "epoch": 0.18419675562532706,
      "grad_norm": 0.15189917385578156,
      "learning_rate": 9.386328089060633e-05,
      "loss": 0.7728,
      "mean_token_accuracy": 0.7693426348268986,
      "num_tokens": 63885584.0,
      "step": 2002
    },
    {
      "entropy": 0.7660795710980892,
      "epoch": 0.18428876199676827,
      "grad_norm": 0.15474411845207214,
      "learning_rate": 9.38602140644647e-05,
      "loss": 0.7498,
      "mean_token_accuracy": 0.7790898159146309,
      "num_tokens": 63917944.0,
      "step": 2003
    },
    {
      "entropy": 0.8617585264146328,
      "epoch": 0.1843807683682095,
      "grad_norm": 0.16178357601165771,
      "learning_rate": 9.385714723832306e-05,
      "loss": 0.8691,
      "mean_token_accuracy": 0.7499283477663994,
      "num_tokens": 63949828.0,
      "step": 2004
    },
    {
      "entropy": 0.9263352155685425,
      "epoch": 0.1844727747396507,
      "grad_norm": 0.1602795422077179,
      "learning_rate": 9.385408041218143e-05,
      "loss": 0.9171,
      "mean_token_accuracy": 0.737714622169733,
      "num_tokens": 63981775.0,
      "step": 2005
    },
    {
      "entropy": 0.8505771532654762,
      "epoch": 0.18456478111109195,
      "grad_norm": 0.15818481147289276,
      "learning_rate": 9.385101358603981e-05,
      "loss": 0.8638,
      "mean_token_accuracy": 0.7530789375305176,
      "num_tokens": 64013962.0,
      "step": 2006
    },
    {
      "entropy": 0.9102665297687054,
      "epoch": 0.18465678748253317,
      "grad_norm": 0.16169792413711548,
      "learning_rate": 9.384794675989819e-05,
      "loss": 0.9135,
      "mean_token_accuracy": 0.7367063835263252,
      "num_tokens": 64045703.0,
      "step": 2007
    },
    {
      "entropy": 0.841713935136795,
      "epoch": 0.18474879385397439,
      "grad_norm": 0.15355545282363892,
      "learning_rate": 9.384487993375656e-05,
      "loss": 0.8317,
      "mean_token_accuracy": 0.75917012616992,
      "num_tokens": 64077171.0,
      "step": 2008
    },
    {
      "entropy": 0.926487859338522,
      "epoch": 0.1848408002254156,
      "grad_norm": 0.15862159430980682,
      "learning_rate": 9.384181310761493e-05,
      "loss": 0.9289,
      "mean_token_accuracy": 0.7370729744434357,
      "num_tokens": 64108900.0,
      "step": 2009
    },
    {
      "entropy": 0.889353197067976,
      "epoch": 0.18493280659685682,
      "grad_norm": 0.16286253929138184,
      "learning_rate": 9.383874628147331e-05,
      "loss": 0.9187,
      "mean_token_accuracy": 0.7435460090637207,
      "num_tokens": 64141547.0,
      "step": 2010
    },
    {
      "entropy": 0.8953009396791458,
      "epoch": 0.18502481296829806,
      "grad_norm": 0.15893928706645966,
      "learning_rate": 9.383567945533168e-05,
      "loss": 0.8839,
      "mean_token_accuracy": 0.7446173205971718,
      "num_tokens": 64172941.0,
      "step": 2011
    },
    {
      "entropy": 0.8864428587257862,
      "epoch": 0.18511681933973928,
      "grad_norm": 0.1566944420337677,
      "learning_rate": 9.383261262919006e-05,
      "loss": 0.8762,
      "mean_token_accuracy": 0.7484751604497433,
      "num_tokens": 64204959.0,
      "step": 2012
    },
    {
      "entropy": 0.804668266326189,
      "epoch": 0.1852088257111805,
      "grad_norm": 0.1523744910955429,
      "learning_rate": 9.382954580304843e-05,
      "loss": 0.797,
      "mean_token_accuracy": 0.7658344246447086,
      "num_tokens": 64237067.0,
      "step": 2013
    },
    {
      "entropy": 0.7890373449772596,
      "epoch": 0.1853008320826217,
      "grad_norm": 0.15298423171043396,
      "learning_rate": 9.382647897690681e-05,
      "loss": 0.7738,
      "mean_token_accuracy": 0.779646098613739,
      "num_tokens": 64269076.0,
      "step": 2014
    },
    {
      "entropy": 0.8391816355288029,
      "epoch": 0.18539283845406293,
      "grad_norm": 0.156507208943367,
      "learning_rate": 9.382341215076518e-05,
      "loss": 0.8507,
      "mean_token_accuracy": 0.7547506541013718,
      "num_tokens": 64301240.0,
      "step": 2015
    },
    {
      "entropy": 0.7176179699599743,
      "epoch": 0.18548484482550417,
      "grad_norm": 0.14936517179012299,
      "learning_rate": 9.382034532462354e-05,
      "loss": 0.6999,
      "mean_token_accuracy": 0.7936183474957943,
      "num_tokens": 64333834.0,
      "step": 2016
    },
    {
      "entropy": 0.9070466496050358,
      "epoch": 0.1855768511969454,
      "grad_norm": 0.16041108965873718,
      "learning_rate": 9.381727849848193e-05,
      "loss": 0.8996,
      "mean_token_accuracy": 0.7438365668058395,
      "num_tokens": 64365959.0,
      "step": 2017
    },
    {
      "entropy": 0.8404087126255035,
      "epoch": 0.1856688575683866,
      "grad_norm": 0.16266557574272156,
      "learning_rate": 9.381421167234031e-05,
      "loss": 0.8547,
      "mean_token_accuracy": 0.7599496878683567,
      "num_tokens": 64398497.0,
      "step": 2018
    },
    {
      "entropy": 0.8685179185122252,
      "epoch": 0.18576086393982782,
      "grad_norm": 0.15677061676979065,
      "learning_rate": 9.381114484619867e-05,
      "loss": 0.8476,
      "mean_token_accuracy": 0.7601738162338734,
      "num_tokens": 64430302.0,
      "step": 2019
    },
    {
      "entropy": 0.8801700137555599,
      "epoch": 0.18585287031126904,
      "grad_norm": 0.1587846726179123,
      "learning_rate": 9.380807802005704e-05,
      "loss": 0.8738,
      "mean_token_accuracy": 0.7542241215705872,
      "num_tokens": 64461727.0,
      "step": 2020
    },
    {
      "entropy": 0.8801525477319956,
      "epoch": 0.1859448766827103,
      "grad_norm": 0.16184459626674652,
      "learning_rate": 9.380501119391542e-05,
      "loss": 0.8692,
      "mean_token_accuracy": 0.7533211074769497,
      "num_tokens": 64494136.0,
      "step": 2021
    },
    {
      "entropy": 0.8940998762845993,
      "epoch": 0.1860368830541515,
      "grad_norm": 0.16499122977256775,
      "learning_rate": 9.380194436777379e-05,
      "loss": 0.8993,
      "mean_token_accuracy": 0.7386087365448475,
      "num_tokens": 64525163.0,
      "step": 2022
    },
    {
      "entropy": 0.8294282592833042,
      "epoch": 0.18612888942559272,
      "grad_norm": 0.15808042883872986,
      "learning_rate": 9.379887754163217e-05,
      "loss": 0.8307,
      "mean_token_accuracy": 0.7567545399069786,
      "num_tokens": 64557009.0,
      "step": 2023
    },
    {
      "entropy": 0.7610705997794867,
      "epoch": 0.18622089579703394,
      "grad_norm": 0.15707454085350037,
      "learning_rate": 9.379581071549054e-05,
      "loss": 0.7777,
      "mean_token_accuracy": 0.7748976908624172,
      "num_tokens": 64589221.0,
      "step": 2024
    },
    {
      "entropy": 0.771922692656517,
      "epoch": 0.18631290216847518,
      "grad_norm": 0.15798892080783844,
      "learning_rate": 9.379274388934892e-05,
      "loss": 0.7958,
      "mean_token_accuracy": 0.7745271436870098,
      "num_tokens": 64621443.0,
      "step": 2025
    },
    {
      "entropy": 0.8313932530581951,
      "epoch": 0.1864049085399164,
      "grad_norm": 0.16411079466342926,
      "learning_rate": 9.378967706320729e-05,
      "loss": 0.8104,
      "mean_token_accuracy": 0.7729224562644958,
      "num_tokens": 64653792.0,
      "step": 2026
    },
    {
      "entropy": 0.8748119957745075,
      "epoch": 0.18649691491135761,
      "grad_norm": 0.15752798318862915,
      "learning_rate": 9.378661023706566e-05,
      "loss": 0.8492,
      "mean_token_accuracy": 0.758167389780283,
      "num_tokens": 64685363.0,
      "step": 2027
    },
    {
      "entropy": 0.9634877406060696,
      "epoch": 0.18658892128279883,
      "grad_norm": 0.1569027304649353,
      "learning_rate": 9.378354341092404e-05,
      "loss": 0.947,
      "mean_token_accuracy": 0.7322870679199696,
      "num_tokens": 64717788.0,
      "step": 2028
    },
    {
      "entropy": 0.8881736323237419,
      "epoch": 0.18668092765424005,
      "grad_norm": 0.15792250633239746,
      "learning_rate": 9.378047658478242e-05,
      "loss": 0.8775,
      "mean_token_accuracy": 0.7494480460882187,
      "num_tokens": 64749228.0,
      "step": 2029
    },
    {
      "entropy": 0.953178521245718,
      "epoch": 0.1867729340256813,
      "grad_norm": 0.1602461338043213,
      "learning_rate": 9.377740975864079e-05,
      "loss": 0.9203,
      "mean_token_accuracy": 0.7399674914777279,
      "num_tokens": 64781497.0,
      "step": 2030
    },
    {
      "entropy": 0.9327591173350811,
      "epoch": 0.1868649403971225,
      "grad_norm": 0.1633158028125763,
      "learning_rate": 9.377434293249916e-05,
      "loss": 0.9441,
      "mean_token_accuracy": 0.7319155223667622,
      "num_tokens": 64812619.0,
      "step": 2031
    },
    {
      "entropy": 0.8882357068359852,
      "epoch": 0.18695694676856373,
      "grad_norm": 0.1570863574743271,
      "learning_rate": 9.377127610635752e-05,
      "loss": 0.8708,
      "mean_token_accuracy": 0.7506612278521061,
      "num_tokens": 64845060.0,
      "step": 2032
    },
    {
      "entropy": 0.8624433614313602,
      "epoch": 0.18704895314000494,
      "grad_norm": 0.1596759408712387,
      "learning_rate": 9.37682092802159e-05,
      "loss": 0.869,
      "mean_token_accuracy": 0.7522710748016834,
      "num_tokens": 64876329.0,
      "step": 2033
    },
    {
      "entropy": 0.9025580398738384,
      "epoch": 0.18714095951144616,
      "grad_norm": 0.15871481597423553,
      "learning_rate": 9.376514245407429e-05,
      "loss": 0.8883,
      "mean_token_accuracy": 0.7478687763214111,
      "num_tokens": 64907958.0,
      "step": 2034
    },
    {
      "entropy": 0.8581497929990292,
      "epoch": 0.1872329658828874,
      "grad_norm": 0.157855823636055,
      "learning_rate": 9.376207562793266e-05,
      "loss": 0.8731,
      "mean_token_accuracy": 0.7531143128871918,
      "num_tokens": 64940726.0,
      "step": 2035
    },
    {
      "entropy": 0.9045156873762608,
      "epoch": 0.18732497225432862,
      "grad_norm": 0.16071809828281403,
      "learning_rate": 9.375900880179102e-05,
      "loss": 0.9337,
      "mean_token_accuracy": 0.7386736422777176,
      "num_tokens": 64972580.0,
      "step": 2036
    },
    {
      "entropy": 0.9036298952996731,
      "epoch": 0.18741697862576984,
      "grad_norm": 0.16352811455726624,
      "learning_rate": 9.37559419756494e-05,
      "loss": 0.9348,
      "mean_token_accuracy": 0.7315878421068192,
      "num_tokens": 65004524.0,
      "step": 2037
    },
    {
      "entropy": 0.8559081815183163,
      "epoch": 0.18750898499721105,
      "grad_norm": 0.16810262203216553,
      "learning_rate": 9.375287514950777e-05,
      "loss": 0.8889,
      "mean_token_accuracy": 0.7461729347705841,
      "num_tokens": 65036454.0,
      "step": 2038
    },
    {
      "entropy": 0.9171912744641304,
      "epoch": 0.18760099136865227,
      "grad_norm": 0.16522151231765747,
      "learning_rate": 9.374980832336615e-05,
      "loss": 0.9429,
      "mean_token_accuracy": 0.7354200817644596,
      "num_tokens": 65069000.0,
      "step": 2039
    },
    {
      "entropy": 0.8885633610188961,
      "epoch": 0.18769299774009351,
      "grad_norm": 0.16042226552963257,
      "learning_rate": 9.374674149722452e-05,
      "loss": 0.8715,
      "mean_token_accuracy": 0.7541116811335087,
      "num_tokens": 65101290.0,
      "step": 2040
    },
    {
      "entropy": 0.8812004756182432,
      "epoch": 0.18778500411153473,
      "grad_norm": 0.15915481746196747,
      "learning_rate": 9.37436746710829e-05,
      "loss": 0.8804,
      "mean_token_accuracy": 0.7449415810406208,
      "num_tokens": 65133413.0,
      "step": 2041
    },
    {
      "entropy": 0.9142667725682259,
      "epoch": 0.18787701048297595,
      "grad_norm": 0.16686473786830902,
      "learning_rate": 9.374060784494127e-05,
      "loss": 0.8904,
      "mean_token_accuracy": 0.7472209744155407,
      "num_tokens": 65164773.0,
      "step": 2042
    },
    {
      "entropy": 0.8553497269749641,
      "epoch": 0.18796901685441716,
      "grad_norm": 0.15235202014446259,
      "learning_rate": 9.373754101879965e-05,
      "loss": 0.8345,
      "mean_token_accuracy": 0.7613614685833454,
      "num_tokens": 65197342.0,
      "step": 2043
    },
    {
      "entropy": 0.9114524945616722,
      "epoch": 0.18806102322585838,
      "grad_norm": 0.15685681998729706,
      "learning_rate": 9.373447419265802e-05,
      "loss": 0.9143,
      "mean_token_accuracy": 0.7399330735206604,
      "num_tokens": 65229364.0,
      "step": 2044
    },
    {
      "entropy": 0.893359798938036,
      "epoch": 0.18815302959729963,
      "grad_norm": 0.16052910685539246,
      "learning_rate": 9.37314073665164e-05,
      "loss": 0.9065,
      "mean_token_accuracy": 0.7452050186693668,
      "num_tokens": 65261563.0,
      "step": 2045
    },
    {
      "entropy": 0.9454722385853529,
      "epoch": 0.18824503596874084,
      "grad_norm": 0.16674014925956726,
      "learning_rate": 9.372834054037477e-05,
      "loss": 0.9533,
      "mean_token_accuracy": 0.7334141582250595,
      "num_tokens": 65292981.0,
      "step": 2046
    },
    {
      "entropy": 0.8630388602614403,
      "epoch": 0.18833704234018206,
      "grad_norm": 0.15543870627880096,
      "learning_rate": 9.372527371423314e-05,
      "loss": 0.8663,
      "mean_token_accuracy": 0.754218615591526,
      "num_tokens": 65325233.0,
      "step": 2047
    },
    {
      "entropy": 0.8885634876787663,
      "epoch": 0.18842904871162328,
      "grad_norm": 0.15675462782382965,
      "learning_rate": 9.372220688809152e-05,
      "loss": 0.8899,
      "mean_token_accuracy": 0.7503672540187836,
      "num_tokens": 65357635.0,
      "step": 2048
    },
    {
      "entropy": 0.9084977805614471,
      "epoch": 0.1885210550830645,
      "grad_norm": 0.15942583978176117,
      "learning_rate": 9.37191400619499e-05,
      "loss": 0.915,
      "mean_token_accuracy": 0.7430415004491806,
      "num_tokens": 65389581.0,
      "step": 2049
    },
    {
      "entropy": 0.8415555544197559,
      "epoch": 0.18861306145450574,
      "grad_norm": 0.15190312266349792,
      "learning_rate": 9.371607323580827e-05,
      "loss": 0.8312,
      "mean_token_accuracy": 0.7616266831755638,
      "num_tokens": 65421699.0,
      "step": 2050
    },
    {
      "entropy": 0.7941470611840487,
      "epoch": 0.18870506782594695,
      "grad_norm": 0.15160848200321198,
      "learning_rate": 9.371300640966664e-05,
      "loss": 0.7695,
      "mean_token_accuracy": 0.7729646824300289,
      "num_tokens": 65453061.0,
      "step": 2051
    },
    {
      "entropy": 0.8557648323476315,
      "epoch": 0.18879707419738817,
      "grad_norm": 0.1504427194595337,
      "learning_rate": 9.370993958352502e-05,
      "loss": 0.8381,
      "mean_token_accuracy": 0.7609410546720028,
      "num_tokens": 65485194.0,
      "step": 2052
    },
    {
      "entropy": 0.891657717525959,
      "epoch": 0.1888890805688294,
      "grad_norm": 0.1631632000207901,
      "learning_rate": 9.370687275738339e-05,
      "loss": 0.8755,
      "mean_token_accuracy": 0.7546772211790085,
      "num_tokens": 65515973.0,
      "step": 2053
    },
    {
      "entropy": 0.8462304472923279,
      "epoch": 0.1889810869402706,
      "grad_norm": 0.15299326181411743,
      "learning_rate": 9.370380593124177e-05,
      "loss": 0.8333,
      "mean_token_accuracy": 0.7614996656775475,
      "num_tokens": 65547531.0,
      "step": 2054
    },
    {
      "entropy": 0.8240957818925381,
      "epoch": 0.18907309331171185,
      "grad_norm": 0.14741098880767822,
      "learning_rate": 9.370073910510013e-05,
      "loss": 0.8186,
      "mean_token_accuracy": 0.7635062858462334,
      "num_tokens": 65579853.0,
      "step": 2055
    },
    {
      "entropy": 0.8090843558311462,
      "epoch": 0.18916509968315307,
      "grad_norm": 0.15631279349327087,
      "learning_rate": 9.369767227895852e-05,
      "loss": 0.8146,
      "mean_token_accuracy": 0.7656925283372402,
      "num_tokens": 65612208.0,
      "step": 2056
    },
    {
      "entropy": 0.7881220206618309,
      "epoch": 0.18925710605459428,
      "grad_norm": 0.16245120763778687,
      "learning_rate": 9.369460545281688e-05,
      "loss": 0.8054,
      "mean_token_accuracy": 0.767612311989069,
      "num_tokens": 65644181.0,
      "step": 2057
    },
    {
      "entropy": 0.9310653619468212,
      "epoch": 0.1893491124260355,
      "grad_norm": 0.16227516531944275,
      "learning_rate": 9.369153862667525e-05,
      "loss": 0.9252,
      "mean_token_accuracy": 0.7376483865082264,
      "num_tokens": 65676166.0,
      "step": 2058
    },
    {
      "entropy": 0.90697106346488,
      "epoch": 0.18944111879747672,
      "grad_norm": 0.16005201637744904,
      "learning_rate": 9.368847180053363e-05,
      "loss": 0.8929,
      "mean_token_accuracy": 0.7444347888231277,
      "num_tokens": 65707727.0,
      "step": 2059
    },
    {
      "entropy": 0.83125857450068,
      "epoch": 0.18953312516891796,
      "grad_norm": 0.15063901245594025,
      "learning_rate": 9.368540497439202e-05,
      "loss": 0.8353,
      "mean_token_accuracy": 0.7557453885674477,
      "num_tokens": 65740071.0,
      "step": 2060
    },
    {
      "entropy": 0.9123864956200123,
      "epoch": 0.18962513154035918,
      "grad_norm": 0.1682862490415573,
      "learning_rate": 9.368233814825038e-05,
      "loss": 0.922,
      "mean_token_accuracy": 0.7390746884047985,
      "num_tokens": 65772177.0,
      "step": 2061
    },
    {
      "entropy": 0.8019394800066948,
      "epoch": 0.1897171379118004,
      "grad_norm": 0.15979711711406708,
      "learning_rate": 9.367927132210875e-05,
      "loss": 0.8201,
      "mean_token_accuracy": 0.7626630812883377,
      "num_tokens": 65804576.0,
      "step": 2062
    },
    {
      "entropy": 0.9125566072762012,
      "epoch": 0.1898091442832416,
      "grad_norm": 0.16224896907806396,
      "learning_rate": 9.367620449596712e-05,
      "loss": 0.9098,
      "mean_token_accuracy": 0.7389998286962509,
      "num_tokens": 65836514.0,
      "step": 2063
    },
    {
      "entropy": 0.9041921626776457,
      "epoch": 0.18990115065468283,
      "grad_norm": 0.1619517207145691,
      "learning_rate": 9.36731376698255e-05,
      "loss": 0.8964,
      "mean_token_accuracy": 0.7419093325734138,
      "num_tokens": 65867852.0,
      "step": 2064
    },
    {
      "entropy": 0.8438561335206032,
      "epoch": 0.18999315702612407,
      "grad_norm": 0.1557321846485138,
      "learning_rate": 9.367007084368388e-05,
      "loss": 0.8432,
      "mean_token_accuracy": 0.7606022581458092,
      "num_tokens": 65899536.0,
      "step": 2065
    },
    {
      "entropy": 0.7718838583678007,
      "epoch": 0.1900851633975653,
      "grad_norm": 0.1508084237575531,
      "learning_rate": 9.366700401754225e-05,
      "loss": 0.7561,
      "mean_token_accuracy": 0.7831364944577217,
      "num_tokens": 65931457.0,
      "step": 2066
    },
    {
      "entropy": 0.903154619038105,
      "epoch": 0.1901771697690065,
      "grad_norm": 0.16352686285972595,
      "learning_rate": 9.366393719140062e-05,
      "loss": 0.8826,
      "mean_token_accuracy": 0.749750267714262,
      "num_tokens": 65963647.0,
      "step": 2067
    },
    {
      "entropy": 0.8343342915177345,
      "epoch": 0.19026917614044772,
      "grad_norm": 0.1577976942062378,
      "learning_rate": 9.3660870365259e-05,
      "loss": 0.848,
      "mean_token_accuracy": 0.7582064867019653,
      "num_tokens": 65995798.0,
      "step": 2068
    },
    {
      "entropy": 0.8333789817988873,
      "epoch": 0.19036118251188894,
      "grad_norm": 0.1567251980304718,
      "learning_rate": 9.365780353911737e-05,
      "loss": 0.8383,
      "mean_token_accuracy": 0.7609939053654671,
      "num_tokens": 66028039.0,
      "step": 2069
    },
    {
      "entropy": 0.8925304748117924,
      "epoch": 0.19045318888333018,
      "grad_norm": 0.16208332777023315,
      "learning_rate": 9.365473671297575e-05,
      "loss": 0.8964,
      "mean_token_accuracy": 0.7444371208548546,
      "num_tokens": 66059910.0,
      "step": 2070
    },
    {
      "entropy": 0.9667200781404972,
      "epoch": 0.1905451952547714,
      "grad_norm": 0.16504539549350739,
      "learning_rate": 9.365166988683412e-05,
      "loss": 0.9628,
      "mean_token_accuracy": 0.729257021099329,
      "num_tokens": 66091502.0,
      "step": 2071
    },
    {
      "entropy": 0.9035168662667274,
      "epoch": 0.19063720162621262,
      "grad_norm": 0.16108198463916779,
      "learning_rate": 9.36486030606925e-05,
      "loss": 0.8965,
      "mean_token_accuracy": 0.7464965023100376,
      "num_tokens": 66123242.0,
      "step": 2072
    },
    {
      "entropy": 0.8675672821700573,
      "epoch": 0.19072920799765383,
      "grad_norm": 0.16194505989551544,
      "learning_rate": 9.364553623455087e-05,
      "loss": 0.8745,
      "mean_token_accuracy": 0.7500729374587536,
      "num_tokens": 66154205.0,
      "step": 2073
    },
    {
      "entropy": 0.9420213140547276,
      "epoch": 0.19082121436909505,
      "grad_norm": 0.1630868911743164,
      "learning_rate": 9.364246940840923e-05,
      "loss": 0.9398,
      "mean_token_accuracy": 0.7348047271370888,
      "num_tokens": 66186523.0,
      "step": 2074
    },
    {
      "entropy": 0.9270667694509029,
      "epoch": 0.1909132207405363,
      "grad_norm": 0.16062036156654358,
      "learning_rate": 9.363940258226761e-05,
      "loss": 0.9097,
      "mean_token_accuracy": 0.7415963970124722,
      "num_tokens": 66218573.0,
      "step": 2075
    },
    {
      "entropy": 0.8215692602097988,
      "epoch": 0.1910052271119775,
      "grad_norm": 0.1590689718723297,
      "learning_rate": 9.3636335756126e-05,
      "loss": 0.8232,
      "mean_token_accuracy": 0.7622773535549641,
      "num_tokens": 66251275.0,
      "step": 2076
    },
    {
      "entropy": 0.795189144089818,
      "epoch": 0.19109723348341873,
      "grad_norm": 0.16071492433547974,
      "learning_rate": 9.363326892998436e-05,
      "loss": 0.7856,
      "mean_token_accuracy": 0.7736237049102783,
      "num_tokens": 66283173.0,
      "step": 2077
    },
    {
      "entropy": 0.9944587536156178,
      "epoch": 0.19118923985485994,
      "grad_norm": 0.16564224660396576,
      "learning_rate": 9.363020210384273e-05,
      "loss": 1.0009,
      "mean_token_accuracy": 0.7175012454390526,
      "num_tokens": 66314753.0,
      "step": 2078
    },
    {
      "entropy": 0.7731790505349636,
      "epoch": 0.19128124622630116,
      "grad_norm": 0.15580476820468903,
      "learning_rate": 9.362713527770111e-05,
      "loss": 0.7721,
      "mean_token_accuracy": 0.7750517874956131,
      "num_tokens": 66347090.0,
      "step": 2079
    },
    {
      "entropy": 0.9598245583474636,
      "epoch": 0.1913732525977424,
      "grad_norm": 0.16624146699905396,
      "learning_rate": 9.36240684515595e-05,
      "loss": 0.9777,
      "mean_token_accuracy": 0.7278884015977383,
      "num_tokens": 66378364.0,
      "step": 2080
    },
    {
      "entropy": 0.9357683062553406,
      "epoch": 0.19146525896918362,
      "grad_norm": 0.1717107743024826,
      "learning_rate": 9.362100162541786e-05,
      "loss": 0.954,
      "mean_token_accuracy": 0.734722763299942,
      "num_tokens": 66409963.0,
      "step": 2081
    },
    {
      "entropy": 0.9204175174236298,
      "epoch": 0.19155726534062484,
      "grad_norm": 0.17001405358314514,
      "learning_rate": 9.361793479927623e-05,
      "loss": 0.9321,
      "mean_token_accuracy": 0.7355771027505398,
      "num_tokens": 66441139.0,
      "step": 2082
    },
    {
      "entropy": 0.9573743641376495,
      "epoch": 0.19164927171206605,
      "grad_norm": 0.16127470135688782,
      "learning_rate": 9.361486797313461e-05,
      "loss": 0.9234,
      "mean_token_accuracy": 0.7395357266068459,
      "num_tokens": 66472796.0,
      "step": 2083
    },
    {
      "entropy": 0.8810067437589169,
      "epoch": 0.19174127808350727,
      "grad_norm": 0.15842416882514954,
      "learning_rate": 9.361180114699298e-05,
      "loss": 0.8777,
      "mean_token_accuracy": 0.7571773789823055,
      "num_tokens": 66505345.0,
      "step": 2084
    },
    {
      "entropy": 0.8114758655428886,
      "epoch": 0.19183328445494852,
      "grad_norm": 0.15940213203430176,
      "learning_rate": 9.360873432085136e-05,
      "loss": 0.8071,
      "mean_token_accuracy": 0.7702943459153175,
      "num_tokens": 66537348.0,
      "step": 2085
    },
    {
      "entropy": 0.836161045357585,
      "epoch": 0.19192529082638973,
      "grad_norm": 0.15598586201667786,
      "learning_rate": 9.360566749470973e-05,
      "loss": 0.8142,
      "mean_token_accuracy": 0.7635610811412334,
      "num_tokens": 66569438.0,
      "step": 2086
    },
    {
      "entropy": 0.9076887965202332,
      "epoch": 0.19201729719783095,
      "grad_norm": 0.16632279753684998,
      "learning_rate": 9.360260066856811e-05,
      "loss": 0.9201,
      "mean_token_accuracy": 0.7434936538338661,
      "num_tokens": 66601333.0,
      "step": 2087
    },
    {
      "entropy": 0.8718653135001659,
      "epoch": 0.19210930356927217,
      "grad_norm": 0.159499853849411,
      "learning_rate": 9.359953384242648e-05,
      "loss": 0.87,
      "mean_token_accuracy": 0.7501581162214279,
      "num_tokens": 66633897.0,
      "step": 2088
    },
    {
      "entropy": 0.9583968780934811,
      "epoch": 0.19220130994071338,
      "grad_norm": 0.1619790494441986,
      "learning_rate": 9.359646701628485e-05,
      "loss": 0.9459,
      "mean_token_accuracy": 0.7350086122751236,
      "num_tokens": 66665979.0,
      "step": 2089
    },
    {
      "entropy": 0.8190605416893959,
      "epoch": 0.19229331631215463,
      "grad_norm": 0.1640845239162445,
      "learning_rate": 9.359340019014323e-05,
      "loss": 0.837,
      "mean_token_accuracy": 0.764364704489708,
      "num_tokens": 66698366.0,
      "step": 2090
    },
    {
      "entropy": 0.9851617068052292,
      "epoch": 0.19238532268359584,
      "grad_norm": 0.16278637945652008,
      "learning_rate": 9.359033336400161e-05,
      "loss": 0.971,
      "mean_token_accuracy": 0.7255047112703323,
      "num_tokens": 66730192.0,
      "step": 2091
    },
    {
      "entropy": 0.8907298073172569,
      "epoch": 0.19247732905503706,
      "grad_norm": 0.1589939147233963,
      "learning_rate": 9.358726653785998e-05,
      "loss": 0.8985,
      "mean_token_accuracy": 0.7452288381755352,
      "num_tokens": 66762923.0,
      "step": 2092
    },
    {
      "entropy": 0.9049146734178066,
      "epoch": 0.19256933542647828,
      "grad_norm": 0.16110602021217346,
      "learning_rate": 9.358419971171834e-05,
      "loss": 0.8795,
      "mean_token_accuracy": 0.7505999431014061,
      "num_tokens": 66794831.0,
      "step": 2093
    },
    {
      "entropy": 0.9162530191242695,
      "epoch": 0.1926613417979195,
      "grad_norm": 0.1666681319475174,
      "learning_rate": 9.358113288557671e-05,
      "loss": 0.924,
      "mean_token_accuracy": 0.7432136125862598,
      "num_tokens": 66826975.0,
      "step": 2094
    },
    {
      "entropy": 0.8475251868367195,
      "epoch": 0.19275334816936074,
      "grad_norm": 0.1590445339679718,
      "learning_rate": 9.35780660594351e-05,
      "loss": 0.8501,
      "mean_token_accuracy": 0.7619863264262676,
      "num_tokens": 66859274.0,
      "step": 2095
    },
    {
      "entropy": 0.875318119302392,
      "epoch": 0.19284535454080196,
      "grad_norm": 0.15862704813480377,
      "learning_rate": 9.357499923329348e-05,
      "loss": 0.8645,
      "mean_token_accuracy": 0.7572305388748646,
      "num_tokens": 66891080.0,
      "step": 2096
    },
    {
      "entropy": 0.9236644580960274,
      "epoch": 0.19293736091224317,
      "grad_norm": 0.16466963291168213,
      "learning_rate": 9.357193240715184e-05,
      "loss": 0.9361,
      "mean_token_accuracy": 0.7440771274268627,
      "num_tokens": 66923076.0,
      "step": 2097
    },
    {
      "entropy": 0.9217009991407394,
      "epoch": 0.1930293672836844,
      "grad_norm": 0.1642724871635437,
      "learning_rate": 9.356886558101021e-05,
      "loss": 0.9196,
      "mean_token_accuracy": 0.7399358674883842,
      "num_tokens": 66955575.0,
      "step": 2098
    },
    {
      "entropy": 0.9660653062164783,
      "epoch": 0.1931213736551256,
      "grad_norm": 0.17173923552036285,
      "learning_rate": 9.356579875486859e-05,
      "loss": 0.9663,
      "mean_token_accuracy": 0.7269208692014217,
      "num_tokens": 66986621.0,
      "step": 2099
    },
    {
      "entropy": 0.9254742749035358,
      "epoch": 0.19321338002656685,
      "grad_norm": 0.16811595857143402,
      "learning_rate": 9.356273192872696e-05,
      "loss": 0.9219,
      "mean_token_accuracy": 0.7349818795919418,
      "num_tokens": 67017736.0,
      "step": 2100
    },
    {
      "entropy": 0.9177574031054974,
      "epoch": 0.19330538639800807,
      "grad_norm": 0.16194204986095428,
      "learning_rate": 9.355966510258534e-05,
      "loss": 0.9118,
      "mean_token_accuracy": 0.740734938532114,
      "num_tokens": 67049272.0,
      "step": 2101
    },
    {
      "entropy": 0.9066963605582714,
      "epoch": 0.19339739276944928,
      "grad_norm": 0.15565623342990875,
      "learning_rate": 9.355659827644371e-05,
      "loss": 0.8792,
      "mean_token_accuracy": 0.7500106282532215,
      "num_tokens": 67081372.0,
      "step": 2102
    },
    {
      "entropy": 0.8991094473749399,
      "epoch": 0.1934893991408905,
      "grad_norm": 0.15841829776763916,
      "learning_rate": 9.355353145030209e-05,
      "loss": 0.9068,
      "mean_token_accuracy": 0.7467436641454697,
      "num_tokens": 67113784.0,
      "step": 2103
    },
    {
      "entropy": 0.8435263633728027,
      "epoch": 0.19358140551233172,
      "grad_norm": 0.1569119542837143,
      "learning_rate": 9.355046462416046e-05,
      "loss": 0.8432,
      "mean_token_accuracy": 0.7595317475497723,
      "num_tokens": 67144935.0,
      "step": 2104
    },
    {
      "entropy": 0.8339877147227526,
      "epoch": 0.19367341188377296,
      "grad_norm": 0.16009721159934998,
      "learning_rate": 9.354739779801883e-05,
      "loss": 0.8514,
      "mean_token_accuracy": 0.7586171142756939,
      "num_tokens": 67177082.0,
      "step": 2105
    },
    {
      "entropy": 0.9030448161065578,
      "epoch": 0.19376541825521418,
      "grad_norm": 0.15736226737499237,
      "learning_rate": 9.354433097187721e-05,
      "loss": 0.9073,
      "mean_token_accuracy": 0.7431066073477268,
      "num_tokens": 67208906.0,
      "step": 2106
    },
    {
      "entropy": 0.9498293995857239,
      "epoch": 0.1938574246266554,
      "grad_norm": 0.160381481051445,
      "learning_rate": 9.354126414573559e-05,
      "loss": 0.9601,
      "mean_token_accuracy": 0.7307383604347706,
      "num_tokens": 67240893.0,
      "step": 2107
    },
    {
      "entropy": 0.8450409136712551,
      "epoch": 0.1939494309980966,
      "grad_norm": 0.1544397920370102,
      "learning_rate": 9.353819731959396e-05,
      "loss": 0.8478,
      "mean_token_accuracy": 0.7589408382773399,
      "num_tokens": 67272560.0,
      "step": 2108
    },
    {
      "entropy": 0.7735142298042774,
      "epoch": 0.19404143736953783,
      "grad_norm": 0.15260325372219086,
      "learning_rate": 9.353513049345233e-05,
      "loss": 0.7582,
      "mean_token_accuracy": 0.7797415517270565,
      "num_tokens": 67304506.0,
      "step": 2109
    },
    {
      "entropy": 0.8527704812586308,
      "epoch": 0.19413344374097907,
      "grad_norm": 0.1555686593055725,
      "learning_rate": 9.35320636673107e-05,
      "loss": 0.8342,
      "mean_token_accuracy": 0.7609553374350071,
      "num_tokens": 67336470.0,
      "step": 2110
    },
    {
      "entropy": 0.9466509278863668,
      "epoch": 0.1942254501124203,
      "grad_norm": 0.1694634109735489,
      "learning_rate": 9.352899684116907e-05,
      "loss": 0.9701,
      "mean_token_accuracy": 0.730896707624197,
      "num_tokens": 67368782.0,
      "step": 2111
    },
    {
      "entropy": 0.9233518056571484,
      "epoch": 0.1943174564838615,
      "grad_norm": 0.16074050962924957,
      "learning_rate": 9.352593001502746e-05,
      "loss": 0.9199,
      "mean_token_accuracy": 0.7435019165277481,
      "num_tokens": 67400104.0,
      "step": 2112
    },
    {
      "entropy": 0.888766223564744,
      "epoch": 0.19440946285530272,
      "grad_norm": 0.16560155153274536,
      "learning_rate": 9.352286318888582e-05,
      "loss": 0.8934,
      "mean_token_accuracy": 0.7464515268802643,
      "num_tokens": 67432392.0,
      "step": 2113
    },
    {
      "entropy": 0.9413719289004803,
      "epoch": 0.19450146922674394,
      "grad_norm": 0.16608722507953644,
      "learning_rate": 9.35197963627442e-05,
      "loss": 0.9378,
      "mean_token_accuracy": 0.732838574796915,
      "num_tokens": 67463850.0,
      "step": 2114
    },
    {
      "entropy": 1.0279192365705967,
      "epoch": 0.19459347559818518,
      "grad_norm": 0.17054377496242523,
      "learning_rate": 9.351672953660257e-05,
      "loss": 1.0286,
      "mean_token_accuracy": 0.7088443785905838,
      "num_tokens": 67496365.0,
      "step": 2115
    },
    {
      "entropy": 0.9449400827288628,
      "epoch": 0.1946854819696264,
      "grad_norm": 0.16602954268455505,
      "learning_rate": 9.351366271046094e-05,
      "loss": 0.9539,
      "mean_token_accuracy": 0.732362899929285,
      "num_tokens": 67527700.0,
      "step": 2116
    },
    {
      "entropy": 0.9624862261116505,
      "epoch": 0.19477748834106762,
      "grad_norm": 0.16469870507717133,
      "learning_rate": 9.351059588431932e-05,
      "loss": 0.9757,
      "mean_token_accuracy": 0.7340248599648476,
      "num_tokens": 67560102.0,
      "step": 2117
    },
    {
      "entropy": 0.9186250232160091,
      "epoch": 0.19486949471250883,
      "grad_norm": 0.15537488460540771,
      "learning_rate": 9.35075290581777e-05,
      "loss": 0.899,
      "mean_token_accuracy": 0.741860531270504,
      "num_tokens": 67592305.0,
      "step": 2118
    },
    {
      "entropy": 0.8950015567243099,
      "epoch": 0.19496150108395005,
      "grad_norm": 0.1594477742910385,
      "learning_rate": 9.350446223203607e-05,
      "loss": 0.8931,
      "mean_token_accuracy": 0.745855625718832,
      "num_tokens": 67624467.0,
      "step": 2119
    },
    {
      "entropy": 0.9548966847360134,
      "epoch": 0.1950535074553913,
      "grad_norm": 0.16676531732082367,
      "learning_rate": 9.350139540589444e-05,
      "loss": 0.9589,
      "mean_token_accuracy": 0.7338507138192654,
      "num_tokens": 67657002.0,
      "step": 2120
    },
    {
      "entropy": 0.8476665187627077,
      "epoch": 0.1951455138268325,
      "grad_norm": 0.16230550408363342,
      "learning_rate": 9.349832857975281e-05,
      "loss": 0.8519,
      "mean_token_accuracy": 0.7569580152630806,
      "num_tokens": 67689004.0,
      "step": 2121
    },
    {
      "entropy": 0.930550042539835,
      "epoch": 0.19523752019827373,
      "grad_norm": 0.15949797630310059,
      "learning_rate": 9.34952617536112e-05,
      "loss": 0.9011,
      "mean_token_accuracy": 0.7426630705595016,
      "num_tokens": 67720904.0,
      "step": 2122
    },
    {
      "entropy": 0.9896409176290035,
      "epoch": 0.19532952656971495,
      "grad_norm": 0.1635100394487381,
      "learning_rate": 9.349219492746957e-05,
      "loss": 0.9785,
      "mean_token_accuracy": 0.7250706627964973,
      "num_tokens": 67753170.0,
      "step": 2123
    },
    {
      "entropy": 0.840200137346983,
      "epoch": 0.19542153294115616,
      "grad_norm": 0.16582147777080536,
      "learning_rate": 9.348912810132794e-05,
      "loss": 0.8608,
      "mean_token_accuracy": 0.7507183030247688,
      "num_tokens": 67785266.0,
      "step": 2124
    },
    {
      "entropy": 0.8174498938024044,
      "epoch": 0.1955135393125974,
      "grad_norm": 0.16237811744213104,
      "learning_rate": 9.34860612751863e-05,
      "loss": 0.8424,
      "mean_token_accuracy": 0.7645053565502167,
      "num_tokens": 67816554.0,
      "step": 2125
    },
    {
      "entropy": 0.8248377479612827,
      "epoch": 0.19560554568403862,
      "grad_norm": 0.1612454205751419,
      "learning_rate": 9.348299444904469e-05,
      "loss": 0.8337,
      "mean_token_accuracy": 0.7661922611296177,
      "num_tokens": 67848157.0,
      "step": 2126
    },
    {
      "entropy": 0.8328925501555204,
      "epoch": 0.19569755205547984,
      "grad_norm": 0.15930548310279846,
      "learning_rate": 9.347992762290307e-05,
      "loss": 0.8412,
      "mean_token_accuracy": 0.7637028507888317,
      "num_tokens": 67880113.0,
      "step": 2127
    },
    {
      "entropy": 0.9146295748651028,
      "epoch": 0.19578955842692106,
      "grad_norm": 0.16118566691875458,
      "learning_rate": 9.347686079676144e-05,
      "loss": 0.9079,
      "mean_token_accuracy": 0.7403504587709904,
      "num_tokens": 67912035.0,
      "step": 2128
    },
    {
      "entropy": 0.9130019657313824,
      "epoch": 0.19588156479836227,
      "grad_norm": 0.1665070354938507,
      "learning_rate": 9.34737939706198e-05,
      "loss": 0.9281,
      "mean_token_accuracy": 0.7367870546877384,
      "num_tokens": 67944028.0,
      "step": 2129
    },
    {
      "entropy": 0.9007689990103245,
      "epoch": 0.19597357116980352,
      "grad_norm": 0.16633763909339905,
      "learning_rate": 9.347072714447819e-05,
      "loss": 0.8951,
      "mean_token_accuracy": 0.7467646636068821,
      "num_tokens": 67976368.0,
      "step": 2130
    },
    {
      "entropy": 0.9687911011278629,
      "epoch": 0.19606557754124473,
      "grad_norm": 0.15961508452892303,
      "learning_rate": 9.346766031833655e-05,
      "loss": 0.9519,
      "mean_token_accuracy": 0.732357207685709,
      "num_tokens": 68008657.0,
      "step": 2131
    },
    {
      "entropy": 0.9267431627959013,
      "epoch": 0.19615758391268595,
      "grad_norm": 0.15613701939582825,
      "learning_rate": 9.346459349219494e-05,
      "loss": 0.9124,
      "mean_token_accuracy": 0.7423344701528549,
      "num_tokens": 68040965.0,
      "step": 2132
    },
    {
      "entropy": 0.9432807043194771,
      "epoch": 0.19624959028412717,
      "grad_norm": 0.161041721701622,
      "learning_rate": 9.34615266660533e-05,
      "loss": 0.9383,
      "mean_token_accuracy": 0.7340038456022739,
      "num_tokens": 68071808.0,
      "step": 2133
    },
    {
      "entropy": 0.9378496669232845,
      "epoch": 0.19634159665556838,
      "grad_norm": 0.1623469740152359,
      "learning_rate": 9.345845983991168e-05,
      "loss": 0.9378,
      "mean_token_accuracy": 0.7317124232649803,
      "num_tokens": 68102660.0,
      "step": 2134
    },
    {
      "entropy": 0.8744136579334736,
      "epoch": 0.19643360302700963,
      "grad_norm": 0.1597345471382141,
      "learning_rate": 9.345539301377005e-05,
      "loss": 0.893,
      "mean_token_accuracy": 0.7470070458948612,
      "num_tokens": 68134663.0,
      "step": 2135
    },
    {
      "entropy": 0.9130709134042263,
      "epoch": 0.19652560939845085,
      "grad_norm": 0.16366080939769745,
      "learning_rate": 9.345232618762842e-05,
      "loss": 0.938,
      "mean_token_accuracy": 0.7319407500326633,
      "num_tokens": 68166285.0,
      "step": 2136
    },
    {
      "entropy": 0.9083885960280895,
      "epoch": 0.19661761576989206,
      "grad_norm": 0.16435208916664124,
      "learning_rate": 9.34492593614868e-05,
      "loss": 0.9145,
      "mean_token_accuracy": 0.7442589700222015,
      "num_tokens": 68198047.0,
      "step": 2137
    },
    {
      "entropy": 0.9266248941421509,
      "epoch": 0.19670962214133328,
      "grad_norm": 0.16171085834503174,
      "learning_rate": 9.344619253534518e-05,
      "loss": 0.934,
      "mean_token_accuracy": 0.7327006570994854,
      "num_tokens": 68229655.0,
      "step": 2138
    },
    {
      "entropy": 0.8541943151503801,
      "epoch": 0.1968016285127745,
      "grad_norm": 0.15876218676567078,
      "learning_rate": 9.344312570920355e-05,
      "loss": 0.8455,
      "mean_token_accuracy": 0.758536271750927,
      "num_tokens": 68262341.0,
      "step": 2139
    },
    {
      "entropy": 0.7640830762684345,
      "epoch": 0.19689363488421574,
      "grad_norm": 0.1471771001815796,
      "learning_rate": 9.344005888306192e-05,
      "loss": 0.7368,
      "mean_token_accuracy": 0.787493884563446,
      "num_tokens": 68295109.0,
      "step": 2140
    },
    {
      "entropy": 0.7940677963197231,
      "epoch": 0.19698564125565696,
      "grad_norm": 0.1550474613904953,
      "learning_rate": 9.34369920569203e-05,
      "loss": 0.7876,
      "mean_token_accuracy": 0.7706819027662277,
      "num_tokens": 68326953.0,
      "step": 2141
    },
    {
      "entropy": 0.9011071883141994,
      "epoch": 0.19707764762709817,
      "grad_norm": 0.1608584076166153,
      "learning_rate": 9.343392523077867e-05,
      "loss": 0.9159,
      "mean_token_accuracy": 0.7381557896733284,
      "num_tokens": 68358967.0,
      "step": 2142
    },
    {
      "entropy": 0.8745359685271978,
      "epoch": 0.1971696539985394,
      "grad_norm": 0.1543223261833191,
      "learning_rate": 9.343085840463705e-05,
      "loss": 0.8401,
      "mean_token_accuracy": 0.7590012811124325,
      "num_tokens": 68390813.0,
      "step": 2143
    },
    {
      "entropy": 0.922530110925436,
      "epoch": 0.19726166036998063,
      "grad_norm": 0.1647340953350067,
      "learning_rate": 9.342779157849542e-05,
      "loss": 0.9291,
      "mean_token_accuracy": 0.7365592271089554,
      "num_tokens": 68423070.0,
      "step": 2144
    },
    {
      "entropy": 0.9111796729266644,
      "epoch": 0.19735366674142185,
      "grad_norm": 0.1610458791255951,
      "learning_rate": 9.34247247523538e-05,
      "loss": 0.9152,
      "mean_token_accuracy": 0.7449517361819744,
      "num_tokens": 68455345.0,
      "step": 2145
    },
    {
      "entropy": 0.7103486321866512,
      "epoch": 0.19744567311286307,
      "grad_norm": 0.1555427461862564,
      "learning_rate": 9.342165792621217e-05,
      "loss": 0.6936,
      "mean_token_accuracy": 0.7970850877463818,
      "num_tokens": 68487310.0,
      "step": 2146
    },
    {
      "entropy": 0.9145626500248909,
      "epoch": 0.19753767948430428,
      "grad_norm": 0.15671616792678833,
      "learning_rate": 9.341859110007053e-05,
      "loss": 0.9028,
      "mean_token_accuracy": 0.7442108914256096,
      "num_tokens": 68519473.0,
      "step": 2147
    },
    {
      "entropy": 0.9443887583911419,
      "epoch": 0.1976296858557455,
      "grad_norm": 0.16379334032535553,
      "learning_rate": 9.341552427392892e-05,
      "loss": 0.9619,
      "mean_token_accuracy": 0.725195188075304,
      "num_tokens": 68552168.0,
      "step": 2148
    },
    {
      "entropy": 0.8554215934127569,
      "epoch": 0.19772169222718675,
      "grad_norm": 0.16314350068569183,
      "learning_rate": 9.34124574477873e-05,
      "loss": 0.8678,
      "mean_token_accuracy": 0.7591111958026886,
      "num_tokens": 68584780.0,
      "step": 2149
    },
    {
      "entropy": 0.9231494348496199,
      "epoch": 0.19781369859862796,
      "grad_norm": 0.15925024449825287,
      "learning_rate": 9.340939062164567e-05,
      "loss": 0.9266,
      "mean_token_accuracy": 0.736957885324955,
      "num_tokens": 68616934.0,
      "step": 2150
    },
    {
      "entropy": 0.856573298573494,
      "epoch": 0.19790570497006918,
      "grad_norm": 0.1621844470500946,
      "learning_rate": 9.340632379550403e-05,
      "loss": 0.8449,
      "mean_token_accuracy": 0.761481236666441,
      "num_tokens": 68648619.0,
      "step": 2151
    },
    {
      "entropy": 0.9310447964817286,
      "epoch": 0.1979977113415104,
      "grad_norm": 0.1645403355360031,
      "learning_rate": 9.34032569693624e-05,
      "loss": 0.9274,
      "mean_token_accuracy": 0.7367376387119293,
      "num_tokens": 68680752.0,
      "step": 2152
    },
    {
      "entropy": 0.8028207886964083,
      "epoch": 0.1980897177129516,
      "grad_norm": 0.15988674759864807,
      "learning_rate": 9.340019014322078e-05,
      "loss": 0.8112,
      "mean_token_accuracy": 0.7647760286927223,
      "num_tokens": 68712478.0,
      "step": 2153
    },
    {
      "entropy": 0.9550810474902391,
      "epoch": 0.19818172408439286,
      "grad_norm": 0.17101852595806122,
      "learning_rate": 9.339712331707916e-05,
      "loss": 0.9796,
      "mean_token_accuracy": 0.7284754142165184,
      "num_tokens": 68744924.0,
      "step": 2154
    },
    {
      "entropy": 0.9150787778198719,
      "epoch": 0.19827373045583407,
      "grad_norm": 0.1661899834871292,
      "learning_rate": 9.339405649093753e-05,
      "loss": 0.8791,
      "mean_token_accuracy": 0.7455488108098507,
      "num_tokens": 68774856.0,
      "step": 2155
    },
    {
      "entropy": 0.8780384138226509,
      "epoch": 0.1983657368272753,
      "grad_norm": 0.15516163408756256,
      "learning_rate": 9.33909896647959e-05,
      "loss": 0.8707,
      "mean_token_accuracy": 0.7496176920831203,
      "num_tokens": 68807066.0,
      "step": 2156
    },
    {
      "entropy": 0.8043510671705008,
      "epoch": 0.1984577431987165,
      "grad_norm": 0.15869726240634918,
      "learning_rate": 9.338792283865428e-05,
      "loss": 0.8052,
      "mean_token_accuracy": 0.7651760093867779,
      "num_tokens": 68837882.0,
      "step": 2157
    },
    {
      "entropy": 0.8685892857611179,
      "epoch": 0.19854974957015772,
      "grad_norm": 0.15946121513843536,
      "learning_rate": 9.338485601251265e-05,
      "loss": 0.8807,
      "mean_token_accuracy": 0.7513917721807957,
      "num_tokens": 68870436.0,
      "step": 2158
    },
    {
      "entropy": 0.8051081895828247,
      "epoch": 0.19864175594159897,
      "grad_norm": 0.15095248818397522,
      "learning_rate": 9.338178918637103e-05,
      "loss": 0.795,
      "mean_token_accuracy": 0.769863449037075,
      "num_tokens": 68902683.0,
      "step": 2159
    },
    {
      "entropy": 0.8500205148011446,
      "epoch": 0.19873376231304019,
      "grad_norm": 0.15605954825878143,
      "learning_rate": 9.33787223602294e-05,
      "loss": 0.8405,
      "mean_token_accuracy": 0.7572056613862514,
      "num_tokens": 68934875.0,
      "step": 2160
    },
    {
      "entropy": 0.8687045499682426,
      "epoch": 0.1988257686844814,
      "grad_norm": 0.15447843074798584,
      "learning_rate": 9.337565553408778e-05,
      "loss": 0.8438,
      "mean_token_accuracy": 0.7601562477648258,
      "num_tokens": 68966974.0,
      "step": 2161
    },
    {
      "entropy": 0.9325909893959761,
      "epoch": 0.19891777505592262,
      "grad_norm": 0.16031193733215332,
      "learning_rate": 9.337258870794615e-05,
      "loss": 0.9235,
      "mean_token_accuracy": 0.7391127869486809,
      "num_tokens": 68998671.0,
      "step": 2162
    },
    {
      "entropy": 0.9575784560292959,
      "epoch": 0.19900978142736384,
      "grad_norm": 0.16326804459095,
      "learning_rate": 9.336952188180453e-05,
      "loss": 0.9586,
      "mean_token_accuracy": 0.7266486249864101,
      "num_tokens": 69030721.0,
      "step": 2163
    },
    {
      "entropy": 0.8920419216156006,
      "epoch": 0.19910178779880508,
      "grad_norm": 0.16042904555797577,
      "learning_rate": 9.33664550556629e-05,
      "loss": 0.8949,
      "mean_token_accuracy": 0.7449088171124458,
      "num_tokens": 69062122.0,
      "step": 2164
    },
    {
      "entropy": 0.9506238624453545,
      "epoch": 0.1991937941702463,
      "grad_norm": 0.16223110258579254,
      "learning_rate": 9.336338822952128e-05,
      "loss": 0.9398,
      "mean_token_accuracy": 0.7357291392982006,
      "num_tokens": 69094483.0,
      "step": 2165
    },
    {
      "entropy": 0.913567716255784,
      "epoch": 0.1992858005416875,
      "grad_norm": 0.1567869484424591,
      "learning_rate": 9.336032140337965e-05,
      "loss": 0.9039,
      "mean_token_accuracy": 0.738114133477211,
      "num_tokens": 69126998.0,
      "step": 2166
    },
    {
      "entropy": 0.840234437957406,
      "epoch": 0.19937780691312873,
      "grad_norm": 0.15466199815273285,
      "learning_rate": 9.335725457723801e-05,
      "loss": 0.8326,
      "mean_token_accuracy": 0.7622020877897739,
      "num_tokens": 69159153.0,
      "step": 2167
    },
    {
      "entropy": 1.0835785642266273,
      "epoch": 0.19946981328456995,
      "grad_norm": 0.17085401713848114,
      "learning_rate": 9.33541877510964e-05,
      "loss": 1.0994,
      "mean_token_accuracy": 0.6930282264947891,
      "num_tokens": 69190751.0,
      "step": 2168
    },
    {
      "entropy": 0.882443169131875,
      "epoch": 0.1995618196560112,
      "grad_norm": 0.15839269757270813,
      "learning_rate": 9.335112092495478e-05,
      "loss": 0.8794,
      "mean_token_accuracy": 0.7450637482106686,
      "num_tokens": 69223258.0,
      "step": 2169
    },
    {
      "entropy": 0.8651303760707378,
      "epoch": 0.1996538260274524,
      "grad_norm": 0.16337448358535767,
      "learning_rate": 9.334805409881314e-05,
      "loss": 0.8839,
      "mean_token_accuracy": 0.7500738985836506,
      "num_tokens": 69255365.0,
      "step": 2170
    },
    {
      "entropy": 0.9112938940525055,
      "epoch": 0.19974583239889362,
      "grad_norm": 0.16801586747169495,
      "learning_rate": 9.334498727267151e-05,
      "loss": 0.9236,
      "mean_token_accuracy": 0.7388861775398254,
      "num_tokens": 69288133.0,
      "step": 2171
    },
    {
      "entropy": 0.8369556702673435,
      "epoch": 0.19983783877033484,
      "grad_norm": 0.1597340703010559,
      "learning_rate": 9.33419204465299e-05,
      "loss": 0.8545,
      "mean_token_accuracy": 0.7557508200407028,
      "num_tokens": 69320183.0,
      "step": 2172
    },
    {
      "entropy": 0.8907689284533262,
      "epoch": 0.19992984514177606,
      "grad_norm": 0.1650434136390686,
      "learning_rate": 9.333885362038826e-05,
      "loss": 0.8837,
      "mean_token_accuracy": 0.7495999112725258,
      "num_tokens": 69352864.0,
      "step": 2173
    },
    {
      "entropy": 1.0019940454512835,
      "epoch": 0.2000218515132173,
      "grad_norm": 0.16670800745487213,
      "learning_rate": 9.333578679424664e-05,
      "loss": 0.9979,
      "mean_token_accuracy": 0.7168303392827511,
      "num_tokens": 69384817.0,
      "step": 2174
    },
    {
      "entropy": 0.9252850078046322,
      "epoch": 0.20011385788465852,
      "grad_norm": 0.1596802920103073,
      "learning_rate": 9.333271996810501e-05,
      "loss": 0.9258,
      "mean_token_accuracy": 0.7425597161054611,
      "num_tokens": 69417365.0,
      "step": 2175
    },
    {
      "entropy": 0.9893336333334446,
      "epoch": 0.20020586425609974,
      "grad_norm": 0.17222565412521362,
      "learning_rate": 9.332965314196339e-05,
      "loss": 0.987,
      "mean_token_accuracy": 0.7197200767695904,
      "num_tokens": 69448995.0,
      "step": 2176
    },
    {
      "entropy": 0.8060479164123535,
      "epoch": 0.20029787062754095,
      "grad_norm": 0.15398778021335602,
      "learning_rate": 9.332658631582176e-05,
      "loss": 0.7987,
      "mean_token_accuracy": 0.7676365748047829,
      "num_tokens": 69481130.0,
      "step": 2177
    },
    {
      "entropy": 0.792138647288084,
      "epoch": 0.20038987699898217,
      "grad_norm": 0.15124332904815674,
      "learning_rate": 9.332351948968013e-05,
      "loss": 0.788,
      "mean_token_accuracy": 0.7721965499222279,
      "num_tokens": 69513498.0,
      "step": 2178
    },
    {
      "entropy": 0.8736269772052765,
      "epoch": 0.2004818833704234,
      "grad_norm": 0.1592663675546646,
      "learning_rate": 9.332045266353851e-05,
      "loss": 0.868,
      "mean_token_accuracy": 0.7570287883281708,
      "num_tokens": 69546258.0,
      "step": 2179
    },
    {
      "entropy": 0.896639421582222,
      "epoch": 0.20057388974186463,
      "grad_norm": 0.1653127372264862,
      "learning_rate": 9.331738583739689e-05,
      "loss": 0.8923,
      "mean_token_accuracy": 0.7461023926734924,
      "num_tokens": 69577216.0,
      "step": 2180
    },
    {
      "entropy": 0.8784413896501064,
      "epoch": 0.20066589611330585,
      "grad_norm": 0.15927815437316895,
      "learning_rate": 9.331431901125526e-05,
      "loss": 0.8641,
      "mean_token_accuracy": 0.754275482147932,
      "num_tokens": 69609588.0,
      "step": 2181
    },
    {
      "entropy": 0.9449828043580055,
      "epoch": 0.20075790248474706,
      "grad_norm": 0.15984024107456207,
      "learning_rate": 9.331125218511363e-05,
      "loss": 0.9373,
      "mean_token_accuracy": 0.7373170740902424,
      "num_tokens": 69640897.0,
      "step": 2182
    },
    {
      "entropy": 0.9587430469691753,
      "epoch": 0.20084990885618828,
      "grad_norm": 0.16778352856636047,
      "learning_rate": 9.3308185358972e-05,
      "loss": 0.9671,
      "mean_token_accuracy": 0.7292763106524944,
      "num_tokens": 69672948.0,
      "step": 2183
    },
    {
      "entropy": 0.9872140549123287,
      "epoch": 0.20094191522762953,
      "grad_norm": 0.17245371639728546,
      "learning_rate": 9.330511853283038e-05,
      "loss": 0.9868,
      "mean_token_accuracy": 0.7216991111636162,
      "num_tokens": 69704728.0,
      "step": 2184
    },
    {
      "entropy": 0.9505719132721424,
      "epoch": 0.20103392159907074,
      "grad_norm": 0.16392174363136292,
      "learning_rate": 9.330205170668876e-05,
      "loss": 0.9544,
      "mean_token_accuracy": 0.7252263575792313,
      "num_tokens": 69736225.0,
      "step": 2185
    },
    {
      "entropy": 0.8538291342556477,
      "epoch": 0.20112592797051196,
      "grad_norm": 0.1608898937702179,
      "learning_rate": 9.329898488054713e-05,
      "loss": 0.8732,
      "mean_token_accuracy": 0.7523510046303272,
      "num_tokens": 69768993.0,
      "step": 2186
    },
    {
      "entropy": 0.9307313039898872,
      "epoch": 0.20121793434195318,
      "grad_norm": 0.1597103625535965,
      "learning_rate": 9.32959180544055e-05,
      "loss": 0.9155,
      "mean_token_accuracy": 0.7391259223222733,
      "num_tokens": 69801103.0,
      "step": 2187
    },
    {
      "entropy": 0.8700545746833086,
      "epoch": 0.2013099407133944,
      "grad_norm": 0.15828531980514526,
      "learning_rate": 9.329285122826388e-05,
      "loss": 0.8428,
      "mean_token_accuracy": 0.7596983499825001,
      "num_tokens": 69833796.0,
      "step": 2188
    },
    {
      "entropy": 0.8197246417403221,
      "epoch": 0.20140194708483564,
      "grad_norm": 0.14586876332759857,
      "learning_rate": 9.328978440212224e-05,
      "loss": 0.7993,
      "mean_token_accuracy": 0.7656061537563801,
      "num_tokens": 69866177.0,
      "step": 2189
    },
    {
      "entropy": 0.8133249562233686,
      "epoch": 0.20149395345627685,
      "grad_norm": 0.15228766202926636,
      "learning_rate": 9.328671757598062e-05,
      "loss": 0.804,
      "mean_token_accuracy": 0.7688422501087189,
      "num_tokens": 69898551.0,
      "step": 2190
    },
    {
      "entropy": 0.8948649019002914,
      "epoch": 0.20158595982771807,
      "grad_norm": 0.15482689440250397,
      "learning_rate": 9.328365074983899e-05,
      "loss": 0.8796,
      "mean_token_accuracy": 0.7462757751345634,
      "num_tokens": 69930101.0,
      "step": 2191
    },
    {
      "entropy": 0.9370650835335255,
      "epoch": 0.2016779661991593,
      "grad_norm": 0.15942099690437317,
      "learning_rate": 9.328058392369737e-05,
      "loss": 0.9488,
      "mean_token_accuracy": 0.7291441224515438,
      "num_tokens": 69961700.0,
      "step": 2192
    },
    {
      "entropy": 0.7694588731974363,
      "epoch": 0.2017699725706005,
      "grad_norm": 0.15336187183856964,
      "learning_rate": 9.327751709755574e-05,
      "loss": 0.7804,
      "mean_token_accuracy": 0.7781623601913452,
      "num_tokens": 69993465.0,
      "step": 2193
    },
    {
      "entropy": 0.9570678174495697,
      "epoch": 0.20186197894204175,
      "grad_norm": 0.16695457696914673,
      "learning_rate": 9.327445027141411e-05,
      "loss": 0.9636,
      "mean_token_accuracy": 0.7288404926657677,
      "num_tokens": 70025375.0,
      "step": 2194
    },
    {
      "entropy": 0.8693357035517693,
      "epoch": 0.20195398531348296,
      "grad_norm": 0.15825155377388,
      "learning_rate": 9.327138344527249e-05,
      "loss": 0.8613,
      "mean_token_accuracy": 0.7554026320576668,
      "num_tokens": 70057465.0,
      "step": 2195
    },
    {
      "entropy": 0.964122112840414,
      "epoch": 0.20204599168492418,
      "grad_norm": 0.16600635647773743,
      "learning_rate": 9.326831661913087e-05,
      "loss": 0.9522,
      "mean_token_accuracy": 0.7324707247316837,
      "num_tokens": 70088934.0,
      "step": 2196
    },
    {
      "entropy": 0.871784832328558,
      "epoch": 0.2021379980563654,
      "grad_norm": 0.15988291800022125,
      "learning_rate": 9.326524979298924e-05,
      "loss": 0.8701,
      "mean_token_accuracy": 0.7508639022707939,
      "num_tokens": 70120984.0,
      "step": 2197
    },
    {
      "entropy": 1.0142599046230316,
      "epoch": 0.20223000442780661,
      "grad_norm": 0.17139887809753418,
      "learning_rate": 9.326218296684761e-05,
      "loss": 1.0144,
      "mean_token_accuracy": 0.7137387879192829,
      "num_tokens": 70152761.0,
      "step": 2198
    },
    {
      "entropy": 0.9462217539548874,
      "epoch": 0.20232201079924786,
      "grad_norm": 0.16658520698547363,
      "learning_rate": 9.325911614070599e-05,
      "loss": 0.9263,
      "mean_token_accuracy": 0.7401720359921455,
      "num_tokens": 70184611.0,
      "step": 2199
    },
    {
      "entropy": 0.778026457875967,
      "epoch": 0.20241401717068908,
      "grad_norm": 0.16702984273433685,
      "learning_rate": 9.325604931456437e-05,
      "loss": 0.826,
      "mean_token_accuracy": 0.7669633738696575,
      "num_tokens": 70216818.0,
      "step": 2200
    },
    {
      "entropy": 0.8951413594186306,
      "epoch": 0.2025060235421303,
      "grad_norm": 0.1673172563314438,
      "learning_rate": 9.325298248842274e-05,
      "loss": 0.8803,
      "mean_token_accuracy": 0.7503014281392097,
      "num_tokens": 70247870.0,
      "step": 2201
    },
    {
      "entropy": 0.8826032392680645,
      "epoch": 0.2025980299135715,
      "grad_norm": 0.1592205911874771,
      "learning_rate": 9.32499156622811e-05,
      "loss": 0.8882,
      "mean_token_accuracy": 0.7465986683964729,
      "num_tokens": 70279373.0,
      "step": 2202
    },
    {
      "entropy": 0.8762294761836529,
      "epoch": 0.20269003628501273,
      "grad_norm": 0.1632188856601715,
      "learning_rate": 9.324684883613949e-05,
      "loss": 0.8716,
      "mean_token_accuracy": 0.7506240978837013,
      "num_tokens": 70311589.0,
      "step": 2203
    },
    {
      "entropy": 0.9174760244786739,
      "epoch": 0.20278204265645397,
      "grad_norm": 0.16115644574165344,
      "learning_rate": 9.324378200999786e-05,
      "loss": 0.9191,
      "mean_token_accuracy": 0.7368734739720821,
      "num_tokens": 70343511.0,
      "step": 2204
    },
    {
      "entropy": 0.8106661178171635,
      "epoch": 0.2028740490278952,
      "grad_norm": 0.15165027976036072,
      "learning_rate": 9.324071518385624e-05,
      "loss": 0.7957,
      "mean_token_accuracy": 0.7712200731039047,
      "num_tokens": 70376279.0,
      "step": 2205
    },
    {
      "entropy": 0.8986492082476616,
      "epoch": 0.2029660553993364,
      "grad_norm": 0.16154593229293823,
      "learning_rate": 9.32376483577146e-05,
      "loss": 0.8951,
      "mean_token_accuracy": 0.7456168755888939,
      "num_tokens": 70407618.0,
      "step": 2206
    },
    {
      "entropy": 0.7975661344826221,
      "epoch": 0.20305806177077762,
      "grad_norm": 0.1589781641960144,
      "learning_rate": 9.323458153157299e-05,
      "loss": 0.8031,
      "mean_token_accuracy": 0.7701114602386951,
      "num_tokens": 70440311.0,
      "step": 2207
    },
    {
      "entropy": 0.957557201385498,
      "epoch": 0.20315006814221884,
      "grad_norm": 0.1652262955904007,
      "learning_rate": 9.323151470543135e-05,
      "loss": 0.9684,
      "mean_token_accuracy": 0.731676310300827,
      "num_tokens": 70471655.0,
      "step": 2208
    },
    {
      "entropy": 0.7807961888611317,
      "epoch": 0.20324207451366008,
      "grad_norm": 0.15728799998760223,
      "learning_rate": 9.322844787928972e-05,
      "loss": 0.7898,
      "mean_token_accuracy": 0.7723208330571651,
      "num_tokens": 70504058.0,
      "step": 2209
    },
    {
      "entropy": 0.8233674541115761,
      "epoch": 0.2033340808851013,
      "grad_norm": 0.15396136045455933,
      "learning_rate": 9.32253810531481e-05,
      "loss": 0.8025,
      "mean_token_accuracy": 0.768945436924696,
      "num_tokens": 70535978.0,
      "step": 2210
    },
    {
      "entropy": 0.8871812038123608,
      "epoch": 0.20342608725654251,
      "grad_norm": 0.15909843146800995,
      "learning_rate": 9.322231422700649e-05,
      "loss": 0.87,
      "mean_token_accuracy": 0.75572819262743,
      "num_tokens": 70568358.0,
      "step": 2211
    },
    {
      "entropy": 0.8215491957962513,
      "epoch": 0.20351809362798373,
      "grad_norm": 0.1542416661977768,
      "learning_rate": 9.321924740086485e-05,
      "loss": 0.8279,
      "mean_token_accuracy": 0.7594820410013199,
      "num_tokens": 70600899.0,
      "step": 2212
    },
    {
      "entropy": 0.8794572614133358,
      "epoch": 0.20361009999942495,
      "grad_norm": 0.15601420402526855,
      "learning_rate": 9.321618057472322e-05,
      "loss": 0.8735,
      "mean_token_accuracy": 0.7527399845421314,
      "num_tokens": 70633092.0,
      "step": 2213
    },
    {
      "entropy": 0.9591118954122066,
      "epoch": 0.2037021063708662,
      "grad_norm": 0.16647869348526,
      "learning_rate": 9.321311374858159e-05,
      "loss": 0.9649,
      "mean_token_accuracy": 0.728139653801918,
      "num_tokens": 70665063.0,
      "step": 2214
    },
    {
      "entropy": 0.9139295145869255,
      "epoch": 0.2037941127423074,
      "grad_norm": 0.16323556005954742,
      "learning_rate": 9.321004692243997e-05,
      "loss": 0.9074,
      "mean_token_accuracy": 0.7435410544276237,
      "num_tokens": 70697295.0,
      "step": 2215
    },
    {
      "entropy": 0.8877883665263653,
      "epoch": 0.20388611911374863,
      "grad_norm": 0.15870517492294312,
      "learning_rate": 9.320698009629835e-05,
      "loss": 0.8932,
      "mean_token_accuracy": 0.7432291954755783,
      "num_tokens": 70729416.0,
      "step": 2216
    },
    {
      "entropy": 1.0039665326476097,
      "epoch": 0.20397812548518984,
      "grad_norm": 0.16851668059825897,
      "learning_rate": 9.320391327015672e-05,
      "loss": 1.0297,
      "mean_token_accuracy": 0.7148955538868904,
      "num_tokens": 70761956.0,
      "step": 2217
    },
    {
      "entropy": 0.8628738764673471,
      "epoch": 0.20407013185663106,
      "grad_norm": 0.1573517769575119,
      "learning_rate": 9.320084644401509e-05,
      "loss": 0.86,
      "mean_token_accuracy": 0.7489518858492374,
      "num_tokens": 70793572.0,
      "step": 2218
    },
    {
      "entropy": 0.8954880312085152,
      "epoch": 0.2041621382280723,
      "grad_norm": 0.16203667223453522,
      "learning_rate": 9.319777961787347e-05,
      "loss": 0.9163,
      "mean_token_accuracy": 0.7381542026996613,
      "num_tokens": 70825000.0,
      "step": 2219
    },
    {
      "entropy": 0.8768757879734039,
      "epoch": 0.20425414459951352,
      "grad_norm": 0.1567380130290985,
      "learning_rate": 9.319471279173184e-05,
      "loss": 0.8761,
      "mean_token_accuracy": 0.7476382367312908,
      "num_tokens": 70857203.0,
      "step": 2220
    },
    {
      "entropy": 0.9052177164703608,
      "epoch": 0.20434615097095474,
      "grad_norm": 0.1660914272069931,
      "learning_rate": 9.319164596559022e-05,
      "loss": 0.9015,
      "mean_token_accuracy": 0.7455748319625854,
      "num_tokens": 70888261.0,
      "step": 2221
    },
    {
      "entropy": 0.888258121907711,
      "epoch": 0.20443815734239595,
      "grad_norm": 0.1576509177684784,
      "learning_rate": 9.318857913944859e-05,
      "loss": 0.869,
      "mean_token_accuracy": 0.7503309175372124,
      "num_tokens": 70920133.0,
      "step": 2222
    },
    {
      "entropy": 1.005892638117075,
      "epoch": 0.20453016371383717,
      "grad_norm": 0.16423732042312622,
      "learning_rate": 9.318551231330697e-05,
      "loss": 1.0109,
      "mean_token_accuracy": 0.7182988561689854,
      "num_tokens": 70952169.0,
      "step": 2223
    },
    {
      "entropy": 1.0240471214056015,
      "epoch": 0.20462217008527842,
      "grad_norm": 0.16790643334388733,
      "learning_rate": 9.318244548716534e-05,
      "loss": 1.0153,
      "mean_token_accuracy": 0.7193193435668945,
      "num_tokens": 70983324.0,
      "step": 2224
    },
    {
      "entropy": 0.888104448094964,
      "epoch": 0.20471417645671963,
      "grad_norm": 0.16220274567604065,
      "learning_rate": 9.31793786610237e-05,
      "loss": 0.8974,
      "mean_token_accuracy": 0.7479278296232224,
      "num_tokens": 71015021.0,
      "step": 2225
    },
    {
      "entropy": 0.8313149120658636,
      "epoch": 0.20480618282816085,
      "grad_norm": 0.1571405977010727,
      "learning_rate": 9.317631183488208e-05,
      "loss": 0.8079,
      "mean_token_accuracy": 0.7616389691829681,
      "num_tokens": 71046675.0,
      "step": 2226
    },
    {
      "entropy": 0.8566624037921429,
      "epoch": 0.20489818919960207,
      "grad_norm": 0.1541277915239334,
      "learning_rate": 9.317324500874047e-05,
      "loss": 0.832,
      "mean_token_accuracy": 0.7621605359017849,
      "num_tokens": 71079001.0,
      "step": 2227
    },
    {
      "entropy": 0.8309692051261663,
      "epoch": 0.20499019557104328,
      "grad_norm": 0.1532411277294159,
      "learning_rate": 9.317017818259883e-05,
      "loss": 0.8238,
      "mean_token_accuracy": 0.7626949213445187,
      "num_tokens": 71111649.0,
      "step": 2228
    },
    {
      "entropy": 0.8722108080983162,
      "epoch": 0.20508220194248453,
      "grad_norm": 0.1641416847705841,
      "learning_rate": 9.31671113564572e-05,
      "loss": 0.8808,
      "mean_token_accuracy": 0.7490550726652145,
      "num_tokens": 71143298.0,
      "step": 2229
    },
    {
      "entropy": 0.792917100712657,
      "epoch": 0.20517420831392574,
      "grad_norm": 0.1514943540096283,
      "learning_rate": 9.316404453031558e-05,
      "loss": 0.7799,
      "mean_token_accuracy": 0.7773265764117241,
      "num_tokens": 71176066.0,
      "step": 2230
    },
    {
      "entropy": 0.9234464429318905,
      "epoch": 0.20526621468536696,
      "grad_norm": 0.16224011778831482,
      "learning_rate": 9.316097770417395e-05,
      "loss": 0.9275,
      "mean_token_accuracy": 0.7358816377818584,
      "num_tokens": 71208262.0,
      "step": 2231
    },
    {
      "entropy": 0.9031125567853451,
      "epoch": 0.20535822105680818,
      "grad_norm": 0.16288326680660248,
      "learning_rate": 9.315791087803233e-05,
      "loss": 0.9073,
      "mean_token_accuracy": 0.740313895046711,
      "num_tokens": 71240120.0,
      "step": 2232
    },
    {
      "entropy": 0.8620570674538612,
      "epoch": 0.2054502274282494,
      "grad_norm": 0.1624261736869812,
      "learning_rate": 9.31548440518907e-05,
      "loss": 0.8648,
      "mean_token_accuracy": 0.751829732209444,
      "num_tokens": 71271996.0,
      "step": 2233
    },
    {
      "entropy": 0.9377483390271664,
      "epoch": 0.20554223379969064,
      "grad_norm": 0.1629192978143692,
      "learning_rate": 9.315177722574908e-05,
      "loss": 0.9414,
      "mean_token_accuracy": 0.7359229177236557,
      "num_tokens": 71304514.0,
      "step": 2234
    },
    {
      "entropy": 0.9029636196792126,
      "epoch": 0.20563424017113185,
      "grad_norm": 0.1612558364868164,
      "learning_rate": 9.314871039960745e-05,
      "loss": 0.9059,
      "mean_token_accuracy": 0.7382849045097828,
      "num_tokens": 71336219.0,
      "step": 2235
    },
    {
      "entropy": 0.8020689897239208,
      "epoch": 0.20572624654257307,
      "grad_norm": 0.15994879603385925,
      "learning_rate": 9.314564357346582e-05,
      "loss": 0.8116,
      "mean_token_accuracy": 0.7646960467100143,
      "num_tokens": 71367058.0,
      "step": 2236
    },
    {
      "entropy": 0.7777710407972336,
      "epoch": 0.2058182529140143,
      "grad_norm": 0.1571982502937317,
      "learning_rate": 9.31425767473242e-05,
      "loss": 0.777,
      "mean_token_accuracy": 0.7706207893788815,
      "num_tokens": 71399256.0,
      "step": 2237
    },
    {
      "entropy": 0.8423745967447758,
      "epoch": 0.2059102592854555,
      "grad_norm": 0.1598420888185501,
      "learning_rate": 9.313950992118258e-05,
      "loss": 0.861,
      "mean_token_accuracy": 0.749571718275547,
      "num_tokens": 71431342.0,
      "step": 2238
    },
    {
      "entropy": 0.9289471209049225,
      "epoch": 0.20600226565689675,
      "grad_norm": 0.16045215725898743,
      "learning_rate": 9.313644309504095e-05,
      "loss": 0.9496,
      "mean_token_accuracy": 0.7349057793617249,
      "num_tokens": 71463427.0,
      "step": 2239
    },
    {
      "entropy": 0.8277268670499325,
      "epoch": 0.20609427202833797,
      "grad_norm": 0.15554259717464447,
      "learning_rate": 9.313337626889932e-05,
      "loss": 0.8149,
      "mean_token_accuracy": 0.7654939852654934,
      "num_tokens": 71494702.0,
      "step": 2240
    },
    {
      "entropy": 0.8088647369295359,
      "epoch": 0.20618627839977918,
      "grad_norm": 0.1524478793144226,
      "learning_rate": 9.313030944275768e-05,
      "loss": 0.8007,
      "mean_token_accuracy": 0.7661332301795483,
      "num_tokens": 71526573.0,
      "step": 2241
    },
    {
      "entropy": 0.9212133027613163,
      "epoch": 0.2062782847712204,
      "grad_norm": 0.15735897421836853,
      "learning_rate": 9.312724261661608e-05,
      "loss": 0.906,
      "mean_token_accuracy": 0.742999579757452,
      "num_tokens": 71558207.0,
      "step": 2242
    },
    {
      "entropy": 0.8286278918385506,
      "epoch": 0.20637029114266162,
      "grad_norm": 0.15218910574913025,
      "learning_rate": 9.312417579047445e-05,
      "loss": 0.8006,
      "mean_token_accuracy": 0.764683086425066,
      "num_tokens": 71590404.0,
      "step": 2243
    },
    {
      "entropy": 0.767413817346096,
      "epoch": 0.20646229751410286,
      "grad_norm": 0.14473795890808105,
      "learning_rate": 9.312110896433281e-05,
      "loss": 0.7429,
      "mean_token_accuracy": 0.7815295867621899,
      "num_tokens": 71622641.0,
      "step": 2244
    },
    {
      "entropy": 0.8499306626617908,
      "epoch": 0.20655430388554408,
      "grad_norm": 0.14879648387432098,
      "learning_rate": 9.311804213819118e-05,
      "loss": 0.8185,
      "mean_token_accuracy": 0.7597612328827381,
      "num_tokens": 71655293.0,
      "step": 2245
    },
    {
      "entropy": 0.9770399704575539,
      "epoch": 0.2066463102569853,
      "grad_norm": 0.17101188004016876,
      "learning_rate": 9.311497531204956e-05,
      "loss": 0.9529,
      "mean_token_accuracy": 0.7336248680949211,
      "num_tokens": 71687348.0,
      "step": 2246
    },
    {
      "entropy": 0.9782465323805809,
      "epoch": 0.2067383166284265,
      "grad_norm": 0.1673109084367752,
      "learning_rate": 9.311190848590795e-05,
      "loss": 0.9806,
      "mean_token_accuracy": 0.725510872900486,
      "num_tokens": 71718869.0,
      "step": 2247
    },
    {
      "entropy": 0.772248474881053,
      "epoch": 0.20683032299986773,
      "grad_norm": 0.15195660293102264,
      "learning_rate": 9.310884165976631e-05,
      "loss": 0.7703,
      "mean_token_accuracy": 0.7755439542233944,
      "num_tokens": 71750581.0,
      "step": 2248
    },
    {
      "entropy": 0.8573681581765413,
      "epoch": 0.20692232937130897,
      "grad_norm": 0.1610807180404663,
      "learning_rate": 9.310577483362468e-05,
      "loss": 0.84,
      "mean_token_accuracy": 0.7638273537158966,
      "num_tokens": 71783227.0,
      "step": 2249
    },
    {
      "entropy": 0.8789233788847923,
      "epoch": 0.2070143357427502,
      "grad_norm": 0.1596616804599762,
      "learning_rate": 9.310270800748306e-05,
      "loss": 0.8927,
      "mean_token_accuracy": 0.7488917447626591,
      "num_tokens": 71814523.0,
      "step": 2250
    },
    {
      "entropy": 0.86962079256773,
      "epoch": 0.2071063421141914,
      "grad_norm": 0.16223444044589996,
      "learning_rate": 9.309964118134143e-05,
      "loss": 0.8801,
      "mean_token_accuracy": 0.7479296550154686,
      "num_tokens": 71846858.0,
      "step": 2251
    },
    {
      "entropy": 0.8553214110434055,
      "epoch": 0.20719834848563262,
      "grad_norm": 0.15519845485687256,
      "learning_rate": 9.309657435519981e-05,
      "loss": 0.8186,
      "mean_token_accuracy": 0.7620874978601933,
      "num_tokens": 71879040.0,
      "step": 2252
    },
    {
      "entropy": 0.8920175693929195,
      "epoch": 0.20729035485707384,
      "grad_norm": 0.1649276316165924,
      "learning_rate": 9.309350752905818e-05,
      "loss": 0.9036,
      "mean_token_accuracy": 0.7448780536651611,
      "num_tokens": 71909372.0,
      "step": 2253
    },
    {
      "entropy": 0.8747664242982864,
      "epoch": 0.20738236122851508,
      "grad_norm": 0.160324826836586,
      "learning_rate": 9.309044070291656e-05,
      "loss": 0.8879,
      "mean_token_accuracy": 0.7416230738162994,
      "num_tokens": 71941069.0,
      "step": 2254
    },
    {
      "entropy": 0.8352093640714884,
      "epoch": 0.2074743675999563,
      "grad_norm": 0.16103175282478333,
      "learning_rate": 9.308737387677493e-05,
      "loss": 0.8455,
      "mean_token_accuracy": 0.7612131349742413,
      "num_tokens": 71972639.0,
      "step": 2255
    },
    {
      "entropy": 0.9325528144836426,
      "epoch": 0.20756637397139752,
      "grad_norm": 0.1638745665550232,
      "learning_rate": 9.30843070506333e-05,
      "loss": 0.9293,
      "mean_token_accuracy": 0.7368008457124233,
      "num_tokens": 72004536.0,
      "step": 2256
    },
    {
      "entropy": 0.8350464589893818,
      "epoch": 0.20765838034283873,
      "grad_norm": 0.15912780165672302,
      "learning_rate": 9.308124022449168e-05,
      "loss": 0.8479,
      "mean_token_accuracy": 0.7581521719694138,
      "num_tokens": 72037304.0,
      "step": 2257
    },
    {
      "entropy": 0.8925025351345539,
      "epoch": 0.20775038671427998,
      "grad_norm": 0.15235424041748047,
      "learning_rate": 9.307817339835006e-05,
      "loss": 0.8969,
      "mean_token_accuracy": 0.7427675537765026,
      "num_tokens": 72070044.0,
      "step": 2258
    },
    {
      "entropy": 0.8922431841492653,
      "epoch": 0.2078423930857212,
      "grad_norm": 0.16025975346565247,
      "learning_rate": 9.307510657220843e-05,
      "loss": 0.8991,
      "mean_token_accuracy": 0.7435870468616486,
      "num_tokens": 72101615.0,
      "step": 2259
    },
    {
      "entropy": 0.9367494266480207,
      "epoch": 0.2079343994571624,
      "grad_norm": 0.166617289185524,
      "learning_rate": 9.30720397460668e-05,
      "loss": 0.9368,
      "mean_token_accuracy": 0.7334547564387321,
      "num_tokens": 72133576.0,
      "step": 2260
    },
    {
      "entropy": 0.8854510430246592,
      "epoch": 0.20802640582860363,
      "grad_norm": 0.15313147008419037,
      "learning_rate": 9.306897291992518e-05,
      "loss": 0.8799,
      "mean_token_accuracy": 0.7545266933739185,
      "num_tokens": 72166309.0,
      "step": 2261
    },
    {
      "entropy": 0.9368469603359699,
      "epoch": 0.20811841220004484,
      "grad_norm": 0.15872061252593994,
      "learning_rate": 9.306590609378354e-05,
      "loss": 0.9126,
      "mean_token_accuracy": 0.7365365065634251,
      "num_tokens": 72197234.0,
      "step": 2262
    },
    {
      "entropy": 0.8913379423320293,
      "epoch": 0.2082104185714861,
      "grad_norm": 0.1670869141817093,
      "learning_rate": 9.306283926764193e-05,
      "loss": 0.8894,
      "mean_token_accuracy": 0.7491629645228386,
      "num_tokens": 72228804.0,
      "step": 2263
    },
    {
      "entropy": 0.9516397789120674,
      "epoch": 0.2083024249429273,
      "grad_norm": 0.16078302264213562,
      "learning_rate": 9.30597724415003e-05,
      "loss": 0.9558,
      "mean_token_accuracy": 0.7303243204951286,
      "num_tokens": 72260739.0,
      "step": 2264
    },
    {
      "entropy": 0.9317420907318592,
      "epoch": 0.20839443131436852,
      "grad_norm": 0.16126899421215057,
      "learning_rate": 9.305670561535868e-05,
      "loss": 0.9081,
      "mean_token_accuracy": 0.7433539219200611,
      "num_tokens": 72292862.0,
      "step": 2265
    },
    {
      "entropy": 0.9710980243980885,
      "epoch": 0.20848643768580974,
      "grad_norm": 0.17002113163471222,
      "learning_rate": 9.305363878921704e-05,
      "loss": 0.9886,
      "mean_token_accuracy": 0.7233421914279461,
      "num_tokens": 72324138.0,
      "step": 2266
    },
    {
      "entropy": 0.8377634603530169,
      "epoch": 0.20857844405725096,
      "grad_norm": 0.15324081480503082,
      "learning_rate": 9.305057196307541e-05,
      "loss": 0.8294,
      "mean_token_accuracy": 0.7573648579418659,
      "num_tokens": 72356310.0,
      "step": 2267
    },
    {
      "entropy": 0.9120050743222237,
      "epoch": 0.2086704504286922,
      "grad_norm": 0.16222664713859558,
      "learning_rate": 9.304750513693379e-05,
      "loss": 0.8971,
      "mean_token_accuracy": 0.7498729303479195,
      "num_tokens": 72388488.0,
      "step": 2268
    },
    {
      "entropy": 0.8644157573580742,
      "epoch": 0.20876245680013342,
      "grad_norm": 0.15668171644210815,
      "learning_rate": 9.304443831079217e-05,
      "loss": 0.8417,
      "mean_token_accuracy": 0.7566785216331482,
      "num_tokens": 72420143.0,
      "step": 2269
    },
    {
      "entropy": 0.781281266361475,
      "epoch": 0.20885446317157463,
      "grad_norm": 0.15086975693702698,
      "learning_rate": 9.304137148465054e-05,
      "loss": 0.8069,
      "mean_token_accuracy": 0.7666926644742489,
      "num_tokens": 72452582.0,
      "step": 2270
    },
    {
      "entropy": 0.9027824606746435,
      "epoch": 0.20894646954301585,
      "grad_norm": 0.16109751164913177,
      "learning_rate": 9.303830465850891e-05,
      "loss": 0.904,
      "mean_token_accuracy": 0.7460171282291412,
      "num_tokens": 72484321.0,
      "step": 2271
    },
    {
      "entropy": 0.8155895974487066,
      "epoch": 0.20903847591445707,
      "grad_norm": 0.15224795043468475,
      "learning_rate": 9.303523783236728e-05,
      "loss": 0.7951,
      "mean_token_accuracy": 0.7695300467312336,
      "num_tokens": 72516459.0,
      "step": 2272
    },
    {
      "entropy": 0.8761390671133995,
      "epoch": 0.2091304822858983,
      "grad_norm": 0.16741062700748444,
      "learning_rate": 9.303217100622566e-05,
      "loss": 0.8874,
      "mean_token_accuracy": 0.7487536258995533,
      "num_tokens": 72548342.0,
      "step": 2273
    },
    {
      "entropy": 0.8773033432662487,
      "epoch": 0.20922248865733953,
      "grad_norm": 0.15922823548316956,
      "learning_rate": 9.302910418008404e-05,
      "loss": 0.8893,
      "mean_token_accuracy": 0.752545677125454,
      "num_tokens": 72580183.0,
      "step": 2274
    },
    {
      "entropy": 0.8742824718356133,
      "epoch": 0.20931449502878074,
      "grad_norm": 0.16291575133800507,
      "learning_rate": 9.302603735394241e-05,
      "loss": 0.8785,
      "mean_token_accuracy": 0.7491361945867538,
      "num_tokens": 72612656.0,
      "step": 2275
    },
    {
      "entropy": 0.9169346243143082,
      "epoch": 0.20940650140022196,
      "grad_norm": 0.1646944135427475,
      "learning_rate": 9.302297052780078e-05,
      "loss": 0.9211,
      "mean_token_accuracy": 0.7414053231477737,
      "num_tokens": 72643937.0,
      "step": 2276
    },
    {
      "entropy": 0.8519726824015379,
      "epoch": 0.20949850777166318,
      "grad_norm": 0.15469864010810852,
      "learning_rate": 9.301990370165916e-05,
      "loss": 0.8582,
      "mean_token_accuracy": 0.7548933997750282,
      "num_tokens": 72676349.0,
      "step": 2277
    },
    {
      "entropy": 0.8380433060228825,
      "epoch": 0.20959051414310442,
      "grad_norm": 0.1525021344423294,
      "learning_rate": 9.301683687551754e-05,
      "loss": 0.848,
      "mean_token_accuracy": 0.7557270415127277,
      "num_tokens": 72708525.0,
      "step": 2278
    },
    {
      "entropy": 0.8926225416362286,
      "epoch": 0.20968252051454564,
      "grad_norm": 0.15022622048854828,
      "learning_rate": 9.301377004937591e-05,
      "loss": 0.8716,
      "mean_token_accuracy": 0.7522858828306198,
      "num_tokens": 72740871.0,
      "step": 2279
    },
    {
      "entropy": 0.8470750600099564,
      "epoch": 0.20977452688598686,
      "grad_norm": 0.15597392618656158,
      "learning_rate": 9.301070322323427e-05,
      "loss": 0.849,
      "mean_token_accuracy": 0.759568840265274,
      "num_tokens": 72773008.0,
      "step": 2280
    },
    {
      "entropy": 0.9882756508886814,
      "epoch": 0.20986653325742807,
      "grad_norm": 0.16542378067970276,
      "learning_rate": 9.300763639709266e-05,
      "loss": 0.9775,
      "mean_token_accuracy": 0.7228252813220024,
      "num_tokens": 72804480.0,
      "step": 2281
    },
    {
      "entropy": 0.93464151956141,
      "epoch": 0.2099585396288693,
      "grad_norm": 0.15854597091674805,
      "learning_rate": 9.300456957095102e-05,
      "loss": 0.931,
      "mean_token_accuracy": 0.7351988777518272,
      "num_tokens": 72836307.0,
      "step": 2282
    },
    {
      "entropy": 0.8390575088560581,
      "epoch": 0.21005054600031053,
      "grad_norm": 0.15083849430084229,
      "learning_rate": 9.30015027448094e-05,
      "loss": 0.8259,
      "mean_token_accuracy": 0.7617522925138474,
      "num_tokens": 72868295.0,
      "step": 2283
    },
    {
      "entropy": 0.8471817430108786,
      "epoch": 0.21014255237175175,
      "grad_norm": 0.1499129980802536,
      "learning_rate": 9.299843591866777e-05,
      "loss": 0.8322,
      "mean_token_accuracy": 0.7625635303556919,
      "num_tokens": 72900440.0,
      "step": 2284
    },
    {
      "entropy": 0.8549118041992188,
      "epoch": 0.21023455874319297,
      "grad_norm": 0.15557771921157837,
      "learning_rate": 9.299536909252615e-05,
      "loss": 0.8702,
      "mean_token_accuracy": 0.7528311014175415,
      "num_tokens": 72932502.0,
      "step": 2285
    },
    {
      "entropy": 0.8649424687027931,
      "epoch": 0.21032656511463418,
      "grad_norm": 0.15368221700191498,
      "learning_rate": 9.299230226638452e-05,
      "loss": 0.8499,
      "mean_token_accuracy": 0.7546233087778091,
      "num_tokens": 72964860.0,
      "step": 2286
    },
    {
      "entropy": 0.8734093382954597,
      "epoch": 0.2104185714860754,
      "grad_norm": 0.1616574376821518,
      "learning_rate": 9.298923544024289e-05,
      "loss": 0.8894,
      "mean_token_accuracy": 0.7474319748580456,
      "num_tokens": 72996730.0,
      "step": 2287
    },
    {
      "entropy": 0.7782254759222269,
      "epoch": 0.21051057785751665,
      "grad_norm": 0.15870293974876404,
      "learning_rate": 9.298616861410127e-05,
      "loss": 0.7732,
      "mean_token_accuracy": 0.772123996168375,
      "num_tokens": 73028728.0,
      "step": 2288
    },
    {
      "entropy": 0.8326247707009315,
      "epoch": 0.21060258422895786,
      "grad_norm": 0.16166287660598755,
      "learning_rate": 9.298310178795965e-05,
      "loss": 0.841,
      "mean_token_accuracy": 0.7570547163486481,
      "num_tokens": 73061124.0,
      "step": 2289
    },
    {
      "entropy": 0.8742987290024757,
      "epoch": 0.21069459060039908,
      "grad_norm": 0.16501806676387787,
      "learning_rate": 9.298003496181802e-05,
      "loss": 0.8592,
      "mean_token_accuracy": 0.7546672336757183,
      "num_tokens": 73093062.0,
      "step": 2290
    },
    {
      "entropy": 0.8095121756196022,
      "epoch": 0.2107865969718403,
      "grad_norm": 0.16070586442947388,
      "learning_rate": 9.297696813567639e-05,
      "loss": 0.8176,
      "mean_token_accuracy": 0.7627358548343182,
      "num_tokens": 73125014.0,
      "step": 2291
    },
    {
      "entropy": 0.8649747185409069,
      "epoch": 0.2108786033432815,
      "grad_norm": 0.15434929728507996,
      "learning_rate": 9.297390130953477e-05,
      "loss": 0.8608,
      "mean_token_accuracy": 0.7530300579965115,
      "num_tokens": 73157677.0,
      "step": 2292
    },
    {
      "entropy": 0.842358898371458,
      "epoch": 0.21097060971472276,
      "grad_norm": 0.1575389951467514,
      "learning_rate": 9.297083448339314e-05,
      "loss": 0.8322,
      "mean_token_accuracy": 0.7612573653459549,
      "num_tokens": 73188954.0,
      "step": 2293
    },
    {
      "entropy": 0.8657597675919533,
      "epoch": 0.21106261608616397,
      "grad_norm": 0.1545700877904892,
      "learning_rate": 9.296776765725152e-05,
      "loss": 0.8701,
      "mean_token_accuracy": 0.7497542388737202,
      "num_tokens": 73220926.0,
      "step": 2294
    },
    {
      "entropy": 0.8375305347144604,
      "epoch": 0.2111546224576052,
      "grad_norm": 0.15892831981182098,
      "learning_rate": 9.296470083110989e-05,
      "loss": 0.8474,
      "mean_token_accuracy": 0.7611798197031021,
      "num_tokens": 73253300.0,
      "step": 2295
    },
    {
      "entropy": 0.8946369141340256,
      "epoch": 0.2112466288290464,
      "grad_norm": 0.1711026132106781,
      "learning_rate": 9.296163400496827e-05,
      "loss": 0.9175,
      "mean_token_accuracy": 0.7415480017662048,
      "num_tokens": 73284946.0,
      "step": 2296
    },
    {
      "entropy": 0.8042214475572109,
      "epoch": 0.21133863520048762,
      "grad_norm": 0.15280771255493164,
      "learning_rate": 9.295856717882664e-05,
      "loss": 0.792,
      "mean_token_accuracy": 0.7685409113764763,
      "num_tokens": 73316638.0,
      "step": 2297
    },
    {
      "entropy": 0.7593127936124802,
      "epoch": 0.21143064157192887,
      "grad_norm": 0.15070487558841705,
      "learning_rate": 9.2955500352685e-05,
      "loss": 0.768,
      "mean_token_accuracy": 0.7812102846801281,
      "num_tokens": 73348759.0,
      "step": 2298
    },
    {
      "entropy": 0.7899413146078587,
      "epoch": 0.21152264794337008,
      "grad_norm": 0.15733014047145844,
      "learning_rate": 9.295243352654339e-05,
      "loss": 0.779,
      "mean_token_accuracy": 0.7734876573085785,
      "num_tokens": 73380889.0,
      "step": 2299
    },
    {
      "entropy": 0.8632377721369267,
      "epoch": 0.2116146543148113,
      "grad_norm": 0.16262930631637573,
      "learning_rate": 9.294936670040177e-05,
      "loss": 0.8576,
      "mean_token_accuracy": 0.7527768313884735,
      "num_tokens": 73412654.0,
      "step": 2300
    },
    {
      "entropy": 0.9495619386434555,
      "epoch": 0.21170666068625252,
      "grad_norm": 0.16623176634311676,
      "learning_rate": 9.294629987426014e-05,
      "loss": 0.9475,
      "mean_token_accuracy": 0.7298998758196831,
      "num_tokens": 73445234.0,
      "step": 2301
    },
    {
      "entropy": 0.8101659417152405,
      "epoch": 0.21179866705769373,
      "grad_norm": 0.15964554250240326,
      "learning_rate": 9.29432330481185e-05,
      "loss": 0.825,
      "mean_token_accuracy": 0.7656626999378204,
      "num_tokens": 73477714.0,
      "step": 2302
    },
    {
      "entropy": 0.8644240833818913,
      "epoch": 0.21189067342913498,
      "grad_norm": 0.15561071038246155,
      "learning_rate": 9.294016622197687e-05,
      "loss": 0.8357,
      "mean_token_accuracy": 0.7615216374397278,
      "num_tokens": 73510164.0,
      "step": 2303
    },
    {
      "entropy": 0.8193222731351852,
      "epoch": 0.2119826798005762,
      "grad_norm": 0.1511794924736023,
      "learning_rate": 9.293709939583525e-05,
      "loss": 0.7889,
      "mean_token_accuracy": 0.7722503058612347,
      "num_tokens": 73542158.0,
      "step": 2304
    },
    {
      "entropy": 0.9018499404191971,
      "epoch": 0.2120746861720174,
      "grad_norm": 0.1594957411289215,
      "learning_rate": 9.293403256969363e-05,
      "loss": 0.8728,
      "mean_token_accuracy": 0.7469648867845535,
      "num_tokens": 73573589.0,
      "step": 2305
    },
    {
      "entropy": 0.8220512084662914,
      "epoch": 0.21216669254345863,
      "grad_norm": 0.1603728085756302,
      "learning_rate": 9.2930965743552e-05,
      "loss": 0.8124,
      "mean_token_accuracy": 0.7637932151556015,
      "num_tokens": 73605323.0,
      "step": 2306
    },
    {
      "entropy": 0.8679672721773386,
      "epoch": 0.21225869891489985,
      "grad_norm": 0.16267962753772736,
      "learning_rate": 9.292789891741037e-05,
      "loss": 0.8783,
      "mean_token_accuracy": 0.7501159161329269,
      "num_tokens": 73636547.0,
      "step": 2307
    },
    {
      "entropy": 0.8786847665905952,
      "epoch": 0.2123507052863411,
      "grad_norm": 0.16042248904705048,
      "learning_rate": 9.292483209126875e-05,
      "loss": 0.8874,
      "mean_token_accuracy": 0.7459948137402534,
      "num_tokens": 73668141.0,
      "step": 2308
    },
    {
      "entropy": 0.8194758594036102,
      "epoch": 0.2124427116577823,
      "grad_norm": 0.16400262713432312,
      "learning_rate": 9.292176526512712e-05,
      "loss": 0.8422,
      "mean_token_accuracy": 0.7582703903317451,
      "num_tokens": 73699835.0,
      "step": 2309
    },
    {
      "entropy": 0.7984047047793865,
      "epoch": 0.21253471802922352,
      "grad_norm": 0.15341556072235107,
      "learning_rate": 9.29186984389855e-05,
      "loss": 0.7955,
      "mean_token_accuracy": 0.765237107872963,
      "num_tokens": 73731684.0,
      "step": 2310
    },
    {
      "entropy": 0.8519179932773113,
      "epoch": 0.21262672440066474,
      "grad_norm": 0.15963229537010193,
      "learning_rate": 9.291563161284387e-05,
      "loss": 0.8656,
      "mean_token_accuracy": 0.7528999224305153,
      "num_tokens": 73763388.0,
      "step": 2311
    },
    {
      "entropy": 0.8162442464381456,
      "epoch": 0.21271873077210596,
      "grad_norm": 0.15015247464179993,
      "learning_rate": 9.291256478670225e-05,
      "loss": 0.8058,
      "mean_token_accuracy": 0.7635101526975632,
      "num_tokens": 73795412.0,
      "step": 2312
    },
    {
      "entropy": 0.8530397266149521,
      "epoch": 0.2128107371435472,
      "grad_norm": 0.15462297201156616,
      "learning_rate": 9.290949796056062e-05,
      "loss": 0.8454,
      "mean_token_accuracy": 0.7592302151024342,
      "num_tokens": 73827598.0,
      "step": 2313
    },
    {
      "entropy": 0.8685322217643261,
      "epoch": 0.21290274351498842,
      "grad_norm": 0.1597013920545578,
      "learning_rate": 9.290643113441899e-05,
      "loss": 0.8376,
      "mean_token_accuracy": 0.7622593753039837,
      "num_tokens": 73859136.0,
      "step": 2314
    },
    {
      "entropy": 0.7380099222064018,
      "epoch": 0.21299474988642964,
      "grad_norm": 0.1455698311328888,
      "learning_rate": 9.290336430827737e-05,
      "loss": 0.7036,
      "mean_token_accuracy": 0.7940162383019924,
      "num_tokens": 73891495.0,
      "step": 2315
    },
    {
      "entropy": 0.8977028522640467,
      "epoch": 0.21308675625787085,
      "grad_norm": 0.1637427657842636,
      "learning_rate": 9.290029748213575e-05,
      "loss": 0.9074,
      "mean_token_accuracy": 0.7449595183134079,
      "num_tokens": 73923166.0,
      "step": 2316
    },
    {
      "entropy": 0.8406349718570709,
      "epoch": 0.21317876262931207,
      "grad_norm": 0.16535893082618713,
      "learning_rate": 9.289723065599412e-05,
      "loss": 0.8519,
      "mean_token_accuracy": 0.7539288587868214,
      "num_tokens": 73953835.0,
      "step": 2317
    },
    {
      "entropy": 0.8481707274913788,
      "epoch": 0.2132707690007533,
      "grad_norm": 0.15947511792182922,
      "learning_rate": 9.289416382985248e-05,
      "loss": 0.852,
      "mean_token_accuracy": 0.7579649947583675,
      "num_tokens": 73985238.0,
      "step": 2318
    },
    {
      "entropy": 0.9580808281898499,
      "epoch": 0.21336277537219453,
      "grad_norm": 0.16335488855838776,
      "learning_rate": 9.289109700371085e-05,
      "loss": 0.9511,
      "mean_token_accuracy": 0.7323847934603691,
      "num_tokens": 74017436.0,
      "step": 2319
    },
    {
      "entropy": 0.9189779311418533,
      "epoch": 0.21345478174363575,
      "grad_norm": 0.16445888578891754,
      "learning_rate": 9.288803017756925e-05,
      "loss": 0.935,
      "mean_token_accuracy": 0.7369248531758785,
      "num_tokens": 74049574.0,
      "step": 2320
    },
    {
      "entropy": 0.8613135479390621,
      "epoch": 0.21354678811507696,
      "grad_norm": 0.15711162984371185,
      "learning_rate": 9.288496335142762e-05,
      "loss": 0.8583,
      "mean_token_accuracy": 0.7517976872622967,
      "num_tokens": 74080706.0,
      "step": 2321
    },
    {
      "entropy": 0.9118535481393337,
      "epoch": 0.21363879448651818,
      "grad_norm": 0.16483868658542633,
      "learning_rate": 9.288189652528598e-05,
      "loss": 0.9364,
      "mean_token_accuracy": 0.7352632023394108,
      "num_tokens": 74112438.0,
      "step": 2322
    },
    {
      "entropy": 0.8229013569653034,
      "epoch": 0.21373080085795942,
      "grad_norm": 0.15979285538196564,
      "learning_rate": 9.287882969914436e-05,
      "loss": 0.8267,
      "mean_token_accuracy": 0.7661837078630924,
      "num_tokens": 74144585.0,
      "step": 2323
    },
    {
      "entropy": 0.7569926884025335,
      "epoch": 0.21382280722940064,
      "grad_norm": 0.1473347693681717,
      "learning_rate": 9.287576287300273e-05,
      "loss": 0.7442,
      "mean_token_accuracy": 0.7849711775779724,
      "num_tokens": 74177028.0,
      "step": 2324
    },
    {
      "entropy": 0.8404094390571117,
      "epoch": 0.21391481360084186,
      "grad_norm": 0.15971264243125916,
      "learning_rate": 9.287269604686111e-05,
      "loss": 0.8493,
      "mean_token_accuracy": 0.7552695386111736,
      "num_tokens": 74208864.0,
      "step": 2325
    },
    {
      "entropy": 0.8926186412572861,
      "epoch": 0.21400681997228307,
      "grad_norm": 0.16540151834487915,
      "learning_rate": 9.286962922071948e-05,
      "loss": 0.8802,
      "mean_token_accuracy": 0.7493114471435547,
      "num_tokens": 74240403.0,
      "step": 2326
    },
    {
      "entropy": 0.7373359240591526,
      "epoch": 0.2140988263437243,
      "grad_norm": 0.14905647933483124,
      "learning_rate": 9.286656239457786e-05,
      "loss": 0.7283,
      "mean_token_accuracy": 0.7870700471103191,
      "num_tokens": 74272259.0,
      "step": 2327
    },
    {
      "entropy": 0.8801760859787464,
      "epoch": 0.21419083271516554,
      "grad_norm": 0.15390077233314514,
      "learning_rate": 9.286349556843623e-05,
      "loss": 0.8536,
      "mean_token_accuracy": 0.7569510787725449,
      "num_tokens": 74304433.0,
      "step": 2328
    },
    {
      "entropy": 0.9202699735760689,
      "epoch": 0.21428283908660675,
      "grad_norm": 0.16361750662326813,
      "learning_rate": 9.28604287422946e-05,
      "loss": 0.8915,
      "mean_token_accuracy": 0.7435425333678722,
      "num_tokens": 74336750.0,
      "step": 2329
    },
    {
      "entropy": 0.9361569508910179,
      "epoch": 0.21437484545804797,
      "grad_norm": 0.1674267053604126,
      "learning_rate": 9.285736191615298e-05,
      "loss": 0.9318,
      "mean_token_accuracy": 0.7335952706634998,
      "num_tokens": 74368579.0,
      "step": 2330
    },
    {
      "entropy": 0.888928234577179,
      "epoch": 0.21446685182948919,
      "grad_norm": 0.15682590007781982,
      "learning_rate": 9.285429509001136e-05,
      "loss": 0.8822,
      "mean_token_accuracy": 0.7466441057622433,
      "num_tokens": 74400507.0,
      "step": 2331
    },
    {
      "entropy": 0.9294977150857449,
      "epoch": 0.2145588582009304,
      "grad_norm": 0.16409148275852203,
      "learning_rate": 9.285122826386973e-05,
      "loss": 0.9231,
      "mean_token_accuracy": 0.7371314279735088,
      "num_tokens": 74432296.0,
      "step": 2332
    },
    {
      "entropy": 0.8925334792584181,
      "epoch": 0.21465086457237165,
      "grad_norm": 0.15958143770694733,
      "learning_rate": 9.28481614377281e-05,
      "loss": 0.8759,
      "mean_token_accuracy": 0.7475364282727242,
      "num_tokens": 74463468.0,
      "step": 2333
    },
    {
      "entropy": 0.9574099369347095,
      "epoch": 0.21474287094381286,
      "grad_norm": 0.16768424212932587,
      "learning_rate": 9.284509461158647e-05,
      "loss": 0.9739,
      "mean_token_accuracy": 0.7254521921277046,
      "num_tokens": 74495697.0,
      "step": 2334
    },
    {
      "entropy": 0.8997032791376114,
      "epoch": 0.21483487731525408,
      "grad_norm": 0.1698794960975647,
      "learning_rate": 9.284202778544485e-05,
      "loss": 0.9067,
      "mean_token_accuracy": 0.7409320212900639,
      "num_tokens": 74527352.0,
      "step": 2335
    },
    {
      "entropy": 0.8077906928956509,
      "epoch": 0.2149268836866953,
      "grad_norm": 0.16342999041080475,
      "learning_rate": 9.283896095930323e-05,
      "loss": 0.8289,
      "mean_token_accuracy": 0.7630523443222046,
      "num_tokens": 74560026.0,
      "step": 2336
    },
    {
      "entropy": 0.9348451122641563,
      "epoch": 0.2150188900581365,
      "grad_norm": 0.1619235724210739,
      "learning_rate": 9.28358941331616e-05,
      "loss": 0.9429,
      "mean_token_accuracy": 0.7320568636059761,
      "num_tokens": 74591763.0,
      "step": 2337
    },
    {
      "entropy": 0.8356334138661623,
      "epoch": 0.21511089642957776,
      "grad_norm": 0.15647444128990173,
      "learning_rate": 9.283282730701996e-05,
      "loss": 0.8435,
      "mean_token_accuracy": 0.7583103813230991,
      "num_tokens": 74623598.0,
      "step": 2338
    },
    {
      "entropy": 0.8339238446205854,
      "epoch": 0.21520290280101897,
      "grad_norm": 0.16077212989330292,
      "learning_rate": 9.282976048087835e-05,
      "loss": 0.8345,
      "mean_token_accuracy": 0.7600149549543858,
      "num_tokens": 74655563.0,
      "step": 2339
    },
    {
      "entropy": 0.8778735399246216,
      "epoch": 0.2152949091724602,
      "grad_norm": 0.15700973570346832,
      "learning_rate": 9.282669365473671e-05,
      "loss": 0.833,
      "mean_token_accuracy": 0.7605788297951221,
      "num_tokens": 74687651.0,
      "step": 2340
    },
    {
      "entropy": 0.9297196641564369,
      "epoch": 0.2153869155439014,
      "grad_norm": 0.15686491131782532,
      "learning_rate": 9.28236268285951e-05,
      "loss": 0.9173,
      "mean_token_accuracy": 0.7417204901576042,
      "num_tokens": 74719457.0,
      "step": 2341
    },
    {
      "entropy": 0.8720533959567547,
      "epoch": 0.21547892191534262,
      "grad_norm": 0.1595546305179596,
      "learning_rate": 9.282056000245346e-05,
      "loss": 0.8623,
      "mean_token_accuracy": 0.7479305639863014,
      "num_tokens": 74751484.0,
      "step": 2342
    },
    {
      "entropy": 0.8389994129538536,
      "epoch": 0.21557092828678387,
      "grad_norm": 0.15346363186836243,
      "learning_rate": 9.281749317631184e-05,
      "loss": 0.8305,
      "mean_token_accuracy": 0.758939802646637,
      "num_tokens": 74783868.0,
      "step": 2343
    },
    {
      "entropy": 0.9312886856496334,
      "epoch": 0.2156629346582251,
      "grad_norm": 0.16573521494865417,
      "learning_rate": 9.281442635017021e-05,
      "loss": 0.9295,
      "mean_token_accuracy": 0.7392262890934944,
      "num_tokens": 74816071.0,
      "step": 2344
    },
    {
      "entropy": 0.9091213829815388,
      "epoch": 0.2157549410296663,
      "grad_norm": 0.15877792239189148,
      "learning_rate": 9.281135952402858e-05,
      "loss": 0.8818,
      "mean_token_accuracy": 0.7479448355734348,
      "num_tokens": 74848227.0,
      "step": 2345
    },
    {
      "entropy": 0.8529758080840111,
      "epoch": 0.21584694740110752,
      "grad_norm": 0.15783029794692993,
      "learning_rate": 9.280829269788696e-05,
      "loss": 0.8357,
      "mean_token_accuracy": 0.758288387209177,
      "num_tokens": 74879920.0,
      "step": 2346
    },
    {
      "entropy": 0.8285192921757698,
      "epoch": 0.21593895377254874,
      "grad_norm": 0.158837229013443,
      "learning_rate": 9.280522587174534e-05,
      "loss": 0.8375,
      "mean_token_accuracy": 0.7608655840158463,
      "num_tokens": 74912300.0,
      "step": 2347
    },
    {
      "entropy": 0.8837095201015472,
      "epoch": 0.21603096014398998,
      "grad_norm": 0.16371788084506989,
      "learning_rate": 9.280215904560371e-05,
      "loss": 0.901,
      "mean_token_accuracy": 0.7441506832838058,
      "num_tokens": 74944138.0,
      "step": 2348
    },
    {
      "entropy": 0.8959066495299339,
      "epoch": 0.2161229665154312,
      "grad_norm": 0.16035102307796478,
      "learning_rate": 9.279909221946208e-05,
      "loss": 0.9126,
      "mean_token_accuracy": 0.7381299212574959,
      "num_tokens": 74976578.0,
      "step": 2349
    },
    {
      "entropy": 0.8476665876805782,
      "epoch": 0.21621497288687241,
      "grad_norm": 0.15813155472278595,
      "learning_rate": 9.279602539332046e-05,
      "loss": 0.8709,
      "mean_token_accuracy": 0.7531716749072075,
      "num_tokens": 75008507.0,
      "step": 2350
    },
    {
      "entropy": 0.8305793777108192,
      "epoch": 0.21630697925831363,
      "grad_norm": 0.1581171452999115,
      "learning_rate": 9.279295856717883e-05,
      "loss": 0.8218,
      "mean_token_accuracy": 0.7596693150699139,
      "num_tokens": 75039476.0,
      "step": 2351
    },
    {
      "entropy": 0.9214646555483341,
      "epoch": 0.21639898562975485,
      "grad_norm": 0.15791237354278564,
      "learning_rate": 9.278989174103721e-05,
      "loss": 0.9353,
      "mean_token_accuracy": 0.7329492978751659,
      "num_tokens": 75070929.0,
      "step": 2352
    },
    {
      "entropy": 0.785088799893856,
      "epoch": 0.2164909920011961,
      "grad_norm": 0.159236341714859,
      "learning_rate": 9.278682491489558e-05,
      "loss": 0.8017,
      "mean_token_accuracy": 0.7744565196335316,
      "num_tokens": 75103697.0,
      "step": 2353
    },
    {
      "entropy": 0.8934606835246086,
      "epoch": 0.2165829983726373,
      "grad_norm": 0.1600653976202011,
      "learning_rate": 9.278375808875396e-05,
      "loss": 0.8874,
      "mean_token_accuracy": 0.7499939389526844,
      "num_tokens": 75135945.0,
      "step": 2354
    },
    {
      "entropy": 0.8991423435509205,
      "epoch": 0.21667500474407853,
      "grad_norm": 0.16743279993534088,
      "learning_rate": 9.278069126261233e-05,
      "loss": 0.898,
      "mean_token_accuracy": 0.7448224611580372,
      "num_tokens": 75167490.0,
      "step": 2355
    },
    {
      "entropy": 0.9201076962053776,
      "epoch": 0.21676701111551974,
      "grad_norm": 0.16606201231479645,
      "learning_rate": 9.27776244364707e-05,
      "loss": 0.9541,
      "mean_token_accuracy": 0.7341414354741573,
      "num_tokens": 75199313.0,
      "step": 2356
    },
    {
      "entropy": 0.9680018480867147,
      "epoch": 0.21685901748696096,
      "grad_norm": 0.1589997112751007,
      "learning_rate": 9.277455761032908e-05,
      "loss": 0.9533,
      "mean_token_accuracy": 0.7317590825259686,
      "num_tokens": 75231759.0,
      "step": 2357
    },
    {
      "entropy": 0.8960989713668823,
      "epoch": 0.2169510238584022,
      "grad_norm": 0.15569919347763062,
      "learning_rate": 9.277149078418746e-05,
      "loss": 0.8886,
      "mean_token_accuracy": 0.7453737258911133,
      "num_tokens": 75263606.0,
      "step": 2358
    },
    {
      "entropy": 0.9482849277555943,
      "epoch": 0.21704303022984342,
      "grad_norm": 0.15724118053913116,
      "learning_rate": 9.276842395804582e-05,
      "loss": 0.9256,
      "mean_token_accuracy": 0.736133761703968,
      "num_tokens": 75295807.0,
      "step": 2359
    },
    {
      "entropy": 1.024195782840252,
      "epoch": 0.21713503660128464,
      "grad_norm": 0.16403645277023315,
      "learning_rate": 9.276535713190419e-05,
      "loss": 1.0137,
      "mean_token_accuracy": 0.7117209136486053,
      "num_tokens": 75327731.0,
      "step": 2360
    },
    {
      "entropy": 0.9239401072263718,
      "epoch": 0.21722704297272585,
      "grad_norm": 0.16181538999080658,
      "learning_rate": 9.276229030576256e-05,
      "loss": 0.93,
      "mean_token_accuracy": 0.7377127967774868,
      "num_tokens": 75360173.0,
      "step": 2361
    },
    {
      "entropy": 0.8845912162214518,
      "epoch": 0.21731904934416707,
      "grad_norm": 0.15625187754631042,
      "learning_rate": 9.275922347962096e-05,
      "loss": 0.8607,
      "mean_token_accuracy": 0.7501765973865986,
      "num_tokens": 75391747.0,
      "step": 2362
    },
    {
      "entropy": 0.8475316539406776,
      "epoch": 0.21741105571560831,
      "grad_norm": 0.1548399180173874,
      "learning_rate": 9.275615665347932e-05,
      "loss": 0.8407,
      "mean_token_accuracy": 0.760573036968708,
      "num_tokens": 75423842.0,
      "step": 2363
    },
    {
      "entropy": 0.8303390741348267,
      "epoch": 0.21750306208704953,
      "grad_norm": 0.15964899957180023,
      "learning_rate": 9.275308982733769e-05,
      "loss": 0.8039,
      "mean_token_accuracy": 0.7697515487670898,
      "num_tokens": 75456004.0,
      "step": 2364
    },
    {
      "entropy": 0.8924689441919327,
      "epoch": 0.21759506845849075,
      "grad_norm": 0.16586615145206451,
      "learning_rate": 9.275002300119606e-05,
      "loss": 0.9201,
      "mean_token_accuracy": 0.7366299442946911,
      "num_tokens": 75488432.0,
      "step": 2365
    },
    {
      "entropy": 0.891871515661478,
      "epoch": 0.21768707482993196,
      "grad_norm": 0.15239176154136658,
      "learning_rate": 9.274695617505444e-05,
      "loss": 0.8565,
      "mean_token_accuracy": 0.7589692138135433,
      "num_tokens": 75520614.0,
      "step": 2366
    },
    {
      "entropy": 0.8693595491349697,
      "epoch": 0.21777908120137318,
      "grad_norm": 0.15834809839725494,
      "learning_rate": 9.274388934891282e-05,
      "loss": 0.8673,
      "mean_token_accuracy": 0.7509845718741417,
      "num_tokens": 75552933.0,
      "step": 2367
    },
    {
      "entropy": 0.9256998375058174,
      "epoch": 0.21787108757281443,
      "grad_norm": 0.16354398429393768,
      "learning_rate": 9.274082252277119e-05,
      "loss": 0.9154,
      "mean_token_accuracy": 0.7432111985981464,
      "num_tokens": 75584888.0,
      "step": 2368
    },
    {
      "entropy": 0.8945051468908787,
      "epoch": 0.21796309394425564,
      "grad_norm": 0.16683915257453918,
      "learning_rate": 9.273775569662956e-05,
      "loss": 0.8942,
      "mean_token_accuracy": 0.7481366842985153,
      "num_tokens": 75616177.0,
      "step": 2369
    },
    {
      "entropy": 0.7913928367197514,
      "epoch": 0.21805510031569686,
      "grad_norm": 0.15472346544265747,
      "learning_rate": 9.273468887048794e-05,
      "loss": 0.8021,
      "mean_token_accuracy": 0.7681840658187866,
      "num_tokens": 75647743.0,
      "step": 2370
    },
    {
      "entropy": 0.9697216413915157,
      "epoch": 0.21814710668713808,
      "grad_norm": 0.16356663405895233,
      "learning_rate": 9.273162204434631e-05,
      "loss": 0.9502,
      "mean_token_accuracy": 0.7296673133969307,
      "num_tokens": 75679604.0,
      "step": 2371
    },
    {
      "entropy": 0.8124455139040947,
      "epoch": 0.21823911305857932,
      "grad_norm": 0.15782524645328522,
      "learning_rate": 9.272855521820469e-05,
      "loss": 0.8105,
      "mean_token_accuracy": 0.7699906527996063,
      "num_tokens": 75712198.0,
      "step": 2372
    },
    {
      "entropy": 0.877621166408062,
      "epoch": 0.21833111943002054,
      "grad_norm": 0.16561342775821686,
      "learning_rate": 9.272548839206306e-05,
      "loss": 0.8674,
      "mean_token_accuracy": 0.7490702718496323,
      "num_tokens": 75744211.0,
      "step": 2373
    },
    {
      "entropy": 0.8295040968805552,
      "epoch": 0.21842312580146175,
      "grad_norm": 0.16530773043632507,
      "learning_rate": 9.272242156592144e-05,
      "loss": 0.8433,
      "mean_token_accuracy": 0.7572709769010544,
      "num_tokens": 75775605.0,
      "step": 2374
    },
    {
      "entropy": 0.8505012206733227,
      "epoch": 0.21851513217290297,
      "grad_norm": 0.15716202557086945,
      "learning_rate": 9.27193547397798e-05,
      "loss": 0.8407,
      "mean_token_accuracy": 0.7608695663511753,
      "num_tokens": 75808373.0,
      "step": 2375
    },
    {
      "entropy": 0.8264030143618584,
      "epoch": 0.2186071385443442,
      "grad_norm": 0.16426382958889008,
      "learning_rate": 9.271628791363817e-05,
      "loss": 0.8271,
      "mean_token_accuracy": 0.7618039399385452,
      "num_tokens": 75839529.0,
      "step": 2376
    },
    {
      "entropy": 0.8500331602990627,
      "epoch": 0.21869914491578543,
      "grad_norm": 0.15977241098880768,
      "learning_rate": 9.271322108749655e-05,
      "loss": 0.8317,
      "mean_token_accuracy": 0.7604336328804493,
      "num_tokens": 75870779.0,
      "step": 2377
    },
    {
      "entropy": 0.8669287972152233,
      "epoch": 0.21879115128722665,
      "grad_norm": 0.16070467233657837,
      "learning_rate": 9.271015426135494e-05,
      "loss": 0.8693,
      "mean_token_accuracy": 0.7538273520767689,
      "num_tokens": 75902487.0,
      "step": 2378
    },
    {
      "entropy": 0.8837708123028278,
      "epoch": 0.21888315765866787,
      "grad_norm": 0.15836681425571442,
      "learning_rate": 9.27070874352133e-05,
      "loss": 0.8707,
      "mean_token_accuracy": 0.7574727237224579,
      "num_tokens": 75934107.0,
      "step": 2379
    },
    {
      "entropy": 0.8982877023518085,
      "epoch": 0.21897516403010908,
      "grad_norm": 0.1636568307876587,
      "learning_rate": 9.270402060907167e-05,
      "loss": 0.893,
      "mean_token_accuracy": 0.7458903677761555,
      "num_tokens": 75965502.0,
      "step": 2380
    },
    {
      "entropy": 0.7785019725561142,
      "epoch": 0.2190671704015503,
      "grad_norm": 0.15349778532981873,
      "learning_rate": 9.270095378293005e-05,
      "loss": 0.7784,
      "mean_token_accuracy": 0.7725294530391693,
      "num_tokens": 75996895.0,
      "step": 2381
    },
    {
      "entropy": 0.8959601409733295,
      "epoch": 0.21915917677299154,
      "grad_norm": 0.16255168616771698,
      "learning_rate": 9.269788695678842e-05,
      "loss": 0.8959,
      "mean_token_accuracy": 0.7454547770321369,
      "num_tokens": 76028648.0,
      "step": 2382
    },
    {
      "entropy": 0.8093692101538181,
      "epoch": 0.21925118314443276,
      "grad_norm": 0.15968404710292816,
      "learning_rate": 9.26948201306468e-05,
      "loss": 0.8335,
      "mean_token_accuracy": 0.7609451897442341,
      "num_tokens": 76061162.0,
      "step": 2383
    },
    {
      "entropy": 0.7668165154755116,
      "epoch": 0.21934318951587398,
      "grad_norm": 0.15506219863891602,
      "learning_rate": 9.269175330450517e-05,
      "loss": 0.7775,
      "mean_token_accuracy": 0.7728948183357716,
      "num_tokens": 76092886.0,
      "step": 2384
    },
    {
      "entropy": 0.7196921799331903,
      "epoch": 0.2194351958873152,
      "grad_norm": 0.15224286913871765,
      "learning_rate": 9.268868647836355e-05,
      "loss": 0.7127,
      "mean_token_accuracy": 0.7899365983903408,
      "num_tokens": 76124763.0,
      "step": 2385
    },
    {
      "entropy": 0.9621920473873615,
      "epoch": 0.2195272022587564,
      "grad_norm": 0.16576825082302094,
      "learning_rate": 9.268561965222192e-05,
      "loss": 0.9783,
      "mean_token_accuracy": 0.7252373807132244,
      "num_tokens": 76156770.0,
      "step": 2386
    },
    {
      "entropy": 0.9017544910311699,
      "epoch": 0.21961920863019765,
      "grad_norm": 0.15667423605918884,
      "learning_rate": 9.268255282608029e-05,
      "loss": 0.897,
      "mean_token_accuracy": 0.743223499506712,
      "num_tokens": 76188949.0,
      "step": 2387
    },
    {
      "entropy": 0.8645553756505251,
      "epoch": 0.21971121500163887,
      "grad_norm": 0.1545802503824234,
      "learning_rate": 9.267948599993867e-05,
      "loss": 0.8434,
      "mean_token_accuracy": 0.7597783878445625,
      "num_tokens": 76221003.0,
      "step": 2388
    },
    {
      "entropy": 0.8304884228855371,
      "epoch": 0.2198032213730801,
      "grad_norm": 0.15363381803035736,
      "learning_rate": 9.267641917379705e-05,
      "loss": 0.8263,
      "mean_token_accuracy": 0.7591828480362892,
      "num_tokens": 76252490.0,
      "step": 2389
    },
    {
      "entropy": 0.8207105826586485,
      "epoch": 0.2198952277445213,
      "grad_norm": 0.16062292456626892,
      "learning_rate": 9.267335234765542e-05,
      "loss": 0.8325,
      "mean_token_accuracy": 0.7621244452893734,
      "num_tokens": 76284991.0,
      "step": 2390
    },
    {
      "entropy": 0.8254145383834839,
      "epoch": 0.21998723411596252,
      "grad_norm": 0.1548829972743988,
      "learning_rate": 9.267028552151379e-05,
      "loss": 0.8092,
      "mean_token_accuracy": 0.7646306343376637,
      "num_tokens": 76316853.0,
      "step": 2391
    },
    {
      "entropy": 0.8714759461581707,
      "epoch": 0.22007924048740377,
      "grad_norm": 0.158981055021286,
      "learning_rate": 9.266721869537215e-05,
      "loss": 0.8641,
      "mean_token_accuracy": 0.7506410293281078,
      "num_tokens": 76347992.0,
      "step": 2392
    },
    {
      "entropy": 0.9102405179291964,
      "epoch": 0.22017124685884498,
      "grad_norm": 0.16308613121509552,
      "learning_rate": 9.266415186923054e-05,
      "loss": 0.9005,
      "mean_token_accuracy": 0.7455661222338676,
      "num_tokens": 76379983.0,
      "step": 2393
    },
    {
      "entropy": 0.8415393512696028,
      "epoch": 0.2202632532302862,
      "grad_norm": 0.15297505259513855,
      "learning_rate": 9.266108504308892e-05,
      "loss": 0.8254,
      "mean_token_accuracy": 0.7634361609816551,
      "num_tokens": 76412733.0,
      "step": 2394
    },
    {
      "entropy": 0.8242152854800224,
      "epoch": 0.22035525960172742,
      "grad_norm": 0.16453465819358826,
      "learning_rate": 9.265801821694728e-05,
      "loss": 0.8515,
      "mean_token_accuracy": 0.7549100369215012,
      "num_tokens": 76445181.0,
      "step": 2395
    },
    {
      "entropy": 0.8393812011927366,
      "epoch": 0.22044726597316863,
      "grad_norm": 0.29978421330451965,
      "learning_rate": 9.265495139080565e-05,
      "loss": 0.8273,
      "mean_token_accuracy": 0.7587360702455044,
      "num_tokens": 76475415.0,
      "step": 2396
    },
    {
      "entropy": 0.7825194150209427,
      "epoch": 0.22053927234460988,
      "grad_norm": 0.1561039835214615,
      "learning_rate": 9.265188456466403e-05,
      "loss": 0.7642,
      "mean_token_accuracy": 0.7783934213221073,
      "num_tokens": 76506859.0,
      "step": 2397
    },
    {
      "entropy": 0.9893200322985649,
      "epoch": 0.2206312787160511,
      "grad_norm": 0.16954182088375092,
      "learning_rate": 9.264881773852242e-05,
      "loss": 0.989,
      "mean_token_accuracy": 0.7232375107705593,
      "num_tokens": 76538538.0,
      "step": 2398
    },
    {
      "entropy": 0.9432032592594624,
      "epoch": 0.2207232850874923,
      "grad_norm": 0.1635211557149887,
      "learning_rate": 9.264575091238078e-05,
      "loss": 0.9333,
      "mean_token_accuracy": 0.7327356748282909,
      "num_tokens": 76569810.0,
      "step": 2399
    },
    {
      "entropy": 0.8307926431298256,
      "epoch": 0.22081529145893353,
      "grad_norm": 0.16534583270549774,
      "learning_rate": 9.264268408623915e-05,
      "loss": 0.8211,
      "mean_token_accuracy": 0.763165008276701,
      "num_tokens": 76600067.0,
      "step": 2400
    },
    {
      "entropy": 0.7885468527674675,
      "epoch": 0.22090729783037474,
      "grad_norm": 0.16105757653713226,
      "learning_rate": 9.263961726009753e-05,
      "loss": 0.7618,
      "mean_token_accuracy": 0.7815434262156487,
      "num_tokens": 76632200.0,
      "step": 2401
    },
    {
      "entropy": 0.911312647163868,
      "epoch": 0.220999304201816,
      "grad_norm": 0.16344329714775085,
      "learning_rate": 9.26365504339559e-05,
      "loss": 0.9257,
      "mean_token_accuracy": 0.7434613108634949,
      "num_tokens": 76663902.0,
      "step": 2402
    },
    {
      "entropy": 0.911921963095665,
      "epoch": 0.2210913105732572,
      "grad_norm": 0.22251935303211212,
      "learning_rate": 9.263348360781428e-05,
      "loss": 0.9317,
      "mean_token_accuracy": 0.7408071756362915,
      "num_tokens": 76695984.0,
      "step": 2403
    },
    {
      "entropy": 0.9112119600176811,
      "epoch": 0.22118331694469842,
      "grad_norm": 0.15777891874313354,
      "learning_rate": 9.263041678167265e-05,
      "loss": 0.9001,
      "mean_token_accuracy": 0.7424022667109966,
      "num_tokens": 76728060.0,
      "step": 2404
    },
    {
      "entropy": 0.7776431851089001,
      "epoch": 0.22127532331613964,
      "grad_norm": 0.15547235310077667,
      "learning_rate": 9.262734995553103e-05,
      "loss": 0.802,
      "mean_token_accuracy": 0.7701295167207718,
      "num_tokens": 76759860.0,
      "step": 2405
    },
    {
      "entropy": 0.7890602611005306,
      "epoch": 0.22136732968758085,
      "grad_norm": 0.15356017649173737,
      "learning_rate": 9.26242831293894e-05,
      "loss": 0.8132,
      "mean_token_accuracy": 0.7622497864067554,
      "num_tokens": 76791792.0,
      "step": 2406
    },
    {
      "entropy": 0.9493694566190243,
      "epoch": 0.2214593360590221,
      "grad_norm": 0.1628301739692688,
      "learning_rate": 9.262121630324777e-05,
      "loss": 0.9592,
      "mean_token_accuracy": 0.7309982180595398,
      "num_tokens": 76824351.0,
      "step": 2407
    },
    {
      "entropy": 0.7940101716667414,
      "epoch": 0.22155134243046332,
      "grad_norm": 0.1532294899225235,
      "learning_rate": 9.261814947710615e-05,
      "loss": 0.7924,
      "mean_token_accuracy": 0.7743231505155563,
      "num_tokens": 76855740.0,
      "step": 2408
    },
    {
      "entropy": 0.8962213099002838,
      "epoch": 0.22164334880190453,
      "grad_norm": 0.1641906201839447,
      "learning_rate": 9.261508265096453e-05,
      "loss": 0.9058,
      "mean_token_accuracy": 0.7448272183537483,
      "num_tokens": 76887358.0,
      "step": 2409
    },
    {
      "entropy": 0.8290924802422523,
      "epoch": 0.22173535517334575,
      "grad_norm": 0.15507254004478455,
      "learning_rate": 9.26120158248229e-05,
      "loss": 0.8163,
      "mean_token_accuracy": 0.7676545009016991,
      "num_tokens": 76919109.0,
      "step": 2410
    },
    {
      "entropy": 0.942939281463623,
      "epoch": 0.22182736154478697,
      "grad_norm": 0.1585475653409958,
      "learning_rate": 9.260894899868127e-05,
      "loss": 0.9197,
      "mean_token_accuracy": 0.735062912106514,
      "num_tokens": 76950923.0,
      "step": 2411
    },
    {
      "entropy": 0.9165598545223475,
      "epoch": 0.2219193679162282,
      "grad_norm": 0.16037026047706604,
      "learning_rate": 9.260588217253965e-05,
      "loss": 0.9187,
      "mean_token_accuracy": 0.7410679310560226,
      "num_tokens": 76983046.0,
      "step": 2412
    },
    {
      "entropy": 0.9298561438918114,
      "epoch": 0.22201137428766943,
      "grad_norm": 0.16303080320358276,
      "learning_rate": 9.260281534639801e-05,
      "loss": 0.8993,
      "mean_token_accuracy": 0.744863398373127,
      "num_tokens": 77015243.0,
      "step": 2413
    },
    {
      "entropy": 0.9026986621320248,
      "epoch": 0.22210338065911064,
      "grad_norm": 0.16461068391799927,
      "learning_rate": 9.25997485202564e-05,
      "loss": 0.8901,
      "mean_token_accuracy": 0.7470667101442814,
      "num_tokens": 77047413.0,
      "step": 2414
    },
    {
      "entropy": 0.8482456654310226,
      "epoch": 0.22219538703055186,
      "grad_norm": 0.15539486706256866,
      "learning_rate": 9.259668169411476e-05,
      "loss": 0.8226,
      "mean_token_accuracy": 0.7664720341563225,
      "num_tokens": 77079735.0,
      "step": 2415
    },
    {
      "entropy": 0.8858373872935772,
      "epoch": 0.22228739340199308,
      "grad_norm": 0.164635568857193,
      "learning_rate": 9.259361486797315e-05,
      "loss": 0.8755,
      "mean_token_accuracy": 0.7483071498572826,
      "num_tokens": 77110753.0,
      "step": 2416
    },
    {
      "entropy": 0.8632417060434818,
      "epoch": 0.22237939977343432,
      "grad_norm": 0.16256655752658844,
      "learning_rate": 9.259054804183151e-05,
      "loss": 0.8537,
      "mean_token_accuracy": 0.7513967789709568,
      "num_tokens": 77142431.0,
      "step": 2417
    },
    {
      "entropy": 0.843621127307415,
      "epoch": 0.22247140614487554,
      "grad_norm": 0.1585281640291214,
      "learning_rate": 9.258748121568988e-05,
      "loss": 0.8364,
      "mean_token_accuracy": 0.7543556876480579,
      "num_tokens": 77174161.0,
      "step": 2418
    },
    {
      "entropy": 0.9568715989589691,
      "epoch": 0.22256341251631676,
      "grad_norm": 0.1773958057165146,
      "learning_rate": 9.258441438954826e-05,
      "loss": 0.9605,
      "mean_token_accuracy": 0.7264894172549248,
      "num_tokens": 77204426.0,
      "step": 2419
    },
    {
      "entropy": 0.781513961032033,
      "epoch": 0.22265541888775797,
      "grad_norm": 0.15730610489845276,
      "learning_rate": 9.258134756340664e-05,
      "loss": 0.7691,
      "mean_token_accuracy": 0.7756846249103546,
      "num_tokens": 77236136.0,
      "step": 2420
    },
    {
      "entropy": 0.8705812357366085,
      "epoch": 0.2227474252591992,
      "grad_norm": 0.16577307879924774,
      "learning_rate": 9.257828073726501e-05,
      "loss": 0.8677,
      "mean_token_accuracy": 0.7493188418447971,
      "num_tokens": 77268415.0,
      "step": 2421
    },
    {
      "entropy": 0.8054774161428213,
      "epoch": 0.22283943163064043,
      "grad_norm": 0.1753743439912796,
      "learning_rate": 9.257521391112338e-05,
      "loss": 0.8094,
      "mean_token_accuracy": 0.7669086866080761,
      "num_tokens": 77300069.0,
      "step": 2422
    },
    {
      "entropy": 0.7800725363194942,
      "epoch": 0.22293143800208165,
      "grad_norm": 0.16658997535705566,
      "learning_rate": 9.257214708498175e-05,
      "loss": 0.8066,
      "mean_token_accuracy": 0.7699519209563732,
      "num_tokens": 77332670.0,
      "step": 2423
    },
    {
      "entropy": 0.8597707152366638,
      "epoch": 0.22302344437352287,
      "grad_norm": 0.1628721058368683,
      "learning_rate": 9.256908025884013e-05,
      "loss": 0.8745,
      "mean_token_accuracy": 0.7497112825512886,
      "num_tokens": 77364845.0,
      "step": 2424
    },
    {
      "entropy": 0.8321843110024929,
      "epoch": 0.22311545074496408,
      "grad_norm": 0.16255098581314087,
      "learning_rate": 9.256601343269851e-05,
      "loss": 0.8526,
      "mean_token_accuracy": 0.7545066028833389,
      "num_tokens": 77396415.0,
      "step": 2425
    },
    {
      "entropy": 0.8522726856172085,
      "epoch": 0.2232074571164053,
      "grad_norm": 0.15373045206069946,
      "learning_rate": 9.256294660655688e-05,
      "loss": 0.8265,
      "mean_token_accuracy": 0.7595599293708801,
      "num_tokens": 77428920.0,
      "step": 2426
    },
    {
      "entropy": 0.8900353945791721,
      "epoch": 0.22329946348784654,
      "grad_norm": 0.15893279016017914,
      "learning_rate": 9.255987978041525e-05,
      "loss": 0.893,
      "mean_token_accuracy": 0.7455858252942562,
      "num_tokens": 77460816.0,
      "step": 2427
    },
    {
      "entropy": 0.9393269494175911,
      "epoch": 0.22339146985928776,
      "grad_norm": 0.1608554571866989,
      "learning_rate": 9.255681295427363e-05,
      "loss": 0.9272,
      "mean_token_accuracy": 0.738957304507494,
      "num_tokens": 77492751.0,
      "step": 2428
    },
    {
      "entropy": 0.8629439491778612,
      "epoch": 0.22348347623072898,
      "grad_norm": 0.15529221296310425,
      "learning_rate": 9.2553746128132e-05,
      "loss": 0.8564,
      "mean_token_accuracy": 0.7540825977921486,
      "num_tokens": 77524799.0,
      "step": 2429
    },
    {
      "entropy": 0.8922556843608618,
      "epoch": 0.2235754826021702,
      "grad_norm": 0.16392524540424347,
      "learning_rate": 9.255067930199038e-05,
      "loss": 0.8724,
      "mean_token_accuracy": 0.7535968795418739,
      "num_tokens": 77555791.0,
      "step": 2430
    },
    {
      "entropy": 0.8145883567631245,
      "epoch": 0.2236674889736114,
      "grad_norm": 0.16306248307228088,
      "learning_rate": 9.254761247584874e-05,
      "loss": 0.7997,
      "mean_token_accuracy": 0.7688535898923874,
      "num_tokens": 77587263.0,
      "step": 2431
    },
    {
      "entropy": 0.8463025651872158,
      "epoch": 0.22375949534505266,
      "grad_norm": 0.1577865034341812,
      "learning_rate": 9.254454564970713e-05,
      "loss": 0.8406,
      "mean_token_accuracy": 0.7589344307780266,
      "num_tokens": 77618796.0,
      "step": 2432
    },
    {
      "entropy": 0.7921186871826649,
      "epoch": 0.22385150171649387,
      "grad_norm": 0.1579602062702179,
      "learning_rate": 9.25414788235655e-05,
      "loss": 0.811,
      "mean_token_accuracy": 0.7654600888490677,
      "num_tokens": 77650371.0,
      "step": 2433
    },
    {
      "entropy": 0.9051210563629866,
      "epoch": 0.2239435080879351,
      "grad_norm": 0.1605272740125656,
      "learning_rate": 9.253841199742386e-05,
      "loss": 0.9165,
      "mean_token_accuracy": 0.7402685470879078,
      "num_tokens": 77682700.0,
      "step": 2434
    },
    {
      "entropy": 0.7928264662623405,
      "epoch": 0.2240355144593763,
      "grad_norm": 0.15692757070064545,
      "learning_rate": 9.253534517128224e-05,
      "loss": 0.7775,
      "mean_token_accuracy": 0.7737283334136009,
      "num_tokens": 77714938.0,
      "step": 2435
    },
    {
      "entropy": 0.8577410690486431,
      "epoch": 0.22412752083081752,
      "grad_norm": 0.16076110303401947,
      "learning_rate": 9.253227834514062e-05,
      "loss": 0.8516,
      "mean_token_accuracy": 0.7559146955609322,
      "num_tokens": 77746761.0,
      "step": 2436
    },
    {
      "entropy": 0.909280925989151,
      "epoch": 0.22421952720225877,
      "grad_norm": 0.16244357824325562,
      "learning_rate": 9.252921151899899e-05,
      "loss": 0.9028,
      "mean_token_accuracy": 0.7423637919127941,
      "num_tokens": 77778751.0,
      "step": 2437
    },
    {
      "entropy": 0.8586148098111153,
      "epoch": 0.22431153357369998,
      "grad_norm": 0.1640169769525528,
      "learning_rate": 9.252614469285736e-05,
      "loss": 0.8577,
      "mean_token_accuracy": 0.7535280324518681,
      "num_tokens": 77810125.0,
      "step": 2438
    },
    {
      "entropy": 0.8394592553377151,
      "epoch": 0.2244035399451412,
      "grad_norm": 0.1634720265865326,
      "learning_rate": 9.252307786671573e-05,
      "loss": 0.8555,
      "mean_token_accuracy": 0.7559535950422287,
      "num_tokens": 77841720.0,
      "step": 2439
    },
    {
      "entropy": 0.832049660384655,
      "epoch": 0.22449554631658242,
      "grad_norm": 0.15731443464756012,
      "learning_rate": 9.252001104057412e-05,
      "loss": 0.8244,
      "mean_token_accuracy": 0.7612277865409851,
      "num_tokens": 77872917.0,
      "step": 2440
    },
    {
      "entropy": 0.8788662552833557,
      "epoch": 0.22458755268802363,
      "grad_norm": 0.1670287549495697,
      "learning_rate": 9.251694421443249e-05,
      "loss": 0.8933,
      "mean_token_accuracy": 0.7500119619071484,
      "num_tokens": 77904058.0,
      "step": 2441
    },
    {
      "entropy": 0.8365580625832081,
      "epoch": 0.22467955905946488,
      "grad_norm": 0.16152448952198029,
      "learning_rate": 9.251387738829086e-05,
      "loss": 0.8353,
      "mean_token_accuracy": 0.75727778673172,
      "num_tokens": 77935611.0,
      "step": 2442
    },
    {
      "entropy": 0.8847270049154758,
      "epoch": 0.2247715654309061,
      "grad_norm": 0.15259428322315216,
      "learning_rate": 9.251081056214924e-05,
      "loss": 0.8595,
      "mean_token_accuracy": 0.7464080639183521,
      "num_tokens": 77967550.0,
      "step": 2443
    },
    {
      "entropy": 0.9056771844625473,
      "epoch": 0.2248635718023473,
      "grad_norm": 0.16158853471279144,
      "learning_rate": 9.250774373600761e-05,
      "loss": 0.902,
      "mean_token_accuracy": 0.741930466145277,
      "num_tokens": 77998382.0,
      "step": 2444
    },
    {
      "entropy": 0.9118116088211536,
      "epoch": 0.22495557817378853,
      "grad_norm": 0.16490763425827026,
      "learning_rate": 9.250467690986599e-05,
      "loss": 0.9093,
      "mean_token_accuracy": 0.7444416992366314,
      "num_tokens": 78030788.0,
      "step": 2445
    },
    {
      "entropy": 0.8060839623212814,
      "epoch": 0.22504758454522975,
      "grad_norm": 0.1540549099445343,
      "learning_rate": 9.250161008372436e-05,
      "loss": 0.7995,
      "mean_token_accuracy": 0.7692074738442898,
      "num_tokens": 78062247.0,
      "step": 2446
    },
    {
      "entropy": 0.9242357388138771,
      "epoch": 0.225139590916671,
      "grad_norm": 0.16591130197048187,
      "learning_rate": 9.249854325758274e-05,
      "loss": 0.9194,
      "mean_token_accuracy": 0.7422970943152905,
      "num_tokens": 78094123.0,
      "step": 2447
    },
    {
      "entropy": 0.8401551768183708,
      "epoch": 0.2252315972881122,
      "grad_norm": 0.15783396363258362,
      "learning_rate": 9.249547643144111e-05,
      "loss": 0.8284,
      "mean_token_accuracy": 0.7632195167243481,
      "num_tokens": 78125511.0,
      "step": 2448
    },
    {
      "entropy": 0.8382232375442982,
      "epoch": 0.22532360365955342,
      "grad_norm": 0.1595248132944107,
      "learning_rate": 9.249240960529948e-05,
      "loss": 0.8209,
      "mean_token_accuracy": 0.75899513438344,
      "num_tokens": 78156576.0,
      "step": 2449
    },
    {
      "entropy": 0.866189980879426,
      "epoch": 0.22541561003099464,
      "grad_norm": 0.15594273805618286,
      "learning_rate": 9.248934277915786e-05,
      "loss": 0.8564,
      "mean_token_accuracy": 0.7557939291000366,
      "num_tokens": 78188655.0,
      "step": 2450
    },
    {
      "entropy": 0.8278778847306967,
      "epoch": 0.22550761640243586,
      "grad_norm": 0.15583886206150055,
      "learning_rate": 9.248627595301624e-05,
      "loss": 0.8286,
      "mean_token_accuracy": 0.7639510966837406,
      "num_tokens": 78221139.0,
      "step": 2451
    },
    {
      "entropy": 0.7921338379383087,
      "epoch": 0.2255996227738771,
      "grad_norm": 0.17282500863075256,
      "learning_rate": 9.24832091268746e-05,
      "loss": 0.7886,
      "mean_token_accuracy": 0.7755043432116508,
      "num_tokens": 78253058.0,
      "step": 2452
    },
    {
      "entropy": 0.8750727027654648,
      "epoch": 0.22569162914531832,
      "grad_norm": 0.16083507239818573,
      "learning_rate": 9.248014230073297e-05,
      "loss": 0.8622,
      "mean_token_accuracy": 0.7477783150970936,
      "num_tokens": 78284959.0,
      "step": 2453
    },
    {
      "entropy": 0.8606866784393787,
      "epoch": 0.22578363551675953,
      "grad_norm": 0.1617298424243927,
      "learning_rate": 9.247707547459134e-05,
      "loss": 0.8514,
      "mean_token_accuracy": 0.755724623799324,
      "num_tokens": 78316204.0,
      "step": 2454
    },
    {
      "entropy": 0.8241336867213249,
      "epoch": 0.22587564188820075,
      "grad_norm": 0.1632377803325653,
      "learning_rate": 9.247400864844972e-05,
      "loss": 0.8382,
      "mean_token_accuracy": 0.7600356824696064,
      "num_tokens": 78348828.0,
      "step": 2455
    },
    {
      "entropy": 0.8431034311652184,
      "epoch": 0.22596764825964197,
      "grad_norm": 0.16564667224884033,
      "learning_rate": 9.24709418223081e-05,
      "loss": 0.8436,
      "mean_token_accuracy": 0.7573892809450626,
      "num_tokens": 78380777.0,
      "step": 2456
    },
    {
      "entropy": 0.7939852811396122,
      "epoch": 0.2260596546310832,
      "grad_norm": 0.17228050529956818,
      "learning_rate": 9.246787499616647e-05,
      "loss": 0.8182,
      "mean_token_accuracy": 0.7668692469596863,
      "num_tokens": 78412211.0,
      "step": 2457
    },
    {
      "entropy": 0.810934241861105,
      "epoch": 0.22615166100252443,
      "grad_norm": 0.1561870276927948,
      "learning_rate": 9.246480817002484e-05,
      "loss": 0.8041,
      "mean_token_accuracy": 0.7616858631372452,
      "num_tokens": 78443779.0,
      "step": 2458
    },
    {
      "entropy": 0.9104740023612976,
      "epoch": 0.22624366737396565,
      "grad_norm": 0.15638893842697144,
      "learning_rate": 9.246174134388322e-05,
      "loss": 0.8923,
      "mean_token_accuracy": 0.7439194172620773,
      "num_tokens": 78475720.0,
      "step": 2459
    },
    {
      "entropy": 0.9017694909125566,
      "epoch": 0.22633567374540686,
      "grad_norm": 0.1610424667596817,
      "learning_rate": 9.245867451774159e-05,
      "loss": 0.8927,
      "mean_token_accuracy": 0.7458363771438599,
      "num_tokens": 78507533.0,
      "step": 2460
    },
    {
      "entropy": 0.8449519742280245,
      "epoch": 0.22642768011684808,
      "grad_norm": 0.15164564549922943,
      "learning_rate": 9.245560769159997e-05,
      "loss": 0.8023,
      "mean_token_accuracy": 0.764660470187664,
      "num_tokens": 78538704.0,
      "step": 2461
    },
    {
      "entropy": 0.894967719912529,
      "epoch": 0.22651968648828932,
      "grad_norm": 0.16110679507255554,
      "learning_rate": 9.245254086545834e-05,
      "loss": 0.8899,
      "mean_token_accuracy": 0.7483405396342278,
      "num_tokens": 78570254.0,
      "step": 2462
    },
    {
      "entropy": 0.9032144881784916,
      "epoch": 0.22661169285973054,
      "grad_norm": 0.1580689251422882,
      "learning_rate": 9.244947403931672e-05,
      "loss": 0.9123,
      "mean_token_accuracy": 0.7412020415067673,
      "num_tokens": 78602565.0,
      "step": 2463
    },
    {
      "entropy": 0.8070832937955856,
      "epoch": 0.22670369923117176,
      "grad_norm": 0.1578044593334198,
      "learning_rate": 9.244640721317509e-05,
      "loss": 0.8092,
      "mean_token_accuracy": 0.7657571919262409,
      "num_tokens": 78634520.0,
      "step": 2464
    },
    {
      "entropy": 0.8746586441993713,
      "epoch": 0.22679570560261297,
      "grad_norm": 0.15881843864917755,
      "learning_rate": 9.244334038703346e-05,
      "loss": 0.8868,
      "mean_token_accuracy": 0.7502185218036175,
      "num_tokens": 78667158.0,
      "step": 2465
    },
    {
      "entropy": 0.9664230234920979,
      "epoch": 0.2268877119740542,
      "grad_norm": 0.16260910034179688,
      "learning_rate": 9.244027356089184e-05,
      "loss": 0.9659,
      "mean_token_accuracy": 0.729084737598896,
      "num_tokens": 78698459.0,
      "step": 2466
    },
    {
      "entropy": 0.905882254242897,
      "epoch": 0.22697971834549543,
      "grad_norm": 0.15757645666599274,
      "learning_rate": 9.243720673475022e-05,
      "loss": 0.8916,
      "mean_token_accuracy": 0.7458305507898331,
      "num_tokens": 78729492.0,
      "step": 2467
    },
    {
      "entropy": 0.9068819880485535,
      "epoch": 0.22707172471693665,
      "grad_norm": 0.15952467918395996,
      "learning_rate": 9.243413990860859e-05,
      "loss": 0.9058,
      "mean_token_accuracy": 0.7448797337710857,
      "num_tokens": 78761035.0,
      "step": 2468
    },
    {
      "entropy": 0.8652681820094585,
      "epoch": 0.22716373108837787,
      "grad_norm": 0.1550253927707672,
      "learning_rate": 9.243107308246695e-05,
      "loss": 0.8412,
      "mean_token_accuracy": 0.7564604356884956,
      "num_tokens": 78792274.0,
      "step": 2469
    },
    {
      "entropy": 0.7865561619400978,
      "epoch": 0.22725573745981908,
      "grad_norm": 0.1493394523859024,
      "learning_rate": 9.242800625632532e-05,
      "loss": 0.7794,
      "mean_token_accuracy": 0.7772038020193577,
      "num_tokens": 78824661.0,
      "step": 2470
    },
    {
      "entropy": 0.8972028810530901,
      "epoch": 0.2273477438312603,
      "grad_norm": 0.16011594235897064,
      "learning_rate": 9.24249394301837e-05,
      "loss": 0.8896,
      "mean_token_accuracy": 0.7418696284294128,
      "num_tokens": 78856461.0,
      "step": 2471
    },
    {
      "entropy": 0.9128634110093117,
      "epoch": 0.22743975020270155,
      "grad_norm": 0.16525974869728088,
      "learning_rate": 9.242187260404209e-05,
      "loss": 0.9069,
      "mean_token_accuracy": 0.7362042888998985,
      "num_tokens": 78887496.0,
      "step": 2472
    },
    {
      "entropy": 0.8340515121817589,
      "epoch": 0.22753175657414276,
      "grad_norm": 0.15634284913539886,
      "learning_rate": 9.241880577790045e-05,
      "loss": 0.8536,
      "mean_token_accuracy": 0.7561912760138512,
      "num_tokens": 78920017.0,
      "step": 2473
    },
    {
      "entropy": 0.8489092029631138,
      "epoch": 0.22762376294558398,
      "grad_norm": 0.16486378014087677,
      "learning_rate": 9.241573895175883e-05,
      "loss": 0.864,
      "mean_token_accuracy": 0.756095115095377,
      "num_tokens": 78951794.0,
      "step": 2474
    },
    {
      "entropy": 0.8770532496273518,
      "epoch": 0.2277157693170252,
      "grad_norm": 0.1694047898054123,
      "learning_rate": 9.24126721256172e-05,
      "loss": 0.9075,
      "mean_token_accuracy": 0.744628768414259,
      "num_tokens": 78983740.0,
      "step": 2475
    },
    {
      "entropy": 0.803309004753828,
      "epoch": 0.2278077756884664,
      "grad_norm": 0.15158061683177948,
      "learning_rate": 9.240960529947557e-05,
      "loss": 0.7867,
      "mean_token_accuracy": 0.7692741863429546,
      "num_tokens": 79015576.0,
      "step": 2476
    },
    {
      "entropy": 0.90396523848176,
      "epoch": 0.22789978205990766,
      "grad_norm": 0.16083337366580963,
      "learning_rate": 9.240653847333395e-05,
      "loss": 0.9069,
      "mean_token_accuracy": 0.7445638254284859,
      "num_tokens": 79047425.0,
      "step": 2477
    },
    {
      "entropy": 0.8284633606672287,
      "epoch": 0.22799178843134887,
      "grad_norm": 0.15636740624904633,
      "learning_rate": 9.240347164719233e-05,
      "loss": 0.8248,
      "mean_token_accuracy": 0.7598037160933018,
      "num_tokens": 79079105.0,
      "step": 2478
    },
    {
      "entropy": 0.9754172451794147,
      "epoch": 0.2280837948027901,
      "grad_norm": 0.162236750125885,
      "learning_rate": 9.24004048210507e-05,
      "loss": 0.9703,
      "mean_token_accuracy": 0.7263271249830723,
      "num_tokens": 79110929.0,
      "step": 2479
    },
    {
      "entropy": 0.8830737844109535,
      "epoch": 0.2281758011742313,
      "grad_norm": 0.15795715153217316,
      "learning_rate": 9.239733799490907e-05,
      "loss": 0.8847,
      "mean_token_accuracy": 0.7462239935994148,
      "num_tokens": 79142481.0,
      "step": 2480
    },
    {
      "entropy": 0.7978446967899799,
      "epoch": 0.22826780754567252,
      "grad_norm": 0.1544060856103897,
      "learning_rate": 9.239427116876744e-05,
      "loss": 0.7808,
      "mean_token_accuracy": 0.7706020884215832,
      "num_tokens": 79174093.0,
      "step": 2481
    },
    {
      "entropy": 0.8742479588836432,
      "epoch": 0.22835981391711377,
      "grad_norm": 0.1600320041179657,
      "learning_rate": 9.239120434262583e-05,
      "loss": 0.8816,
      "mean_token_accuracy": 0.7478441186249256,
      "num_tokens": 79206326.0,
      "step": 2482
    },
    {
      "entropy": 0.8318744078278542,
      "epoch": 0.22845182028855499,
      "grad_norm": 0.15514682233333588,
      "learning_rate": 9.23881375164842e-05,
      "loss": 0.8251,
      "mean_token_accuracy": 0.7598628960549831,
      "num_tokens": 79238651.0,
      "step": 2483
    },
    {
      "entropy": 0.8361840508878231,
      "epoch": 0.2285438266599962,
      "grad_norm": 0.1559528261423111,
      "learning_rate": 9.238507069034257e-05,
      "loss": 0.8412,
      "mean_token_accuracy": 0.7646263167262077,
      "num_tokens": 79270857.0,
      "step": 2484
    },
    {
      "entropy": 0.8149983789771795,
      "epoch": 0.22863583303143742,
      "grad_norm": 0.1584150344133377,
      "learning_rate": 9.238200386420094e-05,
      "loss": 0.8136,
      "mean_token_accuracy": 0.7631149366497993,
      "num_tokens": 79302999.0,
      "step": 2485
    },
    {
      "entropy": 0.7527251671999693,
      "epoch": 0.22872783940287866,
      "grad_norm": 0.15379489958286285,
      "learning_rate": 9.237893703805932e-05,
      "loss": 0.7367,
      "mean_token_accuracy": 0.7819862738251686,
      "num_tokens": 79335013.0,
      "step": 2486
    },
    {
      "entropy": 0.7787428051233292,
      "epoch": 0.22881984577431988,
      "grad_norm": 0.15137594938278198,
      "learning_rate": 9.23758702119177e-05,
      "loss": 0.7838,
      "mean_token_accuracy": 0.7732568792998791,
      "num_tokens": 79367451.0,
      "step": 2487
    },
    {
      "entropy": 0.7821936253458261,
      "epoch": 0.2289118521457611,
      "grad_norm": 0.1501394510269165,
      "learning_rate": 9.237280338577607e-05,
      "loss": 0.7804,
      "mean_token_accuracy": 0.776105273514986,
      "num_tokens": 79400219.0,
      "step": 2488
    },
    {
      "entropy": 0.8999012522399426,
      "epoch": 0.2290038585172023,
      "grad_norm": 0.15447062253952026,
      "learning_rate": 9.236973655963443e-05,
      "loss": 0.8979,
      "mean_token_accuracy": 0.7430298142135143,
      "num_tokens": 79431959.0,
      "step": 2489
    },
    {
      "entropy": 0.6989663504064083,
      "epoch": 0.22909586488864353,
      "grad_norm": 0.15910224616527557,
      "learning_rate": 9.236666973349282e-05,
      "loss": 0.7178,
      "mean_token_accuracy": 0.7912697903811932,
      "num_tokens": 79463770.0,
      "step": 2490
    },
    {
      "entropy": 0.8400491774082184,
      "epoch": 0.22918787126008477,
      "grad_norm": 0.1551157534122467,
      "learning_rate": 9.236360290735118e-05,
      "loss": 0.8357,
      "mean_token_accuracy": 0.7566493302583694,
      "num_tokens": 79495283.0,
      "step": 2491
    },
    {
      "entropy": 0.7601612824946642,
      "epoch": 0.229279877631526,
      "grad_norm": 0.16099302470684052,
      "learning_rate": 9.236053608120956e-05,
      "loss": 0.7667,
      "mean_token_accuracy": 0.7821105308830738,
      "num_tokens": 79526713.0,
      "step": 2492
    },
    {
      "entropy": 0.8660247027873993,
      "epoch": 0.2293718840029672,
      "grad_norm": 0.1597335934638977,
      "learning_rate": 9.235746925506793e-05,
      "loss": 0.8654,
      "mean_token_accuracy": 0.7495826557278633,
      "num_tokens": 79557610.0,
      "step": 2493
    },
    {
      "entropy": 0.7299485690891743,
      "epoch": 0.22946389037440842,
      "grad_norm": 0.14454710483551025,
      "learning_rate": 9.235440242892631e-05,
      "loss": 0.7172,
      "mean_token_accuracy": 0.7882722690701485,
      "num_tokens": 79590237.0,
      "step": 2494
    },
    {
      "entropy": 0.8973509669303894,
      "epoch": 0.22955589674584964,
      "grad_norm": 0.15999411046504974,
      "learning_rate": 9.235133560278468e-05,
      "loss": 0.8687,
      "mean_token_accuracy": 0.7505846209824085,
      "num_tokens": 79621797.0,
      "step": 2495
    },
    {
      "entropy": 0.8143093269318342,
      "epoch": 0.22964790311729089,
      "grad_norm": 0.15850117802619934,
      "learning_rate": 9.234826877664305e-05,
      "loss": 0.8136,
      "mean_token_accuracy": 0.7664641477167606,
      "num_tokens": 79653187.0,
      "step": 2496
    },
    {
      "entropy": 0.8050228506326675,
      "epoch": 0.2297399094887321,
      "grad_norm": 0.15886326134204865,
      "learning_rate": 9.234520195050143e-05,
      "loss": 0.804,
      "mean_token_accuracy": 0.7693191803991795,
      "num_tokens": 79684833.0,
      "step": 2497
    },
    {
      "entropy": 0.80837925337255,
      "epoch": 0.22983191586017332,
      "grad_norm": 0.1535378098487854,
      "learning_rate": 9.234213512435981e-05,
      "loss": 0.7974,
      "mean_token_accuracy": 0.7678496167063713,
      "num_tokens": 79716080.0,
      "step": 2498
    },
    {
      "entropy": 0.9095266684889793,
      "epoch": 0.22992392223161454,
      "grad_norm": 0.15854083001613617,
      "learning_rate": 9.233906829821818e-05,
      "loss": 0.9052,
      "mean_token_accuracy": 0.7428521439433098,
      "num_tokens": 79748342.0,
      "step": 2499
    },
    {
      "entropy": 0.9581333994865417,
      "epoch": 0.23001592860305575,
      "grad_norm": 0.16949696838855743,
      "learning_rate": 9.233600147207655e-05,
      "loss": 0.9494,
      "mean_token_accuracy": 0.7323674149811268,
      "num_tokens": 79780019.0,
      "step": 2500
    },
    {
      "entropy": 0.8045859560370445,
      "epoch": 0.230107934974497,
      "grad_norm": 0.15763889253139496,
      "learning_rate": 9.233293464593492e-05,
      "loss": 0.8236,
      "mean_token_accuracy": 0.760025542229414,
      "num_tokens": 79812211.0,
      "step": 2501
    },
    {
      "entropy": 0.8573220185935497,
      "epoch": 0.2301999413459382,
      "grad_norm": 0.16080912947654724,
      "learning_rate": 9.23298678197933e-05,
      "loss": 0.8673,
      "mean_token_accuracy": 0.7516346983611584,
      "num_tokens": 79844339.0,
      "step": 2502
    },
    {
      "entropy": 0.9637106880545616,
      "epoch": 0.23029194771737943,
      "grad_norm": 0.16696980595588684,
      "learning_rate": 9.232680099365168e-05,
      "loss": 0.9785,
      "mean_token_accuracy": 0.7224063202738762,
      "num_tokens": 79875547.0,
      "step": 2503
    },
    {
      "entropy": 0.811622865498066,
      "epoch": 0.23038395408882065,
      "grad_norm": 0.16427794098854065,
      "learning_rate": 9.232373416751005e-05,
      "loss": 0.8138,
      "mean_token_accuracy": 0.7666120603680611,
      "num_tokens": 79906147.0,
      "step": 2504
    },
    {
      "entropy": 0.9101489596068859,
      "epoch": 0.23047596046026186,
      "grad_norm": 0.1709432154893875,
      "learning_rate": 9.232066734136843e-05,
      "loss": 0.9466,
      "mean_token_accuracy": 0.7294044233858585,
      "num_tokens": 79937569.0,
      "step": 2505
    },
    {
      "entropy": 0.7943923510611057,
      "epoch": 0.2305679668317031,
      "grad_norm": 0.14858651161193848,
      "learning_rate": 9.23176005152268e-05,
      "loss": 0.7764,
      "mean_token_accuracy": 0.7753700651228428,
      "num_tokens": 79969454.0,
      "step": 2506
    },
    {
      "entropy": 0.7729433737695217,
      "epoch": 0.23065997320314433,
      "grad_norm": 0.1531657576560974,
      "learning_rate": 9.231453368908516e-05,
      "loss": 0.7757,
      "mean_token_accuracy": 0.7733856774866581,
      "num_tokens": 80001640.0,
      "step": 2507
    },
    {
      "entropy": 0.7956862337887287,
      "epoch": 0.23075197957458554,
      "grad_norm": 0.15449833869934082,
      "learning_rate": 9.231146686294355e-05,
      "loss": 0.7943,
      "mean_token_accuracy": 0.7683672457933426,
      "num_tokens": 80033919.0,
      "step": 2508
    },
    {
      "entropy": 0.9615556672215462,
      "epoch": 0.23084398594602676,
      "grad_norm": 0.1633494645357132,
      "learning_rate": 9.230840003680193e-05,
      "loss": 0.9569,
      "mean_token_accuracy": 0.7262942269444466,
      "num_tokens": 80064906.0,
      "step": 2509
    },
    {
      "entropy": 0.9393433406949043,
      "epoch": 0.23093599231746798,
      "grad_norm": 0.16427870094776154,
      "learning_rate": 9.23053332106603e-05,
      "loss": 0.9211,
      "mean_token_accuracy": 0.738980770111084,
      "num_tokens": 80096292.0,
      "step": 2510
    },
    {
      "entropy": 0.8870411552488804,
      "epoch": 0.23102799868890922,
      "grad_norm": 0.1578657180070877,
      "learning_rate": 9.230226638451866e-05,
      "loss": 0.8542,
      "mean_token_accuracy": 0.7595647983253002,
      "num_tokens": 80128055.0,
      "step": 2511
    },
    {
      "entropy": 0.8638876006007195,
      "epoch": 0.23112000506035044,
      "grad_norm": 0.15666009485721588,
      "learning_rate": 9.229919955837703e-05,
      "loss": 0.8367,
      "mean_token_accuracy": 0.7591000124812126,
      "num_tokens": 80159705.0,
      "step": 2512
    },
    {
      "entropy": 0.8222678378224373,
      "epoch": 0.23121201143179165,
      "grad_norm": 0.15821228921413422,
      "learning_rate": 9.229613273223541e-05,
      "loss": 0.8111,
      "mean_token_accuracy": 0.7667393311858177,
      "num_tokens": 80192018.0,
      "step": 2513
    },
    {
      "entropy": 0.8936253190040588,
      "epoch": 0.23130401780323287,
      "grad_norm": 0.15716640651226044,
      "learning_rate": 9.22930659060938e-05,
      "loss": 0.8852,
      "mean_token_accuracy": 0.7489055246114731,
      "num_tokens": 80224677.0,
      "step": 2514
    },
    {
      "entropy": 0.7486544344574213,
      "epoch": 0.2313960241746741,
      "grad_norm": 0.1482313871383667,
      "learning_rate": 9.228999907995216e-05,
      "loss": 0.7439,
      "mean_token_accuracy": 0.7806867994368076,
      "num_tokens": 80257075.0,
      "step": 2515
    },
    {
      "entropy": 0.8682393841445446,
      "epoch": 0.23148803054611533,
      "grad_norm": 0.1623973250389099,
      "learning_rate": 9.228693225381053e-05,
      "loss": 0.8814,
      "mean_token_accuracy": 0.7479457445442677,
      "num_tokens": 80289653.0,
      "step": 2516
    },
    {
      "entropy": 0.8468379843980074,
      "epoch": 0.23158003691755655,
      "grad_norm": 0.1609487533569336,
      "learning_rate": 9.228386542766891e-05,
      "loss": 0.8595,
      "mean_token_accuracy": 0.7513856440782547,
      "num_tokens": 80321558.0,
      "step": 2517
    },
    {
      "entropy": 0.8939787559211254,
      "epoch": 0.23167204328899776,
      "grad_norm": 0.16471311450004578,
      "learning_rate": 9.228079860152729e-05,
      "loss": 0.9076,
      "mean_token_accuracy": 0.7450399287045002,
      "num_tokens": 80353986.0,
      "step": 2518
    },
    {
      "entropy": 0.8743475489318371,
      "epoch": 0.23176404966043898,
      "grad_norm": 0.196462020277977,
      "learning_rate": 9.227773177538566e-05,
      "loss": 0.8905,
      "mean_token_accuracy": 0.7494424320757389,
      "num_tokens": 80385871.0,
      "step": 2519
    },
    {
      "entropy": 0.9456807188689709,
      "epoch": 0.2318560560318802,
      "grad_norm": 0.16722087562084198,
      "learning_rate": 9.227466494924403e-05,
      "loss": 0.9627,
      "mean_token_accuracy": 0.7264143973588943,
      "num_tokens": 80417960.0,
      "step": 2520
    },
    {
      "entropy": 0.7453686948865652,
      "epoch": 0.23194806240332144,
      "grad_norm": 0.15357938408851624,
      "learning_rate": 9.227159812310241e-05,
      "loss": 0.7209,
      "mean_token_accuracy": 0.7960108853876591,
      "num_tokens": 80449853.0,
      "step": 2521
    },
    {
      "entropy": 0.8061575628817081,
      "epoch": 0.23204006877476266,
      "grad_norm": 0.1558542251586914,
      "learning_rate": 9.226853129696078e-05,
      "loss": 0.7994,
      "mean_token_accuracy": 0.7679300829768181,
      "num_tokens": 80481749.0,
      "step": 2522
    },
    {
      "entropy": 0.8440524339675903,
      "epoch": 0.23213207514620388,
      "grad_norm": 0.15763160586357117,
      "learning_rate": 9.226546447081916e-05,
      "loss": 0.8427,
      "mean_token_accuracy": 0.7584258764982224,
      "num_tokens": 80513433.0,
      "step": 2523
    },
    {
      "entropy": 0.7367973085492849,
      "epoch": 0.2322240815176451,
      "grad_norm": 0.16486576199531555,
      "learning_rate": 9.226239764467753e-05,
      "loss": 0.7298,
      "mean_token_accuracy": 0.7872848436236382,
      "num_tokens": 80545959.0,
      "step": 2524
    },
    {
      "entropy": 0.8292071986943483,
      "epoch": 0.2323160878890863,
      "grad_norm": 0.16341114044189453,
      "learning_rate": 9.225933081853591e-05,
      "loss": 0.8392,
      "mean_token_accuracy": 0.7631802335381508,
      "num_tokens": 80578350.0,
      "step": 2525
    },
    {
      "entropy": 0.8869740031659603,
      "epoch": 0.23240809426052755,
      "grad_norm": 0.16194956004619598,
      "learning_rate": 9.225626399239428e-05,
      "loss": 0.8752,
      "mean_token_accuracy": 0.7448191121220589,
      "num_tokens": 80609423.0,
      "step": 2526
    },
    {
      "entropy": 0.7774564307183027,
      "epoch": 0.23250010063196877,
      "grad_norm": 0.1541629582643509,
      "learning_rate": 9.225319716625264e-05,
      "loss": 0.7444,
      "mean_token_accuracy": 0.7826799228787422,
      "num_tokens": 80640458.0,
      "step": 2527
    },
    {
      "entropy": 0.8622760400176048,
      "epoch": 0.23259210700341,
      "grad_norm": 0.15932323038578033,
      "learning_rate": 9.225013034011102e-05,
      "loss": 0.8577,
      "mean_token_accuracy": 0.7569322325289249,
      "num_tokens": 80672704.0,
      "step": 2528
    },
    {
      "entropy": 0.7451208382844925,
      "epoch": 0.2326841133748512,
      "grad_norm": 0.15801464021205902,
      "learning_rate": 9.22470635139694e-05,
      "loss": 0.7431,
      "mean_token_accuracy": 0.7878801114857197,
      "num_tokens": 80704324.0,
      "step": 2529
    },
    {
      "entropy": 0.8890659436583519,
      "epoch": 0.23277611974629242,
      "grad_norm": 0.1617988646030426,
      "learning_rate": 9.224399668782777e-05,
      "loss": 0.8699,
      "mean_token_accuracy": 0.7458738535642624,
      "num_tokens": 80735628.0,
      "step": 2530
    },
    {
      "entropy": 0.8315621688961983,
      "epoch": 0.23286812611773366,
      "grad_norm": 0.15676869451999664,
      "learning_rate": 9.224092986168614e-05,
      "loss": 0.8286,
      "mean_token_accuracy": 0.7664830647408962,
      "num_tokens": 80767842.0,
      "step": 2531
    },
    {
      "entropy": 0.815943107008934,
      "epoch": 0.23296013248917488,
      "grad_norm": 0.15587301552295685,
      "learning_rate": 9.223786303554451e-05,
      "loss": 0.8397,
      "mean_token_accuracy": 0.7605793997645378,
      "num_tokens": 80800226.0,
      "step": 2532
    },
    {
      "entropy": 0.9080995880067348,
      "epoch": 0.2330521388606161,
      "grad_norm": 0.16237564384937286,
      "learning_rate": 9.223479620940289e-05,
      "loss": 0.905,
      "mean_token_accuracy": 0.7415450885891914,
      "num_tokens": 80831981.0,
      "step": 2533
    },
    {
      "entropy": 0.8357831444591284,
      "epoch": 0.23314414523205731,
      "grad_norm": 0.15790802240371704,
      "learning_rate": 9.223172938326127e-05,
      "loss": 0.8341,
      "mean_token_accuracy": 0.7609615251421928,
      "num_tokens": 80864258.0,
      "step": 2534
    },
    {
      "entropy": 0.8913528807461262,
      "epoch": 0.23323615160349853,
      "grad_norm": 0.16180554032325745,
      "learning_rate": 9.222866255711964e-05,
      "loss": 0.8751,
      "mean_token_accuracy": 0.7479887343943119,
      "num_tokens": 80895472.0,
      "step": 2535
    },
    {
      "entropy": 0.9673323556780815,
      "epoch": 0.23332815797493978,
      "grad_norm": 0.16437603533267975,
      "learning_rate": 9.222559573097802e-05,
      "loss": 0.9568,
      "mean_token_accuracy": 0.7330957390367985,
      "num_tokens": 80927081.0,
      "step": 2536
    },
    {
      "entropy": 0.8932473044842482,
      "epoch": 0.233420164346381,
      "grad_norm": 0.1582300364971161,
      "learning_rate": 9.222252890483639e-05,
      "loss": 0.8738,
      "mean_token_accuracy": 0.7499231658875942,
      "num_tokens": 80959254.0,
      "step": 2537
    },
    {
      "entropy": 0.7520085498690605,
      "epoch": 0.2335121707178222,
      "grad_norm": 0.15005959570407867,
      "learning_rate": 9.221946207869476e-05,
      "loss": 0.7511,
      "mean_token_accuracy": 0.7800732217729092,
      "num_tokens": 80991236.0,
      "step": 2538
    },
    {
      "entropy": 0.8357239179313183,
      "epoch": 0.23360417708926343,
      "grad_norm": 0.15609849989414215,
      "learning_rate": 9.221639525255314e-05,
      "loss": 0.8377,
      "mean_token_accuracy": 0.7572665959596634,
      "num_tokens": 81023621.0,
      "step": 2539
    },
    {
      "entropy": 0.8552778214216232,
      "epoch": 0.23369618346070464,
      "grad_norm": 0.1668003797531128,
      "learning_rate": 9.221332842641152e-05,
      "loss": 0.8706,
      "mean_token_accuracy": 0.7516941577196121,
      "num_tokens": 81054471.0,
      "step": 2540
    },
    {
      "entropy": 0.9160723946988583,
      "epoch": 0.2337881898321459,
      "grad_norm": 0.1661621630191803,
      "learning_rate": 9.221026160026989e-05,
      "loss": 0.907,
      "mean_token_accuracy": 0.740305420011282,
      "num_tokens": 81085585.0,
      "step": 2541
    },
    {
      "entropy": 0.8386390395462513,
      "epoch": 0.2338801962035871,
      "grad_norm": 0.16242480278015137,
      "learning_rate": 9.220719477412826e-05,
      "loss": 0.8089,
      "mean_token_accuracy": 0.7696345522999763,
      "num_tokens": 81116793.0,
      "step": 2542
    },
    {
      "entropy": 0.87186080776155,
      "epoch": 0.23397220257502832,
      "grad_norm": 0.16557569801807404,
      "learning_rate": 9.220412794798662e-05,
      "loss": 0.8617,
      "mean_token_accuracy": 0.7498132102191448,
      "num_tokens": 81148130.0,
      "step": 2543
    },
    {
      "entropy": 0.919548075646162,
      "epoch": 0.23406420894646954,
      "grad_norm": 0.16033175587654114,
      "learning_rate": 9.2201061121845e-05,
      "loss": 0.9288,
      "mean_token_accuracy": 0.7381046637892723,
      "num_tokens": 81180370.0,
      "step": 2544
    },
    {
      "entropy": 0.8280935920774937,
      "epoch": 0.23415621531791075,
      "grad_norm": 0.1578686535358429,
      "learning_rate": 9.219799429570339e-05,
      "loss": 0.8062,
      "mean_token_accuracy": 0.7656646408140659,
      "num_tokens": 81212160.0,
      "step": 2545
    },
    {
      "entropy": 0.7869646642357111,
      "epoch": 0.234248221689352,
      "grad_norm": 0.15341590344905853,
      "learning_rate": 9.219492746956175e-05,
      "loss": 0.7698,
      "mean_token_accuracy": 0.7751461490988731,
      "num_tokens": 81244552.0,
      "step": 2546
    },
    {
      "entropy": 0.8872454576194286,
      "epoch": 0.23434022806079322,
      "grad_norm": 0.15666957199573517,
      "learning_rate": 9.219186064342012e-05,
      "loss": 0.8933,
      "mean_token_accuracy": 0.7392182908952236,
      "num_tokens": 81276392.0,
      "step": 2547
    },
    {
      "entropy": 0.8826479315757751,
      "epoch": 0.23443223443223443,
      "grad_norm": 0.16041693091392517,
      "learning_rate": 9.21887938172785e-05,
      "loss": 0.8964,
      "mean_token_accuracy": 0.7451517283916473,
      "num_tokens": 81308627.0,
      "step": 2548
    },
    {
      "entropy": 0.8716130517423153,
      "epoch": 0.23452424080367565,
      "grad_norm": 0.16297611594200134,
      "learning_rate": 9.218572699113687e-05,
      "loss": 0.8824,
      "mean_token_accuracy": 0.7461328767240047,
      "num_tokens": 81340635.0,
      "step": 2549
    },
    {
      "entropy": 0.9350419417023659,
      "epoch": 0.23461624717511687,
      "grad_norm": 0.16306786239147186,
      "learning_rate": 9.218266016499525e-05,
      "loss": 0.9446,
      "mean_token_accuracy": 0.7289000004529953,
      "num_tokens": 81372692.0,
      "step": 2550
    },
    {
      "entropy": 0.7892696559429169,
      "epoch": 0.2347082535465581,
      "grad_norm": 0.15358473360538483,
      "learning_rate": 9.217959333885362e-05,
      "loss": 0.7878,
      "mean_token_accuracy": 0.7731362506747246,
      "num_tokens": 81404585.0,
      "step": 2551
    },
    {
      "entropy": 0.970940787345171,
      "epoch": 0.23480025991799933,
      "grad_norm": 0.15692253410816193,
      "learning_rate": 9.2176526512712e-05,
      "loss": 0.976,
      "mean_token_accuracy": 0.7214858792722225,
      "num_tokens": 81437136.0,
      "step": 2552
    },
    {
      "entropy": 0.9183036759495735,
      "epoch": 0.23489226628944054,
      "grad_norm": 0.15874366462230682,
      "learning_rate": 9.217345968657037e-05,
      "loss": 0.9011,
      "mean_token_accuracy": 0.7414589375257492,
      "num_tokens": 81469423.0,
      "step": 2553
    },
    {
      "entropy": 0.8160550743341446,
      "epoch": 0.23498427266088176,
      "grad_norm": 0.15266053378582,
      "learning_rate": 9.217039286042874e-05,
      "loss": 0.7689,
      "mean_token_accuracy": 0.7788615375757217,
      "num_tokens": 81501287.0,
      "step": 2554
    },
    {
      "entropy": 0.9356784634292126,
      "epoch": 0.23507627903232298,
      "grad_norm": 0.16570092737674713,
      "learning_rate": 9.216732603428712e-05,
      "loss": 0.9288,
      "mean_token_accuracy": 0.733412854373455,
      "num_tokens": 81532206.0,
      "step": 2555
    },
    {
      "entropy": 0.7410243358463049,
      "epoch": 0.23516828540376422,
      "grad_norm": 0.1479833573102951,
      "learning_rate": 9.21642592081455e-05,
      "loss": 0.7201,
      "mean_token_accuracy": 0.7879518754780293,
      "num_tokens": 81564974.0,
      "step": 2556
    },
    {
      "entropy": 0.947943814098835,
      "epoch": 0.23526029177520544,
      "grad_norm": 0.1581486612558365,
      "learning_rate": 9.216119238200387e-05,
      "loss": 0.9275,
      "mean_token_accuracy": 0.7373316399753094,
      "num_tokens": 81597111.0,
      "step": 2557
    },
    {
      "entropy": 0.8724577724933624,
      "epoch": 0.23535229814664665,
      "grad_norm": 0.15897764265537262,
      "learning_rate": 9.215812555586224e-05,
      "loss": 0.8761,
      "mean_token_accuracy": 0.7483564466238022,
      "num_tokens": 81629006.0,
      "step": 2558
    },
    {
      "entropy": 0.9332837536931038,
      "epoch": 0.23544430451808787,
      "grad_norm": 0.1694025844335556,
      "learning_rate": 9.21550587297206e-05,
      "loss": 0.9502,
      "mean_token_accuracy": 0.734094500541687,
      "num_tokens": 81661178.0,
      "step": 2559
    },
    {
      "entropy": 0.8224003836512566,
      "epoch": 0.2355363108895291,
      "grad_norm": 0.1574006825685501,
      "learning_rate": 9.2151991903579e-05,
      "loss": 0.815,
      "mean_token_accuracy": 0.7603223137557507,
      "num_tokens": 81692854.0,
      "step": 2560
    },
    {
      "entropy": 0.852325577288866,
      "epoch": 0.23562831726097033,
      "grad_norm": 0.16053256392478943,
      "learning_rate": 9.214892507743737e-05,
      "loss": 0.857,
      "mean_token_accuracy": 0.7526869662106037,
      "num_tokens": 81724837.0,
      "step": 2561
    },
    {
      "entropy": 0.8927282057702541,
      "epoch": 0.23572032363241155,
      "grad_norm": 0.17112049460411072,
      "learning_rate": 9.214585825129574e-05,
      "loss": 0.8998,
      "mean_token_accuracy": 0.747310109436512,
      "num_tokens": 81755583.0,
      "step": 2562
    },
    {
      "entropy": 0.9129526577889919,
      "epoch": 0.23581233000385277,
      "grad_norm": 0.16457533836364746,
      "learning_rate": 9.21427914251541e-05,
      "loss": 0.9135,
      "mean_token_accuracy": 0.7380050681531429,
      "num_tokens": 81787850.0,
      "step": 2563
    },
    {
      "entropy": 0.805546959862113,
      "epoch": 0.23590433637529398,
      "grad_norm": 0.1590842753648758,
      "learning_rate": 9.213972459901249e-05,
      "loss": 0.8048,
      "mean_token_accuracy": 0.7718872167170048,
      "num_tokens": 81819977.0,
      "step": 2564
    },
    {
      "entropy": 0.8548023030161858,
      "epoch": 0.2359963427467352,
      "grad_norm": 0.1820479929447174,
      "learning_rate": 9.213665777287087e-05,
      "loss": 0.9078,
      "mean_token_accuracy": 0.749085821211338,
      "num_tokens": 81851428.0,
      "step": 2565
    },
    {
      "entropy": 0.8179223611950874,
      "epoch": 0.23608834911817644,
      "grad_norm": 0.16495129466056824,
      "learning_rate": 9.213359094672923e-05,
      "loss": 0.8191,
      "mean_token_accuracy": 0.7651279419660568,
      "num_tokens": 81883258.0,
      "step": 2566
    },
    {
      "entropy": 0.9383137710392475,
      "epoch": 0.23618035548961766,
      "grad_norm": 0.16602875292301178,
      "learning_rate": 9.213052412058762e-05,
      "loss": 0.9703,
      "mean_token_accuracy": 0.72605886682868,
      "num_tokens": 81915384.0,
      "step": 2567
    },
    {
      "entropy": 0.8610228858888149,
      "epoch": 0.23627236186105888,
      "grad_norm": 0.1534270942211151,
      "learning_rate": 9.212745729444598e-05,
      "loss": 0.8373,
      "mean_token_accuracy": 0.7573888599872589,
      "num_tokens": 81948152.0,
      "step": 2568
    },
    {
      "entropy": 0.8139781281352043,
      "epoch": 0.2363643682325001,
      "grad_norm": 0.15549415349960327,
      "learning_rate": 9.212439046830435e-05,
      "loss": 0.8038,
      "mean_token_accuracy": 0.771694865077734,
      "num_tokens": 81980149.0,
      "step": 2569
    },
    {
      "entropy": 0.8569931909441948,
      "epoch": 0.2364563746039413,
      "grad_norm": 0.16208238899707794,
      "learning_rate": 9.212132364216273e-05,
      "loss": 0.8288,
      "mean_token_accuracy": 0.7669320553541183,
      "num_tokens": 82011646.0,
      "step": 2570
    },
    {
      "entropy": 0.9148797579109669,
      "epoch": 0.23654838097538256,
      "grad_norm": 0.16228866577148438,
      "learning_rate": 9.211825681602111e-05,
      "loss": 0.8842,
      "mean_token_accuracy": 0.7461399398744106,
      "num_tokens": 82043787.0,
      "step": 2571
    },
    {
      "entropy": 0.9074675999581814,
      "epoch": 0.23664038734682377,
      "grad_norm": 0.1614610254764557,
      "learning_rate": 9.211518998987948e-05,
      "loss": 0.8845,
      "mean_token_accuracy": 0.7449966222047806,
      "num_tokens": 82074095.0,
      "step": 2572
    },
    {
      "entropy": 0.8896848931908607,
      "epoch": 0.236732393718265,
      "grad_norm": 0.16316324472427368,
      "learning_rate": 9.211212316373785e-05,
      "loss": 0.8743,
      "mean_token_accuracy": 0.7568037584424019,
      "num_tokens": 82106285.0,
      "step": 2573
    },
    {
      "entropy": 0.8505956493318081,
      "epoch": 0.2368244000897062,
      "grad_norm": 0.15263493359088898,
      "learning_rate": 9.210905633759622e-05,
      "loss": 0.8526,
      "mean_token_accuracy": 0.7493925057351589,
      "num_tokens": 82137884.0,
      "step": 2574
    },
    {
      "entropy": 0.8491203337907791,
      "epoch": 0.23691640646114742,
      "grad_norm": 0.16360679268836975,
      "learning_rate": 9.21059895114546e-05,
      "loss": 0.8503,
      "mean_token_accuracy": 0.7544326186180115,
      "num_tokens": 82170428.0,
      "step": 2575
    },
    {
      "entropy": 0.8997663520276546,
      "epoch": 0.23700841283258867,
      "grad_norm": 0.16809675097465515,
      "learning_rate": 9.210292268531298e-05,
      "loss": 0.8893,
      "mean_token_accuracy": 0.745012391358614,
      "num_tokens": 82201988.0,
      "step": 2576
    },
    {
      "entropy": 0.7367682419717312,
      "epoch": 0.23710041920402988,
      "grad_norm": 0.16096584498882294,
      "learning_rate": 9.209985585917135e-05,
      "loss": 0.7524,
      "mean_token_accuracy": 0.7816621884703636,
      "num_tokens": 82234756.0,
      "step": 2577
    },
    {
      "entropy": 0.932623352855444,
      "epoch": 0.2371924255754711,
      "grad_norm": 0.16637244820594788,
      "learning_rate": 9.209678903302972e-05,
      "loss": 0.9362,
      "mean_token_accuracy": 0.7314806766808033,
      "num_tokens": 82265461.0,
      "step": 2578
    },
    {
      "entropy": 0.7842476572841406,
      "epoch": 0.23728443194691232,
      "grad_norm": 0.16584260761737823,
      "learning_rate": 9.20937222068881e-05,
      "loss": 0.7983,
      "mean_token_accuracy": 0.7735977880656719,
      "num_tokens": 82297510.0,
      "step": 2579
    },
    {
      "entropy": 0.8925691936165094,
      "epoch": 0.23737643831835353,
      "grad_norm": 0.16861438751220703,
      "learning_rate": 9.209065538074647e-05,
      "loss": 0.8947,
      "mean_token_accuracy": 0.7467829175293446,
      "num_tokens": 82328877.0,
      "step": 2580
    },
    {
      "entropy": 0.7898526359349489,
      "epoch": 0.23746844468979478,
      "grad_norm": 0.15126992762088776,
      "learning_rate": 9.208758855460485e-05,
      "loss": 0.7654,
      "mean_token_accuracy": 0.776801835745573,
      "num_tokens": 82360204.0,
      "step": 2581
    },
    {
      "entropy": 0.8951155059039593,
      "epoch": 0.237560451061236,
      "grad_norm": 0.1577247530221939,
      "learning_rate": 9.208452172846322e-05,
      "loss": 0.8759,
      "mean_token_accuracy": 0.747950479388237,
      "num_tokens": 82392224.0,
      "step": 2582
    },
    {
      "entropy": 0.773871161043644,
      "epoch": 0.2376524574326772,
      "grad_norm": 0.15025973320007324,
      "learning_rate": 9.20814549023216e-05,
      "loss": 0.7613,
      "mean_token_accuracy": 0.7803023681044579,
      "num_tokens": 82424501.0,
      "step": 2583
    },
    {
      "entropy": 0.8289970904588699,
      "epoch": 0.23774446380411843,
      "grad_norm": 0.15839435160160065,
      "learning_rate": 9.207838807617996e-05,
      "loss": 0.8487,
      "mean_token_accuracy": 0.7609178349375725,
      "num_tokens": 82456287.0,
      "step": 2584
    },
    {
      "entropy": 0.8128178790211678,
      "epoch": 0.23783647017555964,
      "grad_norm": 0.15599201619625092,
      "learning_rate": 9.207532125003833e-05,
      "loss": 0.8145,
      "mean_token_accuracy": 0.7643807418644428,
      "num_tokens": 82488032.0,
      "step": 2585
    },
    {
      "entropy": 0.7721768952906132,
      "epoch": 0.2379284765470009,
      "grad_norm": 0.15190310776233673,
      "learning_rate": 9.207225442389671e-05,
      "loss": 0.7587,
      "mean_token_accuracy": 0.7803276404738426,
      "num_tokens": 82520215.0,
      "step": 2586
    },
    {
      "entropy": 0.8277981504797935,
      "epoch": 0.2380204829184421,
      "grad_norm": 0.15725043416023254,
      "learning_rate": 9.20691875977551e-05,
      "loss": 0.8232,
      "mean_token_accuracy": 0.763704463839531,
      "num_tokens": 82552248.0,
      "step": 2587
    },
    {
      "entropy": 0.9363185465335846,
      "epoch": 0.23811248928988332,
      "grad_norm": 0.16090142726898193,
      "learning_rate": 9.206612077161346e-05,
      "loss": 0.9481,
      "mean_token_accuracy": 0.7324553467333317,
      "num_tokens": 82584464.0,
      "step": 2588
    },
    {
      "entropy": 0.8762231264263391,
      "epoch": 0.23820449566132454,
      "grad_norm": 0.15529441833496094,
      "learning_rate": 9.206305394547183e-05,
      "loss": 0.8843,
      "mean_token_accuracy": 0.7425314970314503,
      "num_tokens": 82615614.0,
      "step": 2589
    },
    {
      "entropy": 0.8855576477944851,
      "epoch": 0.23829650203276576,
      "grad_norm": 0.15841518342494965,
      "learning_rate": 9.20599871193302e-05,
      "loss": 0.8907,
      "mean_token_accuracy": 0.74138443171978,
      "num_tokens": 82647220.0,
      "step": 2590
    },
    {
      "entropy": 0.8239800482988358,
      "epoch": 0.238388508404207,
      "grad_norm": 0.15593115985393524,
      "learning_rate": 9.205692029318858e-05,
      "loss": 0.8139,
      "mean_token_accuracy": 0.7625427469611168,
      "num_tokens": 82679644.0,
      "step": 2591
    },
    {
      "entropy": 0.8964928910136223,
      "epoch": 0.23848051477564822,
      "grad_norm": 0.15964682400226593,
      "learning_rate": 9.205385346704696e-05,
      "loss": 0.8927,
      "mean_token_accuracy": 0.7472539842128754,
      "num_tokens": 82712208.0,
      "step": 2592
    },
    {
      "entropy": 0.9038722440600395,
      "epoch": 0.23857252114708943,
      "grad_norm": 0.1596638709306717,
      "learning_rate": 9.205078664090533e-05,
      "loss": 0.8964,
      "mean_token_accuracy": 0.751713864505291,
      "num_tokens": 82744118.0,
      "step": 2593
    },
    {
      "entropy": 0.9071976915001869,
      "epoch": 0.23866452751853065,
      "grad_norm": 0.15747584402561188,
      "learning_rate": 9.204771981476371e-05,
      "loss": 0.891,
      "mean_token_accuracy": 0.7439255230128765,
      "num_tokens": 82776142.0,
      "step": 2594
    },
    {
      "entropy": 0.8433008193969727,
      "epoch": 0.23875653388997187,
      "grad_norm": 0.15473048388957977,
      "learning_rate": 9.204465298862208e-05,
      "loss": 0.8263,
      "mean_token_accuracy": 0.7658606506884098,
      "num_tokens": 82807770.0,
      "step": 2595
    },
    {
      "entropy": 0.863341148942709,
      "epoch": 0.2388485402614131,
      "grad_norm": 0.16441629827022552,
      "learning_rate": 9.204158616248045e-05,
      "loss": 0.8523,
      "mean_token_accuracy": 0.7584327273070812,
      "num_tokens": 82839713.0,
      "step": 2596
    },
    {
      "entropy": 0.7978395354002714,
      "epoch": 0.23894054663285433,
      "grad_norm": 0.15780846774578094,
      "learning_rate": 9.203851933633883e-05,
      "loss": 0.7875,
      "mean_token_accuracy": 0.7738268673419952,
      "num_tokens": 82872152.0,
      "step": 2597
    },
    {
      "entropy": 0.8866297453641891,
      "epoch": 0.23903255300429554,
      "grad_norm": 0.15875214338302612,
      "learning_rate": 9.203545251019721e-05,
      "loss": 0.8871,
      "mean_token_accuracy": 0.746501337736845,
      "num_tokens": 82903483.0,
      "step": 2598
    },
    {
      "entropy": 0.8927215989679098,
      "epoch": 0.23912455937573676,
      "grad_norm": 0.15952229499816895,
      "learning_rate": 9.203238568405558e-05,
      "loss": 0.9062,
      "mean_token_accuracy": 0.7440187446773052,
      "num_tokens": 82935338.0,
      "step": 2599
    },
    {
      "entropy": 0.916540153324604,
      "epoch": 0.23921656574717798,
      "grad_norm": 0.16096587479114532,
      "learning_rate": 9.202931885791395e-05,
      "loss": 0.9103,
      "mean_token_accuracy": 0.7415629252791405,
      "num_tokens": 82966394.0,
      "step": 2600
    },
    {
      "entropy": 0.9564827792346478,
      "epoch": 0.23930857211861922,
      "grad_norm": 0.1726701557636261,
      "learning_rate": 9.202625203177231e-05,
      "loss": 0.9624,
      "mean_token_accuracy": 0.7274199016392231,
      "num_tokens": 82998724.0,
      "step": 2601
    },
    {
      "entropy": 0.8868421614170074,
      "epoch": 0.23940057849006044,
      "grad_norm": 0.15924468636512756,
      "learning_rate": 9.202318520563071e-05,
      "loss": 0.8755,
      "mean_token_accuracy": 0.7485526874661446,
      "num_tokens": 83030381.0,
      "step": 2602
    },
    {
      "entropy": 0.8591272067278624,
      "epoch": 0.23949258486150166,
      "grad_norm": 0.1602860540151596,
      "learning_rate": 9.202011837948908e-05,
      "loss": 0.8572,
      "mean_token_accuracy": 0.755849078297615,
      "num_tokens": 83062356.0,
      "step": 2603
    },
    {
      "entropy": 0.7795567288994789,
      "epoch": 0.23958459123294287,
      "grad_norm": 0.15206968784332275,
      "learning_rate": 9.201705155334744e-05,
      "loss": 0.7886,
      "mean_token_accuracy": 0.7734497338533401,
      "num_tokens": 83094255.0,
      "step": 2604
    },
    {
      "entropy": 0.8585075736045837,
      "epoch": 0.23967659760438412,
      "grad_norm": 0.16777963936328888,
      "learning_rate": 9.201398472720581e-05,
      "loss": 0.8888,
      "mean_token_accuracy": 0.7429855316877365,
      "num_tokens": 83125823.0,
      "step": 2605
    },
    {
      "entropy": 0.8301791995763779,
      "epoch": 0.23976860397582533,
      "grad_norm": 0.16642363369464874,
      "learning_rate": 9.201091790106419e-05,
      "loss": 0.8426,
      "mean_token_accuracy": 0.7573317922651768,
      "num_tokens": 83157752.0,
      "step": 2606
    },
    {
      "entropy": 0.8126261327415705,
      "epoch": 0.23986061034726655,
      "grad_norm": 0.15265722572803497,
      "learning_rate": 9.200785107492257e-05,
      "loss": 0.8093,
      "mean_token_accuracy": 0.7658283673226833,
      "num_tokens": 83190357.0,
      "step": 2607
    },
    {
      "entropy": 0.8466899078339338,
      "epoch": 0.23995261671870777,
      "grad_norm": 0.15767191350460052,
      "learning_rate": 9.200478424878094e-05,
      "loss": 0.8652,
      "mean_token_accuracy": 0.7502413056790829,
      "num_tokens": 83222736.0,
      "step": 2608
    },
    {
      "entropy": 0.8593747317790985,
      "epoch": 0.24004462309014898,
      "grad_norm": 0.15638549625873566,
      "learning_rate": 9.200171742263931e-05,
      "loss": 0.8276,
      "mean_token_accuracy": 0.7605813257396221,
      "num_tokens": 83253962.0,
      "step": 2609
    },
    {
      "entropy": 0.8759680725634098,
      "epoch": 0.24013662946159023,
      "grad_norm": 0.15610621869564056,
      "learning_rate": 9.199865059649769e-05,
      "loss": 0.87,
      "mean_token_accuracy": 0.7533559687435627,
      "num_tokens": 83285365.0,
      "step": 2610
    },
    {
      "entropy": 0.8550430163741112,
      "epoch": 0.24022863583303145,
      "grad_norm": 0.15408548712730408,
      "learning_rate": 9.199558377035606e-05,
      "loss": 0.8358,
      "mean_token_accuracy": 0.7593735866248608,
      "num_tokens": 83317088.0,
      "step": 2611
    },
    {
      "entropy": 0.8566286116838455,
      "epoch": 0.24032064220447266,
      "grad_norm": 0.15875552594661713,
      "learning_rate": 9.199251694421444e-05,
      "loss": 0.8365,
      "mean_token_accuracy": 0.7593943551182747,
      "num_tokens": 83348565.0,
      "step": 2612
    },
    {
      "entropy": 0.8715252727270126,
      "epoch": 0.24041264857591388,
      "grad_norm": 0.15688177943229675,
      "learning_rate": 9.198945011807281e-05,
      "loss": 0.8596,
      "mean_token_accuracy": 0.7547856643795967,
      "num_tokens": 83380388.0,
      "step": 2613
    },
    {
      "entropy": 0.7947793640196323,
      "epoch": 0.2405046549473551,
      "grad_norm": 0.15236447751522064,
      "learning_rate": 9.198638329193119e-05,
      "loss": 0.7808,
      "mean_token_accuracy": 0.7689564675092697,
      "num_tokens": 83412866.0,
      "step": 2614
    },
    {
      "entropy": 0.8756981119513512,
      "epoch": 0.24059666131879634,
      "grad_norm": 0.1606808602809906,
      "learning_rate": 9.198331646578956e-05,
      "loss": 0.8887,
      "mean_token_accuracy": 0.7483172044157982,
      "num_tokens": 83445023.0,
      "step": 2615
    },
    {
      "entropy": 0.7876863349229097,
      "epoch": 0.24068866769023756,
      "grad_norm": 0.15449967980384827,
      "learning_rate": 9.198024963964793e-05,
      "loss": 0.7756,
      "mean_token_accuracy": 0.772075679153204,
      "num_tokens": 83476743.0,
      "step": 2616
    },
    {
      "entropy": 0.820978781208396,
      "epoch": 0.24078067406167877,
      "grad_norm": 0.15552756190299988,
      "learning_rate": 9.197718281350631e-05,
      "loss": 0.8361,
      "mean_token_accuracy": 0.7619118988513947,
      "num_tokens": 83509011.0,
      "step": 2617
    },
    {
      "entropy": 0.886098101735115,
      "epoch": 0.24087268043312,
      "grad_norm": 0.1572766751050949,
      "learning_rate": 9.197411598736469e-05,
      "loss": 0.8783,
      "mean_token_accuracy": 0.7465294003486633,
      "num_tokens": 83540633.0,
      "step": 2618
    },
    {
      "entropy": 0.890236135572195,
      "epoch": 0.2409646868045612,
      "grad_norm": 0.15602342784404755,
      "learning_rate": 9.197104916122306e-05,
      "loss": 0.8864,
      "mean_token_accuracy": 0.7426219657063484,
      "num_tokens": 83573174.0,
      "step": 2619
    },
    {
      "entropy": 0.8652410339564085,
      "epoch": 0.24105669317600245,
      "grad_norm": 0.15635810792446136,
      "learning_rate": 9.196798233508142e-05,
      "loss": 0.8551,
      "mean_token_accuracy": 0.7552597634494305,
      "num_tokens": 83605527.0,
      "step": 2620
    },
    {
      "entropy": 0.9144003298133612,
      "epoch": 0.24114869954744367,
      "grad_norm": 0.16452771425247192,
      "learning_rate": 9.196491550893979e-05,
      "loss": 0.916,
      "mean_token_accuracy": 0.7344015017151833,
      "num_tokens": 83636086.0,
      "step": 2621
    },
    {
      "entropy": 0.8384215757250786,
      "epoch": 0.24124070591888488,
      "grad_norm": 0.15686078369617462,
      "learning_rate": 9.196184868279817e-05,
      "loss": 0.8452,
      "mean_token_accuracy": 0.7584632374346256,
      "num_tokens": 83667763.0,
      "step": 2622
    },
    {
      "entropy": 0.8894981704652309,
      "epoch": 0.2413327122903261,
      "grad_norm": 0.1682279109954834,
      "learning_rate": 9.195878185665656e-05,
      "loss": 0.9143,
      "mean_token_accuracy": 0.7388463541865349,
      "num_tokens": 83699157.0,
      "step": 2623
    },
    {
      "entropy": 0.7924075946211815,
      "epoch": 0.24142471866176732,
      "grad_norm": 0.16573230922222137,
      "learning_rate": 9.195571503051492e-05,
      "loss": 0.7927,
      "mean_token_accuracy": 0.769513163715601,
      "num_tokens": 83731605.0,
      "step": 2624
    },
    {
      "entropy": 0.886124424636364,
      "epoch": 0.24151672503320856,
      "grad_norm": 0.16257694363594055,
      "learning_rate": 9.19526482043733e-05,
      "loss": 0.8837,
      "mean_token_accuracy": 0.7484662346541882,
      "num_tokens": 83763623.0,
      "step": 2625
    },
    {
      "entropy": 0.7490490674972534,
      "epoch": 0.24160873140464978,
      "grad_norm": 0.15446971356868744,
      "learning_rate": 9.194958137823167e-05,
      "loss": 0.7194,
      "mean_token_accuracy": 0.7878251112997532,
      "num_tokens": 83794907.0,
      "step": 2626
    },
    {
      "entropy": 0.8329783119261265,
      "epoch": 0.241700737776091,
      "grad_norm": 0.157354474067688,
      "learning_rate": 9.194651455209004e-05,
      "loss": 0.8399,
      "mean_token_accuracy": 0.7608885243535042,
      "num_tokens": 83827157.0,
      "step": 2627
    },
    {
      "entropy": 0.8635793663561344,
      "epoch": 0.2417927441475322,
      "grad_norm": 0.1505056470632553,
      "learning_rate": 9.194344772594842e-05,
      "loss": 0.8504,
      "mean_token_accuracy": 0.7575216144323349,
      "num_tokens": 83859264.0,
      "step": 2628
    },
    {
      "entropy": 0.8226618655025959,
      "epoch": 0.24188475051897343,
      "grad_norm": 0.15357916057109833,
      "learning_rate": 9.19403808998068e-05,
      "loss": 0.8113,
      "mean_token_accuracy": 0.7639116980135441,
      "num_tokens": 83891362.0,
      "step": 2629
    },
    {
      "entropy": 0.7985224332660437,
      "epoch": 0.24197675689041467,
      "grad_norm": 0.15304464101791382,
      "learning_rate": 9.193731407366517e-05,
      "loss": 0.7773,
      "mean_token_accuracy": 0.7748107723891735,
      "num_tokens": 83924037.0,
      "step": 2630
    },
    {
      "entropy": 0.8947369679808617,
      "epoch": 0.2420687632618559,
      "grad_norm": 0.15743331611156464,
      "learning_rate": 9.193424724752354e-05,
      "loss": 0.8879,
      "mean_token_accuracy": 0.7442633658647537,
      "num_tokens": 83956611.0,
      "step": 2631
    },
    {
      "entropy": 0.9805807694792747,
      "epoch": 0.2421607696332971,
      "grad_norm": 0.16460411250591278,
      "learning_rate": 9.193118042138191e-05,
      "loss": 0.9964,
      "mean_token_accuracy": 0.7210050635039806,
      "num_tokens": 83988339.0,
      "step": 2632
    },
    {
      "entropy": 0.8933026194572449,
      "epoch": 0.24225277600473832,
      "grad_norm": 0.16105754673480988,
      "learning_rate": 9.192811359524029e-05,
      "loss": 0.881,
      "mean_token_accuracy": 0.7452365085482597,
      "num_tokens": 84020021.0,
      "step": 2633
    },
    {
      "entropy": 0.7299044094979763,
      "epoch": 0.24234478237617954,
      "grad_norm": 0.15028993785381317,
      "learning_rate": 9.192504676909867e-05,
      "loss": 0.7228,
      "mean_token_accuracy": 0.7889002077281475,
      "num_tokens": 84051860.0,
      "step": 2634
    },
    {
      "entropy": 0.8793946988880634,
      "epoch": 0.24243678874762079,
      "grad_norm": 0.15434150397777557,
      "learning_rate": 9.192197994295704e-05,
      "loss": 0.8738,
      "mean_token_accuracy": 0.7490848749876022,
      "num_tokens": 84083666.0,
      "step": 2635
    },
    {
      "entropy": 0.8807572089135647,
      "epoch": 0.242528795119062,
      "grad_norm": 0.16311722993850708,
      "learning_rate": 9.19189131168154e-05,
      "loss": 0.9015,
      "mean_token_accuracy": 0.7450370416045189,
      "num_tokens": 84115538.0,
      "step": 2636
    },
    {
      "entropy": 0.8927312400192022,
      "epoch": 0.24262080149050322,
      "grad_norm": 0.16344571113586426,
      "learning_rate": 9.191584629067379e-05,
      "loss": 0.8941,
      "mean_token_accuracy": 0.7414200603961945,
      "num_tokens": 84146747.0,
      "step": 2637
    },
    {
      "entropy": 0.8171936497092247,
      "epoch": 0.24271280786194444,
      "grad_norm": 0.15994304418563843,
      "learning_rate": 9.191277946453217e-05,
      "loss": 0.8354,
      "mean_token_accuracy": 0.7618621848523617,
      "num_tokens": 84179047.0,
      "step": 2638
    },
    {
      "entropy": 0.865275664255023,
      "epoch": 0.24280481423338565,
      "grad_norm": 0.1583852618932724,
      "learning_rate": 9.190971263839054e-05,
      "loss": 0.8557,
      "mean_token_accuracy": 0.7538484781980515,
      "num_tokens": 84211620.0,
      "step": 2639
    },
    {
      "entropy": 0.8012575134634972,
      "epoch": 0.2428968206048269,
      "grad_norm": 0.15329203009605408,
      "learning_rate": 9.19066458122489e-05,
      "loss": 0.7942,
      "mean_token_accuracy": 0.7706322707235813,
      "num_tokens": 84243999.0,
      "step": 2640
    },
    {
      "entropy": 0.8636590801179409,
      "epoch": 0.2429888269762681,
      "grad_norm": 0.16565915942192078,
      "learning_rate": 9.190357898610729e-05,
      "loss": 0.8675,
      "mean_token_accuracy": 0.7586712315678596,
      "num_tokens": 84276767.0,
      "step": 2641
    },
    {
      "entropy": 0.8549296520650387,
      "epoch": 0.24308083334770933,
      "grad_norm": 0.15834064781665802,
      "learning_rate": 9.190051215996565e-05,
      "loss": 0.855,
      "mean_token_accuracy": 0.7535001114010811,
      "num_tokens": 84308903.0,
      "step": 2642
    },
    {
      "entropy": 0.8865865804255009,
      "epoch": 0.24317283971915055,
      "grad_norm": 0.16084182262420654,
      "learning_rate": 9.189744533382403e-05,
      "loss": 0.8803,
      "mean_token_accuracy": 0.7461336590349674,
      "num_tokens": 84340211.0,
      "step": 2643
    },
    {
      "entropy": 0.9162241593003273,
      "epoch": 0.24326484609059176,
      "grad_norm": 0.16639840602874756,
      "learning_rate": 9.18943785076824e-05,
      "loss": 0.9025,
      "mean_token_accuracy": 0.7407877035439014,
      "num_tokens": 84371890.0,
      "step": 2644
    },
    {
      "entropy": 0.9462048560380936,
      "epoch": 0.243356852462033,
      "grad_norm": 0.16633178293704987,
      "learning_rate": 9.189131168154078e-05,
      "loss": 0.9524,
      "mean_token_accuracy": 0.7309900298714638,
      "num_tokens": 84404164.0,
      "step": 2645
    },
    {
      "entropy": 0.9063115231692791,
      "epoch": 0.24344885883347422,
      "grad_norm": 0.15918679535388947,
      "learning_rate": 9.188824485539915e-05,
      "loss": 0.8938,
      "mean_token_accuracy": 0.7438768744468689,
      "num_tokens": 84435801.0,
      "step": 2646
    },
    {
      "entropy": 0.9691201839596033,
      "epoch": 0.24354086520491544,
      "grad_norm": 0.16625137627124786,
      "learning_rate": 9.188517802925752e-05,
      "loss": 0.956,
      "mean_token_accuracy": 0.7303054220974445,
      "num_tokens": 84467645.0,
      "step": 2647
    },
    {
      "entropy": 0.9383071009069681,
      "epoch": 0.24363287157635666,
      "grad_norm": 0.16714230179786682,
      "learning_rate": 9.18821112031159e-05,
      "loss": 0.9549,
      "mean_token_accuracy": 0.7278607599437237,
      "num_tokens": 84499996.0,
      "step": 2648
    },
    {
      "entropy": 0.8844608552753925,
      "epoch": 0.24372487794779787,
      "grad_norm": 0.15624892711639404,
      "learning_rate": 9.187904437697428e-05,
      "loss": 0.8658,
      "mean_token_accuracy": 0.7515738792717457,
      "num_tokens": 84532268.0,
      "step": 2649
    },
    {
      "entropy": 0.8463068790733814,
      "epoch": 0.24381688431923912,
      "grad_norm": 0.16070342063903809,
      "learning_rate": 9.187597755083265e-05,
      "loss": 0.8443,
      "mean_token_accuracy": 0.7610120140016079,
      "num_tokens": 84564612.0,
      "step": 2650
    },
    {
      "entropy": 0.7967936377972364,
      "epoch": 0.24390889069068034,
      "grad_norm": 0.15075263381004333,
      "learning_rate": 9.187291072469102e-05,
      "loss": 0.7572,
      "mean_token_accuracy": 0.7795910462737083,
      "num_tokens": 84596153.0,
      "step": 2651
    },
    {
      "entropy": 0.8193823955953121,
      "epoch": 0.24400089706212155,
      "grad_norm": 0.1579352170228958,
      "learning_rate": 9.186984389854939e-05,
      "loss": 0.8095,
      "mean_token_accuracy": 0.7679202929139137,
      "num_tokens": 84627981.0,
      "step": 2652
    },
    {
      "entropy": 0.8315005972981453,
      "epoch": 0.24409290343356277,
      "grad_norm": 0.1610858142375946,
      "learning_rate": 9.186677707240777e-05,
      "loss": 0.8385,
      "mean_token_accuracy": 0.7590147145092487,
      "num_tokens": 84659711.0,
      "step": 2653
    },
    {
      "entropy": 0.9023727886378765,
      "epoch": 0.24418490980500399,
      "grad_norm": 0.16392652690410614,
      "learning_rate": 9.186371024626615e-05,
      "loss": 0.905,
      "mean_token_accuracy": 0.7444100975990295,
      "num_tokens": 84690847.0,
      "step": 2654
    },
    {
      "entropy": 0.904676515609026,
      "epoch": 0.24427691617644523,
      "grad_norm": 0.16366037726402283,
      "learning_rate": 9.186064342012452e-05,
      "loss": 0.9134,
      "mean_token_accuracy": 0.741337887942791,
      "num_tokens": 84722791.0,
      "step": 2655
    },
    {
      "entropy": 0.7539961952716112,
      "epoch": 0.24436892254788645,
      "grad_norm": 0.15588703751564026,
      "learning_rate": 9.18575765939829e-05,
      "loss": 0.7501,
      "mean_token_accuracy": 0.7780167311429977,
      "num_tokens": 84754413.0,
      "step": 2656
    },
    {
      "entropy": 0.9193182960152626,
      "epoch": 0.24446092891932766,
      "grad_norm": 0.16715286672115326,
      "learning_rate": 9.185450976784127e-05,
      "loss": 0.9504,
      "mean_token_accuracy": 0.7373606488108635,
      "num_tokens": 84786908.0,
      "step": 2657
    },
    {
      "entropy": 0.7816291432827711,
      "epoch": 0.24455293529076888,
      "grad_norm": 0.163258895277977,
      "learning_rate": 9.185144294169963e-05,
      "loss": 0.8092,
      "mean_token_accuracy": 0.7688065059483051,
      "num_tokens": 84818327.0,
      "step": 2658
    },
    {
      "entropy": 0.8555907532572746,
      "epoch": 0.2446449416622101,
      "grad_norm": 0.15940916538238525,
      "learning_rate": 9.184837611555802e-05,
      "loss": 0.857,
      "mean_token_accuracy": 0.7498989216983318,
      "num_tokens": 84850513.0,
      "step": 2659
    },
    {
      "entropy": 0.8879393730312586,
      "epoch": 0.24473694803365134,
      "grad_norm": 0.16076746582984924,
      "learning_rate": 9.18453092894164e-05,
      "loss": 0.8866,
      "mean_token_accuracy": 0.7499151080846786,
      "num_tokens": 84882466.0,
      "step": 2660
    },
    {
      "entropy": 0.8220582772046328,
      "epoch": 0.24482895440509256,
      "grad_norm": 0.15430106222629547,
      "learning_rate": 9.184224246327476e-05,
      "loss": 0.8273,
      "mean_token_accuracy": 0.7593517377972603,
      "num_tokens": 84914291.0,
      "step": 2661
    },
    {
      "entropy": 0.8259285278618336,
      "epoch": 0.24492096077653377,
      "grad_norm": 0.17020905017852783,
      "learning_rate": 9.183917563713313e-05,
      "loss": 0.8434,
      "mean_token_accuracy": 0.7559743858873844,
      "num_tokens": 84945186.0,
      "step": 2662
    },
    {
      "entropy": 0.939699862152338,
      "epoch": 0.245012967147975,
      "grad_norm": 0.16163496673107147,
      "learning_rate": 9.18361088109915e-05,
      "loss": 0.9331,
      "mean_token_accuracy": 0.7366358153522015,
      "num_tokens": 84977528.0,
      "step": 2663
    },
    {
      "entropy": 0.9474322013556957,
      "epoch": 0.2451049735194162,
      "grad_norm": 0.16548722982406616,
      "learning_rate": 9.183304198484988e-05,
      "loss": 0.9361,
      "mean_token_accuracy": 0.737343467772007,
      "num_tokens": 85009208.0,
      "step": 2664
    },
    {
      "entropy": 0.8048810958862305,
      "epoch": 0.24519697989085745,
      "grad_norm": 0.1537661850452423,
      "learning_rate": 9.182997515870826e-05,
      "loss": 0.7864,
      "mean_token_accuracy": 0.7665089927613735,
      "num_tokens": 85041388.0,
      "step": 2665
    },
    {
      "entropy": 0.8987341336905956,
      "epoch": 0.24528898626229867,
      "grad_norm": 0.16110607981681824,
      "learning_rate": 9.182690833256663e-05,
      "loss": 0.9011,
      "mean_token_accuracy": 0.7383300289511681,
      "num_tokens": 85073063.0,
      "step": 2666
    },
    {
      "entropy": 0.7819987926632166,
      "epoch": 0.24538099263373989,
      "grad_norm": 0.15321919322013855,
      "learning_rate": 9.1823841506425e-05,
      "loss": 0.7756,
      "mean_token_accuracy": 0.7751513719558716,
      "num_tokens": 85105799.0,
      "step": 2667
    },
    {
      "entropy": 0.8375424910336733,
      "epoch": 0.2454729990051811,
      "grad_norm": 0.15949204564094543,
      "learning_rate": 9.182077468028338e-05,
      "loss": 0.8321,
      "mean_token_accuracy": 0.7604721635580063,
      "num_tokens": 85138116.0,
      "step": 2668
    },
    {
      "entropy": 0.8406574353575706,
      "epoch": 0.24556500537662232,
      "grad_norm": 0.15258368849754333,
      "learning_rate": 9.181770785414175e-05,
      "loss": 0.8299,
      "mean_token_accuracy": 0.7615390904247761,
      "num_tokens": 85170096.0,
      "step": 2669
    },
    {
      "entropy": 0.9376026522368193,
      "epoch": 0.24565701174806356,
      "grad_norm": 0.16982507705688477,
      "learning_rate": 9.181464102800013e-05,
      "loss": 0.9705,
      "mean_token_accuracy": 0.7290478125214577,
      "num_tokens": 85202261.0,
      "step": 2670
    },
    {
      "entropy": 0.7535418216139078,
      "epoch": 0.24574901811950478,
      "grad_norm": 0.1490194946527481,
      "learning_rate": 9.18115742018585e-05,
      "loss": 0.7343,
      "mean_token_accuracy": 0.7836470864713192,
      "num_tokens": 85233625.0,
      "step": 2671
    },
    {
      "entropy": 0.7677188292145729,
      "epoch": 0.245841024490946,
      "grad_norm": 0.15155711770057678,
      "learning_rate": 9.180850737571688e-05,
      "loss": 0.73,
      "mean_token_accuracy": 0.7867910414934158,
      "num_tokens": 85264415.0,
      "step": 2672
    },
    {
      "entropy": 0.7836126461625099,
      "epoch": 0.24593303086238721,
      "grad_norm": 0.14986424148082733,
      "learning_rate": 9.180544054957525e-05,
      "loss": 0.7615,
      "mean_token_accuracy": 0.7782730758190155,
      "num_tokens": 85297183.0,
      "step": 2673
    },
    {
      "entropy": 0.86301539093256,
      "epoch": 0.24602503723382843,
      "grad_norm": 0.15502800047397614,
      "learning_rate": 9.180237372343361e-05,
      "loss": 0.836,
      "mean_token_accuracy": 0.756617333739996,
      "num_tokens": 85329331.0,
      "step": 2674
    },
    {
      "entropy": 0.8400121983140707,
      "epoch": 0.24611704360526968,
      "grad_norm": 0.15983988344669342,
      "learning_rate": 9.1799306897292e-05,
      "loss": 0.8227,
      "mean_token_accuracy": 0.7603403367102146,
      "num_tokens": 85361436.0,
      "step": 2675
    },
    {
      "entropy": 0.8375260978937149,
      "epoch": 0.2462090499767109,
      "grad_norm": 0.1663489043712616,
      "learning_rate": 9.179624007115038e-05,
      "loss": 0.8639,
      "mean_token_accuracy": 0.7545162625610828,
      "num_tokens": 85392917.0,
      "step": 2676
    },
    {
      "entropy": 0.8764353096485138,
      "epoch": 0.2463010563481521,
      "grad_norm": 0.16168470680713654,
      "learning_rate": 9.179317324500875e-05,
      "loss": 0.8632,
      "mean_token_accuracy": 0.7571360245347023,
      "num_tokens": 85424849.0,
      "step": 2677
    },
    {
      "entropy": 0.8300259746611118,
      "epoch": 0.24639306271959333,
      "grad_norm": 0.15731461346149445,
      "learning_rate": 9.179010641886711e-05,
      "loss": 0.8269,
      "mean_token_accuracy": 0.7601824775338173,
      "num_tokens": 85456402.0,
      "step": 2678
    },
    {
      "entropy": 0.8908086344599724,
      "epoch": 0.24648506909103454,
      "grad_norm": 0.16460981965065002,
      "learning_rate": 9.178703959272548e-05,
      "loss": 0.8767,
      "mean_token_accuracy": 0.7521955110132694,
      "num_tokens": 85487355.0,
      "step": 2679
    },
    {
      "entropy": 0.8082699552178383,
      "epoch": 0.2465770754624758,
      "grad_norm": 0.15998505055904388,
      "learning_rate": 9.178397276658388e-05,
      "loss": 0.8081,
      "mean_token_accuracy": 0.7673908397555351,
      "num_tokens": 85519212.0,
      "step": 2680
    },
    {
      "entropy": 0.8146677520126104,
      "epoch": 0.246669081833917,
      "grad_norm": 0.16086260974407196,
      "learning_rate": 9.178090594044224e-05,
      "loss": 0.8041,
      "mean_token_accuracy": 0.7676817029714584,
      "num_tokens": 85551048.0,
      "step": 2681
    },
    {
      "entropy": 0.9513622373342514,
      "epoch": 0.24676108820535822,
      "grad_norm": 0.1660260260105133,
      "learning_rate": 9.177783911430061e-05,
      "loss": 0.9405,
      "mean_token_accuracy": 0.7350008338689804,
      "num_tokens": 85582757.0,
      "step": 2682
    },
    {
      "entropy": 0.8167126923799515,
      "epoch": 0.24685309457679944,
      "grad_norm": 0.160313680768013,
      "learning_rate": 9.177477228815898e-05,
      "loss": 0.8148,
      "mean_token_accuracy": 0.7635754607617855,
      "num_tokens": 85614080.0,
      "step": 2683
    },
    {
      "entropy": 0.8851228170096874,
      "epoch": 0.24694510094824065,
      "grad_norm": 0.16047237813472748,
      "learning_rate": 9.177170546201736e-05,
      "loss": 0.8664,
      "mean_token_accuracy": 0.7538099028170109,
      "num_tokens": 85646237.0,
      "step": 2684
    },
    {
      "entropy": 0.9340087380260229,
      "epoch": 0.2470371073196819,
      "grad_norm": 0.19659104943275452,
      "learning_rate": 9.176863863587574e-05,
      "loss": 0.9246,
      "mean_token_accuracy": 0.7381389103829861,
      "num_tokens": 85678354.0,
      "step": 2685
    },
    {
      "entropy": 0.8314972892403603,
      "epoch": 0.24712911369112311,
      "grad_norm": 0.15527598559856415,
      "learning_rate": 9.176557180973411e-05,
      "loss": 0.8055,
      "mean_token_accuracy": 0.7687579542398453,
      "num_tokens": 85711053.0,
      "step": 2686
    },
    {
      "entropy": 0.8430221900343895,
      "epoch": 0.24722112006256433,
      "grad_norm": 0.15596242249011993,
      "learning_rate": 9.176250498359249e-05,
      "loss": 0.8394,
      "mean_token_accuracy": 0.7579434104263783,
      "num_tokens": 85743109.0,
      "step": 2687
    },
    {
      "entropy": 0.8469304256141186,
      "epoch": 0.24731312643400555,
      "grad_norm": 0.15543736517429352,
      "learning_rate": 9.175943815745086e-05,
      "loss": 0.8535,
      "mean_token_accuracy": 0.7554105930030346,
      "num_tokens": 85775299.0,
      "step": 2688
    },
    {
      "entropy": 0.8556807320564985,
      "epoch": 0.24740513280544676,
      "grad_norm": 0.167496457695961,
      "learning_rate": 9.175637133130923e-05,
      "loss": 0.8722,
      "mean_token_accuracy": 0.7528364174067974,
      "num_tokens": 85807589.0,
      "step": 2689
    },
    {
      "entropy": 0.8343632966279984,
      "epoch": 0.247497139176888,
      "grad_norm": 0.16537676751613617,
      "learning_rate": 9.175330450516761e-05,
      "loss": 0.8639,
      "mean_token_accuracy": 0.748880248516798,
      "num_tokens": 85839036.0,
      "step": 2690
    },
    {
      "entropy": 0.8901041727513075,
      "epoch": 0.24758914554832923,
      "grad_norm": 0.16190777719020844,
      "learning_rate": 9.175023767902599e-05,
      "loss": 0.901,
      "mean_token_accuracy": 0.7467170841991901,
      "num_tokens": 85871692.0,
      "step": 2691
    },
    {
      "entropy": 0.9419623985886574,
      "epoch": 0.24768115191977044,
      "grad_norm": 0.17066030204296112,
      "learning_rate": 9.174717085288436e-05,
      "loss": 0.9671,
      "mean_token_accuracy": 0.73309026658535,
      "num_tokens": 85903431.0,
      "step": 2692
    },
    {
      "entropy": 0.808463666588068,
      "epoch": 0.24777315829121166,
      "grad_norm": 0.15087269246578217,
      "learning_rate": 9.174410402674273e-05,
      "loss": 0.7967,
      "mean_token_accuracy": 0.7655615918338299,
      "num_tokens": 85935477.0,
      "step": 2693
    },
    {
      "entropy": 0.8564595691859722,
      "epoch": 0.24786516466265288,
      "grad_norm": 0.1576930284500122,
      "learning_rate": 9.17410372006011e-05,
      "loss": 0.8725,
      "mean_token_accuracy": 0.7529133185744286,
      "num_tokens": 85967629.0,
      "step": 2694
    },
    {
      "entropy": 0.8498693853616714,
      "epoch": 0.24795717103409412,
      "grad_norm": 0.15180064737796783,
      "learning_rate": 9.173797037445948e-05,
      "loss": 0.8,
      "mean_token_accuracy": 0.7682793289422989,
      "num_tokens": 85998856.0,
      "step": 2695
    },
    {
      "entropy": 0.8483615983277559,
      "epoch": 0.24804917740553534,
      "grad_norm": 0.15737025439739227,
      "learning_rate": 9.173490354831786e-05,
      "loss": 0.8329,
      "mean_token_accuracy": 0.7662335820496082,
      "num_tokens": 86031532.0,
      "step": 2696
    },
    {
      "entropy": 0.8939174152910709,
      "epoch": 0.24814118377697655,
      "grad_norm": 0.15788894891738892,
      "learning_rate": 9.173183672217623e-05,
      "loss": 0.8853,
      "mean_token_accuracy": 0.7463724762201309,
      "num_tokens": 86064072.0,
      "step": 2697
    },
    {
      "entropy": 0.788768082857132,
      "epoch": 0.24823319014841777,
      "grad_norm": 0.1507626622915268,
      "learning_rate": 9.172876989603459e-05,
      "loss": 0.771,
      "mean_token_accuracy": 0.7781832478940487,
      "num_tokens": 86096184.0,
      "step": 2698
    },
    {
      "entropy": 0.8372474499046803,
      "epoch": 0.248325196519859,
      "grad_norm": 0.15911045670509338,
      "learning_rate": 9.172570306989297e-05,
      "loss": 0.8317,
      "mean_token_accuracy": 0.7566252425312996,
      "num_tokens": 86126761.0,
      "step": 2699
    },
    {
      "entropy": 0.8818000182509422,
      "epoch": 0.24841720289130023,
      "grad_norm": 0.16109924018383026,
      "learning_rate": 9.172263624375134e-05,
      "loss": 0.8838,
      "mean_token_accuracy": 0.7476418167352676,
      "num_tokens": 86159249.0,
      "step": 2700
    },
    {
      "entropy": 0.9387632925063372,
      "epoch": 0.24850920926274145,
      "grad_norm": 0.1576528549194336,
      "learning_rate": 9.171956941760972e-05,
      "loss": 0.9146,
      "mean_token_accuracy": 0.7395677678287029,
      "num_tokens": 86191248.0,
      "step": 2701
    },
    {
      "entropy": 0.7937057130038738,
      "epoch": 0.24860121563418267,
      "grad_norm": 0.15407772362232208,
      "learning_rate": 9.171650259146809e-05,
      "loss": 0.7834,
      "mean_token_accuracy": 0.7748444937169552,
      "num_tokens": 86222909.0,
      "step": 2702
    },
    {
      "entropy": 0.7688965331763029,
      "epoch": 0.24869322200562388,
      "grad_norm": 0.15468192100524902,
      "learning_rate": 9.171343576532647e-05,
      "loss": 0.766,
      "mean_token_accuracy": 0.7780809216201305,
      "num_tokens": 86255114.0,
      "step": 2703
    },
    {
      "entropy": 0.8287820257246494,
      "epoch": 0.2487852283770651,
      "grad_norm": 0.15825626254081726,
      "learning_rate": 9.171036893918484e-05,
      "loss": 0.8464,
      "mean_token_accuracy": 0.7547273337841034,
      "num_tokens": 86287148.0,
      "step": 2704
    },
    {
      "entropy": 0.9749469496309757,
      "epoch": 0.24887723474850634,
      "grad_norm": 0.17305155098438263,
      "learning_rate": 9.170730211304321e-05,
      "loss": 0.9996,
      "mean_token_accuracy": 0.721424475312233,
      "num_tokens": 86318326.0,
      "step": 2705
    },
    {
      "entropy": 0.9338030628859997,
      "epoch": 0.24896924111994756,
      "grad_norm": 0.1631115972995758,
      "learning_rate": 9.170423528690159e-05,
      "loss": 0.9485,
      "mean_token_accuracy": 0.7307279594242573,
      "num_tokens": 86349695.0,
      "step": 2706
    },
    {
      "entropy": 0.9039729163050652,
      "epoch": 0.24906124749138878,
      "grad_norm": 0.17244620621204376,
      "learning_rate": 9.170116846075997e-05,
      "loss": 0.9467,
      "mean_token_accuracy": 0.7364336811006069,
      "num_tokens": 86381605.0,
      "step": 2707
    },
    {
      "entropy": 0.84784154035151,
      "epoch": 0.24915325386283,
      "grad_norm": 0.15662112832069397,
      "learning_rate": 9.169810163461834e-05,
      "loss": 0.8423,
      "mean_token_accuracy": 0.756760623306036,
      "num_tokens": 86413296.0,
      "step": 2708
    },
    {
      "entropy": 0.9118702076375484,
      "epoch": 0.2492452602342712,
      "grad_norm": 0.15864692628383636,
      "learning_rate": 9.169503480847671e-05,
      "loss": 0.9073,
      "mean_token_accuracy": 0.7409906648099422,
      "num_tokens": 86445469.0,
      "step": 2709
    },
    {
      "entropy": 0.8619605824351311,
      "epoch": 0.24933726660571245,
      "grad_norm": 0.15798723697662354,
      "learning_rate": 9.169196798233508e-05,
      "loss": 0.855,
      "mean_token_accuracy": 0.7568154409527779,
      "num_tokens": 86477202.0,
      "step": 2710
    },
    {
      "entropy": 0.8519086167216301,
      "epoch": 0.24942927297715367,
      "grad_norm": 0.1524599939584732,
      "learning_rate": 9.168890115619346e-05,
      "loss": 0.8237,
      "mean_token_accuracy": 0.7607614360749722,
      "num_tokens": 86509471.0,
      "step": 2711
    },
    {
      "entropy": 0.837684839963913,
      "epoch": 0.2495212793485949,
      "grad_norm": 0.15446096658706665,
      "learning_rate": 9.168583433005184e-05,
      "loss": 0.8274,
      "mean_token_accuracy": 0.7607531994581223,
      "num_tokens": 86541173.0,
      "step": 2712
    },
    {
      "entropy": 0.7872753255069256,
      "epoch": 0.2496132857200361,
      "grad_norm": 0.1569407433271408,
      "learning_rate": 9.16827675039102e-05,
      "loss": 0.7812,
      "mean_token_accuracy": 0.7752517983317375,
      "num_tokens": 86572748.0,
      "step": 2713
    },
    {
      "entropy": 0.8225241117179394,
      "epoch": 0.24970529209147732,
      "grad_norm": 0.15181763470172882,
      "learning_rate": 9.167970067776857e-05,
      "loss": 0.8005,
      "mean_token_accuracy": 0.7660845927894115,
      "num_tokens": 86604256.0,
      "step": 2714
    },
    {
      "entropy": 0.7733361534774303,
      "epoch": 0.24979729846291857,
      "grad_norm": 0.1469232141971588,
      "learning_rate": 9.167663385162696e-05,
      "loss": 0.7602,
      "mean_token_accuracy": 0.7729604281485081,
      "num_tokens": 86637024.0,
      "step": 2715
    },
    {
      "entropy": 0.914367388933897,
      "epoch": 0.24988930483435978,
      "grad_norm": 0.1622610241174698,
      "learning_rate": 9.167356702548532e-05,
      "loss": 0.9208,
      "mean_token_accuracy": 0.7436262592673302,
      "num_tokens": 86668742.0,
      "step": 2716
    },
    {
      "entropy": 0.8514330480247736,
      "epoch": 0.249981311205801,
      "grad_norm": 0.1581651121377945,
      "learning_rate": 9.16705001993437e-05,
      "loss": 0.8268,
      "mean_token_accuracy": 0.7594861201941967,
      "num_tokens": 86700155.0,
      "step": 2717
    },
    {
      "entropy": 0.923823282122612,
      "epoch": 0.2500733175772422,
      "grad_norm": 0.16166554391384125,
      "learning_rate": 9.166743337320209e-05,
      "loss": 0.9343,
      "mean_token_accuracy": 0.7373642288148403,
      "num_tokens": 86732750.0,
      "step": 2718
    },
    {
      "entropy": 0.9314404018223286,
      "epoch": 0.25016532394868346,
      "grad_norm": 0.15572430193424225,
      "learning_rate": 9.166436654706045e-05,
      "loss": 0.9279,
      "mean_token_accuracy": 0.7322697043418884,
      "num_tokens": 86765149.0,
      "step": 2719
    },
    {
      "entropy": 0.8539245277643204,
      "epoch": 0.25025733032012465,
      "grad_norm": 0.15266185998916626,
      "learning_rate": 9.166129972091882e-05,
      "loss": 0.8325,
      "mean_token_accuracy": 0.7567691169679165,
      "num_tokens": 86796184.0,
      "step": 2720
    },
    {
      "entropy": 0.6611195225268602,
      "epoch": 0.2503493366915659,
      "grad_norm": 0.14738062024116516,
      "learning_rate": 9.165823289477719e-05,
      "loss": 0.6738,
      "mean_token_accuracy": 0.8000773601233959,
      "num_tokens": 86828339.0,
      "step": 2721
    },
    {
      "entropy": 0.876882754266262,
      "epoch": 0.25044134306300714,
      "grad_norm": 0.17342521250247955,
      "learning_rate": 9.165516606863558e-05,
      "loss": 0.9216,
      "mean_token_accuracy": 0.7442650124430656,
      "num_tokens": 86860580.0,
      "step": 2722
    },
    {
      "entropy": 0.9918787181377411,
      "epoch": 0.2505333494344483,
      "grad_norm": 0.16901317238807678,
      "learning_rate": 9.165209924249395e-05,
      "loss": 0.9885,
      "mean_token_accuracy": 0.7218355387449265,
      "num_tokens": 86892016.0,
      "step": 2723
    },
    {
      "entropy": 0.8815042413771152,
      "epoch": 0.25062535580588957,
      "grad_norm": 0.17050857841968536,
      "learning_rate": 9.164903241635232e-05,
      "loss": 0.9072,
      "mean_token_accuracy": 0.7487012483179569,
      "num_tokens": 86924273.0,
      "step": 2724
    },
    {
      "entropy": 0.8073317855596542,
      "epoch": 0.25071736217733076,
      "grad_norm": 0.15635287761688232,
      "learning_rate": 9.164596559021069e-05,
      "loss": 0.8058,
      "mean_token_accuracy": 0.7733117192983627,
      "num_tokens": 86956293.0,
      "step": 2725
    },
    {
      "entropy": 0.7888832315802574,
      "epoch": 0.250809368548772,
      "grad_norm": 0.1522299200296402,
      "learning_rate": 9.164289876406907e-05,
      "loss": 0.7767,
      "mean_token_accuracy": 0.774725154042244,
      "num_tokens": 86987994.0,
      "step": 2726
    },
    {
      "entropy": 0.7929352559149265,
      "epoch": 0.25090137492021325,
      "grad_norm": 0.16323943436145782,
      "learning_rate": 9.163983193792745e-05,
      "loss": 0.7695,
      "mean_token_accuracy": 0.774309515953064,
      "num_tokens": 87019823.0,
      "step": 2727
    },
    {
      "entropy": 0.8605271484702826,
      "epoch": 0.25099338129165444,
      "grad_norm": 0.15876956284046173,
      "learning_rate": 9.163676511178582e-05,
      "loss": 0.8437,
      "mean_token_accuracy": 0.7572069205343723,
      "num_tokens": 87052005.0,
      "step": 2728
    },
    {
      "entropy": 0.8819307181984186,
      "epoch": 0.2510853876630957,
      "grad_norm": 0.1648988276720047,
      "learning_rate": 9.163369828564419e-05,
      "loss": 0.8864,
      "mean_token_accuracy": 0.7448540218174458,
      "num_tokens": 87083085.0,
      "step": 2729
    },
    {
      "entropy": 0.8360131848603487,
      "epoch": 0.25117739403453687,
      "grad_norm": 0.1561766415834427,
      "learning_rate": 9.163063145950257e-05,
      "loss": 0.8186,
      "mean_token_accuracy": 0.7697000205516815,
      "num_tokens": 87115681.0,
      "step": 2730
    },
    {
      "entropy": 0.8303769044578075,
      "epoch": 0.2512694004059781,
      "grad_norm": 0.1593429446220398,
      "learning_rate": 9.162756463336094e-05,
      "loss": 0.8189,
      "mean_token_accuracy": 0.7611366398632526,
      "num_tokens": 87146868.0,
      "step": 2731
    },
    {
      "entropy": 0.8575064688920975,
      "epoch": 0.25136140677741936,
      "grad_norm": 0.15678897500038147,
      "learning_rate": 9.162449780721932e-05,
      "loss": 0.8593,
      "mean_token_accuracy": 0.7579894997179508,
      "num_tokens": 87178352.0,
      "step": 2732
    },
    {
      "entropy": 0.8398393522948027,
      "epoch": 0.25145341314886055,
      "grad_norm": 0.15339389443397522,
      "learning_rate": 9.162143098107769e-05,
      "loss": 0.8436,
      "mean_token_accuracy": 0.7556849829852581,
      "num_tokens": 87210638.0,
      "step": 2733
    },
    {
      "entropy": 0.9134417176246643,
      "epoch": 0.2515454195203018,
      "grad_norm": 0.1616877168416977,
      "learning_rate": 9.161836415493607e-05,
      "loss": 0.9324,
      "mean_token_accuracy": 0.7390838600695133,
      "num_tokens": 87242266.0,
      "step": 2734
    },
    {
      "entropy": 0.8396364524960518,
      "epoch": 0.251637425891743,
      "grad_norm": 0.15837344527244568,
      "learning_rate": 9.161529732879443e-05,
      "loss": 0.8421,
      "mean_token_accuracy": 0.7594959661364555,
      "num_tokens": 87274311.0,
      "step": 2735
    },
    {
      "entropy": 0.9051467478275299,
      "epoch": 0.2517294322631842,
      "grad_norm": 0.16184209287166595,
      "learning_rate": 9.16122305026528e-05,
      "loss": 0.9088,
      "mean_token_accuracy": 0.7353894077241421,
      "num_tokens": 87306087.0,
      "step": 2736
    },
    {
      "entropy": 0.9203876070678234,
      "epoch": 0.25182143863462547,
      "grad_norm": 0.16413213312625885,
      "learning_rate": 9.160916367651118e-05,
      "loss": 0.9073,
      "mean_token_accuracy": 0.7399497702717781,
      "num_tokens": 87338107.0,
      "step": 2737
    },
    {
      "entropy": 0.8142125271260738,
      "epoch": 0.25191344500606666,
      "grad_norm": 0.1528116762638092,
      "learning_rate": 9.160609685036957e-05,
      "loss": 0.8064,
      "mean_token_accuracy": 0.7665814310312271,
      "num_tokens": 87369259.0,
      "step": 2738
    },
    {
      "entropy": 0.9017786495387554,
      "epoch": 0.2520054513775079,
      "grad_norm": 0.15591572225093842,
      "learning_rate": 9.160303002422793e-05,
      "loss": 0.8811,
      "mean_token_accuracy": 0.7506509199738503,
      "num_tokens": 87401480.0,
      "step": 2739
    },
    {
      "entropy": 0.9262615777552128,
      "epoch": 0.2520974577489491,
      "grad_norm": 0.16294030845165253,
      "learning_rate": 9.15999631980863e-05,
      "loss": 0.9215,
      "mean_token_accuracy": 0.7320987358689308,
      "num_tokens": 87433072.0,
      "step": 2740
    },
    {
      "entropy": 0.839453112334013,
      "epoch": 0.25218946412039034,
      "grad_norm": 0.16228698194026947,
      "learning_rate": 9.159689637194467e-05,
      "loss": 0.8498,
      "mean_token_accuracy": 0.7522500567138195,
      "num_tokens": 87464906.0,
      "step": 2741
    },
    {
      "entropy": 0.7926503606140614,
      "epoch": 0.2522814704918316,
      "grad_norm": 0.1545933037996292,
      "learning_rate": 9.159382954580305e-05,
      "loss": 0.7777,
      "mean_token_accuracy": 0.7721719704568386,
      "num_tokens": 87496753.0,
      "step": 2742
    },
    {
      "entropy": 0.9514006860554218,
      "epoch": 0.25237347686327277,
      "grad_norm": 0.16393670439720154,
      "learning_rate": 9.159076271966143e-05,
      "loss": 0.9565,
      "mean_token_accuracy": 0.7313220538198948,
      "num_tokens": 87528785.0,
      "step": 2743
    },
    {
      "entropy": 0.7859730739146471,
      "epoch": 0.252465483234714,
      "grad_norm": 0.1506839543581009,
      "learning_rate": 9.15876958935198e-05,
      "loss": 0.7688,
      "mean_token_accuracy": 0.7761476598680019,
      "num_tokens": 87560514.0,
      "step": 2744
    },
    {
      "entropy": 0.8212261758744717,
      "epoch": 0.2525574896061552,
      "grad_norm": 0.15564662218093872,
      "learning_rate": 9.158462906737817e-05,
      "loss": 0.8056,
      "mean_token_accuracy": 0.76956931874156,
      "num_tokens": 87593246.0,
      "step": 2745
    },
    {
      "entropy": 0.8865234367549419,
      "epoch": 0.25264949597759645,
      "grad_norm": 0.16233384609222412,
      "learning_rate": 9.158156224123655e-05,
      "loss": 0.8805,
      "mean_token_accuracy": 0.7494681067764759,
      "num_tokens": 87624222.0,
      "step": 2746
    },
    {
      "entropy": 0.8251829110085964,
      "epoch": 0.2527415023490377,
      "grad_norm": 0.15811094641685486,
      "learning_rate": 9.157849541509492e-05,
      "loss": 0.8337,
      "mean_token_accuracy": 0.7609645500779152,
      "num_tokens": 87656300.0,
      "step": 2747
    },
    {
      "entropy": 0.8639294356107712,
      "epoch": 0.2528335087204789,
      "grad_norm": 0.16417552530765533,
      "learning_rate": 9.15754285889533e-05,
      "loss": 0.8618,
      "mean_token_accuracy": 0.7520268931984901,
      "num_tokens": 87687892.0,
      "step": 2748
    },
    {
      "entropy": 0.8229252621531487,
      "epoch": 0.25292551509192013,
      "grad_norm": 0.14978885650634766,
      "learning_rate": 9.157236176281168e-05,
      "loss": 0.7912,
      "mean_token_accuracy": 0.770048625767231,
      "num_tokens": 87719551.0,
      "step": 2749
    },
    {
      "entropy": 0.8497893530875444,
      "epoch": 0.2530175214633613,
      "grad_norm": 0.1640884280204773,
      "learning_rate": 9.156929493667005e-05,
      "loss": 0.8434,
      "mean_token_accuracy": 0.7600831910967827,
      "num_tokens": 87751558.0,
      "step": 2750
    },
    {
      "entropy": 0.9011817872524261,
      "epoch": 0.25310952783480256,
      "grad_norm": 0.16266858577728271,
      "learning_rate": 9.156622811052842e-05,
      "loss": 0.8981,
      "mean_token_accuracy": 0.7486536726355553,
      "num_tokens": 87783161.0,
      "step": 2751
    },
    {
      "entropy": 0.8641266711056232,
      "epoch": 0.2532015342062438,
      "grad_norm": 0.15604686737060547,
      "learning_rate": 9.156316128438678e-05,
      "loss": 0.8593,
      "mean_token_accuracy": 0.7518543601036072,
      "num_tokens": 87815298.0,
      "step": 2752
    },
    {
      "entropy": 0.8162213657051325,
      "epoch": 0.253293540577685,
      "grad_norm": 0.1598665565252304,
      "learning_rate": 9.156009445824516e-05,
      "loss": 0.815,
      "mean_token_accuracy": 0.7613694071769714,
      "num_tokens": 87847382.0,
      "step": 2753
    },
    {
      "entropy": 0.7797237131744623,
      "epoch": 0.25338554694912624,
      "grad_norm": 0.15749722719192505,
      "learning_rate": 9.155702763210355e-05,
      "loss": 0.7883,
      "mean_token_accuracy": 0.7670754678547382,
      "num_tokens": 87879387.0,
      "step": 2754
    },
    {
      "entropy": 0.8318130113184452,
      "epoch": 0.25347755332056743,
      "grad_norm": 0.16052283346652985,
      "learning_rate": 9.155396080596191e-05,
      "loss": 0.824,
      "mean_token_accuracy": 0.7646526545286179,
      "num_tokens": 87911608.0,
      "step": 2755
    },
    {
      "entropy": 0.7953128647059202,
      "epoch": 0.2535695596920087,
      "grad_norm": 0.16091987490653992,
      "learning_rate": 9.155089397982028e-05,
      "loss": 0.7995,
      "mean_token_accuracy": 0.7682630345225334,
      "num_tokens": 87943657.0,
      "step": 2756
    },
    {
      "entropy": 0.9314861632883549,
      "epoch": 0.2536615660634499,
      "grad_norm": 0.16475637257099152,
      "learning_rate": 9.154782715367866e-05,
      "loss": 0.9101,
      "mean_token_accuracy": 0.7399981431663036,
      "num_tokens": 87975979.0,
      "step": 2757
    },
    {
      "entropy": 0.850920557975769,
      "epoch": 0.2537535724348911,
      "grad_norm": 0.1563671976327896,
      "learning_rate": 9.154476032753704e-05,
      "loss": 0.8348,
      "mean_token_accuracy": 0.7597026042640209,
      "num_tokens": 88007923.0,
      "step": 2758
    },
    {
      "entropy": 0.8264209982007742,
      "epoch": 0.25384557880633235,
      "grad_norm": 0.15912050008773804,
      "learning_rate": 9.154169350139541e-05,
      "loss": 0.843,
      "mean_token_accuracy": 0.7628272362053394,
      "num_tokens": 88040388.0,
      "step": 2759
    },
    {
      "entropy": 0.8695603534579277,
      "epoch": 0.25393758517777354,
      "grad_norm": 0.1659490317106247,
      "learning_rate": 9.153862667525378e-05,
      "loss": 0.897,
      "mean_token_accuracy": 0.7512540221214294,
      "num_tokens": 88072083.0,
      "step": 2760
    },
    {
      "entropy": 0.8654303438961506,
      "epoch": 0.2540295915492148,
      "grad_norm": 0.16553208231925964,
      "learning_rate": 9.153555984911216e-05,
      "loss": 0.863,
      "mean_token_accuracy": 0.7508539892733097,
      "num_tokens": 88103928.0,
      "step": 2761
    },
    {
      "entropy": 0.8731610588729382,
      "epoch": 0.25412159792065603,
      "grad_norm": 0.15208066999912262,
      "learning_rate": 9.153249302297053e-05,
      "loss": 0.8486,
      "mean_token_accuracy": 0.7523898929357529,
      "num_tokens": 88136088.0,
      "step": 2762
    },
    {
      "entropy": 0.8120050728321075,
      "epoch": 0.2542136042920972,
      "grad_norm": 0.15259264409542084,
      "learning_rate": 9.152942619682891e-05,
      "loss": 0.7935,
      "mean_token_accuracy": 0.7698288634419441,
      "num_tokens": 88168002.0,
      "step": 2763
    },
    {
      "entropy": 0.8315807059407234,
      "epoch": 0.25430561066353846,
      "grad_norm": 0.14969809353351593,
      "learning_rate": 9.152635937068728e-05,
      "loss": 0.8281,
      "mean_token_accuracy": 0.7619085051119328,
      "num_tokens": 88200194.0,
      "step": 2764
    },
    {
      "entropy": 0.9077835064381361,
      "epoch": 0.25439761703497965,
      "grad_norm": 0.16208165884017944,
      "learning_rate": 9.152329254454566e-05,
      "loss": 0.9078,
      "mean_token_accuracy": 0.7441100254654884,
      "num_tokens": 88231902.0,
      "step": 2765
    },
    {
      "entropy": 0.8087678924202919,
      "epoch": 0.2544896234064209,
      "grad_norm": 0.15620604157447815,
      "learning_rate": 9.152022571840403e-05,
      "loss": 0.81,
      "mean_token_accuracy": 0.768579438328743,
      "num_tokens": 88263541.0,
      "step": 2766
    },
    {
      "entropy": 0.923043992370367,
      "epoch": 0.25458162977786214,
      "grad_norm": 0.16424353420734406,
      "learning_rate": 9.15171588922624e-05,
      "loss": 0.9163,
      "mean_token_accuracy": 0.7369749471545219,
      "num_tokens": 88294967.0,
      "step": 2767
    },
    {
      "entropy": 0.7623236998915672,
      "epoch": 0.25467363614930333,
      "grad_norm": 0.1529853641986847,
      "learning_rate": 9.151409206612078e-05,
      "loss": 0.7751,
      "mean_token_accuracy": 0.7781967483460903,
      "num_tokens": 88326766.0,
      "step": 2768
    },
    {
      "entropy": 0.8622432239353657,
      "epoch": 0.2547656425207446,
      "grad_norm": 0.15968473255634308,
      "learning_rate": 9.151102523997916e-05,
      "loss": 0.8355,
      "mean_token_accuracy": 0.7600646130740643,
      "num_tokens": 88357972.0,
      "step": 2769
    },
    {
      "entropy": 0.7820812333375216,
      "epoch": 0.25485764889218576,
      "grad_norm": 0.1589575856924057,
      "learning_rate": 9.150795841383753e-05,
      "loss": 0.7955,
      "mean_token_accuracy": 0.7730214968323708,
      "num_tokens": 88390740.0,
      "step": 2770
    },
    {
      "entropy": 0.9787991680204868,
      "epoch": 0.254949655263627,
      "grad_norm": 0.16957052052021027,
      "learning_rate": 9.15048915876959e-05,
      "loss": 0.9796,
      "mean_token_accuracy": 0.7220709472894669,
      "num_tokens": 88422812.0,
      "step": 2771
    },
    {
      "entropy": 0.9253401085734367,
      "epoch": 0.25504166163506825,
      "grad_norm": 0.16506490111351013,
      "learning_rate": 9.150182476155426e-05,
      "loss": 0.9265,
      "mean_token_accuracy": 0.73729657381773,
      "num_tokens": 88454165.0,
      "step": 2772
    },
    {
      "entropy": 0.8337426073849201,
      "epoch": 0.25513366800650944,
      "grad_norm": 0.15838368237018585,
      "learning_rate": 9.149875793541264e-05,
      "loss": 0.8421,
      "mean_token_accuracy": 0.7561553418636322,
      "num_tokens": 88486238.0,
      "step": 2773
    },
    {
      "entropy": 0.8948820754885674,
      "epoch": 0.2552256743779507,
      "grad_norm": 0.15800051391124725,
      "learning_rate": 9.149569110927103e-05,
      "loss": 0.894,
      "mean_token_accuracy": 0.7466093897819519,
      "num_tokens": 88518588.0,
      "step": 2774
    },
    {
      "entropy": 0.758627712726593,
      "epoch": 0.2553176807493919,
      "grad_norm": 0.14706432819366455,
      "learning_rate": 9.14926242831294e-05,
      "loss": 0.7322,
      "mean_token_accuracy": 0.7816632241010666,
      "num_tokens": 88550394.0,
      "step": 2775
    },
    {
      "entropy": 0.83804015442729,
      "epoch": 0.2554096871208331,
      "grad_norm": 0.1542797088623047,
      "learning_rate": 9.148955745698776e-05,
      "loss": 0.8478,
      "mean_token_accuracy": 0.7611195370554924,
      "num_tokens": 88582582.0,
      "step": 2776
    },
    {
      "entropy": 0.7838310785591602,
      "epoch": 0.25550169349227436,
      "grad_norm": 0.14777468144893646,
      "learning_rate": 9.148649063084614e-05,
      "loss": 0.7616,
      "mean_token_accuracy": 0.7784951776266098,
      "num_tokens": 88615064.0,
      "step": 2777
    },
    {
      "entropy": 0.9210391268134117,
      "epoch": 0.25559369986371555,
      "grad_norm": 0.15871484577655792,
      "learning_rate": 9.148342380470451e-05,
      "loss": 0.9074,
      "mean_token_accuracy": 0.7398480325937271,
      "num_tokens": 88647071.0,
      "step": 2778
    },
    {
      "entropy": 0.8769982177764177,
      "epoch": 0.2556857062351568,
      "grad_norm": 0.1582174450159073,
      "learning_rate": 9.148035697856289e-05,
      "loss": 0.8692,
      "mean_token_accuracy": 0.7533495537936687,
      "num_tokens": 88679568.0,
      "step": 2779
    },
    {
      "entropy": 0.8068108335137367,
      "epoch": 0.255777712606598,
      "grad_norm": 0.15174080431461334,
      "learning_rate": 9.147729015242127e-05,
      "loss": 0.7979,
      "mean_token_accuracy": 0.7714055180549622,
      "num_tokens": 88711732.0,
      "step": 2780
    },
    {
      "entropy": 0.930365540087223,
      "epoch": 0.25586971897803923,
      "grad_norm": 0.16415907442569733,
      "learning_rate": 9.147422332627964e-05,
      "loss": 0.9321,
      "mean_token_accuracy": 0.7386149913072586,
      "num_tokens": 88743955.0,
      "step": 2781
    },
    {
      "entropy": 0.9443697743117809,
      "epoch": 0.2559617253494805,
      "grad_norm": 0.16549700498580933,
      "learning_rate": 9.147115650013801e-05,
      "loss": 0.9469,
      "mean_token_accuracy": 0.732539489865303,
      "num_tokens": 88776081.0,
      "step": 2782
    },
    {
      "entropy": 0.8253733851015568,
      "epoch": 0.25605373172092166,
      "grad_norm": 0.16491864621639252,
      "learning_rate": 9.146808967399638e-05,
      "loss": 0.8502,
      "mean_token_accuracy": 0.7564817517995834,
      "num_tokens": 88807859.0,
      "step": 2783
    },
    {
      "entropy": 0.9824087023735046,
      "epoch": 0.2561457380923629,
      "grad_norm": 0.16532652080059052,
      "learning_rate": 9.146502284785476e-05,
      "loss": 0.9821,
      "mean_token_accuracy": 0.725590780377388,
      "num_tokens": 88840241.0,
      "step": 2784
    },
    {
      "entropy": 0.8922462798655033,
      "epoch": 0.2562377444638041,
      "grad_norm": 0.16553762555122375,
      "learning_rate": 9.146195602171314e-05,
      "loss": 0.8844,
      "mean_token_accuracy": 0.7482584826648235,
      "num_tokens": 88872359.0,
      "step": 2785
    },
    {
      "entropy": 0.8375036381185055,
      "epoch": 0.25632975083524534,
      "grad_norm": 0.15326903760433197,
      "learning_rate": 9.145888919557151e-05,
      "loss": 0.8218,
      "mean_token_accuracy": 0.7603331170976162,
      "num_tokens": 88905045.0,
      "step": 2786
    },
    {
      "entropy": 1.0006334595382214,
      "epoch": 0.2564217572066866,
      "grad_norm": 0.16587746143341064,
      "learning_rate": 9.145582236942988e-05,
      "loss": 0.9981,
      "mean_token_accuracy": 0.7177365086972713,
      "num_tokens": 88937190.0,
      "step": 2787
    },
    {
      "entropy": 0.884265024214983,
      "epoch": 0.2565137635781278,
      "grad_norm": 0.16339056193828583,
      "learning_rate": 9.145275554328826e-05,
      "loss": 0.8784,
      "mean_token_accuracy": 0.7538694143295288,
      "num_tokens": 88969617.0,
      "step": 2788
    },
    {
      "entropy": 0.8126080632209778,
      "epoch": 0.256605769949569,
      "grad_norm": 0.15208764374256134,
      "learning_rate": 9.144968871714662e-05,
      "loss": 0.7898,
      "mean_token_accuracy": 0.7675426974892616,
      "num_tokens": 89001593.0,
      "step": 2789
    },
    {
      "entropy": 0.8458507135510445,
      "epoch": 0.2566977763210102,
      "grad_norm": 0.1636621057987213,
      "learning_rate": 9.1446621891005e-05,
      "loss": 0.8516,
      "mean_token_accuracy": 0.7577724903821945,
      "num_tokens": 89033060.0,
      "step": 2790
    },
    {
      "entropy": 0.9247744046151638,
      "epoch": 0.25678978269245145,
      "grad_norm": 0.17925265431404114,
      "learning_rate": 9.144355506486337e-05,
      "loss": 0.9719,
      "mean_token_accuracy": 0.7280360944569111,
      "num_tokens": 89065283.0,
      "step": 2791
    },
    {
      "entropy": 0.7767251897603273,
      "epoch": 0.2568817890638927,
      "grad_norm": 0.1604960709810257,
      "learning_rate": 9.144048823872176e-05,
      "loss": 0.7849,
      "mean_token_accuracy": 0.7776144929230213,
      "num_tokens": 89096817.0,
      "step": 2792
    },
    {
      "entropy": 0.8979231901466846,
      "epoch": 0.2569737954353339,
      "grad_norm": 0.16305844485759735,
      "learning_rate": 9.143742141258012e-05,
      "loss": 0.8813,
      "mean_token_accuracy": 0.7517409287393093,
      "num_tokens": 89128922.0,
      "step": 2793
    },
    {
      "entropy": 0.8591989316046238,
      "epoch": 0.25706580180677513,
      "grad_norm": 0.17030738294124603,
      "learning_rate": 9.143435458643849e-05,
      "loss": 0.8765,
      "mean_token_accuracy": 0.7542348392307758,
      "num_tokens": 89161269.0,
      "step": 2794
    },
    {
      "entropy": 0.9160422347486019,
      "epoch": 0.2571578081782163,
      "grad_norm": 0.16399841010570526,
      "learning_rate": 9.143128776029687e-05,
      "loss": 0.9256,
      "mean_token_accuracy": 0.7359187453985214,
      "num_tokens": 89192337.0,
      "step": 2795
    },
    {
      "entropy": 0.8598131239414215,
      "epoch": 0.25724981454965756,
      "grad_norm": 0.1574128121137619,
      "learning_rate": 9.142822093415525e-05,
      "loss": 0.8301,
      "mean_token_accuracy": 0.7646610140800476,
      "num_tokens": 89224314.0,
      "step": 2796
    },
    {
      "entropy": 0.8575816638767719,
      "epoch": 0.2573418209210988,
      "grad_norm": 0.15365314483642578,
      "learning_rate": 9.142515410801362e-05,
      "loss": 0.828,
      "mean_token_accuracy": 0.7632404938340187,
      "num_tokens": 89256333.0,
      "step": 2797
    },
    {
      "entropy": 0.8619565069675446,
      "epoch": 0.25743382729254,
      "grad_norm": 0.15385395288467407,
      "learning_rate": 9.142208728187199e-05,
      "loss": 0.8528,
      "mean_token_accuracy": 0.7576416060328484,
      "num_tokens": 89288952.0,
      "step": 2798
    },
    {
      "entropy": 0.8496936745941639,
      "epoch": 0.25752583366398124,
      "grad_norm": 0.1593405306339264,
      "learning_rate": 9.141902045573036e-05,
      "loss": 0.8401,
      "mean_token_accuracy": 0.7569091878831387,
      "num_tokens": 89320999.0,
      "step": 2799
    },
    {
      "entropy": 0.7744437362998724,
      "epoch": 0.25761784003542243,
      "grad_norm": 0.15686284005641937,
      "learning_rate": 9.141595362958875e-05,
      "loss": 0.7677,
      "mean_token_accuracy": 0.7791652791202068,
      "num_tokens": 89353757.0,
      "step": 2800
    },
    {
      "entropy": 0.8682740293443203,
      "epoch": 0.2577098464068637,
      "grad_norm": 0.1574684977531433,
      "learning_rate": 9.141288680344712e-05,
      "loss": 0.8647,
      "mean_token_accuracy": 0.7575385347008705,
      "num_tokens": 89385455.0,
      "step": 2801
    },
    {
      "entropy": 0.8405726402997971,
      "epoch": 0.2578018527783049,
      "grad_norm": 0.16005155444145203,
      "learning_rate": 9.140981997730549e-05,
      "loss": 0.8504,
      "mean_token_accuracy": 0.7543057426810265,
      "num_tokens": 89416983.0,
      "step": 2802
    },
    {
      "entropy": 0.8528567012399435,
      "epoch": 0.2578938591497461,
      "grad_norm": 0.165125772356987,
      "learning_rate": 9.140675315116386e-05,
      "loss": 0.8644,
      "mean_token_accuracy": 0.7572858333587646,
      "num_tokens": 89448558.0,
      "step": 2803
    },
    {
      "entropy": 0.8738709706813097,
      "epoch": 0.25798586552118735,
      "grad_norm": 0.16953903436660767,
      "learning_rate": 9.140368632502224e-05,
      "loss": 0.8956,
      "mean_token_accuracy": 0.7446806617081165,
      "num_tokens": 89480412.0,
      "step": 2804
    },
    {
      "entropy": 0.9119055541232228,
      "epoch": 0.25807787189262854,
      "grad_norm": 0.16405169665813446,
      "learning_rate": 9.140061949888062e-05,
      "loss": 0.9166,
      "mean_token_accuracy": 0.7436589784920216,
      "num_tokens": 89512419.0,
      "step": 2805
    },
    {
      "entropy": 0.827830420807004,
      "epoch": 0.2581698782640698,
      "grad_norm": 0.15960438549518585,
      "learning_rate": 9.139755267273899e-05,
      "loss": 0.8264,
      "mean_token_accuracy": 0.7627162449061871,
      "num_tokens": 89544378.0,
      "step": 2806
    },
    {
      "entropy": 0.875212762504816,
      "epoch": 0.25826188463551103,
      "grad_norm": 0.16282302141189575,
      "learning_rate": 9.139448584659735e-05,
      "loss": 0.8772,
      "mean_token_accuracy": 0.7505728341639042,
      "num_tokens": 89575776.0,
      "step": 2807
    },
    {
      "entropy": 0.8670352809131145,
      "epoch": 0.2583538910069522,
      "grad_norm": 0.15919779241085052,
      "learning_rate": 9.139141902045574e-05,
      "loss": 0.8541,
      "mean_token_accuracy": 0.7526722960174084,
      "num_tokens": 89607561.0,
      "step": 2808
    },
    {
      "entropy": 0.8465863354504108,
      "epoch": 0.25844589737839346,
      "grad_norm": 0.1605643332004547,
      "learning_rate": 9.13883521943141e-05,
      "loss": 0.8453,
      "mean_token_accuracy": 0.7550167255103588,
      "num_tokens": 89638118.0,
      "step": 2809
    },
    {
      "entropy": 0.8307596147060394,
      "epoch": 0.25853790374983465,
      "grad_norm": 0.1578579545021057,
      "learning_rate": 9.138528536817249e-05,
      "loss": 0.8374,
      "mean_token_accuracy": 0.7645382210612297,
      "num_tokens": 89669260.0,
      "step": 2810
    },
    {
      "entropy": 0.8379404544830322,
      "epoch": 0.2586299101212759,
      "grad_norm": 0.15032045543193817,
      "learning_rate": 9.138221854203087e-05,
      "loss": 0.8149,
      "mean_token_accuracy": 0.764054823666811,
      "num_tokens": 89701647.0,
      "step": 2811
    },
    {
      "entropy": 0.8520856983959675,
      "epoch": 0.25872191649271714,
      "grad_norm": 0.1553380787372589,
      "learning_rate": 9.137915171588923e-05,
      "loss": 0.841,
      "mean_token_accuracy": 0.7565230652689934,
      "num_tokens": 89734393.0,
      "step": 2812
    },
    {
      "entropy": 0.825719028711319,
      "epoch": 0.25881392286415833,
      "grad_norm": 0.15387941896915436,
      "learning_rate": 9.13760848897476e-05,
      "loss": 0.813,
      "mean_token_accuracy": 0.7640046775341034,
      "num_tokens": 89766787.0,
      "step": 2813
    },
    {
      "entropy": 0.9063511025160551,
      "epoch": 0.2589059292355996,
      "grad_norm": 0.16124024987220764,
      "learning_rate": 9.137301806360597e-05,
      "loss": 0.9101,
      "mean_token_accuracy": 0.7403098084032536,
      "num_tokens": 89798988.0,
      "step": 2814
    },
    {
      "entropy": 0.8480651006102562,
      "epoch": 0.25899793560704076,
      "grad_norm": 0.155999094247818,
      "learning_rate": 9.136995123746435e-05,
      "loss": 0.8503,
      "mean_token_accuracy": 0.7546104043722153,
      "num_tokens": 89831756.0,
      "step": 2815
    },
    {
      "entropy": 0.9275361075997353,
      "epoch": 0.259089941978482,
      "grad_norm": 0.15827515721321106,
      "learning_rate": 9.136688441132273e-05,
      "loss": 0.9038,
      "mean_token_accuracy": 0.7390910945832729,
      "num_tokens": 89863952.0,
      "step": 2816
    },
    {
      "entropy": 0.7832501847296953,
      "epoch": 0.25918194834992325,
      "grad_norm": 0.14867666363716125,
      "learning_rate": 9.13638175851811e-05,
      "loss": 0.7451,
      "mean_token_accuracy": 0.7844700329005718,
      "num_tokens": 89895939.0,
      "step": 2817
    },
    {
      "entropy": 0.8502877876162529,
      "epoch": 0.25927395472136444,
      "grad_norm": 0.16363820433616638,
      "learning_rate": 9.136075075903947e-05,
      "loss": 0.8552,
      "mean_token_accuracy": 0.7544577419757843,
      "num_tokens": 89928707.0,
      "step": 2818
    },
    {
      "entropy": 0.9521913155913353,
      "epoch": 0.2593659610928057,
      "grad_norm": 0.1613747626543045,
      "learning_rate": 9.135768393289785e-05,
      "loss": 0.9529,
      "mean_token_accuracy": 0.7349535524845123,
      "num_tokens": 89961093.0,
      "step": 2819
    },
    {
      "entropy": 1.0222590006887913,
      "epoch": 0.2594579674642469,
      "grad_norm": 0.1705067902803421,
      "learning_rate": 9.135461710675622e-05,
      "loss": 1.0366,
      "mean_token_accuracy": 0.7106485925614834,
      "num_tokens": 89993677.0,
      "step": 2820
    },
    {
      "entropy": 0.8609245866537094,
      "epoch": 0.2595499738356881,
      "grad_norm": 0.16584503650665283,
      "learning_rate": 9.13515502806146e-05,
      "loss": 0.8708,
      "mean_token_accuracy": 0.7524748183786869,
      "num_tokens": 90024883.0,
      "step": 2821
    },
    {
      "entropy": 0.7869315836578608,
      "epoch": 0.25964198020712936,
      "grad_norm": 0.15979622304439545,
      "learning_rate": 9.134848345447297e-05,
      "loss": 0.792,
      "mean_token_accuracy": 0.7729918174445629,
      "num_tokens": 90056717.0,
      "step": 2822
    },
    {
      "entropy": 0.8447292745113373,
      "epoch": 0.25973398657857055,
      "grad_norm": 0.16518838703632355,
      "learning_rate": 9.134541662833135e-05,
      "loss": 0.8249,
      "mean_token_accuracy": 0.7591153681278229,
      "num_tokens": 90087681.0,
      "step": 2823
    },
    {
      "entropy": 0.8453956712037325,
      "epoch": 0.2598259929500118,
      "grad_norm": 0.17094330489635468,
      "learning_rate": 9.134234980218972e-05,
      "loss": 0.8656,
      "mean_token_accuracy": 0.7600379884243011,
      "num_tokens": 90119155.0,
      "step": 2824
    },
    {
      "entropy": 0.9652759544551373,
      "epoch": 0.259917999321453,
      "grad_norm": 0.169732466340065,
      "learning_rate": 9.133928297604809e-05,
      "loss": 0.9845,
      "mean_token_accuracy": 0.7246212437748909,
      "num_tokens": 90150489.0,
      "step": 2825
    },
    {
      "entropy": 0.7368095740675926,
      "epoch": 0.26001000569289423,
      "grad_norm": 0.16340185701847076,
      "learning_rate": 9.133621614990647e-05,
      "loss": 0.7579,
      "mean_token_accuracy": 0.7799007222056389,
      "num_tokens": 90183220.0,
      "step": 2826
    },
    {
      "entropy": 0.8660338763147593,
      "epoch": 0.2601020120643355,
      "grad_norm": 0.1599806249141693,
      "learning_rate": 9.133314932376485e-05,
      "loss": 0.8748,
      "mean_token_accuracy": 0.7478078491985798,
      "num_tokens": 90214376.0,
      "step": 2827
    },
    {
      "entropy": 0.881524994969368,
      "epoch": 0.26019401843577666,
      "grad_norm": 0.1620558202266693,
      "learning_rate": 9.133008249762322e-05,
      "loss": 0.885,
      "mean_token_accuracy": 0.7476818300783634,
      "num_tokens": 90247006.0,
      "step": 2828
    },
    {
      "entropy": 0.9129325225949287,
      "epoch": 0.2602860248072179,
      "grad_norm": 0.16515423357486725,
      "learning_rate": 9.132701567148158e-05,
      "loss": 0.9333,
      "mean_token_accuracy": 0.738802719861269,
      "num_tokens": 90279261.0,
      "step": 2829
    },
    {
      "entropy": 0.8334056921303272,
      "epoch": 0.2603780311786591,
      "grad_norm": 0.15752603113651276,
      "learning_rate": 9.132394884533995e-05,
      "loss": 0.827,
      "mean_token_accuracy": 0.7596079520881176,
      "num_tokens": 90311271.0,
      "step": 2830
    },
    {
      "entropy": 0.8350309766829014,
      "epoch": 0.26047003755010034,
      "grad_norm": 0.149174764752388,
      "learning_rate": 9.132088201919833e-05,
      "loss": 0.8027,
      "mean_token_accuracy": 0.769335750490427,
      "num_tokens": 90343648.0,
      "step": 2831
    },
    {
      "entropy": 0.8916435390710831,
      "epoch": 0.2605620439215416,
      "grad_norm": 0.15926620364189148,
      "learning_rate": 9.131781519305671e-05,
      "loss": 0.8466,
      "mean_token_accuracy": 0.7529180906713009,
      "num_tokens": 90375001.0,
      "step": 2832
    },
    {
      "entropy": 0.8382735662162304,
      "epoch": 0.2606540502929828,
      "grad_norm": 0.15188710391521454,
      "learning_rate": 9.131474836691508e-05,
      "loss": 0.8116,
      "mean_token_accuracy": 0.7645600251853466,
      "num_tokens": 90407683.0,
      "step": 2833
    },
    {
      "entropy": 1.0099739655852318,
      "epoch": 0.260746056664424,
      "grad_norm": 0.161537304520607,
      "learning_rate": 9.131168154077345e-05,
      "loss": 0.9965,
      "mean_token_accuracy": 0.7219971977174282,
      "num_tokens": 90439635.0,
      "step": 2834
    },
    {
      "entropy": 0.8736461102962494,
      "epoch": 0.2608380630358652,
      "grad_norm": 0.15905554592609406,
      "learning_rate": 9.130861471463183e-05,
      "loss": 0.86,
      "mean_token_accuracy": 0.7550927214324474,
      "num_tokens": 90470940.0,
      "step": 2835
    },
    {
      "entropy": 0.8264850005507469,
      "epoch": 0.26093006940730645,
      "grad_norm": 0.15622183680534363,
      "learning_rate": 9.13055478884902e-05,
      "loss": 0.828,
      "mean_token_accuracy": 0.7633023373782635,
      "num_tokens": 90502972.0,
      "step": 2836
    },
    {
      "entropy": 0.889338593930006,
      "epoch": 0.2610220757787477,
      "grad_norm": 0.16298504173755646,
      "learning_rate": 9.130248106234858e-05,
      "loss": 0.9079,
      "mean_token_accuracy": 0.745250154286623,
      "num_tokens": 90535197.0,
      "step": 2837
    },
    {
      "entropy": 0.8423227425664663,
      "epoch": 0.2611140821501889,
      "grad_norm": 0.16186237335205078,
      "learning_rate": 9.129941423620696e-05,
      "loss": 0.8446,
      "mean_token_accuracy": 0.7541607320308685,
      "num_tokens": 90567250.0,
      "step": 2838
    },
    {
      "entropy": 0.771701704710722,
      "epoch": 0.26120608852163013,
      "grad_norm": 0.154720738530159,
      "learning_rate": 9.129634741006533e-05,
      "loss": 0.7702,
      "mean_token_accuracy": 0.7786836922168732,
      "num_tokens": 90599231.0,
      "step": 2839
    },
    {
      "entropy": 0.8576270490884781,
      "epoch": 0.2612980948930713,
      "grad_norm": 0.15544119477272034,
      "learning_rate": 9.12932805839237e-05,
      "loss": 0.841,
      "mean_token_accuracy": 0.7576870173215866,
      "num_tokens": 90631456.0,
      "step": 2840
    },
    {
      "entropy": 0.8823127709329128,
      "epoch": 0.26139010126451256,
      "grad_norm": 0.1564403921365738,
      "learning_rate": 9.129021375778207e-05,
      "loss": 0.8673,
      "mean_token_accuracy": 0.7499171607196331,
      "num_tokens": 90663969.0,
      "step": 2841
    },
    {
      "entropy": 0.8288220949470997,
      "epoch": 0.2614821076359538,
      "grad_norm": 0.15910476446151733,
      "learning_rate": 9.128714693164046e-05,
      "loss": 0.8333,
      "mean_token_accuracy": 0.7599842473864555,
      "num_tokens": 90695703.0,
      "step": 2842
    },
    {
      "entropy": 0.8123975992202759,
      "epoch": 0.261574114007395,
      "grad_norm": 0.16509585082530975,
      "learning_rate": 9.128408010549883e-05,
      "loss": 0.8283,
      "mean_token_accuracy": 0.7604980543255806,
      "num_tokens": 90727904.0,
      "step": 2843
    },
    {
      "entropy": 0.7716659530997276,
      "epoch": 0.26166612037883624,
      "grad_norm": 0.16245560348033905,
      "learning_rate": 9.12810132793572e-05,
      "loss": 0.7987,
      "mean_token_accuracy": 0.7716722004115582,
      "num_tokens": 90759828.0,
      "step": 2844
    },
    {
      "entropy": 0.9030359536409378,
      "epoch": 0.26175812675027743,
      "grad_norm": 0.16291511058807373,
      "learning_rate": 9.127794645321556e-05,
      "loss": 0.9219,
      "mean_token_accuracy": 0.7368749380111694,
      "num_tokens": 90791883.0,
      "step": 2845
    },
    {
      "entropy": 0.8616113364696503,
      "epoch": 0.2618501331217187,
      "grad_norm": 0.15755385160446167,
      "learning_rate": 9.127487962707395e-05,
      "loss": 0.8494,
      "mean_token_accuracy": 0.7552045248448849,
      "num_tokens": 90823497.0,
      "step": 2846
    },
    {
      "entropy": 0.7095162216573954,
      "epoch": 0.2619421394931599,
      "grad_norm": 0.15499377250671387,
      "learning_rate": 9.127181280093233e-05,
      "loss": 0.7135,
      "mean_token_accuracy": 0.7911393120884895,
      "num_tokens": 90854812.0,
      "step": 2847
    },
    {
      "entropy": 0.8810529857873917,
      "epoch": 0.2620341458646011,
      "grad_norm": 0.1614808291196823,
      "learning_rate": 9.12687459747907e-05,
      "loss": 0.8451,
      "mean_token_accuracy": 0.7525723315775394,
      "num_tokens": 90885927.0,
      "step": 2848
    },
    {
      "entropy": 0.9191434979438782,
      "epoch": 0.26212615223604235,
      "grad_norm": 0.16453303396701813,
      "learning_rate": 9.126567914864906e-05,
      "loss": 0.9383,
      "mean_token_accuracy": 0.7399155758321285,
      "num_tokens": 90918129.0,
      "step": 2849
    },
    {
      "entropy": 0.8701786287128925,
      "epoch": 0.26221815860748354,
      "grad_norm": 0.16060461103916168,
      "learning_rate": 9.126261232250744e-05,
      "loss": 0.8793,
      "mean_token_accuracy": 0.7525698356330395,
      "num_tokens": 90950224.0,
      "step": 2850
    },
    {
      "entropy": 0.8752622045576572,
      "epoch": 0.2623101649789248,
      "grad_norm": 0.15830542147159576,
      "learning_rate": 9.125954549636581e-05,
      "loss": 0.8575,
      "mean_token_accuracy": 0.7541468217968941,
      "num_tokens": 90982413.0,
      "step": 2851
    },
    {
      "entropy": 0.8845952488481998,
      "epoch": 0.26240217135036603,
      "grad_norm": 0.16241692006587982,
      "learning_rate": 9.12564786702242e-05,
      "loss": 0.885,
      "mean_token_accuracy": 0.7453330457210541,
      "num_tokens": 91014340.0,
      "step": 2852
    },
    {
      "entropy": 0.9421564135700464,
      "epoch": 0.2624941777218072,
      "grad_norm": 0.15729211270809174,
      "learning_rate": 9.125341184408256e-05,
      "loss": 0.9332,
      "mean_token_accuracy": 0.7361602894961834,
      "num_tokens": 91046826.0,
      "step": 2853
    },
    {
      "entropy": 0.7698660269379616,
      "epoch": 0.26258618409324846,
      "grad_norm": 0.15412960946559906,
      "learning_rate": 9.125034501794094e-05,
      "loss": 0.768,
      "mean_token_accuracy": 0.7758978568017483,
      "num_tokens": 91078643.0,
      "step": 2854
    },
    {
      "entropy": 0.8216989003121853,
      "epoch": 0.26267819046468965,
      "grad_norm": 0.15156646072864532,
      "learning_rate": 9.124727819179931e-05,
      "loss": 0.805,
      "mean_token_accuracy": 0.7688013762235641,
      "num_tokens": 91110842.0,
      "step": 2855
    },
    {
      "entropy": 0.6979290321469307,
      "epoch": 0.2627701968361309,
      "grad_norm": 0.15467458963394165,
      "learning_rate": 9.124421136565768e-05,
      "loss": 0.7106,
      "mean_token_accuracy": 0.7901562228798866,
      "num_tokens": 91142585.0,
      "step": 2856
    },
    {
      "entropy": 0.9317745268344879,
      "epoch": 0.26286220320757214,
      "grad_norm": 0.16127899289131165,
      "learning_rate": 9.124114453951606e-05,
      "loss": 0.935,
      "mean_token_accuracy": 0.7366209700703621,
      "num_tokens": 91174741.0,
      "step": 2857
    },
    {
      "entropy": 0.8280125856399536,
      "epoch": 0.26295420957901333,
      "grad_norm": 0.1560603827238083,
      "learning_rate": 9.123807771337444e-05,
      "loss": 0.8045,
      "mean_token_accuracy": 0.766029242426157,
      "num_tokens": 91206888.0,
      "step": 2858
    },
    {
      "entropy": 0.8049558624625206,
      "epoch": 0.2630462159504546,
      "grad_norm": 0.15992172062397003,
      "learning_rate": 9.123501088723281e-05,
      "loss": 0.8128,
      "mean_token_accuracy": 0.7615224979817867,
      "num_tokens": 91238435.0,
      "step": 2859
    },
    {
      "entropy": 0.8293513804674149,
      "epoch": 0.26313822232189576,
      "grad_norm": 0.15440335869789124,
      "learning_rate": 9.123194406109118e-05,
      "loss": 0.8158,
      "mean_token_accuracy": 0.7605469450354576,
      "num_tokens": 91270659.0,
      "step": 2860
    },
    {
      "entropy": 0.9127984158694744,
      "epoch": 0.263230228693337,
      "grad_norm": 0.1617114096879959,
      "learning_rate": 9.122887723494955e-05,
      "loss": 0.9231,
      "mean_token_accuracy": 0.7382904104888439,
      "num_tokens": 91302414.0,
      "step": 2861
    },
    {
      "entropy": 0.8535359762609005,
      "epoch": 0.26332223506477825,
      "grad_norm": 0.15709994733333588,
      "learning_rate": 9.122581040880793e-05,
      "loss": 0.8376,
      "mean_token_accuracy": 0.7590136080980301,
      "num_tokens": 91333701.0,
      "step": 2862
    },
    {
      "entropy": 0.7941982317715883,
      "epoch": 0.26341424143621944,
      "grad_norm": 0.14809665083885193,
      "learning_rate": 9.122274358266631e-05,
      "loss": 0.7822,
      "mean_token_accuracy": 0.7742205001413822,
      "num_tokens": 91366092.0,
      "step": 2863
    },
    {
      "entropy": 0.8632979542016983,
      "epoch": 0.2635062478076607,
      "grad_norm": 0.15998363494873047,
      "learning_rate": 9.121967675652468e-05,
      "loss": 0.8276,
      "mean_token_accuracy": 0.7572339400649071,
      "num_tokens": 91396767.0,
      "step": 2864
    },
    {
      "entropy": 0.824711337685585,
      "epoch": 0.2635982541791019,
      "grad_norm": 0.15612611174583435,
      "learning_rate": 9.121660993038304e-05,
      "loss": 0.8021,
      "mean_token_accuracy": 0.7675869278609753,
      "num_tokens": 91428414.0,
      "step": 2865
    },
    {
      "entropy": 0.7608620505779982,
      "epoch": 0.2636902605505431,
      "grad_norm": 0.15361621975898743,
      "learning_rate": 9.121354310424143e-05,
      "loss": 0.7362,
      "mean_token_accuracy": 0.7831396125257015,
      "num_tokens": 91459325.0,
      "step": 2866
    },
    {
      "entropy": 0.7393491100519896,
      "epoch": 0.26378226692198437,
      "grad_norm": 0.15768975019454956,
      "learning_rate": 9.121047627809979e-05,
      "loss": 0.7437,
      "mean_token_accuracy": 0.7856949120759964,
      "num_tokens": 91491218.0,
      "step": 2867
    },
    {
      "entropy": 0.8564013615250587,
      "epoch": 0.26387427329342555,
      "grad_norm": 0.1622854620218277,
      "learning_rate": 9.120740945195817e-05,
      "loss": 0.8606,
      "mean_token_accuracy": 0.7504638582468033,
      "num_tokens": 91522857.0,
      "step": 2868
    },
    {
      "entropy": 0.7640153151005507,
      "epoch": 0.2639662796648668,
      "grad_norm": 0.1526128351688385,
      "learning_rate": 9.120434262581656e-05,
      "loss": 0.7614,
      "mean_token_accuracy": 0.7821603566408157,
      "num_tokens": 91555225.0,
      "step": 2869
    },
    {
      "entropy": 0.8451892659068108,
      "epoch": 0.264058286036308,
      "grad_norm": 0.15567843616008759,
      "learning_rate": 9.120127579967492e-05,
      "loss": 0.8434,
      "mean_token_accuracy": 0.7570932656526566,
      "num_tokens": 91587585.0,
      "step": 2870
    },
    {
      "entropy": 0.8831662386655807,
      "epoch": 0.26415029240774923,
      "grad_norm": 0.16364054381847382,
      "learning_rate": 9.119820897353329e-05,
      "loss": 0.8989,
      "mean_token_accuracy": 0.7451615929603577,
      "num_tokens": 91618649.0,
      "step": 2871
    },
    {
      "entropy": 0.8390093110501766,
      "epoch": 0.2642422987791905,
      "grad_norm": 0.16145792603492737,
      "learning_rate": 9.119514214739166e-05,
      "loss": 0.8523,
      "mean_token_accuracy": 0.7550773471593857,
      "num_tokens": 91650349.0,
      "step": 2872
    },
    {
      "entropy": 0.785961028188467,
      "epoch": 0.26433430515063167,
      "grad_norm": 0.15803658962249756,
      "learning_rate": 9.119207532125005e-05,
      "loss": 0.7759,
      "mean_token_accuracy": 0.774452917277813,
      "num_tokens": 91681555.0,
      "step": 2873
    },
    {
      "entropy": 0.7837657332420349,
      "epoch": 0.2644263115220729,
      "grad_norm": 0.15829838812351227,
      "learning_rate": 9.118900849510842e-05,
      "loss": 0.7933,
      "mean_token_accuracy": 0.7660886906087399,
      "num_tokens": 91713576.0,
      "step": 2874
    },
    {
      "entropy": 0.7725546509027481,
      "epoch": 0.2645183178935141,
      "grad_norm": 0.15627934038639069,
      "learning_rate": 9.118594166896679e-05,
      "loss": 0.7782,
      "mean_token_accuracy": 0.77275275811553,
      "num_tokens": 91744707.0,
      "step": 2875
    },
    {
      "entropy": 0.7898869682103395,
      "epoch": 0.26461032426495534,
      "grad_norm": 0.1603213995695114,
      "learning_rate": 9.118287484282516e-05,
      "loss": 0.7803,
      "mean_token_accuracy": 0.7797432765364647,
      "num_tokens": 91776712.0,
      "step": 2876
    },
    {
      "entropy": 0.9071152396500111,
      "epoch": 0.2647023306363966,
      "grad_norm": 0.15741148591041565,
      "learning_rate": 9.117980801668354e-05,
      "loss": 0.8936,
      "mean_token_accuracy": 0.7469187155365944,
      "num_tokens": 91808131.0,
      "step": 2877
    },
    {
      "entropy": 0.8437071479856968,
      "epoch": 0.2647943370078378,
      "grad_norm": 0.15946362912654877,
      "learning_rate": 9.117674119054192e-05,
      "loss": 0.8342,
      "mean_token_accuracy": 0.7586522698402405,
      "num_tokens": 91840340.0,
      "step": 2878
    },
    {
      "entropy": 0.7043415140360594,
      "epoch": 0.264886343379279,
      "grad_norm": 0.1533012092113495,
      "learning_rate": 9.117367436440029e-05,
      "loss": 0.697,
      "mean_token_accuracy": 0.7923924773931503,
      "num_tokens": 91871891.0,
      "step": 2879
    },
    {
      "entropy": 0.9167690686881542,
      "epoch": 0.2649783497507202,
      "grad_norm": 0.1651921272277832,
      "learning_rate": 9.117060753825866e-05,
      "loss": 0.9198,
      "mean_token_accuracy": 0.7363271266222,
      "num_tokens": 91903117.0,
      "step": 2880
    },
    {
      "entropy": 0.8919342905282974,
      "epoch": 0.26507035612216145,
      "grad_norm": 0.17178937792778015,
      "learning_rate": 9.116754071211704e-05,
      "loss": 0.8956,
      "mean_token_accuracy": 0.7447965256869793,
      "num_tokens": 91934716.0,
      "step": 2881
    },
    {
      "entropy": 0.8044457715004683,
      "epoch": 0.2651623624936027,
      "grad_norm": 0.16217562556266785,
      "learning_rate": 9.11644738859754e-05,
      "loss": 0.7923,
      "mean_token_accuracy": 0.7724172174930573,
      "num_tokens": 91966084.0,
      "step": 2882
    },
    {
      "entropy": 0.8430944606661797,
      "epoch": 0.2652543688650439,
      "grad_norm": 0.16241076588630676,
      "learning_rate": 9.116140705983379e-05,
      "loss": 0.8501,
      "mean_token_accuracy": 0.7550608702003956,
      "num_tokens": 91998704.0,
      "step": 2883
    },
    {
      "entropy": 0.9000072609633207,
      "epoch": 0.26534637523648513,
      "grad_norm": 0.16053444147109985,
      "learning_rate": 9.115834023369216e-05,
      "loss": 0.8824,
      "mean_token_accuracy": 0.750158466398716,
      "num_tokens": 92030954.0,
      "step": 2884
    },
    {
      "entropy": 0.823117857798934,
      "epoch": 0.2654383816079263,
      "grad_norm": 0.16174916923046112,
      "learning_rate": 9.115527340755054e-05,
      "loss": 0.8031,
      "mean_token_accuracy": 0.7711536437273026,
      "num_tokens": 92063026.0,
      "step": 2885
    },
    {
      "entropy": 0.8553418666124344,
      "epoch": 0.26553038797936757,
      "grad_norm": 0.16697604954242706,
      "learning_rate": 9.11522065814089e-05,
      "loss": 0.8709,
      "mean_token_accuracy": 0.7581527307629585,
      "num_tokens": 92093889.0,
      "step": 2886
    },
    {
      "entropy": 0.7681650277227163,
      "epoch": 0.2656223943508088,
      "grad_norm": 0.15844261646270752,
      "learning_rate": 9.114913975526727e-05,
      "loss": 0.7636,
      "mean_token_accuracy": 0.7791492380201817,
      "num_tokens": 92124839.0,
      "step": 2887
    },
    {
      "entropy": 0.8991984277963638,
      "epoch": 0.26571440072225,
      "grad_norm": 0.15928877890110016,
      "learning_rate": 9.114607292912565e-05,
      "loss": 0.8806,
      "mean_token_accuracy": 0.7502195201814175,
      "num_tokens": 92156031.0,
      "step": 2888
    },
    {
      "entropy": 0.8594846297055483,
      "epoch": 0.26580640709369124,
      "grad_norm": 0.16337630152702332,
      "learning_rate": 9.114300610298404e-05,
      "loss": 0.8592,
      "mean_token_accuracy": 0.7551096454262733,
      "num_tokens": 92187991.0,
      "step": 2889
    },
    {
      "entropy": 0.7958955056965351,
      "epoch": 0.26589841346513243,
      "grad_norm": 0.15973924100399017,
      "learning_rate": 9.11399392768424e-05,
      "loss": 0.7898,
      "mean_token_accuracy": 0.7729382999241352,
      "num_tokens": 92219315.0,
      "step": 2890
    },
    {
      "entropy": 0.90886639803648,
      "epoch": 0.2659904198365737,
      "grad_norm": 0.17257468402385712,
      "learning_rate": 9.113687245070077e-05,
      "loss": 0.9397,
      "mean_token_accuracy": 0.7339711040258408,
      "num_tokens": 92250846.0,
      "step": 2891
    },
    {
      "entropy": 0.8154808077961206,
      "epoch": 0.2660824262080149,
      "grad_norm": 0.15756358206272125,
      "learning_rate": 9.113380562455914e-05,
      "loss": 0.8086,
      "mean_token_accuracy": 0.7690202631056309,
      "num_tokens": 92282916.0,
      "step": 2892
    },
    {
      "entropy": 0.8200621511787176,
      "epoch": 0.2661744325794561,
      "grad_norm": 0.1587061583995819,
      "learning_rate": 9.113073879841752e-05,
      "loss": 0.8145,
      "mean_token_accuracy": 0.7684450931847095,
      "num_tokens": 92314731.0,
      "step": 2893
    },
    {
      "entropy": 0.849332869052887,
      "epoch": 0.26626643895089736,
      "grad_norm": 0.15811751782894135,
      "learning_rate": 9.11276719722759e-05,
      "loss": 0.8649,
      "mean_token_accuracy": 0.7503369934856892,
      "num_tokens": 92346809.0,
      "step": 2894
    },
    {
      "entropy": 0.8367020674049854,
      "epoch": 0.2663584453223386,
      "grad_norm": 0.1598115712404251,
      "learning_rate": 9.112460514613427e-05,
      "loss": 0.8535,
      "mean_token_accuracy": 0.7557095736265182,
      "num_tokens": 92379577.0,
      "step": 2895
    },
    {
      "entropy": 0.8189818188548088,
      "epoch": 0.2664504516937798,
      "grad_norm": 0.1513410061597824,
      "learning_rate": 9.112153831999264e-05,
      "loss": 0.7936,
      "mean_token_accuracy": 0.7660753056406975,
      "num_tokens": 92411369.0,
      "step": 2896
    },
    {
      "entropy": 0.8996575586497784,
      "epoch": 0.26654245806522103,
      "grad_norm": 0.15559642016887665,
      "learning_rate": 9.111847149385102e-05,
      "loss": 0.8901,
      "mean_token_accuracy": 0.7500637471675873,
      "num_tokens": 92443987.0,
      "step": 2897
    },
    {
      "entropy": 0.8338337130844593,
      "epoch": 0.2666344644366622,
      "grad_norm": 0.15110206604003906,
      "learning_rate": 9.111540466770939e-05,
      "loss": 0.8041,
      "mean_token_accuracy": 0.7675247602164745,
      "num_tokens": 92475801.0,
      "step": 2898
    },
    {
      "entropy": 0.8955620750784874,
      "epoch": 0.26672647080810347,
      "grad_norm": 0.16235193610191345,
      "learning_rate": 9.111233784156777e-05,
      "loss": 0.9014,
      "mean_token_accuracy": 0.7497307769954205,
      "num_tokens": 92507474.0,
      "step": 2899
    },
    {
      "entropy": 0.7747979369014502,
      "epoch": 0.2668184771795447,
      "grad_norm": 0.15141193568706512,
      "learning_rate": 9.110927101542615e-05,
      "loss": 0.7558,
      "mean_token_accuracy": 0.7786397077143192,
      "num_tokens": 92539698.0,
      "step": 2900
    },
    {
      "entropy": 0.8788194358348846,
      "epoch": 0.2669104835509859,
      "grad_norm": 0.15956534445285797,
      "learning_rate": 9.110620418928452e-05,
      "loss": 0.8697,
      "mean_token_accuracy": 0.7506467700004578,
      "num_tokens": 92571283.0,
      "step": 2901
    },
    {
      "entropy": 0.9128432534635067,
      "epoch": 0.26700248992242714,
      "grad_norm": 0.16231492161750793,
      "learning_rate": 9.110313736314289e-05,
      "loss": 0.9085,
      "mean_token_accuracy": 0.7441803254187107,
      "num_tokens": 92602349.0,
      "step": 2902
    },
    {
      "entropy": 0.8829511888325214,
      "epoch": 0.26709449629386833,
      "grad_norm": 0.15609730780124664,
      "learning_rate": 9.110007053700125e-05,
      "loss": 0.8773,
      "mean_token_accuracy": 0.7453111037611961,
      "num_tokens": 92634058.0,
      "step": 2903
    },
    {
      "entropy": 0.7983574941754341,
      "epoch": 0.2671865026653096,
      "grad_norm": 0.1558784544467926,
      "learning_rate": 9.109700371085963e-05,
      "loss": 0.8079,
      "mean_token_accuracy": 0.7669933512806892,
      "num_tokens": 92665572.0,
      "step": 2904
    },
    {
      "entropy": 0.7800601720809937,
      "epoch": 0.2672785090367508,
      "grad_norm": 0.15805236995220184,
      "learning_rate": 9.109393688471802e-05,
      "loss": 0.7808,
      "mean_token_accuracy": 0.7773460820317268,
      "num_tokens": 92697710.0,
      "step": 2905
    },
    {
      "entropy": 0.8505289610475302,
      "epoch": 0.267370515408192,
      "grad_norm": 0.1594640612602234,
      "learning_rate": 9.109087005857638e-05,
      "loss": 0.8392,
      "mean_token_accuracy": 0.7597367987036705,
      "num_tokens": 92729589.0,
      "step": 2906
    },
    {
      "entropy": 0.8507708162069321,
      "epoch": 0.26746252177963326,
      "grad_norm": 0.16113723814487457,
      "learning_rate": 9.108780323243475e-05,
      "loss": 0.8554,
      "mean_token_accuracy": 0.7566158138215542,
      "num_tokens": 92761875.0,
      "step": 2907
    },
    {
      "entropy": 0.8846883438527584,
      "epoch": 0.26755452815107444,
      "grad_norm": 0.16272105276584625,
      "learning_rate": 9.108473640629313e-05,
      "loss": 0.9047,
      "mean_token_accuracy": 0.743500791490078,
      "num_tokens": 92793426.0,
      "step": 2908
    },
    {
      "entropy": 0.6752659492194653,
      "epoch": 0.2676465345225157,
      "grad_norm": 0.1510891169309616,
      "learning_rate": 9.10816695801515e-05,
      "loss": 0.6671,
      "mean_token_accuracy": 0.8087419047951698,
      "num_tokens": 92825724.0,
      "step": 2909
    },
    {
      "entropy": 0.7795830480754375,
      "epoch": 0.26773854089395693,
      "grad_norm": 0.15684324502944946,
      "learning_rate": 9.107860275400988e-05,
      "loss": 0.7892,
      "mean_token_accuracy": 0.771674145013094,
      "num_tokens": 92857908.0,
      "step": 2910
    },
    {
      "entropy": 0.9761320091784,
      "epoch": 0.2678305472653981,
      "grad_norm": 0.16719605028629303,
      "learning_rate": 9.107553592786825e-05,
      "loss": 0.9619,
      "mean_token_accuracy": 0.7293858714401722,
      "num_tokens": 92889308.0,
      "step": 2911
    },
    {
      "entropy": 0.823435228317976,
      "epoch": 0.26792255363683937,
      "grad_norm": 0.1571776121854782,
      "learning_rate": 9.107246910172663e-05,
      "loss": 0.8294,
      "mean_token_accuracy": 0.7579174488782883,
      "num_tokens": 92920854.0,
      "step": 2912
    },
    {
      "entropy": 0.8957953471690416,
      "epoch": 0.26801456000828056,
      "grad_norm": 0.15853585302829742,
      "learning_rate": 9.1069402275585e-05,
      "loss": 0.8778,
      "mean_token_accuracy": 0.744007583707571,
      "num_tokens": 92952962.0,
      "step": 2913
    },
    {
      "entropy": 0.8084371648728848,
      "epoch": 0.2681065663797218,
      "grad_norm": 0.16267196834087372,
      "learning_rate": 9.106633544944337e-05,
      "loss": 0.8122,
      "mean_token_accuracy": 0.7675983980298042,
      "num_tokens": 92984400.0,
      "step": 2914
    },
    {
      "entropy": 0.9018083699047565,
      "epoch": 0.26819857275116304,
      "grad_norm": 0.16544367372989655,
      "learning_rate": 9.106326862330175e-05,
      "loss": 0.9078,
      "mean_token_accuracy": 0.742401834577322,
      "num_tokens": 93015908.0,
      "step": 2915
    },
    {
      "entropy": 0.879142738878727,
      "epoch": 0.26829057912260423,
      "grad_norm": 0.16020554304122925,
      "learning_rate": 9.106020179716013e-05,
      "loss": 0.8804,
      "mean_token_accuracy": 0.7469329871237278,
      "num_tokens": 93048321.0,
      "step": 2916
    },
    {
      "entropy": 0.859269879758358,
      "epoch": 0.2683825854940455,
      "grad_norm": 0.1559152901172638,
      "learning_rate": 9.10571349710185e-05,
      "loss": 0.8115,
      "mean_token_accuracy": 0.7651441097259521,
      "num_tokens": 93081089.0,
      "step": 2917
    },
    {
      "entropy": 0.8639744613319635,
      "epoch": 0.26847459186548667,
      "grad_norm": 0.1674472540616989,
      "learning_rate": 9.105406814487687e-05,
      "loss": 0.8427,
      "mean_token_accuracy": 0.7580380998551846,
      "num_tokens": 93112451.0,
      "step": 2918
    },
    {
      "entropy": 0.8322537131607533,
      "epoch": 0.2685665982369279,
      "grad_norm": 0.17146269977092743,
      "learning_rate": 9.105100131873523e-05,
      "loss": 0.8722,
      "mean_token_accuracy": 0.7552945762872696,
      "num_tokens": 93143763.0,
      "step": 2919
    },
    {
      "entropy": 0.7947008460760117,
      "epoch": 0.26865860460836916,
      "grad_norm": 0.15613186359405518,
      "learning_rate": 9.104793449259363e-05,
      "loss": 0.7873,
      "mean_token_accuracy": 0.7650721669197083,
      "num_tokens": 93174816.0,
      "step": 2920
    },
    {
      "entropy": 0.8995070029050112,
      "epoch": 0.26875061097981034,
      "grad_norm": 0.1625315099954605,
      "learning_rate": 9.1044867666452e-05,
      "loss": 0.8832,
      "mean_token_accuracy": 0.7479518763720989,
      "num_tokens": 93206158.0,
      "step": 2921
    },
    {
      "entropy": 0.8300111182034016,
      "epoch": 0.2688426173512516,
      "grad_norm": 0.1622701734304428,
      "learning_rate": 9.104180084031036e-05,
      "loss": 0.8244,
      "mean_token_accuracy": 0.7584727369248867,
      "num_tokens": 93238054.0,
      "step": 2922
    },
    {
      "entropy": 0.9338752888143063,
      "epoch": 0.2689346237226928,
      "grad_norm": 0.16942273080348969,
      "learning_rate": 9.103873401416873e-05,
      "loss": 0.9434,
      "mean_token_accuracy": 0.7347329258918762,
      "num_tokens": 93270044.0,
      "step": 2923
    },
    {
      "entropy": 0.8232596069574356,
      "epoch": 0.269026630094134,
      "grad_norm": 0.16149885952472687,
      "learning_rate": 9.103566718802711e-05,
      "loss": 0.8124,
      "mean_token_accuracy": 0.7658885605633259,
      "num_tokens": 93302047.0,
      "step": 2924
    },
    {
      "entropy": 0.754951748996973,
      "epoch": 0.26911863646557527,
      "grad_norm": 0.15347999334335327,
      "learning_rate": 9.10326003618855e-05,
      "loss": 0.7433,
      "mean_token_accuracy": 0.7825439125299454,
      "num_tokens": 93334438.0,
      "step": 2925
    },
    {
      "entropy": 0.7951746806502342,
      "epoch": 0.26921064283701646,
      "grad_norm": 0.15804728865623474,
      "learning_rate": 9.102953353574386e-05,
      "loss": 0.773,
      "mean_token_accuracy": 0.7748000472784042,
      "num_tokens": 93365996.0,
      "step": 2926
    },
    {
      "entropy": 0.8213299568742514,
      "epoch": 0.2693026492084577,
      "grad_norm": 0.15574048459529877,
      "learning_rate": 9.102646670960223e-05,
      "loss": 0.8194,
      "mean_token_accuracy": 0.7599941231310368,
      "num_tokens": 93398185.0,
      "step": 2927
    },
    {
      "entropy": 0.7681305482983589,
      "epoch": 0.2693946555798989,
      "grad_norm": 0.16501381993293762,
      "learning_rate": 9.102339988346061e-05,
      "loss": 0.7837,
      "mean_token_accuracy": 0.7771062403917313,
      "num_tokens": 93430577.0,
      "step": 2928
    },
    {
      "entropy": 0.8234260417521,
      "epoch": 0.26948666195134013,
      "grad_norm": 0.16909779608249664,
      "learning_rate": 9.102033305731898e-05,
      "loss": 0.8409,
      "mean_token_accuracy": 0.7594892494380474,
      "num_tokens": 93462344.0,
      "step": 2929
    },
    {
      "entropy": 0.8062184769660234,
      "epoch": 0.2695786683227814,
      "grad_norm": 0.1601000875234604,
      "learning_rate": 9.101726623117736e-05,
      "loss": 0.8077,
      "mean_token_accuracy": 0.7674770206212997,
      "num_tokens": 93494457.0,
      "step": 2930
    },
    {
      "entropy": 0.8409732431173325,
      "epoch": 0.26967067469422257,
      "grad_norm": 0.16319899260997772,
      "learning_rate": 9.101419940503574e-05,
      "loss": 0.8725,
      "mean_token_accuracy": 0.751792136579752,
      "num_tokens": 93526280.0,
      "step": 2931
    },
    {
      "entropy": 0.9390648119151592,
      "epoch": 0.2697626810656638,
      "grad_norm": 0.17106667160987854,
      "learning_rate": 9.101113257889411e-05,
      "loss": 0.9398,
      "mean_token_accuracy": 0.7320763543248177,
      "num_tokens": 93557434.0,
      "step": 2932
    },
    {
      "entropy": 0.8833139576017857,
      "epoch": 0.269854687437105,
      "grad_norm": 0.17753319442272186,
      "learning_rate": 9.100806575275248e-05,
      "loss": 0.8603,
      "mean_token_accuracy": 0.7535444386303425,
      "num_tokens": 93589609.0,
      "step": 2933
    },
    {
      "entropy": 0.7969068903476,
      "epoch": 0.26994669380854625,
      "grad_norm": 0.15432360768318176,
      "learning_rate": 9.100499892661085e-05,
      "loss": 0.7808,
      "mean_token_accuracy": 0.772589173167944,
      "num_tokens": 93622012.0,
      "step": 2934
    },
    {
      "entropy": 0.8587327469140291,
      "epoch": 0.2700387001799875,
      "grad_norm": 0.15856295824050903,
      "learning_rate": 9.100193210046923e-05,
      "loss": 0.8264,
      "mean_token_accuracy": 0.7618689090013504,
      "num_tokens": 93654050.0,
      "step": 2935
    },
    {
      "entropy": 0.708247160539031,
      "epoch": 0.2701307065514287,
      "grad_norm": 0.1483694314956665,
      "learning_rate": 9.099886527432761e-05,
      "loss": 0.6861,
      "mean_token_accuracy": 0.7980261296033859,
      "num_tokens": 93685526.0,
      "step": 2936
    },
    {
      "entropy": 0.7031150795519352,
      "epoch": 0.2702227129228699,
      "grad_norm": 0.142785906791687,
      "learning_rate": 9.099579844818598e-05,
      "loss": 0.6696,
      "mean_token_accuracy": 0.798668883740902,
      "num_tokens": 93717977.0,
      "step": 2937
    },
    {
      "entropy": 0.74162114597857,
      "epoch": 0.2703147192943111,
      "grad_norm": 0.14787058532238007,
      "learning_rate": 9.099273162204435e-05,
      "loss": 0.71,
      "mean_token_accuracy": 0.7896817810833454,
      "num_tokens": 93750294.0,
      "step": 2938
    },
    {
      "entropy": 0.788647098466754,
      "epoch": 0.27040672566575236,
      "grad_norm": 0.15811526775360107,
      "learning_rate": 9.098966479590273e-05,
      "loss": 0.7685,
      "mean_token_accuracy": 0.7755924426019192,
      "num_tokens": 93782450.0,
      "step": 2939
    },
    {
      "entropy": 0.8642611242830753,
      "epoch": 0.2704987320371936,
      "grad_norm": 0.1679096817970276,
      "learning_rate": 9.09865979697611e-05,
      "loss": 0.8838,
      "mean_token_accuracy": 0.7484699599444866,
      "num_tokens": 93814362.0,
      "step": 2940
    },
    {
      "entropy": 0.7897633705288172,
      "epoch": 0.2705907384086348,
      "grad_norm": 0.1521085798740387,
      "learning_rate": 9.098353114361948e-05,
      "loss": 0.779,
      "mean_token_accuracy": 0.7745413258671761,
      "num_tokens": 93846662.0,
      "step": 2941
    },
    {
      "entropy": 0.8136439248919487,
      "epoch": 0.27068274478007603,
      "grad_norm": 0.15939153730869293,
      "learning_rate": 9.098046431747784e-05,
      "loss": 0.8047,
      "mean_token_accuracy": 0.767129547894001,
      "num_tokens": 93878885.0,
      "step": 2942
    },
    {
      "entropy": 0.7881902754306793,
      "epoch": 0.2707747511515172,
      "grad_norm": 0.16863900423049927,
      "learning_rate": 9.097739749133623e-05,
      "loss": 0.7945,
      "mean_token_accuracy": 0.7710217796266079,
      "num_tokens": 93911130.0,
      "step": 2943
    },
    {
      "entropy": 0.8621927984058857,
      "epoch": 0.27086675752295847,
      "grad_norm": 0.16594620048999786,
      "learning_rate": 9.09743306651946e-05,
      "loss": 0.8853,
      "mean_token_accuracy": 0.7476134225726128,
      "num_tokens": 93942586.0,
      "step": 2944
    },
    {
      "entropy": 0.782697232440114,
      "epoch": 0.2709587638943997,
      "grad_norm": 0.15686047077178955,
      "learning_rate": 9.097126383905296e-05,
      "loss": 0.8043,
      "mean_token_accuracy": 0.7662477903068066,
      "num_tokens": 93974821.0,
      "step": 2945
    },
    {
      "entropy": 0.7212401553988457,
      "epoch": 0.2710507702658409,
      "grad_norm": 0.14853939414024353,
      "learning_rate": 9.096819701291134e-05,
      "loss": 0.7138,
      "mean_token_accuracy": 0.7913515493273735,
      "num_tokens": 94007336.0,
      "step": 2946
    },
    {
      "entropy": 0.7666271422058344,
      "epoch": 0.27114277663728215,
      "grad_norm": 0.15881362557411194,
      "learning_rate": 9.096513018676972e-05,
      "loss": 0.773,
      "mean_token_accuracy": 0.7732622623443604,
      "num_tokens": 94039504.0,
      "step": 2947
    },
    {
      "entropy": 0.9152071960270405,
      "epoch": 0.27123478300872333,
      "grad_norm": 0.16532039642333984,
      "learning_rate": 9.096206336062809e-05,
      "loss": 0.92,
      "mean_token_accuracy": 0.7424666732549667,
      "num_tokens": 94070964.0,
      "step": 2948
    },
    {
      "entropy": 0.916716668754816,
      "epoch": 0.2713267893801646,
      "grad_norm": 0.16701439023017883,
      "learning_rate": 9.095899653448646e-05,
      "loss": 0.9091,
      "mean_token_accuracy": 0.743901114910841,
      "num_tokens": 94102363.0,
      "step": 2949
    },
    {
      "entropy": 0.8728654514998198,
      "epoch": 0.2714187957516058,
      "grad_norm": 0.15661324560642242,
      "learning_rate": 9.095592970834483e-05,
      "loss": 0.8514,
      "mean_token_accuracy": 0.757886178791523,
      "num_tokens": 94134676.0,
      "step": 2950
    },
    {
      "entropy": 0.8328176867216825,
      "epoch": 0.271510802123047,
      "grad_norm": 0.1591370850801468,
      "learning_rate": 9.095286288220321e-05,
      "loss": 0.8199,
      "mean_token_accuracy": 0.7636412084102631,
      "num_tokens": 94165761.0,
      "step": 2951
    },
    {
      "entropy": 0.8107479028403759,
      "epoch": 0.27160280849448826,
      "grad_norm": 0.15338759124279022,
      "learning_rate": 9.094979605606159e-05,
      "loss": 0.8044,
      "mean_token_accuracy": 0.7673875987529755,
      "num_tokens": 94197842.0,
      "step": 2952
    },
    {
      "entropy": 0.8862663507461548,
      "epoch": 0.27169481486592945,
      "grad_norm": 0.16415540874004364,
      "learning_rate": 9.094672922991996e-05,
      "loss": 0.8867,
      "mean_token_accuracy": 0.7505876943469048,
      "num_tokens": 94229622.0,
      "step": 2953
    },
    {
      "entropy": 0.8763725962489843,
      "epoch": 0.2717868212373707,
      "grad_norm": 0.16790227591991425,
      "learning_rate": 9.094366240377833e-05,
      "loss": 0.8621,
      "mean_token_accuracy": 0.7538400143384933,
      "num_tokens": 94261997.0,
      "step": 2954
    },
    {
      "entropy": 0.884623434394598,
      "epoch": 0.27187882760881193,
      "grad_norm": 0.15819025039672852,
      "learning_rate": 9.094059557763671e-05,
      "loss": 0.8772,
      "mean_token_accuracy": 0.7479198016226292,
      "num_tokens": 94294294.0,
      "step": 2955
    },
    {
      "entropy": 0.8282515369355679,
      "epoch": 0.2719708339802531,
      "grad_norm": 0.16125279664993286,
      "learning_rate": 9.093752875149508e-05,
      "loss": 0.8273,
      "mean_token_accuracy": 0.7549436241388321,
      "num_tokens": 94325637.0,
      "step": 2956
    },
    {
      "entropy": 0.8680758066475391,
      "epoch": 0.27206284035169437,
      "grad_norm": 0.1641145646572113,
      "learning_rate": 9.093446192535346e-05,
      "loss": 0.8718,
      "mean_token_accuracy": 0.744707252830267,
      "num_tokens": 94357289.0,
      "step": 2957
    },
    {
      "entropy": 0.7880359161645174,
      "epoch": 0.27215484672313556,
      "grad_norm": 0.1616339534521103,
      "learning_rate": 9.093139509921183e-05,
      "loss": 0.7846,
      "mean_token_accuracy": 0.7716490402817726,
      "num_tokens": 94389155.0,
      "step": 2958
    },
    {
      "entropy": 0.9108173027634621,
      "epoch": 0.2722468530945768,
      "grad_norm": 0.1686721295118332,
      "learning_rate": 9.09283282730702e-05,
      "loss": 0.9142,
      "mean_token_accuracy": 0.7414024248719215,
      "num_tokens": 94420543.0,
      "step": 2959
    },
    {
      "entropy": 0.78360809199512,
      "epoch": 0.27233885946601805,
      "grad_norm": 0.15779073536396027,
      "learning_rate": 9.092526144692857e-05,
      "loss": 0.7713,
      "mean_token_accuracy": 0.7777864933013916,
      "num_tokens": 94452159.0,
      "step": 2960
    },
    {
      "entropy": 0.8400637730956078,
      "epoch": 0.27243086583745924,
      "grad_norm": 0.15700452029705048,
      "learning_rate": 9.092219462078696e-05,
      "loss": 0.8397,
      "mean_token_accuracy": 0.7588993087410927,
      "num_tokens": 94484159.0,
      "step": 2961
    },
    {
      "entropy": 0.7915164530277252,
      "epoch": 0.2725228722089005,
      "grad_norm": 0.15352746844291687,
      "learning_rate": 9.091912779464534e-05,
      "loss": 0.7864,
      "mean_token_accuracy": 0.7751951143145561,
      "num_tokens": 94516175.0,
      "step": 2962
    },
    {
      "entropy": 0.8328734748065472,
      "epoch": 0.27261487858034167,
      "grad_norm": 0.1576293259859085,
      "learning_rate": 9.09160609685037e-05,
      "loss": 0.8261,
      "mean_token_accuracy": 0.7598410397768021,
      "num_tokens": 94548095.0,
      "step": 2963
    },
    {
      "entropy": 0.8622080404311419,
      "epoch": 0.2727068849517829,
      "grad_norm": 0.1584637314081192,
      "learning_rate": 9.091299414236207e-05,
      "loss": 0.8308,
      "mean_token_accuracy": 0.7570712193846703,
      "num_tokens": 94579536.0,
      "step": 2964
    },
    {
      "entropy": 0.8583197500556707,
      "epoch": 0.27279889132322416,
      "grad_norm": 0.15359093248844147,
      "learning_rate": 9.090992731622044e-05,
      "loss": 0.8507,
      "mean_token_accuracy": 0.7597511447966099,
      "num_tokens": 94612002.0,
      "step": 2965
    },
    {
      "entropy": 0.772807827219367,
      "epoch": 0.27289089769466535,
      "grad_norm": 0.160162091255188,
      "learning_rate": 9.090686049007882e-05,
      "loss": 0.7881,
      "mean_token_accuracy": 0.7743915058672428,
      "num_tokens": 94643756.0,
      "step": 2966
    },
    {
      "entropy": 0.8693925552070141,
      "epoch": 0.2729829040661066,
      "grad_norm": 0.16301539540290833,
      "learning_rate": 9.09037936639372e-05,
      "loss": 0.8823,
      "mean_token_accuracy": 0.7461791448295116,
      "num_tokens": 94675115.0,
      "step": 2967
    },
    {
      "entropy": 0.8348445501178503,
      "epoch": 0.2730749104375478,
      "grad_norm": 0.15176229178905487,
      "learning_rate": 9.090072683779557e-05,
      "loss": 0.8239,
      "mean_token_accuracy": 0.7610224559903145,
      "num_tokens": 94706977.0,
      "step": 2968
    },
    {
      "entropy": 0.7776477113366127,
      "epoch": 0.273166916808989,
      "grad_norm": 0.1524350643157959,
      "learning_rate": 9.089766001165394e-05,
      "loss": 0.7787,
      "mean_token_accuracy": 0.7749852202832699,
      "num_tokens": 94739125.0,
      "step": 2969
    },
    {
      "entropy": 0.9166034385561943,
      "epoch": 0.27325892318043027,
      "grad_norm": 0.16388548910617828,
      "learning_rate": 9.089459318551232e-05,
      "loss": 0.9199,
      "mean_token_accuracy": 0.7411688752472401,
      "num_tokens": 94770736.0,
      "step": 2970
    },
    {
      "entropy": 0.8267360404133797,
      "epoch": 0.27335092955187146,
      "grad_norm": 0.16447369754314423,
      "learning_rate": 9.089152635937069e-05,
      "loss": 0.8417,
      "mean_token_accuracy": 0.7599836103618145,
      "num_tokens": 94802979.0,
      "step": 2971
    },
    {
      "entropy": 0.8375822231173515,
      "epoch": 0.2734429359233127,
      "grad_norm": 0.15226303040981293,
      "learning_rate": 9.088845953322907e-05,
      "loss": 0.8131,
      "mean_token_accuracy": 0.7628697380423546,
      "num_tokens": 94835711.0,
      "step": 2972
    },
    {
      "entropy": 0.8423245623707771,
      "epoch": 0.2735349422947539,
      "grad_norm": 0.1605745106935501,
      "learning_rate": 9.088539270708744e-05,
      "loss": 0.8116,
      "mean_token_accuracy": 0.7631172426044941,
      "num_tokens": 94867889.0,
      "step": 2973
    },
    {
      "entropy": 0.8679430782794952,
      "epoch": 0.27362694866619514,
      "grad_norm": 0.16649280488491058,
      "learning_rate": 9.088232588094582e-05,
      "loss": 0.8791,
      "mean_token_accuracy": 0.7484576813876629,
      "num_tokens": 94900297.0,
      "step": 2974
    },
    {
      "entropy": 0.9263138938695192,
      "epoch": 0.2737189550376364,
      "grad_norm": 0.16073080897331238,
      "learning_rate": 9.087925905480419e-05,
      "loss": 0.9172,
      "mean_token_accuracy": 0.7392488159239292,
      "num_tokens": 94932049.0,
      "step": 2975
    },
    {
      "entropy": 0.827193096280098,
      "epoch": 0.27381096140907757,
      "grad_norm": 0.16261929273605347,
      "learning_rate": 9.087619222866256e-05,
      "loss": 0.81,
      "mean_token_accuracy": 0.7655958086252213,
      "num_tokens": 94963836.0,
      "step": 2976
    },
    {
      "entropy": 0.8339664973318577,
      "epoch": 0.2739029677805188,
      "grad_norm": 0.18173347413539886,
      "learning_rate": 9.087312540252094e-05,
      "loss": 0.8137,
      "mean_token_accuracy": 0.7635317668318748,
      "num_tokens": 94995582.0,
      "step": 2977
    },
    {
      "entropy": 1.0048192217946053,
      "epoch": 0.27399497415196,
      "grad_norm": 0.16850464046001434,
      "learning_rate": 9.087005857637932e-05,
      "loss": 0.9898,
      "mean_token_accuracy": 0.7236749157309532,
      "num_tokens": 95027728.0,
      "step": 2978
    },
    {
      "entropy": 0.972122248262167,
      "epoch": 0.27408698052340125,
      "grad_norm": 0.17038923501968384,
      "learning_rate": 9.086699175023769e-05,
      "loss": 0.9576,
      "mean_token_accuracy": 0.7278411760926247,
      "num_tokens": 95059789.0,
      "step": 2979
    },
    {
      "entropy": 0.8699118886142969,
      "epoch": 0.2741789868948425,
      "grad_norm": 0.16425879299640656,
      "learning_rate": 9.086392492409605e-05,
      "loss": 0.8812,
      "mean_token_accuracy": 0.7451834864914417,
      "num_tokens": 95090763.0,
      "step": 2980
    },
    {
      "entropy": 0.8208758495748043,
      "epoch": 0.2742709932662837,
      "grad_norm": 0.15566222369670868,
      "learning_rate": 9.086085809795442e-05,
      "loss": 0.8041,
      "mean_token_accuracy": 0.7663186825811863,
      "num_tokens": 95121696.0,
      "step": 2981
    },
    {
      "entropy": 0.849691204726696,
      "epoch": 0.2743629996377249,
      "grad_norm": 0.1611594259738922,
      "learning_rate": 9.08577912718128e-05,
      "loss": 0.8578,
      "mean_token_accuracy": 0.7535506673157215,
      "num_tokens": 95153936.0,
      "step": 2982
    },
    {
      "entropy": 0.7956834398210049,
      "epoch": 0.2744550060091661,
      "grad_norm": 0.15793266892433167,
      "learning_rate": 9.085472444567118e-05,
      "loss": 0.7757,
      "mean_token_accuracy": 0.7756148315966129,
      "num_tokens": 95185547.0,
      "step": 2983
    },
    {
      "entropy": 0.8239380456507206,
      "epoch": 0.27454701238060736,
      "grad_norm": 0.1584516167640686,
      "learning_rate": 9.085165761952955e-05,
      "loss": 0.85,
      "mean_token_accuracy": 0.7543034441769123,
      "num_tokens": 95217162.0,
      "step": 2984
    },
    {
      "entropy": 0.8835084550082684,
      "epoch": 0.2746390187520486,
      "grad_norm": 0.16623002290725708,
      "learning_rate": 9.084859079338792e-05,
      "loss": 0.8925,
      "mean_token_accuracy": 0.7468123622238636,
      "num_tokens": 95248737.0,
      "step": 2985
    },
    {
      "entropy": 0.7768028266727924,
      "epoch": 0.2747310251234898,
      "grad_norm": 0.15412400662899017,
      "learning_rate": 9.08455239672463e-05,
      "loss": 0.7761,
      "mean_token_accuracy": 0.7758184373378754,
      "num_tokens": 95281158.0,
      "step": 2986
    },
    {
      "entropy": 0.8305335007607937,
      "epoch": 0.27482303149493104,
      "grad_norm": 0.1570277363061905,
      "learning_rate": 9.084245714110467e-05,
      "loss": 0.8062,
      "mean_token_accuracy": 0.7686524279415607,
      "num_tokens": 95312994.0,
      "step": 2987
    },
    {
      "entropy": 0.9706510528922081,
      "epoch": 0.2749150378663722,
      "grad_norm": 0.16239024698734283,
      "learning_rate": 9.083939031496305e-05,
      "loss": 0.9644,
      "mean_token_accuracy": 0.7262749038636684,
      "num_tokens": 95344817.0,
      "step": 2988
    },
    {
      "entropy": 0.8781954552978277,
      "epoch": 0.27500704423781347,
      "grad_norm": 0.15656183660030365,
      "learning_rate": 9.083632348882142e-05,
      "loss": 0.8761,
      "mean_token_accuracy": 0.7491669654846191,
      "num_tokens": 95376114.0,
      "step": 2989
    },
    {
      "entropy": 0.8731825575232506,
      "epoch": 0.2750990506092547,
      "grad_norm": 0.157260462641716,
      "learning_rate": 9.08332566626798e-05,
      "loss": 0.8535,
      "mean_token_accuracy": 0.7532857693731785,
      "num_tokens": 95408646.0,
      "step": 2990
    },
    {
      "entropy": 0.8857558965682983,
      "epoch": 0.2751910569806959,
      "grad_norm": 0.16027893126010895,
      "learning_rate": 9.083018983653817e-05,
      "loss": 0.8779,
      "mean_token_accuracy": 0.7472330816090107,
      "num_tokens": 95440773.0,
      "step": 2991
    },
    {
      "entropy": 0.9255579076707363,
      "epoch": 0.27528306335213715,
      "grad_norm": 0.1640411615371704,
      "learning_rate": 9.082712301039654e-05,
      "loss": 0.9313,
      "mean_token_accuracy": 0.7356943488121033,
      "num_tokens": 95472810.0,
      "step": 2992
    },
    {
      "entropy": 0.9703857898712158,
      "epoch": 0.27537506972357834,
      "grad_norm": 0.16848722100257874,
      "learning_rate": 9.082405618425493e-05,
      "loss": 0.9862,
      "mean_token_accuracy": 0.7236948944628239,
      "num_tokens": 95505014.0,
      "step": 2993
    },
    {
      "entropy": 0.9352265205234289,
      "epoch": 0.2754670760950196,
      "grad_norm": 0.164889857172966,
      "learning_rate": 9.08209893581133e-05,
      "loss": 0.9474,
      "mean_token_accuracy": 0.7310321778059006,
      "num_tokens": 95535632.0,
      "step": 2994
    },
    {
      "entropy": 0.7133045196533203,
      "epoch": 0.2755590824664608,
      "grad_norm": 0.15168532729148865,
      "learning_rate": 9.081792253197167e-05,
      "loss": 0.6987,
      "mean_token_accuracy": 0.7955734096467495,
      "num_tokens": 95567726.0,
      "step": 2995
    },
    {
      "entropy": 0.8008441478013992,
      "epoch": 0.275651088837902,
      "grad_norm": 0.1496473252773285,
      "learning_rate": 9.081485570583003e-05,
      "loss": 0.7892,
      "mean_token_accuracy": 0.7716951742768288,
      "num_tokens": 95600049.0,
      "step": 2996
    },
    {
      "entropy": 0.8313410133123398,
      "epoch": 0.27574309520934326,
      "grad_norm": 0.16158482432365417,
      "learning_rate": 9.081178887968842e-05,
      "loss": 0.8374,
      "mean_token_accuracy": 0.7552817352116108,
      "num_tokens": 95632380.0,
      "step": 2997
    },
    {
      "entropy": 0.7395443581044674,
      "epoch": 0.27583510158078445,
      "grad_norm": 0.15348933637142181,
      "learning_rate": 9.08087220535468e-05,
      "loss": 0.7413,
      "mean_token_accuracy": 0.7850919403135777,
      "num_tokens": 95664885.0,
      "step": 2998
    },
    {
      "entropy": 0.8483192902058363,
      "epoch": 0.2759271079522257,
      "grad_norm": 0.1603902280330658,
      "learning_rate": 9.080565522740517e-05,
      "loss": 0.8638,
      "mean_token_accuracy": 0.7512321062386036,
      "num_tokens": 95697242.0,
      "step": 2999
    },
    {
      "entropy": 0.8169868588447571,
      "epoch": 0.27601911432366694,
      "grad_norm": 0.16765935719013214,
      "learning_rate": 9.080258840126353e-05,
      "loss": 0.8414,
      "mean_token_accuracy": 0.7566691935062408,
      "num_tokens": 95729269.0,
      "step": 3000
    },
    {
      "entropy": 0.8693693690001965,
      "epoch": 0.2761111206951081,
      "grad_norm": 0.1639304757118225,
      "learning_rate": 9.079952157512191e-05,
      "loss": 0.8652,
      "mean_token_accuracy": 0.7511313483119011,
      "num_tokens": 95761172.0,
      "step": 3001
    },
    {
      "entropy": 0.8800292164087296,
      "epoch": 0.27620312706654937,
      "grad_norm": 0.1628619283437729,
      "learning_rate": 9.079645474898028e-05,
      "loss": 0.8764,
      "mean_token_accuracy": 0.748461801558733,
      "num_tokens": 95793369.0,
      "step": 3002
    },
    {
      "entropy": 0.8230822794139385,
      "epoch": 0.27629513343799056,
      "grad_norm": 0.1606339067220688,
      "learning_rate": 9.079338792283866e-05,
      "loss": 0.8047,
      "mean_token_accuracy": 0.76668830960989,
      "num_tokens": 95825065.0,
      "step": 3003
    },
    {
      "entropy": 0.8763422165066004,
      "epoch": 0.2763871398094318,
      "grad_norm": 0.15909306704998016,
      "learning_rate": 9.079032109669703e-05,
      "loss": 0.8644,
      "mean_token_accuracy": 0.7526733800768852,
      "num_tokens": 95857430.0,
      "step": 3004
    },
    {
      "entropy": 0.8709813579916954,
      "epoch": 0.27647914618087305,
      "grad_norm": 0.16085666418075562,
      "learning_rate": 9.078725427055541e-05,
      "loss": 0.8673,
      "mean_token_accuracy": 0.7520282082259655,
      "num_tokens": 95889812.0,
      "step": 3005
    },
    {
      "entropy": 0.8156521767377853,
      "epoch": 0.27657115255231424,
      "grad_norm": 0.1562260240316391,
      "learning_rate": 9.078418744441378e-05,
      "loss": 0.801,
      "mean_token_accuracy": 0.7634498700499535,
      "num_tokens": 95920754.0,
      "step": 3006
    },
    {
      "entropy": 0.8744444735348225,
      "epoch": 0.2766631589237555,
      "grad_norm": 0.16485761106014252,
      "learning_rate": 9.078112061827215e-05,
      "loss": 0.8779,
      "mean_token_accuracy": 0.7486506588757038,
      "num_tokens": 95952618.0,
      "step": 3007
    },
    {
      "entropy": 0.8764996509999037,
      "epoch": 0.27675516529519667,
      "grad_norm": 0.15946827828884125,
      "learning_rate": 9.077805379213053e-05,
      "loss": 0.8578,
      "mean_token_accuracy": 0.7561634182929993,
      "num_tokens": 95984934.0,
      "step": 3008
    },
    {
      "entropy": 0.8214257098734379,
      "epoch": 0.2768471716666379,
      "grad_norm": 0.15322959423065186,
      "learning_rate": 9.077498696598891e-05,
      "loss": 0.7879,
      "mean_token_accuracy": 0.7733931690454483,
      "num_tokens": 96017012.0,
      "step": 3009
    },
    {
      "entropy": 0.7694719340652227,
      "epoch": 0.27693917803807916,
      "grad_norm": 0.15115885436534882,
      "learning_rate": 9.077192013984728e-05,
      "loss": 0.7501,
      "mean_token_accuracy": 0.7830321080982685,
      "num_tokens": 96049137.0,
      "step": 3010
    },
    {
      "entropy": 0.7702351287007332,
      "epoch": 0.27703118440952035,
      "grad_norm": 0.15306374430656433,
      "learning_rate": 9.076885331370565e-05,
      "loss": 0.7515,
      "mean_token_accuracy": 0.779937133193016,
      "num_tokens": 96080694.0,
      "step": 3011
    },
    {
      "entropy": 0.8132718652486801,
      "epoch": 0.2771231907809616,
      "grad_norm": 0.15643073618412018,
      "learning_rate": 9.076578648756402e-05,
      "loss": 0.8238,
      "mean_token_accuracy": 0.7605291269719601,
      "num_tokens": 96112615.0,
      "step": 3012
    },
    {
      "entropy": 0.809041814878583,
      "epoch": 0.2772151971524028,
      "grad_norm": 0.15925423800945282,
      "learning_rate": 9.07627196614224e-05,
      "loss": 0.8176,
      "mean_token_accuracy": 0.7614431530237198,
      "num_tokens": 96144276.0,
      "step": 3013
    },
    {
      "entropy": 0.8340934030711651,
      "epoch": 0.277307203523844,
      "grad_norm": 0.1632581204175949,
      "learning_rate": 9.075965283528078e-05,
      "loss": 0.83,
      "mean_token_accuracy": 0.7579736672341824,
      "num_tokens": 96175532.0,
      "step": 3014
    },
    {
      "entropy": 0.8343224413692951,
      "epoch": 0.27739920989528527,
      "grad_norm": 0.15912485122680664,
      "learning_rate": 9.075658600913915e-05,
      "loss": 0.8399,
      "mean_token_accuracy": 0.7596706598997116,
      "num_tokens": 96208197.0,
      "step": 3015
    },
    {
      "entropy": 0.8033910486847162,
      "epoch": 0.27749121626672646,
      "grad_norm": 0.15847407281398773,
      "learning_rate": 9.075351918299751e-05,
      "loss": 0.7928,
      "mean_token_accuracy": 0.7677336372435093,
      "num_tokens": 96239988.0,
      "step": 3016
    },
    {
      "entropy": 0.80944231338799,
      "epoch": 0.2775832226381677,
      "grad_norm": 0.16358715295791626,
      "learning_rate": 9.07504523568559e-05,
      "loss": 0.8023,
      "mean_token_accuracy": 0.7688773572444916,
      "num_tokens": 96271670.0,
      "step": 3017
    },
    {
      "entropy": 0.8125239014625549,
      "epoch": 0.2776752290096089,
      "grad_norm": 0.15767213702201843,
      "learning_rate": 9.074738553071426e-05,
      "loss": 0.7968,
      "mean_token_accuracy": 0.7673500031232834,
      "num_tokens": 96303581.0,
      "step": 3018
    },
    {
      "entropy": 0.7828193064779043,
      "epoch": 0.27776723538105014,
      "grad_norm": 0.1558314561843872,
      "learning_rate": 9.074431870457264e-05,
      "loss": 0.7975,
      "mean_token_accuracy": 0.7695734240114689,
      "num_tokens": 96335978.0,
      "step": 3019
    },
    {
      "entropy": 0.9847223199903965,
      "epoch": 0.2778592417524914,
      "grad_norm": 0.17096324265003204,
      "learning_rate": 9.074125187843101e-05,
      "loss": 0.9865,
      "mean_token_accuracy": 0.722423754632473,
      "num_tokens": 96367788.0,
      "step": 3020
    },
    {
      "entropy": 0.8184599839150906,
      "epoch": 0.27795124812393257,
      "grad_norm": 0.15131962299346924,
      "learning_rate": 9.07381850522894e-05,
      "loss": 0.8203,
      "mean_token_accuracy": 0.7622746899724007,
      "num_tokens": 96400436.0,
      "step": 3021
    },
    {
      "entropy": 0.835424954071641,
      "epoch": 0.2780432544953738,
      "grad_norm": 0.16169176995754242,
      "learning_rate": 9.073511822614776e-05,
      "loss": 0.8073,
      "mean_token_accuracy": 0.766181580722332,
      "num_tokens": 96432150.0,
      "step": 3022
    },
    {
      "entropy": 0.8430908862501383,
      "epoch": 0.278135260866815,
      "grad_norm": 0.1576007753610611,
      "learning_rate": 9.073205140000613e-05,
      "loss": 0.8274,
      "mean_token_accuracy": 0.7616921812295914,
      "num_tokens": 96464588.0,
      "step": 3023
    },
    {
      "entropy": 0.9960003159940243,
      "epoch": 0.27822726723825625,
      "grad_norm": 0.170658677816391,
      "learning_rate": 9.072898457386451e-05,
      "loss": 1.0038,
      "mean_token_accuracy": 0.7194048874080181,
      "num_tokens": 96496449.0,
      "step": 3024
    },
    {
      "entropy": 0.9057288207113743,
      "epoch": 0.2783192736096975,
      "grad_norm": 0.1563260406255722,
      "learning_rate": 9.072591774772289e-05,
      "loss": 0.8911,
      "mean_token_accuracy": 0.7481472454965115,
      "num_tokens": 96528644.0,
      "step": 3025
    },
    {
      "entropy": 0.8190557882189751,
      "epoch": 0.2784112799811387,
      "grad_norm": 0.15884803235530853,
      "learning_rate": 9.072285092158126e-05,
      "loss": 0.8293,
      "mean_token_accuracy": 0.7597166784107685,
      "num_tokens": 96560295.0,
      "step": 3026
    },
    {
      "entropy": 0.8881460875272751,
      "epoch": 0.2785032863525799,
      "grad_norm": 0.15954580903053284,
      "learning_rate": 9.071978409543963e-05,
      "loss": 0.879,
      "mean_token_accuracy": 0.7495601177215576,
      "num_tokens": 96592206.0,
      "step": 3027
    },
    {
      "entropy": 0.7762919329106808,
      "epoch": 0.2785952927240211,
      "grad_norm": 0.1603749692440033,
      "learning_rate": 9.071671726929801e-05,
      "loss": 0.7737,
      "mean_token_accuracy": 0.7778223156929016,
      "num_tokens": 96624077.0,
      "step": 3028
    },
    {
      "entropy": 0.9041752405464649,
      "epoch": 0.27868729909546236,
      "grad_norm": 0.1616906374692917,
      "learning_rate": 9.071365044315638e-05,
      "loss": 0.9061,
      "mean_token_accuracy": 0.7449831888079643,
      "num_tokens": 96656327.0,
      "step": 3029
    },
    {
      "entropy": 0.8395270928740501,
      "epoch": 0.2787793054669036,
      "grad_norm": 0.1572946012020111,
      "learning_rate": 9.071058361701476e-05,
      "loss": 0.8253,
      "mean_token_accuracy": 0.7624619826674461,
      "num_tokens": 96688730.0,
      "step": 3030
    },
    {
      "entropy": 0.9404210057109594,
      "epoch": 0.2788713118383448,
      "grad_norm": 0.16468746960163116,
      "learning_rate": 9.070751679087313e-05,
      "loss": 0.9472,
      "mean_token_accuracy": 0.7325521782040596,
      "num_tokens": 96720694.0,
      "step": 3031
    },
    {
      "entropy": 0.8453200533986092,
      "epoch": 0.27896331820978604,
      "grad_norm": 0.15862196683883667,
      "learning_rate": 9.070444996473151e-05,
      "loss": 0.8425,
      "mean_token_accuracy": 0.7599583938717842,
      "num_tokens": 96752851.0,
      "step": 3032
    },
    {
      "entropy": 0.9402693547308445,
      "epoch": 0.2790553245812272,
      "grad_norm": 0.16938772797584534,
      "learning_rate": 9.070138313858988e-05,
      "loss": 0.946,
      "mean_token_accuracy": 0.7348719388246536,
      "num_tokens": 96784744.0,
      "step": 3033
    },
    {
      "entropy": 0.8861788269132376,
      "epoch": 0.27914733095266847,
      "grad_norm": 0.16145122051239014,
      "learning_rate": 9.069831631244824e-05,
      "loss": 0.8833,
      "mean_token_accuracy": 0.7482912875711918,
      "num_tokens": 96816940.0,
      "step": 3034
    },
    {
      "entropy": 0.8151098862290382,
      "epoch": 0.2792393373241097,
      "grad_norm": 0.1536400318145752,
      "learning_rate": 9.069524948630663e-05,
      "loss": 0.8215,
      "mean_token_accuracy": 0.7623922787606716,
      "num_tokens": 96849156.0,
      "step": 3035
    },
    {
      "entropy": 0.798476368188858,
      "epoch": 0.2793313436955509,
      "grad_norm": 0.1630336046218872,
      "learning_rate": 9.069218266016501e-05,
      "loss": 0.8238,
      "mean_token_accuracy": 0.7663250789046288,
      "num_tokens": 96880701.0,
      "step": 3036
    },
    {
      "entropy": 0.8838012348860502,
      "epoch": 0.27942335006699215,
      "grad_norm": 0.167633518576622,
      "learning_rate": 9.068911583402337e-05,
      "loss": 0.8865,
      "mean_token_accuracy": 0.7489401549100876,
      "num_tokens": 96912675.0,
      "step": 3037
    },
    {
      "entropy": 0.8552806749939919,
      "epoch": 0.27951535643843334,
      "grad_norm": 0.16118790209293365,
      "learning_rate": 9.068604900788174e-05,
      "loss": 0.8533,
      "mean_token_accuracy": 0.7546300143003464,
      "num_tokens": 96944602.0,
      "step": 3038
    },
    {
      "entropy": 0.818317661061883,
      "epoch": 0.2796073628098746,
      "grad_norm": 0.15386563539505005,
      "learning_rate": 9.068298218174011e-05,
      "loss": 0.8036,
      "mean_token_accuracy": 0.7656272016465664,
      "num_tokens": 96976155.0,
      "step": 3039
    },
    {
      "entropy": 0.8429694045335054,
      "epoch": 0.2796993691813158,
      "grad_norm": 0.16196073591709137,
      "learning_rate": 9.06799153555985e-05,
      "loss": 0.8482,
      "mean_token_accuracy": 0.756141223013401,
      "num_tokens": 97007825.0,
      "step": 3040
    },
    {
      "entropy": 0.9398678317666054,
      "epoch": 0.279791375552757,
      "grad_norm": 0.16179198026657104,
      "learning_rate": 9.067684852945687e-05,
      "loss": 0.9357,
      "mean_token_accuracy": 0.7334702424705029,
      "num_tokens": 97040003.0,
      "step": 3041
    },
    {
      "entropy": 0.8921387400478125,
      "epoch": 0.27988338192419826,
      "grad_norm": 0.16059865057468414,
      "learning_rate": 9.067378170331524e-05,
      "loss": 0.8662,
      "mean_token_accuracy": 0.7533046007156372,
      "num_tokens": 97071574.0,
      "step": 3042
    },
    {
      "entropy": 0.7835614439100027,
      "epoch": 0.27997538829563945,
      "grad_norm": 0.15247178077697754,
      "learning_rate": 9.067071487717361e-05,
      "loss": 0.7748,
      "mean_token_accuracy": 0.7723787650465965,
      "num_tokens": 97103588.0,
      "step": 3043
    },
    {
      "entropy": 0.7534362487494946,
      "epoch": 0.2800673946670807,
      "grad_norm": 0.15129807591438293,
      "learning_rate": 9.066764805103199e-05,
      "loss": 0.7212,
      "mean_token_accuracy": 0.7908092476427555,
      "num_tokens": 97135374.0,
      "step": 3044
    },
    {
      "entropy": 0.8090875055640936,
      "epoch": 0.28015940103852194,
      "grad_norm": 0.15282028913497925,
      "learning_rate": 9.066458122489037e-05,
      "loss": 0.7916,
      "mean_token_accuracy": 0.7709947191178799,
      "num_tokens": 97167961.0,
      "step": 3045
    },
    {
      "entropy": 0.8359926640987396,
      "epoch": 0.2802514074099631,
      "grad_norm": 0.15902599692344666,
      "learning_rate": 9.066151439874874e-05,
      "loss": 0.8219,
      "mean_token_accuracy": 0.7627169527113438,
      "num_tokens": 97200600.0,
      "step": 3046
    },
    {
      "entropy": 0.8003830555826426,
      "epoch": 0.28034341378140437,
      "grad_norm": 0.16334044933319092,
      "learning_rate": 9.065844757260711e-05,
      "loss": 0.7936,
      "mean_token_accuracy": 0.7708536833524704,
      "num_tokens": 97233368.0,
      "step": 3047
    },
    {
      "entropy": 0.7821346186101437,
      "epoch": 0.28043542015284556,
      "grad_norm": 0.15913254022598267,
      "learning_rate": 9.065538074646549e-05,
      "loss": 0.7749,
      "mean_token_accuracy": 0.7756946943700314,
      "num_tokens": 97264889.0,
      "step": 3048
    },
    {
      "entropy": 0.805832352489233,
      "epoch": 0.2805274265242868,
      "grad_norm": 0.16600418090820312,
      "learning_rate": 9.065231392032386e-05,
      "loss": 0.8469,
      "mean_token_accuracy": 0.7600515447556973,
      "num_tokens": 97296756.0,
      "step": 3049
    },
    {
      "entropy": 0.8462393563240767,
      "epoch": 0.28061943289572805,
      "grad_norm": 0.15984047949314117,
      "learning_rate": 9.064924709418224e-05,
      "loss": 0.8445,
      "mean_token_accuracy": 0.7557237260043621,
      "num_tokens": 97328666.0,
      "step": 3050
    },
    {
      "entropy": 0.8026137910783291,
      "epoch": 0.28071143926716924,
      "grad_norm": 0.15483611822128296,
      "learning_rate": 9.06461802680406e-05,
      "loss": 0.7951,
      "mean_token_accuracy": 0.7683999054133892,
      "num_tokens": 97360986.0,
      "step": 3051
    },
    {
      "entropy": 0.8257872983813286,
      "epoch": 0.2808034456386105,
      "grad_norm": 0.15431053936481476,
      "learning_rate": 9.064311344189899e-05,
      "loss": 0.8256,
      "mean_token_accuracy": 0.7610511183738708,
      "num_tokens": 97393084.0,
      "step": 3052
    },
    {
      "entropy": 0.7798477001488209,
      "epoch": 0.28089545201005167,
      "grad_norm": 0.16909031569957733,
      "learning_rate": 9.064004661575736e-05,
      "loss": 0.779,
      "mean_token_accuracy": 0.7751521952450275,
      "num_tokens": 97424491.0,
      "step": 3053
    },
    {
      "entropy": 0.9191584587097168,
      "epoch": 0.2809874583814929,
      "grad_norm": 0.16286154091358185,
      "learning_rate": 9.063697978961572e-05,
      "loss": 0.9228,
      "mean_token_accuracy": 0.736426517367363,
      "num_tokens": 97456431.0,
      "step": 3054
    },
    {
      "entropy": 0.8285694122314453,
      "epoch": 0.28107946475293416,
      "grad_norm": 0.1614476442337036,
      "learning_rate": 9.06339129634741e-05,
      "loss": 0.8291,
      "mean_token_accuracy": 0.7660128250718117,
      "num_tokens": 97488083.0,
      "step": 3055
    },
    {
      "entropy": 0.8523181714117527,
      "epoch": 0.28117147112437535,
      "grad_norm": 0.15937291085720062,
      "learning_rate": 9.063084613733249e-05,
      "loss": 0.8568,
      "mean_token_accuracy": 0.7619453556835651,
      "num_tokens": 97519872.0,
      "step": 3056
    },
    {
      "entropy": 0.7502105161547661,
      "epoch": 0.2812634774958166,
      "grad_norm": 0.14728102087974548,
      "learning_rate": 9.062777931119085e-05,
      "loss": 0.7266,
      "mean_token_accuracy": 0.7874207384884357,
      "num_tokens": 97552433.0,
      "step": 3057
    },
    {
      "entropy": 0.8058000989258289,
      "epoch": 0.2813554838672578,
      "grad_norm": 0.16382227838039398,
      "learning_rate": 9.062471248504922e-05,
      "loss": 0.7947,
      "mean_token_accuracy": 0.774554904550314,
      "num_tokens": 97583795.0,
      "step": 3058
    },
    {
      "entropy": 0.8055105842649937,
      "epoch": 0.281447490238699,
      "grad_norm": 0.15083789825439453,
      "learning_rate": 9.06216456589076e-05,
      "loss": 0.7987,
      "mean_token_accuracy": 0.7666711732745171,
      "num_tokens": 97616522.0,
      "step": 3059
    },
    {
      "entropy": 0.8919007368385792,
      "epoch": 0.28153949661014027,
      "grad_norm": 0.1586141139268875,
      "learning_rate": 9.061857883276597e-05,
      "loss": 0.891,
      "mean_token_accuracy": 0.7464226670563221,
      "num_tokens": 97648412.0,
      "step": 3060
    },
    {
      "entropy": 0.8744392860680819,
      "epoch": 0.28163150298158146,
      "grad_norm": 0.16142630577087402,
      "learning_rate": 9.061551200662435e-05,
      "loss": 0.8421,
      "mean_token_accuracy": 0.7628076300024986,
      "num_tokens": 97680318.0,
      "step": 3061
    },
    {
      "entropy": 0.895501621067524,
      "epoch": 0.2817235093530227,
      "grad_norm": 0.16055206954479218,
      "learning_rate": 9.061244518048272e-05,
      "loss": 0.9073,
      "mean_token_accuracy": 0.7470449060201645,
      "num_tokens": 97712767.0,
      "step": 3062
    },
    {
      "entropy": 0.8257643133401871,
      "epoch": 0.2818155157244639,
      "grad_norm": 0.1566387414932251,
      "learning_rate": 9.06093783543411e-05,
      "loss": 0.8303,
      "mean_token_accuracy": 0.7620800398290157,
      "num_tokens": 97745220.0,
      "step": 3063
    },
    {
      "entropy": 0.7799400016665459,
      "epoch": 0.28190752209590514,
      "grad_norm": 0.15435372292995453,
      "learning_rate": 9.060631152819947e-05,
      "loss": 0.7895,
      "mean_token_accuracy": 0.7704638950526714,
      "num_tokens": 97777634.0,
      "step": 3064
    },
    {
      "entropy": 0.7662400212138891,
      "epoch": 0.2819995284673464,
      "grad_norm": 0.15818148851394653,
      "learning_rate": 9.060324470205784e-05,
      "loss": 0.7867,
      "mean_token_accuracy": 0.7753710113465786,
      "num_tokens": 97809691.0,
      "step": 3065
    },
    {
      "entropy": 0.8174676727503538,
      "epoch": 0.28209153483878757,
      "grad_norm": 0.15302205085754395,
      "learning_rate": 9.060017787591622e-05,
      "loss": 0.798,
      "mean_token_accuracy": 0.7655254118144512,
      "num_tokens": 97841987.0,
      "step": 3066
    },
    {
      "entropy": 0.9534072820097208,
      "epoch": 0.2821835412102288,
      "grad_norm": 0.16089847683906555,
      "learning_rate": 9.05971110497746e-05,
      "loss": 0.9536,
      "mean_token_accuracy": 0.7315220758318901,
      "num_tokens": 97874530.0,
      "step": 3067
    },
    {
      "entropy": 0.8149995841085911,
      "epoch": 0.28227554758167,
      "grad_norm": 0.16063201427459717,
      "learning_rate": 9.059404422363297e-05,
      "loss": 0.8177,
      "mean_token_accuracy": 0.7629222795367241,
      "num_tokens": 97906517.0,
      "step": 3068
    },
    {
      "entropy": 0.8324655145406723,
      "epoch": 0.28236755395311125,
      "grad_norm": 0.16079318523406982,
      "learning_rate": 9.059097739749134e-05,
      "loss": 0.839,
      "mean_token_accuracy": 0.7547195442020893,
      "num_tokens": 97937548.0,
      "step": 3069
    },
    {
      "entropy": 0.8642386943101883,
      "epoch": 0.2824595603245525,
      "grad_norm": 0.1580801159143448,
      "learning_rate": 9.05879105713497e-05,
      "loss": 0.8412,
      "mean_token_accuracy": 0.7550255581736565,
      "num_tokens": 97969143.0,
      "step": 3070
    },
    {
      "entropy": 0.8299300596117973,
      "epoch": 0.2825515666959937,
      "grad_norm": 0.16148485243320465,
      "learning_rate": 9.058484374520809e-05,
      "loss": 0.8098,
      "mean_token_accuracy": 0.7669722102582455,
      "num_tokens": 98000965.0,
      "step": 3071
    },
    {
      "entropy": 0.9005648866295815,
      "epoch": 0.28264357306743493,
      "grad_norm": 0.16422882676124573,
      "learning_rate": 9.058177691906647e-05,
      "loss": 0.8902,
      "mean_token_accuracy": 0.7457572259008884,
      "num_tokens": 98033260.0,
      "step": 3072
    },
    {
      "entropy": 0.8796969074755907,
      "epoch": 0.2827355794388761,
      "grad_norm": 0.16038139164447784,
      "learning_rate": 9.057871009292484e-05,
      "loss": 0.8418,
      "mean_token_accuracy": 0.7581736892461777,
      "num_tokens": 98064984.0,
      "step": 3073
    },
    {
      "entropy": 0.8651132434606552,
      "epoch": 0.28282758581031736,
      "grad_norm": 0.16356372833251953,
      "learning_rate": 9.05756432667832e-05,
      "loss": 0.8717,
      "mean_token_accuracy": 0.7478417754173279,
      "num_tokens": 98095868.0,
      "step": 3074
    },
    {
      "entropy": 0.7930840700864792,
      "epoch": 0.2829195921817586,
      "grad_norm": 0.1515609174966812,
      "learning_rate": 9.057257644064158e-05,
      "loss": 0.7778,
      "mean_token_accuracy": 0.7674641124904156,
      "num_tokens": 98127444.0,
      "step": 3075
    },
    {
      "entropy": 0.8619761653244495,
      "epoch": 0.2830115985531998,
      "grad_norm": 0.15919014811515808,
      "learning_rate": 9.056950961449995e-05,
      "loss": 0.8613,
      "mean_token_accuracy": 0.7554102391004562,
      "num_tokens": 98159740.0,
      "step": 3076
    },
    {
      "entropy": 0.8028708919882774,
      "epoch": 0.28310360492464104,
      "grad_norm": 0.15689590573310852,
      "learning_rate": 9.056644278835833e-05,
      "loss": 0.8064,
      "mean_token_accuracy": 0.7698930278420448,
      "num_tokens": 98192491.0,
      "step": 3077
    },
    {
      "entropy": 0.9225743003189564,
      "epoch": 0.28319561129608223,
      "grad_norm": 0.16179214417934418,
      "learning_rate": 9.05633759622167e-05,
      "loss": 0.9223,
      "mean_token_accuracy": 0.7406904250383377,
      "num_tokens": 98224647.0,
      "step": 3078
    },
    {
      "entropy": 0.7647452671080828,
      "epoch": 0.2832876176675235,
      "grad_norm": 0.15198369324207306,
      "learning_rate": 9.056030913607508e-05,
      "loss": 0.7501,
      "mean_token_accuracy": 0.7847459651529789,
      "num_tokens": 98257415.0,
      "step": 3079
    },
    {
      "entropy": 0.9177891463041306,
      "epoch": 0.2833796240389647,
      "grad_norm": 0.16940373182296753,
      "learning_rate": 9.055724230993345e-05,
      "loss": 0.9363,
      "mean_token_accuracy": 0.7325453124940395,
      "num_tokens": 98288948.0,
      "step": 3080
    },
    {
      "entropy": 0.7609867490828037,
      "epoch": 0.2834716304104059,
      "grad_norm": 0.16112639009952545,
      "learning_rate": 9.055417548379183e-05,
      "loss": 0.7942,
      "mean_token_accuracy": 0.7682326473295689,
      "num_tokens": 98321475.0,
      "step": 3081
    },
    {
      "entropy": 0.836707029491663,
      "epoch": 0.28356363678184715,
      "grad_norm": 0.16386058926582336,
      "learning_rate": 9.055110865765021e-05,
      "loss": 0.8269,
      "mean_token_accuracy": 0.7591743320226669,
      "num_tokens": 98352572.0,
      "step": 3082
    },
    {
      "entropy": 0.8032928518950939,
      "epoch": 0.28365564315328834,
      "grad_norm": 0.15402673184871674,
      "learning_rate": 9.054804183150858e-05,
      "loss": 0.7981,
      "mean_token_accuracy": 0.7737603895366192,
      "num_tokens": 98384812.0,
      "step": 3083
    },
    {
      "entropy": 0.8499354403465986,
      "epoch": 0.2837476495247296,
      "grad_norm": 0.15929797291755676,
      "learning_rate": 9.054497500536695e-05,
      "loss": 0.8609,
      "mean_token_accuracy": 0.7508604116737843,
      "num_tokens": 98417297.0,
      "step": 3084
    },
    {
      "entropy": 0.810573399066925,
      "epoch": 0.28383965589617083,
      "grad_norm": 0.15518584847450256,
      "learning_rate": 9.054190817922532e-05,
      "loss": 0.7833,
      "mean_token_accuracy": 0.7707192599773407,
      "num_tokens": 98449583.0,
      "step": 3085
    },
    {
      "entropy": 0.7879952937364578,
      "epoch": 0.283931662267612,
      "grad_norm": 0.15002979338169098,
      "learning_rate": 9.05388413530837e-05,
      "loss": 0.7678,
      "mean_token_accuracy": 0.7791833467781544,
      "num_tokens": 98481880.0,
      "step": 3086
    },
    {
      "entropy": 0.7806288655847311,
      "epoch": 0.28402366863905326,
      "grad_norm": 0.14743085205554962,
      "learning_rate": 9.053577452694208e-05,
      "loss": 0.7323,
      "mean_token_accuracy": 0.7849676348268986,
      "num_tokens": 98513224.0,
      "step": 3087
    },
    {
      "entropy": 0.8469027169048786,
      "epoch": 0.28411567501049445,
      "grad_norm": 0.15564970672130585,
      "learning_rate": 9.053270770080045e-05,
      "loss": 0.8235,
      "mean_token_accuracy": 0.7540691643953323,
      "num_tokens": 98545214.0,
      "step": 3088
    },
    {
      "entropy": 0.8932507634162903,
      "epoch": 0.2842076813819357,
      "grad_norm": 0.1692371517419815,
      "learning_rate": 9.052964087465882e-05,
      "loss": 0.9027,
      "mean_token_accuracy": 0.7469326891005039,
      "num_tokens": 98576557.0,
      "step": 3089
    },
    {
      "entropy": 0.8761426098644733,
      "epoch": 0.28429968775337694,
      "grad_norm": 0.15378201007843018,
      "learning_rate": 9.05265740485172e-05,
      "loss": 0.8704,
      "mean_token_accuracy": 0.7483561374247074,
      "num_tokens": 98608677.0,
      "step": 3090
    },
    {
      "entropy": 0.8269048593938351,
      "epoch": 0.28439169412481813,
      "grad_norm": 0.1552077978849411,
      "learning_rate": 9.052350722237557e-05,
      "loss": 0.8306,
      "mean_token_accuracy": 0.757621992379427,
      "num_tokens": 98640232.0,
      "step": 3091
    },
    {
      "entropy": 0.8117869514971972,
      "epoch": 0.2844837004962594,
      "grad_norm": 0.1597934365272522,
      "learning_rate": 9.052044039623395e-05,
      "loss": 0.8182,
      "mean_token_accuracy": 0.7677733525633812,
      "num_tokens": 98670894.0,
      "step": 3092
    },
    {
      "entropy": 0.802549721673131,
      "epoch": 0.28457570686770056,
      "grad_norm": 0.15804684162139893,
      "learning_rate": 9.051737357009231e-05,
      "loss": 0.7812,
      "mean_token_accuracy": 0.7715202569961548,
      "num_tokens": 98703300.0,
      "step": 3093
    },
    {
      "entropy": 0.8448434248566628,
      "epoch": 0.2846677132391418,
      "grad_norm": 0.1557283103466034,
      "learning_rate": 9.05143067439507e-05,
      "loss": 0.8498,
      "mean_token_accuracy": 0.7547202594578266,
      "num_tokens": 98735346.0,
      "step": 3094
    },
    {
      "entropy": 0.8748458810150623,
      "epoch": 0.28475971961058305,
      "grad_norm": 0.15854433178901672,
      "learning_rate": 9.051123991780906e-05,
      "loss": 0.8819,
      "mean_token_accuracy": 0.7493801862001419,
      "num_tokens": 98767853.0,
      "step": 3095
    },
    {
      "entropy": 0.7921118885278702,
      "epoch": 0.28485172598202424,
      "grad_norm": 0.16472604870796204,
      "learning_rate": 9.050817309166743e-05,
      "loss": 0.8245,
      "mean_token_accuracy": 0.7597777023911476,
      "num_tokens": 98799964.0,
      "step": 3096
    },
    {
      "entropy": 0.8888745531439781,
      "epoch": 0.2849437323534655,
      "grad_norm": 0.17468230426311493,
      "learning_rate": 9.050510626552581e-05,
      "loss": 0.9044,
      "mean_token_accuracy": 0.7406248860061169,
      "num_tokens": 98830925.0,
      "step": 3097
    },
    {
      "entropy": 0.7907113265246153,
      "epoch": 0.2850357387249067,
      "grad_norm": 0.16478045284748077,
      "learning_rate": 9.05020394393842e-05,
      "loss": 0.8123,
      "mean_token_accuracy": 0.7664304450154305,
      "num_tokens": 98863175.0,
      "step": 3098
    },
    {
      "entropy": 0.8957884609699249,
      "epoch": 0.2851277450963479,
      "grad_norm": 0.16203847527503967,
      "learning_rate": 9.049897261324256e-05,
      "loss": 0.9008,
      "mean_token_accuracy": 0.7383261807262897,
      "num_tokens": 98894360.0,
      "step": 3099
    },
    {
      "entropy": 0.934861347079277,
      "epoch": 0.28521975146778916,
      "grad_norm": 0.16158874332904816,
      "learning_rate": 9.049590578710093e-05,
      "loss": 0.9106,
      "mean_token_accuracy": 0.7407253086566925,
      "num_tokens": 98926089.0,
      "step": 3100
    },
    {
      "entropy": 0.8484886288642883,
      "epoch": 0.28531175783923035,
      "grad_norm": 0.17152518033981323,
      "learning_rate": 9.04928389609593e-05,
      "loss": 0.8603,
      "mean_token_accuracy": 0.7599891647696495,
      "num_tokens": 98958227.0,
      "step": 3101
    },
    {
      "entropy": 0.8504353277385235,
      "epoch": 0.2854037642106716,
      "grad_norm": 0.15867352485656738,
      "learning_rate": 9.048977213481768e-05,
      "loss": 0.823,
      "mean_token_accuracy": 0.7667888849973679,
      "num_tokens": 98989572.0,
      "step": 3102
    },
    {
      "entropy": 0.7980418540537357,
      "epoch": 0.2854957705821128,
      "grad_norm": 0.15400411188602448,
      "learning_rate": 9.048670530867606e-05,
      "loss": 0.7783,
      "mean_token_accuracy": 0.7748725637793541,
      "num_tokens": 99021186.0,
      "step": 3103
    },
    {
      "entropy": 0.8563730400055647,
      "epoch": 0.28558777695355403,
      "grad_norm": 0.15698377788066864,
      "learning_rate": 9.048363848253443e-05,
      "loss": 0.8549,
      "mean_token_accuracy": 0.752392441034317,
      "num_tokens": 99052942.0,
      "step": 3104
    },
    {
      "entropy": 0.8982325363904238,
      "epoch": 0.2856797833249953,
      "grad_norm": 0.16315749287605286,
      "learning_rate": 9.04805716563928e-05,
      "loss": 0.883,
      "mean_token_accuracy": 0.7439652718603611,
      "num_tokens": 99083816.0,
      "step": 3105
    },
    {
      "entropy": 0.9067273437976837,
      "epoch": 0.28577178969643646,
      "grad_norm": 0.1668548285961151,
      "learning_rate": 9.047750483025118e-05,
      "loss": 0.9069,
      "mean_token_accuracy": 0.743856891989708,
      "num_tokens": 99115222.0,
      "step": 3106
    },
    {
      "entropy": 0.8302781768143177,
      "epoch": 0.2858637960678777,
      "grad_norm": 0.1559470146894455,
      "learning_rate": 9.047443800410955e-05,
      "loss": 0.8076,
      "mean_token_accuracy": 0.7674002721905708,
      "num_tokens": 99147572.0,
      "step": 3107
    },
    {
      "entropy": 0.8517344072461128,
      "epoch": 0.2859558024393189,
      "grad_norm": 0.1592833548784256,
      "learning_rate": 9.047137117796793e-05,
      "loss": 0.85,
      "mean_token_accuracy": 0.7574571371078491,
      "num_tokens": 99178955.0,
      "step": 3108
    },
    {
      "entropy": 0.8412882499396801,
      "epoch": 0.28604780881076014,
      "grad_norm": 0.16000409424304962,
      "learning_rate": 9.04683043518263e-05,
      "loss": 0.8605,
      "mean_token_accuracy": 0.755398653447628,
      "num_tokens": 99211019.0,
      "step": 3109
    },
    {
      "entropy": 0.811328299343586,
      "epoch": 0.2861398151822014,
      "grad_norm": 0.14923787117004395,
      "learning_rate": 9.046523752568468e-05,
      "loss": 0.8138,
      "mean_token_accuracy": 0.7661323547363281,
      "num_tokens": 99243154.0,
      "step": 3110
    },
    {
      "entropy": 0.7492274511605501,
      "epoch": 0.2862318215536426,
      "grad_norm": 0.1505098044872284,
      "learning_rate": 9.046217069954304e-05,
      "loss": 0.7332,
      "mean_token_accuracy": 0.7887174002826214,
      "num_tokens": 99275104.0,
      "step": 3111
    },
    {
      "entropy": 0.798794336616993,
      "epoch": 0.2863238279250838,
      "grad_norm": 0.15642811357975006,
      "learning_rate": 9.045910387340141e-05,
      "loss": 0.7792,
      "mean_token_accuracy": 0.7716614343225956,
      "num_tokens": 99307869.0,
      "step": 3112
    },
    {
      "entropy": 0.9306384548544884,
      "epoch": 0.286415834296525,
      "grad_norm": 0.16430804133415222,
      "learning_rate": 9.045603704725981e-05,
      "loss": 0.9171,
      "mean_token_accuracy": 0.7424902096390724,
      "num_tokens": 99340173.0,
      "step": 3113
    },
    {
      "entropy": 0.8533012494444847,
      "epoch": 0.28650784066796625,
      "grad_norm": 0.15815667808055878,
      "learning_rate": 9.045297022111818e-05,
      "loss": 0.8576,
      "mean_token_accuracy": 0.7527770921587944,
      "num_tokens": 99372770.0,
      "step": 3114
    },
    {
      "entropy": 0.742487357929349,
      "epoch": 0.2865998470394075,
      "grad_norm": 0.15554167330265045,
      "learning_rate": 9.044990339497654e-05,
      "loss": 0.739,
      "mean_token_accuracy": 0.7815855629742146,
      "num_tokens": 99405355.0,
      "step": 3115
    },
    {
      "entropy": 0.9041680805385113,
      "epoch": 0.2866918534108487,
      "grad_norm": 0.16381394863128662,
      "learning_rate": 9.044683656883491e-05,
      "loss": 0.9097,
      "mean_token_accuracy": 0.7415843792259693,
      "num_tokens": 99436881.0,
      "step": 3116
    },
    {
      "entropy": 0.8046462014317513,
      "epoch": 0.28678385978228993,
      "grad_norm": 0.16447478532791138,
      "learning_rate": 9.044376974269329e-05,
      "loss": 0.8135,
      "mean_token_accuracy": 0.7638048529624939,
      "num_tokens": 99468699.0,
      "step": 3117
    },
    {
      "entropy": 0.7955575007945299,
      "epoch": 0.2868758661537311,
      "grad_norm": 0.1626521497964859,
      "learning_rate": 9.044070291655167e-05,
      "loss": 0.8267,
      "mean_token_accuracy": 0.7605165429413319,
      "num_tokens": 99500495.0,
      "step": 3118
    },
    {
      "entropy": 0.7750002276152372,
      "epoch": 0.28696787252517236,
      "grad_norm": 0.1642167568206787,
      "learning_rate": 9.043763609041004e-05,
      "loss": 0.7792,
      "mean_token_accuracy": 0.770627286285162,
      "num_tokens": 99532239.0,
      "step": 3119
    },
    {
      "entropy": 0.9051864929497242,
      "epoch": 0.2870598788966136,
      "grad_norm": 0.1631658375263214,
      "learning_rate": 9.043456926426841e-05,
      "loss": 0.9229,
      "mean_token_accuracy": 0.742281436920166,
      "num_tokens": 99564586.0,
      "step": 3120
    },
    {
      "entropy": 0.7148022688925266,
      "epoch": 0.2871518852680548,
      "grad_norm": 0.15899744629859924,
      "learning_rate": 9.043150243812679e-05,
      "loss": 0.725,
      "mean_token_accuracy": 0.7893040999770164,
      "num_tokens": 99596415.0,
      "step": 3121
    },
    {
      "entropy": 0.8738234303891659,
      "epoch": 0.28724389163949604,
      "grad_norm": 0.1575319766998291,
      "learning_rate": 9.042843561198516e-05,
      "loss": 0.8632,
      "mean_token_accuracy": 0.7496169731020927,
      "num_tokens": 99628689.0,
      "step": 3122
    },
    {
      "entropy": 0.8299461472779512,
      "epoch": 0.2873358980109373,
      "grad_norm": 0.15894727408885956,
      "learning_rate": 9.042536878584354e-05,
      "loss": 0.839,
      "mean_token_accuracy": 0.7582399100065231,
      "num_tokens": 99660230.0,
      "step": 3123
    },
    {
      "entropy": 0.902844401076436,
      "epoch": 0.2874279043823785,
      "grad_norm": 0.15762177109718323,
      "learning_rate": 9.042230195970191e-05,
      "loss": 0.8959,
      "mean_token_accuracy": 0.7449857071042061,
      "num_tokens": 99692163.0,
      "step": 3124
    },
    {
      "entropy": 0.8745828047394753,
      "epoch": 0.2875199107538197,
      "grad_norm": 0.15690195560455322,
      "learning_rate": 9.041923513356029e-05,
      "loss": 0.8675,
      "mean_token_accuracy": 0.750403918325901,
      "num_tokens": 99724072.0,
      "step": 3125
    },
    {
      "entropy": 0.7982974387705326,
      "epoch": 0.2876119171252609,
      "grad_norm": 0.1557229906320572,
      "learning_rate": 9.041616830741866e-05,
      "loss": 0.7942,
      "mean_token_accuracy": 0.7672330401837826,
      "num_tokens": 99756341.0,
      "step": 3126
    },
    {
      "entropy": 0.8323658034205437,
      "epoch": 0.28770392349670215,
      "grad_norm": 0.15505655109882355,
      "learning_rate": 9.041310148127703e-05,
      "loss": 0.7972,
      "mean_token_accuracy": 0.7691407725214958,
      "num_tokens": 99788273.0,
      "step": 3127
    },
    {
      "entropy": 0.7219066843390465,
      "epoch": 0.2877959298681434,
      "grad_norm": 0.15033182501792908,
      "learning_rate": 9.04100346551354e-05,
      "loss": 0.7086,
      "mean_token_accuracy": 0.7939893193542957,
      "num_tokens": 99820488.0,
      "step": 3128
    },
    {
      "entropy": 0.9335199818015099,
      "epoch": 0.2878879362395846,
      "grad_norm": 0.16144677996635437,
      "learning_rate": 9.040696782899379e-05,
      "loss": 0.93,
      "mean_token_accuracy": 0.7353764399886131,
      "num_tokens": 99852238.0,
      "step": 3129
    },
    {
      "entropy": 0.818684671074152,
      "epoch": 0.28797994261102583,
      "grad_norm": 0.1506928950548172,
      "learning_rate": 9.040390100285216e-05,
      "loss": 0.7931,
      "mean_token_accuracy": 0.771225243806839,
      "num_tokens": 99884405.0,
      "step": 3130
    },
    {
      "entropy": 0.8554952591657639,
      "epoch": 0.288071948982467,
      "grad_norm": 0.15345104038715363,
      "learning_rate": 9.040083417671052e-05,
      "loss": 0.8403,
      "mean_token_accuracy": 0.7608412466943264,
      "num_tokens": 99916444.0,
      "step": 3131
    },
    {
      "entropy": 0.8562560528516769,
      "epoch": 0.28816395535390826,
      "grad_norm": 0.16667422652244568,
      "learning_rate": 9.039776735056889e-05,
      "loss": 0.8952,
      "mean_token_accuracy": 0.7496655993163586,
      "num_tokens": 99948896.0,
      "step": 3132
    },
    {
      "entropy": 0.8484035655856133,
      "epoch": 0.2882559617253495,
      "grad_norm": 0.15848588943481445,
      "learning_rate": 9.039470052442727e-05,
      "loss": 0.8445,
      "mean_token_accuracy": 0.7608196809887886,
      "num_tokens": 99980729.0,
      "step": 3133
    },
    {
      "entropy": 0.7678321339190006,
      "epoch": 0.2883479680967907,
      "grad_norm": 0.15099336206912994,
      "learning_rate": 9.039163369828565e-05,
      "loss": 0.7562,
      "mean_token_accuracy": 0.7754376418888569,
      "num_tokens": 100012877.0,
      "step": 3134
    },
    {
      "entropy": 0.8506795130670071,
      "epoch": 0.28843997446823194,
      "grad_norm": 0.16113726794719696,
      "learning_rate": 9.038856687214402e-05,
      "loss": 0.8325,
      "mean_token_accuracy": 0.7597404643893242,
      "num_tokens": 100044568.0,
      "step": 3135
    },
    {
      "entropy": 0.812988517805934,
      "epoch": 0.28853198083967313,
      "grad_norm": 0.15608970820903778,
      "learning_rate": 9.038550004600239e-05,
      "loss": 0.7891,
      "mean_token_accuracy": 0.7712794169783592,
      "num_tokens": 100076323.0,
      "step": 3136
    },
    {
      "entropy": 0.9145891591906548,
      "epoch": 0.2886239872111144,
      "grad_norm": 0.16039247810840607,
      "learning_rate": 9.038243321986077e-05,
      "loss": 0.9037,
      "mean_token_accuracy": 0.7427836582064629,
      "num_tokens": 100108558.0,
      "step": 3137
    },
    {
      "entropy": 0.8768434785306454,
      "epoch": 0.2887159935825556,
      "grad_norm": 0.16368046402931213,
      "learning_rate": 9.037936639371914e-05,
      "loss": 0.8797,
      "mean_token_accuracy": 0.7511985041201115,
      "num_tokens": 100140090.0,
      "step": 3138
    },
    {
      "entropy": 0.8976588435471058,
      "epoch": 0.2888079999539968,
      "grad_norm": 0.1664142906665802,
      "learning_rate": 9.037629956757752e-05,
      "loss": 0.9243,
      "mean_token_accuracy": 0.7401230186223984,
      "num_tokens": 100171801.0,
      "step": 3139
    },
    {
      "entropy": 0.7624607719480991,
      "epoch": 0.28890000632543805,
      "grad_norm": 0.15687142312526703,
      "learning_rate": 9.037323274143589e-05,
      "loss": 0.735,
      "mean_token_accuracy": 0.7891881018877029,
      "num_tokens": 100204565.0,
      "step": 3140
    },
    {
      "entropy": 0.9070580918341875,
      "epoch": 0.28899201269687924,
      "grad_norm": 0.16549333930015564,
      "learning_rate": 9.037016591529427e-05,
      "loss": 0.925,
      "mean_token_accuracy": 0.7363887317478657,
      "num_tokens": 100235869.0,
      "step": 3141
    },
    {
      "entropy": 0.811367467045784,
      "epoch": 0.2890840190683205,
      "grad_norm": 0.15387707948684692,
      "learning_rate": 9.036709908915264e-05,
      "loss": 0.8077,
      "mean_token_accuracy": 0.7617871575057507,
      "num_tokens": 100267640.0,
      "step": 3142
    },
    {
      "entropy": 0.7766444925218821,
      "epoch": 0.28917602543976173,
      "grad_norm": 0.15814250707626343,
      "learning_rate": 9.0364032263011e-05,
      "loss": 0.7795,
      "mean_token_accuracy": 0.775529507547617,
      "num_tokens": 100299818.0,
      "step": 3143
    },
    {
      "entropy": 0.8421365190297365,
      "epoch": 0.2892680318112029,
      "grad_norm": 0.15728148818016052,
      "learning_rate": 9.036096543686939e-05,
      "loss": 0.8478,
      "mean_token_accuracy": 0.7602249793708324,
      "num_tokens": 100332200.0,
      "step": 3144
    },
    {
      "entropy": 0.910310871899128,
      "epoch": 0.28936003818264416,
      "grad_norm": 0.16006198525428772,
      "learning_rate": 9.035789861072777e-05,
      "loss": 0.9137,
      "mean_token_accuracy": 0.7394065633416176,
      "num_tokens": 100364583.0,
      "step": 3145
    },
    {
      "entropy": 0.8640425130724907,
      "epoch": 0.28945204455408535,
      "grad_norm": 0.16084501147270203,
      "learning_rate": 9.035483178458614e-05,
      "loss": 0.8579,
      "mean_token_accuracy": 0.7512971237301826,
      "num_tokens": 100396122.0,
      "step": 3146
    },
    {
      "entropy": 0.782263170927763,
      "epoch": 0.2895440509255266,
      "grad_norm": 0.15656329691410065,
      "learning_rate": 9.03517649584445e-05,
      "loss": 0.7556,
      "mean_token_accuracy": 0.7768679037690163,
      "num_tokens": 100427824.0,
      "step": 3147
    },
    {
      "entropy": 0.8281490355730057,
      "epoch": 0.28963605729696784,
      "grad_norm": 0.15862619876861572,
      "learning_rate": 9.034869813230289e-05,
      "loss": 0.8044,
      "mean_token_accuracy": 0.7640481293201447,
      "num_tokens": 100459580.0,
      "step": 3148
    },
    {
      "entropy": 0.721054220572114,
      "epoch": 0.28972806366840903,
      "grad_norm": 0.1540256291627884,
      "learning_rate": 9.034563130616125e-05,
      "loss": 0.7245,
      "mean_token_accuracy": 0.7868707738816738,
      "num_tokens": 100491559.0,
      "step": 3149
    },
    {
      "entropy": 0.8296151086688042,
      "epoch": 0.2898200700398503,
      "grad_norm": 0.1717168390750885,
      "learning_rate": 9.034256448001964e-05,
      "loss": 0.8483,
      "mean_token_accuracy": 0.7627076655626297,
      "num_tokens": 100523595.0,
      "step": 3150
    },
    {
      "entropy": 0.8158803656697273,
      "epoch": 0.28991207641129146,
      "grad_norm": 0.1532531976699829,
      "learning_rate": 9.0339497653878e-05,
      "loss": 0.8001,
      "mean_token_accuracy": 0.7673046439886093,
      "num_tokens": 100556017.0,
      "step": 3151
    },
    {
      "entropy": 0.8360585309565067,
      "epoch": 0.2900040827827327,
      "grad_norm": 0.15269051492214203,
      "learning_rate": 9.033643082773638e-05,
      "loss": 0.8035,
      "mean_token_accuracy": 0.7627291195094585,
      "num_tokens": 100587575.0,
      "step": 3152
    },
    {
      "entropy": 0.8702558595687151,
      "epoch": 0.29009608915417395,
      "grad_norm": 0.15740007162094116,
      "learning_rate": 9.033336400159475e-05,
      "loss": 0.8343,
      "mean_token_accuracy": 0.7569335848093033,
      "num_tokens": 100619694.0,
      "step": 3153
    },
    {
      "entropy": 0.7989114299416542,
      "epoch": 0.29018809552561514,
      "grad_norm": 0.1598055362701416,
      "learning_rate": 9.033029717545312e-05,
      "loss": 0.7955,
      "mean_token_accuracy": 0.7701336927711964,
      "num_tokens": 100651304.0,
      "step": 3154
    },
    {
      "entropy": 0.8284060247242451,
      "epoch": 0.2902801018970564,
      "grad_norm": 0.15269698202610016,
      "learning_rate": 9.03272303493115e-05,
      "loss": 0.8352,
      "mean_token_accuracy": 0.7624961100518703,
      "num_tokens": 100683360.0,
      "step": 3155
    },
    {
      "entropy": 0.7458656132221222,
      "epoch": 0.2903721082684976,
      "grad_norm": 0.15668541193008423,
      "learning_rate": 9.032416352316988e-05,
      "loss": 0.7614,
      "mean_token_accuracy": 0.7805015929043293,
      "num_tokens": 100714675.0,
      "step": 3156
    },
    {
      "entropy": 0.7555865067988634,
      "epoch": 0.2904641146399388,
      "grad_norm": 0.15557244420051575,
      "learning_rate": 9.032109669702825e-05,
      "loss": 0.7405,
      "mean_token_accuracy": 0.7833556234836578,
      "num_tokens": 100746488.0,
      "step": 3157
    },
    {
      "entropy": 0.7710688766092062,
      "epoch": 0.29055612101138006,
      "grad_norm": 0.1564849615097046,
      "learning_rate": 9.031802987088662e-05,
      "loss": 0.7742,
      "mean_token_accuracy": 0.7772331908345222,
      "num_tokens": 100779056.0,
      "step": 3158
    },
    {
      "entropy": 0.8386783562600613,
      "epoch": 0.29064812738282125,
      "grad_norm": 0.16236209869384766,
      "learning_rate": 9.031496304474499e-05,
      "loss": 0.8389,
      "mean_token_accuracy": 0.761837150901556,
      "num_tokens": 100810594.0,
      "step": 3159
    },
    {
      "entropy": 0.8218664303421974,
      "epoch": 0.2907401337542625,
      "grad_norm": 0.15634682774543762,
      "learning_rate": 9.031189621860338e-05,
      "loss": 0.8091,
      "mean_token_accuracy": 0.7645307630300522,
      "num_tokens": 100842893.0,
      "step": 3160
    },
    {
      "entropy": 0.8912910260260105,
      "epoch": 0.2908321401257037,
      "grad_norm": 0.16268974542617798,
      "learning_rate": 9.030882939246175e-05,
      "loss": 0.8876,
      "mean_token_accuracy": 0.7446510829031467,
      "num_tokens": 100874917.0,
      "step": 3161
    },
    {
      "entropy": 0.7330492604523897,
      "epoch": 0.29092414649714493,
      "grad_norm": 0.15575917065143585,
      "learning_rate": 9.030576256632012e-05,
      "loss": 0.7427,
      "mean_token_accuracy": 0.7834632061421871,
      "num_tokens": 100907268.0,
      "step": 3162
    },
    {
      "entropy": 0.8264786750078201,
      "epoch": 0.2910161528685862,
      "grad_norm": 0.1558249592781067,
      "learning_rate": 9.030269574017849e-05,
      "loss": 0.8243,
      "mean_token_accuracy": 0.7627252340316772,
      "num_tokens": 100939116.0,
      "step": 3163
    },
    {
      "entropy": 0.8212195970118046,
      "epoch": 0.29110815924002736,
      "grad_norm": 0.15763980150222778,
      "learning_rate": 9.029962891403687e-05,
      "loss": 0.805,
      "mean_token_accuracy": 0.7695707604289055,
      "num_tokens": 100971523.0,
      "step": 3164
    },
    {
      "entropy": 0.9046418331563473,
      "epoch": 0.2912001656114686,
      "grad_norm": 0.15962903201580048,
      "learning_rate": 9.029656208789525e-05,
      "loss": 0.8955,
      "mean_token_accuracy": 0.7437741085886955,
      "num_tokens": 101003621.0,
      "step": 3165
    },
    {
      "entropy": 0.923548512160778,
      "epoch": 0.2912921719829098,
      "grad_norm": 0.15732011198997498,
      "learning_rate": 9.029349526175362e-05,
      "loss": 0.9152,
      "mean_token_accuracy": 0.7369136065244675,
      "num_tokens": 101035347.0,
      "step": 3166
    },
    {
      "entropy": 0.8744808994233608,
      "epoch": 0.29138417835435104,
      "grad_norm": 0.15867945551872253,
      "learning_rate": 9.029042843561198e-05,
      "loss": 0.8685,
      "mean_token_accuracy": 0.7499292604625225,
      "num_tokens": 101067417.0,
      "step": 3167
    },
    {
      "entropy": 0.9350709542632103,
      "epoch": 0.2914761847257923,
      "grad_norm": 0.15822337567806244,
      "learning_rate": 9.028736160947037e-05,
      "loss": 0.9177,
      "mean_token_accuracy": 0.7406025230884552,
      "num_tokens": 101099286.0,
      "step": 3168
    },
    {
      "entropy": 0.8968679755926132,
      "epoch": 0.2915681910972335,
      "grad_norm": 0.15883475542068481,
      "learning_rate": 9.028429478332873e-05,
      "loss": 0.9028,
      "mean_token_accuracy": 0.7410541959106922,
      "num_tokens": 101131430.0,
      "step": 3169
    },
    {
      "entropy": 0.8244980126619339,
      "epoch": 0.2916601974686747,
      "grad_norm": 0.16238760948181152,
      "learning_rate": 9.028122795718711e-05,
      "loss": 0.8263,
      "mean_token_accuracy": 0.7644683197140694,
      "num_tokens": 101163920.0,
      "step": 3170
    },
    {
      "entropy": 0.9581815600395203,
      "epoch": 0.2917522038401159,
      "grad_norm": 0.1676633656024933,
      "learning_rate": 9.027816113104548e-05,
      "loss": 0.984,
      "mean_token_accuracy": 0.7196622900664806,
      "num_tokens": 101195755.0,
      "step": 3171
    },
    {
      "entropy": 0.9230419769883156,
      "epoch": 0.29184421021155715,
      "grad_norm": 0.15758293867111206,
      "learning_rate": 9.027509430490386e-05,
      "loss": 0.94,
      "mean_token_accuracy": 0.7371174618601799,
      "num_tokens": 101228323.0,
      "step": 3172
    },
    {
      "entropy": 0.7674350906163454,
      "epoch": 0.2919362165829984,
      "grad_norm": 0.152325838804245,
      "learning_rate": 9.027202747876223e-05,
      "loss": 0.7463,
      "mean_token_accuracy": 0.7821067869663239,
      "num_tokens": 101259955.0,
      "step": 3173
    },
    {
      "entropy": 0.8236006014049053,
      "epoch": 0.2920282229544396,
      "grad_norm": 0.1567000448703766,
      "learning_rate": 9.02689606526206e-05,
      "loss": 0.8186,
      "mean_token_accuracy": 0.7614258453249931,
      "num_tokens": 101290366.0,
      "step": 3174
    },
    {
      "entropy": 0.9573616608977318,
      "epoch": 0.29212022932588083,
      "grad_norm": 0.15783348679542542,
      "learning_rate": 9.026589382647898e-05,
      "loss": 0.9463,
      "mean_token_accuracy": 0.7332042567431927,
      "num_tokens": 101322709.0,
      "step": 3175
    },
    {
      "entropy": 0.7762656304985285,
      "epoch": 0.292212235697322,
      "grad_norm": 0.1583714783191681,
      "learning_rate": 9.026282700033736e-05,
      "loss": 0.7658,
      "mean_token_accuracy": 0.7765906043350697,
      "num_tokens": 101354528.0,
      "step": 3176
    },
    {
      "entropy": 0.8411923963576555,
      "epoch": 0.29230424206876326,
      "grad_norm": 0.15639646351337433,
      "learning_rate": 9.025976017419573e-05,
      "loss": 0.8191,
      "mean_token_accuracy": 0.761646531522274,
      "num_tokens": 101386628.0,
      "step": 3177
    },
    {
      "entropy": 0.8583645783364773,
      "epoch": 0.2923962484402045,
      "grad_norm": 0.1621035784482956,
      "learning_rate": 9.02566933480541e-05,
      "loss": 0.8612,
      "mean_token_accuracy": 0.7497067227959633,
      "num_tokens": 101418760.0,
      "step": 3178
    },
    {
      "entropy": 0.8904569037258625,
      "epoch": 0.2924882548116457,
      "grad_norm": 0.15978878736495972,
      "learning_rate": 9.025362652191248e-05,
      "loss": 0.8917,
      "mean_token_accuracy": 0.7445063628256321,
      "num_tokens": 101451163.0,
      "step": 3179
    },
    {
      "entropy": 0.8493010457605124,
      "epoch": 0.29258026118308694,
      "grad_norm": 0.15679512917995453,
      "learning_rate": 9.025055969577085e-05,
      "loss": 0.831,
      "mean_token_accuracy": 0.7618917338550091,
      "num_tokens": 101483761.0,
      "step": 3180
    },
    {
      "entropy": 0.8157441765069962,
      "epoch": 0.29267226755452813,
      "grad_norm": 0.15882784128189087,
      "learning_rate": 9.024749286962923e-05,
      "loss": 0.8259,
      "mean_token_accuracy": 0.7590639181435108,
      "num_tokens": 101515598.0,
      "step": 3181
    },
    {
      "entropy": 0.8992273658514023,
      "epoch": 0.2927642739259694,
      "grad_norm": 0.1623137891292572,
      "learning_rate": 9.02444260434876e-05,
      "loss": 0.8931,
      "mean_token_accuracy": 0.7444291077554226,
      "num_tokens": 101547985.0,
      "step": 3182
    },
    {
      "entropy": 0.8004123959690332,
      "epoch": 0.2928562802974106,
      "grad_norm": 0.15461677312850952,
      "learning_rate": 9.024135921734598e-05,
      "loss": 0.7981,
      "mean_token_accuracy": 0.7681717798113823,
      "num_tokens": 101580072.0,
      "step": 3183
    },
    {
      "entropy": 0.9203095436096191,
      "epoch": 0.2929482866688518,
      "grad_norm": 0.16571903228759766,
      "learning_rate": 9.023829239120435e-05,
      "loss": 0.9279,
      "mean_token_accuracy": 0.7338174767792225,
      "num_tokens": 101611755.0,
      "step": 3184
    },
    {
      "entropy": 0.8256703801453114,
      "epoch": 0.29304029304029305,
      "grad_norm": 0.15596750378608704,
      "learning_rate": 9.023522556506271e-05,
      "loss": 0.835,
      "mean_token_accuracy": 0.7611192874610424,
      "num_tokens": 101644395.0,
      "step": 3185
    },
    {
      "entropy": 0.796425387263298,
      "epoch": 0.29313229941173424,
      "grad_norm": 0.16226376593112946,
      "learning_rate": 9.02321587389211e-05,
      "loss": 0.7959,
      "mean_token_accuracy": 0.7684643119573593,
      "num_tokens": 101676282.0,
      "step": 3186
    },
    {
      "entropy": 0.8502112217247486,
      "epoch": 0.2932243057831755,
      "grad_norm": 0.15837891399860382,
      "learning_rate": 9.022909191277948e-05,
      "loss": 0.8423,
      "mean_token_accuracy": 0.757843978703022,
      "num_tokens": 101707993.0,
      "step": 3187
    },
    {
      "entropy": 0.8603754565119743,
      "epoch": 0.29331631215461673,
      "grad_norm": 0.15633974969387054,
      "learning_rate": 9.022602508663784e-05,
      "loss": 0.8447,
      "mean_token_accuracy": 0.759591206908226,
      "num_tokens": 101739114.0,
      "step": 3188
    },
    {
      "entropy": 0.8135569915175438,
      "epoch": 0.2934083185260579,
      "grad_norm": 0.15791133046150208,
      "learning_rate": 9.022295826049621e-05,
      "loss": 0.8018,
      "mean_token_accuracy": 0.7678660489618778,
      "num_tokens": 101770845.0,
      "step": 3189
    },
    {
      "entropy": 0.7692399229854345,
      "epoch": 0.29350032489749917,
      "grad_norm": 0.15273478627204895,
      "learning_rate": 9.021989143435458e-05,
      "loss": 0.7319,
      "mean_token_accuracy": 0.7896382436156273,
      "num_tokens": 101802011.0,
      "step": 3190
    },
    {
      "entropy": 0.8852591793984175,
      "epoch": 0.29359233126894035,
      "grad_norm": 0.15941964089870453,
      "learning_rate": 9.021682460821296e-05,
      "loss": 0.8623,
      "mean_token_accuracy": 0.7500354461371899,
      "num_tokens": 101833103.0,
      "step": 3191
    },
    {
      "entropy": 0.7618223074823618,
      "epoch": 0.2936843376403816,
      "grad_norm": 0.15870916843414307,
      "learning_rate": 9.021375778207134e-05,
      "loss": 0.7544,
      "mean_token_accuracy": 0.7763308882713318,
      "num_tokens": 101864513.0,
      "step": 3192
    },
    {
      "entropy": 0.790227148681879,
      "epoch": 0.29377634401182284,
      "grad_norm": 0.1569926142692566,
      "learning_rate": 9.021069095592971e-05,
      "loss": 0.7764,
      "mean_token_accuracy": 0.775145635008812,
      "num_tokens": 101896846.0,
      "step": 3193
    },
    {
      "entropy": 0.7326594367623329,
      "epoch": 0.29386835038326403,
      "grad_norm": 0.1598464995622635,
      "learning_rate": 9.020762412978808e-05,
      "loss": 0.7491,
      "mean_token_accuracy": 0.7835141941905022,
      "num_tokens": 101928808.0,
      "step": 3194
    },
    {
      "entropy": 0.8004922028630972,
      "epoch": 0.2939603567547053,
      "grad_norm": 0.14781293272972107,
      "learning_rate": 9.020455730364646e-05,
      "loss": 0.7889,
      "mean_token_accuracy": 0.7698757201433182,
      "num_tokens": 101961019.0,
      "step": 3195
    },
    {
      "entropy": 0.8491642493754625,
      "epoch": 0.29405236312614647,
      "grad_norm": 0.1701159030199051,
      "learning_rate": 9.020149047750483e-05,
      "loss": 0.8935,
      "mean_token_accuracy": 0.751355841755867,
      "num_tokens": 101993240.0,
      "step": 3196
    },
    {
      "entropy": 0.7565793059766293,
      "epoch": 0.2941443694975877,
      "grad_norm": 0.1574869602918625,
      "learning_rate": 9.019842365136321e-05,
      "loss": 0.7402,
      "mean_token_accuracy": 0.78076047077775,
      "num_tokens": 102025161.0,
      "step": 3197
    },
    {
      "entropy": 0.8273208439350128,
      "epoch": 0.29423637586902895,
      "grad_norm": 0.15284475684165955,
      "learning_rate": 9.019535682522158e-05,
      "loss": 0.8135,
      "mean_token_accuracy": 0.7646205611526966,
      "num_tokens": 102056947.0,
      "step": 3198
    },
    {
      "entropy": 0.7627965211868286,
      "epoch": 0.29432838224047014,
      "grad_norm": 0.15596170723438263,
      "learning_rate": 9.019228999907996e-05,
      "loss": 0.7563,
      "mean_token_accuracy": 0.7792799882590771,
      "num_tokens": 102089150.0,
      "step": 3199
    },
    {
      "entropy": 0.9121864065527916,
      "epoch": 0.2944203886119114,
      "grad_norm": 0.15551190078258514,
      "learning_rate": 9.018922317293833e-05,
      "loss": 0.9059,
      "mean_token_accuracy": 0.738679300993681,
      "num_tokens": 102121145.0,
      "step": 3200
    },
    {
      "entropy": 0.7961977049708366,
      "epoch": 0.2945123949833526,
      "grad_norm": 0.15908531844615936,
      "learning_rate": 9.018615634679671e-05,
      "loss": 0.7856,
      "mean_token_accuracy": 0.7710445411503315,
      "num_tokens": 102153118.0,
      "step": 3201
    },
    {
      "entropy": 0.8160314466804266,
      "epoch": 0.2946044013547938,
      "grad_norm": 0.1605740487575531,
      "learning_rate": 9.018308952065508e-05,
      "loss": 0.8235,
      "mean_token_accuracy": 0.761083547025919,
      "num_tokens": 102185319.0,
      "step": 3202
    },
    {
      "entropy": 0.7755609210580587,
      "epoch": 0.29469640772623507,
      "grad_norm": 0.15232640504837036,
      "learning_rate": 9.018002269451346e-05,
      "loss": 0.7536,
      "mean_token_accuracy": 0.78056875243783,
      "num_tokens": 102216576.0,
      "step": 3203
    },
    {
      "entropy": 0.886900432407856,
      "epoch": 0.29478841409767625,
      "grad_norm": 0.16412122547626495,
      "learning_rate": 9.017695586837183e-05,
      "loss": 0.8757,
      "mean_token_accuracy": 0.7495241351425648,
      "num_tokens": 102248540.0,
      "step": 3204
    },
    {
      "entropy": 0.7630243320018053,
      "epoch": 0.2948804204691175,
      "grad_norm": 0.1493261158466339,
      "learning_rate": 9.01738890422302e-05,
      "loss": 0.7423,
      "mean_token_accuracy": 0.7818890362977982,
      "num_tokens": 102281194.0,
      "step": 3205
    },
    {
      "entropy": 0.8741129636764526,
      "epoch": 0.2949724268405587,
      "grad_norm": 0.1720954030752182,
      "learning_rate": 9.017082221608858e-05,
      "loss": 0.9023,
      "mean_token_accuracy": 0.7533496953547001,
      "num_tokens": 102313052.0,
      "step": 3206
    },
    {
      "entropy": 0.8634685911238194,
      "epoch": 0.29506443321199993,
      "grad_norm": 0.16125012934207916,
      "learning_rate": 9.016775538994696e-05,
      "loss": 0.8375,
      "mean_token_accuracy": 0.7569495067000389,
      "num_tokens": 102344902.0,
      "step": 3207
    },
    {
      "entropy": 0.756110118702054,
      "epoch": 0.2951564395834412,
      "grad_norm": 0.15109305083751678,
      "learning_rate": 9.016468856380532e-05,
      "loss": 0.7457,
      "mean_token_accuracy": 0.7840436510741711,
      "num_tokens": 102377120.0,
      "step": 3208
    },
    {
      "entropy": 0.8631251789629459,
      "epoch": 0.29524844595488237,
      "grad_norm": 0.16220353543758392,
      "learning_rate": 9.016162173766369e-05,
      "loss": 0.8614,
      "mean_token_accuracy": 0.7536821439862251,
      "num_tokens": 102408815.0,
      "step": 3209
    },
    {
      "entropy": 0.7337331175804138,
      "epoch": 0.2953404523263236,
      "grad_norm": 0.15066653490066528,
      "learning_rate": 9.015855491152207e-05,
      "loss": 0.7173,
      "mean_token_accuracy": 0.7897795215249062,
      "num_tokens": 102440895.0,
      "step": 3210
    },
    {
      "entropy": 0.8524109628051519,
      "epoch": 0.2954324586977648,
      "grad_norm": 0.16262073814868927,
      "learning_rate": 9.015548808538044e-05,
      "loss": 0.8585,
      "mean_token_accuracy": 0.7527904510498047,
      "num_tokens": 102472809.0,
      "step": 3211
    },
    {
      "entropy": 0.8847377244383097,
      "epoch": 0.29552446506920604,
      "grad_norm": 0.16650919616222382,
      "learning_rate": 9.015242125923882e-05,
      "loss": 0.8892,
      "mean_token_accuracy": 0.7451476156711578,
      "num_tokens": 102504345.0,
      "step": 3212
    },
    {
      "entropy": 0.8598622307181358,
      "epoch": 0.2956164714406473,
      "grad_norm": 0.1609732061624527,
      "learning_rate": 9.014935443309719e-05,
      "loss": 0.8704,
      "mean_token_accuracy": 0.7526582144200802,
      "num_tokens": 102536389.0,
      "step": 3213
    },
    {
      "entropy": 0.7665574196726084,
      "epoch": 0.2957084778120885,
      "grad_norm": 0.1511075794696808,
      "learning_rate": 9.014628760695557e-05,
      "loss": 0.7671,
      "mean_token_accuracy": 0.7745878249406815,
      "num_tokens": 102568856.0,
      "step": 3214
    },
    {
      "entropy": 0.827289629727602,
      "epoch": 0.2958004841835297,
      "grad_norm": 0.16121956706047058,
      "learning_rate": 9.014322078081394e-05,
      "loss": 0.8011,
      "mean_token_accuracy": 0.7696695402264595,
      "num_tokens": 102601063.0,
      "step": 3215
    },
    {
      "entropy": 0.8316291570663452,
      "epoch": 0.2958924905549709,
      "grad_norm": 0.15940843522548676,
      "learning_rate": 9.014015395467231e-05,
      "loss": 0.8472,
      "mean_token_accuracy": 0.7566188611090183,
      "num_tokens": 102632334.0,
      "step": 3216
    },
    {
      "entropy": 0.8681637272238731,
      "epoch": 0.29598449692641216,
      "grad_norm": 0.1594361513853073,
      "learning_rate": 9.013708712853069e-05,
      "loss": 0.8778,
      "mean_token_accuracy": 0.7511197328567505,
      "num_tokens": 102664576.0,
      "step": 3217
    },
    {
      "entropy": 0.864403922110796,
      "epoch": 0.2960765032978534,
      "grad_norm": 0.15853577852249146,
      "learning_rate": 9.013402030238907e-05,
      "loss": 0.8461,
      "mean_token_accuracy": 0.753927867859602,
      "num_tokens": 102696217.0,
      "step": 3218
    },
    {
      "entropy": 0.8864081986248493,
      "epoch": 0.2961685096692946,
      "grad_norm": 0.26794782280921936,
      "learning_rate": 9.013095347624744e-05,
      "loss": 0.9104,
      "mean_token_accuracy": 0.7467159293591976,
      "num_tokens": 102728656.0,
      "step": 3219
    },
    {
      "entropy": 0.9669194929301739,
      "epoch": 0.29626051604073583,
      "grad_norm": 0.16423971951007843,
      "learning_rate": 9.01278866501058e-05,
      "loss": 0.9515,
      "mean_token_accuracy": 0.7314387038350105,
      "num_tokens": 102760728.0,
      "step": 3220
    },
    {
      "entropy": 0.8133777976036072,
      "epoch": 0.296352522412177,
      "grad_norm": 0.15917067229747772,
      "learning_rate": 9.012481982396417e-05,
      "loss": 0.8178,
      "mean_token_accuracy": 0.7681770361959934,
      "num_tokens": 102792318.0,
      "step": 3221
    },
    {
      "entropy": 0.866450846195221,
      "epoch": 0.29644452878361827,
      "grad_norm": 0.15675264596939087,
      "learning_rate": 9.012175299782256e-05,
      "loss": 0.8361,
      "mean_token_accuracy": 0.7607604898512363,
      "num_tokens": 102824862.0,
      "step": 3222
    },
    {
      "entropy": 0.8351584374904633,
      "epoch": 0.2965365351550595,
      "grad_norm": 0.15654917061328888,
      "learning_rate": 9.011868617168094e-05,
      "loss": 0.818,
      "mean_token_accuracy": 0.7649959996342659,
      "num_tokens": 102856599.0,
      "step": 3223
    },
    {
      "entropy": 0.7767375856637955,
      "epoch": 0.2966285415265007,
      "grad_norm": 0.15343037247657776,
      "learning_rate": 9.01156193455393e-05,
      "loss": 0.7736,
      "mean_token_accuracy": 0.7735678255558014,
      "num_tokens": 102888328.0,
      "step": 3224
    },
    {
      "entropy": 1.0024364590644836,
      "epoch": 0.29672054789794194,
      "grad_norm": 0.16720770299434662,
      "learning_rate": 9.011255251939767e-05,
      "loss": 0.9935,
      "mean_token_accuracy": 0.7193026691675186,
      "num_tokens": 102920029.0,
      "step": 3225
    },
    {
      "entropy": 0.7231214288622141,
      "epoch": 0.29681255426938313,
      "grad_norm": 0.15231072902679443,
      "learning_rate": 9.010948569325605e-05,
      "loss": 0.729,
      "mean_token_accuracy": 0.7829267345368862,
      "num_tokens": 102951839.0,
      "step": 3226
    },
    {
      "entropy": 0.7469484582543373,
      "epoch": 0.2969045606408244,
      "grad_norm": 0.16748405992984772,
      "learning_rate": 9.010641886711442e-05,
      "loss": 0.7611,
      "mean_token_accuracy": 0.7796236276626587,
      "num_tokens": 102983877.0,
      "step": 3227
    },
    {
      "entropy": 0.9165331497788429,
      "epoch": 0.2969965670122656,
      "grad_norm": 0.16665633022785187,
      "learning_rate": 9.01033520409728e-05,
      "loss": 0.9219,
      "mean_token_accuracy": 0.739287830889225,
      "num_tokens": 103015768.0,
      "step": 3228
    },
    {
      "entropy": 0.8059225287288427,
      "epoch": 0.2970885733837068,
      "grad_norm": 0.1589854508638382,
      "learning_rate": 9.010028521483117e-05,
      "loss": 0.8,
      "mean_token_accuracy": 0.7714627273380756,
      "num_tokens": 103047848.0,
      "step": 3229
    },
    {
      "entropy": 0.7455614730715752,
      "epoch": 0.29718057975514806,
      "grad_norm": 0.15252874791622162,
      "learning_rate": 9.009721838868955e-05,
      "loss": 0.7558,
      "mean_token_accuracy": 0.7799040004611015,
      "num_tokens": 103080029.0,
      "step": 3230
    },
    {
      "entropy": 0.8161697797477245,
      "epoch": 0.29727258612658924,
      "grad_norm": 0.15496276319026947,
      "learning_rate": 9.009415156254792e-05,
      "loss": 0.7981,
      "mean_token_accuracy": 0.7746829465031624,
      "num_tokens": 103111793.0,
      "step": 3231
    },
    {
      "entropy": 0.873393353074789,
      "epoch": 0.2973645924980305,
      "grad_norm": 0.15981364250183105,
      "learning_rate": 9.009108473640629e-05,
      "loss": 0.8564,
      "mean_token_accuracy": 0.7516282387077808,
      "num_tokens": 103143846.0,
      "step": 3232
    },
    {
      "entropy": 0.8967332802712917,
      "epoch": 0.29745659886947173,
      "grad_norm": 0.16450969874858856,
      "learning_rate": 9.008801791026467e-05,
      "loss": 0.8841,
      "mean_token_accuracy": 0.7510557547211647,
      "num_tokens": 103176195.0,
      "step": 3233
    },
    {
      "entropy": 0.7989018131047487,
      "epoch": 0.2975486052409129,
      "grad_norm": 0.15022067725658417,
      "learning_rate": 9.008495108412305e-05,
      "loss": 0.7817,
      "mean_token_accuracy": 0.7755111753940582,
      "num_tokens": 103207371.0,
      "step": 3234
    },
    {
      "entropy": 0.8476388491690159,
      "epoch": 0.29764061161235417,
      "grad_norm": 0.15629485249519348,
      "learning_rate": 9.008188425798142e-05,
      "loss": 0.8169,
      "mean_token_accuracy": 0.7614673934876919,
      "num_tokens": 103239172.0,
      "step": 3235
    },
    {
      "entropy": 0.8458954375237226,
      "epoch": 0.29773261798379536,
      "grad_norm": 0.15858782827854156,
      "learning_rate": 9.007881743183979e-05,
      "loss": 0.8192,
      "mean_token_accuracy": 0.7638715282082558,
      "num_tokens": 103271087.0,
      "step": 3236
    },
    {
      "entropy": 0.8923626653850079,
      "epoch": 0.2978246243552366,
      "grad_norm": 0.16321134567260742,
      "learning_rate": 9.007575060569817e-05,
      "loss": 0.8709,
      "mean_token_accuracy": 0.7477518804371357,
      "num_tokens": 103302330.0,
      "step": 3237
    },
    {
      "entropy": 0.7708925027400255,
      "epoch": 0.29791663072667784,
      "grad_norm": 0.1533215492963791,
      "learning_rate": 9.007268377955655e-05,
      "loss": 0.7588,
      "mean_token_accuracy": 0.7757132835686207,
      "num_tokens": 103334043.0,
      "step": 3238
    },
    {
      "entropy": 0.8961350172758102,
      "epoch": 0.29800863709811903,
      "grad_norm": 0.16175471246242523,
      "learning_rate": 9.006961695341492e-05,
      "loss": 0.8834,
      "mean_token_accuracy": 0.7501011155545712,
      "num_tokens": 103366338.0,
      "step": 3239
    },
    {
      "entropy": 0.7792738862335682,
      "epoch": 0.2981006434695603,
      "grad_norm": 0.15869633853435516,
      "learning_rate": 9.006655012727329e-05,
      "loss": 0.7918,
      "mean_token_accuracy": 0.7669379413127899,
      "num_tokens": 103397392.0,
      "step": 3240
    },
    {
      "entropy": 0.9605378359556198,
      "epoch": 0.29819264984100147,
      "grad_norm": 0.16249077022075653,
      "learning_rate": 9.006348330113167e-05,
      "loss": 0.9424,
      "mean_token_accuracy": 0.7302799448370934,
      "num_tokens": 103429845.0,
      "step": 3241
    },
    {
      "entropy": 0.7939936481416225,
      "epoch": 0.2982846562124427,
      "grad_norm": 0.15594905614852905,
      "learning_rate": 9.006041647499004e-05,
      "loss": 0.8002,
      "mean_token_accuracy": 0.7700883150100708,
      "num_tokens": 103461589.0,
      "step": 3242
    },
    {
      "entropy": 0.8614338338375092,
      "epoch": 0.29837666258388396,
      "grad_norm": 0.17068451642990112,
      "learning_rate": 9.005734964884842e-05,
      "loss": 0.8916,
      "mean_token_accuracy": 0.7457436993718147,
      "num_tokens": 103492820.0,
      "step": 3243
    },
    {
      "entropy": 0.9309524521231651,
      "epoch": 0.29846866895532514,
      "grad_norm": 0.17922911047935486,
      "learning_rate": 9.005428282270678e-05,
      "loss": 0.9832,
      "mean_token_accuracy": 0.7208291254937649,
      "num_tokens": 103525029.0,
      "step": 3244
    },
    {
      "entropy": 0.8712481632828712,
      "epoch": 0.2985606753267664,
      "grad_norm": 0.16579996049404144,
      "learning_rate": 9.005121599656517e-05,
      "loss": 0.8696,
      "mean_token_accuracy": 0.7552333362400532,
      "num_tokens": 103556660.0,
      "step": 3245
    },
    {
      "entropy": 0.802677646279335,
      "epoch": 0.2986526816982076,
      "grad_norm": 0.15707947313785553,
      "learning_rate": 9.004814917042353e-05,
      "loss": 0.7871,
      "mean_token_accuracy": 0.772093154489994,
      "num_tokens": 103588081.0,
      "step": 3246
    },
    {
      "entropy": 0.7899724282324314,
      "epoch": 0.2987446880696488,
      "grad_norm": 0.1583571434020996,
      "learning_rate": 9.00450823442819e-05,
      "loss": 0.8022,
      "mean_token_accuracy": 0.7658146806061268,
      "num_tokens": 103620617.0,
      "step": 3247
    },
    {
      "entropy": 0.8739928435534239,
      "epoch": 0.29883669444109007,
      "grad_norm": 0.15742501616477966,
      "learning_rate": 9.004201551814028e-05,
      "loss": 0.8538,
      "mean_token_accuracy": 0.7511718012392521,
      "num_tokens": 103651998.0,
      "step": 3248
    },
    {
      "entropy": 0.8284780085086823,
      "epoch": 0.29892870081253126,
      "grad_norm": 0.1617986112833023,
      "learning_rate": 9.003894869199866e-05,
      "loss": 0.8241,
      "mean_token_accuracy": 0.7647288702428341,
      "num_tokens": 103684052.0,
      "step": 3249
    },
    {
      "entropy": 0.6868641935288906,
      "epoch": 0.2990207071839725,
      "grad_norm": 0.14628863334655762,
      "learning_rate": 9.003588186585703e-05,
      "loss": 0.6508,
      "mean_token_accuracy": 0.8099590204656124,
      "num_tokens": 103715893.0,
      "step": 3250
    },
    {
      "entropy": 0.7928532026708126,
      "epoch": 0.2991127135554137,
      "grad_norm": 0.15125024318695068,
      "learning_rate": 9.00328150397154e-05,
      "loss": 0.7799,
      "mean_token_accuracy": 0.7759184390306473,
      "num_tokens": 103747535.0,
      "step": 3251
    },
    {
      "entropy": 0.8342645354568958,
      "epoch": 0.29920471992685493,
      "grad_norm": 0.154594287276268,
      "learning_rate": 9.002974821357377e-05,
      "loss": 0.8155,
      "mean_token_accuracy": 0.7633600421249866,
      "num_tokens": 103778687.0,
      "step": 3252
    },
    {
      "entropy": 0.782740430906415,
      "epoch": 0.2992967262982962,
      "grad_norm": 0.1500733494758606,
      "learning_rate": 9.002668138743215e-05,
      "loss": 0.7703,
      "mean_token_accuracy": 0.7757139392197132,
      "num_tokens": 103810206.0,
      "step": 3253
    },
    {
      "entropy": 0.8063284046947956,
      "epoch": 0.29938873266973737,
      "grad_norm": 0.1605575531721115,
      "learning_rate": 9.002361456129053e-05,
      "loss": 0.7922,
      "mean_token_accuracy": 0.7727717086672783,
      "num_tokens": 103841743.0,
      "step": 3254
    },
    {
      "entropy": 0.8755191117525101,
      "epoch": 0.2994807390411786,
      "grad_norm": 0.15651896595954895,
      "learning_rate": 9.00205477351489e-05,
      "loss": 0.8523,
      "mean_token_accuracy": 0.750374473631382,
      "num_tokens": 103873952.0,
      "step": 3255
    },
    {
      "entropy": 0.947271253913641,
      "epoch": 0.2995727454126198,
      "grad_norm": 0.16496525704860687,
      "learning_rate": 9.001748090900727e-05,
      "loss": 0.9655,
      "mean_token_accuracy": 0.7281784228980541,
      "num_tokens": 103906532.0,
      "step": 3256
    },
    {
      "entropy": 0.817999791353941,
      "epoch": 0.29966475178406105,
      "grad_norm": 0.1644277721643448,
      "learning_rate": 9.001441408286565e-05,
      "loss": 0.8534,
      "mean_token_accuracy": 0.7581680230796337,
      "num_tokens": 103938595.0,
      "step": 3257
    },
    {
      "entropy": 0.8273706845939159,
      "epoch": 0.2997567581555023,
      "grad_norm": 0.161917582154274,
      "learning_rate": 9.001134725672402e-05,
      "loss": 0.8206,
      "mean_token_accuracy": 0.7648443430662155,
      "num_tokens": 103970772.0,
      "step": 3258
    },
    {
      "entropy": 0.788858350366354,
      "epoch": 0.2998487645269435,
      "grad_norm": 0.15367211401462555,
      "learning_rate": 9.00082804305824e-05,
      "loss": 0.7852,
      "mean_token_accuracy": 0.7717203870415688,
      "num_tokens": 104003232.0,
      "step": 3259
    },
    {
      "entropy": 0.8062245659530163,
      "epoch": 0.2999407708983847,
      "grad_norm": 0.15244629979133606,
      "learning_rate": 9.000521360444077e-05,
      "loss": 0.7818,
      "mean_token_accuracy": 0.7775199860334396,
      "num_tokens": 104035239.0,
      "step": 3260
    },
    {
      "entropy": 0.8418078422546387,
      "epoch": 0.3000327772698259,
      "grad_norm": 0.15364786982536316,
      "learning_rate": 9.000214677829915e-05,
      "loss": 0.8327,
      "mean_token_accuracy": 0.7601509317755699,
      "num_tokens": 104067957.0,
      "step": 3261
    },
    {
      "entropy": 0.8187732361257076,
      "epoch": 0.30012478364126716,
      "grad_norm": 0.16093984246253967,
      "learning_rate": 8.999907995215751e-05,
      "loss": 0.8115,
      "mean_token_accuracy": 0.7645589895546436,
      "num_tokens": 104099722.0,
      "step": 3262
    },
    {
      "entropy": 0.8378640115261078,
      "epoch": 0.3002167900127084,
      "grad_norm": 0.1594831645488739,
      "learning_rate": 8.999601312601588e-05,
      "loss": 0.8374,
      "mean_token_accuracy": 0.7590962573885918,
      "num_tokens": 104132049.0,
      "step": 3263
    },
    {
      "entropy": 0.8633908107876778,
      "epoch": 0.3003087963841496,
      "grad_norm": 0.16856248676776886,
      "learning_rate": 8.999294629987426e-05,
      "loss": 0.8662,
      "mean_token_accuracy": 0.7548037022352219,
      "num_tokens": 104163972.0,
      "step": 3264
    },
    {
      "entropy": 0.7685533501207829,
      "epoch": 0.30040080275559083,
      "grad_norm": 0.16134634613990784,
      "learning_rate": 8.998987947373265e-05,
      "loss": 0.7878,
      "mean_token_accuracy": 0.7701773457229137,
      "num_tokens": 104196410.0,
      "step": 3265
    },
    {
      "entropy": 0.8683384396135807,
      "epoch": 0.300492809127032,
      "grad_norm": 0.16159674525260925,
      "learning_rate": 8.998681264759101e-05,
      "loss": 0.8618,
      "mean_token_accuracy": 0.749833457171917,
      "num_tokens": 104227871.0,
      "step": 3266
    },
    {
      "entropy": 0.846487246453762,
      "epoch": 0.30058481549847327,
      "grad_norm": 0.15977823734283447,
      "learning_rate": 8.998374582144938e-05,
      "loss": 0.8433,
      "mean_token_accuracy": 0.753989364951849,
      "num_tokens": 104259526.0,
      "step": 3267
    },
    {
      "entropy": 0.7987638935446739,
      "epoch": 0.3006768218699145,
      "grad_norm": 0.1621474176645279,
      "learning_rate": 8.998067899530776e-05,
      "loss": 0.8015,
      "mean_token_accuracy": 0.7690377943217754,
      "num_tokens": 104290858.0,
      "step": 3268
    },
    {
      "entropy": 0.9683430157601833,
      "epoch": 0.3007688282413557,
      "grad_norm": 0.1642398089170456,
      "learning_rate": 8.997761216916613e-05,
      "loss": 0.9483,
      "mean_token_accuracy": 0.7298828549683094,
      "num_tokens": 104321311.0,
      "step": 3269
    },
    {
      "entropy": 0.9350246898829937,
      "epoch": 0.30086083461279695,
      "grad_norm": 0.1613132804632187,
      "learning_rate": 8.997454534302451e-05,
      "loss": 0.9359,
      "mean_token_accuracy": 0.7328732721507549,
      "num_tokens": 104352924.0,
      "step": 3270
    },
    {
      "entropy": 0.9168709814548492,
      "epoch": 0.30095284098423813,
      "grad_norm": 0.1685355007648468,
      "learning_rate": 8.997147851688288e-05,
      "loss": 0.923,
      "mean_token_accuracy": 0.7411812767386436,
      "num_tokens": 104384870.0,
      "step": 3271
    },
    {
      "entropy": 0.8733504302799702,
      "epoch": 0.3010448473556794,
      "grad_norm": 0.16202107071876526,
      "learning_rate": 8.996841169074126e-05,
      "loss": 0.8493,
      "mean_token_accuracy": 0.75492849573493,
      "num_tokens": 104416023.0,
      "step": 3272
    },
    {
      "entropy": 0.8365091811865568,
      "epoch": 0.3011368537271206,
      "grad_norm": 0.15835338830947876,
      "learning_rate": 8.996534486459963e-05,
      "loss": 0.8286,
      "mean_token_accuracy": 0.7623091451823711,
      "num_tokens": 104447738.0,
      "step": 3273
    },
    {
      "entropy": 0.8803280927240849,
      "epoch": 0.3012288600985618,
      "grad_norm": 0.15852141380310059,
      "learning_rate": 8.9962278038458e-05,
      "loss": 0.8893,
      "mean_token_accuracy": 0.7468942105770111,
      "num_tokens": 104480423.0,
      "step": 3274
    },
    {
      "entropy": 0.8215044885873795,
      "epoch": 0.30132086647000306,
      "grad_norm": 0.151464581489563,
      "learning_rate": 8.995921121231638e-05,
      "loss": 0.8001,
      "mean_token_accuracy": 0.7666247002780437,
      "num_tokens": 104512608.0,
      "step": 3275
    },
    {
      "entropy": 0.8443125355988741,
      "epoch": 0.30141287284144425,
      "grad_norm": 0.1614406555891037,
      "learning_rate": 8.995614438617476e-05,
      "loss": 0.8351,
      "mean_token_accuracy": 0.758204348385334,
      "num_tokens": 104544281.0,
      "step": 3276
    },
    {
      "entropy": 0.7865425888448954,
      "epoch": 0.3015048792128855,
      "grad_norm": 0.16066083312034607,
      "learning_rate": 8.995307756003313e-05,
      "loss": 0.7874,
      "mean_token_accuracy": 0.771501936018467,
      "num_tokens": 104575544.0,
      "step": 3277
    },
    {
      "entropy": 0.8384427390992641,
      "epoch": 0.30159688558432673,
      "grad_norm": 0.15942394733428955,
      "learning_rate": 8.99500107338915e-05,
      "loss": 0.8358,
      "mean_token_accuracy": 0.7591462694108486,
      "num_tokens": 104606458.0,
      "step": 3278
    },
    {
      "entropy": 0.9054288417100906,
      "epoch": 0.3016888919557679,
      "grad_norm": 0.15813034772872925,
      "learning_rate": 8.994694390774986e-05,
      "loss": 0.8934,
      "mean_token_accuracy": 0.7426137514412403,
      "num_tokens": 104637557.0,
      "step": 3279
    },
    {
      "entropy": 0.7607414610683918,
      "epoch": 0.30178089832720917,
      "grad_norm": 0.14906376600265503,
      "learning_rate": 8.994387708160826e-05,
      "loss": 0.7281,
      "mean_token_accuracy": 0.7850393243134022,
      "num_tokens": 104668919.0,
      "step": 3280
    },
    {
      "entropy": 0.8519710376858711,
      "epoch": 0.30187290469865036,
      "grad_norm": 0.15631885826587677,
      "learning_rate": 8.994081025546663e-05,
      "loss": 0.863,
      "mean_token_accuracy": 0.7527400441467762,
      "num_tokens": 104701655.0,
      "step": 3281
    },
    {
      "entropy": 0.8114689886569977,
      "epoch": 0.3019649110700916,
      "grad_norm": 0.15576130151748657,
      "learning_rate": 8.9937743429325e-05,
      "loss": 0.8063,
      "mean_token_accuracy": 0.7691286876797676,
      "num_tokens": 104734206.0,
      "step": 3282
    },
    {
      "entropy": 0.7976368572562933,
      "epoch": 0.30205691744153285,
      "grad_norm": 0.1560947597026825,
      "learning_rate": 8.993467660318336e-05,
      "loss": 0.8004,
      "mean_token_accuracy": 0.7684306725859642,
      "num_tokens": 104766188.0,
      "step": 3283
    },
    {
      "entropy": 0.7875513788312674,
      "epoch": 0.30214892381297404,
      "grad_norm": 0.16151869297027588,
      "learning_rate": 8.993160977704174e-05,
      "loss": 0.7771,
      "mean_token_accuracy": 0.7791718728840351,
      "num_tokens": 104798414.0,
      "step": 3284
    },
    {
      "entropy": 0.8232519142329693,
      "epoch": 0.3022409301844153,
      "grad_norm": 0.16634659469127655,
      "learning_rate": 8.992854295090012e-05,
      "loss": 0.8405,
      "mean_token_accuracy": 0.7605712227523327,
      "num_tokens": 104830697.0,
      "step": 3285
    },
    {
      "entropy": 0.835929898545146,
      "epoch": 0.30233293655585647,
      "grad_norm": 0.1599072515964508,
      "learning_rate": 8.992547612475849e-05,
      "loss": 0.8207,
      "mean_token_accuracy": 0.7602669596672058,
      "num_tokens": 104862562.0,
      "step": 3286
    },
    {
      "entropy": 0.8086448535323143,
      "epoch": 0.3024249429272977,
      "grad_norm": 0.162196084856987,
      "learning_rate": 8.992240929861686e-05,
      "loss": 0.8029,
      "mean_token_accuracy": 0.7627005912363529,
      "num_tokens": 104893947.0,
      "step": 3287
    },
    {
      "entropy": 0.7711968813091516,
      "epoch": 0.30251694929873896,
      "grad_norm": 0.15262001752853394,
      "learning_rate": 8.991934247247524e-05,
      "loss": 0.7652,
      "mean_token_accuracy": 0.7768972106277943,
      "num_tokens": 104925827.0,
      "step": 3288
    },
    {
      "entropy": 0.7577444668859243,
      "epoch": 0.30260895567018015,
      "grad_norm": 0.15439055860042572,
      "learning_rate": 8.991627564633361e-05,
      "loss": 0.753,
      "mean_token_accuracy": 0.7858549393713474,
      "num_tokens": 104958124.0,
      "step": 3289
    },
    {
      "entropy": 0.9049224965274334,
      "epoch": 0.3027009620416214,
      "grad_norm": 0.16681306064128876,
      "learning_rate": 8.991320882019199e-05,
      "loss": 0.9078,
      "mean_token_accuracy": 0.7460558265447617,
      "num_tokens": 104990189.0,
      "step": 3290
    },
    {
      "entropy": 0.7750443089753389,
      "epoch": 0.3027929684130626,
      "grad_norm": 0.15733176469802856,
      "learning_rate": 8.991014199405036e-05,
      "loss": 0.7684,
      "mean_token_accuracy": 0.7762825153768063,
      "num_tokens": 105022348.0,
      "step": 3291
    },
    {
      "entropy": 0.9550650827586651,
      "epoch": 0.3028849747845038,
      "grad_norm": 0.16315561532974243,
      "learning_rate": 8.990707516790874e-05,
      "loss": 0.9478,
      "mean_token_accuracy": 0.7287241034209728,
      "num_tokens": 105054885.0,
      "step": 3292
    },
    {
      "entropy": 0.9031356237828732,
      "epoch": 0.30297698115594507,
      "grad_norm": 0.1623692363500595,
      "learning_rate": 8.990400834176711e-05,
      "loss": 0.9013,
      "mean_token_accuracy": 0.7411410510540009,
      "num_tokens": 105087024.0,
      "step": 3293
    },
    {
      "entropy": 0.9123191609978676,
      "epoch": 0.30306898752738626,
      "grad_norm": 0.1640353500843048,
      "learning_rate": 8.990094151562548e-05,
      "loss": 0.9174,
      "mean_token_accuracy": 0.7345234416425228,
      "num_tokens": 105118965.0,
      "step": 3294
    },
    {
      "entropy": 0.7764587961137295,
      "epoch": 0.3031609938988275,
      "grad_norm": 0.15043382346630096,
      "learning_rate": 8.989787468948386e-05,
      "loss": 0.7863,
      "mean_token_accuracy": 0.7733436226844788,
      "num_tokens": 105150905.0,
      "step": 3295
    },
    {
      "entropy": 0.7245520725846291,
      "epoch": 0.3032530002702687,
      "grad_norm": 0.14699965715408325,
      "learning_rate": 8.989480786334224e-05,
      "loss": 0.7061,
      "mean_token_accuracy": 0.7925604283809662,
      "num_tokens": 105183522.0,
      "step": 3296
    },
    {
      "entropy": 0.8781691323965788,
      "epoch": 0.30334500664170994,
      "grad_norm": 0.1613631397485733,
      "learning_rate": 8.989174103720061e-05,
      "loss": 0.8785,
      "mean_token_accuracy": 0.7492958977818489,
      "num_tokens": 105215544.0,
      "step": 3297
    },
    {
      "entropy": 0.7716808971017599,
      "epoch": 0.3034370130131512,
      "grad_norm": 0.15590186417102814,
      "learning_rate": 8.988867421105897e-05,
      "loss": 0.7699,
      "mean_token_accuracy": 0.7764951586723328,
      "num_tokens": 105247026.0,
      "step": 3298
    },
    {
      "entropy": 0.9221753589808941,
      "epoch": 0.30352901938459237,
      "grad_norm": 0.1586785465478897,
      "learning_rate": 8.988560738491736e-05,
      "loss": 0.9083,
      "mean_token_accuracy": 0.7405655793845654,
      "num_tokens": 105279177.0,
      "step": 3299
    },
    {
      "entropy": 0.7996969297528267,
      "epoch": 0.3036210257560336,
      "grad_norm": 0.15162025392055511,
      "learning_rate": 8.988254055877572e-05,
      "loss": 0.7897,
      "mean_token_accuracy": 0.7700985707342625,
      "num_tokens": 105311215.0,
      "step": 3300
    },
    {
      "entropy": 0.8502285815775394,
      "epoch": 0.3037130321274748,
      "grad_norm": 0.16252130270004272,
      "learning_rate": 8.98794737326341e-05,
      "loss": 0.8373,
      "mean_token_accuracy": 0.7606907598674297,
      "num_tokens": 105343258.0,
      "step": 3301
    },
    {
      "entropy": 0.8440442532300949,
      "epoch": 0.30380503849891605,
      "grad_norm": 0.16510072350502014,
      "learning_rate": 8.987640690649247e-05,
      "loss": 0.8188,
      "mean_token_accuracy": 0.7636057585477829,
      "num_tokens": 105375090.0,
      "step": 3302
    },
    {
      "entropy": 0.8962010480463505,
      "epoch": 0.3038970448703573,
      "grad_norm": 0.15951000154018402,
      "learning_rate": 8.987334008035085e-05,
      "loss": 0.8818,
      "mean_token_accuracy": 0.7452445812523365,
      "num_tokens": 105407819.0,
      "step": 3303
    },
    {
      "entropy": 0.9074329137802124,
      "epoch": 0.3039890512417985,
      "grad_norm": 0.16500933468341827,
      "learning_rate": 8.987027325420922e-05,
      "loss": 0.8856,
      "mean_token_accuracy": 0.7465296946465969,
      "num_tokens": 105439615.0,
      "step": 3304
    },
    {
      "entropy": 0.6833754424005747,
      "epoch": 0.3040810576132397,
      "grad_norm": 0.1534537822008133,
      "learning_rate": 8.986720642806759e-05,
      "loss": 0.6698,
      "mean_token_accuracy": 0.8015486225485802,
      "num_tokens": 105471659.0,
      "step": 3305
    },
    {
      "entropy": 0.7984276637434959,
      "epoch": 0.3041730639846809,
      "grad_norm": 0.159820556640625,
      "learning_rate": 8.986413960192597e-05,
      "loss": 0.8059,
      "mean_token_accuracy": 0.7688219845294952,
      "num_tokens": 105503751.0,
      "step": 3306
    },
    {
      "entropy": 0.8274598345160484,
      "epoch": 0.30426507035612216,
      "grad_norm": 0.15676307678222656,
      "learning_rate": 8.986107277578435e-05,
      "loss": 0.8248,
      "mean_token_accuracy": 0.761259950697422,
      "num_tokens": 105535513.0,
      "step": 3307
    },
    {
      "entropy": 0.7732108980417252,
      "epoch": 0.3043570767275634,
      "grad_norm": 0.16581346094608307,
      "learning_rate": 8.985800594964272e-05,
      "loss": 0.7847,
      "mean_token_accuracy": 0.7757288031280041,
      "num_tokens": 105568147.0,
      "step": 3308
    },
    {
      "entropy": 0.8576384261250496,
      "epoch": 0.3044490830990046,
      "grad_norm": 0.15907779335975647,
      "learning_rate": 8.985493912350109e-05,
      "loss": 0.8618,
      "mean_token_accuracy": 0.7537677325308323,
      "num_tokens": 105599858.0,
      "step": 3309
    },
    {
      "entropy": 0.8255168981850147,
      "epoch": 0.30454108947044584,
      "grad_norm": 0.16256462037563324,
      "learning_rate": 8.985187229735946e-05,
      "loss": 0.8375,
      "mean_token_accuracy": 0.7583555802702904,
      "num_tokens": 105631613.0,
      "step": 3310
    },
    {
      "entropy": 0.8038213066756725,
      "epoch": 0.304633095841887,
      "grad_norm": 0.164347842335701,
      "learning_rate": 8.984880547121784e-05,
      "loss": 0.813,
      "mean_token_accuracy": 0.7636975757777691,
      "num_tokens": 105664087.0,
      "step": 3311
    },
    {
      "entropy": 0.8426744043827057,
      "epoch": 0.30472510221332827,
      "grad_norm": 0.15932713449001312,
      "learning_rate": 8.984573864507622e-05,
      "loss": 0.8257,
      "mean_token_accuracy": 0.7636202238500118,
      "num_tokens": 105695770.0,
      "step": 3312
    },
    {
      "entropy": 0.8294839784502983,
      "epoch": 0.3048171085847695,
      "grad_norm": 0.1553800255060196,
      "learning_rate": 8.984267181893459e-05,
      "loss": 0.809,
      "mean_token_accuracy": 0.769551333039999,
      "num_tokens": 105727924.0,
      "step": 3313
    },
    {
      "entropy": 0.8986268546432257,
      "epoch": 0.3049091149562107,
      "grad_norm": 0.16098618507385254,
      "learning_rate": 8.983960499279296e-05,
      "loss": 0.8671,
      "mean_token_accuracy": 0.7564220763742924,
      "num_tokens": 105760114.0,
      "step": 3314
    },
    {
      "entropy": 0.8436742760241032,
      "epoch": 0.30500112132765195,
      "grad_norm": 0.1690879464149475,
      "learning_rate": 8.983653816665134e-05,
      "loss": 0.8515,
      "mean_token_accuracy": 0.7549375891685486,
      "num_tokens": 105791865.0,
      "step": 3315
    },
    {
      "entropy": 0.8454981185495853,
      "epoch": 0.30509312769909314,
      "grad_norm": 0.16441774368286133,
      "learning_rate": 8.98334713405097e-05,
      "loss": 0.8501,
      "mean_token_accuracy": 0.7569825872778893,
      "num_tokens": 105823394.0,
      "step": 3316
    },
    {
      "entropy": 0.8073604591190815,
      "epoch": 0.3051851340705344,
      "grad_norm": 0.1652093082666397,
      "learning_rate": 8.983040451436809e-05,
      "loss": 0.8281,
      "mean_token_accuracy": 0.7651664577424526,
      "num_tokens": 105855428.0,
      "step": 3317
    },
    {
      "entropy": 0.7817899733781815,
      "epoch": 0.3052771404419756,
      "grad_norm": 0.1541690230369568,
      "learning_rate": 8.982733768822645e-05,
      "loss": 0.7726,
      "mean_token_accuracy": 0.7757956422865391,
      "num_tokens": 105887548.0,
      "step": 3318
    },
    {
      "entropy": 0.7542545199394226,
      "epoch": 0.3053691468134168,
      "grad_norm": 0.1585613191127777,
      "learning_rate": 8.982427086208484e-05,
      "loss": 0.7595,
      "mean_token_accuracy": 0.7744403779506683,
      "num_tokens": 105919729.0,
      "step": 3319
    },
    {
      "entropy": 0.7945720311254263,
      "epoch": 0.30546115318485806,
      "grad_norm": 0.14960896968841553,
      "learning_rate": 8.98212040359432e-05,
      "loss": 0.7847,
      "mean_token_accuracy": 0.7692020013928413,
      "num_tokens": 105952100.0,
      "step": 3320
    },
    {
      "entropy": 0.8481711223721504,
      "epoch": 0.30555315955629925,
      "grad_norm": 0.15867899358272552,
      "learning_rate": 8.981813720980158e-05,
      "loss": 0.8391,
      "mean_token_accuracy": 0.7576230764389038,
      "num_tokens": 105983718.0,
      "step": 3321
    },
    {
      "entropy": 0.73432806879282,
      "epoch": 0.3056451659277405,
      "grad_norm": 0.1511050909757614,
      "learning_rate": 8.981507038365995e-05,
      "loss": 0.7228,
      "mean_token_accuracy": 0.7931059114634991,
      "num_tokens": 106015090.0,
      "step": 3322
    },
    {
      "entropy": 0.8498538285493851,
      "epoch": 0.30573717229918174,
      "grad_norm": 0.15675461292266846,
      "learning_rate": 8.981200355751833e-05,
      "loss": 0.8378,
      "mean_token_accuracy": 0.7559861354529858,
      "num_tokens": 106047161.0,
      "step": 3323
    },
    {
      "entropy": 0.8771975375711918,
      "epoch": 0.3058291786706229,
      "grad_norm": 0.1592167168855667,
      "learning_rate": 8.98089367313767e-05,
      "loss": 0.8819,
      "mean_token_accuracy": 0.7462339773774147,
      "num_tokens": 106078936.0,
      "step": 3324
    },
    {
      "entropy": 0.8097424507141113,
      "epoch": 0.30592118504206417,
      "grad_norm": 0.16221049427986145,
      "learning_rate": 8.980586990523507e-05,
      "loss": 0.8179,
      "mean_token_accuracy": 0.7635729014873505,
      "num_tokens": 106110577.0,
      "step": 3325
    },
    {
      "entropy": 0.829821765422821,
      "epoch": 0.30601319141350536,
      "grad_norm": 0.16331863403320312,
      "learning_rate": 8.980280307909345e-05,
      "loss": 0.8239,
      "mean_token_accuracy": 0.7648102715611458,
      "num_tokens": 106141365.0,
      "step": 3326
    },
    {
      "entropy": 0.8956534564495087,
      "epoch": 0.3061051977849466,
      "grad_norm": 0.162184938788414,
      "learning_rate": 8.979973625295183e-05,
      "loss": 0.8926,
      "mean_token_accuracy": 0.7476503364741802,
      "num_tokens": 106173400.0,
      "step": 3327
    },
    {
      "entropy": 0.8383217975497246,
      "epoch": 0.30619720415638785,
      "grad_norm": 0.1588013768196106,
      "learning_rate": 8.97966694268102e-05,
      "loss": 0.8311,
      "mean_token_accuracy": 0.7612211406230927,
      "num_tokens": 106205034.0,
      "step": 3328
    },
    {
      "entropy": 0.8862363826483488,
      "epoch": 0.30628921052782904,
      "grad_norm": 0.17446106672286987,
      "learning_rate": 8.979360260066857e-05,
      "loss": 0.9114,
      "mean_token_accuracy": 0.7449915036559105,
      "num_tokens": 106236353.0,
      "step": 3329
    },
    {
      "entropy": 0.8446763232350349,
      "epoch": 0.3063812168992703,
      "grad_norm": 0.16113048791885376,
      "learning_rate": 8.979053577452695e-05,
      "loss": 0.842,
      "mean_token_accuracy": 0.7623230889439583,
      "num_tokens": 106268275.0,
      "step": 3330
    },
    {
      "entropy": 0.7043800875544548,
      "epoch": 0.30647322327071147,
      "grad_norm": 0.14937055110931396,
      "learning_rate": 8.978746894838532e-05,
      "loss": 0.6726,
      "mean_token_accuracy": 0.8059864901006222,
      "num_tokens": 106299876.0,
      "step": 3331
    },
    {
      "entropy": 0.8158455733209848,
      "epoch": 0.3065652296421527,
      "grad_norm": 0.1580374538898468,
      "learning_rate": 8.97844021222437e-05,
      "loss": 0.7844,
      "mean_token_accuracy": 0.7756520546972752,
      "num_tokens": 106330673.0,
      "step": 3332
    },
    {
      "entropy": 0.7675344198942184,
      "epoch": 0.30665723601359396,
      "grad_norm": 0.15294520556926727,
      "learning_rate": 8.978133529610207e-05,
      "loss": 0.7504,
      "mean_token_accuracy": 0.7812991514801979,
      "num_tokens": 106362062.0,
      "step": 3333
    },
    {
      "entropy": 0.8845080770552158,
      "epoch": 0.30674924238503515,
      "grad_norm": 0.15565115213394165,
      "learning_rate": 8.977826846996045e-05,
      "loss": 0.8823,
      "mean_token_accuracy": 0.745218213647604,
      "num_tokens": 106393897.0,
      "step": 3334
    },
    {
      "entropy": 0.9474118612706661,
      "epoch": 0.3068412487564764,
      "grad_norm": 0.1660860776901245,
      "learning_rate": 8.977520164381882e-05,
      "loss": 0.9515,
      "mean_token_accuracy": 0.732705395668745,
      "num_tokens": 106424731.0,
      "step": 3335
    },
    {
      "entropy": 0.8411702234297991,
      "epoch": 0.3069332551279176,
      "grad_norm": 0.16285820305347443,
      "learning_rate": 8.977213481767718e-05,
      "loss": 0.8452,
      "mean_token_accuracy": 0.7572759613394737,
      "num_tokens": 106456487.0,
      "step": 3336
    },
    {
      "entropy": 0.7549398392438889,
      "epoch": 0.3070252614993588,
      "grad_norm": 0.16553448140621185,
      "learning_rate": 8.976906799153557e-05,
      "loss": 0.737,
      "mean_token_accuracy": 0.7847160659730434,
      "num_tokens": 106488223.0,
      "step": 3337
    },
    {
      "entropy": 0.7920087538659573,
      "epoch": 0.30711726787080007,
      "grad_norm": 0.16820314526557922,
      "learning_rate": 8.976600116539395e-05,
      "loss": 0.7914,
      "mean_token_accuracy": 0.7708416096866131,
      "num_tokens": 106520456.0,
      "step": 3338
    },
    {
      "entropy": 0.7634782716631889,
      "epoch": 0.30720927424224126,
      "grad_norm": 0.1590406447649002,
      "learning_rate": 8.976293433925232e-05,
      "loss": 0.7716,
      "mean_token_accuracy": 0.7753103524446487,
      "num_tokens": 106552699.0,
      "step": 3339
    },
    {
      "entropy": 0.8210577815771103,
      "epoch": 0.3073012806136825,
      "grad_norm": 0.16458940505981445,
      "learning_rate": 8.975986751311068e-05,
      "loss": 0.8193,
      "mean_token_accuracy": 0.7632529586553574,
      "num_tokens": 106584265.0,
      "step": 3340
    },
    {
      "entropy": 0.8422010466456413,
      "epoch": 0.3073932869851237,
      "grad_norm": 0.15678757429122925,
      "learning_rate": 8.975680068696905e-05,
      "loss": 0.8309,
      "mean_token_accuracy": 0.7565379999577999,
      "num_tokens": 106616371.0,
      "step": 3341
    },
    {
      "entropy": 0.789811035618186,
      "epoch": 0.30748529335656494,
      "grad_norm": 0.1548515409231186,
      "learning_rate": 8.975373386082743e-05,
      "loss": 0.7771,
      "mean_token_accuracy": 0.7731887884438038,
      "num_tokens": 106648738.0,
      "step": 3342
    },
    {
      "entropy": 0.9151780344545841,
      "epoch": 0.3075772997280062,
      "grad_norm": 0.1640041470527649,
      "learning_rate": 8.975066703468581e-05,
      "loss": 0.9117,
      "mean_token_accuracy": 0.7416977249085903,
      "num_tokens": 106681124.0,
      "step": 3343
    },
    {
      "entropy": 0.9039301872253418,
      "epoch": 0.30766930609944737,
      "grad_norm": 0.16260842978954315,
      "learning_rate": 8.974760020854418e-05,
      "loss": 0.9089,
      "mean_token_accuracy": 0.7424873895943165,
      "num_tokens": 106713021.0,
      "step": 3344
    },
    {
      "entropy": 0.7645845711231232,
      "epoch": 0.3077613124708886,
      "grad_norm": 0.15416578948497772,
      "learning_rate": 8.974453338240255e-05,
      "loss": 0.7609,
      "mean_token_accuracy": 0.776732299476862,
      "num_tokens": 106745129.0,
      "step": 3345
    },
    {
      "entropy": 0.8956094495952129,
      "epoch": 0.3078533188423298,
      "grad_norm": 0.16266463696956635,
      "learning_rate": 8.974146655626093e-05,
      "loss": 0.8859,
      "mean_token_accuracy": 0.7439479492604733,
      "num_tokens": 106777012.0,
      "step": 3346
    },
    {
      "entropy": 0.9455668944865465,
      "epoch": 0.30794532521377105,
      "grad_norm": 0.16309280693531036,
      "learning_rate": 8.97383997301193e-05,
      "loss": 0.9356,
      "mean_token_accuracy": 0.7327968738973141,
      "num_tokens": 106809404.0,
      "step": 3347
    },
    {
      "entropy": 0.9186445437371731,
      "epoch": 0.3080373315852123,
      "grad_norm": 0.15165959298610687,
      "learning_rate": 8.973533290397768e-05,
      "loss": 0.8972,
      "mean_token_accuracy": 0.742682695388794,
      "num_tokens": 106841767.0,
      "step": 3348
    },
    {
      "entropy": 0.9152895957231522,
      "epoch": 0.3081293379566535,
      "grad_norm": 0.15749751031398773,
      "learning_rate": 8.973226607783605e-05,
      "loss": 0.9171,
      "mean_token_accuracy": 0.7377259321510792,
      "num_tokens": 106874273.0,
      "step": 3349
    },
    {
      "entropy": 0.8876442909240723,
      "epoch": 0.3082213443280947,
      "grad_norm": 0.16297698020935059,
      "learning_rate": 8.972919925169443e-05,
      "loss": 0.8887,
      "mean_token_accuracy": 0.7481345348060131,
      "num_tokens": 106906367.0,
      "step": 3350
    },
    {
      "entropy": 0.8780762143433094,
      "epoch": 0.3083133506995359,
      "grad_norm": 0.16475142538547516,
      "learning_rate": 8.97261324255528e-05,
      "loss": 0.8786,
      "mean_token_accuracy": 0.7440071068704128,
      "num_tokens": 106937417.0,
      "step": 3351
    },
    {
      "entropy": 0.8727890364825726,
      "epoch": 0.30840535707097716,
      "grad_norm": 0.1611245572566986,
      "learning_rate": 8.972306559941117e-05,
      "loss": 0.8938,
      "mean_token_accuracy": 0.7447970025241375,
      "num_tokens": 106968851.0,
      "step": 3352
    },
    {
      "entropy": 0.7861177083104849,
      "epoch": 0.3084973634424184,
      "grad_norm": 0.1571521908044815,
      "learning_rate": 8.971999877326955e-05,
      "loss": 0.7753,
      "mean_token_accuracy": 0.7739078663289547,
      "num_tokens": 107000710.0,
      "step": 3353
    },
    {
      "entropy": 0.7040581107139587,
      "epoch": 0.3085893698138596,
      "grad_norm": 0.15606698393821716,
      "learning_rate": 8.971693194712793e-05,
      "loss": 0.7012,
      "mean_token_accuracy": 0.7910261079668999,
      "num_tokens": 107032569.0,
      "step": 3354
    },
    {
      "entropy": 0.8301891945302486,
      "epoch": 0.30868137618530084,
      "grad_norm": 0.17488594353199005,
      "learning_rate": 8.97138651209863e-05,
      "loss": 0.8609,
      "mean_token_accuracy": 0.757118783891201,
      "num_tokens": 107064846.0,
      "step": 3355
    },
    {
      "entropy": 0.8114203121513128,
      "epoch": 0.3087733825567421,
      "grad_norm": 0.16149668395519257,
      "learning_rate": 8.971079829484466e-05,
      "loss": 0.8116,
      "mean_token_accuracy": 0.769741602241993,
      "num_tokens": 107096512.0,
      "step": 3356
    },
    {
      "entropy": 0.8135862872004509,
      "epoch": 0.30886538892818327,
      "grad_norm": 0.15547241270542145,
      "learning_rate": 8.970773146870305e-05,
      "loss": 0.8204,
      "mean_token_accuracy": 0.7613033726811409,
      "num_tokens": 107128743.0,
      "step": 3357
    },
    {
      "entropy": 0.9450870268046856,
      "epoch": 0.3089573952996245,
      "grad_norm": 0.1613178700208664,
      "learning_rate": 8.970466464256143e-05,
      "loss": 0.9335,
      "mean_token_accuracy": 0.7328628338873386,
      "num_tokens": 107160793.0,
      "step": 3358
    },
    {
      "entropy": 0.7558762524276972,
      "epoch": 0.3090494016710657,
      "grad_norm": 0.15947629511356354,
      "learning_rate": 8.97015978164198e-05,
      "loss": 0.7581,
      "mean_token_accuracy": 0.7809092849493027,
      "num_tokens": 107192843.0,
      "step": 3359
    },
    {
      "entropy": 0.874690368771553,
      "epoch": 0.30914140804250695,
      "grad_norm": 0.1612749993801117,
      "learning_rate": 8.969853099027816e-05,
      "loss": 0.8714,
      "mean_token_accuracy": 0.74614367634058,
      "num_tokens": 107224375.0,
      "step": 3360
    },
    {
      "entropy": 0.8191475626081228,
      "epoch": 0.3092334144139482,
      "grad_norm": 0.1571519523859024,
      "learning_rate": 8.969546416413654e-05,
      "loss": 0.8057,
      "mean_token_accuracy": 0.7693458683788776,
      "num_tokens": 107256421.0,
      "step": 3361
    },
    {
      "entropy": 0.916392320767045,
      "epoch": 0.3093254207853894,
      "grad_norm": 0.16228047013282776,
      "learning_rate": 8.969239733799491e-05,
      "loss": 0.9165,
      "mean_token_accuracy": 0.7416674941778183,
      "num_tokens": 107288895.0,
      "step": 3362
    },
    {
      "entropy": 0.8933110572397709,
      "epoch": 0.3094174271568306,
      "grad_norm": 0.15446734428405762,
      "learning_rate": 8.968933051185329e-05,
      "loss": 0.8795,
      "mean_token_accuracy": 0.7469246350228786,
      "num_tokens": 107321521.0,
      "step": 3363
    },
    {
      "entropy": 0.8813019767403603,
      "epoch": 0.3095094335282718,
      "grad_norm": 0.16771674156188965,
      "learning_rate": 8.968626368571166e-05,
      "loss": 0.8756,
      "mean_token_accuracy": 0.7533271200954914,
      "num_tokens": 107352071.0,
      "step": 3364
    },
    {
      "entropy": 0.9212164767086506,
      "epoch": 0.30960143989971306,
      "grad_norm": 0.16714149713516235,
      "learning_rate": 8.968319685957004e-05,
      "loss": 0.907,
      "mean_token_accuracy": 0.7390821017324924,
      "num_tokens": 107383196.0,
      "step": 3365
    },
    {
      "entropy": 0.8212806023657322,
      "epoch": 0.3096934462711543,
      "grad_norm": 0.15265826880931854,
      "learning_rate": 8.968013003342841e-05,
      "loss": 0.8083,
      "mean_token_accuracy": 0.7634923309087753,
      "num_tokens": 107415261.0,
      "step": 3366
    },
    {
      "entropy": 0.9264769703149796,
      "epoch": 0.3097854526425955,
      "grad_norm": 0.16587881743907928,
      "learning_rate": 8.967706320728678e-05,
      "loss": 0.9247,
      "mean_token_accuracy": 0.7352628074586391,
      "num_tokens": 107446861.0,
      "step": 3367
    },
    {
      "entropy": 0.9255437925457954,
      "epoch": 0.30987745901403674,
      "grad_norm": 0.16572177410125732,
      "learning_rate": 8.967399638114516e-05,
      "loss": 0.926,
      "mean_token_accuracy": 0.735027190297842,
      "num_tokens": 107478657.0,
      "step": 3368
    },
    {
      "entropy": 0.6772649940103292,
      "epoch": 0.3099694653854779,
      "grad_norm": 0.14789026975631714,
      "learning_rate": 8.967092955500354e-05,
      "loss": 0.6594,
      "mean_token_accuracy": 0.8099015206098557,
      "num_tokens": 107511179.0,
      "step": 3369
    },
    {
      "entropy": 0.8865250870585442,
      "epoch": 0.31006147175691917,
      "grad_norm": 0.16351258754730225,
      "learning_rate": 8.966786272886191e-05,
      "loss": 0.8762,
      "mean_token_accuracy": 0.745740819722414,
      "num_tokens": 107542324.0,
      "step": 3370
    },
    {
      "entropy": 0.8729602210223675,
      "epoch": 0.3101534781283604,
      "grad_norm": 0.1611730456352234,
      "learning_rate": 8.966479590272028e-05,
      "loss": 0.8871,
      "mean_token_accuracy": 0.7484557814896107,
      "num_tokens": 107574530.0,
      "step": 3371
    },
    {
      "entropy": 0.9120924957096577,
      "epoch": 0.3102454844998016,
      "grad_norm": 0.16307108104228973,
      "learning_rate": 8.966172907657864e-05,
      "loss": 0.9041,
      "mean_token_accuracy": 0.7389633171260357,
      "num_tokens": 107606235.0,
      "step": 3372
    },
    {
      "entropy": 0.8155715521425009,
      "epoch": 0.31033749087124285,
      "grad_norm": 0.15640553832054138,
      "learning_rate": 8.965866225043703e-05,
      "loss": 0.8085,
      "mean_token_accuracy": 0.7675619050860405,
      "num_tokens": 107638389.0,
      "step": 3373
    },
    {
      "entropy": 0.8114733807742596,
      "epoch": 0.31042949724268404,
      "grad_norm": 0.16036543250083923,
      "learning_rate": 8.965559542429541e-05,
      "loss": 0.8197,
      "mean_token_accuracy": 0.7647829875349998,
      "num_tokens": 107670776.0,
      "step": 3374
    },
    {
      "entropy": 0.8628039713948965,
      "epoch": 0.3105215036141253,
      "grad_norm": 0.1548987329006195,
      "learning_rate": 8.965252859815378e-05,
      "loss": 0.8529,
      "mean_token_accuracy": 0.7598339356482029,
      "num_tokens": 107703305.0,
      "step": 3375
    },
    {
      "entropy": 0.8002487942576408,
      "epoch": 0.3106135099855665,
      "grad_norm": 0.15298177301883698,
      "learning_rate": 8.964946177201214e-05,
      "loss": 0.7843,
      "mean_token_accuracy": 0.7678748928010464,
      "num_tokens": 107735468.0,
      "step": 3376
    },
    {
      "entropy": 0.8673310242593288,
      "epoch": 0.3107055163570077,
      "grad_norm": 0.158192440867424,
      "learning_rate": 8.964639494587052e-05,
      "loss": 0.8577,
      "mean_token_accuracy": 0.7484673410654068,
      "num_tokens": 107766588.0,
      "step": 3377
    },
    {
      "entropy": 0.9051303900778294,
      "epoch": 0.31079752272844896,
      "grad_norm": 0.1605863869190216,
      "learning_rate": 8.964332811972889e-05,
      "loss": 0.9021,
      "mean_token_accuracy": 0.7446774244308472,
      "num_tokens": 107799151.0,
      "step": 3378
    },
    {
      "entropy": 0.761414485052228,
      "epoch": 0.31088952909989015,
      "grad_norm": 0.15178439021110535,
      "learning_rate": 8.964026129358727e-05,
      "loss": 0.7433,
      "mean_token_accuracy": 0.7865045443177223,
      "num_tokens": 107831405.0,
      "step": 3379
    },
    {
      "entropy": 0.724536344408989,
      "epoch": 0.3109815354713314,
      "grad_norm": 0.15169699490070343,
      "learning_rate": 8.963719446744564e-05,
      "loss": 0.7212,
      "mean_token_accuracy": 0.7915614955127239,
      "num_tokens": 107862889.0,
      "step": 3380
    },
    {
      "entropy": 0.9335004575550556,
      "epoch": 0.31107354184277264,
      "grad_norm": 0.16302767395973206,
      "learning_rate": 8.963412764130402e-05,
      "loss": 0.9239,
      "mean_token_accuracy": 0.7339739128947258,
      "num_tokens": 107893993.0,
      "step": 3381
    },
    {
      "entropy": 0.8746923618018627,
      "epoch": 0.3111655482142138,
      "grad_norm": 0.1629883348941803,
      "learning_rate": 8.963106081516239e-05,
      "loss": 0.8764,
      "mean_token_accuracy": 0.747954323887825,
      "num_tokens": 107926761.0,
      "step": 3382
    },
    {
      "entropy": 0.8743528239428997,
      "epoch": 0.31125755458565507,
      "grad_norm": 0.16261225938796997,
      "learning_rate": 8.962799398902076e-05,
      "loss": 0.8613,
      "mean_token_accuracy": 0.7521673105657101,
      "num_tokens": 107958810.0,
      "step": 3383
    },
    {
      "entropy": 0.8095735721290112,
      "epoch": 0.31134956095709626,
      "grad_norm": 0.1655687689781189,
      "learning_rate": 8.962492716287914e-05,
      "loss": 0.8165,
      "mean_token_accuracy": 0.765890333801508,
      "num_tokens": 107989736.0,
      "step": 3384
    },
    {
      "entropy": 0.8671386782079935,
      "epoch": 0.3114415673285375,
      "grad_norm": 0.16548436880111694,
      "learning_rate": 8.962186033673752e-05,
      "loss": 0.8556,
      "mean_token_accuracy": 0.7548238039016724,
      "num_tokens": 108022032.0,
      "step": 3385
    },
    {
      "entropy": 0.8222415596246719,
      "epoch": 0.31153357369997875,
      "grad_norm": 0.160286083817482,
      "learning_rate": 8.961879351059589e-05,
      "loss": 0.8153,
      "mean_token_accuracy": 0.760834563523531,
      "num_tokens": 108053502.0,
      "step": 3386
    },
    {
      "entropy": 0.8557467591017485,
      "epoch": 0.31162558007141994,
      "grad_norm": 0.16087114810943604,
      "learning_rate": 8.961572668445426e-05,
      "loss": 0.8414,
      "mean_token_accuracy": 0.7577027343213558,
      "num_tokens": 108085546.0,
      "step": 3387
    },
    {
      "entropy": 0.8516761548817158,
      "epoch": 0.3117175864428612,
      "grad_norm": 0.1561094969511032,
      "learning_rate": 8.961265985831264e-05,
      "loss": 0.8435,
      "mean_token_accuracy": 0.7559720575809479,
      "num_tokens": 108118090.0,
      "step": 3388
    },
    {
      "entropy": 0.7668294813483953,
      "epoch": 0.31180959281430237,
      "grad_norm": 0.15800561010837555,
      "learning_rate": 8.960959303217101e-05,
      "loss": 0.7683,
      "mean_token_accuracy": 0.7804623581469059,
      "num_tokens": 108149626.0,
      "step": 3389
    },
    {
      "entropy": 0.8711364343762398,
      "epoch": 0.3119015991857436,
      "grad_norm": 0.16084899008274078,
      "learning_rate": 8.960652620602939e-05,
      "loss": 0.8811,
      "mean_token_accuracy": 0.7477589920163155,
      "num_tokens": 108181627.0,
      "step": 3390
    },
    {
      "entropy": 0.8789335638284683,
      "epoch": 0.31199360555718486,
      "grad_norm": 0.16048380732536316,
      "learning_rate": 8.960345937988776e-05,
      "loss": 0.869,
      "mean_token_accuracy": 0.7501428350806236,
      "num_tokens": 108213515.0,
      "step": 3391
    },
    {
      "entropy": 0.8345394767820835,
      "epoch": 0.31208561192862605,
      "grad_norm": 0.1597360223531723,
      "learning_rate": 8.960039255374614e-05,
      "loss": 0.8409,
      "mean_token_accuracy": 0.7559681497514248,
      "num_tokens": 108245437.0,
      "step": 3392
    },
    {
      "entropy": 0.8228997774422169,
      "epoch": 0.3121776183000673,
      "grad_norm": 0.15864379703998566,
      "learning_rate": 8.95973257276045e-05,
      "loss": 0.8287,
      "mean_token_accuracy": 0.7638945803046227,
      "num_tokens": 108278027.0,
      "step": 3393
    },
    {
      "entropy": 0.8292393758893013,
      "epoch": 0.3122696246715085,
      "grad_norm": 0.15327489376068115,
      "learning_rate": 8.959425890146287e-05,
      "loss": 0.8118,
      "mean_token_accuracy": 0.7641171775758266,
      "num_tokens": 108310610.0,
      "step": 3394
    },
    {
      "entropy": 0.8673300929367542,
      "epoch": 0.31236163104294973,
      "grad_norm": 0.16689468920230865,
      "learning_rate": 8.959119207532125e-05,
      "loss": 0.8579,
      "mean_token_accuracy": 0.7529419288039207,
      "num_tokens": 108342449.0,
      "step": 3395
    },
    {
      "entropy": 0.8545701242983341,
      "epoch": 0.312453637414391,
      "grad_norm": 0.15719272196292877,
      "learning_rate": 8.958812524917964e-05,
      "loss": 0.841,
      "mean_token_accuracy": 0.7602478265762329,
      "num_tokens": 108373785.0,
      "step": 3396
    },
    {
      "entropy": 0.7741125803440809,
      "epoch": 0.31254564378583216,
      "grad_norm": 0.1587003469467163,
      "learning_rate": 8.9585058423038e-05,
      "loss": 0.7606,
      "mean_token_accuracy": 0.776952039450407,
      "num_tokens": 108406087.0,
      "step": 3397
    },
    {
      "entropy": 0.8209522627294064,
      "epoch": 0.3126376501572734,
      "grad_norm": 0.16815149784088135,
      "learning_rate": 8.958199159689637e-05,
      "loss": 0.8068,
      "mean_token_accuracy": 0.7657389119267464,
      "num_tokens": 108437343.0,
      "step": 3398
    },
    {
      "entropy": 0.8399774860590696,
      "epoch": 0.3127296565287146,
      "grad_norm": 0.16077417135238647,
      "learning_rate": 8.957892477075474e-05,
      "loss": 0.8516,
      "mean_token_accuracy": 0.7590476498007774,
      "num_tokens": 108469587.0,
      "step": 3399
    },
    {
      "entropy": 0.7874167989939451,
      "epoch": 0.31282166290015584,
      "grad_norm": 0.15489107370376587,
      "learning_rate": 8.957585794461313e-05,
      "loss": 0.7498,
      "mean_token_accuracy": 0.7814004495739937,
      "num_tokens": 108501440.0,
      "step": 3400
    },
    {
      "entropy": 0.7249480038881302,
      "epoch": 0.3129136692715971,
      "grad_norm": 0.14852961897850037,
      "learning_rate": 8.95727911184715e-05,
      "loss": 0.7113,
      "mean_token_accuracy": 0.7877282202243805,
      "num_tokens": 108533754.0,
      "step": 3401
    },
    {
      "entropy": 0.8292773254215717,
      "epoch": 0.3130056756430383,
      "grad_norm": 0.16443556547164917,
      "learning_rate": 8.956972429232987e-05,
      "loss": 0.8341,
      "mean_token_accuracy": 0.7582583054900169,
      "num_tokens": 108565300.0,
      "step": 3402
    },
    {
      "entropy": 0.8174557536840439,
      "epoch": 0.3130976820144795,
      "grad_norm": 0.17231027781963348,
      "learning_rate": 8.956665746618824e-05,
      "loss": 0.835,
      "mean_token_accuracy": 0.7617228738963604,
      "num_tokens": 108595529.0,
      "step": 3403
    },
    {
      "entropy": 0.9089966677129269,
      "epoch": 0.3131896883859207,
      "grad_norm": 0.16000422835350037,
      "learning_rate": 8.956359064004662e-05,
      "loss": 0.899,
      "mean_token_accuracy": 0.7424996346235275,
      "num_tokens": 108627359.0,
      "step": 3404
    },
    {
      "entropy": 0.7492682915180922,
      "epoch": 0.31328169475736195,
      "grad_norm": 0.15666493773460388,
      "learning_rate": 8.9560523813905e-05,
      "loss": 0.7399,
      "mean_token_accuracy": 0.7813226282596588,
      "num_tokens": 108659205.0,
      "step": 3405
    },
    {
      "entropy": 0.7816653475165367,
      "epoch": 0.3133737011288032,
      "grad_norm": 0.15676085650920868,
      "learning_rate": 8.955745698776337e-05,
      "loss": 0.7744,
      "mean_token_accuracy": 0.776084840297699,
      "num_tokens": 108690820.0,
      "step": 3406
    },
    {
      "entropy": 0.8218459617346525,
      "epoch": 0.3134657075002444,
      "grad_norm": 0.15737900137901306,
      "learning_rate": 8.955439016162174e-05,
      "loss": 0.8199,
      "mean_token_accuracy": 0.7669998444616795,
      "num_tokens": 108723380.0,
      "step": 3407
    },
    {
      "entropy": 0.774602685123682,
      "epoch": 0.31355771387168563,
      "grad_norm": 0.156219944357872,
      "learning_rate": 8.955132333548012e-05,
      "loss": 0.7567,
      "mean_token_accuracy": 0.773693036288023,
      "num_tokens": 108754305.0,
      "step": 3408
    },
    {
      "entropy": 0.8769824802875519,
      "epoch": 0.3136497202431268,
      "grad_norm": 0.16475169360637665,
      "learning_rate": 8.954825650933849e-05,
      "loss": 0.8583,
      "mean_token_accuracy": 0.7550069615244865,
      "num_tokens": 108785750.0,
      "step": 3409
    },
    {
      "entropy": 0.7020296547561884,
      "epoch": 0.31374172661456806,
      "grad_norm": 0.15911468863487244,
      "learning_rate": 8.954518968319687e-05,
      "loss": 0.7012,
      "mean_token_accuracy": 0.7991133071482182,
      "num_tokens": 108817806.0,
      "step": 3410
    },
    {
      "entropy": 0.885748652741313,
      "epoch": 0.3138337329860093,
      "grad_norm": 0.16307953000068665,
      "learning_rate": 8.954212285705524e-05,
      "loss": 0.8794,
      "mean_token_accuracy": 0.7499202415347099,
      "num_tokens": 108850071.0,
      "step": 3411
    },
    {
      "entropy": 0.9413866214454174,
      "epoch": 0.3139257393574505,
      "grad_norm": 0.1667211502790451,
      "learning_rate": 8.953905603091362e-05,
      "loss": 0.9406,
      "mean_token_accuracy": 0.7375316135585308,
      "num_tokens": 108882091.0,
      "step": 3412
    },
    {
      "entropy": 0.7985147498548031,
      "epoch": 0.31401774572889174,
      "grad_norm": 0.15284083783626556,
      "learning_rate": 8.953598920477198e-05,
      "loss": 0.7967,
      "mean_token_accuracy": 0.7665813155472279,
      "num_tokens": 108914284.0,
      "step": 3413
    },
    {
      "entropy": 0.7827238384634256,
      "epoch": 0.31410975210033293,
      "grad_norm": 0.161714568734169,
      "learning_rate": 8.953292237863035e-05,
      "loss": 0.7675,
      "mean_token_accuracy": 0.7779492922127247,
      "num_tokens": 108946514.0,
      "step": 3414
    },
    {
      "entropy": 0.79734867811203,
      "epoch": 0.3142017584717742,
      "grad_norm": 0.15360012650489807,
      "learning_rate": 8.952985555248873e-05,
      "loss": 0.7806,
      "mean_token_accuracy": 0.7711176313459873,
      "num_tokens": 108977942.0,
      "step": 3415
    },
    {
      "entropy": 0.8050581105053425,
      "epoch": 0.3142937648432154,
      "grad_norm": 0.16392603516578674,
      "learning_rate": 8.952678872634712e-05,
      "loss": 0.7967,
      "mean_token_accuracy": 0.7643947154283524,
      "num_tokens": 109009876.0,
      "step": 3416
    },
    {
      "entropy": 0.8860770743340254,
      "epoch": 0.3143857712146566,
      "grad_norm": 0.15949280560016632,
      "learning_rate": 8.952372190020548e-05,
      "loss": 0.8733,
      "mean_token_accuracy": 0.7473620660603046,
      "num_tokens": 109042403.0,
      "step": 3417
    },
    {
      "entropy": 0.8013679161667824,
      "epoch": 0.31447777758609785,
      "grad_norm": 0.15772996842861176,
      "learning_rate": 8.952065507406385e-05,
      "loss": 0.7906,
      "mean_token_accuracy": 0.7677351757884026,
      "num_tokens": 109074018.0,
      "step": 3418
    },
    {
      "entropy": 0.8261696547269821,
      "epoch": 0.31456978395753904,
      "grad_norm": 0.15970836579799652,
      "learning_rate": 8.951758824792223e-05,
      "loss": 0.8085,
      "mean_token_accuracy": 0.7622645050287247,
      "num_tokens": 109105537.0,
      "step": 3419
    },
    {
      "entropy": 0.9710891451686621,
      "epoch": 0.3146617903289803,
      "grad_norm": 0.16195394098758698,
      "learning_rate": 8.95145214217806e-05,
      "loss": 0.9702,
      "mean_token_accuracy": 0.7267987355589867,
      "num_tokens": 109137858.0,
      "step": 3420
    },
    {
      "entropy": 0.7899205461144447,
      "epoch": 0.31475379670042153,
      "grad_norm": 0.16411779820919037,
      "learning_rate": 8.951145459563898e-05,
      "loss": 0.7969,
      "mean_token_accuracy": 0.7686477825045586,
      "num_tokens": 109169496.0,
      "step": 3421
    },
    {
      "entropy": 0.7515205442905426,
      "epoch": 0.3148458030718627,
      "grad_norm": 0.15749630331993103,
      "learning_rate": 8.950838776949735e-05,
      "loss": 0.7441,
      "mean_token_accuracy": 0.7807129956781864,
      "num_tokens": 109200908.0,
      "step": 3422
    },
    {
      "entropy": 0.7766888216137886,
      "epoch": 0.31493780944330396,
      "grad_norm": 0.15730245411396027,
      "learning_rate": 8.950532094335573e-05,
      "loss": 0.7885,
      "mean_token_accuracy": 0.7683335468173027,
      "num_tokens": 109233009.0,
      "step": 3423
    },
    {
      "entropy": 0.8106716051697731,
      "epoch": 0.31502981581474515,
      "grad_norm": 0.16227684915065765,
      "learning_rate": 8.95022541172141e-05,
      "loss": 0.7942,
      "mean_token_accuracy": 0.7692654058337212,
      "num_tokens": 109264616.0,
      "step": 3424
    },
    {
      "entropy": 0.8341722171753645,
      "epoch": 0.3151218221861864,
      "grad_norm": 0.15843293070793152,
      "learning_rate": 8.949918729107247e-05,
      "loss": 0.8495,
      "mean_token_accuracy": 0.7554115131497383,
      "num_tokens": 109297360.0,
      "step": 3425
    },
    {
      "entropy": 0.7868106588721275,
      "epoch": 0.31521382855762764,
      "grad_norm": 0.16203047335147858,
      "learning_rate": 8.949612046493085e-05,
      "loss": 0.7876,
      "mean_token_accuracy": 0.7723313830792904,
      "num_tokens": 109328396.0,
      "step": 3426
    },
    {
      "entropy": 0.7971117328852415,
      "epoch": 0.31530583492906883,
      "grad_norm": 0.16363993287086487,
      "learning_rate": 8.949305363878923e-05,
      "loss": 0.7896,
      "mean_token_accuracy": 0.7721246965229511,
      "num_tokens": 109360153.0,
      "step": 3427
    },
    {
      "entropy": 0.7811590675264597,
      "epoch": 0.3153978413005101,
      "grad_norm": 0.16658936440944672,
      "learning_rate": 8.94899868126476e-05,
      "loss": 0.7854,
      "mean_token_accuracy": 0.7749624513089657,
      "num_tokens": 109391896.0,
      "step": 3428
    },
    {
      "entropy": 0.7023147456347942,
      "epoch": 0.31548984767195126,
      "grad_norm": 0.15011928975582123,
      "learning_rate": 8.948691998650597e-05,
      "loss": 0.6886,
      "mean_token_accuracy": 0.7978311739861965,
      "num_tokens": 109423152.0,
      "step": 3429
    },
    {
      "entropy": 0.8084341306239367,
      "epoch": 0.3155818540433925,
      "grad_norm": 0.1531917154788971,
      "learning_rate": 8.948385316036433e-05,
      "loss": 0.8004,
      "mean_token_accuracy": 0.7677325420081615,
      "num_tokens": 109455065.0,
      "step": 3430
    },
    {
      "entropy": 0.9036215152591467,
      "epoch": 0.31567386041483375,
      "grad_norm": 0.16358636319637299,
      "learning_rate": 8.948078633422271e-05,
      "loss": 0.8802,
      "mean_token_accuracy": 0.7472815290093422,
      "num_tokens": 109486170.0,
      "step": 3431
    },
    {
      "entropy": 0.8911346234381199,
      "epoch": 0.31576586678627494,
      "grad_norm": 0.16302113234996796,
      "learning_rate": 8.94777195080811e-05,
      "loss": 0.8951,
      "mean_token_accuracy": 0.7453261017799377,
      "num_tokens": 109517229.0,
      "step": 3432
    },
    {
      "entropy": 0.8314700517803431,
      "epoch": 0.3158578731577162,
      "grad_norm": 0.15679825842380524,
      "learning_rate": 8.947465268193946e-05,
      "loss": 0.8138,
      "mean_token_accuracy": 0.7630833238363266,
      "num_tokens": 109549370.0,
      "step": 3433
    },
    {
      "entropy": 0.8138379380106926,
      "epoch": 0.3159498795291574,
      "grad_norm": 0.15530484914779663,
      "learning_rate": 8.947158585579783e-05,
      "loss": 0.8217,
      "mean_token_accuracy": 0.7628655731678009,
      "num_tokens": 109581171.0,
      "step": 3434
    },
    {
      "entropy": 0.8425354920327663,
      "epoch": 0.3160418859005986,
      "grad_norm": 0.1537533700466156,
      "learning_rate": 8.946851902965621e-05,
      "loss": 0.8405,
      "mean_token_accuracy": 0.7589230947196484,
      "num_tokens": 109613504.0,
      "step": 3435
    },
    {
      "entropy": 0.8536132611334324,
      "epoch": 0.31613389227203986,
      "grad_norm": 0.1595718413591385,
      "learning_rate": 8.946545220351458e-05,
      "loss": 0.8703,
      "mean_token_accuracy": 0.7467762492597103,
      "num_tokens": 109646035.0,
      "step": 3436
    },
    {
      "entropy": 0.8575638718903065,
      "epoch": 0.31622589864348105,
      "grad_norm": 0.148703932762146,
      "learning_rate": 8.946238537737296e-05,
      "loss": 0.8521,
      "mean_token_accuracy": 0.7513237930834293,
      "num_tokens": 109677948.0,
      "step": 3437
    },
    {
      "entropy": 0.7438035290688276,
      "epoch": 0.3163179050149223,
      "grad_norm": 0.14720644056797028,
      "learning_rate": 8.945931855123133e-05,
      "loss": 0.7296,
      "mean_token_accuracy": 0.7877347208559513,
      "num_tokens": 109709940.0,
      "step": 3438
    },
    {
      "entropy": 0.7959811296314001,
      "epoch": 0.3164099113863635,
      "grad_norm": 0.15656942129135132,
      "learning_rate": 8.945625172508971e-05,
      "loss": 0.7882,
      "mean_token_accuracy": 0.7710674107074738,
      "num_tokens": 109742708.0,
      "step": 3439
    },
    {
      "entropy": 0.8903893232345581,
      "epoch": 0.31650191775780473,
      "grad_norm": 0.16482454538345337,
      "learning_rate": 8.945318489894808e-05,
      "loss": 0.8739,
      "mean_token_accuracy": 0.7485821358859539,
      "num_tokens": 109773673.0,
      "step": 3440
    },
    {
      "entropy": 0.7849636003375053,
      "epoch": 0.316593924129246,
      "grad_norm": 0.15752696990966797,
      "learning_rate": 8.945011807280646e-05,
      "loss": 0.7602,
      "mean_token_accuracy": 0.7742185704410076,
      "num_tokens": 109804842.0,
      "step": 3441
    },
    {
      "entropy": 0.754973292350769,
      "epoch": 0.31668593050068716,
      "grad_norm": 0.15343701839447021,
      "learning_rate": 8.944705124666483e-05,
      "loss": 0.7311,
      "mean_token_accuracy": 0.7819373346865177,
      "num_tokens": 109836368.0,
      "step": 3442
    },
    {
      "entropy": 0.850654061883688,
      "epoch": 0.3167779368721284,
      "grad_norm": 0.16695885360240936,
      "learning_rate": 8.944398442052321e-05,
      "loss": 0.8481,
      "mean_token_accuracy": 0.7580230347812176,
      "num_tokens": 109868112.0,
      "step": 3443
    },
    {
      "entropy": 0.7854941710829735,
      "epoch": 0.3168699432435696,
      "grad_norm": 0.15840871632099152,
      "learning_rate": 8.944091759438158e-05,
      "loss": 0.7759,
      "mean_token_accuracy": 0.7762467972934246,
      "num_tokens": 109900025.0,
      "step": 3444
    },
    {
      "entropy": 0.8972769118845463,
      "epoch": 0.31696194961501084,
      "grad_norm": 0.1634475737810135,
      "learning_rate": 8.943785076823995e-05,
      "loss": 0.9094,
      "mean_token_accuracy": 0.7431809529662132,
      "num_tokens": 109931903.0,
      "step": 3445
    },
    {
      "entropy": 0.7820693664252758,
      "epoch": 0.3170539559864521,
      "grad_norm": 0.15861153602600098,
      "learning_rate": 8.943478394209833e-05,
      "loss": 0.8021,
      "mean_token_accuracy": 0.768190149217844,
      "num_tokens": 109963893.0,
      "step": 3446
    },
    {
      "entropy": 0.7859980911016464,
      "epoch": 0.3171459623578933,
      "grad_norm": 0.1600302755832672,
      "learning_rate": 8.943171711595671e-05,
      "loss": 0.7814,
      "mean_token_accuracy": 0.7759186439216137,
      "num_tokens": 109996508.0,
      "step": 3447
    },
    {
      "entropy": 0.7639955468475819,
      "epoch": 0.3172379687293345,
      "grad_norm": 0.1620306521654129,
      "learning_rate": 8.942865028981508e-05,
      "loss": 0.7759,
      "mean_token_accuracy": 0.7763983085751534,
      "num_tokens": 110028350.0,
      "step": 3448
    },
    {
      "entropy": 0.89695655554533,
      "epoch": 0.3173299751007757,
      "grad_norm": 0.15601515769958496,
      "learning_rate": 8.942558346367345e-05,
      "loss": 0.8828,
      "mean_token_accuracy": 0.743054922670126,
      "num_tokens": 110060546.0,
      "step": 3449
    },
    {
      "entropy": 0.8009643964469433,
      "epoch": 0.31742198147221695,
      "grad_norm": 0.1575240194797516,
      "learning_rate": 8.942251663753183e-05,
      "loss": 0.7951,
      "mean_token_accuracy": 0.7678246945142746,
      "num_tokens": 110092575.0,
      "step": 3450
    },
    {
      "entropy": 0.8117531165480614,
      "epoch": 0.3175139878436582,
      "grad_norm": 0.1526649445295334,
      "learning_rate": 8.94194498113902e-05,
      "loss": 0.8124,
      "mean_token_accuracy": 0.7645686119794846,
      "num_tokens": 110124985.0,
      "step": 3451
    },
    {
      "entropy": 0.8540495838969946,
      "epoch": 0.3176059942150994,
      "grad_norm": 0.15577557682991028,
      "learning_rate": 8.941638298524858e-05,
      "loss": 0.8407,
      "mean_token_accuracy": 0.7578328773379326,
      "num_tokens": 110157000.0,
      "step": 3452
    },
    {
      "entropy": 0.8964856266975403,
      "epoch": 0.31769800058654063,
      "grad_norm": 0.1592482179403305,
      "learning_rate": 8.941331615910694e-05,
      "loss": 0.8936,
      "mean_token_accuracy": 0.7450084909796715,
      "num_tokens": 110189038.0,
      "step": 3453
    },
    {
      "entropy": 0.771185178309679,
      "epoch": 0.3177900069579818,
      "grad_norm": 0.15692219138145447,
      "learning_rate": 8.941024933296533e-05,
      "loss": 0.7639,
      "mean_token_accuracy": 0.776179626584053,
      "num_tokens": 110219982.0,
      "step": 3454
    },
    {
      "entropy": 0.8270531259477139,
      "epoch": 0.31788201332942306,
      "grad_norm": 0.15755802392959595,
      "learning_rate": 8.940718250682369e-05,
      "loss": 0.8247,
      "mean_token_accuracy": 0.7578254379332066,
      "num_tokens": 110252170.0,
      "step": 3455
    },
    {
      "entropy": 0.8467460498213768,
      "epoch": 0.3179740197008643,
      "grad_norm": 0.1594618409872055,
      "learning_rate": 8.940411568068206e-05,
      "loss": 0.8459,
      "mean_token_accuracy": 0.7530449517071247,
      "num_tokens": 110283798.0,
      "step": 3456
    },
    {
      "entropy": 0.7305431459099054,
      "epoch": 0.3180660260723055,
      "grad_norm": 0.1463945358991623,
      "learning_rate": 8.940104885454044e-05,
      "loss": 0.7016,
      "mean_token_accuracy": 0.793040819466114,
      "num_tokens": 110315813.0,
      "step": 3457
    },
    {
      "entropy": 0.9113414920866489,
      "epoch": 0.31815803244374674,
      "grad_norm": 0.16249766945838928,
      "learning_rate": 8.939798202839882e-05,
      "loss": 0.9153,
      "mean_token_accuracy": 0.742634691298008,
      "num_tokens": 110347286.0,
      "step": 3458
    },
    {
      "entropy": 0.9176751039922237,
      "epoch": 0.31825003881518793,
      "grad_norm": 0.1625015288591385,
      "learning_rate": 8.939491520225719e-05,
      "loss": 0.9226,
      "mean_token_accuracy": 0.7373808845877647,
      "num_tokens": 110378303.0,
      "step": 3459
    },
    {
      "entropy": 0.6770880706608295,
      "epoch": 0.3183420451866292,
      "grad_norm": 0.14814499020576477,
      "learning_rate": 8.939184837611556e-05,
      "loss": 0.6582,
      "mean_token_accuracy": 0.8077810183167458,
      "num_tokens": 110410308.0,
      "step": 3460
    },
    {
      "entropy": 0.9036635607481003,
      "epoch": 0.3184340515580704,
      "grad_norm": 0.16205953061580658,
      "learning_rate": 8.938878154997393e-05,
      "loss": 0.9194,
      "mean_token_accuracy": 0.7385397180914879,
      "num_tokens": 110442674.0,
      "step": 3461
    },
    {
      "entropy": 0.8653042055666447,
      "epoch": 0.3185260579295116,
      "grad_norm": 0.15921372175216675,
      "learning_rate": 8.938571472383231e-05,
      "loss": 0.8467,
      "mean_token_accuracy": 0.7554636783897877,
      "num_tokens": 110474911.0,
      "step": 3462
    },
    {
      "entropy": 0.7399308662861586,
      "epoch": 0.31861806430095285,
      "grad_norm": 0.15623636543750763,
      "learning_rate": 8.938264789769069e-05,
      "loss": 0.7217,
      "mean_token_accuracy": 0.7849855087697506,
      "num_tokens": 110507316.0,
      "step": 3463
    },
    {
      "entropy": 0.8639474343508482,
      "epoch": 0.31871007067239404,
      "grad_norm": 0.16977128386497498,
      "learning_rate": 8.937958107154906e-05,
      "loss": 0.8782,
      "mean_token_accuracy": 0.7500095851719379,
      "num_tokens": 110539855.0,
      "step": 3464
    },
    {
      "entropy": 0.7351997848600149,
      "epoch": 0.3188020770438353,
      "grad_norm": 0.157207652926445,
      "learning_rate": 8.937651424540743e-05,
      "loss": 0.7295,
      "mean_token_accuracy": 0.7833399474620819,
      "num_tokens": 110571474.0,
      "step": 3465
    },
    {
      "entropy": 0.912595372647047,
      "epoch": 0.31889408341527653,
      "grad_norm": 0.1673603057861328,
      "learning_rate": 8.937344741926581e-05,
      "loss": 0.9347,
      "mean_token_accuracy": 0.7377117834985256,
      "num_tokens": 110604097.0,
      "step": 3466
    },
    {
      "entropy": 0.8298111706972122,
      "epoch": 0.3189860897867177,
      "grad_norm": 0.15111568570137024,
      "learning_rate": 8.937038059312418e-05,
      "loss": 0.7933,
      "mean_token_accuracy": 0.7706094235181808,
      "num_tokens": 110636865.0,
      "step": 3467
    },
    {
      "entropy": 0.7930690050125122,
      "epoch": 0.31907809615815896,
      "grad_norm": 0.15420673787593842,
      "learning_rate": 8.936731376698256e-05,
      "loss": 0.7756,
      "mean_token_accuracy": 0.7751842215657234,
      "num_tokens": 110669080.0,
      "step": 3468
    },
    {
      "entropy": 0.838044686242938,
      "epoch": 0.31917010252960015,
      "grad_norm": 0.15985435247421265,
      "learning_rate": 8.936424694084092e-05,
      "loss": 0.8099,
      "mean_token_accuracy": 0.7662044130265713,
      "num_tokens": 110701489.0,
      "step": 3469
    },
    {
      "entropy": 0.7858495488762856,
      "epoch": 0.3192621089010414,
      "grad_norm": 0.1545124650001526,
      "learning_rate": 8.93611801146993e-05,
      "loss": 0.7705,
      "mean_token_accuracy": 0.7765480019152164,
      "num_tokens": 110733636.0,
      "step": 3470
    },
    {
      "entropy": 0.8793566785752773,
      "epoch": 0.31935411527248264,
      "grad_norm": 0.15899978578090668,
      "learning_rate": 8.935811328855767e-05,
      "loss": 0.893,
      "mean_token_accuracy": 0.7393013276159763,
      "num_tokens": 110765667.0,
      "step": 3471
    },
    {
      "entropy": 0.7493473067879677,
      "epoch": 0.31944612164392383,
      "grad_norm": 0.15404105186462402,
      "learning_rate": 8.935504646241604e-05,
      "loss": 0.7255,
      "mean_token_accuracy": 0.7856524996459484,
      "num_tokens": 110797293.0,
      "step": 3472
    },
    {
      "entropy": 0.9172775913029909,
      "epoch": 0.3195381280153651,
      "grad_norm": 0.16059572994709015,
      "learning_rate": 8.935197963627442e-05,
      "loss": 0.9134,
      "mean_token_accuracy": 0.7425948083400726,
      "num_tokens": 110829096.0,
      "step": 3473
    },
    {
      "entropy": 0.8422709237784147,
      "epoch": 0.31963013438680626,
      "grad_norm": 0.16786612570285797,
      "learning_rate": 8.93489128101328e-05,
      "loss": 0.8339,
      "mean_token_accuracy": 0.7546804435551167,
      "num_tokens": 110859754.0,
      "step": 3474
    },
    {
      "entropy": 0.8867340832948685,
      "epoch": 0.3197221407582475,
      "grad_norm": 0.15802140533924103,
      "learning_rate": 8.934584598399117e-05,
      "loss": 0.8771,
      "mean_token_accuracy": 0.7485387660562992,
      "num_tokens": 110892179.0,
      "step": 3475
    },
    {
      "entropy": 0.8563601039350033,
      "epoch": 0.31981414712968875,
      "grad_norm": 0.1600835770368576,
      "learning_rate": 8.934277915784954e-05,
      "loss": 0.8535,
      "mean_token_accuracy": 0.7549396641552448,
      "num_tokens": 110924357.0,
      "step": 3476
    },
    {
      "entropy": 0.8324606735259295,
      "epoch": 0.31990615350112994,
      "grad_norm": 0.1601134091615677,
      "learning_rate": 8.933971233170791e-05,
      "loss": 0.8366,
      "mean_token_accuracy": 0.759260967373848,
      "num_tokens": 110955666.0,
      "step": 3477
    },
    {
      "entropy": 0.7992579881101847,
      "epoch": 0.3199981598725712,
      "grad_norm": 0.15123245120048523,
      "learning_rate": 8.93366455055663e-05,
      "loss": 0.7973,
      "mean_token_accuracy": 0.7695608474314213,
      "num_tokens": 110987899.0,
      "step": 3478
    },
    {
      "entropy": 0.8116626888513565,
      "epoch": 0.3200901662440124,
      "grad_norm": 0.15879400074481964,
      "learning_rate": 8.933357867942467e-05,
      "loss": 0.8062,
      "mean_token_accuracy": 0.7640635073184967,
      "num_tokens": 111019442.0,
      "step": 3479
    },
    {
      "entropy": 0.815089039504528,
      "epoch": 0.3201821726154536,
      "grad_norm": 0.15871118009090424,
      "learning_rate": 8.933051185328304e-05,
      "loss": 0.8125,
      "mean_token_accuracy": 0.7650168091058731,
      "num_tokens": 111051702.0,
      "step": 3480
    },
    {
      "entropy": 0.8192152716219425,
      "epoch": 0.32027417898689486,
      "grad_norm": 0.14853739738464355,
      "learning_rate": 8.932744502714142e-05,
      "loss": 0.8127,
      "mean_token_accuracy": 0.766155794262886,
      "num_tokens": 111083752.0,
      "step": 3481
    },
    {
      "entropy": 0.8583752140402794,
      "epoch": 0.32036618535833605,
      "grad_norm": 0.16112826764583588,
      "learning_rate": 8.932437820099979e-05,
      "loss": 0.8403,
      "mean_token_accuracy": 0.7576774209737778,
      "num_tokens": 111115129.0,
      "step": 3482
    },
    {
      "entropy": 0.8250080049037933,
      "epoch": 0.3204581917297773,
      "grad_norm": 0.15579645335674286,
      "learning_rate": 8.932131137485817e-05,
      "loss": 0.7908,
      "mean_token_accuracy": 0.7734724916517735,
      "num_tokens": 111147335.0,
      "step": 3483
    },
    {
      "entropy": 0.8525599353015423,
      "epoch": 0.3205501981012185,
      "grad_norm": 0.15474143624305725,
      "learning_rate": 8.931824454871654e-05,
      "loss": 0.8387,
      "mean_token_accuracy": 0.7567358165979385,
      "num_tokens": 111179002.0,
      "step": 3484
    },
    {
      "entropy": 0.7751132249832153,
      "epoch": 0.32064220447265973,
      "grad_norm": 0.15190677344799042,
      "learning_rate": 8.931517772257492e-05,
      "loss": 0.759,
      "mean_token_accuracy": 0.7825722768902779,
      "num_tokens": 111210687.0,
      "step": 3485
    },
    {
      "entropy": 0.9125880636274815,
      "epoch": 0.320734210844101,
      "grad_norm": 0.16443994641304016,
      "learning_rate": 8.931211089643329e-05,
      "loss": 0.9096,
      "mean_token_accuracy": 0.7385830767452717,
      "num_tokens": 111242373.0,
      "step": 3486
    },
    {
      "entropy": 0.7925374712795019,
      "epoch": 0.32082621721554216,
      "grad_norm": 0.16025413572788239,
      "learning_rate": 8.930904407029165e-05,
      "loss": 0.7974,
      "mean_token_accuracy": 0.7744494155049324,
      "num_tokens": 111274974.0,
      "step": 3487
    },
    {
      "entropy": 0.8816979937255383,
      "epoch": 0.3209182235869834,
      "grad_norm": 0.16906802356243134,
      "learning_rate": 8.930597724415004e-05,
      "loss": 0.888,
      "mean_token_accuracy": 0.7486885711550713,
      "num_tokens": 111306190.0,
      "step": 3488
    },
    {
      "entropy": 0.7651477418839931,
      "epoch": 0.3210102299584246,
      "grad_norm": 0.16292248666286469,
      "learning_rate": 8.930291041800842e-05,
      "loss": 0.7828,
      "mean_token_accuracy": 0.7700636237859726,
      "num_tokens": 111338046.0,
      "step": 3489
    },
    {
      "entropy": 0.8288666047155857,
      "epoch": 0.32110223632986584,
      "grad_norm": 0.15535566210746765,
      "learning_rate": 8.929984359186679e-05,
      "loss": 0.8166,
      "mean_token_accuracy": 0.7632523365318775,
      "num_tokens": 111370526.0,
      "step": 3490
    },
    {
      "entropy": 0.7950088009238243,
      "epoch": 0.3211942427013071,
      "grad_norm": 0.15817055106163025,
      "learning_rate": 8.929677676572515e-05,
      "loss": 0.7921,
      "mean_token_accuracy": 0.7670589871704578,
      "num_tokens": 111402003.0,
      "step": 3491
    },
    {
      "entropy": 0.7908246032893658,
      "epoch": 0.3212862490727483,
      "grad_norm": 0.15442492067813873,
      "learning_rate": 8.929370993958352e-05,
      "loss": 0.795,
      "mean_token_accuracy": 0.7671869024634361,
      "num_tokens": 111434076.0,
      "step": 3492
    },
    {
      "entropy": 0.7186293508857489,
      "epoch": 0.3213782554441895,
      "grad_norm": 0.14667977392673492,
      "learning_rate": 8.92906431134419e-05,
      "loss": 0.729,
      "mean_token_accuracy": 0.7868706695735455,
      "num_tokens": 111466275.0,
      "step": 3493
    },
    {
      "entropy": 0.816513067111373,
      "epoch": 0.3214702618156307,
      "grad_norm": 0.15883442759513855,
      "learning_rate": 8.928757628730028e-05,
      "loss": 0.7946,
      "mean_token_accuracy": 0.7670467905700207,
      "num_tokens": 111496705.0,
      "step": 3494
    },
    {
      "entropy": 0.8788306880742311,
      "epoch": 0.32156226818707195,
      "grad_norm": 0.15579599142074585,
      "learning_rate": 8.928450946115865e-05,
      "loss": 0.8605,
      "mean_token_accuracy": 0.751182995736599,
      "num_tokens": 111529099.0,
      "step": 3495
    },
    {
      "entropy": 0.9049365520477295,
      "epoch": 0.3216542745585132,
      "grad_norm": 0.15226545929908752,
      "learning_rate": 8.928144263501702e-05,
      "loss": 0.879,
      "mean_token_accuracy": 0.7461988776922226,
      "num_tokens": 111561556.0,
      "step": 3496
    },
    {
      "entropy": 0.817472942173481,
      "epoch": 0.3217462809299544,
      "grad_norm": 0.1591382473707199,
      "learning_rate": 8.92783758088754e-05,
      "loss": 0.807,
      "mean_token_accuracy": 0.7643363811075687,
      "num_tokens": 111593392.0,
      "step": 3497
    },
    {
      "entropy": 0.8467215709388256,
      "epoch": 0.32183828730139563,
      "grad_norm": 0.1576875001192093,
      "learning_rate": 8.927530898273377e-05,
      "loss": 0.8362,
      "mean_token_accuracy": 0.7580426521599293,
      "num_tokens": 111625496.0,
      "step": 3498
    },
    {
      "entropy": 0.8946081697940826,
      "epoch": 0.3219302936728368,
      "grad_norm": 0.16749341785907745,
      "learning_rate": 8.927224215659215e-05,
      "loss": 0.891,
      "mean_token_accuracy": 0.7452874183654785,
      "num_tokens": 111657386.0,
      "step": 3499
    },
    {
      "entropy": 0.8084470666944981,
      "epoch": 0.32202230004427806,
      "grad_norm": 0.16346006095409393,
      "learning_rate": 8.926917533045052e-05,
      "loss": 0.8125,
      "mean_token_accuracy": 0.7639738097786903,
      "num_tokens": 111689531.0,
      "step": 3500
    },
    {
      "entropy": 0.799791282042861,
      "epoch": 0.3221143064157193,
      "grad_norm": 0.15397360920906067,
      "learning_rate": 8.92661085043089e-05,
      "loss": 0.7914,
      "mean_token_accuracy": 0.7676098756492138,
      "num_tokens": 111720915.0,
      "step": 3501
    },
    {
      "entropy": 0.7592886127531528,
      "epoch": 0.3222063127871605,
      "grad_norm": 0.15480400621891022,
      "learning_rate": 8.926304167816727e-05,
      "loss": 0.7333,
      "mean_token_accuracy": 0.7846937440335751,
      "num_tokens": 111752269.0,
      "step": 3502
    },
    {
      "entropy": 0.8349922448396683,
      "epoch": 0.32229831915860174,
      "grad_norm": 0.16307829320430756,
      "learning_rate": 8.925997485202564e-05,
      "loss": 0.8191,
      "mean_token_accuracy": 0.7641830630600452,
      "num_tokens": 111784621.0,
      "step": 3503
    },
    {
      "entropy": 0.8465308099985123,
      "epoch": 0.32239032553004293,
      "grad_norm": 0.1633448302745819,
      "learning_rate": 8.925690802588402e-05,
      "loss": 0.8555,
      "mean_token_accuracy": 0.7556760087609291,
      "num_tokens": 111816828.0,
      "step": 3504
    },
    {
      "entropy": 0.8387850373983383,
      "epoch": 0.3224823319014842,
      "grad_norm": 0.1602678745985031,
      "learning_rate": 8.92538411997424e-05,
      "loss": 0.8541,
      "mean_token_accuracy": 0.7550873458385468,
      "num_tokens": 111849170.0,
      "step": 3505
    },
    {
      "entropy": 0.768605325371027,
      "epoch": 0.3225743382729254,
      "grad_norm": 0.1608756184577942,
      "learning_rate": 8.925077437360077e-05,
      "loss": 0.7761,
      "mean_token_accuracy": 0.7756050117313862,
      "num_tokens": 111881470.0,
      "step": 3506
    },
    {
      "entropy": 0.9586869180202484,
      "epoch": 0.3226663446443666,
      "grad_norm": 0.16286087036132812,
      "learning_rate": 8.924770754745913e-05,
      "loss": 0.9717,
      "mean_token_accuracy": 0.7216954007744789,
      "num_tokens": 111913729.0,
      "step": 3507
    },
    {
      "entropy": 0.8235187102109194,
      "epoch": 0.32275835101580785,
      "grad_norm": 0.15379507839679718,
      "learning_rate": 8.92446407213175e-05,
      "loss": 0.8168,
      "mean_token_accuracy": 0.7663878388702869,
      "num_tokens": 111945757.0,
      "step": 3508
    },
    {
      "entropy": 0.8328001722693443,
      "epoch": 0.32285035738724904,
      "grad_norm": 0.15337993204593658,
      "learning_rate": 8.924157389517588e-05,
      "loss": 0.8116,
      "mean_token_accuracy": 0.7663005292415619,
      "num_tokens": 111977587.0,
      "step": 3509
    },
    {
      "entropy": 0.9084444344043732,
      "epoch": 0.3229423637586903,
      "grad_norm": 0.164095938205719,
      "learning_rate": 8.923850706903426e-05,
      "loss": 0.9008,
      "mean_token_accuracy": 0.7431821785867214,
      "num_tokens": 112009990.0,
      "step": 3510
    },
    {
      "entropy": 0.8985680546611547,
      "epoch": 0.32303437013013153,
      "grad_norm": 0.16000202298164368,
      "learning_rate": 8.923544024289263e-05,
      "loss": 0.8687,
      "mean_token_accuracy": 0.7556611560285091,
      "num_tokens": 112042201.0,
      "step": 3511
    },
    {
      "entropy": 0.7515249755233526,
      "epoch": 0.3231263765015727,
      "grad_norm": 0.15700507164001465,
      "learning_rate": 8.923237341675101e-05,
      "loss": 0.7336,
      "mean_token_accuracy": 0.7887502014636993,
      "num_tokens": 112074794.0,
      "step": 3512
    },
    {
      "entropy": 0.7538326028734446,
      "epoch": 0.32321838287301397,
      "grad_norm": 0.15205712616443634,
      "learning_rate": 8.922930659060938e-05,
      "loss": 0.7432,
      "mean_token_accuracy": 0.7784696035087109,
      "num_tokens": 112106445.0,
      "step": 3513
    },
    {
      "entropy": 0.7234333027154207,
      "epoch": 0.32331038924445515,
      "grad_norm": 0.15145999193191528,
      "learning_rate": 8.922623976446775e-05,
      "loss": 0.7331,
      "mean_token_accuracy": 0.7828084155917168,
      "num_tokens": 112138771.0,
      "step": 3514
    },
    {
      "entropy": 0.7778209578245878,
      "epoch": 0.3234023956158964,
      "grad_norm": 0.15707336366176605,
      "learning_rate": 8.922317293832613e-05,
      "loss": 0.7658,
      "mean_token_accuracy": 0.772449117153883,
      "num_tokens": 112170411.0,
      "step": 3515
    },
    {
      "entropy": 0.8654674328863621,
      "epoch": 0.32349440198733764,
      "grad_norm": 0.15499001741409302,
      "learning_rate": 8.922010611218451e-05,
      "loss": 0.8593,
      "mean_token_accuracy": 0.7483084164559841,
      "num_tokens": 112202381.0,
      "step": 3516
    },
    {
      "entropy": 0.7742456719279289,
      "epoch": 0.32358640835877883,
      "grad_norm": 0.16468708217144012,
      "learning_rate": 8.921703928604288e-05,
      "loss": 0.7878,
      "mean_token_accuracy": 0.7727333419024944,
      "num_tokens": 112234672.0,
      "step": 3517
    },
    {
      "entropy": 0.8356336914002895,
      "epoch": 0.3236784147302201,
      "grad_norm": 0.16786564886569977,
      "learning_rate": 8.921397245990125e-05,
      "loss": 0.8562,
      "mean_token_accuracy": 0.7618610896170139,
      "num_tokens": 112266866.0,
      "step": 3518
    },
    {
      "entropy": 0.7466015908867121,
      "epoch": 0.32377042110166127,
      "grad_norm": 0.16766873002052307,
      "learning_rate": 8.921090563375962e-05,
      "loss": 0.7689,
      "mean_token_accuracy": 0.7753879316151142,
      "num_tokens": 112298048.0,
      "step": 3519
    },
    {
      "entropy": 0.8910118490457535,
      "epoch": 0.3238624274731025,
      "grad_norm": 0.17390671372413635,
      "learning_rate": 8.920783880761801e-05,
      "loss": 0.8941,
      "mean_token_accuracy": 0.7430459000170231,
      "num_tokens": 112330133.0,
      "step": 3520
    },
    {
      "entropy": 0.8590236641466618,
      "epoch": 0.32395443384454375,
      "grad_norm": 0.1568404585123062,
      "learning_rate": 8.920477198147638e-05,
      "loss": 0.8593,
      "mean_token_accuracy": 0.7566804885864258,
      "num_tokens": 112362745.0,
      "step": 3521
    },
    {
      "entropy": 0.79797069914639,
      "epoch": 0.32404644021598494,
      "grad_norm": 0.15595680475234985,
      "learning_rate": 8.920170515533475e-05,
      "loss": 0.7803,
      "mean_token_accuracy": 0.7742975391447544,
      "num_tokens": 112394954.0,
      "step": 3522
    },
    {
      "entropy": 0.9912864379584789,
      "epoch": 0.3241384465874262,
      "grad_norm": 0.16705849766731262,
      "learning_rate": 8.919863832919311e-05,
      "loss": 0.9796,
      "mean_token_accuracy": 0.7240222990512848,
      "num_tokens": 112426996.0,
      "step": 3523
    },
    {
      "entropy": 0.8841212224215269,
      "epoch": 0.3242304529588674,
      "grad_norm": 0.16192105412483215,
      "learning_rate": 8.91955715030515e-05,
      "loss": 0.8658,
      "mean_token_accuracy": 0.7552276067435741,
      "num_tokens": 112459202.0,
      "step": 3524
    },
    {
      "entropy": 0.8577829916030169,
      "epoch": 0.3243224593303086,
      "grad_norm": 0.1575586348772049,
      "learning_rate": 8.919250467690988e-05,
      "loss": 0.8384,
      "mean_token_accuracy": 0.7587824277579784,
      "num_tokens": 112491334.0,
      "step": 3525
    },
    {
      "entropy": 0.8967768251895905,
      "epoch": 0.32441446570174987,
      "grad_norm": 0.15570402145385742,
      "learning_rate": 8.918943785076825e-05,
      "loss": 0.8785,
      "mean_token_accuracy": 0.7474039271473885,
      "num_tokens": 112523596.0,
      "step": 3526
    },
    {
      "entropy": 0.8728480581194162,
      "epoch": 0.32450647207319105,
      "grad_norm": 0.15966250002384186,
      "learning_rate": 8.918637102462661e-05,
      "loss": 0.8518,
      "mean_token_accuracy": 0.7563502043485641,
      "num_tokens": 112554982.0,
      "step": 3527
    },
    {
      "entropy": 0.7137563265860081,
      "epoch": 0.3245984784446323,
      "grad_norm": 0.14698055386543274,
      "learning_rate": 8.9183304198485e-05,
      "loss": 0.7086,
      "mean_token_accuracy": 0.7924289926886559,
      "num_tokens": 112587355.0,
      "step": 3528
    },
    {
      "entropy": 0.8325293138623238,
      "epoch": 0.3246904848160735,
      "grad_norm": 0.1586577147245407,
      "learning_rate": 8.918023737234336e-05,
      "loss": 0.8126,
      "mean_token_accuracy": 0.7621361128985882,
      "num_tokens": 112619706.0,
      "step": 3529
    },
    {
      "entropy": 0.7534335367381573,
      "epoch": 0.32478249118751473,
      "grad_norm": 0.15479594469070435,
      "learning_rate": 8.917717054620174e-05,
      "loss": 0.7409,
      "mean_token_accuracy": 0.7827999591827393,
      "num_tokens": 112650577.0,
      "step": 3530
    },
    {
      "entropy": 0.8307875357568264,
      "epoch": 0.324874497558956,
      "grad_norm": 0.1595810055732727,
      "learning_rate": 8.917410372006011e-05,
      "loss": 0.8322,
      "mean_token_accuracy": 0.7587959840893745,
      "num_tokens": 112682508.0,
      "step": 3531
    },
    {
      "entropy": 0.7566315103322268,
      "epoch": 0.32496650393039717,
      "grad_norm": 0.15443359315395355,
      "learning_rate": 8.91710368939185e-05,
      "loss": 0.7391,
      "mean_token_accuracy": 0.7899300530552864,
      "num_tokens": 112714251.0,
      "step": 3532
    },
    {
      "entropy": 0.8317952565848827,
      "epoch": 0.3250585103018384,
      "grad_norm": 0.16431604325771332,
      "learning_rate": 8.916797006777686e-05,
      "loss": 0.8217,
      "mean_token_accuracy": 0.7693342231214046,
      "num_tokens": 112746129.0,
      "step": 3533
    },
    {
      "entropy": 0.8670476265251637,
      "epoch": 0.3251505166732796,
      "grad_norm": 0.16507373750209808,
      "learning_rate": 8.916490324163523e-05,
      "loss": 0.8813,
      "mean_token_accuracy": 0.7459238618612289,
      "num_tokens": 112778165.0,
      "step": 3534
    },
    {
      "entropy": 0.8264820910990238,
      "epoch": 0.32524252304472084,
      "grad_norm": 0.15872600674629211,
      "learning_rate": 8.916183641549361e-05,
      "loss": 0.8328,
      "mean_token_accuracy": 0.7564338743686676,
      "num_tokens": 112810001.0,
      "step": 3535
    },
    {
      "entropy": 0.7799114724621177,
      "epoch": 0.3253345294161621,
      "grad_norm": 0.15547290444374084,
      "learning_rate": 8.915876958935199e-05,
      "loss": 0.7819,
      "mean_token_accuracy": 0.7712539434432983,
      "num_tokens": 112842077.0,
      "step": 3536
    },
    {
      "entropy": 0.7806409597396851,
      "epoch": 0.3254265357876033,
      "grad_norm": 0.1589875966310501,
      "learning_rate": 8.915570276321036e-05,
      "loss": 0.7979,
      "mean_token_accuracy": 0.7728253044188023,
      "num_tokens": 112874113.0,
      "step": 3537
    },
    {
      "entropy": 0.8154195994138718,
      "epoch": 0.3255185421590445,
      "grad_norm": 0.1644035279750824,
      "learning_rate": 8.915263593706873e-05,
      "loss": 0.818,
      "mean_token_accuracy": 0.7612890712916851,
      "num_tokens": 112905771.0,
      "step": 3538
    },
    {
      "entropy": 0.8208198118954897,
      "epoch": 0.3256105485304857,
      "grad_norm": 0.15620583295822144,
      "learning_rate": 8.91495691109271e-05,
      "loss": 0.801,
      "mean_token_accuracy": 0.7692325785756111,
      "num_tokens": 112937381.0,
      "step": 3539
    },
    {
      "entropy": 0.9078818894922733,
      "epoch": 0.32570255490192696,
      "grad_norm": 0.16435173153877258,
      "learning_rate": 8.914650228478548e-05,
      "loss": 0.9085,
      "mean_token_accuracy": 0.7413159050047398,
      "num_tokens": 112969631.0,
      "step": 3540
    },
    {
      "entropy": 0.7629182226955891,
      "epoch": 0.3257945612733682,
      "grad_norm": 0.16151109337806702,
      "learning_rate": 8.914343545864386e-05,
      "loss": 0.7644,
      "mean_token_accuracy": 0.7763059847056866,
      "num_tokens": 113001877.0,
      "step": 3541
    },
    {
      "entropy": 0.8164016157388687,
      "epoch": 0.3258865676448094,
      "grad_norm": 0.16197903454303741,
      "learning_rate": 8.914036863250223e-05,
      "loss": 0.8168,
      "mean_token_accuracy": 0.7637383826076984,
      "num_tokens": 113034429.0,
      "step": 3542
    },
    {
      "entropy": 0.8725016843527555,
      "epoch": 0.32597857401625063,
      "grad_norm": 0.16132739186286926,
      "learning_rate": 8.913730180636061e-05,
      "loss": 0.872,
      "mean_token_accuracy": 0.7490272372961044,
      "num_tokens": 113065668.0,
      "step": 3543
    },
    {
      "entropy": 0.7546561975032091,
      "epoch": 0.3260705803876918,
      "grad_norm": 0.15967683494091034,
      "learning_rate": 8.913423498021898e-05,
      "loss": 0.734,
      "mean_token_accuracy": 0.7878193147480488,
      "num_tokens": 113098189.0,
      "step": 3544
    },
    {
      "entropy": 0.8953891638666391,
      "epoch": 0.32616258675913307,
      "grad_norm": 0.16531096398830414,
      "learning_rate": 8.913116815407734e-05,
      "loss": 0.8805,
      "mean_token_accuracy": 0.7517305463552475,
      "num_tokens": 113130686.0,
      "step": 3545
    },
    {
      "entropy": 0.8721664287149906,
      "epoch": 0.3262545931305743,
      "grad_norm": 0.16372916102409363,
      "learning_rate": 8.912810132793572e-05,
      "loss": 0.8557,
      "mean_token_accuracy": 0.7556580044329166,
      "num_tokens": 113162599.0,
      "step": 3546
    },
    {
      "entropy": 0.8287416324019432,
      "epoch": 0.3263465995020155,
      "grad_norm": 0.1624942570924759,
      "learning_rate": 8.91250345017941e-05,
      "loss": 0.8272,
      "mean_token_accuracy": 0.7634267359972,
      "num_tokens": 113194321.0,
      "step": 3547
    },
    {
      "entropy": 0.8132980167865753,
      "epoch": 0.32643860587345674,
      "grad_norm": 0.15723322331905365,
      "learning_rate": 8.912196767565247e-05,
      "loss": 0.8059,
      "mean_token_accuracy": 0.7698186300694942,
      "num_tokens": 113226120.0,
      "step": 3548
    },
    {
      "entropy": 0.8703325651586056,
      "epoch": 0.32653061224489793,
      "grad_norm": 0.1609811782836914,
      "learning_rate": 8.911890084951084e-05,
      "loss": 0.8474,
      "mean_token_accuracy": 0.748808678239584,
      "num_tokens": 113257204.0,
      "step": 3549
    },
    {
      "entropy": 0.8273969013243914,
      "epoch": 0.3266226186163392,
      "grad_norm": 0.15259526669979095,
      "learning_rate": 8.911583402336921e-05,
      "loss": 0.8165,
      "mean_token_accuracy": 0.7636073641479015,
      "num_tokens": 113288943.0,
      "step": 3550
    },
    {
      "entropy": 0.8341647144407034,
      "epoch": 0.3267146249877804,
      "grad_norm": 0.1599828004837036,
      "learning_rate": 8.911276719722759e-05,
      "loss": 0.8184,
      "mean_token_accuracy": 0.7662413492798805,
      "num_tokens": 113320947.0,
      "step": 3551
    },
    {
      "entropy": 0.820491585880518,
      "epoch": 0.3268066313592216,
      "grad_norm": 0.16254858672618866,
      "learning_rate": 8.910970037108597e-05,
      "loss": 0.8306,
      "mean_token_accuracy": 0.7629505433142185,
      "num_tokens": 113353391.0,
      "step": 3552
    },
    {
      "entropy": 0.8428402431309223,
      "epoch": 0.32689863773066286,
      "grad_norm": 0.1617119312286377,
      "learning_rate": 8.910663354494434e-05,
      "loss": 0.85,
      "mean_token_accuracy": 0.7559737041592598,
      "num_tokens": 113385617.0,
      "step": 3553
    },
    {
      "entropy": 0.8835074827075005,
      "epoch": 0.32699064410210404,
      "grad_norm": 0.1695723831653595,
      "learning_rate": 8.910356671880271e-05,
      "loss": 0.9011,
      "mean_token_accuracy": 0.7410794869065285,
      "num_tokens": 113417977.0,
      "step": 3554
    },
    {
      "entropy": 0.7775358464568853,
      "epoch": 0.3270826504735453,
      "grad_norm": 0.14951877295970917,
      "learning_rate": 8.910049989266109e-05,
      "loss": 0.7691,
      "mean_token_accuracy": 0.7722590006887913,
      "num_tokens": 113450039.0,
      "step": 3555
    },
    {
      "entropy": 0.838826522231102,
      "epoch": 0.32717465684498653,
      "grad_norm": 0.1645222306251526,
      "learning_rate": 8.909743306651946e-05,
      "loss": 0.842,
      "mean_token_accuracy": 0.7585185430943966,
      "num_tokens": 113481874.0,
      "step": 3556
    },
    {
      "entropy": 0.785923546180129,
      "epoch": 0.3272666632164277,
      "grad_norm": 0.14948385953903198,
      "learning_rate": 8.909436624037784e-05,
      "loss": 0.7845,
      "mean_token_accuracy": 0.7750485092401505,
      "num_tokens": 113514629.0,
      "step": 3557
    },
    {
      "entropy": 0.7786267250776291,
      "epoch": 0.32735866958786897,
      "grad_norm": 0.16133369505405426,
      "learning_rate": 8.909129941423621e-05,
      "loss": 0.7737,
      "mean_token_accuracy": 0.7739647403359413,
      "num_tokens": 113546842.0,
      "step": 3558
    },
    {
      "entropy": 0.8628189750015736,
      "epoch": 0.32745067595931016,
      "grad_norm": 0.1634611338376999,
      "learning_rate": 8.908823258809459e-05,
      "loss": 0.847,
      "mean_token_accuracy": 0.7572952173650265,
      "num_tokens": 113577723.0,
      "step": 3559
    },
    {
      "entropy": 0.8892375193536282,
      "epoch": 0.3275426823307514,
      "grad_norm": 0.15835101902484894,
      "learning_rate": 8.908516576195296e-05,
      "loss": 0.86,
      "mean_token_accuracy": 0.7573098875582218,
      "num_tokens": 113610257.0,
      "step": 3560
    },
    {
      "entropy": 0.8624659813940525,
      "epoch": 0.32763468870219264,
      "grad_norm": 0.16422800719738007,
      "learning_rate": 8.908209893581134e-05,
      "loss": 0.8693,
      "mean_token_accuracy": 0.7504704967141151,
      "num_tokens": 113642156.0,
      "step": 3561
    },
    {
      "entropy": 0.7905363515019417,
      "epoch": 0.32772669507363383,
      "grad_norm": 0.15282107889652252,
      "learning_rate": 8.90790321096697e-05,
      "loss": 0.7884,
      "mean_token_accuracy": 0.7761432118713856,
      "num_tokens": 113673694.0,
      "step": 3562
    },
    {
      "entropy": 0.9458177946507931,
      "epoch": 0.3278187014450751,
      "grad_norm": 0.164636492729187,
      "learning_rate": 8.907596528352809e-05,
      "loss": 0.9374,
      "mean_token_accuracy": 0.7321090772747993,
      "num_tokens": 113705255.0,
      "step": 3563
    },
    {
      "entropy": 0.8458957672119141,
      "epoch": 0.32791070781651627,
      "grad_norm": 0.1638212651014328,
      "learning_rate": 8.907289845738645e-05,
      "loss": 0.8377,
      "mean_token_accuracy": 0.7598610520362854,
      "num_tokens": 113737125.0,
      "step": 3564
    },
    {
      "entropy": 0.9331359639763832,
      "epoch": 0.3280027141879575,
      "grad_norm": 0.15972955524921417,
      "learning_rate": 8.906983163124482e-05,
      "loss": 0.9265,
      "mean_token_accuracy": 0.7373395785689354,
      "num_tokens": 113769752.0,
      "step": 3565
    },
    {
      "entropy": 0.7610961087048054,
      "epoch": 0.32809472055939876,
      "grad_norm": 0.16100288927555084,
      "learning_rate": 8.90667648051032e-05,
      "loss": 0.7753,
      "mean_token_accuracy": 0.775166928768158,
      "num_tokens": 113802219.0,
      "step": 3566
    },
    {
      "entropy": 0.8443037420511246,
      "epoch": 0.32818672693083994,
      "grad_norm": 0.15648706257343292,
      "learning_rate": 8.906369797896159e-05,
      "loss": 0.8454,
      "mean_token_accuracy": 0.7555993236601353,
      "num_tokens": 113834284.0,
      "step": 3567
    },
    {
      "entropy": 0.7570295743644238,
      "epoch": 0.3282787333022812,
      "grad_norm": 0.15555135905742645,
      "learning_rate": 8.906063115281995e-05,
      "loss": 0.7443,
      "mean_token_accuracy": 0.7834891565144062,
      "num_tokens": 113866938.0,
      "step": 3568
    },
    {
      "entropy": 0.8487836010754108,
      "epoch": 0.3283707396737224,
      "grad_norm": 0.16768993437290192,
      "learning_rate": 8.905756432667832e-05,
      "loss": 0.8522,
      "mean_token_accuracy": 0.7549125514924526,
      "num_tokens": 113898630.0,
      "step": 3569
    },
    {
      "entropy": 0.8674600664526224,
      "epoch": 0.3284627460451636,
      "grad_norm": 0.15564127266407013,
      "learning_rate": 8.90544975005367e-05,
      "loss": 0.8593,
      "mean_token_accuracy": 0.7496063709259033,
      "num_tokens": 113930541.0,
      "step": 3570
    },
    {
      "entropy": 0.9221852570772171,
      "epoch": 0.32855475241660487,
      "grad_norm": 0.1600762903690338,
      "learning_rate": 8.905143067439507e-05,
      "loss": 0.9086,
      "mean_token_accuracy": 0.739445049315691,
      "num_tokens": 113962882.0,
      "step": 3571
    },
    {
      "entropy": 0.7250676583498716,
      "epoch": 0.32864675878804606,
      "grad_norm": 0.15269091725349426,
      "learning_rate": 8.904836384825345e-05,
      "loss": 0.7084,
      "mean_token_accuracy": 0.7935532331466675,
      "num_tokens": 113994628.0,
      "step": 3572
    },
    {
      "entropy": 0.8389720786362886,
      "epoch": 0.3287387651594873,
      "grad_norm": 0.15709611773490906,
      "learning_rate": 8.904529702211182e-05,
      "loss": 0.832,
      "mean_token_accuracy": 0.7573861703276634,
      "num_tokens": 114025615.0,
      "step": 3573
    },
    {
      "entropy": 0.8733219243586063,
      "epoch": 0.3288307715309285,
      "grad_norm": 0.16201867163181305,
      "learning_rate": 8.90422301959702e-05,
      "loss": 0.8657,
      "mean_token_accuracy": 0.7476733699440956,
      "num_tokens": 114057883.0,
      "step": 3574
    },
    {
      "entropy": 0.853231031447649,
      "epoch": 0.32892277790236973,
      "grad_norm": 0.16007669270038605,
      "learning_rate": 8.903916336982857e-05,
      "loss": 0.8344,
      "mean_token_accuracy": 0.7592031136155128,
      "num_tokens": 114089632.0,
      "step": 3575
    },
    {
      "entropy": 0.8145501185208559,
      "epoch": 0.329014784273811,
      "grad_norm": 0.15804795920848846,
      "learning_rate": 8.903609654368694e-05,
      "loss": 0.8061,
      "mean_token_accuracy": 0.7648584842681885,
      "num_tokens": 114121610.0,
      "step": 3576
    },
    {
      "entropy": 0.6325392071157694,
      "epoch": 0.32910679064525217,
      "grad_norm": 0.1474388986825943,
      "learning_rate": 8.903302971754532e-05,
      "loss": 0.6398,
      "mean_token_accuracy": 0.8096894472837448,
      "num_tokens": 114154292.0,
      "step": 3577
    },
    {
      "entropy": 0.8600517455488443,
      "epoch": 0.3291987970166934,
      "grad_norm": 0.15634629130363464,
      "learning_rate": 8.90299628914037e-05,
      "loss": 0.8396,
      "mean_token_accuracy": 0.7595234327018261,
      "num_tokens": 114186940.0,
      "step": 3578
    },
    {
      "entropy": 0.9162767138332129,
      "epoch": 0.3292908033881346,
      "grad_norm": 0.16744458675384521,
      "learning_rate": 8.902689606526207e-05,
      "loss": 0.92,
      "mean_token_accuracy": 0.7395855747163296,
      "num_tokens": 114218608.0,
      "step": 3579
    },
    {
      "entropy": 0.8600432090461254,
      "epoch": 0.32938280975957585,
      "grad_norm": 0.1700349599123001,
      "learning_rate": 8.902382923912044e-05,
      "loss": 0.862,
      "mean_token_accuracy": 0.7489110343158245,
      "num_tokens": 114250214.0,
      "step": 3580
    },
    {
      "entropy": 0.8371495828032494,
      "epoch": 0.3294748161310171,
      "grad_norm": 0.15677180886268616,
      "learning_rate": 8.90207624129788e-05,
      "loss": 0.8251,
      "mean_token_accuracy": 0.7632488943636417,
      "num_tokens": 114281961.0,
      "step": 3581
    },
    {
      "entropy": 0.9180954955518246,
      "epoch": 0.3295668225024583,
      "grad_norm": 0.17211078107357025,
      "learning_rate": 8.901769558683719e-05,
      "loss": 0.9174,
      "mean_token_accuracy": 0.7445228174328804,
      "num_tokens": 114312672.0,
      "step": 3582
    },
    {
      "entropy": 0.8097448665648699,
      "epoch": 0.3296588288738995,
      "grad_norm": 0.15775388479232788,
      "learning_rate": 8.901462876069557e-05,
      "loss": 0.7915,
      "mean_token_accuracy": 0.7704235315322876,
      "num_tokens": 114344006.0,
      "step": 3583
    },
    {
      "entropy": 0.8507932834327221,
      "epoch": 0.3297508352453407,
      "grad_norm": 0.1576824188232422,
      "learning_rate": 8.901156193455393e-05,
      "loss": 0.833,
      "mean_token_accuracy": 0.757336188107729,
      "num_tokens": 114376149.0,
      "step": 3584
    },
    {
      "entropy": 0.779567513614893,
      "epoch": 0.32984284161678196,
      "grad_norm": 0.15537048876285553,
      "learning_rate": 8.90084951084123e-05,
      "loss": 0.7794,
      "mean_token_accuracy": 0.772665236145258,
      "num_tokens": 114408467.0,
      "step": 3585
    },
    {
      "entropy": 0.9026730991899967,
      "epoch": 0.3299348479882232,
      "grad_norm": 0.16472385823726654,
      "learning_rate": 8.900542828227068e-05,
      "loss": 0.8843,
      "mean_token_accuracy": 0.747175257652998,
      "num_tokens": 114440093.0,
      "step": 3586
    },
    {
      "entropy": 0.912804052233696,
      "epoch": 0.3300268543596644,
      "grad_norm": 0.16617894172668457,
      "learning_rate": 8.900236145612905e-05,
      "loss": 0.9073,
      "mean_token_accuracy": 0.7404039949178696,
      "num_tokens": 114471856.0,
      "step": 3587
    },
    {
      "entropy": 0.8459425456821918,
      "epoch": 0.33011886073110563,
      "grad_norm": 0.15891116857528687,
      "learning_rate": 8.899929462998743e-05,
      "loss": 0.8554,
      "mean_token_accuracy": 0.7533785291016102,
      "num_tokens": 114503881.0,
      "step": 3588
    },
    {
      "entropy": 0.7839451283216476,
      "epoch": 0.3302108671025469,
      "grad_norm": 0.1545349359512329,
      "learning_rate": 8.89962278038458e-05,
      "loss": 0.7726,
      "mean_token_accuracy": 0.7738772071897984,
      "num_tokens": 114535315.0,
      "step": 3589
    },
    {
      "entropy": 0.9241359271109104,
      "epoch": 0.33030287347398807,
      "grad_norm": 0.1594114601612091,
      "learning_rate": 8.899316097770418e-05,
      "loss": 0.909,
      "mean_token_accuracy": 0.7403079159557819,
      "num_tokens": 114567214.0,
      "step": 3590
    },
    {
      "entropy": 0.8929330017417669,
      "epoch": 0.3303948798454293,
      "grad_norm": 0.16146181523799896,
      "learning_rate": 8.899009415156255e-05,
      "loss": 0.8849,
      "mean_token_accuracy": 0.7452980019152164,
      "num_tokens": 114599982.0,
      "step": 3591
    },
    {
      "entropy": 0.8109452091157436,
      "epoch": 0.3304868862168705,
      "grad_norm": 0.15568256378173828,
      "learning_rate": 8.898702732542092e-05,
      "loss": 0.7889,
      "mean_token_accuracy": 0.7699626237154007,
      "num_tokens": 114631914.0,
      "step": 3592
    },
    {
      "entropy": 0.799917409196496,
      "epoch": 0.33057889258831175,
      "grad_norm": 0.16394449770450592,
      "learning_rate": 8.89839604992793e-05,
      "loss": 0.791,
      "mean_token_accuracy": 0.7750341296195984,
      "num_tokens": 114663073.0,
      "step": 3593
    },
    {
      "entropy": 0.7866846229881048,
      "epoch": 0.330670898959753,
      "grad_norm": 0.15336431562900543,
      "learning_rate": 8.898089367313768e-05,
      "loss": 0.77,
      "mean_token_accuracy": 0.7767357304692268,
      "num_tokens": 114695284.0,
      "step": 3594
    },
    {
      "entropy": 0.8895826525986195,
      "epoch": 0.3307629053311942,
      "grad_norm": 0.16475242376327515,
      "learning_rate": 8.897782684699605e-05,
      "loss": 0.8826,
      "mean_token_accuracy": 0.7471694611012936,
      "num_tokens": 114726789.0,
      "step": 3595
    },
    {
      "entropy": 0.9030837416648865,
      "epoch": 0.3308549117026354,
      "grad_norm": 0.162688747048378,
      "learning_rate": 8.897476002085442e-05,
      "loss": 0.8895,
      "mean_token_accuracy": 0.7419816479086876,
      "num_tokens": 114758342.0,
      "step": 3596
    },
    {
      "entropy": 0.7327923458069563,
      "epoch": 0.3309469180740766,
      "grad_norm": 0.16635191440582275,
      "learning_rate": 8.897169319471278e-05,
      "loss": 0.7657,
      "mean_token_accuracy": 0.778367705643177,
      "num_tokens": 114789754.0,
      "step": 3597
    },
    {
      "entropy": 0.8199177235364914,
      "epoch": 0.33103892444551786,
      "grad_norm": 0.15996363759040833,
      "learning_rate": 8.896862636857118e-05,
      "loss": 0.8194,
      "mean_token_accuracy": 0.7628922201693058,
      "num_tokens": 114821979.0,
      "step": 3598
    },
    {
      "entropy": 0.8251524642109871,
      "epoch": 0.3311309308169591,
      "grad_norm": 0.1591489613056183,
      "learning_rate": 8.896555954242955e-05,
      "loss": 0.8302,
      "mean_token_accuracy": 0.7589161545038223,
      "num_tokens": 114854610.0,
      "step": 3599
    },
    {
      "entropy": 0.9381790906190872,
      "epoch": 0.3312229371884003,
      "grad_norm": 0.1649271696805954,
      "learning_rate": 8.896249271628792e-05,
      "loss": 0.9532,
      "mean_token_accuracy": 0.7337306179106236,
      "num_tokens": 114886999.0,
      "step": 3600
    },
    {
      "entropy": 0.8716475367546082,
      "epoch": 0.33131494355984153,
      "grad_norm": 0.16641946136951447,
      "learning_rate": 8.89594258901463e-05,
      "loss": 0.8675,
      "mean_token_accuracy": 0.7484105825424194,
      "num_tokens": 114919310.0,
      "step": 3601
    },
    {
      "entropy": 0.7869132533669472,
      "epoch": 0.3314069499312827,
      "grad_norm": 0.16145667433738708,
      "learning_rate": 8.895635906400466e-05,
      "loss": 0.787,
      "mean_token_accuracy": 0.777467779815197,
      "num_tokens": 114951815.0,
      "step": 3602
    },
    {
      "entropy": 0.8463872373104095,
      "epoch": 0.33149895630272397,
      "grad_norm": 0.16601453721523285,
      "learning_rate": 8.895329223786305e-05,
      "loss": 0.856,
      "mean_token_accuracy": 0.753614965826273,
      "num_tokens": 114983655.0,
      "step": 3603
    },
    {
      "entropy": 0.8107794895768166,
      "epoch": 0.3315909626741652,
      "grad_norm": 0.15225301682949066,
      "learning_rate": 8.895022541172141e-05,
      "loss": 0.8167,
      "mean_token_accuracy": 0.7627292983233929,
      "num_tokens": 115015932.0,
      "step": 3604
    },
    {
      "entropy": 0.7167086713016033,
      "epoch": 0.3316829690456064,
      "grad_norm": 0.15528762340545654,
      "learning_rate": 8.89471585855798e-05,
      "loss": 0.7158,
      "mean_token_accuracy": 0.7908207811415195,
      "num_tokens": 115048461.0,
      "step": 3605
    },
    {
      "entropy": 0.804124291986227,
      "epoch": 0.33177497541704765,
      "grad_norm": 0.15502788126468658,
      "learning_rate": 8.894409175943816e-05,
      "loss": 0.7901,
      "mean_token_accuracy": 0.7709712535142899,
      "num_tokens": 115080760.0,
      "step": 3606
    },
    {
      "entropy": 0.8811458274722099,
      "epoch": 0.33186698178848884,
      "grad_norm": 0.16447661817073822,
      "learning_rate": 8.894102493329653e-05,
      "loss": 0.8724,
      "mean_token_accuracy": 0.7513988986611366,
      "num_tokens": 115111669.0,
      "step": 3607
    },
    {
      "entropy": 0.8510777726769447,
      "epoch": 0.3319589881599301,
      "grad_norm": 0.16196595132350922,
      "learning_rate": 8.893795810715491e-05,
      "loss": 0.829,
      "mean_token_accuracy": 0.7608736306428909,
      "num_tokens": 115143361.0,
      "step": 3608
    },
    {
      "entropy": 0.8625008836388588,
      "epoch": 0.3320509945313713,
      "grad_norm": 0.16040565073490143,
      "learning_rate": 8.89348912810133e-05,
      "loss": 0.8678,
      "mean_token_accuracy": 0.749535083770752,
      "num_tokens": 115175107.0,
      "step": 3609
    },
    {
      "entropy": 0.8160389997065067,
      "epoch": 0.3321430009028125,
      "grad_norm": 0.16039665043354034,
      "learning_rate": 8.893182445487166e-05,
      "loss": 0.8095,
      "mean_token_accuracy": 0.7668877430260181,
      "num_tokens": 115206975.0,
      "step": 3610
    },
    {
      "entropy": 0.8754699677228928,
      "epoch": 0.33223500727425376,
      "grad_norm": 0.15433968603610992,
      "learning_rate": 8.892875762873003e-05,
      "loss": 0.8562,
      "mean_token_accuracy": 0.7540491670370102,
      "num_tokens": 115238860.0,
      "step": 3611
    },
    {
      "entropy": 0.8209942188113928,
      "epoch": 0.33232701364569495,
      "grad_norm": 0.1554667204618454,
      "learning_rate": 8.89256908025884e-05,
      "loss": 0.8083,
      "mean_token_accuracy": 0.7691846899688244,
      "num_tokens": 115271210.0,
      "step": 3612
    },
    {
      "entropy": 0.735757265239954,
      "epoch": 0.3324190200171362,
      "grad_norm": 0.1488027423620224,
      "learning_rate": 8.892262397644678e-05,
      "loss": 0.7157,
      "mean_token_accuracy": 0.7894121445715427,
      "num_tokens": 115303202.0,
      "step": 3613
    },
    {
      "entropy": 0.7885707840323448,
      "epoch": 0.33251102638857744,
      "grad_norm": 0.16013674437999725,
      "learning_rate": 8.891955715030516e-05,
      "loss": 0.7695,
      "mean_token_accuracy": 0.7755643874406815,
      "num_tokens": 115334566.0,
      "step": 3614
    },
    {
      "entropy": 0.9095710888504982,
      "epoch": 0.3326030327600186,
      "grad_norm": 0.16038651764392853,
      "learning_rate": 8.891649032416353e-05,
      "loss": 0.9148,
      "mean_token_accuracy": 0.7393782325088978,
      "num_tokens": 115367020.0,
      "step": 3615
    },
    {
      "entropy": 0.7996747493743896,
      "epoch": 0.33269503913145987,
      "grad_norm": 0.15485355257987976,
      "learning_rate": 8.89134234980219e-05,
      "loss": 0.7846,
      "mean_token_accuracy": 0.770352341234684,
      "num_tokens": 115399174.0,
      "step": 3616
    },
    {
      "entropy": 0.815029202029109,
      "epoch": 0.33278704550290106,
      "grad_norm": 0.1572398692369461,
      "learning_rate": 8.891035667188028e-05,
      "loss": 0.8159,
      "mean_token_accuracy": 0.7632438465952873,
      "num_tokens": 115431504.0,
      "step": 3617
    },
    {
      "entropy": 0.8200552687048912,
      "epoch": 0.3328790518743423,
      "grad_norm": 0.16495904326438904,
      "learning_rate": 8.890728984573865e-05,
      "loss": 0.8218,
      "mean_token_accuracy": 0.7635450959205627,
      "num_tokens": 115463463.0,
      "step": 3618
    },
    {
      "entropy": 0.7632487062364817,
      "epoch": 0.33297105824578355,
      "grad_norm": 0.16388553380966187,
      "learning_rate": 8.890422301959703e-05,
      "loss": 0.7588,
      "mean_token_accuracy": 0.7748204618692398,
      "num_tokens": 115495498.0,
      "step": 3619
    },
    {
      "entropy": 0.789536215364933,
      "epoch": 0.33306306461722474,
      "grad_norm": 0.1638256162405014,
      "learning_rate": 8.89011561934554e-05,
      "loss": 0.7933,
      "mean_token_accuracy": 0.7691398598253727,
      "num_tokens": 115527710.0,
      "step": 3620
    },
    {
      "entropy": 0.8147595971822739,
      "epoch": 0.333155070988666,
      "grad_norm": 0.1569482386112213,
      "learning_rate": 8.889808936731378e-05,
      "loss": 0.808,
      "mean_token_accuracy": 0.7663806416094303,
      "num_tokens": 115558447.0,
      "step": 3621
    },
    {
      "entropy": 0.8059687651693821,
      "epoch": 0.33324707736010717,
      "grad_norm": 0.16552405059337616,
      "learning_rate": 8.889502254117214e-05,
      "loss": 0.814,
      "mean_token_accuracy": 0.7676874585449696,
      "num_tokens": 115589246.0,
      "step": 3622
    },
    {
      "entropy": 0.686234500259161,
      "epoch": 0.3333390837315484,
      "grad_norm": 0.1551610380411148,
      "learning_rate": 8.889195571503051e-05,
      "loss": 0.6699,
      "mean_token_accuracy": 0.8033677414059639,
      "num_tokens": 115620763.0,
      "step": 3623
    },
    {
      "entropy": 0.8165770806372166,
      "epoch": 0.33343109010298966,
      "grad_norm": 0.15596449375152588,
      "learning_rate": 8.888888888888889e-05,
      "loss": 0.7934,
      "mean_token_accuracy": 0.7694069258868694,
      "num_tokens": 115652576.0,
      "step": 3624
    },
    {
      "entropy": 0.713382177054882,
      "epoch": 0.33352309647443085,
      "grad_norm": 0.15949313342571259,
      "learning_rate": 8.888582206274727e-05,
      "loss": 0.7048,
      "mean_token_accuracy": 0.7907537557184696,
      "num_tokens": 115684727.0,
      "step": 3625
    },
    {
      "entropy": 0.8603894151747227,
      "epoch": 0.3336151028458721,
      "grad_norm": 0.16378435492515564,
      "learning_rate": 8.888275523660564e-05,
      "loss": 0.869,
      "mean_token_accuracy": 0.7471421547234058,
      "num_tokens": 115716403.0,
      "step": 3626
    },
    {
      "entropy": 0.7439581509679556,
      "epoch": 0.3337071092173133,
      "grad_norm": 0.15038296580314636,
      "learning_rate": 8.887968841046401e-05,
      "loss": 0.7325,
      "mean_token_accuracy": 0.7888507023453712,
      "num_tokens": 115748089.0,
      "step": 3627
    },
    {
      "entropy": 0.8052119482308626,
      "epoch": 0.3337991155887545,
      "grad_norm": 0.15734940767288208,
      "learning_rate": 8.887662158432238e-05,
      "loss": 0.7752,
      "mean_token_accuracy": 0.7757416628301144,
      "num_tokens": 115780012.0,
      "step": 3628
    },
    {
      "entropy": 0.7635125108063221,
      "epoch": 0.33389112196019577,
      "grad_norm": 0.1494060903787613,
      "learning_rate": 8.887355475818076e-05,
      "loss": 0.7512,
      "mean_token_accuracy": 0.7756372913718224,
      "num_tokens": 115811683.0,
      "step": 3629
    },
    {
      "entropy": 0.8296754285693169,
      "epoch": 0.33398312833163696,
      "grad_norm": 0.16079352796077728,
      "learning_rate": 8.887048793203914e-05,
      "loss": 0.8154,
      "mean_token_accuracy": 0.7604441605508327,
      "num_tokens": 115843460.0,
      "step": 3630
    },
    {
      "entropy": 0.883204584941268,
      "epoch": 0.3340751347030782,
      "grad_norm": 0.16717183589935303,
      "learning_rate": 8.886742110589751e-05,
      "loss": 0.8728,
      "mean_token_accuracy": 0.7563805729150772,
      "num_tokens": 115874330.0,
      "step": 3631
    },
    {
      "entropy": 0.7770205978304148,
      "epoch": 0.3341671410745194,
      "grad_norm": 0.15396027266979218,
      "learning_rate": 8.886435427975589e-05,
      "loss": 0.7675,
      "mean_token_accuracy": 0.7754591144621372,
      "num_tokens": 115906749.0,
      "step": 3632
    },
    {
      "entropy": 0.7597681414335966,
      "epoch": 0.33425914744596064,
      "grad_norm": 0.15559104084968567,
      "learning_rate": 8.886128745361426e-05,
      "loss": 0.7688,
      "mean_token_accuracy": 0.77598812058568,
      "num_tokens": 115939246.0,
      "step": 3633
    },
    {
      "entropy": 0.7265193667262793,
      "epoch": 0.3343511538174019,
      "grad_norm": 0.15766224265098572,
      "learning_rate": 8.885822062747263e-05,
      "loss": 0.7397,
      "mean_token_accuracy": 0.7814281471073627,
      "num_tokens": 115971375.0,
      "step": 3634
    },
    {
      "entropy": 0.823904063552618,
      "epoch": 0.33444316018884307,
      "grad_norm": 0.1595362424850464,
      "learning_rate": 8.885515380133101e-05,
      "loss": 0.8366,
      "mean_token_accuracy": 0.7565645761787891,
      "num_tokens": 116002635.0,
      "step": 3635
    },
    {
      "entropy": 0.8116504587233067,
      "epoch": 0.3345351665602843,
      "grad_norm": 0.16941915452480316,
      "learning_rate": 8.885208697518939e-05,
      "loss": 0.8411,
      "mean_token_accuracy": 0.7581542544066906,
      "num_tokens": 116034535.0,
      "step": 3636
    },
    {
      "entropy": 0.8121430985629559,
      "epoch": 0.3346271729317255,
      "grad_norm": 0.15128302574157715,
      "learning_rate": 8.884902014904776e-05,
      "loss": 0.8164,
      "mean_token_accuracy": 0.7611706219613552,
      "num_tokens": 116067067.0,
      "step": 3637
    },
    {
      "entropy": 0.7618887759745121,
      "epoch": 0.33471917930316675,
      "grad_norm": 0.15282335877418518,
      "learning_rate": 8.884595332290612e-05,
      "loss": 0.7229,
      "mean_token_accuracy": 0.7903376892209053,
      "num_tokens": 116099234.0,
      "step": 3638
    },
    {
      "entropy": 0.8003894705325365,
      "epoch": 0.334811185674608,
      "grad_norm": 0.1570078283548355,
      "learning_rate": 8.884288649676449e-05,
      "loss": 0.7872,
      "mean_token_accuracy": 0.7701768837869167,
      "num_tokens": 116130838.0,
      "step": 3639
    },
    {
      "entropy": 0.7909752726554871,
      "epoch": 0.3349031920460492,
      "grad_norm": 0.16156211495399475,
      "learning_rate": 8.883981967062289e-05,
      "loss": 0.7824,
      "mean_token_accuracy": 0.7713269516825676,
      "num_tokens": 116161781.0,
      "step": 3640
    },
    {
      "entropy": 0.8605387713760138,
      "epoch": 0.3349951984174904,
      "grad_norm": 0.1603011190891266,
      "learning_rate": 8.883675284448126e-05,
      "loss": 0.8411,
      "mean_token_accuracy": 0.757586769759655,
      "num_tokens": 116194086.0,
      "step": 3641
    },
    {
      "entropy": 0.8223765790462494,
      "epoch": 0.3350872047889316,
      "grad_norm": 0.15769994258880615,
      "learning_rate": 8.883368601833962e-05,
      "loss": 0.803,
      "mean_token_accuracy": 0.7642200514674187,
      "num_tokens": 116225832.0,
      "step": 3642
    },
    {
      "entropy": 0.8542160503566265,
      "epoch": 0.33517921116037286,
      "grad_norm": 0.1579298973083496,
      "learning_rate": 8.883061919219799e-05,
      "loss": 0.842,
      "mean_token_accuracy": 0.7585493512451649,
      "num_tokens": 116258343.0,
      "step": 3643
    },
    {
      "entropy": 0.7841754704713821,
      "epoch": 0.3352712175318141,
      "grad_norm": 0.15193511545658112,
      "learning_rate": 8.882755236605637e-05,
      "loss": 0.7646,
      "mean_token_accuracy": 0.7717516906559467,
      "num_tokens": 116289682.0,
      "step": 3644
    },
    {
      "entropy": 0.8506742939352989,
      "epoch": 0.3353632239032553,
      "grad_norm": 0.16299085319042206,
      "learning_rate": 8.882448553991475e-05,
      "loss": 0.8394,
      "mean_token_accuracy": 0.7581215500831604,
      "num_tokens": 116321679.0,
      "step": 3645
    },
    {
      "entropy": 0.8661758825182915,
      "epoch": 0.33545523027469654,
      "grad_norm": 0.1545385718345642,
      "learning_rate": 8.882141871377312e-05,
      "loss": 0.8659,
      "mean_token_accuracy": 0.7464869432151318,
      "num_tokens": 116353607.0,
      "step": 3646
    },
    {
      "entropy": 0.7757053375244141,
      "epoch": 0.3355472366461377,
      "grad_norm": 0.16197070479393005,
      "learning_rate": 8.881835188763149e-05,
      "loss": 0.7895,
      "mean_token_accuracy": 0.7744596637785435,
      "num_tokens": 116385519.0,
      "step": 3647
    },
    {
      "entropy": 0.8351331688463688,
      "epoch": 0.33563924301757897,
      "grad_norm": 0.17365828156471252,
      "learning_rate": 8.881528506148987e-05,
      "loss": 0.858,
      "mean_token_accuracy": 0.7545035630464554,
      "num_tokens": 116416750.0,
      "step": 3648
    },
    {
      "entropy": 0.8064558673650026,
      "epoch": 0.3357312493890202,
      "grad_norm": 0.15767648816108704,
      "learning_rate": 8.881221823534824e-05,
      "loss": 0.7854,
      "mean_token_accuracy": 0.7716449312865734,
      "num_tokens": 116448748.0,
      "step": 3649
    },
    {
      "entropy": 0.7940638735890388,
      "epoch": 0.3358232557604614,
      "grad_norm": 0.16072049736976624,
      "learning_rate": 8.880915140920662e-05,
      "loss": 0.7939,
      "mean_token_accuracy": 0.7647316791117191,
      "num_tokens": 116480477.0,
      "step": 3650
    },
    {
      "entropy": 0.8334903717041016,
      "epoch": 0.33591526213190265,
      "grad_norm": 0.17076003551483154,
      "learning_rate": 8.880608458306499e-05,
      "loss": 0.8627,
      "mean_token_accuracy": 0.7560299262404442,
      "num_tokens": 116512990.0,
      "step": 3651
    },
    {
      "entropy": 0.7847942523658276,
      "epoch": 0.33600726850334384,
      "grad_norm": 0.15663374960422516,
      "learning_rate": 8.880301775692337e-05,
      "loss": 0.754,
      "mean_token_accuracy": 0.7803404740989208,
      "num_tokens": 116544220.0,
      "step": 3652
    },
    {
      "entropy": 0.830343084409833,
      "epoch": 0.3360992748747851,
      "grad_norm": 0.16040533781051636,
      "learning_rate": 8.879995093078174e-05,
      "loss": 0.8135,
      "mean_token_accuracy": 0.767262015491724,
      "num_tokens": 116575788.0,
      "step": 3653
    },
    {
      "entropy": 0.8874675445258617,
      "epoch": 0.3361912812462263,
      "grad_norm": 0.16346627473831177,
      "learning_rate": 8.87968841046401e-05,
      "loss": 0.8807,
      "mean_token_accuracy": 0.7464034929871559,
      "num_tokens": 116607087.0,
      "step": 3654
    },
    {
      "entropy": 0.7744895126670599,
      "epoch": 0.3362832876176675,
      "grad_norm": 0.16275112330913544,
      "learning_rate": 8.879381727849849e-05,
      "loss": 0.7657,
      "mean_token_accuracy": 0.7825987488031387,
      "num_tokens": 116639502.0,
      "step": 3655
    },
    {
      "entropy": 0.8713775221258402,
      "epoch": 0.33637529398910876,
      "grad_norm": 0.1708855926990509,
      "learning_rate": 8.879075045235687e-05,
      "loss": 0.8679,
      "mean_token_accuracy": 0.7495811469852924,
      "num_tokens": 116670956.0,
      "step": 3656
    },
    {
      "entropy": 0.8958065435290337,
      "epoch": 0.33646730036054995,
      "grad_norm": 0.15959987044334412,
      "learning_rate": 8.878768362621524e-05,
      "loss": 0.8752,
      "mean_token_accuracy": 0.751251831650734,
      "num_tokens": 116703724.0,
      "step": 3657
    },
    {
      "entropy": 0.8154958300292492,
      "epoch": 0.3365593067319912,
      "grad_norm": 0.1482742577791214,
      "learning_rate": 8.87846168000736e-05,
      "loss": 0.7817,
      "mean_token_accuracy": 0.7681431993842125,
      "num_tokens": 116736011.0,
      "step": 3658
    },
    {
      "entropy": 0.8066625036299229,
      "epoch": 0.33665131310343244,
      "grad_norm": 0.15257874131202698,
      "learning_rate": 8.878154997393197e-05,
      "loss": 0.7887,
      "mean_token_accuracy": 0.7676409259438515,
      "num_tokens": 116767483.0,
      "step": 3659
    },
    {
      "entropy": 0.88212014362216,
      "epoch": 0.3367433194748736,
      "grad_norm": 0.15867763757705688,
      "learning_rate": 8.877848314779035e-05,
      "loss": 0.867,
      "mean_token_accuracy": 0.7490551434457302,
      "num_tokens": 116799356.0,
      "step": 3660
    },
    {
      "entropy": 0.7136856466531754,
      "epoch": 0.33683532584631487,
      "grad_norm": 0.1536937803030014,
      "learning_rate": 8.877541632164873e-05,
      "loss": 0.729,
      "mean_token_accuracy": 0.7822437733411789,
      "num_tokens": 116831003.0,
      "step": 3661
    },
    {
      "entropy": 0.7921467293053865,
      "epoch": 0.33692733221775606,
      "grad_norm": 0.15576539933681488,
      "learning_rate": 8.87723494955071e-05,
      "loss": 0.8016,
      "mean_token_accuracy": 0.7680702172219753,
      "num_tokens": 116863005.0,
      "step": 3662
    },
    {
      "entropy": 0.8578747417777777,
      "epoch": 0.3370193385891973,
      "grad_norm": 0.15569129586219788,
      "learning_rate": 8.876928266936548e-05,
      "loss": 0.8606,
      "mean_token_accuracy": 0.7532249167561531,
      "num_tokens": 116894722.0,
      "step": 3663
    },
    {
      "entropy": 0.7921356242150068,
      "epoch": 0.33711134496063855,
      "grad_norm": 0.16004441678524017,
      "learning_rate": 8.876621584322385e-05,
      "loss": 0.7929,
      "mean_token_accuracy": 0.7708720229566097,
      "num_tokens": 116926620.0,
      "step": 3664
    },
    {
      "entropy": 0.7752866856753826,
      "epoch": 0.33720335133207974,
      "grad_norm": 0.15228639543056488,
      "learning_rate": 8.876314901708222e-05,
      "loss": 0.7752,
      "mean_token_accuracy": 0.7743471227586269,
      "num_tokens": 116959232.0,
      "step": 3665
    },
    {
      "entropy": 0.8905178159475327,
      "epoch": 0.337295357703521,
      "grad_norm": 0.16030436754226685,
      "learning_rate": 8.87600821909406e-05,
      "loss": 0.881,
      "mean_token_accuracy": 0.745611596852541,
      "num_tokens": 116991215.0,
      "step": 3666
    },
    {
      "entropy": 0.8775111436843872,
      "epoch": 0.33738736407496217,
      "grad_norm": 0.16572625935077667,
      "learning_rate": 8.875701536479898e-05,
      "loss": 0.8734,
      "mean_token_accuracy": 0.7511332593858242,
      "num_tokens": 117023042.0,
      "step": 3667
    },
    {
      "entropy": 0.8814806453883648,
      "epoch": 0.3374793704464034,
      "grad_norm": 0.17149776220321655,
      "learning_rate": 8.875394853865735e-05,
      "loss": 0.9061,
      "mean_token_accuracy": 0.7410387955605984,
      "num_tokens": 117055219.0,
      "step": 3668
    },
    {
      "entropy": 0.8330910131335258,
      "epoch": 0.33757137681784466,
      "grad_norm": 0.1566135585308075,
      "learning_rate": 8.875088171251572e-05,
      "loss": 0.8066,
      "mean_token_accuracy": 0.7661771029233932,
      "num_tokens": 117087263.0,
      "step": 3669
    },
    {
      "entropy": 0.7134779337793589,
      "epoch": 0.33766338318928585,
      "grad_norm": 0.14967162907123566,
      "learning_rate": 8.874781488637409e-05,
      "loss": 0.6704,
      "mean_token_accuracy": 0.805096797645092,
      "num_tokens": 117119144.0,
      "step": 3670
    },
    {
      "entropy": 0.909952636808157,
      "epoch": 0.3377553895607271,
      "grad_norm": 0.16171784698963165,
      "learning_rate": 8.874474806023247e-05,
      "loss": 0.8929,
      "mean_token_accuracy": 0.742050014436245,
      "num_tokens": 117151536.0,
      "step": 3671
    },
    {
      "entropy": 0.826678778976202,
      "epoch": 0.3378473959321683,
      "grad_norm": 0.15655003488063812,
      "learning_rate": 8.874168123409085e-05,
      "loss": 0.8147,
      "mean_token_accuracy": 0.7604783289134502,
      "num_tokens": 117182264.0,
      "step": 3672
    },
    {
      "entropy": 0.7862184662371874,
      "epoch": 0.3379394023036095,
      "grad_norm": 0.1561945229768753,
      "learning_rate": 8.873861440794922e-05,
      "loss": 0.762,
      "mean_token_accuracy": 0.7792872153222561,
      "num_tokens": 117214331.0,
      "step": 3673
    },
    {
      "entropy": 0.874307245016098,
      "epoch": 0.33803140867505077,
      "grad_norm": 0.16299164295196533,
      "learning_rate": 8.873554758180758e-05,
      "loss": 0.8725,
      "mean_token_accuracy": 0.7515481300652027,
      "num_tokens": 117245751.0,
      "step": 3674
    },
    {
      "entropy": 0.700889153406024,
      "epoch": 0.33812341504649196,
      "grad_norm": 0.1487797498703003,
      "learning_rate": 8.873248075566597e-05,
      "loss": 0.6862,
      "mean_token_accuracy": 0.7942554093897343,
      "num_tokens": 117278158.0,
      "step": 3675
    },
    {
      "entropy": 0.8977634534239769,
      "epoch": 0.3382154214179332,
      "grad_norm": 0.16320493817329407,
      "learning_rate": 8.872941392952435e-05,
      "loss": 0.8939,
      "mean_token_accuracy": 0.7468542940914631,
      "num_tokens": 117309983.0,
      "step": 3676
    },
    {
      "entropy": 0.7923063766211271,
      "epoch": 0.3383074277893744,
      "grad_norm": 0.1535583734512329,
      "learning_rate": 8.872634710338272e-05,
      "loss": 0.7755,
      "mean_token_accuracy": 0.7719611264765263,
      "num_tokens": 117342055.0,
      "step": 3677
    },
    {
      "entropy": 0.8221751414239407,
      "epoch": 0.33839943416081564,
      "grad_norm": 0.161599799990654,
      "learning_rate": 8.872328027724108e-05,
      "loss": 0.8341,
      "mean_token_accuracy": 0.7606333866715431,
      "num_tokens": 117374229.0,
      "step": 3678
    },
    {
      "entropy": 0.7745476104319096,
      "epoch": 0.3384914405322569,
      "grad_norm": 0.15252746641635895,
      "learning_rate": 8.872021345109946e-05,
      "loss": 0.7667,
      "mean_token_accuracy": 0.7754529230296612,
      "num_tokens": 117406610.0,
      "step": 3679
    },
    {
      "entropy": 0.8654115535318851,
      "epoch": 0.33858344690369807,
      "grad_norm": 0.15777841210365295,
      "learning_rate": 8.871714662495783e-05,
      "loss": 0.8681,
      "mean_token_accuracy": 0.7465464733541012,
      "num_tokens": 117438836.0,
      "step": 3680
    },
    {
      "entropy": 0.8312840089201927,
      "epoch": 0.3386754532751393,
      "grad_norm": 0.1638922542333603,
      "learning_rate": 8.871407979881621e-05,
      "loss": 0.8452,
      "mean_token_accuracy": 0.7581473626196384,
      "num_tokens": 117470720.0,
      "step": 3681
    },
    {
      "entropy": 0.6733292415738106,
      "epoch": 0.3387674596465805,
      "grad_norm": 0.15174655616283417,
      "learning_rate": 8.871101297267458e-05,
      "loss": 0.6784,
      "mean_token_accuracy": 0.8034169785678387,
      "num_tokens": 117502760.0,
      "step": 3682
    },
    {
      "entropy": 0.8665666282176971,
      "epoch": 0.33885946601802175,
      "grad_norm": 0.15787923336029053,
      "learning_rate": 8.870794614653296e-05,
      "loss": 0.8562,
      "mean_token_accuracy": 0.7535654231905937,
      "num_tokens": 117534604.0,
      "step": 3683
    },
    {
      "entropy": 0.9273284897208214,
      "epoch": 0.338951472389463,
      "grad_norm": 0.1670336127281189,
      "learning_rate": 8.870487932039133e-05,
      "loss": 0.9107,
      "mean_token_accuracy": 0.74199004098773,
      "num_tokens": 117566828.0,
      "step": 3684
    },
    {
      "entropy": 0.7546845283359289,
      "epoch": 0.3390434787609042,
      "grad_norm": 0.15495595335960388,
      "learning_rate": 8.87018124942497e-05,
      "loss": 0.7548,
      "mean_token_accuracy": 0.7755203880369663,
      "num_tokens": 117599010.0,
      "step": 3685
    },
    {
      "entropy": 0.7731052860617638,
      "epoch": 0.3391354851323454,
      "grad_norm": 0.1596841663122177,
      "learning_rate": 8.869874566810808e-05,
      "loss": 0.7868,
      "mean_token_accuracy": 0.7701879143714905,
      "num_tokens": 117631321.0,
      "step": 3686
    },
    {
      "entropy": 0.7496383264660835,
      "epoch": 0.3392274915037866,
      "grad_norm": 0.16150377690792084,
      "learning_rate": 8.869567884196646e-05,
      "loss": 0.7247,
      "mean_token_accuracy": 0.7845935821533203,
      "num_tokens": 117662202.0,
      "step": 3687
    },
    {
      "entropy": 0.7468512617051601,
      "epoch": 0.33931949787522786,
      "grad_norm": 0.16161344945430756,
      "learning_rate": 8.869261201582483e-05,
      "loss": 0.7361,
      "mean_token_accuracy": 0.7821064479649067,
      "num_tokens": 117693389.0,
      "step": 3688
    },
    {
      "entropy": 0.8006163500249386,
      "epoch": 0.3394115042466691,
      "grad_norm": 0.15974213182926178,
      "learning_rate": 8.86895451896832e-05,
      "loss": 0.7837,
      "mean_token_accuracy": 0.7740936055779457,
      "num_tokens": 117725702.0,
      "step": 3689
    },
    {
      "entropy": 0.7021926436573267,
      "epoch": 0.3395035106181103,
      "grad_norm": 0.1482010781764984,
      "learning_rate": 8.868647836354157e-05,
      "loss": 0.6849,
      "mean_token_accuracy": 0.7978414669632912,
      "num_tokens": 117758096.0,
      "step": 3690
    },
    {
      "entropy": 0.8278223015367985,
      "epoch": 0.33959551698955154,
      "grad_norm": 0.14988671243190765,
      "learning_rate": 8.868341153739995e-05,
      "loss": 0.8171,
      "mean_token_accuracy": 0.7613799311220646,
      "num_tokens": 117790293.0,
      "step": 3691
    },
    {
      "entropy": 0.9387873783707619,
      "epoch": 0.3396875233609927,
      "grad_norm": 0.16463173925876617,
      "learning_rate": 8.868034471125833e-05,
      "loss": 0.9263,
      "mean_token_accuracy": 0.7348545640707016,
      "num_tokens": 117822322.0,
      "step": 3692
    },
    {
      "entropy": 0.8639038391411304,
      "epoch": 0.33977952973243397,
      "grad_norm": 0.15945254266262054,
      "learning_rate": 8.86772778851167e-05,
      "loss": 0.874,
      "mean_token_accuracy": 0.7496752366423607,
      "num_tokens": 117854788.0,
      "step": 3693
    },
    {
      "entropy": 0.7606076449155807,
      "epoch": 0.3398715361038752,
      "grad_norm": 0.15076787769794464,
      "learning_rate": 8.867421105897508e-05,
      "loss": 0.7386,
      "mean_token_accuracy": 0.7830086834728718,
      "num_tokens": 117887322.0,
      "step": 3694
    },
    {
      "entropy": 0.8467345125973225,
      "epoch": 0.3399635424753164,
      "grad_norm": 0.1598215401172638,
      "learning_rate": 8.867114423283345e-05,
      "loss": 0.8512,
      "mean_token_accuracy": 0.7536008879542351,
      "num_tokens": 117919272.0,
      "step": 3695
    },
    {
      "entropy": 0.9232820477336645,
      "epoch": 0.34005554884675765,
      "grad_norm": 0.16981424391269684,
      "learning_rate": 8.866807740669181e-05,
      "loss": 0.9371,
      "mean_token_accuracy": 0.7337960563600063,
      "num_tokens": 117950752.0,
      "step": 3696
    },
    {
      "entropy": 0.905631847679615,
      "epoch": 0.34014755521819884,
      "grad_norm": 0.15749426186084747,
      "learning_rate": 8.86650105805502e-05,
      "loss": 0.8901,
      "mean_token_accuracy": 0.7460664175450802,
      "num_tokens": 117983087.0,
      "step": 3697
    },
    {
      "entropy": 0.8553846329450607,
      "epoch": 0.3402395615896401,
      "grad_norm": 0.16589222848415375,
      "learning_rate": 8.866194375440858e-05,
      "loss": 0.8372,
      "mean_token_accuracy": 0.7556845061480999,
      "num_tokens": 118014979.0,
      "step": 3698
    },
    {
      "entropy": 0.7258607223629951,
      "epoch": 0.3403315679610813,
      "grad_norm": 0.14893867075443268,
      "learning_rate": 8.865887692826694e-05,
      "loss": 0.7104,
      "mean_token_accuracy": 0.7932132892310619,
      "num_tokens": 118047210.0,
      "step": 3699
    },
    {
      "entropy": 0.7842186614871025,
      "epoch": 0.3404235743325225,
      "grad_norm": 0.159607395529747,
      "learning_rate": 8.865581010212531e-05,
      "loss": 0.7777,
      "mean_token_accuracy": 0.7716917842626572,
      "num_tokens": 118079563.0,
      "step": 3700
    },
    {
      "entropy": 0.9354798421263695,
      "epoch": 0.34051558070396376,
      "grad_norm": 0.16963587701320648,
      "learning_rate": 8.865274327598368e-05,
      "loss": 0.9392,
      "mean_token_accuracy": 0.730739526450634,
      "num_tokens": 118109799.0,
      "step": 3701
    },
    {
      "entropy": 0.9464934282004833,
      "epoch": 0.34060758707540495,
      "grad_norm": 0.16352087259292603,
      "learning_rate": 8.864967644984206e-05,
      "loss": 0.9503,
      "mean_token_accuracy": 0.7305676527321339,
      "num_tokens": 118141476.0,
      "step": 3702
    },
    {
      "entropy": 0.7594627905637026,
      "epoch": 0.3406995934468462,
      "grad_norm": 0.15201662480831146,
      "learning_rate": 8.864660962370044e-05,
      "loss": 0.7423,
      "mean_token_accuracy": 0.7831352204084396,
      "num_tokens": 118173791.0,
      "step": 3703
    },
    {
      "entropy": 0.7795003280043602,
      "epoch": 0.34079159981828744,
      "grad_norm": 0.15459637343883514,
      "learning_rate": 8.864354279755881e-05,
      "loss": 0.7769,
      "mean_token_accuracy": 0.7757538892328739,
      "num_tokens": 118206355.0,
      "step": 3704
    },
    {
      "entropy": 0.8717618584632874,
      "epoch": 0.3408836061897286,
      "grad_norm": 0.16742613911628723,
      "learning_rate": 8.864047597141718e-05,
      "loss": 0.8702,
      "mean_token_accuracy": 0.7534622736275196,
      "num_tokens": 118238166.0,
      "step": 3705
    },
    {
      "entropy": 0.8845714628696442,
      "epoch": 0.34097561256116987,
      "grad_norm": 0.16011746227741241,
      "learning_rate": 8.863740914527556e-05,
      "loss": 0.8978,
      "mean_token_accuracy": 0.7447999753057957,
      "num_tokens": 118270769.0,
      "step": 3706
    },
    {
      "entropy": 0.9194144401699305,
      "epoch": 0.34106761893261106,
      "grad_norm": 0.1600901037454605,
      "learning_rate": 8.863434231913393e-05,
      "loss": 0.9215,
      "mean_token_accuracy": 0.7357224337756634,
      "num_tokens": 118302904.0,
      "step": 3707
    },
    {
      "entropy": 0.8001449629664421,
      "epoch": 0.3411596253040523,
      "grad_norm": 0.16112641990184784,
      "learning_rate": 8.863127549299231e-05,
      "loss": 0.7964,
      "mean_token_accuracy": 0.767424400895834,
      "num_tokens": 118335218.0,
      "step": 3708
    },
    {
      "entropy": 0.841883921995759,
      "epoch": 0.34125163167549355,
      "grad_norm": 0.1622907519340515,
      "learning_rate": 8.862820866685068e-05,
      "loss": 0.8175,
      "mean_token_accuracy": 0.7635161690413952,
      "num_tokens": 118366737.0,
      "step": 3709
    },
    {
      "entropy": 0.876302357763052,
      "epoch": 0.34134363804693474,
      "grad_norm": 0.15373189747333527,
      "learning_rate": 8.862514184070906e-05,
      "loss": 0.8401,
      "mean_token_accuracy": 0.7561798356473446,
      "num_tokens": 118398430.0,
      "step": 3710
    },
    {
      "entropy": 0.8893446121364832,
      "epoch": 0.341435644418376,
      "grad_norm": 0.16385892033576965,
      "learning_rate": 8.862207501456743e-05,
      "loss": 0.8668,
      "mean_token_accuracy": 0.752390667796135,
      "num_tokens": 118430478.0,
      "step": 3711
    },
    {
      "entropy": 0.8816887829452753,
      "epoch": 0.34152765078981717,
      "grad_norm": 0.15924251079559326,
      "learning_rate": 8.86190081884258e-05,
      "loss": 0.8794,
      "mean_token_accuracy": 0.7472030259668827,
      "num_tokens": 118463172.0,
      "step": 3712
    },
    {
      "entropy": 0.8651000875979662,
      "epoch": 0.3416196571612584,
      "grad_norm": 0.16222727298736572,
      "learning_rate": 8.861594136228418e-05,
      "loss": 0.8619,
      "mean_token_accuracy": 0.7501156888902187,
      "num_tokens": 118494773.0,
      "step": 3713
    },
    {
      "entropy": 0.7411332316696644,
      "epoch": 0.34171166353269966,
      "grad_norm": 0.1661740094423294,
      "learning_rate": 8.861287453614256e-05,
      "loss": 0.7702,
      "mean_token_accuracy": 0.7777126356959343,
      "num_tokens": 118526520.0,
      "step": 3714
    },
    {
      "entropy": 0.891741719096899,
      "epoch": 0.34180366990414085,
      "grad_norm": 0.1674247831106186,
      "learning_rate": 8.860980771000093e-05,
      "loss": 0.8758,
      "mean_token_accuracy": 0.7437697537243366,
      "num_tokens": 118558430.0,
      "step": 3715
    },
    {
      "entropy": 0.8315691910684109,
      "epoch": 0.3418956762755821,
      "grad_norm": 0.15780077874660492,
      "learning_rate": 8.860674088385929e-05,
      "loss": 0.8308,
      "mean_token_accuracy": 0.7605084739625454,
      "num_tokens": 118590417.0,
      "step": 3716
    },
    {
      "entropy": 0.84157470241189,
      "epoch": 0.3419876826470233,
      "grad_norm": 0.1541234701871872,
      "learning_rate": 8.860367405771766e-05,
      "loss": 0.8301,
      "mean_token_accuracy": 0.757339108735323,
      "num_tokens": 118622980.0,
      "step": 3717
    },
    {
      "entropy": 0.8389298021793365,
      "epoch": 0.34207968901846453,
      "grad_norm": 0.1552625298500061,
      "learning_rate": 8.860060723157606e-05,
      "loss": 0.8543,
      "mean_token_accuracy": 0.7544471733272076,
      "num_tokens": 118655103.0,
      "step": 3718
    },
    {
      "entropy": 0.7533008437603712,
      "epoch": 0.3421716953899058,
      "grad_norm": 0.15825116634368896,
      "learning_rate": 8.859754040543442e-05,
      "loss": 0.7198,
      "mean_token_accuracy": 0.7923881858587265,
      "num_tokens": 118687485.0,
      "step": 3719
    },
    {
      "entropy": 0.7737366054207087,
      "epoch": 0.34226370176134696,
      "grad_norm": 0.1660526990890503,
      "learning_rate": 8.859447357929279e-05,
      "loss": 0.7765,
      "mean_token_accuracy": 0.7746818922460079,
      "num_tokens": 118719323.0,
      "step": 3720
    },
    {
      "entropy": 0.8127138763666153,
      "epoch": 0.3423557081327882,
      "grad_norm": 0.15834467113018036,
      "learning_rate": 8.859140675315116e-05,
      "loss": 0.821,
      "mean_token_accuracy": 0.7602910622954369,
      "num_tokens": 118751008.0,
      "step": 3721
    },
    {
      "entropy": 0.8320991545915604,
      "epoch": 0.3424477145042294,
      "grad_norm": 0.1546429842710495,
      "learning_rate": 8.858833992700954e-05,
      "loss": 0.8176,
      "mean_token_accuracy": 0.7600317597389221,
      "num_tokens": 118783058.0,
      "step": 3722
    },
    {
      "entropy": 0.9133446216583252,
      "epoch": 0.34253972087567064,
      "grad_norm": 0.15908969938755035,
      "learning_rate": 8.858527310086792e-05,
      "loss": 0.9052,
      "mean_token_accuracy": 0.7412885017693043,
      "num_tokens": 118814853.0,
      "step": 3723
    },
    {
      "entropy": 0.842214573174715,
      "epoch": 0.3426317272471119,
      "grad_norm": 0.15964275598526,
      "learning_rate": 8.858220627472629e-05,
      "loss": 0.839,
      "mean_token_accuracy": 0.7534457668662071,
      "num_tokens": 118846325.0,
      "step": 3724
    },
    {
      "entropy": 0.8598306141793728,
      "epoch": 0.3427237336185531,
      "grad_norm": 0.15833033621311188,
      "learning_rate": 8.857913944858467e-05,
      "loss": 0.8419,
      "mean_token_accuracy": 0.7549173831939697,
      "num_tokens": 118878207.0,
      "step": 3725
    },
    {
      "entropy": 0.8039496801793575,
      "epoch": 0.3428157399899943,
      "grad_norm": 0.15657667815685272,
      "learning_rate": 8.857607262244304e-05,
      "loss": 0.7863,
      "mean_token_accuracy": 0.7717664651572704,
      "num_tokens": 118910173.0,
      "step": 3726
    },
    {
      "entropy": 0.8667401075363159,
      "epoch": 0.3429077463614355,
      "grad_norm": 0.16042287647724152,
      "learning_rate": 8.857300579630141e-05,
      "loss": 0.874,
      "mean_token_accuracy": 0.7459614388644695,
      "num_tokens": 118942593.0,
      "step": 3727
    },
    {
      "entropy": 0.8220714125782251,
      "epoch": 0.34299975273287675,
      "grad_norm": 0.16127879917621613,
      "learning_rate": 8.856993897015979e-05,
      "loss": 0.8328,
      "mean_token_accuracy": 0.7599578388035297,
      "num_tokens": 118974974.0,
      "step": 3728
    },
    {
      "entropy": 0.8267379365861416,
      "epoch": 0.343091759104318,
      "grad_norm": 0.1541992872953415,
      "learning_rate": 8.856687214401817e-05,
      "loss": 0.8175,
      "mean_token_accuracy": 0.764610543847084,
      "num_tokens": 119007306.0,
      "step": 3729
    },
    {
      "entropy": 0.8897053375840187,
      "epoch": 0.3431837654757592,
      "grad_norm": 0.1666519194841385,
      "learning_rate": 8.856380531787654e-05,
      "loss": 0.8933,
      "mean_token_accuracy": 0.7457625940442085,
      "num_tokens": 119038185.0,
      "step": 3730
    },
    {
      "entropy": 0.9171824343502522,
      "epoch": 0.34327577184720043,
      "grad_norm": 0.15954819321632385,
      "learning_rate": 8.85607384917349e-05,
      "loss": 0.9003,
      "mean_token_accuracy": 0.7429808229207993,
      "num_tokens": 119070434.0,
      "step": 3731
    },
    {
      "entropy": 0.8549441378563643,
      "epoch": 0.3433677782186416,
      "grad_norm": 0.16388298571109772,
      "learning_rate": 8.855767166559327e-05,
      "loss": 0.8698,
      "mean_token_accuracy": 0.7500322386622429,
      "num_tokens": 119103020.0,
      "step": 3732
    },
    {
      "entropy": 0.9060702845454216,
      "epoch": 0.34345978459008286,
      "grad_norm": 0.16265486180782318,
      "learning_rate": 8.855460483945166e-05,
      "loss": 0.9116,
      "mean_token_accuracy": 0.7359498888254166,
      "num_tokens": 119133535.0,
      "step": 3733
    },
    {
      "entropy": 0.7444387692958117,
      "epoch": 0.3435517909615241,
      "grad_norm": 0.1556912362575531,
      "learning_rate": 8.855153801331004e-05,
      "loss": 0.7338,
      "mean_token_accuracy": 0.7858595922589302,
      "num_tokens": 119165495.0,
      "step": 3734
    },
    {
      "entropy": 0.8029676489531994,
      "epoch": 0.3436437973329653,
      "grad_norm": 0.1566811501979828,
      "learning_rate": 8.85484711871684e-05,
      "loss": 0.7801,
      "mean_token_accuracy": 0.7707334719598293,
      "num_tokens": 119197539.0,
      "step": 3735
    },
    {
      "entropy": 0.831120491027832,
      "epoch": 0.34373580370440654,
      "grad_norm": 0.16096343100070953,
      "learning_rate": 8.854540436102677e-05,
      "loss": 0.8155,
      "mean_token_accuracy": 0.7647062987089157,
      "num_tokens": 119229617.0,
      "step": 3736
    },
    {
      "entropy": 0.8321743197739124,
      "epoch": 0.34382781007584773,
      "grad_norm": 0.15893633663654327,
      "learning_rate": 8.854233753488515e-05,
      "loss": 0.8238,
      "mean_token_accuracy": 0.7573788538575172,
      "num_tokens": 119261848.0,
      "step": 3737
    },
    {
      "entropy": 0.911261148750782,
      "epoch": 0.343919816447289,
      "grad_norm": 0.16171090304851532,
      "learning_rate": 8.853927070874352e-05,
      "loss": 0.8933,
      "mean_token_accuracy": 0.7424199879169464,
      "num_tokens": 119293706.0,
      "step": 3738
    },
    {
      "entropy": 0.844773955643177,
      "epoch": 0.3440118228187302,
      "grad_norm": 0.1561574935913086,
      "learning_rate": 8.85362038826019e-05,
      "loss": 0.8311,
      "mean_token_accuracy": 0.7574186027050018,
      "num_tokens": 119325179.0,
      "step": 3739
    },
    {
      "entropy": 0.784180223941803,
      "epoch": 0.3441038291901714,
      "grad_norm": 0.1520143449306488,
      "learning_rate": 8.853313705646027e-05,
      "loss": 0.7566,
      "mean_token_accuracy": 0.7799860648810863,
      "num_tokens": 119356776.0,
      "step": 3740
    },
    {
      "entropy": 0.8523888625204563,
      "epoch": 0.34419583556161265,
      "grad_norm": 0.1556396782398224,
      "learning_rate": 8.853007023031865e-05,
      "loss": 0.8575,
      "mean_token_accuracy": 0.7518056407570839,
      "num_tokens": 119388479.0,
      "step": 3741
    },
    {
      "entropy": 0.8911075331270695,
      "epoch": 0.34428784193305384,
      "grad_norm": 0.16601233184337616,
      "learning_rate": 8.852700340417702e-05,
      "loss": 0.8863,
      "mean_token_accuracy": 0.7446825057268143,
      "num_tokens": 119421180.0,
      "step": 3742
    },
    {
      "entropy": 0.834352470934391,
      "epoch": 0.3443798483044951,
      "grad_norm": 0.15574300289154053,
      "learning_rate": 8.852393657803539e-05,
      "loss": 0.8386,
      "mean_token_accuracy": 0.7550987936556339,
      "num_tokens": 119453485.0,
      "step": 3743
    },
    {
      "entropy": 0.8214633259922266,
      "epoch": 0.34447185467593633,
      "grad_norm": 0.15685395896434784,
      "learning_rate": 8.852086975189377e-05,
      "loss": 0.8467,
      "mean_token_accuracy": 0.757653072476387,
      "num_tokens": 119485975.0,
      "step": 3744
    },
    {
      "entropy": 0.8120367974042892,
      "epoch": 0.3445638610473775,
      "grad_norm": 0.15783628821372986,
      "learning_rate": 8.851780292575215e-05,
      "loss": 0.8382,
      "mean_token_accuracy": 0.7634787335991859,
      "num_tokens": 119517696.0,
      "step": 3745
    },
    {
      "entropy": 0.8628420624881983,
      "epoch": 0.34465586741881876,
      "grad_norm": 0.15668772161006927,
      "learning_rate": 8.851473609961052e-05,
      "loss": 0.8642,
      "mean_token_accuracy": 0.7524092718958855,
      "num_tokens": 119549886.0,
      "step": 3746
    },
    {
      "entropy": 0.8682076577097178,
      "epoch": 0.34474787379025995,
      "grad_norm": 0.166350319981575,
      "learning_rate": 8.851166927346889e-05,
      "loss": 0.8626,
      "mean_token_accuracy": 0.7560263238847256,
      "num_tokens": 119581372.0,
      "step": 3747
    },
    {
      "entropy": 0.7474403753876686,
      "epoch": 0.3448398801617012,
      "grad_norm": 0.15186481177806854,
      "learning_rate": 8.850860244732725e-05,
      "loss": 0.7217,
      "mean_token_accuracy": 0.7856578379869461,
      "num_tokens": 119613547.0,
      "step": 3748
    },
    {
      "entropy": 0.8332053050398827,
      "epoch": 0.34493188653314244,
      "grad_norm": 0.1625697761774063,
      "learning_rate": 8.850553562118564e-05,
      "loss": 0.8281,
      "mean_token_accuracy": 0.7623979710042477,
      "num_tokens": 119646061.0,
      "step": 3749
    },
    {
      "entropy": 0.825550539419055,
      "epoch": 0.34502389290458363,
      "grad_norm": 0.16239915788173676,
      "learning_rate": 8.850246879504402e-05,
      "loss": 0.8106,
      "mean_token_accuracy": 0.7711258791387081,
      "num_tokens": 119677174.0,
      "step": 3750
    },
    {
      "entropy": 0.948045264929533,
      "epoch": 0.3451158992760249,
      "grad_norm": 0.16361881792545319,
      "learning_rate": 8.849940196890239e-05,
      "loss": 0.9444,
      "mean_token_accuracy": 0.7352516204118729,
      "num_tokens": 119709641.0,
      "step": 3751
    },
    {
      "entropy": 0.8236925341188908,
      "epoch": 0.34520790564746606,
      "grad_norm": 0.15780960023403168,
      "learning_rate": 8.849633514276075e-05,
      "loss": 0.8112,
      "mean_token_accuracy": 0.7655597552657127,
      "num_tokens": 119742033.0,
      "step": 3752
    },
    {
      "entropy": 0.9009114131331444,
      "epoch": 0.3452999120189073,
      "grad_norm": 0.16553598642349243,
      "learning_rate": 8.849326831661913e-05,
      "loss": 0.8739,
      "mean_token_accuracy": 0.7494820989668369,
      "num_tokens": 119773308.0,
      "step": 3753
    },
    {
      "entropy": 0.7686095181852579,
      "epoch": 0.34539191839034855,
      "grad_norm": 0.14496877789497375,
      "learning_rate": 8.84902014904775e-05,
      "loss": 0.7399,
      "mean_token_accuracy": 0.7880070582032204,
      "num_tokens": 119805517.0,
      "step": 3754
    },
    {
      "entropy": 0.8982774820178747,
      "epoch": 0.34548392476178974,
      "grad_norm": 0.15853466093540192,
      "learning_rate": 8.848713466433588e-05,
      "loss": 0.9055,
      "mean_token_accuracy": 0.7445856519043446,
      "num_tokens": 119838090.0,
      "step": 3755
    },
    {
      "entropy": 0.6721353512257338,
      "epoch": 0.345575931133231,
      "grad_norm": 0.1502189338207245,
      "learning_rate": 8.848406783819427e-05,
      "loss": 0.651,
      "mean_token_accuracy": 0.806375939399004,
      "num_tokens": 119870457.0,
      "step": 3756
    },
    {
      "entropy": 0.8785920813679695,
      "epoch": 0.3456679375046722,
      "grad_norm": 0.16576386988162994,
      "learning_rate": 8.848100101205263e-05,
      "loss": 0.8989,
      "mean_token_accuracy": 0.7427621744573116,
      "num_tokens": 119902426.0,
      "step": 3757
    },
    {
      "entropy": 0.8163319602608681,
      "epoch": 0.3457599438761134,
      "grad_norm": 0.16899247467517853,
      "learning_rate": 8.8477934185911e-05,
      "loss": 0.8393,
      "mean_token_accuracy": 0.760418251156807,
      "num_tokens": 119934387.0,
      "step": 3758
    },
    {
      "entropy": 0.8886992279440165,
      "epoch": 0.34585195024755466,
      "grad_norm": 0.17395682632923126,
      "learning_rate": 8.847486735976937e-05,
      "loss": 0.9121,
      "mean_token_accuracy": 0.7410960346460342,
      "num_tokens": 119967149.0,
      "step": 3759
    },
    {
      "entropy": 0.7635833472013474,
      "epoch": 0.34594395661899585,
      "grad_norm": 0.16510646045207977,
      "learning_rate": 8.847180053362776e-05,
      "loss": 0.7714,
      "mean_token_accuracy": 0.7766184322535992,
      "num_tokens": 119999528.0,
      "step": 3760
    },
    {
      "entropy": 0.869922574609518,
      "epoch": 0.3460359629904371,
      "grad_norm": 0.1642490029335022,
      "learning_rate": 8.846873370748613e-05,
      "loss": 0.8486,
      "mean_token_accuracy": 0.7600690573453903,
      "num_tokens": 120031090.0,
      "step": 3761
    },
    {
      "entropy": 0.8930163271725178,
      "epoch": 0.3461279693618783,
      "grad_norm": 0.157710999250412,
      "learning_rate": 8.84656668813445e-05,
      "loss": 0.8956,
      "mean_token_accuracy": 0.7424335554242134,
      "num_tokens": 120063169.0,
      "step": 3762
    },
    {
      "entropy": 0.8254711236804724,
      "epoch": 0.34621997573331953,
      "grad_norm": 0.154154434800148,
      "learning_rate": 8.846260005520287e-05,
      "loss": 0.8042,
      "mean_token_accuracy": 0.7686141692101955,
      "num_tokens": 120095531.0,
      "step": 3763
    },
    {
      "entropy": 0.8259627223014832,
      "epoch": 0.3463119821047608,
      "grad_norm": 0.15411938726902008,
      "learning_rate": 8.845953322906125e-05,
      "loss": 0.805,
      "mean_token_accuracy": 0.7685264982283115,
      "num_tokens": 120128205.0,
      "step": 3764
    },
    {
      "entropy": 0.8197680991142988,
      "epoch": 0.34640398847620196,
      "grad_norm": 0.15063221752643585,
      "learning_rate": 8.845646640291963e-05,
      "loss": 0.8031,
      "mean_token_accuracy": 0.7685605064034462,
      "num_tokens": 120160023.0,
      "step": 3765
    },
    {
      "entropy": 0.800836306065321,
      "epoch": 0.3464959948476432,
      "grad_norm": 0.15559294819831848,
      "learning_rate": 8.8453399576778e-05,
      "loss": 0.7876,
      "mean_token_accuracy": 0.7690862119197845,
      "num_tokens": 120191337.0,
      "step": 3766
    },
    {
      "entropy": 0.8540994934737682,
      "epoch": 0.3465880012190844,
      "grad_norm": 0.16338646411895752,
      "learning_rate": 8.845033275063637e-05,
      "loss": 0.8436,
      "mean_token_accuracy": 0.7573985233902931,
      "num_tokens": 120223213.0,
      "step": 3767
    },
    {
      "entropy": 0.8612070381641388,
      "epoch": 0.34668000759052564,
      "grad_norm": 0.15954744815826416,
      "learning_rate": 8.844726592449475e-05,
      "loss": 0.8312,
      "mean_token_accuracy": 0.7589932307600975,
      "num_tokens": 120254686.0,
      "step": 3768
    },
    {
      "entropy": 0.7959731034934521,
      "epoch": 0.3467720139619669,
      "grad_norm": 0.15368808805942535,
      "learning_rate": 8.844419909835312e-05,
      "loss": 0.7775,
      "mean_token_accuracy": 0.7701177038252354,
      "num_tokens": 120286555.0,
      "step": 3769
    },
    {
      "entropy": 0.8628717139363289,
      "epoch": 0.3468640203334081,
      "grad_norm": 0.16074331104755402,
      "learning_rate": 8.84411322722115e-05,
      "loss": 0.8665,
      "mean_token_accuracy": 0.7494498528540134,
      "num_tokens": 120318196.0,
      "step": 3770
    },
    {
      "entropy": 0.8907548636198044,
      "epoch": 0.3469560267048493,
      "grad_norm": 0.16904211044311523,
      "learning_rate": 8.843806544606986e-05,
      "loss": 0.8771,
      "mean_token_accuracy": 0.7499096691608429,
      "num_tokens": 120349157.0,
      "step": 3771
    },
    {
      "entropy": 0.9293929599225521,
      "epoch": 0.3470480330762905,
      "grad_norm": 0.17182472348213196,
      "learning_rate": 8.843499861992825e-05,
      "loss": 0.9596,
      "mean_token_accuracy": 0.7327613607048988,
      "num_tokens": 120380977.0,
      "step": 3772
    },
    {
      "entropy": 0.8088716734200716,
      "epoch": 0.34714003944773175,
      "grad_norm": 0.15974104404449463,
      "learning_rate": 8.843193179378661e-05,
      "loss": 0.8037,
      "mean_token_accuracy": 0.765363022685051,
      "num_tokens": 120413252.0,
      "step": 3773
    },
    {
      "entropy": 0.8099780157208443,
      "epoch": 0.347232045819173,
      "grad_norm": 0.16105832159519196,
      "learning_rate": 8.842886496764498e-05,
      "loss": 0.8271,
      "mean_token_accuracy": 0.7653442844748497,
      "num_tokens": 120445029.0,
      "step": 3774
    },
    {
      "entropy": 0.8192146420478821,
      "epoch": 0.3473240521906142,
      "grad_norm": 0.16290916502475739,
      "learning_rate": 8.842579814150336e-05,
      "loss": 0.8188,
      "mean_token_accuracy": 0.7618572600185871,
      "num_tokens": 120477159.0,
      "step": 3775
    },
    {
      "entropy": 0.7573690358549356,
      "epoch": 0.34741605856205543,
      "grad_norm": 0.15212994813919067,
      "learning_rate": 8.842273131536174e-05,
      "loss": 0.7343,
      "mean_token_accuracy": 0.7854475863277912,
      "num_tokens": 120509282.0,
      "step": 3776
    },
    {
      "entropy": 0.9038415960967541,
      "epoch": 0.3475080649334966,
      "grad_norm": 0.15961048007011414,
      "learning_rate": 8.841966448922011e-05,
      "loss": 0.9016,
      "mean_token_accuracy": 0.7447250969707966,
      "num_tokens": 120541185.0,
      "step": 3777
    },
    {
      "entropy": 0.8827890865504742,
      "epoch": 0.34760007130493786,
      "grad_norm": 0.15730538964271545,
      "learning_rate": 8.841659766307848e-05,
      "loss": 0.8622,
      "mean_token_accuracy": 0.7533799931406975,
      "num_tokens": 120573732.0,
      "step": 3778
    },
    {
      "entropy": 0.7959932908415794,
      "epoch": 0.3476920776763791,
      "grad_norm": 0.15313290059566498,
      "learning_rate": 8.841353083693685e-05,
      "loss": 0.7793,
      "mean_token_accuracy": 0.7766465917229652,
      "num_tokens": 120605825.0,
      "step": 3779
    },
    {
      "entropy": 0.8593918941915035,
      "epoch": 0.3477840840478203,
      "grad_norm": 0.1587131917476654,
      "learning_rate": 8.841046401079523e-05,
      "loss": 0.8352,
      "mean_token_accuracy": 0.7578673735260963,
      "num_tokens": 120638297.0,
      "step": 3780
    },
    {
      "entropy": 0.7576248534023762,
      "epoch": 0.34787609041926154,
      "grad_norm": 0.15267202258110046,
      "learning_rate": 8.840739718465361e-05,
      "loss": 0.7215,
      "mean_token_accuracy": 0.7895100191235542,
      "num_tokens": 120669867.0,
      "step": 3781
    },
    {
      "entropy": 0.8344754688441753,
      "epoch": 0.34796809679070273,
      "grad_norm": 0.16018937528133392,
      "learning_rate": 8.840433035851198e-05,
      "loss": 0.8208,
      "mean_token_accuracy": 0.762143649160862,
      "num_tokens": 120701102.0,
      "step": 3782
    },
    {
      "entropy": 0.7512946473434567,
      "epoch": 0.348060103162144,
      "grad_norm": 0.1509130895137787,
      "learning_rate": 8.840126353237035e-05,
      "loss": 0.743,
      "mean_token_accuracy": 0.7821798101067543,
      "num_tokens": 120733561.0,
      "step": 3783
    },
    {
      "entropy": 0.855661615729332,
      "epoch": 0.3481521095335852,
      "grad_norm": 0.1672372967004776,
      "learning_rate": 8.839819670622873e-05,
      "loss": 0.8659,
      "mean_token_accuracy": 0.7571824677288532,
      "num_tokens": 120765447.0,
      "step": 3784
    },
    {
      "entropy": 0.7309505622833967,
      "epoch": 0.3482441159050264,
      "grad_norm": 0.1552681177854538,
      "learning_rate": 8.83951298800871e-05,
      "loss": 0.7373,
      "mean_token_accuracy": 0.7844927795231342,
      "num_tokens": 120797992.0,
      "step": 3785
    },
    {
      "entropy": 0.7808074932545424,
      "epoch": 0.34833612227646765,
      "grad_norm": 0.15462271869182587,
      "learning_rate": 8.839206305394548e-05,
      "loss": 0.7773,
      "mean_token_accuracy": 0.773887999355793,
      "num_tokens": 120830167.0,
      "step": 3786
    },
    {
      "entropy": 0.8600835539400578,
      "epoch": 0.34842812864790884,
      "grad_norm": 0.16238686442375183,
      "learning_rate": 8.838899622780386e-05,
      "loss": 0.8783,
      "mean_token_accuracy": 0.7487959004938602,
      "num_tokens": 120861918.0,
      "step": 3787
    },
    {
      "entropy": 0.8208338208496571,
      "epoch": 0.3485201350193501,
      "grad_norm": 0.16467203199863434,
      "learning_rate": 8.838592940166223e-05,
      "loss": 0.8444,
      "mean_token_accuracy": 0.7562086284160614,
      "num_tokens": 120894306.0,
      "step": 3788
    },
    {
      "entropy": 0.7702032849192619,
      "epoch": 0.34861214139079133,
      "grad_norm": 0.15921542048454285,
      "learning_rate": 8.83828625755206e-05,
      "loss": 0.7864,
      "mean_token_accuracy": 0.7693427875638008,
      "num_tokens": 120926313.0,
      "step": 3789
    },
    {
      "entropy": 0.8684881329536438,
      "epoch": 0.3487041477622325,
      "grad_norm": 0.1604766547679901,
      "learning_rate": 8.837979574937896e-05,
      "loss": 0.8625,
      "mean_token_accuracy": 0.7531568855047226,
      "num_tokens": 120958632.0,
      "step": 3790
    },
    {
      "entropy": 0.807744087651372,
      "epoch": 0.34879615413367376,
      "grad_norm": 0.16447189450263977,
      "learning_rate": 8.837672892323734e-05,
      "loss": 0.7942,
      "mean_token_accuracy": 0.7721187360584736,
      "num_tokens": 120990072.0,
      "step": 3791
    },
    {
      "entropy": 0.7503185141831636,
      "epoch": 0.34888816050511495,
      "grad_norm": 0.15391848981380463,
      "learning_rate": 8.837366209709573e-05,
      "loss": 0.7288,
      "mean_token_accuracy": 0.7851504683494568,
      "num_tokens": 121022201.0,
      "step": 3792
    },
    {
      "entropy": 0.8460014723241329,
      "epoch": 0.3489801668765562,
      "grad_norm": 0.1594490110874176,
      "learning_rate": 8.83705952709541e-05,
      "loss": 0.8278,
      "mean_token_accuracy": 0.7578670121729374,
      "num_tokens": 121054601.0,
      "step": 3793
    },
    {
      "entropy": 0.8863828256726265,
      "epoch": 0.34907217324799744,
      "grad_norm": 0.16350792348384857,
      "learning_rate": 8.836752844481246e-05,
      "loss": 0.8778,
      "mean_token_accuracy": 0.7501875571906567,
      "num_tokens": 121086687.0,
      "step": 3794
    },
    {
      "entropy": 0.9087334871292114,
      "epoch": 0.34916417961943863,
      "grad_norm": 0.15923047065734863,
      "learning_rate": 8.836446161867084e-05,
      "loss": 0.8773,
      "mean_token_accuracy": 0.7490927949547768,
      "num_tokens": 121118952.0,
      "step": 3795
    },
    {
      "entropy": 0.8727920632809401,
      "epoch": 0.3492561859908799,
      "grad_norm": 0.15596643090248108,
      "learning_rate": 8.836139479252922e-05,
      "loss": 0.8724,
      "mean_token_accuracy": 0.7515299208462238,
      "num_tokens": 121150862.0,
      "step": 3796
    },
    {
      "entropy": 0.8969176150858402,
      "epoch": 0.34934819236232106,
      "grad_norm": 0.1582353711128235,
      "learning_rate": 8.835832796638759e-05,
      "loss": 0.8813,
      "mean_token_accuracy": 0.7474814653396606,
      "num_tokens": 121183285.0,
      "step": 3797
    },
    {
      "entropy": 0.9163975678384304,
      "epoch": 0.3494401987337623,
      "grad_norm": 0.16705971956253052,
      "learning_rate": 8.835526114024596e-05,
      "loss": 0.9029,
      "mean_token_accuracy": 0.742749709635973,
      "num_tokens": 121214910.0,
      "step": 3798
    },
    {
      "entropy": 0.7214012145996094,
      "epoch": 0.34953220510520355,
      "grad_norm": 0.15191897749900818,
      "learning_rate": 8.835219431410434e-05,
      "loss": 0.719,
      "mean_token_accuracy": 0.7912633307278156,
      "num_tokens": 121246817.0,
      "step": 3799
    },
    {
      "entropy": 0.9089645333588123,
      "epoch": 0.34962421147664474,
      "grad_norm": 0.16077820956707,
      "learning_rate": 8.834912748796271e-05,
      "loss": 0.9325,
      "mean_token_accuracy": 0.7304529994726181,
      "num_tokens": 121279323.0,
      "step": 3800
    },
    {
      "entropy": 0.9179159291088581,
      "epoch": 0.349716217848086,
      "grad_norm": 0.16571374237537384,
      "learning_rate": 8.834606066182109e-05,
      "loss": 0.9107,
      "mean_token_accuracy": 0.7401226945221424,
      "num_tokens": 121311754.0,
      "step": 3801
    },
    {
      "entropy": 0.9194784238934517,
      "epoch": 0.3498082242195272,
      "grad_norm": 0.15678083896636963,
      "learning_rate": 8.834299383567946e-05,
      "loss": 0.9129,
      "mean_token_accuracy": 0.7389914393424988,
      "num_tokens": 121343378.0,
      "step": 3802
    },
    {
      "entropy": 0.7819661758840084,
      "epoch": 0.3499002305909684,
      "grad_norm": 0.16026699542999268,
      "learning_rate": 8.833992700953784e-05,
      "loss": 0.7977,
      "mean_token_accuracy": 0.7694789320230484,
      "num_tokens": 121375688.0,
      "step": 3803
    },
    {
      "entropy": 0.8505206853151321,
      "epoch": 0.34999223696240966,
      "grad_norm": 0.17495007812976837,
      "learning_rate": 8.833686018339621e-05,
      "loss": 0.8629,
      "mean_token_accuracy": 0.7496790401637554,
      "num_tokens": 121408034.0,
      "step": 3804
    },
    {
      "entropy": 0.8036221768707037,
      "epoch": 0.35008424333385085,
      "grad_norm": 0.15931063890457153,
      "learning_rate": 8.833379335725458e-05,
      "loss": 0.8027,
      "mean_token_accuracy": 0.772084679454565,
      "num_tokens": 121440090.0,
      "step": 3805
    },
    {
      "entropy": 0.9043404422700405,
      "epoch": 0.3501762497052921,
      "grad_norm": 0.15718883275985718,
      "learning_rate": 8.833072653111296e-05,
      "loss": 0.9059,
      "mean_token_accuracy": 0.7422502785921097,
      "num_tokens": 121472429.0,
      "step": 3806
    },
    {
      "entropy": 0.7461367975920439,
      "epoch": 0.3502682560767333,
      "grad_norm": 0.15172743797302246,
      "learning_rate": 8.832765970497134e-05,
      "loss": 0.738,
      "mean_token_accuracy": 0.7830536663532257,
      "num_tokens": 121504709.0,
      "step": 3807
    },
    {
      "entropy": 0.8562102783471346,
      "epoch": 0.35036026244817453,
      "grad_norm": 0.1554148942232132,
      "learning_rate": 8.83245928788297e-05,
      "loss": 0.8448,
      "mean_token_accuracy": 0.7566382996737957,
      "num_tokens": 121536509.0,
      "step": 3808
    },
    {
      "entropy": 0.6564690433442593,
      "epoch": 0.3504522688196158,
      "grad_norm": 0.14341983199119568,
      "learning_rate": 8.832152605268807e-05,
      "loss": 0.6289,
      "mean_token_accuracy": 0.8125,
      "num_tokens": 121569277.0,
      "step": 3809
    },
    {
      "entropy": 0.7488473411649466,
      "epoch": 0.35054427519105696,
      "grad_norm": 0.15279901027679443,
      "learning_rate": 8.831845922654644e-05,
      "loss": 0.7093,
      "mean_token_accuracy": 0.7924327105283737,
      "num_tokens": 121601476.0,
      "step": 3810
    },
    {
      "entropy": 0.7706202641129494,
      "epoch": 0.3506362815624982,
      "grad_norm": 0.14755985140800476,
      "learning_rate": 8.831539240040482e-05,
      "loss": 0.7592,
      "mean_token_accuracy": 0.7795416861772537,
      "num_tokens": 121633700.0,
      "step": 3811
    },
    {
      "entropy": 0.9353910367935896,
      "epoch": 0.3507282879339394,
      "grad_norm": 0.16320650279521942,
      "learning_rate": 8.83123255742632e-05,
      "loss": 0.9147,
      "mean_token_accuracy": 0.7432449571788311,
      "num_tokens": 121665911.0,
      "step": 3812
    },
    {
      "entropy": 0.8447041548788548,
      "epoch": 0.35082029430538064,
      "grad_norm": 0.16096341609954834,
      "learning_rate": 8.830925874812157e-05,
      "loss": 0.8401,
      "mean_token_accuracy": 0.7569751925766468,
      "num_tokens": 121697803.0,
      "step": 3813
    },
    {
      "entropy": 0.8196309600025415,
      "epoch": 0.3509123006768219,
      "grad_norm": 0.16276951134204865,
      "learning_rate": 8.830619192197995e-05,
      "loss": 0.8152,
      "mean_token_accuracy": 0.7634732387959957,
      "num_tokens": 121729192.0,
      "step": 3814
    },
    {
      "entropy": 0.9480710439383984,
      "epoch": 0.3510043070482631,
      "grad_norm": 0.17037439346313477,
      "learning_rate": 8.830312509583832e-05,
      "loss": 0.9635,
      "mean_token_accuracy": 0.7302686125040054,
      "num_tokens": 121761412.0,
      "step": 3815
    },
    {
      "entropy": 0.875516664236784,
      "epoch": 0.3510963134197043,
      "grad_norm": 0.17799502611160278,
      "learning_rate": 8.830005826969669e-05,
      "loss": 0.9059,
      "mean_token_accuracy": 0.7403752617537975,
      "num_tokens": 121792513.0,
      "step": 3816
    },
    {
      "entropy": 0.7959139719605446,
      "epoch": 0.35118831979114556,
      "grad_norm": 0.1629268079996109,
      "learning_rate": 8.829699144355507e-05,
      "loss": 0.8054,
      "mean_token_accuracy": 0.7710596211254597,
      "num_tokens": 121824691.0,
      "step": 3817
    },
    {
      "entropy": 0.8259482327848673,
      "epoch": 0.35128032616258675,
      "grad_norm": 0.1591295748949051,
      "learning_rate": 8.829392461741345e-05,
      "loss": 0.8153,
      "mean_token_accuracy": 0.764889270067215,
      "num_tokens": 121855920.0,
      "step": 3818
    },
    {
      "entropy": 0.88987672701478,
      "epoch": 0.351372332534028,
      "grad_norm": 0.16491180658340454,
      "learning_rate": 8.829085779127182e-05,
      "loss": 0.8918,
      "mean_token_accuracy": 0.7427677027881145,
      "num_tokens": 121887524.0,
      "step": 3819
    },
    {
      "entropy": 0.874864848330617,
      "epoch": 0.3514643389054692,
      "grad_norm": 0.16357478499412537,
      "learning_rate": 8.828779096513019e-05,
      "loss": 0.8743,
      "mean_token_accuracy": 0.7484281547367573,
      "num_tokens": 121919180.0,
      "step": 3820
    },
    {
      "entropy": 0.7600622475147247,
      "epoch": 0.35155634527691043,
      "grad_norm": 0.15744197368621826,
      "learning_rate": 8.828472413898856e-05,
      "loss": 0.7557,
      "mean_token_accuracy": 0.7795375213027,
      "num_tokens": 121950833.0,
      "step": 3821
    },
    {
      "entropy": 0.8302732352167368,
      "epoch": 0.3516483516483517,
      "grad_norm": 0.15557779371738434,
      "learning_rate": 8.828165731284694e-05,
      "loss": 0.8259,
      "mean_token_accuracy": 0.76119564473629,
      "num_tokens": 121982481.0,
      "step": 3822
    },
    {
      "entropy": 0.8397612608969212,
      "epoch": 0.35174035801979286,
      "grad_norm": 0.16073249280452728,
      "learning_rate": 8.827859048670532e-05,
      "loss": 0.8408,
      "mean_token_accuracy": 0.7554504722356796,
      "num_tokens": 122015010.0,
      "step": 3823
    },
    {
      "entropy": 0.8164195884019136,
      "epoch": 0.3518323643912341,
      "grad_norm": 0.16060827672481537,
      "learning_rate": 8.827552366056369e-05,
      "loss": 0.7887,
      "mean_token_accuracy": 0.7712399773299694,
      "num_tokens": 122046890.0,
      "step": 3824
    },
    {
      "entropy": 0.7743039652705193,
      "epoch": 0.3519243707626753,
      "grad_norm": 0.15449605882167816,
      "learning_rate": 8.827245683442206e-05,
      "loss": 0.7631,
      "mean_token_accuracy": 0.7814563997089863,
      "num_tokens": 122079017.0,
      "step": 3825
    },
    {
      "entropy": 0.8188587035983801,
      "epoch": 0.35201637713411654,
      "grad_norm": 0.1573369950056076,
      "learning_rate": 8.826939000828044e-05,
      "loss": 0.8159,
      "mean_token_accuracy": 0.7645903825759888,
      "num_tokens": 122110946.0,
      "step": 3826
    },
    {
      "entropy": 0.8627975285053253,
      "epoch": 0.3521083835055578,
      "grad_norm": 0.15811602771282196,
      "learning_rate": 8.82663231821388e-05,
      "loss": 0.8425,
      "mean_token_accuracy": 0.7563510276377201,
      "num_tokens": 122143051.0,
      "step": 3827
    },
    {
      "entropy": 1.0821545831859112,
      "epoch": 0.352200389876999,
      "grad_norm": 0.17075486481189728,
      "learning_rate": 8.826325635599719e-05,
      "loss": 1.0594,
      "mean_token_accuracy": 0.7032840810716152,
      "num_tokens": 122175195.0,
      "step": 3828
    },
    {
      "entropy": 0.8051436506211758,
      "epoch": 0.3522923962484402,
      "grad_norm": 0.1567111760377884,
      "learning_rate": 8.826018952985555e-05,
      "loss": 0.7811,
      "mean_token_accuracy": 0.774740856140852,
      "num_tokens": 122207118.0,
      "step": 3829
    },
    {
      "entropy": 0.8385220132768154,
      "epoch": 0.3523844026198814,
      "grad_norm": 0.15761420130729675,
      "learning_rate": 8.825712270371394e-05,
      "loss": 0.8278,
      "mean_token_accuracy": 0.7542272657155991,
      "num_tokens": 122239210.0,
      "step": 3830
    },
    {
      "entropy": 0.8328091837465763,
      "epoch": 0.35247640899132265,
      "grad_norm": 0.16675327718257904,
      "learning_rate": 8.82540558775723e-05,
      "loss": 0.8342,
      "mean_token_accuracy": 0.7618678510189056,
      "num_tokens": 122271141.0,
      "step": 3831
    },
    {
      "entropy": 0.8425095286220312,
      "epoch": 0.3525684153627639,
      "grad_norm": 0.1652573198080063,
      "learning_rate": 8.825098905143067e-05,
      "loss": 0.8494,
      "mean_token_accuracy": 0.7551990896463394,
      "num_tokens": 122303157.0,
      "step": 3832
    },
    {
      "entropy": 0.8279029037803411,
      "epoch": 0.3526604217342051,
      "grad_norm": 0.16322585940361023,
      "learning_rate": 8.824792222528905e-05,
      "loss": 0.8199,
      "mean_token_accuracy": 0.7626885883510113,
      "num_tokens": 122334204.0,
      "step": 3833
    },
    {
      "entropy": 0.8830248359590769,
      "epoch": 0.35275242810564633,
      "grad_norm": 0.16810938715934753,
      "learning_rate": 8.824485539914743e-05,
      "loss": 0.8726,
      "mean_token_accuracy": 0.7491313070058823,
      "num_tokens": 122365765.0,
      "step": 3834
    },
    {
      "entropy": 0.7702563591301441,
      "epoch": 0.3528444344770875,
      "grad_norm": 0.1581006944179535,
      "learning_rate": 8.82417885730058e-05,
      "loss": 0.7741,
      "mean_token_accuracy": 0.7729658782482147,
      "num_tokens": 122397695.0,
      "step": 3835
    },
    {
      "entropy": 0.8075187727808952,
      "epoch": 0.35293644084852877,
      "grad_norm": 0.1620257943868637,
      "learning_rate": 8.823872174686417e-05,
      "loss": 0.8082,
      "mean_token_accuracy": 0.7631422579288483,
      "num_tokens": 122428785.0,
      "step": 3836
    },
    {
      "entropy": 0.7897501513361931,
      "epoch": 0.35302844721997,
      "grad_norm": 0.15705513954162598,
      "learning_rate": 8.823565492072254e-05,
      "loss": 0.7869,
      "mean_token_accuracy": 0.7673697993159294,
      "num_tokens": 122460000.0,
      "step": 3837
    },
    {
      "entropy": 0.8912017233669758,
      "epoch": 0.3531204535914112,
      "grad_norm": 0.16089780628681183,
      "learning_rate": 8.823258809458093e-05,
      "loss": 0.8863,
      "mean_token_accuracy": 0.7509891092777252,
      "num_tokens": 122492439.0,
      "step": 3838
    },
    {
      "entropy": 0.8066176995635033,
      "epoch": 0.35321245996285244,
      "grad_norm": 0.1527390480041504,
      "learning_rate": 8.82295212684393e-05,
      "loss": 0.8052,
      "mean_token_accuracy": 0.7694367207586765,
      "num_tokens": 122524707.0,
      "step": 3839
    },
    {
      "entropy": 0.8893731161952019,
      "epoch": 0.35330446633429363,
      "grad_norm": 0.16149450838565826,
      "learning_rate": 8.822645444229767e-05,
      "loss": 0.8964,
      "mean_token_accuracy": 0.7384008839726448,
      "num_tokens": 122556170.0,
      "step": 3840
    },
    {
      "entropy": 0.7831693235784769,
      "epoch": 0.3533964727057349,
      "grad_norm": 0.1529574990272522,
      "learning_rate": 8.822338761615604e-05,
      "loss": 0.7465,
      "mean_token_accuracy": 0.7821817770600319,
      "num_tokens": 122587730.0,
      "step": 3841
    },
    {
      "entropy": 0.8369538523256779,
      "epoch": 0.3534884790771761,
      "grad_norm": 0.15401659905910492,
      "learning_rate": 8.822032079001442e-05,
      "loss": 0.8286,
      "mean_token_accuracy": 0.7553616277873516,
      "num_tokens": 122619128.0,
      "step": 3842
    },
    {
      "entropy": 0.8492603972554207,
      "epoch": 0.3535804854486173,
      "grad_norm": 0.15758611261844635,
      "learning_rate": 8.82172539638728e-05,
      "loss": 0.8529,
      "mean_token_accuracy": 0.7506515979766846,
      "num_tokens": 122651039.0,
      "step": 3843
    },
    {
      "entropy": 0.7711849268525839,
      "epoch": 0.35367249182005855,
      "grad_norm": 0.15675005316734314,
      "learning_rate": 8.821418713773117e-05,
      "loss": 0.7737,
      "mean_token_accuracy": 0.7778422795236111,
      "num_tokens": 122682918.0,
      "step": 3844
    },
    {
      "entropy": 0.8688315860927105,
      "epoch": 0.35376449819149974,
      "grad_norm": 0.1648833006620407,
      "learning_rate": 8.821112031158955e-05,
      "loss": 0.8608,
      "mean_token_accuracy": 0.7552086375653744,
      "num_tokens": 122714097.0,
      "step": 3845
    },
    {
      "entropy": 0.8567625191062689,
      "epoch": 0.353856504562941,
      "grad_norm": 0.15732401609420776,
      "learning_rate": 8.820805348544792e-05,
      "loss": 0.8385,
      "mean_token_accuracy": 0.7538790851831436,
      "num_tokens": 122746391.0,
      "step": 3846
    },
    {
      "entropy": 0.8448572624474764,
      "epoch": 0.35394851093438223,
      "grad_norm": 0.15853460133075714,
      "learning_rate": 8.820498665930628e-05,
      "loss": 0.8305,
      "mean_token_accuracy": 0.7594806179404259,
      "num_tokens": 122779093.0,
      "step": 3847
    },
    {
      "entropy": 0.8263257928192616,
      "epoch": 0.3540405173058234,
      "grad_norm": 0.16415563225746155,
      "learning_rate": 8.820191983316467e-05,
      "loss": 0.8207,
      "mean_token_accuracy": 0.7561747431755066,
      "num_tokens": 122810440.0,
      "step": 3848
    },
    {
      "entropy": 0.8424420468509197,
      "epoch": 0.35413252367726467,
      "grad_norm": 0.17686134576797485,
      "learning_rate": 8.819885300702305e-05,
      "loss": 0.8448,
      "mean_token_accuracy": 0.7550505883991718,
      "num_tokens": 122842353.0,
      "step": 3849
    },
    {
      "entropy": 0.8057922217994928,
      "epoch": 0.35422453004870585,
      "grad_norm": 0.16463807225227356,
      "learning_rate": 8.819578618088141e-05,
      "loss": 0.8283,
      "mean_token_accuracy": 0.7633196860551834,
      "num_tokens": 122874782.0,
      "step": 3850
    },
    {
      "entropy": 0.7900019697844982,
      "epoch": 0.3543165364201471,
      "grad_norm": 0.15891534090042114,
      "learning_rate": 8.819271935473978e-05,
      "loss": 0.7923,
      "mean_token_accuracy": 0.7711809761822224,
      "num_tokens": 122907110.0,
      "step": 3851
    },
    {
      "entropy": 0.9383850134909153,
      "epoch": 0.35440854279158834,
      "grad_norm": 0.17464902997016907,
      "learning_rate": 8.818965252859815e-05,
      "loss": 0.9589,
      "mean_token_accuracy": 0.7292192690074444,
      "num_tokens": 122938307.0,
      "step": 3852
    },
    {
      "entropy": 0.762123566120863,
      "epoch": 0.35450054916302953,
      "grad_norm": 0.15573818981647491,
      "learning_rate": 8.818658570245653e-05,
      "loss": 0.7407,
      "mean_token_accuracy": 0.7855227217078209,
      "num_tokens": 122970687.0,
      "step": 3853
    },
    {
      "entropy": 0.8000229820609093,
      "epoch": 0.3545925555344708,
      "grad_norm": 0.15551333129405975,
      "learning_rate": 8.818351887631491e-05,
      "loss": 0.7874,
      "mean_token_accuracy": 0.7688067071139812,
      "num_tokens": 123003298.0,
      "step": 3854
    },
    {
      "entropy": 0.7684974893927574,
      "epoch": 0.35468456190591197,
      "grad_norm": 0.15342703461647034,
      "learning_rate": 8.818045205017328e-05,
      "loss": 0.7457,
      "mean_token_accuracy": 0.7840672135353088,
      "num_tokens": 123034937.0,
      "step": 3855
    },
    {
      "entropy": 0.8323111236095428,
      "epoch": 0.3547765682773532,
      "grad_norm": 0.15924830734729767,
      "learning_rate": 8.817738522403165e-05,
      "loss": 0.7957,
      "mean_token_accuracy": 0.7730872668325901,
      "num_tokens": 123066840.0,
      "step": 3856
    },
    {
      "entropy": 0.8601456582546234,
      "epoch": 0.35486857464879445,
      "grad_norm": 0.15665236115455627,
      "learning_rate": 8.817431839789003e-05,
      "loss": 0.8528,
      "mean_token_accuracy": 0.7530954480171204,
      "num_tokens": 123098401.0,
      "step": 3857
    },
    {
      "entropy": 0.8202933073043823,
      "epoch": 0.35496058102023564,
      "grad_norm": 0.1641482710838318,
      "learning_rate": 8.81712515717484e-05,
      "loss": 0.8201,
      "mean_token_accuracy": 0.7698766589164734,
      "num_tokens": 123130251.0,
      "step": 3858
    },
    {
      "entropy": 0.753490399569273,
      "epoch": 0.3550525873916769,
      "grad_norm": 0.15789639949798584,
      "learning_rate": 8.816818474560678e-05,
      "loss": 0.747,
      "mean_token_accuracy": 0.7909980602562428,
      "num_tokens": 123162356.0,
      "step": 3859
    },
    {
      "entropy": 0.8231643810868263,
      "epoch": 0.3551445937631181,
      "grad_norm": 0.1605677604675293,
      "learning_rate": 8.816511791946515e-05,
      "loss": 0.8077,
      "mean_token_accuracy": 0.7685207389295101,
      "num_tokens": 123193299.0,
      "step": 3860
    },
    {
      "entropy": 0.8217325210571289,
      "epoch": 0.3552366001345593,
      "grad_norm": 0.15912602841854095,
      "learning_rate": 8.816205109332353e-05,
      "loss": 0.8261,
      "mean_token_accuracy": 0.7666271142661572,
      "num_tokens": 123225394.0,
      "step": 3861
    },
    {
      "entropy": 0.8839656822383404,
      "epoch": 0.35532860650600057,
      "grad_norm": 0.1575368493795395,
      "learning_rate": 8.81589842671819e-05,
      "loss": 0.8784,
      "mean_token_accuracy": 0.7431202977895737,
      "num_tokens": 123256934.0,
      "step": 3862
    },
    {
      "entropy": 0.7363398112356663,
      "epoch": 0.35542061287744176,
      "grad_norm": 0.16147452592849731,
      "learning_rate": 8.815591744104026e-05,
      "loss": 0.7504,
      "mean_token_accuracy": 0.7833988666534424,
      "num_tokens": 123288646.0,
      "step": 3863
    },
    {
      "entropy": 0.8046110030263662,
      "epoch": 0.355512619248883,
      "grad_norm": 0.17090420424938202,
      "learning_rate": 8.815285061489865e-05,
      "loss": 0.8261,
      "mean_token_accuracy": 0.7635264359414577,
      "num_tokens": 123320169.0,
      "step": 3864
    },
    {
      "entropy": 0.7578581757843494,
      "epoch": 0.3556046256203242,
      "grad_norm": 0.1523044854402542,
      "learning_rate": 8.814978378875703e-05,
      "loss": 0.7507,
      "mean_token_accuracy": 0.7794568948447704,
      "num_tokens": 123351846.0,
      "step": 3865
    },
    {
      "entropy": 0.8991098701953888,
      "epoch": 0.35569663199176543,
      "grad_norm": 0.16565953195095062,
      "learning_rate": 8.81467169626154e-05,
      "loss": 0.9113,
      "mean_token_accuracy": 0.7422010563313961,
      "num_tokens": 123383091.0,
      "step": 3866
    },
    {
      "entropy": 0.7952282018959522,
      "epoch": 0.3557886383632067,
      "grad_norm": 0.15644925832748413,
      "learning_rate": 8.814365013647376e-05,
      "loss": 0.7705,
      "mean_token_accuracy": 0.7758955471217632,
      "num_tokens": 123415358.0,
      "step": 3867
    },
    {
      "entropy": 0.7681563328951597,
      "epoch": 0.35588064473464787,
      "grad_norm": 0.152495339512825,
      "learning_rate": 8.814058331033213e-05,
      "loss": 0.7655,
      "mean_token_accuracy": 0.7797845304012299,
      "num_tokens": 123447705.0,
      "step": 3868
    },
    {
      "entropy": 0.98968605697155,
      "epoch": 0.3559726511060891,
      "grad_norm": 0.17021581530570984,
      "learning_rate": 8.813751648419051e-05,
      "loss": 0.9859,
      "mean_token_accuracy": 0.7240044362843037,
      "num_tokens": 123479552.0,
      "step": 3869
    },
    {
      "entropy": 0.8605850487947464,
      "epoch": 0.3560646574775303,
      "grad_norm": 0.16424888372421265,
      "learning_rate": 8.81344496580489e-05,
      "loss": 0.8594,
      "mean_token_accuracy": 0.7494294606149197,
      "num_tokens": 123511020.0,
      "step": 3870
    },
    {
      "entropy": 0.7652826737612486,
      "epoch": 0.35615666384897154,
      "grad_norm": 0.14982885122299194,
      "learning_rate": 8.813138283190726e-05,
      "loss": 0.7399,
      "mean_token_accuracy": 0.7799660861492157,
      "num_tokens": 123543271.0,
      "step": 3871
    },
    {
      "entropy": 0.780370470136404,
      "epoch": 0.3562486702204128,
      "grad_norm": 0.14901946485042572,
      "learning_rate": 8.812831600576563e-05,
      "loss": 0.7684,
      "mean_token_accuracy": 0.7744050212204456,
      "num_tokens": 123575396.0,
      "step": 3872
    },
    {
      "entropy": 0.7886612601578236,
      "epoch": 0.356340676591854,
      "grad_norm": 0.1585397571325302,
      "learning_rate": 8.812524917962401e-05,
      "loss": 0.8154,
      "mean_token_accuracy": 0.7665118426084518,
      "num_tokens": 123606221.0,
      "step": 3873
    },
    {
      "entropy": 0.8079205118119717,
      "epoch": 0.3564326829632952,
      "grad_norm": 0.15542809665203094,
      "learning_rate": 8.812218235348238e-05,
      "loss": 0.8041,
      "mean_token_accuracy": 0.7665212340652943,
      "num_tokens": 123637514.0,
      "step": 3874
    },
    {
      "entropy": 0.8762025497853756,
      "epoch": 0.3565246893347364,
      "grad_norm": 0.15812699496746063,
      "learning_rate": 8.811911552734076e-05,
      "loss": 0.8867,
      "mean_token_accuracy": 0.7449298650026321,
      "num_tokens": 123669324.0,
      "step": 3875
    },
    {
      "entropy": 0.7565794810652733,
      "epoch": 0.35661669570617766,
      "grad_norm": 0.15625078976154327,
      "learning_rate": 8.811604870119914e-05,
      "loss": 0.7571,
      "mean_token_accuracy": 0.7785244695842266,
      "num_tokens": 123700850.0,
      "step": 3876
    },
    {
      "entropy": 0.8177671786397696,
      "epoch": 0.3567087020776189,
      "grad_norm": 0.1610877960920334,
      "learning_rate": 8.811298187505751e-05,
      "loss": 0.8147,
      "mean_token_accuracy": 0.7650122717022896,
      "num_tokens": 123732610.0,
      "step": 3877
    },
    {
      "entropy": 0.8457952737808228,
      "epoch": 0.3568007084490601,
      "grad_norm": 0.15711858868598938,
      "learning_rate": 8.810991504891588e-05,
      "loss": 0.829,
      "mean_token_accuracy": 0.7612704336643219,
      "num_tokens": 123764377.0,
      "step": 3878
    },
    {
      "entropy": 0.7525805532932281,
      "epoch": 0.35689271482050133,
      "grad_norm": 0.1481260061264038,
      "learning_rate": 8.810684822277425e-05,
      "loss": 0.7166,
      "mean_token_accuracy": 0.7934305816888809,
      "num_tokens": 123796200.0,
      "step": 3879
    },
    {
      "entropy": 0.8027665577828884,
      "epoch": 0.3569847211919425,
      "grad_norm": 0.1611669361591339,
      "learning_rate": 8.810378139663264e-05,
      "loss": 0.7852,
      "mean_token_accuracy": 0.7702967077493668,
      "num_tokens": 123827815.0,
      "step": 3880
    },
    {
      "entropy": 0.7924179658293724,
      "epoch": 0.35707672756338377,
      "grad_norm": 0.16211912035942078,
      "learning_rate": 8.810071457049101e-05,
      "loss": 0.7705,
      "mean_token_accuracy": 0.7757881432771683,
      "num_tokens": 123860338.0,
      "step": 3881
    },
    {
      "entropy": 0.8279090654104948,
      "epoch": 0.357168733934825,
      "grad_norm": 0.16137219965457916,
      "learning_rate": 8.809764774434938e-05,
      "loss": 0.8329,
      "mean_token_accuracy": 0.7599588297307491,
      "num_tokens": 123891404.0,
      "step": 3882
    },
    {
      "entropy": 0.7953283116221428,
      "epoch": 0.3572607403062662,
      "grad_norm": 0.15649986267089844,
      "learning_rate": 8.809458091820774e-05,
      "loss": 0.7787,
      "mean_token_accuracy": 0.7774722687900066,
      "num_tokens": 123923400.0,
      "step": 3883
    },
    {
      "entropy": 0.8056732006371021,
      "epoch": 0.35735274667770744,
      "grad_norm": 0.15557630360126495,
      "learning_rate": 8.809151409206613e-05,
      "loss": 0.791,
      "mean_token_accuracy": 0.7687728554010391,
      "num_tokens": 123955441.0,
      "step": 3884
    },
    {
      "entropy": 0.8117903433740139,
      "epoch": 0.35744475304914863,
      "grad_norm": 0.16056019067764282,
      "learning_rate": 8.80884472659245e-05,
      "loss": 0.7963,
      "mean_token_accuracy": 0.768803533166647,
      "num_tokens": 123986805.0,
      "step": 3885
    },
    {
      "entropy": 0.9249139949679375,
      "epoch": 0.3575367594205899,
      "grad_norm": 0.17148295044898987,
      "learning_rate": 8.808538043978287e-05,
      "loss": 0.9372,
      "mean_token_accuracy": 0.7280000746250153,
      "num_tokens": 124018612.0,
      "step": 3886
    },
    {
      "entropy": 0.7964979894459248,
      "epoch": 0.3576287657920311,
      "grad_norm": 0.16349273920059204,
      "learning_rate": 8.808231361364124e-05,
      "loss": 0.8081,
      "mean_token_accuracy": 0.7727683857083321,
      "num_tokens": 124051273.0,
      "step": 3887
    },
    {
      "entropy": 0.796128373593092,
      "epoch": 0.3577207721634723,
      "grad_norm": 0.15822678804397583,
      "learning_rate": 8.807924678749962e-05,
      "loss": 0.811,
      "mean_token_accuracy": 0.7639848366379738,
      "num_tokens": 124082442.0,
      "step": 3888
    },
    {
      "entropy": 0.7782258950173855,
      "epoch": 0.35781277853491356,
      "grad_norm": 0.15417996048927307,
      "learning_rate": 8.807617996135799e-05,
      "loss": 0.7559,
      "mean_token_accuracy": 0.7748013138771057,
      "num_tokens": 124114362.0,
      "step": 3889
    },
    {
      "entropy": 0.8669955376535654,
      "epoch": 0.35790478490635474,
      "grad_norm": 0.16038070619106293,
      "learning_rate": 8.807311313521637e-05,
      "loss": 0.855,
      "mean_token_accuracy": 0.7517752014100552,
      "num_tokens": 124146281.0,
      "step": 3890
    },
    {
      "entropy": 0.7577128857374191,
      "epoch": 0.357996791277796,
      "grad_norm": 0.17040887475013733,
      "learning_rate": 8.807004630907474e-05,
      "loss": 0.7998,
      "mean_token_accuracy": 0.7744840122759342,
      "num_tokens": 124177762.0,
      "step": 3891
    },
    {
      "entropy": 0.8198418319225311,
      "epoch": 0.35808879764923723,
      "grad_norm": 0.16359855234622955,
      "learning_rate": 8.806697948293312e-05,
      "loss": 0.8245,
      "mean_token_accuracy": 0.76328269764781,
      "num_tokens": 124210167.0,
      "step": 3892
    },
    {
      "entropy": 0.86313746124506,
      "epoch": 0.3581808040206784,
      "grad_norm": 0.1609724909067154,
      "learning_rate": 8.806391265679149e-05,
      "loss": 0.8447,
      "mean_token_accuracy": 0.7559772469103336,
      "num_tokens": 124242324.0,
      "step": 3893
    },
    {
      "entropy": 0.7952889911830425,
      "epoch": 0.35827281039211967,
      "grad_norm": 0.15515241026878357,
      "learning_rate": 8.806084583064986e-05,
      "loss": 0.8028,
      "mean_token_accuracy": 0.7700570374727249,
      "num_tokens": 124273940.0,
      "step": 3894
    },
    {
      "entropy": 0.940202422440052,
      "epoch": 0.35836481676356086,
      "grad_norm": 0.16045069694519043,
      "learning_rate": 8.805777900450824e-05,
      "loss": 0.9076,
      "mean_token_accuracy": 0.7395432963967323,
      "num_tokens": 124305789.0,
      "step": 3895
    },
    {
      "entropy": 0.8102720454335213,
      "epoch": 0.3584568231350021,
      "grad_norm": 0.1565655618906021,
      "learning_rate": 8.805471217836662e-05,
      "loss": 0.7943,
      "mean_token_accuracy": 0.7699993923306465,
      "num_tokens": 124337794.0,
      "step": 3896
    },
    {
      "entropy": 0.7866607028990984,
      "epoch": 0.35854882950644335,
      "grad_norm": 0.151796355843544,
      "learning_rate": 8.805164535222499e-05,
      "loss": 0.7759,
      "mean_token_accuracy": 0.7778375521302223,
      "num_tokens": 124370175.0,
      "step": 3897
    },
    {
      "entropy": 0.801344882696867,
      "epoch": 0.35864083587788453,
      "grad_norm": 0.15756016969680786,
      "learning_rate": 8.804857852608336e-05,
      "loss": 0.7836,
      "mean_token_accuracy": 0.7658494152128696,
      "num_tokens": 124401377.0,
      "step": 3898
    },
    {
      "entropy": 0.8224347587674856,
      "epoch": 0.3587328422493258,
      "grad_norm": 0.15886300802230835,
      "learning_rate": 8.804551169994172e-05,
      "loss": 0.8198,
      "mean_token_accuracy": 0.7680549398064613,
      "num_tokens": 124433353.0,
      "step": 3899
    },
    {
      "entropy": 0.8984981868416071,
      "epoch": 0.35882484862076697,
      "grad_norm": 0.15991607308387756,
      "learning_rate": 8.80424448738001e-05,
      "loss": 0.8956,
      "mean_token_accuracy": 0.7433252744376659,
      "num_tokens": 124465256.0,
      "step": 3900
    },
    {
      "entropy": 0.7627375181764364,
      "epoch": 0.3589168549922082,
      "grad_norm": 0.15654291212558746,
      "learning_rate": 8.803937804765849e-05,
      "loss": 0.7695,
      "mean_token_accuracy": 0.7747605852782726,
      "num_tokens": 124497419.0,
      "step": 3901
    },
    {
      "entropy": 0.7430278845131397,
      "epoch": 0.35900886136364946,
      "grad_norm": 0.15730933845043182,
      "learning_rate": 8.803631122151686e-05,
      "loss": 0.755,
      "mean_token_accuracy": 0.78126260638237,
      "num_tokens": 124528769.0,
      "step": 3902
    },
    {
      "entropy": 0.8074232116341591,
      "epoch": 0.35910086773509065,
      "grad_norm": 0.15854749083518982,
      "learning_rate": 8.803324439537522e-05,
      "loss": 0.8018,
      "mean_token_accuracy": 0.7708847932517529,
      "num_tokens": 124561307.0,
      "step": 3903
    },
    {
      "entropy": 0.8033623434603214,
      "epoch": 0.3591928741065319,
      "grad_norm": 0.15735717117786407,
      "learning_rate": 8.80301775692336e-05,
      "loss": 0.8017,
      "mean_token_accuracy": 0.7675640843808651,
      "num_tokens": 124592157.0,
      "step": 3904
    },
    {
      "entropy": 0.8066165074706078,
      "epoch": 0.3592848804779731,
      "grad_norm": 0.16374117136001587,
      "learning_rate": 8.802711074309197e-05,
      "loss": 0.8116,
      "mean_token_accuracy": 0.7675763368606567,
      "num_tokens": 124622971.0,
      "step": 3905
    },
    {
      "entropy": 0.9393394850194454,
      "epoch": 0.3593768868494143,
      "grad_norm": 0.16438381373882294,
      "learning_rate": 8.802404391695035e-05,
      "loss": 0.9208,
      "mean_token_accuracy": 0.7313060648739338,
      "num_tokens": 124653666.0,
      "step": 3906
    },
    {
      "entropy": 0.7653329968452454,
      "epoch": 0.35946889322085557,
      "grad_norm": 0.16432248055934906,
      "learning_rate": 8.802097709080874e-05,
      "loss": 0.7749,
      "mean_token_accuracy": 0.7779832072556019,
      "num_tokens": 124685469.0,
      "step": 3907
    },
    {
      "entropy": 0.8396038506180048,
      "epoch": 0.35956089959229676,
      "grad_norm": 0.16249480843544006,
      "learning_rate": 8.80179102646671e-05,
      "loss": 0.8398,
      "mean_token_accuracy": 0.7613067664206028,
      "num_tokens": 124717883.0,
      "step": 3908
    },
    {
      "entropy": 0.7555494364351034,
      "epoch": 0.359652905963738,
      "grad_norm": 0.16898347437381744,
      "learning_rate": 8.801484343852547e-05,
      "loss": 0.7368,
      "mean_token_accuracy": 0.788006916642189,
      "num_tokens": 124750111.0,
      "step": 3909
    },
    {
      "entropy": 0.8156967554241419,
      "epoch": 0.3597449123351792,
      "grad_norm": 0.15486498177051544,
      "learning_rate": 8.801177661238384e-05,
      "loss": 0.8174,
      "mean_token_accuracy": 0.7583503127098083,
      "num_tokens": 124782306.0,
      "step": 3910
    },
    {
      "entropy": 0.9148455001413822,
      "epoch": 0.35983691870662043,
      "grad_norm": 0.16240434348583221,
      "learning_rate": 8.800870978624222e-05,
      "loss": 0.9198,
      "mean_token_accuracy": 0.7346030212938786,
      "num_tokens": 124814314.0,
      "step": 3911
    },
    {
      "entropy": 0.9388261139392853,
      "epoch": 0.3599289250780617,
      "grad_norm": 0.16559064388275146,
      "learning_rate": 8.80056429601006e-05,
      "loss": 0.9354,
      "mean_token_accuracy": 0.7375219129025936,
      "num_tokens": 124846811.0,
      "step": 3912
    },
    {
      "entropy": 0.7667678259313107,
      "epoch": 0.36002093144950287,
      "grad_norm": 0.15229028463363647,
      "learning_rate": 8.800257613395897e-05,
      "loss": 0.7445,
      "mean_token_accuracy": 0.7842138111591339,
      "num_tokens": 124879173.0,
      "step": 3913
    },
    {
      "entropy": 0.8321407083421946,
      "epoch": 0.3601129378209441,
      "grad_norm": 0.15666262805461884,
      "learning_rate": 8.799950930781734e-05,
      "loss": 0.8269,
      "mean_token_accuracy": 0.761312622576952,
      "num_tokens": 124910689.0,
      "step": 3914
    },
    {
      "entropy": 0.780284620821476,
      "epoch": 0.3602049441923853,
      "grad_norm": 0.1554214209318161,
      "learning_rate": 8.799644248167572e-05,
      "loss": 0.7777,
      "mean_token_accuracy": 0.7717700116336346,
      "num_tokens": 124942732.0,
      "step": 3915
    },
    {
      "entropy": 0.8708509281277657,
      "epoch": 0.36029695056382655,
      "grad_norm": 0.1620628982782364,
      "learning_rate": 8.79933756555341e-05,
      "loss": 0.8584,
      "mean_token_accuracy": 0.754249669611454,
      "num_tokens": 124974619.0,
      "step": 3916
    },
    {
      "entropy": 0.7729030977934599,
      "epoch": 0.3603889569352678,
      "grad_norm": 0.15657636523246765,
      "learning_rate": 8.799030882939247e-05,
      "loss": 0.7458,
      "mean_token_accuracy": 0.7769762426614761,
      "num_tokens": 125006406.0,
      "step": 3917
    },
    {
      "entropy": 0.8477798067033291,
      "epoch": 0.360480963306709,
      "grad_norm": 0.16162627935409546,
      "learning_rate": 8.798724200325084e-05,
      "loss": 0.8458,
      "mean_token_accuracy": 0.7513680122792721,
      "num_tokens": 125038542.0,
      "step": 3918
    },
    {
      "entropy": 0.7371966652572155,
      "epoch": 0.3605729696781502,
      "grad_norm": 0.1564147174358368,
      "learning_rate": 8.798417517710922e-05,
      "loss": 0.7367,
      "mean_token_accuracy": 0.7833664864301682,
      "num_tokens": 125070421.0,
      "step": 3919
    },
    {
      "entropy": 0.8346356470137835,
      "epoch": 0.3606649760495914,
      "grad_norm": 0.1642010509967804,
      "learning_rate": 8.798110835096759e-05,
      "loss": 0.8358,
      "mean_token_accuracy": 0.7585852891206741,
      "num_tokens": 125102183.0,
      "step": 3920
    },
    {
      "entropy": 0.8888987563550472,
      "epoch": 0.36075698242103266,
      "grad_norm": 0.16706115007400513,
      "learning_rate": 8.797804152482597e-05,
      "loss": 0.8763,
      "mean_token_accuracy": 0.7465973943471909,
      "num_tokens": 125134094.0,
      "step": 3921
    },
    {
      "entropy": 0.9197887722402811,
      "epoch": 0.3608489887924739,
      "grad_norm": 0.1648307889699936,
      "learning_rate": 8.797497469868433e-05,
      "loss": 0.9441,
      "mean_token_accuracy": 0.7328496798872948,
      "num_tokens": 125166251.0,
      "step": 3922
    },
    {
      "entropy": 0.8836282007396221,
      "epoch": 0.3609409951639151,
      "grad_norm": 0.1686035692691803,
      "learning_rate": 8.797190787254272e-05,
      "loss": 0.9052,
      "mean_token_accuracy": 0.7436863221228123,
      "num_tokens": 125198335.0,
      "step": 3923
    },
    {
      "entropy": 0.8349615186452866,
      "epoch": 0.36103300153535633,
      "grad_norm": 0.16657571494579315,
      "learning_rate": 8.796884104640108e-05,
      "loss": 0.8446,
      "mean_token_accuracy": 0.7593650966882706,
      "num_tokens": 125230139.0,
      "step": 3924
    },
    {
      "entropy": 0.8277018461376429,
      "epoch": 0.3611250079067975,
      "grad_norm": 0.16189897060394287,
      "learning_rate": 8.796577422025945e-05,
      "loss": 0.8236,
      "mean_token_accuracy": 0.7625271864235401,
      "num_tokens": 125262768.0,
      "step": 3925
    },
    {
      "entropy": 0.8756210319697857,
      "epoch": 0.36121701427823877,
      "grad_norm": 0.15879683196544647,
      "learning_rate": 8.796270739411783e-05,
      "loss": 0.8579,
      "mean_token_accuracy": 0.7462161593139172,
      "num_tokens": 125294643.0,
      "step": 3926
    },
    {
      "entropy": 0.9656876921653748,
      "epoch": 0.36130902064968,
      "grad_norm": 0.16486388444900513,
      "learning_rate": 8.795964056797621e-05,
      "loss": 0.9699,
      "mean_token_accuracy": 0.7253120057284832,
      "num_tokens": 125326840.0,
      "step": 3927
    },
    {
      "entropy": 0.7841836679726839,
      "epoch": 0.3614010270211212,
      "grad_norm": 0.15575851500034332,
      "learning_rate": 8.795657374183458e-05,
      "loss": 0.7454,
      "mean_token_accuracy": 0.7858451344072819,
      "num_tokens": 125359608.0,
      "step": 3928
    },
    {
      "entropy": 0.8185913413763046,
      "epoch": 0.36149303339256245,
      "grad_norm": 0.15726853907108307,
      "learning_rate": 8.795350691569295e-05,
      "loss": 0.8027,
      "mean_token_accuracy": 0.768819946795702,
      "num_tokens": 125391571.0,
      "step": 3929
    },
    {
      "entropy": 0.7729168515652418,
      "epoch": 0.36158503976400364,
      "grad_norm": 0.1511806845664978,
      "learning_rate": 8.795044008955132e-05,
      "loss": 0.7349,
      "mean_token_accuracy": 0.7809696346521378,
      "num_tokens": 125423060.0,
      "step": 3930
    },
    {
      "entropy": 0.8200077526271343,
      "epoch": 0.3616770461354449,
      "grad_norm": 0.1585727035999298,
      "learning_rate": 8.79473732634097e-05,
      "loss": 0.8081,
      "mean_token_accuracy": 0.7627638094127178,
      "num_tokens": 125454612.0,
      "step": 3931
    },
    {
      "entropy": 0.7184287458658218,
      "epoch": 0.3617690525068861,
      "grad_norm": 0.14674459397792816,
      "learning_rate": 8.794430643726808e-05,
      "loss": 0.702,
      "mean_token_accuracy": 0.7920012101531029,
      "num_tokens": 125486775.0,
      "step": 3932
    },
    {
      "entropy": 0.8477559015154839,
      "epoch": 0.3618610588783273,
      "grad_norm": 0.16182146966457367,
      "learning_rate": 8.794123961112645e-05,
      "loss": 0.8177,
      "mean_token_accuracy": 0.7663889527320862,
      "num_tokens": 125518723.0,
      "step": 3933
    },
    {
      "entropy": 0.8465719986706972,
      "epoch": 0.36195306524976856,
      "grad_norm": 0.16571323573589325,
      "learning_rate": 8.793817278498482e-05,
      "loss": 0.8556,
      "mean_token_accuracy": 0.7529676929116249,
      "num_tokens": 125549456.0,
      "step": 3934
    },
    {
      "entropy": 0.7466309480369091,
      "epoch": 0.36204507162120975,
      "grad_norm": 0.150594100356102,
      "learning_rate": 8.79351059588432e-05,
      "loss": 0.7284,
      "mean_token_accuracy": 0.7877485118806362,
      "num_tokens": 125582143.0,
      "step": 3935
    },
    {
      "entropy": 0.8676666375249624,
      "epoch": 0.362137077992651,
      "grad_norm": 0.16373896598815918,
      "learning_rate": 8.793203913270157e-05,
      "loss": 0.8734,
      "mean_token_accuracy": 0.7454674988985062,
      "num_tokens": 125613750.0,
      "step": 3936
    },
    {
      "entropy": 0.9189652651548386,
      "epoch": 0.36222908436409224,
      "grad_norm": 0.16698110103607178,
      "learning_rate": 8.792897230655995e-05,
      "loss": 0.9151,
      "mean_token_accuracy": 0.7379536181688309,
      "num_tokens": 125645553.0,
      "step": 3937
    },
    {
      "entropy": 0.7178536970168352,
      "epoch": 0.3623210907355334,
      "grad_norm": 0.167508065700531,
      "learning_rate": 8.792590548041833e-05,
      "loss": 0.7409,
      "mean_token_accuracy": 0.7878397069871426,
      "num_tokens": 125677254.0,
      "step": 3938
    },
    {
      "entropy": 0.8730999324470758,
      "epoch": 0.36241309710697467,
      "grad_norm": 0.16966745257377625,
      "learning_rate": 8.79228386542767e-05,
      "loss": 0.8889,
      "mean_token_accuracy": 0.7526889853179455,
      "num_tokens": 125708894.0,
      "step": 3939
    },
    {
      "entropy": 0.8305307254195213,
      "epoch": 0.36250510347841586,
      "grad_norm": 0.159025639295578,
      "learning_rate": 8.791977182813506e-05,
      "loss": 0.81,
      "mean_token_accuracy": 0.769099909812212,
      "num_tokens": 125740978.0,
      "step": 3940
    },
    {
      "entropy": 0.8325799237936735,
      "epoch": 0.3625971098498571,
      "grad_norm": 0.16129015386104584,
      "learning_rate": 8.791670500199343e-05,
      "loss": 0.8121,
      "mean_token_accuracy": 0.7699800878763199,
      "num_tokens": 125773383.0,
      "step": 3941
    },
    {
      "entropy": 0.9553064852952957,
      "epoch": 0.36268911622129835,
      "grad_norm": 0.17063377797603607,
      "learning_rate": 8.791363817585181e-05,
      "loss": 0.9643,
      "mean_token_accuracy": 0.7318258956074715,
      "num_tokens": 125804558.0,
      "step": 3942
    },
    {
      "entropy": 0.8225627057254314,
      "epoch": 0.36278112259273954,
      "grad_norm": 0.1564221829175949,
      "learning_rate": 8.79105713497102e-05,
      "loss": 0.8056,
      "mean_token_accuracy": 0.7677732706069946,
      "num_tokens": 125836794.0,
      "step": 3943
    },
    {
      "entropy": 0.7901673689484596,
      "epoch": 0.3628731289641808,
      "grad_norm": 0.15048852562904358,
      "learning_rate": 8.790750452356856e-05,
      "loss": 0.7853,
      "mean_token_accuracy": 0.7678054869174957,
      "num_tokens": 125868898.0,
      "step": 3944
    },
    {
      "entropy": 0.8572925068438053,
      "epoch": 0.36296513533562197,
      "grad_norm": 0.16412808001041412,
      "learning_rate": 8.790443769742693e-05,
      "loss": 0.8439,
      "mean_token_accuracy": 0.757486741989851,
      "num_tokens": 125900491.0,
      "step": 3945
    },
    {
      "entropy": 0.7537161912769079,
      "epoch": 0.3630571417070632,
      "grad_norm": 0.15849816799163818,
      "learning_rate": 8.790137087128531e-05,
      "loss": 0.7301,
      "mean_token_accuracy": 0.7860838137567043,
      "num_tokens": 125932702.0,
      "step": 3946
    },
    {
      "entropy": 0.9456305727362633,
      "epoch": 0.36314914807850446,
      "grad_norm": 0.1698368936777115,
      "learning_rate": 8.789830404514368e-05,
      "loss": 0.9381,
      "mean_token_accuracy": 0.7337273880839348,
      "num_tokens": 125965020.0,
      "step": 3947
    },
    {
      "entropy": 0.8197853267192841,
      "epoch": 0.36324115444994565,
      "grad_norm": 0.16335085034370422,
      "learning_rate": 8.789523721900206e-05,
      "loss": 0.8146,
      "mean_token_accuracy": 0.759614210575819,
      "num_tokens": 125997344.0,
      "step": 3948
    },
    {
      "entropy": 0.8169116992503405,
      "epoch": 0.3633331608213869,
      "grad_norm": 0.16037850081920624,
      "learning_rate": 8.789217039286043e-05,
      "loss": 0.7995,
      "mean_token_accuracy": 0.7629402875900269,
      "num_tokens": 126028927.0,
      "step": 3949
    },
    {
      "entropy": 0.8202864974737167,
      "epoch": 0.3634251671928281,
      "grad_norm": 0.15603046119213104,
      "learning_rate": 8.788910356671881e-05,
      "loss": 0.7842,
      "mean_token_accuracy": 0.773261521011591,
      "num_tokens": 126061370.0,
      "step": 3950
    },
    {
      "entropy": 0.8240732736885548,
      "epoch": 0.3635171735642693,
      "grad_norm": 0.1645958572626114,
      "learning_rate": 8.788603674057718e-05,
      "loss": 0.8335,
      "mean_token_accuracy": 0.7601423002779484,
      "num_tokens": 126092818.0,
      "step": 3951
    },
    {
      "entropy": 0.8019191082566977,
      "epoch": 0.36360917993571057,
      "grad_norm": 0.16038531064987183,
      "learning_rate": 8.788296991443555e-05,
      "loss": 0.7995,
      "mean_token_accuracy": 0.7665547505021095,
      "num_tokens": 126124289.0,
      "step": 3952
    },
    {
      "entropy": 0.7882400881499052,
      "epoch": 0.36370118630715176,
      "grad_norm": 0.16171714663505554,
      "learning_rate": 8.787990308829393e-05,
      "loss": 0.783,
      "mean_token_accuracy": 0.7690963633358479,
      "num_tokens": 126155923.0,
      "step": 3953
    },
    {
      "entropy": 0.8058025501668453,
      "epoch": 0.363793192678593,
      "grad_norm": 0.16676580905914307,
      "learning_rate": 8.787683626215231e-05,
      "loss": 0.808,
      "mean_token_accuracy": 0.7621057629585266,
      "num_tokens": 126187428.0,
      "step": 3954
    },
    {
      "entropy": 0.8515272736549377,
      "epoch": 0.3638851990500342,
      "grad_norm": 0.15875963866710663,
      "learning_rate": 8.787376943601068e-05,
      "loss": 0.8405,
      "mean_token_accuracy": 0.7587184645235538,
      "num_tokens": 126218977.0,
      "step": 3955
    },
    {
      "entropy": 0.8993930593132973,
      "epoch": 0.36397720542147544,
      "grad_norm": 0.16499730944633484,
      "learning_rate": 8.787070260986905e-05,
      "loss": 0.9121,
      "mean_token_accuracy": 0.7398312650620937,
      "num_tokens": 126250850.0,
      "step": 3956
    },
    {
      "entropy": 0.7586094569414854,
      "epoch": 0.3640692117929167,
      "grad_norm": 0.15172486007213593,
      "learning_rate": 8.786763578372741e-05,
      "loss": 0.7267,
      "mean_token_accuracy": 0.7871798798441887,
      "num_tokens": 126282869.0,
      "step": 3957
    },
    {
      "entropy": 0.7427377291023731,
      "epoch": 0.36416121816435787,
      "grad_norm": 0.14413316547870636,
      "learning_rate": 8.786456895758581e-05,
      "loss": 0.7007,
      "mean_token_accuracy": 0.7948445379734039,
      "num_tokens": 126315548.0,
      "step": 3958
    },
    {
      "entropy": 0.8012926578521729,
      "epoch": 0.3642532245357991,
      "grad_norm": 0.15583916008472443,
      "learning_rate": 8.786150213144418e-05,
      "loss": 0.7905,
      "mean_token_accuracy": 0.7709271162748337,
      "num_tokens": 126347545.0,
      "step": 3959
    },
    {
      "entropy": 0.8470191173255444,
      "epoch": 0.3643452309072403,
      "grad_norm": 0.16612376272678375,
      "learning_rate": 8.785843530530254e-05,
      "loss": 0.8501,
      "mean_token_accuracy": 0.755960077047348,
      "num_tokens": 126379241.0,
      "step": 3960
    },
    {
      "entropy": 0.7395764514803886,
      "epoch": 0.36443723727868155,
      "grad_norm": 0.15470241010189056,
      "learning_rate": 8.785536847916091e-05,
      "loss": 0.7253,
      "mean_token_accuracy": 0.7868499495089054,
      "num_tokens": 126411472.0,
      "step": 3961
    },
    {
      "entropy": 0.809343047440052,
      "epoch": 0.3645292436501228,
      "grad_norm": 0.15722091495990753,
      "learning_rate": 8.78523016530193e-05,
      "loss": 0.8171,
      "mean_token_accuracy": 0.7633778229355812,
      "num_tokens": 126443427.0,
      "step": 3962
    },
    {
      "entropy": 0.8436812609434128,
      "epoch": 0.364621250021564,
      "grad_norm": 0.16498376429080963,
      "learning_rate": 8.784923482687768e-05,
      "loss": 0.8499,
      "mean_token_accuracy": 0.7517563626170158,
      "num_tokens": 126474651.0,
      "step": 3963
    },
    {
      "entropy": 0.9015608504414558,
      "epoch": 0.3647132563930052,
      "grad_norm": 0.17126184701919556,
      "learning_rate": 8.784616800073604e-05,
      "loss": 0.9344,
      "mean_token_accuracy": 0.7349385879933834,
      "num_tokens": 126506396.0,
      "step": 3964
    },
    {
      "entropy": 0.7757081352174282,
      "epoch": 0.3648052627644464,
      "grad_norm": 0.1657913327217102,
      "learning_rate": 8.784310117459441e-05,
      "loss": 0.7953,
      "mean_token_accuracy": 0.7697840556502342,
      "num_tokens": 126538769.0,
      "step": 3965
    },
    {
      "entropy": 0.7408387679606676,
      "epoch": 0.36489726913588766,
      "grad_norm": 0.15644589066505432,
      "learning_rate": 8.784003434845279e-05,
      "loss": 0.76,
      "mean_token_accuracy": 0.7812956683337688,
      "num_tokens": 126570321.0,
      "step": 3966
    },
    {
      "entropy": 0.823700338602066,
      "epoch": 0.3649892755073289,
      "grad_norm": 0.16799326241016388,
      "learning_rate": 8.783696752231116e-05,
      "loss": 0.8351,
      "mean_token_accuracy": 0.7561098225414753,
      "num_tokens": 126601834.0,
      "step": 3967
    },
    {
      "entropy": 0.7160749975591898,
      "epoch": 0.3650812818787701,
      "grad_norm": 0.15595506131649017,
      "learning_rate": 8.783390069616954e-05,
      "loss": 0.7048,
      "mean_token_accuracy": 0.7934368662536144,
      "num_tokens": 126634090.0,
      "step": 3968
    },
    {
      "entropy": 0.8422998264431953,
      "epoch": 0.36517328825021134,
      "grad_norm": 0.15621601045131683,
      "learning_rate": 8.783083387002792e-05,
      "loss": 0.8249,
      "mean_token_accuracy": 0.7626119032502174,
      "num_tokens": 126665980.0,
      "step": 3969
    },
    {
      "entropy": 0.878008397296071,
      "epoch": 0.3652652946216525,
      "grad_norm": 0.16343845427036285,
      "learning_rate": 8.782776704388629e-05,
      "loss": 0.8551,
      "mean_token_accuracy": 0.7508497163653374,
      "num_tokens": 126697776.0,
      "step": 3970
    },
    {
      "entropy": 0.9104081317782402,
      "epoch": 0.36535730099309377,
      "grad_norm": 0.16033856570720673,
      "learning_rate": 8.782470021774466e-05,
      "loss": 0.8971,
      "mean_token_accuracy": 0.7445614337921143,
      "num_tokens": 126729802.0,
      "step": 3971
    },
    {
      "entropy": 0.8272992968559265,
      "epoch": 0.365449307364535,
      "grad_norm": 0.155325785279274,
      "learning_rate": 8.782163339160303e-05,
      "loss": 0.7941,
      "mean_token_accuracy": 0.7666210010647774,
      "num_tokens": 126762440.0,
      "step": 3972
    },
    {
      "entropy": 0.7247307449579239,
      "epoch": 0.3655413137359762,
      "grad_norm": 0.1501549780368805,
      "learning_rate": 8.781856656546141e-05,
      "loss": 0.696,
      "mean_token_accuracy": 0.7925325185060501,
      "num_tokens": 126794861.0,
      "step": 3973
    },
    {
      "entropy": 0.7979334183037281,
      "epoch": 0.36563332010741745,
      "grad_norm": 0.15051423013210297,
      "learning_rate": 8.781549973931979e-05,
      "loss": 0.7793,
      "mean_token_accuracy": 0.7690415419638157,
      "num_tokens": 126826604.0,
      "step": 3974
    },
    {
      "entropy": 0.7382467985153198,
      "epoch": 0.36572532647885864,
      "grad_norm": 0.15750713646411896,
      "learning_rate": 8.781243291317816e-05,
      "loss": 0.7169,
      "mean_token_accuracy": 0.7872474640607834,
      "num_tokens": 126858529.0,
      "step": 3975
    },
    {
      "entropy": 0.7818368002772331,
      "epoch": 0.3658173328502999,
      "grad_norm": 0.16557520627975464,
      "learning_rate": 8.780936608703653e-05,
      "loss": 0.8038,
      "mean_token_accuracy": 0.7699749544262886,
      "num_tokens": 126890646.0,
      "step": 3976
    },
    {
      "entropy": 0.8099629506468773,
      "epoch": 0.3659093392217411,
      "grad_norm": 0.15552173554897308,
      "learning_rate": 8.78062992608949e-05,
      "loss": 0.808,
      "mean_token_accuracy": 0.764776561409235,
      "num_tokens": 126922944.0,
      "step": 3977
    },
    {
      "entropy": 0.8389763571321964,
      "epoch": 0.3660013455931823,
      "grad_norm": 0.15744782984256744,
      "learning_rate": 8.780323243475327e-05,
      "loss": 0.8621,
      "mean_token_accuracy": 0.7532774657011032,
      "num_tokens": 126954828.0,
      "step": 3978
    },
    {
      "entropy": 0.8193764463067055,
      "epoch": 0.36609335196462356,
      "grad_norm": 0.15994690358638763,
      "learning_rate": 8.780016560861166e-05,
      "loss": 0.8192,
      "mean_token_accuracy": 0.7667301334440708,
      "num_tokens": 126986460.0,
      "step": 3979
    },
    {
      "entropy": 0.8306416347622871,
      "epoch": 0.36618535833606475,
      "grad_norm": 0.16626794636249542,
      "learning_rate": 8.779709878247002e-05,
      "loss": 0.8176,
      "mean_token_accuracy": 0.7676108665764332,
      "num_tokens": 127018125.0,
      "step": 3980
    },
    {
      "entropy": 0.718898693099618,
      "epoch": 0.366277364707506,
      "grad_norm": 0.15648889541625977,
      "learning_rate": 8.77940319563284e-05,
      "loss": 0.7225,
      "mean_token_accuracy": 0.7873343080282211,
      "num_tokens": 127050600.0,
      "step": 3981
    },
    {
      "entropy": 0.869388910010457,
      "epoch": 0.36636937107894724,
      "grad_norm": 0.16109953820705414,
      "learning_rate": 8.779096513018677e-05,
      "loss": 0.8832,
      "mean_token_accuracy": 0.7451267875730991,
      "num_tokens": 127082711.0,
      "step": 3982
    },
    {
      "entropy": 0.8967003263533115,
      "epoch": 0.3664613774503884,
      "grad_norm": 0.16389314830303192,
      "learning_rate": 8.778789830404514e-05,
      "loss": 0.9057,
      "mean_token_accuracy": 0.7395916543900967,
      "num_tokens": 127114932.0,
      "step": 3983
    },
    {
      "entropy": 0.8233585972338915,
      "epoch": 0.36655338382182967,
      "grad_norm": 0.15947499871253967,
      "learning_rate": 8.778483147790352e-05,
      "loss": 0.8254,
      "mean_token_accuracy": 0.7592307440936565,
      "num_tokens": 127146555.0,
      "step": 3984
    },
    {
      "entropy": 0.7994773034006357,
      "epoch": 0.36664539019327086,
      "grad_norm": 0.15546394884586334,
      "learning_rate": 8.77817646517619e-05,
      "loss": 0.7905,
      "mean_token_accuracy": 0.7716978900134563,
      "num_tokens": 127178469.0,
      "step": 3985
    },
    {
      "entropy": 0.9335198998451233,
      "epoch": 0.3667373965647121,
      "grad_norm": 0.17434845864772797,
      "learning_rate": 8.777869782562027e-05,
      "loss": 0.9371,
      "mean_token_accuracy": 0.7312613315880299,
      "num_tokens": 127210279.0,
      "step": 3986
    },
    {
      "entropy": 0.7934442572295666,
      "epoch": 0.36682940293615335,
      "grad_norm": 0.15891015529632568,
      "learning_rate": 8.777563099947864e-05,
      "loss": 0.7809,
      "mean_token_accuracy": 0.7739532068371773,
      "num_tokens": 127242107.0,
      "step": 3987
    },
    {
      "entropy": 0.7143493182957172,
      "epoch": 0.36692140930759454,
      "grad_norm": 0.15759994089603424,
      "learning_rate": 8.777256417333701e-05,
      "loss": 0.7157,
      "mean_token_accuracy": 0.7911524772644043,
      "num_tokens": 127274641.0,
      "step": 3988
    },
    {
      "entropy": 0.9847908318042755,
      "epoch": 0.3670134156790358,
      "grad_norm": 0.16522352397441864,
      "learning_rate": 8.776949734719539e-05,
      "loss": 0.9801,
      "mean_token_accuracy": 0.7210739701986313,
      "num_tokens": 127306222.0,
      "step": 3989
    },
    {
      "entropy": 0.9107610248029232,
      "epoch": 0.36710542205047697,
      "grad_norm": 0.1608375608921051,
      "learning_rate": 8.776643052105377e-05,
      "loss": 0.8852,
      "mean_token_accuracy": 0.7426093928515911,
      "num_tokens": 127337701.0,
      "step": 3990
    },
    {
      "entropy": 0.7996920589357615,
      "epoch": 0.3671974284219182,
      "grad_norm": 0.15838110446929932,
      "learning_rate": 8.776336369491214e-05,
      "loss": 0.7875,
      "mean_token_accuracy": 0.7726570926606655,
      "num_tokens": 127369039.0,
      "step": 3991
    },
    {
      "entropy": 0.8674283102154732,
      "epoch": 0.36728943479335946,
      "grad_norm": 0.1606183499097824,
      "learning_rate": 8.77602968687705e-05,
      "loss": 0.8405,
      "mean_token_accuracy": 0.7546040713787079,
      "num_tokens": 127400843.0,
      "step": 3992
    },
    {
      "entropy": 0.8670284189283848,
      "epoch": 0.36738144116480065,
      "grad_norm": 0.15672333538532257,
      "learning_rate": 8.775723004262889e-05,
      "loss": 0.8523,
      "mean_token_accuracy": 0.7539434768259525,
      "num_tokens": 127432932.0,
      "step": 3993
    },
    {
      "entropy": 0.7564631029963493,
      "epoch": 0.3674734475362419,
      "grad_norm": 0.15088997781276703,
      "learning_rate": 8.775416321648726e-05,
      "loss": 0.7455,
      "mean_token_accuracy": 0.7808534018695354,
      "num_tokens": 127465000.0,
      "step": 3994
    },
    {
      "entropy": 0.8203088920563459,
      "epoch": 0.3675654539076831,
      "grad_norm": 0.16415101289749146,
      "learning_rate": 8.775109639034564e-05,
      "loss": 0.7948,
      "mean_token_accuracy": 0.7714958898723125,
      "num_tokens": 127497242.0,
      "step": 3995
    },
    {
      "entropy": 0.7931037582457066,
      "epoch": 0.3676574602791243,
      "grad_norm": 0.15285944938659668,
      "learning_rate": 8.7748029564204e-05,
      "loss": 0.7765,
      "mean_token_accuracy": 0.7756114676594734,
      "num_tokens": 127529318.0,
      "step": 3996
    },
    {
      "entropy": 0.8569008558988571,
      "epoch": 0.36774946665056557,
      "grad_norm": 0.16136768460273743,
      "learning_rate": 8.774496273806239e-05,
      "loss": 0.8375,
      "mean_token_accuracy": 0.7565362527966499,
      "num_tokens": 127561303.0,
      "step": 3997
    },
    {
      "entropy": 0.7986241858452559,
      "epoch": 0.36784147302200676,
      "grad_norm": 0.1587868183851242,
      "learning_rate": 8.774189591192075e-05,
      "loss": 0.7913,
      "mean_token_accuracy": 0.7689219899475574,
      "num_tokens": 127593097.0,
      "step": 3998
    },
    {
      "entropy": 0.7245506383478642,
      "epoch": 0.367933479393448,
      "grad_norm": 0.15149961411952972,
      "learning_rate": 8.773882908577912e-05,
      "loss": 0.7239,
      "mean_token_accuracy": 0.788293082267046,
      "num_tokens": 127625798.0,
      "step": 3999
    },
    {
      "entropy": 0.8654173016548157,
      "epoch": 0.3680254857648892,
      "grad_norm": 0.15878471732139587,
      "learning_rate": 8.773576225963752e-05,
      "loss": 0.8551,
      "mean_token_accuracy": 0.7552462667226791,
      "num_tokens": 127658194.0,
      "step": 4000
    },
    {
      "entropy": 0.7779717836529016,
      "epoch": 0.36811749213633044,
      "grad_norm": 0.1533113718032837,
      "learning_rate": 8.773269543349588e-05,
      "loss": 0.7801,
      "mean_token_accuracy": 0.7666694782674313,
      "num_tokens": 127690644.0,
      "step": 4001
    },
    {
      "entropy": 0.7450078669935465,
      "epoch": 0.3682094985077717,
      "grad_norm": 0.15874242782592773,
      "learning_rate": 8.772962860735425e-05,
      "loss": 0.7569,
      "mean_token_accuracy": 0.7733174934983253,
      "num_tokens": 127722383.0,
      "step": 4002
    },
    {
      "entropy": 0.6911590807139874,
      "epoch": 0.36830150487921287,
      "grad_norm": 0.15282335877418518,
      "learning_rate": 8.772656178121262e-05,
      "loss": 0.6841,
      "mean_token_accuracy": 0.7981752380728722,
      "num_tokens": 127753886.0,
      "step": 4003
    },
    {
      "entropy": 0.8482338935136795,
      "epoch": 0.3683935112506541,
      "grad_norm": 0.16023407876491547,
      "learning_rate": 8.7723494955071e-05,
      "loss": 0.8472,
      "mean_token_accuracy": 0.755078986287117,
      "num_tokens": 127785918.0,
      "step": 4004
    },
    {
      "entropy": 0.8032221049070358,
      "epoch": 0.3684855176220953,
      "grad_norm": 0.15738379955291748,
      "learning_rate": 8.772042812892938e-05,
      "loss": 0.8047,
      "mean_token_accuracy": 0.7723706886172295,
      "num_tokens": 127818366.0,
      "step": 4005
    },
    {
      "entropy": 0.8651530556380749,
      "epoch": 0.36857752399353655,
      "grad_norm": 0.15708884596824646,
      "learning_rate": 8.771736130278775e-05,
      "loss": 0.8632,
      "mean_token_accuracy": 0.7486366629600525,
      "num_tokens": 127850486.0,
      "step": 4006
    },
    {
      "entropy": 0.8502090573310852,
      "epoch": 0.3686695303649778,
      "grad_norm": 0.16579410433769226,
      "learning_rate": 8.771429447664612e-05,
      "loss": 0.8598,
      "mean_token_accuracy": 0.7554258592426777,
      "num_tokens": 127882393.0,
      "step": 4007
    },
    {
      "entropy": 0.7623420301824808,
      "epoch": 0.368761536736419,
      "grad_norm": 0.15064069628715515,
      "learning_rate": 8.77112276505045e-05,
      "loss": 0.7571,
      "mean_token_accuracy": 0.7771819941699505,
      "num_tokens": 127914420.0,
      "step": 4008
    },
    {
      "entropy": 0.8391737416386604,
      "epoch": 0.3688535431078602,
      "grad_norm": 0.16334427893161774,
      "learning_rate": 8.770816082436287e-05,
      "loss": 0.8239,
      "mean_token_accuracy": 0.7609596252441406,
      "num_tokens": 127945892.0,
      "step": 4009
    },
    {
      "entropy": 0.8878744626417756,
      "epoch": 0.3689455494793014,
      "grad_norm": 0.1660335808992386,
      "learning_rate": 8.770509399822125e-05,
      "loss": 0.9009,
      "mean_token_accuracy": 0.7421878166496754,
      "num_tokens": 127977653.0,
      "step": 4010
    },
    {
      "entropy": 0.7658708207309246,
      "epoch": 0.36903755585074266,
      "grad_norm": 0.16228044033050537,
      "learning_rate": 8.770202717207962e-05,
      "loss": 0.7656,
      "mean_token_accuracy": 0.7790641784667969,
      "num_tokens": 128009022.0,
      "step": 4011
    },
    {
      "entropy": 0.7853386066854,
      "epoch": 0.3691295622221839,
      "grad_norm": 0.15931721031665802,
      "learning_rate": 8.7698960345938e-05,
      "loss": 0.7884,
      "mean_token_accuracy": 0.7741252481937408,
      "num_tokens": 128041616.0,
      "step": 4012
    },
    {
      "entropy": 0.7896102704107761,
      "epoch": 0.3692215685936251,
      "grad_norm": 0.15250766277313232,
      "learning_rate": 8.769589351979637e-05,
      "loss": 0.7671,
      "mean_token_accuracy": 0.7794551067054272,
      "num_tokens": 128074164.0,
      "step": 4013
    },
    {
      "entropy": 0.7469721138477325,
      "epoch": 0.36931357496506634,
      "grad_norm": 0.1566646695137024,
      "learning_rate": 8.769282669365473e-05,
      "loss": 0.7453,
      "mean_token_accuracy": 0.7817753031849861,
      "num_tokens": 128106079.0,
      "step": 4014
    },
    {
      "entropy": 0.8360635489225388,
      "epoch": 0.3694055813365075,
      "grad_norm": 0.16115272045135498,
      "learning_rate": 8.768975986751312e-05,
      "loss": 0.8414,
      "mean_token_accuracy": 0.7564277201890945,
      "num_tokens": 128137997.0,
      "step": 4015
    },
    {
      "entropy": 0.7539717853069305,
      "epoch": 0.36949758770794877,
      "grad_norm": 0.1532340794801712,
      "learning_rate": 8.76866930413715e-05,
      "loss": 0.7451,
      "mean_token_accuracy": 0.7813410125672817,
      "num_tokens": 128170254.0,
      "step": 4016
    },
    {
      "entropy": 0.9055490382015705,
      "epoch": 0.36958959407939,
      "grad_norm": 0.16207289695739746,
      "learning_rate": 8.768362621522987e-05,
      "loss": 0.8968,
      "mean_token_accuracy": 0.7437637560069561,
      "num_tokens": 128202485.0,
      "step": 4017
    },
    {
      "entropy": 0.8353424854576588,
      "epoch": 0.3696816004508312,
      "grad_norm": 0.149844691157341,
      "learning_rate": 8.768055938908823e-05,
      "loss": 0.8274,
      "mean_token_accuracy": 0.7595871984958649,
      "num_tokens": 128235253.0,
      "step": 4018
    },
    {
      "entropy": 0.8888925313949585,
      "epoch": 0.36977360682227245,
      "grad_norm": 0.15933437645435333,
      "learning_rate": 8.76774925629466e-05,
      "loss": 0.8864,
      "mean_token_accuracy": 0.7479917407035828,
      "num_tokens": 128267834.0,
      "step": 4019
    },
    {
      "entropy": 0.7563181798905134,
      "epoch": 0.36986561319371364,
      "grad_norm": 0.15307094156742096,
      "learning_rate": 8.767442573680498e-05,
      "loss": 0.7473,
      "mean_token_accuracy": 0.7790490761399269,
      "num_tokens": 128299732.0,
      "step": 4020
    },
    {
      "entropy": 0.8869237434118986,
      "epoch": 0.3699576195651549,
      "grad_norm": 0.15517838299274445,
      "learning_rate": 8.767135891066336e-05,
      "loss": 0.8707,
      "mean_token_accuracy": 0.748052142560482,
      "num_tokens": 128331732.0,
      "step": 4021
    },
    {
      "entropy": 0.8880502562969923,
      "epoch": 0.3700496259365961,
      "grad_norm": 0.15705333650112152,
      "learning_rate": 8.766829208452173e-05,
      "loss": 0.8742,
      "mean_token_accuracy": 0.7486115619540215,
      "num_tokens": 128363700.0,
      "step": 4022
    },
    {
      "entropy": 0.8855895027518272,
      "epoch": 0.3701416323080373,
      "grad_norm": 0.16336467862129211,
      "learning_rate": 8.76652252583801e-05,
      "loss": 0.8584,
      "mean_token_accuracy": 0.7511954307556152,
      "num_tokens": 128395430.0,
      "step": 4023
    },
    {
      "entropy": 0.8445280641317368,
      "epoch": 0.37023363867947856,
      "grad_norm": 0.1615702211856842,
      "learning_rate": 8.766215843223848e-05,
      "loss": 0.8474,
      "mean_token_accuracy": 0.7495435066521168,
      "num_tokens": 128427698.0,
      "step": 4024
    },
    {
      "entropy": 0.9193143676966429,
      "epoch": 0.37032564505091975,
      "grad_norm": 0.15725642442703247,
      "learning_rate": 8.765909160609685e-05,
      "loss": 0.8861,
      "mean_token_accuracy": 0.7464327961206436,
      "num_tokens": 128459496.0,
      "step": 4025
    },
    {
      "entropy": 0.8091272823512554,
      "epoch": 0.370417651422361,
      "grad_norm": 0.15597863495349884,
      "learning_rate": 8.765602477995523e-05,
      "loss": 0.7801,
      "mean_token_accuracy": 0.7759860940277576,
      "num_tokens": 128491819.0,
      "step": 4026
    },
    {
      "entropy": 0.8476017266511917,
      "epoch": 0.37050965779380224,
      "grad_norm": 0.15518225729465485,
      "learning_rate": 8.76529579538136e-05,
      "loss": 0.8391,
      "mean_token_accuracy": 0.7567649781703949,
      "num_tokens": 128523932.0,
      "step": 4027
    },
    {
      "entropy": 0.7648209035396576,
      "epoch": 0.3706016641652434,
      "grad_norm": 0.16384844481945038,
      "learning_rate": 8.764989112767198e-05,
      "loss": 0.7787,
      "mean_token_accuracy": 0.7793871238827705,
      "num_tokens": 128555479.0,
      "step": 4028
    },
    {
      "entropy": 0.9359458573162556,
      "epoch": 0.37069367053668467,
      "grad_norm": 0.16910851001739502,
      "learning_rate": 8.764682430153035e-05,
      "loss": 0.9556,
      "mean_token_accuracy": 0.7247043922543526,
      "num_tokens": 128587309.0,
      "step": 4029
    },
    {
      "entropy": 0.7990800589323044,
      "epoch": 0.37078567690812586,
      "grad_norm": 0.15622714161872864,
      "learning_rate": 8.764375747538872e-05,
      "loss": 0.8078,
      "mean_token_accuracy": 0.7672034092247486,
      "num_tokens": 128619796.0,
      "step": 4030
    },
    {
      "entropy": 0.7223476376384497,
      "epoch": 0.3708776832795671,
      "grad_norm": 0.16829322278499603,
      "learning_rate": 8.76406906492471e-05,
      "loss": 0.7651,
      "mean_token_accuracy": 0.7796723172068596,
      "num_tokens": 128651508.0,
      "step": 4031
    },
    {
      "entropy": 0.7760967016220093,
      "epoch": 0.37096968965100835,
      "grad_norm": 0.16441710293293,
      "learning_rate": 8.763762382310548e-05,
      "loss": 0.7747,
      "mean_token_accuracy": 0.7797593213617802,
      "num_tokens": 128683821.0,
      "step": 4032
    },
    {
      "entropy": 0.7991883810609579,
      "epoch": 0.37106169602244954,
      "grad_norm": 0.15618565678596497,
      "learning_rate": 8.763455699696385e-05,
      "loss": 0.8029,
      "mean_token_accuracy": 0.7616725340485573,
      "num_tokens": 128715925.0,
      "step": 4033
    },
    {
      "entropy": 0.9622782878577709,
      "epoch": 0.3711537023938908,
      "grad_norm": 0.16306471824645996,
      "learning_rate": 8.763149017082221e-05,
      "loss": 0.9652,
      "mean_token_accuracy": 0.7311449497938156,
      "num_tokens": 128748672.0,
      "step": 4034
    },
    {
      "entropy": 0.8192069604992867,
      "epoch": 0.37124570876533197,
      "grad_norm": 0.1609363555908203,
      "learning_rate": 8.76284233446806e-05,
      "loss": 0.8274,
      "mean_token_accuracy": 0.7647753059864044,
      "num_tokens": 128781120.0,
      "step": 4035
    },
    {
      "entropy": 0.8471022918820381,
      "epoch": 0.3713377151367732,
      "grad_norm": 0.1613018810749054,
      "learning_rate": 8.762535651853898e-05,
      "loss": 0.8313,
      "mean_token_accuracy": 0.7602397464215755,
      "num_tokens": 128813401.0,
      "step": 4036
    },
    {
      "entropy": 0.7957539483904839,
      "epoch": 0.37142972150821446,
      "grad_norm": 0.16099482774734497,
      "learning_rate": 8.762228969239734e-05,
      "loss": 0.7833,
      "mean_token_accuracy": 0.7741830721497536,
      "num_tokens": 128845967.0,
      "step": 4037
    },
    {
      "entropy": 0.9150597043335438,
      "epoch": 0.37152172787965565,
      "grad_norm": 0.16377022862434387,
      "learning_rate": 8.761922286625571e-05,
      "loss": 0.9085,
      "mean_token_accuracy": 0.7403393052518368,
      "num_tokens": 128878248.0,
      "step": 4038
    },
    {
      "entropy": 0.790180642157793,
      "epoch": 0.3716137342510969,
      "grad_norm": 0.1547834873199463,
      "learning_rate": 8.76161560401141e-05,
      "loss": 0.7597,
      "mean_token_accuracy": 0.7792324088513851,
      "num_tokens": 128910046.0,
      "step": 4039
    },
    {
      "entropy": 0.9106619320809841,
      "epoch": 0.3717057406225381,
      "grad_norm": 0.15736722946166992,
      "learning_rate": 8.761308921397246e-05,
      "loss": 0.8997,
      "mean_token_accuracy": 0.7398835495114326,
      "num_tokens": 128942074.0,
      "step": 4040
    },
    {
      "entropy": 0.8467965796589851,
      "epoch": 0.37179774699397933,
      "grad_norm": 0.1559259444475174,
      "learning_rate": 8.761002238783084e-05,
      "loss": 0.8283,
      "mean_token_accuracy": 0.7607864290475845,
      "num_tokens": 128974162.0,
      "step": 4041
    },
    {
      "entropy": 0.8675916939973831,
      "epoch": 0.3718897533654206,
      "grad_norm": 0.15791144967079163,
      "learning_rate": 8.760695556168921e-05,
      "loss": 0.8659,
      "mean_token_accuracy": 0.7543719224631786,
      "num_tokens": 129006504.0,
      "step": 4042
    },
    {
      "entropy": 0.7775794602930546,
      "epoch": 0.37198175973686176,
      "grad_norm": 0.15330635011196136,
      "learning_rate": 8.760388873554759e-05,
      "loss": 0.7686,
      "mean_token_accuracy": 0.7757636941969395,
      "num_tokens": 129039151.0,
      "step": 4043
    },
    {
      "entropy": 0.8813570849597454,
      "epoch": 0.372073766108303,
      "grad_norm": 0.16003143787384033,
      "learning_rate": 8.760082190940596e-05,
      "loss": 0.8516,
      "mean_token_accuracy": 0.7550823129713535,
      "num_tokens": 129070276.0,
      "step": 4044
    },
    {
      "entropy": 0.8972842860966921,
      "epoch": 0.3721657724797442,
      "grad_norm": 0.16518156230449677,
      "learning_rate": 8.759775508326433e-05,
      "loss": 0.8962,
      "mean_token_accuracy": 0.7465511150658131,
      "num_tokens": 129102809.0,
      "step": 4045
    },
    {
      "entropy": 0.7625183276832104,
      "epoch": 0.37225777885118544,
      "grad_norm": 0.1548251062631607,
      "learning_rate": 8.759468825712271e-05,
      "loss": 0.7517,
      "mean_token_accuracy": 0.7774330303072929,
      "num_tokens": 129134453.0,
      "step": 4046
    },
    {
      "entropy": 0.7880821712315083,
      "epoch": 0.3723497852226267,
      "grad_norm": 0.16042819619178772,
      "learning_rate": 8.759162143098109e-05,
      "loss": 0.7721,
      "mean_token_accuracy": 0.7738635800778866,
      "num_tokens": 129165879.0,
      "step": 4047
    },
    {
      "entropy": 0.7492539808154106,
      "epoch": 0.3724417915940679,
      "grad_norm": 0.16722649335861206,
      "learning_rate": 8.758855460483946e-05,
      "loss": 0.7716,
      "mean_token_accuracy": 0.7807111106812954,
      "num_tokens": 129197905.0,
      "step": 4048
    },
    {
      "entropy": 0.8774124756455421,
      "epoch": 0.3725337979655091,
      "grad_norm": 0.15700660645961761,
      "learning_rate": 8.758548777869783e-05,
      "loss": 0.8822,
      "mean_token_accuracy": 0.7454073391854763,
      "num_tokens": 129230217.0,
      "step": 4049
    },
    {
      "entropy": 0.8027145750820637,
      "epoch": 0.37262580433695036,
      "grad_norm": 0.15138351917266846,
      "learning_rate": 8.75824209525562e-05,
      "loss": 0.8094,
      "mean_token_accuracy": 0.7588956840336323,
      "num_tokens": 129262802.0,
      "step": 4050
    },
    {
      "entropy": 0.7893565129488707,
      "epoch": 0.37271781070839155,
      "grad_norm": 0.1602085381746292,
      "learning_rate": 8.757935412641458e-05,
      "loss": 0.7785,
      "mean_token_accuracy": 0.7745846696197987,
      "num_tokens": 129295567.0,
      "step": 4051
    },
    {
      "entropy": 0.8350004088133574,
      "epoch": 0.3728098170798328,
      "grad_norm": 0.1655891090631485,
      "learning_rate": 8.757628730027296e-05,
      "loss": 0.8589,
      "mean_token_accuracy": 0.7530957423150539,
      "num_tokens": 129327807.0,
      "step": 4052
    },
    {
      "entropy": 0.8259093780070543,
      "epoch": 0.372901823451274,
      "grad_norm": 0.16318650543689728,
      "learning_rate": 8.757322047413133e-05,
      "loss": 0.8142,
      "mean_token_accuracy": 0.764505922794342,
      "num_tokens": 129360007.0,
      "step": 4053
    },
    {
      "entropy": 0.8945176340639591,
      "epoch": 0.37299382982271523,
      "grad_norm": 0.16525422036647797,
      "learning_rate": 8.75701536479897e-05,
      "loss": 0.905,
      "mean_token_accuracy": 0.7429056577384472,
      "num_tokens": 129391985.0,
      "step": 4054
    },
    {
      "entropy": 0.7505261488258839,
      "epoch": 0.3730858361941565,
      "grad_norm": 0.16096238791942596,
      "learning_rate": 8.756708682184807e-05,
      "loss": 0.7489,
      "mean_token_accuracy": 0.7848134599626064,
      "num_tokens": 129424187.0,
      "step": 4055
    },
    {
      "entropy": 0.8470068201422691,
      "epoch": 0.37317784256559766,
      "grad_norm": 0.1587616205215454,
      "learning_rate": 8.756401999570644e-05,
      "loss": 0.8423,
      "mean_token_accuracy": 0.7597469314932823,
      "num_tokens": 129455913.0,
      "step": 4056
    },
    {
      "entropy": 0.7282635737210512,
      "epoch": 0.3732698489370389,
      "grad_norm": 0.15393796563148499,
      "learning_rate": 8.756095316956482e-05,
      "loss": 0.7162,
      "mean_token_accuracy": 0.7877933345735073,
      "num_tokens": 129488453.0,
      "step": 4057
    },
    {
      "entropy": 0.7363642081618309,
      "epoch": 0.3733618553084801,
      "grad_norm": 0.15905441343784332,
      "learning_rate": 8.75578863434232e-05,
      "loss": 0.6844,
      "mean_token_accuracy": 0.7976165600121021,
      "num_tokens": 129520398.0,
      "step": 4058
    },
    {
      "entropy": 0.7302055843174458,
      "epoch": 0.37345386167992134,
      "grad_norm": 0.15870577096939087,
      "learning_rate": 8.755481951728157e-05,
      "loss": 0.7131,
      "mean_token_accuracy": 0.7894841656088829,
      "num_tokens": 129552242.0,
      "step": 4059
    },
    {
      "entropy": 0.9103264342993498,
      "epoch": 0.3735458680513626,
      "grad_norm": 0.15969295799732208,
      "learning_rate": 8.755175269113994e-05,
      "loss": 0.8817,
      "mean_token_accuracy": 0.7507480531930923,
      "num_tokens": 129584083.0,
      "step": 4060
    },
    {
      "entropy": 0.7411982771009207,
      "epoch": 0.3736378744228038,
      "grad_norm": 0.1514059603214264,
      "learning_rate": 8.754868586499831e-05,
      "loss": 0.7092,
      "mean_token_accuracy": 0.7924705818295479,
      "num_tokens": 129616369.0,
      "step": 4061
    },
    {
      "entropy": 0.7967141177505255,
      "epoch": 0.373729880794245,
      "grad_norm": 0.16127313673496246,
      "learning_rate": 8.754561903885669e-05,
      "loss": 0.7959,
      "mean_token_accuracy": 0.7705704681575298,
      "num_tokens": 129648679.0,
      "step": 4062
    },
    {
      "entropy": 0.8048204500228167,
      "epoch": 0.3738218871656862,
      "grad_norm": 0.14940133690834045,
      "learning_rate": 8.754255221271507e-05,
      "loss": 0.791,
      "mean_token_accuracy": 0.7697445265948772,
      "num_tokens": 129680929.0,
      "step": 4063
    },
    {
      "entropy": 0.7946144975721836,
      "epoch": 0.37391389353712745,
      "grad_norm": 0.16125668585300446,
      "learning_rate": 8.753948538657344e-05,
      "loss": 0.7954,
      "mean_token_accuracy": 0.7688831649720669,
      "num_tokens": 129712055.0,
      "step": 4064
    },
    {
      "entropy": 0.7706652022898197,
      "epoch": 0.3740058999085687,
      "grad_norm": 0.15872368216514587,
      "learning_rate": 8.753641856043181e-05,
      "loss": 0.764,
      "mean_token_accuracy": 0.7789484858512878,
      "num_tokens": 129744153.0,
      "step": 4065
    },
    {
      "entropy": 0.7404598630964756,
      "epoch": 0.3740979062800099,
      "grad_norm": 0.16774213314056396,
      "learning_rate": 8.753335173429019e-05,
      "loss": 0.7524,
      "mean_token_accuracy": 0.7815217263996601,
      "num_tokens": 129776520.0,
      "step": 4066
    },
    {
      "entropy": 0.8747317586094141,
      "epoch": 0.37418991265145113,
      "grad_norm": 0.15980879962444305,
      "learning_rate": 8.753028490814856e-05,
      "loss": 0.8781,
      "mean_token_accuracy": 0.747841116040945,
      "num_tokens": 129807940.0,
      "step": 4067
    },
    {
      "entropy": 0.8335717562586069,
      "epoch": 0.3742819190228923,
      "grad_norm": 0.15513189136981964,
      "learning_rate": 8.752721808200694e-05,
      "loss": 0.853,
      "mean_token_accuracy": 0.7550608068704605,
      "num_tokens": 129840662.0,
      "step": 4068
    },
    {
      "entropy": 0.8145300298929214,
      "epoch": 0.37437392539433356,
      "grad_norm": 0.16231988370418549,
      "learning_rate": 8.75241512558653e-05,
      "loss": 0.8233,
      "mean_token_accuracy": 0.7579226046800613,
      "num_tokens": 129872006.0,
      "step": 4069
    },
    {
      "entropy": 0.8082912191748619,
      "epoch": 0.3744659317657748,
      "grad_norm": 0.16119495034217834,
      "learning_rate": 8.752108442972369e-05,
      "loss": 0.8183,
      "mean_token_accuracy": 0.7613702304661274,
      "num_tokens": 129902976.0,
      "step": 4070
    },
    {
      "entropy": 0.8024329300969839,
      "epoch": 0.374557938137216,
      "grad_norm": 0.15898005664348602,
      "learning_rate": 8.751801760358206e-05,
      "loss": 0.8,
      "mean_token_accuracy": 0.7685107290744781,
      "num_tokens": 129934919.0,
      "step": 4071
    },
    {
      "entropy": 0.8704750724136829,
      "epoch": 0.37464994450865724,
      "grad_norm": 0.16519534587860107,
      "learning_rate": 8.751495077744042e-05,
      "loss": 0.8611,
      "mean_token_accuracy": 0.7537200264632702,
      "num_tokens": 129966771.0,
      "step": 4072
    },
    {
      "entropy": 0.8093382008373737,
      "epoch": 0.37474195088009843,
      "grad_norm": 0.16282886266708374,
      "learning_rate": 8.75118839512988e-05,
      "loss": 0.7958,
      "mean_token_accuracy": 0.7669232971966267,
      "num_tokens": 129999097.0,
      "step": 4073
    },
    {
      "entropy": 0.7768543362617493,
      "epoch": 0.3748339572515397,
      "grad_norm": 0.15813620388507843,
      "learning_rate": 8.750881712515719e-05,
      "loss": 0.7647,
      "mean_token_accuracy": 0.7797915376722813,
      "num_tokens": 130030474.0,
      "step": 4074
    },
    {
      "entropy": 0.7748862598091364,
      "epoch": 0.3749259636229809,
      "grad_norm": 0.16296915709972382,
      "learning_rate": 8.750575029901555e-05,
      "loss": 0.7521,
      "mean_token_accuracy": 0.7797392792999744,
      "num_tokens": 130062546.0,
      "step": 4075
    },
    {
      "entropy": 0.7884459607303143,
      "epoch": 0.3750179699944221,
      "grad_norm": 0.15664036571979523,
      "learning_rate": 8.750268347287392e-05,
      "loss": 0.7719,
      "mean_token_accuracy": 0.7714450508356094,
      "num_tokens": 130093811.0,
      "step": 4076
    },
    {
      "entropy": 0.8191147223114967,
      "epoch": 0.37510997636586335,
      "grad_norm": 0.1540190875530243,
      "learning_rate": 8.749961664673229e-05,
      "loss": 0.8044,
      "mean_token_accuracy": 0.7618381008505821,
      "num_tokens": 130125831.0,
      "step": 4077
    },
    {
      "entropy": 0.7616619635373354,
      "epoch": 0.37520198273730454,
      "grad_norm": 0.15280002355575562,
      "learning_rate": 8.749654982059068e-05,
      "loss": 0.7552,
      "mean_token_accuracy": 0.7791516147553921,
      "num_tokens": 130157752.0,
      "step": 4078
    },
    {
      "entropy": 0.8310909681022167,
      "epoch": 0.3752939891087458,
      "grad_norm": 0.16477175056934357,
      "learning_rate": 8.749348299444905e-05,
      "loss": 0.8174,
      "mean_token_accuracy": 0.7618382759392262,
      "num_tokens": 130190034.0,
      "step": 4079
    },
    {
      "entropy": 0.8722731210291386,
      "epoch": 0.37538599548018703,
      "grad_norm": 0.15813778340816498,
      "learning_rate": 8.749041616830742e-05,
      "loss": 0.8724,
      "mean_token_accuracy": 0.7494565397500992,
      "num_tokens": 130221971.0,
      "step": 4080
    },
    {
      "entropy": 0.779531717300415,
      "epoch": 0.3754780018516282,
      "grad_norm": 0.16346488893032074,
      "learning_rate": 8.748734934216579e-05,
      "loss": 0.7869,
      "mean_token_accuracy": 0.7727317363023758,
      "num_tokens": 130253487.0,
      "step": 4081
    },
    {
      "entropy": 0.8307377435266972,
      "epoch": 0.37557000822306946,
      "grad_norm": 0.15875351428985596,
      "learning_rate": 8.748428251602417e-05,
      "loss": 0.8234,
      "mean_token_accuracy": 0.7608270347118378,
      "num_tokens": 130285934.0,
      "step": 4082
    },
    {
      "entropy": 0.7421064414083958,
      "epoch": 0.37566201459451065,
      "grad_norm": 0.1532420516014099,
      "learning_rate": 8.748121568988255e-05,
      "loss": 0.7324,
      "mean_token_accuracy": 0.7851414568722248,
      "num_tokens": 130317466.0,
      "step": 4083
    },
    {
      "entropy": 0.8471705559641123,
      "epoch": 0.3757540209659519,
      "grad_norm": 0.16140751540660858,
      "learning_rate": 8.747814886374092e-05,
      "loss": 0.8468,
      "mean_token_accuracy": 0.7471309304237366,
      "num_tokens": 130349688.0,
      "step": 4084
    },
    {
      "entropy": 0.9260601215064526,
      "epoch": 0.37584602733739314,
      "grad_norm": 0.16088037192821503,
      "learning_rate": 8.747508203759929e-05,
      "loss": 0.9173,
      "mean_token_accuracy": 0.7380798310041428,
      "num_tokens": 130381491.0,
      "step": 4085
    },
    {
      "entropy": 0.7290041074156761,
      "epoch": 0.37593803370883433,
      "grad_norm": 0.15418820083141327,
      "learning_rate": 8.747201521145767e-05,
      "loss": 0.7302,
      "mean_token_accuracy": 0.7839409485459328,
      "num_tokens": 130413419.0,
      "step": 4086
    },
    {
      "entropy": 0.7705075647681952,
      "epoch": 0.3760300400802756,
      "grad_norm": 0.16282916069030762,
      "learning_rate": 8.746894838531604e-05,
      "loss": 0.7783,
      "mean_token_accuracy": 0.7732329368591309,
      "num_tokens": 130444448.0,
      "step": 4087
    },
    {
      "entropy": 0.8085124716162682,
      "epoch": 0.37612204645171676,
      "grad_norm": 0.16146279871463776,
      "learning_rate": 8.746588155917442e-05,
      "loss": 0.795,
      "mean_token_accuracy": 0.7661039531230927,
      "num_tokens": 130475638.0,
      "step": 4088
    },
    {
      "entropy": 0.8927824720740318,
      "epoch": 0.376214052823158,
      "grad_norm": 0.15700224041938782,
      "learning_rate": 8.74628147330328e-05,
      "loss": 0.8876,
      "mean_token_accuracy": 0.7466563768684864,
      "num_tokens": 130508191.0,
      "step": 4089
    },
    {
      "entropy": 0.8471049219369888,
      "epoch": 0.37630605919459925,
      "grad_norm": 0.17450976371765137,
      "learning_rate": 8.745974790689117e-05,
      "loss": 0.8437,
      "mean_token_accuracy": 0.7590622045099735,
      "num_tokens": 130539907.0,
      "step": 4090
    },
    {
      "entropy": 0.8682767637073994,
      "epoch": 0.37639806556604044,
      "grad_norm": 0.16888317465782166,
      "learning_rate": 8.745668108074954e-05,
      "loss": 0.8502,
      "mean_token_accuracy": 0.7584325596690178,
      "num_tokens": 130571692.0,
      "step": 4091
    },
    {
      "entropy": 0.7743223123252392,
      "epoch": 0.3764900719374817,
      "grad_norm": 0.15394112467765808,
      "learning_rate": 8.74536142546079e-05,
      "loss": 0.7642,
      "mean_token_accuracy": 0.7766902968287468,
      "num_tokens": 130603633.0,
      "step": 4092
    },
    {
      "entropy": 0.8647866453975439,
      "epoch": 0.3765820783089229,
      "grad_norm": 0.16298164427280426,
      "learning_rate": 8.745054742846628e-05,
      "loss": 0.8424,
      "mean_token_accuracy": 0.7574441656470299,
      "num_tokens": 130635072.0,
      "step": 4093
    },
    {
      "entropy": 0.7816735431551933,
      "epoch": 0.3766740846803641,
      "grad_norm": 0.15835200250148773,
      "learning_rate": 8.744748060232467e-05,
      "loss": 0.7777,
      "mean_token_accuracy": 0.7734944336116314,
      "num_tokens": 130667585.0,
      "step": 4094
    },
    {
      "entropy": 0.7584736179560423,
      "epoch": 0.37676609105180536,
      "grad_norm": 0.15218304097652435,
      "learning_rate": 8.744441377618303e-05,
      "loss": 0.7433,
      "mean_token_accuracy": 0.7788853906095028,
      "num_tokens": 130699952.0,
      "step": 4095
    },
    {
      "entropy": 0.7723041009157896,
      "epoch": 0.37685809742324655,
      "grad_norm": 0.16570191085338593,
      "learning_rate": 8.74413469500414e-05,
      "loss": 0.7971,
      "mean_token_accuracy": 0.7707731164991856,
      "num_tokens": 130731021.0,
      "step": 4096
    },
    {
      "entropy": 0.7806574441492558,
      "epoch": 0.3769501037946878,
      "grad_norm": 0.1601337492465973,
      "learning_rate": 8.743828012389978e-05,
      "loss": 0.7871,
      "mean_token_accuracy": 0.7727965414524078,
      "num_tokens": 130762287.0,
      "step": 4097
    },
    {
      "entropy": 0.8342874199151993,
      "epoch": 0.377042110166129,
      "grad_norm": 0.15651856362819672,
      "learning_rate": 8.743521329775815e-05,
      "loss": 0.8403,
      "mean_token_accuracy": 0.754624355584383,
      "num_tokens": 130794779.0,
      "step": 4098
    },
    {
      "entropy": 0.856488760560751,
      "epoch": 0.37713411653757023,
      "grad_norm": 0.1603168547153473,
      "learning_rate": 8.743214647161653e-05,
      "loss": 0.8493,
      "mean_token_accuracy": 0.7536776922643185,
      "num_tokens": 130827046.0,
      "step": 4099
    },
    {
      "entropy": 0.9053232707083225,
      "epoch": 0.3772261229090115,
      "grad_norm": 0.1598374992609024,
      "learning_rate": 8.74290796454749e-05,
      "loss": 0.8755,
      "mean_token_accuracy": 0.7502192966639996,
      "num_tokens": 130859187.0,
      "step": 4100
    },
    {
      "entropy": 0.7434868915006518,
      "epoch": 0.37731812928045266,
      "grad_norm": 0.1514042615890503,
      "learning_rate": 8.742601281933328e-05,
      "loss": 0.7262,
      "mean_token_accuracy": 0.7874855473637581,
      "num_tokens": 130891826.0,
      "step": 4101
    },
    {
      "entropy": 0.7872999217361212,
      "epoch": 0.3774101356518939,
      "grad_norm": 0.15198375284671783,
      "learning_rate": 8.742294599319165e-05,
      "loss": 0.7935,
      "mean_token_accuracy": 0.7658762149512768,
      "num_tokens": 130923589.0,
      "step": 4102
    },
    {
      "entropy": 0.8784041814506054,
      "epoch": 0.3775021420233351,
      "grad_norm": 0.17920370399951935,
      "learning_rate": 8.741987916705002e-05,
      "loss": 0.92,
      "mean_token_accuracy": 0.7437056973576546,
      "num_tokens": 130955550.0,
      "step": 4103
    },
    {
      "entropy": 0.8522918224334717,
      "epoch": 0.37759414839477634,
      "grad_norm": 0.16588614881038666,
      "learning_rate": 8.74168123409084e-05,
      "loss": 0.8504,
      "mean_token_accuracy": 0.7559499740600586,
      "num_tokens": 130987046.0,
      "step": 4104
    },
    {
      "entropy": 0.8293980024755001,
      "epoch": 0.3776861547662176,
      "grad_norm": 0.1556546539068222,
      "learning_rate": 8.741374551476678e-05,
      "loss": 0.7989,
      "mean_token_accuracy": 0.770696971565485,
      "num_tokens": 131019271.0,
      "step": 4105
    },
    {
      "entropy": 0.722433939576149,
      "epoch": 0.3777781611376588,
      "grad_norm": 0.15722379088401794,
      "learning_rate": 8.741067868862515e-05,
      "loss": 0.6965,
      "mean_token_accuracy": 0.7937810495495796,
      "num_tokens": 131051097.0,
      "step": 4106
    },
    {
      "entropy": 0.7864106316119432,
      "epoch": 0.3778701675091,
      "grad_norm": 0.15509875118732452,
      "learning_rate": 8.740761186248352e-05,
      "loss": 0.757,
      "mean_token_accuracy": 0.7723527364432812,
      "num_tokens": 131083182.0,
      "step": 4107
    },
    {
      "entropy": 0.8425085805356503,
      "epoch": 0.3779621738805412,
      "grad_norm": 0.15373443067073822,
      "learning_rate": 8.740454503634188e-05,
      "loss": 0.8324,
      "mean_token_accuracy": 0.7556495629251003,
      "num_tokens": 131115295.0,
      "step": 4108
    },
    {
      "entropy": 0.7058810722082853,
      "epoch": 0.37805418025198245,
      "grad_norm": 0.14839696884155273,
      "learning_rate": 8.740147821020027e-05,
      "loss": 0.689,
      "mean_token_accuracy": 0.795116625726223,
      "num_tokens": 131147473.0,
      "step": 4109
    },
    {
      "entropy": 0.8314052112400532,
      "epoch": 0.3781461866234237,
      "grad_norm": 0.16211660206317902,
      "learning_rate": 8.739841138405865e-05,
      "loss": 0.8186,
      "mean_token_accuracy": 0.7649254463613033,
      "num_tokens": 131179532.0,
      "step": 4110
    },
    {
      "entropy": 0.8366539943963289,
      "epoch": 0.3782381929948649,
      "grad_norm": 0.16195206344127655,
      "learning_rate": 8.739534455791701e-05,
      "loss": 0.8393,
      "mean_token_accuracy": 0.7534327246248722,
      "num_tokens": 131210061.0,
      "step": 4111
    },
    {
      "entropy": 0.7023038323968649,
      "epoch": 0.37833019936630613,
      "grad_norm": 0.16786795854568481,
      "learning_rate": 8.739227773177538e-05,
      "loss": 0.7133,
      "mean_token_accuracy": 0.7892522253096104,
      "num_tokens": 131241610.0,
      "step": 4112
    },
    {
      "entropy": 0.8322776015847921,
      "epoch": 0.3784222057377473,
      "grad_norm": 0.1598309874534607,
      "learning_rate": 8.738921090563376e-05,
      "loss": 0.8396,
      "mean_token_accuracy": 0.7554603070020676,
      "num_tokens": 131274085.0,
      "step": 4113
    },
    {
      "entropy": 0.8342767022550106,
      "epoch": 0.37851421210918856,
      "grad_norm": 0.16540953516960144,
      "learning_rate": 8.738614407949213e-05,
      "loss": 0.8432,
      "mean_token_accuracy": 0.7570647224783897,
      "num_tokens": 131306003.0,
      "step": 4114
    },
    {
      "entropy": 0.8650121614336967,
      "epoch": 0.3786062184806298,
      "grad_norm": 0.16598913073539734,
      "learning_rate": 8.738307725335051e-05,
      "loss": 0.8455,
      "mean_token_accuracy": 0.7583511993288994,
      "num_tokens": 131338122.0,
      "step": 4115
    },
    {
      "entropy": 0.7809188067913055,
      "epoch": 0.378698224852071,
      "grad_norm": 0.1524384468793869,
      "learning_rate": 8.738001042720888e-05,
      "loss": 0.7599,
      "mean_token_accuracy": 0.7775154635310173,
      "num_tokens": 131370342.0,
      "step": 4116
    },
    {
      "entropy": 0.9356481358408928,
      "epoch": 0.37879023122351224,
      "grad_norm": 0.16167990863323212,
      "learning_rate": 8.737694360106726e-05,
      "loss": 0.9275,
      "mean_token_accuracy": 0.7358379811048508,
      "num_tokens": 131402234.0,
      "step": 4117
    },
    {
      "entropy": 0.8367277402430773,
      "epoch": 0.37888223759495343,
      "grad_norm": 0.15693219006061554,
      "learning_rate": 8.737387677492563e-05,
      "loss": 0.8211,
      "mean_token_accuracy": 0.7628904171288013,
      "num_tokens": 131434484.0,
      "step": 4118
    },
    {
      "entropy": 0.7327806763350964,
      "epoch": 0.3789742439663947,
      "grad_norm": 0.15792328119277954,
      "learning_rate": 8.7370809948784e-05,
      "loss": 0.7233,
      "mean_token_accuracy": 0.7841850072145462,
      "num_tokens": 131466178.0,
      "step": 4119
    },
    {
      "entropy": 0.7994197346270084,
      "epoch": 0.3790662503378359,
      "grad_norm": 0.15808866918087006,
      "learning_rate": 8.736774312264239e-05,
      "loss": 0.8031,
      "mean_token_accuracy": 0.7622867226600647,
      "num_tokens": 131497161.0,
      "step": 4120
    },
    {
      "entropy": 0.6789668146520853,
      "epoch": 0.3791582567092771,
      "grad_norm": 0.1588565856218338,
      "learning_rate": 8.736467629650076e-05,
      "loss": 0.6704,
      "mean_token_accuracy": 0.8008765690028667,
      "num_tokens": 131529241.0,
      "step": 4121
    },
    {
      "entropy": 0.7208697088062763,
      "epoch": 0.37925026308071835,
      "grad_norm": 0.15108740329742432,
      "learning_rate": 8.736160947035913e-05,
      "loss": 0.7238,
      "mean_token_accuracy": 0.7898131720721722,
      "num_tokens": 131560941.0,
      "step": 4122
    },
    {
      "entropy": 0.8742173425853252,
      "epoch": 0.37934226945215954,
      "grad_norm": 0.159368634223938,
      "learning_rate": 8.73585426442175e-05,
      "loss": 0.8713,
      "mean_token_accuracy": 0.7482022941112518,
      "num_tokens": 131592149.0,
      "step": 4123
    },
    {
      "entropy": 0.7938090097159147,
      "epoch": 0.3794342758236008,
      "grad_norm": 0.16456268727779388,
      "learning_rate": 8.735547581807588e-05,
      "loss": 0.8084,
      "mean_token_accuracy": 0.7679577544331551,
      "num_tokens": 131624464.0,
      "step": 4124
    },
    {
      "entropy": 0.8315666187554598,
      "epoch": 0.37952628219504203,
      "grad_norm": 0.16005152463912964,
      "learning_rate": 8.735240899193426e-05,
      "loss": 0.8227,
      "mean_token_accuracy": 0.7579706385731697,
      "num_tokens": 131656031.0,
      "step": 4125
    },
    {
      "entropy": 0.7733435351401567,
      "epoch": 0.3796182885664832,
      "grad_norm": 0.15128354728221893,
      "learning_rate": 8.734934216579263e-05,
      "loss": 0.7573,
      "mean_token_accuracy": 0.775660190731287,
      "num_tokens": 131687578.0,
      "step": 4126
    },
    {
      "entropy": 0.8420027866959572,
      "epoch": 0.37971029493792446,
      "grad_norm": 0.15558390319347382,
      "learning_rate": 8.7346275339651e-05,
      "loss": 0.8349,
      "mean_token_accuracy": 0.7579685561358929,
      "num_tokens": 131720082.0,
      "step": 4127
    },
    {
      "entropy": 0.7162178102880716,
      "epoch": 0.37980230130936565,
      "grad_norm": 0.1485842615365982,
      "learning_rate": 8.734320851350938e-05,
      "loss": 0.6972,
      "mean_token_accuracy": 0.7935210280120373,
      "num_tokens": 131752522.0,
      "step": 4128
    },
    {
      "entropy": 0.7365765869617462,
      "epoch": 0.3798943076808069,
      "grad_norm": 0.15274202823638916,
      "learning_rate": 8.734014168736774e-05,
      "loss": 0.716,
      "mean_token_accuracy": 0.789227794855833,
      "num_tokens": 131784631.0,
      "step": 4129
    },
    {
      "entropy": 0.7751382496207952,
      "epoch": 0.37998631405224814,
      "grad_norm": 0.1555785834789276,
      "learning_rate": 8.733707486122613e-05,
      "loss": 0.7771,
      "mean_token_accuracy": 0.7732851877808571,
      "num_tokens": 131816274.0,
      "step": 4130
    },
    {
      "entropy": 0.7928182948380709,
      "epoch": 0.38007832042368933,
      "grad_norm": 0.15829837322235107,
      "learning_rate": 8.73340080350845e-05,
      "loss": 0.7808,
      "mean_token_accuracy": 0.7697306089103222,
      "num_tokens": 131847966.0,
      "step": 4131
    },
    {
      "entropy": 0.8094328492879868,
      "epoch": 0.3801703267951306,
      "grad_norm": 0.15658582746982574,
      "learning_rate": 8.733094120894288e-05,
      "loss": 0.8086,
      "mean_token_accuracy": 0.7645165920257568,
      "num_tokens": 131880515.0,
      "step": 4132
    },
    {
      "entropy": 0.8359375391155481,
      "epoch": 0.38026233316657176,
      "grad_norm": 0.15940798819065094,
      "learning_rate": 8.732787438280124e-05,
      "loss": 0.8374,
      "mean_token_accuracy": 0.7576212175190449,
      "num_tokens": 131912427.0,
      "step": 4133
    },
    {
      "entropy": 0.8925373665988445,
      "epoch": 0.380354339538013,
      "grad_norm": 0.15960022807121277,
      "learning_rate": 8.732480755665961e-05,
      "loss": 0.8853,
      "mean_token_accuracy": 0.7506990842521191,
      "num_tokens": 131944965.0,
      "step": 4134
    },
    {
      "entropy": 0.8384615890681744,
      "epoch": 0.38044634590945425,
      "grad_norm": 0.1664103865623474,
      "learning_rate": 8.732174073051799e-05,
      "loss": 0.8472,
      "mean_token_accuracy": 0.7566993050277233,
      "num_tokens": 131976810.0,
      "step": 4135
    },
    {
      "entropy": 0.8054969385266304,
      "epoch": 0.38053835228089544,
      "grad_norm": 0.1603134423494339,
      "learning_rate": 8.731867390437637e-05,
      "loss": 0.7989,
      "mean_token_accuracy": 0.7680752314627171,
      "num_tokens": 132009578.0,
      "step": 4136
    },
    {
      "entropy": 0.8264594934880733,
      "epoch": 0.3806303586523367,
      "grad_norm": 0.15449967980384827,
      "learning_rate": 8.731560707823474e-05,
      "loss": 0.8129,
      "mean_token_accuracy": 0.7628625370562077,
      "num_tokens": 132041721.0,
      "step": 4137
    },
    {
      "entropy": 0.7545406706631184,
      "epoch": 0.3807223650237779,
      "grad_norm": 0.16117171943187714,
      "learning_rate": 8.731254025209311e-05,
      "loss": 0.7407,
      "mean_token_accuracy": 0.7841242477297783,
      "num_tokens": 132073488.0,
      "step": 4138
    },
    {
      "entropy": 0.9526731111109257,
      "epoch": 0.3808143713952191,
      "grad_norm": 0.16234929859638214,
      "learning_rate": 8.730947342595148e-05,
      "loss": 0.935,
      "mean_token_accuracy": 0.7343675494194031,
      "num_tokens": 132105883.0,
      "step": 4139
    },
    {
      "entropy": 0.8932273220270872,
      "epoch": 0.38090637776666036,
      "grad_norm": 0.16533347964286804,
      "learning_rate": 8.730640659980986e-05,
      "loss": 0.8865,
      "mean_token_accuracy": 0.746624119579792,
      "num_tokens": 132137664.0,
      "step": 4140
    },
    {
      "entropy": 0.79215775616467,
      "epoch": 0.38099838413810155,
      "grad_norm": 0.16269204020500183,
      "learning_rate": 8.730333977366824e-05,
      "loss": 0.7784,
      "mean_token_accuracy": 0.7718823365867138,
      "num_tokens": 132169713.0,
      "step": 4141
    },
    {
      "entropy": 0.774992972612381,
      "epoch": 0.3810903905095428,
      "grad_norm": 0.15812771022319794,
      "learning_rate": 8.730027294752661e-05,
      "loss": 0.7806,
      "mean_token_accuracy": 0.7724192589521408,
      "num_tokens": 132201227.0,
      "step": 4142
    },
    {
      "entropy": 0.9274743683636189,
      "epoch": 0.381182396880984,
      "grad_norm": 0.16166022419929504,
      "learning_rate": 8.729720612138498e-05,
      "loss": 0.9373,
      "mean_token_accuracy": 0.7321321666240692,
      "num_tokens": 132233241.0,
      "step": 4143
    },
    {
      "entropy": 0.9127565287053585,
      "epoch": 0.38127440325242523,
      "grad_norm": 0.1686488837003708,
      "learning_rate": 8.729413929524336e-05,
      "loss": 0.9194,
      "mean_token_accuracy": 0.7379963733255863,
      "num_tokens": 132265656.0,
      "step": 4144
    },
    {
      "entropy": 0.9322084318846464,
      "epoch": 0.3813664096238665,
      "grad_norm": 0.16754263639450073,
      "learning_rate": 8.729107246910173e-05,
      "loss": 0.9442,
      "mean_token_accuracy": 0.7333414107561111,
      "num_tokens": 132298069.0,
      "step": 4145
    },
    {
      "entropy": 0.8430956322699785,
      "epoch": 0.38145841599530766,
      "grad_norm": 0.1616491824388504,
      "learning_rate": 8.728800564296011e-05,
      "loss": 0.8238,
      "mean_token_accuracy": 0.7610622569918633,
      "num_tokens": 132330186.0,
      "step": 4146
    },
    {
      "entropy": 0.8317676223814487,
      "epoch": 0.3815504223667489,
      "grad_norm": 0.16662909090518951,
      "learning_rate": 8.728493881681847e-05,
      "loss": 0.8363,
      "mean_token_accuracy": 0.7577822990715504,
      "num_tokens": 132361593.0,
      "step": 4147
    },
    {
      "entropy": 0.772182259708643,
      "epoch": 0.3816424287381901,
      "grad_norm": 0.15605215728282928,
      "learning_rate": 8.728187199067686e-05,
      "loss": 0.7435,
      "mean_token_accuracy": 0.7831004858016968,
      "num_tokens": 132392933.0,
      "step": 4148
    },
    {
      "entropy": 0.7948896400630474,
      "epoch": 0.38173443510963134,
      "grad_norm": 0.14800789952278137,
      "learning_rate": 8.727880516453522e-05,
      "loss": 0.7608,
      "mean_token_accuracy": 0.778449110686779,
      "num_tokens": 132425090.0,
      "step": 4149
    },
    {
      "entropy": 0.8528195936232805,
      "epoch": 0.3818264414810726,
      "grad_norm": 0.1587211638689041,
      "learning_rate": 8.727573833839359e-05,
      "loss": 0.8183,
      "mean_token_accuracy": 0.7604418210685253,
      "num_tokens": 132456984.0,
      "step": 4150
    },
    {
      "entropy": 0.7968871407210827,
      "epoch": 0.3819184478525138,
      "grad_norm": 0.16234296560287476,
      "learning_rate": 8.727267151225197e-05,
      "loss": 0.7759,
      "mean_token_accuracy": 0.7780869789421558,
      "num_tokens": 132488797.0,
      "step": 4151
    },
    {
      "entropy": 0.8053414858877659,
      "epoch": 0.382010454223955,
      "grad_norm": 0.161317840218544,
      "learning_rate": 8.726960468611035e-05,
      "loss": 0.7886,
      "mean_token_accuracy": 0.7689677029848099,
      "num_tokens": 132520640.0,
      "step": 4152
    },
    {
      "entropy": 0.7724286131560802,
      "epoch": 0.3821024605953962,
      "grad_norm": 0.16549046337604523,
      "learning_rate": 8.726653785996872e-05,
      "loss": 0.7631,
      "mean_token_accuracy": 0.7779057696461678,
      "num_tokens": 132552066.0,
      "step": 4153
    },
    {
      "entropy": 0.7536732461303473,
      "epoch": 0.38219446696683745,
      "grad_norm": 0.15608333051204681,
      "learning_rate": 8.726347103382709e-05,
      "loss": 0.7359,
      "mean_token_accuracy": 0.7830021567642689,
      "num_tokens": 132584147.0,
      "step": 4154
    },
    {
      "entropy": 0.7125535421073437,
      "epoch": 0.3822864733382787,
      "grad_norm": 0.16355609893798828,
      "learning_rate": 8.726040420768547e-05,
      "loss": 0.7021,
      "mean_token_accuracy": 0.7903895042836666,
      "num_tokens": 132616129.0,
      "step": 4155
    },
    {
      "entropy": 0.8814709167927504,
      "epoch": 0.3823784797097199,
      "grad_norm": 0.15736223757266998,
      "learning_rate": 8.725733738154385e-05,
      "loss": 0.8604,
      "mean_token_accuracy": 0.7498734444379807,
      "num_tokens": 132648060.0,
      "step": 4156
    },
    {
      "entropy": 0.7723355144262314,
      "epoch": 0.38247048608116113,
      "grad_norm": 0.159956157207489,
      "learning_rate": 8.725427055540222e-05,
      "loss": 0.7839,
      "mean_token_accuracy": 0.7703192234039307,
      "num_tokens": 132680468.0,
      "step": 4157
    },
    {
      "entropy": 0.7797469981014729,
      "epoch": 0.3825624924526023,
      "grad_norm": 0.16021586954593658,
      "learning_rate": 8.725120372926059e-05,
      "loss": 0.7861,
      "mean_token_accuracy": 0.7706373110413551,
      "num_tokens": 132712646.0,
      "step": 4158
    },
    {
      "entropy": 0.8370532747358084,
      "epoch": 0.38265449882404357,
      "grad_norm": 0.15802547335624695,
      "learning_rate": 8.724813690311897e-05,
      "loss": 0.831,
      "mean_token_accuracy": 0.759765014052391,
      "num_tokens": 132745195.0,
      "step": 4159
    },
    {
      "entropy": 0.8180717304348946,
      "epoch": 0.3827465051954848,
      "grad_norm": 0.15887928009033203,
      "learning_rate": 8.724507007697734e-05,
      "loss": 0.8137,
      "mean_token_accuracy": 0.7726306021213531,
      "num_tokens": 132777374.0,
      "step": 4160
    },
    {
      "entropy": 0.8574179112911224,
      "epoch": 0.382838511566926,
      "grad_norm": 0.15849226713180542,
      "learning_rate": 8.724200325083572e-05,
      "loss": 0.8452,
      "mean_token_accuracy": 0.7567791454494,
      "num_tokens": 132809902.0,
      "step": 4161
    },
    {
      "entropy": 0.8491595275700092,
      "epoch": 0.38293051793836724,
      "grad_norm": 0.1582030951976776,
      "learning_rate": 8.723893642469409e-05,
      "loss": 0.8444,
      "mean_token_accuracy": 0.7545684017241001,
      "num_tokens": 132841987.0,
      "step": 4162
    },
    {
      "entropy": 0.7679115831851959,
      "epoch": 0.38302252430980843,
      "grad_norm": 0.16203658282756805,
      "learning_rate": 8.723586959855247e-05,
      "loss": 0.7593,
      "mean_token_accuracy": 0.7766945324838161,
      "num_tokens": 132873215.0,
      "step": 4163
    },
    {
      "entropy": 0.8694840222597122,
      "epoch": 0.3831145306812497,
      "grad_norm": 0.15923887491226196,
      "learning_rate": 8.723280277241084e-05,
      "loss": 0.8547,
      "mean_token_accuracy": 0.7525530755519867,
      "num_tokens": 132905176.0,
      "step": 4164
    },
    {
      "entropy": 0.8753567039966583,
      "epoch": 0.3832065370526909,
      "grad_norm": 0.15510301291942596,
      "learning_rate": 8.72297359462692e-05,
      "loss": 0.87,
      "mean_token_accuracy": 0.7478569708764553,
      "num_tokens": 132937133.0,
      "step": 4165
    },
    {
      "entropy": 0.8058976922184229,
      "epoch": 0.3832985434241321,
      "grad_norm": 0.15937575697898865,
      "learning_rate": 8.722666912012759e-05,
      "loss": 0.8218,
      "mean_token_accuracy": 0.7617925554513931,
      "num_tokens": 132969572.0,
      "step": 4166
    },
    {
      "entropy": 0.8335188031196594,
      "epoch": 0.38339054979557335,
      "grad_norm": 0.16086150705814362,
      "learning_rate": 8.722360229398597e-05,
      "loss": 0.8181,
      "mean_token_accuracy": 0.7647471949458122,
      "num_tokens": 133002340.0,
      "step": 4167
    },
    {
      "entropy": 0.7191280722618103,
      "epoch": 0.38348255616701454,
      "grad_norm": 0.14876678586006165,
      "learning_rate": 8.722053546784434e-05,
      "loss": 0.706,
      "mean_token_accuracy": 0.7914744168519974,
      "num_tokens": 133033625.0,
      "step": 4168
    },
    {
      "entropy": 0.8077774122357368,
      "epoch": 0.3835745625384558,
      "grad_norm": 0.15351617336273193,
      "learning_rate": 8.72174686417027e-05,
      "loss": 0.7993,
      "mean_token_accuracy": 0.7709279283881187,
      "num_tokens": 133065256.0,
      "step": 4169
    },
    {
      "entropy": 0.8632537759840488,
      "epoch": 0.38366656890989703,
      "grad_norm": 0.1708492785692215,
      "learning_rate": 8.721440181556107e-05,
      "loss": 0.8543,
      "mean_token_accuracy": 0.7542585916817188,
      "num_tokens": 133097350.0,
      "step": 4170
    },
    {
      "entropy": 0.7806493416428566,
      "epoch": 0.3837585752813382,
      "grad_norm": 0.159913569688797,
      "learning_rate": 8.721133498941945e-05,
      "loss": 0.7608,
      "mean_token_accuracy": 0.7800504975020885,
      "num_tokens": 133129302.0,
      "step": 4171
    },
    {
      "entropy": 0.8996094614267349,
      "epoch": 0.38385058165277947,
      "grad_norm": 0.16581954061985016,
      "learning_rate": 8.720826816327783e-05,
      "loss": 0.8976,
      "mean_token_accuracy": 0.7396076358854771,
      "num_tokens": 133161330.0,
      "step": 4172
    },
    {
      "entropy": 0.8007803354412317,
      "epoch": 0.38394258802422065,
      "grad_norm": 0.1646149456501007,
      "learning_rate": 8.72052013371362e-05,
      "loss": 0.8014,
      "mean_token_accuracy": 0.7737562172114849,
      "num_tokens": 133192004.0,
      "step": 4173
    },
    {
      "entropy": 0.771102687343955,
      "epoch": 0.3840345943956619,
      "grad_norm": 0.14791448414325714,
      "learning_rate": 8.720213451099457e-05,
      "loss": 0.7638,
      "mean_token_accuracy": 0.776521984487772,
      "num_tokens": 133224509.0,
      "step": 4174
    },
    {
      "entropy": 0.7271171193569899,
      "epoch": 0.38412660076710314,
      "grad_norm": 0.15652260184288025,
      "learning_rate": 8.719906768485295e-05,
      "loss": 0.7167,
      "mean_token_accuracy": 0.7882281094789505,
      "num_tokens": 133256381.0,
      "step": 4175
    },
    {
      "entropy": 0.8229997791349888,
      "epoch": 0.38421860713854433,
      "grad_norm": 0.16440297663211823,
      "learning_rate": 8.719600085871132e-05,
      "loss": 0.8398,
      "mean_token_accuracy": 0.7601414658129215,
      "num_tokens": 133288078.0,
      "step": 4176
    },
    {
      "entropy": 0.8403734639286995,
      "epoch": 0.3843106135099856,
      "grad_norm": 0.1602124273777008,
      "learning_rate": 8.71929340325697e-05,
      "loss": 0.8432,
      "mean_token_accuracy": 0.7574491240084171,
      "num_tokens": 133319149.0,
      "step": 4177
    },
    {
      "entropy": 0.7893737833946943,
      "epoch": 0.38440261988142677,
      "grad_norm": 0.1642305552959442,
      "learning_rate": 8.718986720642807e-05,
      "loss": 0.8116,
      "mean_token_accuracy": 0.7656954117119312,
      "num_tokens": 133351066.0,
      "step": 4178
    },
    {
      "entropy": 0.8066931031644344,
      "epoch": 0.384494626252868,
      "grad_norm": 0.15673337876796722,
      "learning_rate": 8.718680038028645e-05,
      "loss": 0.7966,
      "mean_token_accuracy": 0.7729683294892311,
      "num_tokens": 133382325.0,
      "step": 4179
    },
    {
      "entropy": 0.8560374546796083,
      "epoch": 0.38458663262430925,
      "grad_norm": 0.15161894261837006,
      "learning_rate": 8.718373355414482e-05,
      "loss": 0.8333,
      "mean_token_accuracy": 0.754586972296238,
      "num_tokens": 133414463.0,
      "step": 4180
    },
    {
      "entropy": 0.8318451754748821,
      "epoch": 0.38467863899575044,
      "grad_norm": 0.15894195437431335,
      "learning_rate": 8.718066672800319e-05,
      "loss": 0.8083,
      "mean_token_accuracy": 0.765265453606844,
      "num_tokens": 133446187.0,
      "step": 4181
    },
    {
      "entropy": 0.7837329618632793,
      "epoch": 0.3847706453671917,
      "grad_norm": 0.15321305394172668,
      "learning_rate": 8.717759990186157e-05,
      "loss": 0.7597,
      "mean_token_accuracy": 0.779250118881464,
      "num_tokens": 133478955.0,
      "step": 4182
    },
    {
      "entropy": 0.7930565625429153,
      "epoch": 0.3848626517386329,
      "grad_norm": 0.15311159193515778,
      "learning_rate": 8.717453307571995e-05,
      "loss": 0.7742,
      "mean_token_accuracy": 0.7722283154726028,
      "num_tokens": 133511074.0,
      "step": 4183
    },
    {
      "entropy": 0.756275612860918,
      "epoch": 0.3849546581100741,
      "grad_norm": 0.1531313806772232,
      "learning_rate": 8.717146624957832e-05,
      "loss": 0.752,
      "mean_token_accuracy": 0.7799185998737812,
      "num_tokens": 133542583.0,
      "step": 4184
    },
    {
      "entropy": 0.8041997253894806,
      "epoch": 0.38504666448151537,
      "grad_norm": 0.1492578089237213,
      "learning_rate": 8.716839942343668e-05,
      "loss": 0.777,
      "mean_token_accuracy": 0.7715765684843063,
      "num_tokens": 133574603.0,
      "step": 4185
    },
    {
      "entropy": 0.7764155007898808,
      "epoch": 0.38513867085295656,
      "grad_norm": 0.16519571840763092,
      "learning_rate": 8.716533259729507e-05,
      "loss": 0.7761,
      "mean_token_accuracy": 0.7725546061992645,
      "num_tokens": 133606266.0,
      "step": 4186
    },
    {
      "entropy": 0.9361011125147343,
      "epoch": 0.3852306772243978,
      "grad_norm": 0.16318535804748535,
      "learning_rate": 8.716226577115343e-05,
      "loss": 0.9313,
      "mean_token_accuracy": 0.7313161008059978,
      "num_tokens": 133637253.0,
      "step": 4187
    },
    {
      "entropy": 0.8292119055986404,
      "epoch": 0.385322683595839,
      "grad_norm": 0.15760721266269684,
      "learning_rate": 8.715919894501181e-05,
      "loss": 0.8359,
      "mean_token_accuracy": 0.7593198046088219,
      "num_tokens": 133669894.0,
      "step": 4188
    },
    {
      "entropy": 0.9470454677939415,
      "epoch": 0.38541468996728023,
      "grad_norm": 0.16229559481143951,
      "learning_rate": 8.715613211887018e-05,
      "loss": 0.9537,
      "mean_token_accuracy": 0.7292276807129383,
      "num_tokens": 133701230.0,
      "step": 4189
    },
    {
      "entropy": 0.8191677685827017,
      "epoch": 0.3855066963387215,
      "grad_norm": 0.16095027327537537,
      "learning_rate": 8.715306529272856e-05,
      "loss": 0.8263,
      "mean_token_accuracy": 0.7593926750123501,
      "num_tokens": 133733146.0,
      "step": 4190
    },
    {
      "entropy": 0.8708713538944721,
      "epoch": 0.38559870271016267,
      "grad_norm": 0.16134944558143616,
      "learning_rate": 8.714999846658693e-05,
      "loss": 0.8474,
      "mean_token_accuracy": 0.7556372284889221,
      "num_tokens": 133764794.0,
      "step": 4191
    },
    {
      "entropy": 0.8958757929503918,
      "epoch": 0.3856907090816039,
      "grad_norm": 0.1613050103187561,
      "learning_rate": 8.71469316404453e-05,
      "loss": 0.8909,
      "mean_token_accuracy": 0.7433039098978043,
      "num_tokens": 133796529.0,
      "step": 4192
    },
    {
      "entropy": 0.8576117418706417,
      "epoch": 0.3857827154530451,
      "grad_norm": 0.1609002649784088,
      "learning_rate": 8.714386481430368e-05,
      "loss": 0.8385,
      "mean_token_accuracy": 0.7608901597559452,
      "num_tokens": 133828414.0,
      "step": 4193
    },
    {
      "entropy": 0.7335216384381056,
      "epoch": 0.38587472182448634,
      "grad_norm": 0.1527215540409088,
      "learning_rate": 8.714079798816206e-05,
      "loss": 0.7151,
      "mean_token_accuracy": 0.7870680652558804,
      "num_tokens": 133861021.0,
      "step": 4194
    },
    {
      "entropy": 0.8775808624923229,
      "epoch": 0.3859667281959276,
      "grad_norm": 0.16030284762382507,
      "learning_rate": 8.713773116202043e-05,
      "loss": 0.8821,
      "mean_token_accuracy": 0.7464111261069775,
      "num_tokens": 133893185.0,
      "step": 4195
    },
    {
      "entropy": 0.8554101772606373,
      "epoch": 0.3860587345673688,
      "grad_norm": 0.15841154754161835,
      "learning_rate": 8.71346643358788e-05,
      "loss": 0.845,
      "mean_token_accuracy": 0.7548527345061302,
      "num_tokens": 133924555.0,
      "step": 4196
    },
    {
      "entropy": 0.8316023517400026,
      "epoch": 0.38615074093881,
      "grad_norm": 0.15814627707004547,
      "learning_rate": 8.713159750973717e-05,
      "loss": 0.8369,
      "mean_token_accuracy": 0.755095649510622,
      "num_tokens": 133956945.0,
      "step": 4197
    },
    {
      "entropy": 0.7204942740499973,
      "epoch": 0.3862427473102512,
      "grad_norm": 0.15711098909378052,
      "learning_rate": 8.712853068359556e-05,
      "loss": 0.7153,
      "mean_token_accuracy": 0.7887427024543285,
      "num_tokens": 133989344.0,
      "step": 4198
    },
    {
      "entropy": 0.7697148099541664,
      "epoch": 0.38633475368169246,
      "grad_norm": 0.15278293192386627,
      "learning_rate": 8.712546385745393e-05,
      "loss": 0.7593,
      "mean_token_accuracy": 0.7804526649415493,
      "num_tokens": 134021828.0,
      "step": 4199
    },
    {
      "entropy": 0.6813086252659559,
      "epoch": 0.3864267600531337,
      "grad_norm": 0.15416164696216583,
      "learning_rate": 8.71223970313123e-05,
      "loss": 0.6936,
      "mean_token_accuracy": 0.7990364544093609,
      "num_tokens": 134053441.0,
      "step": 4200
    },
    {
      "entropy": 0.7935023605823517,
      "epoch": 0.3865187664245749,
      "grad_norm": 0.15562550723552704,
      "learning_rate": 8.711933020517066e-05,
      "loss": 0.7795,
      "mean_token_accuracy": 0.7743678502738476,
      "num_tokens": 134085110.0,
      "step": 4201
    },
    {
      "entropy": 0.7979561369866133,
      "epoch": 0.38661077279601613,
      "grad_norm": 0.16511832177639008,
      "learning_rate": 8.711626337902905e-05,
      "loss": 0.801,
      "mean_token_accuracy": 0.7701157331466675,
      "num_tokens": 134117111.0,
      "step": 4202
    },
    {
      "entropy": 0.6937302574515343,
      "epoch": 0.3867027791674573,
      "grad_norm": 0.1549079567193985,
      "learning_rate": 8.711319655288743e-05,
      "loss": 0.6892,
      "mean_token_accuracy": 0.7975854426622391,
      "num_tokens": 134149677.0,
      "step": 4203
    },
    {
      "entropy": 0.7984633184969425,
      "epoch": 0.38679478553889857,
      "grad_norm": 0.15964850783348083,
      "learning_rate": 8.71101297267458e-05,
      "loss": 0.8029,
      "mean_token_accuracy": 0.7672463357448578,
      "num_tokens": 134182041.0,
      "step": 4204
    },
    {
      "entropy": 0.7696676105260849,
      "epoch": 0.3868867919103398,
      "grad_norm": 0.15972919762134552,
      "learning_rate": 8.710706290060416e-05,
      "loss": 0.7728,
      "mean_token_accuracy": 0.7753067091107368,
      "num_tokens": 134214413.0,
      "step": 4205
    },
    {
      "entropy": 0.8014027662575245,
      "epoch": 0.386978798281781,
      "grad_norm": 0.15655213594436646,
      "learning_rate": 8.710399607446255e-05,
      "loss": 0.7878,
      "mean_token_accuracy": 0.7748489081859589,
      "num_tokens": 134246380.0,
      "step": 4206
    },
    {
      "entropy": 0.8616602793335915,
      "epoch": 0.38707080465322224,
      "grad_norm": 0.17280034720897675,
      "learning_rate": 8.710092924832091e-05,
      "loss": 0.8864,
      "mean_token_accuracy": 0.7538504488766193,
      "num_tokens": 134278535.0,
      "step": 4207
    },
    {
      "entropy": 0.7511812876909971,
      "epoch": 0.38716281102466343,
      "grad_norm": 0.15573735535144806,
      "learning_rate": 8.70978624221793e-05,
      "loss": 0.732,
      "mean_token_accuracy": 0.7839108295738697,
      "num_tokens": 134310873.0,
      "step": 4208
    },
    {
      "entropy": 0.9528402052819729,
      "epoch": 0.3872548173961047,
      "grad_norm": 0.1644151210784912,
      "learning_rate": 8.709479559603766e-05,
      "loss": 0.9455,
      "mean_token_accuracy": 0.7285618931055069,
      "num_tokens": 134343216.0,
      "step": 4209
    },
    {
      "entropy": 0.8425240777432919,
      "epoch": 0.3873468237675459,
      "grad_norm": 0.16851243376731873,
      "learning_rate": 8.709172876989604e-05,
      "loss": 0.8464,
      "mean_token_accuracy": 0.758952334523201,
      "num_tokens": 134375772.0,
      "step": 4210
    },
    {
      "entropy": 0.8359720408916473,
      "epoch": 0.3874388301389871,
      "grad_norm": 0.16088487207889557,
      "learning_rate": 8.708866194375441e-05,
      "loss": 0.8215,
      "mean_token_accuracy": 0.7593883536756039,
      "num_tokens": 134408138.0,
      "step": 4211
    },
    {
      "entropy": 0.8019145876169205,
      "epoch": 0.38753083651042836,
      "grad_norm": 0.15221558511257172,
      "learning_rate": 8.708559511761278e-05,
      "loss": 0.7716,
      "mean_token_accuracy": 0.774916123598814,
      "num_tokens": 134439957.0,
      "step": 4212
    },
    {
      "entropy": 0.8435503970831633,
      "epoch": 0.38762284288186954,
      "grad_norm": 0.15350355207920074,
      "learning_rate": 8.708252829147116e-05,
      "loss": 0.8274,
      "mean_token_accuracy": 0.7638750895857811,
      "num_tokens": 134472685.0,
      "step": 4213
    },
    {
      "entropy": 0.788871631026268,
      "epoch": 0.3877148492533108,
      "grad_norm": 0.158906951546669,
      "learning_rate": 8.707946146532954e-05,
      "loss": 0.7766,
      "mean_token_accuracy": 0.7702260911464691,
      "num_tokens": 134504008.0,
      "step": 4214
    },
    {
      "entropy": 0.7112585213035345,
      "epoch": 0.38780685562475203,
      "grad_norm": 0.15209490060806274,
      "learning_rate": 8.707639463918791e-05,
      "loss": 0.6941,
      "mean_token_accuracy": 0.7990138903260231,
      "num_tokens": 134535980.0,
      "step": 4215
    },
    {
      "entropy": 0.8328629322350025,
      "epoch": 0.3878988619961932,
      "grad_norm": 0.1673279106616974,
      "learning_rate": 8.707332781304628e-05,
      "loss": 0.8312,
      "mean_token_accuracy": 0.7628801353275776,
      "num_tokens": 134568230.0,
      "step": 4216
    },
    {
      "entropy": 0.8737826719880104,
      "epoch": 0.38799086836763447,
      "grad_norm": 0.1561863124370575,
      "learning_rate": 8.707026098690466e-05,
      "loss": 0.8518,
      "mean_token_accuracy": 0.7539734244346619,
      "num_tokens": 134600532.0,
      "step": 4217
    },
    {
      "entropy": 0.7237005848437548,
      "epoch": 0.38808287473907566,
      "grad_norm": 0.1551409661769867,
      "learning_rate": 8.706719416076303e-05,
      "loss": 0.7228,
      "mean_token_accuracy": 0.7873458750545979,
      "num_tokens": 134632503.0,
      "step": 4218
    },
    {
      "entropy": 0.8341682404279709,
      "epoch": 0.3881748811105169,
      "grad_norm": 0.16514167189598083,
      "learning_rate": 8.706412733462141e-05,
      "loss": 0.8442,
      "mean_token_accuracy": 0.7564951665699482,
      "num_tokens": 134664814.0,
      "step": 4219
    },
    {
      "entropy": 0.8619032017886639,
      "epoch": 0.38826688748195815,
      "grad_norm": 0.1630931794643402,
      "learning_rate": 8.706106050847978e-05,
      "loss": 0.8675,
      "mean_token_accuracy": 0.7489371336996555,
      "num_tokens": 134696456.0,
      "step": 4220
    },
    {
      "entropy": 0.9019888192415237,
      "epoch": 0.38835889385339933,
      "grad_norm": 0.15872278809547424,
      "learning_rate": 8.705799368233816e-05,
      "loss": 0.9141,
      "mean_token_accuracy": 0.7393311262130737,
      "num_tokens": 134728454.0,
      "step": 4221
    },
    {
      "entropy": 0.7539839446544647,
      "epoch": 0.3884509002248406,
      "grad_norm": 0.15352632105350494,
      "learning_rate": 8.705492685619653e-05,
      "loss": 0.7546,
      "mean_token_accuracy": 0.7800309620797634,
      "num_tokens": 134760069.0,
      "step": 4222
    },
    {
      "entropy": 0.8500512726604939,
      "epoch": 0.38854290659628177,
      "grad_norm": 0.16004939377307892,
      "learning_rate": 8.70518600300549e-05,
      "loss": 0.8446,
      "mean_token_accuracy": 0.7574321068823338,
      "num_tokens": 134792397.0,
      "step": 4223
    },
    {
      "entropy": 0.844186196103692,
      "epoch": 0.388634912967723,
      "grad_norm": 0.17345677316188812,
      "learning_rate": 8.704879320391328e-05,
      "loss": 0.8441,
      "mean_token_accuracy": 0.760069228708744,
      "num_tokens": 134824392.0,
      "step": 4224
    },
    {
      "entropy": 0.7186835035681725,
      "epoch": 0.38872691933916426,
      "grad_norm": 0.15049639344215393,
      "learning_rate": 8.704572637777166e-05,
      "loss": 0.6799,
      "mean_token_accuracy": 0.7980187386274338,
      "num_tokens": 134856190.0,
      "step": 4225
    },
    {
      "entropy": 0.7883167490363121,
      "epoch": 0.38881892571060545,
      "grad_norm": 0.15599362552165985,
      "learning_rate": 8.704265955163002e-05,
      "loss": 0.7596,
      "mean_token_accuracy": 0.7773712426424026,
      "num_tokens": 134888507.0,
      "step": 4226
    },
    {
      "entropy": 0.8072188794612885,
      "epoch": 0.3889109320820467,
      "grad_norm": 0.15790718793869019,
      "learning_rate": 8.703959272548839e-05,
      "loss": 0.7887,
      "mean_token_accuracy": 0.7685910761356354,
      "num_tokens": 134919577.0,
      "step": 4227
    },
    {
      "entropy": 0.8350319005548954,
      "epoch": 0.3890029384534879,
      "grad_norm": 0.16331906616687775,
      "learning_rate": 8.703652589934676e-05,
      "loss": 0.8459,
      "mean_token_accuracy": 0.7563123404979706,
      "num_tokens": 134950472.0,
      "step": 4228
    },
    {
      "entropy": 0.7090620063245296,
      "epoch": 0.3890949448249291,
      "grad_norm": 0.1547655463218689,
      "learning_rate": 8.703345907320514e-05,
      "loss": 0.7006,
      "mean_token_accuracy": 0.7906766124069691,
      "num_tokens": 134982207.0,
      "step": 4229
    },
    {
      "entropy": 0.6968223545700312,
      "epoch": 0.38918695119637037,
      "grad_norm": 0.1543329358100891,
      "learning_rate": 8.703039224706352e-05,
      "loss": 0.6781,
      "mean_token_accuracy": 0.7975941747426987,
      "num_tokens": 135013830.0,
      "step": 4230
    },
    {
      "entropy": 0.8722751513123512,
      "epoch": 0.38927895756781156,
      "grad_norm": 0.16849082708358765,
      "learning_rate": 8.702732542092189e-05,
      "loss": 0.8701,
      "mean_token_accuracy": 0.7547032833099365,
      "num_tokens": 135044247.0,
      "step": 4231
    },
    {
      "entropy": 0.6744404323399067,
      "epoch": 0.3893709639392528,
      "grad_norm": 0.15574614703655243,
      "learning_rate": 8.702425859478026e-05,
      "loss": 0.674,
      "mean_token_accuracy": 0.7998658791184425,
      "num_tokens": 135076507.0,
      "step": 4232
    },
    {
      "entropy": 0.6893910281360149,
      "epoch": 0.389462970310694,
      "grad_norm": 0.15297769010066986,
      "learning_rate": 8.702119176863864e-05,
      "loss": 0.6758,
      "mean_token_accuracy": 0.799257967621088,
      "num_tokens": 135108437.0,
      "step": 4233
    },
    {
      "entropy": 0.7742252424359322,
      "epoch": 0.38955497668213523,
      "grad_norm": 0.16956380009651184,
      "learning_rate": 8.701812494249701e-05,
      "loss": 0.7876,
      "mean_token_accuracy": 0.7735006436705589,
      "num_tokens": 135140599.0,
      "step": 4234
    },
    {
      "entropy": 0.8807965721935034,
      "epoch": 0.3896469830535765,
      "grad_norm": 0.1647479236125946,
      "learning_rate": 8.701505811635539e-05,
      "loss": 0.882,
      "mean_token_accuracy": 0.7443690225481987,
      "num_tokens": 135173264.0,
      "step": 4235
    },
    {
      "entropy": 0.72123758867383,
      "epoch": 0.38973898942501767,
      "grad_norm": 0.1532202661037445,
      "learning_rate": 8.701199129021376e-05,
      "loss": 0.7158,
      "mean_token_accuracy": 0.7881754785776138,
      "num_tokens": 135205090.0,
      "step": 4236
    },
    {
      "entropy": 0.7169564235955477,
      "epoch": 0.3898309957964589,
      "grad_norm": 0.15502481162548065,
      "learning_rate": 8.700892446407214e-05,
      "loss": 0.7125,
      "mean_token_accuracy": 0.7953151874244213,
      "num_tokens": 135236394.0,
      "step": 4237
    },
    {
      "entropy": 0.9326854236423969,
      "epoch": 0.3899230021679001,
      "grad_norm": 0.16728663444519043,
      "learning_rate": 8.70058576379305e-05,
      "loss": 0.9067,
      "mean_token_accuracy": 0.7416993714869022,
      "num_tokens": 135268675.0,
      "step": 4238
    },
    {
      "entropy": 0.7945006601512432,
      "epoch": 0.39001500853934135,
      "grad_norm": 0.16337938606739044,
      "learning_rate": 8.700279081178887e-05,
      "loss": 0.7823,
      "mean_token_accuracy": 0.7731934785842896,
      "num_tokens": 135300337.0,
      "step": 4239
    },
    {
      "entropy": 0.8091100268065929,
      "epoch": 0.3901070149107826,
      "grad_norm": 0.15602712333202362,
      "learning_rate": 8.699972398564726e-05,
      "loss": 0.7923,
      "mean_token_accuracy": 0.7680522687733173,
      "num_tokens": 135332602.0,
      "step": 4240
    },
    {
      "entropy": 0.7889046054333448,
      "epoch": 0.3901990212822238,
      "grad_norm": 0.15201322734355927,
      "learning_rate": 8.699665715950564e-05,
      "loss": 0.7693,
      "mean_token_accuracy": 0.7768738605082035,
      "num_tokens": 135364521.0,
      "step": 4241
    },
    {
      "entropy": 0.7764573730528355,
      "epoch": 0.390291027653665,
      "grad_norm": 0.15514345467090607,
      "learning_rate": 8.6993590333364e-05,
      "loss": 0.776,
      "mean_token_accuracy": 0.7737414389848709,
      "num_tokens": 135396997.0,
      "step": 4242
    },
    {
      "entropy": 0.9031853601336479,
      "epoch": 0.3903830340251062,
      "grad_norm": 0.16435615718364716,
      "learning_rate": 8.699052350722237e-05,
      "loss": 0.8924,
      "mean_token_accuracy": 0.7460091523826122,
      "num_tokens": 135428979.0,
      "step": 4243
    },
    {
      "entropy": 0.7554326169192791,
      "epoch": 0.39047504039654746,
      "grad_norm": 0.1573941856622696,
      "learning_rate": 8.698745668108075e-05,
      "loss": 0.7468,
      "mean_token_accuracy": 0.7834243625402451,
      "num_tokens": 135461203.0,
      "step": 4244
    },
    {
      "entropy": 0.7951471656560898,
      "epoch": 0.3905670467679887,
      "grad_norm": 0.15188875794410706,
      "learning_rate": 8.698438985493914e-05,
      "loss": 0.7897,
      "mean_token_accuracy": 0.7691411823034286,
      "num_tokens": 135493177.0,
      "step": 4245
    },
    {
      "entropy": 0.7393810134381056,
      "epoch": 0.3906590531394299,
      "grad_norm": 0.15031833946704865,
      "learning_rate": 8.69813230287975e-05,
      "loss": 0.7328,
      "mean_token_accuracy": 0.78482611104846,
      "num_tokens": 135525442.0,
      "step": 4246
    },
    {
      "entropy": 0.9201915860176086,
      "epoch": 0.39075105951087113,
      "grad_norm": 0.1618070900440216,
      "learning_rate": 8.697825620265587e-05,
      "loss": 0.8931,
      "mean_token_accuracy": 0.7400780506432056,
      "num_tokens": 135557460.0,
      "step": 4247
    },
    {
      "entropy": 0.8050581961870193,
      "epoch": 0.3908430658823123,
      "grad_norm": 0.159433051943779,
      "learning_rate": 8.697518937651425e-05,
      "loss": 0.7785,
      "mean_token_accuracy": 0.775267980992794,
      "num_tokens": 135589226.0,
      "step": 4248
    },
    {
      "entropy": 0.8434535823762417,
      "epoch": 0.39093507225375357,
      "grad_norm": 0.15734606981277466,
      "learning_rate": 8.697212255037262e-05,
      "loss": 0.8154,
      "mean_token_accuracy": 0.7648815251886845,
      "num_tokens": 135621098.0,
      "step": 4249
    },
    {
      "entropy": 0.8726262971758842,
      "epoch": 0.3910270786251948,
      "grad_norm": 0.16172336041927338,
      "learning_rate": 8.6969055724231e-05,
      "loss": 0.8717,
      "mean_token_accuracy": 0.7471059560775757,
      "num_tokens": 135653325.0,
      "step": 4250
    },
    {
      "entropy": 0.7538423966616392,
      "epoch": 0.391119084996636,
      "grad_norm": 0.1652820110321045,
      "learning_rate": 8.696598889808937e-05,
      "loss": 0.7891,
      "mean_token_accuracy": 0.7716167457401752,
      "num_tokens": 135685796.0,
      "step": 4251
    },
    {
      "entropy": 0.8261342085897923,
      "epoch": 0.39121109136807725,
      "grad_norm": 0.1696327030658722,
      "learning_rate": 8.696292207194775e-05,
      "loss": 0.8451,
      "mean_token_accuracy": 0.7563733607530594,
      "num_tokens": 135717904.0,
      "step": 4252
    },
    {
      "entropy": 0.8632139395922422,
      "epoch": 0.39130309773951844,
      "grad_norm": 0.16471731662750244,
      "learning_rate": 8.695985524580612e-05,
      "loss": 0.875,
      "mean_token_accuracy": 0.7469199262559414,
      "num_tokens": 135749616.0,
      "step": 4253
    },
    {
      "entropy": 0.807172367349267,
      "epoch": 0.3913951041109597,
      "grad_norm": 0.15884307026863098,
      "learning_rate": 8.695678841966449e-05,
      "loss": 0.8065,
      "mean_token_accuracy": 0.7703887373209,
      "num_tokens": 135781930.0,
      "step": 4254
    },
    {
      "entropy": 0.8398078698664904,
      "epoch": 0.3914871104824009,
      "grad_norm": 0.15707585215568542,
      "learning_rate": 8.695372159352287e-05,
      "loss": 0.8286,
      "mean_token_accuracy": 0.7569966651499271,
      "num_tokens": 135813167.0,
      "step": 4255
    },
    {
      "entropy": 0.7881499528884888,
      "epoch": 0.3915791168538421,
      "grad_norm": 0.15740911662578583,
      "learning_rate": 8.695065476738125e-05,
      "loss": 0.7795,
      "mean_token_accuracy": 0.7707737535238266,
      "num_tokens": 135845454.0,
      "step": 4256
    },
    {
      "entropy": 0.7546611130237579,
      "epoch": 0.39167112322528336,
      "grad_norm": 0.15941055119037628,
      "learning_rate": 8.694758794123962e-05,
      "loss": 0.7412,
      "mean_token_accuracy": 0.782977219671011,
      "num_tokens": 135877962.0,
      "step": 4257
    },
    {
      "entropy": 0.7950153239071369,
      "epoch": 0.39176312959672455,
      "grad_norm": 0.15787342190742493,
      "learning_rate": 8.694452111509799e-05,
      "loss": 0.799,
      "mean_token_accuracy": 0.7688364051282406,
      "num_tokens": 135909315.0,
      "step": 4258
    },
    {
      "entropy": 0.8668531365692616,
      "epoch": 0.3918551359681658,
      "grad_norm": 0.16240952908992767,
      "learning_rate": 8.694145428895635e-05,
      "loss": 0.8491,
      "mean_token_accuracy": 0.7509497664868832,
      "num_tokens": 135940355.0,
      "step": 4259
    },
    {
      "entropy": 0.8437719494104385,
      "epoch": 0.39194714233960704,
      "grad_norm": 0.1675802767276764,
      "learning_rate": 8.693838746281474e-05,
      "loss": 0.8642,
      "mean_token_accuracy": 0.7551863230764866,
      "num_tokens": 135971550.0,
      "step": 4260
    },
    {
      "entropy": 0.9080169722437859,
      "epoch": 0.3920391487110482,
      "grad_norm": 0.16223080456256866,
      "learning_rate": 8.693532063667312e-05,
      "loss": 0.9004,
      "mean_token_accuracy": 0.741890512406826,
      "num_tokens": 136001356.0,
      "step": 4261
    },
    {
      "entropy": 0.805962735787034,
      "epoch": 0.39213115508248947,
      "grad_norm": 0.16258211433887482,
      "learning_rate": 8.693225381053148e-05,
      "loss": 0.7836,
      "mean_token_accuracy": 0.7722995840013027,
      "num_tokens": 136032132.0,
      "step": 4262
    },
    {
      "entropy": 0.7431719992309809,
      "epoch": 0.39222316145393066,
      "grad_norm": 0.15347407758235931,
      "learning_rate": 8.692918698438985e-05,
      "loss": 0.719,
      "mean_token_accuracy": 0.7942002974450588,
      "num_tokens": 136063480.0,
      "step": 4263
    },
    {
      "entropy": 0.7477964982390404,
      "epoch": 0.3923151678253719,
      "grad_norm": 0.15796522796154022,
      "learning_rate": 8.692612015824823e-05,
      "loss": 0.7399,
      "mean_token_accuracy": 0.7820676788687706,
      "num_tokens": 136095591.0,
      "step": 4264
    },
    {
      "entropy": 0.8695437591522932,
      "epoch": 0.39240717419681315,
      "grad_norm": 0.1583152860403061,
      "learning_rate": 8.69230533321066e-05,
      "loss": 0.8651,
      "mean_token_accuracy": 0.7492977567017078,
      "num_tokens": 136128359.0,
      "step": 4265
    },
    {
      "entropy": 0.783818980678916,
      "epoch": 0.39249918056825434,
      "grad_norm": 0.16042211651802063,
      "learning_rate": 8.691998650596498e-05,
      "loss": 0.7611,
      "mean_token_accuracy": 0.77845149487257,
      "num_tokens": 136159781.0,
      "step": 4266
    },
    {
      "entropy": 0.7877401243895292,
      "epoch": 0.3925911869396956,
      "grad_norm": 0.16819101572036743,
      "learning_rate": 8.691691967982335e-05,
      "loss": 0.8,
      "mean_token_accuracy": 0.7744305208325386,
      "num_tokens": 136191234.0,
      "step": 4267
    },
    {
      "entropy": 0.8319972977042198,
      "epoch": 0.39268319331113677,
      "grad_norm": 0.16077058017253876,
      "learning_rate": 8.691385285368173e-05,
      "loss": 0.8163,
      "mean_token_accuracy": 0.7644079588353634,
      "num_tokens": 136223491.0,
      "step": 4268
    },
    {
      "entropy": 0.8760599363595247,
      "epoch": 0.392775199682578,
      "grad_norm": 0.15950432419776917,
      "learning_rate": 8.69107860275401e-05,
      "loss": 0.8713,
      "mean_token_accuracy": 0.7499559931457043,
      "num_tokens": 136255148.0,
      "step": 4269
    },
    {
      "entropy": 0.8729541264474392,
      "epoch": 0.39286720605401926,
      "grad_norm": 0.16192279756069183,
      "learning_rate": 8.690771920139847e-05,
      "loss": 0.8802,
      "mean_token_accuracy": 0.7461378425359726,
      "num_tokens": 136286731.0,
      "step": 4270
    },
    {
      "entropy": 0.8163797110319138,
      "epoch": 0.39295921242546045,
      "grad_norm": 0.16267050802707672,
      "learning_rate": 8.690465237525685e-05,
      "loss": 0.8156,
      "mean_token_accuracy": 0.768214039504528,
      "num_tokens": 136318831.0,
      "step": 4271
    },
    {
      "entropy": 0.7846856042742729,
      "epoch": 0.3930512187969017,
      "grad_norm": 0.15733559429645538,
      "learning_rate": 8.690158554911523e-05,
      "loss": 0.7822,
      "mean_token_accuracy": 0.7727215215563774,
      "num_tokens": 136351480.0,
      "step": 4272
    },
    {
      "entropy": 0.7652441691607237,
      "epoch": 0.3931432251683429,
      "grad_norm": 0.16149640083312988,
      "learning_rate": 8.68985187229736e-05,
      "loss": 0.7584,
      "mean_token_accuracy": 0.7765516340732574,
      "num_tokens": 136383042.0,
      "step": 4273
    },
    {
      "entropy": 0.7398889735341072,
      "epoch": 0.3932352315397841,
      "grad_norm": 0.1670493185520172,
      "learning_rate": 8.689545189683197e-05,
      "loss": 0.73,
      "mean_token_accuracy": 0.7817765884101391,
      "num_tokens": 136414822.0,
      "step": 4274
    },
    {
      "entropy": 0.7900633383542299,
      "epoch": 0.39332723791122537,
      "grad_norm": 0.1714436411857605,
      "learning_rate": 8.689238507069035e-05,
      "loss": 0.8136,
      "mean_token_accuracy": 0.7675349563360214,
      "num_tokens": 136446619.0,
      "step": 4275
    },
    {
      "entropy": 0.7011450156569481,
      "epoch": 0.39341924428266656,
      "grad_norm": 0.15032297372817993,
      "learning_rate": 8.688931824454873e-05,
      "loss": 0.6977,
      "mean_token_accuracy": 0.791289072483778,
      "num_tokens": 136478991.0,
      "step": 4276
    },
    {
      "entropy": 0.8527170699089766,
      "epoch": 0.3935112506541078,
      "grad_norm": 0.15444126725196838,
      "learning_rate": 8.68862514184071e-05,
      "loss": 0.8293,
      "mean_token_accuracy": 0.7543445117771626,
      "num_tokens": 136510226.0,
      "step": 4277
    },
    {
      "entropy": 0.8368568681180477,
      "epoch": 0.393603257025549,
      "grad_norm": 0.1593276858329773,
      "learning_rate": 8.688318459226547e-05,
      "loss": 0.8056,
      "mean_token_accuracy": 0.7639561183750629,
      "num_tokens": 136541571.0,
      "step": 4278
    },
    {
      "entropy": 0.8088554311543703,
      "epoch": 0.39369526339699024,
      "grad_norm": 0.15476170182228088,
      "learning_rate": 8.688011776612385e-05,
      "loss": 0.7923,
      "mean_token_accuracy": 0.771268505603075,
      "num_tokens": 136573398.0,
      "step": 4279
    },
    {
      "entropy": 0.7582065667957067,
      "epoch": 0.3937872697684315,
      "grad_norm": 0.15158085525035858,
      "learning_rate": 8.687705093998221e-05,
      "loss": 0.743,
      "mean_token_accuracy": 0.7769364193081856,
      "num_tokens": 136604760.0,
      "step": 4280
    },
    {
      "entropy": 0.8048241846263409,
      "epoch": 0.39387927613987267,
      "grad_norm": 0.15986472368240356,
      "learning_rate": 8.68739841138406e-05,
      "loss": 0.7715,
      "mean_token_accuracy": 0.7742272578179836,
      "num_tokens": 136637370.0,
      "step": 4281
    },
    {
      "entropy": 0.7341096103191376,
      "epoch": 0.3939712825113139,
      "grad_norm": 0.165479376912117,
      "learning_rate": 8.687091728769896e-05,
      "loss": 0.7303,
      "mean_token_accuracy": 0.7843291647732258,
      "num_tokens": 136668735.0,
      "step": 4282
    },
    {
      "entropy": 0.7234653905034065,
      "epoch": 0.39406328888275516,
      "grad_norm": 0.14720401167869568,
      "learning_rate": 8.686785046155735e-05,
      "loss": 0.7152,
      "mean_token_accuracy": 0.7899783402681351,
      "num_tokens": 136701323.0,
      "step": 4283
    },
    {
      "entropy": 0.7249206099659204,
      "epoch": 0.39415529525419635,
      "grad_norm": 0.14828458428382874,
      "learning_rate": 8.686478363541571e-05,
      "loss": 0.6987,
      "mean_token_accuracy": 0.7909929603338242,
      "num_tokens": 136733133.0,
      "step": 4284
    },
    {
      "entropy": 0.6868159528821707,
      "epoch": 0.3942473016256376,
      "grad_norm": 0.15236587822437286,
      "learning_rate": 8.686171680927408e-05,
      "loss": 0.7031,
      "mean_token_accuracy": 0.7894863039255142,
      "num_tokens": 136765351.0,
      "step": 4285
    },
    {
      "entropy": 0.6704693287611008,
      "epoch": 0.3943393079970788,
      "grad_norm": 0.16215863823890686,
      "learning_rate": 8.685864998313246e-05,
      "loss": 0.6858,
      "mean_token_accuracy": 0.7942287735641003,
      "num_tokens": 136797200.0,
      "step": 4286
    },
    {
      "entropy": 0.7387065887451172,
      "epoch": 0.39443131436852,
      "grad_norm": 0.1537497639656067,
      "learning_rate": 8.685558315699084e-05,
      "loss": 0.7361,
      "mean_token_accuracy": 0.7873684801161289,
      "num_tokens": 136829161.0,
      "step": 4287
    },
    {
      "entropy": 0.9072910137474537,
      "epoch": 0.39452332073996127,
      "grad_norm": 0.17608384788036346,
      "learning_rate": 8.685251633084921e-05,
      "loss": 0.946,
      "mean_token_accuracy": 0.7311474904417992,
      "num_tokens": 136861045.0,
      "step": 4288
    },
    {
      "entropy": 0.7872186508029699,
      "epoch": 0.39461532711140246,
      "grad_norm": 0.1784304976463318,
      "learning_rate": 8.684944950470758e-05,
      "loss": 0.813,
      "mean_token_accuracy": 0.769575472921133,
      "num_tokens": 136892542.0,
      "step": 4289
    },
    {
      "entropy": 0.684319069609046,
      "epoch": 0.3947073334828437,
      "grad_norm": 0.15635959804058075,
      "learning_rate": 8.684638267856595e-05,
      "loss": 0.6807,
      "mean_token_accuracy": 0.7998765744268894,
      "num_tokens": 136923989.0,
      "step": 4290
    },
    {
      "entropy": 0.7899093423038721,
      "epoch": 0.3947993398542849,
      "grad_norm": 0.1724378913640976,
      "learning_rate": 8.684331585242433e-05,
      "loss": 0.7908,
      "mean_token_accuracy": 0.771404180675745,
      "num_tokens": 136956557.0,
      "step": 4291
    },
    {
      "entropy": 0.7566580157727003,
      "epoch": 0.39489134622572614,
      "grad_norm": 0.1563936471939087,
      "learning_rate": 8.684024902628271e-05,
      "loss": 0.7471,
      "mean_token_accuracy": 0.780902124941349,
      "num_tokens": 136988634.0,
      "step": 4292
    },
    {
      "entropy": 0.8042337726801634,
      "epoch": 0.3949833525971674,
      "grad_norm": 0.16008716821670532,
      "learning_rate": 8.683718220014108e-05,
      "loss": 0.796,
      "mean_token_accuracy": 0.767993874847889,
      "num_tokens": 137021385.0,
      "step": 4293
    },
    {
      "entropy": 0.77345060557127,
      "epoch": 0.39507535896860857,
      "grad_norm": 0.14618268609046936,
      "learning_rate": 8.683411537399945e-05,
      "loss": 0.7497,
      "mean_token_accuracy": 0.7838089652359486,
      "num_tokens": 137052873.0,
      "step": 4294
    },
    {
      "entropy": 0.7985240500420332,
      "epoch": 0.3951673653400498,
      "grad_norm": 0.15207110345363617,
      "learning_rate": 8.683104854785783e-05,
      "loss": 0.8017,
      "mean_token_accuracy": 0.7672784924507141,
      "num_tokens": 137084794.0,
      "step": 4295
    },
    {
      "entropy": 0.937651090323925,
      "epoch": 0.395259371711491,
      "grad_norm": 0.16123130917549133,
      "learning_rate": 8.68279817217162e-05,
      "loss": 0.9136,
      "mean_token_accuracy": 0.7391834929585457,
      "num_tokens": 137116598.0,
      "step": 4296
    },
    {
      "entropy": 0.7977062333375216,
      "epoch": 0.39535137808293225,
      "grad_norm": 0.1509864777326584,
      "learning_rate": 8.682491489557458e-05,
      "loss": 0.7906,
      "mean_token_accuracy": 0.771388117223978,
      "num_tokens": 137148665.0,
      "step": 4297
    },
    {
      "entropy": 0.8245925344526768,
      "epoch": 0.3954433844543735,
      "grad_norm": 0.15272919833660126,
      "learning_rate": 8.682184806943294e-05,
      "loss": 0.7901,
      "mean_token_accuracy": 0.7726667374372482,
      "num_tokens": 137181255.0,
      "step": 4298
    },
    {
      "entropy": 0.8287330232560635,
      "epoch": 0.3955353908258147,
      "grad_norm": 0.15585428476333618,
      "learning_rate": 8.681878124329133e-05,
      "loss": 0.7988,
      "mean_token_accuracy": 0.7664231695234776,
      "num_tokens": 137212924.0,
      "step": 4299
    },
    {
      "entropy": 0.8473609127104282,
      "epoch": 0.3956273971972559,
      "grad_norm": 0.15894542634487152,
      "learning_rate": 8.68157144171497e-05,
      "loss": 0.83,
      "mean_token_accuracy": 0.7594563700258732,
      "num_tokens": 137244845.0,
      "step": 4300
    },
    {
      "entropy": 0.7810115776956081,
      "epoch": 0.3957194035686971,
      "grad_norm": 0.1618606001138687,
      "learning_rate": 8.681264759100806e-05,
      "loss": 0.7702,
      "mean_token_accuracy": 0.7758715488016605,
      "num_tokens": 137276824.0,
      "step": 4301
    },
    {
      "entropy": 0.8447426930069923,
      "epoch": 0.39581140994013836,
      "grad_norm": 0.15919001400470734,
      "learning_rate": 8.680958076486644e-05,
      "loss": 0.8469,
      "mean_token_accuracy": 0.7559487149119377,
      "num_tokens": 137308161.0,
      "step": 4302
    },
    {
      "entropy": 0.8000420778989792,
      "epoch": 0.3959034163115796,
      "grad_norm": 0.1614128053188324,
      "learning_rate": 8.680651393872482e-05,
      "loss": 0.7808,
      "mean_token_accuracy": 0.7753312401473522,
      "num_tokens": 137339822.0,
      "step": 4303
    },
    {
      "entropy": 0.8622737731784582,
      "epoch": 0.3959954226830208,
      "grad_norm": 0.16831503808498383,
      "learning_rate": 8.680344711258319e-05,
      "loss": 0.8613,
      "mean_token_accuracy": 0.7474303208291531,
      "num_tokens": 137370746.0,
      "step": 4304
    },
    {
      "entropy": 0.820012154057622,
      "epoch": 0.39608742905446204,
      "grad_norm": 0.16459505259990692,
      "learning_rate": 8.680038028644156e-05,
      "loss": 0.8215,
      "mean_token_accuracy": 0.7650977186858654,
      "num_tokens": 137401755.0,
      "step": 4305
    },
    {
      "entropy": 0.8205808475613594,
      "epoch": 0.3961794354259032,
      "grad_norm": 0.16316480934619904,
      "learning_rate": 8.679731346029994e-05,
      "loss": 0.8112,
      "mean_token_accuracy": 0.7645847611129284,
      "num_tokens": 137433290.0,
      "step": 4306
    },
    {
      "entropy": 0.7468543220311403,
      "epoch": 0.39627144179734447,
      "grad_norm": 0.15221144258975983,
      "learning_rate": 8.679424663415831e-05,
      "loss": 0.7395,
      "mean_token_accuracy": 0.7862758226692677,
      "num_tokens": 137465177.0,
      "step": 4307
    },
    {
      "entropy": 0.7493326682597399,
      "epoch": 0.3963634481687857,
      "grad_norm": 0.15214569866657257,
      "learning_rate": 8.679117980801669e-05,
      "loss": 0.7443,
      "mean_token_accuracy": 0.7795054204761982,
      "num_tokens": 137497466.0,
      "step": 4308
    },
    {
      "entropy": 0.7695858031511307,
      "epoch": 0.3964554545402269,
      "grad_norm": 0.15855616331100464,
      "learning_rate": 8.678811298187506e-05,
      "loss": 0.7658,
      "mean_token_accuracy": 0.7750889882445335,
      "num_tokens": 137529136.0,
      "step": 4309
    },
    {
      "entropy": 0.7066148780286312,
      "epoch": 0.39654746091166815,
      "grad_norm": 0.1577744334936142,
      "learning_rate": 8.678504615573344e-05,
      "loss": 0.7209,
      "mean_token_accuracy": 0.789254929870367,
      "num_tokens": 137561049.0,
      "step": 4310
    },
    {
      "entropy": 0.8105504773557186,
      "epoch": 0.39663946728310934,
      "grad_norm": 0.1580822914838791,
      "learning_rate": 8.678197932959181e-05,
      "loss": 0.8053,
      "mean_token_accuracy": 0.767706710845232,
      "num_tokens": 137593150.0,
      "step": 4311
    },
    {
      "entropy": 0.7754011638462543,
      "epoch": 0.3967314736545506,
      "grad_norm": 0.15686580538749695,
      "learning_rate": 8.677891250345018e-05,
      "loss": 0.7774,
      "mean_token_accuracy": 0.7751853466033936,
      "num_tokens": 137625751.0,
      "step": 4312
    },
    {
      "entropy": 0.754264336079359,
      "epoch": 0.3968234800259918,
      "grad_norm": 0.160162553191185,
      "learning_rate": 8.677584567730856e-05,
      "loss": 0.7424,
      "mean_token_accuracy": 0.7868561781942844,
      "num_tokens": 137657690.0,
      "step": 4313
    },
    {
      "entropy": 0.8559787590056658,
      "epoch": 0.396915486397433,
      "grad_norm": 0.1588020771741867,
      "learning_rate": 8.677277885116694e-05,
      "loss": 0.8561,
      "mean_token_accuracy": 0.7505843974649906,
      "num_tokens": 137689872.0,
      "step": 4314
    },
    {
      "entropy": 0.8546167500317097,
      "epoch": 0.39700749276887426,
      "grad_norm": 0.15839356184005737,
      "learning_rate": 8.676971202502531e-05,
      "loss": 0.8336,
      "mean_token_accuracy": 0.7555923834443092,
      "num_tokens": 137721259.0,
      "step": 4315
    },
    {
      "entropy": 0.7859054952859879,
      "epoch": 0.39709949914031545,
      "grad_norm": 0.16015814244747162,
      "learning_rate": 8.676664519888367e-05,
      "loss": 0.7713,
      "mean_token_accuracy": 0.7730723023414612,
      "num_tokens": 137752217.0,
      "step": 4316
    },
    {
      "entropy": 0.9331096261739731,
      "epoch": 0.3971915055117567,
      "grad_norm": 0.1636391580104828,
      "learning_rate": 8.676357837274204e-05,
      "loss": 0.9224,
      "mean_token_accuracy": 0.7360465787351131,
      "num_tokens": 137784331.0,
      "step": 4317
    },
    {
      "entropy": 0.8093492407351732,
      "epoch": 0.39728351188319794,
      "grad_norm": 0.15774591267108917,
      "learning_rate": 8.676051154660044e-05,
      "loss": 0.7868,
      "mean_token_accuracy": 0.7688788995146751,
      "num_tokens": 137815917.0,
      "step": 4318
    },
    {
      "entropy": 0.8006084896624088,
      "epoch": 0.3973755182546391,
      "grad_norm": 0.16358770430088043,
      "learning_rate": 8.67574447204588e-05,
      "loss": 0.7896,
      "mean_token_accuracy": 0.7703625224530697,
      "num_tokens": 137847458.0,
      "step": 4319
    },
    {
      "entropy": 0.8266694936901331,
      "epoch": 0.39746752462608037,
      "grad_norm": 0.1625903993844986,
      "learning_rate": 8.675437789431717e-05,
      "loss": 0.8288,
      "mean_token_accuracy": 0.7596943713724613,
      "num_tokens": 137879273.0,
      "step": 4320
    },
    {
      "entropy": 0.7928211949765682,
      "epoch": 0.39755953099752156,
      "grad_norm": 0.1588454693555832,
      "learning_rate": 8.675131106817554e-05,
      "loss": 0.7898,
      "mean_token_accuracy": 0.7695720046758652,
      "num_tokens": 137911378.0,
      "step": 4321
    },
    {
      "entropy": 0.7297004200518131,
      "epoch": 0.3976515373689628,
      "grad_norm": 0.157255619764328,
      "learning_rate": 8.674824424203392e-05,
      "loss": 0.7353,
      "mean_token_accuracy": 0.7844336591660976,
      "num_tokens": 137943323.0,
      "step": 4322
    },
    {
      "entropy": 0.8143177945166826,
      "epoch": 0.39774354374040405,
      "grad_norm": 0.16035771369934082,
      "learning_rate": 8.67451774158923e-05,
      "loss": 0.7984,
      "mean_token_accuracy": 0.7683944590389729,
      "num_tokens": 137974958.0,
      "step": 4323
    },
    {
      "entropy": 0.8592742905020714,
      "epoch": 0.39783555011184524,
      "grad_norm": 0.16616401076316833,
      "learning_rate": 8.674211058975067e-05,
      "loss": 0.839,
      "mean_token_accuracy": 0.7563778348267078,
      "num_tokens": 138007105.0,
      "step": 4324
    },
    {
      "entropy": 0.8701001964509487,
      "epoch": 0.3979275564832865,
      "grad_norm": 0.17079131305217743,
      "learning_rate": 8.673904376360904e-05,
      "loss": 0.8825,
      "mean_token_accuracy": 0.751967042684555,
      "num_tokens": 138038289.0,
      "step": 4325
    },
    {
      "entropy": 0.7534818202257156,
      "epoch": 0.39801956285472767,
      "grad_norm": 0.1567857712507248,
      "learning_rate": 8.673597693746742e-05,
      "loss": 0.7289,
      "mean_token_accuracy": 0.7845746502280235,
      "num_tokens": 138069749.0,
      "step": 4326
    },
    {
      "entropy": 0.770831873640418,
      "epoch": 0.3981115692261689,
      "grad_norm": 0.15538786351680756,
      "learning_rate": 8.673291011132579e-05,
      "loss": 0.7557,
      "mean_token_accuracy": 0.7827656641602516,
      "num_tokens": 138102209.0,
      "step": 4327
    },
    {
      "entropy": 0.8461679965257645,
      "epoch": 0.39820357559761016,
      "grad_norm": 0.1614655703306198,
      "learning_rate": 8.672984328518417e-05,
      "loss": 0.8384,
      "mean_token_accuracy": 0.7592774741351604,
      "num_tokens": 138132764.0,
      "step": 4328
    },
    {
      "entropy": 0.7551915813237429,
      "epoch": 0.39829558196905135,
      "grad_norm": 0.15248925983905792,
      "learning_rate": 8.672677645904254e-05,
      "loss": 0.7443,
      "mean_token_accuracy": 0.7850985825061798,
      "num_tokens": 138165392.0,
      "step": 4329
    },
    {
      "entropy": 0.8505535200238228,
      "epoch": 0.3983875883404926,
      "grad_norm": 0.16108348965644836,
      "learning_rate": 8.672370963290092e-05,
      "loss": 0.8524,
      "mean_token_accuracy": 0.752531997859478,
      "num_tokens": 138197036.0,
      "step": 4330
    },
    {
      "entropy": 0.7483100425451994,
      "epoch": 0.3984795947119338,
      "grad_norm": 0.1522454470396042,
      "learning_rate": 8.672064280675929e-05,
      "loss": 0.7392,
      "mean_token_accuracy": 0.785241074860096,
      "num_tokens": 138229521.0,
      "step": 4331
    },
    {
      "entropy": 0.7946486659348011,
      "epoch": 0.398571601083375,
      "grad_norm": 0.15325070917606354,
      "learning_rate": 8.671757598061766e-05,
      "loss": 0.7745,
      "mean_token_accuracy": 0.7760598808526993,
      "num_tokens": 138261681.0,
      "step": 4332
    },
    {
      "entropy": 0.7220129184424877,
      "epoch": 0.39866360745481627,
      "grad_norm": 0.14982889592647552,
      "learning_rate": 8.671450915447604e-05,
      "loss": 0.7212,
      "mean_token_accuracy": 0.7879708893597126,
      "num_tokens": 138293875.0,
      "step": 4333
    },
    {
      "entropy": 0.7530864644795656,
      "epoch": 0.39875561382625746,
      "grad_norm": 0.151821106672287,
      "learning_rate": 8.671144232833442e-05,
      "loss": 0.7565,
      "mean_token_accuracy": 0.7731948047876358,
      "num_tokens": 138326224.0,
      "step": 4334
    },
    {
      "entropy": 0.7817614804953337,
      "epoch": 0.3988476201976987,
      "grad_norm": 0.16419586539268494,
      "learning_rate": 8.670837550219279e-05,
      "loss": 0.7946,
      "mean_token_accuracy": 0.768190298229456,
      "num_tokens": 138357449.0,
      "step": 4335
    },
    {
      "entropy": 0.7034357935190201,
      "epoch": 0.3989396265691399,
      "grad_norm": 0.16165870428085327,
      "learning_rate": 8.670530867605115e-05,
      "loss": 0.6863,
      "mean_token_accuracy": 0.8008225671947002,
      "num_tokens": 138389783.0,
      "step": 4336
    },
    {
      "entropy": 0.8368466980755329,
      "epoch": 0.39903163294058114,
      "grad_norm": 0.16751877963542938,
      "learning_rate": 8.670224184990954e-05,
      "loss": 0.8205,
      "mean_token_accuracy": 0.7670840099453926,
      "num_tokens": 138420920.0,
      "step": 4337
    },
    {
      "entropy": 0.7278641182929277,
      "epoch": 0.3991236393120224,
      "grad_norm": 0.14888767898082733,
      "learning_rate": 8.66991750237679e-05,
      "loss": 0.7144,
      "mean_token_accuracy": 0.7880249805748463,
      "num_tokens": 138452709.0,
      "step": 4338
    },
    {
      "entropy": 0.8111402746289968,
      "epoch": 0.39921564568346357,
      "grad_norm": 0.15661753714084625,
      "learning_rate": 8.669610819762629e-05,
      "loss": 0.7911,
      "mean_token_accuracy": 0.7730852514505386,
      "num_tokens": 138484540.0,
      "step": 4339
    },
    {
      "entropy": 0.8350012861192226,
      "epoch": 0.3993076520549048,
      "grad_norm": 0.1669192910194397,
      "learning_rate": 8.669304137148465e-05,
      "loss": 0.8223,
      "mean_token_accuracy": 0.7658633068203926,
      "num_tokens": 138516567.0,
      "step": 4340
    },
    {
      "entropy": 0.8203798588365316,
      "epoch": 0.399399658426346,
      "grad_norm": 0.16246037185192108,
      "learning_rate": 8.668997454534303e-05,
      "loss": 0.8139,
      "mean_token_accuracy": 0.7564629353582859,
      "num_tokens": 138547967.0,
      "step": 4341
    },
    {
      "entropy": 0.8222418241202831,
      "epoch": 0.39949166479778725,
      "grad_norm": 0.14983804523944855,
      "learning_rate": 8.66869077192014e-05,
      "loss": 0.8249,
      "mean_token_accuracy": 0.7636446543037891,
      "num_tokens": 138580006.0,
      "step": 4342
    },
    {
      "entropy": 0.7522838991135359,
      "epoch": 0.3995836711692285,
      "grad_norm": 0.1591769903898239,
      "learning_rate": 8.668384089305977e-05,
      "loss": 0.7664,
      "mean_token_accuracy": 0.773526132106781,
      "num_tokens": 138611875.0,
      "step": 4343
    },
    {
      "entropy": 0.8121912702918053,
      "epoch": 0.3996756775406697,
      "grad_norm": 0.15853485465049744,
      "learning_rate": 8.668077406691815e-05,
      "loss": 0.8252,
      "mean_token_accuracy": 0.7632131651043892,
      "num_tokens": 138644196.0,
      "step": 4344
    },
    {
      "entropy": 0.8113891836255789,
      "epoch": 0.3997676839121109,
      "grad_norm": 0.15861161053180695,
      "learning_rate": 8.667770724077653e-05,
      "loss": 0.8039,
      "mean_token_accuracy": 0.7616065517067909,
      "num_tokens": 138676404.0,
      "step": 4345
    },
    {
      "entropy": 0.7169296145439148,
      "epoch": 0.3998596902835521,
      "grad_norm": 0.15718917548656464,
      "learning_rate": 8.66746404146349e-05,
      "loss": 0.737,
      "mean_token_accuracy": 0.7828649990260601,
      "num_tokens": 138709106.0,
      "step": 4346
    },
    {
      "entropy": 0.734160402789712,
      "epoch": 0.39995169665499336,
      "grad_norm": 0.1600690633058548,
      "learning_rate": 8.667157358849327e-05,
      "loss": 0.7356,
      "mean_token_accuracy": 0.7876906543970108,
      "num_tokens": 138740953.0,
      "step": 4347
    },
    {
      "entropy": 0.7680148780345917,
      "epoch": 0.4000437030264346,
      "grad_norm": 0.15159720182418823,
      "learning_rate": 8.666850676235164e-05,
      "loss": 0.74,
      "mean_token_accuracy": 0.7796390876173973,
      "num_tokens": 138772576.0,
      "step": 4348
    },
    {
      "entropy": 0.8035288937389851,
      "epoch": 0.4001357093978758,
      "grad_norm": 0.15289051830768585,
      "learning_rate": 8.666543993621002e-05,
      "loss": 0.7726,
      "mean_token_accuracy": 0.7722129039466381,
      "num_tokens": 138803684.0,
      "step": 4349
    },
    {
      "entropy": 0.7351873833686113,
      "epoch": 0.40022771576931704,
      "grad_norm": 0.15200111269950867,
      "learning_rate": 8.66623731100684e-05,
      "loss": 0.7178,
      "mean_token_accuracy": 0.7889161519706249,
      "num_tokens": 138835837.0,
      "step": 4350
    },
    {
      "entropy": 0.8093269243836403,
      "epoch": 0.4003197221407582,
      "grad_norm": 0.16740332543849945,
      "learning_rate": 8.665930628392677e-05,
      "loss": 0.7901,
      "mean_token_accuracy": 0.769905298948288,
      "num_tokens": 138868097.0,
      "step": 4351
    },
    {
      "entropy": 0.7818273957818747,
      "epoch": 0.40041172851219947,
      "grad_norm": 0.14954563975334167,
      "learning_rate": 8.665623945778514e-05,
      "loss": 0.7804,
      "mean_token_accuracy": 0.7757586017251015,
      "num_tokens": 138900751.0,
      "step": 4352
    },
    {
      "entropy": 0.8011466711759567,
      "epoch": 0.4005037348836407,
      "grad_norm": 0.15709596872329712,
      "learning_rate": 8.665317263164352e-05,
      "loss": 0.7915,
      "mean_token_accuracy": 0.7707890272140503,
      "num_tokens": 138933101.0,
      "step": 4353
    },
    {
      "entropy": 0.7234791070222855,
      "epoch": 0.4005957412550819,
      "grad_norm": 0.15680277347564697,
      "learning_rate": 8.665010580550188e-05,
      "loss": 0.694,
      "mean_token_accuracy": 0.7997357398271561,
      "num_tokens": 138964099.0,
      "step": 4354
    },
    {
      "entropy": 0.7901899572461843,
      "epoch": 0.40068774762652315,
      "grad_norm": 0.14949503540992737,
      "learning_rate": 8.664703897936027e-05,
      "loss": 0.7797,
      "mean_token_accuracy": 0.7754409089684486,
      "num_tokens": 138996405.0,
      "step": 4355
    },
    {
      "entropy": 0.8166088368743658,
      "epoch": 0.40077975399796434,
      "grad_norm": 0.15932045876979828,
      "learning_rate": 8.664397215321863e-05,
      "loss": 0.7967,
      "mean_token_accuracy": 0.7657454647123814,
      "num_tokens": 139028357.0,
      "step": 4356
    },
    {
      "entropy": 0.7152461688965559,
      "epoch": 0.4008717603694056,
      "grad_norm": 0.1614648848772049,
      "learning_rate": 8.664090532707702e-05,
      "loss": 0.701,
      "mean_token_accuracy": 0.7940308749675751,
      "num_tokens": 139060095.0,
      "step": 4357
    },
    {
      "entropy": 0.7782885786145926,
      "epoch": 0.4009637667408468,
      "grad_norm": 0.1671127825975418,
      "learning_rate": 8.663783850093538e-05,
      "loss": 0.7848,
      "mean_token_accuracy": 0.7726522460579872,
      "num_tokens": 139092095.0,
      "step": 4358
    },
    {
      "entropy": 0.898818988353014,
      "epoch": 0.401055773112288,
      "grad_norm": 0.1624075472354889,
      "learning_rate": 8.663477167479375e-05,
      "loss": 0.9155,
      "mean_token_accuracy": 0.737066775560379,
      "num_tokens": 139124475.0,
      "step": 4359
    },
    {
      "entropy": 0.856828112155199,
      "epoch": 0.40114777948372926,
      "grad_norm": 0.1574203222990036,
      "learning_rate": 8.663170484865213e-05,
      "loss": 0.8422,
      "mean_token_accuracy": 0.7536935247480869,
      "num_tokens": 139155822.0,
      "step": 4360
    },
    {
      "entropy": 0.8152277041226625,
      "epoch": 0.40123978585517045,
      "grad_norm": 0.1566011756658554,
      "learning_rate": 8.662863802251051e-05,
      "loss": 0.8225,
      "mean_token_accuracy": 0.7585791945457458,
      "num_tokens": 139187371.0,
      "step": 4361
    },
    {
      "entropy": 0.823297169059515,
      "epoch": 0.4013317922266117,
      "grad_norm": 0.15950189530849457,
      "learning_rate": 8.662557119636888e-05,
      "loss": 0.8298,
      "mean_token_accuracy": 0.7632873877882957,
      "num_tokens": 139219502.0,
      "step": 4362
    },
    {
      "entropy": 0.9366223216056824,
      "epoch": 0.40142379859805294,
      "grad_norm": 0.17130129039287567,
      "learning_rate": 8.662250437022725e-05,
      "loss": 0.9342,
      "mean_token_accuracy": 0.7291664034128189,
      "num_tokens": 139250256.0,
      "step": 4363
    },
    {
      "entropy": 0.8843834362924099,
      "epoch": 0.40151580496949413,
      "grad_norm": 0.16095633804798126,
      "learning_rate": 8.661943754408563e-05,
      "loss": 0.8929,
      "mean_token_accuracy": 0.7426904924213886,
      "num_tokens": 139281990.0,
      "step": 4364
    },
    {
      "entropy": 0.8356434013694525,
      "epoch": 0.4016078113409354,
      "grad_norm": 0.15358003973960876,
      "learning_rate": 8.661637071794401e-05,
      "loss": 0.8058,
      "mean_token_accuracy": 0.7667003460228443,
      "num_tokens": 139314261.0,
      "step": 4365
    },
    {
      "entropy": 0.8377423249185085,
      "epoch": 0.40169981771237656,
      "grad_norm": 0.15498553216457367,
      "learning_rate": 8.661330389180238e-05,
      "loss": 0.8306,
      "mean_token_accuracy": 0.7625199258327484,
      "num_tokens": 139346923.0,
      "step": 4366
    },
    {
      "entropy": 0.6710847076028585,
      "epoch": 0.4017918240838178,
      "grad_norm": 0.15641725063323975,
      "learning_rate": 8.661023706566075e-05,
      "loss": 0.656,
      "mean_token_accuracy": 0.806025180965662,
      "num_tokens": 139379488.0,
      "step": 4367
    },
    {
      "entropy": 0.7980338409543037,
      "epoch": 0.40188383045525905,
      "grad_norm": 0.15308572351932526,
      "learning_rate": 8.660717023951913e-05,
      "loss": 0.7839,
      "mean_token_accuracy": 0.7716925702989101,
      "num_tokens": 139411549.0,
      "step": 4368
    },
    {
      "entropy": 0.8410529643297195,
      "epoch": 0.40197583682670024,
      "grad_norm": 0.1620626449584961,
      "learning_rate": 8.66041034133775e-05,
      "loss": 0.8173,
      "mean_token_accuracy": 0.7642960585653782,
      "num_tokens": 139443055.0,
      "step": 4369
    },
    {
      "entropy": 0.8511195350438356,
      "epoch": 0.4020678431981415,
      "grad_norm": 0.1533229947090149,
      "learning_rate": 8.660103658723588e-05,
      "loss": 0.8262,
      "mean_token_accuracy": 0.7589772045612335,
      "num_tokens": 139474675.0,
      "step": 4370
    },
    {
      "entropy": 0.7632060684263706,
      "epoch": 0.4021598495695827,
      "grad_norm": 0.1501828283071518,
      "learning_rate": 8.659796976109425e-05,
      "loss": 0.7306,
      "mean_token_accuracy": 0.7851455770432949,
      "num_tokens": 139507434.0,
      "step": 4371
    },
    {
      "entropy": 0.7548919897526503,
      "epoch": 0.4022518559410239,
      "grad_norm": 0.15556472539901733,
      "learning_rate": 8.659490293495263e-05,
      "loss": 0.733,
      "mean_token_accuracy": 0.7884932719171047,
      "num_tokens": 139539464.0,
      "step": 4372
    },
    {
      "entropy": 0.7639422863721848,
      "epoch": 0.40234386231246516,
      "grad_norm": 0.16604238748550415,
      "learning_rate": 8.6591836108811e-05,
      "loss": 0.7806,
      "mean_token_accuracy": 0.7796287871897221,
      "num_tokens": 139571503.0,
      "step": 4373
    },
    {
      "entropy": 0.8212047293782234,
      "epoch": 0.40243586868390635,
      "grad_norm": 0.16611728072166443,
      "learning_rate": 8.658876928266936e-05,
      "loss": 0.8379,
      "mean_token_accuracy": 0.7611560299992561,
      "num_tokens": 139603577.0,
      "step": 4374
    },
    {
      "entropy": 0.7269526608288288,
      "epoch": 0.4025278750553476,
      "grad_norm": 0.16269703209400177,
      "learning_rate": 8.658570245652775e-05,
      "loss": 0.738,
      "mean_token_accuracy": 0.7864685878157616,
      "num_tokens": 139636122.0,
      "step": 4375
    },
    {
      "entropy": 0.7711403258144855,
      "epoch": 0.4026198814267888,
      "grad_norm": 0.1556607037782669,
      "learning_rate": 8.658263563038613e-05,
      "loss": 0.7789,
      "mean_token_accuracy": 0.7728221789002419,
      "num_tokens": 139667336.0,
      "step": 4376
    },
    {
      "entropy": 0.8380633220076561,
      "epoch": 0.40271188779823003,
      "grad_norm": 0.1631087362766266,
      "learning_rate": 8.65795688042445e-05,
      "loss": 0.8537,
      "mean_token_accuracy": 0.7554492801427841,
      "num_tokens": 139699293.0,
      "step": 4377
    },
    {
      "entropy": 0.7276904974132776,
      "epoch": 0.4028038941696713,
      "grad_norm": 0.15230897068977356,
      "learning_rate": 8.657650197810286e-05,
      "loss": 0.7195,
      "mean_token_accuracy": 0.7878586314618587,
      "num_tokens": 139731212.0,
      "step": 4378
    },
    {
      "entropy": 0.8463601693511009,
      "epoch": 0.40289590054111246,
      "grad_norm": 0.16354668140411377,
      "learning_rate": 8.657343515196123e-05,
      "loss": 0.84,
      "mean_token_accuracy": 0.7537868209183216,
      "num_tokens": 139763439.0,
      "step": 4379
    },
    {
      "entropy": 0.8674339726567268,
      "epoch": 0.4029879069125537,
      "grad_norm": 0.1607319414615631,
      "learning_rate": 8.657036832581961e-05,
      "loss": 0.854,
      "mean_token_accuracy": 0.754186425358057,
      "num_tokens": 139794530.0,
      "step": 4380
    },
    {
      "entropy": 0.874748582020402,
      "epoch": 0.4030799132839949,
      "grad_norm": 0.15621648728847504,
      "learning_rate": 8.656730149967799e-05,
      "loss": 0.8384,
      "mean_token_accuracy": 0.7574009224772453,
      "num_tokens": 139826587.0,
      "step": 4381
    },
    {
      "entropy": 0.7991771008819342,
      "epoch": 0.40317191965543614,
      "grad_norm": 0.154558926820755,
      "learning_rate": 8.656423467353636e-05,
      "loss": 0.7814,
      "mean_token_accuracy": 0.7703753225505352,
      "num_tokens": 139858899.0,
      "step": 4382
    },
    {
      "entropy": 0.7530431412160397,
      "epoch": 0.4032639260268774,
      "grad_norm": 0.15306991338729858,
      "learning_rate": 8.656116784739473e-05,
      "loss": 0.741,
      "mean_token_accuracy": 0.782896164804697,
      "num_tokens": 139890923.0,
      "step": 4383
    },
    {
      "entropy": 0.7990904599428177,
      "epoch": 0.4033559323983186,
      "grad_norm": 0.15840773284435272,
      "learning_rate": 8.655810102125311e-05,
      "loss": 0.7775,
      "mean_token_accuracy": 0.775368720293045,
      "num_tokens": 139923536.0,
      "step": 4384
    },
    {
      "entropy": 0.8068835251033306,
      "epoch": 0.4034479387697598,
      "grad_norm": 0.16360025107860565,
      "learning_rate": 8.655503419511148e-05,
      "loss": 0.7807,
      "mean_token_accuracy": 0.7728019952774048,
      "num_tokens": 139954054.0,
      "step": 4385
    },
    {
      "entropy": 0.7142554558813572,
      "epoch": 0.403539945141201,
      "grad_norm": 0.15966735780239105,
      "learning_rate": 8.655196736896986e-05,
      "loss": 0.719,
      "mean_token_accuracy": 0.7924164421856403,
      "num_tokens": 139986466.0,
      "step": 4386
    },
    {
      "entropy": 0.7812010534107685,
      "epoch": 0.40363195151264225,
      "grad_norm": 0.1539911925792694,
      "learning_rate": 8.654890054282823e-05,
      "loss": 0.7581,
      "mean_token_accuracy": 0.7777816094458103,
      "num_tokens": 140018757.0,
      "step": 4387
    },
    {
      "entropy": 0.7832820136100054,
      "epoch": 0.4037239578840835,
      "grad_norm": 0.16208499670028687,
      "learning_rate": 8.654583371668661e-05,
      "loss": 0.7933,
      "mean_token_accuracy": 0.7677854746580124,
      "num_tokens": 140051229.0,
      "step": 4388
    },
    {
      "entropy": 0.8392529226839542,
      "epoch": 0.4038159642555247,
      "grad_norm": 0.1629580855369568,
      "learning_rate": 8.654276689054498e-05,
      "loss": 0.8463,
      "mean_token_accuracy": 0.7555507831275463,
      "num_tokens": 140083055.0,
      "step": 4389
    },
    {
      "entropy": 0.8610566146671772,
      "epoch": 0.40390797062696593,
      "grad_norm": 0.16852208971977234,
      "learning_rate": 8.653970006440334e-05,
      "loss": 0.8751,
      "mean_token_accuracy": 0.7478588446974754,
      "num_tokens": 140114323.0,
      "step": 4390
    },
    {
      "entropy": 0.8258617594838142,
      "epoch": 0.4039999769984071,
      "grad_norm": 0.1572355031967163,
      "learning_rate": 8.653663323826173e-05,
      "loss": 0.8065,
      "mean_token_accuracy": 0.7665948644280434,
      "num_tokens": 140146514.0,
      "step": 4391
    },
    {
      "entropy": 0.8448238484561443,
      "epoch": 0.40409198336984836,
      "grad_norm": 0.15524406731128693,
      "learning_rate": 8.653356641212011e-05,
      "loss": 0.8214,
      "mean_token_accuracy": 0.7622006013989449,
      "num_tokens": 140178820.0,
      "step": 4392
    },
    {
      "entropy": 0.7447648253291845,
      "epoch": 0.4041839897412896,
      "grad_norm": 0.15681302547454834,
      "learning_rate": 8.653049958597848e-05,
      "loss": 0.7081,
      "mean_token_accuracy": 0.7899158671498299,
      "num_tokens": 140210336.0,
      "step": 4393
    },
    {
      "entropy": 0.8720273859798908,
      "epoch": 0.4042759961127308,
      "grad_norm": 0.16244713962078094,
      "learning_rate": 8.652743275983684e-05,
      "loss": 0.8817,
      "mean_token_accuracy": 0.7403237372636795,
      "num_tokens": 140242820.0,
      "step": 4394
    },
    {
      "entropy": 0.8931030556559563,
      "epoch": 0.40436800248417204,
      "grad_norm": 0.15651057660579681,
      "learning_rate": 8.652436593369522e-05,
      "loss": 0.868,
      "mean_token_accuracy": 0.7480721473693848,
      "num_tokens": 140274087.0,
      "step": 4395
    },
    {
      "entropy": 0.8616100549697876,
      "epoch": 0.40446000885561323,
      "grad_norm": 0.15598443150520325,
      "learning_rate": 8.65212991075536e-05,
      "loss": 0.8472,
      "mean_token_accuracy": 0.7566724345088005,
      "num_tokens": 140305391.0,
      "step": 4396
    },
    {
      "entropy": 0.8557443954050541,
      "epoch": 0.4045520152270545,
      "grad_norm": 0.16841615736484528,
      "learning_rate": 8.651823228141197e-05,
      "loss": 0.852,
      "mean_token_accuracy": 0.7599308714270592,
      "num_tokens": 140337456.0,
      "step": 4397
    },
    {
      "entropy": 0.8096117936074734,
      "epoch": 0.4046440215984957,
      "grad_norm": 0.1635141372680664,
      "learning_rate": 8.651516545527034e-05,
      "loss": 0.8073,
      "mean_token_accuracy": 0.767000712454319,
      "num_tokens": 140368839.0,
      "step": 4398
    },
    {
      "entropy": 0.7693778425455093,
      "epoch": 0.4047360279699369,
      "grad_norm": 0.16135410964488983,
      "learning_rate": 8.651209862912872e-05,
      "loss": 0.7605,
      "mean_token_accuracy": 0.7752859182655811,
      "num_tokens": 140399639.0,
      "step": 4399
    },
    {
      "entropy": 0.799664419144392,
      "epoch": 0.40482803434137815,
      "grad_norm": 0.1578240990638733,
      "learning_rate": 8.650903180298709e-05,
      "loss": 0.8087,
      "mean_token_accuracy": 0.7661591097712517,
      "num_tokens": 140432346.0,
      "step": 4400
    },
    {
      "entropy": 0.7535984646528959,
      "epoch": 0.40492004071281934,
      "grad_norm": 0.15361344814300537,
      "learning_rate": 8.650596497684547e-05,
      "loss": 0.7441,
      "mean_token_accuracy": 0.7805181406438351,
      "num_tokens": 140463906.0,
      "step": 4401
    },
    {
      "entropy": 0.8459362555295229,
      "epoch": 0.4050120470842606,
      "grad_norm": 0.16780316829681396,
      "learning_rate": 8.650289815070384e-05,
      "loss": 0.8449,
      "mean_token_accuracy": 0.7589607797563076,
      "num_tokens": 140495386.0,
      "step": 4402
    },
    {
      "entropy": 0.7918718364089727,
      "epoch": 0.40510405345570183,
      "grad_norm": 0.1662161499261856,
      "learning_rate": 8.649983132456222e-05,
      "loss": 0.7838,
      "mean_token_accuracy": 0.7718579657375813,
      "num_tokens": 140526714.0,
      "step": 4403
    },
    {
      "entropy": 0.7667257878929377,
      "epoch": 0.405196059827143,
      "grad_norm": 0.15810644626617432,
      "learning_rate": 8.649676449842059e-05,
      "loss": 0.7324,
      "mean_token_accuracy": 0.7862285822629929,
      "num_tokens": 140558435.0,
      "step": 4404
    },
    {
      "entropy": 0.7377495765686035,
      "epoch": 0.40528806619858426,
      "grad_norm": 0.15324118733406067,
      "learning_rate": 8.649369767227896e-05,
      "loss": 0.7326,
      "mean_token_accuracy": 0.7865590341389179,
      "num_tokens": 140590914.0,
      "step": 4405
    },
    {
      "entropy": 0.7309749200940132,
      "epoch": 0.40538007257002545,
      "grad_norm": 0.153891459107399,
      "learning_rate": 8.649063084613734e-05,
      "loss": 0.7066,
      "mean_token_accuracy": 0.7905550599098206,
      "num_tokens": 140622549.0,
      "step": 4406
    },
    {
      "entropy": 0.8043077792972326,
      "epoch": 0.4054720789414667,
      "grad_norm": 0.15959028899669647,
      "learning_rate": 8.648756401999572e-05,
      "loss": 0.8028,
      "mean_token_accuracy": 0.7678774110972881,
      "num_tokens": 140653791.0,
      "step": 4407
    },
    {
      "entropy": 0.6953172907233238,
      "epoch": 0.40556408531290794,
      "grad_norm": 0.15356136858463287,
      "learning_rate": 8.648449719385409e-05,
      "loss": 0.6764,
      "mean_token_accuracy": 0.7992066331207752,
      "num_tokens": 140685933.0,
      "step": 4408
    },
    {
      "entropy": 0.8016958180814981,
      "epoch": 0.40565609168434913,
      "grad_norm": 0.1597042977809906,
      "learning_rate": 8.648143036771246e-05,
      "loss": 0.8183,
      "mean_token_accuracy": 0.7641748376190662,
      "num_tokens": 140718473.0,
      "step": 4409
    },
    {
      "entropy": 0.702772244811058,
      "epoch": 0.4057480980557904,
      "grad_norm": 0.1660853773355484,
      "learning_rate": 8.647836354157082e-05,
      "loss": 0.721,
      "mean_token_accuracy": 0.7909088432788849,
      "num_tokens": 140750967.0,
      "step": 4410
    },
    {
      "entropy": 0.8495804965496063,
      "epoch": 0.40584010442723156,
      "grad_norm": 0.15766669809818268,
      "learning_rate": 8.64752967154292e-05,
      "loss": 0.8481,
      "mean_token_accuracy": 0.7529854439198971,
      "num_tokens": 140783596.0,
      "step": 4411
    },
    {
      "entropy": 0.7856400441378355,
      "epoch": 0.4059321107986728,
      "grad_norm": 0.15926417708396912,
      "learning_rate": 8.647222988928759e-05,
      "loss": 0.7781,
      "mean_token_accuracy": 0.773051880300045,
      "num_tokens": 140815464.0,
      "step": 4412
    },
    {
      "entropy": 0.8292854055762291,
      "epoch": 0.40602411717011405,
      "grad_norm": 0.16010980308055878,
      "learning_rate": 8.646916306314595e-05,
      "loss": 0.8308,
      "mean_token_accuracy": 0.7566285394132137,
      "num_tokens": 140846860.0,
      "step": 4413
    },
    {
      "entropy": 0.7906942293047905,
      "epoch": 0.40611612354155524,
      "grad_norm": 0.16458505392074585,
      "learning_rate": 8.646609623700432e-05,
      "loss": 0.7887,
      "mean_token_accuracy": 0.7710746452212334,
      "num_tokens": 140878690.0,
      "step": 4414
    },
    {
      "entropy": 0.7755707073956728,
      "epoch": 0.4062081299129965,
      "grad_norm": 0.15910686552524567,
      "learning_rate": 8.64630294108627e-05,
      "loss": 0.777,
      "mean_token_accuracy": 0.7752445228397846,
      "num_tokens": 140910747.0,
      "step": 4415
    },
    {
      "entropy": 0.8846572302281857,
      "epoch": 0.4063001362844377,
      "grad_norm": 0.16207630932331085,
      "learning_rate": 8.645996258472107e-05,
      "loss": 0.8993,
      "mean_token_accuracy": 0.7408503852784634,
      "num_tokens": 140943036.0,
      "step": 4416
    },
    {
      "entropy": 0.8487497866153717,
      "epoch": 0.4063921426558789,
      "grad_norm": 0.16231302917003632,
      "learning_rate": 8.645689575857945e-05,
      "loss": 0.839,
      "mean_token_accuracy": 0.7565212175250053,
      "num_tokens": 140974592.0,
      "step": 4417
    },
    {
      "entropy": 0.81348673440516,
      "epoch": 0.40648414902732016,
      "grad_norm": 0.1637900173664093,
      "learning_rate": 8.645382893243782e-05,
      "loss": 0.8181,
      "mean_token_accuracy": 0.7659032456576824,
      "num_tokens": 141006514.0,
      "step": 4418
    },
    {
      "entropy": 0.8530094493180513,
      "epoch": 0.40657615539876135,
      "grad_norm": 0.1574660837650299,
      "learning_rate": 8.64507621062962e-05,
      "loss": 0.8181,
      "mean_token_accuracy": 0.7609489448368549,
      "num_tokens": 141037357.0,
      "step": 4419
    },
    {
      "entropy": 0.8112359046936035,
      "epoch": 0.4066681617702026,
      "grad_norm": 0.15547984838485718,
      "learning_rate": 8.644769528015457e-05,
      "loss": 0.8002,
      "mean_token_accuracy": 0.7652879916131496,
      "num_tokens": 141069297.0,
      "step": 4420
    },
    {
      "entropy": 0.814957071095705,
      "epoch": 0.4067601681416438,
      "grad_norm": 0.15996059775352478,
      "learning_rate": 8.644462845401294e-05,
      "loss": 0.7878,
      "mean_token_accuracy": 0.767077375203371,
      "num_tokens": 141100877.0,
      "step": 4421
    },
    {
      "entropy": 0.8584911096841097,
      "epoch": 0.40685217451308503,
      "grad_norm": 0.1557500958442688,
      "learning_rate": 8.644156162787132e-05,
      "loss": 0.8324,
      "mean_token_accuracy": 0.764381255954504,
      "num_tokens": 141132394.0,
      "step": 4422
    },
    {
      "entropy": 0.7907221205532551,
      "epoch": 0.4069441808845263,
      "grad_norm": 0.1539715975522995,
      "learning_rate": 8.64384948017297e-05,
      "loss": 0.7757,
      "mean_token_accuracy": 0.7714317068457603,
      "num_tokens": 141164845.0,
      "step": 4423
    },
    {
      "entropy": 0.798371834680438,
      "epoch": 0.40703618725596746,
      "grad_norm": 0.15604925155639648,
      "learning_rate": 8.643542797558807e-05,
      "loss": 0.7696,
      "mean_token_accuracy": 0.7783607468008995,
      "num_tokens": 141197453.0,
      "step": 4424
    },
    {
      "entropy": 0.8110921010375023,
      "epoch": 0.4071281936274087,
      "grad_norm": 0.16168037056922913,
      "learning_rate": 8.643236114944644e-05,
      "loss": 0.8162,
      "mean_token_accuracy": 0.76149046048522,
      "num_tokens": 141229987.0,
      "step": 4425
    },
    {
      "entropy": 0.8156020194292068,
      "epoch": 0.4072201999988499,
      "grad_norm": 0.16232316195964813,
      "learning_rate": 8.642929432330482e-05,
      "loss": 0.8246,
      "mean_token_accuracy": 0.762397114187479,
      "num_tokens": 141262159.0,
      "step": 4426
    },
    {
      "entropy": 0.7790936883538961,
      "epoch": 0.40731220637029114,
      "grad_norm": 0.160179004073143,
      "learning_rate": 8.642622749716319e-05,
      "loss": 0.7657,
      "mean_token_accuracy": 0.7771575339138508,
      "num_tokens": 141293935.0,
      "step": 4427
    },
    {
      "entropy": 0.7660750858485699,
      "epoch": 0.4074042127417324,
      "grad_norm": 0.1555587500333786,
      "learning_rate": 8.642316067102157e-05,
      "loss": 0.7526,
      "mean_token_accuracy": 0.77826377004385,
      "num_tokens": 141325719.0,
      "step": 4428
    },
    {
      "entropy": 0.7566420994699001,
      "epoch": 0.4074962191131736,
      "grad_norm": 0.15105167031288147,
      "learning_rate": 8.642009384487994e-05,
      "loss": 0.7496,
      "mean_token_accuracy": 0.7799227237701416,
      "num_tokens": 141357617.0,
      "step": 4429
    },
    {
      "entropy": 0.7919972352683544,
      "epoch": 0.4075882254846148,
      "grad_norm": 0.16174845397472382,
      "learning_rate": 8.641702701873832e-05,
      "loss": 0.8051,
      "mean_token_accuracy": 0.7682539783418179,
      "num_tokens": 141390157.0,
      "step": 4430
    },
    {
      "entropy": 0.7196703441441059,
      "epoch": 0.407680231856056,
      "grad_norm": 0.1548367440700531,
      "learning_rate": 8.641396019259668e-05,
      "loss": 0.709,
      "mean_token_accuracy": 0.788318607956171,
      "num_tokens": 141421525.0,
      "step": 4431
    },
    {
      "entropy": 0.810057956725359,
      "epoch": 0.40777223822749725,
      "grad_norm": 0.1695951372385025,
      "learning_rate": 8.641089336645505e-05,
      "loss": 0.8002,
      "mean_token_accuracy": 0.7691741995513439,
      "num_tokens": 141453070.0,
      "step": 4432
    },
    {
      "entropy": 0.7625819928944111,
      "epoch": 0.4078642445989385,
      "grad_norm": 0.15615350008010864,
      "learning_rate": 8.640782654031343e-05,
      "loss": 0.7474,
      "mean_token_accuracy": 0.7791758738458157,
      "num_tokens": 141485489.0,
      "step": 4433
    },
    {
      "entropy": 0.7726743742823601,
      "epoch": 0.4079562509703797,
      "grad_norm": 0.16402466595172882,
      "learning_rate": 8.640475971417182e-05,
      "loss": 0.7989,
      "mean_token_accuracy": 0.7669868655502796,
      "num_tokens": 141517033.0,
      "step": 4434
    },
    {
      "entropy": 0.7887890134006739,
      "epoch": 0.40804825734182093,
      "grad_norm": 0.15921424329280853,
      "learning_rate": 8.640169288803018e-05,
      "loss": 0.7702,
      "mean_token_accuracy": 0.7757945694029331,
      "num_tokens": 141548409.0,
      "step": 4435
    },
    {
      "entropy": 0.9329372271895409,
      "epoch": 0.4081402637132621,
      "grad_norm": 0.16700181365013123,
      "learning_rate": 8.639862606188855e-05,
      "loss": 0.9567,
      "mean_token_accuracy": 0.7301460392773151,
      "num_tokens": 141580862.0,
      "step": 4436
    },
    {
      "entropy": 0.6948123220354319,
      "epoch": 0.40823227008470336,
      "grad_norm": 0.15304003655910492,
      "learning_rate": 8.639555923574692e-05,
      "loss": 0.6786,
      "mean_token_accuracy": 0.8021107725799084,
      "num_tokens": 141613085.0,
      "step": 4437
    },
    {
      "entropy": 0.8161844778805971,
      "epoch": 0.4083242764561446,
      "grad_norm": 0.15728680789470673,
      "learning_rate": 8.639249240960531e-05,
      "loss": 0.7895,
      "mean_token_accuracy": 0.7682610228657722,
      "num_tokens": 141643808.0,
      "step": 4438
    },
    {
      "entropy": 0.8301820047199726,
      "epoch": 0.4084162828275858,
      "grad_norm": 0.15470553934574127,
      "learning_rate": 8.638942558346368e-05,
      "loss": 0.8201,
      "mean_token_accuracy": 0.7643483802676201,
      "num_tokens": 141676157.0,
      "step": 4439
    },
    {
      "entropy": 0.7751942370086908,
      "epoch": 0.40850828919902704,
      "grad_norm": 0.15568998456001282,
      "learning_rate": 8.638635875732205e-05,
      "loss": 0.7637,
      "mean_token_accuracy": 0.7822869941592216,
      "num_tokens": 141708606.0,
      "step": 4440
    },
    {
      "entropy": 0.7915365658700466,
      "epoch": 0.40860029557046823,
      "grad_norm": 0.16705860197544098,
      "learning_rate": 8.638329193118042e-05,
      "loss": 0.8064,
      "mean_token_accuracy": 0.7688088715076447,
      "num_tokens": 141739206.0,
      "step": 4441
    },
    {
      "entropy": 0.865421686321497,
      "epoch": 0.4086923019419095,
      "grad_norm": 0.1614251285791397,
      "learning_rate": 8.63802251050388e-05,
      "loss": 0.8571,
      "mean_token_accuracy": 0.7511120177805424,
      "num_tokens": 141770978.0,
      "step": 4442
    },
    {
      "entropy": 0.7717647161334753,
      "epoch": 0.4087843083133507,
      "grad_norm": 0.15876436233520508,
      "learning_rate": 8.637715827889718e-05,
      "loss": 0.746,
      "mean_token_accuracy": 0.7820917665958405,
      "num_tokens": 141801692.0,
      "step": 4443
    },
    {
      "entropy": 0.8512538466602564,
      "epoch": 0.4088763146847919,
      "grad_norm": 0.15755128860473633,
      "learning_rate": 8.637409145275555e-05,
      "loss": 0.8286,
      "mean_token_accuracy": 0.7579855173826218,
      "num_tokens": 141833621.0,
      "step": 4444
    },
    {
      "entropy": 0.9346518963575363,
      "epoch": 0.40896832105623315,
      "grad_norm": 0.16812890768051147,
      "learning_rate": 8.637102462661392e-05,
      "loss": 0.9452,
      "mean_token_accuracy": 0.732193510979414,
      "num_tokens": 141864629.0,
      "step": 4445
    },
    {
      "entropy": 0.7931847050786018,
      "epoch": 0.40906032742767434,
      "grad_norm": 0.1558278352022171,
      "learning_rate": 8.63679578004723e-05,
      "loss": 0.7642,
      "mean_token_accuracy": 0.7759575694799423,
      "num_tokens": 141895955.0,
      "step": 4446
    },
    {
      "entropy": 0.7662760373204947,
      "epoch": 0.4091523337991156,
      "grad_norm": 0.1502440869808197,
      "learning_rate": 8.636489097433067e-05,
      "loss": 0.768,
      "mean_token_accuracy": 0.7751294448971748,
      "num_tokens": 141928326.0,
      "step": 4447
    },
    {
      "entropy": 0.8639170415699482,
      "epoch": 0.40924434017055683,
      "grad_norm": 0.16111339628696442,
      "learning_rate": 8.636182414818905e-05,
      "loss": 0.8737,
      "mean_token_accuracy": 0.7456318736076355,
      "num_tokens": 141960291.0,
      "step": 4448
    },
    {
      "entropy": 0.841278214007616,
      "epoch": 0.409336346541998,
      "grad_norm": 0.16120018064975739,
      "learning_rate": 8.635875732204741e-05,
      "loss": 0.8449,
      "mean_token_accuracy": 0.756111066788435,
      "num_tokens": 141992400.0,
      "step": 4449
    },
    {
      "entropy": 0.8230916727334261,
      "epoch": 0.40942835291343926,
      "grad_norm": 0.16274812817573547,
      "learning_rate": 8.63556904959058e-05,
      "loss": 0.8259,
      "mean_token_accuracy": 0.7600231841206551,
      "num_tokens": 142024297.0,
      "step": 4450
    },
    {
      "entropy": 0.7310503367334604,
      "epoch": 0.40952035928488045,
      "grad_norm": 0.1569739431142807,
      "learning_rate": 8.635262366976416e-05,
      "loss": 0.7285,
      "mean_token_accuracy": 0.7884849384427071,
      "num_tokens": 142056662.0,
      "step": 4451
    },
    {
      "entropy": 0.8406836800277233,
      "epoch": 0.4096123656563217,
      "grad_norm": 0.15605272352695465,
      "learning_rate": 8.634955684362253e-05,
      "loss": 0.8343,
      "mean_token_accuracy": 0.7578239478170872,
      "num_tokens": 142089303.0,
      "step": 4452
    },
    {
      "entropy": 0.6994310673326254,
      "epoch": 0.40970437202776294,
      "grad_norm": 0.15623538196086884,
      "learning_rate": 8.634649001748091e-05,
      "loss": 0.6998,
      "mean_token_accuracy": 0.7943422794342041,
      "num_tokens": 142120972.0,
      "step": 4453
    },
    {
      "entropy": 0.7687908485531807,
      "epoch": 0.40979637839920413,
      "grad_norm": 0.15554869174957275,
      "learning_rate": 8.63434231913393e-05,
      "loss": 0.7659,
      "mean_token_accuracy": 0.7805722877383232,
      "num_tokens": 142153041.0,
      "step": 4454
    },
    {
      "entropy": 0.7145817615091801,
      "epoch": 0.4098883847706454,
      "grad_norm": 0.16439685225486755,
      "learning_rate": 8.634035636519766e-05,
      "loss": 0.7235,
      "mean_token_accuracy": 0.7925615310668945,
      "num_tokens": 142185663.0,
      "step": 4455
    },
    {
      "entropy": 0.8337932247668505,
      "epoch": 0.40998039114208656,
      "grad_norm": 0.1622476875782013,
      "learning_rate": 8.633728953905603e-05,
      "loss": 0.8101,
      "mean_token_accuracy": 0.7659592442214489,
      "num_tokens": 142216926.0,
      "step": 4456
    },
    {
      "entropy": 0.9399528801441193,
      "epoch": 0.4100723975135278,
      "grad_norm": 0.16711848974227905,
      "learning_rate": 8.633422271291441e-05,
      "loss": 0.9199,
      "mean_token_accuracy": 0.7328242510557175,
      "num_tokens": 142248984.0,
      "step": 4457
    },
    {
      "entropy": 0.7629917785525322,
      "epoch": 0.41016440388496905,
      "grad_norm": 0.15555231273174286,
      "learning_rate": 8.633115588677278e-05,
      "loss": 0.7509,
      "mean_token_accuracy": 0.7861757166683674,
      "num_tokens": 142280810.0,
      "step": 4458
    },
    {
      "entropy": 0.7829546481370926,
      "epoch": 0.41025641025641024,
      "grad_norm": 0.1643492728471756,
      "learning_rate": 8.632808906063116e-05,
      "loss": 0.7806,
      "mean_token_accuracy": 0.77272704616189,
      "num_tokens": 142311857.0,
      "step": 4459
    },
    {
      "entropy": 0.8448965847492218,
      "epoch": 0.4103484166278515,
      "grad_norm": 0.15490096807479858,
      "learning_rate": 8.632502223448953e-05,
      "loss": 0.819,
      "mean_token_accuracy": 0.7602185644209385,
      "num_tokens": 142343841.0,
      "step": 4460
    },
    {
      "entropy": 0.8499212004244328,
      "epoch": 0.4104404229992927,
      "grad_norm": 0.16372014582157135,
      "learning_rate": 8.632195540834791e-05,
      "loss": 0.8594,
      "mean_token_accuracy": 0.7490151524543762,
      "num_tokens": 142375146.0,
      "step": 4461
    },
    {
      "entropy": 0.7642384711652994,
      "epoch": 0.4105324293707339,
      "grad_norm": 0.15880227088928223,
      "learning_rate": 8.631888858220628e-05,
      "loss": 0.7612,
      "mean_token_accuracy": 0.7778050191700459,
      "num_tokens": 142407551.0,
      "step": 4462
    },
    {
      "entropy": 0.7772066071629524,
      "epoch": 0.41062443574217516,
      "grad_norm": 0.1589951515197754,
      "learning_rate": 8.631582175606465e-05,
      "loss": 0.7653,
      "mean_token_accuracy": 0.7732919342815876,
      "num_tokens": 142438823.0,
      "step": 4463
    },
    {
      "entropy": 0.7889354079961777,
      "epoch": 0.41071644211361635,
      "grad_norm": 0.15745997428894043,
      "learning_rate": 8.631275492992303e-05,
      "loss": 0.7541,
      "mean_token_accuracy": 0.778324194252491,
      "num_tokens": 142470458.0,
      "step": 4464
    },
    {
      "entropy": 0.8007081858813763,
      "epoch": 0.4108084484850576,
      "grad_norm": 0.16317686438560486,
      "learning_rate": 8.630968810378141e-05,
      "loss": 0.8051,
      "mean_token_accuracy": 0.7674171887338161,
      "num_tokens": 142501913.0,
      "step": 4465
    },
    {
      "entropy": 0.7459885701537132,
      "epoch": 0.4109004548564988,
      "grad_norm": 0.16091863811016083,
      "learning_rate": 8.630662127763978e-05,
      "loss": 0.7488,
      "mean_token_accuracy": 0.7804193794727325,
      "num_tokens": 142533497.0,
      "step": 4466
    },
    {
      "entropy": 0.8627739250659943,
      "epoch": 0.41099246122794003,
      "grad_norm": 0.16289503872394562,
      "learning_rate": 8.630355445149815e-05,
      "loss": 0.869,
      "mean_token_accuracy": 0.7490286603569984,
      "num_tokens": 142565384.0,
      "step": 4467
    },
    {
      "entropy": 0.7930417116731405,
      "epoch": 0.4110844675993813,
      "grad_norm": 0.15654434263706207,
      "learning_rate": 8.630048762535651e-05,
      "loss": 0.7969,
      "mean_token_accuracy": 0.764485452324152,
      "num_tokens": 142597316.0,
      "step": 4468
    },
    {
      "entropy": 0.8274433929473162,
      "epoch": 0.41117647397082246,
      "grad_norm": 0.15868796408176422,
      "learning_rate": 8.62974207992149e-05,
      "loss": 0.818,
      "mean_token_accuracy": 0.7668826654553413,
      "num_tokens": 142629168.0,
      "step": 4469
    },
    {
      "entropy": 0.7188860010355711,
      "epoch": 0.4112684803422637,
      "grad_norm": 0.15261772274971008,
      "learning_rate": 8.629435397307328e-05,
      "loss": 0.6944,
      "mean_token_accuracy": 0.7935265153646469,
      "num_tokens": 142661634.0,
      "step": 4470
    },
    {
      "entropy": 0.854890676215291,
      "epoch": 0.4113604867137049,
      "grad_norm": 0.1672695428133011,
      "learning_rate": 8.629128714693164e-05,
      "loss": 0.8614,
      "mean_token_accuracy": 0.7527778595685959,
      "num_tokens": 142693648.0,
      "step": 4471
    },
    {
      "entropy": 0.878272533416748,
      "epoch": 0.41145249308514614,
      "grad_norm": 0.16192686557769775,
      "learning_rate": 8.628822032079001e-05,
      "loss": 0.8524,
      "mean_token_accuracy": 0.7535299062728882,
      "num_tokens": 142725265.0,
      "step": 4472
    },
    {
      "entropy": 0.7458092048764229,
      "epoch": 0.4115444994565874,
      "grad_norm": 0.14748883247375488,
      "learning_rate": 8.628515349464839e-05,
      "loss": 0.7335,
      "mean_token_accuracy": 0.7827360816299915,
      "num_tokens": 142757000.0,
      "step": 4473
    },
    {
      "entropy": 0.7747982535511255,
      "epoch": 0.4116365058280286,
      "grad_norm": 0.15746422111988068,
      "learning_rate": 8.628208666850676e-05,
      "loss": 0.7811,
      "mean_token_accuracy": 0.7764881886541843,
      "num_tokens": 142788758.0,
      "step": 4474
    },
    {
      "entropy": 0.7062945328652859,
      "epoch": 0.4117285121994698,
      "grad_norm": 0.14630766212940216,
      "learning_rate": 8.627901984236514e-05,
      "loss": 0.6772,
      "mean_token_accuracy": 0.7976622954010963,
      "num_tokens": 142821075.0,
      "step": 4475
    },
    {
      "entropy": 0.7564222253859043,
      "epoch": 0.411820518570911,
      "grad_norm": 0.15327119827270508,
      "learning_rate": 8.627595301622351e-05,
      "loss": 0.7294,
      "mean_token_accuracy": 0.784772977232933,
      "num_tokens": 142852421.0,
      "step": 4476
    },
    {
      "entropy": 0.7790822144597769,
      "epoch": 0.41191252494235225,
      "grad_norm": 0.16707418859004974,
      "learning_rate": 8.627288619008189e-05,
      "loss": 0.7648,
      "mean_token_accuracy": 0.7772913165390491,
      "num_tokens": 142884183.0,
      "step": 4477
    },
    {
      "entropy": 0.8511574026197195,
      "epoch": 0.4120045313137935,
      "grad_norm": 0.1737680584192276,
      "learning_rate": 8.626981936394026e-05,
      "loss": 0.8758,
      "mean_token_accuracy": 0.7526406645774841,
      "num_tokens": 142916369.0,
      "step": 4478
    },
    {
      "entropy": 0.8328302353620529,
      "epoch": 0.4120965376852347,
      "grad_norm": 0.15970250964164734,
      "learning_rate": 8.626675253779864e-05,
      "loss": 0.8346,
      "mean_token_accuracy": 0.7587746530771255,
      "num_tokens": 142948823.0,
      "step": 4479
    },
    {
      "entropy": 0.7236127369105816,
      "epoch": 0.41218854405667593,
      "grad_norm": 0.1568552553653717,
      "learning_rate": 8.626368571165701e-05,
      "loss": 0.7313,
      "mean_token_accuracy": 0.7794477865099907,
      "num_tokens": 142980637.0,
      "step": 4480
    },
    {
      "entropy": 0.8552698995918036,
      "epoch": 0.4122805504281171,
      "grad_norm": 0.16580688953399658,
      "learning_rate": 8.626061888551539e-05,
      "loss": 0.8565,
      "mean_token_accuracy": 0.7544540241360664,
      "num_tokens": 143011363.0,
      "step": 4481
    },
    {
      "entropy": 0.852189339697361,
      "epoch": 0.41237255679955837,
      "grad_norm": 0.16307270526885986,
      "learning_rate": 8.625755205937376e-05,
      "loss": 0.8357,
      "mean_token_accuracy": 0.7587158307433128,
      "num_tokens": 143042638.0,
      "step": 4482
    },
    {
      "entropy": 0.8167440257966518,
      "epoch": 0.4124645631709996,
      "grad_norm": 0.15553699433803558,
      "learning_rate": 8.625448523323213e-05,
      "loss": 0.8167,
      "mean_token_accuracy": 0.7622505836188793,
      "num_tokens": 143074522.0,
      "step": 4483
    },
    {
      "entropy": 0.7384087014943361,
      "epoch": 0.4125565695424408,
      "grad_norm": 0.15857434272766113,
      "learning_rate": 8.625141840709051e-05,
      "loss": 0.7225,
      "mean_token_accuracy": 0.7872103899717331,
      "num_tokens": 143106793.0,
      "step": 4484
    },
    {
      "entropy": 0.7382897306233644,
      "epoch": 0.41264857591388204,
      "grad_norm": 0.1553286612033844,
      "learning_rate": 8.624835158094889e-05,
      "loss": 0.7176,
      "mean_token_accuracy": 0.7924973182380199,
      "num_tokens": 143138426.0,
      "step": 4485
    },
    {
      "entropy": 0.8515257630497217,
      "epoch": 0.41274058228532323,
      "grad_norm": 0.159807026386261,
      "learning_rate": 8.624528475480726e-05,
      "loss": 0.8622,
      "mean_token_accuracy": 0.7584990561008453,
      "num_tokens": 143170941.0,
      "step": 4486
    },
    {
      "entropy": 0.8342829793691635,
      "epoch": 0.4128325886567645,
      "grad_norm": 0.159601092338562,
      "learning_rate": 8.624221792866562e-05,
      "loss": 0.8136,
      "mean_token_accuracy": 0.7672271206974983,
      "num_tokens": 143202762.0,
      "step": 4487
    },
    {
      "entropy": 0.8449788987636566,
      "epoch": 0.4129245950282057,
      "grad_norm": 0.16580863296985626,
      "learning_rate": 8.6239151102524e-05,
      "loss": 0.8287,
      "mean_token_accuracy": 0.7591719217598438,
      "num_tokens": 143234118.0,
      "step": 4488
    },
    {
      "entropy": 0.8111384641379118,
      "epoch": 0.4130166013996469,
      "grad_norm": 0.15646781027317047,
      "learning_rate": 8.623608427638237e-05,
      "loss": 0.8173,
      "mean_token_accuracy": 0.7626026049256325,
      "num_tokens": 143265762.0,
      "step": 4489
    },
    {
      "entropy": 0.7839017100632191,
      "epoch": 0.41310860777108815,
      "grad_norm": 0.15555121004581451,
      "learning_rate": 8.623301745024076e-05,
      "loss": 0.7669,
      "mean_token_accuracy": 0.776103425770998,
      "num_tokens": 143297529.0,
      "step": 4490
    },
    {
      "entropy": 0.9295414686203003,
      "epoch": 0.41320061414252934,
      "grad_norm": 0.16158102452754974,
      "learning_rate": 8.622995062409912e-05,
      "loss": 0.9085,
      "mean_token_accuracy": 0.7382762506604195,
      "num_tokens": 143329246.0,
      "step": 4491
    },
    {
      "entropy": 0.8373202793300152,
      "epoch": 0.4132926205139706,
      "grad_norm": 0.16207043826580048,
      "learning_rate": 8.62268837979575e-05,
      "loss": 0.8248,
      "mean_token_accuracy": 0.7603593990206718,
      "num_tokens": 143360714.0,
      "step": 4492
    },
    {
      "entropy": 0.7937959562987089,
      "epoch": 0.41338462688541183,
      "grad_norm": 0.1623910516500473,
      "learning_rate": 8.622381697181587e-05,
      "loss": 0.7879,
      "mean_token_accuracy": 0.7690782472491264,
      "num_tokens": 143392818.0,
      "step": 4493
    },
    {
      "entropy": 0.8386279363185167,
      "epoch": 0.413476633256853,
      "grad_norm": 0.15878570079803467,
      "learning_rate": 8.622075014567424e-05,
      "loss": 0.8322,
      "mean_token_accuracy": 0.7591912224888802,
      "num_tokens": 143424781.0,
      "step": 4494
    },
    {
      "entropy": 0.81102329865098,
      "epoch": 0.41356863962829427,
      "grad_norm": 0.16784195601940155,
      "learning_rate": 8.621768331953262e-05,
      "loss": 0.8082,
      "mean_token_accuracy": 0.7653237469494343,
      "num_tokens": 143457233.0,
      "step": 4495
    },
    {
      "entropy": 0.8917455077171326,
      "epoch": 0.41366064599973545,
      "grad_norm": 0.16201429069042206,
      "learning_rate": 8.6214616493391e-05,
      "loss": 0.8838,
      "mean_token_accuracy": 0.7514631487429142,
      "num_tokens": 143488915.0,
      "step": 4496
    },
    {
      "entropy": 0.8485833611339331,
      "epoch": 0.4137526523711767,
      "grad_norm": 0.1615356206893921,
      "learning_rate": 8.621154966724937e-05,
      "loss": 0.8422,
      "mean_token_accuracy": 0.7511593662202358,
      "num_tokens": 143520828.0,
      "step": 4497
    },
    {
      "entropy": 0.8395666684955359,
      "epoch": 0.41384465874261794,
      "grad_norm": 0.16144776344299316,
      "learning_rate": 8.620848284110774e-05,
      "loss": 0.8411,
      "mean_token_accuracy": 0.7572939731180668,
      "num_tokens": 143552446.0,
      "step": 4498
    },
    {
      "entropy": 0.8123929630964994,
      "epoch": 0.41393666511405913,
      "grad_norm": 0.1597829908132553,
      "learning_rate": 8.62054160149661e-05,
      "loss": 0.7996,
      "mean_token_accuracy": 0.7689096070826054,
      "num_tokens": 143585033.0,
      "step": 4499
    },
    {
      "entropy": 0.8419951759278774,
      "epoch": 0.4140286714855004,
      "grad_norm": 0.16439473628997803,
      "learning_rate": 8.620234918882449e-05,
      "loss": 0.8364,
      "mean_token_accuracy": 0.7572947293519974,
      "num_tokens": 143615931.0,
      "step": 4500
    },
    {
      "entropy": 1.0097515024244785,
      "epoch": 0.41412067785694157,
      "grad_norm": 0.1709626466035843,
      "learning_rate": 8.619928236268287e-05,
      "loss": 1.0012,
      "mean_token_accuracy": 0.7187640853226185,
      "num_tokens": 143647966.0,
      "step": 4501
    },
    {
      "entropy": 0.7984908521175385,
      "epoch": 0.4142126842283828,
      "grad_norm": 0.162941575050354,
      "learning_rate": 8.619621553654124e-05,
      "loss": 0.7841,
      "mean_token_accuracy": 0.7726999334990978,
      "num_tokens": 143679324.0,
      "step": 4502
    },
    {
      "entropy": 0.8631959911435843,
      "epoch": 0.41430469059982405,
      "grad_norm": 0.15573735535144806,
      "learning_rate": 8.61931487103996e-05,
      "loss": 0.8574,
      "mean_token_accuracy": 0.7514755912125111,
      "num_tokens": 143711785.0,
      "step": 4503
    },
    {
      "entropy": 0.8887576498091221,
      "epoch": 0.41439669697126524,
      "grad_norm": 0.16424015164375305,
      "learning_rate": 8.619008188425799e-05,
      "loss": 0.8951,
      "mean_token_accuracy": 0.7431492023169994,
      "num_tokens": 143743975.0,
      "step": 4504
    },
    {
      "entropy": 0.8071123957633972,
      "epoch": 0.4144887033427065,
      "grad_norm": 0.16102947294712067,
      "learning_rate": 8.618701505811635e-05,
      "loss": 0.802,
      "mean_token_accuracy": 0.7655475251376629,
      "num_tokens": 143776047.0,
      "step": 4505
    },
    {
      "entropy": 0.8156405128538609,
      "epoch": 0.4145807097141477,
      "grad_norm": 0.1658691018819809,
      "learning_rate": 8.618394823197474e-05,
      "loss": 0.82,
      "mean_token_accuracy": 0.7663696967065334,
      "num_tokens": 143808309.0,
      "step": 4506
    },
    {
      "entropy": 0.7904166709631681,
      "epoch": 0.4146727160855889,
      "grad_norm": 0.16226786375045776,
      "learning_rate": 8.61808814058331e-05,
      "loss": 0.7983,
      "mean_token_accuracy": 0.772817425429821,
      "num_tokens": 143840210.0,
      "step": 4507
    },
    {
      "entropy": 0.9053803756833076,
      "epoch": 0.41476472245703017,
      "grad_norm": 0.16709619760513306,
      "learning_rate": 8.617781457969149e-05,
      "loss": 0.9014,
      "mean_token_accuracy": 0.7450561746954918,
      "num_tokens": 143871944.0,
      "step": 4508
    },
    {
      "entropy": 0.8557576239109039,
      "epoch": 0.41485672882847136,
      "grad_norm": 0.16033786535263062,
      "learning_rate": 8.617474775354985e-05,
      "loss": 0.8589,
      "mean_token_accuracy": 0.7548679262399673,
      "num_tokens": 143903979.0,
      "step": 4509
    },
    {
      "entropy": 0.738355340436101,
      "epoch": 0.4149487351999126,
      "grad_norm": 0.15662580728530884,
      "learning_rate": 8.617168092740822e-05,
      "loss": 0.7123,
      "mean_token_accuracy": 0.7898793183267117,
      "num_tokens": 143936013.0,
      "step": 4510
    },
    {
      "entropy": 0.7305729389190674,
      "epoch": 0.41504074157135384,
      "grad_norm": 0.15143997967243195,
      "learning_rate": 8.61686141012666e-05,
      "loss": 0.695,
      "mean_token_accuracy": 0.7937044203281403,
      "num_tokens": 143967797.0,
      "step": 4511
    },
    {
      "entropy": 0.7617721986025572,
      "epoch": 0.41513274794279503,
      "grad_norm": 0.16395175457000732,
      "learning_rate": 8.616554727512498e-05,
      "loss": 0.7524,
      "mean_token_accuracy": 0.781597200781107,
      "num_tokens": 143999658.0,
      "step": 4512
    },
    {
      "entropy": 0.7931550070643425,
      "epoch": 0.4152247543142363,
      "grad_norm": 0.158483624458313,
      "learning_rate": 8.616248044898335e-05,
      "loss": 0.7592,
      "mean_token_accuracy": 0.7798235230147839,
      "num_tokens": 144031732.0,
      "step": 4513
    },
    {
      "entropy": 0.7768251746892929,
      "epoch": 0.41531676068567747,
      "grad_norm": 0.16294491291046143,
      "learning_rate": 8.615941362284172e-05,
      "loss": 0.7582,
      "mean_token_accuracy": 0.7802399508655071,
      "num_tokens": 144063960.0,
      "step": 4514
    },
    {
      "entropy": 0.9042362626641989,
      "epoch": 0.4154087670571187,
      "grad_norm": 0.16017252206802368,
      "learning_rate": 8.615634679670009e-05,
      "loss": 0.8925,
      "mean_token_accuracy": 0.7405245676636696,
      "num_tokens": 144096292.0,
      "step": 4515
    },
    {
      "entropy": 0.8998206127434969,
      "epoch": 0.41550077342855996,
      "grad_norm": 0.16232304275035858,
      "learning_rate": 8.615327997055848e-05,
      "loss": 0.9048,
      "mean_token_accuracy": 0.7395599037408829,
      "num_tokens": 144128818.0,
      "step": 4516
    },
    {
      "entropy": 0.8306217007339001,
      "epoch": 0.41559277980000114,
      "grad_norm": 0.1589776873588562,
      "learning_rate": 8.615021314441685e-05,
      "loss": 0.8194,
      "mean_token_accuracy": 0.7641562297940254,
      "num_tokens": 144160875.0,
      "step": 4517
    },
    {
      "entropy": 0.7502652574330568,
      "epoch": 0.4156847861714424,
      "grad_norm": 0.1631874293088913,
      "learning_rate": 8.614714631827522e-05,
      "loss": 0.7449,
      "mean_token_accuracy": 0.7803785391151905,
      "num_tokens": 144192663.0,
      "step": 4518
    },
    {
      "entropy": 0.7109964918345213,
      "epoch": 0.4157767925428836,
      "grad_norm": 0.1630563884973526,
      "learning_rate": 8.61440794921336e-05,
      "loss": 0.6993,
      "mean_token_accuracy": 0.7982547245919704,
      "num_tokens": 144224734.0,
      "step": 4519
    },
    {
      "entropy": 0.8094420395791531,
      "epoch": 0.4158687989143248,
      "grad_norm": 0.1642209142446518,
      "learning_rate": 8.614101266599197e-05,
      "loss": 0.7927,
      "mean_token_accuracy": 0.7671141251921654,
      "num_tokens": 144255909.0,
      "step": 4520
    },
    {
      "entropy": 0.9098432809114456,
      "epoch": 0.41596080528576607,
      "grad_norm": 0.16872932016849518,
      "learning_rate": 8.613794583985035e-05,
      "loss": 0.9126,
      "mean_token_accuracy": 0.7406991347670555,
      "num_tokens": 144288274.0,
      "step": 4521
    },
    {
      "entropy": 0.769041096791625,
      "epoch": 0.41605281165720726,
      "grad_norm": 0.15959005057811737,
      "learning_rate": 8.613487901370872e-05,
      "loss": 0.7536,
      "mean_token_accuracy": 0.7832488566637039,
      "num_tokens": 144320717.0,
      "step": 4522
    },
    {
      "entropy": 0.8012737669050694,
      "epoch": 0.4161448180286485,
      "grad_norm": 0.16060680150985718,
      "learning_rate": 8.61318121875671e-05,
      "loss": 0.8061,
      "mean_token_accuracy": 0.7614594958722591,
      "num_tokens": 144352677.0,
      "step": 4523
    },
    {
      "entropy": 0.6935494672507048,
      "epoch": 0.4162368244000897,
      "grad_norm": 0.156570702791214,
      "learning_rate": 8.612874536142547e-05,
      "loss": 0.6661,
      "mean_token_accuracy": 0.8025020696222782,
      "num_tokens": 144384273.0,
      "step": 4524
    },
    {
      "entropy": 0.7270583268254995,
      "epoch": 0.41632883077153093,
      "grad_norm": 0.154010608792305,
      "learning_rate": 8.612567853528383e-05,
      "loss": 0.732,
      "mean_token_accuracy": 0.7828461267054081,
      "num_tokens": 144416640.0,
      "step": 4525
    },
    {
      "entropy": 0.7875452823936939,
      "epoch": 0.4164208371429722,
      "grad_norm": 0.15570209920406342,
      "learning_rate": 8.612261170914222e-05,
      "loss": 0.7612,
      "mean_token_accuracy": 0.7811396606266499,
      "num_tokens": 144449171.0,
      "step": 4526
    },
    {
      "entropy": 0.7737717665731907,
      "epoch": 0.41651284351441337,
      "grad_norm": 0.1569824516773224,
      "learning_rate": 8.61195448830006e-05,
      "loss": 0.7692,
      "mean_token_accuracy": 0.7778169177472591,
      "num_tokens": 144480871.0,
      "step": 4527
    },
    {
      "entropy": 0.8244694396853447,
      "epoch": 0.4166048498858546,
      "grad_norm": 0.15718339383602142,
      "learning_rate": 8.611647805685896e-05,
      "loss": 0.8241,
      "mean_token_accuracy": 0.7618833892047405,
      "num_tokens": 144513171.0,
      "step": 4528
    },
    {
      "entropy": 0.6618942860513926,
      "epoch": 0.4166968562572958,
      "grad_norm": 0.15114694833755493,
      "learning_rate": 8.611341123071733e-05,
      "loss": 0.645,
      "mean_token_accuracy": 0.8112078756093979,
      "num_tokens": 144545391.0,
      "step": 4529
    },
    {
      "entropy": 0.8036180213093758,
      "epoch": 0.41678886262873704,
      "grad_norm": 0.16020254790782928,
      "learning_rate": 8.61103444045757e-05,
      "loss": 0.7981,
      "mean_token_accuracy": 0.7665360942482948,
      "num_tokens": 144577511.0,
      "step": 4530
    },
    {
      "entropy": 0.8657460249960423,
      "epoch": 0.4168808690001783,
      "grad_norm": 0.16278044879436493,
      "learning_rate": 8.610727757843408e-05,
      "loss": 0.862,
      "mean_token_accuracy": 0.7496981211006641,
      "num_tokens": 144609397.0,
      "step": 4531
    },
    {
      "entropy": 0.7543759532272816,
      "epoch": 0.4169728753716195,
      "grad_norm": 0.1511809080839157,
      "learning_rate": 8.610421075229246e-05,
      "loss": 0.7386,
      "mean_token_accuracy": 0.7819755971431732,
      "num_tokens": 144641649.0,
      "step": 4532
    },
    {
      "entropy": 0.8243298269808292,
      "epoch": 0.4170648817430607,
      "grad_norm": 0.15642566978931427,
      "learning_rate": 8.610114392615083e-05,
      "loss": 0.8188,
      "mean_token_accuracy": 0.7629193924367428,
      "num_tokens": 144673498.0,
      "step": 4533
    },
    {
      "entropy": 0.7835861053317785,
      "epoch": 0.4171568881145019,
      "grad_norm": 0.15588463842868805,
      "learning_rate": 8.60980771000092e-05,
      "loss": 0.7782,
      "mean_token_accuracy": 0.774412713944912,
      "num_tokens": 144705627.0,
      "step": 4534
    },
    {
      "entropy": 0.8346582818776369,
      "epoch": 0.41724889448594316,
      "grad_norm": 0.16522647440433502,
      "learning_rate": 8.609501027386758e-05,
      "loss": 0.8261,
      "mean_token_accuracy": 0.7636561542749405,
      "num_tokens": 144737539.0,
      "step": 4535
    },
    {
      "entropy": 0.7929035145789385,
      "epoch": 0.4173409008573844,
      "grad_norm": 0.16527460515499115,
      "learning_rate": 8.609194344772595e-05,
      "loss": 0.8088,
      "mean_token_accuracy": 0.7660935074090958,
      "num_tokens": 144769540.0,
      "step": 4536
    },
    {
      "entropy": 0.7421628106385469,
      "epoch": 0.4174329072288256,
      "grad_norm": 0.15511679649353027,
      "learning_rate": 8.608887662158433e-05,
      "loss": 0.7106,
      "mean_token_accuracy": 0.7888552136719227,
      "num_tokens": 144801113.0,
      "step": 4537
    },
    {
      "entropy": 0.86304578371346,
      "epoch": 0.41752491360026683,
      "grad_norm": 0.162871316075325,
      "learning_rate": 8.60858097954427e-05,
      "loss": 0.8631,
      "mean_token_accuracy": 0.7484033033251762,
      "num_tokens": 144832481.0,
      "step": 4538
    },
    {
      "entropy": 0.8668824750930071,
      "epoch": 0.417616919971708,
      "grad_norm": 0.15830406546592712,
      "learning_rate": 8.608274296930108e-05,
      "loss": 0.8532,
      "mean_token_accuracy": 0.7538893781602383,
      "num_tokens": 144864705.0,
      "step": 4539
    },
    {
      "entropy": 0.8107856996357441,
      "epoch": 0.41770892634314927,
      "grad_norm": 0.16174452006816864,
      "learning_rate": 8.607967614315945e-05,
      "loss": 0.7891,
      "mean_token_accuracy": 0.7666014470160007,
      "num_tokens": 144896095.0,
      "step": 4540
    },
    {
      "entropy": 0.7459104843437672,
      "epoch": 0.4178009327145905,
      "grad_norm": 0.1629287749528885,
      "learning_rate": 8.607660931701781e-05,
      "loss": 0.7486,
      "mean_token_accuracy": 0.7754403613507748,
      "num_tokens": 144928067.0,
      "step": 4541
    },
    {
      "entropy": 0.89670030772686,
      "epoch": 0.4178929390860317,
      "grad_norm": 0.15871508419513702,
      "learning_rate": 8.60735424908762e-05,
      "loss": 0.8944,
      "mean_token_accuracy": 0.7424206882715225,
      "num_tokens": 144960605.0,
      "step": 4542
    },
    {
      "entropy": 0.7221546322107315,
      "epoch": 0.41798494545747295,
      "grad_norm": 0.1663232296705246,
      "learning_rate": 8.607047566473458e-05,
      "loss": 0.7293,
      "mean_token_accuracy": 0.7870559468865395,
      "num_tokens": 144993128.0,
      "step": 4543
    },
    {
      "entropy": 0.7570828627794981,
      "epoch": 0.41807695182891413,
      "grad_norm": 0.15465998649597168,
      "learning_rate": 8.606740883859295e-05,
      "loss": 0.7387,
      "mean_token_accuracy": 0.7846238315105438,
      "num_tokens": 145025896.0,
      "step": 4544
    },
    {
      "entropy": 0.7791625969111919,
      "epoch": 0.4181689582003554,
      "grad_norm": 0.15877221524715424,
      "learning_rate": 8.606434201245131e-05,
      "loss": 0.7693,
      "mean_token_accuracy": 0.7734597437083721,
      "num_tokens": 145056895.0,
      "step": 4545
    },
    {
      "entropy": 0.8962718639522791,
      "epoch": 0.4182609645717966,
      "grad_norm": 0.16469255089759827,
      "learning_rate": 8.60612751863097e-05,
      "loss": 0.8769,
      "mean_token_accuracy": 0.7464188300073147,
      "num_tokens": 145088588.0,
      "step": 4546
    },
    {
      "entropy": 0.9129232279956341,
      "epoch": 0.4183529709432378,
      "grad_norm": 0.16449591517448425,
      "learning_rate": 8.605820836016806e-05,
      "loss": 0.9188,
      "mean_token_accuracy": 0.7396076433360577,
      "num_tokens": 145120523.0,
      "step": 4547
    },
    {
      "entropy": 0.772522110491991,
      "epoch": 0.41844497731467906,
      "grad_norm": 0.15302924811840057,
      "learning_rate": 8.605514153402644e-05,
      "loss": 0.7434,
      "mean_token_accuracy": 0.7840568087995052,
      "num_tokens": 145152929.0,
      "step": 4548
    },
    {
      "entropy": 0.7632583193480968,
      "epoch": 0.41853698368612025,
      "grad_norm": 0.15166132152080536,
      "learning_rate": 8.605207470788481e-05,
      "loss": 0.7531,
      "mean_token_accuracy": 0.786506824195385,
      "num_tokens": 145185695.0,
      "step": 4549
    },
    {
      "entropy": 0.7220206093043089,
      "epoch": 0.4186289900575615,
      "grad_norm": 0.1549224555492401,
      "learning_rate": 8.60490078817432e-05,
      "loss": 0.7154,
      "mean_token_accuracy": 0.7883147113025188,
      "num_tokens": 145217611.0,
      "step": 4550
    },
    {
      "entropy": 0.8025738559663296,
      "epoch": 0.41872099642900273,
      "grad_norm": 0.16210800409317017,
      "learning_rate": 8.604594105560156e-05,
      "loss": 0.8143,
      "mean_token_accuracy": 0.7626190148293972,
      "num_tokens": 145249022.0,
      "step": 4551
    },
    {
      "entropy": 0.8407811541110277,
      "epoch": 0.4188130028004439,
      "grad_norm": 0.15261705219745636,
      "learning_rate": 8.604287422945993e-05,
      "loss": 0.8249,
      "mean_token_accuracy": 0.7616807594895363,
      "num_tokens": 145280998.0,
      "step": 4552
    },
    {
      "entropy": 0.7695520222187042,
      "epoch": 0.41890500917188517,
      "grad_norm": 0.15632832050323486,
      "learning_rate": 8.603980740331831e-05,
      "loss": 0.7635,
      "mean_token_accuracy": 0.7732005901634693,
      "num_tokens": 145313741.0,
      "step": 4553
    },
    {
      "entropy": 0.8493876922875643,
      "epoch": 0.41899701554332636,
      "grad_norm": 0.16051173210144043,
      "learning_rate": 8.603674057717669e-05,
      "loss": 0.8745,
      "mean_token_accuracy": 0.754030879586935,
      "num_tokens": 145346310.0,
      "step": 4554
    },
    {
      "entropy": 0.8957344107329845,
      "epoch": 0.4190890219147676,
      "grad_norm": 0.1706491857767105,
      "learning_rate": 8.603367375103506e-05,
      "loss": 0.9076,
      "mean_token_accuracy": 0.741692166775465,
      "num_tokens": 145377737.0,
      "step": 4555
    },
    {
      "entropy": 0.759348077699542,
      "epoch": 0.41918102828620885,
      "grad_norm": 0.1669037789106369,
      "learning_rate": 8.603060692489343e-05,
      "loss": 0.7423,
      "mean_token_accuracy": 0.7830957174301147,
      "num_tokens": 145409819.0,
      "step": 4556
    },
    {
      "entropy": 0.9060728810727596,
      "epoch": 0.41927303465765003,
      "grad_norm": 0.16334396600723267,
      "learning_rate": 8.60275400987518e-05,
      "loss": 0.9085,
      "mean_token_accuracy": 0.7380963116884232,
      "num_tokens": 145442328.0,
      "step": 4557
    },
    {
      "entropy": 0.8253596406430006,
      "epoch": 0.4193650410290913,
      "grad_norm": 0.15739575028419495,
      "learning_rate": 8.602447327261019e-05,
      "loss": 0.8029,
      "mean_token_accuracy": 0.7669550180435181,
      "num_tokens": 145474334.0,
      "step": 4558
    },
    {
      "entropy": 0.8775247260928154,
      "epoch": 0.41945704740053247,
      "grad_norm": 0.16090230643749237,
      "learning_rate": 8.602140644646856e-05,
      "loss": 0.8535,
      "mean_token_accuracy": 0.7496736943721771,
      "num_tokens": 145505783.0,
      "step": 4559
    },
    {
      "entropy": 0.8548111971467733,
      "epoch": 0.4195490537719737,
      "grad_norm": 0.15457266569137573,
      "learning_rate": 8.601833962032693e-05,
      "loss": 0.8566,
      "mean_token_accuracy": 0.7514009363949299,
      "num_tokens": 145537987.0,
      "step": 4560
    },
    {
      "entropy": 0.8343433812260628,
      "epoch": 0.41964106014341496,
      "grad_norm": 0.1598043292760849,
      "learning_rate": 8.60152727941853e-05,
      "loss": 0.8243,
      "mean_token_accuracy": 0.7587715871632099,
      "num_tokens": 145570144.0,
      "step": 4561
    },
    {
      "entropy": 0.8947041183710098,
      "epoch": 0.41973306651485615,
      "grad_norm": 0.3732788860797882,
      "learning_rate": 8.601220596804368e-05,
      "loss": 0.876,
      "mean_token_accuracy": 0.7527636922895908,
      "num_tokens": 145602261.0,
      "step": 4562
    },
    {
      "entropy": 0.802547924220562,
      "epoch": 0.4198250728862974,
      "grad_norm": 0.15301138162612915,
      "learning_rate": 8.600913914190206e-05,
      "loss": 0.8104,
      "mean_token_accuracy": 0.765023872256279,
      "num_tokens": 145634723.0,
      "step": 4563
    },
    {
      "entropy": 0.8678766470402479,
      "epoch": 0.4199170792577386,
      "grad_norm": 0.16276900470256805,
      "learning_rate": 8.600607231576042e-05,
      "loss": 0.8773,
      "mean_token_accuracy": 0.7489379607141018,
      "num_tokens": 145667105.0,
      "step": 4564
    },
    {
      "entropy": 0.8827195577323437,
      "epoch": 0.4200090856291798,
      "grad_norm": 0.157405823469162,
      "learning_rate": 8.600300548961879e-05,
      "loss": 0.8636,
      "mean_token_accuracy": 0.7523163370788097,
      "num_tokens": 145699530.0,
      "step": 4565
    },
    {
      "entropy": 0.7445742171257734,
      "epoch": 0.42010109200062107,
      "grad_norm": 0.15601381659507751,
      "learning_rate": 8.599993866347717e-05,
      "loss": 0.7365,
      "mean_token_accuracy": 0.7800416573882103,
      "num_tokens": 145731520.0,
      "step": 4566
    },
    {
      "entropy": 0.8754270896315575,
      "epoch": 0.42019309837206226,
      "grad_norm": 0.1704360544681549,
      "learning_rate": 8.599687183733554e-05,
      "loss": 0.8696,
      "mean_token_accuracy": 0.7472438998520374,
      "num_tokens": 145763951.0,
      "step": 4567
    },
    {
      "entropy": 0.8988167252391577,
      "epoch": 0.4202851047435035,
      "grad_norm": 0.16590768098831177,
      "learning_rate": 8.599380501119392e-05,
      "loss": 0.9036,
      "mean_token_accuracy": 0.7437917403876781,
      "num_tokens": 145796230.0,
      "step": 4568
    },
    {
      "entropy": 0.7939202897250652,
      "epoch": 0.4203771111149447,
      "grad_norm": 0.15913666784763336,
      "learning_rate": 8.599073818505229e-05,
      "loss": 0.7806,
      "mean_token_accuracy": 0.7761568129062653,
      "num_tokens": 145828214.0,
      "step": 4569
    },
    {
      "entropy": 0.7721442673355341,
      "epoch": 0.42046911748638593,
      "grad_norm": 0.16158299148082733,
      "learning_rate": 8.598767135891067e-05,
      "loss": 0.7603,
      "mean_token_accuracy": 0.7767036706209183,
      "num_tokens": 145860323.0,
      "step": 4570
    },
    {
      "entropy": 0.8092829547822475,
      "epoch": 0.4205611238578272,
      "grad_norm": 0.16427403688430786,
      "learning_rate": 8.598460453276904e-05,
      "loss": 0.8007,
      "mean_token_accuracy": 0.7737971805036068,
      "num_tokens": 145892793.0,
      "step": 4571
    },
    {
      "entropy": 0.7533850334584713,
      "epoch": 0.42065313022926837,
      "grad_norm": 0.15690869092941284,
      "learning_rate": 8.598153770662741e-05,
      "loss": 0.7399,
      "mean_token_accuracy": 0.7811124958097935,
      "num_tokens": 145924667.0,
      "step": 4572
    },
    {
      "entropy": 0.7132236901670694,
      "epoch": 0.4207451366007096,
      "grad_norm": 0.14741215109825134,
      "learning_rate": 8.597847088048579e-05,
      "loss": 0.6788,
      "mean_token_accuracy": 0.7985172607004642,
      "num_tokens": 145957002.0,
      "step": 4573
    },
    {
      "entropy": 0.8207326736301184,
      "epoch": 0.4208371429721508,
      "grad_norm": 0.1580919474363327,
      "learning_rate": 8.597540405434417e-05,
      "loss": 0.8078,
      "mean_token_accuracy": 0.7644231207668781,
      "num_tokens": 145988501.0,
      "step": 4574
    },
    {
      "entropy": 0.8459073910489678,
      "epoch": 0.42092914934359205,
      "grad_norm": 0.16118136048316956,
      "learning_rate": 8.597233722820254e-05,
      "loss": 0.8478,
      "mean_token_accuracy": 0.7616735361516476,
      "num_tokens": 146020446.0,
      "step": 4575
    },
    {
      "entropy": 0.6958336066454649,
      "epoch": 0.4210211557150333,
      "grad_norm": 0.16549701988697052,
      "learning_rate": 8.596927040206091e-05,
      "loss": 0.6789,
      "mean_token_accuracy": 0.8016561567783356,
      "num_tokens": 146052344.0,
      "step": 4576
    },
    {
      "entropy": 0.7792667876929045,
      "epoch": 0.4211131620864745,
      "grad_norm": 0.17580147087574005,
      "learning_rate": 8.596620357591929e-05,
      "loss": 0.7685,
      "mean_token_accuracy": 0.7767567038536072,
      "num_tokens": 146084787.0,
      "step": 4577
    },
    {
      "entropy": 0.8031229712069035,
      "epoch": 0.4212051684579157,
      "grad_norm": 0.16030214726924896,
      "learning_rate": 8.596313674977766e-05,
      "loss": 0.7955,
      "mean_token_accuracy": 0.7707584761083126,
      "num_tokens": 146117391.0,
      "step": 4578
    },
    {
      "entropy": 0.7864808011800051,
      "epoch": 0.4212971748293569,
      "grad_norm": 0.1532261073589325,
      "learning_rate": 8.596006992363604e-05,
      "loss": 0.7661,
      "mean_token_accuracy": 0.7768229730427265,
      "num_tokens": 146149470.0,
      "step": 4579
    },
    {
      "entropy": 0.8507726304233074,
      "epoch": 0.42138918120079816,
      "grad_norm": 0.16311869025230408,
      "learning_rate": 8.59570030974944e-05,
      "loss": 0.8239,
      "mean_token_accuracy": 0.7581100091338158,
      "num_tokens": 146181160.0,
      "step": 4580
    },
    {
      "entropy": 0.8729174360632896,
      "epoch": 0.4214811875722394,
      "grad_norm": 0.1688160002231598,
      "learning_rate": 8.595393627135279e-05,
      "loss": 0.8822,
      "mean_token_accuracy": 0.7497265227138996,
      "num_tokens": 146212273.0,
      "step": 4581
    },
    {
      "entropy": 0.7768956888467073,
      "epoch": 0.4215731939436806,
      "grad_norm": 0.16243866086006165,
      "learning_rate": 8.595086944521115e-05,
      "loss": 0.7693,
      "mean_token_accuracy": 0.7730605117976665,
      "num_tokens": 146243856.0,
      "step": 4582
    },
    {
      "entropy": 0.7503443229943514,
      "epoch": 0.42166520031512184,
      "grad_norm": 0.15790176391601562,
      "learning_rate": 8.594780261906952e-05,
      "loss": 0.7482,
      "mean_token_accuracy": 0.7851225398480892,
      "num_tokens": 146275893.0,
      "step": 4583
    },
    {
      "entropy": 0.827079389244318,
      "epoch": 0.421757206686563,
      "grad_norm": 0.1637110412120819,
      "learning_rate": 8.59447357929279e-05,
      "loss": 0.8221,
      "mean_token_accuracy": 0.7603440433740616,
      "num_tokens": 146307485.0,
      "step": 4584
    },
    {
      "entropy": 0.8225524686276913,
      "epoch": 0.42184921305800427,
      "grad_norm": 0.1597427874803543,
      "learning_rate": 8.594166896678629e-05,
      "loss": 0.8323,
      "mean_token_accuracy": 0.7623053230345249,
      "num_tokens": 146339540.0,
      "step": 4585
    },
    {
      "entropy": 0.754946019500494,
      "epoch": 0.4219412194294455,
      "grad_norm": 0.16408030688762665,
      "learning_rate": 8.593860214064465e-05,
      "loss": 0.7481,
      "mean_token_accuracy": 0.7820260152220726,
      "num_tokens": 146370866.0,
      "step": 4586
    },
    {
      "entropy": 0.8852417282760143,
      "epoch": 0.4220332258008867,
      "grad_norm": 0.16907674074172974,
      "learning_rate": 8.593553531450302e-05,
      "loss": 0.8745,
      "mean_token_accuracy": 0.7480325102806091,
      "num_tokens": 146402622.0,
      "step": 4587
    },
    {
      "entropy": 0.7553645018488169,
      "epoch": 0.42212523217232795,
      "grad_norm": 0.16264121234416962,
      "learning_rate": 8.593246848836139e-05,
      "loss": 0.7311,
      "mean_token_accuracy": 0.786100372672081,
      "num_tokens": 146433798.0,
      "step": 4588
    },
    {
      "entropy": 0.7370216995477676,
      "epoch": 0.42221723854376914,
      "grad_norm": 0.16025198996067047,
      "learning_rate": 8.592940166221977e-05,
      "loss": 0.7134,
      "mean_token_accuracy": 0.7892382703721523,
      "num_tokens": 146465035.0,
      "step": 4589
    },
    {
      "entropy": 0.8346895687282085,
      "epoch": 0.4223092449152104,
      "grad_norm": 0.1560012847185135,
      "learning_rate": 8.592633483607815e-05,
      "loss": 0.815,
      "mean_token_accuracy": 0.7643697969615459,
      "num_tokens": 146496542.0,
      "step": 4590
    },
    {
      "entropy": 0.8215970043092966,
      "epoch": 0.4224012512866516,
      "grad_norm": 0.14962022006511688,
      "learning_rate": 8.592326800993652e-05,
      "loss": 0.8142,
      "mean_token_accuracy": 0.7604928500950336,
      "num_tokens": 146529164.0,
      "step": 4591
    },
    {
      "entropy": 0.9242461621761322,
      "epoch": 0.4224932576580928,
      "grad_norm": 0.16236403584480286,
      "learning_rate": 8.592020118379489e-05,
      "loss": 0.9187,
      "mean_token_accuracy": 0.737382810562849,
      "num_tokens": 146561642.0,
      "step": 4592
    },
    {
      "entropy": 0.7443161923438311,
      "epoch": 0.42258526402953406,
      "grad_norm": 0.14938580989837646,
      "learning_rate": 8.591713435765327e-05,
      "loss": 0.7252,
      "mean_token_accuracy": 0.788528323173523,
      "num_tokens": 146594155.0,
      "step": 4593
    },
    {
      "entropy": 0.9029577635228634,
      "epoch": 0.42267727040097525,
      "grad_norm": 0.16229337453842163,
      "learning_rate": 8.591406753151164e-05,
      "loss": 0.8855,
      "mean_token_accuracy": 0.7462423369288445,
      "num_tokens": 146625504.0,
      "step": 4594
    },
    {
      "entropy": 0.7504708841443062,
      "epoch": 0.4227692767724165,
      "grad_norm": 0.15676549077033997,
      "learning_rate": 8.591100070537002e-05,
      "loss": 0.7352,
      "mean_token_accuracy": 0.7805223688483238,
      "num_tokens": 146656498.0,
      "step": 4595
    },
    {
      "entropy": 0.8486317228525877,
      "epoch": 0.42286128314385774,
      "grad_norm": 0.16365309059619904,
      "learning_rate": 8.590793387922839e-05,
      "loss": 0.8401,
      "mean_token_accuracy": 0.7579367011785507,
      "num_tokens": 146689262.0,
      "step": 4596
    },
    {
      "entropy": 0.8185579851269722,
      "epoch": 0.4229532895152989,
      "grad_norm": 0.1649971455335617,
      "learning_rate": 8.590486705308677e-05,
      "loss": 0.8137,
      "mean_token_accuracy": 0.7648798190057278,
      "num_tokens": 146721663.0,
      "step": 4597
    },
    {
      "entropy": 0.7321564666926861,
      "epoch": 0.42304529588674017,
      "grad_norm": 0.1517937183380127,
      "learning_rate": 8.590180022694514e-05,
      "loss": 0.7246,
      "mean_token_accuracy": 0.7859645113348961,
      "num_tokens": 146753776.0,
      "step": 4598
    },
    {
      "entropy": 0.8141267634928226,
      "epoch": 0.42313730225818136,
      "grad_norm": 0.1604725569486618,
      "learning_rate": 8.589873340080352e-05,
      "loss": 0.8075,
      "mean_token_accuracy": 0.7650229185819626,
      "num_tokens": 146785917.0,
      "step": 4599
    },
    {
      "entropy": 0.8870825208723545,
      "epoch": 0.4232293086296226,
      "grad_norm": 0.17333662509918213,
      "learning_rate": 8.589566657466189e-05,
      "loss": 0.9031,
      "mean_token_accuracy": 0.7439095303416252,
      "num_tokens": 146817581.0,
      "step": 4600
    },
    {
      "entropy": 0.8688332196325064,
      "epoch": 0.42332131500106385,
      "grad_norm": 0.17321234941482544,
      "learning_rate": 8.589259974852027e-05,
      "loss": 0.8899,
      "mean_token_accuracy": 0.7457477562129498,
      "num_tokens": 146849858.0,
      "step": 4601
    },
    {
      "entropy": 0.7977531775832176,
      "epoch": 0.42341332137250504,
      "grad_norm": 0.1636018007993698,
      "learning_rate": 8.588953292237863e-05,
      "loss": 0.7777,
      "mean_token_accuracy": 0.7770504988729954,
      "num_tokens": 146881844.0,
      "step": 4602
    },
    {
      "entropy": 0.8689647354185581,
      "epoch": 0.4235053277439463,
      "grad_norm": 0.1621980369091034,
      "learning_rate": 8.5886466096237e-05,
      "loss": 0.8557,
      "mean_token_accuracy": 0.7521816454827785,
      "num_tokens": 146914057.0,
      "step": 4603
    },
    {
      "entropy": 0.9099737778306007,
      "epoch": 0.42359733411538747,
      "grad_norm": 0.186752587556839,
      "learning_rate": 8.588339927009538e-05,
      "loss": 0.9188,
      "mean_token_accuracy": 0.7406060472130775,
      "num_tokens": 146945499.0,
      "step": 4604
    },
    {
      "entropy": 0.8282085247337818,
      "epoch": 0.4236893404868287,
      "grad_norm": 0.16053955256938934,
      "learning_rate": 8.588033244395377e-05,
      "loss": 0.8257,
      "mean_token_accuracy": 0.7623341642320156,
      "num_tokens": 146977719.0,
      "step": 4605
    },
    {
      "entropy": 0.7956900894641876,
      "epoch": 0.42378134685826996,
      "grad_norm": 0.16096076369285583,
      "learning_rate": 8.587726561781213e-05,
      "loss": 0.7598,
      "mean_token_accuracy": 0.7793000601232052,
      "num_tokens": 147010081.0,
      "step": 4606
    },
    {
      "entropy": 0.8362796679139137,
      "epoch": 0.42387335322971115,
      "grad_norm": 0.15440835058689117,
      "learning_rate": 8.58741987916705e-05,
      "loss": 0.8031,
      "mean_token_accuracy": 0.7673660963773727,
      "num_tokens": 147041960.0,
      "step": 4607
    },
    {
      "entropy": 0.7985561452805996,
      "epoch": 0.4239653596011524,
      "grad_norm": 0.154652401804924,
      "learning_rate": 8.587113196552888e-05,
      "loss": 0.7852,
      "mean_token_accuracy": 0.7702245190739632,
      "num_tokens": 147073446.0,
      "step": 4608
    },
    {
      "entropy": 0.7740116193890572,
      "epoch": 0.4240573659725936,
      "grad_norm": 0.15769924223423004,
      "learning_rate": 8.586806513938725e-05,
      "loss": 0.7493,
      "mean_token_accuracy": 0.7826761342585087,
      "num_tokens": 147105783.0,
      "step": 4609
    },
    {
      "entropy": 0.6527638901025057,
      "epoch": 0.4241493723440348,
      "grad_norm": 0.14409498870372772,
      "learning_rate": 8.586499831324563e-05,
      "loss": 0.6167,
      "mean_token_accuracy": 0.8105736933648586,
      "num_tokens": 147137390.0,
      "step": 4610
    },
    {
      "entropy": 0.8632991053164005,
      "epoch": 0.42424137871547607,
      "grad_norm": 0.16945143043994904,
      "learning_rate": 8.5861931487104e-05,
      "loss": 0.8393,
      "mean_token_accuracy": 0.7583488933742046,
      "num_tokens": 147168962.0,
      "step": 4611
    },
    {
      "entropy": 0.7813392616808414,
      "epoch": 0.42433338508691726,
      "grad_norm": 0.16051052510738373,
      "learning_rate": 8.585886466096238e-05,
      "loss": 0.7617,
      "mean_token_accuracy": 0.7804623991250992,
      "num_tokens": 147200602.0,
      "step": 4612
    },
    {
      "entropy": 0.674537681043148,
      "epoch": 0.4244253914583585,
      "grad_norm": 0.1570131480693817,
      "learning_rate": 8.585579783482075e-05,
      "loss": 0.6404,
      "mean_token_accuracy": 0.809480682015419,
      "num_tokens": 147232955.0,
      "step": 4613
    },
    {
      "entropy": 0.7665836736559868,
      "epoch": 0.4245173978297997,
      "grad_norm": 0.17355789244174957,
      "learning_rate": 8.585273100867912e-05,
      "loss": 0.7585,
      "mean_token_accuracy": 0.7749401219189167,
      "num_tokens": 147264436.0,
      "step": 4614
    },
    {
      "entropy": 0.7571151927113533,
      "epoch": 0.42460940420124094,
      "grad_norm": 0.1618274450302124,
      "learning_rate": 8.58496641825375e-05,
      "loss": 0.7518,
      "mean_token_accuracy": 0.7764874435961246,
      "num_tokens": 147295466.0,
      "step": 4615
    },
    {
      "entropy": 0.763299684971571,
      "epoch": 0.4247014105726822,
      "grad_norm": 0.15720979869365692,
      "learning_rate": 8.584659735639588e-05,
      "loss": 0.7553,
      "mean_token_accuracy": 0.7746320143342018,
      "num_tokens": 147327382.0,
      "step": 4616
    },
    {
      "entropy": 0.7637761551886797,
      "epoch": 0.42479341694412337,
      "grad_norm": 0.16678154468536377,
      "learning_rate": 8.584353053025425e-05,
      "loss": 0.7637,
      "mean_token_accuracy": 0.7792493365705013,
      "num_tokens": 147359723.0,
      "step": 4617
    },
    {
      "entropy": 0.7705184109508991,
      "epoch": 0.4248854233155646,
      "grad_norm": 0.16910067200660706,
      "learning_rate": 8.584046370411262e-05,
      "loss": 0.7916,
      "mean_token_accuracy": 0.7755454480648041,
      "num_tokens": 147391738.0,
      "step": 4618
    },
    {
      "entropy": 0.7558711152523756,
      "epoch": 0.4249774296870058,
      "grad_norm": 0.16342554986476898,
      "learning_rate": 8.583739687797098e-05,
      "loss": 0.7571,
      "mean_token_accuracy": 0.7791360728442669,
      "num_tokens": 147422792.0,
      "step": 4619
    },
    {
      "entropy": 0.8463205751031637,
      "epoch": 0.42506943605844705,
      "grad_norm": 0.16216330230236053,
      "learning_rate": 8.583433005182936e-05,
      "loss": 0.8527,
      "mean_token_accuracy": 0.758190106600523,
      "num_tokens": 147455142.0,
      "step": 4620
    },
    {
      "entropy": 0.8014037050306797,
      "epoch": 0.4251614424298883,
      "grad_norm": 0.169124573469162,
      "learning_rate": 8.583126322568775e-05,
      "loss": 0.7993,
      "mean_token_accuracy": 0.7659743055701256,
      "num_tokens": 147485743.0,
      "step": 4621
    },
    {
      "entropy": 0.8713971972465515,
      "epoch": 0.4252534488013295,
      "grad_norm": 0.16011381149291992,
      "learning_rate": 8.582819639954611e-05,
      "loss": 0.8602,
      "mean_token_accuracy": 0.7551388293504715,
      "num_tokens": 147517200.0,
      "step": 4622
    },
    {
      "entropy": 0.7241974268108606,
      "epoch": 0.4253454551727707,
      "grad_norm": 0.15819472074508667,
      "learning_rate": 8.582512957340448e-05,
      "loss": 0.7378,
      "mean_token_accuracy": 0.7868640683591366,
      "num_tokens": 147549634.0,
      "step": 4623
    },
    {
      "entropy": 0.8299671225249767,
      "epoch": 0.4254374615442119,
      "grad_norm": 0.16693611443042755,
      "learning_rate": 8.582206274726286e-05,
      "loss": 0.8287,
      "mean_token_accuracy": 0.7624267227947712,
      "num_tokens": 147582402.0,
      "step": 4624
    },
    {
      "entropy": 0.8488089572638273,
      "epoch": 0.42552946791565316,
      "grad_norm": 0.15589052438735962,
      "learning_rate": 8.581899592112123e-05,
      "loss": 0.8302,
      "mean_token_accuracy": 0.757773045450449,
      "num_tokens": 147614649.0,
      "step": 4625
    },
    {
      "entropy": 0.8596734292805195,
      "epoch": 0.4256214742870944,
      "grad_norm": 0.1586950272321701,
      "learning_rate": 8.581592909497961e-05,
      "loss": 0.8431,
      "mean_token_accuracy": 0.756834264844656,
      "num_tokens": 147646852.0,
      "step": 4626
    },
    {
      "entropy": 0.796437868848443,
      "epoch": 0.4257134806585356,
      "grad_norm": 0.1579326093196869,
      "learning_rate": 8.581286226883798e-05,
      "loss": 0.7795,
      "mean_token_accuracy": 0.7714277654886246,
      "num_tokens": 147678831.0,
      "step": 4627
    },
    {
      "entropy": 0.7353752292692661,
      "epoch": 0.42580548702997684,
      "grad_norm": 0.1520933359861374,
      "learning_rate": 8.580979544269636e-05,
      "loss": 0.7161,
      "mean_token_accuracy": 0.7868848331272602,
      "num_tokens": 147710529.0,
      "step": 4628
    },
    {
      "entropy": 0.8970924913883209,
      "epoch": 0.425897493401418,
      "grad_norm": 0.1580749750137329,
      "learning_rate": 8.580672861655473e-05,
      "loss": 0.8773,
      "mean_token_accuracy": 0.7448320277035236,
      "num_tokens": 147742743.0,
      "step": 4629
    },
    {
      "entropy": 0.9056255221366882,
      "epoch": 0.42598949977285927,
      "grad_norm": 0.15416285395622253,
      "learning_rate": 8.58036617904131e-05,
      "loss": 0.8839,
      "mean_token_accuracy": 0.7499679699540138,
      "num_tokens": 147774989.0,
      "step": 4630
    },
    {
      "entropy": 0.7314015738666058,
      "epoch": 0.4260815061443005,
      "grad_norm": 0.15937337279319763,
      "learning_rate": 8.580059496427148e-05,
      "loss": 0.72,
      "mean_token_accuracy": 0.7853183783590794,
      "num_tokens": 147806909.0,
      "step": 4631
    },
    {
      "entropy": 0.8560720942914486,
      "epoch": 0.4261735125157417,
      "grad_norm": 0.16234341263771057,
      "learning_rate": 8.579752813812986e-05,
      "loss": 0.8259,
      "mean_token_accuracy": 0.7575594745576382,
      "num_tokens": 147839009.0,
      "step": 4632
    },
    {
      "entropy": 0.8501491844654083,
      "epoch": 0.42626551888718295,
      "grad_norm": 0.15691418945789337,
      "learning_rate": 8.579446131198823e-05,
      "loss": 0.8378,
      "mean_token_accuracy": 0.7570584081113338,
      "num_tokens": 147870442.0,
      "step": 4633
    },
    {
      "entropy": 0.790016982704401,
      "epoch": 0.42635752525862414,
      "grad_norm": 0.15450997650623322,
      "learning_rate": 8.57913944858466e-05,
      "loss": 0.766,
      "mean_token_accuracy": 0.7813923172652721,
      "num_tokens": 147901360.0,
      "step": 4634
    },
    {
      "entropy": 0.8079939130693674,
      "epoch": 0.4264495316300654,
      "grad_norm": 0.1593797504901886,
      "learning_rate": 8.578832765970496e-05,
      "loss": 0.7846,
      "mean_token_accuracy": 0.7671005688607693,
      "num_tokens": 147932189.0,
      "step": 4635
    },
    {
      "entropy": 0.6736984439194202,
      "epoch": 0.4265415380015066,
      "grad_norm": 0.16153088212013245,
      "learning_rate": 8.578526083356336e-05,
      "loss": 0.6541,
      "mean_token_accuracy": 0.8061646036803722,
      "num_tokens": 147964215.0,
      "step": 4636
    },
    {
      "entropy": 0.8529234547168016,
      "epoch": 0.4266335443729478,
      "grad_norm": 0.16200190782546997,
      "learning_rate": 8.578219400742173e-05,
      "loss": 0.8574,
      "mean_token_accuracy": 0.7538869194686413,
      "num_tokens": 147996749.0,
      "step": 4637
    },
    {
      "entropy": 0.7433046437799931,
      "epoch": 0.42672555074438906,
      "grad_norm": 0.1579662710428238,
      "learning_rate": 8.57791271812801e-05,
      "loss": 0.7506,
      "mean_token_accuracy": 0.7804654613137245,
      "num_tokens": 148028972.0,
      "step": 4638
    },
    {
      "entropy": 0.8794772103428841,
      "epoch": 0.42681755711583025,
      "grad_norm": 0.17348234355449677,
      "learning_rate": 8.577606035513848e-05,
      "loss": 0.9065,
      "mean_token_accuracy": 0.7435971610248089,
      "num_tokens": 148060644.0,
      "step": 4639
    },
    {
      "entropy": 0.7210108377039433,
      "epoch": 0.4269095634872715,
      "grad_norm": 0.15141186118125916,
      "learning_rate": 8.577299352899684e-05,
      "loss": 0.7198,
      "mean_token_accuracy": 0.789498221129179,
      "num_tokens": 148092549.0,
      "step": 4640
    },
    {
      "entropy": 0.7006548922508955,
      "epoch": 0.42700156985871274,
      "grad_norm": 0.16246777772903442,
      "learning_rate": 8.576992670285523e-05,
      "loss": 0.6937,
      "mean_token_accuracy": 0.794311162084341,
      "num_tokens": 148124122.0,
      "step": 4641
    },
    {
      "entropy": 0.864555686712265,
      "epoch": 0.4270935762301539,
      "grad_norm": 0.1623639315366745,
      "learning_rate": 8.576685987671359e-05,
      "loss": 0.8559,
      "mean_token_accuracy": 0.7492375634610653,
      "num_tokens": 148156682.0,
      "step": 4642
    },
    {
      "entropy": 0.7734821122139692,
      "epoch": 0.42718558260159517,
      "grad_norm": 0.1608583629131317,
      "learning_rate": 8.576379305057197e-05,
      "loss": 0.7771,
      "mean_token_accuracy": 0.7685230076313019,
      "num_tokens": 148188746.0,
      "step": 4643
    },
    {
      "entropy": 0.7568129561841488,
      "epoch": 0.42727758897303636,
      "grad_norm": 0.15591523051261902,
      "learning_rate": 8.576072622443034e-05,
      "loss": 0.7435,
      "mean_token_accuracy": 0.7838151603937149,
      "num_tokens": 148220769.0,
      "step": 4644
    },
    {
      "entropy": 0.7581351287662983,
      "epoch": 0.4273695953444776,
      "grad_norm": 0.15677997469902039,
      "learning_rate": 8.575765939828871e-05,
      "loss": 0.7446,
      "mean_token_accuracy": 0.7806229963898659,
      "num_tokens": 148252716.0,
      "step": 4645
    },
    {
      "entropy": 0.7966628577560186,
      "epoch": 0.42746160171591885,
      "grad_norm": 0.1568538248538971,
      "learning_rate": 8.575459257214709e-05,
      "loss": 0.7801,
      "mean_token_accuracy": 0.7686630040407181,
      "num_tokens": 148283915.0,
      "step": 4646
    },
    {
      "entropy": 0.7955265920609236,
      "epoch": 0.42755360808736004,
      "grad_norm": 0.15458939969539642,
      "learning_rate": 8.575152574600547e-05,
      "loss": 0.7816,
      "mean_token_accuracy": 0.7693360261619091,
      "num_tokens": 148316381.0,
      "step": 4647
    },
    {
      "entropy": 0.7614725586026907,
      "epoch": 0.4276456144588013,
      "grad_norm": 0.1511397808790207,
      "learning_rate": 8.574845891986384e-05,
      "loss": 0.7421,
      "mean_token_accuracy": 0.7843933738768101,
      "num_tokens": 148348191.0,
      "step": 4648
    },
    {
      "entropy": 0.7948442231863737,
      "epoch": 0.42773762083024247,
      "grad_norm": 0.1624593436717987,
      "learning_rate": 8.574539209372221e-05,
      "loss": 0.7795,
      "mean_token_accuracy": 0.7740098387002945,
      "num_tokens": 148380426.0,
      "step": 4649
    },
    {
      "entropy": 0.6832427782937884,
      "epoch": 0.4278296272016837,
      "grad_norm": 0.1473504602909088,
      "learning_rate": 8.574232526758058e-05,
      "loss": 0.6465,
      "mean_token_accuracy": 0.8078856579959393,
      "num_tokens": 148412231.0,
      "step": 4650
    },
    {
      "entropy": 0.6993076112121344,
      "epoch": 0.42792163357312496,
      "grad_norm": 0.1530528962612152,
      "learning_rate": 8.573925844143896e-05,
      "loss": 0.6845,
      "mean_token_accuracy": 0.8020318485796452,
      "num_tokens": 148443810.0,
      "step": 4651
    },
    {
      "entropy": 0.8160107955336571,
      "epoch": 0.42801363994456615,
      "grad_norm": 0.16912797093391418,
      "learning_rate": 8.573619161529734e-05,
      "loss": 0.8322,
      "mean_token_accuracy": 0.7615578509867191,
      "num_tokens": 148475119.0,
      "step": 4652
    },
    {
      "entropy": 0.8919027801603079,
      "epoch": 0.4281056463160074,
      "grad_norm": 0.16659031808376312,
      "learning_rate": 8.573312478915571e-05,
      "loss": 0.9157,
      "mean_token_accuracy": 0.7390161752700806,
      "num_tokens": 148507878.0,
      "step": 4653
    },
    {
      "entropy": 0.7689816355705261,
      "epoch": 0.4281976526874486,
      "grad_norm": 0.16652674973011017,
      "learning_rate": 8.573005796301408e-05,
      "loss": 0.8087,
      "mean_token_accuracy": 0.7708467990159988,
      "num_tokens": 148540054.0,
      "step": 4654
    },
    {
      "entropy": 0.8197981584817171,
      "epoch": 0.4282896590588898,
      "grad_norm": 0.16226164996623993,
      "learning_rate": 8.572699113687246e-05,
      "loss": 0.8203,
      "mean_token_accuracy": 0.7620999440550804,
      "num_tokens": 148572247.0,
      "step": 4655
    },
    {
      "entropy": 0.7783125452697277,
      "epoch": 0.42838166543033107,
      "grad_norm": 0.16882121562957764,
      "learning_rate": 8.572392431073082e-05,
      "loss": 0.7658,
      "mean_token_accuracy": 0.7721019461750984,
      "num_tokens": 148604190.0,
      "step": 4656
    },
    {
      "entropy": 0.7673468887805939,
      "epoch": 0.42847367180177226,
      "grad_norm": 0.15153905749320984,
      "learning_rate": 8.57208574845892e-05,
      "loss": 0.7438,
      "mean_token_accuracy": 0.7774068042635918,
      "num_tokens": 148635535.0,
      "step": 4657
    },
    {
      "entropy": 0.8608374204486609,
      "epoch": 0.4285656781732135,
      "grad_norm": 0.15700213611125946,
      "learning_rate": 8.571779065844757e-05,
      "loss": 0.8488,
      "mean_token_accuracy": 0.7557462975382805,
      "num_tokens": 148667745.0,
      "step": 4658
    },
    {
      "entropy": 0.8797313198447227,
      "epoch": 0.4286576845446547,
      "grad_norm": 0.15872950851917267,
      "learning_rate": 8.571472383230596e-05,
      "loss": 0.8752,
      "mean_token_accuracy": 0.7503047734498978,
      "num_tokens": 148699498.0,
      "step": 4659
    },
    {
      "entropy": 0.8813257776200771,
      "epoch": 0.42874969091609594,
      "grad_norm": 0.15805450081825256,
      "learning_rate": 8.571165700616432e-05,
      "loss": 0.8629,
      "mean_token_accuracy": 0.7508545629680157,
      "num_tokens": 148731612.0,
      "step": 4660
    },
    {
      "entropy": 0.8960307016968727,
      "epoch": 0.4288416972875372,
      "grad_norm": 0.16184832155704498,
      "learning_rate": 8.570859018002269e-05,
      "loss": 0.8627,
      "mean_token_accuracy": 0.7487686835229397,
      "num_tokens": 148763299.0,
      "step": 4661
    },
    {
      "entropy": 0.767580471932888,
      "epoch": 0.42893370365897837,
      "grad_norm": 0.15531283617019653,
      "learning_rate": 8.570552335388107e-05,
      "loss": 0.7526,
      "mean_token_accuracy": 0.7822843641042709,
      "num_tokens": 148795861.0,
      "step": 4662
    },
    {
      "entropy": 0.9275734163820744,
      "epoch": 0.4290257100304196,
      "grad_norm": 0.16419555246829987,
      "learning_rate": 8.570245652773945e-05,
      "loss": 0.9223,
      "mean_token_accuracy": 0.7343013994395733,
      "num_tokens": 148828074.0,
      "step": 4663
    },
    {
      "entropy": 0.8748325183987617,
      "epoch": 0.4291177164018608,
      "grad_norm": 0.16466861963272095,
      "learning_rate": 8.569938970159782e-05,
      "loss": 0.8829,
      "mean_token_accuracy": 0.7459830492734909,
      "num_tokens": 148860298.0,
      "step": 4664
    },
    {
      "entropy": 0.8054137043654919,
      "epoch": 0.42920972277330205,
      "grad_norm": 0.1603316068649292,
      "learning_rate": 8.569632287545619e-05,
      "loss": 0.8158,
      "mean_token_accuracy": 0.7701255418360233,
      "num_tokens": 148892128.0,
      "step": 4665
    },
    {
      "entropy": 0.8207780122756958,
      "epoch": 0.4293017291447433,
      "grad_norm": 0.15869870781898499,
      "learning_rate": 8.569325604931456e-05,
      "loss": 0.819,
      "mean_token_accuracy": 0.7614870592951775,
      "num_tokens": 148924543.0,
      "step": 4666
    },
    {
      "entropy": 0.9153033532202244,
      "epoch": 0.4293937355161845,
      "grad_norm": 0.16945548355579376,
      "learning_rate": 8.569018922317294e-05,
      "loss": 0.932,
      "mean_token_accuracy": 0.7346658669412136,
      "num_tokens": 148956654.0,
      "step": 4667
    },
    {
      "entropy": 0.8417079374194145,
      "epoch": 0.4294857418876257,
      "grad_norm": 0.1542143076658249,
      "learning_rate": 8.568712239703132e-05,
      "loss": 0.8337,
      "mean_token_accuracy": 0.7604818046092987,
      "num_tokens": 148988419.0,
      "step": 4668
    },
    {
      "entropy": 0.8327589016407728,
      "epoch": 0.4295777482590669,
      "grad_norm": 0.15841051936149597,
      "learning_rate": 8.568405557088969e-05,
      "loss": 0.8137,
      "mean_token_accuracy": 0.7665178254246712,
      "num_tokens": 149021183.0,
      "step": 4669
    },
    {
      "entropy": 0.829756261780858,
      "epoch": 0.42966975463050816,
      "grad_norm": 0.15715663135051727,
      "learning_rate": 8.568098874474807e-05,
      "loss": 0.8204,
      "mean_token_accuracy": 0.7636994272470474,
      "num_tokens": 149053230.0,
      "step": 4670
    },
    {
      "entropy": 0.8022433426231146,
      "epoch": 0.4297617610019494,
      "grad_norm": 0.15193212032318115,
      "learning_rate": 8.567792191860644e-05,
      "loss": 0.7735,
      "mean_token_accuracy": 0.7746239453554153,
      "num_tokens": 149085560.0,
      "step": 4671
    },
    {
      "entropy": 0.7933313436806202,
      "epoch": 0.4298537673733906,
      "grad_norm": 0.15679767727851868,
      "learning_rate": 8.56748550924648e-05,
      "loss": 0.771,
      "mean_token_accuracy": 0.7743867337703705,
      "num_tokens": 149117262.0,
      "step": 4672
    },
    {
      "entropy": 0.7440774887800217,
      "epoch": 0.42994577374483184,
      "grad_norm": 0.15412816405296326,
      "learning_rate": 8.567178826632319e-05,
      "loss": 0.7234,
      "mean_token_accuracy": 0.7876635529100895,
      "num_tokens": 149149325.0,
      "step": 4673
    },
    {
      "entropy": 0.8150984980165958,
      "epoch": 0.430037780116273,
      "grad_norm": 0.16799621284008026,
      "learning_rate": 8.566872144018157e-05,
      "loss": 0.8381,
      "mean_token_accuracy": 0.7565007135272026,
      "num_tokens": 149181389.0,
      "step": 4674
    },
    {
      "entropy": 0.8659826964139938,
      "epoch": 0.43012978648771427,
      "grad_norm": 0.15676386654376984,
      "learning_rate": 8.566565461403994e-05,
      "loss": 0.8574,
      "mean_token_accuracy": 0.7593124099075794,
      "num_tokens": 149214157.0,
      "step": 4675
    },
    {
      "entropy": 0.8996595200151205,
      "epoch": 0.4302217928591555,
      "grad_norm": 0.15889915823936462,
      "learning_rate": 8.56625877878983e-05,
      "loss": 0.8849,
      "mean_token_accuracy": 0.7455117516219616,
      "num_tokens": 149245064.0,
      "step": 4676
    },
    {
      "entropy": 0.8075528591871262,
      "epoch": 0.4303137992305967,
      "grad_norm": 0.5708184242248535,
      "learning_rate": 8.565952096175667e-05,
      "loss": 0.7839,
      "mean_token_accuracy": 0.7728626057505608,
      "num_tokens": 149277261.0,
      "step": 4677
    },
    {
      "entropy": 0.8302259277552366,
      "epoch": 0.43040580560203795,
      "grad_norm": 0.1743858903646469,
      "learning_rate": 8.565645413561507e-05,
      "loss": 0.8481,
      "mean_token_accuracy": 0.756588526070118,
      "num_tokens": 149309537.0,
      "step": 4678
    },
    {
      "entropy": 0.8359114155173302,
      "epoch": 0.43049781197347914,
      "grad_norm": 0.15659821033477783,
      "learning_rate": 8.565338730947343e-05,
      "loss": 0.8356,
      "mean_token_accuracy": 0.7582040317356586,
      "num_tokens": 149342018.0,
      "step": 4679
    },
    {
      "entropy": 0.8314057998359203,
      "epoch": 0.4305898183449204,
      "grad_norm": 0.16471658647060394,
      "learning_rate": 8.56503204833318e-05,
      "loss": 0.8184,
      "mean_token_accuracy": 0.7571682445704937,
      "num_tokens": 149374092.0,
      "step": 4680
    },
    {
      "entropy": 0.6771348547190428,
      "epoch": 0.4306818247163616,
      "grad_norm": 0.1696586310863495,
      "learning_rate": 8.564725365719017e-05,
      "loss": 0.67,
      "mean_token_accuracy": 0.8013429567217827,
      "num_tokens": 149406476.0,
      "step": 4681
    },
    {
      "entropy": 0.6836149860173464,
      "epoch": 0.4307738310878028,
      "grad_norm": 0.14521819353103638,
      "learning_rate": 8.564418683104855e-05,
      "loss": 0.6756,
      "mean_token_accuracy": 0.7989351116120815,
      "num_tokens": 149438757.0,
      "step": 4682
    },
    {
      "entropy": 0.7101553231477737,
      "epoch": 0.43086583745924406,
      "grad_norm": 0.16063115000724792,
      "learning_rate": 8.564112000490693e-05,
      "loss": 0.6935,
      "mean_token_accuracy": 0.7916360534727573,
      "num_tokens": 149470042.0,
      "step": 4683
    },
    {
      "entropy": 0.8413604944944382,
      "epoch": 0.43095784383068525,
      "grad_norm": 0.17201068997383118,
      "learning_rate": 8.56380531787653e-05,
      "loss": 0.87,
      "mean_token_accuracy": 0.7559412643313408,
      "num_tokens": 149502682.0,
      "step": 4684
    },
    {
      "entropy": 0.8052816409617662,
      "epoch": 0.4310498502021265,
      "grad_norm": 0.17124561965465546,
      "learning_rate": 8.563498635262367e-05,
      "loss": 0.7881,
      "mean_token_accuracy": 0.7714412696659565,
      "num_tokens": 149534353.0,
      "step": 4685
    },
    {
      "entropy": 0.8808044698089361,
      "epoch": 0.43114185657356774,
      "grad_norm": 0.1582329422235489,
      "learning_rate": 8.563191952648205e-05,
      "loss": 0.8828,
      "mean_token_accuracy": 0.746374350041151,
      "num_tokens": 149566864.0,
      "step": 4686
    },
    {
      "entropy": 0.7647868786007166,
      "epoch": 0.43123386294500893,
      "grad_norm": 0.15566079318523407,
      "learning_rate": 8.562885270034042e-05,
      "loss": 0.7602,
      "mean_token_accuracy": 0.7781843692064285,
      "num_tokens": 149598989.0,
      "step": 4687
    },
    {
      "entropy": 0.8109480738639832,
      "epoch": 0.4313258693164502,
      "grad_norm": 0.15668648481369019,
      "learning_rate": 8.56257858741988e-05,
      "loss": 0.8119,
      "mean_token_accuracy": 0.7627065554261208,
      "num_tokens": 149631193.0,
      "step": 4688
    },
    {
      "entropy": 0.7557604480534792,
      "epoch": 0.43141787568789136,
      "grad_norm": 0.1543303281068802,
      "learning_rate": 8.562271904805717e-05,
      "loss": 0.752,
      "mean_token_accuracy": 0.7749083563685417,
      "num_tokens": 149662727.0,
      "step": 4689
    },
    {
      "entropy": 0.7809812277555466,
      "epoch": 0.4315098820593326,
      "grad_norm": 0.1488116979598999,
      "learning_rate": 8.561965222191555e-05,
      "loss": 0.7749,
      "mean_token_accuracy": 0.7762957289814949,
      "num_tokens": 149695108.0,
      "step": 4690
    },
    {
      "entropy": 0.7836719155311584,
      "epoch": 0.43160188843077385,
      "grad_norm": 0.15261292457580566,
      "learning_rate": 8.561658539577392e-05,
      "loss": 0.758,
      "mean_token_accuracy": 0.7793101519346237,
      "num_tokens": 149726781.0,
      "step": 4691
    },
    {
      "entropy": 0.7687590289860964,
      "epoch": 0.43169389480221504,
      "grad_norm": 0.1536664515733719,
      "learning_rate": 8.561351856963228e-05,
      "loss": 0.7292,
      "mean_token_accuracy": 0.788710705935955,
      "num_tokens": 149758977.0,
      "step": 4692
    },
    {
      "entropy": 0.8875763118267059,
      "epoch": 0.4317859011736563,
      "grad_norm": 0.15894676744937897,
      "learning_rate": 8.561045174349067e-05,
      "loss": 0.8749,
      "mean_token_accuracy": 0.7482263147830963,
      "num_tokens": 149790534.0,
      "step": 4693
    },
    {
      "entropy": 0.8394260071218014,
      "epoch": 0.4318779075450975,
      "grad_norm": 0.16099049150943756,
      "learning_rate": 8.560738491734905e-05,
      "loss": 0.8352,
      "mean_token_accuracy": 0.7609627544879913,
      "num_tokens": 149822956.0,
      "step": 4694
    },
    {
      "entropy": 0.8911239188164473,
      "epoch": 0.4319699139165387,
      "grad_norm": 0.16837310791015625,
      "learning_rate": 8.560431809120742e-05,
      "loss": 0.8788,
      "mean_token_accuracy": 0.7469450980424881,
      "num_tokens": 149854012.0,
      "step": 4695
    },
    {
      "entropy": 0.7764223739504814,
      "epoch": 0.43206192028797996,
      "grad_norm": 0.16536866128444672,
      "learning_rate": 8.560125126506578e-05,
      "loss": 0.7668,
      "mean_token_accuracy": 0.7773840576410294,
      "num_tokens": 149885867.0,
      "step": 4696
    },
    {
      "entropy": 0.7501247599720955,
      "epoch": 0.43215392665942115,
      "grad_norm": 0.16054317355155945,
      "learning_rate": 8.559818443892415e-05,
      "loss": 0.735,
      "mean_token_accuracy": 0.7856106720864773,
      "num_tokens": 149917424.0,
      "step": 4697
    },
    {
      "entropy": 0.8092388957738876,
      "epoch": 0.4322459330308624,
      "grad_norm": 0.15862822532653809,
      "learning_rate": 8.559511761278253e-05,
      "loss": 0.8226,
      "mean_token_accuracy": 0.7631352208554745,
      "num_tokens": 149949523.0,
      "step": 4698
    },
    {
      "entropy": 0.8412404544651508,
      "epoch": 0.4323379394023036,
      "grad_norm": 0.16467128694057465,
      "learning_rate": 8.559205078664091e-05,
      "loss": 0.8307,
      "mean_token_accuracy": 0.7626721747219563,
      "num_tokens": 149981504.0,
      "step": 4699
    },
    {
      "entropy": 0.7238278388977051,
      "epoch": 0.43242994577374483,
      "grad_norm": 0.16265246272087097,
      "learning_rate": 8.558898396049928e-05,
      "loss": 0.7182,
      "mean_token_accuracy": 0.79255535826087,
      "num_tokens": 150013703.0,
      "step": 4700
    },
    {
      "entropy": 0.7189370542764664,
      "epoch": 0.4325219521451861,
      "grad_norm": 0.16005480289459229,
      "learning_rate": 8.558591713435766e-05,
      "loss": 0.7235,
      "mean_token_accuracy": 0.785461850464344,
      "num_tokens": 150045238.0,
      "step": 4701
    },
    {
      "entropy": 0.8666659295558929,
      "epoch": 0.43261395851662726,
      "grad_norm": 0.15947146713733673,
      "learning_rate": 8.558285030821603e-05,
      "loss": 0.8506,
      "mean_token_accuracy": 0.7491436526179314,
      "num_tokens": 150076578.0,
      "step": 4702
    },
    {
      "entropy": 0.8473078981041908,
      "epoch": 0.4327059648880685,
      "grad_norm": 0.15863639116287231,
      "learning_rate": 8.55797834820744e-05,
      "loss": 0.8621,
      "mean_token_accuracy": 0.7505077496170998,
      "num_tokens": 150108473.0,
      "step": 4703
    },
    {
      "entropy": 0.6974059287458658,
      "epoch": 0.4327979712595097,
      "grad_norm": 0.16399797797203064,
      "learning_rate": 8.557671665593278e-05,
      "loss": 0.693,
      "mean_token_accuracy": 0.7972534485161304,
      "num_tokens": 150140853.0,
      "step": 4704
    },
    {
      "entropy": 0.8172098398208618,
      "epoch": 0.43288997763095094,
      "grad_norm": 0.15929129719734192,
      "learning_rate": 8.557364982979116e-05,
      "loss": 0.8007,
      "mean_token_accuracy": 0.7659633941948414,
      "num_tokens": 150172897.0,
      "step": 4705
    },
    {
      "entropy": 0.8716510646045208,
      "epoch": 0.4329819840023922,
      "grad_norm": 0.1725178360939026,
      "learning_rate": 8.557058300364953e-05,
      "loss": 0.8646,
      "mean_token_accuracy": 0.7588584534823895,
      "num_tokens": 150204834.0,
      "step": 4706
    },
    {
      "entropy": 0.7693596314638853,
      "epoch": 0.4330739903738334,
      "grad_norm": 0.15682809054851532,
      "learning_rate": 8.55675161775079e-05,
      "loss": 0.7849,
      "mean_token_accuracy": 0.7706845477223396,
      "num_tokens": 150236154.0,
      "step": 4707
    },
    {
      "entropy": 0.6986318714916706,
      "epoch": 0.4331659967452746,
      "grad_norm": 0.15159638226032257,
      "learning_rate": 8.556444935136627e-05,
      "loss": 0.6791,
      "mean_token_accuracy": 0.7948814146220684,
      "num_tokens": 150267996.0,
      "step": 4708
    },
    {
      "entropy": 0.7980844508856535,
      "epoch": 0.4332580031167158,
      "grad_norm": 0.16267181932926178,
      "learning_rate": 8.556138252522465e-05,
      "loss": 0.8077,
      "mean_token_accuracy": 0.7642902918159962,
      "num_tokens": 150299781.0,
      "step": 4709
    },
    {
      "entropy": 0.7623817529529333,
      "epoch": 0.43335000948815705,
      "grad_norm": 0.162998229265213,
      "learning_rate": 8.555831569908303e-05,
      "loss": 0.7378,
      "mean_token_accuracy": 0.7841429077088833,
      "num_tokens": 150331959.0,
      "step": 4710
    },
    {
      "entropy": 0.8343127518892288,
      "epoch": 0.4334420158595983,
      "grad_norm": 0.16318976879119873,
      "learning_rate": 8.55552488729414e-05,
      "loss": 0.8276,
      "mean_token_accuracy": 0.7603723704814911,
      "num_tokens": 150363171.0,
      "step": 4711
    },
    {
      "entropy": 0.8071335759013891,
      "epoch": 0.4335340222310395,
      "grad_norm": 0.1523260474205017,
      "learning_rate": 8.555218204679976e-05,
      "loss": 0.794,
      "mean_token_accuracy": 0.7686784975230694,
      "num_tokens": 150395115.0,
      "step": 4712
    },
    {
      "entropy": 0.7951165772974491,
      "epoch": 0.43362602860248073,
      "grad_norm": 0.15072591602802277,
      "learning_rate": 8.554911522065815e-05,
      "loss": 0.781,
      "mean_token_accuracy": 0.7701710127294064,
      "num_tokens": 150427435.0,
      "step": 4713
    },
    {
      "entropy": 0.8092553541064262,
      "epoch": 0.4337180349739219,
      "grad_norm": 0.1590370237827301,
      "learning_rate": 8.554604839451651e-05,
      "loss": 0.7984,
      "mean_token_accuracy": 0.7650387957692146,
      "num_tokens": 150459855.0,
      "step": 4714
    },
    {
      "entropy": 0.875880990177393,
      "epoch": 0.43381004134536316,
      "grad_norm": 0.15786035358905792,
      "learning_rate": 8.55429815683749e-05,
      "loss": 0.8631,
      "mean_token_accuracy": 0.7515531927347183,
      "num_tokens": 150490979.0,
      "step": 4715
    },
    {
      "entropy": 0.8424091748893261,
      "epoch": 0.4339020477168044,
      "grad_norm": 0.1619425266981125,
      "learning_rate": 8.553991474223326e-05,
      "loss": 0.8409,
      "mean_token_accuracy": 0.7569734267890453,
      "num_tokens": 150523065.0,
      "step": 4716
    },
    {
      "entropy": 0.8644343465566635,
      "epoch": 0.4339940540882456,
      "grad_norm": 0.16668856143951416,
      "learning_rate": 8.553684791609164e-05,
      "loss": 0.8549,
      "mean_token_accuracy": 0.7514769770205021,
      "num_tokens": 150554467.0,
      "step": 4717
    },
    {
      "entropy": 0.8935752175748348,
      "epoch": 0.43408606045968684,
      "grad_norm": 0.1650361567735672,
      "learning_rate": 8.553378108995001e-05,
      "loss": 0.8838,
      "mean_token_accuracy": 0.7479878291487694,
      "num_tokens": 150586788.0,
      "step": 4718
    },
    {
      "entropy": 0.9369134232401848,
      "epoch": 0.43417806683112803,
      "grad_norm": 0.16090576350688934,
      "learning_rate": 8.55307142638084e-05,
      "loss": 0.9133,
      "mean_token_accuracy": 0.7387399300932884,
      "num_tokens": 150618752.0,
      "step": 4719
    },
    {
      "entropy": 0.7373160682618618,
      "epoch": 0.4342700732025693,
      "grad_norm": 0.15626633167266846,
      "learning_rate": 8.552764743766676e-05,
      "loss": 0.7131,
      "mean_token_accuracy": 0.7910356558859348,
      "num_tokens": 150651520.0,
      "step": 4720
    },
    {
      "entropy": 0.8741193562746048,
      "epoch": 0.4343620795740105,
      "grad_norm": 0.15491756796836853,
      "learning_rate": 8.552458061152514e-05,
      "loss": 0.8747,
      "mean_token_accuracy": 0.7485405132174492,
      "num_tokens": 150683687.0,
      "step": 4721
    },
    {
      "entropy": 0.7261905651539564,
      "epoch": 0.4344540859454517,
      "grad_norm": 0.1493799090385437,
      "learning_rate": 8.552151378538351e-05,
      "loss": 0.714,
      "mean_token_accuracy": 0.7899894341826439,
      "num_tokens": 150715810.0,
      "step": 4722
    },
    {
      "entropy": 0.7515033222734928,
      "epoch": 0.43454609231689295,
      "grad_norm": 0.14950476586818695,
      "learning_rate": 8.551844695924188e-05,
      "loss": 0.7429,
      "mean_token_accuracy": 0.7803672850131989,
      "num_tokens": 150748154.0,
      "step": 4723
    },
    {
      "entropy": 0.8241373021155596,
      "epoch": 0.43463809868833414,
      "grad_norm": 0.1634214222431183,
      "learning_rate": 8.551538013310026e-05,
      "loss": 0.8203,
      "mean_token_accuracy": 0.762108575552702,
      "num_tokens": 150780387.0,
      "step": 4724
    },
    {
      "entropy": 0.7501262072473764,
      "epoch": 0.4347301050597754,
      "grad_norm": 0.15816329419612885,
      "learning_rate": 8.551231330695864e-05,
      "loss": 0.762,
      "mean_token_accuracy": 0.7749011404812336,
      "num_tokens": 150811883.0,
      "step": 4725
    },
    {
      "entropy": 0.7763395365327597,
      "epoch": 0.43482211143121663,
      "grad_norm": 0.15731342136859894,
      "learning_rate": 8.550924648081701e-05,
      "loss": 0.7878,
      "mean_token_accuracy": 0.7759186662733555,
      "num_tokens": 150843841.0,
      "step": 4726
    },
    {
      "entropy": 0.7499911841005087,
      "epoch": 0.4349141178026578,
      "grad_norm": 0.15760089457035065,
      "learning_rate": 8.550617965467538e-05,
      "loss": 0.7528,
      "mean_token_accuracy": 0.7778587676584721,
      "num_tokens": 150875721.0,
      "step": 4727
    },
    {
      "entropy": 0.7814024835824966,
      "epoch": 0.43500612417409906,
      "grad_norm": 0.16215713322162628,
      "learning_rate": 8.550311282853375e-05,
      "loss": 0.7715,
      "mean_token_accuracy": 0.7755322605371475,
      "num_tokens": 150906863.0,
      "step": 4728
    },
    {
      "entropy": 0.8015024289488792,
      "epoch": 0.43509813054554025,
      "grad_norm": 0.16746953129768372,
      "learning_rate": 8.550004600239213e-05,
      "loss": 0.8108,
      "mean_token_accuracy": 0.7679606080055237,
      "num_tokens": 150938454.0,
      "step": 4729
    },
    {
      "entropy": 0.820079093798995,
      "epoch": 0.4351901369169815,
      "grad_norm": 0.1614924967288971,
      "learning_rate": 8.549697917625051e-05,
      "loss": 0.8231,
      "mean_token_accuracy": 0.7601314336061478,
      "num_tokens": 150970893.0,
      "step": 4730
    },
    {
      "entropy": 0.7493061348795891,
      "epoch": 0.43528214328842274,
      "grad_norm": 0.16535025835037231,
      "learning_rate": 8.549391235010888e-05,
      "loss": 0.7603,
      "mean_token_accuracy": 0.7811987660825253,
      "num_tokens": 151003025.0,
      "step": 4731
    },
    {
      "entropy": 0.7720494996756315,
      "epoch": 0.43537414965986393,
      "grad_norm": 0.15942054986953735,
      "learning_rate": 8.549084552396726e-05,
      "loss": 0.7457,
      "mean_token_accuracy": 0.7799035273492336,
      "num_tokens": 151035129.0,
      "step": 4732
    },
    {
      "entropy": 0.7238975688815117,
      "epoch": 0.4354661560313052,
      "grad_norm": 0.1594621241092682,
      "learning_rate": 8.548777869782563e-05,
      "loss": 0.7168,
      "mean_token_accuracy": 0.7831969484686852,
      "num_tokens": 151067341.0,
      "step": 4733
    },
    {
      "entropy": 0.813464317470789,
      "epoch": 0.43555816240274636,
      "grad_norm": 0.1562364250421524,
      "learning_rate": 8.548471187168399e-05,
      "loss": 0.792,
      "mean_token_accuracy": 0.7702169977128506,
      "num_tokens": 151099294.0,
      "step": 4734
    },
    {
      "entropy": 0.7381563633680344,
      "epoch": 0.4356501687741876,
      "grad_norm": 0.1566087007522583,
      "learning_rate": 8.548164504554237e-05,
      "loss": 0.7307,
      "mean_token_accuracy": 0.7879578210413456,
      "num_tokens": 151130898.0,
      "step": 4735
    },
    {
      "entropy": 0.7959471344947815,
      "epoch": 0.43574217514562885,
      "grad_norm": 0.15512970089912415,
      "learning_rate": 8.547857821940076e-05,
      "loss": 0.7804,
      "mean_token_accuracy": 0.7789716832339764,
      "num_tokens": 151163081.0,
      "step": 4736
    },
    {
      "entropy": 0.7835030574351549,
      "epoch": 0.43583418151707004,
      "grad_norm": 0.16127976775169373,
      "learning_rate": 8.547551139325912e-05,
      "loss": 0.7582,
      "mean_token_accuracy": 0.7777202352881432,
      "num_tokens": 151194952.0,
      "step": 4737
    },
    {
      "entropy": 0.8112133536487818,
      "epoch": 0.4359261878885113,
      "grad_norm": 0.15452940762043,
      "learning_rate": 8.547244456711749e-05,
      "loss": 0.8009,
      "mean_token_accuracy": 0.7654952183365822,
      "num_tokens": 151226944.0,
      "step": 4738
    },
    {
      "entropy": 0.861429825425148,
      "epoch": 0.4360181942599525,
      "grad_norm": 0.16686268150806427,
      "learning_rate": 8.546937774097586e-05,
      "loss": 0.8436,
      "mean_token_accuracy": 0.7558138184249401,
      "num_tokens": 151259479.0,
      "step": 4739
    },
    {
      "entropy": 0.7244261261075735,
      "epoch": 0.4361102006313937,
      "grad_norm": 0.1548422873020172,
      "learning_rate": 8.546631091483424e-05,
      "loss": 0.7178,
      "mean_token_accuracy": 0.7959474734961987,
      "num_tokens": 151290877.0,
      "step": 4740
    },
    {
      "entropy": 0.8548797313123941,
      "epoch": 0.43620220700283496,
      "grad_norm": 0.15863099694252014,
      "learning_rate": 8.546324408869262e-05,
      "loss": 0.8562,
      "mean_token_accuracy": 0.750369492918253,
      "num_tokens": 151322917.0,
      "step": 4741
    },
    {
      "entropy": 0.7791571691632271,
      "epoch": 0.43629421337427615,
      "grad_norm": 0.15428714454174042,
      "learning_rate": 8.546017726255099e-05,
      "loss": 0.7638,
      "mean_token_accuracy": 0.7723629921674728,
      "num_tokens": 151355040.0,
      "step": 4742
    },
    {
      "entropy": 0.7898006904870272,
      "epoch": 0.4363862197457174,
      "grad_norm": 0.15450982749462128,
      "learning_rate": 8.545711043640936e-05,
      "loss": 0.7735,
      "mean_token_accuracy": 0.7748834826052189,
      "num_tokens": 151386760.0,
      "step": 4743
    },
    {
      "entropy": 0.6891493797302246,
      "epoch": 0.43647822611715864,
      "grad_norm": 0.15753257274627686,
      "learning_rate": 8.545404361026774e-05,
      "loss": 0.6696,
      "mean_token_accuracy": 0.8030851744115353,
      "num_tokens": 151419037.0,
      "step": 4744
    },
    {
      "entropy": 0.7723507359623909,
      "epoch": 0.43657023248859983,
      "grad_norm": 0.15320102870464325,
      "learning_rate": 8.545097678412611e-05,
      "loss": 0.7773,
      "mean_token_accuracy": 0.7732631638646126,
      "num_tokens": 151451609.0,
      "step": 4745
    },
    {
      "entropy": 0.6957397721707821,
      "epoch": 0.4366622388600411,
      "grad_norm": 0.14998342096805573,
      "learning_rate": 8.544790995798449e-05,
      "loss": 0.6803,
      "mean_token_accuracy": 0.7989898286759853,
      "num_tokens": 151483749.0,
      "step": 4746
    },
    {
      "entropy": 0.7721174769103527,
      "epoch": 0.43675424523148226,
      "grad_norm": 0.15847167372703552,
      "learning_rate": 8.544484313184286e-05,
      "loss": 0.7742,
      "mean_token_accuracy": 0.7717486396431923,
      "num_tokens": 151515888.0,
      "step": 4747
    },
    {
      "entropy": 0.7962457314133644,
      "epoch": 0.4368462516029235,
      "grad_norm": 0.1562553346157074,
      "learning_rate": 8.544177630570124e-05,
      "loss": 0.7967,
      "mean_token_accuracy": 0.7646395154297352,
      "num_tokens": 151547330.0,
      "step": 4748
    },
    {
      "entropy": 0.734290175139904,
      "epoch": 0.43693825797436475,
      "grad_norm": 0.16224025189876556,
      "learning_rate": 8.54387094795596e-05,
      "loss": 0.7402,
      "mean_token_accuracy": 0.7825392819941044,
      "num_tokens": 151579278.0,
      "step": 4749
    },
    {
      "entropy": 0.8370578270405531,
      "epoch": 0.43703026434580594,
      "grad_norm": 0.15599176287651062,
      "learning_rate": 8.543564265341797e-05,
      "loss": 0.8429,
      "mean_token_accuracy": 0.7572980113327503,
      "num_tokens": 151610987.0,
      "step": 4750
    },
    {
      "entropy": 0.796088358387351,
      "epoch": 0.4371222707172472,
      "grad_norm": 0.1607772707939148,
      "learning_rate": 8.543257582727636e-05,
      "loss": 0.8045,
      "mean_token_accuracy": 0.7677928693592548,
      "num_tokens": 151642862.0,
      "step": 4751
    },
    {
      "entropy": 0.8880335167050362,
      "epoch": 0.4372142770886884,
      "grad_norm": 0.16782091557979584,
      "learning_rate": 8.542950900113474e-05,
      "loss": 0.9108,
      "mean_token_accuracy": 0.74068358913064,
      "num_tokens": 151674955.0,
      "step": 4752
    },
    {
      "entropy": 0.7118587270379066,
      "epoch": 0.4373062834601296,
      "grad_norm": 0.14450190961360931,
      "learning_rate": 8.54264421749931e-05,
      "loss": 0.685,
      "mean_token_accuracy": 0.7973864153027534,
      "num_tokens": 151707723.0,
      "step": 4753
    },
    {
      "entropy": 0.8263829201459885,
      "epoch": 0.43739828983157086,
      "grad_norm": 0.15710324048995972,
      "learning_rate": 8.542337534885147e-05,
      "loss": 0.8135,
      "mean_token_accuracy": 0.7598453499376774,
      "num_tokens": 151739447.0,
      "step": 4754
    },
    {
      "entropy": 0.8764786124229431,
      "epoch": 0.43749029620301205,
      "grad_norm": 0.16961397230625153,
      "learning_rate": 8.542030852270984e-05,
      "loss": 0.8815,
      "mean_token_accuracy": 0.7506719306111336,
      "num_tokens": 151771137.0,
      "step": 4755
    },
    {
      "entropy": 0.7924650516360998,
      "epoch": 0.4375823025744533,
      "grad_norm": 0.15813428163528442,
      "learning_rate": 8.541724169656824e-05,
      "loss": 0.7756,
      "mean_token_accuracy": 0.7760673388838768,
      "num_tokens": 151803121.0,
      "step": 4756
    },
    {
      "entropy": 0.7913246769458055,
      "epoch": 0.4376743089458945,
      "grad_norm": 0.1600050926208496,
      "learning_rate": 8.54141748704266e-05,
      "loss": 0.7597,
      "mean_token_accuracy": 0.7773286364972591,
      "num_tokens": 151835602.0,
      "step": 4757
    },
    {
      "entropy": 0.9466858617961407,
      "epoch": 0.43776631531733573,
      "grad_norm": 0.16863282024860382,
      "learning_rate": 8.541110804428497e-05,
      "loss": 0.9319,
      "mean_token_accuracy": 0.7318245656788349,
      "num_tokens": 151867386.0,
      "step": 4758
    },
    {
      "entropy": 0.7729724626988173,
      "epoch": 0.437858321688777,
      "grad_norm": 0.16267797350883484,
      "learning_rate": 8.540804121814334e-05,
      "loss": 0.7672,
      "mean_token_accuracy": 0.7768966518342495,
      "num_tokens": 151899432.0,
      "step": 4759
    },
    {
      "entropy": 0.8050575964152813,
      "epoch": 0.43795032806021816,
      "grad_norm": 0.16061995923519135,
      "learning_rate": 8.540497439200172e-05,
      "loss": 0.8074,
      "mean_token_accuracy": 0.7666183486580849,
      "num_tokens": 151931776.0,
      "step": 4760
    },
    {
      "entropy": 0.7562859170138836,
      "epoch": 0.4380423344316594,
      "grad_norm": 0.15496082603931427,
      "learning_rate": 8.54019075658601e-05,
      "loss": 0.7248,
      "mean_token_accuracy": 0.7888966649770737,
      "num_tokens": 151964229.0,
      "step": 4761
    },
    {
      "entropy": 0.7822975125163794,
      "epoch": 0.4381343408031006,
      "grad_norm": 0.1546972543001175,
      "learning_rate": 8.539884073971847e-05,
      "loss": 0.7662,
      "mean_token_accuracy": 0.7751418277621269,
      "num_tokens": 151996449.0,
      "step": 4762
    },
    {
      "entropy": 0.8116856478154659,
      "epoch": 0.43822634717454184,
      "grad_norm": 0.15661627054214478,
      "learning_rate": 8.539577391357685e-05,
      "loss": 0.8069,
      "mean_token_accuracy": 0.7684183903038502,
      "num_tokens": 152028640.0,
      "step": 4763
    },
    {
      "entropy": 0.7306651398539543,
      "epoch": 0.4383183535459831,
      "grad_norm": 0.1537189483642578,
      "learning_rate": 8.539270708743522e-05,
      "loss": 0.7199,
      "mean_token_accuracy": 0.7879774123430252,
      "num_tokens": 152060189.0,
      "step": 4764
    },
    {
      "entropy": 0.7610232979059219,
      "epoch": 0.4384103599174243,
      "grad_norm": 0.1554393321275711,
      "learning_rate": 8.538964026129359e-05,
      "loss": 0.7526,
      "mean_token_accuracy": 0.7793011665344238,
      "num_tokens": 152091972.0,
      "step": 4765
    },
    {
      "entropy": 0.8267065398395061,
      "epoch": 0.4385023662888655,
      "grad_norm": 0.1648242175579071,
      "learning_rate": 8.538657343515197e-05,
      "loss": 0.8486,
      "mean_token_accuracy": 0.755374476313591,
      "num_tokens": 152123624.0,
      "step": 4766
    },
    {
      "entropy": 0.8233845718204975,
      "epoch": 0.4385943726603067,
      "grad_norm": 0.1624373197555542,
      "learning_rate": 8.538350660901035e-05,
      "loss": 0.8364,
      "mean_token_accuracy": 0.7596427015960217,
      "num_tokens": 152155296.0,
      "step": 4767
    },
    {
      "entropy": 0.5895988140255213,
      "epoch": 0.43868637903174795,
      "grad_norm": 0.1544891744852066,
      "learning_rate": 8.538043978286872e-05,
      "loss": 0.5657,
      "mean_token_accuracy": 0.834157470613718,
      "num_tokens": 152187532.0,
      "step": 4768
    },
    {
      "entropy": 0.7061627823859453,
      "epoch": 0.4387783854031892,
      "grad_norm": 0.14680983126163483,
      "learning_rate": 8.537737295672709e-05,
      "loss": 0.6996,
      "mean_token_accuracy": 0.790699802339077,
      "num_tokens": 152220300.0,
      "step": 4769
    },
    {
      "entropy": 0.9066670760512352,
      "epoch": 0.4388703917746304,
      "grad_norm": 0.16805610060691833,
      "learning_rate": 8.537430613058545e-05,
      "loss": 0.907,
      "mean_token_accuracy": 0.7421443164348602,
      "num_tokens": 152250850.0,
      "step": 4770
    },
    {
      "entropy": 0.782124899327755,
      "epoch": 0.43896239814607163,
      "grad_norm": 0.15229229629039764,
      "learning_rate": 8.537123930444383e-05,
      "loss": 0.7757,
      "mean_token_accuracy": 0.7741700112819672,
      "num_tokens": 152282930.0,
      "step": 4771
    },
    {
      "entropy": 0.7585091777145863,
      "epoch": 0.4390544045175128,
      "grad_norm": 0.15767216682434082,
      "learning_rate": 8.536817247830222e-05,
      "loss": 0.7402,
      "mean_token_accuracy": 0.7815235629677773,
      "num_tokens": 152314701.0,
      "step": 4772
    },
    {
      "entropy": 0.8989594429731369,
      "epoch": 0.43914641088895406,
      "grad_norm": 0.16780716180801392,
      "learning_rate": 8.536510565216058e-05,
      "loss": 0.9019,
      "mean_token_accuracy": 0.7382672131061554,
      "num_tokens": 152347111.0,
      "step": 4773
    },
    {
      "entropy": 0.819849805906415,
      "epoch": 0.4392384172603953,
      "grad_norm": 0.16227132081985474,
      "learning_rate": 8.536203882601895e-05,
      "loss": 0.8265,
      "mean_token_accuracy": 0.7587190493941307,
      "num_tokens": 152379030.0,
      "step": 4774
    },
    {
      "entropy": 0.6876540258526802,
      "epoch": 0.4393304236318365,
      "grad_norm": 0.16583429276943207,
      "learning_rate": 8.535897199987733e-05,
      "loss": 0.662,
      "mean_token_accuracy": 0.8047928959131241,
      "num_tokens": 152411134.0,
      "step": 4775
    },
    {
      "entropy": 0.9311277084052563,
      "epoch": 0.43942243000327774,
      "grad_norm": 0.16684646904468536,
      "learning_rate": 8.53559051737357e-05,
      "loss": 0.9241,
      "mean_token_accuracy": 0.7346858121454716,
      "num_tokens": 152442700.0,
      "step": 4776
    },
    {
      "entropy": 0.744076581671834,
      "epoch": 0.43951443637471893,
      "grad_norm": 0.1595582813024521,
      "learning_rate": 8.535283834759408e-05,
      "loss": 0.7545,
      "mean_token_accuracy": 0.7831289395689964,
      "num_tokens": 152474538.0,
      "step": 4777
    },
    {
      "entropy": 0.894266290590167,
      "epoch": 0.4396064427461602,
      "grad_norm": 0.1583837866783142,
      "learning_rate": 8.534977152145245e-05,
      "loss": 0.8623,
      "mean_token_accuracy": 0.7518083639442921,
      "num_tokens": 152505605.0,
      "step": 4778
    },
    {
      "entropy": 0.7174028418958187,
      "epoch": 0.4396984491176014,
      "grad_norm": 0.15987354516983032,
      "learning_rate": 8.534670469531083e-05,
      "loss": 0.6994,
      "mean_token_accuracy": 0.7943818084895611,
      "num_tokens": 152537582.0,
      "step": 4779
    },
    {
      "entropy": 0.8412028029561043,
      "epoch": 0.4397904554890426,
      "grad_norm": 0.15928657352924347,
      "learning_rate": 8.53436378691692e-05,
      "loss": 0.8357,
      "mean_token_accuracy": 0.7594193778932095,
      "num_tokens": 152569692.0,
      "step": 4780
    },
    {
      "entropy": 0.9015024583786726,
      "epoch": 0.43988246186048385,
      "grad_norm": 0.16087566316127777,
      "learning_rate": 8.534057104302757e-05,
      "loss": 0.8939,
      "mean_token_accuracy": 0.7404895834624767,
      "num_tokens": 152601950.0,
      "step": 4781
    },
    {
      "entropy": 0.817371316254139,
      "epoch": 0.43997446823192504,
      "grad_norm": 0.15927821397781372,
      "learning_rate": 8.533750421688595e-05,
      "loss": 0.8254,
      "mean_token_accuracy": 0.7593819238245487,
      "num_tokens": 152634056.0,
      "step": 4782
    },
    {
      "entropy": 0.7467988897114992,
      "epoch": 0.4400664746033663,
      "grad_norm": 0.15280425548553467,
      "learning_rate": 8.533443739074433e-05,
      "loss": 0.7313,
      "mean_token_accuracy": 0.7812128774821758,
      "num_tokens": 152666571.0,
      "step": 4783
    },
    {
      "entropy": 0.7353355381637812,
      "epoch": 0.44015848097480753,
      "grad_norm": 0.1543206125497818,
      "learning_rate": 8.53313705646027e-05,
      "loss": 0.7125,
      "mean_token_accuracy": 0.7876325733959675,
      "num_tokens": 152698309.0,
      "step": 4784
    },
    {
      "entropy": 0.7685670405626297,
      "epoch": 0.4402504873462487,
      "grad_norm": 0.16929492354393005,
      "learning_rate": 8.532830373846107e-05,
      "loss": 0.759,
      "mean_token_accuracy": 0.7803909406065941,
      "num_tokens": 152730427.0,
      "step": 4785
    },
    {
      "entropy": 0.8975858744233847,
      "epoch": 0.44034249371768996,
      "grad_norm": 0.16315285861492157,
      "learning_rate": 8.532523691231943e-05,
      "loss": 0.8728,
      "mean_token_accuracy": 0.7469593584537506,
      "num_tokens": 152762531.0,
      "step": 4786
    },
    {
      "entropy": 0.7851618155837059,
      "epoch": 0.44043450008913115,
      "grad_norm": 0.15663422644138336,
      "learning_rate": 8.532217008617782e-05,
      "loss": 0.7904,
      "mean_token_accuracy": 0.7655009552836418,
      "num_tokens": 152794100.0,
      "step": 4787
    },
    {
      "entropy": 0.8710892461240292,
      "epoch": 0.4405265064605724,
      "grad_norm": 0.15682177245616913,
      "learning_rate": 8.53191032600362e-05,
      "loss": 0.8362,
      "mean_token_accuracy": 0.7583495788276196,
      "num_tokens": 152826778.0,
      "step": 4788
    },
    {
      "entropy": 0.7671521231532097,
      "epoch": 0.44061851283201364,
      "grad_norm": 0.16404008865356445,
      "learning_rate": 8.531603643389456e-05,
      "loss": 0.7805,
      "mean_token_accuracy": 0.7708914950489998,
      "num_tokens": 152858905.0,
      "step": 4789
    },
    {
      "entropy": 0.7773335240781307,
      "epoch": 0.44071051920345483,
      "grad_norm": 0.16303327679634094,
      "learning_rate": 8.531296960775295e-05,
      "loss": 0.7761,
      "mean_token_accuracy": 0.7693393900990486,
      "num_tokens": 152890889.0,
      "step": 4790
    },
    {
      "entropy": 0.8330507837235928,
      "epoch": 0.4408025255748961,
      "grad_norm": 0.16568523645401,
      "learning_rate": 8.530990278161131e-05,
      "loss": 0.8301,
      "mean_token_accuracy": 0.7616485096514225,
      "num_tokens": 152922855.0,
      "step": 4791
    },
    {
      "entropy": 0.7540663909167051,
      "epoch": 0.44089453194633726,
      "grad_norm": 0.16551201045513153,
      "learning_rate": 8.530683595546968e-05,
      "loss": 0.7689,
      "mean_token_accuracy": 0.7801748365163803,
      "num_tokens": 152955351.0,
      "step": 4792
    },
    {
      "entropy": 0.839316613972187,
      "epoch": 0.4409865383177785,
      "grad_norm": 0.16450689733028412,
      "learning_rate": 8.530376912932806e-05,
      "loss": 0.8429,
      "mean_token_accuracy": 0.760881420224905,
      "num_tokens": 152986761.0,
      "step": 4793
    },
    {
      "entropy": 0.7739088572561741,
      "epoch": 0.44107854468921975,
      "grad_norm": 0.1590813398361206,
      "learning_rate": 8.530070230318644e-05,
      "loss": 0.7729,
      "mean_token_accuracy": 0.7782654091715813,
      "num_tokens": 153019397.0,
      "step": 4794
    },
    {
      "entropy": 0.9111968986690044,
      "epoch": 0.44117055106066094,
      "grad_norm": 0.16821008920669556,
      "learning_rate": 8.529763547704481e-05,
      "loss": 0.906,
      "mean_token_accuracy": 0.740599162876606,
      "num_tokens": 153050580.0,
      "step": 4795
    },
    {
      "entropy": 0.6953868009150028,
      "epoch": 0.4412625574321022,
      "grad_norm": 0.139336958527565,
      "learning_rate": 8.529456865090318e-05,
      "loss": 0.6623,
      "mean_token_accuracy": 0.8026014640927315,
      "num_tokens": 153083201.0,
      "step": 4796
    },
    {
      "entropy": 0.732545044273138,
      "epoch": 0.4413545638035434,
      "grad_norm": 0.1535327285528183,
      "learning_rate": 8.529150182476155e-05,
      "loss": 0.7163,
      "mean_token_accuracy": 0.7888853251934052,
      "num_tokens": 153115387.0,
      "step": 4797
    },
    {
      "entropy": 0.7579482831060886,
      "epoch": 0.4414465701749846,
      "grad_norm": 0.15806682407855988,
      "learning_rate": 8.528843499861994e-05,
      "loss": 0.7713,
      "mean_token_accuracy": 0.7787821255624294,
      "num_tokens": 153147150.0,
      "step": 4798
    },
    {
      "entropy": 0.867162674665451,
      "epoch": 0.44153857654642586,
      "grad_norm": 0.15959234535694122,
      "learning_rate": 8.528536817247831e-05,
      "loss": 0.8489,
      "mean_token_accuracy": 0.7572667375206947,
      "num_tokens": 153179918.0,
      "step": 4799
    },
    {
      "entropy": 0.8269202318042517,
      "epoch": 0.44163058291786705,
      "grad_norm": 0.16405434906482697,
      "learning_rate": 8.528230134633668e-05,
      "loss": 0.8282,
      "mean_token_accuracy": 0.7626763582229614,
      "num_tokens": 153210163.0,
      "step": 4800
    },
    {
      "entropy": 0.9352712724357843,
      "epoch": 0.4417225892893083,
      "grad_norm": 0.1632317155599594,
      "learning_rate": 8.527923452019505e-05,
      "loss": 0.9272,
      "mean_token_accuracy": 0.7346980236470699,
      "num_tokens": 153242711.0,
      "step": 4801
    },
    {
      "entropy": 0.8103016559034586,
      "epoch": 0.4418145956607495,
      "grad_norm": 0.15515807271003723,
      "learning_rate": 8.527616769405343e-05,
      "loss": 0.7953,
      "mean_token_accuracy": 0.7709603905677795,
      "num_tokens": 153274189.0,
      "step": 4802
    },
    {
      "entropy": 0.6749774478375912,
      "epoch": 0.44190660203219073,
      "grad_norm": 0.14617526531219482,
      "learning_rate": 8.527310086791181e-05,
      "loss": 0.6461,
      "mean_token_accuracy": 0.8032250739634037,
      "num_tokens": 153306472.0,
      "step": 4803
    },
    {
      "entropy": 0.7998899780213833,
      "epoch": 0.441998608403632,
      "grad_norm": 0.1535605490207672,
      "learning_rate": 8.527003404177018e-05,
      "loss": 0.7531,
      "mean_token_accuracy": 0.7781371586024761,
      "num_tokens": 153337933.0,
      "step": 4804
    },
    {
      "entropy": 0.8150355480611324,
      "epoch": 0.44209061477507317,
      "grad_norm": 0.15442347526550293,
      "learning_rate": 8.526696721562855e-05,
      "loss": 0.7837,
      "mean_token_accuracy": 0.7746189311146736,
      "num_tokens": 153369760.0,
      "step": 4805
    },
    {
      "entropy": 0.8036406952887774,
      "epoch": 0.4421826211465144,
      "grad_norm": 0.1706598699092865,
      "learning_rate": 8.526390038948693e-05,
      "loss": 0.8086,
      "mean_token_accuracy": 0.7665969058871269,
      "num_tokens": 153400599.0,
      "step": 4806
    },
    {
      "entropy": 0.8397210985422134,
      "epoch": 0.4422746275179556,
      "grad_norm": 0.17091789841651917,
      "learning_rate": 8.52608335633453e-05,
      "loss": 0.8662,
      "mean_token_accuracy": 0.7510612271726131,
      "num_tokens": 153432179.0,
      "step": 4807
    },
    {
      "entropy": 0.7434779964387417,
      "epoch": 0.44236663388939684,
      "grad_norm": 0.15019308030605316,
      "learning_rate": 8.525776673720368e-05,
      "loss": 0.7356,
      "mean_token_accuracy": 0.7825622782111168,
      "num_tokens": 153464533.0,
      "step": 4808
    },
    {
      "entropy": 0.8030031435191631,
      "epoch": 0.4424586402608381,
      "grad_norm": 0.1622098833322525,
      "learning_rate": 8.525469991106204e-05,
      "loss": 0.8145,
      "mean_token_accuracy": 0.764342911541462,
      "num_tokens": 153496005.0,
      "step": 4809
    },
    {
      "entropy": 0.7088978849351406,
      "epoch": 0.4425506466322793,
      "grad_norm": 0.15482394397258759,
      "learning_rate": 8.525163308492043e-05,
      "loss": 0.695,
      "mean_token_accuracy": 0.790329597890377,
      "num_tokens": 153528355.0,
      "step": 4810
    },
    {
      "entropy": 0.7876562289893627,
      "epoch": 0.4426426530037205,
      "grad_norm": 0.1536676585674286,
      "learning_rate": 8.52485662587788e-05,
      "loss": 0.7745,
      "mean_token_accuracy": 0.7722665295004845,
      "num_tokens": 153560343.0,
      "step": 4811
    },
    {
      "entropy": 0.7820138167589903,
      "epoch": 0.4427346593751617,
      "grad_norm": 0.15709665417671204,
      "learning_rate": 8.524549943263716e-05,
      "loss": 0.7758,
      "mean_token_accuracy": 0.7721086628735065,
      "num_tokens": 153592308.0,
      "step": 4812
    },
    {
      "entropy": 0.7945087626576424,
      "epoch": 0.44282666574660295,
      "grad_norm": 0.15640568733215332,
      "learning_rate": 8.524243260649554e-05,
      "loss": 0.7759,
      "mean_token_accuracy": 0.7761771827936172,
      "num_tokens": 153624861.0,
      "step": 4813
    },
    {
      "entropy": 0.7891770135611296,
      "epoch": 0.4429186721180442,
      "grad_norm": 0.1646808385848999,
      "learning_rate": 8.523936578035392e-05,
      "loss": 0.789,
      "mean_token_accuracy": 0.7730864956974983,
      "num_tokens": 153656372.0,
      "step": 4814
    },
    {
      "entropy": 0.8187442570924759,
      "epoch": 0.4430106784894854,
      "grad_norm": 0.17163655161857605,
      "learning_rate": 8.523629895421229e-05,
      "loss": 0.7987,
      "mean_token_accuracy": 0.7684795819222927,
      "num_tokens": 153687410.0,
      "step": 4815
    },
    {
      "entropy": 0.8084456212818623,
      "epoch": 0.44310268486092663,
      "grad_norm": 0.16712243854999542,
      "learning_rate": 8.523323212807066e-05,
      "loss": 0.8077,
      "mean_token_accuracy": 0.7672344036400318,
      "num_tokens": 153719241.0,
      "step": 4816
    },
    {
      "entropy": 0.8988374900072813,
      "epoch": 0.4431946912323678,
      "grad_norm": 0.1655033528804779,
      "learning_rate": 8.523016530192903e-05,
      "loss": 0.8831,
      "mean_token_accuracy": 0.7461752071976662,
      "num_tokens": 153751277.0,
      "step": 4817
    },
    {
      "entropy": 0.8332402221858501,
      "epoch": 0.44328669760380907,
      "grad_norm": 0.15598562359809875,
      "learning_rate": 8.522709847578741e-05,
      "loss": 0.8204,
      "mean_token_accuracy": 0.7584976851940155,
      "num_tokens": 153783048.0,
      "step": 4818
    },
    {
      "entropy": 0.809260094538331,
      "epoch": 0.4433787039752503,
      "grad_norm": 0.15253928303718567,
      "learning_rate": 8.522403164964579e-05,
      "loss": 0.8024,
      "mean_token_accuracy": 0.7701079435646534,
      "num_tokens": 153815797.0,
      "step": 4819
    },
    {
      "entropy": 0.7805536277592182,
      "epoch": 0.4434707103466915,
      "grad_norm": 0.1857926845550537,
      "learning_rate": 8.522096482350416e-05,
      "loss": 0.7872,
      "mean_token_accuracy": 0.7727089859545231,
      "num_tokens": 153848337.0,
      "step": 4820
    },
    {
      "entropy": 0.7551811300218105,
      "epoch": 0.44356271671813274,
      "grad_norm": 0.1534833163022995,
      "learning_rate": 8.521789799736254e-05,
      "loss": 0.7281,
      "mean_token_accuracy": 0.7849056720733643,
      "num_tokens": 153880940.0,
      "step": 4821
    },
    {
      "entropy": 0.6951780840754509,
      "epoch": 0.44365472308957393,
      "grad_norm": 0.15299995243549347,
      "learning_rate": 8.521483117122091e-05,
      "loss": 0.6827,
      "mean_token_accuracy": 0.802257988601923,
      "num_tokens": 153913048.0,
      "step": 4822
    },
    {
      "entropy": 0.7750381920486689,
      "epoch": 0.4437467294610152,
      "grad_norm": 0.16016678512096405,
      "learning_rate": 8.521176434507928e-05,
      "loss": 0.7596,
      "mean_token_accuracy": 0.7760134898126125,
      "num_tokens": 153944521.0,
      "step": 4823
    },
    {
      "entropy": 0.795615928247571,
      "epoch": 0.4438387358324564,
      "grad_norm": 0.16253800690174103,
      "learning_rate": 8.520869751893766e-05,
      "loss": 0.7945,
      "mean_token_accuracy": 0.7685829550027847,
      "num_tokens": 153976858.0,
      "step": 4824
    },
    {
      "entropy": 0.8138229921460152,
      "epoch": 0.4439307422038976,
      "grad_norm": 0.15738600492477417,
      "learning_rate": 8.520563069279604e-05,
      "loss": 0.807,
      "mean_token_accuracy": 0.7632174752652645,
      "num_tokens": 154007587.0,
      "step": 4825
    },
    {
      "entropy": 0.8133902065455914,
      "epoch": 0.44402274857533885,
      "grad_norm": 0.16464722156524658,
      "learning_rate": 8.52025638666544e-05,
      "loss": 0.8393,
      "mean_token_accuracy": 0.7587261982262135,
      "num_tokens": 154039716.0,
      "step": 4826
    },
    {
      "entropy": 0.8841881025582552,
      "epoch": 0.44411475494678004,
      "grad_norm": 0.1696052849292755,
      "learning_rate": 8.519949704051277e-05,
      "loss": 0.8967,
      "mean_token_accuracy": 0.7445732578635216,
      "num_tokens": 154071477.0,
      "step": 4827
    },
    {
      "entropy": 0.8757783621549606,
      "epoch": 0.4442067613182213,
      "grad_norm": 0.1613132506608963,
      "learning_rate": 8.519643021437114e-05,
      "loss": 0.8663,
      "mean_token_accuracy": 0.7496816366910934,
      "num_tokens": 154103907.0,
      "step": 4828
    },
    {
      "entropy": 0.816790921613574,
      "epoch": 0.44429876768966253,
      "grad_norm": 0.15802700817584991,
      "learning_rate": 8.519336338822952e-05,
      "loss": 0.7768,
      "mean_token_accuracy": 0.7692062184214592,
      "num_tokens": 154134896.0,
      "step": 4829
    },
    {
      "entropy": 0.8157556150108576,
      "epoch": 0.4443907740611037,
      "grad_norm": 0.15295107662677765,
      "learning_rate": 8.51902965620879e-05,
      "loss": 0.782,
      "mean_token_accuracy": 0.7730098143219948,
      "num_tokens": 154166831.0,
      "step": 4830
    },
    {
      "entropy": 0.7901348639279604,
      "epoch": 0.44448278043254497,
      "grad_norm": 0.15862926840782166,
      "learning_rate": 8.518722973594627e-05,
      "loss": 0.7817,
      "mean_token_accuracy": 0.7726348824799061,
      "num_tokens": 154198524.0,
      "step": 4831
    },
    {
      "entropy": 0.7571733854711056,
      "epoch": 0.44457478680398616,
      "grad_norm": 0.15488962829113007,
      "learning_rate": 8.518416290980464e-05,
      "loss": 0.7375,
      "mean_token_accuracy": 0.7862406186759472,
      "num_tokens": 154230791.0,
      "step": 4832
    },
    {
      "entropy": 0.7881640270352364,
      "epoch": 0.4446667931754274,
      "grad_norm": 0.15704327821731567,
      "learning_rate": 8.518109608366302e-05,
      "loss": 0.762,
      "mean_token_accuracy": 0.7832503989338875,
      "num_tokens": 154263271.0,
      "step": 4833
    },
    {
      "entropy": 0.727181926369667,
      "epoch": 0.44475879954686864,
      "grad_norm": 0.15162864327430725,
      "learning_rate": 8.517802925752139e-05,
      "loss": 0.7117,
      "mean_token_accuracy": 0.7925432696938515,
      "num_tokens": 154295710.0,
      "step": 4834
    },
    {
      "entropy": 0.6952607836574316,
      "epoch": 0.44485080591830983,
      "grad_norm": 0.15990214049816132,
      "learning_rate": 8.517496243137977e-05,
      "loss": 0.6972,
      "mean_token_accuracy": 0.7933963239192963,
      "num_tokens": 154327424.0,
      "step": 4835
    },
    {
      "entropy": 0.7439710441976786,
      "epoch": 0.4449428122897511,
      "grad_norm": 0.15159298479557037,
      "learning_rate": 8.517189560523814e-05,
      "loss": 0.7388,
      "mean_token_accuracy": 0.7838140167295933,
      "num_tokens": 154359332.0,
      "step": 4836
    },
    {
      "entropy": 0.9150807112455368,
      "epoch": 0.44503481866119227,
      "grad_norm": 0.16906742751598358,
      "learning_rate": 8.516882877909652e-05,
      "loss": 0.9231,
      "mean_token_accuracy": 0.7363668158650398,
      "num_tokens": 154391070.0,
      "step": 4837
    },
    {
      "entropy": 0.8270695153623819,
      "epoch": 0.4451268250326335,
      "grad_norm": 0.16364797949790955,
      "learning_rate": 8.516576195295489e-05,
      "loss": 0.8269,
      "mean_token_accuracy": 0.7605484835803509,
      "num_tokens": 154422344.0,
      "step": 4838
    },
    {
      "entropy": 0.6518621183931828,
      "epoch": 0.44521883140407476,
      "grad_norm": 0.1542326807975769,
      "learning_rate": 8.516269512681327e-05,
      "loss": 0.6457,
      "mean_token_accuracy": 0.8071621507406235,
      "num_tokens": 154454316.0,
      "step": 4839
    },
    {
      "entropy": 0.8840353991836309,
      "epoch": 0.44531083777551594,
      "grad_norm": 0.16026388108730316,
      "learning_rate": 8.515962830067164e-05,
      "loss": 0.8873,
      "mean_token_accuracy": 0.7446371056139469,
      "num_tokens": 154485796.0,
      "step": 4840
    },
    {
      "entropy": 0.8468518685549498,
      "epoch": 0.4454028441469572,
      "grad_norm": 0.16614042222499847,
      "learning_rate": 8.515656147453002e-05,
      "loss": 0.8427,
      "mean_token_accuracy": 0.7535727135837078,
      "num_tokens": 154517263.0,
      "step": 4841
    },
    {
      "entropy": 0.7923102956265211,
      "epoch": 0.4454948505183984,
      "grad_norm": 0.1602209061384201,
      "learning_rate": 8.515349464838839e-05,
      "loss": 0.7865,
      "mean_token_accuracy": 0.774559061974287,
      "num_tokens": 154549641.0,
      "step": 4842
    },
    {
      "entropy": 0.8517488501966,
      "epoch": 0.4455868568898396,
      "grad_norm": 0.1642685979604721,
      "learning_rate": 8.515042782224676e-05,
      "loss": 0.8212,
      "mean_token_accuracy": 0.766651913523674,
      "num_tokens": 154581690.0,
      "step": 4843
    },
    {
      "entropy": 0.7104556653648615,
      "epoch": 0.44567886326128087,
      "grad_norm": 0.15920601785182953,
      "learning_rate": 8.514736099610514e-05,
      "loss": 0.704,
      "mean_token_accuracy": 0.7950877100229263,
      "num_tokens": 154614403.0,
      "step": 4844
    },
    {
      "entropy": 0.7918240129947662,
      "epoch": 0.44577086963272206,
      "grad_norm": 0.16036047041416168,
      "learning_rate": 8.514429416996352e-05,
      "loss": 0.7709,
      "mean_token_accuracy": 0.7785198614001274,
      "num_tokens": 154646893.0,
      "step": 4845
    },
    {
      "entropy": 0.6797931101173162,
      "epoch": 0.4458628760041633,
      "grad_norm": 0.15266665816307068,
      "learning_rate": 8.514122734382189e-05,
      "loss": 0.667,
      "mean_token_accuracy": 0.8034554272890091,
      "num_tokens": 154679352.0,
      "step": 4846
    },
    {
      "entropy": 0.8460178039968014,
      "epoch": 0.4459548823756045,
      "grad_norm": 0.16952532529830933,
      "learning_rate": 8.513816051768025e-05,
      "loss": 0.835,
      "mean_token_accuracy": 0.7573955804109573,
      "num_tokens": 154711240.0,
      "step": 4847
    },
    {
      "entropy": 0.8105768207460642,
      "epoch": 0.44604688874704573,
      "grad_norm": 0.159785196185112,
      "learning_rate": 8.513509369153862e-05,
      "loss": 0.8027,
      "mean_token_accuracy": 0.7618148438632488,
      "num_tokens": 154742580.0,
      "step": 4848
    },
    {
      "entropy": 0.7980357129126787,
      "epoch": 0.446138895118487,
      "grad_norm": 0.16381607949733734,
      "learning_rate": 8.5132026865397e-05,
      "loss": 0.784,
      "mean_token_accuracy": 0.7703238762915134,
      "num_tokens": 154774101.0,
      "step": 4849
    },
    {
      "entropy": 0.7102155387401581,
      "epoch": 0.44623090148992817,
      "grad_norm": 0.1569165587425232,
      "learning_rate": 8.512896003925538e-05,
      "loss": 0.7149,
      "mean_token_accuracy": 0.7875937260687351,
      "num_tokens": 154806228.0,
      "step": 4850
    },
    {
      "entropy": 0.8772140108048916,
      "epoch": 0.4463229078613694,
      "grad_norm": 0.16110923886299133,
      "learning_rate": 8.512589321311375e-05,
      "loss": 0.8694,
      "mean_token_accuracy": 0.7504971511662006,
      "num_tokens": 154838071.0,
      "step": 4851
    },
    {
      "entropy": 0.7623647972941399,
      "epoch": 0.4464149142328106,
      "grad_norm": 0.15623056888580322,
      "learning_rate": 8.512282638697213e-05,
      "loss": 0.7543,
      "mean_token_accuracy": 0.7804408892989159,
      "num_tokens": 154870839.0,
      "step": 4852
    },
    {
      "entropy": 0.6420317720621824,
      "epoch": 0.44650692060425184,
      "grad_norm": 0.15438906848430634,
      "learning_rate": 8.51197595608305e-05,
      "loss": 0.6482,
      "mean_token_accuracy": 0.8088855482637882,
      "num_tokens": 154903320.0,
      "step": 4853
    },
    {
      "entropy": 0.7969608418643475,
      "epoch": 0.4465989269756931,
      "grad_norm": 0.16013763844966888,
      "learning_rate": 8.511669273468887e-05,
      "loss": 0.7997,
      "mean_token_accuracy": 0.767334595322609,
      "num_tokens": 154935406.0,
      "step": 4854
    },
    {
      "entropy": 0.7915902137756348,
      "epoch": 0.4466909333471343,
      "grad_norm": 0.15210317075252533,
      "learning_rate": 8.511362590854725e-05,
      "loss": 0.7655,
      "mean_token_accuracy": 0.7778829485177994,
      "num_tokens": 154966247.0,
      "step": 4855
    },
    {
      "entropy": 0.8273196686059237,
      "epoch": 0.4467829397185755,
      "grad_norm": 0.15457554161548615,
      "learning_rate": 8.511055908240563e-05,
      "loss": 0.8238,
      "mean_token_accuracy": 0.7606394551694393,
      "num_tokens": 154998260.0,
      "step": 4856
    },
    {
      "entropy": 0.7849187050014734,
      "epoch": 0.4468749460900167,
      "grad_norm": 0.16263355314731598,
      "learning_rate": 8.5107492256264e-05,
      "loss": 0.7869,
      "mean_token_accuracy": 0.7740636132657528,
      "num_tokens": 155030406.0,
      "step": 4857
    },
    {
      "entropy": 0.8042354322969913,
      "epoch": 0.44696695246145796,
      "grad_norm": 0.157380148768425,
      "learning_rate": 8.510442543012237e-05,
      "loss": 0.7879,
      "mean_token_accuracy": 0.7694254331290722,
      "num_tokens": 155062102.0,
      "step": 4858
    },
    {
      "entropy": 0.8158824071288109,
      "epoch": 0.4470589588328992,
      "grad_norm": 0.1568228155374527,
      "learning_rate": 8.510135860398074e-05,
      "loss": 0.7933,
      "mean_token_accuracy": 0.7703485377132893,
      "num_tokens": 155093892.0,
      "step": 4859
    },
    {
      "entropy": 0.6760909035801888,
      "epoch": 0.4471509652043404,
      "grad_norm": 0.14594106376171112,
      "learning_rate": 8.509829177783912e-05,
      "loss": 0.6462,
      "mean_token_accuracy": 0.8050715439021587,
      "num_tokens": 155125710.0,
      "step": 4860
    },
    {
      "entropy": 0.734922943636775,
      "epoch": 0.44724297157578163,
      "grad_norm": 0.15435729920864105,
      "learning_rate": 8.50952249516975e-05,
      "loss": 0.7241,
      "mean_token_accuracy": 0.789524506777525,
      "num_tokens": 155158402.0,
      "step": 4861
    },
    {
      "entropy": 0.7771373670548201,
      "epoch": 0.4473349779472228,
      "grad_norm": 0.14880378544330597,
      "learning_rate": 8.509215812555587e-05,
      "loss": 0.7575,
      "mean_token_accuracy": 0.778119407594204,
      "num_tokens": 155190768.0,
      "step": 4862
    },
    {
      "entropy": 0.85196996293962,
      "epoch": 0.44742698431866407,
      "grad_norm": 0.16494189202785492,
      "learning_rate": 8.508909129941423e-05,
      "loss": 0.8478,
      "mean_token_accuracy": 0.7573219873011112,
      "num_tokens": 155222655.0,
      "step": 4863
    },
    {
      "entropy": 0.7911323420703411,
      "epoch": 0.4475189906901053,
      "grad_norm": 0.15604722499847412,
      "learning_rate": 8.508602447327262e-05,
      "loss": 0.7772,
      "mean_token_accuracy": 0.7652524672448635,
      "num_tokens": 155254111.0,
      "step": 4864
    },
    {
      "entropy": 0.8664130792021751,
      "epoch": 0.4476109970615465,
      "grad_norm": 0.1673228144645691,
      "learning_rate": 8.508295764713098e-05,
      "loss": 0.8747,
      "mean_token_accuracy": 0.7477344051003456,
      "num_tokens": 155285715.0,
      "step": 4865
    },
    {
      "entropy": 0.8173853773623705,
      "epoch": 0.44770300343298775,
      "grad_norm": 0.15645544230937958,
      "learning_rate": 8.507989082098937e-05,
      "loss": 0.8097,
      "mean_token_accuracy": 0.7669736780226231,
      "num_tokens": 155318381.0,
      "step": 4866
    },
    {
      "entropy": 0.6629424151033163,
      "epoch": 0.44779500980442893,
      "grad_norm": 0.1579284518957138,
      "learning_rate": 8.507682399484773e-05,
      "loss": 0.6746,
      "mean_token_accuracy": 0.8061664327979088,
      "num_tokens": 155350849.0,
      "step": 4867
    },
    {
      "entropy": 0.7657013032585382,
      "epoch": 0.4478870161758702,
      "grad_norm": 0.15650005638599396,
      "learning_rate": 8.507375716870611e-05,
      "loss": 0.7526,
      "mean_token_accuracy": 0.7816522754728794,
      "num_tokens": 155383520.0,
      "step": 4868
    },
    {
      "entropy": 0.7669861931353807,
      "epoch": 0.4479790225473114,
      "grad_norm": 0.1619252860546112,
      "learning_rate": 8.507069034256448e-05,
      "loss": 0.7777,
      "mean_token_accuracy": 0.7737644501030445,
      "num_tokens": 155415789.0,
      "step": 4869
    },
    {
      "entropy": 0.7797046285122633,
      "epoch": 0.4480710289187526,
      "grad_norm": 0.16412998735904694,
      "learning_rate": 8.506762351642285e-05,
      "loss": 0.8036,
      "mean_token_accuracy": 0.7646700106561184,
      "num_tokens": 155447362.0,
      "step": 4870
    },
    {
      "entropy": 0.7928271759301424,
      "epoch": 0.44816303529019386,
      "grad_norm": 0.15356405079364777,
      "learning_rate": 8.506455669028123e-05,
      "loss": 0.7895,
      "mean_token_accuracy": 0.7720334939658642,
      "num_tokens": 155479318.0,
      "step": 4871
    },
    {
      "entropy": 0.7455516420304775,
      "epoch": 0.44825504166163505,
      "grad_norm": 0.15991194546222687,
      "learning_rate": 8.506148986413961e-05,
      "loss": 0.7499,
      "mean_token_accuracy": 0.7798930108547211,
      "num_tokens": 155510991.0,
      "step": 4872
    },
    {
      "entropy": 0.768953999504447,
      "epoch": 0.4483470480330763,
      "grad_norm": 0.1549295037984848,
      "learning_rate": 8.505842303799798e-05,
      "loss": 0.7289,
      "mean_token_accuracy": 0.7821979783475399,
      "num_tokens": 155543064.0,
      "step": 4873
    },
    {
      "entropy": 0.8156191986054182,
      "epoch": 0.44843905440451753,
      "grad_norm": 0.15766188502311707,
      "learning_rate": 8.505535621185635e-05,
      "loss": 0.7971,
      "mean_token_accuracy": 0.7667980864644051,
      "num_tokens": 155575001.0,
      "step": 4874
    },
    {
      "entropy": 0.826321717351675,
      "epoch": 0.4485310607759587,
      "grad_norm": 0.15598444640636444,
      "learning_rate": 8.505228938571472e-05,
      "loss": 0.8044,
      "mean_token_accuracy": 0.7614794299006462,
      "num_tokens": 155606651.0,
      "step": 4875
    },
    {
      "entropy": 0.7541118608787656,
      "epoch": 0.44862306714739997,
      "grad_norm": 0.15436437726020813,
      "learning_rate": 8.504922255957311e-05,
      "loss": 0.7309,
      "mean_token_accuracy": 0.7829996272921562,
      "num_tokens": 155638130.0,
      "step": 4876
    },
    {
      "entropy": 0.7098454423248768,
      "epoch": 0.44871507351884116,
      "grad_norm": 0.14475159347057343,
      "learning_rate": 8.504615573343148e-05,
      "loss": 0.6946,
      "mean_token_accuracy": 0.7966405861079693,
      "num_tokens": 155670699.0,
      "step": 4877
    },
    {
      "entropy": 0.7488485872745514,
      "epoch": 0.4488070798902824,
      "grad_norm": 0.1481117159128189,
      "learning_rate": 8.504308890728985e-05,
      "loss": 0.73,
      "mean_token_accuracy": 0.7859379202127457,
      "num_tokens": 155702447.0,
      "step": 4878
    },
    {
      "entropy": 0.8189300186932087,
      "epoch": 0.44889908626172365,
      "grad_norm": 0.17654047906398773,
      "learning_rate": 8.504002208114822e-05,
      "loss": 0.8268,
      "mean_token_accuracy": 0.7585215643048286,
      "num_tokens": 155734211.0,
      "step": 4879
    },
    {
      "entropy": 0.7314107194542885,
      "epoch": 0.44899109263316483,
      "grad_norm": 0.16159935295581818,
      "learning_rate": 8.50369552550066e-05,
      "loss": 0.744,
      "mean_token_accuracy": 0.7830612286925316,
      "num_tokens": 155765744.0,
      "step": 4880
    },
    {
      "entropy": 0.7607586421072483,
      "epoch": 0.4490830990046061,
      "grad_norm": 0.15782220661640167,
      "learning_rate": 8.503388842886498e-05,
      "loss": 0.7648,
      "mean_token_accuracy": 0.7827454470098019,
      "num_tokens": 155797945.0,
      "step": 4881
    },
    {
      "entropy": 0.7367046196013689,
      "epoch": 0.44917510537604727,
      "grad_norm": 0.17219842970371246,
      "learning_rate": 8.503082160272335e-05,
      "loss": 0.7488,
      "mean_token_accuracy": 0.7837351784110069,
      "num_tokens": 155830202.0,
      "step": 4882
    },
    {
      "entropy": 0.7623578514903784,
      "epoch": 0.4492671117474885,
      "grad_norm": 0.15690980851650238,
      "learning_rate": 8.502775477658173e-05,
      "loss": 0.7591,
      "mean_token_accuracy": 0.7771518342196941,
      "num_tokens": 155862472.0,
      "step": 4883
    },
    {
      "entropy": 0.92664560303092,
      "epoch": 0.44935911811892976,
      "grad_norm": 0.16047240793704987,
      "learning_rate": 8.50246879504401e-05,
      "loss": 0.8846,
      "mean_token_accuracy": 0.7454448901116848,
      "num_tokens": 155893924.0,
      "step": 4884
    },
    {
      "entropy": 0.8647530004382133,
      "epoch": 0.44945112449037095,
      "grad_norm": 0.16004012525081635,
      "learning_rate": 8.502162112429846e-05,
      "loss": 0.8452,
      "mean_token_accuracy": 0.7586942873895168,
      "num_tokens": 155926008.0,
      "step": 4885
    },
    {
      "entropy": 0.752610856667161,
      "epoch": 0.4495431308618122,
      "grad_norm": 0.16145627200603485,
      "learning_rate": 8.501855429815684e-05,
      "loss": 0.7319,
      "mean_token_accuracy": 0.7872746884822845,
      "num_tokens": 155957831.0,
      "step": 4886
    },
    {
      "entropy": 0.6804722361266613,
      "epoch": 0.4496351372332534,
      "grad_norm": 0.14870552718639374,
      "learning_rate": 8.501548747201523e-05,
      "loss": 0.6637,
      "mean_token_accuracy": 0.8022947162389755,
      "num_tokens": 155990032.0,
      "step": 4887
    },
    {
      "entropy": 0.8845971245318651,
      "epoch": 0.4497271436046946,
      "grad_norm": 0.15914149582386017,
      "learning_rate": 8.50124206458736e-05,
      "loss": 0.8804,
      "mean_token_accuracy": 0.7457185573875904,
      "num_tokens": 156021362.0,
      "step": 4888
    },
    {
      "entropy": 0.7859588526189327,
      "epoch": 0.44981914997613587,
      "grad_norm": 0.160914346575737,
      "learning_rate": 8.500935381973196e-05,
      "loss": 0.7779,
      "mean_token_accuracy": 0.7754861675202847,
      "num_tokens": 156052013.0,
      "step": 4889
    },
    {
      "entropy": 0.7296443898230791,
      "epoch": 0.44991115634757706,
      "grad_norm": 0.15243567526340485,
      "learning_rate": 8.500628699359033e-05,
      "loss": 0.7257,
      "mean_token_accuracy": 0.7858364321291447,
      "num_tokens": 156084333.0,
      "step": 4890
    },
    {
      "entropy": 0.7555712107568979,
      "epoch": 0.4500031627190183,
      "grad_norm": 0.15457119047641754,
      "learning_rate": 8.500322016744871e-05,
      "loss": 0.7334,
      "mean_token_accuracy": 0.7797279097139835,
      "num_tokens": 156116145.0,
      "step": 4891
    },
    {
      "entropy": 0.7276248000562191,
      "epoch": 0.4500951690904595,
      "grad_norm": 0.16081400215625763,
      "learning_rate": 8.500015334130709e-05,
      "loss": 0.7229,
      "mean_token_accuracy": 0.7863832525908947,
      "num_tokens": 156147712.0,
      "step": 4892
    },
    {
      "entropy": 0.7676946260035038,
      "epoch": 0.45018717546190073,
      "grad_norm": 0.16093038022518158,
      "learning_rate": 8.499708651516546e-05,
      "loss": 0.7731,
      "mean_token_accuracy": 0.7737197428941727,
      "num_tokens": 156180157.0,
      "step": 4893
    },
    {
      "entropy": 0.789346257224679,
      "epoch": 0.450279181833342,
      "grad_norm": 0.15983736515045166,
      "learning_rate": 8.499401968902383e-05,
      "loss": 0.8166,
      "mean_token_accuracy": 0.7641306221485138,
      "num_tokens": 156212634.0,
      "step": 4894
    },
    {
      "entropy": 0.8895988930016756,
      "epoch": 0.45037118820478317,
      "grad_norm": 0.16520899534225464,
      "learning_rate": 8.499095286288221e-05,
      "loss": 0.8902,
      "mean_token_accuracy": 0.746096383780241,
      "num_tokens": 156244567.0,
      "step": 4895
    },
    {
      "entropy": 0.8156241234391928,
      "epoch": 0.4504631945762244,
      "grad_norm": 0.16546542942523956,
      "learning_rate": 8.498788603674058e-05,
      "loss": 0.8327,
      "mean_token_accuracy": 0.7573663406074047,
      "num_tokens": 156276945.0,
      "step": 4896
    },
    {
      "entropy": 0.8687437474727631,
      "epoch": 0.4505552009476656,
      "grad_norm": 0.17036612331867218,
      "learning_rate": 8.498481921059896e-05,
      "loss": 0.873,
      "mean_token_accuracy": 0.7489736638963223,
      "num_tokens": 156307206.0,
      "step": 4897
    },
    {
      "entropy": 0.7693162597715855,
      "epoch": 0.45064720731910685,
      "grad_norm": 0.16129259765148163,
      "learning_rate": 8.498175238445733e-05,
      "loss": 0.7652,
      "mean_token_accuracy": 0.7774639651179314,
      "num_tokens": 156338444.0,
      "step": 4898
    },
    {
      "entropy": 0.8498350884765387,
      "epoch": 0.4507392136905481,
      "grad_norm": 0.1626051813364029,
      "learning_rate": 8.497868555831571e-05,
      "loss": 0.8365,
      "mean_token_accuracy": 0.7545686773955822,
      "num_tokens": 156370474.0,
      "step": 4899
    },
    {
      "entropy": 0.8449336923658848,
      "epoch": 0.4508312200619893,
      "grad_norm": 0.1699608862400055,
      "learning_rate": 8.497561873217408e-05,
      "loss": 0.842,
      "mean_token_accuracy": 0.7533034607768059,
      "num_tokens": 156402663.0,
      "step": 4900
    },
    {
      "entropy": 0.8696624413132668,
      "epoch": 0.4509232264334305,
      "grad_norm": 0.1711505949497223,
      "learning_rate": 8.497255190603244e-05,
      "loss": 0.8769,
      "mean_token_accuracy": 0.7507443390786648,
      "num_tokens": 156434304.0,
      "step": 4901
    },
    {
      "entropy": 0.9508023001253605,
      "epoch": 0.4510152328048717,
      "grad_norm": 0.16503740847110748,
      "learning_rate": 8.496948507989083e-05,
      "loss": 0.9523,
      "mean_token_accuracy": 0.7314354367554188,
      "num_tokens": 156466725.0,
      "step": 4902
    },
    {
      "entropy": 0.7872964404523373,
      "epoch": 0.45110723917631296,
      "grad_norm": 0.15788625180721283,
      "learning_rate": 8.496641825374921e-05,
      "loss": 0.7659,
      "mean_token_accuracy": 0.7734062895178795,
      "num_tokens": 156498522.0,
      "step": 4903
    },
    {
      "entropy": 0.7724616602063179,
      "epoch": 0.4511992455477542,
      "grad_norm": 0.14386731386184692,
      "learning_rate": 8.496335142760757e-05,
      "loss": 0.7383,
      "mean_token_accuracy": 0.7847315073013306,
      "num_tokens": 156531040.0,
      "step": 4904
    },
    {
      "entropy": 0.891039103269577,
      "epoch": 0.4512912519191954,
      "grad_norm": 0.2051355242729187,
      "learning_rate": 8.496028460146594e-05,
      "loss": 0.8836,
      "mean_token_accuracy": 0.7485418803989887,
      "num_tokens": 156563030.0,
      "step": 4905
    },
    {
      "entropy": 0.7822888679802418,
      "epoch": 0.45138325829063664,
      "grad_norm": 0.1542302668094635,
      "learning_rate": 8.495721777532431e-05,
      "loss": 0.7507,
      "mean_token_accuracy": 0.7799147479236126,
      "num_tokens": 156594711.0,
      "step": 4906
    },
    {
      "entropy": 0.7839927226305008,
      "epoch": 0.4514752646620778,
      "grad_norm": 0.15807369351387024,
      "learning_rate": 8.495415094918269e-05,
      "loss": 0.778,
      "mean_token_accuracy": 0.7767650298774242,
      "num_tokens": 156626654.0,
      "step": 4907
    },
    {
      "entropy": 0.6946792993694544,
      "epoch": 0.45156727103351907,
      "grad_norm": 0.1486385017633438,
      "learning_rate": 8.495108412304107e-05,
      "loss": 0.6883,
      "mean_token_accuracy": 0.794497549533844,
      "num_tokens": 156658899.0,
      "step": 4908
    },
    {
      "entropy": 0.8562528528273106,
      "epoch": 0.4516592774049603,
      "grad_norm": 0.15857680141925812,
      "learning_rate": 8.494801729689944e-05,
      "loss": 0.8571,
      "mean_token_accuracy": 0.7496035546064377,
      "num_tokens": 156691397.0,
      "step": 4909
    },
    {
      "entropy": 0.8166457675397396,
      "epoch": 0.4517512837764015,
      "grad_norm": 0.15739218890666962,
      "learning_rate": 8.494495047075781e-05,
      "loss": 0.8029,
      "mean_token_accuracy": 0.765981949865818,
      "num_tokens": 156723559.0,
      "step": 4910
    },
    {
      "entropy": 0.8153786193579435,
      "epoch": 0.45184329014784275,
      "grad_norm": 0.15839733183383942,
      "learning_rate": 8.494188364461619e-05,
      "loss": 0.811,
      "mean_token_accuracy": 0.7640904486179352,
      "num_tokens": 156755605.0,
      "step": 4911
    },
    {
      "entropy": 0.7247861959040165,
      "epoch": 0.45193529651928394,
      "grad_norm": 0.15151222050189972,
      "learning_rate": 8.493881681847456e-05,
      "loss": 0.7207,
      "mean_token_accuracy": 0.7882841192185879,
      "num_tokens": 156788330.0,
      "step": 4912
    },
    {
      "entropy": 0.7544763088226318,
      "epoch": 0.4520273028907252,
      "grad_norm": 0.15322192013263702,
      "learning_rate": 8.493574999233294e-05,
      "loss": 0.7357,
      "mean_token_accuracy": 0.783578522503376,
      "num_tokens": 156820840.0,
      "step": 4913
    },
    {
      "entropy": 0.7745220810174942,
      "epoch": 0.4521193092621664,
      "grad_norm": 0.15180513262748718,
      "learning_rate": 8.493268316619132e-05,
      "loss": 0.767,
      "mean_token_accuracy": 0.7723800577223301,
      "num_tokens": 156852200.0,
      "step": 4914
    },
    {
      "entropy": 0.765038637444377,
      "epoch": 0.4522113156336076,
      "grad_norm": 0.15860500931739807,
      "learning_rate": 8.492961634004969e-05,
      "loss": 0.7386,
      "mean_token_accuracy": 0.7801927849650383,
      "num_tokens": 156883512.0,
      "step": 4915
    },
    {
      "entropy": 0.7761003319174051,
      "epoch": 0.45230332200504886,
      "grad_norm": 0.14756500720977783,
      "learning_rate": 8.492654951390806e-05,
      "loss": 0.7587,
      "mean_token_accuracy": 0.7772273868322372,
      "num_tokens": 156915263.0,
      "step": 4916
    },
    {
      "entropy": 0.8133154064416885,
      "epoch": 0.45239532837649005,
      "grad_norm": 0.15990076959133148,
      "learning_rate": 8.492348268776642e-05,
      "loss": 0.7945,
      "mean_token_accuracy": 0.7673600055277348,
      "num_tokens": 156947602.0,
      "step": 4917
    },
    {
      "entropy": 0.8985932637006044,
      "epoch": 0.4524873347479313,
      "grad_norm": 0.16162239015102386,
      "learning_rate": 8.492041586162482e-05,
      "loss": 0.9105,
      "mean_token_accuracy": 0.7414383105933666,
      "num_tokens": 156979652.0,
      "step": 4918
    },
    {
      "entropy": 0.7049762401729822,
      "epoch": 0.45257934111937254,
      "grad_norm": 0.16227681934833527,
      "learning_rate": 8.491734903548319e-05,
      "loss": 0.6909,
      "mean_token_accuracy": 0.7934020459651947,
      "num_tokens": 157011206.0,
      "step": 4919
    },
    {
      "entropy": 0.8394502084702253,
      "epoch": 0.4526713474908137,
      "grad_norm": 0.16198977828025818,
      "learning_rate": 8.491428220934156e-05,
      "loss": 0.8466,
      "mean_token_accuracy": 0.7575390934944153,
      "num_tokens": 157042991.0,
      "step": 4920
    },
    {
      "entropy": 0.9357144460082054,
      "epoch": 0.45276335386225497,
      "grad_norm": 0.16070564091205597,
      "learning_rate": 8.491121538319992e-05,
      "loss": 0.9213,
      "mean_token_accuracy": 0.7340216040611267,
      "num_tokens": 157075376.0,
      "step": 4921
    },
    {
      "entropy": 0.7556687481701374,
      "epoch": 0.45285536023369616,
      "grad_norm": 0.15993590652942657,
      "learning_rate": 8.49081485570583e-05,
      "loss": 0.7425,
      "mean_token_accuracy": 0.7772657461464405,
      "num_tokens": 157107375.0,
      "step": 4922
    },
    {
      "entropy": 0.857011865824461,
      "epoch": 0.4529473666051374,
      "grad_norm": 0.15581227838993073,
      "learning_rate": 8.490508173091669e-05,
      "loss": 0.8346,
      "mean_token_accuracy": 0.7667944356799126,
      "num_tokens": 157140044.0,
      "step": 4923
    },
    {
      "entropy": 0.81588539108634,
      "epoch": 0.45303937297657865,
      "grad_norm": 0.1648908108472824,
      "learning_rate": 8.490201490477505e-05,
      "loss": 0.7966,
      "mean_token_accuracy": 0.7708098143339157,
      "num_tokens": 157172099.0,
      "step": 4924
    },
    {
      "entropy": 0.8084226064383984,
      "epoch": 0.45313137934801984,
      "grad_norm": 0.1569342166185379,
      "learning_rate": 8.489894807863342e-05,
      "loss": 0.7977,
      "mean_token_accuracy": 0.7695537395775318,
      "num_tokens": 157204612.0,
      "step": 4925
    },
    {
      "entropy": 0.6781633701175451,
      "epoch": 0.4532233857194611,
      "grad_norm": 0.1585545688867569,
      "learning_rate": 8.48958812524918e-05,
      "loss": 0.6561,
      "mean_token_accuracy": 0.8033174946904182,
      "num_tokens": 157236902.0,
      "step": 4926
    },
    {
      "entropy": 0.8724102936685085,
      "epoch": 0.45331539209090227,
      "grad_norm": 0.17343389987945557,
      "learning_rate": 8.489281442635017e-05,
      "loss": 0.8982,
      "mean_token_accuracy": 0.7457942105829716,
      "num_tokens": 157269432.0,
      "step": 4927
    },
    {
      "entropy": 0.7988063152879477,
      "epoch": 0.4534073984623435,
      "grad_norm": 0.16237451136112213,
      "learning_rate": 8.488974760020855e-05,
      "loss": 0.8004,
      "mean_token_accuracy": 0.7651961520314217,
      "num_tokens": 157300900.0,
      "step": 4928
    },
    {
      "entropy": 0.7917265444993973,
      "epoch": 0.45349940483378476,
      "grad_norm": 0.16913743317127228,
      "learning_rate": 8.488668077406692e-05,
      "loss": 0.7945,
      "mean_token_accuracy": 0.76840515807271,
      "num_tokens": 157331596.0,
      "step": 4929
    },
    {
      "entropy": 0.7186240274459124,
      "epoch": 0.45359141120522595,
      "grad_norm": 0.1496771275997162,
      "learning_rate": 8.48836139479253e-05,
      "loss": 0.7044,
      "mean_token_accuracy": 0.7885562777519226,
      "num_tokens": 157363218.0,
      "step": 4930
    },
    {
      "entropy": 0.8401139415800571,
      "epoch": 0.4536834175766672,
      "grad_norm": 0.1526770293712616,
      "learning_rate": 8.488054712178367e-05,
      "loss": 0.8289,
      "mean_token_accuracy": 0.7558215670287609,
      "num_tokens": 157395303.0,
      "step": 4931
    },
    {
      "entropy": 0.919913537800312,
      "epoch": 0.4537754239481084,
      "grad_norm": 0.16504880785942078,
      "learning_rate": 8.487748029564204e-05,
      "loss": 0.8916,
      "mean_token_accuracy": 0.7452968135476112,
      "num_tokens": 157427168.0,
      "step": 4932
    },
    {
      "entropy": 0.8632475808262825,
      "epoch": 0.4538674303195496,
      "grad_norm": 0.16186662018299103,
      "learning_rate": 8.487441346950042e-05,
      "loss": 0.8633,
      "mean_token_accuracy": 0.7493872158229351,
      "num_tokens": 157458011.0,
      "step": 4933
    },
    {
      "entropy": 0.8022543750703335,
      "epoch": 0.45395943669099087,
      "grad_norm": 0.15886840224266052,
      "learning_rate": 8.48713466433588e-05,
      "loss": 0.7832,
      "mean_token_accuracy": 0.773474209010601,
      "num_tokens": 157490270.0,
      "step": 4934
    },
    {
      "entropy": 0.8227585628628731,
      "epoch": 0.45405144306243206,
      "grad_norm": 0.16079047322273254,
      "learning_rate": 8.486827981721717e-05,
      "loss": 0.8323,
      "mean_token_accuracy": 0.7558220289647579,
      "num_tokens": 157522157.0,
      "step": 4935
    },
    {
      "entropy": 0.7932940684258938,
      "epoch": 0.4541434494338733,
      "grad_norm": 0.15648671984672546,
      "learning_rate": 8.486521299107554e-05,
      "loss": 0.7824,
      "mean_token_accuracy": 0.7677512466907501,
      "num_tokens": 157553709.0,
      "step": 4936
    },
    {
      "entropy": 0.7609754707664251,
      "epoch": 0.4542354558053145,
      "grad_norm": 0.1577099710702896,
      "learning_rate": 8.48621461649339e-05,
      "loss": 0.7539,
      "mean_token_accuracy": 0.7739433757960796,
      "num_tokens": 157585326.0,
      "step": 4937
    },
    {
      "entropy": 0.8143345508724451,
      "epoch": 0.45432746217675574,
      "grad_norm": 0.15805917978286743,
      "learning_rate": 8.485907933879229e-05,
      "loss": 0.7931,
      "mean_token_accuracy": 0.7668934762477875,
      "num_tokens": 157617489.0,
      "step": 4938
    },
    {
      "entropy": 0.828116474673152,
      "epoch": 0.454419468548197,
      "grad_norm": 0.1620628386735916,
      "learning_rate": 8.485601251265067e-05,
      "loss": 0.8083,
      "mean_token_accuracy": 0.7650489546358585,
      "num_tokens": 157649673.0,
      "step": 4939
    },
    {
      "entropy": 0.7476275488734245,
      "epoch": 0.45451147491963817,
      "grad_norm": 0.15749111771583557,
      "learning_rate": 8.485294568650903e-05,
      "loss": 0.7378,
      "mean_token_accuracy": 0.7832685820758343,
      "num_tokens": 157681556.0,
      "step": 4940
    },
    {
      "entropy": 0.7288476768881083,
      "epoch": 0.4546034812910794,
      "grad_norm": 0.1589738130569458,
      "learning_rate": 8.48498788603674e-05,
      "loss": 0.7284,
      "mean_token_accuracy": 0.7848289757966995,
      "num_tokens": 157712533.0,
      "step": 4941
    },
    {
      "entropy": 0.7447115816175938,
      "epoch": 0.4546954876625206,
      "grad_norm": 0.15665730834007263,
      "learning_rate": 8.484681203422578e-05,
      "loss": 0.7196,
      "mean_token_accuracy": 0.785344984382391,
      "num_tokens": 157744275.0,
      "step": 4942
    },
    {
      "entropy": 0.7620610725134611,
      "epoch": 0.45478749403396185,
      "grad_norm": 0.1551925092935562,
      "learning_rate": 8.484374520808415e-05,
      "loss": 0.7434,
      "mean_token_accuracy": 0.7841373756527901,
      "num_tokens": 157776561.0,
      "step": 4943
    },
    {
      "entropy": 0.7629343550652266,
      "epoch": 0.4548795004054031,
      "grad_norm": 0.14882449805736542,
      "learning_rate": 8.484067838194253e-05,
      "loss": 0.7461,
      "mean_token_accuracy": 0.7836392857134342,
      "num_tokens": 157808841.0,
      "step": 4944
    },
    {
      "entropy": 0.9237363375723362,
      "epoch": 0.4549715067768443,
      "grad_norm": 0.1665066033601761,
      "learning_rate": 8.483761155580091e-05,
      "loss": 0.9194,
      "mean_token_accuracy": 0.736446987837553,
      "num_tokens": 157840592.0,
      "step": 4945
    },
    {
      "entropy": 0.7243937905877829,
      "epoch": 0.4550635131482855,
      "grad_norm": 0.15970967710018158,
      "learning_rate": 8.483454472965928e-05,
      "loss": 0.732,
      "mean_token_accuracy": 0.7815693691372871,
      "num_tokens": 157872498.0,
      "step": 4946
    },
    {
      "entropy": 0.8355274926871061,
      "epoch": 0.4551555195197267,
      "grad_norm": 0.1705227494239807,
      "learning_rate": 8.483147790351765e-05,
      "loss": 0.8671,
      "mean_token_accuracy": 0.7472508028149605,
      "num_tokens": 157904204.0,
      "step": 4947
    },
    {
      "entropy": 0.8463669633492827,
      "epoch": 0.45524752589116796,
      "grad_norm": 0.1645912528038025,
      "learning_rate": 8.482841107737602e-05,
      "loss": 0.8615,
      "mean_token_accuracy": 0.7527628317475319,
      "num_tokens": 157935625.0,
      "step": 4948
    },
    {
      "entropy": 0.8433198444545269,
      "epoch": 0.4553395322626092,
      "grad_norm": 0.1755211055278778,
      "learning_rate": 8.48253442512344e-05,
      "loss": 0.8659,
      "mean_token_accuracy": 0.7511949837207794,
      "num_tokens": 157966933.0,
      "step": 4949
    },
    {
      "entropy": 0.7826541885733604,
      "epoch": 0.4554315386340504,
      "grad_norm": 0.15580059587955475,
      "learning_rate": 8.482227742509278e-05,
      "loss": 0.7854,
      "mean_token_accuracy": 0.7694987431168556,
      "num_tokens": 157999464.0,
      "step": 4950
    },
    {
      "entropy": 0.7577575203031301,
      "epoch": 0.45552354500549164,
      "grad_norm": 0.15584686398506165,
      "learning_rate": 8.481921059895115e-05,
      "loss": 0.7341,
      "mean_token_accuracy": 0.7861675657331944,
      "num_tokens": 158030858.0,
      "step": 4951
    },
    {
      "entropy": 0.7469234243035316,
      "epoch": 0.4556155513769328,
      "grad_norm": 0.15596377849578857,
      "learning_rate": 8.481614377280952e-05,
      "loss": 0.7185,
      "mean_token_accuracy": 0.7863660864531994,
      "num_tokens": 158063315.0,
      "step": 4952
    },
    {
      "entropy": 0.7341910935938358,
      "epoch": 0.45570755774837407,
      "grad_norm": 0.15007731318473816,
      "learning_rate": 8.48130769466679e-05,
      "loss": 0.7129,
      "mean_token_accuracy": 0.7906660251319408,
      "num_tokens": 158095592.0,
      "step": 4953
    },
    {
      "entropy": 0.8013153411448002,
      "epoch": 0.4557995641198153,
      "grad_norm": 0.1604943573474884,
      "learning_rate": 8.481001012052627e-05,
      "loss": 0.7918,
      "mean_token_accuracy": 0.77103116735816,
      "num_tokens": 158127618.0,
      "step": 4954
    },
    {
      "entropy": 0.8691930156201124,
      "epoch": 0.4558915704912565,
      "grad_norm": 0.17321860790252686,
      "learning_rate": 8.480694329438465e-05,
      "loss": 0.8493,
      "mean_token_accuracy": 0.7542638294398785,
      "num_tokens": 158158308.0,
      "step": 4955
    },
    {
      "entropy": 0.7408685758709908,
      "epoch": 0.45598357686269775,
      "grad_norm": 0.15224312245845795,
      "learning_rate": 8.480387646824302e-05,
      "loss": 0.7296,
      "mean_token_accuracy": 0.7859476432204247,
      "num_tokens": 158190556.0,
      "step": 4956
    },
    {
      "entropy": 0.8845944777131081,
      "epoch": 0.45607558323413894,
      "grad_norm": 0.16155152022838593,
      "learning_rate": 8.48008096421014e-05,
      "loss": 0.8776,
      "mean_token_accuracy": 0.7456442676484585,
      "num_tokens": 158222838.0,
      "step": 4957
    },
    {
      "entropy": 0.9239476937800646,
      "epoch": 0.4561675896055802,
      "grad_norm": 0.17286419868469238,
      "learning_rate": 8.479774281595976e-05,
      "loss": 0.9262,
      "mean_token_accuracy": 0.7390604875981808,
      "num_tokens": 158254179.0,
      "step": 4958
    },
    {
      "entropy": 0.7662383355200291,
      "epoch": 0.4562595959770214,
      "grad_norm": 0.15579643845558167,
      "learning_rate": 8.479467598981815e-05,
      "loss": 0.7465,
      "mean_token_accuracy": 0.7801336161792278,
      "num_tokens": 158286237.0,
      "step": 4959
    },
    {
      "entropy": 0.7533433344215155,
      "epoch": 0.4563516023484626,
      "grad_norm": 0.1531396359205246,
      "learning_rate": 8.479160916367651e-05,
      "loss": 0.744,
      "mean_token_accuracy": 0.7793834656476974,
      "num_tokens": 158318265.0,
      "step": 4960
    },
    {
      "entropy": 0.8610907755792141,
      "epoch": 0.45644360871990386,
      "grad_norm": 0.1623634546995163,
      "learning_rate": 8.47885423375349e-05,
      "loss": 0.8477,
      "mean_token_accuracy": 0.7592988982796669,
      "num_tokens": 158350070.0,
      "step": 4961
    },
    {
      "entropy": 0.7440295740962029,
      "epoch": 0.45653561509134505,
      "grad_norm": 0.16241194307804108,
      "learning_rate": 8.478547551139326e-05,
      "loss": 0.7538,
      "mean_token_accuracy": 0.7789885513484478,
      "num_tokens": 158382658.0,
      "step": 4962
    },
    {
      "entropy": 0.7356470432132483,
      "epoch": 0.4566276214627863,
      "grad_norm": 0.14906199276447296,
      "learning_rate": 8.478240868525163e-05,
      "loss": 0.7223,
      "mean_token_accuracy": 0.7856465540826321,
      "num_tokens": 158414941.0,
      "step": 4963
    },
    {
      "entropy": 0.8217952605336905,
      "epoch": 0.45671962783422754,
      "grad_norm": 0.16376474499702454,
      "learning_rate": 8.477934185911001e-05,
      "loss": 0.8176,
      "mean_token_accuracy": 0.7633491903543472,
      "num_tokens": 158447069.0,
      "step": 4964
    },
    {
      "entropy": 0.6384461484849453,
      "epoch": 0.4568116342056687,
      "grad_norm": 0.170317605137825,
      "learning_rate": 8.47762750329684e-05,
      "loss": 0.6509,
      "mean_token_accuracy": 0.8092977963387966,
      "num_tokens": 158478818.0,
      "step": 4965
    },
    {
      "entropy": 0.7379900142550468,
      "epoch": 0.45690364057710997,
      "grad_norm": 0.16735272109508514,
      "learning_rate": 8.477320820682676e-05,
      "loss": 0.7609,
      "mean_token_accuracy": 0.7757553718984127,
      "num_tokens": 158510369.0,
      "step": 4966
    },
    {
      "entropy": 0.8146024905145168,
      "epoch": 0.45699564694855116,
      "grad_norm": 0.16836252808570862,
      "learning_rate": 8.477014138068513e-05,
      "loss": 0.843,
      "mean_token_accuracy": 0.7607084661722183,
      "num_tokens": 158542553.0,
      "step": 4967
    },
    {
      "entropy": 0.751719418913126,
      "epoch": 0.4570876533199924,
      "grad_norm": 0.15320123732089996,
      "learning_rate": 8.47670745545435e-05,
      "loss": 0.7313,
      "mean_token_accuracy": 0.7825224474072456,
      "num_tokens": 158574316.0,
      "step": 4968
    },
    {
      "entropy": 0.7685645092278719,
      "epoch": 0.45717965969143365,
      "grad_norm": 0.1591079831123352,
      "learning_rate": 8.476400772840188e-05,
      "loss": 0.7466,
      "mean_token_accuracy": 0.7814823389053345,
      "num_tokens": 158606691.0,
      "step": 4969
    },
    {
      "entropy": 0.8197613880038261,
      "epoch": 0.45727166606287484,
      "grad_norm": 0.1517079919576645,
      "learning_rate": 8.476094090226026e-05,
      "loss": 0.8112,
      "mean_token_accuracy": 0.7700914926826954,
      "num_tokens": 158638886.0,
      "step": 4970
    },
    {
      "entropy": 0.8741231579333544,
      "epoch": 0.4573636724343161,
      "grad_norm": 0.1566043645143509,
      "learning_rate": 8.475787407611863e-05,
      "loss": 0.8469,
      "mean_token_accuracy": 0.7522480264306068,
      "num_tokens": 158671062.0,
      "step": 4971
    },
    {
      "entropy": 0.749415997415781,
      "epoch": 0.4574556788057573,
      "grad_norm": 0.14971202611923218,
      "learning_rate": 8.4754807249977e-05,
      "loss": 0.7097,
      "mean_token_accuracy": 0.7921625673770905,
      "num_tokens": 158702437.0,
      "step": 4972
    },
    {
      "entropy": 0.6980646755546331,
      "epoch": 0.4575476851771985,
      "grad_norm": 0.1478198766708374,
      "learning_rate": 8.475174042383538e-05,
      "loss": 0.6832,
      "mean_token_accuracy": 0.7955244742333889,
      "num_tokens": 158734246.0,
      "step": 4973
    },
    {
      "entropy": 0.9279488697648048,
      "epoch": 0.45763969154863976,
      "grad_norm": 0.16335900127887726,
      "learning_rate": 8.474867359769375e-05,
      "loss": 0.9242,
      "mean_token_accuracy": 0.7333844117820263,
      "num_tokens": 158766992.0,
      "step": 4974
    },
    {
      "entropy": 0.886914137750864,
      "epoch": 0.45773169792008095,
      "grad_norm": 0.1619744449853897,
      "learning_rate": 8.474560677155213e-05,
      "loss": 0.8783,
      "mean_token_accuracy": 0.7513157315552235,
      "num_tokens": 158799306.0,
      "step": 4975
    },
    {
      "entropy": 0.835733599960804,
      "epoch": 0.4578237042915222,
      "grad_norm": 0.16880559921264648,
      "learning_rate": 8.474253994541051e-05,
      "loss": 0.8349,
      "mean_token_accuracy": 0.7581713758409023,
      "num_tokens": 158830840.0,
      "step": 4976
    },
    {
      "entropy": 0.8531117979437113,
      "epoch": 0.45791571066296344,
      "grad_norm": 0.16628335416316986,
      "learning_rate": 8.473947311926888e-05,
      "loss": 0.86,
      "mean_token_accuracy": 0.7552536092698574,
      "num_tokens": 158862566.0,
      "step": 4977
    },
    {
      "entropy": 0.7892355211079121,
      "epoch": 0.4580077170344046,
      "grad_norm": 0.16317719221115112,
      "learning_rate": 8.473640629312724e-05,
      "loss": 0.7845,
      "mean_token_accuracy": 0.7753724902868271,
      "num_tokens": 158895334.0,
      "step": 4978
    },
    {
      "entropy": 0.7021135296672583,
      "epoch": 0.45809972340584587,
      "grad_norm": 0.15323224663734436,
      "learning_rate": 8.473333946698561e-05,
      "loss": 0.7054,
      "mean_token_accuracy": 0.799522053450346,
      "num_tokens": 158927561.0,
      "step": 4979
    },
    {
      "entropy": 0.8838718049228191,
      "epoch": 0.45819172977728706,
      "grad_norm": 0.15914617478847504,
      "learning_rate": 8.4730272640844e-05,
      "loss": 0.8826,
      "mean_token_accuracy": 0.7447794526815414,
      "num_tokens": 158959523.0,
      "step": 4980
    },
    {
      "entropy": 0.8052560426294804,
      "epoch": 0.4582837361487283,
      "grad_norm": 0.15678222477436066,
      "learning_rate": 8.472720581470238e-05,
      "loss": 0.7692,
      "mean_token_accuracy": 0.7698153108358383,
      "num_tokens": 158991204.0,
      "step": 4981
    },
    {
      "entropy": 0.7459018230438232,
      "epoch": 0.45837574252016955,
      "grad_norm": 0.15334481000900269,
      "learning_rate": 8.472413898856074e-05,
      "loss": 0.7204,
      "mean_token_accuracy": 0.7883375659584999,
      "num_tokens": 159023475.0,
      "step": 4982
    },
    {
      "entropy": 0.8280870951712132,
      "epoch": 0.45846774889161074,
      "grad_norm": 0.1710437536239624,
      "learning_rate": 8.472107216241911e-05,
      "loss": 0.8419,
      "mean_token_accuracy": 0.7563602700829506,
      "num_tokens": 159054955.0,
      "step": 4983
    },
    {
      "entropy": 0.7946829423308372,
      "epoch": 0.458559755263052,
      "grad_norm": 0.1655464470386505,
      "learning_rate": 8.471800533627749e-05,
      "loss": 0.8104,
      "mean_token_accuracy": 0.7644962333142757,
      "num_tokens": 159086963.0,
      "step": 4984
    },
    {
      "entropy": 0.9218880273401737,
      "epoch": 0.45865176163449317,
      "grad_norm": 0.16637694835662842,
      "learning_rate": 8.471493851013586e-05,
      "loss": 0.9254,
      "mean_token_accuracy": 0.7383156679570675,
      "num_tokens": 159118461.0,
      "step": 4985
    },
    {
      "entropy": 0.8164418451488018,
      "epoch": 0.4587437680059344,
      "grad_norm": 0.16028036177158356,
      "learning_rate": 8.471187168399424e-05,
      "loss": 0.7889,
      "mean_token_accuracy": 0.7717165760695934,
      "num_tokens": 159151206.0,
      "step": 4986
    },
    {
      "entropy": 0.7809179360046983,
      "epoch": 0.45883577437737566,
      "grad_norm": 0.1499899923801422,
      "learning_rate": 8.470880485785261e-05,
      "loss": 0.7553,
      "mean_token_accuracy": 0.7809395492076874,
      "num_tokens": 159183363.0,
      "step": 4987
    },
    {
      "entropy": 0.787173219025135,
      "epoch": 0.45892778074881685,
      "grad_norm": 0.15698009729385376,
      "learning_rate": 8.470573803171099e-05,
      "loss": 0.7735,
      "mean_token_accuracy": 0.7719193883240223,
      "num_tokens": 159214701.0,
      "step": 4988
    },
    {
      "entropy": 0.8697910867631435,
      "epoch": 0.4590197871202581,
      "grad_norm": 0.152859628200531,
      "learning_rate": 8.470267120556936e-05,
      "loss": 0.8634,
      "mean_token_accuracy": 0.7489760145545006,
      "num_tokens": 159246597.0,
      "step": 4989
    },
    {
      "entropy": 0.7986658932641149,
      "epoch": 0.4591117934916993,
      "grad_norm": 0.15286745131015778,
      "learning_rate": 8.469960437942773e-05,
      "loss": 0.789,
      "mean_token_accuracy": 0.7686457708477974,
      "num_tokens": 159278426.0,
      "step": 4990
    },
    {
      "entropy": 0.8591982759535313,
      "epoch": 0.4592037998631405,
      "grad_norm": 0.15621401369571686,
      "learning_rate": 8.469653755328611e-05,
      "loss": 0.8506,
      "mean_token_accuracy": 0.7523065842688084,
      "num_tokens": 159310902.0,
      "step": 4991
    },
    {
      "entropy": 0.7922393698245287,
      "epoch": 0.45929580623458177,
      "grad_norm": 0.1575496643781662,
      "learning_rate": 8.469347072714449e-05,
      "loss": 0.7751,
      "mean_token_accuracy": 0.769306093454361,
      "num_tokens": 159342324.0,
      "step": 4992
    },
    {
      "entropy": 0.8131189718842506,
      "epoch": 0.45938781260602296,
      "grad_norm": 0.1608714908361435,
      "learning_rate": 8.469040390100286e-05,
      "loss": 0.7946,
      "mean_token_accuracy": 0.7686200402677059,
      "num_tokens": 159374454.0,
      "step": 4993
    },
    {
      "entropy": 0.8277446441352367,
      "epoch": 0.4594798189774642,
      "grad_norm": 0.1575716882944107,
      "learning_rate": 8.468733707486123e-05,
      "loss": 0.8131,
      "mean_token_accuracy": 0.763876635581255,
      "num_tokens": 159406568.0,
      "step": 4994
    },
    {
      "entropy": 0.7842000126838684,
      "epoch": 0.4595718253489054,
      "grad_norm": 0.15984775125980377,
      "learning_rate": 8.468427024871959e-05,
      "loss": 0.7726,
      "mean_token_accuracy": 0.7750547602772713,
      "num_tokens": 159438622.0,
      "step": 4995
    },
    {
      "entropy": 0.7688613701611757,
      "epoch": 0.45966383172034664,
      "grad_norm": 0.15735682845115662,
      "learning_rate": 8.468120342257799e-05,
      "loss": 0.757,
      "mean_token_accuracy": 0.780013233423233,
      "num_tokens": 159470903.0,
      "step": 4996
    },
    {
      "entropy": 0.7903661970049143,
      "epoch": 0.4597558380917879,
      "grad_norm": 0.15785834193229675,
      "learning_rate": 8.467813659643636e-05,
      "loss": 0.7872,
      "mean_token_accuracy": 0.7753605246543884,
      "num_tokens": 159502173.0,
      "step": 4997
    },
    {
      "entropy": 0.6719992980360985,
      "epoch": 0.45984784446322907,
      "grad_norm": 0.15274180471897125,
      "learning_rate": 8.467506977029472e-05,
      "loss": 0.68,
      "mean_token_accuracy": 0.796235229820013,
      "num_tokens": 159534524.0,
      "step": 4998
    },
    {
      "entropy": 0.7826402969658375,
      "epoch": 0.4599398508346703,
      "grad_norm": 0.16303636133670807,
      "learning_rate": 8.467200294415309e-05,
      "loss": 0.7864,
      "mean_token_accuracy": 0.7690371237695217,
      "num_tokens": 159566076.0,
      "step": 4999
    },
    {
      "entropy": 0.7008463405072689,
      "epoch": 0.4600318572061115,
      "grad_norm": 0.15118035674095154,
      "learning_rate": 8.466893611801147e-05,
      "loss": 0.7054,
      "mean_token_accuracy": 0.7908666916191578,
      "num_tokens": 159597966.0,
      "step": 5000
    },
    {
      "entropy": 0.739465368911624,
      "epoch": 0.46012386357755275,
      "grad_norm": 0.16657303273677826,
      "learning_rate": 8.466586929186985e-05,
      "loss": 0.722,
      "mean_token_accuracy": 0.7915695011615753,
      "num_tokens": 159630543.0,
      "step": 5001
    },
    {
      "entropy": 0.7908969894051552,
      "epoch": 0.460215869948994,
      "grad_norm": 0.17229142785072327,
      "learning_rate": 8.466280246572822e-05,
      "loss": 0.7987,
      "mean_token_accuracy": 0.7734578661620617,
      "num_tokens": 159662419.0,
      "step": 5002
    },
    {
      "entropy": 0.7487604208290577,
      "epoch": 0.4603078763204352,
      "grad_norm": 0.16499359905719757,
      "learning_rate": 8.465973563958659e-05,
      "loss": 0.7452,
      "mean_token_accuracy": 0.7857366092503071,
      "num_tokens": 159695016.0,
      "step": 5003
    },
    {
      "entropy": 0.8081641588360071,
      "epoch": 0.4603998826918764,
      "grad_norm": 0.16752035915851593,
      "learning_rate": 8.465666881344497e-05,
      "loss": 0.805,
      "mean_token_accuracy": 0.7662557475268841,
      "num_tokens": 159726793.0,
      "step": 5004
    },
    {
      "entropy": 0.8523219097405672,
      "epoch": 0.4604918890633176,
      "grad_norm": 0.15513528883457184,
      "learning_rate": 8.465360198730334e-05,
      "loss": 0.8321,
      "mean_token_accuracy": 0.7607531696557999,
      "num_tokens": 159759335.0,
      "step": 5005
    },
    {
      "entropy": 0.695832546800375,
      "epoch": 0.46058389543475886,
      "grad_norm": 0.14557528495788574,
      "learning_rate": 8.465053516116172e-05,
      "loss": 0.6651,
      "mean_token_accuracy": 0.8012549132108688,
      "num_tokens": 159791669.0,
      "step": 5006
    },
    {
      "entropy": 0.8306816555559635,
      "epoch": 0.4606759018062001,
      "grad_norm": 0.15949071943759918,
      "learning_rate": 8.46474683350201e-05,
      "loss": 0.8226,
      "mean_token_accuracy": 0.762861680239439,
      "num_tokens": 159824112.0,
      "step": 5007
    },
    {
      "entropy": 0.8763149101287127,
      "epoch": 0.4607679081776413,
      "grad_norm": 0.1645764261484146,
      "learning_rate": 8.464440150887847e-05,
      "loss": 0.8578,
      "mean_token_accuracy": 0.7500432804226875,
      "num_tokens": 159855774.0,
      "step": 5008
    },
    {
      "entropy": 0.755445521324873,
      "epoch": 0.46085991454908254,
      "grad_norm": 0.16032259166240692,
      "learning_rate": 8.464133468273684e-05,
      "loss": 0.7282,
      "mean_token_accuracy": 0.7884494736790657,
      "num_tokens": 159886796.0,
      "step": 5009
    },
    {
      "entropy": 0.8614304848015308,
      "epoch": 0.46095192092052373,
      "grad_norm": 0.1615578979253769,
      "learning_rate": 8.46382678565952e-05,
      "loss": 0.8649,
      "mean_token_accuracy": 0.7469385489821434,
      "num_tokens": 159918251.0,
      "step": 5010
    },
    {
      "entropy": 0.9040755890309811,
      "epoch": 0.461043927291965,
      "grad_norm": 0.16120243072509766,
      "learning_rate": 8.463520103045359e-05,
      "loss": 0.8907,
      "mean_token_accuracy": 0.7436652779579163,
      "num_tokens": 159950617.0,
      "step": 5011
    },
    {
      "entropy": 0.9057325944304466,
      "epoch": 0.4611359336634062,
      "grad_norm": 0.17213569581508636,
      "learning_rate": 8.463213420431197e-05,
      "loss": 0.9331,
      "mean_token_accuracy": 0.7348585389554501,
      "num_tokens": 159981109.0,
      "step": 5012
    },
    {
      "entropy": 0.7609034944325686,
      "epoch": 0.4612279400348474,
      "grad_norm": 0.1504848748445511,
      "learning_rate": 8.462906737817034e-05,
      "loss": 0.746,
      "mean_token_accuracy": 0.7779707796871662,
      "num_tokens": 160013025.0,
      "step": 5013
    },
    {
      "entropy": 0.7804981945082545,
      "epoch": 0.46131994640628865,
      "grad_norm": 0.1573440581560135,
      "learning_rate": 8.46260005520287e-05,
      "loss": 0.7453,
      "mean_token_accuracy": 0.7820841558277607,
      "num_tokens": 160044829.0,
      "step": 5014
    },
    {
      "entropy": 0.7373662125319242,
      "epoch": 0.46141195277772984,
      "grad_norm": 0.16035370528697968,
      "learning_rate": 8.462293372588709e-05,
      "loss": 0.7238,
      "mean_token_accuracy": 0.7889597974717617,
      "num_tokens": 160075244.0,
      "step": 5015
    },
    {
      "entropy": 0.8745708521455526,
      "epoch": 0.4615039591491711,
      "grad_norm": 0.1637427806854248,
      "learning_rate": 8.461986689974545e-05,
      "loss": 0.8894,
      "mean_token_accuracy": 0.7454567290842533,
      "num_tokens": 160107249.0,
      "step": 5016
    },
    {
      "entropy": 0.8154531316831708,
      "epoch": 0.46159596552061233,
      "grad_norm": 0.1550368368625641,
      "learning_rate": 8.461680007360384e-05,
      "loss": 0.7961,
      "mean_token_accuracy": 0.7711509838700294,
      "num_tokens": 160139527.0,
      "step": 5017
    },
    {
      "entropy": 0.7557669673115015,
      "epoch": 0.4616879718920535,
      "grad_norm": 0.15059620141983032,
      "learning_rate": 8.46137332474622e-05,
      "loss": 0.7428,
      "mean_token_accuracy": 0.7795593403279781,
      "num_tokens": 160171238.0,
      "step": 5018
    },
    {
      "entropy": 0.7723291888833046,
      "epoch": 0.46177997826349476,
      "grad_norm": 0.159902885556221,
      "learning_rate": 8.461066642132058e-05,
      "loss": 0.7699,
      "mean_token_accuracy": 0.7789671123027802,
      "num_tokens": 160202935.0,
      "step": 5019
    },
    {
      "entropy": 0.8009373359382153,
      "epoch": 0.46187198463493595,
      "grad_norm": 0.15756110846996307,
      "learning_rate": 8.460759959517895e-05,
      "loss": 0.8001,
      "mean_token_accuracy": 0.7690596468746662,
      "num_tokens": 160234362.0,
      "step": 5020
    },
    {
      "entropy": 0.7910097911953926,
      "epoch": 0.4619639910063772,
      "grad_norm": 0.16344544291496277,
      "learning_rate": 8.460453276903732e-05,
      "loss": 0.795,
      "mean_token_accuracy": 0.7706142775714397,
      "num_tokens": 160266196.0,
      "step": 5021
    },
    {
      "entropy": 0.712792195379734,
      "epoch": 0.46205599737781844,
      "grad_norm": 0.16067279875278473,
      "learning_rate": 8.46014659428957e-05,
      "loss": 0.688,
      "mean_token_accuracy": 0.7941073440015316,
      "num_tokens": 160298294.0,
      "step": 5022
    },
    {
      "entropy": 0.8253338485956192,
      "epoch": 0.46214800374925963,
      "grad_norm": 0.16087722778320312,
      "learning_rate": 8.459839911675408e-05,
      "loss": 0.8345,
      "mean_token_accuracy": 0.7605328895151615,
      "num_tokens": 160330457.0,
      "step": 5023
    },
    {
      "entropy": 0.7650652043521404,
      "epoch": 0.4622400101207009,
      "grad_norm": 0.15154843032360077,
      "learning_rate": 8.459533229061245e-05,
      "loss": 0.7583,
      "mean_token_accuracy": 0.7802869901061058,
      "num_tokens": 160362076.0,
      "step": 5024
    },
    {
      "entropy": 0.744088577106595,
      "epoch": 0.46233201649214206,
      "grad_norm": 0.15022985637187958,
      "learning_rate": 8.459226546447082e-05,
      "loss": 0.7327,
      "mean_token_accuracy": 0.7830910086631775,
      "num_tokens": 160393730.0,
      "step": 5025
    },
    {
      "entropy": 0.752527391538024,
      "epoch": 0.4624240228635833,
      "grad_norm": 0.15335425734519958,
      "learning_rate": 8.458919863832919e-05,
      "loss": 0.7389,
      "mean_token_accuracy": 0.7833410501480103,
      "num_tokens": 160426295.0,
      "step": 5026
    },
    {
      "entropy": 0.7968931272625923,
      "epoch": 0.46251602923502455,
      "grad_norm": 0.15337516367435455,
      "learning_rate": 8.458613181218757e-05,
      "loss": 0.7984,
      "mean_token_accuracy": 0.7622955031692982,
      "num_tokens": 160457798.0,
      "step": 5027
    },
    {
      "entropy": 0.7746553644537926,
      "epoch": 0.46260803560646574,
      "grad_norm": 0.15719915926456451,
      "learning_rate": 8.458306498604595e-05,
      "loss": 0.7717,
      "mean_token_accuracy": 0.7730823568999767,
      "num_tokens": 160489707.0,
      "step": 5028
    },
    {
      "entropy": 0.7631080467253923,
      "epoch": 0.462700041977907,
      "grad_norm": 0.1607808917760849,
      "learning_rate": 8.457999815990432e-05,
      "loss": 0.7572,
      "mean_token_accuracy": 0.7845116443932056,
      "num_tokens": 160522264.0,
      "step": 5029
    },
    {
      "entropy": 0.8325823694467545,
      "epoch": 0.4627920483493482,
      "grad_norm": 0.15355457365512848,
      "learning_rate": 8.457693133376269e-05,
      "loss": 0.8223,
      "mean_token_accuracy": 0.7630874402821064,
      "num_tokens": 160554591.0,
      "step": 5030
    },
    {
      "entropy": 0.8791149873286486,
      "epoch": 0.4628840547207894,
      "grad_norm": 0.16308540105819702,
      "learning_rate": 8.457386450762107e-05,
      "loss": 0.8488,
      "mean_token_accuracy": 0.7521821819245815,
      "num_tokens": 160585775.0,
      "step": 5031
    },
    {
      "entropy": 0.7605756912380457,
      "epoch": 0.46297606109223066,
      "grad_norm": 0.163997083902359,
      "learning_rate": 8.457079768147943e-05,
      "loss": 0.7717,
      "mean_token_accuracy": 0.7745246961712837,
      "num_tokens": 160617893.0,
      "step": 5032
    },
    {
      "entropy": 0.8529201988130808,
      "epoch": 0.46306806746367185,
      "grad_norm": 0.163122296333313,
      "learning_rate": 8.456773085533782e-05,
      "loss": 0.8394,
      "mean_token_accuracy": 0.7552995532751083,
      "num_tokens": 160649383.0,
      "step": 5033
    },
    {
      "entropy": 0.8562459666281939,
      "epoch": 0.4631600738351131,
      "grad_norm": 0.16897337138652802,
      "learning_rate": 8.45646640291962e-05,
      "loss": 0.8596,
      "mean_token_accuracy": 0.756764505058527,
      "num_tokens": 160680897.0,
      "step": 5034
    },
    {
      "entropy": 0.7273796088993549,
      "epoch": 0.4632520802065543,
      "grad_norm": 0.15887516736984253,
      "learning_rate": 8.456159720305457e-05,
      "loss": 0.7015,
      "mean_token_accuracy": 0.795278862118721,
      "num_tokens": 160712580.0,
      "step": 5035
    },
    {
      "entropy": 0.9323499016463757,
      "epoch": 0.46334408657799553,
      "grad_norm": 0.15859580039978027,
      "learning_rate": 8.455853037691293e-05,
      "loss": 0.8978,
      "mean_token_accuracy": 0.741994708776474,
      "num_tokens": 160744215.0,
      "step": 5036
    },
    {
      "entropy": 0.8817060627043247,
      "epoch": 0.4634360929494368,
      "grad_norm": 0.164163738489151,
      "learning_rate": 8.45554635507713e-05,
      "loss": 0.878,
      "mean_token_accuracy": 0.7443877495825291,
      "num_tokens": 160775543.0,
      "step": 5037
    },
    {
      "entropy": 0.8863855414092541,
      "epoch": 0.46352809932087796,
      "grad_norm": 0.16779300570487976,
      "learning_rate": 8.45523967246297e-05,
      "loss": 0.8719,
      "mean_token_accuracy": 0.7450345605611801,
      "num_tokens": 160806811.0,
      "step": 5038
    },
    {
      "entropy": 0.7863592561334372,
      "epoch": 0.4636201056923192,
      "grad_norm": 0.15563054382801056,
      "learning_rate": 8.454932989848806e-05,
      "loss": 0.7753,
      "mean_token_accuracy": 0.7673834711313248,
      "num_tokens": 160838518.0,
      "step": 5039
    },
    {
      "entropy": 0.8201106004416943,
      "epoch": 0.4637121120637604,
      "grad_norm": 0.15841446816921234,
      "learning_rate": 8.454626307234643e-05,
      "loss": 0.8155,
      "mean_token_accuracy": 0.7632531151175499,
      "num_tokens": 160870130.0,
      "step": 5040
    },
    {
      "entropy": 0.7934239860624075,
      "epoch": 0.46380411843520164,
      "grad_norm": 0.1587304174900055,
      "learning_rate": 8.45431962462048e-05,
      "loss": 0.7924,
      "mean_token_accuracy": 0.7663294225931168,
      "num_tokens": 160902328.0,
      "step": 5041
    },
    {
      "entropy": 0.8437720406800508,
      "epoch": 0.4638961248066429,
      "grad_norm": 0.15671658515930176,
      "learning_rate": 8.454012942006318e-05,
      "loss": 0.8315,
      "mean_token_accuracy": 0.7524590343236923,
      "num_tokens": 160934225.0,
      "step": 5042
    },
    {
      "entropy": 0.7910181879997253,
      "epoch": 0.4639881311780841,
      "grad_norm": 0.16718682646751404,
      "learning_rate": 8.453706259392156e-05,
      "loss": 0.803,
      "mean_token_accuracy": 0.7679471746087074,
      "num_tokens": 160966507.0,
      "step": 5043
    },
    {
      "entropy": 0.7671894561499357,
      "epoch": 0.4640801375495253,
      "grad_norm": 0.16005240380764008,
      "learning_rate": 8.453399576777993e-05,
      "loss": 0.7643,
      "mean_token_accuracy": 0.7717457599937916,
      "num_tokens": 160997604.0,
      "step": 5044
    },
    {
      "entropy": 0.7402691021561623,
      "epoch": 0.4641721439209665,
      "grad_norm": 0.15029050409793854,
      "learning_rate": 8.45309289416383e-05,
      "loss": 0.7355,
      "mean_token_accuracy": 0.782118272036314,
      "num_tokens": 161029222.0,
      "step": 5045
    },
    {
      "entropy": 0.7824805751442909,
      "epoch": 0.46426415029240775,
      "grad_norm": 0.15556032955646515,
      "learning_rate": 8.452786211549668e-05,
      "loss": 0.774,
      "mean_token_accuracy": 0.7716459780931473,
      "num_tokens": 161061631.0,
      "step": 5046
    },
    {
      "entropy": 0.87641785107553,
      "epoch": 0.464356156663849,
      "grad_norm": 0.15664108097553253,
      "learning_rate": 8.452479528935505e-05,
      "loss": 0.8652,
      "mean_token_accuracy": 0.7544499449431896,
      "num_tokens": 161093080.0,
      "step": 5047
    },
    {
      "entropy": 0.6944600380957127,
      "epoch": 0.4644481630352902,
      "grad_norm": 0.1528160721063614,
      "learning_rate": 8.452172846321343e-05,
      "loss": 0.6789,
      "mean_token_accuracy": 0.7991579882800579,
      "num_tokens": 161125095.0,
      "step": 5048
    },
    {
      "entropy": 0.7954096235334873,
      "epoch": 0.46454016940673143,
      "grad_norm": 0.15709035098552704,
      "learning_rate": 8.45186616370718e-05,
      "loss": 0.7888,
      "mean_token_accuracy": 0.7696265168488026,
      "num_tokens": 161157448.0,
      "step": 5049
    },
    {
      "entropy": 0.7263982836157084,
      "epoch": 0.4646321757781726,
      "grad_norm": 0.15453946590423584,
      "learning_rate": 8.451559481093018e-05,
      "loss": 0.6993,
      "mean_token_accuracy": 0.7922462895512581,
      "num_tokens": 161189898.0,
      "step": 5050
    },
    {
      "entropy": 0.8543830756098032,
      "epoch": 0.46472418214961386,
      "grad_norm": 0.16219748556613922,
      "learning_rate": 8.451252798478855e-05,
      "loss": 0.8397,
      "mean_token_accuracy": 0.7570863142609596,
      "num_tokens": 161220896.0,
      "step": 5051
    },
    {
      "entropy": 0.8166995830833912,
      "epoch": 0.4648161885210551,
      "grad_norm": 0.15854328870773315,
      "learning_rate": 8.450946115864691e-05,
      "loss": 0.7935,
      "mean_token_accuracy": 0.770126786082983,
      "num_tokens": 161251915.0,
      "step": 5052
    },
    {
      "entropy": 0.8696986362338066,
      "epoch": 0.4649081948924963,
      "grad_norm": 0.16403073072433472,
      "learning_rate": 8.45063943325053e-05,
      "loss": 0.8457,
      "mean_token_accuracy": 0.757647629827261,
      "num_tokens": 161283659.0,
      "step": 5053
    },
    {
      "entropy": 0.8649376779794693,
      "epoch": 0.46500020126393754,
      "grad_norm": 0.16504338383674622,
      "learning_rate": 8.450332750636368e-05,
      "loss": 0.8688,
      "mean_token_accuracy": 0.7542852535843849,
      "num_tokens": 161315989.0,
      "step": 5054
    },
    {
      "entropy": 0.6854433435946703,
      "epoch": 0.46509220763537873,
      "grad_norm": 0.14882387220859528,
      "learning_rate": 8.450026068022204e-05,
      "loss": 0.67,
      "mean_token_accuracy": 0.8026445433497429,
      "num_tokens": 161348269.0,
      "step": 5055
    },
    {
      "entropy": 0.7460649702697992,
      "epoch": 0.46518421400682,
      "grad_norm": 0.1651267558336258,
      "learning_rate": 8.449719385408041e-05,
      "loss": 0.743,
      "mean_token_accuracy": 0.7825619839131832,
      "num_tokens": 161380651.0,
      "step": 5056
    },
    {
      "entropy": 0.8191963322460651,
      "epoch": 0.4652762203782612,
      "grad_norm": 0.15682126581668854,
      "learning_rate": 8.449412702793878e-05,
      "loss": 0.8007,
      "mean_token_accuracy": 0.7669363357126713,
      "num_tokens": 161412479.0,
      "step": 5057
    },
    {
      "entropy": 0.8125718627125025,
      "epoch": 0.4653682267497024,
      "grad_norm": 0.16396218538284302,
      "learning_rate": 8.449106020179716e-05,
      "loss": 0.8136,
      "mean_token_accuracy": 0.7651096843183041,
      "num_tokens": 161444442.0,
      "step": 5058
    },
    {
      "entropy": 0.7050856277346611,
      "epoch": 0.46546023312114365,
      "grad_norm": 0.15926367044448853,
      "learning_rate": 8.448799337565554e-05,
      "loss": 0.7041,
      "mean_token_accuracy": 0.7950782738626003,
      "num_tokens": 161477140.0,
      "step": 5059
    },
    {
      "entropy": 0.8309361506253481,
      "epoch": 0.46555223949258484,
      "grad_norm": 0.16195176541805267,
      "learning_rate": 8.448492654951391e-05,
      "loss": 0.8138,
      "mean_token_accuracy": 0.7637379691004753,
      "num_tokens": 161508082.0,
      "step": 5060
    },
    {
      "entropy": 0.8478915505111217,
      "epoch": 0.4656442458640261,
      "grad_norm": 0.17002321779727936,
      "learning_rate": 8.448185972337228e-05,
      "loss": 0.8491,
      "mean_token_accuracy": 0.7531306855380535,
      "num_tokens": 161540426.0,
      "step": 5061
    },
    {
      "entropy": 0.7300110515207052,
      "epoch": 0.46573625223546733,
      "grad_norm": 0.184764564037323,
      "learning_rate": 8.447879289723066e-05,
      "loss": 0.733,
      "mean_token_accuracy": 0.7867409363389015,
      "num_tokens": 161572369.0,
      "step": 5062
    },
    {
      "entropy": 0.7423068881034851,
      "epoch": 0.4658282586069085,
      "grad_norm": 0.16041815280914307,
      "learning_rate": 8.447572607108903e-05,
      "loss": 0.7502,
      "mean_token_accuracy": 0.7800585739314556,
      "num_tokens": 161603255.0,
      "step": 5063
    },
    {
      "entropy": 0.6960385739803314,
      "epoch": 0.46592026497834976,
      "grad_norm": 0.15815313160419464,
      "learning_rate": 8.447265924494741e-05,
      "loss": 0.684,
      "mean_token_accuracy": 0.7943488694727421,
      "num_tokens": 161634485.0,
      "step": 5064
    },
    {
      "entropy": 0.7278556395322084,
      "epoch": 0.46601227134979095,
      "grad_norm": 0.1549481600522995,
      "learning_rate": 8.446959241880579e-05,
      "loss": 0.7106,
      "mean_token_accuracy": 0.7874641045928001,
      "num_tokens": 161666895.0,
      "step": 5065
    },
    {
      "entropy": 0.7588498592376709,
      "epoch": 0.4661042777212322,
      "grad_norm": 0.16073711216449738,
      "learning_rate": 8.446652559266416e-05,
      "loss": 0.753,
      "mean_token_accuracy": 0.7785639017820358,
      "num_tokens": 161698961.0,
      "step": 5066
    },
    {
      "entropy": 0.8026615427806973,
      "epoch": 0.46619628409267344,
      "grad_norm": 0.16757701337337494,
      "learning_rate": 8.446345876652253e-05,
      "loss": 0.8109,
      "mean_token_accuracy": 0.7689132019877434,
      "num_tokens": 161730570.0,
      "step": 5067
    },
    {
      "entropy": 0.8004765994846821,
      "epoch": 0.46628829046411463,
      "grad_norm": 0.1595907211303711,
      "learning_rate": 8.44603919403809e-05,
      "loss": 0.8053,
      "mean_token_accuracy": 0.7708982229232788,
      "num_tokens": 161762195.0,
      "step": 5068
    },
    {
      "entropy": 0.7903331574052572,
      "epoch": 0.4663802968355559,
      "grad_norm": 0.15254689753055573,
      "learning_rate": 8.445732511423928e-05,
      "loss": 0.7692,
      "mean_token_accuracy": 0.7766623869538307,
      "num_tokens": 161794523.0,
      "step": 5069
    },
    {
      "entropy": 0.8946295380592346,
      "epoch": 0.46647230320699706,
      "grad_norm": 0.15978294610977173,
      "learning_rate": 8.445425828809766e-05,
      "loss": 0.8902,
      "mean_token_accuracy": 0.7435061559081078,
      "num_tokens": 161827250.0,
      "step": 5070
    },
    {
      "entropy": 0.7569933068007231,
      "epoch": 0.4665643095784383,
      "grad_norm": 0.1501714289188385,
      "learning_rate": 8.445119146195603e-05,
      "loss": 0.7421,
      "mean_token_accuracy": 0.7804436944425106,
      "num_tokens": 161859240.0,
      "step": 5071
    },
    {
      "entropy": 0.7937517985701561,
      "epoch": 0.46665631594987955,
      "grad_norm": 0.15777145326137543,
      "learning_rate": 8.44481246358144e-05,
      "loss": 0.7712,
      "mean_token_accuracy": 0.7744518555700779,
      "num_tokens": 161891506.0,
      "step": 5072
    },
    {
      "entropy": 0.7873483784496784,
      "epoch": 0.46674832232132074,
      "grad_norm": 0.15736189484596252,
      "learning_rate": 8.444505780967277e-05,
      "loss": 0.7798,
      "mean_token_accuracy": 0.7706999517977238,
      "num_tokens": 161923968.0,
      "step": 5073
    },
    {
      "entropy": 0.8645448759198189,
      "epoch": 0.466840328692762,
      "grad_norm": 0.15921159088611603,
      "learning_rate": 8.444199098353116e-05,
      "loss": 0.8308,
      "mean_token_accuracy": 0.7642213553190231,
      "num_tokens": 161955844.0,
      "step": 5074
    },
    {
      "entropy": 0.8656888455152512,
      "epoch": 0.4669323350642032,
      "grad_norm": 0.15402851998806,
      "learning_rate": 8.443892415738952e-05,
      "loss": 0.8468,
      "mean_token_accuracy": 0.7523282393813133,
      "num_tokens": 161988253.0,
      "step": 5075
    },
    {
      "entropy": 0.7576217409223318,
      "epoch": 0.4670243414356444,
      "grad_norm": 0.1502183973789215,
      "learning_rate": 8.443585733124789e-05,
      "loss": 0.7435,
      "mean_token_accuracy": 0.7797163277864456,
      "num_tokens": 162020123.0,
      "step": 5076
    },
    {
      "entropy": 0.8505578842014074,
      "epoch": 0.46711634780708566,
      "grad_norm": 0.16080515086650848,
      "learning_rate": 8.443279050510627e-05,
      "loss": 0.8337,
      "mean_token_accuracy": 0.7622686959803104,
      "num_tokens": 162052442.0,
      "step": 5077
    },
    {
      "entropy": 0.8296422548592091,
      "epoch": 0.46720835417852685,
      "grad_norm": 0.1587674766778946,
      "learning_rate": 8.442972367896464e-05,
      "loss": 0.8137,
      "mean_token_accuracy": 0.7642454765737057,
      "num_tokens": 162084557.0,
      "step": 5078
    },
    {
      "entropy": 0.6598470341414213,
      "epoch": 0.4673003605499681,
      "grad_norm": 0.1531548947095871,
      "learning_rate": 8.442665685282302e-05,
      "loss": 0.6385,
      "mean_token_accuracy": 0.8097478076815605,
      "num_tokens": 162116242.0,
      "step": 5079
    },
    {
      "entropy": 0.8580582812428474,
      "epoch": 0.4673923669214093,
      "grad_norm": 0.17188270390033722,
      "learning_rate": 8.442359002668139e-05,
      "loss": 0.8791,
      "mean_token_accuracy": 0.751896895468235,
      "num_tokens": 162148552.0,
      "step": 5080
    },
    {
      "entropy": 0.776639349758625,
      "epoch": 0.46748437329285053,
      "grad_norm": 0.1554788500070572,
      "learning_rate": 8.442052320053977e-05,
      "loss": 0.7665,
      "mean_token_accuracy": 0.7756278924643993,
      "num_tokens": 162180571.0,
      "step": 5081
    },
    {
      "entropy": 0.8275728933513165,
      "epoch": 0.4675763796642918,
      "grad_norm": 0.16730491816997528,
      "learning_rate": 8.441745637439814e-05,
      "loss": 0.8374,
      "mean_token_accuracy": 0.7586451582610607,
      "num_tokens": 162212531.0,
      "step": 5082
    },
    {
      "entropy": 0.7876517362892628,
      "epoch": 0.46766838603573296,
      "grad_norm": 0.16049638390541077,
      "learning_rate": 8.441438954825651e-05,
      "loss": 0.7776,
      "mean_token_accuracy": 0.7749780155718327,
      "num_tokens": 162244499.0,
      "step": 5083
    },
    {
      "entropy": 0.7168771028518677,
      "epoch": 0.4677603924071742,
      "grad_norm": 0.1620548814535141,
      "learning_rate": 8.441132272211489e-05,
      "loss": 0.7093,
      "mean_token_accuracy": 0.792482890188694,
      "num_tokens": 162276614.0,
      "step": 5084
    },
    {
      "entropy": 0.6342466324567795,
      "epoch": 0.4678523987786154,
      "grad_norm": 0.15715546905994415,
      "learning_rate": 8.440825589597327e-05,
      "loss": 0.612,
      "mean_token_accuracy": 0.8195695020258427,
      "num_tokens": 162308069.0,
      "step": 5085
    },
    {
      "entropy": 0.7389107178896666,
      "epoch": 0.46794440515005664,
      "grad_norm": 0.16185656189918518,
      "learning_rate": 8.440518906983164e-05,
      "loss": 0.7167,
      "mean_token_accuracy": 0.785258088260889,
      "num_tokens": 162338997.0,
      "step": 5086
    },
    {
      "entropy": 0.6843371614813805,
      "epoch": 0.4680364115214979,
      "grad_norm": 0.15772201120853424,
      "learning_rate": 8.440212224369e-05,
      "loss": 0.6824,
      "mean_token_accuracy": 0.79690932482481,
      "num_tokens": 162371705.0,
      "step": 5087
    },
    {
      "entropy": 0.808480478823185,
      "epoch": 0.4681284178929391,
      "grad_norm": 0.16113406419754028,
      "learning_rate": 8.439905541754837e-05,
      "loss": 0.8126,
      "mean_token_accuracy": 0.7635851316154003,
      "num_tokens": 162404041.0,
      "step": 5088
    },
    {
      "entropy": 0.8566452525556087,
      "epoch": 0.4682204242643803,
      "grad_norm": 0.1560635268688202,
      "learning_rate": 8.439598859140676e-05,
      "loss": 0.8459,
      "mean_token_accuracy": 0.7565485760569572,
      "num_tokens": 162436611.0,
      "step": 5089
    },
    {
      "entropy": 0.7234358694404364,
      "epoch": 0.4683124306358215,
      "grad_norm": 0.15065594017505646,
      "learning_rate": 8.439292176526514e-05,
      "loss": 0.7103,
      "mean_token_accuracy": 0.7921523861587048,
      "num_tokens": 162468676.0,
      "step": 5090
    },
    {
      "entropy": 0.7159870527684689,
      "epoch": 0.46840443700726275,
      "grad_norm": 0.15306779742240906,
      "learning_rate": 8.43898549391235e-05,
      "loss": 0.7205,
      "mean_token_accuracy": 0.7852265797555447,
      "num_tokens": 162500968.0,
      "step": 5091
    },
    {
      "entropy": 0.8000919632613659,
      "epoch": 0.468496443378704,
      "grad_norm": 0.16727635264396667,
      "learning_rate": 8.438678811298187e-05,
      "loss": 0.7828,
      "mean_token_accuracy": 0.7749819122254848,
      "num_tokens": 162531977.0,
      "step": 5092
    },
    {
      "entropy": 0.7782155722379684,
      "epoch": 0.4685884497501452,
      "grad_norm": 0.1669309139251709,
      "learning_rate": 8.438372128684025e-05,
      "loss": 0.7836,
      "mean_token_accuracy": 0.7678747363388538,
      "num_tokens": 162563735.0,
      "step": 5093
    },
    {
      "entropy": 0.8788433894515038,
      "epoch": 0.46868045612158643,
      "grad_norm": 0.16553540527820587,
      "learning_rate": 8.438065446069862e-05,
      "loss": 0.8704,
      "mean_token_accuracy": 0.751151580363512,
      "num_tokens": 162595327.0,
      "step": 5094
    },
    {
      "entropy": 0.6940530687570572,
      "epoch": 0.4687724624930276,
      "grad_norm": 0.14830735325813293,
      "learning_rate": 8.4377587634557e-05,
      "loss": 0.6621,
      "mean_token_accuracy": 0.8035543337464333,
      "num_tokens": 162627924.0,
      "step": 5095
    },
    {
      "entropy": 0.8866175040602684,
      "epoch": 0.46886446886446886,
      "grad_norm": 0.16345511376857758,
      "learning_rate": 8.437452080841539e-05,
      "loss": 0.877,
      "mean_token_accuracy": 0.7503594160079956,
      "num_tokens": 162659962.0,
      "step": 5096
    },
    {
      "entropy": 0.8630284313112497,
      "epoch": 0.4689564752359101,
      "grad_norm": 0.16132450103759766,
      "learning_rate": 8.437145398227375e-05,
      "loss": 0.8782,
      "mean_token_accuracy": 0.7456680834293365,
      "num_tokens": 162691166.0,
      "step": 5097
    },
    {
      "entropy": 0.7531932778656483,
      "epoch": 0.4690484816073513,
      "grad_norm": 0.15542426705360413,
      "learning_rate": 8.436838715613212e-05,
      "loss": 0.7309,
      "mean_token_accuracy": 0.7924000471830368,
      "num_tokens": 162722711.0,
      "step": 5098
    },
    {
      "entropy": 0.797323590144515,
      "epoch": 0.46914048797879254,
      "grad_norm": 0.16265109181404114,
      "learning_rate": 8.436532032999049e-05,
      "loss": 0.791,
      "mean_token_accuracy": 0.7683176957070827,
      "num_tokens": 162755026.0,
      "step": 5099
    },
    {
      "entropy": 0.6291262358427048,
      "epoch": 0.46923249435023373,
      "grad_norm": 0.1579522341489792,
      "learning_rate": 8.436225350384887e-05,
      "loss": 0.6198,
      "mean_token_accuracy": 0.8128615580499172,
      "num_tokens": 162786488.0,
      "step": 5100
    },
    {
      "entropy": 0.8017503656446934,
      "epoch": 0.469324500721675,
      "grad_norm": 0.15622247755527496,
      "learning_rate": 8.435918667770725e-05,
      "loss": 0.8064,
      "mean_token_accuracy": 0.765576209872961,
      "num_tokens": 162818045.0,
      "step": 5101
    },
    {
      "entropy": 0.7100887726992369,
      "epoch": 0.4694165070931162,
      "grad_norm": 0.15721169114112854,
      "learning_rate": 8.435611985156562e-05,
      "loss": 0.7009,
      "mean_token_accuracy": 0.7933211363852024,
      "num_tokens": 162849414.0,
      "step": 5102
    },
    {
      "entropy": 0.6853756885975599,
      "epoch": 0.4695085134645574,
      "grad_norm": 0.16406817734241486,
      "learning_rate": 8.435305302542399e-05,
      "loss": 0.6778,
      "mean_token_accuracy": 0.7973126024007797,
      "num_tokens": 162881804.0,
      "step": 5103
    },
    {
      "entropy": 0.8263371586799622,
      "epoch": 0.46960051983599865,
      "grad_norm": 0.16099749505519867,
      "learning_rate": 8.434998619928237e-05,
      "loss": 0.8119,
      "mean_token_accuracy": 0.7674429081380367,
      "num_tokens": 162913272.0,
      "step": 5104
    },
    {
      "entropy": 0.757466021925211,
      "epoch": 0.46969252620743984,
      "grad_norm": 0.16136032342910767,
      "learning_rate": 8.434691937314074e-05,
      "loss": 0.7365,
      "mean_token_accuracy": 0.7845272235572338,
      "num_tokens": 162944881.0,
      "step": 5105
    },
    {
      "entropy": 0.8134964033961296,
      "epoch": 0.4697845325788811,
      "grad_norm": 0.1645130217075348,
      "learning_rate": 8.434385254699912e-05,
      "loss": 0.8072,
      "mean_token_accuracy": 0.766050785779953,
      "num_tokens": 162976372.0,
      "step": 5106
    },
    {
      "entropy": 0.8226194679737091,
      "epoch": 0.46987653895032233,
      "grad_norm": 0.15732118487358093,
      "learning_rate": 8.434078572085749e-05,
      "loss": 0.8018,
      "mean_token_accuracy": 0.7669225931167603,
      "num_tokens": 163007831.0,
      "step": 5107
    },
    {
      "entropy": 0.8219072222709656,
      "epoch": 0.4699685453217635,
      "grad_norm": 0.1596364825963974,
      "learning_rate": 8.433771889471587e-05,
      "loss": 0.826,
      "mean_token_accuracy": 0.759927049279213,
      "num_tokens": 163039467.0,
      "step": 5108
    },
    {
      "entropy": 0.7832695432007313,
      "epoch": 0.47006055169320476,
      "grad_norm": 0.1681356132030487,
      "learning_rate": 8.433465206857424e-05,
      "loss": 0.7834,
      "mean_token_accuracy": 0.7749642543494701,
      "num_tokens": 163071049.0,
      "step": 5109
    },
    {
      "entropy": 0.8436054699122906,
      "epoch": 0.47015255806464595,
      "grad_norm": 0.17082107067108154,
      "learning_rate": 8.43315852424326e-05,
      "loss": 0.8528,
      "mean_token_accuracy": 0.7528316825628281,
      "num_tokens": 163103620.0,
      "step": 5110
    },
    {
      "entropy": 0.7007008194923401,
      "epoch": 0.4702445644360872,
      "grad_norm": 0.15040522813796997,
      "learning_rate": 8.432851841629098e-05,
      "loss": 0.6654,
      "mean_token_accuracy": 0.8034613281488419,
      "num_tokens": 163135909.0,
      "step": 5111
    },
    {
      "entropy": 0.8438929934054613,
      "epoch": 0.47033657080752844,
      "grad_norm": 0.16711637377738953,
      "learning_rate": 8.432545159014937e-05,
      "loss": 0.8365,
      "mean_token_accuracy": 0.7609852366149426,
      "num_tokens": 163167280.0,
      "step": 5112
    },
    {
      "entropy": 0.9005765020847321,
      "epoch": 0.47042857717896963,
      "grad_norm": 0.16726639866828918,
      "learning_rate": 8.432238476400773e-05,
      "loss": 0.8844,
      "mean_token_accuracy": 0.738565381616354,
      "num_tokens": 163197698.0,
      "step": 5113
    },
    {
      "entropy": 0.8548414893448353,
      "epoch": 0.4705205835504109,
      "grad_norm": 0.15877525508403778,
      "learning_rate": 8.43193179378661e-05,
      "loss": 0.8431,
      "mean_token_accuracy": 0.7564977258443832,
      "num_tokens": 163229862.0,
      "step": 5114
    },
    {
      "entropy": 0.7224305961281061,
      "epoch": 0.47061258992185206,
      "grad_norm": 0.15541839599609375,
      "learning_rate": 8.431625111172447e-05,
      "loss": 0.7007,
      "mean_token_accuracy": 0.7978767417371273,
      "num_tokens": 163261411.0,
      "step": 5115
    },
    {
      "entropy": 0.8591271676123142,
      "epoch": 0.4707045962932933,
      "grad_norm": 0.15814900398254395,
      "learning_rate": 8.431318428558286e-05,
      "loss": 0.8614,
      "mean_token_accuracy": 0.7491197772324085,
      "num_tokens": 163293335.0,
      "step": 5116
    },
    {
      "entropy": 0.7517964392900467,
      "epoch": 0.47079660266473455,
      "grad_norm": 0.15690384805202484,
      "learning_rate": 8.431011745944123e-05,
      "loss": 0.7655,
      "mean_token_accuracy": 0.7754490189254284,
      "num_tokens": 163325604.0,
      "step": 5117
    },
    {
      "entropy": 0.6958468779921532,
      "epoch": 0.47088860903617574,
      "grad_norm": 0.16445787250995636,
      "learning_rate": 8.43070506332996e-05,
      "loss": 0.682,
      "mean_token_accuracy": 0.7978048473596573,
      "num_tokens": 163357374.0,
      "step": 5118
    },
    {
      "entropy": 0.8328113611787558,
      "epoch": 0.470980615407617,
      "grad_norm": 0.16162541508674622,
      "learning_rate": 8.430398380715797e-05,
      "loss": 0.8444,
      "mean_token_accuracy": 0.756453949958086,
      "num_tokens": 163389210.0,
      "step": 5119
    },
    {
      "entropy": 0.8167297802865505,
      "epoch": 0.4710726217790582,
      "grad_norm": 0.1545013040304184,
      "learning_rate": 8.430091698101635e-05,
      "loss": 0.8151,
      "mean_token_accuracy": 0.7667788080871105,
      "num_tokens": 163421271.0,
      "step": 5120
    },
    {
      "entropy": 0.8157068826258183,
      "epoch": 0.4711646281504994,
      "grad_norm": 0.15805213153362274,
      "learning_rate": 8.429785015487473e-05,
      "loss": 0.8208,
      "mean_token_accuracy": 0.7626765929162502,
      "num_tokens": 163453917.0,
      "step": 5121
    },
    {
      "entropy": 0.7248799819499254,
      "epoch": 0.47125663452194066,
      "grad_norm": 0.15965767204761505,
      "learning_rate": 8.42947833287331e-05,
      "loss": 0.7078,
      "mean_token_accuracy": 0.7930687889456749,
      "num_tokens": 163486009.0,
      "step": 5122
    },
    {
      "entropy": 0.7411667536944151,
      "epoch": 0.47134864089338185,
      "grad_norm": 0.16252177953720093,
      "learning_rate": 8.429171650259147e-05,
      "loss": 0.7249,
      "mean_token_accuracy": 0.7865158803761005,
      "num_tokens": 163517610.0,
      "step": 5123
    },
    {
      "entropy": 0.7396797854453325,
      "epoch": 0.4714406472648231,
      "grad_norm": 0.15131886303424835,
      "learning_rate": 8.428864967644985e-05,
      "loss": 0.7145,
      "mean_token_accuracy": 0.7893323004245758,
      "num_tokens": 163549811.0,
      "step": 5124
    },
    {
      "entropy": 0.7995414715260267,
      "epoch": 0.4715326536362643,
      "grad_norm": 0.16474758088588715,
      "learning_rate": 8.428558285030822e-05,
      "loss": 0.7771,
      "mean_token_accuracy": 0.771687675267458,
      "num_tokens": 163580980.0,
      "step": 5125
    },
    {
      "entropy": 0.879821989685297,
      "epoch": 0.47162466000770553,
      "grad_norm": 0.1597052663564682,
      "learning_rate": 8.42825160241666e-05,
      "loss": 0.8573,
      "mean_token_accuracy": 0.7520024217665195,
      "num_tokens": 163612273.0,
      "step": 5126
    },
    {
      "entropy": 0.7675926126539707,
      "epoch": 0.4717166663791468,
      "grad_norm": 0.16118408739566803,
      "learning_rate": 8.427944919802498e-05,
      "loss": 0.77,
      "mean_token_accuracy": 0.7788213305175304,
      "num_tokens": 163644988.0,
      "step": 5127
    },
    {
      "entropy": 0.772876437753439,
      "epoch": 0.47180867275058797,
      "grad_norm": 0.16167204082012177,
      "learning_rate": 8.427638237188335e-05,
      "loss": 0.7646,
      "mean_token_accuracy": 0.7790738269686699,
      "num_tokens": 163677278.0,
      "step": 5128
    },
    {
      "entropy": 0.8704407718032598,
      "epoch": 0.4719006791220292,
      "grad_norm": 0.167560875415802,
      "learning_rate": 8.427331554574171e-05,
      "loss": 0.8622,
      "mean_token_accuracy": 0.7534092292189598,
      "num_tokens": 163707813.0,
      "step": 5129
    },
    {
      "entropy": 0.8373424261808395,
      "epoch": 0.4719926854934704,
      "grad_norm": 0.1625479757785797,
      "learning_rate": 8.427024871960008e-05,
      "loss": 0.8494,
      "mean_token_accuracy": 0.7505929209291935,
      "num_tokens": 163740052.0,
      "step": 5130
    },
    {
      "entropy": 0.851405143737793,
      "epoch": 0.47208469186491164,
      "grad_norm": 0.1646070033311844,
      "learning_rate": 8.426718189345846e-05,
      "loss": 0.8281,
      "mean_token_accuracy": 0.7590467035770416,
      "num_tokens": 163771909.0,
      "step": 5131
    },
    {
      "entropy": 0.7346804551780224,
      "epoch": 0.4721766982363529,
      "grad_norm": 0.15519508719444275,
      "learning_rate": 8.426411506731685e-05,
      "loss": 0.7016,
      "mean_token_accuracy": 0.7963188141584396,
      "num_tokens": 163804215.0,
      "step": 5132
    },
    {
      "entropy": 0.752505823969841,
      "epoch": 0.4722687046077941,
      "grad_norm": 0.15187543630599976,
      "learning_rate": 8.426104824117521e-05,
      "loss": 0.7164,
      "mean_token_accuracy": 0.7903633005917072,
      "num_tokens": 163836458.0,
      "step": 5133
    },
    {
      "entropy": 0.8531786240637302,
      "epoch": 0.4723607109792353,
      "grad_norm": 0.15973873436450958,
      "learning_rate": 8.425798141503358e-05,
      "loss": 0.8335,
      "mean_token_accuracy": 0.7569948881864548,
      "num_tokens": 163868478.0,
      "step": 5134
    },
    {
      "entropy": 0.7369353696703911,
      "epoch": 0.4724527173506765,
      "grad_norm": 0.16890230774879456,
      "learning_rate": 8.425491458889196e-05,
      "loss": 0.7485,
      "mean_token_accuracy": 0.7888448648154736,
      "num_tokens": 163900283.0,
      "step": 5135
    },
    {
      "entropy": 0.7885284163057804,
      "epoch": 0.47254472372211775,
      "grad_norm": 0.1591619998216629,
      "learning_rate": 8.425184776275033e-05,
      "loss": 0.7708,
      "mean_token_accuracy": 0.7784798629581928,
      "num_tokens": 163932416.0,
      "step": 5136
    },
    {
      "entropy": 0.7777881845831871,
      "epoch": 0.472636730093559,
      "grad_norm": 0.16280797123908997,
      "learning_rate": 8.424878093660871e-05,
      "loss": 0.765,
      "mean_token_accuracy": 0.7737102098762989,
      "num_tokens": 163963955.0,
      "step": 5137
    },
    {
      "entropy": 0.7714197374880314,
      "epoch": 0.4727287364650002,
      "grad_norm": 0.15690013766288757,
      "learning_rate": 8.424571411046708e-05,
      "loss": 0.7759,
      "mean_token_accuracy": 0.7761208973824978,
      "num_tokens": 163995596.0,
      "step": 5138
    },
    {
      "entropy": 0.8697222508490086,
      "epoch": 0.47282074283644143,
      "grad_norm": 0.1625160574913025,
      "learning_rate": 8.424264728432546e-05,
      "loss": 0.8741,
      "mean_token_accuracy": 0.7462718822062016,
      "num_tokens": 164027551.0,
      "step": 5139
    },
    {
      "entropy": 0.7837327998131514,
      "epoch": 0.4729127492078826,
      "grad_norm": 0.15998995304107666,
      "learning_rate": 8.423958045818383e-05,
      "loss": 0.7864,
      "mean_token_accuracy": 0.7742840237915516,
      "num_tokens": 164059388.0,
      "step": 5140
    },
    {
      "entropy": 0.8262041844427586,
      "epoch": 0.47300475557932387,
      "grad_norm": 0.16394217312335968,
      "learning_rate": 8.42365136320422e-05,
      "loss": 0.8199,
      "mean_token_accuracy": 0.7604608424007893,
      "num_tokens": 164091282.0,
      "step": 5141
    },
    {
      "entropy": 0.7050326149910688,
      "epoch": 0.4730967619507651,
      "grad_norm": 0.15041328966617584,
      "learning_rate": 8.423344680590058e-05,
      "loss": 0.6881,
      "mean_token_accuracy": 0.7935977838933468,
      "num_tokens": 164123179.0,
      "step": 5142
    },
    {
      "entropy": 0.9351494610309601,
      "epoch": 0.4731887683222063,
      "grad_norm": 0.16334101557731628,
      "learning_rate": 8.423037997975896e-05,
      "loss": 0.9369,
      "mean_token_accuracy": 0.7336917817592621,
      "num_tokens": 164155615.0,
      "step": 5143
    },
    {
      "entropy": 0.8163490332663059,
      "epoch": 0.47328077469364754,
      "grad_norm": 0.1537867933511734,
      "learning_rate": 8.422731315361733e-05,
      "loss": 0.7997,
      "mean_token_accuracy": 0.767871793359518,
      "num_tokens": 164188118.0,
      "step": 5144
    },
    {
      "entropy": 0.8433966524899006,
      "epoch": 0.47337278106508873,
      "grad_norm": 0.1513570249080658,
      "learning_rate": 8.42242463274757e-05,
      "loss": 0.8289,
      "mean_token_accuracy": 0.75904655829072,
      "num_tokens": 164220567.0,
      "step": 5145
    },
    {
      "entropy": 0.816118648275733,
      "epoch": 0.47346478743653,
      "grad_norm": 0.1676976978778839,
      "learning_rate": 8.422117950133406e-05,
      "loss": 0.7917,
      "mean_token_accuracy": 0.7737191915512085,
      "num_tokens": 164252344.0,
      "step": 5146
    },
    {
      "entropy": 0.7849381770938635,
      "epoch": 0.4735567938079712,
      "grad_norm": 0.15803959965705872,
      "learning_rate": 8.421811267519244e-05,
      "loss": 0.7577,
      "mean_token_accuracy": 0.7767266444861889,
      "num_tokens": 164284217.0,
      "step": 5147
    },
    {
      "entropy": 0.8025322929024696,
      "epoch": 0.4736488001794124,
      "grad_norm": 0.16279709339141846,
      "learning_rate": 8.421504584905083e-05,
      "loss": 0.8113,
      "mean_token_accuracy": 0.764786284416914,
      "num_tokens": 164316485.0,
      "step": 5148
    },
    {
      "entropy": 0.790952954441309,
      "epoch": 0.47374080655085365,
      "grad_norm": 0.16059556603431702,
      "learning_rate": 8.42119790229092e-05,
      "loss": 0.7698,
      "mean_token_accuracy": 0.7759455442428589,
      "num_tokens": 164348314.0,
      "step": 5149
    },
    {
      "entropy": 0.8047340009361506,
      "epoch": 0.47383281292229484,
      "grad_norm": 0.16464723646640778,
      "learning_rate": 8.420891219676756e-05,
      "loss": 0.8085,
      "mean_token_accuracy": 0.7652494572103024,
      "num_tokens": 164380184.0,
      "step": 5150
    },
    {
      "entropy": 0.8413729015737772,
      "epoch": 0.4739248192937361,
      "grad_norm": 0.16501320898532867,
      "learning_rate": 8.420584537062594e-05,
      "loss": 0.8473,
      "mean_token_accuracy": 0.7566986717283726,
      "num_tokens": 164412206.0,
      "step": 5151
    },
    {
      "entropy": 0.7286456264555454,
      "epoch": 0.47401682566517733,
      "grad_norm": 0.16662999987602234,
      "learning_rate": 8.420277854448431e-05,
      "loss": 0.736,
      "mean_token_accuracy": 0.7855833992362022,
      "num_tokens": 164443915.0,
      "step": 5152
    },
    {
      "entropy": 0.9453593324869871,
      "epoch": 0.4741088320366185,
      "grad_norm": 0.16937294602394104,
      "learning_rate": 8.419971171834269e-05,
      "loss": 0.9405,
      "mean_token_accuracy": 0.7288645915687084,
      "num_tokens": 164475870.0,
      "step": 5153
    },
    {
      "entropy": 0.8248483445495367,
      "epoch": 0.47420083840805977,
      "grad_norm": 0.16152800619602203,
      "learning_rate": 8.419664489220106e-05,
      "loss": 0.8188,
      "mean_token_accuracy": 0.7640186175704002,
      "num_tokens": 164507785.0,
      "step": 5154
    },
    {
      "entropy": 0.7795240730047226,
      "epoch": 0.47429284477950095,
      "grad_norm": 0.1596006155014038,
      "learning_rate": 8.419357806605944e-05,
      "loss": 0.78,
      "mean_token_accuracy": 0.770627211779356,
      "num_tokens": 164540375.0,
      "step": 5155
    },
    {
      "entropy": 0.7347411755472422,
      "epoch": 0.4743848511509422,
      "grad_norm": 0.15406954288482666,
      "learning_rate": 8.419051123991781e-05,
      "loss": 0.7078,
      "mean_token_accuracy": 0.7831668332219124,
      "num_tokens": 164572417.0,
      "step": 5156
    },
    {
      "entropy": 0.87384332716465,
      "epoch": 0.47447685752238344,
      "grad_norm": 0.16053293645381927,
      "learning_rate": 8.418744441377618e-05,
      "loss": 0.8343,
      "mean_token_accuracy": 0.7554101869463921,
      "num_tokens": 164603964.0,
      "step": 5157
    },
    {
      "entropy": 0.8271229807287455,
      "epoch": 0.47456886389382463,
      "grad_norm": 0.15651504695415497,
      "learning_rate": 8.418437758763457e-05,
      "loss": 0.8172,
      "mean_token_accuracy": 0.7621587514877319,
      "num_tokens": 164636158.0,
      "step": 5158
    },
    {
      "entropy": 0.8567314557731152,
      "epoch": 0.4746608702652659,
      "grad_norm": 0.1760365068912506,
      "learning_rate": 8.418131076149294e-05,
      "loss": 0.8774,
      "mean_token_accuracy": 0.7500869818031788,
      "num_tokens": 164666681.0,
      "step": 5159
    },
    {
      "entropy": 0.9005201607942581,
      "epoch": 0.47475287663670707,
      "grad_norm": 0.17269085347652435,
      "learning_rate": 8.417824393535131e-05,
      "loss": 0.931,
      "mean_token_accuracy": 0.736955925822258,
      "num_tokens": 164699120.0,
      "step": 5160
    },
    {
      "entropy": 0.9000576436519623,
      "epoch": 0.4748448830081483,
      "grad_norm": 0.1688125729560852,
      "learning_rate": 8.417517710920968e-05,
      "loss": 0.9179,
      "mean_token_accuracy": 0.7423493638634682,
      "num_tokens": 164731122.0,
      "step": 5161
    },
    {
      "entropy": 0.7612063251435757,
      "epoch": 0.47493688937958956,
      "grad_norm": 0.14985805749893188,
      "learning_rate": 8.417211028306806e-05,
      "loss": 0.7385,
      "mean_token_accuracy": 0.7804300673305988,
      "num_tokens": 164762349.0,
      "step": 5162
    },
    {
      "entropy": 0.7814712841063738,
      "epoch": 0.47502889575103074,
      "grad_norm": 0.15255466103553772,
      "learning_rate": 8.416904345692644e-05,
      "loss": 0.7382,
      "mean_token_accuracy": 0.7827499955892563,
      "num_tokens": 164794122.0,
      "step": 5163
    },
    {
      "entropy": 0.8280153144150972,
      "epoch": 0.475120902122472,
      "grad_norm": 0.15799066424369812,
      "learning_rate": 8.416597663078481e-05,
      "loss": 0.8087,
      "mean_token_accuracy": 0.7636596709489822,
      "num_tokens": 164826776.0,
      "step": 5164
    },
    {
      "entropy": 0.7485704440623522,
      "epoch": 0.4752129084939132,
      "grad_norm": 0.15493471920490265,
      "learning_rate": 8.416290980464317e-05,
      "loss": 0.7336,
      "mean_token_accuracy": 0.7821888141334057,
      "num_tokens": 164858537.0,
      "step": 5165
    },
    {
      "entropy": 0.7316532451659441,
      "epoch": 0.4753049148653544,
      "grad_norm": 0.14977332949638367,
      "learning_rate": 8.415984297850156e-05,
      "loss": 0.711,
      "mean_token_accuracy": 0.7837161011993885,
      "num_tokens": 164890280.0,
      "step": 5166
    },
    {
      "entropy": 0.7838761191815138,
      "epoch": 0.47539692123679567,
      "grad_norm": 0.16312702000141144,
      "learning_rate": 8.415677615235992e-05,
      "loss": 0.769,
      "mean_token_accuracy": 0.7727215401828289,
      "num_tokens": 164921789.0,
      "step": 5167
    },
    {
      "entropy": 0.8170854616910219,
      "epoch": 0.47548892760823686,
      "grad_norm": 0.16045400500297546,
      "learning_rate": 8.41537093262183e-05,
      "loss": 0.8076,
      "mean_token_accuracy": 0.7662179842591286,
      "num_tokens": 164954417.0,
      "step": 5168
    },
    {
      "entropy": 0.8512692395597696,
      "epoch": 0.4755809339796781,
      "grad_norm": 0.1627604216337204,
      "learning_rate": 8.415064250007667e-05,
      "loss": 0.8605,
      "mean_token_accuracy": 0.7481606043875217,
      "num_tokens": 164985785.0,
      "step": 5169
    },
    {
      "entropy": 0.7794848717749119,
      "epoch": 0.4756729403511193,
      "grad_norm": 0.1683390885591507,
      "learning_rate": 8.414757567393505e-05,
      "loss": 0.8012,
      "mean_token_accuracy": 0.7686905078589916,
      "num_tokens": 165017112.0,
      "step": 5170
    },
    {
      "entropy": 0.7092011142522097,
      "epoch": 0.47576494672256053,
      "grad_norm": 0.15805582702159882,
      "learning_rate": 8.414450884779342e-05,
      "loss": 0.7096,
      "mean_token_accuracy": 0.7879418805241585,
      "num_tokens": 165048617.0,
      "step": 5171
    },
    {
      "entropy": 0.6161089949309826,
      "epoch": 0.4758569530940018,
      "grad_norm": 0.15880484879016876,
      "learning_rate": 8.414144202165179e-05,
      "loss": 0.6005,
      "mean_token_accuracy": 0.8181555978953838,
      "num_tokens": 165080692.0,
      "step": 5172
    },
    {
      "entropy": 0.8725826628506184,
      "epoch": 0.47594895946544297,
      "grad_norm": 0.16248266398906708,
      "learning_rate": 8.413837519551017e-05,
      "loss": 0.8551,
      "mean_token_accuracy": 0.7498167902231216,
      "num_tokens": 165112123.0,
      "step": 5173
    },
    {
      "entropy": 0.8787448070943356,
      "epoch": 0.4760409658368842,
      "grad_norm": 0.1656971573829651,
      "learning_rate": 8.413530836936855e-05,
      "loss": 0.8819,
      "mean_token_accuracy": 0.7467323131859303,
      "num_tokens": 165144044.0,
      "step": 5174
    },
    {
      "entropy": 0.7597235608845949,
      "epoch": 0.4761329722083254,
      "grad_norm": 0.1580447405576706,
      "learning_rate": 8.413224154322692e-05,
      "loss": 0.737,
      "mean_token_accuracy": 0.7860735394060612,
      "num_tokens": 165176396.0,
      "step": 5175
    },
    {
      "entropy": 0.6933110356330872,
      "epoch": 0.47622497857976664,
      "grad_norm": 0.1485605090856552,
      "learning_rate": 8.412917471708529e-05,
      "loss": 0.6815,
      "mean_token_accuracy": 0.7977470122277737,
      "num_tokens": 165208927.0,
      "step": 5176
    },
    {
      "entropy": 0.7888491787016392,
      "epoch": 0.4763169849512079,
      "grad_norm": 0.16265161335468292,
      "learning_rate": 8.412610789094366e-05,
      "loss": 0.7978,
      "mean_token_accuracy": 0.7667700685560703,
      "num_tokens": 165240151.0,
      "step": 5177
    },
    {
      "entropy": 0.7166433166712523,
      "epoch": 0.4764089913226491,
      "grad_norm": 0.1507086157798767,
      "learning_rate": 8.412304106480204e-05,
      "loss": 0.699,
      "mean_token_accuracy": 0.7940157279372215,
      "num_tokens": 165272499.0,
      "step": 5178
    },
    {
      "entropy": 0.8007474131882191,
      "epoch": 0.4765009976940903,
      "grad_norm": 0.1685057282447815,
      "learning_rate": 8.411997423866042e-05,
      "loss": 0.8105,
      "mean_token_accuracy": 0.7624577544629574,
      "num_tokens": 165304699.0,
      "step": 5179
    },
    {
      "entropy": 0.7361320555210114,
      "epoch": 0.4765930040655315,
      "grad_norm": 0.14657635986804962,
      "learning_rate": 8.411690741251879e-05,
      "loss": 0.7088,
      "mean_token_accuracy": 0.7892447263002396,
      "num_tokens": 165336992.0,
      "step": 5180
    },
    {
      "entropy": 0.8062324021011591,
      "epoch": 0.47668501043697276,
      "grad_norm": 0.15435662865638733,
      "learning_rate": 8.411384058637716e-05,
      "loss": 0.7754,
      "mean_token_accuracy": 0.7721446268260479,
      "num_tokens": 165369309.0,
      "step": 5181
    },
    {
      "entropy": 0.913729228079319,
      "epoch": 0.476777016808414,
      "grad_norm": 0.1641283482313156,
      "learning_rate": 8.411077376023554e-05,
      "loss": 0.9132,
      "mean_token_accuracy": 0.7391794621944427,
      "num_tokens": 165401452.0,
      "step": 5182
    },
    {
      "entropy": 0.7880869358778,
      "epoch": 0.4768690231798552,
      "grad_norm": 0.1663503348827362,
      "learning_rate": 8.41077069340939e-05,
      "loss": 0.7723,
      "mean_token_accuracy": 0.7792351879179478,
      "num_tokens": 165434085.0,
      "step": 5183
    },
    {
      "entropy": 0.695388738065958,
      "epoch": 0.47696102955129643,
      "grad_norm": 0.1592782884836197,
      "learning_rate": 8.410464010795229e-05,
      "loss": 0.6823,
      "mean_token_accuracy": 0.8005071766674519,
      "num_tokens": 165466147.0,
      "step": 5184
    },
    {
      "entropy": 0.9131723418831825,
      "epoch": 0.4770530359227376,
      "grad_norm": 0.16474930942058563,
      "learning_rate": 8.410157328181065e-05,
      "loss": 0.8971,
      "mean_token_accuracy": 0.7469351775944233,
      "num_tokens": 165497192.0,
      "step": 5185
    },
    {
      "entropy": 0.6647975463420153,
      "epoch": 0.47714504229417887,
      "grad_norm": 0.14652462303638458,
      "learning_rate": 8.409850645566904e-05,
      "loss": 0.6603,
      "mean_token_accuracy": 0.8080727905035019,
      "num_tokens": 165529960.0,
      "step": 5186
    },
    {
      "entropy": 0.6845179554075003,
      "epoch": 0.4772370486656201,
      "grad_norm": 0.16309674084186554,
      "learning_rate": 8.40954396295274e-05,
      "loss": 0.6854,
      "mean_token_accuracy": 0.7983250766992569,
      "num_tokens": 165561856.0,
      "step": 5187
    },
    {
      "entropy": 0.7287795413285494,
      "epoch": 0.4773290550370613,
      "grad_norm": 0.15810081362724304,
      "learning_rate": 8.409237280338577e-05,
      "loss": 0.729,
      "mean_token_accuracy": 0.7857123129069805,
      "num_tokens": 165594040.0,
      "step": 5188
    },
    {
      "entropy": 0.7643451783806086,
      "epoch": 0.47742106140850255,
      "grad_norm": 0.15535081923007965,
      "learning_rate": 8.408930597724415e-05,
      "loss": 0.7692,
      "mean_token_accuracy": 0.7779422849416733,
      "num_tokens": 165625377.0,
      "step": 5189
    },
    {
      "entropy": 0.6599569823592901,
      "epoch": 0.47751306777994373,
      "grad_norm": 0.16118305921554565,
      "learning_rate": 8.408623915110253e-05,
      "loss": 0.6569,
      "mean_token_accuracy": 0.8063151985406876,
      "num_tokens": 165656922.0,
      "step": 5190
    },
    {
      "entropy": 0.7884930968284607,
      "epoch": 0.477605074151385,
      "grad_norm": 0.16401562094688416,
      "learning_rate": 8.40831723249609e-05,
      "loss": 0.7956,
      "mean_token_accuracy": 0.767137348651886,
      "num_tokens": 165688551.0,
      "step": 5191
    },
    {
      "entropy": 0.7846729643642902,
      "epoch": 0.4776970805228262,
      "grad_norm": 0.16285641491413116,
      "learning_rate": 8.408010549881927e-05,
      "loss": 0.7784,
      "mean_token_accuracy": 0.772902961820364,
      "num_tokens": 165720557.0,
      "step": 5192
    },
    {
      "entropy": 0.6552283130586147,
      "epoch": 0.4777890868942674,
      "grad_norm": 0.15023231506347656,
      "learning_rate": 8.407703867267765e-05,
      "loss": 0.63,
      "mean_token_accuracy": 0.8102524429559708,
      "num_tokens": 165752918.0,
      "step": 5193
    },
    {
      "entropy": 0.6953130988404155,
      "epoch": 0.47788109326570866,
      "grad_norm": 0.1531752198934555,
      "learning_rate": 8.407397184653603e-05,
      "loss": 0.6867,
      "mean_token_accuracy": 0.801024317741394,
      "num_tokens": 165785393.0,
      "step": 5194
    },
    {
      "entropy": 0.7560567799955606,
      "epoch": 0.47797309963714985,
      "grad_norm": 0.15910053253173828,
      "learning_rate": 8.40709050203944e-05,
      "loss": 0.7492,
      "mean_token_accuracy": 0.7873763404786587,
      "num_tokens": 165817982.0,
      "step": 5195
    },
    {
      "entropy": 0.7561619020998478,
      "epoch": 0.4780651060085911,
      "grad_norm": 0.1577204316854477,
      "learning_rate": 8.406783819425277e-05,
      "loss": 0.7379,
      "mean_token_accuracy": 0.7841232307255268,
      "num_tokens": 165849894.0,
      "step": 5196
    },
    {
      "entropy": 0.7823984920978546,
      "epoch": 0.47815711238003233,
      "grad_norm": 0.15557356178760529,
      "learning_rate": 8.406477136811115e-05,
      "loss": 0.7719,
      "mean_token_accuracy": 0.7741900645196438,
      "num_tokens": 165882237.0,
      "step": 5197
    },
    {
      "entropy": 0.8886288665235043,
      "epoch": 0.4782491187514735,
      "grad_norm": 0.1773010492324829,
      "learning_rate": 8.406170454196952e-05,
      "loss": 0.8826,
      "mean_token_accuracy": 0.7463788278400898,
      "num_tokens": 165912968.0,
      "step": 5198
    },
    {
      "entropy": 0.7923057079315186,
      "epoch": 0.47834112512291477,
      "grad_norm": 0.16153579950332642,
      "learning_rate": 8.40586377158279e-05,
      "loss": 0.7861,
      "mean_token_accuracy": 0.7734267003834248,
      "num_tokens": 165945630.0,
      "step": 5199
    },
    {
      "entropy": 0.7887945510447025,
      "epoch": 0.47843313149435596,
      "grad_norm": 0.15586668252944946,
      "learning_rate": 8.405557088968627e-05,
      "loss": 0.7876,
      "mean_token_accuracy": 0.7709155417978764,
      "num_tokens": 165977640.0,
      "step": 5200
    },
    {
      "entropy": 0.7490629069507122,
      "epoch": 0.4785251378657972,
      "grad_norm": 0.1520090401172638,
      "learning_rate": 8.405250406354465e-05,
      "loss": 0.7393,
      "mean_token_accuracy": 0.7813426777720451,
      "num_tokens": 166010177.0,
      "step": 5201
    },
    {
      "entropy": 0.7798090595752001,
      "epoch": 0.47861714423723845,
      "grad_norm": 0.16572289168834686,
      "learning_rate": 8.404943723740302e-05,
      "loss": 0.7682,
      "mean_token_accuracy": 0.7776888832449913,
      "num_tokens": 166042198.0,
      "step": 5202
    },
    {
      "entropy": 0.839749488979578,
      "epoch": 0.47870915060867963,
      "grad_norm": 0.15713703632354736,
      "learning_rate": 8.404637041126138e-05,
      "loss": 0.8375,
      "mean_token_accuracy": 0.7568858079612255,
      "num_tokens": 166074480.0,
      "step": 5203
    },
    {
      "entropy": 0.7719357125461102,
      "epoch": 0.4788011569801209,
      "grad_norm": 0.15542197227478027,
      "learning_rate": 8.404330358511977e-05,
      "loss": 0.7629,
      "mean_token_accuracy": 0.7743195071816444,
      "num_tokens": 166107004.0,
      "step": 5204
    },
    {
      "entropy": 0.7482200749218464,
      "epoch": 0.4788931633515621,
      "grad_norm": 0.1548205018043518,
      "learning_rate": 8.404023675897815e-05,
      "loss": 0.7288,
      "mean_token_accuracy": 0.7803924307227135,
      "num_tokens": 166138410.0,
      "step": 5205
    },
    {
      "entropy": 0.8423316348344088,
      "epoch": 0.4789851697230033,
      "grad_norm": 0.1668292135000229,
      "learning_rate": 8.403716993283651e-05,
      "loss": 0.8487,
      "mean_token_accuracy": 0.7562650851905346,
      "num_tokens": 166170662.0,
      "step": 5206
    },
    {
      "entropy": 0.710219407454133,
      "epoch": 0.47907717609444456,
      "grad_norm": 0.155366450548172,
      "learning_rate": 8.403410310669488e-05,
      "loss": 0.6796,
      "mean_token_accuracy": 0.7988442592322826,
      "num_tokens": 166202374.0,
      "step": 5207
    },
    {
      "entropy": 0.8125598300248384,
      "epoch": 0.47916918246588575,
      "grad_norm": 0.16128143668174744,
      "learning_rate": 8.403103628055325e-05,
      "loss": 0.8089,
      "mean_token_accuracy": 0.7672162428498268,
      "num_tokens": 166235049.0,
      "step": 5208
    },
    {
      "entropy": 0.7864994071424007,
      "epoch": 0.479261188837327,
      "grad_norm": 0.16857819259166718,
      "learning_rate": 8.402796945441163e-05,
      "loss": 0.7942,
      "mean_token_accuracy": 0.7720288299024105,
      "num_tokens": 166267305.0,
      "step": 5209
    },
    {
      "entropy": 0.7633724268525839,
      "epoch": 0.47935319520876823,
      "grad_norm": 0.1657439023256302,
      "learning_rate": 8.402490262827001e-05,
      "loss": 0.7332,
      "mean_token_accuracy": 0.7903573885560036,
      "num_tokens": 166298078.0,
      "step": 5210
    },
    {
      "entropy": 0.9285253621637821,
      "epoch": 0.4794452015802094,
      "grad_norm": 0.16453133523464203,
      "learning_rate": 8.402183580212838e-05,
      "loss": 0.9218,
      "mean_token_accuracy": 0.7368473298847675,
      "num_tokens": 166330079.0,
      "step": 5211
    },
    {
      "entropy": 0.7992825135588646,
      "epoch": 0.47953720795165067,
      "grad_norm": 0.16103699803352356,
      "learning_rate": 8.401876897598675e-05,
      "loss": 0.7924,
      "mean_token_accuracy": 0.7685747854411602,
      "num_tokens": 166362060.0,
      "step": 5212
    },
    {
      "entropy": 0.8144254144281149,
      "epoch": 0.47962921432309186,
      "grad_norm": 0.16080686450004578,
      "learning_rate": 8.401570214984513e-05,
      "loss": 0.8015,
      "mean_token_accuracy": 0.7676109671592712,
      "num_tokens": 166394026.0,
      "step": 5213
    },
    {
      "entropy": 0.7308588568121195,
      "epoch": 0.4797212206945331,
      "grad_norm": 0.15397866070270538,
      "learning_rate": 8.40126353237035e-05,
      "loss": 0.7155,
      "mean_token_accuracy": 0.7919535934925079,
      "num_tokens": 166425351.0,
      "step": 5214
    },
    {
      "entropy": 0.9178040865808725,
      "epoch": 0.47981322706597435,
      "grad_norm": 0.16059650480747223,
      "learning_rate": 8.400956849756188e-05,
      "loss": 0.9032,
      "mean_token_accuracy": 0.7396084330976009,
      "num_tokens": 166457118.0,
      "step": 5215
    },
    {
      "entropy": 0.8147932030260563,
      "epoch": 0.47990523343741553,
      "grad_norm": 0.15835446119308472,
      "learning_rate": 8.400650167142025e-05,
      "loss": 0.7976,
      "mean_token_accuracy": 0.7692598812282085,
      "num_tokens": 166488890.0,
      "step": 5216
    },
    {
      "entropy": 0.8789936006069183,
      "epoch": 0.4799972398088568,
      "grad_norm": 0.15857011079788208,
      "learning_rate": 8.400343484527863e-05,
      "loss": 0.8748,
      "mean_token_accuracy": 0.7473781295120716,
      "num_tokens": 166521038.0,
      "step": 5217
    },
    {
      "entropy": 0.8335211500525475,
      "epoch": 0.48008924618029797,
      "grad_norm": 0.17185120284557343,
      "learning_rate": 8.4000368019137e-05,
      "loss": 0.8564,
      "mean_token_accuracy": 0.7517082057893276,
      "num_tokens": 166552536.0,
      "step": 5218
    },
    {
      "entropy": 0.7287058681249619,
      "epoch": 0.4801812525517392,
      "grad_norm": 0.15316429734230042,
      "learning_rate": 8.399730119299537e-05,
      "loss": 0.7249,
      "mean_token_accuracy": 0.7885411269962788,
      "num_tokens": 166585095.0,
      "step": 5219
    },
    {
      "entropy": 0.9605265371501446,
      "epoch": 0.48027325892318046,
      "grad_norm": 0.15757620334625244,
      "learning_rate": 8.399423436685375e-05,
      "loss": 0.9518,
      "mean_token_accuracy": 0.7258778996765614,
      "num_tokens": 166617120.0,
      "step": 5220
    },
    {
      "entropy": 0.7738384678959846,
      "epoch": 0.48036526529462165,
      "grad_norm": 0.15380239486694336,
      "learning_rate": 8.399116754071213e-05,
      "loss": 0.758,
      "mean_token_accuracy": 0.7777248993515968,
      "num_tokens": 166648741.0,
      "step": 5221
    },
    {
      "entropy": 0.8127052206546068,
      "epoch": 0.4804572716660629,
      "grad_norm": 0.1570986956357956,
      "learning_rate": 8.39881007145705e-05,
      "loss": 0.8074,
      "mean_token_accuracy": 0.7641115076839924,
      "num_tokens": 166680630.0,
      "step": 5222
    },
    {
      "entropy": 0.6785797066986561,
      "epoch": 0.4805492780375041,
      "grad_norm": 0.15366777777671814,
      "learning_rate": 8.398503388842886e-05,
      "loss": 0.6776,
      "mean_token_accuracy": 0.7962942346930504,
      "num_tokens": 166712528.0,
      "step": 5223
    },
    {
      "entropy": 0.7533758170902729,
      "epoch": 0.4806412844089453,
      "grad_norm": 0.16717413067817688,
      "learning_rate": 8.398196706228725e-05,
      "loss": 0.7648,
      "mean_token_accuracy": 0.7762302383780479,
      "num_tokens": 166744152.0,
      "step": 5224
    },
    {
      "entropy": 0.7703113965690136,
      "epoch": 0.48073329078038657,
      "grad_norm": 0.15529121458530426,
      "learning_rate": 8.397890023614561e-05,
      "loss": 0.7567,
      "mean_token_accuracy": 0.7770662866532803,
      "num_tokens": 166775670.0,
      "step": 5225
    },
    {
      "entropy": 0.817438617348671,
      "epoch": 0.48082529715182776,
      "grad_norm": 0.15737801790237427,
      "learning_rate": 8.3975833410004e-05,
      "loss": 0.8121,
      "mean_token_accuracy": 0.7636530660092831,
      "num_tokens": 166807412.0,
      "step": 5226
    },
    {
      "entropy": 0.6860779542475939,
      "epoch": 0.480917303523269,
      "grad_norm": 0.1466493308544159,
      "learning_rate": 8.397276658386236e-05,
      "loss": 0.6548,
      "mean_token_accuracy": 0.8061686009168625,
      "num_tokens": 166839038.0,
      "step": 5227
    },
    {
      "entropy": 0.7151614408940077,
      "epoch": 0.4810093098947102,
      "grad_norm": 0.15497565269470215,
      "learning_rate": 8.396969975772074e-05,
      "loss": 0.7057,
      "mean_token_accuracy": 0.790836114436388,
      "num_tokens": 166870247.0,
      "step": 5228
    },
    {
      "entropy": 0.7009749989956617,
      "epoch": 0.48110131626615144,
      "grad_norm": 0.15243779122829437,
      "learning_rate": 8.396663293157911e-05,
      "loss": 0.6899,
      "mean_token_accuracy": 0.794426042586565,
      "num_tokens": 166901586.0,
      "step": 5229
    },
    {
      "entropy": 0.7841147650033236,
      "epoch": 0.4811933226375927,
      "grad_norm": 0.15437094867229462,
      "learning_rate": 8.396356610543748e-05,
      "loss": 0.7518,
      "mean_token_accuracy": 0.7754592709243298,
      "num_tokens": 166933117.0,
      "step": 5230
    },
    {
      "entropy": 0.6418750379234552,
      "epoch": 0.48128532900903387,
      "grad_norm": 0.17460261285305023,
      "learning_rate": 8.396049927929586e-05,
      "loss": 0.6511,
      "mean_token_accuracy": 0.8084491044282913,
      "num_tokens": 166965500.0,
      "step": 5231
    },
    {
      "entropy": 0.9004889652132988,
      "epoch": 0.4813773353804751,
      "grad_norm": 0.15922404825687408,
      "learning_rate": 8.395743245315424e-05,
      "loss": 0.8837,
      "mean_token_accuracy": 0.7456017471849918,
      "num_tokens": 166997618.0,
      "step": 5232
    },
    {
      "entropy": 0.8107548011466861,
      "epoch": 0.4814693417519163,
      "grad_norm": 0.1590633988380432,
      "learning_rate": 8.395436562701261e-05,
      "loss": 0.8021,
      "mean_token_accuracy": 0.7634168341755867,
      "num_tokens": 167028302.0,
      "step": 5233
    },
    {
      "entropy": 0.7335665635764599,
      "epoch": 0.48156134812335755,
      "grad_norm": 0.15417975187301636,
      "learning_rate": 8.395129880087098e-05,
      "loss": 0.7106,
      "mean_token_accuracy": 0.7954946681857109,
      "num_tokens": 167060181.0,
      "step": 5234
    },
    {
      "entropy": 0.7492597457021475,
      "epoch": 0.4816533544947988,
      "grad_norm": 0.1553354263305664,
      "learning_rate": 8.394823197472935e-05,
      "loss": 0.753,
      "mean_token_accuracy": 0.783217042684555,
      "num_tokens": 167092366.0,
      "step": 5235
    },
    {
      "entropy": 0.8121619448065758,
      "epoch": 0.48174536086624,
      "grad_norm": 0.16081497073173523,
      "learning_rate": 8.394516514858774e-05,
      "loss": 0.8029,
      "mean_token_accuracy": 0.7669978179037571,
      "num_tokens": 167123839.0,
      "step": 5236
    },
    {
      "entropy": 0.787126375362277,
      "epoch": 0.4818373672376812,
      "grad_norm": 0.15852895379066467,
      "learning_rate": 8.394209832244611e-05,
      "loss": 0.7806,
      "mean_token_accuracy": 0.782749842852354,
      "num_tokens": 167155235.0,
      "step": 5237
    },
    {
      "entropy": 0.7457259241491556,
      "epoch": 0.4819293736091224,
      "grad_norm": 0.15617115795612335,
      "learning_rate": 8.393903149630448e-05,
      "loss": 0.7375,
      "mean_token_accuracy": 0.785408541560173,
      "num_tokens": 167187633.0,
      "step": 5238
    },
    {
      "entropy": 0.8020030446350574,
      "epoch": 0.48202137998056366,
      "grad_norm": 0.15182462334632874,
      "learning_rate": 8.393596467016284e-05,
      "loss": 0.786,
      "mean_token_accuracy": 0.7641374357044697,
      "num_tokens": 167219599.0,
      "step": 5239
    },
    {
      "entropy": 0.7289098799228668,
      "epoch": 0.4821133863520049,
      "grad_norm": 0.1607024371623993,
      "learning_rate": 8.393289784402123e-05,
      "loss": 0.727,
      "mean_token_accuracy": 0.7881591133773327,
      "num_tokens": 167251574.0,
      "step": 5240
    },
    {
      "entropy": 0.8949531614780426,
      "epoch": 0.4822053927234461,
      "grad_norm": 0.15943796932697296,
      "learning_rate": 8.392983101787961e-05,
      "loss": 0.8774,
      "mean_token_accuracy": 0.7440531328320503,
      "num_tokens": 167283456.0,
      "step": 5241
    },
    {
      "entropy": 0.7811092622578144,
      "epoch": 0.48229739909488734,
      "grad_norm": 0.15647752583026886,
      "learning_rate": 8.392676419173798e-05,
      "loss": 0.7803,
      "mean_token_accuracy": 0.771081730723381,
      "num_tokens": 167315841.0,
      "step": 5242
    },
    {
      "entropy": 0.5919939707964659,
      "epoch": 0.4823894054663285,
      "grad_norm": 0.14797861874103546,
      "learning_rate": 8.392369736559634e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8275830373167992,
      "num_tokens": 167348609.0,
      "step": 5243
    },
    {
      "entropy": 0.789119279012084,
      "epoch": 0.48248141183776977,
      "grad_norm": 0.16360007226467133,
      "learning_rate": 8.392063053945472e-05,
      "loss": 0.8024,
      "mean_token_accuracy": 0.7670775763690472,
      "num_tokens": 167379873.0,
      "step": 5244
    },
    {
      "entropy": 0.7908155731856823,
      "epoch": 0.482573418209211,
      "grad_norm": 0.16016699373722076,
      "learning_rate": 8.391756371331309e-05,
      "loss": 0.7785,
      "mean_token_accuracy": 0.7764917761087418,
      "num_tokens": 167411716.0,
      "step": 5245
    },
    {
      "entropy": 0.8874424397945404,
      "epoch": 0.4826654245806522,
      "grad_norm": 0.1659286767244339,
      "learning_rate": 8.391449688717147e-05,
      "loss": 0.8852,
      "mean_token_accuracy": 0.7436511591076851,
      "num_tokens": 167443427.0,
      "step": 5246
    },
    {
      "entropy": 0.762246310710907,
      "epoch": 0.48275743095209345,
      "grad_norm": 0.15461896359920502,
      "learning_rate": 8.391143006102984e-05,
      "loss": 0.7634,
      "mean_token_accuracy": 0.7749888896942139,
      "num_tokens": 167474625.0,
      "step": 5247
    },
    {
      "entropy": 0.7282298412173986,
      "epoch": 0.48284943732353464,
      "grad_norm": 0.1693304181098938,
      "learning_rate": 8.390836323488822e-05,
      "loss": 0.7213,
      "mean_token_accuracy": 0.7874235101044178,
      "num_tokens": 167506339.0,
      "step": 5248
    },
    {
      "entropy": 0.8465447872877121,
      "epoch": 0.4829414436949759,
      "grad_norm": 0.17562691867351532,
      "learning_rate": 8.390529640874659e-05,
      "loss": 0.8597,
      "mean_token_accuracy": 0.7518867589533329,
      "num_tokens": 167536572.0,
      "step": 5249
    },
    {
      "entropy": 0.7610751166939735,
      "epoch": 0.4830334500664171,
      "grad_norm": 0.15573570132255554,
      "learning_rate": 8.390222958260496e-05,
      "loss": 0.736,
      "mean_token_accuracy": 0.7869646064937115,
      "num_tokens": 167568978.0,
      "step": 5250
    },
    {
      "entropy": 0.7336054444313049,
      "epoch": 0.4831254564378583,
      "grad_norm": 0.1518692970275879,
      "learning_rate": 8.389916275646334e-05,
      "loss": 0.7272,
      "mean_token_accuracy": 0.7817497849464417,
      "num_tokens": 167600947.0,
      "step": 5251
    },
    {
      "entropy": 0.7490066420286894,
      "epoch": 0.48321746280929956,
      "grad_norm": 0.15595075488090515,
      "learning_rate": 8.389609593032172e-05,
      "loss": 0.7362,
      "mean_token_accuracy": 0.7890356257557869,
      "num_tokens": 167633414.0,
      "step": 5252
    },
    {
      "entropy": 0.7371593900024891,
      "epoch": 0.48330946918074075,
      "grad_norm": 0.1640031337738037,
      "learning_rate": 8.389302910418009e-05,
      "loss": 0.735,
      "mean_token_accuracy": 0.7867949493229389,
      "num_tokens": 167665999.0,
      "step": 5253
    },
    {
      "entropy": 0.6132667511701584,
      "epoch": 0.483401475552182,
      "grad_norm": 0.16526922583580017,
      "learning_rate": 8.388996227803846e-05,
      "loss": 0.6178,
      "mean_token_accuracy": 0.8149363659322262,
      "num_tokens": 167698492.0,
      "step": 5254
    },
    {
      "entropy": 0.72002137824893,
      "epoch": 0.48349348192362324,
      "grad_norm": 0.14845527708530426,
      "learning_rate": 8.388689545189684e-05,
      "loss": 0.6933,
      "mean_token_accuracy": 0.7956993319094181,
      "num_tokens": 167730602.0,
      "step": 5255
    },
    {
      "entropy": 0.7480526939034462,
      "epoch": 0.4835854882950644,
      "grad_norm": 0.1570023000240326,
      "learning_rate": 8.38838286257552e-05,
      "loss": 0.7391,
      "mean_token_accuracy": 0.7815625295042992,
      "num_tokens": 167762852.0,
      "step": 5256
    },
    {
      "entropy": 0.7955609634518623,
      "epoch": 0.48367749466650567,
      "grad_norm": 0.15646754205226898,
      "learning_rate": 8.388076179961359e-05,
      "loss": 0.7705,
      "mean_token_accuracy": 0.7764603570103645,
      "num_tokens": 167795179.0,
      "step": 5257
    },
    {
      "entropy": 0.7876689974218607,
      "epoch": 0.48376950103794686,
      "grad_norm": 0.1620609015226364,
      "learning_rate": 8.387769497347196e-05,
      "loss": 0.7855,
      "mean_token_accuracy": 0.7694186605513096,
      "num_tokens": 167827947.0,
      "step": 5258
    },
    {
      "entropy": 0.7587589137256145,
      "epoch": 0.4838615074093881,
      "grad_norm": 0.1527715027332306,
      "learning_rate": 8.387462814733034e-05,
      "loss": 0.7479,
      "mean_token_accuracy": 0.7806435786187649,
      "num_tokens": 167860193.0,
      "step": 5259
    },
    {
      "entropy": 0.7746089175343513,
      "epoch": 0.48395351378082935,
      "grad_norm": 0.16166149079799652,
      "learning_rate": 8.38715613211887e-05,
      "loss": 0.7757,
      "mean_token_accuracy": 0.7743276469409466,
      "num_tokens": 167891460.0,
      "step": 5260
    },
    {
      "entropy": 0.7353371828794479,
      "epoch": 0.48404552015227054,
      "grad_norm": 0.15472234785556793,
      "learning_rate": 8.386849449504707e-05,
      "loss": 0.7268,
      "mean_token_accuracy": 0.7858091369271278,
      "num_tokens": 167923623.0,
      "step": 5261
    },
    {
      "entropy": 0.8182930871844292,
      "epoch": 0.4841375265237118,
      "grad_norm": 0.16275431215763092,
      "learning_rate": 8.386542766890545e-05,
      "loss": 0.8098,
      "mean_token_accuracy": 0.7654200531542301,
      "num_tokens": 167956233.0,
      "step": 5262
    },
    {
      "entropy": 0.755168816074729,
      "epoch": 0.48422953289515297,
      "grad_norm": 0.16220754384994507,
      "learning_rate": 8.386236084276384e-05,
      "loss": 0.7593,
      "mean_token_accuracy": 0.7823838219046593,
      "num_tokens": 167988197.0,
      "step": 5263
    },
    {
      "entropy": 0.7437452934682369,
      "epoch": 0.4843215392665942,
      "grad_norm": 0.15614238381385803,
      "learning_rate": 8.38592940166222e-05,
      "loss": 0.705,
      "mean_token_accuracy": 0.7942300736904144,
      "num_tokens": 168020836.0,
      "step": 5264
    },
    {
      "entropy": 0.8206591811031103,
      "epoch": 0.48441354563803546,
      "grad_norm": 0.1591477245092392,
      "learning_rate": 8.385622719048057e-05,
      "loss": 0.8047,
      "mean_token_accuracy": 0.7643876299262047,
      "num_tokens": 168052107.0,
      "step": 5265
    },
    {
      "entropy": 0.7948750276118517,
      "epoch": 0.48450555200947665,
      "grad_norm": 0.16878165304660797,
      "learning_rate": 8.385316036433894e-05,
      "loss": 0.7859,
      "mean_token_accuracy": 0.7745554558932781,
      "num_tokens": 168084362.0,
      "step": 5266
    },
    {
      "entropy": 0.8400347884744406,
      "epoch": 0.4845975583809179,
      "grad_norm": 0.15655472874641418,
      "learning_rate": 8.385009353819732e-05,
      "loss": 0.8241,
      "mean_token_accuracy": 0.7621246501803398,
      "num_tokens": 168116833.0,
      "step": 5267
    },
    {
      "entropy": 0.7473954213783145,
      "epoch": 0.4846895647523591,
      "grad_norm": 0.15724386274814606,
      "learning_rate": 8.38470267120557e-05,
      "loss": 0.7258,
      "mean_token_accuracy": 0.7902371622622013,
      "num_tokens": 168148849.0,
      "step": 5268
    },
    {
      "entropy": 0.6220828350633383,
      "epoch": 0.4847815711238003,
      "grad_norm": 0.14897732436656952,
      "learning_rate": 8.384395988591407e-05,
      "loss": 0.6193,
      "mean_token_accuracy": 0.8188070431351662,
      "num_tokens": 168180920.0,
      "step": 5269
    },
    {
      "entropy": 0.8614533469080925,
      "epoch": 0.48487357749524157,
      "grad_norm": 0.1643335074186325,
      "learning_rate": 8.384089305977244e-05,
      "loss": 0.8504,
      "mean_token_accuracy": 0.7544571943581104,
      "num_tokens": 168212917.0,
      "step": 5270
    },
    {
      "entropy": 0.7913817465305328,
      "epoch": 0.48496558386668276,
      "grad_norm": 0.15639027953147888,
      "learning_rate": 8.383782623363082e-05,
      "loss": 0.79,
      "mean_token_accuracy": 0.7677251137793064,
      "num_tokens": 168245280.0,
      "step": 5271
    },
    {
      "entropy": 0.8033524360507727,
      "epoch": 0.485057590238124,
      "grad_norm": 0.1558280736207962,
      "learning_rate": 8.383475940748919e-05,
      "loss": 0.797,
      "mean_token_accuracy": 0.7664866298437119,
      "num_tokens": 168277687.0,
      "step": 5272
    },
    {
      "entropy": 0.7709119506180286,
      "epoch": 0.4851495966095652,
      "grad_norm": 0.15893147885799408,
      "learning_rate": 8.383169258134757e-05,
      "loss": 0.7566,
      "mean_token_accuracy": 0.775879979133606,
      "num_tokens": 168308188.0,
      "step": 5273
    },
    {
      "entropy": 0.7468138812109828,
      "epoch": 0.48524160298100644,
      "grad_norm": 0.15799476206302643,
      "learning_rate": 8.382862575520594e-05,
      "loss": 0.718,
      "mean_token_accuracy": 0.7915283069014549,
      "num_tokens": 168340482.0,
      "step": 5274
    },
    {
      "entropy": 0.765857569873333,
      "epoch": 0.4853336093524477,
      "grad_norm": 0.15909279882907867,
      "learning_rate": 8.382555892906432e-05,
      "loss": 0.7709,
      "mean_token_accuracy": 0.7790596447885036,
      "num_tokens": 168372901.0,
      "step": 5275
    },
    {
      "entropy": 0.794036477804184,
      "epoch": 0.48542561572388887,
      "grad_norm": 0.15433980524539948,
      "learning_rate": 8.382249210292269e-05,
      "loss": 0.8005,
      "mean_token_accuracy": 0.769396897405386,
      "num_tokens": 168405611.0,
      "step": 5276
    },
    {
      "entropy": 0.7326240856200457,
      "epoch": 0.4855176220953301,
      "grad_norm": 0.1558351069688797,
      "learning_rate": 8.381942527678105e-05,
      "loss": 0.7334,
      "mean_token_accuracy": 0.7827485166490078,
      "num_tokens": 168437209.0,
      "step": 5277
    },
    {
      "entropy": 0.7284587118774652,
      "epoch": 0.4856096284667713,
      "grad_norm": 0.1703798472881317,
      "learning_rate": 8.381635845063945e-05,
      "loss": 0.7243,
      "mean_token_accuracy": 0.786476731300354,
      "num_tokens": 168468896.0,
      "step": 5278
    },
    {
      "entropy": 0.7461251970380545,
      "epoch": 0.48570163483821255,
      "grad_norm": 0.16644561290740967,
      "learning_rate": 8.381329162449782e-05,
      "loss": 0.7447,
      "mean_token_accuracy": 0.782730296254158,
      "num_tokens": 168500404.0,
      "step": 5279
    },
    {
      "entropy": 0.7385731600224972,
      "epoch": 0.4857936412096538,
      "grad_norm": 0.16877306997776031,
      "learning_rate": 8.381022479835618e-05,
      "loss": 0.7367,
      "mean_token_accuracy": 0.7801456786692142,
      "num_tokens": 168531916.0,
      "step": 5280
    },
    {
      "entropy": 0.8467577956616879,
      "epoch": 0.485885647581095,
      "grad_norm": 0.16069605946540833,
      "learning_rate": 8.380715797221455e-05,
      "loss": 0.8472,
      "mean_token_accuracy": 0.7562742754817009,
      "num_tokens": 168563959.0,
      "step": 5281
    },
    {
      "entropy": 0.8272961135953665,
      "epoch": 0.4859776539525362,
      "grad_norm": 0.1529092639684677,
      "learning_rate": 8.380409114607293e-05,
      "loss": 0.8043,
      "mean_token_accuracy": 0.7676254287362099,
      "num_tokens": 168595920.0,
      "step": 5282
    },
    {
      "entropy": 0.7120839841663837,
      "epoch": 0.4860696603239774,
      "grad_norm": 0.15097039937973022,
      "learning_rate": 8.380102431993132e-05,
      "loss": 0.6812,
      "mean_token_accuracy": 0.7968635261058807,
      "num_tokens": 168627285.0,
      "step": 5283
    },
    {
      "entropy": 0.9186556003987789,
      "epoch": 0.48616166669541866,
      "grad_norm": 0.16695107519626617,
      "learning_rate": 8.379795749378968e-05,
      "loss": 0.89,
      "mean_token_accuracy": 0.7485988847911358,
      "num_tokens": 168658928.0,
      "step": 5284
    },
    {
      "entropy": 0.7840875647962093,
      "epoch": 0.4862536730668599,
      "grad_norm": 0.15783576667308807,
      "learning_rate": 8.379489066764805e-05,
      "loss": 0.7734,
      "mean_token_accuracy": 0.7769701257348061,
      "num_tokens": 168690230.0,
      "step": 5285
    },
    {
      "entropy": 0.7499577924609184,
      "epoch": 0.4863456794383011,
      "grad_norm": 0.15485651791095734,
      "learning_rate": 8.379182384150643e-05,
      "loss": 0.7256,
      "mean_token_accuracy": 0.7864547073841095,
      "num_tokens": 168722238.0,
      "step": 5286
    },
    {
      "entropy": 0.7360309716314077,
      "epoch": 0.48643768580974234,
      "grad_norm": 0.1574356108903885,
      "learning_rate": 8.37887570153648e-05,
      "loss": 0.7194,
      "mean_token_accuracy": 0.7873851098120213,
      "num_tokens": 168754728.0,
      "step": 5287
    },
    {
      "entropy": 0.7805514466017485,
      "epoch": 0.4865296921811835,
      "grad_norm": 0.16024157404899597,
      "learning_rate": 8.378569018922318e-05,
      "loss": 0.7838,
      "mean_token_accuracy": 0.7737994603812695,
      "num_tokens": 168787077.0,
      "step": 5288
    },
    {
      "entropy": 0.7734700981527567,
      "epoch": 0.48662169855262477,
      "grad_norm": 0.15774419903755188,
      "learning_rate": 8.378262336308155e-05,
      "loss": 0.7516,
      "mean_token_accuracy": 0.7794651016592979,
      "num_tokens": 168819388.0,
      "step": 5289
    },
    {
      "entropy": 0.7379067242145538,
      "epoch": 0.486713704924066,
      "grad_norm": 0.1612413227558136,
      "learning_rate": 8.377955653693993e-05,
      "loss": 0.7138,
      "mean_token_accuracy": 0.7897556349635124,
      "num_tokens": 168850194.0,
      "step": 5290
    },
    {
      "entropy": 0.6549988947808743,
      "epoch": 0.4868057112955072,
      "grad_norm": 0.15943776071071625,
      "learning_rate": 8.37764897107983e-05,
      "loss": 0.6602,
      "mean_token_accuracy": 0.8066932037472725,
      "num_tokens": 168881573.0,
      "step": 5291
    },
    {
      "entropy": 0.7337059173732996,
      "epoch": 0.48689771766694845,
      "grad_norm": 0.14983701705932617,
      "learning_rate": 8.377342288465667e-05,
      "loss": 0.7128,
      "mean_token_accuracy": 0.7889367938041687,
      "num_tokens": 168914264.0,
      "step": 5292
    },
    {
      "entropy": 0.844605278223753,
      "epoch": 0.48698972403838964,
      "grad_norm": 0.16566304862499237,
      "learning_rate": 8.377035605851505e-05,
      "loss": 0.8307,
      "mean_token_accuracy": 0.7604597508907318,
      "num_tokens": 168946358.0,
      "step": 5293
    },
    {
      "entropy": 0.7201855331659317,
      "epoch": 0.4870817304098309,
      "grad_norm": 0.15863245725631714,
      "learning_rate": 8.376728923237343e-05,
      "loss": 0.7345,
      "mean_token_accuracy": 0.7819554917514324,
      "num_tokens": 168977389.0,
      "step": 5294
    },
    {
      "entropy": 0.8196323197335005,
      "epoch": 0.4871737367812721,
      "grad_norm": 0.16080327332019806,
      "learning_rate": 8.37642224062318e-05,
      "loss": 0.8145,
      "mean_token_accuracy": 0.7665210887789726,
      "num_tokens": 169010050.0,
      "step": 5295
    },
    {
      "entropy": 0.7937330435961485,
      "epoch": 0.4872657431527133,
      "grad_norm": 0.16950613260269165,
      "learning_rate": 8.376115558009017e-05,
      "loss": 0.8015,
      "mean_token_accuracy": 0.766660001128912,
      "num_tokens": 169042137.0,
      "step": 5296
    },
    {
      "entropy": 0.7386066764593124,
      "epoch": 0.48735774952415456,
      "grad_norm": 0.16542845964431763,
      "learning_rate": 8.375808875394853e-05,
      "loss": 0.7461,
      "mean_token_accuracy": 0.783776018768549,
      "num_tokens": 169074612.0,
      "step": 5297
    },
    {
      "entropy": 0.8246335424482822,
      "epoch": 0.48744975589559575,
      "grad_norm": 0.1569824069738388,
      "learning_rate": 8.375502192780691e-05,
      "loss": 0.7929,
      "mean_token_accuracy": 0.7683084085583687,
      "num_tokens": 169106045.0,
      "step": 5298
    },
    {
      "entropy": 0.7358281649649143,
      "epoch": 0.487541762267037,
      "grad_norm": 0.15619471669197083,
      "learning_rate": 8.37519551016653e-05,
      "loss": 0.7235,
      "mean_token_accuracy": 0.7873042337596416,
      "num_tokens": 169138370.0,
      "step": 5299
    },
    {
      "entropy": 0.8247027695178986,
      "epoch": 0.48763376863847824,
      "grad_norm": 0.16313207149505615,
      "learning_rate": 8.374888827552366e-05,
      "loss": 0.8098,
      "mean_token_accuracy": 0.7676379978656769,
      "num_tokens": 169169748.0,
      "step": 5300
    },
    {
      "entropy": 0.7582145035266876,
      "epoch": 0.4877257750099194,
      "grad_norm": 0.15540950000286102,
      "learning_rate": 8.374582144938203e-05,
      "loss": 0.7509,
      "mean_token_accuracy": 0.7808375358581543,
      "num_tokens": 169201896.0,
      "step": 5301
    },
    {
      "entropy": 0.6735804937779903,
      "epoch": 0.48781778138136067,
      "grad_norm": 0.15288351476192474,
      "learning_rate": 8.374275462324041e-05,
      "loss": 0.6628,
      "mean_token_accuracy": 0.8006622269749641,
      "num_tokens": 169233673.0,
      "step": 5302
    },
    {
      "entropy": 0.7497870270162821,
      "epoch": 0.48790978775280186,
      "grad_norm": 0.1555968075990677,
      "learning_rate": 8.373968779709878e-05,
      "loss": 0.7452,
      "mean_token_accuracy": 0.7821176201105118,
      "num_tokens": 169265739.0,
      "step": 5303
    },
    {
      "entropy": 0.6872782465070486,
      "epoch": 0.4880017941242431,
      "grad_norm": 0.15121780335903168,
      "learning_rate": 8.373662097095716e-05,
      "loss": 0.6553,
      "mean_token_accuracy": 0.8048043511807919,
      "num_tokens": 169297599.0,
      "step": 5304
    },
    {
      "entropy": 0.7742238529026508,
      "epoch": 0.48809380049568435,
      "grad_norm": 0.1531965434551239,
      "learning_rate": 8.373355414481553e-05,
      "loss": 0.7504,
      "mean_token_accuracy": 0.7785509303212166,
      "num_tokens": 169328947.0,
      "step": 5305
    },
    {
      "entropy": 0.8074370250105858,
      "epoch": 0.48818580686712554,
      "grad_norm": 0.16804486513137817,
      "learning_rate": 8.373048731867391e-05,
      "loss": 0.7918,
      "mean_token_accuracy": 0.7728623300790787,
      "num_tokens": 169359876.0,
      "step": 5306
    },
    {
      "entropy": 0.6962380614131689,
      "epoch": 0.4882778132385668,
      "grad_norm": 0.15348432958126068,
      "learning_rate": 8.372742049253228e-05,
      "loss": 0.6827,
      "mean_token_accuracy": 0.7991633452475071,
      "num_tokens": 169391092.0,
      "step": 5307
    },
    {
      "entropy": 0.7099263872951269,
      "epoch": 0.48836981961000797,
      "grad_norm": 0.15462391078472137,
      "learning_rate": 8.372435366639065e-05,
      "loss": 0.7021,
      "mean_token_accuracy": 0.7939573489129543,
      "num_tokens": 169423416.0,
      "step": 5308
    },
    {
      "entropy": 0.789932819083333,
      "epoch": 0.4884618259814492,
      "grad_norm": 0.16334903240203857,
      "learning_rate": 8.372128684024903e-05,
      "loss": 0.8005,
      "mean_token_accuracy": 0.7662065178155899,
      "num_tokens": 169455700.0,
      "step": 5309
    },
    {
      "entropy": 0.8152282871305943,
      "epoch": 0.48855383235289046,
      "grad_norm": 0.16214345395565033,
      "learning_rate": 8.371822001410741e-05,
      "loss": 0.8344,
      "mean_token_accuracy": 0.7530752792954445,
      "num_tokens": 169487523.0,
      "step": 5310
    },
    {
      "entropy": 0.6866544429212809,
      "epoch": 0.48864583872433165,
      "grad_norm": 0.1518355756998062,
      "learning_rate": 8.371515318796578e-05,
      "loss": 0.6901,
      "mean_token_accuracy": 0.7955197021365166,
      "num_tokens": 169519646.0,
      "step": 5311
    },
    {
      "entropy": 0.7507355455309153,
      "epoch": 0.4887378450957729,
      "grad_norm": 0.15958857536315918,
      "learning_rate": 8.371208636182415e-05,
      "loss": 0.7832,
      "mean_token_accuracy": 0.7727410532534122,
      "num_tokens": 169551226.0,
      "step": 5312
    },
    {
      "entropy": 0.7657914236187935,
      "epoch": 0.4888298514672141,
      "grad_norm": 0.15884065628051758,
      "learning_rate": 8.370901953568253e-05,
      "loss": 0.7516,
      "mean_token_accuracy": 0.7817381285130978,
      "num_tokens": 169583194.0,
      "step": 5313
    },
    {
      "entropy": 0.86006173864007,
      "epoch": 0.4889218578386553,
      "grad_norm": 0.16038258373737335,
      "learning_rate": 8.370595270954091e-05,
      "loss": 0.8258,
      "mean_token_accuracy": 0.7602813430130482,
      "num_tokens": 169615496.0,
      "step": 5314
    },
    {
      "entropy": 0.7403207048773766,
      "epoch": 0.48901386421009657,
      "grad_norm": 0.14470939338207245,
      "learning_rate": 8.370288588339928e-05,
      "loss": 0.7111,
      "mean_token_accuracy": 0.7905456088483334,
      "num_tokens": 169647922.0,
      "step": 5315
    },
    {
      "entropy": 0.812068484723568,
      "epoch": 0.48910587058153776,
      "grad_norm": 0.16960978507995605,
      "learning_rate": 8.369981905725764e-05,
      "loss": 0.8101,
      "mean_token_accuracy": 0.7597853019833565,
      "num_tokens": 169680186.0,
      "step": 5316
    },
    {
      "entropy": 0.7695570737123489,
      "epoch": 0.489197876952979,
      "grad_norm": 0.15346820652484894,
      "learning_rate": 8.369675223111603e-05,
      "loss": 0.7366,
      "mean_token_accuracy": 0.786888562142849,
      "num_tokens": 169712241.0,
      "step": 5317
    },
    {
      "entropy": 0.8264571074396372,
      "epoch": 0.4892898833244202,
      "grad_norm": 0.1706530749797821,
      "learning_rate": 8.36936854049744e-05,
      "loss": 0.823,
      "mean_token_accuracy": 0.7578515261411667,
      "num_tokens": 169743671.0,
      "step": 5318
    },
    {
      "entropy": 0.8738158773630857,
      "epoch": 0.48938188969586144,
      "grad_norm": 0.1591642051935196,
      "learning_rate": 8.369061857883278e-05,
      "loss": 0.8734,
      "mean_token_accuracy": 0.7474250644445419,
      "num_tokens": 169776103.0,
      "step": 5319
    },
    {
      "entropy": 0.844796952791512,
      "epoch": 0.4894738960673027,
      "grad_norm": 0.1597045361995697,
      "learning_rate": 8.368755175269114e-05,
      "loss": 0.845,
      "mean_token_accuracy": 0.7545814700424671,
      "num_tokens": 169808620.0,
      "step": 5320
    },
    {
      "entropy": 0.7762376312166452,
      "epoch": 0.48956590243874387,
      "grad_norm": 0.1571209877729416,
      "learning_rate": 8.368448492654952e-05,
      "loss": 0.7522,
      "mean_token_accuracy": 0.7817677892744541,
      "num_tokens": 169840833.0,
      "step": 5321
    },
    {
      "entropy": 0.6707482766360044,
      "epoch": 0.4896579088101851,
      "grad_norm": 0.15030018985271454,
      "learning_rate": 8.368141810040789e-05,
      "loss": 0.647,
      "mean_token_accuracy": 0.8089657761156559,
      "num_tokens": 169872859.0,
      "step": 5322
    },
    {
      "entropy": 0.76469986140728,
      "epoch": 0.4897499151816263,
      "grad_norm": 0.15592366456985474,
      "learning_rate": 8.367835127426626e-05,
      "loss": 0.7359,
      "mean_token_accuracy": 0.786103006452322,
      "num_tokens": 169904910.0,
      "step": 5323
    },
    {
      "entropy": 0.7736034970730543,
      "epoch": 0.48984192155306755,
      "grad_norm": 0.1705489307641983,
      "learning_rate": 8.367528444812464e-05,
      "loss": 0.7653,
      "mean_token_accuracy": 0.7801216244697571,
      "num_tokens": 169937094.0,
      "step": 5324
    },
    {
      "entropy": 0.7195946201682091,
      "epoch": 0.4899339279245088,
      "grad_norm": 0.15894073247909546,
      "learning_rate": 8.367221762198302e-05,
      "loss": 0.7032,
      "mean_token_accuracy": 0.793419923633337,
      "num_tokens": 169969111.0,
      "step": 5325
    },
    {
      "entropy": 0.7228182936087251,
      "epoch": 0.49002593429595,
      "grad_norm": 0.16348357498645782,
      "learning_rate": 8.366915079584139e-05,
      "loss": 0.6984,
      "mean_token_accuracy": 0.7956139743328094,
      "num_tokens": 170001109.0,
      "step": 5326
    },
    {
      "entropy": 0.789200821891427,
      "epoch": 0.4901179406673912,
      "grad_norm": 0.15927055478096008,
      "learning_rate": 8.366608396969976e-05,
      "loss": 0.7957,
      "mean_token_accuracy": 0.7659727670252323,
      "num_tokens": 170032824.0,
      "step": 5327
    },
    {
      "entropy": 0.6839047782123089,
      "epoch": 0.4902099470388324,
      "grad_norm": 0.16793055832386017,
      "learning_rate": 8.366301714355813e-05,
      "loss": 0.6747,
      "mean_token_accuracy": 0.7994929812848568,
      "num_tokens": 170065087.0,
      "step": 5328
    },
    {
      "entropy": 0.7993769347667694,
      "epoch": 0.49030195341027366,
      "grad_norm": 0.16473019123077393,
      "learning_rate": 8.365995031741651e-05,
      "loss": 0.8085,
      "mean_token_accuracy": 0.7599109522998333,
      "num_tokens": 170096789.0,
      "step": 5329
    },
    {
      "entropy": 0.7389365620911121,
      "epoch": 0.4903939597817149,
      "grad_norm": 0.1516316533088684,
      "learning_rate": 8.365688349127489e-05,
      "loss": 0.7337,
      "mean_token_accuracy": 0.7826805338263512,
      "num_tokens": 170128427.0,
      "step": 5330
    },
    {
      "entropy": 0.8011597860604525,
      "epoch": 0.4904859661531561,
      "grad_norm": 0.15314991772174835,
      "learning_rate": 8.365381666513326e-05,
      "loss": 0.786,
      "mean_token_accuracy": 0.7750254943966866,
      "num_tokens": 170159374.0,
      "step": 5331
    },
    {
      "entropy": 0.8319506645202637,
      "epoch": 0.49057797252459734,
      "grad_norm": 0.16320641338825226,
      "learning_rate": 8.365074983899163e-05,
      "loss": 0.8235,
      "mean_token_accuracy": 0.7664245404303074,
      "num_tokens": 170190539.0,
      "step": 5332
    },
    {
      "entropy": 0.7923352718353271,
      "epoch": 0.49066997889603853,
      "grad_norm": 0.16299836337566376,
      "learning_rate": 8.364768301285001e-05,
      "loss": 0.8113,
      "mean_token_accuracy": 0.7635616585612297,
      "num_tokens": 170222672.0,
      "step": 5333
    },
    {
      "entropy": 0.7411518841981888,
      "epoch": 0.49076198526747977,
      "grad_norm": 0.15562936663627625,
      "learning_rate": 8.364461618670837e-05,
      "loss": 0.7227,
      "mean_token_accuracy": 0.7908164672553539,
      "num_tokens": 170254448.0,
      "step": 5334
    },
    {
      "entropy": 0.7299806792289019,
      "epoch": 0.490853991638921,
      "grad_norm": 0.14832840859889984,
      "learning_rate": 8.364154936056676e-05,
      "loss": 0.7157,
      "mean_token_accuracy": 0.7884033657610416,
      "num_tokens": 170286824.0,
      "step": 5335
    },
    {
      "entropy": 0.7818520087748766,
      "epoch": 0.4909459980103622,
      "grad_norm": 0.1580251008272171,
      "learning_rate": 8.363848253442512e-05,
      "loss": 0.7828,
      "mean_token_accuracy": 0.7717399671673775,
      "num_tokens": 170318554.0,
      "step": 5336
    },
    {
      "entropy": 0.7332432605326176,
      "epoch": 0.49103800438180345,
      "grad_norm": 0.15030716359615326,
      "learning_rate": 8.36354157082835e-05,
      "loss": 0.7279,
      "mean_token_accuracy": 0.7825294882059097,
      "num_tokens": 170350517.0,
      "step": 5337
    },
    {
      "entropy": 0.9258831776678562,
      "epoch": 0.49113001075324464,
      "grad_norm": 0.16450072824954987,
      "learning_rate": 8.363234888214187e-05,
      "loss": 0.926,
      "mean_token_accuracy": 0.7369812838733196,
      "num_tokens": 170382502.0,
      "step": 5338
    },
    {
      "entropy": 0.7290662843734026,
      "epoch": 0.4912220171246859,
      "grad_norm": 0.1571861058473587,
      "learning_rate": 8.362928205600024e-05,
      "loss": 0.7147,
      "mean_token_accuracy": 0.7904851399362087,
      "num_tokens": 170414645.0,
      "step": 5339
    },
    {
      "entropy": 0.7581061404198408,
      "epoch": 0.49131402349612713,
      "grad_norm": 0.16303549706935883,
      "learning_rate": 8.362621522985862e-05,
      "loss": 0.7424,
      "mean_token_accuracy": 0.7826077155768871,
      "num_tokens": 170446220.0,
      "step": 5340
    },
    {
      "entropy": 0.8072613067924976,
      "epoch": 0.4914060298675683,
      "grad_norm": 0.15728162229061127,
      "learning_rate": 8.3623148403717e-05,
      "loss": 0.7985,
      "mean_token_accuracy": 0.77291439473629,
      "num_tokens": 170478318.0,
      "step": 5341
    },
    {
      "entropy": 0.7869525291025639,
      "epoch": 0.49149803623900956,
      "grad_norm": 0.15549041330814362,
      "learning_rate": 8.362008157757537e-05,
      "loss": 0.7807,
      "mean_token_accuracy": 0.7734658680856228,
      "num_tokens": 170510364.0,
      "step": 5342
    },
    {
      "entropy": 0.8162193447351456,
      "epoch": 0.49159004261045075,
      "grad_norm": 0.16021792590618134,
      "learning_rate": 8.361701475143374e-05,
      "loss": 0.8244,
      "mean_token_accuracy": 0.7645485885441303,
      "num_tokens": 170542562.0,
      "step": 5343
    },
    {
      "entropy": 0.7243175953626633,
      "epoch": 0.491682048981892,
      "grad_norm": 0.15131469070911407,
      "learning_rate": 8.361394792529212e-05,
      "loss": 0.7146,
      "mean_token_accuracy": 0.7874338328838348,
      "num_tokens": 170574587.0,
      "step": 5344
    },
    {
      "entropy": 0.714677007868886,
      "epoch": 0.49177405535333324,
      "grad_norm": 0.16157418489456177,
      "learning_rate": 8.361088109915049e-05,
      "loss": 0.7024,
      "mean_token_accuracy": 0.7936412766575813,
      "num_tokens": 170606022.0,
      "step": 5345
    },
    {
      "entropy": 0.7487196680158377,
      "epoch": 0.49186606172477443,
      "grad_norm": 0.16138139367103577,
      "learning_rate": 8.360781427300887e-05,
      "loss": 0.7392,
      "mean_token_accuracy": 0.7847678735852242,
      "num_tokens": 170638711.0,
      "step": 5346
    },
    {
      "entropy": 0.8109553903341293,
      "epoch": 0.4919580680962157,
      "grad_norm": 0.16558264195919037,
      "learning_rate": 8.360474744686724e-05,
      "loss": 0.8279,
      "mean_token_accuracy": 0.760994341224432,
      "num_tokens": 170670410.0,
      "step": 5347
    },
    {
      "entropy": 0.7934878692030907,
      "epoch": 0.49205007446765686,
      "grad_norm": 0.15727227926254272,
      "learning_rate": 8.360168062072562e-05,
      "loss": 0.7921,
      "mean_token_accuracy": 0.770040974020958,
      "num_tokens": 170702354.0,
      "step": 5348
    },
    {
      "entropy": 0.7492975126951933,
      "epoch": 0.4921420808390981,
      "grad_norm": 0.15374375879764557,
      "learning_rate": 8.359861379458399e-05,
      "loss": 0.7378,
      "mean_token_accuracy": 0.7857960313558578,
      "num_tokens": 170734361.0,
      "step": 5349
    },
    {
      "entropy": 0.7299806084483862,
      "epoch": 0.49223408721053935,
      "grad_norm": 0.15844367444515228,
      "learning_rate": 8.359554696844236e-05,
      "loss": 0.7387,
      "mean_token_accuracy": 0.7805391848087311,
      "num_tokens": 170766611.0,
      "step": 5350
    },
    {
      "entropy": 0.785131586715579,
      "epoch": 0.49232609358198054,
      "grad_norm": 0.15909206867218018,
      "learning_rate": 8.359248014230074e-05,
      "loss": 0.7894,
      "mean_token_accuracy": 0.768325112760067,
      "num_tokens": 170798890.0,
      "step": 5351
    },
    {
      "entropy": 0.7684484273195267,
      "epoch": 0.4924180999534218,
      "grad_norm": 0.16294951736927032,
      "learning_rate": 8.358941331615912e-05,
      "loss": 0.7558,
      "mean_token_accuracy": 0.7811664491891861,
      "num_tokens": 170830696.0,
      "step": 5352
    },
    {
      "entropy": 0.8317146580666304,
      "epoch": 0.492510106324863,
      "grad_norm": 0.16160865128040314,
      "learning_rate": 8.358634649001749e-05,
      "loss": 0.8246,
      "mean_token_accuracy": 0.7579181157052517,
      "num_tokens": 170863119.0,
      "step": 5353
    },
    {
      "entropy": 0.8187718074768782,
      "epoch": 0.4926021126963042,
      "grad_norm": 0.16124962270259857,
      "learning_rate": 8.358327966387585e-05,
      "loss": 0.7938,
      "mean_token_accuracy": 0.7641281485557556,
      "num_tokens": 170894832.0,
      "step": 5354
    },
    {
      "entropy": 0.8172686640173197,
      "epoch": 0.49269411906774546,
      "grad_norm": 0.1560146063566208,
      "learning_rate": 8.358021283773422e-05,
      "loss": 0.8093,
      "mean_token_accuracy": 0.7671939693391323,
      "num_tokens": 170926763.0,
      "step": 5355
    },
    {
      "entropy": 0.7173682916909456,
      "epoch": 0.49278612543918665,
      "grad_norm": 0.1677677482366562,
      "learning_rate": 8.357714601159262e-05,
      "loss": 0.7027,
      "mean_token_accuracy": 0.7966635785996914,
      "num_tokens": 170958458.0,
      "step": 5356
    },
    {
      "entropy": 0.8057412505149841,
      "epoch": 0.4928781318106279,
      "grad_norm": 0.1602148413658142,
      "learning_rate": 8.357407918545099e-05,
      "loss": 0.7862,
      "mean_token_accuracy": 0.7696907334029675,
      "num_tokens": 170990040.0,
      "step": 5357
    },
    {
      "entropy": 0.8227789960801601,
      "epoch": 0.4929701381820691,
      "grad_norm": 0.1584843248128891,
      "learning_rate": 8.357101235930935e-05,
      "loss": 0.8094,
      "mean_token_accuracy": 0.7617321982979774,
      "num_tokens": 171022328.0,
      "step": 5358
    },
    {
      "entropy": 0.8012432958930731,
      "epoch": 0.49306214455351033,
      "grad_norm": 0.15553541481494904,
      "learning_rate": 8.356794553316772e-05,
      "loss": 0.7775,
      "mean_token_accuracy": 0.771734070032835,
      "num_tokens": 171054443.0,
      "step": 5359
    },
    {
      "entropy": 0.7871726900339127,
      "epoch": 0.4931541509249516,
      "grad_norm": 0.15671464800834656,
      "learning_rate": 8.35648787070261e-05,
      "loss": 0.7723,
      "mean_token_accuracy": 0.771652489900589,
      "num_tokens": 171086331.0,
      "step": 5360
    },
    {
      "entropy": 0.8555128276348114,
      "epoch": 0.49324615729639276,
      "grad_norm": 0.15918618440628052,
      "learning_rate": 8.356181188088448e-05,
      "loss": 0.8308,
      "mean_token_accuracy": 0.7615698873996735,
      "num_tokens": 171118767.0,
      "step": 5361
    },
    {
      "entropy": 0.6796811167150736,
      "epoch": 0.493338163667834,
      "grad_norm": 0.15324363112449646,
      "learning_rate": 8.355874505474285e-05,
      "loss": 0.6555,
      "mean_token_accuracy": 0.8070594482123852,
      "num_tokens": 171150889.0,
      "step": 5362
    },
    {
      "entropy": 0.8044796958565712,
      "epoch": 0.4934301700392752,
      "grad_norm": 0.1522854119539261,
      "learning_rate": 8.355567822860122e-05,
      "loss": 0.7706,
      "mean_token_accuracy": 0.7725230827927589,
      "num_tokens": 171182361.0,
      "step": 5363
    },
    {
      "entropy": 0.8131247609853745,
      "epoch": 0.49352217641071644,
      "grad_norm": 0.15971961617469788,
      "learning_rate": 8.35526114024596e-05,
      "loss": 0.8234,
      "mean_token_accuracy": 0.7596108764410019,
      "num_tokens": 171214315.0,
      "step": 5364
    },
    {
      "entropy": 0.8448972348123789,
      "epoch": 0.4936141827821577,
      "grad_norm": 0.16306236386299133,
      "learning_rate": 8.354954457631797e-05,
      "loss": 0.8526,
      "mean_token_accuracy": 0.7508204840123653,
      "num_tokens": 171245870.0,
      "step": 5365
    },
    {
      "entropy": 0.6509686950594187,
      "epoch": 0.4937061891535989,
      "grad_norm": 0.1543567180633545,
      "learning_rate": 8.354647775017635e-05,
      "loss": 0.6516,
      "mean_token_accuracy": 0.8064000308513641,
      "num_tokens": 171278160.0,
      "step": 5366
    },
    {
      "entropy": 0.7360285185277462,
      "epoch": 0.4937981955250401,
      "grad_norm": 0.1575927436351776,
      "learning_rate": 8.354341092403472e-05,
      "loss": 0.7326,
      "mean_token_accuracy": 0.7836422584950924,
      "num_tokens": 171309761.0,
      "step": 5367
    },
    {
      "entropy": 0.7725596781820059,
      "epoch": 0.4938902018964813,
      "grad_norm": 0.16348469257354736,
      "learning_rate": 8.35403440978931e-05,
      "loss": 0.7815,
      "mean_token_accuracy": 0.772874753922224,
      "num_tokens": 171342128.0,
      "step": 5368
    },
    {
      "entropy": 0.8566589653491974,
      "epoch": 0.49398220826792255,
      "grad_norm": 0.16245239973068237,
      "learning_rate": 8.353727727175147e-05,
      "loss": 0.8469,
      "mean_token_accuracy": 0.7528105564415455,
      "num_tokens": 171373703.0,
      "step": 5369
    },
    {
      "entropy": 0.7645774967968464,
      "epoch": 0.4940742146393638,
      "grad_norm": 0.16097748279571533,
      "learning_rate": 8.353421044560984e-05,
      "loss": 0.7583,
      "mean_token_accuracy": 0.7764722034335136,
      "num_tokens": 171405457.0,
      "step": 5370
    },
    {
      "entropy": 0.7626279983669519,
      "epoch": 0.494166221010805,
      "grad_norm": 0.16256771981716156,
      "learning_rate": 8.353114361946822e-05,
      "loss": 0.7636,
      "mean_token_accuracy": 0.7757686525583267,
      "num_tokens": 171437349.0,
      "step": 5371
    },
    {
      "entropy": 0.817586038261652,
      "epoch": 0.49425822738224623,
      "grad_norm": 0.1603650599718094,
      "learning_rate": 8.35280767933266e-05,
      "loss": 0.8034,
      "mean_token_accuracy": 0.7649774216115475,
      "num_tokens": 171469263.0,
      "step": 5372
    },
    {
      "entropy": 0.7335409559309483,
      "epoch": 0.4943502337536874,
      "grad_norm": 0.15520086884498596,
      "learning_rate": 8.352500996718497e-05,
      "loss": 0.7052,
      "mean_token_accuracy": 0.7938736006617546,
      "num_tokens": 171501546.0,
      "step": 5373
    },
    {
      "entropy": 0.9644126463681459,
      "epoch": 0.49444224012512866,
      "grad_norm": 0.19047871232032776,
      "learning_rate": 8.352194314104333e-05,
      "loss": 0.9582,
      "mean_token_accuracy": 0.7319890074431896,
      "num_tokens": 171533892.0,
      "step": 5374
    },
    {
      "entropy": 0.7419181056320667,
      "epoch": 0.4945342464965699,
      "grad_norm": 0.15243399143218994,
      "learning_rate": 8.351887631490172e-05,
      "loss": 0.7038,
      "mean_token_accuracy": 0.7902919501066208,
      "num_tokens": 171565477.0,
      "step": 5375
    },
    {
      "entropy": 0.7293214686214924,
      "epoch": 0.4946262528680111,
      "grad_norm": 0.1615508496761322,
      "learning_rate": 8.351580948876008e-05,
      "loss": 0.728,
      "mean_token_accuracy": 0.7896228469908237,
      "num_tokens": 171597406.0,
      "step": 5376
    },
    {
      "entropy": 0.8058056533336639,
      "epoch": 0.49471825923945234,
      "grad_norm": 0.16120825707912445,
      "learning_rate": 8.351274266261846e-05,
      "loss": 0.784,
      "mean_token_accuracy": 0.7706853784620762,
      "num_tokens": 171629441.0,
      "step": 5377
    },
    {
      "entropy": 0.7814044635742903,
      "epoch": 0.49481026561089353,
      "grad_norm": 0.1584048718214035,
      "learning_rate": 8.350967583647683e-05,
      "loss": 0.7746,
      "mean_token_accuracy": 0.7754523567855358,
      "num_tokens": 171661208.0,
      "step": 5378
    },
    {
      "entropy": 0.7007419355213642,
      "epoch": 0.4949022719823348,
      "grad_norm": 0.14645954966545105,
      "learning_rate": 8.350660901033521e-05,
      "loss": 0.6907,
      "mean_token_accuracy": 0.7919119149446487,
      "num_tokens": 171693481.0,
      "step": 5379
    },
    {
      "entropy": 0.7744574621319771,
      "epoch": 0.494994278353776,
      "grad_norm": 0.16595372557640076,
      "learning_rate": 8.350354218419358e-05,
      "loss": 0.7801,
      "mean_token_accuracy": 0.7716067768633366,
      "num_tokens": 171724882.0,
      "step": 5380
    },
    {
      "entropy": 0.7201150376349688,
      "epoch": 0.4950862847252172,
      "grad_norm": 0.16340568661689758,
      "learning_rate": 8.350047535805195e-05,
      "loss": 0.729,
      "mean_token_accuracy": 0.7888047024607658,
      "num_tokens": 171756761.0,
      "step": 5381
    },
    {
      "entropy": 0.761959858238697,
      "epoch": 0.49517829109665845,
      "grad_norm": 0.16312260925769806,
      "learning_rate": 8.349740853191033e-05,
      "loss": 0.7675,
      "mean_token_accuracy": 0.7791012898087502,
      "num_tokens": 171788793.0,
      "step": 5382
    },
    {
      "entropy": 0.7072806153446436,
      "epoch": 0.49527029746809964,
      "grad_norm": 0.15895570814609528,
      "learning_rate": 8.349434170576871e-05,
      "loss": 0.6803,
      "mean_token_accuracy": 0.7985725551843643,
      "num_tokens": 171820087.0,
      "step": 5383
    },
    {
      "entropy": 0.7879549209028482,
      "epoch": 0.4953623038395409,
      "grad_norm": 0.15966059267520905,
      "learning_rate": 8.349127487962708e-05,
      "loss": 0.7912,
      "mean_token_accuracy": 0.7723772674798965,
      "num_tokens": 171852721.0,
      "step": 5384
    },
    {
      "entropy": 0.7745067607611418,
      "epoch": 0.49545431021098213,
      "grad_norm": 0.1549137979745865,
      "learning_rate": 8.348820805348545e-05,
      "loss": 0.7491,
      "mean_token_accuracy": 0.7799713760614395,
      "num_tokens": 171884813.0,
      "step": 5385
    },
    {
      "entropy": 0.7865660600364208,
      "epoch": 0.4955463165824233,
      "grad_norm": 0.16075140237808228,
      "learning_rate": 8.348514122734382e-05,
      "loss": 0.7785,
      "mean_token_accuracy": 0.7733888924121857,
      "num_tokens": 171915983.0,
      "step": 5386
    },
    {
      "entropy": 0.8098998609930277,
      "epoch": 0.49563832295386456,
      "grad_norm": 0.1613669991493225,
      "learning_rate": 8.34820744012022e-05,
      "loss": 0.7887,
      "mean_token_accuracy": 0.7742649018764496,
      "num_tokens": 171948446.0,
      "step": 5387
    },
    {
      "entropy": 0.7188575994223356,
      "epoch": 0.49573032932530575,
      "grad_norm": 0.14650674164295197,
      "learning_rate": 8.347900757506058e-05,
      "loss": 0.6964,
      "mean_token_accuracy": 0.7893833927810192,
      "num_tokens": 171980334.0,
      "step": 5388
    },
    {
      "entropy": 0.7375257518142462,
      "epoch": 0.495822335696747,
      "grad_norm": 0.1581810861825943,
      "learning_rate": 8.347594074891895e-05,
      "loss": 0.7267,
      "mean_token_accuracy": 0.7860930413007736,
      "num_tokens": 172012177.0,
      "step": 5389
    },
    {
      "entropy": 0.7323577962815762,
      "epoch": 0.49591434206818824,
      "grad_norm": 0.15700648725032806,
      "learning_rate": 8.347287392277731e-05,
      "loss": 0.7445,
      "mean_token_accuracy": 0.7813149653375149,
      "num_tokens": 172043675.0,
      "step": 5390
    },
    {
      "entropy": 0.7281894981861115,
      "epoch": 0.49600634843962943,
      "grad_norm": 0.16019320487976074,
      "learning_rate": 8.34698070966357e-05,
      "loss": 0.7387,
      "mean_token_accuracy": 0.7840186506509781,
      "num_tokens": 172074759.0,
      "step": 5391
    },
    {
      "entropy": 0.753687534481287,
      "epoch": 0.4960983548110707,
      "grad_norm": 0.16137902438640594,
      "learning_rate": 8.346674027049406e-05,
      "loss": 0.7548,
      "mean_token_accuracy": 0.7749973647296429,
      "num_tokens": 172106755.0,
      "step": 5392
    },
    {
      "entropy": 0.8238194826990366,
      "epoch": 0.49619036118251186,
      "grad_norm": 0.16655398905277252,
      "learning_rate": 8.346367344435245e-05,
      "loss": 0.8184,
      "mean_token_accuracy": 0.764345608651638,
      "num_tokens": 172139338.0,
      "step": 5393
    },
    {
      "entropy": 0.763019097968936,
      "epoch": 0.4962823675539531,
      "grad_norm": 0.16410624980926514,
      "learning_rate": 8.346060661821081e-05,
      "loss": 0.7716,
      "mean_token_accuracy": 0.7798589691519737,
      "num_tokens": 172171595.0,
      "step": 5394
    },
    {
      "entropy": 0.8207440748810768,
      "epoch": 0.49637437392539435,
      "grad_norm": 0.1531534641981125,
      "learning_rate": 8.34575397920692e-05,
      "loss": 0.79,
      "mean_token_accuracy": 0.766737949103117,
      "num_tokens": 172202660.0,
      "step": 5395
    },
    {
      "entropy": 0.7773783560842276,
      "epoch": 0.49646638029683554,
      "grad_norm": 0.16023550927639008,
      "learning_rate": 8.345447296592756e-05,
      "loss": 0.764,
      "mean_token_accuracy": 0.77473034709692,
      "num_tokens": 172234488.0,
      "step": 5396
    },
    {
      "entropy": 0.8350295312702656,
      "epoch": 0.4965583866682768,
      "grad_norm": 0.15937772393226624,
      "learning_rate": 8.345140613978593e-05,
      "loss": 0.8277,
      "mean_token_accuracy": 0.7658395729959011,
      "num_tokens": 172267157.0,
      "step": 5397
    },
    {
      "entropy": 0.7861018404364586,
      "epoch": 0.496650393039718,
      "grad_norm": 0.16195246577262878,
      "learning_rate": 8.344833931364431e-05,
      "loss": 0.7935,
      "mean_token_accuracy": 0.7721988186240196,
      "num_tokens": 172299021.0,
      "step": 5398
    },
    {
      "entropy": 0.7763461377471685,
      "epoch": 0.4967423994111592,
      "grad_norm": 0.16263467073440552,
      "learning_rate": 8.344527248750269e-05,
      "loss": 0.7683,
      "mean_token_accuracy": 0.7754193507134914,
      "num_tokens": 172331475.0,
      "step": 5399
    },
    {
      "entropy": 0.7207995913922787,
      "epoch": 0.49683440578260046,
      "grad_norm": 0.14975948631763458,
      "learning_rate": 8.344220566136106e-05,
      "loss": 0.6963,
      "mean_token_accuracy": 0.7945947051048279,
      "num_tokens": 172363877.0,
      "step": 5400
    },
    {
      "entropy": 0.7256605383008718,
      "epoch": 0.49692641215404165,
      "grad_norm": 0.1543005406856537,
      "learning_rate": 8.343913883521943e-05,
      "loss": 0.6833,
      "mean_token_accuracy": 0.7969504967331886,
      "num_tokens": 172396384.0,
      "step": 5401
    },
    {
      "entropy": 0.71750583127141,
      "epoch": 0.4970184185254829,
      "grad_norm": 0.15180343389511108,
      "learning_rate": 8.343607200907781e-05,
      "loss": 0.6861,
      "mean_token_accuracy": 0.8034474849700928,
      "num_tokens": 172428063.0,
      "step": 5402
    },
    {
      "entropy": 0.7514093983918428,
      "epoch": 0.4971104248969241,
      "grad_norm": 0.16980807483196259,
      "learning_rate": 8.343300518293619e-05,
      "loss": 0.7624,
      "mean_token_accuracy": 0.7768721543252468,
      "num_tokens": 172458637.0,
      "step": 5403
    },
    {
      "entropy": 0.7915403638035059,
      "epoch": 0.49720243126836533,
      "grad_norm": 0.1665991246700287,
      "learning_rate": 8.342993835679456e-05,
      "loss": 0.8177,
      "mean_token_accuracy": 0.7653665542602539,
      "num_tokens": 172490809.0,
      "step": 5404
    },
    {
      "entropy": 0.768559530377388,
      "epoch": 0.4972944376398066,
      "grad_norm": 0.1577337086200714,
      "learning_rate": 8.342687153065293e-05,
      "loss": 0.7802,
      "mean_token_accuracy": 0.7698793672025204,
      "num_tokens": 172523024.0,
      "step": 5405
    },
    {
      "entropy": 0.7885108282789588,
      "epoch": 0.49738644401124776,
      "grad_norm": 0.15705914795398712,
      "learning_rate": 8.342380470451131e-05,
      "loss": 0.7825,
      "mean_token_accuracy": 0.7710266523063183,
      "num_tokens": 172554648.0,
      "step": 5406
    },
    {
      "entropy": 0.750606544315815,
      "epoch": 0.497478450382689,
      "grad_norm": 0.15456226468086243,
      "learning_rate": 8.342073787836968e-05,
      "loss": 0.7413,
      "mean_token_accuracy": 0.7849370241165161,
      "num_tokens": 172586828.0,
      "step": 5407
    },
    {
      "entropy": 0.8338965512812138,
      "epoch": 0.4975704567541302,
      "grad_norm": 0.15957340598106384,
      "learning_rate": 8.341767105222806e-05,
      "loss": 0.7957,
      "mean_token_accuracy": 0.7687089182436466,
      "num_tokens": 172618236.0,
      "step": 5408
    },
    {
      "entropy": 0.6804021876305342,
      "epoch": 0.49766246312557144,
      "grad_norm": 0.16183984279632568,
      "learning_rate": 8.341460422608643e-05,
      "loss": 0.6577,
      "mean_token_accuracy": 0.8023249953985214,
      "num_tokens": 172650019.0,
      "step": 5409
    },
    {
      "entropy": 0.7590230591595173,
      "epoch": 0.4977544694970127,
      "grad_norm": 0.1660926342010498,
      "learning_rate": 8.341153739994481e-05,
      "loss": 0.755,
      "mean_token_accuracy": 0.7795537784695625,
      "num_tokens": 172681979.0,
      "step": 5410
    },
    {
      "entropy": 0.6825865767896175,
      "epoch": 0.4978464758684539,
      "grad_norm": 0.16134583950042725,
      "learning_rate": 8.340847057380318e-05,
      "loss": 0.6928,
      "mean_token_accuracy": 0.79880141466856,
      "num_tokens": 172714119.0,
      "step": 5411
    },
    {
      "entropy": 0.7243633158504963,
      "epoch": 0.4979384822398951,
      "grad_norm": 0.15393084287643433,
      "learning_rate": 8.340540374766154e-05,
      "loss": 0.696,
      "mean_token_accuracy": 0.7958606071770191,
      "num_tokens": 172746151.0,
      "step": 5412
    },
    {
      "entropy": 0.7677938174456358,
      "epoch": 0.4980304886113363,
      "grad_norm": 0.14916303753852844,
      "learning_rate": 8.340233692151992e-05,
      "loss": 0.7557,
      "mean_token_accuracy": 0.7786747924983501,
      "num_tokens": 172778525.0,
      "step": 5413
    },
    {
      "entropy": 0.74062036909163,
      "epoch": 0.49812249498277755,
      "grad_norm": 0.14797736704349518,
      "learning_rate": 8.33992700953783e-05,
      "loss": 0.7238,
      "mean_token_accuracy": 0.7830693200230598,
      "num_tokens": 172810892.0,
      "step": 5414
    },
    {
      "entropy": 0.6605751905590296,
      "epoch": 0.4982145013542188,
      "grad_norm": 0.1462940126657486,
      "learning_rate": 8.339620326923667e-05,
      "loss": 0.6734,
      "mean_token_accuracy": 0.802493616938591,
      "num_tokens": 172843072.0,
      "step": 5415
    },
    {
      "entropy": 0.7517594955861568,
      "epoch": 0.49830650772566,
      "grad_norm": 0.1554582417011261,
      "learning_rate": 8.339313644309504e-05,
      "loss": 0.7362,
      "mean_token_accuracy": 0.7836536429822445,
      "num_tokens": 172875213.0,
      "step": 5416
    },
    {
      "entropy": 0.8592077791690826,
      "epoch": 0.49839851409710123,
      "grad_norm": 0.16940149664878845,
      "learning_rate": 8.339006961695341e-05,
      "loss": 0.8666,
      "mean_token_accuracy": 0.7567649595439434,
      "num_tokens": 172907366.0,
      "step": 5417
    },
    {
      "entropy": 0.6859346125274897,
      "epoch": 0.4984905204685424,
      "grad_norm": 0.1490919142961502,
      "learning_rate": 8.338700279081179e-05,
      "loss": 0.6599,
      "mean_token_accuracy": 0.8086614646017551,
      "num_tokens": 172939469.0,
      "step": 5418
    },
    {
      "entropy": 0.8711606822907925,
      "epoch": 0.49858252683998366,
      "grad_norm": 0.16334599256515503,
      "learning_rate": 8.338393596467017e-05,
      "loss": 0.8735,
      "mean_token_accuracy": 0.7453362718224525,
      "num_tokens": 172970881.0,
      "step": 5419
    },
    {
      "entropy": 0.771410383284092,
      "epoch": 0.4986745332114249,
      "grad_norm": 0.16147834062576294,
      "learning_rate": 8.338086913852854e-05,
      "loss": 0.7634,
      "mean_token_accuracy": 0.7761236913502216,
      "num_tokens": 173001857.0,
      "step": 5420
    },
    {
      "entropy": 0.791310952976346,
      "epoch": 0.4987665395828661,
      "grad_norm": 0.15745756030082703,
      "learning_rate": 8.337780231238691e-05,
      "loss": 0.7898,
      "mean_token_accuracy": 0.7681823298335075,
      "num_tokens": 173033546.0,
      "step": 5421
    },
    {
      "entropy": 0.703075934201479,
      "epoch": 0.49885854595430734,
      "grad_norm": 0.16616831719875336,
      "learning_rate": 8.337473548624529e-05,
      "loss": 0.6985,
      "mean_token_accuracy": 0.7942847684025764,
      "num_tokens": 173066253.0,
      "step": 5422
    },
    {
      "entropy": 0.7809673622250557,
      "epoch": 0.49895055232574853,
      "grad_norm": 0.16361859440803528,
      "learning_rate": 8.337166866010366e-05,
      "loss": 0.7758,
      "mean_token_accuracy": 0.7704659029841423,
      "num_tokens": 173097919.0,
      "step": 5423
    },
    {
      "entropy": 0.8386838026344776,
      "epoch": 0.4990425586971898,
      "grad_norm": 0.16567979753017426,
      "learning_rate": 8.336860183396204e-05,
      "loss": 0.8336,
      "mean_token_accuracy": 0.7558126375079155,
      "num_tokens": 173129896.0,
      "step": 5424
    },
    {
      "entropy": 0.8547528106719255,
      "epoch": 0.499134565068631,
      "grad_norm": 0.16359993815422058,
      "learning_rate": 8.336553500782041e-05,
      "loss": 0.8292,
      "mean_token_accuracy": 0.7607197389006615,
      "num_tokens": 173161546.0,
      "step": 5425
    },
    {
      "entropy": 0.8487483859062195,
      "epoch": 0.4992265714400722,
      "grad_norm": 0.18711784482002258,
      "learning_rate": 8.336246818167879e-05,
      "loss": 0.854,
      "mean_token_accuracy": 0.7532676495611668,
      "num_tokens": 173192965.0,
      "step": 5426
    },
    {
      "entropy": 0.75078559294343,
      "epoch": 0.49931857781151345,
      "grad_norm": 0.16480225324630737,
      "learning_rate": 8.335940135553716e-05,
      "loss": 0.7523,
      "mean_token_accuracy": 0.7829437181353569,
      "num_tokens": 173224407.0,
      "step": 5427
    },
    {
      "entropy": 0.7175601553171873,
      "epoch": 0.49941058418295464,
      "grad_norm": 0.15829740464687347,
      "learning_rate": 8.335633452939552e-05,
      "loss": 0.7044,
      "mean_token_accuracy": 0.7952868677675724,
      "num_tokens": 173256601.0,
      "step": 5428
    },
    {
      "entropy": 0.6934023536741734,
      "epoch": 0.4995025905543959,
      "grad_norm": 0.15341022610664368,
      "learning_rate": 8.33532677032539e-05,
      "loss": 0.6758,
      "mean_token_accuracy": 0.7967083528637886,
      "num_tokens": 173288067.0,
      "step": 5429
    },
    {
      "entropy": 0.7797111375257373,
      "epoch": 0.49959459692583713,
      "grad_norm": 0.1558935046195984,
      "learning_rate": 8.335020087711229e-05,
      "loss": 0.7673,
      "mean_token_accuracy": 0.7771652191877365,
      "num_tokens": 173319740.0,
      "step": 5430
    },
    {
      "entropy": 0.8359079360961914,
      "epoch": 0.4996866032972783,
      "grad_norm": 0.1605812907218933,
      "learning_rate": 8.334713405097065e-05,
      "loss": 0.8008,
      "mean_token_accuracy": 0.768160916864872,
      "num_tokens": 173351631.0,
      "step": 5431
    },
    {
      "entropy": 0.8369604106992483,
      "epoch": 0.49977860966871956,
      "grad_norm": 0.15950432419776917,
      "learning_rate": 8.334406722482902e-05,
      "loss": 0.8235,
      "mean_token_accuracy": 0.7620037011802197,
      "num_tokens": 173383650.0,
      "step": 5432
    },
    {
      "entropy": 0.7693653516471386,
      "epoch": 0.49987061604016075,
      "grad_norm": 0.15215753018856049,
      "learning_rate": 8.33410003986874e-05,
      "loss": 0.7499,
      "mean_token_accuracy": 0.7836516499519348,
      "num_tokens": 173415498.0,
      "step": 5433
    },
    {
      "entropy": 0.7965945266187191,
      "epoch": 0.499962622411602,
      "grad_norm": 0.15906798839569092,
      "learning_rate": 8.333793357254579e-05,
      "loss": 0.7914,
      "mean_token_accuracy": 0.7677048295736313,
      "num_tokens": 173447948.0,
      "step": 5434
    },
    {
      "entropy": 0.8321841917932034,
      "epoch": 0.5000546287830432,
      "grad_norm": 0.17167828977108002,
      "learning_rate": 8.333486674640415e-05,
      "loss": 0.8432,
      "mean_token_accuracy": 0.7605916894972324,
      "num_tokens": 173479391.0,
      "step": 5435
    },
    {
      "entropy": 0.8165316469967365,
      "epoch": 0.5001466351544844,
      "grad_norm": 0.17919710278511047,
      "learning_rate": 8.333179992026252e-05,
      "loss": 0.8225,
      "mean_token_accuracy": 0.7625509537756443,
      "num_tokens": 173510424.0,
      "step": 5436
    },
    {
      "entropy": 0.783446529880166,
      "epoch": 0.5002386415259257,
      "grad_norm": 0.1580132246017456,
      "learning_rate": 8.33287330941209e-05,
      "loss": 0.7678,
      "mean_token_accuracy": 0.7797777205705643,
      "num_tokens": 173542603.0,
      "step": 5437
    },
    {
      "entropy": 0.827321657910943,
      "epoch": 0.5003306478973669,
      "grad_norm": 0.16230791807174683,
      "learning_rate": 8.332566626797927e-05,
      "loss": 0.8209,
      "mean_token_accuracy": 0.7597803436219692,
      "num_tokens": 173574615.0,
      "step": 5438
    },
    {
      "entropy": 0.7708566728979349,
      "epoch": 0.5004226542688082,
      "grad_norm": 0.1564946174621582,
      "learning_rate": 8.332259944183765e-05,
      "loss": 0.7362,
      "mean_token_accuracy": 0.7879243344068527,
      "num_tokens": 173606905.0,
      "step": 5439
    },
    {
      "entropy": 0.8045233711600304,
      "epoch": 0.5005146606402493,
      "grad_norm": 0.15192078053951263,
      "learning_rate": 8.331953261569602e-05,
      "loss": 0.7882,
      "mean_token_accuracy": 0.7687460146844387,
      "num_tokens": 173638912.0,
      "step": 5440
    },
    {
      "entropy": 0.8032363392412663,
      "epoch": 0.5006066670116905,
      "grad_norm": 0.15296930074691772,
      "learning_rate": 8.33164657895544e-05,
      "loss": 0.7854,
      "mean_token_accuracy": 0.7697148770093918,
      "num_tokens": 173671436.0,
      "step": 5441
    },
    {
      "entropy": 0.7655646670609713,
      "epoch": 0.5006986733831318,
      "grad_norm": 0.16380882263183594,
      "learning_rate": 8.331339896341277e-05,
      "loss": 0.7611,
      "mean_token_accuracy": 0.7799613662064075,
      "num_tokens": 173703657.0,
      "step": 5442
    },
    {
      "entropy": 0.7249550744891167,
      "epoch": 0.500790679754573,
      "grad_norm": 0.1578320860862732,
      "learning_rate": 8.331033213727114e-05,
      "loss": 0.6966,
      "mean_token_accuracy": 0.7913169749081135,
      "num_tokens": 173735497.0,
      "step": 5443
    },
    {
      "entropy": 0.8537628650665283,
      "epoch": 0.5008826861260143,
      "grad_norm": 0.16705453395843506,
      "learning_rate": 8.330726531112952e-05,
      "loss": 0.8446,
      "mean_token_accuracy": 0.7569867372512817,
      "num_tokens": 173768201.0,
      "step": 5444
    },
    {
      "entropy": 0.714986439794302,
      "epoch": 0.5009746924974554,
      "grad_norm": 0.15698587894439697,
      "learning_rate": 8.33041984849879e-05,
      "loss": 0.6988,
      "mean_token_accuracy": 0.7955541461706161,
      "num_tokens": 173800627.0,
      "step": 5445
    },
    {
      "entropy": 0.8453456424176693,
      "epoch": 0.5010666988688967,
      "grad_norm": 0.15917101502418518,
      "learning_rate": 8.330113165884627e-05,
      "loss": 0.8274,
      "mean_token_accuracy": 0.7602743022143841,
      "num_tokens": 173832841.0,
      "step": 5446
    },
    {
      "entropy": 0.731022547930479,
      "epoch": 0.5011587052403379,
      "grad_norm": 0.14856143295764923,
      "learning_rate": 8.329806483270464e-05,
      "loss": 0.7188,
      "mean_token_accuracy": 0.7893058955669403,
      "num_tokens": 173864713.0,
      "step": 5447
    },
    {
      "entropy": 0.723038200289011,
      "epoch": 0.5012507116117791,
      "grad_norm": 0.15102088451385498,
      "learning_rate": 8.3294998006563e-05,
      "loss": 0.7132,
      "mean_token_accuracy": 0.7902619168162346,
      "num_tokens": 173896914.0,
      "step": 5448
    },
    {
      "entropy": 0.7693786881864071,
      "epoch": 0.5013427179832204,
      "grad_norm": 0.15411004424095154,
      "learning_rate": 8.329193118042138e-05,
      "loss": 0.7844,
      "mean_token_accuracy": 0.768676646053791,
      "num_tokens": 173928815.0,
      "step": 5449
    },
    {
      "entropy": 0.7019230239093304,
      "epoch": 0.5014347243546615,
      "grad_norm": 0.1578516811132431,
      "learning_rate": 8.328886435427977e-05,
      "loss": 0.7115,
      "mean_token_accuracy": 0.7899069860577583,
      "num_tokens": 173961165.0,
      "step": 5450
    },
    {
      "entropy": 0.7533403113484383,
      "epoch": 0.5015267307261028,
      "grad_norm": 0.16825105249881744,
      "learning_rate": 8.328579752813813e-05,
      "loss": 0.7577,
      "mean_token_accuracy": 0.7798524498939514,
      "num_tokens": 173992801.0,
      "step": 5451
    },
    {
      "entropy": 0.7329448685050011,
      "epoch": 0.501618737097544,
      "grad_norm": 0.1568758338689804,
      "learning_rate": 8.32827307019965e-05,
      "loss": 0.7101,
      "mean_token_accuracy": 0.7916787303984165,
      "num_tokens": 174023758.0,
      "step": 5452
    },
    {
      "entropy": 0.748884386382997,
      "epoch": 0.5017107434689853,
      "grad_norm": 0.1609378308057785,
      "learning_rate": 8.327966387585488e-05,
      "loss": 0.7289,
      "mean_token_accuracy": 0.7862621732056141,
      "num_tokens": 174055270.0,
      "step": 5453
    },
    {
      "entropy": 0.8652287255972624,
      "epoch": 0.5018027498404265,
      "grad_norm": 0.16038835048675537,
      "learning_rate": 8.327659704971325e-05,
      "loss": 0.8711,
      "mean_token_accuracy": 0.7475010380148888,
      "num_tokens": 174086513.0,
      "step": 5454
    },
    {
      "entropy": 0.792159017175436,
      "epoch": 0.5018947562118676,
      "grad_norm": 0.16817092895507812,
      "learning_rate": 8.327353022357163e-05,
      "loss": 0.7968,
      "mean_token_accuracy": 0.7681858576834202,
      "num_tokens": 174118723.0,
      "step": 5455
    },
    {
      "entropy": 0.7931517623364925,
      "epoch": 0.5019867625833089,
      "grad_norm": 0.15109317004680634,
      "learning_rate": 8.327046339743e-05,
      "loss": 0.7776,
      "mean_token_accuracy": 0.7711290381848812,
      "num_tokens": 174151097.0,
      "step": 5456
    },
    {
      "entropy": 0.7876989245414734,
      "epoch": 0.5020787689547501,
      "grad_norm": 0.1588999629020691,
      "learning_rate": 8.326739657128838e-05,
      "loss": 0.7714,
      "mean_token_accuracy": 0.7725328132510185,
      "num_tokens": 174182793.0,
      "step": 5457
    },
    {
      "entropy": 0.8066707383841276,
      "epoch": 0.5021707753261914,
      "grad_norm": 0.1558055579662323,
      "learning_rate": 8.326432974514675e-05,
      "loss": 0.7912,
      "mean_token_accuracy": 0.77005984634161,
      "num_tokens": 174215561.0,
      "step": 5458
    },
    {
      "entropy": 0.8668371625244617,
      "epoch": 0.5022627816976326,
      "grad_norm": 0.17107869684696198,
      "learning_rate": 8.326126291900512e-05,
      "loss": 0.8543,
      "mean_token_accuracy": 0.7572713010013103,
      "num_tokens": 174247327.0,
      "step": 5459
    },
    {
      "entropy": 0.7310237158089876,
      "epoch": 0.5023547880690737,
      "grad_norm": 0.15770605206489563,
      "learning_rate": 8.32581960928635e-05,
      "loss": 0.7116,
      "mean_token_accuracy": 0.7911750599741936,
      "num_tokens": 174279650.0,
      "step": 5460
    },
    {
      "entropy": 0.8546728231012821,
      "epoch": 0.502446794440515,
      "grad_norm": 0.15958642959594727,
      "learning_rate": 8.325512926672188e-05,
      "loss": 0.8347,
      "mean_token_accuracy": 0.755575992166996,
      "num_tokens": 174311440.0,
      "step": 5461
    },
    {
      "entropy": 0.8476290125399828,
      "epoch": 0.5025388008119562,
      "grad_norm": 0.15710870921611786,
      "learning_rate": 8.325206244058025e-05,
      "loss": 0.8564,
      "mean_token_accuracy": 0.7551259882748127,
      "num_tokens": 174343946.0,
      "step": 5462
    },
    {
      "entropy": 0.6886413358151913,
      "epoch": 0.5026308071833975,
      "grad_norm": 0.1527508795261383,
      "learning_rate": 8.324899561443862e-05,
      "loss": 0.6651,
      "mean_token_accuracy": 0.8057761117815971,
      "num_tokens": 174375566.0,
      "step": 5463
    },
    {
      "entropy": 0.7198618315160275,
      "epoch": 0.5027228135548387,
      "grad_norm": 0.16014787554740906,
      "learning_rate": 8.3245928788297e-05,
      "loss": 0.7125,
      "mean_token_accuracy": 0.7849789261817932,
      "num_tokens": 174408152.0,
      "step": 5464
    },
    {
      "entropy": 0.8484123609960079,
      "epoch": 0.5028148199262799,
      "grad_norm": 0.1637634038925171,
      "learning_rate": 8.324286196215537e-05,
      "loss": 0.8533,
      "mean_token_accuracy": 0.7504561208188534,
      "num_tokens": 174440415.0,
      "step": 5465
    },
    {
      "entropy": 0.7623932138085365,
      "epoch": 0.5029068262977211,
      "grad_norm": 0.15930840373039246,
      "learning_rate": 8.323979513601375e-05,
      "loss": 0.7558,
      "mean_token_accuracy": 0.7783086486160755,
      "num_tokens": 174472497.0,
      "step": 5466
    },
    {
      "entropy": 0.8460042718797922,
      "epoch": 0.5029988326691623,
      "grad_norm": 0.16827435791492462,
      "learning_rate": 8.323672830987212e-05,
      "loss": 0.8449,
      "mean_token_accuracy": 0.7524569556117058,
      "num_tokens": 174503699.0,
      "step": 5467
    },
    {
      "entropy": 0.6993048638105392,
      "epoch": 0.5030908390406036,
      "grad_norm": 0.1634429693222046,
      "learning_rate": 8.32336614837305e-05,
      "loss": 0.6766,
      "mean_token_accuracy": 0.8002345189452171,
      "num_tokens": 174535592.0,
      "step": 5468
    },
    {
      "entropy": 0.8685693964362144,
      "epoch": 0.5031828454120448,
      "grad_norm": 0.16333332657814026,
      "learning_rate": 8.323059465758886e-05,
      "loss": 0.85,
      "mean_token_accuracy": 0.752197228372097,
      "num_tokens": 174567893.0,
      "step": 5469
    },
    {
      "entropy": 0.6913311798125505,
      "epoch": 0.503274851783486,
      "grad_norm": 0.1508670151233673,
      "learning_rate": 8.322752783144723e-05,
      "loss": 0.6801,
      "mean_token_accuracy": 0.7944161705672741,
      "num_tokens": 174600048.0,
      "step": 5470
    },
    {
      "entropy": 0.7733440808951855,
      "epoch": 0.5033668581549272,
      "grad_norm": 0.16022346913814545,
      "learning_rate": 8.322446100530561e-05,
      "loss": 0.7721,
      "mean_token_accuracy": 0.77123898640275,
      "num_tokens": 174632103.0,
      "step": 5471
    },
    {
      "entropy": 0.6205370165407658,
      "epoch": 0.5034588645263685,
      "grad_norm": 0.1527452915906906,
      "learning_rate": 8.3221394179164e-05,
      "loss": 0.5967,
      "mean_token_accuracy": 0.8224657289683819,
      "num_tokens": 174663586.0,
      "step": 5472
    },
    {
      "entropy": 0.712224131450057,
      "epoch": 0.5035508708978097,
      "grad_norm": 0.1549619734287262,
      "learning_rate": 8.321832735302236e-05,
      "loss": 0.7084,
      "mean_token_accuracy": 0.7917978577315807,
      "num_tokens": 174696172.0,
      "step": 5473
    },
    {
      "entropy": 0.8061850350350142,
      "epoch": 0.5036428772692509,
      "grad_norm": 0.16314160823822021,
      "learning_rate": 8.321526052688073e-05,
      "loss": 0.7806,
      "mean_token_accuracy": 0.7719548307359219,
      "num_tokens": 174728042.0,
      "step": 5474
    },
    {
      "entropy": 0.7275212984532118,
      "epoch": 0.5037348836406921,
      "grad_norm": 0.16687346994876862,
      "learning_rate": 8.32121937007391e-05,
      "loss": 0.7257,
      "mean_token_accuracy": 0.7892970144748688,
      "num_tokens": 174759885.0,
      "step": 5475
    },
    {
      "entropy": 0.7252771779894829,
      "epoch": 0.5038268900121333,
      "grad_norm": 0.14994624257087708,
      "learning_rate": 8.32091268745975e-05,
      "loss": 0.6893,
      "mean_token_accuracy": 0.7890426255762577,
      "num_tokens": 174790707.0,
      "step": 5476
    },
    {
      "entropy": 0.8684570603072643,
      "epoch": 0.5039188963835746,
      "grad_norm": 0.16880621016025543,
      "learning_rate": 8.320606004845586e-05,
      "loss": 0.8653,
      "mean_token_accuracy": 0.7509827017784119,
      "num_tokens": 174822736.0,
      "step": 5477
    },
    {
      "entropy": 0.7341880779713392,
      "epoch": 0.5040109027550158,
      "grad_norm": 0.15764853358268738,
      "learning_rate": 8.320299322231423e-05,
      "loss": 0.714,
      "mean_token_accuracy": 0.7891422174870968,
      "num_tokens": 174854436.0,
      "step": 5478
    },
    {
      "entropy": 0.8223336655646563,
      "epoch": 0.504102909126457,
      "grad_norm": 0.16335438191890717,
      "learning_rate": 8.31999263961726e-05,
      "loss": 0.8091,
      "mean_token_accuracy": 0.7622214891016483,
      "num_tokens": 174886965.0,
      "step": 5479
    },
    {
      "entropy": 0.6825045011937618,
      "epoch": 0.5041949154978982,
      "grad_norm": 0.15394768118858337,
      "learning_rate": 8.319685957003098e-05,
      "loss": 0.6584,
      "mean_token_accuracy": 0.8082954324781895,
      "num_tokens": 174919234.0,
      "step": 5480
    },
    {
      "entropy": 0.7519105058163404,
      "epoch": 0.5042869218693394,
      "grad_norm": 0.14855188131332397,
      "learning_rate": 8.319379274388936e-05,
      "loss": 0.7448,
      "mean_token_accuracy": 0.782484881579876,
      "num_tokens": 174951244.0,
      "step": 5481
    },
    {
      "entropy": 0.9033338241279125,
      "epoch": 0.5043789282407807,
      "grad_norm": 0.16802158951759338,
      "learning_rate": 8.319072591774773e-05,
      "loss": 0.9037,
      "mean_token_accuracy": 0.7424999475479126,
      "num_tokens": 174983521.0,
      "step": 5482
    },
    {
      "entropy": 0.8157029561698437,
      "epoch": 0.5044709346122219,
      "grad_norm": 0.15882770717144012,
      "learning_rate": 8.31876590916061e-05,
      "loss": 0.8072,
      "mean_token_accuracy": 0.7630610354244709,
      "num_tokens": 175014672.0,
      "step": 5483
    },
    {
      "entropy": 0.7260832097381353,
      "epoch": 0.5045629409836632,
      "grad_norm": 0.15120424330234528,
      "learning_rate": 8.318459226546448e-05,
      "loss": 0.7145,
      "mean_token_accuracy": 0.7914022989571095,
      "num_tokens": 175046666.0,
      "step": 5484
    },
    {
      "entropy": 0.7913651838898659,
      "epoch": 0.5046549473551043,
      "grad_norm": 0.16018453240394592,
      "learning_rate": 8.318152543932285e-05,
      "loss": 0.7958,
      "mean_token_accuracy": 0.7690196856856346,
      "num_tokens": 175078032.0,
      "step": 5485
    },
    {
      "entropy": 0.6706576589494944,
      "epoch": 0.5047469537265455,
      "grad_norm": 0.15442179143428802,
      "learning_rate": 8.317845861318123e-05,
      "loss": 0.6527,
      "mean_token_accuracy": 0.8050585575401783,
      "num_tokens": 175110771.0,
      "step": 5486
    },
    {
      "entropy": 0.659272987395525,
      "epoch": 0.5048389600979868,
      "grad_norm": 0.15473723411560059,
      "learning_rate": 8.31753917870396e-05,
      "loss": 0.6558,
      "mean_token_accuracy": 0.8056056387722492,
      "num_tokens": 175143027.0,
      "step": 5487
    },
    {
      "entropy": 0.7209903001785278,
      "epoch": 0.504930966469428,
      "grad_norm": 0.16235700249671936,
      "learning_rate": 8.317232496089798e-05,
      "loss": 0.6993,
      "mean_token_accuracy": 0.7925171442329884,
      "num_tokens": 175175044.0,
      "step": 5488
    },
    {
      "entropy": 0.9283970519900322,
      "epoch": 0.5050229728408693,
      "grad_norm": 0.1702759712934494,
      "learning_rate": 8.316925813475634e-05,
      "loss": 0.9323,
      "mean_token_accuracy": 0.7376643009483814,
      "num_tokens": 175206397.0,
      "step": 5489
    },
    {
      "entropy": 0.8011667691171169,
      "epoch": 0.5051149792123104,
      "grad_norm": 0.1624530851840973,
      "learning_rate": 8.316619130861471e-05,
      "loss": 0.7866,
      "mean_token_accuracy": 0.7704578079283237,
      "num_tokens": 175238191.0,
      "step": 5490
    },
    {
      "entropy": 0.8218264970928431,
      "epoch": 0.5052069855837517,
      "grad_norm": 0.15165385603904724,
      "learning_rate": 8.316312448247309e-05,
      "loss": 0.7848,
      "mean_token_accuracy": 0.7754538357257843,
      "num_tokens": 175270345.0,
      "step": 5491
    },
    {
      "entropy": 0.7173209246248007,
      "epoch": 0.5052989919551929,
      "grad_norm": 0.15687821805477142,
      "learning_rate": 8.316005765633147e-05,
      "loss": 0.7107,
      "mean_token_accuracy": 0.7917746156454086,
      "num_tokens": 175302733.0,
      "step": 5492
    },
    {
      "entropy": 0.8604523688554764,
      "epoch": 0.5053909983266341,
      "grad_norm": 0.16326165199279785,
      "learning_rate": 8.315699083018984e-05,
      "loss": 0.8539,
      "mean_token_accuracy": 0.7515549696981907,
      "num_tokens": 175333555.0,
      "step": 5493
    },
    {
      "entropy": 0.8896865993738174,
      "epoch": 0.5054830046980754,
      "grad_norm": 0.1661350578069687,
      "learning_rate": 8.315392400404821e-05,
      "loss": 0.8877,
      "mean_token_accuracy": 0.744792990386486,
      "num_tokens": 175363948.0,
      "step": 5494
    },
    {
      "entropy": 0.8506352603435516,
      "epoch": 0.5055750110695165,
      "grad_norm": 0.15729448199272156,
      "learning_rate": 8.315085717790659e-05,
      "loss": 0.8314,
      "mean_token_accuracy": 0.759013470262289,
      "num_tokens": 175396359.0,
      "step": 5495
    },
    {
      "entropy": 0.829635314643383,
      "epoch": 0.5056670174409578,
      "grad_norm": 0.15934287011623383,
      "learning_rate": 8.314779035176496e-05,
      "loss": 0.8257,
      "mean_token_accuracy": 0.7608772106468678,
      "num_tokens": 175428860.0,
      "step": 5496
    },
    {
      "entropy": 0.8254042454063892,
      "epoch": 0.505759023812399,
      "grad_norm": 0.16357320547103882,
      "learning_rate": 8.314472352562334e-05,
      "loss": 0.8377,
      "mean_token_accuracy": 0.7558125592768192,
      "num_tokens": 175459526.0,
      "step": 5497
    },
    {
      "entropy": 0.8497141040861607,
      "epoch": 0.5058510301838403,
      "grad_norm": 0.1592991054058075,
      "learning_rate": 8.314165669948171e-05,
      "loss": 0.8414,
      "mean_token_accuracy": 0.7509512826800346,
      "num_tokens": 175491965.0,
      "step": 5498
    },
    {
      "entropy": 0.8233323059976101,
      "epoch": 0.5059430365552815,
      "grad_norm": 0.15654486417770386,
      "learning_rate": 8.313858987334009e-05,
      "loss": 0.8231,
      "mean_token_accuracy": 0.7624205946922302,
      "num_tokens": 175523144.0,
      "step": 5499
    },
    {
      "entropy": 0.7174621121957898,
      "epoch": 0.5060350429267226,
      "grad_norm": 0.15525180101394653,
      "learning_rate": 8.313552304719846e-05,
      "loss": 0.7265,
      "mean_token_accuracy": 0.787843719124794,
      "num_tokens": 175555365.0,
      "step": 5500
    },
    {
      "entropy": 0.7978239990770817,
      "epoch": 0.5061270492981639,
      "grad_norm": 0.16789203882217407,
      "learning_rate": 8.313245622105683e-05,
      "loss": 0.8106,
      "mean_token_accuracy": 0.7679928615689278,
      "num_tokens": 175587660.0,
      "step": 5501
    },
    {
      "entropy": 0.789627630263567,
      "epoch": 0.5062190556696051,
      "grad_norm": 0.16265805065631866,
      "learning_rate": 8.312938939491521e-05,
      "loss": 0.7855,
      "mean_token_accuracy": 0.7731609866023064,
      "num_tokens": 175620117.0,
      "step": 5502
    },
    {
      "entropy": 0.7870524767786264,
      "epoch": 0.5063110620410464,
      "grad_norm": 0.15362109243869781,
      "learning_rate": 8.312632256877359e-05,
      "loss": 0.7725,
      "mean_token_accuracy": 0.7721034549176693,
      "num_tokens": 175652048.0,
      "step": 5503
    },
    {
      "entropy": 0.7079046256840229,
      "epoch": 0.5064030684124876,
      "grad_norm": 0.14617502689361572,
      "learning_rate": 8.312325574263196e-05,
      "loss": 0.672,
      "mean_token_accuracy": 0.8002140074968338,
      "num_tokens": 175684351.0,
      "step": 5504
    },
    {
      "entropy": 0.6972979158163071,
      "epoch": 0.5064950747839287,
      "grad_norm": 0.15595321357250214,
      "learning_rate": 8.312018891649032e-05,
      "loss": 0.6863,
      "mean_token_accuracy": 0.8003223463892937,
      "num_tokens": 175716398.0,
      "step": 5505
    },
    {
      "entropy": 0.7678371770307422,
      "epoch": 0.50658708115537,
      "grad_norm": 0.1591530740261078,
      "learning_rate": 8.311712209034869e-05,
      "loss": 0.7336,
      "mean_token_accuracy": 0.782371561974287,
      "num_tokens": 175747958.0,
      "step": 5506
    },
    {
      "entropy": 0.8163204994052649,
      "epoch": 0.5066790875268112,
      "grad_norm": 0.1659257709980011,
      "learning_rate": 8.311405526420707e-05,
      "loss": 0.8228,
      "mean_token_accuracy": 0.761160958558321,
      "num_tokens": 175779651.0,
      "step": 5507
    },
    {
      "entropy": 0.8071562740951777,
      "epoch": 0.5067710938982525,
      "grad_norm": 0.15572895109653473,
      "learning_rate": 8.311098843806546e-05,
      "loss": 0.7843,
      "mean_token_accuracy": 0.7700122371315956,
      "num_tokens": 175810951.0,
      "step": 5508
    },
    {
      "entropy": 0.8382457233965397,
      "epoch": 0.5068631002696937,
      "grad_norm": 0.15699738264083862,
      "learning_rate": 8.310792161192382e-05,
      "loss": 0.8432,
      "mean_token_accuracy": 0.7531238980591297,
      "num_tokens": 175843417.0,
      "step": 5509
    },
    {
      "entropy": 0.6933073550462723,
      "epoch": 0.5069551066411349,
      "grad_norm": 0.14830955862998962,
      "learning_rate": 8.310485478578219e-05,
      "loss": 0.6732,
      "mean_token_accuracy": 0.7971892133355141,
      "num_tokens": 175875547.0,
      "step": 5510
    },
    {
      "entropy": 0.7201114483177662,
      "epoch": 0.5070471130125761,
      "grad_norm": 0.16632261872291565,
      "learning_rate": 8.310178795964057e-05,
      "loss": 0.715,
      "mean_token_accuracy": 0.7933923751115799,
      "num_tokens": 175907514.0,
      "step": 5511
    },
    {
      "entropy": 0.7813945021480322,
      "epoch": 0.5071391193840173,
      "grad_norm": 0.1530279517173767,
      "learning_rate": 8.309872113349894e-05,
      "loss": 0.776,
      "mean_token_accuracy": 0.7739329040050507,
      "num_tokens": 175939598.0,
      "step": 5512
    },
    {
      "entropy": 0.6786966938525438,
      "epoch": 0.5072311257554586,
      "grad_norm": 0.15147055685520172,
      "learning_rate": 8.309565430735732e-05,
      "loss": 0.6708,
      "mean_token_accuracy": 0.7970148921012878,
      "num_tokens": 175971626.0,
      "step": 5513
    },
    {
      "entropy": 0.8485359214246273,
      "epoch": 0.5073231321268998,
      "grad_norm": 0.15953569114208221,
      "learning_rate": 8.309258748121569e-05,
      "loss": 0.8456,
      "mean_token_accuracy": 0.7545517906546593,
      "num_tokens": 176004101.0,
      "step": 5514
    },
    {
      "entropy": 0.7949788942933083,
      "epoch": 0.507415138498341,
      "grad_norm": 0.16350263357162476,
      "learning_rate": 8.308952065507407e-05,
      "loss": 0.7937,
      "mean_token_accuracy": 0.771148469299078,
      "num_tokens": 176036127.0,
      "step": 5515
    },
    {
      "entropy": 0.7973892856389284,
      "epoch": 0.5075071448697822,
      "grad_norm": 0.16603946685791016,
      "learning_rate": 8.308645382893244e-05,
      "loss": 0.7882,
      "mean_token_accuracy": 0.770672932267189,
      "num_tokens": 176067852.0,
      "step": 5516
    },
    {
      "entropy": 0.8583617620170116,
      "epoch": 0.5075991512412235,
      "grad_norm": 0.16477489471435547,
      "learning_rate": 8.30833870027908e-05,
      "loss": 0.8477,
      "mean_token_accuracy": 0.7572463415563107,
      "num_tokens": 176100616.0,
      "step": 5517
    },
    {
      "entropy": 0.7776118330657482,
      "epoch": 0.5076911576126647,
      "grad_norm": 0.1527763158082962,
      "learning_rate": 8.308032017664919e-05,
      "loss": 0.7845,
      "mean_token_accuracy": 0.7681157290935516,
      "num_tokens": 176132792.0,
      "step": 5518
    },
    {
      "entropy": 0.7247757278382778,
      "epoch": 0.507783163984106,
      "grad_norm": 0.15946267545223236,
      "learning_rate": 8.307725335050757e-05,
      "loss": 0.7095,
      "mean_token_accuracy": 0.7921487614512444,
      "num_tokens": 176164985.0,
      "step": 5519
    },
    {
      "entropy": 0.8771599940955639,
      "epoch": 0.5078751703555471,
      "grad_norm": 0.16367687284946442,
      "learning_rate": 8.307418652436594e-05,
      "loss": 0.8762,
      "mean_token_accuracy": 0.7480034604668617,
      "num_tokens": 176197606.0,
      "step": 5520
    },
    {
      "entropy": 0.7251776531338692,
      "epoch": 0.5079671767269883,
      "grad_norm": 0.15520718693733215,
      "learning_rate": 8.30711196982243e-05,
      "loss": 0.6954,
      "mean_token_accuracy": 0.7998888678848743,
      "num_tokens": 176229341.0,
      "step": 5521
    },
    {
      "entropy": 0.649249317124486,
      "epoch": 0.5080591830984296,
      "grad_norm": 0.16396181285381317,
      "learning_rate": 8.306805287208269e-05,
      "loss": 0.6417,
      "mean_token_accuracy": 0.8104956857860088,
      "num_tokens": 176261427.0,
      "step": 5522
    },
    {
      "entropy": 0.8459358792752028,
      "epoch": 0.5081511894698708,
      "grad_norm": 0.15811583399772644,
      "learning_rate": 8.306498604594107e-05,
      "loss": 0.8339,
      "mean_token_accuracy": 0.7572004422545433,
      "num_tokens": 176293703.0,
      "step": 5523
    },
    {
      "entropy": 0.7392279580235481,
      "epoch": 0.5082431958413121,
      "grad_norm": 0.15880708396434784,
      "learning_rate": 8.306191921979944e-05,
      "loss": 0.7132,
      "mean_token_accuracy": 0.7897025719285011,
      "num_tokens": 176325835.0,
      "step": 5524
    },
    {
      "entropy": 0.8143600802868605,
      "epoch": 0.5083352022127532,
      "grad_norm": 0.1635875254869461,
      "learning_rate": 8.30588523936578e-05,
      "loss": 0.8155,
      "mean_token_accuracy": 0.7650389038026333,
      "num_tokens": 176357713.0,
      "step": 5525
    },
    {
      "entropy": 0.8350079394876957,
      "epoch": 0.5084272085841944,
      "grad_norm": 0.16685576736927032,
      "learning_rate": 8.305578556751619e-05,
      "loss": 0.8316,
      "mean_token_accuracy": 0.7599197998642921,
      "num_tokens": 176389058.0,
      "step": 5526
    },
    {
      "entropy": 0.7092581149190664,
      "epoch": 0.5085192149556357,
      "grad_norm": 0.15565289556980133,
      "learning_rate": 8.305271874137455e-05,
      "loss": 0.703,
      "mean_token_accuracy": 0.7932594642043114,
      "num_tokens": 176420692.0,
      "step": 5527
    },
    {
      "entropy": 0.7565216962248087,
      "epoch": 0.5086112213270769,
      "grad_norm": 0.15996073186397552,
      "learning_rate": 8.304965191523293e-05,
      "loss": 0.7377,
      "mean_token_accuracy": 0.7830419279634953,
      "num_tokens": 176451477.0,
      "step": 5528
    },
    {
      "entropy": 0.7763148043304682,
      "epoch": 0.5087032276985182,
      "grad_norm": 0.16250579059123993,
      "learning_rate": 8.30465850890913e-05,
      "loss": 0.7369,
      "mean_token_accuracy": 0.7838722765445709,
      "num_tokens": 176483573.0,
      "step": 5529
    },
    {
      "entropy": 0.7382165491580963,
      "epoch": 0.5087952340699593,
      "grad_norm": 0.15387144684791565,
      "learning_rate": 8.304351826294968e-05,
      "loss": 0.7264,
      "mean_token_accuracy": 0.7853610292077065,
      "num_tokens": 176515722.0,
      "step": 5530
    },
    {
      "entropy": 0.7636826001107693,
      "epoch": 0.5088872404414005,
      "grad_norm": 0.15445475280284882,
      "learning_rate": 8.304045143680805e-05,
      "loss": 0.7591,
      "mean_token_accuracy": 0.7762582786381245,
      "num_tokens": 176548467.0,
      "step": 5531
    },
    {
      "entropy": 0.8285907711833715,
      "epoch": 0.5089792468128418,
      "grad_norm": 0.1682598888874054,
      "learning_rate": 8.303738461066642e-05,
      "loss": 0.8087,
      "mean_token_accuracy": 0.7652899362146854,
      "num_tokens": 176579459.0,
      "step": 5532
    },
    {
      "entropy": 0.8641696497797966,
      "epoch": 0.509071253184283,
      "grad_norm": 0.15965542197227478,
      "learning_rate": 8.30343177845248e-05,
      "loss": 0.8448,
      "mean_token_accuracy": 0.7518090195953846,
      "num_tokens": 176611408.0,
      "step": 5533
    },
    {
      "entropy": 0.6827845126390457,
      "epoch": 0.5091632595557243,
      "grad_norm": 0.1561804562807083,
      "learning_rate": 8.303125095838318e-05,
      "loss": 0.6842,
      "mean_token_accuracy": 0.8003296852111816,
      "num_tokens": 176643029.0,
      "step": 5534
    },
    {
      "entropy": 0.8176420349627733,
      "epoch": 0.5092552659271654,
      "grad_norm": 0.15994824469089508,
      "learning_rate": 8.302818413224155e-05,
      "loss": 0.8185,
      "mean_token_accuracy": 0.7648954391479492,
      "num_tokens": 176675230.0,
      "step": 5535
    },
    {
      "entropy": 0.8838024083524942,
      "epoch": 0.5093472722986067,
      "grad_norm": 0.16821491718292236,
      "learning_rate": 8.302511730609992e-05,
      "loss": 0.8692,
      "mean_token_accuracy": 0.7531268671154976,
      "num_tokens": 176706287.0,
      "step": 5536
    },
    {
      "entropy": 0.7119543291628361,
      "epoch": 0.5094392786700479,
      "grad_norm": 0.15352053940296173,
      "learning_rate": 8.302205047995829e-05,
      "loss": 0.702,
      "mean_token_accuracy": 0.7956056110560894,
      "num_tokens": 176737970.0,
      "step": 5537
    },
    {
      "entropy": 0.9047817662358284,
      "epoch": 0.5095312850414891,
      "grad_norm": 0.16015635430812836,
      "learning_rate": 8.301898365381667e-05,
      "loss": 0.9024,
      "mean_token_accuracy": 0.7417936995625496,
      "num_tokens": 176770505.0,
      "step": 5538
    },
    {
      "entropy": 0.7850410230457783,
      "epoch": 0.5096232914129304,
      "grad_norm": 0.1675969660282135,
      "learning_rate": 8.301591682767505e-05,
      "loss": 0.7791,
      "mean_token_accuracy": 0.7742532789707184,
      "num_tokens": 176802484.0,
      "step": 5539
    },
    {
      "entropy": 0.9063535146415234,
      "epoch": 0.5097152977843715,
      "grad_norm": 0.16024494171142578,
      "learning_rate": 8.301285000153342e-05,
      "loss": 0.893,
      "mean_token_accuracy": 0.741123616695404,
      "num_tokens": 176834510.0,
      "step": 5540
    },
    {
      "entropy": 0.8090183828026056,
      "epoch": 0.5098073041558128,
      "grad_norm": 0.16785728931427002,
      "learning_rate": 8.300978317539178e-05,
      "loss": 0.8085,
      "mean_token_accuracy": 0.7652606777846813,
      "num_tokens": 176866365.0,
      "step": 5541
    },
    {
      "entropy": 0.7362632323056459,
      "epoch": 0.509899310527254,
      "grad_norm": 0.1466924101114273,
      "learning_rate": 8.300671634925017e-05,
      "loss": 0.7167,
      "mean_token_accuracy": 0.7852442488074303,
      "num_tokens": 176898446.0,
      "step": 5542
    },
    {
      "entropy": 0.9371532835066319,
      "epoch": 0.5099913168986953,
      "grad_norm": 0.16720086336135864,
      "learning_rate": 8.300364952310853e-05,
      "loss": 0.9265,
      "mean_token_accuracy": 0.7347133420407772,
      "num_tokens": 176930362.0,
      "step": 5543
    },
    {
      "entropy": 0.6827549040317535,
      "epoch": 0.5100833232701365,
      "grad_norm": 0.1618349254131317,
      "learning_rate": 8.300058269696692e-05,
      "loss": 0.6642,
      "mean_token_accuracy": 0.800431240350008,
      "num_tokens": 176962269.0,
      "step": 5544
    },
    {
      "entropy": 0.7027998585253954,
      "epoch": 0.5101753296415776,
      "grad_norm": 0.15820544958114624,
      "learning_rate": 8.299751587082528e-05,
      "loss": 0.6938,
      "mean_token_accuracy": 0.7953569814562798,
      "num_tokens": 176993929.0,
      "step": 5545
    },
    {
      "entropy": 0.731325825676322,
      "epoch": 0.5102673360130189,
      "grad_norm": 0.15065163373947144,
      "learning_rate": 8.299444904468366e-05,
      "loss": 0.7081,
      "mean_token_accuracy": 0.7916214428842068,
      "num_tokens": 177026588.0,
      "step": 5546
    },
    {
      "entropy": 0.6671143863350153,
      "epoch": 0.5103593423844601,
      "grad_norm": 0.1547168642282486,
      "learning_rate": 8.299138221854203e-05,
      "loss": 0.6656,
      "mean_token_accuracy": 0.8049458600580692,
      "num_tokens": 177058380.0,
      "step": 5547
    },
    {
      "entropy": 0.8802124932408333,
      "epoch": 0.5104513487559014,
      "grad_norm": 0.16162732243537903,
      "learning_rate": 8.29883153924004e-05,
      "loss": 0.8552,
      "mean_token_accuracy": 0.7567197978496552,
      "num_tokens": 177090577.0,
      "step": 5548
    },
    {
      "entropy": 0.7440163195133209,
      "epoch": 0.5105433551273426,
      "grad_norm": 0.15546099841594696,
      "learning_rate": 8.298524856625878e-05,
      "loss": 0.7299,
      "mean_token_accuracy": 0.7861608974635601,
      "num_tokens": 177121949.0,
      "step": 5549
    },
    {
      "entropy": 0.8086006678640842,
      "epoch": 0.5106353614987837,
      "grad_norm": 0.16184239089488983,
      "learning_rate": 8.298218174011716e-05,
      "loss": 0.7851,
      "mean_token_accuracy": 0.7712105326354504,
      "num_tokens": 177154131.0,
      "step": 5550
    },
    {
      "entropy": 0.7903521843254566,
      "epoch": 0.510727367870225,
      "grad_norm": 0.16253960132598877,
      "learning_rate": 8.297911491397553e-05,
      "loss": 0.7681,
      "mean_token_accuracy": 0.7774410285055637,
      "num_tokens": 177186153.0,
      "step": 5551
    },
    {
      "entropy": 0.7394989263266325,
      "epoch": 0.5108193742416662,
      "grad_norm": 0.1589343547821045,
      "learning_rate": 8.29760480878339e-05,
      "loss": 0.7287,
      "mean_token_accuracy": 0.7859515137970448,
      "num_tokens": 177217534.0,
      "step": 5552
    },
    {
      "entropy": 0.7481295671314001,
      "epoch": 0.5109113806131075,
      "grad_norm": 0.1498357355594635,
      "learning_rate": 8.297298126169228e-05,
      "loss": 0.7347,
      "mean_token_accuracy": 0.7848804593086243,
      "num_tokens": 177249430.0,
      "step": 5553
    },
    {
      "entropy": 0.7883544079959393,
      "epoch": 0.5110033869845487,
      "grad_norm": 0.15654593706130981,
      "learning_rate": 8.296991443555066e-05,
      "loss": 0.7782,
      "mean_token_accuracy": 0.7710218764841557,
      "num_tokens": 177281647.0,
      "step": 5554
    },
    {
      "entropy": 0.7791575994342566,
      "epoch": 0.5110953933559899,
      "grad_norm": 0.15643994510173798,
      "learning_rate": 8.296684760940903e-05,
      "loss": 0.7735,
      "mean_token_accuracy": 0.7727659344673157,
      "num_tokens": 177314044.0,
      "step": 5555
    },
    {
      "entropy": 0.776557588018477,
      "epoch": 0.5111873997274311,
      "grad_norm": 0.16437825560569763,
      "learning_rate": 8.29637807832674e-05,
      "loss": 0.7721,
      "mean_token_accuracy": 0.7795406505465508,
      "num_tokens": 177346125.0,
      "step": 5556
    },
    {
      "entropy": 0.7091040629893541,
      "epoch": 0.5112794060988723,
      "grad_norm": 0.1639096587896347,
      "learning_rate": 8.296071395712578e-05,
      "loss": 0.7202,
      "mean_token_accuracy": 0.7873759493231773,
      "num_tokens": 177378117.0,
      "step": 5557
    },
    {
      "entropy": 0.7194349057972431,
      "epoch": 0.5113714124703136,
      "grad_norm": 0.1610925942659378,
      "learning_rate": 8.295764713098415e-05,
      "loss": 0.7181,
      "mean_token_accuracy": 0.7890763655304909,
      "num_tokens": 177410794.0,
      "step": 5558
    },
    {
      "entropy": 0.7043355498462915,
      "epoch": 0.5114634188417548,
      "grad_norm": 0.1628619134426117,
      "learning_rate": 8.295458030484253e-05,
      "loss": 0.7006,
      "mean_token_accuracy": 0.7978890128433704,
      "num_tokens": 177442052.0,
      "step": 5559
    },
    {
      "entropy": 0.7825636845082045,
      "epoch": 0.511555425213196,
      "grad_norm": 0.1836622953414917,
      "learning_rate": 8.29515134787009e-05,
      "loss": 0.7986,
      "mean_token_accuracy": 0.7679179608821869,
      "num_tokens": 177472973.0,
      "step": 5560
    },
    {
      "entropy": 0.7737556714564562,
      "epoch": 0.5116474315846372,
      "grad_norm": 0.15530718863010406,
      "learning_rate": 8.294844665255928e-05,
      "loss": 0.7542,
      "mean_token_accuracy": 0.780400063842535,
      "num_tokens": 177504931.0,
      "step": 5561
    },
    {
      "entropy": 0.7786627374589443,
      "epoch": 0.5117394379560785,
      "grad_norm": 0.16296736896038055,
      "learning_rate": 8.294537982641765e-05,
      "loss": 0.7722,
      "mean_token_accuracy": 0.7754708491265774,
      "num_tokens": 177535573.0,
      "step": 5562
    },
    {
      "entropy": 0.7252042964100838,
      "epoch": 0.5118314443275197,
      "grad_norm": 0.16052228212356567,
      "learning_rate": 8.294231300027601e-05,
      "loss": 0.6912,
      "mean_token_accuracy": 0.7934496402740479,
      "num_tokens": 177566993.0,
      "step": 5563
    },
    {
      "entropy": 0.8092333786189556,
      "epoch": 0.511923450698961,
      "grad_norm": 0.1567164808511734,
      "learning_rate": 8.29392461741344e-05,
      "loss": 0.8105,
      "mean_token_accuracy": 0.7618262320756912,
      "num_tokens": 177599123.0,
      "step": 5564
    },
    {
      "entropy": 0.7739488389343023,
      "epoch": 0.5120154570704021,
      "grad_norm": 0.16455404460430145,
      "learning_rate": 8.293617934799278e-05,
      "loss": 0.7833,
      "mean_token_accuracy": 0.7708249539136887,
      "num_tokens": 177631241.0,
      "step": 5565
    },
    {
      "entropy": 0.8040903322398663,
      "epoch": 0.5121074634418433,
      "grad_norm": 0.16365380585193634,
      "learning_rate": 8.293311252185114e-05,
      "loss": 0.7744,
      "mean_token_accuracy": 0.7730717286467552,
      "num_tokens": 177662746.0,
      "step": 5566
    },
    {
      "entropy": 0.7572248391807079,
      "epoch": 0.5121994698132846,
      "grad_norm": 0.1535940021276474,
      "learning_rate": 8.293004569570951e-05,
      "loss": 0.745,
      "mean_token_accuracy": 0.777117021381855,
      "num_tokens": 177694857.0,
      "step": 5567
    },
    {
      "entropy": 0.6850102227181196,
      "epoch": 0.5122914761847258,
      "grad_norm": 0.14632059633731842,
      "learning_rate": 8.292697886956788e-05,
      "loss": 0.663,
      "mean_token_accuracy": 0.8017014861106873,
      "num_tokens": 177726207.0,
      "step": 5568
    },
    {
      "entropy": 0.7620980739593506,
      "epoch": 0.5123834825561671,
      "grad_norm": 0.16246621310710907,
      "learning_rate": 8.292391204342626e-05,
      "loss": 0.7584,
      "mean_token_accuracy": 0.7774077542126179,
      "num_tokens": 177757497.0,
      "step": 5569
    },
    {
      "entropy": 0.7720983289182186,
      "epoch": 0.5124754889276082,
      "grad_norm": 0.14979298412799835,
      "learning_rate": 8.292084521728464e-05,
      "loss": 0.7544,
      "mean_token_accuracy": 0.7789630405604839,
      "num_tokens": 177789720.0,
      "step": 5570
    },
    {
      "entropy": 0.857785239815712,
      "epoch": 0.5125674952990494,
      "grad_norm": 0.16321483254432678,
      "learning_rate": 8.291777839114301e-05,
      "loss": 0.8587,
      "mean_token_accuracy": 0.750673696398735,
      "num_tokens": 177821721.0,
      "step": 5571
    },
    {
      "entropy": 0.7526744231581688,
      "epoch": 0.5126595016704907,
      "grad_norm": 0.16043740510940552,
      "learning_rate": 8.291471156500138e-05,
      "loss": 0.7428,
      "mean_token_accuracy": 0.7765586599707603,
      "num_tokens": 177853304.0,
      "step": 5572
    },
    {
      "entropy": 0.7225708533078432,
      "epoch": 0.5127515080419319,
      "grad_norm": 0.15786625444889069,
      "learning_rate": 8.291164473885976e-05,
      "loss": 0.722,
      "mean_token_accuracy": 0.7880221717059612,
      "num_tokens": 177885204.0,
      "step": 5573
    },
    {
      "entropy": 0.8323086593300104,
      "epoch": 0.5128435144133732,
      "grad_norm": 0.1601814180612564,
      "learning_rate": 8.290857791271813e-05,
      "loss": 0.8252,
      "mean_token_accuracy": 0.7609341107308865,
      "num_tokens": 177917544.0,
      "step": 5574
    },
    {
      "entropy": 0.7358943689614534,
      "epoch": 0.5129355207848143,
      "grad_norm": 0.16314977407455444,
      "learning_rate": 8.290551108657651e-05,
      "loss": 0.728,
      "mean_token_accuracy": 0.7905942611396313,
      "num_tokens": 177949427.0,
      "step": 5575
    },
    {
      "entropy": 0.7640921212732792,
      "epoch": 0.5130275271562555,
      "grad_norm": 0.15641044080257416,
      "learning_rate": 8.290244426043488e-05,
      "loss": 0.7474,
      "mean_token_accuracy": 0.7806496880948544,
      "num_tokens": 177981158.0,
      "step": 5576
    },
    {
      "entropy": 0.7594138160347939,
      "epoch": 0.5131195335276968,
      "grad_norm": 0.1587541401386261,
      "learning_rate": 8.289937743429326e-05,
      "loss": 0.7432,
      "mean_token_accuracy": 0.7839196808636189,
      "num_tokens": 178013246.0,
      "step": 5577
    },
    {
      "entropy": 0.7717094272375107,
      "epoch": 0.513211539899138,
      "grad_norm": 0.16266405582427979,
      "learning_rate": 8.289631060815163e-05,
      "loss": 0.7485,
      "mean_token_accuracy": 0.7804820463061333,
      "num_tokens": 178044679.0,
      "step": 5578
    },
    {
      "entropy": 0.7552562691271305,
      "epoch": 0.5133035462705793,
      "grad_norm": 0.1577078253030777,
      "learning_rate": 8.289324378201e-05,
      "loss": 0.7309,
      "mean_token_accuracy": 0.7822343334555626,
      "num_tokens": 178076490.0,
      "step": 5579
    },
    {
      "entropy": 0.7264198306947947,
      "epoch": 0.5133955526420204,
      "grad_norm": 0.1498204469680786,
      "learning_rate": 8.289017695586838e-05,
      "loss": 0.7003,
      "mean_token_accuracy": 0.7957865856587887,
      "num_tokens": 178108248.0,
      "step": 5580
    },
    {
      "entropy": 0.8030541036278009,
      "epoch": 0.5134875590134617,
      "grad_norm": 0.15614937245845795,
      "learning_rate": 8.288711012972676e-05,
      "loss": 0.7905,
      "mean_token_accuracy": 0.7664929181337357,
      "num_tokens": 178140263.0,
      "step": 5581
    },
    {
      "entropy": 0.8291088286787271,
      "epoch": 0.5135795653849029,
      "grad_norm": 0.1667584329843521,
      "learning_rate": 8.288404330358512e-05,
      "loss": 0.8448,
      "mean_token_accuracy": 0.756864458322525,
      "num_tokens": 178172172.0,
      "step": 5582
    },
    {
      "entropy": 0.7987677995115519,
      "epoch": 0.5136715717563441,
      "grad_norm": 0.1654706597328186,
      "learning_rate": 8.288097647744349e-05,
      "loss": 0.8082,
      "mean_token_accuracy": 0.7619786784052849,
      "num_tokens": 178203714.0,
      "step": 5583
    },
    {
      "entropy": 0.8312299000099301,
      "epoch": 0.5137635781277854,
      "grad_norm": 0.16484178602695465,
      "learning_rate": 8.287790965130187e-05,
      "loss": 0.8417,
      "mean_token_accuracy": 0.7596008591353893,
      "num_tokens": 178236041.0,
      "step": 5584
    },
    {
      "entropy": 0.8788485582917929,
      "epoch": 0.5138555844992265,
      "grad_norm": 0.17587406933307648,
      "learning_rate": 8.287484282516024e-05,
      "loss": 0.8942,
      "mean_token_accuracy": 0.7456468753516674,
      "num_tokens": 178268304.0,
      "step": 5585
    },
    {
      "entropy": 0.715834753587842,
      "epoch": 0.5139475908706678,
      "grad_norm": 0.16288234293460846,
      "learning_rate": 8.287177599901862e-05,
      "loss": 0.725,
      "mean_token_accuracy": 0.7920199111104012,
      "num_tokens": 178300608.0,
      "step": 5586
    },
    {
      "entropy": 0.7440790925174952,
      "epoch": 0.514039597242109,
      "grad_norm": 0.1506858468055725,
      "learning_rate": 8.286870917287699e-05,
      "loss": 0.7291,
      "mean_token_accuracy": 0.7820029892027378,
      "num_tokens": 178332847.0,
      "step": 5587
    },
    {
      "entropy": 0.7636577747762203,
      "epoch": 0.5141316036135503,
      "grad_norm": 0.15260744094848633,
      "learning_rate": 8.286564234673537e-05,
      "loss": 0.7477,
      "mean_token_accuracy": 0.778972115367651,
      "num_tokens": 178365031.0,
      "step": 5588
    },
    {
      "entropy": 0.7450150419026613,
      "epoch": 0.5142236099849915,
      "grad_norm": 0.15142494440078735,
      "learning_rate": 8.286257552059374e-05,
      "loss": 0.7256,
      "mean_token_accuracy": 0.7859993390738964,
      "num_tokens": 178397100.0,
      "step": 5589
    },
    {
      "entropy": 0.7902210364118218,
      "epoch": 0.5143156163564326,
      "grad_norm": 0.15804021060466766,
      "learning_rate": 8.285950869445211e-05,
      "loss": 0.7724,
      "mean_token_accuracy": 0.7736189775168896,
      "num_tokens": 178428675.0,
      "step": 5590
    },
    {
      "entropy": 0.7685850318521261,
      "epoch": 0.5144076227278739,
      "grad_norm": 0.14909324049949646,
      "learning_rate": 8.285644186831049e-05,
      "loss": 0.7448,
      "mean_token_accuracy": 0.7820509411394596,
      "num_tokens": 178461249.0,
      "step": 5591
    },
    {
      "entropy": 0.776036823168397,
      "epoch": 0.5144996290993151,
      "grad_norm": 0.1571199744939804,
      "learning_rate": 8.285337504216887e-05,
      "loss": 0.7475,
      "mean_token_accuracy": 0.7806911319494247,
      "num_tokens": 178493787.0,
      "step": 5592
    },
    {
      "entropy": 0.7620203588157892,
      "epoch": 0.5145916354707564,
      "grad_norm": 0.15456201136112213,
      "learning_rate": 8.285030821602724e-05,
      "loss": 0.732,
      "mean_token_accuracy": 0.7847346812486649,
      "num_tokens": 178525873.0,
      "step": 5593
    },
    {
      "entropy": 0.9185096696019173,
      "epoch": 0.5146836418421976,
      "grad_norm": 0.15717710554599762,
      "learning_rate": 8.284724138988561e-05,
      "loss": 0.8968,
      "mean_token_accuracy": 0.7410906590521336,
      "num_tokens": 178557251.0,
      "step": 5594
    },
    {
      "entropy": 0.7221684474498034,
      "epoch": 0.5147756482136387,
      "grad_norm": 0.1606985628604889,
      "learning_rate": 8.284417456374398e-05,
      "loss": 0.7264,
      "mean_token_accuracy": 0.7889560386538506,
      "num_tokens": 178589122.0,
      "step": 5595
    },
    {
      "entropy": 0.7130689425393939,
      "epoch": 0.51486765458508,
      "grad_norm": 0.15460146963596344,
      "learning_rate": 8.284110773760237e-05,
      "loss": 0.6958,
      "mean_token_accuracy": 0.7960742935538292,
      "num_tokens": 178620977.0,
      "step": 5596
    },
    {
      "entropy": 0.8710994757711887,
      "epoch": 0.5149596609565212,
      "grad_norm": 0.16297082602977753,
      "learning_rate": 8.283804091146074e-05,
      "loss": 0.8669,
      "mean_token_accuracy": 0.7452164031565189,
      "num_tokens": 178652492.0,
      "step": 5597
    },
    {
      "entropy": 0.7987136729061604,
      "epoch": 0.5150516673279625,
      "grad_norm": 0.15566574037075043,
      "learning_rate": 8.28349740853191e-05,
      "loss": 0.7685,
      "mean_token_accuracy": 0.776330292224884,
      "num_tokens": 178684482.0,
      "step": 5598
    },
    {
      "entropy": 0.7020883429795504,
      "epoch": 0.5151436736994037,
      "grad_norm": 0.15635594725608826,
      "learning_rate": 8.283190725917747e-05,
      "loss": 0.6893,
      "mean_token_accuracy": 0.7954985275864601,
      "num_tokens": 178715902.0,
      "step": 5599
    },
    {
      "entropy": 0.7004642263054848,
      "epoch": 0.5152356800708449,
      "grad_norm": 0.15661366283893585,
      "learning_rate": 8.282884043303586e-05,
      "loss": 0.6999,
      "mean_token_accuracy": 0.797734335064888,
      "num_tokens": 178747814.0,
      "step": 5600
    },
    {
      "entropy": 0.7170673599466681,
      "epoch": 0.5153276864422861,
      "grad_norm": 0.15318408608436584,
      "learning_rate": 8.282577360689424e-05,
      "loss": 0.7313,
      "mean_token_accuracy": 0.7850479781627655,
      "num_tokens": 178779657.0,
      "step": 5601
    },
    {
      "entropy": 0.8155318014323711,
      "epoch": 0.5154196928137273,
      "grad_norm": 0.15939867496490479,
      "learning_rate": 8.28227067807526e-05,
      "loss": 0.804,
      "mean_token_accuracy": 0.7625459432601929,
      "num_tokens": 178811936.0,
      "step": 5602
    },
    {
      "entropy": 0.726182371377945,
      "epoch": 0.5155116991851686,
      "grad_norm": 0.1622529774904251,
      "learning_rate": 8.281963995461097e-05,
      "loss": 0.7407,
      "mean_token_accuracy": 0.7809689380228519,
      "num_tokens": 178843642.0,
      "step": 5603
    },
    {
      "entropy": 0.8067097431048751,
      "epoch": 0.5156037055566098,
      "grad_norm": 0.16237229108810425,
      "learning_rate": 8.281657312846935e-05,
      "loss": 0.8102,
      "mean_token_accuracy": 0.7629373259842396,
      "num_tokens": 178875398.0,
      "step": 5604
    },
    {
      "entropy": 0.7525291405618191,
      "epoch": 0.515695711928051,
      "grad_norm": 0.15189476311206818,
      "learning_rate": 8.281350630232772e-05,
      "loss": 0.7513,
      "mean_token_accuracy": 0.7808424681425095,
      "num_tokens": 178907914.0,
      "step": 5605
    },
    {
      "entropy": 0.7409445960074663,
      "epoch": 0.5157877182994922,
      "grad_norm": 0.15956507623195648,
      "learning_rate": 8.28104394761861e-05,
      "loss": 0.7276,
      "mean_token_accuracy": 0.7821437455713749,
      "num_tokens": 178940219.0,
      "step": 5606
    },
    {
      "entropy": 0.74495661072433,
      "epoch": 0.5158797246709335,
      "grad_norm": 0.16048841178417206,
      "learning_rate": 8.280737265004447e-05,
      "loss": 0.7297,
      "mean_token_accuracy": 0.7874671965837479,
      "num_tokens": 178972214.0,
      "step": 5607
    },
    {
      "entropy": 0.8022854961454868,
      "epoch": 0.5159717310423747,
      "grad_norm": 0.15940147638320923,
      "learning_rate": 8.280430582390285e-05,
      "loss": 0.7722,
      "mean_token_accuracy": 0.7828244380652905,
      "num_tokens": 179004464.0,
      "step": 5608
    },
    {
      "entropy": 0.6624792385846376,
      "epoch": 0.516063737413816,
      "grad_norm": 0.14399422705173492,
      "learning_rate": 8.280123899776122e-05,
      "loss": 0.6421,
      "mean_token_accuracy": 0.8049290552735329,
      "num_tokens": 179036568.0,
      "step": 5609
    },
    {
      "entropy": 0.8407721817493439,
      "epoch": 0.5161557437852571,
      "grad_norm": 0.15696172416210175,
      "learning_rate": 8.279817217161959e-05,
      "loss": 0.8069,
      "mean_token_accuracy": 0.7648395895957947,
      "num_tokens": 179069223.0,
      "step": 5610
    },
    {
      "entropy": 0.7472903206944466,
      "epoch": 0.5162477501566983,
      "grad_norm": 0.15890230238437653,
      "learning_rate": 8.279510534547797e-05,
      "loss": 0.7294,
      "mean_token_accuracy": 0.7849084660410881,
      "num_tokens": 179100986.0,
      "step": 5611
    },
    {
      "entropy": 0.7762463204562664,
      "epoch": 0.5163397565281396,
      "grad_norm": 0.15745417773723602,
      "learning_rate": 8.279203851933635e-05,
      "loss": 0.7882,
      "mean_token_accuracy": 0.7688025012612343,
      "num_tokens": 179132075.0,
      "step": 5612
    },
    {
      "entropy": 0.7025297321379185,
      "epoch": 0.5164317628995808,
      "grad_norm": 0.15459878742694855,
      "learning_rate": 8.278897169319472e-05,
      "loss": 0.7028,
      "mean_token_accuracy": 0.7946708090603352,
      "num_tokens": 179163758.0,
      "step": 5613
    },
    {
      "entropy": 0.8000152930617332,
      "epoch": 0.5165237692710221,
      "grad_norm": 0.14827142655849457,
      "learning_rate": 8.278590486705309e-05,
      "loss": 0.7856,
      "mean_token_accuracy": 0.7675283960998058,
      "num_tokens": 179196348.0,
      "step": 5614
    },
    {
      "entropy": 0.7777385730296373,
      "epoch": 0.5166157756424632,
      "grad_norm": 0.15975840389728546,
      "learning_rate": 8.278283804091147e-05,
      "loss": 0.764,
      "mean_token_accuracy": 0.7751429788768291,
      "num_tokens": 179227690.0,
      "step": 5615
    },
    {
      "entropy": 0.7847190164029598,
      "epoch": 0.5167077820139044,
      "grad_norm": 0.1598164439201355,
      "learning_rate": 8.277977121476984e-05,
      "loss": 0.7772,
      "mean_token_accuracy": 0.7776173315942287,
      "num_tokens": 179259847.0,
      "step": 5616
    },
    {
      "entropy": 0.8337332792580128,
      "epoch": 0.5167997883853457,
      "grad_norm": 0.1584751158952713,
      "learning_rate": 8.277670438862822e-05,
      "loss": 0.8188,
      "mean_token_accuracy": 0.7626777365803719,
      "num_tokens": 179291280.0,
      "step": 5617
    },
    {
      "entropy": 0.7840922120958567,
      "epoch": 0.5168917947567869,
      "grad_norm": 0.16838425397872925,
      "learning_rate": 8.277363756248659e-05,
      "loss": 0.7989,
      "mean_token_accuracy": 0.7662032283842564,
      "num_tokens": 179323405.0,
      "step": 5618
    },
    {
      "entropy": 0.746511947363615,
      "epoch": 0.5169838011282282,
      "grad_norm": 0.14899839460849762,
      "learning_rate": 8.277057073634497e-05,
      "loss": 0.7246,
      "mean_token_accuracy": 0.7847266681492329,
      "num_tokens": 179356002.0,
      "step": 5619
    },
    {
      "entropy": 0.8141305521130562,
      "epoch": 0.5170758074996693,
      "grad_norm": 0.1625165194272995,
      "learning_rate": 8.276750391020333e-05,
      "loss": 0.8072,
      "mean_token_accuracy": 0.7661194540560246,
      "num_tokens": 179387896.0,
      "step": 5620
    },
    {
      "entropy": 0.7145630177110434,
      "epoch": 0.5171678138711105,
      "grad_norm": 0.15786652266979218,
      "learning_rate": 8.27644370840617e-05,
      "loss": 0.7234,
      "mean_token_accuracy": 0.7869623675942421,
      "num_tokens": 179420363.0,
      "step": 5621
    },
    {
      "entropy": 0.6377898715436459,
      "epoch": 0.5172598202425518,
      "grad_norm": 0.1465887725353241,
      "learning_rate": 8.276137025792008e-05,
      "loss": 0.6227,
      "mean_token_accuracy": 0.8146916627883911,
      "num_tokens": 179452535.0,
      "step": 5622
    },
    {
      "entropy": 0.8678732886910439,
      "epoch": 0.517351826613993,
      "grad_norm": 0.15533800423145294,
      "learning_rate": 8.275830343177847e-05,
      "loss": 0.8537,
      "mean_token_accuracy": 0.7501406148076057,
      "num_tokens": 179484560.0,
      "step": 5623
    },
    {
      "entropy": 0.7350630946457386,
      "epoch": 0.5174438329854343,
      "grad_norm": 0.15170548856258392,
      "learning_rate": 8.275523660563683e-05,
      "loss": 0.7233,
      "mean_token_accuracy": 0.7870054356753826,
      "num_tokens": 179516283.0,
      "step": 5624
    },
    {
      "entropy": 0.810594417154789,
      "epoch": 0.5175358393568754,
      "grad_norm": 0.16272224485874176,
      "learning_rate": 8.27521697794952e-05,
      "loss": 0.8177,
      "mean_token_accuracy": 0.7620367221534252,
      "num_tokens": 179548134.0,
      "step": 5625
    },
    {
      "entropy": 0.8137489948421717,
      "epoch": 0.5176278457283167,
      "grad_norm": 0.171220064163208,
      "learning_rate": 8.274910295335357e-05,
      "loss": 0.8183,
      "mean_token_accuracy": 0.7626495063304901,
      "num_tokens": 179580147.0,
      "step": 5626
    },
    {
      "entropy": 0.7179593238979578,
      "epoch": 0.5177198520997579,
      "grad_norm": 0.15387381613254547,
      "learning_rate": 8.274603612721195e-05,
      "loss": 0.7276,
      "mean_token_accuracy": 0.7838428914546967,
      "num_tokens": 179612038.0,
      "step": 5627
    },
    {
      "entropy": 0.7900035865604877,
      "epoch": 0.5178118584711991,
      "grad_norm": 0.1579780876636505,
      "learning_rate": 8.274296930107033e-05,
      "loss": 0.7564,
      "mean_token_accuracy": 0.7786357812583447,
      "num_tokens": 179644214.0,
      "step": 5628
    },
    {
      "entropy": 0.8166593573987484,
      "epoch": 0.5179038648426404,
      "grad_norm": 0.1647643744945526,
      "learning_rate": 8.27399024749287e-05,
      "loss": 0.8081,
      "mean_token_accuracy": 0.7648350484669209,
      "num_tokens": 179675242.0,
      "step": 5629
    },
    {
      "entropy": 0.7898218519985676,
      "epoch": 0.5179958712140815,
      "grad_norm": 0.15826672315597534,
      "learning_rate": 8.273683564878707e-05,
      "loss": 0.7724,
      "mean_token_accuracy": 0.7727367728948593,
      "num_tokens": 179707115.0,
      "step": 5630
    },
    {
      "entropy": 0.7605730928480625,
      "epoch": 0.5180878775855228,
      "grad_norm": 0.15039391815662384,
      "learning_rate": 8.273376882264545e-05,
      "loss": 0.7423,
      "mean_token_accuracy": 0.7833795286715031,
      "num_tokens": 179739095.0,
      "step": 5631
    },
    {
      "entropy": 0.8429079465568066,
      "epoch": 0.518179883956964,
      "grad_norm": 0.1565682739019394,
      "learning_rate": 8.273070199650382e-05,
      "loss": 0.8193,
      "mean_token_accuracy": 0.7606905102729797,
      "num_tokens": 179770638.0,
      "step": 5632
    },
    {
      "entropy": 0.7971814405173063,
      "epoch": 0.5182718903284053,
      "grad_norm": 0.15564623475074768,
      "learning_rate": 8.27276351703622e-05,
      "loss": 0.7745,
      "mean_token_accuracy": 0.7743802890181541,
      "num_tokens": 179802412.0,
      "step": 5633
    },
    {
      "entropy": 0.8952645473182201,
      "epoch": 0.5183638966998465,
      "grad_norm": 0.1587388664484024,
      "learning_rate": 8.272456834422057e-05,
      "loss": 0.8907,
      "mean_token_accuracy": 0.7420218996703625,
      "num_tokens": 179833904.0,
      "step": 5634
    },
    {
      "entropy": 0.8558731935918331,
      "epoch": 0.5184559030712876,
      "grad_norm": 0.15782083570957184,
      "learning_rate": 8.272150151807895e-05,
      "loss": 0.8504,
      "mean_token_accuracy": 0.7496363893151283,
      "num_tokens": 179866235.0,
      "step": 5635
    },
    {
      "entropy": 0.7999930651858449,
      "epoch": 0.5185479094427289,
      "grad_norm": 0.15308870375156403,
      "learning_rate": 8.271843469193732e-05,
      "loss": 0.7877,
      "mean_token_accuracy": 0.767062459141016,
      "num_tokens": 179898508.0,
      "step": 5636
    },
    {
      "entropy": 0.6725904047489166,
      "epoch": 0.5186399158141701,
      "grad_norm": 0.15313467383384705,
      "learning_rate": 8.271536786579568e-05,
      "loss": 0.6653,
      "mean_token_accuracy": 0.8056252636015415,
      "num_tokens": 179929971.0,
      "step": 5637
    },
    {
      "entropy": 0.8427846059203148,
      "epoch": 0.5187319221856114,
      "grad_norm": 0.1676439642906189,
      "learning_rate": 8.271230103965406e-05,
      "loss": 0.8485,
      "mean_token_accuracy": 0.7532826289534569,
      "num_tokens": 179962026.0,
      "step": 5638
    },
    {
      "entropy": 0.8391323480755091,
      "epoch": 0.5188239285570526,
      "grad_norm": 0.15774141252040863,
      "learning_rate": 8.270923421351245e-05,
      "loss": 0.8495,
      "mean_token_accuracy": 0.7543689720332623,
      "num_tokens": 179994025.0,
      "step": 5639
    },
    {
      "entropy": 0.8124000336974859,
      "epoch": 0.5189159349284937,
      "grad_norm": 0.1657446324825287,
      "learning_rate": 8.270616738737081e-05,
      "loss": 0.804,
      "mean_token_accuracy": 0.7666653729975224,
      "num_tokens": 180025903.0,
      "step": 5640
    },
    {
      "entropy": 0.7513125296682119,
      "epoch": 0.519007941299935,
      "grad_norm": 0.16657403111457825,
      "learning_rate": 8.270310056122918e-05,
      "loss": 0.7551,
      "mean_token_accuracy": 0.7791264094412327,
      "num_tokens": 180057745.0,
      "step": 5641
    },
    {
      "entropy": 0.7260773461312056,
      "epoch": 0.5190999476713762,
      "grad_norm": 0.1637420803308487,
      "learning_rate": 8.270003373508756e-05,
      "loss": 0.7321,
      "mean_token_accuracy": 0.7836794778704643,
      "num_tokens": 180089821.0,
      "step": 5642
    },
    {
      "entropy": 0.8216852601617575,
      "epoch": 0.5191919540428175,
      "grad_norm": 0.16604897379875183,
      "learning_rate": 8.269696690894594e-05,
      "loss": 0.8157,
      "mean_token_accuracy": 0.7581232264637947,
      "num_tokens": 180121381.0,
      "step": 5643
    },
    {
      "entropy": 0.7539273854345083,
      "epoch": 0.5192839604142587,
      "grad_norm": 0.16628219187259674,
      "learning_rate": 8.269390008280431e-05,
      "loss": 0.7504,
      "mean_token_accuracy": 0.7806656807661057,
      "num_tokens": 180152199.0,
      "step": 5644
    },
    {
      "entropy": 0.7662339899688959,
      "epoch": 0.5193759667856999,
      "grad_norm": 0.1597549319267273,
      "learning_rate": 8.269083325666268e-05,
      "loss": 0.7621,
      "mean_token_accuracy": 0.7799523696303368,
      "num_tokens": 180184967.0,
      "step": 5645
    },
    {
      "entropy": 0.845312524586916,
      "epoch": 0.5194679731571411,
      "grad_norm": 0.162149578332901,
      "learning_rate": 8.268776643052106e-05,
      "loss": 0.8222,
      "mean_token_accuracy": 0.7613573111593723,
      "num_tokens": 180216741.0,
      "step": 5646
    },
    {
      "entropy": 0.8035998065024614,
      "epoch": 0.5195599795285823,
      "grad_norm": 0.1575022041797638,
      "learning_rate": 8.268469960437943e-05,
      "loss": 0.7888,
      "mean_token_accuracy": 0.7712521553039551,
      "num_tokens": 180248478.0,
      "step": 5647
    },
    {
      "entropy": 0.786101090721786,
      "epoch": 0.5196519859000236,
      "grad_norm": 0.1512508988380432,
      "learning_rate": 8.268163277823781e-05,
      "loss": 0.765,
      "mean_token_accuracy": 0.7814750522375107,
      "num_tokens": 180280961.0,
      "step": 5648
    },
    {
      "entropy": 0.8210403677076101,
      "epoch": 0.5197439922714648,
      "grad_norm": 0.16417557001113892,
      "learning_rate": 8.267856595209618e-05,
      "loss": 0.8127,
      "mean_token_accuracy": 0.7633024379611015,
      "num_tokens": 180312613.0,
      "step": 5649
    },
    {
      "entropy": 0.8788082525134087,
      "epoch": 0.519835998642906,
      "grad_norm": 0.1629500538110733,
      "learning_rate": 8.267549912595456e-05,
      "loss": 0.8793,
      "mean_token_accuracy": 0.7471968978643417,
      "num_tokens": 180344197.0,
      "step": 5650
    },
    {
      "entropy": 0.7833117209374905,
      "epoch": 0.5199280050143472,
      "grad_norm": 0.15587082505226135,
      "learning_rate": 8.267243229981293e-05,
      "loss": 0.76,
      "mean_token_accuracy": 0.7796783074736595,
      "num_tokens": 180376368.0,
      "step": 5651
    },
    {
      "entropy": 0.7783274613320827,
      "epoch": 0.5200200113857885,
      "grad_norm": 0.15250414609909058,
      "learning_rate": 8.26693654736713e-05,
      "loss": 0.7603,
      "mean_token_accuracy": 0.7779425121843815,
      "num_tokens": 180408646.0,
      "step": 5652
    },
    {
      "entropy": 0.7666708249598742,
      "epoch": 0.5201120177572297,
      "grad_norm": 0.16201132535934448,
      "learning_rate": 8.266629864752968e-05,
      "loss": 0.7499,
      "mean_token_accuracy": 0.7739836201071739,
      "num_tokens": 180440302.0,
      "step": 5653
    },
    {
      "entropy": 0.7555548660457134,
      "epoch": 0.520204024128671,
      "grad_norm": 0.15427671372890472,
      "learning_rate": 8.266323182138806e-05,
      "loss": 0.7435,
      "mean_token_accuracy": 0.7843424752354622,
      "num_tokens": 180472464.0,
      "step": 5654
    },
    {
      "entropy": 0.7343436665832996,
      "epoch": 0.5202960305001121,
      "grad_norm": 0.15747348964214325,
      "learning_rate": 8.266016499524643e-05,
      "loss": 0.7289,
      "mean_token_accuracy": 0.7879518792033195,
      "num_tokens": 180505232.0,
      "step": 5655
    },
    {
      "entropy": 0.762737799435854,
      "epoch": 0.5203880368715533,
      "grad_norm": 0.1654900759458542,
      "learning_rate": 8.26570981691048e-05,
      "loss": 0.7604,
      "mean_token_accuracy": 0.7736629657447338,
      "num_tokens": 180537064.0,
      "step": 5656
    },
    {
      "entropy": 0.7946772947907448,
      "epoch": 0.5204800432429946,
      "grad_norm": 0.1665303111076355,
      "learning_rate": 8.265403134296316e-05,
      "loss": 0.8076,
      "mean_token_accuracy": 0.7659123949706554,
      "num_tokens": 180568764.0,
      "step": 5657
    },
    {
      "entropy": 0.8188362866640091,
      "epoch": 0.5205720496144358,
      "grad_norm": 0.16205623745918274,
      "learning_rate": 8.265096451682154e-05,
      "loss": 0.8057,
      "mean_token_accuracy": 0.7686231583356857,
      "num_tokens": 180600723.0,
      "step": 5658
    },
    {
      "entropy": 0.7536797076463699,
      "epoch": 0.5206640559858771,
      "grad_norm": 0.1575290560722351,
      "learning_rate": 8.264789769067993e-05,
      "loss": 0.756,
      "mean_token_accuracy": 0.7797265946865082,
      "num_tokens": 180632971.0,
      "step": 5659
    },
    {
      "entropy": 0.6432328727096319,
      "epoch": 0.5207560623573182,
      "grad_norm": 0.1465015411376953,
      "learning_rate": 8.26448308645383e-05,
      "loss": 0.6232,
      "mean_token_accuracy": 0.8126221261918545,
      "num_tokens": 180665739.0,
      "step": 5660
    },
    {
      "entropy": 0.8719392754137516,
      "epoch": 0.5208480687287594,
      "grad_norm": 0.17105872929096222,
      "learning_rate": 8.264176403839666e-05,
      "loss": 0.8755,
      "mean_token_accuracy": 0.7549887746572495,
      "num_tokens": 180698093.0,
      "step": 5661
    },
    {
      "entropy": 0.7288300488144159,
      "epoch": 0.5209400751002007,
      "grad_norm": 0.15698279440402985,
      "learning_rate": 8.263869721225504e-05,
      "loss": 0.6915,
      "mean_token_accuracy": 0.8010087721049786,
      "num_tokens": 180730659.0,
      "step": 5662
    },
    {
      "entropy": 0.7285200301557779,
      "epoch": 0.5210320814716419,
      "grad_norm": 0.15476995706558228,
      "learning_rate": 8.263563038611341e-05,
      "loss": 0.7055,
      "mean_token_accuracy": 0.794529166072607,
      "num_tokens": 180762758.0,
      "step": 5663
    },
    {
      "entropy": 0.7220368701964617,
      "epoch": 0.5211240878430832,
      "grad_norm": 0.1515016257762909,
      "learning_rate": 8.263256355997179e-05,
      "loss": 0.7215,
      "mean_token_accuracy": 0.7857054732739925,
      "num_tokens": 180794560.0,
      "step": 5664
    },
    {
      "entropy": 0.7539980411529541,
      "epoch": 0.5212160942145243,
      "grad_norm": 0.14796452224254608,
      "learning_rate": 8.262949673383016e-05,
      "loss": 0.7333,
      "mean_token_accuracy": 0.7902855090796947,
      "num_tokens": 180826353.0,
      "step": 5665
    },
    {
      "entropy": 0.7458801846951246,
      "epoch": 0.5213081005859656,
      "grad_norm": 0.153997004032135,
      "learning_rate": 8.262642990768854e-05,
      "loss": 0.7206,
      "mean_token_accuracy": 0.7871546223759651,
      "num_tokens": 180857761.0,
      "step": 5666
    },
    {
      "entropy": 0.6665875855833292,
      "epoch": 0.5214001069574068,
      "grad_norm": 0.1500423699617386,
      "learning_rate": 8.262336308154691e-05,
      "loss": 0.6389,
      "mean_token_accuracy": 0.8110428489744663,
      "num_tokens": 180889706.0,
      "step": 5667
    },
    {
      "entropy": 0.8604093790054321,
      "epoch": 0.521492113328848,
      "grad_norm": 0.16227003931999207,
      "learning_rate": 8.262029625540528e-05,
      "loss": 0.8512,
      "mean_token_accuracy": 0.7513170838356018,
      "num_tokens": 180921102.0,
      "step": 5668
    },
    {
      "entropy": 0.7949501406401396,
      "epoch": 0.5215841197002893,
      "grad_norm": 0.16124482452869415,
      "learning_rate": 8.261722942926366e-05,
      "loss": 0.7886,
      "mean_token_accuracy": 0.7753051333129406,
      "num_tokens": 180952359.0,
      "step": 5669
    },
    {
      "entropy": 0.7667530719190836,
      "epoch": 0.5216761260717304,
      "grad_norm": 0.1551930159330368,
      "learning_rate": 8.261416260312204e-05,
      "loss": 0.7762,
      "mean_token_accuracy": 0.7743146046996117,
      "num_tokens": 180984578.0,
      "step": 5670
    },
    {
      "entropy": 0.6788841541856527,
      "epoch": 0.5217681324431717,
      "grad_norm": 0.16602317988872528,
      "learning_rate": 8.261109577698041e-05,
      "loss": 0.6838,
      "mean_token_accuracy": 0.7961423546075821,
      "num_tokens": 181016996.0,
      "step": 5671
    },
    {
      "entropy": 0.908289585262537,
      "epoch": 0.5218601388146129,
      "grad_norm": 0.16856810450553894,
      "learning_rate": 8.260802895083878e-05,
      "loss": 0.9112,
      "mean_token_accuracy": 0.7379319034516811,
      "num_tokens": 181049335.0,
      "step": 5672
    },
    {
      "entropy": 0.7336226869374514,
      "epoch": 0.5219521451860542,
      "grad_norm": 0.15026211738586426,
      "learning_rate": 8.260496212469714e-05,
      "loss": 0.717,
      "mean_token_accuracy": 0.7926338203251362,
      "num_tokens": 181080754.0,
      "step": 5673
    },
    {
      "entropy": 0.962308507412672,
      "epoch": 0.5220441515574954,
      "grad_norm": 0.16141235828399658,
      "learning_rate": 8.260189529855554e-05,
      "loss": 0.9558,
      "mean_token_accuracy": 0.727350939065218,
      "num_tokens": 181113384.0,
      "step": 5674
    },
    {
      "entropy": 0.7263786531984806,
      "epoch": 0.5221361579289365,
      "grad_norm": 0.15994307398796082,
      "learning_rate": 8.25988284724139e-05,
      "loss": 0.7165,
      "mean_token_accuracy": 0.7899485640227795,
      "num_tokens": 181145872.0,
      "step": 5675
    },
    {
      "entropy": 0.8656177073717117,
      "epoch": 0.5222281643003778,
      "grad_norm": 0.17597834765911102,
      "learning_rate": 8.259576164627227e-05,
      "loss": 0.878,
      "mean_token_accuracy": 0.757008608430624,
      "num_tokens": 181177623.0,
      "step": 5676
    },
    {
      "entropy": 0.6402064952999353,
      "epoch": 0.522320170671819,
      "grad_norm": 0.1400018036365509,
      "learning_rate": 8.259269482013066e-05,
      "loss": 0.6146,
      "mean_token_accuracy": 0.8152494691312313,
      "num_tokens": 181209792.0,
      "step": 5677
    },
    {
      "entropy": 0.7754597030580044,
      "epoch": 0.5224121770432603,
      "grad_norm": 0.15421535074710846,
      "learning_rate": 8.258962799398902e-05,
      "loss": 0.7783,
      "mean_token_accuracy": 0.7730530276894569,
      "num_tokens": 181242225.0,
      "step": 5678
    },
    {
      "entropy": 0.8405924700200558,
      "epoch": 0.5225041834147015,
      "grad_norm": 0.15755626559257507,
      "learning_rate": 8.25865611678474e-05,
      "loss": 0.803,
      "mean_token_accuracy": 0.7683230973780155,
      "num_tokens": 181274433.0,
      "step": 5679
    },
    {
      "entropy": 0.8202827833592892,
      "epoch": 0.5225961897861426,
      "grad_norm": 0.16715143620967865,
      "learning_rate": 8.258349434170577e-05,
      "loss": 0.8211,
      "mean_token_accuracy": 0.7614513002336025,
      "num_tokens": 181306685.0,
      "step": 5680
    },
    {
      "entropy": 0.8367418507114053,
      "epoch": 0.5226881961575839,
      "grad_norm": 0.1577163189649582,
      "learning_rate": 8.258042751556415e-05,
      "loss": 0.8092,
      "mean_token_accuracy": 0.7614225782454014,
      "num_tokens": 181338952.0,
      "step": 5681
    },
    {
      "entropy": 0.8194203265011311,
      "epoch": 0.5227802025290251,
      "grad_norm": 0.1631816178560257,
      "learning_rate": 8.257736068942252e-05,
      "loss": 0.8268,
      "mean_token_accuracy": 0.7639862783253193,
      "num_tokens": 181370874.0,
      "step": 5682
    },
    {
      "entropy": 0.6863573379814625,
      "epoch": 0.5228722089004664,
      "grad_norm": 0.15100312232971191,
      "learning_rate": 8.257429386328089e-05,
      "loss": 0.6633,
      "mean_token_accuracy": 0.8039701879024506,
      "num_tokens": 181403077.0,
      "step": 5683
    },
    {
      "entropy": 0.7782385721802711,
      "epoch": 0.5229642152719076,
      "grad_norm": 0.15932892262935638,
      "learning_rate": 8.257122703713927e-05,
      "loss": 0.7809,
      "mean_token_accuracy": 0.7685554325580597,
      "num_tokens": 181433996.0,
      "step": 5684
    },
    {
      "entropy": 0.7581074945628643,
      "epoch": 0.5230562216433488,
      "grad_norm": 0.15356872975826263,
      "learning_rate": 8.256816021099765e-05,
      "loss": 0.7463,
      "mean_token_accuracy": 0.7829085364937782,
      "num_tokens": 181466227.0,
      "step": 5685
    },
    {
      "entropy": 0.8769489079713821,
      "epoch": 0.52314822801479,
      "grad_norm": 0.15597963333129883,
      "learning_rate": 8.256509338485602e-05,
      "loss": 0.8749,
      "mean_token_accuracy": 0.7481356002390385,
      "num_tokens": 181498287.0,
      "step": 5686
    },
    {
      "entropy": 0.7250389140099287,
      "epoch": 0.5232402343862312,
      "grad_norm": 0.14781154692173004,
      "learning_rate": 8.256202655871439e-05,
      "loss": 0.7051,
      "mean_token_accuracy": 0.7970584146678448,
      "num_tokens": 181530211.0,
      "step": 5687
    },
    {
      "entropy": 0.6713395472615957,
      "epoch": 0.5233322407576725,
      "grad_norm": 0.16719132661819458,
      "learning_rate": 8.255895973257276e-05,
      "loss": 0.6892,
      "mean_token_accuracy": 0.7990577295422554,
      "num_tokens": 181561578.0,
      "step": 5688
    },
    {
      "entropy": 0.795326529070735,
      "epoch": 0.5234242471291137,
      "grad_norm": 0.1602139174938202,
      "learning_rate": 8.255589290643114e-05,
      "loss": 0.7886,
      "mean_token_accuracy": 0.7673952430486679,
      "num_tokens": 181593896.0,
      "step": 5689
    },
    {
      "entropy": 0.8119101990014315,
      "epoch": 0.5235162535005549,
      "grad_norm": 0.1534675508737564,
      "learning_rate": 8.255282608028952e-05,
      "loss": 0.8014,
      "mean_token_accuracy": 0.7683606706559658,
      "num_tokens": 181625767.0,
      "step": 5690
    },
    {
      "entropy": 0.9277946520596743,
      "epoch": 0.5236082598719961,
      "grad_norm": 0.1636352837085724,
      "learning_rate": 8.254975925414789e-05,
      "loss": 0.9189,
      "mean_token_accuracy": 0.7339668199419975,
      "num_tokens": 181657968.0,
      "step": 5691
    },
    {
      "entropy": 0.6610449114814401,
      "epoch": 0.5237002662434374,
      "grad_norm": 0.16134560108184814,
      "learning_rate": 8.254669242800625e-05,
      "loss": 0.6482,
      "mean_token_accuracy": 0.8108754679560661,
      "num_tokens": 181689844.0,
      "step": 5692
    },
    {
      "entropy": 0.8070376813411713,
      "epoch": 0.5237922726148786,
      "grad_norm": 0.1664857268333435,
      "learning_rate": 8.254362560186464e-05,
      "loss": 0.7775,
      "mean_token_accuracy": 0.7777979001402855,
      "num_tokens": 181721710.0,
      "step": 5693
    },
    {
      "entropy": 0.7152484068647027,
      "epoch": 0.5238842789863198,
      "grad_norm": 0.15665829181671143,
      "learning_rate": 8.2540558775723e-05,
      "loss": 0.6903,
      "mean_token_accuracy": 0.7929143235087395,
      "num_tokens": 181753874.0,
      "step": 5694
    },
    {
      "entropy": 0.7498865891247988,
      "epoch": 0.523976285357761,
      "grad_norm": 0.1566300392150879,
      "learning_rate": 8.253749194958139e-05,
      "loss": 0.7384,
      "mean_token_accuracy": 0.7793780602514744,
      "num_tokens": 181785604.0,
      "step": 5695
    },
    {
      "entropy": 0.8400232065469027,
      "epoch": 0.5240682917292022,
      "grad_norm": 0.15991832315921783,
      "learning_rate": 8.253442512343975e-05,
      "loss": 0.8393,
      "mean_token_accuracy": 0.7539766579866409,
      "num_tokens": 181817115.0,
      "step": 5696
    },
    {
      "entropy": 0.8598899506032467,
      "epoch": 0.5241602981006435,
      "grad_norm": 0.1551598608493805,
      "learning_rate": 8.253135829729813e-05,
      "loss": 0.8531,
      "mean_token_accuracy": 0.7492453940212727,
      "num_tokens": 181848838.0,
      "step": 5697
    },
    {
      "entropy": 0.834418248385191,
      "epoch": 0.5242523044720847,
      "grad_norm": 0.16028057038784027,
      "learning_rate": 8.25282914711565e-05,
      "loss": 0.8293,
      "mean_token_accuracy": 0.7580354809761047,
      "num_tokens": 181881055.0,
      "step": 5698
    },
    {
      "entropy": 0.6252011395990849,
      "epoch": 0.524344310843526,
      "grad_norm": 0.14717936515808105,
      "learning_rate": 8.252522464501487e-05,
      "loss": 0.6184,
      "mean_token_accuracy": 0.817102987319231,
      "num_tokens": 181913387.0,
      "step": 5699
    },
    {
      "entropy": 0.7159011550247669,
      "epoch": 0.5244363172149671,
      "grad_norm": 0.16100554168224335,
      "learning_rate": 8.252215781887325e-05,
      "loss": 0.7086,
      "mean_token_accuracy": 0.7906609401106834,
      "num_tokens": 181945722.0,
      "step": 5700
    },
    {
      "entropy": 0.7985095754265785,
      "epoch": 0.5245283235864083,
      "grad_norm": 0.16078974306583405,
      "learning_rate": 8.251909099273163e-05,
      "loss": 0.796,
      "mean_token_accuracy": 0.7650462798774242,
      "num_tokens": 181976494.0,
      "step": 5701
    },
    {
      "entropy": 0.7626328505575657,
      "epoch": 0.5246203299578496,
      "grad_norm": 0.15445570647716522,
      "learning_rate": 8.251602416659e-05,
      "loss": 0.7529,
      "mean_token_accuracy": 0.7754763625562191,
      "num_tokens": 182008471.0,
      "step": 5702
    },
    {
      "entropy": 0.723204342648387,
      "epoch": 0.5247123363292908,
      "grad_norm": 0.15623316168785095,
      "learning_rate": 8.251295734044837e-05,
      "loss": 0.7116,
      "mean_token_accuracy": 0.7902159057557583,
      "num_tokens": 182040215.0,
      "step": 5703
    },
    {
      "entropy": 0.8225897178053856,
      "epoch": 0.5248043427007321,
      "grad_norm": 0.15413391590118408,
      "learning_rate": 8.250989051430674e-05,
      "loss": 0.8255,
      "mean_token_accuracy": 0.7599826417863369,
      "num_tokens": 182072192.0,
      "step": 5704
    },
    {
      "entropy": 0.6702075265347958,
      "epoch": 0.5248963490721732,
      "grad_norm": 0.15023797750473022,
      "learning_rate": 8.250682368816512e-05,
      "loss": 0.6415,
      "mean_token_accuracy": 0.8110703863203526,
      "num_tokens": 182104350.0,
      "step": 5705
    },
    {
      "entropy": 0.809866838157177,
      "epoch": 0.5249883554436144,
      "grad_norm": 0.1649506837129593,
      "learning_rate": 8.25037568620235e-05,
      "loss": 0.8031,
      "mean_token_accuracy": 0.7671602740883827,
      "num_tokens": 182136340.0,
      "step": 5706
    },
    {
      "entropy": 0.8423809558153152,
      "epoch": 0.5250803618150557,
      "grad_norm": 0.15741288661956787,
      "learning_rate": 8.250069003588187e-05,
      "loss": 0.8308,
      "mean_token_accuracy": 0.7579983584582806,
      "num_tokens": 182168440.0,
      "step": 5707
    },
    {
      "entropy": 0.7710105180740356,
      "epoch": 0.5251723681864969,
      "grad_norm": 0.15932422876358032,
      "learning_rate": 8.249762320974025e-05,
      "loss": 0.7744,
      "mean_token_accuracy": 0.7747807390987873,
      "num_tokens": 182199229.0,
      "step": 5708
    },
    {
      "entropy": 0.8394430670887232,
      "epoch": 0.5252643745579382,
      "grad_norm": 0.1701849102973938,
      "learning_rate": 8.249455638359862e-05,
      "loss": 0.8307,
      "mean_token_accuracy": 0.7598147727549076,
      "num_tokens": 182230584.0,
      "step": 5709
    },
    {
      "entropy": 0.735550856217742,
      "epoch": 0.5253563809293793,
      "grad_norm": 0.1541828066110611,
      "learning_rate": 8.249148955745698e-05,
      "loss": 0.7197,
      "mean_token_accuracy": 0.7880874238908291,
      "num_tokens": 182262570.0,
      "step": 5710
    },
    {
      "entropy": 0.9245553500950336,
      "epoch": 0.5254483873008206,
      "grad_norm": 0.17173629999160767,
      "learning_rate": 8.248842273131537e-05,
      "loss": 0.9276,
      "mean_token_accuracy": 0.7335830815136433,
      "num_tokens": 182294411.0,
      "step": 5711
    },
    {
      "entropy": 0.6971088256686926,
      "epoch": 0.5255403936722618,
      "grad_norm": 0.15432144701480865,
      "learning_rate": 8.248535590517375e-05,
      "loss": 0.6864,
      "mean_token_accuracy": 0.7982094436883926,
      "num_tokens": 182326328.0,
      "step": 5712
    },
    {
      "entropy": 0.73395936191082,
      "epoch": 0.525632400043703,
      "grad_norm": 0.14944760501384735,
      "learning_rate": 8.248228907903212e-05,
      "loss": 0.7069,
      "mean_token_accuracy": 0.7892688140273094,
      "num_tokens": 182358157.0,
      "step": 5713
    },
    {
      "entropy": 0.7861816138029099,
      "epoch": 0.5257244064151443,
      "grad_norm": 0.16272170841693878,
      "learning_rate": 8.247922225289048e-05,
      "loss": 0.7688,
      "mean_token_accuracy": 0.7763378843665123,
      "num_tokens": 182390702.0,
      "step": 5714
    },
    {
      "entropy": 0.82231055945158,
      "epoch": 0.5258164127865854,
      "grad_norm": 0.16891993582248688,
      "learning_rate": 8.247615542674885e-05,
      "loss": 0.7947,
      "mean_token_accuracy": 0.7695972993969917,
      "num_tokens": 182422866.0,
      "step": 5715
    },
    {
      "entropy": 0.821586855687201,
      "epoch": 0.5259084191580267,
      "grad_norm": 0.15709219872951508,
      "learning_rate": 8.247308860060725e-05,
      "loss": 0.8029,
      "mean_token_accuracy": 0.766217902302742,
      "num_tokens": 182454599.0,
      "step": 5716
    },
    {
      "entropy": 0.830235043540597,
      "epoch": 0.5260004255294679,
      "grad_norm": 0.15420588850975037,
      "learning_rate": 8.247002177446561e-05,
      "loss": 0.8083,
      "mean_token_accuracy": 0.7652647756040096,
      "num_tokens": 182486411.0,
      "step": 5717
    },
    {
      "entropy": 0.7560066636651754,
      "epoch": 0.5260924319009092,
      "grad_norm": 0.15243610739707947,
      "learning_rate": 8.246695494832398e-05,
      "loss": 0.75,
      "mean_token_accuracy": 0.7794961854815483,
      "num_tokens": 182518969.0,
      "step": 5718
    },
    {
      "entropy": 0.7962136883288622,
      "epoch": 0.5261844382723504,
      "grad_norm": 0.16382592916488647,
      "learning_rate": 8.246388812218235e-05,
      "loss": 0.8134,
      "mean_token_accuracy": 0.7691357620060444,
      "num_tokens": 182550903.0,
      "step": 5719
    },
    {
      "entropy": 0.7530078869313002,
      "epoch": 0.5262764446437915,
      "grad_norm": 0.15735898911952972,
      "learning_rate": 8.246082129604073e-05,
      "loss": 0.743,
      "mean_token_accuracy": 0.7821897640824318,
      "num_tokens": 182583602.0,
      "step": 5720
    },
    {
      "entropy": 0.8360087163746357,
      "epoch": 0.5263684510152328,
      "grad_norm": 0.15538042783737183,
      "learning_rate": 8.245775446989911e-05,
      "loss": 0.8344,
      "mean_token_accuracy": 0.7580294385552406,
      "num_tokens": 182615658.0,
      "step": 5721
    },
    {
      "entropy": 0.8306032717227936,
      "epoch": 0.526460457386674,
      "grad_norm": 0.1585732400417328,
      "learning_rate": 8.245468764375748e-05,
      "loss": 0.8158,
      "mean_token_accuracy": 0.761614978313446,
      "num_tokens": 182648186.0,
      "step": 5722
    },
    {
      "entropy": 0.6927885040640831,
      "epoch": 0.5265524637581153,
      "grad_norm": 0.14936934411525726,
      "learning_rate": 8.245162081761585e-05,
      "loss": 0.6807,
      "mean_token_accuracy": 0.795088566839695,
      "num_tokens": 182680378.0,
      "step": 5723
    },
    {
      "entropy": 0.7420559749007225,
      "epoch": 0.5266444701295565,
      "grad_norm": 0.15661664307117462,
      "learning_rate": 8.244855399147423e-05,
      "loss": 0.7446,
      "mean_token_accuracy": 0.7841933034360409,
      "num_tokens": 182712887.0,
      "step": 5724
    },
    {
      "entropy": 0.782404787838459,
      "epoch": 0.5267364765009976,
      "grad_norm": 0.15601414442062378,
      "learning_rate": 8.24454871653326e-05,
      "loss": 0.7695,
      "mean_token_accuracy": 0.7713127173483372,
      "num_tokens": 182744335.0,
      "step": 5725
    },
    {
      "entropy": 0.8145607933402061,
      "epoch": 0.5268284828724389,
      "grad_norm": 0.1591038852930069,
      "learning_rate": 8.244242033919098e-05,
      "loss": 0.8149,
      "mean_token_accuracy": 0.7601856701076031,
      "num_tokens": 182776130.0,
      "step": 5726
    },
    {
      "entropy": 0.6626110374927521,
      "epoch": 0.5269204892438801,
      "grad_norm": 0.15519241988658905,
      "learning_rate": 8.243935351304935e-05,
      "loss": 0.6337,
      "mean_token_accuracy": 0.811586320400238,
      "num_tokens": 182808654.0,
      "step": 5727
    },
    {
      "entropy": 0.7217606045305729,
      "epoch": 0.5270124956153214,
      "grad_norm": 0.16336527466773987,
      "learning_rate": 8.243628668690773e-05,
      "loss": 0.7031,
      "mean_token_accuracy": 0.793929073959589,
      "num_tokens": 182839626.0,
      "step": 5728
    },
    {
      "entropy": 0.8039039056748152,
      "epoch": 0.5271045019867626,
      "grad_norm": 0.17218926548957825,
      "learning_rate": 8.24332198607661e-05,
      "loss": 0.8187,
      "mean_token_accuracy": 0.7618633396923542,
      "num_tokens": 182871327.0,
      "step": 5729
    },
    {
      "entropy": 0.8211076334118843,
      "epoch": 0.5271965083582038,
      "grad_norm": 0.16556361317634583,
      "learning_rate": 8.243015303462446e-05,
      "loss": 0.8119,
      "mean_token_accuracy": 0.765903826802969,
      "num_tokens": 182903634.0,
      "step": 5730
    },
    {
      "entropy": 0.7255365550518036,
      "epoch": 0.527288514729645,
      "grad_norm": 0.15029233694076538,
      "learning_rate": 8.242708620848285e-05,
      "loss": 0.7188,
      "mean_token_accuracy": 0.7911786213517189,
      "num_tokens": 182936243.0,
      "step": 5731
    },
    {
      "entropy": 0.8284462466835976,
      "epoch": 0.5273805211010862,
      "grad_norm": 0.15908509492874146,
      "learning_rate": 8.242401938234123e-05,
      "loss": 0.8037,
      "mean_token_accuracy": 0.7634780183434486,
      "num_tokens": 182968233.0,
      "step": 5732
    },
    {
      "entropy": 0.742130083963275,
      "epoch": 0.5274725274725275,
      "grad_norm": 0.16566365957260132,
      "learning_rate": 8.24209525561996e-05,
      "loss": 0.7351,
      "mean_token_accuracy": 0.7849501371383667,
      "num_tokens": 183000256.0,
      "step": 5733
    },
    {
      "entropy": 0.7902638651430607,
      "epoch": 0.5275645338439687,
      "grad_norm": 0.15300574898719788,
      "learning_rate": 8.241788573005796e-05,
      "loss": 0.7842,
      "mean_token_accuracy": 0.7691583707928658,
      "num_tokens": 183032041.0,
      "step": 5734
    },
    {
      "entropy": 0.7230034042149782,
      "epoch": 0.5276565402154099,
      "grad_norm": 0.1523195207118988,
      "learning_rate": 8.241481890391633e-05,
      "loss": 0.7119,
      "mean_token_accuracy": 0.7909955233335495,
      "num_tokens": 183064020.0,
      "step": 5735
    },
    {
      "entropy": 0.8178257755935192,
      "epoch": 0.5277485465868511,
      "grad_norm": 0.14907971024513245,
      "learning_rate": 8.241175207777471e-05,
      "loss": 0.7986,
      "mean_token_accuracy": 0.7648104056715965,
      "num_tokens": 183095548.0,
      "step": 5736
    },
    {
      "entropy": 0.7906857114285231,
      "epoch": 0.5278405529582924,
      "grad_norm": 0.16101369261741638,
      "learning_rate": 8.24086852516331e-05,
      "loss": 0.7685,
      "mean_token_accuracy": 0.778038065880537,
      "num_tokens": 183128075.0,
      "step": 5737
    },
    {
      "entropy": 0.8855379372835159,
      "epoch": 0.5279325593297336,
      "grad_norm": 0.1652843952178955,
      "learning_rate": 8.240561842549146e-05,
      "loss": 0.8673,
      "mean_token_accuracy": 0.7505023777484894,
      "num_tokens": 183159154.0,
      "step": 5738
    },
    {
      "entropy": 0.8676322530955076,
      "epoch": 0.5280245657011748,
      "grad_norm": 0.15849988162517548,
      "learning_rate": 8.240255159934984e-05,
      "loss": 0.8337,
      "mean_token_accuracy": 0.756847407668829,
      "num_tokens": 183190893.0,
      "step": 5739
    },
    {
      "entropy": 0.7476626951247454,
      "epoch": 0.528116572072616,
      "grad_norm": 0.15748678147792816,
      "learning_rate": 8.239948477320821e-05,
      "loss": 0.7373,
      "mean_token_accuracy": 0.7869727313518524,
      "num_tokens": 183223510.0,
      "step": 5740
    },
    {
      "entropy": 0.8088985178619623,
      "epoch": 0.5282085784440572,
      "grad_norm": 0.1641015261411667,
      "learning_rate": 8.239641794706658e-05,
      "loss": 0.8121,
      "mean_token_accuracy": 0.7662320993840694,
      "num_tokens": 183255393.0,
      "step": 5741
    },
    {
      "entropy": 0.7855808697640896,
      "epoch": 0.5283005848154985,
      "grad_norm": 0.1767074316740036,
      "learning_rate": 8.239335112092496e-05,
      "loss": 0.7739,
      "mean_token_accuracy": 0.7720897831022739,
      "num_tokens": 183287081.0,
      "step": 5742
    },
    {
      "entropy": 0.8394513726234436,
      "epoch": 0.5283925911869397,
      "grad_norm": 0.16654066741466522,
      "learning_rate": 8.239028429478334e-05,
      "loss": 0.8277,
      "mean_token_accuracy": 0.7588892579078674,
      "num_tokens": 183318802.0,
      "step": 5743
    },
    {
      "entropy": 0.8208817429840565,
      "epoch": 0.528484597558381,
      "grad_norm": 0.15234500169754028,
      "learning_rate": 8.238721746864171e-05,
      "loss": 0.8156,
      "mean_token_accuracy": 0.7586034424602985,
      "num_tokens": 183351257.0,
      "step": 5744
    },
    {
      "entropy": 0.7331023681908846,
      "epoch": 0.5285766039298221,
      "grad_norm": 0.15342238545417786,
      "learning_rate": 8.238415064250008e-05,
      "loss": 0.739,
      "mean_token_accuracy": 0.7853436209261417,
      "num_tokens": 183383708.0,
      "step": 5745
    },
    {
      "entropy": 0.754357224330306,
      "epoch": 0.5286686103012633,
      "grad_norm": 0.16057002544403076,
      "learning_rate": 8.238108381635845e-05,
      "loss": 0.7501,
      "mean_token_accuracy": 0.7781421393156052,
      "num_tokens": 183415494.0,
      "step": 5746
    },
    {
      "entropy": 0.8313983082771301,
      "epoch": 0.5287606166727046,
      "grad_norm": 0.16394264996051788,
      "learning_rate": 8.237801699021683e-05,
      "loss": 0.8288,
      "mean_token_accuracy": 0.7584977932274342,
      "num_tokens": 183447932.0,
      "step": 5747
    },
    {
      "entropy": 0.7889021504670382,
      "epoch": 0.5288526230441458,
      "grad_norm": 0.17030376195907593,
      "learning_rate": 8.237495016407521e-05,
      "loss": 0.7914,
      "mean_token_accuracy": 0.770620983093977,
      "num_tokens": 183479735.0,
      "step": 5748
    },
    {
      "entropy": 0.7926440238952637,
      "epoch": 0.5289446294155871,
      "grad_norm": 0.15700161457061768,
      "learning_rate": 8.237188333793358e-05,
      "loss": 0.7818,
      "mean_token_accuracy": 0.7675088681280613,
      "num_tokens": 183512098.0,
      "step": 5749
    },
    {
      "entropy": 0.7907434124499559,
      "epoch": 0.5290366357870282,
      "grad_norm": 0.15644119679927826,
      "learning_rate": 8.236881651179194e-05,
      "loss": 0.7916,
      "mean_token_accuracy": 0.7724574655294418,
      "num_tokens": 183543516.0,
      "step": 5750
    },
    {
      "entropy": 0.8219442907720804,
      "epoch": 0.5291286421584694,
      "grad_norm": 0.1553800255060196,
      "learning_rate": 8.236574968565033e-05,
      "loss": 0.8143,
      "mean_token_accuracy": 0.7605745755136013,
      "num_tokens": 183576037.0,
      "step": 5751
    },
    {
      "entropy": 0.7025743052363396,
      "epoch": 0.5292206485299107,
      "grad_norm": 0.15745799243450165,
      "learning_rate": 8.236268285950869e-05,
      "loss": 0.7014,
      "mean_token_accuracy": 0.7917101867496967,
      "num_tokens": 183607580.0,
      "step": 5752
    },
    {
      "entropy": 0.7756704716011882,
      "epoch": 0.5293126549013519,
      "grad_norm": 0.1589961051940918,
      "learning_rate": 8.235961603336707e-05,
      "loss": 0.7758,
      "mean_token_accuracy": 0.7750231400132179,
      "num_tokens": 183639604.0,
      "step": 5753
    },
    {
      "entropy": 0.7633599676191807,
      "epoch": 0.5294046612727932,
      "grad_norm": 0.1569579690694809,
      "learning_rate": 8.235654920722544e-05,
      "loss": 0.7549,
      "mean_token_accuracy": 0.777688667178154,
      "num_tokens": 183672190.0,
      "step": 5754
    },
    {
      "entropy": 0.7945439517498016,
      "epoch": 0.5294966676442343,
      "grad_norm": 0.16057761013507843,
      "learning_rate": 8.235348238108382e-05,
      "loss": 0.7947,
      "mean_token_accuracy": 0.7727872580289841,
      "num_tokens": 183704050.0,
      "step": 5755
    },
    {
      "entropy": 0.9294890556484461,
      "epoch": 0.5295886740156756,
      "grad_norm": 0.16348707675933838,
      "learning_rate": 8.235041555494219e-05,
      "loss": 0.9282,
      "mean_token_accuracy": 0.7386168353259563,
      "num_tokens": 183735323.0,
      "step": 5756
    },
    {
      "entropy": 0.7445607408881187,
      "epoch": 0.5296806803871168,
      "grad_norm": 0.16169732809066772,
      "learning_rate": 8.234734872880056e-05,
      "loss": 0.7303,
      "mean_token_accuracy": 0.7829773165285587,
      "num_tokens": 183767175.0,
      "step": 5757
    },
    {
      "entropy": 0.7506187651306391,
      "epoch": 0.529772686758558,
      "grad_norm": 0.15704864263534546,
      "learning_rate": 8.234428190265894e-05,
      "loss": 0.7468,
      "mean_token_accuracy": 0.7801072299480438,
      "num_tokens": 183799457.0,
      "step": 5758
    },
    {
      "entropy": 0.77003313601017,
      "epoch": 0.5298646931299993,
      "grad_norm": 0.15118369460105896,
      "learning_rate": 8.234121507651732e-05,
      "loss": 0.7538,
      "mean_token_accuracy": 0.7799408435821533,
      "num_tokens": 183831782.0,
      "step": 5759
    },
    {
      "entropy": 0.8416907321661711,
      "epoch": 0.5299566995014404,
      "grad_norm": 0.16593335568904877,
      "learning_rate": 8.233814825037569e-05,
      "loss": 0.8231,
      "mean_token_accuracy": 0.7641150169074535,
      "num_tokens": 183863820.0,
      "step": 5760
    },
    {
      "entropy": 0.7721207607537508,
      "epoch": 0.5300487058728817,
      "grad_norm": 0.1577068567276001,
      "learning_rate": 8.233508142423406e-05,
      "loss": 0.7821,
      "mean_token_accuracy": 0.7767854109406471,
      "num_tokens": 183896123.0,
      "step": 5761
    },
    {
      "entropy": 0.7244638372212648,
      "epoch": 0.5301407122443229,
      "grad_norm": 0.15772496163845062,
      "learning_rate": 8.233201459809244e-05,
      "loss": 0.7161,
      "mean_token_accuracy": 0.7937364913523197,
      "num_tokens": 183927831.0,
      "step": 5762
    },
    {
      "entropy": 0.8151576370000839,
      "epoch": 0.5302327186157642,
      "grad_norm": 0.16199104487895966,
      "learning_rate": 8.232894777195082e-05,
      "loss": 0.8016,
      "mean_token_accuracy": 0.7653236538171768,
      "num_tokens": 183960057.0,
      "step": 5763
    },
    {
      "entropy": 0.8882621638476849,
      "epoch": 0.5303247249872054,
      "grad_norm": 0.16495414078235626,
      "learning_rate": 8.232588094580919e-05,
      "loss": 0.8759,
      "mean_token_accuracy": 0.7516795359551907,
      "num_tokens": 183991694.0,
      "step": 5764
    },
    {
      "entropy": 0.8355507757514715,
      "epoch": 0.5304167313586465,
      "grad_norm": 0.15572811663150787,
      "learning_rate": 8.232281411966756e-05,
      "loss": 0.821,
      "mean_token_accuracy": 0.760875653475523,
      "num_tokens": 184023494.0,
      "step": 5765
    },
    {
      "entropy": 0.7863075397908688,
      "epoch": 0.5305087377300878,
      "grad_norm": 0.1589597463607788,
      "learning_rate": 8.231974729352594e-05,
      "loss": 0.7546,
      "mean_token_accuracy": 0.7810706235468388,
      "num_tokens": 184054929.0,
      "step": 5766
    },
    {
      "entropy": 0.8621300384402275,
      "epoch": 0.530600744101529,
      "grad_norm": 0.16335228085517883,
      "learning_rate": 8.23166804673843e-05,
      "loss": 0.845,
      "mean_token_accuracy": 0.7541193589568138,
      "num_tokens": 184087665.0,
      "step": 5767
    },
    {
      "entropy": 0.8615759126842022,
      "epoch": 0.5306927504729703,
      "grad_norm": 0.15992988646030426,
      "learning_rate": 8.231361364124269e-05,
      "loss": 0.8528,
      "mean_token_accuracy": 0.7479202374815941,
      "num_tokens": 184119964.0,
      "step": 5768
    },
    {
      "entropy": 0.7079428043216467,
      "epoch": 0.5307847568444115,
      "grad_norm": 0.16247794032096863,
      "learning_rate": 8.231054681510106e-05,
      "loss": 0.702,
      "mean_token_accuracy": 0.7906114906072617,
      "num_tokens": 184152341.0,
      "step": 5769
    },
    {
      "entropy": 0.761485343798995,
      "epoch": 0.5308767632158526,
      "grad_norm": 0.162199467420578,
      "learning_rate": 8.230747998895944e-05,
      "loss": 0.7404,
      "mean_token_accuracy": 0.7832524366676807,
      "num_tokens": 184183953.0,
      "step": 5770
    },
    {
      "entropy": 0.7647136822342873,
      "epoch": 0.5309687695872939,
      "grad_norm": 0.1608894318342209,
      "learning_rate": 8.23044131628178e-05,
      "loss": 0.7641,
      "mean_token_accuracy": 0.7777692675590515,
      "num_tokens": 184216250.0,
      "step": 5771
    },
    {
      "entropy": 0.9096379242837429,
      "epoch": 0.5310607759587351,
      "grad_norm": 0.17049966752529144,
      "learning_rate": 8.230134633667617e-05,
      "loss": 0.9334,
      "mean_token_accuracy": 0.7309801653027534,
      "num_tokens": 184247754.0,
      "step": 5772
    },
    {
      "entropy": 0.839814092963934,
      "epoch": 0.5311527823301764,
      "grad_norm": 0.15631791949272156,
      "learning_rate": 8.229827951053455e-05,
      "loss": 0.825,
      "mean_token_accuracy": 0.7597974836826324,
      "num_tokens": 184279932.0,
      "step": 5773
    },
    {
      "entropy": 0.7756560407578945,
      "epoch": 0.5312447887016176,
      "grad_norm": 0.1558852344751358,
      "learning_rate": 8.229521268439294e-05,
      "loss": 0.7743,
      "mean_token_accuracy": 0.7734073549509048,
      "num_tokens": 184311750.0,
      "step": 5774
    },
    {
      "entropy": 0.6385166980326176,
      "epoch": 0.5313367950730588,
      "grad_norm": 0.15699374675750732,
      "learning_rate": 8.22921458582513e-05,
      "loss": 0.6251,
      "mean_token_accuracy": 0.8156054466962814,
      "num_tokens": 184343880.0,
      "step": 5775
    },
    {
      "entropy": 0.8113266509026289,
      "epoch": 0.5314288014445,
      "grad_norm": 0.1620723456144333,
      "learning_rate": 8.228907903210967e-05,
      "loss": 0.7923,
      "mean_token_accuracy": 0.7750827670097351,
      "num_tokens": 184375266.0,
      "step": 5776
    },
    {
      "entropy": 0.7881266120821238,
      "epoch": 0.5315208078159412,
      "grad_norm": 0.16766314208507538,
      "learning_rate": 8.228601220596804e-05,
      "loss": 0.775,
      "mean_token_accuracy": 0.7711650654673576,
      "num_tokens": 184407315.0,
      "step": 5777
    },
    {
      "entropy": 0.754855876788497,
      "epoch": 0.5316128141873825,
      "grad_norm": 0.1558908224105835,
      "learning_rate": 8.228294537982642e-05,
      "loss": 0.7262,
      "mean_token_accuracy": 0.7876794002950191,
      "num_tokens": 184438388.0,
      "step": 5778
    },
    {
      "entropy": 0.7058474496006966,
      "epoch": 0.5317048205588237,
      "grad_norm": 0.1510286182165146,
      "learning_rate": 8.22798785536848e-05,
      "loss": 0.6894,
      "mean_token_accuracy": 0.7993376553058624,
      "num_tokens": 184470548.0,
      "step": 5779
    },
    {
      "entropy": 0.8889129683375359,
      "epoch": 0.5317968269302649,
      "grad_norm": 0.1614735871553421,
      "learning_rate": 8.227681172754317e-05,
      "loss": 0.8832,
      "mean_token_accuracy": 0.7462363913655281,
      "num_tokens": 184502460.0,
      "step": 5780
    },
    {
      "entropy": 0.692703727632761,
      "epoch": 0.5318888333017061,
      "grad_norm": 0.14695018529891968,
      "learning_rate": 8.227374490140154e-05,
      "loss": 0.6761,
      "mean_token_accuracy": 0.8015376850962639,
      "num_tokens": 184533967.0,
      "step": 5781
    },
    {
      "entropy": 0.7383866738528013,
      "epoch": 0.5319808396731474,
      "grad_norm": 0.1576484888792038,
      "learning_rate": 8.227067807525992e-05,
      "loss": 0.7227,
      "mean_token_accuracy": 0.7881373874843121,
      "num_tokens": 184565715.0,
      "step": 5782
    },
    {
      "entropy": 0.7011283542960882,
      "epoch": 0.5320728460445886,
      "grad_norm": 0.17187091708183289,
      "learning_rate": 8.226761124911829e-05,
      "loss": 0.714,
      "mean_token_accuracy": 0.7922206223011017,
      "num_tokens": 184597215.0,
      "step": 5783
    },
    {
      "entropy": 0.7228797972202301,
      "epoch": 0.5321648524160298,
      "grad_norm": 0.14683955907821655,
      "learning_rate": 8.226454442297667e-05,
      "loss": 0.7037,
      "mean_token_accuracy": 0.7905731163918972,
      "num_tokens": 184629237.0,
      "step": 5784
    },
    {
      "entropy": 0.7101657725870609,
      "epoch": 0.5322568587874711,
      "grad_norm": 0.14938773214817047,
      "learning_rate": 8.226147759683504e-05,
      "loss": 0.6992,
      "mean_token_accuracy": 0.7944744825363159,
      "num_tokens": 184661159.0,
      "step": 5785
    },
    {
      "entropy": 0.7750175520777702,
      "epoch": 0.5323488651589122,
      "grad_norm": 0.15773066878318787,
      "learning_rate": 8.225841077069342e-05,
      "loss": 0.7593,
      "mean_token_accuracy": 0.779582004994154,
      "num_tokens": 184693195.0,
      "step": 5786
    },
    {
      "entropy": 0.7994931973516941,
      "epoch": 0.5324408715303535,
      "grad_norm": 0.16756078600883484,
      "learning_rate": 8.225534394455179e-05,
      "loss": 0.7777,
      "mean_token_accuracy": 0.7763411849737167,
      "num_tokens": 184725531.0,
      "step": 5787
    },
    {
      "entropy": 0.6484025772660971,
      "epoch": 0.5325328779017947,
      "grad_norm": 0.15101197361946106,
      "learning_rate": 8.225227711841015e-05,
      "loss": 0.6256,
      "mean_token_accuracy": 0.8189463056623936,
      "num_tokens": 184757761.0,
      "step": 5788
    },
    {
      "entropy": 0.8021127972751856,
      "epoch": 0.532624884273236,
      "grad_norm": 0.15812204778194427,
      "learning_rate": 8.224921029226853e-05,
      "loss": 0.8021,
      "mean_token_accuracy": 0.7670551538467407,
      "num_tokens": 184789007.0,
      "step": 5789
    },
    {
      "entropy": 0.7287911800667644,
      "epoch": 0.5327168906446772,
      "grad_norm": 0.157647505402565,
      "learning_rate": 8.224614346612692e-05,
      "loss": 0.7272,
      "mean_token_accuracy": 0.7875673174858093,
      "num_tokens": 184821592.0,
      "step": 5790
    },
    {
      "entropy": 0.6243465952575207,
      "epoch": 0.5328088970161183,
      "grad_norm": 0.15202929079532623,
      "learning_rate": 8.224307663998528e-05,
      "loss": 0.6246,
      "mean_token_accuracy": 0.8159869015216827,
      "num_tokens": 184853416.0,
      "step": 5791
    },
    {
      "entropy": 0.7661978676915169,
      "epoch": 0.5329009033875596,
      "grad_norm": 0.154068723320961,
      "learning_rate": 8.224000981384365e-05,
      "loss": 0.7491,
      "mean_token_accuracy": 0.7763287089765072,
      "num_tokens": 184884705.0,
      "step": 5792
    },
    {
      "entropy": 0.8576608709990978,
      "epoch": 0.5329929097590008,
      "grad_norm": 0.15632180869579315,
      "learning_rate": 8.223694298770202e-05,
      "loss": 0.8567,
      "mean_token_accuracy": 0.7456969134509563,
      "num_tokens": 184916670.0,
      "step": 5793
    },
    {
      "entropy": 0.9565602913498878,
      "epoch": 0.5330849161304421,
      "grad_norm": 0.16451406478881836,
      "learning_rate": 8.223387616156041e-05,
      "loss": 0.964,
      "mean_token_accuracy": 0.7249942570924759,
      "num_tokens": 184948659.0,
      "step": 5794
    },
    {
      "entropy": 0.7953718584030867,
      "epoch": 0.5331769225018833,
      "grad_norm": 0.16317245364189148,
      "learning_rate": 8.223080933541878e-05,
      "loss": 0.7786,
      "mean_token_accuracy": 0.7734069786965847,
      "num_tokens": 184980739.0,
      "step": 5795
    },
    {
      "entropy": 0.7744242306798697,
      "epoch": 0.5332689288733244,
      "grad_norm": 0.16108928620815277,
      "learning_rate": 8.222774250927715e-05,
      "loss": 0.7656,
      "mean_token_accuracy": 0.7770788297057152,
      "num_tokens": 185012999.0,
      "step": 5796
    },
    {
      "entropy": 0.7925224434584379,
      "epoch": 0.5333609352447657,
      "grad_norm": 0.15753114223480225,
      "learning_rate": 8.222467568313553e-05,
      "loss": 0.787,
      "mean_token_accuracy": 0.7695040851831436,
      "num_tokens": 185044991.0,
      "step": 5797
    },
    {
      "entropy": 0.8012562468647957,
      "epoch": 0.5334529416162069,
      "grad_norm": 0.1656530797481537,
      "learning_rate": 8.22216088569939e-05,
      "loss": 0.8102,
      "mean_token_accuracy": 0.766283392906189,
      "num_tokens": 185077233.0,
      "step": 5798
    },
    {
      "entropy": 0.8891956117004156,
      "epoch": 0.5335449479876482,
      "grad_norm": 0.16409164667129517,
      "learning_rate": 8.221854203085228e-05,
      "loss": 0.8899,
      "mean_token_accuracy": 0.7442319914698601,
      "num_tokens": 185109639.0,
      "step": 5799
    },
    {
      "entropy": 0.7759798392653465,
      "epoch": 0.5336369543590894,
      "grad_norm": 0.15283527970314026,
      "learning_rate": 8.221547520471065e-05,
      "loss": 0.7535,
      "mean_token_accuracy": 0.7812405116856098,
      "num_tokens": 185141646.0,
      "step": 5800
    },
    {
      "entropy": 0.8558791168034077,
      "epoch": 0.5337289607305306,
      "grad_norm": 0.15812109410762787,
      "learning_rate": 8.221240837856903e-05,
      "loss": 0.8429,
      "mean_token_accuracy": 0.7541017271578312,
      "num_tokens": 185173406.0,
      "step": 5801
    },
    {
      "entropy": 0.8747222758829594,
      "epoch": 0.5338209671019718,
      "grad_norm": 0.15602602064609528,
      "learning_rate": 8.22093415524274e-05,
      "loss": 0.8529,
      "mean_token_accuracy": 0.7501098401844501,
      "num_tokens": 185205505.0,
      "step": 5802
    },
    {
      "entropy": 0.7581949289888144,
      "epoch": 0.533912973473413,
      "grad_norm": 0.15642602741718292,
      "learning_rate": 8.220627472628577e-05,
      "loss": 0.7488,
      "mean_token_accuracy": 0.777774665504694,
      "num_tokens": 185237343.0,
      "step": 5803
    },
    {
      "entropy": 0.788281999528408,
      "epoch": 0.5340049798448543,
      "grad_norm": 0.16833774745464325,
      "learning_rate": 8.220320790014415e-05,
      "loss": 0.8,
      "mean_token_accuracy": 0.7678908780217171,
      "num_tokens": 185269327.0,
      "step": 5804
    },
    {
      "entropy": 0.8446223363280296,
      "epoch": 0.5340969862162955,
      "grad_norm": 0.1631384938955307,
      "learning_rate": 8.220014107400253e-05,
      "loss": 0.8358,
      "mean_token_accuracy": 0.755794420838356,
      "num_tokens": 185300974.0,
      "step": 5805
    },
    {
      "entropy": 0.7713281232863665,
      "epoch": 0.5341889925877367,
      "grad_norm": 0.15960997343063354,
      "learning_rate": 8.21970742478609e-05,
      "loss": 0.7427,
      "mean_token_accuracy": 0.7828894481062889,
      "num_tokens": 185332493.0,
      "step": 5806
    },
    {
      "entropy": 0.8207415286451578,
      "epoch": 0.5342809989591779,
      "grad_norm": 0.16311292350292206,
      "learning_rate": 8.219400742171926e-05,
      "loss": 0.8082,
      "mean_token_accuracy": 0.7644265294075012,
      "num_tokens": 185363993.0,
      "step": 5807
    },
    {
      "entropy": 0.7605978883802891,
      "epoch": 0.5343730053306192,
      "grad_norm": 0.15698279440402985,
      "learning_rate": 8.219094059557763e-05,
      "loss": 0.765,
      "mean_token_accuracy": 0.7729236893355846,
      "num_tokens": 185396330.0,
      "step": 5808
    },
    {
      "entropy": 0.7481817342340946,
      "epoch": 0.5344650117020604,
      "grad_norm": 0.15782663226127625,
      "learning_rate": 8.218787376943601e-05,
      "loss": 0.7459,
      "mean_token_accuracy": 0.7809555791318417,
      "num_tokens": 185428327.0,
      "step": 5809
    },
    {
      "entropy": 0.8196348771452904,
      "epoch": 0.5345570180735016,
      "grad_norm": 0.16724160313606262,
      "learning_rate": 8.21848069432944e-05,
      "loss": 0.8197,
      "mean_token_accuracy": 0.758535772562027,
      "num_tokens": 185459868.0,
      "step": 5810
    },
    {
      "entropy": 0.7224076967686415,
      "epoch": 0.5346490244449428,
      "grad_norm": 0.1539294719696045,
      "learning_rate": 8.218174011715276e-05,
      "loss": 0.7027,
      "mean_token_accuracy": 0.7938341088593006,
      "num_tokens": 185491615.0,
      "step": 5811
    },
    {
      "entropy": 0.7721347659826279,
      "epoch": 0.534741030816384,
      "grad_norm": 0.16000866889953613,
      "learning_rate": 8.217867329101113e-05,
      "loss": 0.7701,
      "mean_token_accuracy": 0.7739432230591774,
      "num_tokens": 185523798.0,
      "step": 5812
    },
    {
      "entropy": 0.7123412392102182,
      "epoch": 0.5348330371878253,
      "grad_norm": 0.15873681008815765,
      "learning_rate": 8.217560646486951e-05,
      "loss": 0.7198,
      "mean_token_accuracy": 0.7869804985821247,
      "num_tokens": 185555774.0,
      "step": 5813
    },
    {
      "entropy": 0.7439512684941292,
      "epoch": 0.5349250435592665,
      "grad_norm": 0.15809549391269684,
      "learning_rate": 8.217253963872788e-05,
      "loss": 0.7519,
      "mean_token_accuracy": 0.7806971035897732,
      "num_tokens": 185587925.0,
      "step": 5814
    },
    {
      "entropy": 0.712457787245512,
      "epoch": 0.5350170499307078,
      "grad_norm": 0.15401558578014374,
      "learning_rate": 8.216947281258626e-05,
      "loss": 0.6903,
      "mean_token_accuracy": 0.7936535961925983,
      "num_tokens": 185620246.0,
      "step": 5815
    },
    {
      "entropy": 0.7476136442273855,
      "epoch": 0.5351090563021489,
      "grad_norm": 0.16071617603302002,
      "learning_rate": 8.216640598644463e-05,
      "loss": 0.7366,
      "mean_token_accuracy": 0.7824256494641304,
      "num_tokens": 185652617.0,
      "step": 5816
    },
    {
      "entropy": 0.7219822518527508,
      "epoch": 0.5352010626735901,
      "grad_norm": 0.15627896785736084,
      "learning_rate": 8.216333916030301e-05,
      "loss": 0.7153,
      "mean_token_accuracy": 0.7927773967385292,
      "num_tokens": 185685324.0,
      "step": 5817
    },
    {
      "entropy": 0.6895804200321436,
      "epoch": 0.5352930690450314,
      "grad_norm": 0.15463320910930634,
      "learning_rate": 8.216027233416138e-05,
      "loss": 0.6755,
      "mean_token_accuracy": 0.8021435104310513,
      "num_tokens": 185717214.0,
      "step": 5818
    },
    {
      "entropy": 0.7275930941104889,
      "epoch": 0.5353850754164726,
      "grad_norm": 0.15874317288398743,
      "learning_rate": 8.215720550801975e-05,
      "loss": 0.709,
      "mean_token_accuracy": 0.7917944714426994,
      "num_tokens": 185749431.0,
      "step": 5819
    },
    {
      "entropy": 0.7945543993264437,
      "epoch": 0.5354770817879139,
      "grad_norm": 0.16608723998069763,
      "learning_rate": 8.215413868187813e-05,
      "loss": 0.7984,
      "mean_token_accuracy": 0.770495243370533,
      "num_tokens": 185781602.0,
      "step": 5820
    },
    {
      "entropy": 0.8622832093387842,
      "epoch": 0.535569088159355,
      "grad_norm": 0.1567726880311966,
      "learning_rate": 8.215107185573651e-05,
      "loss": 0.8462,
      "mean_token_accuracy": 0.7503764256834984,
      "num_tokens": 185813671.0,
      "step": 5821
    },
    {
      "entropy": 0.7596089374274015,
      "epoch": 0.5356610945307962,
      "grad_norm": 0.15828540921211243,
      "learning_rate": 8.214800502959488e-05,
      "loss": 0.7346,
      "mean_token_accuracy": 0.7875235863029957,
      "num_tokens": 185846210.0,
      "step": 5822
    },
    {
      "entropy": 0.8197419513016939,
      "epoch": 0.5357531009022375,
      "grad_norm": 0.1564648449420929,
      "learning_rate": 8.214493820345325e-05,
      "loss": 0.8099,
      "mean_token_accuracy": 0.7623622082173824,
      "num_tokens": 185878207.0,
      "step": 5823
    },
    {
      "entropy": 0.7712088469415903,
      "epoch": 0.5358451072736787,
      "grad_norm": 0.1649211198091507,
      "learning_rate": 8.214187137731161e-05,
      "loss": 0.741,
      "mean_token_accuracy": 0.7819048389792442,
      "num_tokens": 185910197.0,
      "step": 5824
    },
    {
      "entropy": 0.7264259122312069,
      "epoch": 0.53593711364512,
      "grad_norm": 0.16039322316646576,
      "learning_rate": 8.213880455117e-05,
      "loss": 0.7369,
      "mean_token_accuracy": 0.7826430127024651,
      "num_tokens": 185942546.0,
      "step": 5825
    },
    {
      "entropy": 0.8666879646480083,
      "epoch": 0.5360291200165611,
      "grad_norm": 0.16505859792232513,
      "learning_rate": 8.213573772502838e-05,
      "loss": 0.8682,
      "mean_token_accuracy": 0.7508969567716122,
      "num_tokens": 185975234.0,
      "step": 5826
    },
    {
      "entropy": 0.733958438038826,
      "epoch": 0.5361211263880024,
      "grad_norm": 0.1577727198600769,
      "learning_rate": 8.213267089888674e-05,
      "loss": 0.7088,
      "mean_token_accuracy": 0.7912325859069824,
      "num_tokens": 186006756.0,
      "step": 5827
    },
    {
      "entropy": 0.747377747669816,
      "epoch": 0.5362131327594436,
      "grad_norm": 0.1511530727148056,
      "learning_rate": 8.212960407274513e-05,
      "loss": 0.7235,
      "mean_token_accuracy": 0.7868692241609097,
      "num_tokens": 186039467.0,
      "step": 5828
    },
    {
      "entropy": 0.7768059931695461,
      "epoch": 0.5363051391308848,
      "grad_norm": 0.1615045815706253,
      "learning_rate": 8.21265372466035e-05,
      "loss": 0.7615,
      "mean_token_accuracy": 0.7761957868933678,
      "num_tokens": 186071254.0,
      "step": 5829
    },
    {
      "entropy": 0.7747312281280756,
      "epoch": 0.5363971455023261,
      "grad_norm": 0.15529121458530426,
      "learning_rate": 8.212347042046186e-05,
      "loss": 0.7566,
      "mean_token_accuracy": 0.7809642516076565,
      "num_tokens": 186103692.0,
      "step": 5830
    },
    {
      "entropy": 0.7500335313379765,
      "epoch": 0.5364891518737672,
      "grad_norm": 0.1652008295059204,
      "learning_rate": 8.212040359432024e-05,
      "loss": 0.7599,
      "mean_token_accuracy": 0.7743450999259949,
      "num_tokens": 186135668.0,
      "step": 5831
    },
    {
      "entropy": 0.7558564115315676,
      "epoch": 0.5365811582452085,
      "grad_norm": 0.1578039675951004,
      "learning_rate": 8.211733676817862e-05,
      "loss": 0.7512,
      "mean_token_accuracy": 0.7793004736304283,
      "num_tokens": 186166993.0,
      "step": 5832
    },
    {
      "entropy": 0.7058660984039307,
      "epoch": 0.5366731646166497,
      "grad_norm": 0.15869933366775513,
      "learning_rate": 8.211426994203699e-05,
      "loss": 0.7136,
      "mean_token_accuracy": 0.7898683063685894,
      "num_tokens": 186198165.0,
      "step": 5833
    },
    {
      "entropy": 0.7051579747349024,
      "epoch": 0.536765170988091,
      "grad_norm": 0.1491931974887848,
      "learning_rate": 8.211120311589536e-05,
      "loss": 0.6869,
      "mean_token_accuracy": 0.8002192415297031,
      "num_tokens": 186230395.0,
      "step": 5834
    },
    {
      "entropy": 0.8138199746608734,
      "epoch": 0.5368571773595322,
      "grad_norm": 0.16188111901283264,
      "learning_rate": 8.210813628975373e-05,
      "loss": 0.8087,
      "mean_token_accuracy": 0.7627164758741856,
      "num_tokens": 186261615.0,
      "step": 5835
    },
    {
      "entropy": 0.7136205323040485,
      "epoch": 0.5369491837309733,
      "grad_norm": 0.15461748838424683,
      "learning_rate": 8.210506946361212e-05,
      "loss": 0.6917,
      "mean_token_accuracy": 0.7934921607375145,
      "num_tokens": 186293261.0,
      "step": 5836
    },
    {
      "entropy": 0.8933968227356672,
      "epoch": 0.5370411901024146,
      "grad_norm": 0.15976552665233612,
      "learning_rate": 8.210200263747049e-05,
      "loss": 0.8918,
      "mean_token_accuracy": 0.7404701858758926,
      "num_tokens": 186325548.0,
      "step": 5837
    },
    {
      "entropy": 0.6722179111093283,
      "epoch": 0.5371331964738558,
      "grad_norm": 0.15226596593856812,
      "learning_rate": 8.209893581132886e-05,
      "loss": 0.6564,
      "mean_token_accuracy": 0.798347145318985,
      "num_tokens": 186357388.0,
      "step": 5838
    },
    {
      "entropy": 0.7201992496848106,
      "epoch": 0.5372252028452971,
      "grad_norm": 0.1588502824306488,
      "learning_rate": 8.209586898518723e-05,
      "loss": 0.7131,
      "mean_token_accuracy": 0.7895253747701645,
      "num_tokens": 186389772.0,
      "step": 5839
    },
    {
      "entropy": 0.8920835461467505,
      "epoch": 0.5373172092167383,
      "grad_norm": 0.16072072088718414,
      "learning_rate": 8.209280215904561e-05,
      "loss": 0.8818,
      "mean_token_accuracy": 0.7430052272975445,
      "num_tokens": 186421782.0,
      "step": 5840
    },
    {
      "entropy": 0.8426052257418633,
      "epoch": 0.5374092155881794,
      "grad_norm": 0.16033776104450226,
      "learning_rate": 8.208973533290399e-05,
      "loss": 0.8419,
      "mean_token_accuracy": 0.7545470595359802,
      "num_tokens": 186453703.0,
      "step": 5841
    },
    {
      "entropy": 0.6156984930858016,
      "epoch": 0.5375012219596207,
      "grad_norm": 0.14481262862682343,
      "learning_rate": 8.208666850676236e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.8191554918885231,
      "num_tokens": 186486208.0,
      "step": 5842
    },
    {
      "entropy": 0.7553789652884007,
      "epoch": 0.5375932283310619,
      "grad_norm": 0.15636220574378967,
      "learning_rate": 8.208360168062072e-05,
      "loss": 0.7587,
      "mean_token_accuracy": 0.7766820341348648,
      "num_tokens": 186518218.0,
      "step": 5843
    },
    {
      "entropy": 0.7869174163788557,
      "epoch": 0.5376852347025032,
      "grad_norm": 0.16337697207927704,
      "learning_rate": 8.20805348544791e-05,
      "loss": 0.8067,
      "mean_token_accuracy": 0.7667366527020931,
      "num_tokens": 186549793.0,
      "step": 5844
    },
    {
      "entropy": 0.7396007534116507,
      "epoch": 0.5377772410739444,
      "grad_norm": 0.1515296846628189,
      "learning_rate": 8.207746802833747e-05,
      "loss": 0.7389,
      "mean_token_accuracy": 0.7812394760549068,
      "num_tokens": 186581515.0,
      "step": 5845
    },
    {
      "entropy": 0.8286797218024731,
      "epoch": 0.5378692474453856,
      "grad_norm": 0.1595921516418457,
      "learning_rate": 8.207440120219586e-05,
      "loss": 0.8107,
      "mean_token_accuracy": 0.759458277374506,
      "num_tokens": 186613533.0,
      "step": 5846
    },
    {
      "entropy": 0.726499555632472,
      "epoch": 0.5379612538168268,
      "grad_norm": 0.15632034838199615,
      "learning_rate": 8.207133437605422e-05,
      "loss": 0.7108,
      "mean_token_accuracy": 0.7854499667882919,
      "num_tokens": 186645756.0,
      "step": 5847
    },
    {
      "entropy": 0.7470642309635878,
      "epoch": 0.538053260188268,
      "grad_norm": 0.1487291306257248,
      "learning_rate": 8.20682675499126e-05,
      "loss": 0.7414,
      "mean_token_accuracy": 0.7802420817315578,
      "num_tokens": 186678380.0,
      "step": 5848
    },
    {
      "entropy": 0.767560413107276,
      "epoch": 0.5381452665597093,
      "grad_norm": 0.15363140404224396,
      "learning_rate": 8.206520072377097e-05,
      "loss": 0.7392,
      "mean_token_accuracy": 0.7873119786381721,
      "num_tokens": 186710308.0,
      "step": 5849
    },
    {
      "entropy": 0.9020460546016693,
      "epoch": 0.5382372729311505,
      "grad_norm": 0.1696753352880478,
      "learning_rate": 8.206213389762934e-05,
      "loss": 0.9086,
      "mean_token_accuracy": 0.739350076764822,
      "num_tokens": 186742111.0,
      "step": 5850
    },
    {
      "entropy": 0.8374765180051327,
      "epoch": 0.5383292793025917,
      "grad_norm": 0.17332322895526886,
      "learning_rate": 8.205906707148772e-05,
      "loss": 0.8322,
      "mean_token_accuracy": 0.7558544687926769,
      "num_tokens": 186773944.0,
      "step": 5851
    },
    {
      "entropy": 0.8413414880633354,
      "epoch": 0.5384212856740329,
      "grad_norm": 0.15802694857120514,
      "learning_rate": 8.20560002453461e-05,
      "loss": 0.8294,
      "mean_token_accuracy": 0.7608566582202911,
      "num_tokens": 186805300.0,
      "step": 5852
    },
    {
      "entropy": 0.747914019972086,
      "epoch": 0.5385132920454742,
      "grad_norm": 0.14644400775432587,
      "learning_rate": 8.205293341920447e-05,
      "loss": 0.7178,
      "mean_token_accuracy": 0.7872147634625435,
      "num_tokens": 186837636.0,
      "step": 5853
    },
    {
      "entropy": 0.8528961259871721,
      "epoch": 0.5386052984169154,
      "grad_norm": 0.15719470381736755,
      "learning_rate": 8.204986659306284e-05,
      "loss": 0.8355,
      "mean_token_accuracy": 0.7570614069700241,
      "num_tokens": 186869322.0,
      "step": 5854
    },
    {
      "entropy": 0.8252154476940632,
      "epoch": 0.5386973047883566,
      "grad_norm": 0.16003841161727905,
      "learning_rate": 8.204679976692121e-05,
      "loss": 0.8225,
      "mean_token_accuracy": 0.7637854367494583,
      "num_tokens": 186901660.0,
      "step": 5855
    },
    {
      "entropy": 0.7645006030797958,
      "epoch": 0.5387893111597978,
      "grad_norm": 0.16269372403621674,
      "learning_rate": 8.204373294077959e-05,
      "loss": 0.7462,
      "mean_token_accuracy": 0.7851902693510056,
      "num_tokens": 186933039.0,
      "step": 5856
    },
    {
      "entropy": 0.8265482038259506,
      "epoch": 0.538881317531239,
      "grad_norm": 0.1586083024740219,
      "learning_rate": 8.204066611463797e-05,
      "loss": 0.8165,
      "mean_token_accuracy": 0.7624561749398708,
      "num_tokens": 186965042.0,
      "step": 5857
    },
    {
      "entropy": 0.8122107237577438,
      "epoch": 0.5389733239026803,
      "grad_norm": 0.15972082316875458,
      "learning_rate": 8.203759928849634e-05,
      "loss": 0.7923,
      "mean_token_accuracy": 0.7701973468065262,
      "num_tokens": 186996326.0,
      "step": 5858
    },
    {
      "entropy": 0.7466899082064629,
      "epoch": 0.5390653302741215,
      "grad_norm": 0.15750065445899963,
      "learning_rate": 8.203453246235472e-05,
      "loss": 0.7364,
      "mean_token_accuracy": 0.7812473066151142,
      "num_tokens": 187027869.0,
      "step": 5859
    },
    {
      "entropy": 0.8188632093369961,
      "epoch": 0.5391573366455628,
      "grad_norm": 0.16655106842517853,
      "learning_rate": 8.203146563621309e-05,
      "loss": 0.8256,
      "mean_token_accuracy": 0.760823018848896,
      "num_tokens": 187059995.0,
      "step": 5860
    },
    {
      "entropy": 0.6913590505719185,
      "epoch": 0.5392493430170039,
      "grad_norm": 0.15425269305706024,
      "learning_rate": 8.202839881007146e-05,
      "loss": 0.671,
      "mean_token_accuracy": 0.801261406391859,
      "num_tokens": 187091858.0,
      "step": 5861
    },
    {
      "entropy": 0.707593047991395,
      "epoch": 0.5393413493884451,
      "grad_norm": 0.1650356948375702,
      "learning_rate": 8.202533198392984e-05,
      "loss": 0.7028,
      "mean_token_accuracy": 0.7959205731749535,
      "num_tokens": 187123468.0,
      "step": 5862
    },
    {
      "entropy": 0.7966682724654675,
      "epoch": 0.5394333557598864,
      "grad_norm": 0.15156815946102142,
      "learning_rate": 8.202226515778822e-05,
      "loss": 0.7867,
      "mean_token_accuracy": 0.7672342918813229,
      "num_tokens": 187155397.0,
      "step": 5863
    },
    {
      "entropy": 0.6713064294308424,
      "epoch": 0.5395253621313276,
      "grad_norm": 0.15347035229206085,
      "learning_rate": 8.201919833164659e-05,
      "loss": 0.6706,
      "mean_token_accuracy": 0.7995862849056721,
      "num_tokens": 187186981.0,
      "step": 5864
    },
    {
      "entropy": 0.7118583489209414,
      "epoch": 0.5396173685027689,
      "grad_norm": 0.14986930787563324,
      "learning_rate": 8.201613150550495e-05,
      "loss": 0.695,
      "mean_token_accuracy": 0.7950960136950016,
      "num_tokens": 187219435.0,
      "step": 5865
    },
    {
      "entropy": 0.6935636028647423,
      "epoch": 0.53970937487421,
      "grad_norm": 0.15623502433300018,
      "learning_rate": 8.201306467936332e-05,
      "loss": 0.6963,
      "mean_token_accuracy": 0.7934549488127232,
      "num_tokens": 187251704.0,
      "step": 5866
    },
    {
      "entropy": 0.8328514881432056,
      "epoch": 0.5398013812456512,
      "grad_norm": 0.16100424528121948,
      "learning_rate": 8.20099978532217e-05,
      "loss": 0.8258,
      "mean_token_accuracy": 0.7538568526506424,
      "num_tokens": 187282938.0,
      "step": 5867
    },
    {
      "entropy": 0.694631764665246,
      "epoch": 0.5398933876170925,
      "grad_norm": 0.15304143726825714,
      "learning_rate": 8.200693102708008e-05,
      "loss": 0.6679,
      "mean_token_accuracy": 0.800274420529604,
      "num_tokens": 187315132.0,
      "step": 5868
    },
    {
      "entropy": 0.7723137158900499,
      "epoch": 0.5399853939885337,
      "grad_norm": 0.16311044991016388,
      "learning_rate": 8.200386420093845e-05,
      "loss": 0.7727,
      "mean_token_accuracy": 0.7768317423760891,
      "num_tokens": 187347196.0,
      "step": 5869
    },
    {
      "entropy": 0.770512642338872,
      "epoch": 0.540077400359975,
      "grad_norm": 0.16118472814559937,
      "learning_rate": 8.200079737479682e-05,
      "loss": 0.7591,
      "mean_token_accuracy": 0.7762195989489555,
      "num_tokens": 187378503.0,
      "step": 5870
    },
    {
      "entropy": 0.7380022015422583,
      "epoch": 0.5401694067314161,
      "grad_norm": 0.14892563223838806,
      "learning_rate": 8.19977305486552e-05,
      "loss": 0.7367,
      "mean_token_accuracy": 0.7867210358381271,
      "num_tokens": 187410575.0,
      "step": 5871
    },
    {
      "entropy": 0.785915507003665,
      "epoch": 0.5402614131028574,
      "grad_norm": 0.15104743838310242,
      "learning_rate": 8.199466372251357e-05,
      "loss": 0.7694,
      "mean_token_accuracy": 0.7717872858047485,
      "num_tokens": 187442888.0,
      "step": 5872
    },
    {
      "entropy": 0.7735243812203407,
      "epoch": 0.5403534194742986,
      "grad_norm": 0.15911535918712616,
      "learning_rate": 8.199159689637195e-05,
      "loss": 0.761,
      "mean_token_accuracy": 0.7751880697906017,
      "num_tokens": 187474828.0,
      "step": 5873
    },
    {
      "entropy": 0.8536039050668478,
      "epoch": 0.5404454258457398,
      "grad_norm": 0.16479827463626862,
      "learning_rate": 8.198853007023032e-05,
      "loss": 0.8332,
      "mean_token_accuracy": 0.7584900893270969,
      "num_tokens": 187505894.0,
      "step": 5874
    },
    {
      "entropy": 0.7854172345250845,
      "epoch": 0.5405374322171811,
      "grad_norm": 0.16125966608524323,
      "learning_rate": 8.19854632440887e-05,
      "loss": 0.7837,
      "mean_token_accuracy": 0.7744393795728683,
      "num_tokens": 187537906.0,
      "step": 5875
    },
    {
      "entropy": 0.7825795356184244,
      "epoch": 0.5406294385886222,
      "grad_norm": 0.15767519176006317,
      "learning_rate": 8.198239641794707e-05,
      "loss": 0.7453,
      "mean_token_accuracy": 0.784321703016758,
      "num_tokens": 187570482.0,
      "step": 5876
    },
    {
      "entropy": 0.8421256113797426,
      "epoch": 0.5407214449600635,
      "grad_norm": 0.1625872552394867,
      "learning_rate": 8.197932959180545e-05,
      "loss": 0.834,
      "mean_token_accuracy": 0.7600085996091366,
      "num_tokens": 187602540.0,
      "step": 5877
    },
    {
      "entropy": 0.7419758308678865,
      "epoch": 0.5408134513315047,
      "grad_norm": 0.14451298117637634,
      "learning_rate": 8.197626276566382e-05,
      "loss": 0.7197,
      "mean_token_accuracy": 0.7870784848928452,
      "num_tokens": 187634860.0,
      "step": 5878
    },
    {
      "entropy": 0.7857984285801649,
      "epoch": 0.540905457702946,
      "grad_norm": 0.16802562773227692,
      "learning_rate": 8.19731959395222e-05,
      "loss": 0.7993,
      "mean_token_accuracy": 0.7723062075674534,
      "num_tokens": 187667036.0,
      "step": 5879
    },
    {
      "entropy": 0.7289801295846701,
      "epoch": 0.5409974640743872,
      "grad_norm": 0.15692025423049927,
      "learning_rate": 8.197012911338057e-05,
      "loss": 0.7199,
      "mean_token_accuracy": 0.791054017841816,
      "num_tokens": 187698875.0,
      "step": 5880
    },
    {
      "entropy": 0.7041099537163973,
      "epoch": 0.5410894704458283,
      "grad_norm": 0.15416887402534485,
      "learning_rate": 8.196706228723893e-05,
      "loss": 0.6981,
      "mean_token_accuracy": 0.7969805561006069,
      "num_tokens": 187730071.0,
      "step": 5881
    },
    {
      "entropy": 0.8432863764464855,
      "epoch": 0.5411814768172696,
      "grad_norm": 0.16081881523132324,
      "learning_rate": 8.196399546109732e-05,
      "loss": 0.8478,
      "mean_token_accuracy": 0.7563696652650833,
      "num_tokens": 187761138.0,
      "step": 5882
    },
    {
      "entropy": 0.8001229949295521,
      "epoch": 0.5412734831887108,
      "grad_norm": 0.16103018820285797,
      "learning_rate": 8.19609286349557e-05,
      "loss": 0.8156,
      "mean_token_accuracy": 0.7647743746638298,
      "num_tokens": 187792895.0,
      "step": 5883
    },
    {
      "entropy": 0.8786112889647484,
      "epoch": 0.5413654895601521,
      "grad_norm": 0.16319607198238373,
      "learning_rate": 8.195786180881407e-05,
      "loss": 0.8953,
      "mean_token_accuracy": 0.7427688986063004,
      "num_tokens": 187824705.0,
      "step": 5884
    },
    {
      "entropy": 0.7098510377109051,
      "epoch": 0.5414574959315933,
      "grad_norm": 0.15314920246601105,
      "learning_rate": 8.195479498267243e-05,
      "loss": 0.7102,
      "mean_token_accuracy": 0.7935142889618874,
      "num_tokens": 187856898.0,
      "step": 5885
    },
    {
      "entropy": 0.8759337235242128,
      "epoch": 0.5415495023030344,
      "grad_norm": 0.16363245248794556,
      "learning_rate": 8.19517281565308e-05,
      "loss": 0.8739,
      "mean_token_accuracy": 0.7472553625702858,
      "num_tokens": 187889087.0,
      "step": 5886
    },
    {
      "entropy": 0.7964016459882259,
      "epoch": 0.5416415086744757,
      "grad_norm": 0.16741923987865448,
      "learning_rate": 8.194866133038918e-05,
      "loss": 0.8002,
      "mean_token_accuracy": 0.7689053453505039,
      "num_tokens": 187921111.0,
      "step": 5887
    },
    {
      "entropy": 0.7641935050487518,
      "epoch": 0.5417335150459169,
      "grad_norm": 0.15689386427402496,
      "learning_rate": 8.194559450424756e-05,
      "loss": 0.7569,
      "mean_token_accuracy": 0.7784020006656647,
      "num_tokens": 187952749.0,
      "step": 5888
    },
    {
      "entropy": 0.8202106095850468,
      "epoch": 0.5418255214173582,
      "grad_norm": 0.1583252251148224,
      "learning_rate": 8.194252767810593e-05,
      "loss": 0.7982,
      "mean_token_accuracy": 0.7712692804634571,
      "num_tokens": 187984783.0,
      "step": 5889
    },
    {
      "entropy": 0.7575062364339828,
      "epoch": 0.5419175277887994,
      "grad_norm": 0.1476619839668274,
      "learning_rate": 8.193946085196431e-05,
      "loss": 0.7371,
      "mean_token_accuracy": 0.7801621109247208,
      "num_tokens": 188016029.0,
      "step": 5890
    },
    {
      "entropy": 0.8780279736965895,
      "epoch": 0.5420095341602406,
      "grad_norm": 0.17466923594474792,
      "learning_rate": 8.193639402582268e-05,
      "loss": 0.8713,
      "mean_token_accuracy": 0.7525031752884388,
      "num_tokens": 188047998.0,
      "step": 5891
    },
    {
      "entropy": 0.831896897405386,
      "epoch": 0.5421015405316818,
      "grad_norm": 0.1603645384311676,
      "learning_rate": 8.193332719968105e-05,
      "loss": 0.8337,
      "mean_token_accuracy": 0.7639331668615341,
      "num_tokens": 188079626.0,
      "step": 5892
    },
    {
      "entropy": 0.9563190042972565,
      "epoch": 0.542193546903123,
      "grad_norm": 0.16998320817947388,
      "learning_rate": 8.193026037353943e-05,
      "loss": 0.9432,
      "mean_token_accuracy": 0.7351330295205116,
      "num_tokens": 188111624.0,
      "step": 5893
    },
    {
      "entropy": 0.813875362277031,
      "epoch": 0.5422855532745643,
      "grad_norm": 0.15689437091350555,
      "learning_rate": 8.192719354739781e-05,
      "loss": 0.7949,
      "mean_token_accuracy": 0.7679758369922638,
      "num_tokens": 188142588.0,
      "step": 5894
    },
    {
      "entropy": 0.811417106539011,
      "epoch": 0.5423775596460055,
      "grad_norm": 0.1595936417579651,
      "learning_rate": 8.192412672125618e-05,
      "loss": 0.8048,
      "mean_token_accuracy": 0.7639999985694885,
      "num_tokens": 188174692.0,
      "step": 5895
    },
    {
      "entropy": 0.7436172403395176,
      "epoch": 0.5424695660174467,
      "grad_norm": 0.15250065922737122,
      "learning_rate": 8.192105989511455e-05,
      "loss": 0.7181,
      "mean_token_accuracy": 0.7847483865916729,
      "num_tokens": 188207030.0,
      "step": 5896
    },
    {
      "entropy": 0.7611798103898764,
      "epoch": 0.5425615723888879,
      "grad_norm": 0.1578163206577301,
      "learning_rate": 8.191799306897292e-05,
      "loss": 0.7458,
      "mean_token_accuracy": 0.7789605483412743,
      "num_tokens": 188239119.0,
      "step": 5897
    },
    {
      "entropy": 0.7894473019987345,
      "epoch": 0.5426535787603292,
      "grad_norm": 0.15988454222679138,
      "learning_rate": 8.19149262428313e-05,
      "loss": 0.7913,
      "mean_token_accuracy": 0.7720785290002823,
      "num_tokens": 188271091.0,
      "step": 5898
    },
    {
      "entropy": 0.7239998187869787,
      "epoch": 0.5427455851317704,
      "grad_norm": 0.15587802231311798,
      "learning_rate": 8.191185941668968e-05,
      "loss": 0.7328,
      "mean_token_accuracy": 0.7840808779001236,
      "num_tokens": 188303254.0,
      "step": 5899
    },
    {
      "entropy": 0.6737504620105028,
      "epoch": 0.5428375915032116,
      "grad_norm": 0.15539634227752686,
      "learning_rate": 8.190879259054805e-05,
      "loss": 0.671,
      "mean_token_accuracy": 0.7996317893266678,
      "num_tokens": 188335992.0,
      "step": 5900
    },
    {
      "entropy": 0.7803961932659149,
      "epoch": 0.5429295978746528,
      "grad_norm": 0.1641237586736679,
      "learning_rate": 8.190572576440641e-05,
      "loss": 0.789,
      "mean_token_accuracy": 0.7716703005135059,
      "num_tokens": 188367751.0,
      "step": 5901
    },
    {
      "entropy": 0.7827111799269915,
      "epoch": 0.543021604246094,
      "grad_norm": 0.16071830689907074,
      "learning_rate": 8.19026589382648e-05,
      "loss": 0.7854,
      "mean_token_accuracy": 0.7667513750493526,
      "num_tokens": 188399897.0,
      "step": 5902
    },
    {
      "entropy": 0.7843384705483913,
      "epoch": 0.5431136106175353,
      "grad_norm": 0.16073404252529144,
      "learning_rate": 8.189959211212316e-05,
      "loss": 0.7693,
      "mean_token_accuracy": 0.777114674448967,
      "num_tokens": 188431459.0,
      "step": 5903
    },
    {
      "entropy": 0.7371756695210934,
      "epoch": 0.5432056169889765,
      "grad_norm": 0.15568803250789642,
      "learning_rate": 8.189652528598154e-05,
      "loss": 0.7117,
      "mean_token_accuracy": 0.7895600385963917,
      "num_tokens": 188462442.0,
      "step": 5904
    },
    {
      "entropy": 0.6826624218374491,
      "epoch": 0.5432976233604178,
      "grad_norm": 0.15375441312789917,
      "learning_rate": 8.189345845983991e-05,
      "loss": 0.6514,
      "mean_token_accuracy": 0.8082660771906376,
      "num_tokens": 188494975.0,
      "step": 5905
    },
    {
      "entropy": 0.8226679004728794,
      "epoch": 0.5433896297318589,
      "grad_norm": 0.15510085225105286,
      "learning_rate": 8.18903916336983e-05,
      "loss": 0.796,
      "mean_token_accuracy": 0.7682816758751869,
      "num_tokens": 188526434.0,
      "step": 5906
    },
    {
      "entropy": 0.8487202040851116,
      "epoch": 0.5434816361033001,
      "grad_norm": 0.1589908003807068,
      "learning_rate": 8.188732480755666e-05,
      "loss": 0.8118,
      "mean_token_accuracy": 0.761968731880188,
      "num_tokens": 188559202.0,
      "step": 5907
    },
    {
      "entropy": 0.7577794771641493,
      "epoch": 0.5435736424747414,
      "grad_norm": 0.16000071167945862,
      "learning_rate": 8.188425798141503e-05,
      "loss": 0.7594,
      "mean_token_accuracy": 0.77576994150877,
      "num_tokens": 188591412.0,
      "step": 5908
    },
    {
      "entropy": 0.8353705182671547,
      "epoch": 0.5436656488461826,
      "grad_norm": 0.1569991558790207,
      "learning_rate": 8.188119115527341e-05,
      "loss": 0.8206,
      "mean_token_accuracy": 0.7608648464083672,
      "num_tokens": 188622951.0,
      "step": 5909
    },
    {
      "entropy": 0.7489158939570189,
      "epoch": 0.5437576552176239,
      "grad_norm": 0.1490589827299118,
      "learning_rate": 8.187812432913179e-05,
      "loss": 0.7263,
      "mean_token_accuracy": 0.78627809882164,
      "num_tokens": 188655314.0,
      "step": 5910
    },
    {
      "entropy": 0.8013288956135511,
      "epoch": 0.543849661589065,
      "grad_norm": 0.16687358915805817,
      "learning_rate": 8.187505750299016e-05,
      "loss": 0.7807,
      "mean_token_accuracy": 0.7739051580429077,
      "num_tokens": 188687292.0,
      "step": 5911
    },
    {
      "entropy": 0.771411806344986,
      "epoch": 0.5439416679605062,
      "grad_norm": 0.15679164230823517,
      "learning_rate": 8.187199067684853e-05,
      "loss": 0.7689,
      "mean_token_accuracy": 0.7733996622264385,
      "num_tokens": 188719797.0,
      "step": 5912
    },
    {
      "entropy": 0.7000000681728125,
      "epoch": 0.5440336743319475,
      "grad_norm": 0.16559827327728271,
      "learning_rate": 8.18689238507069e-05,
      "loss": 0.6977,
      "mean_token_accuracy": 0.7969379387795925,
      "num_tokens": 188752281.0,
      "step": 5913
    },
    {
      "entropy": 0.7741272058337927,
      "epoch": 0.5441256807033887,
      "grad_norm": 0.15757760405540466,
      "learning_rate": 8.186585702456529e-05,
      "loss": 0.7699,
      "mean_token_accuracy": 0.7767226062715054,
      "num_tokens": 188784596.0,
      "step": 5914
    },
    {
      "entropy": 0.747570339590311,
      "epoch": 0.54421768707483,
      "grad_norm": 0.16452902555465698,
      "learning_rate": 8.186279019842366e-05,
      "loss": 0.7337,
      "mean_token_accuracy": 0.7827056907117367,
      "num_tokens": 188815564.0,
      "step": 5915
    },
    {
      "entropy": 0.8234516344964504,
      "epoch": 0.5443096934462711,
      "grad_norm": 0.17261259257793427,
      "learning_rate": 8.185972337228203e-05,
      "loss": 0.8112,
      "mean_token_accuracy": 0.7684747092425823,
      "num_tokens": 188847334.0,
      "step": 5916
    },
    {
      "entropy": 0.8607780672609806,
      "epoch": 0.5444016998177124,
      "grad_norm": 0.16004350781440735,
      "learning_rate": 8.18566565461404e-05,
      "loss": 0.8579,
      "mean_token_accuracy": 0.7482473142445087,
      "num_tokens": 188878938.0,
      "step": 5917
    },
    {
      "entropy": 0.826213177293539,
      "epoch": 0.5444937061891536,
      "grad_norm": 0.15926475822925568,
      "learning_rate": 8.185358971999878e-05,
      "loss": 0.8271,
      "mean_token_accuracy": 0.7602788098156452,
      "num_tokens": 188911122.0,
      "step": 5918
    },
    {
      "entropy": 0.7459573373198509,
      "epoch": 0.5445857125605948,
      "grad_norm": 0.1667264699935913,
      "learning_rate": 8.185052289385716e-05,
      "loss": 0.7456,
      "mean_token_accuracy": 0.7835785187780857,
      "num_tokens": 188942613.0,
      "step": 5919
    },
    {
      "entropy": 0.7287417035549879,
      "epoch": 0.5446777189320361,
      "grad_norm": 0.16512829065322876,
      "learning_rate": 8.184745606771553e-05,
      "loss": 0.7274,
      "mean_token_accuracy": 0.7858911976218224,
      "num_tokens": 188974977.0,
      "step": 5920
    },
    {
      "entropy": 0.7054292242974043,
      "epoch": 0.5447697253034772,
      "grad_norm": 0.159592866897583,
      "learning_rate": 8.184438924157391e-05,
      "loss": 0.6879,
      "mean_token_accuracy": 0.7961764819920063,
      "num_tokens": 189007182.0,
      "step": 5921
    },
    {
      "entropy": 0.7231014259159565,
      "epoch": 0.5448617316749185,
      "grad_norm": 0.151780903339386,
      "learning_rate": 8.184132241543227e-05,
      "loss": 0.6796,
      "mean_token_accuracy": 0.7994843050837517,
      "num_tokens": 189039504.0,
      "step": 5922
    },
    {
      "entropy": 0.7166458256542683,
      "epoch": 0.5449537380463597,
      "grad_norm": 0.156014546751976,
      "learning_rate": 8.183825558929064e-05,
      "loss": 0.7184,
      "mean_token_accuracy": 0.7900278940796852,
      "num_tokens": 189070721.0,
      "step": 5923
    },
    {
      "entropy": 0.6785057075321674,
      "epoch": 0.545045744417801,
      "grad_norm": 0.15458762645721436,
      "learning_rate": 8.183518876314902e-05,
      "loss": 0.6714,
      "mean_token_accuracy": 0.8043172769248486,
      "num_tokens": 189103064.0,
      "step": 5924
    },
    {
      "entropy": 0.7915324196219444,
      "epoch": 0.5451377507892422,
      "grad_norm": 0.1516140252351761,
      "learning_rate": 8.18321219370074e-05,
      "loss": 0.7848,
      "mean_token_accuracy": 0.7664353288710117,
      "num_tokens": 189135240.0,
      "step": 5925
    },
    {
      "entropy": 0.7799247447401285,
      "epoch": 0.5452297571606833,
      "grad_norm": 0.15659765899181366,
      "learning_rate": 8.182905511086577e-05,
      "loss": 0.7844,
      "mean_token_accuracy": 0.7699661515653133,
      "num_tokens": 189167741.0,
      "step": 5926
    },
    {
      "entropy": 0.7206703200936317,
      "epoch": 0.5453217635321246,
      "grad_norm": 0.15322043001651764,
      "learning_rate": 8.182598828472414e-05,
      "loss": 0.7166,
      "mean_token_accuracy": 0.7900163196027279,
      "num_tokens": 189200021.0,
      "step": 5927
    },
    {
      "entropy": 0.7147715678438544,
      "epoch": 0.5454137699035658,
      "grad_norm": 0.1598958820104599,
      "learning_rate": 8.182292145858251e-05,
      "loss": 0.7067,
      "mean_token_accuracy": 0.796906378120184,
      "num_tokens": 189232306.0,
      "step": 5928
    },
    {
      "entropy": 0.8964174315333366,
      "epoch": 0.5455057762750071,
      "grad_norm": 0.16555264592170715,
      "learning_rate": 8.181985463244089e-05,
      "loss": 0.8967,
      "mean_token_accuracy": 0.7421503439545631,
      "num_tokens": 189263565.0,
      "step": 5929
    },
    {
      "entropy": 0.7223835587501526,
      "epoch": 0.5455977826464483,
      "grad_norm": 0.15395823121070862,
      "learning_rate": 8.181678780629927e-05,
      "loss": 0.7089,
      "mean_token_accuracy": 0.7951339334249496,
      "num_tokens": 189295819.0,
      "step": 5930
    },
    {
      "entropy": 0.7396577131003141,
      "epoch": 0.5456897890178894,
      "grad_norm": 0.16273778676986694,
      "learning_rate": 8.181372098015764e-05,
      "loss": 0.7304,
      "mean_token_accuracy": 0.7903311811387539,
      "num_tokens": 189327431.0,
      "step": 5931
    },
    {
      "entropy": 0.7338634124025702,
      "epoch": 0.5457817953893307,
      "grad_norm": 0.17111089825630188,
      "learning_rate": 8.181065415401601e-05,
      "loss": 0.7387,
      "mean_token_accuracy": 0.7881190031766891,
      "num_tokens": 189359900.0,
      "step": 5932
    },
    {
      "entropy": 0.799620721489191,
      "epoch": 0.5458738017607719,
      "grad_norm": 0.16098205745220184,
      "learning_rate": 8.180758732787439e-05,
      "loss": 0.7991,
      "mean_token_accuracy": 0.7695279233157635,
      "num_tokens": 189391651.0,
      "step": 5933
    },
    {
      "entropy": 0.8324064500629902,
      "epoch": 0.5459658081322132,
      "grad_norm": 0.16574826836585999,
      "learning_rate": 8.180452050173276e-05,
      "loss": 0.8118,
      "mean_token_accuracy": 0.7670904248952866,
      "num_tokens": 189423459.0,
      "step": 5934
    },
    {
      "entropy": 0.783847002312541,
      "epoch": 0.5460578145036544,
      "grad_norm": 0.15079179406166077,
      "learning_rate": 8.180145367559114e-05,
      "loss": 0.7648,
      "mean_token_accuracy": 0.7740569412708282,
      "num_tokens": 189455519.0,
      "step": 5935
    },
    {
      "entropy": 0.7552102357149124,
      "epoch": 0.5461498208750956,
      "grad_norm": 0.1573300063610077,
      "learning_rate": 8.17983868494495e-05,
      "loss": 0.7439,
      "mean_token_accuracy": 0.7831345722079277,
      "num_tokens": 189487735.0,
      "step": 5936
    },
    {
      "entropy": 0.6264612637460232,
      "epoch": 0.5462418272465368,
      "grad_norm": 0.1484707146883011,
      "learning_rate": 8.179532002330789e-05,
      "loss": 0.6077,
      "mean_token_accuracy": 0.816211149096489,
      "num_tokens": 189519451.0,
      "step": 5937
    },
    {
      "entropy": 0.8330480381846428,
      "epoch": 0.546333833617978,
      "grad_norm": 0.16059932112693787,
      "learning_rate": 8.179225319716626e-05,
      "loss": 0.8381,
      "mean_token_accuracy": 0.7595575302839279,
      "num_tokens": 189550731.0,
      "step": 5938
    },
    {
      "entropy": 0.7569078728556633,
      "epoch": 0.5464258399894193,
      "grad_norm": 0.1631501317024231,
      "learning_rate": 8.178918637102462e-05,
      "loss": 0.7609,
      "mean_token_accuracy": 0.7765948213636875,
      "num_tokens": 189582616.0,
      "step": 5939
    },
    {
      "entropy": 0.7061306163668633,
      "epoch": 0.5465178463608605,
      "grad_norm": 0.15236112475395203,
      "learning_rate": 8.1786119544883e-05,
      "loss": 0.6904,
      "mean_token_accuracy": 0.7943682558834553,
      "num_tokens": 189615050.0,
      "step": 5940
    },
    {
      "entropy": 0.6648759841918945,
      "epoch": 0.5466098527323017,
      "grad_norm": 0.1481602042913437,
      "learning_rate": 8.178305271874139e-05,
      "loss": 0.6512,
      "mean_token_accuracy": 0.8070805333554745,
      "num_tokens": 189647399.0,
      "step": 5941
    },
    {
      "entropy": 0.7537821866571903,
      "epoch": 0.5467018591037429,
      "grad_norm": 0.15628133714199066,
      "learning_rate": 8.177998589259975e-05,
      "loss": 0.7382,
      "mean_token_accuracy": 0.7850614823400974,
      "num_tokens": 189679122.0,
      "step": 5942
    },
    {
      "entropy": 0.7334588058292866,
      "epoch": 0.5467938654751842,
      "grad_norm": 0.1558654010295868,
      "learning_rate": 8.177691906645812e-05,
      "loss": 0.7087,
      "mean_token_accuracy": 0.7898502685129642,
      "num_tokens": 189710645.0,
      "step": 5943
    },
    {
      "entropy": 0.790848683565855,
      "epoch": 0.5468858718466254,
      "grad_norm": 0.1553385704755783,
      "learning_rate": 8.177385224031649e-05,
      "loss": 0.7719,
      "mean_token_accuracy": 0.7784430049359798,
      "num_tokens": 189742418.0,
      "step": 5944
    },
    {
      "entropy": 0.7136029954999685,
      "epoch": 0.5469778782180666,
      "grad_norm": 0.16431604325771332,
      "learning_rate": 8.177078541417487e-05,
      "loss": 0.6874,
      "mean_token_accuracy": 0.7938651107251644,
      "num_tokens": 189774291.0,
      "step": 5945
    },
    {
      "entropy": 0.736778249964118,
      "epoch": 0.5470698845895078,
      "grad_norm": 0.15942849218845367,
      "learning_rate": 8.176771858803325e-05,
      "loss": 0.7436,
      "mean_token_accuracy": 0.7861796133220196,
      "num_tokens": 189806737.0,
      "step": 5946
    },
    {
      "entropy": 0.6865629218518734,
      "epoch": 0.547161890960949,
      "grad_norm": 0.14884160459041595,
      "learning_rate": 8.176465176189162e-05,
      "loss": 0.6656,
      "mean_token_accuracy": 0.803143922239542,
      "num_tokens": 189838824.0,
      "step": 5947
    },
    {
      "entropy": 0.6863450277596712,
      "epoch": 0.5472538973323903,
      "grad_norm": 0.1511482447385788,
      "learning_rate": 8.176158493574999e-05,
      "loss": 0.675,
      "mean_token_accuracy": 0.7961273416876793,
      "num_tokens": 189871296.0,
      "step": 5948
    },
    {
      "entropy": 0.8125578798353672,
      "epoch": 0.5473459037038315,
      "grad_norm": 0.16279666125774384,
      "learning_rate": 8.175851810960837e-05,
      "loss": 0.8211,
      "mean_token_accuracy": 0.7608473524451256,
      "num_tokens": 189903068.0,
      "step": 5949
    },
    {
      "entropy": 0.6833231169730425,
      "epoch": 0.5474379100752728,
      "grad_norm": 0.163343608379364,
      "learning_rate": 8.175545128346674e-05,
      "loss": 0.6793,
      "mean_token_accuracy": 0.7997404038906097,
      "num_tokens": 189935121.0,
      "step": 5950
    },
    {
      "entropy": 0.7700577862560749,
      "epoch": 0.5475299164467139,
      "grad_norm": 0.16466020047664642,
      "learning_rate": 8.175238445732512e-05,
      "loss": 0.773,
      "mean_token_accuracy": 0.7741696909070015,
      "num_tokens": 189966119.0,
      "step": 5951
    },
    {
      "entropy": 0.8340642116963863,
      "epoch": 0.5476219228181551,
      "grad_norm": 0.15206298232078552,
      "learning_rate": 8.17493176311835e-05,
      "loss": 0.8366,
      "mean_token_accuracy": 0.7600961253046989,
      "num_tokens": 189998377.0,
      "step": 5952
    },
    {
      "entropy": 0.800065852701664,
      "epoch": 0.5477139291895964,
      "grad_norm": 0.17626942694187164,
      "learning_rate": 8.174625080504187e-05,
      "loss": 0.8295,
      "mean_token_accuracy": 0.7611846104264259,
      "num_tokens": 190029841.0,
      "step": 5953
    },
    {
      "entropy": 0.7243795469403267,
      "epoch": 0.5478059355610376,
      "grad_norm": 0.15090879797935486,
      "learning_rate": 8.174318397890024e-05,
      "loss": 0.715,
      "mean_token_accuracy": 0.7906467840075493,
      "num_tokens": 190061514.0,
      "step": 5954
    },
    {
      "entropy": 0.7380406372249126,
      "epoch": 0.5478979419324789,
      "grad_norm": 0.1544695496559143,
      "learning_rate": 8.17401171527586e-05,
      "loss": 0.7261,
      "mean_token_accuracy": 0.7893823496997356,
      "num_tokens": 190093816.0,
      "step": 5955
    },
    {
      "entropy": 0.876972483471036,
      "epoch": 0.54798994830392,
      "grad_norm": 0.16142427921295166,
      "learning_rate": 8.1737050326617e-05,
      "loss": 0.8674,
      "mean_token_accuracy": 0.7467220351099968,
      "num_tokens": 190126329.0,
      "step": 5956
    },
    {
      "entropy": 0.7489898111671209,
      "epoch": 0.5480819546753612,
      "grad_norm": 0.15326781570911407,
      "learning_rate": 8.173398350047537e-05,
      "loss": 0.7285,
      "mean_token_accuracy": 0.7890992201864719,
      "num_tokens": 190159056.0,
      "step": 5957
    },
    {
      "entropy": 0.8465461302548647,
      "epoch": 0.5481739610468025,
      "grad_norm": 0.159610778093338,
      "learning_rate": 8.173091667433373e-05,
      "loss": 0.842,
      "mean_token_accuracy": 0.7538911774754524,
      "num_tokens": 190191399.0,
      "step": 5958
    },
    {
      "entropy": 0.6975799165666103,
      "epoch": 0.5482659674182437,
      "grad_norm": 0.14913909137248993,
      "learning_rate": 8.17278498481921e-05,
      "loss": 0.6831,
      "mean_token_accuracy": 0.7986496426165104,
      "num_tokens": 190224061.0,
      "step": 5959
    },
    {
      "entropy": 0.8778853118419647,
      "epoch": 0.548357973789685,
      "grad_norm": 0.16104376316070557,
      "learning_rate": 8.172478302205048e-05,
      "loss": 0.8452,
      "mean_token_accuracy": 0.7531905919313431,
      "num_tokens": 190255741.0,
      "step": 5960
    },
    {
      "entropy": 0.8323763124644756,
      "epoch": 0.5484499801611261,
      "grad_norm": 0.1552756279706955,
      "learning_rate": 8.172171619590887e-05,
      "loss": 0.809,
      "mean_token_accuracy": 0.7628321759402752,
      "num_tokens": 190287382.0,
      "step": 5961
    },
    {
      "entropy": 0.8368029780685902,
      "epoch": 0.5485419865325674,
      "grad_norm": 0.16225722432136536,
      "learning_rate": 8.171864936976723e-05,
      "loss": 0.8156,
      "mean_token_accuracy": 0.7588170729577541,
      "num_tokens": 190318954.0,
      "step": 5962
    },
    {
      "entropy": 0.8904872201383114,
      "epoch": 0.5486339929040086,
      "grad_norm": 0.16282087564468384,
      "learning_rate": 8.17155825436256e-05,
      "loss": 0.8733,
      "mean_token_accuracy": 0.7466446273028851,
      "num_tokens": 190350197.0,
      "step": 5963
    },
    {
      "entropy": 0.7524302285164595,
      "epoch": 0.5487259992754498,
      "grad_norm": 0.14911939203739166,
      "learning_rate": 8.171251571748398e-05,
      "loss": 0.7379,
      "mean_token_accuracy": 0.7864252515137196,
      "num_tokens": 190382965.0,
      "step": 5964
    },
    {
      "entropy": 0.7523713950067759,
      "epoch": 0.5488180056468911,
      "grad_norm": 0.1702347993850708,
      "learning_rate": 8.170944889134235e-05,
      "loss": 0.7483,
      "mean_token_accuracy": 0.7823081277310848,
      "num_tokens": 190413804.0,
      "step": 5965
    },
    {
      "entropy": 0.6707523446530104,
      "epoch": 0.5489100120183322,
      "grad_norm": 0.14974647760391235,
      "learning_rate": 8.170638206520073e-05,
      "loss": 0.6492,
      "mean_token_accuracy": 0.8047818914055824,
      "num_tokens": 190445983.0,
      "step": 5966
    },
    {
      "entropy": 0.6782028563320637,
      "epoch": 0.5490020183897735,
      "grad_norm": 0.1605050414800644,
      "learning_rate": 8.17033152390591e-05,
      "loss": 0.6587,
      "mean_token_accuracy": 0.8097090609371662,
      "num_tokens": 190477952.0,
      "step": 5967
    },
    {
      "entropy": 0.6639335732907057,
      "epoch": 0.5490940247612147,
      "grad_norm": 0.1554328054189682,
      "learning_rate": 8.170024841291748e-05,
      "loss": 0.6569,
      "mean_token_accuracy": 0.804432787001133,
      "num_tokens": 190510694.0,
      "step": 5968
    },
    {
      "entropy": 0.8310938589274883,
      "epoch": 0.549186031132656,
      "grad_norm": 0.16171905398368835,
      "learning_rate": 8.169718158677585e-05,
      "loss": 0.8193,
      "mean_token_accuracy": 0.7589053362607956,
      "num_tokens": 190542505.0,
      "step": 5969
    },
    {
      "entropy": 0.774412602186203,
      "epoch": 0.5492780375040972,
      "grad_norm": 0.15266726911067963,
      "learning_rate": 8.169411476063422e-05,
      "loss": 0.7795,
      "mean_token_accuracy": 0.7744964696466923,
      "num_tokens": 190575029.0,
      "step": 5970
    },
    {
      "entropy": 0.789188627153635,
      "epoch": 0.5493700438755383,
      "grad_norm": 0.16331847012043,
      "learning_rate": 8.16910479344926e-05,
      "loss": 0.7887,
      "mean_token_accuracy": 0.7721030525863171,
      "num_tokens": 190606843.0,
      "step": 5971
    },
    {
      "entropy": 0.8801970444619656,
      "epoch": 0.5494620502469796,
      "grad_norm": 0.1640438735485077,
      "learning_rate": 8.168798110835098e-05,
      "loss": 0.8863,
      "mean_token_accuracy": 0.7420910485088825,
      "num_tokens": 190638213.0,
      "step": 5972
    },
    {
      "entropy": 0.7599710635840893,
      "epoch": 0.5495540566184208,
      "grad_norm": 0.1692570447921753,
      "learning_rate": 8.168491428220935e-05,
      "loss": 0.7748,
      "mean_token_accuracy": 0.7696486860513687,
      "num_tokens": 190669909.0,
      "step": 5973
    },
    {
      "entropy": 0.8596623726189137,
      "epoch": 0.5496460629898621,
      "grad_norm": 0.15719108283519745,
      "learning_rate": 8.168184745606772e-05,
      "loss": 0.8504,
      "mean_token_accuracy": 0.7539977468550205,
      "num_tokens": 190702179.0,
      "step": 5974
    },
    {
      "entropy": 0.7222054917365313,
      "epoch": 0.5497380693613033,
      "grad_norm": 0.15574540197849274,
      "learning_rate": 8.167878062992608e-05,
      "loss": 0.7329,
      "mean_token_accuracy": 0.7852960750460625,
      "num_tokens": 190734908.0,
      "step": 5975
    },
    {
      "entropy": 0.7740677744150162,
      "epoch": 0.5498300757327444,
      "grad_norm": 0.15925899147987366,
      "learning_rate": 8.167571380378447e-05,
      "loss": 0.7683,
      "mean_token_accuracy": 0.7768105678260326,
      "num_tokens": 190767234.0,
      "step": 5976
    },
    {
      "entropy": 0.7972619011998177,
      "epoch": 0.5499220821041857,
      "grad_norm": 0.155472531914711,
      "learning_rate": 8.167264697764285e-05,
      "loss": 0.7908,
      "mean_token_accuracy": 0.7683476619422436,
      "num_tokens": 190799534.0,
      "step": 5977
    },
    {
      "entropy": 0.7257073437795043,
      "epoch": 0.5500140884756269,
      "grad_norm": 0.14671020209789276,
      "learning_rate": 8.166958015150121e-05,
      "loss": 0.7147,
      "mean_token_accuracy": 0.7921796329319477,
      "num_tokens": 190832106.0,
      "step": 5978
    },
    {
      "entropy": 0.7165238987654448,
      "epoch": 0.5501060948470682,
      "grad_norm": 0.1534561812877655,
      "learning_rate": 8.166651332535958e-05,
      "loss": 0.6983,
      "mean_token_accuracy": 0.7955245338380337,
      "num_tokens": 190864563.0,
      "step": 5979
    },
    {
      "entropy": 0.7445898596197367,
      "epoch": 0.5501981012185094,
      "grad_norm": 0.1474943310022354,
      "learning_rate": 8.166344649921796e-05,
      "loss": 0.7218,
      "mean_token_accuracy": 0.7847502157092094,
      "num_tokens": 190896433.0,
      "step": 5980
    },
    {
      "entropy": 0.7592648211866617,
      "epoch": 0.5502901075899506,
      "grad_norm": 0.15504078567028046,
      "learning_rate": 8.166037967307633e-05,
      "loss": 0.7499,
      "mean_token_accuracy": 0.7832855395972729,
      "num_tokens": 190927530.0,
      "step": 5981
    },
    {
      "entropy": 0.7179149799048901,
      "epoch": 0.5503821139613918,
      "grad_norm": 0.1504850536584854,
      "learning_rate": 8.165731284693471e-05,
      "loss": 0.7054,
      "mean_token_accuracy": 0.7960471622645855,
      "num_tokens": 190959152.0,
      "step": 5982
    },
    {
      "entropy": 0.7716169897466898,
      "epoch": 0.550474120332833,
      "grad_norm": 0.15910422801971436,
      "learning_rate": 8.16542460207931e-05,
      "loss": 0.7656,
      "mean_token_accuracy": 0.777254942804575,
      "num_tokens": 190990506.0,
      "step": 5983
    },
    {
      "entropy": 0.9168150946497917,
      "epoch": 0.5505661267042743,
      "grad_norm": 0.16871169209480286,
      "learning_rate": 8.165117919465146e-05,
      "loss": 0.932,
      "mean_token_accuracy": 0.7349510826170444,
      "num_tokens": 191022851.0,
      "step": 5984
    },
    {
      "entropy": 0.8260986674576998,
      "epoch": 0.5506581330757155,
      "grad_norm": 0.16138248145580292,
      "learning_rate": 8.164811236850983e-05,
      "loss": 0.8238,
      "mean_token_accuracy": 0.7588727883994579,
      "num_tokens": 191054441.0,
      "step": 5985
    },
    {
      "entropy": 0.7338607683777809,
      "epoch": 0.5507501394471567,
      "grad_norm": 0.15588732063770294,
      "learning_rate": 8.16450455423682e-05,
      "loss": 0.7437,
      "mean_token_accuracy": 0.7809581123292446,
      "num_tokens": 191086446.0,
      "step": 5986
    },
    {
      "entropy": 0.7654548157006502,
      "epoch": 0.5508421458185979,
      "grad_norm": 0.1687469184398651,
      "learning_rate": 8.164197871622658e-05,
      "loss": 0.7577,
      "mean_token_accuracy": 0.7777836956083775,
      "num_tokens": 191118697.0,
      "step": 5987
    },
    {
      "entropy": 0.6191585035994649,
      "epoch": 0.5509341521900392,
      "grad_norm": 0.1552172303199768,
      "learning_rate": 8.163891189008496e-05,
      "loss": 0.613,
      "mean_token_accuracy": 0.8167121484875679,
      "num_tokens": 191151070.0,
      "step": 5988
    },
    {
      "entropy": 0.7283826656639576,
      "epoch": 0.5510261585614804,
      "grad_norm": 0.1582939773797989,
      "learning_rate": 8.163584506394333e-05,
      "loss": 0.6893,
      "mean_token_accuracy": 0.7968426309525967,
      "num_tokens": 191183556.0,
      "step": 5989
    },
    {
      "entropy": 0.784462746232748,
      "epoch": 0.5511181649329217,
      "grad_norm": 0.15464681386947632,
      "learning_rate": 8.16327782378017e-05,
      "loss": 0.751,
      "mean_token_accuracy": 0.7803607545793056,
      "num_tokens": 191215334.0,
      "step": 5990
    },
    {
      "entropy": 0.7864212393760681,
      "epoch": 0.5512101713043628,
      "grad_norm": 0.16656818985939026,
      "learning_rate": 8.162971141166008e-05,
      "loss": 0.7705,
      "mean_token_accuracy": 0.7753107137978077,
      "num_tokens": 191246880.0,
      "step": 5991
    },
    {
      "entropy": 0.8305101580917835,
      "epoch": 0.551302177675804,
      "grad_norm": 0.15914209187030792,
      "learning_rate": 8.162664458551845e-05,
      "loss": 0.816,
      "mean_token_accuracy": 0.7657095193862915,
      "num_tokens": 191278493.0,
      "step": 5992
    },
    {
      "entropy": 0.7413014397025108,
      "epoch": 0.5513941840472453,
      "grad_norm": 0.1564912647008896,
      "learning_rate": 8.162357775937683e-05,
      "loss": 0.7121,
      "mean_token_accuracy": 0.787505242973566,
      "num_tokens": 191310217.0,
      "step": 5993
    },
    {
      "entropy": 0.8313757330179214,
      "epoch": 0.5514861904186865,
      "grad_norm": 0.16416552662849426,
      "learning_rate": 8.16205109332352e-05,
      "loss": 0.8336,
      "mean_token_accuracy": 0.7576211579144001,
      "num_tokens": 191342175.0,
      "step": 5994
    },
    {
      "entropy": 0.8442597221583128,
      "epoch": 0.5515781967901278,
      "grad_norm": 0.17158855497837067,
      "learning_rate": 8.161744410709358e-05,
      "loss": 0.8485,
      "mean_token_accuracy": 0.755003783851862,
      "num_tokens": 191374368.0,
      "step": 5995
    },
    {
      "entropy": 0.8091115206480026,
      "epoch": 0.5516702031615689,
      "grad_norm": 0.16732440888881683,
      "learning_rate": 8.161437728095194e-05,
      "loss": 0.8167,
      "mean_token_accuracy": 0.7653408236801624,
      "num_tokens": 191407002.0,
      "step": 5996
    },
    {
      "entropy": 0.7015152052044868,
      "epoch": 0.5517622095330101,
      "grad_norm": 0.15845352411270142,
      "learning_rate": 8.161131045481033e-05,
      "loss": 0.6734,
      "mean_token_accuracy": 0.7995651811361313,
      "num_tokens": 191439037.0,
      "step": 5997
    },
    {
      "entropy": 0.8498093299567699,
      "epoch": 0.5518542159044514,
      "grad_norm": 0.1552589237689972,
      "learning_rate": 8.16082436286687e-05,
      "loss": 0.8352,
      "mean_token_accuracy": 0.7569010630249977,
      "num_tokens": 191471182.0,
      "step": 5998
    },
    {
      "entropy": 0.7970709148794413,
      "epoch": 0.5519462222758926,
      "grad_norm": 0.1659885197877884,
      "learning_rate": 8.160517680252708e-05,
      "loss": 0.8099,
      "mean_token_accuracy": 0.7657053582370281,
      "num_tokens": 191503555.0,
      "step": 5999
    },
    {
      "entropy": 0.7965427450835705,
      "epoch": 0.5520382286473339,
      "grad_norm": 0.16310161352157593,
      "learning_rate": 8.160210997638544e-05,
      "loss": 0.7784,
      "mean_token_accuracy": 0.7684974633157253,
      "num_tokens": 191534763.0,
      "step": 6000
    },
    {
      "entropy": 0.7706787437200546,
      "epoch": 0.552130235018775,
      "grad_norm": 0.15562576055526733,
      "learning_rate": 8.159904315024381e-05,
      "loss": 0.7385,
      "mean_token_accuracy": 0.7833890169858932,
      "num_tokens": 191567259.0,
      "step": 6001
    },
    {
      "entropy": 0.8564785588532686,
      "epoch": 0.5522222413902163,
      "grad_norm": 0.1650467813014984,
      "learning_rate": 8.159597632410219e-05,
      "loss": 0.8404,
      "mean_token_accuracy": 0.7567550130188465,
      "num_tokens": 191597976.0,
      "step": 6002
    },
    {
      "entropy": 0.8336704932153225,
      "epoch": 0.5523142477616575,
      "grad_norm": 0.16959384083747864,
      "learning_rate": 8.159290949796057e-05,
      "loss": 0.8366,
      "mean_token_accuracy": 0.7610478885471821,
      "num_tokens": 191629494.0,
      "step": 6003
    },
    {
      "entropy": 0.8957505021244287,
      "epoch": 0.5524062541330987,
      "grad_norm": 0.1650269329547882,
      "learning_rate": 8.158984267181894e-05,
      "loss": 0.8832,
      "mean_token_accuracy": 0.7473922073841095,
      "num_tokens": 191660388.0,
      "step": 6004
    },
    {
      "entropy": 0.7259356360882521,
      "epoch": 0.55249826050454,
      "grad_norm": 0.1571079045534134,
      "learning_rate": 8.158677584567731e-05,
      "loss": 0.7119,
      "mean_token_accuracy": 0.7895214892923832,
      "num_tokens": 191692736.0,
      "step": 6005
    },
    {
      "entropy": 0.739856731146574,
      "epoch": 0.5525902668759811,
      "grad_norm": 0.15249207615852356,
      "learning_rate": 8.158370901953568e-05,
      "loss": 0.7145,
      "mean_token_accuracy": 0.7917918004095554,
      "num_tokens": 191724465.0,
      "step": 6006
    },
    {
      "entropy": 0.8330467566847801,
      "epoch": 0.5526822732474224,
      "grad_norm": 0.15783803164958954,
      "learning_rate": 8.158064219339406e-05,
      "loss": 0.8102,
      "mean_token_accuracy": 0.7623555846512318,
      "num_tokens": 191756663.0,
      "step": 6007
    },
    {
      "entropy": 0.7143422681838274,
      "epoch": 0.5527742796188636,
      "grad_norm": 0.15776567161083221,
      "learning_rate": 8.157757536725244e-05,
      "loss": 0.6895,
      "mean_token_accuracy": 0.7990431413054466,
      "num_tokens": 191788769.0,
      "step": 6008
    },
    {
      "entropy": 0.7926603611558676,
      "epoch": 0.5528662859903049,
      "grad_norm": 0.15281923115253448,
      "learning_rate": 8.157450854111081e-05,
      "loss": 0.7906,
      "mean_token_accuracy": 0.7720540463924408,
      "num_tokens": 191820731.0,
      "step": 6009
    },
    {
      "entropy": 0.8669855333864689,
      "epoch": 0.5529582923617461,
      "grad_norm": 0.1653076559305191,
      "learning_rate": 8.157144171496919e-05,
      "loss": 0.8804,
      "mean_token_accuracy": 0.7433367408812046,
      "num_tokens": 191851947.0,
      "step": 6010
    },
    {
      "entropy": 0.8002363406121731,
      "epoch": 0.5530502987331872,
      "grad_norm": 0.17021411657333374,
      "learning_rate": 8.156837488882756e-05,
      "loss": 0.8085,
      "mean_token_accuracy": 0.7677657529711723,
      "num_tokens": 191883996.0,
      "step": 6011
    },
    {
      "entropy": 0.8670820891857147,
      "epoch": 0.5531423051046285,
      "grad_norm": 0.15995609760284424,
      "learning_rate": 8.156530806268593e-05,
      "loss": 0.8585,
      "mean_token_accuracy": 0.7551304623484612,
      "num_tokens": 191916251.0,
      "step": 6012
    },
    {
      "entropy": 0.7834235690534115,
      "epoch": 0.5532343114760697,
      "grad_norm": 0.16000738739967346,
      "learning_rate": 8.15622412365443e-05,
      "loss": 0.7831,
      "mean_token_accuracy": 0.76929085329175,
      "num_tokens": 191948887.0,
      "step": 6013
    },
    {
      "entropy": 0.7350541241466999,
      "epoch": 0.553326317847511,
      "grad_norm": 0.15425541996955872,
      "learning_rate": 8.155917441040269e-05,
      "loss": 0.7316,
      "mean_token_accuracy": 0.7877219878137112,
      "num_tokens": 191981263.0,
      "step": 6014
    },
    {
      "entropy": 0.860594816505909,
      "epoch": 0.5534183242189522,
      "grad_norm": 0.15892741084098816,
      "learning_rate": 8.155610758426106e-05,
      "loss": 0.8476,
      "mean_token_accuracy": 0.7574911825358868,
      "num_tokens": 192012484.0,
      "step": 6015
    },
    {
      "entropy": 0.7508128182962537,
      "epoch": 0.5535103305903933,
      "grad_norm": 0.1581711769104004,
      "learning_rate": 8.155304075811942e-05,
      "loss": 0.7215,
      "mean_token_accuracy": 0.7928740195930004,
      "num_tokens": 192044297.0,
      "step": 6016
    },
    {
      "entropy": 0.7154111918061972,
      "epoch": 0.5536023369618346,
      "grad_norm": 0.1592387855052948,
      "learning_rate": 8.154997393197779e-05,
      "loss": 0.7082,
      "mean_token_accuracy": 0.7955047711730003,
      "num_tokens": 192077010.0,
      "step": 6017
    },
    {
      "entropy": 0.7269014222547412,
      "epoch": 0.5536943433332758,
      "grad_norm": 0.15175572037696838,
      "learning_rate": 8.154690710583617e-05,
      "loss": 0.6987,
      "mean_token_accuracy": 0.7933903373777866,
      "num_tokens": 192109024.0,
      "step": 6018
    },
    {
      "entropy": 0.746125690639019,
      "epoch": 0.5537863497047171,
      "grad_norm": 0.16021691262722015,
      "learning_rate": 8.154384027969455e-05,
      "loss": 0.7388,
      "mean_token_accuracy": 0.7843107059597969,
      "num_tokens": 192139227.0,
      "step": 6019
    },
    {
      "entropy": 0.8648782018572092,
      "epoch": 0.5538783560761583,
      "grad_norm": 0.16696161031723022,
      "learning_rate": 8.154077345355292e-05,
      "loss": 0.8576,
      "mean_token_accuracy": 0.7544964924454689,
      "num_tokens": 192170868.0,
      "step": 6020
    },
    {
      "entropy": 0.6894812770187855,
      "epoch": 0.5539703624475995,
      "grad_norm": 0.15177986025810242,
      "learning_rate": 8.153770662741129e-05,
      "loss": 0.6809,
      "mean_token_accuracy": 0.7959825806319714,
      "num_tokens": 192203417.0,
      "step": 6021
    },
    {
      "entropy": 0.7893353421241045,
      "epoch": 0.5540623688190407,
      "grad_norm": 0.16263845562934875,
      "learning_rate": 8.153463980126967e-05,
      "loss": 0.7867,
      "mean_token_accuracy": 0.7726844996213913,
      "num_tokens": 192235967.0,
      "step": 6022
    },
    {
      "entropy": 0.810412060469389,
      "epoch": 0.5541543751904819,
      "grad_norm": 0.16554486751556396,
      "learning_rate": 8.153157297512804e-05,
      "loss": 0.8301,
      "mean_token_accuracy": 0.7590293139219284,
      "num_tokens": 192267126.0,
      "step": 6023
    },
    {
      "entropy": 0.7957448363304138,
      "epoch": 0.5542463815619232,
      "grad_norm": 0.18386197090148926,
      "learning_rate": 8.152850614898642e-05,
      "loss": 0.8177,
      "mean_token_accuracy": 0.7706861458718777,
      "num_tokens": 192299111.0,
      "step": 6024
    },
    {
      "entropy": 0.7195996344089508,
      "epoch": 0.5543383879333644,
      "grad_norm": 0.15169855952262878,
      "learning_rate": 8.152543932284479e-05,
      "loss": 0.6937,
      "mean_token_accuracy": 0.7911792770028114,
      "num_tokens": 192330966.0,
      "step": 6025
    },
    {
      "entropy": 0.7095565646886826,
      "epoch": 0.5544303943048056,
      "grad_norm": 0.14465606212615967,
      "learning_rate": 8.152237249670317e-05,
      "loss": 0.7046,
      "mean_token_accuracy": 0.7882008031010628,
      "num_tokens": 192363640.0,
      "step": 6026
    },
    {
      "entropy": 0.7250357652083039,
      "epoch": 0.5545224006762468,
      "grad_norm": 0.15593747794628143,
      "learning_rate": 8.151930567056154e-05,
      "loss": 0.7082,
      "mean_token_accuracy": 0.7878669761121273,
      "num_tokens": 192395817.0,
      "step": 6027
    },
    {
      "entropy": 0.7934261150658131,
      "epoch": 0.554614407047688,
      "grad_norm": 0.16062304377555847,
      "learning_rate": 8.15162388444199e-05,
      "loss": 0.7839,
      "mean_token_accuracy": 0.7674033045768738,
      "num_tokens": 192428236.0,
      "step": 6028
    },
    {
      "entropy": 0.8054333012551069,
      "epoch": 0.5547064134191293,
      "grad_norm": 0.15277118980884552,
      "learning_rate": 8.151317201827829e-05,
      "loss": 0.775,
      "mean_token_accuracy": 0.7710592932999134,
      "num_tokens": 192460231.0,
      "step": 6029
    },
    {
      "entropy": 0.9132533892989159,
      "epoch": 0.5547984197905705,
      "grad_norm": 0.1635074019432068,
      "learning_rate": 8.151010519213667e-05,
      "loss": 0.9137,
      "mean_token_accuracy": 0.7374852299690247,
      "num_tokens": 192491595.0,
      "step": 6030
    },
    {
      "entropy": 0.7676389440894127,
      "epoch": 0.5548904261620117,
      "grad_norm": 0.1618419885635376,
      "learning_rate": 8.150703836599504e-05,
      "loss": 0.7457,
      "mean_token_accuracy": 0.7828066162765026,
      "num_tokens": 192522794.0,
      "step": 6031
    },
    {
      "entropy": 0.6918323133140802,
      "epoch": 0.5549824325334529,
      "grad_norm": 0.16495268046855927,
      "learning_rate": 8.15039715398534e-05,
      "loss": 0.682,
      "mean_token_accuracy": 0.8011768721044064,
      "num_tokens": 192554600.0,
      "step": 6032
    },
    {
      "entropy": 0.6737655261531472,
      "epoch": 0.5550744389048942,
      "grad_norm": 0.15140675008296967,
      "learning_rate": 8.150090471371177e-05,
      "loss": 0.6476,
      "mean_token_accuracy": 0.8064912632107735,
      "num_tokens": 192586591.0,
      "step": 6033
    },
    {
      "entropy": 0.7868568822741508,
      "epoch": 0.5551664452763354,
      "grad_norm": 0.1545213907957077,
      "learning_rate": 8.149783788757017e-05,
      "loss": 0.7785,
      "mean_token_accuracy": 0.7697130851447582,
      "num_tokens": 192619161.0,
      "step": 6034
    },
    {
      "entropy": 0.7533650267869234,
      "epoch": 0.5552584516477767,
      "grad_norm": 0.15891723334789276,
      "learning_rate": 8.149477106142854e-05,
      "loss": 0.7392,
      "mean_token_accuracy": 0.782652672380209,
      "num_tokens": 192651029.0,
      "step": 6035
    },
    {
      "entropy": 0.7167189456522465,
      "epoch": 0.5553504580192178,
      "grad_norm": 0.14931468665599823,
      "learning_rate": 8.14917042352869e-05,
      "loss": 0.6929,
      "mean_token_accuracy": 0.7925790287554264,
      "num_tokens": 192683254.0,
      "step": 6036
    },
    {
      "entropy": 0.7868742290884256,
      "epoch": 0.555442464390659,
      "grad_norm": 0.15820583701133728,
      "learning_rate": 8.148863740914527e-05,
      "loss": 0.7838,
      "mean_token_accuracy": 0.7751549780368805,
      "num_tokens": 192715097.0,
      "step": 6037
    },
    {
      "entropy": 0.8288993276655674,
      "epoch": 0.5555344707621003,
      "grad_norm": 0.1632251739501953,
      "learning_rate": 8.148557058300365e-05,
      "loss": 0.8245,
      "mean_token_accuracy": 0.7565753273665905,
      "num_tokens": 192746453.0,
      "step": 6038
    },
    {
      "entropy": 0.8013939969241619,
      "epoch": 0.5556264771335415,
      "grad_norm": 0.16061708331108093,
      "learning_rate": 8.148250375686203e-05,
      "loss": 0.7873,
      "mean_token_accuracy": 0.7647797577083111,
      "num_tokens": 192777759.0,
      "step": 6039
    },
    {
      "entropy": 0.7201927118003368,
      "epoch": 0.5557184835049828,
      "grad_norm": 0.1563645303249359,
      "learning_rate": 8.14794369307204e-05,
      "loss": 0.7091,
      "mean_token_accuracy": 0.7943665906786919,
      "num_tokens": 192808679.0,
      "step": 6040
    },
    {
      "entropy": 0.7707992177456617,
      "epoch": 0.5558104898764239,
      "grad_norm": 0.16381540894508362,
      "learning_rate": 8.147637010457878e-05,
      "loss": 0.7773,
      "mean_token_accuracy": 0.7706890068948269,
      "num_tokens": 192840593.0,
      "step": 6041
    },
    {
      "entropy": 0.7585370019078255,
      "epoch": 0.5559024962478651,
      "grad_norm": 0.1659216582775116,
      "learning_rate": 8.147330327843715e-05,
      "loss": 0.7464,
      "mean_token_accuracy": 0.7788841165602207,
      "num_tokens": 192872280.0,
      "step": 6042
    },
    {
      "entropy": 0.6856523975729942,
      "epoch": 0.5559945026193064,
      "grad_norm": 0.1508409082889557,
      "learning_rate": 8.147023645229552e-05,
      "loss": 0.6683,
      "mean_token_accuracy": 0.8001473769545555,
      "num_tokens": 192904544.0,
      "step": 6043
    },
    {
      "entropy": 0.7405461082234979,
      "epoch": 0.5560865089907476,
      "grad_norm": 0.16830721497535706,
      "learning_rate": 8.14671696261539e-05,
      "loss": 0.7414,
      "mean_token_accuracy": 0.783707819879055,
      "num_tokens": 192936612.0,
      "step": 6044
    },
    {
      "entropy": 0.7261424474418163,
      "epoch": 0.5561785153621889,
      "grad_norm": 0.15182429552078247,
      "learning_rate": 8.146410280001228e-05,
      "loss": 0.7205,
      "mean_token_accuracy": 0.7876883894205093,
      "num_tokens": 192968501.0,
      "step": 6045
    },
    {
      "entropy": 0.7994981249794364,
      "epoch": 0.55627052173363,
      "grad_norm": 0.1578603833913803,
      "learning_rate": 8.146103597387065e-05,
      "loss": 0.7845,
      "mean_token_accuracy": 0.7720818258821964,
      "num_tokens": 193000645.0,
      "step": 6046
    },
    {
      "entropy": 0.8054140135645866,
      "epoch": 0.5563625281050713,
      "grad_norm": 0.15709611773490906,
      "learning_rate": 8.145796914772902e-05,
      "loss": 0.7849,
      "mean_token_accuracy": 0.7700771577656269,
      "num_tokens": 193032052.0,
      "step": 6047
    },
    {
      "entropy": 0.7401599790900946,
      "epoch": 0.5564545344765125,
      "grad_norm": 0.16071166098117828,
      "learning_rate": 8.145490232158739e-05,
      "loss": 0.7311,
      "mean_token_accuracy": 0.7796068713068962,
      "num_tokens": 193063875.0,
      "step": 6048
    },
    {
      "entropy": 0.761769313365221,
      "epoch": 0.5565465408479537,
      "grad_norm": 0.16225092113018036,
      "learning_rate": 8.145183549544577e-05,
      "loss": 0.7348,
      "mean_token_accuracy": 0.7857695370912552,
      "num_tokens": 193096376.0,
      "step": 6049
    },
    {
      "entropy": 0.820658378303051,
      "epoch": 0.556638547219395,
      "grad_norm": 0.15760204195976257,
      "learning_rate": 8.144876866930415e-05,
      "loss": 0.8012,
      "mean_token_accuracy": 0.764300636947155,
      "num_tokens": 193127258.0,
      "step": 6050
    },
    {
      "entropy": 0.7687272727489471,
      "epoch": 0.5567305535908361,
      "grad_norm": 0.16069215536117554,
      "learning_rate": 8.144570184316252e-05,
      "loss": 0.7464,
      "mean_token_accuracy": 0.7790776714682579,
      "num_tokens": 193159067.0,
      "step": 6051
    },
    {
      "entropy": 0.6728891860693693,
      "epoch": 0.5568225599622774,
      "grad_norm": 0.1568991094827652,
      "learning_rate": 8.144263501702088e-05,
      "loss": 0.6678,
      "mean_token_accuracy": 0.8043879047036171,
      "num_tokens": 193190497.0,
      "step": 6052
    },
    {
      "entropy": 0.8590189758688211,
      "epoch": 0.5569145663337186,
      "grad_norm": 0.15771913528442383,
      "learning_rate": 8.143956819087927e-05,
      "loss": 0.837,
      "mean_token_accuracy": 0.7562801353633404,
      "num_tokens": 193221423.0,
      "step": 6053
    },
    {
      "entropy": 0.7716104164719582,
      "epoch": 0.5570065727051599,
      "grad_norm": 0.15243470668792725,
      "learning_rate": 8.143650136473763e-05,
      "loss": 0.7565,
      "mean_token_accuracy": 0.7778572328388691,
      "num_tokens": 193253067.0,
      "step": 6054
    },
    {
      "entropy": 0.770961906760931,
      "epoch": 0.5570985790766011,
      "grad_norm": 0.15163572132587433,
      "learning_rate": 8.143343453859601e-05,
      "loss": 0.7448,
      "mean_token_accuracy": 0.7815127968788147,
      "num_tokens": 193285185.0,
      "step": 6055
    },
    {
      "entropy": 0.6892071831971407,
      "epoch": 0.5571905854480422,
      "grad_norm": 0.15875332057476044,
      "learning_rate": 8.143036771245438e-05,
      "loss": 0.6867,
      "mean_token_accuracy": 0.8008329421281815,
      "num_tokens": 193317497.0,
      "step": 6056
    },
    {
      "entropy": 0.7086705416440964,
      "epoch": 0.5572825918194835,
      "grad_norm": 0.15463882684707642,
      "learning_rate": 8.142730088631276e-05,
      "loss": 0.6941,
      "mean_token_accuracy": 0.7926849536597729,
      "num_tokens": 193349377.0,
      "step": 6057
    },
    {
      "entropy": 0.7738120164722204,
      "epoch": 0.5573745981909247,
      "grad_norm": 0.15718628466129303,
      "learning_rate": 8.142423406017113e-05,
      "loss": 0.7829,
      "mean_token_accuracy": 0.7677246741950512,
      "num_tokens": 193381964.0,
      "step": 6058
    },
    {
      "entropy": 0.8273743204772472,
      "epoch": 0.557466604562366,
      "grad_norm": 0.1571321189403534,
      "learning_rate": 8.14211672340295e-05,
      "loss": 0.8246,
      "mean_token_accuracy": 0.7584010735154152,
      "num_tokens": 193413839.0,
      "step": 6059
    },
    {
      "entropy": 0.8418916780501604,
      "epoch": 0.5575586109338072,
      "grad_norm": 0.16087044775485992,
      "learning_rate": 8.141810040788788e-05,
      "loss": 0.8389,
      "mean_token_accuracy": 0.7545315697789192,
      "num_tokens": 193446285.0,
      "step": 6060
    },
    {
      "entropy": 0.7391447946429253,
      "epoch": 0.5576506173052483,
      "grad_norm": 0.1579708456993103,
      "learning_rate": 8.141503358174626e-05,
      "loss": 0.7387,
      "mean_token_accuracy": 0.7812765203416348,
      "num_tokens": 193478189.0,
      "step": 6061
    },
    {
      "entropy": 0.7614217661321163,
      "epoch": 0.5577426236766896,
      "grad_norm": 0.15669314563274384,
      "learning_rate": 8.141196675560463e-05,
      "loss": 0.7559,
      "mean_token_accuracy": 0.7809196747839451,
      "num_tokens": 193510211.0,
      "step": 6062
    },
    {
      "entropy": 0.7186033148318529,
      "epoch": 0.5578346300481308,
      "grad_norm": 0.1554379165172577,
      "learning_rate": 8.1408899929463e-05,
      "loss": 0.7098,
      "mean_token_accuracy": 0.7925673387944698,
      "num_tokens": 193542515.0,
      "step": 6063
    },
    {
      "entropy": 0.7733528390526772,
      "epoch": 0.5579266364195721,
      "grad_norm": 0.1541675180196762,
      "learning_rate": 8.140583310332137e-05,
      "loss": 0.7549,
      "mean_token_accuracy": 0.7775937207043171,
      "num_tokens": 193574095.0,
      "step": 6064
    },
    {
      "entropy": 0.7765134368091822,
      "epoch": 0.5580186427910133,
      "grad_norm": 0.16174091398715973,
      "learning_rate": 8.140276627717975e-05,
      "loss": 0.7545,
      "mean_token_accuracy": 0.7748664170503616,
      "num_tokens": 193606067.0,
      "step": 6065
    },
    {
      "entropy": 0.717482591047883,
      "epoch": 0.5581106491624545,
      "grad_norm": 0.15444409847259521,
      "learning_rate": 8.139969945103813e-05,
      "loss": 0.7068,
      "mean_token_accuracy": 0.794789832085371,
      "num_tokens": 193638386.0,
      "step": 6066
    },
    {
      "entropy": 0.7763950023800135,
      "epoch": 0.5582026555338957,
      "grad_norm": 0.1629830151796341,
      "learning_rate": 8.13966326248965e-05,
      "loss": 0.7727,
      "mean_token_accuracy": 0.7730504386126995,
      "num_tokens": 193669385.0,
      "step": 6067
    },
    {
      "entropy": 0.7793804947286844,
      "epoch": 0.5582946619053369,
      "grad_norm": 0.15742641687393188,
      "learning_rate": 8.139356579875486e-05,
      "loss": 0.7701,
      "mean_token_accuracy": 0.7737812921404839,
      "num_tokens": 193701671.0,
      "step": 6068
    },
    {
      "entropy": 0.8005569130182266,
      "epoch": 0.5583866682767782,
      "grad_norm": 0.16777004301548004,
      "learning_rate": 8.139049897261325e-05,
      "loss": 0.7851,
      "mean_token_accuracy": 0.77051692456007,
      "num_tokens": 193733228.0,
      "step": 6069
    },
    {
      "entropy": 0.7552869198843837,
      "epoch": 0.5584786746482194,
      "grad_norm": 0.1584707796573639,
      "learning_rate": 8.138743214647161e-05,
      "loss": 0.7611,
      "mean_token_accuracy": 0.7776471599936485,
      "num_tokens": 193765734.0,
      "step": 6070
    },
    {
      "entropy": 0.760448906570673,
      "epoch": 0.5585706810196606,
      "grad_norm": 0.15794475376605988,
      "learning_rate": 8.138436532033e-05,
      "loss": 0.7619,
      "mean_token_accuracy": 0.7775208540260792,
      "num_tokens": 193797750.0,
      "step": 6071
    },
    {
      "entropy": 0.8069232106208801,
      "epoch": 0.5586626873911018,
      "grad_norm": 0.1689303070306778,
      "learning_rate": 8.138129849418838e-05,
      "loss": 0.8358,
      "mean_token_accuracy": 0.7626712322235107,
      "num_tokens": 193830125.0,
      "step": 6072
    },
    {
      "entropy": 0.7783085573464632,
      "epoch": 0.558754693762543,
      "grad_norm": 0.15558648109436035,
      "learning_rate": 8.137823166804674e-05,
      "loss": 0.7795,
      "mean_token_accuracy": 0.77311896905303,
      "num_tokens": 193861867.0,
      "step": 6073
    },
    {
      "entropy": 0.8830032125115395,
      "epoch": 0.5588467001339843,
      "grad_norm": 0.16073694825172424,
      "learning_rate": 8.137516484190511e-05,
      "loss": 0.8703,
      "mean_token_accuracy": 0.7507479079067707,
      "num_tokens": 193893624.0,
      "step": 6074
    },
    {
      "entropy": 0.7273001223802567,
      "epoch": 0.5589387065054255,
      "grad_norm": 0.1466771513223648,
      "learning_rate": 8.137209801576348e-05,
      "loss": 0.7075,
      "mean_token_accuracy": 0.7915177829563618,
      "num_tokens": 193926108.0,
      "step": 6075
    },
    {
      "entropy": 0.7955648992210627,
      "epoch": 0.5590307128768667,
      "grad_norm": 0.15161876380443573,
      "learning_rate": 8.136903118962188e-05,
      "loss": 0.7725,
      "mean_token_accuracy": 0.773093681782484,
      "num_tokens": 193957795.0,
      "step": 6076
    },
    {
      "entropy": 0.7958554960787296,
      "epoch": 0.5591227192483079,
      "grad_norm": 0.15988466143608093,
      "learning_rate": 8.136596436348024e-05,
      "loss": 0.7726,
      "mean_token_accuracy": 0.7739408612251282,
      "num_tokens": 193990532.0,
      "step": 6077
    },
    {
      "entropy": 0.8251246400177479,
      "epoch": 0.5592147256197492,
      "grad_norm": 0.16133373975753784,
      "learning_rate": 8.136289753733861e-05,
      "loss": 0.8067,
      "mean_token_accuracy": 0.7661683186888695,
      "num_tokens": 194023152.0,
      "step": 6078
    },
    {
      "entropy": 0.8004872854799032,
      "epoch": 0.5593067319911904,
      "grad_norm": 0.15964026749134064,
      "learning_rate": 8.135983071119698e-05,
      "loss": 0.7582,
      "mean_token_accuracy": 0.7764024250209332,
      "num_tokens": 194054468.0,
      "step": 6079
    },
    {
      "entropy": 0.8390591889619827,
      "epoch": 0.5593987383626317,
      "grad_norm": 0.16184179484844208,
      "learning_rate": 8.135676388505536e-05,
      "loss": 0.8428,
      "mean_token_accuracy": 0.756284773349762,
      "num_tokens": 194086108.0,
      "step": 6080
    },
    {
      "entropy": 0.7504545785486698,
      "epoch": 0.5594907447340728,
      "grad_norm": 0.17079655826091766,
      "learning_rate": 8.135369705891374e-05,
      "loss": 0.7518,
      "mean_token_accuracy": 0.7847156934440136,
      "num_tokens": 194118503.0,
      "step": 6081
    },
    {
      "entropy": 0.7952451165765524,
      "epoch": 0.559582751105514,
      "grad_norm": 0.1585853099822998,
      "learning_rate": 8.135063023277211e-05,
      "loss": 0.7668,
      "mean_token_accuracy": 0.776585403829813,
      "num_tokens": 194149894.0,
      "step": 6082
    },
    {
      "entropy": 0.7174165323376656,
      "epoch": 0.5596747574769553,
      "grad_norm": 0.16724087297916412,
      "learning_rate": 8.134756340663048e-05,
      "loss": 0.7061,
      "mean_token_accuracy": 0.7918421775102615,
      "num_tokens": 194180872.0,
      "step": 6083
    },
    {
      "entropy": 0.7597122676670551,
      "epoch": 0.5597667638483965,
      "grad_norm": 0.15820325911045074,
      "learning_rate": 8.134449658048886e-05,
      "loss": 0.7512,
      "mean_token_accuracy": 0.775394145399332,
      "num_tokens": 194212421.0,
      "step": 6084
    },
    {
      "entropy": 0.7825785800814629,
      "epoch": 0.5598587702198378,
      "grad_norm": 0.15800397098064423,
      "learning_rate": 8.134142975434723e-05,
      "loss": 0.791,
      "mean_token_accuracy": 0.7705358862876892,
      "num_tokens": 194244300.0,
      "step": 6085
    },
    {
      "entropy": 0.7351433448493481,
      "epoch": 0.5599507765912789,
      "grad_norm": 0.1540946513414383,
      "learning_rate": 8.133836292820561e-05,
      "loss": 0.7357,
      "mean_token_accuracy": 0.7847430668771267,
      "num_tokens": 194276684.0,
      "step": 6086
    },
    {
      "entropy": 0.725321339443326,
      "epoch": 0.5600427829627201,
      "grad_norm": 0.16692623496055603,
      "learning_rate": 8.133529610206398e-05,
      "loss": 0.733,
      "mean_token_accuracy": 0.7867682576179504,
      "num_tokens": 194308266.0,
      "step": 6087
    },
    {
      "entropy": 0.8335498906672001,
      "epoch": 0.5601347893341614,
      "grad_norm": 0.1632673591375351,
      "learning_rate": 8.133222927592236e-05,
      "loss": 0.8138,
      "mean_token_accuracy": 0.7629985176026821,
      "num_tokens": 194339649.0,
      "step": 6088
    },
    {
      "entropy": 0.6618251884356141,
      "epoch": 0.5602267957056026,
      "grad_norm": 0.15041187405586243,
      "learning_rate": 8.132916244978073e-05,
      "loss": 0.6555,
      "mean_token_accuracy": 0.8010103367269039,
      "num_tokens": 194371659.0,
      "step": 6089
    },
    {
      "entropy": 0.7782188206911087,
      "epoch": 0.5603188020770439,
      "grad_norm": 0.14945530891418457,
      "learning_rate": 8.13260956236391e-05,
      "loss": 0.7655,
      "mean_token_accuracy": 0.778135534375906,
      "num_tokens": 194403613.0,
      "step": 6090
    },
    {
      "entropy": 0.757145032286644,
      "epoch": 0.560410808448485,
      "grad_norm": 0.15921446681022644,
      "learning_rate": 8.132302879749747e-05,
      "loss": 0.7408,
      "mean_token_accuracy": 0.7832129076123238,
      "num_tokens": 194434770.0,
      "step": 6091
    },
    {
      "entropy": 0.6920980978757143,
      "epoch": 0.5605028148199263,
      "grad_norm": 0.1512448638677597,
      "learning_rate": 8.131996197135586e-05,
      "loss": 0.6882,
      "mean_token_accuracy": 0.7951634861528873,
      "num_tokens": 194466855.0,
      "step": 6092
    },
    {
      "entropy": 0.7823742255568504,
      "epoch": 0.5605948211913675,
      "grad_norm": 0.15238213539123535,
      "learning_rate": 8.131689514521422e-05,
      "loss": 0.7809,
      "mean_token_accuracy": 0.7728685215115547,
      "num_tokens": 194498964.0,
      "step": 6093
    },
    {
      "entropy": 0.7769000455737114,
      "epoch": 0.5606868275628087,
      "grad_norm": 0.16768601536750793,
      "learning_rate": 8.131382831907259e-05,
      "loss": 0.7782,
      "mean_token_accuracy": 0.7740562595427036,
      "num_tokens": 194530011.0,
      "step": 6094
    },
    {
      "entropy": 0.6390567254275084,
      "epoch": 0.56077883393425,
      "grad_norm": 0.155293807387352,
      "learning_rate": 8.131076149293096e-05,
      "loss": 0.6364,
      "mean_token_accuracy": 0.8107830584049225,
      "num_tokens": 194562452.0,
      "step": 6095
    },
    {
      "entropy": 0.7069685803726315,
      "epoch": 0.5608708403056911,
      "grad_norm": 0.15363280475139618,
      "learning_rate": 8.130769466678934e-05,
      "loss": 0.6915,
      "mean_token_accuracy": 0.8028281889855862,
      "num_tokens": 194594492.0,
      "step": 6096
    },
    {
      "entropy": 0.8511045668274164,
      "epoch": 0.5609628466771324,
      "grad_norm": 0.1620863527059555,
      "learning_rate": 8.130462784064772e-05,
      "loss": 0.8566,
      "mean_token_accuracy": 0.7509987615048885,
      "num_tokens": 194626349.0,
      "step": 6097
    },
    {
      "entropy": 0.8021703772246838,
      "epoch": 0.5610548530485736,
      "grad_norm": 0.16097944974899292,
      "learning_rate": 8.130156101450609e-05,
      "loss": 0.7863,
      "mean_token_accuracy": 0.7705454304814339,
      "num_tokens": 194658513.0,
      "step": 6098
    },
    {
      "entropy": 0.7856633272022009,
      "epoch": 0.5611468594200149,
      "grad_norm": 0.16191501915454865,
      "learning_rate": 8.129849418836446e-05,
      "loss": 0.7618,
      "mean_token_accuracy": 0.7739258706569672,
      "num_tokens": 194689535.0,
      "step": 6099
    },
    {
      "entropy": 0.6865562293678522,
      "epoch": 0.5612388657914561,
      "grad_norm": 0.16336694359779358,
      "learning_rate": 8.129542736222284e-05,
      "loss": 0.6671,
      "mean_token_accuracy": 0.8052382580935955,
      "num_tokens": 194721543.0,
      "step": 6100
    },
    {
      "entropy": 0.7571423090994358,
      "epoch": 0.5613308721628972,
      "grad_norm": 0.16382035613059998,
      "learning_rate": 8.129236053608121e-05,
      "loss": 0.754,
      "mean_token_accuracy": 0.775886058807373,
      "num_tokens": 194753296.0,
      "step": 6101
    },
    {
      "entropy": 0.7820588517934084,
      "epoch": 0.5614228785343385,
      "grad_norm": 0.16012068092823029,
      "learning_rate": 8.128929370993959e-05,
      "loss": 0.778,
      "mean_token_accuracy": 0.7749639749526978,
      "num_tokens": 194785637.0,
      "step": 6102
    },
    {
      "entropy": 0.7286719297990203,
      "epoch": 0.5615148849057797,
      "grad_norm": 0.15041829645633698,
      "learning_rate": 8.128622688379797e-05,
      "loss": 0.6972,
      "mean_token_accuracy": 0.7892908677458763,
      "num_tokens": 194817186.0,
      "step": 6103
    },
    {
      "entropy": 0.6936855968087912,
      "epoch": 0.561606891277221,
      "grad_norm": 0.24346917867660522,
      "learning_rate": 8.128316005765634e-05,
      "loss": 0.6721,
      "mean_token_accuracy": 0.7980751916766167,
      "num_tokens": 194849424.0,
      "step": 6104
    },
    {
      "entropy": 0.7203829195350409,
      "epoch": 0.5616988976486622,
      "grad_norm": 0.15328355133533478,
      "learning_rate": 8.12800932315147e-05,
      "loss": 0.7129,
      "mean_token_accuracy": 0.7853897400200367,
      "num_tokens": 194880999.0,
      "step": 6105
    },
    {
      "entropy": 0.7144788829609752,
      "epoch": 0.5617909040201033,
      "grad_norm": 0.15113526582717896,
      "learning_rate": 8.127702640537307e-05,
      "loss": 0.6956,
      "mean_token_accuracy": 0.7952021025121212,
      "num_tokens": 194913261.0,
      "step": 6106
    },
    {
      "entropy": 0.7675737217068672,
      "epoch": 0.5618829103915446,
      "grad_norm": 0.152833491563797,
      "learning_rate": 8.127395957923146e-05,
      "loss": 0.7498,
      "mean_token_accuracy": 0.7768226489424706,
      "num_tokens": 194944757.0,
      "step": 6107
    },
    {
      "entropy": 0.8459290526807308,
      "epoch": 0.5619749167629858,
      "grad_norm": 0.15823055803775787,
      "learning_rate": 8.127089275308984e-05,
      "loss": 0.8417,
      "mean_token_accuracy": 0.7560755088925362,
      "num_tokens": 194977166.0,
      "step": 6108
    },
    {
      "entropy": 0.6897333785891533,
      "epoch": 0.5620669231344271,
      "grad_norm": 0.15835346281528473,
      "learning_rate": 8.12678259269482e-05,
      "loss": 0.6736,
      "mean_token_accuracy": 0.8037987314164639,
      "num_tokens": 195009414.0,
      "step": 6109
    },
    {
      "entropy": 0.7118687871843576,
      "epoch": 0.5621589295058683,
      "grad_norm": 0.160564124584198,
      "learning_rate": 8.126475910080657e-05,
      "loss": 0.7048,
      "mean_token_accuracy": 0.7956867627799511,
      "num_tokens": 195040623.0,
      "step": 6110
    },
    {
      "entropy": 0.7350160628557205,
      "epoch": 0.5622509358773095,
      "grad_norm": 0.2017514705657959,
      "learning_rate": 8.126169227466495e-05,
      "loss": 0.7255,
      "mean_token_accuracy": 0.7936176471412182,
      "num_tokens": 195072822.0,
      "step": 6111
    },
    {
      "entropy": 0.7425192967057228,
      "epoch": 0.5623429422487507,
      "grad_norm": 0.1591499298810959,
      "learning_rate": 8.125862544852332e-05,
      "loss": 0.7431,
      "mean_token_accuracy": 0.7791178710758686,
      "num_tokens": 195105270.0,
      "step": 6112
    },
    {
      "entropy": 0.7911757752299309,
      "epoch": 0.5624349486201919,
      "grad_norm": 0.15779735147953033,
      "learning_rate": 8.12555586223817e-05,
      "loss": 0.793,
      "mean_token_accuracy": 0.7651618011295795,
      "num_tokens": 195137674.0,
      "step": 6113
    },
    {
      "entropy": 0.7213008720427752,
      "epoch": 0.5625269549916332,
      "grad_norm": 0.16051360964775085,
      "learning_rate": 8.125249179624007e-05,
      "loss": 0.7387,
      "mean_token_accuracy": 0.7836836390197277,
      "num_tokens": 195169523.0,
      "step": 6114
    },
    {
      "entropy": 0.7668345365673304,
      "epoch": 0.5626189613630744,
      "grad_norm": 0.16195765137672424,
      "learning_rate": 8.124942497009845e-05,
      "loss": 0.7538,
      "mean_token_accuracy": 0.7769808769226074,
      "num_tokens": 195201122.0,
      "step": 6115
    },
    {
      "entropy": 0.7566584125161171,
      "epoch": 0.5627109677345156,
      "grad_norm": 0.16141736507415771,
      "learning_rate": 8.124635814395682e-05,
      "loss": 0.7315,
      "mean_token_accuracy": 0.7870250083506107,
      "num_tokens": 195232421.0,
      "step": 6116
    },
    {
      "entropy": 0.7927650958299637,
      "epoch": 0.5628029741059568,
      "grad_norm": 0.15335889160633087,
      "learning_rate": 8.12432913178152e-05,
      "loss": 0.7823,
      "mean_token_accuracy": 0.7658455111086369,
      "num_tokens": 195264648.0,
      "step": 6117
    },
    {
      "entropy": 0.7746234368532896,
      "epoch": 0.562894980477398,
      "grad_norm": 0.15500664710998535,
      "learning_rate": 8.124022449167357e-05,
      "loss": 0.7568,
      "mean_token_accuracy": 0.7786053195595741,
      "num_tokens": 195296982.0,
      "step": 6118
    },
    {
      "entropy": 0.7485659997910261,
      "epoch": 0.5629869868488393,
      "grad_norm": 0.16710670292377472,
      "learning_rate": 8.123715766553195e-05,
      "loss": 0.757,
      "mean_token_accuracy": 0.7778853476047516,
      "num_tokens": 195328776.0,
      "step": 6119
    },
    {
      "entropy": 0.8094190061092377,
      "epoch": 0.5630789932202805,
      "grad_norm": 0.15617284178733826,
      "learning_rate": 8.123409083939032e-05,
      "loss": 0.7962,
      "mean_token_accuracy": 0.7689793296158314,
      "num_tokens": 195361144.0,
      "step": 6120
    },
    {
      "entropy": 0.7547434903681278,
      "epoch": 0.5631709995917217,
      "grad_norm": 0.15395492315292358,
      "learning_rate": 8.123102401324869e-05,
      "loss": 0.7186,
      "mean_token_accuracy": 0.7870777808129787,
      "num_tokens": 195392482.0,
      "step": 6121
    },
    {
      "entropy": 0.6476460956037045,
      "epoch": 0.5632630059631629,
      "grad_norm": 0.14416851103305817,
      "learning_rate": 8.122795718710707e-05,
      "loss": 0.621,
      "mean_token_accuracy": 0.8144351802766323,
      "num_tokens": 195424751.0,
      "step": 6122
    },
    {
      "entropy": 0.8315981850028038,
      "epoch": 0.5633550123346042,
      "grad_norm": 0.15703003108501434,
      "learning_rate": 8.122489036096545e-05,
      "loss": 0.845,
      "mean_token_accuracy": 0.7573337219655514,
      "num_tokens": 195456311.0,
      "step": 6123
    },
    {
      "entropy": 0.773392241448164,
      "epoch": 0.5634470187060454,
      "grad_norm": 0.15961217880249023,
      "learning_rate": 8.122182353482382e-05,
      "loss": 0.7701,
      "mean_token_accuracy": 0.774012628942728,
      "num_tokens": 195488547.0,
      "step": 6124
    },
    {
      "entropy": 0.7701145112514496,
      "epoch": 0.5635390250774867,
      "grad_norm": 0.15141716599464417,
      "learning_rate": 8.121875670868219e-05,
      "loss": 0.7797,
      "mean_token_accuracy": 0.7686381414532661,
      "num_tokens": 195520526.0,
      "step": 6125
    },
    {
      "entropy": 0.8518513962626457,
      "epoch": 0.5636310314489278,
      "grad_norm": 0.15724773705005646,
      "learning_rate": 8.121568988254055e-05,
      "loss": 0.8525,
      "mean_token_accuracy": 0.7524928711354733,
      "num_tokens": 195553112.0,
      "step": 6126
    },
    {
      "entropy": 0.7912594322115183,
      "epoch": 0.563723037820369,
      "grad_norm": 0.15272276103496552,
      "learning_rate": 8.121262305639894e-05,
      "loss": 0.7818,
      "mean_token_accuracy": 0.7702792026102543,
      "num_tokens": 195585195.0,
      "step": 6127
    },
    {
      "entropy": 0.7622837461531162,
      "epoch": 0.5638150441918103,
      "grad_norm": 0.1554490625858307,
      "learning_rate": 8.120955623025732e-05,
      "loss": 0.7677,
      "mean_token_accuracy": 0.7748788073658943,
      "num_tokens": 195617515.0,
      "step": 6128
    },
    {
      "entropy": 0.8142075520008802,
      "epoch": 0.5639070505632515,
      "grad_norm": 0.1666015088558197,
      "learning_rate": 8.120648940411568e-05,
      "loss": 0.8018,
      "mean_token_accuracy": 0.7649122104048729,
      "num_tokens": 195648355.0,
      "step": 6129
    },
    {
      "entropy": 0.664508618414402,
      "epoch": 0.5639990569346928,
      "grad_norm": 0.15708576142787933,
      "learning_rate": 8.120342257797405e-05,
      "loss": 0.6502,
      "mean_token_accuracy": 0.8045610710978508,
      "num_tokens": 195680278.0,
      "step": 6130
    },
    {
      "entropy": 0.802734263241291,
      "epoch": 0.5640910633061339,
      "grad_norm": 0.15389998257160187,
      "learning_rate": 8.120035575183243e-05,
      "loss": 0.7949,
      "mean_token_accuracy": 0.7690716572105885,
      "num_tokens": 195711905.0,
      "step": 6131
    },
    {
      "entropy": 0.8230801224708557,
      "epoch": 0.5641830696775751,
      "grad_norm": 0.16071496903896332,
      "learning_rate": 8.11972889256908e-05,
      "loss": 0.8097,
      "mean_token_accuracy": 0.7673789635300636,
      "num_tokens": 195743249.0,
      "step": 6132
    },
    {
      "entropy": 0.8152808994054794,
      "epoch": 0.5642750760490164,
      "grad_norm": 0.16399963200092316,
      "learning_rate": 8.119422209954918e-05,
      "loss": 0.8255,
      "mean_token_accuracy": 0.7599371112883091,
      "num_tokens": 195775626.0,
      "step": 6133
    },
    {
      "entropy": 0.7127881590276957,
      "epoch": 0.5643670824204576,
      "grad_norm": 0.16005297005176544,
      "learning_rate": 8.119115527340756e-05,
      "loss": 0.6901,
      "mean_token_accuracy": 0.8012487255036831,
      "num_tokens": 195808117.0,
      "step": 6134
    },
    {
      "entropy": 0.6655048802495003,
      "epoch": 0.5644590887918989,
      "grad_norm": 0.14947667717933655,
      "learning_rate": 8.118808844726593e-05,
      "loss": 0.6523,
      "mean_token_accuracy": 0.805728804320097,
      "num_tokens": 195840450.0,
      "step": 6135
    },
    {
      "entropy": 0.8004254177212715,
      "epoch": 0.56455109516334,
      "grad_norm": 0.15829092264175415,
      "learning_rate": 8.11850216211243e-05,
      "loss": 0.7724,
      "mean_token_accuracy": 0.7731947265565395,
      "num_tokens": 195872119.0,
      "step": 6136
    },
    {
      "entropy": 0.791045100428164,
      "epoch": 0.5646431015347813,
      "grad_norm": 0.17074038088321686,
      "learning_rate": 8.118195479498267e-05,
      "loss": 0.7912,
      "mean_token_accuracy": 0.7718060053884983,
      "num_tokens": 195904228.0,
      "step": 6137
    },
    {
      "entropy": 0.7398163340985775,
      "epoch": 0.5647351079062225,
      "grad_norm": 0.14989621937274933,
      "learning_rate": 8.117888796884105e-05,
      "loss": 0.7365,
      "mean_token_accuracy": 0.778534971177578,
      "num_tokens": 195935386.0,
      "step": 6138
    },
    {
      "entropy": 0.8365665785968304,
      "epoch": 0.5648271142776637,
      "grad_norm": 0.1594756692647934,
      "learning_rate": 8.117582114269943e-05,
      "loss": 0.824,
      "mean_token_accuracy": 0.7674847431480885,
      "num_tokens": 195966527.0,
      "step": 6139
    },
    {
      "entropy": 0.8208596967160702,
      "epoch": 0.564919120649105,
      "grad_norm": 0.16180427372455597,
      "learning_rate": 8.11727543165578e-05,
      "loss": 0.7999,
      "mean_token_accuracy": 0.7673744596540928,
      "num_tokens": 195998389.0,
      "step": 6140
    },
    {
      "entropy": 0.7930564470589161,
      "epoch": 0.5650111270205461,
      "grad_norm": 0.16200894117355347,
      "learning_rate": 8.116968749041617e-05,
      "loss": 0.7811,
      "mean_token_accuracy": 0.772907480597496,
      "num_tokens": 196030837.0,
      "step": 6141
    },
    {
      "entropy": 0.7573768999427557,
      "epoch": 0.5651031333919874,
      "grad_norm": 0.15805624425411224,
      "learning_rate": 8.116662066427455e-05,
      "loss": 0.7376,
      "mean_token_accuracy": 0.7847044616937637,
      "num_tokens": 196062228.0,
      "step": 6142
    },
    {
      "entropy": 0.7109242975711823,
      "epoch": 0.5651951397634286,
      "grad_norm": 0.15726032853126526,
      "learning_rate": 8.116355383813292e-05,
      "loss": 0.7028,
      "mean_token_accuracy": 0.7931651435792446,
      "num_tokens": 196094384.0,
      "step": 6143
    },
    {
      "entropy": 0.7857402358204126,
      "epoch": 0.5652871461348699,
      "grad_norm": 0.16847224533557892,
      "learning_rate": 8.11604870119913e-05,
      "loss": 0.7891,
      "mean_token_accuracy": 0.7690515145659447,
      "num_tokens": 196126530.0,
      "step": 6144
    },
    {
      "entropy": 0.768998121842742,
      "epoch": 0.5653791525063111,
      "grad_norm": 0.15335261821746826,
      "learning_rate": 8.115742018584967e-05,
      "loss": 0.7527,
      "mean_token_accuracy": 0.7813721261918545,
      "num_tokens": 196158769.0,
      "step": 6145
    },
    {
      "entropy": 0.7325848154723644,
      "epoch": 0.5654711588777522,
      "grad_norm": 0.15636910498142242,
      "learning_rate": 8.115435335970805e-05,
      "loss": 0.7309,
      "mean_token_accuracy": 0.7857678011059761,
      "num_tokens": 196191164.0,
      "step": 6146
    },
    {
      "entropy": 0.8123513571918011,
      "epoch": 0.5655631652491935,
      "grad_norm": 0.1604480743408203,
      "learning_rate": 8.115128653356641e-05,
      "loss": 0.8189,
      "mean_token_accuracy": 0.7608227767050266,
      "num_tokens": 196223043.0,
      "step": 6147
    },
    {
      "entropy": 0.7324050441384315,
      "epoch": 0.5656551716206347,
      "grad_norm": 0.15382537245750427,
      "learning_rate": 8.114821970742478e-05,
      "loss": 0.7173,
      "mean_token_accuracy": 0.7896891012787819,
      "num_tokens": 196255263.0,
      "step": 6148
    },
    {
      "entropy": 0.7262407392263412,
      "epoch": 0.565747177992076,
      "grad_norm": 0.15299135446548462,
      "learning_rate": 8.114515288128316e-05,
      "loss": 0.7075,
      "mean_token_accuracy": 0.7901319786906242,
      "num_tokens": 196287767.0,
      "step": 6149
    },
    {
      "entropy": 0.7712275739759207,
      "epoch": 0.5658391843635172,
      "grad_norm": 0.16408874094486237,
      "learning_rate": 8.114208605514155e-05,
      "loss": 0.7518,
      "mean_token_accuracy": 0.7812177427113056,
      "num_tokens": 196319606.0,
      "step": 6150
    },
    {
      "entropy": 0.6844668071717024,
      "epoch": 0.5659311907349583,
      "grad_norm": 0.14928807318210602,
      "learning_rate": 8.113901922899991e-05,
      "loss": 0.6659,
      "mean_token_accuracy": 0.8055904060602188,
      "num_tokens": 196351421.0,
      "step": 6151
    },
    {
      "entropy": 0.8078666068613529,
      "epoch": 0.5660231971063996,
      "grad_norm": 0.17269165813922882,
      "learning_rate": 8.113595240285828e-05,
      "loss": 0.7974,
      "mean_token_accuracy": 0.7649477832019329,
      "num_tokens": 196383334.0,
      "step": 6152
    },
    {
      "entropy": 0.8169114775955677,
      "epoch": 0.5661152034778408,
      "grad_norm": 0.15457792580127716,
      "learning_rate": 8.113288557671665e-05,
      "loss": 0.8126,
      "mean_token_accuracy": 0.7590147852897644,
      "num_tokens": 196415365.0,
      "step": 6153
    },
    {
      "entropy": 0.8264964856207371,
      "epoch": 0.5662072098492821,
      "grad_norm": 0.1545470505952835,
      "learning_rate": 8.112981875057504e-05,
      "loss": 0.8076,
      "mean_token_accuracy": 0.7652953192591667,
      "num_tokens": 196447955.0,
      "step": 6154
    },
    {
      "entropy": 0.7642250824719667,
      "epoch": 0.5662992162207233,
      "grad_norm": 0.15807469189167023,
      "learning_rate": 8.112675192443341e-05,
      "loss": 0.7572,
      "mean_token_accuracy": 0.7753033973276615,
      "num_tokens": 196479885.0,
      "step": 6155
    },
    {
      "entropy": 0.7719323742203414,
      "epoch": 0.5663912225921645,
      "grad_norm": 0.15365390479564667,
      "learning_rate": 8.112368509829178e-05,
      "loss": 0.7639,
      "mean_token_accuracy": 0.78037004545331,
      "num_tokens": 196510979.0,
      "step": 6156
    },
    {
      "entropy": 0.8303781636059284,
      "epoch": 0.5664832289636057,
      "grad_norm": 0.1575808972120285,
      "learning_rate": 8.112061827215015e-05,
      "loss": 0.8326,
      "mean_token_accuracy": 0.7560634985566139,
      "num_tokens": 196542934.0,
      "step": 6157
    },
    {
      "entropy": 0.8315360564738512,
      "epoch": 0.566575235335047,
      "grad_norm": 0.16332711279392242,
      "learning_rate": 8.111755144600853e-05,
      "loss": 0.8325,
      "mean_token_accuracy": 0.7570298574864864,
      "num_tokens": 196575010.0,
      "step": 6158
    },
    {
      "entropy": 0.7758260369300842,
      "epoch": 0.5666672417064882,
      "grad_norm": 0.15531091392040253,
      "learning_rate": 8.111448461986691e-05,
      "loss": 0.7737,
      "mean_token_accuracy": 0.7708043642342091,
      "num_tokens": 196606709.0,
      "step": 6159
    },
    {
      "entropy": 0.6930471230298281,
      "epoch": 0.5667592480779294,
      "grad_norm": 0.15918143093585968,
      "learning_rate": 8.111141779372528e-05,
      "loss": 0.6924,
      "mean_token_accuracy": 0.7974916212260723,
      "num_tokens": 196638608.0,
      "step": 6160
    },
    {
      "entropy": 0.7711673490703106,
      "epoch": 0.5668512544493706,
      "grad_norm": 0.16319622099399567,
      "learning_rate": 8.110835096758365e-05,
      "loss": 0.7727,
      "mean_token_accuracy": 0.7750249803066254,
      "num_tokens": 196670678.0,
      "step": 6161
    },
    {
      "entropy": 0.7396946586668491,
      "epoch": 0.5669432608208118,
      "grad_norm": 0.14308281242847443,
      "learning_rate": 8.110528414144203e-05,
      "loss": 0.7313,
      "mean_token_accuracy": 0.7805506065487862,
      "num_tokens": 196703257.0,
      "step": 6162
    },
    {
      "entropy": 0.7905214931815863,
      "epoch": 0.5670352671922531,
      "grad_norm": 0.16229990124702454,
      "learning_rate": 8.11022173153004e-05,
      "loss": 0.7873,
      "mean_token_accuracy": 0.7704115249216557,
      "num_tokens": 196735782.0,
      "step": 6163
    },
    {
      "entropy": 0.6672728154808283,
      "epoch": 0.5671272735636943,
      "grad_norm": 0.1484958827495575,
      "learning_rate": 8.109915048915878e-05,
      "loss": 0.647,
      "mean_token_accuracy": 0.805192481726408,
      "num_tokens": 196767493.0,
      "step": 6164
    },
    {
      "entropy": 0.6905970051884651,
      "epoch": 0.5672192799351355,
      "grad_norm": 0.17181789875030518,
      "learning_rate": 8.109608366301716e-05,
      "loss": 0.6784,
      "mean_token_accuracy": 0.800811056047678,
      "num_tokens": 196798961.0,
      "step": 6165
    },
    {
      "entropy": 0.7037334218621254,
      "epoch": 0.5673112863065767,
      "grad_norm": 0.15447209775447845,
      "learning_rate": 8.109301683687553e-05,
      "loss": 0.6779,
      "mean_token_accuracy": 0.8004451617598534,
      "num_tokens": 196830068.0,
      "step": 6166
    },
    {
      "entropy": 0.7755861338227987,
      "epoch": 0.5674032926780179,
      "grad_norm": 0.15935882925987244,
      "learning_rate": 8.10899500107339e-05,
      "loss": 0.7515,
      "mean_token_accuracy": 0.7771556824445724,
      "num_tokens": 196861523.0,
      "step": 6167
    },
    {
      "entropy": 0.7684818115085363,
      "epoch": 0.5674952990494592,
      "grad_norm": 0.15518653392791748,
      "learning_rate": 8.108688318459226e-05,
      "loss": 0.7496,
      "mean_token_accuracy": 0.777970939874649,
      "num_tokens": 196893380.0,
      "step": 6168
    },
    {
      "entropy": 0.8566951900720596,
      "epoch": 0.5675873054209004,
      "grad_norm": 0.162309467792511,
      "learning_rate": 8.108381635845064e-05,
      "loss": 0.8541,
      "mean_token_accuracy": 0.7546774558722973,
      "num_tokens": 196925500.0,
      "step": 6169
    },
    {
      "entropy": 0.6807738281786442,
      "epoch": 0.5676793117923417,
      "grad_norm": 0.14741188287734985,
      "learning_rate": 8.108074953230902e-05,
      "loss": 0.6551,
      "mean_token_accuracy": 0.8102722056210041,
      "num_tokens": 196958202.0,
      "step": 6170
    },
    {
      "entropy": 0.7160360049456358,
      "epoch": 0.5677713181637828,
      "grad_norm": 0.16452574729919434,
      "learning_rate": 8.107768270616739e-05,
      "loss": 0.7118,
      "mean_token_accuracy": 0.7907982096076012,
      "num_tokens": 196989337.0,
      "step": 6171
    },
    {
      "entropy": 0.7993422020226717,
      "epoch": 0.567863324535224,
      "grad_norm": 0.1538527011871338,
      "learning_rate": 8.107461588002576e-05,
      "loss": 0.7838,
      "mean_token_accuracy": 0.7752147018909454,
      "num_tokens": 197021778.0,
      "step": 6172
    },
    {
      "entropy": 0.6992855500429869,
      "epoch": 0.5679553309066653,
      "grad_norm": 0.1557975709438324,
      "learning_rate": 8.107154905388414e-05,
      "loss": 0.6968,
      "mean_token_accuracy": 0.7975872084498405,
      "num_tokens": 197054465.0,
      "step": 6173
    },
    {
      "entropy": 0.7260349206626415,
      "epoch": 0.5680473372781065,
      "grad_norm": 0.15975819528102875,
      "learning_rate": 8.106848222774251e-05,
      "loss": 0.7253,
      "mean_token_accuracy": 0.7862066179513931,
      "num_tokens": 197086835.0,
      "step": 6174
    },
    {
      "entropy": 0.8396004531532526,
      "epoch": 0.5681393436495478,
      "grad_norm": 0.16089491546154022,
      "learning_rate": 8.106541540160089e-05,
      "loss": 0.8463,
      "mean_token_accuracy": 0.7555175721645355,
      "num_tokens": 197118356.0,
      "step": 6175
    },
    {
      "entropy": 0.7853314578533173,
      "epoch": 0.5682313500209889,
      "grad_norm": 0.17223531007766724,
      "learning_rate": 8.106234857545926e-05,
      "loss": 0.7877,
      "mean_token_accuracy": 0.7695168107748032,
      "num_tokens": 197149808.0,
      "step": 6176
    },
    {
      "entropy": 0.7925128415226936,
      "epoch": 0.5683233563924301,
      "grad_norm": 0.15103602409362793,
      "learning_rate": 8.105928174931764e-05,
      "loss": 0.7843,
      "mean_token_accuracy": 0.7670269906520844,
      "num_tokens": 197181517.0,
      "step": 6177
    },
    {
      "entropy": 0.7623697705566883,
      "epoch": 0.5684153627638714,
      "grad_norm": 0.1669042855501175,
      "learning_rate": 8.105621492317601e-05,
      "loss": 0.7688,
      "mean_token_accuracy": 0.7763974107801914,
      "num_tokens": 197213470.0,
      "step": 6178
    },
    {
      "entropy": 0.8445915803313255,
      "epoch": 0.5685073691353126,
      "grad_norm": 0.15992748737335205,
      "learning_rate": 8.105314809703438e-05,
      "loss": 0.8423,
      "mean_token_accuracy": 0.7566696777939796,
      "num_tokens": 197244553.0,
      "step": 6179
    },
    {
      "entropy": 0.8577238824218512,
      "epoch": 0.5685993755067539,
      "grad_norm": 0.1553870290517807,
      "learning_rate": 8.105008127089276e-05,
      "loss": 0.8273,
      "mean_token_accuracy": 0.7624795325100422,
      "num_tokens": 197276497.0,
      "step": 6180
    },
    {
      "entropy": 0.7570530418306589,
      "epoch": 0.568691381878195,
      "grad_norm": 0.1537483036518097,
      "learning_rate": 8.104701444475114e-05,
      "loss": 0.7303,
      "mean_token_accuracy": 0.7856259159743786,
      "num_tokens": 197308073.0,
      "step": 6181
    },
    {
      "entropy": 0.8243155311793089,
      "epoch": 0.5687833882496363,
      "grad_norm": 0.15339849889278412,
      "learning_rate": 8.104394761860951e-05,
      "loss": 0.8174,
      "mean_token_accuracy": 0.7635397911071777,
      "num_tokens": 197339180.0,
      "step": 6182
    },
    {
      "entropy": 0.7051878534257412,
      "epoch": 0.5688753946210775,
      "grad_norm": 0.1502407342195511,
      "learning_rate": 8.104088079246787e-05,
      "loss": 0.6839,
      "mean_token_accuracy": 0.7973216101527214,
      "num_tokens": 197371311.0,
      "step": 6183
    },
    {
      "entropy": 0.6502017993479967,
      "epoch": 0.5689674009925187,
      "grad_norm": 0.15851067006587982,
      "learning_rate": 8.103781396632624e-05,
      "loss": 0.6269,
      "mean_token_accuracy": 0.8183737583458424,
      "num_tokens": 197403996.0,
      "step": 6184
    },
    {
      "entropy": 0.7957365158945322,
      "epoch": 0.56905940736396,
      "grad_norm": 0.16907623410224915,
      "learning_rate": 8.103474714018462e-05,
      "loss": 0.7609,
      "mean_token_accuracy": 0.7774110324680805,
      "num_tokens": 197435186.0,
      "step": 6185
    },
    {
      "entropy": 0.8238000199198723,
      "epoch": 0.5691514137354011,
      "grad_norm": 0.15342099964618683,
      "learning_rate": 8.1031680314043e-05,
      "loss": 0.8073,
      "mean_token_accuracy": 0.7619861140847206,
      "num_tokens": 197467334.0,
      "step": 6186
    },
    {
      "entropy": 0.8050150033086538,
      "epoch": 0.5692434201068424,
      "grad_norm": 0.17338941991329193,
      "learning_rate": 8.102861348790137e-05,
      "loss": 0.8069,
      "mean_token_accuracy": 0.7668003216385841,
      "num_tokens": 197498154.0,
      "step": 6187
    },
    {
      "entropy": 0.6624724827706814,
      "epoch": 0.5693354264782836,
      "grad_norm": 0.15020178258419037,
      "learning_rate": 8.102554666175974e-05,
      "loss": 0.6627,
      "mean_token_accuracy": 0.8036194406449795,
      "num_tokens": 197529998.0,
      "step": 6188
    },
    {
      "entropy": 0.7789017595350742,
      "epoch": 0.5694274328497249,
      "grad_norm": 0.16006781160831451,
      "learning_rate": 8.102247983561812e-05,
      "loss": 0.7672,
      "mean_token_accuracy": 0.7727647535502911,
      "num_tokens": 197561274.0,
      "step": 6189
    },
    {
      "entropy": 0.8419913332909346,
      "epoch": 0.5695194392211661,
      "grad_norm": 0.16737529635429382,
      "learning_rate": 8.101941300947649e-05,
      "loss": 0.8571,
      "mean_token_accuracy": 0.7527636252343655,
      "num_tokens": 197593272.0,
      "step": 6190
    },
    {
      "entropy": 0.7439729981124401,
      "epoch": 0.5696114455926072,
      "grad_norm": 0.15263953804969788,
      "learning_rate": 8.101634618333487e-05,
      "loss": 0.7346,
      "mean_token_accuracy": 0.7850265502929688,
      "num_tokens": 197625940.0,
      "step": 6191
    },
    {
      "entropy": 0.8986934423446655,
      "epoch": 0.5697034519640485,
      "grad_norm": 0.16691042482852936,
      "learning_rate": 8.101327935719324e-05,
      "loss": 0.8995,
      "mean_token_accuracy": 0.7432137355208397,
      "num_tokens": 197657286.0,
      "step": 6192
    },
    {
      "entropy": 0.7607014216482639,
      "epoch": 0.5697954583354897,
      "grad_norm": 0.1567021608352661,
      "learning_rate": 8.101021253105162e-05,
      "loss": 0.7439,
      "mean_token_accuracy": 0.7874576821923256,
      "num_tokens": 197689164.0,
      "step": 6193
    },
    {
      "entropy": 0.6582626216113567,
      "epoch": 0.569887464706931,
      "grad_norm": 0.14691008627414703,
      "learning_rate": 8.100714570490999e-05,
      "loss": 0.6282,
      "mean_token_accuracy": 0.8135863244533539,
      "num_tokens": 197720963.0,
      "step": 6194
    },
    {
      "entropy": 0.7956168241798878,
      "epoch": 0.5699794710783722,
      "grad_norm": 0.16702446341514587,
      "learning_rate": 8.100407887876836e-05,
      "loss": 0.8041,
      "mean_token_accuracy": 0.7623897716403008,
      "num_tokens": 197752210.0,
      "step": 6195
    },
    {
      "entropy": 0.6484141442924738,
      "epoch": 0.5700714774498133,
      "grad_norm": 0.1497763842344284,
      "learning_rate": 8.100101205262675e-05,
      "loss": 0.6273,
      "mean_token_accuracy": 0.8148314505815506,
      "num_tokens": 197784582.0,
      "step": 6196
    },
    {
      "entropy": 0.7808452155441046,
      "epoch": 0.5701634838212546,
      "grad_norm": 0.15586259961128235,
      "learning_rate": 8.099794522648512e-05,
      "loss": 0.7685,
      "mean_token_accuracy": 0.7753037884831429,
      "num_tokens": 197816889.0,
      "step": 6197
    },
    {
      "entropy": 0.7790293619036674,
      "epoch": 0.5702554901926958,
      "grad_norm": 0.1523497849702835,
      "learning_rate": 8.099487840034349e-05,
      "loss": 0.765,
      "mean_token_accuracy": 0.7791965566575527,
      "num_tokens": 197848959.0,
      "step": 6198
    },
    {
      "entropy": 0.6646057441830635,
      "epoch": 0.5703474965641371,
      "grad_norm": 0.16700167953968048,
      "learning_rate": 8.099181157420186e-05,
      "loss": 0.6563,
      "mean_token_accuracy": 0.8024658709764481,
      "num_tokens": 197881025.0,
      "step": 6199
    },
    {
      "entropy": 0.7275566309690475,
      "epoch": 0.5704395029355783,
      "grad_norm": 0.15294766426086426,
      "learning_rate": 8.098874474806024e-05,
      "loss": 0.7195,
      "mean_token_accuracy": 0.7877686880528927,
      "num_tokens": 197913793.0,
      "step": 6200
    },
    {
      "entropy": 0.6875928975641727,
      "epoch": 0.5705315093070195,
      "grad_norm": 0.1507503241300583,
      "learning_rate": 8.098567792191862e-05,
      "loss": 0.6772,
      "mean_token_accuracy": 0.7976404242217541,
      "num_tokens": 197946556.0,
      "step": 6201
    },
    {
      "entropy": 0.6837386935949326,
      "epoch": 0.5706235156784607,
      "grad_norm": 0.15485689043998718,
      "learning_rate": 8.098261109577699e-05,
      "loss": 0.6721,
      "mean_token_accuracy": 0.8040144257247448,
      "num_tokens": 197978913.0,
      "step": 6202
    },
    {
      "entropy": 0.7222643317654729,
      "epoch": 0.570715522049902,
      "grad_norm": 0.15804944932460785,
      "learning_rate": 8.097954426963535e-05,
      "loss": 0.6924,
      "mean_token_accuracy": 0.7996182702481747,
      "num_tokens": 198011019.0,
      "step": 6203
    },
    {
      "entropy": 0.7542238840833306,
      "epoch": 0.5708075284213432,
      "grad_norm": 0.16466142237186432,
      "learning_rate": 8.097647744349374e-05,
      "loss": 0.7514,
      "mean_token_accuracy": 0.7767388671636581,
      "num_tokens": 198041299.0,
      "step": 6204
    },
    {
      "entropy": 0.6970021408051252,
      "epoch": 0.5708995347927844,
      "grad_norm": 0.16383104026317596,
      "learning_rate": 8.09734106173521e-05,
      "loss": 0.6905,
      "mean_token_accuracy": 0.796870231628418,
      "num_tokens": 198073807.0,
      "step": 6205
    },
    {
      "entropy": 0.802641874179244,
      "epoch": 0.5709915411642256,
      "grad_norm": 0.15970644354820251,
      "learning_rate": 8.097034379121048e-05,
      "loss": 0.7759,
      "mean_token_accuracy": 0.773629367351532,
      "num_tokens": 198105528.0,
      "step": 6206
    },
    {
      "entropy": 0.770581029355526,
      "epoch": 0.5710835475356668,
      "grad_norm": 0.16254067420959473,
      "learning_rate": 8.096727696506885e-05,
      "loss": 0.7504,
      "mean_token_accuracy": 0.7814521603286266,
      "num_tokens": 198137156.0,
      "step": 6207
    },
    {
      "entropy": 0.6271659880876541,
      "epoch": 0.5711755539071081,
      "grad_norm": 0.15341006219387054,
      "learning_rate": 8.096421013892723e-05,
      "loss": 0.6146,
      "mean_token_accuracy": 0.8198251612484455,
      "num_tokens": 198169177.0,
      "step": 6208
    },
    {
      "entropy": 0.8267356157302856,
      "epoch": 0.5712675602785493,
      "grad_norm": 0.15903159976005554,
      "learning_rate": 8.09611433127856e-05,
      "loss": 0.8095,
      "mean_token_accuracy": 0.7603858299553394,
      "num_tokens": 198201307.0,
      "step": 6209
    },
    {
      "entropy": 0.8074181471019983,
      "epoch": 0.5713595666499905,
      "grad_norm": 0.15600036084651947,
      "learning_rate": 8.095807648664397e-05,
      "loss": 0.7956,
      "mean_token_accuracy": 0.7657137960195541,
      "num_tokens": 198232995.0,
      "step": 6210
    },
    {
      "entropy": 0.6643227264285088,
      "epoch": 0.5714515730214317,
      "grad_norm": 0.14710359275341034,
      "learning_rate": 8.095500966050235e-05,
      "loss": 0.6504,
      "mean_token_accuracy": 0.8034306019544601,
      "num_tokens": 198264548.0,
      "step": 6211
    },
    {
      "entropy": 0.8369471915066242,
      "epoch": 0.5715435793928729,
      "grad_norm": 0.17524977028369904,
      "learning_rate": 8.095194283436073e-05,
      "loss": 0.8321,
      "mean_token_accuracy": 0.7631350979208946,
      "num_tokens": 198296943.0,
      "step": 6212
    },
    {
      "entropy": 0.7134013958275318,
      "epoch": 0.5716355857643142,
      "grad_norm": 0.1606544852256775,
      "learning_rate": 8.09488760082191e-05,
      "loss": 0.6972,
      "mean_token_accuracy": 0.7913184091448784,
      "num_tokens": 198327760.0,
      "step": 6213
    },
    {
      "entropy": 0.7649447033181787,
      "epoch": 0.5717275921357554,
      "grad_norm": 0.15175214409828186,
      "learning_rate": 8.094580918207747e-05,
      "loss": 0.7353,
      "mean_token_accuracy": 0.787051398307085,
      "num_tokens": 198359612.0,
      "step": 6214
    },
    {
      "entropy": 0.7173961289227009,
      "epoch": 0.5718195985071967,
      "grad_norm": 0.15355461835861206,
      "learning_rate": 8.094274235593584e-05,
      "loss": 0.6955,
      "mean_token_accuracy": 0.7894110158085823,
      "num_tokens": 198392145.0,
      "step": 6215
    },
    {
      "entropy": 0.7497940175235271,
      "epoch": 0.5719116048786378,
      "grad_norm": 0.15270863473415375,
      "learning_rate": 8.093967552979422e-05,
      "loss": 0.7295,
      "mean_token_accuracy": 0.7809534445405006,
      "num_tokens": 198424079.0,
      "step": 6216
    },
    {
      "entropy": 0.8851457964628935,
      "epoch": 0.572003611250079,
      "grad_norm": 0.1582627147436142,
      "learning_rate": 8.09366087036526e-05,
      "loss": 0.8776,
      "mean_token_accuracy": 0.7464221902191639,
      "num_tokens": 198456247.0,
      "step": 6217
    },
    {
      "entropy": 0.8530375827103853,
      "epoch": 0.5720956176215203,
      "grad_norm": 0.1583968698978424,
      "learning_rate": 8.093354187751097e-05,
      "loss": 0.8355,
      "mean_token_accuracy": 0.7587568536400795,
      "num_tokens": 198488390.0,
      "step": 6218
    },
    {
      "entropy": 0.7185436654835939,
      "epoch": 0.5721876239929615,
      "grad_norm": 0.15789027512073517,
      "learning_rate": 8.093047505136933e-05,
      "loss": 0.7267,
      "mean_token_accuracy": 0.7878854684531689,
      "num_tokens": 198520778.0,
      "step": 6219
    },
    {
      "entropy": 0.7585103958845139,
      "epoch": 0.5722796303644028,
      "grad_norm": 0.16389408707618713,
      "learning_rate": 8.092740822522772e-05,
      "loss": 0.7468,
      "mean_token_accuracy": 0.7808569818735123,
      "num_tokens": 198552643.0,
      "step": 6220
    },
    {
      "entropy": 0.7793201692402363,
      "epoch": 0.5723716367358439,
      "grad_norm": 0.17141634225845337,
      "learning_rate": 8.092434139908608e-05,
      "loss": 0.7537,
      "mean_token_accuracy": 0.778145357966423,
      "num_tokens": 198583992.0,
      "step": 6221
    },
    {
      "entropy": 0.7885103989392519,
      "epoch": 0.5724636431072851,
      "grad_norm": 0.155290424823761,
      "learning_rate": 8.092127457294447e-05,
      "loss": 0.7806,
      "mean_token_accuracy": 0.7717696614563465,
      "num_tokens": 198616760.0,
      "step": 6222
    },
    {
      "entropy": 0.9061334617435932,
      "epoch": 0.5725556494787264,
      "grad_norm": 0.16027185320854187,
      "learning_rate": 8.091820774680283e-05,
      "loss": 0.9009,
      "mean_token_accuracy": 0.7403997220098972,
      "num_tokens": 198649117.0,
      "step": 6223
    },
    {
      "entropy": 0.7770516835153103,
      "epoch": 0.5726476558501676,
      "grad_norm": 0.1559436321258545,
      "learning_rate": 8.091514092066121e-05,
      "loss": 0.7529,
      "mean_token_accuracy": 0.7775509767234325,
      "num_tokens": 198681580.0,
      "step": 6224
    },
    {
      "entropy": 0.8140580803155899,
      "epoch": 0.5727396622216089,
      "grad_norm": 0.15455497801303864,
      "learning_rate": 8.091207409451958e-05,
      "loss": 0.8115,
      "mean_token_accuracy": 0.76488022133708,
      "num_tokens": 198713211.0,
      "step": 6225
    },
    {
      "entropy": 0.7667206395417452,
      "epoch": 0.57283166859305,
      "grad_norm": 0.15765783190727234,
      "learning_rate": 8.090900726837795e-05,
      "loss": 0.7675,
      "mean_token_accuracy": 0.7767193801701069,
      "num_tokens": 198744970.0,
      "step": 6226
    },
    {
      "entropy": 0.7060567028820515,
      "epoch": 0.5729236749644913,
      "grad_norm": 0.1540021002292633,
      "learning_rate": 8.090594044223633e-05,
      "loss": 0.6862,
      "mean_token_accuracy": 0.7956785336136818,
      "num_tokens": 198777108.0,
      "step": 6227
    },
    {
      "entropy": 0.8524210881441832,
      "epoch": 0.5730156813359325,
      "grad_norm": 0.16292454302310944,
      "learning_rate": 8.090287361609471e-05,
      "loss": 0.8396,
      "mean_token_accuracy": 0.7589609250426292,
      "num_tokens": 198808552.0,
      "step": 6228
    },
    {
      "entropy": 0.6729143830016255,
      "epoch": 0.5731076877073737,
      "grad_norm": 0.15998893976211548,
      "learning_rate": 8.089980678995308e-05,
      "loss": 0.6578,
      "mean_token_accuracy": 0.8012032844126225,
      "num_tokens": 198840560.0,
      "step": 6229
    },
    {
      "entropy": 0.783919271081686,
      "epoch": 0.573199694078815,
      "grad_norm": 0.15653446316719055,
      "learning_rate": 8.089673996381145e-05,
      "loss": 0.7959,
      "mean_token_accuracy": 0.7697239816188812,
      "num_tokens": 198873328.0,
      "step": 6230
    },
    {
      "entropy": 0.7464323192834854,
      "epoch": 0.5732917004502561,
      "grad_norm": 0.16626109182834625,
      "learning_rate": 8.089367313766983e-05,
      "loss": 0.7411,
      "mean_token_accuracy": 0.7789025232195854,
      "num_tokens": 198904148.0,
      "step": 6231
    },
    {
      "entropy": 0.7274770140647888,
      "epoch": 0.5733837068216974,
      "grad_norm": 0.15386395156383514,
      "learning_rate": 8.08906063115282e-05,
      "loss": 0.727,
      "mean_token_accuracy": 0.7847911305725574,
      "num_tokens": 198936559.0,
      "step": 6232
    },
    {
      "entropy": 0.6563061624765396,
      "epoch": 0.5734757131931386,
      "grad_norm": 0.16083309054374695,
      "learning_rate": 8.088753948538658e-05,
      "loss": 0.645,
      "mean_token_accuracy": 0.8080474585294724,
      "num_tokens": 198968702.0,
      "step": 6233
    },
    {
      "entropy": 0.7628333643078804,
      "epoch": 0.5735677195645799,
      "grad_norm": 0.157570019364357,
      "learning_rate": 8.088447265924495e-05,
      "loss": 0.7554,
      "mean_token_accuracy": 0.7772993259131908,
      "num_tokens": 199000651.0,
      "step": 6234
    },
    {
      "entropy": 0.7644247487187386,
      "epoch": 0.5736597259360211,
      "grad_norm": 0.15128572285175323,
      "learning_rate": 8.088140583310333e-05,
      "loss": 0.7376,
      "mean_token_accuracy": 0.7818257994949818,
      "num_tokens": 199032397.0,
      "step": 6235
    },
    {
      "entropy": 0.8237657845020294,
      "epoch": 0.5737517323074622,
      "grad_norm": 0.15344345569610596,
      "learning_rate": 8.08783390069617e-05,
      "loss": 0.803,
      "mean_token_accuracy": 0.7676897905766964,
      "num_tokens": 199065133.0,
      "step": 6236
    },
    {
      "entropy": 0.7129977494478226,
      "epoch": 0.5738437386789035,
      "grad_norm": 0.1528964340686798,
      "learning_rate": 8.087527218082008e-05,
      "loss": 0.6968,
      "mean_token_accuracy": 0.7889960184693336,
      "num_tokens": 199097642.0,
      "step": 6237
    },
    {
      "entropy": 0.8824027255177498,
      "epoch": 0.5739357450503447,
      "grad_norm": 0.16749075055122375,
      "learning_rate": 8.087220535467845e-05,
      "loss": 0.8649,
      "mean_token_accuracy": 0.7490801364183426,
      "num_tokens": 199128664.0,
      "step": 6238
    },
    {
      "entropy": 0.6857462236657739,
      "epoch": 0.574027751421786,
      "grad_norm": 0.15057803690433502,
      "learning_rate": 8.086913852853683e-05,
      "loss": 0.6633,
      "mean_token_accuracy": 0.8048806637525558,
      "num_tokens": 199160869.0,
      "step": 6239
    },
    {
      "entropy": 0.7811956889927387,
      "epoch": 0.5741197577932272,
      "grad_norm": 0.15137752890586853,
      "learning_rate": 8.08660717023952e-05,
      "loss": 0.7645,
      "mean_token_accuracy": 0.7702753134071827,
      "num_tokens": 199192284.0,
      "step": 6240
    },
    {
      "entropy": 0.7460326291620731,
      "epoch": 0.5742117641646683,
      "grad_norm": 0.16718575358390808,
      "learning_rate": 8.086300487625356e-05,
      "loss": 0.7293,
      "mean_token_accuracy": 0.7882943488657475,
      "num_tokens": 199223781.0,
      "step": 6241
    },
    {
      "entropy": 0.8988511227071285,
      "epoch": 0.5743037705361096,
      "grad_norm": 0.1644546389579773,
      "learning_rate": 8.085993805011195e-05,
      "loss": 0.9115,
      "mean_token_accuracy": 0.735262107104063,
      "num_tokens": 199254639.0,
      "step": 6242
    },
    {
      "entropy": 0.8371548857539892,
      "epoch": 0.5743957769075508,
      "grad_norm": 0.15682707726955414,
      "learning_rate": 8.085687122397033e-05,
      "loss": 0.8484,
      "mean_token_accuracy": 0.7512327618896961,
      "num_tokens": 199286450.0,
      "step": 6243
    },
    {
      "entropy": 0.8172141909599304,
      "epoch": 0.5744877832789921,
      "grad_norm": 0.15649136900901794,
      "learning_rate": 8.08538043978287e-05,
      "loss": 0.8156,
      "mean_token_accuracy": 0.7650195769965649,
      "num_tokens": 199317621.0,
      "step": 6244
    },
    {
      "entropy": 0.7647748794406652,
      "epoch": 0.5745797896504333,
      "grad_norm": 0.16243413090705872,
      "learning_rate": 8.085073757168706e-05,
      "loss": 0.7687,
      "mean_token_accuracy": 0.7743542455136776,
      "num_tokens": 199349760.0,
      "step": 6245
    },
    {
      "entropy": 0.6866677738726139,
      "epoch": 0.5746717960218746,
      "grad_norm": 0.15089498460292816,
      "learning_rate": 8.084767074554543e-05,
      "loss": 0.6723,
      "mean_token_accuracy": 0.7983603030443192,
      "num_tokens": 199381388.0,
      "step": 6246
    },
    {
      "entropy": 0.7971693053841591,
      "epoch": 0.5747638023933157,
      "grad_norm": 0.16512492299079895,
      "learning_rate": 8.084460391940381e-05,
      "loss": 0.7905,
      "mean_token_accuracy": 0.7668520994484425,
      "num_tokens": 199412853.0,
      "step": 6247
    },
    {
      "entropy": 0.7705144733190536,
      "epoch": 0.574855808764757,
      "grad_norm": 0.17310816049575806,
      "learning_rate": 8.084153709326219e-05,
      "loss": 0.7584,
      "mean_token_accuracy": 0.7777280397713184,
      "num_tokens": 199443710.0,
      "step": 6248
    },
    {
      "entropy": 0.8104498349130154,
      "epoch": 0.5749478151361982,
      "grad_norm": 0.15700627863407135,
      "learning_rate": 8.083847026712056e-05,
      "loss": 0.8031,
      "mean_token_accuracy": 0.7668367810547352,
      "num_tokens": 199475317.0,
      "step": 6249
    },
    {
      "entropy": 0.7702053599059582,
      "epoch": 0.5750398215076394,
      "grad_norm": 0.15558089315891266,
      "learning_rate": 8.083540344097893e-05,
      "loss": 0.7721,
      "mean_token_accuracy": 0.7722570896148682,
      "num_tokens": 199507302.0,
      "step": 6250
    },
    {
      "entropy": 0.8306268192827702,
      "epoch": 0.5751318278790807,
      "grad_norm": 0.1598258763551712,
      "learning_rate": 8.083233661483731e-05,
      "loss": 0.8092,
      "mean_token_accuracy": 0.7627677880227566,
      "num_tokens": 199539276.0,
      "step": 6251
    },
    {
      "entropy": 0.8169474769383669,
      "epoch": 0.5752238342505218,
      "grad_norm": 0.15492399036884308,
      "learning_rate": 8.082926978869568e-05,
      "loss": 0.7996,
      "mean_token_accuracy": 0.7648691385984421,
      "num_tokens": 199571957.0,
      "step": 6252
    },
    {
      "entropy": 0.7245679926127195,
      "epoch": 0.5753158406219631,
      "grad_norm": 0.16374364495277405,
      "learning_rate": 8.082620296255406e-05,
      "loss": 0.7121,
      "mean_token_accuracy": 0.7892801389098167,
      "num_tokens": 199603520.0,
      "step": 6253
    },
    {
      "entropy": 0.7771807797253132,
      "epoch": 0.5754078469934043,
      "grad_norm": 0.14942139387130737,
      "learning_rate": 8.082313613641244e-05,
      "loss": 0.7548,
      "mean_token_accuracy": 0.7773909978568554,
      "num_tokens": 199635829.0,
      "step": 6254
    },
    {
      "entropy": 0.7479421645402908,
      "epoch": 0.5754998533648455,
      "grad_norm": 0.16987542808055878,
      "learning_rate": 8.082006931027081e-05,
      "loss": 0.7538,
      "mean_token_accuracy": 0.782465361058712,
      "num_tokens": 199667768.0,
      "step": 6255
    },
    {
      "entropy": 0.7988296607509255,
      "epoch": 0.5755918597362868,
      "grad_norm": 0.1660088449716568,
      "learning_rate": 8.081700248412918e-05,
      "loss": 0.7903,
      "mean_token_accuracy": 0.7731588669121265,
      "num_tokens": 199698312.0,
      "step": 6256
    },
    {
      "entropy": 0.6389548163861036,
      "epoch": 0.5756838661077279,
      "grad_norm": 0.16295701265335083,
      "learning_rate": 8.081393565798754e-05,
      "loss": 0.6237,
      "mean_token_accuracy": 0.8134323619306087,
      "num_tokens": 199729978.0,
      "step": 6257
    },
    {
      "entropy": 0.6478856466710567,
      "epoch": 0.5757758724791692,
      "grad_norm": 0.14736109972000122,
      "learning_rate": 8.081086883184593e-05,
      "loss": 0.6277,
      "mean_token_accuracy": 0.8156049735844135,
      "num_tokens": 199762205.0,
      "step": 6258
    },
    {
      "entropy": 0.7816128265112638,
      "epoch": 0.5758678788506104,
      "grad_norm": 0.15775717794895172,
      "learning_rate": 8.080780200570431e-05,
      "loss": 0.7682,
      "mean_token_accuracy": 0.7776088006794453,
      "num_tokens": 199793273.0,
      "step": 6259
    },
    {
      "entropy": 0.7053835839033127,
      "epoch": 0.5759598852220517,
      "grad_norm": 0.1524328589439392,
      "learning_rate": 8.080473517956268e-05,
      "loss": 0.7064,
      "mean_token_accuracy": 0.7927264831960201,
      "num_tokens": 199825915.0,
      "step": 6260
    },
    {
      "entropy": 0.8594235852360725,
      "epoch": 0.5760518915934929,
      "grad_norm": 0.16342085599899292,
      "learning_rate": 8.080166835342104e-05,
      "loss": 0.8436,
      "mean_token_accuracy": 0.7571351453661919,
      "num_tokens": 199857412.0,
      "step": 6261
    },
    {
      "entropy": 0.706976380199194,
      "epoch": 0.576143897964934,
      "grad_norm": 0.16467995941638947,
      "learning_rate": 8.079860152727942e-05,
      "loss": 0.6874,
      "mean_token_accuracy": 0.7986434772610664,
      "num_tokens": 199889118.0,
      "step": 6262
    },
    {
      "entropy": 0.6753663513809443,
      "epoch": 0.5762359043363753,
      "grad_norm": 0.15004122257232666,
      "learning_rate": 8.079553470113779e-05,
      "loss": 0.6525,
      "mean_token_accuracy": 0.8112446069717407,
      "num_tokens": 199920761.0,
      "step": 6263
    },
    {
      "entropy": 0.7758138291537762,
      "epoch": 0.5763279107078165,
      "grad_norm": 0.1565736085176468,
      "learning_rate": 8.079246787499617e-05,
      "loss": 0.7496,
      "mean_token_accuracy": 0.781252097338438,
      "num_tokens": 199951485.0,
      "step": 6264
    },
    {
      "entropy": 0.7535859160125256,
      "epoch": 0.5764199170792578,
      "grad_norm": 0.16182349622249603,
      "learning_rate": 8.078940104885454e-05,
      "loss": 0.7329,
      "mean_token_accuracy": 0.7852434031665325,
      "num_tokens": 199983390.0,
      "step": 6265
    },
    {
      "entropy": 0.7597635425627232,
      "epoch": 0.576511923450699,
      "grad_norm": 0.1602918803691864,
      "learning_rate": 8.078633422271292e-05,
      "loss": 0.7517,
      "mean_token_accuracy": 0.7775381207466125,
      "num_tokens": 200015226.0,
      "step": 6266
    },
    {
      "entropy": 0.736896762624383,
      "epoch": 0.5766039298221401,
      "grad_norm": 0.15868572890758514,
      "learning_rate": 8.078326739657129e-05,
      "loss": 0.7333,
      "mean_token_accuracy": 0.7858630418777466,
      "num_tokens": 200047109.0,
      "step": 6267
    },
    {
      "entropy": 0.798507371917367,
      "epoch": 0.5766959361935814,
      "grad_norm": 0.15632201731204987,
      "learning_rate": 8.078020057042966e-05,
      "loss": 0.7857,
      "mean_token_accuracy": 0.7678864486515522,
      "num_tokens": 200078482.0,
      "step": 6268
    },
    {
      "entropy": 0.7783847693353891,
      "epoch": 0.5767879425650226,
      "grad_norm": 0.16824515163898468,
      "learning_rate": 8.077713374428804e-05,
      "loss": 0.7698,
      "mean_token_accuracy": 0.7800754606723785,
      "num_tokens": 200110232.0,
      "step": 6269
    },
    {
      "entropy": 0.7546468870714307,
      "epoch": 0.5768799489364639,
      "grad_norm": 0.1617763489484787,
      "learning_rate": 8.077406691814642e-05,
      "loss": 0.7445,
      "mean_token_accuracy": 0.7820006348192692,
      "num_tokens": 200142458.0,
      "step": 6270
    },
    {
      "entropy": 0.7753540873527527,
      "epoch": 0.5769719553079051,
      "grad_norm": 0.15921179950237274,
      "learning_rate": 8.077100009200479e-05,
      "loss": 0.7648,
      "mean_token_accuracy": 0.7768374234437943,
      "num_tokens": 200174305.0,
      "step": 6271
    },
    {
      "entropy": 0.7191921314224601,
      "epoch": 0.5770639616793463,
      "grad_norm": 0.15235769748687744,
      "learning_rate": 8.076793326586316e-05,
      "loss": 0.7196,
      "mean_token_accuracy": 0.793431181460619,
      "num_tokens": 200205721.0,
      "step": 6272
    },
    {
      "entropy": 0.7635479914024472,
      "epoch": 0.5771559680507875,
      "grad_norm": 0.15391786396503448,
      "learning_rate": 8.076486643972153e-05,
      "loss": 0.7528,
      "mean_token_accuracy": 0.7801074832677841,
      "num_tokens": 200238178.0,
      "step": 6273
    },
    {
      "entropy": 0.8130598776042461,
      "epoch": 0.5772479744222287,
      "grad_norm": 0.16078191995620728,
      "learning_rate": 8.076179961357992e-05,
      "loss": 0.8056,
      "mean_token_accuracy": 0.7682462595403194,
      "num_tokens": 200269505.0,
      "step": 6274
    },
    {
      "entropy": 0.8972436040639877,
      "epoch": 0.57733998079367,
      "grad_norm": 0.16670973598957062,
      "learning_rate": 8.075873278743829e-05,
      "loss": 0.8888,
      "mean_token_accuracy": 0.744352500885725,
      "num_tokens": 200301445.0,
      "step": 6275
    },
    {
      "entropy": 0.7927631288766861,
      "epoch": 0.5774319871651112,
      "grad_norm": 0.16565723717212677,
      "learning_rate": 8.075566596129666e-05,
      "loss": 0.7816,
      "mean_token_accuracy": 0.7704471051692963,
      "num_tokens": 200333611.0,
      "step": 6276
    },
    {
      "entropy": 0.6953124552965164,
      "epoch": 0.5775239935365524,
      "grad_norm": 0.17032477259635925,
      "learning_rate": 8.075259913515502e-05,
      "loss": 0.6905,
      "mean_token_accuracy": 0.8005638271570206,
      "num_tokens": 200364639.0,
      "step": 6277
    },
    {
      "entropy": 0.7317427024245262,
      "epoch": 0.5776159999079936,
      "grad_norm": 0.1567559391260147,
      "learning_rate": 8.07495323090134e-05,
      "loss": 0.7046,
      "mean_token_accuracy": 0.7928881235420704,
      "num_tokens": 200396171.0,
      "step": 6278
    },
    {
      "entropy": 0.7859720103442669,
      "epoch": 0.5777080062794349,
      "grad_norm": 0.1623605489730835,
      "learning_rate": 8.074646548287179e-05,
      "loss": 0.7863,
      "mean_token_accuracy": 0.7702104859054089,
      "num_tokens": 200428361.0,
      "step": 6279
    },
    {
      "entropy": 0.8877792675048113,
      "epoch": 0.5778000126508761,
      "grad_norm": 0.15888014435768127,
      "learning_rate": 8.074339865673015e-05,
      "loss": 0.8853,
      "mean_token_accuracy": 0.7471574880182743,
      "num_tokens": 200460666.0,
      "step": 6280
    },
    {
      "entropy": 0.7569710407406092,
      "epoch": 0.5778920190223173,
      "grad_norm": 0.15582050383090973,
      "learning_rate": 8.074033183058852e-05,
      "loss": 0.7589,
      "mean_token_accuracy": 0.7775972820818424,
      "num_tokens": 200492816.0,
      "step": 6281
    },
    {
      "entropy": 0.8132050447165966,
      "epoch": 0.5779840253937585,
      "grad_norm": 0.15627801418304443,
      "learning_rate": 8.07372650044469e-05,
      "loss": 0.8176,
      "mean_token_accuracy": 0.7634390294551849,
      "num_tokens": 200525410.0,
      "step": 6282
    },
    {
      "entropy": 0.7583678252995014,
      "epoch": 0.5780760317651997,
      "grad_norm": 0.1570052057504654,
      "learning_rate": 8.073419817830527e-05,
      "loss": 0.748,
      "mean_token_accuracy": 0.7785113342106342,
      "num_tokens": 200557522.0,
      "step": 6283
    },
    {
      "entropy": 0.7631046678870916,
      "epoch": 0.578168038136641,
      "grad_norm": 0.15746019780635834,
      "learning_rate": 8.073113135216365e-05,
      "loss": 0.762,
      "mean_token_accuracy": 0.7783019281923771,
      "num_tokens": 200589719.0,
      "step": 6284
    },
    {
      "entropy": 0.7816500142216682,
      "epoch": 0.5782600445080822,
      "grad_norm": 0.16315355896949768,
      "learning_rate": 8.072806452602203e-05,
      "loss": 0.7835,
      "mean_token_accuracy": 0.7704741135239601,
      "num_tokens": 200621800.0,
      "step": 6285
    },
    {
      "entropy": 0.7770050968974829,
      "epoch": 0.5783520508795235,
      "grad_norm": 0.1652444452047348,
      "learning_rate": 8.07249976998804e-05,
      "loss": 0.7952,
      "mean_token_accuracy": 0.7676160708069801,
      "num_tokens": 200653463.0,
      "step": 6286
    },
    {
      "entropy": 0.8888699933886528,
      "epoch": 0.5784440572509646,
      "grad_norm": 0.16502490639686584,
      "learning_rate": 8.072193087373877e-05,
      "loss": 0.8858,
      "mean_token_accuracy": 0.7468933388590813,
      "num_tokens": 200685362.0,
      "step": 6287
    },
    {
      "entropy": 0.7972737364470959,
      "epoch": 0.5785360636224058,
      "grad_norm": 0.16355226933956146,
      "learning_rate": 8.071886404759714e-05,
      "loss": 0.7893,
      "mean_token_accuracy": 0.7700091116130352,
      "num_tokens": 200717593.0,
      "step": 6288
    },
    {
      "entropy": 0.7796119339764118,
      "epoch": 0.5786280699938471,
      "grad_norm": 0.15401561558246613,
      "learning_rate": 8.071579722145552e-05,
      "loss": 0.7628,
      "mean_token_accuracy": 0.777599386870861,
      "num_tokens": 200749280.0,
      "step": 6289
    },
    {
      "entropy": 0.8845874909311533,
      "epoch": 0.5787200763652883,
      "grad_norm": 0.16354681551456451,
      "learning_rate": 8.07127303953139e-05,
      "loss": 0.8424,
      "mean_token_accuracy": 0.7559351846575737,
      "num_tokens": 200780799.0,
      "step": 6290
    },
    {
      "entropy": 0.8468320723623037,
      "epoch": 0.5788120827367296,
      "grad_norm": 0.1604946255683899,
      "learning_rate": 8.070966356917227e-05,
      "loss": 0.8202,
      "mean_token_accuracy": 0.7621542774140835,
      "num_tokens": 200812229.0,
      "step": 6291
    },
    {
      "entropy": 0.8042459487915039,
      "epoch": 0.5789040891081707,
      "grad_norm": 0.15312406420707703,
      "learning_rate": 8.070659674303064e-05,
      "loss": 0.785,
      "mean_token_accuracy": 0.7734493762254715,
      "num_tokens": 200843993.0,
      "step": 6292
    },
    {
      "entropy": 0.8520974237471819,
      "epoch": 0.578996095479612,
      "grad_norm": 0.15955348312854767,
      "learning_rate": 8.070352991688902e-05,
      "loss": 0.8427,
      "mean_token_accuracy": 0.7540944889187813,
      "num_tokens": 200876239.0,
      "step": 6293
    },
    {
      "entropy": 0.7804214805364609,
      "epoch": 0.5790881018510532,
      "grad_norm": 0.16528089344501495,
      "learning_rate": 8.070046309074739e-05,
      "loss": 0.7818,
      "mean_token_accuracy": 0.7721461318433285,
      "num_tokens": 200907644.0,
      "step": 6294
    },
    {
      "entropy": 0.6066843327134848,
      "epoch": 0.5791801082224944,
      "grad_norm": 0.14624769985675812,
      "learning_rate": 8.069739626460577e-05,
      "loss": 0.5867,
      "mean_token_accuracy": 0.825780563056469,
      "num_tokens": 200939780.0,
      "step": 6295
    },
    {
      "entropy": 0.7197702396661043,
      "epoch": 0.5792721145939357,
      "grad_norm": 0.153267502784729,
      "learning_rate": 8.069432943846414e-05,
      "loss": 0.6924,
      "mean_token_accuracy": 0.7932593263685703,
      "num_tokens": 200971185.0,
      "step": 6296
    },
    {
      "entropy": 0.6499939672648907,
      "epoch": 0.5793641209653768,
      "grad_norm": 0.1522831916809082,
      "learning_rate": 8.069126261232252e-05,
      "loss": 0.6286,
      "mean_token_accuracy": 0.8139964155852795,
      "num_tokens": 201003199.0,
      "step": 6297
    },
    {
      "entropy": 0.7514182664453983,
      "epoch": 0.5794561273368181,
      "grad_norm": 0.1641286164522171,
      "learning_rate": 8.068819578618088e-05,
      "loss": 0.7388,
      "mean_token_accuracy": 0.778253298252821,
      "num_tokens": 201035021.0,
      "step": 6298
    },
    {
      "entropy": 0.8678616210818291,
      "epoch": 0.5795481337082593,
      "grad_norm": 0.16442568600177765,
      "learning_rate": 8.068512896003925e-05,
      "loss": 0.8634,
      "mean_token_accuracy": 0.7499817423522472,
      "num_tokens": 201067106.0,
      "step": 6299
    },
    {
      "entropy": 0.7467313800007105,
      "epoch": 0.5796401400797005,
      "grad_norm": 0.1566130816936493,
      "learning_rate": 8.068206213389763e-05,
      "loss": 0.7581,
      "mean_token_accuracy": 0.777968168258667,
      "num_tokens": 201099739.0,
      "step": 6300
    },
    {
      "entropy": 0.6662234645336866,
      "epoch": 0.5797321464511418,
      "grad_norm": 0.15311498939990997,
      "learning_rate": 8.067899530775602e-05,
      "loss": 0.6511,
      "mean_token_accuracy": 0.809533029794693,
      "num_tokens": 201131358.0,
      "step": 6301
    },
    {
      "entropy": 0.7457143813371658,
      "epoch": 0.5798241528225829,
      "grad_norm": 0.15723876655101776,
      "learning_rate": 8.067592848161438e-05,
      "loss": 0.7375,
      "mean_token_accuracy": 0.7803190276026726,
      "num_tokens": 201163090.0,
      "step": 6302
    },
    {
      "entropy": 0.7041789777576923,
      "epoch": 0.5799161591940242,
      "grad_norm": 0.17103150486946106,
      "learning_rate": 8.067286165547275e-05,
      "loss": 0.712,
      "mean_token_accuracy": 0.7885116450488567,
      "num_tokens": 201195338.0,
      "step": 6303
    },
    {
      "entropy": 0.6171780070289969,
      "epoch": 0.5800081655654654,
      "grad_norm": 0.15399011969566345,
      "learning_rate": 8.066979482933112e-05,
      "loss": 0.6063,
      "mean_token_accuracy": 0.8205934762954712,
      "num_tokens": 201227226.0,
      "step": 6304
    },
    {
      "entropy": 0.736755320802331,
      "epoch": 0.5801001719369067,
      "grad_norm": 0.15683552622795105,
      "learning_rate": 8.06667280031895e-05,
      "loss": 0.7223,
      "mean_token_accuracy": 0.7941903285682201,
      "num_tokens": 201259146.0,
      "step": 6305
    },
    {
      "entropy": 0.7841922156512737,
      "epoch": 0.5801921783083479,
      "grad_norm": 0.15986362099647522,
      "learning_rate": 8.066366117704788e-05,
      "loss": 0.7815,
      "mean_token_accuracy": 0.7699039094150066,
      "num_tokens": 201290568.0,
      "step": 6306
    },
    {
      "entropy": 0.7852820176631212,
      "epoch": 0.580284184679789,
      "grad_norm": 0.15699562430381775,
      "learning_rate": 8.066059435090625e-05,
      "loss": 0.7674,
      "mean_token_accuracy": 0.7744249701499939,
      "num_tokens": 201322438.0,
      "step": 6307
    },
    {
      "entropy": 0.7944781687110662,
      "epoch": 0.5803761910512303,
      "grad_norm": 0.15848669409751892,
      "learning_rate": 8.065752752476462e-05,
      "loss": 0.7781,
      "mean_token_accuracy": 0.773652445524931,
      "num_tokens": 201354728.0,
      "step": 6308
    },
    {
      "entropy": 0.8279907032847404,
      "epoch": 0.5804681974226715,
      "grad_norm": 0.1658608764410019,
      "learning_rate": 8.0654460698623e-05,
      "loss": 0.8279,
      "mean_token_accuracy": 0.7600633203983307,
      "num_tokens": 201386286.0,
      "step": 6309
    },
    {
      "entropy": 0.7419232502579689,
      "epoch": 0.5805602037941128,
      "grad_norm": 0.15257273614406586,
      "learning_rate": 8.065139387248137e-05,
      "loss": 0.7297,
      "mean_token_accuracy": 0.7802859880030155,
      "num_tokens": 201417108.0,
      "step": 6310
    },
    {
      "entropy": 0.741171196103096,
      "epoch": 0.580652210165554,
      "grad_norm": 0.16178913414478302,
      "learning_rate": 8.064832704633975e-05,
      "loss": 0.7339,
      "mean_token_accuracy": 0.7878806889057159,
      "num_tokens": 201448763.0,
      "step": 6311
    },
    {
      "entropy": 0.7838182188570499,
      "epoch": 0.5807442165369952,
      "grad_norm": 0.1563577502965927,
      "learning_rate": 8.064526022019812e-05,
      "loss": 0.7625,
      "mean_token_accuracy": 0.7746467292308807,
      "num_tokens": 201480744.0,
      "step": 6312
    },
    {
      "entropy": 0.7743762787431479,
      "epoch": 0.5808362229084364,
      "grad_norm": 0.16363857686519623,
      "learning_rate": 8.06421933940565e-05,
      "loss": 0.7558,
      "mean_token_accuracy": 0.776181772351265,
      "num_tokens": 201512336.0,
      "step": 6313
    },
    {
      "entropy": 0.75204012170434,
      "epoch": 0.5809282292798776,
      "grad_norm": 0.1579774022102356,
      "learning_rate": 8.063912656791487e-05,
      "loss": 0.7364,
      "mean_token_accuracy": 0.7798038721084595,
      "num_tokens": 201543821.0,
      "step": 6314
    },
    {
      "entropy": 0.7822910724207759,
      "epoch": 0.5810202356513189,
      "grad_norm": 0.16129764914512634,
      "learning_rate": 8.063605974177323e-05,
      "loss": 0.7811,
      "mean_token_accuracy": 0.7714543677866459,
      "num_tokens": 201576078.0,
      "step": 6315
    },
    {
      "entropy": 0.7639681920409203,
      "epoch": 0.5811122420227601,
      "grad_norm": 0.16104336082935333,
      "learning_rate": 8.063299291563163e-05,
      "loss": 0.7438,
      "mean_token_accuracy": 0.7819480486214161,
      "num_tokens": 201608221.0,
      "step": 6316
    },
    {
      "entropy": 0.7768698818981647,
      "epoch": 0.5812042483942013,
      "grad_norm": 0.15105968713760376,
      "learning_rate": 8.062992608949e-05,
      "loss": 0.7699,
      "mean_token_accuracy": 0.7744005955755711,
      "num_tokens": 201640983.0,
      "step": 6317
    },
    {
      "entropy": 0.7412724606692791,
      "epoch": 0.5812962547656425,
      "grad_norm": 0.15405476093292236,
      "learning_rate": 8.062685926334836e-05,
      "loss": 0.7455,
      "mean_token_accuracy": 0.7817461118102074,
      "num_tokens": 201672769.0,
      "step": 6318
    },
    {
      "entropy": 0.7500210348516703,
      "epoch": 0.5813882611370838,
      "grad_norm": 0.16205640137195587,
      "learning_rate": 8.062379243720673e-05,
      "loss": 0.7434,
      "mean_token_accuracy": 0.7774663865566254,
      "num_tokens": 201704955.0,
      "step": 6319
    },
    {
      "entropy": 0.8835061937570572,
      "epoch": 0.581480267508525,
      "grad_norm": 0.16959893703460693,
      "learning_rate": 8.062072561106511e-05,
      "loss": 0.8846,
      "mean_token_accuracy": 0.7440359517931938,
      "num_tokens": 201736497.0,
      "step": 6320
    },
    {
      "entropy": 0.7107734996825457,
      "epoch": 0.5815722738799662,
      "grad_norm": 0.15486393868923187,
      "learning_rate": 8.06176587849235e-05,
      "loss": 0.6949,
      "mean_token_accuracy": 0.7955043911933899,
      "num_tokens": 201768620.0,
      "step": 6321
    },
    {
      "entropy": 0.7030289443209767,
      "epoch": 0.5816642802514074,
      "grad_norm": 0.15442399680614471,
      "learning_rate": 8.061459195878186e-05,
      "loss": 0.6879,
      "mean_token_accuracy": 0.7976690046489239,
      "num_tokens": 201800624.0,
      "step": 6322
    },
    {
      "entropy": 0.8252401277422905,
      "epoch": 0.5817562866228486,
      "grad_norm": 0.16181346774101257,
      "learning_rate": 8.061152513264023e-05,
      "loss": 0.82,
      "mean_token_accuracy": 0.7604249753057957,
      "num_tokens": 201832246.0,
      "step": 6323
    },
    {
      "entropy": 0.7778275422751904,
      "epoch": 0.5818482929942899,
      "grad_norm": 0.15997865796089172,
      "learning_rate": 8.060845830649861e-05,
      "loss": 0.7541,
      "mean_token_accuracy": 0.774393979460001,
      "num_tokens": 201863832.0,
      "step": 6324
    },
    {
      "entropy": 0.7013303618878126,
      "epoch": 0.5819402993657311,
      "grad_norm": 0.15281438827514648,
      "learning_rate": 8.060539148035698e-05,
      "loss": 0.6868,
      "mean_token_accuracy": 0.7972174286842346,
      "num_tokens": 201895685.0,
      "step": 6325
    },
    {
      "entropy": 0.739263353869319,
      "epoch": 0.5820323057371724,
      "grad_norm": 0.1638054996728897,
      "learning_rate": 8.060232465421536e-05,
      "loss": 0.7272,
      "mean_token_accuracy": 0.7870798408985138,
      "num_tokens": 201927221.0,
      "step": 6326
    },
    {
      "entropy": 0.7923410907387733,
      "epoch": 0.5821243121086135,
      "grad_norm": 0.16328993439674377,
      "learning_rate": 8.059925782807373e-05,
      "loss": 0.8155,
      "mean_token_accuracy": 0.7636413872241974,
      "num_tokens": 201959405.0,
      "step": 6327
    },
    {
      "entropy": 0.8175417892634869,
      "epoch": 0.5822163184800547,
      "grad_norm": 0.16284319758415222,
      "learning_rate": 8.059619100193211e-05,
      "loss": 0.8141,
      "mean_token_accuracy": 0.7627188786864281,
      "num_tokens": 201991966.0,
      "step": 6328
    },
    {
      "entropy": 0.7379454039037228,
      "epoch": 0.582308324851496,
      "grad_norm": 0.15973059833049774,
      "learning_rate": 8.059312417579048e-05,
      "loss": 0.7115,
      "mean_token_accuracy": 0.7903411835432053,
      "num_tokens": 202024332.0,
      "step": 6329
    },
    {
      "entropy": 0.7785872016102076,
      "epoch": 0.5824003312229372,
      "grad_norm": 0.15696199238300323,
      "learning_rate": 8.059005734964885e-05,
      "loss": 0.7525,
      "mean_token_accuracy": 0.7813113331794739,
      "num_tokens": 202056723.0,
      "step": 6330
    },
    {
      "entropy": 0.7142103482037783,
      "epoch": 0.5824923375943785,
      "grad_norm": 0.1538364142179489,
      "learning_rate": 8.058699052350723e-05,
      "loss": 0.6892,
      "mean_token_accuracy": 0.7976019233465195,
      "num_tokens": 202088971.0,
      "step": 6331
    },
    {
      "entropy": 0.7651181500405073,
      "epoch": 0.5825843439658196,
      "grad_norm": 0.15773583948612213,
      "learning_rate": 8.058392369736561e-05,
      "loss": 0.7483,
      "mean_token_accuracy": 0.7767143212258816,
      "num_tokens": 202120503.0,
      "step": 6332
    },
    {
      "entropy": 0.6831620540469885,
      "epoch": 0.5826763503372608,
      "grad_norm": 0.15898539125919342,
      "learning_rate": 8.058085687122398e-05,
      "loss": 0.6591,
      "mean_token_accuracy": 0.8021279238164425,
      "num_tokens": 202151604.0,
      "step": 6333
    },
    {
      "entropy": 0.8299389090389013,
      "epoch": 0.5827683567087021,
      "grad_norm": 0.16599726676940918,
      "learning_rate": 8.057779004508234e-05,
      "loss": 0.8367,
      "mean_token_accuracy": 0.7528194151818752,
      "num_tokens": 202183846.0,
      "step": 6334
    },
    {
      "entropy": 0.6999797709286213,
      "epoch": 0.5828603630801433,
      "grad_norm": 0.149987131357193,
      "learning_rate": 8.057472321894071e-05,
      "loss": 0.6943,
      "mean_token_accuracy": 0.7946860827505589,
      "num_tokens": 202216368.0,
      "step": 6335
    },
    {
      "entropy": 0.8266239836812019,
      "epoch": 0.5829523694515846,
      "grad_norm": 0.1553816795349121,
      "learning_rate": 8.05716563927991e-05,
      "loss": 0.8279,
      "mean_token_accuracy": 0.7566586658358574,
      "num_tokens": 202248375.0,
      "step": 6336
    },
    {
      "entropy": 0.7897210288792849,
      "epoch": 0.5830443758230257,
      "grad_norm": 0.16443262994289398,
      "learning_rate": 8.056858956665748e-05,
      "loss": 0.7842,
      "mean_token_accuracy": 0.7707095891237259,
      "num_tokens": 202280347.0,
      "step": 6337
    },
    {
      "entropy": 0.8375037275254726,
      "epoch": 0.583136382194467,
      "grad_norm": 0.15919806063175201,
      "learning_rate": 8.056552274051584e-05,
      "loss": 0.8241,
      "mean_token_accuracy": 0.7626102454960346,
      "num_tokens": 202311045.0,
      "step": 6338
    },
    {
      "entropy": 0.870623491704464,
      "epoch": 0.5832283885659082,
      "grad_norm": 0.15767817199230194,
      "learning_rate": 8.056245591437421e-05,
      "loss": 0.8813,
      "mean_token_accuracy": 0.7450994327664375,
      "num_tokens": 202342726.0,
      "step": 6339
    },
    {
      "entropy": 0.6870536655187607,
      "epoch": 0.5833203949373494,
      "grad_norm": 0.15310555696487427,
      "learning_rate": 8.055938908823259e-05,
      "loss": 0.677,
      "mean_token_accuracy": 0.7991863377392292,
      "num_tokens": 202374133.0,
      "step": 6340
    },
    {
      "entropy": 0.6762062441557646,
      "epoch": 0.5834124013087907,
      "grad_norm": 0.15550066530704498,
      "learning_rate": 8.055632226209096e-05,
      "loss": 0.6548,
      "mean_token_accuracy": 0.8045507110655308,
      "num_tokens": 202405851.0,
      "step": 6341
    },
    {
      "entropy": 0.821690184995532,
      "epoch": 0.5835044076802318,
      "grad_norm": 0.16740159690380096,
      "learning_rate": 8.055325543594934e-05,
      "loss": 0.8168,
      "mean_token_accuracy": 0.7610343135893345,
      "num_tokens": 202438115.0,
      "step": 6342
    },
    {
      "entropy": 0.5365053927525878,
      "epoch": 0.5835964140516731,
      "grad_norm": 0.16305531561374664,
      "learning_rate": 8.055018860980771e-05,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8422117382287979,
      "num_tokens": 202469653.0,
      "step": 6343
    },
    {
      "entropy": 0.6908400990068913,
      "epoch": 0.5836884204231143,
      "grad_norm": 0.15588171780109406,
      "learning_rate": 8.054712178366609e-05,
      "loss": 0.6563,
      "mean_token_accuracy": 0.8058049716055393,
      "num_tokens": 202501655.0,
      "step": 6344
    },
    {
      "entropy": 0.8861616179347038,
      "epoch": 0.5837804267945556,
      "grad_norm": 0.16098451614379883,
      "learning_rate": 8.054405495752446e-05,
      "loss": 0.8687,
      "mean_token_accuracy": 0.7446994036436081,
      "num_tokens": 202533022.0,
      "step": 6345
    },
    {
      "entropy": 0.7419846169650555,
      "epoch": 0.5838724331659968,
      "grad_norm": 0.15696260333061218,
      "learning_rate": 8.054098813138283e-05,
      "loss": 0.7381,
      "mean_token_accuracy": 0.7814257070422173,
      "num_tokens": 202563975.0,
      "step": 6346
    },
    {
      "entropy": 0.8773860763758421,
      "epoch": 0.5839644395374379,
      "grad_norm": 0.1633908897638321,
      "learning_rate": 8.053792130524121e-05,
      "loss": 0.8779,
      "mean_token_accuracy": 0.7505197860300541,
      "num_tokens": 202595844.0,
      "step": 6347
    },
    {
      "entropy": 0.7388794701546431,
      "epoch": 0.5840564459088792,
      "grad_norm": 0.15466415882110596,
      "learning_rate": 8.053485447909959e-05,
      "loss": 0.7076,
      "mean_token_accuracy": 0.7901565581560135,
      "num_tokens": 202628075.0,
      "step": 6348
    },
    {
      "entropy": 0.7846502438187599,
      "epoch": 0.5841484522803204,
      "grad_norm": 0.16298574209213257,
      "learning_rate": 8.053178765295796e-05,
      "loss": 0.7814,
      "mean_token_accuracy": 0.7737542726099491,
      "num_tokens": 202660843.0,
      "step": 6349
    },
    {
      "entropy": 0.7384048812091351,
      "epoch": 0.5842404586517617,
      "grad_norm": 0.1500582993030548,
      "learning_rate": 8.052872082681633e-05,
      "loss": 0.7152,
      "mean_token_accuracy": 0.7902737110853195,
      "num_tokens": 202692509.0,
      "step": 6350
    },
    {
      "entropy": 0.7467363923788071,
      "epoch": 0.5843324650232029,
      "grad_norm": 0.15592637658119202,
      "learning_rate": 8.052565400067471e-05,
      "loss": 0.7412,
      "mean_token_accuracy": 0.7830436155200005,
      "num_tokens": 202724912.0,
      "step": 6351
    },
    {
      "entropy": 0.6637331219390035,
      "epoch": 0.584424471394644,
      "grad_norm": 0.15059129893779755,
      "learning_rate": 8.052258717453308e-05,
      "loss": 0.6447,
      "mean_token_accuracy": 0.8108443357050419,
      "num_tokens": 202756998.0,
      "step": 6352
    },
    {
      "entropy": 0.7915660347789526,
      "epoch": 0.5845164777660853,
      "grad_norm": 0.16551904380321503,
      "learning_rate": 8.051952034839146e-05,
      "loss": 0.7937,
      "mean_token_accuracy": 0.7708525434136391,
      "num_tokens": 202788381.0,
      "step": 6353
    },
    {
      "entropy": 0.7635294329375029,
      "epoch": 0.5846084841375265,
      "grad_norm": 0.15326827764511108,
      "learning_rate": 8.051645352224982e-05,
      "loss": 0.7573,
      "mean_token_accuracy": 0.7775497697293758,
      "num_tokens": 202820424.0,
      "step": 6354
    },
    {
      "entropy": 0.8262108024209738,
      "epoch": 0.5847004905089678,
      "grad_norm": 0.1609928160905838,
      "learning_rate": 8.05133866961082e-05,
      "loss": 0.8188,
      "mean_token_accuracy": 0.7620029002428055,
      "num_tokens": 202851722.0,
      "step": 6355
    },
    {
      "entropy": 0.7839486207813025,
      "epoch": 0.584792496880409,
      "grad_norm": 0.15905141830444336,
      "learning_rate": 8.051031986996657e-05,
      "loss": 0.7763,
      "mean_token_accuracy": 0.7778489217162132,
      "num_tokens": 202883854.0,
      "step": 6356
    },
    {
      "entropy": 0.6824868712574244,
      "epoch": 0.5848845032518502,
      "grad_norm": 0.15173740684986115,
      "learning_rate": 8.050725304382496e-05,
      "loss": 0.666,
      "mean_token_accuracy": 0.7980569675564766,
      "num_tokens": 202915908.0,
      "step": 6357
    },
    {
      "entropy": 0.8612017426639795,
      "epoch": 0.5849765096232914,
      "grad_norm": 0.1608838587999344,
      "learning_rate": 8.050418621768332e-05,
      "loss": 0.8431,
      "mean_token_accuracy": 0.7557514682412148,
      "num_tokens": 202947930.0,
      "step": 6358
    },
    {
      "entropy": 0.6405814066529274,
      "epoch": 0.5850685159947326,
      "grad_norm": 0.15502455830574036,
      "learning_rate": 8.05011193915417e-05,
      "loss": 0.6262,
      "mean_token_accuracy": 0.8079899214208126,
      "num_tokens": 202979140.0,
      "step": 6359
    },
    {
      "entropy": 0.7590779401361942,
      "epoch": 0.5851605223661739,
      "grad_norm": 0.152239590883255,
      "learning_rate": 8.049805256540007e-05,
      "loss": 0.7429,
      "mean_token_accuracy": 0.7800830751657486,
      "num_tokens": 203011158.0,
      "step": 6360
    },
    {
      "entropy": 0.6894998271018267,
      "epoch": 0.5852525287376151,
      "grad_norm": 0.15608155727386475,
      "learning_rate": 8.049498573925844e-05,
      "loss": 0.6609,
      "mean_token_accuracy": 0.8025123327970505,
      "num_tokens": 203043039.0,
      "step": 6361
    },
    {
      "entropy": 0.906234860420227,
      "epoch": 0.5853445351090563,
      "grad_norm": 0.16420932114124298,
      "learning_rate": 8.049191891311682e-05,
      "loss": 0.9233,
      "mean_token_accuracy": 0.7367675267159939,
      "num_tokens": 203075362.0,
      "step": 6362
    },
    {
      "entropy": 0.7392745111137629,
      "epoch": 0.5854365414804975,
      "grad_norm": 0.157680943608284,
      "learning_rate": 8.04888520869752e-05,
      "loss": 0.7351,
      "mean_token_accuracy": 0.7823094427585602,
      "num_tokens": 203107156.0,
      "step": 6363
    },
    {
      "entropy": 0.7138961981981993,
      "epoch": 0.5855285478519388,
      "grad_norm": 0.1503383368253708,
      "learning_rate": 8.048578526083357e-05,
      "loss": 0.7091,
      "mean_token_accuracy": 0.7900563105940819,
      "num_tokens": 203139344.0,
      "step": 6364
    },
    {
      "entropy": 0.7888676207512617,
      "epoch": 0.58562055422338,
      "grad_norm": 0.15872137248516083,
      "learning_rate": 8.048271843469194e-05,
      "loss": 0.7949,
      "mean_token_accuracy": 0.7704670503735542,
      "num_tokens": 203171639.0,
      "step": 6365
    },
    {
      "entropy": 0.7635250296443701,
      "epoch": 0.5857125605948212,
      "grad_norm": 0.16044297814369202,
      "learning_rate": 8.04796516085503e-05,
      "loss": 0.7683,
      "mean_token_accuracy": 0.778932448476553,
      "num_tokens": 203203126.0,
      "step": 6366
    },
    {
      "entropy": 0.7693639490753412,
      "epoch": 0.5858045669662624,
      "grad_norm": 0.15360097587108612,
      "learning_rate": 8.047658478240869e-05,
      "loss": 0.7575,
      "mean_token_accuracy": 0.7755149342119694,
      "num_tokens": 203234596.0,
      "step": 6367
    },
    {
      "entropy": 0.7743924763053656,
      "epoch": 0.5858965733377036,
      "grad_norm": 0.16602282226085663,
      "learning_rate": 8.047351795626707e-05,
      "loss": 0.7811,
      "mean_token_accuracy": 0.7722534723579884,
      "num_tokens": 203267139.0,
      "step": 6368
    },
    {
      "entropy": 0.6295870933681726,
      "epoch": 0.5859885797091449,
      "grad_norm": 0.1539800763130188,
      "learning_rate": 8.047045113012544e-05,
      "loss": 0.6113,
      "mean_token_accuracy": 0.8188748024404049,
      "num_tokens": 203298874.0,
      "step": 6369
    },
    {
      "entropy": 0.7165049742907286,
      "epoch": 0.5860805860805861,
      "grad_norm": 0.15197239816188812,
      "learning_rate": 8.04673843039838e-05,
      "loss": 0.696,
      "mean_token_accuracy": 0.7971337102353573,
      "num_tokens": 203330632.0,
      "step": 6370
    },
    {
      "entropy": 0.7625807486474514,
      "epoch": 0.5861725924520274,
      "grad_norm": 0.15311196446418762,
      "learning_rate": 8.046431747784219e-05,
      "loss": 0.751,
      "mean_token_accuracy": 0.7799668200314045,
      "num_tokens": 203362880.0,
      "step": 6371
    },
    {
      "entropy": 0.73433942720294,
      "epoch": 0.5862645988234685,
      "grad_norm": 0.1531653255224228,
      "learning_rate": 8.046125065170055e-05,
      "loss": 0.7208,
      "mean_token_accuracy": 0.7870123200118542,
      "num_tokens": 203394947.0,
      "step": 6372
    },
    {
      "entropy": 0.7444760743528605,
      "epoch": 0.5863566051949097,
      "grad_norm": 0.15218816697597504,
      "learning_rate": 8.045818382555894e-05,
      "loss": 0.7365,
      "mean_token_accuracy": 0.7847565077245235,
      "num_tokens": 203426347.0,
      "step": 6373
    },
    {
      "entropy": 0.8236728068441153,
      "epoch": 0.586448611566351,
      "grad_norm": 0.17060819268226624,
      "learning_rate": 8.04551169994173e-05,
      "loss": 0.8146,
      "mean_token_accuracy": 0.7641733810305595,
      "num_tokens": 203455925.0,
      "step": 6374
    },
    {
      "entropy": 0.7628369480371475,
      "epoch": 0.5865406179377922,
      "grad_norm": 0.1512911319732666,
      "learning_rate": 8.045205017327569e-05,
      "loss": 0.7561,
      "mean_token_accuracy": 0.7808767482638359,
      "num_tokens": 203488003.0,
      "step": 6375
    },
    {
      "entropy": 0.7747984081506729,
      "epoch": 0.5866326243092335,
      "grad_norm": 0.15370261669158936,
      "learning_rate": 8.044898334713405e-05,
      "loss": 0.7407,
      "mean_token_accuracy": 0.7767431363463402,
      "num_tokens": 203520120.0,
      "step": 6376
    },
    {
      "entropy": 0.7341177128255367,
      "epoch": 0.5867246306806746,
      "grad_norm": 0.15519611537456512,
      "learning_rate": 8.044591652099242e-05,
      "loss": 0.7191,
      "mean_token_accuracy": 0.7889431305229664,
      "num_tokens": 203551922.0,
      "step": 6377
    },
    {
      "entropy": 0.7919953390955925,
      "epoch": 0.5868166370521158,
      "grad_norm": 0.15688063204288483,
      "learning_rate": 8.04428496948508e-05,
      "loss": 0.7991,
      "mean_token_accuracy": 0.7708109579980373,
      "num_tokens": 203583810.0,
      "step": 6378
    },
    {
      "entropy": 0.7032122407108545,
      "epoch": 0.5869086434235571,
      "grad_norm": 0.1605205088853836,
      "learning_rate": 8.043978286870918e-05,
      "loss": 0.7089,
      "mean_token_accuracy": 0.7939116954803467,
      "num_tokens": 203616241.0,
      "step": 6379
    },
    {
      "entropy": 0.6745331063866615,
      "epoch": 0.5870006497949983,
      "grad_norm": 0.15584196150302887,
      "learning_rate": 8.043671604256755e-05,
      "loss": 0.6851,
      "mean_token_accuracy": 0.7991606406867504,
      "num_tokens": 203647915.0,
      "step": 6380
    },
    {
      "entropy": 0.7150711975991726,
      "epoch": 0.5870926561664396,
      "grad_norm": 0.15606476366519928,
      "learning_rate": 8.043364921642592e-05,
      "loss": 0.7021,
      "mean_token_accuracy": 0.7895516455173492,
      "num_tokens": 203679582.0,
      "step": 6381
    },
    {
      "entropy": 0.7812303490936756,
      "epoch": 0.5871846625378807,
      "grad_norm": 0.15351295471191406,
      "learning_rate": 8.04305823902843e-05,
      "loss": 0.7734,
      "mean_token_accuracy": 0.7748121693730354,
      "num_tokens": 203711120.0,
      "step": 6382
    },
    {
      "entropy": 0.805045835673809,
      "epoch": 0.587276668909322,
      "grad_norm": 0.160233274102211,
      "learning_rate": 8.042751556414267e-05,
      "loss": 0.7871,
      "mean_token_accuracy": 0.7678750343620777,
      "num_tokens": 203742326.0,
      "step": 6383
    },
    {
      "entropy": 0.7879580315202475,
      "epoch": 0.5873686752807632,
      "grad_norm": 0.17010660469532013,
      "learning_rate": 8.042444873800105e-05,
      "loss": 0.7866,
      "mean_token_accuracy": 0.7747322842478752,
      "num_tokens": 203772270.0,
      "step": 6384
    },
    {
      "entropy": 0.8195597697049379,
      "epoch": 0.5874606816522044,
      "grad_norm": 0.1639655977487564,
      "learning_rate": 8.042138191185942e-05,
      "loss": 0.7837,
      "mean_token_accuracy": 0.7649114318192005,
      "num_tokens": 203802967.0,
      "step": 6385
    },
    {
      "entropy": 0.616126636043191,
      "epoch": 0.5875526880236457,
      "grad_norm": 0.14149540662765503,
      "learning_rate": 8.04183150857178e-05,
      "loss": 0.5752,
      "mean_token_accuracy": 0.8278421461582184,
      "num_tokens": 203835056.0,
      "step": 6386
    },
    {
      "entropy": 0.8600650746375322,
      "epoch": 0.5876446943950868,
      "grad_norm": 0.1535901576280594,
      "learning_rate": 8.041524825957617e-05,
      "loss": 0.8407,
      "mean_token_accuracy": 0.7520343661308289,
      "num_tokens": 203866619.0,
      "step": 6387
    },
    {
      "entropy": 0.7358041685074568,
      "epoch": 0.5877367007665281,
      "grad_norm": 0.15822476148605347,
      "learning_rate": 8.041218143343454e-05,
      "loss": 0.7192,
      "mean_token_accuracy": 0.7874877154827118,
      "num_tokens": 203898869.0,
      "step": 6388
    },
    {
      "entropy": 0.751721566542983,
      "epoch": 0.5878287071379693,
      "grad_norm": 0.1526876986026764,
      "learning_rate": 8.040911460729292e-05,
      "loss": 0.7388,
      "mean_token_accuracy": 0.7812034040689468,
      "num_tokens": 203931318.0,
      "step": 6389
    },
    {
      "entropy": 0.7054018545895815,
      "epoch": 0.5879207135094106,
      "grad_norm": 0.15154127776622772,
      "learning_rate": 8.04060477811513e-05,
      "loss": 0.6885,
      "mean_token_accuracy": 0.800794929265976,
      "num_tokens": 203962858.0,
      "step": 6390
    },
    {
      "entropy": 0.7364398371428251,
      "epoch": 0.5880127198808518,
      "grad_norm": 0.17091438174247742,
      "learning_rate": 8.040298095500967e-05,
      "loss": 0.7437,
      "mean_token_accuracy": 0.7798398807644844,
      "num_tokens": 203994648.0,
      "step": 6391
    },
    {
      "entropy": 0.8139308709651232,
      "epoch": 0.5881047262522929,
      "grad_norm": 0.15972401201725006,
      "learning_rate": 8.039991412886803e-05,
      "loss": 0.8145,
      "mean_token_accuracy": 0.7606289051473141,
      "num_tokens": 204026699.0,
      "step": 6392
    },
    {
      "entropy": 0.8426349982619286,
      "epoch": 0.5881967326237342,
      "grad_norm": 0.16399815678596497,
      "learning_rate": 8.03968473027264e-05,
      "loss": 0.8324,
      "mean_token_accuracy": 0.7603425160050392,
      "num_tokens": 204058579.0,
      "step": 6393
    },
    {
      "entropy": 0.7175447307527065,
      "epoch": 0.5882887389951754,
      "grad_norm": 0.16015063226222992,
      "learning_rate": 8.03937804765848e-05,
      "loss": 0.7052,
      "mean_token_accuracy": 0.7927054539322853,
      "num_tokens": 204090406.0,
      "step": 6394
    },
    {
      "entropy": 0.7817765194922686,
      "epoch": 0.5883807453666167,
      "grad_norm": 0.15788307785987854,
      "learning_rate": 8.039071365044316e-05,
      "loss": 0.7702,
      "mean_token_accuracy": 0.7757765091955662,
      "num_tokens": 204122441.0,
      "step": 6395
    },
    {
      "entropy": 0.851882541552186,
      "epoch": 0.5884727517380579,
      "grad_norm": 0.15883177518844604,
      "learning_rate": 8.038764682430153e-05,
      "loss": 0.8374,
      "mean_token_accuracy": 0.756239227950573,
      "num_tokens": 204154165.0,
      "step": 6396
    },
    {
      "entropy": 0.7329238504171371,
      "epoch": 0.588564758109499,
      "grad_norm": 0.16448181867599487,
      "learning_rate": 8.03845799981599e-05,
      "loss": 0.7256,
      "mean_token_accuracy": 0.7839644998311996,
      "num_tokens": 204184623.0,
      "step": 6397
    },
    {
      "entropy": 0.7314279358834028,
      "epoch": 0.5886567644809403,
      "grad_norm": 0.15512707829475403,
      "learning_rate": 8.038151317201828e-05,
      "loss": 0.714,
      "mean_token_accuracy": 0.7890373915433884,
      "num_tokens": 204217005.0,
      "step": 6398
    },
    {
      "entropy": 0.7598677836358547,
      "epoch": 0.5887487708523815,
      "grad_norm": 0.15165774524211884,
      "learning_rate": 8.037844634587666e-05,
      "loss": 0.7425,
      "mean_token_accuracy": 0.7812540605664253,
      "num_tokens": 204249033.0,
      "step": 6399
    },
    {
      "entropy": 0.8364827074110508,
      "epoch": 0.5888407772238228,
      "grad_norm": 0.1580135077238083,
      "learning_rate": 8.037537951973503e-05,
      "loss": 0.8125,
      "mean_token_accuracy": 0.7661026492714882,
      "num_tokens": 204280875.0,
      "step": 6400
    },
    {
      "entropy": 0.8443365897983313,
      "epoch": 0.588932783595264,
      "grad_norm": 0.15570776164531708,
      "learning_rate": 8.03723126935934e-05,
      "loss": 0.8314,
      "mean_token_accuracy": 0.7542585767805576,
      "num_tokens": 204312409.0,
      "step": 6401
    },
    {
      "entropy": 0.7378081530332565,
      "epoch": 0.5890247899667052,
      "grad_norm": 0.1603211611509323,
      "learning_rate": 8.036924586745178e-05,
      "loss": 0.7391,
      "mean_token_accuracy": 0.7827631272375584,
      "num_tokens": 204344519.0,
      "step": 6402
    },
    {
      "entropy": 0.7074817214161158,
      "epoch": 0.5891167963381464,
      "grad_norm": 0.16061441600322723,
      "learning_rate": 8.036617904131015e-05,
      "loss": 0.6942,
      "mean_token_accuracy": 0.79604846611619,
      "num_tokens": 204376423.0,
      "step": 6403
    },
    {
      "entropy": 0.7288549989461899,
      "epoch": 0.5892088027095876,
      "grad_norm": 0.1525062918663025,
      "learning_rate": 8.036311221516853e-05,
      "loss": 0.7078,
      "mean_token_accuracy": 0.7958903312683105,
      "num_tokens": 204409191.0,
      "step": 6404
    },
    {
      "entropy": 0.7547742389142513,
      "epoch": 0.5893008090810289,
      "grad_norm": 0.15777218341827393,
      "learning_rate": 8.03600453890269e-05,
      "loss": 0.7452,
      "mean_token_accuracy": 0.7810758352279663,
      "num_tokens": 204441034.0,
      "step": 6405
    },
    {
      "entropy": 0.7528418619185686,
      "epoch": 0.5893928154524701,
      "grad_norm": 0.15047712624073029,
      "learning_rate": 8.035697856288528e-05,
      "loss": 0.7333,
      "mean_token_accuracy": 0.7875479832291603,
      "num_tokens": 204473488.0,
      "step": 6406
    },
    {
      "entropy": 0.8352436497807503,
      "epoch": 0.5894848218239113,
      "grad_norm": 0.16027213633060455,
      "learning_rate": 8.035391173674365e-05,
      "loss": 0.8371,
      "mean_token_accuracy": 0.7593034878373146,
      "num_tokens": 204504704.0,
      "step": 6407
    },
    {
      "entropy": 0.7725705746561289,
      "epoch": 0.5895768281953525,
      "grad_norm": 0.15587875247001648,
      "learning_rate": 8.035084491060201e-05,
      "loss": 0.7468,
      "mean_token_accuracy": 0.7819182649254799,
      "num_tokens": 204535844.0,
      "step": 6408
    },
    {
      "entropy": 0.7913107015192509,
      "epoch": 0.5896688345667938,
      "grad_norm": 0.16046059131622314,
      "learning_rate": 8.03477780844604e-05,
      "loss": 0.7832,
      "mean_token_accuracy": 0.770763672888279,
      "num_tokens": 204567804.0,
      "step": 6409
    },
    {
      "entropy": 0.6442548707127571,
      "epoch": 0.589760840938235,
      "grad_norm": 0.1459970772266388,
      "learning_rate": 8.034471125831878e-05,
      "loss": 0.6187,
      "mean_token_accuracy": 0.8110596686601639,
      "num_tokens": 204598918.0,
      "step": 6410
    },
    {
      "entropy": 0.7625538688153028,
      "epoch": 0.5898528473096762,
      "grad_norm": 0.16120202839374542,
      "learning_rate": 8.034164443217715e-05,
      "loss": 0.77,
      "mean_token_accuracy": 0.7714480310678482,
      "num_tokens": 204630773.0,
      "step": 6411
    },
    {
      "entropy": 0.707021052017808,
      "epoch": 0.5899448536811174,
      "grad_norm": 0.16600696742534637,
      "learning_rate": 8.033857760603551e-05,
      "loss": 0.7272,
      "mean_token_accuracy": 0.789479672908783,
      "num_tokens": 204662153.0,
      "step": 6412
    },
    {
      "entropy": 0.76875707693398,
      "epoch": 0.5900368600525586,
      "grad_norm": 0.16380411386489868,
      "learning_rate": 8.03355107798939e-05,
      "loss": 0.7768,
      "mean_token_accuracy": 0.7704000174999237,
      "num_tokens": 204693634.0,
      "step": 6413
    },
    {
      "entropy": 0.7384569495916367,
      "epoch": 0.5901288664239999,
      "grad_norm": 0.1616789549589157,
      "learning_rate": 8.033244395375226e-05,
      "loss": 0.732,
      "mean_token_accuracy": 0.7846400924026966,
      "num_tokens": 204725004.0,
      "step": 6414
    },
    {
      "entropy": 0.7130808867514133,
      "epoch": 0.5902208727954411,
      "grad_norm": 0.15990842878818512,
      "learning_rate": 8.032937712761064e-05,
      "loss": 0.6788,
      "mean_token_accuracy": 0.8032352514564991,
      "num_tokens": 204756752.0,
      "step": 6415
    },
    {
      "entropy": 0.8804079107940197,
      "epoch": 0.5903128791668824,
      "grad_norm": 0.17252370715141296,
      "learning_rate": 8.032631030146901e-05,
      "loss": 0.8776,
      "mean_token_accuracy": 0.74616714194417,
      "num_tokens": 204787313.0,
      "step": 6416
    },
    {
      "entropy": 0.7448329012840986,
      "epoch": 0.5904048855383235,
      "grad_norm": 0.15691223740577698,
      "learning_rate": 8.03232434753274e-05,
      "loss": 0.7405,
      "mean_token_accuracy": 0.7834462858736515,
      "num_tokens": 204818761.0,
      "step": 6417
    },
    {
      "entropy": 0.9225790128111839,
      "epoch": 0.5904968919097647,
      "grad_norm": 0.16171899437904358,
      "learning_rate": 8.032017664918576e-05,
      "loss": 0.92,
      "mean_token_accuracy": 0.7306326553225517,
      "num_tokens": 204850978.0,
      "step": 6418
    },
    {
      "entropy": 0.8325988091528416,
      "epoch": 0.590588898281206,
      "grad_norm": 0.15790510177612305,
      "learning_rate": 8.031710982304413e-05,
      "loss": 0.8015,
      "mean_token_accuracy": 0.7700189985334873,
      "num_tokens": 204883335.0,
      "step": 6419
    },
    {
      "entropy": 0.7138184383511543,
      "epoch": 0.5906809046526472,
      "grad_norm": 0.15031220018863678,
      "learning_rate": 8.031404299690251e-05,
      "loss": 0.676,
      "mean_token_accuracy": 0.7991909235715866,
      "num_tokens": 204915326.0,
      "step": 6420
    },
    {
      "entropy": 0.8201861176639795,
      "epoch": 0.5907729110240885,
      "grad_norm": 0.15882647037506104,
      "learning_rate": 8.031097617076089e-05,
      "loss": 0.8026,
      "mean_token_accuracy": 0.7660219520330429,
      "num_tokens": 204947177.0,
      "step": 6421
    },
    {
      "entropy": 0.8233970105648041,
      "epoch": 0.5908649173955296,
      "grad_norm": 0.1553497165441513,
      "learning_rate": 8.030790934461926e-05,
      "loss": 0.8012,
      "mean_token_accuracy": 0.7682563848793507,
      "num_tokens": 204979094.0,
      "step": 6422
    },
    {
      "entropy": 0.7394640073180199,
      "epoch": 0.5909569237669708,
      "grad_norm": 0.16526657342910767,
      "learning_rate": 8.030484251847763e-05,
      "loss": 0.7293,
      "mean_token_accuracy": 0.7829725369811058,
      "num_tokens": 205010805.0,
      "step": 6423
    },
    {
      "entropy": 0.7503649201244116,
      "epoch": 0.5910489301384121,
      "grad_norm": 0.15507331490516663,
      "learning_rate": 8.0301775692336e-05,
      "loss": 0.7426,
      "mean_token_accuracy": 0.7835765592753887,
      "num_tokens": 205043363.0,
      "step": 6424
    },
    {
      "entropy": 0.7980811186134815,
      "epoch": 0.5911409365098533,
      "grad_norm": 0.15940827131271362,
      "learning_rate": 8.029870886619438e-05,
      "loss": 0.7896,
      "mean_token_accuracy": 0.7723812349140644,
      "num_tokens": 205075630.0,
      "step": 6425
    },
    {
      "entropy": 0.7251358404755592,
      "epoch": 0.5912329428812946,
      "grad_norm": 0.15874603390693665,
      "learning_rate": 8.029564204005276e-05,
      "loss": 0.7313,
      "mean_token_accuracy": 0.786423496901989,
      "num_tokens": 205108141.0,
      "step": 6426
    },
    {
      "entropy": 0.785711606964469,
      "epoch": 0.5913249492527357,
      "grad_norm": 0.16150999069213867,
      "learning_rate": 8.029257521391113e-05,
      "loss": 0.7667,
      "mean_token_accuracy": 0.7704710848629475,
      "num_tokens": 205139846.0,
      "step": 6427
    },
    {
      "entropy": 0.7223213911056519,
      "epoch": 0.591416955624177,
      "grad_norm": 0.15136899054050446,
      "learning_rate": 8.02895083877695e-05,
      "loss": 0.7154,
      "mean_token_accuracy": 0.7912070713937283,
      "num_tokens": 205172427.0,
      "step": 6428
    },
    {
      "entropy": 0.6839336678385735,
      "epoch": 0.5915089619956182,
      "grad_norm": 0.15855400264263153,
      "learning_rate": 8.028644156162788e-05,
      "loss": 0.6764,
      "mean_token_accuracy": 0.8026933707296848,
      "num_tokens": 205203920.0,
      "step": 6429
    },
    {
      "entropy": 0.7386108711361885,
      "epoch": 0.5916009683670594,
      "grad_norm": 0.151234433054924,
      "learning_rate": 8.028337473548624e-05,
      "loss": 0.7234,
      "mean_token_accuracy": 0.7834854312241077,
      "num_tokens": 205236161.0,
      "step": 6430
    },
    {
      "entropy": 0.8460969924926758,
      "epoch": 0.5916929747385007,
      "grad_norm": 0.15446896851062775,
      "learning_rate": 8.028030790934462e-05,
      "loss": 0.8504,
      "mean_token_accuracy": 0.7510744631290436,
      "num_tokens": 205268114.0,
      "step": 6431
    },
    {
      "entropy": 0.7104700393974781,
      "epoch": 0.5917849811099418,
      "grad_norm": 0.15183159708976746,
      "learning_rate": 8.027724108320299e-05,
      "loss": 0.7073,
      "mean_token_accuracy": 0.7895939387381077,
      "num_tokens": 205300079.0,
      "step": 6432
    },
    {
      "entropy": 0.8104813415557146,
      "epoch": 0.5918769874813831,
      "grad_norm": 0.153885617852211,
      "learning_rate": 8.027417425706137e-05,
      "loss": 0.8108,
      "mean_token_accuracy": 0.7570861168205738,
      "num_tokens": 205332536.0,
      "step": 6433
    },
    {
      "entropy": 0.8034578487277031,
      "epoch": 0.5919689938528243,
      "grad_norm": 0.1600767970085144,
      "learning_rate": 8.027110743091974e-05,
      "loss": 0.7995,
      "mean_token_accuracy": 0.7706472724676132,
      "num_tokens": 205364933.0,
      "step": 6434
    },
    {
      "entropy": 0.7732834294438362,
      "epoch": 0.5920610002242656,
      "grad_norm": 0.16329196095466614,
      "learning_rate": 8.026804060477811e-05,
      "loss": 0.773,
      "mean_token_accuracy": 0.7728634625673294,
      "num_tokens": 205397103.0,
      "step": 6435
    },
    {
      "entropy": 0.732304286211729,
      "epoch": 0.5921530065957068,
      "grad_norm": 0.16454380750656128,
      "learning_rate": 8.026497377863649e-05,
      "loss": 0.7434,
      "mean_token_accuracy": 0.7843661867082119,
      "num_tokens": 205429394.0,
      "step": 6436
    },
    {
      "entropy": 0.7339537870138884,
      "epoch": 0.5922450129671479,
      "grad_norm": 0.1708439290523529,
      "learning_rate": 8.026190695249487e-05,
      "loss": 0.7542,
      "mean_token_accuracy": 0.7789102494716644,
      "num_tokens": 205461460.0,
      "step": 6437
    },
    {
      "entropy": 0.7894754931330681,
      "epoch": 0.5923370193385892,
      "grad_norm": 0.1593955159187317,
      "learning_rate": 8.025884012635324e-05,
      "loss": 0.7707,
      "mean_token_accuracy": 0.7719837538897991,
      "num_tokens": 205493465.0,
      "step": 6438
    },
    {
      "entropy": 0.8040923997759819,
      "epoch": 0.5924290257100304,
      "grad_norm": 0.15896755456924438,
      "learning_rate": 8.025577330021161e-05,
      "loss": 0.766,
      "mean_token_accuracy": 0.7723164483904839,
      "num_tokens": 205524916.0,
      "step": 6439
    },
    {
      "entropy": 0.7599755423143506,
      "epoch": 0.5925210320814717,
      "grad_norm": 0.15846295654773712,
      "learning_rate": 8.025270647406999e-05,
      "loss": 0.7478,
      "mean_token_accuracy": 0.77960454672575,
      "num_tokens": 205557262.0,
      "step": 6440
    },
    {
      "entropy": 0.839160792529583,
      "epoch": 0.5926130384529129,
      "grad_norm": 0.15881659090518951,
      "learning_rate": 8.024963964792837e-05,
      "loss": 0.828,
      "mean_token_accuracy": 0.7554154396057129,
      "num_tokens": 205589931.0,
      "step": 6441
    },
    {
      "entropy": 0.678018232807517,
      "epoch": 0.592705044824354,
      "grad_norm": 0.15597133338451385,
      "learning_rate": 8.024657282178674e-05,
      "loss": 0.6728,
      "mean_token_accuracy": 0.8003105893731117,
      "num_tokens": 205622047.0,
      "step": 6442
    },
    {
      "entropy": 0.733624529093504,
      "epoch": 0.5927970511957953,
      "grad_norm": 0.16245269775390625,
      "learning_rate": 8.024350599564511e-05,
      "loss": 0.7275,
      "mean_token_accuracy": 0.786513913422823,
      "num_tokens": 205653248.0,
      "step": 6443
    },
    {
      "entropy": 0.6857230989262462,
      "epoch": 0.5928890575672365,
      "grad_norm": 0.1549699604511261,
      "learning_rate": 8.024043916950349e-05,
      "loss": 0.6716,
      "mean_token_accuracy": 0.8069267272949219,
      "num_tokens": 205684649.0,
      "step": 6444
    },
    {
      "entropy": 0.8541279435157776,
      "epoch": 0.5929810639386778,
      "grad_norm": 0.1590873897075653,
      "learning_rate": 8.023737234336186e-05,
      "loss": 0.8534,
      "mean_token_accuracy": 0.7507393732666969,
      "num_tokens": 205716443.0,
      "step": 6445
    },
    {
      "entropy": 0.6891656089574099,
      "epoch": 0.593073070310119,
      "grad_norm": 0.14362852275371552,
      "learning_rate": 8.023430551722024e-05,
      "loss": 0.6619,
      "mean_token_accuracy": 0.8026380836963654,
      "num_tokens": 205749086.0,
      "step": 6446
    },
    {
      "entropy": 0.7688812855631113,
      "epoch": 0.5931650766815602,
      "grad_norm": 0.15627062320709229,
      "learning_rate": 8.02312386910786e-05,
      "loss": 0.7526,
      "mean_token_accuracy": 0.7812754325568676,
      "num_tokens": 205781433.0,
      "step": 6447
    },
    {
      "entropy": 0.8092678859829903,
      "epoch": 0.5932570830530014,
      "grad_norm": 0.1584494560956955,
      "learning_rate": 8.022817186493699e-05,
      "loss": 0.7822,
      "mean_token_accuracy": 0.7658618651330471,
      "num_tokens": 205813233.0,
      "step": 6448
    },
    {
      "entropy": 0.7793342731893063,
      "epoch": 0.5933490894244426,
      "grad_norm": 0.15603691339492798,
      "learning_rate": 8.022510503879535e-05,
      "loss": 0.7675,
      "mean_token_accuracy": 0.7766341306269169,
      "num_tokens": 205845052.0,
      "step": 6449
    },
    {
      "entropy": 0.7485304698348045,
      "epoch": 0.5934410957958839,
      "grad_norm": 0.1587413251399994,
      "learning_rate": 8.022203821265372e-05,
      "loss": 0.737,
      "mean_token_accuracy": 0.7876352146267891,
      "num_tokens": 205876690.0,
      "step": 6450
    },
    {
      "entropy": 0.7446821685880423,
      "epoch": 0.5935331021673251,
      "grad_norm": 0.1588190197944641,
      "learning_rate": 8.02189713865121e-05,
      "loss": 0.7442,
      "mean_token_accuracy": 0.781039547175169,
      "num_tokens": 205908441.0,
      "step": 6451
    },
    {
      "entropy": 0.8378266543149948,
      "epoch": 0.5936251085387663,
      "grad_norm": 0.16117030382156372,
      "learning_rate": 8.021590456037049e-05,
      "loss": 0.8255,
      "mean_token_accuracy": 0.7643117606639862,
      "num_tokens": 205940469.0,
      "step": 6452
    },
    {
      "entropy": 0.7739326301962137,
      "epoch": 0.5937171149102075,
      "grad_norm": 0.15677396953105927,
      "learning_rate": 8.021283773422885e-05,
      "loss": 0.7683,
      "mean_token_accuracy": 0.7732589729130268,
      "num_tokens": 205972237.0,
      "step": 6453
    },
    {
      "entropy": 0.8071573786437511,
      "epoch": 0.5938091212816488,
      "grad_norm": 0.16322170197963715,
      "learning_rate": 8.020977090808722e-05,
      "loss": 0.7968,
      "mean_token_accuracy": 0.7742795012891293,
      "num_tokens": 206004097.0,
      "step": 6454
    },
    {
      "entropy": 0.8033352755010128,
      "epoch": 0.59390112765309,
      "grad_norm": 0.16521520912647247,
      "learning_rate": 8.020670408194559e-05,
      "loss": 0.8015,
      "mean_token_accuracy": 0.7749248519539833,
      "num_tokens": 206036613.0,
      "step": 6455
    },
    {
      "entropy": 0.7921232655644417,
      "epoch": 0.5939931340245312,
      "grad_norm": 0.1650679111480713,
      "learning_rate": 8.020363725580397e-05,
      "loss": 0.8015,
      "mean_token_accuracy": 0.7640129625797272,
      "num_tokens": 206068601.0,
      "step": 6456
    },
    {
      "entropy": 0.8793608751147985,
      "epoch": 0.5940851403959724,
      "grad_norm": 0.16414442658424377,
      "learning_rate": 8.020057042966235e-05,
      "loss": 0.8627,
      "mean_token_accuracy": 0.7463245056569576,
      "num_tokens": 206100055.0,
      "step": 6457
    },
    {
      "entropy": 0.8282900582998991,
      "epoch": 0.5941771467674136,
      "grad_norm": 0.17184162139892578,
      "learning_rate": 8.019750360352072e-05,
      "loss": 0.8254,
      "mean_token_accuracy": 0.7634611949324608,
      "num_tokens": 206131384.0,
      "step": 6458
    },
    {
      "entropy": 0.8741166777908802,
      "epoch": 0.5942691531388549,
      "grad_norm": 0.16681066155433655,
      "learning_rate": 8.019443677737909e-05,
      "loss": 0.87,
      "mean_token_accuracy": 0.7491681054234505,
      "num_tokens": 206163259.0,
      "step": 6459
    },
    {
      "entropy": 0.686090137809515,
      "epoch": 0.5943611595102961,
      "grad_norm": 0.1585303544998169,
      "learning_rate": 8.019136995123747e-05,
      "loss": 0.6635,
      "mean_token_accuracy": 0.8042276427149773,
      "num_tokens": 206195320.0,
      "step": 6460
    },
    {
      "entropy": 0.7561408504843712,
      "epoch": 0.5944531658817374,
      "grad_norm": 0.15510870516300201,
      "learning_rate": 8.018830312509584e-05,
      "loss": 0.753,
      "mean_token_accuracy": 0.7756185680627823,
      "num_tokens": 206227493.0,
      "step": 6461
    },
    {
      "entropy": 0.7573105357587337,
      "epoch": 0.5945451722531785,
      "grad_norm": 0.1507454365491867,
      "learning_rate": 8.018523629895422e-05,
      "loss": 0.7413,
      "mean_token_accuracy": 0.7811438180506229,
      "num_tokens": 206259384.0,
      "step": 6462
    },
    {
      "entropy": 0.7145005390048027,
      "epoch": 0.5946371786246197,
      "grad_norm": 0.14745336771011353,
      "learning_rate": 8.018216947281259e-05,
      "loss": 0.7049,
      "mean_token_accuracy": 0.7950799763202667,
      "num_tokens": 206291983.0,
      "step": 6463
    },
    {
      "entropy": 0.7815783564001322,
      "epoch": 0.594729184996061,
      "grad_norm": 0.15834683179855347,
      "learning_rate": 8.017910264667097e-05,
      "loss": 0.7722,
      "mean_token_accuracy": 0.7729985006153584,
      "num_tokens": 206324737.0,
      "step": 6464
    },
    {
      "entropy": 0.7953859902918339,
      "epoch": 0.5948211913675022,
      "grad_norm": 0.15077731013298035,
      "learning_rate": 8.017603582052934e-05,
      "loss": 0.7954,
      "mean_token_accuracy": 0.7683024927973747,
      "num_tokens": 206356706.0,
      "step": 6465
    },
    {
      "entropy": 0.7177160102874041,
      "epoch": 0.5949131977389435,
      "grad_norm": 0.15009845793247223,
      "learning_rate": 8.01729689943877e-05,
      "loss": 0.7053,
      "mean_token_accuracy": 0.7938613221049309,
      "num_tokens": 206389213.0,
      "step": 6466
    },
    {
      "entropy": 0.7058996446430683,
      "epoch": 0.5950052041103846,
      "grad_norm": 0.15857446193695068,
      "learning_rate": 8.016990216824608e-05,
      "loss": 0.7066,
      "mean_token_accuracy": 0.789520837366581,
      "num_tokens": 206420901.0,
      "step": 6467
    },
    {
      "entropy": 0.8687541410326958,
      "epoch": 0.5950972104818258,
      "grad_norm": 0.15851044654846191,
      "learning_rate": 8.016683534210447e-05,
      "loss": 0.8809,
      "mean_token_accuracy": 0.7446808740496635,
      "num_tokens": 206453372.0,
      "step": 6468
    },
    {
      "entropy": 0.9248711578547955,
      "epoch": 0.5951892168532671,
      "grad_norm": 0.16492363810539246,
      "learning_rate": 8.016376851596283e-05,
      "loss": 0.923,
      "mean_token_accuracy": 0.7329346016049385,
      "num_tokens": 206484752.0,
      "step": 6469
    },
    {
      "entropy": 0.7729863626882434,
      "epoch": 0.5952812232247083,
      "grad_norm": 0.16324837505817413,
      "learning_rate": 8.01607016898212e-05,
      "loss": 0.7698,
      "mean_token_accuracy": 0.7768435329198837,
      "num_tokens": 206516937.0,
      "step": 6470
    },
    {
      "entropy": 0.6484729796648026,
      "epoch": 0.5953732295961496,
      "grad_norm": 0.14830641448497772,
      "learning_rate": 8.015763486367958e-05,
      "loss": 0.6434,
      "mean_token_accuracy": 0.8041426092386246,
      "num_tokens": 206548864.0,
      "step": 6471
    },
    {
      "entropy": 0.7331998459994793,
      "epoch": 0.5954652359675907,
      "grad_norm": 0.15013593435287476,
      "learning_rate": 8.015456803753795e-05,
      "loss": 0.7093,
      "mean_token_accuracy": 0.7920070216059685,
      "num_tokens": 206581160.0,
      "step": 6472
    },
    {
      "entropy": 0.6928688902407885,
      "epoch": 0.595557242339032,
      "grad_norm": 0.15729312598705292,
      "learning_rate": 8.015150121139633e-05,
      "loss": 0.6747,
      "mean_token_accuracy": 0.7987152449786663,
      "num_tokens": 206613049.0,
      "step": 6473
    },
    {
      "entropy": 0.7294027004390955,
      "epoch": 0.5956492487104732,
      "grad_norm": 0.15603138506412506,
      "learning_rate": 8.01484343852547e-05,
      "loss": 0.7214,
      "mean_token_accuracy": 0.7897666804492474,
      "num_tokens": 206644910.0,
      "step": 6474
    },
    {
      "entropy": 0.81579883210361,
      "epoch": 0.5957412550819144,
      "grad_norm": 0.1530342400074005,
      "learning_rate": 8.014536755911308e-05,
      "loss": 0.8161,
      "mean_token_accuracy": 0.7641807049512863,
      "num_tokens": 206677405.0,
      "step": 6475
    },
    {
      "entropy": 0.67379380017519,
      "epoch": 0.5958332614533557,
      "grad_norm": 0.15065889060497284,
      "learning_rate": 8.014230073297145e-05,
      "loss": 0.6581,
      "mean_token_accuracy": 0.8074614256620407,
      "num_tokens": 206709582.0,
      "step": 6476
    },
    {
      "entropy": 0.7350539704784751,
      "epoch": 0.5959252678247968,
      "grad_norm": 0.1592596471309662,
      "learning_rate": 8.013923390682983e-05,
      "loss": 0.7092,
      "mean_token_accuracy": 0.7894832864403725,
      "num_tokens": 206742218.0,
      "step": 6477
    },
    {
      "entropy": 0.8554549403488636,
      "epoch": 0.5960172741962381,
      "grad_norm": 0.16293002665042877,
      "learning_rate": 8.01361670806882e-05,
      "loss": 0.8094,
      "mean_token_accuracy": 0.7614176720380783,
      "num_tokens": 206773530.0,
      "step": 6478
    },
    {
      "entropy": 0.6945176515728235,
      "epoch": 0.5961092805676793,
      "grad_norm": 0.14399118721485138,
      "learning_rate": 8.013310025454658e-05,
      "loss": 0.6654,
      "mean_token_accuracy": 0.8009990118443966,
      "num_tokens": 206806093.0,
      "step": 6479
    },
    {
      "entropy": 0.7480811551213264,
      "epoch": 0.5962012869391206,
      "grad_norm": 0.16127833724021912,
      "learning_rate": 8.013003342840495e-05,
      "loss": 0.7282,
      "mean_token_accuracy": 0.7843916937708855,
      "num_tokens": 206837586.0,
      "step": 6480
    },
    {
      "entropy": 0.692300334572792,
      "epoch": 0.5962932933105618,
      "grad_norm": 0.1515669971704483,
      "learning_rate": 8.012696660226332e-05,
      "loss": 0.6842,
      "mean_token_accuracy": 0.7999985851347446,
      "num_tokens": 206869223.0,
      "step": 6481
    },
    {
      "entropy": 0.8406537659466267,
      "epoch": 0.5963852996820029,
      "grad_norm": 0.16202636063098907,
      "learning_rate": 8.01238997761217e-05,
      "loss": 0.8415,
      "mean_token_accuracy": 0.7549130581319332,
      "num_tokens": 206900981.0,
      "step": 6482
    },
    {
      "entropy": 0.7394605241715908,
      "epoch": 0.5964773060534442,
      "grad_norm": 0.1596049666404724,
      "learning_rate": 8.012083294998008e-05,
      "loss": 0.741,
      "mean_token_accuracy": 0.7807878479361534,
      "num_tokens": 206933212.0,
      "step": 6483
    },
    {
      "entropy": 0.7352135572582483,
      "epoch": 0.5965693124248854,
      "grad_norm": 0.15973606705665588,
      "learning_rate": 8.011776612383845e-05,
      "loss": 0.7356,
      "mean_token_accuracy": 0.7828998379409313,
      "num_tokens": 206964443.0,
      "step": 6484
    },
    {
      "entropy": 0.7526090424507856,
      "epoch": 0.5966613187963267,
      "grad_norm": 0.15842664241790771,
      "learning_rate": 8.011469929769682e-05,
      "loss": 0.7556,
      "mean_token_accuracy": 0.7813078053295612,
      "num_tokens": 206996737.0,
      "step": 6485
    },
    {
      "entropy": 0.8538091517984867,
      "epoch": 0.5967533251677679,
      "grad_norm": 0.16166779398918152,
      "learning_rate": 8.011163247155518e-05,
      "loss": 0.8605,
      "mean_token_accuracy": 0.7512096725404263,
      "num_tokens": 207028772.0,
      "step": 6486
    },
    {
      "entropy": 0.734092578291893,
      "epoch": 0.596845331539209,
      "grad_norm": 0.1567302644252777,
      "learning_rate": 8.010856564541356e-05,
      "loss": 0.7246,
      "mean_token_accuracy": 0.7872543409466743,
      "num_tokens": 207061257.0,
      "step": 6487
    },
    {
      "entropy": 0.765211395919323,
      "epoch": 0.5969373379106503,
      "grad_norm": 0.15138205885887146,
      "learning_rate": 8.010549881927195e-05,
      "loss": 0.7577,
      "mean_token_accuracy": 0.7784016281366348,
      "num_tokens": 207092957.0,
      "step": 6488
    },
    {
      "entropy": 0.6274166405200958,
      "epoch": 0.5970293442820915,
      "grad_norm": 0.15672339498996735,
      "learning_rate": 8.010243199313031e-05,
      "loss": 0.6045,
      "mean_token_accuracy": 0.8157878443598747,
      "num_tokens": 207124790.0,
      "step": 6489
    },
    {
      "entropy": 0.8043479789048433,
      "epoch": 0.5971213506535328,
      "grad_norm": 0.15982453525066376,
      "learning_rate": 8.009936516698868e-05,
      "loss": 0.7806,
      "mean_token_accuracy": 0.7697338312864304,
      "num_tokens": 207156186.0,
      "step": 6490
    },
    {
      "entropy": 0.795964571647346,
      "epoch": 0.597213357024974,
      "grad_norm": 0.15761825442314148,
      "learning_rate": 8.009629834084706e-05,
      "loss": 0.776,
      "mean_token_accuracy": 0.7722980082035065,
      "num_tokens": 207188160.0,
      "step": 6491
    },
    {
      "entropy": 0.8558950275182724,
      "epoch": 0.5973053633964152,
      "grad_norm": 0.15556150674819946,
      "learning_rate": 8.009323151470543e-05,
      "loss": 0.8209,
      "mean_token_accuracy": 0.7613931968808174,
      "num_tokens": 207219381.0,
      "step": 6492
    },
    {
      "entropy": 0.7691142968833447,
      "epoch": 0.5973973697678564,
      "grad_norm": 0.1532619148492813,
      "learning_rate": 8.009016468856381e-05,
      "loss": 0.7539,
      "mean_token_accuracy": 0.772418238222599,
      "num_tokens": 207250022.0,
      "step": 6493
    },
    {
      "entropy": 0.7726809997111559,
      "epoch": 0.5974893761392976,
      "grad_norm": 0.16363465785980225,
      "learning_rate": 8.008709786242218e-05,
      "loss": 0.7661,
      "mean_token_accuracy": 0.7791938371956348,
      "num_tokens": 207282525.0,
      "step": 6494
    },
    {
      "entropy": 0.8841302022337914,
      "epoch": 0.5975813825107389,
      "grad_norm": 0.17669934034347534,
      "learning_rate": 8.008403103628056e-05,
      "loss": 0.8985,
      "mean_token_accuracy": 0.7462850511074066,
      "num_tokens": 207314665.0,
      "step": 6495
    },
    {
      "entropy": 0.7799374274909496,
      "epoch": 0.5976733888821801,
      "grad_norm": 0.16811153292655945,
      "learning_rate": 8.008096421013893e-05,
      "loss": 0.7604,
      "mean_token_accuracy": 0.7803197018802166,
      "num_tokens": 207346254.0,
      "step": 6496
    },
    {
      "entropy": 0.8111791536211967,
      "epoch": 0.5977653952536213,
      "grad_norm": 0.1616886854171753,
      "learning_rate": 8.00778973839973e-05,
      "loss": 0.8052,
      "mean_token_accuracy": 0.7667524069547653,
      "num_tokens": 207378884.0,
      "step": 6497
    },
    {
      "entropy": 0.8498926069587469,
      "epoch": 0.5978574016250625,
      "grad_norm": 0.15712971985340118,
      "learning_rate": 8.007483055785568e-05,
      "loss": 0.8353,
      "mean_token_accuracy": 0.7483085542917252,
      "num_tokens": 207410392.0,
      "step": 6498
    },
    {
      "entropy": 0.7078499291092157,
      "epoch": 0.5979494079965038,
      "grad_norm": 0.15936043858528137,
      "learning_rate": 8.007176373171406e-05,
      "loss": 0.695,
      "mean_token_accuracy": 0.7946310080587864,
      "num_tokens": 207441297.0,
      "step": 6499
    },
    {
      "entropy": 0.7039105612784624,
      "epoch": 0.598041414367945,
      "grad_norm": 0.15257161855697632,
      "learning_rate": 8.006869690557243e-05,
      "loss": 0.693,
      "mean_token_accuracy": 0.7945197708904743,
      "num_tokens": 207472051.0,
      "step": 6500
    },
    {
      "entropy": 0.8592873644083738,
      "epoch": 0.5981334207393862,
      "grad_norm": 0.15524490177631378,
      "learning_rate": 8.00656300794308e-05,
      "loss": 0.8512,
      "mean_token_accuracy": 0.7550871409475803,
      "num_tokens": 207504074.0,
      "step": 6501
    },
    {
      "entropy": 0.5241929590702057,
      "epoch": 0.5982254271108274,
      "grad_norm": 0.14207103848457336,
      "learning_rate": 8.006256325328918e-05,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8505542017519474,
      "num_tokens": 207535978.0,
      "step": 6502
    },
    {
      "entropy": 0.8054650630801916,
      "epoch": 0.5983174334822686,
      "grad_norm": 0.16788744926452637,
      "learning_rate": 8.005949642714755e-05,
      "loss": 0.7961,
      "mean_token_accuracy": 0.7715342715382576,
      "num_tokens": 207568003.0,
      "step": 6503
    },
    {
      "entropy": 0.7379892775788903,
      "epoch": 0.5984094398537099,
      "grad_norm": 0.15534688532352448,
      "learning_rate": 8.005642960100593e-05,
      "loss": 0.7266,
      "mean_token_accuracy": 0.7848309576511383,
      "num_tokens": 207599528.0,
      "step": 6504
    },
    {
      "entropy": 0.8020251244306564,
      "epoch": 0.5985014462251511,
      "grad_norm": 0.15531183779239655,
      "learning_rate": 8.00533627748643e-05,
      "loss": 0.8099,
      "mean_token_accuracy": 0.7612819187343121,
      "num_tokens": 207632181.0,
      "step": 6505
    },
    {
      "entropy": 0.754550845362246,
      "epoch": 0.5985934525965924,
      "grad_norm": 0.15638132393360138,
      "learning_rate": 8.005029594872268e-05,
      "loss": 0.7626,
      "mean_token_accuracy": 0.7792627699673176,
      "num_tokens": 207664444.0,
      "step": 6506
    },
    {
      "entropy": 0.7656289162114263,
      "epoch": 0.5986854589680335,
      "grad_norm": 0.15273159742355347,
      "learning_rate": 8.004722912258104e-05,
      "loss": 0.7467,
      "mean_token_accuracy": 0.7817647010087967,
      "num_tokens": 207696734.0,
      "step": 6507
    },
    {
      "entropy": 0.7410564869642258,
      "epoch": 0.5987774653394747,
      "grad_norm": 0.16536611318588257,
      "learning_rate": 8.004416229643941e-05,
      "loss": 0.7425,
      "mean_token_accuracy": 0.7818102762103081,
      "num_tokens": 207728203.0,
      "step": 6508
    },
    {
      "entropy": 0.7252880670130253,
      "epoch": 0.598869471710916,
      "grad_norm": 0.14986376464366913,
      "learning_rate": 8.004109547029779e-05,
      "loss": 0.7267,
      "mean_token_accuracy": 0.7800938338041306,
      "num_tokens": 207760390.0,
      "step": 6509
    },
    {
      "entropy": 0.7285486068576574,
      "epoch": 0.5989614780823572,
      "grad_norm": 0.16146208345890045,
      "learning_rate": 8.003802864415617e-05,
      "loss": 0.7276,
      "mean_token_accuracy": 0.7855895273387432,
      "num_tokens": 207792159.0,
      "step": 6510
    },
    {
      "entropy": 0.6989921238273382,
      "epoch": 0.5990534844537985,
      "grad_norm": 0.1575048267841339,
      "learning_rate": 8.003496181801454e-05,
      "loss": 0.6926,
      "mean_token_accuracy": 0.7970220670104027,
      "num_tokens": 207824234.0,
      "step": 6511
    },
    {
      "entropy": 0.8235620409250259,
      "epoch": 0.5991454908252396,
      "grad_norm": 0.15838590264320374,
      "learning_rate": 8.003189499187291e-05,
      "loss": 0.8223,
      "mean_token_accuracy": 0.761557012796402,
      "num_tokens": 207855754.0,
      "step": 6512
    },
    {
      "entropy": 0.8864719588309526,
      "epoch": 0.5992374971966808,
      "grad_norm": 0.15910327434539795,
      "learning_rate": 8.002882816573128e-05,
      "loss": 0.852,
      "mean_token_accuracy": 0.7525430731475353,
      "num_tokens": 207887266.0,
      "step": 6513
    },
    {
      "entropy": 0.7700284700840712,
      "epoch": 0.5993295035681221,
      "grad_norm": 0.17114853858947754,
      "learning_rate": 8.002576133958967e-05,
      "loss": 0.764,
      "mean_token_accuracy": 0.7813663855195045,
      "num_tokens": 207919081.0,
      "step": 6514
    },
    {
      "entropy": 0.6806915812194347,
      "epoch": 0.5994215099395633,
      "grad_norm": 0.1449906975030899,
      "learning_rate": 8.002269451344804e-05,
      "loss": 0.6561,
      "mean_token_accuracy": 0.8046813495457172,
      "num_tokens": 207951767.0,
      "step": 6515
    },
    {
      "entropy": 0.7461773995310068,
      "epoch": 0.5995135163110046,
      "grad_norm": 0.15247860550880432,
      "learning_rate": 8.001962768730641e-05,
      "loss": 0.7149,
      "mean_token_accuracy": 0.7905261069536209,
      "num_tokens": 207984070.0,
      "step": 6516
    },
    {
      "entropy": 0.7595107778906822,
      "epoch": 0.5996055226824457,
      "grad_norm": 0.15661612153053284,
      "learning_rate": 8.001656086116478e-05,
      "loss": 0.7372,
      "mean_token_accuracy": 0.7782615832984447,
      "num_tokens": 208016117.0,
      "step": 6517
    },
    {
      "entropy": 0.711479127407074,
      "epoch": 0.599697529053887,
      "grad_norm": 0.16032689809799194,
      "learning_rate": 8.001349403502316e-05,
      "loss": 0.6887,
      "mean_token_accuracy": 0.8002009801566601,
      "num_tokens": 208048135.0,
      "step": 6518
    },
    {
      "entropy": 0.8069780524820089,
      "epoch": 0.5997895354253282,
      "grad_norm": 0.15994541347026825,
      "learning_rate": 8.001042720888154e-05,
      "loss": 0.7994,
      "mean_token_accuracy": 0.76594003662467,
      "num_tokens": 208080572.0,
      "step": 6519
    },
    {
      "entropy": 0.7988325357437134,
      "epoch": 0.5998815417967694,
      "grad_norm": 0.15919171273708344,
      "learning_rate": 8.000736038273991e-05,
      "loss": 0.7932,
      "mean_token_accuracy": 0.7671983875334263,
      "num_tokens": 208112502.0,
      "step": 6520
    },
    {
      "entropy": 0.7726119179278612,
      "epoch": 0.5999735481682107,
      "grad_norm": 0.16263021528720856,
      "learning_rate": 8.000429355659828e-05,
      "loss": 0.7704,
      "mean_token_accuracy": 0.7732923589646816,
      "num_tokens": 208143869.0,
      "step": 6521
    },
    {
      "entropy": 0.6723619941622019,
      "epoch": 0.6000655545396518,
      "grad_norm": 0.1568424552679062,
      "learning_rate": 8.000122673045666e-05,
      "loss": 0.6487,
      "mean_token_accuracy": 0.8019725903868675,
      "num_tokens": 208175839.0,
      "step": 6522
    },
    {
      "entropy": 0.7276597954332829,
      "epoch": 0.6001575609110931,
      "grad_norm": 0.16762708127498627,
      "learning_rate": 7.999815990431502e-05,
      "loss": 0.7393,
      "mean_token_accuracy": 0.7822263836860657,
      "num_tokens": 208207830.0,
      "step": 6523
    },
    {
      "entropy": 0.6401676256209612,
      "epoch": 0.6002495672825343,
      "grad_norm": 0.15398135781288147,
      "learning_rate": 7.99950930781734e-05,
      "loss": 0.6401,
      "mean_token_accuracy": 0.8106694482266903,
      "num_tokens": 208239675.0,
      "step": 6524
    },
    {
      "entropy": 0.7679308019578457,
      "epoch": 0.6003415736539756,
      "grad_norm": 0.16490720212459564,
      "learning_rate": 7.999202625203177e-05,
      "loss": 0.7629,
      "mean_token_accuracy": 0.7799038030207157,
      "num_tokens": 208271653.0,
      "step": 6525
    },
    {
      "entropy": 0.7475208314135671,
      "epoch": 0.6004335800254168,
      "grad_norm": 0.1591360718011856,
      "learning_rate": 7.998895942589016e-05,
      "loss": 0.7428,
      "mean_token_accuracy": 0.7813245765864849,
      "num_tokens": 208303857.0,
      "step": 6526
    },
    {
      "entropy": 0.7701478768140078,
      "epoch": 0.6005255863968579,
      "grad_norm": 0.16947530210018158,
      "learning_rate": 7.998589259974852e-05,
      "loss": 0.7691,
      "mean_token_accuracy": 0.77600022777915,
      "num_tokens": 208335824.0,
      "step": 6527
    },
    {
      "entropy": 0.6849273219704628,
      "epoch": 0.6006175927682992,
      "grad_norm": 0.14898930490016937,
      "learning_rate": 7.998282577360689e-05,
      "loss": 0.6804,
      "mean_token_accuracy": 0.7975329756736755,
      "num_tokens": 208368057.0,
      "step": 6528
    },
    {
      "entropy": 0.652855547145009,
      "epoch": 0.6007095991397404,
      "grad_norm": 0.15667785704135895,
      "learning_rate": 7.997975894746527e-05,
      "loss": 0.6412,
      "mean_token_accuracy": 0.8087801225483418,
      "num_tokens": 208400405.0,
      "step": 6529
    },
    {
      "entropy": 0.732154494151473,
      "epoch": 0.6008016055111817,
      "grad_norm": 0.16122844815254211,
      "learning_rate": 7.997669212132365e-05,
      "loss": 0.712,
      "mean_token_accuracy": 0.7900843657553196,
      "num_tokens": 208432281.0,
      "step": 6530
    },
    {
      "entropy": 0.7963873688131571,
      "epoch": 0.6008936118826229,
      "grad_norm": 0.16405391693115234,
      "learning_rate": 7.997362529518202e-05,
      "loss": 0.7982,
      "mean_token_accuracy": 0.7684102021157742,
      "num_tokens": 208464616.0,
      "step": 6531
    },
    {
      "entropy": 0.6842956524342299,
      "epoch": 0.600985618254064,
      "grad_norm": 0.1501532644033432,
      "learning_rate": 7.997055846904039e-05,
      "loss": 0.6572,
      "mean_token_accuracy": 0.8003557734191418,
      "num_tokens": 208496575.0,
      "step": 6532
    },
    {
      "entropy": 0.7394830454140902,
      "epoch": 0.6010776246255053,
      "grad_norm": 0.16090528666973114,
      "learning_rate": 7.996749164289877e-05,
      "loss": 0.7336,
      "mean_token_accuracy": 0.7828326411545277,
      "num_tokens": 208528206.0,
      "step": 6533
    },
    {
      "entropy": 0.7899285778403282,
      "epoch": 0.6011696309969465,
      "grad_norm": 0.15661770105361938,
      "learning_rate": 7.996442481675714e-05,
      "loss": 0.7771,
      "mean_token_accuracy": 0.7706447057425976,
      "num_tokens": 208560502.0,
      "step": 6534
    },
    {
      "entropy": 0.7384083569049835,
      "epoch": 0.6012616373683878,
      "grad_norm": 0.14638780057430267,
      "learning_rate": 7.996135799061552e-05,
      "loss": 0.7061,
      "mean_token_accuracy": 0.7930985987186432,
      "num_tokens": 208592438.0,
      "step": 6535
    },
    {
      "entropy": 0.7083486896008253,
      "epoch": 0.601353643739829,
      "grad_norm": 0.15165138244628906,
      "learning_rate": 7.995829116447389e-05,
      "loss": 0.6945,
      "mean_token_accuracy": 0.796220675110817,
      "num_tokens": 208623836.0,
      "step": 6536
    },
    {
      "entropy": 0.8007223885506392,
      "epoch": 0.6014456501112702,
      "grad_norm": 0.15729112923145294,
      "learning_rate": 7.995522433833227e-05,
      "loss": 0.7812,
      "mean_token_accuracy": 0.7693633362650871,
      "num_tokens": 208655347.0,
      "step": 6537
    },
    {
      "entropy": 0.7703756839036942,
      "epoch": 0.6015376564827114,
      "grad_norm": 0.1580106019973755,
      "learning_rate": 7.995215751219064e-05,
      "loss": 0.7552,
      "mean_token_accuracy": 0.7767476178705692,
      "num_tokens": 208686893.0,
      "step": 6538
    },
    {
      "entropy": 0.7953444868326187,
      "epoch": 0.6016296628541526,
      "grad_norm": 0.16319894790649414,
      "learning_rate": 7.9949090686049e-05,
      "loss": 0.788,
      "mean_token_accuracy": 0.7752577848732471,
      "num_tokens": 208718388.0,
      "step": 6539
    },
    {
      "entropy": 0.7920876611024141,
      "epoch": 0.6017216692255939,
      "grad_norm": 0.16992564499378204,
      "learning_rate": 7.994602385990739e-05,
      "loss": 0.8031,
      "mean_token_accuracy": 0.7662144862115383,
      "num_tokens": 208750604.0,
      "step": 6540
    },
    {
      "entropy": 0.7869543805718422,
      "epoch": 0.6018136755970351,
      "grad_norm": 0.15088659524917603,
      "learning_rate": 7.994295703376577e-05,
      "loss": 0.7704,
      "mean_token_accuracy": 0.7708792164921761,
      "num_tokens": 208783029.0,
      "step": 6541
    },
    {
      "entropy": 0.7942906953394413,
      "epoch": 0.6019056819684763,
      "grad_norm": 0.15826645493507385,
      "learning_rate": 7.993989020762414e-05,
      "loss": 0.7967,
      "mean_token_accuracy": 0.7669575177133083,
      "num_tokens": 208815518.0,
      "step": 6542
    },
    {
      "entropy": 0.7023842185735703,
      "epoch": 0.6019976883399175,
      "grad_norm": 0.1449616253376007,
      "learning_rate": 7.99368233814825e-05,
      "loss": 0.6912,
      "mean_token_accuracy": 0.7946152500808239,
      "num_tokens": 208846769.0,
      "step": 6543
    },
    {
      "entropy": 0.7290205452591181,
      "epoch": 0.6020896947113588,
      "grad_norm": 0.15150055289268494,
      "learning_rate": 7.993375655534087e-05,
      "loss": 0.7203,
      "mean_token_accuracy": 0.7895846106112003,
      "num_tokens": 208878260.0,
      "step": 6544
    },
    {
      "entropy": 0.7711016740649939,
      "epoch": 0.6021817010828,
      "grad_norm": 0.15790529549121857,
      "learning_rate": 7.993068972919925e-05,
      "loss": 0.7668,
      "mean_token_accuracy": 0.7758297249674797,
      "num_tokens": 208909872.0,
      "step": 6545
    },
    {
      "entropy": 0.7453717272728682,
      "epoch": 0.6022737074542412,
      "grad_norm": 0.1471903920173645,
      "learning_rate": 7.992762290305763e-05,
      "loss": 0.7396,
      "mean_token_accuracy": 0.7779952771961689,
      "num_tokens": 208941367.0,
      "step": 6546
    },
    {
      "entropy": 0.8035736083984375,
      "epoch": 0.6023657138256824,
      "grad_norm": 0.16212083399295807,
      "learning_rate": 7.9924556076916e-05,
      "loss": 0.7912,
      "mean_token_accuracy": 0.7693401612341404,
      "num_tokens": 208973755.0,
      "step": 6547
    },
    {
      "entropy": 0.7555980794131756,
      "epoch": 0.6024577201971236,
      "grad_norm": 0.15873612463474274,
      "learning_rate": 7.992148925077437e-05,
      "loss": 0.7308,
      "mean_token_accuracy": 0.7797049097716808,
      "num_tokens": 209005349.0,
      "step": 6548
    },
    {
      "entropy": 0.6920330226421356,
      "epoch": 0.6025497265685649,
      "grad_norm": 0.15456484258174896,
      "learning_rate": 7.991842242463275e-05,
      "loss": 0.6795,
      "mean_token_accuracy": 0.7982978448271751,
      "num_tokens": 209038100.0,
      "step": 6549
    },
    {
      "entropy": 0.7000065557658672,
      "epoch": 0.6026417329400061,
      "grad_norm": 0.15978345274925232,
      "learning_rate": 7.991535559849112e-05,
      "loss": 0.7036,
      "mean_token_accuracy": 0.7948522157967091,
      "num_tokens": 209070868.0,
      "step": 6550
    },
    {
      "entropy": 0.8319433443248272,
      "epoch": 0.6027337393114474,
      "grad_norm": 0.15069535374641418,
      "learning_rate": 7.99122887723495e-05,
      "loss": 0.8208,
      "mean_token_accuracy": 0.7559921182692051,
      "num_tokens": 209102862.0,
      "step": 6551
    },
    {
      "entropy": 0.8474741540849209,
      "epoch": 0.6028257456828885,
      "grad_norm": 0.16358792781829834,
      "learning_rate": 7.990922194620787e-05,
      "loss": 0.8342,
      "mean_token_accuracy": 0.7580398395657539,
      "num_tokens": 209134833.0,
      "step": 6552
    },
    {
      "entropy": 0.7261071326211095,
      "epoch": 0.6029177520543297,
      "grad_norm": 0.14692673087120056,
      "learning_rate": 7.990615512006625e-05,
      "loss": 0.7091,
      "mean_token_accuracy": 0.790307842195034,
      "num_tokens": 209167068.0,
      "step": 6553
    },
    {
      "entropy": 0.8246330507099628,
      "epoch": 0.603009758425771,
      "grad_norm": 0.14882971346378326,
      "learning_rate": 7.990308829392462e-05,
      "loss": 0.8065,
      "mean_token_accuracy": 0.7633235491812229,
      "num_tokens": 209198852.0,
      "step": 6554
    },
    {
      "entropy": 0.80736268684268,
      "epoch": 0.6031017647972122,
      "grad_norm": 0.15595325827598572,
      "learning_rate": 7.990002146778299e-05,
      "loss": 0.7627,
      "mean_token_accuracy": 0.7787143178284168,
      "num_tokens": 209229935.0,
      "step": 6555
    },
    {
      "entropy": 0.7833071053028107,
      "epoch": 0.6031937711686535,
      "grad_norm": 0.16186264157295227,
      "learning_rate": 7.989695464164137e-05,
      "loss": 0.769,
      "mean_token_accuracy": 0.7737236246466637,
      "num_tokens": 209260849.0,
      "step": 6556
    },
    {
      "entropy": 0.8083824478089809,
      "epoch": 0.6032857775400946,
      "grad_norm": 0.159035786986351,
      "learning_rate": 7.989388781549975e-05,
      "loss": 0.8089,
      "mean_token_accuracy": 0.7637511044740677,
      "num_tokens": 209292936.0,
      "step": 6557
    },
    {
      "entropy": 0.767772413790226,
      "epoch": 0.6033777839115358,
      "grad_norm": 0.1527358442544937,
      "learning_rate": 7.989082098935812e-05,
      "loss": 0.7765,
      "mean_token_accuracy": 0.7725660763680935,
      "num_tokens": 209325001.0,
      "step": 6558
    },
    {
      "entropy": 0.7963237352669239,
      "epoch": 0.6034697902829771,
      "grad_norm": 0.17589490115642548,
      "learning_rate": 7.988775416321648e-05,
      "loss": 0.8088,
      "mean_token_accuracy": 0.76608806848526,
      "num_tokens": 209357231.0,
      "step": 6559
    },
    {
      "entropy": 0.8199936710298061,
      "epoch": 0.6035617966544183,
      "grad_norm": 0.1718231439590454,
      "learning_rate": 7.988468733707487e-05,
      "loss": 0.8158,
      "mean_token_accuracy": 0.7602086998522282,
      "num_tokens": 209388168.0,
      "step": 6560
    },
    {
      "entropy": 0.7515507787466049,
      "epoch": 0.6036538030258596,
      "grad_norm": 0.1553138643503189,
      "learning_rate": 7.988162051093325e-05,
      "loss": 0.7479,
      "mean_token_accuracy": 0.7762538380920887,
      "num_tokens": 209420693.0,
      "step": 6561
    },
    {
      "entropy": 0.7911493945866823,
      "epoch": 0.6037458093973007,
      "grad_norm": 0.17194153368473053,
      "learning_rate": 7.987855368479162e-05,
      "loss": 0.797,
      "mean_token_accuracy": 0.7702800184488297,
      "num_tokens": 209452949.0,
      "step": 6562
    },
    {
      "entropy": 0.7690427973866463,
      "epoch": 0.603837815768742,
      "grad_norm": 0.15877549350261688,
      "learning_rate": 7.987548685864998e-05,
      "loss": 0.7507,
      "mean_token_accuracy": 0.7811181582510471,
      "num_tokens": 209484068.0,
      "step": 6563
    },
    {
      "entropy": 0.7800704557448626,
      "epoch": 0.6039298221401832,
      "grad_norm": 0.15914031863212585,
      "learning_rate": 7.987242003250836e-05,
      "loss": 0.7771,
      "mean_token_accuracy": 0.7775832787156105,
      "num_tokens": 209516140.0,
      "step": 6564
    },
    {
      "entropy": 0.7687541898339987,
      "epoch": 0.6040218285116244,
      "grad_norm": 0.15189766883850098,
      "learning_rate": 7.986935320636673e-05,
      "loss": 0.7633,
      "mean_token_accuracy": 0.7798665352165699,
      "num_tokens": 209547744.0,
      "step": 6565
    },
    {
      "entropy": 0.6467344909906387,
      "epoch": 0.6041138348830657,
      "grad_norm": 0.15135617554187775,
      "learning_rate": 7.986628638022511e-05,
      "loss": 0.6197,
      "mean_token_accuracy": 0.8123287409543991,
      "num_tokens": 209579423.0,
      "step": 6566
    },
    {
      "entropy": 0.799315819516778,
      "epoch": 0.6042058412545068,
      "grad_norm": 0.1661297231912613,
      "learning_rate": 7.986321955408348e-05,
      "loss": 0.8011,
      "mean_token_accuracy": 0.7652314454317093,
      "num_tokens": 209611219.0,
      "step": 6567
    },
    {
      "entropy": 0.8238492049276829,
      "epoch": 0.6042978476259481,
      "grad_norm": 0.16145013272762299,
      "learning_rate": 7.986015272794186e-05,
      "loss": 0.8306,
      "mean_token_accuracy": 0.7579727061092854,
      "num_tokens": 209643360.0,
      "step": 6568
    },
    {
      "entropy": 0.7450622040778399,
      "epoch": 0.6043898539973893,
      "grad_norm": 0.15075799822807312,
      "learning_rate": 7.985708590180023e-05,
      "loss": 0.7145,
      "mean_token_accuracy": 0.7881363742053509,
      "num_tokens": 209675167.0,
      "step": 6569
    },
    {
      "entropy": 0.8160505816340446,
      "epoch": 0.6044818603688306,
      "grad_norm": 0.15104889869689941,
      "learning_rate": 7.98540190756586e-05,
      "loss": 0.7932,
      "mean_token_accuracy": 0.7683801054954529,
      "num_tokens": 209707059.0,
      "step": 6570
    },
    {
      "entropy": 0.6034088283777237,
      "epoch": 0.6045738667402718,
      "grad_norm": 0.15436023473739624,
      "learning_rate": 7.985095224951698e-05,
      "loss": 0.573,
      "mean_token_accuracy": 0.8258890509605408,
      "num_tokens": 209739053.0,
      "step": 6571
    },
    {
      "entropy": 0.8816453069448471,
      "epoch": 0.6046658731117129,
      "grad_norm": 0.16121132671833038,
      "learning_rate": 7.984788542337536e-05,
      "loss": 0.8739,
      "mean_token_accuracy": 0.74450958147645,
      "num_tokens": 209770622.0,
      "step": 6572
    },
    {
      "entropy": 0.8748078383505344,
      "epoch": 0.6047578794831542,
      "grad_norm": 0.16073639690876007,
      "learning_rate": 7.984481859723373e-05,
      "loss": 0.8782,
      "mean_token_accuracy": 0.7499445229768753,
      "num_tokens": 209802949.0,
      "step": 6573
    },
    {
      "entropy": 0.7309481259435415,
      "epoch": 0.6048498858545954,
      "grad_norm": 0.15880316495895386,
      "learning_rate": 7.98417517710921e-05,
      "loss": 0.7195,
      "mean_token_accuracy": 0.7940706498920918,
      "num_tokens": 209833553.0,
      "step": 6574
    },
    {
      "entropy": 0.7991013452410698,
      "epoch": 0.6049418922260367,
      "grad_norm": 0.16137157380580902,
      "learning_rate": 7.983868494495047e-05,
      "loss": 0.8063,
      "mean_token_accuracy": 0.7630074173212051,
      "num_tokens": 209866031.0,
      "step": 6575
    },
    {
      "entropy": 0.7924935966730118,
      "epoch": 0.6050338985974779,
      "grad_norm": 0.1531510055065155,
      "learning_rate": 7.983561811880885e-05,
      "loss": 0.7842,
      "mean_token_accuracy": 0.7686602920293808,
      "num_tokens": 209898445.0,
      "step": 6576
    },
    {
      "entropy": 0.7136400081217289,
      "epoch": 0.605125904968919,
      "grad_norm": 0.17016638815402985,
      "learning_rate": 7.983255129266723e-05,
      "loss": 0.7208,
      "mean_token_accuracy": 0.7875397130846977,
      "num_tokens": 209929983.0,
      "step": 6577
    },
    {
      "entropy": 0.6916261352598667,
      "epoch": 0.6052179113403603,
      "grad_norm": 0.15910348296165466,
      "learning_rate": 7.98294844665256e-05,
      "loss": 0.677,
      "mean_token_accuracy": 0.7978952825069427,
      "num_tokens": 209961989.0,
      "step": 6578
    },
    {
      "entropy": 0.8545396365225315,
      "epoch": 0.6053099177118015,
      "grad_norm": 0.16502618789672852,
      "learning_rate": 7.982641764038396e-05,
      "loss": 0.8537,
      "mean_token_accuracy": 0.7536961175501347,
      "num_tokens": 209994185.0,
      "step": 6579
    },
    {
      "entropy": 0.7960409764200449,
      "epoch": 0.6054019240832428,
      "grad_norm": 0.15946035087108612,
      "learning_rate": 7.982335081424235e-05,
      "loss": 0.7823,
      "mean_token_accuracy": 0.7754721455276012,
      "num_tokens": 210025385.0,
      "step": 6580
    },
    {
      "entropy": 0.7195939626544714,
      "epoch": 0.605493930454684,
      "grad_norm": 0.16873058676719666,
      "learning_rate": 7.982028398810071e-05,
      "loss": 0.7383,
      "mean_token_accuracy": 0.783303651958704,
      "num_tokens": 210057334.0,
      "step": 6581
    },
    {
      "entropy": 0.8134278329089284,
      "epoch": 0.6055859368261252,
      "grad_norm": 0.16449128091335297,
      "learning_rate": 7.98172171619591e-05,
      "loss": 0.7977,
      "mean_token_accuracy": 0.7582212723791599,
      "num_tokens": 210089256.0,
      "step": 6582
    },
    {
      "entropy": 0.7313144393265247,
      "epoch": 0.6056779431975664,
      "grad_norm": 0.1528351604938507,
      "learning_rate": 7.981415033581746e-05,
      "loss": 0.7339,
      "mean_token_accuracy": 0.7857807502150536,
      "num_tokens": 210121567.0,
      "step": 6583
    },
    {
      "entropy": 0.7307920046150684,
      "epoch": 0.6057699495690076,
      "grad_norm": 0.14961038529872894,
      "learning_rate": 7.981108350967584e-05,
      "loss": 0.7163,
      "mean_token_accuracy": 0.7887022159993649,
      "num_tokens": 210154231.0,
      "step": 6584
    },
    {
      "entropy": 0.8276907093822956,
      "epoch": 0.6058619559404489,
      "grad_norm": 0.15803076326847076,
      "learning_rate": 7.980801668353421e-05,
      "loss": 0.7983,
      "mean_token_accuracy": 0.7672717496752739,
      "num_tokens": 210185938.0,
      "step": 6585
    },
    {
      "entropy": 0.7541860155761242,
      "epoch": 0.6059539623118901,
      "grad_norm": 0.15463203191757202,
      "learning_rate": 7.980494985739258e-05,
      "loss": 0.7263,
      "mean_token_accuracy": 0.7855614647269249,
      "num_tokens": 210217331.0,
      "step": 6586
    },
    {
      "entropy": 0.7483357824385166,
      "epoch": 0.6060459686833313,
      "grad_norm": 0.15693438053131104,
      "learning_rate": 7.980188303125096e-05,
      "loss": 0.725,
      "mean_token_accuracy": 0.7884011715650558,
      "num_tokens": 210248826.0,
      "step": 6587
    },
    {
      "entropy": 0.648467680439353,
      "epoch": 0.6061379750547725,
      "grad_norm": 0.1536979228258133,
      "learning_rate": 7.979881620510934e-05,
      "loss": 0.6064,
      "mean_token_accuracy": 0.8235781081020832,
      "num_tokens": 210280693.0,
      "step": 6588
    },
    {
      "entropy": 0.9935454949736595,
      "epoch": 0.6062299814262138,
      "grad_norm": 0.16495104134082794,
      "learning_rate": 7.979574937896771e-05,
      "loss": 0.9901,
      "mean_token_accuracy": 0.7140422947704792,
      "num_tokens": 210312184.0,
      "step": 6589
    },
    {
      "entropy": 0.7695052735507488,
      "epoch": 0.606321987797655,
      "grad_norm": 0.15708249807357788,
      "learning_rate": 7.979268255282608e-05,
      "loss": 0.7584,
      "mean_token_accuracy": 0.7799543850123882,
      "num_tokens": 210343868.0,
      "step": 6590
    },
    {
      "entropy": 0.7566957734525204,
      "epoch": 0.6064139941690962,
      "grad_norm": 0.16267789900302887,
      "learning_rate": 7.978961572668446e-05,
      "loss": 0.7575,
      "mean_token_accuracy": 0.7764419615268707,
      "num_tokens": 210374876.0,
      "step": 6591
    },
    {
      "entropy": 0.8630033284425735,
      "epoch": 0.6065060005405374,
      "grad_norm": 0.16665709018707275,
      "learning_rate": 7.978654890054284e-05,
      "loss": 0.864,
      "mean_token_accuracy": 0.750651378184557,
      "num_tokens": 210407103.0,
      "step": 6592
    },
    {
      "entropy": 0.8305176068097353,
      "epoch": 0.6065980069119786,
      "grad_norm": 0.16194674372673035,
      "learning_rate": 7.978348207440121e-05,
      "loss": 0.834,
      "mean_token_accuracy": 0.7584124058485031,
      "num_tokens": 210439139.0,
      "step": 6593
    },
    {
      "entropy": 0.6215958567336202,
      "epoch": 0.6066900132834199,
      "grad_norm": 0.1539551168680191,
      "learning_rate": 7.978041524825958e-05,
      "loss": 0.615,
      "mean_token_accuracy": 0.8170161359012127,
      "num_tokens": 210471287.0,
      "step": 6594
    },
    {
      "entropy": 0.7322047809138894,
      "epoch": 0.6067820196548611,
      "grad_norm": 0.16111448407173157,
      "learning_rate": 7.977734842211796e-05,
      "loss": 0.7472,
      "mean_token_accuracy": 0.7842026464641094,
      "num_tokens": 210503847.0,
      "step": 6595
    },
    {
      "entropy": 0.7706270217895508,
      "epoch": 0.6068740260263024,
      "grad_norm": 0.15339678525924683,
      "learning_rate": 7.977428159597633e-05,
      "loss": 0.7475,
      "mean_token_accuracy": 0.7804178968071938,
      "num_tokens": 210536015.0,
      "step": 6596
    },
    {
      "entropy": 0.7801286298781633,
      "epoch": 0.6069660323977435,
      "grad_norm": 0.15024489164352417,
      "learning_rate": 7.977121476983471e-05,
      "loss": 0.7722,
      "mean_token_accuracy": 0.7684030048549175,
      "num_tokens": 210567648.0,
      "step": 6597
    },
    {
      "entropy": 0.7636658437550068,
      "epoch": 0.6070580387691847,
      "grad_norm": 0.16011592745780945,
      "learning_rate": 7.976814794369308e-05,
      "loss": 0.7678,
      "mean_token_accuracy": 0.7752377428114414,
      "num_tokens": 210599515.0,
      "step": 6598
    },
    {
      "entropy": 0.7189522981643677,
      "epoch": 0.607150045140626,
      "grad_norm": 0.15910989046096802,
      "learning_rate": 7.976508111755146e-05,
      "loss": 0.7176,
      "mean_token_accuracy": 0.7852495163679123,
      "num_tokens": 210630662.0,
      "step": 6599
    },
    {
      "entropy": 0.7190588554367423,
      "epoch": 0.6072420515120672,
      "grad_norm": 0.16316023468971252,
      "learning_rate": 7.976201429140982e-05,
      "loss": 0.7101,
      "mean_token_accuracy": 0.7897082604467869,
      "num_tokens": 210662851.0,
      "step": 6600
    },
    {
      "entropy": 0.785878948867321,
      "epoch": 0.6073340578835085,
      "grad_norm": 0.1580585241317749,
      "learning_rate": 7.975894746526819e-05,
      "loss": 0.7708,
      "mean_token_accuracy": 0.7725542783737183,
      "num_tokens": 210695132.0,
      "step": 6601
    },
    {
      "entropy": 0.7976447883993387,
      "epoch": 0.6074260642549496,
      "grad_norm": 0.1546858698129654,
      "learning_rate": 7.975588063912657e-05,
      "loss": 0.7709,
      "mean_token_accuracy": 0.7693204656243324,
      "num_tokens": 210727504.0,
      "step": 6602
    },
    {
      "entropy": 0.7257482055574656,
      "epoch": 0.6075180706263908,
      "grad_norm": 0.16131651401519775,
      "learning_rate": 7.975281381298496e-05,
      "loss": 0.7118,
      "mean_token_accuracy": 0.7915220968425274,
      "num_tokens": 210759866.0,
      "step": 6603
    },
    {
      "entropy": 0.8481794856488705,
      "epoch": 0.6076100769978321,
      "grad_norm": 0.16166168451309204,
      "learning_rate": 7.974974698684332e-05,
      "loss": 0.825,
      "mean_token_accuracy": 0.7579395733773708,
      "num_tokens": 210791172.0,
      "step": 6604
    },
    {
      "entropy": 0.749772809445858,
      "epoch": 0.6077020833692733,
      "grad_norm": 0.15904855728149414,
      "learning_rate": 7.974668016070169e-05,
      "loss": 0.7353,
      "mean_token_accuracy": 0.7805076166987419,
      "num_tokens": 210822961.0,
      "step": 6605
    },
    {
      "entropy": 0.8271678313612938,
      "epoch": 0.6077940897407146,
      "grad_norm": 0.16505098342895508,
      "learning_rate": 7.974361333456006e-05,
      "loss": 0.8197,
      "mean_token_accuracy": 0.7639963515102863,
      "num_tokens": 210855166.0,
      "step": 6606
    },
    {
      "entropy": 0.783656282350421,
      "epoch": 0.6078860961121557,
      "grad_norm": 0.15656785666942596,
      "learning_rate": 7.974054650841844e-05,
      "loss": 0.7621,
      "mean_token_accuracy": 0.7784202210605145,
      "num_tokens": 210885311.0,
      "step": 6607
    },
    {
      "entropy": 0.7863801829516888,
      "epoch": 0.607978102483597,
      "grad_norm": 0.15652650594711304,
      "learning_rate": 7.973747968227682e-05,
      "loss": 0.7968,
      "mean_token_accuracy": 0.7675429806113243,
      "num_tokens": 210917387.0,
      "step": 6608
    },
    {
      "entropy": 0.779412304982543,
      "epoch": 0.6080701088550382,
      "grad_norm": 0.15526188910007477,
      "learning_rate": 7.973441285613519e-05,
      "loss": 0.7685,
      "mean_token_accuracy": 0.772335659712553,
      "num_tokens": 210949152.0,
      "step": 6609
    },
    {
      "entropy": 0.7853844538331032,
      "epoch": 0.6081621152264794,
      "grad_norm": 0.15987880527973175,
      "learning_rate": 7.973134602999356e-05,
      "loss": 0.7775,
      "mean_token_accuracy": 0.7697658315300941,
      "num_tokens": 210981403.0,
      "step": 6610
    },
    {
      "entropy": 0.8023937176913023,
      "epoch": 0.6082541215979207,
      "grad_norm": 0.1517046093940735,
      "learning_rate": 7.972827920385194e-05,
      "loss": 0.7998,
      "mean_token_accuracy": 0.7672928273677826,
      "num_tokens": 211013675.0,
      "step": 6611
    },
    {
      "entropy": 0.7625174578279257,
      "epoch": 0.6083461279693618,
      "grad_norm": 0.16047334671020508,
      "learning_rate": 7.972521237771031e-05,
      "loss": 0.7647,
      "mean_token_accuracy": 0.7714215964078903,
      "num_tokens": 211043790.0,
      "step": 6612
    },
    {
      "entropy": 0.8347377851605415,
      "epoch": 0.6084381343408031,
      "grad_norm": 0.16077767312526703,
      "learning_rate": 7.972214555156869e-05,
      "loss": 0.8191,
      "mean_token_accuracy": 0.76403384283185,
      "num_tokens": 211075380.0,
      "step": 6613
    },
    {
      "entropy": 0.7236975282430649,
      "epoch": 0.6085301407122443,
      "grad_norm": 0.1595371514558792,
      "learning_rate": 7.971907872542706e-05,
      "loss": 0.7181,
      "mean_token_accuracy": 0.7909240238368511,
      "num_tokens": 211107084.0,
      "step": 6614
    },
    {
      "entropy": 0.7390807420015335,
      "epoch": 0.6086221470836856,
      "grad_norm": 0.18780814111232758,
      "learning_rate": 7.971601189928544e-05,
      "loss": 0.7594,
      "mean_token_accuracy": 0.7792432382702827,
      "num_tokens": 211139245.0,
      "step": 6615
    },
    {
      "entropy": 0.6694026347249746,
      "epoch": 0.6087141534551268,
      "grad_norm": 0.1537383496761322,
      "learning_rate": 7.97129450731438e-05,
      "loss": 0.658,
      "mean_token_accuracy": 0.8036625534296036,
      "num_tokens": 211171262.0,
      "step": 6616
    },
    {
      "entropy": 0.8196073696017265,
      "epoch": 0.6088061598265679,
      "grad_norm": 0.17078201472759247,
      "learning_rate": 7.970987824700217e-05,
      "loss": 0.7871,
      "mean_token_accuracy": 0.7710178308188915,
      "num_tokens": 211203624.0,
      "step": 6617
    },
    {
      "entropy": 0.6513512367382646,
      "epoch": 0.6088981661980092,
      "grad_norm": 0.14339017868041992,
      "learning_rate": 7.970681142086056e-05,
      "loss": 0.6201,
      "mean_token_accuracy": 0.8136495985090733,
      "num_tokens": 211235963.0,
      "step": 6618
    },
    {
      "entropy": 0.7762915194034576,
      "epoch": 0.6089901725694504,
      "grad_norm": 0.1527511030435562,
      "learning_rate": 7.970374459471894e-05,
      "loss": 0.7577,
      "mean_token_accuracy": 0.7786231189966202,
      "num_tokens": 211268137.0,
      "step": 6619
    },
    {
      "entropy": 0.6614007614552975,
      "epoch": 0.6090821789408917,
      "grad_norm": 0.1640542596578598,
      "learning_rate": 7.97006777685773e-05,
      "loss": 0.6399,
      "mean_token_accuracy": 0.8107142671942711,
      "num_tokens": 211300003.0,
      "step": 6620
    },
    {
      "entropy": 0.8088312409818172,
      "epoch": 0.6091741853123329,
      "grad_norm": 0.1562281996011734,
      "learning_rate": 7.969761094243567e-05,
      "loss": 0.7873,
      "mean_token_accuracy": 0.767872080206871,
      "num_tokens": 211331833.0,
      "step": 6621
    },
    {
      "entropy": 0.7271502036601305,
      "epoch": 0.609266191683774,
      "grad_norm": 0.15009111166000366,
      "learning_rate": 7.969454411629405e-05,
      "loss": 0.7064,
      "mean_token_accuracy": 0.7914947010576725,
      "num_tokens": 211364091.0,
      "step": 6622
    },
    {
      "entropy": 0.7459986601024866,
      "epoch": 0.6093581980552153,
      "grad_norm": 0.16261646151542664,
      "learning_rate": 7.969147729015242e-05,
      "loss": 0.7293,
      "mean_token_accuracy": 0.7829660810530186,
      "num_tokens": 211395962.0,
      "step": 6623
    },
    {
      "entropy": 0.7554231937974691,
      "epoch": 0.6094502044266565,
      "grad_norm": 0.15104639530181885,
      "learning_rate": 7.96884104640108e-05,
      "loss": 0.7363,
      "mean_token_accuracy": 0.7805481068789959,
      "num_tokens": 211428036.0,
      "step": 6624
    },
    {
      "entropy": 0.7729437090456486,
      "epoch": 0.6095422107980978,
      "grad_norm": 0.16430875658988953,
      "learning_rate": 7.968534363786917e-05,
      "loss": 0.7639,
      "mean_token_accuracy": 0.7774238772690296,
      "num_tokens": 211459536.0,
      "step": 6625
    },
    {
      "entropy": 0.7495695762336254,
      "epoch": 0.609634217169539,
      "grad_norm": 0.15760618448257446,
      "learning_rate": 7.968227681172755e-05,
      "loss": 0.7515,
      "mean_token_accuracy": 0.777808915823698,
      "num_tokens": 211490149.0,
      "step": 6626
    },
    {
      "entropy": 0.7304324293509126,
      "epoch": 0.6097262235409802,
      "grad_norm": 0.1557258814573288,
      "learning_rate": 7.967920998558592e-05,
      "loss": 0.7252,
      "mean_token_accuracy": 0.7892836853861809,
      "num_tokens": 211521704.0,
      "step": 6627
    },
    {
      "entropy": 0.6953580984845757,
      "epoch": 0.6098182299124214,
      "grad_norm": 0.1574198454618454,
      "learning_rate": 7.967614315944429e-05,
      "loss": 0.6836,
      "mean_token_accuracy": 0.8019516989588737,
      "num_tokens": 211553708.0,
      "step": 6628
    },
    {
      "entropy": 0.7454677280038595,
      "epoch": 0.6099102362838627,
      "grad_norm": 0.15452569723129272,
      "learning_rate": 7.967307633330267e-05,
      "loss": 0.7472,
      "mean_token_accuracy": 0.7771567516028881,
      "num_tokens": 211586272.0,
      "step": 6629
    },
    {
      "entropy": 0.6520652286708355,
      "epoch": 0.6100022426553039,
      "grad_norm": 0.15671966969966888,
      "learning_rate": 7.967000950716105e-05,
      "loss": 0.6524,
      "mean_token_accuracy": 0.8085168153047562,
      "num_tokens": 211618467.0,
      "step": 6630
    },
    {
      "entropy": 0.7755056824535131,
      "epoch": 0.6100942490267451,
      "grad_norm": 0.1639256626367569,
      "learning_rate": 7.966694268101942e-05,
      "loss": 0.7641,
      "mean_token_accuracy": 0.777957409620285,
      "num_tokens": 211650326.0,
      "step": 6631
    },
    {
      "entropy": 0.7169619556516409,
      "epoch": 0.6101862553981863,
      "grad_norm": 0.15627935528755188,
      "learning_rate": 7.966387585487779e-05,
      "loss": 0.7114,
      "mean_token_accuracy": 0.7921593710780144,
      "num_tokens": 211681921.0,
      "step": 6632
    },
    {
      "entropy": 0.769679881632328,
      "epoch": 0.6102782617696275,
      "grad_norm": 0.15452654659748077,
      "learning_rate": 7.966080902873615e-05,
      "loss": 0.7801,
      "mean_token_accuracy": 0.7721533477306366,
      "num_tokens": 211713610.0,
      "step": 6633
    },
    {
      "entropy": 0.8633562289178371,
      "epoch": 0.6103702681410688,
      "grad_norm": 0.17279057204723358,
      "learning_rate": 7.965774220259455e-05,
      "loss": 0.8758,
      "mean_token_accuracy": 0.748928751796484,
      "num_tokens": 211744017.0,
      "step": 6634
    },
    {
      "entropy": 0.7830563988536596,
      "epoch": 0.61046227451251,
      "grad_norm": 0.15531402826309204,
      "learning_rate": 7.965467537645292e-05,
      "loss": 0.7646,
      "mean_token_accuracy": 0.7752265632152557,
      "num_tokens": 211776057.0,
      "step": 6635
    },
    {
      "entropy": 0.7385657504200935,
      "epoch": 0.6105542808839513,
      "grad_norm": 0.1529969722032547,
      "learning_rate": 7.965160855031129e-05,
      "loss": 0.7144,
      "mean_token_accuracy": 0.7915769256651402,
      "num_tokens": 211807034.0,
      "step": 6636
    },
    {
      "entropy": 0.8132250159978867,
      "epoch": 0.6106462872553924,
      "grad_norm": 0.1591726690530777,
      "learning_rate": 7.964854172416965e-05,
      "loss": 0.8221,
      "mean_token_accuracy": 0.7642566226422787,
      "num_tokens": 211838946.0,
      "step": 6637
    },
    {
      "entropy": 0.6804377818480134,
      "epoch": 0.6107382936268336,
      "grad_norm": 0.1447530835866928,
      "learning_rate": 7.964547489802803e-05,
      "loss": 0.6645,
      "mean_token_accuracy": 0.8001934476196766,
      "num_tokens": 211871506.0,
      "step": 6638
    },
    {
      "entropy": 0.649226650595665,
      "epoch": 0.6108302999982749,
      "grad_norm": 0.14411744475364685,
      "learning_rate": 7.964240807188642e-05,
      "loss": 0.6295,
      "mean_token_accuracy": 0.8123362809419632,
      "num_tokens": 211903168.0,
      "step": 6639
    },
    {
      "entropy": 0.7938334830105305,
      "epoch": 0.6109223063697161,
      "grad_norm": 0.1632513850927353,
      "learning_rate": 7.963934124574478e-05,
      "loss": 0.8069,
      "mean_token_accuracy": 0.762355774641037,
      "num_tokens": 211935451.0,
      "step": 6640
    },
    {
      "entropy": 0.8145322296768427,
      "epoch": 0.6110143127411574,
      "grad_norm": 0.164145365357399,
      "learning_rate": 7.963627441960315e-05,
      "loss": 0.8039,
      "mean_token_accuracy": 0.7642817497253418,
      "num_tokens": 211966894.0,
      "step": 6641
    },
    {
      "entropy": 0.7050241213291883,
      "epoch": 0.6111063191125985,
      "grad_norm": 0.15375614166259766,
      "learning_rate": 7.963320759346153e-05,
      "loss": 0.7009,
      "mean_token_accuracy": 0.7912631332874298,
      "num_tokens": 211998401.0,
      "step": 6642
    },
    {
      "entropy": 0.8269395977258682,
      "epoch": 0.6111983254840397,
      "grad_norm": 0.16130326688289642,
      "learning_rate": 7.96301407673199e-05,
      "loss": 0.8125,
      "mean_token_accuracy": 0.7592475898563862,
      "num_tokens": 212029135.0,
      "step": 6643
    },
    {
      "entropy": 0.6851309668272734,
      "epoch": 0.611290331855481,
      "grad_norm": 0.14882177114486694,
      "learning_rate": 7.962707394117828e-05,
      "loss": 0.6593,
      "mean_token_accuracy": 0.8075400367379189,
      "num_tokens": 212060765.0,
      "step": 6644
    },
    {
      "entropy": 0.760255379602313,
      "epoch": 0.6113823382269222,
      "grad_norm": 0.15952490270137787,
      "learning_rate": 7.962400711503665e-05,
      "loss": 0.7481,
      "mean_token_accuracy": 0.7803831249475479,
      "num_tokens": 212093191.0,
      "step": 6645
    },
    {
      "entropy": 0.7954575959593058,
      "epoch": 0.6114743445983635,
      "grad_norm": 0.16364724934101105,
      "learning_rate": 7.962094028889503e-05,
      "loss": 0.7662,
      "mean_token_accuracy": 0.775992713868618,
      "num_tokens": 212124516.0,
      "step": 6646
    },
    {
      "entropy": 0.7328708805143833,
      "epoch": 0.6115663509698046,
      "grad_norm": 0.1567360758781433,
      "learning_rate": 7.96178734627534e-05,
      "loss": 0.7221,
      "mean_token_accuracy": 0.787740346044302,
      "num_tokens": 212156835.0,
      "step": 6647
    },
    {
      "entropy": 0.6800538552924991,
      "epoch": 0.6116583573412459,
      "grad_norm": 0.15419915318489075,
      "learning_rate": 7.961480663661177e-05,
      "loss": 0.6686,
      "mean_token_accuracy": 0.8018447868525982,
      "num_tokens": 212189073.0,
      "step": 6648
    },
    {
      "entropy": 0.8660410940647125,
      "epoch": 0.6117503637126871,
      "grad_norm": 0.1648206263780594,
      "learning_rate": 7.961173981047015e-05,
      "loss": 0.8491,
      "mean_token_accuracy": 0.7545182779431343,
      "num_tokens": 212220178.0,
      "step": 6649
    },
    {
      "entropy": 0.5847712084650993,
      "epoch": 0.6118423700841283,
      "grad_norm": 0.161371648311615,
      "learning_rate": 7.960867298432853e-05,
      "loss": 0.596,
      "mean_token_accuracy": 0.823094766587019,
      "num_tokens": 212252946.0,
      "step": 6650
    },
    {
      "entropy": 0.7672773152589798,
      "epoch": 0.6119343764555696,
      "grad_norm": 0.15938608348369598,
      "learning_rate": 7.96056061581869e-05,
      "loss": 0.7657,
      "mean_token_accuracy": 0.7804109789431095,
      "num_tokens": 212284711.0,
      "step": 6651
    },
    {
      "entropy": 0.8194014988839626,
      "epoch": 0.6120263828270107,
      "grad_norm": 0.16147950291633606,
      "learning_rate": 7.960253933204527e-05,
      "loss": 0.7749,
      "mean_token_accuracy": 0.7744439840316772,
      "num_tokens": 212316064.0,
      "step": 6652
    },
    {
      "entropy": 0.7776915170252323,
      "epoch": 0.612118389198452,
      "grad_norm": 0.15687604248523712,
      "learning_rate": 7.959947250590365e-05,
      "loss": 0.765,
      "mean_token_accuracy": 0.7762802243232727,
      "num_tokens": 212348720.0,
      "step": 6653
    },
    {
      "entropy": 0.7921652868390083,
      "epoch": 0.6122103955698932,
      "grad_norm": 0.16418609023094177,
      "learning_rate": 7.959640567976202e-05,
      "loss": 0.8121,
      "mean_token_accuracy": 0.7621974647045135,
      "num_tokens": 212381304.0,
      "step": 6654
    },
    {
      "entropy": 0.7257687691599131,
      "epoch": 0.6123024019413345,
      "grad_norm": 0.1561390906572342,
      "learning_rate": 7.95933388536204e-05,
      "loss": 0.7158,
      "mean_token_accuracy": 0.7880750223994255,
      "num_tokens": 212413534.0,
      "step": 6655
    },
    {
      "entropy": 0.9156952537596226,
      "epoch": 0.6123944083127757,
      "grad_norm": 0.1660735309123993,
      "learning_rate": 7.959027202747876e-05,
      "loss": 0.9021,
      "mean_token_accuracy": 0.7402945905923843,
      "num_tokens": 212444871.0,
      "step": 6656
    },
    {
      "entropy": 0.7840570621192455,
      "epoch": 0.6124864146842168,
      "grad_norm": 0.15284180641174316,
      "learning_rate": 7.958720520133715e-05,
      "loss": 0.7816,
      "mean_token_accuracy": 0.7694744095206261,
      "num_tokens": 212477324.0,
      "step": 6657
    },
    {
      "entropy": 0.723306629806757,
      "epoch": 0.6125784210556581,
      "grad_norm": 0.15021628141403198,
      "learning_rate": 7.958413837519551e-05,
      "loss": 0.7111,
      "mean_token_accuracy": 0.7915221527218819,
      "num_tokens": 212509819.0,
      "step": 6658
    },
    {
      "entropy": 0.793012723326683,
      "epoch": 0.6126704274270993,
      "grad_norm": 0.1640978902578354,
      "learning_rate": 7.958107154905388e-05,
      "loss": 0.7959,
      "mean_token_accuracy": 0.7666523791849613,
      "num_tokens": 212541503.0,
      "step": 6659
    },
    {
      "entropy": 0.7863712012767792,
      "epoch": 0.6127624337985406,
      "grad_norm": 0.16174615919589996,
      "learning_rate": 7.957800472291226e-05,
      "loss": 0.7695,
      "mean_token_accuracy": 0.7796160690486431,
      "num_tokens": 212573610.0,
      "step": 6660
    },
    {
      "entropy": 0.7861715573817492,
      "epoch": 0.6128544401699818,
      "grad_norm": 0.15364255011081696,
      "learning_rate": 7.957493789677064e-05,
      "loss": 0.7786,
      "mean_token_accuracy": 0.77163016051054,
      "num_tokens": 212605168.0,
      "step": 6661
    },
    {
      "entropy": 0.7250403705984354,
      "epoch": 0.6129464465414229,
      "grad_norm": 0.160390704870224,
      "learning_rate": 7.957187107062901e-05,
      "loss": 0.7354,
      "mean_token_accuracy": 0.7803471237421036,
      "num_tokens": 212636523.0,
      "step": 6662
    },
    {
      "entropy": 0.7669453397393227,
      "epoch": 0.6130384529128642,
      "grad_norm": 0.15723715722560883,
      "learning_rate": 7.956880424448738e-05,
      "loss": 0.7548,
      "mean_token_accuracy": 0.7770597748458385,
      "num_tokens": 212667721.0,
      "step": 6663
    },
    {
      "entropy": 0.7054450530558825,
      "epoch": 0.6131304592843054,
      "grad_norm": 0.16165335476398468,
      "learning_rate": 7.956573741834575e-05,
      "loss": 0.7142,
      "mean_token_accuracy": 0.7949486151337624,
      "num_tokens": 212699495.0,
      "step": 6664
    },
    {
      "entropy": 0.7374550104141235,
      "epoch": 0.6132224656557467,
      "grad_norm": 0.1605551540851593,
      "learning_rate": 7.956267059220413e-05,
      "loss": 0.7364,
      "mean_token_accuracy": 0.7837115414440632,
      "num_tokens": 212731159.0,
      "step": 6665
    },
    {
      "entropy": 0.6429927051067352,
      "epoch": 0.6133144720271879,
      "grad_norm": 0.147960364818573,
      "learning_rate": 7.955960376606251e-05,
      "loss": 0.6179,
      "mean_token_accuracy": 0.8155076615512371,
      "num_tokens": 212763140.0,
      "step": 6666
    },
    {
      "entropy": 0.8386082071810961,
      "epoch": 0.613406478398629,
      "grad_norm": 0.16985975205898285,
      "learning_rate": 7.955653693992088e-05,
      "loss": 0.8459,
      "mean_token_accuracy": 0.7526288740336895,
      "num_tokens": 212794774.0,
      "step": 6667
    },
    {
      "entropy": 0.7906545922160149,
      "epoch": 0.6134984847700703,
      "grad_norm": 0.15552859008312225,
      "learning_rate": 7.955347011377925e-05,
      "loss": 0.784,
      "mean_token_accuracy": 0.7689914889633656,
      "num_tokens": 212827240.0,
      "step": 6668
    },
    {
      "entropy": 0.7791570983827114,
      "epoch": 0.6135904911415115,
      "grad_norm": 0.15880948305130005,
      "learning_rate": 7.955040328763763e-05,
      "loss": 0.7783,
      "mean_token_accuracy": 0.7733309008181095,
      "num_tokens": 212859668.0,
      "step": 6669
    },
    {
      "entropy": 0.7832208685576916,
      "epoch": 0.6136824975129528,
      "grad_norm": 0.16190147399902344,
      "learning_rate": 7.9547336461496e-05,
      "loss": 0.7769,
      "mean_token_accuracy": 0.7737848088145256,
      "num_tokens": 212892436.0,
      "step": 6670
    },
    {
      "entropy": 0.7396881468594074,
      "epoch": 0.613774503884394,
      "grad_norm": 0.1567920744419098,
      "learning_rate": 7.954426963535438e-05,
      "loss": 0.7332,
      "mean_token_accuracy": 0.783530980348587,
      "num_tokens": 212925169.0,
      "step": 6671
    },
    {
      "entropy": 0.8518978077918291,
      "epoch": 0.6138665102558352,
      "grad_norm": 0.16225937008857727,
      "learning_rate": 7.954120280921275e-05,
      "loss": 0.828,
      "mean_token_accuracy": 0.7586348801851273,
      "num_tokens": 212957096.0,
      "step": 6672
    },
    {
      "entropy": 0.8447510488331318,
      "epoch": 0.6139585166272764,
      "grad_norm": 0.15840625762939453,
      "learning_rate": 7.953813598307113e-05,
      "loss": 0.8312,
      "mean_token_accuracy": 0.7543718479573727,
      "num_tokens": 212988049.0,
      "step": 6673
    },
    {
      "entropy": 0.6874769944697618,
      "epoch": 0.6140505229987177,
      "grad_norm": 0.15873205661773682,
      "learning_rate": 7.95350691569295e-05,
      "loss": 0.6803,
      "mean_token_accuracy": 0.8002268224954605,
      "num_tokens": 213019217.0,
      "step": 6674
    },
    {
      "entropy": 0.843929335474968,
      "epoch": 0.6141425293701589,
      "grad_norm": 0.15694262087345123,
      "learning_rate": 7.953200233078786e-05,
      "loss": 0.8279,
      "mean_token_accuracy": 0.7575494684278965,
      "num_tokens": 213050810.0,
      "step": 6675
    },
    {
      "entropy": 0.7572820708155632,
      "epoch": 0.6142345357416001,
      "grad_norm": 0.15708081424236298,
      "learning_rate": 7.952893550464624e-05,
      "loss": 0.7502,
      "mean_token_accuracy": 0.7764142118394375,
      "num_tokens": 213081798.0,
      "step": 6676
    },
    {
      "entropy": 0.8281751293689013,
      "epoch": 0.6143265421130413,
      "grad_norm": 0.1675480306148529,
      "learning_rate": 7.952586867850463e-05,
      "loss": 0.8135,
      "mean_token_accuracy": 0.7639302536845207,
      "num_tokens": 213113267.0,
      "step": 6677
    },
    {
      "entropy": 0.8072092458605766,
      "epoch": 0.6144185484844825,
      "grad_norm": 0.16220667958259583,
      "learning_rate": 7.9522801852363e-05,
      "loss": 0.8232,
      "mean_token_accuracy": 0.7629742473363876,
      "num_tokens": 213145321.0,
      "step": 6678
    },
    {
      "entropy": 0.8377286158502102,
      "epoch": 0.6145105548559238,
      "grad_norm": 0.15433530509471893,
      "learning_rate": 7.951973502622136e-05,
      "loss": 0.8183,
      "mean_token_accuracy": 0.7566869296133518,
      "num_tokens": 213177446.0,
      "step": 6679
    },
    {
      "entropy": 0.7631161399185658,
      "epoch": 0.614602561227365,
      "grad_norm": 0.15426382422447205,
      "learning_rate": 7.951666820007974e-05,
      "loss": 0.7351,
      "mean_token_accuracy": 0.7815495170652866,
      "num_tokens": 213208726.0,
      "step": 6680
    },
    {
      "entropy": 0.7846670318394899,
      "epoch": 0.6146945675988063,
      "grad_norm": 0.15723967552185059,
      "learning_rate": 7.951360137393812e-05,
      "loss": 0.7709,
      "mean_token_accuracy": 0.7736463807523251,
      "num_tokens": 213241003.0,
      "step": 6681
    },
    {
      "entropy": 0.7181870695203543,
      "epoch": 0.6147865739702474,
      "grad_norm": 0.15284429490566254,
      "learning_rate": 7.951053454779649e-05,
      "loss": 0.7204,
      "mean_token_accuracy": 0.7873183973133564,
      "num_tokens": 213273555.0,
      "step": 6682
    },
    {
      "entropy": 0.8903865404427052,
      "epoch": 0.6148785803416886,
      "grad_norm": 0.16499221324920654,
      "learning_rate": 7.950746772165486e-05,
      "loss": 0.8839,
      "mean_token_accuracy": 0.744561068713665,
      "num_tokens": 213305394.0,
      "step": 6683
    },
    {
      "entropy": 0.8554773721843958,
      "epoch": 0.6149705867131299,
      "grad_norm": 0.16370102763175964,
      "learning_rate": 7.950440089551324e-05,
      "loss": 0.8358,
      "mean_token_accuracy": 0.7544255256652832,
      "num_tokens": 213337374.0,
      "step": 6684
    },
    {
      "entropy": 0.8110974840819836,
      "epoch": 0.6150625930845711,
      "grad_norm": 0.1649705022573471,
      "learning_rate": 7.950133406937161e-05,
      "loss": 0.7952,
      "mean_token_accuracy": 0.7710266672074795,
      "num_tokens": 213368833.0,
      "step": 6685
    },
    {
      "entropy": 0.695953119546175,
      "epoch": 0.6151545994560124,
      "grad_norm": 0.1548963338136673,
      "learning_rate": 7.949826724322999e-05,
      "loss": 0.6681,
      "mean_token_accuracy": 0.8007498905062675,
      "num_tokens": 213401416.0,
      "step": 6686
    },
    {
      "entropy": 0.6630786471068859,
      "epoch": 0.6152466058274535,
      "grad_norm": 0.15383850038051605,
      "learning_rate": 7.949520041708836e-05,
      "loss": 0.6485,
      "mean_token_accuracy": 0.8027797676622868,
      "num_tokens": 213433084.0,
      "step": 6687
    },
    {
      "entropy": 0.7375473603606224,
      "epoch": 0.6153386121988947,
      "grad_norm": 0.15622557699680328,
      "learning_rate": 7.949213359094674e-05,
      "loss": 0.7189,
      "mean_token_accuracy": 0.7903090454638004,
      "num_tokens": 213465537.0,
      "step": 6688
    },
    {
      "entropy": 0.809115007519722,
      "epoch": 0.615430618570336,
      "grad_norm": 0.16155385971069336,
      "learning_rate": 7.948906676480511e-05,
      "loss": 0.8165,
      "mean_token_accuracy": 0.7655225470662117,
      "num_tokens": 213497062.0,
      "step": 6689
    },
    {
      "entropy": 0.7216853126883507,
      "epoch": 0.6155226249417772,
      "grad_norm": 0.15770000219345093,
      "learning_rate": 7.948599993866348e-05,
      "loss": 0.7123,
      "mean_token_accuracy": 0.7844072617590427,
      "num_tokens": 213528920.0,
      "step": 6690
    },
    {
      "entropy": 0.805900314822793,
      "epoch": 0.6156146313132185,
      "grad_norm": 0.1637885570526123,
      "learning_rate": 7.948293311252186e-05,
      "loss": 0.8134,
      "mean_token_accuracy": 0.7633961923420429,
      "num_tokens": 213560477.0,
      "step": 6691
    },
    {
      "entropy": 0.6967924982309341,
      "epoch": 0.6157066376846596,
      "grad_norm": 0.15344934165477753,
      "learning_rate": 7.947986628638024e-05,
      "loss": 0.6799,
      "mean_token_accuracy": 0.7987430393695831,
      "num_tokens": 213591965.0,
      "step": 6692
    },
    {
      "entropy": 0.7206979300826788,
      "epoch": 0.6157986440561009,
      "grad_norm": 0.15963315963745117,
      "learning_rate": 7.94767994602386e-05,
      "loss": 0.7064,
      "mean_token_accuracy": 0.792344756424427,
      "num_tokens": 213623869.0,
      "step": 6693
    },
    {
      "entropy": 0.7227212283760309,
      "epoch": 0.6158906504275421,
      "grad_norm": 0.16821858286857605,
      "learning_rate": 7.947373263409697e-05,
      "loss": 0.7216,
      "mean_token_accuracy": 0.7940166145563126,
      "num_tokens": 213655579.0,
      "step": 6694
    },
    {
      "entropy": 0.7848414722830057,
      "epoch": 0.6159826567989833,
      "grad_norm": 0.16077552735805511,
      "learning_rate": 7.947066580795534e-05,
      "loss": 0.7661,
      "mean_token_accuracy": 0.7741433382034302,
      "num_tokens": 213687618.0,
      "step": 6695
    },
    {
      "entropy": 0.8260849863290787,
      "epoch": 0.6160746631704246,
      "grad_norm": 0.1560659110546112,
      "learning_rate": 7.946759898181372e-05,
      "loss": 0.8267,
      "mean_token_accuracy": 0.7596490271389484,
      "num_tokens": 213719067.0,
      "step": 6696
    },
    {
      "entropy": 0.7098764274269342,
      "epoch": 0.6161666695418657,
      "grad_norm": 0.1519925743341446,
      "learning_rate": 7.94645321556721e-05,
      "loss": 0.6994,
      "mean_token_accuracy": 0.7905419431626797,
      "num_tokens": 213750495.0,
      "step": 6697
    },
    {
      "entropy": 0.6637931130826473,
      "epoch": 0.616258675913307,
      "grad_norm": 0.1436561793088913,
      "learning_rate": 7.946146532953047e-05,
      "loss": 0.6518,
      "mean_token_accuracy": 0.8053735196590424,
      "num_tokens": 213782898.0,
      "step": 6698
    },
    {
      "entropy": 0.6287126056849957,
      "epoch": 0.6163506822847482,
      "grad_norm": 0.1506662368774414,
      "learning_rate": 7.945839850338884e-05,
      "loss": 0.6173,
      "mean_token_accuracy": 0.8148605041205883,
      "num_tokens": 213815043.0,
      "step": 6699
    },
    {
      "entropy": 0.7495789509266615,
      "epoch": 0.6164426886561895,
      "grad_norm": 0.1631496697664261,
      "learning_rate": 7.945533167724722e-05,
      "loss": 0.7455,
      "mean_token_accuracy": 0.7836047112941742,
      "num_tokens": 213846564.0,
      "step": 6700
    },
    {
      "entropy": 0.7329547945410013,
      "epoch": 0.6165346950276307,
      "grad_norm": 0.16513626277446747,
      "learning_rate": 7.945226485110559e-05,
      "loss": 0.7499,
      "mean_token_accuracy": 0.7799383290112019,
      "num_tokens": 213878611.0,
      "step": 6701
    },
    {
      "entropy": 0.7501567918807268,
      "epoch": 0.6166267013990718,
      "grad_norm": 0.15389791131019592,
      "learning_rate": 7.944919802496397e-05,
      "loss": 0.739,
      "mean_token_accuracy": 0.7811522856354713,
      "num_tokens": 213910693.0,
      "step": 6702
    },
    {
      "entropy": 0.8840440530329943,
      "epoch": 0.6167187077705131,
      "grad_norm": 0.1656116098165512,
      "learning_rate": 7.944613119882234e-05,
      "loss": 0.8804,
      "mean_token_accuracy": 0.7519614659249783,
      "num_tokens": 213942658.0,
      "step": 6703
    },
    {
      "entropy": 0.7811522874981165,
      "epoch": 0.6168107141419543,
      "grad_norm": 0.16462330520153046,
      "learning_rate": 7.944306437268072e-05,
      "loss": 0.7812,
      "mean_token_accuracy": 0.7719383202493191,
      "num_tokens": 213974784.0,
      "step": 6704
    },
    {
      "entropy": 0.8219312131404877,
      "epoch": 0.6169027205133956,
      "grad_norm": 0.1584533005952835,
      "learning_rate": 7.943999754653909e-05,
      "loss": 0.8231,
      "mean_token_accuracy": 0.7596194297075272,
      "num_tokens": 214006342.0,
      "step": 6705
    },
    {
      "entropy": 0.7447625692002475,
      "epoch": 0.6169947268848368,
      "grad_norm": 0.15236474573612213,
      "learning_rate": 7.943693072039746e-05,
      "loss": 0.7043,
      "mean_token_accuracy": 0.788930419832468,
      "num_tokens": 214037928.0,
      "step": 6706
    },
    {
      "entropy": 0.766594223678112,
      "epoch": 0.617086733256278,
      "grad_norm": 0.15700729191303253,
      "learning_rate": 7.943386389425584e-05,
      "loss": 0.7434,
      "mean_token_accuracy": 0.7784983292222023,
      "num_tokens": 214070030.0,
      "step": 6707
    },
    {
      "entropy": 0.8637153636664152,
      "epoch": 0.6171787396277192,
      "grad_norm": 0.17324534058570862,
      "learning_rate": 7.943079706811422e-05,
      "loss": 0.8466,
      "mean_token_accuracy": 0.756608173251152,
      "num_tokens": 214100445.0,
      "step": 6708
    },
    {
      "entropy": 0.7123164003714919,
      "epoch": 0.6172707459991604,
      "grad_norm": 0.1634848266839981,
      "learning_rate": 7.942773024197259e-05,
      "loss": 0.6874,
      "mean_token_accuracy": 0.7985459677875042,
      "num_tokens": 214132325.0,
      "step": 6709
    },
    {
      "entropy": 0.6442468278110027,
      "epoch": 0.6173627523706017,
      "grad_norm": 0.15998108685016632,
      "learning_rate": 7.942466341583095e-05,
      "loss": 0.6206,
      "mean_token_accuracy": 0.8197076730430126,
      "num_tokens": 214165083.0,
      "step": 6710
    },
    {
      "entropy": 0.7139595858752728,
      "epoch": 0.6174547587420429,
      "grad_norm": 0.16014178097248077,
      "learning_rate": 7.942159658968932e-05,
      "loss": 0.6953,
      "mean_token_accuracy": 0.7940150760114193,
      "num_tokens": 214196818.0,
      "step": 6711
    },
    {
      "entropy": 0.7301435172557831,
      "epoch": 0.6175467651134842,
      "grad_norm": 0.15928678214550018,
      "learning_rate": 7.941852976354772e-05,
      "loss": 0.7151,
      "mean_token_accuracy": 0.7922415211796761,
      "num_tokens": 214229150.0,
      "step": 6712
    },
    {
      "entropy": 0.8504659794270992,
      "epoch": 0.6176387714849253,
      "grad_norm": 0.16543731093406677,
      "learning_rate": 7.941546293740609e-05,
      "loss": 0.8433,
      "mean_token_accuracy": 0.7567204795777798,
      "num_tokens": 214260673.0,
      "step": 6713
    },
    {
      "entropy": 0.6974145155400038,
      "epoch": 0.6177307778563665,
      "grad_norm": 0.15512768924236298,
      "learning_rate": 7.941239611126445e-05,
      "loss": 0.6924,
      "mean_token_accuracy": 0.7959916442632675,
      "num_tokens": 214293313.0,
      "step": 6714
    },
    {
      "entropy": 0.7234057355672121,
      "epoch": 0.6178227842278078,
      "grad_norm": 0.15369302034378052,
      "learning_rate": 7.940932928512283e-05,
      "loss": 0.7305,
      "mean_token_accuracy": 0.7848366536200047,
      "num_tokens": 214325460.0,
      "step": 6715
    },
    {
      "entropy": 0.7263442724943161,
      "epoch": 0.617914790599249,
      "grad_norm": 0.15960194170475006,
      "learning_rate": 7.94062624589812e-05,
      "loss": 0.7135,
      "mean_token_accuracy": 0.7905375324189663,
      "num_tokens": 214357586.0,
      "step": 6716
    },
    {
      "entropy": 0.728422062471509,
      "epoch": 0.6180067969706903,
      "grad_norm": 0.15352429449558258,
      "learning_rate": 7.940319563283958e-05,
      "loss": 0.7258,
      "mean_token_accuracy": 0.7833450548350811,
      "num_tokens": 214389766.0,
      "step": 6717
    },
    {
      "entropy": 0.6801940575242043,
      "epoch": 0.6180988033421314,
      "grad_norm": 0.14927871525287628,
      "learning_rate": 7.940012880669795e-05,
      "loss": 0.6543,
      "mean_token_accuracy": 0.8064284063875675,
      "num_tokens": 214422347.0,
      "step": 6718
    },
    {
      "entropy": 0.7854325901716948,
      "epoch": 0.6181908097135727,
      "grad_norm": 0.1509893834590912,
      "learning_rate": 7.939706198055633e-05,
      "loss": 0.7622,
      "mean_token_accuracy": 0.774943120777607,
      "num_tokens": 214454703.0,
      "step": 6719
    },
    {
      "entropy": 0.5866229049861431,
      "epoch": 0.6182828160850139,
      "grad_norm": 0.14780539274215698,
      "learning_rate": 7.93939951544147e-05,
      "loss": 0.5651,
      "mean_token_accuracy": 0.828710712492466,
      "num_tokens": 214486537.0,
      "step": 6720
    },
    {
      "entropy": 0.7499909270554781,
      "epoch": 0.6183748224564551,
      "grad_norm": 0.15203291177749634,
      "learning_rate": 7.939092832827307e-05,
      "loss": 0.7336,
      "mean_token_accuracy": 0.7833393178880215,
      "num_tokens": 214518618.0,
      "step": 6721
    },
    {
      "entropy": 0.7636207491159439,
      "epoch": 0.6184668288278964,
      "grad_norm": 0.16210748255252838,
      "learning_rate": 7.938786150213145e-05,
      "loss": 0.7752,
      "mean_token_accuracy": 0.7688848040997982,
      "num_tokens": 214549843.0,
      "step": 6722
    },
    {
      "entropy": 0.7248553857207298,
      "epoch": 0.6185588351993375,
      "grad_norm": 0.15208932757377625,
      "learning_rate": 7.938479467598983e-05,
      "loss": 0.7199,
      "mean_token_accuracy": 0.7879243865609169,
      "num_tokens": 214582031.0,
      "step": 6723
    },
    {
      "entropy": 0.7467272058129311,
      "epoch": 0.6186508415707788,
      "grad_norm": 0.15441957116127014,
      "learning_rate": 7.93817278498482e-05,
      "loss": 0.7412,
      "mean_token_accuracy": 0.7828057669103146,
      "num_tokens": 214614395.0,
      "step": 6724
    },
    {
      "entropy": 0.8399667646735907,
      "epoch": 0.61874284794222,
      "grad_norm": 0.16618867218494415,
      "learning_rate": 7.937866102370657e-05,
      "loss": 0.8651,
      "mean_token_accuracy": 0.7485596351325512,
      "num_tokens": 214645854.0,
      "step": 6725
    },
    {
      "entropy": 0.8154331967234612,
      "epoch": 0.6188348543136613,
      "grad_norm": 0.16737845540046692,
      "learning_rate": 7.937559419756494e-05,
      "loss": 0.8157,
      "mean_token_accuracy": 0.761285588145256,
      "num_tokens": 214677162.0,
      "step": 6726
    },
    {
      "entropy": 0.7380770165473223,
      "epoch": 0.6189268606851025,
      "grad_norm": 0.16386739909648895,
      "learning_rate": 7.937252737142332e-05,
      "loss": 0.7067,
      "mean_token_accuracy": 0.7968107536435127,
      "num_tokens": 214708406.0,
      "step": 6727
    },
    {
      "entropy": 0.8599573206156492,
      "epoch": 0.6190188670565436,
      "grad_norm": 0.16392114758491516,
      "learning_rate": 7.93694605452817e-05,
      "loss": 0.8639,
      "mean_token_accuracy": 0.753508448600769,
      "num_tokens": 214740306.0,
      "step": 6728
    },
    {
      "entropy": 0.7544430010020733,
      "epoch": 0.6191108734279849,
      "grad_norm": 0.16276100277900696,
      "learning_rate": 7.936639371914007e-05,
      "loss": 0.7278,
      "mean_token_accuracy": 0.7864058837294579,
      "num_tokens": 214772163.0,
      "step": 6729
    },
    {
      "entropy": 0.8282833360135555,
      "epoch": 0.6192028797994261,
      "grad_norm": 0.16268663108348846,
      "learning_rate": 7.936332689299843e-05,
      "loss": 0.8267,
      "mean_token_accuracy": 0.7565673775970936,
      "num_tokens": 214803806.0,
      "step": 6730
    },
    {
      "entropy": 0.8758626598864794,
      "epoch": 0.6192948861708674,
      "grad_norm": 0.16480042040348053,
      "learning_rate": 7.936026006685682e-05,
      "loss": 0.8766,
      "mean_token_accuracy": 0.7488462403416634,
      "num_tokens": 214835753.0,
      "step": 6731
    },
    {
      "entropy": 0.7346856910735369,
      "epoch": 0.6193868925423086,
      "grad_norm": 0.17141617834568024,
      "learning_rate": 7.935719324071518e-05,
      "loss": 0.7042,
      "mean_token_accuracy": 0.7899126634001732,
      "num_tokens": 214867389.0,
      "step": 6732
    },
    {
      "entropy": 0.7739698495715857,
      "epoch": 0.6194788989137497,
      "grad_norm": 0.15612143278121948,
      "learning_rate": 7.935412641457357e-05,
      "loss": 0.7393,
      "mean_token_accuracy": 0.7784240245819092,
      "num_tokens": 214899036.0,
      "step": 6733
    },
    {
      "entropy": 0.7639559935778379,
      "epoch": 0.619570905285191,
      "grad_norm": 0.15187835693359375,
      "learning_rate": 7.935105958843193e-05,
      "loss": 0.7401,
      "mean_token_accuracy": 0.789454385638237,
      "num_tokens": 214931724.0,
      "step": 6734
    },
    {
      "entropy": 0.7359940651804209,
      "epoch": 0.6196629116566322,
      "grad_norm": 0.15330474078655243,
      "learning_rate": 7.934799276229031e-05,
      "loss": 0.6985,
      "mean_token_accuracy": 0.7952711135149002,
      "num_tokens": 214964342.0,
      "step": 6735
    },
    {
      "entropy": 0.7706716414541006,
      "epoch": 0.6197549180280735,
      "grad_norm": 0.1610095053911209,
      "learning_rate": 7.934492593614868e-05,
      "loss": 0.7558,
      "mean_token_accuracy": 0.7807918637990952,
      "num_tokens": 214995919.0,
      "step": 6736
    },
    {
      "entropy": 0.7308923397213221,
      "epoch": 0.6198469243995147,
      "grad_norm": 0.15977239608764648,
      "learning_rate": 7.934185911000705e-05,
      "loss": 0.7133,
      "mean_token_accuracy": 0.7927614599466324,
      "num_tokens": 215027909.0,
      "step": 6737
    },
    {
      "entropy": 0.8325748760253191,
      "epoch": 0.6199389307709559,
      "grad_norm": 0.16555453836917877,
      "learning_rate": 7.933879228386543e-05,
      "loss": 0.8449,
      "mean_token_accuracy": 0.7589770741760731,
      "num_tokens": 215059692.0,
      "step": 6738
    },
    {
      "entropy": 0.733066389337182,
      "epoch": 0.6200309371423971,
      "grad_norm": 0.14875823259353638,
      "learning_rate": 7.933572545772381e-05,
      "loss": 0.7406,
      "mean_token_accuracy": 0.7842312715947628,
      "num_tokens": 215092030.0,
      "step": 6739
    },
    {
      "entropy": 0.7746361326426268,
      "epoch": 0.6201229435138383,
      "grad_norm": 0.16020643711090088,
      "learning_rate": 7.933265863158218e-05,
      "loss": 0.7616,
      "mean_token_accuracy": 0.777302872389555,
      "num_tokens": 215123860.0,
      "step": 6740
    },
    {
      "entropy": 0.6669119019061327,
      "epoch": 0.6202149498852796,
      "grad_norm": 0.1550193727016449,
      "learning_rate": 7.932959180544055e-05,
      "loss": 0.6593,
      "mean_token_accuracy": 0.8025464415550232,
      "num_tokens": 215155860.0,
      "step": 6741
    },
    {
      "entropy": 0.7029689941555262,
      "epoch": 0.6203069562567208,
      "grad_norm": 0.15483775734901428,
      "learning_rate": 7.932652497929892e-05,
      "loss": 0.6777,
      "mean_token_accuracy": 0.7994899936020374,
      "num_tokens": 215187902.0,
      "step": 6742
    },
    {
      "entropy": 0.7680914085358381,
      "epoch": 0.620398962628162,
      "grad_norm": 0.1619875282049179,
      "learning_rate": 7.93234581531573e-05,
      "loss": 0.7639,
      "mean_token_accuracy": 0.7799112424254417,
      "num_tokens": 215219966.0,
      "step": 6743
    },
    {
      "entropy": 0.7040614988654852,
      "epoch": 0.6204909689996032,
      "grad_norm": 0.1626894623041153,
      "learning_rate": 7.932039132701568e-05,
      "loss": 0.7119,
      "mean_token_accuracy": 0.7879198007285595,
      "num_tokens": 215252562.0,
      "step": 6744
    },
    {
      "entropy": 0.8621592372655869,
      "epoch": 0.6205829753710445,
      "grad_norm": 0.16349738836288452,
      "learning_rate": 7.931732450087405e-05,
      "loss": 0.8755,
      "mean_token_accuracy": 0.7498246692121029,
      "num_tokens": 215284529.0,
      "step": 6745
    },
    {
      "entropy": 0.7614856641739607,
      "epoch": 0.6206749817424857,
      "grad_norm": 0.16177237033843994,
      "learning_rate": 7.931425767473243e-05,
      "loss": 0.789,
      "mean_token_accuracy": 0.770759079605341,
      "num_tokens": 215316655.0,
      "step": 6746
    },
    {
      "entropy": 0.8119928613305092,
      "epoch": 0.6207669881139269,
      "grad_norm": 0.15764383971691132,
      "learning_rate": 7.93111908485908e-05,
      "loss": 0.8014,
      "mean_token_accuracy": 0.766489140689373,
      "num_tokens": 215348512.0,
      "step": 6747
    },
    {
      "entropy": 0.8166375029832125,
      "epoch": 0.6208589944853681,
      "grad_norm": 0.16113634407520294,
      "learning_rate": 7.930812402244916e-05,
      "loss": 0.8123,
      "mean_token_accuracy": 0.7649001777172089,
      "num_tokens": 215380715.0,
      "step": 6748
    },
    {
      "entropy": 0.7794091608375311,
      "epoch": 0.6209510008568093,
      "grad_norm": 0.15605181455612183,
      "learning_rate": 7.930505719630755e-05,
      "loss": 0.7614,
      "mean_token_accuracy": 0.7754446528851986,
      "num_tokens": 215412513.0,
      "step": 6749
    },
    {
      "entropy": 0.8341744430363178,
      "epoch": 0.6210430072282506,
      "grad_norm": 0.15999920666217804,
      "learning_rate": 7.930199037016593e-05,
      "loss": 0.8241,
      "mean_token_accuracy": 0.7546057216823101,
      "num_tokens": 215443788.0,
      "step": 6750
    },
    {
      "entropy": 0.8022352326661348,
      "epoch": 0.6211350135996918,
      "grad_norm": 0.1583276093006134,
      "learning_rate": 7.92989235440243e-05,
      "loss": 0.7907,
      "mean_token_accuracy": 0.7699604369699955,
      "num_tokens": 215475652.0,
      "step": 6751
    },
    {
      "entropy": 0.6870643179863691,
      "epoch": 0.621227019971133,
      "grad_norm": 0.1608155220746994,
      "learning_rate": 7.929585671788266e-05,
      "loss": 0.6843,
      "mean_token_accuracy": 0.7946553379297256,
      "num_tokens": 215507477.0,
      "step": 6752
    },
    {
      "entropy": 0.7289262376725674,
      "epoch": 0.6213190263425742,
      "grad_norm": 0.15069471299648285,
      "learning_rate": 7.929278989174103e-05,
      "loss": 0.7045,
      "mean_token_accuracy": 0.7938489653170109,
      "num_tokens": 215539558.0,
      "step": 6753
    },
    {
      "entropy": 0.8418463692069054,
      "epoch": 0.6214110327140154,
      "grad_norm": 0.16787347197532654,
      "learning_rate": 7.928972306559943e-05,
      "loss": 0.8436,
      "mean_token_accuracy": 0.7530935406684875,
      "num_tokens": 215571690.0,
      "step": 6754
    },
    {
      "entropy": 0.8410144690424204,
      "epoch": 0.6215030390854567,
      "grad_norm": 0.15951718389987946,
      "learning_rate": 7.92866562394578e-05,
      "loss": 0.8278,
      "mean_token_accuracy": 0.7587366290390491,
      "num_tokens": 215603618.0,
      "step": 6755
    },
    {
      "entropy": 0.7827683184295893,
      "epoch": 0.6215950454568979,
      "grad_norm": 0.15764962136745453,
      "learning_rate": 7.928358941331616e-05,
      "loss": 0.7663,
      "mean_token_accuracy": 0.7765426561236382,
      "num_tokens": 215635813.0,
      "step": 6756
    },
    {
      "entropy": 0.7340504880994558,
      "epoch": 0.6216870518283392,
      "grad_norm": 0.15399989485740662,
      "learning_rate": 7.928052258717453e-05,
      "loss": 0.7163,
      "mean_token_accuracy": 0.7819871976971626,
      "num_tokens": 215666970.0,
      "step": 6757
    },
    {
      "entropy": 0.6991034634411335,
      "epoch": 0.6217790581997803,
      "grad_norm": 0.15240360796451569,
      "learning_rate": 7.927745576103291e-05,
      "loss": 0.6916,
      "mean_token_accuracy": 0.7939918711781502,
      "num_tokens": 215698972.0,
      "step": 6758
    },
    {
      "entropy": 0.7273062132298946,
      "epoch": 0.6218710645712215,
      "grad_norm": 0.15145254135131836,
      "learning_rate": 7.927438893489129e-05,
      "loss": 0.7148,
      "mean_token_accuracy": 0.7904459536075592,
      "num_tokens": 215731353.0,
      "step": 6759
    },
    {
      "entropy": 0.8560688812285662,
      "epoch": 0.6219630709426628,
      "grad_norm": 0.16980017721652985,
      "learning_rate": 7.927132210874966e-05,
      "loss": 0.8384,
      "mean_token_accuracy": 0.7537276819348335,
      "num_tokens": 215761936.0,
      "step": 6760
    },
    {
      "entropy": 0.768471222370863,
      "epoch": 0.622055077314104,
      "grad_norm": 0.15603050589561462,
      "learning_rate": 7.926825528260803e-05,
      "loss": 0.7706,
      "mean_token_accuracy": 0.7728297635912895,
      "num_tokens": 215794178.0,
      "step": 6761
    },
    {
      "entropy": 0.7437111400067806,
      "epoch": 0.6221470836855453,
      "grad_norm": 0.16461385786533356,
      "learning_rate": 7.926518845646641e-05,
      "loss": 0.7297,
      "mean_token_accuracy": 0.7809272930026054,
      "num_tokens": 215826106.0,
      "step": 6762
    },
    {
      "entropy": 0.8479232639074326,
      "epoch": 0.6222390900569864,
      "grad_norm": 0.15566486120224,
      "learning_rate": 7.926212163032478e-05,
      "loss": 0.8447,
      "mean_token_accuracy": 0.758796114474535,
      "num_tokens": 215858450.0,
      "step": 6763
    },
    {
      "entropy": 0.8511785808950663,
      "epoch": 0.6223310964284277,
      "grad_norm": 0.16404849290847778,
      "learning_rate": 7.925905480418316e-05,
      "loss": 0.8624,
      "mean_token_accuracy": 0.7485533505678177,
      "num_tokens": 215890660.0,
      "step": 6764
    },
    {
      "entropy": 0.7507437039166689,
      "epoch": 0.6224231027998689,
      "grad_norm": 0.15571902692317963,
      "learning_rate": 7.925598797804153e-05,
      "loss": 0.735,
      "mean_token_accuracy": 0.7806402035057545,
      "num_tokens": 215922618.0,
      "step": 6765
    },
    {
      "entropy": 0.900667816400528,
      "epoch": 0.6225151091713101,
      "grad_norm": 0.15668095648288727,
      "learning_rate": 7.925292115189991e-05,
      "loss": 0.8753,
      "mean_token_accuracy": 0.7424122765660286,
      "num_tokens": 215954544.0,
      "step": 6766
    },
    {
      "entropy": 0.757878364995122,
      "epoch": 0.6226071155427514,
      "grad_norm": 0.15232200920581818,
      "learning_rate": 7.924985432575828e-05,
      "loss": 0.7227,
      "mean_token_accuracy": 0.7854220196604729,
      "num_tokens": 215986205.0,
      "step": 6767
    },
    {
      "entropy": 0.7610921822488308,
      "epoch": 0.6226991219141925,
      "grad_norm": 0.154471293091774,
      "learning_rate": 7.924678749961664e-05,
      "loss": 0.7488,
      "mean_token_accuracy": 0.7755735032260418,
      "num_tokens": 216017767.0,
      "step": 6768
    },
    {
      "entropy": 0.8493087366223335,
      "epoch": 0.6227911282856338,
      "grad_norm": 0.16567689180374146,
      "learning_rate": 7.924372067347503e-05,
      "loss": 0.829,
      "mean_token_accuracy": 0.7559348531067371,
      "num_tokens": 216049538.0,
      "step": 6769
    },
    {
      "entropy": 0.7359444200992584,
      "epoch": 0.622883134657075,
      "grad_norm": 0.16701947152614594,
      "learning_rate": 7.92406538473334e-05,
      "loss": 0.7293,
      "mean_token_accuracy": 0.7925218380987644,
      "num_tokens": 216082047.0,
      "step": 6770
    },
    {
      "entropy": 0.7458947226405144,
      "epoch": 0.6229751410285163,
      "grad_norm": 0.15825822949409485,
      "learning_rate": 7.923758702119177e-05,
      "loss": 0.7342,
      "mean_token_accuracy": 0.7826673537492752,
      "num_tokens": 216114009.0,
      "step": 6771
    },
    {
      "entropy": 0.7519962266087532,
      "epoch": 0.6230671473999575,
      "grad_norm": 0.15698567032814026,
      "learning_rate": 7.923452019505014e-05,
      "loss": 0.7448,
      "mean_token_accuracy": 0.7802510559558868,
      "num_tokens": 216146090.0,
      "step": 6772
    },
    {
      "entropy": 0.6741875223815441,
      "epoch": 0.6231591537713986,
      "grad_norm": 0.1585148721933365,
      "learning_rate": 7.923145336890852e-05,
      "loss": 0.6732,
      "mean_token_accuracy": 0.7993123009800911,
      "num_tokens": 216178167.0,
      "step": 6773
    },
    {
      "entropy": 0.7290232963860035,
      "epoch": 0.6232511601428399,
      "grad_norm": 0.15242509543895721,
      "learning_rate": 7.922838654276689e-05,
      "loss": 0.722,
      "mean_token_accuracy": 0.7827555313706398,
      "num_tokens": 216210340.0,
      "step": 6774
    },
    {
      "entropy": 0.8048685397952795,
      "epoch": 0.6233431665142811,
      "grad_norm": 0.165812149643898,
      "learning_rate": 7.922531971662527e-05,
      "loss": 0.8099,
      "mean_token_accuracy": 0.7663775086402893,
      "num_tokens": 216242438.0,
      "step": 6775
    },
    {
      "entropy": 0.815949697047472,
      "epoch": 0.6234351728857224,
      "grad_norm": 0.16318587958812714,
      "learning_rate": 7.922225289048364e-05,
      "loss": 0.8019,
      "mean_token_accuracy": 0.7687718905508518,
      "num_tokens": 216274538.0,
      "step": 6776
    },
    {
      "entropy": 0.6879681898280978,
      "epoch": 0.6235271792571636,
      "grad_norm": 0.15095527470111847,
      "learning_rate": 7.921918606434202e-05,
      "loss": 0.6641,
      "mean_token_accuracy": 0.8015441931784153,
      "num_tokens": 216306838.0,
      "step": 6777
    },
    {
      "entropy": 0.8062553387135267,
      "epoch": 0.6236191856286047,
      "grad_norm": 0.16492879390716553,
      "learning_rate": 7.921611923820039e-05,
      "loss": 0.8134,
      "mean_token_accuracy": 0.7674841918051243,
      "num_tokens": 216338691.0,
      "step": 6778
    },
    {
      "entropy": 0.7469684053212404,
      "epoch": 0.623711192000046,
      "grad_norm": 0.16101548075675964,
      "learning_rate": 7.921305241205876e-05,
      "loss": 0.7408,
      "mean_token_accuracy": 0.7790977358818054,
      "num_tokens": 216369809.0,
      "step": 6779
    },
    {
      "entropy": 0.7081315964460373,
      "epoch": 0.6238031983714872,
      "grad_norm": 0.17062845826148987,
      "learning_rate": 7.920998558591714e-05,
      "loss": 0.7001,
      "mean_token_accuracy": 0.7941737547516823,
      "num_tokens": 216401922.0,
      "step": 6780
    },
    {
      "entropy": 0.7321723215281963,
      "epoch": 0.6238952047429285,
      "grad_norm": 0.1679207980632782,
      "learning_rate": 7.920691875977552e-05,
      "loss": 0.7281,
      "mean_token_accuracy": 0.7900229394435883,
      "num_tokens": 216434023.0,
      "step": 6781
    },
    {
      "entropy": 0.6677673235535622,
      "epoch": 0.6239872111143697,
      "grad_norm": 0.1644304096698761,
      "learning_rate": 7.920385193363389e-05,
      "loss": 0.6849,
      "mean_token_accuracy": 0.7998153232038021,
      "num_tokens": 216465734.0,
      "step": 6782
    },
    {
      "entropy": 0.7502274494618177,
      "epoch": 0.6240792174858109,
      "grad_norm": 0.15578198432922363,
      "learning_rate": 7.920078510749226e-05,
      "loss": 0.7497,
      "mean_token_accuracy": 0.7836160399019718,
      "num_tokens": 216497723.0,
      "step": 6783
    },
    {
      "entropy": 0.7417934341356158,
      "epoch": 0.6241712238572521,
      "grad_norm": 0.15463289618492126,
      "learning_rate": 7.919771828135062e-05,
      "loss": 0.7271,
      "mean_token_accuracy": 0.7892079576849937,
      "num_tokens": 216529645.0,
      "step": 6784
    },
    {
      "entropy": 0.6994147878140211,
      "epoch": 0.6242632302286933,
      "grad_norm": 0.15036335587501526,
      "learning_rate": 7.9194651455209e-05,
      "loss": 0.6546,
      "mean_token_accuracy": 0.7993792966008186,
      "num_tokens": 216561097.0,
      "step": 6785
    },
    {
      "entropy": 0.7217270713299513,
      "epoch": 0.6243552366001346,
      "grad_norm": 0.15878333151340485,
      "learning_rate": 7.919158462906739e-05,
      "loss": 0.6912,
      "mean_token_accuracy": 0.7951253950595856,
      "num_tokens": 216592793.0,
      "step": 6786
    },
    {
      "entropy": 0.7276176642626524,
      "epoch": 0.6244472429715758,
      "grad_norm": 0.14991648495197296,
      "learning_rate": 7.918851780292576e-05,
      "loss": 0.711,
      "mean_token_accuracy": 0.7878313548862934,
      "num_tokens": 216624558.0,
      "step": 6787
    },
    {
      "entropy": 0.7211491651833057,
      "epoch": 0.624539249343017,
      "grad_norm": 0.1579214334487915,
      "learning_rate": 7.918545097678412e-05,
      "loss": 0.6942,
      "mean_token_accuracy": 0.7944667860865593,
      "num_tokens": 216655880.0,
      "step": 6788
    },
    {
      "entropy": 0.7903389651328325,
      "epoch": 0.6246312557144582,
      "grad_norm": 0.16445203125476837,
      "learning_rate": 7.91823841506425e-05,
      "loss": 0.7642,
      "mean_token_accuracy": 0.7738910727202892,
      "num_tokens": 216687643.0,
      "step": 6789
    },
    {
      "entropy": 0.8529907502233982,
      "epoch": 0.6247232620858995,
      "grad_norm": 0.165459543466568,
      "learning_rate": 7.917931732450087e-05,
      "loss": 0.8491,
      "mean_token_accuracy": 0.7620854116976261,
      "num_tokens": 216719832.0,
      "step": 6790
    },
    {
      "entropy": 0.7955469284206629,
      "epoch": 0.6248152684573407,
      "grad_norm": 0.161915123462677,
      "learning_rate": 7.917625049835925e-05,
      "loss": 0.7935,
      "mean_token_accuracy": 0.7717696614563465,
      "num_tokens": 216752600.0,
      "step": 6791
    },
    {
      "entropy": 0.841705746948719,
      "epoch": 0.624907274828782,
      "grad_norm": 0.16632582247257233,
      "learning_rate": 7.917318367221762e-05,
      "loss": 0.8446,
      "mean_token_accuracy": 0.7487368285655975,
      "num_tokens": 216783251.0,
      "step": 6792
    },
    {
      "entropy": 0.7894716449081898,
      "epoch": 0.6249992812002231,
      "grad_norm": 0.16795331239700317,
      "learning_rate": 7.9170116846076e-05,
      "loss": 0.7955,
      "mean_token_accuracy": 0.7663572318851948,
      "num_tokens": 216814847.0,
      "step": 6793
    },
    {
      "entropy": 0.8296987824141979,
      "epoch": 0.6250912875716643,
      "grad_norm": 0.15142075717449188,
      "learning_rate": 7.916705001993437e-05,
      "loss": 0.8173,
      "mean_token_accuracy": 0.7611150220036507,
      "num_tokens": 216846864.0,
      "step": 6794
    },
    {
      "entropy": 0.7650800440460443,
      "epoch": 0.6251832939431056,
      "grad_norm": 0.1654403656721115,
      "learning_rate": 7.916398319379274e-05,
      "loss": 0.7548,
      "mean_token_accuracy": 0.7747060060501099,
      "num_tokens": 216877779.0,
      "step": 6795
    },
    {
      "entropy": 0.7189792711287737,
      "epoch": 0.6252753003145468,
      "grad_norm": 0.15077261626720428,
      "learning_rate": 7.916091636765112e-05,
      "loss": 0.7079,
      "mean_token_accuracy": 0.7914016470313072,
      "num_tokens": 216909830.0,
      "step": 6796
    },
    {
      "entropy": 0.8146858550608158,
      "epoch": 0.6253673066859881,
      "grad_norm": 0.15654398500919342,
      "learning_rate": 7.91578495415095e-05,
      "loss": 0.7922,
      "mean_token_accuracy": 0.769200723618269,
      "num_tokens": 216942069.0,
      "step": 6797
    },
    {
      "entropy": 0.7513052355498075,
      "epoch": 0.6254593130574292,
      "grad_norm": 0.15517500042915344,
      "learning_rate": 7.915478271536787e-05,
      "loss": 0.7288,
      "mean_token_accuracy": 0.7860465385019779,
      "num_tokens": 216973631.0,
      "step": 6798
    },
    {
      "entropy": 0.7871745154261589,
      "epoch": 0.6255513194288704,
      "grad_norm": 0.15914158523082733,
      "learning_rate": 7.915171588922624e-05,
      "loss": 0.7681,
      "mean_token_accuracy": 0.7780463807284832,
      "num_tokens": 217006192.0,
      "step": 6799
    },
    {
      "entropy": 0.7305289655923843,
      "epoch": 0.6256433258003117,
      "grad_norm": 0.1507120281457901,
      "learning_rate": 7.914864906308462e-05,
      "loss": 0.715,
      "mean_token_accuracy": 0.788003608584404,
      "num_tokens": 217037904.0,
      "step": 6800
    },
    {
      "entropy": 0.7408063225448132,
      "epoch": 0.6257353321717529,
      "grad_norm": 0.15471401810646057,
      "learning_rate": 7.9145582236943e-05,
      "loss": 0.73,
      "mean_token_accuracy": 0.779632568359375,
      "num_tokens": 217069413.0,
      "step": 6801
    },
    {
      "entropy": 0.9204970262944698,
      "epoch": 0.6258273385431942,
      "grad_norm": 0.18115924298763275,
      "learning_rate": 7.914251541080137e-05,
      "loss": 0.9274,
      "mean_token_accuracy": 0.7329103872179985,
      "num_tokens": 217100663.0,
      "step": 6802
    },
    {
      "entropy": 0.7298220545053482,
      "epoch": 0.6259193449146353,
      "grad_norm": 0.1613893061876297,
      "learning_rate": 7.913944858465974e-05,
      "loss": 0.7028,
      "mean_token_accuracy": 0.7888923026621342,
      "num_tokens": 217132094.0,
      "step": 6803
    },
    {
      "entropy": 0.7790422085672617,
      "epoch": 0.6260113512860765,
      "grad_norm": 0.16577018797397614,
      "learning_rate": 7.913638175851812e-05,
      "loss": 0.7909,
      "mean_token_accuracy": 0.7707245722413063,
      "num_tokens": 217164011.0,
      "step": 6804
    },
    {
      "entropy": 0.6506985127925873,
      "epoch": 0.6261033576575178,
      "grad_norm": 0.15459588170051575,
      "learning_rate": 7.913331493237649e-05,
      "loss": 0.6538,
      "mean_token_accuracy": 0.8059298247098923,
      "num_tokens": 217195334.0,
      "step": 6805
    },
    {
      "entropy": 0.7628348264843225,
      "epoch": 0.626195364028959,
      "grad_norm": 0.16382087767124176,
      "learning_rate": 7.913024810623487e-05,
      "loss": 0.7707,
      "mean_token_accuracy": 0.7758900672197342,
      "num_tokens": 217226718.0,
      "step": 6806
    },
    {
      "entropy": 0.7917848154902458,
      "epoch": 0.6262873704004003,
      "grad_norm": 0.15672647953033447,
      "learning_rate": 7.912718128009323e-05,
      "loss": 0.783,
      "mean_token_accuracy": 0.7749223932623863,
      "num_tokens": 217257933.0,
      "step": 6807
    },
    {
      "entropy": 0.718776598572731,
      "epoch": 0.6263793767718414,
      "grad_norm": 0.16079722344875336,
      "learning_rate": 7.912411445395162e-05,
      "loss": 0.717,
      "mean_token_accuracy": 0.788719467818737,
      "num_tokens": 217289259.0,
      "step": 6808
    },
    {
      "entropy": 0.7844743058085442,
      "epoch": 0.6264713831432827,
      "grad_norm": 0.1588796079158783,
      "learning_rate": 7.912104762780998e-05,
      "loss": 0.7585,
      "mean_token_accuracy": 0.7767906226217747,
      "num_tokens": 217321444.0,
      "step": 6809
    },
    {
      "entropy": 0.6078992709517479,
      "epoch": 0.6265633895147239,
      "grad_norm": 0.14387331902980804,
      "learning_rate": 7.911798080166835e-05,
      "loss": 0.5859,
      "mean_token_accuracy": 0.8216252252459526,
      "num_tokens": 217353204.0,
      "step": 6810
    },
    {
      "entropy": 0.7999612856656313,
      "epoch": 0.6266553958861651,
      "grad_norm": 0.15916189551353455,
      "learning_rate": 7.911491397552673e-05,
      "loss": 0.7823,
      "mean_token_accuracy": 0.7695694677531719,
      "num_tokens": 217384448.0,
      "step": 6811
    },
    {
      "entropy": 0.7436028923839331,
      "epoch": 0.6267474022576064,
      "grad_norm": 0.1641596108675003,
      "learning_rate": 7.911184714938511e-05,
      "loss": 0.7348,
      "mean_token_accuracy": 0.7857553139328957,
      "num_tokens": 217416748.0,
      "step": 6812
    },
    {
      "entropy": 0.7308945395052433,
      "epoch": 0.6268394086290475,
      "grad_norm": 0.1633182168006897,
      "learning_rate": 7.910878032324348e-05,
      "loss": 0.7153,
      "mean_token_accuracy": 0.7894206680357456,
      "num_tokens": 217448425.0,
      "step": 6813
    },
    {
      "entropy": 0.6952079758048058,
      "epoch": 0.6269314150004888,
      "grad_norm": 0.1505056470632553,
      "learning_rate": 7.910571349710185e-05,
      "loss": 0.6746,
      "mean_token_accuracy": 0.7966642156243324,
      "num_tokens": 217480411.0,
      "step": 6814
    },
    {
      "entropy": 0.8412358984351158,
      "epoch": 0.62702342137193,
      "grad_norm": 0.16115662455558777,
      "learning_rate": 7.910264667096022e-05,
      "loss": 0.8475,
      "mean_token_accuracy": 0.7538891583681107,
      "num_tokens": 217512734.0,
      "step": 6815
    },
    {
      "entropy": 0.7000873442739248,
      "epoch": 0.6271154277433713,
      "grad_norm": 0.1486492156982422,
      "learning_rate": 7.90995798448186e-05,
      "loss": 0.6867,
      "mean_token_accuracy": 0.7972451560199261,
      "num_tokens": 217544674.0,
      "step": 6816
    },
    {
      "entropy": 0.7547997646033764,
      "epoch": 0.6272074341148125,
      "grad_norm": 0.1567060500383377,
      "learning_rate": 7.909651301867698e-05,
      "loss": 0.7499,
      "mean_token_accuracy": 0.7802022285759449,
      "num_tokens": 217576721.0,
      "step": 6817
    },
    {
      "entropy": 0.8410606849938631,
      "epoch": 0.6272994404862536,
      "grad_norm": 0.15369021892547607,
      "learning_rate": 7.909344619253535e-05,
      "loss": 0.8253,
      "mean_token_accuracy": 0.7601376883685589,
      "num_tokens": 217608912.0,
      "step": 6818
    },
    {
      "entropy": 0.8698541168123484,
      "epoch": 0.6273914468576949,
      "grad_norm": 0.16305553913116455,
      "learning_rate": 7.909037936639372e-05,
      "loss": 0.8457,
      "mean_token_accuracy": 0.7591158896684647,
      "num_tokens": 217641266.0,
      "step": 6819
    },
    {
      "entropy": 0.7007640991359949,
      "epoch": 0.6274834532291361,
      "grad_norm": 0.14975759387016296,
      "learning_rate": 7.90873125402521e-05,
      "loss": 0.7026,
      "mean_token_accuracy": 0.7920554205775261,
      "num_tokens": 217673297.0,
      "step": 6820
    },
    {
      "entropy": 0.7700629774481058,
      "epoch": 0.6275754596005774,
      "grad_norm": 0.14894677698612213,
      "learning_rate": 7.908424571411047e-05,
      "loss": 0.7544,
      "mean_token_accuracy": 0.7757166363298893,
      "num_tokens": 217704569.0,
      "step": 6821
    },
    {
      "entropy": 0.597959715873003,
      "epoch": 0.6276674659720186,
      "grad_norm": 0.1555832475423813,
      "learning_rate": 7.908117888796885e-05,
      "loss": 0.5834,
      "mean_token_accuracy": 0.821891687810421,
      "num_tokens": 217736630.0,
      "step": 6822
    },
    {
      "entropy": 0.7467452492564917,
      "epoch": 0.6277594723434597,
      "grad_norm": 0.1544766128063202,
      "learning_rate": 7.907811206182722e-05,
      "loss": 0.726,
      "mean_token_accuracy": 0.7832550629973412,
      "num_tokens": 217768532.0,
      "step": 6823
    },
    {
      "entropy": 0.7202938888221979,
      "epoch": 0.627851478714901,
      "grad_norm": 0.14746391773223877,
      "learning_rate": 7.90750452356856e-05,
      "loss": 0.6838,
      "mean_token_accuracy": 0.7998668737709522,
      "num_tokens": 217801249.0,
      "step": 6824
    },
    {
      "entropy": 0.8528883941471577,
      "epoch": 0.6279434850863422,
      "grad_norm": 0.16646473109722137,
      "learning_rate": 7.907197840954396e-05,
      "loss": 0.862,
      "mean_token_accuracy": 0.7483476661145687,
      "num_tokens": 217833049.0,
      "step": 6825
    },
    {
      "entropy": 0.7985125426203012,
      "epoch": 0.6280354914577835,
      "grad_norm": 0.1564498096704483,
      "learning_rate": 7.906891158340233e-05,
      "loss": 0.7872,
      "mean_token_accuracy": 0.7687594704329967,
      "num_tokens": 217865500.0,
      "step": 6826
    },
    {
      "entropy": 0.626744874753058,
      "epoch": 0.6281274978292247,
      "grad_norm": 0.1621852070093155,
      "learning_rate": 7.906584475726071e-05,
      "loss": 0.6253,
      "mean_token_accuracy": 0.8161138668656349,
      "num_tokens": 217898155.0,
      "step": 6827
    },
    {
      "entropy": 0.7434835694730282,
      "epoch": 0.6282195042006659,
      "grad_norm": 0.1637037843465805,
      "learning_rate": 7.90627779311191e-05,
      "loss": 0.7473,
      "mean_token_accuracy": 0.7808829993009567,
      "num_tokens": 217930501.0,
      "step": 6828
    },
    {
      "entropy": 0.8433907041326165,
      "epoch": 0.6283115105721071,
      "grad_norm": 0.16378933191299438,
      "learning_rate": 7.905971110497746e-05,
      "loss": 0.8367,
      "mean_token_accuracy": 0.7559948936104774,
      "num_tokens": 217962416.0,
      "step": 6829
    },
    {
      "entropy": 0.6484341975301504,
      "epoch": 0.6284035169435483,
      "grad_norm": 0.1571919173002243,
      "learning_rate": 7.905664427883583e-05,
      "loss": 0.6461,
      "mean_token_accuracy": 0.8007880486547947,
      "num_tokens": 217994561.0,
      "step": 6830
    },
    {
      "entropy": 0.7759687751531601,
      "epoch": 0.6284955233149896,
      "grad_norm": 0.16274134814739227,
      "learning_rate": 7.90535774526942e-05,
      "loss": 0.7651,
      "mean_token_accuracy": 0.7778013944625854,
      "num_tokens": 218026490.0,
      "step": 6831
    },
    {
      "entropy": 0.6445892713963985,
      "epoch": 0.6285875296864308,
      "grad_norm": 0.15866823494434357,
      "learning_rate": 7.90505106265526e-05,
      "loss": 0.633,
      "mean_token_accuracy": 0.8145731277763844,
      "num_tokens": 218059056.0,
      "step": 6832
    },
    {
      "entropy": 0.7827423438429832,
      "epoch": 0.628679536057872,
      "grad_norm": 0.15436352789402008,
      "learning_rate": 7.904744380041096e-05,
      "loss": 0.7584,
      "mean_token_accuracy": 0.7757160887122154,
      "num_tokens": 218091243.0,
      "step": 6833
    },
    {
      "entropy": 0.8432838656008244,
      "epoch": 0.6287715424293132,
      "grad_norm": 0.16385851800441742,
      "learning_rate": 7.904437697426933e-05,
      "loss": 0.8288,
      "mean_token_accuracy": 0.7540723085403442,
      "num_tokens": 218123429.0,
      "step": 6834
    },
    {
      "entropy": 0.754454730078578,
      "epoch": 0.6288635488007545,
      "grad_norm": 0.1590820997953415,
      "learning_rate": 7.904131014812771e-05,
      "loss": 0.7411,
      "mean_token_accuracy": 0.7838918901979923,
      "num_tokens": 218155631.0,
      "step": 6835
    },
    {
      "entropy": 0.7973548173904419,
      "epoch": 0.6289555551721957,
      "grad_norm": 0.16178680956363678,
      "learning_rate": 7.903824332198608e-05,
      "loss": 0.8031,
      "mean_token_accuracy": 0.7671519890427589,
      "num_tokens": 218187805.0,
      "step": 6836
    },
    {
      "entropy": 0.7836390100419521,
      "epoch": 0.629047561543637,
      "grad_norm": 0.15806341171264648,
      "learning_rate": 7.903517649584446e-05,
      "loss": 0.7675,
      "mean_token_accuracy": 0.7758978009223938,
      "num_tokens": 218219142.0,
      "step": 6837
    },
    {
      "entropy": 0.8752648010849953,
      "epoch": 0.6291395679150781,
      "grad_norm": 0.16525588929653168,
      "learning_rate": 7.903210966970283e-05,
      "loss": 0.8566,
      "mean_token_accuracy": 0.7486743927001953,
      "num_tokens": 218250808.0,
      "step": 6838
    },
    {
      "entropy": 0.7539007645100355,
      "epoch": 0.6292315742865193,
      "grad_norm": 0.15580390393733978,
      "learning_rate": 7.902904284356121e-05,
      "loss": 0.7359,
      "mean_token_accuracy": 0.783685214817524,
      "num_tokens": 218281939.0,
      "step": 6839
    },
    {
      "entropy": 0.6400251686573029,
      "epoch": 0.6293235806579606,
      "grad_norm": 0.14696282148361206,
      "learning_rate": 7.902597601741958e-05,
      "loss": 0.6081,
      "mean_token_accuracy": 0.8175961039960384,
      "num_tokens": 218314295.0,
      "step": 6840
    },
    {
      "entropy": 0.78320736810565,
      "epoch": 0.6294155870294018,
      "grad_norm": 0.167501300573349,
      "learning_rate": 7.902290919127795e-05,
      "loss": 0.7691,
      "mean_token_accuracy": 0.774752888828516,
      "num_tokens": 218345953.0,
      "step": 6841
    },
    {
      "entropy": 0.7910546995699406,
      "epoch": 0.6295075934008431,
      "grad_norm": 0.15890119969844818,
      "learning_rate": 7.901984236513633e-05,
      "loss": 0.7779,
      "mean_token_accuracy": 0.7720214873552322,
      "num_tokens": 218378461.0,
      "step": 6842
    },
    {
      "entropy": 0.7603123914450407,
      "epoch": 0.6295995997722842,
      "grad_norm": 0.15796591341495514,
      "learning_rate": 7.901677553899471e-05,
      "loss": 0.7605,
      "mean_token_accuracy": 0.7767181545495987,
      "num_tokens": 218410151.0,
      "step": 6843
    },
    {
      "entropy": 0.8226458206772804,
      "epoch": 0.6296916061437254,
      "grad_norm": 0.17515899240970612,
      "learning_rate": 7.901370871285308e-05,
      "loss": 0.8361,
      "mean_token_accuracy": 0.7612378410995007,
      "num_tokens": 218442120.0,
      "step": 6844
    },
    {
      "entropy": 0.784871868789196,
      "epoch": 0.6297836125151667,
      "grad_norm": 0.15517963469028473,
      "learning_rate": 7.901064188671144e-05,
      "loss": 0.7964,
      "mean_token_accuracy": 0.7674767971038818,
      "num_tokens": 218474713.0,
      "step": 6845
    },
    {
      "entropy": 0.752686619758606,
      "epoch": 0.6298756188866079,
      "grad_norm": 0.16173917055130005,
      "learning_rate": 7.900757506056981e-05,
      "loss": 0.7377,
      "mean_token_accuracy": 0.7814323343336582,
      "num_tokens": 218505943.0,
      "step": 6846
    },
    {
      "entropy": 0.8601666633039713,
      "epoch": 0.6299676252580492,
      "grad_norm": 0.16972991824150085,
      "learning_rate": 7.90045082344282e-05,
      "loss": 0.8801,
      "mean_token_accuracy": 0.7426922656595707,
      "num_tokens": 218538292.0,
      "step": 6847
    },
    {
      "entropy": 0.7626979779452085,
      "epoch": 0.6300596316294903,
      "grad_norm": 0.15815889835357666,
      "learning_rate": 7.900144140828657e-05,
      "loss": 0.7693,
      "mean_token_accuracy": 0.7712207846343517,
      "num_tokens": 218570572.0,
      "step": 6848
    },
    {
      "entropy": 0.650795528665185,
      "epoch": 0.6301516380009315,
      "grad_norm": 0.15809403359889984,
      "learning_rate": 7.899837458214494e-05,
      "loss": 0.6434,
      "mean_token_accuracy": 0.8113416954874992,
      "num_tokens": 218602804.0,
      "step": 6849
    },
    {
      "entropy": 0.8018024731427431,
      "epoch": 0.6302436443723728,
      "grad_norm": 0.16238464415073395,
      "learning_rate": 7.899530775600331e-05,
      "loss": 0.805,
      "mean_token_accuracy": 0.7676852494478226,
      "num_tokens": 218634655.0,
      "step": 6850
    },
    {
      "entropy": 0.6764072310179472,
      "epoch": 0.630335650743814,
      "grad_norm": 0.15016017854213715,
      "learning_rate": 7.899224092986169e-05,
      "loss": 0.6533,
      "mean_token_accuracy": 0.8073914498090744,
      "num_tokens": 218667149.0,
      "step": 6851
    },
    {
      "entropy": 0.7165358234196901,
      "epoch": 0.6304276571152553,
      "grad_norm": 0.15035295486450195,
      "learning_rate": 7.898917410372006e-05,
      "loss": 0.7016,
      "mean_token_accuracy": 0.7940278425812721,
      "num_tokens": 218699733.0,
      "step": 6852
    },
    {
      "entropy": 0.5984758194535971,
      "epoch": 0.6305196634866964,
      "grad_norm": 0.14560063183307648,
      "learning_rate": 7.898610727757844e-05,
      "loss": 0.5822,
      "mean_token_accuracy": 0.8264394551515579,
      "num_tokens": 218732331.0,
      "step": 6853
    },
    {
      "entropy": 0.6276362231001258,
      "epoch": 0.6306116698581377,
      "grad_norm": 0.1614423245191574,
      "learning_rate": 7.898304045143681e-05,
      "loss": 0.6151,
      "mean_token_accuracy": 0.8157286271452904,
      "num_tokens": 218765041.0,
      "step": 6854
    },
    {
      "entropy": 0.730513259768486,
      "epoch": 0.6307036762295789,
      "grad_norm": 0.1567246913909912,
      "learning_rate": 7.897997362529519e-05,
      "loss": 0.7144,
      "mean_token_accuracy": 0.78877829387784,
      "num_tokens": 218797398.0,
      "step": 6855
    },
    {
      "entropy": 0.7758416887372732,
      "epoch": 0.6307956826010201,
      "grad_norm": 0.15309035778045654,
      "learning_rate": 7.897690679915356e-05,
      "loss": 0.7486,
      "mean_token_accuracy": 0.7813841924071312,
      "num_tokens": 218829429.0,
      "step": 6856
    },
    {
      "entropy": 0.711481976788491,
      "epoch": 0.6308876889724614,
      "grad_norm": 0.1503920555114746,
      "learning_rate": 7.897383997301193e-05,
      "loss": 0.689,
      "mean_token_accuracy": 0.7956329733133316,
      "num_tokens": 218860643.0,
      "step": 6857
    },
    {
      "entropy": 0.7894313400611281,
      "epoch": 0.6309796953439025,
      "grad_norm": 0.15467482805252075,
      "learning_rate": 7.897077314687031e-05,
      "loss": 0.7544,
      "mean_token_accuracy": 0.7753642685711384,
      "num_tokens": 218892131.0,
      "step": 6858
    },
    {
      "entropy": 0.7058479264378548,
      "epoch": 0.6310717017153438,
      "grad_norm": 0.15322549641132355,
      "learning_rate": 7.896770632072869e-05,
      "loss": 0.6819,
      "mean_token_accuracy": 0.7979259341955185,
      "num_tokens": 218923818.0,
      "step": 6859
    },
    {
      "entropy": 0.807581352069974,
      "epoch": 0.631163708086785,
      "grad_norm": 0.17349445819854736,
      "learning_rate": 7.896463949458706e-05,
      "loss": 0.8226,
      "mean_token_accuracy": 0.763271376490593,
      "num_tokens": 218955915.0,
      "step": 6860
    },
    {
      "entropy": 0.763004332780838,
      "epoch": 0.6312557144582263,
      "grad_norm": 0.1530408412218094,
      "learning_rate": 7.896157266844543e-05,
      "loss": 0.7528,
      "mean_token_accuracy": 0.7780436724424362,
      "num_tokens": 218988150.0,
      "step": 6861
    },
    {
      "entropy": 0.7093115355819464,
      "epoch": 0.6313477208296675,
      "grad_norm": 0.16528967022895813,
      "learning_rate": 7.895850584230379e-05,
      "loss": 0.7104,
      "mean_token_accuracy": 0.7943553626537323,
      "num_tokens": 219020084.0,
      "step": 6862
    },
    {
      "entropy": 0.7651770375669003,
      "epoch": 0.6314397272011086,
      "grad_norm": 0.16754727065563202,
      "learning_rate": 7.895543901616217e-05,
      "loss": 0.7499,
      "mean_token_accuracy": 0.7790952697396278,
      "num_tokens": 219051099.0,
      "step": 6863
    },
    {
      "entropy": 0.6031515169888735,
      "epoch": 0.6315317335725499,
      "grad_norm": 0.15793846547603607,
      "learning_rate": 7.895237219002056e-05,
      "loss": 0.5826,
      "mean_token_accuracy": 0.8288630768656731,
      "num_tokens": 219083615.0,
      "step": 6864
    },
    {
      "entropy": 0.777523186057806,
      "epoch": 0.6316237399439911,
      "grad_norm": 0.1603538691997528,
      "learning_rate": 7.894930536387892e-05,
      "loss": 0.7568,
      "mean_token_accuracy": 0.7788731455802917,
      "num_tokens": 219116015.0,
      "step": 6865
    },
    {
      "entropy": 0.7567935241386294,
      "epoch": 0.6317157463154324,
      "grad_norm": 0.1436605304479599,
      "learning_rate": 7.89462385377373e-05,
      "loss": 0.7336,
      "mean_token_accuracy": 0.7841677628457546,
      "num_tokens": 219148557.0,
      "step": 6866
    },
    {
      "entropy": 0.7585240807384253,
      "epoch": 0.6318077526868736,
      "grad_norm": 0.15620064735412598,
      "learning_rate": 7.894317171159567e-05,
      "loss": 0.7348,
      "mean_token_accuracy": 0.7816766761243343,
      "num_tokens": 219180882.0,
      "step": 6867
    },
    {
      "entropy": 0.7566265016794205,
      "epoch": 0.6318997590583147,
      "grad_norm": 0.15792877972126007,
      "learning_rate": 7.894010488545404e-05,
      "loss": 0.7257,
      "mean_token_accuracy": 0.7830765843391418,
      "num_tokens": 219212064.0,
      "step": 6868
    },
    {
      "entropy": 0.7926495671272278,
      "epoch": 0.631991765429756,
      "grad_norm": 0.15216009318828583,
      "learning_rate": 7.893703805931242e-05,
      "loss": 0.7934,
      "mean_token_accuracy": 0.7698236890137196,
      "num_tokens": 219244575.0,
      "step": 6869
    },
    {
      "entropy": 0.7504438497126102,
      "epoch": 0.6320837718011972,
      "grad_norm": 0.14170676469802856,
      "learning_rate": 7.89339712331708e-05,
      "loss": 0.7211,
      "mean_token_accuracy": 0.7866550423204899,
      "num_tokens": 219277268.0,
      "step": 6870
    },
    {
      "entropy": 0.6848162338137627,
      "epoch": 0.6321757781726385,
      "grad_norm": 0.1524059921503067,
      "learning_rate": 7.893090440702917e-05,
      "loss": 0.6852,
      "mean_token_accuracy": 0.7993138618767262,
      "num_tokens": 219308165.0,
      "step": 6871
    },
    {
      "entropy": 0.8672268502414227,
      "epoch": 0.6322677845440797,
      "grad_norm": 0.16447299718856812,
      "learning_rate": 7.892783758088754e-05,
      "loss": 0.8695,
      "mean_token_accuracy": 0.7438171096146107,
      "num_tokens": 219339256.0,
      "step": 6872
    },
    {
      "entropy": 0.706127742305398,
      "epoch": 0.6323597909155209,
      "grad_norm": 0.15461266040802002,
      "learning_rate": 7.892477075474591e-05,
      "loss": 0.7034,
      "mean_token_accuracy": 0.7962778583168983,
      "num_tokens": 219371241.0,
      "step": 6873
    },
    {
      "entropy": 0.8098400942981243,
      "epoch": 0.6324517972869621,
      "grad_norm": 0.17031319439411163,
      "learning_rate": 7.89217039286043e-05,
      "loss": 0.8094,
      "mean_token_accuracy": 0.7635058127343655,
      "num_tokens": 219402915.0,
      "step": 6874
    },
    {
      "entropy": 0.6876364629715681,
      "epoch": 0.6325438036584033,
      "grad_norm": 0.15097521245479584,
      "learning_rate": 7.891863710246267e-05,
      "loss": 0.6996,
      "mean_token_accuracy": 0.791767843067646,
      "num_tokens": 219434845.0,
      "step": 6875
    },
    {
      "entropy": 0.6914844606071711,
      "epoch": 0.6326358100298446,
      "grad_norm": 0.1577075868844986,
      "learning_rate": 7.891557027632104e-05,
      "loss": 0.6697,
      "mean_token_accuracy": 0.7997679598629475,
      "num_tokens": 219466591.0,
      "step": 6876
    },
    {
      "entropy": 0.7687974851578474,
      "epoch": 0.6327278164012858,
      "grad_norm": 0.16553539037704468,
      "learning_rate": 7.89125034501794e-05,
      "loss": 0.7825,
      "mean_token_accuracy": 0.7728806212544441,
      "num_tokens": 219498877.0,
      "step": 6877
    },
    {
      "entropy": 0.647903598845005,
      "epoch": 0.632819822772727,
      "grad_norm": 0.14612559974193573,
      "learning_rate": 7.890943662403779e-05,
      "loss": 0.6291,
      "mean_token_accuracy": 0.8138858005404472,
      "num_tokens": 219531030.0,
      "step": 6878
    },
    {
      "entropy": 0.7009576931595802,
      "epoch": 0.6329118291441682,
      "grad_norm": 0.1548846960067749,
      "learning_rate": 7.890636979789617e-05,
      "loss": 0.6845,
      "mean_token_accuracy": 0.7955183312296867,
      "num_tokens": 219563011.0,
      "step": 6879
    },
    {
      "entropy": 0.6400025235489011,
      "epoch": 0.6330038355156095,
      "grad_norm": 0.14393669366836548,
      "learning_rate": 7.890330297175454e-05,
      "loss": 0.6095,
      "mean_token_accuracy": 0.8201652355492115,
      "num_tokens": 219595215.0,
      "step": 6880
    },
    {
      "entropy": 0.7246956834569573,
      "epoch": 0.6330958418870507,
      "grad_norm": 0.15411576628684998,
      "learning_rate": 7.89002361456129e-05,
      "loss": 0.7136,
      "mean_token_accuracy": 0.7890480235219002,
      "num_tokens": 219626876.0,
      "step": 6881
    },
    {
      "entropy": 0.8542935065925121,
      "epoch": 0.633187848258492,
      "grad_norm": 0.16094587743282318,
      "learning_rate": 7.889716931947129e-05,
      "loss": 0.8363,
      "mean_token_accuracy": 0.7578501142561436,
      "num_tokens": 219658077.0,
      "step": 6882
    },
    {
      "entropy": 0.8830975834280252,
      "epoch": 0.6332798546299331,
      "grad_norm": 0.16134829819202423,
      "learning_rate": 7.889410249332965e-05,
      "loss": 0.8728,
      "mean_token_accuracy": 0.7446320690214634,
      "num_tokens": 219690504.0,
      "step": 6883
    },
    {
      "entropy": 0.7369049983099103,
      "epoch": 0.6333718610013743,
      "grad_norm": 0.1492011696100235,
      "learning_rate": 7.889103566718804e-05,
      "loss": 0.7267,
      "mean_token_accuracy": 0.7867975607514381,
      "num_tokens": 219722581.0,
      "step": 6884
    },
    {
      "entropy": 0.7690662890672684,
      "epoch": 0.6334638673728156,
      "grad_norm": 0.15343309938907623,
      "learning_rate": 7.88879688410464e-05,
      "loss": 0.7558,
      "mean_token_accuracy": 0.7762762978672981,
      "num_tokens": 219754010.0,
      "step": 6885
    },
    {
      "entropy": 0.6959979627281427,
      "epoch": 0.6335558737442568,
      "grad_norm": 0.15820489823818207,
      "learning_rate": 7.888490201490478e-05,
      "loss": 0.689,
      "mean_token_accuracy": 0.7933076992630959,
      "num_tokens": 219785509.0,
      "step": 6886
    },
    {
      "entropy": 0.6995438579469919,
      "epoch": 0.6336478801156981,
      "grad_norm": 0.15785536170005798,
      "learning_rate": 7.888183518876315e-05,
      "loss": 0.6959,
      "mean_token_accuracy": 0.7934532426297665,
      "num_tokens": 219817228.0,
      "step": 6887
    },
    {
      "entropy": 0.842024203389883,
      "epoch": 0.6337398864871392,
      "grad_norm": 0.16808946430683136,
      "learning_rate": 7.887876836262152e-05,
      "loss": 0.8447,
      "mean_token_accuracy": 0.7533891052007675,
      "num_tokens": 219849996.0,
      "step": 6888
    },
    {
      "entropy": 0.6365965977311134,
      "epoch": 0.6338318928585804,
      "grad_norm": 0.16134296357631683,
      "learning_rate": 7.88757015364799e-05,
      "loss": 0.6362,
      "mean_token_accuracy": 0.8088445849716663,
      "num_tokens": 219881471.0,
      "step": 6889
    },
    {
      "entropy": 0.7535643242299557,
      "epoch": 0.6339238992300217,
      "grad_norm": 0.16724799573421478,
      "learning_rate": 7.887263471033828e-05,
      "loss": 0.768,
      "mean_token_accuracy": 0.7769832350313663,
      "num_tokens": 219913694.0,
      "step": 6890
    },
    {
      "entropy": 0.7155507206916809,
      "epoch": 0.6340159056014629,
      "grad_norm": 0.1497797966003418,
      "learning_rate": 7.886956788419665e-05,
      "loss": 0.7009,
      "mean_token_accuracy": 0.7904127687215805,
      "num_tokens": 219946153.0,
      "step": 6891
    },
    {
      "entropy": 0.7289273254573345,
      "epoch": 0.6341079119729042,
      "grad_norm": 0.15925821661949158,
      "learning_rate": 7.886650105805502e-05,
      "loss": 0.7227,
      "mean_token_accuracy": 0.7861085422337055,
      "num_tokens": 219977857.0,
      "step": 6892
    },
    {
      "entropy": 0.7606760635972023,
      "epoch": 0.6341999183443453,
      "grad_norm": 0.14937244355678558,
      "learning_rate": 7.886343423191339e-05,
      "loss": 0.7354,
      "mean_token_accuracy": 0.7849611155688763,
      "num_tokens": 220010021.0,
      "step": 6893
    },
    {
      "entropy": 0.7056446466594934,
      "epoch": 0.6342919247157865,
      "grad_norm": 0.15720370411872864,
      "learning_rate": 7.886036740577177e-05,
      "loss": 0.6931,
      "mean_token_accuracy": 0.7940652184188366,
      "num_tokens": 220041530.0,
      "step": 6894
    },
    {
      "entropy": 0.7483748104423285,
      "epoch": 0.6343839310872278,
      "grad_norm": 0.15946544706821442,
      "learning_rate": 7.885730057963015e-05,
      "loss": 0.722,
      "mean_token_accuracy": 0.7928668931126595,
      "num_tokens": 220074063.0,
      "step": 6895
    },
    {
      "entropy": 0.7201498113572598,
      "epoch": 0.634475937458669,
      "grad_norm": 0.14722144603729248,
      "learning_rate": 7.885423375348852e-05,
      "loss": 0.6805,
      "mean_token_accuracy": 0.7994227446615696,
      "num_tokens": 220106247.0,
      "step": 6896
    },
    {
      "entropy": 0.8181928563863039,
      "epoch": 0.6345679438301103,
      "grad_norm": 0.15910127758979797,
      "learning_rate": 7.88511669273469e-05,
      "loss": 0.8035,
      "mean_token_accuracy": 0.7674312442541122,
      "num_tokens": 220137168.0,
      "step": 6897
    },
    {
      "entropy": 0.7863948121666908,
      "epoch": 0.6346599502015514,
      "grad_norm": 0.15939854085445404,
      "learning_rate": 7.884810010120527e-05,
      "loss": 0.7903,
      "mean_token_accuracy": 0.76661242172122,
      "num_tokens": 220169654.0,
      "step": 6898
    },
    {
      "entropy": 0.7124992497265339,
      "epoch": 0.6347519565729927,
      "grad_norm": 0.16192249953746796,
      "learning_rate": 7.884503327506363e-05,
      "loss": 0.7105,
      "mean_token_accuracy": 0.790167510509491,
      "num_tokens": 220202253.0,
      "step": 6899
    },
    {
      "entropy": 0.7630895804613829,
      "epoch": 0.6348439629444339,
      "grad_norm": 0.15378238260746002,
      "learning_rate": 7.884196644892202e-05,
      "loss": 0.7675,
      "mean_token_accuracy": 0.7720111422240734,
      "num_tokens": 220234525.0,
      "step": 6900
    },
    {
      "entropy": 0.7489477945491672,
      "epoch": 0.6349359693158751,
      "grad_norm": 0.17037475109100342,
      "learning_rate": 7.88388996227804e-05,
      "loss": 0.7454,
      "mean_token_accuracy": 0.7808733955025673,
      "num_tokens": 220266179.0,
      "step": 6901
    },
    {
      "entropy": 0.7576781418174505,
      "epoch": 0.6350279756873164,
      "grad_norm": 0.16227898001670837,
      "learning_rate": 7.883583279663877e-05,
      "loss": 0.7732,
      "mean_token_accuracy": 0.7722276411950588,
      "num_tokens": 220298506.0,
      "step": 6902
    },
    {
      "entropy": 0.6916989833116531,
      "epoch": 0.6351199820587575,
      "grad_norm": 0.15824340283870697,
      "learning_rate": 7.883276597049713e-05,
      "loss": 0.6859,
      "mean_token_accuracy": 0.7973580472171307,
      "num_tokens": 220330749.0,
      "step": 6903
    },
    {
      "entropy": 0.7783776931464672,
      "epoch": 0.6352119884301988,
      "grad_norm": 0.15735553205013275,
      "learning_rate": 7.88296991443555e-05,
      "loss": 0.7663,
      "mean_token_accuracy": 0.7793160825967789,
      "num_tokens": 220363433.0,
      "step": 6904
    },
    {
      "entropy": 0.7449226174503565,
      "epoch": 0.63530399480164,
      "grad_norm": 0.155316561460495,
      "learning_rate": 7.882663231821388e-05,
      "loss": 0.7295,
      "mean_token_accuracy": 0.7877831794321537,
      "num_tokens": 220395601.0,
      "step": 6905
    },
    {
      "entropy": 0.7989328354597092,
      "epoch": 0.6353960011730813,
      "grad_norm": 0.1553983837366104,
      "learning_rate": 7.882356549207226e-05,
      "loss": 0.7641,
      "mean_token_accuracy": 0.7741676643490791,
      "num_tokens": 220428017.0,
      "step": 6906
    },
    {
      "entropy": 0.7869943305850029,
      "epoch": 0.6354880075445225,
      "grad_norm": 0.16666008532047272,
      "learning_rate": 7.882049866593063e-05,
      "loss": 0.7747,
      "mean_token_accuracy": 0.7685129009187222,
      "num_tokens": 220459910.0,
      "step": 6907
    },
    {
      "entropy": 0.6044191587716341,
      "epoch": 0.6355800139159636,
      "grad_norm": 0.14952507615089417,
      "learning_rate": 7.8817431839789e-05,
      "loss": 0.5736,
      "mean_token_accuracy": 0.8229855857789516,
      "num_tokens": 220491355.0,
      "step": 6908
    },
    {
      "entropy": 0.7890926245599985,
      "epoch": 0.6356720202874049,
      "grad_norm": 0.16430102288722992,
      "learning_rate": 7.881436501364738e-05,
      "loss": 0.7787,
      "mean_token_accuracy": 0.771955594420433,
      "num_tokens": 220523756.0,
      "step": 6909
    },
    {
      "entropy": 0.6198111455887556,
      "epoch": 0.6357640266588461,
      "grad_norm": 0.15202540159225464,
      "learning_rate": 7.881129818750575e-05,
      "loss": 0.6068,
      "mean_token_accuracy": 0.817258533090353,
      "num_tokens": 220555967.0,
      "step": 6910
    },
    {
      "entropy": 0.7839644812047482,
      "epoch": 0.6358560330302874,
      "grad_norm": 0.15533941984176636,
      "learning_rate": 7.880823136136413e-05,
      "loss": 0.7585,
      "mean_token_accuracy": 0.7738964408636093,
      "num_tokens": 220587447.0,
      "step": 6911
    },
    {
      "entropy": 0.7597628682851791,
      "epoch": 0.6359480394017286,
      "grad_norm": 0.16218988597393036,
      "learning_rate": 7.88051645352225e-05,
      "loss": 0.7562,
      "mean_token_accuracy": 0.7810676880180836,
      "num_tokens": 220618483.0,
      "step": 6912
    },
    {
      "entropy": 0.7581516318023205,
      "epoch": 0.6360400457731697,
      "grad_norm": 0.15725722908973694,
      "learning_rate": 7.880209770908088e-05,
      "loss": 0.7388,
      "mean_token_accuracy": 0.781148586422205,
      "num_tokens": 220649648.0,
      "step": 6913
    },
    {
      "entropy": 0.8058434017002583,
      "epoch": 0.636132052144611,
      "grad_norm": 0.16305983066558838,
      "learning_rate": 7.879903088293925e-05,
      "loss": 0.8036,
      "mean_token_accuracy": 0.7701302319765091,
      "num_tokens": 220681460.0,
      "step": 6914
    },
    {
      "entropy": 0.6640015691518784,
      "epoch": 0.6362240585160522,
      "grad_norm": 0.14963920414447784,
      "learning_rate": 7.879596405679762e-05,
      "loss": 0.6377,
      "mean_token_accuracy": 0.8045287765562534,
      "num_tokens": 220713950.0,
      "step": 6915
    },
    {
      "entropy": 0.8547518812119961,
      "epoch": 0.6363160648874935,
      "grad_norm": 0.15522746741771698,
      "learning_rate": 7.8792897230656e-05,
      "loss": 0.8408,
      "mean_token_accuracy": 0.7520415708422661,
      "num_tokens": 220746150.0,
      "step": 6916
    },
    {
      "entropy": 0.7502295896410942,
      "epoch": 0.6364080712589347,
      "grad_norm": 0.15908190608024597,
      "learning_rate": 7.878983040451438e-05,
      "loss": 0.7292,
      "mean_token_accuracy": 0.7867097072303295,
      "num_tokens": 220778167.0,
      "step": 6917
    },
    {
      "entropy": 0.6475034076720476,
      "epoch": 0.6365000776303759,
      "grad_norm": 0.14975711703300476,
      "learning_rate": 7.878676357837275e-05,
      "loss": 0.638,
      "mean_token_accuracy": 0.8108189776539803,
      "num_tokens": 220810215.0,
      "step": 6918
    },
    {
      "entropy": 0.7320692390203476,
      "epoch": 0.6365920840018171,
      "grad_norm": 0.1596670299768448,
      "learning_rate": 7.878369675223111e-05,
      "loss": 0.7396,
      "mean_token_accuracy": 0.7825630828738213,
      "num_tokens": 220842746.0,
      "step": 6919
    },
    {
      "entropy": 0.7282084096223116,
      "epoch": 0.6366840903732583,
      "grad_norm": 0.15729855000972748,
      "learning_rate": 7.87806299260895e-05,
      "loss": 0.7161,
      "mean_token_accuracy": 0.7874595820903778,
      "num_tokens": 220873975.0,
      "step": 6920
    },
    {
      "entropy": 0.7019317634403706,
      "epoch": 0.6367760967446996,
      "grad_norm": 0.15092310309410095,
      "learning_rate": 7.877756309994788e-05,
      "loss": 0.6996,
      "mean_token_accuracy": 0.7950171530246735,
      "num_tokens": 220906184.0,
      "step": 6921
    },
    {
      "entropy": 0.7185094300657511,
      "epoch": 0.6368681031161408,
      "grad_norm": 0.16186295449733734,
      "learning_rate": 7.877449627380624e-05,
      "loss": 0.7165,
      "mean_token_accuracy": 0.7894270122051239,
      "num_tokens": 220937505.0,
      "step": 6922
    },
    {
      "entropy": 0.8011875506490469,
      "epoch": 0.636960109487582,
      "grad_norm": 0.16271448135375977,
      "learning_rate": 7.877142944766461e-05,
      "loss": 0.8075,
      "mean_token_accuracy": 0.7630279920995235,
      "num_tokens": 220968342.0,
      "step": 6923
    },
    {
      "entropy": 0.8592997640371323,
      "epoch": 0.6370521158590232,
      "grad_norm": 0.16435152292251587,
      "learning_rate": 7.876836262152298e-05,
      "loss": 0.8478,
      "mean_token_accuracy": 0.7527000270783901,
      "num_tokens": 221000042.0,
      "step": 6924
    },
    {
      "entropy": 0.7696901261806488,
      "epoch": 0.6371441222304645,
      "grad_norm": 0.16929267346858978,
      "learning_rate": 7.876529579538136e-05,
      "loss": 0.7737,
      "mean_token_accuracy": 0.7742915563285351,
      "num_tokens": 221032010.0,
      "step": 6925
    },
    {
      "entropy": 0.7291579004377127,
      "epoch": 0.6372361286019057,
      "grad_norm": 0.16437077522277832,
      "learning_rate": 7.876222896923974e-05,
      "loss": 0.7092,
      "mean_token_accuracy": 0.786116998642683,
      "num_tokens": 221064128.0,
      "step": 6926
    },
    {
      "entropy": 0.7003750130534172,
      "epoch": 0.637328134973347,
      "grad_norm": 0.15857945382595062,
      "learning_rate": 7.875916214309811e-05,
      "loss": 0.7099,
      "mean_token_accuracy": 0.7931471616029739,
      "num_tokens": 221096156.0,
      "step": 6927
    },
    {
      "entropy": 0.7965775765478611,
      "epoch": 0.6374201413447881,
      "grad_norm": 0.15748770534992218,
      "learning_rate": 7.875609531695649e-05,
      "loss": 0.7778,
      "mean_token_accuracy": 0.7742580287158489,
      "num_tokens": 221128145.0,
      "step": 6928
    },
    {
      "entropy": 0.7164390534162521,
      "epoch": 0.6375121477162293,
      "grad_norm": 0.1647290587425232,
      "learning_rate": 7.875302849081486e-05,
      "loss": 0.7197,
      "mean_token_accuracy": 0.7878861092031002,
      "num_tokens": 221159980.0,
      "step": 6929
    },
    {
      "entropy": 0.749900771304965,
      "epoch": 0.6376041540876706,
      "grad_norm": 0.16067934036254883,
      "learning_rate": 7.874996166467323e-05,
      "loss": 0.7336,
      "mean_token_accuracy": 0.779102586209774,
      "num_tokens": 221191535.0,
      "step": 6930
    },
    {
      "entropy": 0.7528647445142269,
      "epoch": 0.6376961604591118,
      "grad_norm": 0.15382139384746552,
      "learning_rate": 7.874689483853161e-05,
      "loss": 0.7314,
      "mean_token_accuracy": 0.7839807197451591,
      "num_tokens": 221223728.0,
      "step": 6931
    },
    {
      "entropy": 0.7213712427765131,
      "epoch": 0.6377881668305531,
      "grad_norm": 0.16305390000343323,
      "learning_rate": 7.874382801238999e-05,
      "loss": 0.7135,
      "mean_token_accuracy": 0.7923485487699509,
      "num_tokens": 221256496.0,
      "step": 6932
    },
    {
      "entropy": 0.707587193697691,
      "epoch": 0.6378801732019942,
      "grad_norm": 0.15158496797084808,
      "learning_rate": 7.874076118624836e-05,
      "loss": 0.6781,
      "mean_token_accuracy": 0.8017660863697529,
      "num_tokens": 221288139.0,
      "step": 6933
    },
    {
      "entropy": 0.6838178746402264,
      "epoch": 0.6379721795734354,
      "grad_norm": 0.1562667042016983,
      "learning_rate": 7.873769436010673e-05,
      "loss": 0.673,
      "mean_token_accuracy": 0.7962257191538811,
      "num_tokens": 221320630.0,
      "step": 6934
    },
    {
      "entropy": 0.6883549690246582,
      "epoch": 0.6380641859448767,
      "grad_norm": 0.16248096525669098,
      "learning_rate": 7.87346275339651e-05,
      "loss": 0.6646,
      "mean_token_accuracy": 0.8020828887820244,
      "num_tokens": 221352974.0,
      "step": 6935
    },
    {
      "entropy": 0.8717504180967808,
      "epoch": 0.6381561923163179,
      "grad_norm": 0.1659393459558487,
      "learning_rate": 7.873156070782348e-05,
      "loss": 0.8449,
      "mean_token_accuracy": 0.7571968138217926,
      "num_tokens": 221385126.0,
      "step": 6936
    },
    {
      "entropy": 0.7857749573886395,
      "epoch": 0.6382481986877592,
      "grad_norm": 0.1600053906440735,
      "learning_rate": 7.872849388168186e-05,
      "loss": 0.7848,
      "mean_token_accuracy": 0.766365259885788,
      "num_tokens": 221417263.0,
      "step": 6937
    },
    {
      "entropy": 0.7336103338748217,
      "epoch": 0.6383402050592003,
      "grad_norm": 0.14641864597797394,
      "learning_rate": 7.872542705554023e-05,
      "loss": 0.7202,
      "mean_token_accuracy": 0.7900140732526779,
      "num_tokens": 221449275.0,
      "step": 6938
    },
    {
      "entropy": 0.7350007332861423,
      "epoch": 0.6384322114306415,
      "grad_norm": 0.16301490366458893,
      "learning_rate": 7.87223602293986e-05,
      "loss": 0.7253,
      "mean_token_accuracy": 0.7882105298340321,
      "num_tokens": 221479760.0,
      "step": 6939
    },
    {
      "entropy": 0.732866358011961,
      "epoch": 0.6385242178020828,
      "grad_norm": 0.1690296232700348,
      "learning_rate": 7.871929340325697e-05,
      "loss": 0.7358,
      "mean_token_accuracy": 0.787705522030592,
      "num_tokens": 221511568.0,
      "step": 6940
    },
    {
      "entropy": 0.6378811579197645,
      "epoch": 0.638616224173524,
      "grad_norm": 0.15386812388896942,
      "learning_rate": 7.871622657711534e-05,
      "loss": 0.6305,
      "mean_token_accuracy": 0.8138011172413826,
      "num_tokens": 221543791.0,
      "step": 6941
    },
    {
      "entropy": 0.7490948531776667,
      "epoch": 0.6387082305449653,
      "grad_norm": 0.15550273656845093,
      "learning_rate": 7.871315975097372e-05,
      "loss": 0.7458,
      "mean_token_accuracy": 0.7799200601875782,
      "num_tokens": 221576313.0,
      "step": 6942
    },
    {
      "entropy": 0.6614210242405534,
      "epoch": 0.6388002369164064,
      "grad_norm": 0.15067635476589203,
      "learning_rate": 7.871009292483209e-05,
      "loss": 0.6393,
      "mean_token_accuracy": 0.8155992291867733,
      "num_tokens": 221608461.0,
      "step": 6943
    },
    {
      "entropy": 0.7185911685228348,
      "epoch": 0.6388922432878477,
      "grad_norm": 0.15482744574546814,
      "learning_rate": 7.870702609869047e-05,
      "loss": 0.7196,
      "mean_token_accuracy": 0.7862989902496338,
      "num_tokens": 221640635.0,
      "step": 6944
    },
    {
      "entropy": 0.8105077054351568,
      "epoch": 0.6389842496592889,
      "grad_norm": 0.15946754813194275,
      "learning_rate": 7.870395927254884e-05,
      "loss": 0.8091,
      "mean_token_accuracy": 0.7720139212906361,
      "num_tokens": 221673403.0,
      "step": 6945
    },
    {
      "entropy": 0.6958283763378859,
      "epoch": 0.6390762560307301,
      "grad_norm": 0.15772676467895508,
      "learning_rate": 7.870089244640721e-05,
      "loss": 0.6889,
      "mean_token_accuracy": 0.796492975205183,
      "num_tokens": 221705478.0,
      "step": 6946
    },
    {
      "entropy": 0.7459588032215834,
      "epoch": 0.6391682624021714,
      "grad_norm": 0.15366461873054504,
      "learning_rate": 7.869782562026559e-05,
      "loss": 0.7356,
      "mean_token_accuracy": 0.7801820300519466,
      "num_tokens": 221736622.0,
      "step": 6947
    },
    {
      "entropy": 0.7495031580328941,
      "epoch": 0.6392602687736125,
      "grad_norm": 0.15919840335845947,
      "learning_rate": 7.869475879412397e-05,
      "loss": 0.726,
      "mean_token_accuracy": 0.7853528223931789,
      "num_tokens": 221767982.0,
      "step": 6948
    },
    {
      "entropy": 0.6911295875906944,
      "epoch": 0.6393522751450538,
      "grad_norm": 0.15555141866207123,
      "learning_rate": 7.869169196798234e-05,
      "loss": 0.6548,
      "mean_token_accuracy": 0.807557687163353,
      "num_tokens": 221799377.0,
      "step": 6949
    },
    {
      "entropy": 0.6885363422334194,
      "epoch": 0.639444281516495,
      "grad_norm": 0.15002135932445526,
      "learning_rate": 7.868862514184071e-05,
      "loss": 0.6573,
      "mean_token_accuracy": 0.8050536289811134,
      "num_tokens": 221831476.0,
      "step": 6950
    },
    {
      "entropy": 0.7415374480187893,
      "epoch": 0.6395362878879363,
      "grad_norm": 0.15510840713977814,
      "learning_rate": 7.868555831569908e-05,
      "loss": 0.7327,
      "mean_token_accuracy": 0.7891831248998642,
      "num_tokens": 221863032.0,
      "step": 6951
    },
    {
      "entropy": 0.772614911198616,
      "epoch": 0.6396282942593775,
      "grad_norm": 0.16669751703739166,
      "learning_rate": 7.868249148955747e-05,
      "loss": 0.7673,
      "mean_token_accuracy": 0.7786742299795151,
      "num_tokens": 221895086.0,
      "step": 6952
    },
    {
      "entropy": 0.6795344613492489,
      "epoch": 0.6397203006308186,
      "grad_norm": 0.15866181254386902,
      "learning_rate": 7.867942466341584e-05,
      "loss": 0.6622,
      "mean_token_accuracy": 0.804789274930954,
      "num_tokens": 221926976.0,
      "step": 6953
    },
    {
      "entropy": 0.8312845788896084,
      "epoch": 0.6398123070022599,
      "grad_norm": 0.17144352197647095,
      "learning_rate": 7.86763578372742e-05,
      "loss": 0.835,
      "mean_token_accuracy": 0.754698108881712,
      "num_tokens": 221958620.0,
      "step": 6954
    },
    {
      "entropy": 0.7354001142084599,
      "epoch": 0.6399043133737011,
      "grad_norm": 0.16103647649288177,
      "learning_rate": 7.867329101113257e-05,
      "loss": 0.712,
      "mean_token_accuracy": 0.7875090688467026,
      "num_tokens": 221989838.0,
      "step": 6955
    },
    {
      "entropy": 0.8360791634768248,
      "epoch": 0.6399963197451424,
      "grad_norm": 0.15507563948631287,
      "learning_rate": 7.867022418499096e-05,
      "loss": 0.8071,
      "mean_token_accuracy": 0.7629932016134262,
      "num_tokens": 222021306.0,
      "step": 6956
    },
    {
      "entropy": 0.687332097440958,
      "epoch": 0.6400883261165836,
      "grad_norm": 0.150267094373703,
      "learning_rate": 7.866715735884934e-05,
      "loss": 0.6711,
      "mean_token_accuracy": 0.8006677515804768,
      "num_tokens": 222052261.0,
      "step": 6957
    },
    {
      "entropy": 0.7457604669034481,
      "epoch": 0.6401803324880248,
      "grad_norm": 0.1456674486398697,
      "learning_rate": 7.86640905327077e-05,
      "loss": 0.7276,
      "mean_token_accuracy": 0.7849526852369308,
      "num_tokens": 222084709.0,
      "step": 6958
    },
    {
      "entropy": 0.7927187643945217,
      "epoch": 0.640272338859466,
      "grad_norm": 0.15930810570716858,
      "learning_rate": 7.866102370656609e-05,
      "loss": 0.7835,
      "mean_token_accuracy": 0.7692633122205734,
      "num_tokens": 222116703.0,
      "step": 6959
    },
    {
      "entropy": 0.7339248470962048,
      "epoch": 0.6403643452309072,
      "grad_norm": 0.14879298210144043,
      "learning_rate": 7.865795688042445e-05,
      "loss": 0.7249,
      "mean_token_accuracy": 0.7836687862873077,
      "num_tokens": 222149031.0,
      "step": 6960
    },
    {
      "entropy": 0.6392824426293373,
      "epoch": 0.6404563516023485,
      "grad_norm": 0.15990878641605377,
      "learning_rate": 7.865489005428282e-05,
      "loss": 0.6504,
      "mean_token_accuracy": 0.8066058978438377,
      "num_tokens": 222181453.0,
      "step": 6961
    },
    {
      "entropy": 0.7492011748254299,
      "epoch": 0.6405483579737897,
      "grad_norm": 0.1632310152053833,
      "learning_rate": 7.86518232281412e-05,
      "loss": 0.7323,
      "mean_token_accuracy": 0.785941619426012,
      "num_tokens": 222213770.0,
      "step": 6962
    },
    {
      "entropy": 0.7727239765226841,
      "epoch": 0.6406403643452309,
      "grad_norm": 0.16280412673950195,
      "learning_rate": 7.864875640199958e-05,
      "loss": 0.7865,
      "mean_token_accuracy": 0.7720236070454121,
      "num_tokens": 222245318.0,
      "step": 6963
    },
    {
      "entropy": 0.6552695091813803,
      "epoch": 0.6407323707166721,
      "grad_norm": 0.1512708216905594,
      "learning_rate": 7.864568957585795e-05,
      "loss": 0.6304,
      "mean_token_accuracy": 0.811233002692461,
      "num_tokens": 222276715.0,
      "step": 6964
    },
    {
      "entropy": 0.7738239001482725,
      "epoch": 0.6408243770881134,
      "grad_norm": 0.16418792307376862,
      "learning_rate": 7.864262274971632e-05,
      "loss": 0.7691,
      "mean_token_accuracy": 0.7748025953769684,
      "num_tokens": 222308786.0,
      "step": 6965
    },
    {
      "entropy": 0.6872703023254871,
      "epoch": 0.6409163834595546,
      "grad_norm": 0.1540777087211609,
      "learning_rate": 7.863955592357469e-05,
      "loss": 0.6785,
      "mean_token_accuracy": 0.7967204786837101,
      "num_tokens": 222340832.0,
      "step": 6966
    },
    {
      "entropy": 0.7793655712157488,
      "epoch": 0.6410083898309958,
      "grad_norm": 0.15857237577438354,
      "learning_rate": 7.863648909743307e-05,
      "loss": 0.7823,
      "mean_token_accuracy": 0.773353349417448,
      "num_tokens": 222373146.0,
      "step": 6967
    },
    {
      "entropy": 0.7762088868767023,
      "epoch": 0.641100396202437,
      "grad_norm": 0.15695245563983917,
      "learning_rate": 7.863342227129145e-05,
      "loss": 0.7547,
      "mean_token_accuracy": 0.7766533009707928,
      "num_tokens": 222404776.0,
      "step": 6968
    },
    {
      "entropy": 0.647113611921668,
      "epoch": 0.6411924025738782,
      "grad_norm": 0.15493173897266388,
      "learning_rate": 7.863035544514982e-05,
      "loss": 0.6296,
      "mean_token_accuracy": 0.8125514760613441,
      "num_tokens": 222437068.0,
      "step": 6969
    },
    {
      "entropy": 0.8345769867300987,
      "epoch": 0.6412844089453195,
      "grad_norm": 0.1572309285402298,
      "learning_rate": 7.862728861900819e-05,
      "loss": 0.8085,
      "mean_token_accuracy": 0.7653841860592365,
      "num_tokens": 222469117.0,
      "step": 6970
    },
    {
      "entropy": 0.7175351716578007,
      "epoch": 0.6413764153167607,
      "grad_norm": 0.15733103454113007,
      "learning_rate": 7.862422179286657e-05,
      "loss": 0.7032,
      "mean_token_accuracy": 0.7920929342508316,
      "num_tokens": 222501451.0,
      "step": 6971
    },
    {
      "entropy": 0.7942229211330414,
      "epoch": 0.641468421688202,
      "grad_norm": 0.1532370001077652,
      "learning_rate": 7.862115496672494e-05,
      "loss": 0.7839,
      "mean_token_accuracy": 0.7694723792374134,
      "num_tokens": 222533058.0,
      "step": 6972
    },
    {
      "entropy": 0.7139307390898466,
      "epoch": 0.6415604280596431,
      "grad_norm": 0.1547260880470276,
      "learning_rate": 7.861808814058332e-05,
      "loss": 0.6971,
      "mean_token_accuracy": 0.7940657734870911,
      "num_tokens": 222565308.0,
      "step": 6973
    },
    {
      "entropy": 0.7965372260659933,
      "epoch": 0.6416524344310843,
      "grad_norm": 0.15389075875282288,
      "learning_rate": 7.861502131444169e-05,
      "loss": 0.7843,
      "mean_token_accuracy": 0.7753463238477707,
      "num_tokens": 222597507.0,
      "step": 6974
    },
    {
      "entropy": 0.8037285041064024,
      "epoch": 0.6417444408025256,
      "grad_norm": 0.15977290272712708,
      "learning_rate": 7.861195448830007e-05,
      "loss": 0.7908,
      "mean_token_accuracy": 0.7744931951165199,
      "num_tokens": 222629207.0,
      "step": 6975
    },
    {
      "entropy": 0.7958782557398081,
      "epoch": 0.6418364471739668,
      "grad_norm": 0.15818752348423004,
      "learning_rate": 7.860888766215843e-05,
      "loss": 0.7801,
      "mean_token_accuracy": 0.7691551707684994,
      "num_tokens": 222661467.0,
      "step": 6976
    },
    {
      "entropy": 0.7439003102481365,
      "epoch": 0.6419284535454081,
      "grad_norm": 0.15825903415679932,
      "learning_rate": 7.86058208360168e-05,
      "loss": 0.7257,
      "mean_token_accuracy": 0.784990206360817,
      "num_tokens": 222693519.0,
      "step": 6977
    },
    {
      "entropy": 0.6731985304504633,
      "epoch": 0.6420204599168492,
      "grad_norm": 0.1532893180847168,
      "learning_rate": 7.860275400987518e-05,
      "loss": 0.6562,
      "mean_token_accuracy": 0.8075496405363083,
      "num_tokens": 222725851.0,
      "step": 6978
    },
    {
      "entropy": 0.7621798645704985,
      "epoch": 0.6421124662882904,
      "grad_norm": 0.15318959951400757,
      "learning_rate": 7.859968718373357e-05,
      "loss": 0.7417,
      "mean_token_accuracy": 0.7850859649479389,
      "num_tokens": 222757937.0,
      "step": 6979
    },
    {
      "entropy": 0.6357100512832403,
      "epoch": 0.6422044726597317,
      "grad_norm": 0.15389390289783478,
      "learning_rate": 7.859662035759193e-05,
      "loss": 0.6344,
      "mean_token_accuracy": 0.8108495697379112,
      "num_tokens": 222789912.0,
      "step": 6980
    },
    {
      "entropy": 0.7247553188353777,
      "epoch": 0.6422964790311729,
      "grad_norm": 0.16087166965007782,
      "learning_rate": 7.85935535314503e-05,
      "loss": 0.6981,
      "mean_token_accuracy": 0.7895753011107445,
      "num_tokens": 222821306.0,
      "step": 6981
    },
    {
      "entropy": 0.8516242727637291,
      "epoch": 0.6423884854026142,
      "grad_norm": 0.16525718569755554,
      "learning_rate": 7.859048670530867e-05,
      "loss": 0.8598,
      "mean_token_accuracy": 0.7495944797992706,
      "num_tokens": 222852761.0,
      "step": 6982
    },
    {
      "entropy": 0.719877764582634,
      "epoch": 0.6424804917740553,
      "grad_norm": 0.1557914912700653,
      "learning_rate": 7.858741987916705e-05,
      "loss": 0.7113,
      "mean_token_accuracy": 0.7903487011790276,
      "num_tokens": 222884946.0,
      "step": 6983
    },
    {
      "entropy": 0.6846062242984772,
      "epoch": 0.6425724981454966,
      "grad_norm": 0.1535934954881668,
      "learning_rate": 7.858435305302543e-05,
      "loss": 0.6815,
      "mean_token_accuracy": 0.7947444915771484,
      "num_tokens": 222916689.0,
      "step": 6984
    },
    {
      "entropy": 0.7103770114481449,
      "epoch": 0.6426645045169378,
      "grad_norm": 0.15120014548301697,
      "learning_rate": 7.85812862268838e-05,
      "loss": 0.7063,
      "mean_token_accuracy": 0.7892937287688255,
      "num_tokens": 222949007.0,
      "step": 6985
    },
    {
      "entropy": 0.6969670038670301,
      "epoch": 0.642756510888379,
      "grad_norm": 0.1718427687883377,
      "learning_rate": 7.857821940074217e-05,
      "loss": 0.709,
      "mean_token_accuracy": 0.7921602129936218,
      "num_tokens": 222981036.0,
      "step": 6986
    },
    {
      "entropy": 0.7065334990620613,
      "epoch": 0.6428485172598203,
      "grad_norm": 0.17499487102031708,
      "learning_rate": 7.857515257460055e-05,
      "loss": 0.701,
      "mean_token_accuracy": 0.787717655301094,
      "num_tokens": 223012215.0,
      "step": 6987
    },
    {
      "entropy": 0.7533202134072781,
      "epoch": 0.6429405236312614,
      "grad_norm": 0.15356722474098206,
      "learning_rate": 7.857208574845892e-05,
      "loss": 0.7326,
      "mean_token_accuracy": 0.7873737774789333,
      "num_tokens": 223044114.0,
      "step": 6988
    },
    {
      "entropy": 0.7719446383416653,
      "epoch": 0.6430325300027027,
      "grad_norm": 0.15759645402431488,
      "learning_rate": 7.85690189223173e-05,
      "loss": 0.7593,
      "mean_token_accuracy": 0.7827654667198658,
      "num_tokens": 223075680.0,
      "step": 6989
    },
    {
      "entropy": 0.8331038132309914,
      "epoch": 0.6431245363741439,
      "grad_norm": 0.15807290375232697,
      "learning_rate": 7.856595209617568e-05,
      "loss": 0.8098,
      "mean_token_accuracy": 0.7674301899969578,
      "num_tokens": 223107574.0,
      "step": 6990
    },
    {
      "entropy": 0.7592663504183292,
      "epoch": 0.6432165427455852,
      "grad_norm": 0.15590573847293854,
      "learning_rate": 7.856288527003405e-05,
      "loss": 0.7408,
      "mean_token_accuracy": 0.7804142571985722,
      "num_tokens": 223139874.0,
      "step": 6991
    },
    {
      "entropy": 0.7451607473194599,
      "epoch": 0.6433085491170264,
      "grad_norm": 0.15617635846138,
      "learning_rate": 7.855981844389242e-05,
      "loss": 0.7113,
      "mean_token_accuracy": 0.7901097610592842,
      "num_tokens": 223171429.0,
      "step": 6992
    },
    {
      "entropy": 0.8518300727009773,
      "epoch": 0.6434005554884675,
      "grad_norm": 0.1581299901008606,
      "learning_rate": 7.855675161775078e-05,
      "loss": 0.8249,
      "mean_token_accuracy": 0.7588352896273136,
      "num_tokens": 223203062.0,
      "step": 6993
    },
    {
      "entropy": 0.8336306698620319,
      "epoch": 0.6434925618599088,
      "grad_norm": 0.1553668975830078,
      "learning_rate": 7.855368479160918e-05,
      "loss": 0.8164,
      "mean_token_accuracy": 0.7603516317903996,
      "num_tokens": 223234944.0,
      "step": 6994
    },
    {
      "entropy": 0.8216424584388733,
      "epoch": 0.64358456823135,
      "grad_norm": 0.1589539796113968,
      "learning_rate": 7.855061796546755e-05,
      "loss": 0.8362,
      "mean_token_accuracy": 0.7535314559936523,
      "num_tokens": 223267147.0,
      "step": 6995
    },
    {
      "entropy": 0.6913234740495682,
      "epoch": 0.6436765746027913,
      "grad_norm": 0.15525004267692566,
      "learning_rate": 7.854755113932591e-05,
      "loss": 0.6571,
      "mean_token_accuracy": 0.8058821074664593,
      "num_tokens": 223299366.0,
      "step": 6996
    },
    {
      "entropy": 0.7263008821755648,
      "epoch": 0.6437685809742325,
      "grad_norm": 0.14647334814071655,
      "learning_rate": 7.854448431318428e-05,
      "loss": 0.728,
      "mean_token_accuracy": 0.7829576656222343,
      "num_tokens": 223331835.0,
      "step": 6997
    },
    {
      "entropy": 0.7870141062885523,
      "epoch": 0.6438605873456736,
      "grad_norm": 0.15660057961940765,
      "learning_rate": 7.854141748704266e-05,
      "loss": 0.7795,
      "mean_token_accuracy": 0.774809505790472,
      "num_tokens": 223363781.0,
      "step": 6998
    },
    {
      "entropy": 0.7767510525882244,
      "epoch": 0.6439525937171149,
      "grad_norm": 0.16112786531448364,
      "learning_rate": 7.853835066090105e-05,
      "loss": 0.772,
      "mean_token_accuracy": 0.7787124700844288,
      "num_tokens": 223396025.0,
      "step": 6999
    },
    {
      "entropy": 0.8967073187232018,
      "epoch": 0.6440446000885561,
      "grad_norm": 0.16050510108470917,
      "learning_rate": 7.853528383475941e-05,
      "loss": 0.8987,
      "mean_token_accuracy": 0.740609522908926,
      "num_tokens": 223428225.0,
      "step": 7000
    },
    {
      "entropy": 0.8308091312646866,
      "epoch": 0.6441366064599974,
      "grad_norm": 0.16400142014026642,
      "learning_rate": 7.853221700861778e-05,
      "loss": 0.812,
      "mean_token_accuracy": 0.7660635411739349,
      "num_tokens": 223460299.0,
      "step": 7001
    },
    {
      "entropy": 0.8260846715420485,
      "epoch": 0.6442286128314386,
      "grad_norm": 0.16439177095890045,
      "learning_rate": 7.852915018247616e-05,
      "loss": 0.8171,
      "mean_token_accuracy": 0.7633972950279713,
      "num_tokens": 223492335.0,
      "step": 7002
    },
    {
      "entropy": 0.7950864359736443,
      "epoch": 0.6443206192028798,
      "grad_norm": 0.15909570455551147,
      "learning_rate": 7.852608335633453e-05,
      "loss": 0.8014,
      "mean_token_accuracy": 0.7621428035199642,
      "num_tokens": 223524756.0,
      "step": 7003
    },
    {
      "entropy": 0.7818479835987091,
      "epoch": 0.644412625574321,
      "grad_norm": 0.16049790382385254,
      "learning_rate": 7.852301653019291e-05,
      "loss": 0.7562,
      "mean_token_accuracy": 0.7798853740096092,
      "num_tokens": 223556736.0,
      "step": 7004
    },
    {
      "entropy": 0.7815417293459177,
      "epoch": 0.6445046319457622,
      "grad_norm": 0.16948312520980835,
      "learning_rate": 7.851994970405128e-05,
      "loss": 0.7842,
      "mean_token_accuracy": 0.7708058916032314,
      "num_tokens": 223588961.0,
      "step": 7005
    },
    {
      "entropy": 0.8369145616889,
      "epoch": 0.6445966383172035,
      "grad_norm": 0.16389793157577515,
      "learning_rate": 7.851688287790966e-05,
      "loss": 0.8274,
      "mean_token_accuracy": 0.7620183862745762,
      "num_tokens": 223620892.0,
      "step": 7006
    },
    {
      "entropy": 0.8038438335061073,
      "epoch": 0.6446886446886447,
      "grad_norm": 0.16933207213878632,
      "learning_rate": 7.851381605176803e-05,
      "loss": 0.7973,
      "mean_token_accuracy": 0.7649762853980064,
      "num_tokens": 223652262.0,
      "step": 7007
    },
    {
      "entropy": 0.7376859877258539,
      "epoch": 0.6447806510600859,
      "grad_norm": 0.1517903357744217,
      "learning_rate": 7.85107492256264e-05,
      "loss": 0.7291,
      "mean_token_accuracy": 0.783548466861248,
      "num_tokens": 223684513.0,
      "step": 7008
    },
    {
      "entropy": 0.7690382059663534,
      "epoch": 0.6448726574315271,
      "grad_norm": 0.15151222050189972,
      "learning_rate": 7.850768239948478e-05,
      "loss": 0.7664,
      "mean_token_accuracy": 0.7739690132439137,
      "num_tokens": 223715608.0,
      "step": 7009
    },
    {
      "entropy": 0.6898020375519991,
      "epoch": 0.6449646638029684,
      "grad_norm": 0.16039547324180603,
      "learning_rate": 7.850461557334316e-05,
      "loss": 0.6781,
      "mean_token_accuracy": 0.7983189821243286,
      "num_tokens": 223747842.0,
      "step": 7010
    },
    {
      "entropy": 0.7854726891964674,
      "epoch": 0.6450566701744096,
      "grad_norm": 0.15733622014522552,
      "learning_rate": 7.850154874720153e-05,
      "loss": 0.7826,
      "mean_token_accuracy": 0.770992960780859,
      "num_tokens": 223780045.0,
      "step": 7011
    },
    {
      "entropy": 0.7768614757806063,
      "epoch": 0.6451486765458508,
      "grad_norm": 0.16505606472492218,
      "learning_rate": 7.84984819210599e-05,
      "loss": 0.7709,
      "mean_token_accuracy": 0.773954700678587,
      "num_tokens": 223811993.0,
      "step": 7012
    },
    {
      "entropy": 0.7114319344982505,
      "epoch": 0.645240682917292,
      "grad_norm": 0.15300971269607544,
      "learning_rate": 7.849541509491826e-05,
      "loss": 0.6852,
      "mean_token_accuracy": 0.7968831546604633,
      "num_tokens": 223844185.0,
      "step": 7013
    },
    {
      "entropy": 0.7102359663695097,
      "epoch": 0.6453326892887332,
      "grad_norm": 0.1549244076013565,
      "learning_rate": 7.849234826877664e-05,
      "loss": 0.7004,
      "mean_token_accuracy": 0.791448287665844,
      "num_tokens": 223876155.0,
      "step": 7014
    },
    {
      "entropy": 0.8142467085272074,
      "epoch": 0.6454246956601745,
      "grad_norm": 0.15792810916900635,
      "learning_rate": 7.848928144263503e-05,
      "loss": 0.7866,
      "mean_token_accuracy": 0.7675737217068672,
      "num_tokens": 223907141.0,
      "step": 7015
    },
    {
      "entropy": 0.8202043753117323,
      "epoch": 0.6455167020316157,
      "grad_norm": 0.153945192694664,
      "learning_rate": 7.84862146164934e-05,
      "loss": 0.8024,
      "mean_token_accuracy": 0.7670767866075039,
      "num_tokens": 223938970.0,
      "step": 7016
    },
    {
      "entropy": 0.7377364039421082,
      "epoch": 0.645608708403057,
      "grad_norm": 0.16242384910583496,
      "learning_rate": 7.848314779035178e-05,
      "loss": 0.7366,
      "mean_token_accuracy": 0.7870790399610996,
      "num_tokens": 223970739.0,
      "step": 7017
    },
    {
      "entropy": 0.7534980792552233,
      "epoch": 0.6457007147744981,
      "grad_norm": 0.16097456216812134,
      "learning_rate": 7.848008096421014e-05,
      "loss": 0.749,
      "mean_token_accuracy": 0.7810821607708931,
      "num_tokens": 224002828.0,
      "step": 7018
    },
    {
      "entropy": 0.8382720910012722,
      "epoch": 0.6457927211459393,
      "grad_norm": 0.16667617857456207,
      "learning_rate": 7.847701413806851e-05,
      "loss": 0.8436,
      "mean_token_accuracy": 0.7598011903464794,
      "num_tokens": 224035069.0,
      "step": 7019
    },
    {
      "entropy": 0.8098375760018826,
      "epoch": 0.6458847275173806,
      "grad_norm": 0.15933576226234436,
      "learning_rate": 7.847394731192689e-05,
      "loss": 0.7982,
      "mean_token_accuracy": 0.7639972753822803,
      "num_tokens": 224066838.0,
      "step": 7020
    },
    {
      "entropy": 0.8150272108614445,
      "epoch": 0.6459767338888218,
      "grad_norm": 0.1604910045862198,
      "learning_rate": 7.847088048578527e-05,
      "loss": 0.8071,
      "mean_token_accuracy": 0.7645313069224358,
      "num_tokens": 224098524.0,
      "step": 7021
    },
    {
      "entropy": 0.7874063067138195,
      "epoch": 0.6460687402602631,
      "grad_norm": 0.16045254468917847,
      "learning_rate": 7.846781365964364e-05,
      "loss": 0.7829,
      "mean_token_accuracy": 0.7672378905117512,
      "num_tokens": 224130863.0,
      "step": 7022
    },
    {
      "entropy": 0.7409669253975153,
      "epoch": 0.6461607466317042,
      "grad_norm": 0.15618133544921875,
      "learning_rate": 7.846474683350201e-05,
      "loss": 0.724,
      "mean_token_accuracy": 0.7890619449317455,
      "num_tokens": 224163381.0,
      "step": 7023
    },
    {
      "entropy": 0.7398055717349052,
      "epoch": 0.6462527530031454,
      "grad_norm": 0.1661335974931717,
      "learning_rate": 7.846168000736038e-05,
      "loss": 0.7102,
      "mean_token_accuracy": 0.7945688664913177,
      "num_tokens": 224195420.0,
      "step": 7024
    },
    {
      "entropy": 0.6882862783968449,
      "epoch": 0.6463447593745867,
      "grad_norm": 0.15367881953716278,
      "learning_rate": 7.845861318121876e-05,
      "loss": 0.6809,
      "mean_token_accuracy": 0.7984783947467804,
      "num_tokens": 224227041.0,
      "step": 7025
    },
    {
      "entropy": 0.7503610970452428,
      "epoch": 0.6464367657460279,
      "grad_norm": 0.15369729697704315,
      "learning_rate": 7.845554635507714e-05,
      "loss": 0.7364,
      "mean_token_accuracy": 0.7821775935590267,
      "num_tokens": 224258275.0,
      "step": 7026
    },
    {
      "entropy": 0.8044120371341705,
      "epoch": 0.6465287721174692,
      "grad_norm": 0.16137804090976715,
      "learning_rate": 7.845247952893551e-05,
      "loss": 0.8211,
      "mean_token_accuracy": 0.7653166353702545,
      "num_tokens": 224289613.0,
      "step": 7027
    },
    {
      "entropy": 0.7855161372572184,
      "epoch": 0.6466207784889103,
      "grad_norm": 0.16347208619117737,
      "learning_rate": 7.844941270279388e-05,
      "loss": 0.8044,
      "mean_token_accuracy": 0.76554049924016,
      "num_tokens": 224321993.0,
      "step": 7028
    },
    {
      "entropy": 0.8544645942747593,
      "epoch": 0.6467127848603516,
      "grad_norm": 0.16179513931274414,
      "learning_rate": 7.844634587665226e-05,
      "loss": 0.8609,
      "mean_token_accuracy": 0.748654305934906,
      "num_tokens": 224353155.0,
      "step": 7029
    },
    {
      "entropy": 0.8192231860011816,
      "epoch": 0.6468047912317928,
      "grad_norm": 0.16610221564769745,
      "learning_rate": 7.844327905051063e-05,
      "loss": 0.8051,
      "mean_token_accuracy": 0.7649747431278229,
      "num_tokens": 224384631.0,
      "step": 7030
    },
    {
      "entropy": 0.6485357219353318,
      "epoch": 0.646896797603234,
      "grad_norm": 0.14379265904426575,
      "learning_rate": 7.8440212224369e-05,
      "loss": 0.6239,
      "mean_token_accuracy": 0.8130840137600899,
      "num_tokens": 224416734.0,
      "step": 7031
    },
    {
      "entropy": 0.7893319390714169,
      "epoch": 0.6469888039746753,
      "grad_norm": 0.16190986335277557,
      "learning_rate": 7.843714539822737e-05,
      "loss": 0.7629,
      "mean_token_accuracy": 0.7734259366989136,
      "num_tokens": 224448729.0,
      "step": 7032
    },
    {
      "entropy": 0.7863367609679699,
      "epoch": 0.6470808103461164,
      "grad_norm": 0.15810243785381317,
      "learning_rate": 7.843407857208576e-05,
      "loss": 0.7744,
      "mean_token_accuracy": 0.7716306410729885,
      "num_tokens": 224480856.0,
      "step": 7033
    },
    {
      "entropy": 0.830646850168705,
      "epoch": 0.6471728167175577,
      "grad_norm": 0.1595454066991806,
      "learning_rate": 7.843101174594412e-05,
      "loss": 0.809,
      "mean_token_accuracy": 0.7641596570611,
      "num_tokens": 224512976.0,
      "step": 7034
    },
    {
      "entropy": 0.7034146003425121,
      "epoch": 0.6472648230889989,
      "grad_norm": 0.1543247401714325,
      "learning_rate": 7.842794491980249e-05,
      "loss": 0.6677,
      "mean_token_accuracy": 0.8031925559043884,
      "num_tokens": 224544413.0,
      "step": 7035
    },
    {
      "entropy": 0.6929711252450943,
      "epoch": 0.6473568294604402,
      "grad_norm": 0.162679061293602,
      "learning_rate": 7.842487809366087e-05,
      "loss": 0.695,
      "mean_token_accuracy": 0.7929810062050819,
      "num_tokens": 224577003.0,
      "step": 7036
    },
    {
      "entropy": 0.6633480116724968,
      "epoch": 0.6474488358318814,
      "grad_norm": 0.16138708591461182,
      "learning_rate": 7.842181126751925e-05,
      "loss": 0.6559,
      "mean_token_accuracy": 0.8051009215414524,
      "num_tokens": 224609097.0,
      "step": 7037
    },
    {
      "entropy": 0.6918485574424267,
      "epoch": 0.6475408422033225,
      "grad_norm": 0.15208441019058228,
      "learning_rate": 7.841874444137762e-05,
      "loss": 0.6856,
      "mean_token_accuracy": 0.7947488240897655,
      "num_tokens": 224641478.0,
      "step": 7038
    },
    {
      "entropy": 0.7106715077534318,
      "epoch": 0.6476328485747638,
      "grad_norm": 0.16619913280010223,
      "learning_rate": 7.841567761523599e-05,
      "loss": 0.7156,
      "mean_token_accuracy": 0.7882184349000454,
      "num_tokens": 224672170.0,
      "step": 7039
    },
    {
      "entropy": 0.8245738130062819,
      "epoch": 0.647724854946205,
      "grad_norm": 0.16561740636825562,
      "learning_rate": 7.841261078909437e-05,
      "loss": 0.82,
      "mean_token_accuracy": 0.7640877962112427,
      "num_tokens": 224704753.0,
      "step": 7040
    },
    {
      "entropy": 0.8366010878235102,
      "epoch": 0.6478168613176463,
      "grad_norm": 0.16241921484470367,
      "learning_rate": 7.840954396295275e-05,
      "loss": 0.829,
      "mean_token_accuracy": 0.7554929330945015,
      "num_tokens": 224735627.0,
      "step": 7041
    },
    {
      "entropy": 0.629624355584383,
      "epoch": 0.6479088676890875,
      "grad_norm": 0.14902172982692719,
      "learning_rate": 7.840647713681112e-05,
      "loss": 0.6186,
      "mean_token_accuracy": 0.8152554854750633,
      "num_tokens": 224768012.0,
      "step": 7042
    },
    {
      "entropy": 0.6820451896637678,
      "epoch": 0.6480008740605286,
      "grad_norm": 0.14793610572814941,
      "learning_rate": 7.840341031066949e-05,
      "loss": 0.6695,
      "mean_token_accuracy": 0.7965960837900639,
      "num_tokens": 224800027.0,
      "step": 7043
    },
    {
      "entropy": 0.7285981420427561,
      "epoch": 0.6480928804319699,
      "grad_norm": 0.16306476294994354,
      "learning_rate": 7.840034348452786e-05,
      "loss": 0.6992,
      "mean_token_accuracy": 0.7907245419919491,
      "num_tokens": 224832505.0,
      "step": 7044
    },
    {
      "entropy": 0.6208251174539328,
      "epoch": 0.6481848868034111,
      "grad_norm": 0.1453896313905716,
      "learning_rate": 7.839727665838624e-05,
      "loss": 0.5855,
      "mean_token_accuracy": 0.8259277492761612,
      "num_tokens": 224863643.0,
      "step": 7045
    },
    {
      "entropy": 0.6938970871269703,
      "epoch": 0.6482768931748524,
      "grad_norm": 0.15591667592525482,
      "learning_rate": 7.839420983224462e-05,
      "loss": 0.6838,
      "mean_token_accuracy": 0.7955549880862236,
      "num_tokens": 224895467.0,
      "step": 7046
    },
    {
      "entropy": 0.7088063806295395,
      "epoch": 0.6483688995462936,
      "grad_norm": 0.15757714211940765,
      "learning_rate": 7.839114300610299e-05,
      "loss": 0.703,
      "mean_token_accuracy": 0.7928457446396351,
      "num_tokens": 224927776.0,
      "step": 7047
    },
    {
      "entropy": 0.8180240858346224,
      "epoch": 0.6484609059177348,
      "grad_norm": 0.16017897427082062,
      "learning_rate": 7.838807617996137e-05,
      "loss": 0.8063,
      "mean_token_accuracy": 0.7577881366014481,
      "num_tokens": 224959427.0,
      "step": 7048
    },
    {
      "entropy": 0.7268592547625303,
      "epoch": 0.648552912289176,
      "grad_norm": 0.1593882143497467,
      "learning_rate": 7.838500935381974e-05,
      "loss": 0.7122,
      "mean_token_accuracy": 0.788813591003418,
      "num_tokens": 224991511.0,
      "step": 7049
    },
    {
      "entropy": 0.8182171657681465,
      "epoch": 0.6486449186606172,
      "grad_norm": 0.16388137638568878,
      "learning_rate": 7.83819425276781e-05,
      "loss": 0.8281,
      "mean_token_accuracy": 0.7615449316799641,
      "num_tokens": 225023194.0,
      "step": 7050
    },
    {
      "entropy": 0.6155620329082012,
      "epoch": 0.6487369250320585,
      "grad_norm": 0.14957188069820404,
      "learning_rate": 7.837887570153649e-05,
      "loss": 0.605,
      "mean_token_accuracy": 0.8205740712583065,
      "num_tokens": 225055213.0,
      "step": 7051
    },
    {
      "entropy": 0.7764961589127779,
      "epoch": 0.6488289314034997,
      "grad_norm": 0.16244736313819885,
      "learning_rate": 7.837580887539487e-05,
      "loss": 0.7695,
      "mean_token_accuracy": 0.7767932713031769,
      "num_tokens": 225086827.0,
      "step": 7052
    },
    {
      "entropy": 0.8186197858303785,
      "epoch": 0.6489209377749409,
      "grad_norm": 0.16264373064041138,
      "learning_rate": 7.837274204925324e-05,
      "loss": 0.8066,
      "mean_token_accuracy": 0.7643033266067505,
      "num_tokens": 225118176.0,
      "step": 7053
    },
    {
      "entropy": 0.791192676872015,
      "epoch": 0.6490129441463821,
      "grad_norm": 0.1637670397758484,
      "learning_rate": 7.83696752231116e-05,
      "loss": 0.7914,
      "mean_token_accuracy": 0.7625746987760067,
      "num_tokens": 225149116.0,
      "step": 7054
    },
    {
      "entropy": 0.791076272726059,
      "epoch": 0.6491049505178234,
      "grad_norm": 0.16043800115585327,
      "learning_rate": 7.836660839696997e-05,
      "loss": 0.7869,
      "mean_token_accuracy": 0.7670504674315453,
      "num_tokens": 225181497.0,
      "step": 7055
    },
    {
      "entropy": 0.7719947062432766,
      "epoch": 0.6491969568892646,
      "grad_norm": 0.16315023601055145,
      "learning_rate": 7.836354157082835e-05,
      "loss": 0.7657,
      "mean_token_accuracy": 0.7718561105430126,
      "num_tokens": 225212646.0,
      "step": 7056
    },
    {
      "entropy": 0.6356240045279264,
      "epoch": 0.6492889632607058,
      "grad_norm": 0.15018752217292786,
      "learning_rate": 7.836047474468673e-05,
      "loss": 0.6138,
      "mean_token_accuracy": 0.8190382979810238,
      "num_tokens": 225244943.0,
      "step": 7057
    },
    {
      "entropy": 0.7676794826984406,
      "epoch": 0.649380969632147,
      "grad_norm": 0.15665677189826965,
      "learning_rate": 7.83574079185451e-05,
      "loss": 0.7501,
      "mean_token_accuracy": 0.7841073721647263,
      "num_tokens": 225277148.0,
      "step": 7058
    },
    {
      "entropy": 0.8835291527211666,
      "epoch": 0.6494729760035882,
      "grad_norm": 0.16000792384147644,
      "learning_rate": 7.835434109240347e-05,
      "loss": 0.8633,
      "mean_token_accuracy": 0.7450621500611305,
      "num_tokens": 225308676.0,
      "step": 7059
    },
    {
      "entropy": 0.850252777338028,
      "epoch": 0.6495649823750295,
      "grad_norm": 0.1561698466539383,
      "learning_rate": 7.835127426626185e-05,
      "loss": 0.8393,
      "mean_token_accuracy": 0.7552613876760006,
      "num_tokens": 225340423.0,
      "step": 7060
    },
    {
      "entropy": 0.9002683982253075,
      "epoch": 0.6496569887464707,
      "grad_norm": 0.1639838069677353,
      "learning_rate": 7.834820744012022e-05,
      "loss": 0.909,
      "mean_token_accuracy": 0.7386104986071587,
      "num_tokens": 225372727.0,
      "step": 7061
    },
    {
      "entropy": 0.8928390517830849,
      "epoch": 0.649748995117912,
      "grad_norm": 0.16494756937026978,
      "learning_rate": 7.83451406139786e-05,
      "loss": 0.899,
      "mean_token_accuracy": 0.7445872910320759,
      "num_tokens": 225405285.0,
      "step": 7062
    },
    {
      "entropy": 0.6896291449666023,
      "epoch": 0.6498410014893531,
      "grad_norm": 0.1660977154970169,
      "learning_rate": 7.834207378783697e-05,
      "loss": 0.6936,
      "mean_token_accuracy": 0.7969843447208405,
      "num_tokens": 225437797.0,
      "step": 7063
    },
    {
      "entropy": 0.9013302437961102,
      "epoch": 0.6499330078607943,
      "grad_norm": 0.15838123857975006,
      "learning_rate": 7.833900696169535e-05,
      "loss": 0.8923,
      "mean_token_accuracy": 0.7450406327843666,
      "num_tokens": 225470500.0,
      "step": 7064
    },
    {
      "entropy": 0.8362146746367216,
      "epoch": 0.6500250142322356,
      "grad_norm": 0.16335108876228333,
      "learning_rate": 7.833594013555372e-05,
      "loss": 0.8325,
      "mean_token_accuracy": 0.7571384385228157,
      "num_tokens": 225501900.0,
      "step": 7065
    },
    {
      "entropy": 0.5252052368596196,
      "epoch": 0.6501170206036768,
      "grad_norm": 0.14809775352478027,
      "learning_rate": 7.833287330941209e-05,
      "loss": 0.5075,
      "mean_token_accuracy": 0.846546046435833,
      "num_tokens": 225534496.0,
      "step": 7066
    },
    {
      "entropy": 0.7827712167054415,
      "epoch": 0.6502090269751181,
      "grad_norm": 0.15060476958751678,
      "learning_rate": 7.832980648327047e-05,
      "loss": 0.7492,
      "mean_token_accuracy": 0.7766511961817741,
      "num_tokens": 225565436.0,
      "step": 7067
    },
    {
      "entropy": 0.7452480234205723,
      "epoch": 0.6503010333465592,
      "grad_norm": 0.15516942739486694,
      "learning_rate": 7.832673965712885e-05,
      "loss": 0.7376,
      "mean_token_accuracy": 0.7830190062522888,
      "num_tokens": 225597946.0,
      "step": 7068
    },
    {
      "entropy": 0.6994322594255209,
      "epoch": 0.6503930397180004,
      "grad_norm": 0.14167505502700806,
      "learning_rate": 7.832367283098722e-05,
      "loss": 0.6741,
      "mean_token_accuracy": 0.8013881370425224,
      "num_tokens": 225629993.0,
      "step": 7069
    },
    {
      "entropy": 0.8138234987854958,
      "epoch": 0.6504850460894417,
      "grad_norm": 0.1660451740026474,
      "learning_rate": 7.832060600484558e-05,
      "loss": 0.8132,
      "mean_token_accuracy": 0.7651695646345615,
      "num_tokens": 225660842.0,
      "step": 7070
    },
    {
      "entropy": 0.80012680683285,
      "epoch": 0.6505770524608829,
      "grad_norm": 0.15463247895240784,
      "learning_rate": 7.831753917870395e-05,
      "loss": 0.7883,
      "mean_token_accuracy": 0.769807480275631,
      "num_tokens": 225693130.0,
      "step": 7071
    },
    {
      "entropy": 0.6080197375267744,
      "epoch": 0.6506690588323242,
      "grad_norm": 0.15143421292304993,
      "learning_rate": 7.831447235256235e-05,
      "loss": 0.5837,
      "mean_token_accuracy": 0.8262443244457245,
      "num_tokens": 225724673.0,
      "step": 7072
    },
    {
      "entropy": 0.6860790885984898,
      "epoch": 0.6507610652037653,
      "grad_norm": 0.15604457259178162,
      "learning_rate": 7.831140552642071e-05,
      "loss": 0.6874,
      "mean_token_accuracy": 0.7978916317224503,
      "num_tokens": 225756790.0,
      "step": 7073
    },
    {
      "entropy": 0.7180101033300161,
      "epoch": 0.6508530715752066,
      "grad_norm": 0.15824978053569794,
      "learning_rate": 7.830833870027908e-05,
      "loss": 0.6949,
      "mean_token_accuracy": 0.7972591519355774,
      "num_tokens": 225789325.0,
      "step": 7074
    },
    {
      "entropy": 0.7864622613415122,
      "epoch": 0.6509450779466478,
      "grad_norm": 0.1578536182641983,
      "learning_rate": 7.830527187413745e-05,
      "loss": 0.7716,
      "mean_token_accuracy": 0.7716756388545036,
      "num_tokens": 225821122.0,
      "step": 7075
    },
    {
      "entropy": 0.6746358703821898,
      "epoch": 0.651037084318089,
      "grad_norm": 0.15360479056835175,
      "learning_rate": 7.830220504799583e-05,
      "loss": 0.6701,
      "mean_token_accuracy": 0.8018503896892071,
      "num_tokens": 225853148.0,
      "step": 7076
    },
    {
      "entropy": 0.7843972034752369,
      "epoch": 0.6511290906895303,
      "grad_norm": 0.16458819806575775,
      "learning_rate": 7.829913822185421e-05,
      "loss": 0.7888,
      "mean_token_accuracy": 0.7646307609975338,
      "num_tokens": 225884463.0,
      "step": 7077
    },
    {
      "entropy": 0.6657775714993477,
      "epoch": 0.6512210970609714,
      "grad_norm": 0.16557736694812775,
      "learning_rate": 7.829607139571258e-05,
      "loss": 0.6643,
      "mean_token_accuracy": 0.7988160848617554,
      "num_tokens": 225916612.0,
      "step": 7078
    },
    {
      "entropy": 0.7380835562944412,
      "epoch": 0.6513131034324127,
      "grad_norm": 0.15999022126197815,
      "learning_rate": 7.829300456957096e-05,
      "loss": 0.7411,
      "mean_token_accuracy": 0.7822708301246166,
      "num_tokens": 225949156.0,
      "step": 7079
    },
    {
      "entropy": 0.7809624634683132,
      "epoch": 0.6514051098038539,
      "grad_norm": 0.15731902420520782,
      "learning_rate": 7.828993774342933e-05,
      "loss": 0.7732,
      "mean_token_accuracy": 0.7786605954170227,
      "num_tokens": 225980669.0,
      "step": 7080
    },
    {
      "entropy": 0.7022479325532913,
      "epoch": 0.6514971161752952,
      "grad_norm": 0.15897925198078156,
      "learning_rate": 7.82868709172877e-05,
      "loss": 0.6909,
      "mean_token_accuracy": 0.7995466217398643,
      "num_tokens": 226012469.0,
      "step": 7081
    },
    {
      "entropy": 0.7149181813001633,
      "epoch": 0.6515891225467364,
      "grad_norm": 0.16231432557106018,
      "learning_rate": 7.828380409114608e-05,
      "loss": 0.6955,
      "mean_token_accuracy": 0.796647384762764,
      "num_tokens": 226044080.0,
      "step": 7082
    },
    {
      "entropy": 0.815986467525363,
      "epoch": 0.6516811289181775,
      "grad_norm": 0.15654291212558746,
      "learning_rate": 7.828073726500446e-05,
      "loss": 0.7943,
      "mean_token_accuracy": 0.7679411061108112,
      "num_tokens": 226076325.0,
      "step": 7083
    },
    {
      "entropy": 0.7582853250205517,
      "epoch": 0.6517731352896188,
      "grad_norm": 0.1538000851869583,
      "learning_rate": 7.827767043886283e-05,
      "loss": 0.7228,
      "mean_token_accuracy": 0.787578821182251,
      "num_tokens": 226108473.0,
      "step": 7084
    },
    {
      "entropy": 0.8102413937449455,
      "epoch": 0.65186514166106,
      "grad_norm": 0.1705787181854248,
      "learning_rate": 7.82746036127212e-05,
      "loss": 0.8133,
      "mean_token_accuracy": 0.7634129896759987,
      "num_tokens": 226140561.0,
      "step": 7085
    },
    {
      "entropy": 0.7243900112807751,
      "epoch": 0.6519571480325013,
      "grad_norm": 0.15364818274974823,
      "learning_rate": 7.827153678657956e-05,
      "loss": 0.7187,
      "mean_token_accuracy": 0.7872127220034599,
      "num_tokens": 226171786.0,
      "step": 7086
    },
    {
      "entropy": 0.7850555945187807,
      "epoch": 0.6520491544039425,
      "grad_norm": 0.15606580674648285,
      "learning_rate": 7.826846996043795e-05,
      "loss": 0.7741,
      "mean_token_accuracy": 0.7690297402441502,
      "num_tokens": 226203762.0,
      "step": 7087
    },
    {
      "entropy": 0.7594273686408997,
      "epoch": 0.6521411607753836,
      "grad_norm": 0.15241415798664093,
      "learning_rate": 7.826540313429633e-05,
      "loss": 0.752,
      "mean_token_accuracy": 0.7737353481352329,
      "num_tokens": 226235607.0,
      "step": 7088
    },
    {
      "entropy": 0.7207489348948002,
      "epoch": 0.6522331671468249,
      "grad_norm": 0.14771567285060883,
      "learning_rate": 7.82623363081547e-05,
      "loss": 0.7102,
      "mean_token_accuracy": 0.7929535433650017,
      "num_tokens": 226267985.0,
      "step": 7089
    },
    {
      "entropy": 0.7239457201212645,
      "epoch": 0.6523251735182661,
      "grad_norm": 0.1516944020986557,
      "learning_rate": 7.825926948201306e-05,
      "loss": 0.7107,
      "mean_token_accuracy": 0.7883891128003597,
      "num_tokens": 226299922.0,
      "step": 7090
    },
    {
      "entropy": 0.8303029760718346,
      "epoch": 0.6524171798897074,
      "grad_norm": 0.1606813371181488,
      "learning_rate": 7.825620265587144e-05,
      "loss": 0.8338,
      "mean_token_accuracy": 0.7592242285609245,
      "num_tokens": 226331925.0,
      "step": 7091
    },
    {
      "entropy": 0.6991548333317041,
      "epoch": 0.6525091862611486,
      "grad_norm": 0.1503029614686966,
      "learning_rate": 7.825313582972981e-05,
      "loss": 0.6751,
      "mean_token_accuracy": 0.7997346743941307,
      "num_tokens": 226363463.0,
      "step": 7092
    },
    {
      "entropy": 0.7964683566242456,
      "epoch": 0.6526011926325898,
      "grad_norm": 0.16790121793746948,
      "learning_rate": 7.82500690035882e-05,
      "loss": 0.7896,
      "mean_token_accuracy": 0.7689277976751328,
      "num_tokens": 226395533.0,
      "step": 7093
    },
    {
      "entropy": 0.7461064215749502,
      "epoch": 0.652693199004031,
      "grad_norm": 0.153537780046463,
      "learning_rate": 7.824700217744656e-05,
      "loss": 0.7454,
      "mean_token_accuracy": 0.7819192260503769,
      "num_tokens": 226427286.0,
      "step": 7094
    },
    {
      "entropy": 0.7893575318157673,
      "epoch": 0.6527852053754722,
      "grad_norm": 0.16389746963977814,
      "learning_rate": 7.824393535130494e-05,
      "loss": 0.7848,
      "mean_token_accuracy": 0.772771380841732,
      "num_tokens": 226459009.0,
      "step": 7095
    },
    {
      "entropy": 0.7598089314997196,
      "epoch": 0.6528772117469135,
      "grad_norm": 0.1606905311346054,
      "learning_rate": 7.824086852516331e-05,
      "loss": 0.769,
      "mean_token_accuracy": 0.7777285650372505,
      "num_tokens": 226491053.0,
      "step": 7096
    },
    {
      "entropy": 0.806871498003602,
      "epoch": 0.6529692181183547,
      "grad_norm": 0.16327302157878876,
      "learning_rate": 7.823780169902168e-05,
      "loss": 0.8026,
      "mean_token_accuracy": 0.7690490037202835,
      "num_tokens": 226522485.0,
      "step": 7097
    },
    {
      "entropy": 0.8813015297055244,
      "epoch": 0.6530612244897959,
      "grad_norm": 0.17195051908493042,
      "learning_rate": 7.823473487288006e-05,
      "loss": 0.8827,
      "mean_token_accuracy": 0.7448143437504768,
      "num_tokens": 226553674.0,
      "step": 7098
    },
    {
      "entropy": 0.7509406097233295,
      "epoch": 0.6531532308612371,
      "grad_norm": 0.15672223269939423,
      "learning_rate": 7.823166804673844e-05,
      "loss": 0.7482,
      "mean_token_accuracy": 0.7823903001844883,
      "num_tokens": 226584980.0,
      "step": 7099
    },
    {
      "entropy": 0.7126494795084,
      "epoch": 0.6532452372326784,
      "grad_norm": 0.15457753837108612,
      "learning_rate": 7.822860122059681e-05,
      "loss": 0.7091,
      "mean_token_accuracy": 0.7904754467308521,
      "num_tokens": 226617114.0,
      "step": 7100
    },
    {
      "entropy": 0.9065434522926807,
      "epoch": 0.6533372436041196,
      "grad_norm": 0.1556113064289093,
      "learning_rate": 7.822553439445518e-05,
      "loss": 0.9057,
      "mean_token_accuracy": 0.7389012686908245,
      "num_tokens": 226649692.0,
      "step": 7101
    },
    {
      "entropy": 0.7694163080304861,
      "epoch": 0.6534292499755608,
      "grad_norm": 0.1532258242368698,
      "learning_rate": 7.822246756831355e-05,
      "loss": 0.7454,
      "mean_token_accuracy": 0.7817245721817017,
      "num_tokens": 226681665.0,
      "step": 7102
    },
    {
      "entropy": 0.8536334373056889,
      "epoch": 0.653521256347002,
      "grad_norm": 0.16060929000377655,
      "learning_rate": 7.821940074217193e-05,
      "loss": 0.8386,
      "mean_token_accuracy": 0.7537635192275047,
      "num_tokens": 226713962.0,
      "step": 7103
    },
    {
      "entropy": 0.7036446984857321,
      "epoch": 0.6536132627184432,
      "grad_norm": 0.15057499706745148,
      "learning_rate": 7.821633391603031e-05,
      "loss": 0.6578,
      "mean_token_accuracy": 0.79886220023036,
      "num_tokens": 226745884.0,
      "step": 7104
    },
    {
      "entropy": 0.7841194290667772,
      "epoch": 0.6537052690898845,
      "grad_norm": 0.15599322319030762,
      "learning_rate": 7.821326708988868e-05,
      "loss": 0.7667,
      "mean_token_accuracy": 0.7747890911996365,
      "num_tokens": 226777985.0,
      "step": 7105
    },
    {
      "entropy": 0.7623072750866413,
      "epoch": 0.6537972754613257,
      "grad_norm": 0.14844994246959686,
      "learning_rate": 7.821020026374704e-05,
      "loss": 0.7435,
      "mean_token_accuracy": 0.7830826491117477,
      "num_tokens": 226810229.0,
      "step": 7106
    },
    {
      "entropy": 0.750044509768486,
      "epoch": 0.653889281832767,
      "grad_norm": 0.16266915202140808,
      "learning_rate": 7.820713343760543e-05,
      "loss": 0.7525,
      "mean_token_accuracy": 0.7792252190411091,
      "num_tokens": 226842386.0,
      "step": 7107
    },
    {
      "entropy": 0.8363044429570436,
      "epoch": 0.6539812882042081,
      "grad_norm": 0.15744349360466003,
      "learning_rate": 7.82040666114638e-05,
      "loss": 0.821,
      "mean_token_accuracy": 0.7602660395205021,
      "num_tokens": 226874224.0,
      "step": 7108
    },
    {
      "entropy": 0.747431606054306,
      "epoch": 0.6540732945756493,
      "grad_norm": 0.1589849442243576,
      "learning_rate": 7.820099978532218e-05,
      "loss": 0.7554,
      "mean_token_accuracy": 0.779946606606245,
      "num_tokens": 226906347.0,
      "step": 7109
    },
    {
      "entropy": 0.7376139983534813,
      "epoch": 0.6541653009470906,
      "grad_norm": 0.16852344572544098,
      "learning_rate": 7.819793295918056e-05,
      "loss": 0.7345,
      "mean_token_accuracy": 0.7824297063052654,
      "num_tokens": 226938505.0,
      "step": 7110
    },
    {
      "entropy": 0.7866084631532431,
      "epoch": 0.6542573073185318,
      "grad_norm": 0.15043769776821136,
      "learning_rate": 7.819486613303892e-05,
      "loss": 0.7726,
      "mean_token_accuracy": 0.7695916146039963,
      "num_tokens": 226969912.0,
      "step": 7111
    },
    {
      "entropy": 0.6986659523099661,
      "epoch": 0.6543493136899731,
      "grad_norm": 0.16074837744235992,
      "learning_rate": 7.819179930689729e-05,
      "loss": 0.6685,
      "mean_token_accuracy": 0.7992380075156689,
      "num_tokens": 227000587.0,
      "step": 7112
    },
    {
      "entropy": 0.6779438238590956,
      "epoch": 0.6544413200614142,
      "grad_norm": 0.1619243025779724,
      "learning_rate": 7.818873248075566e-05,
      "loss": 0.6666,
      "mean_token_accuracy": 0.803513415157795,
      "num_tokens": 227032312.0,
      "step": 7113
    },
    {
      "entropy": 0.838173596188426,
      "epoch": 0.6545333264328554,
      "grad_norm": 0.15816262364387512,
      "learning_rate": 7.818566565461406e-05,
      "loss": 0.8132,
      "mean_token_accuracy": 0.7636142820119858,
      "num_tokens": 227064807.0,
      "step": 7114
    },
    {
      "entropy": 0.6487151645123959,
      "epoch": 0.6546253328042967,
      "grad_norm": 0.16073887050151825,
      "learning_rate": 7.818259882847242e-05,
      "loss": 0.6386,
      "mean_token_accuracy": 0.810065183788538,
      "num_tokens": 227096776.0,
      "step": 7115
    },
    {
      "entropy": 0.7658461071550846,
      "epoch": 0.6547173391757379,
      "grad_norm": 0.1529790759086609,
      "learning_rate": 7.817953200233079e-05,
      "loss": 0.7463,
      "mean_token_accuracy": 0.7791707217693329,
      "num_tokens": 227128809.0,
      "step": 7116
    },
    {
      "entropy": 0.6975050717592239,
      "epoch": 0.6548093455471792,
      "grad_norm": 0.14952513575553894,
      "learning_rate": 7.817646517618916e-05,
      "loss": 0.685,
      "mean_token_accuracy": 0.7962159514427185,
      "num_tokens": 227160221.0,
      "step": 7117
    },
    {
      "entropy": 0.7342338748276234,
      "epoch": 0.6549013519186203,
      "grad_norm": 0.1654050052165985,
      "learning_rate": 7.817339835004754e-05,
      "loss": 0.735,
      "mean_token_accuracy": 0.7833146080374718,
      "num_tokens": 227191465.0,
      "step": 7118
    },
    {
      "entropy": 0.8267523311078548,
      "epoch": 0.6549933582900616,
      "grad_norm": 0.1696312129497528,
      "learning_rate": 7.817033152390592e-05,
      "loss": 0.8503,
      "mean_token_accuracy": 0.7583919540047646,
      "num_tokens": 227222111.0,
      "step": 7119
    },
    {
      "entropy": 0.8061349093914032,
      "epoch": 0.6550853646615028,
      "grad_norm": 0.16036270558834076,
      "learning_rate": 7.816726469776429e-05,
      "loss": 0.8092,
      "mean_token_accuracy": 0.7660045400261879,
      "num_tokens": 227253969.0,
      "step": 7120
    },
    {
      "entropy": 0.7092596367001534,
      "epoch": 0.655177371032944,
      "grad_norm": 0.15257269144058228,
      "learning_rate": 7.816419787162266e-05,
      "loss": 0.717,
      "mean_token_accuracy": 0.7864191085100174,
      "num_tokens": 227286015.0,
      "step": 7121
    },
    {
      "entropy": 0.8200550740584731,
      "epoch": 0.6552693774043853,
      "grad_norm": 0.16646024584770203,
      "learning_rate": 7.816113104548104e-05,
      "loss": 0.8253,
      "mean_token_accuracy": 0.758919931948185,
      "num_tokens": 227316978.0,
      "step": 7122
    },
    {
      "entropy": 0.793373316526413,
      "epoch": 0.6553613837758264,
      "grad_norm": 0.15891849994659424,
      "learning_rate": 7.81580642193394e-05,
      "loss": 0.7903,
      "mean_token_accuracy": 0.7695494554936886,
      "num_tokens": 227348249.0,
      "step": 7123
    },
    {
      "entropy": 0.6979718748480082,
      "epoch": 0.6554533901472677,
      "grad_norm": 0.15178626775741577,
      "learning_rate": 7.815499739319779e-05,
      "loss": 0.6739,
      "mean_token_accuracy": 0.8027043715119362,
      "num_tokens": 227380063.0,
      "step": 7124
    },
    {
      "entropy": 0.8111903350800276,
      "epoch": 0.6555453965187089,
      "grad_norm": 0.15992188453674316,
      "learning_rate": 7.815193056705616e-05,
      "loss": 0.81,
      "mean_token_accuracy": 0.7628663964569569,
      "num_tokens": 227412593.0,
      "step": 7125
    },
    {
      "entropy": 0.7737731337547302,
      "epoch": 0.6556374028901502,
      "grad_norm": 0.16401678323745728,
      "learning_rate": 7.814886374091454e-05,
      "loss": 0.7757,
      "mean_token_accuracy": 0.7719688378274441,
      "num_tokens": 227445044.0,
      "step": 7126
    },
    {
      "entropy": 0.6715676952153444,
      "epoch": 0.6557294092615914,
      "grad_norm": 0.16235674917697906,
      "learning_rate": 7.81457969147729e-05,
      "loss": 0.6468,
      "mean_token_accuracy": 0.8069807663559914,
      "num_tokens": 227475955.0,
      "step": 7127
    },
    {
      "entropy": 0.8393130898475647,
      "epoch": 0.6558214156330325,
      "grad_norm": 0.16243021190166473,
      "learning_rate": 7.814273008863127e-05,
      "loss": 0.8044,
      "mean_token_accuracy": 0.7679686136543751,
      "num_tokens": 227507531.0,
      "step": 7128
    },
    {
      "entropy": 0.7822301834821701,
      "epoch": 0.6559134220044738,
      "grad_norm": 0.14876475930213928,
      "learning_rate": 7.813966326248965e-05,
      "loss": 0.7584,
      "mean_token_accuracy": 0.7815239392220974,
      "num_tokens": 227540000.0,
      "step": 7129
    },
    {
      "entropy": 0.7515278235077858,
      "epoch": 0.656005428375915,
      "grad_norm": 0.16253501176834106,
      "learning_rate": 7.813659643634804e-05,
      "loss": 0.7303,
      "mean_token_accuracy": 0.7843698412179947,
      "num_tokens": 227570902.0,
      "step": 7130
    },
    {
      "entropy": 0.7657082192599773,
      "epoch": 0.6560974347473563,
      "grad_norm": 0.1677108258008957,
      "learning_rate": 7.81335296102064e-05,
      "loss": 0.7669,
      "mean_token_accuracy": 0.7755605652928352,
      "num_tokens": 227602729.0,
      "step": 7131
    },
    {
      "entropy": 0.8700656294822693,
      "epoch": 0.6561894411187975,
      "grad_norm": 0.16143253445625305,
      "learning_rate": 7.813046278406477e-05,
      "loss": 0.8554,
      "mean_token_accuracy": 0.7526623904705048,
      "num_tokens": 227635067.0,
      "step": 7132
    },
    {
      "entropy": 0.7202649973332882,
      "epoch": 0.6562814474902386,
      "grad_norm": 0.15129722654819489,
      "learning_rate": 7.812739595792314e-05,
      "loss": 0.6909,
      "mean_token_accuracy": 0.7971073724329472,
      "num_tokens": 227666684.0,
      "step": 7133
    },
    {
      "entropy": 0.8303488381206989,
      "epoch": 0.6563734538616799,
      "grad_norm": 0.1650961935520172,
      "learning_rate": 7.812432913178152e-05,
      "loss": 0.829,
      "mean_token_accuracy": 0.759656872600317,
      "num_tokens": 227697639.0,
      "step": 7134
    },
    {
      "entropy": 0.7962758839130402,
      "epoch": 0.6564654602331211,
      "grad_norm": 0.1480431705713272,
      "learning_rate": 7.81212623056399e-05,
      "loss": 0.7893,
      "mean_token_accuracy": 0.7653258927166462,
      "num_tokens": 227729815.0,
      "step": 7135
    },
    {
      "entropy": 0.7903717067092657,
      "epoch": 0.6565574666045624,
      "grad_norm": 0.1613781899213791,
      "learning_rate": 7.811819547949827e-05,
      "loss": 0.7856,
      "mean_token_accuracy": 0.7696868851780891,
      "num_tokens": 227761834.0,
      "step": 7136
    },
    {
      "entropy": 0.7617421727627516,
      "epoch": 0.6566494729760036,
      "grad_norm": 0.14869464933872223,
      "learning_rate": 7.811512865335664e-05,
      "loss": 0.761,
      "mean_token_accuracy": 0.7741854190826416,
      "num_tokens": 227794050.0,
      "step": 7137
    },
    {
      "entropy": 0.7615968734025955,
      "epoch": 0.6567414793474448,
      "grad_norm": 0.16221199929714203,
      "learning_rate": 7.811206182721502e-05,
      "loss": 0.7481,
      "mean_token_accuracy": 0.7839077152311802,
      "num_tokens": 227825900.0,
      "step": 7138
    },
    {
      "entropy": 0.8148882016539574,
      "epoch": 0.656833485718886,
      "grad_norm": 0.1592821329832077,
      "learning_rate": 7.810899500107339e-05,
      "loss": 0.8076,
      "mean_token_accuracy": 0.766542699187994,
      "num_tokens": 227857438.0,
      "step": 7139
    },
    {
      "entropy": 0.8364590778946877,
      "epoch": 0.6569254920903272,
      "grad_norm": 0.16915088891983032,
      "learning_rate": 7.810592817493177e-05,
      "loss": 0.8418,
      "mean_token_accuracy": 0.757356446236372,
      "num_tokens": 227889388.0,
      "step": 7140
    },
    {
      "entropy": 0.8487864155322313,
      "epoch": 0.6570174984617685,
      "grad_norm": 0.16548284888267517,
      "learning_rate": 7.810286134879015e-05,
      "loss": 0.8311,
      "mean_token_accuracy": 0.7540905363857746,
      "num_tokens": 227921707.0,
      "step": 7141
    },
    {
      "entropy": 0.7111711716279387,
      "epoch": 0.6571095048332097,
      "grad_norm": 0.15191160142421722,
      "learning_rate": 7.809979452264852e-05,
      "loss": 0.7053,
      "mean_token_accuracy": 0.7944554500281811,
      "num_tokens": 227954409.0,
      "step": 7142
    },
    {
      "entropy": 0.7963609825819731,
      "epoch": 0.6572015112046509,
      "grad_norm": 0.15742270648479462,
      "learning_rate": 7.809672769650689e-05,
      "loss": 0.8012,
      "mean_token_accuracy": 0.7700439430773258,
      "num_tokens": 227985929.0,
      "step": 7143
    },
    {
      "entropy": 0.8064995780587196,
      "epoch": 0.6572935175760921,
      "grad_norm": 0.16558301448822021,
      "learning_rate": 7.809366087036525e-05,
      "loss": 0.7874,
      "mean_token_accuracy": 0.7690151147544384,
      "num_tokens": 228016978.0,
      "step": 7144
    },
    {
      "entropy": 0.7271794639527798,
      "epoch": 0.6573855239475334,
      "grad_norm": 0.1698399782180786,
      "learning_rate": 7.809059404422364e-05,
      "loss": 0.7407,
      "mean_token_accuracy": 0.7836697548627853,
      "num_tokens": 228048612.0,
      "step": 7145
    },
    {
      "entropy": 0.8329573757946491,
      "epoch": 0.6574775303189746,
      "grad_norm": 0.16049626469612122,
      "learning_rate": 7.808752721808202e-05,
      "loss": 0.831,
      "mean_token_accuracy": 0.7603574693202972,
      "num_tokens": 228080524.0,
      "step": 7146
    },
    {
      "entropy": 0.7392350435256958,
      "epoch": 0.6575695366904158,
      "grad_norm": 0.15846939384937286,
      "learning_rate": 7.808446039194038e-05,
      "loss": 0.7373,
      "mean_token_accuracy": 0.7785789556801319,
      "num_tokens": 228112299.0,
      "step": 7147
    },
    {
      "entropy": 0.7997567262500525,
      "epoch": 0.657661543061857,
      "grad_norm": 0.1523139625787735,
      "learning_rate": 7.808139356579875e-05,
      "loss": 0.7818,
      "mean_token_accuracy": 0.7687104977667332,
      "num_tokens": 228143869.0,
      "step": 7148
    },
    {
      "entropy": 0.8810387067496777,
      "epoch": 0.6577535494332982,
      "grad_norm": 0.1655472368001938,
      "learning_rate": 7.807832673965713e-05,
      "loss": 0.8763,
      "mean_token_accuracy": 0.7441558986902237,
      "num_tokens": 228175832.0,
      "step": 7149
    },
    {
      "entropy": 0.6920310147106647,
      "epoch": 0.6578455558047395,
      "grad_norm": 0.16436001658439636,
      "learning_rate": 7.80752599135155e-05,
      "loss": 0.6755,
      "mean_token_accuracy": 0.8005007915198803,
      "num_tokens": 228207840.0,
      "step": 7150
    },
    {
      "entropy": 0.7942385412752628,
      "epoch": 0.6579375621761807,
      "grad_norm": 0.15671896934509277,
      "learning_rate": 7.807219308737388e-05,
      "loss": 0.7795,
      "mean_token_accuracy": 0.7723085768520832,
      "num_tokens": 228240122.0,
      "step": 7151
    },
    {
      "entropy": 0.7055822089314461,
      "epoch": 0.658029568547622,
      "grad_norm": 0.16099046170711517,
      "learning_rate": 7.806912626123225e-05,
      "loss": 0.6774,
      "mean_token_accuracy": 0.7954660803079605,
      "num_tokens": 228271513.0,
      "step": 7152
    },
    {
      "entropy": 0.7721216715872288,
      "epoch": 0.6581215749190631,
      "grad_norm": 0.153142511844635,
      "learning_rate": 7.806605943509063e-05,
      "loss": 0.7456,
      "mean_token_accuracy": 0.7818183861672878,
      "num_tokens": 228304244.0,
      "step": 7153
    },
    {
      "entropy": 0.6634752713143826,
      "epoch": 0.6582135812905043,
      "grad_norm": 0.15521030128002167,
      "learning_rate": 7.8062992608949e-05,
      "loss": 0.6428,
      "mean_token_accuracy": 0.8121464103460312,
      "num_tokens": 228335669.0,
      "step": 7154
    },
    {
      "entropy": 0.8342861793935299,
      "epoch": 0.6583055876619456,
      "grad_norm": 0.15964846312999725,
      "learning_rate": 7.805992578280737e-05,
      "loss": 0.8075,
      "mean_token_accuracy": 0.7658800520002842,
      "num_tokens": 228367717.0,
      "step": 7155
    },
    {
      "entropy": 0.7678426820784807,
      "epoch": 0.6583975940333868,
      "grad_norm": 0.15764451026916504,
      "learning_rate": 7.805685895666575e-05,
      "loss": 0.7649,
      "mean_token_accuracy": 0.7800779901444912,
      "num_tokens": 228399661.0,
      "step": 7156
    },
    {
      "entropy": 0.8951945547014475,
      "epoch": 0.6584896004048281,
      "grad_norm": 0.16122445464134216,
      "learning_rate": 7.805379213052413e-05,
      "loss": 0.8827,
      "mean_token_accuracy": 0.7477765940129757,
      "num_tokens": 228431784.0,
      "step": 7157
    },
    {
      "entropy": 0.7791591547429562,
      "epoch": 0.6585816067762692,
      "grad_norm": 0.15376083552837372,
      "learning_rate": 7.80507253043825e-05,
      "loss": 0.7593,
      "mean_token_accuracy": 0.775182943791151,
      "num_tokens": 228464291.0,
      "step": 7158
    },
    {
      "entropy": 0.8684931546449661,
      "epoch": 0.6586736131477104,
      "grad_norm": 0.16561710834503174,
      "learning_rate": 7.804765847824087e-05,
      "loss": 0.8728,
      "mean_token_accuracy": 0.74667838960886,
      "num_tokens": 228496040.0,
      "step": 7159
    },
    {
      "entropy": 0.7728478964418173,
      "epoch": 0.6587656195191517,
      "grad_norm": 0.15508869290351868,
      "learning_rate": 7.804459165209925e-05,
      "loss": 0.7581,
      "mean_token_accuracy": 0.7794061750173569,
      "num_tokens": 228527826.0,
      "step": 7160
    },
    {
      "entropy": 0.6663760859519243,
      "epoch": 0.6588576258905929,
      "grad_norm": 0.16530868411064148,
      "learning_rate": 7.804152482595763e-05,
      "loss": 0.6716,
      "mean_token_accuracy": 0.8000775501132011,
      "num_tokens": 228559577.0,
      "step": 7161
    },
    {
      "entropy": 0.6962447576224804,
      "epoch": 0.6589496322620342,
      "grad_norm": 0.15064620971679688,
      "learning_rate": 7.8038457999816e-05,
      "loss": 0.6876,
      "mean_token_accuracy": 0.7952400334179401,
      "num_tokens": 228592159.0,
      "step": 7162
    },
    {
      "entropy": 0.7101002912968397,
      "epoch": 0.6590416386334753,
      "grad_norm": 0.15734833478927612,
      "learning_rate": 7.803539117367437e-05,
      "loss": 0.7134,
      "mean_token_accuracy": 0.788399413228035,
      "num_tokens": 228624088.0,
      "step": 7163
    },
    {
      "entropy": 0.6936115566641092,
      "epoch": 0.6591336450049166,
      "grad_norm": 0.15692710876464844,
      "learning_rate": 7.803232434753273e-05,
      "loss": 0.6795,
      "mean_token_accuracy": 0.7943127043545246,
      "num_tokens": 228654733.0,
      "step": 7164
    },
    {
      "entropy": 0.7586859464645386,
      "epoch": 0.6592256513763578,
      "grad_norm": 0.15639515221118927,
      "learning_rate": 7.802925752139111e-05,
      "loss": 0.7569,
      "mean_token_accuracy": 0.7766646556556225,
      "num_tokens": 228686971.0,
      "step": 7165
    },
    {
      "entropy": 0.6668936163187027,
      "epoch": 0.659317657747799,
      "grad_norm": 0.15758101642131805,
      "learning_rate": 7.80261906952495e-05,
      "loss": 0.6759,
      "mean_token_accuracy": 0.8008822351694107,
      "num_tokens": 228719574.0,
      "step": 7166
    },
    {
      "entropy": 0.8015042571350932,
      "epoch": 0.6594096641192403,
      "grad_norm": 0.16113179922103882,
      "learning_rate": 7.802312386910786e-05,
      "loss": 0.7774,
      "mean_token_accuracy": 0.7735409550368786,
      "num_tokens": 228751469.0,
      "step": 7167
    },
    {
      "entropy": 0.6989155001938343,
      "epoch": 0.6595016704906814,
      "grad_norm": 0.15595147013664246,
      "learning_rate": 7.802005704296623e-05,
      "loss": 0.6776,
      "mean_token_accuracy": 0.7983729243278503,
      "num_tokens": 228782944.0,
      "step": 7168
    },
    {
      "entropy": 0.7910958435386419,
      "epoch": 0.6595936768621227,
      "grad_norm": 0.15749290585517883,
      "learning_rate": 7.801699021682461e-05,
      "loss": 0.7751,
      "mean_token_accuracy": 0.7736129872500896,
      "num_tokens": 228814157.0,
      "step": 7169
    },
    {
      "entropy": 0.7561138477176428,
      "epoch": 0.6596856832335639,
      "grad_norm": 0.16328124701976776,
      "learning_rate": 7.801392339068298e-05,
      "loss": 0.7504,
      "mean_token_accuracy": 0.7810018211603165,
      "num_tokens": 228845848.0,
      "step": 7170
    },
    {
      "entropy": 0.6389339156448841,
      "epoch": 0.6597776896050052,
      "grad_norm": 0.15751448273658752,
      "learning_rate": 7.801085656454136e-05,
      "loss": 0.6221,
      "mean_token_accuracy": 0.812812514603138,
      "num_tokens": 228877532.0,
      "step": 7171
    },
    {
      "entropy": 0.7899944148957729,
      "epoch": 0.6598696959764464,
      "grad_norm": 0.15759725868701935,
      "learning_rate": 7.800778973839974e-05,
      "loss": 0.7613,
      "mean_token_accuracy": 0.7769978530704975,
      "num_tokens": 228908393.0,
      "step": 7172
    },
    {
      "entropy": 0.7635229267179966,
      "epoch": 0.6599617023478876,
      "grad_norm": 0.15689760446548462,
      "learning_rate": 7.800472291225811e-05,
      "loss": 0.747,
      "mean_token_accuracy": 0.7828491292893887,
      "num_tokens": 228940994.0,
      "step": 7173
    },
    {
      "entropy": 0.7556841839104891,
      "epoch": 0.6600537087193288,
      "grad_norm": 0.16698777675628662,
      "learning_rate": 7.800165608611648e-05,
      "loss": 0.7457,
      "mean_token_accuracy": 0.7864118218421936,
      "num_tokens": 228972566.0,
      "step": 7174
    },
    {
      "entropy": 0.8504532426595688,
      "epoch": 0.66014571509077,
      "grad_norm": 0.16312670707702637,
      "learning_rate": 7.799858925997485e-05,
      "loss": 0.849,
      "mean_token_accuracy": 0.7524514272809029,
      "num_tokens": 229003700.0,
      "step": 7175
    },
    {
      "entropy": 0.7309967949986458,
      "epoch": 0.6602377214622113,
      "grad_norm": 0.16285699605941772,
      "learning_rate": 7.799552243383323e-05,
      "loss": 0.7024,
      "mean_token_accuracy": 0.7909126430749893,
      "num_tokens": 229035137.0,
      "step": 7176
    },
    {
      "entropy": 0.7568534966558218,
      "epoch": 0.6603297278336525,
      "grad_norm": 0.15399937331676483,
      "learning_rate": 7.799245560769161e-05,
      "loss": 0.7497,
      "mean_token_accuracy": 0.7806278616189957,
      "num_tokens": 229066831.0,
      "step": 7177
    },
    {
      "entropy": 0.8422721084207296,
      "epoch": 0.6604217342050938,
      "grad_norm": 0.16169176995754242,
      "learning_rate": 7.798938878154998e-05,
      "loss": 0.8547,
      "mean_token_accuracy": 0.7514215148985386,
      "num_tokens": 229098774.0,
      "step": 7178
    },
    {
      "entropy": 0.6586514282971621,
      "epoch": 0.6605137405765349,
      "grad_norm": 0.146050825715065,
      "learning_rate": 7.798632195540835e-05,
      "loss": 0.633,
      "mean_token_accuracy": 0.8097788989543915,
      "num_tokens": 229130768.0,
      "step": 7179
    },
    {
      "entropy": 0.8131730034947395,
      "epoch": 0.6606057469479761,
      "grad_norm": 0.1617434173822403,
      "learning_rate": 7.798325512926673e-05,
      "loss": 0.791,
      "mean_token_accuracy": 0.7697820626199245,
      "num_tokens": 229161863.0,
      "step": 7180
    },
    {
      "entropy": 0.8482160810381174,
      "epoch": 0.6606977533194174,
      "grad_norm": 0.15285755693912506,
      "learning_rate": 7.79801883031251e-05,
      "loss": 0.8171,
      "mean_token_accuracy": 0.7609328664839268,
      "num_tokens": 229193940.0,
      "step": 7181
    },
    {
      "entropy": 0.7678125239908695,
      "epoch": 0.6607897596908586,
      "grad_norm": 0.16296260058879852,
      "learning_rate": 7.797712147698348e-05,
      "loss": 0.7565,
      "mean_token_accuracy": 0.7764059714972973,
      "num_tokens": 229226123.0,
      "step": 7182
    },
    {
      "entropy": 0.7532285042107105,
      "epoch": 0.6608817660622999,
      "grad_norm": 0.15228091180324554,
      "learning_rate": 7.797405465084184e-05,
      "loss": 0.7459,
      "mean_token_accuracy": 0.7801134809851646,
      "num_tokens": 229258584.0,
      "step": 7183
    },
    {
      "entropy": 0.7734611770138144,
      "epoch": 0.660973772433741,
      "grad_norm": 0.1754855066537857,
      "learning_rate": 7.797098782470023e-05,
      "loss": 0.785,
      "mean_token_accuracy": 0.7693200819194317,
      "num_tokens": 229289549.0,
      "step": 7184
    },
    {
      "entropy": 0.7587405741214752,
      "epoch": 0.6610657788051822,
      "grad_norm": 0.15901802480220795,
      "learning_rate": 7.79679209985586e-05,
      "loss": 0.7668,
      "mean_token_accuracy": 0.7755454555153847,
      "num_tokens": 229322288.0,
      "step": 7185
    },
    {
      "entropy": 0.8187964390963316,
      "epoch": 0.6611577851766235,
      "grad_norm": 0.15612541139125824,
      "learning_rate": 7.796485417241696e-05,
      "loss": 0.8172,
      "mean_token_accuracy": 0.7622045278549194,
      "num_tokens": 229353497.0,
      "step": 7186
    },
    {
      "entropy": 0.7688071131706238,
      "epoch": 0.6612497915480647,
      "grad_norm": 0.15318948030471802,
      "learning_rate": 7.796178734627534e-05,
      "loss": 0.7625,
      "mean_token_accuracy": 0.772628914564848,
      "num_tokens": 229385605.0,
      "step": 7187
    },
    {
      "entropy": 0.7519329935312271,
      "epoch": 0.661341797919506,
      "grad_norm": 0.1640777438879013,
      "learning_rate": 7.795872052013372e-05,
      "loss": 0.7559,
      "mean_token_accuracy": 0.7830728888511658,
      "num_tokens": 229418186.0,
      "step": 7188
    },
    {
      "entropy": 0.8755257148295641,
      "epoch": 0.6614338042909471,
      "grad_norm": 0.16782121360301971,
      "learning_rate": 7.795565369399209e-05,
      "loss": 0.867,
      "mean_token_accuracy": 0.7503605633974075,
      "num_tokens": 229450278.0,
      "step": 7189
    },
    {
      "entropy": 0.6857521533966064,
      "epoch": 0.6615258106623884,
      "grad_norm": 0.14600016176700592,
      "learning_rate": 7.795258686785046e-05,
      "loss": 0.6787,
      "mean_token_accuracy": 0.8026338405907154,
      "num_tokens": 229482847.0,
      "step": 7190
    },
    {
      "entropy": 0.7698119860142469,
      "epoch": 0.6616178170338296,
      "grad_norm": 0.1587170958518982,
      "learning_rate": 7.794952004170883e-05,
      "loss": 0.7586,
      "mean_token_accuracy": 0.7790152691304684,
      "num_tokens": 229514914.0,
      "step": 7191
    },
    {
      "entropy": 0.7597177214920521,
      "epoch": 0.6617098234052708,
      "grad_norm": 0.15318241715431213,
      "learning_rate": 7.794645321556722e-05,
      "loss": 0.7498,
      "mean_token_accuracy": 0.782429963350296,
      "num_tokens": 229547054.0,
      "step": 7192
    },
    {
      "entropy": 0.6977422088384628,
      "epoch": 0.6618018297767121,
      "grad_norm": 0.15084198117256165,
      "learning_rate": 7.794338638942559e-05,
      "loss": 0.6843,
      "mean_token_accuracy": 0.8025103881955147,
      "num_tokens": 229579425.0,
      "step": 7193
    },
    {
      "entropy": 0.8097188249230385,
      "epoch": 0.6618938361481532,
      "grad_norm": 0.16632461547851562,
      "learning_rate": 7.794031956328396e-05,
      "loss": 0.7908,
      "mean_token_accuracy": 0.7720521688461304,
      "num_tokens": 229610961.0,
      "step": 7194
    },
    {
      "entropy": 0.8257662318646908,
      "epoch": 0.6619858425195945,
      "grad_norm": 0.17185279726982117,
      "learning_rate": 7.793725273714233e-05,
      "loss": 0.8335,
      "mean_token_accuracy": 0.7610980644822121,
      "num_tokens": 229643398.0,
      "step": 7195
    },
    {
      "entropy": 0.8637250568717718,
      "epoch": 0.6620778488910357,
      "grad_norm": 0.16018979251384735,
      "learning_rate": 7.793418591100071e-05,
      "loss": 0.8543,
      "mean_token_accuracy": 0.752851665019989,
      "num_tokens": 229675854.0,
      "step": 7196
    },
    {
      "entropy": 0.6332386247813702,
      "epoch": 0.662169855262477,
      "grad_norm": 0.14771336317062378,
      "learning_rate": 7.793111908485909e-05,
      "loss": 0.6136,
      "mean_token_accuracy": 0.8155813962221146,
      "num_tokens": 229708000.0,
      "step": 7197
    },
    {
      "entropy": 0.7217374444007874,
      "epoch": 0.6622618616339182,
      "grad_norm": 0.16745643317699432,
      "learning_rate": 7.792805225871746e-05,
      "loss": 0.7309,
      "mean_token_accuracy": 0.7855882421135902,
      "num_tokens": 229739677.0,
      "step": 7198
    },
    {
      "entropy": 0.7565439250320196,
      "epoch": 0.6623538680053593,
      "grad_norm": 0.1612696647644043,
      "learning_rate": 7.792498543257583e-05,
      "loss": 0.7364,
      "mean_token_accuracy": 0.7805099450051785,
      "num_tokens": 229771985.0,
      "step": 7199
    },
    {
      "entropy": 0.687608472071588,
      "epoch": 0.6624458743768006,
      "grad_norm": 0.1619928926229477,
      "learning_rate": 7.792191860643421e-05,
      "loss": 0.6695,
      "mean_token_accuracy": 0.8019772469997406,
      "num_tokens": 229804463.0,
      "step": 7200
    },
    {
      "entropy": 0.765262559056282,
      "epoch": 0.6625378807482418,
      "grad_norm": 0.15256145596504211,
      "learning_rate": 7.791885178029257e-05,
      "loss": 0.7501,
      "mean_token_accuracy": 0.7798644341528416,
      "num_tokens": 229836915.0,
      "step": 7201
    },
    {
      "entropy": 0.785696055740118,
      "epoch": 0.6626298871196831,
      "grad_norm": 0.15584443509578705,
      "learning_rate": 7.791578495415096e-05,
      "loss": 0.7669,
      "mean_token_accuracy": 0.7734165042638779,
      "num_tokens": 229868491.0,
      "step": 7202
    },
    {
      "entropy": 0.6436558607965708,
      "epoch": 0.6627218934911243,
      "grad_norm": 0.1528477519750595,
      "learning_rate": 7.791271812800934e-05,
      "loss": 0.6174,
      "mean_token_accuracy": 0.8128049001097679,
      "num_tokens": 229900314.0,
      "step": 7203
    },
    {
      "entropy": 0.7378689926117659,
      "epoch": 0.6628138998625654,
      "grad_norm": 0.1630687266588211,
      "learning_rate": 7.79096513018677e-05,
      "loss": 0.7372,
      "mean_token_accuracy": 0.7838071994483471,
      "num_tokens": 229931889.0,
      "step": 7204
    },
    {
      "entropy": 0.7272914927452803,
      "epoch": 0.6629059062340067,
      "grad_norm": 0.15297739207744598,
      "learning_rate": 7.790658447572607e-05,
      "loss": 0.7273,
      "mean_token_accuracy": 0.7843968607485294,
      "num_tokens": 229963449.0,
      "step": 7205
    },
    {
      "entropy": 0.7010822147130966,
      "epoch": 0.6629979126054479,
      "grad_norm": 0.14901895821094513,
      "learning_rate": 7.790351764958444e-05,
      "loss": 0.6932,
      "mean_token_accuracy": 0.7950911223888397,
      "num_tokens": 229995544.0,
      "step": 7206
    },
    {
      "entropy": 0.9184204395860434,
      "epoch": 0.6630899189768892,
      "grad_norm": 0.16302217543125153,
      "learning_rate": 7.790045082344282e-05,
      "loss": 0.9246,
      "mean_token_accuracy": 0.7339715212583542,
      "num_tokens": 230027916.0,
      "step": 7207
    },
    {
      "entropy": 0.782778499647975,
      "epoch": 0.6631819253483304,
      "grad_norm": 0.16153500974178314,
      "learning_rate": 7.78973839973012e-05,
      "loss": 0.7834,
      "mean_token_accuracy": 0.7748010605573654,
      "num_tokens": 230060092.0,
      "step": 7208
    },
    {
      "entropy": 0.7497196458280087,
      "epoch": 0.6632739317197716,
      "grad_norm": 0.15458109974861145,
      "learning_rate": 7.789431717115957e-05,
      "loss": 0.7398,
      "mean_token_accuracy": 0.7788632363080978,
      "num_tokens": 230091745.0,
      "step": 7209
    },
    {
      "entropy": 0.7108439709991217,
      "epoch": 0.6633659380912128,
      "grad_norm": 0.15555335581302643,
      "learning_rate": 7.789125034501794e-05,
      "loss": 0.6868,
      "mean_token_accuracy": 0.7960537299513817,
      "num_tokens": 230122808.0,
      "step": 7210
    },
    {
      "entropy": 0.8568577487021685,
      "epoch": 0.663457944462654,
      "grad_norm": 0.16400636732578278,
      "learning_rate": 7.788818351887632e-05,
      "loss": 0.8646,
      "mean_token_accuracy": 0.7507740780711174,
      "num_tokens": 230153719.0,
      "step": 7211
    },
    {
      "entropy": 0.8009299114346504,
      "epoch": 0.6635499508340953,
      "grad_norm": 0.1686132848262787,
      "learning_rate": 7.788511669273469e-05,
      "loss": 0.7964,
      "mean_token_accuracy": 0.7675279453396797,
      "num_tokens": 230184871.0,
      "step": 7212
    },
    {
      "entropy": 0.7229946926236153,
      "epoch": 0.6636419572055365,
      "grad_norm": 0.14792470633983612,
      "learning_rate": 7.788204986659307e-05,
      "loss": 0.7232,
      "mean_token_accuracy": 0.7845156714320183,
      "num_tokens": 230217253.0,
      "step": 7213
    },
    {
      "entropy": 0.6339634135365486,
      "epoch": 0.6637339635769777,
      "grad_norm": 0.15855824947357178,
      "learning_rate": 7.787898304045144e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.8171676993370056,
      "num_tokens": 230249815.0,
      "step": 7214
    },
    {
      "entropy": 0.7585415150970221,
      "epoch": 0.6638259699484189,
      "grad_norm": 0.16394172608852386,
      "learning_rate": 7.787591621430982e-05,
      "loss": 0.7531,
      "mean_token_accuracy": 0.7804777733981609,
      "num_tokens": 230282068.0,
      "step": 7215
    },
    {
      "entropy": 0.767007140442729,
      "epoch": 0.6639179763198602,
      "grad_norm": 0.16636043787002563,
      "learning_rate": 7.787284938816819e-05,
      "loss": 0.7521,
      "mean_token_accuracy": 0.7795507647097111,
      "num_tokens": 230313877.0,
      "step": 7216
    },
    {
      "entropy": 0.7378750592470169,
      "epoch": 0.6640099826913014,
      "grad_norm": 0.1638476401567459,
      "learning_rate": 7.786978256202656e-05,
      "loss": 0.7381,
      "mean_token_accuracy": 0.7783005423843861,
      "num_tokens": 230345058.0,
      "step": 7217
    },
    {
      "entropy": 0.8114488162100315,
      "epoch": 0.6641019890627426,
      "grad_norm": 0.17333778738975525,
      "learning_rate": 7.786671573588494e-05,
      "loss": 0.8223,
      "mean_token_accuracy": 0.7569247670471668,
      "num_tokens": 230376281.0,
      "step": 7218
    },
    {
      "entropy": 0.792849238961935,
      "epoch": 0.6641939954341838,
      "grad_norm": 0.14968828856945038,
      "learning_rate": 7.786364890974332e-05,
      "loss": 0.7727,
      "mean_token_accuracy": 0.770256768912077,
      "num_tokens": 230408179.0,
      "step": 7219
    },
    {
      "entropy": 0.7686274088919163,
      "epoch": 0.664286001805625,
      "grad_norm": 0.15407994389533997,
      "learning_rate": 7.786058208360169e-05,
      "loss": 0.7427,
      "mean_token_accuracy": 0.7792148664593697,
      "num_tokens": 230439100.0,
      "step": 7220
    },
    {
      "entropy": 0.8619262725114822,
      "epoch": 0.6643780081770663,
      "grad_norm": 0.15729765594005585,
      "learning_rate": 7.785751525746005e-05,
      "loss": 0.84,
      "mean_token_accuracy": 0.7489509657025337,
      "num_tokens": 230470411.0,
      "step": 7221
    },
    {
      "entropy": 0.8669664710760117,
      "epoch": 0.6644700145485075,
      "grad_norm": 0.16317719221115112,
      "learning_rate": 7.785444843131842e-05,
      "loss": 0.8651,
      "mean_token_accuracy": 0.7514809034764767,
      "num_tokens": 230501396.0,
      "step": 7222
    },
    {
      "entropy": 0.7070004008710384,
      "epoch": 0.6645620209199488,
      "grad_norm": 0.148972749710083,
      "learning_rate": 7.78513816051768e-05,
      "loss": 0.6898,
      "mean_token_accuracy": 0.7967566065490246,
      "num_tokens": 230533281.0,
      "step": 7223
    },
    {
      "entropy": 0.7767420634627342,
      "epoch": 0.6646540272913899,
      "grad_norm": 0.15554732084274292,
      "learning_rate": 7.784831477903518e-05,
      "loss": 0.7564,
      "mean_token_accuracy": 0.7761899046599865,
      "num_tokens": 230564327.0,
      "step": 7224
    },
    {
      "entropy": 0.7570941932499409,
      "epoch": 0.6647460336628311,
      "grad_norm": 0.1591244488954544,
      "learning_rate": 7.784524795289355e-05,
      "loss": 0.7378,
      "mean_token_accuracy": 0.7804356254637241,
      "num_tokens": 230596155.0,
      "step": 7225
    },
    {
      "entropy": 0.7202113829553127,
      "epoch": 0.6648380400342724,
      "grad_norm": 0.16432209312915802,
      "learning_rate": 7.784218112675192e-05,
      "loss": 0.7045,
      "mean_token_accuracy": 0.7915908433496952,
      "num_tokens": 230628057.0,
      "step": 7226
    },
    {
      "entropy": 0.7448302991688251,
      "epoch": 0.6649300464057136,
      "grad_norm": 0.15480899810791016,
      "learning_rate": 7.78391143006103e-05,
      "loss": 0.7302,
      "mean_token_accuracy": 0.7820852287113667,
      "num_tokens": 230660270.0,
      "step": 7227
    },
    {
      "entropy": 0.7547538876533508,
      "epoch": 0.6650220527771549,
      "grad_norm": 0.15015776455402374,
      "learning_rate": 7.783604747446867e-05,
      "loss": 0.7195,
      "mean_token_accuracy": 0.7857321836054325,
      "num_tokens": 230692094.0,
      "step": 7228
    },
    {
      "entropy": 0.7435498889535666,
      "epoch": 0.665114059148596,
      "grad_norm": 0.1629895716905594,
      "learning_rate": 7.783298064832705e-05,
      "loss": 0.7288,
      "mean_token_accuracy": 0.7878827787935734,
      "num_tokens": 230724113.0,
      "step": 7229
    },
    {
      "entropy": 0.6922519635409117,
      "epoch": 0.6652060655200372,
      "grad_norm": 0.14989057183265686,
      "learning_rate": 7.782991382218542e-05,
      "loss": 0.6778,
      "mean_token_accuracy": 0.8016266040503979,
      "num_tokens": 230756425.0,
      "step": 7230
    },
    {
      "entropy": 0.6067942073568702,
      "epoch": 0.6652980718914785,
      "grad_norm": 0.1525450497865677,
      "learning_rate": 7.78268469960438e-05,
      "loss": 0.5893,
      "mean_token_accuracy": 0.8213554099202156,
      "num_tokens": 230788827.0,
      "step": 7231
    },
    {
      "entropy": 0.8218413032591343,
      "epoch": 0.6653900782629197,
      "grad_norm": 0.1471240073442459,
      "learning_rate": 7.782378016990217e-05,
      "loss": 0.8208,
      "mean_token_accuracy": 0.759485449641943,
      "num_tokens": 230820767.0,
      "step": 7232
    },
    {
      "entropy": 0.7039004731923342,
      "epoch": 0.665482084634361,
      "grad_norm": 0.15769435465335846,
      "learning_rate": 7.782071334376054e-05,
      "loss": 0.6907,
      "mean_token_accuracy": 0.7986172996461391,
      "num_tokens": 230853173.0,
      "step": 7233
    },
    {
      "entropy": 0.6972402827814221,
      "epoch": 0.6655740910058021,
      "grad_norm": 0.16075727343559265,
      "learning_rate": 7.781764651761893e-05,
      "loss": 0.6968,
      "mean_token_accuracy": 0.792697723954916,
      "num_tokens": 230885381.0,
      "step": 7234
    },
    {
      "entropy": 0.7655957341194153,
      "epoch": 0.6656660973772434,
      "grad_norm": 0.1555275022983551,
      "learning_rate": 7.78145796914773e-05,
      "loss": 0.7591,
      "mean_token_accuracy": 0.7754972130060196,
      "num_tokens": 230916141.0,
      "step": 7235
    },
    {
      "entropy": 0.8468116242438555,
      "epoch": 0.6657581037486846,
      "grad_norm": 0.15967808663845062,
      "learning_rate": 7.781151286533567e-05,
      "loss": 0.835,
      "mean_token_accuracy": 0.7583351843059063,
      "num_tokens": 230948798.0,
      "step": 7236
    },
    {
      "entropy": 0.7673567347228527,
      "epoch": 0.6658501101201258,
      "grad_norm": 0.15242692828178406,
      "learning_rate": 7.780844603919404e-05,
      "loss": 0.7466,
      "mean_token_accuracy": 0.7798983566462994,
      "num_tokens": 230980403.0,
      "step": 7237
    },
    {
      "entropy": 0.7245159335434437,
      "epoch": 0.6659421164915671,
      "grad_norm": 0.15395782887935638,
      "learning_rate": 7.780537921305242e-05,
      "loss": 0.7055,
      "mean_token_accuracy": 0.7868083715438843,
      "num_tokens": 231011831.0,
      "step": 7238
    },
    {
      "entropy": 0.6851564552634954,
      "epoch": 0.6660341228630082,
      "grad_norm": 0.1556026190519333,
      "learning_rate": 7.78023123869108e-05,
      "loss": 0.6913,
      "mean_token_accuracy": 0.7971775084733963,
      "num_tokens": 231043979.0,
      "step": 7239
    },
    {
      "entropy": 0.7445642845705152,
      "epoch": 0.6661261292344495,
      "grad_norm": 0.15672510862350464,
      "learning_rate": 7.779924556076917e-05,
      "loss": 0.7431,
      "mean_token_accuracy": 0.781959991902113,
      "num_tokens": 231076331.0,
      "step": 7240
    },
    {
      "entropy": 0.7675095535814762,
      "epoch": 0.6662181356058907,
      "grad_norm": 0.15259268879890442,
      "learning_rate": 7.779617873462753e-05,
      "loss": 0.7496,
      "mean_token_accuracy": 0.7745702676475048,
      "num_tokens": 231108690.0,
      "step": 7241
    },
    {
      "entropy": 0.7669889908283949,
      "epoch": 0.666310141977332,
      "grad_norm": 0.15139742195606232,
      "learning_rate": 7.779311190848592e-05,
      "loss": 0.7459,
      "mean_token_accuracy": 0.7817570641636848,
      "num_tokens": 231140976.0,
      "step": 7242
    },
    {
      "entropy": 0.6540810726583004,
      "epoch": 0.6664021483487732,
      "grad_norm": 0.1583012044429779,
      "learning_rate": 7.779004508234428e-05,
      "loss": 0.6416,
      "mean_token_accuracy": 0.811349656432867,
      "num_tokens": 231173181.0,
      "step": 7243
    },
    {
      "entropy": 0.7260481892153621,
      "epoch": 0.6664941547202143,
      "grad_norm": 0.15663181245326996,
      "learning_rate": 7.778697825620266e-05,
      "loss": 0.7038,
      "mean_token_accuracy": 0.7892933301627636,
      "num_tokens": 231204821.0,
      "step": 7244
    },
    {
      "entropy": 0.6510884333401918,
      "epoch": 0.6665861610916556,
      "grad_norm": 0.1622650921344757,
      "learning_rate": 7.778391143006103e-05,
      "loss": 0.6577,
      "mean_token_accuracy": 0.8016504943370819,
      "num_tokens": 231237157.0,
      "step": 7245
    },
    {
      "entropy": 0.7531710360199213,
      "epoch": 0.6666781674630968,
      "grad_norm": 0.1651827096939087,
      "learning_rate": 7.778084460391941e-05,
      "loss": 0.7595,
      "mean_token_accuracy": 0.7767657861113548,
      "num_tokens": 231268477.0,
      "step": 7246
    },
    {
      "entropy": 0.7959143836051226,
      "epoch": 0.6667701738345381,
      "grad_norm": 0.16628850996494293,
      "learning_rate": 7.777777777777778e-05,
      "loss": 0.7985,
      "mean_token_accuracy": 0.7650587894022465,
      "num_tokens": 231299802.0,
      "step": 7247
    },
    {
      "entropy": 0.9703712686896324,
      "epoch": 0.6668621802059793,
      "grad_norm": 0.1660739779472351,
      "learning_rate": 7.777471095163615e-05,
      "loss": 0.9668,
      "mean_token_accuracy": 0.7305312603712082,
      "num_tokens": 231331855.0,
      "step": 7248
    },
    {
      "entropy": 0.756624273955822,
      "epoch": 0.6669541865774204,
      "grad_norm": 0.1574835479259491,
      "learning_rate": 7.777164412549453e-05,
      "loss": 0.7427,
      "mean_token_accuracy": 0.7860460840165615,
      "num_tokens": 231364216.0,
      "step": 7249
    },
    {
      "entropy": 0.7411985285580158,
      "epoch": 0.6670461929488617,
      "grad_norm": 0.14998140931129456,
      "learning_rate": 7.776857729935291e-05,
      "loss": 0.7204,
      "mean_token_accuracy": 0.7866833582520485,
      "num_tokens": 231395862.0,
      "step": 7250
    },
    {
      "entropy": 0.7895010877400637,
      "epoch": 0.6671381993203029,
      "grad_norm": 0.15076889097690582,
      "learning_rate": 7.776551047321128e-05,
      "loss": 0.7825,
      "mean_token_accuracy": 0.7661961354315281,
      "num_tokens": 231427168.0,
      "step": 7251
    },
    {
      "entropy": 0.7408136054873466,
      "epoch": 0.6672302056917442,
      "grad_norm": 0.156294584274292,
      "learning_rate": 7.776244364706965e-05,
      "loss": 0.7059,
      "mean_token_accuracy": 0.7922790497541428,
      "num_tokens": 231459394.0,
      "step": 7252
    },
    {
      "entropy": 0.7724322006106377,
      "epoch": 0.6673222120631854,
      "grad_norm": 0.15439550578594208,
      "learning_rate": 7.775937682092802e-05,
      "loss": 0.7649,
      "mean_token_accuracy": 0.7772329859435558,
      "num_tokens": 231491401.0,
      "step": 7253
    },
    {
      "entropy": 0.801052276045084,
      "epoch": 0.6674142184346266,
      "grad_norm": 0.15492068231105804,
      "learning_rate": 7.77563099947864e-05,
      "loss": 0.7956,
      "mean_token_accuracy": 0.7704662978649139,
      "num_tokens": 231523125.0,
      "step": 7254
    },
    {
      "entropy": 0.7652567084878683,
      "epoch": 0.6675062248060678,
      "grad_norm": 0.16172677278518677,
      "learning_rate": 7.775324316864478e-05,
      "loss": 0.7568,
      "mean_token_accuracy": 0.7861545160412788,
      "num_tokens": 231554201.0,
      "step": 7255
    },
    {
      "entropy": 0.8538478650152683,
      "epoch": 0.667598231177509,
      "grad_norm": 0.17005343735218048,
      "learning_rate": 7.775017634250315e-05,
      "loss": 0.8385,
      "mean_token_accuracy": 0.7563003189861774,
      "num_tokens": 231585270.0,
      "step": 7256
    },
    {
      "entropy": 0.7900725472718477,
      "epoch": 0.6676902375489503,
      "grad_norm": 0.16013045608997345,
      "learning_rate": 7.774710951636151e-05,
      "loss": 0.7756,
      "mean_token_accuracy": 0.7709344513714314,
      "num_tokens": 231616688.0,
      "step": 7257
    },
    {
      "entropy": 0.709467800334096,
      "epoch": 0.6677822439203915,
      "grad_norm": 0.1559496372938156,
      "learning_rate": 7.77440426902199e-05,
      "loss": 0.6838,
      "mean_token_accuracy": 0.797319695353508,
      "num_tokens": 231648996.0,
      "step": 7258
    },
    {
      "entropy": 0.753263033926487,
      "epoch": 0.6678742502918327,
      "grad_norm": 0.15340104699134827,
      "learning_rate": 7.774097586407826e-05,
      "loss": 0.7385,
      "mean_token_accuracy": 0.7812292724847794,
      "num_tokens": 231681064.0,
      "step": 7259
    },
    {
      "entropy": 0.6586194131523371,
      "epoch": 0.6679662566632739,
      "grad_norm": 0.15376117825508118,
      "learning_rate": 7.773790903793665e-05,
      "loss": 0.639,
      "mean_token_accuracy": 0.8085504956543446,
      "num_tokens": 231712706.0,
      "step": 7260
    },
    {
      "entropy": 0.7540621794760227,
      "epoch": 0.6680582630347152,
      "grad_norm": 0.16874273121356964,
      "learning_rate": 7.773484221179503e-05,
      "loss": 0.7373,
      "mean_token_accuracy": 0.7819127701222897,
      "num_tokens": 231744091.0,
      "step": 7261
    },
    {
      "entropy": 0.7254529390484095,
      "epoch": 0.6681502694061564,
      "grad_norm": 0.16007472574710846,
      "learning_rate": 7.77317753856534e-05,
      "loss": 0.7129,
      "mean_token_accuracy": 0.7873449809849262,
      "num_tokens": 231775910.0,
      "step": 7262
    },
    {
      "entropy": 0.729896318167448,
      "epoch": 0.6682422757775976,
      "grad_norm": 0.16253864765167236,
      "learning_rate": 7.772870855951176e-05,
      "loss": 0.7312,
      "mean_token_accuracy": 0.7798079028725624,
      "num_tokens": 231807748.0,
      "step": 7263
    },
    {
      "entropy": 0.7525919014587998,
      "epoch": 0.6683342821490388,
      "grad_norm": 0.15485605597496033,
      "learning_rate": 7.772564173337013e-05,
      "loss": 0.7509,
      "mean_token_accuracy": 0.7812438830733299,
      "num_tokens": 231839698.0,
      "step": 7264
    },
    {
      "entropy": 0.7367272283881903,
      "epoch": 0.66842628852048,
      "grad_norm": 0.1737295389175415,
      "learning_rate": 7.772257490722851e-05,
      "loss": 0.7258,
      "mean_token_accuracy": 0.7821048982441425,
      "num_tokens": 231870996.0,
      "step": 7265
    },
    {
      "entropy": 0.7164717838168144,
      "epoch": 0.6685182948919213,
      "grad_norm": 0.15275603532791138,
      "learning_rate": 7.771950808108689e-05,
      "loss": 0.7257,
      "mean_token_accuracy": 0.7813126593828201,
      "num_tokens": 231903527.0,
      "step": 7266
    },
    {
      "entropy": 0.7405095305293798,
      "epoch": 0.6686103012633625,
      "grad_norm": 0.15390346944332123,
      "learning_rate": 7.771644125494526e-05,
      "loss": 0.7317,
      "mean_token_accuracy": 0.78241853043437,
      "num_tokens": 231936147.0,
      "step": 7267
    },
    {
      "entropy": 0.8274130281060934,
      "epoch": 0.6687023076348038,
      "grad_norm": 0.14870071411132812,
      "learning_rate": 7.771337442880363e-05,
      "loss": 0.8194,
      "mean_token_accuracy": 0.7565251179039478,
      "num_tokens": 231968046.0,
      "step": 7268
    },
    {
      "entropy": 0.8527082204818726,
      "epoch": 0.6687943140062449,
      "grad_norm": 0.16197599470615387,
      "learning_rate": 7.771030760266201e-05,
      "loss": 0.8397,
      "mean_token_accuracy": 0.7522659935057163,
      "num_tokens": 231999414.0,
      "step": 7269
    },
    {
      "entropy": 0.7781840432435274,
      "epoch": 0.6688863203776861,
      "grad_norm": 0.15549057722091675,
      "learning_rate": 7.770724077652038e-05,
      "loss": 0.7719,
      "mean_token_accuracy": 0.770779587328434,
      "num_tokens": 232031827.0,
      "step": 7270
    },
    {
      "entropy": 0.8175007048994303,
      "epoch": 0.6689783267491274,
      "grad_norm": 0.15912765264511108,
      "learning_rate": 7.770417395037876e-05,
      "loss": 0.8109,
      "mean_token_accuracy": 0.7656826004385948,
      "num_tokens": 232063921.0,
      "step": 7271
    },
    {
      "entropy": 0.8588179498910904,
      "epoch": 0.6690703331205686,
      "grad_norm": 0.1693842113018036,
      "learning_rate": 7.770110712423713e-05,
      "loss": 0.8531,
      "mean_token_accuracy": 0.7585954181849957,
      "num_tokens": 232094763.0,
      "step": 7272
    },
    {
      "entropy": 0.7388372384011745,
      "epoch": 0.6691623394920099,
      "grad_norm": 0.15534831583499908,
      "learning_rate": 7.769804029809551e-05,
      "loss": 0.7297,
      "mean_token_accuracy": 0.7844401970505714,
      "num_tokens": 232126661.0,
      "step": 7273
    },
    {
      "entropy": 0.7230708561837673,
      "epoch": 0.669254345863451,
      "grad_norm": 0.16624729335308075,
      "learning_rate": 7.769497347195388e-05,
      "loss": 0.704,
      "mean_token_accuracy": 0.794043131172657,
      "num_tokens": 232158795.0,
      "step": 7274
    },
    {
      "entropy": 0.7357456907629967,
      "epoch": 0.6693463522348923,
      "grad_norm": 0.16089074313640594,
      "learning_rate": 7.769190664581226e-05,
      "loss": 0.7224,
      "mean_token_accuracy": 0.7911581546068192,
      "num_tokens": 232190933.0,
      "step": 7275
    },
    {
      "entropy": 0.8205534480512142,
      "epoch": 0.6694383586063335,
      "grad_norm": 0.16480417549610138,
      "learning_rate": 7.768883981967063e-05,
      "loss": 0.8096,
      "mean_token_accuracy": 0.7643824368715286,
      "num_tokens": 232222092.0,
      "step": 7276
    },
    {
      "entropy": 0.7591274045407772,
      "epoch": 0.6695303649777747,
      "grad_norm": 0.15514419972896576,
      "learning_rate": 7.768577299352901e-05,
      "loss": 0.7428,
      "mean_token_accuracy": 0.7807882651686668,
      "num_tokens": 232254249.0,
      "step": 7277
    },
    {
      "entropy": 0.770433759316802,
      "epoch": 0.669622371349216,
      "grad_norm": 0.1576274186372757,
      "learning_rate": 7.768270616738738e-05,
      "loss": 0.7629,
      "mean_token_accuracy": 0.7791608534753323,
      "num_tokens": 232285975.0,
      "step": 7278
    },
    {
      "entropy": 0.7200944554060698,
      "epoch": 0.6697143777206571,
      "grad_norm": 0.15370804071426392,
      "learning_rate": 7.767963934124574e-05,
      "loss": 0.722,
      "mean_token_accuracy": 0.7930160723626614,
      "num_tokens": 232317592.0,
      "step": 7279
    },
    {
      "entropy": 0.6505838204175234,
      "epoch": 0.6698063840920984,
      "grad_norm": 0.1540200561285019,
      "learning_rate": 7.767657251510412e-05,
      "loss": 0.6184,
      "mean_token_accuracy": 0.811643410474062,
      "num_tokens": 232348654.0,
      "step": 7280
    },
    {
      "entropy": 0.9034317396581173,
      "epoch": 0.6698983904635396,
      "grad_norm": 0.16569779813289642,
      "learning_rate": 7.76735056889625e-05,
      "loss": 0.8743,
      "mean_token_accuracy": 0.7481204755604267,
      "num_tokens": 232379648.0,
      "step": 7281
    },
    {
      "entropy": 0.7662601172924042,
      "epoch": 0.6699903968349809,
      "grad_norm": 0.15486879646778107,
      "learning_rate": 7.767043886282087e-05,
      "loss": 0.7577,
      "mean_token_accuracy": 0.7798598520457745,
      "num_tokens": 232411356.0,
      "step": 7282
    },
    {
      "entropy": 0.7969679497182369,
      "epoch": 0.6700824032064221,
      "grad_norm": 0.15850776433944702,
      "learning_rate": 7.766737203667924e-05,
      "loss": 0.7855,
      "mean_token_accuracy": 0.7689728885889053,
      "num_tokens": 232443395.0,
      "step": 7283
    },
    {
      "entropy": 0.8295304859057069,
      "epoch": 0.6701744095778632,
      "grad_norm": 0.16234460473060608,
      "learning_rate": 7.766430521053761e-05,
      "loss": 0.8205,
      "mean_token_accuracy": 0.7664772905409336,
      "num_tokens": 232476022.0,
      "step": 7284
    },
    {
      "entropy": 0.7408042503520846,
      "epoch": 0.6702664159493045,
      "grad_norm": 0.16939599812030792,
      "learning_rate": 7.766123838439599e-05,
      "loss": 0.7251,
      "mean_token_accuracy": 0.7844846732914448,
      "num_tokens": 232507100.0,
      "step": 7285
    },
    {
      "entropy": 0.789796981960535,
      "epoch": 0.6703584223207457,
      "grad_norm": 0.17612802982330322,
      "learning_rate": 7.765817155825437e-05,
      "loss": 0.7857,
      "mean_token_accuracy": 0.7665531747043133,
      "num_tokens": 232539262.0,
      "step": 7286
    },
    {
      "entropy": 0.7131473235785961,
      "epoch": 0.670450428692187,
      "grad_norm": 0.17039760947227478,
      "learning_rate": 7.765510473211274e-05,
      "loss": 0.7113,
      "mean_token_accuracy": 0.7898493371903896,
      "num_tokens": 232570690.0,
      "step": 7287
    },
    {
      "entropy": 0.7258702907711267,
      "epoch": 0.6705424350636282,
      "grad_norm": 0.14657074213027954,
      "learning_rate": 7.765203790597111e-05,
      "loss": 0.6923,
      "mean_token_accuracy": 0.7945442199707031,
      "num_tokens": 232602354.0,
      "step": 7288
    },
    {
      "entropy": 0.6553861517459154,
      "epoch": 0.6706344414350693,
      "grad_norm": 0.14749933779239655,
      "learning_rate": 7.764897107982949e-05,
      "loss": 0.6214,
      "mean_token_accuracy": 0.8122338615357876,
      "num_tokens": 232634007.0,
      "step": 7289
    },
    {
      "entropy": 0.7940464001148939,
      "epoch": 0.6707264478065106,
      "grad_norm": 0.16539432108402252,
      "learning_rate": 7.764590425368786e-05,
      "loss": 0.7981,
      "mean_token_accuracy": 0.7660678997635841,
      "num_tokens": 232666328.0,
      "step": 7290
    },
    {
      "entropy": 0.8126280177384615,
      "epoch": 0.6708184541779518,
      "grad_norm": 0.15779581665992737,
      "learning_rate": 7.764283742754624e-05,
      "loss": 0.8006,
      "mean_token_accuracy": 0.7704895362257957,
      "num_tokens": 232698397.0,
      "step": 7291
    },
    {
      "entropy": 0.7809961251914501,
      "epoch": 0.6709104605493931,
      "grad_norm": 0.15438701212406158,
      "learning_rate": 7.763977060140462e-05,
      "loss": 0.7723,
      "mean_token_accuracy": 0.7715564705431461,
      "num_tokens": 232730518.0,
      "step": 7292
    },
    {
      "entropy": 0.7444243524223566,
      "epoch": 0.6710024669208343,
      "grad_norm": 0.15219172835350037,
      "learning_rate": 7.763670377526299e-05,
      "loss": 0.7441,
      "mean_token_accuracy": 0.7819833122193813,
      "num_tokens": 232762391.0,
      "step": 7293
    },
    {
      "entropy": 0.7519662622362375,
      "epoch": 0.6710944732922755,
      "grad_norm": 0.16198135912418365,
      "learning_rate": 7.763363694912136e-05,
      "loss": 0.7305,
      "mean_token_accuracy": 0.7830474115908146,
      "num_tokens": 232794561.0,
      "step": 7294
    },
    {
      "entropy": 0.787167239934206,
      "epoch": 0.6711864796637167,
      "grad_norm": 0.1500452607870102,
      "learning_rate": 7.763057012297972e-05,
      "loss": 0.7815,
      "mean_token_accuracy": 0.7681902311742306,
      "num_tokens": 232826704.0,
      "step": 7295
    },
    {
      "entropy": 0.7691180109977722,
      "epoch": 0.6712784860351579,
      "grad_norm": 0.15791429579257965,
      "learning_rate": 7.76275032968381e-05,
      "loss": 0.756,
      "mean_token_accuracy": 0.7805138751864433,
      "num_tokens": 232858698.0,
      "step": 7296
    },
    {
      "entropy": 0.6910213679075241,
      "epoch": 0.6713704924065992,
      "grad_norm": 0.16142602264881134,
      "learning_rate": 7.762443647069649e-05,
      "loss": 0.676,
      "mean_token_accuracy": 0.7984922677278519,
      "num_tokens": 232890482.0,
      "step": 7297
    },
    {
      "entropy": 0.7946796268224716,
      "epoch": 0.6714624987780404,
      "grad_norm": 0.16403332352638245,
      "learning_rate": 7.762136964455485e-05,
      "loss": 0.8058,
      "mean_token_accuracy": 0.7669268138706684,
      "num_tokens": 232922181.0,
      "step": 7298
    },
    {
      "entropy": 0.7242791056632996,
      "epoch": 0.6715545051494816,
      "grad_norm": 0.1643524467945099,
      "learning_rate": 7.761830281841322e-05,
      "loss": 0.7237,
      "mean_token_accuracy": 0.7894994392991066,
      "num_tokens": 232954124.0,
      "step": 7299
    },
    {
      "entropy": 0.7184069529175758,
      "epoch": 0.6716465115209228,
      "grad_norm": 0.1661616861820221,
      "learning_rate": 7.76152359922716e-05,
      "loss": 0.6818,
      "mean_token_accuracy": 0.7933006137609482,
      "num_tokens": 232985548.0,
      "step": 7300
    },
    {
      "entropy": 0.8254362158477306,
      "epoch": 0.671738517892364,
      "grad_norm": 0.16149672865867615,
      "learning_rate": 7.761216916612997e-05,
      "loss": 0.8189,
      "mean_token_accuracy": 0.7614533565938473,
      "num_tokens": 233017035.0,
      "step": 7301
    },
    {
      "entropy": 0.7790338639169931,
      "epoch": 0.6718305242638053,
      "grad_norm": 0.16101597249507904,
      "learning_rate": 7.760910233998835e-05,
      "loss": 0.7876,
      "mean_token_accuracy": 0.7710938900709152,
      "num_tokens": 233049380.0,
      "step": 7302
    },
    {
      "entropy": 0.8246599286794662,
      "epoch": 0.6719225306352465,
      "grad_norm": 0.16056320071220398,
      "learning_rate": 7.760603551384672e-05,
      "loss": 0.8248,
      "mean_token_accuracy": 0.7565088495612144,
      "num_tokens": 233081210.0,
      "step": 7303
    },
    {
      "entropy": 0.7745017409324646,
      "epoch": 0.6720145370066877,
      "grad_norm": 0.1651545912027359,
      "learning_rate": 7.76029686877051e-05,
      "loss": 0.7719,
      "mean_token_accuracy": 0.7752514779567719,
      "num_tokens": 233113418.0,
      "step": 7304
    },
    {
      "entropy": 0.7006810754537582,
      "epoch": 0.6721065433781289,
      "grad_norm": 0.16423113644123077,
      "learning_rate": 7.759990186156347e-05,
      "loss": 0.6983,
      "mean_token_accuracy": 0.7969676181674004,
      "num_tokens": 233145965.0,
      "step": 7305
    },
    {
      "entropy": 0.7475933246314526,
      "epoch": 0.6721985497495702,
      "grad_norm": 0.15504081547260284,
      "learning_rate": 7.759683503542184e-05,
      "loss": 0.7368,
      "mean_token_accuracy": 0.7851454839110374,
      "num_tokens": 233178200.0,
      "step": 7306
    },
    {
      "entropy": 0.7681633308529854,
      "epoch": 0.6722905561210114,
      "grad_norm": 0.15190941095352173,
      "learning_rate": 7.759376820928022e-05,
      "loss": 0.7569,
      "mean_token_accuracy": 0.7787247598171234,
      "num_tokens": 233210586.0,
      "step": 7307
    },
    {
      "entropy": 0.7870308756828308,
      "epoch": 0.6723825624924527,
      "grad_norm": 0.15954312682151794,
      "learning_rate": 7.75907013831386e-05,
      "loss": 0.8066,
      "mean_token_accuracy": 0.765832494944334,
      "num_tokens": 233242223.0,
      "step": 7308
    },
    {
      "entropy": 0.7393241375684738,
      "epoch": 0.6724745688638938,
      "grad_norm": 0.1582375019788742,
      "learning_rate": 7.758763455699697e-05,
      "loss": 0.7261,
      "mean_token_accuracy": 0.7837041653692722,
      "num_tokens": 233273859.0,
      "step": 7309
    },
    {
      "entropy": 0.8885337486863136,
      "epoch": 0.672566575235335,
      "grad_norm": 0.1617864966392517,
      "learning_rate": 7.758456773085534e-05,
      "loss": 0.8701,
      "mean_token_accuracy": 0.7481271103024483,
      "num_tokens": 233305106.0,
      "step": 7310
    },
    {
      "entropy": 0.9024527706205845,
      "epoch": 0.6726585816067763,
      "grad_norm": 0.15655925869941711,
      "learning_rate": 7.75815009047137e-05,
      "loss": 0.8935,
      "mean_token_accuracy": 0.7409311793744564,
      "num_tokens": 233337061.0,
      "step": 7311
    },
    {
      "entropy": 0.780497346073389,
      "epoch": 0.6727505879782175,
      "grad_norm": 0.1597587615251541,
      "learning_rate": 7.75784340785721e-05,
      "loss": 0.7581,
      "mean_token_accuracy": 0.7754996679723263,
      "num_tokens": 233369431.0,
      "step": 7312
    },
    {
      "entropy": 0.6709317583590746,
      "epoch": 0.6728425943496588,
      "grad_norm": 0.1514456868171692,
      "learning_rate": 7.757536725243047e-05,
      "loss": 0.6372,
      "mean_token_accuracy": 0.8079072423279285,
      "num_tokens": 233400887.0,
      "step": 7313
    },
    {
      "entropy": 0.7948476057499647,
      "epoch": 0.6729346007210999,
      "grad_norm": 0.1620945781469345,
      "learning_rate": 7.757230042628884e-05,
      "loss": 0.7643,
      "mean_token_accuracy": 0.7804357558488846,
      "num_tokens": 233432503.0,
      "step": 7314
    },
    {
      "entropy": 0.7471798351034522,
      "epoch": 0.6730266070925411,
      "grad_norm": 0.15536236763000488,
      "learning_rate": 7.75692336001472e-05,
      "loss": 0.7263,
      "mean_token_accuracy": 0.7864012233912945,
      "num_tokens": 233464494.0,
      "step": 7315
    },
    {
      "entropy": 0.7459684275090694,
      "epoch": 0.6731186134639824,
      "grad_norm": 0.16447016596794128,
      "learning_rate": 7.756616677400558e-05,
      "loss": 0.7338,
      "mean_token_accuracy": 0.7845535576343536,
      "num_tokens": 233496718.0,
      "step": 7316
    },
    {
      "entropy": 0.7242030501365662,
      "epoch": 0.6732106198354236,
      "grad_norm": 0.1545388400554657,
      "learning_rate": 7.756309994786397e-05,
      "loss": 0.7092,
      "mean_token_accuracy": 0.7924113683402538,
      "num_tokens": 233528621.0,
      "step": 7317
    },
    {
      "entropy": 0.7565589807927608,
      "epoch": 0.6733026262068649,
      "grad_norm": 0.1619139462709427,
      "learning_rate": 7.756003312172233e-05,
      "loss": 0.7654,
      "mean_token_accuracy": 0.7739979512989521,
      "num_tokens": 233561235.0,
      "step": 7318
    },
    {
      "entropy": 0.7542889788746834,
      "epoch": 0.673394632578306,
      "grad_norm": 0.15813805162906647,
      "learning_rate": 7.75569662955807e-05,
      "loss": 0.7379,
      "mean_token_accuracy": 0.7827187590301037,
      "num_tokens": 233593671.0,
      "step": 7319
    },
    {
      "entropy": 0.8347251359373331,
      "epoch": 0.6734866389497473,
      "grad_norm": 0.15935246646404266,
      "learning_rate": 7.755389946943908e-05,
      "loss": 0.8245,
      "mean_token_accuracy": 0.7580458000302315,
      "num_tokens": 233625532.0,
      "step": 7320
    },
    {
      "entropy": 0.6262325011193752,
      "epoch": 0.6735786453211885,
      "grad_norm": 0.16029663383960724,
      "learning_rate": 7.755083264329745e-05,
      "loss": 0.6151,
      "mean_token_accuracy": 0.8174208663403988,
      "num_tokens": 233657050.0,
      "step": 7321
    },
    {
      "entropy": 0.7095934972167015,
      "epoch": 0.6736706516926297,
      "grad_norm": 0.16892899572849274,
      "learning_rate": 7.754776581715583e-05,
      "loss": 0.7134,
      "mean_token_accuracy": 0.790358129888773,
      "num_tokens": 233689335.0,
      "step": 7322
    },
    {
      "entropy": 0.7022652104496956,
      "epoch": 0.673762658064071,
      "grad_norm": 0.15540790557861328,
      "learning_rate": 7.754469899101421e-05,
      "loss": 0.6888,
      "mean_token_accuracy": 0.7960202433168888,
      "num_tokens": 233721436.0,
      "step": 7323
    },
    {
      "entropy": 0.6592752281576395,
      "epoch": 0.6738546644355121,
      "grad_norm": 0.1642361432313919,
      "learning_rate": 7.754163216487258e-05,
      "loss": 0.6457,
      "mean_token_accuracy": 0.8107647337019444,
      "num_tokens": 233753367.0,
      "step": 7324
    },
    {
      "entropy": 0.7589376848191023,
      "epoch": 0.6739466708069534,
      "grad_norm": 0.17393146455287933,
      "learning_rate": 7.753856533873095e-05,
      "loss": 0.7777,
      "mean_token_accuracy": 0.7711551897227764,
      "num_tokens": 233785691.0,
      "step": 7325
    },
    {
      "entropy": 0.8542578332126141,
      "epoch": 0.6740386771783946,
      "grad_norm": 0.15906842052936554,
      "learning_rate": 7.753549851258932e-05,
      "loss": 0.8574,
      "mean_token_accuracy": 0.7518262043595314,
      "num_tokens": 233817109.0,
      "step": 7326
    },
    {
      "entropy": 0.8454954698681831,
      "epoch": 0.6741306835498359,
      "grad_norm": 0.16477827727794647,
      "learning_rate": 7.75324316864477e-05,
      "loss": 0.8379,
      "mean_token_accuracy": 0.7587021142244339,
      "num_tokens": 233849116.0,
      "step": 7327
    },
    {
      "entropy": 0.7006620764732361,
      "epoch": 0.6742226899212771,
      "grad_norm": 0.1546269804239273,
      "learning_rate": 7.752936486030608e-05,
      "loss": 0.6836,
      "mean_token_accuracy": 0.8029907718300819,
      "num_tokens": 233881708.0,
      "step": 7328
    },
    {
      "entropy": 0.7417717780917883,
      "epoch": 0.6743146962927182,
      "grad_norm": 0.1542491912841797,
      "learning_rate": 7.752629803416445e-05,
      "loss": 0.7224,
      "mean_token_accuracy": 0.7890417389571667,
      "num_tokens": 233913194.0,
      "step": 7329
    },
    {
      "entropy": 0.7937721908092499,
      "epoch": 0.6744067026641595,
      "grad_norm": 0.15191398561000824,
      "learning_rate": 7.752323120802282e-05,
      "loss": 0.7896,
      "mean_token_accuracy": 0.7680932022631168,
      "num_tokens": 233945481.0,
      "step": 7330
    },
    {
      "entropy": 0.7105566551908851,
      "epoch": 0.6744987090356007,
      "grad_norm": 0.14469093084335327,
      "learning_rate": 7.75201643818812e-05,
      "loss": 0.7012,
      "mean_token_accuracy": 0.7915611378848553,
      "num_tokens": 233977407.0,
      "step": 7331
    },
    {
      "entropy": 0.7399148195981979,
      "epoch": 0.674590715407042,
      "grad_norm": 0.15764349699020386,
      "learning_rate": 7.751709755573957e-05,
      "loss": 0.7253,
      "mean_token_accuracy": 0.787391897290945,
      "num_tokens": 234009608.0,
      "step": 7332
    },
    {
      "entropy": 0.695987006649375,
      "epoch": 0.6746827217784832,
      "grad_norm": 0.15157905220985413,
      "learning_rate": 7.751403072959795e-05,
      "loss": 0.6776,
      "mean_token_accuracy": 0.7995686531066895,
      "num_tokens": 234041303.0,
      "step": 7333
    },
    {
      "entropy": 0.7774870432913303,
      "epoch": 0.6747747281499243,
      "grad_norm": 0.16255372762680054,
      "learning_rate": 7.751096390345631e-05,
      "loss": 0.7801,
      "mean_token_accuracy": 0.7744915187358856,
      "num_tokens": 234072779.0,
      "step": 7334
    },
    {
      "entropy": 0.7469492480158806,
      "epoch": 0.6748667345213656,
      "grad_norm": 0.15168879926204681,
      "learning_rate": 7.75078970773147e-05,
      "loss": 0.7387,
      "mean_token_accuracy": 0.7755423448979855,
      "num_tokens": 234104201.0,
      "step": 7335
    },
    {
      "entropy": 0.7728137634694576,
      "epoch": 0.6749587408928068,
      "grad_norm": 0.16534842550754547,
      "learning_rate": 7.750483025117306e-05,
      "loss": 0.7635,
      "mean_token_accuracy": 0.7779363542795181,
      "num_tokens": 234134694.0,
      "step": 7336
    },
    {
      "entropy": 0.6825873721390963,
      "epoch": 0.6750507472642481,
      "grad_norm": 0.1499457061290741,
      "learning_rate": 7.750176342503143e-05,
      "loss": 0.6453,
      "mean_token_accuracy": 0.8053931929171085,
      "num_tokens": 234166896.0,
      "step": 7337
    },
    {
      "entropy": 0.7278263419866562,
      "epoch": 0.6751427536356893,
      "grad_norm": 0.1452217698097229,
      "learning_rate": 7.749869659888981e-05,
      "loss": 0.7054,
      "mean_token_accuracy": 0.7886778637766838,
      "num_tokens": 234199231.0,
      "step": 7338
    },
    {
      "entropy": 0.6503347475081682,
      "epoch": 0.6752347600071305,
      "grad_norm": 0.1501016765832901,
      "learning_rate": 7.74956297727482e-05,
      "loss": 0.6265,
      "mean_token_accuracy": 0.809617668390274,
      "num_tokens": 234230784.0,
      "step": 7339
    },
    {
      "entropy": 0.7445249557495117,
      "epoch": 0.6753267663785717,
      "grad_norm": 0.16659149527549744,
      "learning_rate": 7.749256294660656e-05,
      "loss": 0.7416,
      "mean_token_accuracy": 0.7833022475242615,
      "num_tokens": 234262749.0,
      "step": 7340
    },
    {
      "entropy": 0.8911267928779125,
      "epoch": 0.6754187727500129,
      "grad_norm": 0.15881893038749695,
      "learning_rate": 7.748949612046493e-05,
      "loss": 0.8811,
      "mean_token_accuracy": 0.7459121532738209,
      "num_tokens": 234294417.0,
      "step": 7341
    },
    {
      "entropy": 0.7276895046234131,
      "epoch": 0.6755107791214542,
      "grad_norm": 0.15152974426746368,
      "learning_rate": 7.74864292943233e-05,
      "loss": 0.7244,
      "mean_token_accuracy": 0.7848358154296875,
      "num_tokens": 234326935.0,
      "step": 7342
    },
    {
      "entropy": 0.9092724360525608,
      "epoch": 0.6756027854928954,
      "grad_norm": 0.16690486669540405,
      "learning_rate": 7.748336246818168e-05,
      "loss": 0.9065,
      "mean_token_accuracy": 0.7413832768797874,
      "num_tokens": 234359265.0,
      "step": 7343
    },
    {
      "entropy": 0.6926062125712633,
      "epoch": 0.6756947918643366,
      "grad_norm": 0.15047413110733032,
      "learning_rate": 7.748029564204006e-05,
      "loss": 0.6869,
      "mean_token_accuracy": 0.7946112304925919,
      "num_tokens": 234391162.0,
      "step": 7344
    },
    {
      "entropy": 0.6418612617999315,
      "epoch": 0.6757867982357778,
      "grad_norm": 0.15505586564540863,
      "learning_rate": 7.747722881589843e-05,
      "loss": 0.6195,
      "mean_token_accuracy": 0.8165277056396008,
      "num_tokens": 234423453.0,
      "step": 7345
    },
    {
      "entropy": 0.7666782438755035,
      "epoch": 0.675878804607219,
      "grad_norm": 0.15124613046646118,
      "learning_rate": 7.74741619897568e-05,
      "loss": 0.7654,
      "mean_token_accuracy": 0.7746237963438034,
      "num_tokens": 234455528.0,
      "step": 7346
    },
    {
      "entropy": 0.7443847712129354,
      "epoch": 0.6759708109786603,
      "grad_norm": 0.15459635853767395,
      "learning_rate": 7.747109516361518e-05,
      "loss": 0.7255,
      "mean_token_accuracy": 0.7872141450643539,
      "num_tokens": 234488001.0,
      "step": 7347
    },
    {
      "entropy": 0.7299632281064987,
      "epoch": 0.6760628173501015,
      "grad_norm": 0.1517024040222168,
      "learning_rate": 7.746802833747355e-05,
      "loss": 0.7295,
      "mean_token_accuracy": 0.7840221673250198,
      "num_tokens": 234519531.0,
      "step": 7348
    },
    {
      "entropy": 0.7200689939782023,
      "epoch": 0.6761548237215427,
      "grad_norm": 0.15484964847564697,
      "learning_rate": 7.746496151133193e-05,
      "loss": 0.7092,
      "mean_token_accuracy": 0.7893388234078884,
      "num_tokens": 234551563.0,
      "step": 7349
    },
    {
      "entropy": 0.6974947033450007,
      "epoch": 0.6762468300929839,
      "grad_norm": 0.17469030618667603,
      "learning_rate": 7.74618946851903e-05,
      "loss": 0.7009,
      "mean_token_accuracy": 0.7877567000687122,
      "num_tokens": 234583979.0,
      "step": 7350
    },
    {
      "entropy": 0.7674193233251572,
      "epoch": 0.6763388364644252,
      "grad_norm": 0.15580187737941742,
      "learning_rate": 7.745882785904868e-05,
      "loss": 0.7537,
      "mean_token_accuracy": 0.7773704454302788,
      "num_tokens": 234615523.0,
      "step": 7351
    },
    {
      "entropy": 0.786311587318778,
      "epoch": 0.6764308428358664,
      "grad_norm": 0.1600351184606552,
      "learning_rate": 7.745576103290704e-05,
      "loss": 0.7866,
      "mean_token_accuracy": 0.7688491642475128,
      "num_tokens": 234647186.0,
      "step": 7352
    },
    {
      "entropy": 0.7500151451677084,
      "epoch": 0.6765228492073077,
      "grad_norm": 0.15044352412223816,
      "learning_rate": 7.745269420676541e-05,
      "loss": 0.7337,
      "mean_token_accuracy": 0.7849756292998791,
      "num_tokens": 234679599.0,
      "step": 7353
    },
    {
      "entropy": 0.6053790990263224,
      "epoch": 0.6766148555787488,
      "grad_norm": 0.15317362546920776,
      "learning_rate": 7.744962738062381e-05,
      "loss": 0.5804,
      "mean_token_accuracy": 0.8274618983268738,
      "num_tokens": 234712218.0,
      "step": 7354
    },
    {
      "entropy": 0.764061763882637,
      "epoch": 0.67670686195019,
      "grad_norm": 0.15360121428966522,
      "learning_rate": 7.744656055448218e-05,
      "loss": 0.7418,
      "mean_token_accuracy": 0.7790607549250126,
      "num_tokens": 234744565.0,
      "step": 7355
    },
    {
      "entropy": 0.7120874617248774,
      "epoch": 0.6767988683216313,
      "grad_norm": 0.1623067557811737,
      "learning_rate": 7.744349372834054e-05,
      "loss": 0.7111,
      "mean_token_accuracy": 0.7900125980377197,
      "num_tokens": 234776301.0,
      "step": 7356
    },
    {
      "entropy": 0.7327252645045519,
      "epoch": 0.6768908746930725,
      "grad_norm": 0.16525514423847198,
      "learning_rate": 7.744042690219891e-05,
      "loss": 0.7257,
      "mean_token_accuracy": 0.7848835065960884,
      "num_tokens": 234807575.0,
      "step": 7357
    },
    {
      "entropy": 0.7936296705156565,
      "epoch": 0.6769828810645138,
      "grad_norm": 0.1650097370147705,
      "learning_rate": 7.743736007605729e-05,
      "loss": 0.7805,
      "mean_token_accuracy": 0.7738691568374634,
      "num_tokens": 234838840.0,
      "step": 7358
    },
    {
      "entropy": 0.6720875650644302,
      "epoch": 0.6770748874359549,
      "grad_norm": 0.148561030626297,
      "learning_rate": 7.743429324991567e-05,
      "loss": 0.6591,
      "mean_token_accuracy": 0.8040415085852146,
      "num_tokens": 234870990.0,
      "step": 7359
    },
    {
      "entropy": 0.818424366414547,
      "epoch": 0.6771668938073961,
      "grad_norm": 0.15591692924499512,
      "learning_rate": 7.743122642377404e-05,
      "loss": 0.8073,
      "mean_token_accuracy": 0.7626658193767071,
      "num_tokens": 234903076.0,
      "step": 7360
    },
    {
      "entropy": 0.7846440114080906,
      "epoch": 0.6772589001788374,
      "grad_norm": 0.15490517020225525,
      "learning_rate": 7.742815959763241e-05,
      "loss": 0.765,
      "mean_token_accuracy": 0.7750608958303928,
      "num_tokens": 234935552.0,
      "step": 7361
    },
    {
      "entropy": 0.7632138323970139,
      "epoch": 0.6773509065502786,
      "grad_norm": 0.1628345549106598,
      "learning_rate": 7.742509277149079e-05,
      "loss": 0.75,
      "mean_token_accuracy": 0.7810094505548477,
      "num_tokens": 234966948.0,
      "step": 7362
    },
    {
      "entropy": 0.7259329315274954,
      "epoch": 0.6774429129217199,
      "grad_norm": 0.1587395817041397,
      "learning_rate": 7.742202594534916e-05,
      "loss": 0.7132,
      "mean_token_accuracy": 0.788999866694212,
      "num_tokens": 234999017.0,
      "step": 7363
    },
    {
      "entropy": 0.7437162455171347,
      "epoch": 0.677534919293161,
      "grad_norm": 0.16720286011695862,
      "learning_rate": 7.741895911920754e-05,
      "loss": 0.7202,
      "mean_token_accuracy": 0.7864535674452782,
      "num_tokens": 235031420.0,
      "step": 7364
    },
    {
      "entropy": 0.7222743574529886,
      "epoch": 0.6776269256646023,
      "grad_norm": 0.15129750967025757,
      "learning_rate": 7.741589229306591e-05,
      "loss": 0.7156,
      "mean_token_accuracy": 0.7881825044751167,
      "num_tokens": 235062988.0,
      "step": 7365
    },
    {
      "entropy": 0.7101829387247562,
      "epoch": 0.6777189320360435,
      "grad_norm": 0.16149188578128815,
      "learning_rate": 7.741282546692429e-05,
      "loss": 0.6992,
      "mean_token_accuracy": 0.7929912582039833,
      "num_tokens": 235094728.0,
      "step": 7366
    },
    {
      "entropy": 0.8223206494003534,
      "epoch": 0.6778109384074847,
      "grad_norm": 0.16254597902297974,
      "learning_rate": 7.740975864078266e-05,
      "loss": 0.8115,
      "mean_token_accuracy": 0.7622009739279747,
      "num_tokens": 235126443.0,
      "step": 7367
    },
    {
      "entropy": 0.7745052948594093,
      "epoch": 0.677902944778926,
      "grad_norm": 0.1657203733921051,
      "learning_rate": 7.740669181464103e-05,
      "loss": 0.7738,
      "mean_token_accuracy": 0.775004904717207,
      "num_tokens": 235158978.0,
      "step": 7368
    },
    {
      "entropy": 0.7373640164732933,
      "epoch": 0.6779949511503671,
      "grad_norm": 0.15877671539783478,
      "learning_rate": 7.740362498849941e-05,
      "loss": 0.7084,
      "mean_token_accuracy": 0.7906962558627129,
      "num_tokens": 235190163.0,
      "step": 7369
    },
    {
      "entropy": 0.7404811549931765,
      "epoch": 0.6780869575218084,
      "grad_norm": 0.16041944921016693,
      "learning_rate": 7.740055816235779e-05,
      "loss": 0.7158,
      "mean_token_accuracy": 0.7833210043609142,
      "num_tokens": 235220961.0,
      "step": 7370
    },
    {
      "entropy": 0.8122723679989576,
      "epoch": 0.6781789638932496,
      "grad_norm": 0.15437331795692444,
      "learning_rate": 7.739749133621616e-05,
      "loss": 0.8311,
      "mean_token_accuracy": 0.7532959282398224,
      "num_tokens": 235253333.0,
      "step": 7371
    },
    {
      "entropy": 0.7822709502652287,
      "epoch": 0.6782709702646909,
      "grad_norm": 0.15718981623649597,
      "learning_rate": 7.739442451007452e-05,
      "loss": 0.7863,
      "mean_token_accuracy": 0.7692623622715473,
      "num_tokens": 235285383.0,
      "step": 7372
    },
    {
      "entropy": 0.7150296531617641,
      "epoch": 0.6783629766361321,
      "grad_norm": 0.15663668513298035,
      "learning_rate": 7.739135768393289e-05,
      "loss": 0.7051,
      "mean_token_accuracy": 0.7918086498975754,
      "num_tokens": 235317629.0,
      "step": 7373
    },
    {
      "entropy": 0.7110081128776073,
      "epoch": 0.6784549830075732,
      "grad_norm": 0.15520666539669037,
      "learning_rate": 7.738829085779127e-05,
      "loss": 0.7033,
      "mean_token_accuracy": 0.7945964857935905,
      "num_tokens": 235349667.0,
      "step": 7374
    },
    {
      "entropy": 0.8246515765786171,
      "epoch": 0.6785469893790145,
      "grad_norm": 0.1488068848848343,
      "learning_rate": 7.738522403164966e-05,
      "loss": 0.8335,
      "mean_token_accuracy": 0.7549607530236244,
      "num_tokens": 235381116.0,
      "step": 7375
    },
    {
      "entropy": 0.642595935612917,
      "epoch": 0.6786389957504557,
      "grad_norm": 0.15483611822128296,
      "learning_rate": 7.738215720550802e-05,
      "loss": 0.6354,
      "mean_token_accuracy": 0.8108202926814556,
      "num_tokens": 235413672.0,
      "step": 7376
    },
    {
      "entropy": 0.7914865501224995,
      "epoch": 0.678731002121897,
      "grad_norm": 0.15242330729961395,
      "learning_rate": 7.737909037936639e-05,
      "loss": 0.7816,
      "mean_token_accuracy": 0.7714317440986633,
      "num_tokens": 235446224.0,
      "step": 7377
    },
    {
      "entropy": 0.6818441599607468,
      "epoch": 0.6788230084933382,
      "grad_norm": 0.15817490220069885,
      "learning_rate": 7.737602355322477e-05,
      "loss": 0.659,
      "mean_token_accuracy": 0.8017994537949562,
      "num_tokens": 235477985.0,
      "step": 7378
    },
    {
      "entropy": 0.7212690748274326,
      "epoch": 0.6789150148647793,
      "grad_norm": 0.1585027575492859,
      "learning_rate": 7.737295672708314e-05,
      "loss": 0.7308,
      "mean_token_accuracy": 0.7823378071188927,
      "num_tokens": 235510701.0,
      "step": 7379
    },
    {
      "entropy": 0.7681048959493637,
      "epoch": 0.6790070212362206,
      "grad_norm": 0.1648036688566208,
      "learning_rate": 7.736988990094152e-05,
      "loss": 0.7525,
      "mean_token_accuracy": 0.779227364808321,
      "num_tokens": 235541850.0,
      "step": 7380
    },
    {
      "entropy": 0.7719811238348484,
      "epoch": 0.6790990276076618,
      "grad_norm": 0.15684059262275696,
      "learning_rate": 7.736682307479989e-05,
      "loss": 0.737,
      "mean_token_accuracy": 0.7852162830531597,
      "num_tokens": 235573056.0,
      "step": 7381
    },
    {
      "entropy": 0.875713549554348,
      "epoch": 0.6791910339791031,
      "grad_norm": 0.16381128132343292,
      "learning_rate": 7.736375624865827e-05,
      "loss": 0.8604,
      "mean_token_accuracy": 0.7539408914744854,
      "num_tokens": 235605534.0,
      "step": 7382
    },
    {
      "entropy": 0.6171670090407133,
      "epoch": 0.6792830403505443,
      "grad_norm": 0.15169212222099304,
      "learning_rate": 7.736068942251664e-05,
      "loss": 0.61,
      "mean_token_accuracy": 0.8162909895181656,
      "num_tokens": 235637885.0,
      "step": 7383
    },
    {
      "entropy": 0.7256463225930929,
      "epoch": 0.6793750467219855,
      "grad_norm": 0.16893388330936432,
      "learning_rate": 7.7357622596375e-05,
      "loss": 0.7146,
      "mean_token_accuracy": 0.7895727530121803,
      "num_tokens": 235670471.0,
      "step": 7384
    },
    {
      "entropy": 0.823699465021491,
      "epoch": 0.6794670530934267,
      "grad_norm": 0.15953756868839264,
      "learning_rate": 7.735455577023339e-05,
      "loss": 0.804,
      "mean_token_accuracy": 0.7680992223322392,
      "num_tokens": 235702361.0,
      "step": 7385
    },
    {
      "entropy": 0.7268590796738863,
      "epoch": 0.6795590594648679,
      "grad_norm": 0.1624906361103058,
      "learning_rate": 7.735148894409177e-05,
      "loss": 0.6987,
      "mean_token_accuracy": 0.7940700203180313,
      "num_tokens": 235733941.0,
      "step": 7386
    },
    {
      "entropy": 0.766047952696681,
      "epoch": 0.6796510658363092,
      "grad_norm": 0.16432468593120575,
      "learning_rate": 7.734842211795014e-05,
      "loss": 0.743,
      "mean_token_accuracy": 0.77813371270895,
      "num_tokens": 235766045.0,
      "step": 7387
    },
    {
      "entropy": 0.7712325798347592,
      "epoch": 0.6797430722077504,
      "grad_norm": 0.16628652811050415,
      "learning_rate": 7.73453552918085e-05,
      "loss": 0.7426,
      "mean_token_accuracy": 0.7843572646379471,
      "num_tokens": 235797910.0,
      "step": 7388
    },
    {
      "entropy": 0.6798674063757062,
      "epoch": 0.6798350785791916,
      "grad_norm": 0.15020619332790375,
      "learning_rate": 7.734228846566689e-05,
      "loss": 0.6552,
      "mean_token_accuracy": 0.804684616625309,
      "num_tokens": 235829227.0,
      "step": 7389
    },
    {
      "entropy": 0.6828232202678919,
      "epoch": 0.6799270849506328,
      "grad_norm": 0.1659419685602188,
      "learning_rate": 7.733922163952525e-05,
      "loss": 0.6784,
      "mean_token_accuracy": 0.8035678490996361,
      "num_tokens": 235861163.0,
      "step": 7390
    },
    {
      "entropy": 0.6995331533253193,
      "epoch": 0.680019091322074,
      "grad_norm": 0.15544739365577698,
      "learning_rate": 7.733615481338364e-05,
      "loss": 0.6677,
      "mean_token_accuracy": 0.801600743085146,
      "num_tokens": 235891940.0,
      "step": 7391
    },
    {
      "entropy": 0.8392040319740772,
      "epoch": 0.6801110976935153,
      "grad_norm": 0.17398576438426971,
      "learning_rate": 7.7333087987242e-05,
      "loss": 0.83,
      "mean_token_accuracy": 0.76132707670331,
      "num_tokens": 235924036.0,
      "step": 7392
    },
    {
      "entropy": 0.7760454900562763,
      "epoch": 0.6802031040649565,
      "grad_norm": 0.159798264503479,
      "learning_rate": 7.733002116110039e-05,
      "loss": 0.7631,
      "mean_token_accuracy": 0.7747703976929188,
      "num_tokens": 235955524.0,
      "step": 7393
    },
    {
      "entropy": 0.7062018942087889,
      "epoch": 0.6802951104363977,
      "grad_norm": 0.14948631823062897,
      "learning_rate": 7.732695433495875e-05,
      "loss": 0.697,
      "mean_token_accuracy": 0.7931702360510826,
      "num_tokens": 235987625.0,
      "step": 7394
    },
    {
      "entropy": 0.6638649608939886,
      "epoch": 0.6803871168078389,
      "grad_norm": 0.15372896194458008,
      "learning_rate": 7.732388750881713e-05,
      "loss": 0.6632,
      "mean_token_accuracy": 0.8074601106345654,
      "num_tokens": 236019506.0,
      "step": 7395
    },
    {
      "entropy": 0.635661518201232,
      "epoch": 0.6804791231792802,
      "grad_norm": 0.14886021614074707,
      "learning_rate": 7.73208206826755e-05,
      "loss": 0.6187,
      "mean_token_accuracy": 0.8137947209179401,
      "num_tokens": 236051787.0,
      "step": 7396
    },
    {
      "entropy": 0.6119104288518429,
      "epoch": 0.6805711295507214,
      "grad_norm": 0.16049449145793915,
      "learning_rate": 7.731775385653388e-05,
      "loss": 0.6084,
      "mean_token_accuracy": 0.8205111995339394,
      "num_tokens": 236083778.0,
      "step": 7397
    },
    {
      "entropy": 0.7971672639250755,
      "epoch": 0.6806631359221627,
      "grad_norm": 0.15742094814777374,
      "learning_rate": 7.731468703039225e-05,
      "loss": 0.7762,
      "mean_token_accuracy": 0.772256538271904,
      "num_tokens": 236116072.0,
      "step": 7398
    },
    {
      "entropy": 0.6678461786359549,
      "epoch": 0.6807551422936038,
      "grad_norm": 0.16092033684253693,
      "learning_rate": 7.731162020425062e-05,
      "loss": 0.6488,
      "mean_token_accuracy": 0.8057238832116127,
      "num_tokens": 236146714.0,
      "step": 7399
    },
    {
      "entropy": 0.6959844380617142,
      "epoch": 0.680847148665045,
      "grad_norm": 0.15299130976200104,
      "learning_rate": 7.7308553378109e-05,
      "loss": 0.6805,
      "mean_token_accuracy": 0.7999005578458309,
      "num_tokens": 236178782.0,
      "step": 7400
    },
    {
      "entropy": 0.6016603838652372,
      "epoch": 0.6809391550364863,
      "grad_norm": 0.14420616626739502,
      "learning_rate": 7.730548655196738e-05,
      "loss": 0.5867,
      "mean_token_accuracy": 0.8224248997867107,
      "num_tokens": 236210286.0,
      "step": 7401
    },
    {
      "entropy": 0.7560275513678789,
      "epoch": 0.6810311614079275,
      "grad_norm": 0.18273521959781647,
      "learning_rate": 7.730241972582575e-05,
      "loss": 0.7873,
      "mean_token_accuracy": 0.7722882218658924,
      "num_tokens": 236241555.0,
      "step": 7402
    },
    {
      "entropy": 0.6690878123044968,
      "epoch": 0.6811231677793688,
      "grad_norm": 0.15039797127246857,
      "learning_rate": 7.729935289968412e-05,
      "loss": 0.6468,
      "mean_token_accuracy": 0.8096030503511429,
      "num_tokens": 236274066.0,
      "step": 7403
    },
    {
      "entropy": 0.8529013060033321,
      "epoch": 0.6812151741508099,
      "grad_norm": 0.1619715690612793,
      "learning_rate": 7.729628607354249e-05,
      "loss": 0.849,
      "mean_token_accuracy": 0.7575186155736446,
      "num_tokens": 236306373.0,
      "step": 7404
    },
    {
      "entropy": 0.7696916908025742,
      "epoch": 0.6813071805222511,
      "grad_norm": 0.15448977053165436,
      "learning_rate": 7.729321924740087e-05,
      "loss": 0.7757,
      "mean_token_accuracy": 0.7674227505922318,
      "num_tokens": 236337767.0,
      "step": 7405
    },
    {
      "entropy": 0.7396837025880814,
      "epoch": 0.6813991868936924,
      "grad_norm": 0.15564779937267303,
      "learning_rate": 7.729015242125925e-05,
      "loss": 0.7317,
      "mean_token_accuracy": 0.7854445800185204,
      "num_tokens": 236369134.0,
      "step": 7406
    },
    {
      "entropy": 0.7737117586657405,
      "epoch": 0.6814911932651336,
      "grad_norm": 0.15901023149490356,
      "learning_rate": 7.728708559511762e-05,
      "loss": 0.7462,
      "mean_token_accuracy": 0.7778505943715572,
      "num_tokens": 236400351.0,
      "step": 7407
    },
    {
      "entropy": 0.843220129609108,
      "epoch": 0.6815831996365749,
      "grad_norm": 0.15162770450115204,
      "learning_rate": 7.728401876897598e-05,
      "loss": 0.8206,
      "mean_token_accuracy": 0.7597316056489944,
      "num_tokens": 236432546.0,
      "step": 7408
    },
    {
      "entropy": 0.6831870079040527,
      "epoch": 0.681675206008016,
      "grad_norm": 0.1614047735929489,
      "learning_rate": 7.728095194283437e-05,
      "loss": 0.6704,
      "mean_token_accuracy": 0.8023987673223019,
      "num_tokens": 236464975.0,
      "step": 7409
    },
    {
      "entropy": 0.7233686670660973,
      "epoch": 0.6817672123794573,
      "grad_norm": 0.15500709414482117,
      "learning_rate": 7.727788511669273e-05,
      "loss": 0.704,
      "mean_token_accuracy": 0.7906909883022308,
      "num_tokens": 236496013.0,
      "step": 7410
    },
    {
      "entropy": 0.6380361337214708,
      "epoch": 0.6818592187508985,
      "grad_norm": 0.14802676439285278,
      "learning_rate": 7.727481829055112e-05,
      "loss": 0.6116,
      "mean_token_accuracy": 0.8167826272547245,
      "num_tokens": 236528308.0,
      "step": 7411
    },
    {
      "entropy": 0.7743168147280812,
      "epoch": 0.6819512251223397,
      "grad_norm": 0.1636095494031906,
      "learning_rate": 7.727175146440948e-05,
      "loss": 0.7601,
      "mean_token_accuracy": 0.7783494032919407,
      "num_tokens": 236560121.0,
      "step": 7412
    },
    {
      "entropy": 0.660091632977128,
      "epoch": 0.682043231493781,
      "grad_norm": 0.15833014249801636,
      "learning_rate": 7.726868463826786e-05,
      "loss": 0.6498,
      "mean_token_accuracy": 0.804757334291935,
      "num_tokens": 236592123.0,
      "step": 7413
    },
    {
      "entropy": 0.8927335180342197,
      "epoch": 0.6821352378652221,
      "grad_norm": 0.16460241377353668,
      "learning_rate": 7.726561781212623e-05,
      "loss": 0.8852,
      "mean_token_accuracy": 0.7455509267747402,
      "num_tokens": 236624518.0,
      "step": 7414
    },
    {
      "entropy": 0.810508731752634,
      "epoch": 0.6822272442366634,
      "grad_norm": 0.16621184349060059,
      "learning_rate": 7.72625509859846e-05,
      "loss": 0.8154,
      "mean_token_accuracy": 0.7633696272969246,
      "num_tokens": 236656707.0,
      "step": 7415
    },
    {
      "entropy": 0.7233202699571848,
      "epoch": 0.6823192506081046,
      "grad_norm": 0.15641160309314728,
      "learning_rate": 7.725948415984298e-05,
      "loss": 0.7029,
      "mean_token_accuracy": 0.7911736816167831,
      "num_tokens": 236688125.0,
      "step": 7416
    },
    {
      "entropy": 0.689668889157474,
      "epoch": 0.6824112569795459,
      "grad_norm": 0.145370215177536,
      "learning_rate": 7.725641733370136e-05,
      "loss": 0.6715,
      "mean_token_accuracy": 0.8011535443365574,
      "num_tokens": 236720280.0,
      "step": 7417
    },
    {
      "entropy": 0.6536412835121155,
      "epoch": 0.6825032633509871,
      "grad_norm": 0.14593450725078583,
      "learning_rate": 7.725335050755973e-05,
      "loss": 0.6267,
      "mean_token_accuracy": 0.8153686635196209,
      "num_tokens": 236752660.0,
      "step": 7418
    },
    {
      "entropy": 0.6818542815744877,
      "epoch": 0.6825952697224282,
      "grad_norm": 0.15293274819850922,
      "learning_rate": 7.72502836814181e-05,
      "loss": 0.6621,
      "mean_token_accuracy": 0.8011180609464645,
      "num_tokens": 236784462.0,
      "step": 7419
    },
    {
      "entropy": 0.7371944412589073,
      "epoch": 0.6826872760938695,
      "grad_norm": 0.15659622848033905,
      "learning_rate": 7.724721685527648e-05,
      "loss": 0.7305,
      "mean_token_accuracy": 0.7859303876757622,
      "num_tokens": 236816222.0,
      "step": 7420
    },
    {
      "entropy": 0.7761557325720787,
      "epoch": 0.6827792824653107,
      "grad_norm": 0.152645543217659,
      "learning_rate": 7.724415002913485e-05,
      "loss": 0.7738,
      "mean_token_accuracy": 0.7713887766003609,
      "num_tokens": 236848416.0,
      "step": 7421
    },
    {
      "entropy": 0.7194861173629761,
      "epoch": 0.682871288836752,
      "grad_norm": 0.1548376828432083,
      "learning_rate": 7.724108320299323e-05,
      "loss": 0.7069,
      "mean_token_accuracy": 0.787356648594141,
      "num_tokens": 236879887.0,
      "step": 7422
    },
    {
      "entropy": 0.7844161111861467,
      "epoch": 0.6829632952081932,
      "grad_norm": 0.16212278604507446,
      "learning_rate": 7.72380163768516e-05,
      "loss": 0.7581,
      "mean_token_accuracy": 0.7803203538060188,
      "num_tokens": 236911671.0,
      "step": 7423
    },
    {
      "entropy": 0.7387944273650646,
      "epoch": 0.6830553015796343,
      "grad_norm": 0.15842972695827484,
      "learning_rate": 7.723494955070998e-05,
      "loss": 0.7231,
      "mean_token_accuracy": 0.7883609756827354,
      "num_tokens": 236944015.0,
      "step": 7424
    },
    {
      "entropy": 0.8135396763682365,
      "epoch": 0.6831473079510756,
      "grad_norm": 0.16929569840431213,
      "learning_rate": 7.723188272456835e-05,
      "loss": 0.8065,
      "mean_token_accuracy": 0.7703985124826431,
      "num_tokens": 236975772.0,
      "step": 7425
    },
    {
      "entropy": 0.8376027271151543,
      "epoch": 0.6832393143225168,
      "grad_norm": 0.16157400608062744,
      "learning_rate": 7.722881589842671e-05,
      "loss": 0.8216,
      "mean_token_accuracy": 0.7618759609758854,
      "num_tokens": 237007779.0,
      "step": 7426
    },
    {
      "entropy": 0.681392015889287,
      "epoch": 0.6833313206939581,
      "grad_norm": 0.1580905169248581,
      "learning_rate": 7.72257490722851e-05,
      "loss": 0.6804,
      "mean_token_accuracy": 0.7997824586927891,
      "num_tokens": 237039269.0,
      "step": 7427
    },
    {
      "entropy": 0.7521419767290354,
      "epoch": 0.6834233270653993,
      "grad_norm": 0.15025755763053894,
      "learning_rate": 7.722268224614348e-05,
      "loss": 0.7509,
      "mean_token_accuracy": 0.7816542088985443,
      "num_tokens": 237071954.0,
      "step": 7428
    },
    {
      "entropy": 0.7604071628302336,
      "epoch": 0.6835153334368405,
      "grad_norm": 0.15519359707832336,
      "learning_rate": 7.721961542000185e-05,
      "loss": 0.7519,
      "mean_token_accuracy": 0.7756365053355694,
      "num_tokens": 237103813.0,
      "step": 7429
    },
    {
      "entropy": 0.7712708432227373,
      "epoch": 0.6836073398082817,
      "grad_norm": 0.1621960550546646,
      "learning_rate": 7.721654859386021e-05,
      "loss": 0.7664,
      "mean_token_accuracy": 0.7773892544209957,
      "num_tokens": 237135583.0,
      "step": 7430
    },
    {
      "entropy": 0.8782080840319395,
      "epoch": 0.6836993461797229,
      "grad_norm": 0.15484175086021423,
      "learning_rate": 7.721348176771858e-05,
      "loss": 0.872,
      "mean_token_accuracy": 0.7452356480062008,
      "num_tokens": 237167639.0,
      "step": 7431
    },
    {
      "entropy": 0.613685006275773,
      "epoch": 0.6837913525511642,
      "grad_norm": 0.15602408349514008,
      "learning_rate": 7.721041494157698e-05,
      "loss": 0.6008,
      "mean_token_accuracy": 0.8173104822635651,
      "num_tokens": 237199295.0,
      "step": 7432
    },
    {
      "entropy": 0.8362313807010651,
      "epoch": 0.6838833589226054,
      "grad_norm": 0.1590464860200882,
      "learning_rate": 7.720734811543534e-05,
      "loss": 0.83,
      "mean_token_accuracy": 0.7574935555458069,
      "num_tokens": 237229889.0,
      "step": 7433
    },
    {
      "entropy": 0.8630011156201363,
      "epoch": 0.6839753652940466,
      "grad_norm": 0.15885773301124573,
      "learning_rate": 7.720428128929371e-05,
      "loss": 0.8401,
      "mean_token_accuracy": 0.7537279576063156,
      "num_tokens": 237261573.0,
      "step": 7434
    },
    {
      "entropy": 0.7086672317236662,
      "epoch": 0.6840673716654878,
      "grad_norm": 0.15366820991039276,
      "learning_rate": 7.720121446315208e-05,
      "loss": 0.6861,
      "mean_token_accuracy": 0.7990264818072319,
      "num_tokens": 237293991.0,
      "step": 7435
    },
    {
      "entropy": 0.8015408143401146,
      "epoch": 0.6841593780369291,
      "grad_norm": 0.15421231091022491,
      "learning_rate": 7.719814763701046e-05,
      "loss": 0.7818,
      "mean_token_accuracy": 0.7677721455693245,
      "num_tokens": 237325599.0,
      "step": 7436
    },
    {
      "entropy": 0.7562645226716995,
      "epoch": 0.6842513844083703,
      "grad_norm": 0.15480005741119385,
      "learning_rate": 7.719508081086884e-05,
      "loss": 0.7398,
      "mean_token_accuracy": 0.7831414043903351,
      "num_tokens": 237357741.0,
      "step": 7437
    },
    {
      "entropy": 0.8288449142128229,
      "epoch": 0.6843433907798115,
      "grad_norm": 0.16900677978992462,
      "learning_rate": 7.719201398472721e-05,
      "loss": 0.8238,
      "mean_token_accuracy": 0.7623017132282257,
      "num_tokens": 237389364.0,
      "step": 7438
    },
    {
      "entropy": 0.7487089913338423,
      "epoch": 0.6844353971512527,
      "grad_norm": 0.1614963710308075,
      "learning_rate": 7.718894715858558e-05,
      "loss": 0.747,
      "mean_token_accuracy": 0.7840968556702137,
      "num_tokens": 237421197.0,
      "step": 7439
    },
    {
      "entropy": 0.7398779429495335,
      "epoch": 0.6845274035226939,
      "grad_norm": 0.16063272953033447,
      "learning_rate": 7.718588033244396e-05,
      "loss": 0.7289,
      "mean_token_accuracy": 0.7814731486141682,
      "num_tokens": 237453021.0,
      "step": 7440
    },
    {
      "entropy": 0.6963288392871618,
      "epoch": 0.6846194098941352,
      "grad_norm": 0.14929643273353577,
      "learning_rate": 7.718281350630233e-05,
      "loss": 0.6775,
      "mean_token_accuracy": 0.7991269007325172,
      "num_tokens": 237485340.0,
      "step": 7441
    },
    {
      "entropy": 0.614171639084816,
      "epoch": 0.6847114162655764,
      "grad_norm": 0.14689968526363373,
      "learning_rate": 7.717974668016071e-05,
      "loss": 0.5955,
      "mean_token_accuracy": 0.8198962733149529,
      "num_tokens": 237517582.0,
      "step": 7442
    },
    {
      "entropy": 0.8461778312921524,
      "epoch": 0.6848034226370177,
      "grad_norm": 0.15996411442756653,
      "learning_rate": 7.717667985401908e-05,
      "loss": 0.8475,
      "mean_token_accuracy": 0.7552186101675034,
      "num_tokens": 237549742.0,
      "step": 7443
    },
    {
      "entropy": 0.7215804755687714,
      "epoch": 0.6848954290084588,
      "grad_norm": 0.15759599208831787,
      "learning_rate": 7.717361302787746e-05,
      "loss": 0.7057,
      "mean_token_accuracy": 0.7874757796525955,
      "num_tokens": 237581618.0,
      "step": 7444
    },
    {
      "entropy": 0.7372153904289007,
      "epoch": 0.6849874353799,
      "grad_norm": 0.15593519806861877,
      "learning_rate": 7.717054620173583e-05,
      "loss": 0.7182,
      "mean_token_accuracy": 0.7862986139953136,
      "num_tokens": 237613534.0,
      "step": 7445
    },
    {
      "entropy": 0.7750965766608715,
      "epoch": 0.6850794417513413,
      "grad_norm": 0.16167408227920532,
      "learning_rate": 7.71674793755942e-05,
      "loss": 0.7656,
      "mean_token_accuracy": 0.772830180823803,
      "num_tokens": 237645730.0,
      "step": 7446
    },
    {
      "entropy": 0.662138850428164,
      "epoch": 0.6851714481227825,
      "grad_norm": 0.1439901739358902,
      "learning_rate": 7.716441254945258e-05,
      "loss": 0.6411,
      "mean_token_accuracy": 0.8039028346538544,
      "num_tokens": 237678255.0,
      "step": 7447
    },
    {
      "entropy": 0.7756048962473869,
      "epoch": 0.6852634544942238,
      "grad_norm": 0.15135405957698822,
      "learning_rate": 7.716134572331096e-05,
      "loss": 0.7807,
      "mean_token_accuracy": 0.7723841592669487,
      "num_tokens": 237710330.0,
      "step": 7448
    },
    {
      "entropy": 0.6921468386426568,
      "epoch": 0.6853554608656649,
      "grad_norm": 0.15394657850265503,
      "learning_rate": 7.715827889716932e-05,
      "loss": 0.6761,
      "mean_token_accuracy": 0.8006779626011848,
      "num_tokens": 237741931.0,
      "step": 7449
    },
    {
      "entropy": 0.7217855658382177,
      "epoch": 0.6854474672371061,
      "grad_norm": 0.16900531947612762,
      "learning_rate": 7.715521207102769e-05,
      "loss": 0.7169,
      "mean_token_accuracy": 0.7852700874209404,
      "num_tokens": 237773201.0,
      "step": 7450
    },
    {
      "entropy": 0.8128489460796118,
      "epoch": 0.6855394736085474,
      "grad_norm": 0.16201399266719818,
      "learning_rate": 7.715214524488607e-05,
      "loss": 0.8073,
      "mean_token_accuracy": 0.7629707045853138,
      "num_tokens": 237804874.0,
      "step": 7451
    },
    {
      "entropy": 0.7621209193021059,
      "epoch": 0.6856314799799886,
      "grad_norm": 0.17131945490837097,
      "learning_rate": 7.714907841874444e-05,
      "loss": 0.7805,
      "mean_token_accuracy": 0.7762468047440052,
      "num_tokens": 237836613.0,
      "step": 7452
    },
    {
      "entropy": 0.7351378817111254,
      "epoch": 0.6857234863514299,
      "grad_norm": 0.15765134990215302,
      "learning_rate": 7.714601159260282e-05,
      "loss": 0.7285,
      "mean_token_accuracy": 0.7876616418361664,
      "num_tokens": 237868314.0,
      "step": 7453
    },
    {
      "entropy": 0.6485138479620218,
      "epoch": 0.685815492722871,
      "grad_norm": 0.16378600895404816,
      "learning_rate": 7.714294476646119e-05,
      "loss": 0.6361,
      "mean_token_accuracy": 0.8133002072572708,
      "num_tokens": 237900367.0,
      "step": 7454
    },
    {
      "entropy": 0.7004106380045414,
      "epoch": 0.6859074990943123,
      "grad_norm": 0.15493617951869965,
      "learning_rate": 7.713987794031957e-05,
      "loss": 0.6818,
      "mean_token_accuracy": 0.7952115349471569,
      "num_tokens": 237932230.0,
      "step": 7455
    },
    {
      "entropy": 0.7676855400204659,
      "epoch": 0.6859995054657535,
      "grad_norm": 0.1603376269340515,
      "learning_rate": 7.713681111417794e-05,
      "loss": 0.7727,
      "mean_token_accuracy": 0.7751727029681206,
      "num_tokens": 237963818.0,
      "step": 7456
    },
    {
      "entropy": 0.6664930693805218,
      "epoch": 0.6860915118371947,
      "grad_norm": 0.1579933762550354,
      "learning_rate": 7.713374428803631e-05,
      "loss": 0.6419,
      "mean_token_accuracy": 0.8065777234733105,
      "num_tokens": 237995096.0,
      "step": 7457
    },
    {
      "entropy": 0.8157100975513458,
      "epoch": 0.686183518208636,
      "grad_norm": 0.164572075009346,
      "learning_rate": 7.713067746189469e-05,
      "loss": 0.7909,
      "mean_token_accuracy": 0.7692999020218849,
      "num_tokens": 238026777.0,
      "step": 7458
    },
    {
      "entropy": 0.7048759311437607,
      "epoch": 0.6862755245800771,
      "grad_norm": 0.16180042922496796,
      "learning_rate": 7.712761063575307e-05,
      "loss": 0.6838,
      "mean_token_accuracy": 0.7980211228132248,
      "num_tokens": 238058589.0,
      "step": 7459
    },
    {
      "entropy": 0.653591038659215,
      "epoch": 0.6863675309515184,
      "grad_norm": 0.15072211623191833,
      "learning_rate": 7.712454380961144e-05,
      "loss": 0.6423,
      "mean_token_accuracy": 0.8070961721241474,
      "num_tokens": 238090355.0,
      "step": 7460
    },
    {
      "entropy": 0.7329590953886509,
      "epoch": 0.6864595373229596,
      "grad_norm": 0.15462926030158997,
      "learning_rate": 7.712147698346981e-05,
      "loss": 0.7288,
      "mean_token_accuracy": 0.7882757224142551,
      "num_tokens": 238121004.0,
      "step": 7461
    },
    {
      "entropy": 0.7456008084118366,
      "epoch": 0.6865515436944009,
      "grad_norm": 0.15181034803390503,
      "learning_rate": 7.711841015732817e-05,
      "loss": 0.7195,
      "mean_token_accuracy": 0.7877502143383026,
      "num_tokens": 238153245.0,
      "step": 7462
    },
    {
      "entropy": 0.7540139015763998,
      "epoch": 0.6866435500658421,
      "grad_norm": 0.16011552512645721,
      "learning_rate": 7.711534333118656e-05,
      "loss": 0.7287,
      "mean_token_accuracy": 0.7846946939826012,
      "num_tokens": 238185046.0,
      "step": 7463
    },
    {
      "entropy": 0.6886442452669144,
      "epoch": 0.6867355564372832,
      "grad_norm": 0.15527787804603577,
      "learning_rate": 7.711227650504494e-05,
      "loss": 0.6494,
      "mean_token_accuracy": 0.8106651455163956,
      "num_tokens": 238217162.0,
      "step": 7464
    },
    {
      "entropy": 0.8278342336416245,
      "epoch": 0.6868275628087245,
      "grad_norm": 0.15133216977119446,
      "learning_rate": 7.71092096789033e-05,
      "loss": 0.8167,
      "mean_token_accuracy": 0.7618331797420979,
      "num_tokens": 238248623.0,
      "step": 7465
    },
    {
      "entropy": 0.7815176453441381,
      "epoch": 0.6869195691801657,
      "grad_norm": 0.14934538304805756,
      "learning_rate": 7.710614285276167e-05,
      "loss": 0.7606,
      "mean_token_accuracy": 0.7773117274045944,
      "num_tokens": 238281220.0,
      "step": 7466
    },
    {
      "entropy": 0.7725886888802052,
      "epoch": 0.687011575551607,
      "grad_norm": 0.16574878990650177,
      "learning_rate": 7.710307602662005e-05,
      "loss": 0.7708,
      "mean_token_accuracy": 0.7741411328315735,
      "num_tokens": 238313047.0,
      "step": 7467
    },
    {
      "entropy": 0.6681375280022621,
      "epoch": 0.6871035819230482,
      "grad_norm": 0.15152010321617126,
      "learning_rate": 7.710000920047842e-05,
      "loss": 0.6548,
      "mean_token_accuracy": 0.80411396920681,
      "num_tokens": 238345117.0,
      "step": 7468
    },
    {
      "entropy": 0.7411149609833956,
      "epoch": 0.6871955882944893,
      "grad_norm": 0.16032637655735016,
      "learning_rate": 7.70969423743368e-05,
      "loss": 0.7369,
      "mean_token_accuracy": 0.7857363820075989,
      "num_tokens": 238377245.0,
      "step": 7469
    },
    {
      "entropy": 0.7217102255672216,
      "epoch": 0.6872875946659306,
      "grad_norm": 0.1679292619228363,
      "learning_rate": 7.709387554819517e-05,
      "loss": 0.7345,
      "mean_token_accuracy": 0.7837369553744793,
      "num_tokens": 238407837.0,
      "step": 7470
    },
    {
      "entropy": 0.7006789986044168,
      "epoch": 0.6873796010373718,
      "grad_norm": 0.15919755399227142,
      "learning_rate": 7.709080872205355e-05,
      "loss": 0.682,
      "mean_token_accuracy": 0.7975373901426792,
      "num_tokens": 238439755.0,
      "step": 7471
    },
    {
      "entropy": 0.8291097022593021,
      "epoch": 0.6874716074088131,
      "grad_norm": 0.15397202968597412,
      "learning_rate": 7.708774189591192e-05,
      "loss": 0.8148,
      "mean_token_accuracy": 0.7562205530703068,
      "num_tokens": 238471856.0,
      "step": 7472
    },
    {
      "entropy": 0.771112434566021,
      "epoch": 0.6875636137802543,
      "grad_norm": 0.1489265263080597,
      "learning_rate": 7.708467506977029e-05,
      "loss": 0.7608,
      "mean_token_accuracy": 0.7730992585420609,
      "num_tokens": 238504347.0,
      "step": 7473
    },
    {
      "entropy": 0.7641617450863123,
      "epoch": 0.6876556201516955,
      "grad_norm": 0.1521620899438858,
      "learning_rate": 7.708160824362867e-05,
      "loss": 0.7406,
      "mean_token_accuracy": 0.7781996130943298,
      "num_tokens": 238536101.0,
      "step": 7474
    },
    {
      "entropy": 0.839030722156167,
      "epoch": 0.6877476265231367,
      "grad_norm": 0.1656017154455185,
      "learning_rate": 7.707854141748705e-05,
      "loss": 0.8506,
      "mean_token_accuracy": 0.7556817047297955,
      "num_tokens": 238567873.0,
      "step": 7475
    },
    {
      "entropy": 0.7492193542420864,
      "epoch": 0.687839632894578,
      "grad_norm": 0.14777421951293945,
      "learning_rate": 7.707547459134542e-05,
      "loss": 0.728,
      "mean_token_accuracy": 0.7839635349810123,
      "num_tokens": 238600312.0,
      "step": 7476
    },
    {
      "entropy": 0.8094575069844723,
      "epoch": 0.6879316392660192,
      "grad_norm": 0.16447025537490845,
      "learning_rate": 7.707240776520379e-05,
      "loss": 0.8141,
      "mean_token_accuracy": 0.7648382261395454,
      "num_tokens": 238631293.0,
      "step": 7477
    },
    {
      "entropy": 0.718489520251751,
      "epoch": 0.6880236456374604,
      "grad_norm": 0.16492860019207,
      "learning_rate": 7.706934093906217e-05,
      "loss": 0.7078,
      "mean_token_accuracy": 0.7918291874229908,
      "num_tokens": 238662425.0,
      "step": 7478
    },
    {
      "entropy": 0.7185233756899834,
      "epoch": 0.6881156520089016,
      "grad_norm": 0.15475422143936157,
      "learning_rate": 7.706627411292055e-05,
      "loss": 0.7039,
      "mean_token_accuracy": 0.7957345545291901,
      "num_tokens": 238694381.0,
      "step": 7479
    },
    {
      "entropy": 0.5973514290526509,
      "epoch": 0.6882076583803428,
      "grad_norm": 0.14847750961780548,
      "learning_rate": 7.706320728677892e-05,
      "loss": 0.5545,
      "mean_token_accuracy": 0.8376959078013897,
      "num_tokens": 238726617.0,
      "step": 7480
    },
    {
      "entropy": 0.6065159942954779,
      "epoch": 0.6882996647517841,
      "grad_norm": 0.14357946813106537,
      "learning_rate": 7.706014046063729e-05,
      "loss": 0.5989,
      "mean_token_accuracy": 0.8260867893695831,
      "num_tokens": 238758535.0,
      "step": 7481
    },
    {
      "entropy": 0.6056333389133215,
      "epoch": 0.6883916711232253,
      "grad_norm": 0.1484781950712204,
      "learning_rate": 7.705707363449567e-05,
      "loss": 0.5928,
      "mean_token_accuracy": 0.8211191296577454,
      "num_tokens": 238790330.0,
      "step": 7482
    },
    {
      "entropy": 0.7077449876815081,
      "epoch": 0.6884836774946665,
      "grad_norm": 0.15414772927761078,
      "learning_rate": 7.705400680835404e-05,
      "loss": 0.7011,
      "mean_token_accuracy": 0.7922239191830158,
      "num_tokens": 238821558.0,
      "step": 7483
    },
    {
      "entropy": 0.6826113946735859,
      "epoch": 0.6885756838661077,
      "grad_norm": 0.1556154042482376,
      "learning_rate": 7.705093998221242e-05,
      "loss": 0.6691,
      "mean_token_accuracy": 0.8005435205996037,
      "num_tokens": 238854071.0,
      "step": 7484
    },
    {
      "entropy": 0.7046669442206621,
      "epoch": 0.6886676902375489,
      "grad_norm": 0.16468991339206696,
      "learning_rate": 7.704787315607078e-05,
      "loss": 0.6856,
      "mean_token_accuracy": 0.7978657186031342,
      "num_tokens": 238885695.0,
      "step": 7485
    },
    {
      "entropy": 0.7590370960533619,
      "epoch": 0.6887596966089902,
      "grad_norm": 0.16399110853672028,
      "learning_rate": 7.704480632992917e-05,
      "loss": 0.7547,
      "mean_token_accuracy": 0.7787040136754513,
      "num_tokens": 238917346.0,
      "step": 7486
    },
    {
      "entropy": 0.7317015212029219,
      "epoch": 0.6888517029804314,
      "grad_norm": 0.1545829176902771,
      "learning_rate": 7.704173950378753e-05,
      "loss": 0.6994,
      "mean_token_accuracy": 0.7895872667431831,
      "num_tokens": 238949635.0,
      "step": 7487
    },
    {
      "entropy": 0.7005993165075779,
      "epoch": 0.6889437093518727,
      "grad_norm": 0.15294994413852692,
      "learning_rate": 7.70386726776459e-05,
      "loss": 0.6877,
      "mean_token_accuracy": 0.795720137655735,
      "num_tokens": 238981695.0,
      "step": 7488
    },
    {
      "entropy": 0.7138608619570732,
      "epoch": 0.6890357157233138,
      "grad_norm": 0.1591971218585968,
      "learning_rate": 7.703560585150428e-05,
      "loss": 0.7082,
      "mean_token_accuracy": 0.7913409881293774,
      "num_tokens": 239014463.0,
      "step": 7489
    },
    {
      "entropy": 0.6229658424854279,
      "epoch": 0.689127722094755,
      "grad_norm": 0.14771485328674316,
      "learning_rate": 7.703253902536267e-05,
      "loss": 0.5984,
      "mean_token_accuracy": 0.8202571868896484,
      "num_tokens": 239047111.0,
      "step": 7490
    },
    {
      "entropy": 0.6341155739501119,
      "epoch": 0.6892197284661963,
      "grad_norm": 0.15780678391456604,
      "learning_rate": 7.702947219922103e-05,
      "loss": 0.6065,
      "mean_token_accuracy": 0.8172766081988811,
      "num_tokens": 239077919.0,
      "step": 7491
    },
    {
      "entropy": 0.8044809587299824,
      "epoch": 0.6893117348376375,
      "grad_norm": 0.1688428372144699,
      "learning_rate": 7.70264053730794e-05,
      "loss": 0.8195,
      "mean_token_accuracy": 0.7586627006530762,
      "num_tokens": 239109282.0,
      "step": 7492
    },
    {
      "entropy": 0.8008085135370493,
      "epoch": 0.6894037412090788,
      "grad_norm": 0.1586918979883194,
      "learning_rate": 7.702333854693777e-05,
      "loss": 0.8084,
      "mean_token_accuracy": 0.7622111476957798,
      "num_tokens": 239140708.0,
      "step": 7493
    },
    {
      "entropy": 0.6131381839513779,
      "epoch": 0.6894957475805199,
      "grad_norm": 0.16708096861839294,
      "learning_rate": 7.702027172079615e-05,
      "loss": 0.6054,
      "mean_token_accuracy": 0.8225168064236641,
      "num_tokens": 239172892.0,
      "step": 7494
    },
    {
      "entropy": 0.6587367691099644,
      "epoch": 0.6895877539519611,
      "grad_norm": 0.15705080330371857,
      "learning_rate": 7.701720489465453e-05,
      "loss": 0.6464,
      "mean_token_accuracy": 0.8067270293831825,
      "num_tokens": 239203888.0,
      "step": 7495
    },
    {
      "entropy": 0.8460985850542784,
      "epoch": 0.6896797603234024,
      "grad_norm": 0.1828412264585495,
      "learning_rate": 7.70141380685129e-05,
      "loss": 0.8726,
      "mean_token_accuracy": 0.752550520002842,
      "num_tokens": 239234768.0,
      "step": 7496
    },
    {
      "entropy": 0.8009899966418743,
      "epoch": 0.6897717666948436,
      "grad_norm": 0.1607704907655716,
      "learning_rate": 7.701107124237127e-05,
      "loss": 0.7861,
      "mean_token_accuracy": 0.7679641172289848,
      "num_tokens": 239266251.0,
      "step": 7497
    },
    {
      "entropy": 0.713421642780304,
      "epoch": 0.6898637730662849,
      "grad_norm": 0.15544578433036804,
      "learning_rate": 7.700800441622965e-05,
      "loss": 0.6976,
      "mean_token_accuracy": 0.7939934208989143,
      "num_tokens": 239298611.0,
      "step": 7498
    },
    {
      "entropy": 0.7696054354310036,
      "epoch": 0.689955779437726,
      "grad_norm": 0.16657161712646484,
      "learning_rate": 7.700493759008802e-05,
      "loss": 0.7555,
      "mean_token_accuracy": 0.7789471000432968,
      "num_tokens": 239330542.0,
      "step": 7499
    },
    {
      "entropy": 0.7042547911405563,
      "epoch": 0.6900477858091673,
      "grad_norm": 0.15580616891384125,
      "learning_rate": 7.70018707639464e-05,
      "loss": 0.6882,
      "mean_token_accuracy": 0.795498926192522,
      "num_tokens": 239362602.0,
      "step": 7500
    },
    {
      "entropy": 0.5392262395471334,
      "epoch": 0.6901397921806085,
      "grad_norm": 0.13985863327980042,
      "learning_rate": 7.699880393780477e-05,
      "loss": 0.5206,
      "mean_token_accuracy": 0.8438648469746113,
      "num_tokens": 239395091.0,
      "step": 7501
    },
    {
      "entropy": 0.6945722615346313,
      "epoch": 0.6902317985520497,
      "grad_norm": 0.16417311131954193,
      "learning_rate": 7.699573711166315e-05,
      "loss": 0.6893,
      "mean_token_accuracy": 0.8020884208381176,
      "num_tokens": 239427859.0,
      "step": 7502
    },
    {
      "entropy": 0.7865596953779459,
      "epoch": 0.690323804923491,
      "grad_norm": 0.15723706781864166,
      "learning_rate": 7.699267028552152e-05,
      "loss": 0.7775,
      "mean_token_accuracy": 0.7725660502910614,
      "num_tokens": 239459334.0,
      "step": 7503
    },
    {
      "entropy": 0.6482004076242447,
      "epoch": 0.6904158112949321,
      "grad_norm": 0.15044528245925903,
      "learning_rate": 7.698960345937988e-05,
      "loss": 0.6339,
      "mean_token_accuracy": 0.8165746033191681,
      "num_tokens": 239491485.0,
      "step": 7504
    },
    {
      "entropy": 0.7062843013554811,
      "epoch": 0.6905078176663734,
      "grad_norm": 0.16156208515167236,
      "learning_rate": 7.698653663323826e-05,
      "loss": 0.6931,
      "mean_token_accuracy": 0.7947399504482746,
      "num_tokens": 239523785.0,
      "step": 7505
    },
    {
      "entropy": 0.6634152932092547,
      "epoch": 0.6905998240378146,
      "grad_norm": 0.15001702308654785,
      "learning_rate": 7.698346980709665e-05,
      "loss": 0.6357,
      "mean_token_accuracy": 0.8136476464569569,
      "num_tokens": 239556273.0,
      "step": 7506
    },
    {
      "entropy": 0.7142920978367329,
      "epoch": 0.6906918304092559,
      "grad_norm": 0.15689122676849365,
      "learning_rate": 7.698040298095501e-05,
      "loss": 0.6973,
      "mean_token_accuracy": 0.7930347435176373,
      "num_tokens": 239587726.0,
      "step": 7507
    },
    {
      "entropy": 0.7935129534453154,
      "epoch": 0.6907838367806971,
      "grad_norm": 0.15480057895183563,
      "learning_rate": 7.697733615481338e-05,
      "loss": 0.7801,
      "mean_token_accuracy": 0.7697254568338394,
      "num_tokens": 239619697.0,
      "step": 7508
    },
    {
      "entropy": 0.7107953820377588,
      "epoch": 0.6908758431521382,
      "grad_norm": 0.14943666756153107,
      "learning_rate": 7.697426932867176e-05,
      "loss": 0.69,
      "mean_token_accuracy": 0.7949233911931515,
      "num_tokens": 239651920.0,
      "step": 7509
    },
    {
      "entropy": 0.7927297540009022,
      "epoch": 0.6909678495235795,
      "grad_norm": 0.15560214221477509,
      "learning_rate": 7.697120250253013e-05,
      "loss": 0.7802,
      "mean_token_accuracy": 0.7712098471820354,
      "num_tokens": 239683246.0,
      "step": 7510
    },
    {
      "entropy": 0.6849836856126785,
      "epoch": 0.6910598558950207,
      "grad_norm": 0.15472926199436188,
      "learning_rate": 7.696813567638851e-05,
      "loss": 0.6772,
      "mean_token_accuracy": 0.7982728965580463,
      "num_tokens": 239715856.0,
      "step": 7511
    },
    {
      "entropy": 0.7203149888664484,
      "epoch": 0.691151862266462,
      "grad_norm": 0.16690997779369354,
      "learning_rate": 7.696506885024688e-05,
      "loss": 0.7176,
      "mean_token_accuracy": 0.7932063937187195,
      "num_tokens": 239748194.0,
      "step": 7512
    },
    {
      "entropy": 0.7451728126034141,
      "epoch": 0.6912438686379032,
      "grad_norm": 0.164011150598526,
      "learning_rate": 7.696200202410526e-05,
      "loss": 0.725,
      "mean_token_accuracy": 0.7854197137057781,
      "num_tokens": 239780688.0,
      "step": 7513
    },
    {
      "entropy": 0.7592190112918615,
      "epoch": 0.6913358750093443,
      "grad_norm": 0.16486801207065582,
      "learning_rate": 7.695893519796363e-05,
      "loss": 0.7513,
      "mean_token_accuracy": 0.7753321006894112,
      "num_tokens": 239811525.0,
      "step": 7514
    },
    {
      "entropy": 0.7901181951165199,
      "epoch": 0.6914278813807856,
      "grad_norm": 0.15245702862739563,
      "learning_rate": 7.695586837182201e-05,
      "loss": 0.7782,
      "mean_token_accuracy": 0.7743986994028091,
      "num_tokens": 239843836.0,
      "step": 7515
    },
    {
      "entropy": 0.7630588579922915,
      "epoch": 0.6915198877522268,
      "grad_norm": 0.1520344763994217,
      "learning_rate": 7.695280154568038e-05,
      "loss": 0.7582,
      "mean_token_accuracy": 0.7754198312759399,
      "num_tokens": 239875729.0,
      "step": 7516
    },
    {
      "entropy": 0.7234228923916817,
      "epoch": 0.6916118941236681,
      "grad_norm": 0.1472955048084259,
      "learning_rate": 7.694973471953876e-05,
      "loss": 0.7133,
      "mean_token_accuracy": 0.7880950570106506,
      "num_tokens": 239908392.0,
      "step": 7517
    },
    {
      "entropy": 0.6243173871189356,
      "epoch": 0.6917039004951093,
      "grad_norm": 0.15186038613319397,
      "learning_rate": 7.694666789339713e-05,
      "loss": 0.608,
      "mean_token_accuracy": 0.8218021243810654,
      "num_tokens": 239941150.0,
      "step": 7518
    },
    {
      "entropy": 0.5928329704329371,
      "epoch": 0.6917959068665505,
      "grad_norm": 0.15478144586086273,
      "learning_rate": 7.69436010672555e-05,
      "loss": 0.5932,
      "mean_token_accuracy": 0.8207711651921272,
      "num_tokens": 239973568.0,
      "step": 7519
    },
    {
      "entropy": 0.724655469879508,
      "epoch": 0.6918879132379917,
      "grad_norm": 0.1604958027601242,
      "learning_rate": 7.694053424111388e-05,
      "loss": 0.7331,
      "mean_token_accuracy": 0.7871913127601147,
      "num_tokens": 240006113.0,
      "step": 7520
    },
    {
      "entropy": 0.6605938682332635,
      "epoch": 0.691979919609433,
      "grad_norm": 0.165904238820076,
      "learning_rate": 7.693746741497226e-05,
      "loss": 0.6608,
      "mean_token_accuracy": 0.8050528131425381,
      "num_tokens": 240038449.0,
      "step": 7521
    },
    {
      "entropy": 0.7121169250458479,
      "epoch": 0.6920719259808742,
      "grad_norm": 0.16052009165287018,
      "learning_rate": 7.693440058883063e-05,
      "loss": 0.6986,
      "mean_token_accuracy": 0.7932977080345154,
      "num_tokens": 240070531.0,
      "step": 7522
    },
    {
      "entropy": 0.675326582044363,
      "epoch": 0.6921639323523154,
      "grad_norm": 0.16103030741214752,
      "learning_rate": 7.6931333762689e-05,
      "loss": 0.6677,
      "mean_token_accuracy": 0.7980044893920422,
      "num_tokens": 240102805.0,
      "step": 7523
    },
    {
      "entropy": 0.7213351549580693,
      "epoch": 0.6922559387237566,
      "grad_norm": 0.15647290647029877,
      "learning_rate": 7.692826693654736e-05,
      "loss": 0.6981,
      "mean_token_accuracy": 0.7923853471875191,
      "num_tokens": 240134459.0,
      "step": 7524
    },
    {
      "entropy": 0.7662236150354147,
      "epoch": 0.6923479450951978,
      "grad_norm": 0.15225306153297424,
      "learning_rate": 7.692520011040574e-05,
      "loss": 0.7385,
      "mean_token_accuracy": 0.7792158201336861,
      "num_tokens": 240166128.0,
      "step": 7525
    },
    {
      "entropy": 0.6405912861227989,
      "epoch": 0.6924399514666391,
      "grad_norm": 0.14661268889904022,
      "learning_rate": 7.692213328426413e-05,
      "loss": 0.6205,
      "mean_token_accuracy": 0.813658632338047,
      "num_tokens": 240198411.0,
      "step": 7526
    },
    {
      "entropy": 0.7095953784883022,
      "epoch": 0.6925319578380803,
      "grad_norm": 0.14897051453590393,
      "learning_rate": 7.691906645812249e-05,
      "loss": 0.6966,
      "mean_token_accuracy": 0.790154080837965,
      "num_tokens": 240230106.0,
      "step": 7527
    },
    {
      "entropy": 0.8020265363156796,
      "epoch": 0.6926239642095215,
      "grad_norm": 0.15196482837200165,
      "learning_rate": 7.691599963198086e-05,
      "loss": 0.7733,
      "mean_token_accuracy": 0.7740650475025177,
      "num_tokens": 240261616.0,
      "step": 7528
    },
    {
      "entropy": 0.6532169980928302,
      "epoch": 0.6927159705809627,
      "grad_norm": 0.1484462022781372,
      "learning_rate": 7.691293280583924e-05,
      "loss": 0.633,
      "mean_token_accuracy": 0.811723493039608,
      "num_tokens": 240293800.0,
      "step": 7529
    },
    {
      "entropy": 0.8197371698915958,
      "epoch": 0.6928079769524039,
      "grad_norm": 0.15931940078735352,
      "learning_rate": 7.690986597969761e-05,
      "loss": 0.8007,
      "mean_token_accuracy": 0.7682602442800999,
      "num_tokens": 240326237.0,
      "step": 7530
    },
    {
      "entropy": 0.7719225455075502,
      "epoch": 0.6928999833238452,
      "grad_norm": 0.15927593410015106,
      "learning_rate": 7.690679915355599e-05,
      "loss": 0.7471,
      "mean_token_accuracy": 0.7789826989173889,
      "num_tokens": 240358120.0,
      "step": 7531
    },
    {
      "entropy": 0.7349131628870964,
      "epoch": 0.6929919896952864,
      "grad_norm": 0.15869101881980896,
      "learning_rate": 7.690373232741436e-05,
      "loss": 0.719,
      "mean_token_accuracy": 0.7924159504473209,
      "num_tokens": 240390221.0,
      "step": 7532
    },
    {
      "entropy": 0.7274132538586855,
      "epoch": 0.6930839960667277,
      "grad_norm": 0.15190888941287994,
      "learning_rate": 7.690066550127274e-05,
      "loss": 0.714,
      "mean_token_accuracy": 0.7925065122544765,
      "num_tokens": 240422102.0,
      "step": 7533
    },
    {
      "entropy": 0.7322409488260746,
      "epoch": 0.6931760024381688,
      "grad_norm": 0.1600644737482071,
      "learning_rate": 7.689759867513111e-05,
      "loss": 0.73,
      "mean_token_accuracy": 0.7833596542477608,
      "num_tokens": 240454118.0,
      "step": 7534
    },
    {
      "entropy": 0.8652691654860973,
      "epoch": 0.69326800880961,
      "grad_norm": 0.16280117630958557,
      "learning_rate": 7.689453184898948e-05,
      "loss": 0.8799,
      "mean_token_accuracy": 0.7457996308803558,
      "num_tokens": 240486014.0,
      "step": 7535
    },
    {
      "entropy": 0.8264477197080851,
      "epoch": 0.6933600151810513,
      "grad_norm": 0.15992456674575806,
      "learning_rate": 7.689146502284786e-05,
      "loss": 0.8309,
      "mean_token_accuracy": 0.7571189031004906,
      "num_tokens": 240518592.0,
      "step": 7536
    },
    {
      "entropy": 0.7614348530769348,
      "epoch": 0.6934520215524925,
      "grad_norm": 0.16926692426204681,
      "learning_rate": 7.688839819670624e-05,
      "loss": 0.7815,
      "mean_token_accuracy": 0.7685856036841869,
      "num_tokens": 240550572.0,
      "step": 7537
    },
    {
      "entropy": 0.7551673706620932,
      "epoch": 0.6935440279239338,
      "grad_norm": 0.1641397476196289,
      "learning_rate": 7.688533137056461e-05,
      "loss": 0.7711,
      "mean_token_accuracy": 0.772520586848259,
      "num_tokens": 240582736.0,
      "step": 7538
    },
    {
      "entropy": 0.7120810952037573,
      "epoch": 0.6936360342953749,
      "grad_norm": 0.15579865872859955,
      "learning_rate": 7.688226454442298e-05,
      "loss": 0.6989,
      "mean_token_accuracy": 0.791780237108469,
      "num_tokens": 240615074.0,
      "step": 7539
    },
    {
      "entropy": 0.7548304628580809,
      "epoch": 0.6937280406668161,
      "grad_norm": 0.1538027822971344,
      "learning_rate": 7.687919771828136e-05,
      "loss": 0.7403,
      "mean_token_accuracy": 0.7803789116442204,
      "num_tokens": 240647686.0,
      "step": 7540
    },
    {
      "entropy": 0.6946017956361175,
      "epoch": 0.6938200470382574,
      "grad_norm": 0.15522214770317078,
      "learning_rate": 7.687613089213972e-05,
      "loss": 0.6752,
      "mean_token_accuracy": 0.8029736429452896,
      "num_tokens": 240679392.0,
      "step": 7541
    },
    {
      "entropy": 0.6970695685595274,
      "epoch": 0.6939120534096986,
      "grad_norm": 0.16968370974063873,
      "learning_rate": 7.68730640659981e-05,
      "loss": 0.6794,
      "mean_token_accuracy": 0.7954268530011177,
      "num_tokens": 240710858.0,
      "step": 7542
    },
    {
      "entropy": 0.6588600305840373,
      "epoch": 0.6940040597811399,
      "grad_norm": 0.15241603553295135,
      "learning_rate": 7.686999723985647e-05,
      "loss": 0.6367,
      "mean_token_accuracy": 0.8092412538826466,
      "num_tokens": 240743388.0,
      "step": 7543
    },
    {
      "entropy": 0.7449143286794424,
      "epoch": 0.694096066152581,
      "grad_norm": 0.16616688668727875,
      "learning_rate": 7.686693041371486e-05,
      "loss": 0.7174,
      "mean_token_accuracy": 0.7900488376617432,
      "num_tokens": 240775582.0,
      "step": 7544
    },
    {
      "entropy": 0.7168315853923559,
      "epoch": 0.6941880725240223,
      "grad_norm": 0.16117440164089203,
      "learning_rate": 7.686386358757322e-05,
      "loss": 0.7094,
      "mean_token_accuracy": 0.7937523424625397,
      "num_tokens": 240807396.0,
      "step": 7545
    },
    {
      "entropy": 0.6859567668288946,
      "epoch": 0.6942800788954635,
      "grad_norm": 0.15206116437911987,
      "learning_rate": 7.686079676143159e-05,
      "loss": 0.6767,
      "mean_token_accuracy": 0.7981190904974937,
      "num_tokens": 240839185.0,
      "step": 7546
    },
    {
      "entropy": 0.6723775994032621,
      "epoch": 0.6943720852669047,
      "grad_norm": 0.1604377180337906,
      "learning_rate": 7.685772993528997e-05,
      "loss": 0.6451,
      "mean_token_accuracy": 0.8078895844519138,
      "num_tokens": 240871953.0,
      "step": 7547
    },
    {
      "entropy": 0.6663623508065939,
      "epoch": 0.694464091638346,
      "grad_norm": 0.15366703271865845,
      "learning_rate": 7.685466310914835e-05,
      "loss": 0.6453,
      "mean_token_accuracy": 0.807632964104414,
      "num_tokens": 240904226.0,
      "step": 7548
    },
    {
      "entropy": 0.6802077312022448,
      "epoch": 0.6945560980097871,
      "grad_norm": 0.1460757553577423,
      "learning_rate": 7.685159628300672e-05,
      "loss": 0.6656,
      "mean_token_accuracy": 0.8002004660665989,
      "num_tokens": 240935598.0,
      "step": 7549
    },
    {
      "entropy": 0.7875551898032427,
      "epoch": 0.6946481043812284,
      "grad_norm": 0.15767715871334076,
      "learning_rate": 7.684852945686509e-05,
      "loss": 0.7672,
      "mean_token_accuracy": 0.7721450477838516,
      "num_tokens": 240966956.0,
      "step": 7550
    },
    {
      "entropy": 0.854848001152277,
      "epoch": 0.6947401107526696,
      "grad_norm": 0.1625068336725235,
      "learning_rate": 7.684546263072346e-05,
      "loss": 0.8733,
      "mean_token_accuracy": 0.7445134036242962,
      "num_tokens": 240997950.0,
      "step": 7551
    },
    {
      "entropy": 0.7910617664456367,
      "epoch": 0.6948321171241109,
      "grad_norm": 0.16032156348228455,
      "learning_rate": 7.684239580458185e-05,
      "loss": 0.7696,
      "mean_token_accuracy": 0.7674101144075394,
      "num_tokens": 241029473.0,
      "step": 7552
    },
    {
      "entropy": 0.7291016951203346,
      "epoch": 0.6949241234955521,
      "grad_norm": 0.1608472317457199,
      "learning_rate": 7.683932897844022e-05,
      "loss": 0.726,
      "mean_token_accuracy": 0.7854020483791828,
      "num_tokens": 241061399.0,
      "step": 7553
    },
    {
      "entropy": 0.6913976240903139,
      "epoch": 0.6950161298669932,
      "grad_norm": 0.16342705488204956,
      "learning_rate": 7.683626215229859e-05,
      "loss": 0.6856,
      "mean_token_accuracy": 0.7886982858181,
      "num_tokens": 241092652.0,
      "step": 7554
    },
    {
      "entropy": 0.7766660898923874,
      "epoch": 0.6951081362384345,
      "grad_norm": 0.16151392459869385,
      "learning_rate": 7.683319532615696e-05,
      "loss": 0.7699,
      "mean_token_accuracy": 0.776250172406435,
      "num_tokens": 241124370.0,
      "step": 7555
    },
    {
      "entropy": 0.8568740412592888,
      "epoch": 0.6952001426098757,
      "grad_norm": 0.15653185546398163,
      "learning_rate": 7.683012850001534e-05,
      "loss": 0.8631,
      "mean_token_accuracy": 0.7528864815831184,
      "num_tokens": 241156909.0,
      "step": 7556
    },
    {
      "entropy": 0.7649150863289833,
      "epoch": 0.695292148981317,
      "grad_norm": 0.16177113354206085,
      "learning_rate": 7.682706167387372e-05,
      "loss": 0.7534,
      "mean_token_accuracy": 0.77771120890975,
      "num_tokens": 241188755.0,
      "step": 7557
    },
    {
      "entropy": 0.8534499295055866,
      "epoch": 0.6953841553527582,
      "grad_norm": 0.16745978593826294,
      "learning_rate": 7.682399484773209e-05,
      "loss": 0.8479,
      "mean_token_accuracy": 0.7589173130691051,
      "num_tokens": 241220851.0,
      "step": 7558
    },
    {
      "entropy": 0.7003718297928572,
      "epoch": 0.6954761617241993,
      "grad_norm": 0.14728645980358124,
      "learning_rate": 7.682092802159045e-05,
      "loss": 0.6798,
      "mean_token_accuracy": 0.7972778268158436,
      "num_tokens": 241253254.0,
      "step": 7559
    },
    {
      "entropy": 0.7018515858799219,
      "epoch": 0.6955681680956406,
      "grad_norm": 0.1665259748697281,
      "learning_rate": 7.681786119544884e-05,
      "loss": 0.6946,
      "mean_token_accuracy": 0.7910956516861916,
      "num_tokens": 241285172.0,
      "step": 7560
    },
    {
      "entropy": 0.7994505241513252,
      "epoch": 0.6956601744670818,
      "grad_norm": 0.1657220870256424,
      "learning_rate": 7.68147943693072e-05,
      "loss": 0.7743,
      "mean_token_accuracy": 0.7690246365964413,
      "num_tokens": 241317222.0,
      "step": 7561
    },
    {
      "entropy": 0.7231369446963072,
      "epoch": 0.6957521808385231,
      "grad_norm": 0.15103647112846375,
      "learning_rate": 7.681172754316559e-05,
      "loss": 0.7027,
      "mean_token_accuracy": 0.7934114187955856,
      "num_tokens": 241349192.0,
      "step": 7562
    },
    {
      "entropy": 0.8205822445452213,
      "epoch": 0.6958441872099643,
      "grad_norm": 0.16000372171401978,
      "learning_rate": 7.680866071702395e-05,
      "loss": 0.8264,
      "mean_token_accuracy": 0.7578230760991573,
      "num_tokens": 241381668.0,
      "step": 7563
    },
    {
      "entropy": 0.6687055826187134,
      "epoch": 0.6959361935814055,
      "grad_norm": 0.1493268758058548,
      "learning_rate": 7.680559389088233e-05,
      "loss": 0.6452,
      "mean_token_accuracy": 0.808261338621378,
      "num_tokens": 241413385.0,
      "step": 7564
    },
    {
      "entropy": 0.7479179389774799,
      "epoch": 0.6960281999528467,
      "grad_norm": 0.36740005016326904,
      "learning_rate": 7.68025270647407e-05,
      "loss": 0.7574,
      "mean_token_accuracy": 0.7770942784845829,
      "num_tokens": 241445415.0,
      "step": 7565
    },
    {
      "entropy": 0.742334658280015,
      "epoch": 0.696120206324288,
      "grad_norm": 0.1632843017578125,
      "learning_rate": 7.679946023859907e-05,
      "loss": 0.7109,
      "mean_token_accuracy": 0.788420956581831,
      "num_tokens": 241477292.0,
      "step": 7566
    },
    {
      "entropy": 0.7861579144373536,
      "epoch": 0.6962122126957292,
      "grad_norm": 0.16122332215309143,
      "learning_rate": 7.679639341245745e-05,
      "loss": 0.7873,
      "mean_token_accuracy": 0.7680408470332623,
      "num_tokens": 241508735.0,
      "step": 7567
    },
    {
      "entropy": 0.7302915751934052,
      "epoch": 0.6963042190671704,
      "grad_norm": 0.1576634794473648,
      "learning_rate": 7.679332658631583e-05,
      "loss": 0.7096,
      "mean_token_accuracy": 0.7885285131633282,
      "num_tokens": 241540909.0,
      "step": 7568
    },
    {
      "entropy": 0.7766938265413046,
      "epoch": 0.6963962254386116,
      "grad_norm": 0.1673874855041504,
      "learning_rate": 7.67902597601742e-05,
      "loss": 0.7781,
      "mean_token_accuracy": 0.7703855112195015,
      "num_tokens": 241572410.0,
      "step": 7569
    },
    {
      "entropy": 0.7689692564308643,
      "epoch": 0.6964882318100528,
      "grad_norm": 0.1600862592458725,
      "learning_rate": 7.678719293403257e-05,
      "loss": 0.7673,
      "mean_token_accuracy": 0.7805592492222786,
      "num_tokens": 241604808.0,
      "step": 7570
    },
    {
      "entropy": 0.744633880443871,
      "epoch": 0.6965802381814941,
      "grad_norm": 0.15338660776615143,
      "learning_rate": 7.678412610789095e-05,
      "loss": 0.7374,
      "mean_token_accuracy": 0.7810798399150372,
      "num_tokens": 241636549.0,
      "step": 7571
    },
    {
      "entropy": 0.7205295693129301,
      "epoch": 0.6966722445529353,
      "grad_norm": 0.1593475192785263,
      "learning_rate": 7.678105928174932e-05,
      "loss": 0.7227,
      "mean_token_accuracy": 0.787227563560009,
      "num_tokens": 241668354.0,
      "step": 7572
    },
    {
      "entropy": 0.7885692939162254,
      "epoch": 0.6967642509243765,
      "grad_norm": 0.16188636422157288,
      "learning_rate": 7.67779924556077e-05,
      "loss": 0.7859,
      "mean_token_accuracy": 0.7698486894369125,
      "num_tokens": 241700735.0,
      "step": 7573
    },
    {
      "entropy": 0.7219500662758946,
      "epoch": 0.6968562572958177,
      "grad_norm": 0.15191803872585297,
      "learning_rate": 7.677492562946607e-05,
      "loss": 0.6983,
      "mean_token_accuracy": 0.7916063852608204,
      "num_tokens": 241732753.0,
      "step": 7574
    },
    {
      "entropy": 0.6225885953754187,
      "epoch": 0.6969482636672589,
      "grad_norm": 0.1474747359752655,
      "learning_rate": 7.677185880332445e-05,
      "loss": 0.6068,
      "mean_token_accuracy": 0.8180731050670147,
      "num_tokens": 241765091.0,
      "step": 7575
    },
    {
      "entropy": 0.7783480565994978,
      "epoch": 0.6970402700387002,
      "grad_norm": 0.1580721139907837,
      "learning_rate": 7.676879197718282e-05,
      "loss": 0.7654,
      "mean_token_accuracy": 0.7794495299458504,
      "num_tokens": 241797291.0,
      "step": 7576
    },
    {
      "entropy": 0.7180880298838019,
      "epoch": 0.6971322764101414,
      "grad_norm": 0.1569516658782959,
      "learning_rate": 7.676572515104118e-05,
      "loss": 0.6988,
      "mean_token_accuracy": 0.7932194992899895,
      "num_tokens": 241829432.0,
      "step": 7577
    },
    {
      "entropy": 0.8190671559423208,
      "epoch": 0.6972242827815827,
      "grad_norm": 0.16650132834911346,
      "learning_rate": 7.676265832489957e-05,
      "loss": 0.8329,
      "mean_token_accuracy": 0.7573514804244041,
      "num_tokens": 241860606.0,
      "step": 7578
    },
    {
      "entropy": 0.7655077744275331,
      "epoch": 0.6973162891530238,
      "grad_norm": 0.15506044030189514,
      "learning_rate": 7.675959149875795e-05,
      "loss": 0.7541,
      "mean_token_accuracy": 0.7798767127096653,
      "num_tokens": 241892220.0,
      "step": 7579
    },
    {
      "entropy": 0.7305645477026701,
      "epoch": 0.697408295524465,
      "grad_norm": 0.15340057015419006,
      "learning_rate": 7.675652467261632e-05,
      "loss": 0.7266,
      "mean_token_accuracy": 0.7831007800996304,
      "num_tokens": 241923751.0,
      "step": 7580
    },
    {
      "entropy": 0.6844135755673051,
      "epoch": 0.6975003018959063,
      "grad_norm": 0.15518641471862793,
      "learning_rate": 7.675345784647468e-05,
      "loss": 0.6709,
      "mean_token_accuracy": 0.8040879182517529,
      "num_tokens": 241955910.0,
      "step": 7581
    },
    {
      "entropy": 0.7271458152681589,
      "epoch": 0.6975923082673475,
      "grad_norm": 0.16411589086055756,
      "learning_rate": 7.675039102033305e-05,
      "loss": 0.7231,
      "mean_token_accuracy": 0.7825558707118034,
      "num_tokens": 241987976.0,
      "step": 7582
    },
    {
      "entropy": 0.7097965534776449,
      "epoch": 0.6976843146387888,
      "grad_norm": 0.15577435493469238,
      "learning_rate": 7.674732419419143e-05,
      "loss": 0.6988,
      "mean_token_accuracy": 0.7953037656843662,
      "num_tokens": 242020093.0,
      "step": 7583
    },
    {
      "entropy": 0.6957558449357748,
      "epoch": 0.6977763210102299,
      "grad_norm": 0.15555883944034576,
      "learning_rate": 7.674425736804981e-05,
      "loss": 0.6695,
      "mean_token_accuracy": 0.799114391207695,
      "num_tokens": 242051890.0,
      "step": 7584
    },
    {
      "entropy": 0.7859354130923748,
      "epoch": 0.6978683273816711,
      "grad_norm": 0.16423310339450836,
      "learning_rate": 7.674119054190818e-05,
      "loss": 0.7828,
      "mean_token_accuracy": 0.7693586871027946,
      "num_tokens": 242083509.0,
      "step": 7585
    },
    {
      "entropy": 0.6911279242485762,
      "epoch": 0.6979603337531124,
      "grad_norm": 0.1579013615846634,
      "learning_rate": 7.673812371576655e-05,
      "loss": 0.67,
      "mean_token_accuracy": 0.8071000464260578,
      "num_tokens": 242115782.0,
      "step": 7586
    },
    {
      "entropy": 0.7581703849136829,
      "epoch": 0.6980523401245536,
      "grad_norm": 0.16208422183990479,
      "learning_rate": 7.673505688962493e-05,
      "loss": 0.7475,
      "mean_token_accuracy": 0.7812771759927273,
      "num_tokens": 242147008.0,
      "step": 7587
    },
    {
      "entropy": 0.799053848721087,
      "epoch": 0.6981443464959949,
      "grad_norm": 0.15860413014888763,
      "learning_rate": 7.67319900634833e-05,
      "loss": 0.7932,
      "mean_token_accuracy": 0.7673261016607285,
      "num_tokens": 242178997.0,
      "step": 7588
    },
    {
      "entropy": 0.7378744874149561,
      "epoch": 0.698236352867436,
      "grad_norm": 0.15636314451694489,
      "learning_rate": 7.672892323734168e-05,
      "loss": 0.7077,
      "mean_token_accuracy": 0.7912381589412689,
      "num_tokens": 242211363.0,
      "step": 7589
    },
    {
      "entropy": 0.695764034986496,
      "epoch": 0.6983283592388773,
      "grad_norm": 0.16521695256233215,
      "learning_rate": 7.672585641120005e-05,
      "loss": 0.6825,
      "mean_token_accuracy": 0.8008203692734241,
      "num_tokens": 242242302.0,
      "step": 7590
    },
    {
      "entropy": 0.6144939111545682,
      "epoch": 0.6984203656103185,
      "grad_norm": 0.15594683587551117,
      "learning_rate": 7.672278958505843e-05,
      "loss": 0.6037,
      "mean_token_accuracy": 0.8196345679461956,
      "num_tokens": 242274512.0,
      "step": 7591
    },
    {
      "entropy": 0.7684876210987568,
      "epoch": 0.6985123719817597,
      "grad_norm": 0.15255193412303925,
      "learning_rate": 7.67197227589168e-05,
      "loss": 0.7432,
      "mean_token_accuracy": 0.7832121290266514,
      "num_tokens": 242306436.0,
      "step": 7592
    },
    {
      "entropy": 0.8386517446488142,
      "epoch": 0.698604378353201,
      "grad_norm": 0.16076956689357758,
      "learning_rate": 7.671665593277517e-05,
      "loss": 0.8361,
      "mean_token_accuracy": 0.7561880610883236,
      "num_tokens": 242338768.0,
      "step": 7593
    },
    {
      "entropy": 0.7564974930137396,
      "epoch": 0.6986963847246421,
      "grad_norm": 0.16641397774219513,
      "learning_rate": 7.671358910663355e-05,
      "loss": 0.7497,
      "mean_token_accuracy": 0.7773574404418468,
      "num_tokens": 242369979.0,
      "step": 7594
    },
    {
      "entropy": 0.7363650985062122,
      "epoch": 0.6987883910960834,
      "grad_norm": 0.15177984535694122,
      "learning_rate": 7.671052228049193e-05,
      "loss": 0.723,
      "mean_token_accuracy": 0.7853324823081493,
      "num_tokens": 242402610.0,
      "step": 7595
    },
    {
      "entropy": 0.6202778564766049,
      "epoch": 0.6988803974675246,
      "grad_norm": 0.155051589012146,
      "learning_rate": 7.67074554543503e-05,
      "loss": 0.6096,
      "mean_token_accuracy": 0.8184843622148037,
      "num_tokens": 242435378.0,
      "step": 7596
    },
    {
      "entropy": 0.7546891625970602,
      "epoch": 0.6989724038389659,
      "grad_norm": 0.15702076256275177,
      "learning_rate": 7.670438862820866e-05,
      "loss": 0.7659,
      "mean_token_accuracy": 0.7753170765936375,
      "num_tokens": 242467692.0,
      "step": 7597
    },
    {
      "entropy": 0.7038064934313297,
      "epoch": 0.6990644102104071,
      "grad_norm": 0.1545644998550415,
      "learning_rate": 7.670132180206705e-05,
      "loss": 0.6805,
      "mean_token_accuracy": 0.7971108816564083,
      "num_tokens": 242499313.0,
      "step": 7598
    },
    {
      "entropy": 0.7647252511233091,
      "epoch": 0.6991564165818482,
      "grad_norm": 0.16649474203586578,
      "learning_rate": 7.669825497592543e-05,
      "loss": 0.7524,
      "mean_token_accuracy": 0.7781814783811569,
      "num_tokens": 242532081.0,
      "step": 7599
    },
    {
      "entropy": 0.6631166134029627,
      "epoch": 0.6992484229532895,
      "grad_norm": 0.15082499384880066,
      "learning_rate": 7.66951881497838e-05,
      "loss": 0.6389,
      "mean_token_accuracy": 0.807837937027216,
      "num_tokens": 242563515.0,
      "step": 7600
    },
    {
      "entropy": 0.7767475731670856,
      "epoch": 0.6993404293247307,
      "grad_norm": 0.15851366519927979,
      "learning_rate": 7.669212132364216e-05,
      "loss": 0.7727,
      "mean_token_accuracy": 0.7679419293999672,
      "num_tokens": 242595773.0,
      "step": 7601
    },
    {
      "entropy": 0.6794152278453112,
      "epoch": 0.699432435696172,
      "grad_norm": 0.1542198210954666,
      "learning_rate": 7.668905449750054e-05,
      "loss": 0.6671,
      "mean_token_accuracy": 0.8007785491645336,
      "num_tokens": 242627483.0,
      "step": 7602
    },
    {
      "entropy": 0.806675847619772,
      "epoch": 0.6995244420676132,
      "grad_norm": 0.16042479872703552,
      "learning_rate": 7.668598767135891e-05,
      "loss": 0.8027,
      "mean_token_accuracy": 0.760586567223072,
      "num_tokens": 242659436.0,
      "step": 7603
    },
    {
      "entropy": 0.7055428437888622,
      "epoch": 0.6996164484390544,
      "grad_norm": 0.15725025534629822,
      "learning_rate": 7.66829208452173e-05,
      "loss": 0.7112,
      "mean_token_accuracy": 0.7900264114141464,
      "num_tokens": 242691867.0,
      "step": 7604
    },
    {
      "entropy": 0.7164786159992218,
      "epoch": 0.6997084548104956,
      "grad_norm": 0.16067564487457275,
      "learning_rate": 7.667985401907566e-05,
      "loss": 0.7046,
      "mean_token_accuracy": 0.7906749360263348,
      "num_tokens": 242723696.0,
      "step": 7605
    },
    {
      "entropy": 0.6637847619131207,
      "epoch": 0.6998004611819368,
      "grad_norm": 0.1603822559118271,
      "learning_rate": 7.667678719293404e-05,
      "loss": 0.6432,
      "mean_token_accuracy": 0.806567057967186,
      "num_tokens": 242755283.0,
      "step": 7606
    },
    {
      "entropy": 0.6475533973425627,
      "epoch": 0.6998924675533781,
      "grad_norm": 0.14849303662776947,
      "learning_rate": 7.667372036679241e-05,
      "loss": 0.6191,
      "mean_token_accuracy": 0.8133828341960907,
      "num_tokens": 242787076.0,
      "step": 7607
    },
    {
      "entropy": 0.7253998816013336,
      "epoch": 0.6999844739248193,
      "grad_norm": 0.15664871037006378,
      "learning_rate": 7.667065354065078e-05,
      "loss": 0.7134,
      "mean_token_accuracy": 0.7936402149498463,
      "num_tokens": 242818688.0,
      "step": 7608
    },
    {
      "entropy": 0.7337890844792128,
      "epoch": 0.7000764802962605,
      "grad_norm": 0.16120627522468567,
      "learning_rate": 7.666758671450916e-05,
      "loss": 0.723,
      "mean_token_accuracy": 0.78878553211689,
      "num_tokens": 242851201.0,
      "step": 7609
    },
    {
      "entropy": 0.6738234534859657,
      "epoch": 0.7001684866677017,
      "grad_norm": 0.1560145914554596,
      "learning_rate": 7.666451988836754e-05,
      "loss": 0.6581,
      "mean_token_accuracy": 0.8074146173894405,
      "num_tokens": 242882806.0,
      "step": 7610
    },
    {
      "entropy": 0.8461884576827288,
      "epoch": 0.700260493039143,
      "grad_norm": 0.16064859926700592,
      "learning_rate": 7.666145306222591e-05,
      "loss": 0.8491,
      "mean_token_accuracy": 0.7563928514719009,
      "num_tokens": 242914984.0,
      "step": 7611
    },
    {
      "entropy": 0.8005228042602539,
      "epoch": 0.7003524994105842,
      "grad_norm": 0.16478639841079712,
      "learning_rate": 7.665838623608428e-05,
      "loss": 0.7938,
      "mean_token_accuracy": 0.7615905776619911,
      "num_tokens": 242946679.0,
      "step": 7612
    },
    {
      "entropy": 0.8261461295187473,
      "epoch": 0.7004445057820254,
      "grad_norm": 0.16149502992630005,
      "learning_rate": 7.665531940994265e-05,
      "loss": 0.8109,
      "mean_token_accuracy": 0.7635314986109734,
      "num_tokens": 242978778.0,
      "step": 7613
    },
    {
      "entropy": 0.6754465373232961,
      "epoch": 0.7005365121534666,
      "grad_norm": 0.14962594211101532,
      "learning_rate": 7.665225258380103e-05,
      "loss": 0.6415,
      "mean_token_accuracy": 0.8087694570422173,
      "num_tokens": 243011193.0,
      "step": 7614
    },
    {
      "entropy": 0.7069093883037567,
      "epoch": 0.7006285185249078,
      "grad_norm": 0.15102525055408478,
      "learning_rate": 7.664918575765941e-05,
      "loss": 0.6832,
      "mean_token_accuracy": 0.7930694222450256,
      "num_tokens": 243042773.0,
      "step": 7615
    },
    {
      "entropy": 0.6938801780343056,
      "epoch": 0.7007205248963491,
      "grad_norm": 0.1559520810842514,
      "learning_rate": 7.664611893151778e-05,
      "loss": 0.6888,
      "mean_token_accuracy": 0.7912627086043358,
      "num_tokens": 243074905.0,
      "step": 7616
    },
    {
      "entropy": 0.7394914366304874,
      "epoch": 0.7008125312677903,
      "grad_norm": 0.15550178289413452,
      "learning_rate": 7.664305210537614e-05,
      "loss": 0.7206,
      "mean_token_accuracy": 0.7811341546475887,
      "num_tokens": 243106372.0,
      "step": 7617
    },
    {
      "entropy": 0.755211291834712,
      "epoch": 0.7009045376392316,
      "grad_norm": 0.163259819149971,
      "learning_rate": 7.663998527923453e-05,
      "loss": 0.7606,
      "mean_token_accuracy": 0.7754626162350178,
      "num_tokens": 243138403.0,
      "step": 7618
    },
    {
      "entropy": 0.7355961203575134,
      "epoch": 0.7009965440106727,
      "grad_norm": 0.16044124960899353,
      "learning_rate": 7.663691845309289e-05,
      "loss": 0.7292,
      "mean_token_accuracy": 0.785854984074831,
      "num_tokens": 243169935.0,
      "step": 7619
    },
    {
      "entropy": 0.6787296775728464,
      "epoch": 0.7010885503821139,
      "grad_norm": 0.1493343859910965,
      "learning_rate": 7.663385162695127e-05,
      "loss": 0.6699,
      "mean_token_accuracy": 0.8002457469701767,
      "num_tokens": 243202001.0,
      "step": 7620
    },
    {
      "entropy": 0.767768383026123,
      "epoch": 0.7011805567535552,
      "grad_norm": 0.1681603640317917,
      "learning_rate": 7.663078480080964e-05,
      "loss": 0.7661,
      "mean_token_accuracy": 0.7768210545182228,
      "num_tokens": 243234436.0,
      "step": 7621
    },
    {
      "entropy": 0.8000023681670427,
      "epoch": 0.7012725631249964,
      "grad_norm": 0.16010229289531708,
      "learning_rate": 7.662771797466802e-05,
      "loss": 0.8034,
      "mean_token_accuracy": 0.7675610966980457,
      "num_tokens": 243266470.0,
      "step": 7622
    },
    {
      "entropy": 0.6794110964983702,
      "epoch": 0.7013645694964377,
      "grad_norm": 0.1503300964832306,
      "learning_rate": 7.662465114852639e-05,
      "loss": 0.6665,
      "mean_token_accuracy": 0.802143894135952,
      "num_tokens": 243298589.0,
      "step": 7623
    },
    {
      "entropy": 0.8396780733019114,
      "epoch": 0.7014565758678788,
      "grad_norm": 0.15696683526039124,
      "learning_rate": 7.662158432238476e-05,
      "loss": 0.8238,
      "mean_token_accuracy": 0.7561744377017021,
      "num_tokens": 243330409.0,
      "step": 7624
    },
    {
      "entropy": 0.837322536855936,
      "epoch": 0.70154858223932,
      "grad_norm": 0.17641082406044006,
      "learning_rate": 7.661851749624314e-05,
      "loss": 0.8393,
      "mean_token_accuracy": 0.7595140933990479,
      "num_tokens": 243361940.0,
      "step": 7625
    },
    {
      "entropy": 0.80959577485919,
      "epoch": 0.7016405886107613,
      "grad_norm": 0.1627262383699417,
      "learning_rate": 7.661545067010152e-05,
      "loss": 0.7933,
      "mean_token_accuracy": 0.7707135900855064,
      "num_tokens": 243393825.0,
      "step": 7626
    },
    {
      "entropy": 0.8688065409660339,
      "epoch": 0.7017325949822025,
      "grad_norm": 0.157822385430336,
      "learning_rate": 7.661238384395989e-05,
      "loss": 0.861,
      "mean_token_accuracy": 0.749636746942997,
      "num_tokens": 243425631.0,
      "step": 7627
    },
    {
      "entropy": 0.8274762965738773,
      "epoch": 0.7018246013536438,
      "grad_norm": 0.16837841272354126,
      "learning_rate": 7.660931701781826e-05,
      "loss": 0.8172,
      "mean_token_accuracy": 0.7590423114597797,
      "num_tokens": 243457375.0,
      "step": 7628
    },
    {
      "entropy": 0.7912719622254372,
      "epoch": 0.7019166077250849,
      "grad_norm": 0.15320686995983124,
      "learning_rate": 7.660625019167664e-05,
      "loss": 0.7749,
      "mean_token_accuracy": 0.7714890167117119,
      "num_tokens": 243488533.0,
      "step": 7629
    },
    {
      "entropy": 0.7986083664000034,
      "epoch": 0.7020086140965262,
      "grad_norm": 0.1569005399942398,
      "learning_rate": 7.660318336553501e-05,
      "loss": 0.7838,
      "mean_token_accuracy": 0.7688675038516521,
      "num_tokens": 243520784.0,
      "step": 7630
    },
    {
      "entropy": 0.7508621215820312,
      "epoch": 0.7021006204679674,
      "grad_norm": 0.14884549379348755,
      "learning_rate": 7.660011653939339e-05,
      "loss": 0.7252,
      "mean_token_accuracy": 0.7863065302371979,
      "num_tokens": 243552860.0,
      "step": 7631
    },
    {
      "entropy": 0.6673137880861759,
      "epoch": 0.7021926268394086,
      "grad_norm": 0.18313084542751312,
      "learning_rate": 7.659704971325176e-05,
      "loss": 0.6613,
      "mean_token_accuracy": 0.8038823492825031,
      "num_tokens": 243584245.0,
      "step": 7632
    },
    {
      "entropy": 0.7577792480587959,
      "epoch": 0.7022846332108499,
      "grad_norm": 0.15880708396434784,
      "learning_rate": 7.659398288711014e-05,
      "loss": 0.753,
      "mean_token_accuracy": 0.7754039578139782,
      "num_tokens": 243616014.0,
      "step": 7633
    },
    {
      "entropy": 0.7472808677703142,
      "epoch": 0.7023766395822911,
      "grad_norm": 0.1631411910057068,
      "learning_rate": 7.65909160609685e-05,
      "loss": 0.7522,
      "mean_token_accuracy": 0.7805676981806755,
      "num_tokens": 243648388.0,
      "step": 7634
    },
    {
      "entropy": 0.6803871504962444,
      "epoch": 0.7024686459537323,
      "grad_norm": 0.1649216264486313,
      "learning_rate": 7.658784923482689e-05,
      "loss": 0.6824,
      "mean_token_accuracy": 0.8019695430994034,
      "num_tokens": 243681032.0,
      "step": 7635
    },
    {
      "entropy": 0.893890218809247,
      "epoch": 0.7025606523251735,
      "grad_norm": 0.17342601716518402,
      "learning_rate": 7.658478240868526e-05,
      "loss": 0.882,
      "mean_token_accuracy": 0.7472648732364178,
      "num_tokens": 243712771.0,
      "step": 7636
    },
    {
      "entropy": 0.6216355171054602,
      "epoch": 0.7026526586966148,
      "grad_norm": 0.15734365582466125,
      "learning_rate": 7.658171558254364e-05,
      "loss": 0.6106,
      "mean_token_accuracy": 0.8185278289020061,
      "num_tokens": 243744370.0,
      "step": 7637
    },
    {
      "entropy": 0.8754879757761955,
      "epoch": 0.702744665068056,
      "grad_norm": 0.16599132120609283,
      "learning_rate": 7.6578648756402e-05,
      "loss": 0.8698,
      "mean_token_accuracy": 0.7470797635614872,
      "num_tokens": 243776306.0,
      "step": 7638
    },
    {
      "entropy": 0.8465130366384983,
      "epoch": 0.7028366714394972,
      "grad_norm": 0.16264158487319946,
      "learning_rate": 7.657558193026037e-05,
      "loss": 0.8359,
      "mean_token_accuracy": 0.7594253197312355,
      "num_tokens": 243807104.0,
      "step": 7639
    },
    {
      "entropy": 0.8036378622055054,
      "epoch": 0.7029286778109384,
      "grad_norm": 0.15965670347213745,
      "learning_rate": 7.657251510411875e-05,
      "loss": 0.8058,
      "mean_token_accuracy": 0.7677075006067753,
      "num_tokens": 243839475.0,
      "step": 7640
    },
    {
      "entropy": 0.7541009895503521,
      "epoch": 0.7030206841823796,
      "grad_norm": 0.16159340739250183,
      "learning_rate": 7.656944827797714e-05,
      "loss": 0.7415,
      "mean_token_accuracy": 0.7764008119702339,
      "num_tokens": 243871595.0,
      "step": 7641
    },
    {
      "entropy": 0.8762995544821024,
      "epoch": 0.7031126905538209,
      "grad_norm": 0.16330844163894653,
      "learning_rate": 7.65663814518355e-05,
      "loss": 0.8623,
      "mean_token_accuracy": 0.7495631985366344,
      "num_tokens": 243904111.0,
      "step": 7642
    },
    {
      "entropy": 0.7416193261742592,
      "epoch": 0.7032046969252621,
      "grad_norm": 0.15308141708374023,
      "learning_rate": 7.656331462569387e-05,
      "loss": 0.7117,
      "mean_token_accuracy": 0.791195023804903,
      "num_tokens": 243936544.0,
      "step": 7643
    },
    {
      "entropy": 0.7398144025355577,
      "epoch": 0.7032967032967034,
      "grad_norm": 0.162959486246109,
      "learning_rate": 7.656024779955224e-05,
      "loss": 0.7413,
      "mean_token_accuracy": 0.7810872383415699,
      "num_tokens": 243967664.0,
      "step": 7644
    },
    {
      "entropy": 0.8245735354721546,
      "epoch": 0.7033887096681445,
      "grad_norm": 0.15420827269554138,
      "learning_rate": 7.655718097341062e-05,
      "loss": 0.8172,
      "mean_token_accuracy": 0.7579618133604527,
      "num_tokens": 244000069.0,
      "step": 7645
    },
    {
      "entropy": 0.8574688117951155,
      "epoch": 0.7034807160395857,
      "grad_norm": 0.15950457751750946,
      "learning_rate": 7.6554114147269e-05,
      "loss": 0.8605,
      "mean_token_accuracy": 0.7517339773476124,
      "num_tokens": 244031725.0,
      "step": 7646
    },
    {
      "entropy": 0.7952936924993992,
      "epoch": 0.703572722411027,
      "grad_norm": 0.1546565592288971,
      "learning_rate": 7.655104732112737e-05,
      "loss": 0.7975,
      "mean_token_accuracy": 0.7678984962403774,
      "num_tokens": 244063076.0,
      "step": 7647
    },
    {
      "entropy": 0.7031687721610069,
      "epoch": 0.7036647287824682,
      "grad_norm": 0.15763190388679504,
      "learning_rate": 7.654798049498574e-05,
      "loss": 0.7065,
      "mean_token_accuracy": 0.7896900996565819,
      "num_tokens": 244095194.0,
      "step": 7648
    },
    {
      "entropy": 0.7480364516377449,
      "epoch": 0.7037567351539095,
      "grad_norm": 0.15901979804039001,
      "learning_rate": 7.654491366884412e-05,
      "loss": 0.7374,
      "mean_token_accuracy": 0.7852096781134605,
      "num_tokens": 244126684.0,
      "step": 7649
    },
    {
      "entropy": 0.7617840133607388,
      "epoch": 0.7038487415253506,
      "grad_norm": 0.14799635112285614,
      "learning_rate": 7.654184684270249e-05,
      "loss": 0.7412,
      "mean_token_accuracy": 0.7819410488009453,
      "num_tokens": 244158894.0,
      "step": 7650
    },
    {
      "entropy": 0.8097643330693245,
      "epoch": 0.7039407478967918,
      "grad_norm": 0.15385381877422333,
      "learning_rate": 7.653878001656087e-05,
      "loss": 0.8022,
      "mean_token_accuracy": 0.7646013088524342,
      "num_tokens": 244190431.0,
      "step": 7651
    },
    {
      "entropy": 0.7142127454280853,
      "epoch": 0.7040327542682331,
      "grad_norm": 0.16659943759441376,
      "learning_rate": 7.653571319041924e-05,
      "loss": 0.7145,
      "mean_token_accuracy": 0.7863111086189747,
      "num_tokens": 244222283.0,
      "step": 7652
    },
    {
      "entropy": 0.7652881955727935,
      "epoch": 0.7041247606396743,
      "grad_norm": 0.14964427053928375,
      "learning_rate": 7.653264636427762e-05,
      "loss": 0.7547,
      "mean_token_accuracy": 0.7805241197347641,
      "num_tokens": 244254927.0,
      "step": 7653
    },
    {
      "entropy": 0.7991951033473015,
      "epoch": 0.7042167670111156,
      "grad_norm": 0.15512171387672424,
      "learning_rate": 7.652957953813599e-05,
      "loss": 0.7937,
      "mean_token_accuracy": 0.7654928639531136,
      "num_tokens": 244286853.0,
      "step": 7654
    },
    {
      "entropy": 0.7823294922709465,
      "epoch": 0.7043087733825567,
      "grad_norm": 0.15715454518795013,
      "learning_rate": 7.652651271199435e-05,
      "loss": 0.7768,
      "mean_token_accuracy": 0.7727909497916698,
      "num_tokens": 244319063.0,
      "step": 7655
    },
    {
      "entropy": 0.7177119264379144,
      "epoch": 0.704400779753998,
      "grad_norm": 0.15213745832443237,
      "learning_rate": 7.652344588585273e-05,
      "loss": 0.7141,
      "mean_token_accuracy": 0.7892466299235821,
      "num_tokens": 244350352.0,
      "step": 7656
    },
    {
      "entropy": 0.8000492043793201,
      "epoch": 0.7044927861254392,
      "grad_norm": 0.1646660417318344,
      "learning_rate": 7.652037905971112e-05,
      "loss": 0.799,
      "mean_token_accuracy": 0.7665712125599384,
      "num_tokens": 244382573.0,
      "step": 7657
    },
    {
      "entropy": 0.7314578238874674,
      "epoch": 0.7045847924968804,
      "grad_norm": 0.15458540618419647,
      "learning_rate": 7.651731223356948e-05,
      "loss": 0.7077,
      "mean_token_accuracy": 0.79664646089077,
      "num_tokens": 244414600.0,
      "step": 7658
    },
    {
      "entropy": 0.7465539053082466,
      "epoch": 0.7046767988683217,
      "grad_norm": 0.1686175912618637,
      "learning_rate": 7.651424540742785e-05,
      "loss": 0.7421,
      "mean_token_accuracy": 0.7808337435126305,
      "num_tokens": 244446165.0,
      "step": 7659
    },
    {
      "entropy": 0.7496860548853874,
      "epoch": 0.7047688052397628,
      "grad_norm": 0.1515577733516693,
      "learning_rate": 7.651117858128623e-05,
      "loss": 0.7478,
      "mean_token_accuracy": 0.7802695967257023,
      "num_tokens": 244478149.0,
      "step": 7660
    },
    {
      "entropy": 0.6279978100210428,
      "epoch": 0.7048608116112041,
      "grad_norm": 0.16005989909172058,
      "learning_rate": 7.65081117551446e-05,
      "loss": 0.6129,
      "mean_token_accuracy": 0.8216118104755878,
      "num_tokens": 244510713.0,
      "step": 7661
    },
    {
      "entropy": 0.8142352923750877,
      "epoch": 0.7049528179826453,
      "grad_norm": 0.15184378623962402,
      "learning_rate": 7.650504492900298e-05,
      "loss": 0.8038,
      "mean_token_accuracy": 0.7647728845477104,
      "num_tokens": 244542168.0,
      "step": 7662
    },
    {
      "entropy": 0.765898022800684,
      "epoch": 0.7050448243540866,
      "grad_norm": 0.15901324152946472,
      "learning_rate": 7.650197810286135e-05,
      "loss": 0.7483,
      "mean_token_accuracy": 0.7810634337365627,
      "num_tokens": 244573884.0,
      "step": 7663
    },
    {
      "entropy": 0.765862743370235,
      "epoch": 0.7051368307255278,
      "grad_norm": 0.1600445806980133,
      "learning_rate": 7.649891127671973e-05,
      "loss": 0.7638,
      "mean_token_accuracy": 0.7751954458653927,
      "num_tokens": 244606423.0,
      "step": 7664
    },
    {
      "entropy": 0.7707786429673433,
      "epoch": 0.7052288370969689,
      "grad_norm": 0.15818414092063904,
      "learning_rate": 7.64958444505781e-05,
      "loss": 0.7698,
      "mean_token_accuracy": 0.776370219886303,
      "num_tokens": 244638593.0,
      "step": 7665
    },
    {
      "entropy": 0.8006665501743555,
      "epoch": 0.7053208434684102,
      "grad_norm": 0.16076193749904633,
      "learning_rate": 7.649277762443647e-05,
      "loss": 0.782,
      "mean_token_accuracy": 0.7707730457186699,
      "num_tokens": 244670582.0,
      "step": 7666
    },
    {
      "entropy": 0.6718538273125887,
      "epoch": 0.7054128498398514,
      "grad_norm": 0.14689680933952332,
      "learning_rate": 7.648971079829485e-05,
      "loss": 0.6541,
      "mean_token_accuracy": 0.8020724728703499,
      "num_tokens": 244702100.0,
      "step": 7667
    },
    {
      "entropy": 0.7571428865194321,
      "epoch": 0.7055048562112927,
      "grad_norm": 0.1495533436536789,
      "learning_rate": 7.648664397215323e-05,
      "loss": 0.7237,
      "mean_token_accuracy": 0.7815588675439358,
      "num_tokens": 244734488.0,
      "step": 7668
    },
    {
      "entropy": 0.7070229910314083,
      "epoch": 0.7055968625827339,
      "grad_norm": 0.15777841210365295,
      "learning_rate": 7.64835771460116e-05,
      "loss": 0.6962,
      "mean_token_accuracy": 0.7942557521164417,
      "num_tokens": 244766791.0,
      "step": 7669
    },
    {
      "entropy": 0.7642288729548454,
      "epoch": 0.705688868954175,
      "grad_norm": 0.14954622089862823,
      "learning_rate": 7.648051031986997e-05,
      "loss": 0.7541,
      "mean_token_accuracy": 0.7732691168785095,
      "num_tokens": 244798247.0,
      "step": 7670
    },
    {
      "entropy": 0.748133109882474,
      "epoch": 0.7057808753256163,
      "grad_norm": 0.1577063351869583,
      "learning_rate": 7.647744349372833e-05,
      "loss": 0.7243,
      "mean_token_accuracy": 0.7838017120957375,
      "num_tokens": 244829421.0,
      "step": 7671
    },
    {
      "entropy": 0.7034579142928123,
      "epoch": 0.7058728816970575,
      "grad_norm": 0.15316817164421082,
      "learning_rate": 7.647437666758673e-05,
      "loss": 0.7163,
      "mean_token_accuracy": 0.7854543961584568,
      "num_tokens": 244861943.0,
      "step": 7672
    },
    {
      "entropy": 0.6825434062629938,
      "epoch": 0.7059648880684988,
      "grad_norm": 0.16040782630443573,
      "learning_rate": 7.64713098414451e-05,
      "loss": 0.6824,
      "mean_token_accuracy": 0.7968901470303535,
      "num_tokens": 244893825.0,
      "step": 7673
    },
    {
      "entropy": 0.7395736388862133,
      "epoch": 0.70605689443994,
      "grad_norm": 0.15333908796310425,
      "learning_rate": 7.646824301530346e-05,
      "loss": 0.7256,
      "mean_token_accuracy": 0.788584504276514,
      "num_tokens": 244925837.0,
      "step": 7674
    },
    {
      "entropy": 0.6809925455600023,
      "epoch": 0.7061489008113812,
      "grad_norm": 0.1575118452310562,
      "learning_rate": 7.646517618916183e-05,
      "loss": 0.6745,
      "mean_token_accuracy": 0.7932965345680714,
      "num_tokens": 244956408.0,
      "step": 7675
    },
    {
      "entropy": 0.7466193418949842,
      "epoch": 0.7062409071828224,
      "grad_norm": 0.16313517093658447,
      "learning_rate": 7.646210936302021e-05,
      "loss": 0.7575,
      "mean_token_accuracy": 0.7772349752485752,
      "num_tokens": 244989176.0,
      "step": 7676
    },
    {
      "entropy": 0.6508464310318232,
      "epoch": 0.7063329135542636,
      "grad_norm": 0.1589856594800949,
      "learning_rate": 7.64590425368786e-05,
      "loss": 0.6522,
      "mean_token_accuracy": 0.8040228523313999,
      "num_tokens": 245021182.0,
      "step": 7677
    },
    {
      "entropy": 0.7164694629609585,
      "epoch": 0.7064249199257049,
      "grad_norm": 0.1479829102754593,
      "learning_rate": 7.645597571073696e-05,
      "loss": 0.7016,
      "mean_token_accuracy": 0.795869767665863,
      "num_tokens": 245053868.0,
      "step": 7678
    },
    {
      "entropy": 0.6473747193813324,
      "epoch": 0.7065169262971461,
      "grad_norm": 0.1485128402709961,
      "learning_rate": 7.645290888459533e-05,
      "loss": 0.6303,
      "mean_token_accuracy": 0.8104803711175919,
      "num_tokens": 245085587.0,
      "step": 7679
    },
    {
      "entropy": 0.6614797860383987,
      "epoch": 0.7066089326685873,
      "grad_norm": 0.15874788165092468,
      "learning_rate": 7.644984205845371e-05,
      "loss": 0.6269,
      "mean_token_accuracy": 0.8099324591457844,
      "num_tokens": 245117631.0,
      "step": 7680
    },
    {
      "entropy": 0.729758009314537,
      "epoch": 0.7067009390400285,
      "grad_norm": 0.15084581077098846,
      "learning_rate": 7.644677523231208e-05,
      "loss": 0.7007,
      "mean_token_accuracy": 0.7933681979775429,
      "num_tokens": 245149431.0,
      "step": 7681
    },
    {
      "entropy": 0.7810043077915907,
      "epoch": 0.7067929454114698,
      "grad_norm": 0.15181227028369904,
      "learning_rate": 7.644370840617046e-05,
      "loss": 0.7739,
      "mean_token_accuracy": 0.7717147059738636,
      "num_tokens": 245180591.0,
      "step": 7682
    },
    {
      "entropy": 0.7115541249513626,
      "epoch": 0.706884951782911,
      "grad_norm": 0.15495151281356812,
      "learning_rate": 7.644064158002883e-05,
      "loss": 0.6958,
      "mean_token_accuracy": 0.795132365077734,
      "num_tokens": 245213076.0,
      "step": 7683
    },
    {
      "entropy": 0.7706594727933407,
      "epoch": 0.7069769581543522,
      "grad_norm": 0.16620995104312897,
      "learning_rate": 7.643757475388721e-05,
      "loss": 0.7601,
      "mean_token_accuracy": 0.7786032110452652,
      "num_tokens": 245244688.0,
      "step": 7684
    },
    {
      "entropy": 0.6643151976168156,
      "epoch": 0.7070689645257934,
      "grad_norm": 0.16368162631988525,
      "learning_rate": 7.643450792774558e-05,
      "loss": 0.6582,
      "mean_token_accuracy": 0.8038841485977173,
      "num_tokens": 245276443.0,
      "step": 7685
    },
    {
      "entropy": 0.7632371708750725,
      "epoch": 0.7071609708972346,
      "grad_norm": 0.15856462717056274,
      "learning_rate": 7.643144110160395e-05,
      "loss": 0.7757,
      "mean_token_accuracy": 0.7714338041841984,
      "num_tokens": 245309211.0,
      "step": 7686
    },
    {
      "entropy": 0.6090748328715563,
      "epoch": 0.7072529772686759,
      "grad_norm": 0.1495417058467865,
      "learning_rate": 7.642837427546233e-05,
      "loss": 0.5819,
      "mean_token_accuracy": 0.8245352916419506,
      "num_tokens": 245341488.0,
      "step": 7687
    },
    {
      "entropy": 0.7797046536579728,
      "epoch": 0.7073449836401171,
      "grad_norm": 0.15519559383392334,
      "learning_rate": 7.642530744932071e-05,
      "loss": 0.7804,
      "mean_token_accuracy": 0.7702047862112522,
      "num_tokens": 245373468.0,
      "step": 7688
    },
    {
      "entropy": 0.7170884907245636,
      "epoch": 0.7074369900115584,
      "grad_norm": 0.15768687427043915,
      "learning_rate": 7.642224062317908e-05,
      "loss": 0.6959,
      "mean_token_accuracy": 0.7949527502059937,
      "num_tokens": 245404322.0,
      "step": 7689
    },
    {
      "entropy": 0.7725948467850685,
      "epoch": 0.7075289963829995,
      "grad_norm": 0.1593724638223648,
      "learning_rate": 7.641917379703745e-05,
      "loss": 0.7629,
      "mean_token_accuracy": 0.7775341831147671,
      "num_tokens": 245436613.0,
      "step": 7690
    },
    {
      "entropy": 0.6819283608347178,
      "epoch": 0.7076210027544407,
      "grad_norm": 0.15016543865203857,
      "learning_rate": 7.641610697089583e-05,
      "loss": 0.6677,
      "mean_token_accuracy": 0.8024352192878723,
      "num_tokens": 245467465.0,
      "step": 7691
    },
    {
      "entropy": 0.6789554916322231,
      "epoch": 0.707713009125882,
      "grad_norm": 0.1513652503490448,
      "learning_rate": 7.64130401447542e-05,
      "loss": 0.6479,
      "mean_token_accuracy": 0.8103830963373184,
      "num_tokens": 245499412.0,
      "step": 7692
    },
    {
      "entropy": 0.7533810120075941,
      "epoch": 0.7078050154973232,
      "grad_norm": 0.1637014001607895,
      "learning_rate": 7.640997331861258e-05,
      "loss": 0.751,
      "mean_token_accuracy": 0.7792380042374134,
      "num_tokens": 245531019.0,
      "step": 7693
    },
    {
      "entropy": 0.6371631100773811,
      "epoch": 0.7078970218687645,
      "grad_norm": 0.15946808457374573,
      "learning_rate": 7.640690649247094e-05,
      "loss": 0.6179,
      "mean_token_accuracy": 0.8132961466908455,
      "num_tokens": 245563245.0,
      "step": 7694
    },
    {
      "entropy": 0.7735275998711586,
      "epoch": 0.7079890282402056,
      "grad_norm": 0.1603754758834839,
      "learning_rate": 7.640383966632933e-05,
      "loss": 0.762,
      "mean_token_accuracy": 0.7747266441583633,
      "num_tokens": 245595005.0,
      "step": 7695
    },
    {
      "entropy": 0.7088129110634327,
      "epoch": 0.7080810346116468,
      "grad_norm": 0.15929804742336273,
      "learning_rate": 7.64007728401877e-05,
      "loss": 0.6977,
      "mean_token_accuracy": 0.7934445217251778,
      "num_tokens": 245626845.0,
      "step": 7696
    },
    {
      "entropy": 0.7120589893311262,
      "epoch": 0.7081730409830881,
      "grad_norm": 0.1540215015411377,
      "learning_rate": 7.639770601404606e-05,
      "loss": 0.7144,
      "mean_token_accuracy": 0.7856374606490135,
      "num_tokens": 245658535.0,
      "step": 7697
    },
    {
      "entropy": 0.7811185419559479,
      "epoch": 0.7082650473545293,
      "grad_norm": 0.16395685076713562,
      "learning_rate": 7.639463918790444e-05,
      "loss": 0.7937,
      "mean_token_accuracy": 0.7658125683665276,
      "num_tokens": 245690667.0,
      "step": 7698
    },
    {
      "entropy": 0.7430416382849216,
      "epoch": 0.7083570537259706,
      "grad_norm": 0.16229087114334106,
      "learning_rate": 7.639157236176282e-05,
      "loss": 0.7458,
      "mean_token_accuracy": 0.7830075994133949,
      "num_tokens": 245722440.0,
      "step": 7699
    },
    {
      "entropy": 0.8000804651528597,
      "epoch": 0.7084490600974117,
      "grad_norm": 0.17133809626102448,
      "learning_rate": 7.638850553562119e-05,
      "loss": 0.8095,
      "mean_token_accuracy": 0.7619768939912319,
      "num_tokens": 245754159.0,
      "step": 7700
    },
    {
      "entropy": 0.9374744538217783,
      "epoch": 0.708541066468853,
      "grad_norm": 0.16593720018863678,
      "learning_rate": 7.638543870947956e-05,
      "loss": 0.9346,
      "mean_token_accuracy": 0.7316080071032047,
      "num_tokens": 245785997.0,
      "step": 7701
    },
    {
      "entropy": 0.7476457487791777,
      "epoch": 0.7086330728402942,
      "grad_norm": 0.15958671271800995,
      "learning_rate": 7.638237188333793e-05,
      "loss": 0.7393,
      "mean_token_accuracy": 0.7834824584424496,
      "num_tokens": 245817992.0,
      "step": 7702
    },
    {
      "entropy": 0.6874428149312735,
      "epoch": 0.7087250792117354,
      "grad_norm": 0.15295849740505219,
      "learning_rate": 7.637930505719631e-05,
      "loss": 0.6762,
      "mean_token_accuracy": 0.7980634905397892,
      "num_tokens": 245850100.0,
      "step": 7703
    },
    {
      "entropy": 0.7332922741770744,
      "epoch": 0.7088170855831767,
      "grad_norm": 0.15463975071907043,
      "learning_rate": 7.637623823105469e-05,
      "loss": 0.7197,
      "mean_token_accuracy": 0.7853688783943653,
      "num_tokens": 245881562.0,
      "step": 7704
    },
    {
      "entropy": 0.7456270717084408,
      "epoch": 0.7089090919546178,
      "grad_norm": 0.15755492448806763,
      "learning_rate": 7.637317140491306e-05,
      "loss": 0.7334,
      "mean_token_accuracy": 0.7878634668886662,
      "num_tokens": 245913380.0,
      "step": 7705
    },
    {
      "entropy": 0.741539852693677,
      "epoch": 0.7090010983260591,
      "grad_norm": 0.1556846648454666,
      "learning_rate": 7.637010457877143e-05,
      "loss": 0.7296,
      "mean_token_accuracy": 0.7862830348312855,
      "num_tokens": 245945494.0,
      "step": 7706
    },
    {
      "entropy": 0.8660062626004219,
      "epoch": 0.7090931046975003,
      "grad_norm": 0.1610516458749771,
      "learning_rate": 7.636703775262981e-05,
      "loss": 0.8628,
      "mean_token_accuracy": 0.7497411631047726,
      "num_tokens": 245977409.0,
      "step": 7707
    },
    {
      "entropy": 0.768998958170414,
      "epoch": 0.7091851110689416,
      "grad_norm": 0.1633235365152359,
      "learning_rate": 7.636397092648818e-05,
      "loss": 0.7571,
      "mean_token_accuracy": 0.7746807858347893,
      "num_tokens": 246009225.0,
      "step": 7708
    },
    {
      "entropy": 0.7713670842349529,
      "epoch": 0.7092771174403828,
      "grad_norm": 0.15848511457443237,
      "learning_rate": 7.636090410034656e-05,
      "loss": 0.7459,
      "mean_token_accuracy": 0.7745978944003582,
      "num_tokens": 246041136.0,
      "step": 7709
    },
    {
      "entropy": 0.8566139936447144,
      "epoch": 0.7093691238118239,
      "grad_norm": 0.15917418897151947,
      "learning_rate": 7.635783727420492e-05,
      "loss": 0.8426,
      "mean_token_accuracy": 0.7590745314955711,
      "num_tokens": 246073481.0,
      "step": 7710
    },
    {
      "entropy": 0.7720038276165724,
      "epoch": 0.7094611301832652,
      "grad_norm": 0.15933901071548462,
      "learning_rate": 7.63547704480633e-05,
      "loss": 0.7646,
      "mean_token_accuracy": 0.7767154984176159,
      "num_tokens": 246105422.0,
      "step": 7711
    },
    {
      "entropy": 0.7379417940974236,
      "epoch": 0.7095531365547064,
      "grad_norm": 0.15105661749839783,
      "learning_rate": 7.635170362192167e-05,
      "loss": 0.7176,
      "mean_token_accuracy": 0.7872959971427917,
      "num_tokens": 246137318.0,
      "step": 7712
    },
    {
      "entropy": 0.7422634847462177,
      "epoch": 0.7096451429261477,
      "grad_norm": 0.16348151862621307,
      "learning_rate": 7.634863679578004e-05,
      "loss": 0.7415,
      "mean_token_accuracy": 0.7804159075021744,
      "num_tokens": 246168950.0,
      "step": 7713
    },
    {
      "entropy": 0.7926849648356438,
      "epoch": 0.7097371492975889,
      "grad_norm": 0.16398972272872925,
      "learning_rate": 7.634556996963842e-05,
      "loss": 0.7785,
      "mean_token_accuracy": 0.7712567560374737,
      "num_tokens": 246200029.0,
      "step": 7714
    },
    {
      "entropy": 0.7306593675166368,
      "epoch": 0.70982915566903,
      "grad_norm": 0.1523219645023346,
      "learning_rate": 7.63425031434968e-05,
      "loss": 0.7042,
      "mean_token_accuracy": 0.7891686968505383,
      "num_tokens": 246231645.0,
      "step": 7715
    },
    {
      "entropy": 0.7144572343677282,
      "epoch": 0.7099211620404713,
      "grad_norm": 0.15155354142189026,
      "learning_rate": 7.633943631735517e-05,
      "loss": 0.6887,
      "mean_token_accuracy": 0.7955098561942577,
      "num_tokens": 246264147.0,
      "step": 7716
    },
    {
      "entropy": 0.7619734015315771,
      "epoch": 0.7100131684119125,
      "grad_norm": 0.16292253136634827,
      "learning_rate": 7.633636949121354e-05,
      "loss": 0.7467,
      "mean_token_accuracy": 0.7776306755840778,
      "num_tokens": 246295765.0,
      "step": 7717
    },
    {
      "entropy": 0.7643069978803396,
      "epoch": 0.7101051747833538,
      "grad_norm": 0.16074953973293304,
      "learning_rate": 7.633330266507191e-05,
      "loss": 0.7567,
      "mean_token_accuracy": 0.7756357975304127,
      "num_tokens": 246327015.0,
      "step": 7718
    },
    {
      "entropy": 0.7118450421839952,
      "epoch": 0.710197181154795,
      "grad_norm": 0.14952489733695984,
      "learning_rate": 7.63302358389303e-05,
      "loss": 0.7131,
      "mean_token_accuracy": 0.7882188446819782,
      "num_tokens": 246359172.0,
      "step": 7719
    },
    {
      "entropy": 0.778235737234354,
      "epoch": 0.7102891875262362,
      "grad_norm": 0.15898847579956055,
      "learning_rate": 7.632716901278867e-05,
      "loss": 0.7651,
      "mean_token_accuracy": 0.7739639095962048,
      "num_tokens": 246391307.0,
      "step": 7720
    },
    {
      "entropy": 0.7744659408926964,
      "epoch": 0.7103811938976774,
      "grad_norm": 0.16391873359680176,
      "learning_rate": 7.632410218664704e-05,
      "loss": 0.7925,
      "mean_token_accuracy": 0.7749700620770454,
      "num_tokens": 246423206.0,
      "step": 7721
    },
    {
      "entropy": 0.8283637277781963,
      "epoch": 0.7104732002691186,
      "grad_norm": 0.16494673490524292,
      "learning_rate": 7.632103536050542e-05,
      "loss": 0.8487,
      "mean_token_accuracy": 0.7570623457431793,
      "num_tokens": 246454957.0,
      "step": 7722
    },
    {
      "entropy": 0.7564982054755092,
      "epoch": 0.7105652066405599,
      "grad_norm": 0.1586618721485138,
      "learning_rate": 7.631796853436379e-05,
      "loss": 0.7295,
      "mean_token_accuracy": 0.78190703317523,
      "num_tokens": 246486599.0,
      "step": 7723
    },
    {
      "entropy": 0.6854599639773369,
      "epoch": 0.7106572130120011,
      "grad_norm": 0.15351048111915588,
      "learning_rate": 7.631490170822217e-05,
      "loss": 0.68,
      "mean_token_accuracy": 0.7948657311499119,
      "num_tokens": 246518859.0,
      "step": 7724
    },
    {
      "entropy": 0.812454666942358,
      "epoch": 0.7107492193834423,
      "grad_norm": 0.15157049894332886,
      "learning_rate": 7.631183488208054e-05,
      "loss": 0.8059,
      "mean_token_accuracy": 0.7637427859008312,
      "num_tokens": 246551219.0,
      "step": 7725
    },
    {
      "entropy": 0.5767068527638912,
      "epoch": 0.7108412257548835,
      "grad_norm": 0.16225332021713257,
      "learning_rate": 7.630876805593892e-05,
      "loss": 0.566,
      "mean_token_accuracy": 0.8303605765104294,
      "num_tokens": 246582914.0,
      "step": 7726
    },
    {
      "entropy": 0.7537737116217613,
      "epoch": 0.7109332321263248,
      "grad_norm": 0.158100888133049,
      "learning_rate": 7.630570122979729e-05,
      "loss": 0.7421,
      "mean_token_accuracy": 0.7870158441364765,
      "num_tokens": 246614618.0,
      "step": 7727
    },
    {
      "entropy": 0.7858059946447611,
      "epoch": 0.711025238497766,
      "grad_norm": 0.1615440547466278,
      "learning_rate": 7.630263440365565e-05,
      "loss": 0.7528,
      "mean_token_accuracy": 0.7798841521143913,
      "num_tokens": 246646005.0,
      "step": 7728
    },
    {
      "entropy": 0.7244427585974336,
      "epoch": 0.7111172448692072,
      "grad_norm": 0.16013580560684204,
      "learning_rate": 7.629956757751404e-05,
      "loss": 0.7089,
      "mean_token_accuracy": 0.7915889695286751,
      "num_tokens": 246677590.0,
      "step": 7729
    },
    {
      "entropy": 0.7836611066013575,
      "epoch": 0.7112092512406484,
      "grad_norm": 0.16171906888484955,
      "learning_rate": 7.629650075137242e-05,
      "loss": 0.7647,
      "mean_token_accuracy": 0.7724419832229614,
      "num_tokens": 246708726.0,
      "step": 7730
    },
    {
      "entropy": 0.6929294150322676,
      "epoch": 0.7113012576120896,
      "grad_norm": 0.1618535816669464,
      "learning_rate": 7.629343392523079e-05,
      "loss": 0.6894,
      "mean_token_accuracy": 0.795664582401514,
      "num_tokens": 246740806.0,
      "step": 7731
    },
    {
      "entropy": 0.8313948586583138,
      "epoch": 0.7113932639835309,
      "grad_norm": 0.16083213686943054,
      "learning_rate": 7.629036709908915e-05,
      "loss": 0.8155,
      "mean_token_accuracy": 0.7594755440950394,
      "num_tokens": 246771948.0,
      "step": 7732
    },
    {
      "entropy": 0.8021026514470577,
      "epoch": 0.7114852703549721,
      "grad_norm": 0.16246753931045532,
      "learning_rate": 7.628730027294752e-05,
      "loss": 0.7861,
      "mean_token_accuracy": 0.769036553800106,
      "num_tokens": 246803763.0,
      "step": 7733
    },
    {
      "entropy": 0.6498741749674082,
      "epoch": 0.7115772767264134,
      "grad_norm": 0.15244333446025848,
      "learning_rate": 7.62842334468059e-05,
      "loss": 0.6361,
      "mean_token_accuracy": 0.8110232427716255,
      "num_tokens": 246835425.0,
      "step": 7734
    },
    {
      "entropy": 0.7554204948246479,
      "epoch": 0.7116692830978545,
      "grad_norm": 0.15529809892177582,
      "learning_rate": 7.628116662066428e-05,
      "loss": 0.7242,
      "mean_token_accuracy": 0.7877806797623634,
      "num_tokens": 246867437.0,
      "step": 7735
    },
    {
      "entropy": 0.7026443220674992,
      "epoch": 0.7117612894692957,
      "grad_norm": 0.1525433510541916,
      "learning_rate": 7.627809979452265e-05,
      "loss": 0.6854,
      "mean_token_accuracy": 0.7966473810374737,
      "num_tokens": 246899597.0,
      "step": 7736
    },
    {
      "entropy": 0.7714231349527836,
      "epoch": 0.711853295840737,
      "grad_norm": 0.16589736938476562,
      "learning_rate": 7.627503296838102e-05,
      "loss": 0.7716,
      "mean_token_accuracy": 0.7741065993905067,
      "num_tokens": 246931848.0,
      "step": 7737
    },
    {
      "entropy": 0.7292466778308153,
      "epoch": 0.7119453022121782,
      "grad_norm": 0.16648989915847778,
      "learning_rate": 7.62719661422394e-05,
      "loss": 0.7113,
      "mean_token_accuracy": 0.7907077930867672,
      "num_tokens": 246963267.0,
      "step": 7738
    },
    {
      "entropy": 0.6226468198001385,
      "epoch": 0.7120373085836195,
      "grad_norm": 0.14857879281044006,
      "learning_rate": 7.626889931609777e-05,
      "loss": 0.598,
      "mean_token_accuracy": 0.8216148912906647,
      "num_tokens": 246995264.0,
      "step": 7739
    },
    {
      "entropy": 0.8858596757054329,
      "epoch": 0.7121293149550606,
      "grad_norm": 0.1684175729751587,
      "learning_rate": 7.626583248995615e-05,
      "loss": 0.9122,
      "mean_token_accuracy": 0.735633235424757,
      "num_tokens": 247027588.0,
      "step": 7740
    },
    {
      "entropy": 0.7619121950119734,
      "epoch": 0.7122213213265018,
      "grad_norm": 0.15950913727283478,
      "learning_rate": 7.626276566381452e-05,
      "loss": 0.7733,
      "mean_token_accuracy": 0.777202919125557,
      "num_tokens": 247059787.0,
      "step": 7741
    },
    {
      "entropy": 0.7894290555268526,
      "epoch": 0.7123133276979431,
      "grad_norm": 0.15158797800540924,
      "learning_rate": 7.62596988376729e-05,
      "loss": 0.7984,
      "mean_token_accuracy": 0.7638357244431973,
      "num_tokens": 247091558.0,
      "step": 7742
    },
    {
      "entropy": 0.6752070309594274,
      "epoch": 0.7124053340693843,
      "grad_norm": 0.15768490731716156,
      "learning_rate": 7.625663201153127e-05,
      "loss": 0.6747,
      "mean_token_accuracy": 0.8007259666919708,
      "num_tokens": 247122803.0,
      "step": 7743
    },
    {
      "entropy": 0.7668081298470497,
      "epoch": 0.7124973404408256,
      "grad_norm": 0.15152016282081604,
      "learning_rate": 7.625356518538964e-05,
      "loss": 0.754,
      "mean_token_accuracy": 0.7755378596484661,
      "num_tokens": 247154857.0,
      "step": 7744
    },
    {
      "entropy": 0.7013351935893297,
      "epoch": 0.7125893468122667,
      "grad_norm": 0.16090771555900574,
      "learning_rate": 7.625049835924802e-05,
      "loss": 0.6562,
      "mean_token_accuracy": 0.8014717139303684,
      "num_tokens": 247185560.0,
      "step": 7745
    },
    {
      "entropy": 0.7212214414030313,
      "epoch": 0.712681353183708,
      "grad_norm": 0.15394622087478638,
      "learning_rate": 7.62474315331064e-05,
      "loss": 0.692,
      "mean_token_accuracy": 0.7949661128222942,
      "num_tokens": 247217092.0,
      "step": 7746
    },
    {
      "entropy": 0.8994690794497728,
      "epoch": 0.7127733595551492,
      "grad_norm": 0.1590815633535385,
      "learning_rate": 7.624436470696477e-05,
      "loss": 0.9049,
      "mean_token_accuracy": 0.7380979619920254,
      "num_tokens": 247248843.0,
      "step": 7747
    },
    {
      "entropy": 0.6765907481312752,
      "epoch": 0.7128653659265904,
      "grad_norm": 0.16464878618717194,
      "learning_rate": 7.624129788082313e-05,
      "loss": 0.6686,
      "mean_token_accuracy": 0.8038506545126438,
      "num_tokens": 247280891.0,
      "step": 7748
    },
    {
      "entropy": 0.7584703918546438,
      "epoch": 0.7129573722980317,
      "grad_norm": 0.16575177013874054,
      "learning_rate": 7.623823105468152e-05,
      "loss": 0.7498,
      "mean_token_accuracy": 0.7826003953814507,
      "num_tokens": 247312224.0,
      "step": 7749
    },
    {
      "entropy": 0.7871558163315058,
      "epoch": 0.7130493786694728,
      "grad_norm": 0.16131983697414398,
      "learning_rate": 7.623516422853988e-05,
      "loss": 0.7778,
      "mean_token_accuracy": 0.7718077935278416,
      "num_tokens": 247343743.0,
      "step": 7750
    },
    {
      "entropy": 0.7554525760933757,
      "epoch": 0.7131413850409141,
      "grad_norm": 0.1549024134874344,
      "learning_rate": 7.623209740239827e-05,
      "loss": 0.7428,
      "mean_token_accuracy": 0.7801962494850159,
      "num_tokens": 247375739.0,
      "step": 7751
    },
    {
      "entropy": 0.7187416050583124,
      "epoch": 0.7132333914123553,
      "grad_norm": 0.15833783149719238,
      "learning_rate": 7.622903057625663e-05,
      "loss": 0.7132,
      "mean_token_accuracy": 0.7891486622393131,
      "num_tokens": 247408069.0,
      "step": 7752
    },
    {
      "entropy": 0.7869937848299742,
      "epoch": 0.7133253977837966,
      "grad_norm": 0.15364477038383484,
      "learning_rate": 7.622596375011501e-05,
      "loss": 0.7882,
      "mean_token_accuracy": 0.7690650261938572,
      "num_tokens": 247439802.0,
      "step": 7753
    },
    {
      "entropy": 0.7797029986977577,
      "epoch": 0.7134174041552378,
      "grad_norm": 0.1565232276916504,
      "learning_rate": 7.622289692397338e-05,
      "loss": 0.7614,
      "mean_token_accuracy": 0.774081964045763,
      "num_tokens": 247471383.0,
      "step": 7754
    },
    {
      "entropy": 0.7594541795551777,
      "epoch": 0.7135094105266789,
      "grad_norm": 0.15079501271247864,
      "learning_rate": 7.621983009783176e-05,
      "loss": 0.7463,
      "mean_token_accuracy": 0.7770072743296623,
      "num_tokens": 247504120.0,
      "step": 7755
    },
    {
      "entropy": 0.7640374545007944,
      "epoch": 0.7136014168981202,
      "grad_norm": 0.16179980337619781,
      "learning_rate": 7.621676327169013e-05,
      "loss": 0.7589,
      "mean_token_accuracy": 0.7798115313053131,
      "num_tokens": 247535725.0,
      "step": 7756
    },
    {
      "entropy": 0.7364048329181969,
      "epoch": 0.7136934232695614,
      "grad_norm": 0.14941072463989258,
      "learning_rate": 7.621369644554851e-05,
      "loss": 0.7288,
      "mean_token_accuracy": 0.7838046513497829,
      "num_tokens": 247567631.0,
      "step": 7757
    },
    {
      "entropy": 0.748789869248867,
      "epoch": 0.7137854296410027,
      "grad_norm": 0.1541941910982132,
      "learning_rate": 7.621062961940688e-05,
      "loss": 0.7296,
      "mean_token_accuracy": 0.7844168841838837,
      "num_tokens": 247599802.0,
      "step": 7758
    },
    {
      "entropy": 0.7460678378120065,
      "epoch": 0.7138774360124439,
      "grad_norm": 0.16640658676624298,
      "learning_rate": 7.620756279326525e-05,
      "loss": 0.7419,
      "mean_token_accuracy": 0.7892195209860802,
      "num_tokens": 247632051.0,
      "step": 7759
    },
    {
      "entropy": 0.6909842304885387,
      "epoch": 0.713969442383885,
      "grad_norm": 0.16003543138504028,
      "learning_rate": 7.620449596712363e-05,
      "loss": 0.6849,
      "mean_token_accuracy": 0.7961903624236584,
      "num_tokens": 247663979.0,
      "step": 7760
    },
    {
      "entropy": 0.7195100644603372,
      "epoch": 0.7140614487553263,
      "grad_norm": 0.14794141054153442,
      "learning_rate": 7.620142914098201e-05,
      "loss": 0.6999,
      "mean_token_accuracy": 0.7919264361262321,
      "num_tokens": 247695022.0,
      "step": 7761
    },
    {
      "entropy": 0.7100788988173008,
      "epoch": 0.7141534551267675,
      "grad_norm": 0.14736983180046082,
      "learning_rate": 7.619836231484038e-05,
      "loss": 0.7089,
      "mean_token_accuracy": 0.7903619445860386,
      "num_tokens": 247727326.0,
      "step": 7762
    },
    {
      "entropy": 0.6707541197538376,
      "epoch": 0.7142454614982088,
      "grad_norm": 0.1457640826702118,
      "learning_rate": 7.619529548869875e-05,
      "loss": 0.6595,
      "mean_token_accuracy": 0.8016452863812447,
      "num_tokens": 247759639.0,
      "step": 7763
    },
    {
      "entropy": 0.6537999892607331,
      "epoch": 0.71433746786965,
      "grad_norm": 0.15103404223918915,
      "learning_rate": 7.619222866255712e-05,
      "loss": 0.642,
      "mean_token_accuracy": 0.8128191195428371,
      "num_tokens": 247791978.0,
      "step": 7764
    },
    {
      "entropy": 0.712818218395114,
      "epoch": 0.7144294742410912,
      "grad_norm": 0.15465407073497772,
      "learning_rate": 7.61891618364155e-05,
      "loss": 0.686,
      "mean_token_accuracy": 0.7943439707159996,
      "num_tokens": 247823371.0,
      "step": 7765
    },
    {
      "entropy": 0.7231134865432978,
      "epoch": 0.7145214806125324,
      "grad_norm": 0.1691575050354004,
      "learning_rate": 7.618609501027388e-05,
      "loss": 0.7352,
      "mean_token_accuracy": 0.7842858396470547,
      "num_tokens": 247855480.0,
      "step": 7766
    },
    {
      "entropy": 0.8269942924380302,
      "epoch": 0.7146134869839736,
      "grad_norm": 0.16038359701633453,
      "learning_rate": 7.618302818413225e-05,
      "loss": 0.8278,
      "mean_token_accuracy": 0.7608065791428089,
      "num_tokens": 247887195.0,
      "step": 7767
    },
    {
      "entropy": 0.815205991268158,
      "epoch": 0.7147054933554149,
      "grad_norm": 0.15800216794013977,
      "learning_rate": 7.617996135799061e-05,
      "loss": 0.7998,
      "mean_token_accuracy": 0.7682800181210041,
      "num_tokens": 247919806.0,
      "step": 7768
    },
    {
      "entropy": 0.6953230779618025,
      "epoch": 0.7147974997268561,
      "grad_norm": 0.16864758729934692,
      "learning_rate": 7.6176894531849e-05,
      "loss": 0.6905,
      "mean_token_accuracy": 0.796279564499855,
      "num_tokens": 247951979.0,
      "step": 7769
    },
    {
      "entropy": 0.8127613961696625,
      "epoch": 0.7148895060982973,
      "grad_norm": 0.16640840470790863,
      "learning_rate": 7.617382770570736e-05,
      "loss": 0.8178,
      "mean_token_accuracy": 0.7608801908791065,
      "num_tokens": 247983975.0,
      "step": 7770
    },
    {
      "entropy": 0.8499493338167667,
      "epoch": 0.7149815124697385,
      "grad_norm": 0.16478778421878815,
      "learning_rate": 7.617076087956574e-05,
      "loss": 0.8565,
      "mean_token_accuracy": 0.7510617151856422,
      "num_tokens": 248015622.0,
      "step": 7771
    },
    {
      "entropy": 0.6901806257665157,
      "epoch": 0.7150735188411798,
      "grad_norm": 0.15633217990398407,
      "learning_rate": 7.616769405342411e-05,
      "loss": 0.6679,
      "mean_token_accuracy": 0.801559254527092,
      "num_tokens": 248047759.0,
      "step": 7772
    },
    {
      "entropy": 0.8176596965640783,
      "epoch": 0.715165525212621,
      "grad_norm": 0.16402335464954376,
      "learning_rate": 7.61646272272825e-05,
      "loss": 0.8107,
      "mean_token_accuracy": 0.7654799334704876,
      "num_tokens": 248079147.0,
      "step": 7773
    },
    {
      "entropy": 0.8058506604284048,
      "epoch": 0.7152575315840622,
      "grad_norm": 0.15339523553848267,
      "learning_rate": 7.616156040114086e-05,
      "loss": 0.7955,
      "mean_token_accuracy": 0.7651201635599136,
      "num_tokens": 248110555.0,
      "step": 7774
    },
    {
      "entropy": 0.7172575034201145,
      "epoch": 0.7153495379555034,
      "grad_norm": 0.15279529988765717,
      "learning_rate": 7.615849357499923e-05,
      "loss": 0.6832,
      "mean_token_accuracy": 0.7984414994716644,
      "num_tokens": 248142600.0,
      "step": 7775
    },
    {
      "entropy": 0.6991952322423458,
      "epoch": 0.7154415443269446,
      "grad_norm": 0.14327575266361237,
      "learning_rate": 7.615542674885761e-05,
      "loss": 0.6792,
      "mean_token_accuracy": 0.7976426407694817,
      "num_tokens": 248174769.0,
      "step": 7776
    },
    {
      "entropy": 0.7888136841356754,
      "epoch": 0.7155335506983859,
      "grad_norm": 0.15282917022705078,
      "learning_rate": 7.615235992271599e-05,
      "loss": 0.7767,
      "mean_token_accuracy": 0.773173376917839,
      "num_tokens": 248206929.0,
      "step": 7777
    },
    {
      "entropy": 0.6634180210530758,
      "epoch": 0.7156255570698271,
      "grad_norm": 0.15584835410118103,
      "learning_rate": 7.614929309657436e-05,
      "loss": 0.6375,
      "mean_token_accuracy": 0.8113144971430302,
      "num_tokens": 248238802.0,
      "step": 7778
    },
    {
      "entropy": 0.7935320623219013,
      "epoch": 0.7157175634412684,
      "grad_norm": 0.15831786394119263,
      "learning_rate": 7.614622627043273e-05,
      "loss": 0.7763,
      "mean_token_accuracy": 0.7750990763306618,
      "num_tokens": 248270935.0,
      "step": 7779
    },
    {
      "entropy": 0.6737814890220761,
      "epoch": 0.7158095698127095,
      "grad_norm": 0.15074265003204346,
      "learning_rate": 7.614315944429111e-05,
      "loss": 0.6619,
      "mean_token_accuracy": 0.802977092564106,
      "num_tokens": 248303564.0,
      "step": 7780
    },
    {
      "entropy": 0.7043156158179045,
      "epoch": 0.7159015761841507,
      "grad_norm": 0.15215253829956055,
      "learning_rate": 7.614009261814948e-05,
      "loss": 0.6875,
      "mean_token_accuracy": 0.7944777868688107,
      "num_tokens": 248335336.0,
      "step": 7781
    },
    {
      "entropy": 0.7677399143576622,
      "epoch": 0.715993582555592,
      "grad_norm": 0.1607779562473297,
      "learning_rate": 7.613702579200786e-05,
      "loss": 0.7594,
      "mean_token_accuracy": 0.7781425938010216,
      "num_tokens": 248366698.0,
      "step": 7782
    },
    {
      "entropy": 0.705123282968998,
      "epoch": 0.7160855889270332,
      "grad_norm": 0.15246890485286713,
      "learning_rate": 7.613395896586623e-05,
      "loss": 0.6912,
      "mean_token_accuracy": 0.7936547324061394,
      "num_tokens": 248398616.0,
      "step": 7783
    },
    {
      "entropy": 0.8230610694736242,
      "epoch": 0.7161775952984745,
      "grad_norm": 0.16714119911193848,
      "learning_rate": 7.613089213972461e-05,
      "loss": 0.8224,
      "mean_token_accuracy": 0.7576384618878365,
      "num_tokens": 248430186.0,
      "step": 7784
    },
    {
      "entropy": 0.7469585556536913,
      "epoch": 0.7162696016699156,
      "grad_norm": 0.1549912691116333,
      "learning_rate": 7.612782531358298e-05,
      "loss": 0.7524,
      "mean_token_accuracy": 0.7812136970460415,
      "num_tokens": 248462328.0,
      "step": 7785
    },
    {
      "entropy": 0.7058742865920067,
      "epoch": 0.7163616080413568,
      "grad_norm": 0.15642045438289642,
      "learning_rate": 7.612475848744134e-05,
      "loss": 0.7016,
      "mean_token_accuracy": 0.7925045490264893,
      "num_tokens": 248494796.0,
      "step": 7786
    },
    {
      "entropy": 0.7850315570831299,
      "epoch": 0.7164536144127981,
      "grad_norm": 0.16573500633239746,
      "learning_rate": 7.612169166129973e-05,
      "loss": 0.7799,
      "mean_token_accuracy": 0.774956252425909,
      "num_tokens": 248526653.0,
      "step": 7787
    },
    {
      "entropy": 0.771076025441289,
      "epoch": 0.7165456207842393,
      "grad_norm": 0.14662708342075348,
      "learning_rate": 7.61186248351581e-05,
      "loss": 0.7615,
      "mean_token_accuracy": 0.7755336165428162,
      "num_tokens": 248559268.0,
      "step": 7788
    },
    {
      "entropy": 0.8243791311979294,
      "epoch": 0.7166376271556806,
      "grad_norm": 0.16546407341957092,
      "learning_rate": 7.611555800901647e-05,
      "loss": 0.8393,
      "mean_token_accuracy": 0.753466010093689,
      "num_tokens": 248590916.0,
      "step": 7789
    },
    {
      "entropy": 0.7695344649255276,
      "epoch": 0.7167296335271217,
      "grad_norm": 0.15810668468475342,
      "learning_rate": 7.611249118287484e-05,
      "loss": 0.7597,
      "mean_token_accuracy": 0.7773334197700024,
      "num_tokens": 248623487.0,
      "step": 7790
    },
    {
      "entropy": 0.7846469786018133,
      "epoch": 0.716821639898563,
      "grad_norm": 0.15754403173923492,
      "learning_rate": 7.610942435673321e-05,
      "loss": 0.791,
      "mean_token_accuracy": 0.7666886858642101,
      "num_tokens": 248655731.0,
      "step": 7791
    },
    {
      "entropy": 0.7437743861228228,
      "epoch": 0.7169136462700042,
      "grad_norm": 0.15561647713184357,
      "learning_rate": 7.61063575305916e-05,
      "loss": 0.7525,
      "mean_token_accuracy": 0.7852563671767712,
      "num_tokens": 248687635.0,
      "step": 7792
    },
    {
      "entropy": 0.7466721935197711,
      "epoch": 0.7170056526414454,
      "grad_norm": 0.16318325698375702,
      "learning_rate": 7.610329070444997e-05,
      "loss": 0.7206,
      "mean_token_accuracy": 0.7855875976383686,
      "num_tokens": 248718614.0,
      "step": 7793
    },
    {
      "entropy": 0.7026221007108688,
      "epoch": 0.7170976590128867,
      "grad_norm": 0.14973880350589752,
      "learning_rate": 7.610022387830834e-05,
      "loss": 0.6929,
      "mean_token_accuracy": 0.7893313281238079,
      "num_tokens": 248750974.0,
      "step": 7794
    },
    {
      "entropy": 0.8409431278705597,
      "epoch": 0.7171896653843278,
      "grad_norm": 0.1647198647260666,
      "learning_rate": 7.609715705216671e-05,
      "loss": 0.8145,
      "mean_token_accuracy": 0.7611793987452984,
      "num_tokens": 248782565.0,
      "step": 7795
    },
    {
      "entropy": 0.6431141654029489,
      "epoch": 0.7172816717557691,
      "grad_norm": 0.15813638269901276,
      "learning_rate": 7.609409022602509e-05,
      "loss": 0.6277,
      "mean_token_accuracy": 0.8160623461008072,
      "num_tokens": 248815258.0,
      "step": 7796
    },
    {
      "entropy": 0.6888696309179068,
      "epoch": 0.7173736781272103,
      "grad_norm": 0.14767909049987793,
      "learning_rate": 7.609102339988347e-05,
      "loss": 0.6635,
      "mean_token_accuracy": 0.8016304336488247,
      "num_tokens": 248848026.0,
      "step": 7797
    },
    {
      "entropy": 0.7281116340309381,
      "epoch": 0.7174656844986516,
      "grad_norm": 0.1616005152463913,
      "learning_rate": 7.608795657374184e-05,
      "loss": 0.7166,
      "mean_token_accuracy": 0.7896447516977787,
      "num_tokens": 248880096.0,
      "step": 7798
    },
    {
      "entropy": 0.8291907012462616,
      "epoch": 0.7175576908700928,
      "grad_norm": 0.15777286887168884,
      "learning_rate": 7.608488974760021e-05,
      "loss": 0.821,
      "mean_token_accuracy": 0.7602130547165871,
      "num_tokens": 248912473.0,
      "step": 7799
    },
    {
      "entropy": 0.6723835654556751,
      "epoch": 0.7176496972415339,
      "grad_norm": 0.15408095717430115,
      "learning_rate": 7.608182292145859e-05,
      "loss": 0.6399,
      "mean_token_accuracy": 0.8099945895373821,
      "num_tokens": 248944157.0,
      "step": 7800
    },
    {
      "entropy": 0.7166076507419348,
      "epoch": 0.7177417036129752,
      "grad_norm": 0.15199844539165497,
      "learning_rate": 7.607875609531696e-05,
      "loss": 0.7078,
      "mean_token_accuracy": 0.7872540690004826,
      "num_tokens": 248975823.0,
      "step": 7801
    },
    {
      "entropy": 0.8470213040709496,
      "epoch": 0.7178337099844164,
      "grad_norm": 0.15802063047885895,
      "learning_rate": 7.607568926917534e-05,
      "loss": 0.8437,
      "mean_token_accuracy": 0.7559981942176819,
      "num_tokens": 249008023.0,
      "step": 7802
    },
    {
      "entropy": 0.6305005233734846,
      "epoch": 0.7179257163558577,
      "grad_norm": 0.15357202291488647,
      "learning_rate": 7.60726224430337e-05,
      "loss": 0.6068,
      "mean_token_accuracy": 0.8185158036649227,
      "num_tokens": 249039193.0,
      "step": 7803
    },
    {
      "entropy": 0.7103920048102736,
      "epoch": 0.7180177227272989,
      "grad_norm": 0.15432454645633698,
      "learning_rate": 7.606955561689209e-05,
      "loss": 0.6846,
      "mean_token_accuracy": 0.799586396664381,
      "num_tokens": 249070648.0,
      "step": 7804
    },
    {
      "entropy": 0.754861006513238,
      "epoch": 0.71810972909874,
      "grad_norm": 0.16732004284858704,
      "learning_rate": 7.606648879075046e-05,
      "loss": 0.774,
      "mean_token_accuracy": 0.7730255275964737,
      "num_tokens": 249102487.0,
      "step": 7805
    },
    {
      "entropy": 0.7751673795282841,
      "epoch": 0.7182017354701813,
      "grad_norm": 0.1573687195777893,
      "learning_rate": 7.606342196460882e-05,
      "loss": 0.7558,
      "mean_token_accuracy": 0.773172203451395,
      "num_tokens": 249133763.0,
      "step": 7806
    },
    {
      "entropy": 0.7096458189189434,
      "epoch": 0.7182937418416225,
      "grad_norm": 0.14964421093463898,
      "learning_rate": 7.60603551384672e-05,
      "loss": 0.7115,
      "mean_token_accuracy": 0.7864885702729225,
      "num_tokens": 249165800.0,
      "step": 7807
    },
    {
      "entropy": 0.7759762424975634,
      "epoch": 0.7183857482130638,
      "grad_norm": 0.15842406451702118,
      "learning_rate": 7.605728831232559e-05,
      "loss": 0.7591,
      "mean_token_accuracy": 0.774175763130188,
      "num_tokens": 249197494.0,
      "step": 7808
    },
    {
      "entropy": 0.7814515195786953,
      "epoch": 0.718477754584505,
      "grad_norm": 0.16591733694076538,
      "learning_rate": 7.605422148618395e-05,
      "loss": 0.7872,
      "mean_token_accuracy": 0.7737616375088692,
      "num_tokens": 249229582.0,
      "step": 7809
    },
    {
      "entropy": 0.7549196518957615,
      "epoch": 0.7185697609559462,
      "grad_norm": 0.15124650299549103,
      "learning_rate": 7.605115466004232e-05,
      "loss": 0.744,
      "mean_token_accuracy": 0.7792503833770752,
      "num_tokens": 249261513.0,
      "step": 7810
    },
    {
      "entropy": 0.7750103138387203,
      "epoch": 0.7186617673273874,
      "grad_norm": 0.15700888633728027,
      "learning_rate": 7.60480878339007e-05,
      "loss": 0.7689,
      "mean_token_accuracy": 0.7748158536851406,
      "num_tokens": 249293077.0,
      "step": 7811
    },
    {
      "entropy": 0.7348957248032093,
      "epoch": 0.7187537736988286,
      "grad_norm": 0.1551908701658249,
      "learning_rate": 7.604502100775907e-05,
      "loss": 0.7151,
      "mean_token_accuracy": 0.7870091758668423,
      "num_tokens": 249323753.0,
      "step": 7812
    },
    {
      "entropy": 0.7617241647094488,
      "epoch": 0.7188457800702699,
      "grad_norm": 0.14938214421272278,
      "learning_rate": 7.604195418161745e-05,
      "loss": 0.7545,
      "mean_token_accuracy": 0.774508498609066,
      "num_tokens": 249356126.0,
      "step": 7813
    },
    {
      "entropy": 0.6781882327049971,
      "epoch": 0.7189377864417111,
      "grad_norm": 0.15181703865528107,
      "learning_rate": 7.603888735547582e-05,
      "loss": 0.6446,
      "mean_token_accuracy": 0.8086247928440571,
      "num_tokens": 249388749.0,
      "step": 7814
    },
    {
      "entropy": 0.5783234499394894,
      "epoch": 0.7190297928131523,
      "grad_norm": 0.1531478464603424,
      "learning_rate": 7.60358205293342e-05,
      "loss": 0.5554,
      "mean_token_accuracy": 0.832526508718729,
      "num_tokens": 249420654.0,
      "step": 7815
    },
    {
      "entropy": 0.5863795597106218,
      "epoch": 0.7191217991845935,
      "grad_norm": 0.1582360416650772,
      "learning_rate": 7.603275370319257e-05,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8232474401593208,
      "num_tokens": 249453422.0,
      "step": 7816
    },
    {
      "entropy": 0.7413662322796881,
      "epoch": 0.7192138055560348,
      "grad_norm": 0.1634497046470642,
      "learning_rate": 7.602968687705094e-05,
      "loss": 0.729,
      "mean_token_accuracy": 0.7823475748300552,
      "num_tokens": 249484970.0,
      "step": 7817
    },
    {
      "entropy": 0.8403741233050823,
      "epoch": 0.719305811927476,
      "grad_norm": 0.1561255007982254,
      "learning_rate": 7.602662005090932e-05,
      "loss": 0.8253,
      "mean_token_accuracy": 0.7612036541104317,
      "num_tokens": 249517508.0,
      "step": 7818
    },
    {
      "entropy": 0.7198612354695797,
      "epoch": 0.7193978182989172,
      "grad_norm": 0.16850140690803528,
      "learning_rate": 7.60235532247677e-05,
      "loss": 0.7115,
      "mean_token_accuracy": 0.7921142354607582,
      "num_tokens": 249549375.0,
      "step": 7819
    },
    {
      "entropy": 0.8259382862597704,
      "epoch": 0.7194898246703584,
      "grad_norm": 0.16326290369033813,
      "learning_rate": 7.602048639862607e-05,
      "loss": 0.8154,
      "mean_token_accuracy": 0.7632639855146408,
      "num_tokens": 249580809.0,
      "step": 7820
    },
    {
      "entropy": 0.7156979329884052,
      "epoch": 0.7195818310417996,
      "grad_norm": 0.15672911703586578,
      "learning_rate": 7.601741957248444e-05,
      "loss": 0.6825,
      "mean_token_accuracy": 0.7999634258449078,
      "num_tokens": 249612251.0,
      "step": 7821
    },
    {
      "entropy": 0.7891224585473537,
      "epoch": 0.7196738374132409,
      "grad_norm": 0.16958454251289368,
      "learning_rate": 7.60143527463428e-05,
      "loss": 0.7748,
      "mean_token_accuracy": 0.7772296853363514,
      "num_tokens": 249644581.0,
      "step": 7822
    },
    {
      "entropy": 0.71135398093611,
      "epoch": 0.7197658437846821,
      "grad_norm": 0.15872737765312195,
      "learning_rate": 7.601128592020119e-05,
      "loss": 0.7023,
      "mean_token_accuracy": 0.7896397411823273,
      "num_tokens": 249675811.0,
      "step": 7823
    },
    {
      "entropy": 0.6751199588179588,
      "epoch": 0.7198578501561234,
      "grad_norm": 0.1522376388311386,
      "learning_rate": 7.600821909405957e-05,
      "loss": 0.654,
      "mean_token_accuracy": 0.8025815896689892,
      "num_tokens": 249707973.0,
      "step": 7824
    },
    {
      "entropy": 0.7892940994352102,
      "epoch": 0.7199498565275645,
      "grad_norm": 0.15517310798168182,
      "learning_rate": 7.600515226791793e-05,
      "loss": 0.7881,
      "mean_token_accuracy": 0.7612491175532341,
      "num_tokens": 249739698.0,
      "step": 7825
    },
    {
      "entropy": 0.7357265334576368,
      "epoch": 0.7200418628990057,
      "grad_norm": 0.15088872611522675,
      "learning_rate": 7.60020854417763e-05,
      "loss": 0.7266,
      "mean_token_accuracy": 0.7840061336755753,
      "num_tokens": 249771779.0,
      "step": 7826
    },
    {
      "entropy": 0.7462134966626763,
      "epoch": 0.720133869270447,
      "grad_norm": 0.15191501379013062,
      "learning_rate": 7.599901861563468e-05,
      "loss": 0.7493,
      "mean_token_accuracy": 0.7821086719632149,
      "num_tokens": 249803293.0,
      "step": 7827
    },
    {
      "entropy": 0.7180109173059464,
      "epoch": 0.7202258756418882,
      "grad_norm": 0.15315556526184082,
      "learning_rate": 7.599595178949305e-05,
      "loss": 0.7161,
      "mean_token_accuracy": 0.7866601683199406,
      "num_tokens": 249835570.0,
      "step": 7828
    },
    {
      "entropy": 0.7424497622996569,
      "epoch": 0.7203178820133295,
      "grad_norm": 0.16053053736686707,
      "learning_rate": 7.599288496335143e-05,
      "loss": 0.7342,
      "mean_token_accuracy": 0.7813802361488342,
      "num_tokens": 249867875.0,
      "step": 7829
    },
    {
      "entropy": 0.7689687088131905,
      "epoch": 0.7204098883847706,
      "grad_norm": 0.15320421755313873,
      "learning_rate": 7.59898181372098e-05,
      "loss": 0.7541,
      "mean_token_accuracy": 0.7764865010976791,
      "num_tokens": 249899156.0,
      "step": 7830
    },
    {
      "entropy": 0.676791787147522,
      "epoch": 0.7205018947562118,
      "grad_norm": 0.1555180847644806,
      "learning_rate": 7.598675131106818e-05,
      "loss": 0.661,
      "mean_token_accuracy": 0.7996770702302456,
      "num_tokens": 249931581.0,
      "step": 7831
    },
    {
      "entropy": 0.9106304720044136,
      "epoch": 0.7205939011276531,
      "grad_norm": 0.1652318835258484,
      "learning_rate": 7.598368448492655e-05,
      "loss": 0.9204,
      "mean_token_accuracy": 0.7354933805763721,
      "num_tokens": 249962946.0,
      "step": 7832
    },
    {
      "entropy": 0.7655843961983919,
      "epoch": 0.7206859074990943,
      "grad_norm": 0.1514715552330017,
      "learning_rate": 7.598061765878492e-05,
      "loss": 0.7519,
      "mean_token_accuracy": 0.7783109359443188,
      "num_tokens": 249995187.0,
      "step": 7833
    },
    {
      "entropy": 0.6588938608765602,
      "epoch": 0.7207779138705356,
      "grad_norm": 0.1464916169643402,
      "learning_rate": 7.59775508326433e-05,
      "loss": 0.6451,
      "mean_token_accuracy": 0.8086031787097454,
      "num_tokens": 250027933.0,
      "step": 7834
    },
    {
      "entropy": 0.7518705800175667,
      "epoch": 0.7208699202419767,
      "grad_norm": 0.14967560768127441,
      "learning_rate": 7.597448400650168e-05,
      "loss": 0.7414,
      "mean_token_accuracy": 0.785016130656004,
      "num_tokens": 250060636.0,
      "step": 7835
    },
    {
      "entropy": 0.8548146318644285,
      "epoch": 0.720961926613418,
      "grad_norm": 0.15424463152885437,
      "learning_rate": 7.597141718036005e-05,
      "loss": 0.8429,
      "mean_token_accuracy": 0.7532053254544735,
      "num_tokens": 250092266.0,
      "step": 7836
    },
    {
      "entropy": 0.8493393957614899,
      "epoch": 0.7210539329848592,
      "grad_norm": 0.1592412292957306,
      "learning_rate": 7.596835035421842e-05,
      "loss": 0.8622,
      "mean_token_accuracy": 0.7503750175237656,
      "num_tokens": 250124204.0,
      "step": 7837
    },
    {
      "entropy": 0.8120987080037594,
      "epoch": 0.7211459393563004,
      "grad_norm": 0.1668112426996231,
      "learning_rate": 7.596528352807678e-05,
      "loss": 0.8061,
      "mean_token_accuracy": 0.7684736549854279,
      "num_tokens": 250156073.0,
      "step": 7838
    },
    {
      "entropy": 0.8278126604855061,
      "epoch": 0.7212379457277417,
      "grad_norm": 0.15897923707962036,
      "learning_rate": 7.596221670193518e-05,
      "loss": 0.8087,
      "mean_token_accuracy": 0.7571402937173843,
      "num_tokens": 250187655.0,
      "step": 7839
    },
    {
      "entropy": 0.6908633168786764,
      "epoch": 0.7213299520991828,
      "grad_norm": 0.14836974442005157,
      "learning_rate": 7.595914987579355e-05,
      "loss": 0.6701,
      "mean_token_accuracy": 0.8013248704373837,
      "num_tokens": 250219860.0,
      "step": 7840
    },
    {
      "entropy": 0.7341315988451242,
      "epoch": 0.7214219584706241,
      "grad_norm": 0.15764974057674408,
      "learning_rate": 7.595608304965192e-05,
      "loss": 0.7349,
      "mean_token_accuracy": 0.7859954312443733,
      "num_tokens": 250251976.0,
      "step": 7841
    },
    {
      "entropy": 0.7187400236725807,
      "epoch": 0.7215139648420653,
      "grad_norm": 0.15172894299030304,
      "learning_rate": 7.59530162235103e-05,
      "loss": 0.6881,
      "mean_token_accuracy": 0.7949937097728252,
      "num_tokens": 250283533.0,
      "step": 7842
    },
    {
      "entropy": 0.6541743651032448,
      "epoch": 0.7216059712135066,
      "grad_norm": 0.15242579579353333,
      "learning_rate": 7.594994939736866e-05,
      "loss": 0.6465,
      "mean_token_accuracy": 0.8055494651198387,
      "num_tokens": 250315668.0,
      "step": 7843
    },
    {
      "entropy": 0.7217982020229101,
      "epoch": 0.7216979775849478,
      "grad_norm": 0.15646827220916748,
      "learning_rate": 7.594688257122705e-05,
      "loss": 0.7273,
      "mean_token_accuracy": 0.7858561277389526,
      "num_tokens": 250347700.0,
      "step": 7844
    },
    {
      "entropy": 0.5940914358943701,
      "epoch": 0.7217899839563889,
      "grad_norm": 0.16489790380001068,
      "learning_rate": 7.594381574508541e-05,
      "loss": 0.5842,
      "mean_token_accuracy": 0.8202256858348846,
      "num_tokens": 250379959.0,
      "step": 7845
    },
    {
      "entropy": 0.6855104379355907,
      "epoch": 0.7218819903278302,
      "grad_norm": 0.14689376950263977,
      "learning_rate": 7.59407489189438e-05,
      "loss": 0.682,
      "mean_token_accuracy": 0.7930132038891315,
      "num_tokens": 250411770.0,
      "step": 7846
    },
    {
      "entropy": 0.6610247064381838,
      "epoch": 0.7219739966992714,
      "grad_norm": 0.1623591184616089,
      "learning_rate": 7.593768209280216e-05,
      "loss": 0.6559,
      "mean_token_accuracy": 0.8056875579059124,
      "num_tokens": 250443590.0,
      "step": 7847
    },
    {
      "entropy": 0.8891603630036116,
      "epoch": 0.7220660030707127,
      "grad_norm": 0.1627303659915924,
      "learning_rate": 7.593461526666053e-05,
      "loss": 0.8956,
      "mean_token_accuracy": 0.7395302690565586,
      "num_tokens": 250475767.0,
      "step": 7848
    },
    {
      "entropy": 0.6086974814534187,
      "epoch": 0.7221580094421539,
      "grad_norm": 0.16022752225399017,
      "learning_rate": 7.593154844051891e-05,
      "loss": 0.6049,
      "mean_token_accuracy": 0.8214403763413429,
      "num_tokens": 250507707.0,
      "step": 7849
    },
    {
      "entropy": 0.7726941145956516,
      "epoch": 0.722250015813595,
      "grad_norm": 0.16883137822151184,
      "learning_rate": 7.59284816143773e-05,
      "loss": 0.7483,
      "mean_token_accuracy": 0.7789816595613956,
      "num_tokens": 250539620.0,
      "step": 7850
    },
    {
      "entropy": 0.8549030013382435,
      "epoch": 0.7223420221850363,
      "grad_norm": 0.16999629139900208,
      "learning_rate": 7.592541478823566e-05,
      "loss": 0.8415,
      "mean_token_accuracy": 0.7561250850558281,
      "num_tokens": 250570829.0,
      "step": 7851
    },
    {
      "entropy": 0.7578040938824415,
      "epoch": 0.7224340285564775,
      "grad_norm": 0.1520872712135315,
      "learning_rate": 7.592234796209403e-05,
      "loss": 0.7381,
      "mean_token_accuracy": 0.7775016501545906,
      "num_tokens": 250602051.0,
      "step": 7852
    },
    {
      "entropy": 0.7262165462598205,
      "epoch": 0.7225260349279188,
      "grad_norm": 0.15373148024082184,
      "learning_rate": 7.59192811359524e-05,
      "loss": 0.7158,
      "mean_token_accuracy": 0.7854645065963268,
      "num_tokens": 250634576.0,
      "step": 7853
    },
    {
      "entropy": 0.8073814585804939,
      "epoch": 0.72261804129936,
      "grad_norm": 0.1515234112739563,
      "learning_rate": 7.591621430981078e-05,
      "loss": 0.7861,
      "mean_token_accuracy": 0.7672849595546722,
      "num_tokens": 250665847.0,
      "step": 7854
    },
    {
      "entropy": 0.803986269980669,
      "epoch": 0.7227100476708012,
      "grad_norm": 0.15217888355255127,
      "learning_rate": 7.591314748366916e-05,
      "loss": 0.7927,
      "mean_token_accuracy": 0.7678348459303379,
      "num_tokens": 250697817.0,
      "step": 7855
    },
    {
      "entropy": 0.6912982165813446,
      "epoch": 0.7228020540422424,
      "grad_norm": 0.15832598507404327,
      "learning_rate": 7.591008065752753e-05,
      "loss": 0.6998,
      "mean_token_accuracy": 0.7977241687476635,
      "num_tokens": 250728504.0,
      "step": 7856
    },
    {
      "entropy": 0.7579865101724863,
      "epoch": 0.7228940604136836,
      "grad_norm": 0.15762892365455627,
      "learning_rate": 7.59070138313859e-05,
      "loss": 0.7428,
      "mean_token_accuracy": 0.7789614386856556,
      "num_tokens": 250760547.0,
      "step": 7857
    },
    {
      "entropy": 0.7010397501289845,
      "epoch": 0.7229860667851249,
      "grad_norm": 0.15368255972862244,
      "learning_rate": 7.590394700524428e-05,
      "loss": 0.6792,
      "mean_token_accuracy": 0.8007333241403103,
      "num_tokens": 250792175.0,
      "step": 7858
    },
    {
      "entropy": 0.7337883599102497,
      "epoch": 0.7230780731565661,
      "grad_norm": 0.15502247214317322,
      "learning_rate": 7.590088017910265e-05,
      "loss": 0.7242,
      "mean_token_accuracy": 0.788350060582161,
      "num_tokens": 250824080.0,
      "step": 7859
    },
    {
      "entropy": 0.7374857682734728,
      "epoch": 0.7231700795280073,
      "grad_norm": 0.15713320672512054,
      "learning_rate": 7.589781335296103e-05,
      "loss": 0.7259,
      "mean_token_accuracy": 0.7895756177604198,
      "num_tokens": 250856655.0,
      "step": 7860
    },
    {
      "entropy": 0.838067440316081,
      "epoch": 0.7232620858994485,
      "grad_norm": 0.15965953469276428,
      "learning_rate": 7.58947465268194e-05,
      "loss": 0.813,
      "mean_token_accuracy": 0.761335451155901,
      "num_tokens": 250887626.0,
      "step": 7861
    },
    {
      "entropy": 0.6864751502871513,
      "epoch": 0.7233540922708898,
      "grad_norm": 0.15342451632022858,
      "learning_rate": 7.589167970067778e-05,
      "loss": 0.6629,
      "mean_token_accuracy": 0.8021831251680851,
      "num_tokens": 250919335.0,
      "step": 7862
    },
    {
      "entropy": 0.7580422703176737,
      "epoch": 0.723446098642331,
      "grad_norm": 0.1563965231180191,
      "learning_rate": 7.588861287453614e-05,
      "loss": 0.729,
      "mean_token_accuracy": 0.7789822369813919,
      "num_tokens": 250950165.0,
      "step": 7863
    },
    {
      "entropy": 0.7174432445317507,
      "epoch": 0.7235381050137722,
      "grad_norm": 0.15998795628547668,
      "learning_rate": 7.588554604839451e-05,
      "loss": 0.7061,
      "mean_token_accuracy": 0.7942802496254444,
      "num_tokens": 250981510.0,
      "step": 7864
    },
    {
      "entropy": 0.6402461584657431,
      "epoch": 0.7236301113852134,
      "grad_norm": 0.14963191747665405,
      "learning_rate": 7.58824792222529e-05,
      "loss": 0.6345,
      "mean_token_accuracy": 0.8140170685946941,
      "num_tokens": 251013730.0,
      "step": 7865
    },
    {
      "entropy": 0.6158556919544935,
      "epoch": 0.7237221177566546,
      "grad_norm": 0.15309172868728638,
      "learning_rate": 7.587941239611127e-05,
      "loss": 0.6167,
      "mean_token_accuracy": 0.8165701478719711,
      "num_tokens": 251046147.0,
      "step": 7866
    },
    {
      "entropy": 0.7951819393783808,
      "epoch": 0.7238141241280959,
      "grad_norm": 0.16716481745243073,
      "learning_rate": 7.587634556996964e-05,
      "loss": 0.8163,
      "mean_token_accuracy": 0.7628971412777901,
      "num_tokens": 251078337.0,
      "step": 7867
    },
    {
      "entropy": 0.8026394797489047,
      "epoch": 0.7239061304995371,
      "grad_norm": 0.15928445756435394,
      "learning_rate": 7.587327874382801e-05,
      "loss": 0.7945,
      "mean_token_accuracy": 0.7687054090201855,
      "num_tokens": 251109907.0,
      "step": 7868
    },
    {
      "entropy": 0.820055890828371,
      "epoch": 0.7239981368709784,
      "grad_norm": 0.16832119226455688,
      "learning_rate": 7.587021191768638e-05,
      "loss": 0.8176,
      "mean_token_accuracy": 0.7602797076106071,
      "num_tokens": 251141698.0,
      "step": 7869
    },
    {
      "entropy": 0.7625326849520206,
      "epoch": 0.7240901432424195,
      "grad_norm": 0.15650774538516998,
      "learning_rate": 7.586714509154476e-05,
      "loss": 0.7454,
      "mean_token_accuracy": 0.7801604010164738,
      "num_tokens": 251172818.0,
      "step": 7870
    },
    {
      "entropy": 0.827116833999753,
      "epoch": 0.7241821496138607,
      "grad_norm": 0.16842181980609894,
      "learning_rate": 7.586407826540314e-05,
      "loss": 0.8117,
      "mean_token_accuracy": 0.7641343474388123,
      "num_tokens": 251204290.0,
      "step": 7871
    },
    {
      "entropy": 0.5842435583472252,
      "epoch": 0.724274155985302,
      "grad_norm": 0.15183652937412262,
      "learning_rate": 7.586101143926151e-05,
      "loss": 0.5777,
      "mean_token_accuracy": 0.8272777236998081,
      "num_tokens": 251237058.0,
      "step": 7872
    },
    {
      "entropy": 0.7041285885497928,
      "epoch": 0.7243661623567432,
      "grad_norm": 0.15662062168121338,
      "learning_rate": 7.585794461311989e-05,
      "loss": 0.6846,
      "mean_token_accuracy": 0.8001513406634331,
      "num_tokens": 251269151.0,
      "step": 7873
    },
    {
      "entropy": 0.8016780875623226,
      "epoch": 0.7244581687281845,
      "grad_norm": 0.16056226193904877,
      "learning_rate": 7.585487778697826e-05,
      "loss": 0.8056,
      "mean_token_accuracy": 0.7668046727776527,
      "num_tokens": 251301322.0,
      "step": 7874
    },
    {
      "entropy": 0.7764002662152052,
      "epoch": 0.7245501750996256,
      "grad_norm": 0.1575009971857071,
      "learning_rate": 7.585181096083664e-05,
      "loss": 0.7846,
      "mean_token_accuracy": 0.7703137472271919,
      "num_tokens": 251333122.0,
      "step": 7875
    },
    {
      "entropy": 0.8169222567230463,
      "epoch": 0.7246421814710668,
      "grad_norm": 0.1504870057106018,
      "learning_rate": 7.584874413469501e-05,
      "loss": 0.8036,
      "mean_token_accuracy": 0.7608672529459,
      "num_tokens": 251365644.0,
      "step": 7876
    },
    {
      "entropy": 0.766961308196187,
      "epoch": 0.7247341878425081,
      "grad_norm": 0.1525883674621582,
      "learning_rate": 7.584567730855339e-05,
      "loss": 0.7387,
      "mean_token_accuracy": 0.7767328508198261,
      "num_tokens": 251398195.0,
      "step": 7877
    },
    {
      "entropy": 0.7656928356736898,
      "epoch": 0.7248261942139493,
      "grad_norm": 0.1616189330816269,
      "learning_rate": 7.584261048241176e-05,
      "loss": 0.7571,
      "mean_token_accuracy": 0.775357685983181,
      "num_tokens": 251430431.0,
      "step": 7878
    },
    {
      "entropy": 0.775195773690939,
      "epoch": 0.7249182005853906,
      "grad_norm": 0.17015773057937622,
      "learning_rate": 7.583954365627013e-05,
      "loss": 0.7798,
      "mean_token_accuracy": 0.7767594009637833,
      "num_tokens": 251461868.0,
      "step": 7879
    },
    {
      "entropy": 0.683290982618928,
      "epoch": 0.7250102069568317,
      "grad_norm": 0.15220312774181366,
      "learning_rate": 7.58364768301285e-05,
      "loss": 0.6757,
      "mean_token_accuracy": 0.7986709214746952,
      "num_tokens": 251493397.0,
      "step": 7880
    },
    {
      "entropy": 0.7197461575269699,
      "epoch": 0.725102213328273,
      "grad_norm": 0.15147072076797485,
      "learning_rate": 7.583341000398689e-05,
      "loss": 0.7035,
      "mean_token_accuracy": 0.7867350168526173,
      "num_tokens": 251525812.0,
      "step": 7881
    },
    {
      "entropy": 0.6993083907291293,
      "epoch": 0.7251942196997142,
      "grad_norm": 0.15029184520244598,
      "learning_rate": 7.583034317784526e-05,
      "loss": 0.6782,
      "mean_token_accuracy": 0.8030892200767994,
      "num_tokens": 251557553.0,
      "step": 7882
    },
    {
      "entropy": 0.7166417483240366,
      "epoch": 0.7252862260711554,
      "grad_norm": 0.15210284292697906,
      "learning_rate": 7.582727635170362e-05,
      "loss": 0.6908,
      "mean_token_accuracy": 0.7941828928887844,
      "num_tokens": 251589555.0,
      "step": 7883
    },
    {
      "entropy": 0.6310151033103466,
      "epoch": 0.7253782324425967,
      "grad_norm": 0.15015818178653717,
      "learning_rate": 7.582420952556199e-05,
      "loss": 0.6112,
      "mean_token_accuracy": 0.8181525655090809,
      "num_tokens": 251622050.0,
      "step": 7884
    },
    {
      "entropy": 0.8106321711093187,
      "epoch": 0.7254702388140378,
      "grad_norm": 0.15475399792194366,
      "learning_rate": 7.582114269942037e-05,
      "loss": 0.7921,
      "mean_token_accuracy": 0.7731204256415367,
      "num_tokens": 251653919.0,
      "step": 7885
    },
    {
      "entropy": 0.8742558993399143,
      "epoch": 0.7255622451854791,
      "grad_norm": 0.1612847000360489,
      "learning_rate": 7.581807587327875e-05,
      "loss": 0.8618,
      "mean_token_accuracy": 0.7514798901975155,
      "num_tokens": 251686465.0,
      "step": 7886
    },
    {
      "entropy": 0.6950328722596169,
      "epoch": 0.7256542515569203,
      "grad_norm": 0.15197235345840454,
      "learning_rate": 7.581500904713712e-05,
      "loss": 0.6802,
      "mean_token_accuracy": 0.7988152392208576,
      "num_tokens": 251718413.0,
      "step": 7887
    },
    {
      "entropy": 0.6344367666170001,
      "epoch": 0.7257462579283616,
      "grad_norm": 0.1490088254213333,
      "learning_rate": 7.581194222099549e-05,
      "loss": 0.609,
      "mean_token_accuracy": 0.8184834457933903,
      "num_tokens": 251749967.0,
      "step": 7888
    },
    {
      "entropy": 0.7004262674599886,
      "epoch": 0.7258382642998028,
      "grad_norm": 0.1679946482181549,
      "learning_rate": 7.580887539485387e-05,
      "loss": 0.6989,
      "mean_token_accuracy": 0.7959790267050266,
      "num_tokens": 251781446.0,
      "step": 7889
    },
    {
      "entropy": 0.7620374727994204,
      "epoch": 0.7259302706712439,
      "grad_norm": 0.15673412382602692,
      "learning_rate": 7.580580856871224e-05,
      "loss": 0.7728,
      "mean_token_accuracy": 0.7727949246764183,
      "num_tokens": 251813633.0,
      "step": 7890
    },
    {
      "entropy": 0.8456238284707069,
      "epoch": 0.7260222770426852,
      "grad_norm": 0.16465577483177185,
      "learning_rate": 7.580274174257062e-05,
      "loss": 0.8518,
      "mean_token_accuracy": 0.7494657635688782,
      "num_tokens": 251845972.0,
      "step": 7891
    },
    {
      "entropy": 0.7605965677648783,
      "epoch": 0.7261142834141264,
      "grad_norm": 0.15925534069538116,
      "learning_rate": 7.579967491642899e-05,
      "loss": 0.766,
      "mean_token_accuracy": 0.7782953009009361,
      "num_tokens": 251878383.0,
      "step": 7892
    },
    {
      "entropy": 0.7022535521537066,
      "epoch": 0.7262062897855677,
      "grad_norm": 0.15416285395622253,
      "learning_rate": 7.579660809028737e-05,
      "loss": 0.6878,
      "mean_token_accuracy": 0.7924491502344608,
      "num_tokens": 251910960.0,
      "step": 7893
    },
    {
      "entropy": 0.7717578839510679,
      "epoch": 0.7262982961570089,
      "grad_norm": 0.1520114541053772,
      "learning_rate": 7.579354126414574e-05,
      "loss": 0.7641,
      "mean_token_accuracy": 0.7747473306953907,
      "num_tokens": 251943480.0,
      "step": 7894
    },
    {
      "entropy": 0.806653993204236,
      "epoch": 0.72639030252845,
      "grad_norm": 0.16699084639549255,
      "learning_rate": 7.57904744380041e-05,
      "loss": 0.7984,
      "mean_token_accuracy": 0.7667182683944702,
      "num_tokens": 251974954.0,
      "step": 7895
    },
    {
      "entropy": 0.7779343798756599,
      "epoch": 0.7264823088998913,
      "grad_norm": 0.1575072854757309,
      "learning_rate": 7.578740761186249e-05,
      "loss": 0.769,
      "mean_token_accuracy": 0.7684967294335365,
      "num_tokens": 252006937.0,
      "step": 7896
    },
    {
      "entropy": 0.7620575269684196,
      "epoch": 0.7265743152713325,
      "grad_norm": 0.16836895048618317,
      "learning_rate": 7.578434078572087e-05,
      "loss": 0.7492,
      "mean_token_accuracy": 0.7777226157486439,
      "num_tokens": 252038274.0,
      "step": 7897
    },
    {
      "entropy": 0.5793853607028723,
      "epoch": 0.7266663216427738,
      "grad_norm": 0.14838165044784546,
      "learning_rate": 7.578127395957924e-05,
      "loss": 0.5597,
      "mean_token_accuracy": 0.8294431194663048,
      "num_tokens": 252069973.0,
      "step": 7898
    },
    {
      "entropy": 0.821162324398756,
      "epoch": 0.726758328014215,
      "grad_norm": 0.16260945796966553,
      "learning_rate": 7.57782071334376e-05,
      "loss": 0.806,
      "mean_token_accuracy": 0.7652249783277512,
      "num_tokens": 252101774.0,
      "step": 7899
    },
    {
      "entropy": 0.8114989325404167,
      "epoch": 0.7268503343856562,
      "grad_norm": 0.15827545523643494,
      "learning_rate": 7.577514030729597e-05,
      "loss": 0.7845,
      "mean_token_accuracy": 0.770322747528553,
      "num_tokens": 252134102.0,
      "step": 7900
    },
    {
      "entropy": 0.7714655417948961,
      "epoch": 0.7269423407570974,
      "grad_norm": 0.16751447319984436,
      "learning_rate": 7.577207348115435e-05,
      "loss": 0.7497,
      "mean_token_accuracy": 0.7824319936335087,
      "num_tokens": 252165904.0,
      "step": 7901
    },
    {
      "entropy": 0.7357718478888273,
      "epoch": 0.7270343471285386,
      "grad_norm": 0.15308338403701782,
      "learning_rate": 7.576900665501274e-05,
      "loss": 0.7117,
      "mean_token_accuracy": 0.7864997088909149,
      "num_tokens": 252198012.0,
      "step": 7902
    },
    {
      "entropy": 0.737629123032093,
      "epoch": 0.7271263534999799,
      "grad_norm": 0.1595482975244522,
      "learning_rate": 7.57659398288711e-05,
      "loss": 0.7135,
      "mean_token_accuracy": 0.7899434119462967,
      "num_tokens": 252230216.0,
      "step": 7903
    },
    {
      "entropy": 0.806482782587409,
      "epoch": 0.7272183598714211,
      "grad_norm": 0.15723711252212524,
      "learning_rate": 7.576287300272948e-05,
      "loss": 0.7979,
      "mean_token_accuracy": 0.7674141339957714,
      "num_tokens": 252262092.0,
      "step": 7904
    },
    {
      "entropy": 0.7190824449062347,
      "epoch": 0.7273103662428623,
      "grad_norm": 0.15586069226264954,
      "learning_rate": 7.575980617658785e-05,
      "loss": 0.6982,
      "mean_token_accuracy": 0.793573435395956,
      "num_tokens": 252293035.0,
      "step": 7905
    },
    {
      "entropy": 0.7487450372427702,
      "epoch": 0.7274023726143035,
      "grad_norm": 0.16418296098709106,
      "learning_rate": 7.575673935044622e-05,
      "loss": 0.7557,
      "mean_token_accuracy": 0.7759072110056877,
      "num_tokens": 252323521.0,
      "step": 7906
    },
    {
      "entropy": 0.7843949273228645,
      "epoch": 0.7274943789857448,
      "grad_norm": 0.17025581002235413,
      "learning_rate": 7.57536725243046e-05,
      "loss": 0.7659,
      "mean_token_accuracy": 0.775711290538311,
      "num_tokens": 252354627.0,
      "step": 7907
    },
    {
      "entropy": 0.6166856847703457,
      "epoch": 0.727586385357186,
      "grad_norm": 0.14818920195102692,
      "learning_rate": 7.575060569816298e-05,
      "loss": 0.6089,
      "mean_token_accuracy": 0.8168659433722496,
      "num_tokens": 252387259.0,
      "step": 7908
    },
    {
      "entropy": 0.6661929748952389,
      "epoch": 0.7276783917286272,
      "grad_norm": 0.14638425409793854,
      "learning_rate": 7.574753887202135e-05,
      "loss": 0.6538,
      "mean_token_accuracy": 0.8061513751745224,
      "num_tokens": 252419587.0,
      "step": 7909
    },
    {
      "entropy": 0.7729924824088812,
      "epoch": 0.7277703981000684,
      "grad_norm": 0.1573622226715088,
      "learning_rate": 7.574447204587972e-05,
      "loss": 0.7778,
      "mean_token_accuracy": 0.7687584236264229,
      "num_tokens": 252452122.0,
      "step": 7910
    },
    {
      "entropy": 0.6905913837254047,
      "epoch": 0.7278624044715096,
      "grad_norm": 0.15070562064647675,
      "learning_rate": 7.574140521973809e-05,
      "loss": 0.6758,
      "mean_token_accuracy": 0.7991761192679405,
      "num_tokens": 252484433.0,
      "step": 7911
    },
    {
      "entropy": 0.7499027140438557,
      "epoch": 0.7279544108429509,
      "grad_norm": 0.15253068506717682,
      "learning_rate": 7.573833839359648e-05,
      "loss": 0.726,
      "mean_token_accuracy": 0.7845340967178345,
      "num_tokens": 252516213.0,
      "step": 7912
    },
    {
      "entropy": 0.7437638007104397,
      "epoch": 0.7280464172143921,
      "grad_norm": 0.1626688838005066,
      "learning_rate": 7.573527156745485e-05,
      "loss": 0.7377,
      "mean_token_accuracy": 0.7809381894767284,
      "num_tokens": 252548142.0,
      "step": 7913
    },
    {
      "entropy": 0.8698076009750366,
      "epoch": 0.7281384235858334,
      "grad_norm": 0.16642296314239502,
      "learning_rate": 7.573220474131322e-05,
      "loss": 0.8744,
      "mean_token_accuracy": 0.7391108125448227,
      "num_tokens": 252580085.0,
      "step": 7914
    },
    {
      "entropy": 0.7003586068749428,
      "epoch": 0.7282304299572745,
      "grad_norm": 0.15932756662368774,
      "learning_rate": 7.572913791517159e-05,
      "loss": 0.6815,
      "mean_token_accuracy": 0.7978714480996132,
      "num_tokens": 252612688.0,
      "step": 7915
    },
    {
      "entropy": 0.7636811882257462,
      "epoch": 0.7283224363287157,
      "grad_norm": 0.17223118245601654,
      "learning_rate": 7.572607108902997e-05,
      "loss": 0.7478,
      "mean_token_accuracy": 0.7776231840252876,
      "num_tokens": 252644657.0,
      "step": 7916
    },
    {
      "entropy": 0.7494610082358122,
      "epoch": 0.728414442700157,
      "grad_norm": 0.15488079190254211,
      "learning_rate": 7.572300426288835e-05,
      "loss": 0.7524,
      "mean_token_accuracy": 0.778341680765152,
      "num_tokens": 252677126.0,
      "step": 7917
    },
    {
      "entropy": 0.6931417249143124,
      "epoch": 0.7285064490715982,
      "grad_norm": 0.16079236567020416,
      "learning_rate": 7.571993743674672e-05,
      "loss": 0.692,
      "mean_token_accuracy": 0.7970300316810608,
      "num_tokens": 252709654.0,
      "step": 7918
    },
    {
      "entropy": 0.7279869019985199,
      "epoch": 0.7285984554430395,
      "grad_norm": 0.14930546283721924,
      "learning_rate": 7.571687061060508e-05,
      "loss": 0.7316,
      "mean_token_accuracy": 0.7811122015118599,
      "num_tokens": 252741212.0,
      "step": 7919
    },
    {
      "entropy": 0.623401215299964,
      "epoch": 0.7286904618144806,
      "grad_norm": 0.15416908264160156,
      "learning_rate": 7.571380378446347e-05,
      "loss": 0.6128,
      "mean_token_accuracy": 0.8173513375222683,
      "num_tokens": 252773802.0,
      "step": 7920
    },
    {
      "entropy": 0.7422445379197598,
      "epoch": 0.7287824681859219,
      "grad_norm": 0.15258558094501495,
      "learning_rate": 7.571073695832183e-05,
      "loss": 0.7303,
      "mean_token_accuracy": 0.7794930003583431,
      "num_tokens": 252806215.0,
      "step": 7921
    },
    {
      "entropy": 0.7337194718420506,
      "epoch": 0.7288744745573631,
      "grad_norm": 0.1622336506843567,
      "learning_rate": 7.570767013218021e-05,
      "loss": 0.724,
      "mean_token_accuracy": 0.7842049933969975,
      "num_tokens": 252838716.0,
      "step": 7922
    },
    {
      "entropy": 0.7989636845886707,
      "epoch": 0.7289664809288043,
      "grad_norm": 0.1623350977897644,
      "learning_rate": 7.570460330603858e-05,
      "loss": 0.8031,
      "mean_token_accuracy": 0.7682050429284573,
      "num_tokens": 252870555.0,
      "step": 7923
    },
    {
      "entropy": 0.6782918367534876,
      "epoch": 0.7290584873002456,
      "grad_norm": 0.15842172503471375,
      "learning_rate": 7.570153647989696e-05,
      "loss": 0.6519,
      "mean_token_accuracy": 0.8102654293179512,
      "num_tokens": 252903099.0,
      "step": 7924
    },
    {
      "entropy": 0.6524178143590689,
      "epoch": 0.7291504936716867,
      "grad_norm": 0.15155796706676483,
      "learning_rate": 7.569846965375533e-05,
      "loss": 0.6312,
      "mean_token_accuracy": 0.8138269148766994,
      "num_tokens": 252935421.0,
      "step": 7925
    },
    {
      "entropy": 0.7126379311084747,
      "epoch": 0.729242500043128,
      "grad_norm": 0.16040706634521484,
      "learning_rate": 7.56954028276137e-05,
      "loss": 0.6875,
      "mean_token_accuracy": 0.7964474894106388,
      "num_tokens": 252967062.0,
      "step": 7926
    },
    {
      "entropy": 0.76754280179739,
      "epoch": 0.7293345064145692,
      "grad_norm": 0.15099459886550903,
      "learning_rate": 7.569233600147208e-05,
      "loss": 0.7393,
      "mean_token_accuracy": 0.7786346860229969,
      "num_tokens": 252998777.0,
      "step": 7927
    },
    {
      "entropy": 0.6734980046749115,
      "epoch": 0.7294265127860105,
      "grad_norm": 0.1500142514705658,
      "learning_rate": 7.568926917533046e-05,
      "loss": 0.6571,
      "mean_token_accuracy": 0.7942510955035686,
      "num_tokens": 253030663.0,
      "step": 7928
    },
    {
      "entropy": 0.7833382487297058,
      "epoch": 0.7295185191574517,
      "grad_norm": 0.15442423522472382,
      "learning_rate": 7.568620234918883e-05,
      "loss": 0.7519,
      "mean_token_accuracy": 0.7783960066735744,
      "num_tokens": 253062431.0,
      "step": 7929
    },
    {
      "entropy": 0.7628982793539762,
      "epoch": 0.7296105255288928,
      "grad_norm": 0.15740135312080383,
      "learning_rate": 7.56831355230472e-05,
      "loss": 0.7557,
      "mean_token_accuracy": 0.7778056971728802,
      "num_tokens": 253094190.0,
      "step": 7930
    },
    {
      "entropy": 0.6838789284229279,
      "epoch": 0.7297025319003341,
      "grad_norm": 0.15055906772613525,
      "learning_rate": 7.568006869690557e-05,
      "loss": 0.6715,
      "mean_token_accuracy": 0.7983460910618305,
      "num_tokens": 253126742.0,
      "step": 7931
    },
    {
      "entropy": 0.75821821577847,
      "epoch": 0.7297945382717753,
      "grad_norm": 0.15673744678497314,
      "learning_rate": 7.567700187076395e-05,
      "loss": 0.7617,
      "mean_token_accuracy": 0.7718385346233845,
      "num_tokens": 253158260.0,
      "step": 7932
    },
    {
      "entropy": 0.8084640298038721,
      "epoch": 0.7298865446432166,
      "grad_norm": 0.1592136025428772,
      "learning_rate": 7.567393504462233e-05,
      "loss": 0.8255,
      "mean_token_accuracy": 0.7613926120102406,
      "num_tokens": 253189870.0,
      "step": 7933
    },
    {
      "entropy": 0.8548354189842939,
      "epoch": 0.7299785510146578,
      "grad_norm": 0.16113661229610443,
      "learning_rate": 7.56708682184807e-05,
      "loss": 0.8557,
      "mean_token_accuracy": 0.751461286097765,
      "num_tokens": 253221847.0,
      "step": 7934
    },
    {
      "entropy": 0.7176042851060629,
      "epoch": 0.7300705573860989,
      "grad_norm": 0.15872342884540558,
      "learning_rate": 7.566780139233908e-05,
      "loss": 0.7024,
      "mean_token_accuracy": 0.7907408252358437,
      "num_tokens": 253253293.0,
      "step": 7935
    },
    {
      "entropy": 0.6292059011757374,
      "epoch": 0.7301625637575402,
      "grad_norm": 0.15113651752471924,
      "learning_rate": 7.566473456619745e-05,
      "loss": 0.6131,
      "mean_token_accuracy": 0.8188026174902916,
      "num_tokens": 253285199.0,
      "step": 7936
    },
    {
      "entropy": 0.6483762543648481,
      "epoch": 0.7302545701289814,
      "grad_norm": 0.1532796025276184,
      "learning_rate": 7.566166774005581e-05,
      "loss": 0.6347,
      "mean_token_accuracy": 0.8157059140503407,
      "num_tokens": 253317967.0,
      "step": 7937
    },
    {
      "entropy": 0.6353138834238052,
      "epoch": 0.7303465765004227,
      "grad_norm": 0.15265235304832458,
      "learning_rate": 7.56586009139142e-05,
      "loss": 0.6174,
      "mean_token_accuracy": 0.8107044696807861,
      "num_tokens": 253350238.0,
      "step": 7938
    },
    {
      "entropy": 0.6863475609570742,
      "epoch": 0.7304385828718639,
      "grad_norm": 0.16050544381141663,
      "learning_rate": 7.565553408777258e-05,
      "loss": 0.6762,
      "mean_token_accuracy": 0.8000350929796696,
      "num_tokens": 253382773.0,
      "step": 7939
    },
    {
      "entropy": 0.7641975730657578,
      "epoch": 0.730530589243305,
      "grad_norm": 0.15134049952030182,
      "learning_rate": 7.565246726163094e-05,
      "loss": 0.7596,
      "mean_token_accuracy": 0.7772101350128651,
      "num_tokens": 253414889.0,
      "step": 7940
    },
    {
      "entropy": 0.7570315096527338,
      "epoch": 0.7306225956147463,
      "grad_norm": 0.15324823558330536,
      "learning_rate": 7.564940043548931e-05,
      "loss": 0.75,
      "mean_token_accuracy": 0.7740344144403934,
      "num_tokens": 253446728.0,
      "step": 7941
    },
    {
      "entropy": 0.8193526305258274,
      "epoch": 0.7307146019861875,
      "grad_norm": 0.16124288737773895,
      "learning_rate": 7.564633360934768e-05,
      "loss": 0.8132,
      "mean_token_accuracy": 0.7549918219447136,
      "num_tokens": 253477992.0,
      "step": 7942
    },
    {
      "entropy": 0.7826138585805893,
      "epoch": 0.7308066083576288,
      "grad_norm": 0.15467184782028198,
      "learning_rate": 7.564326678320606e-05,
      "loss": 0.7814,
      "mean_token_accuracy": 0.7728930488228798,
      "num_tokens": 253510515.0,
      "step": 7943
    },
    {
      "entropy": 0.789142481982708,
      "epoch": 0.73089861472907,
      "grad_norm": 0.16523361206054688,
      "learning_rate": 7.564019995706444e-05,
      "loss": 0.7777,
      "mean_token_accuracy": 0.7704470157623291,
      "num_tokens": 253542236.0,
      "step": 7944
    },
    {
      "entropy": 0.7951589170843363,
      "epoch": 0.7309906211005112,
      "grad_norm": 0.21056526899337769,
      "learning_rate": 7.563713313092281e-05,
      "loss": 0.8231,
      "mean_token_accuracy": 0.7654098644852638,
      "num_tokens": 253573971.0,
      "step": 7945
    },
    {
      "entropy": 0.7486943807452917,
      "epoch": 0.7310826274719524,
      "grad_norm": 0.15937910974025726,
      "learning_rate": 7.563406630478118e-05,
      "loss": 0.7232,
      "mean_token_accuracy": 0.7888135239481926,
      "num_tokens": 253606203.0,
      "step": 7946
    },
    {
      "entropy": 0.7433761693537235,
      "epoch": 0.7311746338433937,
      "grad_norm": 0.15660560131072998,
      "learning_rate": 7.563099947863956e-05,
      "loss": 0.715,
      "mean_token_accuracy": 0.7913198694586754,
      "num_tokens": 253638314.0,
      "step": 7947
    },
    {
      "entropy": 0.7425393909215927,
      "epoch": 0.7312666402148349,
      "grad_norm": 0.16296958923339844,
      "learning_rate": 7.562793265249793e-05,
      "loss": 0.7483,
      "mean_token_accuracy": 0.7790047861635685,
      "num_tokens": 253670754.0,
      "step": 7948
    },
    {
      "entropy": 0.7805876862257719,
      "epoch": 0.7313586465862761,
      "grad_norm": 0.15797635912895203,
      "learning_rate": 7.562486582635631e-05,
      "loss": 0.7657,
      "mean_token_accuracy": 0.7695169858634472,
      "num_tokens": 253702905.0,
      "step": 7949
    },
    {
      "entropy": 0.7614521440118551,
      "epoch": 0.7314506529577173,
      "grad_norm": 0.15333186089992523,
      "learning_rate": 7.562179900021468e-05,
      "loss": 0.7459,
      "mean_token_accuracy": 0.7785964533686638,
      "num_tokens": 253735074.0,
      "step": 7950
    },
    {
      "entropy": 0.6324459817260504,
      "epoch": 0.7315426593291585,
      "grad_norm": 0.1438017636537552,
      "learning_rate": 7.561873217407306e-05,
      "loss": 0.6065,
      "mean_token_accuracy": 0.8146380670368671,
      "num_tokens": 253766138.0,
      "step": 7951
    },
    {
      "entropy": 0.818536814302206,
      "epoch": 0.7316346657005998,
      "grad_norm": 0.1557816118001938,
      "learning_rate": 7.561566534793143e-05,
      "loss": 0.8059,
      "mean_token_accuracy": 0.7648622393608093,
      "num_tokens": 253798431.0,
      "step": 7952
    },
    {
      "entropy": 0.7679702155292034,
      "epoch": 0.731726672072041,
      "grad_norm": 0.16103729605674744,
      "learning_rate": 7.56125985217898e-05,
      "loss": 0.758,
      "mean_token_accuracy": 0.7805511616170406,
      "num_tokens": 253830367.0,
      "step": 7953
    },
    {
      "entropy": 0.8294976744800806,
      "epoch": 0.7318186784434823,
      "grad_norm": 0.1651262491941452,
      "learning_rate": 7.560953169564818e-05,
      "loss": 0.8107,
      "mean_token_accuracy": 0.7618746869266033,
      "num_tokens": 253861916.0,
      "step": 7954
    },
    {
      "entropy": 0.680297177284956,
      "epoch": 0.7319106848149234,
      "grad_norm": 0.15806886553764343,
      "learning_rate": 7.560646486950656e-05,
      "loss": 0.6644,
      "mean_token_accuracy": 0.8017854616045952,
      "num_tokens": 253893982.0,
      "step": 7955
    },
    {
      "entropy": 0.7435339726507664,
      "epoch": 0.7320026911863646,
      "grad_norm": 0.1549362987279892,
      "learning_rate": 7.560339804336493e-05,
      "loss": 0.7256,
      "mean_token_accuracy": 0.7835787087678909,
      "num_tokens": 253925297.0,
      "step": 7956
    },
    {
      "entropy": 0.7855102196335793,
      "epoch": 0.7320946975578059,
      "grad_norm": 0.15753905475139618,
      "learning_rate": 7.56003312172233e-05,
      "loss": 0.7761,
      "mean_token_accuracy": 0.7737756036221981,
      "num_tokens": 253957177.0,
      "step": 7957
    },
    {
      "entropy": 0.7434988003224134,
      "epoch": 0.7321867039292471,
      "grad_norm": 0.15987050533294678,
      "learning_rate": 7.559726439108166e-05,
      "loss": 0.7237,
      "mean_token_accuracy": 0.784067627042532,
      "num_tokens": 253989011.0,
      "step": 7958
    },
    {
      "entropy": 0.7001366335898638,
      "epoch": 0.7322787103006884,
      "grad_norm": 0.1554851531982422,
      "learning_rate": 7.559419756494006e-05,
      "loss": 0.695,
      "mean_token_accuracy": 0.7912022359669209,
      "num_tokens": 254021465.0,
      "step": 7959
    },
    {
      "entropy": 0.655152378603816,
      "epoch": 0.7323707166721295,
      "grad_norm": 0.16099496185779572,
      "learning_rate": 7.559113073879842e-05,
      "loss": 0.6363,
      "mean_token_accuracy": 0.8110510632395744,
      "num_tokens": 254053696.0,
      "step": 7960
    },
    {
      "entropy": 0.6959823165088892,
      "epoch": 0.7324627230435707,
      "grad_norm": 0.1550913006067276,
      "learning_rate": 7.558806391265679e-05,
      "loss": 0.6889,
      "mean_token_accuracy": 0.7942805290222168,
      "num_tokens": 254084216.0,
      "step": 7961
    },
    {
      "entropy": 0.6347840195521712,
      "epoch": 0.732554729415012,
      "grad_norm": 0.15514636039733887,
      "learning_rate": 7.558499708651516e-05,
      "loss": 0.6399,
      "mean_token_accuracy": 0.8107388615608215,
      "num_tokens": 254115982.0,
      "step": 7962
    },
    {
      "entropy": 0.7805109936743975,
      "epoch": 0.7326467357864532,
      "grad_norm": 0.15526270866394043,
      "learning_rate": 7.558193026037354e-05,
      "loss": 0.7777,
      "mean_token_accuracy": 0.7732713036239147,
      "num_tokens": 254148701.0,
      "step": 7963
    },
    {
      "entropy": 0.6908981762826443,
      "epoch": 0.7327387421578945,
      "grad_norm": 0.15398359298706055,
      "learning_rate": 7.557886343423192e-05,
      "loss": 0.6806,
      "mean_token_accuracy": 0.7986280545592308,
      "num_tokens": 254180770.0,
      "step": 7964
    },
    {
      "entropy": 0.7456530909985304,
      "epoch": 0.7328307485293356,
      "grad_norm": 0.16093984246253967,
      "learning_rate": 7.557579660809029e-05,
      "loss": 0.7142,
      "mean_token_accuracy": 0.7888236753642559,
      "num_tokens": 254212530.0,
      "step": 7965
    },
    {
      "entropy": 0.7020308570936322,
      "epoch": 0.7329227549007769,
      "grad_norm": 0.16295042634010315,
      "learning_rate": 7.557272978194867e-05,
      "loss": 0.7022,
      "mean_token_accuracy": 0.7947042398154736,
      "num_tokens": 254244867.0,
      "step": 7966
    },
    {
      "entropy": 0.7056153975427151,
      "epoch": 0.7330147612722181,
      "grad_norm": 0.14856092631816864,
      "learning_rate": 7.556966295580704e-05,
      "loss": 0.6904,
      "mean_token_accuracy": 0.7949605360627174,
      "num_tokens": 254277206.0,
      "step": 7967
    },
    {
      "entropy": 0.7195652909576893,
      "epoch": 0.7331067676436593,
      "grad_norm": 0.15388768911361694,
      "learning_rate": 7.556659612966541e-05,
      "loss": 0.6967,
      "mean_token_accuracy": 0.795229934155941,
      "num_tokens": 254309193.0,
      "step": 7968
    },
    {
      "entropy": 0.737976549193263,
      "epoch": 0.7331987740151006,
      "grad_norm": 0.15955117344856262,
      "learning_rate": 7.556352930352379e-05,
      "loss": 0.7281,
      "mean_token_accuracy": 0.7848136909306049,
      "num_tokens": 254341496.0,
      "step": 7969
    },
    {
      "entropy": 0.647602709941566,
      "epoch": 0.7332907803865417,
      "grad_norm": 0.15422207117080688,
      "learning_rate": 7.556046247738217e-05,
      "loss": 0.6381,
      "mean_token_accuracy": 0.8123192116618156,
      "num_tokens": 254373845.0,
      "step": 7970
    },
    {
      "entropy": 0.7097981050610542,
      "epoch": 0.733382786757983,
      "grad_norm": 0.15581898391246796,
      "learning_rate": 7.555739565124054e-05,
      "loss": 0.7076,
      "mean_token_accuracy": 0.7933853715658188,
      "num_tokens": 254406104.0,
      "step": 7971
    },
    {
      "entropy": 0.6440607756376266,
      "epoch": 0.7334747931294242,
      "grad_norm": 0.15071403980255127,
      "learning_rate": 7.55543288250989e-05,
      "loss": 0.628,
      "mean_token_accuracy": 0.8126105293631554,
      "num_tokens": 254437552.0,
      "step": 7972
    },
    {
      "entropy": 0.7490789759904146,
      "epoch": 0.7335667995008655,
      "grad_norm": 0.1613229364156723,
      "learning_rate": 7.555126199895727e-05,
      "loss": 0.7243,
      "mean_token_accuracy": 0.7870497144758701,
      "num_tokens": 254469677.0,
      "step": 7973
    },
    {
      "entropy": 0.8512578178197145,
      "epoch": 0.7336588058723067,
      "grad_norm": 0.16155968606472015,
      "learning_rate": 7.554819517281566e-05,
      "loss": 0.8394,
      "mean_token_accuracy": 0.7517282068729401,
      "num_tokens": 254501161.0,
      "step": 7974
    },
    {
      "entropy": 0.7780516855418682,
      "epoch": 0.7337508122437478,
      "grad_norm": 0.1706516146659851,
      "learning_rate": 7.554512834667404e-05,
      "loss": 0.7684,
      "mean_token_accuracy": 0.7738902382552624,
      "num_tokens": 254533701.0,
      "step": 7975
    },
    {
      "entropy": 0.8212194889783859,
      "epoch": 0.7338428186151891,
      "grad_norm": 0.14994770288467407,
      "learning_rate": 7.55420615205324e-05,
      "loss": 0.8189,
      "mean_token_accuracy": 0.7640267796814442,
      "num_tokens": 254566049.0,
      "step": 7976
    },
    {
      "entropy": 0.8171414844691753,
      "epoch": 0.7339348249866303,
      "grad_norm": 0.16187578439712524,
      "learning_rate": 7.553899469439077e-05,
      "loss": 0.8066,
      "mean_token_accuracy": 0.7622466757893562,
      "num_tokens": 254597513.0,
      "step": 7977
    },
    {
      "entropy": 0.7761191260069609,
      "epoch": 0.7340268313580716,
      "grad_norm": 0.16160014271736145,
      "learning_rate": 7.553592786824915e-05,
      "loss": 0.7776,
      "mean_token_accuracy": 0.771003108471632,
      "num_tokens": 254629595.0,
      "step": 7978
    },
    {
      "entropy": 0.7726502437144518,
      "epoch": 0.7341188377295128,
      "grad_norm": 0.15257717669010162,
      "learning_rate": 7.553286104210752e-05,
      "loss": 0.7579,
      "mean_token_accuracy": 0.777103029191494,
      "num_tokens": 254661623.0,
      "step": 7979
    },
    {
      "entropy": 0.736591711640358,
      "epoch": 0.7342108441009539,
      "grad_norm": 0.15118108689785004,
      "learning_rate": 7.55297942159659e-05,
      "loss": 0.7256,
      "mean_token_accuracy": 0.7817702107131481,
      "num_tokens": 254693703.0,
      "step": 7980
    },
    {
      "entropy": 0.7769963908940554,
      "epoch": 0.7343028504723952,
      "grad_norm": 0.15820524096488953,
      "learning_rate": 7.552672738982427e-05,
      "loss": 0.7566,
      "mean_token_accuracy": 0.7763813808560371,
      "num_tokens": 254725922.0,
      "step": 7981
    },
    {
      "entropy": 0.7541712149977684,
      "epoch": 0.7343948568438364,
      "grad_norm": 0.15493535995483398,
      "learning_rate": 7.552366056368265e-05,
      "loss": 0.7461,
      "mean_token_accuracy": 0.7797490023076534,
      "num_tokens": 254757866.0,
      "step": 7982
    },
    {
      "entropy": 0.8331007808446884,
      "epoch": 0.7344868632152777,
      "grad_norm": 0.1810678392648697,
      "learning_rate": 7.552059373754102e-05,
      "loss": 0.8302,
      "mean_token_accuracy": 0.7575288899242878,
      "num_tokens": 254789490.0,
      "step": 7983
    },
    {
      "entropy": 0.7025050073862076,
      "epoch": 0.7345788695867189,
      "grad_norm": 0.1573040634393692,
      "learning_rate": 7.551752691139939e-05,
      "loss": 0.6927,
      "mean_token_accuracy": 0.7961372695863247,
      "num_tokens": 254821215.0,
      "step": 7984
    },
    {
      "entropy": 0.6755422446876764,
      "epoch": 0.73467087595816,
      "grad_norm": 0.15595434606075287,
      "learning_rate": 7.551446008525777e-05,
      "loss": 0.681,
      "mean_token_accuracy": 0.7991278544068336,
      "num_tokens": 254853444.0,
      "step": 7985
    },
    {
      "entropy": 0.6667997930198908,
      "epoch": 0.7347628823296013,
      "grad_norm": 0.16239546239376068,
      "learning_rate": 7.551139325911615e-05,
      "loss": 0.6467,
      "mean_token_accuracy": 0.8062839731574059,
      "num_tokens": 254884040.0,
      "step": 7986
    },
    {
      "entropy": 0.7444989420473576,
      "epoch": 0.7348548887010425,
      "grad_norm": 0.1573944389820099,
      "learning_rate": 7.550832643297452e-05,
      "loss": 0.7339,
      "mean_token_accuracy": 0.7828828357160091,
      "num_tokens": 254916281.0,
      "step": 7987
    },
    {
      "entropy": 0.6780536621809006,
      "epoch": 0.7349468950724838,
      "grad_norm": 0.15516555309295654,
      "learning_rate": 7.550525960683289e-05,
      "loss": 0.6753,
      "mean_token_accuracy": 0.7984740547835827,
      "num_tokens": 254948716.0,
      "step": 7988
    },
    {
      "entropy": 0.6599221378564835,
      "epoch": 0.735038901443925,
      "grad_norm": 0.14724396169185638,
      "learning_rate": 7.550219278069126e-05,
      "loss": 0.6468,
      "mean_token_accuracy": 0.8079675287008286,
      "num_tokens": 254980657.0,
      "step": 7989
    },
    {
      "entropy": 0.7872522165998816,
      "epoch": 0.7351309078153662,
      "grad_norm": 0.1596556007862091,
      "learning_rate": 7.549912595454965e-05,
      "loss": 0.7749,
      "mean_token_accuracy": 0.7741753868758678,
      "num_tokens": 255012262.0,
      "step": 7990
    },
    {
      "entropy": 0.7418296858668327,
      "epoch": 0.7352229141868074,
      "grad_norm": 0.15662319958209991,
      "learning_rate": 7.549605912840802e-05,
      "loss": 0.7263,
      "mean_token_accuracy": 0.7840980179607868,
      "num_tokens": 255044206.0,
      "step": 7991
    },
    {
      "entropy": 0.6964784748852253,
      "epoch": 0.7353149205582487,
      "grad_norm": 0.15638615190982819,
      "learning_rate": 7.549299230226639e-05,
      "loss": 0.6977,
      "mean_token_accuracy": 0.7919233664870262,
      "num_tokens": 255076664.0,
      "step": 7992
    },
    {
      "entropy": 0.8068237975239754,
      "epoch": 0.7354069269296899,
      "grad_norm": 0.1623249650001526,
      "learning_rate": 7.548992547612477e-05,
      "loss": 0.8139,
      "mean_token_accuracy": 0.7629335671663284,
      "num_tokens": 255108806.0,
      "step": 7993
    },
    {
      "entropy": 0.6661587236449122,
      "epoch": 0.7354989333011311,
      "grad_norm": 0.15239734947681427,
      "learning_rate": 7.548685864998314e-05,
      "loss": 0.6564,
      "mean_token_accuracy": 0.7982831485569477,
      "num_tokens": 255140378.0,
      "step": 7994
    },
    {
      "entropy": 0.7915634661912918,
      "epoch": 0.7355909396725723,
      "grad_norm": 0.18126919865608215,
      "learning_rate": 7.548379182384152e-05,
      "loss": 0.7838,
      "mean_token_accuracy": 0.7701074443757534,
      "num_tokens": 255172392.0,
      "step": 7995
    },
    {
      "entropy": 0.665026493370533,
      "epoch": 0.7356829460440135,
      "grad_norm": 0.1556612253189087,
      "learning_rate": 7.548072499769988e-05,
      "loss": 0.6428,
      "mean_token_accuracy": 0.8094350472092628,
      "num_tokens": 255204617.0,
      "step": 7996
    },
    {
      "entropy": 0.8218211624771357,
      "epoch": 0.7357749524154548,
      "grad_norm": 0.1640169620513916,
      "learning_rate": 7.547765817155827e-05,
      "loss": 0.8076,
      "mean_token_accuracy": 0.7655117586255074,
      "num_tokens": 255236661.0,
      "step": 7997
    },
    {
      "entropy": 0.7972464710474014,
      "epoch": 0.735866958786896,
      "grad_norm": 0.17365577816963196,
      "learning_rate": 7.547459134541663e-05,
      "loss": 0.792,
      "mean_token_accuracy": 0.7665201239287853,
      "num_tokens": 255268355.0,
      "step": 7998
    },
    {
      "entropy": 0.6556734591722488,
      "epoch": 0.7359589651583373,
      "grad_norm": 0.15243494510650635,
      "learning_rate": 7.5471524519275e-05,
      "loss": 0.6426,
      "mean_token_accuracy": 0.8068861626088619,
      "num_tokens": 255300154.0,
      "step": 7999
    },
    {
      "entropy": 0.7967630289494991,
      "epoch": 0.7360509715297784,
      "grad_norm": 0.15824931859970093,
      "learning_rate": 7.546845769313338e-05,
      "loss": 0.7971,
      "mean_token_accuracy": 0.768410213291645,
      "num_tokens": 255332193.0,
      "step": 8000
    },
    {
      "entropy": 0.8360033445060253,
      "epoch": 0.7361429779012196,
      "grad_norm": 0.16119720041751862,
      "learning_rate": 7.546539086699176e-05,
      "loss": 0.8121,
      "mean_token_accuracy": 0.758455652743578,
      "num_tokens": 255364273.0,
      "step": 8001
    },
    {
      "entropy": 0.7262035720050335,
      "epoch": 0.7362349842726609,
      "grad_norm": 0.15905381739139557,
      "learning_rate": 7.546232404085013e-05,
      "loss": 0.6994,
      "mean_token_accuracy": 0.7915507592260838,
      "num_tokens": 255396014.0,
      "step": 8002
    },
    {
      "entropy": 0.8733964338898659,
      "epoch": 0.7363269906441021,
      "grad_norm": 0.15840931236743927,
      "learning_rate": 7.54592572147085e-05,
      "loss": 0.8686,
      "mean_token_accuracy": 0.7526811398565769,
      "num_tokens": 255427479.0,
      "step": 8003
    },
    {
      "entropy": 0.8087934069335461,
      "epoch": 0.7364189970155434,
      "grad_norm": 0.16326186060905457,
      "learning_rate": 7.545619038856687e-05,
      "loss": 0.8195,
      "mean_token_accuracy": 0.7594933807849884,
      "num_tokens": 255458843.0,
      "step": 8004
    },
    {
      "entropy": 0.7902308516204357,
      "epoch": 0.7365110033869845,
      "grad_norm": 0.15848040580749512,
      "learning_rate": 7.545312356242525e-05,
      "loss": 0.7982,
      "mean_token_accuracy": 0.7672663033008575,
      "num_tokens": 255491051.0,
      "step": 8005
    },
    {
      "entropy": 0.6576723139733076,
      "epoch": 0.7366030097584257,
      "grad_norm": 0.15184128284454346,
      "learning_rate": 7.545005673628363e-05,
      "loss": 0.6495,
      "mean_token_accuracy": 0.8091097958385944,
      "num_tokens": 255523117.0,
      "step": 8006
    },
    {
      "entropy": 0.8132628202438354,
      "epoch": 0.736695016129867,
      "grad_norm": 0.15818488597869873,
      "learning_rate": 7.5446989910142e-05,
      "loss": 0.8084,
      "mean_token_accuracy": 0.76482904702425,
      "num_tokens": 255554837.0,
      "step": 8007
    },
    {
      "entropy": 0.6521974243223667,
      "epoch": 0.7367870225013082,
      "grad_norm": 0.15826258063316345,
      "learning_rate": 7.544392308400037e-05,
      "loss": 0.647,
      "mean_token_accuracy": 0.8039029464125633,
      "num_tokens": 255586652.0,
      "step": 8008
    },
    {
      "entropy": 0.7723423540592194,
      "epoch": 0.7368790288727495,
      "grad_norm": 0.15902003645896912,
      "learning_rate": 7.544085625785875e-05,
      "loss": 0.7594,
      "mean_token_accuracy": 0.7727055437862873,
      "num_tokens": 255617316.0,
      "step": 8009
    },
    {
      "entropy": 0.6533186715096235,
      "epoch": 0.7369710352441906,
      "grad_norm": 0.15751589834690094,
      "learning_rate": 7.543778943171712e-05,
      "loss": 0.6369,
      "mean_token_accuracy": 0.8121948018670082,
      "num_tokens": 255649912.0,
      "step": 8010
    },
    {
      "entropy": 0.74369134940207,
      "epoch": 0.7370630416156319,
      "grad_norm": 0.15423333644866943,
      "learning_rate": 7.54347226055755e-05,
      "loss": 0.7166,
      "mean_token_accuracy": 0.7868922427296638,
      "num_tokens": 255681358.0,
      "step": 8011
    },
    {
      "entropy": 0.7384578827768564,
      "epoch": 0.7371550479870731,
      "grad_norm": 0.15560251474380493,
      "learning_rate": 7.543165577943387e-05,
      "loss": 0.713,
      "mean_token_accuracy": 0.7907814644277096,
      "num_tokens": 255713630.0,
      "step": 8012
    },
    {
      "entropy": 0.588830528780818,
      "epoch": 0.7372470543585143,
      "grad_norm": 0.14754718542099,
      "learning_rate": 7.542858895329225e-05,
      "loss": 0.562,
      "mean_token_accuracy": 0.8276812620460987,
      "num_tokens": 255745609.0,
      "step": 8013
    },
    {
      "entropy": 0.8926608748733997,
      "epoch": 0.7373390607299556,
      "grad_norm": 0.16081731021404266,
      "learning_rate": 7.542552212715061e-05,
      "loss": 0.8697,
      "mean_token_accuracy": 0.747456032782793,
      "num_tokens": 255777501.0,
      "step": 8014
    },
    {
      "entropy": 0.8441969305276871,
      "epoch": 0.7374310671013967,
      "grad_norm": 0.15751288831233978,
      "learning_rate": 7.542245530100898e-05,
      "loss": 0.8329,
      "mean_token_accuracy": 0.7586066052317619,
      "num_tokens": 255809953.0,
      "step": 8015
    },
    {
      "entropy": 0.6959374882280827,
      "epoch": 0.737523073472838,
      "grad_norm": 0.15334591269493103,
      "learning_rate": 7.541938847486736e-05,
      "loss": 0.6808,
      "mean_token_accuracy": 0.7938166186213493,
      "num_tokens": 255841711.0,
      "step": 8016
    },
    {
      "entropy": 0.6627671495079994,
      "epoch": 0.7376150798442792,
      "grad_norm": 0.15069018304347992,
      "learning_rate": 7.541632164872575e-05,
      "loss": 0.6537,
      "mean_token_accuracy": 0.8049882426857948,
      "num_tokens": 255874148.0,
      "step": 8017
    },
    {
      "entropy": 0.7220156006515026,
      "epoch": 0.7377070862157205,
      "grad_norm": 0.17264895141124725,
      "learning_rate": 7.541325482258411e-05,
      "loss": 0.6891,
      "mean_token_accuracy": 0.7951749302446842,
      "num_tokens": 255905599.0,
      "step": 8018
    },
    {
      "entropy": 0.72829769551754,
      "epoch": 0.7377990925871617,
      "grad_norm": 0.16382116079330444,
      "learning_rate": 7.541018799644248e-05,
      "loss": 0.7345,
      "mean_token_accuracy": 0.784874901175499,
      "num_tokens": 255938288.0,
      "step": 8019
    },
    {
      "entropy": 0.6420098710805178,
      "epoch": 0.7378910989586028,
      "grad_norm": 0.1707415133714676,
      "learning_rate": 7.540712117030085e-05,
      "loss": 0.6486,
      "mean_token_accuracy": 0.8076644279062748,
      "num_tokens": 255970166.0,
      "step": 8020
    },
    {
      "entropy": 0.7593545131385326,
      "epoch": 0.7379831053300441,
      "grad_norm": 0.15128964185714722,
      "learning_rate": 7.540405434415923e-05,
      "loss": 0.75,
      "mean_token_accuracy": 0.7775575295090675,
      "num_tokens": 256001512.0,
      "step": 8021
    },
    {
      "entropy": 0.7964868377894163,
      "epoch": 0.7380751117014853,
      "grad_norm": 0.1687273383140564,
      "learning_rate": 7.540098751801761e-05,
      "loss": 0.7921,
      "mean_token_accuracy": 0.7682682015001774,
      "num_tokens": 256033658.0,
      "step": 8022
    },
    {
      "entropy": 0.7536402512341738,
      "epoch": 0.7381671180729266,
      "grad_norm": 0.15843546390533447,
      "learning_rate": 7.539792069187598e-05,
      "loss": 0.761,
      "mean_token_accuracy": 0.7774581201374531,
      "num_tokens": 256065895.0,
      "step": 8023
    },
    {
      "entropy": 0.7514795828610659,
      "epoch": 0.7382591244443678,
      "grad_norm": 0.15582092106342316,
      "learning_rate": 7.539485386573436e-05,
      "loss": 0.7431,
      "mean_token_accuracy": 0.7838446088135242,
      "num_tokens": 256098438.0,
      "step": 8024
    },
    {
      "entropy": 0.7972057573497295,
      "epoch": 0.7383511308158089,
      "grad_norm": 0.15610793232917786,
      "learning_rate": 7.539178703959273e-05,
      "loss": 0.8007,
      "mean_token_accuracy": 0.7663934044539928,
      "num_tokens": 256130292.0,
      "step": 8025
    },
    {
      "entropy": 0.834581496194005,
      "epoch": 0.7384431371872502,
      "grad_norm": 0.16790282726287842,
      "learning_rate": 7.53887202134511e-05,
      "loss": 0.8175,
      "mean_token_accuracy": 0.7632886916399002,
      "num_tokens": 256161028.0,
      "step": 8026
    },
    {
      "entropy": 0.7239124495536089,
      "epoch": 0.7385351435586914,
      "grad_norm": 0.15775762498378754,
      "learning_rate": 7.538565338730948e-05,
      "loss": 0.7081,
      "mean_token_accuracy": 0.7892268262803555,
      "num_tokens": 256193158.0,
      "step": 8027
    },
    {
      "entropy": 0.6899391654878855,
      "epoch": 0.7386271499301327,
      "grad_norm": 0.15455761551856995,
      "learning_rate": 7.538258656116786e-05,
      "loss": 0.6737,
      "mean_token_accuracy": 0.7978865690529346,
      "num_tokens": 256225024.0,
      "step": 8028
    },
    {
      "entropy": 0.751967029646039,
      "epoch": 0.7387191563015739,
      "grad_norm": 0.15287940204143524,
      "learning_rate": 7.537951973502623e-05,
      "loss": 0.7466,
      "mean_token_accuracy": 0.7832746878266335,
      "num_tokens": 256257050.0,
      "step": 8029
    },
    {
      "entropy": 0.7759305238723755,
      "epoch": 0.738811162673015,
      "grad_norm": 0.16143012046813965,
      "learning_rate": 7.53764529088846e-05,
      "loss": 0.7603,
      "mean_token_accuracy": 0.777433205395937,
      "num_tokens": 256289376.0,
      "step": 8030
    },
    {
      "entropy": 0.838329678401351,
      "epoch": 0.7389031690444563,
      "grad_norm": 0.15767909586429596,
      "learning_rate": 7.537338608274296e-05,
      "loss": 0.8197,
      "mean_token_accuracy": 0.7608492560684681,
      "num_tokens": 256320504.0,
      "step": 8031
    },
    {
      "entropy": 0.7722492758184671,
      "epoch": 0.7389951754158975,
      "grad_norm": 0.15661895275115967,
      "learning_rate": 7.537031925660136e-05,
      "loss": 0.7618,
      "mean_token_accuracy": 0.7751695849001408,
      "num_tokens": 256352472.0,
      "step": 8032
    },
    {
      "entropy": 0.8077216055244207,
      "epoch": 0.7390871817873388,
      "grad_norm": 0.1662754863500595,
      "learning_rate": 7.536725243045973e-05,
      "loss": 0.8137,
      "mean_token_accuracy": 0.7634897232055664,
      "num_tokens": 256384832.0,
      "step": 8033
    },
    {
      "entropy": 0.6778822932392359,
      "epoch": 0.73917918815878,
      "grad_norm": 0.16128943860530853,
      "learning_rate": 7.53641856043181e-05,
      "loss": 0.6771,
      "mean_token_accuracy": 0.7942784056067467,
      "num_tokens": 256416647.0,
      "step": 8034
    },
    {
      "entropy": 0.6921396143734455,
      "epoch": 0.7392711945302212,
      "grad_norm": 0.16028742492198944,
      "learning_rate": 7.536111877817646e-05,
      "loss": 0.6658,
      "mean_token_accuracy": 0.8032757826149464,
      "num_tokens": 256448642.0,
      "step": 8035
    },
    {
      "entropy": 0.708757609128952,
      "epoch": 0.7393632009016624,
      "grad_norm": 0.16418242454528809,
      "learning_rate": 7.535805195203484e-05,
      "loss": 0.6971,
      "mean_token_accuracy": 0.796929806470871,
      "num_tokens": 256480067.0,
      "step": 8036
    },
    {
      "entropy": 0.7324373070150614,
      "epoch": 0.7394552072731037,
      "grad_norm": 0.15333658456802368,
      "learning_rate": 7.535498512589322e-05,
      "loss": 0.7117,
      "mean_token_accuracy": 0.7951207347214222,
      "num_tokens": 256511851.0,
      "step": 8037
    },
    {
      "entropy": 0.7737599983811378,
      "epoch": 0.7395472136445449,
      "grad_norm": 0.1624949872493744,
      "learning_rate": 7.535191829975159e-05,
      "loss": 0.7484,
      "mean_token_accuracy": 0.7751474156975746,
      "num_tokens": 256543602.0,
      "step": 8038
    },
    {
      "entropy": 0.7295498121529818,
      "epoch": 0.7396392200159861,
      "grad_norm": 0.15187206864356995,
      "learning_rate": 7.534885147360996e-05,
      "loss": 0.7112,
      "mean_token_accuracy": 0.7884564958512783,
      "num_tokens": 256575238.0,
      "step": 8039
    },
    {
      "entropy": 0.720642976462841,
      "epoch": 0.7397312263874273,
      "grad_norm": 0.15850234031677246,
      "learning_rate": 7.534578464746834e-05,
      "loss": 0.7235,
      "mean_token_accuracy": 0.7899790219962597,
      "num_tokens": 256607281.0,
      "step": 8040
    },
    {
      "entropy": 0.6390637140721083,
      "epoch": 0.7398232327588685,
      "grad_norm": 0.15619851648807526,
      "learning_rate": 7.534271782132671e-05,
      "loss": 0.5987,
      "mean_token_accuracy": 0.8221164904534817,
      "num_tokens": 256639552.0,
      "step": 8041
    },
    {
      "entropy": 0.8265862008556724,
      "epoch": 0.7399152391303098,
      "grad_norm": 0.16042454540729523,
      "learning_rate": 7.533965099518509e-05,
      "loss": 0.8071,
      "mean_token_accuracy": 0.7679643966257572,
      "num_tokens": 256672290.0,
      "step": 8042
    },
    {
      "entropy": 0.838137798011303,
      "epoch": 0.740007245501751,
      "grad_norm": 0.16002242267131805,
      "learning_rate": 7.533658416904346e-05,
      "loss": 0.8287,
      "mean_token_accuracy": 0.7595362402498722,
      "num_tokens": 256703642.0,
      "step": 8043
    },
    {
      "entropy": 0.8139375662431121,
      "epoch": 0.7400992518731923,
      "grad_norm": 0.16806693375110626,
      "learning_rate": 7.533351734290184e-05,
      "loss": 0.8197,
      "mean_token_accuracy": 0.7625448331236839,
      "num_tokens": 256736349.0,
      "step": 8044
    },
    {
      "entropy": 0.7596361208707094,
      "epoch": 0.7401912582446334,
      "grad_norm": 0.16122758388519287,
      "learning_rate": 7.533045051676021e-05,
      "loss": 0.7695,
      "mean_token_accuracy": 0.7789523676037788,
      "num_tokens": 256767949.0,
      "step": 8045
    },
    {
      "entropy": 0.7712306492030621,
      "epoch": 0.7402832646160746,
      "grad_norm": 0.16040503978729248,
      "learning_rate": 7.532738369061858e-05,
      "loss": 0.7767,
      "mean_token_accuracy": 0.7754655480384827,
      "num_tokens": 256800470.0,
      "step": 8046
    },
    {
      "entropy": 0.7405154220759869,
      "epoch": 0.7403752709875159,
      "grad_norm": 0.1595177948474884,
      "learning_rate": 7.532431686447696e-05,
      "loss": 0.725,
      "mean_token_accuracy": 0.7847346588969231,
      "num_tokens": 256831585.0,
      "step": 8047
    },
    {
      "entropy": 0.6670710239559412,
      "epoch": 0.7404672773589571,
      "grad_norm": 0.1477688103914261,
      "learning_rate": 7.532125003833534e-05,
      "loss": 0.651,
      "mean_token_accuracy": 0.8035143353044987,
      "num_tokens": 256863371.0,
      "step": 8048
    },
    {
      "entropy": 0.6169558223336935,
      "epoch": 0.7405592837303984,
      "grad_norm": 0.1575191617012024,
      "learning_rate": 7.53181832121937e-05,
      "loss": 0.6042,
      "mean_token_accuracy": 0.8210135735571384,
      "num_tokens": 256895673.0,
      "step": 8049
    },
    {
      "entropy": 0.7091694176197052,
      "epoch": 0.7406512901018395,
      "grad_norm": 0.1559368222951889,
      "learning_rate": 7.531511638605207e-05,
      "loss": 0.6977,
      "mean_token_accuracy": 0.7923090308904648,
      "num_tokens": 256927761.0,
      "step": 8050
    },
    {
      "entropy": 0.7540697157382965,
      "epoch": 0.7407432964732807,
      "grad_norm": 0.15680545568466187,
      "learning_rate": 7.531204955991044e-05,
      "loss": 0.7554,
      "mean_token_accuracy": 0.7725899890065193,
      "num_tokens": 256959785.0,
      "step": 8051
    },
    {
      "entropy": 0.842053011059761,
      "epoch": 0.740835302844722,
      "grad_norm": 0.16042739152908325,
      "learning_rate": 7.530898273376882e-05,
      "loss": 0.825,
      "mean_token_accuracy": 0.7556627057492733,
      "num_tokens": 256990868.0,
      "step": 8052
    },
    {
      "entropy": 0.7356868609786034,
      "epoch": 0.7409273092161632,
      "grad_norm": 0.15772175788879395,
      "learning_rate": 7.53059159076272e-05,
      "loss": 0.7257,
      "mean_token_accuracy": 0.7874976508319378,
      "num_tokens": 257023130.0,
      "step": 8053
    },
    {
      "entropy": 0.7597192376852036,
      "epoch": 0.7410193155876045,
      "grad_norm": 0.15751828253269196,
      "learning_rate": 7.530284908148557e-05,
      "loss": 0.738,
      "mean_token_accuracy": 0.783134438097477,
      "num_tokens": 257054938.0,
      "step": 8054
    },
    {
      "entropy": 0.8692819997668266,
      "epoch": 0.7411113219590456,
      "grad_norm": 0.16339243948459625,
      "learning_rate": 7.529978225534395e-05,
      "loss": 0.8647,
      "mean_token_accuracy": 0.7489371113479137,
      "num_tokens": 257086892.0,
      "step": 8055
    },
    {
      "entropy": 0.7597337625920773,
      "epoch": 0.7412033283304869,
      "grad_norm": 0.15756948292255402,
      "learning_rate": 7.529671542920232e-05,
      "loss": 0.7436,
      "mean_token_accuracy": 0.7854396626353264,
      "num_tokens": 257118504.0,
      "step": 8056
    },
    {
      "entropy": 0.7140550184994936,
      "epoch": 0.7412953347019281,
      "grad_norm": 0.15243084728717804,
      "learning_rate": 7.529364860306069e-05,
      "loss": 0.7,
      "mean_token_accuracy": 0.7875637002289295,
      "num_tokens": 257151007.0,
      "step": 8057
    },
    {
      "entropy": 0.8362363632768393,
      "epoch": 0.7413873410733693,
      "grad_norm": 0.16132640838623047,
      "learning_rate": 7.529058177691907e-05,
      "loss": 0.8309,
      "mean_token_accuracy": 0.7603454105556011,
      "num_tokens": 257183137.0,
      "step": 8058
    },
    {
      "entropy": 0.6719320118427277,
      "epoch": 0.7414793474448106,
      "grad_norm": 0.15549570322036743,
      "learning_rate": 7.528751495077745e-05,
      "loss": 0.6632,
      "mean_token_accuracy": 0.8039439544081688,
      "num_tokens": 257215347.0,
      "step": 8059
    },
    {
      "entropy": 0.682900682091713,
      "epoch": 0.7415713538162517,
      "grad_norm": 0.14744041860103607,
      "learning_rate": 7.528444812463582e-05,
      "loss": 0.6738,
      "mean_token_accuracy": 0.8027601353824139,
      "num_tokens": 257248115.0,
      "step": 8060
    },
    {
      "entropy": 0.7384104952216148,
      "epoch": 0.741663360187693,
      "grad_norm": 0.15702170133590698,
      "learning_rate": 7.528138129849419e-05,
      "loss": 0.7383,
      "mean_token_accuracy": 0.7791250422596931,
      "num_tokens": 257280243.0,
      "step": 8061
    },
    {
      "entropy": 0.6402052622288465,
      "epoch": 0.7417553665591342,
      "grad_norm": 0.1548103392124176,
      "learning_rate": 7.527831447235256e-05,
      "loss": 0.6235,
      "mean_token_accuracy": 0.8143100328743458,
      "num_tokens": 257311996.0,
      "step": 8062
    },
    {
      "entropy": 0.8281719572842121,
      "epoch": 0.7418473729305755,
      "grad_norm": 0.1553432047367096,
      "learning_rate": 7.527524764621094e-05,
      "loss": 0.8098,
      "mean_token_accuracy": 0.7612727954983711,
      "num_tokens": 257344129.0,
      "step": 8063
    },
    {
      "entropy": 0.7759141623973846,
      "epoch": 0.7419393793020167,
      "grad_norm": 0.15715515613555908,
      "learning_rate": 7.527218082006932e-05,
      "loss": 0.7596,
      "mean_token_accuracy": 0.7739205621182919,
      "num_tokens": 257375907.0,
      "step": 8064
    },
    {
      "entropy": 0.641970306634903,
      "epoch": 0.7420313856734578,
      "grad_norm": 0.15255849063396454,
      "learning_rate": 7.526911399392769e-05,
      "loss": 0.6294,
      "mean_token_accuracy": 0.8067746721208096,
      "num_tokens": 257407969.0,
      "step": 8065
    },
    {
      "entropy": 0.6814655028283596,
      "epoch": 0.7421233920448991,
      "grad_norm": 0.1522647738456726,
      "learning_rate": 7.526604716778606e-05,
      "loss": 0.6614,
      "mean_token_accuracy": 0.7975085452198982,
      "num_tokens": 257440737.0,
      "step": 8066
    },
    {
      "entropy": 0.7718297075480223,
      "epoch": 0.7422153984163403,
      "grad_norm": 0.15569531917572021,
      "learning_rate": 7.526298034164444e-05,
      "loss": 0.7663,
      "mean_token_accuracy": 0.7768466845154762,
      "num_tokens": 257472610.0,
      "step": 8067
    },
    {
      "entropy": 0.6676082387566566,
      "epoch": 0.7423074047877816,
      "grad_norm": 0.15129660069942474,
      "learning_rate": 7.52599135155028e-05,
      "loss": 0.6596,
      "mean_token_accuracy": 0.8050977177917957,
      "num_tokens": 257504799.0,
      "step": 8068
    },
    {
      "entropy": 0.7652319935150445,
      "epoch": 0.7423994111592228,
      "grad_norm": 0.14887267351150513,
      "learning_rate": 7.525684668936119e-05,
      "loss": 0.7638,
      "mean_token_accuracy": 0.7792015969753265,
      "num_tokens": 257536945.0,
      "step": 8069
    },
    {
      "entropy": 0.8054147539660335,
      "epoch": 0.7424914175306639,
      "grad_norm": 0.16086208820343018,
      "learning_rate": 7.525377986321955e-05,
      "loss": 0.7959,
      "mean_token_accuracy": 0.7661052606999874,
      "num_tokens": 257568884.0,
      "step": 8070
    },
    {
      "entropy": 0.7449812516570091,
      "epoch": 0.7425834239021052,
      "grad_norm": 0.16592009365558624,
      "learning_rate": 7.525071303707794e-05,
      "loss": 0.7478,
      "mean_token_accuracy": 0.7774038128554821,
      "num_tokens": 257599887.0,
      "step": 8071
    },
    {
      "entropy": 0.6622294886037707,
      "epoch": 0.7426754302735464,
      "grad_norm": 0.16200889647006989,
      "learning_rate": 7.52476462109363e-05,
      "loss": 0.652,
      "mean_token_accuracy": 0.8036642335355282,
      "num_tokens": 257631740.0,
      "step": 8072
    },
    {
      "entropy": 0.7042396888136864,
      "epoch": 0.7427674366449877,
      "grad_norm": 0.15181110799312592,
      "learning_rate": 7.524457938479467e-05,
      "loss": 0.6978,
      "mean_token_accuracy": 0.7945433855056763,
      "num_tokens": 257664099.0,
      "step": 8073
    },
    {
      "entropy": 0.8299365118145943,
      "epoch": 0.7428594430164289,
      "grad_norm": 0.1571173071861267,
      "learning_rate": 7.524151255865305e-05,
      "loss": 0.829,
      "mean_token_accuracy": 0.7599273920059204,
      "num_tokens": 257696366.0,
      "step": 8074
    },
    {
      "entropy": 0.6925945971161127,
      "epoch": 0.74295144938787,
      "grad_norm": 0.15408210456371307,
      "learning_rate": 7.523844573251143e-05,
      "loss": 0.6827,
      "mean_token_accuracy": 0.7998567670583725,
      "num_tokens": 257728824.0,
      "step": 8075
    },
    {
      "entropy": 0.7074849177151918,
      "epoch": 0.7430434557593113,
      "grad_norm": 0.15489710867404938,
      "learning_rate": 7.52353789063698e-05,
      "loss": 0.7093,
      "mean_token_accuracy": 0.7920681051909924,
      "num_tokens": 257761296.0,
      "step": 8076
    },
    {
      "entropy": 0.8453859444707632,
      "epoch": 0.7431354621307525,
      "grad_norm": 0.16741155087947845,
      "learning_rate": 7.523231208022817e-05,
      "loss": 0.8664,
      "mean_token_accuracy": 0.7510685995221138,
      "num_tokens": 257793626.0,
      "step": 8077
    },
    {
      "entropy": 0.805937223136425,
      "epoch": 0.7432274685021938,
      "grad_norm": 0.16141533851623535,
      "learning_rate": 7.522924525408655e-05,
      "loss": 0.7993,
      "mean_token_accuracy": 0.7625425793230534,
      "num_tokens": 257825755.0,
      "step": 8078
    },
    {
      "entropy": 0.8037840500473976,
      "epoch": 0.743319474873635,
      "grad_norm": 0.16020327806472778,
      "learning_rate": 7.522617842794493e-05,
      "loss": 0.7885,
      "mean_token_accuracy": 0.7725898250937462,
      "num_tokens": 257857298.0,
      "step": 8079
    },
    {
      "entropy": 0.756387285888195,
      "epoch": 0.7434114812450762,
      "grad_norm": 0.15552634000778198,
      "learning_rate": 7.52231116018033e-05,
      "loss": 0.7234,
      "mean_token_accuracy": 0.7816651649773121,
      "num_tokens": 257888592.0,
      "step": 8080
    },
    {
      "entropy": 0.7709727101027966,
      "epoch": 0.7435034876165174,
      "grad_norm": 0.15203812718391418,
      "learning_rate": 7.522004477566167e-05,
      "loss": 0.7615,
      "mean_token_accuracy": 0.7779220268130302,
      "num_tokens": 257920995.0,
      "step": 8081
    },
    {
      "entropy": 0.7425828594714403,
      "epoch": 0.7435954939879587,
      "grad_norm": 0.1570538431406021,
      "learning_rate": 7.521697794952004e-05,
      "loss": 0.7255,
      "mean_token_accuracy": 0.7850154377520084,
      "num_tokens": 257953275.0,
      "step": 8082
    },
    {
      "entropy": 0.6321696080267429,
      "epoch": 0.7436875003593999,
      "grad_norm": 0.1499525010585785,
      "learning_rate": 7.521391112337842e-05,
      "loss": 0.6091,
      "mean_token_accuracy": 0.8271604552865028,
      "num_tokens": 257985599.0,
      "step": 8083
    },
    {
      "entropy": 0.7483558226376772,
      "epoch": 0.7437795067308411,
      "grad_norm": 0.1541164219379425,
      "learning_rate": 7.52108442972368e-05,
      "loss": 0.7395,
      "mean_token_accuracy": 0.7821946479380131,
      "num_tokens": 258017979.0,
      "step": 8084
    },
    {
      "entropy": 0.7555364649742842,
      "epoch": 0.7438715131022823,
      "grad_norm": 0.1590879261493683,
      "learning_rate": 7.520777747109517e-05,
      "loss": 0.7363,
      "mean_token_accuracy": 0.7879788242280483,
      "num_tokens": 258050624.0,
      "step": 8085
    },
    {
      "entropy": 0.85886000841856,
      "epoch": 0.7439635194737235,
      "grad_norm": 0.16207678616046906,
      "learning_rate": 7.520471064495355e-05,
      "loss": 0.8615,
      "mean_token_accuracy": 0.7507200948894024,
      "num_tokens": 258083095.0,
      "step": 8086
    },
    {
      "entropy": 0.6464110808447003,
      "epoch": 0.7440555258451648,
      "grad_norm": 0.14414264261722565,
      "learning_rate": 7.520164381881192e-05,
      "loss": 0.6219,
      "mean_token_accuracy": 0.8158505409955978,
      "num_tokens": 258115069.0,
      "step": 8087
    },
    {
      "entropy": 0.596128542907536,
      "epoch": 0.744147532216606,
      "grad_norm": 0.14355991780757904,
      "learning_rate": 7.519857699267028e-05,
      "loss": 0.5849,
      "mean_token_accuracy": 0.8217706419527531,
      "num_tokens": 258147492.0,
      "step": 8088
    },
    {
      "entropy": 0.6372934002429247,
      "epoch": 0.7442395385880473,
      "grad_norm": 0.15444093942642212,
      "learning_rate": 7.519551016652867e-05,
      "loss": 0.6106,
      "mean_token_accuracy": 0.8113414607942104,
      "num_tokens": 258179545.0,
      "step": 8089
    },
    {
      "entropy": 0.7854831535369158,
      "epoch": 0.7443315449594884,
      "grad_norm": 0.161748006939888,
      "learning_rate": 7.519244334038705e-05,
      "loss": 0.7721,
      "mean_token_accuracy": 0.768597386777401,
      "num_tokens": 258211379.0,
      "step": 8090
    },
    {
      "entropy": 0.652193927206099,
      "epoch": 0.7444235513309296,
      "grad_norm": 0.15270362794399261,
      "learning_rate": 7.518937651424541e-05,
      "loss": 0.6353,
      "mean_token_accuracy": 0.810158271342516,
      "num_tokens": 258243417.0,
      "step": 8091
    },
    {
      "entropy": 0.7555181086063385,
      "epoch": 0.7445155577023709,
      "grad_norm": 0.16416305303573608,
      "learning_rate": 7.518630968810378e-05,
      "loss": 0.7235,
      "mean_token_accuracy": 0.7894778102636337,
      "num_tokens": 258275436.0,
      "step": 8092
    },
    {
      "entropy": 0.7627019006758928,
      "epoch": 0.7446075640738121,
      "grad_norm": 0.15161822736263275,
      "learning_rate": 7.518324286196215e-05,
      "loss": 0.7525,
      "mean_token_accuracy": 0.7756087072193623,
      "num_tokens": 258307044.0,
      "step": 8093
    },
    {
      "entropy": 0.7639852222055197,
      "epoch": 0.7446995704452534,
      "grad_norm": 0.15704822540283203,
      "learning_rate": 7.518017603582053e-05,
      "loss": 0.7526,
      "mean_token_accuracy": 0.7787619829177856,
      "num_tokens": 258339073.0,
      "step": 8094
    },
    {
      "entropy": 0.7536395639181137,
      "epoch": 0.7447915768166946,
      "grad_norm": 0.15004168450832367,
      "learning_rate": 7.517710920967891e-05,
      "loss": 0.7391,
      "mean_token_accuracy": 0.7810014002025127,
      "num_tokens": 258370985.0,
      "step": 8095
    },
    {
      "entropy": 0.7362917521968484,
      "epoch": 0.7448835831881357,
      "grad_norm": 0.15603694319725037,
      "learning_rate": 7.517404238353728e-05,
      "loss": 0.7295,
      "mean_token_accuracy": 0.7845744341611862,
      "num_tokens": 258402933.0,
      "step": 8096
    },
    {
      "entropy": 0.7420861031860113,
      "epoch": 0.744975589559577,
      "grad_norm": 0.15906183421611786,
      "learning_rate": 7.517097555739565e-05,
      "loss": 0.7372,
      "mean_token_accuracy": 0.7886726520955563,
      "num_tokens": 258434524.0,
      "step": 8097
    },
    {
      "entropy": 0.7217427957803011,
      "epoch": 0.7450675959310182,
      "grad_norm": 0.16248778998851776,
      "learning_rate": 7.516790873125403e-05,
      "loss": 0.7008,
      "mean_token_accuracy": 0.7942018769681454,
      "num_tokens": 258466097.0,
      "step": 8098
    },
    {
      "entropy": 0.8063083682209253,
      "epoch": 0.7451596023024595,
      "grad_norm": 0.16111858189105988,
      "learning_rate": 7.51648419051124e-05,
      "loss": 0.8109,
      "mean_token_accuracy": 0.759985838085413,
      "num_tokens": 258497597.0,
      "step": 8099
    },
    {
      "entropy": 0.718004371970892,
      "epoch": 0.7452516086739007,
      "grad_norm": 0.1605186015367508,
      "learning_rate": 7.516177507897078e-05,
      "loss": 0.7173,
      "mean_token_accuracy": 0.7915062084794044,
      "num_tokens": 258529871.0,
      "step": 8100
    },
    {
      "entropy": 0.662064254283905,
      "epoch": 0.7453436150453419,
      "grad_norm": 0.15533673763275146,
      "learning_rate": 7.515870825282915e-05,
      "loss": 0.6528,
      "mean_token_accuracy": 0.8078296110033989,
      "num_tokens": 258562244.0,
      "step": 8101
    },
    {
      "entropy": 0.8089977037161589,
      "epoch": 0.7454356214167831,
      "grad_norm": 0.15768994390964508,
      "learning_rate": 7.515564142668753e-05,
      "loss": 0.7954,
      "mean_token_accuracy": 0.764208696782589,
      "num_tokens": 258594906.0,
      "step": 8102
    },
    {
      "entropy": 0.7692747730761766,
      "epoch": 0.7455276277882243,
      "grad_norm": 0.15614159405231476,
      "learning_rate": 7.51525746005459e-05,
      "loss": 0.7461,
      "mean_token_accuracy": 0.7799416519701481,
      "num_tokens": 258626128.0,
      "step": 8103
    },
    {
      "entropy": 0.7267665006220341,
      "epoch": 0.7456196341596656,
      "grad_norm": 0.15462851524353027,
      "learning_rate": 7.514950777440426e-05,
      "loss": 0.7195,
      "mean_token_accuracy": 0.7827955558896065,
      "num_tokens": 258657278.0,
      "step": 8104
    },
    {
      "entropy": 0.8321237713098526,
      "epoch": 0.7457116405311068,
      "grad_norm": 0.16024865210056305,
      "learning_rate": 7.514644094826265e-05,
      "loss": 0.8261,
      "mean_token_accuracy": 0.7566846162080765,
      "num_tokens": 258689527.0,
      "step": 8105
    },
    {
      "entropy": 0.6344469245523214,
      "epoch": 0.745803646902548,
      "grad_norm": 0.14674749970436096,
      "learning_rate": 7.514337412212103e-05,
      "loss": 0.622,
      "mean_token_accuracy": 0.811629731208086,
      "num_tokens": 258721898.0,
      "step": 8106
    },
    {
      "entropy": 0.6775523666292429,
      "epoch": 0.7458956532739892,
      "grad_norm": 0.15999835729599,
      "learning_rate": 7.51403072959794e-05,
      "loss": 0.6796,
      "mean_token_accuracy": 0.7979625687003136,
      "num_tokens": 258753328.0,
      "step": 8107
    },
    {
      "entropy": 0.7211298067122698,
      "epoch": 0.7459876596454305,
      "grad_norm": 0.15650928020477295,
      "learning_rate": 7.513724046983776e-05,
      "loss": 0.7188,
      "mean_token_accuracy": 0.7853905782103539,
      "num_tokens": 258785575.0,
      "step": 8108
    },
    {
      "entropy": 0.7258724607527256,
      "epoch": 0.7460796660168717,
      "grad_norm": 0.15271170437335968,
      "learning_rate": 7.513417364369613e-05,
      "loss": 0.7154,
      "mean_token_accuracy": 0.7897887006402016,
      "num_tokens": 258817952.0,
      "step": 8109
    },
    {
      "entropy": 0.705657010897994,
      "epoch": 0.746171672388313,
      "grad_norm": 0.14762578904628754,
      "learning_rate": 7.513110681755453e-05,
      "loss": 0.6902,
      "mean_token_accuracy": 0.7914130389690399,
      "num_tokens": 258850153.0,
      "step": 8110
    },
    {
      "entropy": 0.6190125606954098,
      "epoch": 0.7462636787597541,
      "grad_norm": 0.14790643751621246,
      "learning_rate": 7.51280399914129e-05,
      "loss": 0.5991,
      "mean_token_accuracy": 0.8204900063574314,
      "num_tokens": 258882704.0,
      "step": 8111
    },
    {
      "entropy": 0.6581332441419363,
      "epoch": 0.7463556851311953,
      "grad_norm": 0.16104981303215027,
      "learning_rate": 7.512497316527126e-05,
      "loss": 0.6469,
      "mean_token_accuracy": 0.8084552623331547,
      "num_tokens": 258915065.0,
      "step": 8112
    },
    {
      "entropy": 0.8364161737263203,
      "epoch": 0.7464476915026366,
      "grad_norm": 0.15849189460277557,
      "learning_rate": 7.512190633912963e-05,
      "loss": 0.8229,
      "mean_token_accuracy": 0.7608613818883896,
      "num_tokens": 258947318.0,
      "step": 8113
    },
    {
      "entropy": 0.6827929280698299,
      "epoch": 0.7465396978740778,
      "grad_norm": 0.15612095594406128,
      "learning_rate": 7.511883951298801e-05,
      "loss": 0.6848,
      "mean_token_accuracy": 0.7969828024506569,
      "num_tokens": 258979869.0,
      "step": 8114
    },
    {
      "entropy": 0.6751663982868195,
      "epoch": 0.7466317042455191,
      "grad_norm": 0.15653567016124725,
      "learning_rate": 7.511577268684639e-05,
      "loss": 0.655,
      "mean_token_accuracy": 0.8063732832670212,
      "num_tokens": 259011666.0,
      "step": 8115
    },
    {
      "entropy": 0.7977599520236254,
      "epoch": 0.7467237106169602,
      "grad_norm": 0.16816234588623047,
      "learning_rate": 7.511270586070476e-05,
      "loss": 0.7935,
      "mean_token_accuracy": 0.7684180364012718,
      "num_tokens": 259043581.0,
      "step": 8116
    },
    {
      "entropy": 0.6802265420556068,
      "epoch": 0.7468157169884014,
      "grad_norm": 0.15422099828720093,
      "learning_rate": 7.510963903456314e-05,
      "loss": 0.6937,
      "mean_token_accuracy": 0.7915398590266705,
      "num_tokens": 259075367.0,
      "step": 8117
    },
    {
      "entropy": 0.740668329410255,
      "epoch": 0.7469077233598427,
      "grad_norm": 0.16493543982505798,
      "learning_rate": 7.510657220842151e-05,
      "loss": 0.7377,
      "mean_token_accuracy": 0.7840865850448608,
      "num_tokens": 259107538.0,
      "step": 8118
    },
    {
      "entropy": 0.6747484728693962,
      "epoch": 0.7469997297312839,
      "grad_norm": 0.1540052741765976,
      "learning_rate": 7.510350538227988e-05,
      "loss": 0.648,
      "mean_token_accuracy": 0.8060995526611805,
      "num_tokens": 259140138.0,
      "step": 8119
    },
    {
      "entropy": 0.6715555675327778,
      "epoch": 0.7470917361027252,
      "grad_norm": 0.15101535618305206,
      "learning_rate": 7.510043855613826e-05,
      "loss": 0.6628,
      "mean_token_accuracy": 0.8004560135304928,
      "num_tokens": 259172249.0,
      "step": 8120
    },
    {
      "entropy": 0.6853430271148682,
      "epoch": 0.7471837424741663,
      "grad_norm": 0.1607171893119812,
      "learning_rate": 7.509737172999664e-05,
      "loss": 0.6897,
      "mean_token_accuracy": 0.7959736362099648,
      "num_tokens": 259204263.0,
      "step": 8121
    },
    {
      "entropy": 0.6943592485040426,
      "epoch": 0.7472757488456075,
      "grad_norm": 0.15039412677288055,
      "learning_rate": 7.509430490385501e-05,
      "loss": 0.6727,
      "mean_token_accuracy": 0.7960680723190308,
      "num_tokens": 259236123.0,
      "step": 8122
    },
    {
      "entropy": 0.7259142706170678,
      "epoch": 0.7473677552170488,
      "grad_norm": 0.15020035207271576,
      "learning_rate": 7.509123807771338e-05,
      "loss": 0.711,
      "mean_token_accuracy": 0.7863413281738758,
      "num_tokens": 259267971.0,
      "step": 8123
    },
    {
      "entropy": 0.7711118664592505,
      "epoch": 0.74745976158849,
      "grad_norm": 0.1562512218952179,
      "learning_rate": 7.508817125157174e-05,
      "loss": 0.7516,
      "mean_token_accuracy": 0.7778545208275318,
      "num_tokens": 259299611.0,
      "step": 8124
    },
    {
      "entropy": 0.6396421398967505,
      "epoch": 0.7475517679599313,
      "grad_norm": 0.1627047061920166,
      "learning_rate": 7.508510442543013e-05,
      "loss": 0.6204,
      "mean_token_accuracy": 0.8130744248628616,
      "num_tokens": 259332222.0,
      "step": 8125
    },
    {
      "entropy": 0.7177404016256332,
      "epoch": 0.7476437743313724,
      "grad_norm": 0.15387193858623505,
      "learning_rate": 7.508203759928851e-05,
      "loss": 0.7099,
      "mean_token_accuracy": 0.7916956357657909,
      "num_tokens": 259364587.0,
      "step": 8126
    },
    {
      "entropy": 0.7154343239963055,
      "epoch": 0.7477357807028137,
      "grad_norm": 0.15598678588867188,
      "learning_rate": 7.507897077314688e-05,
      "loss": 0.703,
      "mean_token_accuracy": 0.7937798984348774,
      "num_tokens": 259396189.0,
      "step": 8127
    },
    {
      "entropy": 0.7844130285084248,
      "epoch": 0.7478277870742549,
      "grad_norm": 0.16960446536540985,
      "learning_rate": 7.507590394700524e-05,
      "loss": 0.7876,
      "mean_token_accuracy": 0.7684187963604927,
      "num_tokens": 259427767.0,
      "step": 8128
    },
    {
      "entropy": 0.8070240598171949,
      "epoch": 0.7479197934456961,
      "grad_norm": 0.15786407887935638,
      "learning_rate": 7.507283712086362e-05,
      "loss": 0.826,
      "mean_token_accuracy": 0.7596166245639324,
      "num_tokens": 259459354.0,
      "step": 8129
    },
    {
      "entropy": 0.7648723665624857,
      "epoch": 0.7480117998171374,
      "grad_norm": 0.1556466668844223,
      "learning_rate": 7.506977029472199e-05,
      "loss": 0.7575,
      "mean_token_accuracy": 0.7757137082517147,
      "num_tokens": 259490890.0,
      "step": 8130
    },
    {
      "entropy": 0.8468314036726952,
      "epoch": 0.7481038061885785,
      "grad_norm": 0.1684299260377884,
      "learning_rate": 7.506670346858037e-05,
      "loss": 0.85,
      "mean_token_accuracy": 0.7539239972829819,
      "num_tokens": 259522112.0,
      "step": 8131
    },
    {
      "entropy": 0.7575648091733456,
      "epoch": 0.7481958125600198,
      "grad_norm": 0.1588703840970993,
      "learning_rate": 7.506363664243874e-05,
      "loss": 0.7344,
      "mean_token_accuracy": 0.777477715164423,
      "num_tokens": 259554133.0,
      "step": 8132
    },
    {
      "entropy": 0.6932330839335918,
      "epoch": 0.748287818931461,
      "grad_norm": 0.16398870944976807,
      "learning_rate": 7.506056981629712e-05,
      "loss": 0.6801,
      "mean_token_accuracy": 0.8010055013000965,
      "num_tokens": 259585677.0,
      "step": 8133
    },
    {
      "entropy": 0.716299258172512,
      "epoch": 0.7483798253029023,
      "grad_norm": 0.1555825173854828,
      "learning_rate": 7.505750299015549e-05,
      "loss": 0.6967,
      "mean_token_accuracy": 0.7933918610215187,
      "num_tokens": 259618016.0,
      "step": 8134
    },
    {
      "entropy": 0.8628173861652613,
      "epoch": 0.7484718316743435,
      "grad_norm": 0.15843883156776428,
      "learning_rate": 7.505443616401386e-05,
      "loss": 0.8408,
      "mean_token_accuracy": 0.7538042850792408,
      "num_tokens": 259650015.0,
      "step": 8135
    },
    {
      "entropy": 0.7312694825232029,
      "epoch": 0.7485638380457846,
      "grad_norm": 0.1517000049352646,
      "learning_rate": 7.505136933787224e-05,
      "loss": 0.7226,
      "mean_token_accuracy": 0.7851987443864346,
      "num_tokens": 259682144.0,
      "step": 8136
    },
    {
      "entropy": 0.7021554093807936,
      "epoch": 0.7486558444172259,
      "grad_norm": 0.1586509346961975,
      "learning_rate": 7.504830251173062e-05,
      "loss": 0.6802,
      "mean_token_accuracy": 0.8000686094164848,
      "num_tokens": 259713950.0,
      "step": 8137
    },
    {
      "entropy": 0.7805796656757593,
      "epoch": 0.7487478507886671,
      "grad_norm": 0.1636943370103836,
      "learning_rate": 7.504523568558899e-05,
      "loss": 0.7764,
      "mean_token_accuracy": 0.7721747048199177,
      "num_tokens": 259746139.0,
      "step": 8138
    },
    {
      "entropy": 0.8571721538901329,
      "epoch": 0.7488398571601084,
      "grad_norm": 0.15946200489997864,
      "learning_rate": 7.504216885944736e-05,
      "loss": 0.8484,
      "mean_token_accuracy": 0.757525697350502,
      "num_tokens": 259777280.0,
      "step": 8139
    },
    {
      "entropy": 0.865317665040493,
      "epoch": 0.7489318635315496,
      "grad_norm": 0.16376639902591705,
      "learning_rate": 7.503910203330573e-05,
      "loss": 0.8605,
      "mean_token_accuracy": 0.7500359676778316,
      "num_tokens": 259809941.0,
      "step": 8140
    },
    {
      "entropy": 0.7168524470180273,
      "epoch": 0.7490238699029907,
      "grad_norm": 0.15246321260929108,
      "learning_rate": 7.50360352071641e-05,
      "loss": 0.7067,
      "mean_token_accuracy": 0.7924841865897179,
      "num_tokens": 259841772.0,
      "step": 8141
    },
    {
      "entropy": 0.73684087023139,
      "epoch": 0.749115876274432,
      "grad_norm": 0.15200954675674438,
      "learning_rate": 7.503296838102249e-05,
      "loss": 0.7131,
      "mean_token_accuracy": 0.7905394658446312,
      "num_tokens": 259873987.0,
      "step": 8142
    },
    {
      "entropy": 0.7072981894016266,
      "epoch": 0.7492078826458732,
      "grad_norm": 0.15270838141441345,
      "learning_rate": 7.502990155488086e-05,
      "loss": 0.7007,
      "mean_token_accuracy": 0.7945046313107014,
      "num_tokens": 259906028.0,
      "step": 8143
    },
    {
      "entropy": 0.7505559772253036,
      "epoch": 0.7492998890173145,
      "grad_norm": 0.15338240563869476,
      "learning_rate": 7.502683472873922e-05,
      "loss": 0.7404,
      "mean_token_accuracy": 0.7815869003534317,
      "num_tokens": 259938244.0,
      "step": 8144
    },
    {
      "entropy": 0.791236974298954,
      "epoch": 0.7493918953887557,
      "grad_norm": 0.16219687461853027,
      "learning_rate": 7.50237679025976e-05,
      "loss": 0.7683,
      "mean_token_accuracy": 0.7779770530760288,
      "num_tokens": 259970699.0,
      "step": 8145
    },
    {
      "entropy": 0.7768952641636133,
      "epoch": 0.7494839017601969,
      "grad_norm": 0.15207405388355255,
      "learning_rate": 7.502070107645597e-05,
      "loss": 0.7567,
      "mean_token_accuracy": 0.7748746387660503,
      "num_tokens": 260002375.0,
      "step": 8146
    },
    {
      "entropy": 0.761058734729886,
      "epoch": 0.7495759081316381,
      "grad_norm": 0.15636028349399567,
      "learning_rate": 7.501763425031435e-05,
      "loss": 0.7551,
      "mean_token_accuracy": 0.7784738317131996,
      "num_tokens": 260034864.0,
      "step": 8147
    },
    {
      "entropy": 0.6940254550427198,
      "epoch": 0.7496679145030793,
      "grad_norm": 0.1607494354248047,
      "learning_rate": 7.501456742417274e-05,
      "loss": 0.6746,
      "mean_token_accuracy": 0.805229764431715,
      "num_tokens": 260066891.0,
      "step": 8148
    },
    {
      "entropy": 0.5877781063318253,
      "epoch": 0.7497599208745206,
      "grad_norm": 0.165671244263649,
      "learning_rate": 7.50115005980311e-05,
      "loss": 0.5816,
      "mean_token_accuracy": 0.8223995119333267,
      "num_tokens": 260099037.0,
      "step": 8149
    },
    {
      "entropy": 0.7518214713782072,
      "epoch": 0.7498519272459618,
      "grad_norm": 0.15395142138004303,
      "learning_rate": 7.500843377188947e-05,
      "loss": 0.7357,
      "mean_token_accuracy": 0.7825702764093876,
      "num_tokens": 260130410.0,
      "step": 8150
    },
    {
      "entropy": 0.7256647404283285,
      "epoch": 0.749943933617403,
      "grad_norm": 0.16573216021060944,
      "learning_rate": 7.500536694574784e-05,
      "loss": 0.7186,
      "mean_token_accuracy": 0.7893727868795395,
      "num_tokens": 260162500.0,
      "step": 8151
    },
    {
      "entropy": 0.6716385390609503,
      "epoch": 0.7500359399888442,
      "grad_norm": 0.1532004326581955,
      "learning_rate": 7.500230011960623e-05,
      "loss": 0.6822,
      "mean_token_accuracy": 0.800146572291851,
      "num_tokens": 260194256.0,
      "step": 8152
    },
    {
      "entropy": 0.7212795987725258,
      "epoch": 0.7501279463602855,
      "grad_norm": 0.15752340853214264,
      "learning_rate": 7.49992332934646e-05,
      "loss": 0.706,
      "mean_token_accuracy": 0.78774668648839,
      "num_tokens": 260226464.0,
      "step": 8153
    },
    {
      "entropy": 0.5704594682902098,
      "epoch": 0.7502199527317267,
      "grad_norm": 0.15200984477996826,
      "learning_rate": 7.499616646732297e-05,
      "loss": 0.5553,
      "mean_token_accuracy": 0.835185531526804,
      "num_tokens": 260258919.0,
      "step": 8154
    },
    {
      "entropy": 0.6271281950175762,
      "epoch": 0.750311959103168,
      "grad_norm": 0.15940675139427185,
      "learning_rate": 7.499309964118134e-05,
      "loss": 0.619,
      "mean_token_accuracy": 0.8126987777650356,
      "num_tokens": 260290788.0,
      "step": 8155
    },
    {
      "entropy": 0.654906889423728,
      "epoch": 0.7504039654746091,
      "grad_norm": 0.15506801009178162,
      "learning_rate": 7.499003281503972e-05,
      "loss": 0.6501,
      "mean_token_accuracy": 0.8106148317456245,
      "num_tokens": 260323340.0,
      "step": 8156
    },
    {
      "entropy": 0.7626913078129292,
      "epoch": 0.7504959718460503,
      "grad_norm": 0.15733647346496582,
      "learning_rate": 7.49869659888981e-05,
      "loss": 0.771,
      "mean_token_accuracy": 0.7758265435695648,
      "num_tokens": 260355368.0,
      "step": 8157
    },
    {
      "entropy": 0.705782238394022,
      "epoch": 0.7505879782174916,
      "grad_norm": 0.16183657944202423,
      "learning_rate": 7.498389916275647e-05,
      "loss": 0.6988,
      "mean_token_accuracy": 0.7937086336314678,
      "num_tokens": 260387635.0,
      "step": 8158
    },
    {
      "entropy": 0.7557651335373521,
      "epoch": 0.7506799845889328,
      "grad_norm": 0.14799556136131287,
      "learning_rate": 7.498083233661484e-05,
      "loss": 0.745,
      "mean_token_accuracy": 0.7800781540572643,
      "num_tokens": 260420235.0,
      "step": 8159
    },
    {
      "entropy": 0.808697147294879,
      "epoch": 0.7507719909603741,
      "grad_norm": 0.160725936293602,
      "learning_rate": 7.497776551047322e-05,
      "loss": 0.7938,
      "mean_token_accuracy": 0.7686141096055508,
      "num_tokens": 260452565.0,
      "step": 8160
    },
    {
      "entropy": 0.7439595954492688,
      "epoch": 0.7508639973318152,
      "grad_norm": 0.15518374741077423,
      "learning_rate": 7.497469868433159e-05,
      "loss": 0.7222,
      "mean_token_accuracy": 0.7873161286115646,
      "num_tokens": 260484248.0,
      "step": 8161
    },
    {
      "entropy": 0.7745867688208818,
      "epoch": 0.7509560037032564,
      "grad_norm": 0.15835842490196228,
      "learning_rate": 7.497163185818997e-05,
      "loss": 0.7565,
      "mean_token_accuracy": 0.7708470933139324,
      "num_tokens": 260515768.0,
      "step": 8162
    },
    {
      "entropy": 0.719865957275033,
      "epoch": 0.7510480100746977,
      "grad_norm": 0.1607019156217575,
      "learning_rate": 7.496856503204834e-05,
      "loss": 0.7123,
      "mean_token_accuracy": 0.7936288900673389,
      "num_tokens": 260547572.0,
      "step": 8163
    },
    {
      "entropy": 0.7498037926852703,
      "epoch": 0.7511400164461389,
      "grad_norm": 0.15736305713653564,
      "learning_rate": 7.496549820590672e-05,
      "loss": 0.7372,
      "mean_token_accuracy": 0.7786340452730656,
      "num_tokens": 260579224.0,
      "step": 8164
    },
    {
      "entropy": 0.7100143749266863,
      "epoch": 0.7512320228175802,
      "grad_norm": 0.15127715468406677,
      "learning_rate": 7.496243137976508e-05,
      "loss": 0.6912,
      "mean_token_accuracy": 0.7931430712342262,
      "num_tokens": 260611245.0,
      "step": 8165
    },
    {
      "entropy": 0.7352777756750584,
      "epoch": 0.7513240291890213,
      "grad_norm": 0.15133431553840637,
      "learning_rate": 7.495936455362345e-05,
      "loss": 0.7244,
      "mean_token_accuracy": 0.7867128029465675,
      "num_tokens": 260643627.0,
      "step": 8166
    },
    {
      "entropy": 0.6755937207490206,
      "epoch": 0.7514160355604625,
      "grad_norm": 0.14862285554409027,
      "learning_rate": 7.495629772748183e-05,
      "loss": 0.661,
      "mean_token_accuracy": 0.8028155900537968,
      "num_tokens": 260676286.0,
      "step": 8167
    },
    {
      "entropy": 0.7321220897138119,
      "epoch": 0.7515080419319038,
      "grad_norm": 0.1679999679327011,
      "learning_rate": 7.495323090134022e-05,
      "loss": 0.7467,
      "mean_token_accuracy": 0.7797200083732605,
      "num_tokens": 260707969.0,
      "step": 8168
    },
    {
      "entropy": 0.6624914836138487,
      "epoch": 0.751600048303345,
      "grad_norm": 0.15498585999011993,
      "learning_rate": 7.495016407519858e-05,
      "loss": 0.6543,
      "mean_token_accuracy": 0.8082032464444637,
      "num_tokens": 260740648.0,
      "step": 8169
    },
    {
      "entropy": 0.6353367976844311,
      "epoch": 0.7516920546747863,
      "grad_norm": 0.1652747094631195,
      "learning_rate": 7.494709724905695e-05,
      "loss": 0.6281,
      "mean_token_accuracy": 0.8109821639955044,
      "num_tokens": 260772732.0,
      "step": 8170
    },
    {
      "entropy": 0.704548922367394,
      "epoch": 0.7517840610462274,
      "grad_norm": 0.16201676428318024,
      "learning_rate": 7.494403042291532e-05,
      "loss": 0.703,
      "mean_token_accuracy": 0.7902182415127754,
      "num_tokens": 260803699.0,
      "step": 8171
    },
    {
      "entropy": 0.7624512538313866,
      "epoch": 0.7518760674176687,
      "grad_norm": 0.1571079045534134,
      "learning_rate": 7.49409635967737e-05,
      "loss": 0.7396,
      "mean_token_accuracy": 0.7822989523410797,
      "num_tokens": 260835697.0,
      "step": 8172
    },
    {
      "entropy": 0.8051909171044827,
      "epoch": 0.7519680737891099,
      "grad_norm": 0.16437318921089172,
      "learning_rate": 7.493789677063208e-05,
      "loss": 0.7819,
      "mean_token_accuracy": 0.7690221332013607,
      "num_tokens": 260867151.0,
      "step": 8173
    },
    {
      "entropy": 0.6807244475930929,
      "epoch": 0.7520600801605511,
      "grad_norm": 0.1574520766735077,
      "learning_rate": 7.493482994449045e-05,
      "loss": 0.6684,
      "mean_token_accuracy": 0.7975836619734764,
      "num_tokens": 260898976.0,
      "step": 8174
    },
    {
      "entropy": 0.8008286450058222,
      "epoch": 0.7521520865319924,
      "grad_norm": 0.15357588231563568,
      "learning_rate": 7.493176311834882e-05,
      "loss": 0.8001,
      "mean_token_accuracy": 0.7672271579504013,
      "num_tokens": 260931354.0,
      "step": 8175
    },
    {
      "entropy": 0.7583520617336035,
      "epoch": 0.7522440929034335,
      "grad_norm": 0.15235504508018494,
      "learning_rate": 7.49286962922072e-05,
      "loss": 0.7486,
      "mean_token_accuracy": 0.7772847339510918,
      "num_tokens": 260963805.0,
      "step": 8176
    },
    {
      "entropy": 0.8241102714091539,
      "epoch": 0.7523360992748748,
      "grad_norm": 0.16620345413684845,
      "learning_rate": 7.492562946606557e-05,
      "loss": 0.8247,
      "mean_token_accuracy": 0.7580513507127762,
      "num_tokens": 260995959.0,
      "step": 8177
    },
    {
      "entropy": 0.7683788016438484,
      "epoch": 0.752428105646316,
      "grad_norm": 0.17560122907161713,
      "learning_rate": 7.492256263992395e-05,
      "loss": 0.7553,
      "mean_token_accuracy": 0.7797304391860962,
      "num_tokens": 261026914.0,
      "step": 8178
    },
    {
      "entropy": 0.5798203386366367,
      "epoch": 0.7525201120177573,
      "grad_norm": 0.14537012577056885,
      "learning_rate": 7.491949581378233e-05,
      "loss": 0.5491,
      "mean_token_accuracy": 0.8310942836105824,
      "num_tokens": 261059682.0,
      "step": 8179
    },
    {
      "entropy": 0.700972156599164,
      "epoch": 0.7526121183891985,
      "grad_norm": 0.15300025045871735,
      "learning_rate": 7.49164289876407e-05,
      "loss": 0.6934,
      "mean_token_accuracy": 0.7945052981376648,
      "num_tokens": 261090680.0,
      "step": 8180
    },
    {
      "entropy": 0.7118259388953447,
      "epoch": 0.7527041247606396,
      "grad_norm": 0.15457327663898468,
      "learning_rate": 7.491336216149907e-05,
      "loss": 0.6966,
      "mean_token_accuracy": 0.7910516820847988,
      "num_tokens": 261123033.0,
      "step": 8181
    },
    {
      "entropy": 0.833488205447793,
      "epoch": 0.7527961311320809,
      "grad_norm": 0.15687143802642822,
      "learning_rate": 7.491029533535743e-05,
      "loss": 0.8292,
      "mean_token_accuracy": 0.7570666670799255,
      "num_tokens": 261154073.0,
      "step": 8182
    },
    {
      "entropy": 0.7915023639798164,
      "epoch": 0.7528881375035221,
      "grad_norm": 0.1605619192123413,
      "learning_rate": 7.490722850921581e-05,
      "loss": 0.7951,
      "mean_token_accuracy": 0.7703123278915882,
      "num_tokens": 261185277.0,
      "step": 8183
    },
    {
      "entropy": 0.6752824895083904,
      "epoch": 0.7529801438749634,
      "grad_norm": 0.1528957039117813,
      "learning_rate": 7.49041616830742e-05,
      "loss": 0.6541,
      "mean_token_accuracy": 0.8050530217587948,
      "num_tokens": 261217470.0,
      "step": 8184
    },
    {
      "entropy": 0.6841637883335352,
      "epoch": 0.7530721502464046,
      "grad_norm": 0.15161387622356415,
      "learning_rate": 7.490109485693256e-05,
      "loss": 0.6708,
      "mean_token_accuracy": 0.8002060726284981,
      "num_tokens": 261249707.0,
      "step": 8185
    },
    {
      "entropy": 0.7354242280125618,
      "epoch": 0.7531641566178457,
      "grad_norm": 0.1614283323287964,
      "learning_rate": 7.489802803079093e-05,
      "loss": 0.7426,
      "mean_token_accuracy": 0.780477300286293,
      "num_tokens": 261281341.0,
      "step": 8186
    },
    {
      "entropy": 0.796094972640276,
      "epoch": 0.753256162989287,
      "grad_norm": 0.16440348327159882,
      "learning_rate": 7.489496120464931e-05,
      "loss": 0.7939,
      "mean_token_accuracy": 0.7704532593488693,
      "num_tokens": 261313176.0,
      "step": 8187
    },
    {
      "entropy": 0.6641734135337174,
      "epoch": 0.7533481693607282,
      "grad_norm": 0.16440096497535706,
      "learning_rate": 7.489189437850768e-05,
      "loss": 0.6671,
      "mean_token_accuracy": 0.8028215765953064,
      "num_tokens": 261345393.0,
      "step": 8188
    },
    {
      "entropy": 0.7380818780511618,
      "epoch": 0.7534401757321695,
      "grad_norm": 0.14861905574798584,
      "learning_rate": 7.488882755236606e-05,
      "loss": 0.713,
      "mean_token_accuracy": 0.7873834669589996,
      "num_tokens": 261377389.0,
      "step": 8189
    },
    {
      "entropy": 0.6921974327415228,
      "epoch": 0.7535321821036107,
      "grad_norm": 0.15490610897541046,
      "learning_rate": 7.488576072622443e-05,
      "loss": 0.6779,
      "mean_token_accuracy": 0.7983246371150017,
      "num_tokens": 261410051.0,
      "step": 8190
    },
    {
      "entropy": 0.6894877022132277,
      "epoch": 0.7536241884750519,
      "grad_norm": 0.15635740756988525,
      "learning_rate": 7.488269390008281e-05,
      "loss": 0.6669,
      "mean_token_accuracy": 0.8050928115844727,
      "num_tokens": 261442390.0,
      "step": 8191
    },
    {
      "entropy": 0.706382516771555,
      "epoch": 0.7537161948464931,
      "grad_norm": 0.15374554693698883,
      "learning_rate": 7.487962707394118e-05,
      "loss": 0.6936,
      "mean_token_accuracy": 0.7951792255043983,
      "num_tokens": 261474314.0,
      "step": 8192
    },
    {
      "entropy": 0.7727990485727787,
      "epoch": 0.7538082012179343,
      "grad_norm": 0.15891298651695251,
      "learning_rate": 7.487656024779955e-05,
      "loss": 0.7622,
      "mean_token_accuracy": 0.7765675634145737,
      "num_tokens": 261506437.0,
      "step": 8193
    },
    {
      "entropy": 0.7824715450406075,
      "epoch": 0.7539002075893756,
      "grad_norm": 0.16517873108386993,
      "learning_rate": 7.487349342165793e-05,
      "loss": 0.7778,
      "mean_token_accuracy": 0.7733386419713497,
      "num_tokens": 261537381.0,
      "step": 8194
    },
    {
      "entropy": 0.7080783788114786,
      "epoch": 0.7539922139608168,
      "grad_norm": 0.15313094854354858,
      "learning_rate": 7.487042659551631e-05,
      "loss": 0.7119,
      "mean_token_accuracy": 0.7892317473888397,
      "num_tokens": 261569747.0,
      "step": 8195
    },
    {
      "entropy": 0.660080211237073,
      "epoch": 0.754084220332258,
      "grad_norm": 0.15274693071842194,
      "learning_rate": 7.486735976937468e-05,
      "loss": 0.6464,
      "mean_token_accuracy": 0.803916972130537,
      "num_tokens": 261601785.0,
      "step": 8196
    },
    {
      "entropy": 0.7246275451034307,
      "epoch": 0.7541762267036992,
      "grad_norm": 0.16484051942825317,
      "learning_rate": 7.486429294323305e-05,
      "loss": 0.7259,
      "mean_token_accuracy": 0.7876493521034718,
      "num_tokens": 261633411.0,
      "step": 8197
    },
    {
      "entropy": 0.6805070443078876,
      "epoch": 0.7542682330751405,
      "grad_norm": 0.15904469788074493,
      "learning_rate": 7.486122611709143e-05,
      "loss": 0.6733,
      "mean_token_accuracy": 0.8020042032003403,
      "num_tokens": 261665624.0,
      "step": 8198
    },
    {
      "entropy": 0.753519942983985,
      "epoch": 0.7543602394465817,
      "grad_norm": 0.15709465742111206,
      "learning_rate": 7.485815929094981e-05,
      "loss": 0.7352,
      "mean_token_accuracy": 0.7851568050682545,
      "num_tokens": 261697312.0,
      "step": 8199
    },
    {
      "entropy": 0.7426989357918501,
      "epoch": 0.754452245818023,
      "grad_norm": 0.183659628033638,
      "learning_rate": 7.485509246480818e-05,
      "loss": 0.7365,
      "mean_token_accuracy": 0.7861574292182922,
      "num_tokens": 261729140.0,
      "step": 8200
    },
    {
      "entropy": 0.6448131706565619,
      "epoch": 0.7545442521894641,
      "grad_norm": 0.14870941638946533,
      "learning_rate": 7.485202563866654e-05,
      "loss": 0.6346,
      "mean_token_accuracy": 0.8114809170365334,
      "num_tokens": 261761026.0,
      "step": 8201
    },
    {
      "entropy": 0.7399176862090826,
      "epoch": 0.7546362585609053,
      "grad_norm": 0.15599779784679413,
      "learning_rate": 7.484895881252491e-05,
      "loss": 0.731,
      "mean_token_accuracy": 0.78640441223979,
      "num_tokens": 261792858.0,
      "step": 8202
    },
    {
      "entropy": 0.6809215173125267,
      "epoch": 0.7547282649323466,
      "grad_norm": 0.15540766716003418,
      "learning_rate": 7.48458919863833e-05,
      "loss": 0.6708,
      "mean_token_accuracy": 0.8010811433196068,
      "num_tokens": 261825214.0,
      "step": 8203
    },
    {
      "entropy": 0.7297811843454838,
      "epoch": 0.7548202713037878,
      "grad_norm": 0.15744496881961823,
      "learning_rate": 7.484282516024168e-05,
      "loss": 0.7269,
      "mean_token_accuracy": 0.7868754267692566,
      "num_tokens": 261857765.0,
      "step": 8204
    },
    {
      "entropy": 0.6819806415587664,
      "epoch": 0.7549122776752291,
      "grad_norm": 0.15466612577438354,
      "learning_rate": 7.483975833410004e-05,
      "loss": 0.6538,
      "mean_token_accuracy": 0.8067353963851929,
      "num_tokens": 261889542.0,
      "step": 8205
    },
    {
      "entropy": 0.7598924599587917,
      "epoch": 0.7550042840466702,
      "grad_norm": 0.15165108442306519,
      "learning_rate": 7.483669150795841e-05,
      "loss": 0.7603,
      "mean_token_accuracy": 0.7737213484942913,
      "num_tokens": 261921837.0,
      "step": 8206
    },
    {
      "entropy": 0.7126630079001188,
      "epoch": 0.7550962904181114,
      "grad_norm": 0.15084855258464813,
      "learning_rate": 7.483362468181679e-05,
      "loss": 0.6955,
      "mean_token_accuracy": 0.7913403064012527,
      "num_tokens": 261954380.0,
      "step": 8207
    },
    {
      "entropy": 0.6328637357801199,
      "epoch": 0.7551882967895527,
      "grad_norm": 0.15313968062400818,
      "learning_rate": 7.483055785567516e-05,
      "loss": 0.6222,
      "mean_token_accuracy": 0.8153796195983887,
      "num_tokens": 261985399.0,
      "step": 8208
    },
    {
      "entropy": 0.6268606930971146,
      "epoch": 0.7552803031609939,
      "grad_norm": 0.14737623929977417,
      "learning_rate": 7.482749102953354e-05,
      "loss": 0.6074,
      "mean_token_accuracy": 0.8157814256846905,
      "num_tokens": 262016787.0,
      "step": 8209
    },
    {
      "entropy": 0.6614891029894352,
      "epoch": 0.7553723095324352,
      "grad_norm": 0.15194430947303772,
      "learning_rate": 7.482442420339192e-05,
      "loss": 0.6496,
      "mean_token_accuracy": 0.8074677214026451,
      "num_tokens": 262048393.0,
      "step": 8210
    },
    {
      "entropy": 0.7158561069518328,
      "epoch": 0.7554643159038763,
      "grad_norm": 0.16500596702098846,
      "learning_rate": 7.482135737725029e-05,
      "loss": 0.6987,
      "mean_token_accuracy": 0.7931565903127193,
      "num_tokens": 262080401.0,
      "step": 8211
    },
    {
      "entropy": 0.7224964424967766,
      "epoch": 0.7555563222753175,
      "grad_norm": 0.1551785171031952,
      "learning_rate": 7.481829055110866e-05,
      "loss": 0.7095,
      "mean_token_accuracy": 0.7915256321430206,
      "num_tokens": 262111596.0,
      "step": 8212
    },
    {
      "entropy": 0.7682050969451666,
      "epoch": 0.7556483286467588,
      "grad_norm": 0.15604077279567719,
      "learning_rate": 7.481522372496703e-05,
      "loss": 0.7464,
      "mean_token_accuracy": 0.7791872136294842,
      "num_tokens": 262142477.0,
      "step": 8213
    },
    {
      "entropy": 0.7908107172697783,
      "epoch": 0.7557403350182,
      "grad_norm": 0.1516953855752945,
      "learning_rate": 7.481215689882541e-05,
      "loss": 0.7772,
      "mean_token_accuracy": 0.7702491730451584,
      "num_tokens": 262174069.0,
      "step": 8214
    },
    {
      "entropy": 0.8287880253046751,
      "epoch": 0.7558323413896413,
      "grad_norm": 0.1670948565006256,
      "learning_rate": 7.480909007268379e-05,
      "loss": 0.8408,
      "mean_token_accuracy": 0.752510379999876,
      "num_tokens": 262206368.0,
      "step": 8215
    },
    {
      "entropy": 0.8625408615916967,
      "epoch": 0.7559243477610824,
      "grad_norm": 0.1683575063943863,
      "learning_rate": 7.480602324654216e-05,
      "loss": 0.8541,
      "mean_token_accuracy": 0.7504784017801285,
      "num_tokens": 262238157.0,
      "step": 8216
    },
    {
      "entropy": 0.6215582769364119,
      "epoch": 0.7560163541325237,
      "grad_norm": 0.1479293256998062,
      "learning_rate": 7.480295642040053e-05,
      "loss": 0.6062,
      "mean_token_accuracy": 0.8183351345360279,
      "num_tokens": 262270645.0,
      "step": 8217
    },
    {
      "entropy": 0.7118756771087646,
      "epoch": 0.7561083605039649,
      "grad_norm": 0.15802918374538422,
      "learning_rate": 7.479988959425891e-05,
      "loss": 0.7143,
      "mean_token_accuracy": 0.7851661033928394,
      "num_tokens": 262302084.0,
      "step": 8218
    },
    {
      "entropy": 0.6988493297249079,
      "epoch": 0.7562003668754061,
      "grad_norm": 0.1498042643070221,
      "learning_rate": 7.479682276811727e-05,
      "loss": 0.6697,
      "mean_token_accuracy": 0.7990635484457016,
      "num_tokens": 262333776.0,
      "step": 8219
    },
    {
      "entropy": 0.7575699770823121,
      "epoch": 0.7562923732468474,
      "grad_norm": 0.14928895235061646,
      "learning_rate": 7.479375594197566e-05,
      "loss": 0.7487,
      "mean_token_accuracy": 0.7836466059088707,
      "num_tokens": 262365805.0,
      "step": 8220
    },
    {
      "entropy": 0.641384856775403,
      "epoch": 0.7563843796182885,
      "grad_norm": 0.15321554243564606,
      "learning_rate": 7.479068911583402e-05,
      "loss": 0.6247,
      "mean_token_accuracy": 0.8132530078291893,
      "num_tokens": 262397380.0,
      "step": 8221
    },
    {
      "entropy": 0.8225880246609449,
      "epoch": 0.7564763859897298,
      "grad_norm": 0.17675383388996124,
      "learning_rate": 7.47876222896924e-05,
      "loss": 0.8194,
      "mean_token_accuracy": 0.7577001005411148,
      "num_tokens": 262428889.0,
      "step": 8222
    },
    {
      "entropy": 0.8032648134976625,
      "epoch": 0.756568392361171,
      "grad_norm": 0.16304530203342438,
      "learning_rate": 7.478455546355077e-05,
      "loss": 0.8048,
      "mean_token_accuracy": 0.7689240910112858,
      "num_tokens": 262461147.0,
      "step": 8223
    },
    {
      "entropy": 0.6944338772445917,
      "epoch": 0.7566603987326123,
      "grad_norm": 0.15286129713058472,
      "learning_rate": 7.478148863740914e-05,
      "loss": 0.6698,
      "mean_token_accuracy": 0.8010267689824104,
      "num_tokens": 262493351.0,
      "step": 8224
    },
    {
      "entropy": 0.8632571995258331,
      "epoch": 0.7567524051040535,
      "grad_norm": 0.1684325635433197,
      "learning_rate": 7.477842181126752e-05,
      "loss": 0.8585,
      "mean_token_accuracy": 0.75052360445261,
      "num_tokens": 262524971.0,
      "step": 8225
    },
    {
      "entropy": 0.8165631741285324,
      "epoch": 0.7568444114754946,
      "grad_norm": 0.15757973492145538,
      "learning_rate": 7.47753549851259e-05,
      "loss": 0.8025,
      "mean_token_accuracy": 0.7637781091034412,
      "num_tokens": 262557177.0,
      "step": 8226
    },
    {
      "entropy": 0.7244785409420729,
      "epoch": 0.7569364178469359,
      "grad_norm": 0.15845906734466553,
      "learning_rate": 7.477228815898427e-05,
      "loss": 0.704,
      "mean_token_accuracy": 0.7893138714134693,
      "num_tokens": 262587738.0,
      "step": 8227
    },
    {
      "entropy": 0.5866013076156378,
      "epoch": 0.7570284242183771,
      "grad_norm": 0.1537148356437683,
      "learning_rate": 7.476922133284264e-05,
      "loss": 0.5711,
      "mean_token_accuracy": 0.8287542834877968,
      "num_tokens": 262620437.0,
      "step": 8228
    },
    {
      "entropy": 0.6251798309385777,
      "epoch": 0.7571204305898184,
      "grad_norm": 0.14838430285453796,
      "learning_rate": 7.476615450670101e-05,
      "loss": 0.6101,
      "mean_token_accuracy": 0.8153510019183159,
      "num_tokens": 262652418.0,
      "step": 8229
    },
    {
      "entropy": 0.8390625156462193,
      "epoch": 0.7572124369612596,
      "grad_norm": 0.15559923648834229,
      "learning_rate": 7.47630876805594e-05,
      "loss": 0.8091,
      "mean_token_accuracy": 0.7657324559986591,
      "num_tokens": 262684670.0,
      "step": 8230
    },
    {
      "entropy": 0.656634408980608,
      "epoch": 0.7573044433327007,
      "grad_norm": 0.14747051894664764,
      "learning_rate": 7.476002085441777e-05,
      "loss": 0.6396,
      "mean_token_accuracy": 0.8105392083525658,
      "num_tokens": 262716162.0,
      "step": 8231
    },
    {
      "entropy": 0.7738148588687181,
      "epoch": 0.757396449704142,
      "grad_norm": 0.15811821818351746,
      "learning_rate": 7.475695402827614e-05,
      "loss": 0.7725,
      "mean_token_accuracy": 0.7739506289362907,
      "num_tokens": 262747369.0,
      "step": 8232
    },
    {
      "entropy": 0.803223755210638,
      "epoch": 0.7574884560755832,
      "grad_norm": 0.15423259139060974,
      "learning_rate": 7.47538872021345e-05,
      "loss": 0.807,
      "mean_token_accuracy": 0.7603829652070999,
      "num_tokens": 262779886.0,
      "step": 8233
    },
    {
      "entropy": 0.8037503100931644,
      "epoch": 0.7575804624470245,
      "grad_norm": 0.15880978107452393,
      "learning_rate": 7.475082037599289e-05,
      "loss": 0.8331,
      "mean_token_accuracy": 0.7577344067394733,
      "num_tokens": 262812256.0,
      "step": 8234
    },
    {
      "entropy": 0.6305478252470493,
      "epoch": 0.7576724688184657,
      "grad_norm": 0.14847566187381744,
      "learning_rate": 7.474775354985127e-05,
      "loss": 0.6116,
      "mean_token_accuracy": 0.8217657208442688,
      "num_tokens": 262844697.0,
      "step": 8235
    },
    {
      "entropy": 0.6503655333071947,
      "epoch": 0.7577644751899069,
      "grad_norm": 0.15296663343906403,
      "learning_rate": 7.474468672370964e-05,
      "loss": 0.6443,
      "mean_token_accuracy": 0.8087447956204414,
      "num_tokens": 262876776.0,
      "step": 8236
    },
    {
      "entropy": 0.5900165317580104,
      "epoch": 0.7578564815613481,
      "grad_norm": 0.1531151533126831,
      "learning_rate": 7.474161989756802e-05,
      "loss": 0.5794,
      "mean_token_accuracy": 0.825041726231575,
      "num_tokens": 262908649.0,
      "step": 8237
    },
    {
      "entropy": 0.6174322310835123,
      "epoch": 0.7579484879327893,
      "grad_norm": 0.14951221644878387,
      "learning_rate": 7.473855307142639e-05,
      "loss": 0.6135,
      "mean_token_accuracy": 0.8183983340859413,
      "num_tokens": 262939866.0,
      "step": 8238
    },
    {
      "entropy": 0.7011986877769232,
      "epoch": 0.7580404943042306,
      "grad_norm": 0.1568177193403244,
      "learning_rate": 7.473548624528475e-05,
      "loss": 0.69,
      "mean_token_accuracy": 0.7979583665728569,
      "num_tokens": 262970936.0,
      "step": 8239
    },
    {
      "entropy": 0.7116019409149885,
      "epoch": 0.7581325006756718,
      "grad_norm": 0.15777753293514252,
      "learning_rate": 7.473241941914314e-05,
      "loss": 0.6898,
      "mean_token_accuracy": 0.7889729104936123,
      "num_tokens": 263001568.0,
      "step": 8240
    },
    {
      "entropy": 0.6978958565741777,
      "epoch": 0.758224507047113,
      "grad_norm": 0.15341345965862274,
      "learning_rate": 7.472935259300152e-05,
      "loss": 0.6746,
      "mean_token_accuracy": 0.7986459918320179,
      "num_tokens": 263034102.0,
      "step": 8241
    },
    {
      "entropy": 0.7310754209756851,
      "epoch": 0.7583165134185542,
      "grad_norm": 0.16287925839424133,
      "learning_rate": 7.472628576685988e-05,
      "loss": 0.7344,
      "mean_token_accuracy": 0.7872771918773651,
      "num_tokens": 263066311.0,
      "step": 8242
    },
    {
      "entropy": 0.7436950039118528,
      "epoch": 0.7584085197899955,
      "grad_norm": 0.1499532014131546,
      "learning_rate": 7.472321894071825e-05,
      "loss": 0.7153,
      "mean_token_accuracy": 0.7900819964706898,
      "num_tokens": 263097782.0,
      "step": 8243
    },
    {
      "entropy": 0.7081038653850555,
      "epoch": 0.7585005261614367,
      "grad_norm": 0.15753324329853058,
      "learning_rate": 7.472015211457662e-05,
      "loss": 0.6829,
      "mean_token_accuracy": 0.7988395839929581,
      "num_tokens": 263129743.0,
      "step": 8244
    },
    {
      "entropy": 0.7242036946117878,
      "epoch": 0.758592532532878,
      "grad_norm": 0.16255426406860352,
      "learning_rate": 7.4717085288435e-05,
      "loss": 0.7077,
      "mean_token_accuracy": 0.793377261608839,
      "num_tokens": 263161412.0,
      "step": 8245
    },
    {
      "entropy": 0.8192476816475391,
      "epoch": 0.7586845389043191,
      "grad_norm": 0.15719883143901825,
      "learning_rate": 7.471401846229338e-05,
      "loss": 0.8006,
      "mean_token_accuracy": 0.7623475417494774,
      "num_tokens": 263193249.0,
      "step": 8246
    },
    {
      "entropy": 0.7637277990579605,
      "epoch": 0.7587765452757603,
      "grad_norm": 0.15801171958446503,
      "learning_rate": 7.471095163615175e-05,
      "loss": 0.7644,
      "mean_token_accuracy": 0.7764206156134605,
      "num_tokens": 263225482.0,
      "step": 8247
    },
    {
      "entropy": 0.6744451876729727,
      "epoch": 0.7588685516472016,
      "grad_norm": 0.15485665202140808,
      "learning_rate": 7.470788481001012e-05,
      "loss": 0.6563,
      "mean_token_accuracy": 0.801769319921732,
      "num_tokens": 263257502.0,
      "step": 8248
    },
    {
      "entropy": 0.7247594557702541,
      "epoch": 0.7589605580186428,
      "grad_norm": 0.1554083377122879,
      "learning_rate": 7.47048179838685e-05,
      "loss": 0.6939,
      "mean_token_accuracy": 0.7959292270243168,
      "num_tokens": 263289444.0,
      "step": 8249
    },
    {
      "entropy": 0.7253250889480114,
      "epoch": 0.7590525643900841,
      "grad_norm": 0.16171443462371826,
      "learning_rate": 7.470175115772687e-05,
      "loss": 0.731,
      "mean_token_accuracy": 0.7828046567738056,
      "num_tokens": 263321432.0,
      "step": 8250
    },
    {
      "entropy": 0.7364598736166954,
      "epoch": 0.7591445707615252,
      "grad_norm": 0.1516440361738205,
      "learning_rate": 7.469868433158525e-05,
      "loss": 0.7286,
      "mean_token_accuracy": 0.7803388312458992,
      "num_tokens": 263353091.0,
      "step": 8251
    },
    {
      "entropy": 0.7685550004243851,
      "epoch": 0.7592365771329664,
      "grad_norm": 0.1569109857082367,
      "learning_rate": 7.469561750544362e-05,
      "loss": 0.7551,
      "mean_token_accuracy": 0.7795202918350697,
      "num_tokens": 263383260.0,
      "step": 8252
    },
    {
      "entropy": 0.6977321524173021,
      "epoch": 0.7593285835044077,
      "grad_norm": 0.15398943424224854,
      "learning_rate": 7.4692550679302e-05,
      "loss": 0.6776,
      "mean_token_accuracy": 0.7939479127526283,
      "num_tokens": 263414642.0,
      "step": 8253
    },
    {
      "entropy": 0.5905602443963289,
      "epoch": 0.7594205898758489,
      "grad_norm": 0.17606808245182037,
      "learning_rate": 7.468948385316037e-05,
      "loss": 0.578,
      "mean_token_accuracy": 0.8256075158715248,
      "num_tokens": 263446628.0,
      "step": 8254
    },
    {
      "entropy": 0.670077913440764,
      "epoch": 0.7595125962472902,
      "grad_norm": 0.14777334034442902,
      "learning_rate": 7.468641702701874e-05,
      "loss": 0.6567,
      "mean_token_accuracy": 0.8057082071900368,
      "num_tokens": 263479391.0,
      "step": 8255
    },
    {
      "entropy": 0.6906910687685013,
      "epoch": 0.7596046026187313,
      "grad_norm": 0.15583226084709167,
      "learning_rate": 7.468335020087712e-05,
      "loss": 0.6918,
      "mean_token_accuracy": 0.7932366840541363,
      "num_tokens": 263511175.0,
      "step": 8256
    },
    {
      "entropy": 0.8440969791263342,
      "epoch": 0.7596966089901726,
      "grad_norm": 0.16678713262081146,
      "learning_rate": 7.46802833747355e-05,
      "loss": 0.8379,
      "mean_token_accuracy": 0.7583108358085155,
      "num_tokens": 263543022.0,
      "step": 8257
    },
    {
      "entropy": 0.7866516038775444,
      "epoch": 0.7597886153616138,
      "grad_norm": 0.15734367072582245,
      "learning_rate": 7.467721654859387e-05,
      "loss": 0.7743,
      "mean_token_accuracy": 0.775739099830389,
      "num_tokens": 263574614.0,
      "step": 8258
    },
    {
      "entropy": 0.7567582465708256,
      "epoch": 0.759880621733055,
      "grad_norm": 0.15461865067481995,
      "learning_rate": 7.467414972245223e-05,
      "loss": 0.7291,
      "mean_token_accuracy": 0.7880417332053185,
      "num_tokens": 263606884.0,
      "step": 8259
    },
    {
      "entropy": 0.7718980945646763,
      "epoch": 0.7599726281044963,
      "grad_norm": 0.1584007889032364,
      "learning_rate": 7.46710828963106e-05,
      "loss": 0.7451,
      "mean_token_accuracy": 0.7780902795493603,
      "num_tokens": 263639502.0,
      "step": 8260
    },
    {
      "entropy": 0.7617189045995474,
      "epoch": 0.7600646344759374,
      "grad_norm": 0.15751011669635773,
      "learning_rate": 7.466801607016898e-05,
      "loss": 0.7455,
      "mean_token_accuracy": 0.7796768806874752,
      "num_tokens": 263671409.0,
      "step": 8261
    },
    {
      "entropy": 0.7474177200347185,
      "epoch": 0.7601566408473787,
      "grad_norm": 0.156929150223732,
      "learning_rate": 7.466494924402736e-05,
      "loss": 0.7388,
      "mean_token_accuracy": 0.7842720039188862,
      "num_tokens": 263703444.0,
      "step": 8262
    },
    {
      "entropy": 0.7766599059104919,
      "epoch": 0.7602486472188199,
      "grad_norm": 0.15569151937961578,
      "learning_rate": 7.466188241788573e-05,
      "loss": 0.7757,
      "mean_token_accuracy": 0.7710894010961056,
      "num_tokens": 263735281.0,
      "step": 8263
    },
    {
      "entropy": 0.7847089245915413,
      "epoch": 0.7603406535902612,
      "grad_norm": 0.16274261474609375,
      "learning_rate": 7.46588155917441e-05,
      "loss": 0.7703,
      "mean_token_accuracy": 0.7735401540994644,
      "num_tokens": 263766981.0,
      "step": 8264
    },
    {
      "entropy": 0.78558149933815,
      "epoch": 0.7604326599617024,
      "grad_norm": 0.1562548279762268,
      "learning_rate": 7.465574876560248e-05,
      "loss": 0.7838,
      "mean_token_accuracy": 0.7685136273503304,
      "num_tokens": 263798860.0,
      "step": 8265
    },
    {
      "entropy": 0.7063761539757252,
      "epoch": 0.7605246663331435,
      "grad_norm": 0.163660928606987,
      "learning_rate": 7.465268193946085e-05,
      "loss": 0.7141,
      "mean_token_accuracy": 0.7890478894114494,
      "num_tokens": 263831369.0,
      "step": 8266
    },
    {
      "entropy": 0.7355092484503984,
      "epoch": 0.7606166727045848,
      "grad_norm": 0.16625136137008667,
      "learning_rate": 7.464961511331923e-05,
      "loss": 0.7193,
      "mean_token_accuracy": 0.7888473942875862,
      "num_tokens": 263862820.0,
      "step": 8267
    },
    {
      "entropy": 0.6950621958822012,
      "epoch": 0.760708679076026,
      "grad_norm": 0.15458062291145325,
      "learning_rate": 7.464654828717761e-05,
      "loss": 0.6894,
      "mean_token_accuracy": 0.7909035906195641,
      "num_tokens": 263893995.0,
      "step": 8268
    },
    {
      "entropy": 0.7436295356601477,
      "epoch": 0.7608006854474673,
      "grad_norm": 0.15227586030960083,
      "learning_rate": 7.464348146103598e-05,
      "loss": 0.7264,
      "mean_token_accuracy": 0.7826791629195213,
      "num_tokens": 263925824.0,
      "step": 8269
    },
    {
      "entropy": 0.6037277914583683,
      "epoch": 0.7608926918189085,
      "grad_norm": 0.16530683636665344,
      "learning_rate": 7.464041463489435e-05,
      "loss": 0.591,
      "mean_token_accuracy": 0.8225298039615154,
      "num_tokens": 263957478.0,
      "step": 8270
    },
    {
      "entropy": 0.750943218357861,
      "epoch": 0.7609846981903496,
      "grad_norm": 0.16408447921276093,
      "learning_rate": 7.463734780875272e-05,
      "loss": 0.7589,
      "mean_token_accuracy": 0.7771908864378929,
      "num_tokens": 263989515.0,
      "step": 8271
    },
    {
      "entropy": 0.639009727165103,
      "epoch": 0.7610767045617909,
      "grad_norm": 0.14876317977905273,
      "learning_rate": 7.463428098261111e-05,
      "loss": 0.6329,
      "mean_token_accuracy": 0.8095829300582409,
      "num_tokens": 264022100.0,
      "step": 8272
    },
    {
      "entropy": 0.7378125004470348,
      "epoch": 0.7611687109332321,
      "grad_norm": 0.15150406956672668,
      "learning_rate": 7.463121415646948e-05,
      "loss": 0.73,
      "mean_token_accuracy": 0.7849761582911015,
      "num_tokens": 264054603.0,
      "step": 8273
    },
    {
      "entropy": 0.7756543280556798,
      "epoch": 0.7612607173046734,
      "grad_norm": 0.16021934151649475,
      "learning_rate": 7.462814733032785e-05,
      "loss": 0.7553,
      "mean_token_accuracy": 0.7798147089779377,
      "num_tokens": 264086681.0,
      "step": 8274
    },
    {
      "entropy": 0.6968152038753033,
      "epoch": 0.7613527236761146,
      "grad_norm": 0.15896163880825043,
      "learning_rate": 7.462508050418621e-05,
      "loss": 0.6831,
      "mean_token_accuracy": 0.7991580292582512,
      "num_tokens": 264118858.0,
      "step": 8275
    },
    {
      "entropy": 0.6914545213803649,
      "epoch": 0.7614447300475558,
      "grad_norm": 0.1503048688173294,
      "learning_rate": 7.46220136780446e-05,
      "loss": 0.6762,
      "mean_token_accuracy": 0.8039382323622704,
      "num_tokens": 264151189.0,
      "step": 8276
    },
    {
      "entropy": 0.6963470783084631,
      "epoch": 0.761536736418997,
      "grad_norm": 0.15550266206264496,
      "learning_rate": 7.461894685190298e-05,
      "loss": 0.6979,
      "mean_token_accuracy": 0.7914805263280869,
      "num_tokens": 264183024.0,
      "step": 8277
    },
    {
      "entropy": 0.7470954842865467,
      "epoch": 0.7616287427904382,
      "grad_norm": 0.15412570536136627,
      "learning_rate": 7.461588002576135e-05,
      "loss": 0.7306,
      "mean_token_accuracy": 0.7844413183629513,
      "num_tokens": 264215367.0,
      "step": 8278
    },
    {
      "entropy": 0.8689124509692192,
      "epoch": 0.7617207491618795,
      "grad_norm": 0.15983857214450836,
      "learning_rate": 7.461281319961971e-05,
      "loss": 0.87,
      "mean_token_accuracy": 0.7466276660561562,
      "num_tokens": 264246734.0,
      "step": 8279
    },
    {
      "entropy": 0.6990936677902937,
      "epoch": 0.7618127555333207,
      "grad_norm": 0.158482164144516,
      "learning_rate": 7.46097463734781e-05,
      "loss": 0.6924,
      "mean_token_accuracy": 0.7996118627488613,
      "num_tokens": 264278451.0,
      "step": 8280
    },
    {
      "entropy": 0.7726360410451889,
      "epoch": 0.7619047619047619,
      "grad_norm": 0.16712352633476257,
      "learning_rate": 7.460667954733646e-05,
      "loss": 0.7702,
      "mean_token_accuracy": 0.7735984064638615,
      "num_tokens": 264310405.0,
      "step": 8281
    },
    {
      "entropy": 0.7697792127728462,
      "epoch": 0.7619967682762031,
      "grad_norm": 0.16427171230316162,
      "learning_rate": 7.460361272119484e-05,
      "loss": 0.7543,
      "mean_token_accuracy": 0.7747297063469887,
      "num_tokens": 264341294.0,
      "step": 8282
    },
    {
      "entropy": 0.6959984041750431,
      "epoch": 0.7620887746476444,
      "grad_norm": 0.1569182425737381,
      "learning_rate": 7.460054589505321e-05,
      "loss": 0.675,
      "mean_token_accuracy": 0.803982499986887,
      "num_tokens": 264373318.0,
      "step": 8283
    },
    {
      "entropy": 0.6800618460401893,
      "epoch": 0.7621807810190856,
      "grad_norm": 0.17077399790287018,
      "learning_rate": 7.459747906891159e-05,
      "loss": 0.6878,
      "mean_token_accuracy": 0.803446140140295,
      "num_tokens": 264405818.0,
      "step": 8284
    },
    {
      "entropy": 0.8083992665633559,
      "epoch": 0.7622727873905268,
      "grad_norm": 0.1587560623884201,
      "learning_rate": 7.459441224276996e-05,
      "loss": 0.8022,
      "mean_token_accuracy": 0.7620429247617722,
      "num_tokens": 264437758.0,
      "step": 8285
    },
    {
      "entropy": 0.7558662556111813,
      "epoch": 0.762364793761968,
      "grad_norm": 0.15366318821907043,
      "learning_rate": 7.459134541662833e-05,
      "loss": 0.7329,
      "mean_token_accuracy": 0.7808988317847252,
      "num_tokens": 264469018.0,
      "step": 8286
    },
    {
      "entropy": 0.6517755035310984,
      "epoch": 0.7624568001334092,
      "grad_norm": 0.15608789026737213,
      "learning_rate": 7.458827859048671e-05,
      "loss": 0.6334,
      "mean_token_accuracy": 0.8115746863186359,
      "num_tokens": 264500928.0,
      "step": 8287
    },
    {
      "entropy": 0.7447854783385992,
      "epoch": 0.7625488065048505,
      "grad_norm": 0.15510402619838715,
      "learning_rate": 7.458521176434509e-05,
      "loss": 0.7295,
      "mean_token_accuracy": 0.7839227356016636,
      "num_tokens": 264533184.0,
      "step": 8288
    },
    {
      "entropy": 0.6868742052465677,
      "epoch": 0.7626408128762917,
      "grad_norm": 0.15337343513965607,
      "learning_rate": 7.458214493820346e-05,
      "loss": 0.6734,
      "mean_token_accuracy": 0.8017899505794048,
      "num_tokens": 264564562.0,
      "step": 8289
    },
    {
      "entropy": 0.7096063271164894,
      "epoch": 0.762732819247733,
      "grad_norm": 0.1524781882762909,
      "learning_rate": 7.457907811206183e-05,
      "loss": 0.706,
      "mean_token_accuracy": 0.7920904979109764,
      "num_tokens": 264597227.0,
      "step": 8290
    },
    {
      "entropy": 0.761906323954463,
      "epoch": 0.7628248256191741,
      "grad_norm": 0.15810203552246094,
      "learning_rate": 7.45760112859202e-05,
      "loss": 0.7669,
      "mean_token_accuracy": 0.7728266157209873,
      "num_tokens": 264628496.0,
      "step": 8291
    },
    {
      "entropy": 0.7340310327708721,
      "epoch": 0.7629168319906153,
      "grad_norm": 0.1584690362215042,
      "learning_rate": 7.457294445977858e-05,
      "loss": 0.729,
      "mean_token_accuracy": 0.7859839536249638,
      "num_tokens": 264660395.0,
      "step": 8292
    },
    {
      "entropy": 0.7026560846716166,
      "epoch": 0.7630088383620566,
      "grad_norm": 0.1526041179895401,
      "learning_rate": 7.456987763363696e-05,
      "loss": 0.6894,
      "mean_token_accuracy": 0.7933261133730412,
      "num_tokens": 264691737.0,
      "step": 8293
    },
    {
      "entropy": 0.7341436576098204,
      "epoch": 0.7631008447334978,
      "grad_norm": 0.1534333974123001,
      "learning_rate": 7.456681080749533e-05,
      "loss": 0.7203,
      "mean_token_accuracy": 0.7886669859290123,
      "num_tokens": 264724035.0,
      "step": 8294
    },
    {
      "entropy": 0.793075580149889,
      "epoch": 0.7631928511049391,
      "grad_norm": 0.16162048280239105,
      "learning_rate": 7.45637439813537e-05,
      "loss": 0.7953,
      "mean_token_accuracy": 0.7691131271421909,
      "num_tokens": 264755889.0,
      "step": 8295
    },
    {
      "entropy": 0.6894465573132038,
      "epoch": 0.7632848574763802,
      "grad_norm": 0.14746391773223877,
      "learning_rate": 7.456067715521208e-05,
      "loss": 0.6772,
      "mean_token_accuracy": 0.7981124445796013,
      "num_tokens": 264787166.0,
      "step": 8296
    },
    {
      "entropy": 0.7581575121730566,
      "epoch": 0.7633768638478214,
      "grad_norm": 0.15795835852622986,
      "learning_rate": 7.455761032907044e-05,
      "loss": 0.7521,
      "mean_token_accuracy": 0.7722776420414448,
      "num_tokens": 264818599.0,
      "step": 8297
    },
    {
      "entropy": 0.7230536099523306,
      "epoch": 0.7634688702192627,
      "grad_norm": 0.1462787538766861,
      "learning_rate": 7.455454350292882e-05,
      "loss": 0.7074,
      "mean_token_accuracy": 0.7945027574896812,
      "num_tokens": 264850169.0,
      "step": 8298
    },
    {
      "entropy": 0.7204119320958853,
      "epoch": 0.7635608765907039,
      "grad_norm": 0.15022489428520203,
      "learning_rate": 7.45514766767872e-05,
      "loss": 0.6927,
      "mean_token_accuracy": 0.7934125401079655,
      "num_tokens": 264882048.0,
      "step": 8299
    },
    {
      "entropy": 0.7572776861488819,
      "epoch": 0.7636528829621452,
      "grad_norm": 0.15648382902145386,
      "learning_rate": 7.454840985064557e-05,
      "loss": 0.7345,
      "mean_token_accuracy": 0.7839141748845577,
      "num_tokens": 264913664.0,
      "step": 8300
    },
    {
      "entropy": 0.6772344037890434,
      "epoch": 0.7637448893335863,
      "grad_norm": 0.16265250742435455,
      "learning_rate": 7.454534302450394e-05,
      "loss": 0.6606,
      "mean_token_accuracy": 0.8019225373864174,
      "num_tokens": 264945388.0,
      "step": 8301
    },
    {
      "entropy": 0.8146530501544476,
      "epoch": 0.7638368957050276,
      "grad_norm": 0.15871009230613708,
      "learning_rate": 7.454227619836231e-05,
      "loss": 0.8034,
      "mean_token_accuracy": 0.7633884027600288,
      "num_tokens": 264977175.0,
      "step": 8302
    },
    {
      "entropy": 0.6485998667776585,
      "epoch": 0.7639289020764688,
      "grad_norm": 0.16182856261730194,
      "learning_rate": 7.453920937222069e-05,
      "loss": 0.65,
      "mean_token_accuracy": 0.8045965768396854,
      "num_tokens": 265008517.0,
      "step": 8303
    },
    {
      "entropy": 0.7152379062026739,
      "epoch": 0.76402090844791,
      "grad_norm": 0.16234590113162994,
      "learning_rate": 7.453614254607907e-05,
      "loss": 0.7162,
      "mean_token_accuracy": 0.7890022806823254,
      "num_tokens": 265040098.0,
      "step": 8304
    },
    {
      "entropy": 0.6008580750785768,
      "epoch": 0.7641129148193513,
      "grad_norm": 0.15087386965751648,
      "learning_rate": 7.453307571993744e-05,
      "loss": 0.5929,
      "mean_token_accuracy": 0.8261102475225925,
      "num_tokens": 265072122.0,
      "step": 8305
    },
    {
      "entropy": 0.8285437822341919,
      "epoch": 0.7642049211907924,
      "grad_norm": 0.16193880140781403,
      "learning_rate": 7.453000889379581e-05,
      "loss": 0.833,
      "mean_token_accuracy": 0.7554953992366791,
      "num_tokens": 265103621.0,
      "step": 8306
    },
    {
      "entropy": 0.7091429531574249,
      "epoch": 0.7642969275622337,
      "grad_norm": 0.1662285178899765,
      "learning_rate": 7.452694206765419e-05,
      "loss": 0.7187,
      "mean_token_accuracy": 0.7910330705344677,
      "num_tokens": 265135452.0,
      "step": 8307
    },
    {
      "entropy": 0.6649100165814161,
      "epoch": 0.7643889339336749,
      "grad_norm": 0.16125935316085815,
      "learning_rate": 7.452387524151256e-05,
      "loss": 0.6646,
      "mean_token_accuracy": 0.7997456677258015,
      "num_tokens": 265167776.0,
      "step": 8308
    },
    {
      "entropy": 0.6712693544104695,
      "epoch": 0.7644809403051162,
      "grad_norm": 0.16077309846878052,
      "learning_rate": 7.452080841537094e-05,
      "loss": 0.6771,
      "mean_token_accuracy": 0.7983238250017166,
      "num_tokens": 265199123.0,
      "step": 8309
    },
    {
      "entropy": 0.7733059264719486,
      "epoch": 0.7645729466765574,
      "grad_norm": 0.153147354722023,
      "learning_rate": 7.451774158922931e-05,
      "loss": 0.7645,
      "mean_token_accuracy": 0.7761630080640316,
      "num_tokens": 265231345.0,
      "step": 8310
    },
    {
      "entropy": 0.7287110639736056,
      "epoch": 0.7646649530479985,
      "grad_norm": 0.15086954832077026,
      "learning_rate": 7.451467476308769e-05,
      "loss": 0.7279,
      "mean_token_accuracy": 0.7849809750914574,
      "num_tokens": 265263948.0,
      "step": 8311
    },
    {
      "entropy": 0.7554072607308626,
      "epoch": 0.7647569594194398,
      "grad_norm": 0.1535101979970932,
      "learning_rate": 7.451160793694606e-05,
      "loss": 0.7408,
      "mean_token_accuracy": 0.7762155570089817,
      "num_tokens": 265295731.0,
      "step": 8312
    },
    {
      "entropy": 0.7793130874633789,
      "epoch": 0.764848965790881,
      "grad_norm": 0.15311850607395172,
      "learning_rate": 7.450854111080442e-05,
      "loss": 0.7696,
      "mean_token_accuracy": 0.7743230275809765,
      "num_tokens": 265327853.0,
      "step": 8313
    },
    {
      "entropy": 0.8236570004373789,
      "epoch": 0.7649409721623223,
      "grad_norm": 0.15331685543060303,
      "learning_rate": 7.45054742846628e-05,
      "loss": 0.7977,
      "mean_token_accuracy": 0.7642853520810604,
      "num_tokens": 265359823.0,
      "step": 8314
    },
    {
      "entropy": 0.74894754961133,
      "epoch": 0.7650329785337635,
      "grad_norm": 0.1519385278224945,
      "learning_rate": 7.450240745852119e-05,
      "loss": 0.7419,
      "mean_token_accuracy": 0.7884279116988182,
      "num_tokens": 265391971.0,
      "step": 8315
    },
    {
      "entropy": 0.7658929079771042,
      "epoch": 0.7651249849052046,
      "grad_norm": 0.16059580445289612,
      "learning_rate": 7.449934063237955e-05,
      "loss": 0.7549,
      "mean_token_accuracy": 0.7786662168800831,
      "num_tokens": 265423533.0,
      "step": 8316
    },
    {
      "entropy": 0.7219445053488016,
      "epoch": 0.7652169912766459,
      "grad_norm": 0.14958974719047546,
      "learning_rate": 7.449627380623792e-05,
      "loss": 0.7001,
      "mean_token_accuracy": 0.7941040061414242,
      "num_tokens": 265455895.0,
      "step": 8317
    },
    {
      "entropy": 0.7200271487236023,
      "epoch": 0.7653089976480871,
      "grad_norm": 0.14959660172462463,
      "learning_rate": 7.44932069800963e-05,
      "loss": 0.7113,
      "mean_token_accuracy": 0.7899212688207626,
      "num_tokens": 265488379.0,
      "step": 8318
    },
    {
      "entropy": 0.6612376738339663,
      "epoch": 0.7654010040195284,
      "grad_norm": 0.1489686220884323,
      "learning_rate": 7.449014015395469e-05,
      "loss": 0.6331,
      "mean_token_accuracy": 0.8106556944549084,
      "num_tokens": 265520593.0,
      "step": 8319
    },
    {
      "entropy": 0.6178223937749863,
      "epoch": 0.7654930103909696,
      "grad_norm": 0.15254203975200653,
      "learning_rate": 7.448707332781305e-05,
      "loss": 0.603,
      "mean_token_accuracy": 0.8209312073886395,
      "num_tokens": 265552815.0,
      "step": 8320
    },
    {
      "entropy": 0.7546176770702004,
      "epoch": 0.7655850167624108,
      "grad_norm": 0.1681319624185562,
      "learning_rate": 7.448400650167142e-05,
      "loss": 0.7488,
      "mean_token_accuracy": 0.7782829254865646,
      "num_tokens": 265583962.0,
      "step": 8321
    },
    {
      "entropy": 0.8568318076431751,
      "epoch": 0.765677023133852,
      "grad_norm": 0.16100016236305237,
      "learning_rate": 7.448093967552979e-05,
      "loss": 0.8627,
      "mean_token_accuracy": 0.7485143207013607,
      "num_tokens": 265616052.0,
      "step": 8322
    },
    {
      "entropy": 0.7068897411227226,
      "epoch": 0.7657690295052932,
      "grad_norm": 0.15858453512191772,
      "learning_rate": 7.447787284938817e-05,
      "loss": 0.6946,
      "mean_token_accuracy": 0.7954084575176239,
      "num_tokens": 265648322.0,
      "step": 8323
    },
    {
      "entropy": 0.8316314201802015,
      "epoch": 0.7658610358767345,
      "grad_norm": 0.15685518085956573,
      "learning_rate": 7.447480602324655e-05,
      "loss": 0.8139,
      "mean_token_accuracy": 0.7644484080374241,
      "num_tokens": 265680390.0,
      "step": 8324
    },
    {
      "entropy": 0.634011322632432,
      "epoch": 0.7659530422481757,
      "grad_norm": 0.15188267827033997,
      "learning_rate": 7.447173919710492e-05,
      "loss": 0.6146,
      "mean_token_accuracy": 0.8134104609489441,
      "num_tokens": 265712533.0,
      "step": 8325
    },
    {
      "entropy": 0.7080543898046017,
      "epoch": 0.7660450486196169,
      "grad_norm": 0.1555669903755188,
      "learning_rate": 7.446867237096329e-05,
      "loss": 0.7087,
      "mean_token_accuracy": 0.7920771203935146,
      "num_tokens": 265744961.0,
      "step": 8326
    },
    {
      "entropy": 0.6251947097480297,
      "epoch": 0.7661370549910581,
      "grad_norm": 0.15106333792209625,
      "learning_rate": 7.446560554482167e-05,
      "loss": 0.6039,
      "mean_token_accuracy": 0.8190983347594738,
      "num_tokens": 265777597.0,
      "step": 8327
    },
    {
      "entropy": 0.6958227697759867,
      "epoch": 0.7662290613624994,
      "grad_norm": 0.1474580019712448,
      "learning_rate": 7.446253871868004e-05,
      "loss": 0.6848,
      "mean_token_accuracy": 0.7982716336846352,
      "num_tokens": 265809552.0,
      "step": 8328
    },
    {
      "entropy": 0.6936463639140129,
      "epoch": 0.7663210677339406,
      "grad_norm": 0.15931807458400726,
      "learning_rate": 7.445947189253842e-05,
      "loss": 0.6738,
      "mean_token_accuracy": 0.7984092831611633,
      "num_tokens": 265841604.0,
      "step": 8329
    },
    {
      "entropy": 0.7399959284812212,
      "epoch": 0.7664130741053818,
      "grad_norm": 0.15329481661319733,
      "learning_rate": 7.44564050663968e-05,
      "loss": 0.7547,
      "mean_token_accuracy": 0.7796805612742901,
      "num_tokens": 265874136.0,
      "step": 8330
    },
    {
      "entropy": 0.830005094408989,
      "epoch": 0.766505080476823,
      "grad_norm": 0.157009094953537,
      "learning_rate": 7.445333824025517e-05,
      "loss": 0.8389,
      "mean_token_accuracy": 0.7590945810079575,
      "num_tokens": 265906262.0,
      "step": 8331
    },
    {
      "entropy": 0.6446575485169888,
      "epoch": 0.7665970868482642,
      "grad_norm": 0.15098457038402557,
      "learning_rate": 7.445027141411354e-05,
      "loss": 0.6213,
      "mean_token_accuracy": 0.8141707591712475,
      "num_tokens": 265937904.0,
      "step": 8332
    },
    {
      "entropy": 0.7695706393569708,
      "epoch": 0.7666890932197055,
      "grad_norm": 0.15674686431884766,
      "learning_rate": 7.44472045879719e-05,
      "loss": 0.7692,
      "mean_token_accuracy": 0.7696376219391823,
      "num_tokens": 265969963.0,
      "step": 8333
    },
    {
      "entropy": 0.7019004449248314,
      "epoch": 0.7667810995911467,
      "grad_norm": 0.1583695411682129,
      "learning_rate": 7.444413776183028e-05,
      "loss": 0.6833,
      "mean_token_accuracy": 0.7992760762572289,
      "num_tokens": 266002485.0,
      "step": 8334
    },
    {
      "entropy": 0.5630377801135182,
      "epoch": 0.766873105962588,
      "grad_norm": 0.15089310705661774,
      "learning_rate": 7.444107093568867e-05,
      "loss": 0.5631,
      "mean_token_accuracy": 0.8289602287113667,
      "num_tokens": 266035223.0,
      "step": 8335
    },
    {
      "entropy": 0.6203255215659738,
      "epoch": 0.7669651123340291,
      "grad_norm": 0.1587824523448944,
      "learning_rate": 7.443800410954703e-05,
      "loss": 0.609,
      "mean_token_accuracy": 0.8196348063647747,
      "num_tokens": 266067636.0,
      "step": 8336
    },
    {
      "entropy": 0.6250579366460443,
      "epoch": 0.7670571187054703,
      "grad_norm": 0.1487196385860443,
      "learning_rate": 7.44349372834054e-05,
      "loss": 0.6061,
      "mean_token_accuracy": 0.820033110678196,
      "num_tokens": 266099552.0,
      "step": 8337
    },
    {
      "entropy": 0.7464414052665234,
      "epoch": 0.7671491250769116,
      "grad_norm": 0.1606578826904297,
      "learning_rate": 7.443187045726378e-05,
      "loss": 0.724,
      "mean_token_accuracy": 0.7845488414168358,
      "num_tokens": 266131726.0,
      "step": 8338
    },
    {
      "entropy": 0.6514873653650284,
      "epoch": 0.7672411314483528,
      "grad_norm": 0.15535078942775726,
      "learning_rate": 7.442880363112215e-05,
      "loss": 0.631,
      "mean_token_accuracy": 0.8116038143634796,
      "num_tokens": 266163447.0,
      "step": 8339
    },
    {
      "entropy": 0.6649088272824883,
      "epoch": 0.7673331378197941,
      "grad_norm": 0.14759600162506104,
      "learning_rate": 7.442573680498053e-05,
      "loss": 0.6483,
      "mean_token_accuracy": 0.8098660446703434,
      "num_tokens": 266195505.0,
      "step": 8340
    },
    {
      "entropy": 0.7976309061050415,
      "epoch": 0.7674251441912352,
      "grad_norm": 0.15919998288154602,
      "learning_rate": 7.44226699788389e-05,
      "loss": 0.7729,
      "mean_token_accuracy": 0.7714953981339931,
      "num_tokens": 266226848.0,
      "step": 8341
    },
    {
      "entropy": 0.7108211982995272,
      "epoch": 0.7675171505626764,
      "grad_norm": 0.16107112169265747,
      "learning_rate": 7.441960315269728e-05,
      "loss": 0.6914,
      "mean_token_accuracy": 0.7936585545539856,
      "num_tokens": 266258153.0,
      "step": 8342
    },
    {
      "entropy": 0.8297690823674202,
      "epoch": 0.7676091569341177,
      "grad_norm": 0.15877124667167664,
      "learning_rate": 7.441653632655565e-05,
      "loss": 0.8118,
      "mean_token_accuracy": 0.7593040391802788,
      "num_tokens": 266290506.0,
      "step": 8343
    },
    {
      "entropy": 0.7665696749463677,
      "epoch": 0.7677011633055589,
      "grad_norm": 0.16903752088546753,
      "learning_rate": 7.441346950041402e-05,
      "loss": 0.7364,
      "mean_token_accuracy": 0.7795450948178768,
      "num_tokens": 266321667.0,
      "step": 8344
    },
    {
      "entropy": 0.7519179452210665,
      "epoch": 0.7677931696770002,
      "grad_norm": 0.14469265937805176,
      "learning_rate": 7.44104026742724e-05,
      "loss": 0.728,
      "mean_token_accuracy": 0.7837244234979153,
      "num_tokens": 266353268.0,
      "step": 8345
    },
    {
      "entropy": 0.7066718228161335,
      "epoch": 0.7678851760484413,
      "grad_norm": 0.15169256925582886,
      "learning_rate": 7.440733584813078e-05,
      "loss": 0.688,
      "mean_token_accuracy": 0.7935218997299671,
      "num_tokens": 266385261.0,
      "step": 8346
    },
    {
      "entropy": 0.779655996710062,
      "epoch": 0.7679771824198826,
      "grad_norm": 0.15377488732337952,
      "learning_rate": 7.440426902198915e-05,
      "loss": 0.7799,
      "mean_token_accuracy": 0.7722255364060402,
      "num_tokens": 266417859.0,
      "step": 8347
    },
    {
      "entropy": 0.7509490475058556,
      "epoch": 0.7680691887913238,
      "grad_norm": 0.15091709792613983,
      "learning_rate": 7.440120219584752e-05,
      "loss": 0.751,
      "mean_token_accuracy": 0.7814594320952892,
      "num_tokens": 266450365.0,
      "step": 8348
    },
    {
      "entropy": 0.8135167676955462,
      "epoch": 0.768161195162765,
      "grad_norm": 0.16695556044578552,
      "learning_rate": 7.439813536970588e-05,
      "loss": 0.8176,
      "mean_token_accuracy": 0.7598999068140984,
      "num_tokens": 266482151.0,
      "step": 8349
    },
    {
      "entropy": 0.7835085056722164,
      "epoch": 0.7682532015342063,
      "grad_norm": 0.15918025374412537,
      "learning_rate": 7.439506854356428e-05,
      "loss": 0.7923,
      "mean_token_accuracy": 0.7721270099282265,
      "num_tokens": 266513903.0,
      "step": 8350
    },
    {
      "entropy": 0.7870461903512478,
      "epoch": 0.7683452079056474,
      "grad_norm": 0.1459827423095703,
      "learning_rate": 7.439200171742265e-05,
      "loss": 0.7827,
      "mean_token_accuracy": 0.7675598561763763,
      "num_tokens": 266546524.0,
      "step": 8351
    },
    {
      "entropy": 0.5834093373268843,
      "epoch": 0.7684372142770887,
      "grad_norm": 0.13651293516159058,
      "learning_rate": 7.438893489128101e-05,
      "loss": 0.581,
      "mean_token_accuracy": 0.8257785439491272,
      "num_tokens": 266578898.0,
      "step": 8352
    },
    {
      "entropy": 0.6438481314107776,
      "epoch": 0.7685292206485299,
      "grad_norm": 0.14132794737815857,
      "learning_rate": 7.438586806513938e-05,
      "loss": 0.6401,
      "mean_token_accuracy": 0.8089532516896725,
      "num_tokens": 266610800.0,
      "step": 8353
    },
    {
      "entropy": 0.8305756729096174,
      "epoch": 0.7686212270199712,
      "grad_norm": 0.1588711142539978,
      "learning_rate": 7.438280123899776e-05,
      "loss": 0.8219,
      "mean_token_accuracy": 0.757412564009428,
      "num_tokens": 266642373.0,
      "step": 8354
    },
    {
      "entropy": 0.7732295971363783,
      "epoch": 0.7687132333914124,
      "grad_norm": 0.16805040836334229,
      "learning_rate": 7.437973441285615e-05,
      "loss": 0.7728,
      "mean_token_accuracy": 0.7720735482871532,
      "num_tokens": 266673354.0,
      "step": 8355
    },
    {
      "entropy": 0.7080164942890406,
      "epoch": 0.7688052397628535,
      "grad_norm": 0.14931796491146088,
      "learning_rate": 7.437666758671451e-05,
      "loss": 0.699,
      "mean_token_accuracy": 0.7907474972307682,
      "num_tokens": 266705478.0,
      "step": 8356
    },
    {
      "entropy": 0.7381488606333733,
      "epoch": 0.7688972461342948,
      "grad_norm": 0.15052878856658936,
      "learning_rate": 7.437360076057288e-05,
      "loss": 0.7385,
      "mean_token_accuracy": 0.782787449657917,
      "num_tokens": 266737040.0,
      "step": 8357
    },
    {
      "entropy": 0.7216777577996254,
      "epoch": 0.768989252505736,
      "grad_norm": 0.1579332947731018,
      "learning_rate": 7.437053393443126e-05,
      "loss": 0.7003,
      "mean_token_accuracy": 0.7968901060521603,
      "num_tokens": 266768511.0,
      "step": 8358
    },
    {
      "entropy": 0.8085314258933067,
      "epoch": 0.7690812588771773,
      "grad_norm": 0.15597745776176453,
      "learning_rate": 7.436746710828963e-05,
      "loss": 0.7808,
      "mean_token_accuracy": 0.7711258754134178,
      "num_tokens": 266799919.0,
      "step": 8359
    },
    {
      "entropy": 0.7236270178109407,
      "epoch": 0.7691732652486185,
      "grad_norm": 0.1564657837152481,
      "learning_rate": 7.436440028214801e-05,
      "loss": 0.705,
      "mean_token_accuracy": 0.7913822382688522,
      "num_tokens": 266831954.0,
      "step": 8360
    },
    {
      "entropy": 0.7879636697471142,
      "epoch": 0.7692652716200596,
      "grad_norm": 0.1522190123796463,
      "learning_rate": 7.43613334560064e-05,
      "loss": 0.7814,
      "mean_token_accuracy": 0.768945287913084,
      "num_tokens": 266863959.0,
      "step": 8361
    },
    {
      "entropy": 0.7010805252939463,
      "epoch": 0.7693572779915009,
      "grad_norm": 0.14483608305454254,
      "learning_rate": 7.435826662986476e-05,
      "loss": 0.699,
      "mean_token_accuracy": 0.7925965003669262,
      "num_tokens": 266896087.0,
      "step": 8362
    },
    {
      "entropy": 0.748110705986619,
      "epoch": 0.7694492843629421,
      "grad_norm": 0.15323717892169952,
      "learning_rate": 7.435519980372313e-05,
      "loss": 0.7352,
      "mean_token_accuracy": 0.7839967459440231,
      "num_tokens": 266928043.0,
      "step": 8363
    },
    {
      "entropy": 0.6732020992785692,
      "epoch": 0.7695412907343834,
      "grad_norm": 0.15177102386951447,
      "learning_rate": 7.43521329775815e-05,
      "loss": 0.6602,
      "mean_token_accuracy": 0.8060996830463409,
      "num_tokens": 266960005.0,
      "step": 8364
    },
    {
      "entropy": 0.7234818544238806,
      "epoch": 0.7696332971058246,
      "grad_norm": 0.15801899135112762,
      "learning_rate": 7.434906615143988e-05,
      "loss": 0.7106,
      "mean_token_accuracy": 0.7895690500736237,
      "num_tokens": 266991870.0,
      "step": 8365
    },
    {
      "entropy": 0.779020631685853,
      "epoch": 0.7697253034772658,
      "grad_norm": 0.16592548787593842,
      "learning_rate": 7.434599932529826e-05,
      "loss": 0.7829,
      "mean_token_accuracy": 0.7687381356954575,
      "num_tokens": 267023807.0,
      "step": 8366
    },
    {
      "entropy": 0.6861949134618044,
      "epoch": 0.769817309848707,
      "grad_norm": 0.15504632890224457,
      "learning_rate": 7.434293249915663e-05,
      "loss": 0.6841,
      "mean_token_accuracy": 0.7979201972484589,
      "num_tokens": 267055371.0,
      "step": 8367
    },
    {
      "entropy": 0.6786569897085428,
      "epoch": 0.7699093162201482,
      "grad_norm": 0.1560639888048172,
      "learning_rate": 7.4339865673015e-05,
      "loss": 0.6654,
      "mean_token_accuracy": 0.807722769677639,
      "num_tokens": 267087630.0,
      "step": 8368
    },
    {
      "entropy": 0.7054673042148352,
      "epoch": 0.7700013225915895,
      "grad_norm": 0.1526266485452652,
      "learning_rate": 7.433679884687338e-05,
      "loss": 0.6861,
      "mean_token_accuracy": 0.7976271696388721,
      "num_tokens": 267119933.0,
      "step": 8369
    },
    {
      "entropy": 0.7660769857466221,
      "epoch": 0.7700933289630307,
      "grad_norm": 0.20062820613384247,
      "learning_rate": 7.433373202073175e-05,
      "loss": 0.8114,
      "mean_token_accuracy": 0.7715972177684307,
      "num_tokens": 267151550.0,
      "step": 8370
    },
    {
      "entropy": 0.6571659408509731,
      "epoch": 0.7701853353344719,
      "grad_norm": 0.1633276641368866,
      "learning_rate": 7.433066519459013e-05,
      "loss": 0.6748,
      "mean_token_accuracy": 0.8032484017312527,
      "num_tokens": 267183593.0,
      "step": 8371
    },
    {
      "entropy": 0.6991027807816863,
      "epoch": 0.7702773417059131,
      "grad_norm": 0.15211020410060883,
      "learning_rate": 7.43275983684485e-05,
      "loss": 0.6706,
      "mean_token_accuracy": 0.8042386770248413,
      "num_tokens": 267215480.0,
      "step": 8372
    },
    {
      "entropy": 0.7310404051095247,
      "epoch": 0.7703693480773544,
      "grad_norm": 0.15456819534301758,
      "learning_rate": 7.432453154230688e-05,
      "loss": 0.7127,
      "mean_token_accuracy": 0.7905169874429703,
      "num_tokens": 267246530.0,
      "step": 8373
    },
    {
      "entropy": 0.7456795834004879,
      "epoch": 0.7704613544487956,
      "grad_norm": 0.1552474945783615,
      "learning_rate": 7.432146471616524e-05,
      "loss": 0.7288,
      "mean_token_accuracy": 0.7899028696119785,
      "num_tokens": 267278576.0,
      "step": 8374
    },
    {
      "entropy": 0.7615787796676159,
      "epoch": 0.7705533608202368,
      "grad_norm": 0.15019753575325012,
      "learning_rate": 7.431839789002361e-05,
      "loss": 0.7388,
      "mean_token_accuracy": 0.7843778878450394,
      "num_tokens": 267310335.0,
      "step": 8375
    },
    {
      "entropy": 0.7517201546579599,
      "epoch": 0.770645367191678,
      "grad_norm": 0.15358158946037292,
      "learning_rate": 7.431533106388199e-05,
      "loss": 0.7415,
      "mean_token_accuracy": 0.7779340818524361,
      "num_tokens": 267342673.0,
      "step": 8376
    },
    {
      "entropy": 0.6661055218428373,
      "epoch": 0.7707373735631192,
      "grad_norm": 0.14881256222724915,
      "learning_rate": 7.431226423774037e-05,
      "loss": 0.6434,
      "mean_token_accuracy": 0.8071583323180676,
      "num_tokens": 267374591.0,
      "step": 8377
    },
    {
      "entropy": 0.8148208912461996,
      "epoch": 0.7708293799345605,
      "grad_norm": 0.16142843663692474,
      "learning_rate": 7.430919741159874e-05,
      "loss": 0.8047,
      "mean_token_accuracy": 0.7666492834687233,
      "num_tokens": 267406969.0,
      "step": 8378
    },
    {
      "entropy": 0.7178295273333788,
      "epoch": 0.7709213863060017,
      "grad_norm": 0.14862293004989624,
      "learning_rate": 7.430613058545711e-05,
      "loss": 0.7016,
      "mean_token_accuracy": 0.7926412411034107,
      "num_tokens": 267439006.0,
      "step": 8379
    },
    {
      "entropy": 0.7361892033368349,
      "epoch": 0.771013392677443,
      "grad_norm": 0.16103537380695343,
      "learning_rate": 7.430306375931548e-05,
      "loss": 0.7331,
      "mean_token_accuracy": 0.7853303216397762,
      "num_tokens": 267471151.0,
      "step": 8380
    },
    {
      "entropy": 0.7362309396266937,
      "epoch": 0.7711053990488841,
      "grad_norm": 0.16253434121608734,
      "learning_rate": 7.429999693317386e-05,
      "loss": 0.737,
      "mean_token_accuracy": 0.7820926532149315,
      "num_tokens": 267502628.0,
      "step": 8381
    },
    {
      "entropy": 0.839718010276556,
      "epoch": 0.7711974054203253,
      "grad_norm": 0.15543538331985474,
      "learning_rate": 7.429693010703224e-05,
      "loss": 0.8291,
      "mean_token_accuracy": 0.7520501725375652,
      "num_tokens": 267533841.0,
      "step": 8382
    },
    {
      "entropy": 0.6607535053044558,
      "epoch": 0.7712894117917666,
      "grad_norm": 0.16534920036792755,
      "learning_rate": 7.429386328089061e-05,
      "loss": 0.6536,
      "mean_token_accuracy": 0.807542085647583,
      "num_tokens": 267565633.0,
      "step": 8383
    },
    {
      "entropy": 0.6823731195181608,
      "epoch": 0.7713814181632078,
      "grad_norm": 0.14687055349349976,
      "learning_rate": 7.429079645474898e-05,
      "loss": 0.6638,
      "mean_token_accuracy": 0.801911935210228,
      "num_tokens": 267597499.0,
      "step": 8384
    },
    {
      "entropy": 0.7188094612210989,
      "epoch": 0.7714734245346491,
      "grad_norm": 0.1645669788122177,
      "learning_rate": 7.428772962860736e-05,
      "loss": 0.6995,
      "mean_token_accuracy": 0.7936580181121826,
      "num_tokens": 267630006.0,
      "step": 8385
    },
    {
      "entropy": 0.7523933444172144,
      "epoch": 0.7715654309060902,
      "grad_norm": 0.1695871204137802,
      "learning_rate": 7.428466280246573e-05,
      "loss": 0.7407,
      "mean_token_accuracy": 0.7835961505770683,
      "num_tokens": 267661448.0,
      "step": 8386
    },
    {
      "entropy": 0.7812820728868246,
      "epoch": 0.7716574372775314,
      "grad_norm": 0.1635599136352539,
      "learning_rate": 7.428159597632411e-05,
      "loss": 0.7721,
      "mean_token_accuracy": 0.7736227586865425,
      "num_tokens": 267693180.0,
      "step": 8387
    },
    {
      "entropy": 0.7909341994673014,
      "epoch": 0.7717494436489727,
      "grad_norm": 0.16374804079532623,
      "learning_rate": 7.427852915018248e-05,
      "loss": 0.7878,
      "mean_token_accuracy": 0.767100241035223,
      "num_tokens": 267725432.0,
      "step": 8388
    },
    {
      "entropy": 0.781392976641655,
      "epoch": 0.7718414500204139,
      "grad_norm": 0.15995852649211884,
      "learning_rate": 7.427546232404086e-05,
      "loss": 0.7832,
      "mean_token_accuracy": 0.7706314921379089,
      "num_tokens": 267757643.0,
      "step": 8389
    },
    {
      "entropy": 0.6286614276468754,
      "epoch": 0.7719334563918552,
      "grad_norm": 0.15534387528896332,
      "learning_rate": 7.427239549789922e-05,
      "loss": 0.6389,
      "mean_token_accuracy": 0.8102282099425793,
      "num_tokens": 267789423.0,
      "step": 8390
    },
    {
      "entropy": 0.7082630582153797,
      "epoch": 0.7720254627632963,
      "grad_norm": 0.1587345004081726,
      "learning_rate": 7.426932867175759e-05,
      "loss": 0.6951,
      "mean_token_accuracy": 0.7956516481935978,
      "num_tokens": 267821565.0,
      "step": 8391
    },
    {
      "entropy": 0.7853483837097883,
      "epoch": 0.7721174691347376,
      "grad_norm": 0.1617041528224945,
      "learning_rate": 7.426626184561599e-05,
      "loss": 0.7805,
      "mean_token_accuracy": 0.7714457102119923,
      "num_tokens": 267853769.0,
      "step": 8392
    },
    {
      "entropy": 0.7127526625990868,
      "epoch": 0.7722094755061788,
      "grad_norm": 0.14588497579097748,
      "learning_rate": 7.426319501947436e-05,
      "loss": 0.6954,
      "mean_token_accuracy": 0.7936808690428734,
      "num_tokens": 267885232.0,
      "step": 8393
    },
    {
      "entropy": 0.6181660573929548,
      "epoch": 0.77230148187762,
      "grad_norm": 0.14580833911895752,
      "learning_rate": 7.426012819333272e-05,
      "loss": 0.6007,
      "mean_token_accuracy": 0.8203944116830826,
      "num_tokens": 267917069.0,
      "step": 8394
    },
    {
      "entropy": 0.7774047702550888,
      "epoch": 0.7723934882490613,
      "grad_norm": 0.17007975280284882,
      "learning_rate": 7.425706136719109e-05,
      "loss": 0.7858,
      "mean_token_accuracy": 0.7729325853288174,
      "num_tokens": 267948859.0,
      "step": 8395
    },
    {
      "entropy": 0.6570176146924496,
      "epoch": 0.7724854946205024,
      "grad_norm": 0.15304981172084808,
      "learning_rate": 7.425399454104947e-05,
      "loss": 0.6418,
      "mean_token_accuracy": 0.811908170580864,
      "num_tokens": 267980920.0,
      "step": 8396
    },
    {
      "entropy": 0.7374170143157244,
      "epoch": 0.7725775009919437,
      "grad_norm": 0.16340811550617218,
      "learning_rate": 7.425092771490785e-05,
      "loss": 0.7247,
      "mean_token_accuracy": 0.7855722308158875,
      "num_tokens": 268012282.0,
      "step": 8397
    },
    {
      "entropy": 0.7774721495807171,
      "epoch": 0.7726695073633849,
      "grad_norm": 0.1516706645488739,
      "learning_rate": 7.424786088876622e-05,
      "loss": 0.7771,
      "mean_token_accuracy": 0.7698515504598618,
      "num_tokens": 268044643.0,
      "step": 8398
    },
    {
      "entropy": 0.7082904018461704,
      "epoch": 0.7727615137348262,
      "grad_norm": 0.15444369614124298,
      "learning_rate": 7.424479406262459e-05,
      "loss": 0.6937,
      "mean_token_accuracy": 0.7902578972280025,
      "num_tokens": 268076582.0,
      "step": 8399
    },
    {
      "entropy": 0.698989525437355,
      "epoch": 0.7728535201062674,
      "grad_norm": 0.16233238577842712,
      "learning_rate": 7.424172723648297e-05,
      "loss": 0.6716,
      "mean_token_accuracy": 0.7985612154006958,
      "num_tokens": 268108673.0,
      "step": 8400
    },
    {
      "entropy": 0.7437477223575115,
      "epoch": 0.7729455264777085,
      "grad_norm": 0.16020254790782928,
      "learning_rate": 7.423866041034134e-05,
      "loss": 0.7358,
      "mean_token_accuracy": 0.7831477522850037,
      "num_tokens": 268140097.0,
      "step": 8401
    },
    {
      "entropy": 0.6760464897379279,
      "epoch": 0.7730375328491498,
      "grad_norm": 0.15099237859249115,
      "learning_rate": 7.423559358419972e-05,
      "loss": 0.6643,
      "mean_token_accuracy": 0.8015177249908447,
      "num_tokens": 268172578.0,
      "step": 8402
    },
    {
      "entropy": 0.7208482464775443,
      "epoch": 0.773129539220591,
      "grad_norm": 0.1616649329662323,
      "learning_rate": 7.423252675805809e-05,
      "loss": 0.7097,
      "mean_token_accuracy": 0.7909321188926697,
      "num_tokens": 268204554.0,
      "step": 8403
    },
    {
      "entropy": 0.6507297614589334,
      "epoch": 0.7732215455920323,
      "grad_norm": 0.14895480871200562,
      "learning_rate": 7.422945993191647e-05,
      "loss": 0.6308,
      "mean_token_accuracy": 0.8121646195650101,
      "num_tokens": 268237270.0,
      "step": 8404
    },
    {
      "entropy": 0.7136127837002277,
      "epoch": 0.7733135519634735,
      "grad_norm": 0.15435150265693665,
      "learning_rate": 7.422639310577484e-05,
      "loss": 0.6788,
      "mean_token_accuracy": 0.7917367406189442,
      "num_tokens": 268269139.0,
      "step": 8405
    },
    {
      "entropy": 0.7983676381409168,
      "epoch": 0.7734055583349146,
      "grad_norm": 0.15589064359664917,
      "learning_rate": 7.42233262796332e-05,
      "loss": 0.7876,
      "mean_token_accuracy": 0.7664480283856392,
      "num_tokens": 268300641.0,
      "step": 8406
    },
    {
      "entropy": 0.7326412294059992,
      "epoch": 0.7734975647063559,
      "grad_norm": 0.15702013671398163,
      "learning_rate": 7.422025945349159e-05,
      "loss": 0.7223,
      "mean_token_accuracy": 0.7930477634072304,
      "num_tokens": 268332664.0,
      "step": 8407
    },
    {
      "entropy": 0.744467981159687,
      "epoch": 0.7735895710777971,
      "grad_norm": 0.1516851782798767,
      "learning_rate": 7.421719262734997e-05,
      "loss": 0.7394,
      "mean_token_accuracy": 0.7802860587835312,
      "num_tokens": 268364447.0,
      "step": 8408
    },
    {
      "entropy": 0.6528158709406853,
      "epoch": 0.7736815774492384,
      "grad_norm": 0.15242217481136322,
      "learning_rate": 7.421412580120834e-05,
      "loss": 0.6363,
      "mean_token_accuracy": 0.8104123063385487,
      "num_tokens": 268396395.0,
      "step": 8409
    },
    {
      "entropy": 0.8747371733188629,
      "epoch": 0.7737735838206796,
      "grad_norm": 0.16340754926204681,
      "learning_rate": 7.42110589750667e-05,
      "loss": 0.8649,
      "mean_token_accuracy": 0.7425920218229294,
      "num_tokens": 268428219.0,
      "step": 8410
    },
    {
      "entropy": 0.6805642284452915,
      "epoch": 0.7738655901921208,
      "grad_norm": 0.15514503419399261,
      "learning_rate": 7.420799214892507e-05,
      "loss": 0.6856,
      "mean_token_accuracy": 0.7941848672926426,
      "num_tokens": 268460760.0,
      "step": 8411
    },
    {
      "entropy": 0.8018915485590696,
      "epoch": 0.773957596563562,
      "grad_norm": 0.17150849103927612,
      "learning_rate": 7.420492532278345e-05,
      "loss": 0.8062,
      "mean_token_accuracy": 0.7656481973826885,
      "num_tokens": 268492396.0,
      "step": 8412
    },
    {
      "entropy": 0.811549074947834,
      "epoch": 0.7740496029350032,
      "grad_norm": 0.16509371995925903,
      "learning_rate": 7.420185849664183e-05,
      "loss": 0.8171,
      "mean_token_accuracy": 0.768900703638792,
      "num_tokens": 268524137.0,
      "step": 8413
    },
    {
      "entropy": 0.804741969332099,
      "epoch": 0.7741416093064445,
      "grad_norm": 0.15574100613594055,
      "learning_rate": 7.41987916705002e-05,
      "loss": 0.7892,
      "mean_token_accuracy": 0.770672045648098,
      "num_tokens": 268556588.0,
      "step": 8414
    },
    {
      "entropy": 0.6789482589811087,
      "epoch": 0.7742336156778857,
      "grad_norm": 0.15257465839385986,
      "learning_rate": 7.419572484435857e-05,
      "loss": 0.6697,
      "mean_token_accuracy": 0.8008944615721703,
      "num_tokens": 268589336.0,
      "step": 8415
    },
    {
      "entropy": 0.7495875135064125,
      "epoch": 0.7743256220493269,
      "grad_norm": 0.15332476794719696,
      "learning_rate": 7.419265801821695e-05,
      "loss": 0.7368,
      "mean_token_accuracy": 0.7812209986150265,
      "num_tokens": 268621415.0,
      "step": 8416
    },
    {
      "entropy": 0.706055847927928,
      "epoch": 0.7744176284207681,
      "grad_norm": 0.16794997453689575,
      "learning_rate": 7.418959119207532e-05,
      "loss": 0.6824,
      "mean_token_accuracy": 0.7991289906203747,
      "num_tokens": 268652911.0,
      "step": 8417
    },
    {
      "entropy": 0.773205304518342,
      "epoch": 0.7745096347922094,
      "grad_norm": 0.1523023396730423,
      "learning_rate": 7.41865243659337e-05,
      "loss": 0.7437,
      "mean_token_accuracy": 0.7784441001713276,
      "num_tokens": 268684718.0,
      "step": 8418
    },
    {
      "entropy": 0.7247454188764095,
      "epoch": 0.7746016411636506,
      "grad_norm": 0.15889571607112885,
      "learning_rate": 7.418345753979207e-05,
      "loss": 0.6893,
      "mean_token_accuracy": 0.7917864210903645,
      "num_tokens": 268716471.0,
      "step": 8419
    },
    {
      "entropy": 0.6239937506616116,
      "epoch": 0.7746936475350918,
      "grad_norm": 0.1607053577899933,
      "learning_rate": 7.418039071365045e-05,
      "loss": 0.6058,
      "mean_token_accuracy": 0.814903263002634,
      "num_tokens": 268747354.0,
      "step": 8420
    },
    {
      "entropy": 0.8182337768375874,
      "epoch": 0.774785653906533,
      "grad_norm": 0.1533818393945694,
      "learning_rate": 7.417732388750882e-05,
      "loss": 0.7899,
      "mean_token_accuracy": 0.7646048963069916,
      "num_tokens": 268779156.0,
      "step": 8421
    },
    {
      "entropy": 0.676281651481986,
      "epoch": 0.7748776602779742,
      "grad_norm": 0.145756334066391,
      "learning_rate": 7.417425706136719e-05,
      "loss": 0.6648,
      "mean_token_accuracy": 0.8038845546543598,
      "num_tokens": 268811824.0,
      "step": 8422
    },
    {
      "entropy": 0.624363262206316,
      "epoch": 0.7749696666494155,
      "grad_norm": 0.15423481166362762,
      "learning_rate": 7.417119023522557e-05,
      "loss": 0.6086,
      "mean_token_accuracy": 0.8195128068327904,
      "num_tokens": 268844058.0,
      "step": 8423
    },
    {
      "entropy": 0.7527207992970943,
      "epoch": 0.7750616730208567,
      "grad_norm": 0.15597715973854065,
      "learning_rate": 7.416812340908395e-05,
      "loss": 0.74,
      "mean_token_accuracy": 0.7835301011800766,
      "num_tokens": 268875604.0,
      "step": 8424
    },
    {
      "entropy": 0.7664725463837385,
      "epoch": 0.775153679392298,
      "grad_norm": 0.15973465144634247,
      "learning_rate": 7.416505658294232e-05,
      "loss": 0.7539,
      "mean_token_accuracy": 0.7776155136525631,
      "num_tokens": 268908029.0,
      "step": 8425
    },
    {
      "entropy": 0.647901950404048,
      "epoch": 0.7752456857637391,
      "grad_norm": 0.15824313461780548,
      "learning_rate": 7.416198975680068e-05,
      "loss": 0.6518,
      "mean_token_accuracy": 0.808040514588356,
      "num_tokens": 268940459.0,
      "step": 8426
    },
    {
      "entropy": 0.6865790598094463,
      "epoch": 0.7753376921351803,
      "grad_norm": 0.15223608911037445,
      "learning_rate": 7.415892293065907e-05,
      "loss": 0.6833,
      "mean_token_accuracy": 0.7954586446285248,
      "num_tokens": 268973004.0,
      "step": 8427
    },
    {
      "entropy": 0.6617205729708076,
      "epoch": 0.7754296985066216,
      "grad_norm": 0.16892513632774353,
      "learning_rate": 7.415585610451743e-05,
      "loss": 0.6636,
      "mean_token_accuracy": 0.8024937957525253,
      "num_tokens": 269004714.0,
      "step": 8428
    },
    {
      "entropy": 0.6701563904061913,
      "epoch": 0.7755217048780628,
      "grad_norm": 0.1651470959186554,
      "learning_rate": 7.415278927837582e-05,
      "loss": 0.6755,
      "mean_token_accuracy": 0.7967627011239529,
      "num_tokens": 269036436.0,
      "step": 8429
    },
    {
      "entropy": 0.6397349163889885,
      "epoch": 0.7756137112495041,
      "grad_norm": 0.1605667620897293,
      "learning_rate": 7.414972245223418e-05,
      "loss": 0.6352,
      "mean_token_accuracy": 0.8115903958678246,
      "num_tokens": 269069003.0,
      "step": 8430
    },
    {
      "entropy": 0.7480891365557909,
      "epoch": 0.7757057176209452,
      "grad_norm": 0.17378509044647217,
      "learning_rate": 7.414665562609256e-05,
      "loss": 0.7647,
      "mean_token_accuracy": 0.7769638188183308,
      "num_tokens": 269101445.0,
      "step": 8431
    },
    {
      "entropy": 0.6993854241445661,
      "epoch": 0.7757977239923864,
      "grad_norm": 0.1493859887123108,
      "learning_rate": 7.414358879995093e-05,
      "loss": 0.6987,
      "mean_token_accuracy": 0.7969317212700844,
      "num_tokens": 269133741.0,
      "step": 8432
    },
    {
      "entropy": 0.7414909256622195,
      "epoch": 0.7758897303638277,
      "grad_norm": 0.15440583229064941,
      "learning_rate": 7.41405219738093e-05,
      "loss": 0.7312,
      "mean_token_accuracy": 0.7865445613861084,
      "num_tokens": 269165924.0,
      "step": 8433
    },
    {
      "entropy": 0.7491185963153839,
      "epoch": 0.7759817367352689,
      "grad_norm": 0.15726524591445923,
      "learning_rate": 7.413745514766768e-05,
      "loss": 0.743,
      "mean_token_accuracy": 0.7796621471643448,
      "num_tokens": 269197881.0,
      "step": 8434
    },
    {
      "entropy": 0.7120674252510071,
      "epoch": 0.7760737431067102,
      "grad_norm": 0.15494239330291748,
      "learning_rate": 7.413438832152606e-05,
      "loss": 0.6836,
      "mean_token_accuracy": 0.7957655601203442,
      "num_tokens": 269229619.0,
      "step": 8435
    },
    {
      "entropy": 0.7821333706378937,
      "epoch": 0.7761657494781513,
      "grad_norm": 0.15843357145786285,
      "learning_rate": 7.413132149538443e-05,
      "loss": 0.7732,
      "mean_token_accuracy": 0.7794236205518246,
      "num_tokens": 269261485.0,
      "step": 8436
    },
    {
      "entropy": 0.8496836870908737,
      "epoch": 0.7762577558495926,
      "grad_norm": 0.16164863109588623,
      "learning_rate": 7.41282546692428e-05,
      "loss": 0.8361,
      "mean_token_accuracy": 0.759065069258213,
      "num_tokens": 269293162.0,
      "step": 8437
    },
    {
      "entropy": 0.7163041215389967,
      "epoch": 0.7763497622210338,
      "grad_norm": 0.15455052256584167,
      "learning_rate": 7.412518784310118e-05,
      "loss": 0.6976,
      "mean_token_accuracy": 0.7955849766731262,
      "num_tokens": 269324573.0,
      "step": 8438
    },
    {
      "entropy": 0.7603961061686277,
      "epoch": 0.776441768592475,
      "grad_norm": 0.14962905645370483,
      "learning_rate": 7.412212101695956e-05,
      "loss": 0.7478,
      "mean_token_accuracy": 0.7774780206382275,
      "num_tokens": 269356825.0,
      "step": 8439
    },
    {
      "entropy": 0.721658450551331,
      "epoch": 0.7765337749639163,
      "grad_norm": 0.15850821137428284,
      "learning_rate": 7.411905419081793e-05,
      "loss": 0.7136,
      "mean_token_accuracy": 0.7853643223643303,
      "num_tokens": 269388485.0,
      "step": 8440
    },
    {
      "entropy": 0.6860156077891588,
      "epoch": 0.7766257813353574,
      "grad_norm": 0.15196369588375092,
      "learning_rate": 7.41159873646763e-05,
      "loss": 0.6673,
      "mean_token_accuracy": 0.8040055260062218,
      "num_tokens": 269420597.0,
      "step": 8441
    },
    {
      "entropy": 0.7453662604093552,
      "epoch": 0.7767177877067987,
      "grad_norm": 0.15306375920772552,
      "learning_rate": 7.411292053853467e-05,
      "loss": 0.7273,
      "mean_token_accuracy": 0.7843505889177322,
      "num_tokens": 269452246.0,
      "step": 8442
    },
    {
      "entropy": 0.8109282851219177,
      "epoch": 0.7768097940782399,
      "grad_norm": 0.1569751352071762,
      "learning_rate": 7.410985371239305e-05,
      "loss": 0.7935,
      "mean_token_accuracy": 0.7694814093410969,
      "num_tokens": 269484528.0,
      "step": 8443
    },
    {
      "entropy": 0.7253571376204491,
      "epoch": 0.7769018004496812,
      "grad_norm": 0.15636396408081055,
      "learning_rate": 7.410678688625143e-05,
      "loss": 0.7152,
      "mean_token_accuracy": 0.7856834307312965,
      "num_tokens": 269515675.0,
      "step": 8444
    },
    {
      "entropy": 0.7576875127851963,
      "epoch": 0.7769938068211224,
      "grad_norm": 0.16931787133216858,
      "learning_rate": 7.41037200601098e-05,
      "loss": 0.7682,
      "mean_token_accuracy": 0.7744540050625801,
      "num_tokens": 269546952.0,
      "step": 8445
    },
    {
      "entropy": 0.703269524499774,
      "epoch": 0.7770858131925635,
      "grad_norm": 0.15099526941776276,
      "learning_rate": 7.410065323396816e-05,
      "loss": 0.6956,
      "mean_token_accuracy": 0.7973133437335491,
      "num_tokens": 269579619.0,
      "step": 8446
    },
    {
      "entropy": 0.728420639410615,
      "epoch": 0.7771778195640048,
      "grad_norm": 0.15892623364925385,
      "learning_rate": 7.409758640782655e-05,
      "loss": 0.7059,
      "mean_token_accuracy": 0.7905838266015053,
      "num_tokens": 269611800.0,
      "step": 8447
    },
    {
      "entropy": 0.6598333939909935,
      "epoch": 0.777269825935446,
      "grad_norm": 0.14920447766780853,
      "learning_rate": 7.409451958168491e-05,
      "loss": 0.6568,
      "mean_token_accuracy": 0.8047174587845802,
      "num_tokens": 269643835.0,
      "step": 8448
    },
    {
      "entropy": 0.6449388330802321,
      "epoch": 0.7773618323068873,
      "grad_norm": 0.14903821051120758,
      "learning_rate": 7.40914527555433e-05,
      "loss": 0.6238,
      "mean_token_accuracy": 0.8118040449917316,
      "num_tokens": 269676037.0,
      "step": 8449
    },
    {
      "entropy": 0.7799408324062824,
      "epoch": 0.7774538386783285,
      "grad_norm": 0.16331203281879425,
      "learning_rate": 7.408838592940166e-05,
      "loss": 0.7593,
      "mean_token_accuracy": 0.7755193933844566,
      "num_tokens": 269707674.0,
      "step": 8450
    },
    {
      "entropy": 0.6844371100887656,
      "epoch": 0.7775458450497696,
      "grad_norm": 0.16319283843040466,
      "learning_rate": 7.408531910326004e-05,
      "loss": 0.6512,
      "mean_token_accuracy": 0.8027563989162445,
      "num_tokens": 269739513.0,
      "step": 8451
    },
    {
      "entropy": 0.8299889918416739,
      "epoch": 0.7776378514212109,
      "grad_norm": 0.1683456003665924,
      "learning_rate": 7.408225227711841e-05,
      "loss": 0.82,
      "mean_token_accuracy": 0.764246441423893,
      "num_tokens": 269770924.0,
      "step": 8452
    },
    {
      "entropy": 0.69473946839571,
      "epoch": 0.7777298577926521,
      "grad_norm": 0.15853320062160492,
      "learning_rate": 7.407918545097678e-05,
      "loss": 0.6938,
      "mean_token_accuracy": 0.7943790480494499,
      "num_tokens": 269802912.0,
      "step": 8453
    },
    {
      "entropy": 0.749951945617795,
      "epoch": 0.7778218641640934,
      "grad_norm": 0.16599124670028687,
      "learning_rate": 7.407611862483516e-05,
      "loss": 0.7334,
      "mean_token_accuracy": 0.7830946706235409,
      "num_tokens": 269835258.0,
      "step": 8454
    },
    {
      "entropy": 0.7060797568410635,
      "epoch": 0.7779138705355346,
      "grad_norm": 0.15782001614570618,
      "learning_rate": 7.407305179869354e-05,
      "loss": 0.6933,
      "mean_token_accuracy": 0.794991496950388,
      "num_tokens": 269867218.0,
      "step": 8455
    },
    {
      "entropy": 0.7606798368506134,
      "epoch": 0.7780058769069758,
      "grad_norm": 0.16436246037483215,
      "learning_rate": 7.406998497255191e-05,
      "loss": 0.7581,
      "mean_token_accuracy": 0.7799652777612209,
      "num_tokens": 269899779.0,
      "step": 8456
    },
    {
      "entropy": 0.7768608257174492,
      "epoch": 0.778097883278417,
      "grad_norm": 0.15776051580905914,
      "learning_rate": 7.406691814641028e-05,
      "loss": 0.7885,
      "mean_token_accuracy": 0.772331528365612,
      "num_tokens": 269931805.0,
      "step": 8457
    },
    {
      "entropy": 0.8023094311356544,
      "epoch": 0.7781898896498582,
      "grad_norm": 0.1557949334383011,
      "learning_rate": 7.406385132026866e-05,
      "loss": 0.8036,
      "mean_token_accuracy": 0.765748854726553,
      "num_tokens": 269963768.0,
      "step": 8458
    },
    {
      "entropy": 0.660432493314147,
      "epoch": 0.7782818960212995,
      "grad_norm": 0.1551281362771988,
      "learning_rate": 7.406078449412703e-05,
      "loss": 0.6485,
      "mean_token_accuracy": 0.808917049318552,
      "num_tokens": 269995486.0,
      "step": 8459
    },
    {
      "entropy": 0.8294745422899723,
      "epoch": 0.7783739023927407,
      "grad_norm": 0.159877210855484,
      "learning_rate": 7.405771766798541e-05,
      "loss": 0.8322,
      "mean_token_accuracy": 0.7561885118484497,
      "num_tokens": 270027342.0,
      "step": 8460
    },
    {
      "entropy": 0.6893805228173733,
      "epoch": 0.7784659087641819,
      "grad_norm": 0.16560526192188263,
      "learning_rate": 7.405465084184378e-05,
      "loss": 0.6933,
      "mean_token_accuracy": 0.7971676252782345,
      "num_tokens": 270058939.0,
      "step": 8461
    },
    {
      "entropy": 0.6560484943911433,
      "epoch": 0.7785579151356231,
      "grad_norm": 0.14560219645500183,
      "learning_rate": 7.405158401570216e-05,
      "loss": 0.6457,
      "mean_token_accuracy": 0.8097125664353371,
      "num_tokens": 270090464.0,
      "step": 8462
    },
    {
      "entropy": 0.8274881131947041,
      "epoch": 0.7786499215070644,
      "grad_norm": 0.15921354293823242,
      "learning_rate": 7.404851718956053e-05,
      "loss": 0.8297,
      "mean_token_accuracy": 0.7549255415797234,
      "num_tokens": 270122212.0,
      "step": 8463
    },
    {
      "entropy": 0.8583139702677727,
      "epoch": 0.7787419278785056,
      "grad_norm": 0.15971405804157257,
      "learning_rate": 7.40454503634189e-05,
      "loss": 0.865,
      "mean_token_accuracy": 0.7482916079461575,
      "num_tokens": 270154056.0,
      "step": 8464
    },
    {
      "entropy": 0.6062721610069275,
      "epoch": 0.7788339342499468,
      "grad_norm": 0.15652082860469818,
      "learning_rate": 7.404238353727728e-05,
      "loss": 0.6033,
      "mean_token_accuracy": 0.8164522051811218,
      "num_tokens": 270185827.0,
      "step": 8465
    },
    {
      "entropy": 0.7899483107030392,
      "epoch": 0.778925940621388,
      "grad_norm": 0.15123353898525238,
      "learning_rate": 7.403931671113566e-05,
      "loss": 0.7686,
      "mean_token_accuracy": 0.772488608956337,
      "num_tokens": 270217926.0,
      "step": 8466
    },
    {
      "entropy": 0.7451774161309004,
      "epoch": 0.7790179469928292,
      "grad_norm": 0.16442224383354187,
      "learning_rate": 7.403624988499402e-05,
      "loss": 0.7359,
      "mean_token_accuracy": 0.7820848263800144,
      "num_tokens": 270250395.0,
      "step": 8467
    },
    {
      "entropy": 0.7349152509123087,
      "epoch": 0.7791099533642705,
      "grad_norm": 0.159428671002388,
      "learning_rate": 7.403318305885239e-05,
      "loss": 0.7184,
      "mean_token_accuracy": 0.7865876220166683,
      "num_tokens": 270282273.0,
      "step": 8468
    },
    {
      "entropy": 0.7473154626786709,
      "epoch": 0.7792019597357117,
      "grad_norm": 0.15249498188495636,
      "learning_rate": 7.403011623271076e-05,
      "loss": 0.7317,
      "mean_token_accuracy": 0.7817752994596958,
      "num_tokens": 270314116.0,
      "step": 8469
    },
    {
      "entropy": 0.821124036796391,
      "epoch": 0.779293966107153,
      "grad_norm": 0.15880122780799866,
      "learning_rate": 7.402704940656916e-05,
      "loss": 0.807,
      "mean_token_accuracy": 0.7658600509166718,
      "num_tokens": 270345775.0,
      "step": 8470
    },
    {
      "entropy": 0.6152582271024585,
      "epoch": 0.7793859724785941,
      "grad_norm": 0.1558084636926651,
      "learning_rate": 7.402398258042752e-05,
      "loss": 0.6031,
      "mean_token_accuracy": 0.822352547198534,
      "num_tokens": 270378041.0,
      "step": 8471
    },
    {
      "entropy": 0.7573912721127272,
      "epoch": 0.7794779788500353,
      "grad_norm": 0.15942133963108063,
      "learning_rate": 7.402091575428589e-05,
      "loss": 0.7306,
      "mean_token_accuracy": 0.7812934517860413,
      "num_tokens": 270409158.0,
      "step": 8472
    },
    {
      "entropy": 0.7820795644074678,
      "epoch": 0.7795699852214766,
      "grad_norm": 0.16440118849277496,
      "learning_rate": 7.401784892814426e-05,
      "loss": 0.7886,
      "mean_token_accuracy": 0.7727218642830849,
      "num_tokens": 270441079.0,
      "step": 8473
    },
    {
      "entropy": 0.6023328276351094,
      "epoch": 0.7796619915929178,
      "grad_norm": 0.1477334350347519,
      "learning_rate": 7.401478210200264e-05,
      "loss": 0.6074,
      "mean_token_accuracy": 0.8193027675151825,
      "num_tokens": 270473452.0,
      "step": 8474
    },
    {
      "entropy": 0.6819146797060966,
      "epoch": 0.7797539979643591,
      "grad_norm": 0.1617940217256546,
      "learning_rate": 7.401171527586102e-05,
      "loss": 0.682,
      "mean_token_accuracy": 0.7973816134035587,
      "num_tokens": 270505435.0,
      "step": 8475
    },
    {
      "entropy": 0.7415379174053669,
      "epoch": 0.7798460043358002,
      "grad_norm": 0.1503383070230484,
      "learning_rate": 7.400864844971939e-05,
      "loss": 0.7236,
      "mean_token_accuracy": 0.7812758013606071,
      "num_tokens": 270537302.0,
      "step": 8476
    },
    {
      "entropy": 0.7329591037705541,
      "epoch": 0.7799380107072414,
      "grad_norm": 0.162094384431839,
      "learning_rate": 7.400558162357776e-05,
      "loss": 0.7329,
      "mean_token_accuracy": 0.7854157090187073,
      "num_tokens": 270568987.0,
      "step": 8477
    },
    {
      "entropy": 0.6281170938163996,
      "epoch": 0.7800300170786827,
      "grad_norm": 0.15970712900161743,
      "learning_rate": 7.400251479743614e-05,
      "loss": 0.6069,
      "mean_token_accuracy": 0.8172166459262371,
      "num_tokens": 270600607.0,
      "step": 8478
    },
    {
      "entropy": 0.6942751817405224,
      "epoch": 0.7801220234501239,
      "grad_norm": 0.15127450227737427,
      "learning_rate": 7.399944797129451e-05,
      "loss": 0.6602,
      "mean_token_accuracy": 0.8010799400508404,
      "num_tokens": 270632233.0,
      "step": 8479
    },
    {
      "entropy": 0.5796406287699938,
      "epoch": 0.7802140298215652,
      "grad_norm": 0.14087823033332825,
      "learning_rate": 7.399638114515289e-05,
      "loss": 0.5536,
      "mean_token_accuracy": 0.8326584026217461,
      "num_tokens": 270663820.0,
      "step": 8480
    },
    {
      "entropy": 0.7461374122649431,
      "epoch": 0.7803060361930063,
      "grad_norm": 0.1537722498178482,
      "learning_rate": 7.399331431901127e-05,
      "loss": 0.7338,
      "mean_token_accuracy": 0.7823897749185562,
      "num_tokens": 270696101.0,
      "step": 8481
    },
    {
      "entropy": 0.7235535820946097,
      "epoch": 0.7803980425644476,
      "grad_norm": 0.14775879681110382,
      "learning_rate": 7.399024749286964e-05,
      "loss": 0.7051,
      "mean_token_accuracy": 0.7885327786207199,
      "num_tokens": 270728082.0,
      "step": 8482
    },
    {
      "entropy": 0.710578840225935,
      "epoch": 0.7804900489358888,
      "grad_norm": 0.15709859132766724,
      "learning_rate": 7.3987180666728e-05,
      "loss": 0.6902,
      "mean_token_accuracy": 0.7955599315464497,
      "num_tokens": 270760207.0,
      "step": 8483
    },
    {
      "entropy": 0.6714707110077143,
      "epoch": 0.78058205530733,
      "grad_norm": 0.1521081030368805,
      "learning_rate": 7.398411384058637e-05,
      "loss": 0.6514,
      "mean_token_accuracy": 0.8052070327103138,
      "num_tokens": 270792303.0,
      "step": 8484
    },
    {
      "entropy": 0.7547136694192886,
      "epoch": 0.7806740616787713,
      "grad_norm": 0.15353070199489594,
      "learning_rate": 7.398104701444475e-05,
      "loss": 0.742,
      "mean_token_accuracy": 0.7807872779667377,
      "num_tokens": 270824160.0,
      "step": 8485
    },
    {
      "entropy": 0.6752267684787512,
      "epoch": 0.7807660680502124,
      "grad_norm": 0.15177907049655914,
      "learning_rate": 7.397798018830314e-05,
      "loss": 0.6653,
      "mean_token_accuracy": 0.8009885810315609,
      "num_tokens": 270856212.0,
      "step": 8486
    },
    {
      "entropy": 0.7236415408551693,
      "epoch": 0.7808580744216537,
      "grad_norm": 0.1551021784543991,
      "learning_rate": 7.39749133621615e-05,
      "loss": 0.7076,
      "mean_token_accuracy": 0.7911282926797867,
      "num_tokens": 270888239.0,
      "step": 8487
    },
    {
      "entropy": 0.8197584003210068,
      "epoch": 0.7809500807930949,
      "grad_norm": 0.16155073046684265,
      "learning_rate": 7.397184653601987e-05,
      "loss": 0.845,
      "mean_token_accuracy": 0.7525942772626877,
      "num_tokens": 270920318.0,
      "step": 8488
    },
    {
      "entropy": 0.6505640102550387,
      "epoch": 0.7810420871645362,
      "grad_norm": 0.15686604380607605,
      "learning_rate": 7.396877970987825e-05,
      "loss": 0.6559,
      "mean_token_accuracy": 0.803996704518795,
      "num_tokens": 270952352.0,
      "step": 8489
    },
    {
      "entropy": 0.7489370983093977,
      "epoch": 0.7811340935359774,
      "grad_norm": 0.16374242305755615,
      "learning_rate": 7.396571288373662e-05,
      "loss": 0.7674,
      "mean_token_accuracy": 0.7726884260773659,
      "num_tokens": 270983430.0,
      "step": 8490
    },
    {
      "entropy": 0.7370063737034798,
      "epoch": 0.7812260999074185,
      "grad_norm": 0.15555284917354584,
      "learning_rate": 7.3962646057595e-05,
      "loss": 0.7268,
      "mean_token_accuracy": 0.7846766151487827,
      "num_tokens": 271015554.0,
      "step": 8491
    },
    {
      "entropy": 0.7605577427893877,
      "epoch": 0.7813181062788598,
      "grad_norm": 0.16150332987308502,
      "learning_rate": 7.395957923145337e-05,
      "loss": 0.7382,
      "mean_token_accuracy": 0.7833043895661831,
      "num_tokens": 271046966.0,
      "step": 8492
    },
    {
      "entropy": 0.7652390394359827,
      "epoch": 0.781410112650301,
      "grad_norm": 0.16582337021827698,
      "learning_rate": 7.395651240531175e-05,
      "loss": 0.7703,
      "mean_token_accuracy": 0.7722799442708492,
      "num_tokens": 271079658.0,
      "step": 8493
    },
    {
      "entropy": 0.7359480950981379,
      "epoch": 0.7815021190217423,
      "grad_norm": 0.15548856556415558,
      "learning_rate": 7.395344557917012e-05,
      "loss": 0.7149,
      "mean_token_accuracy": 0.7870174422860146,
      "num_tokens": 271111847.0,
      "step": 8494
    },
    {
      "entropy": 0.813851110637188,
      "epoch": 0.7815941253931835,
      "grad_norm": 0.1645001620054245,
      "learning_rate": 7.395037875302849e-05,
      "loss": 0.7949,
      "mean_token_accuracy": 0.7679225727915764,
      "num_tokens": 271144615.0,
      "step": 8495
    },
    {
      "entropy": 0.7741543902084231,
      "epoch": 0.7816861317646246,
      "grad_norm": 0.15749868750572205,
      "learning_rate": 7.394731192688687e-05,
      "loss": 0.7429,
      "mean_token_accuracy": 0.7856639660894871,
      "num_tokens": 271176698.0,
      "step": 8496
    },
    {
      "entropy": 0.6690422166138887,
      "epoch": 0.7817781381360659,
      "grad_norm": 0.1473979949951172,
      "learning_rate": 7.394424510074525e-05,
      "loss": 0.6338,
      "mean_token_accuracy": 0.8119029626250267,
      "num_tokens": 271208119.0,
      "step": 8497
    },
    {
      "entropy": 0.7703954912722111,
      "epoch": 0.7818701445075071,
      "grad_norm": 0.16149114072322845,
      "learning_rate": 7.394117827460362e-05,
      "loss": 0.7588,
      "mean_token_accuracy": 0.7768103219568729,
      "num_tokens": 271239981.0,
      "step": 8498
    },
    {
      "entropy": 0.8638346008956432,
      "epoch": 0.7819621508789484,
      "grad_norm": 0.1668161004781723,
      "learning_rate": 7.393811144846199e-05,
      "loss": 0.8496,
      "mean_token_accuracy": 0.7564315907657146,
      "num_tokens": 271272640.0,
      "step": 8499
    },
    {
      "entropy": 0.7306911759078503,
      "epoch": 0.7820541572503896,
      "grad_norm": 0.15903018414974213,
      "learning_rate": 7.393504462232035e-05,
      "loss": 0.708,
      "mean_token_accuracy": 0.7972352243959904,
      "num_tokens": 271305336.0,
      "step": 8500
    },
    {
      "entropy": 0.8422243520617485,
      "epoch": 0.7821461636218308,
      "grad_norm": 0.15904955565929413,
      "learning_rate": 7.393197779617874e-05,
      "loss": 0.8345,
      "mean_token_accuracy": 0.7586082704365253,
      "num_tokens": 271337500.0,
      "step": 8501
    },
    {
      "entropy": 0.7553521385416389,
      "epoch": 0.782238169993272,
      "grad_norm": 0.1546938121318817,
      "learning_rate": 7.392891097003712e-05,
      "loss": 0.7327,
      "mean_token_accuracy": 0.7838312797248363,
      "num_tokens": 271369567.0,
      "step": 8502
    },
    {
      "entropy": 0.7145666358992457,
      "epoch": 0.7823301763647132,
      "grad_norm": 0.157041534781456,
      "learning_rate": 7.392584414389549e-05,
      "loss": 0.6836,
      "mean_token_accuracy": 0.7986959107220173,
      "num_tokens": 271400604.0,
      "step": 8503
    },
    {
      "entropy": 0.6821659300476313,
      "epoch": 0.7824221827361545,
      "grad_norm": 0.1578352302312851,
      "learning_rate": 7.392277731775385e-05,
      "loss": 0.6741,
      "mean_token_accuracy": 0.8030373752117157,
      "num_tokens": 271433103.0,
      "step": 8504
    },
    {
      "entropy": 0.6402300521731377,
      "epoch": 0.7825141891075957,
      "grad_norm": 0.14524619281291962,
      "learning_rate": 7.391971049161223e-05,
      "loss": 0.6303,
      "mean_token_accuracy": 0.8119591623544693,
      "num_tokens": 271465334.0,
      "step": 8505
    },
    {
      "entropy": 0.7315555009990931,
      "epoch": 0.7826061954790369,
      "grad_norm": 0.15939107537269592,
      "learning_rate": 7.39166436654706e-05,
      "loss": 0.7317,
      "mean_token_accuracy": 0.7829719893634319,
      "num_tokens": 271497943.0,
      "step": 8506
    },
    {
      "entropy": 0.6612913329154253,
      "epoch": 0.7826982018504781,
      "grad_norm": 0.15721625089645386,
      "learning_rate": 7.391357683932898e-05,
      "loss": 0.6618,
      "mean_token_accuracy": 0.8063562214374542,
      "num_tokens": 271530275.0,
      "step": 8507
    },
    {
      "entropy": 0.7663223035633564,
      "epoch": 0.7827902082219194,
      "grad_norm": 0.1534901261329651,
      "learning_rate": 7.391051001318735e-05,
      "loss": 0.7584,
      "mean_token_accuracy": 0.7761588357388973,
      "num_tokens": 271562666.0,
      "step": 8508
    },
    {
      "entropy": 0.6189523302018642,
      "epoch": 0.7828822145933606,
      "grad_norm": 0.15249097347259521,
      "learning_rate": 7.390744318704573e-05,
      "loss": 0.6063,
      "mean_token_accuracy": 0.8206370808184147,
      "num_tokens": 271594924.0,
      "step": 8509
    },
    {
      "entropy": 0.7302311901003122,
      "epoch": 0.7829742209648018,
      "grad_norm": 0.1509684920310974,
      "learning_rate": 7.39043763609041e-05,
      "loss": 0.7252,
      "mean_token_accuracy": 0.7856609635055065,
      "num_tokens": 271627524.0,
      "step": 8510
    },
    {
      "entropy": 0.781963337212801,
      "epoch": 0.783066227336243,
      "grad_norm": 0.1696271151304245,
      "learning_rate": 7.390130953476247e-05,
      "loss": 0.7941,
      "mean_token_accuracy": 0.7706146836280823,
      "num_tokens": 271659662.0,
      "step": 8511
    },
    {
      "entropy": 0.7020242903381586,
      "epoch": 0.7831582337076842,
      "grad_norm": 0.15776827931404114,
      "learning_rate": 7.389824270862086e-05,
      "loss": 0.7094,
      "mean_token_accuracy": 0.7899495922029018,
      "num_tokens": 271692129.0,
      "step": 8512
    },
    {
      "entropy": 0.8258201871067286,
      "epoch": 0.7832502400791255,
      "grad_norm": 0.16377785801887512,
      "learning_rate": 7.389517588247923e-05,
      "loss": 0.8263,
      "mean_token_accuracy": 0.7578641101717949,
      "num_tokens": 271724164.0,
      "step": 8513
    },
    {
      "entropy": 0.6663247980177402,
      "epoch": 0.7833422464505667,
      "grad_norm": 0.15748898684978485,
      "learning_rate": 7.38921090563376e-05,
      "loss": 0.6493,
      "mean_token_accuracy": 0.8084777519106865,
      "num_tokens": 271756043.0,
      "step": 8514
    },
    {
      "entropy": 0.9589138105511665,
      "epoch": 0.783434252822008,
      "grad_norm": 0.16445249319076538,
      "learning_rate": 7.388904223019597e-05,
      "loss": 0.946,
      "mean_token_accuracy": 0.7243949808180332,
      "num_tokens": 271787703.0,
      "step": 8515
    },
    {
      "entropy": 0.7770838253200054,
      "epoch": 0.7835262591934491,
      "grad_norm": 0.15561869740486145,
      "learning_rate": 7.388597540405435e-05,
      "loss": 0.7506,
      "mean_token_accuracy": 0.777635969221592,
      "num_tokens": 271819349.0,
      "step": 8516
    },
    {
      "entropy": 0.8095159409567714,
      "epoch": 0.7836182655648903,
      "grad_norm": 0.15029647946357727,
      "learning_rate": 7.388290857791273e-05,
      "loss": 0.7919,
      "mean_token_accuracy": 0.7685501016676426,
      "num_tokens": 271851611.0,
      "step": 8517
    },
    {
      "entropy": 0.6154816746711731,
      "epoch": 0.7837102719363316,
      "grad_norm": 0.14434169232845306,
      "learning_rate": 7.38798417517711e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.820955153554678,
      "num_tokens": 271883948.0,
      "step": 8518
    },
    {
      "entropy": 0.7195031903684139,
      "epoch": 0.7838022783077728,
      "grad_norm": 0.15756195783615112,
      "learning_rate": 7.387677492562947e-05,
      "loss": 0.7001,
      "mean_token_accuracy": 0.7950352430343628,
      "num_tokens": 271916120.0,
      "step": 8519
    },
    {
      "entropy": 0.6525544952601194,
      "epoch": 0.7838942846792141,
      "grad_norm": 0.1632777452468872,
      "learning_rate": 7.387370809948785e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.8131441846489906,
      "num_tokens": 271948199.0,
      "step": 8520
    },
    {
      "entropy": 0.6348206531256437,
      "epoch": 0.7839862910506552,
      "grad_norm": 0.14695827662944794,
      "learning_rate": 7.387064127334622e-05,
      "loss": 0.6236,
      "mean_token_accuracy": 0.8122722953557968,
      "num_tokens": 271980028.0,
      "step": 8521
    },
    {
      "entropy": 0.7266394607722759,
      "epoch": 0.7840782974220964,
      "grad_norm": 0.15805917978286743,
      "learning_rate": 7.38675744472046e-05,
      "loss": 0.7227,
      "mean_token_accuracy": 0.7861533723771572,
      "num_tokens": 272011999.0,
      "step": 8522
    },
    {
      "entropy": 0.6927889306098223,
      "epoch": 0.7841703037935377,
      "grad_norm": 0.14924904704093933,
      "learning_rate": 7.386450762106296e-05,
      "loss": 0.672,
      "mean_token_accuracy": 0.7981198951601982,
      "num_tokens": 272044096.0,
      "step": 8523
    },
    {
      "entropy": 0.7051739394664764,
      "epoch": 0.7842623101649789,
      "grad_norm": 0.15370120108127594,
      "learning_rate": 7.386144079492135e-05,
      "loss": 0.7012,
      "mean_token_accuracy": 0.7923919223248959,
      "num_tokens": 272076351.0,
      "step": 8524
    },
    {
      "entropy": 0.7693591872230172,
      "epoch": 0.7843543165364202,
      "grad_norm": 0.16013094782829285,
      "learning_rate": 7.385837396877971e-05,
      "loss": 0.7628,
      "mean_token_accuracy": 0.7763490565121174,
      "num_tokens": 272108842.0,
      "step": 8525
    },
    {
      "entropy": 0.764161292463541,
      "epoch": 0.7844463229078613,
      "grad_norm": 0.16010309755802155,
      "learning_rate": 7.385530714263808e-05,
      "loss": 0.7566,
      "mean_token_accuracy": 0.7780318781733513,
      "num_tokens": 272140067.0,
      "step": 8526
    },
    {
      "entropy": 0.7340683452785015,
      "epoch": 0.7845383292793026,
      "grad_norm": 0.15414585173130035,
      "learning_rate": 7.385224031649646e-05,
      "loss": 0.7323,
      "mean_token_accuracy": 0.7888293638825417,
      "num_tokens": 272171754.0,
      "step": 8527
    },
    {
      "entropy": 0.6302158441394567,
      "epoch": 0.7846303356507438,
      "grad_norm": 0.16700467467308044,
      "learning_rate": 7.384917349035484e-05,
      "loss": 0.6152,
      "mean_token_accuracy": 0.8170327842235565,
      "num_tokens": 272203933.0,
      "step": 8528
    },
    {
      "entropy": 0.6855208408087492,
      "epoch": 0.784722342022185,
      "grad_norm": 0.1547228991985321,
      "learning_rate": 7.384610666421321e-05,
      "loss": 0.6852,
      "mean_token_accuracy": 0.8005902580916882,
      "num_tokens": 272235288.0,
      "step": 8529
    },
    {
      "entropy": 0.762965808622539,
      "epoch": 0.7848143483936263,
      "grad_norm": 0.15182481706142426,
      "learning_rate": 7.384303983807158e-05,
      "loss": 0.7735,
      "mean_token_accuracy": 0.7717209383845329,
      "num_tokens": 272267118.0,
      "step": 8530
    },
    {
      "entropy": 0.6949124690145254,
      "epoch": 0.7849063547650674,
      "grad_norm": 0.15113437175750732,
      "learning_rate": 7.383997301192995e-05,
      "loss": 0.6886,
      "mean_token_accuracy": 0.7972819842398167,
      "num_tokens": 272298744.0,
      "step": 8531
    },
    {
      "entropy": 0.7210070807486773,
      "epoch": 0.7849983611365087,
      "grad_norm": 0.1545727699995041,
      "learning_rate": 7.383690618578833e-05,
      "loss": 0.7095,
      "mean_token_accuracy": 0.7871234193444252,
      "num_tokens": 272330898.0,
      "step": 8532
    },
    {
      "entropy": 0.6758428867906332,
      "epoch": 0.7850903675079499,
      "grad_norm": 0.15054763853549957,
      "learning_rate": 7.383383935964671e-05,
      "loss": 0.6722,
      "mean_token_accuracy": 0.8015005029737949,
      "num_tokens": 272362978.0,
      "step": 8533
    },
    {
      "entropy": 0.7831039894372225,
      "epoch": 0.7851823738793912,
      "grad_norm": 0.17746292054653168,
      "learning_rate": 7.383077253350508e-05,
      "loss": 0.8009,
      "mean_token_accuracy": 0.7686121910810471,
      "num_tokens": 272394546.0,
      "step": 8534
    },
    {
      "entropy": 0.7062401492148638,
      "epoch": 0.7852743802508324,
      "grad_norm": 0.15497301518917084,
      "learning_rate": 7.382770570736345e-05,
      "loss": 0.6925,
      "mean_token_accuracy": 0.7953660041093826,
      "num_tokens": 272427008.0,
      "step": 8535
    },
    {
      "entropy": 0.7116552088409662,
      "epoch": 0.7853663866222735,
      "grad_norm": 0.1621994823217392,
      "learning_rate": 7.382463888122183e-05,
      "loss": 0.7049,
      "mean_token_accuracy": 0.7932934835553169,
      "num_tokens": 272459330.0,
      "step": 8536
    },
    {
      "entropy": 0.6832623807713389,
      "epoch": 0.7854583929937148,
      "grad_norm": 0.15525010228157043,
      "learning_rate": 7.38215720550802e-05,
      "loss": 0.6573,
      "mean_token_accuracy": 0.8068063668906689,
      "num_tokens": 272491436.0,
      "step": 8537
    },
    {
      "entropy": 0.8715250566601753,
      "epoch": 0.785550399365156,
      "grad_norm": 0.16261185705661774,
      "learning_rate": 7.381850522893858e-05,
      "loss": 0.8601,
      "mean_token_accuracy": 0.7456566356122494,
      "num_tokens": 272523811.0,
      "step": 8538
    },
    {
      "entropy": 0.7553435936570168,
      "epoch": 0.7856424057365973,
      "grad_norm": 0.1507043093442917,
      "learning_rate": 7.381543840279695e-05,
      "loss": 0.7321,
      "mean_token_accuracy": 0.7832330614328384,
      "num_tokens": 272555698.0,
      "step": 8539
    },
    {
      "entropy": 0.8116856049746275,
      "epoch": 0.7857344121080385,
      "grad_norm": 0.1720178872346878,
      "learning_rate": 7.381237157665533e-05,
      "loss": 0.7936,
      "mean_token_accuracy": 0.76827522367239,
      "num_tokens": 272587957.0,
      "step": 8540
    },
    {
      "entropy": 0.710743760690093,
      "epoch": 0.7858264184794796,
      "grad_norm": 0.1596609205007553,
      "learning_rate": 7.38093047505137e-05,
      "loss": 0.6992,
      "mean_token_accuracy": 0.7910830117762089,
      "num_tokens": 272619823.0,
      "step": 8541
    },
    {
      "entropy": 0.8098485693335533,
      "epoch": 0.7859184248509209,
      "grad_norm": 0.15769857168197632,
      "learning_rate": 7.380623792437206e-05,
      "loss": 0.8016,
      "mean_token_accuracy": 0.7665514089167118,
      "num_tokens": 272651833.0,
      "step": 8542
    },
    {
      "entropy": 0.6909339167177677,
      "epoch": 0.7860104312223621,
      "grad_norm": 0.14901579916477203,
      "learning_rate": 7.380317109823044e-05,
      "loss": 0.6736,
      "mean_token_accuracy": 0.799611072987318,
      "num_tokens": 272683808.0,
      "step": 8543
    },
    {
      "entropy": 0.6987296976149082,
      "epoch": 0.7861024375938034,
      "grad_norm": 0.16125808656215668,
      "learning_rate": 7.380010427208883e-05,
      "loss": 0.6998,
      "mean_token_accuracy": 0.795626800507307,
      "num_tokens": 272714937.0,
      "step": 8544
    },
    {
      "entropy": 0.7419259492307901,
      "epoch": 0.7861944439652446,
      "grad_norm": 0.15823298692703247,
      "learning_rate": 7.379703744594719e-05,
      "loss": 0.7386,
      "mean_token_accuracy": 0.7841432765126228,
      "num_tokens": 272746846.0,
      "step": 8545
    },
    {
      "entropy": 0.7214630153030157,
      "epoch": 0.7862864503366858,
      "grad_norm": 0.16559213399887085,
      "learning_rate": 7.379397061980556e-05,
      "loss": 0.7164,
      "mean_token_accuracy": 0.7933027371764183,
      "num_tokens": 272777860.0,
      "step": 8546
    },
    {
      "entropy": 0.7438743440434337,
      "epoch": 0.786378456708127,
      "grad_norm": 0.15091730654239655,
      "learning_rate": 7.379090379366394e-05,
      "loss": 0.7359,
      "mean_token_accuracy": 0.7864509858191013,
      "num_tokens": 272809775.0,
      "step": 8547
    },
    {
      "entropy": 0.6060006059706211,
      "epoch": 0.7864704630795682,
      "grad_norm": 0.14475904405117035,
      "learning_rate": 7.378783696752231e-05,
      "loss": 0.5996,
      "mean_token_accuracy": 0.820652425289154,
      "num_tokens": 272841791.0,
      "step": 8548
    },
    {
      "entropy": 0.6943454025313258,
      "epoch": 0.7865624694510095,
      "grad_norm": 0.16151531040668488,
      "learning_rate": 7.378477014138069e-05,
      "loss": 0.7013,
      "mean_token_accuracy": 0.7942841239273548,
      "num_tokens": 272874053.0,
      "step": 8549
    },
    {
      "entropy": 0.8227536808699369,
      "epoch": 0.7866544758224507,
      "grad_norm": 0.15640684962272644,
      "learning_rate": 7.378170331523906e-05,
      "loss": 0.8054,
      "mean_token_accuracy": 0.7611001990735531,
      "num_tokens": 272906116.0,
      "step": 8550
    },
    {
      "entropy": 0.7192155793309212,
      "epoch": 0.7867464821938919,
      "grad_norm": 0.15136462450027466,
      "learning_rate": 7.377863648909744e-05,
      "loss": 0.6903,
      "mean_token_accuracy": 0.7924254685640335,
      "num_tokens": 272937777.0,
      "step": 8551
    },
    {
      "entropy": 0.6969116292893887,
      "epoch": 0.7868384885653331,
      "grad_norm": 0.15180860459804535,
      "learning_rate": 7.377556966295581e-05,
      "loss": 0.6899,
      "mean_token_accuracy": 0.7938534393906593,
      "num_tokens": 272970074.0,
      "step": 8552
    },
    {
      "entropy": 0.6636876277625561,
      "epoch": 0.7869304949367744,
      "grad_norm": 0.15112192928791046,
      "learning_rate": 7.377250283681418e-05,
      "loss": 0.6419,
      "mean_token_accuracy": 0.8111265935003757,
      "num_tokens": 273001693.0,
      "step": 8553
    },
    {
      "entropy": 0.7692442703992128,
      "epoch": 0.7870225013082156,
      "grad_norm": 0.15780681371688843,
      "learning_rate": 7.376943601067256e-05,
      "loss": 0.7451,
      "mean_token_accuracy": 0.7805689461529255,
      "num_tokens": 273033984.0,
      "step": 8554
    },
    {
      "entropy": 0.7713220082223415,
      "epoch": 0.7871145076796568,
      "grad_norm": 0.1523052155971527,
      "learning_rate": 7.376636918453094e-05,
      "loss": 0.7664,
      "mean_token_accuracy": 0.7744520530104637,
      "num_tokens": 273065581.0,
      "step": 8555
    },
    {
      "entropy": 0.7557821571826935,
      "epoch": 0.787206514051098,
      "grad_norm": 0.17176759243011475,
      "learning_rate": 7.376330235838931e-05,
      "loss": 0.7349,
      "mean_token_accuracy": 0.7784265726804733,
      "num_tokens": 273096214.0,
      "step": 8556
    },
    {
      "entropy": 0.5424125287681818,
      "epoch": 0.7872985204225392,
      "grad_norm": 0.1521076261997223,
      "learning_rate": 7.376023553224768e-05,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8428789675235748,
      "num_tokens": 273128360.0,
      "step": 8557
    },
    {
      "entropy": 0.7493317387998104,
      "epoch": 0.7873905267939805,
      "grad_norm": 0.16830655932426453,
      "learning_rate": 7.375716870610606e-05,
      "loss": 0.7362,
      "mean_token_accuracy": 0.7796186208724976,
      "num_tokens": 273158895.0,
      "step": 8558
    },
    {
      "entropy": 0.9297135733067989,
      "epoch": 0.7874825331654217,
      "grad_norm": 0.1698494404554367,
      "learning_rate": 7.375410187996444e-05,
      "loss": 0.9378,
      "mean_token_accuracy": 0.7319910824298859,
      "num_tokens": 273190871.0,
      "step": 8559
    },
    {
      "entropy": 0.7792004533112049,
      "epoch": 0.787574539536863,
      "grad_norm": 0.1557946652173996,
      "learning_rate": 7.37510350538228e-05,
      "loss": 0.7726,
      "mean_token_accuracy": 0.771990992128849,
      "num_tokens": 273222967.0,
      "step": 8560
    },
    {
      "entropy": 0.6873204540461302,
      "epoch": 0.7876665459083042,
      "grad_norm": 0.1579616218805313,
      "learning_rate": 7.374796822768117e-05,
      "loss": 0.6731,
      "mean_token_accuracy": 0.800992626696825,
      "num_tokens": 273255177.0,
      "step": 8561
    },
    {
      "entropy": 0.6091364892199636,
      "epoch": 0.7877585522797453,
      "grad_norm": 0.14666375517845154,
      "learning_rate": 7.374490140153954e-05,
      "loss": 0.6074,
      "mean_token_accuracy": 0.8194530941545963,
      "num_tokens": 273287452.0,
      "step": 8562
    },
    {
      "entropy": 0.761012252420187,
      "epoch": 0.7878505586511866,
      "grad_norm": 0.1611592024564743,
      "learning_rate": 7.374183457539792e-05,
      "loss": 0.752,
      "mean_token_accuracy": 0.7788048163056374,
      "num_tokens": 273319558.0,
      "step": 8563
    },
    {
      "entropy": 0.7336780186742544,
      "epoch": 0.7879425650226278,
      "grad_norm": 0.15773077309131622,
      "learning_rate": 7.37387677492563e-05,
      "loss": 0.7008,
      "mean_token_accuracy": 0.7894185967743397,
      "num_tokens": 273351841.0,
      "step": 8564
    },
    {
      "entropy": 0.6742825247347355,
      "epoch": 0.7880345713940691,
      "grad_norm": 0.15692102909088135,
      "learning_rate": 7.373570092311467e-05,
      "loss": 0.653,
      "mean_token_accuracy": 0.8098753802478313,
      "num_tokens": 273384445.0,
      "step": 8565
    },
    {
      "entropy": 0.7086341287940741,
      "epoch": 0.7881265777655103,
      "grad_norm": 0.168618306517601,
      "learning_rate": 7.373263409697304e-05,
      "loss": 0.6986,
      "mean_token_accuracy": 0.7984604984521866,
      "num_tokens": 273415622.0,
      "step": 8566
    },
    {
      "entropy": 0.6993858758360147,
      "epoch": 0.7882185841369515,
      "grad_norm": 0.1541106402873993,
      "learning_rate": 7.372956727083142e-05,
      "loss": 0.6788,
      "mean_token_accuracy": 0.800257284194231,
      "num_tokens": 273447966.0,
      "step": 8567
    },
    {
      "entropy": 0.6398835983127356,
      "epoch": 0.7883105905083927,
      "grad_norm": 0.15397000312805176,
      "learning_rate": 7.372650044468979e-05,
      "loss": 0.6321,
      "mean_token_accuracy": 0.8098541162908077,
      "num_tokens": 273480400.0,
      "step": 8568
    },
    {
      "entropy": 0.7735233772546053,
      "epoch": 0.7884025968798339,
      "grad_norm": 0.1581113636493683,
      "learning_rate": 7.372343361854817e-05,
      "loss": 0.7603,
      "mean_token_accuracy": 0.7799490764737129,
      "num_tokens": 273512572.0,
      "step": 8569
    },
    {
      "entropy": 0.6825246205553412,
      "epoch": 0.7884946032512752,
      "grad_norm": 0.15679603815078735,
      "learning_rate": 7.372036679240654e-05,
      "loss": 0.6687,
      "mean_token_accuracy": 0.8012077212333679,
      "num_tokens": 273544472.0,
      "step": 8570
    },
    {
      "entropy": 0.7472797520458698,
      "epoch": 0.7885866096227164,
      "grad_norm": 0.16353359818458557,
      "learning_rate": 7.371729996626492e-05,
      "loss": 0.7328,
      "mean_token_accuracy": 0.7833852767944336,
      "num_tokens": 273577015.0,
      "step": 8571
    },
    {
      "entropy": 0.7118648011237383,
      "epoch": 0.7886786159941576,
      "grad_norm": 0.15723274648189545,
      "learning_rate": 7.371423314012329e-05,
      "loss": 0.7108,
      "mean_token_accuracy": 0.7905420139431953,
      "num_tokens": 273608493.0,
      "step": 8572
    },
    {
      "entropy": 0.6758079566061497,
      "epoch": 0.7887706223655988,
      "grad_norm": 0.15429018437862396,
      "learning_rate": 7.371116631398166e-05,
      "loss": 0.6551,
      "mean_token_accuracy": 0.802077766507864,
      "num_tokens": 273639914.0,
      "step": 8573
    },
    {
      "entropy": 0.8466695938259363,
      "epoch": 0.78886262873704,
      "grad_norm": 0.1673181802034378,
      "learning_rate": 7.370809948784004e-05,
      "loss": 0.8454,
      "mean_token_accuracy": 0.7590932324528694,
      "num_tokens": 273671420.0,
      "step": 8574
    },
    {
      "entropy": 0.6785710789263248,
      "epoch": 0.7889546351084813,
      "grad_norm": 0.14531457424163818,
      "learning_rate": 7.370503266169842e-05,
      "loss": 0.6663,
      "mean_token_accuracy": 0.7995629571378231,
      "num_tokens": 273704137.0,
      "step": 8575
    },
    {
      "entropy": 0.7604596018791199,
      "epoch": 0.7890466414799225,
      "grad_norm": 0.15616276860237122,
      "learning_rate": 7.370196583555679e-05,
      "loss": 0.7574,
      "mean_token_accuracy": 0.775534201413393,
      "num_tokens": 273735916.0,
      "step": 8576
    },
    {
      "entropy": 0.7103800345212221,
      "epoch": 0.7891386478513637,
      "grad_norm": 0.1590428650379181,
      "learning_rate": 7.369889900941515e-05,
      "loss": 0.6938,
      "mean_token_accuracy": 0.7975443340837955,
      "num_tokens": 273768428.0,
      "step": 8577
    },
    {
      "entropy": 0.7106568459421396,
      "epoch": 0.7892306542228049,
      "grad_norm": 0.16185952723026276,
      "learning_rate": 7.369583218327354e-05,
      "loss": 0.6966,
      "mean_token_accuracy": 0.791290320456028,
      "num_tokens": 273799578.0,
      "step": 8578
    },
    {
      "entropy": 0.7845104336738586,
      "epoch": 0.7893226605942462,
      "grad_norm": 0.15051478147506714,
      "learning_rate": 7.36927653571319e-05,
      "loss": 0.7757,
      "mean_token_accuracy": 0.7671909630298615,
      "num_tokens": 273831159.0,
      "step": 8579
    },
    {
      "entropy": 0.7506093196570873,
      "epoch": 0.7894146669656874,
      "grad_norm": 0.14986221492290497,
      "learning_rate": 7.368969853099029e-05,
      "loss": 0.7372,
      "mean_token_accuracy": 0.7796771936118603,
      "num_tokens": 273863224.0,
      "step": 8580
    },
    {
      "entropy": 0.6893134415149689,
      "epoch": 0.7895066733371287,
      "grad_norm": 0.16436223685741425,
      "learning_rate": 7.368663170484865e-05,
      "loss": 0.6623,
      "mean_token_accuracy": 0.8005218096077442,
      "num_tokens": 273894310.0,
      "step": 8581
    },
    {
      "entropy": 0.7511411318555474,
      "epoch": 0.7895986797085698,
      "grad_norm": 0.15950360894203186,
      "learning_rate": 7.368356487870703e-05,
      "loss": 0.7631,
      "mean_token_accuracy": 0.7788693904876709,
      "num_tokens": 273925707.0,
      "step": 8582
    },
    {
      "entropy": 0.7684105727821589,
      "epoch": 0.789690686080011,
      "grad_norm": 0.15488223731517792,
      "learning_rate": 7.36804980525654e-05,
      "loss": 0.7449,
      "mean_token_accuracy": 0.7757253386080265,
      "num_tokens": 273957201.0,
      "step": 8583
    },
    {
      "entropy": 0.7630168162286282,
      "epoch": 0.7897826924514523,
      "grad_norm": 0.16555221378803253,
      "learning_rate": 7.367743122642377e-05,
      "loss": 0.7574,
      "mean_token_accuracy": 0.7739398628473282,
      "num_tokens": 273988961.0,
      "step": 8584
    },
    {
      "entropy": 0.6836388884112239,
      "epoch": 0.7898746988228935,
      "grad_norm": 0.15380649268627167,
      "learning_rate": 7.367436440028215e-05,
      "loss": 0.668,
      "mean_token_accuracy": 0.8092549256980419,
      "num_tokens": 274020423.0,
      "step": 8585
    },
    {
      "entropy": 0.7630134802311659,
      "epoch": 0.7899667051943348,
      "grad_norm": 0.15431813895702362,
      "learning_rate": 7.367129757414053e-05,
      "loss": 0.7358,
      "mean_token_accuracy": 0.7857904434204102,
      "num_tokens": 274052784.0,
      "step": 8586
    },
    {
      "entropy": 0.7531762197613716,
      "epoch": 0.7900587115657759,
      "grad_norm": 0.15231549739837646,
      "learning_rate": 7.36682307479989e-05,
      "loss": 0.7497,
      "mean_token_accuracy": 0.7743534557521343,
      "num_tokens": 274084779.0,
      "step": 8587
    },
    {
      "entropy": 0.8413972146809101,
      "epoch": 0.7901507179372171,
      "grad_norm": 0.1625741869211197,
      "learning_rate": 7.366516392185727e-05,
      "loss": 0.8383,
      "mean_token_accuracy": 0.7559649273753166,
      "num_tokens": 274116736.0,
      "step": 8588
    },
    {
      "entropy": 0.7303318157792091,
      "epoch": 0.7902427243086584,
      "grad_norm": 0.15725639462471008,
      "learning_rate": 7.366209709571564e-05,
      "loss": 0.7258,
      "mean_token_accuracy": 0.786669846624136,
      "num_tokens": 274148477.0,
      "step": 8589
    },
    {
      "entropy": 0.79659328982234,
      "epoch": 0.7903347306800996,
      "grad_norm": 0.15633517503738403,
      "learning_rate": 7.365903026957403e-05,
      "loss": 0.7846,
      "mean_token_accuracy": 0.7710376977920532,
      "num_tokens": 274180887.0,
      "step": 8590
    },
    {
      "entropy": 0.8377163615077734,
      "epoch": 0.7904267370515409,
      "grad_norm": 0.15696901082992554,
      "learning_rate": 7.36559634434324e-05,
      "loss": 0.8217,
      "mean_token_accuracy": 0.759906567633152,
      "num_tokens": 274212053.0,
      "step": 8591
    },
    {
      "entropy": 0.672072040848434,
      "epoch": 0.790518743422982,
      "grad_norm": 0.1511387676000595,
      "learning_rate": 7.365289661729077e-05,
      "loss": 0.6648,
      "mean_token_accuracy": 0.8023151494562626,
      "num_tokens": 274243564.0,
      "step": 8592
    },
    {
      "entropy": 0.7287766924127936,
      "epoch": 0.7906107497944233,
      "grad_norm": 0.15541014075279236,
      "learning_rate": 7.364982979114914e-05,
      "loss": 0.7169,
      "mean_token_accuracy": 0.7881781384348869,
      "num_tokens": 274275027.0,
      "step": 8593
    },
    {
      "entropy": 0.7706855647265911,
      "epoch": 0.7907027561658645,
      "grad_norm": 0.1663215011358261,
      "learning_rate": 7.364676296500752e-05,
      "loss": 0.7681,
      "mean_token_accuracy": 0.7778508588671684,
      "num_tokens": 274307609.0,
      "step": 8594
    },
    {
      "entropy": 0.8780867643654346,
      "epoch": 0.7907947625373057,
      "grad_norm": 0.17531369626522064,
      "learning_rate": 7.36436961388659e-05,
      "loss": 0.8556,
      "mean_token_accuracy": 0.752273440361023,
      "num_tokens": 274339613.0,
      "step": 8595
    },
    {
      "entropy": 0.8039433620870113,
      "epoch": 0.790886768908747,
      "grad_norm": 0.16240191459655762,
      "learning_rate": 7.364062931272427e-05,
      "loss": 0.7905,
      "mean_token_accuracy": 0.7687235586345196,
      "num_tokens": 274370788.0,
      "step": 8596
    },
    {
      "entropy": 0.8562704306095839,
      "epoch": 0.7909787752801881,
      "grad_norm": 0.1710420697927475,
      "learning_rate": 7.363756248658263e-05,
      "loss": 0.8586,
      "mean_token_accuracy": 0.7530887424945831,
      "num_tokens": 274403177.0,
      "step": 8597
    },
    {
      "entropy": 0.6948898583650589,
      "epoch": 0.7910707816516294,
      "grad_norm": 0.14617110788822174,
      "learning_rate": 7.363449566044102e-05,
      "loss": 0.6783,
      "mean_token_accuracy": 0.8022138699889183,
      "num_tokens": 274435079.0,
      "step": 8598
    },
    {
      "entropy": 0.7582244519144297,
      "epoch": 0.7911627880230706,
      "grad_norm": 0.16229483485221863,
      "learning_rate": 7.363142883429938e-05,
      "loss": 0.754,
      "mean_token_accuracy": 0.7741826437413692,
      "num_tokens": 274466575.0,
      "step": 8599
    },
    {
      "entropy": 0.5704895593225956,
      "epoch": 0.7912547943945119,
      "grad_norm": 0.15632276237010956,
      "learning_rate": 7.362836200815776e-05,
      "loss": 0.5609,
      "mean_token_accuracy": 0.833602212369442,
      "num_tokens": 274497831.0,
      "step": 8600
    },
    {
      "entropy": 0.654584439471364,
      "epoch": 0.7913468007659531,
      "grad_norm": 0.15747042000293732,
      "learning_rate": 7.362529518201613e-05,
      "loss": 0.6282,
      "mean_token_accuracy": 0.811036005616188,
      "num_tokens": 274529557.0,
      "step": 8601
    },
    {
      "entropy": 0.7592297773808241,
      "epoch": 0.7914388071373942,
      "grad_norm": 0.15971432626247406,
      "learning_rate": 7.362222835587451e-05,
      "loss": 0.7523,
      "mean_token_accuracy": 0.7802796475589275,
      "num_tokens": 274561419.0,
      "step": 8602
    },
    {
      "entropy": 0.7047729417681694,
      "epoch": 0.7915308135088355,
      "grad_norm": 0.15369847416877747,
      "learning_rate": 7.361916152973288e-05,
      "loss": 0.7046,
      "mean_token_accuracy": 0.7960044331848621,
      "num_tokens": 274593959.0,
      "step": 8603
    },
    {
      "entropy": 0.6620243433862925,
      "epoch": 0.7916228198802767,
      "grad_norm": 0.1461433619260788,
      "learning_rate": 7.361609470359125e-05,
      "loss": 0.627,
      "mean_token_accuracy": 0.8089936338365078,
      "num_tokens": 274626045.0,
      "step": 8604
    },
    {
      "entropy": 0.7821628358215094,
      "epoch": 0.791714826251718,
      "grad_norm": 0.16493122279644012,
      "learning_rate": 7.361302787744963e-05,
      "loss": 0.7936,
      "mean_token_accuracy": 0.7713088393211365,
      "num_tokens": 274658114.0,
      "step": 8605
    },
    {
      "entropy": 0.7641425207257271,
      "epoch": 0.7918068326231592,
      "grad_norm": 0.1649802327156067,
      "learning_rate": 7.360996105130801e-05,
      "loss": 0.7715,
      "mean_token_accuracy": 0.7772484458982944,
      "num_tokens": 274689996.0,
      "step": 8606
    },
    {
      "entropy": 0.6146721132099628,
      "epoch": 0.7918988389946003,
      "grad_norm": 0.14977332949638367,
      "learning_rate": 7.360689422516638e-05,
      "loss": 0.5977,
      "mean_token_accuracy": 0.8185565508902073,
      "num_tokens": 274721803.0,
      "step": 8607
    },
    {
      "entropy": 0.6639824332669377,
      "epoch": 0.7919908453660416,
      "grad_norm": 0.15179799497127533,
      "learning_rate": 7.360382739902475e-05,
      "loss": 0.6614,
      "mean_token_accuracy": 0.8071005567908287,
      "num_tokens": 274753821.0,
      "step": 8608
    },
    {
      "entropy": 0.6675180271267891,
      "epoch": 0.7920828517374828,
      "grad_norm": 0.1461772620677948,
      "learning_rate": 7.360076057288313e-05,
      "loss": 0.6551,
      "mean_token_accuracy": 0.8007979057729244,
      "num_tokens": 274785928.0,
      "step": 8609
    },
    {
      "entropy": 0.6975451232865453,
      "epoch": 0.7921748581089241,
      "grad_norm": 0.15732145309448242,
      "learning_rate": 7.35976937467415e-05,
      "loss": 0.6887,
      "mean_token_accuracy": 0.7989288605749607,
      "num_tokens": 274817900.0,
      "step": 8610
    },
    {
      "entropy": 0.6818988434970379,
      "epoch": 0.7922668644803653,
      "grad_norm": 0.15937431156635284,
      "learning_rate": 7.359462692059988e-05,
      "loss": 0.647,
      "mean_token_accuracy": 0.8076740428805351,
      "num_tokens": 274850071.0,
      "step": 8611
    },
    {
      "entropy": 0.6721364799886942,
      "epoch": 0.7923588708518065,
      "grad_norm": 0.15472657978534698,
      "learning_rate": 7.359156009445825e-05,
      "loss": 0.6612,
      "mean_token_accuracy": 0.8026912622153759,
      "num_tokens": 274880973.0,
      "step": 8612
    },
    {
      "entropy": 0.7747740875929594,
      "epoch": 0.7924508772232477,
      "grad_norm": 0.1608923077583313,
      "learning_rate": 7.358849326831663e-05,
      "loss": 0.7806,
      "mean_token_accuracy": 0.775224182754755,
      "num_tokens": 274913066.0,
      "step": 8613
    },
    {
      "entropy": 0.6955502862110734,
      "epoch": 0.7925428835946889,
      "grad_norm": 0.14988668262958527,
      "learning_rate": 7.3585426442175e-05,
      "loss": 0.671,
      "mean_token_accuracy": 0.8050806038081646,
      "num_tokens": 274945834.0,
      "step": 8614
    },
    {
      "entropy": 0.7889116518199444,
      "epoch": 0.7926348899661302,
      "grad_norm": 0.15525148808956146,
      "learning_rate": 7.358235961603336e-05,
      "loss": 0.7707,
      "mean_token_accuracy": 0.770535834133625,
      "num_tokens": 274978281.0,
      "step": 8615
    },
    {
      "entropy": 0.6177563574165106,
      "epoch": 0.7927268963375714,
      "grad_norm": 0.16196255385875702,
      "learning_rate": 7.357929278989175e-05,
      "loss": 0.5885,
      "mean_token_accuracy": 0.8139913156628609,
      "num_tokens": 275009828.0,
      "step": 8616
    },
    {
      "entropy": 0.7713784947991371,
      "epoch": 0.7928189027090126,
      "grad_norm": 0.16558709740638733,
      "learning_rate": 7.357622596375013e-05,
      "loss": 0.7718,
      "mean_token_accuracy": 0.7705737873911858,
      "num_tokens": 275041522.0,
      "step": 8617
    },
    {
      "entropy": 0.8728860784322023,
      "epoch": 0.7929109090804538,
      "grad_norm": 0.16619348526000977,
      "learning_rate": 7.35731591376085e-05,
      "loss": 0.8853,
      "mean_token_accuracy": 0.7450478672981262,
      "num_tokens": 275073346.0,
      "step": 8618
    },
    {
      "entropy": 0.7422376433387399,
      "epoch": 0.793002915451895,
      "grad_norm": 0.1575368046760559,
      "learning_rate": 7.357009231146686e-05,
      "loss": 0.7432,
      "mean_token_accuracy": 0.7843665890395641,
      "num_tokens": 275105301.0,
      "step": 8619
    },
    {
      "entropy": 0.7688485234975815,
      "epoch": 0.7930949218233363,
      "grad_norm": 0.1515977531671524,
      "learning_rate": 7.356702548532523e-05,
      "loss": 0.7555,
      "mean_token_accuracy": 0.7753106392920017,
      "num_tokens": 275137393.0,
      "step": 8620
    },
    {
      "entropy": 0.8319140635430813,
      "epoch": 0.7931869281947775,
      "grad_norm": 0.1590123325586319,
      "learning_rate": 7.356395865918361e-05,
      "loss": 0.8211,
      "mean_token_accuracy": 0.7576138712465763,
      "num_tokens": 275169175.0,
      "step": 8621
    },
    {
      "entropy": 0.7218499965965748,
      "epoch": 0.7932789345662187,
      "grad_norm": 0.16193337738513947,
      "learning_rate": 7.3560891833042e-05,
      "loss": 0.7092,
      "mean_token_accuracy": 0.7880433686077595,
      "num_tokens": 275201523.0,
      "step": 8622
    },
    {
      "entropy": 0.8927723821252584,
      "epoch": 0.7933709409376599,
      "grad_norm": 0.16460934281349182,
      "learning_rate": 7.355782500690036e-05,
      "loss": 0.8835,
      "mean_token_accuracy": 0.7409804463386536,
      "num_tokens": 275232969.0,
      "step": 8623
    },
    {
      "entropy": 0.733483949676156,
      "epoch": 0.7934629473091012,
      "grad_norm": 0.16281366348266602,
      "learning_rate": 7.355475818075873e-05,
      "loss": 0.7135,
      "mean_token_accuracy": 0.7930534966289997,
      "num_tokens": 275264713.0,
      "step": 8624
    },
    {
      "entropy": 0.5529687674716115,
      "epoch": 0.7935549536805424,
      "grad_norm": 0.15843969583511353,
      "learning_rate": 7.355169135461711e-05,
      "loss": 0.532,
      "mean_token_accuracy": 0.8385082222521305,
      "num_tokens": 275296850.0,
      "step": 8625
    },
    {
      "entropy": 0.7217599246650934,
      "epoch": 0.7936469600519837,
      "grad_norm": 0.14874383807182312,
      "learning_rate": 7.354862452847548e-05,
      "loss": 0.6937,
      "mean_token_accuracy": 0.7927995882928371,
      "num_tokens": 275327921.0,
      "step": 8626
    },
    {
      "entropy": 0.6552297063171864,
      "epoch": 0.7937389664234248,
      "grad_norm": 0.15190809965133667,
      "learning_rate": 7.354555770233386e-05,
      "loss": 0.6329,
      "mean_token_accuracy": 0.8115664087235928,
      "num_tokens": 275359583.0,
      "step": 8627
    },
    {
      "entropy": 0.7160813882946968,
      "epoch": 0.793830972794866,
      "grad_norm": 0.15713636577129364,
      "learning_rate": 7.354249087619223e-05,
      "loss": 0.7008,
      "mean_token_accuracy": 0.7943427078425884,
      "num_tokens": 275391372.0,
      "step": 8628
    },
    {
      "entropy": 0.7218711394816637,
      "epoch": 0.7939229791663073,
      "grad_norm": 0.16313666105270386,
      "learning_rate": 7.353942405005061e-05,
      "loss": 0.7227,
      "mean_token_accuracy": 0.7872022613883018,
      "num_tokens": 275423353.0,
      "step": 8629
    },
    {
      "entropy": 0.8043688908219337,
      "epoch": 0.7940149855377485,
      "grad_norm": 0.15272633731365204,
      "learning_rate": 7.353635722390898e-05,
      "loss": 0.7739,
      "mean_token_accuracy": 0.7699203416705132,
      "num_tokens": 275455592.0,
      "step": 8630
    },
    {
      "entropy": 0.7237706948071718,
      "epoch": 0.7941069919091898,
      "grad_norm": 0.15718333423137665,
      "learning_rate": 7.353329039776735e-05,
      "loss": 0.6937,
      "mean_token_accuracy": 0.7940209098160267,
      "num_tokens": 275488058.0,
      "step": 8631
    },
    {
      "entropy": 0.6940217334777117,
      "epoch": 0.7941989982806309,
      "grad_norm": 0.146015003323555,
      "learning_rate": 7.353022357162573e-05,
      "loss": 0.6791,
      "mean_token_accuracy": 0.7960436791181564,
      "num_tokens": 275520431.0,
      "step": 8632
    },
    {
      "entropy": 0.7181407762691379,
      "epoch": 0.7942910046520721,
      "grad_norm": 0.16015562415122986,
      "learning_rate": 7.352715674548411e-05,
      "loss": 0.7121,
      "mean_token_accuracy": 0.7899916768074036,
      "num_tokens": 275552922.0,
      "step": 8633
    },
    {
      "entropy": 0.7375677917152643,
      "epoch": 0.7943830110235134,
      "grad_norm": 0.14811429381370544,
      "learning_rate": 7.352408991934248e-05,
      "loss": 0.7249,
      "mean_token_accuracy": 0.7809288054704666,
      "num_tokens": 275584740.0,
      "step": 8634
    },
    {
      "entropy": 0.8046506606042385,
      "epoch": 0.7944750173949546,
      "grad_norm": 0.17203399538993835,
      "learning_rate": 7.352102309320084e-05,
      "loss": 0.8253,
      "mean_token_accuracy": 0.7674262449145317,
      "num_tokens": 275616788.0,
      "step": 8635
    },
    {
      "entropy": 0.749617600813508,
      "epoch": 0.7945670237663959,
      "grad_norm": 0.15952078998088837,
      "learning_rate": 7.351795626705923e-05,
      "loss": 0.7386,
      "mean_token_accuracy": 0.7817649841308594,
      "num_tokens": 275648832.0,
      "step": 8636
    },
    {
      "entropy": 0.6641843467950821,
      "epoch": 0.794659030137837,
      "grad_norm": 0.1518545001745224,
      "learning_rate": 7.35148894409176e-05,
      "loss": 0.6458,
      "mean_token_accuracy": 0.8047825545072556,
      "num_tokens": 275681337.0,
      "step": 8637
    },
    {
      "entropy": 0.8102107010781765,
      "epoch": 0.7947510365092783,
      "grad_norm": 0.17020367085933685,
      "learning_rate": 7.351182261477597e-05,
      "loss": 0.7951,
      "mean_token_accuracy": 0.7689100131392479,
      "num_tokens": 275713722.0,
      "step": 8638
    },
    {
      "entropy": 0.7284613084048033,
      "epoch": 0.7948430428807195,
      "grad_norm": 0.1512845903635025,
      "learning_rate": 7.350875578863434e-05,
      "loss": 0.7239,
      "mean_token_accuracy": 0.7883278056979179,
      "num_tokens": 275745345.0,
      "step": 8639
    },
    {
      "entropy": 0.7759668901562691,
      "epoch": 0.7949350492521607,
      "grad_norm": 0.1513255089521408,
      "learning_rate": 7.350568896249272e-05,
      "loss": 0.751,
      "mean_token_accuracy": 0.7834436632692814,
      "num_tokens": 275777339.0,
      "step": 8640
    },
    {
      "entropy": 0.8371236640959978,
      "epoch": 0.795027055623602,
      "grad_norm": 0.1595459282398224,
      "learning_rate": 7.350262213635109e-05,
      "loss": 0.8263,
      "mean_token_accuracy": 0.7535664848983288,
      "num_tokens": 275809106.0,
      "step": 8641
    },
    {
      "entropy": 0.7633695490658283,
      "epoch": 0.7951190619950431,
      "grad_norm": 0.1672288030385971,
      "learning_rate": 7.349955531020947e-05,
      "loss": 0.7468,
      "mean_token_accuracy": 0.7804617285728455,
      "num_tokens": 275840997.0,
      "step": 8642
    },
    {
      "entropy": 0.7817872315645218,
      "epoch": 0.7952110683664844,
      "grad_norm": 0.1579856276512146,
      "learning_rate": 7.349648848406784e-05,
      "loss": 0.7712,
      "mean_token_accuracy": 0.7789361625909805,
      "num_tokens": 275873216.0,
      "step": 8643
    },
    {
      "entropy": 0.7378325667232275,
      "epoch": 0.7953030747379256,
      "grad_norm": 0.15226268768310547,
      "learning_rate": 7.349342165792622e-05,
      "loss": 0.7281,
      "mean_token_accuracy": 0.7871750332415104,
      "num_tokens": 275905806.0,
      "step": 8644
    },
    {
      "entropy": 0.8145085666328669,
      "epoch": 0.7953950811093669,
      "grad_norm": 0.1576419323682785,
      "learning_rate": 7.349035483178459e-05,
      "loss": 0.8043,
      "mean_token_accuracy": 0.7631847225129604,
      "num_tokens": 275937687.0,
      "step": 8645
    },
    {
      "entropy": 0.7110970262438059,
      "epoch": 0.7954870874808081,
      "grad_norm": 0.16469496488571167,
      "learning_rate": 7.348728800564296e-05,
      "loss": 0.7176,
      "mean_token_accuracy": 0.7880043126642704,
      "num_tokens": 275969271.0,
      "step": 8646
    },
    {
      "entropy": 0.7856228332966566,
      "epoch": 0.7955790938522492,
      "grad_norm": 0.16110827028751373,
      "learning_rate": 7.348422117950134e-05,
      "loss": 0.7953,
      "mean_token_accuracy": 0.7684356980025768,
      "num_tokens": 276001116.0,
      "step": 8647
    },
    {
      "entropy": 0.7973865792155266,
      "epoch": 0.7956711002236905,
      "grad_norm": 0.16899259388446808,
      "learning_rate": 7.348115435335972e-05,
      "loss": 0.8199,
      "mean_token_accuracy": 0.7635311968624592,
      "num_tokens": 276032945.0,
      "step": 8648
    },
    {
      "entropy": 0.7216665148735046,
      "epoch": 0.7957631065951317,
      "grad_norm": 0.16332603991031647,
      "learning_rate": 7.347808752721809e-05,
      "loss": 0.7262,
      "mean_token_accuracy": 0.7846609950065613,
      "num_tokens": 276064138.0,
      "step": 8649
    },
    {
      "entropy": 0.6718251667916775,
      "epoch": 0.795855112966573,
      "grad_norm": 0.15148352086544037,
      "learning_rate": 7.347502070107646e-05,
      "loss": 0.6642,
      "mean_token_accuracy": 0.801823478192091,
      "num_tokens": 276096105.0,
      "step": 8650
    },
    {
      "entropy": 0.8296542838215828,
      "epoch": 0.7959471193380142,
      "grad_norm": 0.16838936507701874,
      "learning_rate": 7.347195387493482e-05,
      "loss": 0.8328,
      "mean_token_accuracy": 0.755806315690279,
      "num_tokens": 276128253.0,
      "step": 8651
    },
    {
      "entropy": 0.6998727340251207,
      "epoch": 0.7960391257094553,
      "grad_norm": 0.1535198986530304,
      "learning_rate": 7.34688870487932e-05,
      "loss": 0.6857,
      "mean_token_accuracy": 0.7980379797518253,
      "num_tokens": 276160397.0,
      "step": 8652
    },
    {
      "entropy": 0.7737140357494354,
      "epoch": 0.7961311320808966,
      "grad_norm": 0.16579480469226837,
      "learning_rate": 7.346582022265159e-05,
      "loss": 0.7749,
      "mean_token_accuracy": 0.7687743678689003,
      "num_tokens": 276192401.0,
      "step": 8653
    },
    {
      "entropy": 0.7474843673408031,
      "epoch": 0.7962231384523378,
      "grad_norm": 0.161149263381958,
      "learning_rate": 7.346275339650996e-05,
      "loss": 0.7327,
      "mean_token_accuracy": 0.7795789241790771,
      "num_tokens": 276224390.0,
      "step": 8654
    },
    {
      "entropy": 0.7095134984701872,
      "epoch": 0.7963151448237791,
      "grad_norm": 0.15165315568447113,
      "learning_rate": 7.345968657036832e-05,
      "loss": 0.6852,
      "mean_token_accuracy": 0.7921774163842201,
      "num_tokens": 276256795.0,
      "step": 8655
    },
    {
      "entropy": 0.7780853118747473,
      "epoch": 0.7964071511952203,
      "grad_norm": 0.15648497641086578,
      "learning_rate": 7.34566197442267e-05,
      "loss": 0.7547,
      "mean_token_accuracy": 0.7774479165673256,
      "num_tokens": 276289205.0,
      "step": 8656
    },
    {
      "entropy": 0.7889130860567093,
      "epoch": 0.7964991575666615,
      "grad_norm": 0.16835397481918335,
      "learning_rate": 7.345355291808507e-05,
      "loss": 0.76,
      "mean_token_accuracy": 0.7705518156290054,
      "num_tokens": 276320263.0,
      "step": 8657
    },
    {
      "entropy": 0.7816012986004353,
      "epoch": 0.7965911639381027,
      "grad_norm": 0.15050090849399567,
      "learning_rate": 7.345048609194345e-05,
      "loss": 0.7611,
      "mean_token_accuracy": 0.7725670672953129,
      "num_tokens": 276352490.0,
      "step": 8658
    },
    {
      "entropy": 0.815913874655962,
      "epoch": 0.7966831703095439,
      "grad_norm": 0.15915587544441223,
      "learning_rate": 7.344741926580182e-05,
      "loss": 0.7886,
      "mean_token_accuracy": 0.7668149881064892,
      "num_tokens": 276384998.0,
      "step": 8659
    },
    {
      "entropy": 0.7811201587319374,
      "epoch": 0.7967751766809852,
      "grad_norm": 0.1592133790254593,
      "learning_rate": 7.34443524396602e-05,
      "loss": 0.7706,
      "mean_token_accuracy": 0.7704102098941803,
      "num_tokens": 276416417.0,
      "step": 8660
    },
    {
      "entropy": 0.6987581262364984,
      "epoch": 0.7968671830524264,
      "grad_norm": 0.15221545100212097,
      "learning_rate": 7.344128561351857e-05,
      "loss": 0.6641,
      "mean_token_accuracy": 0.8054456077516079,
      "num_tokens": 276448874.0,
      "step": 8661
    },
    {
      "entropy": 0.7686122385784984,
      "epoch": 0.7969591894238676,
      "grad_norm": 0.15034542977809906,
      "learning_rate": 7.343821878737694e-05,
      "loss": 0.777,
      "mean_token_accuracy": 0.7690579667687416,
      "num_tokens": 276481291.0,
      "step": 8662
    },
    {
      "entropy": 0.8562644035555422,
      "epoch": 0.7970511957953088,
      "grad_norm": 0.16038477420806885,
      "learning_rate": 7.343515196123532e-05,
      "loss": 0.8535,
      "mean_token_accuracy": 0.7549153566360474,
      "num_tokens": 276513667.0,
      "step": 8663
    },
    {
      "entropy": 0.7356301285326481,
      "epoch": 0.79714320216675,
      "grad_norm": 0.15742000937461853,
      "learning_rate": 7.34320851350937e-05,
      "loss": 0.7437,
      "mean_token_accuracy": 0.7810259982943535,
      "num_tokens": 276545662.0,
      "step": 8664
    },
    {
      "entropy": 0.6638048551976681,
      "epoch": 0.7972352085381913,
      "grad_norm": 0.1561233252286911,
      "learning_rate": 7.342901830895207e-05,
      "loss": 0.6501,
      "mean_token_accuracy": 0.8066902607679367,
      "num_tokens": 276577897.0,
      "step": 8665
    },
    {
      "entropy": 0.6835841406136751,
      "epoch": 0.7973272149096325,
      "grad_norm": 0.1634513884782791,
      "learning_rate": 7.342595148281044e-05,
      "loss": 0.6619,
      "mean_token_accuracy": 0.798008993268013,
      "num_tokens": 276609292.0,
      "step": 8666
    },
    {
      "entropy": 0.6109954845160246,
      "epoch": 0.7974192212810737,
      "grad_norm": 0.1543463170528412,
      "learning_rate": 7.342288465666882e-05,
      "loss": 0.597,
      "mean_token_accuracy": 0.8212644495069981,
      "num_tokens": 276641200.0,
      "step": 8667
    },
    {
      "entropy": 0.7426529694348574,
      "epoch": 0.7975112276525149,
      "grad_norm": 0.1565074324607849,
      "learning_rate": 7.341981783052719e-05,
      "loss": 0.7131,
      "mean_token_accuracy": 0.7879211641848087,
      "num_tokens": 276672906.0,
      "step": 8668
    },
    {
      "entropy": 0.7236555814743042,
      "epoch": 0.7976032340239562,
      "grad_norm": 0.1670968383550644,
      "learning_rate": 7.341675100438557e-05,
      "loss": 0.6841,
      "mean_token_accuracy": 0.7947533316910267,
      "num_tokens": 276704391.0,
      "step": 8669
    },
    {
      "entropy": 0.6885641515254974,
      "epoch": 0.7976952403953974,
      "grad_norm": 0.1544513702392578,
      "learning_rate": 7.341368417824394e-05,
      "loss": 0.6749,
      "mean_token_accuracy": 0.8012617863714695,
      "num_tokens": 276735687.0,
      "step": 8670
    },
    {
      "entropy": 0.7186551298946142,
      "epoch": 0.7977872467668387,
      "grad_norm": 0.15337605774402618,
      "learning_rate": 7.341061735210232e-05,
      "loss": 0.6968,
      "mean_token_accuracy": 0.7893872745335102,
      "num_tokens": 276767169.0,
      "step": 8671
    },
    {
      "entropy": 0.7140410942956805,
      "epoch": 0.7978792531382798,
      "grad_norm": 0.15784403681755066,
      "learning_rate": 7.340755052596069e-05,
      "loss": 0.717,
      "mean_token_accuracy": 0.7908207699656487,
      "num_tokens": 276799307.0,
      "step": 8672
    },
    {
      "entropy": 0.7336580008268356,
      "epoch": 0.797971259509721,
      "grad_norm": 0.16007699072360992,
      "learning_rate": 7.340448369981905e-05,
      "loss": 0.7118,
      "mean_token_accuracy": 0.7888613939285278,
      "num_tokens": 276831128.0,
      "step": 8673
    },
    {
      "entropy": 0.6848418023437262,
      "epoch": 0.7980632658811623,
      "grad_norm": 0.1710948646068573,
      "learning_rate": 7.340141687367743e-05,
      "loss": 0.6559,
      "mean_token_accuracy": 0.80562574416399,
      "num_tokens": 276862990.0,
      "step": 8674
    },
    {
      "entropy": 0.7612132085487247,
      "epoch": 0.7981552722526035,
      "grad_norm": 0.15439791977405548,
      "learning_rate": 7.339835004753582e-05,
      "loss": 0.7532,
      "mean_token_accuracy": 0.7762806378304958,
      "num_tokens": 276894773.0,
      "step": 8675
    },
    {
      "entropy": 0.7526262812316418,
      "epoch": 0.7982472786240448,
      "grad_norm": 0.16133055090904236,
      "learning_rate": 7.339528322139418e-05,
      "loss": 0.7359,
      "mean_token_accuracy": 0.7771245688199997,
      "num_tokens": 276926299.0,
      "step": 8676
    },
    {
      "entropy": 0.7019932009279728,
      "epoch": 0.7983392849954859,
      "grad_norm": 0.16710074245929718,
      "learning_rate": 7.339221639525255e-05,
      "loss": 0.702,
      "mean_token_accuracy": 0.7942386940121651,
      "num_tokens": 276958765.0,
      "step": 8677
    },
    {
      "entropy": 0.8070729710161686,
      "epoch": 0.7984312913669271,
      "grad_norm": 0.15946589410305023,
      "learning_rate": 7.338914956911093e-05,
      "loss": 0.7962,
      "mean_token_accuracy": 0.7678358778357506,
      "num_tokens": 276990599.0,
      "step": 8678
    },
    {
      "entropy": 0.7685145139694214,
      "epoch": 0.7985232977383684,
      "grad_norm": 0.1535957157611847,
      "learning_rate": 7.338608274296931e-05,
      "loss": 0.7524,
      "mean_token_accuracy": 0.7762715555727482,
      "num_tokens": 277022549.0,
      "step": 8679
    },
    {
      "entropy": 0.7281819079071283,
      "epoch": 0.7986153041098096,
      "grad_norm": 0.1561039388179779,
      "learning_rate": 7.338301591682768e-05,
      "loss": 0.7003,
      "mean_token_accuracy": 0.7871804945170879,
      "num_tokens": 277053972.0,
      "step": 8680
    },
    {
      "entropy": 0.7546409610658884,
      "epoch": 0.7987073104812509,
      "grad_norm": 0.1641690731048584,
      "learning_rate": 7.337994909068605e-05,
      "loss": 0.7469,
      "mean_token_accuracy": 0.7793191596865654,
      "num_tokens": 277086634.0,
      "step": 8681
    },
    {
      "entropy": 0.7258099932223558,
      "epoch": 0.798799316852692,
      "grad_norm": 0.15330316126346588,
      "learning_rate": 7.337688226454442e-05,
      "loss": 0.7087,
      "mean_token_accuracy": 0.7897977940738201,
      "num_tokens": 277118126.0,
      "step": 8682
    },
    {
      "entropy": 0.6473147533833981,
      "epoch": 0.7988913232241333,
      "grad_norm": 0.16976962983608246,
      "learning_rate": 7.33738154384028e-05,
      "loss": 0.6343,
      "mean_token_accuracy": 0.8130328804254532,
      "num_tokens": 277150788.0,
      "step": 8683
    },
    {
      "entropy": 0.7093620523810387,
      "epoch": 0.7989833295955745,
      "grad_norm": 0.15558768808841705,
      "learning_rate": 7.337074861226118e-05,
      "loss": 0.7172,
      "mean_token_accuracy": 0.7904948033392429,
      "num_tokens": 277183412.0,
      "step": 8684
    },
    {
      "entropy": 0.6786512937396765,
      "epoch": 0.7990753359670157,
      "grad_norm": 0.1613267958164215,
      "learning_rate": 7.336768178611955e-05,
      "loss": 0.6826,
      "mean_token_accuracy": 0.7981942854821682,
      "num_tokens": 277215441.0,
      "step": 8685
    },
    {
      "entropy": 0.6748366234824061,
      "epoch": 0.799167342338457,
      "grad_norm": 0.15456122159957886,
      "learning_rate": 7.336461495997792e-05,
      "loss": 0.6541,
      "mean_token_accuracy": 0.8098076432943344,
      "num_tokens": 277247000.0,
      "step": 8686
    },
    {
      "entropy": 0.7227402869611979,
      "epoch": 0.7992593487098981,
      "grad_norm": 0.159025177359581,
      "learning_rate": 7.33615481338363e-05,
      "loss": 0.7123,
      "mean_token_accuracy": 0.7895055040717125,
      "num_tokens": 277277958.0,
      "step": 8687
    },
    {
      "entropy": 0.7220110092312098,
      "epoch": 0.7993513550813394,
      "grad_norm": 0.16631846129894257,
      "learning_rate": 7.335848130769467e-05,
      "loss": 0.7306,
      "mean_token_accuracy": 0.7850747294723988,
      "num_tokens": 277309642.0,
      "step": 8688
    },
    {
      "entropy": 0.7130556087940931,
      "epoch": 0.7994433614527806,
      "grad_norm": 0.16178148984909058,
      "learning_rate": 7.335541448155305e-05,
      "loss": 0.7003,
      "mean_token_accuracy": 0.7939271405339241,
      "num_tokens": 277341644.0,
      "step": 8689
    },
    {
      "entropy": 0.7425155211240053,
      "epoch": 0.7995353678242219,
      "grad_norm": 0.15410476922988892,
      "learning_rate": 7.335234765541142e-05,
      "loss": 0.728,
      "mean_token_accuracy": 0.7828730829060078,
      "num_tokens": 277373582.0,
      "step": 8690
    },
    {
      "entropy": 0.6797556346282363,
      "epoch": 0.7996273741956631,
      "grad_norm": 0.14900648593902588,
      "learning_rate": 7.33492808292698e-05,
      "loss": 0.6587,
      "mean_token_accuracy": 0.8034528084099293,
      "num_tokens": 277405720.0,
      "step": 8691
    },
    {
      "entropy": 0.7183944098651409,
      "epoch": 0.7997193805671042,
      "grad_norm": 0.15787851810455322,
      "learning_rate": 7.334621400312816e-05,
      "loss": 0.7065,
      "mean_token_accuracy": 0.7908086962997913,
      "num_tokens": 277437819.0,
      "step": 8692
    },
    {
      "entropy": 0.8576727956533432,
      "epoch": 0.7998113869385455,
      "grad_norm": 0.16370825469493866,
      "learning_rate": 7.334314717698653e-05,
      "loss": 0.8472,
      "mean_token_accuracy": 0.756616685539484,
      "num_tokens": 277469928.0,
      "step": 8693
    },
    {
      "entropy": 0.6558112646453083,
      "epoch": 0.7999033933099867,
      "grad_norm": 0.14706553518772125,
      "learning_rate": 7.334008035084491e-05,
      "loss": 0.6421,
      "mean_token_accuracy": 0.8108944296836853,
      "num_tokens": 277502167.0,
      "step": 8694
    },
    {
      "entropy": 0.6259125154465437,
      "epoch": 0.799995399681428,
      "grad_norm": 0.15183736383914948,
      "learning_rate": 7.33370135247033e-05,
      "loss": 0.6169,
      "mean_token_accuracy": 0.8158285468816757,
      "num_tokens": 277534499.0,
      "step": 8695
    },
    {
      "entropy": 0.7401563841849566,
      "epoch": 0.8000874060528692,
      "grad_norm": 0.16472914814949036,
      "learning_rate": 7.333394669856166e-05,
      "loss": 0.7298,
      "mean_token_accuracy": 0.7833366319537163,
      "num_tokens": 277566539.0,
      "step": 8696
    },
    {
      "entropy": 0.7483876124024391,
      "epoch": 0.8001794124243103,
      "grad_norm": 0.1498461365699768,
      "learning_rate": 7.333087987242003e-05,
      "loss": 0.7454,
      "mean_token_accuracy": 0.779013030230999,
      "num_tokens": 277598643.0,
      "step": 8697
    },
    {
      "entropy": 0.6960453726351261,
      "epoch": 0.8002714187957516,
      "grad_norm": 0.14938275516033173,
      "learning_rate": 7.332781304627841e-05,
      "loss": 0.673,
      "mean_token_accuracy": 0.7992946580052376,
      "num_tokens": 277630138.0,
      "step": 8698
    },
    {
      "entropy": 0.7879940047860146,
      "epoch": 0.8003634251671928,
      "grad_norm": 0.1641312688589096,
      "learning_rate": 7.332474622013678e-05,
      "loss": 0.7906,
      "mean_token_accuracy": 0.7687594965100288,
      "num_tokens": 277662010.0,
      "step": 8699
    },
    {
      "entropy": 0.6813267953693867,
      "epoch": 0.8004554315386341,
      "grad_norm": 0.15466268360614777,
      "learning_rate": 7.332167939399516e-05,
      "loss": 0.681,
      "mean_token_accuracy": 0.7962678074836731,
      "num_tokens": 277693905.0,
      "step": 8700
    },
    {
      "entropy": 0.6661997158080339,
      "epoch": 0.8005474379100753,
      "grad_norm": 0.16913190484046936,
      "learning_rate": 7.331861256785353e-05,
      "loss": 0.6475,
      "mean_token_accuracy": 0.8062079288065434,
      "num_tokens": 277725455.0,
      "step": 8701
    },
    {
      "entropy": 0.7527295667678118,
      "epoch": 0.8006394442815165,
      "grad_norm": 0.14732486009597778,
      "learning_rate": 7.331554574171191e-05,
      "loss": 0.7459,
      "mean_token_accuracy": 0.7816265635192394,
      "num_tokens": 277756909.0,
      "step": 8702
    },
    {
      "entropy": 0.7220632694661617,
      "epoch": 0.8007314506529577,
      "grad_norm": 0.15043924748897552,
      "learning_rate": 7.331247891557028e-05,
      "loss": 0.718,
      "mean_token_accuracy": 0.7884098626673222,
      "num_tokens": 277789677.0,
      "step": 8703
    },
    {
      "entropy": 0.8219965286552906,
      "epoch": 0.8008234570243989,
      "grad_norm": 0.1551295518875122,
      "learning_rate": 7.330941208942865e-05,
      "loss": 0.8172,
      "mean_token_accuracy": 0.7622773312032223,
      "num_tokens": 277821924.0,
      "step": 8704
    },
    {
      "entropy": 0.5846951603889465,
      "epoch": 0.8009154633958402,
      "grad_norm": 0.14840760827064514,
      "learning_rate": 7.330634526328703e-05,
      "loss": 0.583,
      "mean_token_accuracy": 0.8244711756706238,
      "num_tokens": 277853997.0,
      "step": 8705
    },
    {
      "entropy": 0.7002903008833528,
      "epoch": 0.8010074697672814,
      "grad_norm": 0.1630277782678604,
      "learning_rate": 7.330327843714541e-05,
      "loss": 0.6791,
      "mean_token_accuracy": 0.8004847094416618,
      "num_tokens": 277886204.0,
      "step": 8706
    },
    {
      "entropy": 0.7941499054431915,
      "epoch": 0.8010994761387226,
      "grad_norm": 0.15977531671524048,
      "learning_rate": 7.330021161100378e-05,
      "loss": 0.7817,
      "mean_token_accuracy": 0.7736492678523064,
      "num_tokens": 277918007.0,
      "step": 8707
    },
    {
      "entropy": 0.7483111601322889,
      "epoch": 0.8011914825101638,
      "grad_norm": 0.1610647737979889,
      "learning_rate": 7.329714478486215e-05,
      "loss": 0.7288,
      "mean_token_accuracy": 0.7870013453066349,
      "num_tokens": 277950047.0,
      "step": 8708
    },
    {
      "entropy": 0.8663951978087425,
      "epoch": 0.801283488881605,
      "grad_norm": 0.18526670336723328,
      "learning_rate": 7.329407795872051e-05,
      "loss": 0.8572,
      "mean_token_accuracy": 0.7490371391177177,
      "num_tokens": 277981906.0,
      "step": 8709
    },
    {
      "entropy": 0.6642525680363178,
      "epoch": 0.8013754952530463,
      "grad_norm": 0.15457867085933685,
      "learning_rate": 7.329101113257891e-05,
      "loss": 0.6494,
      "mean_token_accuracy": 0.8068163990974426,
      "num_tokens": 278014213.0,
      "step": 8710
    },
    {
      "entropy": 0.7144221179187298,
      "epoch": 0.8014675016244875,
      "grad_norm": 0.15216219425201416,
      "learning_rate": 7.328794430643728e-05,
      "loss": 0.7072,
      "mean_token_accuracy": 0.7916690111160278,
      "num_tokens": 278046474.0,
      "step": 8711
    },
    {
      "entropy": 0.6468470860272646,
      "epoch": 0.8015595079959287,
      "grad_norm": 0.1518307328224182,
      "learning_rate": 7.328487748029564e-05,
      "loss": 0.6423,
      "mean_token_accuracy": 0.8126125521957874,
      "num_tokens": 278079038.0,
      "step": 8712
    },
    {
      "entropy": 0.7223338698968291,
      "epoch": 0.8016515143673699,
      "grad_norm": 0.15718403458595276,
      "learning_rate": 7.328181065415401e-05,
      "loss": 0.7252,
      "mean_token_accuracy": 0.7875550799071789,
      "num_tokens": 278111035.0,
      "step": 8713
    },
    {
      "entropy": 0.7767902743071318,
      "epoch": 0.8017435207388112,
      "grad_norm": 0.1617385298013687,
      "learning_rate": 7.32787438280124e-05,
      "loss": 0.762,
      "mean_token_accuracy": 0.7805095911026001,
      "num_tokens": 278143517.0,
      "step": 8714
    },
    {
      "entropy": 0.8084343895316124,
      "epoch": 0.8018355271102524,
      "grad_norm": 0.16689491271972656,
      "learning_rate": 7.327567700187077e-05,
      "loss": 0.8036,
      "mean_token_accuracy": 0.7646379433572292,
      "num_tokens": 278175949.0,
      "step": 8715
    },
    {
      "entropy": 0.677094804123044,
      "epoch": 0.8019275334816937,
      "grad_norm": 0.15737442672252655,
      "learning_rate": 7.327261017572914e-05,
      "loss": 0.6745,
      "mean_token_accuracy": 0.8010976389050484,
      "num_tokens": 278206930.0,
      "step": 8716
    },
    {
      "entropy": 0.6587237641215324,
      "epoch": 0.8020195398531348,
      "grad_norm": 0.15219268202781677,
      "learning_rate": 7.326954334958751e-05,
      "loss": 0.6382,
      "mean_token_accuracy": 0.8060249201953411,
      "num_tokens": 278238985.0,
      "step": 8717
    },
    {
      "entropy": 0.7126543559134007,
      "epoch": 0.802111546224576,
      "grad_norm": 0.1618020236492157,
      "learning_rate": 7.326647652344589e-05,
      "loss": 0.6923,
      "mean_token_accuracy": 0.7932676561176777,
      "num_tokens": 278270673.0,
      "step": 8718
    },
    {
      "entropy": 0.8643290400505066,
      "epoch": 0.8022035525960173,
      "grad_norm": 0.16150318086147308,
      "learning_rate": 7.326340969730426e-05,
      "loss": 0.8641,
      "mean_token_accuracy": 0.7541092038154602,
      "num_tokens": 278302999.0,
      "step": 8719
    },
    {
      "entropy": 0.7101536775007844,
      "epoch": 0.8022955589674585,
      "grad_norm": 0.1548486351966858,
      "learning_rate": 7.326034287116264e-05,
      "loss": 0.6871,
      "mean_token_accuracy": 0.799552146345377,
      "num_tokens": 278334911.0,
      "step": 8720
    },
    {
      "entropy": 0.701247725635767,
      "epoch": 0.8023875653388998,
      "grad_norm": 0.15377475321292877,
      "learning_rate": 7.325727604502101e-05,
      "loss": 0.6835,
      "mean_token_accuracy": 0.799418430775404,
      "num_tokens": 278366696.0,
      "step": 8721
    },
    {
      "entropy": 0.6602160832844675,
      "epoch": 0.8024795717103409,
      "grad_norm": 0.15044865012168884,
      "learning_rate": 7.325420921887939e-05,
      "loss": 0.6539,
      "mean_token_accuracy": 0.803980864584446,
      "num_tokens": 278398595.0,
      "step": 8722
    },
    {
      "entropy": 0.8016211502254009,
      "epoch": 0.8025715780817821,
      "grad_norm": 0.16784293949604034,
      "learning_rate": 7.325114239273776e-05,
      "loss": 0.7936,
      "mean_token_accuracy": 0.7657866030931473,
      "num_tokens": 278430127.0,
      "step": 8723
    },
    {
      "entropy": 0.7656143456697464,
      "epoch": 0.8026635844532234,
      "grad_norm": 0.15474581718444824,
      "learning_rate": 7.324807556659613e-05,
      "loss": 0.7466,
      "mean_token_accuracy": 0.7843096181750298,
      "num_tokens": 278461639.0,
      "step": 8724
    },
    {
      "entropy": 0.6726931966841221,
      "epoch": 0.8027555908246646,
      "grad_norm": 0.15069837868213654,
      "learning_rate": 7.324500874045451e-05,
      "loss": 0.6505,
      "mean_token_accuracy": 0.8045010343194008,
      "num_tokens": 278494081.0,
      "step": 8725
    },
    {
      "entropy": 0.6026321370154619,
      "epoch": 0.8028475971961059,
      "grad_norm": 0.1512342244386673,
      "learning_rate": 7.324194191431289e-05,
      "loss": 0.589,
      "mean_token_accuracy": 0.8220278434455395,
      "num_tokens": 278525866.0,
      "step": 8726
    },
    {
      "entropy": 0.8075893130153418,
      "epoch": 0.802939603567547,
      "grad_norm": 0.14818674325942993,
      "learning_rate": 7.323887508817126e-05,
      "loss": 0.7935,
      "mean_token_accuracy": 0.765643373131752,
      "num_tokens": 278558359.0,
      "step": 8727
    },
    {
      "entropy": 0.6202487647533417,
      "epoch": 0.8030316099389883,
      "grad_norm": 0.14054229855537415,
      "learning_rate": 7.323580826202962e-05,
      "loss": 0.608,
      "mean_token_accuracy": 0.8147233612835407,
      "num_tokens": 278590276.0,
      "step": 8728
    },
    {
      "entropy": 0.761592710390687,
      "epoch": 0.8031236163104295,
      "grad_norm": 0.156532883644104,
      "learning_rate": 7.3232741435888e-05,
      "loss": 0.7647,
      "mean_token_accuracy": 0.7748839184641838,
      "num_tokens": 278622357.0,
      "step": 8729
    },
    {
      "entropy": 0.7081341221928596,
      "epoch": 0.8032156226818707,
      "grad_norm": 0.15648281574249268,
      "learning_rate": 7.322967460974637e-05,
      "loss": 0.6899,
      "mean_token_accuracy": 0.7950776666402817,
      "num_tokens": 278653591.0,
      "step": 8730
    },
    {
      "entropy": 0.7878997065126896,
      "epoch": 0.803307629053312,
      "grad_norm": 0.15642079710960388,
      "learning_rate": 7.322660778360476e-05,
      "loss": 0.7683,
      "mean_token_accuracy": 0.770071305334568,
      "num_tokens": 278684869.0,
      "step": 8731
    },
    {
      "entropy": 0.6786263203248382,
      "epoch": 0.8033996354247531,
      "grad_norm": 0.15891626477241516,
      "learning_rate": 7.322354095746312e-05,
      "loss": 0.6695,
      "mean_token_accuracy": 0.7994782254099846,
      "num_tokens": 278716387.0,
      "step": 8732
    },
    {
      "entropy": 0.7461670525372028,
      "epoch": 0.8034916417961944,
      "grad_norm": 0.159740149974823,
      "learning_rate": 7.32204741313215e-05,
      "loss": 0.7274,
      "mean_token_accuracy": 0.7908696383237839,
      "num_tokens": 278748481.0,
      "step": 8733
    },
    {
      "entropy": 0.6884655393660069,
      "epoch": 0.8035836481676356,
      "grad_norm": 0.1571081280708313,
      "learning_rate": 7.321740730517987e-05,
      "loss": 0.6948,
      "mean_token_accuracy": 0.7933191321790218,
      "num_tokens": 278780179.0,
      "step": 8734
    },
    {
      "entropy": 0.7051366921514273,
      "epoch": 0.8036756545390769,
      "grad_norm": 0.15204043686389923,
      "learning_rate": 7.321434047903824e-05,
      "loss": 0.6995,
      "mean_token_accuracy": 0.7939106188714504,
      "num_tokens": 278812392.0,
      "step": 8735
    },
    {
      "entropy": 0.7418066514655948,
      "epoch": 0.8037676609105181,
      "grad_norm": 0.1719076782464981,
      "learning_rate": 7.321127365289662e-05,
      "loss": 0.7269,
      "mean_token_accuracy": 0.7858679890632629,
      "num_tokens": 278843833.0,
      "step": 8736
    },
    {
      "entropy": 0.7101117242127657,
      "epoch": 0.8038596672819592,
      "grad_norm": 0.15575529634952545,
      "learning_rate": 7.3208206826755e-05,
      "loss": 0.7052,
      "mean_token_accuracy": 0.7888794057071209,
      "num_tokens": 278876084.0,
      "step": 8737
    },
    {
      "entropy": 0.6791186928749084,
      "epoch": 0.8039516736534005,
      "grad_norm": 0.16253961622714996,
      "learning_rate": 7.320514000061337e-05,
      "loss": 0.6616,
      "mean_token_accuracy": 0.8061065338551998,
      "num_tokens": 278907970.0,
      "step": 8738
    },
    {
      "entropy": 0.794282941147685,
      "epoch": 0.8040436800248417,
      "grad_norm": 0.16508202254772186,
      "learning_rate": 7.320207317447174e-05,
      "loss": 0.7876,
      "mean_token_accuracy": 0.7680122330784798,
      "num_tokens": 278939596.0,
      "step": 8739
    },
    {
      "entropy": 0.6267684977501631,
      "epoch": 0.804135686396283,
      "grad_norm": 0.16101735830307007,
      "learning_rate": 7.319900634833011e-05,
      "loss": 0.6036,
      "mean_token_accuracy": 0.8127685971558094,
      "num_tokens": 278970763.0,
      "step": 8740
    },
    {
      "entropy": 0.5681100264191628,
      "epoch": 0.8042276927677242,
      "grad_norm": 0.1554991602897644,
      "learning_rate": 7.319593952218849e-05,
      "loss": 0.5582,
      "mean_token_accuracy": 0.8313648700714111,
      "num_tokens": 279002883.0,
      "step": 8741
    },
    {
      "entropy": 0.649780485779047,
      "epoch": 0.8043196991391653,
      "grad_norm": 0.15616890788078308,
      "learning_rate": 7.319287269604687e-05,
      "loss": 0.6348,
      "mean_token_accuracy": 0.8151688501238823,
      "num_tokens": 279034984.0,
      "step": 8742
    },
    {
      "entropy": 0.6754046753048897,
      "epoch": 0.8044117055106066,
      "grad_norm": 0.14419010281562805,
      "learning_rate": 7.318980586990524e-05,
      "loss": 0.6615,
      "mean_token_accuracy": 0.8046204261481762,
      "num_tokens": 279067698.0,
      "step": 8743
    },
    {
      "entropy": 0.6780106793157756,
      "epoch": 0.8045037118820478,
      "grad_norm": 0.15571212768554688,
      "learning_rate": 7.31867390437636e-05,
      "loss": 0.665,
      "mean_token_accuracy": 0.8009248077869415,
      "num_tokens": 279099656.0,
      "step": 8744
    },
    {
      "entropy": 0.8162395134568214,
      "epoch": 0.8045957182534891,
      "grad_norm": 0.17385460436344147,
      "learning_rate": 7.318367221762199e-05,
      "loss": 0.8123,
      "mean_token_accuracy": 0.762108501046896,
      "num_tokens": 279131443.0,
      "step": 8745
    },
    {
      "entropy": 0.7638941761106253,
      "epoch": 0.8046877246249303,
      "grad_norm": 0.1583751142024994,
      "learning_rate": 7.318060539148035e-05,
      "loss": 0.7468,
      "mean_token_accuracy": 0.7787830047309399,
      "num_tokens": 279162949.0,
      "step": 8746
    },
    {
      "entropy": 0.7053752020001411,
      "epoch": 0.8047797309963715,
      "grad_norm": 0.15668143332004547,
      "learning_rate": 7.317753856533874e-05,
      "loss": 0.6965,
      "mean_token_accuracy": 0.7940533608198166,
      "num_tokens": 279195648.0,
      "step": 8747
    },
    {
      "entropy": 0.764478761702776,
      "epoch": 0.8048717373678127,
      "grad_norm": 0.15658257901668549,
      "learning_rate": 7.31744717391971e-05,
      "loss": 0.7563,
      "mean_token_accuracy": 0.7804531790316105,
      "num_tokens": 279228105.0,
      "step": 8748
    },
    {
      "entropy": 0.7640081867575645,
      "epoch": 0.804963743739254,
      "grad_norm": 0.1552649736404419,
      "learning_rate": 7.317140491305549e-05,
      "loss": 0.7618,
      "mean_token_accuracy": 0.7718638107180595,
      "num_tokens": 279259051.0,
      "step": 8749
    },
    {
      "entropy": 0.7290884321555495,
      "epoch": 0.8050557501106952,
      "grad_norm": 0.15624821186065674,
      "learning_rate": 7.316833808691385e-05,
      "loss": 0.7274,
      "mean_token_accuracy": 0.78649527952075,
      "num_tokens": 279290600.0,
      "step": 8750
    },
    {
      "entropy": 0.7289764527231455,
      "epoch": 0.8051477564821364,
      "grad_norm": 0.1598057895898819,
      "learning_rate": 7.316527126077222e-05,
      "loss": 0.722,
      "mean_token_accuracy": 0.7894418723881245,
      "num_tokens": 279321742.0,
      "step": 8751
    },
    {
      "entropy": 0.727693947032094,
      "epoch": 0.8052397628535776,
      "grad_norm": 0.16058385372161865,
      "learning_rate": 7.31622044346306e-05,
      "loss": 0.7063,
      "mean_token_accuracy": 0.792288426309824,
      "num_tokens": 279353049.0,
      "step": 8752
    },
    {
      "entropy": 0.8461899943649769,
      "epoch": 0.8053317692250188,
      "grad_norm": 0.16353468596935272,
      "learning_rate": 7.315913760848898e-05,
      "loss": 0.8337,
      "mean_token_accuracy": 0.7580994479358196,
      "num_tokens": 279384153.0,
      "step": 8753
    },
    {
      "entropy": 0.717794869095087,
      "epoch": 0.8054237755964601,
      "grad_norm": 0.1647447794675827,
      "learning_rate": 7.315607078234735e-05,
      "loss": 0.7018,
      "mean_token_accuracy": 0.7961878068745136,
      "num_tokens": 279416197.0,
      "step": 8754
    },
    {
      "entropy": 0.815524410456419,
      "epoch": 0.8055157819679013,
      "grad_norm": 0.15477491915225983,
      "learning_rate": 7.315300395620572e-05,
      "loss": 0.7831,
      "mean_token_accuracy": 0.7716096118092537,
      "num_tokens": 279447460.0,
      "step": 8755
    },
    {
      "entropy": 0.6157420948147774,
      "epoch": 0.8056077883393425,
      "grad_norm": 0.1662493348121643,
      "learning_rate": 7.31499371300641e-05,
      "loss": 0.5984,
      "mean_token_accuracy": 0.8204924613237381,
      "num_tokens": 279479643.0,
      "step": 8756
    },
    {
      "entropy": 0.7537859617732465,
      "epoch": 0.8056997947107837,
      "grad_norm": 0.1528196483850479,
      "learning_rate": 7.314687030392248e-05,
      "loss": 0.7304,
      "mean_token_accuracy": 0.785501517355442,
      "num_tokens": 279512129.0,
      "step": 8757
    },
    {
      "entropy": 0.674949717707932,
      "epoch": 0.8057918010822249,
      "grad_norm": 0.15661700069904327,
      "learning_rate": 7.314380347778085e-05,
      "loss": 0.6565,
      "mean_token_accuracy": 0.8029296509921551,
      "num_tokens": 279544417.0,
      "step": 8758
    },
    {
      "entropy": 0.6889066845178604,
      "epoch": 0.8058838074536662,
      "grad_norm": 0.1549614816904068,
      "learning_rate": 7.314073665163922e-05,
      "loss": 0.6653,
      "mean_token_accuracy": 0.8001908957958221,
      "num_tokens": 279576002.0,
      "step": 8759
    },
    {
      "entropy": 0.7251291740685701,
      "epoch": 0.8059758138251074,
      "grad_norm": 0.1630224585533142,
      "learning_rate": 7.31376698254976e-05,
      "loss": 0.7217,
      "mean_token_accuracy": 0.7858989797532558,
      "num_tokens": 279607020.0,
      "step": 8760
    },
    {
      "entropy": 0.7478988692164421,
      "epoch": 0.8060678201965487,
      "grad_norm": 0.15109646320343018,
      "learning_rate": 7.313460299935597e-05,
      "loss": 0.7468,
      "mean_token_accuracy": 0.7814122959971428,
      "num_tokens": 279639067.0,
      "step": 8761
    },
    {
      "entropy": 0.6018506120890379,
      "epoch": 0.8061598265679898,
      "grad_norm": 0.15372908115386963,
      "learning_rate": 7.313153617321435e-05,
      "loss": 0.5829,
      "mean_token_accuracy": 0.8228732533752918,
      "num_tokens": 279670714.0,
      "step": 8762
    },
    {
      "entropy": 0.6822966281324625,
      "epoch": 0.806251832939431,
      "grad_norm": 0.15836812555789948,
      "learning_rate": 7.312846934707272e-05,
      "loss": 0.6695,
      "mean_token_accuracy": 0.8001716956496239,
      "num_tokens": 279702562.0,
      "step": 8763
    },
    {
      "entropy": 0.6144524896517396,
      "epoch": 0.8063438393108723,
      "grad_norm": 0.1546555459499359,
      "learning_rate": 7.31254025209311e-05,
      "loss": 0.5958,
      "mean_token_accuracy": 0.821291096508503,
      "num_tokens": 279734457.0,
      "step": 8764
    },
    {
      "entropy": 0.6787190530449152,
      "epoch": 0.8064358456823135,
      "grad_norm": 0.16779188811779022,
      "learning_rate": 7.312233569478947e-05,
      "loss": 0.6765,
      "mean_token_accuracy": 0.7989127673208714,
      "num_tokens": 279766373.0,
      "step": 8765
    },
    {
      "entropy": 0.6551671717315912,
      "epoch": 0.8065278520537548,
      "grad_norm": 0.1707802265882492,
      "learning_rate": 7.311926886864783e-05,
      "loss": 0.6326,
      "mean_token_accuracy": 0.8082478009164333,
      "num_tokens": 279797342.0,
      "step": 8766
    },
    {
      "entropy": 0.7121847751550376,
      "epoch": 0.8066198584251959,
      "grad_norm": 0.16339369118213654,
      "learning_rate": 7.311620204250622e-05,
      "loss": 0.7078,
      "mean_token_accuracy": 0.7933313101530075,
      "num_tokens": 279829178.0,
      "step": 8767
    },
    {
      "entropy": 0.7780629508197308,
      "epoch": 0.8067118647966371,
      "grad_norm": 0.16245633363723755,
      "learning_rate": 7.31131352163646e-05,
      "loss": 0.7664,
      "mean_token_accuracy": 0.7753198742866516,
      "num_tokens": 279861716.0,
      "step": 8768
    },
    {
      "entropy": 0.6603033030405641,
      "epoch": 0.8068038711680784,
      "grad_norm": 0.15072911977767944,
      "learning_rate": 7.311006839022297e-05,
      "loss": 0.6491,
      "mean_token_accuracy": 0.808539692312479,
      "num_tokens": 279894083.0,
      "step": 8769
    },
    {
      "entropy": 0.7447666190564632,
      "epoch": 0.8068958775395196,
      "grad_norm": 0.15414448082447052,
      "learning_rate": 7.310700156408133e-05,
      "loss": 0.7261,
      "mean_token_accuracy": 0.7853901386260986,
      "num_tokens": 279925804.0,
      "step": 8770
    },
    {
      "entropy": 0.7642114870250225,
      "epoch": 0.8069878839109609,
      "grad_norm": 0.16136810183525085,
      "learning_rate": 7.31039347379397e-05,
      "loss": 0.7493,
      "mean_token_accuracy": 0.7821925990283489,
      "num_tokens": 279957513.0,
      "step": 8771
    },
    {
      "entropy": 0.6411701454780996,
      "epoch": 0.807079890282402,
      "grad_norm": 0.1461721658706665,
      "learning_rate": 7.310086791179808e-05,
      "loss": 0.6155,
      "mean_token_accuracy": 0.8121973536908627,
      "num_tokens": 279989492.0,
      "step": 8772
    },
    {
      "entropy": 0.716025922447443,
      "epoch": 0.8071718966538433,
      "grad_norm": 0.15767790377140045,
      "learning_rate": 7.309780108565646e-05,
      "loss": 0.6904,
      "mean_token_accuracy": 0.7948335558176041,
      "num_tokens": 280021064.0,
      "step": 8773
    },
    {
      "entropy": 0.7803935967385769,
      "epoch": 0.8072639030252845,
      "grad_norm": 0.15702877938747406,
      "learning_rate": 7.309473425951483e-05,
      "loss": 0.7722,
      "mean_token_accuracy": 0.7742666453123093,
      "num_tokens": 280052760.0,
      "step": 8774
    },
    {
      "entropy": 0.7312551867216825,
      "epoch": 0.8073559093967257,
      "grad_norm": 0.153609499335289,
      "learning_rate": 7.30916674333732e-05,
      "loss": 0.7124,
      "mean_token_accuracy": 0.7903361357748508,
      "num_tokens": 280083883.0,
      "step": 8775
    },
    {
      "entropy": 0.6171490252017975,
      "epoch": 0.807447915768167,
      "grad_norm": 0.16608518362045288,
      "learning_rate": 7.308860060723158e-05,
      "loss": 0.6033,
      "mean_token_accuracy": 0.8196253105998039,
      "num_tokens": 280115977.0,
      "step": 8776
    },
    {
      "entropy": 0.6944465292617679,
      "epoch": 0.8075399221396081,
      "grad_norm": 0.15697593986988068,
      "learning_rate": 7.308553378108995e-05,
      "loss": 0.6537,
      "mean_token_accuracy": 0.8060253337025642,
      "num_tokens": 280148085.0,
      "step": 8777
    },
    {
      "entropy": 0.7477563815191388,
      "epoch": 0.8076319285110494,
      "grad_norm": 0.16390866041183472,
      "learning_rate": 7.308246695494833e-05,
      "loss": 0.7298,
      "mean_token_accuracy": 0.7943333126604557,
      "num_tokens": 280178993.0,
      "step": 8778
    },
    {
      "entropy": 0.7890052832663059,
      "epoch": 0.8077239348824906,
      "grad_norm": 0.15958985686302185,
      "learning_rate": 7.30794001288067e-05,
      "loss": 0.7784,
      "mean_token_accuracy": 0.7717009298503399,
      "num_tokens": 280210723.0,
      "step": 8779
    },
    {
      "entropy": 0.6532865557819605,
      "epoch": 0.8078159412539319,
      "grad_norm": 0.14981915056705475,
      "learning_rate": 7.307633330266508e-05,
      "loss": 0.6362,
      "mean_token_accuracy": 0.8062329888343811,
      "num_tokens": 280242276.0,
      "step": 8780
    },
    {
      "entropy": 0.7250798530876637,
      "epoch": 0.8079079476253731,
      "grad_norm": 0.14457906782627106,
      "learning_rate": 7.307326647652345e-05,
      "loss": 0.7094,
      "mean_token_accuracy": 0.7900694347918034,
      "num_tokens": 280274448.0,
      "step": 8781
    },
    {
      "entropy": 0.8043064400553703,
      "epoch": 0.8079999539968142,
      "grad_norm": 0.16484344005584717,
      "learning_rate": 7.307019965038182e-05,
      "loss": 0.786,
      "mean_token_accuracy": 0.7698336355388165,
      "num_tokens": 280305154.0,
      "step": 8782
    },
    {
      "entropy": 0.6740787765011191,
      "epoch": 0.8080919603682555,
      "grad_norm": 0.16965307295322418,
      "learning_rate": 7.30671328242402e-05,
      "loss": 0.6563,
      "mean_token_accuracy": 0.8050937242805958,
      "num_tokens": 280336962.0,
      "step": 8783
    },
    {
      "entropy": 0.8254942670464516,
      "epoch": 0.8081839667396967,
      "grad_norm": 0.1642749309539795,
      "learning_rate": 7.306406599809858e-05,
      "loss": 0.8321,
      "mean_token_accuracy": 0.7613161467015743,
      "num_tokens": 280368441.0,
      "step": 8784
    },
    {
      "entropy": 0.700109576806426,
      "epoch": 0.808275973111138,
      "grad_norm": 0.15932878851890564,
      "learning_rate": 7.306099917195695e-05,
      "loss": 0.6966,
      "mean_token_accuracy": 0.7926087751984596,
      "num_tokens": 280400048.0,
      "step": 8785
    },
    {
      "entropy": 0.653129430487752,
      "epoch": 0.8083679794825792,
      "grad_norm": 0.16596299409866333,
      "learning_rate": 7.305793234581531e-05,
      "loss": 0.6577,
      "mean_token_accuracy": 0.8042387515306473,
      "num_tokens": 280431656.0,
      "step": 8786
    },
    {
      "entropy": 0.6785263521596789,
      "epoch": 0.8084599858540203,
      "grad_norm": 0.16019345819950104,
      "learning_rate": 7.30548655196737e-05,
      "loss": 0.6685,
      "mean_token_accuracy": 0.7992553859949112,
      "num_tokens": 280463416.0,
      "step": 8787
    },
    {
      "entropy": 0.7235721964389086,
      "epoch": 0.8085519922254616,
      "grad_norm": 0.15358476340770721,
      "learning_rate": 7.305179869353206e-05,
      "loss": 0.7255,
      "mean_token_accuracy": 0.7882106862962246,
      "num_tokens": 280494869.0,
      "step": 8788
    },
    {
      "entropy": 0.7660503722727299,
      "epoch": 0.8086439985969028,
      "grad_norm": 0.1604904681444168,
      "learning_rate": 7.304873186739044e-05,
      "loss": 0.7835,
      "mean_token_accuracy": 0.7719584293663502,
      "num_tokens": 280526490.0,
      "step": 8789
    },
    {
      "entropy": 0.8399364203214645,
      "epoch": 0.8087360049683441,
      "grad_norm": 0.16268514096736908,
      "learning_rate": 7.304566504124881e-05,
      "loss": 0.8268,
      "mean_token_accuracy": 0.7574987187981606,
      "num_tokens": 280558782.0,
      "step": 8790
    },
    {
      "entropy": 0.6631929287686944,
      "epoch": 0.8088280113397853,
      "grad_norm": 0.16678792238235474,
      "learning_rate": 7.30425982151072e-05,
      "loss": 0.626,
      "mean_token_accuracy": 0.813741896301508,
      "num_tokens": 280590043.0,
      "step": 8791
    },
    {
      "entropy": 0.7218763995915651,
      "epoch": 0.8089200177112265,
      "grad_norm": 0.15815667808055878,
      "learning_rate": 7.303953138896556e-05,
      "loss": 0.7193,
      "mean_token_accuracy": 0.7891359701752663,
      "num_tokens": 280622045.0,
      "step": 8792
    },
    {
      "entropy": 0.787384893745184,
      "epoch": 0.8090120240826677,
      "grad_norm": 0.1612640768289566,
      "learning_rate": 7.303646456282394e-05,
      "loss": 0.7731,
      "mean_token_accuracy": 0.7696536928415298,
      "num_tokens": 280654042.0,
      "step": 8793
    },
    {
      "entropy": 0.7347054043784738,
      "epoch": 0.809104030454109,
      "grad_norm": 0.1639278680086136,
      "learning_rate": 7.303339773668231e-05,
      "loss": 0.718,
      "mean_token_accuracy": 0.7882935702800751,
      "num_tokens": 280686283.0,
      "step": 8794
    },
    {
      "entropy": 0.6602313974872231,
      "epoch": 0.8091960368255502,
      "grad_norm": 0.1640673130750656,
      "learning_rate": 7.303033091054069e-05,
      "loss": 0.6514,
      "mean_token_accuracy": 0.8017388619482517,
      "num_tokens": 280717815.0,
      "step": 8795
    },
    {
      "entropy": 0.6770196817815304,
      "epoch": 0.8092880431969914,
      "grad_norm": 0.15212877094745636,
      "learning_rate": 7.302726408439906e-05,
      "loss": 0.658,
      "mean_token_accuracy": 0.8074035309255123,
      "num_tokens": 280748805.0,
      "step": 8796
    },
    {
      "entropy": 0.6483152583241463,
      "epoch": 0.8093800495684326,
      "grad_norm": 0.15645915269851685,
      "learning_rate": 7.302419725825743e-05,
      "loss": 0.6275,
      "mean_token_accuracy": 0.8094264045357704,
      "num_tokens": 280779950.0,
      "step": 8797
    },
    {
      "entropy": 0.7575468113645911,
      "epoch": 0.8094720559398738,
      "grad_norm": 0.15636046230793,
      "learning_rate": 7.302113043211581e-05,
      "loss": 0.7385,
      "mean_token_accuracy": 0.7793776132166386,
      "num_tokens": 280811220.0,
      "step": 8798
    },
    {
      "entropy": 0.7843654099851847,
      "epoch": 0.8095640623113151,
      "grad_norm": 0.1499468982219696,
      "learning_rate": 7.301806360597419e-05,
      "loss": 0.7695,
      "mean_token_accuracy": 0.7757474184036255,
      "num_tokens": 280843696.0,
      "step": 8799
    },
    {
      "entropy": 0.8951494973152876,
      "epoch": 0.8096560686827563,
      "grad_norm": 0.161033034324646,
      "learning_rate": 7.301499677983256e-05,
      "loss": 0.8789,
      "mean_token_accuracy": 0.7480482794344425,
      "num_tokens": 280875576.0,
      "step": 8800
    },
    {
      "entropy": 0.7089257780462503,
      "epoch": 0.8097480750541975,
      "grad_norm": 0.14815685153007507,
      "learning_rate": 7.301192995369093e-05,
      "loss": 0.6875,
      "mean_token_accuracy": 0.7991449274122715,
      "num_tokens": 280907370.0,
      "step": 8801
    },
    {
      "entropy": 0.7315994389355183,
      "epoch": 0.8098400814256387,
      "grad_norm": 0.15197136998176575,
      "learning_rate": 7.30088631275493e-05,
      "loss": 0.7115,
      "mean_token_accuracy": 0.7908039651811123,
      "num_tokens": 280938918.0,
      "step": 8802
    },
    {
      "entropy": 0.7551451288163662,
      "epoch": 0.8099320877970799,
      "grad_norm": 0.16854926943778992,
      "learning_rate": 7.300579630140768e-05,
      "loss": 0.7533,
      "mean_token_accuracy": 0.7842795662581921,
      "num_tokens": 280971001.0,
      "step": 8803
    },
    {
      "entropy": 0.6859685089439154,
      "epoch": 0.8100240941685212,
      "grad_norm": 0.15417101979255676,
      "learning_rate": 7.300272947526606e-05,
      "loss": 0.6827,
      "mean_token_accuracy": 0.797606784850359,
      "num_tokens": 281003000.0,
      "step": 8804
    },
    {
      "entropy": 0.7523537967354059,
      "epoch": 0.8101161005399624,
      "grad_norm": 0.15922929346561432,
      "learning_rate": 7.299966264912443e-05,
      "loss": 0.7624,
      "mean_token_accuracy": 0.7778226025402546,
      "num_tokens": 281035075.0,
      "step": 8805
    },
    {
      "entropy": 0.8452366832643747,
      "epoch": 0.8102081069114037,
      "grad_norm": 0.16026504337787628,
      "learning_rate": 7.299659582298279e-05,
      "loss": 0.8599,
      "mean_token_accuracy": 0.7525277324020863,
      "num_tokens": 281066755.0,
      "step": 8806
    },
    {
      "entropy": 0.622067048214376,
      "epoch": 0.8103001132828448,
      "grad_norm": 0.1551104336977005,
      "learning_rate": 7.299352899684117e-05,
      "loss": 0.6246,
      "mean_token_accuracy": 0.8134521245956421,
      "num_tokens": 281099159.0,
      "step": 8807
    },
    {
      "entropy": 0.7403935818001628,
      "epoch": 0.810392119654286,
      "grad_norm": 0.1558847725391388,
      "learning_rate": 7.299046217069954e-05,
      "loss": 0.7175,
      "mean_token_accuracy": 0.7894258201122284,
      "num_tokens": 281130653.0,
      "step": 8808
    },
    {
      "entropy": 0.7819300182163715,
      "epoch": 0.8104841260257273,
      "grad_norm": 0.15648329257965088,
      "learning_rate": 7.298739534455792e-05,
      "loss": 0.78,
      "mean_token_accuracy": 0.7712905146181583,
      "num_tokens": 281162626.0,
      "step": 8809
    },
    {
      "entropy": 0.8214421309530735,
      "epoch": 0.8105761323971685,
      "grad_norm": 0.16528865694999695,
      "learning_rate": 7.298432851841629e-05,
      "loss": 0.8148,
      "mean_token_accuracy": 0.7626269571483135,
      "num_tokens": 281194707.0,
      "step": 8810
    },
    {
      "entropy": 0.699939239770174,
      "epoch": 0.8106681387686098,
      "grad_norm": 0.16284100711345673,
      "learning_rate": 7.298126169227467e-05,
      "loss": 0.6923,
      "mean_token_accuracy": 0.793315514922142,
      "num_tokens": 281227045.0,
      "step": 8811
    },
    {
      "entropy": 0.6121280323714018,
      "epoch": 0.8107601451400509,
      "grad_norm": 0.1594342589378357,
      "learning_rate": 7.297819486613304e-05,
      "loss": 0.6061,
      "mean_token_accuracy": 0.8195356205105782,
      "num_tokens": 281258076.0,
      "step": 8812
    },
    {
      "entropy": 0.7171043753623962,
      "epoch": 0.8108521515114921,
      "grad_norm": 0.1467992514371872,
      "learning_rate": 7.297512803999141e-05,
      "loss": 0.6985,
      "mean_token_accuracy": 0.7929666042327881,
      "num_tokens": 281290740.0,
      "step": 8813
    },
    {
      "entropy": 0.801898998208344,
      "epoch": 0.8109441578829334,
      "grad_norm": 0.16155773401260376,
      "learning_rate": 7.297206121384979e-05,
      "loss": 0.7811,
      "mean_token_accuracy": 0.7693518958985806,
      "num_tokens": 281322973.0,
      "step": 8814
    },
    {
      "entropy": 0.7128304913640022,
      "epoch": 0.8110361642543746,
      "grad_norm": 0.16628646850585938,
      "learning_rate": 7.296899438770817e-05,
      "loss": 0.7003,
      "mean_token_accuracy": 0.7896296568214893,
      "num_tokens": 281354153.0,
      "step": 8815
    },
    {
      "entropy": 0.7155962474644184,
      "epoch": 0.8111281706258159,
      "grad_norm": 0.15674036741256714,
      "learning_rate": 7.296592756156654e-05,
      "loss": 0.6922,
      "mean_token_accuracy": 0.7932563982903957,
      "num_tokens": 281385560.0,
      "step": 8816
    },
    {
      "entropy": 0.6640739282593131,
      "epoch": 0.811220176997257,
      "grad_norm": 0.15892334282398224,
      "learning_rate": 7.296286073542491e-05,
      "loss": 0.6397,
      "mean_token_accuracy": 0.8089822381734848,
      "num_tokens": 281417126.0,
      "step": 8817
    },
    {
      "entropy": 0.7285239677876234,
      "epoch": 0.8113121833686983,
      "grad_norm": 0.15154895186424255,
      "learning_rate": 7.295979390928329e-05,
      "loss": 0.6989,
      "mean_token_accuracy": 0.7905909679830074,
      "num_tokens": 281449499.0,
      "step": 8818
    },
    {
      "entropy": 0.6988730654120445,
      "epoch": 0.8114041897401395,
      "grad_norm": 0.16293446719646454,
      "learning_rate": 7.295672708314166e-05,
      "loss": 0.6938,
      "mean_token_accuracy": 0.8006265833973885,
      "num_tokens": 281481138.0,
      "step": 8819
    },
    {
      "entropy": 0.6785936690866947,
      "epoch": 0.8114961961115807,
      "grad_norm": 0.15084905922412872,
      "learning_rate": 7.295366025700004e-05,
      "loss": 0.6632,
      "mean_token_accuracy": 0.8048894926905632,
      "num_tokens": 281512548.0,
      "step": 8820
    },
    {
      "entropy": 0.904801931232214,
      "epoch": 0.811588202483022,
      "grad_norm": 0.1611488312482834,
      "learning_rate": 7.29505934308584e-05,
      "loss": 0.9216,
      "mean_token_accuracy": 0.7363728359341621,
      "num_tokens": 281542872.0,
      "step": 8821
    },
    {
      "entropy": 0.7331694019958377,
      "epoch": 0.8116802088544631,
      "grad_norm": 0.15519411861896515,
      "learning_rate": 7.294752660471679e-05,
      "loss": 0.7329,
      "mean_token_accuracy": 0.7804453186690807,
      "num_tokens": 281574563.0,
      "step": 8822
    },
    {
      "entropy": 0.7524434942752123,
      "epoch": 0.8117722152259044,
      "grad_norm": 0.1662304401397705,
      "learning_rate": 7.294445977857516e-05,
      "loss": 0.7508,
      "mean_token_accuracy": 0.7835780717432499,
      "num_tokens": 281605687.0,
      "step": 8823
    },
    {
      "entropy": 0.6548904003575444,
      "epoch": 0.8118642215973456,
      "grad_norm": 0.14741657674312592,
      "learning_rate": 7.294139295243352e-05,
      "loss": 0.6439,
      "mean_token_accuracy": 0.8046582750976086,
      "num_tokens": 281637997.0,
      "step": 8824
    },
    {
      "entropy": 0.6722556035965681,
      "epoch": 0.8119562279687869,
      "grad_norm": 0.1545913964509964,
      "learning_rate": 7.29383261262919e-05,
      "loss": 0.651,
      "mean_token_accuracy": 0.8040770292282104,
      "num_tokens": 281669512.0,
      "step": 8825
    },
    {
      "entropy": 0.7071272470057011,
      "epoch": 0.8120482343402281,
      "grad_norm": 0.15020997822284698,
      "learning_rate": 7.293525930015029e-05,
      "loss": 0.7098,
      "mean_token_accuracy": 0.7856381423771381,
      "num_tokens": 281701917.0,
      "step": 8826
    },
    {
      "entropy": 0.7113747503608465,
      "epoch": 0.8121402407116692,
      "grad_norm": 0.15359772741794586,
      "learning_rate": 7.293219247400865e-05,
      "loss": 0.7015,
      "mean_token_accuracy": 0.7938238866627216,
      "num_tokens": 281733629.0,
      "step": 8827
    },
    {
      "entropy": 0.7640140671283007,
      "epoch": 0.8122322470831105,
      "grad_norm": 0.1623331606388092,
      "learning_rate": 7.292912564786702e-05,
      "loss": 0.7643,
      "mean_token_accuracy": 0.774017907679081,
      "num_tokens": 281765070.0,
      "step": 8828
    },
    {
      "entropy": 0.7643362637609243,
      "epoch": 0.8123242534545517,
      "grad_norm": 0.1622413694858551,
      "learning_rate": 7.292605882172539e-05,
      "loss": 0.7616,
      "mean_token_accuracy": 0.7808681279420853,
      "num_tokens": 281795780.0,
      "step": 8829
    },
    {
      "entropy": 0.6669468283653259,
      "epoch": 0.812416259825993,
      "grad_norm": 0.15858986973762512,
      "learning_rate": 7.292299199558378e-05,
      "loss": 0.6692,
      "mean_token_accuracy": 0.8038977384567261,
      "num_tokens": 281827192.0,
      "step": 8830
    },
    {
      "entropy": 0.6859957072883844,
      "epoch": 0.8125082661974342,
      "grad_norm": 0.16934099793434143,
      "learning_rate": 7.291992516944215e-05,
      "loss": 0.6855,
      "mean_token_accuracy": 0.8003430031239986,
      "num_tokens": 281859505.0,
      "step": 8831
    },
    {
      "entropy": 0.7651117453351617,
      "epoch": 0.8126002725688753,
      "grad_norm": 0.15542635321617126,
      "learning_rate": 7.291685834330052e-05,
      "loss": 0.7392,
      "mean_token_accuracy": 0.7810560949146748,
      "num_tokens": 281891863.0,
      "step": 8832
    },
    {
      "entropy": 0.7125754971057177,
      "epoch": 0.8126922789403166,
      "grad_norm": 0.1505800485610962,
      "learning_rate": 7.291379151715889e-05,
      "loss": 0.702,
      "mean_token_accuracy": 0.7918311543762684,
      "num_tokens": 281923474.0,
      "step": 8833
    },
    {
      "entropy": 0.6695321649312973,
      "epoch": 0.8127842853117578,
      "grad_norm": 0.15358196198940277,
      "learning_rate": 7.291072469101727e-05,
      "loss": 0.6542,
      "mean_token_accuracy": 0.801561001688242,
      "num_tokens": 281955472.0,
      "step": 8834
    },
    {
      "entropy": 0.7819194737821817,
      "epoch": 0.8128762916831991,
      "grad_norm": 0.15350036323070526,
      "learning_rate": 7.290765786487565e-05,
      "loss": 0.7873,
      "mean_token_accuracy": 0.7748149819672108,
      "num_tokens": 281987309.0,
      "step": 8835
    },
    {
      "entropy": 0.8379936423152685,
      "epoch": 0.8129682980546403,
      "grad_norm": 0.1584431678056717,
      "learning_rate": 7.290459103873402e-05,
      "loss": 0.8213,
      "mean_token_accuracy": 0.7596557810902596,
      "num_tokens": 282018670.0,
      "step": 8836
    },
    {
      "entropy": 0.7872576862573624,
      "epoch": 0.8130603044260815,
      "grad_norm": 0.1558152139186859,
      "learning_rate": 7.290152421259239e-05,
      "loss": 0.7674,
      "mean_token_accuracy": 0.7778166681528091,
      "num_tokens": 282051007.0,
      "step": 8837
    },
    {
      "entropy": 0.8222018182277679,
      "epoch": 0.8131523107975227,
      "grad_norm": 0.16916054487228394,
      "learning_rate": 7.289845738645077e-05,
      "loss": 0.8544,
      "mean_token_accuracy": 0.7578879036009312,
      "num_tokens": 282082275.0,
      "step": 8838
    },
    {
      "entropy": 0.7269451608881354,
      "epoch": 0.813244317168964,
      "grad_norm": 0.14502111077308655,
      "learning_rate": 7.289539056030914e-05,
      "loss": 0.7085,
      "mean_token_accuracy": 0.7902436442673206,
      "num_tokens": 282114526.0,
      "step": 8839
    },
    {
      "entropy": 0.8404010888189077,
      "epoch": 0.8133363235404052,
      "grad_norm": 0.1541755646467209,
      "learning_rate": 7.289232373416752e-05,
      "loss": 0.8231,
      "mean_token_accuracy": 0.7614155113697052,
      "num_tokens": 282147011.0,
      "step": 8840
    },
    {
      "entropy": 0.6437424458563328,
      "epoch": 0.8134283299118464,
      "grad_norm": 0.15187010169029236,
      "learning_rate": 7.288925690802589e-05,
      "loss": 0.6295,
      "mean_token_accuracy": 0.812542337924242,
      "num_tokens": 282179263.0,
      "step": 8841
    },
    {
      "entropy": 0.6070180898532271,
      "epoch": 0.8135203362832876,
      "grad_norm": 0.1475210338830948,
      "learning_rate": 7.288619008188427e-05,
      "loss": 0.5942,
      "mean_token_accuracy": 0.8224943578243256,
      "num_tokens": 282211224.0,
      "step": 8842
    },
    {
      "entropy": 0.7110947258770466,
      "epoch": 0.8136123426547288,
      "grad_norm": 0.16115175187587738,
      "learning_rate": 7.288312325574263e-05,
      "loss": 0.7015,
      "mean_token_accuracy": 0.7943053729832172,
      "num_tokens": 282242767.0,
      "step": 8843
    },
    {
      "entropy": 0.678065774962306,
      "epoch": 0.8137043490261701,
      "grad_norm": 0.15735198557376862,
      "learning_rate": 7.2880056429601e-05,
      "loss": 0.6599,
      "mean_token_accuracy": 0.8014461509883404,
      "num_tokens": 282275012.0,
      "step": 8844
    },
    {
      "entropy": 0.8109607920050621,
      "epoch": 0.8137963553976113,
      "grad_norm": 0.15819770097732544,
      "learning_rate": 7.287698960345938e-05,
      "loss": 0.8024,
      "mean_token_accuracy": 0.7647063657641411,
      "num_tokens": 282306918.0,
      "step": 8845
    },
    {
      "entropy": 0.6647379416972399,
      "epoch": 0.8138883617690525,
      "grad_norm": 0.16598786413669586,
      "learning_rate": 7.287392277731777e-05,
      "loss": 0.6671,
      "mean_token_accuracy": 0.8026313185691833,
      "num_tokens": 282338782.0,
      "step": 8846
    },
    {
      "entropy": 0.758580032736063,
      "epoch": 0.8139803681404937,
      "grad_norm": 0.16530585289001465,
      "learning_rate": 7.287085595117613e-05,
      "loss": 0.7545,
      "mean_token_accuracy": 0.7776653543114662,
      "num_tokens": 282370983.0,
      "step": 8847
    },
    {
      "entropy": 0.7231135200709105,
      "epoch": 0.8140723745119349,
      "grad_norm": 0.1657288670539856,
      "learning_rate": 7.28677891250345e-05,
      "loss": 0.7079,
      "mean_token_accuracy": 0.7904610820114613,
      "num_tokens": 282402300.0,
      "step": 8848
    },
    {
      "entropy": 0.7315331194549799,
      "epoch": 0.8141643808833762,
      "grad_norm": 0.15555812418460846,
      "learning_rate": 7.286472229889288e-05,
      "loss": 0.7367,
      "mean_token_accuracy": 0.7794448807835579,
      "num_tokens": 282434154.0,
      "step": 8849
    },
    {
      "entropy": 0.6935278903692961,
      "epoch": 0.8142563872548174,
      "grad_norm": 0.1470392495393753,
      "learning_rate": 7.286165547275125e-05,
      "loss": 0.6689,
      "mean_token_accuracy": 0.8053352497518063,
      "num_tokens": 282466117.0,
      "step": 8850
    },
    {
      "entropy": 0.7131359148770571,
      "epoch": 0.8143483936262587,
      "grad_norm": 0.15581178665161133,
      "learning_rate": 7.285858864660963e-05,
      "loss": 0.6987,
      "mean_token_accuracy": 0.7938751801848412,
      "num_tokens": 282498885.0,
      "step": 8851
    },
    {
      "entropy": 0.7702209986746311,
      "epoch": 0.8144403999976998,
      "grad_norm": 0.16139881312847137,
      "learning_rate": 7.2855521820468e-05,
      "loss": 0.7574,
      "mean_token_accuracy": 0.7769039571285248,
      "num_tokens": 282530625.0,
      "step": 8852
    },
    {
      "entropy": 0.6232226341962814,
      "epoch": 0.814532406369141,
      "grad_norm": 0.14995503425598145,
      "learning_rate": 7.285245499432638e-05,
      "loss": 0.5903,
      "mean_token_accuracy": 0.8245407342910767,
      "num_tokens": 282562706.0,
      "step": 8853
    },
    {
      "entropy": 0.6644568182528019,
      "epoch": 0.8146244127405823,
      "grad_norm": 0.14936047792434692,
      "learning_rate": 7.284938816818475e-05,
      "loss": 0.6439,
      "mean_token_accuracy": 0.8056899718940258,
      "num_tokens": 282594792.0,
      "step": 8854
    },
    {
      "entropy": 0.7307546157389879,
      "epoch": 0.8147164191120235,
      "grad_norm": 0.15565499663352966,
      "learning_rate": 7.284632134204312e-05,
      "loss": 0.7231,
      "mean_token_accuracy": 0.7867359705269337,
      "num_tokens": 282626779.0,
      "step": 8855
    },
    {
      "entropy": 0.7716090036556125,
      "epoch": 0.8148084254834648,
      "grad_norm": 0.15901805460453033,
      "learning_rate": 7.28432545159015e-05,
      "loss": 0.7809,
      "mean_token_accuracy": 0.7755277641117573,
      "num_tokens": 282658859.0,
      "step": 8856
    },
    {
      "entropy": 0.6166004845872521,
      "epoch": 0.8149004318549059,
      "grad_norm": 0.16082599759101868,
      "learning_rate": 7.284018768975988e-05,
      "loss": 0.6174,
      "mean_token_accuracy": 0.8097791485488415,
      "num_tokens": 282690466.0,
      "step": 8857
    },
    {
      "entropy": 0.8147182650864124,
      "epoch": 0.8149924382263471,
      "grad_norm": 0.16122965514659882,
      "learning_rate": 7.283712086361825e-05,
      "loss": 0.8048,
      "mean_token_accuracy": 0.7634128518402576,
      "num_tokens": 282722569.0,
      "step": 8858
    },
    {
      "entropy": 0.731562914326787,
      "epoch": 0.8150844445977884,
      "grad_norm": 0.1655273288488388,
      "learning_rate": 7.283405403747662e-05,
      "loss": 0.7309,
      "mean_token_accuracy": 0.7874297760426998,
      "num_tokens": 282754199.0,
      "step": 8859
    },
    {
      "entropy": 0.7536404319107533,
      "epoch": 0.8151764509692296,
      "grad_norm": 0.15255899727344513,
      "learning_rate": 7.283098721133498e-05,
      "loss": 0.7516,
      "mean_token_accuracy": 0.7792046628892422,
      "num_tokens": 282786019.0,
      "step": 8860
    },
    {
      "entropy": 0.8362527154386044,
      "epoch": 0.8152684573406709,
      "grad_norm": 0.16592524945735931,
      "learning_rate": 7.282792038519336e-05,
      "loss": 0.8198,
      "mean_token_accuracy": 0.7620096392929554,
      "num_tokens": 282816859.0,
      "step": 8861
    },
    {
      "entropy": 0.6645166017115116,
      "epoch": 0.815360463712112,
      "grad_norm": 0.16417311131954193,
      "learning_rate": 7.282485355905175e-05,
      "loss": 0.6569,
      "mean_token_accuracy": 0.8073108308017254,
      "num_tokens": 282849142.0,
      "step": 8862
    },
    {
      "entropy": 0.6329911462962627,
      "epoch": 0.8154524700835533,
      "grad_norm": 0.15983767807483673,
      "learning_rate": 7.282178673291011e-05,
      "loss": 0.6213,
      "mean_token_accuracy": 0.8162010423839092,
      "num_tokens": 282881517.0,
      "step": 8863
    },
    {
      "entropy": 0.6390299266204238,
      "epoch": 0.8155444764549945,
      "grad_norm": 0.15358687937259674,
      "learning_rate": 7.281871990676848e-05,
      "loss": 0.6178,
      "mean_token_accuracy": 0.8206056840717793,
      "num_tokens": 282912432.0,
      "step": 8864
    },
    {
      "entropy": 0.6728829611092806,
      "epoch": 0.8156364828264357,
      "grad_norm": 0.1428249478340149,
      "learning_rate": 7.281565308062686e-05,
      "loss": 0.6426,
      "mean_token_accuracy": 0.8059559054672718,
      "num_tokens": 282944525.0,
      "step": 8865
    },
    {
      "entropy": 0.6589282006025314,
      "epoch": 0.815728489197877,
      "grad_norm": 0.19384068250656128,
      "learning_rate": 7.281258625448523e-05,
      "loss": 0.6248,
      "mean_token_accuracy": 0.8146298788487911,
      "num_tokens": 282976787.0,
      "step": 8866
    },
    {
      "entropy": 0.8172745518386364,
      "epoch": 0.8158204955693181,
      "grad_norm": 0.1610552817583084,
      "learning_rate": 7.280951942834361e-05,
      "loss": 0.8039,
      "mean_token_accuracy": 0.7623845748603344,
      "num_tokens": 283008220.0,
      "step": 8867
    },
    {
      "entropy": 0.7432925309985876,
      "epoch": 0.8159125019407594,
      "grad_norm": 0.15912580490112305,
      "learning_rate": 7.280645260220198e-05,
      "loss": 0.7302,
      "mean_token_accuracy": 0.7871149815618992,
      "num_tokens": 283040063.0,
      "step": 8868
    },
    {
      "entropy": 0.6837506256997585,
      "epoch": 0.8160045083122006,
      "grad_norm": 0.14766494929790497,
      "learning_rate": 7.280338577606036e-05,
      "loss": 0.6779,
      "mean_token_accuracy": 0.7994626276195049,
      "num_tokens": 283072831.0,
      "step": 8869
    },
    {
      "entropy": 0.7056467570364475,
      "epoch": 0.8160965146836419,
      "grad_norm": 0.16047316789627075,
      "learning_rate": 7.280031894991873e-05,
      "loss": 0.7022,
      "mean_token_accuracy": 0.7967587821185589,
      "num_tokens": 283105326.0,
      "step": 8870
    },
    {
      "entropy": 0.6693502105772495,
      "epoch": 0.8161885210550831,
      "grad_norm": 0.1468183994293213,
      "learning_rate": 7.27972521237771e-05,
      "loss": 0.6532,
      "mean_token_accuracy": 0.8070458211004734,
      "num_tokens": 283137491.0,
      "step": 8871
    },
    {
      "entropy": 0.6425915788859129,
      "epoch": 0.8162805274265242,
      "grad_norm": 0.15711542963981628,
      "learning_rate": 7.279418529763548e-05,
      "loss": 0.6211,
      "mean_token_accuracy": 0.8133961446583271,
      "num_tokens": 283169069.0,
      "step": 8872
    },
    {
      "entropy": 0.7631380688399076,
      "epoch": 0.8163725337979655,
      "grad_norm": 0.1499480903148651,
      "learning_rate": 7.279111847149386e-05,
      "loss": 0.746,
      "mean_token_accuracy": 0.7797861881554127,
      "num_tokens": 283200583.0,
      "step": 8873
    },
    {
      "entropy": 0.7319032046943903,
      "epoch": 0.8164645401694067,
      "grad_norm": 0.14943814277648926,
      "learning_rate": 7.278805164535223e-05,
      "loss": 0.7159,
      "mean_token_accuracy": 0.7835667803883553,
      "num_tokens": 283231683.0,
      "step": 8874
    },
    {
      "entropy": 0.7668934892863035,
      "epoch": 0.816556546540848,
      "grad_norm": 0.1626661717891693,
      "learning_rate": 7.27849848192106e-05,
      "loss": 0.7824,
      "mean_token_accuracy": 0.7682063542306423,
      "num_tokens": 283263539.0,
      "step": 8875
    },
    {
      "entropy": 0.7543725930154324,
      "epoch": 0.8166485529122892,
      "grad_norm": 0.15735207498073578,
      "learning_rate": 7.278191799306896e-05,
      "loss": 0.7517,
      "mean_token_accuracy": 0.7768127359449863,
      "num_tokens": 283295465.0,
      "step": 8876
    },
    {
      "entropy": 0.7828283980488777,
      "epoch": 0.8167405592837303,
      "grad_norm": 0.17547361552715302,
      "learning_rate": 7.277885116692736e-05,
      "loss": 0.7944,
      "mean_token_accuracy": 0.7684010416269302,
      "num_tokens": 283327845.0,
      "step": 8877
    },
    {
      "entropy": 0.6212687860243022,
      "epoch": 0.8168325656551716,
      "grad_norm": 0.15409885346889496,
      "learning_rate": 7.277578434078573e-05,
      "loss": 0.603,
      "mean_token_accuracy": 0.8212668597698212,
      "num_tokens": 283360352.0,
      "step": 8878
    },
    {
      "entropy": 0.707438513636589,
      "epoch": 0.8169245720266128,
      "grad_norm": 0.1521175503730774,
      "learning_rate": 7.27727175146441e-05,
      "loss": 0.6913,
      "mean_token_accuracy": 0.7997340001165867,
      "num_tokens": 283392354.0,
      "step": 8879
    },
    {
      "entropy": 0.6796687673777342,
      "epoch": 0.8170165783980541,
      "grad_norm": 0.15466517210006714,
      "learning_rate": 7.276965068850248e-05,
      "loss": 0.6733,
      "mean_token_accuracy": 0.7985201738774776,
      "num_tokens": 283424696.0,
      "step": 8880
    },
    {
      "entropy": 0.632179755717516,
      "epoch": 0.8171085847694953,
      "grad_norm": 0.1584043800830841,
      "learning_rate": 7.276658386236084e-05,
      "loss": 0.6279,
      "mean_token_accuracy": 0.8177241794764996,
      "num_tokens": 283456226.0,
      "step": 8881
    },
    {
      "entropy": 0.7687338851392269,
      "epoch": 0.8172005911409365,
      "grad_norm": 0.15326014161109924,
      "learning_rate": 7.276351703621923e-05,
      "loss": 0.7649,
      "mean_token_accuracy": 0.772969413548708,
      "num_tokens": 283488873.0,
      "step": 8882
    },
    {
      "entropy": 0.8191963210701942,
      "epoch": 0.8172925975123777,
      "grad_norm": 0.15639886260032654,
      "learning_rate": 7.27604502100776e-05,
      "loss": 0.8097,
      "mean_token_accuracy": 0.762598343193531,
      "num_tokens": 283519839.0,
      "step": 8883
    },
    {
      "entropy": 0.6770821567624807,
      "epoch": 0.817384603883819,
      "grad_norm": 0.16495661437511444,
      "learning_rate": 7.275738338393598e-05,
      "loss": 0.663,
      "mean_token_accuracy": 0.8018505983054638,
      "num_tokens": 283551198.0,
      "step": 8884
    },
    {
      "entropy": 0.7852525822818279,
      "epoch": 0.8174766102552602,
      "grad_norm": 0.15878231823444366,
      "learning_rate": 7.275431655779434e-05,
      "loss": 0.7688,
      "mean_token_accuracy": 0.7753232270479202,
      "num_tokens": 283583695.0,
      "step": 8885
    },
    {
      "entropy": 0.6230907812714577,
      "epoch": 0.8175686166267014,
      "grad_norm": 0.14271393418312073,
      "learning_rate": 7.275124973165271e-05,
      "loss": 0.6037,
      "mean_token_accuracy": 0.8190775848925114,
      "num_tokens": 283616167.0,
      "step": 8886
    },
    {
      "entropy": 0.6761398669332266,
      "epoch": 0.8176606229981426,
      "grad_norm": 0.1524570882320404,
      "learning_rate": 7.274818290551109e-05,
      "loss": 0.6614,
      "mean_token_accuracy": 0.8088189102709293,
      "num_tokens": 283648307.0,
      "step": 8887
    },
    {
      "entropy": 0.8470855671912432,
      "epoch": 0.8177526293695838,
      "grad_norm": 0.15934288501739502,
      "learning_rate": 7.274511607936947e-05,
      "loss": 0.8388,
      "mean_token_accuracy": 0.753144808113575,
      "num_tokens": 283679785.0,
      "step": 8888
    },
    {
      "entropy": 0.7899370566010475,
      "epoch": 0.8178446357410251,
      "grad_norm": 0.16054481267929077,
      "learning_rate": 7.274204925322784e-05,
      "loss": 0.7806,
      "mean_token_accuracy": 0.7681625857949257,
      "num_tokens": 283710872.0,
      "step": 8889
    },
    {
      "entropy": 0.6818925896659493,
      "epoch": 0.8179366421124663,
      "grad_norm": 0.15273325145244598,
      "learning_rate": 7.273898242708621e-05,
      "loss": 0.6567,
      "mean_token_accuracy": 0.8085002452135086,
      "num_tokens": 283743640.0,
      "step": 8890
    },
    {
      "entropy": 0.6839016936719418,
      "epoch": 0.8180286484839076,
      "grad_norm": 0.14513365924358368,
      "learning_rate": 7.273591560094458e-05,
      "loss": 0.6807,
      "mean_token_accuracy": 0.7950051240622997,
      "num_tokens": 283775178.0,
      "step": 8891
    },
    {
      "entropy": 0.6777231749147177,
      "epoch": 0.8181206548553487,
      "grad_norm": 0.15922227501869202,
      "learning_rate": 7.273284877480296e-05,
      "loss": 0.6891,
      "mean_token_accuracy": 0.7911290563642979,
      "num_tokens": 283806881.0,
      "step": 8892
    },
    {
      "entropy": 0.7743386402726173,
      "epoch": 0.8182126612267899,
      "grad_norm": 0.16230830550193787,
      "learning_rate": 7.272978194866134e-05,
      "loss": 0.7637,
      "mean_token_accuracy": 0.7732686400413513,
      "num_tokens": 283838376.0,
      "step": 8893
    },
    {
      "entropy": 0.7602397426962852,
      "epoch": 0.8183046675982312,
      "grad_norm": 0.16219723224639893,
      "learning_rate": 7.272671512251971e-05,
      "loss": 0.7601,
      "mean_token_accuracy": 0.7753823287785053,
      "num_tokens": 283870061.0,
      "step": 8894
    },
    {
      "entropy": 0.7483861483633518,
      "epoch": 0.8183966739696724,
      "grad_norm": 0.16454194486141205,
      "learning_rate": 7.272364829637808e-05,
      "loss": 0.7412,
      "mean_token_accuracy": 0.7803338058292866,
      "num_tokens": 283901540.0,
      "step": 8895
    },
    {
      "entropy": 0.6490075001493096,
      "epoch": 0.8184886803411137,
      "grad_norm": 0.15494303405284882,
      "learning_rate": 7.272058147023646e-05,
      "loss": 0.6433,
      "mean_token_accuracy": 0.8092268817126751,
      "num_tokens": 283934005.0,
      "step": 8896
    },
    {
      "entropy": 0.7823300156742334,
      "epoch": 0.8185806867125548,
      "grad_norm": 0.1656738966703415,
      "learning_rate": 7.271751464409483e-05,
      "loss": 0.7828,
      "mean_token_accuracy": 0.7707355432212353,
      "num_tokens": 283966036.0,
      "step": 8897
    },
    {
      "entropy": 0.7003092532977462,
      "epoch": 0.818672693083996,
      "grad_norm": 0.15831705927848816,
      "learning_rate": 7.27144478179532e-05,
      "loss": 0.6956,
      "mean_token_accuracy": 0.7982044890522957,
      "num_tokens": 283998372.0,
      "step": 8898
    },
    {
      "entropy": 0.6349435402080417,
      "epoch": 0.8187646994554373,
      "grad_norm": 0.15170037746429443,
      "learning_rate": 7.271138099181157e-05,
      "loss": 0.6156,
      "mean_token_accuracy": 0.8178951218724251,
      "num_tokens": 284030737.0,
      "step": 8899
    },
    {
      "entropy": 0.6046836357563734,
      "epoch": 0.8188567058268785,
      "grad_norm": 0.15465371310710907,
      "learning_rate": 7.270831416566996e-05,
      "loss": 0.5849,
      "mean_token_accuracy": 0.8256338685750961,
      "num_tokens": 284063099.0,
      "step": 8900
    },
    {
      "entropy": 0.8339173123240471,
      "epoch": 0.8189487121983198,
      "grad_norm": 0.1565534770488739,
      "learning_rate": 7.270524733952832e-05,
      "loss": 0.8298,
      "mean_token_accuracy": 0.7543990351259708,
      "num_tokens": 284094130.0,
      "step": 8901
    },
    {
      "entropy": 0.5927000986412168,
      "epoch": 0.8190407185697609,
      "grad_norm": 0.1431489735841751,
      "learning_rate": 7.270218051338669e-05,
      "loss": 0.5678,
      "mean_token_accuracy": 0.8250194899737835,
      "num_tokens": 284125871.0,
      "step": 8902
    },
    {
      "entropy": 0.6909135021269321,
      "epoch": 0.8191327249412022,
      "grad_norm": 0.15771518647670746,
      "learning_rate": 7.269911368724507e-05,
      "loss": 0.6752,
      "mean_token_accuracy": 0.8001469746232033,
      "num_tokens": 284156764.0,
      "step": 8903
    },
    {
      "entropy": 0.7788947112858295,
      "epoch": 0.8192247313126434,
      "grad_norm": 0.1580466628074646,
      "learning_rate": 7.269604686110345e-05,
      "loss": 0.7576,
      "mean_token_accuracy": 0.7775906845927238,
      "num_tokens": 284188732.0,
      "step": 8904
    },
    {
      "entropy": 0.8277196306735277,
      "epoch": 0.8193167376840846,
      "grad_norm": 0.15928709506988525,
      "learning_rate": 7.269298003496182e-05,
      "loss": 0.8189,
      "mean_token_accuracy": 0.7573447786271572,
      "num_tokens": 284220414.0,
      "step": 8905
    },
    {
      "entropy": 0.7262854222208261,
      "epoch": 0.8194087440555259,
      "grad_norm": 0.155595600605011,
      "learning_rate": 7.268991320882019e-05,
      "loss": 0.7249,
      "mean_token_accuracy": 0.7876815795898438,
      "num_tokens": 284252444.0,
      "step": 8906
    },
    {
      "entropy": 0.7615687288343906,
      "epoch": 0.819500750426967,
      "grad_norm": 0.15496665239334106,
      "learning_rate": 7.268684638267856e-05,
      "loss": 0.7622,
      "mean_token_accuracy": 0.7784506678581238,
      "num_tokens": 284284776.0,
      "step": 8907
    },
    {
      "entropy": 0.7765335142612457,
      "epoch": 0.8195927567984083,
      "grad_norm": 0.16071389615535736,
      "learning_rate": 7.268377955653694e-05,
      "loss": 0.7699,
      "mean_token_accuracy": 0.7732838615775108,
      "num_tokens": 284317059.0,
      "step": 8908
    },
    {
      "entropy": 0.8150903768837452,
      "epoch": 0.8196847631698495,
      "grad_norm": 0.15790818631649017,
      "learning_rate": 7.268071273039532e-05,
      "loss": 0.8074,
      "mean_token_accuracy": 0.7639236263930798,
      "num_tokens": 284349221.0,
      "step": 8909
    },
    {
      "entropy": 0.5948092974722385,
      "epoch": 0.8197767695412908,
      "grad_norm": 0.14612305164337158,
      "learning_rate": 7.267764590425369e-05,
      "loss": 0.5938,
      "mean_token_accuracy": 0.8201647475361824,
      "num_tokens": 284381463.0,
      "step": 8910
    },
    {
      "entropy": 0.6842238791286945,
      "epoch": 0.819868775912732,
      "grad_norm": 0.1586386263370514,
      "learning_rate": 7.267457907811207e-05,
      "loss": 0.67,
      "mean_token_accuracy": 0.8074273951351643,
      "num_tokens": 284413617.0,
      "step": 8911
    },
    {
      "entropy": 0.795402780175209,
      "epoch": 0.8199607822841731,
      "grad_norm": 0.15735557675361633,
      "learning_rate": 7.267151225197044e-05,
      "loss": 0.793,
      "mean_token_accuracy": 0.7653169073164463,
      "num_tokens": 284445509.0,
      "step": 8912
    },
    {
      "entropy": 0.7542642150074244,
      "epoch": 0.8200527886556144,
      "grad_norm": 0.16248662769794464,
      "learning_rate": 7.266844542582882e-05,
      "loss": 0.7362,
      "mean_token_accuracy": 0.7828256152570248,
      "num_tokens": 284477508.0,
      "step": 8913
    },
    {
      "entropy": 0.7218400090932846,
      "epoch": 0.8201447950270556,
      "grad_norm": 0.16699829697608948,
      "learning_rate": 7.266537859968719e-05,
      "loss": 0.7212,
      "mean_token_accuracy": 0.7859941199421883,
      "num_tokens": 284509304.0,
      "step": 8914
    },
    {
      "entropy": 0.5568763841874897,
      "epoch": 0.8202368013984969,
      "grad_norm": 0.1467861831188202,
      "learning_rate": 7.266231177354557e-05,
      "loss": 0.551,
      "mean_token_accuracy": 0.8323766440153122,
      "num_tokens": 284542072.0,
      "step": 8915
    },
    {
      "entropy": 0.727360425516963,
      "epoch": 0.8203288077699381,
      "grad_norm": 0.15941035747528076,
      "learning_rate": 7.265924494740394e-05,
      "loss": 0.7232,
      "mean_token_accuracy": 0.786419115960598,
      "num_tokens": 284574426.0,
      "step": 8916
    },
    {
      "entropy": 0.7960215043276548,
      "epoch": 0.8204208141413792,
      "grad_norm": 0.1623135805130005,
      "learning_rate": 7.26561781212623e-05,
      "loss": 0.7839,
      "mean_token_accuracy": 0.7684144712984562,
      "num_tokens": 284605889.0,
      "step": 8917
    },
    {
      "entropy": 0.7514284271746874,
      "epoch": 0.8205128205128205,
      "grad_norm": 0.16022086143493652,
      "learning_rate": 7.265311129512069e-05,
      "loss": 0.7532,
      "mean_token_accuracy": 0.7790559493005276,
      "num_tokens": 284638231.0,
      "step": 8918
    },
    {
      "entropy": 0.6776708979159594,
      "epoch": 0.8206048268842617,
      "grad_norm": 0.15381906926631927,
      "learning_rate": 7.265004446897907e-05,
      "loss": 0.6619,
      "mean_token_accuracy": 0.8008718490600586,
      "num_tokens": 284669805.0,
      "step": 8919
    },
    {
      "entropy": 0.6677124500274658,
      "epoch": 0.820696833255703,
      "grad_norm": 0.15749070048332214,
      "learning_rate": 7.264697764283744e-05,
      "loss": 0.6484,
      "mean_token_accuracy": 0.8037423938512802,
      "num_tokens": 284701582.0,
      "step": 8920
    },
    {
      "entropy": 0.8012014366686344,
      "epoch": 0.8207888396271442,
      "grad_norm": 0.157455712556839,
      "learning_rate": 7.26439108166958e-05,
      "loss": 0.788,
      "mean_token_accuracy": 0.7713336236774921,
      "num_tokens": 284733371.0,
      "step": 8921
    },
    {
      "entropy": 0.6651942543685436,
      "epoch": 0.8208808459985854,
      "grad_norm": 0.15206952393054962,
      "learning_rate": 7.264084399055417e-05,
      "loss": 0.6295,
      "mean_token_accuracy": 0.814245879650116,
      "num_tokens": 284764480.0,
      "step": 8922
    },
    {
      "entropy": 0.6840065103024244,
      "epoch": 0.8209728523700266,
      "grad_norm": 0.15396681427955627,
      "learning_rate": 7.263777716441255e-05,
      "loss": 0.669,
      "mean_token_accuracy": 0.8046112731099129,
      "num_tokens": 284796607.0,
      "step": 8923
    },
    {
      "entropy": 0.6036889012902975,
      "epoch": 0.8210648587414678,
      "grad_norm": 0.15107502043247223,
      "learning_rate": 7.263471033827093e-05,
      "loss": 0.5755,
      "mean_token_accuracy": 0.8231262862682343,
      "num_tokens": 284827436.0,
      "step": 8924
    },
    {
      "entropy": 0.7014011442661285,
      "epoch": 0.8211568651129091,
      "grad_norm": 0.1591929793357849,
      "learning_rate": 7.26316435121293e-05,
      "loss": 0.6805,
      "mean_token_accuracy": 0.8001134544610977,
      "num_tokens": 284857763.0,
      "step": 8925
    },
    {
      "entropy": 0.772796293720603,
      "epoch": 0.8212488714843503,
      "grad_norm": 0.16058947145938873,
      "learning_rate": 7.262857668598767e-05,
      "loss": 0.76,
      "mean_token_accuracy": 0.7762425802648067,
      "num_tokens": 284890099.0,
      "step": 8926
    },
    {
      "entropy": 0.7709118444472551,
      "epoch": 0.8213408778557915,
      "grad_norm": 0.16814643144607544,
      "learning_rate": 7.262550985984605e-05,
      "loss": 0.7643,
      "mean_token_accuracy": 0.779382660984993,
      "num_tokens": 284921596.0,
      "step": 8927
    },
    {
      "entropy": 0.7489400487393141,
      "epoch": 0.8214328842272327,
      "grad_norm": 0.1546616405248642,
      "learning_rate": 7.262244303370442e-05,
      "loss": 0.7133,
      "mean_token_accuracy": 0.7882344946265221,
      "num_tokens": 284952818.0,
      "step": 8928
    },
    {
      "entropy": 0.6782505381852388,
      "epoch": 0.821524890598674,
      "grad_norm": 0.16030468046665192,
      "learning_rate": 7.26193762075628e-05,
      "loss": 0.6813,
      "mean_token_accuracy": 0.7978207990527153,
      "num_tokens": 284985371.0,
      "step": 8929
    },
    {
      "entropy": 0.8155511673539877,
      "epoch": 0.8216168969701152,
      "grad_norm": 0.16628038883209229,
      "learning_rate": 7.261630938142117e-05,
      "loss": 0.8032,
      "mean_token_accuracy": 0.7669973969459534,
      "num_tokens": 285017775.0,
      "step": 8930
    },
    {
      "entropy": 0.7562508452683687,
      "epoch": 0.8217089033415564,
      "grad_norm": 0.15087632834911346,
      "learning_rate": 7.261324255527955e-05,
      "loss": 0.7538,
      "mean_token_accuracy": 0.7731324769556522,
      "num_tokens": 285049922.0,
      "step": 8931
    },
    {
      "entropy": 0.7298192102462053,
      "epoch": 0.8218009097129976,
      "grad_norm": 0.15326179563999176,
      "learning_rate": 7.261017572913792e-05,
      "loss": 0.7171,
      "mean_token_accuracy": 0.791311077773571,
      "num_tokens": 285081585.0,
      "step": 8932
    },
    {
      "entropy": 0.709159241989255,
      "epoch": 0.8218929160844388,
      "grad_norm": 0.14941658079624176,
      "learning_rate": 7.260710890299629e-05,
      "loss": 0.6976,
      "mean_token_accuracy": 0.7960095331072807,
      "num_tokens": 285113673.0,
      "step": 8933
    },
    {
      "entropy": 0.7204658314585686,
      "epoch": 0.8219849224558801,
      "grad_norm": 0.1611117422580719,
      "learning_rate": 7.260404207685467e-05,
      "loss": 0.7157,
      "mean_token_accuracy": 0.7881926000118256,
      "num_tokens": 285145820.0,
      "step": 8934
    },
    {
      "entropy": 0.6702667865902185,
      "epoch": 0.8220769288273213,
      "grad_norm": 0.15612970292568207,
      "learning_rate": 7.260097525071305e-05,
      "loss": 0.6556,
      "mean_token_accuracy": 0.8039199933409691,
      "num_tokens": 285176903.0,
      "step": 8935
    },
    {
      "entropy": 0.8001235947012901,
      "epoch": 0.8221689351987626,
      "grad_norm": 0.16293250024318695,
      "learning_rate": 7.259790842457142e-05,
      "loss": 0.7823,
      "mean_token_accuracy": 0.7663587965071201,
      "num_tokens": 285208319.0,
      "step": 8936
    },
    {
      "entropy": 0.8077472299337387,
      "epoch": 0.8222609415702037,
      "grad_norm": 0.1669590175151825,
      "learning_rate": 7.259484159842978e-05,
      "loss": 0.7978,
      "mean_token_accuracy": 0.7628472745418549,
      "num_tokens": 285239371.0,
      "step": 8937
    },
    {
      "entropy": 0.6016338150948286,
      "epoch": 0.8223529479416449,
      "grad_norm": 0.14659234881401062,
      "learning_rate": 7.259177477228815e-05,
      "loss": 0.5841,
      "mean_token_accuracy": 0.8247440494596958,
      "num_tokens": 285271497.0,
      "step": 8938
    },
    {
      "entropy": 0.7416516542434692,
      "epoch": 0.8224449543130862,
      "grad_norm": 0.15762391686439514,
      "learning_rate": 7.258870794614653e-05,
      "loss": 0.7424,
      "mean_token_accuracy": 0.7811893559992313,
      "num_tokens": 285303552.0,
      "step": 8939
    },
    {
      "entropy": 0.6924250656738877,
      "epoch": 0.8225369606845274,
      "grad_norm": 0.1547284871339798,
      "learning_rate": 7.258564112000491e-05,
      "loss": 0.6934,
      "mean_token_accuracy": 0.7937106117606163,
      "num_tokens": 285336042.0,
      "step": 8940
    },
    {
      "entropy": 0.8373758736997843,
      "epoch": 0.8226289670559687,
      "grad_norm": 0.1577252596616745,
      "learning_rate": 7.258257429386328e-05,
      "loss": 0.826,
      "mean_token_accuracy": 0.7588753290474415,
      "num_tokens": 285367816.0,
      "step": 8941
    },
    {
      "entropy": 0.6646565785631537,
      "epoch": 0.8227209734274098,
      "grad_norm": 0.15737617015838623,
      "learning_rate": 7.257950746772166e-05,
      "loss": 0.6371,
      "mean_token_accuracy": 0.8029040358960629,
      "num_tokens": 285399229.0,
      "step": 8942
    },
    {
      "entropy": 0.6973677277565002,
      "epoch": 0.822812979798851,
      "grad_norm": 0.16932177543640137,
      "learning_rate": 7.257644064158003e-05,
      "loss": 0.6869,
      "mean_token_accuracy": 0.7948487438261509,
      "num_tokens": 285430940.0,
      "step": 8943
    },
    {
      "entropy": 0.6323857214301825,
      "epoch": 0.8229049861702923,
      "grad_norm": 0.15756022930145264,
      "learning_rate": 7.25733738154384e-05,
      "loss": 0.6271,
      "mean_token_accuracy": 0.8107408173382282,
      "num_tokens": 285463640.0,
      "step": 8944
    },
    {
      "entropy": 0.727318050339818,
      "epoch": 0.8229969925417335,
      "grad_norm": 0.15722672641277313,
      "learning_rate": 7.257030698929678e-05,
      "loss": 0.7159,
      "mean_token_accuracy": 0.7904860787093639,
      "num_tokens": 285496408.0,
      "step": 8945
    },
    {
      "entropy": 0.6942963246256113,
      "epoch": 0.8230889989131748,
      "grad_norm": 0.15469060838222504,
      "learning_rate": 7.256724016315516e-05,
      "loss": 0.6769,
      "mean_token_accuracy": 0.795751940459013,
      "num_tokens": 285527420.0,
      "step": 8946
    },
    {
      "entropy": 0.797913920134306,
      "epoch": 0.8231810052846159,
      "grad_norm": 0.154256671667099,
      "learning_rate": 7.256417333701353e-05,
      "loss": 0.7867,
      "mean_token_accuracy": 0.7687107436358929,
      "num_tokens": 285559426.0,
      "step": 8947
    },
    {
      "entropy": 0.6184217445552349,
      "epoch": 0.8232730116560572,
      "grad_norm": 0.15080685913562775,
      "learning_rate": 7.25611065108719e-05,
      "loss": 0.5773,
      "mean_token_accuracy": 0.8293310329318047,
      "num_tokens": 285591222.0,
      "step": 8948
    },
    {
      "entropy": 0.7895031571388245,
      "epoch": 0.8233650180274984,
      "grad_norm": 0.15076005458831787,
      "learning_rate": 7.255803968473027e-05,
      "loss": 0.7894,
      "mean_token_accuracy": 0.7685685604810715,
      "num_tokens": 285623439.0,
      "step": 8949
    },
    {
      "entropy": 0.7949450332671404,
      "epoch": 0.8234570243989396,
      "grad_norm": 0.15249180793762207,
      "learning_rate": 7.255497285858866e-05,
      "loss": 0.7858,
      "mean_token_accuracy": 0.7663241326808929,
      "num_tokens": 285654721.0,
      "step": 8950
    },
    {
      "entropy": 0.7760120406746864,
      "epoch": 0.8235490307703809,
      "grad_norm": 0.16734275221824646,
      "learning_rate": 7.255190603244703e-05,
      "loss": 0.776,
      "mean_token_accuracy": 0.7697091773152351,
      "num_tokens": 285686503.0,
      "step": 8951
    },
    {
      "entropy": 0.6996889263391495,
      "epoch": 0.823641037141822,
      "grad_norm": 0.16084517538547516,
      "learning_rate": 7.25488392063054e-05,
      "loss": 0.6933,
      "mean_token_accuracy": 0.7947257719933987,
      "num_tokens": 285717840.0,
      "step": 8952
    },
    {
      "entropy": 0.6514554033055902,
      "epoch": 0.8237330435132633,
      "grad_norm": 0.14601847529411316,
      "learning_rate": 7.254577238016376e-05,
      "loss": 0.6285,
      "mean_token_accuracy": 0.8144664205610752,
      "num_tokens": 285750133.0,
      "step": 8953
    },
    {
      "entropy": 0.6074075791984797,
      "epoch": 0.8238250498847045,
      "grad_norm": 0.1509881168603897,
      "learning_rate": 7.254270555402215e-05,
      "loss": 0.574,
      "mean_token_accuracy": 0.8255472555756569,
      "num_tokens": 285782005.0,
      "step": 8954
    },
    {
      "entropy": 0.7604123763740063,
      "epoch": 0.8239170562561458,
      "grad_norm": 0.15656660497188568,
      "learning_rate": 7.253963872788053e-05,
      "loss": 0.753,
      "mean_token_accuracy": 0.7780224904417992,
      "num_tokens": 285811894.0,
      "step": 8955
    },
    {
      "entropy": 0.6791110895574093,
      "epoch": 0.824009062627587,
      "grad_norm": 0.1494625061750412,
      "learning_rate": 7.25365719017389e-05,
      "loss": 0.6759,
      "mean_token_accuracy": 0.800452470779419,
      "num_tokens": 285844652.0,
      "step": 8956
    },
    {
      "entropy": 0.708913279697299,
      "epoch": 0.8241010689990281,
      "grad_norm": 0.159681037068367,
      "learning_rate": 7.253350507559726e-05,
      "loss": 0.7018,
      "mean_token_accuracy": 0.7922932095825672,
      "num_tokens": 285877028.0,
      "step": 8957
    },
    {
      "entropy": 0.8011132031679153,
      "epoch": 0.8241930753704694,
      "grad_norm": 0.16026878356933594,
      "learning_rate": 7.253043824945564e-05,
      "loss": 0.8154,
      "mean_token_accuracy": 0.7628692090511322,
      "num_tokens": 285908947.0,
      "step": 8958
    },
    {
      "entropy": 0.7142870854586363,
      "epoch": 0.8242850817419106,
      "grad_norm": 0.15641997754573822,
      "learning_rate": 7.252737142331401e-05,
      "loss": 0.7215,
      "mean_token_accuracy": 0.7857906110584736,
      "num_tokens": 285940825.0,
      "step": 8959
    },
    {
      "entropy": 0.7827363703399897,
      "epoch": 0.8243770881133519,
      "grad_norm": 0.16898083686828613,
      "learning_rate": 7.25243045971724e-05,
      "loss": 0.776,
      "mean_token_accuracy": 0.7749377563595772,
      "num_tokens": 285972864.0,
      "step": 8960
    },
    {
      "entropy": 0.7203196417540312,
      "epoch": 0.8244690944847931,
      "grad_norm": 0.150071382522583,
      "learning_rate": 7.252123777103076e-05,
      "loss": 0.706,
      "mean_token_accuracy": 0.7900875471532345,
      "num_tokens": 286005278.0,
      "step": 8961
    },
    {
      "entropy": 0.7627804595977068,
      "epoch": 0.8245611008562342,
      "grad_norm": 0.15748925507068634,
      "learning_rate": 7.251817094488914e-05,
      "loss": 0.7599,
      "mean_token_accuracy": 0.7751562520861626,
      "num_tokens": 286036762.0,
      "step": 8962
    },
    {
      "entropy": 0.7508614212274551,
      "epoch": 0.8246531072276755,
      "grad_norm": 0.15928609669208527,
      "learning_rate": 7.251510411874751e-05,
      "loss": 0.7457,
      "mean_token_accuracy": 0.7802472375333309,
      "num_tokens": 286068398.0,
      "step": 8963
    },
    {
      "entropy": 0.8576329164206982,
      "epoch": 0.8247451135991167,
      "grad_norm": 0.16591401398181915,
      "learning_rate": 7.251203729260588e-05,
      "loss": 0.8672,
      "mean_token_accuracy": 0.7490975223481655,
      "num_tokens": 286100586.0,
      "step": 8964
    },
    {
      "entropy": 0.6751609351485968,
      "epoch": 0.824837119970558,
      "grad_norm": 0.15011581778526306,
      "learning_rate": 7.250897046646426e-05,
      "loss": 0.6525,
      "mean_token_accuracy": 0.8008700162172318,
      "num_tokens": 286132187.0,
      "step": 8965
    },
    {
      "entropy": 0.7018976230174303,
      "epoch": 0.8249291263419992,
      "grad_norm": 0.1536310911178589,
      "learning_rate": 7.250590364032264e-05,
      "loss": 0.6859,
      "mean_token_accuracy": 0.8027660734951496,
      "num_tokens": 286164451.0,
      "step": 8966
    },
    {
      "entropy": 0.6982177179306746,
      "epoch": 0.8250211327134404,
      "grad_norm": 0.15565235912799835,
      "learning_rate": 7.250283681418101e-05,
      "loss": 0.685,
      "mean_token_accuracy": 0.7973880246281624,
      "num_tokens": 286196394.0,
      "step": 8967
    },
    {
      "entropy": 0.6583628598600626,
      "epoch": 0.8251131390848816,
      "grad_norm": 0.14899961650371552,
      "learning_rate": 7.249976998803938e-05,
      "loss": 0.6433,
      "mean_token_accuracy": 0.8089684806764126,
      "num_tokens": 286228722.0,
      "step": 8968
    },
    {
      "entropy": 0.7483910918235779,
      "epoch": 0.8252051454563228,
      "grad_norm": 0.1580926477909088,
      "learning_rate": 7.249670316189776e-05,
      "loss": 0.7257,
      "mean_token_accuracy": 0.783611886203289,
      "num_tokens": 286260455.0,
      "step": 8969
    },
    {
      "entropy": 0.6762728448957205,
      "epoch": 0.8252971518277641,
      "grad_norm": 0.1476806253194809,
      "learning_rate": 7.249363633575613e-05,
      "loss": 0.6494,
      "mean_token_accuracy": 0.8038259074091911,
      "num_tokens": 286292402.0,
      "step": 8970
    },
    {
      "entropy": 0.7565533183515072,
      "epoch": 0.8253891581992053,
      "grad_norm": 0.15069164335727692,
      "learning_rate": 7.249056950961451e-05,
      "loss": 0.7405,
      "mean_token_accuracy": 0.7831461913883686,
      "num_tokens": 286324378.0,
      "step": 8971
    },
    {
      "entropy": 0.7935268431901932,
      "epoch": 0.8254811645706465,
      "grad_norm": 0.16202698647975922,
      "learning_rate": 7.248750268347288e-05,
      "loss": 0.7871,
      "mean_token_accuracy": 0.7696957215666771,
      "num_tokens": 286356856.0,
      "step": 8972
    },
    {
      "entropy": 0.7028046706691384,
      "epoch": 0.8255731709420877,
      "grad_norm": 0.1535670906305313,
      "learning_rate": 7.248443585733126e-05,
      "loss": 0.6757,
      "mean_token_accuracy": 0.8015553168952465,
      "num_tokens": 286388603.0,
      "step": 8973
    },
    {
      "entropy": 0.73134471103549,
      "epoch": 0.825665177313529,
      "grad_norm": 0.153033047914505,
      "learning_rate": 7.248136903118963e-05,
      "loss": 0.7026,
      "mean_token_accuracy": 0.7926597632467747,
      "num_tokens": 286420444.0,
      "step": 8974
    },
    {
      "entropy": 0.6885848473757505,
      "epoch": 0.8257571836849702,
      "grad_norm": 0.14651906490325928,
      "learning_rate": 7.2478302205048e-05,
      "loss": 0.6767,
      "mean_token_accuracy": 0.7995542213320732,
      "num_tokens": 286453212.0,
      "step": 8975
    },
    {
      "entropy": 0.6635593939572573,
      "epoch": 0.8258491900564114,
      "grad_norm": 0.15837140381336212,
      "learning_rate": 7.247523537890637e-05,
      "loss": 0.6388,
      "mean_token_accuracy": 0.8086057156324387,
      "num_tokens": 286484692.0,
      "step": 8976
    },
    {
      "entropy": 0.7110080998390913,
      "epoch": 0.8259411964278526,
      "grad_norm": 0.15549886226654053,
      "learning_rate": 7.247216855276476e-05,
      "loss": 0.6728,
      "mean_token_accuracy": 0.8004459366202354,
      "num_tokens": 286515413.0,
      "step": 8977
    },
    {
      "entropy": 0.674189105629921,
      "epoch": 0.8260332027992938,
      "grad_norm": 0.16654551029205322,
      "learning_rate": 7.246910172662312e-05,
      "loss": 0.6683,
      "mean_token_accuracy": 0.7973058447241783,
      "num_tokens": 286547442.0,
      "step": 8978
    },
    {
      "entropy": 0.6516083627939224,
      "epoch": 0.8261252091707351,
      "grad_norm": 0.16132307052612305,
      "learning_rate": 7.246603490048149e-05,
      "loss": 0.6406,
      "mean_token_accuracy": 0.8065245486795902,
      "num_tokens": 286579295.0,
      "step": 8979
    },
    {
      "entropy": 0.6889814017340541,
      "epoch": 0.8262172155421763,
      "grad_norm": 0.16747808456420898,
      "learning_rate": 7.246296807433986e-05,
      "loss": 0.6819,
      "mean_token_accuracy": 0.7993496395647526,
      "num_tokens": 286611723.0,
      "step": 8980
    },
    {
      "entropy": 0.8625480607151985,
      "epoch": 0.8263092219136176,
      "grad_norm": 0.166040301322937,
      "learning_rate": 7.245990124819824e-05,
      "loss": 0.872,
      "mean_token_accuracy": 0.7457972131669521,
      "num_tokens": 286643020.0,
      "step": 8981
    },
    {
      "entropy": 0.7233224231749773,
      "epoch": 0.8264012282850587,
      "grad_norm": 0.16105417907238007,
      "learning_rate": 7.245683442205662e-05,
      "loss": 0.7018,
      "mean_token_accuracy": 0.7930994182825089,
      "num_tokens": 286675159.0,
      "step": 8982
    },
    {
      "entropy": 0.7569531016051769,
      "epoch": 0.8264932346564999,
      "grad_norm": 0.15916679799556732,
      "learning_rate": 7.245376759591499e-05,
      "loss": 0.747,
      "mean_token_accuracy": 0.7799199931323528,
      "num_tokens": 286707059.0,
      "step": 8983
    },
    {
      "entropy": 0.8399435617029667,
      "epoch": 0.8265852410279412,
      "grad_norm": 0.16573628783226013,
      "learning_rate": 7.245070076977336e-05,
      "loss": 0.8544,
      "mean_token_accuracy": 0.7536828257143497,
      "num_tokens": 286738657.0,
      "step": 8984
    },
    {
      "entropy": 0.6747440714389086,
      "epoch": 0.8266772473993824,
      "grad_norm": 0.14598838984966278,
      "learning_rate": 7.244763394363174e-05,
      "loss": 0.6611,
      "mean_token_accuracy": 0.802978154271841,
      "num_tokens": 286770884.0,
      "step": 8985
    },
    {
      "entropy": 0.7300641424953938,
      "epoch": 0.8267692537708237,
      "grad_norm": 0.16483032703399658,
      "learning_rate": 7.244456711749011e-05,
      "loss": 0.715,
      "mean_token_accuracy": 0.7867937013506889,
      "num_tokens": 286802568.0,
      "step": 8986
    },
    {
      "entropy": 0.6254475926980376,
      "epoch": 0.8268612601422648,
      "grad_norm": 0.15986885130405426,
      "learning_rate": 7.244150029134849e-05,
      "loss": 0.6174,
      "mean_token_accuracy": 0.8146114684641361,
      "num_tokens": 286834977.0,
      "step": 8987
    },
    {
      "entropy": 0.7209993782453239,
      "epoch": 0.826953266513706,
      "grad_norm": 0.15098638832569122,
      "learning_rate": 7.243843346520686e-05,
      "loss": 0.6954,
      "mean_token_accuracy": 0.7921662516891956,
      "num_tokens": 286867672.0,
      "step": 8988
    },
    {
      "entropy": 0.6096637770533562,
      "epoch": 0.8270452728851473,
      "grad_norm": 0.16232706606388092,
      "learning_rate": 7.243536663906524e-05,
      "loss": 0.6009,
      "mean_token_accuracy": 0.8198185078799725,
      "num_tokens": 286899597.0,
      "step": 8989
    },
    {
      "entropy": 0.6847195085138083,
      "epoch": 0.8271372792565885,
      "grad_norm": 0.15189285576343536,
      "learning_rate": 7.24322998129236e-05,
      "loss": 0.6854,
      "mean_token_accuracy": 0.7910325825214386,
      "num_tokens": 286931363.0,
      "step": 8990
    },
    {
      "entropy": 0.689436299726367,
      "epoch": 0.8272292856280298,
      "grad_norm": 0.15207020938396454,
      "learning_rate": 7.242923298678197e-05,
      "loss": 0.6807,
      "mean_token_accuracy": 0.800985861569643,
      "num_tokens": 286963686.0,
      "step": 8991
    },
    {
      "entropy": 0.5709239523857832,
      "epoch": 0.8273212919994709,
      "grad_norm": 0.15486657619476318,
      "learning_rate": 7.242616616064036e-05,
      "loss": 0.5578,
      "mean_token_accuracy": 0.8337900452315807,
      "num_tokens": 286996389.0,
      "step": 8992
    },
    {
      "entropy": 0.8054125048220158,
      "epoch": 0.8274132983709122,
      "grad_norm": 0.1560627669095993,
      "learning_rate": 7.242309933449874e-05,
      "loss": 0.8023,
      "mean_token_accuracy": 0.7602116726338863,
      "num_tokens": 287028265.0,
      "step": 8993
    },
    {
      "entropy": 0.7632589526474476,
      "epoch": 0.8275053047423534,
      "grad_norm": 0.16058342158794403,
      "learning_rate": 7.24200325083571e-05,
      "loss": 0.7617,
      "mean_token_accuracy": 0.7748209238052368,
      "num_tokens": 287060666.0,
      "step": 8994
    },
    {
      "entropy": 0.6896180994808674,
      "epoch": 0.8275973111137946,
      "grad_norm": 0.14826738834381104,
      "learning_rate": 7.241696568221547e-05,
      "loss": 0.677,
      "mean_token_accuracy": 0.7938706204295158,
      "num_tokens": 287092071.0,
      "step": 8995
    },
    {
      "entropy": 0.7506554163992405,
      "epoch": 0.8276893174852359,
      "grad_norm": 0.16319486498832703,
      "learning_rate": 7.241389885607384e-05,
      "loss": 0.726,
      "mean_token_accuracy": 0.782239269465208,
      "num_tokens": 287123049.0,
      "step": 8996
    },
    {
      "entropy": 0.7898458410054445,
      "epoch": 0.827781323856677,
      "grad_norm": 0.16183070838451385,
      "learning_rate": 7.241083202993224e-05,
      "loss": 0.785,
      "mean_token_accuracy": 0.7679465562105179,
      "num_tokens": 287154123.0,
      "step": 8997
    },
    {
      "entropy": 0.6768246106803417,
      "epoch": 0.8278733302281183,
      "grad_norm": 0.15750867128372192,
      "learning_rate": 7.24077652037906e-05,
      "loss": 0.6541,
      "mean_token_accuracy": 0.8055593781173229,
      "num_tokens": 287186186.0,
      "step": 8998
    },
    {
      "entropy": 0.8642590716481209,
      "epoch": 0.8279653365995595,
      "grad_norm": 0.15480400621891022,
      "learning_rate": 7.240469837764897e-05,
      "loss": 0.8455,
      "mean_token_accuracy": 0.7585528083145618,
      "num_tokens": 287218538.0,
      "step": 8999
    },
    {
      "entropy": 0.7152856905013323,
      "epoch": 0.8280573429710008,
      "grad_norm": 0.15999771654605865,
      "learning_rate": 7.240163155150735e-05,
      "loss": 0.6866,
      "mean_token_accuracy": 0.8001035042107105,
      "num_tokens": 287249962.0,
      "step": 9000
    },
    {
      "entropy": 0.761523213237524,
      "epoch": 0.828149349342442,
      "grad_norm": 0.15745222568511963,
      "learning_rate": 7.239856472536572e-05,
      "loss": 0.748,
      "mean_token_accuracy": 0.7786234319210052,
      "num_tokens": 287281880.0,
      "step": 9001
    },
    {
      "entropy": 0.708950687199831,
      "epoch": 0.8282413557138831,
      "grad_norm": 0.15754638612270355,
      "learning_rate": 7.23954978992241e-05,
      "loss": 0.6758,
      "mean_token_accuracy": 0.8008165247738361,
      "num_tokens": 287313665.0,
      "step": 9002
    },
    {
      "entropy": 0.7151319161057472,
      "epoch": 0.8283333620853244,
      "grad_norm": 0.16598650813102722,
      "learning_rate": 7.239243107308247e-05,
      "loss": 0.6825,
      "mean_token_accuracy": 0.7917914390563965,
      "num_tokens": 287345145.0,
      "step": 9003
    },
    {
      "entropy": 0.6185675030574203,
      "epoch": 0.8284253684567656,
      "grad_norm": 0.15404513478279114,
      "learning_rate": 7.238936424694085e-05,
      "loss": 0.6059,
      "mean_token_accuracy": 0.8185132965445518,
      "num_tokens": 287377088.0,
      "step": 9004
    },
    {
      "entropy": 0.7598221022635698,
      "epoch": 0.8285173748282069,
      "grad_norm": 0.15361785888671875,
      "learning_rate": 7.238629742079922e-05,
      "loss": 0.7444,
      "mean_token_accuracy": 0.7812954969704151,
      "num_tokens": 287409339.0,
      "step": 9005
    },
    {
      "entropy": 0.7528890855610371,
      "epoch": 0.8286093811996481,
      "grad_norm": 0.16694988310337067,
      "learning_rate": 7.238323059465759e-05,
      "loss": 0.7562,
      "mean_token_accuracy": 0.7786045223474503,
      "num_tokens": 287440897.0,
      "step": 9006
    },
    {
      "entropy": 0.7257866002619267,
      "epoch": 0.8287013875710892,
      "grad_norm": 0.15898717939853668,
      "learning_rate": 7.238016376851597e-05,
      "loss": 0.7266,
      "mean_token_accuracy": 0.7842355668544769,
      "num_tokens": 287472597.0,
      "step": 9007
    },
    {
      "entropy": 0.6250223647803068,
      "epoch": 0.8287933939425305,
      "grad_norm": 0.14458304643630981,
      "learning_rate": 7.237709694237435e-05,
      "loss": 0.6133,
      "mean_token_accuracy": 0.8166712820529938,
      "num_tokens": 287504918.0,
      "step": 9008
    },
    {
      "entropy": 0.8765357509255409,
      "epoch": 0.8288854003139717,
      "grad_norm": 0.16685186326503754,
      "learning_rate": 7.237403011623272e-05,
      "loss": 0.8893,
      "mean_token_accuracy": 0.7396270148456097,
      "num_tokens": 287535928.0,
      "step": 9009
    },
    {
      "entropy": 0.7099541109055281,
      "epoch": 0.828977406685413,
      "grad_norm": 0.15405835211277008,
      "learning_rate": 7.237096329009109e-05,
      "loss": 0.7122,
      "mean_token_accuracy": 0.7868079245090485,
      "num_tokens": 287568198.0,
      "step": 9010
    },
    {
      "entropy": 0.7425329498946667,
      "epoch": 0.8290694130568542,
      "grad_norm": 0.1563442200422287,
      "learning_rate": 7.236789646394945e-05,
      "loss": 0.7287,
      "mean_token_accuracy": 0.783229898661375,
      "num_tokens": 287600229.0,
      "step": 9011
    },
    {
      "entropy": 0.6890686508268118,
      "epoch": 0.8291614194282954,
      "grad_norm": 0.15635515749454498,
      "learning_rate": 7.236482963780784e-05,
      "loss": 0.6665,
      "mean_token_accuracy": 0.7989141792058945,
      "num_tokens": 287631852.0,
      "step": 9012
    },
    {
      "entropy": 0.7994140386581421,
      "epoch": 0.8292534257997366,
      "grad_norm": 0.1601131409406662,
      "learning_rate": 7.236176281166622e-05,
      "loss": 0.7844,
      "mean_token_accuracy": 0.769573662430048,
      "num_tokens": 287663757.0,
      "step": 9013
    },
    {
      "entropy": 0.7297971583902836,
      "epoch": 0.8293454321711778,
      "grad_norm": 0.14865612983703613,
      "learning_rate": 7.235869598552458e-05,
      "loss": 0.7207,
      "mean_token_accuracy": 0.7857004143297672,
      "num_tokens": 287695144.0,
      "step": 9014
    },
    {
      "entropy": 0.7098129745572805,
      "epoch": 0.8294374385426191,
      "grad_norm": 0.1574014574289322,
      "learning_rate": 7.235562915938295e-05,
      "loss": 0.6903,
      "mean_token_accuracy": 0.7947061210870743,
      "num_tokens": 287727332.0,
      "step": 9015
    },
    {
      "entropy": 0.7665156293660402,
      "epoch": 0.8295294449140603,
      "grad_norm": 0.16708172857761383,
      "learning_rate": 7.235256233324133e-05,
      "loss": 0.7575,
      "mean_token_accuracy": 0.77857830747962,
      "num_tokens": 287759056.0,
      "step": 9016
    },
    {
      "entropy": 0.6984675414860249,
      "epoch": 0.8296214512855015,
      "grad_norm": 0.1575230360031128,
      "learning_rate": 7.23494955070997e-05,
      "loss": 0.6886,
      "mean_token_accuracy": 0.7969398684799671,
      "num_tokens": 287791415.0,
      "step": 9017
    },
    {
      "entropy": 0.5786485997959971,
      "epoch": 0.8297134576569427,
      "grad_norm": 0.14953181147575378,
      "learning_rate": 7.234642868095808e-05,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8253297582268715,
      "num_tokens": 287823751.0,
      "step": 9018
    },
    {
      "entropy": 0.7793207466602325,
      "epoch": 0.829805464028384,
      "grad_norm": 0.15791882574558258,
      "learning_rate": 7.234336185481645e-05,
      "loss": 0.7631,
      "mean_token_accuracy": 0.7762205973267555,
      "num_tokens": 287854817.0,
      "step": 9019
    },
    {
      "entropy": 0.8050232790410519,
      "epoch": 0.8298974703998252,
      "grad_norm": 0.15131767094135284,
      "learning_rate": 7.234029502867483e-05,
      "loss": 0.8078,
      "mean_token_accuracy": 0.7669456787407398,
      "num_tokens": 287887337.0,
      "step": 9020
    },
    {
      "entropy": 0.6526709124445915,
      "epoch": 0.8299894767712664,
      "grad_norm": 0.1452147662639618,
      "learning_rate": 7.23372282025332e-05,
      "loss": 0.6218,
      "mean_token_accuracy": 0.8102943859994411,
      "num_tokens": 287919502.0,
      "step": 9021
    },
    {
      "entropy": 0.8400962855666876,
      "epoch": 0.8300814831427077,
      "grad_norm": 0.16540585458278656,
      "learning_rate": 7.233416137639157e-05,
      "loss": 0.8527,
      "mean_token_accuracy": 0.7511075586080551,
      "num_tokens": 287951788.0,
      "step": 9022
    },
    {
      "entropy": 0.7471245685592294,
      "epoch": 0.8301734895141488,
      "grad_norm": 0.15939123928546906,
      "learning_rate": 7.233109455024995e-05,
      "loss": 0.7364,
      "mean_token_accuracy": 0.7833173461258411,
      "num_tokens": 287983948.0,
      "step": 9023
    },
    {
      "entropy": 0.7751171141862869,
      "epoch": 0.8302654958855901,
      "grad_norm": 0.1564231961965561,
      "learning_rate": 7.232802772410833e-05,
      "loss": 0.7696,
      "mean_token_accuracy": 0.7723954319953918,
      "num_tokens": 288015114.0,
      "step": 9024
    },
    {
      "entropy": 0.7902029156684875,
      "epoch": 0.8303575022570313,
      "grad_norm": 0.1552777737379074,
      "learning_rate": 7.23249608979667e-05,
      "loss": 0.7754,
      "mean_token_accuracy": 0.779893308877945,
      "num_tokens": 288046118.0,
      "step": 9025
    },
    {
      "entropy": 0.8213692829012871,
      "epoch": 0.8304495086284726,
      "grad_norm": 0.16438235342502594,
      "learning_rate": 7.232189407182507e-05,
      "loss": 0.8117,
      "mean_token_accuracy": 0.759682409465313,
      "num_tokens": 288077844.0,
      "step": 9026
    },
    {
      "entropy": 0.6879009269177914,
      "epoch": 0.8305415149999138,
      "grad_norm": 0.16105374693870544,
      "learning_rate": 7.231882724568343e-05,
      "loss": 0.6672,
      "mean_token_accuracy": 0.8013769239187241,
      "num_tokens": 288110194.0,
      "step": 9027
    },
    {
      "entropy": 0.7640702128410339,
      "epoch": 0.8306335213713549,
      "grad_norm": 0.15473443269729614,
      "learning_rate": 7.231576041954182e-05,
      "loss": 0.7554,
      "mean_token_accuracy": 0.7801374867558479,
      "num_tokens": 288142266.0,
      "step": 9028
    },
    {
      "entropy": 0.7252395208925009,
      "epoch": 0.8307255277427962,
      "grad_norm": 0.16033917665481567,
      "learning_rate": 7.23126935934002e-05,
      "loss": 0.7202,
      "mean_token_accuracy": 0.7915887907147408,
      "num_tokens": 288173513.0,
      "step": 9029
    },
    {
      "entropy": 0.7625559717416763,
      "epoch": 0.8308175341142374,
      "grad_norm": 0.155028834939003,
      "learning_rate": 7.230962676725857e-05,
      "loss": 0.7613,
      "mean_token_accuracy": 0.775902770459652,
      "num_tokens": 288204848.0,
      "step": 9030
    },
    {
      "entropy": 0.7226546164602041,
      "epoch": 0.8309095404856787,
      "grad_norm": 0.1539127230644226,
      "learning_rate": 7.230655994111695e-05,
      "loss": 0.6962,
      "mean_token_accuracy": 0.7994554825127125,
      "num_tokens": 288236687.0,
      "step": 9031
    },
    {
      "entropy": 0.7526086550205946,
      "epoch": 0.8310015468571199,
      "grad_norm": 0.14946207404136658,
      "learning_rate": 7.230349311497531e-05,
      "loss": 0.7499,
      "mean_token_accuracy": 0.7801882028579712,
      "num_tokens": 288268732.0,
      "step": 9032
    },
    {
      "entropy": 0.7035164404660463,
      "epoch": 0.831093553228561,
      "grad_norm": 0.16737692058086395,
      "learning_rate": 7.23004262888337e-05,
      "loss": 0.7129,
      "mean_token_accuracy": 0.7900338843464851,
      "num_tokens": 288300009.0,
      "step": 9033
    },
    {
      "entropy": 0.7043463783338666,
      "epoch": 0.8311855596000023,
      "grad_norm": 0.16487069427967072,
      "learning_rate": 7.229735946269206e-05,
      "loss": 0.697,
      "mean_token_accuracy": 0.7920679710805416,
      "num_tokens": 288332406.0,
      "step": 9034
    },
    {
      "entropy": 0.7247197423130274,
      "epoch": 0.8312775659714435,
      "grad_norm": 0.1542634814977646,
      "learning_rate": 7.229429263655045e-05,
      "loss": 0.7259,
      "mean_token_accuracy": 0.785319097340107,
      "num_tokens": 288363877.0,
      "step": 9035
    },
    {
      "entropy": 0.6642403639853001,
      "epoch": 0.8313695723428848,
      "grad_norm": 0.15734757483005524,
      "learning_rate": 7.229122581040881e-05,
      "loss": 0.6246,
      "mean_token_accuracy": 0.8100782036781311,
      "num_tokens": 288394914.0,
      "step": 9036
    },
    {
      "entropy": 0.6057250201702118,
      "epoch": 0.831461578714326,
      "grad_norm": 0.15744243562221527,
      "learning_rate": 7.228815898426718e-05,
      "loss": 0.603,
      "mean_token_accuracy": 0.8201303109526634,
      "num_tokens": 288426716.0,
      "step": 9037
    },
    {
      "entropy": 0.602662481367588,
      "epoch": 0.8315535850857672,
      "grad_norm": 0.1442810446023941,
      "learning_rate": 7.228509215812556e-05,
      "loss": 0.5744,
      "mean_token_accuracy": 0.8270961083471775,
      "num_tokens": 288458868.0,
      "step": 9038
    },
    {
      "entropy": 0.7822940368205309,
      "epoch": 0.8316455914572084,
      "grad_norm": 0.16748479008674622,
      "learning_rate": 7.228202533198394e-05,
      "loss": 0.7671,
      "mean_token_accuracy": 0.7802121862769127,
      "num_tokens": 288491024.0,
      "step": 9039
    },
    {
      "entropy": 0.7370800450444221,
      "epoch": 0.8317375978286496,
      "grad_norm": 0.15442612767219543,
      "learning_rate": 7.227895850584231e-05,
      "loss": 0.7262,
      "mean_token_accuracy": 0.7869678661227226,
      "num_tokens": 288523344.0,
      "step": 9040
    },
    {
      "entropy": 0.65129888150841,
      "epoch": 0.8318296042000909,
      "grad_norm": 0.1575363129377365,
      "learning_rate": 7.227589167970068e-05,
      "loss": 0.6418,
      "mean_token_accuracy": 0.8064895048737526,
      "num_tokens": 288556047.0,
      "step": 9041
    },
    {
      "entropy": 0.6010574614629149,
      "epoch": 0.8319216105715321,
      "grad_norm": 0.15597344934940338,
      "learning_rate": 7.227282485355905e-05,
      "loss": 0.571,
      "mean_token_accuracy": 0.8293974287807941,
      "num_tokens": 288587878.0,
      "step": 9042
    },
    {
      "entropy": 0.796992477029562,
      "epoch": 0.8320136169429733,
      "grad_norm": 0.15801484882831573,
      "learning_rate": 7.226975802741743e-05,
      "loss": 0.7919,
      "mean_token_accuracy": 0.7683635912835598,
      "num_tokens": 288620160.0,
      "step": 9043
    },
    {
      "entropy": 0.7380119673907757,
      "epoch": 0.8321056233144145,
      "grad_norm": 0.14947862923145294,
      "learning_rate": 7.226669120127581e-05,
      "loss": 0.7232,
      "mean_token_accuracy": 0.7856890857219696,
      "num_tokens": 288652329.0,
      "step": 9044
    },
    {
      "entropy": 0.679066464304924,
      "epoch": 0.8321976296858558,
      "grad_norm": 0.16159687936306,
      "learning_rate": 7.226362437513418e-05,
      "loss": 0.6788,
      "mean_token_accuracy": 0.7998374439775944,
      "num_tokens": 288684826.0,
      "step": 9045
    },
    {
      "entropy": 0.6982147675007582,
      "epoch": 0.832289636057297,
      "grad_norm": 0.15861134231090546,
      "learning_rate": 7.226055754899255e-05,
      "loss": 0.6792,
      "mean_token_accuracy": 0.7959214597940445,
      "num_tokens": 288715861.0,
      "step": 9046
    },
    {
      "entropy": 0.6829829849302769,
      "epoch": 0.8323816424287382,
      "grad_norm": 0.14946940541267395,
      "learning_rate": 7.225749072285093e-05,
      "loss": 0.6613,
      "mean_token_accuracy": 0.8035935424268246,
      "num_tokens": 288748191.0,
      "step": 9047
    },
    {
      "entropy": 0.7032972620800138,
      "epoch": 0.8324736488001794,
      "grad_norm": 0.1501259058713913,
      "learning_rate": 7.22544238967093e-05,
      "loss": 0.6875,
      "mean_token_accuracy": 0.7969786040484905,
      "num_tokens": 288780224.0,
      "step": 9048
    },
    {
      "entropy": 0.7577467486262321,
      "epoch": 0.8325656551716206,
      "grad_norm": 0.15988539159297943,
      "learning_rate": 7.225135707056768e-05,
      "loss": 0.7389,
      "mean_token_accuracy": 0.7871299050748348,
      "num_tokens": 288812769.0,
      "step": 9049
    },
    {
      "entropy": 0.8337681796401739,
      "epoch": 0.8326576615430619,
      "grad_norm": 0.16370603442192078,
      "learning_rate": 7.224829024442604e-05,
      "loss": 0.8125,
      "mean_token_accuracy": 0.7598007544875145,
      "num_tokens": 288843713.0,
      "step": 9050
    },
    {
      "entropy": 0.7212961576879025,
      "epoch": 0.8327496679145031,
      "grad_norm": 0.2335633933544159,
      "learning_rate": 7.224522341828443e-05,
      "loss": 0.7276,
      "mean_token_accuracy": 0.7881589718163013,
      "num_tokens": 288875768.0,
      "step": 9051
    },
    {
      "entropy": 0.7176481820642948,
      "epoch": 0.8328416742859444,
      "grad_norm": 0.15961724519729614,
      "learning_rate": 7.22421565921428e-05,
      "loss": 0.6965,
      "mean_token_accuracy": 0.7944708205759525,
      "num_tokens": 288908371.0,
      "step": 9052
    },
    {
      "entropy": 0.7095932513475418,
      "epoch": 0.8329336806573855,
      "grad_norm": 0.15164814889431,
      "learning_rate": 7.223908976600116e-05,
      "loss": 0.7097,
      "mean_token_accuracy": 0.790350180119276,
      "num_tokens": 288940948.0,
      "step": 9053
    },
    {
      "entropy": 0.6601394917815924,
      "epoch": 0.8330256870288267,
      "grad_norm": 0.14953729510307312,
      "learning_rate": 7.223602293985954e-05,
      "loss": 0.6484,
      "mean_token_accuracy": 0.8044304214417934,
      "num_tokens": 288972525.0,
      "step": 9054
    },
    {
      "entropy": 0.7356527559459209,
      "epoch": 0.833117693400268,
      "grad_norm": 0.15007644891738892,
      "learning_rate": 7.223295611371792e-05,
      "loss": 0.7261,
      "mean_token_accuracy": 0.7848861217498779,
      "num_tokens": 289004656.0,
      "step": 9055
    },
    {
      "entropy": 0.6611402835696936,
      "epoch": 0.8332096997717092,
      "grad_norm": 0.1511750966310501,
      "learning_rate": 7.222988928757629e-05,
      "loss": 0.642,
      "mean_token_accuracy": 0.8087352104485035,
      "num_tokens": 289036981.0,
      "step": 9056
    },
    {
      "entropy": 0.72827123478055,
      "epoch": 0.8333017061431505,
      "grad_norm": 0.15757159888744354,
      "learning_rate": 7.222682246143466e-05,
      "loss": 0.7119,
      "mean_token_accuracy": 0.7875059805810452,
      "num_tokens": 289068261.0,
      "step": 9057
    },
    {
      "entropy": 0.7024492742493749,
      "epoch": 0.8333937125145916,
      "grad_norm": 0.15552684664726257,
      "learning_rate": 7.222375563529303e-05,
      "loss": 0.6958,
      "mean_token_accuracy": 0.7982423640787601,
      "num_tokens": 289100155.0,
      "step": 9058
    },
    {
      "entropy": 0.7100191693753004,
      "epoch": 0.8334857188860328,
      "grad_norm": 0.16387872397899628,
      "learning_rate": 7.222068880915141e-05,
      "loss": 0.7039,
      "mean_token_accuracy": 0.79165318608284,
      "num_tokens": 289132534.0,
      "step": 9059
    },
    {
      "entropy": 0.7278943229466677,
      "epoch": 0.8335777252574741,
      "grad_norm": 0.1531963348388672,
      "learning_rate": 7.221762198300979e-05,
      "loss": 0.7248,
      "mean_token_accuracy": 0.7876242324709892,
      "num_tokens": 289164482.0,
      "step": 9060
    },
    {
      "entropy": 0.8459995836019516,
      "epoch": 0.8336697316289153,
      "grad_norm": 0.15800246596336365,
      "learning_rate": 7.221455515686816e-05,
      "loss": 0.8338,
      "mean_token_accuracy": 0.7521229311823845,
      "num_tokens": 289196679.0,
      "step": 9061
    },
    {
      "entropy": 0.7535284496843815,
      "epoch": 0.8337617380003566,
      "grad_norm": 0.15873971581459045,
      "learning_rate": 7.221148833072654e-05,
      "loss": 0.732,
      "mean_token_accuracy": 0.7823768369853497,
      "num_tokens": 289228110.0,
      "step": 9062
    },
    {
      "entropy": 0.7912134602665901,
      "epoch": 0.8338537443717977,
      "grad_norm": 0.16510143876075745,
      "learning_rate": 7.220842150458491e-05,
      "loss": 0.7828,
      "mean_token_accuracy": 0.7714928016066551,
      "num_tokens": 289259687.0,
      "step": 9063
    },
    {
      "entropy": 0.7188374362885952,
      "epoch": 0.833945750743239,
      "grad_norm": 0.1576336920261383,
      "learning_rate": 7.220535467844328e-05,
      "loss": 0.7222,
      "mean_token_accuracy": 0.7913253493607044,
      "num_tokens": 289292306.0,
      "step": 9064
    },
    {
      "entropy": 0.6333608757704496,
      "epoch": 0.8340377571146802,
      "grad_norm": 0.15111960470676422,
      "learning_rate": 7.220228785230166e-05,
      "loss": 0.6252,
      "mean_token_accuracy": 0.8152494542300701,
      "num_tokens": 289324022.0,
      "step": 9065
    },
    {
      "entropy": 0.6546909417957067,
      "epoch": 0.8341297634861214,
      "grad_norm": 0.15430863201618195,
      "learning_rate": 7.219922102616004e-05,
      "loss": 0.6515,
      "mean_token_accuracy": 0.802647054195404,
      "num_tokens": 289356660.0,
      "step": 9066
    },
    {
      "entropy": 0.6999685764312744,
      "epoch": 0.8342217698575627,
      "grad_norm": 0.1590089499950409,
      "learning_rate": 7.219615420001841e-05,
      "loss": 0.6781,
      "mean_token_accuracy": 0.7941929548978806,
      "num_tokens": 289387870.0,
      "step": 9067
    },
    {
      "entropy": 0.752943177241832,
      "epoch": 0.8343137762290038,
      "grad_norm": 0.16467207670211792,
      "learning_rate": 7.219308737387677e-05,
      "loss": 0.7364,
      "mean_token_accuracy": 0.7855724282562733,
      "num_tokens": 289419938.0,
      "step": 9068
    },
    {
      "entropy": 0.8011746732518077,
      "epoch": 0.8344057826004451,
      "grad_norm": 0.15543636679649353,
      "learning_rate": 7.219002054773514e-05,
      "loss": 0.7952,
      "mean_token_accuracy": 0.7682080380618572,
      "num_tokens": 289452242.0,
      "step": 9069
    },
    {
      "entropy": 0.7420968264341354,
      "epoch": 0.8344977889718863,
      "grad_norm": 0.15573789179325104,
      "learning_rate": 7.218695372159354e-05,
      "loss": 0.7404,
      "mean_token_accuracy": 0.7851684652268887,
      "num_tokens": 289483012.0,
      "step": 9070
    },
    {
      "entropy": 0.7893108427524567,
      "epoch": 0.8345897953433276,
      "grad_norm": 0.15170513093471527,
      "learning_rate": 7.21838868954519e-05,
      "loss": 0.7567,
      "mean_token_accuracy": 0.7760877497494221,
      "num_tokens": 289514820.0,
      "step": 9071
    },
    {
      "entropy": 0.6985017471015453,
      "epoch": 0.8346818017147688,
      "grad_norm": 0.14756865799427032,
      "learning_rate": 7.218082006931027e-05,
      "loss": 0.6931,
      "mean_token_accuracy": 0.7948386967182159,
      "num_tokens": 289547096.0,
      "step": 9072
    },
    {
      "entropy": 0.7017037626355886,
      "epoch": 0.8347738080862099,
      "grad_norm": 0.15140597522258759,
      "learning_rate": 7.217775324316864e-05,
      "loss": 0.6785,
      "mean_token_accuracy": 0.7986987791955471,
      "num_tokens": 289579199.0,
      "step": 9073
    },
    {
      "entropy": 0.7091028904542327,
      "epoch": 0.8348658144576512,
      "grad_norm": 0.16450288891792297,
      "learning_rate": 7.217468641702702e-05,
      "loss": 0.6771,
      "mean_token_accuracy": 0.8006963580846786,
      "num_tokens": 289610801.0,
      "step": 9074
    },
    {
      "entropy": 0.7282332964241505,
      "epoch": 0.8349578208290924,
      "grad_norm": 0.155295267701149,
      "learning_rate": 7.21716195908854e-05,
      "loss": 0.7169,
      "mean_token_accuracy": 0.790201723575592,
      "num_tokens": 289642645.0,
      "step": 9075
    },
    {
      "entropy": 0.7702591791749,
      "epoch": 0.8350498272005337,
      "grad_norm": 0.16323600709438324,
      "learning_rate": 7.216855276474377e-05,
      "loss": 0.74,
      "mean_token_accuracy": 0.780291423201561,
      "num_tokens": 289674698.0,
      "step": 9076
    },
    {
      "entropy": 0.7883815765380859,
      "epoch": 0.8351418335719749,
      "grad_norm": 0.16881293058395386,
      "learning_rate": 7.216548593860214e-05,
      "loss": 0.8054,
      "mean_token_accuracy": 0.7657389044761658,
      "num_tokens": 289705687.0,
      "step": 9077
    },
    {
      "entropy": 0.7313745729625225,
      "epoch": 0.835233839943416,
      "grad_norm": 0.1524861603975296,
      "learning_rate": 7.216241911246052e-05,
      "loss": 0.7197,
      "mean_token_accuracy": 0.7861227318644524,
      "num_tokens": 289736818.0,
      "step": 9078
    },
    {
      "entropy": 0.8070585038512945,
      "epoch": 0.8353258463148573,
      "grad_norm": 0.15937016904354095,
      "learning_rate": 7.215935228631889e-05,
      "loss": 0.7981,
      "mean_token_accuracy": 0.7645954303443432,
      "num_tokens": 289767905.0,
      "step": 9079
    },
    {
      "entropy": 0.5677783973515034,
      "epoch": 0.8354178526862985,
      "grad_norm": 0.16189704835414886,
      "learning_rate": 7.215628546017727e-05,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8356296569108963,
      "num_tokens": 289800053.0,
      "step": 9080
    },
    {
      "entropy": 0.7145978566259146,
      "epoch": 0.8355098590577398,
      "grad_norm": 0.16118332743644714,
      "learning_rate": 7.215321863403564e-05,
      "loss": 0.6958,
      "mean_token_accuracy": 0.7911686673760414,
      "num_tokens": 289832008.0,
      "step": 9081
    },
    {
      "entropy": 0.7620208309963346,
      "epoch": 0.835601865429181,
      "grad_norm": 0.1525549441576004,
      "learning_rate": 7.215015180789402e-05,
      "loss": 0.7387,
      "mean_token_accuracy": 0.7821249067783356,
      "num_tokens": 289864246.0,
      "step": 9082
    },
    {
      "entropy": 0.6497007552534342,
      "epoch": 0.8356938718006222,
      "grad_norm": 0.14238062500953674,
      "learning_rate": 7.214708498175239e-05,
      "loss": 0.6277,
      "mean_token_accuracy": 0.8129583597183228,
      "num_tokens": 289895952.0,
      "step": 9083
    },
    {
      "entropy": 0.7372745517641306,
      "epoch": 0.8357858781720634,
      "grad_norm": 0.1538478583097458,
      "learning_rate": 7.214401815561076e-05,
      "loss": 0.7108,
      "mean_token_accuracy": 0.7909903712570667,
      "num_tokens": 289928402.0,
      "step": 9084
    },
    {
      "entropy": 0.6672486402094364,
      "epoch": 0.8358778845435046,
      "grad_norm": 0.15175490081310272,
      "learning_rate": 7.214095132946914e-05,
      "loss": 0.6577,
      "mean_token_accuracy": 0.7982964813709259,
      "num_tokens": 289959607.0,
      "step": 9085
    },
    {
      "entropy": 0.7003800440579653,
      "epoch": 0.8359698909149459,
      "grad_norm": 0.15945777297019958,
      "learning_rate": 7.213788450332752e-05,
      "loss": 0.6851,
      "mean_token_accuracy": 0.7994478903710842,
      "num_tokens": 289990766.0,
      "step": 9086
    },
    {
      "entropy": 0.7068077716976404,
      "epoch": 0.8360618972863871,
      "grad_norm": 0.15663355588912964,
      "learning_rate": 7.213481767718589e-05,
      "loss": 0.7013,
      "mean_token_accuracy": 0.7938034310936928,
      "num_tokens": 290022675.0,
      "step": 9087
    },
    {
      "entropy": 0.6730928458273411,
      "epoch": 0.8361539036578283,
      "grad_norm": 0.16182364523410797,
      "learning_rate": 7.213175085104425e-05,
      "loss": 0.6678,
      "mean_token_accuracy": 0.7998382449150085,
      "num_tokens": 290055151.0,
      "step": 9088
    },
    {
      "entropy": 0.7336231525987387,
      "epoch": 0.8362459100292695,
      "grad_norm": 0.15595953166484833,
      "learning_rate": 7.212868402490262e-05,
      "loss": 0.7375,
      "mean_token_accuracy": 0.7802033089101315,
      "num_tokens": 290087012.0,
      "step": 9089
    },
    {
      "entropy": 0.6424590721726418,
      "epoch": 0.8363379164007108,
      "grad_norm": 0.1499582827091217,
      "learning_rate": 7.2125617198761e-05,
      "loss": 0.6244,
      "mean_token_accuracy": 0.8121199309825897,
      "num_tokens": 290118774.0,
      "step": 9090
    },
    {
      "entropy": 0.7046530190855265,
      "epoch": 0.836429922772152,
      "grad_norm": 0.1526794135570526,
      "learning_rate": 7.212255037261938e-05,
      "loss": 0.7037,
      "mean_token_accuracy": 0.7907492406666279,
      "num_tokens": 290151073.0,
      "step": 9091
    },
    {
      "entropy": 0.6194196175783873,
      "epoch": 0.8365219291435932,
      "grad_norm": 0.14288388192653656,
      "learning_rate": 7.211948354647775e-05,
      "loss": 0.5984,
      "mean_token_accuracy": 0.8217645697295666,
      "num_tokens": 290182854.0,
      "step": 9092
    },
    {
      "entropy": 0.641222364269197,
      "epoch": 0.8366139355150344,
      "grad_norm": 0.15027932822704315,
      "learning_rate": 7.211641672033613e-05,
      "loss": 0.6288,
      "mean_token_accuracy": 0.8117330186069012,
      "num_tokens": 290215089.0,
      "step": 9093
    },
    {
      "entropy": 0.7702732374891639,
      "epoch": 0.8367059418864756,
      "grad_norm": 0.16285118460655212,
      "learning_rate": 7.21133498941945e-05,
      "loss": 0.7643,
      "mean_token_accuracy": 0.7727273516356945,
      "num_tokens": 290246879.0,
      "step": 9094
    },
    {
      "entropy": 0.6996157485991716,
      "epoch": 0.8367979482579169,
      "grad_norm": 0.1448875069618225,
      "learning_rate": 7.211028306805287e-05,
      "loss": 0.6898,
      "mean_token_accuracy": 0.7903244458138943,
      "num_tokens": 290279596.0,
      "step": 9095
    },
    {
      "entropy": 0.7506060730665922,
      "epoch": 0.8368899546293581,
      "grad_norm": 0.15412555634975433,
      "learning_rate": 7.210721624191125e-05,
      "loss": 0.7507,
      "mean_token_accuracy": 0.7809656746685505,
      "num_tokens": 290311272.0,
      "step": 9096
    },
    {
      "entropy": 0.7312762401998043,
      "epoch": 0.8369819610007994,
      "grad_norm": 0.1577703356742859,
      "learning_rate": 7.210414941576963e-05,
      "loss": 0.7324,
      "mean_token_accuracy": 0.7790478616952896,
      "num_tokens": 290343641.0,
      "step": 9097
    },
    {
      "entropy": 0.8387475442141294,
      "epoch": 0.8370739673722405,
      "grad_norm": 0.16553093492984772,
      "learning_rate": 7.2101082589628e-05,
      "loss": 0.8237,
      "mean_token_accuracy": 0.7642966508865356,
      "num_tokens": 290374845.0,
      "step": 9098
    },
    {
      "entropy": 0.7379847355186939,
      "epoch": 0.8371659737436817,
      "grad_norm": 0.15322065353393555,
      "learning_rate": 7.209801576348637e-05,
      "loss": 0.7349,
      "mean_token_accuracy": 0.7838339321315289,
      "num_tokens": 290406960.0,
      "step": 9099
    },
    {
      "entropy": 0.6216367911547422,
      "epoch": 0.837257980115123,
      "grad_norm": 0.15819482505321503,
      "learning_rate": 7.209494893734474e-05,
      "loss": 0.596,
      "mean_token_accuracy": 0.8189467936754227,
      "num_tokens": 290438372.0,
      "step": 9100
    },
    {
      "entropy": 0.7426403919234872,
      "epoch": 0.8373499864865642,
      "grad_norm": 0.1579977571964264,
      "learning_rate": 7.209188211120312e-05,
      "loss": 0.7343,
      "mean_token_accuracy": 0.7812631540000439,
      "num_tokens": 290470854.0,
      "step": 9101
    },
    {
      "entropy": 0.6809677165001631,
      "epoch": 0.8374419928580055,
      "grad_norm": 0.15011511743068695,
      "learning_rate": 7.20888152850615e-05,
      "loss": 0.6706,
      "mean_token_accuracy": 0.8011860884726048,
      "num_tokens": 290503406.0,
      "step": 9102
    },
    {
      "entropy": 0.7277039233595133,
      "epoch": 0.8375339992294466,
      "grad_norm": 0.1522298902273178,
      "learning_rate": 7.208574845891987e-05,
      "loss": 0.7198,
      "mean_token_accuracy": 0.7868996374309063,
      "num_tokens": 290534696.0,
      "step": 9103
    },
    {
      "entropy": 0.7818477936089039,
      "epoch": 0.8376260056008878,
      "grad_norm": 0.16164185106754303,
      "learning_rate": 7.208268163277823e-05,
      "loss": 0.7753,
      "mean_token_accuracy": 0.7724071629345417,
      "num_tokens": 290566204.0,
      "step": 9104
    },
    {
      "entropy": 0.7703542821109295,
      "epoch": 0.8377180119723291,
      "grad_norm": 0.15275658667087555,
      "learning_rate": 7.207961480663662e-05,
      "loss": 0.7656,
      "mean_token_accuracy": 0.772015817463398,
      "num_tokens": 290597830.0,
      "step": 9105
    },
    {
      "entropy": 0.7607256602495909,
      "epoch": 0.8378100183437703,
      "grad_norm": 0.1534416675567627,
      "learning_rate": 7.207654798049498e-05,
      "loss": 0.7529,
      "mean_token_accuracy": 0.7785833179950714,
      "num_tokens": 290629931.0,
      "step": 9106
    },
    {
      "entropy": 0.8179734088480473,
      "epoch": 0.8379020247152116,
      "grad_norm": 0.16052763164043427,
      "learning_rate": 7.207348115435337e-05,
      "loss": 0.8058,
      "mean_token_accuracy": 0.7636258080601692,
      "num_tokens": 290662072.0,
      "step": 9107
    },
    {
      "entropy": 0.6669319029897451,
      "epoch": 0.8379940310866527,
      "grad_norm": 0.1633872240781784,
      "learning_rate": 7.207041432821173e-05,
      "loss": 0.6702,
      "mean_token_accuracy": 0.8028389737010002,
      "num_tokens": 290694318.0,
      "step": 9108
    },
    {
      "entropy": 0.7575189173221588,
      "epoch": 0.838086037458094,
      "grad_norm": 0.15163789689540863,
      "learning_rate": 7.206734750207011e-05,
      "loss": 0.7474,
      "mean_token_accuracy": 0.7725635282695293,
      "num_tokens": 290726059.0,
      "step": 9109
    },
    {
      "entropy": 0.7028163289651275,
      "epoch": 0.8381780438295352,
      "grad_norm": 0.15795832872390747,
      "learning_rate": 7.206428067592848e-05,
      "loss": 0.6891,
      "mean_token_accuracy": 0.8006570860743523,
      "num_tokens": 290757574.0,
      "step": 9110
    },
    {
      "entropy": 0.8231040649116039,
      "epoch": 0.8382700502009764,
      "grad_norm": 0.15850788354873657,
      "learning_rate": 7.206121384978685e-05,
      "loss": 0.8029,
      "mean_token_accuracy": 0.7611682675778866,
      "num_tokens": 290789197.0,
      "step": 9111
    },
    {
      "entropy": 0.6808954635635018,
      "epoch": 0.8383620565724177,
      "grad_norm": 0.14355963468551636,
      "learning_rate": 7.205814702364523e-05,
      "loss": 0.6668,
      "mean_token_accuracy": 0.8031413070857525,
      "num_tokens": 290821618.0,
      "step": 9112
    },
    {
      "entropy": 0.824683353304863,
      "epoch": 0.8384540629438588,
      "grad_norm": 0.15829217433929443,
      "learning_rate": 7.205508019750361e-05,
      "loss": 0.8353,
      "mean_token_accuracy": 0.7573675513267517,
      "num_tokens": 290854199.0,
      "step": 9113
    },
    {
      "entropy": 0.655140820890665,
      "epoch": 0.8385460693153001,
      "grad_norm": 0.14977408945560455,
      "learning_rate": 7.205201337136198e-05,
      "loss": 0.6284,
      "mean_token_accuracy": 0.8129732832312584,
      "num_tokens": 290886562.0,
      "step": 9114
    },
    {
      "entropy": 0.7693132124841213,
      "epoch": 0.8386380756867413,
      "grad_norm": 0.1705305427312851,
      "learning_rate": 7.204894654522035e-05,
      "loss": 0.7726,
      "mean_token_accuracy": 0.7719013281166553,
      "num_tokens": 290917860.0,
      "step": 9115
    },
    {
      "entropy": 0.7068646363914013,
      "epoch": 0.8387300820581826,
      "grad_norm": 0.16985681653022766,
      "learning_rate": 7.204587971907872e-05,
      "loss": 0.7114,
      "mean_token_accuracy": 0.7856886014342308,
      "num_tokens": 290949963.0,
      "step": 9116
    },
    {
      "entropy": 0.72588218934834,
      "epoch": 0.8388220884296238,
      "grad_norm": 0.16153481602668762,
      "learning_rate": 7.204281289293711e-05,
      "loss": 0.7233,
      "mean_token_accuracy": 0.7865076884627342,
      "num_tokens": 290981436.0,
      "step": 9117
    },
    {
      "entropy": 0.6672923490405083,
      "epoch": 0.8389140948010649,
      "grad_norm": 0.1622495800256729,
      "learning_rate": 7.203974606679548e-05,
      "loss": 0.6636,
      "mean_token_accuracy": 0.8006380833685398,
      "num_tokens": 291012455.0,
      "step": 9118
    },
    {
      "entropy": 0.7106853816658258,
      "epoch": 0.8390061011725062,
      "grad_norm": 0.16657352447509766,
      "learning_rate": 7.203667924065385e-05,
      "loss": 0.7224,
      "mean_token_accuracy": 0.78583649918437,
      "num_tokens": 291043099.0,
      "step": 9119
    },
    {
      "entropy": 0.6950165051966906,
      "epoch": 0.8390981075439474,
      "grad_norm": 0.15076442062854767,
      "learning_rate": 7.203361241451222e-05,
      "loss": 0.6803,
      "mean_token_accuracy": 0.7965792417526245,
      "num_tokens": 291075162.0,
      "step": 9120
    },
    {
      "entropy": 0.6909131482243538,
      "epoch": 0.8391901139153887,
      "grad_norm": 0.15015725791454315,
      "learning_rate": 7.20305455883706e-05,
      "loss": 0.6501,
      "mean_token_accuracy": 0.8066425733268261,
      "num_tokens": 291107338.0,
      "step": 9121
    },
    {
      "entropy": 0.8037039646878839,
      "epoch": 0.8392821202868299,
      "grad_norm": 0.16066840291023254,
      "learning_rate": 7.202747876222898e-05,
      "loss": 0.7796,
      "mean_token_accuracy": 0.774312362074852,
      "num_tokens": 291138768.0,
      "step": 9122
    },
    {
      "entropy": 0.795921178534627,
      "epoch": 0.839374126658271,
      "grad_norm": 0.16759628057479858,
      "learning_rate": 7.202441193608735e-05,
      "loss": 0.7814,
      "mean_token_accuracy": 0.7702923193573952,
      "num_tokens": 291170495.0,
      "step": 9123
    },
    {
      "entropy": 0.7076387796550989,
      "epoch": 0.8394661330297123,
      "grad_norm": 0.15972813963890076,
      "learning_rate": 7.202134510994573e-05,
      "loss": 0.6906,
      "mean_token_accuracy": 0.7926385439932346,
      "num_tokens": 291201209.0,
      "step": 9124
    },
    {
      "entropy": 0.6522480361163616,
      "epoch": 0.8395581394011535,
      "grad_norm": 0.15658971667289734,
      "learning_rate": 7.20182782838041e-05,
      "loss": 0.6438,
      "mean_token_accuracy": 0.8082863688468933,
      "num_tokens": 291232479.0,
      "step": 9125
    },
    {
      "entropy": 0.7804186958819628,
      "epoch": 0.8396501457725948,
      "grad_norm": 0.15578942000865936,
      "learning_rate": 7.201521145766246e-05,
      "loss": 0.7543,
      "mean_token_accuracy": 0.7833068482577801,
      "num_tokens": 291264594.0,
      "step": 9126
    },
    {
      "entropy": 0.7531399819999933,
      "epoch": 0.839742152144036,
      "grad_norm": 0.16203129291534424,
      "learning_rate": 7.201214463152084e-05,
      "loss": 0.7453,
      "mean_token_accuracy": 0.778734628111124,
      "num_tokens": 291296092.0,
      "step": 9127
    },
    {
      "entropy": 0.7967815641313791,
      "epoch": 0.8398341585154772,
      "grad_norm": 0.16317884624004364,
      "learning_rate": 7.200907780537923e-05,
      "loss": 0.7801,
      "mean_token_accuracy": 0.7698479257524014,
      "num_tokens": 291327205.0,
      "step": 9128
    },
    {
      "entropy": 0.675640681758523,
      "epoch": 0.8399261648869184,
      "grad_norm": 0.15583230555057526,
      "learning_rate": 7.20060109792376e-05,
      "loss": 0.6721,
      "mean_token_accuracy": 0.8022660538554192,
      "num_tokens": 291359155.0,
      "step": 9129
    },
    {
      "entropy": 0.7380182724446058,
      "epoch": 0.8400181712583596,
      "grad_norm": 0.14910826086997986,
      "learning_rate": 7.200294415309596e-05,
      "loss": 0.7199,
      "mean_token_accuracy": 0.7871597744524479,
      "num_tokens": 291391020.0,
      "step": 9130
    },
    {
      "entropy": 0.6940079070627689,
      "epoch": 0.8401101776298009,
      "grad_norm": 0.15100473165512085,
      "learning_rate": 7.199987732695433e-05,
      "loss": 0.675,
      "mean_token_accuracy": 0.7931666970252991,
      "num_tokens": 291423005.0,
      "step": 9131
    },
    {
      "entropy": 0.7716233115643263,
      "epoch": 0.8402021840012421,
      "grad_norm": 0.16140922904014587,
      "learning_rate": 7.199681050081271e-05,
      "loss": 0.7455,
      "mean_token_accuracy": 0.7789973504841328,
      "num_tokens": 291454862.0,
      "step": 9132
    },
    {
      "entropy": 0.7918229848146439,
      "epoch": 0.8402941903726833,
      "grad_norm": 0.16213946044445038,
      "learning_rate": 7.199374367467109e-05,
      "loss": 0.7791,
      "mean_token_accuracy": 0.7711617462337017,
      "num_tokens": 291486798.0,
      "step": 9133
    },
    {
      "entropy": 0.7759582027792931,
      "epoch": 0.8403861967441245,
      "grad_norm": 0.15595367550849915,
      "learning_rate": 7.199067684852946e-05,
      "loss": 0.7614,
      "mean_token_accuracy": 0.7725962623953819,
      "num_tokens": 291518655.0,
      "step": 9134
    },
    {
      "entropy": 0.7150936499238014,
      "epoch": 0.8404782031155658,
      "grad_norm": 0.15224827826023102,
      "learning_rate": 7.198761002238783e-05,
      "loss": 0.7056,
      "mean_token_accuracy": 0.7914420291781425,
      "num_tokens": 291551316.0,
      "step": 9135
    },
    {
      "entropy": 0.6603256389498711,
      "epoch": 0.840570209487007,
      "grad_norm": 0.1530485302209854,
      "learning_rate": 7.198454319624621e-05,
      "loss": 0.6539,
      "mean_token_accuracy": 0.8050058484077454,
      "num_tokens": 291583157.0,
      "step": 9136
    },
    {
      "entropy": 0.710412859916687,
      "epoch": 0.8406622158584482,
      "grad_norm": 0.161378413438797,
      "learning_rate": 7.198147637010458e-05,
      "loss": 0.7088,
      "mean_token_accuracy": 0.7920417338609695,
      "num_tokens": 291615244.0,
      "step": 9137
    },
    {
      "entropy": 0.7094080299139023,
      "epoch": 0.8407542222298894,
      "grad_norm": 0.1612183302640915,
      "learning_rate": 7.197840954396296e-05,
      "loss": 0.7155,
      "mean_token_accuracy": 0.7852180972695351,
      "num_tokens": 291647202.0,
      "step": 9138
    },
    {
      "entropy": 0.6052472312003374,
      "epoch": 0.8408462286013306,
      "grad_norm": 0.14680461585521698,
      "learning_rate": 7.197534271782133e-05,
      "loss": 0.5818,
      "mean_token_accuracy": 0.8263604193925858,
      "num_tokens": 291679447.0,
      "step": 9139
    },
    {
      "entropy": 0.7243910431861877,
      "epoch": 0.8409382349727719,
      "grad_norm": 0.15329138934612274,
      "learning_rate": 7.197227589167971e-05,
      "loss": 0.7085,
      "mean_token_accuracy": 0.7877175360918045,
      "num_tokens": 291711428.0,
      "step": 9140
    },
    {
      "entropy": 0.6879602484405041,
      "epoch": 0.8410302413442131,
      "grad_norm": 0.1530250608921051,
      "learning_rate": 7.196920906553808e-05,
      "loss": 0.6899,
      "mean_token_accuracy": 0.7967805787920952,
      "num_tokens": 291743326.0,
      "step": 9141
    },
    {
      "entropy": 0.7535233972594142,
      "epoch": 0.8411222477156544,
      "grad_norm": 0.15617775917053223,
      "learning_rate": 7.196614223939644e-05,
      "loss": 0.7602,
      "mean_token_accuracy": 0.7801307067275047,
      "num_tokens": 291775639.0,
      "step": 9142
    },
    {
      "entropy": 0.7731720898300409,
      "epoch": 0.8412142540870955,
      "grad_norm": 0.16273672878742218,
      "learning_rate": 7.196307541325483e-05,
      "loss": 0.7562,
      "mean_token_accuracy": 0.7743697874248028,
      "num_tokens": 291807481.0,
      "step": 9143
    },
    {
      "entropy": 0.8433875795453787,
      "epoch": 0.8413062604585367,
      "grad_norm": 0.15419432520866394,
      "learning_rate": 7.196000858711321e-05,
      "loss": 0.8247,
      "mean_token_accuracy": 0.7588876411318779,
      "num_tokens": 291839453.0,
      "step": 9144
    },
    {
      "entropy": 0.7226981520652771,
      "epoch": 0.841398266829978,
      "grad_norm": 0.1483229547739029,
      "learning_rate": 7.195694176097158e-05,
      "loss": 0.7016,
      "mean_token_accuracy": 0.7941842451691628,
      "num_tokens": 291871312.0,
      "step": 9145
    },
    {
      "entropy": 0.6813469771295786,
      "epoch": 0.8414902732014192,
      "grad_norm": 0.1489291936159134,
      "learning_rate": 7.195387493482994e-05,
      "loss": 0.6662,
      "mean_token_accuracy": 0.8007392100989819,
      "num_tokens": 291903543.0,
      "step": 9146
    },
    {
      "entropy": 0.6625216286629438,
      "epoch": 0.8415822795728605,
      "grad_norm": 0.1504790335893631,
      "learning_rate": 7.195080810868831e-05,
      "loss": 0.6453,
      "mean_token_accuracy": 0.8096517100930214,
      "num_tokens": 291935159.0,
      "step": 9147
    },
    {
      "entropy": 0.7267012242227793,
      "epoch": 0.8416742859443016,
      "grad_norm": 0.16325423121452332,
      "learning_rate": 7.194774128254669e-05,
      "loss": 0.7182,
      "mean_token_accuracy": 0.7912342511117458,
      "num_tokens": 291967169.0,
      "step": 9148
    },
    {
      "entropy": 0.709007864817977,
      "epoch": 0.8417662923157428,
      "grad_norm": 0.1657252311706543,
      "learning_rate": 7.194467445640507e-05,
      "loss": 0.7039,
      "mean_token_accuracy": 0.7942780181765556,
      "num_tokens": 291999746.0,
      "step": 9149
    },
    {
      "entropy": 0.7095518782734871,
      "epoch": 0.8418582986871841,
      "grad_norm": 0.16132722795009613,
      "learning_rate": 7.194160763026344e-05,
      "loss": 0.7033,
      "mean_token_accuracy": 0.7924651354551315,
      "num_tokens": 292031823.0,
      "step": 9150
    },
    {
      "entropy": 0.6440283069387078,
      "epoch": 0.8419503050586253,
      "grad_norm": 0.1572652906179428,
      "learning_rate": 7.193854080412181e-05,
      "loss": 0.6184,
      "mean_token_accuracy": 0.8158353306353092,
      "num_tokens": 292063784.0,
      "step": 9151
    },
    {
      "entropy": 0.7242790758609772,
      "epoch": 0.8420423114300666,
      "grad_norm": 0.15191426873207092,
      "learning_rate": 7.193547397798019e-05,
      "loss": 0.7255,
      "mean_token_accuracy": 0.7837333753705025,
      "num_tokens": 292096225.0,
      "step": 9152
    },
    {
      "entropy": 0.8049300238490105,
      "epoch": 0.8421343178015077,
      "grad_norm": 0.16763795912265778,
      "learning_rate": 7.193240715183857e-05,
      "loss": 0.8077,
      "mean_token_accuracy": 0.7666965015232563,
      "num_tokens": 292127685.0,
      "step": 9153
    },
    {
      "entropy": 0.7672038618475199,
      "epoch": 0.842226324172949,
      "grad_norm": 0.1521313339471817,
      "learning_rate": 7.192934032569694e-05,
      "loss": 0.7606,
      "mean_token_accuracy": 0.7699453718960285,
      "num_tokens": 292159234.0,
      "step": 9154
    },
    {
      "entropy": 0.670178510248661,
      "epoch": 0.8423183305443902,
      "grad_norm": 0.15014037489891052,
      "learning_rate": 7.192627349955532e-05,
      "loss": 0.6628,
      "mean_token_accuracy": 0.8012550435960293,
      "num_tokens": 292191241.0,
      "step": 9155
    },
    {
      "entropy": 0.7219825275242329,
      "epoch": 0.8424103369158314,
      "grad_norm": 0.1495664119720459,
      "learning_rate": 7.192320667341369e-05,
      "loss": 0.7103,
      "mean_token_accuracy": 0.7907122112810612,
      "num_tokens": 292222894.0,
      "step": 9156
    },
    {
      "entropy": 0.6905241142958403,
      "epoch": 0.8425023432872727,
      "grad_norm": 0.15638595819473267,
      "learning_rate": 7.192013984727206e-05,
      "loss": 0.668,
      "mean_token_accuracy": 0.7961458154022694,
      "num_tokens": 292254580.0,
      "step": 9157
    },
    {
      "entropy": 0.7114584296941757,
      "epoch": 0.8425943496587138,
      "grad_norm": 0.15870784223079681,
      "learning_rate": 7.191707302113044e-05,
      "loss": 0.6985,
      "mean_token_accuracy": 0.7920849099755287,
      "num_tokens": 292286119.0,
      "step": 9158
    },
    {
      "entropy": 0.6766656888648868,
      "epoch": 0.8426863560301551,
      "grad_norm": 0.14660494029521942,
      "learning_rate": 7.191400619498882e-05,
      "loss": 0.6636,
      "mean_token_accuracy": 0.8052758872509003,
      "num_tokens": 292318391.0,
      "step": 9159
    },
    {
      "entropy": 0.6700955852866173,
      "epoch": 0.8427783624015963,
      "grad_norm": 0.15591922402381897,
      "learning_rate": 7.191093936884719e-05,
      "loss": 0.6697,
      "mean_token_accuracy": 0.8001781888306141,
      "num_tokens": 292350652.0,
      "step": 9160
    },
    {
      "entropy": 0.7005115170031786,
      "epoch": 0.8428703687730376,
      "grad_norm": 0.14557160437107086,
      "learning_rate": 7.190787254270556e-05,
      "loss": 0.6929,
      "mean_token_accuracy": 0.7915661409497261,
      "num_tokens": 292382140.0,
      "step": 9161
    },
    {
      "entropy": 0.8136640302836895,
      "epoch": 0.8429623751444788,
      "grad_norm": 0.15356279909610748,
      "learning_rate": 7.190480571656392e-05,
      "loss": 0.8031,
      "mean_token_accuracy": 0.764214813709259,
      "num_tokens": 292414626.0,
      "step": 9162
    },
    {
      "entropy": 0.7441516611725092,
      "epoch": 0.8430543815159199,
      "grad_norm": 0.15078510344028473,
      "learning_rate": 7.19017388904223e-05,
      "loss": 0.7471,
      "mean_token_accuracy": 0.7783933505415916,
      "num_tokens": 292446321.0,
      "step": 9163
    },
    {
      "entropy": 0.7189911454916,
      "epoch": 0.8431463878873612,
      "grad_norm": 0.1564454883337021,
      "learning_rate": 7.189867206428069e-05,
      "loss": 0.7183,
      "mean_token_accuracy": 0.7875682562589645,
      "num_tokens": 292478299.0,
      "step": 9164
    },
    {
      "entropy": 0.7024230156093836,
      "epoch": 0.8432383942588024,
      "grad_norm": 0.14831531047821045,
      "learning_rate": 7.189560523813905e-05,
      "loss": 0.7018,
      "mean_token_accuracy": 0.7926139272749424,
      "num_tokens": 292509887.0,
      "step": 9165
    },
    {
      "entropy": 0.6817587539553642,
      "epoch": 0.8433304006302437,
      "grad_norm": 0.16079320013523102,
      "learning_rate": 7.189253841199742e-05,
      "loss": 0.6752,
      "mean_token_accuracy": 0.7978444024920464,
      "num_tokens": 292542655.0,
      "step": 9166
    },
    {
      "entropy": 0.7140501905232668,
      "epoch": 0.8434224070016849,
      "grad_norm": 0.1615847796201706,
      "learning_rate": 7.18894715858558e-05,
      "loss": 0.6957,
      "mean_token_accuracy": 0.7915410958230495,
      "num_tokens": 292574690.0,
      "step": 9167
    },
    {
      "entropy": 0.714583944529295,
      "epoch": 0.843514413373126,
      "grad_norm": 0.15474404394626617,
      "learning_rate": 7.188640475971417e-05,
      "loss": 0.7085,
      "mean_token_accuracy": 0.7922707013785839,
      "num_tokens": 292607010.0,
      "step": 9168
    },
    {
      "entropy": 0.6972763650119305,
      "epoch": 0.8436064197445673,
      "grad_norm": 0.15226253867149353,
      "learning_rate": 7.188333793357255e-05,
      "loss": 0.6742,
      "mean_token_accuracy": 0.799869854003191,
      "num_tokens": 292638734.0,
      "step": 9169
    },
    {
      "entropy": 0.7433112189173698,
      "epoch": 0.8436984261160085,
      "grad_norm": 0.15649564564228058,
      "learning_rate": 7.188027110743092e-05,
      "loss": 0.7257,
      "mean_token_accuracy": 0.7860777266323566,
      "num_tokens": 292670088.0,
      "step": 9170
    },
    {
      "entropy": 0.7848769146949053,
      "epoch": 0.8437904324874498,
      "grad_norm": 0.15170952677726746,
      "learning_rate": 7.18772042812893e-05,
      "loss": 0.7732,
      "mean_token_accuracy": 0.7689503319561481,
      "num_tokens": 292702167.0,
      "step": 9171
    },
    {
      "entropy": 0.7587796999141574,
      "epoch": 0.843882438858891,
      "grad_norm": 0.1496972143650055,
      "learning_rate": 7.187413745514767e-05,
      "loss": 0.7477,
      "mean_token_accuracy": 0.7812251336872578,
      "num_tokens": 292734372.0,
      "step": 9172
    },
    {
      "entropy": 0.7783654741942883,
      "epoch": 0.8439744452303322,
      "grad_norm": 0.1524667590856552,
      "learning_rate": 7.187107062900604e-05,
      "loss": 0.7745,
      "mean_token_accuracy": 0.7744183391332626,
      "num_tokens": 292766538.0,
      "step": 9173
    },
    {
      "entropy": 0.8053816445171833,
      "epoch": 0.8440664516017734,
      "grad_norm": 0.16029632091522217,
      "learning_rate": 7.186800380286442e-05,
      "loss": 0.7924,
      "mean_token_accuracy": 0.7657095827162266,
      "num_tokens": 292798105.0,
      "step": 9174
    },
    {
      "entropy": 0.6884124297648668,
      "epoch": 0.8441584579732146,
      "grad_norm": 0.15505832433700562,
      "learning_rate": 7.18649369767228e-05,
      "loss": 0.6768,
      "mean_token_accuracy": 0.7969431504607201,
      "num_tokens": 292830508.0,
      "step": 9175
    },
    {
      "entropy": 0.7801714912056923,
      "epoch": 0.8442504643446559,
      "grad_norm": 0.15843702852725983,
      "learning_rate": 7.186187015058117e-05,
      "loss": 0.7614,
      "mean_token_accuracy": 0.7750273942947388,
      "num_tokens": 292863198.0,
      "step": 9176
    },
    {
      "entropy": 0.7092513535171747,
      "epoch": 0.8443424707160971,
      "grad_norm": 0.15084293484687805,
      "learning_rate": 7.185880332443954e-05,
      "loss": 0.7052,
      "mean_token_accuracy": 0.7908965721726418,
      "num_tokens": 292895221.0,
      "step": 9177
    },
    {
      "entropy": 0.6667038910090923,
      "epoch": 0.8444344770875383,
      "grad_norm": 0.16076156497001648,
      "learning_rate": 7.18557364982979e-05,
      "loss": 0.6615,
      "mean_token_accuracy": 0.8025096170604229,
      "num_tokens": 292927745.0,
      "step": 9178
    },
    {
      "entropy": 0.7156208157539368,
      "epoch": 0.8445264834589795,
      "grad_norm": 0.15464666485786438,
      "learning_rate": 7.185266967215629e-05,
      "loss": 0.7064,
      "mean_token_accuracy": 0.7906319051980972,
      "num_tokens": 292959867.0,
      "step": 9179
    },
    {
      "entropy": 0.7562061324715614,
      "epoch": 0.8446184898304208,
      "grad_norm": 0.1567748486995697,
      "learning_rate": 7.184960284601467e-05,
      "loss": 0.7583,
      "mean_token_accuracy": 0.7699157074093819,
      "num_tokens": 292991694.0,
      "step": 9180
    },
    {
      "entropy": 0.6906786076724529,
      "epoch": 0.844710496201862,
      "grad_norm": 0.1554396152496338,
      "learning_rate": 7.184653601987304e-05,
      "loss": 0.6751,
      "mean_token_accuracy": 0.8002968057990074,
      "num_tokens": 293023034.0,
      "step": 9181
    },
    {
      "entropy": 0.7586347591131926,
      "epoch": 0.8448025025733032,
      "grad_norm": 0.16306738555431366,
      "learning_rate": 7.18434691937314e-05,
      "loss": 0.7556,
      "mean_token_accuracy": 0.7768150754272938,
      "num_tokens": 293054438.0,
      "step": 9182
    },
    {
      "entropy": 0.635008642449975,
      "epoch": 0.8448945089447444,
      "grad_norm": 0.15982696413993835,
      "learning_rate": 7.184040236758978e-05,
      "loss": 0.624,
      "mean_token_accuracy": 0.815783154219389,
      "num_tokens": 293086881.0,
      "step": 9183
    },
    {
      "entropy": 0.7008657902479172,
      "epoch": 0.8449865153161856,
      "grad_norm": 0.14395911991596222,
      "learning_rate": 7.183733554144815e-05,
      "loss": 0.6954,
      "mean_token_accuracy": 0.7891611680388451,
      "num_tokens": 293119638.0,
      "step": 9184
    },
    {
      "entropy": 0.7335260156542063,
      "epoch": 0.8450785216876269,
      "grad_norm": 0.1568959355354309,
      "learning_rate": 7.183426871530653e-05,
      "loss": 0.7146,
      "mean_token_accuracy": 0.7843048386275768,
      "num_tokens": 293151287.0,
      "step": 9185
    },
    {
      "entropy": 0.780246414244175,
      "epoch": 0.8451705280590681,
      "grad_norm": 0.16298745572566986,
      "learning_rate": 7.183120188916492e-05,
      "loss": 0.7725,
      "mean_token_accuracy": 0.7741758935153484,
      "num_tokens": 293183965.0,
      "step": 9186
    },
    {
      "entropy": 0.6956896670162678,
      "epoch": 0.8452625344305094,
      "grad_norm": 0.14600618183612823,
      "learning_rate": 7.182813506302328e-05,
      "loss": 0.6812,
      "mean_token_accuracy": 0.799797210842371,
      "num_tokens": 293215603.0,
      "step": 9187
    },
    {
      "entropy": 0.7747749499976635,
      "epoch": 0.8453545408019505,
      "grad_norm": 0.16897286474704742,
      "learning_rate": 7.182506823688165e-05,
      "loss": 0.7541,
      "mean_token_accuracy": 0.7771247625350952,
      "num_tokens": 293247747.0,
      "step": 9188
    },
    {
      "entropy": 0.7749623879790306,
      "epoch": 0.8454465471733917,
      "grad_norm": 0.14899472892284393,
      "learning_rate": 7.182200141074002e-05,
      "loss": 0.7676,
      "mean_token_accuracy": 0.7753770686686039,
      "num_tokens": 293279830.0,
      "step": 9189
    },
    {
      "entropy": 0.7032394637353718,
      "epoch": 0.845538553544833,
      "grad_norm": 0.17058227956295013,
      "learning_rate": 7.181893458459841e-05,
      "loss": 0.6996,
      "mean_token_accuracy": 0.7961457781493664,
      "num_tokens": 293312001.0,
      "step": 9190
    },
    {
      "entropy": 0.6026732185855508,
      "epoch": 0.8456305599162742,
      "grad_norm": 0.15841256082057953,
      "learning_rate": 7.181586775845678e-05,
      "loss": 0.573,
      "mean_token_accuracy": 0.8272872976958752,
      "num_tokens": 293342572.0,
      "step": 9191
    },
    {
      "entropy": 0.7739413361996412,
      "epoch": 0.8457225662877155,
      "grad_norm": 0.16036087274551392,
      "learning_rate": 7.181280093231515e-05,
      "loss": 0.752,
      "mean_token_accuracy": 0.7801231928169727,
      "num_tokens": 293374683.0,
      "step": 9192
    },
    {
      "entropy": 0.8565878160297871,
      "epoch": 0.8458145726591566,
      "grad_norm": 0.15790407359600067,
      "learning_rate": 7.180973410617352e-05,
      "loss": 0.8495,
      "mean_token_accuracy": 0.7525311261415482,
      "num_tokens": 293406247.0,
      "step": 9193
    },
    {
      "entropy": 0.7613676395267248,
      "epoch": 0.8459065790305978,
      "grad_norm": 0.16311173141002655,
      "learning_rate": 7.18066672800319e-05,
      "loss": 0.7545,
      "mean_token_accuracy": 0.7755700945854187,
      "num_tokens": 293438064.0,
      "step": 9194
    },
    {
      "entropy": 0.7220560852438211,
      "epoch": 0.8459985854020391,
      "grad_norm": 0.1587047278881073,
      "learning_rate": 7.180360045389028e-05,
      "loss": 0.6937,
      "mean_token_accuracy": 0.7891317345201969,
      "num_tokens": 293469022.0,
      "step": 9195
    },
    {
      "entropy": 0.6695957779884338,
      "epoch": 0.8460905917734803,
      "grad_norm": 0.14823909103870392,
      "learning_rate": 7.180053362774865e-05,
      "loss": 0.6519,
      "mean_token_accuracy": 0.8060564361512661,
      "num_tokens": 293501229.0,
      "step": 9196
    },
    {
      "entropy": 0.7239779178053141,
      "epoch": 0.8461825981449216,
      "grad_norm": 0.15822318196296692,
      "learning_rate": 7.179746680160702e-05,
      "loss": 0.7184,
      "mean_token_accuracy": 0.7914208956062794,
      "num_tokens": 293533596.0,
      "step": 9197
    },
    {
      "entropy": 0.757901300676167,
      "epoch": 0.8462746045163627,
      "grad_norm": 0.15856708586215973,
      "learning_rate": 7.17943999754654e-05,
      "loss": 0.7521,
      "mean_token_accuracy": 0.7806506454944611,
      "num_tokens": 293565209.0,
      "step": 9198
    },
    {
      "entropy": 0.7164069004356861,
      "epoch": 0.846366610887804,
      "grad_norm": 0.15840736031532288,
      "learning_rate": 7.179133314932377e-05,
      "loss": 0.7248,
      "mean_token_accuracy": 0.7903875485062599,
      "num_tokens": 293597372.0,
      "step": 9199
    },
    {
      "entropy": 0.6900338064879179,
      "epoch": 0.8464586172592452,
      "grad_norm": 0.1546599417924881,
      "learning_rate": 7.178826632318215e-05,
      "loss": 0.6783,
      "mean_token_accuracy": 0.7990505546331406,
      "num_tokens": 293629283.0,
      "step": 9200
    },
    {
      "entropy": 0.7004851587116718,
      "epoch": 0.8465506236306864,
      "grad_norm": 0.14961956441402435,
      "learning_rate": 7.178519949704051e-05,
      "loss": 0.6976,
      "mean_token_accuracy": 0.7914286889135838,
      "num_tokens": 293661918.0,
      "step": 9201
    },
    {
      "entropy": 0.7097783256322145,
      "epoch": 0.8466426300021277,
      "grad_norm": 0.15354159474372864,
      "learning_rate": 7.17821326708989e-05,
      "loss": 0.7,
      "mean_token_accuracy": 0.7904226407408714,
      "num_tokens": 293694087.0,
      "step": 9202
    },
    {
      "entropy": 0.741591271944344,
      "epoch": 0.8467346363735688,
      "grad_norm": 0.15400955080986023,
      "learning_rate": 7.177906584475726e-05,
      "loss": 0.7103,
      "mean_token_accuracy": 0.7888648398220539,
      "num_tokens": 293724404.0,
      "step": 9203
    },
    {
      "entropy": 0.7345251608639956,
      "epoch": 0.8468266427450101,
      "grad_norm": 0.15243278443813324,
      "learning_rate": 7.177599901861563e-05,
      "loss": 0.7211,
      "mean_token_accuracy": 0.7900500409305096,
      "num_tokens": 293757044.0,
      "step": 9204
    },
    {
      "entropy": 0.7452695947140455,
      "epoch": 0.8469186491164513,
      "grad_norm": 0.15192413330078125,
      "learning_rate": 7.177293219247401e-05,
      "loss": 0.7316,
      "mean_token_accuracy": 0.7796895615756512,
      "num_tokens": 293788278.0,
      "step": 9205
    },
    {
      "entropy": 0.8226054962724447,
      "epoch": 0.8470106554878926,
      "grad_norm": 0.16392016410827637,
      "learning_rate": 7.17698653663324e-05,
      "loss": 0.8133,
      "mean_token_accuracy": 0.762117300182581,
      "num_tokens": 293820182.0,
      "step": 9206
    },
    {
      "entropy": 0.8894792981445789,
      "epoch": 0.8471026618593338,
      "grad_norm": 0.158228799700737,
      "learning_rate": 7.176679854019076e-05,
      "loss": 0.8844,
      "mean_token_accuracy": 0.7434817440807819,
      "num_tokens": 293852571.0,
      "step": 9207
    },
    {
      "entropy": 0.7162426970899105,
      "epoch": 0.8471946682307749,
      "grad_norm": 0.15706899762153625,
      "learning_rate": 7.176373171404913e-05,
      "loss": 0.7117,
      "mean_token_accuracy": 0.7904890291392803,
      "num_tokens": 293884041.0,
      "step": 9208
    },
    {
      "entropy": 0.7407329604029655,
      "epoch": 0.8472866746022162,
      "grad_norm": 0.1547851264476776,
      "learning_rate": 7.17606648879075e-05,
      "loss": 0.7184,
      "mean_token_accuracy": 0.7909626513719559,
      "num_tokens": 293915184.0,
      "step": 9209
    },
    {
      "entropy": 0.6786611713469028,
      "epoch": 0.8473786809736574,
      "grad_norm": 0.15295594930648804,
      "learning_rate": 7.175759806176588e-05,
      "loss": 0.6758,
      "mean_token_accuracy": 0.7944788783788681,
      "num_tokens": 293947105.0,
      "step": 9210
    },
    {
      "entropy": 0.7477181470021605,
      "epoch": 0.8474706873450987,
      "grad_norm": 0.1601785272359848,
      "learning_rate": 7.175453123562426e-05,
      "loss": 0.7402,
      "mean_token_accuracy": 0.7802557982504368,
      "num_tokens": 293978556.0,
      "step": 9211
    },
    {
      "entropy": 0.7973382696509361,
      "epoch": 0.8475626937165399,
      "grad_norm": 0.16166172921657562,
      "learning_rate": 7.175146440948263e-05,
      "loss": 0.8051,
      "mean_token_accuracy": 0.7658268734812737,
      "num_tokens": 294010068.0,
      "step": 9212
    },
    {
      "entropy": 0.8655599392950535,
      "epoch": 0.847654700087981,
      "grad_norm": 0.16238465905189514,
      "learning_rate": 7.174839758334101e-05,
      "loss": 0.8801,
      "mean_token_accuracy": 0.7446059957146645,
      "num_tokens": 294042233.0,
      "step": 9213
    },
    {
      "entropy": 0.6477555762976408,
      "epoch": 0.8477467064594223,
      "grad_norm": 0.1534823179244995,
      "learning_rate": 7.174533075719938e-05,
      "loss": 0.6409,
      "mean_token_accuracy": 0.8092668391764164,
      "num_tokens": 294074457.0,
      "step": 9214
    },
    {
      "entropy": 0.7489454736933112,
      "epoch": 0.8478387128308635,
      "grad_norm": 0.15878349542617798,
      "learning_rate": 7.174226393105775e-05,
      "loss": 0.74,
      "mean_token_accuracy": 0.7840565331280231,
      "num_tokens": 294105899.0,
      "step": 9215
    },
    {
      "entropy": 0.6779028549790382,
      "epoch": 0.8479307192023048,
      "grad_norm": 0.1509767472743988,
      "learning_rate": 7.173919710491613e-05,
      "loss": 0.6725,
      "mean_token_accuracy": 0.7995847463607788,
      "num_tokens": 294138667.0,
      "step": 9216
    },
    {
      "entropy": 0.739797854796052,
      "epoch": 0.848022725573746,
      "grad_norm": 0.1614757776260376,
      "learning_rate": 7.173613027877451e-05,
      "loss": 0.7082,
      "mean_token_accuracy": 0.7898979373276234,
      "num_tokens": 294170744.0,
      "step": 9217
    },
    {
      "entropy": 0.6172679867595434,
      "epoch": 0.8481147319451872,
      "grad_norm": 0.1549183428287506,
      "learning_rate": 7.173306345263288e-05,
      "loss": 0.6059,
      "mean_token_accuracy": 0.819999560713768,
      "num_tokens": 294202989.0,
      "step": 9218
    },
    {
      "entropy": 0.6989898486062884,
      "epoch": 0.8482067383166284,
      "grad_norm": 0.1525004506111145,
      "learning_rate": 7.172999662649124e-05,
      "loss": 0.6895,
      "mean_token_accuracy": 0.7935312762856483,
      "num_tokens": 294235088.0,
      "step": 9219
    },
    {
      "entropy": 0.8720769342035055,
      "epoch": 0.8482987446880697,
      "grad_norm": 0.16067087650299072,
      "learning_rate": 7.172692980034961e-05,
      "loss": 0.8566,
      "mean_token_accuracy": 0.7469127289950848,
      "num_tokens": 294267601.0,
      "step": 9220
    },
    {
      "entropy": 0.7055386696010828,
      "epoch": 0.8483907510595109,
      "grad_norm": 0.16139115393161774,
      "learning_rate": 7.1723862974208e-05,
      "loss": 0.6876,
      "mean_token_accuracy": 0.7937961295247078,
      "num_tokens": 294298909.0,
      "step": 9221
    },
    {
      "entropy": 0.754301531240344,
      "epoch": 0.8484827574309521,
      "grad_norm": 0.1617162972688675,
      "learning_rate": 7.172079614806638e-05,
      "loss": 0.7479,
      "mean_token_accuracy": 0.7778256572782993,
      "num_tokens": 294330151.0,
      "step": 9222
    },
    {
      "entropy": 0.7079686895012856,
      "epoch": 0.8485747638023933,
      "grad_norm": 0.14493924379348755,
      "learning_rate": 7.171772932192474e-05,
      "loss": 0.6809,
      "mean_token_accuracy": 0.7959273457527161,
      "num_tokens": 294361846.0,
      "step": 9223
    },
    {
      "entropy": 0.7169918306171894,
      "epoch": 0.8486667701738345,
      "grad_norm": 0.1569448560476303,
      "learning_rate": 7.171466249578311e-05,
      "loss": 0.7105,
      "mean_token_accuracy": 0.791932437568903,
      "num_tokens": 294394010.0,
      "step": 9224
    },
    {
      "entropy": 0.5775814140215516,
      "epoch": 0.8487587765452758,
      "grad_norm": 0.14086732268333435,
      "learning_rate": 7.171159566964149e-05,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8286913707852364,
      "num_tokens": 294425757.0,
      "step": 9225
    },
    {
      "entropy": 0.6783013325184584,
      "epoch": 0.848850782916717,
      "grad_norm": 0.15347783267498016,
      "learning_rate": 7.170852884349986e-05,
      "loss": 0.6786,
      "mean_token_accuracy": 0.7930466346442699,
      "num_tokens": 294457949.0,
      "step": 9226
    },
    {
      "entropy": 0.7157650254666805,
      "epoch": 0.8489427892881583,
      "grad_norm": 0.15602463483810425,
      "learning_rate": 7.170546201735824e-05,
      "loss": 0.7132,
      "mean_token_accuracy": 0.7869654111564159,
      "num_tokens": 294489631.0,
      "step": 9227
    },
    {
      "entropy": 0.796334408223629,
      "epoch": 0.8490347956595994,
      "grad_norm": 0.15323390066623688,
      "learning_rate": 7.170239519121661e-05,
      "loss": 0.7835,
      "mean_token_accuracy": 0.769356083124876,
      "num_tokens": 294522010.0,
      "step": 9228
    },
    {
      "entropy": 0.7483775354921818,
      "epoch": 0.8491268020310406,
      "grad_norm": 0.1554114818572998,
      "learning_rate": 7.169932836507499e-05,
      "loss": 0.7401,
      "mean_token_accuracy": 0.7841345146298409,
      "num_tokens": 294554351.0,
      "step": 9229
    },
    {
      "entropy": 0.7027828544378281,
      "epoch": 0.8492188084024819,
      "grad_norm": 0.15870484709739685,
      "learning_rate": 7.169626153893336e-05,
      "loss": 0.6685,
      "mean_token_accuracy": 0.7998899780213833,
      "num_tokens": 294586382.0,
      "step": 9230
    },
    {
      "entropy": 0.5796954156830907,
      "epoch": 0.8493108147739231,
      "grad_norm": 0.1509189009666443,
      "learning_rate": 7.169319471279173e-05,
      "loss": 0.5654,
      "mean_token_accuracy": 0.8270756639540195,
      "num_tokens": 294618941.0,
      "step": 9231
    },
    {
      "entropy": 0.781430296599865,
      "epoch": 0.8494028211453644,
      "grad_norm": 0.16551601886749268,
      "learning_rate": 7.169012788665011e-05,
      "loss": 0.7618,
      "mean_token_accuracy": 0.776944775134325,
      "num_tokens": 294651423.0,
      "step": 9232
    },
    {
      "entropy": 0.7833967749029398,
      "epoch": 0.8494948275168055,
      "grad_norm": 0.16358345746994019,
      "learning_rate": 7.168706106050849e-05,
      "loss": 0.7835,
      "mean_token_accuracy": 0.7671772055327892,
      "num_tokens": 294683499.0,
      "step": 9233
    },
    {
      "entropy": 0.7059527970850468,
      "epoch": 0.8495868338882467,
      "grad_norm": 0.15662719309329987,
      "learning_rate": 7.168399423436686e-05,
      "loss": 0.6956,
      "mean_token_accuracy": 0.7943080626428127,
      "num_tokens": 294715362.0,
      "step": 9234
    },
    {
      "entropy": 0.5616456512361765,
      "epoch": 0.849678840259688,
      "grad_norm": 0.1501058042049408,
      "learning_rate": 7.168092740822523e-05,
      "loss": 0.5292,
      "mean_token_accuracy": 0.8450856730341911,
      "num_tokens": 294747523.0,
      "step": 9235
    },
    {
      "entropy": 0.9459944777190685,
      "epoch": 0.8497708466311292,
      "grad_norm": 0.16790860891342163,
      "learning_rate": 7.16778605820836e-05,
      "loss": 0.9577,
      "mean_token_accuracy": 0.7286823466420174,
      "num_tokens": 294778826.0,
      "step": 9236
    },
    {
      "entropy": 0.6862558834254742,
      "epoch": 0.8498628530025705,
      "grad_norm": 0.1446121782064438,
      "learning_rate": 7.167479375594199e-05,
      "loss": 0.6548,
      "mean_token_accuracy": 0.8032681532204151,
      "num_tokens": 294810755.0,
      "step": 9237
    },
    {
      "entropy": 0.6454623285681009,
      "epoch": 0.8499548593740116,
      "grad_norm": 0.16041584312915802,
      "learning_rate": 7.167172692980036e-05,
      "loss": 0.6511,
      "mean_token_accuracy": 0.809081431478262,
      "num_tokens": 294843454.0,
      "step": 9238
    },
    {
      "entropy": 0.7591703534126282,
      "epoch": 0.8500468657454529,
      "grad_norm": 0.15679171681404114,
      "learning_rate": 7.166866010365872e-05,
      "loss": 0.7436,
      "mean_token_accuracy": 0.7814613804221153,
      "num_tokens": 294874727.0,
      "step": 9239
    },
    {
      "entropy": 0.6383411642163992,
      "epoch": 0.8501388721168941,
      "grad_norm": 0.15572838485240936,
      "learning_rate": 7.166559327751709e-05,
      "loss": 0.6067,
      "mean_token_accuracy": 0.82059196382761,
      "num_tokens": 294905749.0,
      "step": 9240
    },
    {
      "entropy": 0.7301694918423891,
      "epoch": 0.8502308784883353,
      "grad_norm": 0.15705886483192444,
      "learning_rate": 7.166252645137547e-05,
      "loss": 0.7166,
      "mean_token_accuracy": 0.7865882441401482,
      "num_tokens": 294937795.0,
      "step": 9241
    },
    {
      "entropy": 0.7536682542413473,
      "epoch": 0.8503228848597766,
      "grad_norm": 0.14913347363471985,
      "learning_rate": 7.165945962523385e-05,
      "loss": 0.7223,
      "mean_token_accuracy": 0.7870372459292412,
      "num_tokens": 294970446.0,
      "step": 9242
    },
    {
      "entropy": 0.7899387031793594,
      "epoch": 0.8504148912312177,
      "grad_norm": 0.15425072610378265,
      "learning_rate": 7.165639279909222e-05,
      "loss": 0.7946,
      "mean_token_accuracy": 0.7616885751485825,
      "num_tokens": 295002270.0,
      "step": 9243
    },
    {
      "entropy": 0.8850183393806219,
      "epoch": 0.850506897602659,
      "grad_norm": 0.16105367243289948,
      "learning_rate": 7.16533259729506e-05,
      "loss": 0.8826,
      "mean_token_accuracy": 0.74454340711236,
      "num_tokens": 295034421.0,
      "step": 9244
    },
    {
      "entropy": 0.7768974304199219,
      "epoch": 0.8505989039741002,
      "grad_norm": 0.15736828744411469,
      "learning_rate": 7.165025914680897e-05,
      "loss": 0.7578,
      "mean_token_accuracy": 0.77474220469594,
      "num_tokens": 295066990.0,
      "step": 9245
    },
    {
      "entropy": 0.6810788493603468,
      "epoch": 0.8506909103455415,
      "grad_norm": 0.15527667105197906,
      "learning_rate": 7.164719232066734e-05,
      "loss": 0.6708,
      "mean_token_accuracy": 0.7998138964176178,
      "num_tokens": 295099623.0,
      "step": 9246
    },
    {
      "entropy": 0.7248140070587397,
      "epoch": 0.8507829167169827,
      "grad_norm": 0.16763654351234436,
      "learning_rate": 7.164412549452572e-05,
      "loss": 0.7444,
      "mean_token_accuracy": 0.781427763402462,
      "num_tokens": 295132121.0,
      "step": 9247
    },
    {
      "entropy": 0.7623528596013784,
      "epoch": 0.8508749230884238,
      "grad_norm": 0.15902139246463776,
      "learning_rate": 7.16410586683841e-05,
      "loss": 0.7576,
      "mean_token_accuracy": 0.776562511920929,
      "num_tokens": 295163978.0,
      "step": 9248
    },
    {
      "entropy": 0.6874351017177105,
      "epoch": 0.8509669294598651,
      "grad_norm": 0.16207319498062134,
      "learning_rate": 7.163799184224247e-05,
      "loss": 0.6808,
      "mean_token_accuracy": 0.8005663715302944,
      "num_tokens": 295196126.0,
      "step": 9249
    },
    {
      "entropy": 0.749596981331706,
      "epoch": 0.8510589358313063,
      "grad_norm": 0.16014590859413147,
      "learning_rate": 7.163492501610084e-05,
      "loss": 0.7521,
      "mean_token_accuracy": 0.7778571136295795,
      "num_tokens": 295228463.0,
      "step": 9250
    },
    {
      "entropy": 0.7579569686204195,
      "epoch": 0.8511509422027476,
      "grad_norm": 0.15806153416633606,
      "learning_rate": 7.16318581899592e-05,
      "loss": 0.753,
      "mean_token_accuracy": 0.7787609025835991,
      "num_tokens": 295259871.0,
      "step": 9251
    },
    {
      "entropy": 0.6719326321035624,
      "epoch": 0.8512429485741888,
      "grad_norm": 0.15345031023025513,
      "learning_rate": 7.162879136381759e-05,
      "loss": 0.6571,
      "mean_token_accuracy": 0.8050297535955906,
      "num_tokens": 295291724.0,
      "step": 9252
    },
    {
      "entropy": 0.8009656332433224,
      "epoch": 0.8513349549456299,
      "grad_norm": 0.1620047241449356,
      "learning_rate": 7.162572453767597e-05,
      "loss": 0.7865,
      "mean_token_accuracy": 0.7740407884120941,
      "num_tokens": 295323233.0,
      "step": 9253
    },
    {
      "entropy": 0.6484742127358913,
      "epoch": 0.8514269613170712,
      "grad_norm": 0.15092961490154266,
      "learning_rate": 7.162265771153434e-05,
      "loss": 0.6225,
      "mean_token_accuracy": 0.8164214119315147,
      "num_tokens": 295355382.0,
      "step": 9254
    },
    {
      "entropy": 0.7230396438390017,
      "epoch": 0.8515189676885124,
      "grad_norm": 0.15424619615077972,
      "learning_rate": 7.16195908853927e-05,
      "loss": 0.7,
      "mean_token_accuracy": 0.7940858714282513,
      "num_tokens": 295387453.0,
      "step": 9255
    },
    {
      "entropy": 0.691457667388022,
      "epoch": 0.8516109740599537,
      "grad_norm": 0.1491248905658722,
      "learning_rate": 7.161652405925109e-05,
      "loss": 0.6775,
      "mean_token_accuracy": 0.7946472838521004,
      "num_tokens": 295419534.0,
      "step": 9256
    },
    {
      "entropy": 0.6256690109148622,
      "epoch": 0.8517029804313949,
      "grad_norm": 0.14928953349590302,
      "learning_rate": 7.161345723310945e-05,
      "loss": 0.6065,
      "mean_token_accuracy": 0.8230440691113472,
      "num_tokens": 295451606.0,
      "step": 9257
    },
    {
      "entropy": 0.6933057215064764,
      "epoch": 0.851794986802836,
      "grad_norm": 0.16009092330932617,
      "learning_rate": 7.161039040696784e-05,
      "loss": 0.6912,
      "mean_token_accuracy": 0.7939947471022606,
      "num_tokens": 295483788.0,
      "step": 9258
    },
    {
      "entropy": 0.5817230446264148,
      "epoch": 0.8518869931742773,
      "grad_norm": 0.14524132013320923,
      "learning_rate": 7.16073235808262e-05,
      "loss": 0.5614,
      "mean_token_accuracy": 0.8316090516746044,
      "num_tokens": 295516417.0,
      "step": 9259
    },
    {
      "entropy": 0.7339549222961068,
      "epoch": 0.8519789995457185,
      "grad_norm": 0.15771357715129852,
      "learning_rate": 7.160425675468459e-05,
      "loss": 0.7215,
      "mean_token_accuracy": 0.7868035212159157,
      "num_tokens": 295548632.0,
      "step": 9260
    },
    {
      "entropy": 0.6919783968478441,
      "epoch": 0.8520710059171598,
      "grad_norm": 0.14992010593414307,
      "learning_rate": 7.160118992854295e-05,
      "loss": 0.6974,
      "mean_token_accuracy": 0.7997874282300472,
      "num_tokens": 295580414.0,
      "step": 9261
    },
    {
      "entropy": 0.7695006765425205,
      "epoch": 0.852163012288601,
      "grad_norm": 0.15162672102451324,
      "learning_rate": 7.159812310240132e-05,
      "loss": 0.7702,
      "mean_token_accuracy": 0.7708836980164051,
      "num_tokens": 295612548.0,
      "step": 9262
    },
    {
      "entropy": 0.6566467825323343,
      "epoch": 0.8522550186600422,
      "grad_norm": 0.16399939358234406,
      "learning_rate": 7.15950562762597e-05,
      "loss": 0.6397,
      "mean_token_accuracy": 0.8097750842571259,
      "num_tokens": 295644055.0,
      "step": 9263
    },
    {
      "entropy": 0.7831311319023371,
      "epoch": 0.8523470250314834,
      "grad_norm": 0.15169775485992432,
      "learning_rate": 7.159198945011808e-05,
      "loss": 0.7795,
      "mean_token_accuracy": 0.7710037939250469,
      "num_tokens": 295676543.0,
      "step": 9264
    },
    {
      "entropy": 0.8629758954048157,
      "epoch": 0.8524390314029247,
      "grad_norm": 0.15710625052452087,
      "learning_rate": 7.158892262397645e-05,
      "loss": 0.8738,
      "mean_token_accuracy": 0.7468511760234833,
      "num_tokens": 295708711.0,
      "step": 9265
    },
    {
      "entropy": 0.5679602976888418,
      "epoch": 0.8525310377743659,
      "grad_norm": 0.1501118540763855,
      "learning_rate": 7.158585579783482e-05,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8323536403477192,
      "num_tokens": 295741165.0,
      "step": 9266
    },
    {
      "entropy": 0.6966983992606401,
      "epoch": 0.8526230441458071,
      "grad_norm": 0.15750013291835785,
      "learning_rate": 7.158278897169319e-05,
      "loss": 0.6892,
      "mean_token_accuracy": 0.7943028844892979,
      "num_tokens": 295773739.0,
      "step": 9267
    },
    {
      "entropy": 0.7111820443533361,
      "epoch": 0.8527150505172483,
      "grad_norm": 0.16041125357151031,
      "learning_rate": 7.157972214555157e-05,
      "loss": 0.7089,
      "mean_token_accuracy": 0.7973058894276619,
      "num_tokens": 295806237.0,
      "step": 9268
    },
    {
      "entropy": 0.7450157999992371,
      "epoch": 0.8528070568886895,
      "grad_norm": 0.1591140180826187,
      "learning_rate": 7.157665531940995e-05,
      "loss": 0.7288,
      "mean_token_accuracy": 0.7871914654970169,
      "num_tokens": 295838060.0,
      "step": 9269
    },
    {
      "entropy": 0.7893609739840031,
      "epoch": 0.8528990632601308,
      "grad_norm": 0.153928741812706,
      "learning_rate": 7.157358849326832e-05,
      "loss": 0.7618,
      "mean_token_accuracy": 0.7739544101059437,
      "num_tokens": 295870126.0,
      "step": 9270
    },
    {
      "entropy": 0.6509202048182487,
      "epoch": 0.852991069631572,
      "grad_norm": 0.14922577142715454,
      "learning_rate": 7.157052166712669e-05,
      "loss": 0.635,
      "mean_token_accuracy": 0.8083102628588676,
      "num_tokens": 295902491.0,
      "step": 9271
    },
    {
      "entropy": 0.7199104931205511,
      "epoch": 0.8530830760030133,
      "grad_norm": 0.16797427833080292,
      "learning_rate": 7.156745484098507e-05,
      "loss": 0.6947,
      "mean_token_accuracy": 0.7988784722983837,
      "num_tokens": 295934066.0,
      "step": 9272
    },
    {
      "entropy": 0.6855895835906267,
      "epoch": 0.8531750823744544,
      "grad_norm": 0.1487516164779663,
      "learning_rate": 7.156438801484345e-05,
      "loss": 0.6674,
      "mean_token_accuracy": 0.7969290241599083,
      "num_tokens": 295965448.0,
      "step": 9273
    },
    {
      "entropy": 0.7808198388665915,
      "epoch": 0.8532670887458956,
      "grad_norm": 0.17414379119873047,
      "learning_rate": 7.156132118870182e-05,
      "loss": 0.7968,
      "mean_token_accuracy": 0.7671060785651207,
      "num_tokens": 295997469.0,
      "step": 9274
    },
    {
      "entropy": 0.6734881959855556,
      "epoch": 0.8533590951173369,
      "grad_norm": 0.16426751017570496,
      "learning_rate": 7.15582543625602e-05,
      "loss": 0.6638,
      "mean_token_accuracy": 0.8001768924295902,
      "num_tokens": 296029407.0,
      "step": 9275
    },
    {
      "entropy": 0.7721558157354593,
      "epoch": 0.8534511014887781,
      "grad_norm": 0.15801751613616943,
      "learning_rate": 7.155518753641857e-05,
      "loss": 0.775,
      "mean_token_accuracy": 0.7685927860438824,
      "num_tokens": 296061382.0,
      "step": 9276
    },
    {
      "entropy": 0.7388074900954962,
      "epoch": 0.8535431078602194,
      "grad_norm": 0.1522987186908722,
      "learning_rate": 7.155212071027693e-05,
      "loss": 0.7131,
      "mean_token_accuracy": 0.7860884107649326,
      "num_tokens": 296093806.0,
      "step": 9277
    },
    {
      "entropy": 0.6497070249170065,
      "epoch": 0.8536351142316605,
      "grad_norm": 0.150343656539917,
      "learning_rate": 7.154905388413532e-05,
      "loss": 0.6276,
      "mean_token_accuracy": 0.8116071000695229,
      "num_tokens": 296125728.0,
      "step": 9278
    },
    {
      "entropy": 0.758166566491127,
      "epoch": 0.8537271206031017,
      "grad_norm": 0.15080419182777405,
      "learning_rate": 7.15459870579937e-05,
      "loss": 0.743,
      "mean_token_accuracy": 0.7800386287271976,
      "num_tokens": 296158097.0,
      "step": 9279
    },
    {
      "entropy": 0.8311229813843966,
      "epoch": 0.853819126974543,
      "grad_norm": 0.16156117618083954,
      "learning_rate": 7.154292023185206e-05,
      "loss": 0.8152,
      "mean_token_accuracy": 0.7652864903211594,
      "num_tokens": 296189649.0,
      "step": 9280
    },
    {
      "entropy": 0.6889754477888346,
      "epoch": 0.8539111333459842,
      "grad_norm": 0.14824868738651276,
      "learning_rate": 7.153985340571043e-05,
      "loss": 0.6763,
      "mean_token_accuracy": 0.7970256134867668,
      "num_tokens": 296222065.0,
      "step": 9281
    },
    {
      "entropy": 0.7700135577470064,
      "epoch": 0.8540031397174255,
      "grad_norm": 0.15271902084350586,
      "learning_rate": 7.15367865795688e-05,
      "loss": 0.7547,
      "mean_token_accuracy": 0.7767580300569534,
      "num_tokens": 296253660.0,
      "step": 9282
    },
    {
      "entropy": 0.7261857818812132,
      "epoch": 0.8540951460888666,
      "grad_norm": 0.15626661479473114,
      "learning_rate": 7.153371975342718e-05,
      "loss": 0.713,
      "mean_token_accuracy": 0.7889329828321934,
      "num_tokens": 296283566.0,
      "step": 9283
    },
    {
      "entropy": 0.7403159290552139,
      "epoch": 0.8541871524603079,
      "grad_norm": 0.14912141859531403,
      "learning_rate": 7.153065292728556e-05,
      "loss": 0.7177,
      "mean_token_accuracy": 0.784930158406496,
      "num_tokens": 296315486.0,
      "step": 9284
    },
    {
      "entropy": 0.597387183457613,
      "epoch": 0.8542791588317491,
      "grad_norm": 0.15372668206691742,
      "learning_rate": 7.152758610114393e-05,
      "loss": 0.5865,
      "mean_token_accuracy": 0.8229139111936092,
      "num_tokens": 296347417.0,
      "step": 9285
    },
    {
      "entropy": 0.7338540926575661,
      "epoch": 0.8543711652031903,
      "grad_norm": 0.15401272475719452,
      "learning_rate": 7.15245192750023e-05,
      "loss": 0.7348,
      "mean_token_accuracy": 0.7827333174645901,
      "num_tokens": 296379020.0,
      "step": 9286
    },
    {
      "entropy": 0.7744400314986706,
      "epoch": 0.8544631715746316,
      "grad_norm": 0.15605783462524414,
      "learning_rate": 7.152145244886068e-05,
      "loss": 0.7674,
      "mean_token_accuracy": 0.7725348919630051,
      "num_tokens": 296411381.0,
      "step": 9287
    },
    {
      "entropy": 0.7000164426863194,
      "epoch": 0.8545551779460727,
      "grad_norm": 0.1573788821697235,
      "learning_rate": 7.151838562271905e-05,
      "loss": 0.678,
      "mean_token_accuracy": 0.7972492650151253,
      "num_tokens": 296443775.0,
      "step": 9288
    },
    {
      "entropy": 0.5995199102908373,
      "epoch": 0.854647184317514,
      "grad_norm": 0.146518275141716,
      "learning_rate": 7.151531879657743e-05,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8261918239295483,
      "num_tokens": 296475700.0,
      "step": 9289
    },
    {
      "entropy": 0.671977061778307,
      "epoch": 0.8547391906889552,
      "grad_norm": 0.15351417660713196,
      "learning_rate": 7.15122519704358e-05,
      "loss": 0.6654,
      "mean_token_accuracy": 0.8041447624564171,
      "num_tokens": 296507587.0,
      "step": 9290
    },
    {
      "entropy": 0.7320583015680313,
      "epoch": 0.8548311970603965,
      "grad_norm": 0.15575651824474335,
      "learning_rate": 7.150918514429418e-05,
      "loss": 0.7102,
      "mean_token_accuracy": 0.7875051312148571,
      "num_tokens": 296538690.0,
      "step": 9291
    },
    {
      "entropy": 0.6746419947594404,
      "epoch": 0.8549232034318377,
      "grad_norm": 0.16007113456726074,
      "learning_rate": 7.150611831815255e-05,
      "loss": 0.659,
      "mean_token_accuracy": 0.8036817871034145,
      "num_tokens": 296570488.0,
      "step": 9292
    },
    {
      "entropy": 0.7095679063349962,
      "epoch": 0.8550152098032788,
      "grad_norm": 0.1506062000989914,
      "learning_rate": 7.150305149201091e-05,
      "loss": 0.7039,
      "mean_token_accuracy": 0.7885853685438633,
      "num_tokens": 296602368.0,
      "step": 9293
    },
    {
      "entropy": 0.7157891988754272,
      "epoch": 0.8551072161747201,
      "grad_norm": 0.16648705303668976,
      "learning_rate": 7.14999846658693e-05,
      "loss": 0.719,
      "mean_token_accuracy": 0.7912423275411129,
      "num_tokens": 296634538.0,
      "step": 9294
    },
    {
      "entropy": 0.6871769111603498,
      "epoch": 0.8551992225461613,
      "grad_norm": 0.1517302691936493,
      "learning_rate": 7.149691783972768e-05,
      "loss": 0.6693,
      "mean_token_accuracy": 0.8038580641150475,
      "num_tokens": 296666072.0,
      "step": 9295
    },
    {
      "entropy": 0.6385317072272301,
      "epoch": 0.8552912289176026,
      "grad_norm": 0.16230399906635284,
      "learning_rate": 7.149385101358605e-05,
      "loss": 0.6326,
      "mean_token_accuracy": 0.8112977407872677,
      "num_tokens": 296697392.0,
      "step": 9296
    },
    {
      "entropy": 0.6222475552931428,
      "epoch": 0.8553832352890438,
      "grad_norm": 0.1588875651359558,
      "learning_rate": 7.149078418744441e-05,
      "loss": 0.6018,
      "mean_token_accuracy": 0.8197043016552925,
      "num_tokens": 296729589.0,
      "step": 9297
    },
    {
      "entropy": 0.675569910556078,
      "epoch": 0.8554752416604849,
      "grad_norm": 0.1537511944770813,
      "learning_rate": 7.148771736130278e-05,
      "loss": 0.6678,
      "mean_token_accuracy": 0.8056351952254772,
      "num_tokens": 296762232.0,
      "step": 9298
    },
    {
      "entropy": 0.7032335121184587,
      "epoch": 0.8555672480319262,
      "grad_norm": 0.15394003689289093,
      "learning_rate": 7.148465053516116e-05,
      "loss": 0.6881,
      "mean_token_accuracy": 0.7948860339820385,
      "num_tokens": 296794249.0,
      "step": 9299
    },
    {
      "entropy": 0.6535628940910101,
      "epoch": 0.8556592544033674,
      "grad_norm": 0.165849506855011,
      "learning_rate": 7.148158370901954e-05,
      "loss": 0.6368,
      "mean_token_accuracy": 0.8083375096321106,
      "num_tokens": 296826576.0,
      "step": 9300
    },
    {
      "entropy": 0.781649649143219,
      "epoch": 0.8557512607748087,
      "grad_norm": 0.15680566430091858,
      "learning_rate": 7.147851688287791e-05,
      "loss": 0.7747,
      "mean_token_accuracy": 0.7753229737281799,
      "num_tokens": 296858490.0,
      "step": 9301
    },
    {
      "entropy": 0.6535873916000128,
      "epoch": 0.8558432671462499,
      "grad_norm": 0.15006360411643982,
      "learning_rate": 7.147545005673628e-05,
      "loss": 0.6499,
      "mean_token_accuracy": 0.8052450530230999,
      "num_tokens": 296890838.0,
      "step": 9302
    },
    {
      "entropy": 0.7341656144708395,
      "epoch": 0.855935273517691,
      "grad_norm": 0.14776501059532166,
      "learning_rate": 7.147238323059466e-05,
      "loss": 0.7303,
      "mean_token_accuracy": 0.7858513966202736,
      "num_tokens": 296922041.0,
      "step": 9303
    },
    {
      "entropy": 0.7369539747014642,
      "epoch": 0.8560272798891323,
      "grad_norm": 0.15657730400562286,
      "learning_rate": 7.146931640445303e-05,
      "loss": 0.7234,
      "mean_token_accuracy": 0.782026369124651,
      "num_tokens": 296953915.0,
      "step": 9304
    },
    {
      "entropy": 0.6443697698414326,
      "epoch": 0.8561192862605735,
      "grad_norm": 0.1637953370809555,
      "learning_rate": 7.146624957831141e-05,
      "loss": 0.63,
      "mean_token_accuracy": 0.8153681457042694,
      "num_tokens": 296985803.0,
      "step": 9305
    },
    {
      "entropy": 0.7375577334314585,
      "epoch": 0.8562112926320148,
      "grad_norm": 0.1566283106803894,
      "learning_rate": 7.146318275216979e-05,
      "loss": 0.7277,
      "mean_token_accuracy": 0.779588907957077,
      "num_tokens": 297018058.0,
      "step": 9306
    },
    {
      "entropy": 0.6474961824715137,
      "epoch": 0.856303299003456,
      "grad_norm": 0.1541268229484558,
      "learning_rate": 7.146011592602816e-05,
      "loss": 0.6225,
      "mean_token_accuracy": 0.8131704144179821,
      "num_tokens": 297049941.0,
      "step": 9307
    },
    {
      "entropy": 0.6624164399690926,
      "epoch": 0.8563953053748972,
      "grad_norm": 0.1625298261642456,
      "learning_rate": 7.145704909988653e-05,
      "loss": 0.6429,
      "mean_token_accuracy": 0.8085928037762642,
      "num_tokens": 297081206.0,
      "step": 9308
    },
    {
      "entropy": 0.7358057871460915,
      "epoch": 0.8564873117463384,
      "grad_norm": 0.16583764553070068,
      "learning_rate": 7.14539822737449e-05,
      "loss": 0.7238,
      "mean_token_accuracy": 0.7829283513128757,
      "num_tokens": 297111810.0,
      "step": 9309
    },
    {
      "entropy": 0.8249125834554434,
      "epoch": 0.8565793181177797,
      "grad_norm": 0.159418523311615,
      "learning_rate": 7.145091544760329e-05,
      "loss": 0.8297,
      "mean_token_accuracy": 0.7627994902431965,
      "num_tokens": 297144039.0,
      "step": 9310
    },
    {
      "entropy": 0.7495959587395191,
      "epoch": 0.8566713244892209,
      "grad_norm": 0.15243613719940186,
      "learning_rate": 7.144784862146166e-05,
      "loss": 0.7445,
      "mean_token_accuracy": 0.7796294949948788,
      "num_tokens": 297175922.0,
      "step": 9311
    },
    {
      "entropy": 0.66484734416008,
      "epoch": 0.8567633308606621,
      "grad_norm": 0.14784260094165802,
      "learning_rate": 7.144478179532003e-05,
      "loss": 0.6458,
      "mean_token_accuracy": 0.8074377216398716,
      "num_tokens": 297207862.0,
      "step": 9312
    },
    {
      "entropy": 0.524963092058897,
      "epoch": 0.8568553372321033,
      "grad_norm": 0.14393942058086395,
      "learning_rate": 7.14417149691784e-05,
      "loss": 0.4977,
      "mean_token_accuracy": 0.8467649072408676,
      "num_tokens": 297239881.0,
      "step": 9313
    },
    {
      "entropy": 0.7670757714658976,
      "epoch": 0.8569473436035445,
      "grad_norm": 0.16007854044437408,
      "learning_rate": 7.143864814303678e-05,
      "loss": 0.7652,
      "mean_token_accuracy": 0.7755994573235512,
      "num_tokens": 297272133.0,
      "step": 9314
    },
    {
      "entropy": 0.788808261975646,
      "epoch": 0.8570393499749858,
      "grad_norm": 0.16021373867988586,
      "learning_rate": 7.143558131689516e-05,
      "loss": 0.7935,
      "mean_token_accuracy": 0.7649881131947041,
      "num_tokens": 297302834.0,
      "step": 9315
    },
    {
      "entropy": 0.7095064520835876,
      "epoch": 0.857131356346427,
      "grad_norm": 0.15681712329387665,
      "learning_rate": 7.143251449075352e-05,
      "loss": 0.6946,
      "mean_token_accuracy": 0.7923196665942669,
      "num_tokens": 297334593.0,
      "step": 9316
    },
    {
      "entropy": 0.7174708247184753,
      "epoch": 0.8572233627178683,
      "grad_norm": 0.14910614490509033,
      "learning_rate": 7.142944766461189e-05,
      "loss": 0.6964,
      "mean_token_accuracy": 0.7940537966787815,
      "num_tokens": 297367193.0,
      "step": 9317
    },
    {
      "entropy": 0.799740944057703,
      "epoch": 0.8573153690893094,
      "grad_norm": 0.15455390512943268,
      "learning_rate": 7.142638083847027e-05,
      "loss": 0.8013,
      "mean_token_accuracy": 0.7644482888281345,
      "num_tokens": 297398588.0,
      "step": 9318
    },
    {
      "entropy": 0.6612803675234318,
      "epoch": 0.8574073754607506,
      "grad_norm": 0.150625541806221,
      "learning_rate": 7.142331401232864e-05,
      "loss": 0.6372,
      "mean_token_accuracy": 0.8108151815831661,
      "num_tokens": 297430719.0,
      "step": 9319
    },
    {
      "entropy": 0.6725093647837639,
      "epoch": 0.8574993818321919,
      "grad_norm": 0.14883765578269958,
      "learning_rate": 7.142024718618702e-05,
      "loss": 0.657,
      "mean_token_accuracy": 0.8061027936637402,
      "num_tokens": 297463444.0,
      "step": 9320
    },
    {
      "entropy": 0.7818607799708843,
      "epoch": 0.8575913882036331,
      "grad_norm": 0.15874233841896057,
      "learning_rate": 7.141718036004539e-05,
      "loss": 0.7717,
      "mean_token_accuracy": 0.767365001142025,
      "num_tokens": 297494937.0,
      "step": 9321
    },
    {
      "entropy": 0.8251926526427269,
      "epoch": 0.8576833945750744,
      "grad_norm": 0.16131265461444855,
      "learning_rate": 7.141411353390377e-05,
      "loss": 0.8204,
      "mean_token_accuracy": 0.7624752931296825,
      "num_tokens": 297527260.0,
      "step": 9322
    },
    {
      "entropy": 0.6777688665315509,
      "epoch": 0.8577754009465155,
      "grad_norm": 0.15160194039344788,
      "learning_rate": 7.141104670776214e-05,
      "loss": 0.6813,
      "mean_token_accuracy": 0.796684168279171,
      "num_tokens": 297560028.0,
      "step": 9323
    },
    {
      "entropy": 0.5320757003501058,
      "epoch": 0.8578674073179567,
      "grad_norm": 0.14485399425029755,
      "learning_rate": 7.140797988162051e-05,
      "loss": 0.5263,
      "mean_token_accuracy": 0.8383158072829247,
      "num_tokens": 297592738.0,
      "step": 9324
    },
    {
      "entropy": 0.8263344466686249,
      "epoch": 0.857959413689398,
      "grad_norm": 0.15485422313213348,
      "learning_rate": 7.140491305547889e-05,
      "loss": 0.8078,
      "mean_token_accuracy": 0.7633046358823776,
      "num_tokens": 297624795.0,
      "step": 9325
    },
    {
      "entropy": 0.685953488573432,
      "epoch": 0.8580514200608392,
      "grad_norm": 0.15089641511440277,
      "learning_rate": 7.140184622933727e-05,
      "loss": 0.6751,
      "mean_token_accuracy": 0.797943726181984,
      "num_tokens": 297657198.0,
      "step": 9326
    },
    {
      "entropy": 0.6988447234034538,
      "epoch": 0.8581434264322805,
      "grad_norm": 0.1652422845363617,
      "learning_rate": 7.139877940319564e-05,
      "loss": 0.71,
      "mean_token_accuracy": 0.790430162101984,
      "num_tokens": 297689874.0,
      "step": 9327
    },
    {
      "entropy": 0.735746905207634,
      "epoch": 0.8582354328037216,
      "grad_norm": 0.15638157725334167,
      "learning_rate": 7.139571257705401e-05,
      "loss": 0.7236,
      "mean_token_accuracy": 0.787425022572279,
      "num_tokens": 297722180.0,
      "step": 9328
    },
    {
      "entropy": 0.7233849866315722,
      "epoch": 0.8583274391751629,
      "grad_norm": 0.1499679833650589,
      "learning_rate": 7.139264575091237e-05,
      "loss": 0.7049,
      "mean_token_accuracy": 0.7887495830655098,
      "num_tokens": 297754052.0,
      "step": 9329
    },
    {
      "entropy": 0.5786712244153023,
      "epoch": 0.8584194455466041,
      "grad_norm": 0.14468227326869965,
      "learning_rate": 7.138957892477076e-05,
      "loss": 0.5668,
      "mean_token_accuracy": 0.828399296849966,
      "num_tokens": 297786252.0,
      "step": 9330
    },
    {
      "entropy": 0.8570959703065455,
      "epoch": 0.8585114519180453,
      "grad_norm": 0.1602494716644287,
      "learning_rate": 7.138651209862914e-05,
      "loss": 0.8495,
      "mean_token_accuracy": 0.7536162473261356,
      "num_tokens": 297817549.0,
      "step": 9331
    },
    {
      "entropy": 0.8060774747282267,
      "epoch": 0.8586034582894866,
      "grad_norm": 0.15578515827655792,
      "learning_rate": 7.13834452724875e-05,
      "loss": 0.7998,
      "mean_token_accuracy": 0.7633193731307983,
      "num_tokens": 297848894.0,
      "step": 9332
    },
    {
      "entropy": 0.6526872832328081,
      "epoch": 0.8586954646609277,
      "grad_norm": 0.1524016261100769,
      "learning_rate": 7.138037844634587e-05,
      "loss": 0.6351,
      "mean_token_accuracy": 0.8095323108136654,
      "num_tokens": 297881286.0,
      "step": 9333
    },
    {
      "entropy": 0.8380382470786572,
      "epoch": 0.858787471032369,
      "grad_norm": 0.15755334496498108,
      "learning_rate": 7.137731162020425e-05,
      "loss": 0.8228,
      "mean_token_accuracy": 0.7612464055418968,
      "num_tokens": 297913102.0,
      "step": 9334
    },
    {
      "entropy": 0.5744717475026846,
      "epoch": 0.8588794774038102,
      "grad_norm": 0.13873127102851868,
      "learning_rate": 7.137424479406262e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8298993892967701,
      "num_tokens": 297945193.0,
      "step": 9335
    },
    {
      "entropy": 0.7727782651782036,
      "epoch": 0.8589714837752515,
      "grad_norm": 0.1567884385585785,
      "learning_rate": 7.1371177967921e-05,
      "loss": 0.7712,
      "mean_token_accuracy": 0.767839390784502,
      "num_tokens": 297977209.0,
      "step": 9336
    },
    {
      "entropy": 0.6931013837456703,
      "epoch": 0.8590634901466927,
      "grad_norm": 0.14930345118045807,
      "learning_rate": 7.136811114177939e-05,
      "loss": 0.6697,
      "mean_token_accuracy": 0.7944237478077412,
      "num_tokens": 298009347.0,
      "step": 9337
    },
    {
      "entropy": 0.6730757746845484,
      "epoch": 0.8591554965181338,
      "grad_norm": 0.1625451296567917,
      "learning_rate": 7.136504431563775e-05,
      "loss": 0.6614,
      "mean_token_accuracy": 0.7989698499441147,
      "num_tokens": 298041492.0,
      "step": 9338
    },
    {
      "entropy": 0.6938877943903208,
      "epoch": 0.8592475028895751,
      "grad_norm": 0.16013438999652863,
      "learning_rate": 7.136197748949612e-05,
      "loss": 0.6807,
      "mean_token_accuracy": 0.7982243038713932,
      "num_tokens": 298073165.0,
      "step": 9339
    },
    {
      "entropy": 0.7943490128964186,
      "epoch": 0.8593395092610163,
      "grad_norm": 0.1565820276737213,
      "learning_rate": 7.135891066335449e-05,
      "loss": 0.786,
      "mean_token_accuracy": 0.7694966346025467,
      "num_tokens": 298105192.0,
      "step": 9340
    },
    {
      "entropy": 0.6411199234426022,
      "epoch": 0.8594315156324576,
      "grad_norm": 0.14995883405208588,
      "learning_rate": 7.135584383721287e-05,
      "loss": 0.6226,
      "mean_token_accuracy": 0.816396925598383,
      "num_tokens": 298137728.0,
      "step": 9341
    },
    {
      "entropy": 0.8068310432136059,
      "epoch": 0.8595235220038988,
      "grad_norm": 0.1554294377565384,
      "learning_rate": 7.135277701107125e-05,
      "loss": 0.7934,
      "mean_token_accuracy": 0.7703254967927933,
      "num_tokens": 298170036.0,
      "step": 9342
    },
    {
      "entropy": 0.6797912418842316,
      "epoch": 0.8596155283753399,
      "grad_norm": 0.148912712931633,
      "learning_rate": 7.134971018492962e-05,
      "loss": 0.6753,
      "mean_token_accuracy": 0.7967238947749138,
      "num_tokens": 298201924.0,
      "step": 9343
    },
    {
      "entropy": 0.7927571497857571,
      "epoch": 0.8597075347467812,
      "grad_norm": 0.1606292724609375,
      "learning_rate": 7.134664335878799e-05,
      "loss": 0.7832,
      "mean_token_accuracy": 0.7697999328374863,
      "num_tokens": 298233843.0,
      "step": 9344
    },
    {
      "entropy": 0.6731222905218601,
      "epoch": 0.8597995411182224,
      "grad_norm": 0.16124464571475983,
      "learning_rate": 7.134357653264637e-05,
      "loss": 0.6531,
      "mean_token_accuracy": 0.8043259158730507,
      "num_tokens": 298266037.0,
      "step": 9345
    },
    {
      "entropy": 0.6806737147271633,
      "epoch": 0.8598915474896637,
      "grad_norm": 0.16043823957443237,
      "learning_rate": 7.134050970650474e-05,
      "loss": 0.6785,
      "mean_token_accuracy": 0.7972434684634209,
      "num_tokens": 298298716.0,
      "step": 9346
    },
    {
      "entropy": 0.7487082909792662,
      "epoch": 0.8599835538611049,
      "grad_norm": 0.1544087827205658,
      "learning_rate": 7.133744288036312e-05,
      "loss": 0.7452,
      "mean_token_accuracy": 0.7802769765257835,
      "num_tokens": 298330337.0,
      "step": 9347
    },
    {
      "entropy": 0.7610769718885422,
      "epoch": 0.860075560232546,
      "grad_norm": 0.1525411456823349,
      "learning_rate": 7.133437605422149e-05,
      "loss": 0.765,
      "mean_token_accuracy": 0.7772616781294346,
      "num_tokens": 298361512.0,
      "step": 9348
    },
    {
      "entropy": 0.7371473023667932,
      "epoch": 0.8601675666039873,
      "grad_norm": 0.16762474179267883,
      "learning_rate": 7.133130922807987e-05,
      "loss": 0.7159,
      "mean_token_accuracy": 0.7869123592972755,
      "num_tokens": 298393924.0,
      "step": 9349
    },
    {
      "entropy": 0.6777955461293459,
      "epoch": 0.8602595729754285,
      "grad_norm": 0.15613555908203125,
      "learning_rate": 7.132824240193824e-05,
      "loss": 0.6776,
      "mean_token_accuracy": 0.7975696586072445,
      "num_tokens": 298425347.0,
      "step": 9350
    },
    {
      "entropy": 0.7873505000025034,
      "epoch": 0.8603515793468698,
      "grad_norm": 0.15739117562770844,
      "learning_rate": 7.13251755757966e-05,
      "loss": 0.7793,
      "mean_token_accuracy": 0.7702473998069763,
      "num_tokens": 298456839.0,
      "step": 9351
    },
    {
      "entropy": 0.6488509960472584,
      "epoch": 0.860443585718311,
      "grad_norm": 0.1496942788362503,
      "learning_rate": 7.132210874965498e-05,
      "loss": 0.6264,
      "mean_token_accuracy": 0.8130797371268272,
      "num_tokens": 298489241.0,
      "step": 9352
    },
    {
      "entropy": 0.7225528489798307,
      "epoch": 0.8605355920897522,
      "grad_norm": 0.16193963587284088,
      "learning_rate": 7.131904192351337e-05,
      "loss": 0.6987,
      "mean_token_accuracy": 0.7941003106534481,
      "num_tokens": 298521252.0,
      "step": 9353
    },
    {
      "entropy": 0.5414722617715597,
      "epoch": 0.8606275984611934,
      "grad_norm": 0.14265698194503784,
      "learning_rate": 7.131597509737173e-05,
      "loss": 0.5172,
      "mean_token_accuracy": 0.8444129787385464,
      "num_tokens": 298553878.0,
      "step": 9354
    },
    {
      "entropy": 0.6608944507315755,
      "epoch": 0.8607196048326347,
      "grad_norm": 0.1455209106206894,
      "learning_rate": 7.13129082712301e-05,
      "loss": 0.6398,
      "mean_token_accuracy": 0.8110570982098579,
      "num_tokens": 298585013.0,
      "step": 9355
    },
    {
      "entropy": 0.7261070720851421,
      "epoch": 0.8608116112040759,
      "grad_norm": 0.14705535769462585,
      "learning_rate": 7.130984144508847e-05,
      "loss": 0.7187,
      "mean_token_accuracy": 0.7898045480251312,
      "num_tokens": 298616758.0,
      "step": 9356
    },
    {
      "entropy": 0.7563739940524101,
      "epoch": 0.8609036175755171,
      "grad_norm": 0.1523788869380951,
      "learning_rate": 7.130677461894686e-05,
      "loss": 0.744,
      "mean_token_accuracy": 0.78176524117589,
      "num_tokens": 298649307.0,
      "step": 9357
    },
    {
      "entropy": 0.7396757546812296,
      "epoch": 0.8609956239469583,
      "grad_norm": 0.15923769772052765,
      "learning_rate": 7.130370779280523e-05,
      "loss": 0.7254,
      "mean_token_accuracy": 0.7868451625108719,
      "num_tokens": 298680773.0,
      "step": 9358
    },
    {
      "entropy": 0.7327138874679804,
      "epoch": 0.8610876303183995,
      "grad_norm": 0.16041822731494904,
      "learning_rate": 7.13006409666636e-05,
      "loss": 0.7241,
      "mean_token_accuracy": 0.7845996841788292,
      "num_tokens": 298713203.0,
      "step": 9359
    },
    {
      "entropy": 0.7311303205788136,
      "epoch": 0.8611796366898408,
      "grad_norm": 0.15769672393798828,
      "learning_rate": 7.129757414052197e-05,
      "loss": 0.7296,
      "mean_token_accuracy": 0.7863708809018135,
      "num_tokens": 298745361.0,
      "step": 9360
    },
    {
      "entropy": 0.7008637581020594,
      "epoch": 0.861271643061282,
      "grad_norm": 0.16315065324306488,
      "learning_rate": 7.129450731438035e-05,
      "loss": 0.6812,
      "mean_token_accuracy": 0.801989059895277,
      "num_tokens": 298777295.0,
      "step": 9361
    },
    {
      "entropy": 0.6692836005240679,
      "epoch": 0.8613636494327233,
      "grad_norm": 0.16072410345077515,
      "learning_rate": 7.129144048823873e-05,
      "loss": 0.6372,
      "mean_token_accuracy": 0.81234135851264,
      "num_tokens": 298808145.0,
      "step": 9362
    },
    {
      "entropy": 0.7899717148393393,
      "epoch": 0.8614556558041644,
      "grad_norm": 0.15900303423404694,
      "learning_rate": 7.12883736620971e-05,
      "loss": 0.7668,
      "mean_token_accuracy": 0.7736680582165718,
      "num_tokens": 298839733.0,
      "step": 9363
    },
    {
      "entropy": 0.7437034863978624,
      "epoch": 0.8615476621756056,
      "grad_norm": 0.16146042943000793,
      "learning_rate": 7.128530683595547e-05,
      "loss": 0.7417,
      "mean_token_accuracy": 0.7858252041041851,
      "num_tokens": 298871766.0,
      "step": 9364
    },
    {
      "entropy": 0.6597649594768882,
      "epoch": 0.8616396685470469,
      "grad_norm": 0.14669270813465118,
      "learning_rate": 7.128224000981385e-05,
      "loss": 0.6551,
      "mean_token_accuracy": 0.8035356514155865,
      "num_tokens": 298904463.0,
      "step": 9365
    },
    {
      "entropy": 0.6583692217245698,
      "epoch": 0.8617316749184881,
      "grad_norm": 0.16212201118469238,
      "learning_rate": 7.127917318367222e-05,
      "loss": 0.6427,
      "mean_token_accuracy": 0.8038212023675442,
      "num_tokens": 298936578.0,
      "step": 9366
    },
    {
      "entropy": 0.7809443064033985,
      "epoch": 0.8618236812899294,
      "grad_norm": 0.16503342986106873,
      "learning_rate": 7.12761063575306e-05,
      "loss": 0.7898,
      "mean_token_accuracy": 0.7645572423934937,
      "num_tokens": 298969097.0,
      "step": 9367
    },
    {
      "entropy": 0.7508597746491432,
      "epoch": 0.8619156876613705,
      "grad_norm": 0.16139595210552216,
      "learning_rate": 7.127303953138898e-05,
      "loss": 0.739,
      "mean_token_accuracy": 0.7800043001770973,
      "num_tokens": 299000038.0,
      "step": 9368
    },
    {
      "entropy": 0.608279675245285,
      "epoch": 0.8620076940328117,
      "grad_norm": 0.15536369383335114,
      "learning_rate": 7.126997270524735e-05,
      "loss": 0.6135,
      "mean_token_accuracy": 0.8134923093020916,
      "num_tokens": 299031974.0,
      "step": 9369
    },
    {
      "entropy": 0.7602689657360315,
      "epoch": 0.862099700404253,
      "grad_norm": 0.16456453502178192,
      "learning_rate": 7.126690587910571e-05,
      "loss": 0.7495,
      "mean_token_accuracy": 0.7787371687591076,
      "num_tokens": 299063732.0,
      "step": 9370
    },
    {
      "entropy": 0.7024401854723692,
      "epoch": 0.8621917067756942,
      "grad_norm": 0.15249143540859222,
      "learning_rate": 7.126383905296408e-05,
      "loss": 0.6939,
      "mean_token_accuracy": 0.7948904447257519,
      "num_tokens": 299095359.0,
      "step": 9371
    },
    {
      "entropy": 0.7347941920161247,
      "epoch": 0.8622837131471355,
      "grad_norm": 0.15575668215751648,
      "learning_rate": 7.126077222682246e-05,
      "loss": 0.7198,
      "mean_token_accuracy": 0.7877590842545033,
      "num_tokens": 299127133.0,
      "step": 9372
    },
    {
      "entropy": 0.8200506959110498,
      "epoch": 0.8623757195185766,
      "grad_norm": 0.15288935601711273,
      "learning_rate": 7.125770540068085e-05,
      "loss": 0.8032,
      "mean_token_accuracy": 0.7619360275566578,
      "num_tokens": 299159114.0,
      "step": 9373
    },
    {
      "entropy": 0.6757137970998883,
      "epoch": 0.8624677258900179,
      "grad_norm": 0.15401972830295563,
      "learning_rate": 7.125463857453921e-05,
      "loss": 0.6488,
      "mean_token_accuracy": 0.8064882047474384,
      "num_tokens": 299191251.0,
      "step": 9374
    },
    {
      "entropy": 0.6827469021081924,
      "epoch": 0.8625597322614591,
      "grad_norm": 0.14911216497421265,
      "learning_rate": 7.125157174839758e-05,
      "loss": 0.6601,
      "mean_token_accuracy": 0.8016263656318188,
      "num_tokens": 299223336.0,
      "step": 9375
    },
    {
      "entropy": 0.6900500860065222,
      "epoch": 0.8626517386329003,
      "grad_norm": 0.1534057855606079,
      "learning_rate": 7.124850492225596e-05,
      "loss": 0.6792,
      "mean_token_accuracy": 0.7968048267066479,
      "num_tokens": 299255386.0,
      "step": 9376
    },
    {
      "entropy": 0.8208514656871557,
      "epoch": 0.8627437450043416,
      "grad_norm": 0.1681324541568756,
      "learning_rate": 7.124543809611433e-05,
      "loss": 0.8067,
      "mean_token_accuracy": 0.7633748315274715,
      "num_tokens": 299287630.0,
      "step": 9377
    },
    {
      "entropy": 0.6749212071299553,
      "epoch": 0.8628357513757827,
      "grad_norm": 0.15617109835147858,
      "learning_rate": 7.124237126997271e-05,
      "loss": 0.6577,
      "mean_token_accuracy": 0.8082833215594292,
      "num_tokens": 299318556.0,
      "step": 9378
    },
    {
      "entropy": 0.6894717160612345,
      "epoch": 0.862927757747224,
      "grad_norm": 0.15473876893520355,
      "learning_rate": 7.123930444383108e-05,
      "loss": 0.6817,
      "mean_token_accuracy": 0.802324466407299,
      "num_tokens": 299350235.0,
      "step": 9379
    },
    {
      "entropy": 0.7316184937953949,
      "epoch": 0.8630197641186652,
      "grad_norm": 0.14969295263290405,
      "learning_rate": 7.123623761768946e-05,
      "loss": 0.7092,
      "mean_token_accuracy": 0.7856973484158516,
      "num_tokens": 299381138.0,
      "step": 9380
    },
    {
      "entropy": 0.685287032276392,
      "epoch": 0.8631117704901065,
      "grad_norm": 0.1465553492307663,
      "learning_rate": 7.123317079154783e-05,
      "loss": 0.6777,
      "mean_token_accuracy": 0.8046531528234482,
      "num_tokens": 299413906.0,
      "step": 9381
    },
    {
      "entropy": 0.830245666205883,
      "epoch": 0.8632037768615477,
      "grad_norm": 0.1538570076227188,
      "learning_rate": 7.12301039654062e-05,
      "loss": 0.8143,
      "mean_token_accuracy": 0.7580981180071831,
      "num_tokens": 299445110.0,
      "step": 9382
    },
    {
      "entropy": 0.663673710078001,
      "epoch": 0.8632957832329888,
      "grad_norm": 0.1493903398513794,
      "learning_rate": 7.122703713926458e-05,
      "loss": 0.6467,
      "mean_token_accuracy": 0.8079764321446419,
      "num_tokens": 299477801.0,
      "step": 9383
    },
    {
      "entropy": 0.5991844329982996,
      "epoch": 0.8633877896044301,
      "grad_norm": 0.15614666044712067,
      "learning_rate": 7.122397031312296e-05,
      "loss": 0.5927,
      "mean_token_accuracy": 0.8255625069141388,
      "num_tokens": 299510211.0,
      "step": 9384
    },
    {
      "entropy": 0.6595040559768677,
      "epoch": 0.8634797959758713,
      "grad_norm": 0.1559057980775833,
      "learning_rate": 7.122090348698133e-05,
      "loss": 0.6559,
      "mean_token_accuracy": 0.8054000027477741,
      "num_tokens": 299541724.0,
      "step": 9385
    },
    {
      "entropy": 0.7625353578478098,
      "epoch": 0.8635718023473126,
      "grad_norm": 0.16446596384048462,
      "learning_rate": 7.12178366608397e-05,
      "loss": 0.7778,
      "mean_token_accuracy": 0.773286622017622,
      "num_tokens": 299572628.0,
      "step": 9386
    },
    {
      "entropy": 0.5428614635020494,
      "epoch": 0.8636638087187538,
      "grad_norm": 0.1433032602071762,
      "learning_rate": 7.121476983469806e-05,
      "loss": 0.5211,
      "mean_token_accuracy": 0.8432894051074982,
      "num_tokens": 299604547.0,
      "step": 9387
    },
    {
      "entropy": 0.7461486291140318,
      "epoch": 0.863755815090195,
      "grad_norm": 0.1581535041332245,
      "learning_rate": 7.121170300855645e-05,
      "loss": 0.7509,
      "mean_token_accuracy": 0.7813522480428219,
      "num_tokens": 299636795.0,
      "step": 9388
    },
    {
      "entropy": 0.7161935400217772,
      "epoch": 0.8638478214616362,
      "grad_norm": 0.15428520739078522,
      "learning_rate": 7.120863618241483e-05,
      "loss": 0.7205,
      "mean_token_accuracy": 0.7893827147781849,
      "num_tokens": 299668364.0,
      "step": 9389
    },
    {
      "entropy": 0.7543684802949429,
      "epoch": 0.8639398278330774,
      "grad_norm": 0.15689656138420105,
      "learning_rate": 7.12055693562732e-05,
      "loss": 0.7489,
      "mean_token_accuracy": 0.7772361226379871,
      "num_tokens": 299700033.0,
      "step": 9390
    },
    {
      "entropy": 0.5843158727511764,
      "epoch": 0.8640318342045187,
      "grad_norm": 0.1481853723526001,
      "learning_rate": 7.120250253013156e-05,
      "loss": 0.5747,
      "mean_token_accuracy": 0.8267658092081547,
      "num_tokens": 299732518.0,
      "step": 9391
    },
    {
      "entropy": 0.6819134410470724,
      "epoch": 0.8641238405759599,
      "grad_norm": 0.15792417526245117,
      "learning_rate": 7.119943570398994e-05,
      "loss": 0.6716,
      "mean_token_accuracy": 0.800098180770874,
      "num_tokens": 299764937.0,
      "step": 9392
    },
    {
      "entropy": 0.6880963798612356,
      "epoch": 0.8642158469474011,
      "grad_norm": 0.15414679050445557,
      "learning_rate": 7.119636887784833e-05,
      "loss": 0.6806,
      "mean_token_accuracy": 0.7967887744307518,
      "num_tokens": 299796771.0,
      "step": 9393
    },
    {
      "entropy": 0.7324457168579102,
      "epoch": 0.8643078533188423,
      "grad_norm": 0.1551772803068161,
      "learning_rate": 7.119330205170669e-05,
      "loss": 0.7312,
      "mean_token_accuracy": 0.7846244312822819,
      "num_tokens": 299829273.0,
      "step": 9394
    },
    {
      "entropy": 0.6403395067900419,
      "epoch": 0.8643998596902835,
      "grad_norm": 0.14846108853816986,
      "learning_rate": 7.119023522556506e-05,
      "loss": 0.6251,
      "mean_token_accuracy": 0.8137858547270298,
      "num_tokens": 299861954.0,
      "step": 9395
    },
    {
      "entropy": 0.6851215865463018,
      "epoch": 0.8644918660617248,
      "grad_norm": 0.17375381290912628,
      "learning_rate": 7.118716839942344e-05,
      "loss": 0.6693,
      "mean_token_accuracy": 0.802326925098896,
      "num_tokens": 299893999.0,
      "step": 9396
    },
    {
      "entropy": 0.6878429874777794,
      "epoch": 0.864583872433166,
      "grad_norm": 0.16416507959365845,
      "learning_rate": 7.118410157328181e-05,
      "loss": 0.6636,
      "mean_token_accuracy": 0.8006279990077019,
      "num_tokens": 299926234.0,
      "step": 9397
    },
    {
      "entropy": 0.7307764273136854,
      "epoch": 0.8646758788046072,
      "grad_norm": 0.1567036211490631,
      "learning_rate": 7.118103474714019e-05,
      "loss": 0.6986,
      "mean_token_accuracy": 0.7965372949838638,
      "num_tokens": 299957921.0,
      "step": 9398
    },
    {
      "entropy": 0.7893817499279976,
      "epoch": 0.8647678851760484,
      "grad_norm": 0.16812321543693542,
      "learning_rate": 7.117796792099857e-05,
      "loss": 0.7868,
      "mean_token_accuracy": 0.7669800966978073,
      "num_tokens": 299989422.0,
      "step": 9399
    },
    {
      "entropy": 0.7787327207624912,
      "epoch": 0.8648598915474897,
      "grad_norm": 0.15346363186836243,
      "learning_rate": 7.117490109485694e-05,
      "loss": 0.7582,
      "mean_token_accuracy": 0.7759929969906807,
      "num_tokens": 300020798.0,
      "step": 9400
    },
    {
      "entropy": 0.745806310325861,
      "epoch": 0.8649518979189309,
      "grad_norm": 0.15797533094882965,
      "learning_rate": 7.117183426871531e-05,
      "loss": 0.7332,
      "mean_token_accuracy": 0.7849910296499729,
      "num_tokens": 300053146.0,
      "step": 9401
    },
    {
      "entropy": 0.6929074302315712,
      "epoch": 0.8650439042903721,
      "grad_norm": 0.147673100233078,
      "learning_rate": 7.116876744257368e-05,
      "loss": 0.6846,
      "mean_token_accuracy": 0.7943231239914894,
      "num_tokens": 300085695.0,
      "step": 9402
    },
    {
      "entropy": 0.7214850448071957,
      "epoch": 0.8651359106618133,
      "grad_norm": 0.16046148538589478,
      "learning_rate": 7.116570061643206e-05,
      "loss": 0.6853,
      "mean_token_accuracy": 0.799546480178833,
      "num_tokens": 300117980.0,
      "step": 9403
    },
    {
      "entropy": 0.6318915672600269,
      "epoch": 0.8652279170332545,
      "grad_norm": 0.1728653907775879,
      "learning_rate": 7.116263379029044e-05,
      "loss": 0.5989,
      "mean_token_accuracy": 0.8181651458144188,
      "num_tokens": 300150090.0,
      "step": 9404
    },
    {
      "entropy": 0.710662767291069,
      "epoch": 0.8653199234046958,
      "grad_norm": 0.14905354380607605,
      "learning_rate": 7.115956696414881e-05,
      "loss": 0.7073,
      "mean_token_accuracy": 0.7903906628489494,
      "num_tokens": 300180597.0,
      "step": 9405
    },
    {
      "entropy": 0.7089619562029839,
      "epoch": 0.865411929776137,
      "grad_norm": 0.15632876753807068,
      "learning_rate": 7.115650013800718e-05,
      "loss": 0.6911,
      "mean_token_accuracy": 0.7940429858863354,
      "num_tokens": 300212967.0,
      "step": 9406
    },
    {
      "entropy": 0.6816257676109672,
      "epoch": 0.8655039361475783,
      "grad_norm": 0.15161395072937012,
      "learning_rate": 7.115343331186556e-05,
      "loss": 0.6645,
      "mean_token_accuracy": 0.8036345206201077,
      "num_tokens": 300245089.0,
      "step": 9407
    },
    {
      "entropy": 0.7538397312164307,
      "epoch": 0.8655959425190194,
      "grad_norm": 0.1569976955652237,
      "learning_rate": 7.115036648572392e-05,
      "loss": 0.7761,
      "mean_token_accuracy": 0.7740913853049278,
      "num_tokens": 300277378.0,
      "step": 9408
    },
    {
      "entropy": 0.7171483132988214,
      "epoch": 0.8656879488904606,
      "grad_norm": 0.15576738119125366,
      "learning_rate": 7.11472996595823e-05,
      "loss": 0.7049,
      "mean_token_accuracy": 0.7906314693391323,
      "num_tokens": 300308930.0,
      "step": 9409
    },
    {
      "entropy": 0.8319399114698172,
      "epoch": 0.8657799552619019,
      "grad_norm": 0.1570492535829544,
      "learning_rate": 7.114423283344067e-05,
      "loss": 0.836,
      "mean_token_accuracy": 0.7544476129114628,
      "num_tokens": 300340483.0,
      "step": 9410
    },
    {
      "entropy": 0.7999623846262693,
      "epoch": 0.8658719616333431,
      "grad_norm": 0.1617983877658844,
      "learning_rate": 7.114116600729906e-05,
      "loss": 0.7885,
      "mean_token_accuracy": 0.7660381384193897,
      "num_tokens": 300372086.0,
      "step": 9411
    },
    {
      "entropy": 0.7237855978310108,
      "epoch": 0.8659639680047844,
      "grad_norm": 0.1587868332862854,
      "learning_rate": 7.113809918115742e-05,
      "loss": 0.7234,
      "mean_token_accuracy": 0.7842574529349804,
      "num_tokens": 300404854.0,
      "step": 9412
    },
    {
      "entropy": 0.7814938835799694,
      "epoch": 0.8660559743762255,
      "grad_norm": 0.15497761964797974,
      "learning_rate": 7.113503235501579e-05,
      "loss": 0.7816,
      "mean_token_accuracy": 0.7745209895074368,
      "num_tokens": 300436938.0,
      "step": 9413
    },
    {
      "entropy": 0.8089552875608206,
      "epoch": 0.8661479807476667,
      "grad_norm": 0.16208279132843018,
      "learning_rate": 7.113196552887417e-05,
      "loss": 0.8008,
      "mean_token_accuracy": 0.7689556889235973,
      "num_tokens": 300468854.0,
      "step": 9414
    },
    {
      "entropy": 0.6917135594412684,
      "epoch": 0.866239987119108,
      "grad_norm": 0.1519940346479416,
      "learning_rate": 7.112889870273255e-05,
      "loss": 0.6855,
      "mean_token_accuracy": 0.7955516390502453,
      "num_tokens": 300500916.0,
      "step": 9415
    },
    {
      "entropy": 0.6809289772063494,
      "epoch": 0.8663319934905492,
      "grad_norm": 0.1549980640411377,
      "learning_rate": 7.112583187659092e-05,
      "loss": 0.6611,
      "mean_token_accuracy": 0.8039533533155918,
      "num_tokens": 300532881.0,
      "step": 9416
    },
    {
      "entropy": 0.7079488737508655,
      "epoch": 0.8664239998619905,
      "grad_norm": 0.1466134637594223,
      "learning_rate": 7.112276505044929e-05,
      "loss": 0.6889,
      "mean_token_accuracy": 0.7950500212609768,
      "num_tokens": 300565119.0,
      "step": 9417
    },
    {
      "entropy": 0.7793791815638542,
      "epoch": 0.8665160062334316,
      "grad_norm": 0.15578745305538177,
      "learning_rate": 7.111969822430766e-05,
      "loss": 0.7616,
      "mean_token_accuracy": 0.776256438344717,
      "num_tokens": 300597056.0,
      "step": 9418
    },
    {
      "entropy": 0.7559075858443975,
      "epoch": 0.8666080126048729,
      "grad_norm": 0.15704002976417542,
      "learning_rate": 7.111663139816604e-05,
      "loss": 0.7598,
      "mean_token_accuracy": 0.7764304131269455,
      "num_tokens": 300627644.0,
      "step": 9419
    },
    {
      "entropy": 0.7665963768959045,
      "epoch": 0.8667000189763141,
      "grad_norm": 0.1515263319015503,
      "learning_rate": 7.111356457202442e-05,
      "loss": 0.7593,
      "mean_token_accuracy": 0.7741331905126572,
      "num_tokens": 300659634.0,
      "step": 9420
    },
    {
      "entropy": 0.7888748310506344,
      "epoch": 0.8667920253477553,
      "grad_norm": 0.1598089039325714,
      "learning_rate": 7.111049774588279e-05,
      "loss": 0.7709,
      "mean_token_accuracy": 0.7720173187553883,
      "num_tokens": 300690817.0,
      "step": 9421
    },
    {
      "entropy": 0.6942821424454451,
      "epoch": 0.8668840317191966,
      "grad_norm": 0.1557421088218689,
      "learning_rate": 7.110743091974116e-05,
      "loss": 0.6766,
      "mean_token_accuracy": 0.7931774668395519,
      "num_tokens": 300722051.0,
      "step": 9422
    },
    {
      "entropy": 0.6799720376729965,
      "epoch": 0.8669760380906377,
      "grad_norm": 0.1579056680202484,
      "learning_rate": 7.110436409359954e-05,
      "loss": 0.6656,
      "mean_token_accuracy": 0.8029154315590858,
      "num_tokens": 300754358.0,
      "step": 9423
    },
    {
      "entropy": 0.7226154636591673,
      "epoch": 0.867068044462079,
      "grad_norm": 0.1556926816701889,
      "learning_rate": 7.11012972674579e-05,
      "loss": 0.7135,
      "mean_token_accuracy": 0.7895899824798107,
      "num_tokens": 300785930.0,
      "step": 9424
    },
    {
      "entropy": 0.7345237042754889,
      "epoch": 0.8671600508335202,
      "grad_norm": 0.15676768124103546,
      "learning_rate": 7.109823044131629e-05,
      "loss": 0.7236,
      "mean_token_accuracy": 0.7871703617274761,
      "num_tokens": 300818235.0,
      "step": 9425
    },
    {
      "entropy": 0.8337792176753283,
      "epoch": 0.8672520572049615,
      "grad_norm": 0.15863996744155884,
      "learning_rate": 7.109516361517465e-05,
      "loss": 0.8368,
      "mean_token_accuracy": 0.7580298408865929,
      "num_tokens": 300849671.0,
      "step": 9426
    },
    {
      "entropy": 0.6408861503005028,
      "epoch": 0.8673440635764027,
      "grad_norm": 0.14432469010353088,
      "learning_rate": 7.109209678903304e-05,
      "loss": 0.6275,
      "mean_token_accuracy": 0.8110961392521858,
      "num_tokens": 300881384.0,
      "step": 9427
    },
    {
      "entropy": 0.6639566514641047,
      "epoch": 0.8674360699478438,
      "grad_norm": 0.15528973937034607,
      "learning_rate": 7.10890299628914e-05,
      "loss": 0.6414,
      "mean_token_accuracy": 0.8050641529262066,
      "num_tokens": 300912498.0,
      "step": 9428
    },
    {
      "entropy": 0.7425696616992354,
      "epoch": 0.8675280763192851,
      "grad_norm": 0.15245935320854187,
      "learning_rate": 7.108596313674977e-05,
      "loss": 0.7414,
      "mean_token_accuracy": 0.7844297215342522,
      "num_tokens": 300945129.0,
      "step": 9429
    },
    {
      "entropy": 0.672296429052949,
      "epoch": 0.8676200826907263,
      "grad_norm": 0.15642182528972626,
      "learning_rate": 7.108289631060817e-05,
      "loss": 0.6713,
      "mean_token_accuracy": 0.7971320003271103,
      "num_tokens": 300976687.0,
      "step": 9430
    },
    {
      "entropy": 0.7621972765773535,
      "epoch": 0.8677120890621676,
      "grad_norm": 0.15754960477352142,
      "learning_rate": 7.107982948446653e-05,
      "loss": 0.7624,
      "mean_token_accuracy": 0.7788499556481838,
      "num_tokens": 301008572.0,
      "step": 9431
    },
    {
      "entropy": 0.7649450586177409,
      "epoch": 0.8678040954336088,
      "grad_norm": 0.15176703035831451,
      "learning_rate": 7.10767626583249e-05,
      "loss": 0.7536,
      "mean_token_accuracy": 0.7805104739964008,
      "num_tokens": 301040225.0,
      "step": 9432
    },
    {
      "entropy": 0.6658989861607552,
      "epoch": 0.86789610180505,
      "grad_norm": 0.15933527052402496,
      "learning_rate": 7.107369583218327e-05,
      "loss": 0.6628,
      "mean_token_accuracy": 0.8045866303145885,
      "num_tokens": 301072022.0,
      "step": 9433
    },
    {
      "entropy": 0.6720318980515003,
      "epoch": 0.8679881081764912,
      "grad_norm": 0.16329513490200043,
      "learning_rate": 7.107062900604165e-05,
      "loss": 0.6586,
      "mean_token_accuracy": 0.8015291132032871,
      "num_tokens": 301102631.0,
      "step": 9434
    },
    {
      "entropy": 0.8501088488847017,
      "epoch": 0.8680801145479324,
      "grad_norm": 0.1487184315919876,
      "learning_rate": 7.106756217990003e-05,
      "loss": 0.8296,
      "mean_token_accuracy": 0.7561903186142445,
      "num_tokens": 301134243.0,
      "step": 9435
    },
    {
      "entropy": 0.7197634484618902,
      "epoch": 0.8681721209193737,
      "grad_norm": 0.15416289865970612,
      "learning_rate": 7.10644953537584e-05,
      "loss": 0.722,
      "mean_token_accuracy": 0.7869780398905277,
      "num_tokens": 301166875.0,
      "step": 9436
    },
    {
      "entropy": 0.7680638814345002,
      "epoch": 0.8682641272908149,
      "grad_norm": 0.15282733738422394,
      "learning_rate": 7.106142852761677e-05,
      "loss": 0.7508,
      "mean_token_accuracy": 0.7838510535657406,
      "num_tokens": 301198403.0,
      "step": 9437
    },
    {
      "entropy": 0.6490823617205024,
      "epoch": 0.8683561336622561,
      "grad_norm": 0.15472978353500366,
      "learning_rate": 7.105836170147515e-05,
      "loss": 0.634,
      "mean_token_accuracy": 0.8103040084242821,
      "num_tokens": 301229929.0,
      "step": 9438
    },
    {
      "entropy": 0.6991249807178974,
      "epoch": 0.8684481400336973,
      "grad_norm": 0.1649285852909088,
      "learning_rate": 7.105529487533352e-05,
      "loss": 0.6925,
      "mean_token_accuracy": 0.7925313599407673,
      "num_tokens": 301261339.0,
      "step": 9439
    },
    {
      "entropy": 0.6567652514204383,
      "epoch": 0.8685401464051385,
      "grad_norm": 0.14103394746780396,
      "learning_rate": 7.10522280491919e-05,
      "loss": 0.6385,
      "mean_token_accuracy": 0.8115219697356224,
      "num_tokens": 301292477.0,
      "step": 9440
    },
    {
      "entropy": 0.713728504255414,
      "epoch": 0.8686321527765798,
      "grad_norm": 0.15785759687423706,
      "learning_rate": 7.104916122305027e-05,
      "loss": 0.701,
      "mean_token_accuracy": 0.7906535379588604,
      "num_tokens": 301324513.0,
      "step": 9441
    },
    {
      "entropy": 0.7993347812443972,
      "epoch": 0.868724159148021,
      "grad_norm": 0.15786348283290863,
      "learning_rate": 7.104609439690865e-05,
      "loss": 0.8035,
      "mean_token_accuracy": 0.7655744925141335,
      "num_tokens": 301356866.0,
      "step": 9442
    },
    {
      "entropy": 0.6744106747210026,
      "epoch": 0.8688161655194622,
      "grad_norm": 0.1456708014011383,
      "learning_rate": 7.104302757076702e-05,
      "loss": 0.6442,
      "mean_token_accuracy": 0.8072120472788811,
      "num_tokens": 301388685.0,
      "step": 9443
    },
    {
      "entropy": 0.6963634938001633,
      "epoch": 0.8689081718909034,
      "grad_norm": 0.1476767659187317,
      "learning_rate": 7.103996074462538e-05,
      "loss": 0.6734,
      "mean_token_accuracy": 0.8007310293614864,
      "num_tokens": 301420365.0,
      "step": 9444
    },
    {
      "entropy": 0.6776910927146673,
      "epoch": 0.8690001782623447,
      "grad_norm": 0.1538129597902298,
      "learning_rate": 7.103689391848377e-05,
      "loss": 0.6619,
      "mean_token_accuracy": 0.8012172840535641,
      "num_tokens": 301452318.0,
      "step": 9445
    },
    {
      "entropy": 0.6588343000039458,
      "epoch": 0.8690921846337859,
      "grad_norm": 0.14567917585372925,
      "learning_rate": 7.103382709234215e-05,
      "loss": 0.6584,
      "mean_token_accuracy": 0.8032059147953987,
      "num_tokens": 301484725.0,
      "step": 9446
    },
    {
      "entropy": 0.7252768520265818,
      "epoch": 0.8691841910052271,
      "grad_norm": 0.156536266207695,
      "learning_rate": 7.103076026620052e-05,
      "loss": 0.7083,
      "mean_token_accuracy": 0.790626086294651,
      "num_tokens": 301516500.0,
      "step": 9447
    },
    {
      "entropy": 0.7714630998671055,
      "epoch": 0.8692761973766683,
      "grad_norm": 0.15624500811100006,
      "learning_rate": 7.102769344005888e-05,
      "loss": 0.7654,
      "mean_token_accuracy": 0.7741681709885597,
      "num_tokens": 301548497.0,
      "step": 9448
    },
    {
      "entropy": 0.8233904056251049,
      "epoch": 0.8693682037481095,
      "grad_norm": 0.15213371813297272,
      "learning_rate": 7.102462661391725e-05,
      "loss": 0.8139,
      "mean_token_accuracy": 0.760954562574625,
      "num_tokens": 301581174.0,
      "step": 9449
    },
    {
      "entropy": 0.7955761272460222,
      "epoch": 0.8694602101195508,
      "grad_norm": 0.15440776944160461,
      "learning_rate": 7.102155978777563e-05,
      "loss": 0.7766,
      "mean_token_accuracy": 0.7702567242085934,
      "num_tokens": 301612881.0,
      "step": 9450
    },
    {
      "entropy": 0.7331493068486452,
      "epoch": 0.869552216490992,
      "grad_norm": 0.14947505295276642,
      "learning_rate": 7.101849296163401e-05,
      "loss": 0.7208,
      "mean_token_accuracy": 0.7852564416825771,
      "num_tokens": 301644200.0,
      "step": 9451
    },
    {
      "entropy": 0.8410804308950901,
      "epoch": 0.8696442228624333,
      "grad_norm": 0.16650955379009247,
      "learning_rate": 7.101542613549238e-05,
      "loss": 0.8431,
      "mean_token_accuracy": 0.7546150535345078,
      "num_tokens": 301675943.0,
      "step": 9452
    },
    {
      "entropy": 0.6401942614465952,
      "epoch": 0.8697362292338744,
      "grad_norm": 0.15919563174247742,
      "learning_rate": 7.101235930935075e-05,
      "loss": 0.6269,
      "mean_token_accuracy": 0.8127207942306995,
      "num_tokens": 301707853.0,
      "step": 9453
    },
    {
      "entropy": 0.8252610377967358,
      "epoch": 0.8698282356053156,
      "grad_norm": 0.1646946668624878,
      "learning_rate": 7.100929248320913e-05,
      "loss": 0.805,
      "mean_token_accuracy": 0.7659855931997299,
      "num_tokens": 301738893.0,
      "step": 9454
    },
    {
      "entropy": 0.6691693309694529,
      "epoch": 0.8699202419767569,
      "grad_norm": 0.15261279046535492,
      "learning_rate": 7.10062256570675e-05,
      "loss": 0.6467,
      "mean_token_accuracy": 0.8016528151929379,
      "num_tokens": 301770299.0,
      "step": 9455
    },
    {
      "entropy": 0.7216642834246159,
      "epoch": 0.8700122483481981,
      "grad_norm": 0.15584354102611542,
      "learning_rate": 7.100315883092588e-05,
      "loss": 0.7063,
      "mean_token_accuracy": 0.7929015196859837,
      "num_tokens": 301802424.0,
      "step": 9456
    },
    {
      "entropy": 0.6661333963274956,
      "epoch": 0.8701042547196394,
      "grad_norm": 0.16474248468875885,
      "learning_rate": 7.100009200478426e-05,
      "loss": 0.6508,
      "mean_token_accuracy": 0.8026233203709126,
      "num_tokens": 301834087.0,
      "step": 9457
    },
    {
      "entropy": 0.6420670691877604,
      "epoch": 0.8701962610910805,
      "grad_norm": 0.16124258935451508,
      "learning_rate": 7.099702517864263e-05,
      "loss": 0.6323,
      "mean_token_accuracy": 0.8115709386765957,
      "num_tokens": 301865725.0,
      "step": 9458
    },
    {
      "entropy": 0.7117118127644062,
      "epoch": 0.8702882674625217,
      "grad_norm": 0.1626591682434082,
      "learning_rate": 7.0993958352501e-05,
      "loss": 0.7155,
      "mean_token_accuracy": 0.7894459404051304,
      "num_tokens": 301897623.0,
      "step": 9459
    },
    {
      "entropy": 0.8022860232740641,
      "epoch": 0.870380273833963,
      "grad_norm": 0.15779733657836914,
      "learning_rate": 7.099089152635937e-05,
      "loss": 0.7964,
      "mean_token_accuracy": 0.7696094401180744,
      "num_tokens": 301929872.0,
      "step": 9460
    },
    {
      "entropy": 0.6929803611710668,
      "epoch": 0.8704722802054042,
      "grad_norm": 0.14567877352237701,
      "learning_rate": 7.098782470021775e-05,
      "loss": 0.6806,
      "mean_token_accuracy": 0.7963567636907101,
      "num_tokens": 301962160.0,
      "step": 9461
    },
    {
      "entropy": 0.7078960966318846,
      "epoch": 0.8705642865768455,
      "grad_norm": 0.14545206725597382,
      "learning_rate": 7.098475787407613e-05,
      "loss": 0.6837,
      "mean_token_accuracy": 0.7955137118697166,
      "num_tokens": 301994159.0,
      "step": 9462
    },
    {
      "entropy": 0.7412412539124489,
      "epoch": 0.8706562929482866,
      "grad_norm": 0.15763337910175323,
      "learning_rate": 7.09816910479345e-05,
      "loss": 0.7311,
      "mean_token_accuracy": 0.7856837324798107,
      "num_tokens": 302026388.0,
      "step": 9463
    },
    {
      "entropy": 0.7852325700223446,
      "epoch": 0.8707482993197279,
      "grad_norm": 0.15953169763088226,
      "learning_rate": 7.097862422179286e-05,
      "loss": 0.7829,
      "mean_token_accuracy": 0.7693373784422874,
      "num_tokens": 302057964.0,
      "step": 9464
    },
    {
      "entropy": 0.726873017847538,
      "epoch": 0.8708403056911691,
      "grad_norm": 0.15617746114730835,
      "learning_rate": 7.097555739565125e-05,
      "loss": 0.721,
      "mean_token_accuracy": 0.7840094231069088,
      "num_tokens": 302089561.0,
      "step": 9465
    },
    {
      "entropy": 0.6638039676472545,
      "epoch": 0.8709323120626103,
      "grad_norm": 0.1514037400484085,
      "learning_rate": 7.097249056950961e-05,
      "loss": 0.6418,
      "mean_token_accuracy": 0.8058439232409,
      "num_tokens": 302122329.0,
      "step": 9466
    },
    {
      "entropy": 0.7913656998425722,
      "epoch": 0.8710243184340516,
      "grad_norm": 0.15442600846290588,
      "learning_rate": 7.0969423743368e-05,
      "loss": 0.7787,
      "mean_token_accuracy": 0.7721321769058704,
      "num_tokens": 302153580.0,
      "step": 9467
    },
    {
      "entropy": 0.5929783284664154,
      "epoch": 0.8711163248054927,
      "grad_norm": 0.15177573263645172,
      "learning_rate": 7.096635691722636e-05,
      "loss": 0.5859,
      "mean_token_accuracy": 0.8234858475625515,
      "num_tokens": 302186167.0,
      "step": 9468
    },
    {
      "entropy": 0.7608021926134825,
      "epoch": 0.871208331176934,
      "grad_norm": 0.1640535593032837,
      "learning_rate": 7.096329009108474e-05,
      "loss": 0.7715,
      "mean_token_accuracy": 0.7742783315479755,
      "num_tokens": 302218133.0,
      "step": 9469
    },
    {
      "entropy": 0.7814521789550781,
      "epoch": 0.8713003375483752,
      "grad_norm": 0.16345754265785217,
      "learning_rate": 7.096022326494311e-05,
      "loss": 0.7781,
      "mean_token_accuracy": 0.7698923051357269,
      "num_tokens": 302250380.0,
      "step": 9470
    },
    {
      "entropy": 0.7607823945581913,
      "epoch": 0.8713923439198165,
      "grad_norm": 0.15596403181552887,
      "learning_rate": 7.095715643880148e-05,
      "loss": 0.7392,
      "mean_token_accuracy": 0.7812750414013863,
      "num_tokens": 302280933.0,
      "step": 9471
    },
    {
      "entropy": 0.7376413093879819,
      "epoch": 0.8714843502912577,
      "grad_norm": 0.1518380045890808,
      "learning_rate": 7.095408961265986e-05,
      "loss": 0.7225,
      "mean_token_accuracy": 0.7882676310837269,
      "num_tokens": 302312598.0,
      "step": 9472
    },
    {
      "entropy": 0.6656134016811848,
      "epoch": 0.8715763566626988,
      "grad_norm": 0.16171549260616302,
      "learning_rate": 7.095102278651824e-05,
      "loss": 0.6452,
      "mean_token_accuracy": 0.8056312650442123,
      "num_tokens": 302344887.0,
      "step": 9473
    },
    {
      "entropy": 0.6262186537496746,
      "epoch": 0.8716683630341401,
      "grad_norm": 0.15904594957828522,
      "learning_rate": 7.094795596037661e-05,
      "loss": 0.6195,
      "mean_token_accuracy": 0.8143936358392239,
      "num_tokens": 302376036.0,
      "step": 9474
    },
    {
      "entropy": 0.6454485580325127,
      "epoch": 0.8717603694055813,
      "grad_norm": 0.1557372808456421,
      "learning_rate": 7.094488913423498e-05,
      "loss": 0.6202,
      "mean_token_accuracy": 0.8129967749118805,
      "num_tokens": 302408286.0,
      "step": 9475
    },
    {
      "entropy": 0.6696409042924643,
      "epoch": 0.8718523757770226,
      "grad_norm": 0.1508333683013916,
      "learning_rate": 7.094182230809336e-05,
      "loss": 0.6528,
      "mean_token_accuracy": 0.8086920753121376,
      "num_tokens": 302440266.0,
      "step": 9476
    },
    {
      "entropy": 0.6822449387982488,
      "epoch": 0.8719443821484638,
      "grad_norm": 0.1521729677915573,
      "learning_rate": 7.093875548195174e-05,
      "loss": 0.6549,
      "mean_token_accuracy": 0.8037151731550694,
      "num_tokens": 302471017.0,
      "step": 9477
    },
    {
      "entropy": 0.7009687088429928,
      "epoch": 0.872036388519905,
      "grad_norm": 0.14898468554019928,
      "learning_rate": 7.093568865581011e-05,
      "loss": 0.686,
      "mean_token_accuracy": 0.8025824874639511,
      "num_tokens": 302503447.0,
      "step": 9478
    },
    {
      "entropy": 0.7051727827638388,
      "epoch": 0.8721283948913462,
      "grad_norm": 0.14980155229568481,
      "learning_rate": 7.093262182966848e-05,
      "loss": 0.6974,
      "mean_token_accuracy": 0.7939973175525665,
      "num_tokens": 302536215.0,
      "step": 9479
    },
    {
      "entropy": 0.7416692562401295,
      "epoch": 0.8722204012627874,
      "grad_norm": 0.15849530696868896,
      "learning_rate": 7.092955500352684e-05,
      "loss": 0.7331,
      "mean_token_accuracy": 0.7852247841656208,
      "num_tokens": 302568007.0,
      "step": 9480
    },
    {
      "entropy": 0.7598546296358109,
      "epoch": 0.8723124076342287,
      "grad_norm": 0.16181041300296783,
      "learning_rate": 7.092648817738523e-05,
      "loss": 0.7489,
      "mean_token_accuracy": 0.7797464281320572,
      "num_tokens": 302599626.0,
      "step": 9481
    },
    {
      "entropy": 0.7342812791466713,
      "epoch": 0.8724044140056699,
      "grad_norm": 0.1571923792362213,
      "learning_rate": 7.092342135124361e-05,
      "loss": 0.7414,
      "mean_token_accuracy": 0.7795138843357563,
      "num_tokens": 302631806.0,
      "step": 9482
    },
    {
      "entropy": 0.6307021770626307,
      "epoch": 0.8724964203771112,
      "grad_norm": 0.14694327116012573,
      "learning_rate": 7.092035452510198e-05,
      "loss": 0.6169,
      "mean_token_accuracy": 0.8152629658579826,
      "num_tokens": 302663375.0,
      "step": 9483
    },
    {
      "entropy": 0.7866385309025645,
      "epoch": 0.8725884267485523,
      "grad_norm": 0.17012441158294678,
      "learning_rate": 7.091728769896034e-05,
      "loss": 0.7879,
      "mean_token_accuracy": 0.7737286984920502,
      "num_tokens": 302695871.0,
      "step": 9484
    },
    {
      "entropy": 0.8208592589944601,
      "epoch": 0.8726804331199935,
      "grad_norm": 0.15719197690486908,
      "learning_rate": 7.091422087281872e-05,
      "loss": 0.8054,
      "mean_token_accuracy": 0.7606742791831493,
      "num_tokens": 302727767.0,
      "step": 9485
    },
    {
      "entropy": 0.6976297805085778,
      "epoch": 0.8727724394914348,
      "grad_norm": 0.15792900323867798,
      "learning_rate": 7.091115404667709e-05,
      "loss": 0.6907,
      "mean_token_accuracy": 0.7941882386803627,
      "num_tokens": 302760066.0,
      "step": 9486
    },
    {
      "entropy": 0.6005914583802223,
      "epoch": 0.872864445862876,
      "grad_norm": 0.15012696385383606,
      "learning_rate": 7.090808722053547e-05,
      "loss": 0.5867,
      "mean_token_accuracy": 0.8238208778202534,
      "num_tokens": 302792125.0,
      "step": 9487
    },
    {
      "entropy": 0.7557874228805304,
      "epoch": 0.8729564522343173,
      "grad_norm": 0.15534748136997223,
      "learning_rate": 7.090502039439386e-05,
      "loss": 0.7298,
      "mean_token_accuracy": 0.7841922529041767,
      "num_tokens": 302824161.0,
      "step": 9488
    },
    {
      "entropy": 0.7220633719116449,
      "epoch": 0.8730484586057584,
      "grad_norm": 0.15059684216976166,
      "learning_rate": 7.090195356825222e-05,
      "loss": 0.7123,
      "mean_token_accuracy": 0.7838424891233444,
      "num_tokens": 302856369.0,
      "step": 9489
    },
    {
      "entropy": 0.8074232637882233,
      "epoch": 0.8731404649771997,
      "grad_norm": 0.165130615234375,
      "learning_rate": 7.089888674211059e-05,
      "loss": 0.7978,
      "mean_token_accuracy": 0.7648448422551155,
      "num_tokens": 302888490.0,
      "step": 9490
    },
    {
      "entropy": 0.6839218540117145,
      "epoch": 0.8732324713486409,
      "grad_norm": 0.1515226811170578,
      "learning_rate": 7.089581991596896e-05,
      "loss": 0.6795,
      "mean_token_accuracy": 0.8004763498902321,
      "num_tokens": 302920991.0,
      "step": 9491
    },
    {
      "entropy": 0.7039325479418039,
      "epoch": 0.8733244777200821,
      "grad_norm": 0.14803825318813324,
      "learning_rate": 7.089275308982734e-05,
      "loss": 0.6783,
      "mean_token_accuracy": 0.8033909387886524,
      "num_tokens": 302952852.0,
      "step": 9492
    },
    {
      "entropy": 0.7303701546043158,
      "epoch": 0.8734164840915234,
      "grad_norm": 0.1571943610906601,
      "learning_rate": 7.088968626368572e-05,
      "loss": 0.7177,
      "mean_token_accuracy": 0.7898227423429489,
      "num_tokens": 302985169.0,
      "step": 9493
    },
    {
      "entropy": 0.7497800588607788,
      "epoch": 0.8735084904629645,
      "grad_norm": 0.15568050742149353,
      "learning_rate": 7.088661943754409e-05,
      "loss": 0.734,
      "mean_token_accuracy": 0.7785241790115833,
      "num_tokens": 303016828.0,
      "step": 9494
    },
    {
      "entropy": 0.8248034520074725,
      "epoch": 0.8736004968344058,
      "grad_norm": 0.16465412080287933,
      "learning_rate": 7.088355261140246e-05,
      "loss": 0.8115,
      "mean_token_accuracy": 0.7644734270870686,
      "num_tokens": 303047837.0,
      "step": 9495
    },
    {
      "entropy": 0.627840319648385,
      "epoch": 0.873692503205847,
      "grad_norm": 0.15349248051643372,
      "learning_rate": 7.088048578526084e-05,
      "loss": 0.6159,
      "mean_token_accuracy": 0.8142041862010956,
      "num_tokens": 303079733.0,
      "step": 9496
    },
    {
      "entropy": 0.772350424900651,
      "epoch": 0.8737845095772883,
      "grad_norm": 0.16255183517932892,
      "learning_rate": 7.087741895911921e-05,
      "loss": 0.7685,
      "mean_token_accuracy": 0.7737188674509525,
      "num_tokens": 303111165.0,
      "step": 9497
    },
    {
      "entropy": 0.7049489915370941,
      "epoch": 0.8738765159487295,
      "grad_norm": 0.15161405503749847,
      "learning_rate": 7.087435213297759e-05,
      "loss": 0.6967,
      "mean_token_accuracy": 0.7890214063227177,
      "num_tokens": 303143047.0,
      "step": 9498
    },
    {
      "entropy": 0.6338754156604409,
      "epoch": 0.8739685223201706,
      "grad_norm": 0.1465035229921341,
      "learning_rate": 7.087128530683596e-05,
      "loss": 0.6281,
      "mean_token_accuracy": 0.8130107074975967,
      "num_tokens": 303175624.0,
      "step": 9499
    },
    {
      "entropy": 0.5907882805913687,
      "epoch": 0.8740605286916119,
      "grad_norm": 0.14817287027835846,
      "learning_rate": 7.086821848069434e-05,
      "loss": 0.5946,
      "mean_token_accuracy": 0.8173217549920082,
      "num_tokens": 303207194.0,
      "step": 9500
    },
    {
      "entropy": 0.7150478037074208,
      "epoch": 0.8741525350630531,
      "grad_norm": 0.15299271047115326,
      "learning_rate": 7.08651516545527e-05,
      "loss": 0.7247,
      "mean_token_accuracy": 0.7894177548587322,
      "num_tokens": 303239446.0,
      "step": 9501
    },
    {
      "entropy": 0.7302867900580168,
      "epoch": 0.8742445414344944,
      "grad_norm": 0.15194149315357208,
      "learning_rate": 7.086208482841107e-05,
      "loss": 0.7316,
      "mean_token_accuracy": 0.7842911146581173,
      "num_tokens": 303270824.0,
      "step": 9502
    },
    {
      "entropy": 0.7121652271598577,
      "epoch": 0.8743365478059356,
      "grad_norm": 0.151997908949852,
      "learning_rate": 7.085901800226945e-05,
      "loss": 0.6932,
      "mean_token_accuracy": 0.795502308756113,
      "num_tokens": 303303475.0,
      "step": 9503
    },
    {
      "entropy": 0.8328878059983253,
      "epoch": 0.8744285541773767,
      "grad_norm": 0.16419878602027893,
      "learning_rate": 7.085595117612784e-05,
      "loss": 0.8211,
      "mean_token_accuracy": 0.7633174434304237,
      "num_tokens": 303333182.0,
      "step": 9504
    },
    {
      "entropy": 0.79943189304322,
      "epoch": 0.874520560548818,
      "grad_norm": 0.1580924242734909,
      "learning_rate": 7.08528843499862e-05,
      "loss": 0.784,
      "mean_token_accuracy": 0.7668987177312374,
      "num_tokens": 303365586.0,
      "step": 9505
    },
    {
      "entropy": 0.8312976695597172,
      "epoch": 0.8746125669202592,
      "grad_norm": 0.15992482006549835,
      "learning_rate": 7.084981752384457e-05,
      "loss": 0.8125,
      "mean_token_accuracy": 0.7608719170093536,
      "num_tokens": 303396963.0,
      "step": 9506
    },
    {
      "entropy": 0.6103994213044643,
      "epoch": 0.8747045732917005,
      "grad_norm": 0.15517641603946686,
      "learning_rate": 7.084675069770294e-05,
      "loss": 0.5827,
      "mean_token_accuracy": 0.8206404112279415,
      "num_tokens": 303429133.0,
      "step": 9507
    },
    {
      "entropy": 0.7239912189543247,
      "epoch": 0.8747965796631417,
      "grad_norm": 0.15731917321681976,
      "learning_rate": 7.084368387156133e-05,
      "loss": 0.7247,
      "mean_token_accuracy": 0.7886646389961243,
      "num_tokens": 303460958.0,
      "step": 9508
    },
    {
      "entropy": 0.7113727852702141,
      "epoch": 0.8748885860345829,
      "grad_norm": 0.15942971408367157,
      "learning_rate": 7.08406170454197e-05,
      "loss": 0.7116,
      "mean_token_accuracy": 0.7900479920208454,
      "num_tokens": 303492760.0,
      "step": 9509
    },
    {
      "entropy": 0.6349211614578962,
      "epoch": 0.8749805924060241,
      "grad_norm": 0.154306098818779,
      "learning_rate": 7.083755021927807e-05,
      "loss": 0.6107,
      "mean_token_accuracy": 0.8134110309183598,
      "num_tokens": 303524079.0,
      "step": 9510
    },
    {
      "entropy": 0.7812917716801167,
      "epoch": 0.8750725987774653,
      "grad_norm": 0.16597890853881836,
      "learning_rate": 7.083448339313644e-05,
      "loss": 0.7678,
      "mean_token_accuracy": 0.7731347307562828,
      "num_tokens": 303555940.0,
      "step": 9511
    },
    {
      "entropy": 0.6509244758635759,
      "epoch": 0.8751646051489066,
      "grad_norm": 0.15379466116428375,
      "learning_rate": 7.083141656699482e-05,
      "loss": 0.6337,
      "mean_token_accuracy": 0.813650906085968,
      "num_tokens": 303587628.0,
      "step": 9512
    },
    {
      "entropy": 0.7601792849600315,
      "epoch": 0.8752566115203478,
      "grad_norm": 0.15863455832004547,
      "learning_rate": 7.08283497408532e-05,
      "loss": 0.7545,
      "mean_token_accuracy": 0.7763179279863834,
      "num_tokens": 303620039.0,
      "step": 9513
    },
    {
      "entropy": 0.6592027638107538,
      "epoch": 0.875348617891789,
      "grad_norm": 0.15765076875686646,
      "learning_rate": 7.082528291471157e-05,
      "loss": 0.6572,
      "mean_token_accuracy": 0.8023984767496586,
      "num_tokens": 303651341.0,
      "step": 9514
    },
    {
      "entropy": 0.7158920411020517,
      "epoch": 0.8754406242632302,
      "grad_norm": 0.16139797866344452,
      "learning_rate": 7.082221608856994e-05,
      "loss": 0.7184,
      "mean_token_accuracy": 0.7928129099309444,
      "num_tokens": 303683377.0,
      "step": 9515
    },
    {
      "entropy": 0.7304631043225527,
      "epoch": 0.8755326306346715,
      "grad_norm": 0.15378303825855255,
      "learning_rate": 7.081914926242832e-05,
      "loss": 0.7266,
      "mean_token_accuracy": 0.7834684029221535,
      "num_tokens": 303715598.0,
      "step": 9516
    },
    {
      "entropy": 0.5726066790521145,
      "epoch": 0.8756246370061127,
      "grad_norm": 0.14335691928863525,
      "learning_rate": 7.081608243628669e-05,
      "loss": 0.551,
      "mean_token_accuracy": 0.8371103517711163,
      "num_tokens": 303748072.0,
      "step": 9517
    },
    {
      "entropy": 0.6243804022669792,
      "epoch": 0.875716643377554,
      "grad_norm": 0.15344542264938354,
      "learning_rate": 7.081301561014507e-05,
      "loss": 0.5926,
      "mean_token_accuracy": 0.8210651613771915,
      "num_tokens": 303780397.0,
      "step": 9518
    },
    {
      "entropy": 0.6964067425578833,
      "epoch": 0.8758086497489951,
      "grad_norm": 0.15025892853736877,
      "learning_rate": 7.080994878400345e-05,
      "loss": 0.6894,
      "mean_token_accuracy": 0.7878014147281647,
      "num_tokens": 303812356.0,
      "step": 9519
    },
    {
      "entropy": 0.7476435974240303,
      "epoch": 0.8759006561204363,
      "grad_norm": 0.1675640195608139,
      "learning_rate": 7.080688195786182e-05,
      "loss": 0.7486,
      "mean_token_accuracy": 0.7771080583333969,
      "num_tokens": 303843675.0,
      "step": 9520
    },
    {
      "entropy": 0.7572608385235071,
      "epoch": 0.8759926624918776,
      "grad_norm": 0.15893831849098206,
      "learning_rate": 7.080381513172019e-05,
      "loss": 0.7543,
      "mean_token_accuracy": 0.7785063795745373,
      "num_tokens": 303874727.0,
      "step": 9521
    },
    {
      "entropy": 0.5448395553976297,
      "epoch": 0.8760846688633188,
      "grad_norm": 0.1518707573413849,
      "learning_rate": 7.080074830557855e-05,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8382115177810192,
      "num_tokens": 303906306.0,
      "step": 9522
    },
    {
      "entropy": 0.7088436875492334,
      "epoch": 0.8761766752347601,
      "grad_norm": 0.15108579397201538,
      "learning_rate": 7.079768147943693e-05,
      "loss": 0.6995,
      "mean_token_accuracy": 0.7904795184731483,
      "num_tokens": 303937922.0,
      "step": 9523
    },
    {
      "entropy": 0.6687766704708338,
      "epoch": 0.8762686816062012,
      "grad_norm": 0.15074141323566437,
      "learning_rate": 7.079461465329532e-05,
      "loss": 0.6559,
      "mean_token_accuracy": 0.8020180836319923,
      "num_tokens": 303969414.0,
      "step": 9524
    },
    {
      "entropy": 0.7106018010526896,
      "epoch": 0.8763606879776424,
      "grad_norm": 0.16008706390857697,
      "learning_rate": 7.079154782715368e-05,
      "loss": 0.7008,
      "mean_token_accuracy": 0.7958316840231419,
      "num_tokens": 304001595.0,
      "step": 9525
    },
    {
      "entropy": 0.6645860392600298,
      "epoch": 0.8764526943490837,
      "grad_norm": 0.15013879537582397,
      "learning_rate": 7.078848100101205e-05,
      "loss": 0.6435,
      "mean_token_accuracy": 0.8059172332286835,
      "num_tokens": 304033319.0,
      "step": 9526
    },
    {
      "entropy": 0.6375437248498201,
      "epoch": 0.8765447007205249,
      "grad_norm": 0.1546163409948349,
      "learning_rate": 7.078541417487043e-05,
      "loss": 0.637,
      "mean_token_accuracy": 0.809252142906189,
      "num_tokens": 304065807.0,
      "step": 9527
    },
    {
      "entropy": 0.6969247292727232,
      "epoch": 0.8766367070919662,
      "grad_norm": 0.15995441377162933,
      "learning_rate": 7.07823473487288e-05,
      "loss": 0.6783,
      "mean_token_accuracy": 0.7970060482621193,
      "num_tokens": 304098160.0,
      "step": 9528
    },
    {
      "entropy": 0.6319955633953214,
      "epoch": 0.8767287134634073,
      "grad_norm": 0.15977586805820465,
      "learning_rate": 7.077928052258718e-05,
      "loss": 0.6376,
      "mean_token_accuracy": 0.8142426311969757,
      "num_tokens": 304129956.0,
      "step": 9529
    },
    {
      "entropy": 0.7331339698284864,
      "epoch": 0.8768207198348485,
      "grad_norm": 0.1546429991722107,
      "learning_rate": 7.077621369644555e-05,
      "loss": 0.7196,
      "mean_token_accuracy": 0.7916033565998077,
      "num_tokens": 304161372.0,
      "step": 9530
    },
    {
      "entropy": 0.6793500334024429,
      "epoch": 0.8769127262062898,
      "grad_norm": 0.15140970051288605,
      "learning_rate": 7.077314687030393e-05,
      "loss": 0.6807,
      "mean_token_accuracy": 0.8008390516042709,
      "num_tokens": 304192704.0,
      "step": 9531
    },
    {
      "entropy": 0.7725203596055508,
      "epoch": 0.877004732577731,
      "grad_norm": 0.16519008576869965,
      "learning_rate": 7.07700800441623e-05,
      "loss": 0.7693,
      "mean_token_accuracy": 0.7742812857031822,
      "num_tokens": 304224470.0,
      "step": 9532
    },
    {
      "entropy": 0.6746007576584816,
      "epoch": 0.8770967389491723,
      "grad_norm": 0.16027650237083435,
      "learning_rate": 7.076701321802067e-05,
      "loss": 0.6422,
      "mean_token_accuracy": 0.8093331754207611,
      "num_tokens": 304256645.0,
      "step": 9533
    },
    {
      "entropy": 0.6952027110382915,
      "epoch": 0.8771887453206134,
      "grad_norm": 0.15714499354362488,
      "learning_rate": 7.076394639187905e-05,
      "loss": 0.6863,
      "mean_token_accuracy": 0.7975052148103714,
      "num_tokens": 304289177.0,
      "step": 9534
    },
    {
      "entropy": 0.7841057237237692,
      "epoch": 0.8772807516920547,
      "grad_norm": 0.15390750765800476,
      "learning_rate": 7.076087956573743e-05,
      "loss": 0.7577,
      "mean_token_accuracy": 0.7760691493749619,
      "num_tokens": 304321032.0,
      "step": 9535
    },
    {
      "entropy": 0.6576796937733889,
      "epoch": 0.8773727580634959,
      "grad_norm": 0.15209747850894928,
      "learning_rate": 7.07578127395958e-05,
      "loss": 0.6382,
      "mean_token_accuracy": 0.8096005134284496,
      "num_tokens": 304353254.0,
      "step": 9536
    },
    {
      "entropy": 0.5831970311701298,
      "epoch": 0.8774647644349372,
      "grad_norm": 0.14003942906856537,
      "learning_rate": 7.075474591345417e-05,
      "loss": 0.5493,
      "mean_token_accuracy": 0.8314490914344788,
      "num_tokens": 304384977.0,
      "step": 9537
    },
    {
      "entropy": 0.8488270658999681,
      "epoch": 0.8775567708063784,
      "grad_norm": 0.15711307525634766,
      "learning_rate": 7.075167908731253e-05,
      "loss": 0.8387,
      "mean_token_accuracy": 0.7548383884131908,
      "num_tokens": 304416693.0,
      "step": 9538
    },
    {
      "entropy": 0.8086378648877144,
      "epoch": 0.8776487771778195,
      "grad_norm": 0.1660158336162567,
      "learning_rate": 7.074861226117092e-05,
      "loss": 0.8098,
      "mean_token_accuracy": 0.7595930621027946,
      "num_tokens": 304446535.0,
      "step": 9539
    },
    {
      "entropy": 0.650096689350903,
      "epoch": 0.8777407835492608,
      "grad_norm": 0.1478722244501114,
      "learning_rate": 7.07455454350293e-05,
      "loss": 0.6471,
      "mean_token_accuracy": 0.8055305071175098,
      "num_tokens": 304478807.0,
      "step": 9540
    },
    {
      "entropy": 0.7322646770626307,
      "epoch": 0.877832789920702,
      "grad_norm": 0.15152607858181,
      "learning_rate": 7.074247860888766e-05,
      "loss": 0.7127,
      "mean_token_accuracy": 0.7860793098807335,
      "num_tokens": 304510853.0,
      "step": 9541
    },
    {
      "entropy": 0.7049979325383902,
      "epoch": 0.8779247962921433,
      "grad_norm": 0.16131335496902466,
      "learning_rate": 7.073941178274603e-05,
      "loss": 0.688,
      "mean_token_accuracy": 0.7958362214267254,
      "num_tokens": 304541843.0,
      "step": 9542
    },
    {
      "entropy": 0.6153056267648935,
      "epoch": 0.8780168026635845,
      "grad_norm": 0.14713409543037415,
      "learning_rate": 7.073634495660441e-05,
      "loss": 0.5953,
      "mean_token_accuracy": 0.8205070532858372,
      "num_tokens": 304574066.0,
      "step": 9543
    },
    {
      "entropy": 0.695370988920331,
      "epoch": 0.8781088090350256,
      "grad_norm": 0.16348977386951447,
      "learning_rate": 7.073327813046278e-05,
      "loss": 0.6887,
      "mean_token_accuracy": 0.7939359992742538,
      "num_tokens": 304606003.0,
      "step": 9544
    },
    {
      "entropy": 0.6828480269759893,
      "epoch": 0.8782008154064669,
      "grad_norm": 0.15607891976833344,
      "learning_rate": 7.073021130432116e-05,
      "loss": 0.6714,
      "mean_token_accuracy": 0.7972989045083523,
      "num_tokens": 304638102.0,
      "step": 9545
    },
    {
      "entropy": 0.7219934361055493,
      "epoch": 0.8782928217779081,
      "grad_norm": 0.15484113991260529,
      "learning_rate": 7.072714447817953e-05,
      "loss": 0.7179,
      "mean_token_accuracy": 0.7887686230242252,
      "num_tokens": 304670603.0,
      "step": 9546
    },
    {
      "entropy": 0.6545900348573923,
      "epoch": 0.8783848281493494,
      "grad_norm": 0.15405647456645966,
      "learning_rate": 7.072407765203791e-05,
      "loss": 0.6474,
      "mean_token_accuracy": 0.8071141727268696,
      "num_tokens": 304702427.0,
      "step": 9547
    },
    {
      "entropy": 0.7335512097924948,
      "epoch": 0.8784768345207906,
      "grad_norm": 0.16458065807819366,
      "learning_rate": 7.072101082589628e-05,
      "loss": 0.7187,
      "mean_token_accuracy": 0.7906699627637863,
      "num_tokens": 304734417.0,
      "step": 9548
    },
    {
      "entropy": 0.6286568231880665,
      "epoch": 0.8785688408922318,
      "grad_norm": 0.16361847519874573,
      "learning_rate": 7.071794399975465e-05,
      "loss": 0.6187,
      "mean_token_accuracy": 0.8142566941678524,
      "num_tokens": 304766931.0,
      "step": 9549
    },
    {
      "entropy": 0.7094953618943691,
      "epoch": 0.878660847263673,
      "grad_norm": 0.15366291999816895,
      "learning_rate": 7.071487717361304e-05,
      "loss": 0.6924,
      "mean_token_accuracy": 0.79409234598279,
      "num_tokens": 304799064.0,
      "step": 9550
    },
    {
      "entropy": 0.7175127118825912,
      "epoch": 0.8787528536351142,
      "grad_norm": 0.15633535385131836,
      "learning_rate": 7.071181034747141e-05,
      "loss": 0.6977,
      "mean_token_accuracy": 0.7924710810184479,
      "num_tokens": 304831003.0,
      "step": 9551
    },
    {
      "entropy": 0.7346137873828411,
      "epoch": 0.8788448600065555,
      "grad_norm": 0.15282303094863892,
      "learning_rate": 7.070874352132978e-05,
      "loss": 0.7248,
      "mean_token_accuracy": 0.7809320874512196,
      "num_tokens": 304862758.0,
      "step": 9552
    },
    {
      "entropy": 0.6817892752587795,
      "epoch": 0.8789368663779967,
      "grad_norm": 0.15604519844055176,
      "learning_rate": 7.070567669518815e-05,
      "loss": 0.6618,
      "mean_token_accuracy": 0.7998245432972908,
      "num_tokens": 304894622.0,
      "step": 9553
    },
    {
      "entropy": 0.723484443500638,
      "epoch": 0.8790288727494379,
      "grad_norm": 0.16464875638484955,
      "learning_rate": 7.070260986904653e-05,
      "loss": 0.7041,
      "mean_token_accuracy": 0.7906196415424347,
      "num_tokens": 304925536.0,
      "step": 9554
    },
    {
      "entropy": 0.697942391037941,
      "epoch": 0.8791208791208791,
      "grad_norm": 0.15247201919555664,
      "learning_rate": 7.069954304290491e-05,
      "loss": 0.6786,
      "mean_token_accuracy": 0.7965099066495895,
      "num_tokens": 304957455.0,
      "step": 9555
    },
    {
      "entropy": 0.6624684957787395,
      "epoch": 0.8792128854923204,
      "grad_norm": 0.169123575091362,
      "learning_rate": 7.069647621676328e-05,
      "loss": 0.6768,
      "mean_token_accuracy": 0.7993257828056812,
      "num_tokens": 304989255.0,
      "step": 9556
    },
    {
      "entropy": 0.6705937888473272,
      "epoch": 0.8793048918637616,
      "grad_norm": 0.1642393320798874,
      "learning_rate": 7.069340939062165e-05,
      "loss": 0.6844,
      "mean_token_accuracy": 0.7943196631968021,
      "num_tokens": 305021235.0,
      "step": 9557
    },
    {
      "entropy": 0.700545059517026,
      "epoch": 0.8793968982352028,
      "grad_norm": 0.15919914841651917,
      "learning_rate": 7.069034256448003e-05,
      "loss": 0.6855,
      "mean_token_accuracy": 0.7980202548205853,
      "num_tokens": 305052778.0,
      "step": 9558
    },
    {
      "entropy": 0.7639635377563536,
      "epoch": 0.879488904606644,
      "grad_norm": 0.15614260733127594,
      "learning_rate": 7.06872757383384e-05,
      "loss": 0.7606,
      "mean_token_accuracy": 0.7758842557668686,
      "num_tokens": 305085434.0,
      "step": 9559
    },
    {
      "entropy": 0.842604722827673,
      "epoch": 0.8795809109780852,
      "grad_norm": 0.17533108592033386,
      "learning_rate": 7.068420891219678e-05,
      "loss": 0.8526,
      "mean_token_accuracy": 0.7539642453193665,
      "num_tokens": 305117143.0,
      "step": 9560
    },
    {
      "entropy": 0.6878068819642067,
      "epoch": 0.8796729173495265,
      "grad_norm": 0.14985522627830505,
      "learning_rate": 7.068114208605514e-05,
      "loss": 0.672,
      "mean_token_accuracy": 0.8000341206789017,
      "num_tokens": 305148829.0,
      "step": 9561
    },
    {
      "entropy": 0.7177100293338299,
      "epoch": 0.8797649237209677,
      "grad_norm": 0.16333860158920288,
      "learning_rate": 7.067807525991353e-05,
      "loss": 0.686,
      "mean_token_accuracy": 0.7952856086194515,
      "num_tokens": 305181026.0,
      "step": 9562
    },
    {
      "entropy": 0.7882575541734695,
      "epoch": 0.879856930092409,
      "grad_norm": 0.1619870960712433,
      "learning_rate": 7.067500843377189e-05,
      "loss": 0.7757,
      "mean_token_accuracy": 0.7745972685515881,
      "num_tokens": 305213107.0,
      "step": 9563
    },
    {
      "entropy": 0.7209370099008083,
      "epoch": 0.8799489364638501,
      "grad_norm": 0.14826011657714844,
      "learning_rate": 7.067194160763026e-05,
      "loss": 0.7027,
      "mean_token_accuracy": 0.7890089526772499,
      "num_tokens": 305245819.0,
      "step": 9564
    },
    {
      "entropy": 0.6584218647330999,
      "epoch": 0.8800409428352913,
      "grad_norm": 0.15914234519004822,
      "learning_rate": 7.066887478148864e-05,
      "loss": 0.6378,
      "mean_token_accuracy": 0.8086672201752663,
      "num_tokens": 305277585.0,
      "step": 9565
    },
    {
      "entropy": 0.711791854351759,
      "epoch": 0.8801329492067326,
      "grad_norm": 0.15279118716716766,
      "learning_rate": 7.066580795534702e-05,
      "loss": 0.7018,
      "mean_token_accuracy": 0.7897481620311737,
      "num_tokens": 305309491.0,
      "step": 9566
    },
    {
      "entropy": 0.6851588087156415,
      "epoch": 0.8802249555781738,
      "grad_norm": 0.1519990861415863,
      "learning_rate": 7.066274112920539e-05,
      "loss": 0.6698,
      "mean_token_accuracy": 0.7983748987317085,
      "num_tokens": 305341455.0,
      "step": 9567
    },
    {
      "entropy": 0.7258566878736019,
      "epoch": 0.8803169619496151,
      "grad_norm": 0.15734143555164337,
      "learning_rate": 7.065967430306376e-05,
      "loss": 0.7146,
      "mean_token_accuracy": 0.7904340252280235,
      "num_tokens": 305372176.0,
      "step": 9568
    },
    {
      "entropy": 0.7794321831315756,
      "epoch": 0.8804089683210562,
      "grad_norm": 0.15431247651576996,
      "learning_rate": 7.065660747692213e-05,
      "loss": 0.7717,
      "mean_token_accuracy": 0.7700676396489143,
      "num_tokens": 305403372.0,
      "step": 9569
    },
    {
      "entropy": 0.7341389786452055,
      "epoch": 0.8805009746924974,
      "grad_norm": 0.15281113982200623,
      "learning_rate": 7.065354065078051e-05,
      "loss": 0.7194,
      "mean_token_accuracy": 0.784868448972702,
      "num_tokens": 305435996.0,
      "step": 9570
    },
    {
      "entropy": 0.6772367209196091,
      "epoch": 0.8805929810639387,
      "grad_norm": 0.15415403246879578,
      "learning_rate": 7.065047382463889e-05,
      "loss": 0.6655,
      "mean_token_accuracy": 0.7964773736894131,
      "num_tokens": 305468496.0,
      "step": 9571
    },
    {
      "entropy": 0.772213576361537,
      "epoch": 0.8806849874353799,
      "grad_norm": 0.15304136276245117,
      "learning_rate": 7.064740699849726e-05,
      "loss": 0.7689,
      "mean_token_accuracy": 0.7759356983006,
      "num_tokens": 305500680.0,
      "step": 9572
    },
    {
      "entropy": 0.6494280062615871,
      "epoch": 0.8807769938068212,
      "grad_norm": 0.15364941954612732,
      "learning_rate": 7.064434017235563e-05,
      "loss": 0.6396,
      "mean_token_accuracy": 0.8086218014359474,
      "num_tokens": 305532418.0,
      "step": 9573
    },
    {
      "entropy": 0.7636018097400665,
      "epoch": 0.8808690001782623,
      "grad_norm": 0.1532343626022339,
      "learning_rate": 7.064127334621401e-05,
      "loss": 0.7641,
      "mean_token_accuracy": 0.7767276838421822,
      "num_tokens": 305563953.0,
      "step": 9574
    },
    {
      "entropy": 0.8379837796092033,
      "epoch": 0.8809610065497036,
      "grad_norm": 0.17686010897159576,
      "learning_rate": 7.063820652007238e-05,
      "loss": 0.8228,
      "mean_token_accuracy": 0.7615292631089687,
      "num_tokens": 305595217.0,
      "step": 9575
    },
    {
      "entropy": 0.7482713218778372,
      "epoch": 0.8810530129211448,
      "grad_norm": 0.16134639084339142,
      "learning_rate": 7.063513969393076e-05,
      "loss": 0.7419,
      "mean_token_accuracy": 0.7820266485214233,
      "num_tokens": 305626796.0,
      "step": 9576
    },
    {
      "entropy": 0.7349460925906897,
      "epoch": 0.881145019292586,
      "grad_norm": 0.1599193513393402,
      "learning_rate": 7.063207286778912e-05,
      "loss": 0.743,
      "mean_token_accuracy": 0.7772751338779926,
      "num_tokens": 305658647.0,
      "step": 9577
    },
    {
      "entropy": 0.6220851093530655,
      "epoch": 0.8812370256640273,
      "grad_norm": 0.1495019644498825,
      "learning_rate": 7.06290060416475e-05,
      "loss": 0.5915,
      "mean_token_accuracy": 0.8149818181991577,
      "num_tokens": 305689594.0,
      "step": 9578
    },
    {
      "entropy": 0.7069403622299433,
      "epoch": 0.8813290320354684,
      "grad_norm": 0.15599536895751953,
      "learning_rate": 7.062593921550587e-05,
      "loss": 0.7127,
      "mean_token_accuracy": 0.7884751707315445,
      "num_tokens": 305720514.0,
      "step": 9579
    },
    {
      "entropy": 0.7973044961690903,
      "epoch": 0.8814210384069097,
      "grad_norm": 0.1660662293434143,
      "learning_rate": 7.062287238936424e-05,
      "loss": 0.7929,
      "mean_token_accuracy": 0.7679124400019646,
      "num_tokens": 305752609.0,
      "step": 9580
    },
    {
      "entropy": 0.6725745461881161,
      "epoch": 0.8815130447783509,
      "grad_norm": 0.15989981591701508,
      "learning_rate": 7.061980556322262e-05,
      "loss": 0.6736,
      "mean_token_accuracy": 0.7972501367330551,
      "num_tokens": 305784120.0,
      "step": 9581
    },
    {
      "entropy": 0.7357601672410965,
      "epoch": 0.8816050511497922,
      "grad_norm": 0.15187720954418182,
      "learning_rate": 7.0616738737081e-05,
      "loss": 0.7201,
      "mean_token_accuracy": 0.7890167385339737,
      "num_tokens": 305816411.0,
      "step": 9582
    },
    {
      "entropy": 0.7976577691733837,
      "epoch": 0.8816970575212334,
      "grad_norm": 0.16306032240390778,
      "learning_rate": 7.061367191093937e-05,
      "loss": 0.7851,
      "mean_token_accuracy": 0.769863173365593,
      "num_tokens": 305848448.0,
      "step": 9583
    },
    {
      "entropy": 0.7200176101177931,
      "epoch": 0.8817890638926745,
      "grad_norm": 0.15728668868541718,
      "learning_rate": 7.061060508479774e-05,
      "loss": 0.7086,
      "mean_token_accuracy": 0.7898626551032066,
      "num_tokens": 305880723.0,
      "step": 9584
    },
    {
      "entropy": 0.6750348499044776,
      "epoch": 0.8818810702641158,
      "grad_norm": 0.15257231891155243,
      "learning_rate": 7.060753825865612e-05,
      "loss": 0.6685,
      "mean_token_accuracy": 0.8024095892906189,
      "num_tokens": 305912753.0,
      "step": 9585
    },
    {
      "entropy": 0.7211212618276477,
      "epoch": 0.881973076635557,
      "grad_norm": 0.16150115430355072,
      "learning_rate": 7.060447143251449e-05,
      "loss": 0.6981,
      "mean_token_accuracy": 0.7975541763007641,
      "num_tokens": 305944994.0,
      "step": 9586
    },
    {
      "entropy": 0.7207027338445187,
      "epoch": 0.8820650830069983,
      "grad_norm": 0.15708687901496887,
      "learning_rate": 7.060140460637287e-05,
      "loss": 0.7042,
      "mean_token_accuracy": 0.786486316472292,
      "num_tokens": 305977762.0,
      "step": 9587
    },
    {
      "entropy": 0.6825596634298563,
      "epoch": 0.8821570893784395,
      "grad_norm": 0.15406645834445953,
      "learning_rate": 7.059833778023124e-05,
      "loss": 0.6765,
      "mean_token_accuracy": 0.8008824326097965,
      "num_tokens": 306009898.0,
      "step": 9588
    },
    {
      "entropy": 0.631110624410212,
      "epoch": 0.8822490957498806,
      "grad_norm": 0.15859338641166687,
      "learning_rate": 7.059527095408962e-05,
      "loss": 0.6051,
      "mean_token_accuracy": 0.8166765533387661,
      "num_tokens": 306041518.0,
      "step": 9589
    },
    {
      "entropy": 0.7579452823847532,
      "epoch": 0.8823411021213219,
      "grad_norm": 0.15527045726776123,
      "learning_rate": 7.059220412794799e-05,
      "loss": 0.7543,
      "mean_token_accuracy": 0.7754621207714081,
      "num_tokens": 306072798.0,
      "step": 9590
    },
    {
      "entropy": 0.6816126480698586,
      "epoch": 0.8824331084927631,
      "grad_norm": 0.14849595725536346,
      "learning_rate": 7.058913730180636e-05,
      "loss": 0.6508,
      "mean_token_accuracy": 0.8018495924770832,
      "num_tokens": 306103540.0,
      "step": 9591
    },
    {
      "entropy": 0.6608305121771991,
      "epoch": 0.8825251148642044,
      "grad_norm": 0.15592695772647858,
      "learning_rate": 7.058607047566474e-05,
      "loss": 0.6346,
      "mean_token_accuracy": 0.8111604452133179,
      "num_tokens": 306135674.0,
      "step": 9592
    },
    {
      "entropy": 0.7069302629679441,
      "epoch": 0.8826171212356456,
      "grad_norm": 0.1518353372812271,
      "learning_rate": 7.058300364952312e-05,
      "loss": 0.6839,
      "mean_token_accuracy": 0.7966080754995346,
      "num_tokens": 306167878.0,
      "step": 9593
    },
    {
      "entropy": 0.5615766393020749,
      "epoch": 0.8827091276070868,
      "grad_norm": 0.15556932985782623,
      "learning_rate": 7.057993682338149e-05,
      "loss": 0.5397,
      "mean_token_accuracy": 0.8409388288855553,
      "num_tokens": 306200468.0,
      "step": 9594
    },
    {
      "entropy": 0.8224303536117077,
      "epoch": 0.882801133978528,
      "grad_norm": 0.16242380440235138,
      "learning_rate": 7.057686999723985e-05,
      "loss": 0.8171,
      "mean_token_accuracy": 0.7578083872795105,
      "num_tokens": 306232299.0,
      "step": 9595
    },
    {
      "entropy": 0.7419270575046539,
      "epoch": 0.8828931403499692,
      "grad_norm": 0.15116052329540253,
      "learning_rate": 7.057380317109824e-05,
      "loss": 0.7228,
      "mean_token_accuracy": 0.7841936200857162,
      "num_tokens": 306264380.0,
      "step": 9596
    },
    {
      "entropy": 0.6420835200697184,
      "epoch": 0.8829851467214105,
      "grad_norm": 0.1500396430492401,
      "learning_rate": 7.057073634495662e-05,
      "loss": 0.6345,
      "mean_token_accuracy": 0.8108150362968445,
      "num_tokens": 306296526.0,
      "step": 9597
    },
    {
      "entropy": 0.6646807929500937,
      "epoch": 0.8830771530928517,
      "grad_norm": 0.16234050691127777,
      "learning_rate": 7.056766951881499e-05,
      "loss": 0.6606,
      "mean_token_accuracy": 0.8077064156532288,
      "num_tokens": 306328974.0,
      "step": 9598
    },
    {
      "entropy": 0.5449506044387817,
      "epoch": 0.8831691594642929,
      "grad_norm": 0.1515217423439026,
      "learning_rate": 7.056460269267335e-05,
      "loss": 0.5176,
      "mean_token_accuracy": 0.8450939990580082,
      "num_tokens": 306361252.0,
      "step": 9599
    },
    {
      "entropy": 0.7388967173174024,
      "epoch": 0.8832611658357341,
      "grad_norm": 0.16020016372203827,
      "learning_rate": 7.056153586653172e-05,
      "loss": 0.7474,
      "mean_token_accuracy": 0.7790093123912811,
      "num_tokens": 306393779.0,
      "step": 9600
    },
    {
      "entropy": 0.5780046246945858,
      "epoch": 0.8833531722071754,
      "grad_norm": 0.14695534110069275,
      "learning_rate": 7.05584690403901e-05,
      "loss": 0.562,
      "mean_token_accuracy": 0.8340211138129234,
      "num_tokens": 306425696.0,
      "step": 9601
    },
    {
      "entropy": 0.7053318284451962,
      "epoch": 0.8834451785786166,
      "grad_norm": 0.17010842263698578,
      "learning_rate": 7.055540221424848e-05,
      "loss": 0.7127,
      "mean_token_accuracy": 0.786475881934166,
      "num_tokens": 306458023.0,
      "step": 9602
    },
    {
      "entropy": 0.7090697921812534,
      "epoch": 0.8835371849500578,
      "grad_norm": 0.17311067879199982,
      "learning_rate": 7.055233538810685e-05,
      "loss": 0.7047,
      "mean_token_accuracy": 0.7845826670527458,
      "num_tokens": 306490241.0,
      "step": 9603
    },
    {
      "entropy": 0.6207144651561975,
      "epoch": 0.883629191321499,
      "grad_norm": 0.14848639070987701,
      "learning_rate": 7.054926856196522e-05,
      "loss": 0.608,
      "mean_token_accuracy": 0.8195124082267284,
      "num_tokens": 306521619.0,
      "step": 9604
    },
    {
      "entropy": 0.7567181400954723,
      "epoch": 0.8837211976929402,
      "grad_norm": 0.15678168833255768,
      "learning_rate": 7.05462017358236e-05,
      "loss": 0.7598,
      "mean_token_accuracy": 0.7753489427268505,
      "num_tokens": 306553935.0,
      "step": 9605
    },
    {
      "entropy": 0.7205708604305983,
      "epoch": 0.8838132040643815,
      "grad_norm": 0.15944786369800568,
      "learning_rate": 7.054313490968197e-05,
      "loss": 0.723,
      "mean_token_accuracy": 0.7871533744037151,
      "num_tokens": 306585324.0,
      "step": 9606
    },
    {
      "entropy": 0.685234347358346,
      "epoch": 0.8839052104358227,
      "grad_norm": 0.15606766939163208,
      "learning_rate": 7.054006808354035e-05,
      "loss": 0.676,
      "mean_token_accuracy": 0.7969435527920723,
      "num_tokens": 306617282.0,
      "step": 9607
    },
    {
      "entropy": 0.7071074079722166,
      "epoch": 0.883997216807264,
      "grad_norm": 0.16143502295017242,
      "learning_rate": 7.053700125739872e-05,
      "loss": 0.6727,
      "mean_token_accuracy": 0.7973864190280437,
      "num_tokens": 306648515.0,
      "step": 9608
    },
    {
      "entropy": 0.6476562339812517,
      "epoch": 0.8840892231787051,
      "grad_norm": 0.16689592599868774,
      "learning_rate": 7.05339344312571e-05,
      "loss": 0.6442,
      "mean_token_accuracy": 0.8100879341363907,
      "num_tokens": 306681283.0,
      "step": 9609
    },
    {
      "entropy": 0.8510579708963633,
      "epoch": 0.8841812295501463,
      "grad_norm": 0.1565975844860077,
      "learning_rate": 7.053086760511547e-05,
      "loss": 0.8218,
      "mean_token_accuracy": 0.7536366172134876,
      "num_tokens": 306713300.0,
      "step": 9610
    },
    {
      "entropy": 0.7068846505135298,
      "epoch": 0.8842732359215876,
      "grad_norm": 0.15173284709453583,
      "learning_rate": 7.052780077897384e-05,
      "loss": 0.6873,
      "mean_token_accuracy": 0.7952795773744583,
      "num_tokens": 306744987.0,
      "step": 9611
    },
    {
      "entropy": 0.7015085322782397,
      "epoch": 0.8843652422930288,
      "grad_norm": 0.1513458639383316,
      "learning_rate": 7.052473395283222e-05,
      "loss": 0.6731,
      "mean_token_accuracy": 0.8009830936789513,
      "num_tokens": 306776767.0,
      "step": 9612
    },
    {
      "entropy": 0.6880677714943886,
      "epoch": 0.8844572486644701,
      "grad_norm": 0.15192002058029175,
      "learning_rate": 7.05216671266906e-05,
      "loss": 0.6735,
      "mean_token_accuracy": 0.8005766123533249,
      "num_tokens": 306808226.0,
      "step": 9613
    },
    {
      "entropy": 0.7348237801343203,
      "epoch": 0.8845492550359112,
      "grad_norm": 0.15143539011478424,
      "learning_rate": 7.051860030054897e-05,
      "loss": 0.7205,
      "mean_token_accuracy": 0.7876403294503689,
      "num_tokens": 306840793.0,
      "step": 9614
    },
    {
      "entropy": 0.8486295025795698,
      "epoch": 0.8846412614073524,
      "grad_norm": 0.1617109626531601,
      "learning_rate": 7.051553347440733e-05,
      "loss": 0.8454,
      "mean_token_accuracy": 0.7556461542844772,
      "num_tokens": 306873217.0,
      "step": 9615
    },
    {
      "entropy": 0.7039809357374907,
      "epoch": 0.8847332677787937,
      "grad_norm": 0.15391035377979279,
      "learning_rate": 7.051246664826572e-05,
      "loss": 0.6919,
      "mean_token_accuracy": 0.7946344502270222,
      "num_tokens": 306905480.0,
      "step": 9616
    },
    {
      "entropy": 0.6885203514248133,
      "epoch": 0.8848252741502349,
      "grad_norm": 0.15029747784137726,
      "learning_rate": 7.050939982212408e-05,
      "loss": 0.6604,
      "mean_token_accuracy": 0.8015260472893715,
      "num_tokens": 306937649.0,
      "step": 9617
    },
    {
      "entropy": 0.7678579147905111,
      "epoch": 0.8849172805216762,
      "grad_norm": 0.15394465625286102,
      "learning_rate": 7.050633299598246e-05,
      "loss": 0.7411,
      "mean_token_accuracy": 0.7791832126677036,
      "num_tokens": 306969296.0,
      "step": 9618
    },
    {
      "entropy": 0.6200787276029587,
      "epoch": 0.8850092868931173,
      "grad_norm": 0.14847645163536072,
      "learning_rate": 7.050326616984083e-05,
      "loss": 0.6154,
      "mean_token_accuracy": 0.8171067722141743,
      "num_tokens": 307001429.0,
      "step": 9619
    },
    {
      "entropy": 0.7037163451313972,
      "epoch": 0.8851012932645586,
      "grad_norm": 0.1482899785041809,
      "learning_rate": 7.050019934369921e-05,
      "loss": 0.6964,
      "mean_token_accuracy": 0.7918742038309574,
      "num_tokens": 307033639.0,
      "step": 9620
    },
    {
      "entropy": 0.6921079633757472,
      "epoch": 0.8851932996359998,
      "grad_norm": 0.15904779732227325,
      "learning_rate": 7.049713251755758e-05,
      "loss": 0.6755,
      "mean_token_accuracy": 0.7942922227084637,
      "num_tokens": 307064550.0,
      "step": 9621
    },
    {
      "entropy": 0.6354828104376793,
      "epoch": 0.885285306007441,
      "grad_norm": 0.15782587230205536,
      "learning_rate": 7.049406569141595e-05,
      "loss": 0.6261,
      "mean_token_accuracy": 0.8133354783058167,
      "num_tokens": 307096193.0,
      "step": 9622
    },
    {
      "entropy": 0.631239041686058,
      "epoch": 0.8853773123788823,
      "grad_norm": 0.1499142050743103,
      "learning_rate": 7.049099886527433e-05,
      "loss": 0.6183,
      "mean_token_accuracy": 0.8168279752135277,
      "num_tokens": 307128426.0,
      "step": 9623
    },
    {
      "entropy": 0.7271452508866787,
      "epoch": 0.8854693187503234,
      "grad_norm": 0.15853729844093323,
      "learning_rate": 7.048793203913271e-05,
      "loss": 0.7235,
      "mean_token_accuracy": 0.7892913706600666,
      "num_tokens": 307160577.0,
      "step": 9624
    },
    {
      "entropy": 0.7904794998466969,
      "epoch": 0.8855613251217647,
      "grad_norm": 0.1556849628686905,
      "learning_rate": 7.048486521299108e-05,
      "loss": 0.7855,
      "mean_token_accuracy": 0.7649829909205437,
      "num_tokens": 307192343.0,
      "step": 9625
    },
    {
      "entropy": 0.6398868421092629,
      "epoch": 0.8856533314932059,
      "grad_norm": 0.15467418730258942,
      "learning_rate": 7.048179838684945e-05,
      "loss": 0.6325,
      "mean_token_accuracy": 0.8111378140747547,
      "num_tokens": 307224596.0,
      "step": 9626
    },
    {
      "entropy": 0.6884300597012043,
      "epoch": 0.8857453378646472,
      "grad_norm": 0.15638403594493866,
      "learning_rate": 7.047873156070782e-05,
      "loss": 0.6921,
      "mean_token_accuracy": 0.7944143079221249,
      "num_tokens": 307257081.0,
      "step": 9627
    },
    {
      "entropy": 0.7966320477426052,
      "epoch": 0.8858373442360884,
      "grad_norm": 0.16164684295654297,
      "learning_rate": 7.047566473456621e-05,
      "loss": 0.7902,
      "mean_token_accuracy": 0.7674634791910648,
      "num_tokens": 307288424.0,
      "step": 9628
    },
    {
      "entropy": 0.6897305082529783,
      "epoch": 0.8859293506075295,
      "grad_norm": 0.15129077434539795,
      "learning_rate": 7.047259790842458e-05,
      "loss": 0.6624,
      "mean_token_accuracy": 0.7998327985405922,
      "num_tokens": 307320449.0,
      "step": 9629
    },
    {
      "entropy": 0.6863480685278773,
      "epoch": 0.8860213569789708,
      "grad_norm": 0.16064195334911346,
      "learning_rate": 7.046953108228295e-05,
      "loss": 0.6621,
      "mean_token_accuracy": 0.8018710948526859,
      "num_tokens": 307351949.0,
      "step": 9630
    },
    {
      "entropy": 0.7443727757781744,
      "epoch": 0.886113363350412,
      "grad_norm": 0.15474838018417358,
      "learning_rate": 7.046646425614132e-05,
      "loss": 0.7346,
      "mean_token_accuracy": 0.7842414639890194,
      "num_tokens": 307383964.0,
      "step": 9631
    },
    {
      "entropy": 0.7420056629925966,
      "epoch": 0.8862053697218533,
      "grad_norm": 0.1548931896686554,
      "learning_rate": 7.04633974299997e-05,
      "loss": 0.7319,
      "mean_token_accuracy": 0.783065851777792,
      "num_tokens": 307416124.0,
      "step": 9632
    },
    {
      "entropy": 0.7689073923975229,
      "epoch": 0.8862973760932945,
      "grad_norm": 0.17796577513217926,
      "learning_rate": 7.046033060385808e-05,
      "loss": 0.7717,
      "mean_token_accuracy": 0.7735776305198669,
      "num_tokens": 307448160.0,
      "step": 9633
    },
    {
      "entropy": 0.7537177801132202,
      "epoch": 0.8863893824647356,
      "grad_norm": 0.16050490736961365,
      "learning_rate": 7.045726377771645e-05,
      "loss": 0.738,
      "mean_token_accuracy": 0.7874788492918015,
      "num_tokens": 307480576.0,
      "step": 9634
    },
    {
      "entropy": 0.6680786870419979,
      "epoch": 0.8864813888361769,
      "grad_norm": 0.15357573330402374,
      "learning_rate": 7.045419695157481e-05,
      "loss": 0.6459,
      "mean_token_accuracy": 0.8082058243453503,
      "num_tokens": 307512767.0,
      "step": 9635
    },
    {
      "entropy": 0.8138674423098564,
      "epoch": 0.8865733952076181,
      "grad_norm": 0.157337948679924,
      "learning_rate": 7.04511301254332e-05,
      "loss": 0.8035,
      "mean_token_accuracy": 0.7625872753560543,
      "num_tokens": 307545086.0,
      "step": 9636
    },
    {
      "entropy": 0.8721104953438044,
      "epoch": 0.8866654015790594,
      "grad_norm": 0.15350845456123352,
      "learning_rate": 7.044806329929156e-05,
      "loss": 0.8649,
      "mean_token_accuracy": 0.7445906549692154,
      "num_tokens": 307576822.0,
      "step": 9637
    },
    {
      "entropy": 0.6852910667657852,
      "epoch": 0.8867574079505006,
      "grad_norm": 0.15122373402118683,
      "learning_rate": 7.044499647314994e-05,
      "loss": 0.6694,
      "mean_token_accuracy": 0.8008206076920033,
      "num_tokens": 307608601.0,
      "step": 9638
    },
    {
      "entropy": 0.7585507407784462,
      "epoch": 0.8868494143219418,
      "grad_norm": 0.15271273255348206,
      "learning_rate": 7.044192964700831e-05,
      "loss": 0.7423,
      "mean_token_accuracy": 0.7788259871304035,
      "num_tokens": 307640378.0,
      "step": 9639
    },
    {
      "entropy": 0.6123691461980343,
      "epoch": 0.886941420693383,
      "grad_norm": 0.1540195792913437,
      "learning_rate": 7.04388628208667e-05,
      "loss": 0.6186,
      "mean_token_accuracy": 0.8174866251647472,
      "num_tokens": 307673100.0,
      "step": 9640
    },
    {
      "entropy": 0.7415809240192175,
      "epoch": 0.8870334270648242,
      "grad_norm": 0.1600474715232849,
      "learning_rate": 7.043579599472506e-05,
      "loss": 0.7375,
      "mean_token_accuracy": 0.781391803175211,
      "num_tokens": 307705302.0,
      "step": 9641
    },
    {
      "entropy": 0.7192567307502031,
      "epoch": 0.8871254334362655,
      "grad_norm": 0.15500742197036743,
      "learning_rate": 7.043272916858343e-05,
      "loss": 0.6981,
      "mean_token_accuracy": 0.7894995287060738,
      "num_tokens": 307736543.0,
      "step": 9642
    },
    {
      "entropy": 0.8774393331259489,
      "epoch": 0.8872174398077067,
      "grad_norm": 0.15801529586315155,
      "learning_rate": 7.042966234244181e-05,
      "loss": 0.879,
      "mean_token_accuracy": 0.7401497736573219,
      "num_tokens": 307768357.0,
      "step": 9643
    },
    {
      "entropy": 0.750929213128984,
      "epoch": 0.8873094461791479,
      "grad_norm": 0.15612201392650604,
      "learning_rate": 7.042659551630019e-05,
      "loss": 0.7519,
      "mean_token_accuracy": 0.7809238322079182,
      "num_tokens": 307800330.0,
      "step": 9644
    },
    {
      "entropy": 0.710002563893795,
      "epoch": 0.8874014525505891,
      "grad_norm": 0.1508781611919403,
      "learning_rate": 7.042352869015856e-05,
      "loss": 0.697,
      "mean_token_accuracy": 0.7970291972160339,
      "num_tokens": 307832344.0,
      "step": 9645
    },
    {
      "entropy": 0.6953284312039614,
      "epoch": 0.8874934589220304,
      "grad_norm": 0.15276481211185455,
      "learning_rate": 7.042046186401693e-05,
      "loss": 0.6815,
      "mean_token_accuracy": 0.7960510365664959,
      "num_tokens": 307864510.0,
      "step": 9646
    },
    {
      "entropy": 0.779976237565279,
      "epoch": 0.8875854652934716,
      "grad_norm": 0.16111548244953156,
      "learning_rate": 7.041739503787531e-05,
      "loss": 0.7591,
      "mean_token_accuracy": 0.7786975204944611,
      "num_tokens": 307896436.0,
      "step": 9647
    },
    {
      "entropy": 0.6979182213544846,
      "epoch": 0.8876774716649128,
      "grad_norm": 0.1454891413450241,
      "learning_rate": 7.041432821173368e-05,
      "loss": 0.6787,
      "mean_token_accuracy": 0.7943682707846165,
      "num_tokens": 307928453.0,
      "step": 9648
    },
    {
      "entropy": 0.8555288389325142,
      "epoch": 0.887769478036354,
      "grad_norm": 0.1619478464126587,
      "learning_rate": 7.041126138559206e-05,
      "loss": 0.824,
      "mean_token_accuracy": 0.7568857483565807,
      "num_tokens": 307959347.0,
      "step": 9649
    },
    {
      "entropy": 0.7843047678470612,
      "epoch": 0.8878614844077952,
      "grad_norm": 0.15852880477905273,
      "learning_rate": 7.040819455945043e-05,
      "loss": 0.7641,
      "mean_token_accuracy": 0.7725748494267464,
      "num_tokens": 307991254.0,
      "step": 9650
    },
    {
      "entropy": 0.6946888603270054,
      "epoch": 0.8879534907792365,
      "grad_norm": 0.16466876864433289,
      "learning_rate": 7.040512773330881e-05,
      "loss": 0.6824,
      "mean_token_accuracy": 0.7976789213716984,
      "num_tokens": 308023699.0,
      "step": 9651
    },
    {
      "entropy": 0.6915777707472444,
      "epoch": 0.8880454971506777,
      "grad_norm": 0.15777607262134552,
      "learning_rate": 7.040206090716718e-05,
      "loss": 0.6804,
      "mean_token_accuracy": 0.7980469018220901,
      "num_tokens": 308054643.0,
      "step": 9652
    },
    {
      "entropy": 0.8185752313584089,
      "epoch": 0.888137503522119,
      "grad_norm": 0.16076898574829102,
      "learning_rate": 7.039899408102554e-05,
      "loss": 0.8131,
      "mean_token_accuracy": 0.7611190862953663,
      "num_tokens": 308086345.0,
      "step": 9653
    },
    {
      "entropy": 0.6776997521519661,
      "epoch": 0.8882295098935601,
      "grad_norm": 0.15106405317783356,
      "learning_rate": 7.039592725488393e-05,
      "loss": 0.669,
      "mean_token_accuracy": 0.8023823499679565,
      "num_tokens": 308117795.0,
      "step": 9654
    },
    {
      "entropy": 0.7409691456705332,
      "epoch": 0.8883215162650013,
      "grad_norm": 0.15260465443134308,
      "learning_rate": 7.03928604287423e-05,
      "loss": 0.7311,
      "mean_token_accuracy": 0.7796307131648064,
      "num_tokens": 308149371.0,
      "step": 9655
    },
    {
      "entropy": 0.7403335925191641,
      "epoch": 0.8884135226364426,
      "grad_norm": 0.15359261631965637,
      "learning_rate": 7.038979360260067e-05,
      "loss": 0.7314,
      "mean_token_accuracy": 0.7831679284572601,
      "num_tokens": 308181211.0,
      "step": 9656
    },
    {
      "entropy": 0.643894228618592,
      "epoch": 0.8885055290078838,
      "grad_norm": 0.1439363807439804,
      "learning_rate": 7.038672677645904e-05,
      "loss": 0.6393,
      "mean_token_accuracy": 0.8130439072847366,
      "num_tokens": 308213541.0,
      "step": 9657
    },
    {
      "entropy": 0.6610963018611073,
      "epoch": 0.8885975353793251,
      "grad_norm": 0.1623806208372116,
      "learning_rate": 7.038365995031741e-05,
      "loss": 0.6561,
      "mean_token_accuracy": 0.8034290261566639,
      "num_tokens": 308245867.0,
      "step": 9658
    },
    {
      "entropy": 0.6358518749475479,
      "epoch": 0.8886895417507662,
      "grad_norm": 0.15814737975597382,
      "learning_rate": 7.038059312417579e-05,
      "loss": 0.6289,
      "mean_token_accuracy": 0.8143282979726791,
      "num_tokens": 308277483.0,
      "step": 9659
    },
    {
      "entropy": 0.7284986805170774,
      "epoch": 0.8887815481222074,
      "grad_norm": 0.1522912085056305,
      "learning_rate": 7.037752629803417e-05,
      "loss": 0.7362,
      "mean_token_accuracy": 0.7802562788128853,
      "num_tokens": 308309678.0,
      "step": 9660
    },
    {
      "entropy": 0.7684119921177626,
      "epoch": 0.8888735544936487,
      "grad_norm": 0.16549603641033173,
      "learning_rate": 7.037445947189254e-05,
      "loss": 0.7648,
      "mean_token_accuracy": 0.7730023451149464,
      "num_tokens": 308340022.0,
      "step": 9661
    },
    {
      "entropy": 0.8047299459576607,
      "epoch": 0.8889655608650899,
      "grad_norm": 0.15636329352855682,
      "learning_rate": 7.037139264575091e-05,
      "loss": 0.8074,
      "mean_token_accuracy": 0.7663075886666775,
      "num_tokens": 308371224.0,
      "step": 9662
    },
    {
      "entropy": 0.7457189299166203,
      "epoch": 0.8890575672365312,
      "grad_norm": 0.16062182188034058,
      "learning_rate": 7.036832581960929e-05,
      "loss": 0.7475,
      "mean_token_accuracy": 0.7765879109501839,
      "num_tokens": 308402151.0,
      "step": 9663
    },
    {
      "entropy": 0.733982166275382,
      "epoch": 0.8891495736079723,
      "grad_norm": 0.1596907377243042,
      "learning_rate": 7.036525899346766e-05,
      "loss": 0.7347,
      "mean_token_accuracy": 0.783097505569458,
      "num_tokens": 308433897.0,
      "step": 9664
    },
    {
      "entropy": 0.6538415336981416,
      "epoch": 0.8892415799794136,
      "grad_norm": 0.150006964802742,
      "learning_rate": 7.036219216732604e-05,
      "loss": 0.6344,
      "mean_token_accuracy": 0.8115242794156075,
      "num_tokens": 308465712.0,
      "step": 9665
    },
    {
      "entropy": 0.7513416139408946,
      "epoch": 0.8893335863508548,
      "grad_norm": 0.1560031920671463,
      "learning_rate": 7.035912534118441e-05,
      "loss": 0.75,
      "mean_token_accuracy": 0.7817319631576538,
      "num_tokens": 308498038.0,
      "step": 9666
    },
    {
      "entropy": 0.6389488037675619,
      "epoch": 0.889425592722296,
      "grad_norm": 0.1570833921432495,
      "learning_rate": 7.035605851504279e-05,
      "loss": 0.6258,
      "mean_token_accuracy": 0.811767291277647,
      "num_tokens": 308530118.0,
      "step": 9667
    },
    {
      "entropy": 0.7172889616340399,
      "epoch": 0.8895175990937373,
      "grad_norm": 0.16678012907505035,
      "learning_rate": 7.035299168890116e-05,
      "loss": 0.6919,
      "mean_token_accuracy": 0.7969530299305916,
      "num_tokens": 308561316.0,
      "step": 9668
    },
    {
      "entropy": 0.8401995971798897,
      "epoch": 0.8896096054651784,
      "grad_norm": 0.1594872921705246,
      "learning_rate": 7.034992486275952e-05,
      "loss": 0.8227,
      "mean_token_accuracy": 0.7671616598963737,
      "num_tokens": 308593980.0,
      "step": 9669
    },
    {
      "entropy": 0.6393571980297565,
      "epoch": 0.8897016118366197,
      "grad_norm": 0.15289156138896942,
      "learning_rate": 7.03468580366179e-05,
      "loss": 0.6156,
      "mean_token_accuracy": 0.8200877197086811,
      "num_tokens": 308625602.0,
      "step": 9670
    },
    {
      "entropy": 0.754255760461092,
      "epoch": 0.8897936182080609,
      "grad_norm": 0.15577618777751923,
      "learning_rate": 7.034379121047629e-05,
      "loss": 0.7331,
      "mean_token_accuracy": 0.7813347168266773,
      "num_tokens": 308657255.0,
      "step": 9671
    },
    {
      "entropy": 0.7062916681170464,
      "epoch": 0.8898856245795022,
      "grad_norm": 0.1502193659543991,
      "learning_rate": 7.034072438433466e-05,
      "loss": 0.689,
      "mean_token_accuracy": 0.7964572459459305,
      "num_tokens": 308689796.0,
      "step": 9672
    },
    {
      "entropy": 0.7181137204170227,
      "epoch": 0.8899776309509434,
      "grad_norm": 0.15331952273845673,
      "learning_rate": 7.033765755819302e-05,
      "loss": 0.7018,
      "mean_token_accuracy": 0.7910886146128178,
      "num_tokens": 308722275.0,
      "step": 9673
    },
    {
      "entropy": 0.7655514925718307,
      "epoch": 0.8900696373223845,
      "grad_norm": 0.15565571188926697,
      "learning_rate": 7.03345907320514e-05,
      "loss": 0.7558,
      "mean_token_accuracy": 0.7787880413234234,
      "num_tokens": 308754930.0,
      "step": 9674
    },
    {
      "entropy": 0.6045652609318495,
      "epoch": 0.8901616436938258,
      "grad_norm": 0.15035521984100342,
      "learning_rate": 7.033152390590979e-05,
      "loss": 0.5887,
      "mean_token_accuracy": 0.8217860087752342,
      "num_tokens": 308787271.0,
      "step": 9675
    },
    {
      "entropy": 0.7893930040299892,
      "epoch": 0.890253650065267,
      "grad_norm": 0.1615239828824997,
      "learning_rate": 7.032845707976815e-05,
      "loss": 0.8004,
      "mean_token_accuracy": 0.767118263989687,
      "num_tokens": 308818650.0,
      "step": 9676
    },
    {
      "entropy": 0.7899424079805613,
      "epoch": 0.8903456564367083,
      "grad_norm": 0.15457771718502045,
      "learning_rate": 7.032539025362652e-05,
      "loss": 0.7891,
      "mean_token_accuracy": 0.7674966230988503,
      "num_tokens": 308851082.0,
      "step": 9677
    },
    {
      "entropy": 0.7123268488794565,
      "epoch": 0.8904376628081495,
      "grad_norm": 0.1542898416519165,
      "learning_rate": 7.03223234274849e-05,
      "loss": 0.7152,
      "mean_token_accuracy": 0.7856649607419968,
      "num_tokens": 308882969.0,
      "step": 9678
    },
    {
      "entropy": 0.671601340174675,
      "epoch": 0.8905296691795906,
      "grad_norm": 0.1521022617816925,
      "learning_rate": 7.031925660134327e-05,
      "loss": 0.6467,
      "mean_token_accuracy": 0.8102095723152161,
      "num_tokens": 308914592.0,
      "step": 9679
    },
    {
      "entropy": 0.7700145319104195,
      "epoch": 0.8906216755510319,
      "grad_norm": 0.16368262469768524,
      "learning_rate": 7.031618977520165e-05,
      "loss": 0.7614,
      "mean_token_accuracy": 0.7761921808123589,
      "num_tokens": 308946621.0,
      "step": 9680
    },
    {
      "entropy": 0.7867386545985937,
      "epoch": 0.8907136819224731,
      "grad_norm": 0.15613657236099243,
      "learning_rate": 7.031312294906002e-05,
      "loss": 0.7856,
      "mean_token_accuracy": 0.7718074694275856,
      "num_tokens": 308978299.0,
      "step": 9681
    },
    {
      "entropy": 0.6076631024479866,
      "epoch": 0.8908056882939144,
      "grad_norm": 0.139872744679451,
      "learning_rate": 7.03100561229184e-05,
      "loss": 0.5967,
      "mean_token_accuracy": 0.8183725886046886,
      "num_tokens": 309010039.0,
      "step": 9682
    },
    {
      "entropy": 0.6448656460270286,
      "epoch": 0.8908976946653556,
      "grad_norm": 0.1484212875366211,
      "learning_rate": 7.030698929677677e-05,
      "loss": 0.6396,
      "mean_token_accuracy": 0.8079468868672848,
      "num_tokens": 309041210.0,
      "step": 9683
    },
    {
      "entropy": 0.6965286415070295,
      "epoch": 0.8909897010367968,
      "grad_norm": 0.1567315310239792,
      "learning_rate": 7.030392247063514e-05,
      "loss": 0.6912,
      "mean_token_accuracy": 0.7947283424437046,
      "num_tokens": 309072755.0,
      "step": 9684
    },
    {
      "entropy": 0.6721262969076633,
      "epoch": 0.891081707408238,
      "grad_norm": 0.15393488109111786,
      "learning_rate": 7.030085564449352e-05,
      "loss": 0.6546,
      "mean_token_accuracy": 0.8023935668170452,
      "num_tokens": 309104482.0,
      "step": 9685
    },
    {
      "entropy": 0.6860969737172127,
      "epoch": 0.8911737137796792,
      "grad_norm": 0.14783485233783722,
      "learning_rate": 7.02977888183519e-05,
      "loss": 0.6768,
      "mean_token_accuracy": 0.7961494214832783,
      "num_tokens": 309136460.0,
      "step": 9686
    },
    {
      "entropy": 0.7602259572595358,
      "epoch": 0.8912657201511205,
      "grad_norm": 0.15878857672214508,
      "learning_rate": 7.029472199221027e-05,
      "loss": 0.7471,
      "mean_token_accuracy": 0.7782161384820938,
      "num_tokens": 309167261.0,
      "step": 9687
    },
    {
      "entropy": 0.82300647161901,
      "epoch": 0.8913577265225617,
      "grad_norm": 0.15816228091716766,
      "learning_rate": 7.029165516606864e-05,
      "loss": 0.8085,
      "mean_token_accuracy": 0.7625592947006226,
      "num_tokens": 309197879.0,
      "step": 9688
    },
    {
      "entropy": 0.7312655746936798,
      "epoch": 0.8914497328940029,
      "grad_norm": 0.14812161028385162,
      "learning_rate": 7.0288588339927e-05,
      "loss": 0.7205,
      "mean_token_accuracy": 0.7878422029316425,
      "num_tokens": 309230505.0,
      "step": 9689
    },
    {
      "entropy": 0.6669285157695413,
      "epoch": 0.8915417392654441,
      "grad_norm": 0.15501949191093445,
      "learning_rate": 7.028552151378539e-05,
      "loss": 0.653,
      "mean_token_accuracy": 0.8059894107282162,
      "num_tokens": 309262588.0,
      "step": 9690
    },
    {
      "entropy": 0.774366999976337,
      "epoch": 0.8916337456368854,
      "grad_norm": 0.17113016545772552,
      "learning_rate": 7.028245468764377e-05,
      "loss": 0.7782,
      "mean_token_accuracy": 0.776549905538559,
      "num_tokens": 309294526.0,
      "step": 9691
    },
    {
      "entropy": 0.7849632147699594,
      "epoch": 0.8917257520083266,
      "grad_norm": 0.15285179018974304,
      "learning_rate": 7.027938786150213e-05,
      "loss": 0.7714,
      "mean_token_accuracy": 0.7762327343225479,
      "num_tokens": 309326443.0,
      "step": 9692
    },
    {
      "entropy": 0.6538647655397654,
      "epoch": 0.8918177583797678,
      "grad_norm": 0.1666766107082367,
      "learning_rate": 7.02763210353605e-05,
      "loss": 0.6397,
      "mean_token_accuracy": 0.8115736916661263,
      "num_tokens": 309358514.0,
      "step": 9693
    },
    {
      "entropy": 0.6936702486127615,
      "epoch": 0.891909764751209,
      "grad_norm": 0.16218049824237823,
      "learning_rate": 7.027325420921888e-05,
      "loss": 0.6896,
      "mean_token_accuracy": 0.7987096086144447,
      "num_tokens": 309390289.0,
      "step": 9694
    },
    {
      "entropy": 0.7407143972814083,
      "epoch": 0.8920017711226502,
      "grad_norm": 0.15160223841667175,
      "learning_rate": 7.027018738307725e-05,
      "loss": 0.716,
      "mean_token_accuracy": 0.7893079370260239,
      "num_tokens": 309422558.0,
      "step": 9695
    },
    {
      "entropy": 0.7109206933528185,
      "epoch": 0.8920937774940915,
      "grad_norm": 0.14840839803218842,
      "learning_rate": 7.026712055693563e-05,
      "loss": 0.7091,
      "mean_token_accuracy": 0.793302558362484,
      "num_tokens": 309454858.0,
      "step": 9696
    },
    {
      "entropy": 0.6923775132745504,
      "epoch": 0.8921857838655327,
      "grad_norm": 0.15844149887561798,
      "learning_rate": 7.0264053730794e-05,
      "loss": 0.6678,
      "mean_token_accuracy": 0.799572303891182,
      "num_tokens": 309487006.0,
      "step": 9697
    },
    {
      "entropy": 0.6767344493418932,
      "epoch": 0.892277790236974,
      "grad_norm": 0.1544109731912613,
      "learning_rate": 7.026098690465238e-05,
      "loss": 0.6708,
      "mean_token_accuracy": 0.797864057123661,
      "num_tokens": 309519351.0,
      "step": 9698
    },
    {
      "entropy": 0.6757818590849638,
      "epoch": 0.8923697966084151,
      "grad_norm": 0.155557319521904,
      "learning_rate": 7.025792007851075e-05,
      "loss": 0.6642,
      "mean_token_accuracy": 0.8014401309192181,
      "num_tokens": 309551049.0,
      "step": 9699
    },
    {
      "entropy": 0.732369152829051,
      "epoch": 0.8924618029798563,
      "grad_norm": 0.15993107855319977,
      "learning_rate": 7.025485325236912e-05,
      "loss": 0.7014,
      "mean_token_accuracy": 0.7903265058994293,
      "num_tokens": 309583335.0,
      "step": 9700
    },
    {
      "entropy": 0.7009381111711264,
      "epoch": 0.8925538093512976,
      "grad_norm": 0.15613125264644623,
      "learning_rate": 7.02517864262275e-05,
      "loss": 0.689,
      "mean_token_accuracy": 0.7954860590398312,
      "num_tokens": 309615426.0,
      "step": 9701
    },
    {
      "entropy": 0.7938085328787565,
      "epoch": 0.8926458157227388,
      "grad_norm": 0.16545656323432922,
      "learning_rate": 7.024871960008588e-05,
      "loss": 0.802,
      "mean_token_accuracy": 0.7650130093097687,
      "num_tokens": 309647163.0,
      "step": 9702
    },
    {
      "entropy": 0.742486952804029,
      "epoch": 0.8927378220941801,
      "grad_norm": 0.1516485959291458,
      "learning_rate": 7.024565277394425e-05,
      "loss": 0.7277,
      "mean_token_accuracy": 0.7857805043458939,
      "num_tokens": 309679229.0,
      "step": 9703
    },
    {
      "entropy": 0.7627192828804255,
      "epoch": 0.8928298284656212,
      "grad_norm": 0.173240527510643,
      "learning_rate": 7.024258594780262e-05,
      "loss": 0.7486,
      "mean_token_accuracy": 0.7774785198271275,
      "num_tokens": 309711683.0,
      "step": 9704
    },
    {
      "entropy": 0.8229615800082684,
      "epoch": 0.8929218348370624,
      "grad_norm": 0.15834133327007294,
      "learning_rate": 7.0239519121661e-05,
      "loss": 0.8054,
      "mean_token_accuracy": 0.7608070410788059,
      "num_tokens": 309742692.0,
      "step": 9705
    },
    {
      "entropy": 0.863924540579319,
      "epoch": 0.8930138412085037,
      "grad_norm": 0.16179075837135315,
      "learning_rate": 7.023645229551937e-05,
      "loss": 0.8633,
      "mean_token_accuracy": 0.750846866518259,
      "num_tokens": 309775143.0,
      "step": 9706
    },
    {
      "entropy": 0.6475413935258985,
      "epoch": 0.8931058475799449,
      "grad_norm": 0.16078592836856842,
      "learning_rate": 7.023338546937775e-05,
      "loss": 0.6409,
      "mean_token_accuracy": 0.8031401410698891,
      "num_tokens": 309806873.0,
      "step": 9707
    },
    {
      "entropy": 0.786225214600563,
      "epoch": 0.8931978539513862,
      "grad_norm": 0.15629799664020538,
      "learning_rate": 7.023031864323612e-05,
      "loss": 0.7649,
      "mean_token_accuracy": 0.7725161090493202,
      "num_tokens": 309838556.0,
      "step": 9708
    },
    {
      "entropy": 0.7441845908761024,
      "epoch": 0.8932898603228273,
      "grad_norm": 0.1564975380897522,
      "learning_rate": 7.02272518170945e-05,
      "loss": 0.7398,
      "mean_token_accuracy": 0.7840327546000481,
      "num_tokens": 309870233.0,
      "step": 9709
    },
    {
      "entropy": 0.773976749740541,
      "epoch": 0.8933818666942686,
      "grad_norm": 0.1620330512523651,
      "learning_rate": 7.022418499095286e-05,
      "loss": 0.7657,
      "mean_token_accuracy": 0.7740952782332897,
      "num_tokens": 309901735.0,
      "step": 9710
    },
    {
      "entropy": 0.7285696119070053,
      "epoch": 0.8934738730657098,
      "grad_norm": 0.16926395893096924,
      "learning_rate": 7.022111816481123e-05,
      "loss": 0.7164,
      "mean_token_accuracy": 0.7893508598208427,
      "num_tokens": 309933674.0,
      "step": 9711
    },
    {
      "entropy": 0.6906839329749346,
      "epoch": 0.893565879437151,
      "grad_norm": 0.15170729160308838,
      "learning_rate": 7.021805133866961e-05,
      "loss": 0.6695,
      "mean_token_accuracy": 0.7969261668622494,
      "num_tokens": 309965300.0,
      "step": 9712
    },
    {
      "entropy": 0.6893171723932028,
      "epoch": 0.8936578858085923,
      "grad_norm": 0.15844608843326569,
      "learning_rate": 7.0214984512528e-05,
      "loss": 0.6721,
      "mean_token_accuracy": 0.8009597659111023,
      "num_tokens": 309997146.0,
      "step": 9713
    },
    {
      "entropy": 0.7376275844871998,
      "epoch": 0.8937498921800334,
      "grad_norm": 0.157451331615448,
      "learning_rate": 7.021191768638636e-05,
      "loss": 0.7189,
      "mean_token_accuracy": 0.7831058837473392,
      "num_tokens": 310029582.0,
      "step": 9714
    },
    {
      "entropy": 0.7229223754256964,
      "epoch": 0.8938418985514747,
      "grad_norm": 0.16410456597805023,
      "learning_rate": 7.020885086024473e-05,
      "loss": 0.7358,
      "mean_token_accuracy": 0.7827345654368401,
      "num_tokens": 310061450.0,
      "step": 9715
    },
    {
      "entropy": 0.7455324344336987,
      "epoch": 0.8939339049229159,
      "grad_norm": 0.16147375106811523,
      "learning_rate": 7.020578403410311e-05,
      "loss": 0.7386,
      "mean_token_accuracy": 0.7806043401360512,
      "num_tokens": 310093784.0,
      "step": 9716
    },
    {
      "entropy": 0.734908202663064,
      "epoch": 0.8940259112943572,
      "grad_norm": 0.16121307015419006,
      "learning_rate": 7.02027172079615e-05,
      "loss": 0.7307,
      "mean_token_accuracy": 0.7865401096642017,
      "num_tokens": 310125479.0,
      "step": 9717
    },
    {
      "entropy": 0.7254662308841944,
      "epoch": 0.8941179176657984,
      "grad_norm": 0.16434839367866516,
      "learning_rate": 7.019965038181986e-05,
      "loss": 0.721,
      "mean_token_accuracy": 0.792930644005537,
      "num_tokens": 310157732.0,
      "step": 9718
    },
    {
      "entropy": 0.6506439801305532,
      "epoch": 0.8942099240372395,
      "grad_norm": 0.14761841297149658,
      "learning_rate": 7.019658355567823e-05,
      "loss": 0.6479,
      "mean_token_accuracy": 0.803863599896431,
      "num_tokens": 310190187.0,
      "step": 9719
    },
    {
      "entropy": 0.6933459471911192,
      "epoch": 0.8943019304086808,
      "grad_norm": 0.1621619015932083,
      "learning_rate": 7.01935167295366e-05,
      "loss": 0.6873,
      "mean_token_accuracy": 0.7908600829541683,
      "num_tokens": 310219906.0,
      "step": 9720
    },
    {
      "entropy": 0.6701205931603909,
      "epoch": 0.894393936780122,
      "grad_norm": 0.1533735692501068,
      "learning_rate": 7.019044990339498e-05,
      "loss": 0.6653,
      "mean_token_accuracy": 0.8015193976461887,
      "num_tokens": 310252212.0,
      "step": 9721
    },
    {
      "entropy": 0.7149121016263962,
      "epoch": 0.8944859431515633,
      "grad_norm": 0.14680945873260498,
      "learning_rate": 7.018738307725336e-05,
      "loss": 0.7011,
      "mean_token_accuracy": 0.793262105435133,
      "num_tokens": 310284149.0,
      "step": 9722
    },
    {
      "entropy": 0.7621957194060087,
      "epoch": 0.8945779495230045,
      "grad_norm": 0.15533535182476044,
      "learning_rate": 7.018431625111173e-05,
      "loss": 0.746,
      "mean_token_accuracy": 0.7809430584311485,
      "num_tokens": 310316286.0,
      "step": 9723
    },
    {
      "entropy": 0.7084170933812857,
      "epoch": 0.8946699558944456,
      "grad_norm": 0.1523558348417282,
      "learning_rate": 7.01812494249701e-05,
      "loss": 0.7005,
      "mean_token_accuracy": 0.7889360673725605,
      "num_tokens": 310348228.0,
      "step": 9724
    },
    {
      "entropy": 0.6464080289006233,
      "epoch": 0.8947619622658869,
      "grad_norm": 0.15981796383857727,
      "learning_rate": 7.017818259882848e-05,
      "loss": 0.6321,
      "mean_token_accuracy": 0.8140261173248291,
      "num_tokens": 310380829.0,
      "step": 9725
    },
    {
      "entropy": 0.7443840745836496,
      "epoch": 0.8948539686373281,
      "grad_norm": 0.1542026698589325,
      "learning_rate": 7.017511577268685e-05,
      "loss": 0.7342,
      "mean_token_accuracy": 0.7863458432257175,
      "num_tokens": 310412856.0,
      "step": 9726
    },
    {
      "entropy": 0.5919651426374912,
      "epoch": 0.8949459750087694,
      "grad_norm": 0.15372680127620697,
      "learning_rate": 7.017204894654523e-05,
      "loss": 0.5706,
      "mean_token_accuracy": 0.8285414129495621,
      "num_tokens": 310444619.0,
      "step": 9727
    },
    {
      "entropy": 0.7046142499893904,
      "epoch": 0.8950379813802106,
      "grad_norm": 0.14843183755874634,
      "learning_rate": 7.01689821204036e-05,
      "loss": 0.6816,
      "mean_token_accuracy": 0.7974228486418724,
      "num_tokens": 310476636.0,
      "step": 9728
    },
    {
      "entropy": 0.7461630403995514,
      "epoch": 0.8951299877516518,
      "grad_norm": 0.1592244952917099,
      "learning_rate": 7.016591529426198e-05,
      "loss": 0.7458,
      "mean_token_accuracy": 0.7812934070825577,
      "num_tokens": 310508940.0,
      "step": 9729
    },
    {
      "entropy": 0.7049547769129276,
      "epoch": 0.895221994123093,
      "grad_norm": 0.15638230741024017,
      "learning_rate": 7.016284846812034e-05,
      "loss": 0.6929,
      "mean_token_accuracy": 0.7955847531557083,
      "num_tokens": 310541132.0,
      "step": 9730
    },
    {
      "entropy": 0.7301756367087364,
      "epoch": 0.8953140004945342,
      "grad_norm": 0.15731576085090637,
      "learning_rate": 7.015978164197871e-05,
      "loss": 0.7368,
      "mean_token_accuracy": 0.7872087098658085,
      "num_tokens": 310572090.0,
      "step": 9731
    },
    {
      "entropy": 0.71028982847929,
      "epoch": 0.8954060068659755,
      "grad_norm": 0.15627345442771912,
      "learning_rate": 7.01567148158371e-05,
      "loss": 0.7017,
      "mean_token_accuracy": 0.7907323129475117,
      "num_tokens": 310604442.0,
      "step": 9732
    },
    {
      "entropy": 0.7490415498614311,
      "epoch": 0.8954980132374167,
      "grad_norm": 0.14790017902851105,
      "learning_rate": 7.015364798969547e-05,
      "loss": 0.7409,
      "mean_token_accuracy": 0.7839168235659599,
      "num_tokens": 310636845.0,
      "step": 9733
    },
    {
      "entropy": 0.7454306595027447,
      "epoch": 0.8955900196088579,
      "grad_norm": 0.15182456374168396,
      "learning_rate": 7.015058116355384e-05,
      "loss": 0.7399,
      "mean_token_accuracy": 0.7841071374714375,
      "num_tokens": 310668385.0,
      "step": 9734
    },
    {
      "entropy": 0.7743997294455767,
      "epoch": 0.8956820259802991,
      "grad_norm": 0.14572659134864807,
      "learning_rate": 7.014751433741221e-05,
      "loss": 0.7662,
      "mean_token_accuracy": 0.7715252749621868,
      "num_tokens": 310700773.0,
      "step": 9735
    },
    {
      "entropy": 0.7429249472916126,
      "epoch": 0.8957740323517404,
      "grad_norm": 0.16099292039871216,
      "learning_rate": 7.014444751127059e-05,
      "loss": 0.7288,
      "mean_token_accuracy": 0.7899978458881378,
      "num_tokens": 310732956.0,
      "step": 9736
    },
    {
      "entropy": 0.6262855222448707,
      "epoch": 0.8958660387231816,
      "grad_norm": 0.14752885699272156,
      "learning_rate": 7.014138068512896e-05,
      "loss": 0.6191,
      "mean_token_accuracy": 0.8102495074272156,
      "num_tokens": 310765203.0,
      "step": 9737
    },
    {
      "entropy": 0.6266257651150227,
      "epoch": 0.8959580450946228,
      "grad_norm": 0.15453769266605377,
      "learning_rate": 7.013831385898734e-05,
      "loss": 0.6204,
      "mean_token_accuracy": 0.8117399401962757,
      "num_tokens": 310796769.0,
      "step": 9738
    },
    {
      "entropy": 0.7293918682262301,
      "epoch": 0.896050051466064,
      "grad_norm": 0.16430437564849854,
      "learning_rate": 7.013524703284571e-05,
      "loss": 0.724,
      "mean_token_accuracy": 0.7888215221464634,
      "num_tokens": 310828562.0,
      "step": 9739
    },
    {
      "entropy": 0.729425510391593,
      "epoch": 0.8961420578375052,
      "grad_norm": 0.16010093688964844,
      "learning_rate": 7.013218020670409e-05,
      "loss": 0.722,
      "mean_token_accuracy": 0.7841552682220936,
      "num_tokens": 310860021.0,
      "step": 9740
    },
    {
      "entropy": 0.7023179419338703,
      "epoch": 0.8962340642089465,
      "grad_norm": 0.16280686855316162,
      "learning_rate": 7.012911338056246e-05,
      "loss": 0.6983,
      "mean_token_accuracy": 0.7929897345602512,
      "num_tokens": 310892789.0,
      "step": 9741
    },
    {
      "entropy": 0.6688303053379059,
      "epoch": 0.8963260705803877,
      "grad_norm": 0.15089653432369232,
      "learning_rate": 7.012604655442083e-05,
      "loss": 0.6606,
      "mean_token_accuracy": 0.8056538589298725,
      "num_tokens": 310924535.0,
      "step": 9742
    },
    {
      "entropy": 0.6134938094764948,
      "epoch": 0.896418076951829,
      "grad_norm": 0.14048050343990326,
      "learning_rate": 7.012297972827921e-05,
      "loss": 0.5918,
      "mean_token_accuracy": 0.817150890827179,
      "num_tokens": 310956434.0,
      "step": 9743
    },
    {
      "entropy": 0.6552703473716974,
      "epoch": 0.8965100833232701,
      "grad_norm": 0.1549362689256668,
      "learning_rate": 7.011991290213759e-05,
      "loss": 0.6462,
      "mean_token_accuracy": 0.8065964467823505,
      "num_tokens": 310987645.0,
      "step": 9744
    },
    {
      "entropy": 0.7720081247389317,
      "epoch": 0.8966020896947113,
      "grad_norm": 0.1624019742012024,
      "learning_rate": 7.011684607599596e-05,
      "loss": 0.7734,
      "mean_token_accuracy": 0.7732260227203369,
      "num_tokens": 311018703.0,
      "step": 9745
    },
    {
      "entropy": 0.7445423752069473,
      "epoch": 0.8966940960661526,
      "grad_norm": 0.15188178420066833,
      "learning_rate": 7.011377924985432e-05,
      "loss": 0.7359,
      "mean_token_accuracy": 0.7817879542708397,
      "num_tokens": 311050517.0,
      "step": 9746
    },
    {
      "entropy": 0.7473039664328098,
      "epoch": 0.8967861024375938,
      "grad_norm": 0.153280109167099,
      "learning_rate": 7.011071242371269e-05,
      "loss": 0.7395,
      "mean_token_accuracy": 0.783237561583519,
      "num_tokens": 311082959.0,
      "step": 9747
    },
    {
      "entropy": 0.7875000387430191,
      "epoch": 0.8968781088090351,
      "grad_norm": 0.16047151386737823,
      "learning_rate": 7.010764559757109e-05,
      "loss": 0.767,
      "mean_token_accuracy": 0.776949055492878,
      "num_tokens": 311113927.0,
      "step": 9748
    },
    {
      "entropy": 0.612736240029335,
      "epoch": 0.8969701151804762,
      "grad_norm": 0.162144273519516,
      "learning_rate": 7.010457877142946e-05,
      "loss": 0.5987,
      "mean_token_accuracy": 0.8213411867618561,
      "num_tokens": 311145636.0,
      "step": 9749
    },
    {
      "entropy": 0.7511744797229767,
      "epoch": 0.8970621215519174,
      "grad_norm": 0.15910568833351135,
      "learning_rate": 7.010151194528782e-05,
      "loss": 0.7437,
      "mean_token_accuracy": 0.7773890271782875,
      "num_tokens": 311178202.0,
      "step": 9750
    },
    {
      "entropy": 0.7898722123354673,
      "epoch": 0.8971541279233587,
      "grad_norm": 0.16849593818187714,
      "learning_rate": 7.009844511914619e-05,
      "loss": 0.7845,
      "mean_token_accuracy": 0.7724143080413342,
      "num_tokens": 311210095.0,
      "step": 9751
    },
    {
      "entropy": 0.8026852607727051,
      "epoch": 0.8972461342947999,
      "grad_norm": 0.1586998999118805,
      "learning_rate": 7.009537829300457e-05,
      "loss": 0.7859,
      "mean_token_accuracy": 0.7673180811107159,
      "num_tokens": 311242403.0,
      "step": 9752
    },
    {
      "entropy": 0.7365549225360155,
      "epoch": 0.8973381406662412,
      "grad_norm": 0.15504398941993713,
      "learning_rate": 7.009231146686295e-05,
      "loss": 0.708,
      "mean_token_accuracy": 0.7883164882659912,
      "num_tokens": 311274587.0,
      "step": 9753
    },
    {
      "entropy": 0.7108295187354088,
      "epoch": 0.8974301470376823,
      "grad_norm": 0.1589454561471939,
      "learning_rate": 7.008924464072132e-05,
      "loss": 0.685,
      "mean_token_accuracy": 0.7934074178338051,
      "num_tokens": 311305470.0,
      "step": 9754
    },
    {
      "entropy": 0.7785720536485314,
      "epoch": 0.8975221534091236,
      "grad_norm": 0.15640532970428467,
      "learning_rate": 7.008617781457969e-05,
      "loss": 0.7811,
      "mean_token_accuracy": 0.773921512067318,
      "num_tokens": 311337920.0,
      "step": 9755
    },
    {
      "entropy": 0.6715652821585536,
      "epoch": 0.8976141597805648,
      "grad_norm": 0.1498286873102188,
      "learning_rate": 7.008311098843807e-05,
      "loss": 0.6603,
      "mean_token_accuracy": 0.8060934692621231,
      "num_tokens": 311369668.0,
      "step": 9756
    },
    {
      "entropy": 0.7154117226600647,
      "epoch": 0.897706166152006,
      "grad_norm": 0.15789103507995605,
      "learning_rate": 7.008004416229644e-05,
      "loss": 0.7015,
      "mean_token_accuracy": 0.7950894124805927,
      "num_tokens": 311402305.0,
      "step": 9757
    },
    {
      "entropy": 0.8201962653547525,
      "epoch": 0.8977981725234473,
      "grad_norm": 0.16052629053592682,
      "learning_rate": 7.007697733615482e-05,
      "loss": 0.8304,
      "mean_token_accuracy": 0.7590490020811558,
      "num_tokens": 311434954.0,
      "step": 9758
    },
    {
      "entropy": 0.815439485013485,
      "epoch": 0.8978901788948884,
      "grad_norm": 0.15913765132427216,
      "learning_rate": 7.007391051001319e-05,
      "loss": 0.8204,
      "mean_token_accuracy": 0.7591370642185211,
      "num_tokens": 311466374.0,
      "step": 9759
    },
    {
      "entropy": 0.7482439130544662,
      "epoch": 0.8979821852663297,
      "grad_norm": 0.1558789610862732,
      "learning_rate": 7.007084368387157e-05,
      "loss": 0.7387,
      "mean_token_accuracy": 0.7842269204556942,
      "num_tokens": 311499142.0,
      "step": 9760
    },
    {
      "entropy": 0.655074667185545,
      "epoch": 0.8980741916377709,
      "grad_norm": 0.14644883573055267,
      "learning_rate": 7.006777685772994e-05,
      "loss": 0.6333,
      "mean_token_accuracy": 0.8123593628406525,
      "num_tokens": 311530691.0,
      "step": 9761
    },
    {
      "entropy": 0.7210860699415207,
      "epoch": 0.8981661980092122,
      "grad_norm": 0.15433555841445923,
      "learning_rate": 7.00647100315883e-05,
      "loss": 0.7118,
      "mean_token_accuracy": 0.788340013474226,
      "num_tokens": 311563015.0,
      "step": 9762
    },
    {
      "entropy": 0.7040946641936898,
      "epoch": 0.8982582043806534,
      "grad_norm": 0.15444137156009674,
      "learning_rate": 7.006164320544669e-05,
      "loss": 0.6759,
      "mean_token_accuracy": 0.7983924373984337,
      "num_tokens": 311594974.0,
      "step": 9763
    },
    {
      "entropy": 0.7302393466234207,
      "epoch": 0.8983502107520945,
      "grad_norm": 0.15846437215805054,
      "learning_rate": 7.005857637930507e-05,
      "loss": 0.7137,
      "mean_token_accuracy": 0.7890910655260086,
      "num_tokens": 311627458.0,
      "step": 9764
    },
    {
      "entropy": 0.6590369679033756,
      "epoch": 0.8984422171235358,
      "grad_norm": 0.14430320262908936,
      "learning_rate": 7.005550955316344e-05,
      "loss": 0.6379,
      "mean_token_accuracy": 0.8080693148076534,
      "num_tokens": 311659688.0,
      "step": 9765
    },
    {
      "entropy": 0.7413441371172667,
      "epoch": 0.898534223494977,
      "grad_norm": 0.14801038801670074,
      "learning_rate": 7.00524427270218e-05,
      "loss": 0.7236,
      "mean_token_accuracy": 0.7884866632521152,
      "num_tokens": 311692010.0,
      "step": 9766
    },
    {
      "entropy": 0.629822576418519,
      "epoch": 0.8986262298664183,
      "grad_norm": 0.1455378532409668,
      "learning_rate": 7.004937590088019e-05,
      "loss": 0.619,
      "mean_token_accuracy": 0.8130967542529106,
      "num_tokens": 311724003.0,
      "step": 9767
    },
    {
      "entropy": 0.6452059149742126,
      "epoch": 0.8987182362378595,
      "grad_norm": 0.15701407194137573,
      "learning_rate": 7.004630907473855e-05,
      "loss": 0.6288,
      "mean_token_accuracy": 0.8120154365897179,
      "num_tokens": 311756676.0,
      "step": 9768
    },
    {
      "entropy": 0.7190279345959425,
      "epoch": 0.8988102426093006,
      "grad_norm": 0.15230204164981842,
      "learning_rate": 7.004324224859694e-05,
      "loss": 0.7144,
      "mean_token_accuracy": 0.7880436964333057,
      "num_tokens": 311788056.0,
      "step": 9769
    },
    {
      "entropy": 0.7450856510549784,
      "epoch": 0.8989022489807419,
      "grad_norm": 0.15815109014511108,
      "learning_rate": 7.00401754224553e-05,
      "loss": 0.7435,
      "mean_token_accuracy": 0.783076073974371,
      "num_tokens": 311820108.0,
      "step": 9770
    },
    {
      "entropy": 0.7825587298721075,
      "epoch": 0.8989942553521831,
      "grad_norm": 0.16036280989646912,
      "learning_rate": 7.003710859631368e-05,
      "loss": 0.7701,
      "mean_token_accuracy": 0.7700859978795052,
      "num_tokens": 311851805.0,
      "step": 9771
    },
    {
      "entropy": 0.7319088838994503,
      "epoch": 0.8990862617236244,
      "grad_norm": 0.16398406028747559,
      "learning_rate": 7.003404177017205e-05,
      "loss": 0.7311,
      "mean_token_accuracy": 0.7801793999969959,
      "num_tokens": 311884199.0,
      "step": 9772
    },
    {
      "entropy": 0.7769899535924196,
      "epoch": 0.8991782680950656,
      "grad_norm": 0.16605359315872192,
      "learning_rate": 7.003097494403042e-05,
      "loss": 0.7743,
      "mean_token_accuracy": 0.7745270244777203,
      "num_tokens": 311914683.0,
      "step": 9773
    },
    {
      "entropy": 0.6506979689002037,
      "epoch": 0.8992702744665068,
      "grad_norm": 0.14960184693336487,
      "learning_rate": 7.00279081178888e-05,
      "loss": 0.6397,
      "mean_token_accuracy": 0.8094101101160049,
      "num_tokens": 311947027.0,
      "step": 9774
    },
    {
      "entropy": 0.6745307296514511,
      "epoch": 0.899362280837948,
      "grad_norm": 0.15842914581298828,
      "learning_rate": 7.002484129174718e-05,
      "loss": 0.6541,
      "mean_token_accuracy": 0.805201031267643,
      "num_tokens": 311978365.0,
      "step": 9775
    },
    {
      "entropy": 0.6936991307884455,
      "epoch": 0.8994542872093892,
      "grad_norm": 0.16436250507831573,
      "learning_rate": 7.002177446560555e-05,
      "loss": 0.696,
      "mean_token_accuracy": 0.7945244126021862,
      "num_tokens": 312010449.0,
      "step": 9776
    },
    {
      "entropy": 0.6627129632979631,
      "epoch": 0.8995462935808305,
      "grad_norm": 0.15756377577781677,
      "learning_rate": 7.001870763946392e-05,
      "loss": 0.6718,
      "mean_token_accuracy": 0.7990637272596359,
      "num_tokens": 312041869.0,
      "step": 9777
    },
    {
      "entropy": 0.7327678613364697,
      "epoch": 0.8996382999522717,
      "grad_norm": 0.160724475979805,
      "learning_rate": 7.001564081332229e-05,
      "loss": 0.7223,
      "mean_token_accuracy": 0.7868462279438972,
      "num_tokens": 312074120.0,
      "step": 9778
    },
    {
      "entropy": 0.7500945702195168,
      "epoch": 0.8997303063237129,
      "grad_norm": 0.1520105004310608,
      "learning_rate": 7.001257398718067e-05,
      "loss": 0.7442,
      "mean_token_accuracy": 0.7810527011752129,
      "num_tokens": 312106380.0,
      "step": 9779
    },
    {
      "entropy": 0.6477057933807373,
      "epoch": 0.8998223126951541,
      "grad_norm": 0.15816226601600647,
      "learning_rate": 7.000950716103905e-05,
      "loss": 0.6332,
      "mean_token_accuracy": 0.8058095760643482,
      "num_tokens": 312138353.0,
      "step": 9780
    },
    {
      "entropy": 0.776277968659997,
      "epoch": 0.8999143190665954,
      "grad_norm": 0.16119402647018433,
      "learning_rate": 7.000644033489742e-05,
      "loss": 0.7653,
      "mean_token_accuracy": 0.7704412788152695,
      "num_tokens": 312170640.0,
      "step": 9781
    },
    {
      "entropy": 0.7974539082497358,
      "epoch": 0.9000063254380366,
      "grad_norm": 0.15462848544120789,
      "learning_rate": 7.000337350875579e-05,
      "loss": 0.7738,
      "mean_token_accuracy": 0.7724439986050129,
      "num_tokens": 312203184.0,
      "step": 9782
    },
    {
      "entropy": 0.7750661857426167,
      "epoch": 0.9000983318094778,
      "grad_norm": 0.14983734488487244,
      "learning_rate": 7.000030668261417e-05,
      "loss": 0.7488,
      "mean_token_accuracy": 0.7768608294427395,
      "num_tokens": 312235086.0,
      "step": 9783
    },
    {
      "entropy": 0.8131045922636986,
      "epoch": 0.900190338180919,
      "grad_norm": 0.15743198990821838,
      "learning_rate": 6.999723985647253e-05,
      "loss": 0.7858,
      "mean_token_accuracy": 0.7686615660786629,
      "num_tokens": 312266360.0,
      "step": 9784
    },
    {
      "entropy": 0.5807728487998247,
      "epoch": 0.9002823445523602,
      "grad_norm": 0.14609964191913605,
      "learning_rate": 6.999417303033092e-05,
      "loss": 0.5671,
      "mean_token_accuracy": 0.8288491927087307,
      "num_tokens": 312298178.0,
      "step": 9785
    },
    {
      "entropy": 0.6535271164029837,
      "epoch": 0.9003743509238015,
      "grad_norm": 0.15155813097953796,
      "learning_rate": 6.999110620418928e-05,
      "loss": 0.649,
      "mean_token_accuracy": 0.8075745217502117,
      "num_tokens": 312329880.0,
      "step": 9786
    },
    {
      "entropy": 0.7291507609188557,
      "epoch": 0.9004663572952427,
      "grad_norm": 0.15699797868728638,
      "learning_rate": 6.998803937804767e-05,
      "loss": 0.7096,
      "mean_token_accuracy": 0.792345691472292,
      "num_tokens": 312362351.0,
      "step": 9787
    },
    {
      "entropy": 0.6720915921032429,
      "epoch": 0.900558363666684,
      "grad_norm": 0.1570935845375061,
      "learning_rate": 6.998497255190603e-05,
      "loss": 0.6651,
      "mean_token_accuracy": 0.8029627352952957,
      "num_tokens": 312394630.0,
      "step": 9788
    },
    {
      "entropy": 0.6465042699128389,
      "epoch": 0.9006503700381251,
      "grad_norm": 0.1733008474111557,
      "learning_rate": 6.99819057257644e-05,
      "loss": 0.6292,
      "mean_token_accuracy": 0.812382772564888,
      "num_tokens": 312425805.0,
      "step": 9789
    },
    {
      "entropy": 0.5883707981556654,
      "epoch": 0.9007423764095663,
      "grad_norm": 0.15316803753376007,
      "learning_rate": 6.997883889962278e-05,
      "loss": 0.5572,
      "mean_token_accuracy": 0.83474525436759,
      "num_tokens": 312458037.0,
      "step": 9790
    },
    {
      "entropy": 0.6612536823377013,
      "epoch": 0.9008343827810076,
      "grad_norm": 0.15339699387550354,
      "learning_rate": 6.997577207348116e-05,
      "loss": 0.658,
      "mean_token_accuracy": 0.8016945905983448,
      "num_tokens": 312489392.0,
      "step": 9791
    },
    {
      "entropy": 0.6629131333902478,
      "epoch": 0.9009263891524488,
      "grad_norm": 0.16087733209133148,
      "learning_rate": 6.997270524733953e-05,
      "loss": 0.6499,
      "mean_token_accuracy": 0.806018091738224,
      "num_tokens": 312521759.0,
      "step": 9792
    },
    {
      "entropy": 0.7252996694296598,
      "epoch": 0.9010183955238901,
      "grad_norm": 0.16281099617481232,
      "learning_rate": 6.99696384211979e-05,
      "loss": 0.7238,
      "mean_token_accuracy": 0.7881408557295799,
      "num_tokens": 312553615.0,
      "step": 9793
    },
    {
      "entropy": 0.6771484483033419,
      "epoch": 0.9011104018953312,
      "grad_norm": 0.1725471466779709,
      "learning_rate": 6.996657159505628e-05,
      "loss": 0.676,
      "mean_token_accuracy": 0.8005300164222717,
      "num_tokens": 312585859.0,
      "step": 9794
    },
    {
      "entropy": 0.6295735482126474,
      "epoch": 0.9012024082667724,
      "grad_norm": 0.14799141883850098,
      "learning_rate": 6.996350476891466e-05,
      "loss": 0.6131,
      "mean_token_accuracy": 0.8164939843118191,
      "num_tokens": 312618247.0,
      "step": 9795
    },
    {
      "entropy": 0.7400970757007599,
      "epoch": 0.9012944146382137,
      "grad_norm": 0.15446551144123077,
      "learning_rate": 6.996043794277303e-05,
      "loss": 0.7329,
      "mean_token_accuracy": 0.7860733978450298,
      "num_tokens": 312650316.0,
      "step": 9796
    },
    {
      "entropy": 0.6991653218865395,
      "epoch": 0.9013864210096549,
      "grad_norm": 0.15732482075691223,
      "learning_rate": 6.99573711166314e-05,
      "loss": 0.6703,
      "mean_token_accuracy": 0.8016501441597939,
      "num_tokens": 312682133.0,
      "step": 9797
    },
    {
      "entropy": 0.7175726126879454,
      "epoch": 0.9014784273810962,
      "grad_norm": 0.1599174588918686,
      "learning_rate": 6.995430429048978e-05,
      "loss": 0.7128,
      "mean_token_accuracy": 0.7875833995640278,
      "num_tokens": 312714815.0,
      "step": 9798
    },
    {
      "entropy": 0.6470885593444109,
      "epoch": 0.9015704337525373,
      "grad_norm": 0.15912048518657684,
      "learning_rate": 6.995123746434815e-05,
      "loss": 0.644,
      "mean_token_accuracy": 0.8081750795245171,
      "num_tokens": 312747105.0,
      "step": 9799
    },
    {
      "entropy": 0.6245025247335434,
      "epoch": 0.9016624401239786,
      "grad_norm": 0.15743839740753174,
      "learning_rate": 6.994817063820653e-05,
      "loss": 0.6203,
      "mean_token_accuracy": 0.8138361908495426,
      "num_tokens": 312779253.0,
      "step": 9800
    },
    {
      "entropy": 0.6972022131085396,
      "epoch": 0.9017544464954198,
      "grad_norm": 0.1532084345817566,
      "learning_rate": 6.99451038120649e-05,
      "loss": 0.6957,
      "mean_token_accuracy": 0.7960177324712276,
      "num_tokens": 312811245.0,
      "step": 9801
    },
    {
      "entropy": 0.6872608549892902,
      "epoch": 0.901846452866861,
      "grad_norm": 0.15582983195781708,
      "learning_rate": 6.994203698592328e-05,
      "loss": 0.6725,
      "mean_token_accuracy": 0.8043621443212032,
      "num_tokens": 312843394.0,
      "step": 9802
    },
    {
      "entropy": 0.7419478930532932,
      "epoch": 0.9019384592383023,
      "grad_norm": 0.16141456365585327,
      "learning_rate": 6.993897015978165e-05,
      "loss": 0.7433,
      "mean_token_accuracy": 0.7835108824074268,
      "num_tokens": 312875048.0,
      "step": 9803
    },
    {
      "entropy": 0.7473692037165165,
      "epoch": 0.9020304656097434,
      "grad_norm": 0.14331267774105072,
      "learning_rate": 6.993590333364001e-05,
      "loss": 0.7292,
      "mean_token_accuracy": 0.7796697542071342,
      "num_tokens": 312906807.0,
      "step": 9804
    },
    {
      "entropy": 0.7191784307360649,
      "epoch": 0.9021224719811847,
      "grad_norm": 0.152790829539299,
      "learning_rate": 6.99328365074984e-05,
      "loss": 0.7207,
      "mean_token_accuracy": 0.7888144962489605,
      "num_tokens": 312939246.0,
      "step": 9805
    },
    {
      "entropy": 0.8391008079051971,
      "epoch": 0.9022144783526259,
      "grad_norm": 0.15501053631305695,
      "learning_rate": 6.992976968135678e-05,
      "loss": 0.8304,
      "mean_token_accuracy": 0.7601708620786667,
      "num_tokens": 312971301.0,
      "step": 9806
    },
    {
      "entropy": 0.6553949918597937,
      "epoch": 0.9023064847240672,
      "grad_norm": 0.1474887728691101,
      "learning_rate": 6.992670285521514e-05,
      "loss": 0.6344,
      "mean_token_accuracy": 0.8092664740979671,
      "num_tokens": 313003646.0,
      "step": 9807
    },
    {
      "entropy": 0.7481492627412081,
      "epoch": 0.9023984910955084,
      "grad_norm": 0.16327811777591705,
      "learning_rate": 6.992363602907351e-05,
      "loss": 0.7432,
      "mean_token_accuracy": 0.7837705723941326,
      "num_tokens": 313035179.0,
      "step": 9808
    },
    {
      "entropy": 0.6252400306984782,
      "epoch": 0.9024904974669495,
      "grad_norm": 0.14379404485225677,
      "learning_rate": 6.992056920293188e-05,
      "loss": 0.5895,
      "mean_token_accuracy": 0.8220630213618279,
      "num_tokens": 313067182.0,
      "step": 9809
    },
    {
      "entropy": 0.6840851828455925,
      "epoch": 0.9025825038383908,
      "grad_norm": 0.15334004163742065,
      "learning_rate": 6.991750237679026e-05,
      "loss": 0.664,
      "mean_token_accuracy": 0.8028981685638428,
      "num_tokens": 313098989.0,
      "step": 9810
    },
    {
      "entropy": 0.6761615984141827,
      "epoch": 0.902674510209832,
      "grad_norm": 0.15886333584785461,
      "learning_rate": 6.991443555064864e-05,
      "loss": 0.6617,
      "mean_token_accuracy": 0.8001260198652744,
      "num_tokens": 313131130.0,
      "step": 9811
    },
    {
      "entropy": 0.7350510582327843,
      "epoch": 0.9027665165812733,
      "grad_norm": 0.16361670196056366,
      "learning_rate": 6.991136872450701e-05,
      "loss": 0.7398,
      "mean_token_accuracy": 0.7817340828478336,
      "num_tokens": 313162607.0,
      "step": 9812
    },
    {
      "entropy": 0.7427537720650434,
      "epoch": 0.9028585229527145,
      "grad_norm": 0.15304584801197052,
      "learning_rate": 6.990830189836538e-05,
      "loss": 0.7334,
      "mean_token_accuracy": 0.7836622074246407,
      "num_tokens": 313194621.0,
      "step": 9813
    },
    {
      "entropy": 0.6531521026045084,
      "epoch": 0.9029505293241556,
      "grad_norm": 0.15353186428546906,
      "learning_rate": 6.990523507222376e-05,
      "loss": 0.6472,
      "mean_token_accuracy": 0.8031473755836487,
      "num_tokens": 313225453.0,
      "step": 9814
    },
    {
      "entropy": 0.607645879033953,
      "epoch": 0.9030425356955969,
      "grad_norm": 0.14837847650051117,
      "learning_rate": 6.990216824608213e-05,
      "loss": 0.5902,
      "mean_token_accuracy": 0.8241151012480259,
      "num_tokens": 313257586.0,
      "step": 9815
    },
    {
      "entropy": 0.8098444268107414,
      "epoch": 0.9031345420670381,
      "grad_norm": 0.1632319986820221,
      "learning_rate": 6.989910141994051e-05,
      "loss": 0.8093,
      "mean_token_accuracy": 0.7615335360169411,
      "num_tokens": 313290234.0,
      "step": 9816
    },
    {
      "entropy": 0.7256257403641939,
      "epoch": 0.9032265484384794,
      "grad_norm": 0.1497650295495987,
      "learning_rate": 6.989603459379888e-05,
      "loss": 0.7246,
      "mean_token_accuracy": 0.7849978767335415,
      "num_tokens": 313322354.0,
      "step": 9817
    },
    {
      "entropy": 0.6722328960895538,
      "epoch": 0.9033185548099206,
      "grad_norm": 0.15327788889408112,
      "learning_rate": 6.989296776765726e-05,
      "loss": 0.6553,
      "mean_token_accuracy": 0.8070054352283478,
      "num_tokens": 313354224.0,
      "step": 9818
    },
    {
      "entropy": 0.7043386250734329,
      "epoch": 0.9034105611813618,
      "grad_norm": 0.16479474306106567,
      "learning_rate": 6.988990094151563e-05,
      "loss": 0.6931,
      "mean_token_accuracy": 0.7894992344081402,
      "num_tokens": 313385581.0,
      "step": 9819
    },
    {
      "entropy": 0.6068550273776054,
      "epoch": 0.903502567552803,
      "grad_norm": 0.14737345278263092,
      "learning_rate": 6.9886834115374e-05,
      "loss": 0.5975,
      "mean_token_accuracy": 0.8187743425369263,
      "num_tokens": 313418144.0,
      "step": 9820
    },
    {
      "entropy": 0.7568378485739231,
      "epoch": 0.9035945739242442,
      "grad_norm": 0.16324695944786072,
      "learning_rate": 6.988376728923238e-05,
      "loss": 0.7385,
      "mean_token_accuracy": 0.7820223942399025,
      "num_tokens": 313450364.0,
      "step": 9821
    },
    {
      "entropy": 0.7064705528318882,
      "epoch": 0.9036865802956855,
      "grad_norm": 0.1449199765920639,
      "learning_rate": 6.988070046309076e-05,
      "loss": 0.6922,
      "mean_token_accuracy": 0.795671783387661,
      "num_tokens": 313482466.0,
      "step": 9822
    },
    {
      "entropy": 0.7375322915613651,
      "epoch": 0.9037785866671267,
      "grad_norm": 0.15684305131435394,
      "learning_rate": 6.987763363694913e-05,
      "loss": 0.7161,
      "mean_token_accuracy": 0.7874334454536438,
      "num_tokens": 313514436.0,
      "step": 9823
    },
    {
      "entropy": 0.6678204871714115,
      "epoch": 0.9038705930385679,
      "grad_norm": 0.15447697043418884,
      "learning_rate": 6.98745668108075e-05,
      "loss": 0.6413,
      "mean_token_accuracy": 0.8077951297163963,
      "num_tokens": 313546445.0,
      "step": 9824
    },
    {
      "entropy": 0.680232772603631,
      "epoch": 0.9039625994100091,
      "grad_norm": 0.15950535237789154,
      "learning_rate": 6.987149998466587e-05,
      "loss": 0.654,
      "mean_token_accuracy": 0.805111538618803,
      "num_tokens": 313578221.0,
      "step": 9825
    },
    {
      "entropy": 0.6654059663414955,
      "epoch": 0.9040546057814504,
      "grad_norm": 0.15601997077465057,
      "learning_rate": 6.986843315852424e-05,
      "loss": 0.661,
      "mean_token_accuracy": 0.8031200058758259,
      "num_tokens": 313610267.0,
      "step": 9826
    },
    {
      "entropy": 0.7716039847582579,
      "epoch": 0.9041466121528916,
      "grad_norm": 0.15488244593143463,
      "learning_rate": 6.986536633238262e-05,
      "loss": 0.7729,
      "mean_token_accuracy": 0.7733152210712433,
      "num_tokens": 313642445.0,
      "step": 9827
    },
    {
      "entropy": 0.7857575621455908,
      "epoch": 0.9042386185243328,
      "grad_norm": 0.1599627435207367,
      "learning_rate": 6.986229950624099e-05,
      "loss": 0.7726,
      "mean_token_accuracy": 0.7767398208379745,
      "num_tokens": 313674039.0,
      "step": 9828
    },
    {
      "entropy": 0.7677183151245117,
      "epoch": 0.904330624895774,
      "grad_norm": 0.1578005999326706,
      "learning_rate": 6.985923268009937e-05,
      "loss": 0.7502,
      "mean_token_accuracy": 0.7838185988366604,
      "num_tokens": 313706567.0,
      "step": 9829
    },
    {
      "entropy": 0.6432377453893423,
      "epoch": 0.9044226312672152,
      "grad_norm": 0.1565162092447281,
      "learning_rate": 6.985616585395774e-05,
      "loss": 0.6294,
      "mean_token_accuracy": 0.8118269480764866,
      "num_tokens": 313738660.0,
      "step": 9830
    },
    {
      "entropy": 0.8904267698526382,
      "epoch": 0.9045146376386565,
      "grad_norm": 0.1676774024963379,
      "learning_rate": 6.985309902781611e-05,
      "loss": 0.8943,
      "mean_token_accuracy": 0.7428970113396645,
      "num_tokens": 313770304.0,
      "step": 9831
    },
    {
      "entropy": 0.6594960726797581,
      "epoch": 0.9046066440100977,
      "grad_norm": 0.15930035710334778,
      "learning_rate": 6.985003220167449e-05,
      "loss": 0.6535,
      "mean_token_accuracy": 0.8064328283071518,
      "num_tokens": 313802441.0,
      "step": 9832
    },
    {
      "entropy": 0.6588748693466187,
      "epoch": 0.904698650381539,
      "grad_norm": 0.15119968354701996,
      "learning_rate": 6.984696537553287e-05,
      "loss": 0.635,
      "mean_token_accuracy": 0.809274435043335,
      "num_tokens": 313834461.0,
      "step": 9833
    },
    {
      "entropy": 0.7757204324007034,
      "epoch": 0.9047906567529801,
      "grad_norm": 0.16427789628505707,
      "learning_rate": 6.984389854939124e-05,
      "loss": 0.7457,
      "mean_token_accuracy": 0.7782172560691833,
      "num_tokens": 313865772.0,
      "step": 9834
    },
    {
      "entropy": 0.7172437068074942,
      "epoch": 0.9048826631244213,
      "grad_norm": 0.1589546650648117,
      "learning_rate": 6.984083172324961e-05,
      "loss": 0.7024,
      "mean_token_accuracy": 0.795226190239191,
      "num_tokens": 313897738.0,
      "step": 9835
    },
    {
      "entropy": 0.7014139154925942,
      "epoch": 0.9049746694958626,
      "grad_norm": 0.1567389965057373,
      "learning_rate": 6.983776489710799e-05,
      "loss": 0.6927,
      "mean_token_accuracy": 0.7946244217455387,
      "num_tokens": 313930135.0,
      "step": 9836
    },
    {
      "entropy": 0.8073681280948222,
      "epoch": 0.9050666758673038,
      "grad_norm": 0.16256089508533478,
      "learning_rate": 6.983469807096637e-05,
      "loss": 0.805,
      "mean_token_accuracy": 0.7723867669701576,
      "num_tokens": 313961878.0,
      "step": 9837
    },
    {
      "entropy": 0.6758123897016048,
      "epoch": 0.9051586822387451,
      "grad_norm": 0.14744770526885986,
      "learning_rate": 6.983163124482474e-05,
      "loss": 0.6657,
      "mean_token_accuracy": 0.8023252449929714,
      "num_tokens": 313994052.0,
      "step": 9838
    },
    {
      "entropy": 0.6766122914850712,
      "epoch": 0.9052506886101862,
      "grad_norm": 0.1522417962551117,
      "learning_rate": 6.98285644186831e-05,
      "loss": 0.6679,
      "mean_token_accuracy": 0.8034966364502907,
      "num_tokens": 314026061.0,
      "step": 9839
    },
    {
      "entropy": 0.7053922209888697,
      "epoch": 0.9053426949816274,
      "grad_norm": 0.15152977406978607,
      "learning_rate": 6.982549759254147e-05,
      "loss": 0.676,
      "mean_token_accuracy": 0.796904381364584,
      "num_tokens": 314057709.0,
      "step": 9840
    },
    {
      "entropy": 0.6486853901296854,
      "epoch": 0.9054347013530687,
      "grad_norm": 0.16112056374549866,
      "learning_rate": 6.982243076639986e-05,
      "loss": 0.646,
      "mean_token_accuracy": 0.8058718405663967,
      "num_tokens": 314089199.0,
      "step": 9841
    },
    {
      "entropy": 0.8038075286895037,
      "epoch": 0.9055267077245099,
      "grad_norm": 0.15666624903678894,
      "learning_rate": 6.981936394025824e-05,
      "loss": 0.7966,
      "mean_token_accuracy": 0.7635356895625591,
      "num_tokens": 314120475.0,
      "step": 9842
    },
    {
      "entropy": 0.722719170153141,
      "epoch": 0.9056187140959512,
      "grad_norm": 0.16917426884174347,
      "learning_rate": 6.98162971141166e-05,
      "loss": 0.7227,
      "mean_token_accuracy": 0.7864150293171406,
      "num_tokens": 314152509.0,
      "step": 9843
    },
    {
      "entropy": 0.7015506383031607,
      "epoch": 0.9057107204673923,
      "grad_norm": 0.15085852146148682,
      "learning_rate": 6.981323028797497e-05,
      "loss": 0.6922,
      "mean_token_accuracy": 0.7976479381322861,
      "num_tokens": 314185142.0,
      "step": 9844
    },
    {
      "entropy": 0.6653766501694918,
      "epoch": 0.9058027268388336,
      "grad_norm": 0.15368232131004333,
      "learning_rate": 6.981016346183335e-05,
      "loss": 0.6559,
      "mean_token_accuracy": 0.8042340315878391,
      "num_tokens": 314217607.0,
      "step": 9845
    },
    {
      "entropy": 0.7172192875295877,
      "epoch": 0.9058947332102748,
      "grad_norm": 0.15314981341362,
      "learning_rate": 6.980709663569172e-05,
      "loss": 0.7099,
      "mean_token_accuracy": 0.7914319187402725,
      "num_tokens": 314250107.0,
      "step": 9846
    },
    {
      "entropy": 0.7581367753446102,
      "epoch": 0.905986739581716,
      "grad_norm": 0.1598554253578186,
      "learning_rate": 6.98040298095501e-05,
      "loss": 0.7593,
      "mean_token_accuracy": 0.7729617692530155,
      "num_tokens": 314282445.0,
      "step": 9847
    },
    {
      "entropy": 0.6987941190600395,
      "epoch": 0.9060787459531573,
      "grad_norm": 0.16324779391288757,
      "learning_rate": 6.980096298340847e-05,
      "loss": 0.6874,
      "mean_token_accuracy": 0.7975902669131756,
      "num_tokens": 314314341.0,
      "step": 9848
    },
    {
      "entropy": 0.7214428950101137,
      "epoch": 0.9061707523245984,
      "grad_norm": 0.15869629383087158,
      "learning_rate": 6.979789615726685e-05,
      "loss": 0.7138,
      "mean_token_accuracy": 0.7877066023647785,
      "num_tokens": 314345873.0,
      "step": 9849
    },
    {
      "entropy": 0.6545142456889153,
      "epoch": 0.9062627586960397,
      "grad_norm": 0.15035808086395264,
      "learning_rate": 6.979482933112522e-05,
      "loss": 0.6478,
      "mean_token_accuracy": 0.8063101731240749,
      "num_tokens": 314378110.0,
      "step": 9850
    },
    {
      "entropy": 0.5233671786263585,
      "epoch": 0.9063547650674809,
      "grad_norm": 0.1542322188615799,
      "learning_rate": 6.979176250498359e-05,
      "loss": 0.5021,
      "mean_token_accuracy": 0.8456229828298092,
      "num_tokens": 314410406.0,
      "step": 9851
    },
    {
      "entropy": 0.7427673805505037,
      "epoch": 0.9064467714389222,
      "grad_norm": 0.15971937775611877,
      "learning_rate": 6.978869567884197e-05,
      "loss": 0.7328,
      "mean_token_accuracy": 0.7830657213926315,
      "num_tokens": 314442603.0,
      "step": 9852
    },
    {
      "entropy": 0.6459930203855038,
      "epoch": 0.9065387778103634,
      "grad_norm": 0.1416565626859665,
      "learning_rate": 6.978562885270035e-05,
      "loss": 0.6359,
      "mean_token_accuracy": 0.8107081167399883,
      "num_tokens": 314474189.0,
      "step": 9853
    },
    {
      "entropy": 0.7264706417918205,
      "epoch": 0.9066307841818045,
      "grad_norm": 0.1594800353050232,
      "learning_rate": 6.978256202655872e-05,
      "loss": 0.6939,
      "mean_token_accuracy": 0.794347170740366,
      "num_tokens": 314506536.0,
      "step": 9854
    },
    {
      "entropy": 0.715314551256597,
      "epoch": 0.9067227905532458,
      "grad_norm": 0.1519995480775833,
      "learning_rate": 6.977949520041709e-05,
      "loss": 0.6853,
      "mean_token_accuracy": 0.7924241572618484,
      "num_tokens": 314537914.0,
      "step": 9855
    },
    {
      "entropy": 0.5832235384732485,
      "epoch": 0.906814796924687,
      "grad_norm": 0.1499062180519104,
      "learning_rate": 6.977642837427547e-05,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8342979177832603,
      "num_tokens": 314569650.0,
      "step": 9856
    },
    {
      "entropy": 0.7124111466109753,
      "epoch": 0.9069068032961283,
      "grad_norm": 0.16589726507663727,
      "learning_rate": 6.977336154813384e-05,
      "loss": 0.7084,
      "mean_token_accuracy": 0.7929227687418461,
      "num_tokens": 314601231.0,
      "step": 9857
    },
    {
      "entropy": 0.8205553442239761,
      "epoch": 0.9069988096675695,
      "grad_norm": 0.15954890847206116,
      "learning_rate": 6.977029472199222e-05,
      "loss": 0.8035,
      "mean_token_accuracy": 0.7618141658604145,
      "num_tokens": 314632984.0,
      "step": 9858
    },
    {
      "entropy": 0.6925916187465191,
      "epoch": 0.9070908160390107,
      "grad_norm": 0.15048789978027344,
      "learning_rate": 6.976722789585059e-05,
      "loss": 0.6834,
      "mean_token_accuracy": 0.7987072877585888,
      "num_tokens": 314665658.0,
      "step": 9859
    },
    {
      "entropy": 0.6845429269596934,
      "epoch": 0.9071828224104519,
      "grad_norm": 0.151750847697258,
      "learning_rate": 6.976416106970897e-05,
      "loss": 0.6759,
      "mean_token_accuracy": 0.8009379543364048,
      "num_tokens": 314697623.0,
      "step": 9860
    },
    {
      "entropy": 0.8663131296634674,
      "epoch": 0.9072748287818931,
      "grad_norm": 0.16075517237186432,
      "learning_rate": 6.976109424356733e-05,
      "loss": 0.8613,
      "mean_token_accuracy": 0.7487509027123451,
      "num_tokens": 314729330.0,
      "step": 9861
    },
    {
      "entropy": 0.7049398459494114,
      "epoch": 0.9073668351533344,
      "grad_norm": 0.15750139951705933,
      "learning_rate": 6.97580274174257e-05,
      "loss": 0.6897,
      "mean_token_accuracy": 0.7961486726999283,
      "num_tokens": 314760419.0,
      "step": 9862
    },
    {
      "entropy": 0.7131492383778095,
      "epoch": 0.9074588415247756,
      "grad_norm": 0.15170803666114807,
      "learning_rate": 6.975496059128408e-05,
      "loss": 0.6942,
      "mean_token_accuracy": 0.7899488434195518,
      "num_tokens": 314792242.0,
      "step": 9863
    },
    {
      "entropy": 0.6848900057375431,
      "epoch": 0.9075508478962168,
      "grad_norm": 0.15193608403205872,
      "learning_rate": 6.975189376514247e-05,
      "loss": 0.6766,
      "mean_token_accuracy": 0.7997283153235912,
      "num_tokens": 314824522.0,
      "step": 9864
    },
    {
      "entropy": 0.7477631997317076,
      "epoch": 0.907642854267658,
      "grad_norm": 0.1609216332435608,
      "learning_rate": 6.974882693900083e-05,
      "loss": 0.7463,
      "mean_token_accuracy": 0.780682161450386,
      "num_tokens": 314855885.0,
      "step": 9865
    },
    {
      "entropy": 0.8301909863948822,
      "epoch": 0.9077348606390993,
      "grad_norm": 0.16159681975841522,
      "learning_rate": 6.97457601128592e-05,
      "loss": 0.8381,
      "mean_token_accuracy": 0.7612665742635727,
      "num_tokens": 314888505.0,
      "step": 9866
    },
    {
      "entropy": 0.7299970611929893,
      "epoch": 0.9078268670105405,
      "grad_norm": 0.15771140158176422,
      "learning_rate": 6.974269328671757e-05,
      "loss": 0.7352,
      "mean_token_accuracy": 0.7817654125392437,
      "num_tokens": 314920186.0,
      "step": 9867
    },
    {
      "entropy": 0.6804525926709175,
      "epoch": 0.9079188733819817,
      "grad_norm": 0.16383729875087738,
      "learning_rate": 6.973962646057596e-05,
      "loss": 0.6758,
      "mean_token_accuracy": 0.8056800402700901,
      "num_tokens": 314952891.0,
      "step": 9868
    },
    {
      "entropy": 0.7460151314735413,
      "epoch": 0.9080108797534229,
      "grad_norm": 0.16657377779483795,
      "learning_rate": 6.973655963443433e-05,
      "loss": 0.7559,
      "mean_token_accuracy": 0.7824374884366989,
      "num_tokens": 314985123.0,
      "step": 9869
    },
    {
      "entropy": 0.7902086544781923,
      "epoch": 0.9081028861248641,
      "grad_norm": 0.15634115040302277,
      "learning_rate": 6.97334928082927e-05,
      "loss": 0.7933,
      "mean_token_accuracy": 0.7682635448873043,
      "num_tokens": 315017221.0,
      "step": 9870
    },
    {
      "entropy": 0.6860272064805031,
      "epoch": 0.9081948924963054,
      "grad_norm": 0.14233368635177612,
      "learning_rate": 6.973042598215107e-05,
      "loss": 0.671,
      "mean_token_accuracy": 0.7971482761204243,
      "num_tokens": 315048919.0,
      "step": 9871
    },
    {
      "entropy": 0.7441874211654067,
      "epoch": 0.9082868988677466,
      "grad_norm": 0.15530644357204437,
      "learning_rate": 6.972735915600945e-05,
      "loss": 0.7195,
      "mean_token_accuracy": 0.7853357866406441,
      "num_tokens": 315081237.0,
      "step": 9872
    },
    {
      "entropy": 0.667265513446182,
      "epoch": 0.9083789052391879,
      "grad_norm": 0.1541728526353836,
      "learning_rate": 6.972429232986783e-05,
      "loss": 0.6428,
      "mean_token_accuracy": 0.8077193722128868,
      "num_tokens": 315113250.0,
      "step": 9873
    },
    {
      "entropy": 0.7807818818837404,
      "epoch": 0.908470911610629,
      "grad_norm": 0.163929283618927,
      "learning_rate": 6.97212255037262e-05,
      "loss": 0.7767,
      "mean_token_accuracy": 0.7719280682504177,
      "num_tokens": 315144935.0,
      "step": 9874
    },
    {
      "entropy": 0.6509709265083075,
      "epoch": 0.9085629179820702,
      "grad_norm": 0.15504074096679688,
      "learning_rate": 6.971815867758457e-05,
      "loss": 0.636,
      "mean_token_accuracy": 0.8091997765004635,
      "num_tokens": 315176776.0,
      "step": 9875
    },
    {
      "entropy": 0.6958451168611646,
      "epoch": 0.9086549243535115,
      "grad_norm": 0.15417562425136566,
      "learning_rate": 6.971509185144295e-05,
      "loss": 0.6752,
      "mean_token_accuracy": 0.8002155870199203,
      "num_tokens": 315208384.0,
      "step": 9876
    },
    {
      "entropy": 0.7998518766835332,
      "epoch": 0.9087469307249527,
      "grad_norm": 0.154031902551651,
      "learning_rate": 6.971202502530132e-05,
      "loss": 0.7922,
      "mean_token_accuracy": 0.7664785161614418,
      "num_tokens": 315239557.0,
      "step": 9877
    },
    {
      "entropy": 0.7380704842507839,
      "epoch": 0.908838937096394,
      "grad_norm": 0.15023864805698395,
      "learning_rate": 6.97089581991597e-05,
      "loss": 0.7334,
      "mean_token_accuracy": 0.7818404249846935,
      "num_tokens": 315271558.0,
      "step": 9878
    },
    {
      "entropy": 0.730588911101222,
      "epoch": 0.9089309434678351,
      "grad_norm": 0.15773837268352509,
      "learning_rate": 6.970589137301806e-05,
      "loss": 0.7112,
      "mean_token_accuracy": 0.791744876652956,
      "num_tokens": 315303612.0,
      "step": 9879
    },
    {
      "entropy": 0.7386471899226308,
      "epoch": 0.9090229498392763,
      "grad_norm": 0.150690957903862,
      "learning_rate": 6.970282454687645e-05,
      "loss": 0.7303,
      "mean_token_accuracy": 0.7835359014570713,
      "num_tokens": 315335941.0,
      "step": 9880
    },
    {
      "entropy": 0.772407777607441,
      "epoch": 0.9091149562107176,
      "grad_norm": 0.16093012690544128,
      "learning_rate": 6.969975772073481e-05,
      "loss": 0.7736,
      "mean_token_accuracy": 0.7700521759688854,
      "num_tokens": 315368059.0,
      "step": 9881
    },
    {
      "entropy": 0.7880233377218246,
      "epoch": 0.9092069625821588,
      "grad_norm": 0.169601708650589,
      "learning_rate": 6.969669089459318e-05,
      "loss": 0.7941,
      "mean_token_accuracy": 0.7673871666193008,
      "num_tokens": 315400769.0,
      "step": 9882
    },
    {
      "entropy": 0.79029724560678,
      "epoch": 0.9092989689536001,
      "grad_norm": 0.15437515079975128,
      "learning_rate": 6.969362406845156e-05,
      "loss": 0.7856,
      "mean_token_accuracy": 0.7692475579679012,
      "num_tokens": 315432357.0,
      "step": 9883
    },
    {
      "entropy": 0.7224316317588091,
      "epoch": 0.9093909753250412,
      "grad_norm": 0.15637865662574768,
      "learning_rate": 6.969055724230994e-05,
      "loss": 0.695,
      "mean_token_accuracy": 0.7957786284387112,
      "num_tokens": 315464452.0,
      "step": 9884
    },
    {
      "entropy": 0.7149686850607395,
      "epoch": 0.9094829816964825,
      "grad_norm": 0.14739848673343658,
      "learning_rate": 6.968749041616831e-05,
      "loss": 0.7005,
      "mean_token_accuracy": 0.7913227789103985,
      "num_tokens": 315496361.0,
      "step": 9885
    },
    {
      "entropy": 0.6809814795851707,
      "epoch": 0.9095749880679237,
      "grad_norm": 0.15764528512954712,
      "learning_rate": 6.968442359002668e-05,
      "loss": 0.672,
      "mean_token_accuracy": 0.796702403575182,
      "num_tokens": 315527718.0,
      "step": 9886
    },
    {
      "entropy": 0.8055642284452915,
      "epoch": 0.9096669944393649,
      "grad_norm": 0.15871556103229523,
      "learning_rate": 6.968135676388506e-05,
      "loss": 0.7961,
      "mean_token_accuracy": 0.7646455578505993,
      "num_tokens": 315559147.0,
      "step": 9887
    },
    {
      "entropy": 0.7623544055968523,
      "epoch": 0.9097590008108062,
      "grad_norm": 0.161536306142807,
      "learning_rate": 6.967828993774343e-05,
      "loss": 0.7483,
      "mean_token_accuracy": 0.7818508930504322,
      "num_tokens": 315591304.0,
      "step": 9888
    },
    {
      "entropy": 0.7663570307195187,
      "epoch": 0.9098510071822473,
      "grad_norm": 0.1599097102880478,
      "learning_rate": 6.967522311160181e-05,
      "loss": 0.7547,
      "mean_token_accuracy": 0.7736945524811745,
      "num_tokens": 315622378.0,
      "step": 9889
    },
    {
      "entropy": 0.7320497836917639,
      "epoch": 0.9099430135536886,
      "grad_norm": 0.1612945795059204,
      "learning_rate": 6.967215628546018e-05,
      "loss": 0.7083,
      "mean_token_accuracy": 0.7909064516425133,
      "num_tokens": 315653684.0,
      "step": 9890
    },
    {
      "entropy": 0.6795080602169037,
      "epoch": 0.9100350199251298,
      "grad_norm": 0.1550017148256302,
      "learning_rate": 6.966908945931856e-05,
      "loss": 0.6768,
      "mean_token_accuracy": 0.7976425476372242,
      "num_tokens": 315685617.0,
      "step": 9891
    },
    {
      "entropy": 0.7301352173089981,
      "epoch": 0.910127026296571,
      "grad_norm": 0.1498243510723114,
      "learning_rate": 6.966602263317693e-05,
      "loss": 0.7021,
      "mean_token_accuracy": 0.795051995664835,
      "num_tokens": 315717018.0,
      "step": 9892
    },
    {
      "entropy": 0.7375743966549635,
      "epoch": 0.9102190326680123,
      "grad_norm": 0.1569632887840271,
      "learning_rate": 6.96629558070353e-05,
      "loss": 0.7273,
      "mean_token_accuracy": 0.7890201769769192,
      "num_tokens": 315748712.0,
      "step": 9893
    },
    {
      "entropy": 0.724428853020072,
      "epoch": 0.9103110390394534,
      "grad_norm": 0.1519523561000824,
      "learning_rate": 6.965988898089368e-05,
      "loss": 0.7305,
      "mean_token_accuracy": 0.7794889062643051,
      "num_tokens": 315780436.0,
      "step": 9894
    },
    {
      "entropy": 0.6701399367302656,
      "epoch": 0.9104030454108947,
      "grad_norm": 0.15622501075267792,
      "learning_rate": 6.965682215475206e-05,
      "loss": 0.6547,
      "mean_token_accuracy": 0.8031398020684719,
      "num_tokens": 315811945.0,
      "step": 9895
    },
    {
      "entropy": 0.6753231612965465,
      "epoch": 0.9104950517823359,
      "grad_norm": 0.16504019498825073,
      "learning_rate": 6.965375532861043e-05,
      "loss": 0.6669,
      "mean_token_accuracy": 0.804222110658884,
      "num_tokens": 315844160.0,
      "step": 9896
    },
    {
      "entropy": 0.7780500873923302,
      "epoch": 0.9105870581537772,
      "grad_norm": 0.15847866237163544,
      "learning_rate": 6.96506885024688e-05,
      "loss": 0.7623,
      "mean_token_accuracy": 0.7760356739163399,
      "num_tokens": 315876419.0,
      "step": 9897
    },
    {
      "entropy": 0.6107933362945914,
      "epoch": 0.9106790645252184,
      "grad_norm": 0.15424805879592896,
      "learning_rate": 6.964762167632716e-05,
      "loss": 0.5928,
      "mean_token_accuracy": 0.8241737969219685,
      "num_tokens": 315909108.0,
      "step": 9898
    },
    {
      "entropy": 0.7619552724063396,
      "epoch": 0.9107710708966595,
      "grad_norm": 0.1561582386493683,
      "learning_rate": 6.964455485018554e-05,
      "loss": 0.7644,
      "mean_token_accuracy": 0.7754602394998074,
      "num_tokens": 315941185.0,
      "step": 9899
    },
    {
      "entropy": 0.6786413658410311,
      "epoch": 0.9108630772681008,
      "grad_norm": 0.15292607247829437,
      "learning_rate": 6.964148802404393e-05,
      "loss": 0.6688,
      "mean_token_accuracy": 0.8009535260498524,
      "num_tokens": 315972650.0,
      "step": 9900
    },
    {
      "entropy": 0.6805456914007664,
      "epoch": 0.910955083639542,
      "grad_norm": 0.15290777385234833,
      "learning_rate": 6.96384211979023e-05,
      "loss": 0.6769,
      "mean_token_accuracy": 0.8034671321511269,
      "num_tokens": 316005178.0,
      "step": 9901
    },
    {
      "entropy": 0.7266144827008247,
      "epoch": 0.9110470900109833,
      "grad_norm": 0.16328927874565125,
      "learning_rate": 6.963535437176066e-05,
      "loss": 0.7242,
      "mean_token_accuracy": 0.7884104773402214,
      "num_tokens": 316036794.0,
      "step": 9902
    },
    {
      "entropy": 0.6110139088705182,
      "epoch": 0.9111390963824245,
      "grad_norm": 0.1522422581911087,
      "learning_rate": 6.963228754561904e-05,
      "loss": 0.5998,
      "mean_token_accuracy": 0.8203346952795982,
      "num_tokens": 316068843.0,
      "step": 9903
    },
    {
      "entropy": 0.7500724289566278,
      "epoch": 0.9112311027538657,
      "grad_norm": 0.15557074546813965,
      "learning_rate": 6.962922071947741e-05,
      "loss": 0.7369,
      "mean_token_accuracy": 0.7861584685742855,
      "num_tokens": 316100142.0,
      "step": 9904
    },
    {
      "entropy": 0.8024079306051135,
      "epoch": 0.9113231091253069,
      "grad_norm": 0.159196138381958,
      "learning_rate": 6.962615389333579e-05,
      "loss": 0.8012,
      "mean_token_accuracy": 0.767073854804039,
      "num_tokens": 316132682.0,
      "step": 9905
    },
    {
      "entropy": 0.6969160605221987,
      "epoch": 0.9114151154967481,
      "grad_norm": 0.14915065467357635,
      "learning_rate": 6.962308706719416e-05,
      "loss": 0.6752,
      "mean_token_accuracy": 0.7976376675069332,
      "num_tokens": 316164745.0,
      "step": 9906
    },
    {
      "entropy": 0.6650022082030773,
      "epoch": 0.9115071218681894,
      "grad_norm": 0.14503569900989532,
      "learning_rate": 6.962002024105254e-05,
      "loss": 0.6454,
      "mean_token_accuracy": 0.8059107214212418,
      "num_tokens": 316197030.0,
      "step": 9907
    },
    {
      "entropy": 0.7657595723867416,
      "epoch": 0.9115991282396306,
      "grad_norm": 0.15983769297599792,
      "learning_rate": 6.961695341491091e-05,
      "loss": 0.7438,
      "mean_token_accuracy": 0.7838235534727573,
      "num_tokens": 316229323.0,
      "step": 9908
    },
    {
      "entropy": 0.6134783467277884,
      "epoch": 0.9116911346110718,
      "grad_norm": 0.156590536236763,
      "learning_rate": 6.961388658876928e-05,
      "loss": 0.5958,
      "mean_token_accuracy": 0.8208448886871338,
      "num_tokens": 316261351.0,
      "step": 9909
    },
    {
      "entropy": 0.7356185307726264,
      "epoch": 0.911783140982513,
      "grad_norm": 0.15087851881980896,
      "learning_rate": 6.961081976262766e-05,
      "loss": 0.7078,
      "mean_token_accuracy": 0.7868572361767292,
      "num_tokens": 316293122.0,
      "step": 9910
    },
    {
      "entropy": 0.5966942282393575,
      "epoch": 0.9118751473539543,
      "grad_norm": 0.14734426140785217,
      "learning_rate": 6.960775293648604e-05,
      "loss": 0.5789,
      "mean_token_accuracy": 0.825727041810751,
      "num_tokens": 316325292.0,
      "step": 9911
    },
    {
      "entropy": 0.676892539486289,
      "epoch": 0.9119671537253955,
      "grad_norm": 0.15715257823467255,
      "learning_rate": 6.960468611034441e-05,
      "loss": 0.6701,
      "mean_token_accuracy": 0.79983951151371,
      "num_tokens": 316356630.0,
      "step": 9912
    },
    {
      "entropy": 0.7365404553711414,
      "epoch": 0.9120591600968367,
      "grad_norm": 0.1572246551513672,
      "learning_rate": 6.960161928420278e-05,
      "loss": 0.735,
      "mean_token_accuracy": 0.7826876901090145,
      "num_tokens": 316389050.0,
      "step": 9913
    },
    {
      "entropy": 0.7741962485015392,
      "epoch": 0.9121511664682779,
      "grad_norm": 0.15014219284057617,
      "learning_rate": 6.959855245806114e-05,
      "loss": 0.7641,
      "mean_token_accuracy": 0.7709750719368458,
      "num_tokens": 316421038.0,
      "step": 9914
    },
    {
      "entropy": 0.8368560876697302,
      "epoch": 0.9122431728397191,
      "grad_norm": 0.16129636764526367,
      "learning_rate": 6.959548563191954e-05,
      "loss": 0.8407,
      "mean_token_accuracy": 0.7535016313195229,
      "num_tokens": 316453116.0,
      "step": 9915
    },
    {
      "entropy": 0.6642161794006824,
      "epoch": 0.9123351792111604,
      "grad_norm": 0.14657005667686462,
      "learning_rate": 6.95924188057779e-05,
      "loss": 0.6519,
      "mean_token_accuracy": 0.8058983273804188,
      "num_tokens": 316484808.0,
      "step": 9916
    },
    {
      "entropy": 0.7492212830111384,
      "epoch": 0.9124271855826016,
      "grad_norm": 0.1566193848848343,
      "learning_rate": 6.958935197963627e-05,
      "loss": 0.7365,
      "mean_token_accuracy": 0.7807813212275505,
      "num_tokens": 316516443.0,
      "step": 9917
    },
    {
      "entropy": 0.68883216381073,
      "epoch": 0.9125191919540429,
      "grad_norm": 0.15228170156478882,
      "learning_rate": 6.958628515349466e-05,
      "loss": 0.6752,
      "mean_token_accuracy": 0.8013165891170502,
      "num_tokens": 316548731.0,
      "step": 9918
    },
    {
      "entropy": 0.7703688591718674,
      "epoch": 0.912611198325484,
      "grad_norm": 0.16678579151630402,
      "learning_rate": 6.958321832735302e-05,
      "loss": 0.7711,
      "mean_token_accuracy": 0.7749440148472786,
      "num_tokens": 316581055.0,
      "step": 9919
    },
    {
      "entropy": 0.6916721183806658,
      "epoch": 0.9127032046969252,
      "grad_norm": 0.1576404571533203,
      "learning_rate": 6.95801515012114e-05,
      "loss": 0.69,
      "mean_token_accuracy": 0.7959622405469418,
      "num_tokens": 316612336.0,
      "step": 9920
    },
    {
      "entropy": 0.7263891790062189,
      "epoch": 0.9127952110683665,
      "grad_norm": 0.15550215542316437,
      "learning_rate": 6.957708467506977e-05,
      "loss": 0.7052,
      "mean_token_accuracy": 0.7916035950183868,
      "num_tokens": 316644814.0,
      "step": 9921
    },
    {
      "entropy": 0.6432506218552589,
      "epoch": 0.9128872174398077,
      "grad_norm": 0.16024400293827057,
      "learning_rate": 6.957401784892815e-05,
      "loss": 0.6359,
      "mean_token_accuracy": 0.8109251148998737,
      "num_tokens": 316676767.0,
      "step": 9922
    },
    {
      "entropy": 0.6867733504623175,
      "epoch": 0.912979223811249,
      "grad_norm": 0.14714451134204865,
      "learning_rate": 6.957095102278652e-05,
      "loss": 0.6704,
      "mean_token_accuracy": 0.7995147295296192,
      "num_tokens": 316709245.0,
      "step": 9923
    },
    {
      "entropy": 0.6895167902112007,
      "epoch": 0.9130712301826901,
      "grad_norm": 0.15938501060009003,
      "learning_rate": 6.956788419664489e-05,
      "loss": 0.6853,
      "mean_token_accuracy": 0.8007963374257088,
      "num_tokens": 316741559.0,
      "step": 9924
    },
    {
      "entropy": 0.7356048617511988,
      "epoch": 0.9131632365541313,
      "grad_norm": 0.1528223603963852,
      "learning_rate": 6.956481737050327e-05,
      "loss": 0.7408,
      "mean_token_accuracy": 0.7815223149955273,
      "num_tokens": 316773015.0,
      "step": 9925
    },
    {
      "entropy": 0.6348440367728472,
      "epoch": 0.9132552429255726,
      "grad_norm": 0.17086344957351685,
      "learning_rate": 6.956175054436165e-05,
      "loss": 0.6267,
      "mean_token_accuracy": 0.8143207542598248,
      "num_tokens": 316804567.0,
      "step": 9926
    },
    {
      "entropy": 0.7385774347931147,
      "epoch": 0.9133472492970138,
      "grad_norm": 0.14791327714920044,
      "learning_rate": 6.955868371822002e-05,
      "loss": 0.7265,
      "mean_token_accuracy": 0.7874362953007221,
      "num_tokens": 316836472.0,
      "step": 9927
    },
    {
      "entropy": 0.7776539586484432,
      "epoch": 0.9134392556684551,
      "grad_norm": 0.15791702270507812,
      "learning_rate": 6.955561689207839e-05,
      "loss": 0.7629,
      "mean_token_accuracy": 0.7753236778080463,
      "num_tokens": 316868662.0,
      "step": 9928
    },
    {
      "entropy": 0.6638348661363125,
      "epoch": 0.9135312620398962,
      "grad_norm": 0.15338610112667084,
      "learning_rate": 6.955255006593676e-05,
      "loss": 0.6422,
      "mean_token_accuracy": 0.807020153850317,
      "num_tokens": 316901054.0,
      "step": 9929
    },
    {
      "entropy": 0.7215070929378271,
      "epoch": 0.9136232684113375,
      "grad_norm": 0.15438160300254822,
      "learning_rate": 6.954948323979514e-05,
      "loss": 0.7167,
      "mean_token_accuracy": 0.7853221409022808,
      "num_tokens": 316932458.0,
      "step": 9930
    },
    {
      "entropy": 0.7091897567734122,
      "epoch": 0.9137152747827787,
      "grad_norm": 0.1550876647233963,
      "learning_rate": 6.954641641365352e-05,
      "loss": 0.7005,
      "mean_token_accuracy": 0.7920494973659515,
      "num_tokens": 316964040.0,
      "step": 9931
    },
    {
      "entropy": 0.783055561594665,
      "epoch": 0.9138072811542199,
      "grad_norm": 0.15218020975589752,
      "learning_rate": 6.954334958751189e-05,
      "loss": 0.767,
      "mean_token_accuracy": 0.7738208547234535,
      "num_tokens": 316996390.0,
      "step": 9932
    },
    {
      "entropy": 0.7263622842729092,
      "epoch": 0.9138992875256612,
      "grad_norm": 0.15355050563812256,
      "learning_rate": 6.954028276137026e-05,
      "loss": 0.7013,
      "mean_token_accuracy": 0.7907178401947021,
      "num_tokens": 317028433.0,
      "step": 9933
    },
    {
      "entropy": 0.6992142908275127,
      "epoch": 0.9139912938971023,
      "grad_norm": 0.1508665382862091,
      "learning_rate": 6.953721593522864e-05,
      "loss": 0.6867,
      "mean_token_accuracy": 0.793359499424696,
      "num_tokens": 317059441.0,
      "step": 9934
    },
    {
      "entropy": 0.7658503837883472,
      "epoch": 0.9140833002685436,
      "grad_norm": 0.15848466753959656,
      "learning_rate": 6.9534149109087e-05,
      "loss": 0.7643,
      "mean_token_accuracy": 0.7756367437541485,
      "num_tokens": 317091865.0,
      "step": 9935
    },
    {
      "entropy": 0.6607688553631306,
      "epoch": 0.9141753066399848,
      "grad_norm": 0.1535189300775528,
      "learning_rate": 6.953108228294539e-05,
      "loss": 0.6651,
      "mean_token_accuracy": 0.8036913499236107,
      "num_tokens": 317123515.0,
      "step": 9936
    },
    {
      "entropy": 0.7675842326134443,
      "epoch": 0.914267313011426,
      "grad_norm": 0.157068133354187,
      "learning_rate": 6.952801545680375e-05,
      "loss": 0.7702,
      "mean_token_accuracy": 0.7753793857991695,
      "num_tokens": 317155534.0,
      "step": 9937
    },
    {
      "entropy": 0.6453035892918706,
      "epoch": 0.9143593193828673,
      "grad_norm": 0.15001808106899261,
      "learning_rate": 6.952494863066214e-05,
      "loss": 0.6292,
      "mean_token_accuracy": 0.8124967105686665,
      "num_tokens": 317187207.0,
      "step": 9938
    },
    {
      "entropy": 0.7825527247041464,
      "epoch": 0.9144513257543084,
      "grad_norm": 0.16372056305408478,
      "learning_rate": 6.95218818045205e-05,
      "loss": 0.7802,
      "mean_token_accuracy": 0.7702504396438599,
      "num_tokens": 317219125.0,
      "step": 9939
    },
    {
      "entropy": 0.7626891061663628,
      "epoch": 0.9145433321257497,
      "grad_norm": 0.1572502851486206,
      "learning_rate": 6.951881497837887e-05,
      "loss": 0.753,
      "mean_token_accuracy": 0.7739312537014484,
      "num_tokens": 317250975.0,
      "step": 9940
    },
    {
      "entropy": 0.6499994248151779,
      "epoch": 0.9146353384971909,
      "grad_norm": 0.1503913253545761,
      "learning_rate": 6.951574815223725e-05,
      "loss": 0.6217,
      "mean_token_accuracy": 0.8124854378402233,
      "num_tokens": 317282834.0,
      "step": 9941
    },
    {
      "entropy": 0.6257903105579317,
      "epoch": 0.9147273448686322,
      "grad_norm": 0.1486394703388214,
      "learning_rate": 6.951268132609563e-05,
      "loss": 0.6044,
      "mean_token_accuracy": 0.8177580423653126,
      "num_tokens": 317314146.0,
      "step": 9942
    },
    {
      "entropy": 0.7766043171286583,
      "epoch": 0.9148193512400734,
      "grad_norm": 0.16053040325641632,
      "learning_rate": 6.9509614499954e-05,
      "loss": 0.7622,
      "mean_token_accuracy": 0.7725018784403801,
      "num_tokens": 317346160.0,
      "step": 9943
    },
    {
      "entropy": 0.7040560031309724,
      "epoch": 0.9149113576115147,
      "grad_norm": 0.15694858133792877,
      "learning_rate": 6.950654767381237e-05,
      "loss": 0.7067,
      "mean_token_accuracy": 0.7939725518226624,
      "num_tokens": 317378457.0,
      "step": 9944
    },
    {
      "entropy": 0.8519250098615885,
      "epoch": 0.9150033639829558,
      "grad_norm": 0.16589556634426117,
      "learning_rate": 6.950348084767075e-05,
      "loss": 0.8374,
      "mean_token_accuracy": 0.7552572302520275,
      "num_tokens": 317410322.0,
      "step": 9945
    },
    {
      "entropy": 0.5933488458395004,
      "epoch": 0.915095370354397,
      "grad_norm": 0.1727464497089386,
      "learning_rate": 6.950041402152912e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.8226212970912457,
      "num_tokens": 317442262.0,
      "step": 9946
    },
    {
      "entropy": 0.5894409464672208,
      "epoch": 0.9151873767258383,
      "grad_norm": 0.15855944156646729,
      "learning_rate": 6.94973471953875e-05,
      "loss": 0.5688,
      "mean_token_accuracy": 0.8311855345964432,
      "num_tokens": 317474451.0,
      "step": 9947
    },
    {
      "entropy": 0.7775322217494249,
      "epoch": 0.9152793830972795,
      "grad_norm": 0.15149272978305817,
      "learning_rate": 6.949428036924587e-05,
      "loss": 0.7684,
      "mean_token_accuracy": 0.7776103429496288,
      "num_tokens": 317507168.0,
      "step": 9948
    },
    {
      "entropy": 0.6891325674951077,
      "epoch": 0.9153713894687208,
      "grad_norm": 0.15433643758296967,
      "learning_rate": 6.949121354310425e-05,
      "loss": 0.6834,
      "mean_token_accuracy": 0.8002696111798286,
      "num_tokens": 317538591.0,
      "step": 9949
    },
    {
      "entropy": 0.7655931916087866,
      "epoch": 0.9154633958401619,
      "grad_norm": 0.15250885486602783,
      "learning_rate": 6.948814671696262e-05,
      "loss": 0.7512,
      "mean_token_accuracy": 0.7792609818279743,
      "num_tokens": 317570465.0,
      "step": 9950
    },
    {
      "entropy": 0.7524932948872447,
      "epoch": 0.9155554022116031,
      "grad_norm": 0.15422415733337402,
      "learning_rate": 6.948507989082099e-05,
      "loss": 0.7364,
      "mean_token_accuracy": 0.7903403602540493,
      "num_tokens": 317603111.0,
      "step": 9951
    },
    {
      "entropy": 0.6946470532566309,
      "epoch": 0.9156474085830444,
      "grad_norm": 0.1624579131603241,
      "learning_rate": 6.948201306467937e-05,
      "loss": 0.7003,
      "mean_token_accuracy": 0.7896745726466179,
      "num_tokens": 317635705.0,
      "step": 9952
    },
    {
      "entropy": 0.7080320604145527,
      "epoch": 0.9157394149544856,
      "grad_norm": 0.1512802392244339,
      "learning_rate": 6.947894623853775e-05,
      "loss": 0.6894,
      "mean_token_accuracy": 0.7951046787202358,
      "num_tokens": 317668220.0,
      "step": 9953
    },
    {
      "entropy": 0.6640472151339054,
      "epoch": 0.9158314213259269,
      "grad_norm": 0.1492760181427002,
      "learning_rate": 6.947587941239612e-05,
      "loss": 0.6406,
      "mean_token_accuracy": 0.8076999634504318,
      "num_tokens": 317700645.0,
      "step": 9954
    },
    {
      "entropy": 0.6919131334871054,
      "epoch": 0.915923427697368,
      "grad_norm": 0.1529681235551834,
      "learning_rate": 6.947281258625448e-05,
      "loss": 0.6832,
      "mean_token_accuracy": 0.7965882793068886,
      "num_tokens": 317732390.0,
      "step": 9955
    },
    {
      "entropy": 0.7654892429709435,
      "epoch": 0.9160154340688093,
      "grad_norm": 0.15937885642051697,
      "learning_rate": 6.946974576011287e-05,
      "loss": 0.766,
      "mean_token_accuracy": 0.7724371924996376,
      "num_tokens": 317764658.0,
      "step": 9956
    },
    {
      "entropy": 0.8604301046580076,
      "epoch": 0.9161074404402505,
      "grad_norm": 0.1602267175912857,
      "learning_rate": 6.946667893397125e-05,
      "loss": 0.8426,
      "mean_token_accuracy": 0.7526186183094978,
      "num_tokens": 317796124.0,
      "step": 9957
    },
    {
      "entropy": 0.7300421670079231,
      "epoch": 0.9161994468116917,
      "grad_norm": 0.16083726286888123,
      "learning_rate": 6.946361210782961e-05,
      "loss": 0.7055,
      "mean_token_accuracy": 0.794402714818716,
      "num_tokens": 317828502.0,
      "step": 9958
    },
    {
      "entropy": 0.7887132130563259,
      "epoch": 0.916291453183133,
      "grad_norm": 0.15655331313610077,
      "learning_rate": 6.946054528168798e-05,
      "loss": 0.7823,
      "mean_token_accuracy": 0.7685951963067055,
      "num_tokens": 317859744.0,
      "step": 9959
    },
    {
      "entropy": 0.7913587242364883,
      "epoch": 0.9163834595545741,
      "grad_norm": 0.15333998203277588,
      "learning_rate": 6.945747845554635e-05,
      "loss": 0.7762,
      "mean_token_accuracy": 0.7680510580539703,
      "num_tokens": 317892113.0,
      "step": 9960
    },
    {
      "entropy": 0.6925205439329147,
      "epoch": 0.9164754659260154,
      "grad_norm": 0.14994674921035767,
      "learning_rate": 6.945441162940473e-05,
      "loss": 0.6745,
      "mean_token_accuracy": 0.7951846681535244,
      "num_tokens": 317924314.0,
      "step": 9961
    },
    {
      "entropy": 0.7596172243356705,
      "epoch": 0.9165674722974566,
      "grad_norm": 0.15743465721607208,
      "learning_rate": 6.945134480326311e-05,
      "loss": 0.7489,
      "mean_token_accuracy": 0.7793570682406425,
      "num_tokens": 317956737.0,
      "step": 9962
    },
    {
      "entropy": 0.6805580146610737,
      "epoch": 0.9166594786688979,
      "grad_norm": 0.15861579775810242,
      "learning_rate": 6.944827797712148e-05,
      "loss": 0.6794,
      "mean_token_accuracy": 0.8001701943576336,
      "num_tokens": 317988283.0,
      "step": 9963
    },
    {
      "entropy": 0.7812789753079414,
      "epoch": 0.9167514850403391,
      "grad_norm": 0.15527063608169556,
      "learning_rate": 6.944521115097985e-05,
      "loss": 0.7733,
      "mean_token_accuracy": 0.7679052799940109,
      "num_tokens": 318020353.0,
      "step": 9964
    },
    {
      "entropy": 0.706810999661684,
      "epoch": 0.9168434914117802,
      "grad_norm": 0.1532265692949295,
      "learning_rate": 6.944214432483823e-05,
      "loss": 0.7033,
      "mean_token_accuracy": 0.7902745679020882,
      "num_tokens": 318052699.0,
      "step": 9965
    },
    {
      "entropy": 0.6937880013138056,
      "epoch": 0.9169354977832215,
      "grad_norm": 0.16165407001972198,
      "learning_rate": 6.94390774986966e-05,
      "loss": 0.6804,
      "mean_token_accuracy": 0.7978475093841553,
      "num_tokens": 318083970.0,
      "step": 9966
    },
    {
      "entropy": 0.6965666711330414,
      "epoch": 0.9170275041546627,
      "grad_norm": 0.16809342801570892,
      "learning_rate": 6.943601067255498e-05,
      "loss": 0.6745,
      "mean_token_accuracy": 0.7992452010512352,
      "num_tokens": 318115294.0,
      "step": 9967
    },
    {
      "entropy": 0.6503252424299717,
      "epoch": 0.917119510526104,
      "grad_norm": 0.16171392798423767,
      "learning_rate": 6.943294384641335e-05,
      "loss": 0.6514,
      "mean_token_accuracy": 0.8005641810595989,
      "num_tokens": 318147273.0,
      "step": 9968
    },
    {
      "entropy": 0.8111811578273773,
      "epoch": 0.9172115168975452,
      "grad_norm": 0.16488702595233917,
      "learning_rate": 6.942987702027173e-05,
      "loss": 0.8047,
      "mean_token_accuracy": 0.7628498487174511,
      "num_tokens": 318178734.0,
      "step": 9969
    },
    {
      "entropy": 0.7196056060492992,
      "epoch": 0.9173035232689863,
      "grad_norm": 0.15344573557376862,
      "learning_rate": 6.94268101941301e-05,
      "loss": 0.6999,
      "mean_token_accuracy": 0.7865083701908588,
      "num_tokens": 318210522.0,
      "step": 9970
    },
    {
      "entropy": 0.6921303365379572,
      "epoch": 0.9173955296404276,
      "grad_norm": 0.15634600818157196,
      "learning_rate": 6.942374336798846e-05,
      "loss": 0.6817,
      "mean_token_accuracy": 0.7995149195194244,
      "num_tokens": 318242558.0,
      "step": 9971
    },
    {
      "entropy": 0.7527589909732342,
      "epoch": 0.9174875360118688,
      "grad_norm": 0.16255822777748108,
      "learning_rate": 6.942067654184685e-05,
      "loss": 0.747,
      "mean_token_accuracy": 0.7785642631351948,
      "num_tokens": 318273794.0,
      "step": 9972
    },
    {
      "entropy": 0.6917078038677573,
      "epoch": 0.9175795423833101,
      "grad_norm": 0.15654051303863525,
      "learning_rate": 6.941760971570523e-05,
      "loss": 0.6774,
      "mean_token_accuracy": 0.7994005270302296,
      "num_tokens": 318305236.0,
      "step": 9973
    },
    {
      "entropy": 0.6421024547889829,
      "epoch": 0.9176715487547513,
      "grad_norm": 0.14953060448169708,
      "learning_rate": 6.94145428895636e-05,
      "loss": 0.6374,
      "mean_token_accuracy": 0.8121487237513065,
      "num_tokens": 318337496.0,
      "step": 9974
    },
    {
      "entropy": 0.727041058242321,
      "epoch": 0.9177635551261925,
      "grad_norm": 0.15349675714969635,
      "learning_rate": 6.941147606342196e-05,
      "loss": 0.7083,
      "mean_token_accuracy": 0.7896493934094906,
      "num_tokens": 318369099.0,
      "step": 9975
    },
    {
      "entropy": 0.6205169139429927,
      "epoch": 0.9178555614976337,
      "grad_norm": 0.15125249326229095,
      "learning_rate": 6.940840923728034e-05,
      "loss": 0.5933,
      "mean_token_accuracy": 0.8223515525460243,
      "num_tokens": 318401036.0,
      "step": 9976
    },
    {
      "entropy": 0.7604425735771656,
      "epoch": 0.9179475678690749,
      "grad_norm": 0.16217675805091858,
      "learning_rate": 6.940534241113871e-05,
      "loss": 0.7692,
      "mean_token_accuracy": 0.7745232656598091,
      "num_tokens": 318433104.0,
      "step": 9977
    },
    {
      "entropy": 0.6258802730590105,
      "epoch": 0.9180395742405162,
      "grad_norm": 0.14057479798793793,
      "learning_rate": 6.94022755849971e-05,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8177808783948421,
      "num_tokens": 318464797.0,
      "step": 9978
    },
    {
      "entropy": 0.6739685516804457,
      "epoch": 0.9181315806119574,
      "grad_norm": 0.15312393009662628,
      "learning_rate": 6.939920875885546e-05,
      "loss": 0.6619,
      "mean_token_accuracy": 0.8062943294644356,
      "num_tokens": 318496623.0,
      "step": 9979
    },
    {
      "entropy": 0.7782473489642143,
      "epoch": 0.9182235869833986,
      "grad_norm": 0.14988350868225098,
      "learning_rate": 6.939614193271384e-05,
      "loss": 0.7629,
      "mean_token_accuracy": 0.7718226723372936,
      "num_tokens": 318528792.0,
      "step": 9980
    },
    {
      "entropy": 0.7389786336570978,
      "epoch": 0.9183155933548398,
      "grad_norm": 0.1520860493183136,
      "learning_rate": 6.939307510657221e-05,
      "loss": 0.7349,
      "mean_token_accuracy": 0.78506725654006,
      "num_tokens": 318561086.0,
      "step": 9981
    },
    {
      "entropy": 0.6920284256339073,
      "epoch": 0.918407599726281,
      "grad_norm": 0.1512753814458847,
      "learning_rate": 6.939000828043058e-05,
      "loss": 0.6761,
      "mean_token_accuracy": 0.7995144613087177,
      "num_tokens": 318593072.0,
      "step": 9982
    },
    {
      "entropy": 0.7174007277935743,
      "epoch": 0.9184996060977223,
      "grad_norm": 0.14879091084003448,
      "learning_rate": 6.938694145428896e-05,
      "loss": 0.6933,
      "mean_token_accuracy": 0.7916452772915363,
      "num_tokens": 318624840.0,
      "step": 9983
    },
    {
      "entropy": 0.7099181059747934,
      "epoch": 0.9185916124691635,
      "grad_norm": 0.1496090143918991,
      "learning_rate": 6.938387462814734e-05,
      "loss": 0.69,
      "mean_token_accuracy": 0.7962145544588566,
      "num_tokens": 318656532.0,
      "step": 9984
    },
    {
      "entropy": 0.7843203265219927,
      "epoch": 0.9186836188406047,
      "grad_norm": 0.16409288346767426,
      "learning_rate": 6.938080780200571e-05,
      "loss": 0.7783,
      "mean_token_accuracy": 0.7751600220799446,
      "num_tokens": 318688757.0,
      "step": 9985
    },
    {
      "entropy": 0.7237826623022556,
      "epoch": 0.9187756252120459,
      "grad_norm": 0.17175886034965515,
      "learning_rate": 6.937774097586408e-05,
      "loss": 0.7393,
      "mean_token_accuracy": 0.781003650277853,
      "num_tokens": 318721371.0,
      "step": 9986
    },
    {
      "entropy": 0.6601234581321478,
      "epoch": 0.9188676315834872,
      "grad_norm": 0.15273530781269073,
      "learning_rate": 6.937467414972245e-05,
      "loss": 0.656,
      "mean_token_accuracy": 0.8029108494520187,
      "num_tokens": 318753413.0,
      "step": 9987
    },
    {
      "entropy": 0.7051912695169449,
      "epoch": 0.9189596379549284,
      "grad_norm": 0.15639397501945496,
      "learning_rate": 6.937160732358084e-05,
      "loss": 0.6943,
      "mean_token_accuracy": 0.789123423397541,
      "num_tokens": 318784883.0,
      "step": 9988
    },
    {
      "entropy": 0.7705016061663628,
      "epoch": 0.9190516443263697,
      "grad_norm": 0.1670760214328766,
      "learning_rate": 6.936854049743921e-05,
      "loss": 0.7645,
      "mean_token_accuracy": 0.7700841501355171,
      "num_tokens": 318816124.0,
      "step": 9989
    },
    {
      "entropy": 0.799666678532958,
      "epoch": 0.9191436506978108,
      "grad_norm": 0.1554257571697235,
      "learning_rate": 6.936547367129758e-05,
      "loss": 0.8133,
      "mean_token_accuracy": 0.761804610490799,
      "num_tokens": 318847286.0,
      "step": 9990
    },
    {
      "entropy": 0.7293779980391264,
      "epoch": 0.919235657069252,
      "grad_norm": 0.15962086617946625,
      "learning_rate": 6.936240684515594e-05,
      "loss": 0.7247,
      "mean_token_accuracy": 0.7837360315024853,
      "num_tokens": 318878933.0,
      "step": 9991
    },
    {
      "entropy": 0.7169304648414254,
      "epoch": 0.9193276634406933,
      "grad_norm": 0.1580122858285904,
      "learning_rate": 6.935934001901433e-05,
      "loss": 0.6972,
      "mean_token_accuracy": 0.7959704324603081,
      "num_tokens": 318910208.0,
      "step": 9992
    },
    {
      "entropy": 0.7372545003890991,
      "epoch": 0.9194196698121345,
      "grad_norm": 0.16247539222240448,
      "learning_rate": 6.935627319287271e-05,
      "loss": 0.7333,
      "mean_token_accuracy": 0.7839266993105412,
      "num_tokens": 318941539.0,
      "step": 9993
    },
    {
      "entropy": 0.7418679129332304,
      "epoch": 0.9195116761835758,
      "grad_norm": 0.152090921998024,
      "learning_rate": 6.935320636673107e-05,
      "loss": 0.7195,
      "mean_token_accuracy": 0.7925521582365036,
      "num_tokens": 318974161.0,
      "step": 9994
    },
    {
      "entropy": 0.7352519817650318,
      "epoch": 0.9196036825550169,
      "grad_norm": 0.15448540449142456,
      "learning_rate": 6.935013954058944e-05,
      "loss": 0.7049,
      "mean_token_accuracy": 0.7905003167688847,
      "num_tokens": 319006771.0,
      "step": 9995
    },
    {
      "entropy": 0.7006584238260984,
      "epoch": 0.9196956889264581,
      "grad_norm": 0.1504005491733551,
      "learning_rate": 6.934707271444782e-05,
      "loss": 0.6759,
      "mean_token_accuracy": 0.7969538532197475,
      "num_tokens": 319038676.0,
      "step": 9996
    },
    {
      "entropy": 0.7115880157798529,
      "epoch": 0.9197876952978994,
      "grad_norm": 0.1473681628704071,
      "learning_rate": 6.934400588830619e-05,
      "loss": 0.6931,
      "mean_token_accuracy": 0.7941428981721401,
      "num_tokens": 319070059.0,
      "step": 9997
    },
    {
      "entropy": 0.7326098997145891,
      "epoch": 0.9198797016693406,
      "grad_norm": 0.15123046934604645,
      "learning_rate": 6.934093906216457e-05,
      "loss": 0.722,
      "mean_token_accuracy": 0.7821482755243778,
      "num_tokens": 319101855.0,
      "step": 9998
    },
    {
      "entropy": 0.7548060845583677,
      "epoch": 0.9199717080407819,
      "grad_norm": 0.15556462109088898,
      "learning_rate": 6.933787223602294e-05,
      "loss": 0.7412,
      "mean_token_accuracy": 0.7825354300439358,
      "num_tokens": 319134332.0,
      "step": 9999
    },
    {
      "entropy": 0.7195710502564907,
      "epoch": 0.920063714412223,
      "grad_norm": 0.15761610865592957,
      "learning_rate": 6.933480540988132e-05,
      "loss": 0.7045,
      "mean_token_accuracy": 0.7937524653971195,
      "num_tokens": 319166406.0,
      "step": 10000
    },
    {
      "entropy": 0.7317797038704157,
      "epoch": 0.9201557207836643,
      "grad_norm": 0.15185149013996124,
      "learning_rate": 6.933173858373969e-05,
      "loss": 0.7069,
      "mean_token_accuracy": 0.7913852073252201,
      "num_tokens": 319198252.0,
      "step": 10001
    },
    {
      "entropy": 0.6521715186536312,
      "epoch": 0.9202477271551055,
      "grad_norm": 0.15052390098571777,
      "learning_rate": 6.932867175759806e-05,
      "loss": 0.638,
      "mean_token_accuracy": 0.8097384683787823,
      "num_tokens": 319230588.0,
      "step": 10002
    },
    {
      "entropy": 0.4935324275866151,
      "epoch": 0.9203397335265467,
      "grad_norm": 0.14617522060871124,
      "learning_rate": 6.932560493145644e-05,
      "loss": 0.466,
      "mean_token_accuracy": 0.8562178798019886,
      "num_tokens": 319263190.0,
      "step": 10003
    },
    {
      "entropy": 0.8986342828720808,
      "epoch": 0.920431739897988,
      "grad_norm": 0.1668703407049179,
      "learning_rate": 6.932253810531482e-05,
      "loss": 0.9063,
      "mean_token_accuracy": 0.7376941479742527,
      "num_tokens": 319293931.0,
      "step": 10004
    },
    {
      "entropy": 0.726238583214581,
      "epoch": 0.9205237462694291,
      "grad_norm": 0.15867480635643005,
      "learning_rate": 6.931947127917319e-05,
      "loss": 0.7188,
      "mean_token_accuracy": 0.7832648456096649,
      "num_tokens": 319325237.0,
      "step": 10005
    },
    {
      "entropy": 0.7880181837826967,
      "epoch": 0.9206157526408704,
      "grad_norm": 0.15475909411907196,
      "learning_rate": 6.931640445303156e-05,
      "loss": 0.7896,
      "mean_token_accuracy": 0.7691408470273018,
      "num_tokens": 319357390.0,
      "step": 10006
    },
    {
      "entropy": 0.6472890563309193,
      "epoch": 0.9207077590123116,
      "grad_norm": 0.17355003952980042,
      "learning_rate": 6.931333762688994e-05,
      "loss": 0.6537,
      "mean_token_accuracy": 0.8059832043945789,
      "num_tokens": 319389233.0,
      "step": 10007
    },
    {
      "entropy": 0.6652557821944356,
      "epoch": 0.9207997653837529,
      "grad_norm": 0.15902040898799896,
      "learning_rate": 6.93102708007483e-05,
      "loss": 0.6542,
      "mean_token_accuracy": 0.8054920472204685,
      "num_tokens": 319421180.0,
      "step": 10008
    },
    {
      "entropy": 0.8067612629383802,
      "epoch": 0.9208917717551941,
      "grad_norm": 0.15871120989322662,
      "learning_rate": 6.930720397460669e-05,
      "loss": 0.7976,
      "mean_token_accuracy": 0.769433569163084,
      "num_tokens": 319452721.0,
      "step": 10009
    },
    {
      "entropy": 0.686876630410552,
      "epoch": 0.9209837781266352,
      "grad_norm": 0.15559357404708862,
      "learning_rate": 6.930413714846506e-05,
      "loss": 0.6894,
      "mean_token_accuracy": 0.7939366772770882,
      "num_tokens": 319485171.0,
      "step": 10010
    },
    {
      "entropy": 0.717473616823554,
      "epoch": 0.9210757844980765,
      "grad_norm": 0.1565064936876297,
      "learning_rate": 6.930107032232344e-05,
      "loss": 0.706,
      "mean_token_accuracy": 0.7886949703097343,
      "num_tokens": 319517831.0,
      "step": 10011
    },
    {
      "entropy": 0.7296528425067663,
      "epoch": 0.9211677908695177,
      "grad_norm": 0.1593588888645172,
      "learning_rate": 6.92980034961818e-05,
      "loss": 0.733,
      "mean_token_accuracy": 0.7807900942862034,
      "num_tokens": 319549961.0,
      "step": 10012
    },
    {
      "entropy": 0.6601831577718258,
      "epoch": 0.921259797240959,
      "grad_norm": 0.15349198877811432,
      "learning_rate": 6.929493667004017e-05,
      "loss": 0.6455,
      "mean_token_accuracy": 0.806370671838522,
      "num_tokens": 319582374.0,
      "step": 10013
    },
    {
      "entropy": 0.702958638779819,
      "epoch": 0.9213518036124002,
      "grad_norm": 0.16017575562000275,
      "learning_rate": 6.929186984389855e-05,
      "loss": 0.6699,
      "mean_token_accuracy": 0.7966594696044922,
      "num_tokens": 319614102.0,
      "step": 10014
    },
    {
      "entropy": 0.8034601304680109,
      "epoch": 0.9214438099838413,
      "grad_norm": 0.17059485614299774,
      "learning_rate": 6.928880301775694e-05,
      "loss": 0.8025,
      "mean_token_accuracy": 0.767827495932579,
      "num_tokens": 319646415.0,
      "step": 10015
    },
    {
      "entropy": 0.7202405901625752,
      "epoch": 0.9215358163552826,
      "grad_norm": 0.15718232095241547,
      "learning_rate": 6.92857361916153e-05,
      "loss": 0.7104,
      "mean_token_accuracy": 0.7853466831147671,
      "num_tokens": 319678372.0,
      "step": 10016
    },
    {
      "entropy": 0.6869063312187791,
      "epoch": 0.9216278227267238,
      "grad_norm": 0.1456579715013504,
      "learning_rate": 6.928266936547367e-05,
      "loss": 0.6666,
      "mean_token_accuracy": 0.796948991715908,
      "num_tokens": 319709651.0,
      "step": 10017
    },
    {
      "entropy": 0.5935928355902433,
      "epoch": 0.9217198290981651,
      "grad_norm": 0.14472149312496185,
      "learning_rate": 6.927960253933204e-05,
      "loss": 0.5947,
      "mean_token_accuracy": 0.8241494409739971,
      "num_tokens": 319741248.0,
      "step": 10018
    },
    {
      "entropy": 0.5481100119650364,
      "epoch": 0.9218118354696063,
      "grad_norm": 0.14352266490459442,
      "learning_rate": 6.927653571319042e-05,
      "loss": 0.5254,
      "mean_token_accuracy": 0.8393064960837364,
      "num_tokens": 319773413.0,
      "step": 10019
    },
    {
      "entropy": 0.7654887363314629,
      "epoch": 0.9219038418410475,
      "grad_norm": 0.15309712290763855,
      "learning_rate": 6.92734688870488e-05,
      "loss": 0.7453,
      "mean_token_accuracy": 0.7766719982028008,
      "num_tokens": 319805879.0,
      "step": 10020
    },
    {
      "entropy": 0.6656090337783098,
      "epoch": 0.9219958482124887,
      "grad_norm": 0.15048709511756897,
      "learning_rate": 6.927040206090717e-05,
      "loss": 0.657,
      "mean_token_accuracy": 0.8118128255009651,
      "num_tokens": 319837603.0,
      "step": 10021
    },
    {
      "entropy": 0.7887123208492994,
      "epoch": 0.92208785458393,
      "grad_norm": 0.15650716423988342,
      "learning_rate": 6.926733523476554e-05,
      "loss": 0.7843,
      "mean_token_accuracy": 0.7723917625844479,
      "num_tokens": 319869083.0,
      "step": 10022
    },
    {
      "entropy": 0.7293807994574308,
      "epoch": 0.9221798609553712,
      "grad_norm": 0.16031105816364288,
      "learning_rate": 6.926426840862392e-05,
      "loss": 0.7085,
      "mean_token_accuracy": 0.7865784354507923,
      "num_tokens": 319900780.0,
      "step": 10023
    },
    {
      "entropy": 0.6331095211207867,
      "epoch": 0.9222718673268124,
      "grad_norm": 0.14732055366039276,
      "learning_rate": 6.926120158248229e-05,
      "loss": 0.6129,
      "mean_token_accuracy": 0.8129370175302029,
      "num_tokens": 319932763.0,
      "step": 10024
    },
    {
      "entropy": 0.7142865313217044,
      "epoch": 0.9223638736982536,
      "grad_norm": 0.16352733969688416,
      "learning_rate": 6.925813475634067e-05,
      "loss": 0.7038,
      "mean_token_accuracy": 0.7933197468519211,
      "num_tokens": 319964355.0,
      "step": 10025
    },
    {
      "entropy": 0.6586469803005457,
      "epoch": 0.9224558800696948,
      "grad_norm": 0.1499454528093338,
      "learning_rate": 6.925506793019904e-05,
      "loss": 0.6473,
      "mean_token_accuracy": 0.8058638796210289,
      "num_tokens": 319996677.0,
      "step": 10026
    },
    {
      "entropy": 0.6870859954506159,
      "epoch": 0.9225478864411361,
      "grad_norm": 0.16267329454421997,
      "learning_rate": 6.925200110405742e-05,
      "loss": 0.6868,
      "mean_token_accuracy": 0.800089493393898,
      "num_tokens": 320028680.0,
      "step": 10027
    },
    {
      "entropy": 0.8489757478237152,
      "epoch": 0.9226398928125773,
      "grad_norm": 0.16786479949951172,
      "learning_rate": 6.924893427791579e-05,
      "loss": 0.8526,
      "mean_token_accuracy": 0.7477334588766098,
      "num_tokens": 320059721.0,
      "step": 10028
    },
    {
      "entropy": 0.8023544810712337,
      "epoch": 0.9227318991840185,
      "grad_norm": 0.1651490479707718,
      "learning_rate": 6.924586745177415e-05,
      "loss": 0.7845,
      "mean_token_accuracy": 0.7682928405702114,
      "num_tokens": 320091155.0,
      "step": 10029
    },
    {
      "entropy": 0.7299495283514261,
      "epoch": 0.9228239055554597,
      "grad_norm": 0.16199935972690582,
      "learning_rate": 6.924280062563254e-05,
      "loss": 0.7192,
      "mean_token_accuracy": 0.7855132147669792,
      "num_tokens": 320123115.0,
      "step": 10030
    },
    {
      "entropy": 0.6984968576580286,
      "epoch": 0.9229159119269009,
      "grad_norm": 0.15615074336528778,
      "learning_rate": 6.923973379949092e-05,
      "loss": 0.678,
      "mean_token_accuracy": 0.7948505915701389,
      "num_tokens": 320155016.0,
      "step": 10031
    },
    {
      "entropy": 0.7193333669565618,
      "epoch": 0.9230079182983422,
      "grad_norm": 0.163724884390831,
      "learning_rate": 6.923666697334928e-05,
      "loss": 0.6986,
      "mean_token_accuracy": 0.7952575646340847,
      "num_tokens": 320186271.0,
      "step": 10032
    },
    {
      "entropy": 0.7473000008612871,
      "epoch": 0.9230999246697834,
      "grad_norm": 0.15048228204250336,
      "learning_rate": 6.923360014720765e-05,
      "loss": 0.7434,
      "mean_token_accuracy": 0.777312234044075,
      "num_tokens": 320218893.0,
      "step": 10033
    },
    {
      "entropy": 0.683626439422369,
      "epoch": 0.9231919310412247,
      "grad_norm": 0.16222909092903137,
      "learning_rate": 6.923053332106602e-05,
      "loss": 0.6761,
      "mean_token_accuracy": 0.8029282540082932,
      "num_tokens": 320250361.0,
      "step": 10034
    },
    {
      "entropy": 0.7030383627861738,
      "epoch": 0.9232839374126658,
      "grad_norm": 0.15259341895580292,
      "learning_rate": 6.922746649492442e-05,
      "loss": 0.6907,
      "mean_token_accuracy": 0.7928045578300953,
      "num_tokens": 320282752.0,
      "step": 10035
    },
    {
      "entropy": 0.6658203289844096,
      "epoch": 0.923375943784107,
      "grad_norm": 0.14650294184684753,
      "learning_rate": 6.922439966878278e-05,
      "loss": 0.6503,
      "mean_token_accuracy": 0.8053961433470249,
      "num_tokens": 320315192.0,
      "step": 10036
    },
    {
      "entropy": 0.6745353452861309,
      "epoch": 0.9234679501555483,
      "grad_norm": 0.14765432476997375,
      "learning_rate": 6.922133284264115e-05,
      "loss": 0.6809,
      "mean_token_accuracy": 0.7994432747364044,
      "num_tokens": 320347401.0,
      "step": 10037
    },
    {
      "entropy": 0.7337830904871225,
      "epoch": 0.9235599565269895,
      "grad_norm": 0.1632947474718094,
      "learning_rate": 6.921826601649953e-05,
      "loss": 0.7259,
      "mean_token_accuracy": 0.7865215912461281,
      "num_tokens": 320380000.0,
      "step": 10038
    },
    {
      "entropy": 0.5459805894643068,
      "epoch": 0.9236519628984308,
      "grad_norm": 0.14175595343112946,
      "learning_rate": 6.92151991903579e-05,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8329222090542316,
      "num_tokens": 320412171.0,
      "step": 10039
    },
    {
      "entropy": 0.6332358927465975,
      "epoch": 0.9237439692698719,
      "grad_norm": 0.15180563926696777,
      "learning_rate": 6.921213236421628e-05,
      "loss": 0.6158,
      "mean_token_accuracy": 0.8139631822705269,
      "num_tokens": 320443525.0,
      "step": 10040
    },
    {
      "entropy": 0.776739465072751,
      "epoch": 0.9238359756413131,
      "grad_norm": 0.16606175899505615,
      "learning_rate": 6.920906553807465e-05,
      "loss": 0.7694,
      "mean_token_accuracy": 0.7764557749032974,
      "num_tokens": 320476127.0,
      "step": 10041
    },
    {
      "entropy": 0.792960787191987,
      "epoch": 0.9239279820127544,
      "grad_norm": 0.16206376254558563,
      "learning_rate": 6.920599871193303e-05,
      "loss": 0.8016,
      "mean_token_accuracy": 0.7664768546819687,
      "num_tokens": 320508380.0,
      "step": 10042
    },
    {
      "entropy": 0.8171132728457451,
      "epoch": 0.9240199883841956,
      "grad_norm": 0.16963684558868408,
      "learning_rate": 6.92029318857914e-05,
      "loss": 0.8092,
      "mean_token_accuracy": 0.7664848938584328,
      "num_tokens": 320540349.0,
      "step": 10043
    },
    {
      "entropy": 0.7006282955408096,
      "epoch": 0.9241119947556369,
      "grad_norm": 0.15460680425167084,
      "learning_rate": 6.919986505964977e-05,
      "loss": 0.6817,
      "mean_token_accuracy": 0.7925930991768837,
      "num_tokens": 320571745.0,
      "step": 10044
    },
    {
      "entropy": 0.7206188198179007,
      "epoch": 0.924204001127078,
      "grad_norm": 0.16092151403427124,
      "learning_rate": 6.919679823350815e-05,
      "loss": 0.7058,
      "mean_token_accuracy": 0.791772086173296,
      "num_tokens": 320603961.0,
      "step": 10045
    },
    {
      "entropy": 0.8085630759596825,
      "epoch": 0.9242960074985193,
      "grad_norm": 0.16168299317359924,
      "learning_rate": 6.919373140736653e-05,
      "loss": 0.7921,
      "mean_token_accuracy": 0.7689031213521957,
      "num_tokens": 320635564.0,
      "step": 10046
    },
    {
      "entropy": 0.5516939535737038,
      "epoch": 0.9243880138699605,
      "grad_norm": 0.1392057240009308,
      "learning_rate": 6.91906645812249e-05,
      "loss": 0.5169,
      "mean_token_accuracy": 0.8427882194519043,
      "num_tokens": 320668332.0,
      "step": 10047
    },
    {
      "entropy": 0.7709966078400612,
      "epoch": 0.9244800202414017,
      "grad_norm": 0.1561095416545868,
      "learning_rate": 6.918759775508327e-05,
      "loss": 0.7561,
      "mean_token_accuracy": 0.7786037921905518,
      "num_tokens": 320700353.0,
      "step": 10048
    },
    {
      "entropy": 0.7292346861213446,
      "epoch": 0.924572026612843,
      "grad_norm": 0.15002547204494476,
      "learning_rate": 6.918453092894163e-05,
      "loss": 0.7192,
      "mean_token_accuracy": 0.7879794277250767,
      "num_tokens": 320732605.0,
      "step": 10049
    },
    {
      "entropy": 0.6692859251052141,
      "epoch": 0.9246640329842841,
      "grad_norm": 0.15094876289367676,
      "learning_rate": 6.918146410280001e-05,
      "loss": 0.6628,
      "mean_token_accuracy": 0.8035629540681839,
      "num_tokens": 320764760.0,
      "step": 10050
    },
    {
      "entropy": 0.6223231581971049,
      "epoch": 0.9247560393557254,
      "grad_norm": 0.15153905749320984,
      "learning_rate": 6.91783972766584e-05,
      "loss": 0.6004,
      "mean_token_accuracy": 0.8175779841840267,
      "num_tokens": 320796949.0,
      "step": 10051
    },
    {
      "entropy": 0.5421300865709782,
      "epoch": 0.9248480457271666,
      "grad_norm": 0.15300601720809937,
      "learning_rate": 6.917533045051676e-05,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8371445313096046,
      "num_tokens": 320829601.0,
      "step": 10052
    },
    {
      "entropy": 0.7616028375923634,
      "epoch": 0.9249400520986079,
      "grad_norm": 0.16030330955982208,
      "learning_rate": 6.917226362437513e-05,
      "loss": 0.7567,
      "mean_token_accuracy": 0.7761100940406322,
      "num_tokens": 320860946.0,
      "step": 10053
    },
    {
      "entropy": 0.7592726144939661,
      "epoch": 0.9250320584700491,
      "grad_norm": 0.15871545672416687,
      "learning_rate": 6.916919679823351e-05,
      "loss": 0.7458,
      "mean_token_accuracy": 0.7781198844313622,
      "num_tokens": 320892670.0,
      "step": 10054
    },
    {
      "entropy": 0.6296313665807247,
      "epoch": 0.9251240648414902,
      "grad_norm": 0.1460118442773819,
      "learning_rate": 6.916612997209188e-05,
      "loss": 0.6213,
      "mean_token_accuracy": 0.8208964318037033,
      "num_tokens": 320925438.0,
      "step": 10055
    },
    {
      "entropy": 0.636111730709672,
      "epoch": 0.9252160712129315,
      "grad_norm": 0.15088102221488953,
      "learning_rate": 6.916306314595026e-05,
      "loss": 0.6288,
      "mean_token_accuracy": 0.8146796897053719,
      "num_tokens": 320957486.0,
      "step": 10056
    },
    {
      "entropy": 0.6198417283594608,
      "epoch": 0.9253080775843727,
      "grad_norm": 0.16705523431301117,
      "learning_rate": 6.915999631980863e-05,
      "loss": 0.6201,
      "mean_token_accuracy": 0.8150073625147343,
      "num_tokens": 320989902.0,
      "step": 10057
    },
    {
      "entropy": 0.7687426842749119,
      "epoch": 0.925400083955814,
      "grad_norm": 0.16118517518043518,
      "learning_rate": 6.915692949366701e-05,
      "loss": 0.7681,
      "mean_token_accuracy": 0.7754175923764706,
      "num_tokens": 321021899.0,
      "step": 10058
    },
    {
      "entropy": 0.7426656968891621,
      "epoch": 0.9254920903272552,
      "grad_norm": 0.16048260033130646,
      "learning_rate": 6.915386266752538e-05,
      "loss": 0.7363,
      "mean_token_accuracy": 0.7848457545042038,
      "num_tokens": 321054153.0,
      "step": 10059
    },
    {
      "entropy": 0.6714558964595199,
      "epoch": 0.9255840966986963,
      "grad_norm": 0.16212117671966553,
      "learning_rate": 6.915079584138375e-05,
      "loss": 0.669,
      "mean_token_accuracy": 0.8003207370638847,
      "num_tokens": 321085844.0,
      "step": 10060
    },
    {
      "entropy": 0.7232849858701229,
      "epoch": 0.9256761030701376,
      "grad_norm": 0.16266725957393646,
      "learning_rate": 6.914772901524213e-05,
      "loss": 0.7197,
      "mean_token_accuracy": 0.7878721281886101,
      "num_tokens": 321117393.0,
      "step": 10061
    },
    {
      "entropy": 0.7654467169195414,
      "epoch": 0.9257681094415788,
      "grad_norm": 0.16172921657562256,
      "learning_rate": 6.914466218910051e-05,
      "loss": 0.7577,
      "mean_token_accuracy": 0.7756998836994171,
      "num_tokens": 321149551.0,
      "step": 10062
    },
    {
      "entropy": 0.7012363411486149,
      "epoch": 0.9258601158130201,
      "grad_norm": 0.15066835284233093,
      "learning_rate": 6.914159536295888e-05,
      "loss": 0.6926,
      "mean_token_accuracy": 0.7967973276972771,
      "num_tokens": 321181428.0,
      "step": 10063
    },
    {
      "entropy": 0.7100589666515589,
      "epoch": 0.9259521221844613,
      "grad_norm": 0.15040528774261475,
      "learning_rate": 6.913852853681725e-05,
      "loss": 0.6978,
      "mean_token_accuracy": 0.7941379621624947,
      "num_tokens": 321214096.0,
      "step": 10064
    },
    {
      "entropy": 0.7416942454874516,
      "epoch": 0.9260441285559025,
      "grad_norm": 0.1543944925069809,
      "learning_rate": 6.913546171067561e-05,
      "loss": 0.7258,
      "mean_token_accuracy": 0.7829320654273033,
      "num_tokens": 321245992.0,
      "step": 10065
    },
    {
      "entropy": 0.6624593399465084,
      "epoch": 0.9261361349273437,
      "grad_norm": 0.1581188589334488,
      "learning_rate": 6.9132394884534e-05,
      "loss": 0.6601,
      "mean_token_accuracy": 0.800947543233633,
      "num_tokens": 321278191.0,
      "step": 10066
    },
    {
      "entropy": 0.6524279695004225,
      "epoch": 0.926228141298785,
      "grad_norm": 0.15083719789981842,
      "learning_rate": 6.912932805839238e-05,
      "loss": 0.6444,
      "mean_token_accuracy": 0.8062034510076046,
      "num_tokens": 321310061.0,
      "step": 10067
    },
    {
      "entropy": 0.8177984319627285,
      "epoch": 0.9263201476702262,
      "grad_norm": 0.16416975855827332,
      "learning_rate": 6.912626123225074e-05,
      "loss": 0.81,
      "mean_token_accuracy": 0.7623271904885769,
      "num_tokens": 321342348.0,
      "step": 10068
    },
    {
      "entropy": 0.7641009818762541,
      "epoch": 0.9264121540416674,
      "grad_norm": 0.16724248230457306,
      "learning_rate": 6.912319440610913e-05,
      "loss": 0.7443,
      "mean_token_accuracy": 0.7768818587064743,
      "num_tokens": 321374000.0,
      "step": 10069
    },
    {
      "entropy": 0.8039760161191225,
      "epoch": 0.9265041604131086,
      "grad_norm": 0.1725081205368042,
      "learning_rate": 6.91201275799675e-05,
      "loss": 0.8071,
      "mean_token_accuracy": 0.7714586034417152,
      "num_tokens": 321405629.0,
      "step": 10070
    },
    {
      "entropy": 0.7091279597952962,
      "epoch": 0.9265961667845498,
      "grad_norm": 0.1555512696504593,
      "learning_rate": 6.911706075382586e-05,
      "loss": 0.7149,
      "mean_token_accuracy": 0.7893836311995983,
      "num_tokens": 321437119.0,
      "step": 10071
    },
    {
      "entropy": 0.6160593489184976,
      "epoch": 0.9266881731559911,
      "grad_norm": 0.1687343418598175,
      "learning_rate": 6.911399392768424e-05,
      "loss": 0.6024,
      "mean_token_accuracy": 0.82093221321702,
      "num_tokens": 321468748.0,
      "step": 10072
    },
    {
      "entropy": 0.7596363481134176,
      "epoch": 0.9267801795274323,
      "grad_norm": 0.1489769071340561,
      "learning_rate": 6.911092710154262e-05,
      "loss": 0.7353,
      "mean_token_accuracy": 0.7873107083141804,
      "num_tokens": 321501516.0,
      "step": 10073
    },
    {
      "entropy": 0.7282904125750065,
      "epoch": 0.9268721858988735,
      "grad_norm": 0.1611119508743286,
      "learning_rate": 6.910786027540099e-05,
      "loss": 0.709,
      "mean_token_accuracy": 0.790121428668499,
      "num_tokens": 321533588.0,
      "step": 10074
    },
    {
      "entropy": 0.7752249278128147,
      "epoch": 0.9269641922703147,
      "grad_norm": 0.16347157955169678,
      "learning_rate": 6.910479344925936e-05,
      "loss": 0.7557,
      "mean_token_accuracy": 0.7764816731214523,
      "num_tokens": 321565391.0,
      "step": 10075
    },
    {
      "entropy": 0.6411216538399458,
      "epoch": 0.9270561986417559,
      "grad_norm": 0.14603546261787415,
      "learning_rate": 6.910172662311774e-05,
      "loss": 0.6175,
      "mean_token_accuracy": 0.8123501688241959,
      "num_tokens": 321597450.0,
      "step": 10076
    },
    {
      "entropy": 0.6722378823906183,
      "epoch": 0.9271482050131972,
      "grad_norm": 0.15343627333641052,
      "learning_rate": 6.909865979697612e-05,
      "loss": 0.6387,
      "mean_token_accuracy": 0.8089510388672352,
      "num_tokens": 321629569.0,
      "step": 10077
    },
    {
      "entropy": 0.6157147493213415,
      "epoch": 0.9272402113846384,
      "grad_norm": 0.15549327433109283,
      "learning_rate": 6.909559297083449e-05,
      "loss": 0.614,
      "mean_token_accuracy": 0.8133190684020519,
      "num_tokens": 321661007.0,
      "step": 10078
    },
    {
      "entropy": 0.6817674655467272,
      "epoch": 0.9273322177560797,
      "grad_norm": 0.15572255849838257,
      "learning_rate": 6.909252614469286e-05,
      "loss": 0.6621,
      "mean_token_accuracy": 0.8019202537834644,
      "num_tokens": 321693308.0,
      "step": 10079
    },
    {
      "entropy": 0.5709645822644234,
      "epoch": 0.9274242241275208,
      "grad_norm": 0.15026229619979858,
      "learning_rate": 6.908945931855123e-05,
      "loss": 0.5674,
      "mean_token_accuracy": 0.8296796195209026,
      "num_tokens": 321725509.0,
      "step": 10080
    },
    {
      "entropy": 0.6937322095036507,
      "epoch": 0.927516230498962,
      "grad_norm": 0.1592530608177185,
      "learning_rate": 6.908639249240961e-05,
      "loss": 0.6775,
      "mean_token_accuracy": 0.7991457246243954,
      "num_tokens": 321756088.0,
      "step": 10081
    },
    {
      "entropy": 0.8693338520824909,
      "epoch": 0.9276082368704033,
      "grad_norm": 0.17022302746772766,
      "learning_rate": 6.908332566626799e-05,
      "loss": 0.8982,
      "mean_token_accuracy": 0.7372478023171425,
      "num_tokens": 321787396.0,
      "step": 10082
    },
    {
      "entropy": 0.5100908642634749,
      "epoch": 0.9277002432418445,
      "grad_norm": 0.14729033410549164,
      "learning_rate": 6.908025884012636e-05,
      "loss": 0.49,
      "mean_token_accuracy": 0.8506980910897255,
      "num_tokens": 321819915.0,
      "step": 10083
    },
    {
      "entropy": 0.7158867958933115,
      "epoch": 0.9277922496132858,
      "grad_norm": 0.16711753606796265,
      "learning_rate": 6.907719201398473e-05,
      "loss": 0.7103,
      "mean_token_accuracy": 0.7906990014016628,
      "num_tokens": 321851682.0,
      "step": 10084
    },
    {
      "entropy": 0.759673647582531,
      "epoch": 0.9278842559847269,
      "grad_norm": 0.15763312578201294,
      "learning_rate": 6.907412518784311e-05,
      "loss": 0.7668,
      "mean_token_accuracy": 0.7754876054823399,
      "num_tokens": 321883458.0,
      "step": 10085
    },
    {
      "entropy": 0.5400787517428398,
      "epoch": 0.9279762623561681,
      "grad_norm": 0.14842349290847778,
      "learning_rate": 6.907105836170147e-05,
      "loss": 0.525,
      "mean_token_accuracy": 0.8430475071072578,
      "num_tokens": 321915650.0,
      "step": 10086
    },
    {
      "entropy": 0.682932311668992,
      "epoch": 0.9280682687276094,
      "grad_norm": 0.15183192491531372,
      "learning_rate": 6.906799153555986e-05,
      "loss": 0.6657,
      "mean_token_accuracy": 0.8029856085777283,
      "num_tokens": 321947450.0,
      "step": 10087
    },
    {
      "entropy": 0.7603417783975601,
      "epoch": 0.9281602750990506,
      "grad_norm": 0.15779916942119598,
      "learning_rate": 6.906492470941822e-05,
      "loss": 0.7518,
      "mean_token_accuracy": 0.777743998914957,
      "num_tokens": 321978921.0,
      "step": 10088
    },
    {
      "entropy": 0.7225582785904408,
      "epoch": 0.9282522814704919,
      "grad_norm": 0.1613718718290329,
      "learning_rate": 6.90618578832766e-05,
      "loss": 0.7098,
      "mean_token_accuracy": 0.7906680107116699,
      "num_tokens": 322011382.0,
      "step": 10089
    },
    {
      "entropy": 0.6892097359523177,
      "epoch": 0.928344287841933,
      "grad_norm": 0.15986180305480957,
      "learning_rate": 6.905879105713497e-05,
      "loss": 0.6745,
      "mean_token_accuracy": 0.8047891855239868,
      "num_tokens": 322042877.0,
      "step": 10090
    },
    {
      "entropy": 0.7807013662531972,
      "epoch": 0.9284362942133743,
      "grad_norm": 0.17247657477855682,
      "learning_rate": 6.905572423099334e-05,
      "loss": 0.7465,
      "mean_token_accuracy": 0.7813421562314034,
      "num_tokens": 322074667.0,
      "step": 10091
    },
    {
      "entropy": 0.7125077228993177,
      "epoch": 0.9285283005848155,
      "grad_norm": 0.15110737085342407,
      "learning_rate": 6.905265740485172e-05,
      "loss": 0.6947,
      "mean_token_accuracy": 0.794600386172533,
      "num_tokens": 322106867.0,
      "step": 10092
    },
    {
      "entropy": 0.8483287543058395,
      "epoch": 0.9286203069562567,
      "grad_norm": 0.158844456076622,
      "learning_rate": 6.90495905787101e-05,
      "loss": 0.8439,
      "mean_token_accuracy": 0.7530940696597099,
      "num_tokens": 322139597.0,
      "step": 10093
    },
    {
      "entropy": 0.691204534843564,
      "epoch": 0.928712313327698,
      "grad_norm": 0.14293403923511505,
      "learning_rate": 6.904652375256847e-05,
      "loss": 0.6701,
      "mean_token_accuracy": 0.8008691109716892,
      "num_tokens": 322171592.0,
      "step": 10094
    },
    {
      "entropy": 0.7054439028725028,
      "epoch": 0.9288043196991391,
      "grad_norm": 0.15565888583660126,
      "learning_rate": 6.904345692642684e-05,
      "loss": 0.6871,
      "mean_token_accuracy": 0.7925667501986027,
      "num_tokens": 322203525.0,
      "step": 10095
    },
    {
      "entropy": 0.8024335382506251,
      "epoch": 0.9288963260705804,
      "grad_norm": 0.17689859867095947,
      "learning_rate": 6.904039010028521e-05,
      "loss": 0.7982,
      "mean_token_accuracy": 0.7648606970906258,
      "num_tokens": 322235056.0,
      "step": 10096
    },
    {
      "entropy": 0.7258170172572136,
      "epoch": 0.9289883324420216,
      "grad_norm": 0.15618418157100677,
      "learning_rate": 6.903732327414359e-05,
      "loss": 0.7267,
      "mean_token_accuracy": 0.7877820767462254,
      "num_tokens": 322267715.0,
      "step": 10097
    },
    {
      "entropy": 0.6603909423574805,
      "epoch": 0.9290803388134629,
      "grad_norm": 0.15847380459308624,
      "learning_rate": 6.903425644800197e-05,
      "loss": 0.6474,
      "mean_token_accuracy": 0.8042293973267078,
      "num_tokens": 322300170.0,
      "step": 10098
    },
    {
      "entropy": 0.7146832384169102,
      "epoch": 0.9291723451849041,
      "grad_norm": 0.1563357412815094,
      "learning_rate": 6.903118962186034e-05,
      "loss": 0.7051,
      "mean_token_accuracy": 0.7918129824101925,
      "num_tokens": 322332190.0,
      "step": 10099
    },
    {
      "entropy": 0.6367739280685782,
      "epoch": 0.9292643515563452,
      "grad_norm": 0.15132346749305725,
      "learning_rate": 6.902812279571872e-05,
      "loss": 0.6183,
      "mean_token_accuracy": 0.8168084993958473,
      "num_tokens": 322364699.0,
      "step": 10100
    },
    {
      "entropy": 0.7458810023963451,
      "epoch": 0.9293563579277865,
      "grad_norm": 0.14752796292304993,
      "learning_rate": 6.902505596957709e-05,
      "loss": 0.7355,
      "mean_token_accuracy": 0.7832323051989079,
      "num_tokens": 322396003.0,
      "step": 10101
    },
    {
      "entropy": 0.7932907827198505,
      "epoch": 0.9294483642992277,
      "grad_norm": 0.159769669175148,
      "learning_rate": 6.902198914343546e-05,
      "loss": 0.7916,
      "mean_token_accuracy": 0.7681764774024487,
      "num_tokens": 322427616.0,
      "step": 10102
    },
    {
      "entropy": 0.7277450393885374,
      "epoch": 0.929540370670669,
      "grad_norm": 0.15636759996414185,
      "learning_rate": 6.901892231729384e-05,
      "loss": 0.7224,
      "mean_token_accuracy": 0.7815062776207924,
      "num_tokens": 322459277.0,
      "step": 10103
    },
    {
      "entropy": 0.7303236108273268,
      "epoch": 0.9296323770421102,
      "grad_norm": 0.14794310927391052,
      "learning_rate": 6.901585549115222e-05,
      "loss": 0.7227,
      "mean_token_accuracy": 0.782582875341177,
      "num_tokens": 322490545.0,
      "step": 10104
    },
    {
      "entropy": 0.6138364970684052,
      "epoch": 0.9297243834135513,
      "grad_norm": 0.15385182201862335,
      "learning_rate": 6.901278866501059e-05,
      "loss": 0.5986,
      "mean_token_accuracy": 0.8169780932366848,
      "num_tokens": 322521946.0,
      "step": 10105
    },
    {
      "entropy": 0.7211675699800253,
      "epoch": 0.9298163897849926,
      "grad_norm": 0.14555470645427704,
      "learning_rate": 6.900972183886895e-05,
      "loss": 0.6931,
      "mean_token_accuracy": 0.795242715626955,
      "num_tokens": 322554713.0,
      "step": 10106
    },
    {
      "entropy": 0.8182109426707029,
      "epoch": 0.9299083961564338,
      "grad_norm": 0.15608720481395721,
      "learning_rate": 6.900665501272732e-05,
      "loss": 0.8163,
      "mean_token_accuracy": 0.763112235814333,
      "num_tokens": 322585861.0,
      "step": 10107
    },
    {
      "entropy": 0.7560088392347097,
      "epoch": 0.9300004025278751,
      "grad_norm": 0.16317932307720184,
      "learning_rate": 6.900358818658572e-05,
      "loss": 0.7444,
      "mean_token_accuracy": 0.7805887050926685,
      "num_tokens": 322617983.0,
      "step": 10108
    },
    {
      "entropy": 0.7740400936454535,
      "epoch": 0.9300924088993163,
      "grad_norm": 0.1623762547969818,
      "learning_rate": 6.900052136044408e-05,
      "loss": 0.7363,
      "mean_token_accuracy": 0.7853161320090294,
      "num_tokens": 322648925.0,
      "step": 10109
    },
    {
      "entropy": 0.765366405248642,
      "epoch": 0.9301844152707575,
      "grad_norm": 0.15867240726947784,
      "learning_rate": 6.899745453430245e-05,
      "loss": 0.7486,
      "mean_token_accuracy": 0.7841012179851532,
      "num_tokens": 322681095.0,
      "step": 10110
    },
    {
      "entropy": 0.7334907166659832,
      "epoch": 0.9302764216421987,
      "grad_norm": 0.15616540610790253,
      "learning_rate": 6.899438770816082e-05,
      "loss": 0.7262,
      "mean_token_accuracy": 0.7884567119181156,
      "num_tokens": 322713377.0,
      "step": 10111
    },
    {
      "entropy": 0.6746008209884167,
      "epoch": 0.93036842801364,
      "grad_norm": 0.16372458636760712,
      "learning_rate": 6.89913208820192e-05,
      "loss": 0.6816,
      "mean_token_accuracy": 0.8004342392086983,
      "num_tokens": 322745854.0,
      "step": 10112
    },
    {
      "entropy": 0.7266215058043599,
      "epoch": 0.9304604343850812,
      "grad_norm": 0.15073148906230927,
      "learning_rate": 6.898825405587758e-05,
      "loss": 0.7193,
      "mean_token_accuracy": 0.791246373206377,
      "num_tokens": 322777886.0,
      "step": 10113
    },
    {
      "entropy": 0.7629630714654922,
      "epoch": 0.9305524407565224,
      "grad_norm": 0.1556529402732849,
      "learning_rate": 6.898518722973595e-05,
      "loss": 0.7397,
      "mean_token_accuracy": 0.7767633721232414,
      "num_tokens": 322809621.0,
      "step": 10114
    },
    {
      "entropy": 0.7241054177284241,
      "epoch": 0.9306444471279636,
      "grad_norm": 0.15834400057792664,
      "learning_rate": 6.898212040359432e-05,
      "loss": 0.7215,
      "mean_token_accuracy": 0.7874822579324245,
      "num_tokens": 322841756.0,
      "step": 10115
    },
    {
      "entropy": 0.7104441709816456,
      "epoch": 0.9307364534994048,
      "grad_norm": 0.15885885059833527,
      "learning_rate": 6.89790535774527e-05,
      "loss": 0.7204,
      "mean_token_accuracy": 0.7840985618531704,
      "num_tokens": 322874200.0,
      "step": 10116
    },
    {
      "entropy": 0.7823220239952207,
      "epoch": 0.9308284598708461,
      "grad_norm": 0.15713804960250854,
      "learning_rate": 6.897598675131107e-05,
      "loss": 0.7845,
      "mean_token_accuracy": 0.7690116576850414,
      "num_tokens": 322905858.0,
      "step": 10117
    },
    {
      "entropy": 0.644562229514122,
      "epoch": 0.9309204662422873,
      "grad_norm": 0.14216068387031555,
      "learning_rate": 6.897291992516945e-05,
      "loss": 0.6314,
      "mean_token_accuracy": 0.8086646683514118,
      "num_tokens": 322938430.0,
      "step": 10118
    },
    {
      "entropy": 0.6619132664054632,
      "epoch": 0.9310124726137285,
      "grad_norm": 0.14979533851146698,
      "learning_rate": 6.896985309902782e-05,
      "loss": 0.6519,
      "mean_token_accuracy": 0.8048888295888901,
      "num_tokens": 322970487.0,
      "step": 10119
    },
    {
      "entropy": 0.7058610543608665,
      "epoch": 0.9311044789851697,
      "grad_norm": 0.14937850832939148,
      "learning_rate": 6.89667862728862e-05,
      "loss": 0.6877,
      "mean_token_accuracy": 0.7940788231790066,
      "num_tokens": 323002026.0,
      "step": 10120
    },
    {
      "entropy": 0.6250061132013798,
      "epoch": 0.9311964853566109,
      "grad_norm": 0.1475614458322525,
      "learning_rate": 6.896371944674457e-05,
      "loss": 0.6026,
      "mean_token_accuracy": 0.816873911768198,
      "num_tokens": 323034226.0,
      "step": 10121
    },
    {
      "entropy": 0.8196586258709431,
      "epoch": 0.9312884917280522,
      "grad_norm": 0.15701286494731903,
      "learning_rate": 6.896065262060293e-05,
      "loss": 0.8121,
      "mean_token_accuracy": 0.762927919626236,
      "num_tokens": 323065884.0,
      "step": 10122
    },
    {
      "entropy": 0.6183875780552626,
      "epoch": 0.9313804980994934,
      "grad_norm": 0.1496785283088684,
      "learning_rate": 6.895758579446132e-05,
      "loss": 0.601,
      "mean_token_accuracy": 0.8224759958684444,
      "num_tokens": 323098275.0,
      "step": 10123
    },
    {
      "entropy": 0.5530614103190601,
      "epoch": 0.9314725044709347,
      "grad_norm": 0.15011315047740936,
      "learning_rate": 6.89545189683197e-05,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8397418446838856,
      "num_tokens": 323130399.0,
      "step": 10124
    },
    {
      "entropy": 0.8053175006061792,
      "epoch": 0.9315645108423758,
      "grad_norm": 0.15631148219108582,
      "learning_rate": 6.895145214217807e-05,
      "loss": 0.8189,
      "mean_token_accuracy": 0.7570472210645676,
      "num_tokens": 323162873.0,
      "step": 10125
    },
    {
      "entropy": 0.7704017050564289,
      "epoch": 0.931656517213817,
      "grad_norm": 0.16695156693458557,
      "learning_rate": 6.894838531603643e-05,
      "loss": 0.7839,
      "mean_token_accuracy": 0.7731046676635742,
      "num_tokens": 323195271.0,
      "step": 10126
    },
    {
      "entropy": 0.6949842497706413,
      "epoch": 0.9317485235852583,
      "grad_norm": 0.1598910093307495,
      "learning_rate": 6.89453184898948e-05,
      "loss": 0.711,
      "mean_token_accuracy": 0.7931217178702354,
      "num_tokens": 323227290.0,
      "step": 10127
    },
    {
      "entropy": 0.8004434136673808,
      "epoch": 0.9318405299566995,
      "grad_norm": 0.15888231992721558,
      "learning_rate": 6.894225166375318e-05,
      "loss": 0.8015,
      "mean_token_accuracy": 0.7681481316685677,
      "num_tokens": 323259320.0,
      "step": 10128
    },
    {
      "entropy": 0.7190520148724318,
      "epoch": 0.9319325363281408,
      "grad_norm": 0.1558990478515625,
      "learning_rate": 6.893918483761156e-05,
      "loss": 0.7105,
      "mean_token_accuracy": 0.7867806144058704,
      "num_tokens": 323290308.0,
      "step": 10129
    },
    {
      "entropy": 0.6921202074736357,
      "epoch": 0.9320245426995819,
      "grad_norm": 0.1484592854976654,
      "learning_rate": 6.893611801146993e-05,
      "loss": 0.6783,
      "mean_token_accuracy": 0.7970200255513191,
      "num_tokens": 323323076.0,
      "step": 10130
    },
    {
      "entropy": 0.6590983057394624,
      "epoch": 0.9321165490710231,
      "grad_norm": 0.150732159614563,
      "learning_rate": 6.893305118532831e-05,
      "loss": 0.64,
      "mean_token_accuracy": 0.8110101819038391,
      "num_tokens": 323355079.0,
      "step": 10131
    },
    {
      "entropy": 0.6195534877479076,
      "epoch": 0.9322085554424644,
      "grad_norm": 0.15210844576358795,
      "learning_rate": 6.892998435918668e-05,
      "loss": 0.5865,
      "mean_token_accuracy": 0.8238384313881397,
      "num_tokens": 323386867.0,
      "step": 10132
    },
    {
      "entropy": 0.776025390252471,
      "epoch": 0.9323005618139056,
      "grad_norm": 0.16005708277225494,
      "learning_rate": 6.892691753304505e-05,
      "loss": 0.7591,
      "mean_token_accuracy": 0.7756818532943726,
      "num_tokens": 323418719.0,
      "step": 10133
    },
    {
      "entropy": 0.8539554663002491,
      "epoch": 0.9323925681853469,
      "grad_norm": 0.1715088039636612,
      "learning_rate": 6.892385070690343e-05,
      "loss": 0.8241,
      "mean_token_accuracy": 0.759704414755106,
      "num_tokens": 323451450.0,
      "step": 10134
    },
    {
      "entropy": 0.7603825498372316,
      "epoch": 0.932484574556788,
      "grad_norm": 0.1562788486480713,
      "learning_rate": 6.892078388076181e-05,
      "loss": 0.7351,
      "mean_token_accuracy": 0.7794371359050274,
      "num_tokens": 323483358.0,
      "step": 10135
    },
    {
      "entropy": 0.8624521512538195,
      "epoch": 0.9325765809282293,
      "grad_norm": 0.1549885869026184,
      "learning_rate": 6.891771705462018e-05,
      "loss": 0.8645,
      "mean_token_accuracy": 0.7486150413751602,
      "num_tokens": 323514972.0,
      "step": 10136
    },
    {
      "entropy": 0.7578782495111227,
      "epoch": 0.9326685872996705,
      "grad_norm": 0.15582281351089478,
      "learning_rate": 6.891465022847855e-05,
      "loss": 0.7449,
      "mean_token_accuracy": 0.7813914306461811,
      "num_tokens": 323546457.0,
      "step": 10137
    },
    {
      "entropy": 0.7676836643368006,
      "epoch": 0.9327605936711117,
      "grad_norm": 0.16688348352909088,
      "learning_rate": 6.891158340233692e-05,
      "loss": 0.7515,
      "mean_token_accuracy": 0.7786552645266056,
      "num_tokens": 323577796.0,
      "step": 10138
    },
    {
      "entropy": 0.7014904525130987,
      "epoch": 0.932852600042553,
      "grad_norm": 0.15529468655586243,
      "learning_rate": 6.89085165761953e-05,
      "loss": 0.6961,
      "mean_token_accuracy": 0.7971718683838844,
      "num_tokens": 323610083.0,
      "step": 10139
    },
    {
      "entropy": 0.717237657867372,
      "epoch": 0.9329446064139941,
      "grad_norm": 0.15512606501579285,
      "learning_rate": 6.890544975005368e-05,
      "loss": 0.7001,
      "mean_token_accuracy": 0.7861653417348862,
      "num_tokens": 323641966.0,
      "step": 10140
    },
    {
      "entropy": 0.6810935847461224,
      "epoch": 0.9330366127854354,
      "grad_norm": 0.16409756243228912,
      "learning_rate": 6.890238292391205e-05,
      "loss": 0.6777,
      "mean_token_accuracy": 0.8009281866252422,
      "num_tokens": 323674734.0,
      "step": 10141
    },
    {
      "entropy": 0.6225458662956953,
      "epoch": 0.9331286191568766,
      "grad_norm": 0.14883823692798615,
      "learning_rate": 6.889931609777041e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.814807690680027,
      "num_tokens": 323707188.0,
      "step": 10142
    },
    {
      "entropy": 0.8596897907555103,
      "epoch": 0.9332206255283179,
      "grad_norm": 0.1700042188167572,
      "learning_rate": 6.88962492716288e-05,
      "loss": 0.8573,
      "mean_token_accuracy": 0.7492248527705669,
      "num_tokens": 323738780.0,
      "step": 10143
    },
    {
      "entropy": 0.7128691719844937,
      "epoch": 0.9333126318997591,
      "grad_norm": 0.1589837223291397,
      "learning_rate": 6.889318244548716e-05,
      "loss": 0.7114,
      "mean_token_accuracy": 0.788653701543808,
      "num_tokens": 323771125.0,
      "step": 10144
    },
    {
      "entropy": 0.7672869302332401,
      "epoch": 0.9334046382712002,
      "grad_norm": 0.15095055103302002,
      "learning_rate": 6.889011561934555e-05,
      "loss": 0.765,
      "mean_token_accuracy": 0.7725601233541965,
      "num_tokens": 323803270.0,
      "step": 10145
    },
    {
      "entropy": 0.7803751453757286,
      "epoch": 0.9334966446426415,
      "grad_norm": 0.16506415605545044,
      "learning_rate": 6.888704879320391e-05,
      "loss": 0.7867,
      "mean_token_accuracy": 0.7696881890296936,
      "num_tokens": 323835076.0,
      "step": 10146
    },
    {
      "entropy": 0.6314508384093642,
      "epoch": 0.9335886510140827,
      "grad_norm": 0.16868750751018524,
      "learning_rate": 6.88839819670623e-05,
      "loss": 0.6079,
      "mean_token_accuracy": 0.8173891380429268,
      "num_tokens": 323867063.0,
      "step": 10147
    },
    {
      "entropy": 0.6421156749129295,
      "epoch": 0.933680657385524,
      "grad_norm": 0.1488923877477646,
      "learning_rate": 6.888091514092066e-05,
      "loss": 0.6179,
      "mean_token_accuracy": 0.8155077770352364,
      "num_tokens": 323899164.0,
      "step": 10148
    },
    {
      "entropy": 0.7294517625123262,
      "epoch": 0.9337726637569652,
      "grad_norm": 0.14206893742084503,
      "learning_rate": 6.887784831477903e-05,
      "loss": 0.707,
      "mean_token_accuracy": 0.7897939309477806,
      "num_tokens": 323931353.0,
      "step": 10149
    },
    {
      "entropy": 0.6980528580024838,
      "epoch": 0.9338646701284063,
      "grad_norm": 0.15744632482528687,
      "learning_rate": 6.887478148863741e-05,
      "loss": 0.698,
      "mean_token_accuracy": 0.7966393753886223,
      "num_tokens": 323963729.0,
      "step": 10150
    },
    {
      "entropy": 0.6543837338685989,
      "epoch": 0.9339566764998476,
      "grad_norm": 0.153764009475708,
      "learning_rate": 6.887171466249579e-05,
      "loss": 0.6382,
      "mean_token_accuracy": 0.8058384172618389,
      "num_tokens": 323995235.0,
      "step": 10151
    },
    {
      "entropy": 0.6282222801819444,
      "epoch": 0.9340486828712888,
      "grad_norm": 0.1475261151790619,
      "learning_rate": 6.886864783635416e-05,
      "loss": 0.6113,
      "mean_token_accuracy": 0.8183899000287056,
      "num_tokens": 324027014.0,
      "step": 10152
    },
    {
      "entropy": 0.7818011604249477,
      "epoch": 0.9341406892427301,
      "grad_norm": 0.1548357456922531,
      "learning_rate": 6.886558101021253e-05,
      "loss": 0.7746,
      "mean_token_accuracy": 0.7727334536612034,
      "num_tokens": 324057996.0,
      "step": 10153
    },
    {
      "entropy": 0.6588866021484137,
      "epoch": 0.9342326956141713,
      "grad_norm": 0.1516200453042984,
      "learning_rate": 6.88625141840709e-05,
      "loss": 0.6477,
      "mean_token_accuracy": 0.8066430427134037,
      "num_tokens": 324090108.0,
      "step": 10154
    },
    {
      "entropy": 0.7376813665032387,
      "epoch": 0.9343247019856125,
      "grad_norm": 0.15880410373210907,
      "learning_rate": 6.885944735792929e-05,
      "loss": 0.7388,
      "mean_token_accuracy": 0.7802061587572098,
      "num_tokens": 324122230.0,
      "step": 10155
    },
    {
      "entropy": 0.7880259342491627,
      "epoch": 0.9344167083570537,
      "grad_norm": 0.16056005656719208,
      "learning_rate": 6.885638053178766e-05,
      "loss": 0.7592,
      "mean_token_accuracy": 0.7759217992424965,
      "num_tokens": 324153994.0,
      "step": 10156
    },
    {
      "entropy": 0.70884801261127,
      "epoch": 0.934508714728495,
      "grad_norm": 0.15792158246040344,
      "learning_rate": 6.885331370564603e-05,
      "loss": 0.7001,
      "mean_token_accuracy": 0.7906686328351498,
      "num_tokens": 324185175.0,
      "step": 10157
    },
    {
      "entropy": 0.7124788127839565,
      "epoch": 0.9346007210999362,
      "grad_norm": 0.15364503860473633,
      "learning_rate": 6.88502468795044e-05,
      "loss": 0.7086,
      "mean_token_accuracy": 0.7886479161679745,
      "num_tokens": 324216999.0,
      "step": 10158
    },
    {
      "entropy": 0.782972227782011,
      "epoch": 0.9346927274713774,
      "grad_norm": 0.15658265352249146,
      "learning_rate": 6.884718005336278e-05,
      "loss": 0.7813,
      "mean_token_accuracy": 0.7722495757043362,
      "num_tokens": 324248306.0,
      "step": 10159
    },
    {
      "entropy": 0.6909549608826637,
      "epoch": 0.9347847338428186,
      "grad_norm": 0.155491903424263,
      "learning_rate": 6.884411322722116e-05,
      "loss": 0.6806,
      "mean_token_accuracy": 0.794401329010725,
      "num_tokens": 324279928.0,
      "step": 10160
    },
    {
      "entropy": 0.7220604456961155,
      "epoch": 0.9348767402142598,
      "grad_norm": 0.14911699295043945,
      "learning_rate": 6.884104640107953e-05,
      "loss": 0.7107,
      "mean_token_accuracy": 0.7912051305174828,
      "num_tokens": 324312175.0,
      "step": 10161
    },
    {
      "entropy": 0.7396451802924275,
      "epoch": 0.9349687465857011,
      "grad_norm": 0.15460941195487976,
      "learning_rate": 6.883797957493791e-05,
      "loss": 0.7313,
      "mean_token_accuracy": 0.781394399702549,
      "num_tokens": 324343276.0,
      "step": 10162
    },
    {
      "entropy": 0.7455081949010491,
      "epoch": 0.9350607529571423,
      "grad_norm": 0.1670181006193161,
      "learning_rate": 6.883491274879628e-05,
      "loss": 0.7275,
      "mean_token_accuracy": 0.7860452309250832,
      "num_tokens": 324374529.0,
      "step": 10163
    },
    {
      "entropy": 0.6297427164390683,
      "epoch": 0.9351527593285835,
      "grad_norm": 0.15406657755374908,
      "learning_rate": 6.883184592265464e-05,
      "loss": 0.6323,
      "mean_token_accuracy": 0.8127997629344463,
      "num_tokens": 324406188.0,
      "step": 10164
    },
    {
      "entropy": 0.7169634252786636,
      "epoch": 0.9352447657000247,
      "grad_norm": 0.1581365019083023,
      "learning_rate": 6.882877909651302e-05,
      "loss": 0.7209,
      "mean_token_accuracy": 0.7916728295385838,
      "num_tokens": 324438335.0,
      "step": 10165
    },
    {
      "entropy": 0.7583516109734774,
      "epoch": 0.9353367720714659,
      "grad_norm": 0.15043190121650696,
      "learning_rate": 6.88257122703714e-05,
      "loss": 0.7485,
      "mean_token_accuracy": 0.7763632200658321,
      "num_tokens": 324471040.0,
      "step": 10166
    },
    {
      "entropy": 0.6892903102561831,
      "epoch": 0.9354287784429072,
      "grad_norm": 0.1483832597732544,
      "learning_rate": 6.882264544422977e-05,
      "loss": 0.6856,
      "mean_token_accuracy": 0.798353299498558,
      "num_tokens": 324503035.0,
      "step": 10167
    },
    {
      "entropy": 0.6980393305420876,
      "epoch": 0.9355207848143484,
      "grad_norm": 0.15539246797561646,
      "learning_rate": 6.881957861808814e-05,
      "loss": 0.684,
      "mean_token_accuracy": 0.7951562851667404,
      "num_tokens": 324534827.0,
      "step": 10168
    },
    {
      "entropy": 0.7157390704378486,
      "epoch": 0.9356127911857897,
      "grad_norm": 0.1677287369966507,
      "learning_rate": 6.881651179194651e-05,
      "loss": 0.7261,
      "mean_token_accuracy": 0.7815553657710552,
      "num_tokens": 324566348.0,
      "step": 10169
    },
    {
      "entropy": 0.6539084045216441,
      "epoch": 0.9357047975572308,
      "grad_norm": 0.14853854477405548,
      "learning_rate": 6.881344496580489e-05,
      "loss": 0.6392,
      "mean_token_accuracy": 0.8074598982930183,
      "num_tokens": 324597858.0,
      "step": 10170
    },
    {
      "entropy": 0.719328049570322,
      "epoch": 0.935796803928672,
      "grad_norm": 0.1542857587337494,
      "learning_rate": 6.881037813966327e-05,
      "loss": 0.7078,
      "mean_token_accuracy": 0.7895593754947186,
      "num_tokens": 324630485.0,
      "step": 10171
    },
    {
      "entropy": 0.6121994610875845,
      "epoch": 0.9358888103001133,
      "grad_norm": 0.14621984958648682,
      "learning_rate": 6.880731131352164e-05,
      "loss": 0.5957,
      "mean_token_accuracy": 0.8193430453538895,
      "num_tokens": 324662617.0,
      "step": 10172
    },
    {
      "entropy": 0.6609339145943522,
      "epoch": 0.9359808166715545,
      "grad_norm": 0.1538531482219696,
      "learning_rate": 6.880424448738001e-05,
      "loss": 0.6466,
      "mean_token_accuracy": 0.8095223233103752,
      "num_tokens": 324695003.0,
      "step": 10173
    },
    {
      "entropy": 0.7483523469418287,
      "epoch": 0.9360728230429958,
      "grad_norm": 0.15791191160678864,
      "learning_rate": 6.880117766123839e-05,
      "loss": 0.724,
      "mean_token_accuracy": 0.782615102827549,
      "num_tokens": 324727065.0,
      "step": 10174
    },
    {
      "entropy": 0.6919484511017799,
      "epoch": 0.9361648294144369,
      "grad_norm": 0.15736140310764313,
      "learning_rate": 6.879811083509676e-05,
      "loss": 0.6629,
      "mean_token_accuracy": 0.8003854900598526,
      "num_tokens": 324758359.0,
      "step": 10175
    },
    {
      "entropy": 0.6770732365548611,
      "epoch": 0.9362568357858781,
      "grad_norm": 0.15773750841617584,
      "learning_rate": 6.879504400895514e-05,
      "loss": 0.6757,
      "mean_token_accuracy": 0.7995035089552402,
      "num_tokens": 324790754.0,
      "step": 10176
    },
    {
      "entropy": 0.6493137767538428,
      "epoch": 0.9363488421573194,
      "grad_norm": 0.14820776879787445,
      "learning_rate": 6.87919771828135e-05,
      "loss": 0.6368,
      "mean_token_accuracy": 0.8095439001917839,
      "num_tokens": 324823297.0,
      "step": 10177
    },
    {
      "entropy": 0.7998994179069996,
      "epoch": 0.9364408485287606,
      "grad_norm": 0.16295860707759857,
      "learning_rate": 6.878891035667189e-05,
      "loss": 0.7915,
      "mean_token_accuracy": 0.768079724162817,
      "num_tokens": 324855642.0,
      "step": 10178
    },
    {
      "entropy": 0.7966304123401642,
      "epoch": 0.9365328549002019,
      "grad_norm": 0.1547269970178604,
      "learning_rate": 6.878584353053026e-05,
      "loss": 0.7895,
      "mean_token_accuracy": 0.7683715932071209,
      "num_tokens": 324887751.0,
      "step": 10179
    },
    {
      "entropy": 0.7996522262692451,
      "epoch": 0.936624861271643,
      "grad_norm": 0.16232949495315552,
      "learning_rate": 6.878277670438862e-05,
      "loss": 0.8022,
      "mean_token_accuracy": 0.7622728087007999,
      "num_tokens": 324919118.0,
      "step": 10180
    },
    {
      "entropy": 0.7506463248282671,
      "epoch": 0.9367168676430843,
      "grad_norm": 0.15489354729652405,
      "learning_rate": 6.8779709878247e-05,
      "loss": 0.7506,
      "mean_token_accuracy": 0.7797782458364964,
      "num_tokens": 324951301.0,
      "step": 10181
    },
    {
      "entropy": 0.6750179613009095,
      "epoch": 0.9368088740145255,
      "grad_norm": 0.14765667915344238,
      "learning_rate": 6.877664305210539e-05,
      "loss": 0.6298,
      "mean_token_accuracy": 0.8105587214231491,
      "num_tokens": 324983116.0,
      "step": 10182
    },
    {
      "entropy": 0.6595414532348514,
      "epoch": 0.9369008803859668,
      "grad_norm": 0.149120032787323,
      "learning_rate": 6.877357622596375e-05,
      "loss": 0.6368,
      "mean_token_accuracy": 0.8143734186887741,
      "num_tokens": 325014901.0,
      "step": 10183
    },
    {
      "entropy": 0.8206261266022921,
      "epoch": 0.936992886757408,
      "grad_norm": 0.15874889492988586,
      "learning_rate": 6.877050939982212e-05,
      "loss": 0.8178,
      "mean_token_accuracy": 0.7595624849200249,
      "num_tokens": 325047506.0,
      "step": 10184
    },
    {
      "entropy": 0.7795734163373709,
      "epoch": 0.9370848931288491,
      "grad_norm": 0.16267234086990356,
      "learning_rate": 6.876744257368049e-05,
      "loss": 0.767,
      "mean_token_accuracy": 0.7766121290624142,
      "num_tokens": 325080107.0,
      "step": 10185
    },
    {
      "entropy": 0.7462374651804566,
      "epoch": 0.9371768995002904,
      "grad_norm": 0.16366128623485565,
      "learning_rate": 6.876437574753887e-05,
      "loss": 0.7377,
      "mean_token_accuracy": 0.7854417487978935,
      "num_tokens": 325111868.0,
      "step": 10186
    },
    {
      "entropy": 0.8118211440742016,
      "epoch": 0.9372689058717316,
      "grad_norm": 0.15623357892036438,
      "learning_rate": 6.876130892139725e-05,
      "loss": 0.8175,
      "mean_token_accuracy": 0.7594134844839573,
      "num_tokens": 325143353.0,
      "step": 10187
    },
    {
      "entropy": 0.7579035516828299,
      "epoch": 0.9373609122431729,
      "grad_norm": 0.1516595035791397,
      "learning_rate": 6.875824209525562e-05,
      "loss": 0.7555,
      "mean_token_accuracy": 0.7753207422792912,
      "num_tokens": 325175851.0,
      "step": 10188
    },
    {
      "entropy": 0.6672601513564587,
      "epoch": 0.9374529186146141,
      "grad_norm": 0.15672117471694946,
      "learning_rate": 6.8755175269114e-05,
      "loss": 0.6365,
      "mean_token_accuracy": 0.8100520260632038,
      "num_tokens": 325207912.0,
      "step": 10189
    },
    {
      "entropy": 0.7568550575524569,
      "epoch": 0.9375449249860552,
      "grad_norm": 0.15948538482189178,
      "learning_rate": 6.875210844297237e-05,
      "loss": 0.7611,
      "mean_token_accuracy": 0.7762533910572529,
      "num_tokens": 325239278.0,
      "step": 10190
    },
    {
      "entropy": 0.7432481255382299,
      "epoch": 0.9376369313574965,
      "grad_norm": 0.15322010219097137,
      "learning_rate": 6.874904161683075e-05,
      "loss": 0.7387,
      "mean_token_accuracy": 0.7802418768405914,
      "num_tokens": 325272033.0,
      "step": 10191
    },
    {
      "entropy": 0.8094010036438704,
      "epoch": 0.9377289377289377,
      "grad_norm": 0.16251105070114136,
      "learning_rate": 6.874597479068912e-05,
      "loss": 0.8141,
      "mean_token_accuracy": 0.7640089616179466,
      "num_tokens": 325303767.0,
      "step": 10192
    },
    {
      "entropy": 0.7476240340620279,
      "epoch": 0.937820944100379,
      "grad_norm": 0.16165699064731598,
      "learning_rate": 6.87429079645475e-05,
      "loss": 0.7532,
      "mean_token_accuracy": 0.7776557728648186,
      "num_tokens": 325336128.0,
      "step": 10193
    },
    {
      "entropy": 0.7146274521946907,
      "epoch": 0.9379129504718202,
      "grad_norm": 0.1552668958902359,
      "learning_rate": 6.873984113840587e-05,
      "loss": 0.6966,
      "mean_token_accuracy": 0.7956897877156734,
      "num_tokens": 325368271.0,
      "step": 10194
    },
    {
      "entropy": 0.7658624183386564,
      "epoch": 0.9380049568432614,
      "grad_norm": 0.15073393285274506,
      "learning_rate": 6.873677431226424e-05,
      "loss": 0.7648,
      "mean_token_accuracy": 0.7740136459469795,
      "num_tokens": 325400076.0,
      "step": 10195
    },
    {
      "entropy": 0.6907831970602274,
      "epoch": 0.9380969632147026,
      "grad_norm": 0.15667015314102173,
      "learning_rate": 6.873370748612262e-05,
      "loss": 0.6796,
      "mean_token_accuracy": 0.7960389107465744,
      "num_tokens": 325431578.0,
      "step": 10196
    },
    {
      "entropy": 0.6457981672137976,
      "epoch": 0.9381889695861438,
      "grad_norm": 0.16848178207874298,
      "learning_rate": 6.8730640659981e-05,
      "loss": 0.6379,
      "mean_token_accuracy": 0.8074898980557919,
      "num_tokens": 325463900.0,
      "step": 10197
    },
    {
      "entropy": 0.7639705017209053,
      "epoch": 0.9382809759575851,
      "grad_norm": 0.15461745858192444,
      "learning_rate": 6.872757383383937e-05,
      "loss": 0.7577,
      "mean_token_accuracy": 0.7753758355975151,
      "num_tokens": 325495522.0,
      "step": 10198
    },
    {
      "entropy": 0.6272282041609287,
      "epoch": 0.9383729823290263,
      "grad_norm": 0.15350225567817688,
      "learning_rate": 6.872450700769774e-05,
      "loss": 0.6071,
      "mean_token_accuracy": 0.8177929371595383,
      "num_tokens": 325527962.0,
      "step": 10199
    },
    {
      "entropy": 0.6298452988266945,
      "epoch": 0.9384649887004675,
      "grad_norm": 0.1537771373987198,
      "learning_rate": 6.87214401815561e-05,
      "loss": 0.6006,
      "mean_token_accuracy": 0.8172840066254139,
      "num_tokens": 325559899.0,
      "step": 10200
    },
    {
      "entropy": 0.6930432487279177,
      "epoch": 0.9385569950719087,
      "grad_norm": 0.15250229835510254,
      "learning_rate": 6.871837335541448e-05,
      "loss": 0.6814,
      "mean_token_accuracy": 0.7980617173016071,
      "num_tokens": 325591216.0,
      "step": 10201
    },
    {
      "entropy": 0.7052438016980886,
      "epoch": 0.93864900144335,
      "grad_norm": 0.15518039464950562,
      "learning_rate": 6.871530652927287e-05,
      "loss": 0.6943,
      "mean_token_accuracy": 0.7916542179882526,
      "num_tokens": 325622936.0,
      "step": 10202
    },
    {
      "entropy": 0.7359449686482549,
      "epoch": 0.9387410078147912,
      "grad_norm": 0.15378747880458832,
      "learning_rate": 6.871223970313123e-05,
      "loss": 0.7193,
      "mean_token_accuracy": 0.783555518835783,
      "num_tokens": 325654725.0,
      "step": 10203
    },
    {
      "entropy": 0.707669734954834,
      "epoch": 0.9388330141862324,
      "grad_norm": 0.1648680716753006,
      "learning_rate": 6.87091728769896e-05,
      "loss": 0.706,
      "mean_token_accuracy": 0.7928328439593315,
      "num_tokens": 325686674.0,
      "step": 10204
    },
    {
      "entropy": 0.8753211302682757,
      "epoch": 0.9389250205576736,
      "grad_norm": 0.16553765535354614,
      "learning_rate": 6.870610605084798e-05,
      "loss": 0.8593,
      "mean_token_accuracy": 0.7489217035472393,
      "num_tokens": 325718465.0,
      "step": 10205
    },
    {
      "entropy": 0.6870389878749847,
      "epoch": 0.9390170269291148,
      "grad_norm": 0.14939694106578827,
      "learning_rate": 6.870303922470635e-05,
      "loss": 0.6649,
      "mean_token_accuracy": 0.8017483241856098,
      "num_tokens": 325751000.0,
      "step": 10206
    },
    {
      "entropy": 0.7904809061437845,
      "epoch": 0.9391090333005561,
      "grad_norm": 0.15569579601287842,
      "learning_rate": 6.869997239856473e-05,
      "loss": 0.785,
      "mean_token_accuracy": 0.772856842726469,
      "num_tokens": 325783253.0,
      "step": 10207
    },
    {
      "entropy": 0.754199156537652,
      "epoch": 0.9392010396719973,
      "grad_norm": 0.1519538313150406,
      "learning_rate": 6.86969055724231e-05,
      "loss": 0.7388,
      "mean_token_accuracy": 0.7801513671875,
      "num_tokens": 325815141.0,
      "step": 10208
    },
    {
      "entropy": 0.793554050847888,
      "epoch": 0.9392930460434386,
      "grad_norm": 0.1515205353498459,
      "learning_rate": 6.869383874628148e-05,
      "loss": 0.7907,
      "mean_token_accuracy": 0.76847705245018,
      "num_tokens": 325847189.0,
      "step": 10209
    },
    {
      "entropy": 0.6491711316630244,
      "epoch": 0.9393850524148797,
      "grad_norm": 0.1513260304927826,
      "learning_rate": 6.869077192013985e-05,
      "loss": 0.646,
      "mean_token_accuracy": 0.8057085201144218,
      "num_tokens": 325878219.0,
      "step": 10210
    },
    {
      "entropy": 0.591742028016597,
      "epoch": 0.9394770587863209,
      "grad_norm": 0.14532308280467987,
      "learning_rate": 6.868770509399822e-05,
      "loss": 0.5879,
      "mean_token_accuracy": 0.8245221525430679,
      "num_tokens": 325910778.0,
      "step": 10211
    },
    {
      "entropy": 0.7751869149506092,
      "epoch": 0.9395690651577622,
      "grad_norm": 0.1556798666715622,
      "learning_rate": 6.86846382678566e-05,
      "loss": 0.7773,
      "mean_token_accuracy": 0.7691521495580673,
      "num_tokens": 325942731.0,
      "step": 10212
    },
    {
      "entropy": 0.7364711677655578,
      "epoch": 0.9396610715292034,
      "grad_norm": 0.14878953993320465,
      "learning_rate": 6.868157144171498e-05,
      "loss": 0.7262,
      "mean_token_accuracy": 0.787368930876255,
      "num_tokens": 325974589.0,
      "step": 10213
    },
    {
      "entropy": 0.6361248884350061,
      "epoch": 0.9397530779006447,
      "grad_norm": 0.14425182342529297,
      "learning_rate": 6.867850461557335e-05,
      "loss": 0.6209,
      "mean_token_accuracy": 0.8126473240554333,
      "num_tokens": 326006856.0,
      "step": 10214
    },
    {
      "entropy": 0.6959371343255043,
      "epoch": 0.9398450842720858,
      "grad_norm": 0.1496036946773529,
      "learning_rate": 6.867543778943172e-05,
      "loss": 0.6664,
      "mean_token_accuracy": 0.8056738451123238,
      "num_tokens": 326038716.0,
      "step": 10215
    },
    {
      "entropy": 0.5770297516137362,
      "epoch": 0.939937090643527,
      "grad_norm": 0.15520043671131134,
      "learning_rate": 6.867237096329008e-05,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8293508477509022,
      "num_tokens": 326071190.0,
      "step": 10216
    },
    {
      "entropy": 0.8251059278845787,
      "epoch": 0.9400290970149683,
      "grad_norm": 0.16532674431800842,
      "learning_rate": 6.866930413714847e-05,
      "loss": 0.8174,
      "mean_token_accuracy": 0.7588772214949131,
      "num_tokens": 326103244.0,
      "step": 10217
    },
    {
      "entropy": 0.8426848892122507,
      "epoch": 0.9401211033864095,
      "grad_norm": 0.15378418564796448,
      "learning_rate": 6.866623731100685e-05,
      "loss": 0.8361,
      "mean_token_accuracy": 0.7598234564065933,
      "num_tokens": 326135086.0,
      "step": 10218
    },
    {
      "entropy": 0.6852692160755396,
      "epoch": 0.9402131097578508,
      "grad_norm": 0.15558287501335144,
      "learning_rate": 6.866317048486521e-05,
      "loss": 0.6728,
      "mean_token_accuracy": 0.7990985848009586,
      "num_tokens": 326166321.0,
      "step": 10219
    },
    {
      "entropy": 0.7526608817279339,
      "epoch": 0.9403051161292919,
      "grad_norm": 0.15306012332439423,
      "learning_rate": 6.86601036587236e-05,
      "loss": 0.7433,
      "mean_token_accuracy": 0.7744256146252155,
      "num_tokens": 326197363.0,
      "step": 10220
    },
    {
      "entropy": 0.6354440115392208,
      "epoch": 0.9403971225007332,
      "grad_norm": 0.14972570538520813,
      "learning_rate": 6.865703683258196e-05,
      "loss": 0.6269,
      "mean_token_accuracy": 0.8109714351594448,
      "num_tokens": 326229706.0,
      "step": 10221
    },
    {
      "entropy": 0.7949503306299448,
      "epoch": 0.9404891288721744,
      "grad_norm": 0.1568681299686432,
      "learning_rate": 6.865397000644033e-05,
      "loss": 0.7911,
      "mean_token_accuracy": 0.7654209993779659,
      "num_tokens": 326261586.0,
      "step": 10222
    },
    {
      "entropy": 0.8159114345908165,
      "epoch": 0.9405811352436156,
      "grad_norm": 0.16404984891414642,
      "learning_rate": 6.865090318029871e-05,
      "loss": 0.8163,
      "mean_token_accuracy": 0.7590224631130695,
      "num_tokens": 326293642.0,
      "step": 10223
    },
    {
      "entropy": 0.6509511629119515,
      "epoch": 0.9406731416150569,
      "grad_norm": 0.1594287008047104,
      "learning_rate": 6.86478363541571e-05,
      "loss": 0.6476,
      "mean_token_accuracy": 0.8082477636635303,
      "num_tokens": 326325709.0,
      "step": 10224
    },
    {
      "entropy": 0.7140600923448801,
      "epoch": 0.940765147986498,
      "grad_norm": 0.15128649771213531,
      "learning_rate": 6.864476952801546e-05,
      "loss": 0.715,
      "mean_token_accuracy": 0.7873488441109657,
      "num_tokens": 326357792.0,
      "step": 10225
    },
    {
      "entropy": 0.7309846244752407,
      "epoch": 0.9408571543579393,
      "grad_norm": 0.1568520963191986,
      "learning_rate": 6.864170270187383e-05,
      "loss": 0.731,
      "mean_token_accuracy": 0.782810527831316,
      "num_tokens": 326389270.0,
      "step": 10226
    },
    {
      "entropy": 0.7410568911582232,
      "epoch": 0.9409491607293805,
      "grad_norm": 0.16111113131046295,
      "learning_rate": 6.86386358757322e-05,
      "loss": 0.7327,
      "mean_token_accuracy": 0.7835461907088757,
      "num_tokens": 326421231.0,
      "step": 10227
    },
    {
      "entropy": 0.61137705296278,
      "epoch": 0.9410411671008218,
      "grad_norm": 0.18584522604942322,
      "learning_rate": 6.86355690495906e-05,
      "loss": 0.5844,
      "mean_token_accuracy": 0.8221396431326866,
      "num_tokens": 326452670.0,
      "step": 10228
    },
    {
      "entropy": 0.7358296941965818,
      "epoch": 0.941133173472263,
      "grad_norm": 0.1519259363412857,
      "learning_rate": 6.863250222344896e-05,
      "loss": 0.7308,
      "mean_token_accuracy": 0.7850101701915264,
      "num_tokens": 326484271.0,
      "step": 10229
    },
    {
      "entropy": 0.6720579452812672,
      "epoch": 0.9412251798437041,
      "grad_norm": 0.16491609811782837,
      "learning_rate": 6.862943539730733e-05,
      "loss": 0.6794,
      "mean_token_accuracy": 0.7996390797197819,
      "num_tokens": 326516700.0,
      "step": 10230
    },
    {
      "entropy": 0.5724028656259179,
      "epoch": 0.9413171862151454,
      "grad_norm": 0.14736048877239227,
      "learning_rate": 6.86263685711657e-05,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8353749327361584,
      "num_tokens": 326548432.0,
      "step": 10231
    },
    {
      "entropy": 0.6638455335050821,
      "epoch": 0.9414091925865866,
      "grad_norm": 0.15493811666965485,
      "learning_rate": 6.862330174502408e-05,
      "loss": 0.6523,
      "mean_token_accuracy": 0.8107175230979919,
      "num_tokens": 326579834.0,
      "step": 10232
    },
    {
      "entropy": 0.7767457216978073,
      "epoch": 0.9415011989580279,
      "grad_norm": 0.15576167404651642,
      "learning_rate": 6.862023491888246e-05,
      "loss": 0.7685,
      "mean_token_accuracy": 0.7726990208029747,
      "num_tokens": 326611599.0,
      "step": 10233
    },
    {
      "entropy": 0.826952775940299,
      "epoch": 0.9415932053294691,
      "grad_norm": 0.16090598702430725,
      "learning_rate": 6.861716809274083e-05,
      "loss": 0.8191,
      "mean_token_accuracy": 0.7627857215702534,
      "num_tokens": 326643211.0,
      "step": 10234
    },
    {
      "entropy": 0.721154673025012,
      "epoch": 0.9416852117009102,
      "grad_norm": 0.15333406627178192,
      "learning_rate": 6.86141012665992e-05,
      "loss": 0.7068,
      "mean_token_accuracy": 0.7920170053839684,
      "num_tokens": 326675329.0,
      "step": 10235
    },
    {
      "entropy": 0.7062386628240347,
      "epoch": 0.9417772180723515,
      "grad_norm": 0.1497337818145752,
      "learning_rate": 6.861103444045758e-05,
      "loss": 0.6738,
      "mean_token_accuracy": 0.7980586849153042,
      "num_tokens": 326707576.0,
      "step": 10236
    },
    {
      "entropy": 0.8105028811842203,
      "epoch": 0.9418692244437927,
      "grad_norm": 0.15584246814250946,
      "learning_rate": 6.860796761431594e-05,
      "loss": 0.8146,
      "mean_token_accuracy": 0.7640279568731785,
      "num_tokens": 326739817.0,
      "step": 10237
    },
    {
      "entropy": 0.7024173736572266,
      "epoch": 0.941961230815234,
      "grad_norm": 0.15445125102996826,
      "learning_rate": 6.860490078817433e-05,
      "loss": 0.6932,
      "mean_token_accuracy": 0.7956416234374046,
      "num_tokens": 326771023.0,
      "step": 10238
    },
    {
      "entropy": 0.7596748024225235,
      "epoch": 0.9420532371866752,
      "grad_norm": 0.1645587831735611,
      "learning_rate": 6.86018339620327e-05,
      "loss": 0.7459,
      "mean_token_accuracy": 0.7764328494668007,
      "num_tokens": 326802482.0,
      "step": 10239
    },
    {
      "entropy": 0.7880463358014822,
      "epoch": 0.9421452435581164,
      "grad_norm": 0.15590660274028778,
      "learning_rate": 6.859876713589108e-05,
      "loss": 0.7595,
      "mean_token_accuracy": 0.7775643356144428,
      "num_tokens": 326834063.0,
      "step": 10240
    },
    {
      "entropy": 0.5685094054788351,
      "epoch": 0.9422372499295576,
      "grad_norm": 0.14119035005569458,
      "learning_rate": 6.859570030974944e-05,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8341074474155903,
      "num_tokens": 326865632.0,
      "step": 10241
    },
    {
      "entropy": 0.7712514717131853,
      "epoch": 0.9423292563009988,
      "grad_norm": 0.1505737602710724,
      "learning_rate": 6.859263348360781e-05,
      "loss": 0.7671,
      "mean_token_accuracy": 0.77292275801301,
      "num_tokens": 326898026.0,
      "step": 10242
    },
    {
      "entropy": 0.6329084280878305,
      "epoch": 0.9424212626724401,
      "grad_norm": 0.15849122405052185,
      "learning_rate": 6.858956665746619e-05,
      "loss": 0.6247,
      "mean_token_accuracy": 0.8136962912976742,
      "num_tokens": 326930051.0,
      "step": 10243
    },
    {
      "entropy": 0.743819460272789,
      "epoch": 0.9425132690438813,
      "grad_norm": 0.15172813832759857,
      "learning_rate": 6.858649983132457e-05,
      "loss": 0.7261,
      "mean_token_accuracy": 0.7860636338591576,
      "num_tokens": 326962416.0,
      "step": 10244
    },
    {
      "entropy": 0.7510775960981846,
      "epoch": 0.9426052754153225,
      "grad_norm": 0.16291096806526184,
      "learning_rate": 6.858343300518294e-05,
      "loss": 0.7501,
      "mean_token_accuracy": 0.7797754406929016,
      "num_tokens": 326993703.0,
      "step": 10245
    },
    {
      "entropy": 0.7046045921742916,
      "epoch": 0.9426972817867637,
      "grad_norm": 0.155190110206604,
      "learning_rate": 6.858036617904131e-05,
      "loss": 0.6967,
      "mean_token_accuracy": 0.7907453961670399,
      "num_tokens": 327026385.0,
      "step": 10246
    },
    {
      "entropy": 0.7297260612249374,
      "epoch": 0.942789288158205,
      "grad_norm": 0.15695008635520935,
      "learning_rate": 6.857729935289968e-05,
      "loss": 0.7208,
      "mean_token_accuracy": 0.7900940291583538,
      "num_tokens": 327058209.0,
      "step": 10247
    },
    {
      "entropy": 0.7553333342075348,
      "epoch": 0.9428812945296462,
      "grad_norm": 0.1624874323606491,
      "learning_rate": 6.857423252675806e-05,
      "loss": 0.743,
      "mean_token_accuracy": 0.7820018902420998,
      "num_tokens": 327090377.0,
      "step": 10248
    },
    {
      "entropy": 0.7852437272667885,
      "epoch": 0.9429733009010874,
      "grad_norm": 0.16023242473602295,
      "learning_rate": 6.857116570061644e-05,
      "loss": 0.791,
      "mean_token_accuracy": 0.7702346965670586,
      "num_tokens": 327122481.0,
      "step": 10249
    },
    {
      "entropy": 0.5893347314558923,
      "epoch": 0.9430653072725286,
      "grad_norm": 0.15100273489952087,
      "learning_rate": 6.856809887447481e-05,
      "loss": 0.5867,
      "mean_token_accuracy": 0.8231504112482071,
      "num_tokens": 327154866.0,
      "step": 10250
    },
    {
      "entropy": 0.761835565790534,
      "epoch": 0.9431573136439698,
      "grad_norm": 0.1613786369562149,
      "learning_rate": 6.856503204833319e-05,
      "loss": 0.7721,
      "mean_token_accuracy": 0.7739069387316704,
      "num_tokens": 327187634.0,
      "step": 10251
    },
    {
      "entropy": 0.752957072108984,
      "epoch": 0.9432493200154111,
      "grad_norm": 0.15474851429462433,
      "learning_rate": 6.856196522219156e-05,
      "loss": 0.7517,
      "mean_token_accuracy": 0.7795467264950275,
      "num_tokens": 327219329.0,
      "step": 10252
    },
    {
      "entropy": 0.7854017056524754,
      "epoch": 0.9433413263868523,
      "grad_norm": 0.155962735414505,
      "learning_rate": 6.855889839604993e-05,
      "loss": 0.7875,
      "mean_token_accuracy": 0.7715791165828705,
      "num_tokens": 327251298.0,
      "step": 10253
    },
    {
      "entropy": 0.6386408880352974,
      "epoch": 0.9434333327582936,
      "grad_norm": 0.16241240501403809,
      "learning_rate": 6.855583156990831e-05,
      "loss": 0.6396,
      "mean_token_accuracy": 0.810828872025013,
      "num_tokens": 327283726.0,
      "step": 10254
    },
    {
      "entropy": 0.7947721257805824,
      "epoch": 0.9435253391297347,
      "grad_norm": 0.15866787731647491,
      "learning_rate": 6.855276474376669e-05,
      "loss": 0.7956,
      "mean_token_accuracy": 0.7664678990840912,
      "num_tokens": 327315789.0,
      "step": 10255
    },
    {
      "entropy": 0.6443138346076012,
      "epoch": 0.9436173455011759,
      "grad_norm": 0.14558905363082886,
      "learning_rate": 6.854969791762506e-05,
      "loss": 0.6357,
      "mean_token_accuracy": 0.8125547766685486,
      "num_tokens": 327347300.0,
      "step": 10256
    },
    {
      "entropy": 0.5813431283459067,
      "epoch": 0.9437093518726172,
      "grad_norm": 0.15731364488601685,
      "learning_rate": 6.854663109148342e-05,
      "loss": 0.5543,
      "mean_token_accuracy": 0.831223264336586,
      "num_tokens": 327379998.0,
      "step": 10257
    },
    {
      "entropy": 0.759287353605032,
      "epoch": 0.9438013582440584,
      "grad_norm": 0.15261337161064148,
      "learning_rate": 6.854356426534179e-05,
      "loss": 0.747,
      "mean_token_accuracy": 0.7787977829575539,
      "num_tokens": 327412331.0,
      "step": 10258
    },
    {
      "entropy": 0.7852251995354891,
      "epoch": 0.9438933646154997,
      "grad_norm": 0.15769319236278534,
      "learning_rate": 6.854049743920017e-05,
      "loss": 0.7803,
      "mean_token_accuracy": 0.7729426510632038,
      "num_tokens": 327444869.0,
      "step": 10259
    },
    {
      "entropy": 0.7724103406071663,
      "epoch": 0.9439853709869408,
      "grad_norm": 0.16673992574214935,
      "learning_rate": 6.853743061305855e-05,
      "loss": 0.7574,
      "mean_token_accuracy": 0.7779949344694614,
      "num_tokens": 327476918.0,
      "step": 10260
    },
    {
      "entropy": 0.7917922977358103,
      "epoch": 0.944077377358382,
      "grad_norm": 0.14707933366298676,
      "learning_rate": 6.853436378691692e-05,
      "loss": 0.7784,
      "mean_token_accuracy": 0.7681001462042332,
      "num_tokens": 327508881.0,
      "step": 10261
    },
    {
      "entropy": 0.7699980642646551,
      "epoch": 0.9441693837298233,
      "grad_norm": 0.16266222298145294,
      "learning_rate": 6.853129696077529e-05,
      "loss": 0.7351,
      "mean_token_accuracy": 0.7825841307640076,
      "num_tokens": 327540175.0,
      "step": 10262
    },
    {
      "entropy": 0.6927220430225134,
      "epoch": 0.9442613901012645,
      "grad_norm": 0.1594981700181961,
      "learning_rate": 6.852823013463367e-05,
      "loss": 0.6825,
      "mean_token_accuracy": 0.7973579801619053,
      "num_tokens": 327571640.0,
      "step": 10263
    },
    {
      "entropy": 0.6439159344881773,
      "epoch": 0.9443533964727058,
      "grad_norm": 0.14696477353572845,
      "learning_rate": 6.852516330849204e-05,
      "loss": 0.6286,
      "mean_token_accuracy": 0.8115537203848362,
      "num_tokens": 327603228.0,
      "step": 10264
    },
    {
      "entropy": 0.6609776299446821,
      "epoch": 0.9444454028441469,
      "grad_norm": 0.1481802612543106,
      "learning_rate": 6.852209648235042e-05,
      "loss": 0.6576,
      "mean_token_accuracy": 0.807225052267313,
      "num_tokens": 327635725.0,
      "step": 10265
    },
    {
      "entropy": 0.6196193341165781,
      "epoch": 0.9445374092155882,
      "grad_norm": 0.16480277478694916,
      "learning_rate": 6.851902965620879e-05,
      "loss": 0.6046,
      "mean_token_accuracy": 0.817658469080925,
      "num_tokens": 327667572.0,
      "step": 10266
    },
    {
      "entropy": 0.6424077972769737,
      "epoch": 0.9446294155870294,
      "grad_norm": 0.15999138355255127,
      "learning_rate": 6.851596283006717e-05,
      "loss": 0.6267,
      "mean_token_accuracy": 0.8194380663335323,
      "num_tokens": 327699107.0,
      "step": 10267
    },
    {
      "entropy": 0.7649979684501886,
      "epoch": 0.9447214219584706,
      "grad_norm": 0.16527801752090454,
      "learning_rate": 6.851289600392554e-05,
      "loss": 0.7517,
      "mean_token_accuracy": 0.7740330100059509,
      "num_tokens": 327730024.0,
      "step": 10268
    },
    {
      "entropy": 0.7123095486313105,
      "epoch": 0.9448134283299119,
      "grad_norm": 0.15189369022846222,
      "learning_rate": 6.85098291777839e-05,
      "loss": 0.6942,
      "mean_token_accuracy": 0.7916049435734749,
      "num_tokens": 327761463.0,
      "step": 10269
    },
    {
      "entropy": 0.7089124452322721,
      "epoch": 0.944905434701353,
      "grad_norm": 0.14614856243133545,
      "learning_rate": 6.850676235164229e-05,
      "loss": 0.7034,
      "mean_token_accuracy": 0.7924223393201828,
      "num_tokens": 327794115.0,
      "step": 10270
    },
    {
      "entropy": 0.7367655672132969,
      "epoch": 0.9449974410727943,
      "grad_norm": 0.15745142102241516,
      "learning_rate": 6.850369552550067e-05,
      "loss": 0.7246,
      "mean_token_accuracy": 0.7848134115338326,
      "num_tokens": 327825966.0,
      "step": 10271
    },
    {
      "entropy": 0.6902289371937513,
      "epoch": 0.9450894474442355,
      "grad_norm": 0.1657087504863739,
      "learning_rate": 6.850062869935904e-05,
      "loss": 0.6875,
      "mean_token_accuracy": 0.7951385080814362,
      "num_tokens": 327858549.0,
      "step": 10272
    },
    {
      "entropy": 0.6843404471874237,
      "epoch": 0.9451814538156768,
      "grad_norm": 0.15135960280895233,
      "learning_rate": 6.84975618732174e-05,
      "loss": 0.6628,
      "mean_token_accuracy": 0.8016533628106117,
      "num_tokens": 327890522.0,
      "step": 10273
    },
    {
      "entropy": 0.6690133027732372,
      "epoch": 0.945273460187118,
      "grad_norm": 0.15363626182079315,
      "learning_rate": 6.849449504707577e-05,
      "loss": 0.6627,
      "mean_token_accuracy": 0.8010441064834595,
      "num_tokens": 327922681.0,
      "step": 10274
    },
    {
      "entropy": 0.7162890676409006,
      "epoch": 0.9453654665585591,
      "grad_norm": 0.15165838599205017,
      "learning_rate": 6.849142822093417e-05,
      "loss": 0.7008,
      "mean_token_accuracy": 0.7939362563192844,
      "num_tokens": 327955449.0,
      "step": 10275
    },
    {
      "entropy": 0.7244958505034447,
      "epoch": 0.9454574729300004,
      "grad_norm": 0.15537843108177185,
      "learning_rate": 6.848836139479254e-05,
      "loss": 0.7249,
      "mean_token_accuracy": 0.7845965512096882,
      "num_tokens": 327987646.0,
      "step": 10276
    },
    {
      "entropy": 0.8555167317390442,
      "epoch": 0.9455494793014416,
      "grad_norm": 0.16201376914978027,
      "learning_rate": 6.84852945686509e-05,
      "loss": 0.8579,
      "mean_token_accuracy": 0.748564638197422,
      "num_tokens": 328019355.0,
      "step": 10277
    },
    {
      "entropy": 0.7270487304776907,
      "epoch": 0.9456414856728829,
      "grad_norm": 0.1523723602294922,
      "learning_rate": 6.848222774250927e-05,
      "loss": 0.6963,
      "mean_token_accuracy": 0.7890530750155449,
      "num_tokens": 328051237.0,
      "step": 10278
    },
    {
      "entropy": 0.7126069571822882,
      "epoch": 0.9457334920443241,
      "grad_norm": 0.15443724393844604,
      "learning_rate": 6.847916091636765e-05,
      "loss": 0.7062,
      "mean_token_accuracy": 0.7893201671540737,
      "num_tokens": 328083064.0,
      "step": 10279
    },
    {
      "entropy": 0.8223051410168409,
      "epoch": 0.9458254984157652,
      "grad_norm": 0.16107648611068726,
      "learning_rate": 6.847609409022603e-05,
      "loss": 0.8095,
      "mean_token_accuracy": 0.7585391663014889,
      "num_tokens": 328115063.0,
      "step": 10280
    },
    {
      "entropy": 0.7284814957529306,
      "epoch": 0.9459175047872065,
      "grad_norm": 0.15903064608573914,
      "learning_rate": 6.84730272640844e-05,
      "loss": 0.7145,
      "mean_token_accuracy": 0.7879007384181023,
      "num_tokens": 328146940.0,
      "step": 10281
    },
    {
      "entropy": 0.6974259093403816,
      "epoch": 0.9460095111586477,
      "grad_norm": 0.15483807027339935,
      "learning_rate": 6.846996043794278e-05,
      "loss": 0.7135,
      "mean_token_accuracy": 0.7872941009700298,
      "num_tokens": 328178362.0,
      "step": 10282
    },
    {
      "entropy": 0.6267587635666132,
      "epoch": 0.946101517530089,
      "grad_norm": 0.14746347069740295,
      "learning_rate": 6.846689361180115e-05,
      "loss": 0.6103,
      "mean_token_accuracy": 0.8171828649938107,
      "num_tokens": 328210724.0,
      "step": 10283
    },
    {
      "entropy": 0.7485371679067612,
      "epoch": 0.9461935239015302,
      "grad_norm": 0.15496504306793213,
      "learning_rate": 6.846382678565952e-05,
      "loss": 0.7328,
      "mean_token_accuracy": 0.7825896814465523,
      "num_tokens": 328242593.0,
      "step": 10284
    },
    {
      "entropy": 0.7129318099468946,
      "epoch": 0.9462855302729714,
      "grad_norm": 0.15777646005153656,
      "learning_rate": 6.84607599595179e-05,
      "loss": 0.703,
      "mean_token_accuracy": 0.7922272272408009,
      "num_tokens": 328274529.0,
      "step": 10285
    },
    {
      "entropy": 0.6156075457111001,
      "epoch": 0.9463775366444126,
      "grad_norm": 0.1528434455394745,
      "learning_rate": 6.845769313337628e-05,
      "loss": 0.6147,
      "mean_token_accuracy": 0.8134769909083843,
      "num_tokens": 328306954.0,
      "step": 10286
    },
    {
      "entropy": 0.7400847058743238,
      "epoch": 0.9464695430158538,
      "grad_norm": 0.16170033812522888,
      "learning_rate": 6.845462630723465e-05,
      "loss": 0.7483,
      "mean_token_accuracy": 0.7786560244858265,
      "num_tokens": 328339409.0,
      "step": 10287
    },
    {
      "entropy": 0.6700083632022142,
      "epoch": 0.9465615493872951,
      "grad_norm": 0.16100279986858368,
      "learning_rate": 6.845155948109302e-05,
      "loss": 0.6473,
      "mean_token_accuracy": 0.8059649765491486,
      "num_tokens": 328371384.0,
      "step": 10288
    },
    {
      "entropy": 0.8378865458071232,
      "epoch": 0.9466535557587363,
      "grad_norm": 0.163504496216774,
      "learning_rate": 6.844849265495139e-05,
      "loss": 0.8432,
      "mean_token_accuracy": 0.7534159682691097,
      "num_tokens": 328403752.0,
      "step": 10289
    },
    {
      "entropy": 0.7190609239041805,
      "epoch": 0.9467455621301775,
      "grad_norm": 0.15694665908813477,
      "learning_rate": 6.844542582880977e-05,
      "loss": 0.692,
      "mean_token_accuracy": 0.7948270030319691,
      "num_tokens": 328435020.0,
      "step": 10290
    },
    {
      "entropy": 0.5856334585696459,
      "epoch": 0.9468375685016187,
      "grad_norm": 0.1510893553495407,
      "learning_rate": 6.844235900266815e-05,
      "loss": 0.5779,
      "mean_token_accuracy": 0.8267110958695412,
      "num_tokens": 328467029.0,
      "step": 10291
    },
    {
      "entropy": 0.8239528499543667,
      "epoch": 0.94692957487306,
      "grad_norm": 0.15996317565441132,
      "learning_rate": 6.843929217652652e-05,
      "loss": 0.8141,
      "mean_token_accuracy": 0.7610402218997478,
      "num_tokens": 328498349.0,
      "step": 10292
    },
    {
      "entropy": 0.6855325326323509,
      "epoch": 0.9470215812445012,
      "grad_norm": 0.15924198925495148,
      "learning_rate": 6.843622535038488e-05,
      "loss": 0.6572,
      "mean_token_accuracy": 0.802146390080452,
      "num_tokens": 328530289.0,
      "step": 10293
    },
    {
      "entropy": 0.8513432946056128,
      "epoch": 0.9471135876159424,
      "grad_norm": 0.16020697355270386,
      "learning_rate": 6.843315852424327e-05,
      "loss": 0.8259,
      "mean_token_accuracy": 0.7598823942244053,
      "num_tokens": 328562920.0,
      "step": 10294
    },
    {
      "entropy": 0.6496013347059488,
      "epoch": 0.9472055939873836,
      "grad_norm": 0.14726974070072174,
      "learning_rate": 6.843009169810163e-05,
      "loss": 0.6245,
      "mean_token_accuracy": 0.8131667338311672,
      "num_tokens": 328594486.0,
      "step": 10295
    },
    {
      "entropy": 0.8018520437180996,
      "epoch": 0.9472976003588248,
      "grad_norm": 0.15916411578655243,
      "learning_rate": 6.842702487196002e-05,
      "loss": 0.7838,
      "mean_token_accuracy": 0.7722727358341217,
      "num_tokens": 328625552.0,
      "step": 10296
    },
    {
      "entropy": 0.6795486472547054,
      "epoch": 0.9473896067302661,
      "grad_norm": 0.15456868708133698,
      "learning_rate": 6.842395804581838e-05,
      "loss": 0.6575,
      "mean_token_accuracy": 0.8035749197006226,
      "num_tokens": 328656515.0,
      "step": 10297
    },
    {
      "entropy": 0.7545772902667522,
      "epoch": 0.9474816131017073,
      "grad_norm": 0.15400910377502441,
      "learning_rate": 6.842089121967676e-05,
      "loss": 0.734,
      "mean_token_accuracy": 0.7799118459224701,
      "num_tokens": 328689198.0,
      "step": 10298
    },
    {
      "entropy": 0.6595485396683216,
      "epoch": 0.9475736194731486,
      "grad_norm": 0.14935563504695892,
      "learning_rate": 6.841782439353513e-05,
      "loss": 0.6536,
      "mean_token_accuracy": 0.8041111901402473,
      "num_tokens": 328721933.0,
      "step": 10299
    },
    {
      "entropy": 0.7250062450766563,
      "epoch": 0.9476656258445897,
      "grad_norm": 0.15180587768554688,
      "learning_rate": 6.84147575673935e-05,
      "loss": 0.7164,
      "mean_token_accuracy": 0.7907767035067081,
      "num_tokens": 328754509.0,
      "step": 10300
    },
    {
      "entropy": 0.7398843094706535,
      "epoch": 0.9477576322160309,
      "grad_norm": 0.15686509013175964,
      "learning_rate": 6.841169074125188e-05,
      "loss": 0.7182,
      "mean_token_accuracy": 0.7886806055903435,
      "num_tokens": 328786927.0,
      "step": 10301
    },
    {
      "entropy": 0.7243082784116268,
      "epoch": 0.9478496385874722,
      "grad_norm": 0.15350216627120972,
      "learning_rate": 6.840862391511026e-05,
      "loss": 0.7169,
      "mean_token_accuracy": 0.7848932966589928,
      "num_tokens": 328818987.0,
      "step": 10302
    },
    {
      "entropy": 0.7942668218165636,
      "epoch": 0.9479416449589134,
      "grad_norm": 0.1576240360736847,
      "learning_rate": 6.840555708896863e-05,
      "loss": 0.7869,
      "mean_token_accuracy": 0.7669716812670231,
      "num_tokens": 328850132.0,
      "step": 10303
    },
    {
      "entropy": 0.6700397096574306,
      "epoch": 0.9480336513303547,
      "grad_norm": 0.15084582567214966,
      "learning_rate": 6.8402490262827e-05,
      "loss": 0.6588,
      "mean_token_accuracy": 0.8035118989646435,
      "num_tokens": 328882710.0,
      "step": 10304
    },
    {
      "entropy": 0.6904528848826885,
      "epoch": 0.9481256577017958,
      "grad_norm": 0.14988377690315247,
      "learning_rate": 6.839942343668537e-05,
      "loss": 0.681,
      "mean_token_accuracy": 0.797615434974432,
      "num_tokens": 328915090.0,
      "step": 10305
    },
    {
      "entropy": 0.676810160279274,
      "epoch": 0.948217664073237,
      "grad_norm": 0.14834477007389069,
      "learning_rate": 6.839635661054375e-05,
      "loss": 0.676,
      "mean_token_accuracy": 0.7989091947674751,
      "num_tokens": 328947446.0,
      "step": 10306
    },
    {
      "entropy": 0.7534959763288498,
      "epoch": 0.9483096704446783,
      "grad_norm": 0.16255152225494385,
      "learning_rate": 6.839328978440213e-05,
      "loss": 0.7619,
      "mean_token_accuracy": 0.7753321677446365,
      "num_tokens": 328979069.0,
      "step": 10307
    },
    {
      "entropy": 0.7448869049549103,
      "epoch": 0.9484016768161195,
      "grad_norm": 0.15851105749607086,
      "learning_rate": 6.83902229582605e-05,
      "loss": 0.7283,
      "mean_token_accuracy": 0.7881651520729065,
      "num_tokens": 329010987.0,
      "step": 10308
    },
    {
      "entropy": 0.666586569044739,
      "epoch": 0.9484936831875608,
      "grad_norm": 0.15496236085891724,
      "learning_rate": 6.838715613211887e-05,
      "loss": 0.6617,
      "mean_token_accuracy": 0.8076634965837002,
      "num_tokens": 329042548.0,
      "step": 10309
    },
    {
      "entropy": 0.7742556240409613,
      "epoch": 0.9485856895590019,
      "grad_norm": 0.15406490862369537,
      "learning_rate": 6.838408930597725e-05,
      "loss": 0.7638,
      "mean_token_accuracy": 0.7759901210665703,
      "num_tokens": 329074038.0,
      "step": 10310
    },
    {
      "entropy": 0.7554226946085691,
      "epoch": 0.9486776959304432,
      "grad_norm": 0.16249322891235352,
      "learning_rate": 6.838102247983563e-05,
      "loss": 0.7476,
      "mean_token_accuracy": 0.7810341082513332,
      "num_tokens": 329105042.0,
      "step": 10311
    },
    {
      "entropy": 0.747472308576107,
      "epoch": 0.9487697023018844,
      "grad_norm": 0.154374897480011,
      "learning_rate": 6.8377955653694e-05,
      "loss": 0.7413,
      "mean_token_accuracy": 0.786244984716177,
      "num_tokens": 329136481.0,
      "step": 10312
    },
    {
      "entropy": 0.8249295577406883,
      "epoch": 0.9488617086733256,
      "grad_norm": 0.15922503173351288,
      "learning_rate": 6.837488882755238e-05,
      "loss": 0.8218,
      "mean_token_accuracy": 0.7583281099796295,
      "num_tokens": 329168143.0,
      "step": 10313
    },
    {
      "entropy": 0.6936363000422716,
      "epoch": 0.9489537150447669,
      "grad_norm": 0.15313883125782013,
      "learning_rate": 6.837182200141075e-05,
      "loss": 0.6774,
      "mean_token_accuracy": 0.8020981848239899,
      "num_tokens": 329200062.0,
      "step": 10314
    },
    {
      "entropy": 0.711804574355483,
      "epoch": 0.949045721416208,
      "grad_norm": 0.1499633640050888,
      "learning_rate": 6.836875517526911e-05,
      "loss": 0.7012,
      "mean_token_accuracy": 0.7911785654723644,
      "num_tokens": 329231467.0,
      "step": 10315
    },
    {
      "entropy": 0.7636668048799038,
      "epoch": 0.9491377277876493,
      "grad_norm": 0.15395349264144897,
      "learning_rate": 6.83656883491275e-05,
      "loss": 0.764,
      "mean_token_accuracy": 0.7757800482213497,
      "num_tokens": 329263723.0,
      "step": 10316
    },
    {
      "entropy": 0.7769700158387423,
      "epoch": 0.9492297341590905,
      "grad_norm": 0.15673257410526276,
      "learning_rate": 6.836262152298588e-05,
      "loss": 0.7578,
      "mean_token_accuracy": 0.7797809429466724,
      "num_tokens": 329295839.0,
      "step": 10317
    },
    {
      "entropy": 0.7039863262325525,
      "epoch": 0.9493217405305318,
      "grad_norm": 0.1644108146429062,
      "learning_rate": 6.835955469684424e-05,
      "loss": 0.6958,
      "mean_token_accuracy": 0.7866733558475971,
      "num_tokens": 329326825.0,
      "step": 10318
    },
    {
      "entropy": 0.7463262835517526,
      "epoch": 0.949413746901973,
      "grad_norm": 0.15772227942943573,
      "learning_rate": 6.835648787070261e-05,
      "loss": 0.7329,
      "mean_token_accuracy": 0.7812688238918781,
      "num_tokens": 329358645.0,
      "step": 10319
    },
    {
      "entropy": 0.6704513747245073,
      "epoch": 0.9495057532734141,
      "grad_norm": 0.1560581922531128,
      "learning_rate": 6.835342104456098e-05,
      "loss": 0.6369,
      "mean_token_accuracy": 0.8078330755233765,
      "num_tokens": 329389703.0,
      "step": 10320
    },
    {
      "entropy": 0.8076649457216263,
      "epoch": 0.9495977596448554,
      "grad_norm": 0.1702118217945099,
      "learning_rate": 6.835035421841936e-05,
      "loss": 0.8077,
      "mean_token_accuracy": 0.7648176550865173,
      "num_tokens": 329421659.0,
      "step": 10321
    },
    {
      "entropy": 0.6783352103084326,
      "epoch": 0.9496897660162966,
      "grad_norm": 0.15081091225147247,
      "learning_rate": 6.834728739227774e-05,
      "loss": 0.6386,
      "mean_token_accuracy": 0.8096504062414169,
      "num_tokens": 329453698.0,
      "step": 10322
    },
    {
      "entropy": 0.6910475520417094,
      "epoch": 0.9497817723877379,
      "grad_norm": 0.15269412100315094,
      "learning_rate": 6.834422056613611e-05,
      "loss": 0.6633,
      "mean_token_accuracy": 0.8057943508028984,
      "num_tokens": 329486055.0,
      "step": 10323
    },
    {
      "entropy": 0.6840794142335653,
      "epoch": 0.9498737787591791,
      "grad_norm": 0.15452580153942108,
      "learning_rate": 6.834115373999448e-05,
      "loss": 0.6629,
      "mean_token_accuracy": 0.8011512719094753,
      "num_tokens": 329518356.0,
      "step": 10324
    },
    {
      "entropy": 0.7017823867499828,
      "epoch": 0.9499657851306202,
      "grad_norm": 0.15440934896469116,
      "learning_rate": 6.833808691385286e-05,
      "loss": 0.6912,
      "mean_token_accuracy": 0.7929501384496689,
      "num_tokens": 329550302.0,
      "step": 10325
    },
    {
      "entropy": 0.6907314918935299,
      "epoch": 0.9500577915020615,
      "grad_norm": 0.1529570072889328,
      "learning_rate": 6.833502008771123e-05,
      "loss": 0.6827,
      "mean_token_accuracy": 0.7938347607851028,
      "num_tokens": 329581932.0,
      "step": 10326
    },
    {
      "entropy": 0.6697320286184549,
      "epoch": 0.9501497978735027,
      "grad_norm": 0.1545184999704361,
      "learning_rate": 6.833195326156961e-05,
      "loss": 0.6656,
      "mean_token_accuracy": 0.8026415407657623,
      "num_tokens": 329613485.0,
      "step": 10327
    },
    {
      "entropy": 0.718242515809834,
      "epoch": 0.950241804244944,
      "grad_norm": 0.17164349555969238,
      "learning_rate": 6.832888643542798e-05,
      "loss": 0.7213,
      "mean_token_accuracy": 0.7849012687802315,
      "num_tokens": 329644830.0,
      "step": 10328
    },
    {
      "entropy": 0.8030125461518764,
      "epoch": 0.9503338106163852,
      "grad_norm": 0.15431763231754303,
      "learning_rate": 6.832581960928636e-05,
      "loss": 0.7851,
      "mean_token_accuracy": 0.7674397937953472,
      "num_tokens": 329676834.0,
      "step": 10329
    },
    {
      "entropy": 0.629848881624639,
      "epoch": 0.9504258169878264,
      "grad_norm": 0.1494298279285431,
      "learning_rate": 6.832275278314473e-05,
      "loss": 0.6271,
      "mean_token_accuracy": 0.8117061480879784,
      "num_tokens": 329709602.0,
      "step": 10330
    },
    {
      "entropy": 0.5959222931414843,
      "epoch": 0.9505178233592676,
      "grad_norm": 0.15927089750766754,
      "learning_rate": 6.83196859570031e-05,
      "loss": 0.583,
      "mean_token_accuracy": 0.8255954794585705,
      "num_tokens": 329741358.0,
      "step": 10331
    },
    {
      "entropy": 0.7131360387429595,
      "epoch": 0.9506098297307088,
      "grad_norm": 0.15496760606765747,
      "learning_rate": 6.831661913086148e-05,
      "loss": 0.6978,
      "mean_token_accuracy": 0.7916103638708591,
      "num_tokens": 329773700.0,
      "step": 10332
    },
    {
      "entropy": 0.6765552032738924,
      "epoch": 0.9507018361021501,
      "grad_norm": 0.15333731472492218,
      "learning_rate": 6.831355230471986e-05,
      "loss": 0.6813,
      "mean_token_accuracy": 0.7967240139842033,
      "num_tokens": 329806434.0,
      "step": 10333
    },
    {
      "entropy": 0.5991768725216389,
      "epoch": 0.9507938424735913,
      "grad_norm": 0.15126122534275055,
      "learning_rate": 6.831048547857822e-05,
      "loss": 0.5852,
      "mean_token_accuracy": 0.8215456381440163,
      "num_tokens": 329837821.0,
      "step": 10334
    },
    {
      "entropy": 0.6920499000698328,
      "epoch": 0.9508858488450325,
      "grad_norm": 0.1542934775352478,
      "learning_rate": 6.830741865243659e-05,
      "loss": 0.6846,
      "mean_token_accuracy": 0.8013885542750359,
      "num_tokens": 329869128.0,
      "step": 10335
    },
    {
      "entropy": 0.7834386443719268,
      "epoch": 0.9509778552164737,
      "grad_norm": 0.1553724706172943,
      "learning_rate": 6.830435182629496e-05,
      "loss": 0.7711,
      "mean_token_accuracy": 0.7717783525586128,
      "num_tokens": 329901315.0,
      "step": 10336
    },
    {
      "entropy": 0.751272976398468,
      "epoch": 0.951069861587915,
      "grad_norm": 0.16046187281608582,
      "learning_rate": 6.830128500015334e-05,
      "loss": 0.7444,
      "mean_token_accuracy": 0.7774026989936829,
      "num_tokens": 329933243.0,
      "step": 10337
    },
    {
      "entropy": 0.72953275218606,
      "epoch": 0.9511618679593562,
      "grad_norm": 0.15783913433551788,
      "learning_rate": 6.829821817401172e-05,
      "loss": 0.7315,
      "mean_token_accuracy": 0.779413241893053,
      "num_tokens": 329964825.0,
      "step": 10338
    },
    {
      "entropy": 0.7632519965991378,
      "epoch": 0.9512538743307974,
      "grad_norm": 0.15950921177864075,
      "learning_rate": 6.829515134787009e-05,
      "loss": 0.7403,
      "mean_token_accuracy": 0.7844377271831036,
      "num_tokens": 329996911.0,
      "step": 10339
    },
    {
      "entropy": 0.7003697939217091,
      "epoch": 0.9513458807022386,
      "grad_norm": 0.1544104665517807,
      "learning_rate": 6.829208452172846e-05,
      "loss": 0.6899,
      "mean_token_accuracy": 0.7947129681706429,
      "num_tokens": 330028137.0,
      "step": 10340
    },
    {
      "entropy": 0.699734952300787,
      "epoch": 0.9514378870736798,
      "grad_norm": 0.1592847853899002,
      "learning_rate": 6.828901769558684e-05,
      "loss": 0.6784,
      "mean_token_accuracy": 0.799206305295229,
      "num_tokens": 330060519.0,
      "step": 10341
    },
    {
      "entropy": 0.7755158264189959,
      "epoch": 0.9515298934451211,
      "grad_norm": 0.15504035353660583,
      "learning_rate": 6.828595086944521e-05,
      "loss": 0.757,
      "mean_token_accuracy": 0.7730560526251793,
      "num_tokens": 330091834.0,
      "step": 10342
    },
    {
      "entropy": 0.6983494674786925,
      "epoch": 0.9516218998165623,
      "grad_norm": 0.1663280576467514,
      "learning_rate": 6.828288404330359e-05,
      "loss": 0.6949,
      "mean_token_accuracy": 0.7930177189409733,
      "num_tokens": 330123821.0,
      "step": 10343
    },
    {
      "entropy": 0.6716302875429392,
      "epoch": 0.9517139061880036,
      "grad_norm": 0.14703315496444702,
      "learning_rate": 6.827981721716197e-05,
      "loss": 0.6578,
      "mean_token_accuracy": 0.8033426031470299,
      "num_tokens": 330155946.0,
      "step": 10344
    },
    {
      "entropy": 0.6715113278478384,
      "epoch": 0.9518059125594447,
      "grad_norm": 0.17111022770404816,
      "learning_rate": 6.827675039102034e-05,
      "loss": 0.6627,
      "mean_token_accuracy": 0.8027927540242672,
      "num_tokens": 330186146.0,
      "step": 10345
    },
    {
      "entropy": 0.7144035925157368,
      "epoch": 0.9518979189308859,
      "grad_norm": 0.15415389835834503,
      "learning_rate": 6.827368356487871e-05,
      "loss": 0.6963,
      "mean_token_accuracy": 0.7901563346385956,
      "num_tokens": 330217569.0,
      "step": 10346
    },
    {
      "entropy": 0.5822206549346447,
      "epoch": 0.9519899253023272,
      "grad_norm": 0.1554756909608841,
      "learning_rate": 6.827061673873707e-05,
      "loss": 0.5481,
      "mean_token_accuracy": 0.8331722132861614,
      "num_tokens": 330249549.0,
      "step": 10347
    },
    {
      "entropy": 0.6906355638056993,
      "epoch": 0.9520819316737684,
      "grad_norm": 0.14547686278820038,
      "learning_rate": 6.826754991259547e-05,
      "loss": 0.671,
      "mean_token_accuracy": 0.8048881441354752,
      "num_tokens": 330281272.0,
      "step": 10348
    },
    {
      "entropy": 0.5610641781240702,
      "epoch": 0.9521739380452097,
      "grad_norm": 0.14641766250133514,
      "learning_rate": 6.826448308645384e-05,
      "loss": 0.5414,
      "mean_token_accuracy": 0.8341645672917366,
      "num_tokens": 330312976.0,
      "step": 10349
    },
    {
      "entropy": 0.7535094954073429,
      "epoch": 0.9522659444166508,
      "grad_norm": 0.1585170179605484,
      "learning_rate": 6.82614162603122e-05,
      "loss": 0.7485,
      "mean_token_accuracy": 0.7790782302618027,
      "num_tokens": 330345151.0,
      "step": 10350
    },
    {
      "entropy": 0.7332879174500704,
      "epoch": 0.952357950788092,
      "grad_norm": 0.1494898796081543,
      "learning_rate": 6.825834943417057e-05,
      "loss": 0.7317,
      "mean_token_accuracy": 0.7851615622639656,
      "num_tokens": 330377260.0,
      "step": 10351
    },
    {
      "entropy": 0.7531126365065575,
      "epoch": 0.9524499571595333,
      "grad_norm": 0.1499609500169754,
      "learning_rate": 6.825528260802895e-05,
      "loss": 0.7477,
      "mean_token_accuracy": 0.7790671400725842,
      "num_tokens": 330409412.0,
      "step": 10352
    },
    {
      "entropy": 0.7712447289377451,
      "epoch": 0.9525419635309745,
      "grad_norm": 0.16035711765289307,
      "learning_rate": 6.825221578188734e-05,
      "loss": 0.7622,
      "mean_token_accuracy": 0.7748049907386303,
      "num_tokens": 330440666.0,
      "step": 10353
    },
    {
      "entropy": 0.7784079490229487,
      "epoch": 0.9526339699024158,
      "grad_norm": 0.15622101724147797,
      "learning_rate": 6.82491489557457e-05,
      "loss": 0.755,
      "mean_token_accuracy": 0.7761527560651302,
      "num_tokens": 330472824.0,
      "step": 10354
    },
    {
      "entropy": 0.7365318667143583,
      "epoch": 0.9527259762738569,
      "grad_norm": 0.15841032564640045,
      "learning_rate": 6.824608212960407e-05,
      "loss": 0.7031,
      "mean_token_accuracy": 0.7872166335582733,
      "num_tokens": 330503659.0,
      "step": 10355
    },
    {
      "entropy": 0.7545578312128782,
      "epoch": 0.9528179826452982,
      "grad_norm": 0.15412180125713348,
      "learning_rate": 6.824301530346245e-05,
      "loss": 0.7503,
      "mean_token_accuracy": 0.777005348354578,
      "num_tokens": 330534614.0,
      "step": 10356
    },
    {
      "entropy": 0.7202246319502592,
      "epoch": 0.9529099890167394,
      "grad_norm": 0.1544807404279709,
      "learning_rate": 6.823994847732082e-05,
      "loss": 0.7035,
      "mean_token_accuracy": 0.7904775477945805,
      "num_tokens": 330566358.0,
      "step": 10357
    },
    {
      "entropy": 0.6793606076389551,
      "epoch": 0.9530019953881806,
      "grad_norm": 0.15205572545528412,
      "learning_rate": 6.82368816511792e-05,
      "loss": 0.6576,
      "mean_token_accuracy": 0.8032529763877392,
      "num_tokens": 330598257.0,
      "step": 10358
    },
    {
      "entropy": 0.6657685688696802,
      "epoch": 0.9530940017596219,
      "grad_norm": 0.14847935736179352,
      "learning_rate": 6.823381482503757e-05,
      "loss": 0.6576,
      "mean_token_accuracy": 0.8071863390505314,
      "num_tokens": 330629656.0,
      "step": 10359
    },
    {
      "entropy": 0.8413185738027096,
      "epoch": 0.953186008131063,
      "grad_norm": 0.1589173525571823,
      "learning_rate": 6.823074799889595e-05,
      "loss": 0.8278,
      "mean_token_accuracy": 0.7557625398039818,
      "num_tokens": 330660471.0,
      "step": 10360
    },
    {
      "entropy": 0.5988645711913705,
      "epoch": 0.9532780145025043,
      "grad_norm": 0.15095260739326477,
      "learning_rate": 6.822768117275432e-05,
      "loss": 0.6038,
      "mean_token_accuracy": 0.8185758404433727,
      "num_tokens": 330692540.0,
      "step": 10361
    },
    {
      "entropy": 0.7170590162277222,
      "epoch": 0.9533700208739455,
      "grad_norm": 0.1654355674982071,
      "learning_rate": 6.822461434661269e-05,
      "loss": 0.717,
      "mean_token_accuracy": 0.7872217446565628,
      "num_tokens": 330724251.0,
      "step": 10362
    },
    {
      "entropy": 0.7426900770515203,
      "epoch": 0.9534620272453868,
      "grad_norm": 0.1547955721616745,
      "learning_rate": 6.822154752047107e-05,
      "loss": 0.743,
      "mean_token_accuracy": 0.7797574996948242,
      "num_tokens": 330756705.0,
      "step": 10363
    },
    {
      "entropy": 0.7068341113626957,
      "epoch": 0.953554033616828,
      "grad_norm": 0.15573282539844513,
      "learning_rate": 6.821848069432945e-05,
      "loss": 0.6923,
      "mean_token_accuracy": 0.7947261370718479,
      "num_tokens": 330788313.0,
      "step": 10364
    },
    {
      "entropy": 0.6831525526940823,
      "epoch": 0.9536460399882691,
      "grad_norm": 0.15726906061172485,
      "learning_rate": 6.821541386818782e-05,
      "loss": 0.6689,
      "mean_token_accuracy": 0.7964128144085407,
      "num_tokens": 330819628.0,
      "step": 10365
    },
    {
      "entropy": 0.7024758234620094,
      "epoch": 0.9537380463597104,
      "grad_norm": 0.1533324271440506,
      "learning_rate": 6.821234704204619e-05,
      "loss": 0.6983,
      "mean_token_accuracy": 0.794379822909832,
      "num_tokens": 330850323.0,
      "step": 10366
    },
    {
      "entropy": 0.7304856879636645,
      "epoch": 0.9538300527311516,
      "grad_norm": 0.1612171232700348,
      "learning_rate": 6.820928021590455e-05,
      "loss": 0.7181,
      "mean_token_accuracy": 0.7858642600476742,
      "num_tokens": 330882397.0,
      "step": 10367
    },
    {
      "entropy": 0.7194500174373388,
      "epoch": 0.9539220591025929,
      "grad_norm": 0.16906310617923737,
      "learning_rate": 6.820621338976294e-05,
      "loss": 0.7083,
      "mean_token_accuracy": 0.79090765863657,
      "num_tokens": 330914400.0,
      "step": 10368
    },
    {
      "entropy": 0.7798740491271019,
      "epoch": 0.9540140654740341,
      "grad_norm": 0.16032767295837402,
      "learning_rate": 6.820314656362132e-05,
      "loss": 0.7843,
      "mean_token_accuracy": 0.770420353859663,
      "num_tokens": 330946518.0,
      "step": 10369
    },
    {
      "entropy": 0.7835494428873062,
      "epoch": 0.9541060718454752,
      "grad_norm": 0.15782110393047333,
      "learning_rate": 6.820007973747968e-05,
      "loss": 0.7558,
      "mean_token_accuracy": 0.7751386798918247,
      "num_tokens": 330978750.0,
      "step": 10370
    },
    {
      "entropy": 0.7999539375305176,
      "epoch": 0.9541980782169165,
      "grad_norm": 0.16153496503829956,
      "learning_rate": 6.819701291133805e-05,
      "loss": 0.7845,
      "mean_token_accuracy": 0.7636932954192162,
      "num_tokens": 331010946.0,
      "step": 10371
    },
    {
      "entropy": 0.720542473718524,
      "epoch": 0.9542900845883577,
      "grad_norm": 0.16235938668251038,
      "learning_rate": 6.819394608519643e-05,
      "loss": 0.7214,
      "mean_token_accuracy": 0.7825980111956596,
      "num_tokens": 331042486.0,
      "step": 10372
    },
    {
      "entropy": 0.687037056311965,
      "epoch": 0.954382090959799,
      "grad_norm": 0.16294613480567932,
      "learning_rate": 6.81908792590548e-05,
      "loss": 0.653,
      "mean_token_accuracy": 0.8043670579791069,
      "num_tokens": 331073518.0,
      "step": 10373
    },
    {
      "entropy": 0.6109270863234997,
      "epoch": 0.9544740973312402,
      "grad_norm": 0.14858472347259521,
      "learning_rate": 6.818781243291318e-05,
      "loss": 0.6053,
      "mean_token_accuracy": 0.8198653161525726,
      "num_tokens": 331105482.0,
      "step": 10374
    },
    {
      "entropy": 0.7381099797785282,
      "epoch": 0.9545661037026814,
      "grad_norm": 0.15823811292648315,
      "learning_rate": 6.818474560677156e-05,
      "loss": 0.734,
      "mean_token_accuracy": 0.7846692986786366,
      "num_tokens": 331137739.0,
      "step": 10375
    },
    {
      "entropy": 0.739355331286788,
      "epoch": 0.9546581100741226,
      "grad_norm": 0.1624433547258377,
      "learning_rate": 6.818167878062993e-05,
      "loss": 0.7325,
      "mean_token_accuracy": 0.7864391021430492,
      "num_tokens": 331169520.0,
      "step": 10376
    },
    {
      "entropy": 0.7441974394023418,
      "epoch": 0.9547501164455638,
      "grad_norm": 0.15793685615062714,
      "learning_rate": 6.81786119544883e-05,
      "loss": 0.7435,
      "mean_token_accuracy": 0.7881110832095146,
      "num_tokens": 331201539.0,
      "step": 10377
    },
    {
      "entropy": 0.6673200316727161,
      "epoch": 0.9548421228170051,
      "grad_norm": 0.14708095788955688,
      "learning_rate": 6.817554512834667e-05,
      "loss": 0.6446,
      "mean_token_accuracy": 0.8022946082055569,
      "num_tokens": 331232662.0,
      "step": 10378
    },
    {
      "entropy": 0.7148380745202303,
      "epoch": 0.9549341291884463,
      "grad_norm": 0.16837042570114136,
      "learning_rate": 6.817247830220505e-05,
      "loss": 0.7191,
      "mean_token_accuracy": 0.7895751893520355,
      "num_tokens": 331264172.0,
      "step": 10379
    },
    {
      "entropy": 0.6926519572734833,
      "epoch": 0.9550261355598875,
      "grad_norm": 0.14734429121017456,
      "learning_rate": 6.816941147606343e-05,
      "loss": 0.6904,
      "mean_token_accuracy": 0.7962998487055302,
      "num_tokens": 331296645.0,
      "step": 10380
    },
    {
      "entropy": 0.6049151439219713,
      "epoch": 0.9551181419313287,
      "grad_norm": 0.14170698821544647,
      "learning_rate": 6.81663446499218e-05,
      "loss": 0.5888,
      "mean_token_accuracy": 0.824204184114933,
      "num_tokens": 331328779.0,
      "step": 10381
    },
    {
      "entropy": 0.6051322063431144,
      "epoch": 0.95521014830277,
      "grad_norm": 0.16112157702445984,
      "learning_rate": 6.816327782378017e-05,
      "loss": 0.5763,
      "mean_token_accuracy": 0.8225452862679958,
      "num_tokens": 331360472.0,
      "step": 10382
    },
    {
      "entropy": 0.6930302958935499,
      "epoch": 0.9553021546742112,
      "grad_norm": 0.15064461529254913,
      "learning_rate": 6.816021099763855e-05,
      "loss": 0.6908,
      "mean_token_accuracy": 0.7972472757101059,
      "num_tokens": 331392625.0,
      "step": 10383
    },
    {
      "entropy": 0.7574808867648244,
      "epoch": 0.9553941610456524,
      "grad_norm": 0.14804288744926453,
      "learning_rate": 6.815714417149692e-05,
      "loss": 0.7498,
      "mean_token_accuracy": 0.7766664922237396,
      "num_tokens": 331424942.0,
      "step": 10384
    },
    {
      "entropy": 0.7384954579174519,
      "epoch": 0.9554861674170936,
      "grad_norm": 0.15282543003559113,
      "learning_rate": 6.81540773453553e-05,
      "loss": 0.7183,
      "mean_token_accuracy": 0.7845971398055553,
      "num_tokens": 331457123.0,
      "step": 10385
    },
    {
      "entropy": 0.7885136064141989,
      "epoch": 0.9555781737885348,
      "grad_norm": 0.162718266248703,
      "learning_rate": 6.815101051921367e-05,
      "loss": 0.7665,
      "mean_token_accuracy": 0.7750266641378403,
      "num_tokens": 331488590.0,
      "step": 10386
    },
    {
      "entropy": 0.7341670067980886,
      "epoch": 0.9556701801599761,
      "grad_norm": 0.1648067831993103,
      "learning_rate": 6.814794369307205e-05,
      "loss": 0.7344,
      "mean_token_accuracy": 0.7846819348633289,
      "num_tokens": 331520471.0,
      "step": 10387
    },
    {
      "entropy": 0.7197143957018852,
      "epoch": 0.9557621865314173,
      "grad_norm": 0.15455861389636993,
      "learning_rate": 6.814487686693041e-05,
      "loss": 0.6978,
      "mean_token_accuracy": 0.7926844172179699,
      "num_tokens": 331553239.0,
      "step": 10388
    },
    {
      "entropy": 0.6906748618930578,
      "epoch": 0.9558541929028586,
      "grad_norm": 0.1478624939918518,
      "learning_rate": 6.814181004078878e-05,
      "loss": 0.6789,
      "mean_token_accuracy": 0.7994207590818405,
      "num_tokens": 331585543.0,
      "step": 10389
    },
    {
      "entropy": 0.6510620024055243,
      "epoch": 0.9559461992742997,
      "grad_norm": 0.15086205303668976,
      "learning_rate": 6.813874321464716e-05,
      "loss": 0.6364,
      "mean_token_accuracy": 0.8057578913867474,
      "num_tokens": 331616406.0,
      "step": 10390
    },
    {
      "entropy": 0.6300481921061873,
      "epoch": 0.9560382056457409,
      "grad_norm": 0.14670509099960327,
      "learning_rate": 6.813567638850555e-05,
      "loss": 0.5895,
      "mean_token_accuracy": 0.8223542906343937,
      "num_tokens": 331648611.0,
      "step": 10391
    },
    {
      "entropy": 0.805447407066822,
      "epoch": 0.9561302120171822,
      "grad_norm": 0.15985581278800964,
      "learning_rate": 6.813260956236391e-05,
      "loss": 0.8057,
      "mean_token_accuracy": 0.7615090124309063,
      "num_tokens": 331680226.0,
      "step": 10392
    },
    {
      "entropy": 0.7411986812949181,
      "epoch": 0.9562222183886234,
      "grad_norm": 0.1552794724702835,
      "learning_rate": 6.812954273622228e-05,
      "loss": 0.7273,
      "mean_token_accuracy": 0.7850522138178349,
      "num_tokens": 331711885.0,
      "step": 10393
    },
    {
      "entropy": 0.7745989188551903,
      "epoch": 0.9563142247600647,
      "grad_norm": 0.167907252907753,
      "learning_rate": 6.812647591008065e-05,
      "loss": 0.7616,
      "mean_token_accuracy": 0.7739268839359283,
      "num_tokens": 331743836.0,
      "step": 10394
    },
    {
      "entropy": 0.7046617604792118,
      "epoch": 0.9564062311315058,
      "grad_norm": 0.15686923265457153,
      "learning_rate": 6.812340908393904e-05,
      "loss": 0.6926,
      "mean_token_accuracy": 0.7946471460163593,
      "num_tokens": 331776090.0,
      "step": 10395
    },
    {
      "entropy": 0.6639353046193719,
      "epoch": 0.956498237502947,
      "grad_norm": 0.16695290803909302,
      "learning_rate": 6.812034225779741e-05,
      "loss": 0.6543,
      "mean_token_accuracy": 0.8067062087357044,
      "num_tokens": 331808302.0,
      "step": 10396
    },
    {
      "entropy": 0.6614685468375683,
      "epoch": 0.9565902438743883,
      "grad_norm": 0.1518082320690155,
      "learning_rate": 6.811727543165578e-05,
      "loss": 0.6507,
      "mean_token_accuracy": 0.8040320798754692,
      "num_tokens": 331840909.0,
      "step": 10397
    },
    {
      "entropy": 0.6607112372294068,
      "epoch": 0.9566822502458295,
      "grad_norm": 0.14555080235004425,
      "learning_rate": 6.811420860551415e-05,
      "loss": 0.6531,
      "mean_token_accuracy": 0.8050027564167976,
      "num_tokens": 331872952.0,
      "step": 10398
    },
    {
      "entropy": 0.7946364339441061,
      "epoch": 0.9567742566172708,
      "grad_norm": 0.1572858989238739,
      "learning_rate": 6.811114177937253e-05,
      "loss": 0.7981,
      "mean_token_accuracy": 0.7643549516797066,
      "num_tokens": 331905109.0,
      "step": 10399
    },
    {
      "entropy": 0.6281259199604392,
      "epoch": 0.9568662629887119,
      "grad_norm": 0.1459970921278,
      "learning_rate": 6.810807495323091e-05,
      "loss": 0.6115,
      "mean_token_accuracy": 0.8164280354976654,
      "num_tokens": 331936399.0,
      "step": 10400
    },
    {
      "entropy": 0.6092888973653316,
      "epoch": 0.9569582693601532,
      "grad_norm": 0.15422403812408447,
      "learning_rate": 6.810500812708928e-05,
      "loss": 0.6034,
      "mean_token_accuracy": 0.820247009396553,
      "num_tokens": 331967757.0,
      "step": 10401
    },
    {
      "entropy": 0.7659487500786781,
      "epoch": 0.9570502757315944,
      "grad_norm": 0.1627964824438095,
      "learning_rate": 6.810194130094765e-05,
      "loss": 0.7881,
      "mean_token_accuracy": 0.7730962820351124,
      "num_tokens": 331998907.0,
      "step": 10402
    },
    {
      "entropy": 0.6827439526095986,
      "epoch": 0.9571422821030356,
      "grad_norm": 0.15675345063209534,
      "learning_rate": 6.809887447480603e-05,
      "loss": 0.6569,
      "mean_token_accuracy": 0.8044237121939659,
      "num_tokens": 332030623.0,
      "step": 10403
    },
    {
      "entropy": 0.7115828655660152,
      "epoch": 0.9572342884744769,
      "grad_norm": 0.15495802462100983,
      "learning_rate": 6.80958076486644e-05,
      "loss": 0.7083,
      "mean_token_accuracy": 0.7914004512131214,
      "num_tokens": 332063261.0,
      "step": 10404
    },
    {
      "entropy": 0.776873804628849,
      "epoch": 0.957326294845918,
      "grad_norm": 0.15778258442878723,
      "learning_rate": 6.809274082252278e-05,
      "loss": 0.7605,
      "mean_token_accuracy": 0.7788488157093525,
      "num_tokens": 332094767.0,
      "step": 10405
    },
    {
      "entropy": 0.7298601940274239,
      "epoch": 0.9574183012173593,
      "grad_norm": 0.15925809741020203,
      "learning_rate": 6.808967399638116e-05,
      "loss": 0.7103,
      "mean_token_accuracy": 0.7893152721226215,
      "num_tokens": 332127269.0,
      "step": 10406
    },
    {
      "entropy": 0.6768114045262337,
      "epoch": 0.9575103075888005,
      "grad_norm": 0.1655510663986206,
      "learning_rate": 6.808660717023953e-05,
      "loss": 0.6667,
      "mean_token_accuracy": 0.8067215792834759,
      "num_tokens": 332158612.0,
      "step": 10407
    },
    {
      "entropy": 0.6903613340109587,
      "epoch": 0.9576023139602418,
      "grad_norm": 0.15687842667102814,
      "learning_rate": 6.80835403440979e-05,
      "loss": 0.6745,
      "mean_token_accuracy": 0.8008562661707401,
      "num_tokens": 332190577.0,
      "step": 10408
    },
    {
      "entropy": 0.5958782425150275,
      "epoch": 0.957694320331683,
      "grad_norm": 0.14586122334003448,
      "learning_rate": 6.808047351795626e-05,
      "loss": 0.5837,
      "mean_token_accuracy": 0.8267413713037968,
      "num_tokens": 332222689.0,
      "step": 10409
    },
    {
      "entropy": 0.615108042024076,
      "epoch": 0.9577863267031242,
      "grad_norm": 0.15335136651992798,
      "learning_rate": 6.807740669181464e-05,
      "loss": 0.5906,
      "mean_token_accuracy": 0.8237563110888004,
      "num_tokens": 332254877.0,
      "step": 10410
    },
    {
      "entropy": 0.7984740901738405,
      "epoch": 0.9578783330745654,
      "grad_norm": 0.16249743103981018,
      "learning_rate": 6.807433986567303e-05,
      "loss": 0.7843,
      "mean_token_accuracy": 0.7708918713033199,
      "num_tokens": 332286589.0,
      "step": 10411
    },
    {
      "entropy": 0.7338308542966843,
      "epoch": 0.9579703394460066,
      "grad_norm": 0.15629039704799652,
      "learning_rate": 6.807127303953139e-05,
      "loss": 0.731,
      "mean_token_accuracy": 0.7853407189249992,
      "num_tokens": 332318788.0,
      "step": 10412
    },
    {
      "entropy": 0.565003227442503,
      "epoch": 0.9580623458174479,
      "grad_norm": 0.15016482770442963,
      "learning_rate": 6.806820621338976e-05,
      "loss": 0.5496,
      "mean_token_accuracy": 0.8307982720434666,
      "num_tokens": 332350863.0,
      "step": 10413
    },
    {
      "entropy": 0.7665461823344231,
      "epoch": 0.9581543521888891,
      "grad_norm": 0.15949726104736328,
      "learning_rate": 6.806513938724814e-05,
      "loss": 0.7501,
      "mean_token_accuracy": 0.7824391908943653,
      "num_tokens": 332382336.0,
      "step": 10414
    },
    {
      "entropy": 0.7284893468022346,
      "epoch": 0.9582463585603304,
      "grad_norm": 0.15831369161605835,
      "learning_rate": 6.806207256110651e-05,
      "loss": 0.7211,
      "mean_token_accuracy": 0.7870792783796787,
      "num_tokens": 332414933.0,
      "step": 10415
    },
    {
      "entropy": 0.7411925792694092,
      "epoch": 0.9583383649317715,
      "grad_norm": 0.1565663069486618,
      "learning_rate": 6.805900573496489e-05,
      "loss": 0.7145,
      "mean_token_accuracy": 0.7858598716557026,
      "num_tokens": 332445947.0,
      "step": 10416
    },
    {
      "entropy": 0.7581945564597845,
      "epoch": 0.9584303713032127,
      "grad_norm": 0.16065749526023865,
      "learning_rate": 6.805593890882326e-05,
      "loss": 0.7283,
      "mean_token_accuracy": 0.7843787483870983,
      "num_tokens": 332477643.0,
      "step": 10417
    },
    {
      "entropy": 0.581575183197856,
      "epoch": 0.958522377674654,
      "grad_norm": 0.1426018923521042,
      "learning_rate": 6.805287208268164e-05,
      "loss": 0.5639,
      "mean_token_accuracy": 0.8259046338498592,
      "num_tokens": 332509291.0,
      "step": 10418
    },
    {
      "entropy": 0.79386773891747,
      "epoch": 0.9586143840460952,
      "grad_norm": 0.152455136179924,
      "learning_rate": 6.804980525654001e-05,
      "loss": 0.7889,
      "mean_token_accuracy": 0.7661974132061005,
      "num_tokens": 332540783.0,
      "step": 10419
    },
    {
      "entropy": 0.6863946178928018,
      "epoch": 0.9587063904175365,
      "grad_norm": 0.15365168452262878,
      "learning_rate": 6.804673843039838e-05,
      "loss": 0.6698,
      "mean_token_accuracy": 0.8025994226336479,
      "num_tokens": 332572731.0,
      "step": 10420
    },
    {
      "entropy": 0.6888884101063013,
      "epoch": 0.9587983967889776,
      "grad_norm": 0.1536545604467392,
      "learning_rate": 6.804367160425676e-05,
      "loss": 0.6787,
      "mean_token_accuracy": 0.7975575514137745,
      "num_tokens": 332604878.0,
      "step": 10421
    },
    {
      "entropy": 0.7861401326954365,
      "epoch": 0.9588904031604188,
      "grad_norm": 0.15668457746505737,
      "learning_rate": 6.804060477811514e-05,
      "loss": 0.7828,
      "mean_token_accuracy": 0.7684304490685463,
      "num_tokens": 332636410.0,
      "step": 10422
    },
    {
      "entropy": 0.7559653669595718,
      "epoch": 0.9589824095318601,
      "grad_norm": 0.157333105802536,
      "learning_rate": 6.803753795197351e-05,
      "loss": 0.7536,
      "mean_token_accuracy": 0.777506485581398,
      "num_tokens": 332667583.0,
      "step": 10423
    },
    {
      "entropy": 0.72782345674932,
      "epoch": 0.9590744159033013,
      "grad_norm": 0.15118083357810974,
      "learning_rate": 6.803447112583188e-05,
      "loss": 0.7152,
      "mean_token_accuracy": 0.7906947396695614,
      "num_tokens": 332698780.0,
      "step": 10424
    },
    {
      "entropy": 0.7501499224454165,
      "epoch": 0.9591664222747426,
      "grad_norm": 0.15752235054969788,
      "learning_rate": 6.803140429969024e-05,
      "loss": 0.7405,
      "mean_token_accuracy": 0.7784449532628059,
      "num_tokens": 332730944.0,
      "step": 10425
    },
    {
      "entropy": 0.7977633578702807,
      "epoch": 0.9592584286461837,
      "grad_norm": 0.15883994102478027,
      "learning_rate": 6.802833747354862e-05,
      "loss": 0.8016,
      "mean_token_accuracy": 0.7701890729367733,
      "num_tokens": 332762971.0,
      "step": 10426
    },
    {
      "entropy": 0.6104874247685075,
      "epoch": 0.959350435017625,
      "grad_norm": 0.1530519425868988,
      "learning_rate": 6.8025270647407e-05,
      "loss": 0.6047,
      "mean_token_accuracy": 0.8181096278131008,
      "num_tokens": 332795422.0,
      "step": 10427
    },
    {
      "entropy": 0.7724782451987267,
      "epoch": 0.9594424413890662,
      "grad_norm": 0.15339310467243195,
      "learning_rate": 6.802220382126537e-05,
      "loss": 0.7635,
      "mean_token_accuracy": 0.7774289511144161,
      "num_tokens": 332826653.0,
      "step": 10428
    },
    {
      "entropy": 0.5871471241116524,
      "epoch": 0.9595344477605074,
      "grad_norm": 0.1512991040945053,
      "learning_rate": 6.801913699512374e-05,
      "loss": 0.5773,
      "mean_token_accuracy": 0.8248698525130749,
      "num_tokens": 332858534.0,
      "step": 10429
    },
    {
      "entropy": 0.7078781519085169,
      "epoch": 0.9596264541319487,
      "grad_norm": 0.16004393994808197,
      "learning_rate": 6.801607016898212e-05,
      "loss": 0.6915,
      "mean_token_accuracy": 0.7923189625144005,
      "num_tokens": 332890323.0,
      "step": 10430
    },
    {
      "entropy": 0.6549162659794092,
      "epoch": 0.9597184605033898,
      "grad_norm": 0.16350652277469635,
      "learning_rate": 6.80130033428405e-05,
      "loss": 0.641,
      "mean_token_accuracy": 0.8091638162732124,
      "num_tokens": 332921677.0,
      "step": 10431
    },
    {
      "entropy": 0.7088744603097439,
      "epoch": 0.9598104668748311,
      "grad_norm": 0.15665943920612335,
      "learning_rate": 6.800993651669887e-05,
      "loss": 0.7025,
      "mean_token_accuracy": 0.7882219031453133,
      "num_tokens": 332954044.0,
      "step": 10432
    },
    {
      "entropy": 0.6643048394471407,
      "epoch": 0.9599024732462723,
      "grad_norm": 0.1399981528520584,
      "learning_rate": 6.800686969055725e-05,
      "loss": 0.6277,
      "mean_token_accuracy": 0.8100103288888931,
      "num_tokens": 332986037.0,
      "step": 10433
    },
    {
      "entropy": 0.6828327290713787,
      "epoch": 0.9599944796177136,
      "grad_norm": 0.1556127965450287,
      "learning_rate": 6.800380286441562e-05,
      "loss": 0.6529,
      "mean_token_accuracy": 0.8025740087032318,
      "num_tokens": 333017697.0,
      "step": 10434
    },
    {
      "entropy": 0.6876280698925257,
      "epoch": 0.9600864859891548,
      "grad_norm": 0.14695671200752258,
      "learning_rate": 6.800073603827399e-05,
      "loss": 0.6682,
      "mean_token_accuracy": 0.79716607183218,
      "num_tokens": 333050015.0,
      "step": 10435
    },
    {
      "entropy": 0.722772516310215,
      "epoch": 0.9601784923605959,
      "grad_norm": 0.15421341359615326,
      "learning_rate": 6.799766921213237e-05,
      "loss": 0.7003,
      "mean_token_accuracy": 0.7907210811972618,
      "num_tokens": 333082250.0,
      "step": 10436
    },
    {
      "entropy": 0.7943957708775997,
      "epoch": 0.9602704987320372,
      "grad_norm": 0.1545526534318924,
      "learning_rate": 6.799460238599075e-05,
      "loss": 0.8002,
      "mean_token_accuracy": 0.7659740000963211,
      "num_tokens": 333114301.0,
      "step": 10437
    },
    {
      "entropy": 0.677141573280096,
      "epoch": 0.9603625051034784,
      "grad_norm": 0.16246284544467926,
      "learning_rate": 6.799153555984912e-05,
      "loss": 0.6625,
      "mean_token_accuracy": 0.8043157421052456,
      "num_tokens": 333146283.0,
      "step": 10438
    },
    {
      "entropy": 0.6915840785950422,
      "epoch": 0.9604545114749197,
      "grad_norm": 0.15433791279792786,
      "learning_rate": 6.798846873370749e-05,
      "loss": 0.6775,
      "mean_token_accuracy": 0.7966926023364067,
      "num_tokens": 333176660.0,
      "step": 10439
    },
    {
      "entropy": 0.7411697376519442,
      "epoch": 0.9605465178463609,
      "grad_norm": 0.1519503891468048,
      "learning_rate": 6.798540190756586e-05,
      "loss": 0.7219,
      "mean_token_accuracy": 0.7849065884947777,
      "num_tokens": 333209179.0,
      "step": 10440
    },
    {
      "entropy": 0.6435529552400112,
      "epoch": 0.960638524217802,
      "grad_norm": 0.1507902294397354,
      "learning_rate": 6.798233508142424e-05,
      "loss": 0.6269,
      "mean_token_accuracy": 0.8134612999856472,
      "num_tokens": 333241838.0,
      "step": 10441
    },
    {
      "entropy": 0.6608381439000368,
      "epoch": 0.9607305305892433,
      "grad_norm": 0.14766854047775269,
      "learning_rate": 6.797926825528262e-05,
      "loss": 0.6562,
      "mean_token_accuracy": 0.8034663610160351,
      "num_tokens": 333273871.0,
      "step": 10442
    },
    {
      "entropy": 0.7911554463207722,
      "epoch": 0.9608225369606845,
      "grad_norm": 0.17438198626041412,
      "learning_rate": 6.797620142914099e-05,
      "loss": 0.8096,
      "mean_token_accuracy": 0.7595423609018326,
      "num_tokens": 333305056.0,
      "step": 10443
    },
    {
      "entropy": 0.6451566787436604,
      "epoch": 0.9609145433321258,
      "grad_norm": 0.1651897430419922,
      "learning_rate": 6.797313460299935e-05,
      "loss": 0.619,
      "mean_token_accuracy": 0.811078991740942,
      "num_tokens": 333336013.0,
      "step": 10444
    },
    {
      "entropy": 0.6484313337132335,
      "epoch": 0.961006549703567,
      "grad_norm": 0.14922724664211273,
      "learning_rate": 6.797006777685774e-05,
      "loss": 0.6326,
      "mean_token_accuracy": 0.812554620206356,
      "num_tokens": 333368530.0,
      "step": 10445
    },
    {
      "entropy": 0.677259162068367,
      "epoch": 0.9610985560750082,
      "grad_norm": 0.15088209509849548,
      "learning_rate": 6.79670009507161e-05,
      "loss": 0.6688,
      "mean_token_accuracy": 0.8069727644324303,
      "num_tokens": 333399776.0,
      "step": 10446
    },
    {
      "entropy": 0.7230394370853901,
      "epoch": 0.9611905624464494,
      "grad_norm": 0.15892080962657928,
      "learning_rate": 6.796393412457449e-05,
      "loss": 0.7206,
      "mean_token_accuracy": 0.7893907316029072,
      "num_tokens": 333432421.0,
      "step": 10447
    },
    {
      "entropy": 0.7247852198779583,
      "epoch": 0.9612825688178906,
      "grad_norm": 0.15304149687290192,
      "learning_rate": 6.796086729843285e-05,
      "loss": 0.7032,
      "mean_token_accuracy": 0.789087351411581,
      "num_tokens": 333464421.0,
      "step": 10448
    },
    {
      "entropy": 0.7091844407841563,
      "epoch": 0.9613745751893319,
      "grad_norm": 0.14794373512268066,
      "learning_rate": 6.795780047229123e-05,
      "loss": 0.6895,
      "mean_token_accuracy": 0.7961963899433613,
      "num_tokens": 333497036.0,
      "step": 10449
    },
    {
      "entropy": 0.8307904303073883,
      "epoch": 0.9614665815607731,
      "grad_norm": 0.16876192390918732,
      "learning_rate": 6.79547336461496e-05,
      "loss": 0.8162,
      "mean_token_accuracy": 0.7629802115261555,
      "num_tokens": 333527936.0,
      "step": 10450
    },
    {
      "entropy": 0.61202597245574,
      "epoch": 0.9615585879322143,
      "grad_norm": 0.15299445390701294,
      "learning_rate": 6.795166682000797e-05,
      "loss": 0.5922,
      "mean_token_accuracy": 0.8218981325626373,
      "num_tokens": 333559559.0,
      "step": 10451
    },
    {
      "entropy": 0.6905034258961678,
      "epoch": 0.9616505943036555,
      "grad_norm": 0.14619655907154083,
      "learning_rate": 6.794859999386635e-05,
      "loss": 0.6845,
      "mean_token_accuracy": 0.7963788509368896,
      "num_tokens": 333592327.0,
      "step": 10452
    },
    {
      "entropy": 0.6509590670466423,
      "epoch": 0.9617426006750968,
      "grad_norm": 0.14761729538440704,
      "learning_rate": 6.794553316772473e-05,
      "loss": 0.6445,
      "mean_token_accuracy": 0.8057989962399006,
      "num_tokens": 333624612.0,
      "step": 10453
    },
    {
      "entropy": 0.7407727334648371,
      "epoch": 0.961834607046538,
      "grad_norm": 0.15885180234909058,
      "learning_rate": 6.79424663415831e-05,
      "loss": 0.7374,
      "mean_token_accuracy": 0.7923368662595749,
      "num_tokens": 333656873.0,
      "step": 10454
    },
    {
      "entropy": 0.7330195922404528,
      "epoch": 0.9619266134179792,
      "grad_norm": 0.15252956748008728,
      "learning_rate": 6.793939951544147e-05,
      "loss": 0.745,
      "mean_token_accuracy": 0.7784949913620949,
      "num_tokens": 333689106.0,
      "step": 10455
    },
    {
      "entropy": 0.8433579467236996,
      "epoch": 0.9620186197894204,
      "grad_norm": 0.15758295357227325,
      "learning_rate": 6.793633268929984e-05,
      "loss": 0.835,
      "mean_token_accuracy": 0.7578773461282253,
      "num_tokens": 333720586.0,
      "step": 10456
    },
    {
      "entropy": 0.6438830140978098,
      "epoch": 0.9621106261608616,
      "grad_norm": 0.15245388448238373,
      "learning_rate": 6.793326586315822e-05,
      "loss": 0.6187,
      "mean_token_accuracy": 0.8142603449523449,
      "num_tokens": 333751605.0,
      "step": 10457
    },
    {
      "entropy": 0.5749893975444138,
      "epoch": 0.9622026325323029,
      "grad_norm": 0.14912867546081543,
      "learning_rate": 6.79301990370166e-05,
      "loss": 0.5528,
      "mean_token_accuracy": 0.8318054974079132,
      "num_tokens": 333783178.0,
      "step": 10458
    },
    {
      "entropy": 0.8241154234856367,
      "epoch": 0.9622946389037441,
      "grad_norm": 0.16769270598888397,
      "learning_rate": 6.792713221087497e-05,
      "loss": 0.8144,
      "mean_token_accuracy": 0.7648150883615017,
      "num_tokens": 333815392.0,
      "step": 10459
    },
    {
      "entropy": 0.6926734913140535,
      "epoch": 0.9623866452751854,
      "grad_norm": 0.16429536044597626,
      "learning_rate": 6.792406538473334e-05,
      "loss": 0.6688,
      "mean_token_accuracy": 0.8050403967499733,
      "num_tokens": 333847564.0,
      "step": 10460
    },
    {
      "entropy": 0.7573100328445435,
      "epoch": 0.9624786516466265,
      "grad_norm": 0.16124336421489716,
      "learning_rate": 6.792099855859172e-05,
      "loss": 0.7445,
      "mean_token_accuracy": 0.7816483192145824,
      "num_tokens": 333878429.0,
      "step": 10461
    },
    {
      "entropy": 0.7634071614593267,
      "epoch": 0.9625706580180677,
      "grad_norm": 0.1559850424528122,
      "learning_rate": 6.791793173245008e-05,
      "loss": 0.7672,
      "mean_token_accuracy": 0.7726938351988792,
      "num_tokens": 333910147.0,
      "step": 10462
    },
    {
      "entropy": 0.7460916619747877,
      "epoch": 0.962662664389509,
      "grad_norm": 0.14970946311950684,
      "learning_rate": 6.791486490630847e-05,
      "loss": 0.7361,
      "mean_token_accuracy": 0.7832610569894314,
      "num_tokens": 333942204.0,
      "step": 10463
    },
    {
      "entropy": 0.6872613839805126,
      "epoch": 0.9627546707609502,
      "grad_norm": 0.1590845137834549,
      "learning_rate": 6.791179808016685e-05,
      "loss": 0.6706,
      "mean_token_accuracy": 0.8016099818050861,
      "num_tokens": 333973635.0,
      "step": 10464
    },
    {
      "entropy": 0.689918352290988,
      "epoch": 0.9628466771323915,
      "grad_norm": 0.15602441132068634,
      "learning_rate": 6.790873125402522e-05,
      "loss": 0.6797,
      "mean_token_accuracy": 0.8007300086319447,
      "num_tokens": 334004586.0,
      "step": 10465
    },
    {
      "entropy": 0.6297848941758275,
      "epoch": 0.9629386835038326,
      "grad_norm": 0.15350694954395294,
      "learning_rate": 6.790566442788358e-05,
      "loss": 0.6207,
      "mean_token_accuracy": 0.8169109933078289,
      "num_tokens": 334036017.0,
      "step": 10466
    },
    {
      "entropy": 0.6790546178817749,
      "epoch": 0.9630306898752738,
      "grad_norm": 0.14826563000679016,
      "learning_rate": 6.790259760174195e-05,
      "loss": 0.6736,
      "mean_token_accuracy": 0.8005842305719852,
      "num_tokens": 334068468.0,
      "step": 10467
    },
    {
      "entropy": 0.763245502486825,
      "epoch": 0.9631226962467151,
      "grad_norm": 0.15922261774539948,
      "learning_rate": 6.789953077560035e-05,
      "loss": 0.7532,
      "mean_token_accuracy": 0.7740016654133797,
      "num_tokens": 334100530.0,
      "step": 10468
    },
    {
      "entropy": 0.6301877871155739,
      "epoch": 0.9632147026181563,
      "grad_norm": 0.15146134793758392,
      "learning_rate": 6.789646394945871e-05,
      "loss": 0.6173,
      "mean_token_accuracy": 0.8153077512979507,
      "num_tokens": 334132988.0,
      "step": 10469
    },
    {
      "entropy": 0.6215807218104601,
      "epoch": 0.9633067089895976,
      "grad_norm": 0.1521523892879486,
      "learning_rate": 6.789339712331708e-05,
      "loss": 0.6056,
      "mean_token_accuracy": 0.8174677565693855,
      "num_tokens": 334164803.0,
      "step": 10470
    },
    {
      "entropy": 0.7652538754045963,
      "epoch": 0.9633987153610387,
      "grad_norm": 0.159673273563385,
      "learning_rate": 6.789033029717545e-05,
      "loss": 0.7563,
      "mean_token_accuracy": 0.7800348252058029,
      "num_tokens": 334196473.0,
      "step": 10471
    },
    {
      "entropy": 0.7872004620730877,
      "epoch": 0.96349072173248,
      "grad_norm": 0.16594651341438293,
      "learning_rate": 6.788726347103383e-05,
      "loss": 0.7896,
      "mean_token_accuracy": 0.7692038491368294,
      "num_tokens": 334227811.0,
      "step": 10472
    },
    {
      "entropy": 0.7559331152588129,
      "epoch": 0.9635827281039212,
      "grad_norm": 0.16109560430049896,
      "learning_rate": 6.788419664489221e-05,
      "loss": 0.7386,
      "mean_token_accuracy": 0.7833860367536545,
      "num_tokens": 334259663.0,
      "step": 10473
    },
    {
      "entropy": 0.7351145297288895,
      "epoch": 0.9636747344753624,
      "grad_norm": 0.15827137231826782,
      "learning_rate": 6.788112981875058e-05,
      "loss": 0.7137,
      "mean_token_accuracy": 0.7903101481497288,
      "num_tokens": 334291502.0,
      "step": 10474
    },
    {
      "entropy": 0.58737812936306,
      "epoch": 0.9637667408468037,
      "grad_norm": 0.1464805155992508,
      "learning_rate": 6.787806299260895e-05,
      "loss": 0.5773,
      "mean_token_accuracy": 0.8271737806499004,
      "num_tokens": 334323973.0,
      "step": 10475
    },
    {
      "entropy": 0.723706005141139,
      "epoch": 0.9638587472182448,
      "grad_norm": 0.15872900187969208,
      "learning_rate": 6.787499616646733e-05,
      "loss": 0.6969,
      "mean_token_accuracy": 0.7944507971405983,
      "num_tokens": 334355911.0,
      "step": 10476
    },
    {
      "entropy": 0.8124489095062017,
      "epoch": 0.9639507535896861,
      "grad_norm": 0.15314558148384094,
      "learning_rate": 6.78719293403257e-05,
      "loss": 0.8097,
      "mean_token_accuracy": 0.7616039328277111,
      "num_tokens": 334387506.0,
      "step": 10477
    },
    {
      "entropy": 0.7241396177560091,
      "epoch": 0.9640427599611273,
      "grad_norm": 0.16408225893974304,
      "learning_rate": 6.786886251418408e-05,
      "loss": 0.7134,
      "mean_token_accuracy": 0.7834387645125389,
      "num_tokens": 334418895.0,
      "step": 10478
    },
    {
      "entropy": 0.6646340470761061,
      "epoch": 0.9641347663325686,
      "grad_norm": 0.1573539674282074,
      "learning_rate": 6.786579568804245e-05,
      "loss": 0.6465,
      "mean_token_accuracy": 0.8048742078244686,
      "num_tokens": 334450820.0,
      "step": 10479
    },
    {
      "entropy": 0.756290883757174,
      "epoch": 0.9642267727040098,
      "grad_norm": 0.15764503180980682,
      "learning_rate": 6.786272886190083e-05,
      "loss": 0.7451,
      "mean_token_accuracy": 0.7791400663554668,
      "num_tokens": 334481992.0,
      "step": 10480
    },
    {
      "entropy": 0.689883591607213,
      "epoch": 0.9643187790754509,
      "grad_norm": 0.16057968139648438,
      "learning_rate": 6.78596620357592e-05,
      "loss": 0.6795,
      "mean_token_accuracy": 0.8027783185243607,
      "num_tokens": 334514301.0,
      "step": 10481
    },
    {
      "entropy": 0.7971299737691879,
      "epoch": 0.9644107854468922,
      "grad_norm": 0.16522438824176788,
      "learning_rate": 6.785659520961756e-05,
      "loss": 0.8021,
      "mean_token_accuracy": 0.7668251693248749,
      "num_tokens": 334546545.0,
      "step": 10482
    },
    {
      "entropy": 0.6983304657042027,
      "epoch": 0.9645027918183334,
      "grad_norm": 0.15243415534496307,
      "learning_rate": 6.785352838347595e-05,
      "loss": 0.6916,
      "mean_token_accuracy": 0.7944676466286182,
      "num_tokens": 334578667.0,
      "step": 10483
    },
    {
      "entropy": 0.6750128492712975,
      "epoch": 0.9645947981897747,
      "grad_norm": 0.16067010164260864,
      "learning_rate": 6.785046155733433e-05,
      "loss": 0.6625,
      "mean_token_accuracy": 0.8025503419339657,
      "num_tokens": 334610400.0,
      "step": 10484
    },
    {
      "entropy": 0.6921940613538027,
      "epoch": 0.9646868045612159,
      "grad_norm": 0.15375195443630219,
      "learning_rate": 6.78473947311927e-05,
      "loss": 0.6897,
      "mean_token_accuracy": 0.7993061542510986,
      "num_tokens": 334642222.0,
      "step": 10485
    },
    {
      "entropy": 0.6781353885307908,
      "epoch": 0.964778810932657,
      "grad_norm": 0.14459428191184998,
      "learning_rate": 6.784432790505106e-05,
      "loss": 0.6618,
      "mean_token_accuracy": 0.7988126128911972,
      "num_tokens": 334674508.0,
      "step": 10486
    },
    {
      "entropy": 0.7060933643952012,
      "epoch": 0.9648708173040983,
      "grad_norm": 0.15126557648181915,
      "learning_rate": 6.784126107890943e-05,
      "loss": 0.7011,
      "mean_token_accuracy": 0.7963321059942245,
      "num_tokens": 334706666.0,
      "step": 10487
    },
    {
      "entropy": 0.7005090061575174,
      "epoch": 0.9649628236755395,
      "grad_norm": 0.15567128360271454,
      "learning_rate": 6.783819425276781e-05,
      "loss": 0.6837,
      "mean_token_accuracy": 0.7974599115550518,
      "num_tokens": 334738274.0,
      "step": 10488
    },
    {
      "entropy": 0.8124092109501362,
      "epoch": 0.9650548300469808,
      "grad_norm": 0.15776626765727997,
      "learning_rate": 6.78351274266262e-05,
      "loss": 0.8094,
      "mean_token_accuracy": 0.7676772437989712,
      "num_tokens": 334770078.0,
      "step": 10489
    },
    {
      "entropy": 0.7968130074441433,
      "epoch": 0.965146836418422,
      "grad_norm": 0.15507860481739044,
      "learning_rate": 6.783206060048456e-05,
      "loss": 0.7812,
      "mean_token_accuracy": 0.7698117047548294,
      "num_tokens": 334801687.0,
      "step": 10490
    },
    {
      "entropy": 0.701839642599225,
      "epoch": 0.9652388427898632,
      "grad_norm": 0.15600761771202087,
      "learning_rate": 6.782899377434293e-05,
      "loss": 0.6867,
      "mean_token_accuracy": 0.7955822199583054,
      "num_tokens": 334833665.0,
      "step": 10491
    },
    {
      "entropy": 0.76115339435637,
      "epoch": 0.9653308491613044,
      "grad_norm": 0.15360896289348602,
      "learning_rate": 6.782592694820131e-05,
      "loss": 0.7422,
      "mean_token_accuracy": 0.7804295979440212,
      "num_tokens": 334865904.0,
      "step": 10492
    },
    {
      "entropy": 0.8110752096399665,
      "epoch": 0.9654228555327456,
      "grad_norm": 0.16557522118091583,
      "learning_rate": 6.782286012205968e-05,
      "loss": 0.8061,
      "mean_token_accuracy": 0.7615419961512089,
      "num_tokens": 334897583.0,
      "step": 10493
    },
    {
      "entropy": 0.5876595973968506,
      "epoch": 0.9655148619041869,
      "grad_norm": 0.14943335950374603,
      "learning_rate": 6.781979329591806e-05,
      "loss": 0.5703,
      "mean_token_accuracy": 0.8286009095609188,
      "num_tokens": 334930090.0,
      "step": 10494
    },
    {
      "entropy": 0.7785910312086344,
      "epoch": 0.9656068682756281,
      "grad_norm": 0.16083769500255585,
      "learning_rate": 6.781672646977644e-05,
      "loss": 0.7621,
      "mean_token_accuracy": 0.7721336595714092,
      "num_tokens": 334961679.0,
      "step": 10495
    },
    {
      "entropy": 0.6549607086926699,
      "epoch": 0.9656988746470693,
      "grad_norm": 0.14495083689689636,
      "learning_rate": 6.781365964363481e-05,
      "loss": 0.6427,
      "mean_token_accuracy": 0.8061939179897308,
      "num_tokens": 334994414.0,
      "step": 10496
    },
    {
      "entropy": 0.6576637476682663,
      "epoch": 0.9657908810185105,
      "grad_norm": 0.14652591943740845,
      "learning_rate": 6.781059281749318e-05,
      "loss": 0.6509,
      "mean_token_accuracy": 0.8034491129219532,
      "num_tokens": 335027027.0,
      "step": 10497
    },
    {
      "entropy": 0.6730944393202662,
      "epoch": 0.9658828873899518,
      "grad_norm": 0.16199244558811188,
      "learning_rate": 6.780752599135154e-05,
      "loss": 0.6671,
      "mean_token_accuracy": 0.8030440472066402,
      "num_tokens": 335058830.0,
      "step": 10498
    },
    {
      "entropy": 0.6369539201259613,
      "epoch": 0.965974893761393,
      "grad_norm": 0.15112310647964478,
      "learning_rate": 6.780445916520993e-05,
      "loss": 0.6214,
      "mean_token_accuracy": 0.8111187517642975,
      "num_tokens": 335091223.0,
      "step": 10499
    },
    {
      "entropy": 0.6441938569769263,
      "epoch": 0.9660669001328342,
      "grad_norm": 0.14958570897579193,
      "learning_rate": 6.780139233906831e-05,
      "loss": 0.6208,
      "mean_token_accuracy": 0.817092277109623,
      "num_tokens": 335123418.0,
      "step": 10500
    },
    {
      "entropy": 0.7213664762675762,
      "epoch": 0.9661589065042754,
      "grad_norm": 0.16534079611301422,
      "learning_rate": 6.779832551292668e-05,
      "loss": 0.7149,
      "mean_token_accuracy": 0.7902813032269478,
      "num_tokens": 335155393.0,
      "step": 10501
    },
    {
      "entropy": 0.6956989075988531,
      "epoch": 0.9662509128757166,
      "grad_norm": 0.15969474613666534,
      "learning_rate": 6.779525868678504e-05,
      "loss": 0.6992,
      "mean_token_accuracy": 0.789272490888834,
      "num_tokens": 335187039.0,
      "step": 10502
    },
    {
      "entropy": 0.6833267044275999,
      "epoch": 0.9663429192471579,
      "grad_norm": 0.16233769059181213,
      "learning_rate": 6.779219186064342e-05,
      "loss": 0.6636,
      "mean_token_accuracy": 0.8015560880303383,
      "num_tokens": 335219136.0,
      "step": 10503
    },
    {
      "entropy": 0.529009566642344,
      "epoch": 0.9664349256185991,
      "grad_norm": 0.14842405915260315,
      "learning_rate": 6.778912503450179e-05,
      "loss": 0.4936,
      "mean_token_accuracy": 0.8472982719540596,
      "num_tokens": 335250354.0,
      "step": 10504
    },
    {
      "entropy": 0.7445050636306405,
      "epoch": 0.9665269319900404,
      "grad_norm": 0.15409502387046814,
      "learning_rate": 6.778605820836017e-05,
      "loss": 0.7323,
      "mean_token_accuracy": 0.782936979085207,
      "num_tokens": 335281442.0,
      "step": 10505
    },
    {
      "entropy": 0.7050412502139807,
      "epoch": 0.9666189383614815,
      "grad_norm": 0.14970716834068298,
      "learning_rate": 6.778299138221854e-05,
      "loss": 0.6885,
      "mean_token_accuracy": 0.7987059950828552,
      "num_tokens": 335312516.0,
      "step": 10506
    },
    {
      "entropy": 0.7794883642345667,
      "epoch": 0.9667109447329227,
      "grad_norm": 0.1615133285522461,
      "learning_rate": 6.777992455607692e-05,
      "loss": 0.7788,
      "mean_token_accuracy": 0.7707688137888908,
      "num_tokens": 335344903.0,
      "step": 10507
    },
    {
      "entropy": 0.547856415156275,
      "epoch": 0.966802951104364,
      "grad_norm": 0.14840613305568695,
      "learning_rate": 6.777685772993529e-05,
      "loss": 0.5339,
      "mean_token_accuracy": 0.836481012403965,
      "num_tokens": 335377084.0,
      "step": 10508
    },
    {
      "entropy": 0.6776507953181863,
      "epoch": 0.9668949574758052,
      "grad_norm": 0.15564417839050293,
      "learning_rate": 6.777379090379366e-05,
      "loss": 0.6722,
      "mean_token_accuracy": 0.7986418195068836,
      "num_tokens": 335409562.0,
      "step": 10509
    },
    {
      "entropy": 0.8422871958464384,
      "epoch": 0.9669869638472465,
      "grad_norm": 0.16530828177928925,
      "learning_rate": 6.777072407765204e-05,
      "loss": 0.8396,
      "mean_token_accuracy": 0.7559425458312035,
      "num_tokens": 335441210.0,
      "step": 10510
    },
    {
      "entropy": 0.8178528174757957,
      "epoch": 0.9670789702186876,
      "grad_norm": 0.1576511263847351,
      "learning_rate": 6.776765725151042e-05,
      "loss": 0.8227,
      "mean_token_accuracy": 0.7563239857554436,
      "num_tokens": 335473219.0,
      "step": 10511
    },
    {
      "entropy": 0.7150242924690247,
      "epoch": 0.9671709765901289,
      "grad_norm": 0.15693390369415283,
      "learning_rate": 6.776459042536879e-05,
      "loss": 0.6961,
      "mean_token_accuracy": 0.7976838611066341,
      "num_tokens": 335505227.0,
      "step": 10512
    },
    {
      "entropy": 0.7683178726583719,
      "epoch": 0.9672629829615701,
      "grad_norm": 0.16684956848621368,
      "learning_rate": 6.776152359922716e-05,
      "loss": 0.7442,
      "mean_token_accuracy": 0.7772562317550182,
      "num_tokens": 335535898.0,
      "step": 10513
    },
    {
      "entropy": 0.6385474409908056,
      "epoch": 0.9673549893330113,
      "grad_norm": 0.15457671880722046,
      "learning_rate": 6.775845677308553e-05,
      "loss": 0.6272,
      "mean_token_accuracy": 0.8160071671009064,
      "num_tokens": 335568226.0,
      "step": 10514
    },
    {
      "entropy": 0.7511184774339199,
      "epoch": 0.9674469957044526,
      "grad_norm": 0.1553974449634552,
      "learning_rate": 6.775538994694392e-05,
      "loss": 0.7577,
      "mean_token_accuracy": 0.7737687900662422,
      "num_tokens": 335600387.0,
      "step": 10515
    },
    {
      "entropy": 0.7939011920243502,
      "epoch": 0.9675390020758937,
      "grad_norm": 0.15741904079914093,
      "learning_rate": 6.775232312080229e-05,
      "loss": 0.8032,
      "mean_token_accuracy": 0.7667487226426601,
      "num_tokens": 335631918.0,
      "step": 10516
    },
    {
      "entropy": 0.6814411263912916,
      "epoch": 0.967631008447335,
      "grad_norm": 0.16158193349838257,
      "learning_rate": 6.774925629466066e-05,
      "loss": 0.6684,
      "mean_token_accuracy": 0.8035847283899784,
      "num_tokens": 335664155.0,
      "step": 10517
    },
    {
      "entropy": 0.7025232128798962,
      "epoch": 0.9677230148187762,
      "grad_norm": 0.15779879689216614,
      "learning_rate": 6.774618946851902e-05,
      "loss": 0.6913,
      "mean_token_accuracy": 0.7940782159566879,
      "num_tokens": 335696656.0,
      "step": 10518
    },
    {
      "entropy": 0.9049936197698116,
      "epoch": 0.9678150211902175,
      "grad_norm": 0.16284102201461792,
      "learning_rate": 6.77431226423774e-05,
      "loss": 0.889,
      "mean_token_accuracy": 0.7431676089763641,
      "num_tokens": 335728739.0,
      "step": 10519
    },
    {
      "entropy": 0.631651584059,
      "epoch": 0.9679070275616587,
      "grad_norm": 0.15819275379180908,
      "learning_rate": 6.774005581623579e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8150783814489841,
      "num_tokens": 335760492.0,
      "step": 10520
    },
    {
      "entropy": 0.8658122289925814,
      "epoch": 0.9679990339330998,
      "grad_norm": 0.15652528405189514,
      "learning_rate": 6.773698899009416e-05,
      "loss": 0.8444,
      "mean_token_accuracy": 0.7536990381777287,
      "num_tokens": 335792448.0,
      "step": 10521
    },
    {
      "entropy": 0.6179084610193968,
      "epoch": 0.9680910403045411,
      "grad_norm": 0.1564515382051468,
      "learning_rate": 6.773392216395252e-05,
      "loss": 0.6004,
      "mean_token_accuracy": 0.818936463445425,
      "num_tokens": 335824677.0,
      "step": 10522
    },
    {
      "entropy": 0.760549645870924,
      "epoch": 0.9681830466759823,
      "grad_norm": 0.15943452715873718,
      "learning_rate": 6.77308553378109e-05,
      "loss": 0.7578,
      "mean_token_accuracy": 0.7739357426762581,
      "num_tokens": 335856425.0,
      "step": 10523
    },
    {
      "entropy": 0.6810291092842817,
      "epoch": 0.9682750530474236,
      "grad_norm": 0.15459364652633667,
      "learning_rate": 6.772778851166927e-05,
      "loss": 0.6507,
      "mean_token_accuracy": 0.8053512200713158,
      "num_tokens": 335888521.0,
      "step": 10524
    },
    {
      "entropy": 0.7077015172690153,
      "epoch": 0.9683670594188648,
      "grad_norm": 0.159923255443573,
      "learning_rate": 6.772472168552765e-05,
      "loss": 0.7067,
      "mean_token_accuracy": 0.7855750471353531,
      "num_tokens": 335919483.0,
      "step": 10525
    },
    {
      "entropy": 0.6317358752712607,
      "epoch": 0.9684590657903059,
      "grad_norm": 0.14460024237632751,
      "learning_rate": 6.772165485938604e-05,
      "loss": 0.625,
      "mean_token_accuracy": 0.8154569007456303,
      "num_tokens": 335952141.0,
      "step": 10526
    },
    {
      "entropy": 0.6819276548922062,
      "epoch": 0.9685510721617472,
      "grad_norm": 0.15570661425590515,
      "learning_rate": 6.77185880332444e-05,
      "loss": 0.6736,
      "mean_token_accuracy": 0.7989084310829639,
      "num_tokens": 335984464.0,
      "step": 10527
    },
    {
      "entropy": 0.7514672316610813,
      "epoch": 0.9686430785331884,
      "grad_norm": 0.15718428790569305,
      "learning_rate": 6.771552120710277e-05,
      "loss": 0.7499,
      "mean_token_accuracy": 0.7783442176878452,
      "num_tokens": 336016688.0,
      "step": 10528
    },
    {
      "entropy": 0.7650087494403124,
      "epoch": 0.9687350849046297,
      "grad_norm": 0.152118980884552,
      "learning_rate": 6.771245438096114e-05,
      "loss": 0.7629,
      "mean_token_accuracy": 0.7765888459980488,
      "num_tokens": 336048290.0,
      "step": 10529
    },
    {
      "entropy": 0.6672723554074764,
      "epoch": 0.9688270912760709,
      "grad_norm": 0.14140962064266205,
      "learning_rate": 6.770938755481952e-05,
      "loss": 0.6496,
      "mean_token_accuracy": 0.8048717826604843,
      "num_tokens": 336080450.0,
      "step": 10530
    },
    {
      "entropy": 0.7344309277832508,
      "epoch": 0.968919097647512,
      "grad_norm": 0.154737189412117,
      "learning_rate": 6.77063207286779e-05,
      "loss": 0.725,
      "mean_token_accuracy": 0.7861641272902489,
      "num_tokens": 336111962.0,
      "step": 10531
    },
    {
      "entropy": 0.617815419100225,
      "epoch": 0.9690111040189533,
      "grad_norm": 0.15271960198879242,
      "learning_rate": 6.770325390253627e-05,
      "loss": 0.5973,
      "mean_token_accuracy": 0.8184418566524982,
      "num_tokens": 336143098.0,
      "step": 10532
    },
    {
      "entropy": 0.7587207145988941,
      "epoch": 0.9691031103903945,
      "grad_norm": 0.15506324172019958,
      "learning_rate": 6.770018707639464e-05,
      "loss": 0.7521,
      "mean_token_accuracy": 0.7766907922923565,
      "num_tokens": 336174943.0,
      "step": 10533
    },
    {
      "entropy": 0.635185282677412,
      "epoch": 0.9691951167618358,
      "grad_norm": 0.15268121659755707,
      "learning_rate": 6.769712025025302e-05,
      "loss": 0.6158,
      "mean_token_accuracy": 0.8147372081875801,
      "num_tokens": 336206792.0,
      "step": 10534
    },
    {
      "entropy": 0.7282321695238352,
      "epoch": 0.969287123133277,
      "grad_norm": 0.15430523455142975,
      "learning_rate": 6.769405342411139e-05,
      "loss": 0.7022,
      "mean_token_accuracy": 0.7884497977793217,
      "num_tokens": 336239178.0,
      "step": 10535
    },
    {
      "entropy": 0.6893883086740971,
      "epoch": 0.9693791295047182,
      "grad_norm": 0.1543431133031845,
      "learning_rate": 6.769098659796977e-05,
      "loss": 0.6852,
      "mean_token_accuracy": 0.7944175004959106,
      "num_tokens": 336271403.0,
      "step": 10536
    },
    {
      "entropy": 0.7006131149828434,
      "epoch": 0.9694711358761594,
      "grad_norm": 0.15469494462013245,
      "learning_rate": 6.768791977182814e-05,
      "loss": 0.6969,
      "mean_token_accuracy": 0.7928788810968399,
      "num_tokens": 336303543.0,
      "step": 10537
    },
    {
      "entropy": 0.763069299980998,
      "epoch": 0.9695631422476007,
      "grad_norm": 0.1571149230003357,
      "learning_rate": 6.768485294568652e-05,
      "loss": 0.7407,
      "mean_token_accuracy": 0.7801626138389111,
      "num_tokens": 336335889.0,
      "step": 10538
    },
    {
      "entropy": 0.7859314158558846,
      "epoch": 0.9696551486190419,
      "grad_norm": 0.16027677059173584,
      "learning_rate": 6.768178611954489e-05,
      "loss": 0.7867,
      "mean_token_accuracy": 0.7688223384320736,
      "num_tokens": 336367763.0,
      "step": 10539
    },
    {
      "entropy": 0.7926706578582525,
      "epoch": 0.9697471549904831,
      "grad_norm": 0.16399525105953217,
      "learning_rate": 6.767871929340325e-05,
      "loss": 0.7744,
      "mean_token_accuracy": 0.7742527239024639,
      "num_tokens": 336399058.0,
      "step": 10540
    },
    {
      "entropy": 0.6328762136399746,
      "epoch": 0.9698391613619243,
      "grad_norm": 0.15599016845226288,
      "learning_rate": 6.767565246726163e-05,
      "loss": 0.6177,
      "mean_token_accuracy": 0.8139687068760395,
      "num_tokens": 336430847.0,
      "step": 10541
    },
    {
      "entropy": 0.7601782195270061,
      "epoch": 0.9699311677333655,
      "grad_norm": 0.16370613873004913,
      "learning_rate": 6.767258564112002e-05,
      "loss": 0.7616,
      "mean_token_accuracy": 0.7771149538457394,
      "num_tokens": 336463096.0,
      "step": 10542
    },
    {
      "entropy": 0.7432701271027327,
      "epoch": 0.9700231741048068,
      "grad_norm": 0.15369117259979248,
      "learning_rate": 6.766951881497838e-05,
      "loss": 0.7337,
      "mean_token_accuracy": 0.7882528565824032,
      "num_tokens": 336495497.0,
      "step": 10543
    },
    {
      "entropy": 0.6366763459518552,
      "epoch": 0.970115180476248,
      "grad_norm": 0.1529141217470169,
      "learning_rate": 6.766645198883675e-05,
      "loss": 0.6158,
      "mean_token_accuracy": 0.8155534453690052,
      "num_tokens": 336527607.0,
      "step": 10544
    },
    {
      "entropy": 0.6689321752637625,
      "epoch": 0.9702071868476893,
      "grad_norm": 0.14516805112361908,
      "learning_rate": 6.766338516269512e-05,
      "loss": 0.6536,
      "mean_token_accuracy": 0.8025650680065155,
      "num_tokens": 336559789.0,
      "step": 10545
    },
    {
      "entropy": 0.6339048258960247,
      "epoch": 0.9702991932191304,
      "grad_norm": 0.14677385985851288,
      "learning_rate": 6.76603183365535e-05,
      "loss": 0.6268,
      "mean_token_accuracy": 0.8107163906097412,
      "num_tokens": 336591608.0,
      "step": 10546
    },
    {
      "entropy": 0.7093491535633802,
      "epoch": 0.9703911995905716,
      "grad_norm": 0.15798795223236084,
      "learning_rate": 6.765725151041188e-05,
      "loss": 0.6965,
      "mean_token_accuracy": 0.7889992520213127,
      "num_tokens": 336623841.0,
      "step": 10547
    },
    {
      "entropy": 0.6402760352939367,
      "epoch": 0.9704832059620129,
      "grad_norm": 0.146479532122612,
      "learning_rate": 6.765418468427025e-05,
      "loss": 0.624,
      "mean_token_accuracy": 0.8124172240495682,
      "num_tokens": 336655701.0,
      "step": 10548
    },
    {
      "entropy": 0.7333541251718998,
      "epoch": 0.9705752123334541,
      "grad_norm": 0.1628318727016449,
      "learning_rate": 6.765111785812862e-05,
      "loss": 0.7401,
      "mean_token_accuracy": 0.7826445251703262,
      "num_tokens": 336687603.0,
      "step": 10549
    },
    {
      "entropy": 0.7352830395102501,
      "epoch": 0.9706672187048954,
      "grad_norm": 0.159367173910141,
      "learning_rate": 6.7648051031987e-05,
      "loss": 0.743,
      "mean_token_accuracy": 0.7770535945892334,
      "num_tokens": 336719182.0,
      "step": 10550
    },
    {
      "entropy": 0.73453239351511,
      "epoch": 0.9707592250763365,
      "grad_norm": 0.15433871746063232,
      "learning_rate": 6.764498420584538e-05,
      "loss": 0.7265,
      "mean_token_accuracy": 0.7847617827355862,
      "num_tokens": 336751330.0,
      "step": 10551
    },
    {
      "entropy": 0.7827771287411451,
      "epoch": 0.9708512314477777,
      "grad_norm": 0.1558324545621872,
      "learning_rate": 6.764191737970375e-05,
      "loss": 0.7821,
      "mean_token_accuracy": 0.7697421684861183,
      "num_tokens": 336782330.0,
      "step": 10552
    },
    {
      "entropy": 0.7272303551435471,
      "epoch": 0.970943237819219,
      "grad_norm": 0.15019746124744415,
      "learning_rate": 6.763885055356212e-05,
      "loss": 0.7161,
      "mean_token_accuracy": 0.787222295999527,
      "num_tokens": 336814120.0,
      "step": 10553
    },
    {
      "entropy": 0.7602425906807184,
      "epoch": 0.9710352441906602,
      "grad_norm": 0.15721143782138824,
      "learning_rate": 6.76357837274205e-05,
      "loss": 0.7527,
      "mean_token_accuracy": 0.7822928056120872,
      "num_tokens": 336846226.0,
      "step": 10554
    },
    {
      "entropy": 0.6896449904888868,
      "epoch": 0.9711272505621015,
      "grad_norm": 0.1520087718963623,
      "learning_rate": 6.763271690127887e-05,
      "loss": 0.6713,
      "mean_token_accuracy": 0.7949179336428642,
      "num_tokens": 336878179.0,
      "step": 10555
    },
    {
      "entropy": 0.7484518066048622,
      "epoch": 0.9712192569335426,
      "grad_norm": 0.15979091823101044,
      "learning_rate": 6.762965007513725e-05,
      "loss": 0.7342,
      "mean_token_accuracy": 0.7846382446587086,
      "num_tokens": 336909936.0,
      "step": 10556
    },
    {
      "entropy": 0.7735441243276,
      "epoch": 0.9713112633049839,
      "grad_norm": 0.15075522661209106,
      "learning_rate": 6.762658324899563e-05,
      "loss": 0.7571,
      "mean_token_accuracy": 0.7785229831933975,
      "num_tokens": 336941415.0,
      "step": 10557
    },
    {
      "entropy": 0.7245197277516127,
      "epoch": 0.9714032696764251,
      "grad_norm": 0.15866287052631378,
      "learning_rate": 6.7623516422854e-05,
      "loss": 0.703,
      "mean_token_accuracy": 0.792338814586401,
      "num_tokens": 336972010.0,
      "step": 10558
    },
    {
      "entropy": 0.5938303470611572,
      "epoch": 0.9714952760478663,
      "grad_norm": 0.14990799129009247,
      "learning_rate": 6.762044959671236e-05,
      "loss": 0.5778,
      "mean_token_accuracy": 0.8244544118642807,
      "num_tokens": 337003296.0,
      "step": 10559
    },
    {
      "entropy": 0.8035529144108295,
      "epoch": 0.9715872824193076,
      "grad_norm": 0.168071448802948,
      "learning_rate": 6.761738277057073e-05,
      "loss": 0.8102,
      "mean_token_accuracy": 0.7637681253254414,
      "num_tokens": 337034950.0,
      "step": 10560
    },
    {
      "entropy": 0.714344646781683,
      "epoch": 0.9716792887907487,
      "grad_norm": 0.14511479437351227,
      "learning_rate": 6.761431594442911e-05,
      "loss": 0.7085,
      "mean_token_accuracy": 0.7860003374516964,
      "num_tokens": 337067096.0,
      "step": 10561
    },
    {
      "entropy": 0.6640689074993134,
      "epoch": 0.97177129516219,
      "grad_norm": 0.15830907225608826,
      "learning_rate": 6.76112491182875e-05,
      "loss": 0.6559,
      "mean_token_accuracy": 0.8070737756788731,
      "num_tokens": 337098938.0,
      "step": 10562
    },
    {
      "entropy": 0.6577910138294101,
      "epoch": 0.9718633015336312,
      "grad_norm": 0.14559032022953033,
      "learning_rate": 6.760818229214586e-05,
      "loss": 0.6407,
      "mean_token_accuracy": 0.806329995393753,
      "num_tokens": 337131128.0,
      "step": 10563
    },
    {
      "entropy": 0.7828767523169518,
      "epoch": 0.9719553079050725,
      "grad_norm": 0.1558733582496643,
      "learning_rate": 6.760511546600423e-05,
      "loss": 0.7755,
      "mean_token_accuracy": 0.7661004476249218,
      "num_tokens": 337161815.0,
      "step": 10564
    },
    {
      "entropy": 0.7576594091951847,
      "epoch": 0.9720473142765137,
      "grad_norm": 0.16401514410972595,
      "learning_rate": 6.760204863986261e-05,
      "loss": 0.7662,
      "mean_token_accuracy": 0.7728305123746395,
      "num_tokens": 337193743.0,
      "step": 10565
    },
    {
      "entropy": 0.7198924580588937,
      "epoch": 0.9721393206479548,
      "grad_norm": 0.15998481214046478,
      "learning_rate": 6.759898181372098e-05,
      "loss": 0.7252,
      "mean_token_accuracy": 0.7881961390376091,
      "num_tokens": 337226511.0,
      "step": 10566
    },
    {
      "entropy": 0.7353869210928679,
      "epoch": 0.9722313270193961,
      "grad_norm": 0.15878520905971527,
      "learning_rate": 6.759591498757936e-05,
      "loss": 0.7366,
      "mean_token_accuracy": 0.7761143408715725,
      "num_tokens": 337258310.0,
      "step": 10567
    },
    {
      "entropy": 0.604597101919353,
      "epoch": 0.9723233333908373,
      "grad_norm": 0.16367650032043457,
      "learning_rate": 6.759284816143773e-05,
      "loss": 0.5905,
      "mean_token_accuracy": 0.8208009898662567,
      "num_tokens": 337289985.0,
      "step": 10568
    },
    {
      "entropy": 0.7860522139817476,
      "epoch": 0.9724153397622786,
      "grad_norm": 0.1582975685596466,
      "learning_rate": 6.758978133529611e-05,
      "loss": 0.7767,
      "mean_token_accuracy": 0.7756835408508778,
      "num_tokens": 337321629.0,
      "step": 10569
    },
    {
      "entropy": 0.7250996846705675,
      "epoch": 0.9725073461337198,
      "grad_norm": 0.15886923670768738,
      "learning_rate": 6.758671450915448e-05,
      "loss": 0.7192,
      "mean_token_accuracy": 0.7956133633852005,
      "num_tokens": 337353503.0,
      "step": 10570
    },
    {
      "entropy": 0.6459373431280255,
      "epoch": 0.9725993525051609,
      "grad_norm": 0.15018582344055176,
      "learning_rate": 6.758364768301285e-05,
      "loss": 0.6367,
      "mean_token_accuracy": 0.8100015446543694,
      "num_tokens": 337385958.0,
      "step": 10571
    },
    {
      "entropy": 0.7130837645381689,
      "epoch": 0.9726913588766022,
      "grad_norm": 0.15575556457042694,
      "learning_rate": 6.758058085687123e-05,
      "loss": 0.6877,
      "mean_token_accuracy": 0.7917790003120899,
      "num_tokens": 337417093.0,
      "step": 10572
    },
    {
      "entropy": 0.6063654711470008,
      "epoch": 0.9727833652480434,
      "grad_norm": 0.15825489163398743,
      "learning_rate": 6.757751403072961e-05,
      "loss": 0.5897,
      "mean_token_accuracy": 0.8247350193560123,
      "num_tokens": 337449298.0,
      "step": 10573
    },
    {
      "entropy": 0.7501623965799809,
      "epoch": 0.9728753716194847,
      "grad_norm": 0.1513490527868271,
      "learning_rate": 6.757444720458798e-05,
      "loss": 0.7374,
      "mean_token_accuracy": 0.7860096953809261,
      "num_tokens": 337481102.0,
      "step": 10574
    },
    {
      "entropy": 0.7025008499622345,
      "epoch": 0.9729673779909259,
      "grad_norm": 0.15990649163722992,
      "learning_rate": 6.757138037844635e-05,
      "loss": 0.6813,
      "mean_token_accuracy": 0.7993027158081532,
      "num_tokens": 337512603.0,
      "step": 10575
    },
    {
      "entropy": 0.7116616852581501,
      "epoch": 0.973059384362367,
      "grad_norm": 0.14640459418296814,
      "learning_rate": 6.756831355230471e-05,
      "loss": 0.6923,
      "mean_token_accuracy": 0.7975670658051968,
      "num_tokens": 337545009.0,
      "step": 10576
    },
    {
      "entropy": 0.6928593069314957,
      "epoch": 0.9731513907338083,
      "grad_norm": 0.15537488460540771,
      "learning_rate": 6.75652467261631e-05,
      "loss": 0.6787,
      "mean_token_accuracy": 0.7972130440175533,
      "num_tokens": 337577630.0,
      "step": 10577
    },
    {
      "entropy": 0.5907162073999643,
      "epoch": 0.9732433971052495,
      "grad_norm": 0.1512099951505661,
      "learning_rate": 6.756217990002148e-05,
      "loss": 0.571,
      "mean_token_accuracy": 0.8313009403645992,
      "num_tokens": 337609902.0,
      "step": 10578
    },
    {
      "entropy": 0.5856647323817015,
      "epoch": 0.9733354034766908,
      "grad_norm": 0.14588406682014465,
      "learning_rate": 6.755911307387984e-05,
      "loss": 0.5528,
      "mean_token_accuracy": 0.8307448625564575,
      "num_tokens": 337641615.0,
      "step": 10579
    },
    {
      "entropy": 0.6770801078528166,
      "epoch": 0.973427409848132,
      "grad_norm": 0.15753860771656036,
      "learning_rate": 6.755604624773821e-05,
      "loss": 0.6552,
      "mean_token_accuracy": 0.802685908973217,
      "num_tokens": 337672856.0,
      "step": 10580
    },
    {
      "entropy": 0.7444513887166977,
      "epoch": 0.9735194162195732,
      "grad_norm": 0.1575642079114914,
      "learning_rate": 6.75529794215966e-05,
      "loss": 0.7375,
      "mean_token_accuracy": 0.7873792387545109,
      "num_tokens": 337704739.0,
      "step": 10581
    },
    {
      "entropy": 0.8119419813156128,
      "epoch": 0.9736114225910144,
      "grad_norm": 0.1580500453710556,
      "learning_rate": 6.754991259545496e-05,
      "loss": 0.8008,
      "mean_token_accuracy": 0.765133935958147,
      "num_tokens": 337736349.0,
      "step": 10582
    },
    {
      "entropy": 0.7209542170166969,
      "epoch": 0.9737034289624557,
      "grad_norm": 0.15572994947433472,
      "learning_rate": 6.754684576931334e-05,
      "loss": 0.7072,
      "mean_token_accuracy": 0.7868721708655357,
      "num_tokens": 337767975.0,
      "step": 10583
    },
    {
      "entropy": 0.8187835067510605,
      "epoch": 0.9737954353338969,
      "grad_norm": 0.16079941391944885,
      "learning_rate": 6.754377894317171e-05,
      "loss": 0.8213,
      "mean_token_accuracy": 0.7564080730080605,
      "num_tokens": 337799902.0,
      "step": 10584
    },
    {
      "entropy": 0.6487867031246424,
      "epoch": 0.9738874417053381,
      "grad_norm": 0.16482362151145935,
      "learning_rate": 6.754071211703009e-05,
      "loss": 0.6293,
      "mean_token_accuracy": 0.8066891022026539,
      "num_tokens": 337830715.0,
      "step": 10585
    },
    {
      "entropy": 0.721027122810483,
      "epoch": 0.9739794480767793,
      "grad_norm": 0.15432745218276978,
      "learning_rate": 6.753764529088846e-05,
      "loss": 0.7129,
      "mean_token_accuracy": 0.7894077450037003,
      "num_tokens": 337862582.0,
      "step": 10586
    },
    {
      "entropy": 0.7256790231913328,
      "epoch": 0.9740714544482205,
      "grad_norm": 0.1622440367937088,
      "learning_rate": 6.753457846474683e-05,
      "loss": 0.7309,
      "mean_token_accuracy": 0.7840545065701008,
      "num_tokens": 337894778.0,
      "step": 10587
    },
    {
      "entropy": 0.6176075688563287,
      "epoch": 0.9741634608196618,
      "grad_norm": 0.15290309488773346,
      "learning_rate": 6.753151163860522e-05,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8250227607786655,
      "num_tokens": 337926485.0,
      "step": 10588
    },
    {
      "entropy": 0.7470509316772223,
      "epoch": 0.974255467191103,
      "grad_norm": 0.15583010017871857,
      "learning_rate": 6.752844481246359e-05,
      "loss": 0.7303,
      "mean_token_accuracy": 0.7845859676599503,
      "num_tokens": 337958345.0,
      "step": 10589
    },
    {
      "entropy": 0.8283224366605282,
      "epoch": 0.9743474735625443,
      "grad_norm": 0.1700611114501953,
      "learning_rate": 6.752537798632196e-05,
      "loss": 0.8414,
      "mean_token_accuracy": 0.7552184946835041,
      "num_tokens": 337990402.0,
      "step": 10590
    },
    {
      "entropy": 0.6968209920451045,
      "epoch": 0.9744394799339854,
      "grad_norm": 0.1519262045621872,
      "learning_rate": 6.752231116018033e-05,
      "loss": 0.6766,
      "mean_token_accuracy": 0.7971009574830532,
      "num_tokens": 338021582.0,
      "step": 10591
    },
    {
      "entropy": 0.7682728972285986,
      "epoch": 0.9745314863054266,
      "grad_norm": 0.15699885785579681,
      "learning_rate": 6.751924433403871e-05,
      "loss": 0.7555,
      "mean_token_accuracy": 0.7757477797567844,
      "num_tokens": 338053241.0,
      "step": 10592
    },
    {
      "entropy": 0.686125093139708,
      "epoch": 0.9746234926768679,
      "grad_norm": 0.14954133331775665,
      "learning_rate": 6.751617750789709e-05,
      "loss": 0.6667,
      "mean_token_accuracy": 0.8009371906518936,
      "num_tokens": 338085521.0,
      "step": 10593
    },
    {
      "entropy": 0.6611022315919399,
      "epoch": 0.9747154990483091,
      "grad_norm": 0.1567360907793045,
      "learning_rate": 6.751311068175546e-05,
      "loss": 0.6515,
      "mean_token_accuracy": 0.8074835985898972,
      "num_tokens": 338116504.0,
      "step": 10594
    },
    {
      "entropy": 0.6066690571606159,
      "epoch": 0.9748075054197504,
      "grad_norm": 0.15125270187854767,
      "learning_rate": 6.751004385561382e-05,
      "loss": 0.5823,
      "mean_token_accuracy": 0.823846310377121,
      "num_tokens": 338148032.0,
      "step": 10595
    },
    {
      "entropy": 0.7638347633183002,
      "epoch": 0.9748995117911915,
      "grad_norm": 0.15812021493911743,
      "learning_rate": 6.75069770294722e-05,
      "loss": 0.7556,
      "mean_token_accuracy": 0.7782583720982075,
      "num_tokens": 338179410.0,
      "step": 10596
    },
    {
      "entropy": 0.66310172714293,
      "epoch": 0.9749915181626327,
      "grad_norm": 0.15262490510940552,
      "learning_rate": 6.750391020333057e-05,
      "loss": 0.6549,
      "mean_token_accuracy": 0.8073478266596794,
      "num_tokens": 338211807.0,
      "step": 10597
    },
    {
      "entropy": 0.7040813155472279,
      "epoch": 0.975083524534074,
      "grad_norm": 0.1516890972852707,
      "learning_rate": 6.750084337718896e-05,
      "loss": 0.6937,
      "mean_token_accuracy": 0.7970133163034916,
      "num_tokens": 338243940.0,
      "step": 10598
    },
    {
      "entropy": 0.768902562558651,
      "epoch": 0.9751755309055152,
      "grad_norm": 0.16012081503868103,
      "learning_rate": 6.749777655104732e-05,
      "loss": 0.7579,
      "mean_token_accuracy": 0.7748105749487877,
      "num_tokens": 338275724.0,
      "step": 10599
    },
    {
      "entropy": 0.740759564563632,
      "epoch": 0.9752675372769565,
      "grad_norm": 0.1601741760969162,
      "learning_rate": 6.74947097249057e-05,
      "loss": 0.7245,
      "mean_token_accuracy": 0.7854767851531506,
      "num_tokens": 338307712.0,
      "step": 10600
    },
    {
      "entropy": 0.6790946256369352,
      "epoch": 0.9753595436483976,
      "grad_norm": 0.1545310765504837,
      "learning_rate": 6.749164289876407e-05,
      "loss": 0.6504,
      "mean_token_accuracy": 0.8050965704023838,
      "num_tokens": 338339797.0,
      "step": 10601
    },
    {
      "entropy": 0.6184138711541891,
      "epoch": 0.9754515500198389,
      "grad_norm": 0.15956680476665497,
      "learning_rate": 6.748857607262244e-05,
      "loss": 0.5919,
      "mean_token_accuracy": 0.8268303573131561,
      "num_tokens": 338372166.0,
      "step": 10602
    },
    {
      "entropy": 0.7288716584444046,
      "epoch": 0.9755435563912801,
      "grad_norm": 0.1587485522031784,
      "learning_rate": 6.748550924648082e-05,
      "loss": 0.7217,
      "mean_token_accuracy": 0.787134513258934,
      "num_tokens": 338404240.0,
      "step": 10603
    },
    {
      "entropy": 0.7027798295021057,
      "epoch": 0.9756355627627213,
      "grad_norm": 0.15404194593429565,
      "learning_rate": 6.74824424203392e-05,
      "loss": 0.6775,
      "mean_token_accuracy": 0.798168808221817,
      "num_tokens": 338436459.0,
      "step": 10604
    },
    {
      "entropy": 0.6786873713135719,
      "epoch": 0.9757275691341626,
      "grad_norm": 0.15494772791862488,
      "learning_rate": 6.747937559419757e-05,
      "loss": 0.6518,
      "mean_token_accuracy": 0.8029133826494217,
      "num_tokens": 338468371.0,
      "step": 10605
    },
    {
      "entropy": 0.6840687282383442,
      "epoch": 0.9758195755056037,
      "grad_norm": 0.16339637339115143,
      "learning_rate": 6.747630876805594e-05,
      "loss": 0.6789,
      "mean_token_accuracy": 0.8003688789904118,
      "num_tokens": 338499987.0,
      "step": 10606
    },
    {
      "entropy": 0.5773518849164248,
      "epoch": 0.975911581877045,
      "grad_norm": 0.16283704340457916,
      "learning_rate": 6.747324194191431e-05,
      "loss": 0.5655,
      "mean_token_accuracy": 0.8299007453024387,
      "num_tokens": 338531895.0,
      "step": 10607
    },
    {
      "entropy": 0.7104271287098527,
      "epoch": 0.9760035882484862,
      "grad_norm": 0.15442417562007904,
      "learning_rate": 6.747017511577269e-05,
      "loss": 0.6861,
      "mean_token_accuracy": 0.7947802282869816,
      "num_tokens": 338563482.0,
      "step": 10608
    },
    {
      "entropy": 0.7202060278505087,
      "epoch": 0.9760955946199275,
      "grad_norm": 0.15835075080394745,
      "learning_rate": 6.746710828963107e-05,
      "loss": 0.7168,
      "mean_token_accuracy": 0.7835622690618038,
      "num_tokens": 338595797.0,
      "step": 10609
    },
    {
      "entropy": 0.6228368226438761,
      "epoch": 0.9761876009913687,
      "grad_norm": 0.145578995347023,
      "learning_rate": 6.746404146348944e-05,
      "loss": 0.5996,
      "mean_token_accuracy": 0.818638265132904,
      "num_tokens": 338627788.0,
      "step": 10610
    },
    {
      "entropy": 0.7269191946834326,
      "epoch": 0.9762796073628098,
      "grad_norm": 0.16028504073619843,
      "learning_rate": 6.74609746373478e-05,
      "loss": 0.7151,
      "mean_token_accuracy": 0.7966817729175091,
      "num_tokens": 338660178.0,
      "step": 10611
    },
    {
      "entropy": 0.625467075034976,
      "epoch": 0.9763716137342511,
      "grad_norm": 0.1483326405286789,
      "learning_rate": 6.745790781120619e-05,
      "loss": 0.6032,
      "mean_token_accuracy": 0.8201422095298767,
      "num_tokens": 338692021.0,
      "step": 10612
    },
    {
      "entropy": 0.5790224429219961,
      "epoch": 0.9764636201056923,
      "grad_norm": 0.15252497792243958,
      "learning_rate": 6.745484098506455e-05,
      "loss": 0.5489,
      "mean_token_accuracy": 0.8353726975619793,
      "num_tokens": 338722423.0,
      "step": 10613
    },
    {
      "entropy": 0.657885329797864,
      "epoch": 0.9765556264771336,
      "grad_norm": 0.15731024742126465,
      "learning_rate": 6.745177415892294e-05,
      "loss": 0.6452,
      "mean_token_accuracy": 0.8062548413872719,
      "num_tokens": 338753316.0,
      "step": 10614
    },
    {
      "entropy": 0.5910470262169838,
      "epoch": 0.9766476328485748,
      "grad_norm": 0.1453896313905716,
      "learning_rate": 6.74487073327813e-05,
      "loss": 0.5699,
      "mean_token_accuracy": 0.8236535526812077,
      "num_tokens": 338785476.0,
      "step": 10615
    },
    {
      "entropy": 0.7188417613506317,
      "epoch": 0.9767396392200159,
      "grad_norm": 0.15789790451526642,
      "learning_rate": 6.744564050663969e-05,
      "loss": 0.7195,
      "mean_token_accuracy": 0.787387166172266,
      "num_tokens": 338817717.0,
      "step": 10616
    },
    {
      "entropy": 0.7821421977132559,
      "epoch": 0.9768316455914572,
      "grad_norm": 0.15802080929279327,
      "learning_rate": 6.744257368049805e-05,
      "loss": 0.7866,
      "mean_token_accuracy": 0.7660809271037579,
      "num_tokens": 338850324.0,
      "step": 10617
    },
    {
      "entropy": 0.6357427015900612,
      "epoch": 0.9769236519628984,
      "grad_norm": 0.14881528913974762,
      "learning_rate": 6.743950685435642e-05,
      "loss": 0.629,
      "mean_token_accuracy": 0.8089667595922947,
      "num_tokens": 338882051.0,
      "step": 10618
    },
    {
      "entropy": 0.7958768159151077,
      "epoch": 0.9770156583343397,
      "grad_norm": 0.1581508070230484,
      "learning_rate": 6.74364400282148e-05,
      "loss": 0.782,
      "mean_token_accuracy": 0.7707370892167091,
      "num_tokens": 338913666.0,
      "step": 10619
    },
    {
      "entropy": 0.5519986972212791,
      "epoch": 0.9771076647057809,
      "grad_norm": 0.14208228886127472,
      "learning_rate": 6.743337320207318e-05,
      "loss": 0.5332,
      "mean_token_accuracy": 0.8371062837541103,
      "num_tokens": 338945747.0,
      "step": 10620
    },
    {
      "entropy": 0.6252443306148052,
      "epoch": 0.977199671077222,
      "grad_norm": 0.15168790519237518,
      "learning_rate": 6.743030637593155e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.814689364284277,
      "num_tokens": 338978223.0,
      "step": 10621
    },
    {
      "entropy": 0.8075909893959761,
      "epoch": 0.9772916774486633,
      "grad_norm": 0.16244572401046753,
      "learning_rate": 6.742723954978992e-05,
      "loss": 0.7914,
      "mean_token_accuracy": 0.7665093913674355,
      "num_tokens": 339009491.0,
      "step": 10622
    },
    {
      "entropy": 0.7375062312930822,
      "epoch": 0.9773836838201045,
      "grad_norm": 0.1657152771949768,
      "learning_rate": 6.74241727236483e-05,
      "loss": 0.7021,
      "mean_token_accuracy": 0.7887800745666027,
      "num_tokens": 339041471.0,
      "step": 10623
    },
    {
      "entropy": 0.8323513269424438,
      "epoch": 0.9774756901915458,
      "grad_norm": 0.1657683551311493,
      "learning_rate": 6.742110589750667e-05,
      "loss": 0.8351,
      "mean_token_accuracy": 0.7566789612174034,
      "num_tokens": 339073151.0,
      "step": 10624
    },
    {
      "entropy": 0.7555996123701334,
      "epoch": 0.977567696562987,
      "grad_norm": 0.16382910311222076,
      "learning_rate": 6.741803907136505e-05,
      "loss": 0.7508,
      "mean_token_accuracy": 0.7851800583302975,
      "num_tokens": 339105495.0,
      "step": 10625
    },
    {
      "entropy": 0.6278354432433844,
      "epoch": 0.9776597029344282,
      "grad_norm": 0.17628969252109528,
      "learning_rate": 6.741497224522342e-05,
      "loss": 0.6147,
      "mean_token_accuracy": 0.8158922232687473,
      "num_tokens": 339136935.0,
      "step": 10626
    },
    {
      "entropy": 0.7004539538174868,
      "epoch": 0.9777517093058694,
      "grad_norm": 0.15758752822875977,
      "learning_rate": 6.74119054190818e-05,
      "loss": 0.675,
      "mean_token_accuracy": 0.7957786843180656,
      "num_tokens": 339168820.0,
      "step": 10627
    },
    {
      "entropy": 0.8613657411187887,
      "epoch": 0.9778437156773107,
      "grad_norm": 0.16373014450073242,
      "learning_rate": 6.740883859294017e-05,
      "loss": 0.8516,
      "mean_token_accuracy": 0.7560464255511761,
      "num_tokens": 339201005.0,
      "step": 10628
    },
    {
      "entropy": 0.6481874091550708,
      "epoch": 0.9779357220487519,
      "grad_norm": 0.15759406983852386,
      "learning_rate": 6.740577176679854e-05,
      "loss": 0.6185,
      "mean_token_accuracy": 0.8152636997401714,
      "num_tokens": 339232792.0,
      "step": 10629
    },
    {
      "entropy": 0.7250134348869324,
      "epoch": 0.9780277284201931,
      "grad_norm": 0.1499382108449936,
      "learning_rate": 6.740270494065692e-05,
      "loss": 0.72,
      "mean_token_accuracy": 0.788966242223978,
      "num_tokens": 339265366.0,
      "step": 10630
    },
    {
      "entropy": 0.7430151924490929,
      "epoch": 0.9781197347916343,
      "grad_norm": 0.15708082914352417,
      "learning_rate": 6.73996381145153e-05,
      "loss": 0.7319,
      "mean_token_accuracy": 0.7837865762412548,
      "num_tokens": 339296611.0,
      "step": 10631
    },
    {
      "entropy": 0.6650301441550255,
      "epoch": 0.9782117411630755,
      "grad_norm": 0.15015451610088348,
      "learning_rate": 6.739657128837367e-05,
      "loss": 0.65,
      "mean_token_accuracy": 0.8012652955949306,
      "num_tokens": 339328519.0,
      "step": 10632
    },
    {
      "entropy": 0.631567862816155,
      "epoch": 0.9783037475345168,
      "grad_norm": 0.14712867140769958,
      "learning_rate": 6.739350446223203e-05,
      "loss": 0.621,
      "mean_token_accuracy": 0.8192693889141083,
      "num_tokens": 339360456.0,
      "step": 10633
    },
    {
      "entropy": 0.6720016375184059,
      "epoch": 0.978395753905958,
      "grad_norm": 0.15067559480667114,
      "learning_rate": 6.73904376360904e-05,
      "loss": 0.6683,
      "mean_token_accuracy": 0.8012412339448929,
      "num_tokens": 339391976.0,
      "step": 10634
    },
    {
      "entropy": 0.680339464917779,
      "epoch": 0.9784877602773993,
      "grad_norm": 0.14898161590099335,
      "learning_rate": 6.73873708099488e-05,
      "loss": 0.6676,
      "mean_token_accuracy": 0.8007022142410278,
      "num_tokens": 339423961.0,
      "step": 10635
    },
    {
      "entropy": 0.8026669584214687,
      "epoch": 0.9785797666488404,
      "grad_norm": 0.1564677506685257,
      "learning_rate": 6.738430398380716e-05,
      "loss": 0.804,
      "mean_token_accuracy": 0.7714842669665813,
      "num_tokens": 339455771.0,
      "step": 10636
    },
    {
      "entropy": 0.7806021086871624,
      "epoch": 0.9786717730202816,
      "grad_norm": 0.1547614485025406,
      "learning_rate": 6.738123715766553e-05,
      "loss": 0.7759,
      "mean_token_accuracy": 0.7706379182636738,
      "num_tokens": 339487413.0,
      "step": 10637
    },
    {
      "entropy": 0.7693557813763618,
      "epoch": 0.9787637793917229,
      "grad_norm": 0.1616917997598648,
      "learning_rate": 6.73781703315239e-05,
      "loss": 0.7694,
      "mean_token_accuracy": 0.7737730257213116,
      "num_tokens": 339519297.0,
      "step": 10638
    },
    {
      "entropy": 0.58122737146914,
      "epoch": 0.9788557857631641,
      "grad_norm": 0.17865176498889923,
      "learning_rate": 6.737510350538228e-05,
      "loss": 0.5745,
      "mean_token_accuracy": 0.8251509852707386,
      "num_tokens": 339551253.0,
      "step": 10639
    },
    {
      "entropy": 0.6680003860965371,
      "epoch": 0.9789477921346054,
      "grad_norm": 0.15373562276363373,
      "learning_rate": 6.737203667924066e-05,
      "loss": 0.657,
      "mean_token_accuracy": 0.8019241280853748,
      "num_tokens": 339583151.0,
      "step": 10640
    },
    {
      "entropy": 0.6653646603226662,
      "epoch": 0.9790397985060465,
      "grad_norm": 0.15480153262615204,
      "learning_rate": 6.736896985309903e-05,
      "loss": 0.6639,
      "mean_token_accuracy": 0.8039392158389091,
      "num_tokens": 339615407.0,
      "step": 10641
    },
    {
      "entropy": 0.6428008042275906,
      "epoch": 0.9791318048774877,
      "grad_norm": 0.1487569659948349,
      "learning_rate": 6.73659030269574e-05,
      "loss": 0.6276,
      "mean_token_accuracy": 0.8083941899240017,
      "num_tokens": 339646864.0,
      "step": 10642
    },
    {
      "entropy": 0.742633230984211,
      "epoch": 0.979223811248929,
      "grad_norm": 0.1569424569606781,
      "learning_rate": 6.736283620081578e-05,
      "loss": 0.726,
      "mean_token_accuracy": 0.7838422544300556,
      "num_tokens": 339677868.0,
      "step": 10643
    },
    {
      "entropy": 0.7661631442606449,
      "epoch": 0.9793158176203702,
      "grad_norm": 0.1544497311115265,
      "learning_rate": 6.735976937467415e-05,
      "loss": 0.7621,
      "mean_token_accuracy": 0.7773768864572048,
      "num_tokens": 339710329.0,
      "step": 10644
    },
    {
      "entropy": 0.717248871922493,
      "epoch": 0.9794078239918115,
      "grad_norm": 0.16199244558811188,
      "learning_rate": 6.735670254853253e-05,
      "loss": 0.709,
      "mean_token_accuracy": 0.7878887839615345,
      "num_tokens": 339742010.0,
      "step": 10645
    },
    {
      "entropy": 0.8198088780045509,
      "epoch": 0.9794998303632526,
      "grad_norm": 0.1567835807800293,
      "learning_rate": 6.73536357223909e-05,
      "loss": 0.8061,
      "mean_token_accuracy": 0.7660497799515724,
      "num_tokens": 339773956.0,
      "step": 10646
    },
    {
      "entropy": 0.6666793972253799,
      "epoch": 0.9795918367346939,
      "grad_norm": 0.15395958721637726,
      "learning_rate": 6.735056889624928e-05,
      "loss": 0.6455,
      "mean_token_accuracy": 0.8097910322248936,
      "num_tokens": 339805915.0,
      "step": 10647
    },
    {
      "entropy": 0.6658129752613604,
      "epoch": 0.9796838431061351,
      "grad_norm": 0.15029458701610565,
      "learning_rate": 6.734750207010765e-05,
      "loss": 0.6407,
      "mean_token_accuracy": 0.8115718178451061,
      "num_tokens": 339838271.0,
      "step": 10648
    },
    {
      "entropy": 0.7826550453901291,
      "epoch": 0.9797758494775763,
      "grad_norm": 0.1553201824426651,
      "learning_rate": 6.734443524396602e-05,
      "loss": 0.7785,
      "mean_token_accuracy": 0.7709270305931568,
      "num_tokens": 339870878.0,
      "step": 10649
    },
    {
      "entropy": 0.7195312557742,
      "epoch": 0.9798678558490176,
      "grad_norm": 0.1492643505334854,
      "learning_rate": 6.73413684178244e-05,
      "loss": 0.7059,
      "mean_token_accuracy": 0.7891606464982033,
      "num_tokens": 339902985.0,
      "step": 10650
    },
    {
      "entropy": 0.6838386887684464,
      "epoch": 0.9799598622204587,
      "grad_norm": 0.1702621579170227,
      "learning_rate": 6.733830159168278e-05,
      "loss": 0.6513,
      "mean_token_accuracy": 0.8040628433227539,
      "num_tokens": 339934794.0,
      "step": 10651
    },
    {
      "entropy": 0.7162658143788576,
      "epoch": 0.9800518685919,
      "grad_norm": 0.16132943332195282,
      "learning_rate": 6.733523476554115e-05,
      "loss": 0.7166,
      "mean_token_accuracy": 0.784760408103466,
      "num_tokens": 339966584.0,
      "step": 10652
    },
    {
      "entropy": 0.7376744374632835,
      "epoch": 0.9801438749633412,
      "grad_norm": 0.15600518882274628,
      "learning_rate": 6.733216793939951e-05,
      "loss": 0.7287,
      "mean_token_accuracy": 0.7850036844611168,
      "num_tokens": 339998155.0,
      "step": 10653
    },
    {
      "entropy": 0.708761977031827,
      "epoch": 0.9802358813347825,
      "grad_norm": 0.1618315726518631,
      "learning_rate": 6.73291011132579e-05,
      "loss": 0.6971,
      "mean_token_accuracy": 0.79369742795825,
      "num_tokens": 340029903.0,
      "step": 10654
    },
    {
      "entropy": 0.6835060566663742,
      "epoch": 0.9803278877062237,
      "grad_norm": 0.15703396499156952,
      "learning_rate": 6.732603428711626e-05,
      "loss": 0.6872,
      "mean_token_accuracy": 0.7945076413452625,
      "num_tokens": 340061691.0,
      "step": 10655
    },
    {
      "entropy": 0.6371063506230712,
      "epoch": 0.9804198940776648,
      "grad_norm": 0.15215326845645905,
      "learning_rate": 6.732296746097464e-05,
      "loss": 0.6381,
      "mean_token_accuracy": 0.8198533095419407,
      "num_tokens": 340093831.0,
      "step": 10656
    },
    {
      "entropy": 0.7259279992431402,
      "epoch": 0.9805119004491061,
      "grad_norm": 0.15343718230724335,
      "learning_rate": 6.731990063483301e-05,
      "loss": 0.7261,
      "mean_token_accuracy": 0.7826810516417027,
      "num_tokens": 340125873.0,
      "step": 10657
    },
    {
      "entropy": 0.6564228171482682,
      "epoch": 0.9806039068205473,
      "grad_norm": 0.14669272303581238,
      "learning_rate": 6.73168338086914e-05,
      "loss": 0.6523,
      "mean_token_accuracy": 0.803366232663393,
      "num_tokens": 340158341.0,
      "step": 10658
    },
    {
      "entropy": 0.6347621958702803,
      "epoch": 0.9806959131919886,
      "grad_norm": 0.15829682350158691,
      "learning_rate": 6.731376698254976e-05,
      "loss": 0.6215,
      "mean_token_accuracy": 0.8155990391969681,
      "num_tokens": 340191061.0,
      "step": 10659
    },
    {
      "entropy": 0.6350116906687617,
      "epoch": 0.9807879195634298,
      "grad_norm": 0.14971430599689484,
      "learning_rate": 6.731070015640813e-05,
      "loss": 0.6211,
      "mean_token_accuracy": 0.8175940215587616,
      "num_tokens": 340222913.0,
      "step": 10660
    },
    {
      "entropy": 0.6956046745181084,
      "epoch": 0.9808799259348709,
      "grad_norm": 0.14982624351978302,
      "learning_rate": 6.730763333026651e-05,
      "loss": 0.6863,
      "mean_token_accuracy": 0.7971460521221161,
      "num_tokens": 340255144.0,
      "step": 10661
    },
    {
      "entropy": 0.6208344101905823,
      "epoch": 0.9809719323063122,
      "grad_norm": 0.1541215032339096,
      "learning_rate": 6.730456650412489e-05,
      "loss": 0.603,
      "mean_token_accuracy": 0.8185001760721207,
      "num_tokens": 340287486.0,
      "step": 10662
    },
    {
      "entropy": 0.7035593800246716,
      "epoch": 0.9810639386777534,
      "grad_norm": 0.15996944904327393,
      "learning_rate": 6.730149967798326e-05,
      "loss": 0.6927,
      "mean_token_accuracy": 0.7992163375020027,
      "num_tokens": 340319257.0,
      "step": 10663
    },
    {
      "entropy": 0.6475210161879659,
      "epoch": 0.9811559450491947,
      "grad_norm": 0.14944489300251007,
      "learning_rate": 6.729843285184163e-05,
      "loss": 0.624,
      "mean_token_accuracy": 0.8116655796766281,
      "num_tokens": 340351278.0,
      "step": 10664
    },
    {
      "entropy": 0.7395369587466121,
      "epoch": 0.9812479514206359,
      "grad_norm": 0.16104325652122498,
      "learning_rate": 6.72953660257e-05,
      "loss": 0.7257,
      "mean_token_accuracy": 0.7832578308880329,
      "num_tokens": 340382340.0,
      "step": 10665
    },
    {
      "entropy": 0.5317175472155213,
      "epoch": 0.9813399577920771,
      "grad_norm": 0.1439451426267624,
      "learning_rate": 6.729229919955838e-05,
      "loss": 0.4974,
      "mean_token_accuracy": 0.8496740385890007,
      "num_tokens": 340414834.0,
      "step": 10666
    },
    {
      "entropy": 0.811916159465909,
      "epoch": 0.9814319641635183,
      "grad_norm": 0.15855883061885834,
      "learning_rate": 6.728923237341676e-05,
      "loss": 0.8043,
      "mean_token_accuracy": 0.7632581554353237,
      "num_tokens": 340446888.0,
      "step": 10667
    },
    {
      "entropy": 0.6885384954512119,
      "epoch": 0.9815239705349595,
      "grad_norm": 0.15737544000148773,
      "learning_rate": 6.728616554727513e-05,
      "loss": 0.6847,
      "mean_token_accuracy": 0.7977384068071842,
      "num_tokens": 340478909.0,
      "step": 10668
    },
    {
      "entropy": 0.7664950825273991,
      "epoch": 0.9816159769064008,
      "grad_norm": 0.15581783652305603,
      "learning_rate": 6.72830987211335e-05,
      "loss": 0.7523,
      "mean_token_accuracy": 0.780837893486023,
      "num_tokens": 340511373.0,
      "step": 10669
    },
    {
      "entropy": 0.78896295838058,
      "epoch": 0.981707983277842,
      "grad_norm": 0.1547204852104187,
      "learning_rate": 6.728003189499188e-05,
      "loss": 0.7766,
      "mean_token_accuracy": 0.7691548354923725,
      "num_tokens": 340543360.0,
      "step": 10670
    },
    {
      "entropy": 0.7590831182897091,
      "epoch": 0.9817999896492832,
      "grad_norm": 0.16884124279022217,
      "learning_rate": 6.727696506885026e-05,
      "loss": 0.7463,
      "mean_token_accuracy": 0.7795246317982674,
      "num_tokens": 340574398.0,
      "step": 10671
    },
    {
      "entropy": 0.6365676112473011,
      "epoch": 0.9818919960207244,
      "grad_norm": 0.16068392992019653,
      "learning_rate": 6.727389824270863e-05,
      "loss": 0.6211,
      "mean_token_accuracy": 0.8154824897646904,
      "num_tokens": 340606610.0,
      "step": 10672
    },
    {
      "entropy": 0.6746859988197684,
      "epoch": 0.9819840023921657,
      "grad_norm": 0.15184547007083893,
      "learning_rate": 6.727083141656699e-05,
      "loss": 0.6637,
      "mean_token_accuracy": 0.8043478205800056,
      "num_tokens": 340639378.0,
      "step": 10673
    },
    {
      "entropy": 0.8352789860218763,
      "epoch": 0.9820760087636069,
      "grad_norm": 0.15795911848545074,
      "learning_rate": 6.726776459042537e-05,
      "loss": 0.8354,
      "mean_token_accuracy": 0.7557978183031082,
      "num_tokens": 340671123.0,
      "step": 10674
    },
    {
      "entropy": 0.5773513941094279,
      "epoch": 0.9821680151350481,
      "grad_norm": 0.15004587173461914,
      "learning_rate": 6.726469776428374e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8305414877831936,
      "num_tokens": 340703411.0,
      "step": 10675
    },
    {
      "entropy": 0.6845098277553916,
      "epoch": 0.9822600215064893,
      "grad_norm": 0.14998675882816315,
      "learning_rate": 6.726163093814212e-05,
      "loss": 0.6661,
      "mean_token_accuracy": 0.8026098683476448,
      "num_tokens": 340735087.0,
      "step": 10676
    },
    {
      "entropy": 0.7267597084864974,
      "epoch": 0.9823520278779305,
      "grad_norm": 0.15167909860610962,
      "learning_rate": 6.72585641120005e-05,
      "loss": 0.716,
      "mean_token_accuracy": 0.7872472107410431,
      "num_tokens": 340767294.0,
      "step": 10677
    },
    {
      "entropy": 0.6687714830040932,
      "epoch": 0.9824440342493718,
      "grad_norm": 0.14570148289203644,
      "learning_rate": 6.725549728585887e-05,
      "loss": 0.6371,
      "mean_token_accuracy": 0.8044130727648735,
      "num_tokens": 340799377.0,
      "step": 10678
    },
    {
      "entropy": 0.7174546960741282,
      "epoch": 0.982536040620813,
      "grad_norm": 0.15991050004959106,
      "learning_rate": 6.725243045971724e-05,
      "loss": 0.7121,
      "mean_token_accuracy": 0.7910387143492699,
      "num_tokens": 340831641.0,
      "step": 10679
    },
    {
      "entropy": 0.7398865707218647,
      "epoch": 0.9826280469922543,
      "grad_norm": 0.1677507609128952,
      "learning_rate": 6.724936363357561e-05,
      "loss": 0.7363,
      "mean_token_accuracy": 0.7848191149532795,
      "num_tokens": 340863639.0,
      "step": 10680
    },
    {
      "entropy": 0.60740558616817,
      "epoch": 0.9827200533636954,
      "grad_norm": 0.15986360609531403,
      "learning_rate": 6.724629680743399e-05,
      "loss": 0.5821,
      "mean_token_accuracy": 0.8246910274028778,
      "num_tokens": 340895603.0,
      "step": 10681
    },
    {
      "entropy": 0.6945192981511354,
      "epoch": 0.9828120597351366,
      "grad_norm": 0.15827642381191254,
      "learning_rate": 6.724322998129237e-05,
      "loss": 0.692,
      "mean_token_accuracy": 0.7932898104190826,
      "num_tokens": 340927587.0,
      "step": 10682
    },
    {
      "entropy": 0.6769187282770872,
      "epoch": 0.9829040661065779,
      "grad_norm": 0.15207941830158234,
      "learning_rate": 6.724016315515074e-05,
      "loss": 0.6746,
      "mean_token_accuracy": 0.7956702373921871,
      "num_tokens": 340959220.0,
      "step": 10683
    },
    {
      "entropy": 0.6096819452941418,
      "epoch": 0.9829960724780191,
      "grad_norm": 0.14726604521274567,
      "learning_rate": 6.723709632900911e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.823176022619009,
      "num_tokens": 340991277.0,
      "step": 10684
    },
    {
      "entropy": 0.8270751256495714,
      "epoch": 0.9830880788494604,
      "grad_norm": 0.1612326204776764,
      "learning_rate": 6.723402950286749e-05,
      "loss": 0.8155,
      "mean_token_accuracy": 0.7606712207198143,
      "num_tokens": 341022717.0,
      "step": 10685
    },
    {
      "entropy": 0.701416989788413,
      "epoch": 0.9831800852209015,
      "grad_norm": 0.15029048919677734,
      "learning_rate": 6.723096267672586e-05,
      "loss": 0.6809,
      "mean_token_accuracy": 0.7946742922067642,
      "num_tokens": 341055221.0,
      "step": 10686
    },
    {
      "entropy": 0.8002241607755423,
      "epoch": 0.9832720915923427,
      "grad_norm": 0.15925928950309753,
      "learning_rate": 6.722789585058424e-05,
      "loss": 0.8066,
      "mean_token_accuracy": 0.7639710269868374,
      "num_tokens": 341087423.0,
      "step": 10687
    },
    {
      "entropy": 0.7077586334198713,
      "epoch": 0.983364097963784,
      "grad_norm": 0.15047641098499298,
      "learning_rate": 6.72248290244426e-05,
      "loss": 0.6983,
      "mean_token_accuracy": 0.7926603928208351,
      "num_tokens": 341119931.0,
      "step": 10688
    },
    {
      "entropy": 0.6931829694658518,
      "epoch": 0.9834561043352252,
      "grad_norm": 0.15825660526752472,
      "learning_rate": 6.722176219830099e-05,
      "loss": 0.6822,
      "mean_token_accuracy": 0.7966234050691128,
      "num_tokens": 341152489.0,
      "step": 10689
    },
    {
      "entropy": 0.6130394442006946,
      "epoch": 0.9835481107066665,
      "grad_norm": 0.15189363062381744,
      "learning_rate": 6.721869537215936e-05,
      "loss": 0.5959,
      "mean_token_accuracy": 0.8189002573490143,
      "num_tokens": 341183335.0,
      "step": 10690
    },
    {
      "entropy": 0.7160324510186911,
      "epoch": 0.9836401170781076,
      "grad_norm": 0.14388436079025269,
      "learning_rate": 6.721562854601772e-05,
      "loss": 0.7016,
      "mean_token_accuracy": 0.7919887155294418,
      "num_tokens": 341215554.0,
      "step": 10691
    },
    {
      "entropy": 0.79881826415658,
      "epoch": 0.9837321234495489,
      "grad_norm": 0.15636423230171204,
      "learning_rate": 6.72125617198761e-05,
      "loss": 0.7942,
      "mean_token_accuracy": 0.7649000771343708,
      "num_tokens": 341246080.0,
      "step": 10692
    },
    {
      "entropy": 0.6974719683639705,
      "epoch": 0.9838241298209901,
      "grad_norm": 0.15531207621097565,
      "learning_rate": 6.720949489373449e-05,
      "loss": 0.6898,
      "mean_token_accuracy": 0.7952987514436245,
      "num_tokens": 341278221.0,
      "step": 10693
    },
    {
      "entropy": 0.6482951007783413,
      "epoch": 0.9839161361924313,
      "grad_norm": 0.16126228868961334,
      "learning_rate": 6.720642806759285e-05,
      "loss": 0.6307,
      "mean_token_accuracy": 0.8080011010169983,
      "num_tokens": 341310323.0,
      "step": 10694
    },
    {
      "entropy": 0.6941235959529877,
      "epoch": 0.9840081425638726,
      "grad_norm": 0.14790797233581543,
      "learning_rate": 6.720336124145122e-05,
      "loss": 0.6926,
      "mean_token_accuracy": 0.7953289970755577,
      "num_tokens": 341342559.0,
      "step": 10695
    },
    {
      "entropy": 0.7635149266570807,
      "epoch": 0.9841001489353137,
      "grad_norm": 0.15529868006706238,
      "learning_rate": 6.720029441530959e-05,
      "loss": 0.7727,
      "mean_token_accuracy": 0.7701333314180374,
      "num_tokens": 341374586.0,
      "step": 10696
    },
    {
      "entropy": 0.6859048530459404,
      "epoch": 0.984192155306755,
      "grad_norm": 0.15027043223381042,
      "learning_rate": 6.719722758916797e-05,
      "loss": 0.6823,
      "mean_token_accuracy": 0.7976969107985497,
      "num_tokens": 341405874.0,
      "step": 10697
    },
    {
      "entropy": 0.7358623333275318,
      "epoch": 0.9842841616781962,
      "grad_norm": 0.16095931828022003,
      "learning_rate": 6.719416076302635e-05,
      "loss": 0.7305,
      "mean_token_accuracy": 0.7860477045178413,
      "num_tokens": 341437892.0,
      "step": 10698
    },
    {
      "entropy": 0.644883319735527,
      "epoch": 0.9843761680496375,
      "grad_norm": 0.14940939843654633,
      "learning_rate": 6.719109393688472e-05,
      "loss": 0.6192,
      "mean_token_accuracy": 0.8117432072758675,
      "num_tokens": 341469712.0,
      "step": 10699
    },
    {
      "entropy": 0.6901897192001343,
      "epoch": 0.9844681744210787,
      "grad_norm": 0.15333226323127747,
      "learning_rate": 6.718802711074309e-05,
      "loss": 0.6882,
      "mean_token_accuracy": 0.7910367660224438,
      "num_tokens": 341501171.0,
      "step": 10700
    },
    {
      "entropy": 0.7035043984651566,
      "epoch": 0.9845601807925198,
      "grad_norm": 0.1523108333349228,
      "learning_rate": 6.718496028460147e-05,
      "loss": 0.7149,
      "mean_token_accuracy": 0.7885532379150391,
      "num_tokens": 341532905.0,
      "step": 10701
    },
    {
      "entropy": 0.7175933048129082,
      "epoch": 0.9846521871639611,
      "grad_norm": 0.15437746047973633,
      "learning_rate": 6.718189345845984e-05,
      "loss": 0.6884,
      "mean_token_accuracy": 0.7947478108108044,
      "num_tokens": 341565121.0,
      "step": 10702
    },
    {
      "entropy": 0.5460159163922071,
      "epoch": 0.9847441935354023,
      "grad_norm": 0.14516349136829376,
      "learning_rate": 6.717882663231822e-05,
      "loss": 0.5284,
      "mean_token_accuracy": 0.8421624004840851,
      "num_tokens": 341596625.0,
      "step": 10703
    },
    {
      "entropy": 0.6963630188256502,
      "epoch": 0.9848361999068436,
      "grad_norm": 0.1680130660533905,
      "learning_rate": 6.717575980617659e-05,
      "loss": 0.6906,
      "mean_token_accuracy": 0.7966231033205986,
      "num_tokens": 341629393.0,
      "step": 10704
    },
    {
      "entropy": 0.6057257149368525,
      "epoch": 0.9849282062782848,
      "grad_norm": 0.14986005425453186,
      "learning_rate": 6.717269298003497e-05,
      "loss": 0.5893,
      "mean_token_accuracy": 0.8268276751041412,
      "num_tokens": 341661229.0,
      "step": 10705
    },
    {
      "entropy": 0.591246142052114,
      "epoch": 0.985020212649726,
      "grad_norm": 0.13910962641239166,
      "learning_rate": 6.716962615389334e-05,
      "loss": 0.5706,
      "mean_token_accuracy": 0.8290906623005867,
      "num_tokens": 341693930.0,
      "step": 10706
    },
    {
      "entropy": 0.634125187061727,
      "epoch": 0.9851122190211672,
      "grad_norm": 0.14676249027252197,
      "learning_rate": 6.71665593277517e-05,
      "loss": 0.6112,
      "mean_token_accuracy": 0.8137099519371986,
      "num_tokens": 341725483.0,
      "step": 10707
    },
    {
      "entropy": 0.6431856136769056,
      "epoch": 0.9852042253926084,
      "grad_norm": 0.15026801824569702,
      "learning_rate": 6.71634925016101e-05,
      "loss": 0.6192,
      "mean_token_accuracy": 0.8166288249194622,
      "num_tokens": 341757497.0,
      "step": 10708
    },
    {
      "entropy": 0.7354996390640736,
      "epoch": 0.9852962317640497,
      "grad_norm": 0.1569245606660843,
      "learning_rate": 6.716042567546847e-05,
      "loss": 0.7282,
      "mean_token_accuracy": 0.7839795649051666,
      "num_tokens": 341789399.0,
      "step": 10709
    },
    {
      "entropy": 0.6744760097935796,
      "epoch": 0.9853882381354909,
      "grad_norm": 0.15067873895168304,
      "learning_rate": 6.715735884932683e-05,
      "loss": 0.6615,
      "mean_token_accuracy": 0.8031509630382061,
      "num_tokens": 341821859.0,
      "step": 10710
    },
    {
      "entropy": 0.7357632834464312,
      "epoch": 0.9854802445069321,
      "grad_norm": 0.17317534983158112,
      "learning_rate": 6.71542920231852e-05,
      "loss": 0.7538,
      "mean_token_accuracy": 0.7801604233682156,
      "num_tokens": 341854475.0,
      "step": 10711
    },
    {
      "entropy": 0.675763227045536,
      "epoch": 0.9855722508783733,
      "grad_norm": 0.15033851563930511,
      "learning_rate": 6.715122519704358e-05,
      "loss": 0.6585,
      "mean_token_accuracy": 0.8007294982671738,
      "num_tokens": 341886185.0,
      "step": 10712
    },
    {
      "entropy": 0.744746396318078,
      "epoch": 0.9856642572498145,
      "grad_norm": 0.16166594624519348,
      "learning_rate": 6.714815837090197e-05,
      "loss": 0.7347,
      "mean_token_accuracy": 0.7871145308017731,
      "num_tokens": 341917609.0,
      "step": 10713
    },
    {
      "entropy": 0.6038727629929781,
      "epoch": 0.9857562636212558,
      "grad_norm": 0.1602073609828949,
      "learning_rate": 6.714509154476033e-05,
      "loss": 0.5895,
      "mean_token_accuracy": 0.8248329311609268,
      "num_tokens": 341949266.0,
      "step": 10714
    },
    {
      "entropy": 0.6711984761059284,
      "epoch": 0.985848269992697,
      "grad_norm": 0.14792989194393158,
      "learning_rate": 6.71420247186187e-05,
      "loss": 0.6539,
      "mean_token_accuracy": 0.8015061318874359,
      "num_tokens": 341981378.0,
      "step": 10715
    },
    {
      "entropy": 0.7429183339700103,
      "epoch": 0.9859402763641382,
      "grad_norm": 0.15745660662651062,
      "learning_rate": 6.713895789247708e-05,
      "loss": 0.7442,
      "mean_token_accuracy": 0.7791816145181656,
      "num_tokens": 342012920.0,
      "step": 10716
    },
    {
      "entropy": 0.6340972045436502,
      "epoch": 0.9860322827355794,
      "grad_norm": 0.15954308211803436,
      "learning_rate": 6.713589106633545e-05,
      "loss": 0.621,
      "mean_token_accuracy": 0.8151038847863674,
      "num_tokens": 342045155.0,
      "step": 10717
    },
    {
      "entropy": 0.6251770742237568,
      "epoch": 0.9861242891070207,
      "grad_norm": 0.1496426910161972,
      "learning_rate": 6.713282424019383e-05,
      "loss": 0.5937,
      "mean_token_accuracy": 0.8226956687867641,
      "num_tokens": 342077225.0,
      "step": 10718
    },
    {
      "entropy": 0.7192734554409981,
      "epoch": 0.9862162954784619,
      "grad_norm": 0.15465523302555084,
      "learning_rate": 6.71297574140522e-05,
      "loss": 0.6936,
      "mean_token_accuracy": 0.7935804724693298,
      "num_tokens": 342108606.0,
      "step": 10719
    },
    {
      "entropy": 0.7618325632065535,
      "epoch": 0.9863083018499031,
      "grad_norm": 0.1574353724718094,
      "learning_rate": 6.712669058791058e-05,
      "loss": 0.7527,
      "mean_token_accuracy": 0.7777319364249706,
      "num_tokens": 342140247.0,
      "step": 10720
    },
    {
      "entropy": 0.8123092018067837,
      "epoch": 0.9864003082213443,
      "grad_norm": 0.16468827426433563,
      "learning_rate": 6.712362376176895e-05,
      "loss": 0.7864,
      "mean_token_accuracy": 0.767015628516674,
      "num_tokens": 342171689.0,
      "step": 10721
    },
    {
      "entropy": 0.6636488363146782,
      "epoch": 0.9864923145927855,
      "grad_norm": 0.14738401770591736,
      "learning_rate": 6.712055693562732e-05,
      "loss": 0.6353,
      "mean_token_accuracy": 0.8134344816207886,
      "num_tokens": 342204417.0,
      "step": 10722
    },
    {
      "entropy": 0.7238919101655483,
      "epoch": 0.9865843209642268,
      "grad_norm": 0.16132871806621552,
      "learning_rate": 6.71174901094857e-05,
      "loss": 0.7095,
      "mean_token_accuracy": 0.7883936651051044,
      "num_tokens": 342235678.0,
      "step": 10723
    },
    {
      "entropy": 0.7347845742478967,
      "epoch": 0.986676327335668,
      "grad_norm": 0.15566398203372955,
      "learning_rate": 6.711442328334408e-05,
      "loss": 0.7315,
      "mean_token_accuracy": 0.7783857360482216,
      "num_tokens": 342267217.0,
      "step": 10724
    },
    {
      "entropy": 0.6981153748929501,
      "epoch": 0.9867683337071093,
      "grad_norm": 0.1591978520154953,
      "learning_rate": 6.711135645720245e-05,
      "loss": 0.6952,
      "mean_token_accuracy": 0.795062966644764,
      "num_tokens": 342299673.0,
      "step": 10725
    },
    {
      "entropy": 0.8025952689349651,
      "epoch": 0.9868603400785504,
      "grad_norm": 0.1569969356060028,
      "learning_rate": 6.710828963106082e-05,
      "loss": 0.8002,
      "mean_token_accuracy": 0.7668984606862068,
      "num_tokens": 342330915.0,
      "step": 10726
    },
    {
      "entropy": 0.6547057349234819,
      "epoch": 0.9869523464499916,
      "grad_norm": 0.1571449339389801,
      "learning_rate": 6.710522280491918e-05,
      "loss": 0.6643,
      "mean_token_accuracy": 0.8040328361093998,
      "num_tokens": 342362591.0,
      "step": 10727
    },
    {
      "entropy": 0.5869275350123644,
      "epoch": 0.9870443528214329,
      "grad_norm": 0.15489153563976288,
      "learning_rate": 6.710215597877756e-05,
      "loss": 0.5786,
      "mean_token_accuracy": 0.8224467448890209,
      "num_tokens": 342393934.0,
      "step": 10728
    },
    {
      "entropy": 0.7637748457491398,
      "epoch": 0.9871363591928741,
      "grad_norm": 0.1553264558315277,
      "learning_rate": 6.709908915263595e-05,
      "loss": 0.7409,
      "mean_token_accuracy": 0.7803970724344254,
      "num_tokens": 342425696.0,
      "step": 10729
    },
    {
      "entropy": 0.6619486603885889,
      "epoch": 0.9872283655643154,
      "grad_norm": 0.15594930946826935,
      "learning_rate": 6.709602232649431e-05,
      "loss": 0.6476,
      "mean_token_accuracy": 0.8037876412272453,
      "num_tokens": 342457194.0,
      "step": 10730
    },
    {
      "entropy": 0.6627030745148659,
      "epoch": 0.9873203719357565,
      "grad_norm": 0.14798647165298462,
      "learning_rate": 6.709295550035268e-05,
      "loss": 0.6452,
      "mean_token_accuracy": 0.8029203861951828,
      "num_tokens": 342489676.0,
      "step": 10731
    },
    {
      "entropy": 0.7620030697435141,
      "epoch": 0.9874123783071977,
      "grad_norm": 0.15814760327339172,
      "learning_rate": 6.708988867421106e-05,
      "loss": 0.7562,
      "mean_token_accuracy": 0.7775351032614708,
      "num_tokens": 342521050.0,
      "step": 10732
    },
    {
      "entropy": 0.7577329576015472,
      "epoch": 0.987504384678639,
      "grad_norm": 0.15735454857349396,
      "learning_rate": 6.708682184806943e-05,
      "loss": 0.7389,
      "mean_token_accuracy": 0.7791989520192146,
      "num_tokens": 342552676.0,
      "step": 10733
    },
    {
      "entropy": 0.6861260319128633,
      "epoch": 0.9875963910500802,
      "grad_norm": 0.14998988807201385,
      "learning_rate": 6.708375502192781e-05,
      "loss": 0.6745,
      "mean_token_accuracy": 0.8019385971128941,
      "num_tokens": 342584694.0,
      "step": 10734
    },
    {
      "entropy": 0.7339207902550697,
      "epoch": 0.9876883974215215,
      "grad_norm": 0.16129682958126068,
      "learning_rate": 6.708068819578618e-05,
      "loss": 0.7093,
      "mean_token_accuracy": 0.7891803681850433,
      "num_tokens": 342616261.0,
      "step": 10735
    },
    {
      "entropy": 0.6689471062272787,
      "epoch": 0.9877804037929626,
      "grad_norm": 0.14422205090522766,
      "learning_rate": 6.707762136964456e-05,
      "loss": 0.6555,
      "mean_token_accuracy": 0.8031104765832424,
      "num_tokens": 342648029.0,
      "step": 10736
    },
    {
      "entropy": 0.7065583765506744,
      "epoch": 0.9878724101644039,
      "grad_norm": 0.14937640726566315,
      "learning_rate": 6.707455454350293e-05,
      "loss": 0.7076,
      "mean_token_accuracy": 0.7923536412417889,
      "num_tokens": 342680291.0,
      "step": 10737
    },
    {
      "entropy": 0.8047258090227842,
      "epoch": 0.9879644165358451,
      "grad_norm": 0.1558113992214203,
      "learning_rate": 6.70714877173613e-05,
      "loss": 0.8052,
      "mean_token_accuracy": 0.763905830681324,
      "num_tokens": 342712259.0,
      "step": 10738
    },
    {
      "entropy": 0.6521598761901259,
      "epoch": 0.9880564229072863,
      "grad_norm": 0.14718444645404816,
      "learning_rate": 6.706842089121968e-05,
      "loss": 0.6284,
      "mean_token_accuracy": 0.8183500468730927,
      "num_tokens": 342744109.0,
      "step": 10739
    },
    {
      "entropy": 0.7290212623775005,
      "epoch": 0.9881484292787276,
      "grad_norm": 0.1555648148059845,
      "learning_rate": 6.706535406507806e-05,
      "loss": 0.7004,
      "mean_token_accuracy": 0.7870186977088451,
      "num_tokens": 342775787.0,
      "step": 10740
    },
    {
      "entropy": 0.5731196776032448,
      "epoch": 0.9882404356501687,
      "grad_norm": 0.14721882343292236,
      "learning_rate": 6.706228723893643e-05,
      "loss": 0.557,
      "mean_token_accuracy": 0.8357735462486744,
      "num_tokens": 342808185.0,
      "step": 10741
    },
    {
      "entropy": 0.7122553996741772,
      "epoch": 0.98833244202161,
      "grad_norm": 0.1551104336977005,
      "learning_rate": 6.70592204127948e-05,
      "loss": 0.7042,
      "mean_token_accuracy": 0.7947388589382172,
      "num_tokens": 342840595.0,
      "step": 10742
    },
    {
      "entropy": 0.6804009452462196,
      "epoch": 0.9884244483930512,
      "grad_norm": 0.15448178350925446,
      "learning_rate": 6.705615358665318e-05,
      "loss": 0.6689,
      "mean_token_accuracy": 0.7970963716506958,
      "num_tokens": 342872682.0,
      "step": 10743
    },
    {
      "entropy": 0.6680690040811896,
      "epoch": 0.9885164547644925,
      "grad_norm": 0.1461191028356552,
      "learning_rate": 6.705308676051155e-05,
      "loss": 0.6473,
      "mean_token_accuracy": 0.8055110238492489,
      "num_tokens": 342904291.0,
      "step": 10744
    },
    {
      "entropy": 0.6767096184194088,
      "epoch": 0.9886084611359337,
      "grad_norm": 0.174092099070549,
      "learning_rate": 6.705001993436993e-05,
      "loss": 0.6848,
      "mean_token_accuracy": 0.7942872121930122,
      "num_tokens": 342936682.0,
      "step": 10745
    },
    {
      "entropy": 0.7388930274173617,
      "epoch": 0.9887004675073748,
      "grad_norm": 0.15219688415527344,
      "learning_rate": 6.70469531082283e-05,
      "loss": 0.7438,
      "mean_token_accuracy": 0.7853900454938412,
      "num_tokens": 342968630.0,
      "step": 10746
    },
    {
      "entropy": 0.6837972048670053,
      "epoch": 0.9887924738788161,
      "grad_norm": 0.15207479894161224,
      "learning_rate": 6.704388628208668e-05,
      "loss": 0.663,
      "mean_token_accuracy": 0.800333920866251,
      "num_tokens": 343000494.0,
      "step": 10747
    },
    {
      "entropy": 0.7458132058382034,
      "epoch": 0.9888844802502573,
      "grad_norm": 0.15495915710926056,
      "learning_rate": 6.704081945594504e-05,
      "loss": 0.7506,
      "mean_token_accuracy": 0.7780111655592918,
      "num_tokens": 343031924.0,
      "step": 10748
    },
    {
      "entropy": 0.785844910889864,
      "epoch": 0.9889764866216986,
      "grad_norm": 0.15896983444690704,
      "learning_rate": 6.703775262980341e-05,
      "loss": 0.7833,
      "mean_token_accuracy": 0.765913300216198,
      "num_tokens": 343063819.0,
      "step": 10749
    },
    {
      "entropy": 0.6084613744169474,
      "epoch": 0.9890684929931398,
      "grad_norm": 0.1541769653558731,
      "learning_rate": 6.70346858036618e-05,
      "loss": 0.5962,
      "mean_token_accuracy": 0.8217114247381687,
      "num_tokens": 343095811.0,
      "step": 10750
    },
    {
      "entropy": 0.722863707691431,
      "epoch": 0.989160499364581,
      "grad_norm": 0.15148432552814484,
      "learning_rate": 6.703161897752017e-05,
      "loss": 0.7109,
      "mean_token_accuracy": 0.7864315584301949,
      "num_tokens": 343127055.0,
      "step": 10751
    },
    {
      "entropy": 0.6440486563369632,
      "epoch": 0.9892525057360222,
      "grad_norm": 0.14748679101467133,
      "learning_rate": 6.702855215137854e-05,
      "loss": 0.6347,
      "mean_token_accuracy": 0.8150553852319717,
      "num_tokens": 343158974.0,
      "step": 10752
    },
    {
      "entropy": 0.6829834477975965,
      "epoch": 0.9893445121074634,
      "grad_norm": 0.15272484719753265,
      "learning_rate": 6.702548532523691e-05,
      "loss": 0.6612,
      "mean_token_accuracy": 0.8025909699499607,
      "num_tokens": 343191567.0,
      "step": 10753
    },
    {
      "entropy": 0.8635164052248001,
      "epoch": 0.9894365184789047,
      "grad_norm": 0.1613001972436905,
      "learning_rate": 6.702241849909528e-05,
      "loss": 0.8574,
      "mean_token_accuracy": 0.7545499727129936,
      "num_tokens": 343222875.0,
      "step": 10754
    },
    {
      "entropy": 0.7369737103581429,
      "epoch": 0.9895285248503459,
      "grad_norm": 0.16535404324531555,
      "learning_rate": 6.701935167295367e-05,
      "loss": 0.7381,
      "mean_token_accuracy": 0.7782619521021843,
      "num_tokens": 343254319.0,
      "step": 10755
    },
    {
      "entropy": 0.7962956931442022,
      "epoch": 0.9896205312217871,
      "grad_norm": 0.16246114671230316,
      "learning_rate": 6.701628484681204e-05,
      "loss": 0.7784,
      "mean_token_accuracy": 0.7698286324739456,
      "num_tokens": 343286009.0,
      "step": 10756
    },
    {
      "entropy": 0.7675060164183378,
      "epoch": 0.9897125375932283,
      "grad_norm": 0.15703585743904114,
      "learning_rate": 6.701321802067041e-05,
      "loss": 0.7514,
      "mean_token_accuracy": 0.7756641656160355,
      "num_tokens": 343317736.0,
      "step": 10757
    },
    {
      "entropy": 0.7040892308577895,
      "epoch": 0.9898045439646695,
      "grad_norm": 0.15862968564033508,
      "learning_rate": 6.701015119452878e-05,
      "loss": 0.6907,
      "mean_token_accuracy": 0.7917059138417244,
      "num_tokens": 343350182.0,
      "step": 10758
    },
    {
      "entropy": 0.7692784853279591,
      "epoch": 0.9898965503361108,
      "grad_norm": 0.17698609828948975,
      "learning_rate": 6.700708436838716e-05,
      "loss": 0.758,
      "mean_token_accuracy": 0.7762006856501102,
      "num_tokens": 343382289.0,
      "step": 10759
    },
    {
      "entropy": 0.5945233330130577,
      "epoch": 0.989988556707552,
      "grad_norm": 0.1400427371263504,
      "learning_rate": 6.700401754224554e-05,
      "loss": 0.59,
      "mean_token_accuracy": 0.8221310190856457,
      "num_tokens": 343414178.0,
      "step": 10760
    },
    {
      "entropy": 0.6152955498546362,
      "epoch": 0.9900805630789932,
      "grad_norm": 0.14927798509597778,
      "learning_rate": 6.700095071610391e-05,
      "loss": 0.5952,
      "mean_token_accuracy": 0.8193485029041767,
      "num_tokens": 343446489.0,
      "step": 10761
    },
    {
      "entropy": 0.7141527123749256,
      "epoch": 0.9901725694504344,
      "grad_norm": 0.1583973914384842,
      "learning_rate": 6.699788388996228e-05,
      "loss": 0.6887,
      "mean_token_accuracy": 0.7976505011320114,
      "num_tokens": 343478569.0,
      "step": 10762
    },
    {
      "entropy": 0.6849739067256451,
      "epoch": 0.9902645758218757,
      "grad_norm": 0.14715206623077393,
      "learning_rate": 6.699481706382066e-05,
      "loss": 0.6653,
      "mean_token_accuracy": 0.8029139712452888,
      "num_tokens": 343510746.0,
      "step": 10763
    },
    {
      "entropy": 0.6760705541819334,
      "epoch": 0.9903565821933169,
      "grad_norm": 0.16324540972709656,
      "learning_rate": 6.699175023767902e-05,
      "loss": 0.6649,
      "mean_token_accuracy": 0.7997146211564541,
      "num_tokens": 343542612.0,
      "step": 10764
    },
    {
      "entropy": 0.6761240754276514,
      "epoch": 0.9904485885647581,
      "grad_norm": 0.14761677384376526,
      "learning_rate": 6.69886834115374e-05,
      "loss": 0.6609,
      "mean_token_accuracy": 0.8030462078750134,
      "num_tokens": 343575000.0,
      "step": 10765
    },
    {
      "entropy": 0.7475274335592985,
      "epoch": 0.9905405949361993,
      "grad_norm": 0.1581384688615799,
      "learning_rate": 6.698561658539577e-05,
      "loss": 0.7501,
      "mean_token_accuracy": 0.7808474749326706,
      "num_tokens": 343607176.0,
      "step": 10766
    },
    {
      "entropy": 0.7130754292011261,
      "epoch": 0.9906326013076405,
      "grad_norm": 0.1629365086555481,
      "learning_rate": 6.698254975925416e-05,
      "loss": 0.6959,
      "mean_token_accuracy": 0.7947229705750942,
      "num_tokens": 343639420.0,
      "step": 10767
    },
    {
      "entropy": 0.718221090734005,
      "epoch": 0.9907246076790818,
      "grad_norm": 0.16604968905448914,
      "learning_rate": 6.697948293311252e-05,
      "loss": 0.731,
      "mean_token_accuracy": 0.7878916822373867,
      "num_tokens": 343671486.0,
      "step": 10768
    },
    {
      "entropy": 0.7898632325232029,
      "epoch": 0.990816614050523,
      "grad_norm": 0.1605367213487625,
      "learning_rate": 6.697641610697089e-05,
      "loss": 0.7842,
      "mean_token_accuracy": 0.7727605551481247,
      "num_tokens": 343702667.0,
      "step": 10769
    },
    {
      "entropy": 0.6296033356338739,
      "epoch": 0.9909086204219643,
      "grad_norm": 0.16042953729629517,
      "learning_rate": 6.697334928082927e-05,
      "loss": 0.6232,
      "mean_token_accuracy": 0.8151320889592171,
      "num_tokens": 343734741.0,
      "step": 10770
    },
    {
      "entropy": 0.6668254286050797,
      "epoch": 0.9910006267934054,
      "grad_norm": 0.154945507645607,
      "learning_rate": 6.697028245468765e-05,
      "loss": 0.6637,
      "mean_token_accuracy": 0.804306261241436,
      "num_tokens": 343766882.0,
      "step": 10771
    },
    {
      "entropy": 0.6241018762812018,
      "epoch": 0.9910926331648466,
      "grad_norm": 0.1460465043783188,
      "learning_rate": 6.696721562854602e-05,
      "loss": 0.6013,
      "mean_token_accuracy": 0.8168587759137154,
      "num_tokens": 343798548.0,
      "step": 10772
    },
    {
      "entropy": 0.8466452434659004,
      "epoch": 0.9911846395362879,
      "grad_norm": 0.16484686732292175,
      "learning_rate": 6.696414880240439e-05,
      "loss": 0.8489,
      "mean_token_accuracy": 0.7563196681439877,
      "num_tokens": 343831123.0,
      "step": 10773
    },
    {
      "entropy": 0.5859060045331717,
      "epoch": 0.9912766459077291,
      "grad_norm": 0.14794258773326874,
      "learning_rate": 6.696108197626277e-05,
      "loss": 0.587,
      "mean_token_accuracy": 0.8244335204362869,
      "num_tokens": 343862952.0,
      "step": 10774
    },
    {
      "entropy": 0.7442157361656427,
      "epoch": 0.9913686522791704,
      "grad_norm": 0.15119753777980804,
      "learning_rate": 6.695801515012114e-05,
      "loss": 0.7307,
      "mean_token_accuracy": 0.7822629623115063,
      "num_tokens": 343895433.0,
      "step": 10775
    },
    {
      "entropy": 0.6606582626700401,
      "epoch": 0.9914606586506115,
      "grad_norm": 0.14821286499500275,
      "learning_rate": 6.695494832397952e-05,
      "loss": 0.6588,
      "mean_token_accuracy": 0.8001754730939865,
      "num_tokens": 343927663.0,
      "step": 10776
    },
    {
      "entropy": 0.5735927168279886,
      "epoch": 0.9915526650220527,
      "grad_norm": 0.15325070917606354,
      "learning_rate": 6.695188149783789e-05,
      "loss": 0.5497,
      "mean_token_accuracy": 0.8329834789037704,
      "num_tokens": 343960365.0,
      "step": 10777
    },
    {
      "entropy": 0.7826302107423544,
      "epoch": 0.991644671393494,
      "grad_norm": 0.16105858981609344,
      "learning_rate": 6.694881467169627e-05,
      "loss": 0.7761,
      "mean_token_accuracy": 0.7726006545126438,
      "num_tokens": 343992610.0,
      "step": 10778
    },
    {
      "entropy": 0.7275161556899548,
      "epoch": 0.9917366777649352,
      "grad_norm": 0.15739943087100983,
      "learning_rate": 6.694574784555464e-05,
      "loss": 0.7163,
      "mean_token_accuracy": 0.7864337973296642,
      "num_tokens": 344023823.0,
      "step": 10779
    },
    {
      "entropy": 0.6856575161218643,
      "epoch": 0.9918286841363765,
      "grad_norm": 0.1538325995206833,
      "learning_rate": 6.6942681019413e-05,
      "loss": 0.6909,
      "mean_token_accuracy": 0.7922129519283772,
      "num_tokens": 344055534.0,
      "step": 10780
    },
    {
      "entropy": 0.5891643664799631,
      "epoch": 0.9919206905078176,
      "grad_norm": 0.1508904993534088,
      "learning_rate": 6.693961419327139e-05,
      "loss": 0.5691,
      "mean_token_accuracy": 0.8265917561948299,
      "num_tokens": 344087687.0,
      "step": 10781
    },
    {
      "entropy": 0.7669612839818001,
      "epoch": 0.9920126968792589,
      "grad_norm": 0.15659593045711517,
      "learning_rate": 6.693654736712977e-05,
      "loss": 0.759,
      "mean_token_accuracy": 0.779259979724884,
      "num_tokens": 344118586.0,
      "step": 10782
    },
    {
      "entropy": 0.6041261241771281,
      "epoch": 0.9921047032507001,
      "grad_norm": 0.1446591466665268,
      "learning_rate": 6.693348054098814e-05,
      "loss": 0.5998,
      "mean_token_accuracy": 0.8241670578718185,
      "num_tokens": 344151048.0,
      "step": 10783
    },
    {
      "entropy": 0.7308912593871355,
      "epoch": 0.9921967096221413,
      "grad_norm": 0.15078559517860413,
      "learning_rate": 6.69304137148465e-05,
      "loss": 0.7187,
      "mean_token_accuracy": 0.7892863526940346,
      "num_tokens": 344183467.0,
      "step": 10784
    },
    {
      "entropy": 0.7304187547415495,
      "epoch": 0.9922887159935826,
      "grad_norm": 0.1534116566181183,
      "learning_rate": 6.692734688870487e-05,
      "loss": 0.7044,
      "mean_token_accuracy": 0.7886729873716831,
      "num_tokens": 344215047.0,
      "step": 10785
    },
    {
      "entropy": 0.7387849614024162,
      "epoch": 0.9923807223650237,
      "grad_norm": 0.15662816166877747,
      "learning_rate": 6.692428006256325e-05,
      "loss": 0.7188,
      "mean_token_accuracy": 0.7899422906339169,
      "num_tokens": 344247176.0,
      "step": 10786
    },
    {
      "entropy": 0.7515231482684612,
      "epoch": 0.992472728736465,
      "grad_norm": 0.15318962931632996,
      "learning_rate": 6.692121323642164e-05,
      "loss": 0.7333,
      "mean_token_accuracy": 0.783893309533596,
      "num_tokens": 344279531.0,
      "step": 10787
    },
    {
      "entropy": 0.6724366955459118,
      "epoch": 0.9925647351079062,
      "grad_norm": 0.15184269845485687,
      "learning_rate": 6.691814641028e-05,
      "loss": 0.6546,
      "mean_token_accuracy": 0.8029724359512329,
      "num_tokens": 344311630.0,
      "step": 10788
    },
    {
      "entropy": 0.5736624905839562,
      "epoch": 0.9926567414793475,
      "grad_norm": 0.14328216016292572,
      "learning_rate": 6.691507958413837e-05,
      "loss": 0.5554,
      "mean_token_accuracy": 0.8324397876858711,
      "num_tokens": 344343846.0,
      "step": 10789
    },
    {
      "entropy": 0.5934606399387121,
      "epoch": 0.9927487478507887,
      "grad_norm": 0.14868777990341187,
      "learning_rate": 6.691201275799675e-05,
      "loss": 0.5955,
      "mean_token_accuracy": 0.819635983556509,
      "num_tokens": 344375666.0,
      "step": 10790
    },
    {
      "entropy": 0.7812086325138807,
      "epoch": 0.9928407542222298,
      "grad_norm": 0.16050119698047638,
      "learning_rate": 6.690894593185513e-05,
      "loss": 0.7734,
      "mean_token_accuracy": 0.7699658460915089,
      "num_tokens": 344408012.0,
      "step": 10791
    },
    {
      "entropy": 0.7579163946211338,
      "epoch": 0.9929327605936711,
      "grad_norm": 0.16025415062904358,
      "learning_rate": 6.69058791057135e-05,
      "loss": 0.7752,
      "mean_token_accuracy": 0.7685063816606998,
      "num_tokens": 344438517.0,
      "step": 10792
    },
    {
      "entropy": 0.7749667949974537,
      "epoch": 0.9930247669651123,
      "grad_norm": 0.16385215520858765,
      "learning_rate": 6.690281227957187e-05,
      "loss": 0.7791,
      "mean_token_accuracy": 0.7710829079151154,
      "num_tokens": 344470875.0,
      "step": 10793
    },
    {
      "entropy": 0.7205776497721672,
      "epoch": 0.9931167733365536,
      "grad_norm": 0.1588430106639862,
      "learning_rate": 6.689974545343025e-05,
      "loss": 0.7155,
      "mean_token_accuracy": 0.785976879298687,
      "num_tokens": 344502448.0,
      "step": 10794
    },
    {
      "entropy": 0.7119633732363582,
      "epoch": 0.9932087797079948,
      "grad_norm": 0.16248252987861633,
      "learning_rate": 6.689667862728862e-05,
      "loss": 0.7128,
      "mean_token_accuracy": 0.7936295755207539,
      "num_tokens": 344533219.0,
      "step": 10795
    },
    {
      "entropy": 0.6819195002317429,
      "epoch": 0.993300786079436,
      "grad_norm": 0.14672689139842987,
      "learning_rate": 6.6893611801147e-05,
      "loss": 0.6651,
      "mean_token_accuracy": 0.8005755506455898,
      "num_tokens": 344565447.0,
      "step": 10796
    },
    {
      "entropy": 0.7185427099466324,
      "epoch": 0.9933927924508772,
      "grad_norm": 0.16160903871059418,
      "learning_rate": 6.689054497500537e-05,
      "loss": 0.7186,
      "mean_token_accuracy": 0.7911116778850555,
      "num_tokens": 344596008.0,
      "step": 10797
    },
    {
      "entropy": 0.8016707301139832,
      "epoch": 0.9934847988223184,
      "grad_norm": 0.16602005064487457,
      "learning_rate": 6.688747814886375e-05,
      "loss": 0.7904,
      "mean_token_accuracy": 0.7696431279182434,
      "num_tokens": 344628248.0,
      "step": 10798
    },
    {
      "entropy": 0.6824175361543894,
      "epoch": 0.9935768051937597,
      "grad_norm": 0.1552087664604187,
      "learning_rate": 6.688441132272212e-05,
      "loss": 0.6732,
      "mean_token_accuracy": 0.8006101995706558,
      "num_tokens": 344660742.0,
      "step": 10799
    },
    {
      "entropy": 0.670639562420547,
      "epoch": 0.9936688115652009,
      "grad_norm": 0.1425110399723053,
      "learning_rate": 6.688134449658049e-05,
      "loss": 0.6636,
      "mean_token_accuracy": 0.8016855455935001,
      "num_tokens": 344693179.0,
      "step": 10800
    },
    {
      "entropy": 0.6599009223282337,
      "epoch": 0.9937608179366421,
      "grad_norm": 0.1553841382265091,
      "learning_rate": 6.687827767043887e-05,
      "loss": 0.6551,
      "mean_token_accuracy": 0.8074255250394344,
      "num_tokens": 344725469.0,
      "step": 10801
    },
    {
      "entropy": 0.7757883109152317,
      "epoch": 0.9938528243080833,
      "grad_norm": 0.15855452418327332,
      "learning_rate": 6.687521084429725e-05,
      "loss": 0.7829,
      "mean_token_accuracy": 0.7689474076032639,
      "num_tokens": 344757175.0,
      "step": 10802
    },
    {
      "entropy": 0.5795407295227051,
      "epoch": 0.9939448306795245,
      "grad_norm": 0.15758000314235687,
      "learning_rate": 6.687214401815562e-05,
      "loss": 0.5767,
      "mean_token_accuracy": 0.8233759813010693,
      "num_tokens": 344789081.0,
      "step": 10803
    },
    {
      "entropy": 0.7402180396020412,
      "epoch": 0.9940368370509658,
      "grad_norm": 0.16361987590789795,
      "learning_rate": 6.686907719201398e-05,
      "loss": 0.7282,
      "mean_token_accuracy": 0.7837566845119,
      "num_tokens": 344820962.0,
      "step": 10804
    },
    {
      "entropy": 0.7793048862367868,
      "epoch": 0.994128843422407,
      "grad_norm": 0.15453974902629852,
      "learning_rate": 6.686601036587237e-05,
      "loss": 0.7523,
      "mean_token_accuracy": 0.77679792791605,
      "num_tokens": 344852834.0,
      "step": 10805
    },
    {
      "entropy": 0.7493541445583105,
      "epoch": 0.9942208497938482,
      "grad_norm": 0.160062775015831,
      "learning_rate": 6.686294353973073e-05,
      "loss": 0.7524,
      "mean_token_accuracy": 0.77812734618783,
      "num_tokens": 344884693.0,
      "step": 10806
    },
    {
      "entropy": 0.6082068625837564,
      "epoch": 0.9943128561652894,
      "grad_norm": 0.13591396808624268,
      "learning_rate": 6.685987671358911e-05,
      "loss": 0.5738,
      "mean_token_accuracy": 0.8299529291689396,
      "num_tokens": 344916889.0,
      "step": 10807
    },
    {
      "entropy": 0.5910112503916025,
      "epoch": 0.9944048625367307,
      "grad_norm": 0.1520845592021942,
      "learning_rate": 6.685680988744748e-05,
      "loss": 0.5641,
      "mean_token_accuracy": 0.8321896269917488,
      "num_tokens": 344948757.0,
      "step": 10808
    },
    {
      "entropy": 0.6240068860352039,
      "epoch": 0.9944968689081719,
      "grad_norm": 0.1496381014585495,
      "learning_rate": 6.685374306130586e-05,
      "loss": 0.6019,
      "mean_token_accuracy": 0.8181179650127888,
      "num_tokens": 344981525.0,
      "step": 10809
    },
    {
      "entropy": 0.6905934661626816,
      "epoch": 0.9945888752796131,
      "grad_norm": 0.14948153495788574,
      "learning_rate": 6.685067623516423e-05,
      "loss": 0.6743,
      "mean_token_accuracy": 0.7986595891416073,
      "num_tokens": 345013412.0,
      "step": 10810
    },
    {
      "entropy": 0.8332785591483116,
      "epoch": 0.9946808816510543,
      "grad_norm": 0.16550293564796448,
      "learning_rate": 6.68476094090226e-05,
      "loss": 0.8409,
      "mean_token_accuracy": 0.756006445735693,
      "num_tokens": 345045533.0,
      "step": 10811
    },
    {
      "entropy": 0.6532058566808701,
      "epoch": 0.9947728880224955,
      "grad_norm": 0.15743549168109894,
      "learning_rate": 6.684454258288098e-05,
      "loss": 0.6317,
      "mean_token_accuracy": 0.8141423799097538,
      "num_tokens": 345077519.0,
      "step": 10812
    },
    {
      "entropy": 0.7174852453172207,
      "epoch": 0.9948648943939368,
      "grad_norm": 0.15358835458755493,
      "learning_rate": 6.684147575673936e-05,
      "loss": 0.7026,
      "mean_token_accuracy": 0.791305385529995,
      "num_tokens": 345109832.0,
      "step": 10813
    },
    {
      "entropy": 0.6976348096504807,
      "epoch": 0.994956900765378,
      "grad_norm": 0.15170064568519592,
      "learning_rate": 6.683840893059773e-05,
      "loss": 0.6899,
      "mean_token_accuracy": 0.7940831072628498,
      "num_tokens": 345142155.0,
      "step": 10814
    },
    {
      "entropy": 0.6204143255017698,
      "epoch": 0.9950489071368193,
      "grad_norm": 0.1409311443567276,
      "learning_rate": 6.68353421044561e-05,
      "loss": 0.6152,
      "mean_token_accuracy": 0.8114605993032455,
      "num_tokens": 345174078.0,
      "step": 10815
    },
    {
      "entropy": 0.6792154386639595,
      "epoch": 0.9951409135082604,
      "grad_norm": 0.14964395761489868,
      "learning_rate": 6.683227527831447e-05,
      "loss": 0.6795,
      "mean_token_accuracy": 0.7974983379244804,
      "num_tokens": 345206264.0,
      "step": 10816
    },
    {
      "entropy": 0.6952597107738256,
      "epoch": 0.9952329198797016,
      "grad_norm": 0.14977137744426727,
      "learning_rate": 6.682920845217285e-05,
      "loss": 0.6768,
      "mean_token_accuracy": 0.7977260127663612,
      "num_tokens": 345238423.0,
      "step": 10817
    },
    {
      "entropy": 0.6377477552741766,
      "epoch": 0.9953249262511429,
      "grad_norm": 0.16152425110340118,
      "learning_rate": 6.682614162603123e-05,
      "loss": 0.6385,
      "mean_token_accuracy": 0.8105308338999748,
      "num_tokens": 345270641.0,
      "step": 10818
    },
    {
      "entropy": 0.750324472784996,
      "epoch": 0.9954169326225841,
      "grad_norm": 0.15432511270046234,
      "learning_rate": 6.68230747998896e-05,
      "loss": 0.7503,
      "mean_token_accuracy": 0.7783770971000195,
      "num_tokens": 345302362.0,
      "step": 10819
    },
    {
      "entropy": 0.6472237631678581,
      "epoch": 0.9955089389940254,
      "grad_norm": 0.1531330943107605,
      "learning_rate": 6.682000797374796e-05,
      "loss": 0.6438,
      "mean_token_accuracy": 0.8036659955978394,
      "num_tokens": 345334238.0,
      "step": 10820
    },
    {
      "entropy": 0.7867288943380117,
      "epoch": 0.9956009453654665,
      "grad_norm": 0.17016279697418213,
      "learning_rate": 6.681694114760635e-05,
      "loss": 0.7866,
      "mean_token_accuracy": 0.7710942961275578,
      "num_tokens": 345366258.0,
      "step": 10821
    },
    {
      "entropy": 0.7326446920633316,
      "epoch": 0.9956929517369077,
      "grad_norm": 0.15269401669502258,
      "learning_rate": 6.681387432146471e-05,
      "loss": 0.7269,
      "mean_token_accuracy": 0.7834556810557842,
      "num_tokens": 345398404.0,
      "step": 10822
    },
    {
      "entropy": 0.7246920168399811,
      "epoch": 0.995784958108349,
      "grad_norm": 0.1623348444700241,
      "learning_rate": 6.68108074953231e-05,
      "loss": 0.6891,
      "mean_token_accuracy": 0.7902564890682697,
      "num_tokens": 345428884.0,
      "step": 10823
    },
    {
      "entropy": 0.7372211925685406,
      "epoch": 0.9958769644797902,
      "grad_norm": 0.16010963916778564,
      "learning_rate": 6.680774066918146e-05,
      "loss": 0.7156,
      "mean_token_accuracy": 0.7888367958366871,
      "num_tokens": 345460916.0,
      "step": 10824
    },
    {
      "entropy": 0.6847714437171817,
      "epoch": 0.9959689708512315,
      "grad_norm": 0.14638179540634155,
      "learning_rate": 6.680467384303984e-05,
      "loss": 0.6611,
      "mean_token_accuracy": 0.8020917288959026,
      "num_tokens": 345493175.0,
      "step": 10825
    },
    {
      "entropy": 0.7340292874723673,
      "epoch": 0.9960609772226726,
      "grad_norm": 0.1611267328262329,
      "learning_rate": 6.680160701689821e-05,
      "loss": 0.7166,
      "mean_token_accuracy": 0.7841636501252651,
      "num_tokens": 345525508.0,
      "step": 10826
    },
    {
      "entropy": 0.7063327599316835,
      "epoch": 0.9961529835941139,
      "grad_norm": 0.1553497016429901,
      "learning_rate": 6.679854019075658e-05,
      "loss": 0.6722,
      "mean_token_accuracy": 0.7971266470849514,
      "num_tokens": 345557212.0,
      "step": 10827
    },
    {
      "entropy": 0.5671592466533184,
      "epoch": 0.9962449899655551,
      "grad_norm": 0.14719940721988678,
      "learning_rate": 6.679547336461496e-05,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8348774202167988,
      "num_tokens": 345589095.0,
      "step": 10828
    },
    {
      "entropy": 0.6739183831959963,
      "epoch": 0.9963369963369964,
      "grad_norm": 0.1585342288017273,
      "learning_rate": 6.679240653847334e-05,
      "loss": 0.6964,
      "mean_token_accuracy": 0.793883703649044,
      "num_tokens": 345621538.0,
      "step": 10829
    },
    {
      "entropy": 0.7472993321716785,
      "epoch": 0.9964290027084376,
      "grad_norm": 0.15805289149284363,
      "learning_rate": 6.678933971233171e-05,
      "loss": 0.7516,
      "mean_token_accuracy": 0.7810756228864193,
      "num_tokens": 345653324.0,
      "step": 10830
    },
    {
      "entropy": 0.7354216519743204,
      "epoch": 0.9965210090798787,
      "grad_norm": 0.15132851898670197,
      "learning_rate": 6.678627288619008e-05,
      "loss": 0.7332,
      "mean_token_accuracy": 0.7836678400635719,
      "num_tokens": 345685800.0,
      "step": 10831
    },
    {
      "entropy": 0.7084527146071196,
      "epoch": 0.99661301545132,
      "grad_norm": 0.14873826503753662,
      "learning_rate": 6.678320606004846e-05,
      "loss": 0.7066,
      "mean_token_accuracy": 0.7900439985096455,
      "num_tokens": 345717989.0,
      "step": 10832
    },
    {
      "entropy": 0.6555616389960051,
      "epoch": 0.9967050218227612,
      "grad_norm": 0.150567427277565,
      "learning_rate": 6.678013923390684e-05,
      "loss": 0.6597,
      "mean_token_accuracy": 0.8023892752826214,
      "num_tokens": 345750675.0,
      "step": 10833
    },
    {
      "entropy": 0.7393480315804482,
      "epoch": 0.9967970281942025,
      "grad_norm": 0.15898364782333374,
      "learning_rate": 6.677707240776521e-05,
      "loss": 0.7391,
      "mean_token_accuracy": 0.7841736897826195,
      "num_tokens": 345782037.0,
      "step": 10834
    },
    {
      "entropy": 0.6871941313147545,
      "epoch": 0.9968890345656437,
      "grad_norm": 0.15557844936847687,
      "learning_rate": 6.677400558162358e-05,
      "loss": 0.667,
      "mean_token_accuracy": 0.8017421141266823,
      "num_tokens": 345814048.0,
      "step": 10835
    },
    {
      "entropy": 0.5712172649800777,
      "epoch": 0.9969810409370848,
      "grad_norm": 0.15374541282653809,
      "learning_rate": 6.677093875548196e-05,
      "loss": 0.5558,
      "mean_token_accuracy": 0.8307640366256237,
      "num_tokens": 345846267.0,
      "step": 10836
    },
    {
      "entropy": 0.7518662754446268,
      "epoch": 0.9970730473085261,
      "grad_norm": 0.15640655159950256,
      "learning_rate": 6.676787192934033e-05,
      "loss": 0.7421,
      "mean_token_accuracy": 0.7804466374218464,
      "num_tokens": 345877996.0,
      "step": 10837
    },
    {
      "entropy": 0.5528900772333145,
      "epoch": 0.9971650536799673,
      "grad_norm": 0.14041969180107117,
      "learning_rate": 6.676480510319871e-05,
      "loss": 0.5214,
      "mean_token_accuracy": 0.839076291769743,
      "num_tokens": 345909585.0,
      "step": 10838
    },
    {
      "entropy": 0.7652533873915672,
      "epoch": 0.9972570600514086,
      "grad_norm": 0.1574499011039734,
      "learning_rate": 6.676173827705708e-05,
      "loss": 0.759,
      "mean_token_accuracy": 0.7736484445631504,
      "num_tokens": 345940890.0,
      "step": 10839
    },
    {
      "entropy": 0.7549010366201401,
      "epoch": 0.9973490664228498,
      "grad_norm": 0.15799598395824432,
      "learning_rate": 6.675867145091546e-05,
      "loss": 0.7562,
      "mean_token_accuracy": 0.7742687463760376,
      "num_tokens": 345972440.0,
      "step": 10840
    },
    {
      "entropy": 0.6636172076687217,
      "epoch": 0.997441072794291,
      "grad_norm": 0.1587314009666443,
      "learning_rate": 6.675560462477383e-05,
      "loss": 0.6378,
      "mean_token_accuracy": 0.8088156096637249,
      "num_tokens": 346004548.0,
      "step": 10841
    },
    {
      "entropy": 0.7224581986665726,
      "epoch": 0.9975330791657322,
      "grad_norm": 0.16612033545970917,
      "learning_rate": 6.67525377986322e-05,
      "loss": 0.7329,
      "mean_token_accuracy": 0.7866311147809029,
      "num_tokens": 346036676.0,
      "step": 10842
    },
    {
      "entropy": 0.8078804817050695,
      "epoch": 0.9976250855371734,
      "grad_norm": 0.1617724746465683,
      "learning_rate": 6.674947097249057e-05,
      "loss": 0.821,
      "mean_token_accuracy": 0.7615036852657795,
      "num_tokens": 346067749.0,
      "step": 10843
    },
    {
      "entropy": 0.7221355978399515,
      "epoch": 0.9977170919086147,
      "grad_norm": 0.1590796262025833,
      "learning_rate": 6.674640414634896e-05,
      "loss": 0.7142,
      "mean_token_accuracy": 0.7887849994003773,
      "num_tokens": 346099804.0,
      "step": 10844
    },
    {
      "entropy": 0.7533706314861774,
      "epoch": 0.9978090982800559,
      "grad_norm": 0.14854960143566132,
      "learning_rate": 6.674333732020732e-05,
      "loss": 0.7346,
      "mean_token_accuracy": 0.7839143127202988,
      "num_tokens": 346132217.0,
      "step": 10845
    },
    {
      "entropy": 0.6553160981275141,
      "epoch": 0.9979011046514971,
      "grad_norm": 0.1495293825864792,
      "learning_rate": 6.674027049406569e-05,
      "loss": 0.656,
      "mean_token_accuracy": 0.8066457323729992,
      "num_tokens": 346163954.0,
      "step": 10846
    },
    {
      "entropy": 0.69985631108284,
      "epoch": 0.9979931110229383,
      "grad_norm": 0.16674433648586273,
      "learning_rate": 6.673720366792406e-05,
      "loss": 0.6905,
      "mean_token_accuracy": 0.799196682870388,
      "num_tokens": 346196049.0,
      "step": 10847
    },
    {
      "entropy": 0.7850935608148575,
      "epoch": 0.9980851173943796,
      "grad_norm": 0.1593358963727951,
      "learning_rate": 6.673413684178244e-05,
      "loss": 0.7757,
      "mean_token_accuracy": 0.773831445723772,
      "num_tokens": 346228017.0,
      "step": 10848
    },
    {
      "entropy": 0.6615530867129564,
      "epoch": 0.9981771237658208,
      "grad_norm": 0.14902935922145844,
      "learning_rate": 6.673107001564082e-05,
      "loss": 0.6512,
      "mean_token_accuracy": 0.8073189482092857,
      "num_tokens": 346259724.0,
      "step": 10849
    },
    {
      "entropy": 0.5788842597976327,
      "epoch": 0.998269130137262,
      "grad_norm": 0.146217942237854,
      "learning_rate": 6.672800318949919e-05,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8317787833511829,
      "num_tokens": 346292081.0,
      "step": 10850
    },
    {
      "entropy": 0.7186735086143017,
      "epoch": 0.9983611365087032,
      "grad_norm": 0.158821240067482,
      "learning_rate": 6.672493636335756e-05,
      "loss": 0.7038,
      "mean_token_accuracy": 0.7934148982167244,
      "num_tokens": 346323841.0,
      "step": 10851
    },
    {
      "entropy": 0.6809679251164198,
      "epoch": 0.9984531428801444,
      "grad_norm": 0.1526089757680893,
      "learning_rate": 6.672186953721594e-05,
      "loss": 0.6837,
      "mean_token_accuracy": 0.7969912327826023,
      "num_tokens": 346355518.0,
      "step": 10852
    },
    {
      "entropy": 0.6653263624757528,
      "epoch": 0.9985451492515857,
      "grad_norm": 0.14160585403442383,
      "learning_rate": 6.671880271107431e-05,
      "loss": 0.6549,
      "mean_token_accuracy": 0.8040918707847595,
      "num_tokens": 346387794.0,
      "step": 10853
    },
    {
      "entropy": 0.7460291385650635,
      "epoch": 0.9986371556230269,
      "grad_norm": 0.15650811791419983,
      "learning_rate": 6.671573588493269e-05,
      "loss": 0.7252,
      "mean_token_accuracy": 0.7846652790904045,
      "num_tokens": 346419886.0,
      "step": 10854
    },
    {
      "entropy": 0.7169933561235666,
      "epoch": 0.9987291619944682,
      "grad_norm": 0.160558819770813,
      "learning_rate": 6.671266905879106e-05,
      "loss": 0.6987,
      "mean_token_accuracy": 0.7922168411314487,
      "num_tokens": 346451385.0,
      "step": 10855
    },
    {
      "entropy": 0.7304557710886002,
      "epoch": 0.9988211683659093,
      "grad_norm": 0.16046153008937836,
      "learning_rate": 6.670960223264944e-05,
      "loss": 0.7253,
      "mean_token_accuracy": 0.7871816568076611,
      "num_tokens": 346481725.0,
      "step": 10856
    },
    {
      "entropy": 0.742641270160675,
      "epoch": 0.9989131747373505,
      "grad_norm": 0.15400442481040955,
      "learning_rate": 6.67065354065078e-05,
      "loss": 0.729,
      "mean_token_accuracy": 0.780948419123888,
      "num_tokens": 346513390.0,
      "step": 10857
    },
    {
      "entropy": 0.7093084985390306,
      "epoch": 0.9990051811087918,
      "grad_norm": 0.15645098686218262,
      "learning_rate": 6.670346858036617e-05,
      "loss": 0.6976,
      "mean_token_accuracy": 0.7937665283679962,
      "num_tokens": 346545048.0,
      "step": 10858
    },
    {
      "entropy": 0.7258321680128574,
      "epoch": 0.999097187480233,
      "grad_norm": 0.1696975976228714,
      "learning_rate": 6.670040175422456e-05,
      "loss": 0.7287,
      "mean_token_accuracy": 0.7844110876321793,
      "num_tokens": 346576290.0,
      "step": 10859
    },
    {
      "entropy": 0.7598477322608232,
      "epoch": 0.9991891938516743,
      "grad_norm": 0.17301052808761597,
      "learning_rate": 6.669733492808294e-05,
      "loss": 0.7596,
      "mean_token_accuracy": 0.7780938781797886,
      "num_tokens": 346608522.0,
      "step": 10860
    },
    {
      "entropy": 0.6982610691338778,
      "epoch": 0.9992812002231154,
      "grad_norm": 0.15256164968013763,
      "learning_rate": 6.66942681019413e-05,
      "loss": 0.6761,
      "mean_token_accuracy": 0.7958274148404598,
      "num_tokens": 346640195.0,
      "step": 10861
    },
    {
      "entropy": 0.6985594127327204,
      "epoch": 0.9993732065945566,
      "grad_norm": 0.15505656599998474,
      "learning_rate": 6.669120127579967e-05,
      "loss": 0.6809,
      "mean_token_accuracy": 0.7996110133826733,
      "num_tokens": 346672638.0,
      "step": 10862
    },
    {
      "entropy": 0.7910066060721874,
      "epoch": 0.9994652129659979,
      "grad_norm": 0.1599981039762497,
      "learning_rate": 6.668813444965805e-05,
      "loss": 0.7626,
      "mean_token_accuracy": 0.7733800001442432,
      "num_tokens": 346704739.0,
      "step": 10863
    },
    {
      "entropy": 0.700426023453474,
      "epoch": 0.9995572193374391,
      "grad_norm": 0.1500086933374405,
      "learning_rate": 6.668506762351642e-05,
      "loss": 0.6784,
      "mean_token_accuracy": 0.7938801124691963,
      "num_tokens": 346736204.0,
      "step": 10864
    },
    {
      "entropy": 0.7300129123032093,
      "epoch": 0.9996492257088804,
      "grad_norm": 0.15329793095588684,
      "learning_rate": 6.66820007973748e-05,
      "loss": 0.7098,
      "mean_token_accuracy": 0.7903252132236958,
      "num_tokens": 346767693.0,
      "step": 10865
    },
    {
      "entropy": 0.777059517800808,
      "epoch": 0.9997412320803215,
      "grad_norm": 0.15498676896095276,
      "learning_rate": 6.667893397123317e-05,
      "loss": 0.7614,
      "mean_token_accuracy": 0.7808580324053764,
      "num_tokens": 346799524.0,
      "step": 10866
    },
    {
      "entropy": 0.7458492089062929,
      "epoch": 0.9998332384517628,
      "grad_norm": 0.1522398740053177,
      "learning_rate": 6.667586714509155e-05,
      "loss": 0.7442,
      "mean_token_accuracy": 0.779858510941267,
      "num_tokens": 346831308.0,
      "step": 10867
    },
    {
      "entropy": 0.7047930676490068,
      "epoch": 0.999925244823204,
      "grad_norm": 0.15304671227931976,
      "learning_rate": 6.667280031894992e-05,
      "loss": 0.6983,
      "mean_token_accuracy": 0.7935652211308479,
      "num_tokens": 346863175.0,
      "step": 10868
    },
    {
      "entropy": 0.698239232485111,
      "epoch": 1.0,
      "grad_norm": 0.1780262291431427,
      "learning_rate": 6.666973349280829e-05,
      "loss": 0.6984,
      "mean_token_accuracy": 0.7843229082914499,
      "num_tokens": 346887762.0,
      "step": 10869
    },
    {
      "entropy": 0.7446477059274912,
      "epoch": 1.0000920063714411,
      "grad_norm": 0.15724828839302063,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.659,
      "mean_token_accuracy": 0.7963673584163189,
      "num_tokens": 346919023.0,
      "step": 10870
    },
    {
      "entropy": 0.7004071082919836,
      "epoch": 1.0001840127428825,
      "grad_norm": 0.14970508217811584,
      "learning_rate": 6.666359984052505e-05,
      "loss": 0.626,
      "mean_token_accuracy": 0.8065500594675541,
      "num_tokens": 346950643.0,
      "step": 10871
    },
    {
      "entropy": 0.6081130795646459,
      "epoch": 1.0002760191143236,
      "grad_norm": 0.1362941712141037,
      "learning_rate": 6.666053301438342e-05,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8381778001785278,
      "num_tokens": 346982590.0,
      "step": 10872
    },
    {
      "entropy": 0.6558470688760281,
      "epoch": 1.000368025485765,
      "grad_norm": 0.14820900559425354,
      "learning_rate": 6.665746618824179e-05,
      "loss": 0.6029,
      "mean_token_accuracy": 0.8164119645953178,
      "num_tokens": 347014855.0,
      "step": 10873
    },
    {
      "entropy": 0.6195752928033471,
      "epoch": 1.0004600318572061,
      "grad_norm": 0.1447901874780655,
      "learning_rate": 6.665439936210015e-05,
      "loss": 0.5941,
      "mean_token_accuracy": 0.8206167258322239,
      "num_tokens": 347046900.0,
      "step": 10874
    },
    {
      "entropy": 0.6441139224916697,
      "epoch": 1.0005520382286472,
      "grad_norm": 0.15413637459278107,
      "learning_rate": 6.665133253595855e-05,
      "loss": 0.6265,
      "mean_token_accuracy": 0.8087260015308857,
      "num_tokens": 347078559.0,
      "step": 10875
    },
    {
      "entropy": 0.6043624803423882,
      "epoch": 1.0006440446000886,
      "grad_norm": 0.1498301923274994,
      "learning_rate": 6.664826570981692e-05,
      "loss": 0.5655,
      "mean_token_accuracy": 0.8259975053369999,
      "num_tokens": 347110399.0,
      "step": 10876
    },
    {
      "entropy": 0.6130339885130525,
      "epoch": 1.0007360509715297,
      "grad_norm": 0.15572820603847504,
      "learning_rate": 6.664519888367529e-05,
      "loss": 0.5898,
      "mean_token_accuracy": 0.8225942514836788,
      "num_tokens": 347141606.0,
      "step": 10877
    },
    {
      "entropy": 0.5133166969753802,
      "epoch": 1.000828057342971,
      "grad_norm": 0.15125444531440735,
      "learning_rate": 6.664213205753365e-05,
      "loss": 0.5191,
      "mean_token_accuracy": 0.8405295945703983,
      "num_tokens": 347174125.0,
      "step": 10878
    },
    {
      "entropy": 0.594632226973772,
      "epoch": 1.0009200637144122,
      "grad_norm": 0.1516452431678772,
      "learning_rate": 6.663906523139203e-05,
      "loss": 0.5749,
      "mean_token_accuracy": 0.8192759715020657,
      "num_tokens": 347206200.0,
      "step": 10879
    },
    {
      "entropy": 0.4803753327578306,
      "epoch": 1.0010120700858534,
      "grad_norm": 0.1329870969057083,
      "learning_rate": 6.663599840525042e-05,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8517781309783459,
      "num_tokens": 347238457.0,
      "step": 10880
    },
    {
      "entropy": 0.5796688850969076,
      "epoch": 1.0011040764572947,
      "grad_norm": 0.15478304028511047,
      "learning_rate": 6.663293157910878e-05,
      "loss": 0.5547,
      "mean_token_accuracy": 0.8299643285572529,
      "num_tokens": 347270534.0,
      "step": 10881
    },
    {
      "entropy": 0.4750945996493101,
      "epoch": 1.0011960828287358,
      "grad_norm": 0.15997923910617828,
      "learning_rate": 6.662986475296715e-05,
      "loss": 0.4664,
      "mean_token_accuracy": 0.8528170362114906,
      "num_tokens": 347302506.0,
      "step": 10882
    },
    {
      "entropy": 0.6370038818567991,
      "epoch": 1.0012880892001772,
      "grad_norm": 0.16028724610805511,
      "learning_rate": 6.662679792682553e-05,
      "loss": 0.6246,
      "mean_token_accuracy": 0.8069759272038937,
      "num_tokens": 347333637.0,
      "step": 10883
    },
    {
      "entropy": 0.6827068906277418,
      "epoch": 1.0013800955716183,
      "grad_norm": 0.16374889016151428,
      "learning_rate": 6.66237311006839e-05,
      "loss": 0.6759,
      "mean_token_accuracy": 0.7931073792278767,
      "num_tokens": 347366277.0,
      "step": 10884
    },
    {
      "entropy": 0.6622396018356085,
      "epoch": 1.0014721019430595,
      "grad_norm": 0.16930747032165527,
      "learning_rate": 6.662066427454228e-05,
      "loss": 0.6423,
      "mean_token_accuracy": 0.8058211281895638,
      "num_tokens": 347398373.0,
      "step": 10885
    },
    {
      "entropy": 0.6117447586730123,
      "epoch": 1.0015641083145008,
      "grad_norm": 0.1564471274614334,
      "learning_rate": 6.661759744840065e-05,
      "loss": 0.6182,
      "mean_token_accuracy": 0.8085714802145958,
      "num_tokens": 347430808.0,
      "step": 10886
    },
    {
      "entropy": 0.6259973328560591,
      "epoch": 1.001656114685942,
      "grad_norm": 0.15397760272026062,
      "learning_rate": 6.661453062225903e-05,
      "loss": 0.594,
      "mean_token_accuracy": 0.8136333264410496,
      "num_tokens": 347462019.0,
      "step": 10887
    },
    {
      "entropy": 0.6274276450276375,
      "epoch": 1.0017481210573833,
      "grad_norm": 0.15608319640159607,
      "learning_rate": 6.66114637961174e-05,
      "loss": 0.6136,
      "mean_token_accuracy": 0.8106551915407181,
      "num_tokens": 347494206.0,
      "step": 10888
    },
    {
      "entropy": 0.6654837047681212,
      "epoch": 1.0018401274288244,
      "grad_norm": 0.1577060967683792,
      "learning_rate": 6.660839696997577e-05,
      "loss": 0.6583,
      "mean_token_accuracy": 0.7992147840559483,
      "num_tokens": 347526569.0,
      "step": 10889
    },
    {
      "entropy": 0.5051075536757708,
      "epoch": 1.0019321338002656,
      "grad_norm": 0.14683876931667328,
      "learning_rate": 6.660533014383415e-05,
      "loss": 0.4758,
      "mean_token_accuracy": 0.850743792951107,
      "num_tokens": 347558911.0,
      "step": 10890
    },
    {
      "entropy": 0.5545512214303017,
      "epoch": 1.002024140171707,
      "grad_norm": 0.1533421277999878,
      "learning_rate": 6.660226331769253e-05,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8411615788936615,
      "num_tokens": 347591465.0,
      "step": 10891
    },
    {
      "entropy": 0.6190305072814226,
      "epoch": 1.002116146543148,
      "grad_norm": 0.15913037955760956,
      "learning_rate": 6.65991964915509e-05,
      "loss": 0.605,
      "mean_token_accuracy": 0.8129717595875263,
      "num_tokens": 347623322.0,
      "step": 10892
    },
    {
      "entropy": 0.7353445859625936,
      "epoch": 1.0022081529145894,
      "grad_norm": 0.16494370996952057,
      "learning_rate": 6.659612966540927e-05,
      "loss": 0.7207,
      "mean_token_accuracy": 0.7826841808855534,
      "num_tokens": 347655137.0,
      "step": 10893
    },
    {
      "entropy": 0.5702814906835556,
      "epoch": 1.0023001592860306,
      "grad_norm": 0.1446094661951065,
      "learning_rate": 6.659306283926765e-05,
      "loss": 0.5394,
      "mean_token_accuracy": 0.828872561454773,
      "num_tokens": 347687077.0,
      "step": 10894
    },
    {
      "entropy": 0.6183013552799821,
      "epoch": 1.0023921656574717,
      "grad_norm": 0.15070860087871552,
      "learning_rate": 6.658999601312602e-05,
      "loss": 0.599,
      "mean_token_accuracy": 0.8137944340705872,
      "num_tokens": 347719439.0,
      "step": 10895
    },
    {
      "entropy": 0.5126731246709824,
      "epoch": 1.002484172028913,
      "grad_norm": 0.15130530297756195,
      "learning_rate": 6.65869291869844e-05,
      "loss": 0.4797,
      "mean_token_accuracy": 0.853540800511837,
      "num_tokens": 347751554.0,
      "step": 10896
    },
    {
      "entropy": 0.5193057125434279,
      "epoch": 1.0025761784003542,
      "grad_norm": 0.1568012237548828,
      "learning_rate": 6.658386236084277e-05,
      "loss": 0.4941,
      "mean_token_accuracy": 0.843054324388504,
      "num_tokens": 347783614.0,
      "step": 10897
    },
    {
      "entropy": 0.6923736315220594,
      "epoch": 1.0026681847717955,
      "grad_norm": 0.15880896151065826,
      "learning_rate": 6.658079553470115e-05,
      "loss": 0.6719,
      "mean_token_accuracy": 0.798379585146904,
      "num_tokens": 347815268.0,
      "step": 10898
    },
    {
      "entropy": 0.5896908743306994,
      "epoch": 1.0027601911432367,
      "grad_norm": 0.15078572928905487,
      "learning_rate": 6.657772870855951e-05,
      "loss": 0.5704,
      "mean_token_accuracy": 0.8250398524105549,
      "num_tokens": 347847687.0,
      "step": 10899
    },
    {
      "entropy": 0.6614209776744246,
      "epoch": 1.0028521975146778,
      "grad_norm": 0.15421143174171448,
      "learning_rate": 6.657466188241788e-05,
      "loss": 0.6306,
      "mean_token_accuracy": 0.8069374673068523,
      "num_tokens": 347878858.0,
      "step": 10900
    },
    {
      "entropy": 0.5209066150709987,
      "epoch": 1.0029442038861192,
      "grad_norm": 0.15083561837673187,
      "learning_rate": 6.657159505627626e-05,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8474457375705242,
      "num_tokens": 347911391.0,
      "step": 10901
    },
    {
      "entropy": 0.5923266038298607,
      "epoch": 1.0030362102575603,
      "grad_norm": 0.15395836532115936,
      "learning_rate": 6.656852823013465e-05,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8244790397584438,
      "num_tokens": 347943336.0,
      "step": 10902
    },
    {
      "entropy": 0.6281817005947232,
      "epoch": 1.0031282166290016,
      "grad_norm": 0.14728213846683502,
      "learning_rate": 6.656546140399301e-05,
      "loss": 0.6082,
      "mean_token_accuracy": 0.8163062073290348,
      "num_tokens": 347975307.0,
      "step": 10903
    },
    {
      "entropy": 0.5962233780883253,
      "epoch": 1.0032202230004428,
      "grad_norm": 0.1607588231563568,
      "learning_rate": 6.656239457785138e-05,
      "loss": 0.6033,
      "mean_token_accuracy": 0.8131915666162968,
      "num_tokens": 348006981.0,
      "step": 10904
    },
    {
      "entropy": 0.692023690789938,
      "epoch": 1.003312229371884,
      "grad_norm": 0.16088750958442688,
      "learning_rate": 6.655932775170975e-05,
      "loss": 0.692,
      "mean_token_accuracy": 0.7893695011734962,
      "num_tokens": 348039449.0,
      "step": 10905
    },
    {
      "entropy": 0.5861635897308588,
      "epoch": 1.0034042357433253,
      "grad_norm": 0.1503043919801712,
      "learning_rate": 6.655626092556814e-05,
      "loss": 0.5589,
      "mean_token_accuracy": 0.8313687667250633,
      "num_tokens": 348071351.0,
      "step": 10906
    },
    {
      "entropy": 0.5883157383650541,
      "epoch": 1.0034962421147664,
      "grad_norm": 0.15152022242546082,
      "learning_rate": 6.655319409942651e-05,
      "loss": 0.575,
      "mean_token_accuracy": 0.8221167214214802,
      "num_tokens": 348103150.0,
      "step": 10907
    },
    {
      "entropy": 0.6713458551093936,
      "epoch": 1.0035882484862078,
      "grad_norm": 0.15495944023132324,
      "learning_rate": 6.655012727328488e-05,
      "loss": 0.6616,
      "mean_token_accuracy": 0.79629060998559,
      "num_tokens": 348135452.0,
      "step": 10908
    },
    {
      "entropy": 0.6781594976782799,
      "epoch": 1.003680254857649,
      "grad_norm": 0.16349343955516815,
      "learning_rate": 6.654706044714325e-05,
      "loss": 0.6397,
      "mean_token_accuracy": 0.80039456859231,
      "num_tokens": 348166667.0,
      "step": 10909
    },
    {
      "entropy": 0.6976355519145727,
      "epoch": 1.00377226122909,
      "grad_norm": 0.15499527752399445,
      "learning_rate": 6.654399362100163e-05,
      "loss": 0.6805,
      "mean_token_accuracy": 0.7914190366864204,
      "num_tokens": 348198390.0,
      "step": 10910
    },
    {
      "entropy": 0.5866344040259719,
      "epoch": 1.0038642676005314,
      "grad_norm": 0.1617622822523117,
      "learning_rate": 6.654092679486001e-05,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8248719200491905,
      "num_tokens": 348229341.0,
      "step": 10911
    },
    {
      "entropy": 0.5959839019924402,
      "epoch": 1.0039562739719725,
      "grad_norm": 0.15586143732070923,
      "learning_rate": 6.653785996871838e-05,
      "loss": 0.5756,
      "mean_token_accuracy": 0.8203299529850483,
      "num_tokens": 348261323.0,
      "step": 10912
    },
    {
      "entropy": 0.6106792259961367,
      "epoch": 1.0040482803434139,
      "grad_norm": 0.16252771019935608,
      "learning_rate": 6.653479314257675e-05,
      "loss": 0.6014,
      "mean_token_accuracy": 0.8139155842363834,
      "num_tokens": 348292923.0,
      "step": 10913
    },
    {
      "entropy": 0.6509055066853762,
      "epoch": 1.004140286714855,
      "grad_norm": 0.1542721390724182,
      "learning_rate": 6.653172631643513e-05,
      "loss": 0.6325,
      "mean_token_accuracy": 0.8085167855024338,
      "num_tokens": 348325484.0,
      "step": 10914
    },
    {
      "entropy": 0.6438561379909515,
      "epoch": 1.0042322930862961,
      "grad_norm": 0.15605264902114868,
      "learning_rate": 6.65286594902935e-05,
      "loss": 0.6184,
      "mean_token_accuracy": 0.8126851692795753,
      "num_tokens": 348357574.0,
      "step": 10915
    },
    {
      "entropy": 0.694303534924984,
      "epoch": 1.0043242994577375,
      "grad_norm": 0.1615842580795288,
      "learning_rate": 6.652559266415188e-05,
      "loss": 0.6871,
      "mean_token_accuracy": 0.793173111975193,
      "num_tokens": 348390114.0,
      "step": 10916
    },
    {
      "entropy": 0.6850280910730362,
      "epoch": 1.0044163058291786,
      "grad_norm": 0.1665077954530716,
      "learning_rate": 6.652252583801024e-05,
      "loss": 0.6687,
      "mean_token_accuracy": 0.7972176931798458,
      "num_tokens": 348422166.0,
      "step": 10917
    },
    {
      "entropy": 0.677992528770119,
      "epoch": 1.00450831220062,
      "grad_norm": 0.17399096488952637,
      "learning_rate": 6.651945901186863e-05,
      "loss": 0.6568,
      "mean_token_accuracy": 0.803054366260767,
      "num_tokens": 348453129.0,
      "step": 10918
    },
    {
      "entropy": 0.5873249433934689,
      "epoch": 1.0046003185720611,
      "grad_norm": 0.1605038046836853,
      "learning_rate": 6.6516392185727e-05,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8361441493034363,
      "num_tokens": 348484638.0,
      "step": 10919
    },
    {
      "entropy": 0.5736432652920485,
      "epoch": 1.0046923249435022,
      "grad_norm": 0.16228340566158295,
      "learning_rate": 6.651332535958536e-05,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8298228569328785,
      "num_tokens": 348516680.0,
      "step": 10920
    },
    {
      "entropy": 0.6720968224108219,
      "epoch": 1.0047843313149436,
      "grad_norm": 0.16339623928070068,
      "learning_rate": 6.651025853344374e-05,
      "loss": 0.6649,
      "mean_token_accuracy": 0.7987001724541187,
      "num_tokens": 348548613.0,
      "step": 10921
    },
    {
      "entropy": 0.6046053692698479,
      "epoch": 1.0048763376863847,
      "grad_norm": 0.15105615556240082,
      "learning_rate": 6.650719170730212e-05,
      "loss": 0.5797,
      "mean_token_accuracy": 0.8208046741783619,
      "num_tokens": 348580248.0,
      "step": 10922
    },
    {
      "entropy": 0.6224609753116965,
      "epoch": 1.004968344057826,
      "grad_norm": 0.15655265748500824,
      "learning_rate": 6.650412488116049e-05,
      "loss": 0.5969,
      "mean_token_accuracy": 0.8175101950764656,
      "num_tokens": 348611676.0,
      "step": 10923
    },
    {
      "entropy": 0.7369170933961868,
      "epoch": 1.0050603504292672,
      "grad_norm": 0.16962487995624542,
      "learning_rate": 6.650105805501886e-05,
      "loss": 0.7153,
      "mean_token_accuracy": 0.7849998101592064,
      "num_tokens": 348643748.0,
      "step": 10924
    },
    {
      "entropy": 0.6571086160838604,
      "epoch": 1.0051523568007084,
      "grad_norm": 0.16149339079856873,
      "learning_rate": 6.649799122887724e-05,
      "loss": 0.6357,
      "mean_token_accuracy": 0.805961936712265,
      "num_tokens": 348675135.0,
      "step": 10925
    },
    {
      "entropy": 0.5634472728706896,
      "epoch": 1.0052443631721497,
      "grad_norm": 0.15198832750320435,
      "learning_rate": 6.649492440273561e-05,
      "loss": 0.5348,
      "mean_token_accuracy": 0.831954836845398,
      "num_tokens": 348706089.0,
      "step": 10926
    },
    {
      "entropy": 0.6117124482989311,
      "epoch": 1.0053363695435908,
      "grad_norm": 0.14948128163814545,
      "learning_rate": 6.649185757659399e-05,
      "loss": 0.6008,
      "mean_token_accuracy": 0.813456580042839,
      "num_tokens": 348738573.0,
      "step": 10927
    },
    {
      "entropy": 0.7032616883516312,
      "epoch": 1.0054283759150322,
      "grad_norm": 0.16513986885547638,
      "learning_rate": 6.648879075045236e-05,
      "loss": 0.695,
      "mean_token_accuracy": 0.7891263775527477,
      "num_tokens": 348769841.0,
      "step": 10928
    },
    {
      "entropy": 0.6866722702980042,
      "epoch": 1.0055203822864733,
      "grad_norm": 0.16456320881843567,
      "learning_rate": 6.648572392431074e-05,
      "loss": 0.6934,
      "mean_token_accuracy": 0.7922024503350258,
      "num_tokens": 348802480.0,
      "step": 10929
    },
    {
      "entropy": 0.6138346623629332,
      "epoch": 1.0056123886579145,
      "grad_norm": 0.1599079817533493,
      "learning_rate": 6.648265709816911e-05,
      "loss": 0.5939,
      "mean_token_accuracy": 0.8168669529259205,
      "num_tokens": 348834569.0,
      "step": 10930
    },
    {
      "entropy": 0.6708566974848509,
      "epoch": 1.0057043950293558,
      "grad_norm": 0.16316886246204376,
      "learning_rate": 6.647959027202748e-05,
      "loss": 0.6351,
      "mean_token_accuracy": 0.807753648608923,
      "num_tokens": 348866471.0,
      "step": 10931
    },
    {
      "entropy": 0.6261600940488279,
      "epoch": 1.005796401400797,
      "grad_norm": 0.16501733660697937,
      "learning_rate": 6.647652344588586e-05,
      "loss": 0.6299,
      "mean_token_accuracy": 0.8082340285181999,
      "num_tokens": 348898985.0,
      "step": 10932
    },
    {
      "entropy": 0.6217933064326644,
      "epoch": 1.0058884077722383,
      "grad_norm": 0.1578802913427353,
      "learning_rate": 6.647345661974424e-05,
      "loss": 0.6093,
      "mean_token_accuracy": 0.8130737766623497,
      "num_tokens": 348931141.0,
      "step": 10933
    },
    {
      "entropy": 0.694828761741519,
      "epoch": 1.0059804141436794,
      "grad_norm": 0.16271106898784637,
      "learning_rate": 6.64703897936026e-05,
      "loss": 0.6619,
      "mean_token_accuracy": 0.7969559989869595,
      "num_tokens": 348962766.0,
      "step": 10934
    },
    {
      "entropy": 0.5367849660106003,
      "epoch": 1.0060724205151206,
      "grad_norm": 0.15316835045814514,
      "learning_rate": 6.646732296746097e-05,
      "loss": 0.5211,
      "mean_token_accuracy": 0.8399878926575184,
      "num_tokens": 348995029.0,
      "step": 10935
    },
    {
      "entropy": 0.7776718828827143,
      "epoch": 1.006164426886562,
      "grad_norm": 0.16826890408992767,
      "learning_rate": 6.646425614131934e-05,
      "loss": 0.7814,
      "mean_token_accuracy": 0.7707139998674393,
      "num_tokens": 349026844.0,
      "step": 10936
    },
    {
      "entropy": 0.5574790416285396,
      "epoch": 1.006256433258003,
      "grad_norm": 0.14805026352405548,
      "learning_rate": 6.646118931517772e-05,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8367219753563404,
      "num_tokens": 349059172.0,
      "step": 10937
    },
    {
      "entropy": 0.5391229232773185,
      "epoch": 1.0063484396294444,
      "grad_norm": 0.14946506917476654,
      "learning_rate": 6.64581224890361e-05,
      "loss": 0.5059,
      "mean_token_accuracy": 0.8429195284843445,
      "num_tokens": 349090077.0,
      "step": 10938
    },
    {
      "entropy": 0.6838505603373051,
      "epoch": 1.0064404460008856,
      "grad_norm": 0.1615060269832611,
      "learning_rate": 6.645505566289447e-05,
      "loss": 0.6675,
      "mean_token_accuracy": 0.7931743487715721,
      "num_tokens": 349121816.0,
      "step": 10939
    },
    {
      "entropy": 0.6093126703053713,
      "epoch": 1.0065324523723267,
      "grad_norm": 0.15486103296279907,
      "learning_rate": 6.645198883675284e-05,
      "loss": 0.5863,
      "mean_token_accuracy": 0.8203622475266457,
      "num_tokens": 349153562.0,
      "step": 10940
    },
    {
      "entropy": 0.6195192225277424,
      "epoch": 1.006624458743768,
      "grad_norm": 0.15664668381214142,
      "learning_rate": 6.644892201061122e-05,
      "loss": 0.6182,
      "mean_token_accuracy": 0.8057975433766842,
      "num_tokens": 349185009.0,
      "step": 10941
    },
    {
      "entropy": 0.5902737872675061,
      "epoch": 1.0067164651152092,
      "grad_norm": 0.15882104635238647,
      "learning_rate": 6.644585518446959e-05,
      "loss": 0.5701,
      "mean_token_accuracy": 0.82512466609478,
      "num_tokens": 349216477.0,
      "step": 10942
    },
    {
      "entropy": 0.5902024358510971,
      "epoch": 1.0068084714866505,
      "grad_norm": 0.15452225506305695,
      "learning_rate": 6.644278835832797e-05,
      "loss": 0.5698,
      "mean_token_accuracy": 0.820309292525053,
      "num_tokens": 349248388.0,
      "step": 10943
    },
    {
      "entropy": 0.645845515653491,
      "epoch": 1.0069004778580917,
      "grad_norm": 0.15055815875530243,
      "learning_rate": 6.643972153218634e-05,
      "loss": 0.6184,
      "mean_token_accuracy": 0.8088449202477932,
      "num_tokens": 349280657.0,
      "step": 10944
    },
    {
      "entropy": 0.6308116912841797,
      "epoch": 1.0069924842295328,
      "grad_norm": 0.14734742045402527,
      "learning_rate": 6.643665470604472e-05,
      "loss": 0.599,
      "mean_token_accuracy": 0.8171343021094799,
      "num_tokens": 349312645.0,
      "step": 10945
    },
    {
      "entropy": 0.6608874909579754,
      "epoch": 1.0070844906009742,
      "grad_norm": 0.16142140328884125,
      "learning_rate": 6.643358787990309e-05,
      "loss": 0.6502,
      "mean_token_accuracy": 0.8034578897058964,
      "num_tokens": 349344435.0,
      "step": 10946
    },
    {
      "entropy": 0.6839134944602847,
      "epoch": 1.0071764969724153,
      "grad_norm": 0.15933284163475037,
      "learning_rate": 6.643052105376146e-05,
      "loss": 0.6572,
      "mean_token_accuracy": 0.7983086295425892,
      "num_tokens": 349376854.0,
      "step": 10947
    },
    {
      "entropy": 0.5546489339321852,
      "epoch": 1.0072685033438566,
      "grad_norm": 0.1474950760602951,
      "learning_rate": 6.642745422761984e-05,
      "loss": 0.5513,
      "mean_token_accuracy": 0.8296702243387699,
      "num_tokens": 349408945.0,
      "step": 10948
    },
    {
      "entropy": 0.6891835574060678,
      "epoch": 1.0073605097152978,
      "grad_norm": 0.16101473569869995,
      "learning_rate": 6.642438740147822e-05,
      "loss": 0.6848,
      "mean_token_accuracy": 0.788262203335762,
      "num_tokens": 349441292.0,
      "step": 10949
    },
    {
      "entropy": 0.5566583937034011,
      "epoch": 1.007452516086739,
      "grad_norm": 0.1552775353193283,
      "learning_rate": 6.642132057533659e-05,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8339374624192715,
      "num_tokens": 349473859.0,
      "step": 10950
    },
    {
      "entropy": 0.6355675682425499,
      "epoch": 1.0075445224581803,
      "grad_norm": 0.1586921215057373,
      "learning_rate": 6.641825374919496e-05,
      "loss": 0.6116,
      "mean_token_accuracy": 0.8097077757120132,
      "num_tokens": 349505523.0,
      "step": 10951
    },
    {
      "entropy": 0.5602410202845931,
      "epoch": 1.0076365288296214,
      "grad_norm": 0.15822173655033112,
      "learning_rate": 6.641518692305334e-05,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8315250277519226,
      "num_tokens": 349537651.0,
      "step": 10952
    },
    {
      "entropy": 0.6235916065052152,
      "epoch": 1.0077285352010628,
      "grad_norm": 0.14932972192764282,
      "learning_rate": 6.641212009691172e-05,
      "loss": 0.5953,
      "mean_token_accuracy": 0.8160252645611763,
      "num_tokens": 349569504.0,
      "step": 10953
    },
    {
      "entropy": 0.6084201773628592,
      "epoch": 1.007820541572504,
      "grad_norm": 0.1614835262298584,
      "learning_rate": 6.640905327077009e-05,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8183329068124294,
      "num_tokens": 349600560.0,
      "step": 10954
    },
    {
      "entropy": 0.5823383899405599,
      "epoch": 1.007912547943945,
      "grad_norm": 0.15394338965415955,
      "learning_rate": 6.640598644462845e-05,
      "loss": 0.5813,
      "mean_token_accuracy": 0.8194487243890762,
      "num_tokens": 349633013.0,
      "step": 10955
    },
    {
      "entropy": 0.6999311614781618,
      "epoch": 1.0080045543153864,
      "grad_norm": 0.1714368313550949,
      "learning_rate": 6.640291961848684e-05,
      "loss": 0.7028,
      "mean_token_accuracy": 0.7857029214501381,
      "num_tokens": 349664999.0,
      "step": 10956
    },
    {
      "entropy": 0.6741777062416077,
      "epoch": 1.0080965606868275,
      "grad_norm": 0.16806791722774506,
      "learning_rate": 6.63998527923452e-05,
      "loss": 0.6471,
      "mean_token_accuracy": 0.8027965165674686,
      "num_tokens": 349697099.0,
      "step": 10957
    },
    {
      "entropy": 0.6516195759177208,
      "epoch": 1.0081885670582689,
      "grad_norm": 0.15710417926311493,
      "learning_rate": 6.639678596620358e-05,
      "loss": 0.6365,
      "mean_token_accuracy": 0.8075391985476017,
      "num_tokens": 349729114.0,
      "step": 10958
    },
    {
      "entropy": 0.531251510605216,
      "epoch": 1.00828057342971,
      "grad_norm": 0.15802329778671265,
      "learning_rate": 6.639371914006195e-05,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8457512930035591,
      "num_tokens": 349760772.0,
      "step": 10959
    },
    {
      "entropy": 0.5609109727665782,
      "epoch": 1.0083725798011511,
      "grad_norm": 0.15072046220302582,
      "learning_rate": 6.639065231392033e-05,
      "loss": 0.5403,
      "mean_token_accuracy": 0.8319435901939869,
      "num_tokens": 349793135.0,
      "step": 10960
    },
    {
      "entropy": 0.5749802771024406,
      "epoch": 1.0084645861725925,
      "grad_norm": 0.15694987773895264,
      "learning_rate": 6.63875854877787e-05,
      "loss": 0.5394,
      "mean_token_accuracy": 0.8342496640980244,
      "num_tokens": 349824686.0,
      "step": 10961
    },
    {
      "entropy": 0.5721326526254416,
      "epoch": 1.0085565925440336,
      "grad_norm": 0.15424829721450806,
      "learning_rate": 6.638451866163707e-05,
      "loss": 0.5503,
      "mean_token_accuracy": 0.8304428905248642,
      "num_tokens": 349855901.0,
      "step": 10962
    },
    {
      "entropy": 0.7054499797523022,
      "epoch": 1.008648598915475,
      "grad_norm": 0.16125360131263733,
      "learning_rate": 6.638145183549545e-05,
      "loss": 0.6925,
      "mean_token_accuracy": 0.7898795008659363,
      "num_tokens": 349888121.0,
      "step": 10963
    },
    {
      "entropy": 0.5305404011160135,
      "epoch": 1.0087406052869161,
      "grad_norm": 0.14959126710891724,
      "learning_rate": 6.637838500935383e-05,
      "loss": 0.5203,
      "mean_token_accuracy": 0.8383235782384872,
      "num_tokens": 349919218.0,
      "step": 10964
    },
    {
      "entropy": 0.562034641392529,
      "epoch": 1.0088326116583572,
      "grad_norm": 0.1537920981645584,
      "learning_rate": 6.63753181832122e-05,
      "loss": 0.541,
      "mean_token_accuracy": 0.8300891071557999,
      "num_tokens": 349951335.0,
      "step": 10965
    },
    {
      "entropy": 0.610174274072051,
      "epoch": 1.0089246180297986,
      "grad_norm": 0.159748375415802,
      "learning_rate": 6.637225135707057e-05,
      "loss": 0.6039,
      "mean_token_accuracy": 0.8146920464932919,
      "num_tokens": 349983505.0,
      "step": 10966
    },
    {
      "entropy": 0.6934318616986275,
      "epoch": 1.0090166244012397,
      "grad_norm": 0.16524679958820343,
      "learning_rate": 6.636918453092894e-05,
      "loss": 0.6862,
      "mean_token_accuracy": 0.7928786352276802,
      "num_tokens": 350015068.0,
      "step": 10967
    },
    {
      "entropy": 0.5932855010032654,
      "epoch": 1.009108630772681,
      "grad_norm": 0.16383042931556702,
      "learning_rate": 6.636611770478732e-05,
      "loss": 0.5807,
      "mean_token_accuracy": 0.8195142410695553,
      "num_tokens": 350047588.0,
      "step": 10968
    },
    {
      "entropy": 0.5998427723534405,
      "epoch": 1.0092006371441222,
      "grad_norm": 0.15979379415512085,
      "learning_rate": 6.63630508786457e-05,
      "loss": 0.584,
      "mean_token_accuracy": 0.8185329884290695,
      "num_tokens": 350079461.0,
      "step": 10969
    },
    {
      "entropy": 0.6051503457129002,
      "epoch": 1.0092926435155634,
      "grad_norm": 0.14905397593975067,
      "learning_rate": 6.635998405250407e-05,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8277823403477669,
      "num_tokens": 350111813.0,
      "step": 10970
    },
    {
      "entropy": 0.5949586927890778,
      "epoch": 1.0093846498870047,
      "grad_norm": 0.1488446444272995,
      "learning_rate": 6.635691722636243e-05,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8241960927844048,
      "num_tokens": 350143895.0,
      "step": 10971
    },
    {
      "entropy": 0.5749419275671244,
      "epoch": 1.0094766562584458,
      "grad_norm": 0.1522548794746399,
      "learning_rate": 6.635385040022082e-05,
      "loss": 0.536,
      "mean_token_accuracy": 0.8355530314147472,
      "num_tokens": 350175794.0,
      "step": 10972
    },
    {
      "entropy": 0.5915492009371519,
      "epoch": 1.0095686626298872,
      "grad_norm": 0.14872200787067413,
      "learning_rate": 6.635078357407918e-05,
      "loss": 0.5713,
      "mean_token_accuracy": 0.8215517066419125,
      "num_tokens": 350207935.0,
      "step": 10973
    },
    {
      "entropy": 0.6406087949872017,
      "epoch": 1.0096606690013283,
      "grad_norm": 0.15580610930919647,
      "learning_rate": 6.634771674793757e-05,
      "loss": 0.6191,
      "mean_token_accuracy": 0.8094131574034691,
      "num_tokens": 350240098.0,
      "step": 10974
    },
    {
      "entropy": 0.6285090772435069,
      "epoch": 1.0097526753727695,
      "grad_norm": 0.16348868608474731,
      "learning_rate": 6.634464992179593e-05,
      "loss": 0.6109,
      "mean_token_accuracy": 0.8122440874576569,
      "num_tokens": 350271751.0,
      "step": 10975
    },
    {
      "entropy": 0.5136460815556347,
      "epoch": 1.0098446817442108,
      "grad_norm": 0.15793493390083313,
      "learning_rate": 6.634158309565431e-05,
      "loss": 0.5075,
      "mean_token_accuracy": 0.8440858349204063,
      "num_tokens": 350304167.0,
      "step": 10976
    },
    {
      "entropy": 0.5701566743664443,
      "epoch": 1.009936688115652,
      "grad_norm": 0.1520206332206726,
      "learning_rate": 6.633851626951268e-05,
      "loss": 0.5587,
      "mean_token_accuracy": 0.826384648680687,
      "num_tokens": 350335943.0,
      "step": 10977
    },
    {
      "entropy": 0.609862931072712,
      "epoch": 1.0100286944870933,
      "grad_norm": 0.15355715155601501,
      "learning_rate": 6.633544944337105e-05,
      "loss": 0.6039,
      "mean_token_accuracy": 0.8127379417419434,
      "num_tokens": 350367999.0,
      "step": 10978
    },
    {
      "entropy": 0.62664251960814,
      "epoch": 1.0101207008585344,
      "grad_norm": 0.16613921523094177,
      "learning_rate": 6.633238261722943e-05,
      "loss": 0.6371,
      "mean_token_accuracy": 0.8063435256481171,
      "num_tokens": 350399827.0,
      "step": 10979
    },
    {
      "entropy": 0.5475966851226985,
      "epoch": 1.0102127072299756,
      "grad_norm": 0.1591777503490448,
      "learning_rate": 6.632931579108781e-05,
      "loss": 0.5401,
      "mean_token_accuracy": 0.8286989256739616,
      "num_tokens": 350431104.0,
      "step": 10980
    },
    {
      "entropy": 0.5925411880016327,
      "epoch": 1.010304713601417,
      "grad_norm": 0.16003826260566711,
      "learning_rate": 6.632624896494618e-05,
      "loss": 0.5589,
      "mean_token_accuracy": 0.8251691609621048,
      "num_tokens": 350462814.0,
      "step": 10981
    },
    {
      "entropy": 0.6301700696349144,
      "epoch": 1.010396719972858,
      "grad_norm": 0.15587154030799866,
      "learning_rate": 6.632318213880455e-05,
      "loss": 0.5886,
      "mean_token_accuracy": 0.8166570253670216,
      "num_tokens": 350494381.0,
      "step": 10982
    },
    {
      "entropy": 0.7035482106730342,
      "epoch": 1.0104887263442994,
      "grad_norm": 0.15579040348529816,
      "learning_rate": 6.632011531266293e-05,
      "loss": 0.6676,
      "mean_token_accuracy": 0.7928411737084389,
      "num_tokens": 350526920.0,
      "step": 10983
    },
    {
      "entropy": 0.6481243735179305,
      "epoch": 1.0105807327157406,
      "grad_norm": 0.16516435146331787,
      "learning_rate": 6.63170484865213e-05,
      "loss": 0.6364,
      "mean_token_accuracy": 0.8096832558512688,
      "num_tokens": 350558414.0,
      "step": 10984
    },
    {
      "entropy": 0.6193045303225517,
      "epoch": 1.0106727390871817,
      "grad_norm": 0.15473751723766327,
      "learning_rate": 6.631398166037968e-05,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8105091564357281,
      "num_tokens": 350589818.0,
      "step": 10985
    },
    {
      "entropy": 0.6055394820868969,
      "epoch": 1.010764745458623,
      "grad_norm": 0.16377896070480347,
      "learning_rate": 6.631091483423805e-05,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8150523714721203,
      "num_tokens": 350620985.0,
      "step": 10986
    },
    {
      "entropy": 0.7058002580888569,
      "epoch": 1.0108567518300642,
      "grad_norm": 0.16672080755233765,
      "learning_rate": 6.630784800809643e-05,
      "loss": 0.7181,
      "mean_token_accuracy": 0.7820626869797707,
      "num_tokens": 350652939.0,
      "step": 10987
    },
    {
      "entropy": 0.6873657703399658,
      "epoch": 1.0109487582015055,
      "grad_norm": 0.16129228472709656,
      "learning_rate": 6.63047811819548e-05,
      "loss": 0.6638,
      "mean_token_accuracy": 0.7962864749133587,
      "num_tokens": 350685444.0,
      "step": 10988
    },
    {
      "entropy": 0.5078576104715466,
      "epoch": 1.0110407645729467,
      "grad_norm": 0.14289747178554535,
      "learning_rate": 6.630171435581316e-05,
      "loss": 0.4963,
      "mean_token_accuracy": 0.8459775745868683,
      "num_tokens": 350717967.0,
      "step": 10989
    },
    {
      "entropy": 0.6421841215342283,
      "epoch": 1.0111327709443878,
      "grad_norm": 0.1612612009048462,
      "learning_rate": 6.629864752967155e-05,
      "loss": 0.6254,
      "mean_token_accuracy": 0.8093783594667912,
      "num_tokens": 350749790.0,
      "step": 10990
    },
    {
      "entropy": 0.5625727577134967,
      "epoch": 1.0112247773158292,
      "grad_norm": 0.1508132815361023,
      "learning_rate": 6.629558070352993e-05,
      "loss": 0.5538,
      "mean_token_accuracy": 0.8309045024216175,
      "num_tokens": 350782415.0,
      "step": 10991
    },
    {
      "entropy": 0.6060150675475597,
      "epoch": 1.0113167836872703,
      "grad_norm": 0.1535290628671646,
      "learning_rate": 6.62925138773883e-05,
      "loss": 0.6005,
      "mean_token_accuracy": 0.8175417482852936,
      "num_tokens": 350814096.0,
      "step": 10992
    },
    {
      "entropy": 0.644313232973218,
      "epoch": 1.0114087900587116,
      "grad_norm": 0.1566474288702011,
      "learning_rate": 6.628944705124666e-05,
      "loss": 0.6384,
      "mean_token_accuracy": 0.803574126213789,
      "num_tokens": 350846264.0,
      "step": 10993
    },
    {
      "entropy": 0.58039206918329,
      "epoch": 1.0115007964301528,
      "grad_norm": 0.14902934432029724,
      "learning_rate": 6.628638022510504e-05,
      "loss": 0.5546,
      "mean_token_accuracy": 0.827618919312954,
      "num_tokens": 350878714.0,
      "step": 10994
    },
    {
      "entropy": 0.6707050614058971,
      "epoch": 1.011592802801594,
      "grad_norm": 0.17209863662719727,
      "learning_rate": 6.628331339896343e-05,
      "loss": 0.662,
      "mean_token_accuracy": 0.7986576817929745,
      "num_tokens": 350910076.0,
      "step": 10995
    },
    {
      "entropy": 0.49567429441958666,
      "epoch": 1.0116848091730353,
      "grad_norm": 0.15040670335292816,
      "learning_rate": 6.62802465728218e-05,
      "loss": 0.4668,
      "mean_token_accuracy": 0.8537387624382973,
      "num_tokens": 350942144.0,
      "step": 10996
    },
    {
      "entropy": 0.5120819117873907,
      "epoch": 1.0117768155444764,
      "grad_norm": 0.15633948147296906,
      "learning_rate": 6.627717974668016e-05,
      "loss": 0.4977,
      "mean_token_accuracy": 0.841265644878149,
      "num_tokens": 350974570.0,
      "step": 10997
    },
    {
      "entropy": 0.5664507579058409,
      "epoch": 1.0118688219159178,
      "grad_norm": 0.15833806991577148,
      "learning_rate": 6.627411292053853e-05,
      "loss": 0.5393,
      "mean_token_accuracy": 0.83293117582798,
      "num_tokens": 351006677.0,
      "step": 10998
    },
    {
      "entropy": 0.6160113224759698,
      "epoch": 1.011960828287359,
      "grad_norm": 0.15336838364601135,
      "learning_rate": 6.627104609439691e-05,
      "loss": 0.6148,
      "mean_token_accuracy": 0.8110036738216877,
      "num_tokens": 351039263.0,
      "step": 10999
    },
    {
      "entropy": 0.5967137366533279,
      "epoch": 1.0120528346588,
      "grad_norm": 0.15615053474903107,
      "learning_rate": 6.626797926825529e-05,
      "loss": 0.5948,
      "mean_token_accuracy": 0.8153305947780609,
      "num_tokens": 351071671.0,
      "step": 11000
    },
    {
      "entropy": 0.666102284565568,
      "epoch": 1.0121448410302414,
      "grad_norm": 0.15762457251548767,
      "learning_rate": 6.626491244211366e-05,
      "loss": 0.6487,
      "mean_token_accuracy": 0.8029063232243061,
      "num_tokens": 351103622.0,
      "step": 11001
    },
    {
      "entropy": 0.6005322355777025,
      "epoch": 1.0122368474016825,
      "grad_norm": 0.15800757706165314,
      "learning_rate": 6.626184561597203e-05,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8302970752120018,
      "num_tokens": 351135486.0,
      "step": 11002
    },
    {
      "entropy": 0.6356082502752542,
      "epoch": 1.0123288537731239,
      "grad_norm": 0.1611367017030716,
      "learning_rate": 6.625877878983041e-05,
      "loss": 0.6451,
      "mean_token_accuracy": 0.8019732125103474,
      "num_tokens": 351167625.0,
      "step": 11003
    },
    {
      "entropy": 0.6717119161039591,
      "epoch": 1.012420860144565,
      "grad_norm": 0.1592823714017868,
      "learning_rate": 6.625571196368878e-05,
      "loss": 0.67,
      "mean_token_accuracy": 0.795979417860508,
      "num_tokens": 351199377.0,
      "step": 11004
    },
    {
      "entropy": 0.7132770977914333,
      "epoch": 1.0125128665160061,
      "grad_norm": 0.16297759115695953,
      "learning_rate": 6.625264513754716e-05,
      "loss": 0.7032,
      "mean_token_accuracy": 0.7886040546000004,
      "num_tokens": 351230978.0,
      "step": 11005
    },
    {
      "entropy": 0.6637182980775833,
      "epoch": 1.0126048728874475,
      "grad_norm": 0.16253751516342163,
      "learning_rate": 6.624957831140553e-05,
      "loss": 0.654,
      "mean_token_accuracy": 0.8014400750398636,
      "num_tokens": 351263702.0,
      "step": 11006
    },
    {
      "entropy": 0.6757617415860295,
      "epoch": 1.0126968792588886,
      "grad_norm": 0.1623353660106659,
      "learning_rate": 6.624651148526391e-05,
      "loss": 0.6515,
      "mean_token_accuracy": 0.7981824949383736,
      "num_tokens": 351295243.0,
      "step": 11007
    },
    {
      "entropy": 0.6406693495810032,
      "epoch": 1.01278888563033,
      "grad_norm": 0.15957331657409668,
      "learning_rate": 6.624344465912228e-05,
      "loss": 0.637,
      "mean_token_accuracy": 0.8013305403292179,
      "num_tokens": 351326849.0,
      "step": 11008
    },
    {
      "entropy": 0.5494940634816885,
      "epoch": 1.0128808920017711,
      "grad_norm": 0.16706079244613647,
      "learning_rate": 6.624037783298064e-05,
      "loss": 0.5326,
      "mean_token_accuracy": 0.8353089094161987,
      "num_tokens": 351358129.0,
      "step": 11009
    },
    {
      "entropy": 0.5769694522023201,
      "epoch": 1.0129728983732122,
      "grad_norm": 0.15138742327690125,
      "learning_rate": 6.623731100683903e-05,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8291705846786499,
      "num_tokens": 351390505.0,
      "step": 11010
    },
    {
      "entropy": 0.6034260299056768,
      "epoch": 1.0130649047446536,
      "grad_norm": 0.15104998648166656,
      "learning_rate": 6.623424418069741e-05,
      "loss": 0.5797,
      "mean_token_accuracy": 0.8206226788461208,
      "num_tokens": 351422595.0,
      "step": 11011
    },
    {
      "entropy": 0.6304898988455534,
      "epoch": 1.0131569111160947,
      "grad_norm": 0.15169267356395721,
      "learning_rate": 6.623117735455577e-05,
      "loss": 0.6111,
      "mean_token_accuracy": 0.8102647289633751,
      "num_tokens": 351454072.0,
      "step": 11012
    },
    {
      "entropy": 0.5484110731631517,
      "epoch": 1.013248917487536,
      "grad_norm": 0.15564821660518646,
      "learning_rate": 6.622811052841414e-05,
      "loss": 0.5263,
      "mean_token_accuracy": 0.835785299539566,
      "num_tokens": 351484801.0,
      "step": 11013
    },
    {
      "entropy": 0.6683779153972864,
      "epoch": 1.0133409238589772,
      "grad_norm": 0.15732859075069427,
      "learning_rate": 6.622504370227252e-05,
      "loss": 0.6509,
      "mean_token_accuracy": 0.8044260814785957,
      "num_tokens": 351517014.0,
      "step": 11014
    },
    {
      "entropy": 0.641270337626338,
      "epoch": 1.0134329302304184,
      "grad_norm": 0.15394315123558044,
      "learning_rate": 6.622197687613089e-05,
      "loss": 0.6334,
      "mean_token_accuracy": 0.8023986294865608,
      "num_tokens": 351549521.0,
      "step": 11015
    },
    {
      "entropy": 0.6790159214287996,
      "epoch": 1.0135249366018597,
      "grad_norm": 0.16354414820671082,
      "learning_rate": 6.621891004998927e-05,
      "loss": 0.6653,
      "mean_token_accuracy": 0.7977433577179909,
      "num_tokens": 351581321.0,
      "step": 11016
    },
    {
      "entropy": 0.6376621685922146,
      "epoch": 1.0136169429733008,
      "grad_norm": 0.1522216945886612,
      "learning_rate": 6.621584322384764e-05,
      "loss": 0.6021,
      "mean_token_accuracy": 0.8155758269131184,
      "num_tokens": 351612829.0,
      "step": 11017
    },
    {
      "entropy": 0.5702707208693027,
      "epoch": 1.0137089493447422,
      "grad_norm": 0.15142840147018433,
      "learning_rate": 6.621277639770602e-05,
      "loss": 0.5606,
      "mean_token_accuracy": 0.8264832831919193,
      "num_tokens": 351644458.0,
      "step": 11018
    },
    {
      "entropy": 0.6537159495055676,
      "epoch": 1.0138009557161833,
      "grad_norm": 0.1609329879283905,
      "learning_rate": 6.620970957156439e-05,
      "loss": 0.6308,
      "mean_token_accuracy": 0.8043663874268532,
      "num_tokens": 351676551.0,
      "step": 11019
    },
    {
      "entropy": 0.6457998380064964,
      "epoch": 1.0138929620876245,
      "grad_norm": 0.15384197235107422,
      "learning_rate": 6.620664274542276e-05,
      "loss": 0.6386,
      "mean_token_accuracy": 0.8026502579450607,
      "num_tokens": 351707895.0,
      "step": 11020
    },
    {
      "entropy": 0.523009404540062,
      "epoch": 1.0139849684590658,
      "grad_norm": 0.14022475481033325,
      "learning_rate": 6.620357591928114e-05,
      "loss": 0.493,
      "mean_token_accuracy": 0.8502502925693989,
      "num_tokens": 351739588.0,
      "step": 11021
    },
    {
      "entropy": 0.6179595161229372,
      "epoch": 1.014076974830507,
      "grad_norm": 0.15183477103710175,
      "learning_rate": 6.620050909313952e-05,
      "loss": 0.6009,
      "mean_token_accuracy": 0.8168337419629097,
      "num_tokens": 351772001.0,
      "step": 11022
    },
    {
      "entropy": 0.6887759380042553,
      "epoch": 1.0141689812019483,
      "grad_norm": 0.15569652616977692,
      "learning_rate": 6.619744226699789e-05,
      "loss": 0.6744,
      "mean_token_accuracy": 0.7908788993954659,
      "num_tokens": 351804445.0,
      "step": 11023
    },
    {
      "entropy": 0.5930067300796509,
      "epoch": 1.0142609875733894,
      "grad_norm": 0.14936436712741852,
      "learning_rate": 6.619437544085626e-05,
      "loss": 0.5644,
      "mean_token_accuracy": 0.8288798965513706,
      "num_tokens": 351836266.0,
      "step": 11024
    },
    {
      "entropy": 0.6786847859621048,
      "epoch": 1.0143529939448306,
      "grad_norm": 0.15696190297603607,
      "learning_rate": 6.619130861471463e-05,
      "loss": 0.6397,
      "mean_token_accuracy": 0.8030541352927685,
      "num_tokens": 351868130.0,
      "step": 11025
    },
    {
      "entropy": 0.5820025503635406,
      "epoch": 1.014445000316272,
      "grad_norm": 0.1530900001525879,
      "learning_rate": 6.618824178857302e-05,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8288661725819111,
      "num_tokens": 351900647.0,
      "step": 11026
    },
    {
      "entropy": 0.6647491734474897,
      "epoch": 1.014537006687713,
      "grad_norm": 0.15899589657783508,
      "learning_rate": 6.618517496243139e-05,
      "loss": 0.643,
      "mean_token_accuracy": 0.8025966063141823,
      "num_tokens": 351932774.0,
      "step": 11027
    },
    {
      "entropy": 0.6610039565712214,
      "epoch": 1.0146290130591544,
      "grad_norm": 0.16163918375968933,
      "learning_rate": 6.618210813628976e-05,
      "loss": 0.6559,
      "mean_token_accuracy": 0.8008415885269642,
      "num_tokens": 351964528.0,
      "step": 11028
    },
    {
      "entropy": 0.6319720186293125,
      "epoch": 1.0147210194305956,
      "grad_norm": 0.1667981892824173,
      "learning_rate": 6.617904131014812e-05,
      "loss": 0.6249,
      "mean_token_accuracy": 0.8077455908060074,
      "num_tokens": 351996427.0,
      "step": 11029
    },
    {
      "entropy": 0.7051077205687761,
      "epoch": 1.0148130258020367,
      "grad_norm": 0.1610659807920456,
      "learning_rate": 6.61759744840065e-05,
      "loss": 0.6961,
      "mean_token_accuracy": 0.7895935513079166,
      "num_tokens": 352027774.0,
      "step": 11030
    },
    {
      "entropy": 0.6064496878534555,
      "epoch": 1.014905032173478,
      "grad_norm": 0.15710285305976868,
      "learning_rate": 6.617290765786489e-05,
      "loss": 0.6014,
      "mean_token_accuracy": 0.8130242973566055,
      "num_tokens": 352059263.0,
      "step": 11031
    },
    {
      "entropy": 0.6607928704470396,
      "epoch": 1.0149970385449192,
      "grad_norm": 0.15530021488666534,
      "learning_rate": 6.616984083172325e-05,
      "loss": 0.6525,
      "mean_token_accuracy": 0.8018679060041904,
      "num_tokens": 352091093.0,
      "step": 11032
    },
    {
      "entropy": 0.6263420656323433,
      "epoch": 1.0150890449163605,
      "grad_norm": 0.17862440645694733,
      "learning_rate": 6.616677400558162e-05,
      "loss": 0.6186,
      "mean_token_accuracy": 0.8107434585690498,
      "num_tokens": 352122184.0,
      "step": 11033
    },
    {
      "entropy": 0.6126333978027105,
      "epoch": 1.0151810512878017,
      "grad_norm": 0.15853089094161987,
      "learning_rate": 6.616370717944e-05,
      "loss": 0.5992,
      "mean_token_accuracy": 0.8119876272976398,
      "num_tokens": 352153216.0,
      "step": 11034
    },
    {
      "entropy": 0.6092394208535552,
      "epoch": 1.0152730576592428,
      "grad_norm": 0.15811309218406677,
      "learning_rate": 6.616064035329837e-05,
      "loss": 0.5874,
      "mean_token_accuracy": 0.8164487555623055,
      "num_tokens": 352184968.0,
      "step": 11035
    },
    {
      "entropy": 0.6375425755977631,
      "epoch": 1.0153650640306842,
      "grad_norm": 0.1635967642068863,
      "learning_rate": 6.615757352715675e-05,
      "loss": 0.6192,
      "mean_token_accuracy": 0.810476865619421,
      "num_tokens": 352217595.0,
      "step": 11036
    },
    {
      "entropy": 0.5916134575381875,
      "epoch": 1.0154570704021253,
      "grad_norm": 0.17333415150642395,
      "learning_rate": 6.615450670101512e-05,
      "loss": 0.5762,
      "mean_token_accuracy": 0.8232266828417778,
      "num_tokens": 352249816.0,
      "step": 11037
    },
    {
      "entropy": 0.6880120318382978,
      "epoch": 1.0155490767735667,
      "grad_norm": 0.15956367552280426,
      "learning_rate": 6.61514398748735e-05,
      "loss": 0.6615,
      "mean_token_accuracy": 0.7998226620256901,
      "num_tokens": 352280868.0,
      "step": 11038
    },
    {
      "entropy": 0.6034432090818882,
      "epoch": 1.0156410831450078,
      "grad_norm": 0.1485728919506073,
      "learning_rate": 6.614837304873187e-05,
      "loss": 0.5773,
      "mean_token_accuracy": 0.8206797204911709,
      "num_tokens": 352313438.0,
      "step": 11039
    },
    {
      "entropy": 0.520239258185029,
      "epoch": 1.015733089516449,
      "grad_norm": 0.15643763542175293,
      "learning_rate": 6.614530622259024e-05,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8435681946575642,
      "num_tokens": 352345533.0,
      "step": 11040
    },
    {
      "entropy": 0.584861671552062,
      "epoch": 1.0158250958878903,
      "grad_norm": 0.14934131503105164,
      "learning_rate": 6.614223939644862e-05,
      "loss": 0.5503,
      "mean_token_accuracy": 0.828768890351057,
      "num_tokens": 352377612.0,
      "step": 11041
    },
    {
      "entropy": 0.6946605909615755,
      "epoch": 1.0159171022593314,
      "grad_norm": 0.1678958237171173,
      "learning_rate": 6.6139172570307e-05,
      "loss": 0.6912,
      "mean_token_accuracy": 0.7862155400216579,
      "num_tokens": 352408858.0,
      "step": 11042
    },
    {
      "entropy": 0.6227648006752133,
      "epoch": 1.0160091086307728,
      "grad_norm": 0.16486507654190063,
      "learning_rate": 6.613610574416537e-05,
      "loss": 0.5947,
      "mean_token_accuracy": 0.8178276941180229,
      "num_tokens": 352440565.0,
      "step": 11043
    },
    {
      "entropy": 0.49876074213534594,
      "epoch": 1.016101115002214,
      "grad_norm": 0.15790380537509918,
      "learning_rate": 6.613303891802374e-05,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8522159047424793,
      "num_tokens": 352472008.0,
      "step": 11044
    },
    {
      "entropy": 0.6395889539271593,
      "epoch": 1.016193121373655,
      "grad_norm": 0.1610896736383438,
      "learning_rate": 6.612997209188212e-05,
      "loss": 0.6201,
      "mean_token_accuracy": 0.8103026896715164,
      "num_tokens": 352503770.0,
      "step": 11045
    },
    {
      "entropy": 0.6597904898226261,
      "epoch": 1.0162851277450964,
      "grad_norm": 0.1929241269826889,
      "learning_rate": 6.612690526574049e-05,
      "loss": 0.6576,
      "mean_token_accuracy": 0.7993140108883381,
      "num_tokens": 352535352.0,
      "step": 11046
    },
    {
      "entropy": 0.5410220241174102,
      "epoch": 1.0163771341165375,
      "grad_norm": 0.15348826348781586,
      "learning_rate": 6.612383843959887e-05,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8335313722491264,
      "num_tokens": 352567581.0,
      "step": 11047
    },
    {
      "entropy": 0.6072694230824709,
      "epoch": 1.0164691404879789,
      "grad_norm": 0.15070106089115143,
      "learning_rate": 6.612077161345724e-05,
      "loss": 0.5987,
      "mean_token_accuracy": 0.817675918340683,
      "num_tokens": 352600221.0,
      "step": 11048
    },
    {
      "entropy": 0.7117509618401527,
      "epoch": 1.01656114685942,
      "grad_norm": 0.1611892431974411,
      "learning_rate": 6.611770478731562e-05,
      "loss": 0.6934,
      "mean_token_accuracy": 0.78482910618186,
      "num_tokens": 352632064.0,
      "step": 11049
    },
    {
      "entropy": 0.6598198786377907,
      "epoch": 1.0166531532308611,
      "grad_norm": 0.15813371539115906,
      "learning_rate": 6.611463796117398e-05,
      "loss": 0.6489,
      "mean_token_accuracy": 0.8054358623921871,
      "num_tokens": 352663853.0,
      "step": 11050
    },
    {
      "entropy": 0.6303966827690601,
      "epoch": 1.0167451596023025,
      "grad_norm": 0.1589166671037674,
      "learning_rate": 6.611157113503235e-05,
      "loss": 0.6335,
      "mean_token_accuracy": 0.8059895001351833,
      "num_tokens": 352696118.0,
      "step": 11051
    },
    {
      "entropy": 0.5621340526267886,
      "epoch": 1.0168371659737436,
      "grad_norm": 0.16869229078292847,
      "learning_rate": 6.610850430889073e-05,
      "loss": 0.5503,
      "mean_token_accuracy": 0.8305464386940002,
      "num_tokens": 352727426.0,
      "step": 11052
    },
    {
      "entropy": 0.7095607733353972,
      "epoch": 1.016929172345185,
      "grad_norm": 0.1702626794576645,
      "learning_rate": 6.610543748274912e-05,
      "loss": 0.6961,
      "mean_token_accuracy": 0.7891273498535156,
      "num_tokens": 352759182.0,
      "step": 11053
    },
    {
      "entropy": 0.6976561062037945,
      "epoch": 1.0170211787166261,
      "grad_norm": 0.1655324101448059,
      "learning_rate": 6.610237065660748e-05,
      "loss": 0.6938,
      "mean_token_accuracy": 0.7891455292701721,
      "num_tokens": 352791901.0,
      "step": 11054
    },
    {
      "entropy": 0.5602178629487753,
      "epoch": 1.0171131850880673,
      "grad_norm": 0.15013176202774048,
      "learning_rate": 6.609930383046585e-05,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8342827260494232,
      "num_tokens": 352823951.0,
      "step": 11055
    },
    {
      "entropy": 0.7577825877815485,
      "epoch": 1.0172051914595086,
      "grad_norm": 0.16746200621128082,
      "learning_rate": 6.609623700432422e-05,
      "loss": 0.7421,
      "mean_token_accuracy": 0.7731116600334644,
      "num_tokens": 352854960.0,
      "step": 11056
    },
    {
      "entropy": 0.7051164079457521,
      "epoch": 1.0172971978309497,
      "grad_norm": 0.1649230271577835,
      "learning_rate": 6.60931701781826e-05,
      "loss": 0.6703,
      "mean_token_accuracy": 0.7957754954695702,
      "num_tokens": 352886238.0,
      "step": 11057
    },
    {
      "entropy": 0.671159066259861,
      "epoch": 1.017389204202391,
      "grad_norm": 0.1660454273223877,
      "learning_rate": 6.609010335204098e-05,
      "loss": 0.6707,
      "mean_token_accuracy": 0.7966479770839214,
      "num_tokens": 352918501.0,
      "step": 11058
    },
    {
      "entropy": 0.6223780568689108,
      "epoch": 1.0174812105738322,
      "grad_norm": 0.15739114582538605,
      "learning_rate": 6.608703652589935e-05,
      "loss": 0.5924,
      "mean_token_accuracy": 0.8179436028003693,
      "num_tokens": 352949502.0,
      "step": 11059
    },
    {
      "entropy": 0.5718491282314062,
      "epoch": 1.0175732169452734,
      "grad_norm": 0.15061341226100922,
      "learning_rate": 6.608396969975772e-05,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8309592977166176,
      "num_tokens": 352981143.0,
      "step": 11060
    },
    {
      "entropy": 0.6058708969503641,
      "epoch": 1.0176652233167147,
      "grad_norm": 0.16067953407764435,
      "learning_rate": 6.60809028736161e-05,
      "loss": 0.5875,
      "mean_token_accuracy": 0.8155613988637924,
      "num_tokens": 353012103.0,
      "step": 11061
    },
    {
      "entropy": 0.6546070370823145,
      "epoch": 1.0177572296881559,
      "grad_norm": 0.15826952457427979,
      "learning_rate": 6.607783604747447e-05,
      "loss": 0.6181,
      "mean_token_accuracy": 0.8102425299584866,
      "num_tokens": 353043852.0,
      "step": 11062
    },
    {
      "entropy": 0.7171711754053831,
      "epoch": 1.0178492360595972,
      "grad_norm": 0.16422703862190247,
      "learning_rate": 6.607476922133285e-05,
      "loss": 0.7044,
      "mean_token_accuracy": 0.7853160835802555,
      "num_tokens": 353075750.0,
      "step": 11063
    },
    {
      "entropy": 0.6430658120661974,
      "epoch": 1.0179412424310383,
      "grad_norm": 0.16197578608989716,
      "learning_rate": 6.607170239519122e-05,
      "loss": 0.6183,
      "mean_token_accuracy": 0.8090569227933884,
      "num_tokens": 353108035.0,
      "step": 11064
    },
    {
      "entropy": 0.6222786456346512,
      "epoch": 1.0180332488024795,
      "grad_norm": 0.15631158649921417,
      "learning_rate": 6.60686355690496e-05,
      "loss": 0.6212,
      "mean_token_accuracy": 0.8132646605372429,
      "num_tokens": 353139976.0,
      "step": 11065
    },
    {
      "entropy": 0.7424217462539673,
      "epoch": 1.0181252551739208,
      "grad_norm": 0.16745689511299133,
      "learning_rate": 6.606556874290797e-05,
      "loss": 0.736,
      "mean_token_accuracy": 0.7771452255547047,
      "num_tokens": 353171960.0,
      "step": 11066
    },
    {
      "entropy": 0.5685632722452283,
      "epoch": 1.018217261545362,
      "grad_norm": 0.1556652933359146,
      "learning_rate": 6.606250191676633e-05,
      "loss": 0.558,
      "mean_token_accuracy": 0.8258872367441654,
      "num_tokens": 353203958.0,
      "step": 11067
    },
    {
      "entropy": 0.6338471611961722,
      "epoch": 1.0183092679168033,
      "grad_norm": 0.1676892191171646,
      "learning_rate": 6.605943509062471e-05,
      "loss": 0.6208,
      "mean_token_accuracy": 0.8105869218707085,
      "num_tokens": 353236110.0,
      "step": 11068
    },
    {
      "entropy": 0.6272645900025964,
      "epoch": 1.0184012742882445,
      "grad_norm": 0.15850459039211273,
      "learning_rate": 6.60563682644831e-05,
      "loss": 0.6221,
      "mean_token_accuracy": 0.8093082569539547,
      "num_tokens": 353268695.0,
      "step": 11069
    },
    {
      "entropy": 0.6914626397192478,
      "epoch": 1.0184932806596856,
      "grad_norm": 0.16093946993350983,
      "learning_rate": 6.605330143834146e-05,
      "loss": 0.6729,
      "mean_token_accuracy": 0.8020366281270981,
      "num_tokens": 353300987.0,
      "step": 11070
    },
    {
      "entropy": 0.6326673608273268,
      "epoch": 1.018585287031127,
      "grad_norm": 0.1513955295085907,
      "learning_rate": 6.605023461219983e-05,
      "loss": 0.6181,
      "mean_token_accuracy": 0.8095074705779552,
      "num_tokens": 353332672.0,
      "step": 11071
    },
    {
      "entropy": 0.5841182675212622,
      "epoch": 1.018677293402568,
      "grad_norm": 0.14835195243358612,
      "learning_rate": 6.60471677860582e-05,
      "loss": 0.572,
      "mean_token_accuracy": 0.8215572461485863,
      "num_tokens": 353363721.0,
      "step": 11072
    },
    {
      "entropy": 0.6309042535722256,
      "epoch": 1.0187692997740094,
      "grad_norm": 0.1626461297273636,
      "learning_rate": 6.60441009599166e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.8123874999582767,
      "num_tokens": 353394777.0,
      "step": 11073
    },
    {
      "entropy": 0.6438672076910734,
      "epoch": 1.0188613061454506,
      "grad_norm": 0.16023921966552734,
      "learning_rate": 6.604103413377496e-05,
      "loss": 0.627,
      "mean_token_accuracy": 0.8075745515525341,
      "num_tokens": 353427491.0,
      "step": 11074
    },
    {
      "entropy": 0.5389050720259547,
      "epoch": 1.0189533125168917,
      "grad_norm": 0.15460754930973053,
      "learning_rate": 6.603796730763333e-05,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8352979086339474,
      "num_tokens": 353458566.0,
      "step": 11075
    },
    {
      "entropy": 0.525908867828548,
      "epoch": 1.019045318888333,
      "grad_norm": 0.1620456576347351,
      "learning_rate": 6.603490048149171e-05,
      "loss": 0.5184,
      "mean_token_accuracy": 0.8396496847271919,
      "num_tokens": 353490866.0,
      "step": 11076
    },
    {
      "entropy": 0.556233175098896,
      "epoch": 1.0191373252597742,
      "grad_norm": 0.15255925059318542,
      "learning_rate": 6.603183365535008e-05,
      "loss": 0.5375,
      "mean_token_accuracy": 0.836491834372282,
      "num_tokens": 353522873.0,
      "step": 11077
    },
    {
      "entropy": 0.6604852434247732,
      "epoch": 1.0192293316312155,
      "grad_norm": 0.16136911511421204,
      "learning_rate": 6.602876682920846e-05,
      "loss": 0.6423,
      "mean_token_accuracy": 0.8049019947648048,
      "num_tokens": 353555373.0,
      "step": 11078
    },
    {
      "entropy": 0.6759150288999081,
      "epoch": 1.0193213380026567,
      "grad_norm": 0.16325074434280396,
      "learning_rate": 6.602570000306683e-05,
      "loss": 0.6668,
      "mean_token_accuracy": 0.7975920960307121,
      "num_tokens": 353586424.0,
      "step": 11079
    },
    {
      "entropy": 0.6414695112034678,
      "epoch": 1.0194133443740978,
      "grad_norm": 0.15506097674369812,
      "learning_rate": 6.602263317692521e-05,
      "loss": 0.6327,
      "mean_token_accuracy": 0.8071647062897682,
      "num_tokens": 353617784.0,
      "step": 11080
    },
    {
      "entropy": 0.6260799784213305,
      "epoch": 1.0195053507455392,
      "grad_norm": 0.15215231478214264,
      "learning_rate": 6.601956635078358e-05,
      "loss": 0.6022,
      "mean_token_accuracy": 0.8158596158027649,
      "num_tokens": 353650116.0,
      "step": 11081
    },
    {
      "entropy": 0.6704672025516629,
      "epoch": 1.0195973571169803,
      "grad_norm": 0.15849238634109497,
      "learning_rate": 6.601649952464195e-05,
      "loss": 0.6467,
      "mean_token_accuracy": 0.803951159119606,
      "num_tokens": 353682256.0,
      "step": 11082
    },
    {
      "entropy": 0.665842049755156,
      "epoch": 1.0196893634884217,
      "grad_norm": 0.15945442020893097,
      "learning_rate": 6.601343269850033e-05,
      "loss": 0.6464,
      "mean_token_accuracy": 0.8055992983281612,
      "num_tokens": 353714573.0,
      "step": 11083
    },
    {
      "entropy": 0.6950315218418837,
      "epoch": 1.0197813698598628,
      "grad_norm": 0.15773607790470123,
      "learning_rate": 6.601036587235871e-05,
      "loss": 0.6515,
      "mean_token_accuracy": 0.792545922100544,
      "num_tokens": 353746099.0,
      "step": 11084
    },
    {
      "entropy": 0.6785743068903685,
      "epoch": 1.019873376231304,
      "grad_norm": 0.16946280002593994,
      "learning_rate": 6.600729904621708e-05,
      "loss": 0.666,
      "mean_token_accuracy": 0.7984946109354496,
      "num_tokens": 353778522.0,
      "step": 11085
    },
    {
      "entropy": 0.6115368194878101,
      "epoch": 1.0199653826027453,
      "grad_norm": 0.16101215779781342,
      "learning_rate": 6.600423222007544e-05,
      "loss": 0.5914,
      "mean_token_accuracy": 0.8179872743785381,
      "num_tokens": 353810381.0,
      "step": 11086
    },
    {
      "entropy": 0.6322286948561668,
      "epoch": 1.0200573889741864,
      "grad_norm": 0.15679097175598145,
      "learning_rate": 6.600116539393381e-05,
      "loss": 0.6324,
      "mean_token_accuracy": 0.8060303702950478,
      "num_tokens": 353842391.0,
      "step": 11087
    },
    {
      "entropy": 0.6115294303745031,
      "epoch": 1.0201493953456278,
      "grad_norm": 0.16463007032871246,
      "learning_rate": 6.59980985677922e-05,
      "loss": 0.5904,
      "mean_token_accuracy": 0.8196417726576328,
      "num_tokens": 353874222.0,
      "step": 11088
    },
    {
      "entropy": 0.5328229684382677,
      "epoch": 1.020241401717069,
      "grad_norm": 0.16088560223579407,
      "learning_rate": 6.599503174165058e-05,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8386833630502224,
      "num_tokens": 353905412.0,
      "step": 11089
    },
    {
      "entropy": 0.6908662151545286,
      "epoch": 1.02033340808851,
      "grad_norm": 0.16650280356407166,
      "learning_rate": 6.599196491550894e-05,
      "loss": 0.6678,
      "mean_token_accuracy": 0.7993615232408047,
      "num_tokens": 353936790.0,
      "step": 11090
    },
    {
      "entropy": 0.6337641887366772,
      "epoch": 1.0204254144599514,
      "grad_norm": 0.16658952832221985,
      "learning_rate": 6.598889808936731e-05,
      "loss": 0.6196,
      "mean_token_accuracy": 0.8101487830281258,
      "num_tokens": 353969118.0,
      "step": 11091
    },
    {
      "entropy": 0.5568565288558602,
      "epoch": 1.0205174208313925,
      "grad_norm": 0.15186917781829834,
      "learning_rate": 6.598583126322569e-05,
      "loss": 0.5538,
      "mean_token_accuracy": 0.830848153680563,
      "num_tokens": 354000755.0,
      "step": 11092
    },
    {
      "entropy": 0.6763074500486255,
      "epoch": 1.0206094272028339,
      "grad_norm": 0.1632208377122879,
      "learning_rate": 6.598276443708406e-05,
      "loss": 0.6561,
      "mean_token_accuracy": 0.7987980842590332,
      "num_tokens": 354032457.0,
      "step": 11093
    },
    {
      "entropy": 0.5611438155174255,
      "epoch": 1.020701433574275,
      "grad_norm": 0.14878307282924652,
      "learning_rate": 6.597969761094244e-05,
      "loss": 0.5432,
      "mean_token_accuracy": 0.8337280079722404,
      "num_tokens": 354064795.0,
      "step": 11094
    },
    {
      "entropy": 0.6106539517641068,
      "epoch": 1.0207934399457161,
      "grad_norm": 0.15188337862491608,
      "learning_rate": 6.597663078480081e-05,
      "loss": 0.5782,
      "mean_token_accuracy": 0.8243507705628872,
      "num_tokens": 354096675.0,
      "step": 11095
    },
    {
      "entropy": 0.6159883178770542,
      "epoch": 1.0208854463171575,
      "grad_norm": 0.1543053239583969,
      "learning_rate": 6.597356395865919e-05,
      "loss": 0.6025,
      "mean_token_accuracy": 0.8131205439567566,
      "num_tokens": 354128792.0,
      "step": 11096
    },
    {
      "entropy": 0.5276921344920993,
      "epoch": 1.0209774526885986,
      "grad_norm": 0.1519629955291748,
      "learning_rate": 6.597049713251756e-05,
      "loss": 0.5013,
      "mean_token_accuracy": 0.8412129320204258,
      "num_tokens": 354160859.0,
      "step": 11097
    },
    {
      "entropy": 0.5842585302889347,
      "epoch": 1.02106945906004,
      "grad_norm": 0.1553962528705597,
      "learning_rate": 6.596743030637593e-05,
      "loss": 0.5869,
      "mean_token_accuracy": 0.8183368258178234,
      "num_tokens": 354192692.0,
      "step": 11098
    },
    {
      "entropy": 0.548811499029398,
      "epoch": 1.0211614654314811,
      "grad_norm": 0.16364656388759613,
      "learning_rate": 6.596436348023431e-05,
      "loss": 0.5419,
      "mean_token_accuracy": 0.8357716351747513,
      "num_tokens": 354223994.0,
      "step": 11099
    },
    {
      "entropy": 0.654418108984828,
      "epoch": 1.0212534718029223,
      "grad_norm": 0.1650916039943695,
      "learning_rate": 6.596129665409269e-05,
      "loss": 0.6564,
      "mean_token_accuracy": 0.8031101487576962,
      "num_tokens": 354255905.0,
      "step": 11100
    },
    {
      "entropy": 0.6365551818162203,
      "epoch": 1.0213454781743636,
      "grad_norm": 0.15012800693511963,
      "learning_rate": 6.595822982795106e-05,
      "loss": 0.6028,
      "mean_token_accuracy": 0.8129064179956913,
      "num_tokens": 354287931.0,
      "step": 11101
    },
    {
      "entropy": 0.5609106216579676,
      "epoch": 1.0214374845458047,
      "grad_norm": 0.14666463434696198,
      "learning_rate": 6.595516300180943e-05,
      "loss": 0.5467,
      "mean_token_accuracy": 0.8333580866456032,
      "num_tokens": 354319990.0,
      "step": 11102
    },
    {
      "entropy": 0.6174989379942417,
      "epoch": 1.021529490917246,
      "grad_norm": 0.15569953620433807,
      "learning_rate": 6.59520961756678e-05,
      "loss": 0.6041,
      "mean_token_accuracy": 0.8138704113662243,
      "num_tokens": 354352317.0,
      "step": 11103
    },
    {
      "entropy": 0.6093851318582892,
      "epoch": 1.0216214972886872,
      "grad_norm": 0.1589273363351822,
      "learning_rate": 6.594902934952617e-05,
      "loss": 0.6056,
      "mean_token_accuracy": 0.8145916312932968,
      "num_tokens": 354384415.0,
      "step": 11104
    },
    {
      "entropy": 0.6420521875843406,
      "epoch": 1.0217135036601284,
      "grad_norm": 0.15510326623916626,
      "learning_rate": 6.594596252338456e-05,
      "loss": 0.6177,
      "mean_token_accuracy": 0.8103531524538994,
      "num_tokens": 354416747.0,
      "step": 11105
    },
    {
      "entropy": 0.6994703970849514,
      "epoch": 1.0218055100315697,
      "grad_norm": 0.1632402092218399,
      "learning_rate": 6.594289569724292e-05,
      "loss": 0.6826,
      "mean_token_accuracy": 0.792529035359621,
      "num_tokens": 354448328.0,
      "step": 11106
    },
    {
      "entropy": 0.6300539821386337,
      "epoch": 1.0218975164030109,
      "grad_norm": 0.1669742316007614,
      "learning_rate": 6.59398288711013e-05,
      "loss": 0.6314,
      "mean_token_accuracy": 0.8051268570125103,
      "num_tokens": 354479817.0,
      "step": 11107
    },
    {
      "entropy": 0.5499340174719691,
      "epoch": 1.0219895227744522,
      "grad_norm": 0.1526762694120407,
      "learning_rate": 6.593676204495967e-05,
      "loss": 0.5286,
      "mean_token_accuracy": 0.8377416767179966,
      "num_tokens": 354512339.0,
      "step": 11108
    },
    {
      "entropy": 0.6236615059897304,
      "epoch": 1.0220815291458933,
      "grad_norm": 0.16852763295173645,
      "learning_rate": 6.593369521881804e-05,
      "loss": 0.605,
      "mean_token_accuracy": 0.8152129165828228,
      "num_tokens": 354543776.0,
      "step": 11109
    },
    {
      "entropy": 0.5942573705688119,
      "epoch": 1.0221735355173345,
      "grad_norm": 0.15385645627975464,
      "learning_rate": 6.593062839267642e-05,
      "loss": 0.5855,
      "mean_token_accuracy": 0.8229169584810734,
      "num_tokens": 354575433.0,
      "step": 11110
    },
    {
      "entropy": 0.6644972711801529,
      "epoch": 1.0222655418887758,
      "grad_norm": 0.16244442760944366,
      "learning_rate": 6.59275615665348e-05,
      "loss": 0.6657,
      "mean_token_accuracy": 0.7971116192638874,
      "num_tokens": 354608201.0,
      "step": 11111
    },
    {
      "entropy": 0.5990500869229436,
      "epoch": 1.022357548260217,
      "grad_norm": 0.1672416776418686,
      "learning_rate": 6.592449474039317e-05,
      "loss": 0.5618,
      "mean_token_accuracy": 0.8262341171503067,
      "num_tokens": 354640379.0,
      "step": 11112
    },
    {
      "entropy": 0.6213775053620338,
      "epoch": 1.0224495546316583,
      "grad_norm": 0.1602761298418045,
      "learning_rate": 6.592142791425154e-05,
      "loss": 0.5926,
      "mean_token_accuracy": 0.8193947710096836,
      "num_tokens": 354671808.0,
      "step": 11113
    },
    {
      "entropy": 0.554726124741137,
      "epoch": 1.0225415610030995,
      "grad_norm": 0.14929072558879852,
      "learning_rate": 6.591836108810992e-05,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8341713026165962,
      "num_tokens": 354704395.0,
      "step": 11114
    },
    {
      "entropy": 0.6081827320158482,
      "epoch": 1.0226335673745406,
      "grad_norm": 0.15853779017925262,
      "learning_rate": 6.59152942619683e-05,
      "loss": 0.5915,
      "mean_token_accuracy": 0.8142186738550663,
      "num_tokens": 354736508.0,
      "step": 11115
    },
    {
      "entropy": 0.7144319303333759,
      "epoch": 1.022725573745982,
      "grad_norm": 0.15909773111343384,
      "learning_rate": 6.591222743582667e-05,
      "loss": 0.7067,
      "mean_token_accuracy": 0.7827513255178928,
      "num_tokens": 354767476.0,
      "step": 11116
    },
    {
      "entropy": 0.6858319491147995,
      "epoch": 1.022817580117423,
      "grad_norm": 0.16167880594730377,
      "learning_rate": 6.590916060968504e-05,
      "loss": 0.6763,
      "mean_token_accuracy": 0.7939940094947815,
      "num_tokens": 354800094.0,
      "step": 11117
    },
    {
      "entropy": 0.7416092287749052,
      "epoch": 1.0229095864888644,
      "grad_norm": 0.16189830005168915,
      "learning_rate": 6.59060937835434e-05,
      "loss": 0.758,
      "mean_token_accuracy": 0.7699565440416336,
      "num_tokens": 354831949.0,
      "step": 11118
    },
    {
      "entropy": 0.5553849888965487,
      "epoch": 1.0230015928603056,
      "grad_norm": 0.1494748294353485,
      "learning_rate": 6.590302695740179e-05,
      "loss": 0.5556,
      "mean_token_accuracy": 0.8281549476087093,
      "num_tokens": 354864360.0,
      "step": 11119
    },
    {
      "entropy": 0.5576512366533279,
      "epoch": 1.0230935992317467,
      "grad_norm": 0.15282440185546875,
      "learning_rate": 6.589996013126017e-05,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8285101875662804,
      "num_tokens": 354895345.0,
      "step": 11120
    },
    {
      "entropy": 0.639838669449091,
      "epoch": 1.023185605603188,
      "grad_norm": 0.16308560967445374,
      "learning_rate": 6.589689330511854e-05,
      "loss": 0.629,
      "mean_token_accuracy": 0.8073063679039478,
      "num_tokens": 354926781.0,
      "step": 11121
    },
    {
      "entropy": 0.5961843840777874,
      "epoch": 1.0232776119746292,
      "grad_norm": 0.15467730164527893,
      "learning_rate": 6.58938264789769e-05,
      "loss": 0.5703,
      "mean_token_accuracy": 0.820558562874794,
      "num_tokens": 354958189.0,
      "step": 11122
    },
    {
      "entropy": 0.6378604657948017,
      "epoch": 1.0233696183460705,
      "grad_norm": 0.16523084044456482,
      "learning_rate": 6.589075965283529e-05,
      "loss": 0.6168,
      "mean_token_accuracy": 0.8114181160926819,
      "num_tokens": 354990512.0,
      "step": 11123
    },
    {
      "entropy": 0.5497778244316578,
      "epoch": 1.0234616247175117,
      "grad_norm": 0.14505857229232788,
      "learning_rate": 6.588769282669365e-05,
      "loss": 0.5407,
      "mean_token_accuracy": 0.8324793092906475,
      "num_tokens": 355022828.0,
      "step": 11124
    },
    {
      "entropy": 0.6446080766618252,
      "epoch": 1.0235536310889528,
      "grad_norm": 0.16275165975093842,
      "learning_rate": 6.588462600055204e-05,
      "loss": 0.6512,
      "mean_token_accuracy": 0.8038684166967869,
      "num_tokens": 355054782.0,
      "step": 11125
    },
    {
      "entropy": 0.6478744447231293,
      "epoch": 1.0236456374603942,
      "grad_norm": 0.16452698409557343,
      "learning_rate": 6.58815591744104e-05,
      "loss": 0.6437,
      "mean_token_accuracy": 0.8083246238529682,
      "num_tokens": 355086282.0,
      "step": 11126
    },
    {
      "entropy": 0.7061208710074425,
      "epoch": 1.0237376438318353,
      "grad_norm": 0.16302618384361267,
      "learning_rate": 6.587849234826878e-05,
      "loss": 0.6976,
      "mean_token_accuracy": 0.7891222164034843,
      "num_tokens": 355118019.0,
      "step": 11127
    },
    {
      "entropy": 0.6330856643617153,
      "epoch": 1.0238296502032767,
      "grad_norm": 0.15950924158096313,
      "learning_rate": 6.587542552212715e-05,
      "loss": 0.6095,
      "mean_token_accuracy": 0.8113800063729286,
      "num_tokens": 355149298.0,
      "step": 11128
    },
    {
      "entropy": 0.5797736402601004,
      "epoch": 1.0239216565747178,
      "grad_norm": 0.1527262181043625,
      "learning_rate": 6.587235869598552e-05,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8283810466527939,
      "num_tokens": 355181761.0,
      "step": 11129
    },
    {
      "entropy": 0.6083549065515399,
      "epoch": 1.024013662946159,
      "grad_norm": 0.15909124910831451,
      "learning_rate": 6.58692918698439e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8239563554525375,
      "num_tokens": 355213396.0,
      "step": 11130
    },
    {
      "entropy": 0.6916986331343651,
      "epoch": 1.0241056693176003,
      "grad_norm": 0.16330231726169586,
      "learning_rate": 6.586622504370228e-05,
      "loss": 0.6693,
      "mean_token_accuracy": 0.7952448911964893,
      "num_tokens": 355245327.0,
      "step": 11131
    },
    {
      "entropy": 0.4228186681866646,
      "epoch": 1.0241976756890414,
      "grad_norm": 0.1542735993862152,
      "learning_rate": 6.586315821756065e-05,
      "loss": 0.397,
      "mean_token_accuracy": 0.8761871457099915,
      "num_tokens": 355277589.0,
      "step": 11132
    },
    {
      "entropy": 0.5805315556935966,
      "epoch": 1.0242896820604828,
      "grad_norm": 0.15418881177902222,
      "learning_rate": 6.586009139141902e-05,
      "loss": 0.5668,
      "mean_token_accuracy": 0.8258967883884907,
      "num_tokens": 355309576.0,
      "step": 11133
    },
    {
      "entropy": 0.5890538031235337,
      "epoch": 1.024381688431924,
      "grad_norm": 0.159847691655159,
      "learning_rate": 6.585702456527739e-05,
      "loss": 0.5903,
      "mean_token_accuracy": 0.8201438225805759,
      "num_tokens": 355341444.0,
      "step": 11134
    },
    {
      "entropy": 0.5169252487830818,
      "epoch": 1.024473694803365,
      "grad_norm": 0.1514231264591217,
      "learning_rate": 6.585395773913577e-05,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8441071957349777,
      "num_tokens": 355373290.0,
      "step": 11135
    },
    {
      "entropy": 0.5332802999764681,
      "epoch": 1.0245657011748064,
      "grad_norm": 0.1653662621974945,
      "learning_rate": 6.585089091299415e-05,
      "loss": 0.5388,
      "mean_token_accuracy": 0.8336604200303555,
      "num_tokens": 355404485.0,
      "step": 11136
    },
    {
      "entropy": 0.5673503521829844,
      "epoch": 1.0246577075462475,
      "grad_norm": 0.15026113390922546,
      "learning_rate": 6.584782408685252e-05,
      "loss": 0.555,
      "mean_token_accuracy": 0.8300462700426579,
      "num_tokens": 355436915.0,
      "step": 11137
    },
    {
      "entropy": 0.7351520285010338,
      "epoch": 1.0247497139176889,
      "grad_norm": 0.17579299211502075,
      "learning_rate": 6.58447572607109e-05,
      "loss": 0.7412,
      "mean_token_accuracy": 0.7737283930182457,
      "num_tokens": 355467525.0,
      "step": 11138
    },
    {
      "entropy": 0.5895990002900362,
      "epoch": 1.02484172028913,
      "grad_norm": 0.15308117866516113,
      "learning_rate": 6.584169043456927e-05,
      "loss": 0.5711,
      "mean_token_accuracy": 0.8244980722665787,
      "num_tokens": 355499095.0,
      "step": 11139
    },
    {
      "entropy": 0.6220831498503685,
      "epoch": 1.0249337266605711,
      "grad_norm": 0.15484720468521118,
      "learning_rate": 6.583862360842763e-05,
      "loss": 0.6053,
      "mean_token_accuracy": 0.814725685864687,
      "num_tokens": 355531297.0,
      "step": 11140
    },
    {
      "entropy": 0.626759547740221,
      "epoch": 1.0250257330320125,
      "grad_norm": 0.16055339574813843,
      "learning_rate": 6.583555678228602e-05,
      "loss": 0.6192,
      "mean_token_accuracy": 0.8114000484347343,
      "num_tokens": 355563566.0,
      "step": 11141
    },
    {
      "entropy": 0.6269311467185616,
      "epoch": 1.0251177394034536,
      "grad_norm": 0.15989360213279724,
      "learning_rate": 6.58324899561444e-05,
      "loss": 0.6078,
      "mean_token_accuracy": 0.8111033551394939,
      "num_tokens": 355594774.0,
      "step": 11142
    },
    {
      "entropy": 0.61144311260432,
      "epoch": 1.025209745774895,
      "grad_norm": 0.15761882066726685,
      "learning_rate": 6.582942313000277e-05,
      "loss": 0.5668,
      "mean_token_accuracy": 0.8230639398097992,
      "num_tokens": 355626205.0,
      "step": 11143
    },
    {
      "entropy": 0.5285051185637712,
      "epoch": 1.0253017521463361,
      "grad_norm": 0.15169763565063477,
      "learning_rate": 6.582635630386113e-05,
      "loss": 0.5204,
      "mean_token_accuracy": 0.8407036922872066,
      "num_tokens": 355658393.0,
      "step": 11144
    },
    {
      "entropy": 0.5779129816219211,
      "epoch": 1.0253937585177773,
      "grad_norm": 0.15228958427906036,
      "learning_rate": 6.58232894777195e-05,
      "loss": 0.5263,
      "mean_token_accuracy": 0.835735984146595,
      "num_tokens": 355689392.0,
      "step": 11145
    },
    {
      "entropy": 0.5383781120181084,
      "epoch": 1.0254857648892186,
      "grad_norm": 0.1535869836807251,
      "learning_rate": 6.58202226515779e-05,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8412112295627594,
      "num_tokens": 355721868.0,
      "step": 11146
    },
    {
      "entropy": 0.7336319303140044,
      "epoch": 1.0255777712606597,
      "grad_norm": 0.16432741284370422,
      "learning_rate": 6.581715582543626e-05,
      "loss": 0.7195,
      "mean_token_accuracy": 0.7799639701843262,
      "num_tokens": 355753401.0,
      "step": 11147
    },
    {
      "entropy": 0.5101771764457226,
      "epoch": 1.025669777632101,
      "grad_norm": 0.14977982640266418,
      "learning_rate": 6.581408899929463e-05,
      "loss": 0.4835,
      "mean_token_accuracy": 0.852300614118576,
      "num_tokens": 355785785.0,
      "step": 11148
    },
    {
      "entropy": 0.6322557460516691,
      "epoch": 1.0257617840035422,
      "grad_norm": 0.15779227018356323,
      "learning_rate": 6.5811022173153e-05,
      "loss": 0.62,
      "mean_token_accuracy": 0.806129727512598,
      "num_tokens": 355818003.0,
      "step": 11149
    },
    {
      "entropy": 0.6521858964115381,
      "epoch": 1.0258537903749834,
      "grad_norm": 0.15728923678398132,
      "learning_rate": 6.580795534701138e-05,
      "loss": 0.637,
      "mean_token_accuracy": 0.8029757365584373,
      "num_tokens": 355850281.0,
      "step": 11150
    },
    {
      "entropy": 0.6446063239127398,
      "epoch": 1.0259457967464247,
      "grad_norm": 0.15990093350410461,
      "learning_rate": 6.580488852086976e-05,
      "loss": 0.6356,
      "mean_token_accuracy": 0.8107896000146866,
      "num_tokens": 355882222.0,
      "step": 11151
    },
    {
      "entropy": 0.7422805540263653,
      "epoch": 1.0260378031178659,
      "grad_norm": 0.16674324870109558,
      "learning_rate": 6.580182169472813e-05,
      "loss": 0.7178,
      "mean_token_accuracy": 0.7809755206108093,
      "num_tokens": 355913017.0,
      "step": 11152
    },
    {
      "entropy": 0.6274510212242603,
      "epoch": 1.0261298094893072,
      "grad_norm": 0.1666877418756485,
      "learning_rate": 6.57987548685865e-05,
      "loss": 0.6,
      "mean_token_accuracy": 0.8150048740208149,
      "num_tokens": 355944821.0,
      "step": 11153
    },
    {
      "entropy": 0.6204209718853235,
      "epoch": 1.0262218158607483,
      "grad_norm": 0.15575242042541504,
      "learning_rate": 6.579568804244488e-05,
      "loss": 0.6016,
      "mean_token_accuracy": 0.8170488774776459,
      "num_tokens": 355976735.0,
      "step": 11154
    },
    {
      "entropy": 0.6272664051502943,
      "epoch": 1.0263138222321895,
      "grad_norm": 0.15776260197162628,
      "learning_rate": 6.579262121630325e-05,
      "loss": 0.6144,
      "mean_token_accuracy": 0.809675183147192,
      "num_tokens": 356009080.0,
      "step": 11155
    },
    {
      "entropy": 0.7152967974543571,
      "epoch": 1.0264058286036308,
      "grad_norm": 0.15648090839385986,
      "learning_rate": 6.578955439016163e-05,
      "loss": 0.7097,
      "mean_token_accuracy": 0.7846948616206646,
      "num_tokens": 356041241.0,
      "step": 11156
    },
    {
      "entropy": 0.607933247461915,
      "epoch": 1.026497834975072,
      "grad_norm": 0.1574198305606842,
      "learning_rate": 6.578648756402e-05,
      "loss": 0.5806,
      "mean_token_accuracy": 0.8226118944585323,
      "num_tokens": 356073140.0,
      "step": 11157
    },
    {
      "entropy": 0.5809630993753672,
      "epoch": 1.0265898413465133,
      "grad_norm": 0.15651533007621765,
      "learning_rate": 6.578342073787838e-05,
      "loss": 0.5683,
      "mean_token_accuracy": 0.822663675993681,
      "num_tokens": 356104886.0,
      "step": 11158
    },
    {
      "entropy": 0.5935042882338166,
      "epoch": 1.0266818477179545,
      "grad_norm": 0.1581466943025589,
      "learning_rate": 6.578035391173675e-05,
      "loss": 0.5926,
      "mean_token_accuracy": 0.8201813958585262,
      "num_tokens": 356136496.0,
      "step": 11159
    },
    {
      "entropy": 0.6058900514617562,
      "epoch": 1.0267738540893956,
      "grad_norm": 0.16271446645259857,
      "learning_rate": 6.577728708559511e-05,
      "loss": 0.6074,
      "mean_token_accuracy": 0.8187516145408154,
      "num_tokens": 356168684.0,
      "step": 11160
    },
    {
      "entropy": 0.605561375617981,
      "epoch": 1.026865860460837,
      "grad_norm": 0.1518544703722,
      "learning_rate": 6.57742202594535e-05,
      "loss": 0.6003,
      "mean_token_accuracy": 0.8132413737475872,
      "num_tokens": 356200527.0,
      "step": 11161
    },
    {
      "entropy": 0.5402805926278234,
      "epoch": 1.026957866832278,
      "grad_norm": 0.15425775945186615,
      "learning_rate": 6.577115343331188e-05,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8370953500270844,
      "num_tokens": 356232645.0,
      "step": 11162
    },
    {
      "entropy": 0.6417617425322533,
      "epoch": 1.0270498732037194,
      "grad_norm": 0.15694303810596466,
      "learning_rate": 6.576808660717025e-05,
      "loss": 0.6393,
      "mean_token_accuracy": 0.8047176338732243,
      "num_tokens": 356264013.0,
      "step": 11163
    },
    {
      "entropy": 0.4644618174061179,
      "epoch": 1.0271418795751606,
      "grad_norm": 0.14796100556850433,
      "learning_rate": 6.576501978102861e-05,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8599135056138039,
      "num_tokens": 356295799.0,
      "step": 11164
    },
    {
      "entropy": 0.645774464122951,
      "epoch": 1.0272338859466017,
      "grad_norm": 0.16234184801578522,
      "learning_rate": 6.5761952954887e-05,
      "loss": 0.6159,
      "mean_token_accuracy": 0.8134028017520905,
      "num_tokens": 356327594.0,
      "step": 11165
    },
    {
      "entropy": 0.6662279199808836,
      "epoch": 1.027325892318043,
      "grad_norm": 0.1563371866941452,
      "learning_rate": 6.575888612874536e-05,
      "loss": 0.6676,
      "mean_token_accuracy": 0.7951232865452766,
      "num_tokens": 356359917.0,
      "step": 11166
    },
    {
      "entropy": 0.6526262909173965,
      "epoch": 1.0274178986894842,
      "grad_norm": 0.1650128811597824,
      "learning_rate": 6.575581930260374e-05,
      "loss": 0.6237,
      "mean_token_accuracy": 0.807604618370533,
      "num_tokens": 356391474.0,
      "step": 11167
    },
    {
      "entropy": 0.7018301654607058,
      "epoch": 1.0275099050609255,
      "grad_norm": 0.15965121984481812,
      "learning_rate": 6.575275247646211e-05,
      "loss": 0.6862,
      "mean_token_accuracy": 0.7937578149139881,
      "num_tokens": 356422865.0,
      "step": 11168
    },
    {
      "entropy": 0.668730091303587,
      "epoch": 1.0276019114323667,
      "grad_norm": 0.15969325602054596,
      "learning_rate": 6.574968565032049e-05,
      "loss": 0.6634,
      "mean_token_accuracy": 0.8020192347466946,
      "num_tokens": 356454167.0,
      "step": 11169
    },
    {
      "entropy": 0.645359518006444,
      "epoch": 1.0276939178038078,
      "grad_norm": 0.1597966104745865,
      "learning_rate": 6.574661882417886e-05,
      "loss": 0.6255,
      "mean_token_accuracy": 0.8072432056069374,
      "num_tokens": 356486346.0,
      "step": 11170
    },
    {
      "entropy": 0.6369847320020199,
      "epoch": 1.0277859241752492,
      "grad_norm": 0.1603982150554657,
      "learning_rate": 6.574355199803723e-05,
      "loss": 0.6381,
      "mean_token_accuracy": 0.8047345019876957,
      "num_tokens": 356519077.0,
      "step": 11171
    },
    {
      "entropy": 0.7073330264538527,
      "epoch": 1.0278779305466903,
      "grad_norm": 0.16839192807674408,
      "learning_rate": 6.574048517189561e-05,
      "loss": 0.6954,
      "mean_token_accuracy": 0.7881055399775505,
      "num_tokens": 356551347.0,
      "step": 11172
    },
    {
      "entropy": 0.5894474070519209,
      "epoch": 1.0279699369181317,
      "grad_norm": 0.15522634983062744,
      "learning_rate": 6.573741834575399e-05,
      "loss": 0.5851,
      "mean_token_accuracy": 0.8214107230305672,
      "num_tokens": 356582550.0,
      "step": 11173
    },
    {
      "entropy": 0.5580227412283421,
      "epoch": 1.0280619432895728,
      "grad_norm": 0.15298543870449066,
      "learning_rate": 6.573435151961236e-05,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8335065133869648,
      "num_tokens": 356614481.0,
      "step": 11174
    },
    {
      "entropy": 0.6567193642258644,
      "epoch": 1.028153949661014,
      "grad_norm": 0.1749880462884903,
      "learning_rate": 6.573128469347073e-05,
      "loss": 0.671,
      "mean_token_accuracy": 0.7964727506041527,
      "num_tokens": 356646260.0,
      "step": 11175
    },
    {
      "entropy": 0.6559860613197088,
      "epoch": 1.0282459560324553,
      "grad_norm": 0.1611291915178299,
      "learning_rate": 6.57282178673291e-05,
      "loss": 0.6433,
      "mean_token_accuracy": 0.802121389657259,
      "num_tokens": 356677914.0,
      "step": 11176
    },
    {
      "entropy": 0.6459015142172575,
      "epoch": 1.0283379624038964,
      "grad_norm": 0.16649475693702698,
      "learning_rate": 6.572515104118748e-05,
      "loss": 0.6125,
      "mean_token_accuracy": 0.8158500678837299,
      "num_tokens": 356708940.0,
      "step": 11177
    },
    {
      "entropy": 0.5533565804362297,
      "epoch": 1.0284299687753378,
      "grad_norm": 0.1525716334581375,
      "learning_rate": 6.572208421504586e-05,
      "loss": 0.5449,
      "mean_token_accuracy": 0.830286517739296,
      "num_tokens": 356740612.0,
      "step": 11178
    },
    {
      "entropy": 0.6642203498631716,
      "epoch": 1.028521975146779,
      "grad_norm": 0.1557031124830246,
      "learning_rate": 6.571901738890423e-05,
      "loss": 0.6405,
      "mean_token_accuracy": 0.7998861521482468,
      "num_tokens": 356771821.0,
      "step": 11179
    },
    {
      "entropy": 0.5018505984917283,
      "epoch": 1.02861398151822,
      "grad_norm": 0.14705398678779602,
      "learning_rate": 6.57159505627626e-05,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8459547534584999,
      "num_tokens": 356803784.0,
      "step": 11180
    },
    {
      "entropy": 0.7198075894266367,
      "epoch": 1.0287059878896614,
      "grad_norm": 0.1608545184135437,
      "learning_rate": 6.571288373662098e-05,
      "loss": 0.6991,
      "mean_token_accuracy": 0.7895880490541458,
      "num_tokens": 356835893.0,
      "step": 11181
    },
    {
      "entropy": 0.5911889020353556,
      "epoch": 1.0287979942611025,
      "grad_norm": 0.15623465180397034,
      "learning_rate": 6.570981691047934e-05,
      "loss": 0.5583,
      "mean_token_accuracy": 0.8299335092306137,
      "num_tokens": 356867014.0,
      "step": 11182
    },
    {
      "entropy": 0.567258894443512,
      "epoch": 1.0288900006325439,
      "grad_norm": 0.15450255572795868,
      "learning_rate": 6.570675008433772e-05,
      "loss": 0.5487,
      "mean_token_accuracy": 0.8321421071887016,
      "num_tokens": 356898359.0,
      "step": 11183
    },
    {
      "entropy": 0.654614357277751,
      "epoch": 1.028982007003985,
      "grad_norm": 0.16758786141872406,
      "learning_rate": 6.570368325819609e-05,
      "loss": 0.6474,
      "mean_token_accuracy": 0.8009790740907192,
      "num_tokens": 356930251.0,
      "step": 11184
    },
    {
      "entropy": 0.621184716001153,
      "epoch": 1.0290740133754261,
      "grad_norm": 0.15916162729263306,
      "learning_rate": 6.570061643205447e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.8188925497233868,
      "num_tokens": 356961399.0,
      "step": 11185
    },
    {
      "entropy": 0.6081996322609484,
      "epoch": 1.0291660197468675,
      "grad_norm": 0.15683822333812714,
      "learning_rate": 6.569754960591284e-05,
      "loss": 0.5952,
      "mean_token_accuracy": 0.815809078514576,
      "num_tokens": 356992701.0,
      "step": 11186
    },
    {
      "entropy": 0.6773356478661299,
      "epoch": 1.0292580261183086,
      "grad_norm": 0.1669316440820694,
      "learning_rate": 6.569448277977121e-05,
      "loss": 0.6482,
      "mean_token_accuracy": 0.8018272966146469,
      "num_tokens": 357023993.0,
      "step": 11187
    },
    {
      "entropy": 0.5948080448433757,
      "epoch": 1.02935003248975,
      "grad_norm": 0.1556406319141388,
      "learning_rate": 6.569141595362959e-05,
      "loss": 0.5848,
      "mean_token_accuracy": 0.8235202133655548,
      "num_tokens": 357055663.0,
      "step": 11188
    },
    {
      "entropy": 0.6246946761384606,
      "epoch": 1.0294420388611911,
      "grad_norm": 0.16084134578704834,
      "learning_rate": 6.568834912748797e-05,
      "loss": 0.6281,
      "mean_token_accuracy": 0.8092124089598656,
      "num_tokens": 357087280.0,
      "step": 11189
    },
    {
      "entropy": 0.5400379505008459,
      "epoch": 1.0295340452326323,
      "grad_norm": 0.17175115644931793,
      "learning_rate": 6.568528230134634e-05,
      "loss": 0.5372,
      "mean_token_accuracy": 0.832184374332428,
      "num_tokens": 357118916.0,
      "step": 11190
    },
    {
      "entropy": 0.7275659963488579,
      "epoch": 1.0296260516040736,
      "grad_norm": 0.1684601753950119,
      "learning_rate": 6.568221547520471e-05,
      "loss": 0.727,
      "mean_token_accuracy": 0.7834472320973873,
      "num_tokens": 357151347.0,
      "step": 11191
    },
    {
      "entropy": 0.4981008544564247,
      "epoch": 1.0297180579755147,
      "grad_norm": 0.14512968063354492,
      "learning_rate": 6.567914864906308e-05,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8478180356323719,
      "num_tokens": 357183785.0,
      "step": 11192
    },
    {
      "entropy": 0.6259238012135029,
      "epoch": 1.029810064346956,
      "grad_norm": 0.15392276644706726,
      "learning_rate": 6.567608182292147e-05,
      "loss": 0.6181,
      "mean_token_accuracy": 0.8112550191581249,
      "num_tokens": 357215483.0,
      "step": 11193
    },
    {
      "entropy": 0.5913379192352295,
      "epoch": 1.0299020707183972,
      "grad_norm": 0.15633432567119598,
      "learning_rate": 6.567301499677984e-05,
      "loss": 0.5709,
      "mean_token_accuracy": 0.8223900608718395,
      "num_tokens": 357248140.0,
      "step": 11194
    },
    {
      "entropy": 0.6151008568704128,
      "epoch": 1.0299940770898384,
      "grad_norm": 0.1543024778366089,
      "learning_rate": 6.56699481706382e-05,
      "loss": 0.6061,
      "mean_token_accuracy": 0.8108697421848774,
      "num_tokens": 357279760.0,
      "step": 11195
    },
    {
      "entropy": 0.6578126419335604,
      "epoch": 1.0300860834612797,
      "grad_norm": 0.15859423577785492,
      "learning_rate": 6.566688134449659e-05,
      "loss": 0.633,
      "mean_token_accuracy": 0.8001300320029259,
      "num_tokens": 357311794.0,
      "step": 11196
    },
    {
      "entropy": 0.5486818440258503,
      "epoch": 1.0301780898327209,
      "grad_norm": 0.14729826152324677,
      "learning_rate": 6.566381451835496e-05,
      "loss": 0.5255,
      "mean_token_accuracy": 0.8373872339725494,
      "num_tokens": 357343452.0,
      "step": 11197
    },
    {
      "entropy": 0.5791084785014391,
      "epoch": 1.0302700962041622,
      "grad_norm": 0.15082857012748718,
      "learning_rate": 6.566074769221334e-05,
      "loss": 0.5642,
      "mean_token_accuracy": 0.824513491243124,
      "num_tokens": 357375466.0,
      "step": 11198
    },
    {
      "entropy": 0.7106267549097538,
      "epoch": 1.0303621025756033,
      "grad_norm": 0.16655319929122925,
      "learning_rate": 6.56576808660717e-05,
      "loss": 0.6922,
      "mean_token_accuracy": 0.7931222505867481,
      "num_tokens": 357407557.0,
      "step": 11199
    },
    {
      "entropy": 0.6002542097121477,
      "epoch": 1.0304541089470445,
      "grad_norm": 0.16402719914913177,
      "learning_rate": 6.565461403993009e-05,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8153147138655186,
      "num_tokens": 357439023.0,
      "step": 11200
    },
    {
      "entropy": 0.6549103818833828,
      "epoch": 1.0305461153184858,
      "grad_norm": 0.16113047301769257,
      "learning_rate": 6.565154721378845e-05,
      "loss": 0.6457,
      "mean_token_accuracy": 0.802578192204237,
      "num_tokens": 357470499.0,
      "step": 11201
    },
    {
      "entropy": 0.6089485604315996,
      "epoch": 1.030638121689927,
      "grad_norm": 0.15870225429534912,
      "learning_rate": 6.564848038764682e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8211034312844276,
      "num_tokens": 357502486.0,
      "step": 11202
    },
    {
      "entropy": 0.5897171534597874,
      "epoch": 1.0307301280613683,
      "grad_norm": 0.1540016382932663,
      "learning_rate": 6.56454135615052e-05,
      "loss": 0.583,
      "mean_token_accuracy": 0.8176778480410576,
      "num_tokens": 357534318.0,
      "step": 11203
    },
    {
      "entropy": 0.5221086088567972,
      "epoch": 1.0308221344328095,
      "grad_norm": 0.15392553806304932,
      "learning_rate": 6.564234673536359e-05,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8457939140498638,
      "num_tokens": 357566278.0,
      "step": 11204
    },
    {
      "entropy": 0.7097489228472114,
      "epoch": 1.0309141408042506,
      "grad_norm": 0.16394098103046417,
      "learning_rate": 6.563927990922195e-05,
      "loss": 0.6867,
      "mean_token_accuracy": 0.7933954671025276,
      "num_tokens": 357598294.0,
      "step": 11205
    },
    {
      "entropy": 0.64798878505826,
      "epoch": 1.031006147175692,
      "grad_norm": 0.1587889939546585,
      "learning_rate": 6.563621308308032e-05,
      "loss": 0.6393,
      "mean_token_accuracy": 0.8037106767296791,
      "num_tokens": 357630898.0,
      "step": 11206
    },
    {
      "entropy": 0.5797769781202078,
      "epoch": 1.031098153547133,
      "grad_norm": 0.15837378799915314,
      "learning_rate": 6.563314625693869e-05,
      "loss": 0.5766,
      "mean_token_accuracy": 0.8215299993753433,
      "num_tokens": 357662721.0,
      "step": 11207
    },
    {
      "entropy": 0.5368738728575408,
      "epoch": 1.0311901599185744,
      "grad_norm": 0.1563686579465866,
      "learning_rate": 6.563007943079707e-05,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8394439779222012,
      "num_tokens": 357694753.0,
      "step": 11208
    },
    {
      "entropy": 0.5386973861604929,
      "epoch": 1.0312821662900156,
      "grad_norm": 0.16020961105823517,
      "learning_rate": 6.562701260465545e-05,
      "loss": 0.5377,
      "mean_token_accuracy": 0.8332789540290833,
      "num_tokens": 357727002.0,
      "step": 11209
    },
    {
      "entropy": 0.6422846084460616,
      "epoch": 1.0313741726614567,
      "grad_norm": 0.16949813067913055,
      "learning_rate": 6.562394577851382e-05,
      "loss": 0.6338,
      "mean_token_accuracy": 0.8028657957911491,
      "num_tokens": 357757457.0,
      "step": 11210
    },
    {
      "entropy": 0.7769491225481033,
      "epoch": 1.031466179032898,
      "grad_norm": 0.16516825556755066,
      "learning_rate": 6.562087895237219e-05,
      "loss": 0.7646,
      "mean_token_accuracy": 0.7700598202645779,
      "num_tokens": 357789698.0,
      "step": 11211
    },
    {
      "entropy": 0.571486652828753,
      "epoch": 1.0315581854043392,
      "grad_norm": 0.1581244319677353,
      "learning_rate": 6.561781212623057e-05,
      "loss": 0.5504,
      "mean_token_accuracy": 0.8295733295381069,
      "num_tokens": 357822100.0,
      "step": 11212
    },
    {
      "entropy": 0.6360442079603672,
      "epoch": 1.0316501917757805,
      "grad_norm": 0.15768976509571075,
      "learning_rate": 6.561474530008894e-05,
      "loss": 0.6183,
      "mean_token_accuracy": 0.8086828477680683,
      "num_tokens": 357854459.0,
      "step": 11213
    },
    {
      "entropy": 0.6084075421094894,
      "epoch": 1.0317421981472217,
      "grad_norm": 0.15869195759296417,
      "learning_rate": 6.561167847394732e-05,
      "loss": 0.601,
      "mean_token_accuracy": 0.8134176656603813,
      "num_tokens": 357886180.0,
      "step": 11214
    },
    {
      "entropy": 0.6195746809244156,
      "epoch": 1.0318342045186628,
      "grad_norm": 0.15605174005031586,
      "learning_rate": 6.560861164780569e-05,
      "loss": 0.6161,
      "mean_token_accuracy": 0.8086886256933212,
      "num_tokens": 357917969.0,
      "step": 11215
    },
    {
      "entropy": 0.630449203774333,
      "epoch": 1.0319262108901042,
      "grad_norm": 0.15317851305007935,
      "learning_rate": 6.560554482166407e-05,
      "loss": 0.6109,
      "mean_token_accuracy": 0.811018094420433,
      "num_tokens": 357950075.0,
      "step": 11216
    },
    {
      "entropy": 0.6461335280910134,
      "epoch": 1.0320182172615453,
      "grad_norm": 0.153901606798172,
      "learning_rate": 6.560247799552244e-05,
      "loss": 0.6072,
      "mean_token_accuracy": 0.8142009600996971,
      "num_tokens": 357982398.0,
      "step": 11217
    },
    {
      "entropy": 0.5709450431168079,
      "epoch": 1.0321102236329867,
      "grad_norm": 0.15393422544002533,
      "learning_rate": 6.55994111693808e-05,
      "loss": 0.54,
      "mean_token_accuracy": 0.8329514563083649,
      "num_tokens": 358014437.0,
      "step": 11218
    },
    {
      "entropy": 0.5758693758398294,
      "epoch": 1.0322022300044278,
      "grad_norm": 0.15092776715755463,
      "learning_rate": 6.559634434323918e-05,
      "loss": 0.5688,
      "mean_token_accuracy": 0.8273100033402443,
      "num_tokens": 358046446.0,
      "step": 11219
    },
    {
      "entropy": 0.5527088111266494,
      "epoch": 1.032294236375869,
      "grad_norm": 0.164415642619133,
      "learning_rate": 6.559327751709757e-05,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8316296450793743,
      "num_tokens": 358078677.0,
      "step": 11220
    },
    {
      "entropy": 0.6317416243255138,
      "epoch": 1.0323862427473103,
      "grad_norm": 0.16037510335445404,
      "learning_rate": 6.559021069095593e-05,
      "loss": 0.6309,
      "mean_token_accuracy": 0.807031974196434,
      "num_tokens": 358110715.0,
      "step": 11221
    },
    {
      "entropy": 0.604914229363203,
      "epoch": 1.0324782491187514,
      "grad_norm": 0.16620273888111115,
      "learning_rate": 6.55871438648143e-05,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8212810680270195,
      "num_tokens": 358141371.0,
      "step": 11222
    },
    {
      "entropy": 0.6269135307520628,
      "epoch": 1.0325702554901928,
      "grad_norm": 0.1611940860748291,
      "learning_rate": 6.558407703867267e-05,
      "loss": 0.61,
      "mean_token_accuracy": 0.816243339329958,
      "num_tokens": 358173763.0,
      "step": 11223
    },
    {
      "entropy": 0.5860498826950788,
      "epoch": 1.032662261861634,
      "grad_norm": 0.15154896676540375,
      "learning_rate": 6.558101021253105e-05,
      "loss": 0.5868,
      "mean_token_accuracy": 0.8206036798655987,
      "num_tokens": 358206210.0,
      "step": 11224
    },
    {
      "entropy": 0.6190072223544121,
      "epoch": 1.032754268233075,
      "grad_norm": 0.15720811486244202,
      "learning_rate": 6.557794338638943e-05,
      "loss": 0.603,
      "mean_token_accuracy": 0.8138999864459038,
      "num_tokens": 358237507.0,
      "step": 11225
    },
    {
      "entropy": 0.6829803958535194,
      "epoch": 1.0328462746045164,
      "grad_norm": 0.16133752465248108,
      "learning_rate": 6.55748765602478e-05,
      "loss": 0.6793,
      "mean_token_accuracy": 0.7945583015680313,
      "num_tokens": 358268864.0,
      "step": 11226
    },
    {
      "entropy": 0.6116115637123585,
      "epoch": 1.0329382809759575,
      "grad_norm": 0.15127769112586975,
      "learning_rate": 6.557180973410618e-05,
      "loss": 0.6114,
      "mean_token_accuracy": 0.8105361126363277,
      "num_tokens": 358300669.0,
      "step": 11227
    },
    {
      "entropy": 0.6963477935642004,
      "epoch": 1.0330302873473989,
      "grad_norm": 0.16403399407863617,
      "learning_rate": 6.556874290796455e-05,
      "loss": 0.6848,
      "mean_token_accuracy": 0.788776483386755,
      "num_tokens": 358331782.0,
      "step": 11228
    },
    {
      "entropy": 0.6402001734822989,
      "epoch": 1.03312229371884,
      "grad_norm": 0.15336227416992188,
      "learning_rate": 6.556567608182292e-05,
      "loss": 0.6216,
      "mean_token_accuracy": 0.8093182891607285,
      "num_tokens": 358363215.0,
      "step": 11229
    },
    {
      "entropy": 0.5721560167148709,
      "epoch": 1.0332143000902811,
      "grad_norm": 0.15806904435157776,
      "learning_rate": 6.55626092556813e-05,
      "loss": 0.5662,
      "mean_token_accuracy": 0.8244564384222031,
      "num_tokens": 358395452.0,
      "step": 11230
    },
    {
      "entropy": 0.6152497567236423,
      "epoch": 1.0333063064617225,
      "grad_norm": 0.16084058582782745,
      "learning_rate": 6.555954242953968e-05,
      "loss": 0.5992,
      "mean_token_accuracy": 0.8163048177957535,
      "num_tokens": 358427134.0,
      "step": 11231
    },
    {
      "entropy": 0.6443797256797552,
      "epoch": 1.0333983128331636,
      "grad_norm": 0.15951676666736603,
      "learning_rate": 6.555647560339805e-05,
      "loss": 0.6258,
      "mean_token_accuracy": 0.8042616210877895,
      "num_tokens": 358459253.0,
      "step": 11232
    },
    {
      "entropy": 0.5831717289984226,
      "epoch": 1.033490319204605,
      "grad_norm": 0.17002196609973907,
      "learning_rate": 6.555340877725642e-05,
      "loss": 0.5623,
      "mean_token_accuracy": 0.828153844922781,
      "num_tokens": 358491017.0,
      "step": 11233
    },
    {
      "entropy": 0.5530284061096609,
      "epoch": 1.0335823255760461,
      "grad_norm": 0.1540166437625885,
      "learning_rate": 6.55503419511148e-05,
      "loss": 0.541,
      "mean_token_accuracy": 0.8326274491846561,
      "num_tokens": 358523006.0,
      "step": 11234
    },
    {
      "entropy": 0.5790864285081625,
      "epoch": 1.0336743319474873,
      "grad_norm": 0.1564413458108902,
      "learning_rate": 6.554727512497318e-05,
      "loss": 0.5606,
      "mean_token_accuracy": 0.8304401077330112,
      "num_tokens": 358555048.0,
      "step": 11235
    },
    {
      "entropy": 0.6096401680260897,
      "epoch": 1.0337663383189286,
      "grad_norm": 0.16089259088039398,
      "learning_rate": 6.554420829883155e-05,
      "loss": 0.5717,
      "mean_token_accuracy": 0.820878766477108,
      "num_tokens": 358586031.0,
      "step": 11236
    },
    {
      "entropy": 0.6338641350157559,
      "epoch": 1.0338583446903697,
      "grad_norm": 0.15699462592601776,
      "learning_rate": 6.554114147268991e-05,
      "loss": 0.609,
      "mean_token_accuracy": 0.8068219944834709,
      "num_tokens": 358616302.0,
      "step": 11237
    },
    {
      "entropy": 0.6513941399753094,
      "epoch": 1.033950351061811,
      "grad_norm": 0.15782621502876282,
      "learning_rate": 6.553807464654828e-05,
      "loss": 0.6242,
      "mean_token_accuracy": 0.8060525581240654,
      "num_tokens": 358647592.0,
      "step": 11238
    },
    {
      "entropy": 0.6388822924345732,
      "epoch": 1.0340423574332522,
      "grad_norm": 0.15884581208229065,
      "learning_rate": 6.553500782040666e-05,
      "loss": 0.6252,
      "mean_token_accuracy": 0.8081190139055252,
      "num_tokens": 358678706.0,
      "step": 11239
    },
    {
      "entropy": 0.5580443013459444,
      "epoch": 1.0341343638046934,
      "grad_norm": 0.1513216346502304,
      "learning_rate": 6.553194099426505e-05,
      "loss": 0.5403,
      "mean_token_accuracy": 0.8324749507009983,
      "num_tokens": 358710400.0,
      "step": 11240
    },
    {
      "entropy": 0.5876111406832933,
      "epoch": 1.0342263701761347,
      "grad_norm": 0.15666908025741577,
      "learning_rate": 6.552887416812341e-05,
      "loss": 0.567,
      "mean_token_accuracy": 0.8224202543497086,
      "num_tokens": 358741781.0,
      "step": 11241
    },
    {
      "entropy": 0.5780525393784046,
      "epoch": 1.0343183765475759,
      "grad_norm": 0.15570874512195587,
      "learning_rate": 6.552580734198178e-05,
      "loss": 0.569,
      "mean_token_accuracy": 0.8256644941866398,
      "num_tokens": 358773650.0,
      "step": 11242
    },
    {
      "entropy": 0.5698785241693258,
      "epoch": 1.0344103829190172,
      "grad_norm": 0.15447472035884857,
      "learning_rate": 6.552274051584016e-05,
      "loss": 0.5684,
      "mean_token_accuracy": 0.825314212590456,
      "num_tokens": 358806053.0,
      "step": 11243
    },
    {
      "entropy": 0.6346882805228233,
      "epoch": 1.0345023892904583,
      "grad_norm": 0.16997532546520233,
      "learning_rate": 6.551967368969853e-05,
      "loss": 0.6221,
      "mean_token_accuracy": 0.8105118460953236,
      "num_tokens": 358838508.0,
      "step": 11244
    },
    {
      "entropy": 0.70526616089046,
      "epoch": 1.0345943956618995,
      "grad_norm": 0.16605310142040253,
      "learning_rate": 6.551660686355691e-05,
      "loss": 0.6929,
      "mean_token_accuracy": 0.794137179851532,
      "num_tokens": 358869664.0,
      "step": 11245
    },
    {
      "entropy": 0.606017735786736,
      "epoch": 1.0346864020333408,
      "grad_norm": 0.16007481515407562,
      "learning_rate": 6.551354003741528e-05,
      "loss": 0.5849,
      "mean_token_accuracy": 0.8190787136554718,
      "num_tokens": 358901556.0,
      "step": 11246
    },
    {
      "entropy": 0.5502998121082783,
      "epoch": 1.034778408404782,
      "grad_norm": 0.15124063193798065,
      "learning_rate": 6.551047321127366e-05,
      "loss": 0.5417,
      "mean_token_accuracy": 0.8319225795567036,
      "num_tokens": 358933974.0,
      "step": 11247
    },
    {
      "entropy": 0.5398168936371803,
      "epoch": 1.0348704147762233,
      "grad_norm": 0.16363489627838135,
      "learning_rate": 6.550740638513203e-05,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8414734937250614,
      "num_tokens": 358965991.0,
      "step": 11248
    },
    {
      "entropy": 0.5966099416837096,
      "epoch": 1.0349624211476645,
      "grad_norm": 0.15708203613758087,
      "learning_rate": 6.55043395589904e-05,
      "loss": 0.5989,
      "mean_token_accuracy": 0.8157765045762062,
      "num_tokens": 358997330.0,
      "step": 11249
    },
    {
      "entropy": 0.5581850782036781,
      "epoch": 1.0350544275191056,
      "grad_norm": 0.1491868495941162,
      "learning_rate": 6.550127273284878e-05,
      "loss": 0.5493,
      "mean_token_accuracy": 0.8297613821923733,
      "num_tokens": 359029452.0,
      "step": 11250
    },
    {
      "entropy": 0.5660249562934041,
      "epoch": 1.035146433890547,
      "grad_norm": 0.15340019762516022,
      "learning_rate": 6.549820590670716e-05,
      "loss": 0.5654,
      "mean_token_accuracy": 0.8309675864875317,
      "num_tokens": 359061345.0,
      "step": 11251
    },
    {
      "entropy": 0.701406167820096,
      "epoch": 1.035238440261988,
      "grad_norm": 0.16169241070747375,
      "learning_rate": 6.549513908056553e-05,
      "loss": 0.6852,
      "mean_token_accuracy": 0.7912253178656101,
      "num_tokens": 359093734.0,
      "step": 11252
    },
    {
      "entropy": 0.6404104381799698,
      "epoch": 1.0353304466334294,
      "grad_norm": 0.15897880494594574,
      "learning_rate": 6.54920722544239e-05,
      "loss": 0.6011,
      "mean_token_accuracy": 0.8135656714439392,
      "num_tokens": 359125194.0,
      "step": 11253
    },
    {
      "entropy": 0.6517519950866699,
      "epoch": 1.0354224530048706,
      "grad_norm": 0.15694375336170197,
      "learning_rate": 6.548900542828226e-05,
      "loss": 0.6181,
      "mean_token_accuracy": 0.8080190718173981,
      "num_tokens": 359157613.0,
      "step": 11254
    },
    {
      "entropy": 0.6393206678330898,
      "epoch": 1.0355144593763117,
      "grad_norm": 0.1631609946489334,
      "learning_rate": 6.548593860214064e-05,
      "loss": 0.627,
      "mean_token_accuracy": 0.8089647404849529,
      "num_tokens": 359188725.0,
      "step": 11255
    },
    {
      "entropy": 0.6649221004918218,
      "epoch": 1.035606465747753,
      "grad_norm": 0.1545964777469635,
      "learning_rate": 6.548287177599903e-05,
      "loss": 0.6485,
      "mean_token_accuracy": 0.8058474212884903,
      "num_tokens": 359221188.0,
      "step": 11256
    },
    {
      "entropy": 0.5919221267104149,
      "epoch": 1.0356984721191942,
      "grad_norm": 0.16434337198734283,
      "learning_rate": 6.54798049498574e-05,
      "loss": 0.5749,
      "mean_token_accuracy": 0.8225748315453529,
      "num_tokens": 359253632.0,
      "step": 11257
    },
    {
      "entropy": 0.6446048617362976,
      "epoch": 1.0357904784906355,
      "grad_norm": 0.16433022916316986,
      "learning_rate": 6.547673812371578e-05,
      "loss": 0.6352,
      "mean_token_accuracy": 0.8067122548818588,
      "num_tokens": 359285562.0,
      "step": 11258
    },
    {
      "entropy": 0.4880861770361662,
      "epoch": 1.0358824848620767,
      "grad_norm": 0.15272651612758636,
      "learning_rate": 6.547367129757414e-05,
      "loss": 0.4717,
      "mean_token_accuracy": 0.8546700999140739,
      "num_tokens": 359317435.0,
      "step": 11259
    },
    {
      "entropy": 0.6602481165900826,
      "epoch": 1.0359744912335178,
      "grad_norm": 0.17034833133220673,
      "learning_rate": 6.547060447143251e-05,
      "loss": 0.6497,
      "mean_token_accuracy": 0.8063477501273155,
      "num_tokens": 359349608.0,
      "step": 11260
    },
    {
      "entropy": 0.6174950916320086,
      "epoch": 1.0360664976049592,
      "grad_norm": 0.15856313705444336,
      "learning_rate": 6.546753764529089e-05,
      "loss": 0.6065,
      "mean_token_accuracy": 0.8154224939644337,
      "num_tokens": 359381570.0,
      "step": 11261
    },
    {
      "entropy": 0.6575037855654955,
      "epoch": 1.0361585039764003,
      "grad_norm": 0.15837803483009338,
      "learning_rate": 6.546447081914927e-05,
      "loss": 0.6592,
      "mean_token_accuracy": 0.7976585887372494,
      "num_tokens": 359412831.0,
      "step": 11262
    },
    {
      "entropy": 0.6367569994181395,
      "epoch": 1.0362505103478417,
      "grad_norm": 0.15342766046524048,
      "learning_rate": 6.546140399300764e-05,
      "loss": 0.6256,
      "mean_token_accuracy": 0.8035937286913395,
      "num_tokens": 359444938.0,
      "step": 11263
    },
    {
      "entropy": 0.6461022458970547,
      "epoch": 1.0363425167192828,
      "grad_norm": 0.15990446507930756,
      "learning_rate": 6.545833716686601e-05,
      "loss": 0.6421,
      "mean_token_accuracy": 0.8040594421327114,
      "num_tokens": 359476673.0,
      "step": 11264
    },
    {
      "entropy": 0.5588886719197035,
      "epoch": 1.036434523090724,
      "grad_norm": 0.15579748153686523,
      "learning_rate": 6.545527034072438e-05,
      "loss": 0.551,
      "mean_token_accuracy": 0.8322730138897896,
      "num_tokens": 359508016.0,
      "step": 11265
    },
    {
      "entropy": 0.69586774520576,
      "epoch": 1.0365265294621653,
      "grad_norm": 0.16706958413124084,
      "learning_rate": 6.545220351458277e-05,
      "loss": 0.6695,
      "mean_token_accuracy": 0.7957456111907959,
      "num_tokens": 359540020.0,
      "step": 11266
    },
    {
      "entropy": 0.5514724077656865,
      "epoch": 1.0366185358336064,
      "grad_norm": 0.14992979168891907,
      "learning_rate": 6.544913668844114e-05,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8351389616727829,
      "num_tokens": 359572546.0,
      "step": 11267
    },
    {
      "entropy": 0.562500074505806,
      "epoch": 1.0367105422050478,
      "grad_norm": 0.15632300078868866,
      "learning_rate": 6.544606986229951e-05,
      "loss": 0.552,
      "mean_token_accuracy": 0.8297572359442711,
      "num_tokens": 359604578.0,
      "step": 11268
    },
    {
      "entropy": 0.6200545467436314,
      "epoch": 1.036802548576489,
      "grad_norm": 0.1573575884103775,
      "learning_rate": 6.544300303615788e-05,
      "loss": 0.6053,
      "mean_token_accuracy": 0.8134637139737606,
      "num_tokens": 359636581.0,
      "step": 11269
    },
    {
      "entropy": 0.6691756835207343,
      "epoch": 1.03689455494793,
      "grad_norm": 0.1584833413362503,
      "learning_rate": 6.543993621001626e-05,
      "loss": 0.6589,
      "mean_token_accuracy": 0.7942923083901405,
      "num_tokens": 359668434.0,
      "step": 11270
    },
    {
      "entropy": 0.690344556234777,
      "epoch": 1.0369865613193714,
      "grad_norm": 0.1697334498167038,
      "learning_rate": 6.543686938387464e-05,
      "loss": 0.6797,
      "mean_token_accuracy": 0.7929201610386372,
      "num_tokens": 359699635.0,
      "step": 11271
    },
    {
      "entropy": 0.5330392420291901,
      "epoch": 1.0370785676908125,
      "grad_norm": 0.15341712534427643,
      "learning_rate": 6.543380255773301e-05,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8450674377381802,
      "num_tokens": 359731307.0,
      "step": 11272
    },
    {
      "entropy": 0.5980792324990034,
      "epoch": 1.0371705740622539,
      "grad_norm": 0.15647125244140625,
      "learning_rate": 6.543073573159138e-05,
      "loss": 0.5847,
      "mean_token_accuracy": 0.8207773976027966,
      "num_tokens": 359762994.0,
      "step": 11273
    },
    {
      "entropy": 0.653015517629683,
      "epoch": 1.037262580433695,
      "grad_norm": 0.15488234162330627,
      "learning_rate": 6.542766890544976e-05,
      "loss": 0.6242,
      "mean_token_accuracy": 0.8086663410067558,
      "num_tokens": 359795515.0,
      "step": 11274
    },
    {
      "entropy": 0.6369661763310432,
      "epoch": 1.0373545868051361,
      "grad_norm": 0.15562483668327332,
      "learning_rate": 6.542460207930812e-05,
      "loss": 0.6216,
      "mean_token_accuracy": 0.8173789456486702,
      "num_tokens": 359827015.0,
      "step": 11275
    },
    {
      "entropy": 0.5964508280158043,
      "epoch": 1.0374465931765775,
      "grad_norm": 0.16231490671634674,
      "learning_rate": 6.54215352531665e-05,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8185994289815426,
      "num_tokens": 359858963.0,
      "step": 11276
    },
    {
      "entropy": 0.619480574503541,
      "epoch": 1.0375385995480186,
      "grad_norm": 0.15553130209445953,
      "learning_rate": 6.541846842702487e-05,
      "loss": 0.6081,
      "mean_token_accuracy": 0.8121733702719212,
      "num_tokens": 359891032.0,
      "step": 11277
    },
    {
      "entropy": 0.6400401666760445,
      "epoch": 1.03763060591946,
      "grad_norm": 0.16834978759288788,
      "learning_rate": 6.541540160088326e-05,
      "loss": 0.6386,
      "mean_token_accuracy": 0.8061617240309715,
      "num_tokens": 359922495.0,
      "step": 11278
    },
    {
      "entropy": 0.6658603632822633,
      "epoch": 1.0377226122909011,
      "grad_norm": 0.1533675193786621,
      "learning_rate": 6.541233477474162e-05,
      "loss": 0.6561,
      "mean_token_accuracy": 0.8024588413536549,
      "num_tokens": 359954806.0,
      "step": 11279
    },
    {
      "entropy": 0.6245888397097588,
      "epoch": 1.0378146186623423,
      "grad_norm": 0.15304048359394073,
      "learning_rate": 6.540926794859999e-05,
      "loss": 0.6084,
      "mean_token_accuracy": 0.814521424472332,
      "num_tokens": 359987008.0,
      "step": 11280
    },
    {
      "entropy": 0.5239458573050797,
      "epoch": 1.0379066250337836,
      "grad_norm": 0.14923307299613953,
      "learning_rate": 6.540620112245837e-05,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8460754230618477,
      "num_tokens": 360019338.0,
      "step": 11281
    },
    {
      "entropy": 0.6646715607494116,
      "epoch": 1.0379986314052247,
      "grad_norm": 0.16378210484981537,
      "learning_rate": 6.540313429631675e-05,
      "loss": 0.6419,
      "mean_token_accuracy": 0.8059790506958961,
      "num_tokens": 360051043.0,
      "step": 11282
    },
    {
      "entropy": 0.6406269576400518,
      "epoch": 1.038090637776666,
      "grad_norm": 0.15901625156402588,
      "learning_rate": 6.540006747017512e-05,
      "loss": 0.6262,
      "mean_token_accuracy": 0.8079768121242523,
      "num_tokens": 360082741.0,
      "step": 11283
    },
    {
      "entropy": 0.6018247604370117,
      "epoch": 1.0381826441481072,
      "grad_norm": 0.16391518712043762,
      "learning_rate": 6.539700064403349e-05,
      "loss": 0.5822,
      "mean_token_accuracy": 0.8171807266771793,
      "num_tokens": 360115152.0,
      "step": 11284
    },
    {
      "entropy": 0.6460073199123144,
      "epoch": 1.0382746505195484,
      "grad_norm": 0.162327840924263,
      "learning_rate": 6.539393381789186e-05,
      "loss": 0.6528,
      "mean_token_accuracy": 0.8023933693766594,
      "num_tokens": 360147139.0,
      "step": 11285
    },
    {
      "entropy": 0.688604798167944,
      "epoch": 1.0383666568909897,
      "grad_norm": 0.16144128143787384,
      "learning_rate": 6.539086699175024e-05,
      "loss": 0.6799,
      "mean_token_accuracy": 0.7882529571652412,
      "num_tokens": 360178442.0,
      "step": 11286
    },
    {
      "entropy": 0.6879567429423332,
      "epoch": 1.0384586632624309,
      "grad_norm": 0.17104949057102203,
      "learning_rate": 6.538780016560862e-05,
      "loss": 0.6813,
      "mean_token_accuracy": 0.7949122898280621,
      "num_tokens": 360210505.0,
      "step": 11287
    },
    {
      "entropy": 0.6244544330984354,
      "epoch": 1.0385506696338722,
      "grad_norm": 0.16270744800567627,
      "learning_rate": 6.538473333946699e-05,
      "loss": 0.6059,
      "mean_token_accuracy": 0.8130488395690918,
      "num_tokens": 360242382.0,
      "step": 11288
    },
    {
      "entropy": 0.5901247579604387,
      "epoch": 1.0386426760053133,
      "grad_norm": 0.15785330533981323,
      "learning_rate": 6.538166651332537e-05,
      "loss": 0.5798,
      "mean_token_accuracy": 0.8239158503711224,
      "num_tokens": 360274921.0,
      "step": 11289
    },
    {
      "entropy": 0.4621091363951564,
      "epoch": 1.0387346823767545,
      "grad_norm": 0.14036832749843597,
      "learning_rate": 6.537859968718374e-05,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8578538224101067,
      "num_tokens": 360307055.0,
      "step": 11290
    },
    {
      "entropy": 0.6233379542827606,
      "epoch": 1.0388266887481958,
      "grad_norm": 0.1605856865644455,
      "learning_rate": 6.53755328610421e-05,
      "loss": 0.6058,
      "mean_token_accuracy": 0.8112677708268166,
      "num_tokens": 360339254.0,
      "step": 11291
    },
    {
      "entropy": 0.688158331438899,
      "epoch": 1.038918695119637,
      "grad_norm": 0.1578766256570816,
      "learning_rate": 6.537246603490049e-05,
      "loss": 0.6851,
      "mean_token_accuracy": 0.7953712716698647,
      "num_tokens": 360371231.0,
      "step": 11292
    },
    {
      "entropy": 0.6501013059169054,
      "epoch": 1.0390107014910783,
      "grad_norm": 0.15800419449806213,
      "learning_rate": 6.536939920875887e-05,
      "loss": 0.6536,
      "mean_token_accuracy": 0.8028835691511631,
      "num_tokens": 360402856.0,
      "step": 11293
    },
    {
      "entropy": 0.6760437907651067,
      "epoch": 1.0391027078625195,
      "grad_norm": 0.17113277316093445,
      "learning_rate": 6.536633238261724e-05,
      "loss": 0.6519,
      "mean_token_accuracy": 0.8010628744959831,
      "num_tokens": 360434381.0,
      "step": 11294
    },
    {
      "entropy": 0.6072567403316498,
      "epoch": 1.0391947142339606,
      "grad_norm": 0.15749794244766235,
      "learning_rate": 6.53632655564756e-05,
      "loss": 0.5879,
      "mean_token_accuracy": 0.819013349711895,
      "num_tokens": 360466860.0,
      "step": 11295
    },
    {
      "entropy": 0.6136963907629251,
      "epoch": 1.039286720605402,
      "grad_norm": 0.15559901297092438,
      "learning_rate": 6.536019873033397e-05,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8124320469796658,
      "num_tokens": 360497822.0,
      "step": 11296
    },
    {
      "entropy": 0.6504846680909395,
      "epoch": 1.039378726976843,
      "grad_norm": 0.1632031798362732,
      "learning_rate": 6.535713190419235e-05,
      "loss": 0.6295,
      "mean_token_accuracy": 0.8070244565606117,
      "num_tokens": 360529743.0,
      "step": 11297
    },
    {
      "entropy": 0.6693718936294317,
      "epoch": 1.0394707333482844,
      "grad_norm": 0.16140252351760864,
      "learning_rate": 6.535406507805073e-05,
      "loss": 0.654,
      "mean_token_accuracy": 0.798920065164566,
      "num_tokens": 360561471.0,
      "step": 11298
    },
    {
      "entropy": 0.5929766409099102,
      "epoch": 1.0395627397197256,
      "grad_norm": 0.15122537314891815,
      "learning_rate": 6.53509982519091e-05,
      "loss": 0.5681,
      "mean_token_accuracy": 0.822842862457037,
      "num_tokens": 360592714.0,
      "step": 11299
    },
    {
      "entropy": 0.5161897800862789,
      "epoch": 1.0396547460911667,
      "grad_norm": 0.14950916171073914,
      "learning_rate": 6.534793142576747e-05,
      "loss": 0.4944,
      "mean_token_accuracy": 0.8455421663820744,
      "num_tokens": 360624510.0,
      "step": 11300
    },
    {
      "entropy": 0.685538413003087,
      "epoch": 1.039746752462608,
      "grad_norm": 0.16347506642341614,
      "learning_rate": 6.534486459962585e-05,
      "loss": 0.6835,
      "mean_token_accuracy": 0.7920996509492397,
      "num_tokens": 360656490.0,
      "step": 11301
    },
    {
      "entropy": 0.6164305992424488,
      "epoch": 1.0398387588340492,
      "grad_norm": 0.16771110892295837,
      "learning_rate": 6.534179777348422e-05,
      "loss": 0.6109,
      "mean_token_accuracy": 0.8160485401749611,
      "num_tokens": 360688251.0,
      "step": 11302
    },
    {
      "entropy": 0.6097148424014449,
      "epoch": 1.0399307652054905,
      "grad_norm": 0.1593731939792633,
      "learning_rate": 6.53387309473426e-05,
      "loss": 0.6116,
      "mean_token_accuracy": 0.8085715658962727,
      "num_tokens": 360719920.0,
      "step": 11303
    },
    {
      "entropy": 0.6904276125133038,
      "epoch": 1.0400227715769317,
      "grad_norm": 0.16248266398906708,
      "learning_rate": 6.533566412120097e-05,
      "loss": 0.684,
      "mean_token_accuracy": 0.7899565510451794,
      "num_tokens": 360751855.0,
      "step": 11304
    },
    {
      "entropy": 0.6903837993741035,
      "epoch": 1.0401147779483728,
      "grad_norm": 0.16958564519882202,
      "learning_rate": 6.533259729505935e-05,
      "loss": 0.6707,
      "mean_token_accuracy": 0.795820277184248,
      "num_tokens": 360780946.0,
      "step": 11305
    },
    {
      "entropy": 0.5451074438169599,
      "epoch": 1.0402067843198142,
      "grad_norm": 0.15899071097373962,
      "learning_rate": 6.532953046891772e-05,
      "loss": 0.5125,
      "mean_token_accuracy": 0.8433113135397434,
      "num_tokens": 360812477.0,
      "step": 11306
    },
    {
      "entropy": 0.5934247951954603,
      "epoch": 1.0402987906912553,
      "grad_norm": 0.16354793310165405,
      "learning_rate": 6.532646364277609e-05,
      "loss": 0.5752,
      "mean_token_accuracy": 0.822447132319212,
      "num_tokens": 360844302.0,
      "step": 11307
    },
    {
      "entropy": 0.621164882555604,
      "epoch": 1.0403907970626967,
      "grad_norm": 0.16061946749687195,
      "learning_rate": 6.532339681663447e-05,
      "loss": 0.5972,
      "mean_token_accuracy": 0.8134232349693775,
      "num_tokens": 360875259.0,
      "step": 11308
    },
    {
      "entropy": 0.619119611568749,
      "epoch": 1.0404828034341378,
      "grad_norm": 0.15735647082328796,
      "learning_rate": 6.532032999049285e-05,
      "loss": 0.5985,
      "mean_token_accuracy": 0.8188568539917469,
      "num_tokens": 360907636.0,
      "step": 11309
    },
    {
      "entropy": 0.6336795762181282,
      "epoch": 1.040574809805579,
      "grad_norm": 0.1538573056459427,
      "learning_rate": 6.531726316435122e-05,
      "loss": 0.6269,
      "mean_token_accuracy": 0.8041358664631844,
      "num_tokens": 360939708.0,
      "step": 11310
    },
    {
      "entropy": 0.5527670681476593,
      "epoch": 1.0406668161770203,
      "grad_norm": 0.15867435932159424,
      "learning_rate": 6.531419633820958e-05,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8394958190619946,
      "num_tokens": 360971524.0,
      "step": 11311
    },
    {
      "entropy": 0.6583319511264563,
      "epoch": 1.0407588225484614,
      "grad_norm": 0.1653902679681778,
      "learning_rate": 6.531112951206795e-05,
      "loss": 0.6392,
      "mean_token_accuracy": 0.8088385798037052,
      "num_tokens": 361003587.0,
      "step": 11312
    },
    {
      "entropy": 0.5506768226623535,
      "epoch": 1.0408508289199028,
      "grad_norm": 0.15013745427131653,
      "learning_rate": 6.530806268592635e-05,
      "loss": 0.5389,
      "mean_token_accuracy": 0.8332906141877174,
      "num_tokens": 361036059.0,
      "step": 11313
    },
    {
      "entropy": 0.6330753713846207,
      "epoch": 1.040942835291344,
      "grad_norm": 0.15820913016796112,
      "learning_rate": 6.530499585978472e-05,
      "loss": 0.6107,
      "mean_token_accuracy": 0.8124703541398048,
      "num_tokens": 361067457.0,
      "step": 11314
    },
    {
      "entropy": 0.6449433751404285,
      "epoch": 1.041034841662785,
      "grad_norm": 0.1595258265733719,
      "learning_rate": 6.530192903364308e-05,
      "loss": 0.6179,
      "mean_token_accuracy": 0.8071821182966232,
      "num_tokens": 361098839.0,
      "step": 11315
    },
    {
      "entropy": 0.6982866749167442,
      "epoch": 1.0411268480342264,
      "grad_norm": 0.16514776647090912,
      "learning_rate": 6.529886220750145e-05,
      "loss": 0.693,
      "mean_token_accuracy": 0.7937782146036625,
      "num_tokens": 361130824.0,
      "step": 11316
    },
    {
      "entropy": 0.5698676090687513,
      "epoch": 1.0412188544056675,
      "grad_norm": 0.15596292912960052,
      "learning_rate": 6.529579538135983e-05,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8325263448059559,
      "num_tokens": 361163127.0,
      "step": 11317
    },
    {
      "entropy": 0.625254325568676,
      "epoch": 1.0413108607771089,
      "grad_norm": 0.15437157452106476,
      "learning_rate": 6.529272855521821e-05,
      "loss": 0.6102,
      "mean_token_accuracy": 0.8098956756293774,
      "num_tokens": 361195119.0,
      "step": 11318
    },
    {
      "entropy": 0.6990798246115446,
      "epoch": 1.04140286714855,
      "grad_norm": 0.16002075374126434,
      "learning_rate": 6.528966172907658e-05,
      "loss": 0.6801,
      "mean_token_accuracy": 0.7919263541698456,
      "num_tokens": 361226919.0,
      "step": 11319
    },
    {
      "entropy": 0.6328364587388933,
      "epoch": 1.0414948735199911,
      "grad_norm": 0.1617008000612259,
      "learning_rate": 6.528659490293496e-05,
      "loss": 0.6248,
      "mean_token_accuracy": 0.8060440346598625,
      "num_tokens": 361258536.0,
      "step": 11320
    },
    {
      "entropy": 0.6231993236579001,
      "epoch": 1.0415868798914325,
      "grad_norm": 0.1587372124195099,
      "learning_rate": 6.528352807679333e-05,
      "loss": 0.6188,
      "mean_token_accuracy": 0.8130705654621124,
      "num_tokens": 361291055.0,
      "step": 11321
    },
    {
      "entropy": 0.6681632632389665,
      "epoch": 1.0416788862628736,
      "grad_norm": 0.15742380917072296,
      "learning_rate": 6.52804612506517e-05,
      "loss": 0.6682,
      "mean_token_accuracy": 0.799884207546711,
      "num_tokens": 361322840.0,
      "step": 11322
    },
    {
      "entropy": 0.6115213511511683,
      "epoch": 1.041770892634315,
      "grad_norm": 0.15813057124614716,
      "learning_rate": 6.527739442451008e-05,
      "loss": 0.6119,
      "mean_token_accuracy": 0.8124778531491756,
      "num_tokens": 361355206.0,
      "step": 11323
    },
    {
      "entropy": 0.5987418964505196,
      "epoch": 1.0418628990057561,
      "grad_norm": 0.1527065634727478,
      "learning_rate": 6.527432759836846e-05,
      "loss": 0.5738,
      "mean_token_accuracy": 0.8227873370051384,
      "num_tokens": 361387562.0,
      "step": 11324
    },
    {
      "entropy": 0.6084656808525324,
      "epoch": 1.0419549053771973,
      "grad_norm": 0.15500381588935852,
      "learning_rate": 6.527126077222683e-05,
      "loss": 0.5884,
      "mean_token_accuracy": 0.8183947093784809,
      "num_tokens": 361419880.0,
      "step": 11325
    },
    {
      "entropy": 0.5397191988304257,
      "epoch": 1.0420469117486386,
      "grad_norm": 0.15201863646507263,
      "learning_rate": 6.52681939460852e-05,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8353799507021904,
      "num_tokens": 361452254.0,
      "step": 11326
    },
    {
      "entropy": 0.5283767757937312,
      "epoch": 1.0421389181200797,
      "grad_norm": 0.1452038288116455,
      "learning_rate": 6.526512711994357e-05,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8404021486639977,
      "num_tokens": 361484408.0,
      "step": 11327
    },
    {
      "entropy": 0.614433865994215,
      "epoch": 1.042230924491521,
      "grad_norm": 0.15625466406345367,
      "learning_rate": 6.526206029380195e-05,
      "loss": 0.6024,
      "mean_token_accuracy": 0.8171016424894333,
      "num_tokens": 361515803.0,
      "step": 11328
    },
    {
      "entropy": 0.6728823930025101,
      "epoch": 1.0423229308629622,
      "grad_norm": 0.1606377512216568,
      "learning_rate": 6.525899346766033e-05,
      "loss": 0.669,
      "mean_token_accuracy": 0.7937152348458767,
      "num_tokens": 361547632.0,
      "step": 11329
    },
    {
      "entropy": 0.6121555352583528,
      "epoch": 1.0424149372344034,
      "grad_norm": 0.16202230751514435,
      "learning_rate": 6.52559266415187e-05,
      "loss": 0.5905,
      "mean_token_accuracy": 0.8182537108659744,
      "num_tokens": 361580131.0,
      "step": 11330
    },
    {
      "entropy": 0.6749308267608285,
      "epoch": 1.0425069436058447,
      "grad_norm": 0.1637076586484909,
      "learning_rate": 6.525285981537706e-05,
      "loss": 0.6715,
      "mean_token_accuracy": 0.79473427683115,
      "num_tokens": 361611929.0,
      "step": 11331
    },
    {
      "entropy": 0.7386414967477322,
      "epoch": 1.0425989499772859,
      "grad_norm": 0.1592947244644165,
      "learning_rate": 6.524979298923545e-05,
      "loss": 0.7352,
      "mean_token_accuracy": 0.7800780273973942,
      "num_tokens": 361644598.0,
      "step": 11332
    },
    {
      "entropy": 0.5869670892134309,
      "epoch": 1.0426909563487272,
      "grad_norm": 0.1562202423810959,
      "learning_rate": 6.524672616309381e-05,
      "loss": 0.577,
      "mean_token_accuracy": 0.8255320638418198,
      "num_tokens": 361676673.0,
      "step": 11333
    },
    {
      "entropy": 0.6389273125678301,
      "epoch": 1.0427829627201683,
      "grad_norm": 0.16067086160182953,
      "learning_rate": 6.52436593369522e-05,
      "loss": 0.6314,
      "mean_token_accuracy": 0.804293479770422,
      "num_tokens": 361708831.0,
      "step": 11334
    },
    {
      "entropy": 0.6098017580807209,
      "epoch": 1.0428749690916095,
      "grad_norm": 0.15540757775306702,
      "learning_rate": 6.524059251081056e-05,
      "loss": 0.6138,
      "mean_token_accuracy": 0.8104639053344727,
      "num_tokens": 361741131.0,
      "step": 11335
    },
    {
      "entropy": 0.635205615311861,
      "epoch": 1.0429669754630508,
      "grad_norm": 0.15789088606834412,
      "learning_rate": 6.523752568466894e-05,
      "loss": 0.6281,
      "mean_token_accuracy": 0.8059217110276222,
      "num_tokens": 361773254.0,
      "step": 11336
    },
    {
      "entropy": 0.6679215580224991,
      "epoch": 1.043058981834492,
      "grad_norm": 0.1597139835357666,
      "learning_rate": 6.523445885852731e-05,
      "loss": 0.6566,
      "mean_token_accuracy": 0.8011646382510662,
      "num_tokens": 361804183.0,
      "step": 11337
    },
    {
      "entropy": 0.4984524268656969,
      "epoch": 1.0431509882059333,
      "grad_norm": 0.1567150056362152,
      "learning_rate": 6.523139203238568e-05,
      "loss": 0.4729,
      "mean_token_accuracy": 0.8550131805241108,
      "num_tokens": 361836146.0,
      "step": 11338
    },
    {
      "entropy": 0.558585025370121,
      "epoch": 1.0432429945773745,
      "grad_norm": 0.16429337859153748,
      "learning_rate": 6.522832520624406e-05,
      "loss": 0.5525,
      "mean_token_accuracy": 0.828677162528038,
      "num_tokens": 361868150.0,
      "step": 11339
    },
    {
      "entropy": 0.6236891150474548,
      "epoch": 1.0433350009488156,
      "grad_norm": 0.15528947114944458,
      "learning_rate": 6.522525838010244e-05,
      "loss": 0.6038,
      "mean_token_accuracy": 0.8156319819390774,
      "num_tokens": 361900687.0,
      "step": 11340
    },
    {
      "entropy": 0.6857223361730576,
      "epoch": 1.043427007320257,
      "grad_norm": 0.16497768461704254,
      "learning_rate": 6.522219155396081e-05,
      "loss": 0.6589,
      "mean_token_accuracy": 0.7993718199431896,
      "num_tokens": 361931862.0,
      "step": 11341
    },
    {
      "entropy": 0.6730406787246466,
      "epoch": 1.043519013691698,
      "grad_norm": 0.16159197688102722,
      "learning_rate": 6.521912472781918e-05,
      "loss": 0.6489,
      "mean_token_accuracy": 0.8008205257356167,
      "num_tokens": 361963376.0,
      "step": 11342
    },
    {
      "entropy": 0.6018912214785814,
      "epoch": 1.0436110200631394,
      "grad_norm": 0.14345425367355347,
      "learning_rate": 6.521605790167755e-05,
      "loss": 0.5745,
      "mean_token_accuracy": 0.8200779370963573,
      "num_tokens": 361995695.0,
      "step": 11343
    },
    {
      "entropy": 0.6843041144311428,
      "epoch": 1.0437030264345806,
      "grad_norm": 0.15901626646518707,
      "learning_rate": 6.521299107553593e-05,
      "loss": 0.6538,
      "mean_token_accuracy": 0.800620011985302,
      "num_tokens": 362028388.0,
      "step": 11344
    },
    {
      "entropy": 0.7259683404117823,
      "epoch": 1.0437950328060217,
      "grad_norm": 0.16313642263412476,
      "learning_rate": 6.520992424939431e-05,
      "loss": 0.7008,
      "mean_token_accuracy": 0.7865181341767311,
      "num_tokens": 362059186.0,
      "step": 11345
    },
    {
      "entropy": 0.5867699924856424,
      "epoch": 1.043887039177463,
      "grad_norm": 0.14732445776462555,
      "learning_rate": 6.520685742325268e-05,
      "loss": 0.5743,
      "mean_token_accuracy": 0.820958212018013,
      "num_tokens": 362091150.0,
      "step": 11346
    },
    {
      "entropy": 0.5222950987517834,
      "epoch": 1.0439790455489042,
      "grad_norm": 0.15597349405288696,
      "learning_rate": 6.520379059711104e-05,
      "loss": 0.498,
      "mean_token_accuracy": 0.8413230404257774,
      "num_tokens": 362122687.0,
      "step": 11347
    },
    {
      "entropy": 0.5966097163036466,
      "epoch": 1.0440710519203455,
      "grad_norm": 0.15881267189979553,
      "learning_rate": 6.520072377096943e-05,
      "loss": 0.5828,
      "mean_token_accuracy": 0.8181601874530315,
      "num_tokens": 362155076.0,
      "step": 11348
    },
    {
      "entropy": 0.6202506273984909,
      "epoch": 1.0441630582917867,
      "grad_norm": 0.1615038514137268,
      "learning_rate": 6.51976569448278e-05,
      "loss": 0.6196,
      "mean_token_accuracy": 0.8113898076117039,
      "num_tokens": 362187177.0,
      "step": 11349
    },
    {
      "entropy": 0.6609824821352959,
      "epoch": 1.0442550646632278,
      "grad_norm": 0.1627250611782074,
      "learning_rate": 6.519459011868618e-05,
      "loss": 0.6555,
      "mean_token_accuracy": 0.7982389740645885,
      "num_tokens": 362218877.0,
      "step": 11350
    },
    {
      "entropy": 0.5744073176756501,
      "epoch": 1.0443470710346692,
      "grad_norm": 0.16115599870681763,
      "learning_rate": 6.519152329254456e-05,
      "loss": 0.5658,
      "mean_token_accuracy": 0.8283259235322475,
      "num_tokens": 362249974.0,
      "step": 11351
    },
    {
      "entropy": 0.664430333301425,
      "epoch": 1.0444390774061103,
      "grad_norm": 0.15930384397506714,
      "learning_rate": 6.518845646640292e-05,
      "loss": 0.6545,
      "mean_token_accuracy": 0.7999523878097534,
      "num_tokens": 362282651.0,
      "step": 11352
    },
    {
      "entropy": 0.5951716313138604,
      "epoch": 1.0445310837775517,
      "grad_norm": 0.15495075285434723,
      "learning_rate": 6.518538964026129e-05,
      "loss": 0.5834,
      "mean_token_accuracy": 0.8222058266401291,
      "num_tokens": 362315042.0,
      "step": 11353
    },
    {
      "entropy": 0.5670714639127254,
      "epoch": 1.0446230901489928,
      "grad_norm": 0.15432991087436676,
      "learning_rate": 6.518232281411967e-05,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8288856185972691,
      "num_tokens": 362347079.0,
      "step": 11354
    },
    {
      "entropy": 0.5420350264757872,
      "epoch": 1.044715096520434,
      "grad_norm": 0.15763238072395325,
      "learning_rate": 6.517925598797806e-05,
      "loss": 0.5306,
      "mean_token_accuracy": 0.832607489079237,
      "num_tokens": 362378829.0,
      "step": 11355
    },
    {
      "entropy": 0.7444331981241703,
      "epoch": 1.0448071028918753,
      "grad_norm": 0.16290809214115143,
      "learning_rate": 6.517618916183642e-05,
      "loss": 0.7268,
      "mean_token_accuracy": 0.7748517915606499,
      "num_tokens": 362410812.0,
      "step": 11356
    },
    {
      "entropy": 0.6367619577795267,
      "epoch": 1.0448991092633164,
      "grad_norm": 0.15845172107219696,
      "learning_rate": 6.517312233569479e-05,
      "loss": 0.6426,
      "mean_token_accuracy": 0.8022008091211319,
      "num_tokens": 362442926.0,
      "step": 11357
    },
    {
      "entropy": 0.6414101924747229,
      "epoch": 1.0449911156347578,
      "grad_norm": 0.16263112425804138,
      "learning_rate": 6.517005550955316e-05,
      "loss": 0.6286,
      "mean_token_accuracy": 0.8061984591186047,
      "num_tokens": 362474452.0,
      "step": 11358
    },
    {
      "entropy": 0.710134282708168,
      "epoch": 1.045083122006199,
      "grad_norm": 0.1584492176771164,
      "learning_rate": 6.516698868341154e-05,
      "loss": 0.6816,
      "mean_token_accuracy": 0.7893397472798824,
      "num_tokens": 362505779.0,
      "step": 11359
    },
    {
      "entropy": 0.6443398483097553,
      "epoch": 1.04517512837764,
      "grad_norm": 0.15380682051181793,
      "learning_rate": 6.516392185726992e-05,
      "loss": 0.6293,
      "mean_token_accuracy": 0.8028711937367916,
      "num_tokens": 362538264.0,
      "step": 11360
    },
    {
      "entropy": 0.7032506875693798,
      "epoch": 1.0452671347490814,
      "grad_norm": 0.16242197155952454,
      "learning_rate": 6.516085503112829e-05,
      "loss": 0.685,
      "mean_token_accuracy": 0.7926701866090298,
      "num_tokens": 362570176.0,
      "step": 11361
    },
    {
      "entropy": 0.7444228231906891,
      "epoch": 1.0453591411205225,
      "grad_norm": 0.1611459106206894,
      "learning_rate": 6.515778820498666e-05,
      "loss": 0.7216,
      "mean_token_accuracy": 0.7821303308010101,
      "num_tokens": 362601704.0,
      "step": 11362
    },
    {
      "entropy": 0.6525474116206169,
      "epoch": 1.0454511474919639,
      "grad_norm": 0.16241173446178436,
      "learning_rate": 6.515472137884504e-05,
      "loss": 0.638,
      "mean_token_accuracy": 0.8045470342040062,
      "num_tokens": 362633346.0,
      "step": 11363
    },
    {
      "entropy": 0.674727838486433,
      "epoch": 1.045543153863405,
      "grad_norm": 0.15949514508247375,
      "learning_rate": 6.515165455270341e-05,
      "loss": 0.6417,
      "mean_token_accuracy": 0.8001524098217487,
      "num_tokens": 362664704.0,
      "step": 11364
    },
    {
      "entropy": 0.6382793318480253,
      "epoch": 1.0456351602348462,
      "grad_norm": 0.1525222659111023,
      "learning_rate": 6.514858772656179e-05,
      "loss": 0.6254,
      "mean_token_accuracy": 0.8101888038218021,
      "num_tokens": 362697091.0,
      "step": 11365
    },
    {
      "entropy": 0.5885003674775362,
      "epoch": 1.0457271666062875,
      "grad_norm": 0.15260553359985352,
      "learning_rate": 6.514552090042016e-05,
      "loss": 0.5657,
      "mean_token_accuracy": 0.8247752003371716,
      "num_tokens": 362729310.0,
      "step": 11366
    },
    {
      "entropy": 0.5693422611802816,
      "epoch": 1.0458191729777286,
      "grad_norm": 0.16273964941501617,
      "learning_rate": 6.514245407427854e-05,
      "loss": 0.5588,
      "mean_token_accuracy": 0.8279097117483616,
      "num_tokens": 362761728.0,
      "step": 11367
    },
    {
      "entropy": 0.6291062645614147,
      "epoch": 1.04591117934917,
      "grad_norm": 0.16098758578300476,
      "learning_rate": 6.51393872481369e-05,
      "loss": 0.6321,
      "mean_token_accuracy": 0.8051988780498505,
      "num_tokens": 362793996.0,
      "step": 11368
    },
    {
      "entropy": 0.6343740336596966,
      "epoch": 1.0460031857206111,
      "grad_norm": 0.15987972915172577,
      "learning_rate": 6.513632042199527e-05,
      "loss": 0.6323,
      "mean_token_accuracy": 0.8104198686778545,
      "num_tokens": 362826372.0,
      "step": 11369
    },
    {
      "entropy": 0.652833603322506,
      "epoch": 1.0460951920920523,
      "grad_norm": 0.16730745136737823,
      "learning_rate": 6.513325359585365e-05,
      "loss": 0.6474,
      "mean_token_accuracy": 0.803126435726881,
      "num_tokens": 362858765.0,
      "step": 11370
    },
    {
      "entropy": 0.6223719641566277,
      "epoch": 1.0461871984634936,
      "grad_norm": 0.16728921234607697,
      "learning_rate": 6.513018676971204e-05,
      "loss": 0.6459,
      "mean_token_accuracy": 0.8078217208385468,
      "num_tokens": 362889181.0,
      "step": 11371
    },
    {
      "entropy": 0.6167031144723296,
      "epoch": 1.0462792048349348,
      "grad_norm": 0.15332530438899994,
      "learning_rate": 6.51271199435704e-05,
      "loss": 0.5923,
      "mean_token_accuracy": 0.8164460808038712,
      "num_tokens": 362920844.0,
      "step": 11372
    },
    {
      "entropy": 0.651158350519836,
      "epoch": 1.046371211206376,
      "grad_norm": 0.1573869287967682,
      "learning_rate": 6.512405311742877e-05,
      "loss": 0.6492,
      "mean_token_accuracy": 0.801630437374115,
      "num_tokens": 362953612.0,
      "step": 11373
    },
    {
      "entropy": 0.6780338976532221,
      "epoch": 1.0464632175778172,
      "grad_norm": 0.16182108223438263,
      "learning_rate": 6.512098629128714e-05,
      "loss": 0.6621,
      "mean_token_accuracy": 0.800847515463829,
      "num_tokens": 362986105.0,
      "step": 11374
    },
    {
      "entropy": 0.561512011103332,
      "epoch": 1.0465552239492584,
      "grad_norm": 0.15617334842681885,
      "learning_rate": 6.511791946514552e-05,
      "loss": 0.5632,
      "mean_token_accuracy": 0.8264914900064468,
      "num_tokens": 363017510.0,
      "step": 11375
    },
    {
      "entropy": 0.6154231652617455,
      "epoch": 1.0466472303206997,
      "grad_norm": 0.1526922881603241,
      "learning_rate": 6.51148526390039e-05,
      "loss": 0.5986,
      "mean_token_accuracy": 0.8132420852780342,
      "num_tokens": 363049881.0,
      "step": 11376
    },
    {
      "entropy": 0.5458917832002044,
      "epoch": 1.0467392366921409,
      "grad_norm": 0.1463434249162674,
      "learning_rate": 6.511178581286227e-05,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8412224054336548,
      "num_tokens": 363081988.0,
      "step": 11377
    },
    {
      "entropy": 0.534812000580132,
      "epoch": 1.0468312430635822,
      "grad_norm": 0.14850884675979614,
      "learning_rate": 6.510871898672064e-05,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8429783210158348,
      "num_tokens": 363114476.0,
      "step": 11378
    },
    {
      "entropy": 0.6627360340207815,
      "epoch": 1.0469232494350234,
      "grad_norm": 0.16060467064380646,
      "learning_rate": 6.510565216057902e-05,
      "loss": 0.6509,
      "mean_token_accuracy": 0.8042361736297607,
      "num_tokens": 363146937.0,
      "step": 11379
    },
    {
      "entropy": 0.6090762475505471,
      "epoch": 1.0470152558064645,
      "grad_norm": 0.16238924860954285,
      "learning_rate": 6.510258533443739e-05,
      "loss": 0.589,
      "mean_token_accuracy": 0.8206882327795029,
      "num_tokens": 363178988.0,
      "step": 11380
    },
    {
      "entropy": 0.6503971740603447,
      "epoch": 1.0471072621779058,
      "grad_norm": 0.1513805091381073,
      "learning_rate": 6.509951850829577e-05,
      "loss": 0.6182,
      "mean_token_accuracy": 0.8078080676496029,
      "num_tokens": 363211104.0,
      "step": 11381
    },
    {
      "entropy": 0.5487975468859076,
      "epoch": 1.047199268549347,
      "grad_norm": 0.1512773633003235,
      "learning_rate": 6.509645168215415e-05,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8322135470807552,
      "num_tokens": 363242291.0,
      "step": 11382
    },
    {
      "entropy": 0.579563157632947,
      "epoch": 1.0472912749207883,
      "grad_norm": 0.15228909254074097,
      "learning_rate": 6.509338485601252e-05,
      "loss": 0.5629,
      "mean_token_accuracy": 0.8264209479093552,
      "num_tokens": 363274152.0,
      "step": 11383
    },
    {
      "entropy": 0.6162634734064341,
      "epoch": 1.0473832812922295,
      "grad_norm": 0.14861229062080383,
      "learning_rate": 6.509031802987089e-05,
      "loss": 0.611,
      "mean_token_accuracy": 0.8126746490597725,
      "num_tokens": 363306817.0,
      "step": 11384
    },
    {
      "entropy": 0.6078766463324428,
      "epoch": 1.0474752876636706,
      "grad_norm": 0.1599850356578827,
      "learning_rate": 6.508725120372925e-05,
      "loss": 0.5947,
      "mean_token_accuracy": 0.8157589621841908,
      "num_tokens": 363339314.0,
      "step": 11385
    },
    {
      "entropy": 0.6052808333188295,
      "epoch": 1.047567294035112,
      "grad_norm": 0.15287961065769196,
      "learning_rate": 6.508418437758765e-05,
      "loss": 0.5901,
      "mean_token_accuracy": 0.8184751123189926,
      "num_tokens": 363371759.0,
      "step": 11386
    },
    {
      "entropy": 0.5558225614950061,
      "epoch": 1.047659300406553,
      "grad_norm": 0.1629764437675476,
      "learning_rate": 6.508111755144602e-05,
      "loss": 0.5609,
      "mean_token_accuracy": 0.8249412551522255,
      "num_tokens": 363402221.0,
      "step": 11387
    },
    {
      "entropy": 0.5525262001901865,
      "epoch": 1.0477513067779944,
      "grad_norm": 0.15481343865394592,
      "learning_rate": 6.507805072530438e-05,
      "loss": 0.5477,
      "mean_token_accuracy": 0.8304296247661114,
      "num_tokens": 363433731.0,
      "step": 11388
    },
    {
      "entropy": 0.5358664095401764,
      "epoch": 1.0478433131494356,
      "grad_norm": 0.15451189875602722,
      "learning_rate": 6.507498389916275e-05,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8407602943480015,
      "num_tokens": 363466165.0,
      "step": 11389
    },
    {
      "entropy": 0.5839177342131734,
      "epoch": 1.0479353195208767,
      "grad_norm": 0.15513983368873596,
      "learning_rate": 6.507191707302113e-05,
      "loss": 0.5796,
      "mean_token_accuracy": 0.8211032710969448,
      "num_tokens": 363498189.0,
      "step": 11390
    },
    {
      "entropy": 0.6270215706899762,
      "epoch": 1.048027325892318,
      "grad_norm": 0.15259040892124176,
      "learning_rate": 6.506885024687952e-05,
      "loss": 0.6222,
      "mean_token_accuracy": 0.8118256889283657,
      "num_tokens": 363529917.0,
      "step": 11391
    },
    {
      "entropy": 0.6200581416487694,
      "epoch": 1.0481193322637592,
      "grad_norm": 0.1616324782371521,
      "learning_rate": 6.506578342073788e-05,
      "loss": 0.6077,
      "mean_token_accuracy": 0.811031848192215,
      "num_tokens": 363561095.0,
      "step": 11392
    },
    {
      "entropy": 0.5321970013901591,
      "epoch": 1.0482113386352006,
      "grad_norm": 0.1482274979352951,
      "learning_rate": 6.506271659459625e-05,
      "loss": 0.5276,
      "mean_token_accuracy": 0.8340355828404427,
      "num_tokens": 363593459.0,
      "step": 11393
    },
    {
      "entropy": 0.6453676223754883,
      "epoch": 1.0483033450066417,
      "grad_norm": 0.16885577142238617,
      "learning_rate": 6.505964976845463e-05,
      "loss": 0.6489,
      "mean_token_accuracy": 0.8024999983608723,
      "num_tokens": 363624855.0,
      "step": 11394
    },
    {
      "entropy": 0.6548293363302946,
      "epoch": 1.0483953513780828,
      "grad_norm": 0.1533888727426529,
      "learning_rate": 6.5056582942313e-05,
      "loss": 0.6468,
      "mean_token_accuracy": 0.8014822714030743,
      "num_tokens": 363656498.0,
      "step": 11395
    },
    {
      "entropy": 0.6523930812254548,
      "epoch": 1.0484873577495242,
      "grad_norm": 0.17752349376678467,
      "learning_rate": 6.505351611617138e-05,
      "loss": 0.6574,
      "mean_token_accuracy": 0.7999953739345074,
      "num_tokens": 363688632.0,
      "step": 11396
    },
    {
      "entropy": 0.6131954109296203,
      "epoch": 1.0485793641209653,
      "grad_norm": 0.1550181359052658,
      "learning_rate": 6.505044929002975e-05,
      "loss": 0.6216,
      "mean_token_accuracy": 0.8130483664572239,
      "num_tokens": 363720558.0,
      "step": 11397
    },
    {
      "entropy": 0.5840537641197443,
      "epoch": 1.0486713704924067,
      "grad_norm": 0.1643660068511963,
      "learning_rate": 6.504738246388813e-05,
      "loss": 0.5664,
      "mean_token_accuracy": 0.8201086819171906,
      "num_tokens": 363752629.0,
      "step": 11398
    },
    {
      "entropy": 0.6238615792244673,
      "epoch": 1.0487633768638478,
      "grad_norm": 0.16380460560321808,
      "learning_rate": 6.50443156377465e-05,
      "loss": 0.5857,
      "mean_token_accuracy": 0.8153411746025085,
      "num_tokens": 363784473.0,
      "step": 11399
    },
    {
      "entropy": 0.5458680093288422,
      "epoch": 1.048855383235289,
      "grad_norm": 0.15095660090446472,
      "learning_rate": 6.504124881160487e-05,
      "loss": 0.5101,
      "mean_token_accuracy": 0.8369912542402744,
      "num_tokens": 363815448.0,
      "step": 11400
    },
    {
      "entropy": 0.6456878865137696,
      "epoch": 1.0489473896067303,
      "grad_norm": 0.1582343429327011,
      "learning_rate": 6.503818198546325e-05,
      "loss": 0.6186,
      "mean_token_accuracy": 0.81131911277771,
      "num_tokens": 363847809.0,
      "step": 11401
    },
    {
      "entropy": 0.7851273603737354,
      "epoch": 1.0490393959781714,
      "grad_norm": 0.15656065940856934,
      "learning_rate": 6.503511515932163e-05,
      "loss": 0.7698,
      "mean_token_accuracy": 0.7678992412984371,
      "num_tokens": 363880185.0,
      "step": 11402
    },
    {
      "entropy": 0.6963729951530695,
      "epoch": 1.0491314023496128,
      "grad_norm": 0.15824255347251892,
      "learning_rate": 6.503204833318e-05,
      "loss": 0.6689,
      "mean_token_accuracy": 0.7995501272380352,
      "num_tokens": 363911308.0,
      "step": 11403
    },
    {
      "entropy": 0.5890619242563844,
      "epoch": 1.049223408721054,
      "grad_norm": 0.1517166644334793,
      "learning_rate": 6.502898150703837e-05,
      "loss": 0.5709,
      "mean_token_accuracy": 0.8220900520682335,
      "num_tokens": 363943887.0,
      "step": 11404
    },
    {
      "entropy": 0.6881731264293194,
      "epoch": 1.0493154150924953,
      "grad_norm": 0.15675227344036102,
      "learning_rate": 6.502591468089673e-05,
      "loss": 0.674,
      "mean_token_accuracy": 0.7975470162928104,
      "num_tokens": 363975760.0,
      "step": 11405
    },
    {
      "entropy": 0.6471852473914623,
      "epoch": 1.0494074214639364,
      "grad_norm": 0.16173015534877777,
      "learning_rate": 6.502284785475512e-05,
      "loss": 0.6438,
      "mean_token_accuracy": 0.8027476146817207,
      "num_tokens": 364008410.0,
      "step": 11406
    },
    {
      "entropy": 0.6363593637943268,
      "epoch": 1.0494994278353775,
      "grad_norm": 0.14968813955783844,
      "learning_rate": 6.50197810286135e-05,
      "loss": 0.6204,
      "mean_token_accuracy": 0.8104485012590885,
      "num_tokens": 364040943.0,
      "step": 11407
    },
    {
      "entropy": 0.6776473093777895,
      "epoch": 1.0495914342068189,
      "grad_norm": 0.15755045413970947,
      "learning_rate": 6.501671420247186e-05,
      "loss": 0.6605,
      "mean_token_accuracy": 0.7987811975181103,
      "num_tokens": 364073062.0,
      "step": 11408
    },
    {
      "entropy": 0.5704047251492739,
      "epoch": 1.04968344057826,
      "grad_norm": 0.15131087601184845,
      "learning_rate": 6.501364737633025e-05,
      "loss": 0.5478,
      "mean_token_accuracy": 0.8290121592581272,
      "num_tokens": 364105009.0,
      "step": 11409
    },
    {
      "entropy": 0.538573021069169,
      "epoch": 1.0497754469497014,
      "grad_norm": 0.14987801015377045,
      "learning_rate": 6.501058055018861e-05,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8344681486487389,
      "num_tokens": 364136480.0,
      "step": 11410
    },
    {
      "entropy": 0.6767888758331537,
      "epoch": 1.0498674533211425,
      "grad_norm": 0.160080686211586,
      "learning_rate": 6.500751372404698e-05,
      "loss": 0.6603,
      "mean_token_accuracy": 0.7958000786602497,
      "num_tokens": 364168478.0,
      "step": 11411
    },
    {
      "entropy": 0.6714940965175629,
      "epoch": 1.0499594596925836,
      "grad_norm": 0.16177238523960114,
      "learning_rate": 6.500444689790536e-05,
      "loss": 0.6688,
      "mean_token_accuracy": 0.79579097032547,
      "num_tokens": 364200784.0,
      "step": 11412
    },
    {
      "entropy": 0.47658340679481626,
      "epoch": 1.050051466064025,
      "grad_norm": 0.1525873988866806,
      "learning_rate": 6.500138007176374e-05,
      "loss": 0.4714,
      "mean_token_accuracy": 0.8549815379083157,
      "num_tokens": 364233096.0,
      "step": 11413
    },
    {
      "entropy": 0.6362222153693438,
      "epoch": 1.0501434724354661,
      "grad_norm": 0.1554228514432907,
      "learning_rate": 6.499831324562211e-05,
      "loss": 0.6216,
      "mean_token_accuracy": 0.8082991279661655,
      "num_tokens": 364265077.0,
      "step": 11414
    },
    {
      "entropy": 0.6152723683044314,
      "epoch": 1.0502354788069075,
      "grad_norm": 0.16414757072925568,
      "learning_rate": 6.499524641948048e-05,
      "loss": 0.5918,
      "mean_token_accuracy": 0.8192642405629158,
      "num_tokens": 364297195.0,
      "step": 11415
    },
    {
      "entropy": 0.5170491999015212,
      "epoch": 1.0503274851783486,
      "grad_norm": 0.14998678863048553,
      "learning_rate": 6.499217959333885e-05,
      "loss": 0.5039,
      "mean_token_accuracy": 0.8393073379993439,
      "num_tokens": 364329451.0,
      "step": 11416
    },
    {
      "entropy": 0.5881760595366359,
      "epoch": 1.0504194915497898,
      "grad_norm": 0.15749900043010712,
      "learning_rate": 6.498911276719723e-05,
      "loss": 0.5719,
      "mean_token_accuracy": 0.8258674256503582,
      "num_tokens": 364361404.0,
      "step": 11417
    },
    {
      "entropy": 0.6628879606723785,
      "epoch": 1.050511497921231,
      "grad_norm": 0.1667446345090866,
      "learning_rate": 6.498604594105561e-05,
      "loss": 0.6403,
      "mean_token_accuracy": 0.8028820119798183,
      "num_tokens": 364393103.0,
      "step": 11418
    },
    {
      "entropy": 0.5620206650346518,
      "epoch": 1.0506035042926722,
      "grad_norm": 0.1549670249223709,
      "learning_rate": 6.498297911491398e-05,
      "loss": 0.538,
      "mean_token_accuracy": 0.8325112462043762,
      "num_tokens": 364425325.0,
      "step": 11419
    },
    {
      "entropy": 0.6222828179597855,
      "epoch": 1.0506955106641136,
      "grad_norm": 0.15272343158721924,
      "learning_rate": 6.497991228877235e-05,
      "loss": 0.6132,
      "mean_token_accuracy": 0.8119345381855965,
      "num_tokens": 364457391.0,
      "step": 11420
    },
    {
      "entropy": 0.670191453769803,
      "epoch": 1.0507875170355547,
      "grad_norm": 0.15741407871246338,
      "learning_rate": 6.497684546263073e-05,
      "loss": 0.6711,
      "mean_token_accuracy": 0.7968404665589333,
      "num_tokens": 364489238.0,
      "step": 11421
    },
    {
      "entropy": 0.6151786129921675,
      "epoch": 1.0508795234069959,
      "grad_norm": 0.1545073688030243,
      "learning_rate": 6.49737786364891e-05,
      "loss": 0.6065,
      "mean_token_accuracy": 0.8171534761786461,
      "num_tokens": 364521310.0,
      "step": 11422
    },
    {
      "entropy": 0.48143685702234507,
      "epoch": 1.0509715297784372,
      "grad_norm": 0.15412047505378723,
      "learning_rate": 6.497071181034748e-05,
      "loss": 0.4708,
      "mean_token_accuracy": 0.856441680341959,
      "num_tokens": 364553208.0,
      "step": 11423
    },
    {
      "entropy": 0.5518038030713797,
      "epoch": 1.0510635361498784,
      "grad_norm": 0.16590151190757751,
      "learning_rate": 6.496764498420585e-05,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8327039889991283,
      "num_tokens": 364585093.0,
      "step": 11424
    },
    {
      "entropy": 0.6453018076717854,
      "epoch": 1.0511555425213197,
      "grad_norm": 0.15579816699028015,
      "learning_rate": 6.496457815806423e-05,
      "loss": 0.62,
      "mean_token_accuracy": 0.8109638430178165,
      "num_tokens": 364617247.0,
      "step": 11425
    },
    {
      "entropy": 0.599403889849782,
      "epoch": 1.0512475488927608,
      "grad_norm": 0.16371679306030273,
      "learning_rate": 6.49615113319226e-05,
      "loss": 0.5949,
      "mean_token_accuracy": 0.8201431669294834,
      "num_tokens": 364648988.0,
      "step": 11426
    },
    {
      "entropy": 0.6985520031303167,
      "epoch": 1.051339555264202,
      "grad_norm": 0.16643120348453522,
      "learning_rate": 6.495844450578096e-05,
      "loss": 0.6841,
      "mean_token_accuracy": 0.7913945242762566,
      "num_tokens": 364680305.0,
      "step": 11427
    },
    {
      "entropy": 0.66095139272511,
      "epoch": 1.0514315616356433,
      "grad_norm": 0.15751783549785614,
      "learning_rate": 6.495537767963934e-05,
      "loss": 0.6525,
      "mean_token_accuracy": 0.8006252124905586,
      "num_tokens": 364712065.0,
      "step": 11428
    },
    {
      "entropy": 0.5553001649677753,
      "epoch": 1.0515235680070845,
      "grad_norm": 0.15427406132221222,
      "learning_rate": 6.495231085349773e-05,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8348609693348408,
      "num_tokens": 364743705.0,
      "step": 11429
    },
    {
      "entropy": 0.5332914674654603,
      "epoch": 1.0516155743785258,
      "grad_norm": 0.1523066759109497,
      "learning_rate": 6.494924402735609e-05,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8388533480465412,
      "num_tokens": 364776381.0,
      "step": 11430
    },
    {
      "entropy": 0.618366239592433,
      "epoch": 1.051707580749967,
      "grad_norm": 0.1564176082611084,
      "learning_rate": 6.494617720121446e-05,
      "loss": 0.6135,
      "mean_token_accuracy": 0.8139421977102757,
      "num_tokens": 364808335.0,
      "step": 11431
    },
    {
      "entropy": 0.6617803759872913,
      "epoch": 1.051799587121408,
      "grad_norm": 0.16230730712413788,
      "learning_rate": 6.494311037507283e-05,
      "loss": 0.645,
      "mean_token_accuracy": 0.8004652857780457,
      "num_tokens": 364840423.0,
      "step": 11432
    },
    {
      "entropy": 0.6150843249633908,
      "epoch": 1.0518915934928494,
      "grad_norm": 0.15335525572299957,
      "learning_rate": 6.494004354893122e-05,
      "loss": 0.5966,
      "mean_token_accuracy": 0.818917952477932,
      "num_tokens": 364872048.0,
      "step": 11433
    },
    {
      "entropy": 0.6092608934268355,
      "epoch": 1.0519835998642906,
      "grad_norm": 0.15284200012683868,
      "learning_rate": 6.493697672278959e-05,
      "loss": 0.5849,
      "mean_token_accuracy": 0.819420475512743,
      "num_tokens": 364904228.0,
      "step": 11434
    },
    {
      "entropy": 0.5778341609984636,
      "epoch": 1.052075606235732,
      "grad_norm": 0.16060051321983337,
      "learning_rate": 6.493390989664796e-05,
      "loss": 0.5625,
      "mean_token_accuracy": 0.8297654539346695,
      "num_tokens": 364935623.0,
      "step": 11435
    },
    {
      "entropy": 0.5792207634076476,
      "epoch": 1.052167612607173,
      "grad_norm": 0.1688866764307022,
      "learning_rate": 6.493084307050633e-05,
      "loss": 0.5635,
      "mean_token_accuracy": 0.8281609378755093,
      "num_tokens": 364967590.0,
      "step": 11436
    },
    {
      "entropy": 0.6499683316797018,
      "epoch": 1.0522596189786142,
      "grad_norm": 0.16252432763576508,
      "learning_rate": 6.492777624436471e-05,
      "loss": 0.6222,
      "mean_token_accuracy": 0.8106291890144348,
      "num_tokens": 364999876.0,
      "step": 11437
    },
    {
      "entropy": 0.5336073017679155,
      "epoch": 1.0523516253500556,
      "grad_norm": 0.15273095667362213,
      "learning_rate": 6.492470941822309e-05,
      "loss": 0.516,
      "mean_token_accuracy": 0.8439262621104717,
      "num_tokens": 365031580.0,
      "step": 11438
    },
    {
      "entropy": 0.6080591045320034,
      "epoch": 1.0524436317214967,
      "grad_norm": 0.18678493797779083,
      "learning_rate": 6.492164259208146e-05,
      "loss": 0.5816,
      "mean_token_accuracy": 0.820291556417942,
      "num_tokens": 365063629.0,
      "step": 11439
    },
    {
      "entropy": 0.6720459721982479,
      "epoch": 1.052535638092938,
      "grad_norm": 0.15711310505867004,
      "learning_rate": 6.491857576593984e-05,
      "loss": 0.6603,
      "mean_token_accuracy": 0.7955826036632061,
      "num_tokens": 365094763.0,
      "step": 11440
    },
    {
      "entropy": 0.6508864387869835,
      "epoch": 1.0526276444643792,
      "grad_norm": 0.16412372887134552,
      "learning_rate": 6.491550893979821e-05,
      "loss": 0.6383,
      "mean_token_accuracy": 0.8075405061244965,
      "num_tokens": 365126589.0,
      "step": 11441
    },
    {
      "entropy": 0.5311248833313584,
      "epoch": 1.0527196508358203,
      "grad_norm": 0.15657122433185577,
      "learning_rate": 6.491244211365658e-05,
      "loss": 0.5121,
      "mean_token_accuracy": 0.8408185541629791,
      "num_tokens": 365157666.0,
      "step": 11442
    },
    {
      "entropy": 0.6200449662283063,
      "epoch": 1.0528116572072617,
      "grad_norm": 0.15368010103702545,
      "learning_rate": 6.490937528751496e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8181966580450535,
      "num_tokens": 365189876.0,
      "step": 11443
    },
    {
      "entropy": 0.6649383716285229,
      "epoch": 1.0529036635787028,
      "grad_norm": 0.16044892370700836,
      "learning_rate": 6.490630846137334e-05,
      "loss": 0.6582,
      "mean_token_accuracy": 0.7969166412949562,
      "num_tokens": 365222098.0,
      "step": 11444
    },
    {
      "entropy": 0.5905695641413331,
      "epoch": 1.0529956699501442,
      "grad_norm": 0.15867389738559723,
      "learning_rate": 6.49032416352317e-05,
      "loss": 0.5749,
      "mean_token_accuracy": 0.8222676552832127,
      "num_tokens": 365254560.0,
      "step": 11445
    },
    {
      "entropy": 0.5995423160493374,
      "epoch": 1.0530876763215853,
      "grad_norm": 0.16280873119831085,
      "learning_rate": 6.490017480909007e-05,
      "loss": 0.567,
      "mean_token_accuracy": 0.8222132846713066,
      "num_tokens": 365286094.0,
      "step": 11446
    },
    {
      "entropy": 0.7026238515973091,
      "epoch": 1.0531796826930264,
      "grad_norm": 0.15886031091213226,
      "learning_rate": 6.489710798294844e-05,
      "loss": 0.6742,
      "mean_token_accuracy": 0.789745669811964,
      "num_tokens": 365317490.0,
      "step": 11447
    },
    {
      "entropy": 0.62220817245543,
      "epoch": 1.0532716890644678,
      "grad_norm": 0.17237821221351624,
      "learning_rate": 6.489404115680682e-05,
      "loss": 0.5958,
      "mean_token_accuracy": 0.815756231546402,
      "num_tokens": 365348776.0,
      "step": 11448
    },
    {
      "entropy": 0.603271808475256,
      "epoch": 1.053363695435909,
      "grad_norm": 0.1522413194179535,
      "learning_rate": 6.48909743306652e-05,
      "loss": 0.5843,
      "mean_token_accuracy": 0.8188003227114677,
      "num_tokens": 365380953.0,
      "step": 11449
    },
    {
      "entropy": 0.744919516146183,
      "epoch": 1.0534557018073503,
      "grad_norm": 0.16073289513587952,
      "learning_rate": 6.488790750452357e-05,
      "loss": 0.7415,
      "mean_token_accuracy": 0.7769869454205036,
      "num_tokens": 365413211.0,
      "step": 11450
    },
    {
      "entropy": 0.7036219676956534,
      "epoch": 1.0535477081787914,
      "grad_norm": 0.1667218655347824,
      "learning_rate": 6.488484067838194e-05,
      "loss": 0.6928,
      "mean_token_accuracy": 0.7896980419754982,
      "num_tokens": 365444622.0,
      "step": 11451
    },
    {
      "entropy": 0.6616778858006,
      "epoch": 1.0536397145502325,
      "grad_norm": 0.15957853198051453,
      "learning_rate": 6.488177385224032e-05,
      "loss": 0.6421,
      "mean_token_accuracy": 0.8012591674923897,
      "num_tokens": 365476332.0,
      "step": 11452
    },
    {
      "entropy": 0.7366012325510383,
      "epoch": 1.0537317209216739,
      "grad_norm": 0.16776402294635773,
      "learning_rate": 6.487870702609869e-05,
      "loss": 0.7209,
      "mean_token_accuracy": 0.784818485379219,
      "num_tokens": 365508678.0,
      "step": 11453
    },
    {
      "entropy": 0.6036098822951317,
      "epoch": 1.053823727293115,
      "grad_norm": 0.15691295266151428,
      "learning_rate": 6.487564019995707e-05,
      "loss": 0.5903,
      "mean_token_accuracy": 0.818221140652895,
      "num_tokens": 365540986.0,
      "step": 11454
    },
    {
      "entropy": 0.5597520302981138,
      "epoch": 1.0539157336645564,
      "grad_norm": 0.1555941104888916,
      "learning_rate": 6.487257337381544e-05,
      "loss": 0.5523,
      "mean_token_accuracy": 0.8299459777772427,
      "num_tokens": 365573357.0,
      "step": 11455
    },
    {
      "entropy": 0.7120181117206812,
      "epoch": 1.0540077400359975,
      "grad_norm": 0.1715201735496521,
      "learning_rate": 6.486950654767382e-05,
      "loss": 0.7061,
      "mean_token_accuracy": 0.7834617793560028,
      "num_tokens": 365604784.0,
      "step": 11456
    },
    {
      "entropy": 0.6079757381230593,
      "epoch": 1.0540997464074386,
      "grad_norm": 0.16301171481609344,
      "learning_rate": 6.486643972153219e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.821682084351778,
      "num_tokens": 365637345.0,
      "step": 11457
    },
    {
      "entropy": 0.6263370662927628,
      "epoch": 1.05419175277888,
      "grad_norm": 0.16043967008590698,
      "learning_rate": 6.486337289539056e-05,
      "loss": 0.6326,
      "mean_token_accuracy": 0.8075398653745651,
      "num_tokens": 365669251.0,
      "step": 11458
    },
    {
      "entropy": 0.696970833465457,
      "epoch": 1.0542837591503211,
      "grad_norm": 0.1733854115009308,
      "learning_rate": 6.486030606924894e-05,
      "loss": 0.6878,
      "mean_token_accuracy": 0.793230127543211,
      "num_tokens": 365700533.0,
      "step": 11459
    },
    {
      "entropy": 0.6202390226535499,
      "epoch": 1.0543757655217625,
      "grad_norm": 0.15885810554027557,
      "learning_rate": 6.485723924310732e-05,
      "loss": 0.6072,
      "mean_token_accuracy": 0.8134983032941818,
      "num_tokens": 365732259.0,
      "step": 11460
    },
    {
      "entropy": 0.6466219034045935,
      "epoch": 1.0544677718932036,
      "grad_norm": 0.16060087084770203,
      "learning_rate": 6.485417241696569e-05,
      "loss": 0.6261,
      "mean_token_accuracy": 0.8088178224861622,
      "num_tokens": 365764769.0,
      "step": 11461
    },
    {
      "entropy": 0.5861257575452328,
      "epoch": 1.0545597782646448,
      "grad_norm": 0.1702958047389984,
      "learning_rate": 6.485110559082405e-05,
      "loss": 0.5813,
      "mean_token_accuracy": 0.8228589184582233,
      "num_tokens": 365796961.0,
      "step": 11462
    },
    {
      "entropy": 0.592181982472539,
      "epoch": 1.054651784636086,
      "grad_norm": 0.15447190403938293,
      "learning_rate": 6.484803876468242e-05,
      "loss": 0.5678,
      "mean_token_accuracy": 0.8252751305699348,
      "num_tokens": 365828871.0,
      "step": 11463
    },
    {
      "entropy": 0.5563176143914461,
      "epoch": 1.0547437910075272,
      "grad_norm": 0.15176764130592346,
      "learning_rate": 6.48449719385408e-05,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8302204087376595,
      "num_tokens": 365861186.0,
      "step": 11464
    },
    {
      "entropy": 0.7178669571876526,
      "epoch": 1.0548357973789686,
      "grad_norm": 0.15300790965557098,
      "learning_rate": 6.484190511239919e-05,
      "loss": 0.6955,
      "mean_token_accuracy": 0.7860890477895737,
      "num_tokens": 365893808.0,
      "step": 11465
    },
    {
      "entropy": 0.538854718208313,
      "epoch": 1.0549278037504097,
      "grad_norm": 0.1524760127067566,
      "learning_rate": 6.483883828625755e-05,
      "loss": 0.5255,
      "mean_token_accuracy": 0.8355808183550835,
      "num_tokens": 365925625.0,
      "step": 11466
    },
    {
      "entropy": 0.5457146437838674,
      "epoch": 1.0550198101218509,
      "grad_norm": 0.1523897796869278,
      "learning_rate": 6.483577146011592e-05,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8391873240470886,
      "num_tokens": 365957959.0,
      "step": 11467
    },
    {
      "entropy": 0.6861556228250265,
      "epoch": 1.0551118164932922,
      "grad_norm": 0.1609906405210495,
      "learning_rate": 6.48327046339743e-05,
      "loss": 0.6871,
      "mean_token_accuracy": 0.7926273979246616,
      "num_tokens": 365990294.0,
      "step": 11468
    },
    {
      "entropy": 0.5232018399983644,
      "epoch": 1.0552038228647334,
      "grad_norm": 0.14434270560741425,
      "learning_rate": 6.482963780783267e-05,
      "loss": 0.5139,
      "mean_token_accuracy": 0.8402024358510971,
      "num_tokens": 366022784.0,
      "step": 11469
    },
    {
      "entropy": 0.6005763709545135,
      "epoch": 1.0552958292361747,
      "grad_norm": 0.15614847838878632,
      "learning_rate": 6.482657098169105e-05,
      "loss": 0.5917,
      "mean_token_accuracy": 0.8170373439788818,
      "num_tokens": 366055147.0,
      "step": 11470
    },
    {
      "entropy": 0.6486760023981333,
      "epoch": 1.0553878356076158,
      "grad_norm": 0.15694381296634674,
      "learning_rate": 6.482350415554943e-05,
      "loss": 0.6266,
      "mean_token_accuracy": 0.8116098307073116,
      "num_tokens": 366086875.0,
      "step": 11471
    },
    {
      "entropy": 0.617978673428297,
      "epoch": 1.055479841979057,
      "grad_norm": 0.15466374158859253,
      "learning_rate": 6.48204373294078e-05,
      "loss": 0.5993,
      "mean_token_accuracy": 0.8134466893970966,
      "num_tokens": 366118299.0,
      "step": 11472
    },
    {
      "entropy": 0.6101607289165258,
      "epoch": 1.0555718483504983,
      "grad_norm": 0.16175709664821625,
      "learning_rate": 6.481737050326617e-05,
      "loss": 0.597,
      "mean_token_accuracy": 0.8210935331881046,
      "num_tokens": 366150041.0,
      "step": 11473
    },
    {
      "entropy": 0.6031465716660023,
      "epoch": 1.0556638547219395,
      "grad_norm": 0.16499513387680054,
      "learning_rate": 6.481430367712455e-05,
      "loss": 0.5939,
      "mean_token_accuracy": 0.8155688717961311,
      "num_tokens": 366182210.0,
      "step": 11474
    },
    {
      "entropy": 0.5840566302649677,
      "epoch": 1.0557558610933808,
      "grad_norm": 0.15060190856456757,
      "learning_rate": 6.481123685098293e-05,
      "loss": 0.5679,
      "mean_token_accuracy": 0.8234964236617088,
      "num_tokens": 366214412.0,
      "step": 11475
    },
    {
      "entropy": 0.6640490079298615,
      "epoch": 1.055847867464822,
      "grad_norm": 0.16135062277317047,
      "learning_rate": 6.48081700248413e-05,
      "loss": 0.6414,
      "mean_token_accuracy": 0.8058769255876541,
      "num_tokens": 366246492.0,
      "step": 11476
    },
    {
      "entropy": 0.6991221755743027,
      "epoch": 1.055939873836263,
      "grad_norm": 0.165602907538414,
      "learning_rate": 6.480510319869967e-05,
      "loss": 0.6986,
      "mean_token_accuracy": 0.7885986492037773,
      "num_tokens": 366278350.0,
      "step": 11477
    },
    {
      "entropy": 0.5016881646588445,
      "epoch": 1.0560318802077044,
      "grad_norm": 0.14411412179470062,
      "learning_rate": 6.480203637255804e-05,
      "loss": 0.4753,
      "mean_token_accuracy": 0.848770622164011,
      "num_tokens": 366310489.0,
      "step": 11478
    },
    {
      "entropy": 0.5861703921109438,
      "epoch": 1.0561238865791456,
      "grad_norm": 0.1534552276134491,
      "learning_rate": 6.479896954641642e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.8210376724600792,
      "num_tokens": 366342427.0,
      "step": 11479
    },
    {
      "entropy": 0.6797659546136856,
      "epoch": 1.056215892950587,
      "grad_norm": 0.1593485176563263,
      "learning_rate": 6.47959027202748e-05,
      "loss": 0.6635,
      "mean_token_accuracy": 0.7960436344146729,
      "num_tokens": 366374871.0,
      "step": 11480
    },
    {
      "entropy": 0.6361502222716808,
      "epoch": 1.056307899322028,
      "grad_norm": 0.15999962389469147,
      "learning_rate": 6.479283589413317e-05,
      "loss": 0.625,
      "mean_token_accuracy": 0.808661513030529,
      "num_tokens": 366406572.0,
      "step": 11481
    },
    {
      "entropy": 0.6209884798154235,
      "epoch": 1.0563999056934692,
      "grad_norm": 0.15616589784622192,
      "learning_rate": 6.478976906799153e-05,
      "loss": 0.6061,
      "mean_token_accuracy": 0.8120869100093842,
      "num_tokens": 366438949.0,
      "step": 11482
    },
    {
      "entropy": 0.6283762296661735,
      "epoch": 1.0564919120649106,
      "grad_norm": 0.157831072807312,
      "learning_rate": 6.478670224184992e-05,
      "loss": 0.6113,
      "mean_token_accuracy": 0.8126362971961498,
      "num_tokens": 366470022.0,
      "step": 11483
    },
    {
      "entropy": 0.6398956999182701,
      "epoch": 1.0565839184363517,
      "grad_norm": 0.1520375907421112,
      "learning_rate": 6.478363541570828e-05,
      "loss": 0.614,
      "mean_token_accuracy": 0.8123543560504913,
      "num_tokens": 366502119.0,
      "step": 11484
    },
    {
      "entropy": 0.6121013532392681,
      "epoch": 1.056675924807793,
      "grad_norm": 0.15726704895496368,
      "learning_rate": 6.478056858956666e-05,
      "loss": 0.6074,
      "mean_token_accuracy": 0.810021635144949,
      "num_tokens": 366533564.0,
      "step": 11485
    },
    {
      "entropy": 0.639546986669302,
      "epoch": 1.0567679311792342,
      "grad_norm": 0.16478601098060608,
      "learning_rate": 6.477750176342503e-05,
      "loss": 0.6224,
      "mean_token_accuracy": 0.8084999397397041,
      "num_tokens": 366565626.0,
      "step": 11486
    },
    {
      "entropy": 0.5554078258574009,
      "epoch": 1.0568599375506753,
      "grad_norm": 0.1678774207830429,
      "learning_rate": 6.477443493728341e-05,
      "loss": 0.5486,
      "mean_token_accuracy": 0.8334452845156193,
      "num_tokens": 366598394.0,
      "step": 11487
    },
    {
      "entropy": 0.6281374003738165,
      "epoch": 1.0569519439221167,
      "grad_norm": 0.1551823616027832,
      "learning_rate": 6.477136811114178e-05,
      "loss": 0.6221,
      "mean_token_accuracy": 0.8076685853302479,
      "num_tokens": 366630410.0,
      "step": 11488
    },
    {
      "entropy": 0.6575850276276469,
      "epoch": 1.0570439502935578,
      "grad_norm": 0.16264264285564423,
      "learning_rate": 6.476830128500015e-05,
      "loss": 0.6367,
      "mean_token_accuracy": 0.8046320974826813,
      "num_tokens": 366661907.0,
      "step": 11489
    },
    {
      "entropy": 0.6085373666137457,
      "epoch": 1.0571359566649992,
      "grad_norm": 0.16177916526794434,
      "learning_rate": 6.476523445885853e-05,
      "loss": 0.5967,
      "mean_token_accuracy": 0.8159943632781506,
      "num_tokens": 366693539.0,
      "step": 11490
    },
    {
      "entropy": 0.6654332475736737,
      "epoch": 1.0572279630364403,
      "grad_norm": 0.16917279362678528,
      "learning_rate": 6.476216763271691e-05,
      "loss": 0.6498,
      "mean_token_accuracy": 0.8033435679972172,
      "num_tokens": 366725702.0,
      "step": 11491
    },
    {
      "entropy": 0.6620098669081926,
      "epoch": 1.0573199694078814,
      "grad_norm": 0.16638052463531494,
      "learning_rate": 6.475910080657528e-05,
      "loss": 0.663,
      "mean_token_accuracy": 0.7951421327888966,
      "num_tokens": 366757061.0,
      "step": 11492
    },
    {
      "entropy": 0.6006533335894346,
      "epoch": 1.0574119757793228,
      "grad_norm": 0.15395039319992065,
      "learning_rate": 6.475603398043365e-05,
      "loss": 0.5814,
      "mean_token_accuracy": 0.8244869560003281,
      "num_tokens": 366789202.0,
      "step": 11493
    },
    {
      "entropy": 0.566552366130054,
      "epoch": 1.057503982150764,
      "grad_norm": 0.15357200801372528,
      "learning_rate": 6.475296715429202e-05,
      "loss": 0.5426,
      "mean_token_accuracy": 0.8338248766958714,
      "num_tokens": 366821412.0,
      "step": 11494
    },
    {
      "entropy": 0.6269537284970284,
      "epoch": 1.0575959885222053,
      "grad_norm": 0.1563967913389206,
      "learning_rate": 6.47499003281504e-05,
      "loss": 0.626,
      "mean_token_accuracy": 0.8061735145747662,
      "num_tokens": 366853778.0,
      "step": 11495
    },
    {
      "entropy": 0.6168107502162457,
      "epoch": 1.0576879948936464,
      "grad_norm": 0.1582166701555252,
      "learning_rate": 6.474683350200878e-05,
      "loss": 0.6086,
      "mean_token_accuracy": 0.8098671138286591,
      "num_tokens": 366885912.0,
      "step": 11496
    },
    {
      "entropy": 0.6277752406895161,
      "epoch": 1.0577800012650875,
      "grad_norm": 0.1716056764125824,
      "learning_rate": 6.474376667586715e-05,
      "loss": 0.6149,
      "mean_token_accuracy": 0.8101401478052139,
      "num_tokens": 366918278.0,
      "step": 11497
    },
    {
      "entropy": 0.6201385837048292,
      "epoch": 1.0578720076365289,
      "grad_norm": 0.1522868424654007,
      "learning_rate": 6.474069984972551e-05,
      "loss": 0.6066,
      "mean_token_accuracy": 0.8133906610310078,
      "num_tokens": 366950558.0,
      "step": 11498
    },
    {
      "entropy": 0.6781242080032825,
      "epoch": 1.05796401400797,
      "grad_norm": 0.15657739341259003,
      "learning_rate": 6.47376330235839e-05,
      "loss": 0.6528,
      "mean_token_accuracy": 0.8008961118757725,
      "num_tokens": 366983252.0,
      "step": 11499
    },
    {
      "entropy": 0.5758312745019794,
      "epoch": 1.0580560203794114,
      "grad_norm": 0.15020142495632172,
      "learning_rate": 6.473456619744226e-05,
      "loss": 0.5635,
      "mean_token_accuracy": 0.8250300697982311,
      "num_tokens": 367015221.0,
      "step": 11500
    },
    {
      "entropy": 0.5810476038604975,
      "epoch": 1.0581480267508525,
      "grad_norm": 0.15708905458450317,
      "learning_rate": 6.473149937130065e-05,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8290615305304527,
      "num_tokens": 367047542.0,
      "step": 11501
    },
    {
      "entropy": 0.7193429972976446,
      "epoch": 1.0582400331222936,
      "grad_norm": 0.17409391701221466,
      "learning_rate": 6.472843254515903e-05,
      "loss": 0.7157,
      "mean_token_accuracy": 0.7840334326028824,
      "num_tokens": 367079043.0,
      "step": 11502
    },
    {
      "entropy": 0.6937735229730606,
      "epoch": 1.058332039493735,
      "grad_norm": 0.17210322618484497,
      "learning_rate": 6.47253657190174e-05,
      "loss": 0.7032,
      "mean_token_accuracy": 0.7882221229374409,
      "num_tokens": 367111121.0,
      "step": 11503
    },
    {
      "entropy": 0.6074908766895533,
      "epoch": 1.0584240458651761,
      "grad_norm": 0.1623196005821228,
      "learning_rate": 6.472229889287576e-05,
      "loss": 0.5926,
      "mean_token_accuracy": 0.8200472295284271,
      "num_tokens": 367143101.0,
      "step": 11504
    },
    {
      "entropy": 0.6606934145092964,
      "epoch": 1.0585160522366175,
      "grad_norm": 0.15984536707401276,
      "learning_rate": 6.471923206673413e-05,
      "loss": 0.647,
      "mean_token_accuracy": 0.8020777218043804,
      "num_tokens": 367174491.0,
      "step": 11505
    },
    {
      "entropy": 0.6960956836119294,
      "epoch": 1.0586080586080586,
      "grad_norm": 0.16457054018974304,
      "learning_rate": 6.471616524059253e-05,
      "loss": 0.6818,
      "mean_token_accuracy": 0.7928864322602749,
      "num_tokens": 367206931.0,
      "step": 11506
    },
    {
      "entropy": 0.5857688244432211,
      "epoch": 1.0587000649794998,
      "grad_norm": 0.15569522976875305,
      "learning_rate": 6.47130984144509e-05,
      "loss": 0.5803,
      "mean_token_accuracy": 0.8186010271310806,
      "num_tokens": 367238592.0,
      "step": 11507
    },
    {
      "entropy": 0.647826362401247,
      "epoch": 1.058792071350941,
      "grad_norm": 0.16065716743469238,
      "learning_rate": 6.471003158830926e-05,
      "loss": 0.6399,
      "mean_token_accuracy": 0.8006436787545681,
      "num_tokens": 367269341.0,
      "step": 11508
    },
    {
      "entropy": 0.5808493252843618,
      "epoch": 1.0588840777223822,
      "grad_norm": 0.172527015209198,
      "learning_rate": 6.470696476216763e-05,
      "loss": 0.5793,
      "mean_token_accuracy": 0.823421835899353,
      "num_tokens": 367300071.0,
      "step": 11509
    },
    {
      "entropy": 0.6119796521961689,
      "epoch": 1.0589760840938236,
      "grad_norm": 0.15544529259204865,
      "learning_rate": 6.470389793602601e-05,
      "loss": 0.5996,
      "mean_token_accuracy": 0.8123333007097244,
      "num_tokens": 367332618.0,
      "step": 11510
    },
    {
      "entropy": 0.5743270721286535,
      "epoch": 1.0590680904652647,
      "grad_norm": 0.16254781186580658,
      "learning_rate": 6.470083110988439e-05,
      "loss": 0.5646,
      "mean_token_accuracy": 0.8273755013942719,
      "num_tokens": 367365326.0,
      "step": 11511
    },
    {
      "entropy": 0.565783528611064,
      "epoch": 1.0591600968367059,
      "grad_norm": 0.1497086137533188,
      "learning_rate": 6.469776428374276e-05,
      "loss": 0.5606,
      "mean_token_accuracy": 0.8292074464261532,
      "num_tokens": 367397555.0,
      "step": 11512
    },
    {
      "entropy": 0.6008274229243398,
      "epoch": 1.0592521032081472,
      "grad_norm": 0.15726228058338165,
      "learning_rate": 6.469469745760113e-05,
      "loss": 0.5978,
      "mean_token_accuracy": 0.8188272677361965,
      "num_tokens": 367429956.0,
      "step": 11513
    },
    {
      "entropy": 0.6261375658214092,
      "epoch": 1.0593441095795884,
      "grad_norm": 0.15375672280788422,
      "learning_rate": 6.469163063145951e-05,
      "loss": 0.6151,
      "mean_token_accuracy": 0.8107660971581936,
      "num_tokens": 367462690.0,
      "step": 11514
    },
    {
      "entropy": 0.6685594618320465,
      "epoch": 1.0594361159510297,
      "grad_norm": 0.15579469501972198,
      "learning_rate": 6.468856380531788e-05,
      "loss": 0.6301,
      "mean_token_accuracy": 0.8053967989981174,
      "num_tokens": 367494380.0,
      "step": 11515
    },
    {
      "entropy": 0.5597274685278535,
      "epoch": 1.0595281223224708,
      "grad_norm": 0.1679033488035202,
      "learning_rate": 6.468549697917626e-05,
      "loss": 0.5402,
      "mean_token_accuracy": 0.8351445868611336,
      "num_tokens": 367526558.0,
      "step": 11516
    },
    {
      "entropy": 0.6005108635872602,
      "epoch": 1.059620128693912,
      "grad_norm": 0.14408685266971588,
      "learning_rate": 6.468243015303463e-05,
      "loss": 0.5716,
      "mean_token_accuracy": 0.8209397010505199,
      "num_tokens": 367559145.0,
      "step": 11517
    },
    {
      "entropy": 0.6345889009535313,
      "epoch": 1.0597121350653533,
      "grad_norm": 0.1554761379957199,
      "learning_rate": 6.467936332689301e-05,
      "loss": 0.6177,
      "mean_token_accuracy": 0.8124108463525772,
      "num_tokens": 367591398.0,
      "step": 11518
    },
    {
      "entropy": 0.592604398727417,
      "epoch": 1.0598041414367945,
      "grad_norm": 0.16029484570026398,
      "learning_rate": 6.467629650075138e-05,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8208790346980095,
      "num_tokens": 367622636.0,
      "step": 11519
    },
    {
      "entropy": 0.6592225264757872,
      "epoch": 1.0598961478082358,
      "grad_norm": 0.15441328287124634,
      "learning_rate": 6.467322967460974e-05,
      "loss": 0.6358,
      "mean_token_accuracy": 0.8039691224694252,
      "num_tokens": 367655182.0,
      "step": 11520
    },
    {
      "entropy": 0.6515316413715482,
      "epoch": 1.059988154179677,
      "grad_norm": 0.16234493255615234,
      "learning_rate": 6.467016284846812e-05,
      "loss": 0.6406,
      "mean_token_accuracy": 0.8038494065403938,
      "num_tokens": 367687580.0,
      "step": 11521
    },
    {
      "entropy": 0.5243578841909766,
      "epoch": 1.060080160551118,
      "grad_norm": 0.15379688143730164,
      "learning_rate": 6.46670960223265e-05,
      "loss": 0.5073,
      "mean_token_accuracy": 0.8426318764686584,
      "num_tokens": 367719253.0,
      "step": 11522
    },
    {
      "entropy": 0.6626626662909985,
      "epoch": 1.0601721669225594,
      "grad_norm": 0.16241474449634552,
      "learning_rate": 6.466402919618487e-05,
      "loss": 0.6613,
      "mean_token_accuracy": 0.7971420437097549,
      "num_tokens": 367751279.0,
      "step": 11523
    },
    {
      "entropy": 0.5880952775478363,
      "epoch": 1.0602641732940006,
      "grad_norm": 0.15458934009075165,
      "learning_rate": 6.466096237004324e-05,
      "loss": 0.5671,
      "mean_token_accuracy": 0.8219123519957066,
      "num_tokens": 367783003.0,
      "step": 11524
    },
    {
      "entropy": 0.5741651421412826,
      "epoch": 1.060356179665442,
      "grad_norm": 0.15259051322937012,
      "learning_rate": 6.465789554390161e-05,
      "loss": 0.5657,
      "mean_token_accuracy": 0.8259706199169159,
      "num_tokens": 367815339.0,
      "step": 11525
    },
    {
      "entropy": 0.6429525800049305,
      "epoch": 1.060448186036883,
      "grad_norm": 0.16050109267234802,
      "learning_rate": 6.465482871775999e-05,
      "loss": 0.6298,
      "mean_token_accuracy": 0.8096207231283188,
      "num_tokens": 367847687.0,
      "step": 11526
    },
    {
      "entropy": 0.5803962098434567,
      "epoch": 1.0605401924083242,
      "grad_norm": 0.1490607112646103,
      "learning_rate": 6.465176189161837e-05,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8200720474123955,
      "num_tokens": 367880455.0,
      "step": 11527
    },
    {
      "entropy": 0.7066553300246596,
      "epoch": 1.0606321987797656,
      "grad_norm": 0.16702444851398468,
      "learning_rate": 6.464869506547674e-05,
      "loss": 0.6905,
      "mean_token_accuracy": 0.7912511266767979,
      "num_tokens": 367912305.0,
      "step": 11528
    },
    {
      "entropy": 0.662000959739089,
      "epoch": 1.0607242051512067,
      "grad_norm": 0.1649908423423767,
      "learning_rate": 6.464562823933511e-05,
      "loss": 0.6654,
      "mean_token_accuracy": 0.8030708357691765,
      "num_tokens": 367944649.0,
      "step": 11529
    },
    {
      "entropy": 0.614757064729929,
      "epoch": 1.060816211522648,
      "grad_norm": 0.1681906282901764,
      "learning_rate": 6.464256141319349e-05,
      "loss": 0.6062,
      "mean_token_accuracy": 0.8122021295130253,
      "num_tokens": 367976473.0,
      "step": 11530
    },
    {
      "entropy": 0.709778256714344,
      "epoch": 1.0609082178940892,
      "grad_norm": 0.18513475358486176,
      "learning_rate": 6.463949458705186e-05,
      "loss": 0.7127,
      "mean_token_accuracy": 0.785390317440033,
      "num_tokens": 368007624.0,
      "step": 11531
    },
    {
      "entropy": 0.72544976323843,
      "epoch": 1.0610002242655303,
      "grad_norm": 0.1678742617368698,
      "learning_rate": 6.463642776091024e-05,
      "loss": 0.7133,
      "mean_token_accuracy": 0.7822572253644466,
      "num_tokens": 368039530.0,
      "step": 11532
    },
    {
      "entropy": 0.5854614544659853,
      "epoch": 1.0610922306369717,
      "grad_norm": 0.1498972326517105,
      "learning_rate": 6.463336093476862e-05,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8293047808110714,
      "num_tokens": 368071885.0,
      "step": 11533
    },
    {
      "entropy": 0.6775152441114187,
      "epoch": 1.0611842370084128,
      "grad_norm": 0.16292040050029755,
      "learning_rate": 6.463029410862699e-05,
      "loss": 0.6669,
      "mean_token_accuracy": 0.7977136448025703,
      "num_tokens": 368104294.0,
      "step": 11534
    },
    {
      "entropy": 0.6887227948755026,
      "epoch": 1.0612762433798542,
      "grad_norm": 0.16800954937934875,
      "learning_rate": 6.462722728248536e-05,
      "loss": 0.6714,
      "mean_token_accuracy": 0.7981694936752319,
      "num_tokens": 368136326.0,
      "step": 11535
    },
    {
      "entropy": 0.48423549998551607,
      "epoch": 1.0613682497512953,
      "grad_norm": 0.1507328301668167,
      "learning_rate": 6.462416045634372e-05,
      "loss": 0.4735,
      "mean_token_accuracy": 0.8536249957978725,
      "num_tokens": 368168384.0,
      "step": 11536
    },
    {
      "entropy": 0.682150486856699,
      "epoch": 1.0614602561227364,
      "grad_norm": 0.16000638902187347,
      "learning_rate": 6.46210936302021e-05,
      "loss": 0.6529,
      "mean_token_accuracy": 0.7986220642924309,
      "num_tokens": 368200479.0,
      "step": 11537
    },
    {
      "entropy": 0.5524875465780497,
      "epoch": 1.0615522624941778,
      "grad_norm": 0.1488371044397354,
      "learning_rate": 6.461802680406049e-05,
      "loss": 0.5316,
      "mean_token_accuracy": 0.8352935537695885,
      "num_tokens": 368232817.0,
      "step": 11538
    },
    {
      "entropy": 0.7003696374595165,
      "epoch": 1.061644268865619,
      "grad_norm": 0.1670401394367218,
      "learning_rate": 6.461495997791886e-05,
      "loss": 0.7014,
      "mean_token_accuracy": 0.7914176657795906,
      "num_tokens": 368264439.0,
      "step": 11539
    },
    {
      "entropy": 0.712976137176156,
      "epoch": 1.0617362752370603,
      "grad_norm": 0.16605478525161743,
      "learning_rate": 6.461189315177722e-05,
      "loss": 0.7033,
      "mean_token_accuracy": 0.7839868366718292,
      "num_tokens": 368296348.0,
      "step": 11540
    },
    {
      "entropy": 0.6857746206223965,
      "epoch": 1.0618282816085014,
      "grad_norm": 0.1612548828125,
      "learning_rate": 6.46088263256356e-05,
      "loss": 0.6716,
      "mean_token_accuracy": 0.7954084314405918,
      "num_tokens": 368327583.0,
      "step": 11541
    },
    {
      "entropy": 0.6686652880162001,
      "epoch": 1.0619202879799425,
      "grad_norm": 0.15874992311000824,
      "learning_rate": 6.460575949949397e-05,
      "loss": 0.6458,
      "mean_token_accuracy": 0.8022376485168934,
      "num_tokens": 368359666.0,
      "step": 11542
    },
    {
      "entropy": 0.6560273822396994,
      "epoch": 1.062012294351384,
      "grad_norm": 0.1555083841085434,
      "learning_rate": 6.460269267335235e-05,
      "loss": 0.6463,
      "mean_token_accuracy": 0.8036568462848663,
      "num_tokens": 368391916.0,
      "step": 11543
    },
    {
      "entropy": 0.621699258685112,
      "epoch": 1.062104300722825,
      "grad_norm": 0.1514313966035843,
      "learning_rate": 6.459962584721072e-05,
      "loss": 0.609,
      "mean_token_accuracy": 0.809114620089531,
      "num_tokens": 368424051.0,
      "step": 11544
    },
    {
      "entropy": 0.6723566427826881,
      "epoch": 1.0621963070942664,
      "grad_norm": 0.16564278304576874,
      "learning_rate": 6.45965590210691e-05,
      "loss": 0.6762,
      "mean_token_accuracy": 0.794536616653204,
      "num_tokens": 368456070.0,
      "step": 11545
    },
    {
      "entropy": 0.49571497552096844,
      "epoch": 1.0622883134657075,
      "grad_norm": 0.15372534096240997,
      "learning_rate": 6.459349219492747e-05,
      "loss": 0.4739,
      "mean_token_accuracy": 0.8527606278657913,
      "num_tokens": 368488501.0,
      "step": 11546
    },
    {
      "entropy": 0.5850958302617073,
      "epoch": 1.0623803198371486,
      "grad_norm": 0.1599045991897583,
      "learning_rate": 6.459042536878584e-05,
      "loss": 0.571,
      "mean_token_accuracy": 0.8237001411616802,
      "num_tokens": 368519721.0,
      "step": 11547
    },
    {
      "entropy": 0.6303936038166285,
      "epoch": 1.06247232620859,
      "grad_norm": 0.16546770930290222,
      "learning_rate": 6.458735854264422e-05,
      "loss": 0.6179,
      "mean_token_accuracy": 0.8090708926320076,
      "num_tokens": 368551608.0,
      "step": 11548
    },
    {
      "entropy": 0.5401257737539709,
      "epoch": 1.0625643325800311,
      "grad_norm": 0.14839521050453186,
      "learning_rate": 6.45842917165026e-05,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8353700935840607,
      "num_tokens": 368583341.0,
      "step": 11549
    },
    {
      "entropy": 0.6618144661188126,
      "epoch": 1.0626563389514725,
      "grad_norm": 0.16384179890155792,
      "learning_rate": 6.458122489036097e-05,
      "loss": 0.6481,
      "mean_token_accuracy": 0.804459985345602,
      "num_tokens": 368615699.0,
      "step": 11550
    },
    {
      "entropy": 0.5788997188210487,
      "epoch": 1.0627483453229136,
      "grad_norm": 0.17502906918525696,
      "learning_rate": 6.457815806421934e-05,
      "loss": 0.5706,
      "mean_token_accuracy": 0.8244873769581318,
      "num_tokens": 368647475.0,
      "step": 11551
    },
    {
      "entropy": 0.650476067326963,
      "epoch": 1.0628403516943548,
      "grad_norm": 0.15769656002521515,
      "learning_rate": 6.45750912380777e-05,
      "loss": 0.6432,
      "mean_token_accuracy": 0.8050087504088879,
      "num_tokens": 368679875.0,
      "step": 11552
    },
    {
      "entropy": 0.6292110364884138,
      "epoch": 1.0629323580657961,
      "grad_norm": 0.15632644295692444,
      "learning_rate": 6.45720244119361e-05,
      "loss": 0.6217,
      "mean_token_accuracy": 0.8057939074933529,
      "num_tokens": 368711479.0,
      "step": 11553
    },
    {
      "entropy": 0.6091015841811895,
      "epoch": 1.0630243644372372,
      "grad_norm": 0.16359750926494598,
      "learning_rate": 6.456895758579447e-05,
      "loss": 0.5859,
      "mean_token_accuracy": 0.8249770291149616,
      "num_tokens": 368743450.0,
      "step": 11554
    },
    {
      "entropy": 0.5672963233664632,
      "epoch": 1.0631163708086786,
      "grad_norm": 0.1573534607887268,
      "learning_rate": 6.456589075965284e-05,
      "loss": 0.5481,
      "mean_token_accuracy": 0.8354072757065296,
      "num_tokens": 368775766.0,
      "step": 11555
    },
    {
      "entropy": 0.6002369802445173,
      "epoch": 1.0632083771801197,
      "grad_norm": 0.1582612693309784,
      "learning_rate": 6.45628239335112e-05,
      "loss": 0.5879,
      "mean_token_accuracy": 0.8213973343372345,
      "num_tokens": 368807946.0,
      "step": 11556
    },
    {
      "entropy": 0.5902994852513075,
      "epoch": 1.0633003835515609,
      "grad_norm": 0.15616469085216522,
      "learning_rate": 6.455975710736959e-05,
      "loss": 0.5644,
      "mean_token_accuracy": 0.8290104493498802,
      "num_tokens": 368840276.0,
      "step": 11557
    },
    {
      "entropy": 0.5583338905125856,
      "epoch": 1.0633923899230022,
      "grad_norm": 0.14505352079868317,
      "learning_rate": 6.455669028122797e-05,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8319997787475586,
      "num_tokens": 368872348.0,
      "step": 11558
    },
    {
      "entropy": 0.5943026561290026,
      "epoch": 1.0634843962944434,
      "grad_norm": 0.16385997831821442,
      "learning_rate": 6.455362345508633e-05,
      "loss": 0.5974,
      "mean_token_accuracy": 0.8142881616950035,
      "num_tokens": 368903893.0,
      "step": 11559
    },
    {
      "entropy": 0.6429596859961748,
      "epoch": 1.0635764026658847,
      "grad_norm": 0.15957783162593842,
      "learning_rate": 6.45505566289447e-05,
      "loss": 0.6251,
      "mean_token_accuracy": 0.8071904890239239,
      "num_tokens": 368936016.0,
      "step": 11560
    },
    {
      "entropy": 0.6018205843865871,
      "epoch": 1.0636684090373258,
      "grad_norm": 0.15753695368766785,
      "learning_rate": 6.454748980280308e-05,
      "loss": 0.5725,
      "mean_token_accuracy": 0.8207607604563236,
      "num_tokens": 368967920.0,
      "step": 11561
    },
    {
      "entropy": 0.5891563557088375,
      "epoch": 1.063760415408767,
      "grad_norm": 0.16548871994018555,
      "learning_rate": 6.454442297666145e-05,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8260586075484753,
      "num_tokens": 369000312.0,
      "step": 11562
    },
    {
      "entropy": 0.6878458056598902,
      "epoch": 1.0638524217802083,
      "grad_norm": 0.1632123589515686,
      "learning_rate": 6.454135615051983e-05,
      "loss": 0.6915,
      "mean_token_accuracy": 0.7913385592401028,
      "num_tokens": 369031893.0,
      "step": 11563
    },
    {
      "entropy": 0.637447876855731,
      "epoch": 1.0639444281516495,
      "grad_norm": 0.15807922184467316,
      "learning_rate": 6.453828932437821e-05,
      "loss": 0.6322,
      "mean_token_accuracy": 0.8061257973313332,
      "num_tokens": 369064098.0,
      "step": 11564
    },
    {
      "entropy": 0.5456499997526407,
      "epoch": 1.0640364345230908,
      "grad_norm": 0.1534956842660904,
      "learning_rate": 6.453522249823658e-05,
      "loss": 0.5214,
      "mean_token_accuracy": 0.8427030295133591,
      "num_tokens": 369095869.0,
      "step": 11565
    },
    {
      "entropy": 0.615098100155592,
      "epoch": 1.064128440894532,
      "grad_norm": 0.15734048187732697,
      "learning_rate": 6.453215567209495e-05,
      "loss": 0.5752,
      "mean_token_accuracy": 0.8208525814116001,
      "num_tokens": 369127800.0,
      "step": 11566
    },
    {
      "entropy": 0.5516714630648494,
      "epoch": 1.064220447265973,
      "grad_norm": 0.15995261073112488,
      "learning_rate": 6.452908884595332e-05,
      "loss": 0.5278,
      "mean_token_accuracy": 0.8397361077368259,
      "num_tokens": 369160550.0,
      "step": 11567
    },
    {
      "entropy": 0.647506988607347,
      "epoch": 1.0643124536374144,
      "grad_norm": 0.16031694412231445,
      "learning_rate": 6.45260220198117e-05,
      "loss": 0.6372,
      "mean_token_accuracy": 0.8070603795349598,
      "num_tokens": 369192989.0,
      "step": 11568
    },
    {
      "entropy": 0.767314225435257,
      "epoch": 1.0644044600088556,
      "grad_norm": 0.17240594327449799,
      "learning_rate": 6.452295519367008e-05,
      "loss": 0.7847,
      "mean_token_accuracy": 0.7663186155259609,
      "num_tokens": 369225030.0,
      "step": 11569
    },
    {
      "entropy": 0.555538974236697,
      "epoch": 1.064496466380297,
      "grad_norm": 0.15742136538028717,
      "learning_rate": 6.451988836752845e-05,
      "loss": 0.5352,
      "mean_token_accuracy": 0.829916350543499,
      "num_tokens": 369257117.0,
      "step": 11570
    },
    {
      "entropy": 0.5898737031966448,
      "epoch": 1.064588472751738,
      "grad_norm": 0.15741810202598572,
      "learning_rate": 6.451682154138682e-05,
      "loss": 0.5583,
      "mean_token_accuracy": 0.831045351922512,
      "num_tokens": 369289261.0,
      "step": 11571
    },
    {
      "entropy": 0.6720075402408838,
      "epoch": 1.0646804791231792,
      "grad_norm": 0.15796403586864471,
      "learning_rate": 6.45137547152452e-05,
      "loss": 0.6707,
      "mean_token_accuracy": 0.7944017387926579,
      "num_tokens": 369321346.0,
      "step": 11572
    },
    {
      "entropy": 0.6900858022272587,
      "epoch": 1.0647724854946206,
      "grad_norm": 0.16090330481529236,
      "learning_rate": 6.451068788910357e-05,
      "loss": 0.6893,
      "mean_token_accuracy": 0.7871978394687176,
      "num_tokens": 369353311.0,
      "step": 11573
    },
    {
      "entropy": 0.5939967669546604,
      "epoch": 1.0648644918660617,
      "grad_norm": 0.15052542090415955,
      "learning_rate": 6.450762106296195e-05,
      "loss": 0.5765,
      "mean_token_accuracy": 0.8234154172241688,
      "num_tokens": 369385728.0,
      "step": 11574
    },
    {
      "entropy": 0.559375380165875,
      "epoch": 1.064956498237503,
      "grad_norm": 0.16718606650829315,
      "learning_rate": 6.450455423682032e-05,
      "loss": 0.5417,
      "mean_token_accuracy": 0.8295697569847107,
      "num_tokens": 369418305.0,
      "step": 11575
    },
    {
      "entropy": 0.6062845326960087,
      "epoch": 1.0650485046089442,
      "grad_norm": 0.1607314497232437,
      "learning_rate": 6.45014874106787e-05,
      "loss": 0.6106,
      "mean_token_accuracy": 0.8114840276539326,
      "num_tokens": 369450054.0,
      "step": 11576
    },
    {
      "entropy": 0.7128580845892429,
      "epoch": 1.0651405109803853,
      "grad_norm": 0.16722913086414337,
      "learning_rate": 6.449842058453706e-05,
      "loss": 0.7072,
      "mean_token_accuracy": 0.7807188369333744,
      "num_tokens": 369480895.0,
      "step": 11577
    },
    {
      "entropy": 0.5691962484270334,
      "epoch": 1.0652325173518267,
      "grad_norm": 0.15324120223522186,
      "learning_rate": 6.449535375839543e-05,
      "loss": 0.5566,
      "mean_token_accuracy": 0.8271311186254025,
      "num_tokens": 369513257.0,
      "step": 11578
    },
    {
      "entropy": 0.5664912760257721,
      "epoch": 1.0653245237232678,
      "grad_norm": 0.1563396155834198,
      "learning_rate": 6.449228693225381e-05,
      "loss": 0.5581,
      "mean_token_accuracy": 0.825756810605526,
      "num_tokens": 369544801.0,
      "step": 11579
    },
    {
      "entropy": 0.5900455117225647,
      "epoch": 1.0654165300947092,
      "grad_norm": 0.15260006487369537,
      "learning_rate": 6.44892201061122e-05,
      "loss": 0.5903,
      "mean_token_accuracy": 0.8226679153740406,
      "num_tokens": 369576719.0,
      "step": 11580
    },
    {
      "entropy": 0.6430571107193828,
      "epoch": 1.0655085364661503,
      "grad_norm": 0.1590586155653,
      "learning_rate": 6.448615327997056e-05,
      "loss": 0.6126,
      "mean_token_accuracy": 0.8135054856538773,
      "num_tokens": 369608531.0,
      "step": 11581
    },
    {
      "entropy": 0.6649076258763671,
      "epoch": 1.0656005428375914,
      "grad_norm": 0.15925273299217224,
      "learning_rate": 6.448308645382893e-05,
      "loss": 0.6595,
      "mean_token_accuracy": 0.7977683991193771,
      "num_tokens": 369640678.0,
      "step": 11582
    },
    {
      "entropy": 0.6090736901387572,
      "epoch": 1.0656925492090328,
      "grad_norm": 0.1547267884016037,
      "learning_rate": 6.44800196276873e-05,
      "loss": 0.5851,
      "mean_token_accuracy": 0.8188849836587906,
      "num_tokens": 369672712.0,
      "step": 11583
    },
    {
      "entropy": 0.6969467448070645,
      "epoch": 1.065784555580474,
      "grad_norm": 0.16730377078056335,
      "learning_rate": 6.447695280154568e-05,
      "loss": 0.6897,
      "mean_token_accuracy": 0.7922865226864815,
      "num_tokens": 369704494.0,
      "step": 11584
    },
    {
      "entropy": 0.6088058119639754,
      "epoch": 1.0658765619519153,
      "grad_norm": 0.18118621408939362,
      "learning_rate": 6.447388597540406e-05,
      "loss": 0.5905,
      "mean_token_accuracy": 0.8186697587370872,
      "num_tokens": 369736658.0,
      "step": 11585
    },
    {
      "entropy": 0.6558212758973241,
      "epoch": 1.0659685683233564,
      "grad_norm": 0.15899904072284698,
      "learning_rate": 6.447081914926243e-05,
      "loss": 0.6433,
      "mean_token_accuracy": 0.8040133379399776,
      "num_tokens": 369769184.0,
      "step": 11586
    },
    {
      "entropy": 0.5583428917452693,
      "epoch": 1.0660605746947975,
      "grad_norm": 0.1533270627260208,
      "learning_rate": 6.44677523231208e-05,
      "loss": 0.5419,
      "mean_token_accuracy": 0.8313360139727592,
      "num_tokens": 369801520.0,
      "step": 11587
    },
    {
      "entropy": 0.6326333777979016,
      "epoch": 1.066152581066239,
      "grad_norm": 0.1565520465373993,
      "learning_rate": 6.446468549697918e-05,
      "loss": 0.6127,
      "mean_token_accuracy": 0.8109364099800587,
      "num_tokens": 369832591.0,
      "step": 11588
    },
    {
      "entropy": 0.6455669403076172,
      "epoch": 1.06624458743768,
      "grad_norm": 0.15606072545051575,
      "learning_rate": 6.446161867083755e-05,
      "loss": 0.6252,
      "mean_token_accuracy": 0.8080160468816757,
      "num_tokens": 369863775.0,
      "step": 11589
    },
    {
      "entropy": 0.5302552524954081,
      "epoch": 1.0663365938091214,
      "grad_norm": 0.15763446688652039,
      "learning_rate": 6.445855184469593e-05,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8459961749613285,
      "num_tokens": 369896540.0,
      "step": 11590
    },
    {
      "entropy": 0.6353875696659088,
      "epoch": 1.0664286001805625,
      "grad_norm": 0.15829941630363464,
      "learning_rate": 6.44554850185543e-05,
      "loss": 0.6176,
      "mean_token_accuracy": 0.8118090406060219,
      "num_tokens": 369928453.0,
      "step": 11591
    },
    {
      "entropy": 0.6264742873609066,
      "epoch": 1.0665206065520036,
      "grad_norm": 0.16131293773651123,
      "learning_rate": 6.445241819241268e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.8079387471079826,
      "num_tokens": 369960356.0,
      "step": 11592
    },
    {
      "entropy": 0.5894994395785034,
      "epoch": 1.066612612923445,
      "grad_norm": 0.1642971783876419,
      "learning_rate": 6.444935136627105e-05,
      "loss": 0.5828,
      "mean_token_accuracy": 0.82105478271842,
      "num_tokens": 369993061.0,
      "step": 11593
    },
    {
      "entropy": 0.6404398996382952,
      "epoch": 1.0667046192948861,
      "grad_norm": 0.15767104923725128,
      "learning_rate": 6.444628454012943e-05,
      "loss": 0.6251,
      "mean_token_accuracy": 0.8108609318733215,
      "num_tokens": 370025464.0,
      "step": 11594
    },
    {
      "entropy": 0.5835713660344481,
      "epoch": 1.0667966256663275,
      "grad_norm": 0.15280288457870483,
      "learning_rate": 6.444321771398781e-05,
      "loss": 0.5588,
      "mean_token_accuracy": 0.8281281478703022,
      "num_tokens": 370056669.0,
      "step": 11595
    },
    {
      "entropy": 0.5828414503484964,
      "epoch": 1.0668886320377686,
      "grad_norm": 0.1503535360097885,
      "learning_rate": 6.444015088784618e-05,
      "loss": 0.5608,
      "mean_token_accuracy": 0.8225713223218918,
      "num_tokens": 370088631.0,
      "step": 11596
    },
    {
      "entropy": 0.5970825366675854,
      "epoch": 1.0669806384092098,
      "grad_norm": 0.15723933279514313,
      "learning_rate": 6.443708406170454e-05,
      "loss": 0.5801,
      "mean_token_accuracy": 0.8205765560269356,
      "num_tokens": 370120977.0,
      "step": 11597
    },
    {
      "entropy": 0.6746200639754534,
      "epoch": 1.0670726447806511,
      "grad_norm": 0.1663256585597992,
      "learning_rate": 6.443401723556291e-05,
      "loss": 0.6786,
      "mean_token_accuracy": 0.7944399639964104,
      "num_tokens": 370153249.0,
      "step": 11598
    },
    {
      "entropy": 0.6403799429535866,
      "epoch": 1.0671646511520922,
      "grad_norm": 0.16980081796646118,
      "learning_rate": 6.44309504094213e-05,
      "loss": 0.6317,
      "mean_token_accuracy": 0.8080658130347729,
      "num_tokens": 370184667.0,
      "step": 11599
    },
    {
      "entropy": 0.5597822768613696,
      "epoch": 1.0672566575235336,
      "grad_norm": 0.1617022454738617,
      "learning_rate": 6.442788358327967e-05,
      "loss": 0.5429,
      "mean_token_accuracy": 0.8287263438105583,
      "num_tokens": 370216757.0,
      "step": 11600
    },
    {
      "entropy": 0.6803330257534981,
      "epoch": 1.0673486638949747,
      "grad_norm": 0.17978239059448242,
      "learning_rate": 6.442481675713804e-05,
      "loss": 0.6781,
      "mean_token_accuracy": 0.7932637929916382,
      "num_tokens": 370249243.0,
      "step": 11601
    },
    {
      "entropy": 0.6011397335678339,
      "epoch": 1.0674406702664159,
      "grad_norm": 0.15760914981365204,
      "learning_rate": 6.442174993099641e-05,
      "loss": 0.5936,
      "mean_token_accuracy": 0.8169428333640099,
      "num_tokens": 370281429.0,
      "step": 11602
    },
    {
      "entropy": 0.6530337929725647,
      "epoch": 1.0675326766378572,
      "grad_norm": 0.16836531460285187,
      "learning_rate": 6.441868310485479e-05,
      "loss": 0.6453,
      "mean_token_accuracy": 0.8027511462569237,
      "num_tokens": 370313200.0,
      "step": 11603
    },
    {
      "entropy": 0.5887935077771544,
      "epoch": 1.0676246830092984,
      "grad_norm": 0.16241472959518433,
      "learning_rate": 6.441561627871316e-05,
      "loss": 0.57,
      "mean_token_accuracy": 0.8275366984307766,
      "num_tokens": 370344831.0,
      "step": 11604
    },
    {
      "entropy": 0.6944999657571316,
      "epoch": 1.0677166893807397,
      "grad_norm": 0.16555608808994293,
      "learning_rate": 6.441254945257154e-05,
      "loss": 0.6811,
      "mean_token_accuracy": 0.7931646034121513,
      "num_tokens": 370377153.0,
      "step": 11605
    },
    {
      "entropy": 0.6516879517585039,
      "epoch": 1.0678086957521808,
      "grad_norm": 0.1566437929868698,
      "learning_rate": 6.440948262642991e-05,
      "loss": 0.6446,
      "mean_token_accuracy": 0.8013469353318214,
      "num_tokens": 370408740.0,
      "step": 11606
    },
    {
      "entropy": 0.6698861718177795,
      "epoch": 1.067900702123622,
      "grad_norm": 0.16139575839042664,
      "learning_rate": 6.440641580028829e-05,
      "loss": 0.6533,
      "mean_token_accuracy": 0.7994333803653717,
      "num_tokens": 370440616.0,
      "step": 11607
    },
    {
      "entropy": 0.6483898544684052,
      "epoch": 1.0679927084950633,
      "grad_norm": 0.1566259115934372,
      "learning_rate": 6.440334897414666e-05,
      "loss": 0.6244,
      "mean_token_accuracy": 0.8104402869939804,
      "num_tokens": 370472904.0,
      "step": 11608
    },
    {
      "entropy": 0.6192226856946945,
      "epoch": 1.0680847148665045,
      "grad_norm": 0.15247716009616852,
      "learning_rate": 6.440028214800503e-05,
      "loss": 0.5839,
      "mean_token_accuracy": 0.8177198097109795,
      "num_tokens": 370504984.0,
      "step": 11609
    },
    {
      "entropy": 0.6160843316465616,
      "epoch": 1.0681767212379458,
      "grad_norm": 0.1577625721693039,
      "learning_rate": 6.439721532186341e-05,
      "loss": 0.5791,
      "mean_token_accuracy": 0.8213581740856171,
      "num_tokens": 370537512.0,
      "step": 11610
    },
    {
      "entropy": 0.6012045694515109,
      "epoch": 1.068268727609387,
      "grad_norm": 0.15591289103031158,
      "learning_rate": 6.439414849572179e-05,
      "loss": 0.5804,
      "mean_token_accuracy": 0.8205518648028374,
      "num_tokens": 370569771.0,
      "step": 11611
    },
    {
      "entropy": 0.5452516302466393,
      "epoch": 1.068360733980828,
      "grad_norm": 0.16142398118972778,
      "learning_rate": 6.439108166958016e-05,
      "loss": 0.5311,
      "mean_token_accuracy": 0.8356574550271034,
      "num_tokens": 370601518.0,
      "step": 11612
    },
    {
      "entropy": 0.6195458937436342,
      "epoch": 1.0684527403522694,
      "grad_norm": 0.15897707641124725,
      "learning_rate": 6.438801484343852e-05,
      "loss": 0.6123,
      "mean_token_accuracy": 0.8138773292303085,
      "num_tokens": 370634141.0,
      "step": 11613
    },
    {
      "entropy": 0.6130415098741651,
      "epoch": 1.0685447467237106,
      "grad_norm": 0.15422165393829346,
      "learning_rate": 6.438494801729689e-05,
      "loss": 0.5883,
      "mean_token_accuracy": 0.8171565905213356,
      "num_tokens": 370665188.0,
      "step": 11614
    },
    {
      "entropy": 0.5854126764461398,
      "epoch": 1.068636753095152,
      "grad_norm": 0.15695688128471375,
      "learning_rate": 6.438188119115527e-05,
      "loss": 0.575,
      "mean_token_accuracy": 0.8211011476814747,
      "num_tokens": 370696909.0,
      "step": 11615
    },
    {
      "entropy": 0.6403062231838703,
      "epoch": 1.068728759466593,
      "grad_norm": 0.17341773211956024,
      "learning_rate": 6.437881436501366e-05,
      "loss": 0.6307,
      "mean_token_accuracy": 0.8051998317241669,
      "num_tokens": 370728990.0,
      "step": 11616
    },
    {
      "entropy": 0.5872672903351486,
      "epoch": 1.0688207658380342,
      "grad_norm": 0.16143976151943207,
      "learning_rate": 6.437574753887202e-05,
      "loss": 0.5951,
      "mean_token_accuracy": 0.8212840557098389,
      "num_tokens": 370760894.0,
      "step": 11617
    },
    {
      "entropy": 0.5832388438284397,
      "epoch": 1.0689127722094756,
      "grad_norm": 0.14635750651359558,
      "learning_rate": 6.437268071273039e-05,
      "loss": 0.5513,
      "mean_token_accuracy": 0.8277604542672634,
      "num_tokens": 370792683.0,
      "step": 11618
    },
    {
      "entropy": 0.6354620549827814,
      "epoch": 1.0690047785809167,
      "grad_norm": 0.15574628114700317,
      "learning_rate": 6.436961388658877e-05,
      "loss": 0.6212,
      "mean_token_accuracy": 0.8097851425409317,
      "num_tokens": 370824140.0,
      "step": 11619
    },
    {
      "entropy": 0.6053505456075072,
      "epoch": 1.069096784952358,
      "grad_norm": 0.15059928596019745,
      "learning_rate": 6.436654706044714e-05,
      "loss": 0.6033,
      "mean_token_accuracy": 0.8171188794076443,
      "num_tokens": 370856307.0,
      "step": 11620
    },
    {
      "entropy": 0.6567016672343016,
      "epoch": 1.0691887913237992,
      "grad_norm": 0.15713395178318024,
      "learning_rate": 6.436348023430552e-05,
      "loss": 0.6484,
      "mean_token_accuracy": 0.8029917851090431,
      "num_tokens": 370888616.0,
      "step": 11621
    },
    {
      "entropy": 0.6678876467049122,
      "epoch": 1.0692807976952403,
      "grad_norm": 0.15961851179599762,
      "learning_rate": 6.436041340816389e-05,
      "loss": 0.6463,
      "mean_token_accuracy": 0.8007687367498875,
      "num_tokens": 370920547.0,
      "step": 11622
    },
    {
      "entropy": 0.5847564060240984,
      "epoch": 1.0693728040666817,
      "grad_norm": 0.15291441977024078,
      "learning_rate": 6.435734658202227e-05,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8281379826366901,
      "num_tokens": 370952592.0,
      "step": 11623
    },
    {
      "entropy": 0.5737136611714959,
      "epoch": 1.0694648104381228,
      "grad_norm": 0.1535821259021759,
      "learning_rate": 6.435427975588064e-05,
      "loss": 0.5841,
      "mean_token_accuracy": 0.8234042897820473,
      "num_tokens": 370984410.0,
      "step": 11624
    },
    {
      "entropy": 0.6355892457067966,
      "epoch": 1.0695568168095642,
      "grad_norm": 0.16014781594276428,
      "learning_rate": 6.435121292973901e-05,
      "loss": 0.6125,
      "mean_token_accuracy": 0.8098964467644691,
      "num_tokens": 371015087.0,
      "step": 11625
    },
    {
      "entropy": 0.5878320289775729,
      "epoch": 1.0696488231810053,
      "grad_norm": 0.16403815150260925,
      "learning_rate": 6.43481461035974e-05,
      "loss": 0.579,
      "mean_token_accuracy": 0.8227477297186852,
      "num_tokens": 371047162.0,
      "step": 11626
    },
    {
      "entropy": 0.6461958643049002,
      "epoch": 1.0697408295524464,
      "grad_norm": 0.15947087109088898,
      "learning_rate": 6.434507927745577e-05,
      "loss": 0.6406,
      "mean_token_accuracy": 0.803675826638937,
      "num_tokens": 371079389.0,
      "step": 11627
    },
    {
      "entropy": 0.6425506919622421,
      "epoch": 1.0698328359238878,
      "grad_norm": 0.15793295204639435,
      "learning_rate": 6.434201245131414e-05,
      "loss": 0.6246,
      "mean_token_accuracy": 0.8072685413062572,
      "num_tokens": 371110891.0,
      "step": 11628
    },
    {
      "entropy": 0.6717328149825335,
      "epoch": 1.069924842295329,
      "grad_norm": 0.16214148700237274,
      "learning_rate": 6.43389456251725e-05,
      "loss": 0.6644,
      "mean_token_accuracy": 0.7973414398729801,
      "num_tokens": 371142965.0,
      "step": 11629
    },
    {
      "entropy": 0.5758958272635937,
      "epoch": 1.0700168486667703,
      "grad_norm": 0.16641193628311157,
      "learning_rate": 6.433587879903089e-05,
      "loss": 0.5741,
      "mean_token_accuracy": 0.8242321759462357,
      "num_tokens": 371175038.0,
      "step": 11630
    },
    {
      "entropy": 0.6270127706229687,
      "epoch": 1.0701088550382114,
      "grad_norm": 0.15508399903774261,
      "learning_rate": 6.433281197288927e-05,
      "loss": 0.6184,
      "mean_token_accuracy": 0.8094885647296906,
      "num_tokens": 371207239.0,
      "step": 11631
    },
    {
      "entropy": 0.6146877398714423,
      "epoch": 1.0702008614096525,
      "grad_norm": 0.16273066401481628,
      "learning_rate": 6.432974514674764e-05,
      "loss": 0.6083,
      "mean_token_accuracy": 0.8209579512476921,
      "num_tokens": 371239090.0,
      "step": 11632
    },
    {
      "entropy": 0.5980602204799652,
      "epoch": 1.070292867781094,
      "grad_norm": 0.15322989225387573,
      "learning_rate": 6.4326678320606e-05,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8264445066452026,
      "num_tokens": 371271078.0,
      "step": 11633
    },
    {
      "entropy": 0.5868565202690661,
      "epoch": 1.070384874152535,
      "grad_norm": 0.15082858502864838,
      "learning_rate": 6.432361149446439e-05,
      "loss": 0.578,
      "mean_token_accuracy": 0.822708822786808,
      "num_tokens": 371303819.0,
      "step": 11634
    },
    {
      "entropy": 0.5132240001112223,
      "epoch": 1.0704768805239764,
      "grad_norm": 0.15319772064685822,
      "learning_rate": 6.432054466832275e-05,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8497599959373474,
      "num_tokens": 371336368.0,
      "step": 11635
    },
    {
      "entropy": 0.5943759717047215,
      "epoch": 1.0705688868954175,
      "grad_norm": 0.15097403526306152,
      "learning_rate": 6.431747784218113e-05,
      "loss": 0.5803,
      "mean_token_accuracy": 0.8227888271212578,
      "num_tokens": 371368830.0,
      "step": 11636
    },
    {
      "entropy": 0.7277845181524754,
      "epoch": 1.0706608932668586,
      "grad_norm": 0.1622588336467743,
      "learning_rate": 6.43144110160395e-05,
      "loss": 0.7167,
      "mean_token_accuracy": 0.7844566218554974,
      "num_tokens": 371401053.0,
      "step": 11637
    },
    {
      "entropy": 0.6677747908979654,
      "epoch": 1.0707528996383,
      "grad_norm": 0.16205556690692902,
      "learning_rate": 6.431134418989788e-05,
      "loss": 0.6451,
      "mean_token_accuracy": 0.7985911555588245,
      "num_tokens": 371432586.0,
      "step": 11638
    },
    {
      "entropy": 0.7139922603964806,
      "epoch": 1.0708449060097411,
      "grad_norm": 0.1656411737203598,
      "learning_rate": 6.430827736375625e-05,
      "loss": 0.6955,
      "mean_token_accuracy": 0.7915518283843994,
      "num_tokens": 371464579.0,
      "step": 11639
    },
    {
      "entropy": 0.6666705030947924,
      "epoch": 1.0709369123811825,
      "grad_norm": 0.1606798619031906,
      "learning_rate": 6.430521053761462e-05,
      "loss": 0.6467,
      "mean_token_accuracy": 0.8009248860180378,
      "num_tokens": 371497044.0,
      "step": 11640
    },
    {
      "entropy": 0.665817528963089,
      "epoch": 1.0710289187526236,
      "grad_norm": 0.1618146002292633,
      "learning_rate": 6.4302143711473e-05,
      "loss": 0.6439,
      "mean_token_accuracy": 0.8006245121359825,
      "num_tokens": 371528861.0,
      "step": 11641
    },
    {
      "entropy": 0.5409806370735168,
      "epoch": 1.0711209251240648,
      "grad_norm": 0.1538720279932022,
      "learning_rate": 6.429907688533138e-05,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8366112150251865,
      "num_tokens": 371561116.0,
      "step": 11642
    },
    {
      "entropy": 0.6165948491543531,
      "epoch": 1.0712129314955061,
      "grad_norm": 0.1574084609746933,
      "learning_rate": 6.429601005918975e-05,
      "loss": 0.5972,
      "mean_token_accuracy": 0.8117814436554909,
      "num_tokens": 371592310.0,
      "step": 11643
    },
    {
      "entropy": 0.5868088752031326,
      "epoch": 1.0713049378669472,
      "grad_norm": 0.15445466339588165,
      "learning_rate": 6.429294323304812e-05,
      "loss": 0.5703,
      "mean_token_accuracy": 0.8265461958944798,
      "num_tokens": 371624201.0,
      "step": 11644
    },
    {
      "entropy": 0.5644895620644093,
      "epoch": 1.0713969442383886,
      "grad_norm": 0.14991754293441772,
      "learning_rate": 6.428987640690649e-05,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8284119628369808,
      "num_tokens": 371655788.0,
      "step": 11645
    },
    {
      "entropy": 0.662418432533741,
      "epoch": 1.0714889506098297,
      "grad_norm": 0.16713286936283112,
      "learning_rate": 6.428680958076487e-05,
      "loss": 0.6594,
      "mean_token_accuracy": 0.7997461296617985,
      "num_tokens": 371687923.0,
      "step": 11646
    },
    {
      "entropy": 0.6255277022719383,
      "epoch": 1.0715809569812709,
      "grad_norm": 0.15264049172401428,
      "learning_rate": 6.428374275462325e-05,
      "loss": 0.5939,
      "mean_token_accuracy": 0.817864615470171,
      "num_tokens": 371719696.0,
      "step": 11647
    },
    {
      "entropy": 0.6617561466991901,
      "epoch": 1.0716729633527122,
      "grad_norm": 0.15623360872268677,
      "learning_rate": 6.428067592848162e-05,
      "loss": 0.6489,
      "mean_token_accuracy": 0.8019847050309181,
      "num_tokens": 371752181.0,
      "step": 11648
    },
    {
      "entropy": 0.7208687141537666,
      "epoch": 1.0717649697241534,
      "grad_norm": 0.1636999100446701,
      "learning_rate": 6.427760910233998e-05,
      "loss": 0.7052,
      "mean_token_accuracy": 0.7812629975378513,
      "num_tokens": 371784399.0,
      "step": 11649
    },
    {
      "entropy": 0.7220601756125689,
      "epoch": 1.0718569760955947,
      "grad_norm": 0.17494699358940125,
      "learning_rate": 6.427454227619837e-05,
      "loss": 0.7014,
      "mean_token_accuracy": 0.7870981208980083,
      "num_tokens": 371815671.0,
      "step": 11650
    },
    {
      "entropy": 0.5734339393675327,
      "epoch": 1.0719489824670358,
      "grad_norm": 0.15487957000732422,
      "learning_rate": 6.427147545005673e-05,
      "loss": 0.5636,
      "mean_token_accuracy": 0.8201901018619537,
      "num_tokens": 371846928.0,
      "step": 11651
    },
    {
      "entropy": 0.5684410966932774,
      "epoch": 1.072040988838477,
      "grad_norm": 0.1559053510427475,
      "learning_rate": 6.426840862391512e-05,
      "loss": 0.5576,
      "mean_token_accuracy": 0.8256592266261578,
      "num_tokens": 371878845.0,
      "step": 11652
    },
    {
      "entropy": 0.6935500595718622,
      "epoch": 1.0721329952099183,
      "grad_norm": 0.17136777937412262,
      "learning_rate": 6.42653417977735e-05,
      "loss": 0.6889,
      "mean_token_accuracy": 0.792652640491724,
      "num_tokens": 371910011.0,
      "step": 11653
    },
    {
      "entropy": 0.6034531462937593,
      "epoch": 1.0722250015813595,
      "grad_norm": 0.1571730524301529,
      "learning_rate": 6.426227497163187e-05,
      "loss": 0.5756,
      "mean_token_accuracy": 0.8270491994917393,
      "num_tokens": 371941803.0,
      "step": 11654
    },
    {
      "entropy": 0.4832216026261449,
      "epoch": 1.0723170079528008,
      "grad_norm": 0.15239962935447693,
      "learning_rate": 6.425920814549023e-05,
      "loss": 0.4859,
      "mean_token_accuracy": 0.8490453958511353,
      "num_tokens": 371973696.0,
      "step": 11655
    },
    {
      "entropy": 0.5356352021917701,
      "epoch": 1.072409014324242,
      "grad_norm": 0.14684809744358063,
      "learning_rate": 6.42561413193486e-05,
      "loss": 0.5267,
      "mean_token_accuracy": 0.8387629315257072,
      "num_tokens": 372006067.0,
      "step": 11656
    },
    {
      "entropy": 0.6342138964682817,
      "epoch": 1.072501020695683,
      "grad_norm": 0.16580000519752502,
      "learning_rate": 6.425307449320698e-05,
      "loss": 0.6293,
      "mean_token_accuracy": 0.8081441037356853,
      "num_tokens": 372037729.0,
      "step": 11657
    },
    {
      "entropy": 0.6093416903167963,
      "epoch": 1.0725930270671244,
      "grad_norm": 0.1567615568637848,
      "learning_rate": 6.425000766706536e-05,
      "loss": 0.6019,
      "mean_token_accuracy": 0.8144979923963547,
      "num_tokens": 372069624.0,
      "step": 11658
    },
    {
      "entropy": 0.7311144080013037,
      "epoch": 1.0726850334385656,
      "grad_norm": 0.16963379085063934,
      "learning_rate": 6.424694084092373e-05,
      "loss": 0.7187,
      "mean_token_accuracy": 0.7841745913028717,
      "num_tokens": 372101335.0,
      "step": 11659
    },
    {
      "entropy": 0.6017667837440968,
      "epoch": 1.072777039810007,
      "grad_norm": 0.1557818055152893,
      "learning_rate": 6.42438740147821e-05,
      "loss": 0.5702,
      "mean_token_accuracy": 0.8236939124763012,
      "num_tokens": 372133852.0,
      "step": 11660
    },
    {
      "entropy": 0.7382163126021624,
      "epoch": 1.072869046181448,
      "grad_norm": 0.16939115524291992,
      "learning_rate": 6.424080718864048e-05,
      "loss": 0.7294,
      "mean_token_accuracy": 0.7779907286167145,
      "num_tokens": 372166089.0,
      "step": 11661
    },
    {
      "entropy": 0.5989611567929387,
      "epoch": 1.0729610525528892,
      "grad_norm": 0.15835238993167877,
      "learning_rate": 6.423774036249885e-05,
      "loss": 0.5729,
      "mean_token_accuracy": 0.8217437602579594,
      "num_tokens": 372197428.0,
      "step": 11662
    },
    {
      "entropy": 0.6735384985804558,
      "epoch": 1.0730530589243306,
      "grad_norm": 0.1611841470003128,
      "learning_rate": 6.423467353635723e-05,
      "loss": 0.66,
      "mean_token_accuracy": 0.799114752560854,
      "num_tokens": 372229958.0,
      "step": 11663
    },
    {
      "entropy": 0.6216978002339602,
      "epoch": 1.0731450652957717,
      "grad_norm": 0.16140873730182648,
      "learning_rate": 6.42316067102156e-05,
      "loss": 0.5985,
      "mean_token_accuracy": 0.8176837153732777,
      "num_tokens": 372261426.0,
      "step": 11664
    },
    {
      "entropy": 0.4692254848778248,
      "epoch": 1.073237071667213,
      "grad_norm": 0.1544477641582489,
      "learning_rate": 6.422853988407398e-05,
      "loss": 0.4482,
      "mean_token_accuracy": 0.8587408103048801,
      "num_tokens": 372294163.0,
      "step": 11665
    },
    {
      "entropy": 0.7470983369275928,
      "epoch": 1.0733290780386542,
      "grad_norm": 0.16874828934669495,
      "learning_rate": 6.422547305793235e-05,
      "loss": 0.7416,
      "mean_token_accuracy": 0.7774449624121189,
      "num_tokens": 372325913.0,
      "step": 11666
    },
    {
      "entropy": 0.6699259858578444,
      "epoch": 1.0734210844100953,
      "grad_norm": 0.16273334622383118,
      "learning_rate": 6.422240623179072e-05,
      "loss": 0.6633,
      "mean_token_accuracy": 0.7977919578552246,
      "num_tokens": 372357384.0,
      "step": 11667
    },
    {
      "entropy": 0.5773841487243772,
      "epoch": 1.0735130907815367,
      "grad_norm": 0.15742623805999756,
      "learning_rate": 6.42193394056491e-05,
      "loss": 0.5718,
      "mean_token_accuracy": 0.8233840987086296,
      "num_tokens": 372389653.0,
      "step": 11668
    },
    {
      "entropy": 0.6137265376746655,
      "epoch": 1.0736050971529778,
      "grad_norm": 0.16545487940311432,
      "learning_rate": 6.421627257950748e-05,
      "loss": 0.6095,
      "mean_token_accuracy": 0.8161028325557709,
      "num_tokens": 372422421.0,
      "step": 11669
    },
    {
      "entropy": 0.6465345080941916,
      "epoch": 1.0736971035244192,
      "grad_norm": 0.1570456475019455,
      "learning_rate": 6.421320575336585e-05,
      "loss": 0.6417,
      "mean_token_accuracy": 0.8044513501226902,
      "num_tokens": 372454654.0,
      "step": 11670
    },
    {
      "entropy": 0.6513697914779186,
      "epoch": 1.0737891098958603,
      "grad_norm": 0.17072387039661407,
      "learning_rate": 6.421013892722421e-05,
      "loss": 0.6457,
      "mean_token_accuracy": 0.8026051633059978,
      "num_tokens": 372486717.0,
      "step": 11671
    },
    {
      "entropy": 0.6555443042889237,
      "epoch": 1.0738811162673014,
      "grad_norm": 0.15131521224975586,
      "learning_rate": 6.420707210108258e-05,
      "loss": 0.6333,
      "mean_token_accuracy": 0.8073174431920052,
      "num_tokens": 372519456.0,
      "step": 11672
    },
    {
      "entropy": 0.6106728091835976,
      "epoch": 1.0739731226387428,
      "grad_norm": 0.15392036736011505,
      "learning_rate": 6.420400527494098e-05,
      "loss": 0.5722,
      "mean_token_accuracy": 0.8221507221460342,
      "num_tokens": 372550626.0,
      "step": 11673
    },
    {
      "entropy": 0.6584234954789281,
      "epoch": 1.074065129010184,
      "grad_norm": 0.15587778389453888,
      "learning_rate": 6.420093844879934e-05,
      "loss": 0.6374,
      "mean_token_accuracy": 0.8042694739997387,
      "num_tokens": 372583231.0,
      "step": 11674
    },
    {
      "entropy": 0.6802284941077232,
      "epoch": 1.0741571353816253,
      "grad_norm": 0.15879328548908234,
      "learning_rate": 6.419787162265771e-05,
      "loss": 0.657,
      "mean_token_accuracy": 0.7968066595494747,
      "num_tokens": 372614650.0,
      "step": 11675
    },
    {
      "entropy": 0.6051208730787039,
      "epoch": 1.0742491417530664,
      "grad_norm": 0.15669173002243042,
      "learning_rate": 6.419480479651608e-05,
      "loss": 0.5898,
      "mean_token_accuracy": 0.8167270570993423,
      "num_tokens": 372646220.0,
      "step": 11676
    },
    {
      "entropy": 0.6843086816370487,
      "epoch": 1.0743411481245075,
      "grad_norm": 0.1585986465215683,
      "learning_rate": 6.419173797037446e-05,
      "loss": 0.6735,
      "mean_token_accuracy": 0.7942273169755936,
      "num_tokens": 372678569.0,
      "step": 11677
    },
    {
      "entropy": 0.5794034227728844,
      "epoch": 1.074433154495949,
      "grad_norm": 0.1572801172733307,
      "learning_rate": 6.418867114423284e-05,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8260965347290039,
      "num_tokens": 372709552.0,
      "step": 11678
    },
    {
      "entropy": 0.7362527623772621,
      "epoch": 1.07452516086739,
      "grad_norm": 0.16612538695335388,
      "learning_rate": 6.418560431809121e-05,
      "loss": 0.7214,
      "mean_token_accuracy": 0.777455747127533,
      "num_tokens": 372741968.0,
      "step": 11679
    },
    {
      "entropy": 0.5268832305446267,
      "epoch": 1.0746171672388314,
      "grad_norm": 0.15639866888523102,
      "learning_rate": 6.418253749194958e-05,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8371351882815361,
      "num_tokens": 372773129.0,
      "step": 11680
    },
    {
      "entropy": 0.6240232419222593,
      "epoch": 1.0747091736102725,
      "grad_norm": 0.15827402472496033,
      "learning_rate": 6.417947066580796e-05,
      "loss": 0.6388,
      "mean_token_accuracy": 0.8072079680860043,
      "num_tokens": 372805327.0,
      "step": 11681
    },
    {
      "entropy": 0.6210119649767876,
      "epoch": 1.0748011799817137,
      "grad_norm": 0.1573728770017624,
      "learning_rate": 6.417640383966633e-05,
      "loss": 0.6273,
      "mean_token_accuracy": 0.8104009330272675,
      "num_tokens": 372837017.0,
      "step": 11682
    },
    {
      "entropy": 0.668768297880888,
      "epoch": 1.074893186353155,
      "grad_norm": 0.16070453822612762,
      "learning_rate": 6.417333701352471e-05,
      "loss": 0.6685,
      "mean_token_accuracy": 0.7970470897853374,
      "num_tokens": 372869022.0,
      "step": 11683
    },
    {
      "entropy": 0.6882192399352789,
      "epoch": 1.0749851927245961,
      "grad_norm": 0.16847407817840576,
      "learning_rate": 6.417027018738309e-05,
      "loss": 0.6946,
      "mean_token_accuracy": 0.7911431007087231,
      "num_tokens": 372901695.0,
      "step": 11684
    },
    {
      "entropy": 0.6460868776775897,
      "epoch": 1.0750771990960375,
      "grad_norm": 0.16570855677127838,
      "learning_rate": 6.416720336124146e-05,
      "loss": 0.626,
      "mean_token_accuracy": 0.8065807968378067,
      "num_tokens": 372933700.0,
      "step": 11685
    },
    {
      "entropy": 0.5836671404540539,
      "epoch": 1.0751692054674786,
      "grad_norm": 0.1585967093706131,
      "learning_rate": 6.416413653509983e-05,
      "loss": 0.5519,
      "mean_token_accuracy": 0.8293144889175892,
      "num_tokens": 372966119.0,
      "step": 11686
    },
    {
      "entropy": 0.635142408311367,
      "epoch": 1.0752612118389198,
      "grad_norm": 0.16564702987670898,
      "learning_rate": 6.41610697089582e-05,
      "loss": 0.6003,
      "mean_token_accuracy": 0.8144719786942005,
      "num_tokens": 372997866.0,
      "step": 11687
    },
    {
      "entropy": 0.6749577783048153,
      "epoch": 1.0753532182103611,
      "grad_norm": 0.1644732654094696,
      "learning_rate": 6.415800288281658e-05,
      "loss": 0.6651,
      "mean_token_accuracy": 0.7964833229780197,
      "num_tokens": 373029801.0,
      "step": 11688
    },
    {
      "entropy": 0.6247204663231969,
      "epoch": 1.0754452245818023,
      "grad_norm": 0.16805550456047058,
      "learning_rate": 6.415493605667496e-05,
      "loss": 0.6051,
      "mean_token_accuracy": 0.8139891177415848,
      "num_tokens": 373061928.0,
      "step": 11689
    },
    {
      "entropy": 0.6906015705317259,
      "epoch": 1.0755372309532436,
      "grad_norm": 0.1623317152261734,
      "learning_rate": 6.415186923053333e-05,
      "loss": 0.6799,
      "mean_token_accuracy": 0.7907254993915558,
      "num_tokens": 373093253.0,
      "step": 11690
    },
    {
      "entropy": 0.579786934889853,
      "epoch": 1.0756292373246847,
      "grad_norm": 0.15861250460147858,
      "learning_rate": 6.414880240439169e-05,
      "loss": 0.5683,
      "mean_token_accuracy": 0.8221667781472206,
      "num_tokens": 373124660.0,
      "step": 11691
    },
    {
      "entropy": 0.6043314589187503,
      "epoch": 1.0757212436961259,
      "grad_norm": 0.15751422941684723,
      "learning_rate": 6.414573557825007e-05,
      "loss": 0.5821,
      "mean_token_accuracy": 0.8245944194495678,
      "num_tokens": 373156823.0,
      "step": 11692
    },
    {
      "entropy": 0.6465194504708052,
      "epoch": 1.0758132500675672,
      "grad_norm": 0.1655506193637848,
      "learning_rate": 6.414266875210844e-05,
      "loss": 0.6421,
      "mean_token_accuracy": 0.8039284721016884,
      "num_tokens": 373188308.0,
      "step": 11693
    },
    {
      "entropy": 0.6066624019294977,
      "epoch": 1.0759052564390084,
      "grad_norm": 0.16162000596523285,
      "learning_rate": 6.413960192596682e-05,
      "loss": 0.5986,
      "mean_token_accuracy": 0.8157803528010845,
      "num_tokens": 373220635.0,
      "step": 11694
    },
    {
      "entropy": 0.6411512149497867,
      "epoch": 1.0759972628104497,
      "grad_norm": 0.15885984897613525,
      "learning_rate": 6.413653509982519e-05,
      "loss": 0.6431,
      "mean_token_accuracy": 0.804802268743515,
      "num_tokens": 373252370.0,
      "step": 11695
    },
    {
      "entropy": 0.5395876783877611,
      "epoch": 1.0760892691818909,
      "grad_norm": 0.14879904687404633,
      "learning_rate": 6.413346827368357e-05,
      "loss": 0.5231,
      "mean_token_accuracy": 0.8406795114278793,
      "num_tokens": 373284677.0,
      "step": 11696
    },
    {
      "entropy": 0.5941870249807835,
      "epoch": 1.076181275553332,
      "grad_norm": 0.16138924658298492,
      "learning_rate": 6.413040144754194e-05,
      "loss": 0.5653,
      "mean_token_accuracy": 0.8243887983262539,
      "num_tokens": 373316864.0,
      "step": 11697
    },
    {
      "entropy": 0.6280372608453035,
      "epoch": 1.0762732819247733,
      "grad_norm": 0.1628701388835907,
      "learning_rate": 6.412733462140031e-05,
      "loss": 0.6159,
      "mean_token_accuracy": 0.8085381388664246,
      "num_tokens": 373348895.0,
      "step": 11698
    },
    {
      "entropy": 0.6196625353768468,
      "epoch": 1.0763652882962145,
      "grad_norm": 0.16152949631214142,
      "learning_rate": 6.412426779525869e-05,
      "loss": 0.6125,
      "mean_token_accuracy": 0.8142991699278355,
      "num_tokens": 373380747.0,
      "step": 11699
    },
    {
      "entropy": 0.543799951672554,
      "epoch": 1.0764572946676558,
      "grad_norm": 0.14794422686100006,
      "learning_rate": 6.412120096911707e-05,
      "loss": 0.5402,
      "mean_token_accuracy": 0.8324489891529083,
      "num_tokens": 373412942.0,
      "step": 11700
    },
    {
      "entropy": 0.6389734642580152,
      "epoch": 1.076549301039097,
      "grad_norm": 0.15725553035736084,
      "learning_rate": 6.411813414297544e-05,
      "loss": 0.6317,
      "mean_token_accuracy": 0.8069495521485806,
      "num_tokens": 373445707.0,
      "step": 11701
    },
    {
      "entropy": 0.5901557691395283,
      "epoch": 1.076641307410538,
      "grad_norm": 0.16474151611328125,
      "learning_rate": 6.411506731683381e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8211846575140953,
      "num_tokens": 373477267.0,
      "step": 11702
    },
    {
      "entropy": 0.5357124069705606,
      "epoch": 1.0767333137819795,
      "grad_norm": 0.14520716667175293,
      "learning_rate": 6.411200049069218e-05,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8376220911741257,
      "num_tokens": 373509501.0,
      "step": 11703
    },
    {
      "entropy": 0.7386556882411242,
      "epoch": 1.0768253201534206,
      "grad_norm": 0.1662726104259491,
      "learning_rate": 6.410893366455056e-05,
      "loss": 0.7268,
      "mean_token_accuracy": 0.778567660599947,
      "num_tokens": 373541777.0,
      "step": 11704
    },
    {
      "entropy": 0.6760766142979264,
      "epoch": 1.076917326524862,
      "grad_norm": 0.16994205117225647,
      "learning_rate": 6.410586683840894e-05,
      "loss": 0.6634,
      "mean_token_accuracy": 0.7965070232748985,
      "num_tokens": 373574165.0,
      "step": 11705
    },
    {
      "entropy": 0.622443369589746,
      "epoch": 1.077009332896303,
      "grad_norm": 0.15661582350730896,
      "learning_rate": 6.41028000122673e-05,
      "loss": 0.6122,
      "mean_token_accuracy": 0.8122677281498909,
      "num_tokens": 373606901.0,
      "step": 11706
    },
    {
      "entropy": 0.6906845476478338,
      "epoch": 1.0771013392677442,
      "grad_norm": 0.16382743418216705,
      "learning_rate": 6.409973318612567e-05,
      "loss": 0.6738,
      "mean_token_accuracy": 0.7939139790832996,
      "num_tokens": 373637913.0,
      "step": 11707
    },
    {
      "entropy": 0.6488401954993606,
      "epoch": 1.0771933456391856,
      "grad_norm": 0.16932536661624908,
      "learning_rate": 6.409666635998406e-05,
      "loss": 0.6361,
      "mean_token_accuracy": 0.8039263971149921,
      "num_tokens": 373669878.0,
      "step": 11708
    },
    {
      "entropy": 0.5591107867658138,
      "epoch": 1.0772853520106267,
      "grad_norm": 0.15846744179725647,
      "learning_rate": 6.409359953384244e-05,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8307424038648605,
      "num_tokens": 373702095.0,
      "step": 11709
    },
    {
      "entropy": 0.6959099369123578,
      "epoch": 1.077377358382068,
      "grad_norm": 0.17049813270568848,
      "learning_rate": 6.40905327077008e-05,
      "loss": 0.6999,
      "mean_token_accuracy": 0.7890578880906105,
      "num_tokens": 373733302.0,
      "step": 11710
    },
    {
      "entropy": 0.6172031722962856,
      "epoch": 1.0774693647535092,
      "grad_norm": 0.1581602692604065,
      "learning_rate": 6.408746588155917e-05,
      "loss": 0.6148,
      "mean_token_accuracy": 0.8139040395617485,
      "num_tokens": 373765687.0,
      "step": 11711
    },
    {
      "entropy": 0.5916288513690233,
      "epoch": 1.0775613711249503,
      "grad_norm": 0.1563778966665268,
      "learning_rate": 6.408439905541755e-05,
      "loss": 0.5751,
      "mean_token_accuracy": 0.8207953944802284,
      "num_tokens": 373797898.0,
      "step": 11712
    },
    {
      "entropy": 0.5357247502543032,
      "epoch": 1.0776533774963917,
      "grad_norm": 0.15581223368644714,
      "learning_rate": 6.408133222927592e-05,
      "loss": 0.5181,
      "mean_token_accuracy": 0.8385773524641991,
      "num_tokens": 373829983.0,
      "step": 11713
    },
    {
      "entropy": 0.6172058880329132,
      "epoch": 1.0777453838678328,
      "grad_norm": 0.15671876072883606,
      "learning_rate": 6.40782654031343e-05,
      "loss": 0.5971,
      "mean_token_accuracy": 0.8125688172876835,
      "num_tokens": 373861991.0,
      "step": 11714
    },
    {
      "entropy": 0.6001919414848089,
      "epoch": 1.0778373902392742,
      "grad_norm": 0.15659430623054504,
      "learning_rate": 6.407519857699268e-05,
      "loss": 0.5843,
      "mean_token_accuracy": 0.8176098130643368,
      "num_tokens": 373894646.0,
      "step": 11715
    },
    {
      "entropy": 0.6181738134473562,
      "epoch": 1.0779293966107153,
      "grad_norm": 0.1609310358762741,
      "learning_rate": 6.407213175085105e-05,
      "loss": 0.6063,
      "mean_token_accuracy": 0.8155347108840942,
      "num_tokens": 373926078.0,
      "step": 11716
    },
    {
      "entropy": 0.6466336529701948,
      "epoch": 1.0780214029821564,
      "grad_norm": 0.1649147868156433,
      "learning_rate": 6.406906492470942e-05,
      "loss": 0.6399,
      "mean_token_accuracy": 0.8025315850973129,
      "num_tokens": 373957884.0,
      "step": 11717
    },
    {
      "entropy": 0.6858938876539469,
      "epoch": 1.0781134093535978,
      "grad_norm": 0.15868540108203888,
      "learning_rate": 6.406599809856779e-05,
      "loss": 0.6716,
      "mean_token_accuracy": 0.7954978942871094,
      "num_tokens": 373989692.0,
      "step": 11718
    },
    {
      "entropy": 0.5647694440558553,
      "epoch": 1.078205415725039,
      "grad_norm": 0.15769553184509277,
      "learning_rate": 6.406293127242617e-05,
      "loss": 0.5393,
      "mean_token_accuracy": 0.8311141580343246,
      "num_tokens": 374021608.0,
      "step": 11719
    },
    {
      "entropy": 0.6561401188373566,
      "epoch": 1.0782974220964803,
      "grad_norm": 0.15798857808113098,
      "learning_rate": 6.405986444628455e-05,
      "loss": 0.6377,
      "mean_token_accuracy": 0.8050548359751701,
      "num_tokens": 374053396.0,
      "step": 11720
    },
    {
      "entropy": 0.5258029857650399,
      "epoch": 1.0783894284679214,
      "grad_norm": 0.1575159728527069,
      "learning_rate": 6.405679762014292e-05,
      "loss": 0.5044,
      "mean_token_accuracy": 0.8436098657548428,
      "num_tokens": 374085384.0,
      "step": 11721
    },
    {
      "entropy": 0.7223948407918215,
      "epoch": 1.0784814348393625,
      "grad_norm": 0.16089577972888947,
      "learning_rate": 6.405373079400129e-05,
      "loss": 0.7162,
      "mean_token_accuracy": 0.7820792719721794,
      "num_tokens": 374117380.0,
      "step": 11722
    },
    {
      "entropy": 0.6862637661397457,
      "epoch": 1.078573441210804,
      "grad_norm": 0.15714521706104279,
      "learning_rate": 6.405066396785967e-05,
      "loss": 0.6663,
      "mean_token_accuracy": 0.7974637448787689,
      "num_tokens": 374149269.0,
      "step": 11723
    },
    {
      "entropy": 0.570886267349124,
      "epoch": 1.078665447582245,
      "grad_norm": 0.15285691618919373,
      "learning_rate": 6.404759714171804e-05,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8322359137237072,
      "num_tokens": 374180772.0,
      "step": 11724
    },
    {
      "entropy": 0.6274473071098328,
      "epoch": 1.0787574539536864,
      "grad_norm": 0.1592620611190796,
      "learning_rate": 6.404453031557642e-05,
      "loss": 0.6252,
      "mean_token_accuracy": 0.806906346231699,
      "num_tokens": 374212919.0,
      "step": 11725
    },
    {
      "entropy": 0.691464563831687,
      "epoch": 1.0788494603251275,
      "grad_norm": 0.16094227135181427,
      "learning_rate": 6.404146348943479e-05,
      "loss": 0.6879,
      "mean_token_accuracy": 0.7941335998475552,
      "num_tokens": 374245195.0,
      "step": 11726
    },
    {
      "entropy": 0.5801641009747982,
      "epoch": 1.0789414666965687,
      "grad_norm": 0.158652663230896,
      "learning_rate": 6.403839666329317e-05,
      "loss": 0.5836,
      "mean_token_accuracy": 0.8184558488428593,
      "num_tokens": 374277262.0,
      "step": 11727
    },
    {
      "entropy": 0.58926004730165,
      "epoch": 1.07903347306801,
      "grad_norm": 0.15594734251499176,
      "learning_rate": 6.403532983715153e-05,
      "loss": 0.5928,
      "mean_token_accuracy": 0.8175482079386711,
      "num_tokens": 374309335.0,
      "step": 11728
    },
    {
      "entropy": 0.6833104323595762,
      "epoch": 1.0791254794394511,
      "grad_norm": 0.1611059159040451,
      "learning_rate": 6.40322630110099e-05,
      "loss": 0.6744,
      "mean_token_accuracy": 0.795548502355814,
      "num_tokens": 374341648.0,
      "step": 11729
    },
    {
      "entropy": 0.6849204711616039,
      "epoch": 1.0792174858108925,
      "grad_norm": 0.16746802628040314,
      "learning_rate": 6.402919618486828e-05,
      "loss": 0.6824,
      "mean_token_accuracy": 0.7932930327951908,
      "num_tokens": 374373170.0,
      "step": 11730
    },
    {
      "entropy": 0.606786260381341,
      "epoch": 1.0793094921823336,
      "grad_norm": 0.1579751968383789,
      "learning_rate": 6.402612935872667e-05,
      "loss": 0.5938,
      "mean_token_accuracy": 0.8170256465673447,
      "num_tokens": 374405101.0,
      "step": 11731
    },
    {
      "entropy": 0.6111011300235987,
      "epoch": 1.0794014985537748,
      "grad_norm": 0.16019785404205322,
      "learning_rate": 6.402306253258503e-05,
      "loss": 0.5943,
      "mean_token_accuracy": 0.8169584386050701,
      "num_tokens": 374436884.0,
      "step": 11732
    },
    {
      "entropy": 0.6522915977984667,
      "epoch": 1.0794935049252161,
      "grad_norm": 0.16081653535366058,
      "learning_rate": 6.40199957064434e-05,
      "loss": 0.6368,
      "mean_token_accuracy": 0.8040214516222477,
      "num_tokens": 374468373.0,
      "step": 11733
    },
    {
      "entropy": 0.5367119582369924,
      "epoch": 1.0795855112966573,
      "grad_norm": 0.15747742354869843,
      "learning_rate": 6.401692888030177e-05,
      "loss": 0.5168,
      "mean_token_accuracy": 0.8424864262342453,
      "num_tokens": 374500243.0,
      "step": 11734
    },
    {
      "entropy": 0.6577965412288904,
      "epoch": 1.0796775176680986,
      "grad_norm": 0.16539588570594788,
      "learning_rate": 6.401386205416015e-05,
      "loss": 0.6353,
      "mean_token_accuracy": 0.8040721379220486,
      "num_tokens": 374531781.0,
      "step": 11735
    },
    {
      "entropy": 0.6665035430341959,
      "epoch": 1.0797695240395397,
      "grad_norm": 0.1610693484544754,
      "learning_rate": 6.401079522801853e-05,
      "loss": 0.6519,
      "mean_token_accuracy": 0.7998344711959362,
      "num_tokens": 374564037.0,
      "step": 11736
    },
    {
      "entropy": 0.683708418160677,
      "epoch": 1.0798615304109809,
      "grad_norm": 0.1689496487379074,
      "learning_rate": 6.40077284018769e-05,
      "loss": 0.6804,
      "mean_token_accuracy": 0.7935049757361412,
      "num_tokens": 374595433.0,
      "step": 11737
    },
    {
      "entropy": 0.7313556149601936,
      "epoch": 1.0799535367824222,
      "grad_norm": 0.16346123814582825,
      "learning_rate": 6.400466157573527e-05,
      "loss": 0.7171,
      "mean_token_accuracy": 0.7807749137282372,
      "num_tokens": 374627563.0,
      "step": 11738
    },
    {
      "entropy": 0.6428047604858875,
      "epoch": 1.0800455431538634,
      "grad_norm": 0.1648382693529129,
      "learning_rate": 6.400159474959365e-05,
      "loss": 0.6194,
      "mean_token_accuracy": 0.8126425035297871,
      "num_tokens": 374659101.0,
      "step": 11739
    },
    {
      "entropy": 0.6115284692496061,
      "epoch": 1.0801375495253047,
      "grad_norm": 0.16023080050945282,
      "learning_rate": 6.399852792345202e-05,
      "loss": 0.5879,
      "mean_token_accuracy": 0.8198267668485641,
      "num_tokens": 374690746.0,
      "step": 11740
    },
    {
      "entropy": 0.6809972617775202,
      "epoch": 1.0802295558967459,
      "grad_norm": 0.17866069078445435,
      "learning_rate": 6.39954610973104e-05,
      "loss": 0.6794,
      "mean_token_accuracy": 0.7939873933792114,
      "num_tokens": 374722746.0,
      "step": 11741
    },
    {
      "entropy": 0.5035707801580429,
      "epoch": 1.080321562268187,
      "grad_norm": 0.14383529126644135,
      "learning_rate": 6.399239427116877e-05,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8562127761542797,
      "num_tokens": 374755161.0,
      "step": 11742
    },
    {
      "entropy": 0.5770151428878307,
      "epoch": 1.0804135686396283,
      "grad_norm": 0.15496684610843658,
      "learning_rate": 6.398932744502715e-05,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8324104733765125,
      "num_tokens": 374786848.0,
      "step": 11743
    },
    {
      "entropy": 0.7236248850822449,
      "epoch": 1.0805055750110695,
      "grad_norm": 0.16471372544765472,
      "learning_rate": 6.398626061888552e-05,
      "loss": 0.7064,
      "mean_token_accuracy": 0.7812177613377571,
      "num_tokens": 374818379.0,
      "step": 11744
    },
    {
      "entropy": 0.6799155566841364,
      "epoch": 1.0805975813825108,
      "grad_norm": 0.16531991958618164,
      "learning_rate": 6.398319379274388e-05,
      "loss": 0.6445,
      "mean_token_accuracy": 0.801029771566391,
      "num_tokens": 374850171.0,
      "step": 11745
    },
    {
      "entropy": 0.5802870597690344,
      "epoch": 1.080689587753952,
      "grad_norm": 0.16401973366737366,
      "learning_rate": 6.398012696660228e-05,
      "loss": 0.5674,
      "mean_token_accuracy": 0.825612086802721,
      "num_tokens": 374882124.0,
      "step": 11746
    },
    {
      "entropy": 0.5862169098109007,
      "epoch": 1.080781594125393,
      "grad_norm": 0.16241145133972168,
      "learning_rate": 6.397706014046065e-05,
      "loss": 0.5901,
      "mean_token_accuracy": 0.8191198222339153,
      "num_tokens": 374914414.0,
      "step": 11747
    },
    {
      "entropy": 0.5489927642047405,
      "epoch": 1.0808736004968345,
      "grad_norm": 0.14926348626613617,
      "learning_rate": 6.397399331431901e-05,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8329838290810585,
      "num_tokens": 374946377.0,
      "step": 11748
    },
    {
      "entropy": 0.5640841564163566,
      "epoch": 1.0809656068682756,
      "grad_norm": 0.16352137923240662,
      "learning_rate": 6.397092648817738e-05,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8306830860674381,
      "num_tokens": 374978435.0,
      "step": 11749
    },
    {
      "entropy": 0.593018758110702,
      "epoch": 1.081057613239717,
      "grad_norm": 0.15238766372203827,
      "learning_rate": 6.396785966203576e-05,
      "loss": 0.5758,
      "mean_token_accuracy": 0.824846874922514,
      "num_tokens": 375010489.0,
      "step": 11750
    },
    {
      "entropy": 0.651926701888442,
      "epoch": 1.081149619611158,
      "grad_norm": 0.15925391018390656,
      "learning_rate": 6.396479283589414e-05,
      "loss": 0.6397,
      "mean_token_accuracy": 0.8056251220405102,
      "num_tokens": 375042666.0,
      "step": 11751
    },
    {
      "entropy": 0.723782742395997,
      "epoch": 1.0812416259825992,
      "grad_norm": 0.1667809933423996,
      "learning_rate": 6.396172600975251e-05,
      "loss": 0.7168,
      "mean_token_accuracy": 0.7810357362031937,
      "num_tokens": 375074583.0,
      "step": 11752
    },
    {
      "entropy": 0.584253991022706,
      "epoch": 1.0813336323540406,
      "grad_norm": 0.1472121626138687,
      "learning_rate": 6.395865918361088e-05,
      "loss": 0.5614,
      "mean_token_accuracy": 0.8271334990859032,
      "num_tokens": 375107089.0,
      "step": 11753
    },
    {
      "entropy": 0.503829631023109,
      "epoch": 1.0814256387254817,
      "grad_norm": 0.14604830741882324,
      "learning_rate": 6.395559235746926e-05,
      "loss": 0.4839,
      "mean_token_accuracy": 0.8493506237864494,
      "num_tokens": 375139522.0,
      "step": 11754
    },
    {
      "entropy": 0.6669149193912745,
      "epoch": 1.081517645096923,
      "grad_norm": 0.1582387238740921,
      "learning_rate": 6.395252553132763e-05,
      "loss": 0.6544,
      "mean_token_accuracy": 0.8034355640411377,
      "num_tokens": 375171045.0,
      "step": 11755
    },
    {
      "entropy": 0.5165864825248718,
      "epoch": 1.0816096514683642,
      "grad_norm": 0.15395529568195343,
      "learning_rate": 6.394945870518601e-05,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8473911955952644,
      "num_tokens": 375202491.0,
      "step": 11756
    },
    {
      "entropy": 0.559201542288065,
      "epoch": 1.0817016578398053,
      "grad_norm": 0.15521331131458282,
      "learning_rate": 6.394639187904438e-05,
      "loss": 0.5427,
      "mean_token_accuracy": 0.8326383419334888,
      "num_tokens": 375234731.0,
      "step": 11757
    },
    {
      "entropy": 0.6123787835240364,
      "epoch": 1.0817936642112467,
      "grad_norm": 0.16565163433551788,
      "learning_rate": 6.394332505290276e-05,
      "loss": 0.5906,
      "mean_token_accuracy": 0.8179084286093712,
      "num_tokens": 375266878.0,
      "step": 11758
    },
    {
      "entropy": 0.633357607293874,
      "epoch": 1.0818856705826878,
      "grad_norm": 0.16025903820991516,
      "learning_rate": 6.394025822676113e-05,
      "loss": 0.6135,
      "mean_token_accuracy": 0.8124375194311142,
      "num_tokens": 375298181.0,
      "step": 11759
    },
    {
      "entropy": 0.5702516864985228,
      "epoch": 1.0819776769541292,
      "grad_norm": 0.15988892316818237,
      "learning_rate": 6.39371914006195e-05,
      "loss": 0.5553,
      "mean_token_accuracy": 0.8259819149971008,
      "num_tokens": 375330026.0,
      "step": 11760
    },
    {
      "entropy": 0.70073620416224,
      "epoch": 1.0820696833255703,
      "grad_norm": 0.1668975055217743,
      "learning_rate": 6.393412457447788e-05,
      "loss": 0.672,
      "mean_token_accuracy": 0.7962789684534073,
      "num_tokens": 375362085.0,
      "step": 11761
    },
    {
      "entropy": 0.6902743279933929,
      "epoch": 1.0821616896970114,
      "grad_norm": 0.165801540017128,
      "learning_rate": 6.393105774833626e-05,
      "loss": 0.6789,
      "mean_token_accuracy": 0.7910769134759903,
      "num_tokens": 375393639.0,
      "step": 11762
    },
    {
      "entropy": 0.6549540460109711,
      "epoch": 1.0822536960684528,
      "grad_norm": 0.1663365513086319,
      "learning_rate": 6.392799092219463e-05,
      "loss": 0.642,
      "mean_token_accuracy": 0.8004297912120819,
      "num_tokens": 375425178.0,
      "step": 11763
    },
    {
      "entropy": 0.598234080709517,
      "epoch": 1.082345702439894,
      "grad_norm": 0.15080296993255615,
      "learning_rate": 6.3924924096053e-05,
      "loss": 0.585,
      "mean_token_accuracy": 0.818576168268919,
      "num_tokens": 375456347.0,
      "step": 11764
    },
    {
      "entropy": 0.5719208531081676,
      "epoch": 1.0824377088113353,
      "grad_norm": 0.15750490128993988,
      "learning_rate": 6.392185726991136e-05,
      "loss": 0.553,
      "mean_token_accuracy": 0.8261391185224056,
      "num_tokens": 375488106.0,
      "step": 11765
    },
    {
      "entropy": 0.6328635588288307,
      "epoch": 1.0825297151827764,
      "grad_norm": 0.16458599269390106,
      "learning_rate": 6.391879044376974e-05,
      "loss": 0.6354,
      "mean_token_accuracy": 0.8084033206105232,
      "num_tokens": 375520210.0,
      "step": 11766
    },
    {
      "entropy": 0.6358799375593662,
      "epoch": 1.0826217215542175,
      "grad_norm": 0.16993597149848938,
      "learning_rate": 6.391572361762813e-05,
      "loss": 0.6218,
      "mean_token_accuracy": 0.8100812211632729,
      "num_tokens": 375551873.0,
      "step": 11767
    },
    {
      "entropy": 0.6474271975457668,
      "epoch": 1.082713727925659,
      "grad_norm": 0.15808071196079254,
      "learning_rate": 6.39126567914865e-05,
      "loss": 0.6454,
      "mean_token_accuracy": 0.8012251369655132,
      "num_tokens": 375583834.0,
      "step": 11768
    },
    {
      "entropy": 0.5774373169988394,
      "epoch": 1.0828057342971,
      "grad_norm": 0.1522171050310135,
      "learning_rate": 6.390958996534486e-05,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8236134313046932,
      "num_tokens": 375615446.0,
      "step": 11769
    },
    {
      "entropy": 0.5978965517133474,
      "epoch": 1.0828977406685414,
      "grad_norm": 0.15926171839237213,
      "learning_rate": 6.390652313920324e-05,
      "loss": 0.5777,
      "mean_token_accuracy": 0.8208948411047459,
      "num_tokens": 375646698.0,
      "step": 11770
    },
    {
      "entropy": 0.719023184850812,
      "epoch": 1.0829897470399825,
      "grad_norm": 0.17156969010829926,
      "learning_rate": 6.390345631306161e-05,
      "loss": 0.7229,
      "mean_token_accuracy": 0.7815680578351021,
      "num_tokens": 375679083.0,
      "step": 11771
    },
    {
      "entropy": 0.6838039569556713,
      "epoch": 1.0830817534114237,
      "grad_norm": 0.16294178366661072,
      "learning_rate": 6.390038948691999e-05,
      "loss": 0.6475,
      "mean_token_accuracy": 0.7991169206798077,
      "num_tokens": 375710565.0,
      "step": 11772
    },
    {
      "entropy": 0.6562689561396837,
      "epoch": 1.083173759782865,
      "grad_norm": 0.1599910855293274,
      "learning_rate": 6.389732266077836e-05,
      "loss": 0.6465,
      "mean_token_accuracy": 0.8029497303068638,
      "num_tokens": 375742844.0,
      "step": 11773
    },
    {
      "entropy": 0.7388902436941862,
      "epoch": 1.0832657661543061,
      "grad_norm": 0.16091124713420868,
      "learning_rate": 6.389425583463674e-05,
      "loss": 0.7147,
      "mean_token_accuracy": 0.7823551222681999,
      "num_tokens": 375775477.0,
      "step": 11774
    },
    {
      "entropy": 0.5729527734220028,
      "epoch": 1.0833577725257475,
      "grad_norm": 0.15771004557609558,
      "learning_rate": 6.389118900849511e-05,
      "loss": 0.556,
      "mean_token_accuracy": 0.8295835256576538,
      "num_tokens": 375807258.0,
      "step": 11775
    },
    {
      "entropy": 0.5137728303670883,
      "epoch": 1.0834497788971886,
      "grad_norm": 0.1577623039484024,
      "learning_rate": 6.388812218235348e-05,
      "loss": 0.5118,
      "mean_token_accuracy": 0.8413833901286125,
      "num_tokens": 375839908.0,
      "step": 11776
    },
    {
      "entropy": 0.5750700999051332,
      "epoch": 1.0835417852686298,
      "grad_norm": 0.15577207505702972,
      "learning_rate": 6.388505535621186e-05,
      "loss": 0.5725,
      "mean_token_accuracy": 0.8260916620492935,
      "num_tokens": 375871020.0,
      "step": 11777
    },
    {
      "entropy": 0.5857322262600064,
      "epoch": 1.0836337916400711,
      "grad_norm": 0.15482491254806519,
      "learning_rate": 6.388198853007024e-05,
      "loss": 0.5724,
      "mean_token_accuracy": 0.8233047425746918,
      "num_tokens": 375903508.0,
      "step": 11778
    },
    {
      "entropy": 0.5956699112430215,
      "epoch": 1.0837257980115123,
      "grad_norm": 0.16710110008716583,
      "learning_rate": 6.387892170392861e-05,
      "loss": 0.5761,
      "mean_token_accuracy": 0.8207273669540882,
      "num_tokens": 375935219.0,
      "step": 11779
    },
    {
      "entropy": 0.6313351141288877,
      "epoch": 1.0838178043829536,
      "grad_norm": 0.1707102507352829,
      "learning_rate": 6.387585487778698e-05,
      "loss": 0.6143,
      "mean_token_accuracy": 0.8142065443098545,
      "num_tokens": 375967622.0,
      "step": 11780
    },
    {
      "entropy": 0.5764442877843976,
      "epoch": 1.0839098107543947,
      "grad_norm": 0.15347634255886078,
      "learning_rate": 6.387278805164536e-05,
      "loss": 0.5719,
      "mean_token_accuracy": 0.8226444609463215,
      "num_tokens": 375999957.0,
      "step": 11781
    },
    {
      "entropy": 0.6312475106678903,
      "epoch": 1.0840018171258359,
      "grad_norm": 0.16079185903072357,
      "learning_rate": 6.386972122550373e-05,
      "loss": 0.6213,
      "mean_token_accuracy": 0.810955498367548,
      "num_tokens": 376032209.0,
      "step": 11782
    },
    {
      "entropy": 0.6419861521571875,
      "epoch": 1.0840938234972772,
      "grad_norm": 0.16185930371284485,
      "learning_rate": 6.38666543993621e-05,
      "loss": 0.629,
      "mean_token_accuracy": 0.8080579414963722,
      "num_tokens": 376064189.0,
      "step": 11783
    },
    {
      "entropy": 0.5867998069152236,
      "epoch": 1.0841858298687184,
      "grad_norm": 0.1510140746831894,
      "learning_rate": 6.386358757322047e-05,
      "loss": 0.5743,
      "mean_token_accuracy": 0.823955874890089,
      "num_tokens": 376096277.0,
      "step": 11784
    },
    {
      "entropy": 0.6134626567363739,
      "epoch": 1.0842778362401597,
      "grad_norm": 0.16317972540855408,
      "learning_rate": 6.386052074707886e-05,
      "loss": 0.6113,
      "mean_token_accuracy": 0.8155191838741302,
      "num_tokens": 376128705.0,
      "step": 11785
    },
    {
      "entropy": 0.5754716135561466,
      "epoch": 1.0843698426116009,
      "grad_norm": 0.16159048676490784,
      "learning_rate": 6.385745392093722e-05,
      "loss": 0.5689,
      "mean_token_accuracy": 0.8224696554243565,
      "num_tokens": 376160738.0,
      "step": 11786
    },
    {
      "entropy": 0.7537786988541484,
      "epoch": 1.084461848983042,
      "grad_norm": 0.16870161890983582,
      "learning_rate": 6.385438709479559e-05,
      "loss": 0.7502,
      "mean_token_accuracy": 0.7718683890998363,
      "num_tokens": 376192881.0,
      "step": 11787
    },
    {
      "entropy": 0.6943440362811089,
      "epoch": 1.0845538553544833,
      "grad_norm": 0.1617308408021927,
      "learning_rate": 6.385132026865397e-05,
      "loss": 0.6837,
      "mean_token_accuracy": 0.7940441630780697,
      "num_tokens": 376225208.0,
      "step": 11788
    },
    {
      "entropy": 0.6903794705867767,
      "epoch": 1.0846458617259245,
      "grad_norm": 0.16157376766204834,
      "learning_rate": 6.384825344251235e-05,
      "loss": 0.6729,
      "mean_token_accuracy": 0.7952468991279602,
      "num_tokens": 376257151.0,
      "step": 11789
    },
    {
      "entropy": 0.6922485362738371,
      "epoch": 1.0847378680973658,
      "grad_norm": 0.16355061531066895,
      "learning_rate": 6.384518661637072e-05,
      "loss": 0.6782,
      "mean_token_accuracy": 0.7914164401590824,
      "num_tokens": 376288757.0,
      "step": 11790
    },
    {
      "entropy": 0.6380477342754602,
      "epoch": 1.084829874468807,
      "grad_norm": 0.151520773768425,
      "learning_rate": 6.384211979022909e-05,
      "loss": 0.6148,
      "mean_token_accuracy": 0.8103494495153427,
      "num_tokens": 376320860.0,
      "step": 11791
    },
    {
      "entropy": 0.5560551723465323,
      "epoch": 1.084921880840248,
      "grad_norm": 0.15167897939682007,
      "learning_rate": 6.383905296408746e-05,
      "loss": 0.5432,
      "mean_token_accuracy": 0.8329638056457043,
      "num_tokens": 376353194.0,
      "step": 11792
    },
    {
      "entropy": 0.6765514742583036,
      "epoch": 1.0850138872116895,
      "grad_norm": 0.15592429041862488,
      "learning_rate": 6.383598613794585e-05,
      "loss": 0.6683,
      "mean_token_accuracy": 0.7910661920905113,
      "num_tokens": 376385962.0,
      "step": 11793
    },
    {
      "entropy": 0.6914942022413015,
      "epoch": 1.0851058935831306,
      "grad_norm": 0.173335000872612,
      "learning_rate": 6.383291931180422e-05,
      "loss": 0.6717,
      "mean_token_accuracy": 0.7949753701686859,
      "num_tokens": 376417099.0,
      "step": 11794
    },
    {
      "entropy": 0.5509589836001396,
      "epoch": 1.085197899954572,
      "grad_norm": 0.1592782586812973,
      "learning_rate": 6.382985248566259e-05,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8333919607102871,
      "num_tokens": 376449553.0,
      "step": 11795
    },
    {
      "entropy": 0.6295468835160136,
      "epoch": 1.085289906326013,
      "grad_norm": 0.16777212917804718,
      "learning_rate": 6.382678565952096e-05,
      "loss": 0.6238,
      "mean_token_accuracy": 0.8107971623539925,
      "num_tokens": 376481215.0,
      "step": 11796
    },
    {
      "entropy": 0.6963593261316419,
      "epoch": 1.0853819126974542,
      "grad_norm": 0.16792689263820648,
      "learning_rate": 6.382371883337934e-05,
      "loss": 0.6786,
      "mean_token_accuracy": 0.7922318428754807,
      "num_tokens": 376513405.0,
      "step": 11797
    },
    {
      "entropy": 0.595171058550477,
      "epoch": 1.0854739190688956,
      "grad_norm": 0.15940074622631073,
      "learning_rate": 6.382065200723772e-05,
      "loss": 0.5754,
      "mean_token_accuracy": 0.8211839906871319,
      "num_tokens": 376544968.0,
      "step": 11798
    },
    {
      "entropy": 0.6112273498438299,
      "epoch": 1.0855659254403367,
      "grad_norm": 0.15919646620750427,
      "learning_rate": 6.381758518109609e-05,
      "loss": 0.5885,
      "mean_token_accuracy": 0.8215779475867748,
      "num_tokens": 376577189.0,
      "step": 11799
    },
    {
      "entropy": 0.6287319604307413,
      "epoch": 1.085657931811778,
      "grad_norm": 0.16673554480075836,
      "learning_rate": 6.381451835495446e-05,
      "loss": 0.6225,
      "mean_token_accuracy": 0.810223463922739,
      "num_tokens": 376609710.0,
      "step": 11800
    },
    {
      "entropy": 0.6164457481354475,
      "epoch": 1.0857499381832192,
      "grad_norm": 0.1661093831062317,
      "learning_rate": 6.381145152881284e-05,
      "loss": 0.5814,
      "mean_token_accuracy": 0.8213748782873154,
      "num_tokens": 376640952.0,
      "step": 11801
    },
    {
      "entropy": 0.5408950820565224,
      "epoch": 1.0858419445546603,
      "grad_norm": 0.15195204317569733,
      "learning_rate": 6.38083847026712e-05,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8332676589488983,
      "num_tokens": 376673498.0,
      "step": 11802
    },
    {
      "entropy": 0.5958103351294994,
      "epoch": 1.0859339509261017,
      "grad_norm": 0.15839973092079163,
      "learning_rate": 6.380531787652959e-05,
      "loss": 0.5782,
      "mean_token_accuracy": 0.8195355273783207,
      "num_tokens": 376705510.0,
      "step": 11803
    },
    {
      "entropy": 0.5740511221811175,
      "epoch": 1.0860259572975428,
      "grad_norm": 0.15318162739276886,
      "learning_rate": 6.380225105038795e-05,
      "loss": 0.5569,
      "mean_token_accuracy": 0.8239482939243317,
      "num_tokens": 376736914.0,
      "step": 11804
    },
    {
      "entropy": 0.543744646012783,
      "epoch": 1.0861179636689842,
      "grad_norm": 0.15479439496994019,
      "learning_rate": 6.379918422424634e-05,
      "loss": 0.5365,
      "mean_token_accuracy": 0.836836252361536,
      "num_tokens": 376769065.0,
      "step": 11805
    },
    {
      "entropy": 0.7060813158750534,
      "epoch": 1.0862099700404253,
      "grad_norm": 0.16534006595611572,
      "learning_rate": 6.37961173981047e-05,
      "loss": 0.712,
      "mean_token_accuracy": 0.7835715934634209,
      "num_tokens": 376801548.0,
      "step": 11806
    },
    {
      "entropy": 0.6591257154941559,
      "epoch": 1.0863019764118664,
      "grad_norm": 0.1598830223083496,
      "learning_rate": 6.379305057196307e-05,
      "loss": 0.6404,
      "mean_token_accuracy": 0.8013137727975845,
      "num_tokens": 376833249.0,
      "step": 11807
    },
    {
      "entropy": 0.6096173841506243,
      "epoch": 1.0863939827833078,
      "grad_norm": 0.150283545255661,
      "learning_rate": 6.378998374582145e-05,
      "loss": 0.5994,
      "mean_token_accuracy": 0.8144221119582653,
      "num_tokens": 376864941.0,
      "step": 11808
    },
    {
      "entropy": 0.6189433019608259,
      "epoch": 1.086485989154749,
      "grad_norm": 0.1530759483575821,
      "learning_rate": 6.378691691967983e-05,
      "loss": 0.6229,
      "mean_token_accuracy": 0.8097016662359238,
      "num_tokens": 376897352.0,
      "step": 11809
    },
    {
      "entropy": 0.6356941782869399,
      "epoch": 1.0865779955261903,
      "grad_norm": 0.16342733800411224,
      "learning_rate": 6.37838500935382e-05,
      "loss": 0.6154,
      "mean_token_accuracy": 0.812426932156086,
      "num_tokens": 376929494.0,
      "step": 11810
    },
    {
      "entropy": 0.6463294783607125,
      "epoch": 1.0866700018976314,
      "grad_norm": 0.15352314710617065,
      "learning_rate": 6.378078326739657e-05,
      "loss": 0.626,
      "mean_token_accuracy": 0.8107717968523502,
      "num_tokens": 376962152.0,
      "step": 11811
    },
    {
      "entropy": 0.5646679294295609,
      "epoch": 1.0867620082690725,
      "grad_norm": 0.15849809348583221,
      "learning_rate": 6.377771644125495e-05,
      "loss": 0.5387,
      "mean_token_accuracy": 0.8309707082808018,
      "num_tokens": 376994669.0,
      "step": 11812
    },
    {
      "entropy": 0.6231055818498135,
      "epoch": 1.086854014640514,
      "grad_norm": 0.15940357744693756,
      "learning_rate": 6.377464961511332e-05,
      "loss": 0.6116,
      "mean_token_accuracy": 0.8114352598786354,
      "num_tokens": 377026863.0,
      "step": 11813
    },
    {
      "entropy": 0.5599014973267913,
      "epoch": 1.086946021011955,
      "grad_norm": 0.15187931060791016,
      "learning_rate": 6.37715827889717e-05,
      "loss": 0.5561,
      "mean_token_accuracy": 0.826855082064867,
      "num_tokens": 377058765.0,
      "step": 11814
    },
    {
      "entropy": 0.6668716743588448,
      "epoch": 1.0870380273833964,
      "grad_norm": 0.16179227828979492,
      "learning_rate": 6.376851596283007e-05,
      "loss": 0.6506,
      "mean_token_accuracy": 0.8001952692866325,
      "num_tokens": 377090933.0,
      "step": 11815
    },
    {
      "entropy": 0.6488898620009422,
      "epoch": 1.0871300337548375,
      "grad_norm": 0.16214345395565033,
      "learning_rate": 6.376544913668845e-05,
      "loss": 0.6334,
      "mean_token_accuracy": 0.8060245178639889,
      "num_tokens": 377123044.0,
      "step": 11816
    },
    {
      "entropy": 0.582689109724015,
      "epoch": 1.0872220401262787,
      "grad_norm": 0.14868660271167755,
      "learning_rate": 6.376238231054682e-05,
      "loss": 0.5574,
      "mean_token_accuracy": 0.8261349573731422,
      "num_tokens": 377155177.0,
      "step": 11817
    },
    {
      "entropy": 0.654613914899528,
      "epoch": 1.08731404649772,
      "grad_norm": 0.16075226664543152,
      "learning_rate": 6.375931548440519e-05,
      "loss": 0.6496,
      "mean_token_accuracy": 0.8046221174299717,
      "num_tokens": 377186929.0,
      "step": 11818
    },
    {
      "entropy": 0.6399179734289646,
      "epoch": 1.0874060528691611,
      "grad_norm": 0.15997624397277832,
      "learning_rate": 6.375624865826357e-05,
      "loss": 0.6344,
      "mean_token_accuracy": 0.8075433447957039,
      "num_tokens": 377218525.0,
      "step": 11819
    },
    {
      "entropy": 0.5570125784724951,
      "epoch": 1.0874980592406025,
      "grad_norm": 0.15449099242687225,
      "learning_rate": 6.375318183212195e-05,
      "loss": 0.5445,
      "mean_token_accuracy": 0.8329515308141708,
      "num_tokens": 377250922.0,
      "step": 11820
    },
    {
      "entropy": 0.6963294874876738,
      "epoch": 1.0875900656120436,
      "grad_norm": 0.16133925318717957,
      "learning_rate": 6.375011500598032e-05,
      "loss": 0.6949,
      "mean_token_accuracy": 0.791571594774723,
      "num_tokens": 377282793.0,
      "step": 11821
    },
    {
      "entropy": 0.5753876967355609,
      "epoch": 1.0876820719834848,
      "grad_norm": 0.15964502096176147,
      "learning_rate": 6.374704817983868e-05,
      "loss": 0.5566,
      "mean_token_accuracy": 0.8308269456028938,
      "num_tokens": 377314252.0,
      "step": 11822
    },
    {
      "entropy": 0.710295083001256,
      "epoch": 1.0877740783549261,
      "grad_norm": 0.1624717116355896,
      "learning_rate": 6.374398135369705e-05,
      "loss": 0.7057,
      "mean_token_accuracy": 0.7850716784596443,
      "num_tokens": 377346386.0,
      "step": 11823
    },
    {
      "entropy": 0.5638949093408883,
      "epoch": 1.0878660847263673,
      "grad_norm": 0.1534026563167572,
      "learning_rate": 6.374091452755543e-05,
      "loss": 0.5518,
      "mean_token_accuracy": 0.8320102505385876,
      "num_tokens": 377379154.0,
      "step": 11824
    },
    {
      "entropy": 0.6616087984293699,
      "epoch": 1.0879580910978086,
      "grad_norm": 0.15585793554782867,
      "learning_rate": 6.373784770141381e-05,
      "loss": 0.6547,
      "mean_token_accuracy": 0.8000896386802197,
      "num_tokens": 377411720.0,
      "step": 11825
    },
    {
      "entropy": 0.595786145888269,
      "epoch": 1.0880500974692497,
      "grad_norm": 0.1554359793663025,
      "learning_rate": 6.373478087527218e-05,
      "loss": 0.5926,
      "mean_token_accuracy": 0.8173374608159065,
      "num_tokens": 377443695.0,
      "step": 11826
    },
    {
      "entropy": 0.6172572961077094,
      "epoch": 1.0881421038406909,
      "grad_norm": 0.16243292391300201,
      "learning_rate": 6.373171404913055e-05,
      "loss": 0.6062,
      "mean_token_accuracy": 0.814652569591999,
      "num_tokens": 377475888.0,
      "step": 11827
    },
    {
      "entropy": 0.6341392006725073,
      "epoch": 1.0882341102121322,
      "grad_norm": 0.15935252606868744,
      "learning_rate": 6.372864722298893e-05,
      "loss": 0.6313,
      "mean_token_accuracy": 0.8085414431989193,
      "num_tokens": 377508015.0,
      "step": 11828
    },
    {
      "entropy": 0.6114749824628234,
      "epoch": 1.0883261165835734,
      "grad_norm": 0.16626609861850739,
      "learning_rate": 6.372558039684731e-05,
      "loss": 0.5906,
      "mean_token_accuracy": 0.8172632902860641,
      "num_tokens": 377539225.0,
      "step": 11829
    },
    {
      "entropy": 0.5811828393489122,
      "epoch": 1.0884181229550147,
      "grad_norm": 0.15159127116203308,
      "learning_rate": 6.372251357070568e-05,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8246563598513603,
      "num_tokens": 377571178.0,
      "step": 11830
    },
    {
      "entropy": 0.6238956274464726,
      "epoch": 1.0885101293264559,
      "grad_norm": 0.15772631764411926,
      "learning_rate": 6.371944674456405e-05,
      "loss": 0.6029,
      "mean_token_accuracy": 0.8146367929875851,
      "num_tokens": 377603623.0,
      "step": 11831
    },
    {
      "entropy": 0.6585776899009943,
      "epoch": 1.088602135697897,
      "grad_norm": 0.16078349947929382,
      "learning_rate": 6.371637991842243e-05,
      "loss": 0.6449,
      "mean_token_accuracy": 0.8000638112425804,
      "num_tokens": 377634612.0,
      "step": 11832
    },
    {
      "entropy": 0.6242024768143892,
      "epoch": 1.0886941420693383,
      "grad_norm": 0.15967023372650146,
      "learning_rate": 6.37133130922808e-05,
      "loss": 0.6032,
      "mean_token_accuracy": 0.814521599560976,
      "num_tokens": 377666215.0,
      "step": 11833
    },
    {
      "entropy": 0.5499474909156561,
      "epoch": 1.0887861484407795,
      "grad_norm": 0.1524859070777893,
      "learning_rate": 6.371024626613918e-05,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8345196098089218,
      "num_tokens": 377698897.0,
      "step": 11834
    },
    {
      "entropy": 0.6115487404167652,
      "epoch": 1.0888781548122208,
      "grad_norm": 0.15343812108039856,
      "learning_rate": 6.370717943999755e-05,
      "loss": 0.6023,
      "mean_token_accuracy": 0.8147511668503284,
      "num_tokens": 377731124.0,
      "step": 11835
    },
    {
      "entropy": 0.5902906097471714,
      "epoch": 1.088970161183662,
      "grad_norm": 0.16336536407470703,
      "learning_rate": 6.370411261385593e-05,
      "loss": 0.5896,
      "mean_token_accuracy": 0.8209645338356495,
      "num_tokens": 377763259.0,
      "step": 11836
    },
    {
      "entropy": 0.5768775437027216,
      "epoch": 1.089062167555103,
      "grad_norm": 0.16705627739429474,
      "learning_rate": 6.37010457877143e-05,
      "loss": 0.5683,
      "mean_token_accuracy": 0.8276519179344177,
      "num_tokens": 377795629.0,
      "step": 11837
    },
    {
      "entropy": 0.6371710132807493,
      "epoch": 1.0891541739265445,
      "grad_norm": 0.16061976552009583,
      "learning_rate": 6.369797896157266e-05,
      "loss": 0.6241,
      "mean_token_accuracy": 0.8102742210030556,
      "num_tokens": 377827745.0,
      "step": 11838
    },
    {
      "entropy": 0.625950681976974,
      "epoch": 1.0892461802979856,
      "grad_norm": 0.14815214276313782,
      "learning_rate": 6.369491213543105e-05,
      "loss": 0.6121,
      "mean_token_accuracy": 0.8113702908158302,
      "num_tokens": 377860513.0,
      "step": 11839
    },
    {
      "entropy": 0.48567596822977066,
      "epoch": 1.089338186669427,
      "grad_norm": 0.15238462388515472,
      "learning_rate": 6.369184530928943e-05,
      "loss": 0.4836,
      "mean_token_accuracy": 0.8475342690944672,
      "num_tokens": 377891260.0,
      "step": 11840
    },
    {
      "entropy": 0.6768936011940241,
      "epoch": 1.089430193040868,
      "grad_norm": 0.1704729050397873,
      "learning_rate": 6.36887784831478e-05,
      "loss": 0.6646,
      "mean_token_accuracy": 0.7968404032289982,
      "num_tokens": 377922251.0,
      "step": 11841
    },
    {
      "entropy": 0.6482561454176903,
      "epoch": 1.0895221994123092,
      "grad_norm": 0.16142717003822327,
      "learning_rate": 6.368571165700616e-05,
      "loss": 0.6194,
      "mean_token_accuracy": 0.8093739002943039,
      "num_tokens": 377954052.0,
      "step": 11842
    },
    {
      "entropy": 0.6828020848333836,
      "epoch": 1.0896142057837506,
      "grad_norm": 0.15785259008407593,
      "learning_rate": 6.368264483086454e-05,
      "loss": 0.6705,
      "mean_token_accuracy": 0.7946175038814545,
      "num_tokens": 377985983.0,
      "step": 11843
    },
    {
      "entropy": 0.66024349629879,
      "epoch": 1.0897062121551917,
      "grad_norm": 0.16369426250457764,
      "learning_rate": 6.367957800472291e-05,
      "loss": 0.6563,
      "mean_token_accuracy": 0.795457262545824,
      "num_tokens": 378017111.0,
      "step": 11844
    },
    {
      "entropy": 0.6313123181462288,
      "epoch": 1.089798218526633,
      "grad_norm": 0.16134080290794373,
      "learning_rate": 6.36765111785813e-05,
      "loss": 0.6171,
      "mean_token_accuracy": 0.8116951696574688,
      "num_tokens": 378048770.0,
      "step": 11845
    },
    {
      "entropy": 0.6302424734458327,
      "epoch": 1.0898902248980742,
      "grad_norm": 0.1585814654827118,
      "learning_rate": 6.367344435243966e-05,
      "loss": 0.6285,
      "mean_token_accuracy": 0.808411817997694,
      "num_tokens": 378081034.0,
      "step": 11846
    },
    {
      "entropy": 0.6219961559399962,
      "epoch": 1.0899822312695153,
      "grad_norm": 0.1623236984014511,
      "learning_rate": 6.367037752629804e-05,
      "loss": 0.62,
      "mean_token_accuracy": 0.8121599033474922,
      "num_tokens": 378113181.0,
      "step": 11847
    },
    {
      "entropy": 0.6958265285938978,
      "epoch": 1.0900742376409567,
      "grad_norm": 0.15989619493484497,
      "learning_rate": 6.366731070015641e-05,
      "loss": 0.6824,
      "mean_token_accuracy": 0.7943555042147636,
      "num_tokens": 378145544.0,
      "step": 11848
    },
    {
      "entropy": 0.623894952237606,
      "epoch": 1.0901662440123978,
      "grad_norm": 0.1613728553056717,
      "learning_rate": 6.366424387401478e-05,
      "loss": 0.6088,
      "mean_token_accuracy": 0.813900675624609,
      "num_tokens": 378177425.0,
      "step": 11849
    },
    {
      "entropy": 0.5608063414692879,
      "epoch": 1.0902582503838392,
      "grad_norm": 0.1504739224910736,
      "learning_rate": 6.366117704787316e-05,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8286383338272572,
      "num_tokens": 378208741.0,
      "step": 11850
    },
    {
      "entropy": 0.6403994522988796,
      "epoch": 1.0903502567552803,
      "grad_norm": 0.16095352172851562,
      "learning_rate": 6.365811022173154e-05,
      "loss": 0.622,
      "mean_token_accuracy": 0.807411827147007,
      "num_tokens": 378240175.0,
      "step": 11851
    },
    {
      "entropy": 0.5221989210695028,
      "epoch": 1.0904422631267214,
      "grad_norm": 0.14686469733715057,
      "learning_rate": 6.365504339558991e-05,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8402607142925262,
      "num_tokens": 378272465.0,
      "step": 11852
    },
    {
      "entropy": 0.6351076997816563,
      "epoch": 1.0905342694981628,
      "grad_norm": 0.18785202503204346,
      "learning_rate": 6.365197656944828e-05,
      "loss": 0.639,
      "mean_token_accuracy": 0.8071022890508175,
      "num_tokens": 378304275.0,
      "step": 11853
    },
    {
      "entropy": 0.6887276321649551,
      "epoch": 1.090626275869604,
      "grad_norm": 0.16600863635540009,
      "learning_rate": 6.364890974330665e-05,
      "loss": 0.6819,
      "mean_token_accuracy": 0.7903700992465019,
      "num_tokens": 378336228.0,
      "step": 11854
    },
    {
      "entropy": 0.6592941768467426,
      "epoch": 1.0907182822410453,
      "grad_norm": 0.1626938134431839,
      "learning_rate": 6.364584291716503e-05,
      "loss": 0.6451,
      "mean_token_accuracy": 0.8050465658307076,
      "num_tokens": 378368574.0,
      "step": 11855
    },
    {
      "entropy": 0.5970623008906841,
      "epoch": 1.0908102886124864,
      "grad_norm": 0.17228686809539795,
      "learning_rate": 6.364277609102341e-05,
      "loss": 0.5926,
      "mean_token_accuracy": 0.8181871324777603,
      "num_tokens": 378400439.0,
      "step": 11856
    },
    {
      "entropy": 0.7098241131752729,
      "epoch": 1.0909022949839275,
      "grad_norm": 0.16100846230983734,
      "learning_rate": 6.363970926488178e-05,
      "loss": 0.6825,
      "mean_token_accuracy": 0.7906316071748734,
      "num_tokens": 378431807.0,
      "step": 11857
    },
    {
      "entropy": 0.5943564316257834,
      "epoch": 1.090994301355369,
      "grad_norm": 0.15678684413433075,
      "learning_rate": 6.363664243874014e-05,
      "loss": 0.5809,
      "mean_token_accuracy": 0.8255210556089878,
      "num_tokens": 378463177.0,
      "step": 11858
    },
    {
      "entropy": 0.6412585265934467,
      "epoch": 1.09108630772681,
      "grad_norm": 0.16088879108428955,
      "learning_rate": 6.363357561259853e-05,
      "loss": 0.6185,
      "mean_token_accuracy": 0.8156879171729088,
      "num_tokens": 378494781.0,
      "step": 11859
    },
    {
      "entropy": 0.6407788181677461,
      "epoch": 1.0911783140982514,
      "grad_norm": 0.16286110877990723,
      "learning_rate": 6.36305087864569e-05,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8116719760000706,
      "num_tokens": 378526583.0,
      "step": 11860
    },
    {
      "entropy": 0.7015941385179758,
      "epoch": 1.0912703204696925,
      "grad_norm": 0.16328366100788116,
      "learning_rate": 6.362744196031527e-05,
      "loss": 0.6998,
      "mean_token_accuracy": 0.7857847027480602,
      "num_tokens": 378558409.0,
      "step": 11861
    },
    {
      "entropy": 0.6057896791025996,
      "epoch": 1.0913623268411337,
      "grad_norm": 0.1538843810558319,
      "learning_rate": 6.362437513417364e-05,
      "loss": 0.5833,
      "mean_token_accuracy": 0.8179833926260471,
      "num_tokens": 378590489.0,
      "step": 11862
    },
    {
      "entropy": 0.6356091890484095,
      "epoch": 1.091454333212575,
      "grad_norm": 0.1570235639810562,
      "learning_rate": 6.362130830803202e-05,
      "loss": 0.6193,
      "mean_token_accuracy": 0.8117100298404694,
      "num_tokens": 378622472.0,
      "step": 11863
    },
    {
      "entropy": 0.6545014195144176,
      "epoch": 1.0915463395840161,
      "grad_norm": 0.15841738879680634,
      "learning_rate": 6.361824148189039e-05,
      "loss": 0.6402,
      "mean_token_accuracy": 0.8080545961856842,
      "num_tokens": 378654311.0,
      "step": 11864
    },
    {
      "entropy": 0.7073232959955931,
      "epoch": 1.0916383459554575,
      "grad_norm": 0.1680661141872406,
      "learning_rate": 6.361517465574876e-05,
      "loss": 0.6992,
      "mean_token_accuracy": 0.7856518775224686,
      "num_tokens": 378685815.0,
      "step": 11865
    },
    {
      "entropy": 0.5345931760966778,
      "epoch": 1.0917303523268986,
      "grad_norm": 0.15238210558891296,
      "learning_rate": 6.361210782960714e-05,
      "loss": 0.5138,
      "mean_token_accuracy": 0.8388416804373264,
      "num_tokens": 378718126.0,
      "step": 11866
    },
    {
      "entropy": 0.7491922378540039,
      "epoch": 1.0918223586983398,
      "grad_norm": 0.16459356248378754,
      "learning_rate": 6.360904100346552e-05,
      "loss": 0.7484,
      "mean_token_accuracy": 0.7729172259569168,
      "num_tokens": 378749772.0,
      "step": 11867
    },
    {
      "entropy": 0.5692598968744278,
      "epoch": 1.0919143650697811,
      "grad_norm": 0.1507723480463028,
      "learning_rate": 6.360597417732389e-05,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8286955691874027,
      "num_tokens": 378781391.0,
      "step": 11868
    },
    {
      "entropy": 0.6322497744113207,
      "epoch": 1.0920063714412223,
      "grad_norm": 0.16114069521427155,
      "learning_rate": 6.360290735118226e-05,
      "loss": 0.6232,
      "mean_token_accuracy": 0.8088402338325977,
      "num_tokens": 378813440.0,
      "step": 11869
    },
    {
      "entropy": 0.6849557375535369,
      "epoch": 1.0920983778126636,
      "grad_norm": 0.1578359752893448,
      "learning_rate": 6.359984052504064e-05,
      "loss": 0.6842,
      "mean_token_accuracy": 0.790432270616293,
      "num_tokens": 378845674.0,
      "step": 11870
    },
    {
      "entropy": 0.6041748402640224,
      "epoch": 1.0921903841841047,
      "grad_norm": 0.15823747217655182,
      "learning_rate": 6.359677369889902e-05,
      "loss": 0.5856,
      "mean_token_accuracy": 0.8263432011008263,
      "num_tokens": 378877468.0,
      "step": 11871
    },
    {
      "entropy": 0.6527170920744538,
      "epoch": 1.0922823905555459,
      "grad_norm": 0.15922623872756958,
      "learning_rate": 6.359370687275739e-05,
      "loss": 0.6499,
      "mean_token_accuracy": 0.8033493272960186,
      "num_tokens": 378909037.0,
      "step": 11872
    },
    {
      "entropy": 0.6133285984396935,
      "epoch": 1.0923743969269872,
      "grad_norm": 0.16054663062095642,
      "learning_rate": 6.359064004661576e-05,
      "loss": 0.5828,
      "mean_token_accuracy": 0.8199496753513813,
      "num_tokens": 378940553.0,
      "step": 11873
    },
    {
      "entropy": 0.5334213301539421,
      "epoch": 1.0924664032984284,
      "grad_norm": 0.14483703672885895,
      "learning_rate": 6.358757322047414e-05,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8400212377309799,
      "num_tokens": 378972710.0,
      "step": 11874
    },
    {
      "entropy": 0.5739539950154722,
      "epoch": 1.0925584096698697,
      "grad_norm": 0.15711325407028198,
      "learning_rate": 6.35845063943325e-05,
      "loss": 0.5468,
      "mean_token_accuracy": 0.8297875933349133,
      "num_tokens": 379004270.0,
      "step": 11875
    },
    {
      "entropy": 0.6816957900300622,
      "epoch": 1.0926504160413109,
      "grad_norm": 0.1719539612531662,
      "learning_rate": 6.358143956819089e-05,
      "loss": 0.6783,
      "mean_token_accuracy": 0.7912918031215668,
      "num_tokens": 379034613.0,
      "step": 11876
    },
    {
      "entropy": 0.6312663005664945,
      "epoch": 1.092742422412752,
      "grad_norm": 0.16140031814575195,
      "learning_rate": 6.357837274204926e-05,
      "loss": 0.6085,
      "mean_token_accuracy": 0.8158974200487137,
      "num_tokens": 379067179.0,
      "step": 11877
    },
    {
      "entropy": 0.6287515396252275,
      "epoch": 1.0928344287841933,
      "grad_norm": 0.15753239393234253,
      "learning_rate": 6.357530591590764e-05,
      "loss": 0.6053,
      "mean_token_accuracy": 0.8140251971781254,
      "num_tokens": 379098390.0,
      "step": 11878
    },
    {
      "entropy": 0.6936844848096371,
      "epoch": 1.0929264351556345,
      "grad_norm": 0.16414448618888855,
      "learning_rate": 6.3572239089766e-05,
      "loss": 0.6931,
      "mean_token_accuracy": 0.7904758974909782,
      "num_tokens": 379130266.0,
      "step": 11879
    },
    {
      "entropy": 0.6365681039169431,
      "epoch": 1.0930184415270758,
      "grad_norm": 0.16383248567581177,
      "learning_rate": 6.356917226362437e-05,
      "loss": 0.638,
      "mean_token_accuracy": 0.8029492981731892,
      "num_tokens": 379162004.0,
      "step": 11880
    },
    {
      "entropy": 0.6984271183609962,
      "epoch": 1.093110447898517,
      "grad_norm": 0.16711117327213287,
      "learning_rate": 6.356610543748275e-05,
      "loss": 0.6858,
      "mean_token_accuracy": 0.7925179190933704,
      "num_tokens": 379193813.0,
      "step": 11881
    },
    {
      "entropy": 0.5855126492679119,
      "epoch": 1.093202454269958,
      "grad_norm": 0.14848856627941132,
      "learning_rate": 6.356303861134114e-05,
      "loss": 0.564,
      "mean_token_accuracy": 0.8260251767933369,
      "num_tokens": 379226348.0,
      "step": 11882
    },
    {
      "entropy": 0.6863495949655771,
      "epoch": 1.0932944606413995,
      "grad_norm": 0.16180817782878876,
      "learning_rate": 6.35599717851995e-05,
      "loss": 0.6652,
      "mean_token_accuracy": 0.7973889671266079,
      "num_tokens": 379258644.0,
      "step": 11883
    },
    {
      "entropy": 0.5489746038801968,
      "epoch": 1.0933864670128406,
      "grad_norm": 0.15349440276622772,
      "learning_rate": 6.355690495905787e-05,
      "loss": 0.5294,
      "mean_token_accuracy": 0.8369076997041702,
      "num_tokens": 379290313.0,
      "step": 11884
    },
    {
      "entropy": 0.7045743782073259,
      "epoch": 1.093478473384282,
      "grad_norm": 0.16506676375865936,
      "learning_rate": 6.355383813291624e-05,
      "loss": 0.6802,
      "mean_token_accuracy": 0.7882198803126812,
      "num_tokens": 379322554.0,
      "step": 11885
    },
    {
      "entropy": 0.5571859739720821,
      "epoch": 1.093570479755723,
      "grad_norm": 0.15464730560779572,
      "learning_rate": 6.355077130677462e-05,
      "loss": 0.5486,
      "mean_token_accuracy": 0.8340722061693668,
      "num_tokens": 379354605.0,
      "step": 11886
    },
    {
      "entropy": 0.5994292218238115,
      "epoch": 1.0936624861271642,
      "grad_norm": 0.15893878042697906,
      "learning_rate": 6.3547704480633e-05,
      "loss": 0.5871,
      "mean_token_accuracy": 0.8213625699281693,
      "num_tokens": 379386720.0,
      "step": 11887
    },
    {
      "entropy": 0.5660076774656773,
      "epoch": 1.0937544924986056,
      "grad_norm": 0.15512599050998688,
      "learning_rate": 6.354463765449137e-05,
      "loss": 0.5677,
      "mean_token_accuracy": 0.8256422989070415,
      "num_tokens": 379418886.0,
      "step": 11888
    },
    {
      "entropy": 0.6108412016183138,
      "epoch": 1.0938464988700467,
      "grad_norm": 0.1618339866399765,
      "learning_rate": 6.354157082834974e-05,
      "loss": 0.614,
      "mean_token_accuracy": 0.8121086359024048,
      "num_tokens": 379450308.0,
      "step": 11889
    },
    {
      "entropy": 0.7267041895538568,
      "epoch": 1.093938505241488,
      "grad_norm": 0.16621871292591095,
      "learning_rate": 6.353850400220812e-05,
      "loss": 0.7144,
      "mean_token_accuracy": 0.783916037529707,
      "num_tokens": 379482352.0,
      "step": 11890
    },
    {
      "entropy": 0.624447213485837,
      "epoch": 1.0940305116129292,
      "grad_norm": 0.15580452978610992,
      "learning_rate": 6.353543717606649e-05,
      "loss": 0.6132,
      "mean_token_accuracy": 0.8115372993052006,
      "num_tokens": 379513794.0,
      "step": 11891
    },
    {
      "entropy": 0.6038223877549171,
      "epoch": 1.0941225179843703,
      "grad_norm": 0.15664193034172058,
      "learning_rate": 6.353237034992487e-05,
      "loss": 0.5854,
      "mean_token_accuracy": 0.8154805824160576,
      "num_tokens": 379545957.0,
      "step": 11892
    },
    {
      "entropy": 0.7064747326076031,
      "epoch": 1.0942145243558117,
      "grad_norm": 0.1612911820411682,
      "learning_rate": 6.352930352378324e-05,
      "loss": 0.6911,
      "mean_token_accuracy": 0.7874097228050232,
      "num_tokens": 379577931.0,
      "step": 11893
    },
    {
      "entropy": 0.5516957920044661,
      "epoch": 1.0943065307272528,
      "grad_norm": 0.15884675085544586,
      "learning_rate": 6.352623669764162e-05,
      "loss": 0.5387,
      "mean_token_accuracy": 0.8342841863632202,
      "num_tokens": 379610635.0,
      "step": 11894
    },
    {
      "entropy": 0.5521543910726905,
      "epoch": 1.0943985370986942,
      "grad_norm": 0.16161219775676727,
      "learning_rate": 6.352316987149999e-05,
      "loss": 0.5425,
      "mean_token_accuracy": 0.8341159448027611,
      "num_tokens": 379642804.0,
      "step": 11895
    },
    {
      "entropy": 0.6960644358769059,
      "epoch": 1.0944905434701353,
      "grad_norm": 0.1627708226442337,
      "learning_rate": 6.352010304535835e-05,
      "loss": 0.6753,
      "mean_token_accuracy": 0.7909561395645142,
      "num_tokens": 379674422.0,
      "step": 11896
    },
    {
      "entropy": 0.6857427936047316,
      "epoch": 1.0945825498415764,
      "grad_norm": 0.16451624035835266,
      "learning_rate": 6.351703621921673e-05,
      "loss": 0.6739,
      "mean_token_accuracy": 0.7908727452158928,
      "num_tokens": 379706043.0,
      "step": 11897
    },
    {
      "entropy": 0.6903490703552961,
      "epoch": 1.0946745562130178,
      "grad_norm": 0.1691029816865921,
      "learning_rate": 6.351396939307512e-05,
      "loss": 0.6904,
      "mean_token_accuracy": 0.7908220030367374,
      "num_tokens": 379737657.0,
      "step": 11898
    },
    {
      "entropy": 0.571671093814075,
      "epoch": 1.094766562584459,
      "grad_norm": 0.15447935461997986,
      "learning_rate": 6.351090256693348e-05,
      "loss": 0.5612,
      "mean_token_accuracy": 0.8282539881765842,
      "num_tokens": 379770110.0,
      "step": 11899
    },
    {
      "entropy": 0.5005371966399252,
      "epoch": 1.0948585689559003,
      "grad_norm": 0.14355427026748657,
      "learning_rate": 6.350783574079185e-05,
      "loss": 0.4773,
      "mean_token_accuracy": 0.8529961593449116,
      "num_tokens": 379802659.0,
      "step": 11900
    },
    {
      "entropy": 0.6480695400387049,
      "epoch": 1.0949505753273414,
      "grad_norm": 0.16121533513069153,
      "learning_rate": 6.350476891465023e-05,
      "loss": 0.6431,
      "mean_token_accuracy": 0.8069467656314373,
      "num_tokens": 379834158.0,
      "step": 11901
    },
    {
      "entropy": 0.682750079780817,
      "epoch": 1.0950425816987825,
      "grad_norm": 0.16380590200424194,
      "learning_rate": 6.35017020885086e-05,
      "loss": 0.6832,
      "mean_token_accuracy": 0.7971929013729095,
      "num_tokens": 379866475.0,
      "step": 11902
    },
    {
      "entropy": 0.5542347654700279,
      "epoch": 1.095134588070224,
      "grad_norm": 0.14995677769184113,
      "learning_rate": 6.349863526236698e-05,
      "loss": 0.5355,
      "mean_token_accuracy": 0.8333236984908581,
      "num_tokens": 379898249.0,
      "step": 11903
    },
    {
      "entropy": 0.7697122450917959,
      "epoch": 1.095226594441665,
      "grad_norm": 0.16773483157157898,
      "learning_rate": 6.349556843622535e-05,
      "loss": 0.7668,
      "mean_token_accuracy": 0.7720367833971977,
      "num_tokens": 379930354.0,
      "step": 11904
    },
    {
      "entropy": 0.573528666049242,
      "epoch": 1.0953186008131064,
      "grad_norm": 0.15291699767112732,
      "learning_rate": 6.349250161008373e-05,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8280670791864395,
      "num_tokens": 379962523.0,
      "step": 11905
    },
    {
      "entropy": 0.54703782312572,
      "epoch": 1.0954106071845475,
      "grad_norm": 0.1627333015203476,
      "learning_rate": 6.34894347839421e-05,
      "loss": 0.539,
      "mean_token_accuracy": 0.8358115553855896,
      "num_tokens": 379994885.0,
      "step": 11906
    },
    {
      "entropy": 0.6639720872044563,
      "epoch": 1.0955026135559887,
      "grad_norm": 0.1630607396364212,
      "learning_rate": 6.348636795780047e-05,
      "loss": 0.6285,
      "mean_token_accuracy": 0.8077680803835392,
      "num_tokens": 380026875.0,
      "step": 11907
    },
    {
      "entropy": 0.7446363605558872,
      "epoch": 1.09559461992743,
      "grad_norm": 0.16203485429286957,
      "learning_rate": 6.348330113165885e-05,
      "loss": 0.7108,
      "mean_token_accuracy": 0.781511727720499,
      "num_tokens": 380058921.0,
      "step": 11908
    },
    {
      "entropy": 0.619807327631861,
      "epoch": 1.0956866262988711,
      "grad_norm": 0.15737125277519226,
      "learning_rate": 6.348023430551723e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.8116484619677067,
      "num_tokens": 380090674.0,
      "step": 11909
    },
    {
      "entropy": 0.4731813268736005,
      "epoch": 1.0957786326703125,
      "grad_norm": 0.15189902484416962,
      "learning_rate": 6.34771674793756e-05,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8579883277416229,
      "num_tokens": 380123013.0,
      "step": 11910
    },
    {
      "entropy": 0.6631059646606445,
      "epoch": 1.0958706390417536,
      "grad_norm": 0.16551554203033447,
      "learning_rate": 6.347410065323397e-05,
      "loss": 0.6595,
      "mean_token_accuracy": 0.8003782071173191,
      "num_tokens": 380155201.0,
      "step": 11911
    },
    {
      "entropy": 0.6518605723977089,
      "epoch": 1.0959626454131948,
      "grad_norm": 0.16248954832553864,
      "learning_rate": 6.347103382709233e-05,
      "loss": 0.6416,
      "mean_token_accuracy": 0.8038158304989338,
      "num_tokens": 380186862.0,
      "step": 11912
    },
    {
      "entropy": 0.5858313236385584,
      "epoch": 1.0960546517846361,
      "grad_norm": 0.15565459430217743,
      "learning_rate": 6.346796700095073e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8248612321913242,
      "num_tokens": 380218460.0,
      "step": 11913
    },
    {
      "entropy": 0.5003202706575394,
      "epoch": 1.0961466581560773,
      "grad_norm": 0.15193094313144684,
      "learning_rate": 6.34649001748091e-05,
      "loss": 0.4906,
      "mean_token_accuracy": 0.8449903354048729,
      "num_tokens": 380249646.0,
      "step": 11914
    },
    {
      "entropy": 0.5564770624041557,
      "epoch": 1.0962386645275186,
      "grad_norm": 0.15088704228401184,
      "learning_rate": 6.346183334866747e-05,
      "loss": 0.5299,
      "mean_token_accuracy": 0.8355957008898258,
      "num_tokens": 380281728.0,
      "step": 11915
    },
    {
      "entropy": 0.6279805079102516,
      "epoch": 1.0963306708989597,
      "grad_norm": 0.15901923179626465,
      "learning_rate": 6.345876652252583e-05,
      "loss": 0.6178,
      "mean_token_accuracy": 0.812587209045887,
      "num_tokens": 380313833.0,
      "step": 11916
    },
    {
      "entropy": 0.6072670286521316,
      "epoch": 1.0964226772704009,
      "grad_norm": 0.15542295575141907,
      "learning_rate": 6.345569969638421e-05,
      "loss": 0.6014,
      "mean_token_accuracy": 0.8137932792305946,
      "num_tokens": 380344958.0,
      "step": 11917
    },
    {
      "entropy": 0.6041550878435373,
      "epoch": 1.0965146836418422,
      "grad_norm": 0.15693417191505432,
      "learning_rate": 6.34526328702426e-05,
      "loss": 0.5847,
      "mean_token_accuracy": 0.8227629661560059,
      "num_tokens": 380376924.0,
      "step": 11918
    },
    {
      "entropy": 0.5667641870677471,
      "epoch": 1.0966066900132834,
      "grad_norm": 0.16103875637054443,
      "learning_rate": 6.344956604410096e-05,
      "loss": 0.5381,
      "mean_token_accuracy": 0.8330463729798794,
      "num_tokens": 380408693.0,
      "step": 11919
    },
    {
      "entropy": 0.603081788867712,
      "epoch": 1.0966986963847247,
      "grad_norm": 0.16133926808834076,
      "learning_rate": 6.344649921795933e-05,
      "loss": 0.6005,
      "mean_token_accuracy": 0.815698079764843,
      "num_tokens": 380441005.0,
      "step": 11920
    },
    {
      "entropy": 0.6703418409451842,
      "epoch": 1.0967907027561659,
      "grad_norm": 0.16668850183486938,
      "learning_rate": 6.344343239181771e-05,
      "loss": 0.6515,
      "mean_token_accuracy": 0.7969161048531532,
      "num_tokens": 380472804.0,
      "step": 11921
    },
    {
      "entropy": 0.5311009967699647,
      "epoch": 1.096882709127607,
      "grad_norm": 0.15422064065933228,
      "learning_rate": 6.344036556567608e-05,
      "loss": 0.514,
      "mean_token_accuracy": 0.8387220241129398,
      "num_tokens": 380504786.0,
      "step": 11922
    },
    {
      "entropy": 0.588913245126605,
      "epoch": 1.0969747154990483,
      "grad_norm": 0.15502120554447174,
      "learning_rate": 6.343729873953446e-05,
      "loss": 0.5805,
      "mean_token_accuracy": 0.8271403945982456,
      "num_tokens": 380537377.0,
      "step": 11923
    },
    {
      "entropy": 0.7171002868562937,
      "epoch": 1.0970667218704895,
      "grad_norm": 0.1582155078649521,
      "learning_rate": 6.343423191339283e-05,
      "loss": 0.6997,
      "mean_token_accuracy": 0.7899828962981701,
      "num_tokens": 380569718.0,
      "step": 11924
    },
    {
      "entropy": 0.6979609187692404,
      "epoch": 1.0971587282419308,
      "grad_norm": 0.16427826881408691,
      "learning_rate": 6.343116508725121e-05,
      "loss": 0.6905,
      "mean_token_accuracy": 0.7904232703149319,
      "num_tokens": 380601255.0,
      "step": 11925
    },
    {
      "entropy": 0.6573767345398664,
      "epoch": 1.097250734613372,
      "grad_norm": 0.16941416263580322,
      "learning_rate": 6.342809826110958e-05,
      "loss": 0.6397,
      "mean_token_accuracy": 0.8011911362409592,
      "num_tokens": 380632831.0,
      "step": 11926
    },
    {
      "entropy": 0.6188094457611442,
      "epoch": 1.097342740984813,
      "grad_norm": 0.1624142974615097,
      "learning_rate": 6.342503143496795e-05,
      "loss": 0.6143,
      "mean_token_accuracy": 0.811731856316328,
      "num_tokens": 380665105.0,
      "step": 11927
    },
    {
      "entropy": 0.8326420709490776,
      "epoch": 1.0974347473562545,
      "grad_norm": 0.16903331875801086,
      "learning_rate": 6.342196460882633e-05,
      "loss": 0.8308,
      "mean_token_accuracy": 0.7535650096833706,
      "num_tokens": 380697180.0,
      "step": 11928
    },
    {
      "entropy": 0.6229967633262277,
      "epoch": 1.0975267537276956,
      "grad_norm": 0.15552392601966858,
      "learning_rate": 6.341889778268471e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.813365500420332,
      "num_tokens": 380728257.0,
      "step": 11929
    },
    {
      "entropy": 0.7006345577538013,
      "epoch": 1.097618760099137,
      "grad_norm": 0.16479505598545074,
      "learning_rate": 6.341583095654308e-05,
      "loss": 0.6939,
      "mean_token_accuracy": 0.7903308942914009,
      "num_tokens": 380759940.0,
      "step": 11930
    },
    {
      "entropy": 0.6507134847342968,
      "epoch": 1.097710766470578,
      "grad_norm": 0.15576288104057312,
      "learning_rate": 6.341276413040145e-05,
      "loss": 0.6419,
      "mean_token_accuracy": 0.8047078065574169,
      "num_tokens": 380792013.0,
      "step": 11931
    },
    {
      "entropy": 0.5872362591326237,
      "epoch": 1.0978027728420192,
      "grad_norm": 0.16941745579242706,
      "learning_rate": 6.340969730425983e-05,
      "loss": 0.5828,
      "mean_token_accuracy": 0.8229987509548664,
      "num_tokens": 380823409.0,
      "step": 11932
    },
    {
      "entropy": 0.6890586353838444,
      "epoch": 1.0978947792134606,
      "grad_norm": 0.17416314780712128,
      "learning_rate": 6.34066304781182e-05,
      "loss": 0.677,
      "mean_token_accuracy": 0.7941144444048405,
      "num_tokens": 380854788.0,
      "step": 11933
    },
    {
      "entropy": 0.7644546665251255,
      "epoch": 1.0979867855849017,
      "grad_norm": 0.16856995224952698,
      "learning_rate": 6.340356365197658e-05,
      "loss": 0.7454,
      "mean_token_accuracy": 0.7745496071875095,
      "num_tokens": 380886975.0,
      "step": 11934
    },
    {
      "entropy": 0.6490277796983719,
      "epoch": 1.098078791956343,
      "grad_norm": 0.16346299648284912,
      "learning_rate": 6.340049682583494e-05,
      "loss": 0.6491,
      "mean_token_accuracy": 0.8026683926582336,
      "num_tokens": 380918924.0,
      "step": 11935
    },
    {
      "entropy": 0.542554852552712,
      "epoch": 1.0981707983277842,
      "grad_norm": 0.15981094539165497,
      "learning_rate": 6.339742999969333e-05,
      "loss": 0.5327,
      "mean_token_accuracy": 0.837152224034071,
      "num_tokens": 380950819.0,
      "step": 11936
    },
    {
      "entropy": 0.6335855768993497,
      "epoch": 1.0982628046992253,
      "grad_norm": 0.15793505311012268,
      "learning_rate": 6.33943631735517e-05,
      "loss": 0.6277,
      "mean_token_accuracy": 0.8090160898864269,
      "num_tokens": 380983174.0,
      "step": 11937
    },
    {
      "entropy": 0.708765484392643,
      "epoch": 1.0983548110706667,
      "grad_norm": 0.16256099939346313,
      "learning_rate": 6.339129634741006e-05,
      "loss": 0.7045,
      "mean_token_accuracy": 0.7873159572482109,
      "num_tokens": 381014450.0,
      "step": 11938
    },
    {
      "entropy": 0.5869412049651146,
      "epoch": 1.0984468174421078,
      "grad_norm": 0.15997889637947083,
      "learning_rate": 6.338822952126844e-05,
      "loss": 0.572,
      "mean_token_accuracy": 0.8239392302930355,
      "num_tokens": 381045122.0,
      "step": 11939
    },
    {
      "entropy": 0.628399464301765,
      "epoch": 1.0985388238135492,
      "grad_norm": 0.15890677273273468,
      "learning_rate": 6.338516269512682e-05,
      "loss": 0.6092,
      "mean_token_accuracy": 0.811918817460537,
      "num_tokens": 381076882.0,
      "step": 11940
    },
    {
      "entropy": 0.6837815158069134,
      "epoch": 1.0986308301849903,
      "grad_norm": 0.1632758527994156,
      "learning_rate": 6.338209586898519e-05,
      "loss": 0.6656,
      "mean_token_accuracy": 0.8014768958091736,
      "num_tokens": 381109174.0,
      "step": 11941
    },
    {
      "entropy": 0.568936213850975,
      "epoch": 1.0987228365564314,
      "grad_norm": 0.15907713770866394,
      "learning_rate": 6.337902904284356e-05,
      "loss": 0.554,
      "mean_token_accuracy": 0.8282151706516743,
      "num_tokens": 381141611.0,
      "step": 11942
    },
    {
      "entropy": 0.6307819448411465,
      "epoch": 1.0988148429278728,
      "grad_norm": 0.15843963623046875,
      "learning_rate": 6.337596221670193e-05,
      "loss": 0.6035,
      "mean_token_accuracy": 0.8143111281096935,
      "num_tokens": 381173022.0,
      "step": 11943
    },
    {
      "entropy": 0.6631115898489952,
      "epoch": 1.098906849299314,
      "grad_norm": 0.16678956151008606,
      "learning_rate": 6.337289539056031e-05,
      "loss": 0.6536,
      "mean_token_accuracy": 0.8045660257339478,
      "num_tokens": 381203971.0,
      "step": 11944
    },
    {
      "entropy": 0.6286183726042509,
      "epoch": 1.0989988556707553,
      "grad_norm": 0.1566656231880188,
      "learning_rate": 6.336982856441869e-05,
      "loss": 0.6151,
      "mean_token_accuracy": 0.8071400374174118,
      "num_tokens": 381235485.0,
      "step": 11945
    },
    {
      "entropy": 0.6711976109072566,
      "epoch": 1.0990908620421964,
      "grad_norm": 0.16222146153450012,
      "learning_rate": 6.336676173827706e-05,
      "loss": 0.6531,
      "mean_token_accuracy": 0.8032673485577106,
      "num_tokens": 381267699.0,
      "step": 11946
    },
    {
      "entropy": 0.6371101979166269,
      "epoch": 1.0991828684136375,
      "grad_norm": 0.1563735157251358,
      "learning_rate": 6.336369491213543e-05,
      "loss": 0.6293,
      "mean_token_accuracy": 0.8034535050392151,
      "num_tokens": 381300110.0,
      "step": 11947
    },
    {
      "entropy": 0.6191577352583408,
      "epoch": 1.099274874785079,
      "grad_norm": 0.17147879302501678,
      "learning_rate": 6.336062808599381e-05,
      "loss": 0.6026,
      "mean_token_accuracy": 0.8132533207535744,
      "num_tokens": 381331583.0,
      "step": 11948
    },
    {
      "entropy": 0.6173294028267264,
      "epoch": 1.09936688115652,
      "grad_norm": 0.1576484590768814,
      "learning_rate": 6.335756125985219e-05,
      "loss": 0.6031,
      "mean_token_accuracy": 0.8149766698479652,
      "num_tokens": 381363224.0,
      "step": 11949
    },
    {
      "entropy": 0.6226806100457907,
      "epoch": 1.0994588875279614,
      "grad_norm": 0.14877699315547943,
      "learning_rate": 6.335449443371056e-05,
      "loss": 0.6181,
      "mean_token_accuracy": 0.8086972497403622,
      "num_tokens": 381395877.0,
      "step": 11950
    },
    {
      "entropy": 0.4571898225694895,
      "epoch": 1.0995508938994025,
      "grad_norm": 0.1599913090467453,
      "learning_rate": 6.335142760756893e-05,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8645159490406513,
      "num_tokens": 381428187.0,
      "step": 11951
    },
    {
      "entropy": 0.630360608920455,
      "epoch": 1.0996429002708437,
      "grad_norm": 0.16002051532268524,
      "learning_rate": 6.33483607814273e-05,
      "loss": 0.6119,
      "mean_token_accuracy": 0.808208454400301,
      "num_tokens": 381460215.0,
      "step": 11952
    },
    {
      "entropy": 0.6757314652204514,
      "epoch": 1.099734906642285,
      "grad_norm": 0.15840883553028107,
      "learning_rate": 6.334529395528567e-05,
      "loss": 0.6679,
      "mean_token_accuracy": 0.7953606210649014,
      "num_tokens": 381492695.0,
      "step": 11953
    },
    {
      "entropy": 0.6195656713098288,
      "epoch": 1.0998269130137261,
      "grad_norm": 0.15680500864982605,
      "learning_rate": 6.334222712914406e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.8105012588202953,
      "num_tokens": 381525363.0,
      "step": 11954
    },
    {
      "entropy": 0.791356410831213,
      "epoch": 1.0999189193851675,
      "grad_norm": 0.17451339960098267,
      "learning_rate": 6.333916030300242e-05,
      "loss": 0.7783,
      "mean_token_accuracy": 0.7643240690231323,
      "num_tokens": 381556429.0,
      "step": 11955
    },
    {
      "entropy": 0.6430069906637073,
      "epoch": 1.1000109257566086,
      "grad_norm": 0.1677086502313614,
      "learning_rate": 6.33360934768608e-05,
      "loss": 0.6234,
      "mean_token_accuracy": 0.8096400983631611,
      "num_tokens": 381588348.0,
      "step": 11956
    },
    {
      "entropy": 0.594147227704525,
      "epoch": 1.1001029321280498,
      "grad_norm": 0.15782730281352997,
      "learning_rate": 6.333302665071917e-05,
      "loss": 0.5719,
      "mean_token_accuracy": 0.8182265236973763,
      "num_tokens": 381619908.0,
      "step": 11957
    },
    {
      "entropy": 0.6506288843229413,
      "epoch": 1.1001949384994911,
      "grad_norm": 0.1632167249917984,
      "learning_rate": 6.332995982457754e-05,
      "loss": 0.6325,
      "mean_token_accuracy": 0.8108272813260555,
      "num_tokens": 381652147.0,
      "step": 11958
    },
    {
      "entropy": 0.586180504411459,
      "epoch": 1.1002869448709323,
      "grad_norm": 0.1586458683013916,
      "learning_rate": 6.332689299843592e-05,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8243829756975174,
      "num_tokens": 381682857.0,
      "step": 11959
    },
    {
      "entropy": 0.6522474158555269,
      "epoch": 1.1003789512423736,
      "grad_norm": 0.15678897500038147,
      "learning_rate": 6.33238261722943e-05,
      "loss": 0.6392,
      "mean_token_accuracy": 0.803611233830452,
      "num_tokens": 381714784.0,
      "step": 11960
    },
    {
      "entropy": 0.5447798287495971,
      "epoch": 1.1004709576138147,
      "grad_norm": 0.1668032556772232,
      "learning_rate": 6.332075934615267e-05,
      "loss": 0.5348,
      "mean_token_accuracy": 0.8351547345519066,
      "num_tokens": 381746834.0,
      "step": 11961
    },
    {
      "entropy": 0.6587655022740364,
      "epoch": 1.1005629639852559,
      "grad_norm": 0.16587460041046143,
      "learning_rate": 6.331769252001104e-05,
      "loss": 0.6653,
      "mean_token_accuracy": 0.7994803078472614,
      "num_tokens": 381779096.0,
      "step": 11962
    },
    {
      "entropy": 0.5222017737105489,
      "epoch": 1.1006549703566972,
      "grad_norm": 0.15956899523735046,
      "learning_rate": 6.331462569386942e-05,
      "loss": 0.5029,
      "mean_token_accuracy": 0.8452509231865406,
      "num_tokens": 381811006.0,
      "step": 11963
    },
    {
      "entropy": 0.6545507432892919,
      "epoch": 1.1007469767281384,
      "grad_norm": 0.1564396470785141,
      "learning_rate": 6.331155886772779e-05,
      "loss": 0.6398,
      "mean_token_accuracy": 0.8037366680800915,
      "num_tokens": 381843183.0,
      "step": 11964
    },
    {
      "entropy": 0.6679178923368454,
      "epoch": 1.1008389830995797,
      "grad_norm": 0.16202862560749054,
      "learning_rate": 6.330849204158617e-05,
      "loss": 0.6647,
      "mean_token_accuracy": 0.798310961574316,
      "num_tokens": 381874374.0,
      "step": 11965
    },
    {
      "entropy": 0.6865989421494305,
      "epoch": 1.1009309894710209,
      "grad_norm": 0.1588037610054016,
      "learning_rate": 6.330542521544454e-05,
      "loss": 0.6805,
      "mean_token_accuracy": 0.7994368523359299,
      "num_tokens": 381906392.0,
      "step": 11966
    },
    {
      "entropy": 0.7967787981033325,
      "epoch": 1.101022995842462,
      "grad_norm": 0.1654234528541565,
      "learning_rate": 6.330235838930292e-05,
      "loss": 0.8002,
      "mean_token_accuracy": 0.7629780098795891,
      "num_tokens": 381938530.0,
      "step": 11967
    },
    {
      "entropy": 0.6995074637234211,
      "epoch": 1.1011150022139033,
      "grad_norm": 0.1583458036184311,
      "learning_rate": 6.329929156316129e-05,
      "loss": 0.6901,
      "mean_token_accuracy": 0.7887689024209976,
      "num_tokens": 381970562.0,
      "step": 11968
    },
    {
      "entropy": 0.6859185509383678,
      "epoch": 1.1012070085853445,
      "grad_norm": 0.16259123384952545,
      "learning_rate": 6.329622473701966e-05,
      "loss": 0.687,
      "mean_token_accuracy": 0.7919819094240665,
      "num_tokens": 382001960.0,
      "step": 11969
    },
    {
      "entropy": 0.721027310937643,
      "epoch": 1.1012990149567858,
      "grad_norm": 0.1599176526069641,
      "learning_rate": 6.329315791087804e-05,
      "loss": 0.7018,
      "mean_token_accuracy": 0.7831134870648384,
      "num_tokens": 382034210.0,
      "step": 11970
    },
    {
      "entropy": 0.5695859137922525,
      "epoch": 1.101391021328227,
      "grad_norm": 0.1586872786283493,
      "learning_rate": 6.329009108473642e-05,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8345566801726818,
      "num_tokens": 382066654.0,
      "step": 11971
    },
    {
      "entropy": 0.5998999401926994,
      "epoch": 1.101483027699668,
      "grad_norm": 0.16278602182865143,
      "learning_rate": 6.328702425859479e-05,
      "loss": 0.5892,
      "mean_token_accuracy": 0.8229323662817478,
      "num_tokens": 382099237.0,
      "step": 11972
    },
    {
      "entropy": 0.6706593707203865,
      "epoch": 1.1015750340711095,
      "grad_norm": 0.16380295157432556,
      "learning_rate": 6.328395743245315e-05,
      "loss": 0.6433,
      "mean_token_accuracy": 0.806949183344841,
      "num_tokens": 382131522.0,
      "step": 11973
    },
    {
      "entropy": 0.5711197732016444,
      "epoch": 1.1016670404425506,
      "grad_norm": 0.15031611919403076,
      "learning_rate": 6.328089060631152e-05,
      "loss": 0.5574,
      "mean_token_accuracy": 0.8254853449761868,
      "num_tokens": 382163985.0,
      "step": 11974
    },
    {
      "entropy": 0.6687174532562494,
      "epoch": 1.101759046813992,
      "grad_norm": 0.16317394375801086,
      "learning_rate": 6.32778237801699e-05,
      "loss": 0.6375,
      "mean_token_accuracy": 0.8033119663596153,
      "num_tokens": 382195717.0,
      "step": 11975
    },
    {
      "entropy": 0.7160703279078007,
      "epoch": 1.101851053185433,
      "grad_norm": 0.15577921271324158,
      "learning_rate": 6.327475695402828e-05,
      "loss": 0.6903,
      "mean_token_accuracy": 0.7890835590660572,
      "num_tokens": 382227732.0,
      "step": 11976
    },
    {
      "entropy": 0.520350469276309,
      "epoch": 1.1019430595568742,
      "grad_norm": 0.15034909546375275,
      "learning_rate": 6.327169012788665e-05,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8396522663533688,
      "num_tokens": 382260270.0,
      "step": 11977
    },
    {
      "entropy": 0.627936314791441,
      "epoch": 1.1020350659283156,
      "grad_norm": 0.1531105488538742,
      "learning_rate": 6.326862330174502e-05,
      "loss": 0.6072,
      "mean_token_accuracy": 0.8124030530452728,
      "num_tokens": 382292011.0,
      "step": 11978
    },
    {
      "entropy": 0.5766855692490935,
      "epoch": 1.1021270722997567,
      "grad_norm": 0.15559548139572144,
      "learning_rate": 6.32655564756034e-05,
      "loss": 0.554,
      "mean_token_accuracy": 0.8248195685446262,
      "num_tokens": 382323030.0,
      "step": 11979
    },
    {
      "entropy": 0.58954325504601,
      "epoch": 1.102219078671198,
      "grad_norm": 0.15747672319412231,
      "learning_rate": 6.326248964946177e-05,
      "loss": 0.5898,
      "mean_token_accuracy": 0.8143748603761196,
      "num_tokens": 382354424.0,
      "step": 11980
    },
    {
      "entropy": 0.5682789254933596,
      "epoch": 1.1023110850426392,
      "grad_norm": 0.1492757946252823,
      "learning_rate": 6.325942282332015e-05,
      "loss": 0.5569,
      "mean_token_accuracy": 0.8241319172084332,
      "num_tokens": 382385905.0,
      "step": 11981
    },
    {
      "entropy": 0.5835724920034409,
      "epoch": 1.1024030914140803,
      "grad_norm": 0.160687655210495,
      "learning_rate": 6.325635599717852e-05,
      "loss": 0.5693,
      "mean_token_accuracy": 0.8245355524122715,
      "num_tokens": 382418162.0,
      "step": 11982
    },
    {
      "entropy": 0.5771518899127841,
      "epoch": 1.1024950977855217,
      "grad_norm": 0.17331644892692566,
      "learning_rate": 6.32532891710369e-05,
      "loss": 0.5639,
      "mean_token_accuracy": 0.8248582482337952,
      "num_tokens": 382449937.0,
      "step": 11983
    },
    {
      "entropy": 0.6031269682571292,
      "epoch": 1.1025871041569628,
      "grad_norm": 0.15962396562099457,
      "learning_rate": 6.325022234489527e-05,
      "loss": 0.5979,
      "mean_token_accuracy": 0.8146620728075504,
      "num_tokens": 382482540.0,
      "step": 11984
    },
    {
      "entropy": 0.6717560794204473,
      "epoch": 1.1026791105284042,
      "grad_norm": 0.16206161677837372,
      "learning_rate": 6.324715551875364e-05,
      "loss": 0.6582,
      "mean_token_accuracy": 0.7978860847651958,
      "num_tokens": 382514603.0,
      "step": 11985
    },
    {
      "entropy": 0.6009036041796207,
      "epoch": 1.1027711168998453,
      "grad_norm": 0.1585412621498108,
      "learning_rate": 6.324408869261202e-05,
      "loss": 0.5925,
      "mean_token_accuracy": 0.8179527893662453,
      "num_tokens": 382546082.0,
      "step": 11986
    },
    {
      "entropy": 0.6890031443908811,
      "epoch": 1.1028631232712864,
      "grad_norm": 0.16296343505382538,
      "learning_rate": 6.32410218664704e-05,
      "loss": 0.6788,
      "mean_token_accuracy": 0.7952912747859955,
      "num_tokens": 382578475.0,
      "step": 11987
    },
    {
      "entropy": 0.5952843073755503,
      "epoch": 1.1029551296427278,
      "grad_norm": 0.1603439897298813,
      "learning_rate": 6.323795504032877e-05,
      "loss": 0.5849,
      "mean_token_accuracy": 0.8173780180513859,
      "num_tokens": 382610096.0,
      "step": 11988
    },
    {
      "entropy": 0.5689821811392903,
      "epoch": 1.103047136014169,
      "grad_norm": 0.15598522126674652,
      "learning_rate": 6.323488821418713e-05,
      "loss": 0.554,
      "mean_token_accuracy": 0.8271706104278564,
      "num_tokens": 382641970.0,
      "step": 11989
    },
    {
      "entropy": 0.6854412276297808,
      "epoch": 1.1031391423856103,
      "grad_norm": 0.17028091847896576,
      "learning_rate": 6.323182138804552e-05,
      "loss": 0.6727,
      "mean_token_accuracy": 0.7930016592144966,
      "num_tokens": 382673252.0,
      "step": 11990
    },
    {
      "entropy": 0.6088428245857358,
      "epoch": 1.1032311487570514,
      "grad_norm": 0.16098040342330933,
      "learning_rate": 6.32287545619039e-05,
      "loss": 0.6026,
      "mean_token_accuracy": 0.8148465901613235,
      "num_tokens": 382704948.0,
      "step": 11991
    },
    {
      "entropy": 0.6013772636651993,
      "epoch": 1.1033231551284925,
      "grad_norm": 0.15744563937187195,
      "learning_rate": 6.322568773576227e-05,
      "loss": 0.5959,
      "mean_token_accuracy": 0.8200106769800186,
      "num_tokens": 382737144.0,
      "step": 11992
    },
    {
      "entropy": 0.6751821972429752,
      "epoch": 1.103415161499934,
      "grad_norm": 0.16724829375743866,
      "learning_rate": 6.322262090962063e-05,
      "loss": 0.6633,
      "mean_token_accuracy": 0.7977272048592567,
      "num_tokens": 382769424.0,
      "step": 11993
    },
    {
      "entropy": 0.5878638164140284,
      "epoch": 1.103507167871375,
      "grad_norm": 0.16226014494895935,
      "learning_rate": 6.321955408347901e-05,
      "loss": 0.5756,
      "mean_token_accuracy": 0.8262227773666382,
      "num_tokens": 382801588.0,
      "step": 11994
    },
    {
      "entropy": 0.7246158961206675,
      "epoch": 1.1035991742428164,
      "grad_norm": 0.17125514149665833,
      "learning_rate": 6.321648725733738e-05,
      "loss": 0.712,
      "mean_token_accuracy": 0.7883660085499287,
      "num_tokens": 382833664.0,
      "step": 11995
    },
    {
      "entropy": 0.7271316573023796,
      "epoch": 1.1036911806142575,
      "grad_norm": 0.17550769448280334,
      "learning_rate": 6.321342043119576e-05,
      "loss": 0.719,
      "mean_token_accuracy": 0.7800336070358753,
      "num_tokens": 382864504.0,
      "step": 11996
    },
    {
      "entropy": 0.5514501570723951,
      "epoch": 1.1037831869856987,
      "grad_norm": 0.1550539880990982,
      "learning_rate": 6.321035360505413e-05,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8344113454222679,
      "num_tokens": 382896820.0,
      "step": 11997
    },
    {
      "entropy": 0.6465871091932058,
      "epoch": 1.10387519335714,
      "grad_norm": 0.16534429788589478,
      "learning_rate": 6.320728677891251e-05,
      "loss": 0.6318,
      "mean_token_accuracy": 0.8025892004370689,
      "num_tokens": 382927918.0,
      "step": 11998
    },
    {
      "entropy": 0.6107212156057358,
      "epoch": 1.1039671997285811,
      "grad_norm": 0.15835008025169373,
      "learning_rate": 6.320421995277088e-05,
      "loss": 0.5948,
      "mean_token_accuracy": 0.8161421529948711,
      "num_tokens": 382959738.0,
      "step": 11999
    },
    {
      "entropy": 0.5931877382099628,
      "epoch": 1.1040592061000225,
      "grad_norm": 0.15515421330928802,
      "learning_rate": 6.320115312662925e-05,
      "loss": 0.5693,
      "mean_token_accuracy": 0.8228791356086731,
      "num_tokens": 382992079.0,
      "step": 12000
    },
    {
      "entropy": 0.6094473283737898,
      "epoch": 1.1041512124714636,
      "grad_norm": 0.16292496025562286,
      "learning_rate": 6.319808630048763e-05,
      "loss": 0.5945,
      "mean_token_accuracy": 0.8226769156754017,
      "num_tokens": 383024661.0,
      "step": 12001
    },
    {
      "entropy": 0.6634793616831303,
      "epoch": 1.1042432188429048,
      "grad_norm": 0.15795069932937622,
      "learning_rate": 6.319501947434601e-05,
      "loss": 0.6425,
      "mean_token_accuracy": 0.8029317706823349,
      "num_tokens": 383057068.0,
      "step": 12002
    },
    {
      "entropy": 0.6376544609665871,
      "epoch": 1.1043352252143461,
      "grad_norm": 0.1588907092809677,
      "learning_rate": 6.319195264820438e-05,
      "loss": 0.6273,
      "mean_token_accuracy": 0.8094801492989063,
      "num_tokens": 383089272.0,
      "step": 12003
    },
    {
      "entropy": 0.5416455622762442,
      "epoch": 1.1044272315857873,
      "grad_norm": 0.15834888815879822,
      "learning_rate": 6.318888582206275e-05,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8383303321897984,
      "num_tokens": 383120982.0,
      "step": 12004
    },
    {
      "entropy": 0.5758941103704274,
      "epoch": 1.1045192379572286,
      "grad_norm": 0.15816035866737366,
      "learning_rate": 6.318581899592112e-05,
      "loss": 0.561,
      "mean_token_accuracy": 0.8268836252391338,
      "num_tokens": 383153463.0,
      "step": 12005
    },
    {
      "entropy": 0.5869226567447186,
      "epoch": 1.1046112443286698,
      "grad_norm": 0.1621537059545517,
      "learning_rate": 6.31827521697795e-05,
      "loss": 0.5612,
      "mean_token_accuracy": 0.826030895113945,
      "num_tokens": 383185408.0,
      "step": 12006
    },
    {
      "entropy": 0.5731405122205615,
      "epoch": 1.1047032507001109,
      "grad_norm": 0.15770556032657623,
      "learning_rate": 6.317968534363788e-05,
      "loss": 0.5562,
      "mean_token_accuracy": 0.829423077404499,
      "num_tokens": 383217114.0,
      "step": 12007
    },
    {
      "entropy": 0.7190863359719515,
      "epoch": 1.1047952570715522,
      "grad_norm": 0.17109417915344238,
      "learning_rate": 6.317661851749625e-05,
      "loss": 0.7156,
      "mean_token_accuracy": 0.7829580716788769,
      "num_tokens": 383248402.0,
      "step": 12008
    },
    {
      "entropy": 0.5466786315664649,
      "epoch": 1.1048872634429934,
      "grad_norm": 0.14585916697978973,
      "learning_rate": 6.317355169135461e-05,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8320934809744358,
      "num_tokens": 383279870.0,
      "step": 12009
    },
    {
      "entropy": 0.6163487443700433,
      "epoch": 1.1049792698144347,
      "grad_norm": 0.16105104982852936,
      "learning_rate": 6.3170484865213e-05,
      "loss": 0.6207,
      "mean_token_accuracy": 0.8124691024422646,
      "num_tokens": 383312010.0,
      "step": 12010
    },
    {
      "entropy": 0.4792287051677704,
      "epoch": 1.1050712761858759,
      "grad_norm": 0.15534374117851257,
      "learning_rate": 6.316741803907136e-05,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8504989221692085,
      "num_tokens": 383343525.0,
      "step": 12011
    },
    {
      "entropy": 0.5634260438382626,
      "epoch": 1.105163282557317,
      "grad_norm": 0.16159643232822418,
      "learning_rate": 6.316435121292974e-05,
      "loss": 0.5568,
      "mean_token_accuracy": 0.827612716704607,
      "num_tokens": 383375313.0,
      "step": 12012
    },
    {
      "entropy": 0.6184499431401491,
      "epoch": 1.1052552889287584,
      "grad_norm": 0.15579114854335785,
      "learning_rate": 6.316128438678811e-05,
      "loss": 0.6204,
      "mean_token_accuracy": 0.8122921325266361,
      "num_tokens": 383407591.0,
      "step": 12013
    },
    {
      "entropy": 0.6359606646001339,
      "epoch": 1.1053472953001995,
      "grad_norm": 0.1558522880077362,
      "learning_rate": 6.31582175606465e-05,
      "loss": 0.608,
      "mean_token_accuracy": 0.8139837309718132,
      "num_tokens": 383439347.0,
      "step": 12014
    },
    {
      "entropy": 0.6228372287005186,
      "epoch": 1.1054393016716408,
      "grad_norm": 0.1675879806280136,
      "learning_rate": 6.315515073450486e-05,
      "loss": 0.6175,
      "mean_token_accuracy": 0.8073762468993664,
      "num_tokens": 383471846.0,
      "step": 12015
    },
    {
      "entropy": 0.6497380696237087,
      "epoch": 1.105531308043082,
      "grad_norm": 0.16210828721523285,
      "learning_rate": 6.315208390836323e-05,
      "loss": 0.6289,
      "mean_token_accuracy": 0.8060955815017223,
      "num_tokens": 383503628.0,
      "step": 12016
    },
    {
      "entropy": 0.5616650767624378,
      "epoch": 1.105623314414523,
      "grad_norm": 0.14947886765003204,
      "learning_rate": 6.314901708222161e-05,
      "loss": 0.5419,
      "mean_token_accuracy": 0.8297236263751984,
      "num_tokens": 383535217.0,
      "step": 12017
    },
    {
      "entropy": 0.521446275524795,
      "epoch": 1.1057153207859645,
      "grad_norm": 0.14871874451637268,
      "learning_rate": 6.314595025607999e-05,
      "loss": 0.515,
      "mean_token_accuracy": 0.8470627814531326,
      "num_tokens": 383567985.0,
      "step": 12018
    },
    {
      "entropy": 0.5831917012110353,
      "epoch": 1.1058073271574056,
      "grad_norm": 0.15154601633548737,
      "learning_rate": 6.314288342993836e-05,
      "loss": 0.5665,
      "mean_token_accuracy": 0.8252966180443764,
      "num_tokens": 383599727.0,
      "step": 12019
    },
    {
      "entropy": 0.6270369291305542,
      "epoch": 1.105899333528847,
      "grad_norm": 0.1628337949514389,
      "learning_rate": 6.313981660379673e-05,
      "loss": 0.6018,
      "mean_token_accuracy": 0.8165284357964993,
      "num_tokens": 383631824.0,
      "step": 12020
    },
    {
      "entropy": 0.6122035942971706,
      "epoch": 1.105991339900288,
      "grad_norm": 0.16548223793506622,
      "learning_rate": 6.313674977765511e-05,
      "loss": 0.5983,
      "mean_token_accuracy": 0.8123471476137638,
      "num_tokens": 383663174.0,
      "step": 12021
    },
    {
      "entropy": 0.634098244830966,
      "epoch": 1.1060833462717292,
      "grad_norm": 0.15254315733909607,
      "learning_rate": 6.313368295151348e-05,
      "loss": 0.6152,
      "mean_token_accuracy": 0.8106791637837887,
      "num_tokens": 383695543.0,
      "step": 12022
    },
    {
      "entropy": 0.5666808634996414,
      "epoch": 1.1061753526431706,
      "grad_norm": 0.15094567835330963,
      "learning_rate": 6.313061612537186e-05,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8299771957099438,
      "num_tokens": 383727049.0,
      "step": 12023
    },
    {
      "entropy": 0.6612866334617138,
      "epoch": 1.1062673590146117,
      "grad_norm": 0.1594214141368866,
      "learning_rate": 6.312754929923023e-05,
      "loss": 0.6425,
      "mean_token_accuracy": 0.8053527288138866,
      "num_tokens": 383759768.0,
      "step": 12024
    },
    {
      "entropy": 0.5612829886376858,
      "epoch": 1.106359365386053,
      "grad_norm": 0.15662670135498047,
      "learning_rate": 6.312448247308861e-05,
      "loss": 0.549,
      "mean_token_accuracy": 0.8318228349089622,
      "num_tokens": 383791320.0,
      "step": 12025
    },
    {
      "entropy": 0.6920330030843616,
      "epoch": 1.1064513717574942,
      "grad_norm": 0.16984252631664276,
      "learning_rate": 6.312141564694698e-05,
      "loss": 0.6573,
      "mean_token_accuracy": 0.8036328256130219,
      "num_tokens": 383822869.0,
      "step": 12026
    },
    {
      "entropy": 0.6836907491087914,
      "epoch": 1.1065433781289353,
      "grad_norm": 0.16968178749084473,
      "learning_rate": 6.311834882080534e-05,
      "loss": 0.6672,
      "mean_token_accuracy": 0.7990001924335957,
      "num_tokens": 383854050.0,
      "step": 12027
    },
    {
      "entropy": 0.6578105939552188,
      "epoch": 1.1066353845003767,
      "grad_norm": 0.1600983589887619,
      "learning_rate": 6.311528199466373e-05,
      "loss": 0.6363,
      "mean_token_accuracy": 0.8074147403240204,
      "num_tokens": 383886326.0,
      "step": 12028
    },
    {
      "entropy": 0.6586362197995186,
      "epoch": 1.1067273908718178,
      "grad_norm": 0.16297981142997742,
      "learning_rate": 6.311221516852211e-05,
      "loss": 0.6473,
      "mean_token_accuracy": 0.8048132210969925,
      "num_tokens": 383918382.0,
      "step": 12029
    },
    {
      "entropy": 0.6342771155759692,
      "epoch": 1.1068193972432592,
      "grad_norm": 0.15569394826889038,
      "learning_rate": 6.310914834238047e-05,
      "loss": 0.6259,
      "mean_token_accuracy": 0.8054095506668091,
      "num_tokens": 383951080.0,
      "step": 12030
    },
    {
      "entropy": 0.5376173630356789,
      "epoch": 1.1069114036147003,
      "grad_norm": 0.15627673268318176,
      "learning_rate": 6.310608151623884e-05,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8374940678477287,
      "num_tokens": 383983710.0,
      "step": 12031
    },
    {
      "entropy": 0.5390506722033024,
      "epoch": 1.1070034099861414,
      "grad_norm": 0.154277503490448,
      "learning_rate": 6.310301469009721e-05,
      "loss": 0.525,
      "mean_token_accuracy": 0.8363459333777428,
      "num_tokens": 384015217.0,
      "step": 12032
    },
    {
      "entropy": 0.6866044532507658,
      "epoch": 1.1070954163575828,
      "grad_norm": 0.1676957905292511,
      "learning_rate": 6.30999478639556e-05,
      "loss": 0.6742,
      "mean_token_accuracy": 0.7978359200060368,
      "num_tokens": 384047529.0,
      "step": 12033
    },
    {
      "entropy": 0.8098901025950909,
      "epoch": 1.107187422729024,
      "grad_norm": 0.17234432697296143,
      "learning_rate": 6.309688103781397e-05,
      "loss": 0.814,
      "mean_token_accuracy": 0.7552516683936119,
      "num_tokens": 384079634.0,
      "step": 12034
    },
    {
      "entropy": 0.5860696099698544,
      "epoch": 1.1072794291004653,
      "grad_norm": 0.1615850031375885,
      "learning_rate": 6.309381421167234e-05,
      "loss": 0.5695,
      "mean_token_accuracy": 0.8246820792555809,
      "num_tokens": 384111600.0,
      "step": 12035
    },
    {
      "entropy": 0.5879472447559237,
      "epoch": 1.1073714354719064,
      "grad_norm": 0.1563267707824707,
      "learning_rate": 6.309074738553071e-05,
      "loss": 0.5956,
      "mean_token_accuracy": 0.817957628518343,
      "num_tokens": 384142864.0,
      "step": 12036
    },
    {
      "entropy": 0.6467847560998052,
      "epoch": 1.1074634418433476,
      "grad_norm": 0.14972388744354248,
      "learning_rate": 6.308768055938909e-05,
      "loss": 0.629,
      "mean_token_accuracy": 0.8058601655066013,
      "num_tokens": 384174910.0,
      "step": 12037
    },
    {
      "entropy": 0.6174280717968941,
      "epoch": 1.107555448214789,
      "grad_norm": 0.16824491322040558,
      "learning_rate": 6.308461373324747e-05,
      "loss": 0.601,
      "mean_token_accuracy": 0.8205983266234398,
      "num_tokens": 384207075.0,
      "step": 12038
    },
    {
      "entropy": 0.5935301668941975,
      "epoch": 1.10764745458623,
      "grad_norm": 0.15920902788639069,
      "learning_rate": 6.308154690710584e-05,
      "loss": 0.5781,
      "mean_token_accuracy": 0.8200115971267223,
      "num_tokens": 384237643.0,
      "step": 12039
    },
    {
      "entropy": 0.6400418179109693,
      "epoch": 1.1077394609576714,
      "grad_norm": 0.1610008180141449,
      "learning_rate": 6.307848008096421e-05,
      "loss": 0.6301,
      "mean_token_accuracy": 0.8065466955304146,
      "num_tokens": 384269897.0,
      "step": 12040
    },
    {
      "entropy": 0.6438258457928896,
      "epoch": 1.1078314673291125,
      "grad_norm": 0.15998513996601105,
      "learning_rate": 6.307541325482259e-05,
      "loss": 0.6279,
      "mean_token_accuracy": 0.8082853630185127,
      "num_tokens": 384301933.0,
      "step": 12041
    },
    {
      "entropy": 0.6157977469265461,
      "epoch": 1.1079234737005537,
      "grad_norm": 0.1536858081817627,
      "learning_rate": 6.307234642868096e-05,
      "loss": 0.6005,
      "mean_token_accuracy": 0.8104097284376621,
      "num_tokens": 384334184.0,
      "step": 12042
    },
    {
      "entropy": 0.5102575048804283,
      "epoch": 1.108015480071995,
      "grad_norm": 0.15157130360603333,
      "learning_rate": 6.306927960253934e-05,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8468540459871292,
      "num_tokens": 384366385.0,
      "step": 12043
    },
    {
      "entropy": 0.6354872509837151,
      "epoch": 1.1081074864434362,
      "grad_norm": 0.15808744728565216,
      "learning_rate": 6.30662127763977e-05,
      "loss": 0.6181,
      "mean_token_accuracy": 0.80919548869133,
      "num_tokens": 384398113.0,
      "step": 12044
    },
    {
      "entropy": 0.6714009754359722,
      "epoch": 1.1081994928148775,
      "grad_norm": 0.16778218746185303,
      "learning_rate": 6.306314595025609e-05,
      "loss": 0.6487,
      "mean_token_accuracy": 0.7991227395832539,
      "num_tokens": 384429285.0,
      "step": 12045
    },
    {
      "entropy": 0.6176077537238598,
      "epoch": 1.1082914991863186,
      "grad_norm": 0.1582685112953186,
      "learning_rate": 6.306007912411446e-05,
      "loss": 0.5954,
      "mean_token_accuracy": 0.8169086761772633,
      "num_tokens": 384461300.0,
      "step": 12046
    },
    {
      "entropy": 0.7011028490960598,
      "epoch": 1.1083835055577598,
      "grad_norm": 0.16058394312858582,
      "learning_rate": 6.305701229797282e-05,
      "loss": 0.6925,
      "mean_token_accuracy": 0.7841862663626671,
      "num_tokens": 384492985.0,
      "step": 12047
    },
    {
      "entropy": 0.5950198192149401,
      "epoch": 1.1084755119292011,
      "grad_norm": 0.1590910702943802,
      "learning_rate": 6.30539454718312e-05,
      "loss": 0.5765,
      "mean_token_accuracy": 0.8234287537634373,
      "num_tokens": 384524605.0,
      "step": 12048
    },
    {
      "entropy": 0.6356653682887554,
      "epoch": 1.1085675183006423,
      "grad_norm": 0.1606191247701645,
      "learning_rate": 6.305087864568959e-05,
      "loss": 0.6306,
      "mean_token_accuracy": 0.8040268942713737,
      "num_tokens": 384556859.0,
      "step": 12049
    },
    {
      "entropy": 0.6857656296342611,
      "epoch": 1.1086595246720836,
      "grad_norm": 0.17119748890399933,
      "learning_rate": 6.304781181954795e-05,
      "loss": 0.6875,
      "mean_token_accuracy": 0.790591225028038,
      "num_tokens": 384588413.0,
      "step": 12050
    },
    {
      "entropy": 0.5832212101668119,
      "epoch": 1.1087515310435248,
      "grad_norm": 0.16215600073337555,
      "learning_rate": 6.304474499340632e-05,
      "loss": 0.5729,
      "mean_token_accuracy": 0.8214516714215279,
      "num_tokens": 384620101.0,
      "step": 12051
    },
    {
      "entropy": 0.5890526305884123,
      "epoch": 1.1088435374149659,
      "grad_norm": 0.14780138432979584,
      "learning_rate": 6.30416781672647e-05,
      "loss": 0.5795,
      "mean_token_accuracy": 0.8232779689133167,
      "num_tokens": 384652869.0,
      "step": 12052
    },
    {
      "entropy": 0.6441827863454819,
      "epoch": 1.1089355437864072,
      "grad_norm": 0.15569570660591125,
      "learning_rate": 6.303861134112307e-05,
      "loss": 0.6308,
      "mean_token_accuracy": 0.8067709356546402,
      "num_tokens": 384685100.0,
      "step": 12053
    },
    {
      "entropy": 0.5635609570890665,
      "epoch": 1.1090275501578484,
      "grad_norm": 0.15884777903556824,
      "learning_rate": 6.303554451498145e-05,
      "loss": 0.5635,
      "mean_token_accuracy": 0.8309917040169239,
      "num_tokens": 384716757.0,
      "step": 12054
    },
    {
      "entropy": 0.6347595984116197,
      "epoch": 1.1091195565292897,
      "grad_norm": 0.16154800355434418,
      "learning_rate": 6.303247768883982e-05,
      "loss": 0.6229,
      "mean_token_accuracy": 0.808841023594141,
      "num_tokens": 384747956.0,
      "step": 12055
    },
    {
      "entropy": 0.5751896044239402,
      "epoch": 1.1092115629007309,
      "grad_norm": 0.1590966433286667,
      "learning_rate": 6.30294108626982e-05,
      "loss": 0.5567,
      "mean_token_accuracy": 0.8329343982040882,
      "num_tokens": 384780374.0,
      "step": 12056
    },
    {
      "entropy": 0.6791312582790852,
      "epoch": 1.109303569272172,
      "grad_norm": 0.15807975828647614,
      "learning_rate": 6.302634403655657e-05,
      "loss": 0.6627,
      "mean_token_accuracy": 0.8023855537176132,
      "num_tokens": 384812671.0,
      "step": 12057
    },
    {
      "entropy": 0.6411737520247698,
      "epoch": 1.1093955756436134,
      "grad_norm": 0.16682231426239014,
      "learning_rate": 6.302327721041494e-05,
      "loss": 0.6242,
      "mean_token_accuracy": 0.8070715256035328,
      "num_tokens": 384843865.0,
      "step": 12058
    },
    {
      "entropy": 0.7089442908763885,
      "epoch": 1.1094875820150545,
      "grad_norm": 0.15862254798412323,
      "learning_rate": 6.302021038427332e-05,
      "loss": 0.6886,
      "mean_token_accuracy": 0.7916867621243,
      "num_tokens": 384875765.0,
      "step": 12059
    },
    {
      "entropy": 0.608628480695188,
      "epoch": 1.1095795883864958,
      "grad_norm": 0.15795181691646576,
      "learning_rate": 6.30171435581317e-05,
      "loss": 0.604,
      "mean_token_accuracy": 0.8168815039098263,
      "num_tokens": 384907664.0,
      "step": 12060
    },
    {
      "entropy": 0.5973829217255116,
      "epoch": 1.109671594757937,
      "grad_norm": 0.15960852801799774,
      "learning_rate": 6.301407673199007e-05,
      "loss": 0.596,
      "mean_token_accuracy": 0.817391712218523,
      "num_tokens": 384940167.0,
      "step": 12061
    },
    {
      "entropy": 0.6763542108237743,
      "epoch": 1.109763601129378,
      "grad_norm": 0.1610650271177292,
      "learning_rate": 6.301100990584844e-05,
      "loss": 0.6545,
      "mean_token_accuracy": 0.7973170392215252,
      "num_tokens": 384971494.0,
      "step": 12062
    },
    {
      "entropy": 0.6185367219150066,
      "epoch": 1.1098556075008195,
      "grad_norm": 0.16985516250133514,
      "learning_rate": 6.30079430797068e-05,
      "loss": 0.6103,
      "mean_token_accuracy": 0.8093392848968506,
      "num_tokens": 385002866.0,
      "step": 12063
    },
    {
      "entropy": 0.6328936573117971,
      "epoch": 1.1099476138722606,
      "grad_norm": 0.15534666180610657,
      "learning_rate": 6.300487625356519e-05,
      "loss": 0.611,
      "mean_token_accuracy": 0.8115195259451866,
      "num_tokens": 385034976.0,
      "step": 12064
    },
    {
      "entropy": 0.5021874122321606,
      "epoch": 1.110039620243702,
      "grad_norm": 0.15914025902748108,
      "learning_rate": 6.300180942742357e-05,
      "loss": 0.4812,
      "mean_token_accuracy": 0.8445672281086445,
      "num_tokens": 385066781.0,
      "step": 12065
    },
    {
      "entropy": 0.5888270568102598,
      "epoch": 1.110131626615143,
      "grad_norm": 0.1593174785375595,
      "learning_rate": 6.299874260128194e-05,
      "loss": 0.5886,
      "mean_token_accuracy": 0.8180723711848259,
      "num_tokens": 385097489.0,
      "step": 12066
    },
    {
      "entropy": 0.5332597168162465,
      "epoch": 1.1102236329865842,
      "grad_norm": 0.15605047345161438,
      "learning_rate": 6.29956757751403e-05,
      "loss": 0.51,
      "mean_token_accuracy": 0.8395263999700546,
      "num_tokens": 385128876.0,
      "step": 12067
    },
    {
      "entropy": 0.5677589206025004,
      "epoch": 1.1103156393580256,
      "grad_norm": 0.15770789980888367,
      "learning_rate": 6.299260894899868e-05,
      "loss": 0.5587,
      "mean_token_accuracy": 0.8278328068554401,
      "num_tokens": 385160900.0,
      "step": 12068
    },
    {
      "entropy": 0.6055639870464802,
      "epoch": 1.1104076457294667,
      "grad_norm": 0.1557881087064743,
      "learning_rate": 6.298954212285707e-05,
      "loss": 0.5836,
      "mean_token_accuracy": 0.8241402618587017,
      "num_tokens": 385192959.0,
      "step": 12069
    },
    {
      "entropy": 0.6115036141127348,
      "epoch": 1.110499652100908,
      "grad_norm": 0.1516178399324417,
      "learning_rate": 6.298647529671543e-05,
      "loss": 0.5775,
      "mean_token_accuracy": 0.8222318068146706,
      "num_tokens": 385224536.0,
      "step": 12070
    },
    {
      "entropy": 0.5980748003348708,
      "epoch": 1.1105916584723492,
      "grad_norm": 0.15914016962051392,
      "learning_rate": 6.29834084705738e-05,
      "loss": 0.5856,
      "mean_token_accuracy": 0.8204961195588112,
      "num_tokens": 385256190.0,
      "step": 12071
    },
    {
      "entropy": 0.6639385856688023,
      "epoch": 1.1106836648437903,
      "grad_norm": 0.17071960866451263,
      "learning_rate": 6.298034164443218e-05,
      "loss": 0.6518,
      "mean_token_accuracy": 0.8006869219243526,
      "num_tokens": 385287698.0,
      "step": 12072
    },
    {
      "entropy": 0.5065830959938467,
      "epoch": 1.1107756712152317,
      "grad_norm": 0.15250568091869354,
      "learning_rate": 6.297727481829055e-05,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8461786471307278,
      "num_tokens": 385318748.0,
      "step": 12073
    },
    {
      "entropy": 0.6243943981826305,
      "epoch": 1.1108676775866728,
      "grad_norm": 0.1636541485786438,
      "learning_rate": 6.297420799214893e-05,
      "loss": 0.6215,
      "mean_token_accuracy": 0.8069710396230221,
      "num_tokens": 385350675.0,
      "step": 12074
    },
    {
      "entropy": 0.5877062268555164,
      "epoch": 1.1109596839581142,
      "grad_norm": 0.15758270025253296,
      "learning_rate": 6.29711411660073e-05,
      "loss": 0.5656,
      "mean_token_accuracy": 0.8242346234619617,
      "num_tokens": 385381797.0,
      "step": 12075
    },
    {
      "entropy": 0.7365130633115768,
      "epoch": 1.1110516903295553,
      "grad_norm": 0.16390351951122284,
      "learning_rate": 6.296807433986568e-05,
      "loss": 0.7175,
      "mean_token_accuracy": 0.7788845226168633,
      "num_tokens": 385413528.0,
      "step": 12076
    },
    {
      "entropy": 0.6805877778679132,
      "epoch": 1.1111436967009964,
      "grad_norm": 0.16608503460884094,
      "learning_rate": 6.296500751372405e-05,
      "loss": 0.6708,
      "mean_token_accuracy": 0.7958868220448494,
      "num_tokens": 385445327.0,
      "step": 12077
    },
    {
      "entropy": 0.6072870269417763,
      "epoch": 1.1112357030724378,
      "grad_norm": 0.16156841814517975,
      "learning_rate": 6.296194068758242e-05,
      "loss": 0.602,
      "mean_token_accuracy": 0.8164609670639038,
      "num_tokens": 385477541.0,
      "step": 12078
    },
    {
      "entropy": 0.5547613725066185,
      "epoch": 1.111327709443879,
      "grad_norm": 0.15913015604019165,
      "learning_rate": 6.29588738614408e-05,
      "loss": 0.5363,
      "mean_token_accuracy": 0.8373364247381687,
      "num_tokens": 385509264.0,
      "step": 12079
    },
    {
      "entropy": 0.745368167757988,
      "epoch": 1.1114197158153203,
      "grad_norm": 0.18012483417987823,
      "learning_rate": 6.295580703529918e-05,
      "loss": 0.7398,
      "mean_token_accuracy": 0.7756605222821236,
      "num_tokens": 385540722.0,
      "step": 12080
    },
    {
      "entropy": 0.634011797606945,
      "epoch": 1.1115117221867614,
      "grad_norm": 0.15897272527217865,
      "learning_rate": 6.295274020915755e-05,
      "loss": 0.6227,
      "mean_token_accuracy": 0.8134532012045383,
      "num_tokens": 385572362.0,
      "step": 12081
    },
    {
      "entropy": 0.6445634625852108,
      "epoch": 1.1116037285582026,
      "grad_norm": 0.1582200676202774,
      "learning_rate": 6.294967338301592e-05,
      "loss": 0.6337,
      "mean_token_accuracy": 0.8045937344431877,
      "num_tokens": 385604165.0,
      "step": 12082
    },
    {
      "entropy": 0.5892518889158964,
      "epoch": 1.111695734929644,
      "grad_norm": 0.16232439875602722,
      "learning_rate": 6.29466065568743e-05,
      "loss": 0.5813,
      "mean_token_accuracy": 0.8241766504943371,
      "num_tokens": 385636004.0,
      "step": 12083
    },
    {
      "entropy": 0.7152683772146702,
      "epoch": 1.111787741301085,
      "grad_norm": 0.16115818917751312,
      "learning_rate": 6.294353973073267e-05,
      "loss": 0.6971,
      "mean_token_accuracy": 0.7877790480852127,
      "num_tokens": 385668036.0,
      "step": 12084
    },
    {
      "entropy": 0.603926032781601,
      "epoch": 1.1118797476725264,
      "grad_norm": 0.16697411239147186,
      "learning_rate": 6.294047290459105e-05,
      "loss": 0.5965,
      "mean_token_accuracy": 0.8185357227921486,
      "num_tokens": 385700476.0,
      "step": 12085
    },
    {
      "entropy": 0.6513371579349041,
      "epoch": 1.1119717540439675,
      "grad_norm": 0.16492721438407898,
      "learning_rate": 6.293740607844941e-05,
      "loss": 0.6447,
      "mean_token_accuracy": 0.8028602339327335,
      "num_tokens": 385732739.0,
      "step": 12086
    },
    {
      "entropy": 0.639233355410397,
      "epoch": 1.1120637604154087,
      "grad_norm": 0.1632261574268341,
      "learning_rate": 6.29343392523078e-05,
      "loss": 0.6153,
      "mean_token_accuracy": 0.8130492232739925,
      "num_tokens": 385764607.0,
      "step": 12087
    },
    {
      "entropy": 0.5286483764648438,
      "epoch": 1.11215576678685,
      "grad_norm": 0.1603332757949829,
      "learning_rate": 6.293127242616616e-05,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8382335156202316,
      "num_tokens": 385795858.0,
      "step": 12088
    },
    {
      "entropy": 0.6704924628138542,
      "epoch": 1.1122477731582912,
      "grad_norm": 0.16073234379291534,
      "learning_rate": 6.292820560002453e-05,
      "loss": 0.6503,
      "mean_token_accuracy": 0.7995770908892155,
      "num_tokens": 385827478.0,
      "step": 12089
    },
    {
      "entropy": 0.5992990285158157,
      "epoch": 1.1123397795297325,
      "grad_norm": 0.16053877770900726,
      "learning_rate": 6.292513877388291e-05,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8183028623461723,
      "num_tokens": 385859546.0,
      "step": 12090
    },
    {
      "entropy": 0.5971203148365021,
      "epoch": 1.1124317859011736,
      "grad_norm": 0.15299256145954132,
      "learning_rate": 6.29220719477413e-05,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8222013302147388,
      "num_tokens": 385891725.0,
      "step": 12091
    },
    {
      "entropy": 0.513954627327621,
      "epoch": 1.1125237922726148,
      "grad_norm": 0.15446777641773224,
      "learning_rate": 6.291900512159966e-05,
      "loss": 0.4845,
      "mean_token_accuracy": 0.8469610586762428,
      "num_tokens": 385923212.0,
      "step": 12092
    },
    {
      "entropy": 0.592799499630928,
      "epoch": 1.1126157986440561,
      "grad_norm": 0.15526577830314636,
      "learning_rate": 6.291593829545803e-05,
      "loss": 0.57,
      "mean_token_accuracy": 0.8247033096849918,
      "num_tokens": 385955183.0,
      "step": 12093
    },
    {
      "entropy": 0.5560277886688709,
      "epoch": 1.1127078050154973,
      "grad_norm": 0.16551519930362701,
      "learning_rate": 6.29128714693164e-05,
      "loss": 0.5475,
      "mean_token_accuracy": 0.830170139670372,
      "num_tokens": 385987388.0,
      "step": 12094
    },
    {
      "entropy": 0.5858641555532813,
      "epoch": 1.1127998113869386,
      "grad_norm": 0.1553163081407547,
      "learning_rate": 6.290980464317478e-05,
      "loss": 0.5719,
      "mean_token_accuracy": 0.8243082240223885,
      "num_tokens": 386019070.0,
      "step": 12095
    },
    {
      "entropy": 0.49657947942614555,
      "epoch": 1.1128918177583798,
      "grad_norm": 0.14877478778362274,
      "learning_rate": 6.290673781703316e-05,
      "loss": 0.4744,
      "mean_token_accuracy": 0.852279718965292,
      "num_tokens": 386051190.0,
      "step": 12096
    },
    {
      "entropy": 0.49571618996560574,
      "epoch": 1.1129838241298209,
      "grad_norm": 0.15356607735157013,
      "learning_rate": 6.290367099089153e-05,
      "loss": 0.4906,
      "mean_token_accuracy": 0.8490956276655197,
      "num_tokens": 386083757.0,
      "step": 12097
    },
    {
      "entropy": 0.610845829360187,
      "epoch": 1.1130758305012622,
      "grad_norm": 0.16548465192317963,
      "learning_rate": 6.29006041647499e-05,
      "loss": 0.6154,
      "mean_token_accuracy": 0.8142615146934986,
      "num_tokens": 386115949.0,
      "step": 12098
    },
    {
      "entropy": 0.6495084166526794,
      "epoch": 1.1131678368727034,
      "grad_norm": 0.16703087091445923,
      "learning_rate": 6.289753733860828e-05,
      "loss": 0.635,
      "mean_token_accuracy": 0.8040970116853714,
      "num_tokens": 386147785.0,
      "step": 12099
    },
    {
      "entropy": 0.6116006933152676,
      "epoch": 1.1132598432441447,
      "grad_norm": 0.15784059464931488,
      "learning_rate": 6.289447051246665e-05,
      "loss": 0.5992,
      "mean_token_accuracy": 0.8136153481900692,
      "num_tokens": 386180114.0,
      "step": 12100
    },
    {
      "entropy": 0.7258870489895344,
      "epoch": 1.1133518496155859,
      "grad_norm": 0.16162638366222382,
      "learning_rate": 6.289140368632503e-05,
      "loss": 0.7196,
      "mean_token_accuracy": 0.7803900502622128,
      "num_tokens": 386212335.0,
      "step": 12101
    },
    {
      "entropy": 0.6928859241306782,
      "epoch": 1.113443855987027,
      "grad_norm": 0.1571415513753891,
      "learning_rate": 6.28883368601834e-05,
      "loss": 0.6659,
      "mean_token_accuracy": 0.7935296520590782,
      "num_tokens": 386243877.0,
      "step": 12102
    },
    {
      "entropy": 0.7203040756285191,
      "epoch": 1.1135358623584684,
      "grad_norm": 0.17155055701732635,
      "learning_rate": 6.288527003404178e-05,
      "loss": 0.7218,
      "mean_token_accuracy": 0.7856583930552006,
      "num_tokens": 386275524.0,
      "step": 12103
    },
    {
      "entropy": 0.6444077342748642,
      "epoch": 1.1136278687299095,
      "grad_norm": 0.1619299203157425,
      "learning_rate": 6.288220320790014e-05,
      "loss": 0.6356,
      "mean_token_accuracy": 0.8048135675489902,
      "num_tokens": 386307687.0,
      "step": 12104
    },
    {
      "entropy": 0.7044136375188828,
      "epoch": 1.1137198751013508,
      "grad_norm": 0.1568893939256668,
      "learning_rate": 6.287913638175851e-05,
      "loss": 0.6983,
      "mean_token_accuracy": 0.7863843701779842,
      "num_tokens": 386340364.0,
      "step": 12105
    },
    {
      "entropy": 0.712964816018939,
      "epoch": 1.113811881472792,
      "grad_norm": 0.16076979041099548,
      "learning_rate": 6.28760695556169e-05,
      "loss": 0.6993,
      "mean_token_accuracy": 0.7847051285207272,
      "num_tokens": 386372477.0,
      "step": 12106
    },
    {
      "entropy": 0.7715111188590527,
      "epoch": 1.113903887844233,
      "grad_norm": 0.16902223229408264,
      "learning_rate": 6.287300272947528e-05,
      "loss": 0.7667,
      "mean_token_accuracy": 0.7725570313632488,
      "num_tokens": 386405029.0,
      "step": 12107
    },
    {
      "entropy": 0.5362414438277483,
      "epoch": 1.1139958942156745,
      "grad_norm": 0.161681666970253,
      "learning_rate": 6.286993590333364e-05,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8353296890854836,
      "num_tokens": 386437264.0,
      "step": 12108
    },
    {
      "entropy": 0.6114154495298862,
      "epoch": 1.1140879005871156,
      "grad_norm": 0.15641936659812927,
      "learning_rate": 6.286686907719201e-05,
      "loss": 0.5965,
      "mean_token_accuracy": 0.8170955069363117,
      "num_tokens": 386469581.0,
      "step": 12109
    },
    {
      "entropy": 0.7023135907948017,
      "epoch": 1.114179906958557,
      "grad_norm": 0.1648755967617035,
      "learning_rate": 6.286380225105038e-05,
      "loss": 0.6858,
      "mean_token_accuracy": 0.7898243553936481,
      "num_tokens": 386501491.0,
      "step": 12110
    },
    {
      "entropy": 0.6267715697176754,
      "epoch": 1.114271913329998,
      "grad_norm": 0.1617051661014557,
      "learning_rate": 6.286073542490877e-05,
      "loss": 0.6268,
      "mean_token_accuracy": 0.8148301541805267,
      "num_tokens": 386533389.0,
      "step": 12111
    },
    {
      "entropy": 0.5712135126814246,
      "epoch": 1.1143639197014394,
      "grad_norm": 0.15543478727340698,
      "learning_rate": 6.285766859876714e-05,
      "loss": 0.565,
      "mean_token_accuracy": 0.8258372843265533,
      "num_tokens": 386565881.0,
      "step": 12112
    },
    {
      "entropy": 0.6347584556788206,
      "epoch": 1.1144559260728806,
      "grad_norm": 0.1605682671070099,
      "learning_rate": 6.285460177262551e-05,
      "loss": 0.6317,
      "mean_token_accuracy": 0.8068779595196247,
      "num_tokens": 386596367.0,
      "step": 12113
    },
    {
      "entropy": 0.6157841943204403,
      "epoch": 1.1145479324443217,
      "grad_norm": 0.15796270966529846,
      "learning_rate": 6.285153494648389e-05,
      "loss": 0.5943,
      "mean_token_accuracy": 0.8196347020566463,
      "num_tokens": 386628375.0,
      "step": 12114
    },
    {
      "entropy": 0.6086333505809307,
      "epoch": 1.114639938815763,
      "grad_norm": 0.15174700319766998,
      "learning_rate": 6.284846812034226e-05,
      "loss": 0.5777,
      "mean_token_accuracy": 0.8203182220458984,
      "num_tokens": 386660227.0,
      "step": 12115
    },
    {
      "entropy": 0.584507321473211,
      "epoch": 1.1147319451872042,
      "grad_norm": 0.16160091757774353,
      "learning_rate": 6.284540129420064e-05,
      "loss": 0.5603,
      "mean_token_accuracy": 0.824185773730278,
      "num_tokens": 386691707.0,
      "step": 12116
    },
    {
      "entropy": 0.5869407271966338,
      "epoch": 1.1148239515586456,
      "grad_norm": 0.15383706986904144,
      "learning_rate": 6.284233446805901e-05,
      "loss": 0.5733,
      "mean_token_accuracy": 0.8195316158235073,
      "num_tokens": 386722783.0,
      "step": 12117
    },
    {
      "entropy": 0.6151750516146421,
      "epoch": 1.1149159579300867,
      "grad_norm": 0.1599535495042801,
      "learning_rate": 6.283926764191739e-05,
      "loss": 0.6152,
      "mean_token_accuracy": 0.8101671971380711,
      "num_tokens": 386754859.0,
      "step": 12118
    },
    {
      "entropy": 0.5275798444636166,
      "epoch": 1.1150079643015278,
      "grad_norm": 0.15003614127635956,
      "learning_rate": 6.283620081577576e-05,
      "loss": 0.5095,
      "mean_token_accuracy": 0.844604566693306,
      "num_tokens": 386786607.0,
      "step": 12119
    },
    {
      "entropy": 0.6119045969098806,
      "epoch": 1.1150999706729692,
      "grad_norm": 0.1641126275062561,
      "learning_rate": 6.283313398963413e-05,
      "loss": 0.6014,
      "mean_token_accuracy": 0.812938753515482,
      "num_tokens": 386818819.0,
      "step": 12120
    },
    {
      "entropy": 0.6752929100766778,
      "epoch": 1.1151919770444103,
      "grad_norm": 0.1740967035293579,
      "learning_rate": 6.283006716349251e-05,
      "loss": 0.6691,
      "mean_token_accuracy": 0.7962855510413647,
      "num_tokens": 386850562.0,
      "step": 12121
    },
    {
      "entropy": 0.5891275107860565,
      "epoch": 1.1152839834158517,
      "grad_norm": 0.1566610485315323,
      "learning_rate": 6.282700033735089e-05,
      "loss": 0.5752,
      "mean_token_accuracy": 0.8209591582417488,
      "num_tokens": 386882826.0,
      "step": 12122
    },
    {
      "entropy": 0.662980911321938,
      "epoch": 1.1153759897872928,
      "grad_norm": 0.16918280720710754,
      "learning_rate": 6.282393351120926e-05,
      "loss": 0.6387,
      "mean_token_accuracy": 0.8079873733222485,
      "num_tokens": 386914988.0,
      "step": 12123
    },
    {
      "entropy": 0.6518758940510452,
      "epoch": 1.115467996158734,
      "grad_norm": 0.15698593854904175,
      "learning_rate": 6.282086668506762e-05,
      "loss": 0.6333,
      "mean_token_accuracy": 0.8060925044119358,
      "num_tokens": 386946778.0,
      "step": 12124
    },
    {
      "entropy": 0.5729522109031677,
      "epoch": 1.1155600025301753,
      "grad_norm": 0.15559199452400208,
      "learning_rate": 6.281779985892599e-05,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8238819502294064,
      "num_tokens": 386977130.0,
      "step": 12125
    },
    {
      "entropy": 0.5950844474136829,
      "epoch": 1.1156520089016164,
      "grad_norm": 0.15603800117969513,
      "learning_rate": 6.281473303278437e-05,
      "loss": 0.5765,
      "mean_token_accuracy": 0.8239938877522945,
      "num_tokens": 387009484.0,
      "step": 12126
    },
    {
      "entropy": 0.628826797939837,
      "epoch": 1.1157440152730578,
      "grad_norm": 0.16120390594005585,
      "learning_rate": 6.281166620664275e-05,
      "loss": 0.6236,
      "mean_token_accuracy": 0.8113384433090687,
      "num_tokens": 387041520.0,
      "step": 12127
    },
    {
      "entropy": 0.6565228719264269,
      "epoch": 1.115836021644499,
      "grad_norm": 0.16070224344730377,
      "learning_rate": 6.280859938050112e-05,
      "loss": 0.6266,
      "mean_token_accuracy": 0.8089266419410706,
      "num_tokens": 387072606.0,
      "step": 12128
    },
    {
      "entropy": 0.6354540046304464,
      "epoch": 1.11592802801594,
      "grad_norm": 0.160100057721138,
      "learning_rate": 6.280553255435949e-05,
      "loss": 0.6235,
      "mean_token_accuracy": 0.809100303798914,
      "num_tokens": 387103867.0,
      "step": 12129
    },
    {
      "entropy": 0.6412583729252219,
      "epoch": 1.1160200343873814,
      "grad_norm": 0.15302447974681854,
      "learning_rate": 6.280246572821787e-05,
      "loss": 0.6155,
      "mean_token_accuracy": 0.8110044375061989,
      "num_tokens": 387135620.0,
      "step": 12130
    },
    {
      "entropy": 0.6145969144999981,
      "epoch": 1.1161120407588225,
      "grad_norm": 0.16689516603946686,
      "learning_rate": 6.279939890207624e-05,
      "loss": 0.5942,
      "mean_token_accuracy": 0.818646427243948,
      "num_tokens": 387168135.0,
      "step": 12131
    },
    {
      "entropy": 0.5946705425158143,
      "epoch": 1.1162040471302639,
      "grad_norm": 0.15712812542915344,
      "learning_rate": 6.279633207593462e-05,
      "loss": 0.5984,
      "mean_token_accuracy": 0.8178642056882381,
      "num_tokens": 387199581.0,
      "step": 12132
    },
    {
      "entropy": 0.6124974638223648,
      "epoch": 1.116296053501705,
      "grad_norm": 0.16141897439956665,
      "learning_rate": 6.279326524979299e-05,
      "loss": 0.6105,
      "mean_token_accuracy": 0.8164975084364414,
      "num_tokens": 387230827.0,
      "step": 12133
    },
    {
      "entropy": 0.6307868026196957,
      "epoch": 1.1163880598731462,
      "grad_norm": 0.16376706957817078,
      "learning_rate": 6.279019842365137e-05,
      "loss": 0.6145,
      "mean_token_accuracy": 0.8133719153702259,
      "num_tokens": 387262456.0,
      "step": 12134
    },
    {
      "entropy": 0.6082110162824392,
      "epoch": 1.1164800662445875,
      "grad_norm": 0.15629763901233673,
      "learning_rate": 6.278713159750974e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8167837485671043,
      "num_tokens": 387294823.0,
      "step": 12135
    },
    {
      "entropy": 0.6834202315658331,
      "epoch": 1.1165720726160286,
      "grad_norm": 0.1594376564025879,
      "learning_rate": 6.27840647713681e-05,
      "loss": 0.6729,
      "mean_token_accuracy": 0.8009903095662594,
      "num_tokens": 387327108.0,
      "step": 12136
    },
    {
      "entropy": 0.6697252821177244,
      "epoch": 1.11666407898747,
      "grad_norm": 0.1640646755695343,
      "learning_rate": 6.278099794522649e-05,
      "loss": 0.6534,
      "mean_token_accuracy": 0.7998804226517677,
      "num_tokens": 387358694.0,
      "step": 12137
    },
    {
      "entropy": 0.6750289872288704,
      "epoch": 1.1167560853589111,
      "grad_norm": 0.16358605027198792,
      "learning_rate": 6.277793111908487e-05,
      "loss": 0.6634,
      "mean_token_accuracy": 0.7975106500089169,
      "num_tokens": 387391267.0,
      "step": 12138
    },
    {
      "entropy": 0.5953036565333605,
      "epoch": 1.1168480917303523,
      "grad_norm": 0.15361717343330383,
      "learning_rate": 6.277486429294324e-05,
      "loss": 0.5885,
      "mean_token_accuracy": 0.8131055645644665,
      "num_tokens": 387423520.0,
      "step": 12139
    },
    {
      "entropy": 0.669703827239573,
      "epoch": 1.1169400981017936,
      "grad_norm": 0.15775735676288605,
      "learning_rate": 6.27717974668016e-05,
      "loss": 0.6607,
      "mean_token_accuracy": 0.7955698780715466,
      "num_tokens": 387456133.0,
      "step": 12140
    },
    {
      "entropy": 0.5598648055456579,
      "epoch": 1.1170321044732348,
      "grad_norm": 0.154282808303833,
      "learning_rate": 6.276873064065999e-05,
      "loss": 0.5546,
      "mean_token_accuracy": 0.8318857960402966,
      "num_tokens": 387488527.0,
      "step": 12141
    },
    {
      "entropy": 0.5998812578618526,
      "epoch": 1.117124110844676,
      "grad_norm": 0.15575256943702698,
      "learning_rate": 6.276566381451835e-05,
      "loss": 0.5863,
      "mean_token_accuracy": 0.8160900957882404,
      "num_tokens": 387519840.0,
      "step": 12142
    },
    {
      "entropy": 0.6064626509323716,
      "epoch": 1.1172161172161172,
      "grad_norm": 0.1547597199678421,
      "learning_rate": 6.276259698837674e-05,
      "loss": 0.5907,
      "mean_token_accuracy": 0.8208630494773388,
      "num_tokens": 387552044.0,
      "step": 12143
    },
    {
      "entropy": 0.617518630810082,
      "epoch": 1.1173081235875584,
      "grad_norm": 0.1556403487920761,
      "learning_rate": 6.27595301622351e-05,
      "loss": 0.5999,
      "mean_token_accuracy": 0.8116064593195915,
      "num_tokens": 387582898.0,
      "step": 12144
    },
    {
      "entropy": 0.688717607408762,
      "epoch": 1.1174001299589997,
      "grad_norm": 0.16368821263313293,
      "learning_rate": 6.275646333609348e-05,
      "loss": 0.6707,
      "mean_token_accuracy": 0.7948914989829063,
      "num_tokens": 387614351.0,
      "step": 12145
    },
    {
      "entropy": 0.5783405778929591,
      "epoch": 1.1174921363304409,
      "grad_norm": 0.1554708182811737,
      "learning_rate": 6.275339650995185e-05,
      "loss": 0.5673,
      "mean_token_accuracy": 0.8279650919139385,
      "num_tokens": 387646481.0,
      "step": 12146
    },
    {
      "entropy": 0.6014251187443733,
      "epoch": 1.1175841427018822,
      "grad_norm": 0.15658484399318695,
      "learning_rate": 6.275032968381022e-05,
      "loss": 0.5926,
      "mean_token_accuracy": 0.818914782255888,
      "num_tokens": 387678413.0,
      "step": 12147
    },
    {
      "entropy": 0.6247548405081034,
      "epoch": 1.1176761490733234,
      "grad_norm": 0.15972864627838135,
      "learning_rate": 6.27472628576686e-05,
      "loss": 0.6097,
      "mean_token_accuracy": 0.8106067627668381,
      "num_tokens": 387710526.0,
      "step": 12148
    },
    {
      "entropy": 0.4939286857843399,
      "epoch": 1.1177681554447645,
      "grad_norm": 0.14794684946537018,
      "learning_rate": 6.274419603152698e-05,
      "loss": 0.4781,
      "mean_token_accuracy": 0.8507971689105034,
      "num_tokens": 387742767.0,
      "step": 12149
    },
    {
      "entropy": 0.6942765265703201,
      "epoch": 1.1178601618162058,
      "grad_norm": 0.1545189768075943,
      "learning_rate": 6.274112920538535e-05,
      "loss": 0.6774,
      "mean_token_accuracy": 0.7935648001730442,
      "num_tokens": 387775071.0,
      "step": 12150
    },
    {
      "entropy": 0.6898792739957571,
      "epoch": 1.117952168187647,
      "grad_norm": 0.16135825216770172,
      "learning_rate": 6.273806237924372e-05,
      "loss": 0.685,
      "mean_token_accuracy": 0.7922096401453018,
      "num_tokens": 387806904.0,
      "step": 12151
    },
    {
      "entropy": 0.5689001921564341,
      "epoch": 1.1180441745590883,
      "grad_norm": 0.15459120273590088,
      "learning_rate": 6.273499555310209e-05,
      "loss": 0.5616,
      "mean_token_accuracy": 0.8267132453620434,
      "num_tokens": 387839130.0,
      "step": 12152
    },
    {
      "entropy": 0.637919107452035,
      "epoch": 1.1181361809305295,
      "grad_norm": 0.15679892897605896,
      "learning_rate": 6.273192872696048e-05,
      "loss": 0.6085,
      "mean_token_accuracy": 0.8112007975578308,
      "num_tokens": 387871520.0,
      "step": 12153
    },
    {
      "entropy": 0.5916770715266466,
      "epoch": 1.1182281873019706,
      "grad_norm": 0.15254662930965424,
      "learning_rate": 6.272886190081885e-05,
      "loss": 0.5719,
      "mean_token_accuracy": 0.8228974901139736,
      "num_tokens": 387903882.0,
      "step": 12154
    },
    {
      "entropy": 0.5573037397116423,
      "epoch": 1.118320193673412,
      "grad_norm": 0.1562248021364212,
      "learning_rate": 6.272579507467722e-05,
      "loss": 0.5426,
      "mean_token_accuracy": 0.8321195282042027,
      "num_tokens": 387936035.0,
      "step": 12155
    },
    {
      "entropy": 0.594511535950005,
      "epoch": 1.118412200044853,
      "grad_norm": 0.15622200071811676,
      "learning_rate": 6.272272824853559e-05,
      "loss": 0.5878,
      "mean_token_accuracy": 0.818859875202179,
      "num_tokens": 387968047.0,
      "step": 12156
    },
    {
      "entropy": 0.6732793087139726,
      "epoch": 1.1185042064162944,
      "grad_norm": 0.16584879159927368,
      "learning_rate": 6.271966142239397e-05,
      "loss": 0.661,
      "mean_token_accuracy": 0.8013495579361916,
      "num_tokens": 387999640.0,
      "step": 12157
    },
    {
      "entropy": 0.47935498133301735,
      "epoch": 1.1185962127877356,
      "grad_norm": 0.15279844403266907,
      "learning_rate": 6.271659459625235e-05,
      "loss": 0.4541,
      "mean_token_accuracy": 0.857850156724453,
      "num_tokens": 388032136.0,
      "step": 12158
    },
    {
      "entropy": 0.6747901178896427,
      "epoch": 1.1186882191591767,
      "grad_norm": 0.16625641286373138,
      "learning_rate": 6.271352777011072e-05,
      "loss": 0.6802,
      "mean_token_accuracy": 0.7956645600497723,
      "num_tokens": 388064591.0,
      "step": 12159
    },
    {
      "entropy": 0.639413096010685,
      "epoch": 1.118780225530618,
      "grad_norm": 0.1676376312971115,
      "learning_rate": 6.271046094396908e-05,
      "loss": 0.6429,
      "mean_token_accuracy": 0.7984349094331264,
      "num_tokens": 388095546.0,
      "step": 12160
    },
    {
      "entropy": 0.6277893390506506,
      "epoch": 1.1188722319020592,
      "grad_norm": 0.1615757942199707,
      "learning_rate": 6.270739411782747e-05,
      "loss": 0.6274,
      "mean_token_accuracy": 0.8045448586344719,
      "num_tokens": 388127824.0,
      "step": 12161
    },
    {
      "entropy": 0.5879512634128332,
      "epoch": 1.1189642382735006,
      "grad_norm": 0.1514732390642166,
      "learning_rate": 6.270432729168583e-05,
      "loss": 0.5805,
      "mean_token_accuracy": 0.8243176266551018,
      "num_tokens": 388160135.0,
      "step": 12162
    },
    {
      "entropy": 0.5362866842187941,
      "epoch": 1.1190562446449417,
      "grad_norm": 0.14763420820236206,
      "learning_rate": 6.270126046554422e-05,
      "loss": 0.5139,
      "mean_token_accuracy": 0.8411957174539566,
      "num_tokens": 388192129.0,
      "step": 12163
    },
    {
      "entropy": 0.5136353923007846,
      "epoch": 1.1191482510163828,
      "grad_norm": 0.1436835527420044,
      "learning_rate": 6.269819363940258e-05,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8488614819943905,
      "num_tokens": 388224690.0,
      "step": 12164
    },
    {
      "entropy": 0.6844372563064098,
      "epoch": 1.1192402573878242,
      "grad_norm": 0.1573692411184311,
      "learning_rate": 6.269512681326096e-05,
      "loss": 0.6688,
      "mean_token_accuracy": 0.7958404645323753,
      "num_tokens": 388256399.0,
      "step": 12165
    },
    {
      "entropy": 0.64595952257514,
      "epoch": 1.1193322637592653,
      "grad_norm": 0.1553141474723816,
      "learning_rate": 6.269205998711933e-05,
      "loss": 0.6281,
      "mean_token_accuracy": 0.807477455586195,
      "num_tokens": 388288022.0,
      "step": 12166
    },
    {
      "entropy": 0.6765026412904263,
      "epoch": 1.1194242701307067,
      "grad_norm": 0.1715923547744751,
      "learning_rate": 6.26889931609777e-05,
      "loss": 0.6674,
      "mean_token_accuracy": 0.7944790758192539,
      "num_tokens": 388319506.0,
      "step": 12167
    },
    {
      "entropy": 0.6138639729470015,
      "epoch": 1.1195162765021478,
      "grad_norm": 0.1587565690279007,
      "learning_rate": 6.268592633483608e-05,
      "loss": 0.6089,
      "mean_token_accuracy": 0.8126843087375164,
      "num_tokens": 388351635.0,
      "step": 12168
    },
    {
      "entropy": 0.6335076568648219,
      "epoch": 1.119608282873589,
      "grad_norm": 0.16094516217708588,
      "learning_rate": 6.268285950869446e-05,
      "loss": 0.6195,
      "mean_token_accuracy": 0.808599054813385,
      "num_tokens": 388383027.0,
      "step": 12169
    },
    {
      "entropy": 0.5859657041728497,
      "epoch": 1.1197002892450303,
      "grad_norm": 0.15644702315330505,
      "learning_rate": 6.267979268255283e-05,
      "loss": 0.5692,
      "mean_token_accuracy": 0.8202797807753086,
      "num_tokens": 388414509.0,
      "step": 12170
    },
    {
      "entropy": 0.5795813612639904,
      "epoch": 1.1197922956164714,
      "grad_norm": 0.1516784429550171,
      "learning_rate": 6.26767258564112e-05,
      "loss": 0.5833,
      "mean_token_accuracy": 0.8172140456736088,
      "num_tokens": 388446511.0,
      "step": 12171
    },
    {
      "entropy": 0.7137757875025272,
      "epoch": 1.1198843019879128,
      "grad_norm": 0.16189798712730408,
      "learning_rate": 6.267365903026958e-05,
      "loss": 0.706,
      "mean_token_accuracy": 0.7858299985527992,
      "num_tokens": 388478282.0,
      "step": 12172
    },
    {
      "entropy": 0.5247584069147706,
      "epoch": 1.119976308359354,
      "grad_norm": 0.15752921998500824,
      "learning_rate": 6.267059220412795e-05,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8385604359209538,
      "num_tokens": 388510588.0,
      "step": 12173
    },
    {
      "entropy": 0.6179256662726402,
      "epoch": 1.120068314730795,
      "grad_norm": 0.16154305636882782,
      "learning_rate": 6.266752537798633e-05,
      "loss": 0.6108,
      "mean_token_accuracy": 0.8124524131417274,
      "num_tokens": 388542288.0,
      "step": 12174
    },
    {
      "entropy": 0.6170281525701284,
      "epoch": 1.1201603211022364,
      "grad_norm": 0.16369417309761047,
      "learning_rate": 6.26644585518447e-05,
      "loss": 0.5978,
      "mean_token_accuracy": 0.8173465207219124,
      "num_tokens": 388573837.0,
      "step": 12175
    },
    {
      "entropy": 0.5631396099925041,
      "epoch": 1.1202523274736775,
      "grad_norm": 0.15140175819396973,
      "learning_rate": 6.266139172570308e-05,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8358767703175545,
      "num_tokens": 388605063.0,
      "step": 12176
    },
    {
      "entropy": 0.6698744408786297,
      "epoch": 1.120344333845119,
      "grad_norm": 0.16333924233913422,
      "learning_rate": 6.265832489956145e-05,
      "loss": 0.6656,
      "mean_token_accuracy": 0.7953897379338741,
      "num_tokens": 388636438.0,
      "step": 12177
    },
    {
      "entropy": 0.5665833340026438,
      "epoch": 1.12043634021656,
      "grad_norm": 0.14915554225444794,
      "learning_rate": 6.265525807341981e-05,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8250980749726295,
      "num_tokens": 388667934.0,
      "step": 12178
    },
    {
      "entropy": 0.7399786822497845,
      "epoch": 1.1205283465880012,
      "grad_norm": 0.16753000020980835,
      "learning_rate": 6.26521912472782e-05,
      "loss": 0.717,
      "mean_token_accuracy": 0.7829414941370487,
      "num_tokens": 388699620.0,
      "step": 12179
    },
    {
      "entropy": 0.6163069549947977,
      "epoch": 1.1206203529594425,
      "grad_norm": 0.16178862750530243,
      "learning_rate": 6.264912442113658e-05,
      "loss": 0.61,
      "mean_token_accuracy": 0.8143032044172287,
      "num_tokens": 388731321.0,
      "step": 12180
    },
    {
      "entropy": 0.6611706521362066,
      "epoch": 1.1207123593308836,
      "grad_norm": 0.16338525712490082,
      "learning_rate": 6.264605759499495e-05,
      "loss": 0.66,
      "mean_token_accuracy": 0.7975606173276901,
      "num_tokens": 388762422.0,
      "step": 12181
    },
    {
      "entropy": 0.6730274092406034,
      "epoch": 1.120804365702325,
      "grad_norm": 0.16613337397575378,
      "learning_rate": 6.264299076885331e-05,
      "loss": 0.6696,
      "mean_token_accuracy": 0.7926733493804932,
      "num_tokens": 388794903.0,
      "step": 12182
    },
    {
      "entropy": 0.577995334751904,
      "epoch": 1.1208963720737661,
      "grad_norm": 0.15582917630672455,
      "learning_rate": 6.263992394271168e-05,
      "loss": 0.5446,
      "mean_token_accuracy": 0.8291815221309662,
      "num_tokens": 388826467.0,
      "step": 12183
    },
    {
      "entropy": 0.676032311283052,
      "epoch": 1.1209883784452073,
      "grad_norm": 0.15753516554832458,
      "learning_rate": 6.263685711657006e-05,
      "loss": 0.6612,
      "mean_token_accuracy": 0.8009379357099533,
      "num_tokens": 388857819.0,
      "step": 12184
    },
    {
      "entropy": 0.5534293325617909,
      "epoch": 1.1210803848166486,
      "grad_norm": 0.1476096361875534,
      "learning_rate": 6.263379029042844e-05,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8311251699924469,
      "num_tokens": 388889772.0,
      "step": 12185
    },
    {
      "entropy": 0.6437073741108179,
      "epoch": 1.1211723911880898,
      "grad_norm": 0.1558183878660202,
      "learning_rate": 6.263072346428681e-05,
      "loss": 0.6186,
      "mean_token_accuracy": 0.8068449981510639,
      "num_tokens": 388921173.0,
      "step": 12186
    },
    {
      "entropy": 0.6951745059341192,
      "epoch": 1.1212643975595311,
      "grad_norm": 0.16562555730342865,
      "learning_rate": 6.262765663814518e-05,
      "loss": 0.6783,
      "mean_token_accuracy": 0.7942397929728031,
      "num_tokens": 388953177.0,
      "step": 12187
    },
    {
      "entropy": 0.6709417197853327,
      "epoch": 1.1213564039309722,
      "grad_norm": 0.16643516719341278,
      "learning_rate": 6.262458981200356e-05,
      "loss": 0.6553,
      "mean_token_accuracy": 0.8011669963598251,
      "num_tokens": 388984773.0,
      "step": 12188
    },
    {
      "entropy": 0.5872824713587761,
      "epoch": 1.1214484103024134,
      "grad_norm": 0.15302881598472595,
      "learning_rate": 6.262152298586194e-05,
      "loss": 0.5629,
      "mean_token_accuracy": 0.8240437619388103,
      "num_tokens": 389016812.0,
      "step": 12189
    },
    {
      "entropy": 0.6656054444611073,
      "epoch": 1.1215404166738547,
      "grad_norm": 0.1595534235239029,
      "learning_rate": 6.261845615972031e-05,
      "loss": 0.6513,
      "mean_token_accuracy": 0.801630437374115,
      "num_tokens": 389049580.0,
      "step": 12190
    },
    {
      "entropy": 0.6417529303580523,
      "epoch": 1.1216324230452959,
      "grad_norm": 0.1581224799156189,
      "learning_rate": 6.261538933357868e-05,
      "loss": 0.6285,
      "mean_token_accuracy": 0.8053717613220215,
      "num_tokens": 389082070.0,
      "step": 12191
    },
    {
      "entropy": 0.6214770022779703,
      "epoch": 1.1217244294167372,
      "grad_norm": 0.16677717864513397,
      "learning_rate": 6.261232250743706e-05,
      "loss": 0.6192,
      "mean_token_accuracy": 0.8124306686222553,
      "num_tokens": 389113650.0,
      "step": 12192
    },
    {
      "entropy": 0.5405152402818203,
      "epoch": 1.1218164357881784,
      "grad_norm": 0.16381323337554932,
      "learning_rate": 6.260925568129543e-05,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8452176228165627,
      "num_tokens": 389145691.0,
      "step": 12193
    },
    {
      "entropy": 0.5511918552219868,
      "epoch": 1.1219084421596195,
      "grad_norm": 0.160098597407341,
      "learning_rate": 6.260618885515381e-05,
      "loss": 0.5536,
      "mean_token_accuracy": 0.8262425251305103,
      "num_tokens": 389178302.0,
      "step": 12194
    },
    {
      "entropy": 0.6714460095390677,
      "epoch": 1.1220004485310608,
      "grad_norm": 0.16266730427742004,
      "learning_rate": 6.260312202901218e-05,
      "loss": 0.6659,
      "mean_token_accuracy": 0.798232726752758,
      "num_tokens": 389211068.0,
      "step": 12195
    },
    {
      "entropy": 0.5970324017107487,
      "epoch": 1.122092454902502,
      "grad_norm": 0.16167142987251282,
      "learning_rate": 6.260005520287056e-05,
      "loss": 0.5817,
      "mean_token_accuracy": 0.8234329856932163,
      "num_tokens": 389242931.0,
      "step": 12196
    },
    {
      "entropy": 0.5851993151009083,
      "epoch": 1.1221844612739433,
      "grad_norm": 0.14655552804470062,
      "learning_rate": 6.259698837672893e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.821891900151968,
      "num_tokens": 389274990.0,
      "step": 12197
    },
    {
      "entropy": 0.6075099362060428,
      "epoch": 1.1222764676453845,
      "grad_norm": 0.15676119923591614,
      "learning_rate": 6.25939215505873e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8161752112209797,
      "num_tokens": 389307023.0,
      "step": 12198
    },
    {
      "entropy": 0.68764977902174,
      "epoch": 1.1223684740168256,
      "grad_norm": 0.16272112727165222,
      "learning_rate": 6.259085472444568e-05,
      "loss": 0.6737,
      "mean_token_accuracy": 0.7934075258672237,
      "num_tokens": 389338503.0,
      "step": 12199
    },
    {
      "entropy": 0.6170589029788971,
      "epoch": 1.122460480388267,
      "grad_norm": 0.15361036360263824,
      "learning_rate": 6.258778789830406e-05,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8176712989807129,
      "num_tokens": 389369952.0,
      "step": 12200
    },
    {
      "entropy": 0.5627485029399395,
      "epoch": 1.122552486759708,
      "grad_norm": 0.15487124025821686,
      "learning_rate": 6.258472107216242e-05,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8273320123553276,
      "num_tokens": 389401973.0,
      "step": 12201
    },
    {
      "entropy": 0.6434088665992022,
      "epoch": 1.1226444931311494,
      "grad_norm": 0.1612452119588852,
      "learning_rate": 6.258165424602079e-05,
      "loss": 0.628,
      "mean_token_accuracy": 0.8064448572695255,
      "num_tokens": 389434092.0,
      "step": 12202
    },
    {
      "entropy": 0.5978073216974735,
      "epoch": 1.1227364995025906,
      "grad_norm": 0.16088344156742096,
      "learning_rate": 6.257858741987917e-05,
      "loss": 0.5929,
      "mean_token_accuracy": 0.8173680528998375,
      "num_tokens": 389465948.0,
      "step": 12203
    },
    {
      "entropy": 0.7285236036404967,
      "epoch": 1.1228285058740317,
      "grad_norm": 0.16285350918769836,
      "learning_rate": 6.257552059373754e-05,
      "loss": 0.7293,
      "mean_token_accuracy": 0.7769272066652775,
      "num_tokens": 389497869.0,
      "step": 12204
    },
    {
      "entropy": 0.665761923417449,
      "epoch": 1.122920512245473,
      "grad_norm": 0.16509081423282623,
      "learning_rate": 6.257245376759592e-05,
      "loss": 0.6589,
      "mean_token_accuracy": 0.7984505072236061,
      "num_tokens": 389529883.0,
      "step": 12205
    },
    {
      "entropy": 0.7141081038862467,
      "epoch": 1.1230125186169142,
      "grad_norm": 0.16843369603157043,
      "learning_rate": 6.256938694145429e-05,
      "loss": 0.7162,
      "mean_token_accuracy": 0.7843811325728893,
      "num_tokens": 389562042.0,
      "step": 12206
    },
    {
      "entropy": 0.6978280358016491,
      "epoch": 1.1231045249883556,
      "grad_norm": 0.16192954778671265,
      "learning_rate": 6.256632011531267e-05,
      "loss": 0.692,
      "mean_token_accuracy": 0.7893532961606979,
      "num_tokens": 389593498.0,
      "step": 12207
    },
    {
      "entropy": 0.6409865152090788,
      "epoch": 1.1231965313597967,
      "grad_norm": 0.15829631686210632,
      "learning_rate": 6.256325328917104e-05,
      "loss": 0.6331,
      "mean_token_accuracy": 0.8051211312413216,
      "num_tokens": 389624599.0,
      "step": 12208
    },
    {
      "entropy": 0.56433873064816,
      "epoch": 1.1232885377312378,
      "grad_norm": 0.1503930240869522,
      "learning_rate": 6.256018646302941e-05,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8304724842309952,
      "num_tokens": 389657244.0,
      "step": 12209
    },
    {
      "entropy": 0.6610716003924608,
      "epoch": 1.1233805441026792,
      "grad_norm": 0.16245786845684052,
      "learning_rate": 6.255711963688779e-05,
      "loss": 0.6567,
      "mean_token_accuracy": 0.7977386452257633,
      "num_tokens": 389689408.0,
      "step": 12210
    },
    {
      "entropy": 0.6332115896511823,
      "epoch": 1.1234725504741203,
      "grad_norm": 0.15981976687908173,
      "learning_rate": 6.255405281074617e-05,
      "loss": 0.6223,
      "mean_token_accuracy": 0.8124847002327442,
      "num_tokens": 389721095.0,
      "step": 12211
    },
    {
      "entropy": 0.6031040474772453,
      "epoch": 1.1235645568455617,
      "grad_norm": 0.16927991807460785,
      "learning_rate": 6.255098598460454e-05,
      "loss": 0.6039,
      "mean_token_accuracy": 0.8171661905944347,
      "num_tokens": 389753251.0,
      "step": 12212
    },
    {
      "entropy": 0.6134344469755888,
      "epoch": 1.1236565632170028,
      "grad_norm": 0.1520242840051651,
      "learning_rate": 6.25479191584629e-05,
      "loss": 0.6002,
      "mean_token_accuracy": 0.8175582550466061,
      "num_tokens": 389785702.0,
      "step": 12213
    },
    {
      "entropy": 0.5569830322638154,
      "epoch": 1.123748569588444,
      "grad_norm": 0.15373452007770538,
      "learning_rate": 6.254485233232127e-05,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8345274478197098,
      "num_tokens": 389817673.0,
      "step": 12214
    },
    {
      "entropy": 0.7192040309309959,
      "epoch": 1.1238405759598853,
      "grad_norm": 0.16082864999771118,
      "learning_rate": 6.254178550617966e-05,
      "loss": 0.703,
      "mean_token_accuracy": 0.7867340371012688,
      "num_tokens": 389849546.0,
      "step": 12215
    },
    {
      "entropy": 0.6580533515661955,
      "epoch": 1.1239325823313264,
      "grad_norm": 0.164226233959198,
      "learning_rate": 6.253871868003804e-05,
      "loss": 0.644,
      "mean_token_accuracy": 0.8042950667440891,
      "num_tokens": 389881086.0,
      "step": 12216
    },
    {
      "entropy": 0.6970789153128862,
      "epoch": 1.1240245887027678,
      "grad_norm": 0.17435042560100555,
      "learning_rate": 6.25356518538964e-05,
      "loss": 0.7027,
      "mean_token_accuracy": 0.7875924967229366,
      "num_tokens": 389912722.0,
      "step": 12217
    },
    {
      "entropy": 0.5997174102813005,
      "epoch": 1.124116595074209,
      "grad_norm": 0.15608170628547668,
      "learning_rate": 6.253258502775477e-05,
      "loss": 0.5787,
      "mean_token_accuracy": 0.8174318112432957,
      "num_tokens": 389945208.0,
      "step": 12218
    },
    {
      "entropy": 0.716785401571542,
      "epoch": 1.12420860144565,
      "grad_norm": 0.16981111466884613,
      "learning_rate": 6.252951820161315e-05,
      "loss": 0.695,
      "mean_token_accuracy": 0.7889684215188026,
      "num_tokens": 389976501.0,
      "step": 12219
    },
    {
      "entropy": 0.5622415896505117,
      "epoch": 1.1243006078170914,
      "grad_norm": 0.15238194167613983,
      "learning_rate": 6.252645137547152e-05,
      "loss": 0.5302,
      "mean_token_accuracy": 0.8353838212788105,
      "num_tokens": 390008762.0,
      "step": 12220
    },
    {
      "entropy": 0.6009800117462873,
      "epoch": 1.1243926141885325,
      "grad_norm": 0.14949263632297516,
      "learning_rate": 6.25233845493299e-05,
      "loss": 0.5847,
      "mean_token_accuracy": 0.8182501681149006,
      "num_tokens": 390040446.0,
      "step": 12221
    },
    {
      "entropy": 0.6074634864926338,
      "epoch": 1.124484620559974,
      "grad_norm": 0.1486591100692749,
      "learning_rate": 6.252031772318827e-05,
      "loss": 0.5813,
      "mean_token_accuracy": 0.8240891396999359,
      "num_tokens": 390072520.0,
      "step": 12222
    },
    {
      "entropy": 0.6077162539586425,
      "epoch": 1.124576626931415,
      "grad_norm": 0.154767245054245,
      "learning_rate": 6.251725089704665e-05,
      "loss": 0.5853,
      "mean_token_accuracy": 0.822782926261425,
      "num_tokens": 390103105.0,
      "step": 12223
    },
    {
      "entropy": 0.6045178370550275,
      "epoch": 1.1246686333028562,
      "grad_norm": 0.15919719636440277,
      "learning_rate": 6.251418407090502e-05,
      "loss": 0.5967,
      "mean_token_accuracy": 0.8211998082697392,
      "num_tokens": 390135748.0,
      "step": 12224
    },
    {
      "entropy": 0.6325082806870341,
      "epoch": 1.1247606396742975,
      "grad_norm": 0.15800632536411285,
      "learning_rate": 6.251111724476339e-05,
      "loss": 0.628,
      "mean_token_accuracy": 0.8069018237292767,
      "num_tokens": 390168316.0,
      "step": 12225
    },
    {
      "entropy": 0.5615962669253349,
      "epoch": 1.1248526460457386,
      "grad_norm": 0.15300825238227844,
      "learning_rate": 6.250805041862177e-05,
      "loss": 0.5508,
      "mean_token_accuracy": 0.8291165083646774,
      "num_tokens": 390200338.0,
      "step": 12226
    },
    {
      "entropy": 0.6629042997956276,
      "epoch": 1.12494465241718,
      "grad_norm": 0.16105742752552032,
      "learning_rate": 6.250498359248015e-05,
      "loss": 0.6545,
      "mean_token_accuracy": 0.7970714196562767,
      "num_tokens": 390231864.0,
      "step": 12227
    },
    {
      "entropy": 0.7259157840162516,
      "epoch": 1.1250366587886211,
      "grad_norm": 0.16760748624801636,
      "learning_rate": 6.250191676633852e-05,
      "loss": 0.7279,
      "mean_token_accuracy": 0.7833708822727203,
      "num_tokens": 390263620.0,
      "step": 12228
    },
    {
      "entropy": 0.5146128749474883,
      "epoch": 1.1251286651600623,
      "grad_norm": 0.14703534543514252,
      "learning_rate": 6.249884994019689e-05,
      "loss": 0.4988,
      "mean_token_accuracy": 0.8434556610882282,
      "num_tokens": 390295923.0,
      "step": 12229
    },
    {
      "entropy": 0.6674160938709974,
      "epoch": 1.1252206715315036,
      "grad_norm": 0.172599196434021,
      "learning_rate": 6.249578311405526e-05,
      "loss": 0.6525,
      "mean_token_accuracy": 0.796290498226881,
      "num_tokens": 390328071.0,
      "step": 12230
    },
    {
      "entropy": 0.6661576898768544,
      "epoch": 1.1253126779029448,
      "grad_norm": 0.16179761290550232,
      "learning_rate": 6.249271628791365e-05,
      "loss": 0.6548,
      "mean_token_accuracy": 0.799326665699482,
      "num_tokens": 390358931.0,
      "step": 12231
    },
    {
      "entropy": 0.6105111055076122,
      "epoch": 1.1254046842743861,
      "grad_norm": 0.16021029651165009,
      "learning_rate": 6.248964946177202e-05,
      "loss": 0.6094,
      "mean_token_accuracy": 0.8159906603395939,
      "num_tokens": 390391289.0,
      "step": 12232
    },
    {
      "entropy": 0.5769012458622456,
      "epoch": 1.1254966906458272,
      "grad_norm": 0.1568443924188614,
      "learning_rate": 6.248658263563039e-05,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8283979035913944,
      "num_tokens": 390423423.0,
      "step": 12233
    },
    {
      "entropy": 0.5749664455652237,
      "epoch": 1.1255886970172684,
      "grad_norm": 0.15459245443344116,
      "learning_rate": 6.248351580948877e-05,
      "loss": 0.5632,
      "mean_token_accuracy": 0.8255990743637085,
      "num_tokens": 390455385.0,
      "step": 12234
    },
    {
      "entropy": 0.5859174244105816,
      "epoch": 1.1256807033887097,
      "grad_norm": 0.1622537225484848,
      "learning_rate": 6.248044898334714e-05,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8236128054559231,
      "num_tokens": 390487360.0,
      "step": 12235
    },
    {
      "entropy": 0.5148732149973512,
      "epoch": 1.1257727097601509,
      "grad_norm": 0.14699053764343262,
      "learning_rate": 6.247738215720552e-05,
      "loss": 0.4954,
      "mean_token_accuracy": 0.8430797047913074,
      "num_tokens": 390518705.0,
      "step": 12236
    },
    {
      "entropy": 0.6685750521719456,
      "epoch": 1.1258647161315922,
      "grad_norm": 0.16246701776981354,
      "learning_rate": 6.247431533106388e-05,
      "loss": 0.6447,
      "mean_token_accuracy": 0.805591206997633,
      "num_tokens": 390550450.0,
      "step": 12237
    },
    {
      "entropy": 0.6810858305543661,
      "epoch": 1.1259567225030334,
      "grad_norm": 0.16081364452838898,
      "learning_rate": 6.247124850492227e-05,
      "loss": 0.6732,
      "mean_token_accuracy": 0.7930239252746105,
      "num_tokens": 390582695.0,
      "step": 12238
    },
    {
      "entropy": 0.6287307413294911,
      "epoch": 1.1260487288744745,
      "grad_norm": 0.16652996838092804,
      "learning_rate": 6.246818167878063e-05,
      "loss": 0.61,
      "mean_token_accuracy": 0.8160664066672325,
      "num_tokens": 390614726.0,
      "step": 12239
    },
    {
      "entropy": 0.5816456750035286,
      "epoch": 1.1261407352459158,
      "grad_norm": 0.16290725767612457,
      "learning_rate": 6.2465114852639e-05,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8268264979124069,
      "num_tokens": 390645955.0,
      "step": 12240
    },
    {
      "entropy": 0.5464858710765839,
      "epoch": 1.126232741617357,
      "grad_norm": 0.15299299359321594,
      "learning_rate": 6.246204802649738e-05,
      "loss": 0.5327,
      "mean_token_accuracy": 0.8365838453173637,
      "num_tokens": 390677624.0,
      "step": 12241
    },
    {
      "entropy": 0.5896412301808596,
      "epoch": 1.1263247479887983,
      "grad_norm": 0.15495622158050537,
      "learning_rate": 6.245898120035576e-05,
      "loss": 0.5957,
      "mean_token_accuracy": 0.8178696222603321,
      "num_tokens": 390710155.0,
      "step": 12242
    },
    {
      "entropy": 0.6161240022629499,
      "epoch": 1.1264167543602395,
      "grad_norm": 0.15014274418354034,
      "learning_rate": 6.245591437421413e-05,
      "loss": 0.6102,
      "mean_token_accuracy": 0.8095197826623917,
      "num_tokens": 390742317.0,
      "step": 12243
    },
    {
      "entropy": 0.5838295798748732,
      "epoch": 1.1265087607316806,
      "grad_norm": 0.15939417481422424,
      "learning_rate": 6.24528475480725e-05,
      "loss": 0.5695,
      "mean_token_accuracy": 0.8251358233392239,
      "num_tokens": 390774363.0,
      "step": 12244
    },
    {
      "entropy": 0.6968494728207588,
      "epoch": 1.126600767103122,
      "grad_norm": 0.16109409928321838,
      "learning_rate": 6.244978072193087e-05,
      "loss": 0.6912,
      "mean_token_accuracy": 0.7895634137094021,
      "num_tokens": 390806749.0,
      "step": 12245
    },
    {
      "entropy": 0.6989008700475097,
      "epoch": 1.126692773474563,
      "grad_norm": 0.1662316769361496,
      "learning_rate": 6.244671389578925e-05,
      "loss": 0.6922,
      "mean_token_accuracy": 0.7898368127644062,
      "num_tokens": 390838636.0,
      "step": 12246
    },
    {
      "entropy": 0.7415228523313999,
      "epoch": 1.1267847798460044,
      "grad_norm": 0.17028383910655975,
      "learning_rate": 6.244364706964763e-05,
      "loss": 0.7454,
      "mean_token_accuracy": 0.7799937315285206,
      "num_tokens": 390871280.0,
      "step": 12247
    },
    {
      "entropy": 0.6201130882836878,
      "epoch": 1.1268767862174456,
      "grad_norm": 0.16090355813503265,
      "learning_rate": 6.2440580243506e-05,
      "loss": 0.6055,
      "mean_token_accuracy": 0.8188715316355228,
      "num_tokens": 390903324.0,
      "step": 12248
    },
    {
      "entropy": 0.6971054095774889,
      "epoch": 1.1269687925888867,
      "grad_norm": 0.1662106215953827,
      "learning_rate": 6.243751341736437e-05,
      "loss": 0.6787,
      "mean_token_accuracy": 0.7913970574736595,
      "num_tokens": 390935325.0,
      "step": 12249
    },
    {
      "entropy": 0.6025424031540751,
      "epoch": 1.127060798960328,
      "grad_norm": 0.16041292250156403,
      "learning_rate": 6.243444659122275e-05,
      "loss": 0.5765,
      "mean_token_accuracy": 0.8202284388244152,
      "num_tokens": 390967363.0,
      "step": 12250
    },
    {
      "entropy": 0.6157169044017792,
      "epoch": 1.1271528053317692,
      "grad_norm": 0.15637730062007904,
      "learning_rate": 6.243137976508112e-05,
      "loss": 0.6,
      "mean_token_accuracy": 0.8115477040410042,
      "num_tokens": 390999312.0,
      "step": 12251
    },
    {
      "entropy": 0.5608044024556875,
      "epoch": 1.1272448117032106,
      "grad_norm": 0.1550743281841278,
      "learning_rate": 6.24283129389395e-05,
      "loss": 0.5413,
      "mean_token_accuracy": 0.8320026025176048,
      "num_tokens": 391031614.0,
      "step": 12252
    },
    {
      "entropy": 0.5441435761749744,
      "epoch": 1.1273368180746517,
      "grad_norm": 0.14884009957313538,
      "learning_rate": 6.242524611279787e-05,
      "loss": 0.5189,
      "mean_token_accuracy": 0.8389177918434143,
      "num_tokens": 391063462.0,
      "step": 12253
    },
    {
      "entropy": 0.5669227130711079,
      "epoch": 1.1274288244460928,
      "grad_norm": 0.15523508191108704,
      "learning_rate": 6.242217928665625e-05,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8288879506289959,
      "num_tokens": 391095638.0,
      "step": 12254
    },
    {
      "entropy": 0.5445665931329131,
      "epoch": 1.1275208308175342,
      "grad_norm": 0.1615658551454544,
      "learning_rate": 6.241911246051461e-05,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8373457528650761,
      "num_tokens": 391127284.0,
      "step": 12255
    },
    {
      "entropy": 0.6436649551615119,
      "epoch": 1.1276128371889753,
      "grad_norm": 0.1582445353269577,
      "learning_rate": 6.241604563437298e-05,
      "loss": 0.6252,
      "mean_token_accuracy": 0.8118716813623905,
      "num_tokens": 391159558.0,
      "step": 12256
    },
    {
      "entropy": 0.6492670746520162,
      "epoch": 1.1277048435604167,
      "grad_norm": 0.15533041954040527,
      "learning_rate": 6.241297880823136e-05,
      "loss": 0.6328,
      "mean_token_accuracy": 0.8028708770871162,
      "num_tokens": 391191680.0,
      "step": 12257
    },
    {
      "entropy": 0.5094097340479493,
      "epoch": 1.1277968499318578,
      "grad_norm": 0.14867985248565674,
      "learning_rate": 6.240991198208975e-05,
      "loss": 0.492,
      "mean_token_accuracy": 0.8480989597737789,
      "num_tokens": 391224399.0,
      "step": 12258
    },
    {
      "entropy": 0.6849898025393486,
      "epoch": 1.127888856303299,
      "grad_norm": 0.1595989167690277,
      "learning_rate": 6.240684515594811e-05,
      "loss": 0.6627,
      "mean_token_accuracy": 0.7951829880475998,
      "num_tokens": 391256506.0,
      "step": 12259
    },
    {
      "entropy": 0.5531642870046198,
      "epoch": 1.1279808626747403,
      "grad_norm": 0.1533711701631546,
      "learning_rate": 6.240377832980648e-05,
      "loss": 0.5253,
      "mean_token_accuracy": 0.8372022397816181,
      "num_tokens": 391289166.0,
      "step": 12260
    },
    {
      "entropy": 0.6410122103989124,
      "epoch": 1.1280728690461814,
      "grad_norm": 0.17411935329437256,
      "learning_rate": 6.240071150366485e-05,
      "loss": 0.632,
      "mean_token_accuracy": 0.8086837418377399,
      "num_tokens": 391321157.0,
      "step": 12261
    },
    {
      "entropy": 0.5701414011418819,
      "epoch": 1.1281648754176228,
      "grad_norm": 0.1608845293521881,
      "learning_rate": 6.239764467752323e-05,
      "loss": 0.5522,
      "mean_token_accuracy": 0.8262173868715763,
      "num_tokens": 391352263.0,
      "step": 12262
    },
    {
      "entropy": 0.6169297434389591,
      "epoch": 1.128256881789064,
      "grad_norm": 0.16567569971084595,
      "learning_rate": 6.239457785138161e-05,
      "loss": 0.6126,
      "mean_token_accuracy": 0.8083662278950214,
      "num_tokens": 391383317.0,
      "step": 12263
    },
    {
      "entropy": 0.6738693192601204,
      "epoch": 1.128348888160505,
      "grad_norm": 0.1625688076019287,
      "learning_rate": 6.239151102523998e-05,
      "loss": 0.6743,
      "mean_token_accuracy": 0.7947109267115593,
      "num_tokens": 391415486.0,
      "step": 12264
    },
    {
      "entropy": 0.635108906775713,
      "epoch": 1.1284408945319464,
      "grad_norm": 0.16503790020942688,
      "learning_rate": 6.238844419909836e-05,
      "loss": 0.6293,
      "mean_token_accuracy": 0.8061422929167747,
      "num_tokens": 391446815.0,
      "step": 12265
    },
    {
      "entropy": 0.756386436522007,
      "epoch": 1.1285329009033875,
      "grad_norm": 0.17043514549732208,
      "learning_rate": 6.238537737295673e-05,
      "loss": 0.7578,
      "mean_token_accuracy": 0.7703295536339283,
      "num_tokens": 391478638.0,
      "step": 12266
    },
    {
      "entropy": 0.6198518201708794,
      "epoch": 1.128624907274829,
      "grad_norm": 0.1685205101966858,
      "learning_rate": 6.23823105468151e-05,
      "loss": 0.6236,
      "mean_token_accuracy": 0.8107348829507828,
      "num_tokens": 391511233.0,
      "step": 12267
    },
    {
      "entropy": 0.5656331572681665,
      "epoch": 1.12871691364627,
      "grad_norm": 0.15342174470424652,
      "learning_rate": 6.237924372067348e-05,
      "loss": 0.5651,
      "mean_token_accuracy": 0.8288760334253311,
      "num_tokens": 391543361.0,
      "step": 12268
    },
    {
      "entropy": 0.7125058528035879,
      "epoch": 1.1288089200177112,
      "grad_norm": 0.16686077415943146,
      "learning_rate": 6.237617689453186e-05,
      "loss": 0.6901,
      "mean_token_accuracy": 0.7905103489756584,
      "num_tokens": 391574870.0,
      "step": 12269
    },
    {
      "entropy": 0.7230622079223394,
      "epoch": 1.1289009263891525,
      "grad_norm": 0.17126809060573578,
      "learning_rate": 6.237311006839023e-05,
      "loss": 0.7099,
      "mean_token_accuracy": 0.7864061295986176,
      "num_tokens": 391606905.0,
      "step": 12270
    },
    {
      "entropy": 0.6679568160325289,
      "epoch": 1.1289929327605936,
      "grad_norm": 0.15742310881614685,
      "learning_rate": 6.23700432422486e-05,
      "loss": 0.6547,
      "mean_token_accuracy": 0.802689041942358,
      "num_tokens": 391639118.0,
      "step": 12271
    },
    {
      "entropy": 0.6688907369971275,
      "epoch": 1.129084939132035,
      "grad_norm": 0.15886905789375305,
      "learning_rate": 6.236697641610696e-05,
      "loss": 0.6545,
      "mean_token_accuracy": 0.7996447272598743,
      "num_tokens": 391671210.0,
      "step": 12272
    },
    {
      "entropy": 0.6525553874671459,
      "epoch": 1.1291769455034761,
      "grad_norm": 0.15672554075717926,
      "learning_rate": 6.236390958996536e-05,
      "loss": 0.6266,
      "mean_token_accuracy": 0.8062328174710274,
      "num_tokens": 391702190.0,
      "step": 12273
    },
    {
      "entropy": 0.6818134430795908,
      "epoch": 1.1292689518749173,
      "grad_norm": 0.15690231323242188,
      "learning_rate": 6.236084276382373e-05,
      "loss": 0.668,
      "mean_token_accuracy": 0.7970216386020184,
      "num_tokens": 391734596.0,
      "step": 12274
    },
    {
      "entropy": 0.5972980633378029,
      "epoch": 1.1293609582463586,
      "grad_norm": 0.16030140221118927,
      "learning_rate": 6.23577759376821e-05,
      "loss": 0.6009,
      "mean_token_accuracy": 0.8130358196794987,
      "num_tokens": 391766817.0,
      "step": 12275
    },
    {
      "entropy": 0.5398918054997921,
      "epoch": 1.1294529646177998,
      "grad_norm": 0.15327294170856476,
      "learning_rate": 6.235470911154046e-05,
      "loss": 0.5197,
      "mean_token_accuracy": 0.8355904147028923,
      "num_tokens": 391797287.0,
      "step": 12276
    },
    {
      "entropy": 0.7549906354397535,
      "epoch": 1.1295449709892411,
      "grad_norm": 0.17108669877052307,
      "learning_rate": 6.235164228539884e-05,
      "loss": 0.7564,
      "mean_token_accuracy": 0.7738033346831799,
      "num_tokens": 391828951.0,
      "step": 12277
    },
    {
      "entropy": 0.7085132971405983,
      "epoch": 1.1296369773606822,
      "grad_norm": 0.16037341952323914,
      "learning_rate": 6.234857545925722e-05,
      "loss": 0.7076,
      "mean_token_accuracy": 0.7828568406403065,
      "num_tokens": 391861159.0,
      "step": 12278
    },
    {
      "entropy": 0.6033173822797835,
      "epoch": 1.1297289837321234,
      "grad_norm": 0.15771757066249847,
      "learning_rate": 6.234550863311559e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.8244365006685257,
      "num_tokens": 391891866.0,
      "step": 12279
    },
    {
      "entropy": 0.6454706005752087,
      "epoch": 1.1298209901035647,
      "grad_norm": 0.15523184835910797,
      "learning_rate": 6.234244180697396e-05,
      "loss": 0.6479,
      "mean_token_accuracy": 0.802466843277216,
      "num_tokens": 391923978.0,
      "step": 12280
    },
    {
      "entropy": 0.6947300154715776,
      "epoch": 1.1299129964750059,
      "grad_norm": 0.15530936419963837,
      "learning_rate": 6.233937498083234e-05,
      "loss": 0.6647,
      "mean_token_accuracy": 0.8013516366481781,
      "num_tokens": 391956199.0,
      "step": 12281
    },
    {
      "entropy": 0.5993171334266663,
      "epoch": 1.1300050028464472,
      "grad_norm": 0.15336376428604126,
      "learning_rate": 6.233630815469071e-05,
      "loss": 0.5874,
      "mean_token_accuracy": 0.8163923919200897,
      "num_tokens": 391987547.0,
      "step": 12282
    },
    {
      "entropy": 0.6359727792441845,
      "epoch": 1.1300970092178884,
      "grad_norm": 0.16467930376529694,
      "learning_rate": 6.233324132854909e-05,
      "loss": 0.6234,
      "mean_token_accuracy": 0.8121641390025616,
      "num_tokens": 392020315.0,
      "step": 12283
    },
    {
      "entropy": 0.669388061389327,
      "epoch": 1.1301890155893295,
      "grad_norm": 0.16219037771224976,
      "learning_rate": 6.233017450240746e-05,
      "loss": 0.654,
      "mean_token_accuracy": 0.8029518499970436,
      "num_tokens": 392052290.0,
      "step": 12284
    },
    {
      "entropy": 0.6203744802623987,
      "epoch": 1.1302810219607708,
      "grad_norm": 0.17093099653720856,
      "learning_rate": 6.232710767626584e-05,
      "loss": 0.6204,
      "mean_token_accuracy": 0.8112956024706364,
      "num_tokens": 392084605.0,
      "step": 12285
    },
    {
      "entropy": 0.6478435602039099,
      "epoch": 1.130373028332212,
      "grad_norm": 0.1681499481201172,
      "learning_rate": 6.232404085012421e-05,
      "loss": 0.6482,
      "mean_token_accuracy": 0.8050568625330925,
      "num_tokens": 392116152.0,
      "step": 12286
    },
    {
      "entropy": 0.5536952610127628,
      "epoch": 1.1304650347036533,
      "grad_norm": 0.15010304749011993,
      "learning_rate": 6.232097402398258e-05,
      "loss": 0.5475,
      "mean_token_accuracy": 0.8329035602509975,
      "num_tokens": 392148519.0,
      "step": 12287
    },
    {
      "entropy": 0.592045309022069,
      "epoch": 1.1305570410750945,
      "grad_norm": 0.1644749492406845,
      "learning_rate": 6.231790719784096e-05,
      "loss": 0.5935,
      "mean_token_accuracy": 0.8201674669981003,
      "num_tokens": 392181137.0,
      "step": 12288
    },
    {
      "entropy": 0.6090902891010046,
      "epoch": 1.1306490474465356,
      "grad_norm": 0.15194542706012726,
      "learning_rate": 6.231484037169934e-05,
      "loss": 0.5905,
      "mean_token_accuracy": 0.8170220665633678,
      "num_tokens": 392212525.0,
      "step": 12289
    },
    {
      "entropy": 0.7356106955558062,
      "epoch": 1.130741053817977,
      "grad_norm": 0.16237938404083252,
      "learning_rate": 6.231177354555771e-05,
      "loss": 0.7234,
      "mean_token_accuracy": 0.7812386862933636,
      "num_tokens": 392244301.0,
      "step": 12290
    },
    {
      "entropy": 0.6696268729865551,
      "epoch": 1.130833060189418,
      "grad_norm": 0.17072536051273346,
      "learning_rate": 6.230870671941608e-05,
      "loss": 0.6566,
      "mean_token_accuracy": 0.7986288517713547,
      "num_tokens": 392275808.0,
      "step": 12291
    },
    {
      "entropy": 0.5718462448567152,
      "epoch": 1.1309250665608594,
      "grad_norm": 0.163856640458107,
      "learning_rate": 6.230563989327444e-05,
      "loss": 0.5495,
      "mean_token_accuracy": 0.8313373439013958,
      "num_tokens": 392307974.0,
      "step": 12292
    },
    {
      "entropy": 0.5822297548875213,
      "epoch": 1.1310170729323006,
      "grad_norm": 0.16134408116340637,
      "learning_rate": 6.230257306713282e-05,
      "loss": 0.5579,
      "mean_token_accuracy": 0.8248727954924107,
      "num_tokens": 392338496.0,
      "step": 12293
    },
    {
      "entropy": 0.6565506514161825,
      "epoch": 1.1311090793037417,
      "grad_norm": 0.15848511457443237,
      "learning_rate": 6.22995062409912e-05,
      "loss": 0.647,
      "mean_token_accuracy": 0.8036904372274876,
      "num_tokens": 392370505.0,
      "step": 12294
    },
    {
      "entropy": 0.6814506938681006,
      "epoch": 1.131201085675183,
      "grad_norm": 0.16233666241168976,
      "learning_rate": 6.229643941484957e-05,
      "loss": 0.6496,
      "mean_token_accuracy": 0.7972255758941174,
      "num_tokens": 392402625.0,
      "step": 12295
    },
    {
      "entropy": 0.5858832020312548,
      "epoch": 1.1312930920466242,
      "grad_norm": 0.1519584357738495,
      "learning_rate": 6.229337258870796e-05,
      "loss": 0.5691,
      "mean_token_accuracy": 0.822447158396244,
      "num_tokens": 392434005.0,
      "step": 12296
    },
    {
      "entropy": 0.6114946603775024,
      "epoch": 1.1313850984180656,
      "grad_norm": 0.15280985832214355,
      "learning_rate": 6.229030576256632e-05,
      "loss": 0.5771,
      "mean_token_accuracy": 0.8222698383033276,
      "num_tokens": 392466222.0,
      "step": 12297
    },
    {
      "entropy": 0.7092425469309092,
      "epoch": 1.1314771047895067,
      "grad_norm": 0.17062333226203918,
      "learning_rate": 6.228723893642469e-05,
      "loss": 0.6947,
      "mean_token_accuracy": 0.78279609978199,
      "num_tokens": 392497873.0,
      "step": 12298
    },
    {
      "entropy": 0.6452382635325193,
      "epoch": 1.1315691111609478,
      "grad_norm": 0.165867418050766,
      "learning_rate": 6.228417211028307e-05,
      "loss": 0.6268,
      "mean_token_accuracy": 0.8062387481331825,
      "num_tokens": 392530209.0,
      "step": 12299
    },
    {
      "entropy": 0.6940925903618336,
      "epoch": 1.1316611175323892,
      "grad_norm": 0.16175763309001923,
      "learning_rate": 6.228110528414145e-05,
      "loss": 0.6701,
      "mean_token_accuracy": 0.7976093515753746,
      "num_tokens": 392561802.0,
      "step": 12300
    },
    {
      "entropy": 0.6780228149145842,
      "epoch": 1.1317531239038303,
      "grad_norm": 0.16283129155635834,
      "learning_rate": 6.227803845799982e-05,
      "loss": 0.6599,
      "mean_token_accuracy": 0.8007563203573227,
      "num_tokens": 392593687.0,
      "step": 12301
    },
    {
      "entropy": 0.549600345082581,
      "epoch": 1.1318451302752717,
      "grad_norm": 0.15096589922904968,
      "learning_rate": 6.227497163185819e-05,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8328261785209179,
      "num_tokens": 392626145.0,
      "step": 12302
    },
    {
      "entropy": 0.6032944666221738,
      "epoch": 1.1319371366467128,
      "grad_norm": 0.1649762988090515,
      "learning_rate": 6.227190480571656e-05,
      "loss": 0.5887,
      "mean_token_accuracy": 0.8194115571677685,
      "num_tokens": 392658168.0,
      "step": 12303
    },
    {
      "entropy": 0.6156762689352036,
      "epoch": 1.132029143018154,
      "grad_norm": 0.1598474234342575,
      "learning_rate": 6.226883797957495e-05,
      "loss": 0.6075,
      "mean_token_accuracy": 0.8116383142769337,
      "num_tokens": 392689879.0,
      "step": 12304
    },
    {
      "entropy": 0.6119699608534575,
      "epoch": 1.1321211493895953,
      "grad_norm": 0.16154032945632935,
      "learning_rate": 6.226577115343332e-05,
      "loss": 0.6072,
      "mean_token_accuracy": 0.8156466633081436,
      "num_tokens": 392721716.0,
      "step": 12305
    },
    {
      "entropy": 0.725898589938879,
      "epoch": 1.1322131557610364,
      "grad_norm": 0.16482435166835785,
      "learning_rate": 6.226270432729169e-05,
      "loss": 0.7116,
      "mean_token_accuracy": 0.7843592576682568,
      "num_tokens": 392753711.0,
      "step": 12306
    },
    {
      "entropy": 0.5690128142014146,
      "epoch": 1.1323051621324778,
      "grad_norm": 0.15671518445014954,
      "learning_rate": 6.225963750115006e-05,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8250868245959282,
      "num_tokens": 392785340.0,
      "step": 12307
    },
    {
      "entropy": 0.5909899696707726,
      "epoch": 1.132397168503919,
      "grad_norm": 0.15450182557106018,
      "learning_rate": 6.225657067500844e-05,
      "loss": 0.5838,
      "mean_token_accuracy": 0.8215879872441292,
      "num_tokens": 392817160.0,
      "step": 12308
    },
    {
      "entropy": 0.6804636958986521,
      "epoch": 1.13248917487536,
      "grad_norm": 0.16278193891048431,
      "learning_rate": 6.225350384886682e-05,
      "loss": 0.6697,
      "mean_token_accuracy": 0.7934175655245781,
      "num_tokens": 392848902.0,
      "step": 12309
    },
    {
      "entropy": 0.5952820926904678,
      "epoch": 1.1325811812468014,
      "grad_norm": 0.16166257858276367,
      "learning_rate": 6.225043702272519e-05,
      "loss": 0.5943,
      "mean_token_accuracy": 0.817878071218729,
      "num_tokens": 392880910.0,
      "step": 12310
    },
    {
      "entropy": 0.5270485626533628,
      "epoch": 1.1326731876182425,
      "grad_norm": 0.155563622713089,
      "learning_rate": 6.224737019658355e-05,
      "loss": 0.5,
      "mean_token_accuracy": 0.845644161105156,
      "num_tokens": 392913068.0,
      "step": 12311
    },
    {
      "entropy": 0.557780809700489,
      "epoch": 1.132765193989684,
      "grad_norm": 0.1632363498210907,
      "learning_rate": 6.224430337044194e-05,
      "loss": 0.5491,
      "mean_token_accuracy": 0.8315729200839996,
      "num_tokens": 392944872.0,
      "step": 12312
    },
    {
      "entropy": 0.6755315978080034,
      "epoch": 1.132857200361125,
      "grad_norm": 0.16163857281208038,
      "learning_rate": 6.22412365443003e-05,
      "loss": 0.6596,
      "mean_token_accuracy": 0.7964269742369652,
      "num_tokens": 392976581.0,
      "step": 12313
    },
    {
      "entropy": 0.7083921693265438,
      "epoch": 1.1329492067325662,
      "grad_norm": 0.17328019440174103,
      "learning_rate": 6.223816971815869e-05,
      "loss": 0.6783,
      "mean_token_accuracy": 0.7884764261543751,
      "num_tokens": 393007457.0,
      "step": 12314
    },
    {
      "entropy": 0.6149065038189292,
      "epoch": 1.1330412131040075,
      "grad_norm": 0.15218201279640198,
      "learning_rate": 6.223510289201705e-05,
      "loss": 0.5915,
      "mean_token_accuracy": 0.8180170729756355,
      "num_tokens": 393039107.0,
      "step": 12315
    },
    {
      "entropy": 0.5917170001193881,
      "epoch": 1.1331332194754486,
      "grad_norm": 0.15937185287475586,
      "learning_rate": 6.223203606587543e-05,
      "loss": 0.5812,
      "mean_token_accuracy": 0.8222680985927582,
      "num_tokens": 393071059.0,
      "step": 12316
    },
    {
      "entropy": 0.645594896748662,
      "epoch": 1.13322522584689,
      "grad_norm": 0.163922518491745,
      "learning_rate": 6.22289692397338e-05,
      "loss": 0.6242,
      "mean_token_accuracy": 0.8085721917450428,
      "num_tokens": 393102602.0,
      "step": 12317
    },
    {
      "entropy": 0.6096079703420401,
      "epoch": 1.1333172322183311,
      "grad_norm": 0.15040433406829834,
      "learning_rate": 6.222590241359217e-05,
      "loss": 0.6009,
      "mean_token_accuracy": 0.81539436429739,
      "num_tokens": 393134835.0,
      "step": 12318
    },
    {
      "entropy": 0.7109447922557592,
      "epoch": 1.1334092385897723,
      "grad_norm": 0.16296744346618652,
      "learning_rate": 6.222283558745055e-05,
      "loss": 0.701,
      "mean_token_accuracy": 0.7903465665876865,
      "num_tokens": 393166846.0,
      "step": 12319
    },
    {
      "entropy": 0.47809869796037674,
      "epoch": 1.1335012449612136,
      "grad_norm": 0.14495256543159485,
      "learning_rate": 6.221976876130893e-05,
      "loss": 0.4629,
      "mean_token_accuracy": 0.8533494099974632,
      "num_tokens": 393199161.0,
      "step": 12320
    },
    {
      "entropy": 0.6820257194340229,
      "epoch": 1.1335932513326548,
      "grad_norm": 0.1631256341934204,
      "learning_rate": 6.22167019351673e-05,
      "loss": 0.6646,
      "mean_token_accuracy": 0.7959784865379333,
      "num_tokens": 393231242.0,
      "step": 12321
    },
    {
      "entropy": 0.6646714527159929,
      "epoch": 1.1336852577040961,
      "grad_norm": 0.1563953310251236,
      "learning_rate": 6.221363510902567e-05,
      "loss": 0.6422,
      "mean_token_accuracy": 0.8023679293692112,
      "num_tokens": 393263487.0,
      "step": 12322
    },
    {
      "entropy": 0.5923055149614811,
      "epoch": 1.1337772640755372,
      "grad_norm": 0.15488485991954803,
      "learning_rate": 6.221056828288404e-05,
      "loss": 0.5726,
      "mean_token_accuracy": 0.8231687434017658,
      "num_tokens": 393295586.0,
      "step": 12323
    },
    {
      "entropy": 0.6159800440073013,
      "epoch": 1.1338692704469784,
      "grad_norm": 0.16243302822113037,
      "learning_rate": 6.220750145674242e-05,
      "loss": 0.6039,
      "mean_token_accuracy": 0.8157337754964828,
      "num_tokens": 393328026.0,
      "step": 12324
    },
    {
      "entropy": 0.7097544008865952,
      "epoch": 1.1339612768184197,
      "grad_norm": 0.16421106457710266,
      "learning_rate": 6.22044346306008e-05,
      "loss": 0.6928,
      "mean_token_accuracy": 0.7934701964259148,
      "num_tokens": 393360137.0,
      "step": 12325
    },
    {
      "entropy": 0.6259119063615799,
      "epoch": 1.1340532831898609,
      "grad_norm": 0.15588416159152985,
      "learning_rate": 6.220136780445917e-05,
      "loss": 0.6239,
      "mean_token_accuracy": 0.8059559427201748,
      "num_tokens": 393391821.0,
      "step": 12326
    },
    {
      "entropy": 0.5606862790882587,
      "epoch": 1.1341452895613022,
      "grad_norm": 0.15644219517707825,
      "learning_rate": 6.219830097831755e-05,
      "loss": 0.5589,
      "mean_token_accuracy": 0.8297991566359997,
      "num_tokens": 393423829.0,
      "step": 12327
    },
    {
      "entropy": 0.6628409549593925,
      "epoch": 1.1342372959327434,
      "grad_norm": 0.1585996448993683,
      "learning_rate": 6.219523415217592e-05,
      "loss": 0.6398,
      "mean_token_accuracy": 0.8046713508665562,
      "num_tokens": 393455721.0,
      "step": 12328
    },
    {
      "entropy": 0.5748065458610654,
      "epoch": 1.1343293023041845,
      "grad_norm": 0.16057245433330536,
      "learning_rate": 6.219216732603428e-05,
      "loss": 0.5625,
      "mean_token_accuracy": 0.8266722522675991,
      "num_tokens": 393487102.0,
      "step": 12329
    },
    {
      "entropy": 0.6759069124236703,
      "epoch": 1.1344213086756259,
      "grad_norm": 0.16487883031368256,
      "learning_rate": 6.218910049989267e-05,
      "loss": 0.6596,
      "mean_token_accuracy": 0.795551523566246,
      "num_tokens": 393519257.0,
      "step": 12330
    },
    {
      "entropy": 0.6647465527057648,
      "epoch": 1.134513315047067,
      "grad_norm": 0.16044220328330994,
      "learning_rate": 6.218603367375105e-05,
      "loss": 0.6716,
      "mean_token_accuracy": 0.8005103543400764,
      "num_tokens": 393551520.0,
      "step": 12331
    },
    {
      "entropy": 0.6527817221358418,
      "epoch": 1.1346053214185083,
      "grad_norm": 0.16175054013729095,
      "learning_rate": 6.218296684760942e-05,
      "loss": 0.645,
      "mean_token_accuracy": 0.7996819466352463,
      "num_tokens": 393583598.0,
      "step": 12332
    },
    {
      "entropy": 0.5309668425470591,
      "epoch": 1.1346973277899495,
      "grad_norm": 0.1475021094083786,
      "learning_rate": 6.217990002146778e-05,
      "loss": 0.5242,
      "mean_token_accuracy": 0.8411750048398972,
      "num_tokens": 393614884.0,
      "step": 12333
    },
    {
      "entropy": 0.6677129622548819,
      "epoch": 1.1347893341613906,
      "grad_norm": 0.1703447550535202,
      "learning_rate": 6.217683319532615e-05,
      "loss": 0.667,
      "mean_token_accuracy": 0.7977585159242153,
      "num_tokens": 393646487.0,
      "step": 12334
    },
    {
      "entropy": 0.6559971887618303,
      "epoch": 1.134881340532832,
      "grad_norm": 0.15778139233589172,
      "learning_rate": 6.217376636918453e-05,
      "loss": 0.6561,
      "mean_token_accuracy": 0.7999578714370728,
      "num_tokens": 393678910.0,
      "step": 12335
    },
    {
      "entropy": 0.7940126471221447,
      "epoch": 1.134973346904273,
      "grad_norm": 0.16828560829162598,
      "learning_rate": 6.217069954304291e-05,
      "loss": 0.7909,
      "mean_token_accuracy": 0.7593330852687359,
      "num_tokens": 393710642.0,
      "step": 12336
    },
    {
      "entropy": 0.6143331676721573,
      "epoch": 1.1350653532757145,
      "grad_norm": 0.15553882718086243,
      "learning_rate": 6.216763271690128e-05,
      "loss": 0.5983,
      "mean_token_accuracy": 0.8199006617069244,
      "num_tokens": 393743088.0,
      "step": 12337
    },
    {
      "entropy": 0.7092504762113094,
      "epoch": 1.1351573596471556,
      "grad_norm": 0.16143575310707092,
      "learning_rate": 6.216456589075965e-05,
      "loss": 0.6979,
      "mean_token_accuracy": 0.7874245569109917,
      "num_tokens": 393775821.0,
      "step": 12338
    },
    {
      "entropy": 0.623607499524951,
      "epoch": 1.1352493660185967,
      "grad_norm": 0.16616837680339813,
      "learning_rate": 6.216149906461803e-05,
      "loss": 0.6001,
      "mean_token_accuracy": 0.8166167177259922,
      "num_tokens": 393807521.0,
      "step": 12339
    },
    {
      "entropy": 0.47232686914503574,
      "epoch": 1.135341372390038,
      "grad_norm": 0.15281306207180023,
      "learning_rate": 6.21584322384764e-05,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8569556064903736,
      "num_tokens": 393839926.0,
      "step": 12340
    },
    {
      "entropy": 0.6409442238509655,
      "epoch": 1.1354333787614792,
      "grad_norm": 0.1584843397140503,
      "learning_rate": 6.215536541233478e-05,
      "loss": 0.6046,
      "mean_token_accuracy": 0.811288133263588,
      "num_tokens": 393872352.0,
      "step": 12341
    },
    {
      "entropy": 0.5975676514208317,
      "epoch": 1.1355253851329206,
      "grad_norm": 0.1589736044406891,
      "learning_rate": 6.215229858619315e-05,
      "loss": 0.5804,
      "mean_token_accuracy": 0.8229115717113018,
      "num_tokens": 393905120.0,
      "step": 12342
    },
    {
      "entropy": 0.7488523554056883,
      "epoch": 1.1356173915043617,
      "grad_norm": 0.16878309845924377,
      "learning_rate": 6.214923176005153e-05,
      "loss": 0.7575,
      "mean_token_accuracy": 0.7745724394917488,
      "num_tokens": 393936891.0,
      "step": 12343
    },
    {
      "entropy": 0.5846832673996687,
      "epoch": 1.1357093978758028,
      "grad_norm": 0.15462666749954224,
      "learning_rate": 6.21461649339099e-05,
      "loss": 0.5698,
      "mean_token_accuracy": 0.8257516659796238,
      "num_tokens": 393968738.0,
      "step": 12344
    },
    {
      "entropy": 0.6460323520004749,
      "epoch": 1.1358014042472442,
      "grad_norm": 0.1643526554107666,
      "learning_rate": 6.214309810776827e-05,
      "loss": 0.637,
      "mean_token_accuracy": 0.8036103807389736,
      "num_tokens": 394000792.0,
      "step": 12345
    },
    {
      "entropy": 0.5478709973394871,
      "epoch": 1.1358934106186853,
      "grad_norm": 0.15844713151454926,
      "learning_rate": 6.214003128162665e-05,
      "loss": 0.5244,
      "mean_token_accuracy": 0.8396775051951408,
      "num_tokens": 394032710.0,
      "step": 12346
    },
    {
      "entropy": 0.6726537328213453,
      "epoch": 1.1359854169901267,
      "grad_norm": 0.15965703129768372,
      "learning_rate": 6.213696445548503e-05,
      "loss": 0.6478,
      "mean_token_accuracy": 0.8027587458491325,
      "num_tokens": 394064001.0,
      "step": 12347
    },
    {
      "entropy": 0.6116947904229164,
      "epoch": 1.1360774233615678,
      "grad_norm": 0.15971821546554565,
      "learning_rate": 6.21338976293434e-05,
      "loss": 0.6073,
      "mean_token_accuracy": 0.8177916407585144,
      "num_tokens": 394095997.0,
      "step": 12348
    },
    {
      "entropy": 0.5889474134892225,
      "epoch": 1.136169429733009,
      "grad_norm": 0.1539253443479538,
      "learning_rate": 6.213083080320176e-05,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8261889554560184,
      "num_tokens": 394127908.0,
      "step": 12349
    },
    {
      "entropy": 0.6284814616665244,
      "epoch": 1.1362614361044503,
      "grad_norm": 0.15516634285449982,
      "learning_rate": 6.212776397706013e-05,
      "loss": 0.6266,
      "mean_token_accuracy": 0.8106968179345131,
      "num_tokens": 394159791.0,
      "step": 12350
    },
    {
      "entropy": 0.5618092929944396,
      "epoch": 1.1363534424758914,
      "grad_norm": 0.15653608739376068,
      "learning_rate": 6.212469715091853e-05,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8352963402867317,
      "num_tokens": 394191169.0,
      "step": 12351
    },
    {
      "entropy": 0.6074608974158764,
      "epoch": 1.1364454488473328,
      "grad_norm": 0.15839944779872894,
      "learning_rate": 6.21216303247769e-05,
      "loss": 0.5906,
      "mean_token_accuracy": 0.8172150924801826,
      "num_tokens": 394223007.0,
      "step": 12352
    },
    {
      "entropy": 0.5178135049063712,
      "epoch": 1.136537455218774,
      "grad_norm": 0.15937285125255585,
      "learning_rate": 6.211856349863526e-05,
      "loss": 0.521,
      "mean_token_accuracy": 0.8415673784911633,
      "num_tokens": 394255110.0,
      "step": 12353
    },
    {
      "entropy": 0.5548597415909171,
      "epoch": 1.136629461590215,
      "grad_norm": 0.1631753295660019,
      "learning_rate": 6.211549667249363e-05,
      "loss": 0.5315,
      "mean_token_accuracy": 0.8350154794752598,
      "num_tokens": 394286681.0,
      "step": 12354
    },
    {
      "entropy": 0.6128473374992609,
      "epoch": 1.1367214679616564,
      "grad_norm": 0.16613374650478363,
      "learning_rate": 6.211242984635201e-05,
      "loss": 0.5962,
      "mean_token_accuracy": 0.8200141340494156,
      "num_tokens": 394318077.0,
      "step": 12355
    },
    {
      "entropy": 0.569280824624002,
      "epoch": 1.1368134743330975,
      "grad_norm": 0.15525124967098236,
      "learning_rate": 6.21093630202104e-05,
      "loss": 0.547,
      "mean_token_accuracy": 0.8332198448479176,
      "num_tokens": 394350157.0,
      "step": 12356
    },
    {
      "entropy": 0.6814850363880396,
      "epoch": 1.136905480704539,
      "grad_norm": 0.164445698261261,
      "learning_rate": 6.210629619406876e-05,
      "loss": 0.6776,
      "mean_token_accuracy": 0.7943492047488689,
      "num_tokens": 394382144.0,
      "step": 12357
    },
    {
      "entropy": 0.5923316897824407,
      "epoch": 1.13699748707598,
      "grad_norm": 0.15584328770637512,
      "learning_rate": 6.210322936792714e-05,
      "loss": 0.5602,
      "mean_token_accuracy": 0.829811342060566,
      "num_tokens": 394414839.0,
      "step": 12358
    },
    {
      "entropy": 0.6012179283425212,
      "epoch": 1.1370894934474212,
      "grad_norm": 0.158112034201622,
      "learning_rate": 6.210016254178551e-05,
      "loss": 0.5766,
      "mean_token_accuracy": 0.8261687010526657,
      "num_tokens": 394447549.0,
      "step": 12359
    },
    {
      "entropy": 0.5495352195575833,
      "epoch": 1.1371814998188625,
      "grad_norm": 0.1625681221485138,
      "learning_rate": 6.209709571564388e-05,
      "loss": 0.5278,
      "mean_token_accuracy": 0.836388636380434,
      "num_tokens": 394480095.0,
      "step": 12360
    },
    {
      "entropy": 0.5346544748172164,
      "epoch": 1.1372735061903037,
      "grad_norm": 0.1533183753490448,
      "learning_rate": 6.209402888950226e-05,
      "loss": 0.533,
      "mean_token_accuracy": 0.8325377479195595,
      "num_tokens": 394512162.0,
      "step": 12361
    },
    {
      "entropy": 0.6372593753039837,
      "epoch": 1.137365512561745,
      "grad_norm": 0.16514630615711212,
      "learning_rate": 6.209096206336064e-05,
      "loss": 0.6227,
      "mean_token_accuracy": 0.8112387247383595,
      "num_tokens": 394544538.0,
      "step": 12362
    },
    {
      "entropy": 0.636776078492403,
      "epoch": 1.1374575189331861,
      "grad_norm": 0.15730862319469452,
      "learning_rate": 6.208789523721901e-05,
      "loss": 0.6248,
      "mean_token_accuracy": 0.8073217160999775,
      "num_tokens": 394576703.0,
      "step": 12363
    },
    {
      "entropy": 0.5491024050861597,
      "epoch": 1.1375495253046273,
      "grad_norm": 0.15518303215503693,
      "learning_rate": 6.208482841107738e-05,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8334970772266388,
      "num_tokens": 394608784.0,
      "step": 12364
    },
    {
      "entropy": 0.5659561809152365,
      "epoch": 1.1376415316760686,
      "grad_norm": 0.15096664428710938,
      "learning_rate": 6.208176158493574e-05,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8334191478788853,
      "num_tokens": 394641295.0,
      "step": 12365
    },
    {
      "entropy": 0.7255323380231857,
      "epoch": 1.1377335380475098,
      "grad_norm": 0.1694958359003067,
      "learning_rate": 6.207869475879413e-05,
      "loss": 0.7274,
      "mean_token_accuracy": 0.7827072367072105,
      "num_tokens": 394673563.0,
      "step": 12366
    },
    {
      "entropy": 0.6965818954631686,
      "epoch": 1.1378255444189511,
      "grad_norm": 0.16082830727100372,
      "learning_rate": 6.207562793265251e-05,
      "loss": 0.6954,
      "mean_token_accuracy": 0.7916997969150543,
      "num_tokens": 394705939.0,
      "step": 12367
    },
    {
      "entropy": 0.6449702396057546,
      "epoch": 1.1379175507903923,
      "grad_norm": 0.16915425658226013,
      "learning_rate": 6.207256110651088e-05,
      "loss": 0.6404,
      "mean_token_accuracy": 0.8040197491645813,
      "num_tokens": 394737170.0,
      "step": 12368
    },
    {
      "entropy": 0.623005174100399,
      "epoch": 1.1380095571618334,
      "grad_norm": 0.15564432740211487,
      "learning_rate": 6.206949428036924e-05,
      "loss": 0.6222,
      "mean_token_accuracy": 0.8127247802913189,
      "num_tokens": 394769661.0,
      "step": 12369
    },
    {
      "entropy": 0.5177079252898693,
      "epoch": 1.1381015635332747,
      "grad_norm": 0.1447088122367859,
      "learning_rate": 6.206642745422762e-05,
      "loss": 0.4954,
      "mean_token_accuracy": 0.846622172743082,
      "num_tokens": 394801756.0,
      "step": 12370
    },
    {
      "entropy": 0.6313479994423687,
      "epoch": 1.1381935699047159,
      "grad_norm": 0.16320443153381348,
      "learning_rate": 6.206336062808599e-05,
      "loss": 0.5945,
      "mean_token_accuracy": 0.8157147988677025,
      "num_tokens": 394833117.0,
      "step": 12371
    },
    {
      "entropy": 0.6934815645217896,
      "epoch": 1.1382855762761572,
      "grad_norm": 0.1595602184534073,
      "learning_rate": 6.206029380194437e-05,
      "loss": 0.6847,
      "mean_token_accuracy": 0.7909411303699017,
      "num_tokens": 394864840.0,
      "step": 12372
    },
    {
      "entropy": 0.7203186266124249,
      "epoch": 1.1383775826475984,
      "grad_norm": 0.16784615814685822,
      "learning_rate": 6.205722697580274e-05,
      "loss": 0.722,
      "mean_token_accuracy": 0.7836071215569973,
      "num_tokens": 394896396.0,
      "step": 12373
    },
    {
      "entropy": 0.5842782338149846,
      "epoch": 1.1384695890190395,
      "grad_norm": 0.15299874544143677,
      "learning_rate": 6.205416014966112e-05,
      "loss": 0.5765,
      "mean_token_accuracy": 0.8256901353597641,
      "num_tokens": 394928333.0,
      "step": 12374
    },
    {
      "entropy": 0.49247184954583645,
      "epoch": 1.1385615953904809,
      "grad_norm": 0.1537412405014038,
      "learning_rate": 6.205109332351949e-05,
      "loss": 0.4844,
      "mean_token_accuracy": 0.8498180881142616,
      "num_tokens": 394960560.0,
      "step": 12375
    },
    {
      "entropy": 0.6812575459480286,
      "epoch": 1.138653601761922,
      "grad_norm": 0.1651483178138733,
      "learning_rate": 6.204802649737786e-05,
      "loss": 0.6623,
      "mean_token_accuracy": 0.7990099191665649,
      "num_tokens": 394992982.0,
      "step": 12376
    },
    {
      "entropy": 0.5233149817213416,
      "epoch": 1.1387456081333633,
      "grad_norm": 0.14749421179294586,
      "learning_rate": 6.204495967123624e-05,
      "loss": 0.5052,
      "mean_token_accuracy": 0.8410474695265293,
      "num_tokens": 395025335.0,
      "step": 12377
    },
    {
      "entropy": 0.6642313320189714,
      "epoch": 1.1388376145048045,
      "grad_norm": 0.15975506603717804,
      "learning_rate": 6.204189284509462e-05,
      "loss": 0.6555,
      "mean_token_accuracy": 0.8007282987236977,
      "num_tokens": 395057882.0,
      "step": 12378
    },
    {
      "entropy": 0.7056918069720268,
      "epoch": 1.1389296208762456,
      "grad_norm": 0.16620883345603943,
      "learning_rate": 6.203882601895299e-05,
      "loss": 0.7022,
      "mean_token_accuracy": 0.7887178026139736,
      "num_tokens": 395089403.0,
      "step": 12379
    },
    {
      "entropy": 0.6630691699683666,
      "epoch": 1.139021627247687,
      "grad_norm": 0.16039837896823883,
      "learning_rate": 6.203575919281136e-05,
      "loss": 0.6589,
      "mean_token_accuracy": 0.8020371422171593,
      "num_tokens": 395121803.0,
      "step": 12380
    },
    {
      "entropy": 0.5776183507405221,
      "epoch": 1.139113633619128,
      "grad_norm": 0.15313725173473358,
      "learning_rate": 6.203269236666973e-05,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8285687863826752,
      "num_tokens": 395153815.0,
      "step": 12381
    },
    {
      "entropy": 0.5933971544727683,
      "epoch": 1.1392056399905695,
      "grad_norm": 0.15539084374904633,
      "learning_rate": 6.202962554052811e-05,
      "loss": 0.5912,
      "mean_token_accuracy": 0.8183011710643768,
      "num_tokens": 395186583.0,
      "step": 12382
    },
    {
      "entropy": 0.6600109804421663,
      "epoch": 1.1392976463620106,
      "grad_norm": 0.16000936925411224,
      "learning_rate": 6.202655871438649e-05,
      "loss": 0.6475,
      "mean_token_accuracy": 0.8024699278175831,
      "num_tokens": 395218282.0,
      "step": 12383
    },
    {
      "entropy": 0.6397528704255819,
      "epoch": 1.1393896527334517,
      "grad_norm": 0.15698426961898804,
      "learning_rate": 6.202349188824486e-05,
      "loss": 0.6377,
      "mean_token_accuracy": 0.8022607937455177,
      "num_tokens": 395250696.0,
      "step": 12384
    },
    {
      "entropy": 0.7728920914232731,
      "epoch": 1.139481659104893,
      "grad_norm": 0.17086392641067505,
      "learning_rate": 6.202042506210324e-05,
      "loss": 0.7666,
      "mean_token_accuracy": 0.7714737839996815,
      "num_tokens": 395282889.0,
      "step": 12385
    },
    {
      "entropy": 0.5656840410083532,
      "epoch": 1.1395736654763342,
      "grad_norm": 0.15899968147277832,
      "learning_rate": 6.20173582359616e-05,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8328882455825806,
      "num_tokens": 395314592.0,
      "step": 12386
    },
    {
      "entropy": 0.6959112454205751,
      "epoch": 1.1396656718477756,
      "grad_norm": 0.1642943173646927,
      "learning_rate": 6.201429140981997e-05,
      "loss": 0.6899,
      "mean_token_accuracy": 0.7895363718271255,
      "num_tokens": 395346408.0,
      "step": 12387
    },
    {
      "entropy": 0.6443367004394531,
      "epoch": 1.1397576782192167,
      "grad_norm": 0.15642262995243073,
      "learning_rate": 6.201122458367835e-05,
      "loss": 0.6335,
      "mean_token_accuracy": 0.8100607134401798,
      "num_tokens": 395378415.0,
      "step": 12388
    },
    {
      "entropy": 0.6084187878295779,
      "epoch": 1.1398496845906578,
      "grad_norm": 0.16397123038768768,
      "learning_rate": 6.200815775753674e-05,
      "loss": 0.596,
      "mean_token_accuracy": 0.8163613267242908,
      "num_tokens": 395410787.0,
      "step": 12389
    },
    {
      "entropy": 0.6252110991626978,
      "epoch": 1.1399416909620992,
      "grad_norm": 0.1681101769208908,
      "learning_rate": 6.20050909313951e-05,
      "loss": 0.6079,
      "mean_token_accuracy": 0.8155026882886887,
      "num_tokens": 395442642.0,
      "step": 12390
    },
    {
      "entropy": 0.6527020689100027,
      "epoch": 1.1400336973335403,
      "grad_norm": 0.1557142287492752,
      "learning_rate": 6.200202410525347e-05,
      "loss": 0.6353,
      "mean_token_accuracy": 0.8036376014351845,
      "num_tokens": 395474674.0,
      "step": 12391
    },
    {
      "entropy": 0.6265429016202688,
      "epoch": 1.1401257037049817,
      "grad_norm": 0.17157284915447235,
      "learning_rate": 6.199895727911185e-05,
      "loss": 0.6184,
      "mean_token_accuracy": 0.8115414418280125,
      "num_tokens": 395506699.0,
      "step": 12392
    },
    {
      "entropy": 0.6391171962022781,
      "epoch": 1.1402177100764228,
      "grad_norm": 0.16240745782852173,
      "learning_rate": 6.199589045297023e-05,
      "loss": 0.6246,
      "mean_token_accuracy": 0.8117387555539608,
      "num_tokens": 395538451.0,
      "step": 12393
    },
    {
      "entropy": 0.6765471533872187,
      "epoch": 1.140309716447864,
      "grad_norm": 0.15574510395526886,
      "learning_rate": 6.19928236268286e-05,
      "loss": 0.6784,
      "mean_token_accuracy": 0.7974761202931404,
      "num_tokens": 395570761.0,
      "step": 12394
    },
    {
      "entropy": 0.5132903782650828,
      "epoch": 1.1404017228193053,
      "grad_norm": 0.1527033895254135,
      "learning_rate": 6.198975680068697e-05,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8457493036985397,
      "num_tokens": 395602243.0,
      "step": 12395
    },
    {
      "entropy": 0.7308641150593758,
      "epoch": 1.1404937291907464,
      "grad_norm": 0.16139931976795197,
      "learning_rate": 6.198668997454534e-05,
      "loss": 0.7161,
      "mean_token_accuracy": 0.7823752611875534,
      "num_tokens": 395634149.0,
      "step": 12396
    },
    {
      "entropy": 0.6607214901596308,
      "epoch": 1.1405857355621878,
      "grad_norm": 0.15887364745140076,
      "learning_rate": 6.198362314840372e-05,
      "loss": 0.6374,
      "mean_token_accuracy": 0.8045762553811073,
      "num_tokens": 395665038.0,
      "step": 12397
    },
    {
      "entropy": 0.6812430191785097,
      "epoch": 1.140677741933629,
      "grad_norm": 0.15610195696353912,
      "learning_rate": 6.19805563222621e-05,
      "loss": 0.6661,
      "mean_token_accuracy": 0.8014470711350441,
      "num_tokens": 395697399.0,
      "step": 12398
    },
    {
      "entropy": 0.6197273870930076,
      "epoch": 1.14076974830507,
      "grad_norm": 0.15428097546100616,
      "learning_rate": 6.197748949612047e-05,
      "loss": 0.6039,
      "mean_token_accuracy": 0.8158012628555298,
      "num_tokens": 395729387.0,
      "step": 12399
    },
    {
      "entropy": 0.6714733485132456,
      "epoch": 1.1408617546765114,
      "grad_norm": 0.16260893642902374,
      "learning_rate": 6.197442266997884e-05,
      "loss": 0.6502,
      "mean_token_accuracy": 0.8007790222764015,
      "num_tokens": 395760688.0,
      "step": 12400
    },
    {
      "entropy": 0.6364411385729909,
      "epoch": 1.1409537610479525,
      "grad_norm": 0.16774992644786835,
      "learning_rate": 6.197135584383722e-05,
      "loss": 0.6145,
      "mean_token_accuracy": 0.8105788603425026,
      "num_tokens": 395792550.0,
      "step": 12401
    },
    {
      "entropy": 0.6053524827584624,
      "epoch": 1.141045767419394,
      "grad_norm": 0.15748253464698792,
      "learning_rate": 6.196828901769559e-05,
      "loss": 0.5907,
      "mean_token_accuracy": 0.8182736672461033,
      "num_tokens": 395824994.0,
      "step": 12402
    },
    {
      "entropy": 0.6314315591007471,
      "epoch": 1.141137773790835,
      "grad_norm": 0.16423700749874115,
      "learning_rate": 6.196522219155397e-05,
      "loss": 0.6295,
      "mean_token_accuracy": 0.8057046234607697,
      "num_tokens": 395857043.0,
      "step": 12403
    },
    {
      "entropy": 0.626120975241065,
      "epoch": 1.1412297801622762,
      "grad_norm": 0.15485729277133942,
      "learning_rate": 6.196215536541234e-05,
      "loss": 0.6101,
      "mean_token_accuracy": 0.8086957708001137,
      "num_tokens": 395888750.0,
      "step": 12404
    },
    {
      "entropy": 0.5744456993415952,
      "epoch": 1.1413217865337175,
      "grad_norm": 0.15394450724124908,
      "learning_rate": 6.195908853927072e-05,
      "loss": 0.566,
      "mean_token_accuracy": 0.8244447112083435,
      "num_tokens": 395920743.0,
      "step": 12405
    },
    {
      "entropy": 0.6690436210483313,
      "epoch": 1.1414137929051587,
      "grad_norm": 0.17307709157466888,
      "learning_rate": 6.195602171312908e-05,
      "loss": 0.6623,
      "mean_token_accuracy": 0.8025307431817055,
      "num_tokens": 395952462.0,
      "step": 12406
    },
    {
      "entropy": 0.5079704192467034,
      "epoch": 1.1415057992766,
      "grad_norm": 0.1569303274154663,
      "learning_rate": 6.195295488698745e-05,
      "loss": 0.5005,
      "mean_token_accuracy": 0.8436812199652195,
      "num_tokens": 395984578.0,
      "step": 12407
    },
    {
      "entropy": 0.6563872192054987,
      "epoch": 1.1415978056480411,
      "grad_norm": 0.164134219288826,
      "learning_rate": 6.194988806084583e-05,
      "loss": 0.6393,
      "mean_token_accuracy": 0.8059910237789154,
      "num_tokens": 396015527.0,
      "step": 12408
    },
    {
      "entropy": 0.6510522160679102,
      "epoch": 1.1416898120194823,
      "grad_norm": 0.16496874392032623,
      "learning_rate": 6.194682123470422e-05,
      "loss": 0.6432,
      "mean_token_accuracy": 0.805675595998764,
      "num_tokens": 396047430.0,
      "step": 12409
    },
    {
      "entropy": 0.5234911367297173,
      "epoch": 1.1417818183909236,
      "grad_norm": 0.15643589198589325,
      "learning_rate": 6.194375440856258e-05,
      "loss": 0.5166,
      "mean_token_accuracy": 0.84093327075243,
      "num_tokens": 396079347.0,
      "step": 12410
    },
    {
      "entropy": 0.5810280684381723,
      "epoch": 1.1418738247623648,
      "grad_norm": 0.14827196300029755,
      "learning_rate": 6.194068758242095e-05,
      "loss": 0.5632,
      "mean_token_accuracy": 0.8262479528784752,
      "num_tokens": 396111392.0,
      "step": 12411
    },
    {
      "entropy": 0.6840307265520096,
      "epoch": 1.1419658311338061,
      "grad_norm": 0.16011863946914673,
      "learning_rate": 6.193762075627932e-05,
      "loss": 0.6995,
      "mean_token_accuracy": 0.7873546667397022,
      "num_tokens": 396143632.0,
      "step": 12412
    },
    {
      "entropy": 0.7505455948412418,
      "epoch": 1.1420578375052473,
      "grad_norm": 0.16815003752708435,
      "learning_rate": 6.19345539301377e-05,
      "loss": 0.7472,
      "mean_token_accuracy": 0.7729654647409916,
      "num_tokens": 396175456.0,
      "step": 12413
    },
    {
      "entropy": 0.602573552634567,
      "epoch": 1.1421498438766884,
      "grad_norm": 0.16156314313411713,
      "learning_rate": 6.193148710399608e-05,
      "loss": 0.584,
      "mean_token_accuracy": 0.8187119178473949,
      "num_tokens": 396207025.0,
      "step": 12414
    },
    {
      "entropy": 0.614264965057373,
      "epoch": 1.1422418502481297,
      "grad_norm": 0.1573498249053955,
      "learning_rate": 6.192842027785445e-05,
      "loss": 0.6035,
      "mean_token_accuracy": 0.816615529358387,
      "num_tokens": 396238786.0,
      "step": 12415
    },
    {
      "entropy": 0.6730538504198194,
      "epoch": 1.1423338566195709,
      "grad_norm": 0.16414783895015717,
      "learning_rate": 6.192535345171283e-05,
      "loss": 0.6549,
      "mean_token_accuracy": 0.7979723960161209,
      "num_tokens": 396270507.0,
      "step": 12416
    },
    {
      "entropy": 0.6716275569051504,
      "epoch": 1.1424258629910122,
      "grad_norm": 0.16298367083072662,
      "learning_rate": 6.19222866255712e-05,
      "loss": 0.668,
      "mean_token_accuracy": 0.8001120686531067,
      "num_tokens": 396302803.0,
      "step": 12417
    },
    {
      "entropy": 0.6313197873532772,
      "epoch": 1.1425178693624534,
      "grad_norm": 0.15750151872634888,
      "learning_rate": 6.191921979942957e-05,
      "loss": 0.6045,
      "mean_token_accuracy": 0.814838096499443,
      "num_tokens": 396335077.0,
      "step": 12418
    },
    {
      "entropy": 0.674794614315033,
      "epoch": 1.1426098757338945,
      "grad_norm": 0.17322558164596558,
      "learning_rate": 6.191615297328795e-05,
      "loss": 0.6694,
      "mean_token_accuracy": 0.7997473627328873,
      "num_tokens": 396366753.0,
      "step": 12419
    },
    {
      "entropy": 0.6394802406430244,
      "epoch": 1.1427018821053359,
      "grad_norm": 0.1666450798511505,
      "learning_rate": 6.191308614714633e-05,
      "loss": 0.6407,
      "mean_token_accuracy": 0.804548192769289,
      "num_tokens": 396398223.0,
      "step": 12420
    },
    {
      "entropy": 0.5241368832066655,
      "epoch": 1.142793888476777,
      "grad_norm": 0.14741474390029907,
      "learning_rate": 6.19100193210047e-05,
      "loss": 0.5145,
      "mean_token_accuracy": 0.8422299735248089,
      "num_tokens": 396430766.0,
      "step": 12421
    },
    {
      "entropy": 0.5981762809678912,
      "epoch": 1.1428858948482183,
      "grad_norm": 0.15144409239292145,
      "learning_rate": 6.190695249486307e-05,
      "loss": 0.5935,
      "mean_token_accuracy": 0.8188077695667744,
      "num_tokens": 396462482.0,
      "step": 12422
    },
    {
      "entropy": 0.6483870055526495,
      "epoch": 1.1429779012196595,
      "grad_norm": 0.15216001868247986,
      "learning_rate": 6.190388566872143e-05,
      "loss": 0.642,
      "mean_token_accuracy": 0.8039689064025879,
      "num_tokens": 396495229.0,
      "step": 12423
    },
    {
      "entropy": 0.626111158169806,
      "epoch": 1.1430699075911006,
      "grad_norm": 0.1528548002243042,
      "learning_rate": 6.190081884257983e-05,
      "loss": 0.5906,
      "mean_token_accuracy": 0.8184879720211029,
      "num_tokens": 396526348.0,
      "step": 12424
    },
    {
      "entropy": 0.6445505935698748,
      "epoch": 1.143161913962542,
      "grad_norm": 0.1545259803533554,
      "learning_rate": 6.18977520164382e-05,
      "loss": 0.6086,
      "mean_token_accuracy": 0.8142320699989796,
      "num_tokens": 396557920.0,
      "step": 12425
    },
    {
      "entropy": 0.5231994986534119,
      "epoch": 1.143253920333983,
      "grad_norm": 0.1490478366613388,
      "learning_rate": 6.189468519029656e-05,
      "loss": 0.4945,
      "mean_token_accuracy": 0.8463121838867664,
      "num_tokens": 396590245.0,
      "step": 12426
    },
    {
      "entropy": 0.597865954041481,
      "epoch": 1.1433459267054245,
      "grad_norm": 0.15328308939933777,
      "learning_rate": 6.189161836415493e-05,
      "loss": 0.5912,
      "mean_token_accuracy": 0.8180431164801121,
      "num_tokens": 396621638.0,
      "step": 12427
    },
    {
      "entropy": 0.6105694342404604,
      "epoch": 1.1434379330768656,
      "grad_norm": 0.15902957320213318,
      "learning_rate": 6.188855153801331e-05,
      "loss": 0.6054,
      "mean_token_accuracy": 0.8103019297122955,
      "num_tokens": 396652592.0,
      "step": 12428
    },
    {
      "entropy": 0.6478136051446199,
      "epoch": 1.1435299394483067,
      "grad_norm": 0.15949390828609467,
      "learning_rate": 6.18854847118717e-05,
      "loss": 0.6427,
      "mean_token_accuracy": 0.8072457611560822,
      "num_tokens": 396684764.0,
      "step": 12429
    },
    {
      "entropy": 0.7376069463789463,
      "epoch": 1.143621945819748,
      "grad_norm": 0.16188976168632507,
      "learning_rate": 6.188241788573006e-05,
      "loss": 0.7275,
      "mean_token_accuracy": 0.7805806025862694,
      "num_tokens": 396716822.0,
      "step": 12430
    },
    {
      "entropy": 0.5785601790994406,
      "epoch": 1.1437139521911892,
      "grad_norm": 0.15297441184520721,
      "learning_rate": 6.187935105958843e-05,
      "loss": 0.5713,
      "mean_token_accuracy": 0.8203974552452564,
      "num_tokens": 396749515.0,
      "step": 12431
    },
    {
      "entropy": 0.6487518176436424,
      "epoch": 1.1438059585626306,
      "grad_norm": 0.15585385262966156,
      "learning_rate": 6.187628423344681e-05,
      "loss": 0.637,
      "mean_token_accuracy": 0.8035667091608047,
      "num_tokens": 396781738.0,
      "step": 12432
    },
    {
      "entropy": 0.6606129016727209,
      "epoch": 1.1438979649340717,
      "grad_norm": 0.17502190172672272,
      "learning_rate": 6.187321740730518e-05,
      "loss": 0.66,
      "mean_token_accuracy": 0.8012001700699329,
      "num_tokens": 396812944.0,
      "step": 12433
    },
    {
      "entropy": 0.5841666264459491,
      "epoch": 1.1439899713055128,
      "grad_norm": 0.15342099964618683,
      "learning_rate": 6.187015058116356e-05,
      "loss": 0.5586,
      "mean_token_accuracy": 0.8259284608066082,
      "num_tokens": 396845026.0,
      "step": 12434
    },
    {
      "entropy": 0.6372024081647396,
      "epoch": 1.1440819776769542,
      "grad_norm": 0.16139303147792816,
      "learning_rate": 6.186708375502193e-05,
      "loss": 0.6424,
      "mean_token_accuracy": 0.8031755350530148,
      "num_tokens": 396876875.0,
      "step": 12435
    },
    {
      "entropy": 0.5965050058439374,
      "epoch": 1.1441739840483953,
      "grad_norm": 0.15972968935966492,
      "learning_rate": 6.186401692888031e-05,
      "loss": 0.579,
      "mean_token_accuracy": 0.8212312608957291,
      "num_tokens": 396908045.0,
      "step": 12436
    },
    {
      "entropy": 0.5275425743311644,
      "epoch": 1.1442659904198367,
      "grad_norm": 0.14700770378112793,
      "learning_rate": 6.186095010273868e-05,
      "loss": 0.5171,
      "mean_token_accuracy": 0.841869805008173,
      "num_tokens": 396939487.0,
      "step": 12437
    },
    {
      "entropy": 0.6869044788181782,
      "epoch": 1.1443579967912778,
      "grad_norm": 0.16122180223464966,
      "learning_rate": 6.185788327659705e-05,
      "loss": 0.6751,
      "mean_token_accuracy": 0.793569665402174,
      "num_tokens": 396971462.0,
      "step": 12438
    },
    {
      "entropy": 0.6682485826313496,
      "epoch": 1.144450003162719,
      "grad_norm": 0.15900689363479614,
      "learning_rate": 6.185481645045543e-05,
      "loss": 0.657,
      "mean_token_accuracy": 0.800269428640604,
      "num_tokens": 397003721.0,
      "step": 12439
    },
    {
      "entropy": 0.535219163633883,
      "epoch": 1.1445420095341603,
      "grad_norm": 0.15656521916389465,
      "learning_rate": 6.185174962431381e-05,
      "loss": 0.5204,
      "mean_token_accuracy": 0.8369112685322762,
      "num_tokens": 397035415.0,
      "step": 12440
    },
    {
      "entropy": 0.6274595763534307,
      "epoch": 1.1446340159056014,
      "grad_norm": 0.1586485207080841,
      "learning_rate": 6.184868279817218e-05,
      "loss": 0.6109,
      "mean_token_accuracy": 0.8136193193495274,
      "num_tokens": 397067442.0,
      "step": 12441
    },
    {
      "entropy": 0.5496135056018829,
      "epoch": 1.1447260222770428,
      "grad_norm": 0.15448737144470215,
      "learning_rate": 6.184561597203055e-05,
      "loss": 0.5474,
      "mean_token_accuracy": 0.831415094435215,
      "num_tokens": 397099481.0,
      "step": 12442
    },
    {
      "entropy": 0.6107170972973108,
      "epoch": 1.144818028648484,
      "grad_norm": 0.15926627814769745,
      "learning_rate": 6.184254914588891e-05,
      "loss": 0.6001,
      "mean_token_accuracy": 0.8182693310081959,
      "num_tokens": 397130776.0,
      "step": 12443
    },
    {
      "entropy": 0.626948868855834,
      "epoch": 1.144910035019925,
      "grad_norm": 0.15877458453178406,
      "learning_rate": 6.18394823197473e-05,
      "loss": 0.6308,
      "mean_token_accuracy": 0.8096575401723385,
      "num_tokens": 397162968.0,
      "step": 12444
    },
    {
      "entropy": 0.6335325594991446,
      "epoch": 1.1450020413913664,
      "grad_norm": 0.15942305326461792,
      "learning_rate": 6.183641549360568e-05,
      "loss": 0.6031,
      "mean_token_accuracy": 0.8117259591817856,
      "num_tokens": 397194387.0,
      "step": 12445
    },
    {
      "entropy": 0.6468245983123779,
      "epoch": 1.1450940477628075,
      "grad_norm": 0.16397881507873535,
      "learning_rate": 6.183334866746404e-05,
      "loss": 0.6396,
      "mean_token_accuracy": 0.8031845763325691,
      "num_tokens": 397227092.0,
      "step": 12446
    },
    {
      "entropy": 0.5567029081285,
      "epoch": 1.145186054134249,
      "grad_norm": 0.1593295782804489,
      "learning_rate": 6.183028184132243e-05,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8342839442193508,
      "num_tokens": 397258949.0,
      "step": 12447
    },
    {
      "entropy": 0.719151969300583,
      "epoch": 1.14527806050569,
      "grad_norm": 0.16210778057575226,
      "learning_rate": 6.182721501518079e-05,
      "loss": 0.7047,
      "mean_token_accuracy": 0.7893036119639874,
      "num_tokens": 397290974.0,
      "step": 12448
    },
    {
      "entropy": 0.6243793591856956,
      "epoch": 1.1453700668771312,
      "grad_norm": 0.16469493508338928,
      "learning_rate": 6.182414818903916e-05,
      "loss": 0.6145,
      "mean_token_accuracy": 0.8118052929639816,
      "num_tokens": 397322501.0,
      "step": 12449
    },
    {
      "entropy": 0.7673214953392744,
      "epoch": 1.1454620732485725,
      "grad_norm": 0.17715772986412048,
      "learning_rate": 6.182108136289754e-05,
      "loss": 0.7632,
      "mean_token_accuracy": 0.7720635905861855,
      "num_tokens": 397352509.0,
      "step": 12450
    },
    {
      "entropy": 0.5671604666858912,
      "epoch": 1.1455540796200137,
      "grad_norm": 0.15873225033283234,
      "learning_rate": 6.181801453675592e-05,
      "loss": 0.5487,
      "mean_token_accuracy": 0.8291239477694035,
      "num_tokens": 397383632.0,
      "step": 12451
    },
    {
      "entropy": 0.64148950856179,
      "epoch": 1.145646085991455,
      "grad_norm": 0.16528421640396118,
      "learning_rate": 6.181494771061429e-05,
      "loss": 0.6381,
      "mean_token_accuracy": 0.80607820302248,
      "num_tokens": 397415320.0,
      "step": 12452
    },
    {
      "entropy": 0.5881329821422696,
      "epoch": 1.1457380923628961,
      "grad_norm": 0.14985701441764832,
      "learning_rate": 6.181188088447266e-05,
      "loss": 0.564,
      "mean_token_accuracy": 0.8275515250861645,
      "num_tokens": 397447712.0,
      "step": 12453
    },
    {
      "entropy": 0.6130380611866713,
      "epoch": 1.1458300987343373,
      "grad_norm": 0.15602156519889832,
      "learning_rate": 6.180881405833103e-05,
      "loss": 0.5872,
      "mean_token_accuracy": 0.8153445608913898,
      "num_tokens": 397479118.0,
      "step": 12454
    },
    {
      "entropy": 0.6398752778768539,
      "epoch": 1.1459221051057786,
      "grad_norm": 0.1588774472475052,
      "learning_rate": 6.180574723218941e-05,
      "loss": 0.645,
      "mean_token_accuracy": 0.8023729138076305,
      "num_tokens": 397511073.0,
      "step": 12455
    },
    {
      "entropy": 0.6934632528573275,
      "epoch": 1.1460141114772198,
      "grad_norm": 0.16544952988624573,
      "learning_rate": 6.180268040604779e-05,
      "loss": 0.6846,
      "mean_token_accuracy": 0.7880833148956299,
      "num_tokens": 397542600.0,
      "step": 12456
    },
    {
      "entropy": 0.6474995664320886,
      "epoch": 1.1461061178486611,
      "grad_norm": 0.15495172142982483,
      "learning_rate": 6.179961357990616e-05,
      "loss": 0.6274,
      "mean_token_accuracy": 0.8121814765036106,
      "num_tokens": 397574654.0,
      "step": 12457
    },
    {
      "entropy": 0.6714616566896439,
      "epoch": 1.1461981242201023,
      "grad_norm": 0.1672070324420929,
      "learning_rate": 6.179654675376453e-05,
      "loss": 0.6538,
      "mean_token_accuracy": 0.7967824377119541,
      "num_tokens": 397606699.0,
      "step": 12458
    },
    {
      "entropy": 0.549614348448813,
      "epoch": 1.1462901305915434,
      "grad_norm": 0.15922768414020538,
      "learning_rate": 6.179347992762291e-05,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8367918692529202,
      "num_tokens": 397639086.0,
      "step": 12459
    },
    {
      "entropy": 0.6286148531362414,
      "epoch": 1.1463821369629847,
      "grad_norm": 0.1581532210111618,
      "learning_rate": 6.179041310148128e-05,
      "loss": 0.6067,
      "mean_token_accuracy": 0.8136859908699989,
      "num_tokens": 397671019.0,
      "step": 12460
    },
    {
      "entropy": 0.6622096672654152,
      "epoch": 1.1464741433344259,
      "grad_norm": 0.1564919650554657,
      "learning_rate": 6.178734627533966e-05,
      "loss": 0.6501,
      "mean_token_accuracy": 0.8028609678149223,
      "num_tokens": 397703618.0,
      "step": 12461
    },
    {
      "entropy": 0.6285829041153193,
      "epoch": 1.1465661497058672,
      "grad_norm": 0.16266392171382904,
      "learning_rate": 6.178427944919802e-05,
      "loss": 0.6193,
      "mean_token_accuracy": 0.8093726336956024,
      "num_tokens": 397735406.0,
      "step": 12462
    },
    {
      "entropy": 0.5709601864218712,
      "epoch": 1.1466581560773084,
      "grad_norm": 0.1632539927959442,
      "learning_rate": 6.17812126230564e-05,
      "loss": 0.5578,
      "mean_token_accuracy": 0.8280309215188026,
      "num_tokens": 397768119.0,
      "step": 12463
    },
    {
      "entropy": 0.5672345133498311,
      "epoch": 1.1467501624487495,
      "grad_norm": 0.1526334285736084,
      "learning_rate": 6.177814579691477e-05,
      "loss": 0.5538,
      "mean_token_accuracy": 0.8329168409109116,
      "num_tokens": 397799848.0,
      "step": 12464
    },
    {
      "entropy": 0.7530194912105799,
      "epoch": 1.1468421688201909,
      "grad_norm": 0.17113281786441803,
      "learning_rate": 6.177507897077314e-05,
      "loss": 0.7477,
      "mean_token_accuracy": 0.7727098241448402,
      "num_tokens": 397831450.0,
      "step": 12465
    },
    {
      "entropy": 0.5958067318424582,
      "epoch": 1.146934175191632,
      "grad_norm": 0.15715409815311432,
      "learning_rate": 6.177201214463152e-05,
      "loss": 0.577,
      "mean_token_accuracy": 0.823171753436327,
      "num_tokens": 397864153.0,
      "step": 12466
    },
    {
      "entropy": 0.7724650744348764,
      "epoch": 1.1470261815630733,
      "grad_norm": 0.1815497726202011,
      "learning_rate": 6.17689453184899e-05,
      "loss": 0.7739,
      "mean_token_accuracy": 0.7725968882441521,
      "num_tokens": 397895538.0,
      "step": 12467
    },
    {
      "entropy": 0.7119228504598141,
      "epoch": 1.1471181879345145,
      "grad_norm": 0.16057145595550537,
      "learning_rate": 6.176587849234827e-05,
      "loss": 0.7052,
      "mean_token_accuracy": 0.7864797785878181,
      "num_tokens": 397928046.0,
      "step": 12468
    },
    {
      "entropy": 0.6067257300019264,
      "epoch": 1.1472101943059556,
      "grad_norm": 0.17007967829704285,
      "learning_rate": 6.176281166620664e-05,
      "loss": 0.5952,
      "mean_token_accuracy": 0.8169468231499195,
      "num_tokens": 397960290.0,
      "step": 12469
    },
    {
      "entropy": 0.6825837679207325,
      "epoch": 1.147302200677397,
      "grad_norm": 0.1590149700641632,
      "learning_rate": 6.175974484006501e-05,
      "loss": 0.6716,
      "mean_token_accuracy": 0.7958759926259518,
      "num_tokens": 397992403.0,
      "step": 12470
    },
    {
      "entropy": 0.5713729206472635,
      "epoch": 1.147394207048838,
      "grad_norm": 0.15989749133586884,
      "learning_rate": 6.17566780139234e-05,
      "loss": 0.5627,
      "mean_token_accuracy": 0.8258596546947956,
      "num_tokens": 398024538.0,
      "step": 12471
    },
    {
      "entropy": 0.5521664377301931,
      "epoch": 1.1474862134202795,
      "grad_norm": 0.15311789512634277,
      "learning_rate": 6.175361118778177e-05,
      "loss": 0.544,
      "mean_token_accuracy": 0.8299490287899971,
      "num_tokens": 398056880.0,
      "step": 12472
    },
    {
      "entropy": 0.650708282366395,
      "epoch": 1.1475782197917206,
      "grad_norm": 0.1533706784248352,
      "learning_rate": 6.175054436164014e-05,
      "loss": 0.6515,
      "mean_token_accuracy": 0.8027827478945255,
      "num_tokens": 398089456.0,
      "step": 12473
    },
    {
      "entropy": 0.5914721507579088,
      "epoch": 1.1476702261631617,
      "grad_norm": 0.15770556032657623,
      "learning_rate": 6.174747753549851e-05,
      "loss": 0.5753,
      "mean_token_accuracy": 0.8224270455539227,
      "num_tokens": 398122028.0,
      "step": 12474
    },
    {
      "entropy": 0.6822161367163062,
      "epoch": 1.147762232534603,
      "grad_norm": 0.1652955561876297,
      "learning_rate": 6.174441070935689e-05,
      "loss": 0.667,
      "mean_token_accuracy": 0.7966043576598167,
      "num_tokens": 398154141.0,
      "step": 12475
    },
    {
      "entropy": 0.6649995986372232,
      "epoch": 1.1478542389060442,
      "grad_norm": 0.1520155966281891,
      "learning_rate": 6.174134388321527e-05,
      "loss": 0.6369,
      "mean_token_accuracy": 0.806101318448782,
      "num_tokens": 398185588.0,
      "step": 12476
    },
    {
      "entropy": 0.5684383641928434,
      "epoch": 1.1479462452774856,
      "grad_norm": 0.15477295219898224,
      "learning_rate": 6.173827705707364e-05,
      "loss": 0.5574,
      "mean_token_accuracy": 0.8276895880699158,
      "num_tokens": 398217445.0,
      "step": 12477
    },
    {
      "entropy": 0.582215444650501,
      "epoch": 1.1480382516489267,
      "grad_norm": 0.1571342498064041,
      "learning_rate": 6.173521023093202e-05,
      "loss": 0.5617,
      "mean_token_accuracy": 0.8311056047677994,
      "num_tokens": 398249142.0,
      "step": 12478
    },
    {
      "entropy": 0.5368361109867692,
      "epoch": 1.1481302580203678,
      "grad_norm": 0.14717097580432892,
      "learning_rate": 6.173214340479039e-05,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8470903001725674,
      "num_tokens": 398281654.0,
      "step": 12479
    },
    {
      "entropy": 0.6618140856735408,
      "epoch": 1.1482222643918092,
      "grad_norm": 0.16528595983982086,
      "learning_rate": 6.172907657864875e-05,
      "loss": 0.6439,
      "mean_token_accuracy": 0.8004782237112522,
      "num_tokens": 398313412.0,
      "step": 12480
    },
    {
      "entropy": 0.6499750092625618,
      "epoch": 1.1483142707632503,
      "grad_norm": 0.16269880533218384,
      "learning_rate": 6.172600975250714e-05,
      "loss": 0.645,
      "mean_token_accuracy": 0.8054536692798138,
      "num_tokens": 398346073.0,
      "step": 12481
    },
    {
      "entropy": 0.7597648613154888,
      "epoch": 1.1484062771346917,
      "grad_norm": 0.16651137173175812,
      "learning_rate": 6.172294292636552e-05,
      "loss": 0.7567,
      "mean_token_accuracy": 0.7689599134027958,
      "num_tokens": 398377474.0,
      "step": 12482
    },
    {
      "entropy": 0.6769728194922209,
      "epoch": 1.1484982835061328,
      "grad_norm": 0.15804097056388855,
      "learning_rate": 6.171987610022389e-05,
      "loss": 0.6708,
      "mean_token_accuracy": 0.7977951355278492,
      "num_tokens": 398409830.0,
      "step": 12483
    },
    {
      "entropy": 0.6610789932310581,
      "epoch": 1.148590289877574,
      "grad_norm": 0.1585070937871933,
      "learning_rate": 6.171680927408225e-05,
      "loss": 0.638,
      "mean_token_accuracy": 0.8068955317139626,
      "num_tokens": 398442209.0,
      "step": 12484
    },
    {
      "entropy": 0.6604223642498255,
      "epoch": 1.1486822962490153,
      "grad_norm": 0.16365469992160797,
      "learning_rate": 6.171374244794062e-05,
      "loss": 0.6522,
      "mean_token_accuracy": 0.8032886795699596,
      "num_tokens": 398473556.0,
      "step": 12485
    },
    {
      "entropy": 0.5711758444085717,
      "epoch": 1.1487743026204564,
      "grad_norm": 0.16323789954185486,
      "learning_rate": 6.1710675621799e-05,
      "loss": 0.5641,
      "mean_token_accuracy": 0.8258028887212276,
      "num_tokens": 398505182.0,
      "step": 12486
    },
    {
      "entropy": 0.5793720250949264,
      "epoch": 1.1488663089918978,
      "grad_norm": 0.1455877125263214,
      "learning_rate": 6.170760879565738e-05,
      "loss": 0.5809,
      "mean_token_accuracy": 0.8219956010580063,
      "num_tokens": 398537950.0,
      "step": 12487
    },
    {
      "entropy": 0.6082525751553476,
      "epoch": 1.148958315363339,
      "grad_norm": 0.16114647686481476,
      "learning_rate": 6.170454196951575e-05,
      "loss": 0.6148,
      "mean_token_accuracy": 0.8132952339947224,
      "num_tokens": 398570696.0,
      "step": 12488
    },
    {
      "entropy": 0.5676320679485798,
      "epoch": 1.14905032173478,
      "grad_norm": 0.15358665585517883,
      "learning_rate": 6.170147514337412e-05,
      "loss": 0.5605,
      "mean_token_accuracy": 0.8285260796546936,
      "num_tokens": 398602688.0,
      "step": 12489
    },
    {
      "entropy": 0.4463285179808736,
      "epoch": 1.1491423281062214,
      "grad_norm": 0.14859817922115326,
      "learning_rate": 6.16984083172325e-05,
      "loss": 0.4339,
      "mean_token_accuracy": 0.8661365173757076,
      "num_tokens": 398634806.0,
      "step": 12490
    },
    {
      "entropy": 0.6790928561240435,
      "epoch": 1.1492343344776625,
      "grad_norm": 0.16621901094913483,
      "learning_rate": 6.169534149109087e-05,
      "loss": 0.6862,
      "mean_token_accuracy": 0.7919354885816574,
      "num_tokens": 398666020.0,
      "step": 12491
    },
    {
      "entropy": 0.6160495243966579,
      "epoch": 1.149326340849104,
      "grad_norm": 0.1578884720802307,
      "learning_rate": 6.169227466494925e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.8129483312368393,
      "num_tokens": 398698115.0,
      "step": 12492
    },
    {
      "entropy": 0.5682510752230883,
      "epoch": 1.149418347220545,
      "grad_norm": 0.15616066753864288,
      "learning_rate": 6.168920783880762e-05,
      "loss": 0.5587,
      "mean_token_accuracy": 0.8240585811436176,
      "num_tokens": 398730520.0,
      "step": 12493
    },
    {
      "entropy": 0.647894062101841,
      "epoch": 1.1495103535919862,
      "grad_norm": 0.15358972549438477,
      "learning_rate": 6.1686141012666e-05,
      "loss": 0.635,
      "mean_token_accuracy": 0.8075027950108051,
      "num_tokens": 398762717.0,
      "step": 12494
    },
    {
      "entropy": 0.6773941516876221,
      "epoch": 1.1496023599634275,
      "grad_norm": 0.1613815575838089,
      "learning_rate": 6.168307418652437e-05,
      "loss": 0.6687,
      "mean_token_accuracy": 0.797627966850996,
      "num_tokens": 398794920.0,
      "step": 12495
    },
    {
      "entropy": 0.640712053515017,
      "epoch": 1.1496943663348687,
      "grad_norm": 0.15429727733135223,
      "learning_rate": 6.168000736038274e-05,
      "loss": 0.6252,
      "mean_token_accuracy": 0.8098922222852707,
      "num_tokens": 398826624.0,
      "step": 12496
    },
    {
      "entropy": 0.5796362049877644,
      "epoch": 1.14978637270631,
      "grad_norm": 0.15390892326831818,
      "learning_rate": 6.167694053424112e-05,
      "loss": 0.566,
      "mean_token_accuracy": 0.824682779610157,
      "num_tokens": 398858537.0,
      "step": 12497
    },
    {
      "entropy": 0.5819000769406557,
      "epoch": 1.1498783790777511,
      "grad_norm": 0.1557639241218567,
      "learning_rate": 6.16738737080995e-05,
      "loss": 0.5593,
      "mean_token_accuracy": 0.8310986831784248,
      "num_tokens": 398891153.0,
      "step": 12498
    },
    {
      "entropy": 0.6530608218163252,
      "epoch": 1.1499703854491923,
      "grad_norm": 0.1570030003786087,
      "learning_rate": 6.167080688195787e-05,
      "loss": 0.6251,
      "mean_token_accuracy": 0.8092974834144115,
      "num_tokens": 398923665.0,
      "step": 12499
    },
    {
      "entropy": 0.5882731191813946,
      "epoch": 1.1500623918206336,
      "grad_norm": 0.17019732296466827,
      "learning_rate": 6.166774005581623e-05,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8248474784195423,
      "num_tokens": 398955693.0,
      "step": 12500
    },
    {
      "entropy": 0.60167570784688,
      "epoch": 1.1501543981920748,
      "grad_norm": 0.1530766636133194,
      "learning_rate": 6.16646732296746e-05,
      "loss": 0.5843,
      "mean_token_accuracy": 0.8181661292910576,
      "num_tokens": 398987704.0,
      "step": 12501
    },
    {
      "entropy": 0.584281818009913,
      "epoch": 1.1502464045635161,
      "grad_norm": 0.15858326852321625,
      "learning_rate": 6.166160640353298e-05,
      "loss": 0.5743,
      "mean_token_accuracy": 0.819682240486145,
      "num_tokens": 399020102.0,
      "step": 12502
    },
    {
      "entropy": 0.6961255669593811,
      "epoch": 1.1503384109349573,
      "grad_norm": 0.16187046468257904,
      "learning_rate": 6.165853957739136e-05,
      "loss": 0.696,
      "mean_token_accuracy": 0.7883487120270729,
      "num_tokens": 399052085.0,
      "step": 12503
    },
    {
      "entropy": 0.6256786119192839,
      "epoch": 1.1504304173063984,
      "grad_norm": 0.1624433696269989,
      "learning_rate": 6.165547275124973e-05,
      "loss": 0.6198,
      "mean_token_accuracy": 0.8117522895336151,
      "num_tokens": 399083969.0,
      "step": 12504
    },
    {
      "entropy": 0.7229004427790642,
      "epoch": 1.1505224236778397,
      "grad_norm": 0.16991500556468964,
      "learning_rate": 6.16524059251081e-05,
      "loss": 0.722,
      "mean_token_accuracy": 0.7804803922772408,
      "num_tokens": 399116262.0,
      "step": 12505
    },
    {
      "entropy": 0.5329565550200641,
      "epoch": 1.1506144300492809,
      "grad_norm": 0.16053129732608795,
      "learning_rate": 6.164933909896648e-05,
      "loss": 0.5204,
      "mean_token_accuracy": 0.840036004781723,
      "num_tokens": 399148414.0,
      "step": 12506
    },
    {
      "entropy": 0.5931178545579314,
      "epoch": 1.1507064364207222,
      "grad_norm": 0.1590176224708557,
      "learning_rate": 6.164627227282485e-05,
      "loss": 0.579,
      "mean_token_accuracy": 0.8202678374946117,
      "num_tokens": 399180674.0,
      "step": 12507
    },
    {
      "entropy": 0.6207096977159381,
      "epoch": 1.1507984427921634,
      "grad_norm": 0.15645700693130493,
      "learning_rate": 6.164320544668323e-05,
      "loss": 0.6124,
      "mean_token_accuracy": 0.8154609836637974,
      "num_tokens": 399213001.0,
      "step": 12508
    },
    {
      "entropy": 0.6490658614784479,
      "epoch": 1.1508904491636045,
      "grad_norm": 0.16453945636749268,
      "learning_rate": 6.164013862054161e-05,
      "loss": 0.6296,
      "mean_token_accuracy": 0.8059286363422871,
      "num_tokens": 399243912.0,
      "step": 12509
    },
    {
      "entropy": 0.6588023798540235,
      "epoch": 1.1509824555350459,
      "grad_norm": 0.16029897332191467,
      "learning_rate": 6.163707179439998e-05,
      "loss": 0.6563,
      "mean_token_accuracy": 0.7998053357005119,
      "num_tokens": 399276268.0,
      "step": 12510
    },
    {
      "entropy": 0.5743353818543255,
      "epoch": 1.151074461906487,
      "grad_norm": 0.158537819981575,
      "learning_rate": 6.163400496825835e-05,
      "loss": 0.5661,
      "mean_token_accuracy": 0.8280709683895111,
      "num_tokens": 399308419.0,
      "step": 12511
    },
    {
      "entropy": 0.6334912721067667,
      "epoch": 1.1511664682779283,
      "grad_norm": 0.16213703155517578,
      "learning_rate": 6.163093814211673e-05,
      "loss": 0.6256,
      "mean_token_accuracy": 0.8092435412108898,
      "num_tokens": 399340745.0,
      "step": 12512
    },
    {
      "entropy": 0.7522643022239208,
      "epoch": 1.1512584746493695,
      "grad_norm": 0.16518191993236542,
      "learning_rate": 6.162787131597511e-05,
      "loss": 0.7449,
      "mean_token_accuracy": 0.7792953550815582,
      "num_tokens": 399373217.0,
      "step": 12513
    },
    {
      "entropy": 0.6375181330367923,
      "epoch": 1.1513504810208106,
      "grad_norm": 0.1620526760816574,
      "learning_rate": 6.162480448983348e-05,
      "loss": 0.611,
      "mean_token_accuracy": 0.8144148774445057,
      "num_tokens": 399405631.0,
      "step": 12514
    },
    {
      "entropy": 0.6473494535312057,
      "epoch": 1.151442487392252,
      "grad_norm": 0.16530899703502655,
      "learning_rate": 6.162173766369185e-05,
      "loss": 0.644,
      "mean_token_accuracy": 0.7997112348675728,
      "num_tokens": 399437823.0,
      "step": 12515
    },
    {
      "entropy": 0.5499817440286279,
      "epoch": 1.151534493763693,
      "grad_norm": 0.15687783062458038,
      "learning_rate": 6.161867083755021e-05,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8332377672195435,
      "num_tokens": 399469243.0,
      "step": 12516
    },
    {
      "entropy": 0.6338571077212691,
      "epoch": 1.1516265001351345,
      "grad_norm": 0.15736311674118042,
      "learning_rate": 6.16156040114086e-05,
      "loss": 0.63,
      "mean_token_accuracy": 0.806858267635107,
      "num_tokens": 399501193.0,
      "step": 12517
    },
    {
      "entropy": 0.667673422023654,
      "epoch": 1.1517185065065756,
      "grad_norm": 0.1574450135231018,
      "learning_rate": 6.161253718526698e-05,
      "loss": 0.6458,
      "mean_token_accuracy": 0.7998207993805408,
      "num_tokens": 399533242.0,
      "step": 12518
    },
    {
      "entropy": 0.5393115812912583,
      "epoch": 1.1518105128780167,
      "grad_norm": 0.1480705589056015,
      "learning_rate": 6.160947035912535e-05,
      "loss": 0.516,
      "mean_token_accuracy": 0.8414127640426159,
      "num_tokens": 399565048.0,
      "step": 12519
    },
    {
      "entropy": 0.6692915260791779,
      "epoch": 1.151902519249458,
      "grad_norm": 0.16342847049236298,
      "learning_rate": 6.160640353298371e-05,
      "loss": 0.6648,
      "mean_token_accuracy": 0.8019160255789757,
      "num_tokens": 399597162.0,
      "step": 12520
    },
    {
      "entropy": 0.5546263311989605,
      "epoch": 1.1519945256208992,
      "grad_norm": 0.1662762612104416,
      "learning_rate": 6.16033367068421e-05,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8373949751257896,
      "num_tokens": 399629090.0,
      "step": 12521
    },
    {
      "entropy": 0.6724150124937296,
      "epoch": 1.1520865319923406,
      "grad_norm": 0.160316601395607,
      "learning_rate": 6.160026988070046e-05,
      "loss": 0.6645,
      "mean_token_accuracy": 0.8011331968009472,
      "num_tokens": 399660844.0,
      "step": 12522
    },
    {
      "entropy": 0.5720565123483539,
      "epoch": 1.1521785383637817,
      "grad_norm": 0.15511725842952728,
      "learning_rate": 6.159720305455884e-05,
      "loss": 0.5516,
      "mean_token_accuracy": 0.829716645181179,
      "num_tokens": 399692722.0,
      "step": 12523
    },
    {
      "entropy": 0.7351727737113833,
      "epoch": 1.1522705447352228,
      "grad_norm": 0.16435250639915466,
      "learning_rate": 6.159413622841721e-05,
      "loss": 0.7435,
      "mean_token_accuracy": 0.7797145992517471,
      "num_tokens": 399724481.0,
      "step": 12524
    },
    {
      "entropy": 0.5913975834846497,
      "epoch": 1.1523625511066642,
      "grad_norm": 0.15161928534507751,
      "learning_rate": 6.15910694022756e-05,
      "loss": 0.5678,
      "mean_token_accuracy": 0.8246057666838169,
      "num_tokens": 399756807.0,
      "step": 12525
    },
    {
      "entropy": 0.7019187090918422,
      "epoch": 1.1524545574781053,
      "grad_norm": 0.16494333744049072,
      "learning_rate": 6.158800257613396e-05,
      "loss": 0.6965,
      "mean_token_accuracy": 0.7897191643714905,
      "num_tokens": 399788506.0,
      "step": 12526
    },
    {
      "entropy": 0.5857827886939049,
      "epoch": 1.1525465638495467,
      "grad_norm": 0.1537531167268753,
      "learning_rate": 6.158493574999233e-05,
      "loss": 0.5833,
      "mean_token_accuracy": 0.8228163085877895,
      "num_tokens": 399821216.0,
      "step": 12527
    },
    {
      "entropy": 0.5228724647313356,
      "epoch": 1.1526385702209878,
      "grad_norm": 0.1561654657125473,
      "learning_rate": 6.158186892385071e-05,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8409807719290257,
      "num_tokens": 399852988.0,
      "step": 12528
    },
    {
      "entropy": 0.6890172716230154,
      "epoch": 1.152730576592429,
      "grad_norm": 0.16496893763542175,
      "learning_rate": 6.157880209770909e-05,
      "loss": 0.685,
      "mean_token_accuracy": 0.7920877523720264,
      "num_tokens": 399883998.0,
      "step": 12529
    },
    {
      "entropy": 0.7213133657351136,
      "epoch": 1.1528225829638703,
      "grad_norm": 0.16401194036006927,
      "learning_rate": 6.157573527156746e-05,
      "loss": 0.7345,
      "mean_token_accuracy": 0.7776467055082321,
      "num_tokens": 399915931.0,
      "step": 12530
    },
    {
      "entropy": 0.7536927722394466,
      "epoch": 1.1529145893353114,
      "grad_norm": 0.1661224514245987,
      "learning_rate": 6.157266844542583e-05,
      "loss": 0.7468,
      "mean_token_accuracy": 0.7718115001916885,
      "num_tokens": 399947589.0,
      "step": 12531
    },
    {
      "entropy": 0.6651991233229637,
      "epoch": 1.1530065957067528,
      "grad_norm": 0.15460124611854553,
      "learning_rate": 6.15696016192842e-05,
      "loss": 0.6557,
      "mean_token_accuracy": 0.8005007319152355,
      "num_tokens": 399980357.0,
      "step": 12532
    },
    {
      "entropy": 0.6461223671212792,
      "epoch": 1.153098602078194,
      "grad_norm": 0.16007381677627563,
      "learning_rate": 6.156653479314258e-05,
      "loss": 0.6205,
      "mean_token_accuracy": 0.8111306577920914,
      "num_tokens": 400011699.0,
      "step": 12533
    },
    {
      "entropy": 0.597927650436759,
      "epoch": 1.153190608449635,
      "grad_norm": 0.16182544827461243,
      "learning_rate": 6.156346796700096e-05,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8288383185863495,
      "num_tokens": 400043360.0,
      "step": 12534
    },
    {
      "entropy": 0.7442982625216246,
      "epoch": 1.1532826148210764,
      "grad_norm": 0.16018398106098175,
      "learning_rate": 6.156040114085933e-05,
      "loss": 0.7328,
      "mean_token_accuracy": 0.7771634161472321,
      "num_tokens": 400075292.0,
      "step": 12535
    },
    {
      "entropy": 0.6614075060933828,
      "epoch": 1.1533746211925175,
      "grad_norm": 0.9762125015258789,
      "learning_rate": 6.15573343147177e-05,
      "loss": 0.7105,
      "mean_token_accuracy": 0.8012426644563675,
      "num_tokens": 400107453.0,
      "step": 12536
    },
    {
      "entropy": 0.5775686390697956,
      "epoch": 1.153466627563959,
      "grad_norm": 0.1544279009103775,
      "learning_rate": 6.155426748857608e-05,
      "loss": 0.557,
      "mean_token_accuracy": 0.8284998014569283,
      "num_tokens": 400139846.0,
      "step": 12537
    },
    {
      "entropy": 0.6102480255067348,
      "epoch": 1.1535586339354,
      "grad_norm": 0.17959213256835938,
      "learning_rate": 6.155120066243444e-05,
      "loss": 0.5893,
      "mean_token_accuracy": 0.8196839578449726,
      "num_tokens": 400172246.0,
      "step": 12538
    },
    {
      "entropy": 0.6789997331798077,
      "epoch": 1.1536506403068412,
      "grad_norm": 0.17623034119606018,
      "learning_rate": 6.154813383629283e-05,
      "loss": 0.6738,
      "mean_token_accuracy": 0.8028445951640606,
      "num_tokens": 400203144.0,
      "step": 12539
    },
    {
      "entropy": 0.6175279673188925,
      "epoch": 1.1537426466782825,
      "grad_norm": 0.31593430042266846,
      "learning_rate": 6.15450670101512e-05,
      "loss": 0.6102,
      "mean_token_accuracy": 0.8166398108005524,
      "num_tokens": 400235345.0,
      "step": 12540
    },
    {
      "entropy": 0.7198263350874186,
      "epoch": 1.1538346530497237,
      "grad_norm": 0.16532424092292786,
      "learning_rate": 6.154200018400957e-05,
      "loss": 0.7108,
      "mean_token_accuracy": 0.7864139229059219,
      "num_tokens": 400267742.0,
      "step": 12541
    },
    {
      "entropy": 0.6043607504107058,
      "epoch": 1.153926659421165,
      "grad_norm": 0.1636044681072235,
      "learning_rate": 6.153893335786794e-05,
      "loss": 0.5897,
      "mean_token_accuracy": 0.8219213038682938,
      "num_tokens": 400299081.0,
      "step": 12542
    },
    {
      "entropy": 0.5588542148470879,
      "epoch": 1.1540186657926061,
      "grad_norm": 0.16600896418094635,
      "learning_rate": 6.153586653172631e-05,
      "loss": 0.5634,
      "mean_token_accuracy": 0.8266470804810524,
      "num_tokens": 400330547.0,
      "step": 12543
    },
    {
      "entropy": 0.6366480784490705,
      "epoch": 1.1541106721640473,
      "grad_norm": 0.16014273464679718,
      "learning_rate": 6.15327997055847e-05,
      "loss": 0.6238,
      "mean_token_accuracy": 0.8074254579842091,
      "num_tokens": 400362880.0,
      "step": 12544
    },
    {
      "entropy": 0.4666228275746107,
      "epoch": 1.1542026785354886,
      "grad_norm": 0.1453714817762375,
      "learning_rate": 6.152973287944307e-05,
      "loss": 0.4579,
      "mean_token_accuracy": 0.8573523610830307,
      "num_tokens": 400394902.0,
      "step": 12545
    },
    {
      "entropy": 0.5390175860375166,
      "epoch": 1.1542946849069298,
      "grad_norm": 0.14768849313259125,
      "learning_rate": 6.152666605330144e-05,
      "loss": 0.5184,
      "mean_token_accuracy": 0.8354245834052563,
      "num_tokens": 400426737.0,
      "step": 12546
    },
    {
      "entropy": 0.6629142444580793,
      "epoch": 1.1543866912783711,
      "grad_norm": 0.16559794545173645,
      "learning_rate": 6.152359922715981e-05,
      "loss": 0.6521,
      "mean_token_accuracy": 0.7994280233979225,
      "num_tokens": 400458828.0,
      "step": 12547
    },
    {
      "entropy": 0.6264387415722013,
      "epoch": 1.1544786976498123,
      "grad_norm": 0.15769243240356445,
      "learning_rate": 6.152053240101819e-05,
      "loss": 0.6211,
      "mean_token_accuracy": 0.8078948110342026,
      "num_tokens": 400490391.0,
      "step": 12548
    },
    {
      "entropy": 0.5609429571777582,
      "epoch": 1.1545707040212534,
      "grad_norm": 0.15816394984722137,
      "learning_rate": 6.151746557487657e-05,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8288928158581257,
      "num_tokens": 400522218.0,
      "step": 12549
    },
    {
      "entropy": 0.6014771512709558,
      "epoch": 1.1546627103926947,
      "grad_norm": 0.15566343069076538,
      "learning_rate": 6.151439874873494e-05,
      "loss": 0.5937,
      "mean_token_accuracy": 0.8201383166015148,
      "num_tokens": 400554225.0,
      "step": 12550
    },
    {
      "entropy": 0.6809070874005556,
      "epoch": 1.1547547167641359,
      "grad_norm": 0.15970923006534576,
      "learning_rate": 6.151133192259331e-05,
      "loss": 0.6565,
      "mean_token_accuracy": 0.8001220971345901,
      "num_tokens": 400586460.0,
      "step": 12551
    },
    {
      "entropy": 0.584039100445807,
      "epoch": 1.1548467231355772,
      "grad_norm": 0.1590511053800583,
      "learning_rate": 6.150826509645169e-05,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8237068094313145,
      "num_tokens": 400617908.0,
      "step": 12552
    },
    {
      "entropy": 0.6425019688904285,
      "epoch": 1.1549387295070184,
      "grad_norm": 0.15902400016784668,
      "learning_rate": 6.150519827031006e-05,
      "loss": 0.6328,
      "mean_token_accuracy": 0.8060521185398102,
      "num_tokens": 400649504.0,
      "step": 12553
    },
    {
      "entropy": 0.634262578561902,
      "epoch": 1.1550307358784595,
      "grad_norm": 0.16272272169589996,
      "learning_rate": 6.150213144416844e-05,
      "loss": 0.6281,
      "mean_token_accuracy": 0.8065553940832615,
      "num_tokens": 400681948.0,
      "step": 12554
    },
    {
      "entropy": 0.7281483337283134,
      "epoch": 1.1551227422499009,
      "grad_norm": 0.16158601641654968,
      "learning_rate": 6.14990646180268e-05,
      "loss": 0.729,
      "mean_token_accuracy": 0.7813884653151035,
      "num_tokens": 400713567.0,
      "step": 12555
    },
    {
      "entropy": 0.6343301106244326,
      "epoch": 1.155214748621342,
      "grad_norm": 0.15761631727218628,
      "learning_rate": 6.149599779188519e-05,
      "loss": 0.6247,
      "mean_token_accuracy": 0.8086086474359035,
      "num_tokens": 400745389.0,
      "step": 12556
    },
    {
      "entropy": 0.5284028798341751,
      "epoch": 1.1553067549927833,
      "grad_norm": 0.15922819077968597,
      "learning_rate": 6.149293096574356e-05,
      "loss": 0.5125,
      "mean_token_accuracy": 0.8446672074496746,
      "num_tokens": 400776640.0,
      "step": 12557
    },
    {
      "entropy": 0.5586043866351247,
      "epoch": 1.1553987613642245,
      "grad_norm": 0.19897381961345673,
      "learning_rate": 6.148986413960192e-05,
      "loss": 0.5488,
      "mean_token_accuracy": 0.8349239490926266,
      "num_tokens": 400808498.0,
      "step": 12558
    },
    {
      "entropy": 0.5805538278073072,
      "epoch": 1.1554907677356656,
      "grad_norm": 0.15442386269569397,
      "learning_rate": 6.14867973134603e-05,
      "loss": 0.5736,
      "mean_token_accuracy": 0.8229454793035984,
      "num_tokens": 400841176.0,
      "step": 12559
    },
    {
      "entropy": 0.5896794870495796,
      "epoch": 1.155582774107107,
      "grad_norm": 0.15235577523708344,
      "learning_rate": 6.148373048731869e-05,
      "loss": 0.5811,
      "mean_token_accuracy": 0.8182911761105061,
      "num_tokens": 400873231.0,
      "step": 12560
    },
    {
      "entropy": 0.6123478738591075,
      "epoch": 1.155674780478548,
      "grad_norm": 0.16093820333480835,
      "learning_rate": 6.148066366117705e-05,
      "loss": 0.6042,
      "mean_token_accuracy": 0.8170471973717213,
      "num_tokens": 400905655.0,
      "step": 12561
    },
    {
      "entropy": 0.5621436964720488,
      "epoch": 1.1557667868499895,
      "grad_norm": 0.16095098853111267,
      "learning_rate": 6.147759683503542e-05,
      "loss": 0.5575,
      "mean_token_accuracy": 0.8271565288305283,
      "num_tokens": 400937423.0,
      "step": 12562
    },
    {
      "entropy": 0.643779506906867,
      "epoch": 1.1558587932214306,
      "grad_norm": 0.15668343007564545,
      "learning_rate": 6.147453000889379e-05,
      "loss": 0.6418,
      "mean_token_accuracy": 0.802987165749073,
      "num_tokens": 400969520.0,
      "step": 12563
    },
    {
      "entropy": 0.571981268003583,
      "epoch": 1.1559507995928717,
      "grad_norm": 0.15695954859256744,
      "learning_rate": 6.147146318275217e-05,
      "loss": 0.5555,
      "mean_token_accuracy": 0.8307585045695305,
      "num_tokens": 401001779.0,
      "step": 12564
    },
    {
      "entropy": 0.6169859888032079,
      "epoch": 1.156042805964313,
      "grad_norm": 0.15969878435134888,
      "learning_rate": 6.146839635661055e-05,
      "loss": 0.6041,
      "mean_token_accuracy": 0.8126465529203415,
      "num_tokens": 401033237.0,
      "step": 12565
    },
    {
      "entropy": 0.7689934968948364,
      "epoch": 1.1561348123357542,
      "grad_norm": 0.1705353707075119,
      "learning_rate": 6.146532953046892e-05,
      "loss": 0.7616,
      "mean_token_accuracy": 0.7709814384579659,
      "num_tokens": 401065059.0,
      "step": 12566
    },
    {
      "entropy": 0.6564987916499376,
      "epoch": 1.1562268187071956,
      "grad_norm": 0.16360372304916382,
      "learning_rate": 6.146226270432729e-05,
      "loss": 0.6343,
      "mean_token_accuracy": 0.8023303672671318,
      "num_tokens": 401096387.0,
      "step": 12567
    },
    {
      "entropy": 0.6653701290488243,
      "epoch": 1.1563188250786367,
      "grad_norm": 0.15895193815231323,
      "learning_rate": 6.145919587818567e-05,
      "loss": 0.6436,
      "mean_token_accuracy": 0.8040464967489243,
      "num_tokens": 401127996.0,
      "step": 12568
    },
    {
      "entropy": 0.5966701973229647,
      "epoch": 1.1564108314500778,
      "grad_norm": 0.15566350519657135,
      "learning_rate": 6.145612905204404e-05,
      "loss": 0.5933,
      "mean_token_accuracy": 0.8207437768578529,
      "num_tokens": 401160764.0,
      "step": 12569
    },
    {
      "entropy": 0.6979442723095417,
      "epoch": 1.1565028378215192,
      "grad_norm": 0.16105620563030243,
      "learning_rate": 6.145306222590242e-05,
      "loss": 0.6813,
      "mean_token_accuracy": 0.7937680669128895,
      "num_tokens": 401191793.0,
      "step": 12570
    },
    {
      "entropy": 0.678867019712925,
      "epoch": 1.1565948441929603,
      "grad_norm": 0.15583637356758118,
      "learning_rate": 6.14499953997608e-05,
      "loss": 0.6607,
      "mean_token_accuracy": 0.7966363579034805,
      "num_tokens": 401223946.0,
      "step": 12571
    },
    {
      "entropy": 0.6271709781140089,
      "epoch": 1.1566868505644017,
      "grad_norm": 0.157325878739357,
      "learning_rate": 6.144692857361917e-05,
      "loss": 0.618,
      "mean_token_accuracy": 0.8074054382741451,
      "num_tokens": 401255857.0,
      "step": 12572
    },
    {
      "entropy": 0.688151016831398,
      "epoch": 1.1567788569358428,
      "grad_norm": 0.16152985394001007,
      "learning_rate": 6.144386174747754e-05,
      "loss": 0.676,
      "mean_token_accuracy": 0.7991058938205242,
      "num_tokens": 401287468.0,
      "step": 12573
    },
    {
      "entropy": 0.6466993670910597,
      "epoch": 1.156870863307284,
      "grad_norm": 0.15849439799785614,
      "learning_rate": 6.14407949213359e-05,
      "loss": 0.6193,
      "mean_token_accuracy": 0.8097205758094788,
      "num_tokens": 401319442.0,
      "step": 12574
    },
    {
      "entropy": 0.5906429355964065,
      "epoch": 1.1569628696787253,
      "grad_norm": 0.16175219416618347,
      "learning_rate": 6.143772809519429e-05,
      "loss": 0.5696,
      "mean_token_accuracy": 0.823595192283392,
      "num_tokens": 401350918.0,
      "step": 12575
    },
    {
      "entropy": 0.6266635814681649,
      "epoch": 1.1570548760501664,
      "grad_norm": 0.15555544197559357,
      "learning_rate": 6.143466126905267e-05,
      "loss": 0.627,
      "mean_token_accuracy": 0.8125366009771824,
      "num_tokens": 401383571.0,
      "step": 12576
    },
    {
      "entropy": 0.7180796694010496,
      "epoch": 1.1571468824216078,
      "grad_norm": 0.1607608199119568,
      "learning_rate": 6.143159444291103e-05,
      "loss": 0.7156,
      "mean_token_accuracy": 0.7855488173663616,
      "num_tokens": 401415140.0,
      "step": 12577
    },
    {
      "entropy": 0.6287715081125498,
      "epoch": 1.157238888793049,
      "grad_norm": 0.16212141513824463,
      "learning_rate": 6.14285276167694e-05,
      "loss": 0.6176,
      "mean_token_accuracy": 0.8103734217584133,
      "num_tokens": 401446975.0,
      "step": 12578
    },
    {
      "entropy": 0.6359396188054234,
      "epoch": 1.15733089516449,
      "grad_norm": 0.15376703441143036,
      "learning_rate": 6.142546079062778e-05,
      "loss": 0.6243,
      "mean_token_accuracy": 0.8074772469699383,
      "num_tokens": 401478906.0,
      "step": 12579
    },
    {
      "entropy": 0.6734068319201469,
      "epoch": 1.1574229015359314,
      "grad_norm": 0.16696463525295258,
      "learning_rate": 6.142239396448615e-05,
      "loss": 0.6611,
      "mean_token_accuracy": 0.796523205935955,
      "num_tokens": 401510913.0,
      "step": 12580
    },
    {
      "entropy": 0.6459511108696461,
      "epoch": 1.1575149079073725,
      "grad_norm": 0.15991081297397614,
      "learning_rate": 6.141932713834453e-05,
      "loss": 0.6387,
      "mean_token_accuracy": 0.80333561450243,
      "num_tokens": 401543136.0,
      "step": 12581
    },
    {
      "entropy": 0.7846213448792696,
      "epoch": 1.157606914278814,
      "grad_norm": 0.17464055120944977,
      "learning_rate": 6.14162603122029e-05,
      "loss": 0.7719,
      "mean_token_accuracy": 0.7684539444744587,
      "num_tokens": 401574764.0,
      "step": 12582
    },
    {
      "entropy": 0.6234679650515318,
      "epoch": 1.157698920650255,
      "grad_norm": 0.15831829607486725,
      "learning_rate": 6.141319348606128e-05,
      "loss": 0.6095,
      "mean_token_accuracy": 0.8102762959897518,
      "num_tokens": 401606517.0,
      "step": 12583
    },
    {
      "entropy": 0.5717992410063744,
      "epoch": 1.1577909270216962,
      "grad_norm": 0.20177678763866425,
      "learning_rate": 6.141012665991965e-05,
      "loss": 0.5756,
      "mean_token_accuracy": 0.8219264410436153,
      "num_tokens": 401638926.0,
      "step": 12584
    },
    {
      "entropy": 0.5372052034363151,
      "epoch": 1.1578829333931375,
      "grad_norm": 0.15601781010627747,
      "learning_rate": 6.140705983377802e-05,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8421565406024456,
      "num_tokens": 401671367.0,
      "step": 12585
    },
    {
      "entropy": 0.6953169452026486,
      "epoch": 1.1579749397645787,
      "grad_norm": 0.1638512909412384,
      "learning_rate": 6.14039930076364e-05,
      "loss": 0.673,
      "mean_token_accuracy": 0.7946280762553215,
      "num_tokens": 401703707.0,
      "step": 12586
    },
    {
      "entropy": 0.554560977499932,
      "epoch": 1.15806694613602,
      "grad_norm": 0.16423475742340088,
      "learning_rate": 6.140092618149478e-05,
      "loss": 0.5551,
      "mean_token_accuracy": 0.8282148912549019,
      "num_tokens": 401735470.0,
      "step": 12587
    },
    {
      "entropy": 0.6296832822263241,
      "epoch": 1.1581589525074611,
      "grad_norm": 0.1663064807653427,
      "learning_rate": 6.139785935535315e-05,
      "loss": 0.6105,
      "mean_token_accuracy": 0.8114977069199085,
      "num_tokens": 401767359.0,
      "step": 12588
    },
    {
      "entropy": 0.5633092131465673,
      "epoch": 1.1582509588789023,
      "grad_norm": 0.1477556824684143,
      "learning_rate": 6.139479252921152e-05,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8293136432766914,
      "num_tokens": 401799736.0,
      "step": 12589
    },
    {
      "entropy": 0.6159085724502802,
      "epoch": 1.1583429652503436,
      "grad_norm": 0.1688009649515152,
      "learning_rate": 6.139172570306988e-05,
      "loss": 0.6017,
      "mean_token_accuracy": 0.8152236565947533,
      "num_tokens": 401831881.0,
      "step": 12590
    },
    {
      "entropy": 0.6485949447378516,
      "epoch": 1.1584349716217848,
      "grad_norm": 0.16668526828289032,
      "learning_rate": 6.138865887692828e-05,
      "loss": 0.6428,
      "mean_token_accuracy": 0.8064571470022202,
      "num_tokens": 401862938.0,
      "step": 12591
    },
    {
      "entropy": 0.6474290629848838,
      "epoch": 1.1585269779932261,
      "grad_norm": 0.1665302813053131,
      "learning_rate": 6.138559205078665e-05,
      "loss": 0.6374,
      "mean_token_accuracy": 0.8042043820023537,
      "num_tokens": 401894977.0,
      "step": 12592
    },
    {
      "entropy": 0.5788237918168306,
      "epoch": 1.1586189843646673,
      "grad_norm": 0.16325215995311737,
      "learning_rate": 6.138252522464502e-05,
      "loss": 0.5558,
      "mean_token_accuracy": 0.8246861472725868,
      "num_tokens": 401925248.0,
      "step": 12593
    },
    {
      "entropy": 0.6510245949029922,
      "epoch": 1.1587109907361084,
      "grad_norm": 0.1652635931968689,
      "learning_rate": 6.137945839850338e-05,
      "loss": 0.6296,
      "mean_token_accuracy": 0.808032289147377,
      "num_tokens": 401957180.0,
      "step": 12594
    },
    {
      "entropy": 0.5245853550732136,
      "epoch": 1.1588029971075497,
      "grad_norm": 0.14271168410778046,
      "learning_rate": 6.137639157236176e-05,
      "loss": 0.5125,
      "mean_token_accuracy": 0.8423182666301727,
      "num_tokens": 401989493.0,
      "step": 12595
    },
    {
      "entropy": 0.6381749417632818,
      "epoch": 1.1588950034789909,
      "grad_norm": 0.16046643257141113,
      "learning_rate": 6.137332474622015e-05,
      "loss": 0.6306,
      "mean_token_accuracy": 0.8073310814797878,
      "num_tokens": 402021034.0,
      "step": 12596
    },
    {
      "entropy": 0.6103674778714776,
      "epoch": 1.1589870098504322,
      "grad_norm": 0.15521511435508728,
      "learning_rate": 6.137025792007851e-05,
      "loss": 0.5986,
      "mean_token_accuracy": 0.8160270936787128,
      "num_tokens": 402052524.0,
      "step": 12597
    },
    {
      "entropy": 0.6625061351805925,
      "epoch": 1.1590790162218734,
      "grad_norm": 0.1674436628818512,
      "learning_rate": 6.136719109393688e-05,
      "loss": 0.6588,
      "mean_token_accuracy": 0.8000541962683201,
      "num_tokens": 402084195.0,
      "step": 12598
    },
    {
      "entropy": 0.6060051023960114,
      "epoch": 1.1591710225933145,
      "grad_norm": 0.1557805836200714,
      "learning_rate": 6.136412426779526e-05,
      "loss": 0.5951,
      "mean_token_accuracy": 0.8197619356215,
      "num_tokens": 402116535.0,
      "step": 12599
    },
    {
      "entropy": 0.5846638893708587,
      "epoch": 1.1592630289647559,
      "grad_norm": 0.1565539836883545,
      "learning_rate": 6.136105744165363e-05,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8268745020031929,
      "num_tokens": 402148115.0,
      "step": 12600
    },
    {
      "entropy": 0.6131699224933982,
      "epoch": 1.159355035336197,
      "grad_norm": 0.1559438407421112,
      "learning_rate": 6.135799061551201e-05,
      "loss": 0.6093,
      "mean_token_accuracy": 0.8155316822230816,
      "num_tokens": 402179324.0,
      "step": 12601
    },
    {
      "entropy": 0.6681925021111965,
      "epoch": 1.1594470417076383,
      "grad_norm": 0.15712720155715942,
      "learning_rate": 6.13549237893704e-05,
      "loss": 0.6565,
      "mean_token_accuracy": 0.8017873652279377,
      "num_tokens": 402211515.0,
      "step": 12602
    },
    {
      "entropy": 0.6545578520745039,
      "epoch": 1.1595390480790795,
      "grad_norm": 0.15620771050453186,
      "learning_rate": 6.135185696322876e-05,
      "loss": 0.6375,
      "mean_token_accuracy": 0.8064906410872936,
      "num_tokens": 402243336.0,
      "step": 12603
    },
    {
      "entropy": 0.5536239501088858,
      "epoch": 1.1596310544505206,
      "grad_norm": 0.17193558812141418,
      "learning_rate": 6.134879013708713e-05,
      "loss": 0.5493,
      "mean_token_accuracy": 0.8306376077234745,
      "num_tokens": 402275702.0,
      "step": 12604
    },
    {
      "entropy": 0.6858114115893841,
      "epoch": 1.159723060821962,
      "grad_norm": 0.16404224932193756,
      "learning_rate": 6.13457233109455e-05,
      "loss": 0.6805,
      "mean_token_accuracy": 0.7914594635367393,
      "num_tokens": 402307741.0,
      "step": 12605
    },
    {
      "entropy": 0.6188222859054804,
      "epoch": 1.159815067193403,
      "grad_norm": 0.16406695544719696,
      "learning_rate": 6.134265648480388e-05,
      "loss": 0.611,
      "mean_token_accuracy": 0.8166882582008839,
      "num_tokens": 402340355.0,
      "step": 12606
    },
    {
      "entropy": 0.6386982412077487,
      "epoch": 1.1599070735648445,
      "grad_norm": 0.15928398072719574,
      "learning_rate": 6.133958965866226e-05,
      "loss": 0.6166,
      "mean_token_accuracy": 0.8090976662933826,
      "num_tokens": 402371685.0,
      "step": 12607
    },
    {
      "entropy": 0.5292554451152682,
      "epoch": 1.1599990799362856,
      "grad_norm": 0.1552577167749405,
      "learning_rate": 6.133652283252063e-05,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8441266380250454,
      "num_tokens": 402403179.0,
      "step": 12608
    },
    {
      "entropy": 0.6152433753013611,
      "epoch": 1.1600910863077267,
      "grad_norm": 0.1563802808523178,
      "learning_rate": 6.1333456006379e-05,
      "loss": 0.5966,
      "mean_token_accuracy": 0.8131097182631493,
      "num_tokens": 402434582.0,
      "step": 12609
    },
    {
      "entropy": 0.6548888105899096,
      "epoch": 1.160183092679168,
      "grad_norm": 0.16312217712402344,
      "learning_rate": 6.133038918023738e-05,
      "loss": 0.6488,
      "mean_token_accuracy": 0.8055039681494236,
      "num_tokens": 402466946.0,
      "step": 12610
    },
    {
      "entropy": 0.5758315958082676,
      "epoch": 1.1602750990506092,
      "grad_norm": 0.1509566307067871,
      "learning_rate": 6.132732235409575e-05,
      "loss": 0.5581,
      "mean_token_accuracy": 0.8257107511162758,
      "num_tokens": 402498846.0,
      "step": 12611
    },
    {
      "entropy": 0.7015667259693146,
      "epoch": 1.1603671054220506,
      "grad_norm": 0.17107658088207245,
      "learning_rate": 6.132425552795413e-05,
      "loss": 0.6825,
      "mean_token_accuracy": 0.7925826422870159,
      "num_tokens": 402530964.0,
      "step": 12612
    },
    {
      "entropy": 0.5461074290797114,
      "epoch": 1.1604591117934917,
      "grad_norm": 0.15512371063232422,
      "learning_rate": 6.13211887018125e-05,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8334028795361519,
      "num_tokens": 402562891.0,
      "step": 12613
    },
    {
      "entropy": 0.7169230543076992,
      "epoch": 1.1605511181649328,
      "grad_norm": 0.1634054034948349,
      "learning_rate": 6.131812187567088e-05,
      "loss": 0.7007,
      "mean_token_accuracy": 0.7883635461330414,
      "num_tokens": 402593000.0,
      "step": 12614
    },
    {
      "entropy": 0.7332150191068649,
      "epoch": 1.1606431245363742,
      "grad_norm": 0.16382350027561188,
      "learning_rate": 6.131505504952924e-05,
      "loss": 0.7446,
      "mean_token_accuracy": 0.773632138967514,
      "num_tokens": 402625768.0,
      "step": 12615
    },
    {
      "entropy": 0.6379959671758115,
      "epoch": 1.1607351309078153,
      "grad_norm": 0.16594724357128143,
      "learning_rate": 6.131198822338761e-05,
      "loss": 0.6268,
      "mean_token_accuracy": 0.8069444298744202,
      "num_tokens": 402658303.0,
      "step": 12616
    },
    {
      "entropy": 0.5497546847909689,
      "epoch": 1.1608271372792567,
      "grad_norm": 0.15443538129329681,
      "learning_rate": 6.1308921397246e-05,
      "loss": 0.5312,
      "mean_token_accuracy": 0.8350522369146347,
      "num_tokens": 402690236.0,
      "step": 12617
    },
    {
      "entropy": 0.6416899021714926,
      "epoch": 1.1609191436506978,
      "grad_norm": 0.15975037217140198,
      "learning_rate": 6.130585457110437e-05,
      "loss": 0.6233,
      "mean_token_accuracy": 0.8141794055700302,
      "num_tokens": 402721743.0,
      "step": 12618
    },
    {
      "entropy": 0.6891572885215282,
      "epoch": 1.161011150022139,
      "grad_norm": 0.1577051281929016,
      "learning_rate": 6.130278774496274e-05,
      "loss": 0.6733,
      "mean_token_accuracy": 0.7924104332923889,
      "num_tokens": 402753098.0,
      "step": 12619
    },
    {
      "entropy": 0.5602606711909175,
      "epoch": 1.1611031563935803,
      "grad_norm": 0.1521521955728531,
      "learning_rate": 6.129972091882111e-05,
      "loss": 0.542,
      "mean_token_accuracy": 0.8335167095065117,
      "num_tokens": 402784800.0,
      "step": 12620
    },
    {
      "entropy": 0.5860346588306129,
      "epoch": 1.1611951627650214,
      "grad_norm": 0.16238930821418762,
      "learning_rate": 6.129665409267948e-05,
      "loss": 0.5695,
      "mean_token_accuracy": 0.8219230808317661,
      "num_tokens": 402816894.0,
      "step": 12621
    },
    {
      "entropy": 0.6699978150427341,
      "epoch": 1.1612871691364628,
      "grad_norm": 0.1629120409488678,
      "learning_rate": 6.129358726653786e-05,
      "loss": 0.659,
      "mean_token_accuracy": 0.8031052015721798,
      "num_tokens": 402849492.0,
      "step": 12622
    },
    {
      "entropy": 0.4735582135617733,
      "epoch": 1.161379175507904,
      "grad_norm": 0.15115344524383545,
      "learning_rate": 6.129052044039624e-05,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8598279133439064,
      "num_tokens": 402881450.0,
      "step": 12623
    },
    {
      "entropy": 0.5911038368940353,
      "epoch": 1.161471181879345,
      "grad_norm": 0.16885314881801605,
      "learning_rate": 6.128745361425461e-05,
      "loss": 0.5825,
      "mean_token_accuracy": 0.8189284577965736,
      "num_tokens": 402913458.0,
      "step": 12624
    },
    {
      "entropy": 0.6674218215048313,
      "epoch": 1.1615631882507864,
      "grad_norm": 0.16606995463371277,
      "learning_rate": 6.128438678811298e-05,
      "loss": 0.6494,
      "mean_token_accuracy": 0.800078496336937,
      "num_tokens": 402945014.0,
      "step": 12625
    },
    {
      "entropy": 0.7274369169026613,
      "epoch": 1.1616551946222275,
      "grad_norm": 0.16600994765758514,
      "learning_rate": 6.128131996197136e-05,
      "loss": 0.7267,
      "mean_token_accuracy": 0.777463685721159,
      "num_tokens": 402976938.0,
      "step": 12626
    },
    {
      "entropy": 0.5529254321008921,
      "epoch": 1.161747200993669,
      "grad_norm": 0.16516871750354767,
      "learning_rate": 6.127825313582973e-05,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8283937871456146,
      "num_tokens": 403009050.0,
      "step": 12627
    },
    {
      "entropy": 0.6669282019138336,
      "epoch": 1.16183920736511,
      "grad_norm": 0.16189001500606537,
      "learning_rate": 6.127518630968811e-05,
      "loss": 0.6617,
      "mean_token_accuracy": 0.7996062487363815,
      "num_tokens": 403041136.0,
      "step": 12628
    },
    {
      "entropy": 0.6757936794310808,
      "epoch": 1.1619312137365512,
      "grad_norm": 0.16339272260665894,
      "learning_rate": 6.127211948354649e-05,
      "loss": 0.6646,
      "mean_token_accuracy": 0.797191709280014,
      "num_tokens": 403072796.0,
      "step": 12629
    },
    {
      "entropy": 0.6777433324605227,
      "epoch": 1.1620232201079925,
      "grad_norm": 0.16179117560386658,
      "learning_rate": 6.126905265740486e-05,
      "loss": 0.6474,
      "mean_token_accuracy": 0.7984489351511002,
      "num_tokens": 403104640.0,
      "step": 12630
    },
    {
      "entropy": 0.6064231302589178,
      "epoch": 1.1621152264794337,
      "grad_norm": 0.15480133891105652,
      "learning_rate": 6.126598583126322e-05,
      "loss": 0.5992,
      "mean_token_accuracy": 0.8180192559957504,
      "num_tokens": 403137269.0,
      "step": 12631
    },
    {
      "entropy": 0.5867715105414391,
      "epoch": 1.162207232850875,
      "grad_norm": 0.15712150931358337,
      "learning_rate": 6.12629190051216e-05,
      "loss": 0.5703,
      "mean_token_accuracy": 0.8252596594393253,
      "num_tokens": 403168339.0,
      "step": 12632
    },
    {
      "entropy": 0.7177749257534742,
      "epoch": 1.1622992392223161,
      "grad_norm": 0.15992599725723267,
      "learning_rate": 6.125985217897999e-05,
      "loss": 0.6908,
      "mean_token_accuracy": 0.789518591016531,
      "num_tokens": 403200460.0,
      "step": 12633
    },
    {
      "entropy": 0.6387641089968383,
      "epoch": 1.1623912455937573,
      "grad_norm": 0.160330668091774,
      "learning_rate": 6.125678535283836e-05,
      "loss": 0.6289,
      "mean_token_accuracy": 0.8124796003103256,
      "num_tokens": 403232364.0,
      "step": 12634
    },
    {
      "entropy": 0.6075585447251797,
      "epoch": 1.1624832519651986,
      "grad_norm": 0.15176504850387573,
      "learning_rate": 6.125371852669672e-05,
      "loss": 0.6041,
      "mean_token_accuracy": 0.8130687177181244,
      "num_tokens": 403264060.0,
      "step": 12635
    },
    {
      "entropy": 0.6591655239462852,
      "epoch": 1.1625752583366398,
      "grad_norm": 0.16431885957717896,
      "learning_rate": 6.125065170055509e-05,
      "loss": 0.6712,
      "mean_token_accuracy": 0.7990915812551975,
      "num_tokens": 403296491.0,
      "step": 12636
    },
    {
      "entropy": 0.6523353457450867,
      "epoch": 1.1626672647080811,
      "grad_norm": 0.16331468522548676,
      "learning_rate": 6.124758487441347e-05,
      "loss": 0.6632,
      "mean_token_accuracy": 0.7947211638092995,
      "num_tokens": 403328050.0,
      "step": 12637
    },
    {
      "entropy": 0.5675372797995806,
      "epoch": 1.1627592710795223,
      "grad_norm": 0.15957719087600708,
      "learning_rate": 6.124451804827185e-05,
      "loss": 0.5703,
      "mean_token_accuracy": 0.8233926855027676,
      "num_tokens": 403359756.0,
      "step": 12638
    },
    {
      "entropy": 0.5891177952289581,
      "epoch": 1.1628512774509634,
      "grad_norm": 0.15044298768043518,
      "learning_rate": 6.124145122213022e-05,
      "loss": 0.5805,
      "mean_token_accuracy": 0.8235832937061787,
      "num_tokens": 403392524.0,
      "step": 12639
    },
    {
      "entropy": 0.6632768521085382,
      "epoch": 1.1629432838224047,
      "grad_norm": 0.1667858213186264,
      "learning_rate": 6.123838439598859e-05,
      "loss": 0.6306,
      "mean_token_accuracy": 0.8088501840829849,
      "num_tokens": 403424850.0,
      "step": 12640
    },
    {
      "entropy": 0.5708592971786857,
      "epoch": 1.1630352901938459,
      "grad_norm": 0.15563218295574188,
      "learning_rate": 6.123531756984697e-05,
      "loss": 0.5585,
      "mean_token_accuracy": 0.8259898163378239,
      "num_tokens": 403456095.0,
      "step": 12641
    },
    {
      "entropy": 0.659967178478837,
      "epoch": 1.1631272965652872,
      "grad_norm": 0.15621967613697052,
      "learning_rate": 6.123225074370534e-05,
      "loss": 0.6579,
      "mean_token_accuracy": 0.8017999306321144,
      "num_tokens": 403487847.0,
      "step": 12642
    },
    {
      "entropy": 0.5758118350058794,
      "epoch": 1.1632193029367284,
      "grad_norm": 0.1479964554309845,
      "learning_rate": 6.122918391756372e-05,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8248954825103283,
      "num_tokens": 403519744.0,
      "step": 12643
    },
    {
      "entropy": 0.646555919200182,
      "epoch": 1.1633113093081695,
      "grad_norm": 0.16675490140914917,
      "learning_rate": 6.122611709142209e-05,
      "loss": 0.6225,
      "mean_token_accuracy": 0.8092065416276455,
      "num_tokens": 403551566.0,
      "step": 12644
    },
    {
      "entropy": 0.549762412905693,
      "epoch": 1.1634033156796109,
      "grad_norm": 0.1533418595790863,
      "learning_rate": 6.122305026528047e-05,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8345587626099586,
      "num_tokens": 403583329.0,
      "step": 12645
    },
    {
      "entropy": 0.5639318842440844,
      "epoch": 1.163495322051052,
      "grad_norm": 0.15899145603179932,
      "learning_rate": 6.121998343913884e-05,
      "loss": 0.5533,
      "mean_token_accuracy": 0.8298796378076077,
      "num_tokens": 403615993.0,
      "step": 12646
    },
    {
      "entropy": 0.7105090525001287,
      "epoch": 1.1635873284224933,
      "grad_norm": 0.16259311139583588,
      "learning_rate": 6.12169166129972e-05,
      "loss": 0.7069,
      "mean_token_accuracy": 0.7882751785218716,
      "num_tokens": 403646971.0,
      "step": 12647
    },
    {
      "entropy": 0.641019074479118,
      "epoch": 1.1636793347939345,
      "grad_norm": 0.16683407127857208,
      "learning_rate": 6.121384978685559e-05,
      "loss": 0.6319,
      "mean_token_accuracy": 0.8091555088758469,
      "num_tokens": 403679157.0,
      "step": 12648
    },
    {
      "entropy": 0.7378407251089811,
      "epoch": 1.1637713411653756,
      "grad_norm": 0.1687498837709427,
      "learning_rate": 6.121078296071397e-05,
      "loss": 0.7314,
      "mean_token_accuracy": 0.7814208902418613,
      "num_tokens": 403711600.0,
      "step": 12649
    },
    {
      "entropy": 0.6564894486218691,
      "epoch": 1.163863347536817,
      "grad_norm": 0.1586129069328308,
      "learning_rate": 6.120771613457234e-05,
      "loss": 0.6424,
      "mean_token_accuracy": 0.8047425858676434,
      "num_tokens": 403744219.0,
      "step": 12650
    },
    {
      "entropy": 0.6221248954534531,
      "epoch": 1.163955353908258,
      "grad_norm": 0.17044657468795776,
      "learning_rate": 6.12046493084307e-05,
      "loss": 0.6189,
      "mean_token_accuracy": 0.8106247559189796,
      "num_tokens": 403776316.0,
      "step": 12651
    },
    {
      "entropy": 0.6802108474075794,
      "epoch": 1.1640473602796995,
      "grad_norm": 0.1665278822183609,
      "learning_rate": 6.120158248228907e-05,
      "loss": 0.6866,
      "mean_token_accuracy": 0.7926533594727516,
      "num_tokens": 403808540.0,
      "step": 12652
    },
    {
      "entropy": 0.560454600956291,
      "epoch": 1.1641393666511406,
      "grad_norm": 0.16332679986953735,
      "learning_rate": 6.119851565614745e-05,
      "loss": 0.548,
      "mean_token_accuracy": 0.8286487273871899,
      "num_tokens": 403840168.0,
      "step": 12653
    },
    {
      "entropy": 0.5741901365108788,
      "epoch": 1.1642313730225817,
      "grad_norm": 0.1515815407037735,
      "learning_rate": 6.119544883000583e-05,
      "loss": 0.5571,
      "mean_token_accuracy": 0.826216708868742,
      "num_tokens": 403872743.0,
      "step": 12654
    },
    {
      "entropy": 0.6217704375740141,
      "epoch": 1.164323379394023,
      "grad_norm": 0.15673713386058807,
      "learning_rate": 6.11923820038642e-05,
      "loss": 0.625,
      "mean_token_accuracy": 0.8122352473437786,
      "num_tokens": 403904745.0,
      "step": 12655
    },
    {
      "entropy": 0.6009446289390326,
      "epoch": 1.1644153857654642,
      "grad_norm": 0.14699195325374603,
      "learning_rate": 6.118931517772257e-05,
      "loss": 0.5972,
      "mean_token_accuracy": 0.8161117769777775,
      "num_tokens": 403937385.0,
      "step": 12656
    },
    {
      "entropy": 0.6320998501032591,
      "epoch": 1.1645073921369056,
      "grad_norm": 0.15535758435726166,
      "learning_rate": 6.118624835158095e-05,
      "loss": 0.6165,
      "mean_token_accuracy": 0.8123835250735283,
      "num_tokens": 403969840.0,
      "step": 12657
    },
    {
      "entropy": 0.6674230434000492,
      "epoch": 1.1645993985083467,
      "grad_norm": 0.15902099013328552,
      "learning_rate": 6.118318152543932e-05,
      "loss": 0.6517,
      "mean_token_accuracy": 0.8011654429137707,
      "num_tokens": 404001721.0,
      "step": 12658
    },
    {
      "entropy": 0.6135320942848921,
      "epoch": 1.1646914048797878,
      "grad_norm": 0.1540921926498413,
      "learning_rate": 6.11801146992977e-05,
      "loss": 0.5965,
      "mean_token_accuracy": 0.819852102547884,
      "num_tokens": 404033447.0,
      "step": 12659
    },
    {
      "entropy": 0.5212987754493952,
      "epoch": 1.1647834112512292,
      "grad_norm": 0.15379109978675842,
      "learning_rate": 6.117704787315608e-05,
      "loss": 0.4946,
      "mean_token_accuracy": 0.8456364646553993,
      "num_tokens": 404065821.0,
      "step": 12660
    },
    {
      "entropy": 0.6714410148561001,
      "epoch": 1.1648754176226703,
      "grad_norm": 0.15691517293453217,
      "learning_rate": 6.117398104701445e-05,
      "loss": 0.6509,
      "mean_token_accuracy": 0.798434428870678,
      "num_tokens": 404097740.0,
      "step": 12661
    },
    {
      "entropy": 0.6144701475277543,
      "epoch": 1.1649674239941117,
      "grad_norm": 0.1618310511112213,
      "learning_rate": 6.117091422087282e-05,
      "loss": 0.6119,
      "mean_token_accuracy": 0.813060887157917,
      "num_tokens": 404130240.0,
      "step": 12662
    },
    {
      "entropy": 0.608712911605835,
      "epoch": 1.1650594303655528,
      "grad_norm": 0.16153372824192047,
      "learning_rate": 6.116784739473119e-05,
      "loss": 0.6079,
      "mean_token_accuracy": 0.8141636066138744,
      "num_tokens": 404162060.0,
      "step": 12663
    },
    {
      "entropy": 0.5777365323156118,
      "epoch": 1.165151436736994,
      "grad_norm": 0.1538206934928894,
      "learning_rate": 6.116478056858958e-05,
      "loss": 0.5659,
      "mean_token_accuracy": 0.82506338134408,
      "num_tokens": 404193959.0,
      "step": 12664
    },
    {
      "entropy": 0.6580703388899565,
      "epoch": 1.1652434431084353,
      "grad_norm": 0.16361920535564423,
      "learning_rate": 6.116171374244795e-05,
      "loss": 0.6473,
      "mean_token_accuracy": 0.8052406348288059,
      "num_tokens": 404225707.0,
      "step": 12665
    },
    {
      "entropy": 0.7011438459157944,
      "epoch": 1.1653354494798764,
      "grad_norm": 0.15812768042087555,
      "learning_rate": 6.115864691630632e-05,
      "loss": 0.6831,
      "mean_token_accuracy": 0.7925545200705528,
      "num_tokens": 404257593.0,
      "step": 12666
    },
    {
      "entropy": 0.6018199357204139,
      "epoch": 1.1654274558513178,
      "grad_norm": 0.15394315123558044,
      "learning_rate": 6.115558009016469e-05,
      "loss": 0.5992,
      "mean_token_accuracy": 0.818795770406723,
      "num_tokens": 404289610.0,
      "step": 12667
    },
    {
      "entropy": 0.6380753749981523,
      "epoch": 1.165519462222759,
      "grad_norm": 0.15824416279792786,
      "learning_rate": 6.115251326402307e-05,
      "loss": 0.6218,
      "mean_token_accuracy": 0.8073582835495472,
      "num_tokens": 404322076.0,
      "step": 12668
    },
    {
      "entropy": 0.6977597274817526,
      "epoch": 1.1656114685942,
      "grad_norm": 0.16361482441425323,
      "learning_rate": 6.114944643788145e-05,
      "loss": 0.6925,
      "mean_token_accuracy": 0.7896884307265282,
      "num_tokens": 404354199.0,
      "step": 12669
    },
    {
      "entropy": 0.5628902763128281,
      "epoch": 1.1657034749656414,
      "grad_norm": 0.1568567305803299,
      "learning_rate": 6.114637961173982e-05,
      "loss": 0.5496,
      "mean_token_accuracy": 0.829556155949831,
      "num_tokens": 404386014.0,
      "step": 12670
    },
    {
      "entropy": 0.6752585154026747,
      "epoch": 1.1657954813370826,
      "grad_norm": 0.16327977180480957,
      "learning_rate": 6.114331278559818e-05,
      "loss": 0.664,
      "mean_token_accuracy": 0.7973078265786171,
      "num_tokens": 404418174.0,
      "step": 12671
    },
    {
      "entropy": 0.4471165919676423,
      "epoch": 1.165887487708524,
      "grad_norm": 0.14580433070659637,
      "learning_rate": 6.114024595945657e-05,
      "loss": 0.4156,
      "mean_token_accuracy": 0.8724851086735725,
      "num_tokens": 404450339.0,
      "step": 12672
    },
    {
      "entropy": 0.6734555047005415,
      "epoch": 1.165979494079965,
      "grad_norm": 0.16563880443572998,
      "learning_rate": 6.113717913331493e-05,
      "loss": 0.6591,
      "mean_token_accuracy": 0.7983401603996754,
      "num_tokens": 404483073.0,
      "step": 12673
    },
    {
      "entropy": 0.5652454383671284,
      "epoch": 1.1660715004514062,
      "grad_norm": 0.1579791158437729,
      "learning_rate": 6.113411230717331e-05,
      "loss": 0.5591,
      "mean_token_accuracy": 0.8271444290876389,
      "num_tokens": 404514749.0,
      "step": 12674
    },
    {
      "entropy": 0.5862823133356869,
      "epoch": 1.1661635068228475,
      "grad_norm": 0.15216027200222015,
      "learning_rate": 6.113104548103168e-05,
      "loss": 0.5729,
      "mean_token_accuracy": 0.8219812475144863,
      "num_tokens": 404547323.0,
      "step": 12675
    },
    {
      "entropy": 0.6413582796230912,
      "epoch": 1.1662555131942887,
      "grad_norm": 0.1496254950761795,
      "learning_rate": 6.112797865489006e-05,
      "loss": 0.6281,
      "mean_token_accuracy": 0.8068566396832466,
      "num_tokens": 404579120.0,
      "step": 12676
    },
    {
      "entropy": 0.592612381093204,
      "epoch": 1.16634751956573,
      "grad_norm": 0.15614409744739532,
      "learning_rate": 6.112491182874843e-05,
      "loss": 0.5742,
      "mean_token_accuracy": 0.8259659297764301,
      "num_tokens": 404611334.0,
      "step": 12677
    },
    {
      "entropy": 0.5636953180655837,
      "epoch": 1.1664395259371712,
      "grad_norm": 0.1621299833059311,
      "learning_rate": 6.11218450026068e-05,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8317964673042297,
      "num_tokens": 404642704.0,
      "step": 12678
    },
    {
      "entropy": 0.5607934216968715,
      "epoch": 1.1665315323086123,
      "grad_norm": 0.1600496619939804,
      "learning_rate": 6.111877817646518e-05,
      "loss": 0.5477,
      "mean_token_accuracy": 0.8355367295444012,
      "num_tokens": 404675204.0,
      "step": 12679
    },
    {
      "entropy": 0.6474362146109343,
      "epoch": 1.1666235386800536,
      "grad_norm": 0.16199065744876862,
      "learning_rate": 6.111571135032356e-05,
      "loss": 0.6474,
      "mean_token_accuracy": 0.8057184889912605,
      "num_tokens": 404707417.0,
      "step": 12680
    },
    {
      "entropy": 0.7304923003539443,
      "epoch": 1.1667155450514948,
      "grad_norm": 0.16311100125312805,
      "learning_rate": 6.111264452418193e-05,
      "loss": 0.7275,
      "mean_token_accuracy": 0.7831490971148014,
      "num_tokens": 404739108.0,
      "step": 12681
    },
    {
      "entropy": 0.6142390742897987,
      "epoch": 1.1668075514229361,
      "grad_norm": 0.15486204624176025,
      "learning_rate": 6.11095776980403e-05,
      "loss": 0.5906,
      "mean_token_accuracy": 0.8203126154839993,
      "num_tokens": 404771017.0,
      "step": 12682
    },
    {
      "entropy": 0.6322346180677414,
      "epoch": 1.1668995577943773,
      "grad_norm": 0.15507927536964417,
      "learning_rate": 6.110651087189867e-05,
      "loss": 0.611,
      "mean_token_accuracy": 0.8168717697262764,
      "num_tokens": 404803208.0,
      "step": 12683
    },
    {
      "entropy": 0.601616065017879,
      "epoch": 1.1669915641658184,
      "grad_norm": 0.15389446914196014,
      "learning_rate": 6.110344404575705e-05,
      "loss": 0.5963,
      "mean_token_accuracy": 0.8153219632804394,
      "num_tokens": 404834844.0,
      "step": 12684
    },
    {
      "entropy": 0.5695150550454855,
      "epoch": 1.1670835705372598,
      "grad_norm": 0.15109197795391083,
      "learning_rate": 6.110037721961543e-05,
      "loss": 0.5579,
      "mean_token_accuracy": 0.8298392854630947,
      "num_tokens": 404866628.0,
      "step": 12685
    },
    {
      "entropy": 0.6998587939888239,
      "epoch": 1.1671755769087009,
      "grad_norm": 0.1677805632352829,
      "learning_rate": 6.10973103934738e-05,
      "loss": 0.6711,
      "mean_token_accuracy": 0.7927443347871304,
      "num_tokens": 404898827.0,
      "step": 12686
    },
    {
      "entropy": 0.5125611866824329,
      "epoch": 1.1672675832801422,
      "grad_norm": 0.14718079566955566,
      "learning_rate": 6.109424356733216e-05,
      "loss": 0.4997,
      "mean_token_accuracy": 0.847861610352993,
      "num_tokens": 404930339.0,
      "step": 12687
    },
    {
      "entropy": 0.5793621968477964,
      "epoch": 1.1673595896515834,
      "grad_norm": 0.14715348184108734,
      "learning_rate": 6.109117674119055e-05,
      "loss": 0.5481,
      "mean_token_accuracy": 0.8329680562019348,
      "num_tokens": 404962415.0,
      "step": 12688
    },
    {
      "entropy": 0.6750026112422347,
      "epoch": 1.1674515960230245,
      "grad_norm": 0.15986061096191406,
      "learning_rate": 6.108810991504891e-05,
      "loss": 0.669,
      "mean_token_accuracy": 0.7972557470202446,
      "num_tokens": 404994946.0,
      "step": 12689
    },
    {
      "entropy": 0.6691183932125568,
      "epoch": 1.1675436023944659,
      "grad_norm": 0.16457925736904144,
      "learning_rate": 6.10850430889073e-05,
      "loss": 0.653,
      "mean_token_accuracy": 0.7992641925811768,
      "num_tokens": 405025948.0,
      "step": 12690
    },
    {
      "entropy": 0.69153908547014,
      "epoch": 1.167635608765907,
      "grad_norm": 0.1672302931547165,
      "learning_rate": 6.108197626276568e-05,
      "loss": 0.6848,
      "mean_token_accuracy": 0.7918420918285847,
      "num_tokens": 405058687.0,
      "step": 12691
    },
    {
      "entropy": 0.6538674384355545,
      "epoch": 1.1677276151373484,
      "grad_norm": 0.16448603570461273,
      "learning_rate": 6.107890943662404e-05,
      "loss": 0.6372,
      "mean_token_accuracy": 0.8070193119347095,
      "num_tokens": 405090307.0,
      "step": 12692
    },
    {
      "entropy": 0.6837374959141016,
      "epoch": 1.1678196215087895,
      "grad_norm": 0.16034099459648132,
      "learning_rate": 6.107584261048241e-05,
      "loss": 0.6805,
      "mean_token_accuracy": 0.7939751408994198,
      "num_tokens": 405122414.0,
      "step": 12693
    },
    {
      "entropy": 0.6063252398744226,
      "epoch": 1.1679116278802306,
      "grad_norm": 0.15882453322410583,
      "learning_rate": 6.107277578434078e-05,
      "loss": 0.6038,
      "mean_token_accuracy": 0.8174955956637859,
      "num_tokens": 405154786.0,
      "step": 12694
    },
    {
      "entropy": 0.5916091427206993,
      "epoch": 1.168003634251672,
      "grad_norm": 0.15548671782016754,
      "learning_rate": 6.106970895819916e-05,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8216841705143452,
      "num_tokens": 405186926.0,
      "step": 12695
    },
    {
      "entropy": 0.5510899936780334,
      "epoch": 1.168095640623113,
      "grad_norm": 0.16780242323875427,
      "learning_rate": 6.106664213205754e-05,
      "loss": 0.5244,
      "mean_token_accuracy": 0.8385966010391712,
      "num_tokens": 405219035.0,
      "step": 12696
    },
    {
      "entropy": 0.6381193613633513,
      "epoch": 1.1681876469945545,
      "grad_norm": 0.15909194946289062,
      "learning_rate": 6.106357530591591e-05,
      "loss": 0.626,
      "mean_token_accuracy": 0.8069191500544548,
      "num_tokens": 405251016.0,
      "step": 12697
    },
    {
      "entropy": 0.577943948097527,
      "epoch": 1.1682796533659956,
      "grad_norm": 0.16247142851352692,
      "learning_rate": 6.106050847977428e-05,
      "loss": 0.5656,
      "mean_token_accuracy": 0.8297855630517006,
      "num_tokens": 405283421.0,
      "step": 12698
    },
    {
      "entropy": 0.6232094801962376,
      "epoch": 1.1683716597374367,
      "grad_norm": 0.163001149892807,
      "learning_rate": 6.105744165363266e-05,
      "loss": 0.6227,
      "mean_token_accuracy": 0.8108972422778606,
      "num_tokens": 405315818.0,
      "step": 12699
    },
    {
      "entropy": 0.5645198412239552,
      "epoch": 1.168463666108878,
      "grad_norm": 0.15448720753192902,
      "learning_rate": 6.105437482749103e-05,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8275416195392609,
      "num_tokens": 405347715.0,
      "step": 12700
    },
    {
      "entropy": 0.7325584646314383,
      "epoch": 1.1685556724803192,
      "grad_norm": 0.1644921749830246,
      "learning_rate": 6.105130800134941e-05,
      "loss": 0.7175,
      "mean_token_accuracy": 0.7844203747808933,
      "num_tokens": 405380012.0,
      "step": 12701
    },
    {
      "entropy": 0.5394227816723287,
      "epoch": 1.1686476788517606,
      "grad_norm": 0.15931645035743713,
      "learning_rate": 6.104824117520778e-05,
      "loss": 0.5283,
      "mean_token_accuracy": 0.8359356969594955,
      "num_tokens": 405411795.0,
      "step": 12702
    },
    {
      "entropy": 0.5291157746687531,
      "epoch": 1.1687396852232017,
      "grad_norm": 0.15077249705791473,
      "learning_rate": 6.104517434906616e-05,
      "loss": 0.518,
      "mean_token_accuracy": 0.8352096937596798,
      "num_tokens": 405443260.0,
      "step": 12703
    },
    {
      "entropy": 0.5970950108021498,
      "epoch": 1.1688316915946428,
      "grad_norm": 0.15855784714221954,
      "learning_rate": 6.104210752292453e-05,
      "loss": 0.5745,
      "mean_token_accuracy": 0.8241670988500118,
      "num_tokens": 405475148.0,
      "step": 12704
    },
    {
      "entropy": 0.5294525912031531,
      "epoch": 1.1689236979660842,
      "grad_norm": 0.14721767604351044,
      "learning_rate": 6.10390406967829e-05,
      "loss": 0.5253,
      "mean_token_accuracy": 0.8384719714522362,
      "num_tokens": 405507243.0,
      "step": 12705
    },
    {
      "entropy": 0.7737531643360853,
      "epoch": 1.1690157043375253,
      "grad_norm": 0.1737314760684967,
      "learning_rate": 6.103597387064128e-05,
      "loss": 0.7732,
      "mean_token_accuracy": 0.7665000967681408,
      "num_tokens": 405539581.0,
      "step": 12706
    },
    {
      "entropy": 0.6552637871354818,
      "epoch": 1.1691077107089667,
      "grad_norm": 0.1599409431219101,
      "learning_rate": 6.103290704449966e-05,
      "loss": 0.6549,
      "mean_token_accuracy": 0.8027216158807278,
      "num_tokens": 405571167.0,
      "step": 12707
    },
    {
      "entropy": 0.665824631229043,
      "epoch": 1.1691997170804078,
      "grad_norm": 0.16301193833351135,
      "learning_rate": 6.1029840218358025e-05,
      "loss": 0.6513,
      "mean_token_accuracy": 0.798660896718502,
      "num_tokens": 405602802.0,
      "step": 12708
    },
    {
      "entropy": 0.6641330327838659,
      "epoch": 1.169291723451849,
      "grad_norm": 0.16691391170024872,
      "learning_rate": 6.102677339221639e-05,
      "loss": 0.6341,
      "mean_token_accuracy": 0.8060023710131645,
      "num_tokens": 405633892.0,
      "step": 12709
    },
    {
      "entropy": 0.6230142265558243,
      "epoch": 1.1693837298232903,
      "grad_norm": 0.1579827070236206,
      "learning_rate": 6.102370656607477e-05,
      "loss": 0.5967,
      "mean_token_accuracy": 0.8148881606757641,
      "num_tokens": 405665082.0,
      "step": 12710
    },
    {
      "entropy": 0.5628484822809696,
      "epoch": 1.1694757361947314,
      "grad_norm": 0.15280374884605408,
      "learning_rate": 6.102063973993315e-05,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8325560763478279,
      "num_tokens": 405697675.0,
      "step": 12711
    },
    {
      "entropy": 0.6220531985163689,
      "epoch": 1.1695677425661728,
      "grad_norm": 0.15543076395988464,
      "learning_rate": 6.1017572913791524e-05,
      "loss": 0.6018,
      "mean_token_accuracy": 0.8119761720299721,
      "num_tokens": 405729707.0,
      "step": 12712
    },
    {
      "entropy": 0.6699754875153303,
      "epoch": 1.169659748937614,
      "grad_norm": 0.1673119217157364,
      "learning_rate": 6.101450608764989e-05,
      "loss": 0.6494,
      "mean_token_accuracy": 0.8008839264512062,
      "num_tokens": 405761566.0,
      "step": 12713
    },
    {
      "entropy": 0.49218873074278235,
      "epoch": 1.169751755309055,
      "grad_norm": 0.14342951774597168,
      "learning_rate": 6.1011439261508266e-05,
      "loss": 0.4804,
      "mean_token_accuracy": 0.8518491089344025,
      "num_tokens": 405793737.0,
      "step": 12714
    },
    {
      "entropy": 0.6767038386315107,
      "epoch": 1.1698437616804964,
      "grad_norm": 0.1637333780527115,
      "learning_rate": 6.100837243536665e-05,
      "loss": 0.6675,
      "mean_token_accuracy": 0.7993181273341179,
      "num_tokens": 405826126.0,
      "step": 12715
    },
    {
      "entropy": 0.5841982550919056,
      "epoch": 1.1699357680519376,
      "grad_norm": 0.15549936890602112,
      "learning_rate": 6.1005305609225016e-05,
      "loss": 0.5759,
      "mean_token_accuracy": 0.8216799758374691,
      "num_tokens": 405858233.0,
      "step": 12716
    },
    {
      "entropy": 0.6161065488122404,
      "epoch": 1.170027774423379,
      "grad_norm": 0.15177151560783386,
      "learning_rate": 6.100223878308339e-05,
      "loss": 0.6179,
      "mean_token_accuracy": 0.8109329454600811,
      "num_tokens": 405889928.0,
      "step": 12717
    },
    {
      "entropy": 0.6372742149978876,
      "epoch": 1.17011978079482,
      "grad_norm": 0.16659508645534515,
      "learning_rate": 6.099917195694176e-05,
      "loss": 0.6327,
      "mean_token_accuracy": 0.8055512197315693,
      "num_tokens": 405921831.0,
      "step": 12718
    },
    {
      "entropy": 0.6809663232415915,
      "epoch": 1.1702117871662612,
      "grad_norm": 0.15969699621200562,
      "learning_rate": 6.099610513080014e-05,
      "loss": 0.6758,
      "mean_token_accuracy": 0.7934380993247032,
      "num_tokens": 405953110.0,
      "step": 12719
    },
    {
      "entropy": 0.6315272403880954,
      "epoch": 1.1703037935377025,
      "grad_norm": 0.15657590329647064,
      "learning_rate": 6.0993038304658514e-05,
      "loss": 0.6255,
      "mean_token_accuracy": 0.8075342699885368,
      "num_tokens": 405985258.0,
      "step": 12720
    },
    {
      "entropy": 0.5875232629477978,
      "epoch": 1.1703957999091437,
      "grad_norm": 0.15818507969379425,
      "learning_rate": 6.098997147851688e-05,
      "loss": 0.5753,
      "mean_token_accuracy": 0.8245118819177151,
      "num_tokens": 406017739.0,
      "step": 12721
    },
    {
      "entropy": 0.5230427719652653,
      "epoch": 1.170487806280585,
      "grad_norm": 0.15181657671928406,
      "learning_rate": 6.0986904652375264e-05,
      "loss": 0.5192,
      "mean_token_accuracy": 0.8410500250756741,
      "num_tokens": 406049559.0,
      "step": 12722
    },
    {
      "entropy": 0.6033643372356892,
      "epoch": 1.1705798126520262,
      "grad_norm": 0.1589975357055664,
      "learning_rate": 6.098383782623364e-05,
      "loss": 0.5912,
      "mean_token_accuracy": 0.8182668760418892,
      "num_tokens": 406081360.0,
      "step": 12723
    },
    {
      "entropy": 0.6110208742320538,
      "epoch": 1.1706718190234673,
      "grad_norm": 0.15479236841201782,
      "learning_rate": 6.0980771000092006e-05,
      "loss": 0.5928,
      "mean_token_accuracy": 0.8161911591887474,
      "num_tokens": 406113531.0,
      "step": 12724
    },
    {
      "entropy": 0.49226576648652554,
      "epoch": 1.1707638253949086,
      "grad_norm": 0.15131348371505737,
      "learning_rate": 6.097770417395038e-05,
      "loss": 0.4863,
      "mean_token_accuracy": 0.8496181145310402,
      "num_tokens": 406145237.0,
      "step": 12725
    },
    {
      "entropy": 0.6405715327709913,
      "epoch": 1.1708558317663498,
      "grad_norm": 0.15978476405143738,
      "learning_rate": 6.097463734780876e-05,
      "loss": 0.6347,
      "mean_token_accuracy": 0.8054680824279785,
      "num_tokens": 406177541.0,
      "step": 12726
    },
    {
      "entropy": 0.6383450021967292,
      "epoch": 1.1709478381377911,
      "grad_norm": 0.166026771068573,
      "learning_rate": 6.097157052166713e-05,
      "loss": 0.628,
      "mean_token_accuracy": 0.8074133545160294,
      "num_tokens": 406209056.0,
      "step": 12727
    },
    {
      "entropy": 0.5835109576582909,
      "epoch": 1.1710398445092323,
      "grad_norm": 0.15402765572071075,
      "learning_rate": 6.0968503695525505e-05,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8260383307933807,
      "num_tokens": 406240757.0,
      "step": 12728
    },
    {
      "entropy": 0.579437181353569,
      "epoch": 1.1711318508806734,
      "grad_norm": 0.15311574935913086,
      "learning_rate": 6.096543686938387e-05,
      "loss": 0.56,
      "mean_token_accuracy": 0.8240049965679646,
      "num_tokens": 406272101.0,
      "step": 12729
    },
    {
      "entropy": 0.5860629230737686,
      "epoch": 1.1712238572521148,
      "grad_norm": 0.1642579287290573,
      "learning_rate": 6.0962370043242254e-05,
      "loss": 0.5817,
      "mean_token_accuracy": 0.8216734230518341,
      "num_tokens": 406304163.0,
      "step": 12730
    },
    {
      "entropy": 0.6786310449242592,
      "epoch": 1.1713158636235559,
      "grad_norm": 0.16082462668418884,
      "learning_rate": 6.095930321710063e-05,
      "loss": 0.6625,
      "mean_token_accuracy": 0.7992794252932072,
      "num_tokens": 406336931.0,
      "step": 12731
    },
    {
      "entropy": 0.5668971692211926,
      "epoch": 1.1714078699949972,
      "grad_norm": 0.15580295026302338,
      "learning_rate": 6.0956236390958996e-05,
      "loss": 0.553,
      "mean_token_accuracy": 0.832791656255722,
      "num_tokens": 406369662.0,
      "step": 12732
    },
    {
      "entropy": 0.6670769080519676,
      "epoch": 1.1714998763664384,
      "grad_norm": 0.1614105999469757,
      "learning_rate": 6.095316956481737e-05,
      "loss": 0.6382,
      "mean_token_accuracy": 0.8110053464770317,
      "num_tokens": 406402076.0,
      "step": 12733
    },
    {
      "entropy": 0.679668053984642,
      "epoch": 1.1715918827378795,
      "grad_norm": 0.16015096008777618,
      "learning_rate": 6.095010273867575e-05,
      "loss": 0.6608,
      "mean_token_accuracy": 0.7990810461342335,
      "num_tokens": 406433934.0,
      "step": 12734
    },
    {
      "entropy": 0.6125657539814711,
      "epoch": 1.1716838891093209,
      "grad_norm": 0.15865980088710785,
      "learning_rate": 6.094703591253412e-05,
      "loss": 0.5911,
      "mean_token_accuracy": 0.8176275789737701,
      "num_tokens": 406465888.0,
      "step": 12735
    },
    {
      "entropy": 0.5641652224585414,
      "epoch": 1.171775895480762,
      "grad_norm": 0.15833981335163116,
      "learning_rate": 6.0943969086392495e-05,
      "loss": 0.561,
      "mean_token_accuracy": 0.8337734863162041,
      "num_tokens": 406497513.0,
      "step": 12736
    },
    {
      "entropy": 0.5846628863364458,
      "epoch": 1.1718679018522034,
      "grad_norm": 0.1574895679950714,
      "learning_rate": 6.094090226025086e-05,
      "loss": 0.5661,
      "mean_token_accuracy": 0.820073951035738,
      "num_tokens": 406528617.0,
      "step": 12737
    },
    {
      "entropy": 0.5478672115132213,
      "epoch": 1.1719599082236445,
      "grad_norm": 0.14957524836063385,
      "learning_rate": 6.0937835434109244e-05,
      "loss": 0.5267,
      "mean_token_accuracy": 0.8354589864611626,
      "num_tokens": 406560428.0,
      "step": 12738
    },
    {
      "entropy": 0.6382253263145685,
      "epoch": 1.1720519145950856,
      "grad_norm": 0.16551603376865387,
      "learning_rate": 6.093476860796762e-05,
      "loss": 0.6368,
      "mean_token_accuracy": 0.8070534393191338,
      "num_tokens": 406593062.0,
      "step": 12739
    },
    {
      "entropy": 0.6000440642237663,
      "epoch": 1.172143920966527,
      "grad_norm": 0.15175332129001617,
      "learning_rate": 6.093170178182599e-05,
      "loss": 0.5949,
      "mean_token_accuracy": 0.8158478401601315,
      "num_tokens": 406624832.0,
      "step": 12740
    },
    {
      "entropy": 0.6065329629927874,
      "epoch": 1.172235927337968,
      "grad_norm": 0.165225088596344,
      "learning_rate": 6.092863495568436e-05,
      "loss": 0.5966,
      "mean_token_accuracy": 0.8116251826286316,
      "num_tokens": 406656785.0,
      "step": 12741
    },
    {
      "entropy": 0.6450487766414881,
      "epoch": 1.1723279337094095,
      "grad_norm": 0.16548706591129303,
      "learning_rate": 6.092556812954274e-05,
      "loss": 0.6158,
      "mean_token_accuracy": 0.8075078092515469,
      "num_tokens": 406688574.0,
      "step": 12742
    },
    {
      "entropy": 0.6206421134993434,
      "epoch": 1.1724199400808506,
      "grad_norm": 0.1592569500207901,
      "learning_rate": 6.092250130340111e-05,
      "loss": 0.6124,
      "mean_token_accuracy": 0.8134020455181599,
      "num_tokens": 406720528.0,
      "step": 12743
    },
    {
      "entropy": 0.5082128681242466,
      "epoch": 1.1725119464522917,
      "grad_norm": 0.1503574401140213,
      "learning_rate": 6.0919434477259485e-05,
      "loss": 0.4932,
      "mean_token_accuracy": 0.8447812087833881,
      "num_tokens": 406752599.0,
      "step": 12744
    },
    {
      "entropy": 0.7528585232794285,
      "epoch": 1.172603952823733,
      "grad_norm": 0.16674205660820007,
      "learning_rate": 6.091636765111785e-05,
      "loss": 0.7379,
      "mean_token_accuracy": 0.7771284990012646,
      "num_tokens": 406784681.0,
      "step": 12745
    },
    {
      "entropy": 0.6262429784983397,
      "epoch": 1.1726959591951742,
      "grad_norm": 0.15923461318016052,
      "learning_rate": 6.091330082497624e-05,
      "loss": 0.6251,
      "mean_token_accuracy": 0.8109865337610245,
      "num_tokens": 406816422.0,
      "step": 12746
    },
    {
      "entropy": 0.6536807790398598,
      "epoch": 1.1727879655666156,
      "grad_norm": 0.15723265707492828,
      "learning_rate": 6.091023399883461e-05,
      "loss": 0.6286,
      "mean_token_accuracy": 0.8063134253025055,
      "num_tokens": 406848408.0,
      "step": 12747
    },
    {
      "entropy": 0.6540792919695377,
      "epoch": 1.1728799719380567,
      "grad_norm": 0.15710540115833282,
      "learning_rate": 6.090716717269298e-05,
      "loss": 0.6304,
      "mean_token_accuracy": 0.8074497878551483,
      "num_tokens": 406880507.0,
      "step": 12748
    },
    {
      "entropy": 0.6005316935479641,
      "epoch": 1.1729719783094978,
      "grad_norm": 0.16284377872943878,
      "learning_rate": 6.090410034655135e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8189321793615818,
      "num_tokens": 406912288.0,
      "step": 12749
    },
    {
      "entropy": 0.6312586665153503,
      "epoch": 1.1730639846809392,
      "grad_norm": 0.15743710100650787,
      "learning_rate": 6.090103352040973e-05,
      "loss": 0.6247,
      "mean_token_accuracy": 0.8107891827821732,
      "num_tokens": 406944617.0,
      "step": 12750
    },
    {
      "entropy": 0.6123976269736886,
      "epoch": 1.1731559910523803,
      "grad_norm": 0.1626615673303604,
      "learning_rate": 6.089796669426811e-05,
      "loss": 0.6047,
      "mean_token_accuracy": 0.815016470849514,
      "num_tokens": 406976434.0,
      "step": 12751
    },
    {
      "entropy": 0.6256641624495387,
      "epoch": 1.1732479974238217,
      "grad_norm": 0.15701745450496674,
      "learning_rate": 6.0894899868126476e-05,
      "loss": 0.6056,
      "mean_token_accuracy": 0.8141790814697742,
      "num_tokens": 407008630.0,
      "step": 12752
    },
    {
      "entropy": 0.541200352832675,
      "epoch": 1.1733400037952628,
      "grad_norm": 0.14727966487407684,
      "learning_rate": 6.089183304198486e-05,
      "loss": 0.538,
      "mean_token_accuracy": 0.8364496938884258,
      "num_tokens": 407040857.0,
      "step": 12753
    },
    {
      "entropy": 0.7370462836697698,
      "epoch": 1.173432010166704,
      "grad_norm": 0.17198871076107025,
      "learning_rate": 6.088876621584323e-05,
      "loss": 0.747,
      "mean_token_accuracy": 0.7782877162098885,
      "num_tokens": 407071911.0,
      "step": 12754
    },
    {
      "entropy": 0.5717560742050409,
      "epoch": 1.1735240165381453,
      "grad_norm": 0.15651968121528625,
      "learning_rate": 6.08856993897016e-05,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8280287571251392,
      "num_tokens": 407103481.0,
      "step": 12755
    },
    {
      "entropy": 0.5864193048328161,
      "epoch": 1.1736160229095864,
      "grad_norm": 0.15488377213478088,
      "learning_rate": 6.0882632563559974e-05,
      "loss": 0.5782,
      "mean_token_accuracy": 0.81834477186203,
      "num_tokens": 407135099.0,
      "step": 12756
    },
    {
      "entropy": 0.5608217688277364,
      "epoch": 1.1737080292810278,
      "grad_norm": 0.15422746539115906,
      "learning_rate": 6.0879565737418356e-05,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8386127538979053,
      "num_tokens": 407167397.0,
      "step": 12757
    },
    {
      "entropy": 0.6659384202212095,
      "epoch": 1.173800035652469,
      "grad_norm": 0.17444126307964325,
      "learning_rate": 6.0876498911276724e-05,
      "loss": 0.6622,
      "mean_token_accuracy": 0.7979831509292126,
      "num_tokens": 407198594.0,
      "step": 12758
    },
    {
      "entropy": 0.6683011446148157,
      "epoch": 1.17389204202391,
      "grad_norm": 0.1631225198507309,
      "learning_rate": 6.08734320851351e-05,
      "loss": 0.6597,
      "mean_token_accuracy": 0.7949567437171936,
      "num_tokens": 407229630.0,
      "step": 12759
    },
    {
      "entropy": 0.5211081746965647,
      "epoch": 1.1739840483953514,
      "grad_norm": 0.15671630203723907,
      "learning_rate": 6.0870365258993466e-05,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8415449261665344,
      "num_tokens": 407261655.0,
      "step": 12760
    },
    {
      "entropy": 0.746961148455739,
      "epoch": 1.1740760547667926,
      "grad_norm": 0.16738229990005493,
      "learning_rate": 6.086729843285185e-05,
      "loss": 0.7464,
      "mean_token_accuracy": 0.77591847255826,
      "num_tokens": 407293024.0,
      "step": 12761
    },
    {
      "entropy": 0.5106835281476378,
      "epoch": 1.174168061138234,
      "grad_norm": 0.1558929830789566,
      "learning_rate": 6.086423160671022e-05,
      "loss": 0.4979,
      "mean_token_accuracy": 0.847391851246357,
      "num_tokens": 407325325.0,
      "step": 12762
    },
    {
      "entropy": 0.5200215689837933,
      "epoch": 1.174260067509675,
      "grad_norm": 0.16874666512012482,
      "learning_rate": 6.086116478056859e-05,
      "loss": 0.5082,
      "mean_token_accuracy": 0.8433544561266899,
      "num_tokens": 407357822.0,
      "step": 12763
    },
    {
      "entropy": 0.6859582932665944,
      "epoch": 1.1743520738811162,
      "grad_norm": 0.16303814947605133,
      "learning_rate": 6.0858097954426965e-05,
      "loss": 0.6778,
      "mean_token_accuracy": 0.7951883934438229,
      "num_tokens": 407389774.0,
      "step": 12764
    },
    {
      "entropy": 0.7917574141174555,
      "epoch": 1.1744440802525575,
      "grad_norm": 0.17307135462760925,
      "learning_rate": 6.0855031128285346e-05,
      "loss": 0.7946,
      "mean_token_accuracy": 0.7602671384811401,
      "num_tokens": 407421997.0,
      "step": 12765
    },
    {
      "entropy": 0.6167860440909863,
      "epoch": 1.1745360866239987,
      "grad_norm": 0.1579926460981369,
      "learning_rate": 6.0851964302143714e-05,
      "loss": 0.6094,
      "mean_token_accuracy": 0.8138345591723919,
      "num_tokens": 407453959.0,
      "step": 12766
    },
    {
      "entropy": 0.6673713196069002,
      "epoch": 1.17462809299544,
      "grad_norm": 0.15836398303508759,
      "learning_rate": 6.084889747600209e-05,
      "loss": 0.6512,
      "mean_token_accuracy": 0.7983118519186974,
      "num_tokens": 407486515.0,
      "step": 12767
    },
    {
      "entropy": 0.5780379790812731,
      "epoch": 1.1747200993668812,
      "grad_norm": 0.1591811180114746,
      "learning_rate": 6.084583064986046e-05,
      "loss": 0.5746,
      "mean_token_accuracy": 0.8228588849306107,
      "num_tokens": 407517789.0,
      "step": 12768
    },
    {
      "entropy": 0.6533560464158654,
      "epoch": 1.1748121057383223,
      "grad_norm": 0.1558014154434204,
      "learning_rate": 6.084276382371884e-05,
      "loss": 0.6386,
      "mean_token_accuracy": 0.80068514123559,
      "num_tokens": 407549182.0,
      "step": 12769
    },
    {
      "entropy": 0.5427117787767202,
      "epoch": 1.1749041121097636,
      "grad_norm": 0.14908117055892944,
      "learning_rate": 6.083969699757721e-05,
      "loss": 0.521,
      "mean_token_accuracy": 0.8413454480469227,
      "num_tokens": 407580883.0,
      "step": 12770
    },
    {
      "entropy": 0.6261754110455513,
      "epoch": 1.1749961184812048,
      "grad_norm": 0.15955790877342224,
      "learning_rate": 6.083663017143558e-05,
      "loss": 0.6152,
      "mean_token_accuracy": 0.8095375150442123,
      "num_tokens": 407612996.0,
      "step": 12771
    },
    {
      "entropy": 0.6684248084202409,
      "epoch": 1.1750881248526461,
      "grad_norm": 0.1621030569076538,
      "learning_rate": 6.0833563345293955e-05,
      "loss": 0.6484,
      "mean_token_accuracy": 0.7985404320061207,
      "num_tokens": 407644111.0,
      "step": 12772
    },
    {
      "entropy": 0.569434130564332,
      "epoch": 1.1751801312240873,
      "grad_norm": 0.15266858041286469,
      "learning_rate": 6.083049651915234e-05,
      "loss": 0.5556,
      "mean_token_accuracy": 0.8286662511527538,
      "num_tokens": 407676260.0,
      "step": 12773
    },
    {
      "entropy": 0.5449368534609675,
      "epoch": 1.1752721375955284,
      "grad_norm": 0.14624515175819397,
      "learning_rate": 6.0827429693010705e-05,
      "loss": 0.5277,
      "mean_token_accuracy": 0.8365836888551712,
      "num_tokens": 407708697.0,
      "step": 12774
    },
    {
      "entropy": 0.6322678960859776,
      "epoch": 1.1753641439669698,
      "grad_norm": 0.15531550347805023,
      "learning_rate": 6.082436286686908e-05,
      "loss": 0.6111,
      "mean_token_accuracy": 0.8170810453593731,
      "num_tokens": 407741058.0,
      "step": 12775
    },
    {
      "entropy": 0.700784420594573,
      "epoch": 1.1754561503384109,
      "grad_norm": 0.16285955905914307,
      "learning_rate": 6.082129604072745e-05,
      "loss": 0.6938,
      "mean_token_accuracy": 0.7916183769702911,
      "num_tokens": 407772970.0,
      "step": 12776
    },
    {
      "entropy": 0.6470303107053041,
      "epoch": 1.1755481567098522,
      "grad_norm": 0.15695512294769287,
      "learning_rate": 6.081822921458583e-05,
      "loss": 0.6242,
      "mean_token_accuracy": 0.8102377317845821,
      "num_tokens": 407804803.0,
      "step": 12777
    },
    {
      "entropy": 0.5581740560010076,
      "epoch": 1.1756401630812934,
      "grad_norm": 0.15515203773975372,
      "learning_rate": 6.08151623884442e-05,
      "loss": 0.5446,
      "mean_token_accuracy": 0.8319455981254578,
      "num_tokens": 407836273.0,
      "step": 12778
    },
    {
      "entropy": 0.62677101790905,
      "epoch": 1.1757321694527345,
      "grad_norm": 0.15673108398914337,
      "learning_rate": 6.081209556230257e-05,
      "loss": 0.6023,
      "mean_token_accuracy": 0.8144170716404915,
      "num_tokens": 407867990.0,
      "step": 12779
    },
    {
      "entropy": 0.6344245970249176,
      "epoch": 1.1758241758241759,
      "grad_norm": 0.16233284771442413,
      "learning_rate": 6.0809028736160946e-05,
      "loss": 0.6249,
      "mean_token_accuracy": 0.8097932524979115,
      "num_tokens": 407899190.0,
      "step": 12780
    },
    {
      "entropy": 0.6026475699618459,
      "epoch": 1.175916182195617,
      "grad_norm": 0.1590389907360077,
      "learning_rate": 6.080596191001933e-05,
      "loss": 0.5813,
      "mean_token_accuracy": 0.8210140652954578,
      "num_tokens": 407931440.0,
      "step": 12781
    },
    {
      "entropy": 0.559157052077353,
      "epoch": 1.1760081885670584,
      "grad_norm": 0.16486454010009766,
      "learning_rate": 6.0802895083877695e-05,
      "loss": 0.5478,
      "mean_token_accuracy": 0.8316208980977535,
      "num_tokens": 407963820.0,
      "step": 12782
    },
    {
      "entropy": 0.5830582622438669,
      "epoch": 1.1761001949384995,
      "grad_norm": 0.16232247650623322,
      "learning_rate": 6.079982825773607e-05,
      "loss": 0.5722,
      "mean_token_accuracy": 0.8205007761716843,
      "num_tokens": 407994513.0,
      "step": 12783
    },
    {
      "entropy": 0.6326200356706977,
      "epoch": 1.1761922013099406,
      "grad_norm": 0.15721987187862396,
      "learning_rate": 6.079676143159445e-05,
      "loss": 0.6175,
      "mean_token_accuracy": 0.8051367402076721,
      "num_tokens": 408026298.0,
      "step": 12784
    },
    {
      "entropy": 0.6334924604743719,
      "epoch": 1.176284207681382,
      "grad_norm": 0.15888625383377075,
      "learning_rate": 6.079369460545282e-05,
      "loss": 0.6294,
      "mean_token_accuracy": 0.8097102493047714,
      "num_tokens": 408058770.0,
      "step": 12785
    },
    {
      "entropy": 0.6473674811422825,
      "epoch": 1.176376214052823,
      "grad_norm": 0.16651202738285065,
      "learning_rate": 6.0790627779311194e-05,
      "loss": 0.6455,
      "mean_token_accuracy": 0.8025207854807377,
      "num_tokens": 408090520.0,
      "step": 12786
    },
    {
      "entropy": 0.7553344294428825,
      "epoch": 1.1764682204242645,
      "grad_norm": 0.170662522315979,
      "learning_rate": 6.078756095316956e-05,
      "loss": 0.7724,
      "mean_token_accuracy": 0.7711690701544285,
      "num_tokens": 408122335.0,
      "step": 12787
    },
    {
      "entropy": 0.6359975654631853,
      "epoch": 1.1765602267957056,
      "grad_norm": 0.1631392389535904,
      "learning_rate": 6.078449412702795e-05,
      "loss": 0.6031,
      "mean_token_accuracy": 0.8163022436201572,
      "num_tokens": 408153677.0,
      "step": 12788
    },
    {
      "entropy": 0.6694142054766417,
      "epoch": 1.1766522331671467,
      "grad_norm": 0.16092340648174286,
      "learning_rate": 6.078142730088632e-05,
      "loss": 0.6626,
      "mean_token_accuracy": 0.7988073192536831,
      "num_tokens": 408184893.0,
      "step": 12789
    },
    {
      "entropy": 0.6702719815075397,
      "epoch": 1.176744239538588,
      "grad_norm": 0.16322875022888184,
      "learning_rate": 6.077836047474469e-05,
      "loss": 0.6471,
      "mean_token_accuracy": 0.8003043942153454,
      "num_tokens": 408216305.0,
      "step": 12790
    },
    {
      "entropy": 0.7211842183023691,
      "epoch": 1.1768362459100292,
      "grad_norm": 0.17251920700073242,
      "learning_rate": 6.077529364860306e-05,
      "loss": 0.6992,
      "mean_token_accuracy": 0.7845070362091064,
      "num_tokens": 408247185.0,
      "step": 12791
    },
    {
      "entropy": 0.5805152431130409,
      "epoch": 1.1769282522814706,
      "grad_norm": 0.159579336643219,
      "learning_rate": 6.077222682246144e-05,
      "loss": 0.5655,
      "mean_token_accuracy": 0.8252116292715073,
      "num_tokens": 408279222.0,
      "step": 12792
    },
    {
      "entropy": 0.7079675663262606,
      "epoch": 1.1770202586529117,
      "grad_norm": 0.16709178686141968,
      "learning_rate": 6.0769159996319816e-05,
      "loss": 0.7056,
      "mean_token_accuracy": 0.7900221087038517,
      "num_tokens": 408311305.0,
      "step": 12793
    },
    {
      "entropy": 0.7560301003977656,
      "epoch": 1.1771122650243528,
      "grad_norm": 0.17371408641338348,
      "learning_rate": 6.0766093170178184e-05,
      "loss": 0.7465,
      "mean_token_accuracy": 0.7768257856369019,
      "num_tokens": 408342265.0,
      "step": 12794
    },
    {
      "entropy": 0.539337164722383,
      "epoch": 1.1772042713957942,
      "grad_norm": 0.15573018789291382,
      "learning_rate": 6.076302634403656e-05,
      "loss": 0.5236,
      "mean_token_accuracy": 0.8371158577501774,
      "num_tokens": 408374784.0,
      "step": 12795
    },
    {
      "entropy": 0.64905821159482,
      "epoch": 1.1772962777672353,
      "grad_norm": 0.16486600041389465,
      "learning_rate": 6.075995951789494e-05,
      "loss": 0.6599,
      "mean_token_accuracy": 0.7991401068866253,
      "num_tokens": 408406808.0,
      "step": 12796
    },
    {
      "entropy": 0.6805592412129045,
      "epoch": 1.1773882841386767,
      "grad_norm": 0.16640853881835938,
      "learning_rate": 6.075689269175331e-05,
      "loss": 0.6802,
      "mean_token_accuracy": 0.7973605021834373,
      "num_tokens": 408438421.0,
      "step": 12797
    },
    {
      "entropy": 0.6323010083287954,
      "epoch": 1.1774802905101178,
      "grad_norm": 0.15358193218708038,
      "learning_rate": 6.075382586561168e-05,
      "loss": 0.6252,
      "mean_token_accuracy": 0.8140072375535965,
      "num_tokens": 408470679.0,
      "step": 12798
    },
    {
      "entropy": 0.6846966538578272,
      "epoch": 1.177572296881559,
      "grad_norm": 0.15838094055652618,
      "learning_rate": 6.075075903947005e-05,
      "loss": 0.6599,
      "mean_token_accuracy": 0.8009763434529305,
      "num_tokens": 408502661.0,
      "step": 12799
    },
    {
      "entropy": 0.6293739769607782,
      "epoch": 1.1776643032530003,
      "grad_norm": 0.15254352986812592,
      "learning_rate": 6.074769221332843e-05,
      "loss": 0.6231,
      "mean_token_accuracy": 0.8071268796920776,
      "num_tokens": 408535104.0,
      "step": 12800
    },
    {
      "entropy": 0.6550117917358875,
      "epoch": 1.1777563096244414,
      "grad_norm": 0.15782584249973297,
      "learning_rate": 6.0744625387186806e-05,
      "loss": 0.6426,
      "mean_token_accuracy": 0.8051098883152008,
      "num_tokens": 408567210.0,
      "step": 12801
    },
    {
      "entropy": 0.6591804204508662,
      "epoch": 1.1778483159958828,
      "grad_norm": 0.15834955871105194,
      "learning_rate": 6.0741558561045174e-05,
      "loss": 0.6487,
      "mean_token_accuracy": 0.8050226606428623,
      "num_tokens": 408599761.0,
      "step": 12802
    },
    {
      "entropy": 0.6123838163912296,
      "epoch": 1.177940322367324,
      "grad_norm": 0.15773789584636688,
      "learning_rate": 6.073849173490355e-05,
      "loss": 0.587,
      "mean_token_accuracy": 0.8169838227331638,
      "num_tokens": 408631734.0,
      "step": 12803
    },
    {
      "entropy": 0.7924594320356846,
      "epoch": 1.178032328738765,
      "grad_norm": 0.16763360798358917,
      "learning_rate": 6.073542490876193e-05,
      "loss": 0.7741,
      "mean_token_accuracy": 0.7658708766102791,
      "num_tokens": 408664365.0,
      "step": 12804
    },
    {
      "entropy": 0.6407094392925501,
      "epoch": 1.1781243351102064,
      "grad_norm": 0.1614852100610733,
      "learning_rate": 6.07323580826203e-05,
      "loss": 0.6377,
      "mean_token_accuracy": 0.8016873486340046,
      "num_tokens": 408696640.0,
      "step": 12805
    },
    {
      "entropy": 0.6743394583463669,
      "epoch": 1.1782163414816476,
      "grad_norm": 0.15859590470790863,
      "learning_rate": 6.072929125647867e-05,
      "loss": 0.6583,
      "mean_token_accuracy": 0.7972917072474957,
      "num_tokens": 408728564.0,
      "step": 12806
    },
    {
      "entropy": 0.5978240319527686,
      "epoch": 1.178308347853089,
      "grad_norm": 0.16349764168262482,
      "learning_rate": 6.072622443033704e-05,
      "loss": 0.5788,
      "mean_token_accuracy": 0.8271327540278435,
      "num_tokens": 408760315.0,
      "step": 12807
    },
    {
      "entropy": 0.6137932622805238,
      "epoch": 1.17840035422453,
      "grad_norm": 0.15444286167621613,
      "learning_rate": 6.072315760419542e-05,
      "loss": 0.6032,
      "mean_token_accuracy": 0.8114595152437687,
      "num_tokens": 408792360.0,
      "step": 12808
    },
    {
      "entropy": 0.5973086170852184,
      "epoch": 1.1784923605959712,
      "grad_norm": 0.16439910233020782,
      "learning_rate": 6.07200907780538e-05,
      "loss": 0.5815,
      "mean_token_accuracy": 0.8181856982409954,
      "num_tokens": 408824498.0,
      "step": 12809
    },
    {
      "entropy": 0.6200208365917206,
      "epoch": 1.1785843669674125,
      "grad_norm": 0.1558399796485901,
      "learning_rate": 6.0717023951912165e-05,
      "loss": 0.6123,
      "mean_token_accuracy": 0.8115965574979782,
      "num_tokens": 408856726.0,
      "step": 12810
    },
    {
      "entropy": 0.6186262127012014,
      "epoch": 1.1786763733388537,
      "grad_norm": 0.16148318350315094,
      "learning_rate": 6.071395712577054e-05,
      "loss": 0.6148,
      "mean_token_accuracy": 0.8094556629657745,
      "num_tokens": 408888046.0,
      "step": 12811
    },
    {
      "entropy": 0.5552530330605805,
      "epoch": 1.178768379710295,
      "grad_norm": 0.15198737382888794,
      "learning_rate": 6.071089029962892e-05,
      "loss": 0.5502,
      "mean_token_accuracy": 0.8334855102002621,
      "num_tokens": 408920566.0,
      "step": 12812
    },
    {
      "entropy": 0.49385589826852083,
      "epoch": 1.1788603860817362,
      "grad_norm": 0.15056167542934418,
      "learning_rate": 6.070782347348729e-05,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8463666774332523,
      "num_tokens": 408952739.0,
      "step": 12813
    },
    {
      "entropy": 0.5825871592387557,
      "epoch": 1.1789523924531773,
      "grad_norm": 0.1565801054239273,
      "learning_rate": 6.070475664734566e-05,
      "loss": 0.5766,
      "mean_token_accuracy": 0.8210620433092117,
      "num_tokens": 408984536.0,
      "step": 12814
    },
    {
      "entropy": 0.597351485863328,
      "epoch": 1.1790443988246186,
      "grad_norm": 0.15211522579193115,
      "learning_rate": 6.0701689821204045e-05,
      "loss": 0.5951,
      "mean_token_accuracy": 0.8166293688118458,
      "num_tokens": 409016840.0,
      "step": 12815
    },
    {
      "entropy": 0.5530285304412246,
      "epoch": 1.1791364051960598,
      "grad_norm": 0.15017487108707428,
      "learning_rate": 6.069862299506241e-05,
      "loss": 0.5278,
      "mean_token_accuracy": 0.8351617269217968,
      "num_tokens": 409048706.0,
      "step": 12816
    },
    {
      "entropy": 0.6663478035479784,
      "epoch": 1.1792284115675011,
      "grad_norm": 0.1654634177684784,
      "learning_rate": 6.069555616892079e-05,
      "loss": 0.6551,
      "mean_token_accuracy": 0.7962551712989807,
      "num_tokens": 409080301.0,
      "step": 12817
    },
    {
      "entropy": 0.6501547060906887,
      "epoch": 1.1793204179389423,
      "grad_norm": 0.16839751601219177,
      "learning_rate": 6.0692489342779155e-05,
      "loss": 0.6382,
      "mean_token_accuracy": 0.802005086094141,
      "num_tokens": 409110370.0,
      "step": 12818
    },
    {
      "entropy": 0.6803659070283175,
      "epoch": 1.1794124243103834,
      "grad_norm": 0.1635662168264389,
      "learning_rate": 6.0689422516637537e-05,
      "loss": 0.6742,
      "mean_token_accuracy": 0.7968026846647263,
      "num_tokens": 409142517.0,
      "step": 12819
    },
    {
      "entropy": 0.595958886668086,
      "epoch": 1.1795044306818248,
      "grad_norm": 0.15990608930587769,
      "learning_rate": 6.068635569049591e-05,
      "loss": 0.5993,
      "mean_token_accuracy": 0.8205725662410259,
      "num_tokens": 409174208.0,
      "step": 12820
    },
    {
      "entropy": 0.6585810147225857,
      "epoch": 1.179596437053266,
      "grad_norm": 0.15683068335056305,
      "learning_rate": 6.068328886435428e-05,
      "loss": 0.6385,
      "mean_token_accuracy": 0.8041860498487949,
      "num_tokens": 409206248.0,
      "step": 12821
    },
    {
      "entropy": 0.5681253466755152,
      "epoch": 1.1796884434247072,
      "grad_norm": 0.15520286560058594,
      "learning_rate": 6.0680222038212654e-05,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8302231058478355,
      "num_tokens": 409237770.0,
      "step": 12822
    },
    {
      "entropy": 0.5851716613397002,
      "epoch": 1.1797804497961484,
      "grad_norm": 0.16266874969005585,
      "learning_rate": 6.0677155212071035e-05,
      "loss": 0.559,
      "mean_token_accuracy": 0.829734243452549,
      "num_tokens": 409269284.0,
      "step": 12823
    },
    {
      "entropy": 0.5951076112687588,
      "epoch": 1.1798724561675895,
      "grad_norm": 0.1545606106519699,
      "learning_rate": 6.06740883859294e-05,
      "loss": 0.5787,
      "mean_token_accuracy": 0.8188728988170624,
      "num_tokens": 409301265.0,
      "step": 12824
    },
    {
      "entropy": 0.584199097007513,
      "epoch": 1.1799644625390309,
      "grad_norm": 0.1673383265733719,
      "learning_rate": 6.067102155978778e-05,
      "loss": 0.5627,
      "mean_token_accuracy": 0.824126411229372,
      "num_tokens": 409332534.0,
      "step": 12825
    },
    {
      "entropy": 0.6272962661460042,
      "epoch": 1.180056468910472,
      "grad_norm": 0.5330429077148438,
      "learning_rate": 6.0667954733646146e-05,
      "loss": 0.6107,
      "mean_token_accuracy": 0.8134483620524406,
      "num_tokens": 409364547.0,
      "step": 12826
    },
    {
      "entropy": 0.6910545136779547,
      "epoch": 1.1801484752819134,
      "grad_norm": 0.16435980796813965,
      "learning_rate": 6.0664887907504534e-05,
      "loss": 0.6675,
      "mean_token_accuracy": 0.7955349199473858,
      "num_tokens": 409396502.0,
      "step": 12827
    },
    {
      "entropy": 0.6493249610066414,
      "epoch": 1.1802404816533545,
      "grad_norm": 0.16249294579029083,
      "learning_rate": 6.06618210813629e-05,
      "loss": 0.6421,
      "mean_token_accuracy": 0.8033401966094971,
      "num_tokens": 409428620.0,
      "step": 12828
    },
    {
      "entropy": 0.5851630941033363,
      "epoch": 1.1803324880247956,
      "grad_norm": 0.16093024611473083,
      "learning_rate": 6.065875425522127e-05,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8238081894814968,
      "num_tokens": 409460173.0,
      "step": 12829
    },
    {
      "entropy": 0.586370088160038,
      "epoch": 1.180424494396237,
      "grad_norm": 0.173745796084404,
      "learning_rate": 6.0655687429079644e-05,
      "loss": 0.5713,
      "mean_token_accuracy": 0.8222965188324451,
      "num_tokens": 409492289.0,
      "step": 12830
    },
    {
      "entropy": 0.651757107116282,
      "epoch": 1.1805165007676781,
      "grad_norm": 0.1597006767988205,
      "learning_rate": 6.0652620602938026e-05,
      "loss": 0.6346,
      "mean_token_accuracy": 0.807503592222929,
      "num_tokens": 409524690.0,
      "step": 12831
    },
    {
      "entropy": 0.7561479639261961,
      "epoch": 1.1806085071391195,
      "grad_norm": 0.17377810180187225,
      "learning_rate": 6.06495537767964e-05,
      "loss": 0.7528,
      "mean_token_accuracy": 0.7723657861351967,
      "num_tokens": 409556481.0,
      "step": 12832
    },
    {
      "entropy": 0.6703077079728246,
      "epoch": 1.1807005135105606,
      "grad_norm": 0.16391624510288239,
      "learning_rate": 6.064648695065477e-05,
      "loss": 0.6618,
      "mean_token_accuracy": 0.7988701611757278,
      "num_tokens": 409588856.0,
      "step": 12833
    },
    {
      "entropy": 0.7681279890239239,
      "epoch": 1.1807925198820017,
      "grad_norm": 0.16730621457099915,
      "learning_rate": 6.064342012451314e-05,
      "loss": 0.7754,
      "mean_token_accuracy": 0.7650698311626911,
      "num_tokens": 409620813.0,
      "step": 12834
    },
    {
      "entropy": 0.5949154337868094,
      "epoch": 1.180884526253443,
      "grad_norm": 0.15510797500610352,
      "learning_rate": 6.0640353298371524e-05,
      "loss": 0.5938,
      "mean_token_accuracy": 0.8206617794930935,
      "num_tokens": 409653120.0,
      "step": 12835
    },
    {
      "entropy": 0.6225858014076948,
      "epoch": 1.1809765326248842,
      "grad_norm": 0.17176562547683716,
      "learning_rate": 6.063728647222989e-05,
      "loss": 0.6082,
      "mean_token_accuracy": 0.8125547543168068,
      "num_tokens": 409684909.0,
      "step": 12836
    },
    {
      "entropy": 0.5959023199975491,
      "epoch": 1.1810685389963256,
      "grad_norm": 0.15855368971824646,
      "learning_rate": 6.063421964608827e-05,
      "loss": 0.5804,
      "mean_token_accuracy": 0.8192933574318886,
      "num_tokens": 409716216.0,
      "step": 12837
    },
    {
      "entropy": 0.5967301726341248,
      "epoch": 1.1811605453677667,
      "grad_norm": 0.1518947035074234,
      "learning_rate": 6.0631152819946635e-05,
      "loss": 0.5823,
      "mean_token_accuracy": 0.8197972550988197,
      "num_tokens": 409748984.0,
      "step": 12838
    },
    {
      "entropy": 0.6170119373127818,
      "epoch": 1.1812525517392078,
      "grad_norm": 0.1576225757598877,
      "learning_rate": 6.0628085993805016e-05,
      "loss": 0.5819,
      "mean_token_accuracy": 0.8212177976965904,
      "num_tokens": 409780658.0,
      "step": 12839
    },
    {
      "entropy": 0.6137290522456169,
      "epoch": 1.1813445581106492,
      "grad_norm": 0.16177381575107574,
      "learning_rate": 6.062501916766339e-05,
      "loss": 0.5956,
      "mean_token_accuracy": 0.8147336095571518,
      "num_tokens": 409812440.0,
      "step": 12840
    },
    {
      "entropy": 0.5905677545815706,
      "epoch": 1.1814365644820903,
      "grad_norm": 0.16221065819263458,
      "learning_rate": 6.062195234152176e-05,
      "loss": 0.5878,
      "mean_token_accuracy": 0.8201001062989235,
      "num_tokens": 409844682.0,
      "step": 12841
    },
    {
      "entropy": 0.5305453427135944,
      "epoch": 1.1815285708535317,
      "grad_norm": 0.15989354252815247,
      "learning_rate": 6.061888551538013e-05,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8421405963599682,
      "num_tokens": 409876595.0,
      "step": 12842
    },
    {
      "entropy": 0.6618895437568426,
      "epoch": 1.1816205772249728,
      "grad_norm": 0.15954725444316864,
      "learning_rate": 6.0615818689238515e-05,
      "loss": 0.6595,
      "mean_token_accuracy": 0.8014923669397831,
      "num_tokens": 409908596.0,
      "step": 12843
    },
    {
      "entropy": 0.5636222250759602,
      "epoch": 1.181712583596414,
      "grad_norm": 0.1591787040233612,
      "learning_rate": 6.061275186309688e-05,
      "loss": 0.5548,
      "mean_token_accuracy": 0.8279785439372063,
      "num_tokens": 409940480.0,
      "step": 12844
    },
    {
      "entropy": 0.6284146150574088,
      "epoch": 1.1818045899678553,
      "grad_norm": 0.15905113518238068,
      "learning_rate": 6.060968503695526e-05,
      "loss": 0.6055,
      "mean_token_accuracy": 0.8118184767663479,
      "num_tokens": 409973102.0,
      "step": 12845
    },
    {
      "entropy": 0.6240399079397321,
      "epoch": 1.1818965963392964,
      "grad_norm": 0.16031309962272644,
      "learning_rate": 6.060661821081364e-05,
      "loss": 0.6269,
      "mean_token_accuracy": 0.8104398436844349,
      "num_tokens": 410005135.0,
      "step": 12846
    },
    {
      "entropy": 0.692963657900691,
      "epoch": 1.1819886027107378,
      "grad_norm": 0.16286158561706543,
      "learning_rate": 6.0603551384672006e-05,
      "loss": 0.6905,
      "mean_token_accuracy": 0.791682954877615,
      "num_tokens": 410037282.0,
      "step": 12847
    },
    {
      "entropy": 0.6297562716063112,
      "epoch": 1.182080609082179,
      "grad_norm": 0.16225384175777435,
      "learning_rate": 6.060048455853038e-05,
      "loss": 0.6173,
      "mean_token_accuracy": 0.8165185563266277,
      "num_tokens": 410069810.0,
      "step": 12848
    },
    {
      "entropy": 0.6574691254645586,
      "epoch": 1.18217261545362,
      "grad_norm": 0.15992605686187744,
      "learning_rate": 6.059741773238875e-05,
      "loss": 0.6475,
      "mean_token_accuracy": 0.8047504611313343,
      "num_tokens": 410101293.0,
      "step": 12849
    },
    {
      "entropy": 0.6484873332083225,
      "epoch": 1.1822646218250614,
      "grad_norm": 0.1608210653066635,
      "learning_rate": 6.059435090624713e-05,
      "loss": 0.6397,
      "mean_token_accuracy": 0.8085398897528648,
      "num_tokens": 410133379.0,
      "step": 12850
    },
    {
      "entropy": 0.6523200552910566,
      "epoch": 1.1823566281965026,
      "grad_norm": 0.15733548998832703,
      "learning_rate": 6.0591284080105505e-05,
      "loss": 0.6448,
      "mean_token_accuracy": 0.7999961413443089,
      "num_tokens": 410165367.0,
      "step": 12851
    },
    {
      "entropy": 0.6119341384619474,
      "epoch": 1.182448634567944,
      "grad_norm": 0.1558966338634491,
      "learning_rate": 6.058821725396387e-05,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8138749077916145,
      "num_tokens": 410197125.0,
      "step": 12852
    },
    {
      "entropy": 0.668057007715106,
      "epoch": 1.182540640939385,
      "grad_norm": 0.1627742350101471,
      "learning_rate": 6.058515042782225e-05,
      "loss": 0.6543,
      "mean_token_accuracy": 0.8024899363517761,
      "num_tokens": 410228970.0,
      "step": 12853
    },
    {
      "entropy": 0.6217446802183986,
      "epoch": 1.1826326473108262,
      "grad_norm": 0.15942099690437317,
      "learning_rate": 6.058208360168063e-05,
      "loss": 0.6128,
      "mean_token_accuracy": 0.8124975971877575,
      "num_tokens": 410261721.0,
      "step": 12854
    },
    {
      "entropy": 0.5217242427170277,
      "epoch": 1.1827246536822675,
      "grad_norm": 0.14671209454536438,
      "learning_rate": 6.0579016775539e-05,
      "loss": 0.5066,
      "mean_token_accuracy": 0.8398961536586285,
      "num_tokens": 410293778.0,
      "step": 12855
    },
    {
      "entropy": 0.5325000714510679,
      "epoch": 1.1828166600537087,
      "grad_norm": 0.14626295864582062,
      "learning_rate": 6.057594994939737e-05,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8411106616258621,
      "num_tokens": 410325886.0,
      "step": 12856
    },
    {
      "entropy": 0.6917587053030729,
      "epoch": 1.18290866642515,
      "grad_norm": 0.17106474936008453,
      "learning_rate": 6.057288312325574e-05,
      "loss": 0.6793,
      "mean_token_accuracy": 0.7928517796099186,
      "num_tokens": 410358196.0,
      "step": 12857
    },
    {
      "entropy": 0.6211818046867847,
      "epoch": 1.1830006727965912,
      "grad_norm": 0.16705621778964996,
      "learning_rate": 6.056981629711412e-05,
      "loss": 0.6118,
      "mean_token_accuracy": 0.8095204122364521,
      "num_tokens": 410389947.0,
      "step": 12858
    },
    {
      "entropy": 0.632463064044714,
      "epoch": 1.1830926791680323,
      "grad_norm": 0.15922266244888306,
      "learning_rate": 6.0566749470972495e-05,
      "loss": 0.6383,
      "mean_token_accuracy": 0.805241946130991,
      "num_tokens": 410421658.0,
      "step": 12859
    },
    {
      "entropy": 0.5693967835977674,
      "epoch": 1.1831846855394736,
      "grad_norm": 0.15652354061603546,
      "learning_rate": 6.056368264483086e-05,
      "loss": 0.553,
      "mean_token_accuracy": 0.8316059969365597,
      "num_tokens": 410453110.0,
      "step": 12860
    },
    {
      "entropy": 0.6711183656007051,
      "epoch": 1.1832766919109148,
      "grad_norm": 0.16308577358722687,
      "learning_rate": 6.056061581868924e-05,
      "loss": 0.6678,
      "mean_token_accuracy": 0.8025546073913574,
      "num_tokens": 410485087.0,
      "step": 12861
    },
    {
      "entropy": 0.5163745004683733,
      "epoch": 1.1833686982823561,
      "grad_norm": 0.14833982288837433,
      "learning_rate": 6.055754899254762e-05,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8495572544634342,
      "num_tokens": 410517494.0,
      "step": 12862
    },
    {
      "entropy": 0.5666032275184989,
      "epoch": 1.1834607046537973,
      "grad_norm": 0.15680193901062012,
      "learning_rate": 6.055448216640599e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8290773667395115,
      "num_tokens": 410549603.0,
      "step": 12863
    },
    {
      "entropy": 0.6440720614045858,
      "epoch": 1.1835527110252384,
      "grad_norm": 0.15361739695072174,
      "learning_rate": 6.055141534026436e-05,
      "loss": 0.627,
      "mean_token_accuracy": 0.8083376027643681,
      "num_tokens": 410581813.0,
      "step": 12864
    },
    {
      "entropy": 0.7263589613139629,
      "epoch": 1.1836447173966798,
      "grad_norm": 0.16079625487327576,
      "learning_rate": 6.054834851412273e-05,
      "loss": 0.7167,
      "mean_token_accuracy": 0.780796654522419,
      "num_tokens": 410613857.0,
      "step": 12865
    },
    {
      "entropy": 0.6702664494514465,
      "epoch": 1.183736723768121,
      "grad_norm": 0.16401860117912292,
      "learning_rate": 6.054528168798112e-05,
      "loss": 0.652,
      "mean_token_accuracy": 0.8004784658551216,
      "num_tokens": 410646062.0,
      "step": 12866
    },
    {
      "entropy": 0.6296720616519451,
      "epoch": 1.1838287301395622,
      "grad_norm": 0.16353578865528107,
      "learning_rate": 6.0542214861839486e-05,
      "loss": 0.6307,
      "mean_token_accuracy": 0.8045900650322437,
      "num_tokens": 410677903.0,
      "step": 12867
    },
    {
      "entropy": 0.5165645405650139,
      "epoch": 1.1839207365110034,
      "grad_norm": 0.15600460767745972,
      "learning_rate": 6.0539148035697854e-05,
      "loss": 0.5076,
      "mean_token_accuracy": 0.840537529438734,
      "num_tokens": 410709359.0,
      "step": 12868
    },
    {
      "entropy": 0.6606485582888126,
      "epoch": 1.1840127428824445,
      "grad_norm": 0.16115926206111908,
      "learning_rate": 6.053608120955623e-05,
      "loss": 0.6601,
      "mean_token_accuracy": 0.7964460477232933,
      "num_tokens": 410741965.0,
      "step": 12869
    },
    {
      "entropy": 0.5955555159598589,
      "epoch": 1.1841047492538859,
      "grad_norm": 0.1585136204957962,
      "learning_rate": 6.053301438341461e-05,
      "loss": 0.5734,
      "mean_token_accuracy": 0.8212171904742718,
      "num_tokens": 410773647.0,
      "step": 12870
    },
    {
      "entropy": 0.6964815147221088,
      "epoch": 1.184196755625327,
      "grad_norm": 0.16256742179393768,
      "learning_rate": 6.0529947557272984e-05,
      "loss": 0.683,
      "mean_token_accuracy": 0.7917600646615028,
      "num_tokens": 410805533.0,
      "step": 12871
    },
    {
      "entropy": 0.6459903568029404,
      "epoch": 1.1842887619967684,
      "grad_norm": 0.15680153667926788,
      "learning_rate": 6.052688073113135e-05,
      "loss": 0.6289,
      "mean_token_accuracy": 0.8103121109306812,
      "num_tokens": 410836554.0,
      "step": 12872
    },
    {
      "entropy": 0.6041005849838257,
      "epoch": 1.1843807683682095,
      "grad_norm": 0.15702031552791595,
      "learning_rate": 6.0523813904989734e-05,
      "loss": 0.5823,
      "mean_token_accuracy": 0.8161187246441841,
      "num_tokens": 410867934.0,
      "step": 12873
    },
    {
      "entropy": 0.565071483142674,
      "epoch": 1.1844727747396506,
      "grad_norm": 0.1616765707731247,
      "learning_rate": 6.052074707884811e-05,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8305428549647331,
      "num_tokens": 410899689.0,
      "step": 12874
    },
    {
      "entropy": 0.7267187163233757,
      "epoch": 1.184564781111092,
      "grad_norm": 0.1724214255809784,
      "learning_rate": 6.0517680252706476e-05,
      "loss": 0.7079,
      "mean_token_accuracy": 0.7849208824336529,
      "num_tokens": 410930729.0,
      "step": 12875
    },
    {
      "entropy": 0.6075108367949724,
      "epoch": 1.1846567874825331,
      "grad_norm": 0.1635294109582901,
      "learning_rate": 6.051461342656485e-05,
      "loss": 0.6006,
      "mean_token_accuracy": 0.8122540712356567,
      "num_tokens": 410962011.0,
      "step": 12876
    },
    {
      "entropy": 0.6876761121675372,
      "epoch": 1.1847487938539745,
      "grad_norm": 0.16272957623004913,
      "learning_rate": 6.051154660042323e-05,
      "loss": 0.6719,
      "mean_token_accuracy": 0.796604111790657,
      "num_tokens": 410993577.0,
      "step": 12877
    },
    {
      "entropy": 0.6693268194794655,
      "epoch": 1.1848408002254156,
      "grad_norm": 0.16332793235778809,
      "learning_rate": 6.05084797742816e-05,
      "loss": 0.6638,
      "mean_token_accuracy": 0.7996008507907391,
      "num_tokens": 411025515.0,
      "step": 12878
    },
    {
      "entropy": 0.5904426574707031,
      "epoch": 1.1849328065968567,
      "grad_norm": 0.15868507325649261,
      "learning_rate": 6.0505412948139975e-05,
      "loss": 0.5804,
      "mean_token_accuracy": 0.8229742012917995,
      "num_tokens": 411057894.0,
      "step": 12879
    },
    {
      "entropy": 0.6249990556389093,
      "epoch": 1.185024812968298,
      "grad_norm": 0.1704150140285492,
      "learning_rate": 6.050234612199834e-05,
      "loss": 0.6314,
      "mean_token_accuracy": 0.808990228921175,
      "num_tokens": 411088683.0,
      "step": 12880
    },
    {
      "entropy": 0.6249595563858747,
      "epoch": 1.1851168193397392,
      "grad_norm": 0.15521781146526337,
      "learning_rate": 6.0499279295856724e-05,
      "loss": 0.6063,
      "mean_token_accuracy": 0.8146154135465622,
      "num_tokens": 411121422.0,
      "step": 12881
    },
    {
      "entropy": 0.637336503714323,
      "epoch": 1.1852088257111806,
      "grad_norm": 0.15958553552627563,
      "learning_rate": 6.04962124697151e-05,
      "loss": 0.6275,
      "mean_token_accuracy": 0.8068297505378723,
      "num_tokens": 411153028.0,
      "step": 12882
    },
    {
      "entropy": 0.6513865813612938,
      "epoch": 1.1853008320826217,
      "grad_norm": 0.1579613834619522,
      "learning_rate": 6.0493145643573467e-05,
      "loss": 0.6412,
      "mean_token_accuracy": 0.8037446066737175,
      "num_tokens": 411185438.0,
      "step": 12883
    },
    {
      "entropy": 0.763935498893261,
      "epoch": 1.1853928384540628,
      "grad_norm": 0.1638147383928299,
      "learning_rate": 6.049007881743184e-05,
      "loss": 0.7571,
      "mean_token_accuracy": 0.7690280377864838,
      "num_tokens": 411217637.0,
      "step": 12884
    },
    {
      "entropy": 0.5395246469415724,
      "epoch": 1.1854848448255042,
      "grad_norm": 0.1650964468717575,
      "learning_rate": 6.048701199129022e-05,
      "loss": 0.5117,
      "mean_token_accuracy": 0.8423259668052197,
      "num_tokens": 411250060.0,
      "step": 12885
    },
    {
      "entropy": 0.6249136356636882,
      "epoch": 1.1855768511969453,
      "grad_norm": 0.15146276354789734,
      "learning_rate": 6.048394516514859e-05,
      "loss": 0.6057,
      "mean_token_accuracy": 0.8123128600418568,
      "num_tokens": 411282384.0,
      "step": 12886
    },
    {
      "entropy": 0.6015826789662242,
      "epoch": 1.1856688575683867,
      "grad_norm": 0.1524311751127243,
      "learning_rate": 6.0480878339006965e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8154684491455555,
      "num_tokens": 411314800.0,
      "step": 12887
    },
    {
      "entropy": 0.6336343679577112,
      "epoch": 1.1857608639398278,
      "grad_norm": 0.15778134763240814,
      "learning_rate": 6.047781151286533e-05,
      "loss": 0.6216,
      "mean_token_accuracy": 0.8124285191297531,
      "num_tokens": 411347498.0,
      "step": 12888
    },
    {
      "entropy": 0.656800439581275,
      "epoch": 1.185852870311269,
      "grad_norm": 0.17137038707733154,
      "learning_rate": 6.0474744686723714e-05,
      "loss": 0.6398,
      "mean_token_accuracy": 0.8018249236047268,
      "num_tokens": 411378589.0,
      "step": 12889
    },
    {
      "entropy": 0.5184663943946362,
      "epoch": 1.1859448766827103,
      "grad_norm": 0.15502318739891052,
      "learning_rate": 6.047167786058209e-05,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8438099250197411,
      "num_tokens": 411410861.0,
      "step": 12890
    },
    {
      "entropy": 0.57224484346807,
      "epoch": 1.1860368830541514,
      "grad_norm": 0.1550436168909073,
      "learning_rate": 6.046861103444046e-05,
      "loss": 0.5621,
      "mean_token_accuracy": 0.823591735213995,
      "num_tokens": 411441985.0,
      "step": 12891
    },
    {
      "entropy": 0.5151236054953188,
      "epoch": 1.1861288894255928,
      "grad_norm": 0.1508413702249527,
      "learning_rate": 6.046554420829883e-05,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8448976539075375,
      "num_tokens": 411474241.0,
      "step": 12892
    },
    {
      "entropy": 0.6307276543229818,
      "epoch": 1.186220895797034,
      "grad_norm": 0.15391504764556885,
      "learning_rate": 6.046247738215721e-05,
      "loss": 0.6233,
      "mean_token_accuracy": 0.8089801073074341,
      "num_tokens": 411506545.0,
      "step": 12893
    },
    {
      "entropy": 0.49785149935632944,
      "epoch": 1.186312902168475,
      "grad_norm": 0.14616230130195618,
      "learning_rate": 6.045941055601558e-05,
      "loss": 0.4871,
      "mean_token_accuracy": 0.8483933508396149,
      "num_tokens": 411538877.0,
      "step": 12894
    },
    {
      "entropy": 0.5104107651859522,
      "epoch": 1.1864049085399164,
      "grad_norm": 0.15420880913734436,
      "learning_rate": 6.0456343729873956e-05,
      "loss": 0.5012,
      "mean_token_accuracy": 0.8421576805412769,
      "num_tokens": 411570987.0,
      "step": 12895
    },
    {
      "entropy": 0.7047312166541815,
      "epoch": 1.1864969149113576,
      "grad_norm": 0.1630563586950302,
      "learning_rate": 6.0453276903732323e-05,
      "loss": 0.7044,
      "mean_token_accuracy": 0.7878213562071323,
      "num_tokens": 411602622.0,
      "step": 12896
    },
    {
      "entropy": 0.5895881745964289,
      "epoch": 1.186588921282799,
      "grad_norm": 0.15224012732505798,
      "learning_rate": 6.0450210077590705e-05,
      "loss": 0.5677,
      "mean_token_accuracy": 0.8246490731835365,
      "num_tokens": 411634863.0,
      "step": 12897
    },
    {
      "entropy": 0.5681608496233821,
      "epoch": 1.18668092765424,
      "grad_norm": 0.15649960935115814,
      "learning_rate": 6.044714325144908e-05,
      "loss": 0.564,
      "mean_token_accuracy": 0.8295067325234413,
      "num_tokens": 411667097.0,
      "step": 12898
    },
    {
      "entropy": 0.4878831682726741,
      "epoch": 1.1867729340256812,
      "grad_norm": 0.15333670377731323,
      "learning_rate": 6.044407642530745e-05,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8531825803220272,
      "num_tokens": 411699794.0,
      "step": 12899
    },
    {
      "entropy": 0.6228166781365871,
      "epoch": 1.1868649403971225,
      "grad_norm": 0.15943261981010437,
      "learning_rate": 6.044100959916582e-05,
      "loss": 0.6308,
      "mean_token_accuracy": 0.8057497106492519,
      "num_tokens": 411731879.0,
      "step": 12900
    },
    {
      "entropy": 0.5792943835258484,
      "epoch": 1.1869569467685637,
      "grad_norm": 0.1730756163597107,
      "learning_rate": 6.0437942773024203e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8226151540875435,
      "num_tokens": 411763009.0,
      "step": 12901
    },
    {
      "entropy": 0.5071085756644607,
      "epoch": 1.187048953140005,
      "grad_norm": 0.1501636952161789,
      "learning_rate": 6.043487594688257e-05,
      "loss": 0.4948,
      "mean_token_accuracy": 0.8451801724731922,
      "num_tokens": 411794866.0,
      "step": 12902
    },
    {
      "entropy": 0.605363305658102,
      "epoch": 1.1871409595114462,
      "grad_norm": 0.15578654408454895,
      "learning_rate": 6.0431809120740946e-05,
      "loss": 0.5982,
      "mean_token_accuracy": 0.8189945891499519,
      "num_tokens": 411827051.0,
      "step": 12903
    },
    {
      "entropy": 0.6800463758409023,
      "epoch": 1.1872329658828873,
      "grad_norm": 0.16238172352313995,
      "learning_rate": 6.042874229459933e-05,
      "loss": 0.6742,
      "mean_token_accuracy": 0.792614821344614,
      "num_tokens": 411858884.0,
      "step": 12904
    },
    {
      "entropy": 0.6642596386373043,
      "epoch": 1.1873249722543286,
      "grad_norm": 0.1641029417514801,
      "learning_rate": 6.0425675468457695e-05,
      "loss": 0.6329,
      "mean_token_accuracy": 0.8061506077647209,
      "num_tokens": 411890849.0,
      "step": 12905
    },
    {
      "entropy": 0.6642412804067135,
      "epoch": 1.1874169786257698,
      "grad_norm": 0.16158293187618256,
      "learning_rate": 6.042260864231607e-05,
      "loss": 0.6577,
      "mean_token_accuracy": 0.7977614924311638,
      "num_tokens": 411922118.0,
      "step": 12906
    },
    {
      "entropy": 0.7027017809450626,
      "epoch": 1.1875089849972111,
      "grad_norm": 0.16589194536209106,
      "learning_rate": 6.041954181617444e-05,
      "loss": 0.688,
      "mean_token_accuracy": 0.7875435464084148,
      "num_tokens": 411953674.0,
      "step": 12907
    },
    {
      "entropy": 0.7225943394005299,
      "epoch": 1.1876009913686523,
      "grad_norm": 0.15984661877155304,
      "learning_rate": 6.0416474990032826e-05,
      "loss": 0.714,
      "mean_token_accuracy": 0.777837660163641,
      "num_tokens": 411985646.0,
      "step": 12908
    },
    {
      "entropy": 0.6836789008229971,
      "epoch": 1.1876929977400934,
      "grad_norm": 0.16519048810005188,
      "learning_rate": 6.0413408163891194e-05,
      "loss": 0.6816,
      "mean_token_accuracy": 0.7985980473458767,
      "num_tokens": 412017163.0,
      "step": 12909
    },
    {
      "entropy": 0.6219694651663303,
      "epoch": 1.1877850041115348,
      "grad_norm": 0.15134617686271667,
      "learning_rate": 6.041034133774957e-05,
      "loss": 0.6115,
      "mean_token_accuracy": 0.8125102818012238,
      "num_tokens": 412048951.0,
      "step": 12910
    },
    {
      "entropy": 0.5172385200858116,
      "epoch": 1.187877010482976,
      "grad_norm": 0.15703244507312775,
      "learning_rate": 6.0407274511607936e-05,
      "loss": 0.4982,
      "mean_token_accuracy": 0.8470037803053856,
      "num_tokens": 412081263.0,
      "step": 12911
    },
    {
      "entropy": 0.5755518544465303,
      "epoch": 1.1879690168544172,
      "grad_norm": 0.15334543585777283,
      "learning_rate": 6.040420768546632e-05,
      "loss": 0.5418,
      "mean_token_accuracy": 0.8329243287444115,
      "num_tokens": 412113554.0,
      "step": 12912
    },
    {
      "entropy": 0.5638619577512145,
      "epoch": 1.1880610232258584,
      "grad_norm": 0.15300965309143066,
      "learning_rate": 6.040114085932469e-05,
      "loss": 0.5426,
      "mean_token_accuracy": 0.8337432183325291,
      "num_tokens": 412146164.0,
      "step": 12913
    },
    {
      "entropy": 0.43261874839663506,
      "epoch": 1.1881530295972995,
      "grad_norm": 0.14375944435596466,
      "learning_rate": 6.039807403318306e-05,
      "loss": 0.4098,
      "mean_token_accuracy": 0.8715882264077663,
      "num_tokens": 412178700.0,
      "step": 12914
    },
    {
      "entropy": 0.7158865150995553,
      "epoch": 1.1882450359687409,
      "grad_norm": 0.15856191515922546,
      "learning_rate": 6.0395007207041435e-05,
      "loss": 0.6962,
      "mean_token_accuracy": 0.7887215055525303,
      "num_tokens": 412210773.0,
      "step": 12915
    },
    {
      "entropy": 0.574910094961524,
      "epoch": 1.188337042340182,
      "grad_norm": 0.15066885948181152,
      "learning_rate": 6.0391940380899816e-05,
      "loss": 0.5638,
      "mean_token_accuracy": 0.825688362121582,
      "num_tokens": 412242769.0,
      "step": 12916
    },
    {
      "entropy": 0.6499043228104711,
      "epoch": 1.1884290487116234,
      "grad_norm": 0.15248294174671173,
      "learning_rate": 6.0388873554758184e-05,
      "loss": 0.6215,
      "mean_token_accuracy": 0.8051152005791664,
      "num_tokens": 412274541.0,
      "step": 12917
    },
    {
      "entropy": 0.529246206395328,
      "epoch": 1.1885210550830645,
      "grad_norm": 0.14718180894851685,
      "learning_rate": 6.038580672861656e-05,
      "loss": 0.517,
      "mean_token_accuracy": 0.8402014710009098,
      "num_tokens": 412306339.0,
      "step": 12918
    },
    {
      "entropy": 0.5973511505872011,
      "epoch": 1.1886130614545056,
      "grad_norm": 0.15314503014087677,
      "learning_rate": 6.038273990247493e-05,
      "loss": 0.5849,
      "mean_token_accuracy": 0.8170783035457134,
      "num_tokens": 412338577.0,
      "step": 12919
    },
    {
      "entropy": 0.6615813542157412,
      "epoch": 1.188705067825947,
      "grad_norm": 0.16858428716659546,
      "learning_rate": 6.037967307633331e-05,
      "loss": 0.6643,
      "mean_token_accuracy": 0.8000758141279221,
      "num_tokens": 412371032.0,
      "step": 12920
    },
    {
      "entropy": 0.5204954547807574,
      "epoch": 1.1887970741973881,
      "grad_norm": 0.15875308215618134,
      "learning_rate": 6.037660625019168e-05,
      "loss": 0.5109,
      "mean_token_accuracy": 0.8478487357497215,
      "num_tokens": 412403362.0,
      "step": 12921
    },
    {
      "entropy": 0.7020768877118826,
      "epoch": 1.1888890805688295,
      "grad_norm": 0.16914820671081543,
      "learning_rate": 6.037353942405005e-05,
      "loss": 0.7051,
      "mean_token_accuracy": 0.7854323647916317,
      "num_tokens": 412435880.0,
      "step": 12922
    },
    {
      "entropy": 0.585023395717144,
      "epoch": 1.1889810869402706,
      "grad_norm": 0.15374670922756195,
      "learning_rate": 6.0370472597908425e-05,
      "loss": 0.5803,
      "mean_token_accuracy": 0.8225524872541428,
      "num_tokens": 412468057.0,
      "step": 12923
    },
    {
      "entropy": 0.5923198889940977,
      "epoch": 1.1890730933117117,
      "grad_norm": 0.16419124603271484,
      "learning_rate": 6.036740577176681e-05,
      "loss": 0.5808,
      "mean_token_accuracy": 0.823333777487278,
      "num_tokens": 412499873.0,
      "step": 12924
    },
    {
      "entropy": 0.5354879545047879,
      "epoch": 1.189165099683153,
      "grad_norm": 0.1537798047065735,
      "learning_rate": 6.0364338945625175e-05,
      "loss": 0.5235,
      "mean_token_accuracy": 0.8346861153841019,
      "num_tokens": 412531700.0,
      "step": 12925
    },
    {
      "entropy": 0.599360466003418,
      "epoch": 1.1892571060545942,
      "grad_norm": 0.1624981015920639,
      "learning_rate": 6.036127211948355e-05,
      "loss": 0.5901,
      "mean_token_accuracy": 0.8133135288953781,
      "num_tokens": 412563212.0,
      "step": 12926
    },
    {
      "entropy": 0.6972933802753687,
      "epoch": 1.1893491124260356,
      "grad_norm": 0.1620873510837555,
      "learning_rate": 6.035820529334192e-05,
      "loss": 0.6938,
      "mean_token_accuracy": 0.7911777086555958,
      "num_tokens": 412595665.0,
      "step": 12927
    },
    {
      "entropy": 0.6904404703527689,
      "epoch": 1.1894411187974767,
      "grad_norm": 0.16163800656795502,
      "learning_rate": 6.03551384672003e-05,
      "loss": 0.6779,
      "mean_token_accuracy": 0.7978106290102005,
      "num_tokens": 412628013.0,
      "step": 12928
    },
    {
      "entropy": 0.5955011984333396,
      "epoch": 1.1895331251689178,
      "grad_norm": 0.1536337286233902,
      "learning_rate": 6.035207164105867e-05,
      "loss": 0.5732,
      "mean_token_accuracy": 0.8185814246535301,
      "num_tokens": 412659795.0,
      "step": 12929
    },
    {
      "entropy": 0.6482262620702386,
      "epoch": 1.1896251315403592,
      "grad_norm": 0.15930841863155365,
      "learning_rate": 6.034900481491704e-05,
      "loss": 0.6413,
      "mean_token_accuracy": 0.8056476823985577,
      "num_tokens": 412692114.0,
      "step": 12930
    },
    {
      "entropy": 0.5667450567707419,
      "epoch": 1.1897171379118003,
      "grad_norm": 0.1492510288953781,
      "learning_rate": 6.0345937988775416e-05,
      "loss": 0.5496,
      "mean_token_accuracy": 0.8315381407737732,
      "num_tokens": 412724637.0,
      "step": 12931
    },
    {
      "entropy": 0.5658194813877344,
      "epoch": 1.1898091442832417,
      "grad_norm": 0.16022823750972748,
      "learning_rate": 6.03428711626338e-05,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8321414180099964,
      "num_tokens": 412755185.0,
      "step": 12932
    },
    {
      "entropy": 0.7076150085777044,
      "epoch": 1.1899011506546828,
      "grad_norm": 0.16214579343795776,
      "learning_rate": 6.0339804336492165e-05,
      "loss": 0.6902,
      "mean_token_accuracy": 0.7908612452447414,
      "num_tokens": 412787316.0,
      "step": 12933
    },
    {
      "entropy": 0.5672255093231797,
      "epoch": 1.189993157026124,
      "grad_norm": 0.15646225214004517,
      "learning_rate": 6.033673751035054e-05,
      "loss": 0.5393,
      "mean_token_accuracy": 0.8313247710466385,
      "num_tokens": 412819320.0,
      "step": 12934
    },
    {
      "entropy": 0.6194953545928001,
      "epoch": 1.1900851633975653,
      "grad_norm": 0.14914673566818237,
      "learning_rate": 6.033367068420892e-05,
      "loss": 0.6006,
      "mean_token_accuracy": 0.8151112124323845,
      "num_tokens": 412852079.0,
      "step": 12935
    },
    {
      "entropy": 0.554384870454669,
      "epoch": 1.1901771697690064,
      "grad_norm": 0.15027625858783722,
      "learning_rate": 6.033060385806729e-05,
      "loss": 0.5401,
      "mean_token_accuracy": 0.8311931975185871,
      "num_tokens": 412884673.0,
      "step": 12936
    },
    {
      "entropy": 0.6140545941889286,
      "epoch": 1.1902691761404478,
      "grad_norm": 0.15064236521720886,
      "learning_rate": 6.0327537031925664e-05,
      "loss": 0.6126,
      "mean_token_accuracy": 0.8163776211440563,
      "num_tokens": 412917441.0,
      "step": 12937
    },
    {
      "entropy": 0.6443721447139978,
      "epoch": 1.190361182511889,
      "grad_norm": 0.16096869111061096,
      "learning_rate": 6.032447020578403e-05,
      "loss": 0.6306,
      "mean_token_accuracy": 0.8117424994707108,
      "num_tokens": 412949587.0,
      "step": 12938
    },
    {
      "entropy": 0.6607138067483902,
      "epoch": 1.19045318888333,
      "grad_norm": 0.15992142260074615,
      "learning_rate": 6.032140337964241e-05,
      "loss": 0.654,
      "mean_token_accuracy": 0.7976130805909634,
      "num_tokens": 412981056.0,
      "step": 12939
    },
    {
      "entropy": 0.6168176387436688,
      "epoch": 1.1905451952547714,
      "grad_norm": 0.16099253296852112,
      "learning_rate": 6.031833655350079e-05,
      "loss": 0.6145,
      "mean_token_accuracy": 0.814532645046711,
      "num_tokens": 413013257.0,
      "step": 12940
    },
    {
      "entropy": 0.652438866905868,
      "epoch": 1.1906372016262126,
      "grad_norm": 0.1594868004322052,
      "learning_rate": 6.0315269727359156e-05,
      "loss": 0.6484,
      "mean_token_accuracy": 0.8016856946051121,
      "num_tokens": 413045466.0,
      "step": 12941
    },
    {
      "entropy": 0.6601704070344567,
      "epoch": 1.190729207997654,
      "grad_norm": 0.1644580066204071,
      "learning_rate": 6.031220290121753e-05,
      "loss": 0.6521,
      "mean_token_accuracy": 0.8019727505743504,
      "num_tokens": 413077630.0,
      "step": 12942
    },
    {
      "entropy": 0.46943290857598186,
      "epoch": 1.190821214369095,
      "grad_norm": 0.15409138798713684,
      "learning_rate": 6.030913607507591e-05,
      "loss": 0.4603,
      "mean_token_accuracy": 0.8595850579440594,
      "num_tokens": 413109952.0,
      "step": 12943
    },
    {
      "entropy": 0.6207207459956408,
      "epoch": 1.1909132207405362,
      "grad_norm": 0.15547607839107513,
      "learning_rate": 6.030606924893428e-05,
      "loss": 0.6104,
      "mean_token_accuracy": 0.8116990178823471,
      "num_tokens": 413142329.0,
      "step": 12944
    },
    {
      "entropy": 0.6771057089790702,
      "epoch": 1.1910052271119775,
      "grad_norm": 0.1608951836824417,
      "learning_rate": 6.0303002422792654e-05,
      "loss": 0.6758,
      "mean_token_accuracy": 0.794353611767292,
      "num_tokens": 413174712.0,
      "step": 12945
    },
    {
      "entropy": 0.6286241225898266,
      "epoch": 1.1910972334834187,
      "grad_norm": 0.15560559928417206,
      "learning_rate": 6.029993559665102e-05,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8155966363847256,
      "num_tokens": 413206802.0,
      "step": 12946
    },
    {
      "entropy": 0.6505820993334055,
      "epoch": 1.19118923985486,
      "grad_norm": 0.15819257497787476,
      "learning_rate": 6.029686877050941e-05,
      "loss": 0.6438,
      "mean_token_accuracy": 0.8049435093998909,
      "num_tokens": 413238659.0,
      "step": 12947
    },
    {
      "entropy": 0.6965616121888161,
      "epoch": 1.1912812462263012,
      "grad_norm": 0.16269536316394806,
      "learning_rate": 6.029380194436778e-05,
      "loss": 0.6991,
      "mean_token_accuracy": 0.7852487191557884,
      "num_tokens": 413270944.0,
      "step": 12948
    },
    {
      "entropy": 0.7449693977832794,
      "epoch": 1.1913732525977423,
      "grad_norm": 0.17631401121616364,
      "learning_rate": 6.0290735118226146e-05,
      "loss": 0.7466,
      "mean_token_accuracy": 0.7783134765923023,
      "num_tokens": 413302901.0,
      "step": 12949
    },
    {
      "entropy": 0.5778220500797033,
      "epoch": 1.1914652589691836,
      "grad_norm": 0.15685111284255981,
      "learning_rate": 6.028766829208452e-05,
      "loss": 0.5706,
      "mean_token_accuracy": 0.82271683588624,
      "num_tokens": 413335080.0,
      "step": 12950
    },
    {
      "entropy": 0.6429454232566059,
      "epoch": 1.1915572653406248,
      "grad_norm": 0.15303926169872284,
      "learning_rate": 6.02846014659429e-05,
      "loss": 0.6286,
      "mean_token_accuracy": 0.8093688488006592,
      "num_tokens": 413367056.0,
      "step": 12951
    },
    {
      "entropy": 0.6288506258279085,
      "epoch": 1.1916492717120661,
      "grad_norm": 0.17288829386234283,
      "learning_rate": 6.028153463980128e-05,
      "loss": 0.6229,
      "mean_token_accuracy": 0.8067662194371223,
      "num_tokens": 413397795.0,
      "step": 12952
    },
    {
      "entropy": 0.5449880911037326,
      "epoch": 1.1917412780835073,
      "grad_norm": 0.14552177488803864,
      "learning_rate": 6.0278467813659645e-05,
      "loss": 0.5312,
      "mean_token_accuracy": 0.8336432352662086,
      "num_tokens": 413430063.0,
      "step": 12953
    },
    {
      "entropy": 0.5225226059556007,
      "epoch": 1.1918332844549484,
      "grad_norm": 0.1491377353668213,
      "learning_rate": 6.027540098751802e-05,
      "loss": 0.5105,
      "mean_token_accuracy": 0.842152439057827,
      "num_tokens": 413462011.0,
      "step": 12954
    },
    {
      "entropy": 0.6508317068219185,
      "epoch": 1.1919252908263898,
      "grad_norm": 0.15631957352161407,
      "learning_rate": 6.02723341613764e-05,
      "loss": 0.6286,
      "mean_token_accuracy": 0.8078235238790512,
      "num_tokens": 413494055.0,
      "step": 12955
    },
    {
      "entropy": 0.5017553996294737,
      "epoch": 1.192017297197831,
      "grad_norm": 0.15128152072429657,
      "learning_rate": 6.026926733523477e-05,
      "loss": 0.4914,
      "mean_token_accuracy": 0.8508582040667534,
      "num_tokens": 413526061.0,
      "step": 12956
    },
    {
      "entropy": 0.6323916986584663,
      "epoch": 1.1921093035692722,
      "grad_norm": 0.15381073951721191,
      "learning_rate": 6.026620050909314e-05,
      "loss": 0.618,
      "mean_token_accuracy": 0.8141316287219524,
      "num_tokens": 413558185.0,
      "step": 12957
    },
    {
      "entropy": 0.7761156475171447,
      "epoch": 1.1922013099407134,
      "grad_norm": 0.18158479034900665,
      "learning_rate": 6.026313368295151e-05,
      "loss": 0.7665,
      "mean_token_accuracy": 0.7713797762989998,
      "num_tokens": 413589469.0,
      "step": 12958
    },
    {
      "entropy": 0.5305654630064964,
      "epoch": 1.1922933163121545,
      "grad_norm": 0.1479751020669937,
      "learning_rate": 6.026006685680989e-05,
      "loss": 0.518,
      "mean_token_accuracy": 0.8370233215391636,
      "num_tokens": 413620930.0,
      "step": 12959
    },
    {
      "entropy": 0.5754859037697315,
      "epoch": 1.1923853226835959,
      "grad_norm": 0.1643238216638565,
      "learning_rate": 6.025700003066827e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8246095813810825,
      "num_tokens": 413652116.0,
      "step": 12960
    },
    {
      "entropy": 0.6407639570534229,
      "epoch": 1.192477329055037,
      "grad_norm": 0.1532324105501175,
      "learning_rate": 6.0253933204526635e-05,
      "loss": 0.6321,
      "mean_token_accuracy": 0.8065248057246208,
      "num_tokens": 413684793.0,
      "step": 12961
    },
    {
      "entropy": 0.7331072986125946,
      "epoch": 1.1925693354264784,
      "grad_norm": 0.16691647469997406,
      "learning_rate": 6.025086637838501e-05,
      "loss": 0.7173,
      "mean_token_accuracy": 0.7824029810726643,
      "num_tokens": 413716677.0,
      "step": 12962
    },
    {
      "entropy": 0.6126660360023379,
      "epoch": 1.1926613417979195,
      "grad_norm": 0.1586424559354782,
      "learning_rate": 6.024779955224339e-05,
      "loss": 0.5937,
      "mean_token_accuracy": 0.819816593080759,
      "num_tokens": 413749018.0,
      "step": 12963
    },
    {
      "entropy": 0.5948704043403268,
      "epoch": 1.1927533481693606,
      "grad_norm": 0.15694181621074677,
      "learning_rate": 6.024473272610176e-05,
      "loss": 0.5736,
      "mean_token_accuracy": 0.8243765011429787,
      "num_tokens": 413781594.0,
      "step": 12964
    },
    {
      "entropy": 0.653703186661005,
      "epoch": 1.192845354540802,
      "grad_norm": 0.16050399839878082,
      "learning_rate": 6.0241665899960134e-05,
      "loss": 0.6263,
      "mean_token_accuracy": 0.8067592680454254,
      "num_tokens": 413813207.0,
      "step": 12965
    },
    {
      "entropy": 0.7070904653519392,
      "epoch": 1.1929373609122431,
      "grad_norm": 0.1708565056324005,
      "learning_rate": 6.0238599073818515e-05,
      "loss": 0.6957,
      "mean_token_accuracy": 0.7871012203395367,
      "num_tokens": 413844901.0,
      "step": 12966
    },
    {
      "entropy": 0.5429828404448926,
      "epoch": 1.1930293672836845,
      "grad_norm": 0.1629900187253952,
      "learning_rate": 6.023553224767688e-05,
      "loss": 0.5357,
      "mean_token_accuracy": 0.8366168662905693,
      "num_tokens": 413877448.0,
      "step": 12967
    },
    {
      "entropy": 0.6169707104563713,
      "epoch": 1.1931213736551256,
      "grad_norm": 0.15831728279590607,
      "learning_rate": 6.023246542153526e-05,
      "loss": 0.6085,
      "mean_token_accuracy": 0.8144526183605194,
      "num_tokens": 413908208.0,
      "step": 12968
    },
    {
      "entropy": 0.5788307599723339,
      "epoch": 1.1932133800265667,
      "grad_norm": 0.1558813750743866,
      "learning_rate": 6.0229398595393625e-05,
      "loss": 0.5672,
      "mean_token_accuracy": 0.8235832899808884,
      "num_tokens": 413940976.0,
      "step": 12969
    },
    {
      "entropy": 0.6377137303352356,
      "epoch": 1.193305386398008,
      "grad_norm": 0.15645325183868408,
      "learning_rate": 6.022633176925201e-05,
      "loss": 0.6272,
      "mean_token_accuracy": 0.8082258366048336,
      "num_tokens": 413973003.0,
      "step": 12970
    },
    {
      "entropy": 0.5512858163565397,
      "epoch": 1.1933973927694492,
      "grad_norm": 0.15477624535560608,
      "learning_rate": 6.022326494311038e-05,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8314606621861458,
      "num_tokens": 414005771.0,
      "step": 12971
    },
    {
      "entropy": 0.6227812208235264,
      "epoch": 1.1934893991408906,
      "grad_norm": 0.15793730318546295,
      "learning_rate": 6.022019811696875e-05,
      "loss": 0.6081,
      "mean_token_accuracy": 0.8114351406693459,
      "num_tokens": 414037530.0,
      "step": 12972
    },
    {
      "entropy": 0.728292353451252,
      "epoch": 1.1935814055123317,
      "grad_norm": 0.16534411907196045,
      "learning_rate": 6.0217131290827124e-05,
      "loss": 0.7256,
      "mean_token_accuracy": 0.775928758084774,
      "num_tokens": 414069593.0,
      "step": 12973
    },
    {
      "entropy": 0.5524968486279249,
      "epoch": 1.1936734118837729,
      "grad_norm": 0.15492941439151764,
      "learning_rate": 6.0214064464685505e-05,
      "loss": 0.5284,
      "mean_token_accuracy": 0.8390820994973183,
      "num_tokens": 414101352.0,
      "step": 12974
    },
    {
      "entropy": 0.6708187200129032,
      "epoch": 1.1937654182552142,
      "grad_norm": 0.16469743847846985,
      "learning_rate": 6.021099763854387e-05,
      "loss": 0.6495,
      "mean_token_accuracy": 0.8006170988082886,
      "num_tokens": 414133668.0,
      "step": 12975
    },
    {
      "entropy": 0.5528062689118087,
      "epoch": 1.1938574246266553,
      "grad_norm": 0.14786450564861298,
      "learning_rate": 6.020793081240225e-05,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8277317807078362,
      "num_tokens": 414166221.0,
      "step": 12976
    },
    {
      "entropy": 0.6161195635795593,
      "epoch": 1.1939494309980967,
      "grad_norm": 0.1597299724817276,
      "learning_rate": 6.0204863986260616e-05,
      "loss": 0.5944,
      "mean_token_accuracy": 0.8179482519626617,
      "num_tokens": 414198025.0,
      "step": 12977
    },
    {
      "entropy": 0.6442152243107557,
      "epoch": 1.1940414373695378,
      "grad_norm": 0.1617545783519745,
      "learning_rate": 6.0201797160119e-05,
      "loss": 0.6154,
      "mean_token_accuracy": 0.8121930174529552,
      "num_tokens": 414229527.0,
      "step": 12978
    },
    {
      "entropy": 0.620650639757514,
      "epoch": 1.194133443740979,
      "grad_norm": 0.1524384468793869,
      "learning_rate": 6.019873033397737e-05,
      "loss": 0.6045,
      "mean_token_accuracy": 0.815290343016386,
      "num_tokens": 414261464.0,
      "step": 12979
    },
    {
      "entropy": 0.6087453765794635,
      "epoch": 1.1942254501124203,
      "grad_norm": 0.15629859268665314,
      "learning_rate": 6.019566350783574e-05,
      "loss": 0.5835,
      "mean_token_accuracy": 0.8214310519397259,
      "num_tokens": 414292966.0,
      "step": 12980
    },
    {
      "entropy": 0.550206933170557,
      "epoch": 1.1943174564838615,
      "grad_norm": 0.14983920753002167,
      "learning_rate": 6.0192596681694114e-05,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8345733024179935,
      "num_tokens": 414325321.0,
      "step": 12981
    },
    {
      "entropy": 0.563748930580914,
      "epoch": 1.1944094628553028,
      "grad_norm": 0.15943624079227448,
      "learning_rate": 6.0189529855552496e-05,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8265962265431881,
      "num_tokens": 414357853.0,
      "step": 12982
    },
    {
      "entropy": 0.5706585557200015,
      "epoch": 1.194501469226744,
      "grad_norm": 0.15355169773101807,
      "learning_rate": 6.0186463029410864e-05,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8310533054172993,
      "num_tokens": 414390458.0,
      "step": 12983
    },
    {
      "entropy": 0.6356290578842163,
      "epoch": 1.194593475598185,
      "grad_norm": 0.15525540709495544,
      "learning_rate": 6.018339620326924e-05,
      "loss": 0.6317,
      "mean_token_accuracy": 0.8059317618608475,
      "num_tokens": 414422999.0,
      "step": 12984
    },
    {
      "entropy": 0.7037357222288847,
      "epoch": 1.1946854819696264,
      "grad_norm": 0.16261082887649536,
      "learning_rate": 6.0180329377127606e-05,
      "loss": 0.7067,
      "mean_token_accuracy": 0.7860014848411083,
      "num_tokens": 414454800.0,
      "step": 12985
    },
    {
      "entropy": 0.6401643604040146,
      "epoch": 1.1947774883410676,
      "grad_norm": 0.16666653752326965,
      "learning_rate": 6.0177262550985994e-05,
      "loss": 0.6419,
      "mean_token_accuracy": 0.8058981075882912,
      "num_tokens": 414486000.0,
      "step": 12986
    },
    {
      "entropy": 0.6324488995596766,
      "epoch": 1.194869494712509,
      "grad_norm": 0.16603893041610718,
      "learning_rate": 6.017419572484436e-05,
      "loss": 0.6157,
      "mean_token_accuracy": 0.8092028759419918,
      "num_tokens": 414517813.0,
      "step": 12987
    },
    {
      "entropy": 0.484440291300416,
      "epoch": 1.19496150108395,
      "grad_norm": 0.16720198094844818,
      "learning_rate": 6.017112889870273e-05,
      "loss": 0.453,
      "mean_token_accuracy": 0.8541475683450699,
      "num_tokens": 414549084.0,
      "step": 12988
    },
    {
      "entropy": 0.593599483370781,
      "epoch": 1.1950535074553912,
      "grad_norm": 0.1477828323841095,
      "learning_rate": 6.0168062072561105e-05,
      "loss": 0.5892,
      "mean_token_accuracy": 0.8196864910423756,
      "num_tokens": 414581150.0,
      "step": 12989
    },
    {
      "entropy": 0.6493975836783648,
      "epoch": 1.1951455138268325,
      "grad_norm": 0.1598268747329712,
      "learning_rate": 6.0164995246419486e-05,
      "loss": 0.6394,
      "mean_token_accuracy": 0.8071234002709389,
      "num_tokens": 414613333.0,
      "step": 12990
    },
    {
      "entropy": 0.6309031648561358,
      "epoch": 1.1952375201982737,
      "grad_norm": 0.1608072817325592,
      "learning_rate": 6.016192842027786e-05,
      "loss": 0.6215,
      "mean_token_accuracy": 0.8098122328519821,
      "num_tokens": 414645258.0,
      "step": 12991
    },
    {
      "entropy": 0.6688140891492367,
      "epoch": 1.195329526569715,
      "grad_norm": 0.15593448281288147,
      "learning_rate": 6.015886159413623e-05,
      "loss": 0.6524,
      "mean_token_accuracy": 0.8010559342801571,
      "num_tokens": 414677662.0,
      "step": 12992
    },
    {
      "entropy": 0.646027434617281,
      "epoch": 1.1954215329411562,
      "grad_norm": 0.1600857973098755,
      "learning_rate": 6.0155794767994597e-05,
      "loss": 0.6352,
      "mean_token_accuracy": 0.8087330274283886,
      "num_tokens": 414709818.0,
      "step": 12993
    },
    {
      "entropy": 0.6340792719274759,
      "epoch": 1.1955135393125973,
      "grad_norm": 0.162253275513649,
      "learning_rate": 6.0152727941852985e-05,
      "loss": 0.6171,
      "mean_token_accuracy": 0.8109401538968086,
      "num_tokens": 414742208.0,
      "step": 12994
    },
    {
      "entropy": 0.5893305670469999,
      "epoch": 1.1956055456840387,
      "grad_norm": 0.16127710044384003,
      "learning_rate": 6.014966111571135e-05,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8225369602441788,
      "num_tokens": 414773761.0,
      "step": 12995
    },
    {
      "entropy": 0.6146195866167545,
      "epoch": 1.1956975520554798,
      "grad_norm": 0.15361468493938446,
      "learning_rate": 6.014659428956973e-05,
      "loss": 0.5848,
      "mean_token_accuracy": 0.8218158185482025,
      "num_tokens": 414806223.0,
      "step": 12996
    },
    {
      "entropy": 0.629583285190165,
      "epoch": 1.1957895584269211,
      "grad_norm": 0.16131363809108734,
      "learning_rate": 6.014352746342811e-05,
      "loss": 0.6104,
      "mean_token_accuracy": 0.8109550848603249,
      "num_tokens": 414838225.0,
      "step": 12997
    },
    {
      "entropy": 0.647470273077488,
      "epoch": 1.1958815647983623,
      "grad_norm": 0.16246531903743744,
      "learning_rate": 6.0140460637286477e-05,
      "loss": 0.6428,
      "mean_token_accuracy": 0.8032623305916786,
      "num_tokens": 414870740.0,
      "step": 12998
    },
    {
      "entropy": 0.5697713326662779,
      "epoch": 1.1959735711698034,
      "grad_norm": 0.15811394155025482,
      "learning_rate": 6.013739381114485e-05,
      "loss": 0.5427,
      "mean_token_accuracy": 0.829469345510006,
      "num_tokens": 414901847.0,
      "step": 12999
    },
    {
      "entropy": 0.5800871169194579,
      "epoch": 1.1960655775412448,
      "grad_norm": 0.15875762701034546,
      "learning_rate": 6.013432698500322e-05,
      "loss": 0.5583,
      "mean_token_accuracy": 0.8283914625644684,
      "num_tokens": 414933982.0,
      "step": 13000
    },
    {
      "entropy": 0.6328763710334897,
      "epoch": 1.196157583912686,
      "grad_norm": 0.15315742790699005,
      "learning_rate": 6.01312601588616e-05,
      "loss": 0.6234,
      "mean_token_accuracy": 0.8105231374502182,
      "num_tokens": 414966123.0,
      "step": 13001
    },
    {
      "entropy": 0.6992759639397264,
      "epoch": 1.1962495902841273,
      "grad_norm": 0.1617761105298996,
      "learning_rate": 6.0128193332719975e-05,
      "loss": 0.6992,
      "mean_token_accuracy": 0.7894046194851398,
      "num_tokens": 414998477.0,
      "step": 13002
    },
    {
      "entropy": 0.495380621869117,
      "epoch": 1.1963415966555684,
      "grad_norm": 0.15228545665740967,
      "learning_rate": 6.012512650657834e-05,
      "loss": 0.4885,
      "mean_token_accuracy": 0.8497373796999454,
      "num_tokens": 415030505.0,
      "step": 13003
    },
    {
      "entropy": 0.6274855546653271,
      "epoch": 1.1964336030270095,
      "grad_norm": 0.1522865891456604,
      "learning_rate": 6.012205968043672e-05,
      "loss": 0.6216,
      "mean_token_accuracy": 0.8081307671964169,
      "num_tokens": 415062354.0,
      "step": 13004
    },
    {
      "entropy": 0.6410403046756983,
      "epoch": 1.1965256093984509,
      "grad_norm": 0.15856176614761353,
      "learning_rate": 6.01189928542951e-05,
      "loss": 0.6415,
      "mean_token_accuracy": 0.8016859143972397,
      "num_tokens": 415094897.0,
      "step": 13005
    },
    {
      "entropy": 0.5958133451640606,
      "epoch": 1.196617615769892,
      "grad_norm": 0.15684495866298676,
      "learning_rate": 6.011592602815347e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8189065344631672,
      "num_tokens": 415126804.0,
      "step": 13006
    },
    {
      "entropy": 0.56841048784554,
      "epoch": 1.1967096221413334,
      "grad_norm": 0.15678708255290985,
      "learning_rate": 6.011285920201184e-05,
      "loss": 0.5666,
      "mean_token_accuracy": 0.8218335025012493,
      "num_tokens": 415159285.0,
      "step": 13007
    },
    {
      "entropy": 0.6145685957744718,
      "epoch": 1.1968016285127745,
      "grad_norm": 0.1700088083744049,
      "learning_rate": 6.010979237587021e-05,
      "loss": 0.6069,
      "mean_token_accuracy": 0.8185309320688248,
      "num_tokens": 415191435.0,
      "step": 13008
    },
    {
      "entropy": 0.641692440956831,
      "epoch": 1.1968936348842156,
      "grad_norm": 0.17840273678302765,
      "learning_rate": 6.010672554972859e-05,
      "loss": 0.6486,
      "mean_token_accuracy": 0.8080082423985004,
      "num_tokens": 415221777.0,
      "step": 13009
    },
    {
      "entropy": 0.5901474598795176,
      "epoch": 1.196985641255657,
      "grad_norm": 0.16320621967315674,
      "learning_rate": 6.0103658723586966e-05,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8232305683195591,
      "num_tokens": 415253249.0,
      "step": 13010
    },
    {
      "entropy": 0.6393881645053625,
      "epoch": 1.1970776476270981,
      "grad_norm": 0.166458398103714,
      "learning_rate": 6.0100591897445333e-05,
      "loss": 0.6463,
      "mean_token_accuracy": 0.8000488430261612,
      "num_tokens": 415284619.0,
      "step": 13011
    },
    {
      "entropy": 0.5633505834266543,
      "epoch": 1.1971696539985395,
      "grad_norm": 0.15751788020133972,
      "learning_rate": 6.009752507130371e-05,
      "loss": 0.5502,
      "mean_token_accuracy": 0.8285314626991749,
      "num_tokens": 415316613.0,
      "step": 13012
    },
    {
      "entropy": 0.6663244599476457,
      "epoch": 1.1972616603699806,
      "grad_norm": 0.16832461953163147,
      "learning_rate": 6.009445824516209e-05,
      "loss": 0.6473,
      "mean_token_accuracy": 0.8000671900808811,
      "num_tokens": 415348759.0,
      "step": 13013
    },
    {
      "entropy": 0.6139925876632333,
      "epoch": 1.1973536667414217,
      "grad_norm": 0.15311630070209503,
      "learning_rate": 6.009139141902046e-05,
      "loss": 0.5984,
      "mean_token_accuracy": 0.8202281147241592,
      "num_tokens": 415381391.0,
      "step": 13014
    },
    {
      "entropy": 0.6013573408126831,
      "epoch": 1.197445673112863,
      "grad_norm": 0.15590256452560425,
      "learning_rate": 6.008832459287883e-05,
      "loss": 0.583,
      "mean_token_accuracy": 0.8191929049789906,
      "num_tokens": 415413297.0,
      "step": 13015
    },
    {
      "entropy": 0.6387728881090879,
      "epoch": 1.1975376794843042,
      "grad_norm": 0.16446565091609955,
      "learning_rate": 6.00852577667372e-05,
      "loss": 0.6368,
      "mean_token_accuracy": 0.8049489594995975,
      "num_tokens": 415445540.0,
      "step": 13016
    },
    {
      "entropy": 0.68042266741395,
      "epoch": 1.1976296858557456,
      "grad_norm": 0.1559692919254303,
      "learning_rate": 6.008219094059558e-05,
      "loss": 0.6665,
      "mean_token_accuracy": 0.797811858355999,
      "num_tokens": 415477324.0,
      "step": 13017
    },
    {
      "entropy": 0.6310700550675392,
      "epoch": 1.1977216922271867,
      "grad_norm": 0.1557679921388626,
      "learning_rate": 6.0079124114453956e-05,
      "loss": 0.6157,
      "mean_token_accuracy": 0.8141250610351562,
      "num_tokens": 415509585.0,
      "step": 13018
    },
    {
      "entropy": 0.6321336114779115,
      "epoch": 1.1978136985986279,
      "grad_norm": 0.15582680702209473,
      "learning_rate": 6.0076057288312324e-05,
      "loss": 0.6201,
      "mean_token_accuracy": 0.8059694431722164,
      "num_tokens": 415541673.0,
      "step": 13019
    },
    {
      "entropy": 0.6103196330368519,
      "epoch": 1.1979057049700692,
      "grad_norm": 0.15349751710891724,
      "learning_rate": 6.00729904621707e-05,
      "loss": 0.5924,
      "mean_token_accuracy": 0.8189001232385635,
      "num_tokens": 415573567.0,
      "step": 13020
    },
    {
      "entropy": 0.6831368785351515,
      "epoch": 1.1979977113415103,
      "grad_norm": 0.164847269654274,
      "learning_rate": 6.006992363602908e-05,
      "loss": 0.6666,
      "mean_token_accuracy": 0.7969490811228752,
      "num_tokens": 415605644.0,
      "step": 13021
    },
    {
      "entropy": 0.6885925643146038,
      "epoch": 1.1980897177129517,
      "grad_norm": 0.15915772318840027,
      "learning_rate": 6.006685680988745e-05,
      "loss": 0.6905,
      "mean_token_accuracy": 0.7946601919829845,
      "num_tokens": 415636969.0,
      "step": 13022
    },
    {
      "entropy": 0.6556102000176907,
      "epoch": 1.1981817240843928,
      "grad_norm": 0.1618194282054901,
      "learning_rate": 6.006378998374582e-05,
      "loss": 0.6414,
      "mean_token_accuracy": 0.8020149134099483,
      "num_tokens": 415668978.0,
      "step": 13023
    },
    {
      "entropy": 0.6273304326459765,
      "epoch": 1.198273730455834,
      "grad_norm": 0.1561550498008728,
      "learning_rate": 6.006072315760419e-05,
      "loss": 0.5971,
      "mean_token_accuracy": 0.8146014921367168,
      "num_tokens": 415700407.0,
      "step": 13024
    },
    {
      "entropy": 0.6700379997491837,
      "epoch": 1.1983657368272753,
      "grad_norm": 0.17570564150810242,
      "learning_rate": 6.005765633146257e-05,
      "loss": 0.6472,
      "mean_token_accuracy": 0.8037039041519165,
      "num_tokens": 415731753.0,
      "step": 13025
    },
    {
      "entropy": 0.6506435163319111,
      "epoch": 1.1984577431987165,
      "grad_norm": 0.16119031608104706,
      "learning_rate": 6.0054589505320946e-05,
      "loss": 0.647,
      "mean_token_accuracy": 0.8017200939357281,
      "num_tokens": 415764052.0,
      "step": 13026
    },
    {
      "entropy": 0.7035203166306019,
      "epoch": 1.1985497495701578,
      "grad_norm": 0.15784604847431183,
      "learning_rate": 6.0051522679179314e-05,
      "loss": 0.6848,
      "mean_token_accuracy": 0.7890656664967537,
      "num_tokens": 415796643.0,
      "step": 13027
    },
    {
      "entropy": 0.49096013605594635,
      "epoch": 1.198641755941599,
      "grad_norm": 0.16812551021575928,
      "learning_rate": 6.00484558530377e-05,
      "loss": 0.4913,
      "mean_token_accuracy": 0.8497197069227695,
      "num_tokens": 415828688.0,
      "step": 13028
    },
    {
      "entropy": 0.6662360150367022,
      "epoch": 1.19873376231304,
      "grad_norm": 0.15924043953418732,
      "learning_rate": 6.004538902689607e-05,
      "loss": 0.6591,
      "mean_token_accuracy": 0.795759528875351,
      "num_tokens": 415861096.0,
      "step": 13029
    },
    {
      "entropy": 0.581687806174159,
      "epoch": 1.1988257686844814,
      "grad_norm": 0.15644729137420654,
      "learning_rate": 6.0042322200754445e-05,
      "loss": 0.5591,
      "mean_token_accuracy": 0.8290119245648384,
      "num_tokens": 415892992.0,
      "step": 13030
    },
    {
      "entropy": 0.6425761356949806,
      "epoch": 1.1989177750559226,
      "grad_norm": 0.16236570477485657,
      "learning_rate": 6.003925537461281e-05,
      "loss": 0.6457,
      "mean_token_accuracy": 0.8059611432254314,
      "num_tokens": 415925063.0,
      "step": 13031
    },
    {
      "entropy": 0.6094631515443325,
      "epoch": 1.199009781427364,
      "grad_norm": 0.15907356142997742,
      "learning_rate": 6.0036188548471194e-05,
      "loss": 0.5974,
      "mean_token_accuracy": 0.8095955923199654,
      "num_tokens": 415956325.0,
      "step": 13032
    },
    {
      "entropy": 0.6236458830535412,
      "epoch": 1.199101787798805,
      "grad_norm": 0.1632297933101654,
      "learning_rate": 6.003312172232957e-05,
      "loss": 0.6237,
      "mean_token_accuracy": 0.8093317821621895,
      "num_tokens": 415988253.0,
      "step": 13033
    },
    {
      "entropy": 0.6454971022903919,
      "epoch": 1.1991937941702462,
      "grad_norm": 0.15786227583885193,
      "learning_rate": 6.003005489618794e-05,
      "loss": 0.636,
      "mean_token_accuracy": 0.8103420399129391,
      "num_tokens": 416020234.0,
      "step": 13034
    },
    {
      "entropy": 0.6108095832169056,
      "epoch": 1.1992858005416875,
      "grad_norm": 0.15627451241016388,
      "learning_rate": 6.002698807004631e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8241627402603626,
      "num_tokens": 416052062.0,
      "step": 13035
    },
    {
      "entropy": 0.6200011745095253,
      "epoch": 1.1993778069131287,
      "grad_norm": 0.15605860948562622,
      "learning_rate": 6.002392124390469e-05,
      "loss": 0.5873,
      "mean_token_accuracy": 0.8201241306960583,
      "num_tokens": 416083978.0,
      "step": 13036
    },
    {
      "entropy": 0.780923081561923,
      "epoch": 1.19946981328457,
      "grad_norm": 0.1712142825126648,
      "learning_rate": 6.002085441776306e-05,
      "loss": 0.7676,
      "mean_token_accuracy": 0.7683721370995045,
      "num_tokens": 416116296.0,
      "step": 13037
    },
    {
      "entropy": 0.5971446130424738,
      "epoch": 1.1995618196560112,
      "grad_norm": 0.15581689774990082,
      "learning_rate": 6.0017787591621435e-05,
      "loss": 0.5903,
      "mean_token_accuracy": 0.8223257698118687,
      "num_tokens": 416148293.0,
      "step": 13038
    },
    {
      "entropy": 0.585293160751462,
      "epoch": 1.1996538260274523,
      "grad_norm": 0.15093782544136047,
      "learning_rate": 6.00147207654798e-05,
      "loss": 0.5627,
      "mean_token_accuracy": 0.8243624344468117,
      "num_tokens": 416179624.0,
      "step": 13039
    },
    {
      "entropy": 0.6821936741471291,
      "epoch": 1.1997458323988937,
      "grad_norm": 0.15975630283355713,
      "learning_rate": 6.0011653939338185e-05,
      "loss": 0.652,
      "mean_token_accuracy": 0.8044241145253181,
      "num_tokens": 416211621.0,
      "step": 13040
    },
    {
      "entropy": 0.5972270015627146,
      "epoch": 1.1998378387703348,
      "grad_norm": 0.16343489289283752,
      "learning_rate": 6.000858711319656e-05,
      "loss": 0.5837,
      "mean_token_accuracy": 0.8217406161129475,
      "num_tokens": 416244172.0,
      "step": 13041
    },
    {
      "entropy": 0.6135218366980553,
      "epoch": 1.1999298451417761,
      "grad_norm": 0.17141760885715485,
      "learning_rate": 6.000552028705493e-05,
      "loss": 0.6062,
      "mean_token_accuracy": 0.8165040202438831,
      "num_tokens": 416276375.0,
      "step": 13042
    },
    {
      "entropy": 0.7249039113521576,
      "epoch": 1.2000218515132173,
      "grad_norm": 0.1679985523223877,
      "learning_rate": 6.00024534609133e-05,
      "loss": 0.7264,
      "mean_token_accuracy": 0.778686385601759,
      "num_tokens": 416308682.0,
      "step": 13043
    },
    {
      "entropy": 0.5981201231479645,
      "epoch": 1.2001138578846584,
      "grad_norm": 0.16005094349384308,
      "learning_rate": 5.999938663477168e-05,
      "loss": 0.5882,
      "mean_token_accuracy": 0.8155602775514126,
      "num_tokens": 416340670.0,
      "step": 13044
    },
    {
      "entropy": 0.6112986002117395,
      "epoch": 1.2002058642560998,
      "grad_norm": 0.1578228920698166,
      "learning_rate": 5.999631980863005e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8118961714208126,
      "num_tokens": 416372559.0,
      "step": 13045
    },
    {
      "entropy": 0.5449804961681366,
      "epoch": 1.200297870627541,
      "grad_norm": 0.16628816723823547,
      "learning_rate": 5.9993252982488426e-05,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8462403155863285,
      "num_tokens": 416404689.0,
      "step": 13046
    },
    {
      "entropy": 0.5885756397619843,
      "epoch": 1.2003898769989823,
      "grad_norm": 0.15005575120449066,
      "learning_rate": 5.9990186156346794e-05,
      "loss": 0.5691,
      "mean_token_accuracy": 0.8245609775185585,
      "num_tokens": 416437036.0,
      "step": 13047
    },
    {
      "entropy": 0.6448211232200265,
      "epoch": 1.2004818833704234,
      "grad_norm": 0.16171927750110626,
      "learning_rate": 5.9987119330205175e-05,
      "loss": 0.6533,
      "mean_token_accuracy": 0.8031815588474274,
      "num_tokens": 416468984.0,
      "step": 13048
    },
    {
      "entropy": 0.596752991899848,
      "epoch": 1.2005738897418645,
      "grad_norm": 0.165516197681427,
      "learning_rate": 5.998405250406355e-05,
      "loss": 0.5912,
      "mean_token_accuracy": 0.8196251913905144,
      "num_tokens": 416501475.0,
      "step": 13049
    },
    {
      "entropy": 0.6446545775979757,
      "epoch": 1.2006658961133059,
      "grad_norm": 0.15637964010238647,
      "learning_rate": 5.998098567792192e-05,
      "loss": 0.6324,
      "mean_token_accuracy": 0.8092103265225887,
      "num_tokens": 416533159.0,
      "step": 13050
    },
    {
      "entropy": 0.5756786717101932,
      "epoch": 1.200757902484747,
      "grad_norm": 0.1502281129360199,
      "learning_rate": 5.997791885178029e-05,
      "loss": 0.559,
      "mean_token_accuracy": 0.8283563777804375,
      "num_tokens": 416565371.0,
      "step": 13051
    },
    {
      "entropy": 0.6191781722009182,
      "epoch": 1.2008499088561884,
      "grad_norm": 0.1534678190946579,
      "learning_rate": 5.9974852025638674e-05,
      "loss": 0.6019,
      "mean_token_accuracy": 0.8105527497828007,
      "num_tokens": 416597246.0,
      "step": 13052
    },
    {
      "entropy": 0.6787472246214747,
      "epoch": 1.2009419152276295,
      "grad_norm": 0.15239636600017548,
      "learning_rate": 5.997178519949704e-05,
      "loss": 0.658,
      "mean_token_accuracy": 0.7987881079316139,
      "num_tokens": 416628701.0,
      "step": 13053
    },
    {
      "entropy": 0.7060671411454678,
      "epoch": 1.2010339215990706,
      "grad_norm": 0.15692664682865143,
      "learning_rate": 5.9968718373355416e-05,
      "loss": 0.6754,
      "mean_token_accuracy": 0.7938053980469704,
      "num_tokens": 416660860.0,
      "step": 13054
    },
    {
      "entropy": 0.6640402004122734,
      "epoch": 1.201125927970512,
      "grad_norm": 0.16413982212543488,
      "learning_rate": 5.9965651547213784e-05,
      "loss": 0.6602,
      "mean_token_accuracy": 0.8016851842403412,
      "num_tokens": 416692554.0,
      "step": 13055
    },
    {
      "entropy": 0.6695271041244268,
      "epoch": 1.2012179343419531,
      "grad_norm": 0.15800701081752777,
      "learning_rate": 5.9962584721072165e-05,
      "loss": 0.6504,
      "mean_token_accuracy": 0.7998920567333698,
      "num_tokens": 416724988.0,
      "step": 13056
    },
    {
      "entropy": 0.6623428091406822,
      "epoch": 1.2013099407133945,
      "grad_norm": 0.15953543782234192,
      "learning_rate": 5.995951789493054e-05,
      "loss": 0.6327,
      "mean_token_accuracy": 0.8050685673952103,
      "num_tokens": 416757089.0,
      "step": 13057
    },
    {
      "entropy": 0.5610321583226323,
      "epoch": 1.2014019470848356,
      "grad_norm": 0.1519128531217575,
      "learning_rate": 5.995645106878891e-05,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8318494111299515,
      "num_tokens": 416789054.0,
      "step": 13058
    },
    {
      "entropy": 0.7393439337611198,
      "epoch": 1.2014939534562767,
      "grad_norm": 0.16864150762557983,
      "learning_rate": 5.995338424264729e-05,
      "loss": 0.7429,
      "mean_token_accuracy": 0.7747665084898472,
      "num_tokens": 416819818.0,
      "step": 13059
    },
    {
      "entropy": 0.6477660369127989,
      "epoch": 1.201585959827718,
      "grad_norm": 0.1624513864517212,
      "learning_rate": 5.9950317416505664e-05,
      "loss": 0.6263,
      "mean_token_accuracy": 0.805779404938221,
      "num_tokens": 416851335.0,
      "step": 13060
    },
    {
      "entropy": 0.6810052879154682,
      "epoch": 1.2016779661991592,
      "grad_norm": 0.1598825454711914,
      "learning_rate": 5.994725059036403e-05,
      "loss": 0.6679,
      "mean_token_accuracy": 0.7951637990772724,
      "num_tokens": 416883547.0,
      "step": 13061
    },
    {
      "entropy": 0.5024714269675314,
      "epoch": 1.2017699725706006,
      "grad_norm": 0.15066565573215485,
      "learning_rate": 5.9944183764222407e-05,
      "loss": 0.494,
      "mean_token_accuracy": 0.8478314653038979,
      "num_tokens": 416916064.0,
      "step": 13062
    },
    {
      "entropy": 0.688546996563673,
      "epoch": 1.2018619789420417,
      "grad_norm": 0.16151192784309387,
      "learning_rate": 5.994111693808079e-05,
      "loss": 0.6878,
      "mean_token_accuracy": 0.79188646748662,
      "num_tokens": 416948114.0,
      "step": 13063
    },
    {
      "entropy": 0.6685505770146847,
      "epoch": 1.2019539853134829,
      "grad_norm": 0.16030554473400116,
      "learning_rate": 5.9938050111939156e-05,
      "loss": 0.6708,
      "mean_token_accuracy": 0.7989250384271145,
      "num_tokens": 416980245.0,
      "step": 13064
    },
    {
      "entropy": 0.6248587593436241,
      "epoch": 1.2020459916849242,
      "grad_norm": 0.16989301145076752,
      "learning_rate": 5.993498328579753e-05,
      "loss": 0.6203,
      "mean_token_accuracy": 0.8118893466889858,
      "num_tokens": 417013013.0,
      "step": 13065
    },
    {
      "entropy": 0.5620099706575274,
      "epoch": 1.2021379980563653,
      "grad_norm": 0.15042123198509216,
      "learning_rate": 5.99319164596559e-05,
      "loss": 0.5508,
      "mean_token_accuracy": 0.8276017047464848,
      "num_tokens": 417044875.0,
      "step": 13066
    },
    {
      "entropy": 0.6327706314623356,
      "epoch": 1.2022300044278067,
      "grad_norm": 0.15852420032024384,
      "learning_rate": 5.9928849633514287e-05,
      "loss": 0.6233,
      "mean_token_accuracy": 0.8049004264175892,
      "num_tokens": 417076132.0,
      "step": 13067
    },
    {
      "entropy": 0.5271862936206162,
      "epoch": 1.2023220107992478,
      "grad_norm": 0.15647000074386597,
      "learning_rate": 5.9925782807372654e-05,
      "loss": 0.5176,
      "mean_token_accuracy": 0.8389548510313034,
      "num_tokens": 417108401.0,
      "step": 13068
    },
    {
      "entropy": 0.5281755165196955,
      "epoch": 1.202414017170689,
      "grad_norm": 0.15575921535491943,
      "learning_rate": 5.992271598123103e-05,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8438719026744366,
      "num_tokens": 417140821.0,
      "step": 13069
    },
    {
      "entropy": 0.5119760571978986,
      "epoch": 1.2025060235421303,
      "grad_norm": 0.1571093052625656,
      "learning_rate": 5.99196491550894e-05,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8428016379475594,
      "num_tokens": 417173223.0,
      "step": 13070
    },
    {
      "entropy": 0.5582075212150812,
      "epoch": 1.2025980299135715,
      "grad_norm": 0.15184567868709564,
      "learning_rate": 5.991658232894778e-05,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8269798457622528,
      "num_tokens": 417205637.0,
      "step": 13071
    },
    {
      "entropy": 0.7545397281646729,
      "epoch": 1.2026900362850128,
      "grad_norm": 0.17350968718528748,
      "learning_rate": 5.991351550280615e-05,
      "loss": 0.7554,
      "mean_token_accuracy": 0.7714822739362717,
      "num_tokens": 417237252.0,
      "step": 13072
    },
    {
      "entropy": 0.7508737705647945,
      "epoch": 1.202782042656454,
      "grad_norm": 0.17599104344844818,
      "learning_rate": 5.991044867666452e-05,
      "loss": 0.7468,
      "mean_token_accuracy": 0.7762726806104183,
      "num_tokens": 417268587.0,
      "step": 13073
    },
    {
      "entropy": 0.5915879029780626,
      "epoch": 1.202874049027895,
      "grad_norm": 0.15614014863967896,
      "learning_rate": 5.9907381850522896e-05,
      "loss": 0.5811,
      "mean_token_accuracy": 0.8259908966720104,
      "num_tokens": 417300671.0,
      "step": 13074
    },
    {
      "entropy": 0.5367394760251045,
      "epoch": 1.2029660553993364,
      "grad_norm": 0.1457509696483612,
      "learning_rate": 5.990431502438128e-05,
      "loss": 0.5327,
      "mean_token_accuracy": 0.8388849496841431,
      "num_tokens": 417332736.0,
      "step": 13075
    },
    {
      "entropy": 0.623305331915617,
      "epoch": 1.2030580617707776,
      "grad_norm": 0.16302768886089325,
      "learning_rate": 5.9901248198239645e-05,
      "loss": 0.6082,
      "mean_token_accuracy": 0.8165913559496403,
      "num_tokens": 417365504.0,
      "step": 13076
    },
    {
      "entropy": 0.6208368628285825,
      "epoch": 1.203150068142219,
      "grad_norm": 0.15475411713123322,
      "learning_rate": 5.989818137209802e-05,
      "loss": 0.6032,
      "mean_token_accuracy": 0.8149573653936386,
      "num_tokens": 417397274.0,
      "step": 13077
    },
    {
      "entropy": 0.6887991279363632,
      "epoch": 1.20324207451366,
      "grad_norm": 0.1625126749277115,
      "learning_rate": 5.989511454595639e-05,
      "loss": 0.6759,
      "mean_token_accuracy": 0.7968224734067917,
      "num_tokens": 417429477.0,
      "step": 13078
    },
    {
      "entropy": 0.5465079648420215,
      "epoch": 1.2033340808851012,
      "grad_norm": 0.14786958694458008,
      "learning_rate": 5.989204771981477e-05,
      "loss": 0.5298,
      "mean_token_accuracy": 0.8412254452705383,
      "num_tokens": 417462195.0,
      "step": 13079
    },
    {
      "entropy": 0.6721578305587173,
      "epoch": 1.2034260872565425,
      "grad_norm": 0.16892589628696442,
      "learning_rate": 5.9888980893673143e-05,
      "loss": 0.6517,
      "mean_token_accuracy": 0.800596222281456,
      "num_tokens": 417494490.0,
      "step": 13080
    },
    {
      "entropy": 0.5193083323538303,
      "epoch": 1.2035180936279837,
      "grad_norm": 0.149741530418396,
      "learning_rate": 5.988591406753151e-05,
      "loss": 0.4979,
      "mean_token_accuracy": 0.8465625569224358,
      "num_tokens": 417526462.0,
      "step": 13081
    },
    {
      "entropy": 0.7250908724963665,
      "epoch": 1.203610099999425,
      "grad_norm": 0.16351507604122162,
      "learning_rate": 5.9882847241389886e-05,
      "loss": 0.71,
      "mean_token_accuracy": 0.7879924550652504,
      "num_tokens": 417558904.0,
      "step": 13082
    },
    {
      "entropy": 0.6176714599132538,
      "epoch": 1.2037021063708662,
      "grad_norm": 0.15854254364967346,
      "learning_rate": 5.987978041524827e-05,
      "loss": 0.6047,
      "mean_token_accuracy": 0.8166135922074318,
      "num_tokens": 417590198.0,
      "step": 13083
    },
    {
      "entropy": 0.7308749295771122,
      "epoch": 1.2037941127423073,
      "grad_norm": 0.16058477759361267,
      "learning_rate": 5.9876713589106635e-05,
      "loss": 0.7152,
      "mean_token_accuracy": 0.7858859673142433,
      "num_tokens": 417621551.0,
      "step": 13084
    },
    {
      "entropy": 0.6275713755749166,
      "epoch": 1.2038861191137487,
      "grad_norm": 0.1582873910665512,
      "learning_rate": 5.987364676296501e-05,
      "loss": 0.6134,
      "mean_token_accuracy": 0.8091071136295795,
      "num_tokens": 417652974.0,
      "step": 13085
    },
    {
      "entropy": 0.7231127433478832,
      "epoch": 1.2039781254851898,
      "grad_norm": 0.17213678359985352,
      "learning_rate": 5.987057993682338e-05,
      "loss": 0.722,
      "mean_token_accuracy": 0.7874291613698006,
      "num_tokens": 417684123.0,
      "step": 13086
    },
    {
      "entropy": 0.7487799450755119,
      "epoch": 1.2040701318566311,
      "grad_norm": 0.17559406161308289,
      "learning_rate": 5.986751311068176e-05,
      "loss": 0.7538,
      "mean_token_accuracy": 0.7727548256516457,
      "num_tokens": 417714737.0,
      "step": 13087
    },
    {
      "entropy": 0.660110654309392,
      "epoch": 1.2041621382280723,
      "grad_norm": 0.16147683560848236,
      "learning_rate": 5.9864446284540134e-05,
      "loss": 0.6556,
      "mean_token_accuracy": 0.8001503981649876,
      "num_tokens": 417747217.0,
      "step": 13088
    },
    {
      "entropy": 0.6091721314005554,
      "epoch": 1.2042541445995134,
      "grad_norm": 0.15969803929328918,
      "learning_rate": 5.98613794583985e-05,
      "loss": 0.5931,
      "mean_token_accuracy": 0.8209819160401821,
      "num_tokens": 417779087.0,
      "step": 13089
    },
    {
      "entropy": 0.6081715244799852,
      "epoch": 1.2043461509709548,
      "grad_norm": 0.15693315863609314,
      "learning_rate": 5.985831263225688e-05,
      "loss": 0.5969,
      "mean_token_accuracy": 0.8159861862659454,
      "num_tokens": 417811231.0,
      "step": 13090
    },
    {
      "entropy": 0.5634792316704988,
      "epoch": 1.204438157342396,
      "grad_norm": 0.16645294427871704,
      "learning_rate": 5.985524580611526e-05,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8276128396391869,
      "num_tokens": 417843381.0,
      "step": 13091
    },
    {
      "entropy": 0.56101318821311,
      "epoch": 1.2045301637138373,
      "grad_norm": 0.15301276743412018,
      "learning_rate": 5.9852178979973626e-05,
      "loss": 0.5567,
      "mean_token_accuracy": 0.8297299891710281,
      "num_tokens": 417875168.0,
      "step": 13092
    },
    {
      "entropy": 0.6568753328174353,
      "epoch": 1.2046221700852784,
      "grad_norm": 0.16681376099586487,
      "learning_rate": 5.9849112153832e-05,
      "loss": 0.6438,
      "mean_token_accuracy": 0.804804153740406,
      "num_tokens": 417906282.0,
      "step": 13093
    },
    {
      "entropy": 0.5852841213345528,
      "epoch": 1.2047141764567195,
      "grad_norm": 0.15899331867694855,
      "learning_rate": 5.984604532769038e-05,
      "loss": 0.5742,
      "mean_token_accuracy": 0.8202630393207073,
      "num_tokens": 417937722.0,
      "step": 13094
    },
    {
      "entropy": 0.5972239701077342,
      "epoch": 1.2048061828281609,
      "grad_norm": 0.1582976132631302,
      "learning_rate": 5.984297850154875e-05,
      "loss": 0.5807,
      "mean_token_accuracy": 0.823598712682724,
      "num_tokens": 417970488.0,
      "step": 13095
    },
    {
      "entropy": 0.5944761876016855,
      "epoch": 1.204898189199602,
      "grad_norm": 0.1553552895784378,
      "learning_rate": 5.9839911675407124e-05,
      "loss": 0.5779,
      "mean_token_accuracy": 0.8202619552612305,
      "num_tokens": 418003116.0,
      "step": 13096
    },
    {
      "entropy": 0.6459754481911659,
      "epoch": 1.2049901955710434,
      "grad_norm": 0.15979954600334167,
      "learning_rate": 5.983684484926549e-05,
      "loss": 0.6303,
      "mean_token_accuracy": 0.8100377023220062,
      "num_tokens": 418035554.0,
      "step": 13097
    },
    {
      "entropy": 0.6333734886720777,
      "epoch": 1.2050822019424845,
      "grad_norm": 0.162678524851799,
      "learning_rate": 5.9833778023123874e-05,
      "loss": 0.6142,
      "mean_token_accuracy": 0.8106794208288193,
      "num_tokens": 418067504.0,
      "step": 13098
    },
    {
      "entropy": 0.7060635285452008,
      "epoch": 1.2051742083139256,
      "grad_norm": 0.15542009472846985,
      "learning_rate": 5.983071119698225e-05,
      "loss": 0.6828,
      "mean_token_accuracy": 0.7953747250139713,
      "num_tokens": 418099572.0,
      "step": 13099
    },
    {
      "entropy": 0.6663775537163019,
      "epoch": 1.205266214685367,
      "grad_norm": 0.16628192365169525,
      "learning_rate": 5.9827644370840616e-05,
      "loss": 0.6529,
      "mean_token_accuracy": 0.8012927360832691,
      "num_tokens": 418131454.0,
      "step": 13100
    },
    {
      "entropy": 0.6782145984470844,
      "epoch": 1.2053582210568081,
      "grad_norm": 0.16178931295871735,
      "learning_rate": 5.982457754469899e-05,
      "loss": 0.6736,
      "mean_token_accuracy": 0.7949395179748535,
      "num_tokens": 418163682.0,
      "step": 13101
    },
    {
      "entropy": 0.4830636531114578,
      "epoch": 1.2054502274282495,
      "grad_norm": 0.16171038150787354,
      "learning_rate": 5.982151071855737e-05,
      "loss": 0.4642,
      "mean_token_accuracy": 0.8538331761956215,
      "num_tokens": 418196127.0,
      "step": 13102
    },
    {
      "entropy": 0.6658601588569582,
      "epoch": 1.2055422337996906,
      "grad_norm": 0.16121065616607666,
      "learning_rate": 5.981844389241574e-05,
      "loss": 0.6406,
      "mean_token_accuracy": 0.8073973543941975,
      "num_tokens": 418227754.0,
      "step": 13103
    },
    {
      "entropy": 0.6698483880609274,
      "epoch": 1.2056342401711317,
      "grad_norm": 0.1608714908361435,
      "learning_rate": 5.9815377066274115e-05,
      "loss": 0.6615,
      "mean_token_accuracy": 0.8007297180593014,
      "num_tokens": 418259396.0,
      "step": 13104
    },
    {
      "entropy": 0.60167545825243,
      "epoch": 1.205726246542573,
      "grad_norm": 0.15650604665279388,
      "learning_rate": 5.981231024013248e-05,
      "loss": 0.596,
      "mean_token_accuracy": 0.8215013705193996,
      "num_tokens": 418290772.0,
      "step": 13105
    },
    {
      "entropy": 0.5868304669857025,
      "epoch": 1.2058182529140142,
      "grad_norm": 0.16162465512752533,
      "learning_rate": 5.980924341399087e-05,
      "loss": 0.581,
      "mean_token_accuracy": 0.823226910084486,
      "num_tokens": 418322692.0,
      "step": 13106
    },
    {
      "entropy": 0.6229428080841899,
      "epoch": 1.2059102592854556,
      "grad_norm": 0.161623015999794,
      "learning_rate": 5.980617658784924e-05,
      "loss": 0.6157,
      "mean_token_accuracy": 0.811893817037344,
      "num_tokens": 418355220.0,
      "step": 13107
    },
    {
      "entropy": 0.631968361325562,
      "epoch": 1.2060022656568967,
      "grad_norm": 0.16255250573158264,
      "learning_rate": 5.9803109761707606e-05,
      "loss": 0.6185,
      "mean_token_accuracy": 0.8139911890029907,
      "num_tokens": 418387535.0,
      "step": 13108
    },
    {
      "entropy": 0.636125510558486,
      "epoch": 1.2060942720283379,
      "grad_norm": 0.15837161242961884,
      "learning_rate": 5.980004293556598e-05,
      "loss": 0.6165,
      "mean_token_accuracy": 0.8129058107733727,
      "num_tokens": 418419712.0,
      "step": 13109
    },
    {
      "entropy": 0.6087225396186113,
      "epoch": 1.2061862783997792,
      "grad_norm": 0.15673337876796722,
      "learning_rate": 5.979697610942436e-05,
      "loss": 0.6095,
      "mean_token_accuracy": 0.8136234953999519,
      "num_tokens": 418452073.0,
      "step": 13110
    },
    {
      "entropy": 0.6492832293733954,
      "epoch": 1.2062782847712203,
      "grad_norm": 0.1652195006608963,
      "learning_rate": 5.979390928328274e-05,
      "loss": 0.6437,
      "mean_token_accuracy": 0.8031660318374634,
      "num_tokens": 418484482.0,
      "step": 13111
    },
    {
      "entropy": 0.6306798998266459,
      "epoch": 1.2063702911426617,
      "grad_norm": 0.15662643313407898,
      "learning_rate": 5.9790842457141105e-05,
      "loss": 0.6168,
      "mean_token_accuracy": 0.8107193969190121,
      "num_tokens": 418516413.0,
      "step": 13112
    },
    {
      "entropy": 0.6813495475798845,
      "epoch": 1.2064622975141028,
      "grad_norm": 0.16090594232082367,
      "learning_rate": 5.978777563099948e-05,
      "loss": 0.6724,
      "mean_token_accuracy": 0.7922053635120392,
      "num_tokens": 418547955.0,
      "step": 13113
    },
    {
      "entropy": 0.5972009934484959,
      "epoch": 1.206554303885544,
      "grad_norm": 0.1505705565214157,
      "learning_rate": 5.978470880485786e-05,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8222817927598953,
      "num_tokens": 418580312.0,
      "step": 13114
    },
    {
      "entropy": 0.6117472760379314,
      "epoch": 1.2066463102569853,
      "grad_norm": 0.15710972249507904,
      "learning_rate": 5.978164197871623e-05,
      "loss": 0.6025,
      "mean_token_accuracy": 0.8162397108972073,
      "num_tokens": 418612802.0,
      "step": 13115
    },
    {
      "entropy": 0.552031728439033,
      "epoch": 1.2067383166284265,
      "grad_norm": 0.15427550673484802,
      "learning_rate": 5.9778575152574604e-05,
      "loss": 0.5367,
      "mean_token_accuracy": 0.83394780382514,
      "num_tokens": 418644657.0,
      "step": 13116
    },
    {
      "entropy": 0.6701981835067272,
      "epoch": 1.2068303229998678,
      "grad_norm": 0.16571323573589325,
      "learning_rate": 5.977550832643297e-05,
      "loss": 0.6539,
      "mean_token_accuracy": 0.80174845084548,
      "num_tokens": 418676238.0,
      "step": 13117
    },
    {
      "entropy": 0.6613078825175762,
      "epoch": 1.206922329371309,
      "grad_norm": 0.15801458060741425,
      "learning_rate": 5.977244150029135e-05,
      "loss": 0.6497,
      "mean_token_accuracy": 0.8020570911467075,
      "num_tokens": 418708330.0,
      "step": 13118
    },
    {
      "entropy": 0.5330276535823941,
      "epoch": 1.20701433574275,
      "grad_norm": 0.15914183855056763,
      "learning_rate": 5.976937467414973e-05,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8397204801440239,
      "num_tokens": 418740819.0,
      "step": 13119
    },
    {
      "entropy": 0.5191608141176403,
      "epoch": 1.2071063421141914,
      "grad_norm": 0.14689156413078308,
      "learning_rate": 5.9766307848008095e-05,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8463682532310486,
      "num_tokens": 418773223.0,
      "step": 13120
    },
    {
      "entropy": 0.5385568495839834,
      "epoch": 1.2071983484856326,
      "grad_norm": 0.1587272584438324,
      "learning_rate": 5.976324102186648e-05,
      "loss": 0.5293,
      "mean_token_accuracy": 0.8369233533740044,
      "num_tokens": 418804845.0,
      "step": 13121
    },
    {
      "entropy": 0.6722131948918104,
      "epoch": 1.207290354857074,
      "grad_norm": 0.16234353184700012,
      "learning_rate": 5.976017419572485e-05,
      "loss": 0.6671,
      "mean_token_accuracy": 0.7957969680428505,
      "num_tokens": 418836788.0,
      "step": 13122
    },
    {
      "entropy": 0.5006232541054487,
      "epoch": 1.207382361228515,
      "grad_norm": 0.15010125935077667,
      "learning_rate": 5.975710736958322e-05,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8475238271057606,
      "num_tokens": 418868993.0,
      "step": 13123
    },
    {
      "entropy": 0.6865468416363001,
      "epoch": 1.2074743675999562,
      "grad_norm": 0.16745220124721527,
      "learning_rate": 5.9754040543441594e-05,
      "loss": 0.6731,
      "mean_token_accuracy": 0.7963152639567852,
      "num_tokens": 418901206.0,
      "step": 13124
    },
    {
      "entropy": 0.6356700304895639,
      "epoch": 1.2075663739713975,
      "grad_norm": 0.16255518794059753,
      "learning_rate": 5.9750973717299975e-05,
      "loss": 0.623,
      "mean_token_accuracy": 0.8104881532490253,
      "num_tokens": 418933341.0,
      "step": 13125
    },
    {
      "entropy": 0.6473942832089961,
      "epoch": 1.2076583803428387,
      "grad_norm": 0.16581399738788605,
      "learning_rate": 5.974790689115834e-05,
      "loss": 0.6324,
      "mean_token_accuracy": 0.8042908161878586,
      "num_tokens": 418964819.0,
      "step": 13126
    },
    {
      "entropy": 0.5330698247998953,
      "epoch": 1.20775038671428,
      "grad_norm": 0.1471177041530609,
      "learning_rate": 5.974484006501672e-05,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8353292532265186,
      "num_tokens": 418996896.0,
      "step": 13127
    },
    {
      "entropy": 0.6411269661039114,
      "epoch": 1.2078423930857212,
      "grad_norm": 0.1592719852924347,
      "learning_rate": 5.9741773238875086e-05,
      "loss": 0.6415,
      "mean_token_accuracy": 0.8082250021398067,
      "num_tokens": 419029014.0,
      "step": 13128
    },
    {
      "entropy": 0.5957448156550527,
      "epoch": 1.2079343994571623,
      "grad_norm": 0.15885396301746368,
      "learning_rate": 5.973870641273347e-05,
      "loss": 0.5829,
      "mean_token_accuracy": 0.8242083750665188,
      "num_tokens": 419060550.0,
      "step": 13129
    },
    {
      "entropy": 0.6489674113690853,
      "epoch": 1.2080264058286037,
      "grad_norm": 0.16194912791252136,
      "learning_rate": 5.973563958659184e-05,
      "loss": 0.6215,
      "mean_token_accuracy": 0.8085448369383812,
      "num_tokens": 419092523.0,
      "step": 13130
    },
    {
      "entropy": 0.6992375236004591,
      "epoch": 1.2081184122000448,
      "grad_norm": 0.1658024936914444,
      "learning_rate": 5.973257276045021e-05,
      "loss": 0.685,
      "mean_token_accuracy": 0.7920464314520359,
      "num_tokens": 419124614.0,
      "step": 13131
    },
    {
      "entropy": 0.6435550637543201,
      "epoch": 1.2082104185714861,
      "grad_norm": 0.15352317690849304,
      "learning_rate": 5.9729505934308584e-05,
      "loss": 0.6307,
      "mean_token_accuracy": 0.8087445050477982,
      "num_tokens": 419157382.0,
      "step": 13132
    },
    {
      "entropy": 0.6043731011450291,
      "epoch": 1.2083024249429273,
      "grad_norm": 0.15727946162223816,
      "learning_rate": 5.9726439108166966e-05,
      "loss": 0.5862,
      "mean_token_accuracy": 0.8186298310756683,
      "num_tokens": 419189409.0,
      "step": 13133
    },
    {
      "entropy": 0.5368176102638245,
      "epoch": 1.2083944313143684,
      "grad_norm": 0.15075229108333588,
      "learning_rate": 5.9723372282025334e-05,
      "loss": 0.5077,
      "mean_token_accuracy": 0.8420897796750069,
      "num_tokens": 419221855.0,
      "step": 13134
    },
    {
      "entropy": 0.6630297787487507,
      "epoch": 1.2084864376858098,
      "grad_norm": 0.1603853702545166,
      "learning_rate": 5.972030545588371e-05,
      "loss": 0.6371,
      "mean_token_accuracy": 0.8008332177996635,
      "num_tokens": 419253351.0,
      "step": 13135
    },
    {
      "entropy": 0.6657328065484762,
      "epoch": 1.208578444057251,
      "grad_norm": 0.15949048101902008,
      "learning_rate": 5.9717238629742076e-05,
      "loss": 0.6569,
      "mean_token_accuracy": 0.8032860308885574,
      "num_tokens": 419284274.0,
      "step": 13136
    },
    {
      "entropy": 0.6224249973893166,
      "epoch": 1.2086704504286923,
      "grad_norm": 0.1591549813747406,
      "learning_rate": 5.971417180360046e-05,
      "loss": 0.5983,
      "mean_token_accuracy": 0.8144010715186596,
      "num_tokens": 419316347.0,
      "step": 13137
    },
    {
      "entropy": 0.5499319527298212,
      "epoch": 1.2087624568001334,
      "grad_norm": 0.14890260994434357,
      "learning_rate": 5.971110497745883e-05,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8347436599433422,
      "num_tokens": 419348751.0,
      "step": 13138
    },
    {
      "entropy": 0.5938344839960337,
      "epoch": 1.2088544631715745,
      "grad_norm": 0.15981389582157135,
      "learning_rate": 5.97080381513172e-05,
      "loss": 0.5982,
      "mean_token_accuracy": 0.8153700456023216,
      "num_tokens": 419381519.0,
      "step": 13139
    },
    {
      "entropy": 0.608480763155967,
      "epoch": 1.2089464695430159,
      "grad_norm": 0.17269013822078705,
      "learning_rate": 5.9704971325175575e-05,
      "loss": 0.6106,
      "mean_token_accuracy": 0.8159881681203842,
      "num_tokens": 419412932.0,
      "step": 13140
    },
    {
      "entropy": 0.6091998512856662,
      "epoch": 1.209038475914457,
      "grad_norm": 0.15953688323497772,
      "learning_rate": 5.9701904499033956e-05,
      "loss": 0.6103,
      "mean_token_accuracy": 0.8091586530208588,
      "num_tokens": 419445355.0,
      "step": 13141
    },
    {
      "entropy": 0.5396599248051643,
      "epoch": 1.2091304822858984,
      "grad_norm": 0.15257325768470764,
      "learning_rate": 5.9698837672892324e-05,
      "loss": 0.5286,
      "mean_token_accuracy": 0.8359788544476032,
      "num_tokens": 419476827.0,
      "step": 13142
    },
    {
      "entropy": 0.6095498017966747,
      "epoch": 1.2092224886573395,
      "grad_norm": 0.16600647568702698,
      "learning_rate": 5.96957708467507e-05,
      "loss": 0.5981,
      "mean_token_accuracy": 0.8077440857887268,
      "num_tokens": 419508446.0,
      "step": 13143
    },
    {
      "entropy": 0.5441438062116504,
      "epoch": 1.2093144950287806,
      "grad_norm": 0.16232797503471375,
      "learning_rate": 5.969270402060907e-05,
      "loss": 0.5326,
      "mean_token_accuracy": 0.8381537161767483,
      "num_tokens": 419539637.0,
      "step": 13144
    },
    {
      "entropy": 0.5281860725954175,
      "epoch": 1.209406501400222,
      "grad_norm": 0.1536368429660797,
      "learning_rate": 5.9689637194467455e-05,
      "loss": 0.5325,
      "mean_token_accuracy": 0.8371071554720402,
      "num_tokens": 419571936.0,
      "step": 13145
    },
    {
      "entropy": 0.6750898659229279,
      "epoch": 1.2094985077716631,
      "grad_norm": 0.17552302777767181,
      "learning_rate": 5.968657036832582e-05,
      "loss": 0.6598,
      "mean_token_accuracy": 0.7996794916689396,
      "num_tokens": 419603072.0,
      "step": 13146
    },
    {
      "entropy": 0.5616124616935849,
      "epoch": 1.2095905141431045,
      "grad_norm": 0.15635329484939575,
      "learning_rate": 5.968350354218419e-05,
      "loss": 0.546,
      "mean_token_accuracy": 0.8317404873669147,
      "num_tokens": 419635663.0,
      "step": 13147
    },
    {
      "entropy": 0.8008070606738329,
      "epoch": 1.2096825205145456,
      "grad_norm": 0.16728059947490692,
      "learning_rate": 5.968043671604258e-05,
      "loss": 0.7814,
      "mean_token_accuracy": 0.7668768912553787,
      "num_tokens": 419667416.0,
      "step": 13148
    },
    {
      "entropy": 0.5894123110920191,
      "epoch": 1.2097745268859867,
      "grad_norm": 0.15319043397903442,
      "learning_rate": 5.967736988990095e-05,
      "loss": 0.5675,
      "mean_token_accuracy": 0.8264666870236397,
      "num_tokens": 419699534.0,
      "step": 13149
    },
    {
      "entropy": 0.6921615414321423,
      "epoch": 1.209866533257428,
      "grad_norm": 0.16260379552841187,
      "learning_rate": 5.967430306375932e-05,
      "loss": 0.6793,
      "mean_token_accuracy": 0.7930688709020615,
      "num_tokens": 419731468.0,
      "step": 13150
    },
    {
      "entropy": 0.6317856265231967,
      "epoch": 1.2099585396288692,
      "grad_norm": 0.16068755090236664,
      "learning_rate": 5.967123623761769e-05,
      "loss": 0.6169,
      "mean_token_accuracy": 0.8085537627339363,
      "num_tokens": 419763327.0,
      "step": 13151
    },
    {
      "entropy": 0.4640753131825477,
      "epoch": 1.2100505460003106,
      "grad_norm": 0.14517125487327576,
      "learning_rate": 5.966816941147607e-05,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8630923293530941,
      "num_tokens": 419796095.0,
      "step": 13152
    },
    {
      "entropy": 0.6687251450493932,
      "epoch": 1.2101425523717517,
      "grad_norm": 0.15870478749275208,
      "learning_rate": 5.9665102585334445e-05,
      "loss": 0.658,
      "mean_token_accuracy": 0.8002379909157753,
      "num_tokens": 419828214.0,
      "step": 13153
    },
    {
      "entropy": 0.6557367062196136,
      "epoch": 1.2102345587431929,
      "grad_norm": 0.15799404680728912,
      "learning_rate": 5.966203575919281e-05,
      "loss": 0.6461,
      "mean_token_accuracy": 0.8031837232410908,
      "num_tokens": 419860529.0,
      "step": 13154
    },
    {
      "entropy": 0.7007526494562626,
      "epoch": 1.2103265651146342,
      "grad_norm": 0.15655775368213654,
      "learning_rate": 5.965896893305119e-05,
      "loss": 0.6791,
      "mean_token_accuracy": 0.7904522567987442,
      "num_tokens": 419892821.0,
      "step": 13155
    },
    {
      "entropy": 0.5189750827848911,
      "epoch": 1.2104185714860753,
      "grad_norm": 0.16523636877536774,
      "learning_rate": 5.965590210690957e-05,
      "loss": 0.5134,
      "mean_token_accuracy": 0.843421820551157,
      "num_tokens": 419925050.0,
      "step": 13156
    },
    {
      "entropy": 0.495705708861351,
      "epoch": 1.2105105778575167,
      "grad_norm": 0.1454704850912094,
      "learning_rate": 5.965283528076794e-05,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8529044091701508,
      "num_tokens": 419957130.0,
      "step": 13157
    },
    {
      "entropy": 0.6107633970677853,
      "epoch": 1.2106025842289578,
      "grad_norm": 0.16592176258563995,
      "learning_rate": 5.964976845462631e-05,
      "loss": 0.6007,
      "mean_token_accuracy": 0.8127967268228531,
      "num_tokens": 419988173.0,
      "step": 13158
    },
    {
      "entropy": 0.6614165939390659,
      "epoch": 1.210694590600399,
      "grad_norm": 0.16310779750347137,
      "learning_rate": 5.964670162848468e-05,
      "loss": 0.6437,
      "mean_token_accuracy": 0.8059839643537998,
      "num_tokens": 420019590.0,
      "step": 13159
    },
    {
      "entropy": 0.6659792228601873,
      "epoch": 1.2107865969718403,
      "grad_norm": 0.16182632744312286,
      "learning_rate": 5.964363480234306e-05,
      "loss": 0.6543,
      "mean_token_accuracy": 0.8005308844149113,
      "num_tokens": 420052206.0,
      "step": 13160
    },
    {
      "entropy": 0.5891882958821952,
      "epoch": 1.2108786033432815,
      "grad_norm": 0.15923456847667694,
      "learning_rate": 5.9640567976201436e-05,
      "loss": 0.5788,
      "mean_token_accuracy": 0.8227400444447994,
      "num_tokens": 420083710.0,
      "step": 13161
    },
    {
      "entropy": 0.6711738146841526,
      "epoch": 1.2109706097147228,
      "grad_norm": 0.15900318324565887,
      "learning_rate": 5.9637501150059804e-05,
      "loss": 0.6595,
      "mean_token_accuracy": 0.7946132980287075,
      "num_tokens": 420115671.0,
      "step": 13162
    },
    {
      "entropy": 0.5277845861855894,
      "epoch": 1.211062616086164,
      "grad_norm": 0.15055802464485168,
      "learning_rate": 5.963443432391818e-05,
      "loss": 0.5235,
      "mean_token_accuracy": 0.8387626260519028,
      "num_tokens": 420147685.0,
      "step": 13163
    },
    {
      "entropy": 0.5997943393886089,
      "epoch": 1.211154622457605,
      "grad_norm": 0.15726159512996674,
      "learning_rate": 5.963136749777656e-05,
      "loss": 0.5801,
      "mean_token_accuracy": 0.8219645619392395,
      "num_tokens": 420179475.0,
      "step": 13164
    },
    {
      "entropy": 0.604179747402668,
      "epoch": 1.2112466288290464,
      "grad_norm": 0.15921853482723236,
      "learning_rate": 5.962830067163493e-05,
      "loss": 0.6009,
      "mean_token_accuracy": 0.81578004732728,
      "num_tokens": 420211288.0,
      "step": 13165
    },
    {
      "entropy": 0.6161545189097524,
      "epoch": 1.2113386352004876,
      "grad_norm": 0.1591368317604065,
      "learning_rate": 5.96252338454933e-05,
      "loss": 0.6077,
      "mean_token_accuracy": 0.8154373839497566,
      "num_tokens": 420243075.0,
      "step": 13166
    },
    {
      "entropy": 0.6892130058258772,
      "epoch": 1.211430641571929,
      "grad_norm": 0.16130900382995605,
      "learning_rate": 5.962216701935167e-05,
      "loss": 0.6831,
      "mean_token_accuracy": 0.7928603477776051,
      "num_tokens": 420275160.0,
      "step": 13167
    },
    {
      "entropy": 0.6408870257437229,
      "epoch": 1.21152264794337,
      "grad_norm": 0.1560058742761612,
      "learning_rate": 5.961910019321005e-05,
      "loss": 0.6221,
      "mean_token_accuracy": 0.8095366954803467,
      "num_tokens": 420307583.0,
      "step": 13168
    },
    {
      "entropy": 0.5973208574578166,
      "epoch": 1.2116146543148112,
      "grad_norm": 0.15026362240314484,
      "learning_rate": 5.9616033367068426e-05,
      "loss": 0.5796,
      "mean_token_accuracy": 0.8198951296508312,
      "num_tokens": 420339859.0,
      "step": 13169
    },
    {
      "entropy": 0.5604884852655232,
      "epoch": 1.2117066606862525,
      "grad_norm": 0.15843282639980316,
      "learning_rate": 5.9612966540926794e-05,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8286520093679428,
      "num_tokens": 420370860.0,
      "step": 13170
    },
    {
      "entropy": 0.6225533969700336,
      "epoch": 1.2117986670576937,
      "grad_norm": 0.16125939786434174,
      "learning_rate": 5.960989971478517e-05,
      "loss": 0.6297,
      "mean_token_accuracy": 0.8119363151490688,
      "num_tokens": 420402124.0,
      "step": 13171
    },
    {
      "entropy": 0.6353972051292658,
      "epoch": 1.211890673429135,
      "grad_norm": 0.16152624785900116,
      "learning_rate": 5.960683288864355e-05,
      "loss": 0.6346,
      "mean_token_accuracy": 0.8086857125163078,
      "num_tokens": 420434336.0,
      "step": 13172
    },
    {
      "entropy": 0.6990877315402031,
      "epoch": 1.2119826798005762,
      "grad_norm": 0.1614162176847458,
      "learning_rate": 5.960376606250192e-05,
      "loss": 0.6883,
      "mean_token_accuracy": 0.7884212993085384,
      "num_tokens": 420465254.0,
      "step": 13173
    },
    {
      "entropy": 0.545277604367584,
      "epoch": 1.2120746861720173,
      "grad_norm": 0.15011556446552277,
      "learning_rate": 5.960069923636029e-05,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8352583721280098,
      "num_tokens": 420497337.0,
      "step": 13174
    },
    {
      "entropy": 0.6150883426889777,
      "epoch": 1.2121666925434587,
      "grad_norm": 0.1706770658493042,
      "learning_rate": 5.959763241021866e-05,
      "loss": 0.6025,
      "mean_token_accuracy": 0.8143964037299156,
      "num_tokens": 420529001.0,
      "step": 13175
    },
    {
      "entropy": 0.5865510310977697,
      "epoch": 1.2122586989148998,
      "grad_norm": 0.1594560444355011,
      "learning_rate": 5.959456558407704e-05,
      "loss": 0.5697,
      "mean_token_accuracy": 0.8226295411586761,
      "num_tokens": 420560835.0,
      "step": 13176
    },
    {
      "entropy": 0.5820660162717104,
      "epoch": 1.2123507052863411,
      "grad_norm": 0.1571420282125473,
      "learning_rate": 5.9591498757935416e-05,
      "loss": 0.581,
      "mean_token_accuracy": 0.823383629322052,
      "num_tokens": 420593111.0,
      "step": 13177
    },
    {
      "entropy": 0.5779969384893775,
      "epoch": 1.2124427116577823,
      "grad_norm": 0.16471336781978607,
      "learning_rate": 5.9588431931793784e-05,
      "loss": 0.5481,
      "mean_token_accuracy": 0.8337027840316296,
      "num_tokens": 420625866.0,
      "step": 13178
    },
    {
      "entropy": 0.570180376060307,
      "epoch": 1.2125347180292234,
      "grad_norm": 0.14994558691978455,
      "learning_rate": 5.9585365105652166e-05,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8291242010891438,
      "num_tokens": 420657881.0,
      "step": 13179
    },
    {
      "entropy": 0.6189155261963606,
      "epoch": 1.2126267244006648,
      "grad_norm": 0.1533227264881134,
      "learning_rate": 5.958229827951054e-05,
      "loss": 0.5971,
      "mean_token_accuracy": 0.815012764185667,
      "num_tokens": 420689067.0,
      "step": 13180
    },
    {
      "entropy": 0.5355353699997067,
      "epoch": 1.212718730772106,
      "grad_norm": 0.1646324247121811,
      "learning_rate": 5.957923145336891e-05,
      "loss": 0.5277,
      "mean_token_accuracy": 0.8390000052750111,
      "num_tokens": 420721349.0,
      "step": 13181
    },
    {
      "entropy": 0.6885867659002542,
      "epoch": 1.2128107371435473,
      "grad_norm": 0.1551678627729416,
      "learning_rate": 5.957616462722728e-05,
      "loss": 0.6638,
      "mean_token_accuracy": 0.7974307276308537,
      "num_tokens": 420753302.0,
      "step": 13182
    },
    {
      "entropy": 0.5309297172352672,
      "epoch": 1.2129027435149884,
      "grad_norm": 0.14730092883110046,
      "learning_rate": 5.9573097801085664e-05,
      "loss": 0.5126,
      "mean_token_accuracy": 0.840465996414423,
      "num_tokens": 420784646.0,
      "step": 13183
    },
    {
      "entropy": 0.6721677668392658,
      "epoch": 1.2129947498864295,
      "grad_norm": 0.1633022278547287,
      "learning_rate": 5.957003097494403e-05,
      "loss": 0.6659,
      "mean_token_accuracy": 0.7973197139799595,
      "num_tokens": 420816889.0,
      "step": 13184
    },
    {
      "entropy": 0.6845377087593079,
      "epoch": 1.2130867562578709,
      "grad_norm": 0.1645459234714508,
      "learning_rate": 5.956696414880241e-05,
      "loss": 0.6742,
      "mean_token_accuracy": 0.792406789958477,
      "num_tokens": 420847867.0,
      "step": 13185
    },
    {
      "entropy": 0.6871895547956228,
      "epoch": 1.213178762629312,
      "grad_norm": 0.16293787956237793,
      "learning_rate": 5.9563897322660775e-05,
      "loss": 0.6788,
      "mean_token_accuracy": 0.7983747757971287,
      "num_tokens": 420879086.0,
      "step": 13186
    },
    {
      "entropy": 0.6728473398834467,
      "epoch": 1.2132707690007534,
      "grad_norm": 0.15618187189102173,
      "learning_rate": 5.956083049651916e-05,
      "loss": 0.6653,
      "mean_token_accuracy": 0.7997690550982952,
      "num_tokens": 420911367.0,
      "step": 13187
    },
    {
      "entropy": 0.66188869997859,
      "epoch": 1.2133627753721945,
      "grad_norm": 0.1577100306749344,
      "learning_rate": 5.955776367037753e-05,
      "loss": 0.6523,
      "mean_token_accuracy": 0.8025219701230526,
      "num_tokens": 420943085.0,
      "step": 13188
    },
    {
      "entropy": 0.7008387669920921,
      "epoch": 1.2134547817436356,
      "grad_norm": 0.16246075928211212,
      "learning_rate": 5.9554696844235905e-05,
      "loss": 0.6845,
      "mean_token_accuracy": 0.7880585081875324,
      "num_tokens": 420975392.0,
      "step": 13189
    },
    {
      "entropy": 0.6924092248082161,
      "epoch": 1.213546788115077,
      "grad_norm": 0.16932572424411774,
      "learning_rate": 5.955163001809427e-05,
      "loss": 0.6835,
      "mean_token_accuracy": 0.794053465127945,
      "num_tokens": 421006602.0,
      "step": 13190
    },
    {
      "entropy": 0.65396648645401,
      "epoch": 1.2136387944865181,
      "grad_norm": 0.16560587286949158,
      "learning_rate": 5.9548563191952655e-05,
      "loss": 0.6418,
      "mean_token_accuracy": 0.8013336211442947,
      "num_tokens": 421038421.0,
      "step": 13191
    },
    {
      "entropy": 0.6298087043687701,
      "epoch": 1.2137308008579595,
      "grad_norm": 0.1561332643032074,
      "learning_rate": 5.954549636581103e-05,
      "loss": 0.6261,
      "mean_token_accuracy": 0.8081695064902306,
      "num_tokens": 421070701.0,
      "step": 13192
    },
    {
      "entropy": 0.6009201379492879,
      "epoch": 1.2138228072294006,
      "grad_norm": 0.1654665470123291,
      "learning_rate": 5.95424295396694e-05,
      "loss": 0.5873,
      "mean_token_accuracy": 0.8188036419451237,
      "num_tokens": 421103195.0,
      "step": 13193
    },
    {
      "entropy": 0.5859109833836555,
      "epoch": 1.2139148136008417,
      "grad_norm": 0.15753674507141113,
      "learning_rate": 5.953936271352777e-05,
      "loss": 0.5822,
      "mean_token_accuracy": 0.8202537931501865,
      "num_tokens": 421135045.0,
      "step": 13194
    },
    {
      "entropy": 0.6098553761839867,
      "epoch": 1.214006819972283,
      "grad_norm": 0.1570158749818802,
      "learning_rate": 5.953629588738615e-05,
      "loss": 0.6072,
      "mean_token_accuracy": 0.8150880336761475,
      "num_tokens": 421166337.0,
      "step": 13195
    },
    {
      "entropy": 0.6207606662064791,
      "epoch": 1.2140988263437242,
      "grad_norm": 0.1558356136083603,
      "learning_rate": 5.953322906124452e-05,
      "loss": 0.6005,
      "mean_token_accuracy": 0.8126688487827778,
      "num_tokens": 421198112.0,
      "step": 13196
    },
    {
      "entropy": 0.7565501360222697,
      "epoch": 1.2141908327151656,
      "grad_norm": 0.17019186913967133,
      "learning_rate": 5.9530162235102896e-05,
      "loss": 0.7447,
      "mean_token_accuracy": 0.7801199518144131,
      "num_tokens": 421230506.0,
      "step": 13197
    },
    {
      "entropy": 0.8065432403236628,
      "epoch": 1.2142828390866067,
      "grad_norm": 0.16805680096149445,
      "learning_rate": 5.9527095408961264e-05,
      "loss": 0.7751,
      "mean_token_accuracy": 0.7643074430525303,
      "num_tokens": 421262659.0,
      "step": 13198
    },
    {
      "entropy": 0.566320457495749,
      "epoch": 1.2143748454580479,
      "grad_norm": 0.1570814698934555,
      "learning_rate": 5.9524028582819645e-05,
      "loss": 0.5545,
      "mean_token_accuracy": 0.8294464088976383,
      "num_tokens": 421294422.0,
      "step": 13199
    },
    {
      "entropy": 0.5337793417274952,
      "epoch": 1.2144668518294892,
      "grad_norm": 0.14923767745494843,
      "learning_rate": 5.952096175667802e-05,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8335201144218445,
      "num_tokens": 421326352.0,
      "step": 13200
    },
    {
      "entropy": 0.6049358937889338,
      "epoch": 1.2145588582009303,
      "grad_norm": 0.15790985524654388,
      "learning_rate": 5.951789493053639e-05,
      "loss": 0.5976,
      "mean_token_accuracy": 0.8151747845113277,
      "num_tokens": 421358615.0,
      "step": 13201
    },
    {
      "entropy": 0.47603102400898933,
      "epoch": 1.2146508645723717,
      "grad_norm": 0.14663077890872955,
      "learning_rate": 5.951482810439476e-05,
      "loss": 0.4557,
      "mean_token_accuracy": 0.858923863619566,
      "num_tokens": 421390829.0,
      "step": 13202
    },
    {
      "entropy": 0.6390418782830238,
      "epoch": 1.2147428709438128,
      "grad_norm": 0.16124825179576874,
      "learning_rate": 5.9511761278253144e-05,
      "loss": 0.6299,
      "mean_token_accuracy": 0.8057404197752476,
      "num_tokens": 421422835.0,
      "step": 13203
    },
    {
      "entropy": 0.5791002810001373,
      "epoch": 1.214834877315254,
      "grad_norm": 0.15714548528194427,
      "learning_rate": 5.950869445211151e-05,
      "loss": 0.562,
      "mean_token_accuracy": 0.8299914076924324,
      "num_tokens": 421454738.0,
      "step": 13204
    },
    {
      "entropy": 0.779263511300087,
      "epoch": 1.2149268836866953,
      "grad_norm": 0.1733836531639099,
      "learning_rate": 5.9505627625969886e-05,
      "loss": 0.7733,
      "mean_token_accuracy": 0.7672752477228642,
      "num_tokens": 421485507.0,
      "step": 13205
    },
    {
      "entropy": 0.6525705587118864,
      "epoch": 1.2150188900581365,
      "grad_norm": 0.15816059708595276,
      "learning_rate": 5.9502560799828254e-05,
      "loss": 0.6518,
      "mean_token_accuracy": 0.8034031204879284,
      "num_tokens": 421517848.0,
      "step": 13206
    },
    {
      "entropy": 0.7017074059695005,
      "epoch": 1.2151108964295778,
      "grad_norm": 0.16651801764965057,
      "learning_rate": 5.9499493973686636e-05,
      "loss": 0.7039,
      "mean_token_accuracy": 0.7884657382965088,
      "num_tokens": 421550108.0,
      "step": 13207
    },
    {
      "entropy": 0.5221341550350189,
      "epoch": 1.215202902801019,
      "grad_norm": 0.15361745655536652,
      "learning_rate": 5.949642714754501e-05,
      "loss": 0.506,
      "mean_token_accuracy": 0.8440750502049923,
      "num_tokens": 421581096.0,
      "step": 13208
    },
    {
      "entropy": 0.5780196525156498,
      "epoch": 1.21529490917246,
      "grad_norm": 0.152821347117424,
      "learning_rate": 5.949336032140338e-05,
      "loss": 0.5521,
      "mean_token_accuracy": 0.8289959169924259,
      "num_tokens": 421612687.0,
      "step": 13209
    },
    {
      "entropy": 0.7700272649526596,
      "epoch": 1.2153869155439014,
      "grad_norm": 0.16088373959064484,
      "learning_rate": 5.949029349526176e-05,
      "loss": 0.7539,
      "mean_token_accuracy": 0.7697839327156544,
      "num_tokens": 421645244.0,
      "step": 13210
    },
    {
      "entropy": 0.6325645297765732,
      "epoch": 1.2154789219153426,
      "grad_norm": 0.1575164794921875,
      "learning_rate": 5.9487226669120134e-05,
      "loss": 0.6184,
      "mean_token_accuracy": 0.8106268234550953,
      "num_tokens": 421677013.0,
      "step": 13211
    },
    {
      "entropy": 0.6189095620065928,
      "epoch": 1.215570928286784,
      "grad_norm": 0.16264349222183228,
      "learning_rate": 5.94841598429785e-05,
      "loss": 0.6116,
      "mean_token_accuracy": 0.8151322901248932,
      "num_tokens": 421708853.0,
      "step": 13212
    },
    {
      "entropy": 0.6454381626099348,
      "epoch": 1.215662934658225,
      "grad_norm": 0.16490787267684937,
      "learning_rate": 5.948109301683688e-05,
      "loss": 0.6283,
      "mean_token_accuracy": 0.8060055859386921,
      "num_tokens": 421740338.0,
      "step": 13213
    },
    {
      "entropy": 0.6473144944757223,
      "epoch": 1.2157549410296662,
      "grad_norm": 0.1603682041168213,
      "learning_rate": 5.947802619069526e-05,
      "loss": 0.6193,
      "mean_token_accuracy": 0.8094655312597752,
      "num_tokens": 421771119.0,
      "step": 13214
    },
    {
      "entropy": 0.6937107555568218,
      "epoch": 1.2158469474011075,
      "grad_norm": 0.16511720418930054,
      "learning_rate": 5.9474959364553626e-05,
      "loss": 0.6931,
      "mean_token_accuracy": 0.7933748438954353,
      "num_tokens": 421803234.0,
      "step": 13215
    },
    {
      "entropy": 0.6870383638888597,
      "epoch": 1.2159389537725487,
      "grad_norm": 0.1638563871383667,
      "learning_rate": 5.9471892538412e-05,
      "loss": 0.6668,
      "mean_token_accuracy": 0.7930885180830956,
      "num_tokens": 421835542.0,
      "step": 13216
    },
    {
      "entropy": 0.5076667852699757,
      "epoch": 1.21603096014399,
      "grad_norm": 0.14463815093040466,
      "learning_rate": 5.946882571227037e-05,
      "loss": 0.492,
      "mean_token_accuracy": 0.8463995829224586,
      "num_tokens": 421868026.0,
      "step": 13217
    },
    {
      "entropy": 0.5187691897153854,
      "epoch": 1.2161229665154312,
      "grad_norm": 0.1505492925643921,
      "learning_rate": 5.946575888612875e-05,
      "loss": 0.5041,
      "mean_token_accuracy": 0.8448237366974354,
      "num_tokens": 421899946.0,
      "step": 13218
    },
    {
      "entropy": 0.5844136066734791,
      "epoch": 1.2162149728868723,
      "grad_norm": 0.15577493607997894,
      "learning_rate": 5.9462692059987125e-05,
      "loss": 0.5632,
      "mean_token_accuracy": 0.8235294297337532,
      "num_tokens": 421931559.0,
      "step": 13219
    },
    {
      "entropy": 0.6109859347343445,
      "epoch": 1.2163069792583137,
      "grad_norm": 0.15961863100528717,
      "learning_rate": 5.945962523384549e-05,
      "loss": 0.6005,
      "mean_token_accuracy": 0.8143456354737282,
      "num_tokens": 421962912.0,
      "step": 13220
    },
    {
      "entropy": 0.7015581391751766,
      "epoch": 1.2163989856297548,
      "grad_norm": 0.1673239916563034,
      "learning_rate": 5.945655840770387e-05,
      "loss": 0.7011,
      "mean_token_accuracy": 0.7903411351144314,
      "num_tokens": 421994355.0,
      "step": 13221
    },
    {
      "entropy": 0.5846061781048775,
      "epoch": 1.2164909920011961,
      "grad_norm": 0.1551455557346344,
      "learning_rate": 5.945349158156225e-05,
      "loss": 0.5813,
      "mean_token_accuracy": 0.8195104524493217,
      "num_tokens": 422026034.0,
      "step": 13222
    },
    {
      "entropy": 0.7132098507136106,
      "epoch": 1.2165829983726373,
      "grad_norm": 0.16304659843444824,
      "learning_rate": 5.9450424755420616e-05,
      "loss": 0.7041,
      "mean_token_accuracy": 0.7862912304699421,
      "num_tokens": 422058524.0,
      "step": 13223
    },
    {
      "entropy": 0.6514510437846184,
      "epoch": 1.2166750047440784,
      "grad_norm": 0.16076365113258362,
      "learning_rate": 5.944735792927899e-05,
      "loss": 0.6419,
      "mean_token_accuracy": 0.8042267300188541,
      "num_tokens": 422090713.0,
      "step": 13224
    },
    {
      "entropy": 0.5922899637371302,
      "epoch": 1.2167670111155198,
      "grad_norm": 0.157011479139328,
      "learning_rate": 5.944429110313736e-05,
      "loss": 0.5781,
      "mean_token_accuracy": 0.8240313194692135,
      "num_tokens": 422122579.0,
      "step": 13225
    },
    {
      "entropy": 0.45005336636677384,
      "epoch": 1.216859017486961,
      "grad_norm": 0.1414766162633896,
      "learning_rate": 5.944122427699575e-05,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8678591400384903,
      "num_tokens": 422154043.0,
      "step": 13226
    },
    {
      "entropy": 0.7117372453212738,
      "epoch": 1.2169510238584023,
      "grad_norm": 0.1689269244670868,
      "learning_rate": 5.9438157450854115e-05,
      "loss": 0.7099,
      "mean_token_accuracy": 0.7835213057696819,
      "num_tokens": 422186028.0,
      "step": 13227
    },
    {
      "entropy": 0.46597175672650337,
      "epoch": 1.2170430302298434,
      "grad_norm": 0.15331073105335236,
      "learning_rate": 5.943509062471248e-05,
      "loss": 0.4503,
      "mean_token_accuracy": 0.8615503124892712,
      "num_tokens": 422218444.0,
      "step": 13228
    },
    {
      "entropy": 0.5455353325232863,
      "epoch": 1.2171350366012845,
      "grad_norm": 0.16372524201869965,
      "learning_rate": 5.943202379857086e-05,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8344626799225807,
      "num_tokens": 422250418.0,
      "step": 13229
    },
    {
      "entropy": 0.6683600293472409,
      "epoch": 1.2172270429727259,
      "grad_norm": 0.161394402384758,
      "learning_rate": 5.942895697242924e-05,
      "loss": 0.674,
      "mean_token_accuracy": 0.7932555601000786,
      "num_tokens": 422282346.0,
      "step": 13230
    },
    {
      "entropy": 0.6059288010001183,
      "epoch": 1.217319049344167,
      "grad_norm": 0.15742287039756775,
      "learning_rate": 5.9425890146287614e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.8221736550331116,
      "num_tokens": 422314746.0,
      "step": 13231
    },
    {
      "entropy": 0.7229148577898741,
      "epoch": 1.2174110557156084,
      "grad_norm": 0.15794193744659424,
      "learning_rate": 5.942282332014598e-05,
      "loss": 0.7091,
      "mean_token_accuracy": 0.7856264039874077,
      "num_tokens": 422346853.0,
      "step": 13232
    },
    {
      "entropy": 0.6867092866450548,
      "epoch": 1.2175030620870495,
      "grad_norm": 0.1615864485502243,
      "learning_rate": 5.9419756494004356e-05,
      "loss": 0.6853,
      "mean_token_accuracy": 0.7939198724925518,
      "num_tokens": 422379581.0,
      "step": 13233
    },
    {
      "entropy": 0.5260039838030934,
      "epoch": 1.2175950684584906,
      "grad_norm": 0.15458272397518158,
      "learning_rate": 5.941668966786274e-05,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8420055508613586,
      "num_tokens": 422411722.0,
      "step": 13234
    },
    {
      "entropy": 0.5656318170949817,
      "epoch": 1.217687074829932,
      "grad_norm": 0.1488407701253891,
      "learning_rate": 5.9413622841721105e-05,
      "loss": 0.5489,
      "mean_token_accuracy": 0.8296033777296543,
      "num_tokens": 422442949.0,
      "step": 13235
    },
    {
      "entropy": 0.591316469013691,
      "epoch": 1.2177790812013731,
      "grad_norm": 0.16527453064918518,
      "learning_rate": 5.941055601557948e-05,
      "loss": 0.5781,
      "mean_token_accuracy": 0.8222857229411602,
      "num_tokens": 422475522.0,
      "step": 13236
    },
    {
      "entropy": 0.6896599009633064,
      "epoch": 1.2178710875728145,
      "grad_norm": 0.15875661373138428,
      "learning_rate": 5.940748918943785e-05,
      "loss": 0.6771,
      "mean_token_accuracy": 0.7936260513961315,
      "num_tokens": 422507947.0,
      "step": 13237
    },
    {
      "entropy": 0.5861258422955871,
      "epoch": 1.2179630939442556,
      "grad_norm": 0.15529315173625946,
      "learning_rate": 5.940442236329623e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8217477537691593,
      "num_tokens": 422538831.0,
      "step": 13238
    },
    {
      "entropy": 0.5921025136485696,
      "epoch": 1.2180551003156967,
      "grad_norm": 0.1608409434556961,
      "learning_rate": 5.9401355537154604e-05,
      "loss": 0.5912,
      "mean_token_accuracy": 0.821340274065733,
      "num_tokens": 422571074.0,
      "step": 13239
    },
    {
      "entropy": 0.5060474742203951,
      "epoch": 1.218147106687138,
      "grad_norm": 0.15965023636817932,
      "learning_rate": 5.939828871101297e-05,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8479134552180767,
      "num_tokens": 422603299.0,
      "step": 13240
    },
    {
      "entropy": 0.5989187890663743,
      "epoch": 1.2182391130585792,
      "grad_norm": 0.15627093613147736,
      "learning_rate": 5.939522188487135e-05,
      "loss": 0.5813,
      "mean_token_accuracy": 0.8221620693802834,
      "num_tokens": 422635343.0,
      "step": 13241
    },
    {
      "entropy": 0.6167837642133236,
      "epoch": 1.2183311194300206,
      "grad_norm": 0.1592186838388443,
      "learning_rate": 5.939215505872973e-05,
      "loss": 0.6185,
      "mean_token_accuracy": 0.8095571175217628,
      "num_tokens": 422666426.0,
      "step": 13242
    },
    {
      "entropy": 0.6356431394815445,
      "epoch": 1.2184231258014617,
      "grad_norm": 0.15438540279865265,
      "learning_rate": 5.9389088232588096e-05,
      "loss": 0.6301,
      "mean_token_accuracy": 0.805605411529541,
      "num_tokens": 422698455.0,
      "step": 13243
    },
    {
      "entropy": 0.6907157646492124,
      "epoch": 1.2185151321729029,
      "grad_norm": 0.16241972148418427,
      "learning_rate": 5.938602140644647e-05,
      "loss": 0.6771,
      "mean_token_accuracy": 0.7903539836406708,
      "num_tokens": 422730511.0,
      "step": 13244
    },
    {
      "entropy": 0.6216133944690228,
      "epoch": 1.2186071385443442,
      "grad_norm": 0.15086473524570465,
      "learning_rate": 5.938295458030485e-05,
      "loss": 0.6128,
      "mean_token_accuracy": 0.8094703406095505,
      "num_tokens": 422762228.0,
      "step": 13245
    },
    {
      "entropy": 0.5431929272599518,
      "epoch": 1.2186991449157853,
      "grad_norm": 0.15260355174541473,
      "learning_rate": 5.937988775416322e-05,
      "loss": 0.5257,
      "mean_token_accuracy": 0.838477112352848,
      "num_tokens": 422794288.0,
      "step": 13246
    },
    {
      "entropy": 0.5848347721621394,
      "epoch": 1.2187911512872267,
      "grad_norm": 0.1615760326385498,
      "learning_rate": 5.9376820928021594e-05,
      "loss": 0.5717,
      "mean_token_accuracy": 0.8239231035113335,
      "num_tokens": 422825968.0,
      "step": 13247
    },
    {
      "entropy": 0.6065728329122066,
      "epoch": 1.2188831576586678,
      "grad_norm": 0.1607409119606018,
      "learning_rate": 5.937375410187996e-05,
      "loss": 0.6055,
      "mean_token_accuracy": 0.8165749460458755,
      "num_tokens": 422858295.0,
      "step": 13248
    },
    {
      "entropy": 0.6256109569221735,
      "epoch": 1.218975164030109,
      "grad_norm": 0.1616056263446808,
      "learning_rate": 5.9370687275738344e-05,
      "loss": 0.6169,
      "mean_token_accuracy": 0.8084888346493244,
      "num_tokens": 422890109.0,
      "step": 13249
    },
    {
      "entropy": 0.6861296966671944,
      "epoch": 1.2190671704015503,
      "grad_norm": 0.15910793840885162,
      "learning_rate": 5.936762044959672e-05,
      "loss": 0.6662,
      "mean_token_accuracy": 0.7964163348078728,
      "num_tokens": 422921631.0,
      "step": 13250
    },
    {
      "entropy": 0.5608882596716285,
      "epoch": 1.2191591767729915,
      "grad_norm": 0.15002284944057465,
      "learning_rate": 5.9364553623455086e-05,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8285998925566673,
      "num_tokens": 422954105.0,
      "step": 13251
    },
    {
      "entropy": 0.6169448886066675,
      "epoch": 1.2192511831444328,
      "grad_norm": 0.16046175360679626,
      "learning_rate": 5.936148679731346e-05,
      "loss": 0.5949,
      "mean_token_accuracy": 0.8198980502784252,
      "num_tokens": 422986381.0,
      "step": 13252
    },
    {
      "entropy": 0.787552597001195,
      "epoch": 1.219343189515874,
      "grad_norm": 0.17527951300144196,
      "learning_rate": 5.935841997117184e-05,
      "loss": 0.7891,
      "mean_token_accuracy": 0.7644544392824173,
      "num_tokens": 423017135.0,
      "step": 13253
    },
    {
      "entropy": 0.6096754772588611,
      "epoch": 1.219435195887315,
      "grad_norm": 0.16240687668323517,
      "learning_rate": 5.935535314503021e-05,
      "loss": 0.6082,
      "mean_token_accuracy": 0.8174018897116184,
      "num_tokens": 423049187.0,
      "step": 13254
    },
    {
      "entropy": 0.6286130829248577,
      "epoch": 1.2195272022587564,
      "grad_norm": 0.16265273094177246,
      "learning_rate": 5.9352286318888585e-05,
      "loss": 0.6206,
      "mean_token_accuracy": 0.8120665661990643,
      "num_tokens": 423081308.0,
      "step": 13255
    },
    {
      "entropy": 0.6046443283557892,
      "epoch": 1.2196192086301976,
      "grad_norm": 0.15579424798488617,
      "learning_rate": 5.934921949274695e-05,
      "loss": 0.5873,
      "mean_token_accuracy": 0.8165584914386272,
      "num_tokens": 423113554.0,
      "step": 13256
    },
    {
      "entropy": 0.6825486626476049,
      "epoch": 1.219711215001639,
      "grad_norm": 0.15856119990348816,
      "learning_rate": 5.9346152666605334e-05,
      "loss": 0.6681,
      "mean_token_accuracy": 0.794383205473423,
      "num_tokens": 423145321.0,
      "step": 13257
    },
    {
      "entropy": 0.6805654317140579,
      "epoch": 1.21980322137308,
      "grad_norm": 0.16322481632232666,
      "learning_rate": 5.934308584046371e-05,
      "loss": 0.6612,
      "mean_token_accuracy": 0.7985414192080498,
      "num_tokens": 423177378.0,
      "step": 13258
    },
    {
      "entropy": 0.6004644008353353,
      "epoch": 1.2198952277445212,
      "grad_norm": 0.15809760987758636,
      "learning_rate": 5.9340019014322077e-05,
      "loss": 0.5747,
      "mean_token_accuracy": 0.8274427689611912,
      "num_tokens": 423208866.0,
      "step": 13259
    },
    {
      "entropy": 0.5872793542221189,
      "epoch": 1.2199872341159625,
      "grad_norm": 0.15757128596305847,
      "learning_rate": 5.933695218818045e-05,
      "loss": 0.5672,
      "mean_token_accuracy": 0.8287952952086926,
      "num_tokens": 423241034.0,
      "step": 13260
    },
    {
      "entropy": 0.6808096319437027,
      "epoch": 1.2200792404874037,
      "grad_norm": 0.1610376536846161,
      "learning_rate": 5.933388536203883e-05,
      "loss": 0.6747,
      "mean_token_accuracy": 0.7928323745727539,
      "num_tokens": 423273141.0,
      "step": 13261
    },
    {
      "entropy": 0.6853478075936437,
      "epoch": 1.220171246858845,
      "grad_norm": 0.16996759176254272,
      "learning_rate": 5.93308185358972e-05,
      "loss": 0.6804,
      "mean_token_accuracy": 0.7938939072191715,
      "num_tokens": 423304362.0,
      "step": 13262
    },
    {
      "entropy": 0.6484235916286707,
      "epoch": 1.2202632532302862,
      "grad_norm": 0.1636662632226944,
      "learning_rate": 5.9327751709755575e-05,
      "loss": 0.6363,
      "mean_token_accuracy": 0.8050144016742706,
      "num_tokens": 423336492.0,
      "step": 13263
    },
    {
      "entropy": 0.5832020714879036,
      "epoch": 1.2203552596017273,
      "grad_norm": 0.15882118046283722,
      "learning_rate": 5.932468488361394e-05,
      "loss": 0.5509,
      "mean_token_accuracy": 0.8329656608402729,
      "num_tokens": 423368561.0,
      "step": 13264
    },
    {
      "entropy": 0.6571524441242218,
      "epoch": 1.2204472659731687,
      "grad_norm": 0.1546737551689148,
      "learning_rate": 5.932161805747233e-05,
      "loss": 0.6537,
      "mean_token_accuracy": 0.806542482227087,
      "num_tokens": 423399787.0,
      "step": 13265
    },
    {
      "entropy": 0.7110765445977449,
      "epoch": 1.2205392723446098,
      "grad_norm": 0.16509751975536346,
      "learning_rate": 5.93185512313307e-05,
      "loss": 0.6833,
      "mean_token_accuracy": 0.7908631414175034,
      "num_tokens": 423431188.0,
      "step": 13266
    },
    {
      "entropy": 0.6630082316696644,
      "epoch": 1.2206312787160511,
      "grad_norm": 0.16446444392204285,
      "learning_rate": 5.931548440518907e-05,
      "loss": 0.6595,
      "mean_token_accuracy": 0.7986820004880428,
      "num_tokens": 423463470.0,
      "step": 13267
    },
    {
      "entropy": 0.6154363257810473,
      "epoch": 1.2207232850874923,
      "grad_norm": 0.15435446798801422,
      "learning_rate": 5.931241757904744e-05,
      "loss": 0.6053,
      "mean_token_accuracy": 0.8097293451428413,
      "num_tokens": 423495379.0,
      "step": 13268
    },
    {
      "entropy": 0.6813861448317766,
      "epoch": 1.2208152914589334,
      "grad_norm": 0.15811850130558014,
      "learning_rate": 5.930935075290582e-05,
      "loss": 0.6667,
      "mean_token_accuracy": 0.7969845570623875,
      "num_tokens": 423527833.0,
      "step": 13269
    },
    {
      "entropy": 0.6247764499858022,
      "epoch": 1.2209072978303748,
      "grad_norm": 0.15718822181224823,
      "learning_rate": 5.93062839267642e-05,
      "loss": 0.6215,
      "mean_token_accuracy": 0.8105971701443195,
      "num_tokens": 423559879.0,
      "step": 13270
    },
    {
      "entropy": 0.6388178905472159,
      "epoch": 1.220999304201816,
      "grad_norm": 0.15834683179855347,
      "learning_rate": 5.9303217100622566e-05,
      "loss": 0.6494,
      "mean_token_accuracy": 0.8017720580101013,
      "num_tokens": 423591391.0,
      "step": 13271
    },
    {
      "entropy": 0.703703448176384,
      "epoch": 1.2210913105732573,
      "grad_norm": 0.16179794073104858,
      "learning_rate": 5.930015027448095e-05,
      "loss": 0.698,
      "mean_token_accuracy": 0.7832212932407856,
      "num_tokens": 423623539.0,
      "step": 13272
    },
    {
      "entropy": 0.6407062280923128,
      "epoch": 1.2211833169446984,
      "grad_norm": 0.15483751893043518,
      "learning_rate": 5.929708344833932e-05,
      "loss": 0.6258,
      "mean_token_accuracy": 0.8067551255226135,
      "num_tokens": 423655638.0,
      "step": 13273
    },
    {
      "entropy": 0.5788387283682823,
      "epoch": 1.2212753233161395,
      "grad_norm": 0.15763266384601593,
      "learning_rate": 5.929401662219769e-05,
      "loss": 0.5652,
      "mean_token_accuracy": 0.8233742788434029,
      "num_tokens": 423687197.0,
      "step": 13274
    },
    {
      "entropy": 0.5182957630604506,
      "epoch": 1.2213673296875809,
      "grad_norm": 0.15383364260196686,
      "learning_rate": 5.9290949796056064e-05,
      "loss": 0.5107,
      "mean_token_accuracy": 0.8432344384491444,
      "num_tokens": 423719176.0,
      "step": 13275
    },
    {
      "entropy": 0.5531255044043064,
      "epoch": 1.221459336059022,
      "grad_norm": 0.15275870263576508,
      "learning_rate": 5.9287882969914446e-05,
      "loss": 0.5398,
      "mean_token_accuracy": 0.8307754136621952,
      "num_tokens": 423751297.0,
      "step": 13276
    },
    {
      "entropy": 0.5827916823327541,
      "epoch": 1.2215513424304634,
      "grad_norm": 0.15564055740833282,
      "learning_rate": 5.9284816143772813e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.8195070140063763,
      "num_tokens": 423783520.0,
      "step": 13277
    },
    {
      "entropy": 0.6373172700405121,
      "epoch": 1.2216433488019045,
      "grad_norm": 0.16858592629432678,
      "learning_rate": 5.928174931763119e-05,
      "loss": 0.632,
      "mean_token_accuracy": 0.8067202642560005,
      "num_tokens": 423814803.0,
      "step": 13278
    },
    {
      "entropy": 0.6005971096456051,
      "epoch": 1.2217353551733456,
      "grad_norm": 0.15612612664699554,
      "learning_rate": 5.9278682491489556e-05,
      "loss": 0.5801,
      "mean_token_accuracy": 0.8188204914331436,
      "num_tokens": 423846504.0,
      "step": 13279
    },
    {
      "entropy": 0.7343875300139189,
      "epoch": 1.221827361544787,
      "grad_norm": 0.17205718159675598,
      "learning_rate": 5.927561566534794e-05,
      "loss": 0.7153,
      "mean_token_accuracy": 0.7797991782426834,
      "num_tokens": 423877855.0,
      "step": 13280
    },
    {
      "entropy": 0.6111818067729473,
      "epoch": 1.2219193679162281,
      "grad_norm": 0.15693014860153198,
      "learning_rate": 5.927254883920631e-05,
      "loss": 0.6041,
      "mean_token_accuracy": 0.8150610513985157,
      "num_tokens": 423910078.0,
      "step": 13281
    },
    {
      "entropy": 0.6725980285555124,
      "epoch": 1.2220113742876695,
      "grad_norm": 0.16279277205467224,
      "learning_rate": 5.926948201306468e-05,
      "loss": 0.6609,
      "mean_token_accuracy": 0.7990440912544727,
      "num_tokens": 423942104.0,
      "step": 13282
    },
    {
      "entropy": 0.6496825031936169,
      "epoch": 1.2221033806591106,
      "grad_norm": 0.15206550061702728,
      "learning_rate": 5.9266415186923055e-05,
      "loss": 0.6377,
      "mean_token_accuracy": 0.8079261817038059,
      "num_tokens": 423974280.0,
      "step": 13283
    },
    {
      "entropy": 0.5866645034402609,
      "epoch": 1.2221953870305517,
      "grad_norm": 0.15781864523887634,
      "learning_rate": 5.9263348360781436e-05,
      "loss": 0.5743,
      "mean_token_accuracy": 0.825641043484211,
      "num_tokens": 424006168.0,
      "step": 13284
    },
    {
      "entropy": 0.6561309872195125,
      "epoch": 1.222287393401993,
      "grad_norm": 0.16060130298137665,
      "learning_rate": 5.9260281534639804e-05,
      "loss": 0.6479,
      "mean_token_accuracy": 0.8011677786707878,
      "num_tokens": 424038263.0,
      "step": 13285
    },
    {
      "entropy": 0.6638198047876358,
      "epoch": 1.2223793997734342,
      "grad_norm": 0.16226515173912048,
      "learning_rate": 5.925721470849818e-05,
      "loss": 0.6423,
      "mean_token_accuracy": 0.8016659542918205,
      "num_tokens": 424069664.0,
      "step": 13286
    },
    {
      "entropy": 0.5951927993446589,
      "epoch": 1.2224714061448756,
      "grad_norm": 0.15549592673778534,
      "learning_rate": 5.9254147882356546e-05,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8255579806864262,
      "num_tokens": 424101602.0,
      "step": 13287
    },
    {
      "entropy": 0.6420920165255666,
      "epoch": 1.2225634125163167,
      "grad_norm": 0.15599846839904785,
      "learning_rate": 5.925108105621493e-05,
      "loss": 0.6409,
      "mean_token_accuracy": 0.8079594522714615,
      "num_tokens": 424133782.0,
      "step": 13288
    },
    {
      "entropy": 0.7115699192509055,
      "epoch": 1.2226554188877579,
      "grad_norm": 0.16599400341510773,
      "learning_rate": 5.92480142300733e-05,
      "loss": 0.7002,
      "mean_token_accuracy": 0.7890862710773945,
      "num_tokens": 424165287.0,
      "step": 13289
    },
    {
      "entropy": 0.6538106156513095,
      "epoch": 1.2227474252591992,
      "grad_norm": 0.15968981385231018,
      "learning_rate": 5.924494740393167e-05,
      "loss": 0.6517,
      "mean_token_accuracy": 0.7999206110835075,
      "num_tokens": 424198055.0,
      "step": 13290
    },
    {
      "entropy": 0.509689318947494,
      "epoch": 1.2228394316306403,
      "grad_norm": 0.15109454095363617,
      "learning_rate": 5.9241880577790045e-05,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8444932214915752,
      "num_tokens": 424230231.0,
      "step": 13291
    },
    {
      "entropy": 0.655466478317976,
      "epoch": 1.2229314380020817,
      "grad_norm": 0.15936696529388428,
      "learning_rate": 5.9238813751648426e-05,
      "loss": 0.6464,
      "mean_token_accuracy": 0.7997689284384251,
      "num_tokens": 424262224.0,
      "step": 13292
    },
    {
      "entropy": 0.6421069204807281,
      "epoch": 1.2230234443735228,
      "grad_norm": 0.15716524422168732,
      "learning_rate": 5.9235746925506794e-05,
      "loss": 0.6376,
      "mean_token_accuracy": 0.8082860633730888,
      "num_tokens": 424294367.0,
      "step": 13293
    },
    {
      "entropy": 0.712106860242784,
      "epoch": 1.223115450744964,
      "grad_norm": 0.16483370959758759,
      "learning_rate": 5.923268009936517e-05,
      "loss": 0.7086,
      "mean_token_accuracy": 0.7838376574218273,
      "num_tokens": 424326256.0,
      "step": 13294
    },
    {
      "entropy": 0.6938832672312856,
      "epoch": 1.2232074571164053,
      "grad_norm": 0.1662072241306305,
      "learning_rate": 5.922961327322354e-05,
      "loss": 0.6905,
      "mean_token_accuracy": 0.7902839370071888,
      "num_tokens": 424357790.0,
      "step": 13295
    },
    {
      "entropy": 0.591149341315031,
      "epoch": 1.2232994634878465,
      "grad_norm": 0.1515239030122757,
      "learning_rate": 5.922654644708192e-05,
      "loss": 0.566,
      "mean_token_accuracy": 0.8235814422369003,
      "num_tokens": 424389978.0,
      "step": 13296
    },
    {
      "entropy": 0.5949892178177834,
      "epoch": 1.2233914698592878,
      "grad_norm": 0.154016375541687,
      "learning_rate": 5.922347962094029e-05,
      "loss": 0.5655,
      "mean_token_accuracy": 0.8274503387510777,
      "num_tokens": 424421637.0,
      "step": 13297
    },
    {
      "entropy": 0.6292326506227255,
      "epoch": 1.223483476230729,
      "grad_norm": 0.1541510820388794,
      "learning_rate": 5.922041279479866e-05,
      "loss": 0.6007,
      "mean_token_accuracy": 0.8127983175218105,
      "num_tokens": 424453804.0,
      "step": 13298
    },
    {
      "entropy": 0.5210873689502478,
      "epoch": 1.22357548260217,
      "grad_norm": 0.16149000823497772,
      "learning_rate": 5.9217345968657035e-05,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8442587628960609,
      "num_tokens": 424485725.0,
      "step": 13299
    },
    {
      "entropy": 0.6080566989257932,
      "epoch": 1.2236674889736114,
      "grad_norm": 0.15156790614128113,
      "learning_rate": 5.921427914251542e-05,
      "loss": 0.5995,
      "mean_token_accuracy": 0.8163835592567921,
      "num_tokens": 424518314.0,
      "step": 13300
    },
    {
      "entropy": 0.6840624809265137,
      "epoch": 1.2237594953450526,
      "grad_norm": 0.1691131442785263,
      "learning_rate": 5.9211212316373785e-05,
      "loss": 0.6543,
      "mean_token_accuracy": 0.7983358204364777,
      "num_tokens": 424549119.0,
      "step": 13301
    },
    {
      "entropy": 0.6604460645467043,
      "epoch": 1.223851501716494,
      "grad_norm": 0.16297346353530884,
      "learning_rate": 5.920814549023216e-05,
      "loss": 0.6477,
      "mean_token_accuracy": 0.8056505434215069,
      "num_tokens": 424581423.0,
      "step": 13302
    },
    {
      "entropy": 0.5822379551827908,
      "epoch": 1.223943508087935,
      "grad_norm": 0.15241938829421997,
      "learning_rate": 5.920507866409054e-05,
      "loss": 0.575,
      "mean_token_accuracy": 0.8244989700615406,
      "num_tokens": 424613307.0,
      "step": 13303
    },
    {
      "entropy": 0.5679228929802775,
      "epoch": 1.2240355144593762,
      "grad_norm": 0.15793141722679138,
      "learning_rate": 5.920201183794891e-05,
      "loss": 0.5678,
      "mean_token_accuracy": 0.8227443993091583,
      "num_tokens": 424645384.0,
      "step": 13304
    },
    {
      "entropy": 0.6312033054418862,
      "epoch": 1.2241275208308176,
      "grad_norm": 0.15747016668319702,
      "learning_rate": 5.919894501180728e-05,
      "loss": 0.6236,
      "mean_token_accuracy": 0.810293011367321,
      "num_tokens": 424677673.0,
      "step": 13305
    },
    {
      "entropy": 0.6446540113538504,
      "epoch": 1.2242195272022587,
      "grad_norm": 0.16051417589187622,
      "learning_rate": 5.919587818566565e-05,
      "loss": 0.6466,
      "mean_token_accuracy": 0.8037542402744293,
      "num_tokens": 424709376.0,
      "step": 13306
    },
    {
      "entropy": 0.5704477392137051,
      "epoch": 1.2243115335737,
      "grad_norm": 0.1591799408197403,
      "learning_rate": 5.919281135952404e-05,
      "loss": 0.5699,
      "mean_token_accuracy": 0.8236147463321686,
      "num_tokens": 424742139.0,
      "step": 13307
    },
    {
      "entropy": 0.7185067683458328,
      "epoch": 1.2244035399451412,
      "grad_norm": 0.16554509103298187,
      "learning_rate": 5.918974453338241e-05,
      "loss": 0.7186,
      "mean_token_accuracy": 0.7821395434439182,
      "num_tokens": 424774304.0,
      "step": 13308
    },
    {
      "entropy": 0.6679869899526238,
      "epoch": 1.2244955463165823,
      "grad_norm": 0.15626779198646545,
      "learning_rate": 5.918667770724078e-05,
      "loss": 0.6568,
      "mean_token_accuracy": 0.7991255037486553,
      "num_tokens": 424806193.0,
      "step": 13309
    },
    {
      "entropy": 0.6482338048517704,
      "epoch": 1.2245875526880237,
      "grad_norm": 0.15557511150836945,
      "learning_rate": 5.918361088109915e-05,
      "loss": 0.6364,
      "mean_token_accuracy": 0.8059840649366379,
      "num_tokens": 424838926.0,
      "step": 13310
    },
    {
      "entropy": 0.6260678917169571,
      "epoch": 1.2246795590594648,
      "grad_norm": 0.15879209339618683,
      "learning_rate": 5.918054405495753e-05,
      "loss": 0.6246,
      "mean_token_accuracy": 0.8151314668357372,
      "num_tokens": 424870294.0,
      "step": 13311
    },
    {
      "entropy": 0.7280023209750652,
      "epoch": 1.2247715654309062,
      "grad_norm": 0.16401898860931396,
      "learning_rate": 5.9177477228815906e-05,
      "loss": 0.7148,
      "mean_token_accuracy": 0.7866154052317142,
      "num_tokens": 424902270.0,
      "step": 13312
    },
    {
      "entropy": 0.5865366412326694,
      "epoch": 1.2248635718023473,
      "grad_norm": 0.1552533358335495,
      "learning_rate": 5.9174410402674274e-05,
      "loss": 0.5663,
      "mean_token_accuracy": 0.8225958533585072,
      "num_tokens": 424934201.0,
      "step": 13313
    },
    {
      "entropy": 0.630867188796401,
      "epoch": 1.2249555781737884,
      "grad_norm": 0.1584150493144989,
      "learning_rate": 5.917134357653265e-05,
      "loss": 0.6332,
      "mean_token_accuracy": 0.801986251026392,
      "num_tokens": 424966035.0,
      "step": 13314
    },
    {
      "entropy": 0.616697751916945,
      "epoch": 1.2250475845452298,
      "grad_norm": 0.15613235533237457,
      "learning_rate": 5.916827675039103e-05,
      "loss": 0.595,
      "mean_token_accuracy": 0.8158542513847351,
      "num_tokens": 424997866.0,
      "step": 13315
    },
    {
      "entropy": 0.617148666176945,
      "epoch": 1.225139590916671,
      "grad_norm": 0.15616922080516815,
      "learning_rate": 5.91652099242494e-05,
      "loss": 0.5921,
      "mean_token_accuracy": 0.8161198571324348,
      "num_tokens": 425029757.0,
      "step": 13316
    },
    {
      "entropy": 0.6369333807379007,
      "epoch": 1.2252315972881123,
      "grad_norm": 0.15592791140079498,
      "learning_rate": 5.916214309810777e-05,
      "loss": 0.6275,
      "mean_token_accuracy": 0.8124227672815323,
      "num_tokens": 425062496.0,
      "step": 13317
    },
    {
      "entropy": 0.5576633121818304,
      "epoch": 1.2253236036595534,
      "grad_norm": 0.1559525728225708,
      "learning_rate": 5.915907627196614e-05,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8353893309831619,
      "num_tokens": 425094916.0,
      "step": 13318
    },
    {
      "entropy": 0.7277183569967747,
      "epoch": 1.2254156100309945,
      "grad_norm": 0.1634381115436554,
      "learning_rate": 5.915600944582452e-05,
      "loss": 0.7142,
      "mean_token_accuracy": 0.784584853798151,
      "num_tokens": 425126990.0,
      "step": 13319
    },
    {
      "entropy": 0.5925385016016662,
      "epoch": 1.2255076164024359,
      "grad_norm": 0.16518183052539825,
      "learning_rate": 5.9152942619682896e-05,
      "loss": 0.5709,
      "mean_token_accuracy": 0.8225282579660416,
      "num_tokens": 425158737.0,
      "step": 13320
    },
    {
      "entropy": 0.7147017642855644,
      "epoch": 1.225599622773877,
      "grad_norm": 0.17119652032852173,
      "learning_rate": 5.9149875793541264e-05,
      "loss": 0.7113,
      "mean_token_accuracy": 0.7818463258445263,
      "num_tokens": 425191053.0,
      "step": 13321
    },
    {
      "entropy": 0.605428060516715,
      "epoch": 1.2256916291453184,
      "grad_norm": 0.160367950797081,
      "learning_rate": 5.914680896739964e-05,
      "loss": 0.5986,
      "mean_token_accuracy": 0.8168773911893368,
      "num_tokens": 425222892.0,
      "step": 13322
    },
    {
      "entropy": 0.5328482538461685,
      "epoch": 1.2257836355167595,
      "grad_norm": 0.15394939482212067,
      "learning_rate": 5.914374214125802e-05,
      "loss": 0.5221,
      "mean_token_accuracy": 0.8365905843675137,
      "num_tokens": 425255016.0,
      "step": 13323
    },
    {
      "entropy": 0.6407016646116972,
      "epoch": 1.2258756418882006,
      "grad_norm": 0.15582320094108582,
      "learning_rate": 5.914067531511639e-05,
      "loss": 0.6321,
      "mean_token_accuracy": 0.79974365234375,
      "num_tokens": 425286574.0,
      "step": 13324
    },
    {
      "entropy": 0.6724757570773363,
      "epoch": 1.225967648259642,
      "grad_norm": 0.16080127656459808,
      "learning_rate": 5.913760848897476e-05,
      "loss": 0.6602,
      "mean_token_accuracy": 0.7975413277745247,
      "num_tokens": 425317702.0,
      "step": 13325
    },
    {
      "entropy": 0.712267505005002,
      "epoch": 1.2260596546310831,
      "grad_norm": 0.16733334958553314,
      "learning_rate": 5.913454166283313e-05,
      "loss": 0.7073,
      "mean_token_accuracy": 0.7884644903242588,
      "num_tokens": 425350438.0,
      "step": 13326
    },
    {
      "entropy": 0.5657169595360756,
      "epoch": 1.2261516610025245,
      "grad_norm": 0.15143859386444092,
      "learning_rate": 5.913147483669151e-05,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8287386745214462,
      "num_tokens": 425382875.0,
      "step": 13327
    },
    {
      "entropy": 0.7421667128801346,
      "epoch": 1.2262436673739656,
      "grad_norm": 0.16227975487709045,
      "learning_rate": 5.912840801054989e-05,
      "loss": 0.7207,
      "mean_token_accuracy": 0.7774351835250854,
      "num_tokens": 425414209.0,
      "step": 13328
    },
    {
      "entropy": 0.5151744782924652,
      "epoch": 1.2263356737454068,
      "grad_norm": 0.15794603526592255,
      "learning_rate": 5.9125341184408254e-05,
      "loss": 0.5077,
      "mean_token_accuracy": 0.8426604680716991,
      "num_tokens": 425445904.0,
      "step": 13329
    },
    {
      "entropy": 0.5242103096097708,
      "epoch": 1.226427680116848,
      "grad_norm": 0.15942376852035522,
      "learning_rate": 5.912227435826663e-05,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8446247912943363,
      "num_tokens": 425477522.0,
      "step": 13330
    },
    {
      "entropy": 0.6315458128228784,
      "epoch": 1.2265196864882892,
      "grad_norm": 0.17078381776809692,
      "learning_rate": 5.911920753212501e-05,
      "loss": 0.6262,
      "mean_token_accuracy": 0.8083125725388527,
      "num_tokens": 425508984.0,
      "step": 13331
    },
    {
      "entropy": 0.6080804672092199,
      "epoch": 1.2266116928597306,
      "grad_norm": 0.16085967421531677,
      "learning_rate": 5.911614070598338e-05,
      "loss": 0.6016,
      "mean_token_accuracy": 0.8140720874071121,
      "num_tokens": 425541029.0,
      "step": 13332
    },
    {
      "entropy": 0.6671762000769377,
      "epoch": 1.2267036992311717,
      "grad_norm": 0.16012664139270782,
      "learning_rate": 5.911307387984175e-05,
      "loss": 0.6455,
      "mean_token_accuracy": 0.7963589653372765,
      "num_tokens": 425572422.0,
      "step": 13333
    },
    {
      "entropy": 0.6727479537948966,
      "epoch": 1.2267957056026129,
      "grad_norm": 0.16353200376033783,
      "learning_rate": 5.9110007053700135e-05,
      "loss": 0.6637,
      "mean_token_accuracy": 0.7944327183067799,
      "num_tokens": 425604145.0,
      "step": 13334
    },
    {
      "entropy": 0.7255304511636496,
      "epoch": 1.2268877119740542,
      "grad_norm": 0.17065928876399994,
      "learning_rate": 5.91069402275585e-05,
      "loss": 0.7176,
      "mean_token_accuracy": 0.7861846499145031,
      "num_tokens": 425635460.0,
      "step": 13335
    },
    {
      "entropy": 0.6721388194710016,
      "epoch": 1.2269797183454954,
      "grad_norm": 0.15175551176071167,
      "learning_rate": 5.910387340141688e-05,
      "loss": 0.6509,
      "mean_token_accuracy": 0.802678607404232,
      "num_tokens": 425667961.0,
      "step": 13336
    },
    {
      "entropy": 0.7160092443227768,
      "epoch": 1.2270717247169367,
      "grad_norm": 0.16596639156341553,
      "learning_rate": 5.9100806575275245e-05,
      "loss": 0.6955,
      "mean_token_accuracy": 0.7871794700622559,
      "num_tokens": 425699685.0,
      "step": 13337
    },
    {
      "entropy": 0.6388903446495533,
      "epoch": 1.2271637310883778,
      "grad_norm": 0.1592939794063568,
      "learning_rate": 5.9097739749133626e-05,
      "loss": 0.6093,
      "mean_token_accuracy": 0.8132065758109093,
      "num_tokens": 425731414.0,
      "step": 13338
    },
    {
      "entropy": 0.6156607698649168,
      "epoch": 1.227255737459819,
      "grad_norm": 0.1550336480140686,
      "learning_rate": 5.9094672922992e-05,
      "loss": 0.6021,
      "mean_token_accuracy": 0.8190678209066391,
      "num_tokens": 425763706.0,
      "step": 13339
    },
    {
      "entropy": 0.6066476292908192,
      "epoch": 1.2273477438312603,
      "grad_norm": 0.15787625312805176,
      "learning_rate": 5.909160609685037e-05,
      "loss": 0.5894,
      "mean_token_accuracy": 0.821702629327774,
      "num_tokens": 425795978.0,
      "step": 13340
    },
    {
      "entropy": 0.6392934024333954,
      "epoch": 1.2274397502027015,
      "grad_norm": 0.160056471824646,
      "learning_rate": 5.9088539270708743e-05,
      "loss": 0.6256,
      "mean_token_accuracy": 0.8096638880670071,
      "num_tokens": 425828219.0,
      "step": 13341
    },
    {
      "entropy": 0.580087929032743,
      "epoch": 1.2275317565741428,
      "grad_norm": 0.16108758747577667,
      "learning_rate": 5.9085472444567125e-05,
      "loss": 0.5769,
      "mean_token_accuracy": 0.8224281519651413,
      "num_tokens": 425860207.0,
      "step": 13342
    },
    {
      "entropy": 0.6850283723324537,
      "epoch": 1.227623762945584,
      "grad_norm": 0.15674585103988647,
      "learning_rate": 5.908240561842549e-05,
      "loss": 0.6918,
      "mean_token_accuracy": 0.7901696600019932,
      "num_tokens": 425892408.0,
      "step": 13343
    },
    {
      "entropy": 0.5963471373543143,
      "epoch": 1.227715769317025,
      "grad_norm": 0.16165992617607117,
      "learning_rate": 5.907933879228387e-05,
      "loss": 0.5983,
      "mean_token_accuracy": 0.8193343542516232,
      "num_tokens": 425924529.0,
      "step": 13344
    },
    {
      "entropy": 0.7035087645053864,
      "epoch": 1.2278077756884664,
      "grad_norm": 0.16333112120628357,
      "learning_rate": 5.9076271966142235e-05,
      "loss": 0.7037,
      "mean_token_accuracy": 0.7896149009466171,
      "num_tokens": 425956273.0,
      "step": 13345
    },
    {
      "entropy": 0.5357664204202592,
      "epoch": 1.2278997820599076,
      "grad_norm": 0.15413935482501984,
      "learning_rate": 5.9073205140000624e-05,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8394184224307537,
      "num_tokens": 425987656.0,
      "step": 13346
    },
    {
      "entropy": 0.6357546499930322,
      "epoch": 1.227991788431349,
      "grad_norm": 0.15975087881088257,
      "learning_rate": 5.907013831385899e-05,
      "loss": 0.6408,
      "mean_token_accuracy": 0.8081224225461483,
      "num_tokens": 426020014.0,
      "step": 13347
    },
    {
      "entropy": 0.6448170784860849,
      "epoch": 1.22808379480279,
      "grad_norm": 0.16045024991035461,
      "learning_rate": 5.906707148771736e-05,
      "loss": 0.6503,
      "mean_token_accuracy": 0.8024361580610275,
      "num_tokens": 426051338.0,
      "step": 13348
    },
    {
      "entropy": 0.6408550888299942,
      "epoch": 1.2281758011742312,
      "grad_norm": 0.18761789798736572,
      "learning_rate": 5.9064004661575734e-05,
      "loss": 0.6277,
      "mean_token_accuracy": 0.8049575239419937,
      "num_tokens": 426082548.0,
      "step": 13349
    },
    {
      "entropy": 0.6361253252252936,
      "epoch": 1.2282678075456726,
      "grad_norm": 0.1514665186405182,
      "learning_rate": 5.9060937835434115e-05,
      "loss": 0.61,
      "mean_token_accuracy": 0.8129489533603191,
      "num_tokens": 426114623.0,
      "step": 13350
    },
    {
      "entropy": 0.7049540914595127,
      "epoch": 1.2283598139171137,
      "grad_norm": 0.1613485962152481,
      "learning_rate": 5.905787100929249e-05,
      "loss": 0.6918,
      "mean_token_accuracy": 0.7906777039170265,
      "num_tokens": 426146189.0,
      "step": 13351
    },
    {
      "entropy": 0.5962607031688094,
      "epoch": 1.228451820288555,
      "grad_norm": 0.15101967751979828,
      "learning_rate": 5.905480418315086e-05,
      "loss": 0.5855,
      "mean_token_accuracy": 0.8202161304652691,
      "num_tokens": 426178136.0,
      "step": 13352
    },
    {
      "entropy": 0.6235365602187812,
      "epoch": 1.2285438266599962,
      "grad_norm": 0.15396109223365784,
      "learning_rate": 5.905173735700923e-05,
      "loss": 0.5942,
      "mean_token_accuracy": 0.8163008391857147,
      "num_tokens": 426210399.0,
      "step": 13353
    },
    {
      "entropy": 0.5649724560789764,
      "epoch": 1.2286358330314373,
      "grad_norm": 0.15346978604793549,
      "learning_rate": 5.9048670530867614e-05,
      "loss": 0.545,
      "mean_token_accuracy": 0.8327162824571133,
      "num_tokens": 426242369.0,
      "step": 13354
    },
    {
      "entropy": 0.6138121178373694,
      "epoch": 1.2287278394028787,
      "grad_norm": 0.15507015585899353,
      "learning_rate": 5.904560370472598e-05,
      "loss": 0.5846,
      "mean_token_accuracy": 0.8159619122743607,
      "num_tokens": 426273627.0,
      "step": 13355
    },
    {
      "entropy": 0.6994784139096737,
      "epoch": 1.2288198457743198,
      "grad_norm": 0.16331180930137634,
      "learning_rate": 5.9042536878584356e-05,
      "loss": 0.704,
      "mean_token_accuracy": 0.7885402888059616,
      "num_tokens": 426305192.0,
      "step": 13356
    },
    {
      "entropy": 0.5847479924559593,
      "epoch": 1.2289118521457612,
      "grad_norm": 0.1534118950366974,
      "learning_rate": 5.9039470052442724e-05,
      "loss": 0.5706,
      "mean_token_accuracy": 0.8240877687931061,
      "num_tokens": 426337255.0,
      "step": 13357
    },
    {
      "entropy": 0.5608043968677521,
      "epoch": 1.2290038585172023,
      "grad_norm": 0.1574990600347519,
      "learning_rate": 5.9036403226301106e-05,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8366935849189758,
      "num_tokens": 426369412.0,
      "step": 13358
    },
    {
      "entropy": 0.654835844412446,
      "epoch": 1.2290958648886434,
      "grad_norm": 0.15898708999156952,
      "learning_rate": 5.903333640015948e-05,
      "loss": 0.6472,
      "mean_token_accuracy": 0.80640684440732,
      "num_tokens": 426401527.0,
      "step": 13359
    },
    {
      "entropy": 0.6058022957295179,
      "epoch": 1.2291878712600848,
      "grad_norm": 0.1632753163576126,
      "learning_rate": 5.903026957401785e-05,
      "loss": 0.6126,
      "mean_token_accuracy": 0.8139821961522102,
      "num_tokens": 426433355.0,
      "step": 13360
    },
    {
      "entropy": 0.6849205773323774,
      "epoch": 1.229279877631526,
      "grad_norm": 0.1627674698829651,
      "learning_rate": 5.902720274787622e-05,
      "loss": 0.6726,
      "mean_token_accuracy": 0.7937182225286961,
      "num_tokens": 426465198.0,
      "step": 13361
    },
    {
      "entropy": 0.6311476640403271,
      "epoch": 1.2293718840029673,
      "grad_norm": 0.15702849626541138,
      "learning_rate": 5.9024135921734604e-05,
      "loss": 0.6273,
      "mean_token_accuracy": 0.8068927861750126,
      "num_tokens": 426497227.0,
      "step": 13362
    },
    {
      "entropy": 0.5852786004543304,
      "epoch": 1.2294638903744084,
      "grad_norm": 0.1542612761259079,
      "learning_rate": 5.902106909559297e-05,
      "loss": 0.5747,
      "mean_token_accuracy": 0.8233715631067753,
      "num_tokens": 426529905.0,
      "step": 13363
    },
    {
      "entropy": 0.6928582396358252,
      "epoch": 1.2295558967458495,
      "grad_norm": 0.16771124303340912,
      "learning_rate": 5.901800226945135e-05,
      "loss": 0.6715,
      "mean_token_accuracy": 0.7954389974474907,
      "num_tokens": 426561461.0,
      "step": 13364
    },
    {
      "entropy": 0.677357817068696,
      "epoch": 1.2296479031172909,
      "grad_norm": 0.16103987395763397,
      "learning_rate": 5.901493544330973e-05,
      "loss": 0.6635,
      "mean_token_accuracy": 0.7942529171705246,
      "num_tokens": 426593163.0,
      "step": 13365
    },
    {
      "entropy": 0.7107107155025005,
      "epoch": 1.229739909488732,
      "grad_norm": 0.1655282825231552,
      "learning_rate": 5.9011868617168096e-05,
      "loss": 0.6985,
      "mean_token_accuracy": 0.7852502688765526,
      "num_tokens": 426624915.0,
      "step": 13366
    },
    {
      "entropy": 0.5517902337014675,
      "epoch": 1.2298319158601734,
      "grad_norm": 0.15202024579048157,
      "learning_rate": 5.900880179102647e-05,
      "loss": 0.5317,
      "mean_token_accuracy": 0.8354974426329136,
      "num_tokens": 426656853.0,
      "step": 13367
    },
    {
      "entropy": 0.5638036672025919,
      "epoch": 1.2299239222316145,
      "grad_norm": 0.15200500190258026,
      "learning_rate": 5.900573496488484e-05,
      "loss": 0.543,
      "mean_token_accuracy": 0.828588355332613,
      "num_tokens": 426688495.0,
      "step": 13368
    },
    {
      "entropy": 0.6123722027987242,
      "epoch": 1.2300159286030556,
      "grad_norm": 0.15515395998954773,
      "learning_rate": 5.900266813874322e-05,
      "loss": 0.6048,
      "mean_token_accuracy": 0.8165422081947327,
      "num_tokens": 426720650.0,
      "step": 13369
    },
    {
      "entropy": 0.7282486734911799,
      "epoch": 1.230107934974497,
      "grad_norm": 0.16943633556365967,
      "learning_rate": 5.8999601312601595e-05,
      "loss": 0.7409,
      "mean_token_accuracy": 0.781453225761652,
      "num_tokens": 426752238.0,
      "step": 13370
    },
    {
      "entropy": 0.641569621860981,
      "epoch": 1.2301999413459381,
      "grad_norm": 0.1601875275373459,
      "learning_rate": 5.899653448645996e-05,
      "loss": 0.6353,
      "mean_token_accuracy": 0.8070286363363266,
      "num_tokens": 426784451.0,
      "step": 13371
    },
    {
      "entropy": 0.5546399550512433,
      "epoch": 1.2302919477173795,
      "grad_norm": 0.14879170060157776,
      "learning_rate": 5.899346766031834e-05,
      "loss": 0.5382,
      "mean_token_accuracy": 0.833292480558157,
      "num_tokens": 426815724.0,
      "step": 13372
    },
    {
      "entropy": 0.6534878313541412,
      "epoch": 1.2303839540888206,
      "grad_norm": 0.16082307696342468,
      "learning_rate": 5.899040083417672e-05,
      "loss": 0.6404,
      "mean_token_accuracy": 0.8054520301520824,
      "num_tokens": 426847265.0,
      "step": 13373
    },
    {
      "entropy": 0.6118086706846952,
      "epoch": 1.2304759604602618,
      "grad_norm": 0.1539972424507141,
      "learning_rate": 5.8987334008035087e-05,
      "loss": 0.593,
      "mean_token_accuracy": 0.8168966770172119,
      "num_tokens": 426880033.0,
      "step": 13374
    },
    {
      "entropy": 0.5712413275614381,
      "epoch": 1.230567966831703,
      "grad_norm": 0.15226846933364868,
      "learning_rate": 5.898426718189346e-05,
      "loss": 0.559,
      "mean_token_accuracy": 0.828665878623724,
      "num_tokens": 426912716.0,
      "step": 13375
    },
    {
      "entropy": 0.6920671174302697,
      "epoch": 1.2306599732031442,
      "grad_norm": 0.16717354953289032,
      "learning_rate": 5.898120035575183e-05,
      "loss": 0.6742,
      "mean_token_accuracy": 0.7904736399650574,
      "num_tokens": 426944336.0,
      "step": 13376
    },
    {
      "entropy": 0.5485686613246799,
      "epoch": 1.2307519795745856,
      "grad_norm": 0.149138942360878,
      "learning_rate": 5.897813352961021e-05,
      "loss": 0.5218,
      "mean_token_accuracy": 0.8299591578543186,
      "num_tokens": 426975604.0,
      "step": 13377
    },
    {
      "entropy": 0.6474285125732422,
      "epoch": 1.2308439859460267,
      "grad_norm": 0.16157355904579163,
      "learning_rate": 5.8975066703468585e-05,
      "loss": 0.6404,
      "mean_token_accuracy": 0.8050435557961464,
      "num_tokens": 427007646.0,
      "step": 13378
    },
    {
      "entropy": 0.5441933227702975,
      "epoch": 1.2309359923174679,
      "grad_norm": 0.1448923498392105,
      "learning_rate": 5.897199987732695e-05,
      "loss": 0.5275,
      "mean_token_accuracy": 0.8336702026426792,
      "num_tokens": 427039830.0,
      "step": 13379
    },
    {
      "entropy": 0.5837899108882993,
      "epoch": 1.2310279986889092,
      "grad_norm": 0.1630827635526657,
      "learning_rate": 5.896893305118533e-05,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8264226242899895,
      "num_tokens": 427072154.0,
      "step": 13380
    },
    {
      "entropy": 0.6489161662757397,
      "epoch": 1.2311200050603504,
      "grad_norm": 0.17024490237236023,
      "learning_rate": 5.896586622504371e-05,
      "loss": 0.63,
      "mean_token_accuracy": 0.8099381662905216,
      "num_tokens": 427104750.0,
      "step": 13381
    },
    {
      "entropy": 0.6255656601861119,
      "epoch": 1.2312120114317917,
      "grad_norm": 0.16282996535301208,
      "learning_rate": 5.896279939890208e-05,
      "loss": 0.6131,
      "mean_token_accuracy": 0.814533032476902,
      "num_tokens": 427136891.0,
      "step": 13382
    },
    {
      "entropy": 0.6081263720989227,
      "epoch": 1.2313040178032328,
      "grad_norm": 0.15642359852790833,
      "learning_rate": 5.895973257276045e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8175378926098347,
      "num_tokens": 427168272.0,
      "step": 13383
    },
    {
      "entropy": 0.6317267213016748,
      "epoch": 1.231396024174674,
      "grad_norm": 0.16005392372608185,
      "learning_rate": 5.895666574661882e-05,
      "loss": 0.6138,
      "mean_token_accuracy": 0.809424877166748,
      "num_tokens": 427200333.0,
      "step": 13384
    },
    {
      "entropy": 0.6076961811631918,
      "epoch": 1.2314880305461153,
      "grad_norm": 0.16352249681949615,
      "learning_rate": 5.895359892047721e-05,
      "loss": 0.6035,
      "mean_token_accuracy": 0.8102075718343258,
      "num_tokens": 427232903.0,
      "step": 13385
    },
    {
      "entropy": 0.6812859512865543,
      "epoch": 1.2315800369175565,
      "grad_norm": 0.1632111519575119,
      "learning_rate": 5.8950532094335576e-05,
      "loss": 0.6788,
      "mean_token_accuracy": 0.7908568792045116,
      "num_tokens": 427264764.0,
      "step": 13386
    },
    {
      "entropy": 0.5922729410231113,
      "epoch": 1.2316720432889978,
      "grad_norm": 0.15526621043682098,
      "learning_rate": 5.8947465268193943e-05,
      "loss": 0.5741,
      "mean_token_accuracy": 0.8206134997308254,
      "num_tokens": 427296471.0,
      "step": 13387
    },
    {
      "entropy": 0.6697286628186703,
      "epoch": 1.231764049660439,
      "grad_norm": 0.1571321338415146,
      "learning_rate": 5.894439844205232e-05,
      "loss": 0.6656,
      "mean_token_accuracy": 0.7992173507809639,
      "num_tokens": 427328355.0,
      "step": 13388
    },
    {
      "entropy": 0.5421639811247587,
      "epoch": 1.23185605603188,
      "grad_norm": 0.15129193663597107,
      "learning_rate": 5.89413316159107e-05,
      "loss": 0.5292,
      "mean_token_accuracy": 0.8400687500834465,
      "num_tokens": 427360151.0,
      "step": 13389
    },
    {
      "entropy": 0.551855793222785,
      "epoch": 1.2319480624033214,
      "grad_norm": 0.15533652901649475,
      "learning_rate": 5.8938264789769074e-05,
      "loss": 0.5477,
      "mean_token_accuracy": 0.8330289721488953,
      "num_tokens": 427392574.0,
      "step": 13390
    },
    {
      "entropy": 0.6398058831691742,
      "epoch": 1.2320400687747626,
      "grad_norm": 0.1638793796300888,
      "learning_rate": 5.893519796362744e-05,
      "loss": 0.6327,
      "mean_token_accuracy": 0.8064146637916565,
      "num_tokens": 427424534.0,
      "step": 13391
    },
    {
      "entropy": 0.5811767689883709,
      "epoch": 1.232132075146204,
      "grad_norm": 0.15566757321357727,
      "learning_rate": 5.8932131137485823e-05,
      "loss": 0.5715,
      "mean_token_accuracy": 0.8231801763176918,
      "num_tokens": 427455474.0,
      "step": 13392
    },
    {
      "entropy": 0.5370157277211547,
      "epoch": 1.232224081517645,
      "grad_norm": 0.15402750670909882,
      "learning_rate": 5.89290643113442e-05,
      "loss": 0.5126,
      "mean_token_accuracy": 0.8442453481256962,
      "num_tokens": 427486392.0,
      "step": 13393
    },
    {
      "entropy": 0.5986258676275611,
      "epoch": 1.2323160878890862,
      "grad_norm": 0.1559285968542099,
      "learning_rate": 5.8925997485202566e-05,
      "loss": 0.5858,
      "mean_token_accuracy": 0.82054078951478,
      "num_tokens": 427518592.0,
      "step": 13394
    },
    {
      "entropy": 0.5570327062159777,
      "epoch": 1.2324080942605276,
      "grad_norm": 0.16505245864391327,
      "learning_rate": 5.892293065906094e-05,
      "loss": 0.547,
      "mean_token_accuracy": 0.8328560143709183,
      "num_tokens": 427549448.0,
      "step": 13395
    },
    {
      "entropy": 0.6364135071635246,
      "epoch": 1.2325001006319687,
      "grad_norm": 0.15946276485919952,
      "learning_rate": 5.891986383291932e-05,
      "loss": 0.6398,
      "mean_token_accuracy": 0.8053694069385529,
      "num_tokens": 427581085.0,
      "step": 13396
    },
    {
      "entropy": 0.6437008082866669,
      "epoch": 1.23259210700341,
      "grad_norm": 0.17442260682582855,
      "learning_rate": 5.891679700677769e-05,
      "loss": 0.6526,
      "mean_token_accuracy": 0.8046580702066422,
      "num_tokens": 427612880.0,
      "step": 13397
    },
    {
      "entropy": 0.6567103583365679,
      "epoch": 1.2326841133748512,
      "grad_norm": 0.1602054238319397,
      "learning_rate": 5.8913730180636065e-05,
      "loss": 0.6495,
      "mean_token_accuracy": 0.8016152791678905,
      "num_tokens": 427644677.0,
      "step": 13398
    },
    {
      "entropy": 0.6793159395456314,
      "epoch": 1.2327761197462923,
      "grad_norm": 0.15781457722187042,
      "learning_rate": 5.891066335449443e-05,
      "loss": 0.6901,
      "mean_token_accuracy": 0.7942089289426804,
      "num_tokens": 427677328.0,
      "step": 13399
    },
    {
      "entropy": 0.6408419609069824,
      "epoch": 1.2328681261177337,
      "grad_norm": 0.16033349931240082,
      "learning_rate": 5.8907596528352814e-05,
      "loss": 0.6317,
      "mean_token_accuracy": 0.8058685064315796,
      "num_tokens": 427709109.0,
      "step": 13400
    },
    {
      "entropy": 0.7008167654275894,
      "epoch": 1.2329601324891748,
      "grad_norm": 0.16167660057544708,
      "learning_rate": 5.890452970221119e-05,
      "loss": 0.669,
      "mean_token_accuracy": 0.7948584817349911,
      "num_tokens": 427739988.0,
      "step": 13401
    },
    {
      "entropy": 0.6364569123834372,
      "epoch": 1.2330521388606162,
      "grad_norm": 0.1615370362997055,
      "learning_rate": 5.8901462876069556e-05,
      "loss": 0.6203,
      "mean_token_accuracy": 0.8084508590400219,
      "num_tokens": 427771894.0,
      "step": 13402
    },
    {
      "entropy": 0.5639573941007257,
      "epoch": 1.2331441452320573,
      "grad_norm": 0.16622672975063324,
      "learning_rate": 5.889839604992793e-05,
      "loss": 0.5556,
      "mean_token_accuracy": 0.8281468078494072,
      "num_tokens": 427803544.0,
      "step": 13403
    },
    {
      "entropy": 0.7297418583184481,
      "epoch": 1.2332361516034984,
      "grad_norm": 0.16418509185314178,
      "learning_rate": 5.889532922378631e-05,
      "loss": 0.7203,
      "mean_token_accuracy": 0.7822202369570732,
      "num_tokens": 427835301.0,
      "step": 13404
    },
    {
      "entropy": 0.6153154857456684,
      "epoch": 1.2333281579749398,
      "grad_norm": 0.1638842523097992,
      "learning_rate": 5.889226239764468e-05,
      "loss": 0.6078,
      "mean_token_accuracy": 0.8150176592171192,
      "num_tokens": 427867463.0,
      "step": 13405
    },
    {
      "entropy": 0.5948311076499522,
      "epoch": 1.233420164346381,
      "grad_norm": 0.15492583811283112,
      "learning_rate": 5.8889195571503055e-05,
      "loss": 0.572,
      "mean_token_accuracy": 0.8211899437010288,
      "num_tokens": 427899287.0,
      "step": 13406
    },
    {
      "entropy": 0.6639568619430065,
      "epoch": 1.2335121707178223,
      "grad_norm": 0.15321284532546997,
      "learning_rate": 5.888612874536142e-05,
      "loss": 0.6485,
      "mean_token_accuracy": 0.8008996918797493,
      "num_tokens": 427931141.0,
      "step": 13407
    },
    {
      "entropy": 0.5952378152869642,
      "epoch": 1.2336041770892634,
      "grad_norm": 0.15837092697620392,
      "learning_rate": 5.8883061919219804e-05,
      "loss": 0.5876,
      "mean_token_accuracy": 0.821228601038456,
      "num_tokens": 427962791.0,
      "step": 13408
    },
    {
      "entropy": 0.5730472784489393,
      "epoch": 1.2336961834607045,
      "grad_norm": 0.15889808535575867,
      "learning_rate": 5.887999509307818e-05,
      "loss": 0.5509,
      "mean_token_accuracy": 0.8294570222496986,
      "num_tokens": 427993954.0,
      "step": 13409
    },
    {
      "entropy": 0.6535204313695431,
      "epoch": 1.2337881898321459,
      "grad_norm": 0.15871579945087433,
      "learning_rate": 5.887692826693655e-05,
      "loss": 0.6525,
      "mean_token_accuracy": 0.7964617237448692,
      "num_tokens": 428025991.0,
      "step": 13410
    },
    {
      "entropy": 0.78292072750628,
      "epoch": 1.233880196203587,
      "grad_norm": 0.16531136631965637,
      "learning_rate": 5.887386144079492e-05,
      "loss": 0.7741,
      "mean_token_accuracy": 0.7668449468910694,
      "num_tokens": 428058119.0,
      "step": 13411
    },
    {
      "entropy": 0.7157217133790255,
      "epoch": 1.2339722025750284,
      "grad_norm": 0.16059309244155884,
      "learning_rate": 5.88707946146533e-05,
      "loss": 0.7092,
      "mean_token_accuracy": 0.7856179513037205,
      "num_tokens": 428090827.0,
      "step": 13412
    },
    {
      "entropy": 0.6306106336414814,
      "epoch": 1.2340642089464695,
      "grad_norm": 0.1547502726316452,
      "learning_rate": 5.886772778851167e-05,
      "loss": 0.6082,
      "mean_token_accuracy": 0.8132478073239326,
      "num_tokens": 428122681.0,
      "step": 13413
    },
    {
      "entropy": 0.5968686724081635,
      "epoch": 1.2341562153179106,
      "grad_norm": 0.1574159860610962,
      "learning_rate": 5.8864660962370045e-05,
      "loss": 0.5806,
      "mean_token_accuracy": 0.8223661594092846,
      "num_tokens": 428154982.0,
      "step": 13414
    },
    {
      "entropy": 0.5733635872602463,
      "epoch": 1.234248221689352,
      "grad_norm": 0.15315109491348267,
      "learning_rate": 5.886159413622841e-05,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8276055604219437,
      "num_tokens": 428186640.0,
      "step": 13415
    },
    {
      "entropy": 0.4284244840964675,
      "epoch": 1.2343402280607931,
      "grad_norm": 0.14110495150089264,
      "learning_rate": 5.8858527310086795e-05,
      "loss": 0.4047,
      "mean_token_accuracy": 0.8730539157986641,
      "num_tokens": 428218775.0,
      "step": 13416
    },
    {
      "entropy": 0.6498561315238476,
      "epoch": 1.2344322344322345,
      "grad_norm": 0.16389070451259613,
      "learning_rate": 5.885546048394517e-05,
      "loss": 0.6575,
      "mean_token_accuracy": 0.8046083487570286,
      "num_tokens": 428250472.0,
      "step": 13417
    },
    {
      "entropy": 0.6827135356143117,
      "epoch": 1.2345242408036756,
      "grad_norm": 0.16425132751464844,
      "learning_rate": 5.885239365780354e-05,
      "loss": 0.6677,
      "mean_token_accuracy": 0.7981232367455959,
      "num_tokens": 428282320.0,
      "step": 13418
    },
    {
      "entropy": 0.6045280378311872,
      "epoch": 1.2346162471751168,
      "grad_norm": 0.15002217888832092,
      "learning_rate": 5.884932683166191e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8203990906476974,
      "num_tokens": 428314324.0,
      "step": 13419
    },
    {
      "entropy": 0.6142421923577785,
      "epoch": 1.234708253546558,
      "grad_norm": 0.15734440088272095,
      "learning_rate": 5.884626000552029e-05,
      "loss": 0.5978,
      "mean_token_accuracy": 0.8177132196724415,
      "num_tokens": 428346286.0,
      "step": 13420
    },
    {
      "entropy": 0.5867367526516318,
      "epoch": 1.2348002599179992,
      "grad_norm": 0.15563786029815674,
      "learning_rate": 5.884319317937866e-05,
      "loss": 0.5748,
      "mean_token_accuracy": 0.8251416087150574,
      "num_tokens": 428378403.0,
      "step": 13421
    },
    {
      "entropy": 0.685653967782855,
      "epoch": 1.2348922662894406,
      "grad_norm": 0.16297423839569092,
      "learning_rate": 5.8840126353237036e-05,
      "loss": 0.6827,
      "mean_token_accuracy": 0.7950894385576248,
      "num_tokens": 428410308.0,
      "step": 13422
    },
    {
      "entropy": 0.6087547289207578,
      "epoch": 1.2349842726608817,
      "grad_norm": 0.16080696880817413,
      "learning_rate": 5.883705952709542e-05,
      "loss": 0.5859,
      "mean_token_accuracy": 0.8212596476078033,
      "num_tokens": 428442287.0,
      "step": 13423
    },
    {
      "entropy": 0.6213243789970875,
      "epoch": 1.2350762790323229,
      "grad_norm": 0.15535570681095123,
      "learning_rate": 5.8833992700953785e-05,
      "loss": 0.6173,
      "mean_token_accuracy": 0.8134682476520538,
      "num_tokens": 428474235.0,
      "step": 13424
    },
    {
      "entropy": 0.606918117031455,
      "epoch": 1.2351682854037642,
      "grad_norm": 0.1594035029411316,
      "learning_rate": 5.883092587481216e-05,
      "loss": 0.6078,
      "mean_token_accuracy": 0.8113574422895908,
      "num_tokens": 428505943.0,
      "step": 13425
    },
    {
      "entropy": 0.5537551711313426,
      "epoch": 1.2352602917752054,
      "grad_norm": 0.14994794130325317,
      "learning_rate": 5.882785904867053e-05,
      "loss": 0.5398,
      "mean_token_accuracy": 0.8369018509984016,
      "num_tokens": 428538251.0,
      "step": 13426
    },
    {
      "entropy": 0.5898460019379854,
      "epoch": 1.2353522981466467,
      "grad_norm": 0.1651187539100647,
      "learning_rate": 5.8824792222528916e-05,
      "loss": 0.5815,
      "mean_token_accuracy": 0.8248382024466991,
      "num_tokens": 428569706.0,
      "step": 13427
    },
    {
      "entropy": 0.6874966211616993,
      "epoch": 1.2354443045180878,
      "grad_norm": 0.16442926228046417,
      "learning_rate": 5.8821725396387284e-05,
      "loss": 0.6709,
      "mean_token_accuracy": 0.7965231202542782,
      "num_tokens": 428601164.0,
      "step": 13428
    },
    {
      "entropy": 0.6269114012829959,
      "epoch": 1.235536310889529,
      "grad_norm": 0.16209842264652252,
      "learning_rate": 5.881865857024566e-05,
      "loss": 0.6217,
      "mean_token_accuracy": 0.8113307729363441,
      "num_tokens": 428633253.0,
      "step": 13429
    },
    {
      "entropy": 0.5877078818157315,
      "epoch": 1.2356283172609703,
      "grad_norm": 0.15749339759349823,
      "learning_rate": 5.8815591744104026e-05,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8181546628475189,
      "num_tokens": 428665619.0,
      "step": 13430
    },
    {
      "entropy": 0.6448710830882192,
      "epoch": 1.2357203236324115,
      "grad_norm": 0.1613968163728714,
      "learning_rate": 5.881252491796241e-05,
      "loss": 0.6371,
      "mean_token_accuracy": 0.8045681975781918,
      "num_tokens": 428697559.0,
      "step": 13431
    },
    {
      "entropy": 0.5813704542815685,
      "epoch": 1.2358123300038528,
      "grad_norm": 0.14814673364162445,
      "learning_rate": 5.880945809182078e-05,
      "loss": 0.5696,
      "mean_token_accuracy": 0.8258935064077377,
      "num_tokens": 428730211.0,
      "step": 13432
    },
    {
      "entropy": 0.6281665274873376,
      "epoch": 1.235904336375294,
      "grad_norm": 0.1561862975358963,
      "learning_rate": 5.880639126567915e-05,
      "loss": 0.6188,
      "mean_token_accuracy": 0.8080056607723236,
      "num_tokens": 428762235.0,
      "step": 13433
    },
    {
      "entropy": 0.6228580344468355,
      "epoch": 1.235996342746735,
      "grad_norm": 0.1522064357995987,
      "learning_rate": 5.8803324439537525e-05,
      "loss": 0.6125,
      "mean_token_accuracy": 0.8112704940140247,
      "num_tokens": 428794856.0,
      "step": 13434
    },
    {
      "entropy": 0.6038763206452131,
      "epoch": 1.2360883491181764,
      "grad_norm": 0.16917534172534943,
      "learning_rate": 5.8800257613395906e-05,
      "loss": 0.5934,
      "mean_token_accuracy": 0.814390029758215,
      "num_tokens": 428827056.0,
      "step": 13435
    },
    {
      "entropy": 0.6157181784510612,
      "epoch": 1.2361803554896176,
      "grad_norm": 0.1603897213935852,
      "learning_rate": 5.8797190787254274e-05,
      "loss": 0.6069,
      "mean_token_accuracy": 0.8134528808295727,
      "num_tokens": 428858814.0,
      "step": 13436
    },
    {
      "entropy": 0.5979391690343618,
      "epoch": 1.236272361861059,
      "grad_norm": 0.15228715538978577,
      "learning_rate": 5.879412396111265e-05,
      "loss": 0.5855,
      "mean_token_accuracy": 0.8186348266899586,
      "num_tokens": 428890301.0,
      "step": 13437
    },
    {
      "entropy": 0.5802043909206986,
      "epoch": 1.2363643682325,
      "grad_norm": 0.1535598784685135,
      "learning_rate": 5.8791057134971017e-05,
      "loss": 0.5706,
      "mean_token_accuracy": 0.8219574838876724,
      "num_tokens": 428922355.0,
      "step": 13438
    },
    {
      "entropy": 0.6903489604592323,
      "epoch": 1.2364563746039412,
      "grad_norm": 0.16552825272083282,
      "learning_rate": 5.87879903088294e-05,
      "loss": 0.6844,
      "mean_token_accuracy": 0.7965548522770405,
      "num_tokens": 428953857.0,
      "step": 13439
    },
    {
      "entropy": 0.6793220490217209,
      "epoch": 1.2365483809753826,
      "grad_norm": 0.15532168745994568,
      "learning_rate": 5.878492348268777e-05,
      "loss": 0.6553,
      "mean_token_accuracy": 0.7984054014086723,
      "num_tokens": 428985334.0,
      "step": 13440
    },
    {
      "entropy": 0.48755977442488074,
      "epoch": 1.2366403873468237,
      "grad_norm": 0.1503746211528778,
      "learning_rate": 5.878185665654614e-05,
      "loss": 0.4632,
      "mean_token_accuracy": 0.855194266885519,
      "num_tokens": 429017096.0,
      "step": 13441
    },
    {
      "entropy": 0.7664265111088753,
      "epoch": 1.236732393718265,
      "grad_norm": 0.1839134842157364,
      "learning_rate": 5.8778789830404515e-05,
      "loss": 0.7722,
      "mean_token_accuracy": 0.7698239870369434,
      "num_tokens": 429048422.0,
      "step": 13442
    },
    {
      "entropy": 0.6720030680298805,
      "epoch": 1.2368244000897062,
      "grad_norm": 0.16844002902507782,
      "learning_rate": 5.8775723004262897e-05,
      "loss": 0.6689,
      "mean_token_accuracy": 0.7959161177277565,
      "num_tokens": 429080146.0,
      "step": 13443
    },
    {
      "entropy": 0.61621009465307,
      "epoch": 1.2369164064611473,
      "grad_norm": 0.14858682453632355,
      "learning_rate": 5.8772656178121264e-05,
      "loss": 0.6047,
      "mean_token_accuracy": 0.8154445551335812,
      "num_tokens": 429112381.0,
      "step": 13444
    },
    {
      "entropy": 0.5884604137390852,
      "epoch": 1.2370084128325887,
      "grad_norm": 0.15703490376472473,
      "learning_rate": 5.876958935197964e-05,
      "loss": 0.5705,
      "mean_token_accuracy": 0.8243416398763657,
      "num_tokens": 429144882.0,
      "step": 13445
    },
    {
      "entropy": 0.7382212169468403,
      "epoch": 1.2371004192040298,
      "grad_norm": 0.16359958052635193,
      "learning_rate": 5.876652252583801e-05,
      "loss": 0.7215,
      "mean_token_accuracy": 0.7818636298179626,
      "num_tokens": 429176331.0,
      "step": 13446
    },
    {
      "entropy": 0.6211536824703217,
      "epoch": 1.2371924255754712,
      "grad_norm": 0.15039919316768646,
      "learning_rate": 5.876345569969639e-05,
      "loss": 0.6024,
      "mean_token_accuracy": 0.8184629417955875,
      "num_tokens": 429208886.0,
      "step": 13447
    },
    {
      "entropy": 0.6394818816334009,
      "epoch": 1.2372844319469123,
      "grad_norm": 0.1498263031244278,
      "learning_rate": 5.876038887355476e-05,
      "loss": 0.61,
      "mean_token_accuracy": 0.8136590085923672,
      "num_tokens": 429241290.0,
      "step": 13448
    },
    {
      "entropy": 0.6410828968510032,
      "epoch": 1.2373764383183534,
      "grad_norm": 0.16106201708316803,
      "learning_rate": 5.875732204741313e-05,
      "loss": 0.6299,
      "mean_token_accuracy": 0.8064023852348328,
      "num_tokens": 429271859.0,
      "step": 13449
    },
    {
      "entropy": 0.5971753392368555,
      "epoch": 1.2374684446897948,
      "grad_norm": 0.15658250451087952,
      "learning_rate": 5.8754255221271506e-05,
      "loss": 0.591,
      "mean_token_accuracy": 0.8208846971392632,
      "num_tokens": 429303761.0,
      "step": 13450
    },
    {
      "entropy": 0.6655870713293552,
      "epoch": 1.237560451061236,
      "grad_norm": 0.1669778972864151,
      "learning_rate": 5.875118839512989e-05,
      "loss": 0.663,
      "mean_token_accuracy": 0.802204929292202,
      "num_tokens": 429335923.0,
      "step": 13451
    },
    {
      "entropy": 0.6755134668201208,
      "epoch": 1.2376524574326773,
      "grad_norm": 0.16663379967212677,
      "learning_rate": 5.8748121568988255e-05,
      "loss": 0.6598,
      "mean_token_accuracy": 0.8006080277264118,
      "num_tokens": 429367066.0,
      "step": 13452
    },
    {
      "entropy": 0.6869114227592945,
      "epoch": 1.2377444638041184,
      "grad_norm": 0.15935376286506653,
      "learning_rate": 5.874505474284663e-05,
      "loss": 0.6846,
      "mean_token_accuracy": 0.7967069633305073,
      "num_tokens": 429398944.0,
      "step": 13453
    },
    {
      "entropy": 0.4430586341768503,
      "epoch": 1.2378364701755595,
      "grad_norm": 0.14387726783752441,
      "learning_rate": 5.874198791670501e-05,
      "loss": 0.425,
      "mean_token_accuracy": 0.8665709421038628,
      "num_tokens": 429430954.0,
      "step": 13454
    },
    {
      "entropy": 0.5024450430646539,
      "epoch": 1.2379284765470009,
      "grad_norm": 0.15187714993953705,
      "learning_rate": 5.873892109056338e-05,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8467492870986462,
      "num_tokens": 429462999.0,
      "step": 13455
    },
    {
      "entropy": 0.525319155305624,
      "epoch": 1.238020482918442,
      "grad_norm": 0.16442181169986725,
      "learning_rate": 5.8735854264421753e-05,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8388881199061871,
      "num_tokens": 429494702.0,
      "step": 13456
    },
    {
      "entropy": 0.6814172305166721,
      "epoch": 1.2381124892898834,
      "grad_norm": 0.1648281216621399,
      "learning_rate": 5.873278743828012e-05,
      "loss": 0.6832,
      "mean_token_accuracy": 0.797504972666502,
      "num_tokens": 429526319.0,
      "step": 13457
    },
    {
      "entropy": 0.5915819238871336,
      "epoch": 1.2382044956613245,
      "grad_norm": 0.15487509965896606,
      "learning_rate": 5.87297206121385e-05,
      "loss": 0.5857,
      "mean_token_accuracy": 0.8224320933222771,
      "num_tokens": 429559056.0,
      "step": 13458
    },
    {
      "entropy": 0.7147410474717617,
      "epoch": 1.2382965020327656,
      "grad_norm": 0.16500882804393768,
      "learning_rate": 5.872665378599688e-05,
      "loss": 0.7098,
      "mean_token_accuracy": 0.784333985298872,
      "num_tokens": 429590716.0,
      "step": 13459
    },
    {
      "entropy": 0.5925349583849311,
      "epoch": 1.238388508404207,
      "grad_norm": 0.1522141844034195,
      "learning_rate": 5.8723586959855245e-05,
      "loss": 0.5772,
      "mean_token_accuracy": 0.822451014071703,
      "num_tokens": 429622382.0,
      "step": 13460
    },
    {
      "entropy": 0.5974573204293847,
      "epoch": 1.2384805147756481,
      "grad_norm": 0.17175450921058655,
      "learning_rate": 5.872052013371362e-05,
      "loss": 0.5711,
      "mean_token_accuracy": 0.8247728906571865,
      "num_tokens": 429654451.0,
      "step": 13461
    },
    {
      "entropy": 0.6241476866416633,
      "epoch": 1.2385725211470895,
      "grad_norm": 0.155990868806839,
      "learning_rate": 5.8717453307572e-05,
      "loss": 0.6091,
      "mean_token_accuracy": 0.8129206337034702,
      "num_tokens": 429686628.0,
      "step": 13462
    },
    {
      "entropy": 0.6264461409300566,
      "epoch": 1.2386645275185306,
      "grad_norm": 0.15187351405620575,
      "learning_rate": 5.871438648143037e-05,
      "loss": 0.6221,
      "mean_token_accuracy": 0.8115541897714138,
      "num_tokens": 429718698.0,
      "step": 13463
    },
    {
      "entropy": 0.6454753447324038,
      "epoch": 1.2387565338899718,
      "grad_norm": 0.16145268082618713,
      "learning_rate": 5.8711319655288744e-05,
      "loss": 0.6436,
      "mean_token_accuracy": 0.8044914677739143,
      "num_tokens": 429751120.0,
      "step": 13464
    },
    {
      "entropy": 0.622252251021564,
      "epoch": 1.2388485402614131,
      "grad_norm": 0.16505315899848938,
      "learning_rate": 5.870825282914711e-05,
      "loss": 0.6078,
      "mean_token_accuracy": 0.8140260055661201,
      "num_tokens": 429783091.0,
      "step": 13465
    },
    {
      "entropy": 0.6603395789861679,
      "epoch": 1.2389405466328542,
      "grad_norm": 0.15763749182224274,
      "learning_rate": 5.87051860030055e-05,
      "loss": 0.6683,
      "mean_token_accuracy": 0.7963934950530529,
      "num_tokens": 429815158.0,
      "step": 13466
    },
    {
      "entropy": 0.6292084511369467,
      "epoch": 1.2390325530042956,
      "grad_norm": 0.15660446882247925,
      "learning_rate": 5.870211917686387e-05,
      "loss": 0.604,
      "mean_token_accuracy": 0.8138937205076218,
      "num_tokens": 429847411.0,
      "step": 13467
    },
    {
      "entropy": 0.6929601635783911,
      "epoch": 1.2391245593757367,
      "grad_norm": 0.16702772676944733,
      "learning_rate": 5.8699052350722236e-05,
      "loss": 0.6735,
      "mean_token_accuracy": 0.7974533513188362,
      "num_tokens": 429879789.0,
      "step": 13468
    },
    {
      "entropy": 0.5929981898516417,
      "epoch": 1.2392165657471779,
      "grad_norm": 0.15737242996692657,
      "learning_rate": 5.869598552458061e-05,
      "loss": 0.567,
      "mean_token_accuracy": 0.8246527835726738,
      "num_tokens": 429910892.0,
      "step": 13469
    },
    {
      "entropy": 0.6852457271888852,
      "epoch": 1.2393085721186192,
      "grad_norm": 0.1531020998954773,
      "learning_rate": 5.869291869843899e-05,
      "loss": 0.6565,
      "mean_token_accuracy": 0.7986118644475937,
      "num_tokens": 429942843.0,
      "step": 13470
    },
    {
      "entropy": 0.5994620807468891,
      "epoch": 1.2394005784900604,
      "grad_norm": 0.15968815982341766,
      "learning_rate": 5.8689851872297366e-05,
      "loss": 0.5917,
      "mean_token_accuracy": 0.8210483007133007,
      "num_tokens": 429974658.0,
      "step": 13471
    },
    {
      "entropy": 0.5744737051427364,
      "epoch": 1.2394925848615017,
      "grad_norm": 0.15830989181995392,
      "learning_rate": 5.8686785046155734e-05,
      "loss": 0.5574,
      "mean_token_accuracy": 0.8312143497169018,
      "num_tokens": 430006718.0,
      "step": 13472
    },
    {
      "entropy": 0.5611130632460117,
      "epoch": 1.2395845912329428,
      "grad_norm": 0.15984316170215607,
      "learning_rate": 5.868371822001411e-05,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8309947364032269,
      "num_tokens": 430038457.0,
      "step": 13473
    },
    {
      "entropy": 0.6283718515187502,
      "epoch": 1.2396765976043842,
      "grad_norm": 0.15987084805965424,
      "learning_rate": 5.868065139387249e-05,
      "loss": 0.6223,
      "mean_token_accuracy": 0.8128928653895855,
      "num_tokens": 430069990.0,
      "step": 13474
    },
    {
      "entropy": 0.6126599796116352,
      "epoch": 1.2397686039758253,
      "grad_norm": 0.16323666274547577,
      "learning_rate": 5.867758456773086e-05,
      "loss": 0.6099,
      "mean_token_accuracy": 0.8128771595656872,
      "num_tokens": 430102070.0,
      "step": 13475
    },
    {
      "entropy": 0.61398970708251,
      "epoch": 1.2398606103472665,
      "grad_norm": 0.15428270399570465,
      "learning_rate": 5.867451774158923e-05,
      "loss": 0.6042,
      "mean_token_accuracy": 0.8196262270212173,
      "num_tokens": 430133628.0,
      "step": 13476
    },
    {
      "entropy": 0.6784104518592358,
      "epoch": 1.2399526167187078,
      "grad_norm": 0.1593165248632431,
      "learning_rate": 5.86714509154476e-05,
      "loss": 0.6886,
      "mean_token_accuracy": 0.7955339699983597,
      "num_tokens": 430165915.0,
      "step": 13477
    },
    {
      "entropy": 0.6261078184470534,
      "epoch": 1.240044623090149,
      "grad_norm": 0.1614503562450409,
      "learning_rate": 5.866838408930598e-05,
      "loss": 0.6,
      "mean_token_accuracy": 0.811595194041729,
      "num_tokens": 430197602.0,
      "step": 13478
    },
    {
      "entropy": 0.6156491423025727,
      "epoch": 1.2401366294615903,
      "grad_norm": 0.1542300134897232,
      "learning_rate": 5.866531726316436e-05,
      "loss": 0.6096,
      "mean_token_accuracy": 0.8139000423252583,
      "num_tokens": 430229673.0,
      "step": 13479
    },
    {
      "entropy": 0.5326061146333814,
      "epoch": 1.2402286358330314,
      "grad_norm": 0.1501654088497162,
      "learning_rate": 5.8662250437022725e-05,
      "loss": 0.527,
      "mean_token_accuracy": 0.8387362398207188,
      "num_tokens": 430261864.0,
      "step": 13480
    },
    {
      "entropy": 0.6247670073062181,
      "epoch": 1.2403206422044726,
      "grad_norm": 0.1645587980747223,
      "learning_rate": 5.86591836108811e-05,
      "loss": 0.6136,
      "mean_token_accuracy": 0.8184154257178307,
      "num_tokens": 430294479.0,
      "step": 13481
    },
    {
      "entropy": 0.6511237621307373,
      "epoch": 1.240412648575914,
      "grad_norm": 0.15750113129615784,
      "learning_rate": 5.865611678473948e-05,
      "loss": 0.6449,
      "mean_token_accuracy": 0.8056518547236919,
      "num_tokens": 430326135.0,
      "step": 13482
    },
    {
      "entropy": 0.6746707521378994,
      "epoch": 1.240504654947355,
      "grad_norm": 0.15803878009319305,
      "learning_rate": 5.865304995859785e-05,
      "loss": 0.6562,
      "mean_token_accuracy": 0.7943121716380119,
      "num_tokens": 430357579.0,
      "step": 13483
    },
    {
      "entropy": 0.6278254128992558,
      "epoch": 1.2405966613187964,
      "grad_norm": 0.15380606055259705,
      "learning_rate": 5.864998313245622e-05,
      "loss": 0.6063,
      "mean_token_accuracy": 0.8104276396334171,
      "num_tokens": 430389745.0,
      "step": 13484
    },
    {
      "entropy": 0.6316166166216135,
      "epoch": 1.2406886676902376,
      "grad_norm": 0.16159023344516754,
      "learning_rate": 5.8646916306314605e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8113133423030376,
      "num_tokens": 430421528.0,
      "step": 13485
    },
    {
      "entropy": 0.6673262547701597,
      "epoch": 1.2407806740616787,
      "grad_norm": 0.1618277132511139,
      "learning_rate": 5.864384948017297e-05,
      "loss": 0.6514,
      "mean_token_accuracy": 0.8040045537054539,
      "num_tokens": 430453176.0,
      "step": 13486
    },
    {
      "entropy": 0.6783617623150349,
      "epoch": 1.24087268043312,
      "grad_norm": 0.16010984778404236,
      "learning_rate": 5.864078265403135e-05,
      "loss": 0.6781,
      "mean_token_accuracy": 0.7945582866668701,
      "num_tokens": 430484796.0,
      "step": 13487
    },
    {
      "entropy": 0.5898780040442944,
      "epoch": 1.2409646868045612,
      "grad_norm": 0.15143224596977234,
      "learning_rate": 5.8637715827889715e-05,
      "loss": 0.5729,
      "mean_token_accuracy": 0.8222760297358036,
      "num_tokens": 430516996.0,
      "step": 13488
    },
    {
      "entropy": 0.6978292223066092,
      "epoch": 1.2410566931760025,
      "grad_norm": 0.1664000153541565,
      "learning_rate": 5.8634649001748096e-05,
      "loss": 0.6927,
      "mean_token_accuracy": 0.7915643006563187,
      "num_tokens": 430549277.0,
      "step": 13489
    },
    {
      "entropy": 0.6091406084597111,
      "epoch": 1.2411486995474437,
      "grad_norm": 0.15479923784732819,
      "learning_rate": 5.863158217560647e-05,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8188314735889435,
      "num_tokens": 430580762.0,
      "step": 13490
    },
    {
      "entropy": 0.594437725841999,
      "epoch": 1.2412407059188848,
      "grad_norm": 0.16201473772525787,
      "learning_rate": 5.862851534946484e-05,
      "loss": 0.5759,
      "mean_token_accuracy": 0.8223603032529354,
      "num_tokens": 430612947.0,
      "step": 13491
    },
    {
      "entropy": 0.6373399738222361,
      "epoch": 1.2413327122903262,
      "grad_norm": 0.15996097028255463,
      "learning_rate": 5.8625448523323214e-05,
      "loss": 0.6343,
      "mean_token_accuracy": 0.810515072196722,
      "num_tokens": 430645690.0,
      "step": 13492
    },
    {
      "entropy": 0.6680008042603731,
      "epoch": 1.2414247186617673,
      "grad_norm": 0.1606743186712265,
      "learning_rate": 5.8622381697181595e-05,
      "loss": 0.6613,
      "mean_token_accuracy": 0.8028510212898254,
      "num_tokens": 430677029.0,
      "step": 13493
    },
    {
      "entropy": 0.6866645254194736,
      "epoch": 1.2415167250332086,
      "grad_norm": 0.16693685948848724,
      "learning_rate": 5.861931487103996e-05,
      "loss": 0.6837,
      "mean_token_accuracy": 0.7900399193167686,
      "num_tokens": 430708532.0,
      "step": 13494
    },
    {
      "entropy": 0.5533930072560906,
      "epoch": 1.2416087314046498,
      "grad_norm": 0.14948934316635132,
      "learning_rate": 5.861624804489834e-05,
      "loss": 0.5316,
      "mean_token_accuracy": 0.8347533456981182,
      "num_tokens": 430740954.0,
      "step": 13495
    },
    {
      "entropy": 0.6100653186440468,
      "epoch": 1.241700737776091,
      "grad_norm": 0.15542016923427582,
      "learning_rate": 5.8613181218756705e-05,
      "loss": 0.6067,
      "mean_token_accuracy": 0.8136220425367355,
      "num_tokens": 430773385.0,
      "step": 13496
    },
    {
      "entropy": 0.5539974332787097,
      "epoch": 1.2417927441475323,
      "grad_norm": 0.1562240868806839,
      "learning_rate": 5.861011439261509e-05,
      "loss": 0.5397,
      "mean_token_accuracy": 0.8311381302773952,
      "num_tokens": 430805635.0,
      "step": 13497
    },
    {
      "entropy": 0.6147636752575636,
      "epoch": 1.2418847505189734,
      "grad_norm": 0.15571017563343048,
      "learning_rate": 5.860704756647346e-05,
      "loss": 0.6061,
      "mean_token_accuracy": 0.8136125691235065,
      "num_tokens": 430837810.0,
      "step": 13498
    },
    {
      "entropy": 0.5584591748192906,
      "epoch": 1.2419767568904148,
      "grad_norm": 0.15161803364753723,
      "learning_rate": 5.860398074033183e-05,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8314989320933819,
      "num_tokens": 430870273.0,
      "step": 13499
    },
    {
      "entropy": 0.55952771846205,
      "epoch": 1.242068763261856,
      "grad_norm": 0.15429715812206268,
      "learning_rate": 5.8600913914190204e-05,
      "loss": 0.5493,
      "mean_token_accuracy": 0.8326489962637424,
      "num_tokens": 430901963.0,
      "step": 13500
    },
    {
      "entropy": 0.6509762965142727,
      "epoch": 1.242160769633297,
      "grad_norm": 0.1644594967365265,
      "learning_rate": 5.8597847088048585e-05,
      "loss": 0.6607,
      "mean_token_accuracy": 0.7979962453246117,
      "num_tokens": 430934177.0,
      "step": 13501
    },
    {
      "entropy": 0.5861798273399472,
      "epoch": 1.2422527760047384,
      "grad_norm": 0.15759772062301636,
      "learning_rate": 5.859478026190695e-05,
      "loss": 0.5902,
      "mean_token_accuracy": 0.8203009366989136,
      "num_tokens": 430965095.0,
      "step": 13502
    },
    {
      "entropy": 0.6360366027802229,
      "epoch": 1.2423447823761795,
      "grad_norm": 0.15371251106262207,
      "learning_rate": 5.859171343576533e-05,
      "loss": 0.6239,
      "mean_token_accuracy": 0.8065202832221985,
      "num_tokens": 430996740.0,
      "step": 13503
    },
    {
      "entropy": 0.6301224865019321,
      "epoch": 1.2424367887476209,
      "grad_norm": 0.15903310477733612,
      "learning_rate": 5.8588646609623696e-05,
      "loss": 0.6207,
      "mean_token_accuracy": 0.8122281059622765,
      "num_tokens": 431028509.0,
      "step": 13504
    },
    {
      "entropy": 0.6031939322128892,
      "epoch": 1.242528795119062,
      "grad_norm": 0.1818179190158844,
      "learning_rate": 5.8585579783482084e-05,
      "loss": 0.5998,
      "mean_token_accuracy": 0.8156940452754498,
      "num_tokens": 431060791.0,
      "step": 13505
    },
    {
      "entropy": 0.5747685520909727,
      "epoch": 1.2426208014905031,
      "grad_norm": 0.14987117052078247,
      "learning_rate": 5.858251295734045e-05,
      "loss": 0.5723,
      "mean_token_accuracy": 0.8217327520251274,
      "num_tokens": 431093174.0,
      "step": 13506
    },
    {
      "entropy": 0.7109661204740405,
      "epoch": 1.2427128078619445,
      "grad_norm": 0.16305503249168396,
      "learning_rate": 5.857944613119882e-05,
      "loss": 0.7108,
      "mean_token_accuracy": 0.7846770100295544,
      "num_tokens": 431125761.0,
      "step": 13507
    },
    {
      "entropy": 0.6081267427653074,
      "epoch": 1.2428048142333856,
      "grad_norm": 0.1624978929758072,
      "learning_rate": 5.8576379305057194e-05,
      "loss": 0.5718,
      "mean_token_accuracy": 0.821160614490509,
      "num_tokens": 431157716.0,
      "step": 13508
    },
    {
      "entropy": 0.6069595832377672,
      "epoch": 1.242896820604827,
      "grad_norm": 0.15559162199497223,
      "learning_rate": 5.8573312478915576e-05,
      "loss": 0.5897,
      "mean_token_accuracy": 0.8154036365449429,
      "num_tokens": 431189386.0,
      "step": 13509
    },
    {
      "entropy": 0.5925675109028816,
      "epoch": 1.2429888269762681,
      "grad_norm": 0.15497367084026337,
      "learning_rate": 5.857024565277395e-05,
      "loss": 0.5772,
      "mean_token_accuracy": 0.8204798735678196,
      "num_tokens": 431221617.0,
      "step": 13510
    },
    {
      "entropy": 0.5995412264019251,
      "epoch": 1.2430808333477092,
      "grad_norm": 0.15700607001781464,
      "learning_rate": 5.856717882663232e-05,
      "loss": 0.5829,
      "mean_token_accuracy": 0.8299849927425385,
      "num_tokens": 431253552.0,
      "step": 13511
    },
    {
      "entropy": 0.7118108198046684,
      "epoch": 1.2431728397191506,
      "grad_norm": 0.16879823803901672,
      "learning_rate": 5.8564112000490686e-05,
      "loss": 0.7217,
      "mean_token_accuracy": 0.7904781959950924,
      "num_tokens": 431285363.0,
      "step": 13512
    },
    {
      "entropy": 0.7546670511364937,
      "epoch": 1.2432648460905917,
      "grad_norm": 0.16552932560443878,
      "learning_rate": 5.8561045174349074e-05,
      "loss": 0.7376,
      "mean_token_accuracy": 0.774103756994009,
      "num_tokens": 431317364.0,
      "step": 13513
    },
    {
      "entropy": 0.6725046169012785,
      "epoch": 1.243356852462033,
      "grad_norm": 0.1628207564353943,
      "learning_rate": 5.855797834820744e-05,
      "loss": 0.6386,
      "mean_token_accuracy": 0.8026120886206627,
      "num_tokens": 431349354.0,
      "step": 13514
    },
    {
      "entropy": 0.6258168276399374,
      "epoch": 1.2434488588334742,
      "grad_norm": 0.15652517974376678,
      "learning_rate": 5.855491152206582e-05,
      "loss": 0.6184,
      "mean_token_accuracy": 0.8095082901418209,
      "num_tokens": 431381569.0,
      "step": 13515
    },
    {
      "entropy": 0.5421446277759969,
      "epoch": 1.2435408652049154,
      "grad_norm": 0.14782394468784332,
      "learning_rate": 5.85518446959242e-05,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8353825844824314,
      "num_tokens": 431413051.0,
      "step": 13516
    },
    {
      "entropy": 0.5443812552839518,
      "epoch": 1.2436328715763567,
      "grad_norm": 0.15323564410209656,
      "learning_rate": 5.8548777869782566e-05,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8387059271335602,
      "num_tokens": 431444810.0,
      "step": 13517
    },
    {
      "entropy": 0.5383858426939696,
      "epoch": 1.2437248779477978,
      "grad_norm": 0.1580013781785965,
      "learning_rate": 5.854571104364094e-05,
      "loss": 0.5177,
      "mean_token_accuracy": 0.8379182070493698,
      "num_tokens": 431476120.0,
      "step": 13518
    },
    {
      "entropy": 0.6624132320284843,
      "epoch": 1.2438168843192392,
      "grad_norm": 0.15867194533348083,
      "learning_rate": 5.854264421749931e-05,
      "loss": 0.6551,
      "mean_token_accuracy": 0.7974934577941895,
      "num_tokens": 431507931.0,
      "step": 13519
    },
    {
      "entropy": 0.6221994804218411,
      "epoch": 1.2439088906906803,
      "grad_norm": 0.16077858209609985,
      "learning_rate": 5.853957739135769e-05,
      "loss": 0.6098,
      "mean_token_accuracy": 0.8130468800663948,
      "num_tokens": 431539696.0,
      "step": 13520
    },
    {
      "entropy": 0.5002462472766638,
      "epoch": 1.2440008970621215,
      "grad_norm": 0.15034490823745728,
      "learning_rate": 5.8536510565216065e-05,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8507877327501774,
      "num_tokens": 431572464.0,
      "step": 13521
    },
    {
      "entropy": 0.6002569161355495,
      "epoch": 1.2440929034335628,
      "grad_norm": 0.16642369329929352,
      "learning_rate": 5.853344373907443e-05,
      "loss": 0.5861,
      "mean_token_accuracy": 0.818130623549223,
      "num_tokens": 431602825.0,
      "step": 13522
    },
    {
      "entropy": 0.6698072999715805,
      "epoch": 1.244184909805004,
      "grad_norm": 0.1584080159664154,
      "learning_rate": 5.853037691293281e-05,
      "loss": 0.6646,
      "mean_token_accuracy": 0.7988357730209827,
      "num_tokens": 431635024.0,
      "step": 13523
    },
    {
      "entropy": 0.5946244345977902,
      "epoch": 1.2442769161764453,
      "grad_norm": 0.16195467114448547,
      "learning_rate": 5.852731008679119e-05,
      "loss": 0.5976,
      "mean_token_accuracy": 0.8153934702277184,
      "num_tokens": 431667377.0,
      "step": 13524
    },
    {
      "entropy": 0.6994889825582504,
      "epoch": 1.2443689225478864,
      "grad_norm": 0.16479472815990448,
      "learning_rate": 5.852424326064956e-05,
      "loss": 0.6929,
      "mean_token_accuracy": 0.7915644906461239,
      "num_tokens": 431699960.0,
      "step": 13525
    },
    {
      "entropy": 0.6136886011809111,
      "epoch": 1.2444609289193276,
      "grad_norm": 0.1560410112142563,
      "learning_rate": 5.852117643450793e-05,
      "loss": 0.6005,
      "mean_token_accuracy": 0.8140053115785122,
      "num_tokens": 431732196.0,
      "step": 13526
    },
    {
      "entropy": 0.6628198903053999,
      "epoch": 1.244552935290769,
      "grad_norm": 0.17455020546913147,
      "learning_rate": 5.85181096083663e-05,
      "loss": 0.6631,
      "mean_token_accuracy": 0.8004330284893513,
      "num_tokens": 431763686.0,
      "step": 13527
    },
    {
      "entropy": 0.6616682009771466,
      "epoch": 1.24464494166221,
      "grad_norm": 0.16122765839099884,
      "learning_rate": 5.851504278222468e-05,
      "loss": 0.6509,
      "mean_token_accuracy": 0.8018053993582726,
      "num_tokens": 431795627.0,
      "step": 13528
    },
    {
      "entropy": 0.5279988199472427,
      "epoch": 1.2447369480336514,
      "grad_norm": 0.14384829998016357,
      "learning_rate": 5.8511975956083055e-05,
      "loss": 0.5126,
      "mean_token_accuracy": 0.8391779847443104,
      "num_tokens": 431827924.0,
      "step": 13529
    },
    {
      "entropy": 0.686291141435504,
      "epoch": 1.2448289544050926,
      "grad_norm": 0.15945489704608917,
      "learning_rate": 5.850890912994142e-05,
      "loss": 0.6797,
      "mean_token_accuracy": 0.7910711392760277,
      "num_tokens": 431860022.0,
      "step": 13530
    },
    {
      "entropy": 0.6336824279278517,
      "epoch": 1.2449209607765337,
      "grad_norm": 0.15919534862041473,
      "learning_rate": 5.85058423037998e-05,
      "loss": 0.6127,
      "mean_token_accuracy": 0.8124319687485695,
      "num_tokens": 431891055.0,
      "step": 13531
    },
    {
      "entropy": 0.6611185558140278,
      "epoch": 1.245012967147975,
      "grad_norm": 0.16616344451904297,
      "learning_rate": 5.850277547765818e-05,
      "loss": 0.6547,
      "mean_token_accuracy": 0.8019640780985355,
      "num_tokens": 431922095.0,
      "step": 13532
    },
    {
      "entropy": 0.5612351316958666,
      "epoch": 1.2451049735194162,
      "grad_norm": 0.16020068526268005,
      "learning_rate": 5.849970865151655e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.831519365310669,
      "num_tokens": 431953182.0,
      "step": 13533
    },
    {
      "entropy": 0.6171158980578184,
      "epoch": 1.2451969798908575,
      "grad_norm": 0.1496899425983429,
      "learning_rate": 5.849664182537492e-05,
      "loss": 0.5882,
      "mean_token_accuracy": 0.8193084709346294,
      "num_tokens": 431985610.0,
      "step": 13534
    },
    {
      "entropy": 0.5575939491391182,
      "epoch": 1.2452889862622987,
      "grad_norm": 0.14726081490516663,
      "learning_rate": 5.849357499923329e-05,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8296798467636108,
      "num_tokens": 432018293.0,
      "step": 13535
    },
    {
      "entropy": 0.7294934410601854,
      "epoch": 1.2453809926337398,
      "grad_norm": 0.1609850525856018,
      "learning_rate": 5.849050817309167e-05,
      "loss": 0.7239,
      "mean_token_accuracy": 0.7790153957903385,
      "num_tokens": 432050819.0,
      "step": 13536
    },
    {
      "entropy": 0.7276347652077675,
      "epoch": 1.2454729990051812,
      "grad_norm": 0.16883541643619537,
      "learning_rate": 5.8487441346950046e-05,
      "loss": 0.7262,
      "mean_token_accuracy": 0.7852255366742611,
      "num_tokens": 432083218.0,
      "step": 13537
    },
    {
      "entropy": 0.6129519483074546,
      "epoch": 1.2455650053766223,
      "grad_norm": 0.15405769646167755,
      "learning_rate": 5.8484374520808414e-05,
      "loss": 0.6022,
      "mean_token_accuracy": 0.8117906153202057,
      "num_tokens": 432115129.0,
      "step": 13538
    },
    {
      "entropy": 0.6494015343487263,
      "epoch": 1.2456570117480636,
      "grad_norm": 0.16114167869091034,
      "learning_rate": 5.848130769466679e-05,
      "loss": 0.6573,
      "mean_token_accuracy": 0.7980438992381096,
      "num_tokens": 432147437.0,
      "step": 13539
    },
    {
      "entropy": 0.5814390173181891,
      "epoch": 1.2457490181195048,
      "grad_norm": 0.15859095752239227,
      "learning_rate": 5.847824086852517e-05,
      "loss": 0.5612,
      "mean_token_accuracy": 0.8298780098557472,
      "num_tokens": 432179286.0,
      "step": 13540
    },
    {
      "entropy": 0.7119233086705208,
      "epoch": 1.245841024490946,
      "grad_norm": 0.16384780406951904,
      "learning_rate": 5.847517404238354e-05,
      "loss": 0.6981,
      "mean_token_accuracy": 0.7904137782752514,
      "num_tokens": 432211833.0,
      "step": 13541
    },
    {
      "entropy": 0.6574539472348988,
      "epoch": 1.2459330308623873,
      "grad_norm": 0.16414231061935425,
      "learning_rate": 5.847210721624191e-05,
      "loss": 0.64,
      "mean_token_accuracy": 0.8050792068243027,
      "num_tokens": 432243817.0,
      "step": 13542
    },
    {
      "entropy": 0.5379533302038908,
      "epoch": 1.2460250372338284,
      "grad_norm": 0.15731212496757507,
      "learning_rate": 5.846904039010028e-05,
      "loss": 0.5172,
      "mean_token_accuracy": 0.843318447470665,
      "num_tokens": 432276033.0,
      "step": 13543
    },
    {
      "entropy": 0.7644574795849621,
      "epoch": 1.2461170436052698,
      "grad_norm": 0.1699763834476471,
      "learning_rate": 5.846597356395866e-05,
      "loss": 0.7542,
      "mean_token_accuracy": 0.7725793644785881,
      "num_tokens": 432308491.0,
      "step": 13544
    },
    {
      "entropy": 0.586843641474843,
      "epoch": 1.246209049976711,
      "grad_norm": 0.16509824991226196,
      "learning_rate": 5.8462906737817036e-05,
      "loss": 0.5691,
      "mean_token_accuracy": 0.825155645608902,
      "num_tokens": 432340658.0,
      "step": 13545
    },
    {
      "entropy": 0.6004185006022453,
      "epoch": 1.246301056348152,
      "grad_norm": 0.1559741199016571,
      "learning_rate": 5.8459839911675404e-05,
      "loss": 0.5841,
      "mean_token_accuracy": 0.8157529011368752,
      "num_tokens": 432372886.0,
      "step": 13546
    },
    {
      "entropy": 0.6839172430336475,
      "epoch": 1.2463930627195934,
      "grad_norm": 0.15709196031093597,
      "learning_rate": 5.845677308553379e-05,
      "loss": 0.6697,
      "mean_token_accuracy": 0.7995891459286213,
      "num_tokens": 432404862.0,
      "step": 13547
    },
    {
      "entropy": 0.6588228270411491,
      "epoch": 1.2464850690910345,
      "grad_norm": 0.1621522456407547,
      "learning_rate": 5.845370625939216e-05,
      "loss": 0.6358,
      "mean_token_accuracy": 0.804885882884264,
      "num_tokens": 432436849.0,
      "step": 13548
    },
    {
      "entropy": 0.6215848447754979,
      "epoch": 1.2465770754624759,
      "grad_norm": 0.1582815945148468,
      "learning_rate": 5.8450639433250535e-05,
      "loss": 0.6115,
      "mean_token_accuracy": 0.8139011152088642,
      "num_tokens": 432468997.0,
      "step": 13549
    },
    {
      "entropy": 0.5450646262615919,
      "epoch": 1.246669081833917,
      "grad_norm": 0.14948125183582306,
      "learning_rate": 5.84475726071089e-05,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8343516476452351,
      "num_tokens": 432501095.0,
      "step": 13550
    },
    {
      "entropy": 0.7273653531447053,
      "epoch": 1.2467610882053581,
      "grad_norm": 0.1691914200782776,
      "learning_rate": 5.8444505780967284e-05,
      "loss": 0.7166,
      "mean_token_accuracy": 0.7832535654306412,
      "num_tokens": 432532852.0,
      "step": 13551
    },
    {
      "entropy": 0.6479080226272345,
      "epoch": 1.2468530945767995,
      "grad_norm": 0.16146038472652435,
      "learning_rate": 5.844143895482566e-05,
      "loss": 0.6443,
      "mean_token_accuracy": 0.802228257060051,
      "num_tokens": 432564945.0,
      "step": 13552
    },
    {
      "entropy": 0.6518581379204988,
      "epoch": 1.2469451009482406,
      "grad_norm": 0.16434897482395172,
      "learning_rate": 5.8438372128684026e-05,
      "loss": 0.6414,
      "mean_token_accuracy": 0.8001020289957523,
      "num_tokens": 432596399.0,
      "step": 13553
    },
    {
      "entropy": 0.6322071086615324,
      "epoch": 1.247037107319682,
      "grad_norm": 0.15968717634677887,
      "learning_rate": 5.84353053025424e-05,
      "loss": 0.6326,
      "mean_token_accuracy": 0.8093116357922554,
      "num_tokens": 432628525.0,
      "step": 13554
    },
    {
      "entropy": 0.6333552682772279,
      "epoch": 1.2471291136911231,
      "grad_norm": 0.15790694952011108,
      "learning_rate": 5.843223847640078e-05,
      "loss": 0.6236,
      "mean_token_accuracy": 0.8092686757445335,
      "num_tokens": 432660280.0,
      "step": 13555
    },
    {
      "entropy": 0.6164904367178679,
      "epoch": 1.2472211200625642,
      "grad_norm": 0.15962274372577667,
      "learning_rate": 5.842917165025915e-05,
      "loss": 0.5984,
      "mean_token_accuracy": 0.8168250732123852,
      "num_tokens": 432692017.0,
      "step": 13556
    },
    {
      "entropy": 0.6411498105153441,
      "epoch": 1.2473131264340056,
      "grad_norm": 0.1553613245487213,
      "learning_rate": 5.8426104824117525e-05,
      "loss": 0.6315,
      "mean_token_accuracy": 0.8056334219872952,
      "num_tokens": 432723394.0,
      "step": 13557
    },
    {
      "entropy": 0.5788328826893121,
      "epoch": 1.2474051328054467,
      "grad_norm": 0.14970356225967407,
      "learning_rate": 5.842303799797589e-05,
      "loss": 0.5674,
      "mean_token_accuracy": 0.8263806141912937,
      "num_tokens": 432755699.0,
      "step": 13558
    },
    {
      "entropy": 0.6461909841746092,
      "epoch": 1.247497139176888,
      "grad_norm": 0.16533806920051575,
      "learning_rate": 5.8419971171834274e-05,
      "loss": 0.6438,
      "mean_token_accuracy": 0.8047874644398689,
      "num_tokens": 432787697.0,
      "step": 13559
    },
    {
      "entropy": 0.6061988472938538,
      "epoch": 1.2475891455483292,
      "grad_norm": 0.1686784327030182,
      "learning_rate": 5.841690434569265e-05,
      "loss": 0.6032,
      "mean_token_accuracy": 0.817499216645956,
      "num_tokens": 432820129.0,
      "step": 13560
    },
    {
      "entropy": 0.6989496685564518,
      "epoch": 1.2476811519197704,
      "grad_norm": 0.16431134939193726,
      "learning_rate": 5.841383751955102e-05,
      "loss": 0.6845,
      "mean_token_accuracy": 0.7900674566626549,
      "num_tokens": 432852755.0,
      "step": 13561
    },
    {
      "entropy": 0.7207445949316025,
      "epoch": 1.2477731582912117,
      "grad_norm": 0.15922293066978455,
      "learning_rate": 5.841077069340939e-05,
      "loss": 0.6999,
      "mean_token_accuracy": 0.786980614066124,
      "num_tokens": 432884306.0,
      "step": 13562
    },
    {
      "entropy": 0.7563261780887842,
      "epoch": 1.2478651646626528,
      "grad_norm": 0.1663479208946228,
      "learning_rate": 5.840770386726777e-05,
      "loss": 0.7477,
      "mean_token_accuracy": 0.7790796905755997,
      "num_tokens": 432915948.0,
      "step": 13563
    },
    {
      "entropy": 0.6050158776342869,
      "epoch": 1.2479571710340942,
      "grad_norm": 0.15667405724525452,
      "learning_rate": 5.840463704112614e-05,
      "loss": 0.5929,
      "mean_token_accuracy": 0.8154220320284367,
      "num_tokens": 432947138.0,
      "step": 13564
    },
    {
      "entropy": 0.6108365841209888,
      "epoch": 1.2480491774055353,
      "grad_norm": 0.15548670291900635,
      "learning_rate": 5.8401570214984515e-05,
      "loss": 0.5965,
      "mean_token_accuracy": 0.8151691146194935,
      "num_tokens": 432978424.0,
      "step": 13565
    },
    {
      "entropy": 0.6168243177235126,
      "epoch": 1.2481411837769765,
      "grad_norm": 0.1603458672761917,
      "learning_rate": 5.839850338884288e-05,
      "loss": 0.5926,
      "mean_token_accuracy": 0.8190787695348263,
      "num_tokens": 433010929.0,
      "step": 13566
    },
    {
      "entropy": 0.6113046547397971,
      "epoch": 1.2482331901484178,
      "grad_norm": 0.16739551723003387,
      "learning_rate": 5.8395436562701265e-05,
      "loss": 0.5945,
      "mean_token_accuracy": 0.8172762021422386,
      "num_tokens": 433042187.0,
      "step": 13567
    },
    {
      "entropy": 0.6769581325352192,
      "epoch": 1.248325196519859,
      "grad_norm": 0.16136081516742706,
      "learning_rate": 5.839236973655964e-05,
      "loss": 0.6658,
      "mean_token_accuracy": 0.7979142256081104,
      "num_tokens": 433073685.0,
      "step": 13568
    },
    {
      "entropy": 0.5736592635512352,
      "epoch": 1.2484172028913003,
      "grad_norm": 0.18366113305091858,
      "learning_rate": 5.838930291041801e-05,
      "loss": 0.5593,
      "mean_token_accuracy": 0.8324122056365013,
      "num_tokens": 433105814.0,
      "step": 13569
    },
    {
      "entropy": 0.6664004549384117,
      "epoch": 1.2485092092627414,
      "grad_norm": 0.15995897352695465,
      "learning_rate": 5.838623608427638e-05,
      "loss": 0.6602,
      "mean_token_accuracy": 0.7985290139913559,
      "num_tokens": 433137631.0,
      "step": 13570
    },
    {
      "entropy": 0.650577811524272,
      "epoch": 1.2486012156341826,
      "grad_norm": 0.15970934927463531,
      "learning_rate": 5.838316925813476e-05,
      "loss": 0.6321,
      "mean_token_accuracy": 0.8083401210606098,
      "num_tokens": 433170280.0,
      "step": 13571
    },
    {
      "entropy": 0.5982727948576212,
      "epoch": 1.248693222005624,
      "grad_norm": 0.15630251169204712,
      "learning_rate": 5.838010243199313e-05,
      "loss": 0.5835,
      "mean_token_accuracy": 0.8208181895315647,
      "num_tokens": 433202634.0,
      "step": 13572
    },
    {
      "entropy": 0.5072833825834095,
      "epoch": 1.248785228377065,
      "grad_norm": 0.14842307567596436,
      "learning_rate": 5.8377035605851506e-05,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8466499447822571,
      "num_tokens": 433234315.0,
      "step": 13573
    },
    {
      "entropy": 0.6545376516878605,
      "epoch": 1.2488772347485064,
      "grad_norm": 0.1619809865951538,
      "learning_rate": 5.8373968779709874e-05,
      "loss": 0.6402,
      "mean_token_accuracy": 0.8020425513386726,
      "num_tokens": 433266692.0,
      "step": 13574
    },
    {
      "entropy": 0.48188381362706423,
      "epoch": 1.2489692411199476,
      "grad_norm": 0.14114414155483246,
      "learning_rate": 5.8370901953568255e-05,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8553413450717926,
      "num_tokens": 433298818.0,
      "step": 13575
    },
    {
      "entropy": 0.6324389539659023,
      "epoch": 1.2490612474913887,
      "grad_norm": 0.1502128541469574,
      "learning_rate": 5.836783512742663e-05,
      "loss": 0.6118,
      "mean_token_accuracy": 0.8137650936841965,
      "num_tokens": 433331362.0,
      "step": 13576
    },
    {
      "entropy": 0.7108718436211348,
      "epoch": 1.24915325386283,
      "grad_norm": 0.16905659437179565,
      "learning_rate": 5.8364768301285e-05,
      "loss": 0.7078,
      "mean_token_accuracy": 0.7848527431488037,
      "num_tokens": 433362791.0,
      "step": 13577
    },
    {
      "entropy": 0.5801659333519638,
      "epoch": 1.2492452602342712,
      "grad_norm": 0.16089190542697906,
      "learning_rate": 5.836170147514338e-05,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8291189558804035,
      "num_tokens": 433394817.0,
      "step": 13578
    },
    {
      "entropy": 0.7265507224947214,
      "epoch": 1.2493372666057125,
      "grad_norm": 0.16762475669384003,
      "learning_rate": 5.8358634649001754e-05,
      "loss": 0.7268,
      "mean_token_accuracy": 0.7822379544377327,
      "num_tokens": 433426503.0,
      "step": 13579
    },
    {
      "entropy": 0.5870698569342494,
      "epoch": 1.2494292729771537,
      "grad_norm": 0.15566863119602203,
      "learning_rate": 5.835556782286012e-05,
      "loss": 0.5776,
      "mean_token_accuracy": 0.8229421079158783,
      "num_tokens": 433459271.0,
      "step": 13580
    },
    {
      "entropy": 0.5152214355766773,
      "epoch": 1.2495212793485948,
      "grad_norm": 0.16356495022773743,
      "learning_rate": 5.8352500996718496e-05,
      "loss": 0.5098,
      "mean_token_accuracy": 0.8405087701976299,
      "num_tokens": 433491760.0,
      "step": 13581
    },
    {
      "entropy": 0.5941928261891007,
      "epoch": 1.2496132857200362,
      "grad_norm": 0.15254025161266327,
      "learning_rate": 5.834943417057688e-05,
      "loss": 0.5827,
      "mean_token_accuracy": 0.8201040327548981,
      "num_tokens": 433524243.0,
      "step": 13582
    },
    {
      "entropy": 0.627136318013072,
      "epoch": 1.2497052920914773,
      "grad_norm": 0.1646062284708023,
      "learning_rate": 5.8346367344435246e-05,
      "loss": 0.6356,
      "mean_token_accuracy": 0.8024653010070324,
      "num_tokens": 433555926.0,
      "step": 13583
    },
    {
      "entropy": 0.6342028006911278,
      "epoch": 1.2497972984629186,
      "grad_norm": 0.158645361661911,
      "learning_rate": 5.834330051829362e-05,
      "loss": 0.6284,
      "mean_token_accuracy": 0.8067393749952316,
      "num_tokens": 433588047.0,
      "step": 13584
    },
    {
      "entropy": 0.7749241404235363,
      "epoch": 1.2498893048343598,
      "grad_norm": 0.1737881451845169,
      "learning_rate": 5.834023369215199e-05,
      "loss": 0.7686,
      "mean_token_accuracy": 0.7678365670144558,
      "num_tokens": 433619710.0,
      "step": 13585
    },
    {
      "entropy": 0.7444337513297796,
      "epoch": 1.249981311205801,
      "grad_norm": 0.16746576130390167,
      "learning_rate": 5.8337166866010376e-05,
      "loss": 0.718,
      "mean_token_accuracy": 0.7798621952533722,
      "num_tokens": 433650814.0,
      "step": 13586
    },
    {
      "entropy": 0.6744423303753138,
      "epoch": 1.2500733175772423,
      "grad_norm": 0.1580801159143448,
      "learning_rate": 5.8334100039868744e-05,
      "loss": 0.6737,
      "mean_token_accuracy": 0.7917984873056412,
      "num_tokens": 433682831.0,
      "step": 13587
    },
    {
      "entropy": 0.6697842674329877,
      "epoch": 1.2501653239486834,
      "grad_norm": 0.16270826756954193,
      "learning_rate": 5.833103321372711e-05,
      "loss": 0.6634,
      "mean_token_accuracy": 0.8004623837769032,
      "num_tokens": 433714700.0,
      "step": 13588
    },
    {
      "entropy": 0.6101587247103453,
      "epoch": 1.2502573303201245,
      "grad_norm": 0.15851427614688873,
      "learning_rate": 5.832796638758549e-05,
      "loss": 0.5917,
      "mean_token_accuracy": 0.8199047036468983,
      "num_tokens": 433745303.0,
      "step": 13589
    },
    {
      "entropy": 0.6925514582544565,
      "epoch": 1.250349336691566,
      "grad_norm": 0.16202689707279205,
      "learning_rate": 5.832489956144387e-05,
      "loss": 0.6874,
      "mean_token_accuracy": 0.7926772609353065,
      "num_tokens": 433776908.0,
      "step": 13590
    },
    {
      "entropy": 0.5882102251052856,
      "epoch": 1.2504413430630072,
      "grad_norm": 0.150215283036232,
      "learning_rate": 5.832183273530224e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.825865413993597,
      "num_tokens": 433808954.0,
      "step": 13591
    },
    {
      "entropy": 0.6289986539632082,
      "epoch": 1.2505333494344484,
      "grad_norm": 0.1580798476934433,
      "learning_rate": 5.831876590916061e-05,
      "loss": 0.5952,
      "mean_token_accuracy": 0.8144492879509926,
      "num_tokens": 433840428.0,
      "step": 13592
    },
    {
      "entropy": 0.6850251480937004,
      "epoch": 1.2506253558058895,
      "grad_norm": 0.15445822477340698,
      "learning_rate": 5.8315699083018985e-05,
      "loss": 0.6685,
      "mean_token_accuracy": 0.7940789982676506,
      "num_tokens": 433871940.0,
      "step": 13593
    },
    {
      "entropy": 0.6937677906826138,
      "epoch": 1.2507173621773306,
      "grad_norm": 0.17444536089897156,
      "learning_rate": 5.831263225687737e-05,
      "loss": 0.6839,
      "mean_token_accuracy": 0.7875981554389,
      "num_tokens": 433903266.0,
      "step": 13594
    },
    {
      "entropy": 0.6051407810300589,
      "epoch": 1.250809368548772,
      "grad_norm": 0.15545816719532013,
      "learning_rate": 5.8309565430735735e-05,
      "loss": 0.5876,
      "mean_token_accuracy": 0.8209045305848122,
      "num_tokens": 433935716.0,
      "step": 13595
    },
    {
      "entropy": 0.5564764784649014,
      "epoch": 1.2509013749202134,
      "grad_norm": 0.1519545316696167,
      "learning_rate": 5.830649860459411e-05,
      "loss": 0.5396,
      "mean_token_accuracy": 0.8310263156890869,
      "num_tokens": 433968092.0,
      "step": 13596
    },
    {
      "entropy": 0.5316421296447515,
      "epoch": 1.2509933812916545,
      "grad_norm": 0.15023034811019897,
      "learning_rate": 5.830343177845248e-05,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8434770032763481,
      "num_tokens": 434000154.0,
      "step": 13597
    },
    {
      "entropy": 0.6941378936171532,
      "epoch": 1.2510853876630956,
      "grad_norm": 0.1642829328775406,
      "learning_rate": 5.830036495231086e-05,
      "loss": 0.6805,
      "mean_token_accuracy": 0.7958221882581711,
      "num_tokens": 434031785.0,
      "step": 13598
    },
    {
      "entropy": 0.6135280653834343,
      "epoch": 1.2511773940345368,
      "grad_norm": 0.15915237367153168,
      "learning_rate": 5.829729812616923e-05,
      "loss": 0.6144,
      "mean_token_accuracy": 0.8115504011511803,
      "num_tokens": 434063940.0,
      "step": 13599
    },
    {
      "entropy": 0.55869853310287,
      "epoch": 1.2512694004059781,
      "grad_norm": 0.1501564234495163,
      "learning_rate": 5.82942313000276e-05,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8276578187942505,
      "num_tokens": 434095579.0,
      "step": 13600
    },
    {
      "entropy": 0.5352748651057482,
      "epoch": 1.2513614067774195,
      "grad_norm": 0.1491609662771225,
      "learning_rate": 5.8291164473885976e-05,
      "loss": 0.5315,
      "mean_token_accuracy": 0.8346877247095108,
      "num_tokens": 434127494.0,
      "step": 13601
    },
    {
      "entropy": 0.589847844094038,
      "epoch": 1.2514534131488606,
      "grad_norm": 0.1564723402261734,
      "learning_rate": 5.828809764774436e-05,
      "loss": 0.5847,
      "mean_token_accuracy": 0.8182928673923016,
      "num_tokens": 434159801.0,
      "step": 13602
    },
    {
      "entropy": 0.6257422268390656,
      "epoch": 1.2515454195203017,
      "grad_norm": 0.1558103859424591,
      "learning_rate": 5.8285030821602725e-05,
      "loss": 0.6148,
      "mean_token_accuracy": 0.8159666918218136,
      "num_tokens": 434191091.0,
      "step": 13603
    },
    {
      "entropy": 0.5929930591955781,
      "epoch": 1.2516374258917429,
      "grad_norm": 0.14978595077991486,
      "learning_rate": 5.82819639954611e-05,
      "loss": 0.574,
      "mean_token_accuracy": 0.8183021433651447,
      "num_tokens": 434223511.0,
      "step": 13604
    },
    {
      "entropy": 0.545425727032125,
      "epoch": 1.2517294322631842,
      "grad_norm": 0.15911944210529327,
      "learning_rate": 5.827889716931947e-05,
      "loss": 0.5417,
      "mean_token_accuracy": 0.8324592709541321,
      "num_tokens": 434255512.0,
      "step": 13605
    },
    {
      "entropy": 0.5835385480895638,
      "epoch": 1.2518214386346256,
      "grad_norm": 0.15905387699604034,
      "learning_rate": 5.827583034317785e-05,
      "loss": 0.5679,
      "mean_token_accuracy": 0.8223843239247799,
      "num_tokens": 434287601.0,
      "step": 13606
    },
    {
      "entropy": 0.5735831633210182,
      "epoch": 1.2519134450060667,
      "grad_norm": 0.15574988722801208,
      "learning_rate": 5.8272763517036224e-05,
      "loss": 0.5626,
      "mean_token_accuracy": 0.8262788243591785,
      "num_tokens": 434319540.0,
      "step": 13607
    },
    {
      "entropy": 0.6436597658321261,
      "epoch": 1.2520054513775078,
      "grad_norm": 0.15553101897239685,
      "learning_rate": 5.826969669089459e-05,
      "loss": 0.6406,
      "mean_token_accuracy": 0.8024387545883656,
      "num_tokens": 434351094.0,
      "step": 13608
    },
    {
      "entropy": 0.5639484319835901,
      "epoch": 1.252097457748949,
      "grad_norm": 0.1593909114599228,
      "learning_rate": 5.826662986475297e-05,
      "loss": 0.5607,
      "mean_token_accuracy": 0.827514573931694,
      "num_tokens": 434383452.0,
      "step": 13609
    },
    {
      "entropy": 0.5543096344918013,
      "epoch": 1.2521894641203903,
      "grad_norm": 0.15684348344802856,
      "learning_rate": 5.826356303861135e-05,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8274409435689449,
      "num_tokens": 434415146.0,
      "step": 13610
    },
    {
      "entropy": 0.5915232533589005,
      "epoch": 1.2522814704918317,
      "grad_norm": 0.152741938829422,
      "learning_rate": 5.8260496212469715e-05,
      "loss": 0.572,
      "mean_token_accuracy": 0.8244687281548977,
      "num_tokens": 434447914.0,
      "step": 13611
    },
    {
      "entropy": 0.6014915350824594,
      "epoch": 1.2523734768632728,
      "grad_norm": 0.15750928223133087,
      "learning_rate": 5.825742938632809e-05,
      "loss": 0.5983,
      "mean_token_accuracy": 0.8186559975147247,
      "num_tokens": 434480279.0,
      "step": 13612
    },
    {
      "entropy": 0.5507243783213198,
      "epoch": 1.252465483234714,
      "grad_norm": 0.15290828049182892,
      "learning_rate": 5.825436256018647e-05,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8368194289505482,
      "num_tokens": 434512097.0,
      "step": 13613
    },
    {
      "entropy": 0.6398242292925715,
      "epoch": 1.252557489606155,
      "grad_norm": 0.16185437142848969,
      "learning_rate": 5.825129573404484e-05,
      "loss": 0.6143,
      "mean_token_accuracy": 0.8103312142193317,
      "num_tokens": 434543929.0,
      "step": 13614
    },
    {
      "entropy": 0.5687246369197965,
      "epoch": 1.2526494959775964,
      "grad_norm": 0.15386071801185608,
      "learning_rate": 5.8248228907903214e-05,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8309563584625721,
      "num_tokens": 434576303.0,
      "step": 13615
    },
    {
      "entropy": 0.6424074955284595,
      "epoch": 1.2527415023490378,
      "grad_norm": 0.16532748937606812,
      "learning_rate": 5.824516208176158e-05,
      "loss": 0.6213,
      "mean_token_accuracy": 0.8094907663762569,
      "num_tokens": 434608449.0,
      "step": 13616
    },
    {
      "entropy": 0.6610729414969683,
      "epoch": 1.252833508720479,
      "grad_norm": 0.1580122411251068,
      "learning_rate": 5.824209525561996e-05,
      "loss": 0.6518,
      "mean_token_accuracy": 0.7996850423514843,
      "num_tokens": 434640146.0,
      "step": 13617
    },
    {
      "entropy": 0.6328689716756344,
      "epoch": 1.25292551509192,
      "grad_norm": 0.1583128124475479,
      "learning_rate": 5.823902842947834e-05,
      "loss": 0.6289,
      "mean_token_accuracy": 0.8111495934426785,
      "num_tokens": 434671957.0,
      "step": 13618
    },
    {
      "entropy": 0.6377339735627174,
      "epoch": 1.2530175214633612,
      "grad_norm": 0.15829838812351227,
      "learning_rate": 5.8235961603336706e-05,
      "loss": 0.6302,
      "mean_token_accuracy": 0.8068484589457512,
      "num_tokens": 434704279.0,
      "step": 13619
    },
    {
      "entropy": 0.7601367719471455,
      "epoch": 1.2531095278348026,
      "grad_norm": 0.16766640543937683,
      "learning_rate": 5.823289477719508e-05,
      "loss": 0.7652,
      "mean_token_accuracy": 0.7722020521759987,
      "num_tokens": 434735636.0,
      "step": 13620
    },
    {
      "entropy": 0.6200429936870933,
      "epoch": 1.253201534206244,
      "grad_norm": 0.15310364961624146,
      "learning_rate": 5.822982795105346e-05,
      "loss": 0.5906,
      "mean_token_accuracy": 0.8164769411087036,
      "num_tokens": 434767257.0,
      "step": 13621
    },
    {
      "entropy": 0.6331547386944294,
      "epoch": 1.253293540577685,
      "grad_norm": 0.1562199741601944,
      "learning_rate": 5.822676112491183e-05,
      "loss": 0.6159,
      "mean_token_accuracy": 0.8146170899271965,
      "num_tokens": 434800000.0,
      "step": 13622
    },
    {
      "entropy": 0.5620622523128986,
      "epoch": 1.2533855469491262,
      "grad_norm": 0.15183494985103607,
      "learning_rate": 5.8223694298770204e-05,
      "loss": 0.5481,
      "mean_token_accuracy": 0.8336027972400188,
      "num_tokens": 434831412.0,
      "step": 13623
    },
    {
      "entropy": 0.5046822112053633,
      "epoch": 1.2534775533205673,
      "grad_norm": 0.15086548030376434,
      "learning_rate": 5.822062747262857e-05,
      "loss": 0.4968,
      "mean_token_accuracy": 0.8424895741045475,
      "num_tokens": 434862680.0,
      "step": 13624
    },
    {
      "entropy": 0.7350164316594601,
      "epoch": 1.2535695596920087,
      "grad_norm": 0.16614554822444916,
      "learning_rate": 5.821756064648696e-05,
      "loss": 0.7416,
      "mean_token_accuracy": 0.7771963551640511,
      "num_tokens": 434895107.0,
      "step": 13625
    },
    {
      "entropy": 0.4535769382491708,
      "epoch": 1.25366156606345,
      "grad_norm": 0.15286265313625336,
      "learning_rate": 5.821449382034533e-05,
      "loss": 0.4235,
      "mean_token_accuracy": 0.8690435811877251,
      "num_tokens": 434926515.0,
      "step": 13626
    },
    {
      "entropy": 0.690495241433382,
      "epoch": 1.2537535724348912,
      "grad_norm": 0.15991295874118805,
      "learning_rate": 5.8211426994203696e-05,
      "loss": 0.6857,
      "mean_token_accuracy": 0.7944705300033092,
      "num_tokens": 434958587.0,
      "step": 13627
    },
    {
      "entropy": 0.6072365585714579,
      "epoch": 1.2538455788063323,
      "grad_norm": 0.16131538152694702,
      "learning_rate": 5.820836016806207e-05,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8207561261951923,
      "num_tokens": 434990694.0,
      "step": 13628
    },
    {
      "entropy": 0.5765265468508005,
      "epoch": 1.2539375851777734,
      "grad_norm": 0.15532363951206207,
      "learning_rate": 5.820529334192045e-05,
      "loss": 0.5632,
      "mean_token_accuracy": 0.8255302868783474,
      "num_tokens": 435022751.0,
      "step": 13629
    },
    {
      "entropy": 0.5916787218302488,
      "epoch": 1.2540295915492148,
      "grad_norm": 0.15699619054794312,
      "learning_rate": 5.820222651577883e-05,
      "loss": 0.5792,
      "mean_token_accuracy": 0.8206822089850903,
      "num_tokens": 435054583.0,
      "step": 13630
    },
    {
      "entropy": 0.5737063260748982,
      "epoch": 1.2541215979206561,
      "grad_norm": 0.16422803699970245,
      "learning_rate": 5.8199159689637195e-05,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8264249265193939,
      "num_tokens": 435085980.0,
      "step": 13631
    },
    {
      "entropy": 0.5542237930931151,
      "epoch": 1.2542136042920973,
      "grad_norm": 0.1538374423980713,
      "learning_rate": 5.819609286349556e-05,
      "loss": 0.5268,
      "mean_token_accuracy": 0.8390670120716095,
      "num_tokens": 435117863.0,
      "step": 13632
    },
    {
      "entropy": 0.6469243662431836,
      "epoch": 1.2543056106635384,
      "grad_norm": 0.17331239581108093,
      "learning_rate": 5.819302603735395e-05,
      "loss": 0.6459,
      "mean_token_accuracy": 0.8025184758007526,
      "num_tokens": 435150044.0,
      "step": 13633
    },
    {
      "entropy": 0.6178058413788676,
      "epoch": 1.2543976170349795,
      "grad_norm": 0.1676267832517624,
      "learning_rate": 5.818995921121232e-05,
      "loss": 0.6005,
      "mean_token_accuracy": 0.8149045445024967,
      "num_tokens": 435181661.0,
      "step": 13634
    },
    {
      "entropy": 0.7169063650071621,
      "epoch": 1.254489623406421,
      "grad_norm": 0.17151640355587006,
      "learning_rate": 5.818689238507069e-05,
      "loss": 0.7197,
      "mean_token_accuracy": 0.7799645736813545,
      "num_tokens": 435213863.0,
      "step": 13635
    },
    {
      "entropy": 0.5815444542095065,
      "epoch": 1.2545816297778623,
      "grad_norm": 0.15722180902957916,
      "learning_rate": 5.8183825558929075e-05,
      "loss": 0.577,
      "mean_token_accuracy": 0.8260765410959721,
      "num_tokens": 435245696.0,
      "step": 13636
    },
    {
      "entropy": 0.6548863518983126,
      "epoch": 1.2546736361493034,
      "grad_norm": 0.15795059502124786,
      "learning_rate": 5.818075873278744e-05,
      "loss": 0.6337,
      "mean_token_accuracy": 0.8069775514304638,
      "num_tokens": 435278207.0,
      "step": 13637
    },
    {
      "entropy": 0.6099499766714871,
      "epoch": 1.2547656425207445,
      "grad_norm": 0.15754881501197815,
      "learning_rate": 5.817769190664582e-05,
      "loss": 0.6135,
      "mean_token_accuracy": 0.8145431280136108,
      "num_tokens": 435309919.0,
      "step": 13638
    },
    {
      "entropy": 0.5571518521755934,
      "epoch": 1.2548576488921857,
      "grad_norm": 0.15907052159309387,
      "learning_rate": 5.8174625080504185e-05,
      "loss": 0.5377,
      "mean_token_accuracy": 0.8327736221253872,
      "num_tokens": 435341316.0,
      "step": 13639
    },
    {
      "entropy": 0.6047813538461924,
      "epoch": 1.254949655263627,
      "grad_norm": 0.159969300031662,
      "learning_rate": 5.8171558254362567e-05,
      "loss": 0.5808,
      "mean_token_accuracy": 0.8212082013487816,
      "num_tokens": 435371983.0,
      "step": 13640
    },
    {
      "entropy": 0.6903152912855148,
      "epoch": 1.2550416616350684,
      "grad_norm": 0.16970178484916687,
      "learning_rate": 5.816849142822094e-05,
      "loss": 0.6762,
      "mean_token_accuracy": 0.7924971356987953,
      "num_tokens": 435404065.0,
      "step": 13641
    },
    {
      "entropy": 0.6632866561412811,
      "epoch": 1.2551336680065095,
      "grad_norm": 0.16707375645637512,
      "learning_rate": 5.816542460207931e-05,
      "loss": 0.635,
      "mean_token_accuracy": 0.8061380200088024,
      "num_tokens": 435436161.0,
      "step": 13642
    },
    {
      "entropy": 0.648351401090622,
      "epoch": 1.2552256743779506,
      "grad_norm": 0.15764057636260986,
      "learning_rate": 5.8162357775937684e-05,
      "loss": 0.6425,
      "mean_token_accuracy": 0.8063574358820915,
      "num_tokens": 435468491.0,
      "step": 13643
    },
    {
      "entropy": 0.5398991145193577,
      "epoch": 1.2553176807493918,
      "grad_norm": 0.1520121842622757,
      "learning_rate": 5.8159290949796065e-05,
      "loss": 0.522,
      "mean_token_accuracy": 0.8354054801166058,
      "num_tokens": 435500557.0,
      "step": 13644
    },
    {
      "entropy": 0.5607970524579287,
      "epoch": 1.2554096871208331,
      "grad_norm": 0.14940722286701202,
      "learning_rate": 5.815622412365443e-05,
      "loss": 0.5327,
      "mean_token_accuracy": 0.8338676989078522,
      "num_tokens": 435532323.0,
      "step": 13645
    },
    {
      "entropy": 0.6134951133280993,
      "epoch": 1.2555016934922745,
      "grad_norm": 0.15334729850292206,
      "learning_rate": 5.815315729751281e-05,
      "loss": 0.6013,
      "mean_token_accuracy": 0.8160414695739746,
      "num_tokens": 435564863.0,
      "step": 13646
    },
    {
      "entropy": 0.6523516476154327,
      "epoch": 1.2555936998637156,
      "grad_norm": 0.16623999178409576,
      "learning_rate": 5.8150090471371176e-05,
      "loss": 0.6539,
      "mean_token_accuracy": 0.8021928891539574,
      "num_tokens": 435597382.0,
      "step": 13647
    },
    {
      "entropy": 0.6487322021275759,
      "epoch": 1.2556857062351567,
      "grad_norm": 0.15549160540103912,
      "learning_rate": 5.814702364522956e-05,
      "loss": 0.6355,
      "mean_token_accuracy": 0.8048530407249928,
      "num_tokens": 435629464.0,
      "step": 13648
    },
    {
      "entropy": 0.5477808117866516,
      "epoch": 1.2557777126065979,
      "grad_norm": 0.151653453707695,
      "learning_rate": 5.814395681908793e-05,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8316429704427719,
      "num_tokens": 435661085.0,
      "step": 13649
    },
    {
      "entropy": 0.5554218746256083,
      "epoch": 1.2558697189780392,
      "grad_norm": 0.16039544343948364,
      "learning_rate": 5.81408899929463e-05,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8348941579461098,
      "num_tokens": 435693759.0,
      "step": 13650
    },
    {
      "entropy": 0.6816946798935533,
      "epoch": 1.2559617253494806,
      "grad_norm": 0.15961383283138275,
      "learning_rate": 5.8137823166804674e-05,
      "loss": 0.6705,
      "mean_token_accuracy": 0.7961686104536057,
      "num_tokens": 435725983.0,
      "step": 13651
    },
    {
      "entropy": 0.5324064139276743,
      "epoch": 1.2560537317209217,
      "grad_norm": 0.16104111075401306,
      "learning_rate": 5.8134756340663056e-05,
      "loss": 0.5171,
      "mean_token_accuracy": 0.843619741499424,
      "num_tokens": 435758524.0,
      "step": 13652
    },
    {
      "entropy": 0.4992150692269206,
      "epoch": 1.2561457380923629,
      "grad_norm": 0.1461208164691925,
      "learning_rate": 5.8131689514521423e-05,
      "loss": 0.4941,
      "mean_token_accuracy": 0.8505524694919586,
      "num_tokens": 435790780.0,
      "step": 13653
    },
    {
      "entropy": 0.6311277132481337,
      "epoch": 1.256237744463804,
      "grad_norm": 0.16256491839885712,
      "learning_rate": 5.81286226883798e-05,
      "loss": 0.6259,
      "mean_token_accuracy": 0.8056371062994003,
      "num_tokens": 435821599.0,
      "step": 13654
    },
    {
      "entropy": 0.656229505315423,
      "epoch": 1.2563297508352453,
      "grad_norm": 0.1670663058757782,
      "learning_rate": 5.8125555862238166e-05,
      "loss": 0.6422,
      "mean_token_accuracy": 0.8041480630636215,
      "num_tokens": 435853041.0,
      "step": 13655
    },
    {
      "entropy": 0.5945656355470419,
      "epoch": 1.2564217572066867,
      "grad_norm": 0.1514093279838562,
      "learning_rate": 5.812248903609655e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8246347196400166,
      "num_tokens": 435885026.0,
      "step": 13656
    },
    {
      "entropy": 0.6275304760783911,
      "epoch": 1.2565137635781278,
      "grad_norm": 0.1531461924314499,
      "learning_rate": 5.811942220995492e-05,
      "loss": 0.5962,
      "mean_token_accuracy": 0.8172629922628403,
      "num_tokens": 435916604.0,
      "step": 13657
    },
    {
      "entropy": 0.5275810770690441,
      "epoch": 1.256605769949569,
      "grad_norm": 0.14519594609737396,
      "learning_rate": 5.811635538381329e-05,
      "loss": 0.521,
      "mean_token_accuracy": 0.8366330862045288,
      "num_tokens": 435948700.0,
      "step": 13658
    },
    {
      "entropy": 0.6745608430355787,
      "epoch": 1.25669777632101,
      "grad_norm": 0.1546560674905777,
      "learning_rate": 5.8113288557671665e-05,
      "loss": 0.6563,
      "mean_token_accuracy": 0.796975564211607,
      "num_tokens": 435980868.0,
      "step": 13659
    },
    {
      "entropy": 0.5804679868742824,
      "epoch": 1.2567897826924515,
      "grad_norm": 0.15185654163360596,
      "learning_rate": 5.8110221731530046e-05,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8243133202195168,
      "num_tokens": 436012439.0,
      "step": 13660
    },
    {
      "entropy": 0.5385894356295466,
      "epoch": 1.2568817890638928,
      "grad_norm": 0.1520770937204361,
      "learning_rate": 5.8107154905388414e-05,
      "loss": 0.5211,
      "mean_token_accuracy": 0.8389720879495144,
      "num_tokens": 436044282.0,
      "step": 13661
    },
    {
      "entropy": 0.5961839761584997,
      "epoch": 1.256973795435334,
      "grad_norm": 0.15826700627803802,
      "learning_rate": 5.810408807924679e-05,
      "loss": 0.5906,
      "mean_token_accuracy": 0.8164738900959492,
      "num_tokens": 436076291.0,
      "step": 13662
    },
    {
      "entropy": 0.6370399966835976,
      "epoch": 1.257065801806775,
      "grad_norm": 0.15838630497455597,
      "learning_rate": 5.8101021253105156e-05,
      "loss": 0.6312,
      "mean_token_accuracy": 0.8057442866265774,
      "num_tokens": 436107489.0,
      "step": 13663
    },
    {
      "entropy": 0.6417255224660039,
      "epoch": 1.2571578081782162,
      "grad_norm": 0.16165275871753693,
      "learning_rate": 5.809795442696354e-05,
      "loss": 0.6345,
      "mean_token_accuracy": 0.8025124408304691,
      "num_tokens": 436139705.0,
      "step": 13664
    },
    {
      "entropy": 0.5125674800947309,
      "epoch": 1.2572498145496576,
      "grad_norm": 0.15870796144008636,
      "learning_rate": 5.809488760082191e-05,
      "loss": 0.4999,
      "mean_token_accuracy": 0.844484843313694,
      "num_tokens": 436172232.0,
      "step": 13665
    },
    {
      "entropy": 0.6268217032775283,
      "epoch": 1.257341820921099,
      "grad_norm": 0.1629694253206253,
      "learning_rate": 5.809182077468028e-05,
      "loss": 0.6223,
      "mean_token_accuracy": 0.8087681941688061,
      "num_tokens": 436204099.0,
      "step": 13666
    },
    {
      "entropy": 0.7425359562039375,
      "epoch": 1.25743382729254,
      "grad_norm": 0.17244724929332733,
      "learning_rate": 5.808875394853867e-05,
      "loss": 0.7357,
      "mean_token_accuracy": 0.7784502729773521,
      "num_tokens": 436236194.0,
      "step": 13667
    },
    {
      "entropy": 0.5878154039382935,
      "epoch": 1.2575258336639812,
      "grad_norm": 0.14707857370376587,
      "learning_rate": 5.8085687122397036e-05,
      "loss": 0.5736,
      "mean_token_accuracy": 0.825842697173357,
      "num_tokens": 436268962.0,
      "step": 13668
    },
    {
      "entropy": 0.6562045160681009,
      "epoch": 1.2576178400354223,
      "grad_norm": 0.167812779545784,
      "learning_rate": 5.808262029625541e-05,
      "loss": 0.6421,
      "mean_token_accuracy": 0.8023116551339626,
      "num_tokens": 436300919.0,
      "step": 13669
    },
    {
      "entropy": 0.617229632101953,
      "epoch": 1.2577098464068637,
      "grad_norm": 0.16002872586250305,
      "learning_rate": 5.807955347011378e-05,
      "loss": 0.6142,
      "mean_token_accuracy": 0.8117799274623394,
      "num_tokens": 436332920.0,
      "step": 13670
    },
    {
      "entropy": 0.6809809934347868,
      "epoch": 1.257801852778305,
      "grad_norm": 0.16578036546707153,
      "learning_rate": 5.807648664397216e-05,
      "loss": 0.6773,
      "mean_token_accuracy": 0.7949528284370899,
      "num_tokens": 436364810.0,
      "step": 13671
    },
    {
      "entropy": 0.48816477367654443,
      "epoch": 1.2578938591497462,
      "grad_norm": 0.15387794375419617,
      "learning_rate": 5.8073419817830535e-05,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8503233194351196,
      "num_tokens": 436396537.0,
      "step": 13672
    },
    {
      "entropy": 0.7457143012434244,
      "epoch": 1.2579858655211873,
      "grad_norm": 0.16682220995426178,
      "learning_rate": 5.80703529916889e-05,
      "loss": 0.7428,
      "mean_token_accuracy": 0.7788374833762646,
      "num_tokens": 436428405.0,
      "step": 13673
    },
    {
      "entropy": 0.6844955421984196,
      "epoch": 1.2580778718926284,
      "grad_norm": 0.16952720284461975,
      "learning_rate": 5.806728616554728e-05,
      "loss": 0.6845,
      "mean_token_accuracy": 0.7922520935535431,
      "num_tokens": 436459260.0,
      "step": 13674
    },
    {
      "entropy": 0.6277901027351618,
      "epoch": 1.2581698782640698,
      "grad_norm": 0.15775327384471893,
      "learning_rate": 5.806421933940566e-05,
      "loss": 0.6068,
      "mean_token_accuracy": 0.8096248544752598,
      "num_tokens": 436490763.0,
      "step": 13675
    },
    {
      "entropy": 0.5837818700820208,
      "epoch": 1.2582618846355111,
      "grad_norm": 0.14718598127365112,
      "learning_rate": 5.806115251326403e-05,
      "loss": 0.5589,
      "mean_token_accuracy": 0.8289042450487614,
      "num_tokens": 436522433.0,
      "step": 13676
    },
    {
      "entropy": 0.6897852420806885,
      "epoch": 1.2583538910069523,
      "grad_norm": 0.16030238568782806,
      "learning_rate": 5.80580856871224e-05,
      "loss": 0.6635,
      "mean_token_accuracy": 0.7934593819081783,
      "num_tokens": 436554070.0,
      "step": 13677
    },
    {
      "entropy": 0.6991717759519815,
      "epoch": 1.2584458973783934,
      "grad_norm": 0.16114352643489838,
      "learning_rate": 5.805501886098077e-05,
      "loss": 0.6893,
      "mean_token_accuracy": 0.7939013764262199,
      "num_tokens": 436586085.0,
      "step": 13678
    },
    {
      "entropy": 0.6184375127777457,
      "epoch": 1.2585379037498345,
      "grad_norm": 0.15728577971458435,
      "learning_rate": 5.805195203483915e-05,
      "loss": 0.6163,
      "mean_token_accuracy": 0.8129852078855038,
      "num_tokens": 436618062.0,
      "step": 13679
    },
    {
      "entropy": 0.604900823906064,
      "epoch": 1.258629910121276,
      "grad_norm": 0.15387053787708282,
      "learning_rate": 5.8048885208697525e-05,
      "loss": 0.5847,
      "mean_token_accuracy": 0.825279738754034,
      "num_tokens": 436650575.0,
      "step": 13680
    },
    {
      "entropy": 0.6279559843242168,
      "epoch": 1.2587219164927173,
      "grad_norm": 0.15896327793598175,
      "learning_rate": 5.804581838255589e-05,
      "loss": 0.6081,
      "mean_token_accuracy": 0.8137033991515636,
      "num_tokens": 436682150.0,
      "step": 13681
    },
    {
      "entropy": 0.6318195350468159,
      "epoch": 1.2588139228641584,
      "grad_norm": 0.15740148723125458,
      "learning_rate": 5.804275155641427e-05,
      "loss": 0.6212,
      "mean_token_accuracy": 0.8106242902576923,
      "num_tokens": 436713969.0,
      "step": 13682
    },
    {
      "entropy": 0.5899330638349056,
      "epoch": 1.2589059292355995,
      "grad_norm": 0.15041781961917877,
      "learning_rate": 5.803968473027265e-05,
      "loss": 0.5824,
      "mean_token_accuracy": 0.8250339962542057,
      "num_tokens": 436745715.0,
      "step": 13683
    },
    {
      "entropy": 0.6316450517624617,
      "epoch": 1.2589979356070407,
      "grad_norm": 0.16030316054821014,
      "learning_rate": 5.803661790413102e-05,
      "loss": 0.6204,
      "mean_token_accuracy": 0.8085256665945053,
      "num_tokens": 436777391.0,
      "step": 13684
    },
    {
      "entropy": 0.6858752369880676,
      "epoch": 1.259089941978482,
      "grad_norm": 0.16567544639110565,
      "learning_rate": 5.803355107798939e-05,
      "loss": 0.6837,
      "mean_token_accuracy": 0.7938522696495056,
      "num_tokens": 436809627.0,
      "step": 13685
    },
    {
      "entropy": 0.6066624373197556,
      "epoch": 1.2591819483499234,
      "grad_norm": 0.15981805324554443,
      "learning_rate": 5.803048425184776e-05,
      "loss": 0.6002,
      "mean_token_accuracy": 0.8136745616793633,
      "num_tokens": 436841478.0,
      "step": 13686
    },
    {
      "entropy": 0.6505708508193493,
      "epoch": 1.2592739547213645,
      "grad_norm": 0.1591271311044693,
      "learning_rate": 5.802741742570614e-05,
      "loss": 0.6378,
      "mean_token_accuracy": 0.8060159794986248,
      "num_tokens": 436873201.0,
      "step": 13687
    },
    {
      "entropy": 0.659436721354723,
      "epoch": 1.2593659610928056,
      "grad_norm": 0.15871307253837585,
      "learning_rate": 5.8024350599564516e-05,
      "loss": 0.6637,
      "mean_token_accuracy": 0.799128532409668,
      "num_tokens": 436905406.0,
      "step": 13688
    },
    {
      "entropy": 0.7083176057785749,
      "epoch": 1.2594579674642468,
      "grad_norm": 0.1756422370672226,
      "learning_rate": 5.8021283773422884e-05,
      "loss": 0.7067,
      "mean_token_accuracy": 0.7909508123993874,
      "num_tokens": 436936683.0,
      "step": 13689
    },
    {
      "entropy": 0.6399471312761307,
      "epoch": 1.2595499738356881,
      "grad_norm": 0.15628892183303833,
      "learning_rate": 5.801821694728126e-05,
      "loss": 0.6334,
      "mean_token_accuracy": 0.8068052195012569,
      "num_tokens": 436968985.0,
      "step": 13690
    },
    {
      "entropy": 0.6842237245291471,
      "epoch": 1.2596419802071295,
      "grad_norm": 0.1685868352651596,
      "learning_rate": 5.801515012113964e-05,
      "loss": 0.6651,
      "mean_token_accuracy": 0.7957841157913208,
      "num_tokens": 437000473.0,
      "step": 13691
    },
    {
      "entropy": 0.7616059444844723,
      "epoch": 1.2597339865785706,
      "grad_norm": 0.17196066677570343,
      "learning_rate": 5.801208329499801e-05,
      "loss": 0.7618,
      "mean_token_accuracy": 0.7682157531380653,
      "num_tokens": 437031704.0,
      "step": 13692
    },
    {
      "entropy": 0.6974148098379374,
      "epoch": 1.2598259929500117,
      "grad_norm": 0.16710494458675385,
      "learning_rate": 5.800901646885638e-05,
      "loss": 0.7014,
      "mean_token_accuracy": 0.787652999162674,
      "num_tokens": 437063066.0,
      "step": 13693
    },
    {
      "entropy": 0.7034330125898123,
      "epoch": 1.2599179993214529,
      "grad_norm": 0.1627243459224701,
      "learning_rate": 5.800594964271475e-05,
      "loss": 0.6788,
      "mean_token_accuracy": 0.7947148494422436,
      "num_tokens": 437095105.0,
      "step": 13694
    },
    {
      "entropy": 0.6151662520132959,
      "epoch": 1.2600100056928942,
      "grad_norm": 0.1584627479314804,
      "learning_rate": 5.800288281657313e-05,
      "loss": 0.6078,
      "mean_token_accuracy": 0.8148611858487129,
      "num_tokens": 437126681.0,
      "step": 13695
    },
    {
      "entropy": 0.65718699619174,
      "epoch": 1.2601020120643356,
      "grad_norm": 0.1549854576587677,
      "learning_rate": 5.7999815990431506e-05,
      "loss": 0.646,
      "mean_token_accuracy": 0.80357675999403,
      "num_tokens": 437158273.0,
      "step": 13696
    },
    {
      "entropy": 0.6510269418358803,
      "epoch": 1.2601940184357767,
      "grad_norm": 0.1586599498987198,
      "learning_rate": 5.7996749164289874e-05,
      "loss": 0.6401,
      "mean_token_accuracy": 0.8094366751611233,
      "num_tokens": 437190132.0,
      "step": 13697
    },
    {
      "entropy": 0.7824202496558428,
      "epoch": 1.2602860248072179,
      "grad_norm": 0.16092392802238464,
      "learning_rate": 5.7993682338148256e-05,
      "loss": 0.7623,
      "mean_token_accuracy": 0.7709265388548374,
      "num_tokens": 437222651.0,
      "step": 13698
    },
    {
      "entropy": 0.6032309448346496,
      "epoch": 1.260378031178659,
      "grad_norm": 0.15560069680213928,
      "learning_rate": 5.799061551200663e-05,
      "loss": 0.5842,
      "mean_token_accuracy": 0.8212880231440067,
      "num_tokens": 437254548.0,
      "step": 13699
    },
    {
      "entropy": 0.6152170039713383,
      "epoch": 1.2604700375501003,
      "grad_norm": 0.15077722072601318,
      "learning_rate": 5.7987548685865e-05,
      "loss": 0.5935,
      "mean_token_accuracy": 0.8138209618628025,
      "num_tokens": 437286417.0,
      "step": 13700
    },
    {
      "entropy": 0.6444226931780577,
      "epoch": 1.2605620439215417,
      "grad_norm": 0.15743428468704224,
      "learning_rate": 5.798448185972337e-05,
      "loss": 0.6344,
      "mean_token_accuracy": 0.8081530071794987,
      "num_tokens": 437318811.0,
      "step": 13701
    },
    {
      "entropy": 0.6261066533625126,
      "epoch": 1.2606540502929828,
      "grad_norm": 0.1607842743396759,
      "learning_rate": 5.7981415033581754e-05,
      "loss": 0.6102,
      "mean_token_accuracy": 0.8134374730288982,
      "num_tokens": 437350524.0,
      "step": 13702
    },
    {
      "entropy": 0.5927297053858638,
      "epoch": 1.260746056664424,
      "grad_norm": 0.1764361411333084,
      "learning_rate": 5.797834820744012e-05,
      "loss": 0.5823,
      "mean_token_accuracy": 0.8252977021038532,
      "num_tokens": 437382272.0,
      "step": 13703
    },
    {
      "entropy": 0.6681660795584321,
      "epoch": 1.260838063035865,
      "grad_norm": 0.15843342244625092,
      "learning_rate": 5.7975281381298497e-05,
      "loss": 0.6715,
      "mean_token_accuracy": 0.7967858389019966,
      "num_tokens": 437414876.0,
      "step": 13704
    },
    {
      "entropy": 0.5811425466090441,
      "epoch": 1.2609300694073065,
      "grad_norm": 0.1580074429512024,
      "learning_rate": 5.7972214555156864e-05,
      "loss": 0.5633,
      "mean_token_accuracy": 0.8302118293941021,
      "num_tokens": 437447199.0,
      "step": 13705
    },
    {
      "entropy": 0.4959310023114085,
      "epoch": 1.2610220757787478,
      "grad_norm": 0.14028817415237427,
      "learning_rate": 5.796914772901525e-05,
      "loss": 0.4769,
      "mean_token_accuracy": 0.8501442708075047,
      "num_tokens": 437478867.0,
      "step": 13706
    },
    {
      "entropy": 0.590019004419446,
      "epoch": 1.261114082150189,
      "grad_norm": 0.1543416529893875,
      "learning_rate": 5.796608090287362e-05,
      "loss": 0.5658,
      "mean_token_accuracy": 0.8218632414937019,
      "num_tokens": 437510292.0,
      "step": 13707
    },
    {
      "entropy": 0.5404981728643179,
      "epoch": 1.26120608852163,
      "grad_norm": 0.1579848974943161,
      "learning_rate": 5.796301407673199e-05,
      "loss": 0.538,
      "mean_token_accuracy": 0.8332445994019508,
      "num_tokens": 437542073.0,
      "step": 13708
    },
    {
      "entropy": 0.5911095780320466,
      "epoch": 1.2612980948930712,
      "grad_norm": 0.15382631123065948,
      "learning_rate": 5.795994725059036e-05,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8210938684642315,
      "num_tokens": 437574060.0,
      "step": 13709
    },
    {
      "entropy": 0.5179063575342298,
      "epoch": 1.2613901012645126,
      "grad_norm": 0.14957548677921295,
      "learning_rate": 5.7956880424448744e-05,
      "loss": 0.5161,
      "mean_token_accuracy": 0.839840468019247,
      "num_tokens": 437605390.0,
      "step": 13710
    },
    {
      "entropy": 0.7480579558759928,
      "epoch": 1.261482107635954,
      "grad_norm": 0.16457128524780273,
      "learning_rate": 5.795381359830712e-05,
      "loss": 0.7465,
      "mean_token_accuracy": 0.7777762115001678,
      "num_tokens": 437637350.0,
      "step": 13711
    },
    {
      "entropy": 0.5635834722779691,
      "epoch": 1.261574114007395,
      "grad_norm": 0.1625271439552307,
      "learning_rate": 5.795074677216549e-05,
      "loss": 0.5579,
      "mean_token_accuracy": 0.825296375900507,
      "num_tokens": 437669111.0,
      "step": 13712
    },
    {
      "entropy": 0.6037453711032867,
      "epoch": 1.2616661203788362,
      "grad_norm": 0.1583925038576126,
      "learning_rate": 5.794767994602386e-05,
      "loss": 0.6002,
      "mean_token_accuracy": 0.815705519169569,
      "num_tokens": 437701152.0,
      "step": 13713
    },
    {
      "entropy": 0.5784766646102071,
      "epoch": 1.2617581267502773,
      "grad_norm": 0.15420205891132355,
      "learning_rate": 5.794461311988224e-05,
      "loss": 0.554,
      "mean_token_accuracy": 0.8271432407200336,
      "num_tokens": 437732977.0,
      "step": 13714
    },
    {
      "entropy": 0.4909834433346987,
      "epoch": 1.2618501331217187,
      "grad_norm": 0.15331459045410156,
      "learning_rate": 5.794154629374061e-05,
      "loss": 0.4763,
      "mean_token_accuracy": 0.8518258295953274,
      "num_tokens": 437764984.0,
      "step": 13715
    },
    {
      "entropy": 0.6318513322621584,
      "epoch": 1.26194213949316,
      "grad_norm": 0.16049198806285858,
      "learning_rate": 5.7938479467598986e-05,
      "loss": 0.6117,
      "mean_token_accuracy": 0.8129026256501675,
      "num_tokens": 437796441.0,
      "step": 13716
    },
    {
      "entropy": 0.5694343615323305,
      "epoch": 1.2620341458646012,
      "grad_norm": 0.15513081848621368,
      "learning_rate": 5.7935412641457353e-05,
      "loss": 0.5625,
      "mean_token_accuracy": 0.8271933756768703,
      "num_tokens": 437829002.0,
      "step": 13717
    },
    {
      "entropy": 0.6629965286701918,
      "epoch": 1.2621261522360423,
      "grad_norm": 0.16043683886528015,
      "learning_rate": 5.7932345815315735e-05,
      "loss": 0.6527,
      "mean_token_accuracy": 0.8001281134784222,
      "num_tokens": 437861201.0,
      "step": 13718
    },
    {
      "entropy": 0.6111866626888514,
      "epoch": 1.2622181586074834,
      "grad_norm": 0.16301684081554413,
      "learning_rate": 5.792927898917411e-05,
      "loss": 0.5938,
      "mean_token_accuracy": 0.8203914351761341,
      "num_tokens": 437893300.0,
      "step": 13719
    },
    {
      "entropy": 0.5580829447135329,
      "epoch": 1.2623101649789248,
      "grad_norm": 0.15420153737068176,
      "learning_rate": 5.792621216303248e-05,
      "loss": 0.5538,
      "mean_token_accuracy": 0.8314461708068848,
      "num_tokens": 437924987.0,
      "step": 13720
    },
    {
      "entropy": 0.7987187765538692,
      "epoch": 1.2624021713503661,
      "grad_norm": 0.17007331550121307,
      "learning_rate": 5.792314533689085e-05,
      "loss": 0.8081,
      "mean_token_accuracy": 0.762097679078579,
      "num_tokens": 437956775.0,
      "step": 13721
    },
    {
      "entropy": 0.6621308773756027,
      "epoch": 1.2624941777218073,
      "grad_norm": 0.15915058553218842,
      "learning_rate": 5.7920078510749233e-05,
      "loss": 0.6596,
      "mean_token_accuracy": 0.7980025112628937,
      "num_tokens": 437988988.0,
      "step": 13722
    },
    {
      "entropy": 0.6815325375646353,
      "epoch": 1.2625861840932484,
      "grad_norm": 0.16443856060504913,
      "learning_rate": 5.79170116846076e-05,
      "loss": 0.6692,
      "mean_token_accuracy": 0.7929310724139214,
      "num_tokens": 438020582.0,
      "step": 13723
    },
    {
      "entropy": 0.611801678314805,
      "epoch": 1.2626781904646895,
      "grad_norm": 0.15760627388954163,
      "learning_rate": 5.7913944858465976e-05,
      "loss": 0.6135,
      "mean_token_accuracy": 0.8129325099289417,
      "num_tokens": 438053162.0,
      "step": 13724
    },
    {
      "entropy": 0.6250302214175463,
      "epoch": 1.262770196836131,
      "grad_norm": 0.15671208500862122,
      "learning_rate": 5.7910878032324344e-05,
      "loss": 0.6093,
      "mean_token_accuracy": 0.8161828592419624,
      "num_tokens": 438085166.0,
      "step": 13725
    },
    {
      "entropy": 0.535996078979224,
      "epoch": 1.2628622032075723,
      "grad_norm": 0.1484290361404419,
      "learning_rate": 5.7907811206182725e-05,
      "loss": 0.511,
      "mean_token_accuracy": 0.8399997055530548,
      "num_tokens": 438116630.0,
      "step": 13726
    },
    {
      "entropy": 0.5999484313651919,
      "epoch": 1.2629542095790134,
      "grad_norm": 0.1809384673833847,
      "learning_rate": 5.79047443800411e-05,
      "loss": 0.6049,
      "mean_token_accuracy": 0.8168606795370579,
      "num_tokens": 438147215.0,
      "step": 13727
    },
    {
      "entropy": 0.7309411130845547,
      "epoch": 1.2630462159504545,
      "grad_norm": 0.16507674753665924,
      "learning_rate": 5.790167755389947e-05,
      "loss": 0.7257,
      "mean_token_accuracy": 0.7836959548294544,
      "num_tokens": 438178981.0,
      "step": 13728
    },
    {
      "entropy": 0.619749067351222,
      "epoch": 1.2631382223218957,
      "grad_norm": 0.15703168511390686,
      "learning_rate": 5.789861072775785e-05,
      "loss": 0.617,
      "mean_token_accuracy": 0.811744213104248,
      "num_tokens": 438210666.0,
      "step": 13729
    },
    {
      "entropy": 0.6262092581018806,
      "epoch": 1.263230228693337,
      "grad_norm": 0.15438854694366455,
      "learning_rate": 5.7895543901616224e-05,
      "loss": 0.6084,
      "mean_token_accuracy": 0.8138249963521957,
      "num_tokens": 438242920.0,
      "step": 13730
    },
    {
      "entropy": 0.550021992996335,
      "epoch": 1.2633222350647784,
      "grad_norm": 0.144331693649292,
      "learning_rate": 5.789247707547459e-05,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8353813290596008,
      "num_tokens": 438275185.0,
      "step": 13731
    },
    {
      "entropy": 0.6006852071732283,
      "epoch": 1.2634142414362195,
      "grad_norm": 0.150711789727211,
      "learning_rate": 5.7889410249332966e-05,
      "loss": 0.5778,
      "mean_token_accuracy": 0.8199211917817593,
      "num_tokens": 438307470.0,
      "step": 13732
    },
    {
      "entropy": 0.6936467550694942,
      "epoch": 1.2635062478076606,
      "grad_norm": 0.16807857155799866,
      "learning_rate": 5.788634342319135e-05,
      "loss": 0.676,
      "mean_token_accuracy": 0.7937901169061661,
      "num_tokens": 438339396.0,
      "step": 13733
    },
    {
      "entropy": 0.5733952317386866,
      "epoch": 1.2635982541791018,
      "grad_norm": 0.16455870866775513,
      "learning_rate": 5.7883276597049716e-05,
      "loss": 0.5755,
      "mean_token_accuracy": 0.8244554027915001,
      "num_tokens": 438371402.0,
      "step": 13734
    },
    {
      "entropy": 0.7072537019848824,
      "epoch": 1.2636902605505431,
      "grad_norm": 0.16083458065986633,
      "learning_rate": 5.788020977090809e-05,
      "loss": 0.7113,
      "mean_token_accuracy": 0.7834815122187138,
      "num_tokens": 438403195.0,
      "step": 13735
    },
    {
      "entropy": 0.6744230575859547,
      "epoch": 1.2637822669219845,
      "grad_norm": 0.1583990752696991,
      "learning_rate": 5.787714294476646e-05,
      "loss": 0.6747,
      "mean_token_accuracy": 0.7974455766379833,
      "num_tokens": 438435265.0,
      "step": 13736
    },
    {
      "entropy": 0.6464333217591047,
      "epoch": 1.2638742732934256,
      "grad_norm": 0.15935631096363068,
      "learning_rate": 5.787407611862484e-05,
      "loss": 0.6315,
      "mean_token_accuracy": 0.805219043046236,
      "num_tokens": 438466347.0,
      "step": 13737
    },
    {
      "entropy": 0.6575862150639296,
      "epoch": 1.2639662796648667,
      "grad_norm": 0.15972796082496643,
      "learning_rate": 5.7871009292483214e-05,
      "loss": 0.635,
      "mean_token_accuracy": 0.80533467233181,
      "num_tokens": 438498306.0,
      "step": 13738
    },
    {
      "entropy": 0.6307687070220709,
      "epoch": 1.2640582860363079,
      "grad_norm": 0.1547071635723114,
      "learning_rate": 5.786794246634158e-05,
      "loss": 0.6132,
      "mean_token_accuracy": 0.812846727669239,
      "num_tokens": 438530393.0,
      "step": 13739
    },
    {
      "entropy": 0.6325311362743378,
      "epoch": 1.2641502924077492,
      "grad_norm": 0.1573491096496582,
      "learning_rate": 5.786487564019996e-05,
      "loss": 0.6147,
      "mean_token_accuracy": 0.8123775310814381,
      "num_tokens": 438562360.0,
      "step": 13740
    },
    {
      "entropy": 0.5498445695266128,
      "epoch": 1.2642422987791906,
      "grad_norm": 0.1593708097934723,
      "learning_rate": 5.786180881405834e-05,
      "loss": 0.5356,
      "mean_token_accuracy": 0.8309050276875496,
      "num_tokens": 438594073.0,
      "step": 13741
    },
    {
      "entropy": 0.5159841305576265,
      "epoch": 1.2643343051506317,
      "grad_norm": 0.14529584348201752,
      "learning_rate": 5.7858741987916706e-05,
      "loss": 0.4882,
      "mean_token_accuracy": 0.8471213132143021,
      "num_tokens": 438626063.0,
      "step": 13742
    },
    {
      "entropy": 0.5967738665640354,
      "epoch": 1.2644263115220729,
      "grad_norm": 0.15732207894325256,
      "learning_rate": 5.785567516177508e-05,
      "loss": 0.5969,
      "mean_token_accuracy": 0.8148504793643951,
      "num_tokens": 438658138.0,
      "step": 13743
    },
    {
      "entropy": 0.6380666252225637,
      "epoch": 1.264518317893514,
      "grad_norm": 0.16364844143390656,
      "learning_rate": 5.785260833563345e-05,
      "loss": 0.6383,
      "mean_token_accuracy": 0.8086763918399811,
      "num_tokens": 438690089.0,
      "step": 13744
    },
    {
      "entropy": 0.5800126646645367,
      "epoch": 1.2646103242649553,
      "grad_norm": 0.15401408076286316,
      "learning_rate": 5.784954150949184e-05,
      "loss": 0.5798,
      "mean_token_accuracy": 0.8222378268837929,
      "num_tokens": 438722330.0,
      "step": 13745
    },
    {
      "entropy": 0.6779597923159599,
      "epoch": 1.2647023306363967,
      "grad_norm": 0.1594392955303192,
      "learning_rate": 5.7846474683350205e-05,
      "loss": 0.6623,
      "mean_token_accuracy": 0.7983884960412979,
      "num_tokens": 438754640.0,
      "step": 13746
    },
    {
      "entropy": 0.6125987693667412,
      "epoch": 1.2647943370078378,
      "grad_norm": 0.15857064723968506,
      "learning_rate": 5.784340785720857e-05,
      "loss": 0.5979,
      "mean_token_accuracy": 0.8155373483896255,
      "num_tokens": 438784685.0,
      "step": 13747
    },
    {
      "entropy": 0.6178328525274992,
      "epoch": 1.264886343379279,
      "grad_norm": 0.15251502394676208,
      "learning_rate": 5.784034103106695e-05,
      "loss": 0.6082,
      "mean_token_accuracy": 0.8116765692830086,
      "num_tokens": 438816879.0,
      "step": 13748
    },
    {
      "entropy": 0.6052141319960356,
      "epoch": 1.26497834975072,
      "grad_norm": 0.15773345530033112,
      "learning_rate": 5.783727420492533e-05,
      "loss": 0.5921,
      "mean_token_accuracy": 0.8171080984175205,
      "num_tokens": 438848902.0,
      "step": 13749
    },
    {
      "entropy": 0.6929060164839029,
      "epoch": 1.2650703561221615,
      "grad_norm": 0.1620091199874878,
      "learning_rate": 5.78342073787837e-05,
      "loss": 0.6715,
      "mean_token_accuracy": 0.7947232015430927,
      "num_tokens": 438880581.0,
      "step": 13750
    },
    {
      "entropy": 0.5717438710853457,
      "epoch": 1.2651623624936028,
      "grad_norm": 0.15394631028175354,
      "learning_rate": 5.783114055264207e-05,
      "loss": 0.5644,
      "mean_token_accuracy": 0.8252474144101143,
      "num_tokens": 438912207.0,
      "step": 13751
    },
    {
      "entropy": 0.67267812974751,
      "epoch": 1.265254368865044,
      "grad_norm": 0.16387198865413666,
      "learning_rate": 5.782807372650044e-05,
      "loss": 0.668,
      "mean_token_accuracy": 0.7966343238949776,
      "num_tokens": 438943667.0,
      "step": 13752
    },
    {
      "entropy": 0.58520183339715,
      "epoch": 1.265346375236485,
      "grad_norm": 0.15540838241577148,
      "learning_rate": 5.782500690035883e-05,
      "loss": 0.5706,
      "mean_token_accuracy": 0.8235589042305946,
      "num_tokens": 438975262.0,
      "step": 13753
    },
    {
      "entropy": 0.5135431690141559,
      "epoch": 1.2654383816079262,
      "grad_norm": 0.14271491765975952,
      "learning_rate": 5.7821940074217195e-05,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8461661450564861,
      "num_tokens": 439007120.0,
      "step": 13754
    },
    {
      "entropy": 0.5401266701519489,
      "epoch": 1.2655303879793676,
      "grad_norm": 0.1541070193052292,
      "learning_rate": 5.781887324807557e-05,
      "loss": 0.5424,
      "mean_token_accuracy": 0.8327365443110466,
      "num_tokens": 439039664.0,
      "step": 13755
    },
    {
      "entropy": 0.5715267099440098,
      "epoch": 1.265622394350809,
      "grad_norm": 0.15987472236156464,
      "learning_rate": 5.781580642193394e-05,
      "loss": 0.5693,
      "mean_token_accuracy": 0.8261524923145771,
      "num_tokens": 439070994.0,
      "step": 13756
    },
    {
      "entropy": 0.6786004714667797,
      "epoch": 1.26571440072225,
      "grad_norm": 0.16592134535312653,
      "learning_rate": 5.781273959579232e-05,
      "loss": 0.6789,
      "mean_token_accuracy": 0.7923277094960213,
      "num_tokens": 439102076.0,
      "step": 13757
    },
    {
      "entropy": 0.6381137333810329,
      "epoch": 1.2658064070936912,
      "grad_norm": 0.15315087139606476,
      "learning_rate": 5.7809672769650694e-05,
      "loss": 0.6259,
      "mean_token_accuracy": 0.8073647134006023,
      "num_tokens": 439134782.0,
      "step": 13758
    },
    {
      "entropy": 0.6173055893741548,
      "epoch": 1.2658984134651323,
      "grad_norm": 0.15498214960098267,
      "learning_rate": 5.780660594350906e-05,
      "loss": 0.6031,
      "mean_token_accuracy": 0.8111415319144726,
      "num_tokens": 439166603.0,
      "step": 13759
    },
    {
      "entropy": 0.6344747766852379,
      "epoch": 1.2659904198365737,
      "grad_norm": 0.16453328728675842,
      "learning_rate": 5.780353911736744e-05,
      "loss": 0.6171,
      "mean_token_accuracy": 0.8125417307019234,
      "num_tokens": 439198148.0,
      "step": 13760
    },
    {
      "entropy": 0.5708293532952666,
      "epoch": 1.266082426208015,
      "grad_norm": 0.15702217817306519,
      "learning_rate": 5.780047229122582e-05,
      "loss": 0.573,
      "mean_token_accuracy": 0.8258732296526432,
      "num_tokens": 439230916.0,
      "step": 13761
    },
    {
      "entropy": 0.53183813393116,
      "epoch": 1.2661744325794562,
      "grad_norm": 0.15059877932071686,
      "learning_rate": 5.7797405465084186e-05,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8385734669864178,
      "num_tokens": 439263657.0,
      "step": 13762
    },
    {
      "entropy": 0.5990386549383402,
      "epoch": 1.2662664389508973,
      "grad_norm": 0.15963777899742126,
      "learning_rate": 5.779433863894256e-05,
      "loss": 0.5785,
      "mean_token_accuracy": 0.8243114911019802,
      "num_tokens": 439295182.0,
      "step": 13763
    },
    {
      "entropy": 0.5023461068049073,
      "epoch": 1.2663584453223387,
      "grad_norm": 0.14742983877658844,
      "learning_rate": 5.779127181280094e-05,
      "loss": 0.481,
      "mean_token_accuracy": 0.8517886810004711,
      "num_tokens": 439327654.0,
      "step": 13764
    },
    {
      "entropy": 0.628225376829505,
      "epoch": 1.2664504516937798,
      "grad_norm": 0.15992984175682068,
      "learning_rate": 5.778820498665931e-05,
      "loss": 0.6117,
      "mean_token_accuracy": 0.8113311231136322,
      "num_tokens": 439359488.0,
      "step": 13765
    },
    {
      "entropy": 0.6744086258113384,
      "epoch": 1.2665424580652211,
      "grad_norm": 0.16198669373989105,
      "learning_rate": 5.7785138160517684e-05,
      "loss": 0.6522,
      "mean_token_accuracy": 0.8017571456730366,
      "num_tokens": 439390828.0,
      "step": 13766
    },
    {
      "entropy": 0.6353954020887613,
      "epoch": 1.2666344644366623,
      "grad_norm": 0.15533164143562317,
      "learning_rate": 5.778207133437605e-05,
      "loss": 0.6304,
      "mean_token_accuracy": 0.8063975311815739,
      "num_tokens": 439422855.0,
      "step": 13767
    },
    {
      "entropy": 0.6986438315361738,
      "epoch": 1.2667264708081034,
      "grad_norm": 0.16634538769721985,
      "learning_rate": 5.7779004508234433e-05,
      "loss": 0.6957,
      "mean_token_accuracy": 0.7895130775868893,
      "num_tokens": 439455003.0,
      "step": 13768
    },
    {
      "entropy": 0.5912506282329559,
      "epoch": 1.2668184771795448,
      "grad_norm": 0.15831808745861053,
      "learning_rate": 5.777593768209281e-05,
      "loss": 0.575,
      "mean_token_accuracy": 0.8252931125462055,
      "num_tokens": 439487771.0,
      "step": 13769
    },
    {
      "entropy": 0.5404855851083994,
      "epoch": 1.266910483550986,
      "grad_norm": 0.15732689201831818,
      "learning_rate": 5.7772870855951176e-05,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8361400105059147,
      "num_tokens": 439519406.0,
      "step": 13770
    },
    {
      "entropy": 0.55109510011971,
      "epoch": 1.2670024899224273,
      "grad_norm": 0.16485466063022614,
      "learning_rate": 5.776980402980955e-05,
      "loss": 0.5415,
      "mean_token_accuracy": 0.8304088115692139,
      "num_tokens": 439551247.0,
      "step": 13771
    },
    {
      "entropy": 0.5657030129805207,
      "epoch": 1.2670944962938684,
      "grad_norm": 0.15448838472366333,
      "learning_rate": 5.776673720366793e-05,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8313202150166035,
      "num_tokens": 439583643.0,
      "step": 13772
    },
    {
      "entropy": 0.6541113965213299,
      "epoch": 1.2671865026653095,
      "grad_norm": 0.15886668860912323,
      "learning_rate": 5.77636703775263e-05,
      "loss": 0.6332,
      "mean_token_accuracy": 0.8088391534984112,
      "num_tokens": 439615012.0,
      "step": 13773
    },
    {
      "entropy": 0.5947929695248604,
      "epoch": 1.2672785090367509,
      "grad_norm": 0.15301169455051422,
      "learning_rate": 5.7760603551384675e-05,
      "loss": 0.5626,
      "mean_token_accuracy": 0.823301937431097,
      "num_tokens": 439646275.0,
      "step": 13774
    },
    {
      "entropy": 0.5852472428232431,
      "epoch": 1.267370515408192,
      "grad_norm": 0.15075983107089996,
      "learning_rate": 5.775753672524304e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8242523483932018,
      "num_tokens": 439678123.0,
      "step": 13775
    },
    {
      "entropy": 0.6696411073207855,
      "epoch": 1.2674625217796334,
      "grad_norm": 0.16363607347011566,
      "learning_rate": 5.7754469899101424e-05,
      "loss": 0.657,
      "mean_token_accuracy": 0.8039701879024506,
      "num_tokens": 439709978.0,
      "step": 13776
    },
    {
      "entropy": 0.48017821181565523,
      "epoch": 1.2675545281510745,
      "grad_norm": 0.14806224405765533,
      "learning_rate": 5.77514030729598e-05,
      "loss": 0.4612,
      "mean_token_accuracy": 0.8567595668137074,
      "num_tokens": 439741840.0,
      "step": 13777
    },
    {
      "entropy": 0.6841826140880585,
      "epoch": 1.2676465345225156,
      "grad_norm": 0.16064894199371338,
      "learning_rate": 5.7748336246818166e-05,
      "loss": 0.6713,
      "mean_token_accuracy": 0.7966771461069584,
      "num_tokens": 439773549.0,
      "step": 13778
    },
    {
      "entropy": 0.7108347751200199,
      "epoch": 1.267738540893957,
      "grad_norm": 0.16078463196754456,
      "learning_rate": 5.774526942067654e-05,
      "loss": 0.6958,
      "mean_token_accuracy": 0.784707210958004,
      "num_tokens": 439805252.0,
      "step": 13779
    },
    {
      "entropy": 0.5878602098673582,
      "epoch": 1.2678305472653981,
      "grad_norm": 0.1499243676662445,
      "learning_rate": 5.774220259453492e-05,
      "loss": 0.5769,
      "mean_token_accuracy": 0.8213383071124554,
      "num_tokens": 439837150.0,
      "step": 13780
    },
    {
      "entropy": 0.5603157030418515,
      "epoch": 1.2679225536368395,
      "grad_norm": 0.16476786136627197,
      "learning_rate": 5.773913576839329e-05,
      "loss": 0.553,
      "mean_token_accuracy": 0.8286260962486267,
      "num_tokens": 439869084.0,
      "step": 13781
    },
    {
      "entropy": 0.6063054464757442,
      "epoch": 1.2680145600082806,
      "grad_norm": 0.154469832777977,
      "learning_rate": 5.7736068942251665e-05,
      "loss": 0.5927,
      "mean_token_accuracy": 0.8164966627955437,
      "num_tokens": 439901221.0,
      "step": 13782
    },
    {
      "entropy": 0.6917987745255232,
      "epoch": 1.2681065663797217,
      "grad_norm": 0.17053069174289703,
      "learning_rate": 5.773300211611003e-05,
      "loss": 0.6942,
      "mean_token_accuracy": 0.7879476137459278,
      "num_tokens": 439932692.0,
      "step": 13783
    },
    {
      "entropy": 0.6701811198145151,
      "epoch": 1.268198572751163,
      "grad_norm": 0.161770761013031,
      "learning_rate": 5.772993528996842e-05,
      "loss": 0.6552,
      "mean_token_accuracy": 0.7974055372178555,
      "num_tokens": 439965031.0,
      "step": 13784
    },
    {
      "entropy": 0.5110015431419015,
      "epoch": 1.2682905791226042,
      "grad_norm": 0.15486016869544983,
      "learning_rate": 5.772686846382679e-05,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8452044799923897,
      "num_tokens": 439997176.0,
      "step": 13785
    },
    {
      "entropy": 0.7225799364969134,
      "epoch": 1.2683825854940456,
      "grad_norm": 0.16789081692695618,
      "learning_rate": 5.772380163768516e-05,
      "loss": 0.726,
      "mean_token_accuracy": 0.7797723710536957,
      "num_tokens": 440029698.0,
      "step": 13786
    },
    {
      "entropy": 0.5401289416477084,
      "epoch": 1.2684745918654867,
      "grad_norm": 0.14539852738380432,
      "learning_rate": 5.772073481154353e-05,
      "loss": 0.5276,
      "mean_token_accuracy": 0.8366796672344208,
      "num_tokens": 440062152.0,
      "step": 13787
    },
    {
      "entropy": 0.5128947738558054,
      "epoch": 1.2685665982369279,
      "grad_norm": 0.15809257328510284,
      "learning_rate": 5.771766798540191e-05,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8475893102586269,
      "num_tokens": 440094133.0,
      "step": 13788
    },
    {
      "entropy": 0.7198819555342197,
      "epoch": 1.2686586046083692,
      "grad_norm": 0.1667206585407257,
      "learning_rate": 5.771460115926029e-05,
      "loss": 0.7148,
      "mean_token_accuracy": 0.7824502997100353,
      "num_tokens": 440125674.0,
      "step": 13789
    },
    {
      "entropy": 0.5739311098586768,
      "epoch": 1.2687506109798103,
      "grad_norm": 0.15329621732234955,
      "learning_rate": 5.7711534333118655e-05,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8260896615684032,
      "num_tokens": 440158236.0,
      "step": 13790
    },
    {
      "entropy": 0.5350831290706992,
      "epoch": 1.2688426173512517,
      "grad_norm": 0.1457105427980423,
      "learning_rate": 5.770846750697704e-05,
      "loss": 0.524,
      "mean_token_accuracy": 0.8380444534122944,
      "num_tokens": 440190145.0,
      "step": 13791
    },
    {
      "entropy": 0.6712054796516895,
      "epoch": 1.2689346237226928,
      "grad_norm": 0.15702159702777863,
      "learning_rate": 5.770540068083541e-05,
      "loss": 0.6661,
      "mean_token_accuracy": 0.799128495156765,
      "num_tokens": 440222333.0,
      "step": 13792
    },
    {
      "entropy": 0.5866163615137339,
      "epoch": 1.269026630094134,
      "grad_norm": 0.15363295376300812,
      "learning_rate": 5.770233385469378e-05,
      "loss": 0.583,
      "mean_token_accuracy": 0.8196604922413826,
      "num_tokens": 440254763.0,
      "step": 13793
    },
    {
      "entropy": 0.5611393190920353,
      "epoch": 1.2691186364655753,
      "grad_norm": 0.15867279469966888,
      "learning_rate": 5.7699267028552154e-05,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8289593532681465,
      "num_tokens": 440287162.0,
      "step": 13794
    },
    {
      "entropy": 0.550787543412298,
      "epoch": 1.2692106428370165,
      "grad_norm": 0.15062953531742096,
      "learning_rate": 5.7696200202410535e-05,
      "loss": 0.5381,
      "mean_token_accuracy": 0.833231259137392,
      "num_tokens": 440319130.0,
      "step": 13795
    },
    {
      "entropy": 0.6171670462936163,
      "epoch": 1.2693026492084578,
      "grad_norm": 0.15959621965885162,
      "learning_rate": 5.76931333762689e-05,
      "loss": 0.6047,
      "mean_token_accuracy": 0.8137450814247131,
      "num_tokens": 440350427.0,
      "step": 13796
    },
    {
      "entropy": 0.6375608462840319,
      "epoch": 1.269394655579899,
      "grad_norm": 0.15607643127441406,
      "learning_rate": 5.769006655012728e-05,
      "loss": 0.6298,
      "mean_token_accuracy": 0.8073320500552654,
      "num_tokens": 440382711.0,
      "step": 13797
    },
    {
      "entropy": 0.6464697252959013,
      "epoch": 1.26948666195134,
      "grad_norm": 0.16145503520965576,
      "learning_rate": 5.7686999723985646e-05,
      "loss": 0.6488,
      "mean_token_accuracy": 0.8052172251045704,
      "num_tokens": 440414759.0,
      "step": 13798
    },
    {
      "entropy": 0.7107817232608795,
      "epoch": 1.2695786683227814,
      "grad_norm": 0.16713950037956238,
      "learning_rate": 5.768393289784403e-05,
      "loss": 0.7128,
      "mean_token_accuracy": 0.7838229387998581,
      "num_tokens": 440446505.0,
      "step": 13799
    },
    {
      "entropy": 0.5620288867503405,
      "epoch": 1.2696706746942226,
      "grad_norm": 0.1476304978132248,
      "learning_rate": 5.76808660717024e-05,
      "loss": 0.5473,
      "mean_token_accuracy": 0.8285129889845848,
      "num_tokens": 440478783.0,
      "step": 13800
    },
    {
      "entropy": 0.650095958262682,
      "epoch": 1.269762681065664,
      "grad_norm": 0.16378401219844818,
      "learning_rate": 5.767779924556077e-05,
      "loss": 0.6366,
      "mean_token_accuracy": 0.8033820278942585,
      "num_tokens": 440510398.0,
      "step": 13801
    },
    {
      "entropy": 0.6074390653520823,
      "epoch": 1.269854687437105,
      "grad_norm": 0.1628562957048416,
      "learning_rate": 5.7674732419419144e-05,
      "loss": 0.6069,
      "mean_token_accuracy": 0.8150371760129929,
      "num_tokens": 440542505.0,
      "step": 13802
    },
    {
      "entropy": 0.5744472052901983,
      "epoch": 1.2699466938085462,
      "grad_norm": 0.14746052026748657,
      "learning_rate": 5.7671665593277526e-05,
      "loss": 0.5593,
      "mean_token_accuracy": 0.8255939744412899,
      "num_tokens": 440574672.0,
      "step": 13803
    },
    {
      "entropy": 0.6232615932822227,
      "epoch": 1.2700387001799875,
      "grad_norm": 0.15896300971508026,
      "learning_rate": 5.7668598767135894e-05,
      "loss": 0.617,
      "mean_token_accuracy": 0.8133557327091694,
      "num_tokens": 440606929.0,
      "step": 13804
    },
    {
      "entropy": 0.6685177534818649,
      "epoch": 1.2701307065514287,
      "grad_norm": 0.16299718618392944,
      "learning_rate": 5.766553194099427e-05,
      "loss": 0.6569,
      "mean_token_accuracy": 0.7992502339184284,
      "num_tokens": 440638659.0,
      "step": 13805
    },
    {
      "entropy": 0.634759115986526,
      "epoch": 1.27022271292287,
      "grad_norm": 0.15785636007785797,
      "learning_rate": 5.7662465114852636e-05,
      "loss": 0.6071,
      "mean_token_accuracy": 0.8161735199391842,
      "num_tokens": 440670593.0,
      "step": 13806
    },
    {
      "entropy": 0.5525024672970176,
      "epoch": 1.2703147192943112,
      "grad_norm": 0.15504798293113708,
      "learning_rate": 5.765939828871102e-05,
      "loss": 0.5308,
      "mean_token_accuracy": 0.835678543895483,
      "num_tokens": 440701870.0,
      "step": 13807
    },
    {
      "entropy": 0.6227516289800406,
      "epoch": 1.2704067256657523,
      "grad_norm": 0.1634073406457901,
      "learning_rate": 5.765633146256939e-05,
      "loss": 0.606,
      "mean_token_accuracy": 0.8096674010157585,
      "num_tokens": 440732597.0,
      "step": 13808
    },
    {
      "entropy": 0.5335129173472524,
      "epoch": 1.2704987320371937,
      "grad_norm": 0.14506670832633972,
      "learning_rate": 5.765326463642776e-05,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8409722782671452,
      "num_tokens": 440764385.0,
      "step": 13809
    },
    {
      "entropy": 0.6415604399517179,
      "epoch": 1.2705907384086348,
      "grad_norm": 0.16487589478492737,
      "learning_rate": 5.7650197810286135e-05,
      "loss": 0.6348,
      "mean_token_accuracy": 0.8093107342720032,
      "num_tokens": 440796999.0,
      "step": 13810
    },
    {
      "entropy": 0.6221687346696854,
      "epoch": 1.2706827447800761,
      "grad_norm": 0.15995191037654877,
      "learning_rate": 5.7647130984144516e-05,
      "loss": 0.6131,
      "mean_token_accuracy": 0.8156876191496849,
      "num_tokens": 440828026.0,
      "step": 13811
    },
    {
      "entropy": 0.5349616473540664,
      "epoch": 1.2707747511515173,
      "grad_norm": 0.1497558206319809,
      "learning_rate": 5.7644064158002884e-05,
      "loss": 0.5218,
      "mean_token_accuracy": 0.8411870077252388,
      "num_tokens": 440859383.0,
      "step": 13812
    },
    {
      "entropy": 0.6019812077283859,
      "epoch": 1.2708667575229584,
      "grad_norm": 0.15618813037872314,
      "learning_rate": 5.764099733186126e-05,
      "loss": 0.5923,
      "mean_token_accuracy": 0.8220997974276543,
      "num_tokens": 440891467.0,
      "step": 13813
    },
    {
      "entropy": 0.6835436644032598,
      "epoch": 1.2709587638943998,
      "grad_norm": 0.1668415665626526,
      "learning_rate": 5.7637930505719627e-05,
      "loss": 0.6684,
      "mean_token_accuracy": 0.7953656986355782,
      "num_tokens": 440923900.0,
      "step": 13814
    },
    {
      "entropy": 0.5516601167619228,
      "epoch": 1.271050770265841,
      "grad_norm": 0.14585795998573303,
      "learning_rate": 5.763486367957801e-05,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8363810889422894,
      "num_tokens": 440956574.0,
      "step": 13815
    },
    {
      "entropy": 0.5600680224597454,
      "epoch": 1.2711427766372823,
      "grad_norm": 0.1488608866930008,
      "learning_rate": 5.763179685343638e-05,
      "loss": 0.5586,
      "mean_token_accuracy": 0.8300561681389809,
      "num_tokens": 440989342.0,
      "step": 13816
    },
    {
      "entropy": 0.650835670530796,
      "epoch": 1.2712347830087234,
      "grad_norm": 0.16844171285629272,
      "learning_rate": 5.762873002729475e-05,
      "loss": 0.6319,
      "mean_token_accuracy": 0.8081373013556004,
      "num_tokens": 441020410.0,
      "step": 13817
    },
    {
      "entropy": 0.6224777325987816,
      "epoch": 1.2713267893801645,
      "grad_norm": 0.15846839547157288,
      "learning_rate": 5.7625663201153125e-05,
      "loss": 0.6169,
      "mean_token_accuracy": 0.8111876398324966,
      "num_tokens": 441053089.0,
      "step": 13818
    },
    {
      "entropy": 0.5860644560307264,
      "epoch": 1.2714187957516059,
      "grad_norm": 0.15950800478458405,
      "learning_rate": 5.7622596375011507e-05,
      "loss": 0.5616,
      "mean_token_accuracy": 0.8268010690808296,
      "num_tokens": 441084119.0,
      "step": 13819
    },
    {
      "entropy": 0.5717830564826727,
      "epoch": 1.271510802123047,
      "grad_norm": 0.15866833925247192,
      "learning_rate": 5.7619529548869874e-05,
      "loss": 0.5589,
      "mean_token_accuracy": 0.8255448527634144,
      "num_tokens": 441116638.0,
      "step": 13820
    },
    {
      "entropy": 0.6927671115845442,
      "epoch": 1.2716028084944884,
      "grad_norm": 0.16553106904029846,
      "learning_rate": 5.761646272272825e-05,
      "loss": 0.6892,
      "mean_token_accuracy": 0.7919283732771873,
      "num_tokens": 441148654.0,
      "step": 13821
    },
    {
      "entropy": 0.6401946377009153,
      "epoch": 1.2716948148659295,
      "grad_norm": 0.16350410878658295,
      "learning_rate": 5.761339589658663e-05,
      "loss": 0.6218,
      "mean_token_accuracy": 0.8058386035263538,
      "num_tokens": 441180723.0,
      "step": 13822
    },
    {
      "entropy": 0.6923335511237383,
      "epoch": 1.2717868212373706,
      "grad_norm": 0.15983091294765472,
      "learning_rate": 5.7610329070445e-05,
      "loss": 0.6888,
      "mean_token_accuracy": 0.7915571220219135,
      "num_tokens": 441212586.0,
      "step": 13823
    },
    {
      "entropy": 0.594536162680015,
      "epoch": 1.271878827608812,
      "grad_norm": 0.16395476460456848,
      "learning_rate": 5.760726224430337e-05,
      "loss": 0.5974,
      "mean_token_accuracy": 0.8201429657638073,
      "num_tokens": 441244210.0,
      "step": 13824
    },
    {
      "entropy": 0.532375868409872,
      "epoch": 1.2719708339802531,
      "grad_norm": 0.1584663987159729,
      "learning_rate": 5.760419541816174e-05,
      "loss": 0.5353,
      "mean_token_accuracy": 0.8369136229157448,
      "num_tokens": 441276670.0,
      "step": 13825
    },
    {
      "entropy": 0.6258563362061977,
      "epoch": 1.2720628403516945,
      "grad_norm": 0.15396229922771454,
      "learning_rate": 5.760112859202013e-05,
      "loss": 0.6147,
      "mean_token_accuracy": 0.8093373440206051,
      "num_tokens": 441308622.0,
      "step": 13826
    },
    {
      "entropy": 0.6061709001660347,
      "epoch": 1.2721548467231356,
      "grad_norm": 0.1604192703962326,
      "learning_rate": 5.75980617658785e-05,
      "loss": 0.6013,
      "mean_token_accuracy": 0.8163116089999676,
      "num_tokens": 441341158.0,
      "step": 13827
    },
    {
      "entropy": 0.6623376775532961,
      "epoch": 1.2722468530945767,
      "grad_norm": 0.1596689075231552,
      "learning_rate": 5.759499493973687e-05,
      "loss": 0.652,
      "mean_token_accuracy": 0.8022796660661697,
      "num_tokens": 441372450.0,
      "step": 13828
    },
    {
      "entropy": 0.6190747730433941,
      "epoch": 1.272338859466018,
      "grad_norm": 0.15410862863063812,
      "learning_rate": 5.759192811359524e-05,
      "loss": 0.604,
      "mean_token_accuracy": 0.8139769770205021,
      "num_tokens": 441404544.0,
      "step": 13829
    },
    {
      "entropy": 0.6470602639019489,
      "epoch": 1.2724308658374592,
      "grad_norm": 0.15431171655654907,
      "learning_rate": 5.758886128745362e-05,
      "loss": 0.6374,
      "mean_token_accuracy": 0.8021760955452919,
      "num_tokens": 441435727.0,
      "step": 13830
    },
    {
      "entropy": 0.7247574552893639,
      "epoch": 1.2725228722089006,
      "grad_norm": 0.16241970658302307,
      "learning_rate": 5.7585794461311996e-05,
      "loss": 0.7313,
      "mean_token_accuracy": 0.7763128317892551,
      "num_tokens": 441467254.0,
      "step": 13831
    },
    {
      "entropy": 0.6637845207005739,
      "epoch": 1.2726148785803417,
      "grad_norm": 0.15826597809791565,
      "learning_rate": 5.7582727635170363e-05,
      "loss": 0.6417,
      "mean_token_accuracy": 0.8029331788420677,
      "num_tokens": 441499442.0,
      "step": 13832
    },
    {
      "entropy": 0.7051487434655428,
      "epoch": 1.2727068849517829,
      "grad_norm": 0.16533924639225006,
      "learning_rate": 5.757966080902874e-05,
      "loss": 0.7007,
      "mean_token_accuracy": 0.7870677225291729,
      "num_tokens": 441530694.0,
      "step": 13833
    },
    {
      "entropy": 0.6898352075368166,
      "epoch": 1.2727988913232242,
      "grad_norm": 0.16541637480258942,
      "learning_rate": 5.757659398288712e-05,
      "loss": 0.6829,
      "mean_token_accuracy": 0.7883370779454708,
      "num_tokens": 441561544.0,
      "step": 13834
    },
    {
      "entropy": 0.5695726941339672,
      "epoch": 1.2728908976946653,
      "grad_norm": 0.15103237330913544,
      "learning_rate": 5.757352715674549e-05,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8321606516838074,
      "num_tokens": 441593435.0,
      "step": 13835
    },
    {
      "entropy": 0.6152792545035481,
      "epoch": 1.2729829040661067,
      "grad_norm": 0.16093946993350983,
      "learning_rate": 5.757046033060386e-05,
      "loss": 0.5966,
      "mean_token_accuracy": 0.8158233687281609,
      "num_tokens": 441624298.0,
      "step": 13836
    },
    {
      "entropy": 0.6104194968938828,
      "epoch": 1.2730749104375478,
      "grad_norm": 0.15467391908168793,
      "learning_rate": 5.756739350446223e-05,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8173552937805653,
      "num_tokens": 441656191.0,
      "step": 13837
    },
    {
      "entropy": 0.6196712795644999,
      "epoch": 1.273166916808989,
      "grad_norm": 0.15923695266246796,
      "learning_rate": 5.756432667832061e-05,
      "loss": 0.6007,
      "mean_token_accuracy": 0.8188120722770691,
      "num_tokens": 441688543.0,
      "step": 13838
    },
    {
      "entropy": 0.5993935372680426,
      "epoch": 1.2732589231804303,
      "grad_norm": 0.14865665137767792,
      "learning_rate": 5.7561259852178986e-05,
      "loss": 0.5965,
      "mean_token_accuracy": 0.8170304633677006,
      "num_tokens": 441721149.0,
      "step": 13839
    },
    {
      "entropy": 0.5539134088903666,
      "epoch": 1.2733509295518715,
      "grad_norm": 0.15042364597320557,
      "learning_rate": 5.7558193026037354e-05,
      "loss": 0.5409,
      "mean_token_accuracy": 0.832875233143568,
      "num_tokens": 441752936.0,
      "step": 13840
    },
    {
      "entropy": 0.6769829438999295,
      "epoch": 1.2734429359233128,
      "grad_norm": 0.16352351009845734,
      "learning_rate": 5.755512619989573e-05,
      "loss": 0.6649,
      "mean_token_accuracy": 0.7944390848278999,
      "num_tokens": 441784372.0,
      "step": 13841
    },
    {
      "entropy": 0.5919586946256459,
      "epoch": 1.273534942294754,
      "grad_norm": 0.14974480867385864,
      "learning_rate": 5.755205937375411e-05,
      "loss": 0.5845,
      "mean_token_accuracy": 0.8220030665397644,
      "num_tokens": 441816644.0,
      "step": 13842
    },
    {
      "entropy": 0.5995071521028876,
      "epoch": 1.273626948666195,
      "grad_norm": 0.1582026332616806,
      "learning_rate": 5.754899254761248e-05,
      "loss": 0.5775,
      "mean_token_accuracy": 0.8212923519313335,
      "num_tokens": 441847007.0,
      "step": 13843
    },
    {
      "entropy": 0.587158776819706,
      "epoch": 1.2737189550376364,
      "grad_norm": 0.15900833904743195,
      "learning_rate": 5.754592572147085e-05,
      "loss": 0.5863,
      "mean_token_accuracy": 0.8201338388025761,
      "num_tokens": 441878543.0,
      "step": 13844
    },
    {
      "entropy": 0.593153971247375,
      "epoch": 1.2738109614090776,
      "grad_norm": 0.16789330542087555,
      "learning_rate": 5.754285889532922e-05,
      "loss": 0.5985,
      "mean_token_accuracy": 0.8173963613808155,
      "num_tokens": 441910133.0,
      "step": 13845
    },
    {
      "entropy": 0.5071455137804151,
      "epoch": 1.273902967780519,
      "grad_norm": 0.1521548181772232,
      "learning_rate": 5.75397920691876e-05,
      "loss": 0.5046,
      "mean_token_accuracy": 0.8444474749267101,
      "num_tokens": 441942432.0,
      "step": 13846
    },
    {
      "entropy": 0.5231405235826969,
      "epoch": 1.27399497415196,
      "grad_norm": 0.14768217504024506,
      "learning_rate": 5.7536725243045976e-05,
      "loss": 0.5231,
      "mean_token_accuracy": 0.8339936845004559,
      "num_tokens": 441974624.0,
      "step": 13847
    },
    {
      "entropy": 0.6174324341118336,
      "epoch": 1.2740869805234012,
      "grad_norm": 0.15763483941555023,
      "learning_rate": 5.7533658416904344e-05,
      "loss": 0.6032,
      "mean_token_accuracy": 0.8130255676805973,
      "num_tokens": 442007003.0,
      "step": 13848
    },
    {
      "entropy": 0.662416135892272,
      "epoch": 1.2741789868948425,
      "grad_norm": 0.1658257246017456,
      "learning_rate": 5.753059159076272e-05,
      "loss": 0.6518,
      "mean_token_accuracy": 0.8001807779073715,
      "num_tokens": 442038555.0,
      "step": 13849
    },
    {
      "entropy": 0.7047492824494839,
      "epoch": 1.2742709932662837,
      "grad_norm": 0.16315844655036926,
      "learning_rate": 5.75275247646211e-05,
      "loss": 0.6942,
      "mean_token_accuracy": 0.787362840026617,
      "num_tokens": 442069928.0,
      "step": 13850
    },
    {
      "entropy": 0.606521500274539,
      "epoch": 1.274362999637725,
      "grad_norm": 0.151242196559906,
      "learning_rate": 5.752445793847947e-05,
      "loss": 0.5902,
      "mean_token_accuracy": 0.8212295360863209,
      "num_tokens": 442101840.0,
      "step": 13851
    },
    {
      "entropy": 0.6789348497986794,
      "epoch": 1.2744550060091662,
      "grad_norm": 0.17026516795158386,
      "learning_rate": 5.752139111233784e-05,
      "loss": 0.6799,
      "mean_token_accuracy": 0.795574814081192,
      "num_tokens": 442132987.0,
      "step": 13852
    },
    {
      "entropy": 0.6980616003274918,
      "epoch": 1.2745470123806073,
      "grad_norm": 0.16185380518436432,
      "learning_rate": 5.7518324286196224e-05,
      "loss": 0.691,
      "mean_token_accuracy": 0.7918191477656364,
      "num_tokens": 442165120.0,
      "step": 13853
    },
    {
      "entropy": 0.5436262395232916,
      "epoch": 1.2746390187520487,
      "grad_norm": 0.15807361900806427,
      "learning_rate": 5.751525746005459e-05,
      "loss": 0.5203,
      "mean_token_accuracy": 0.8398622013628483,
      "num_tokens": 442196792.0,
      "step": 13854
    },
    {
      "entropy": 0.6259259423241019,
      "epoch": 1.2747310251234898,
      "grad_norm": 0.1555822640657425,
      "learning_rate": 5.751219063391297e-05,
      "loss": 0.6205,
      "mean_token_accuracy": 0.8086319752037525,
      "num_tokens": 442229118.0,
      "step": 13855
    },
    {
      "entropy": 0.691930714994669,
      "epoch": 1.2748230314949311,
      "grad_norm": 0.161603644490242,
      "learning_rate": 5.7509123807771335e-05,
      "loss": 0.6866,
      "mean_token_accuracy": 0.7947201356291771,
      "num_tokens": 442261567.0,
      "step": 13856
    },
    {
      "entropy": 0.6193616297096014,
      "epoch": 1.2749150378663723,
      "grad_norm": 0.15491046011447906,
      "learning_rate": 5.7506056981629716e-05,
      "loss": 0.6135,
      "mean_token_accuracy": 0.8129960596561432,
      "num_tokens": 442293814.0,
      "step": 13857
    },
    {
      "entropy": 0.6622995901852846,
      "epoch": 1.2750070442378134,
      "grad_norm": 0.1638595312833786,
      "learning_rate": 5.750299015548809e-05,
      "loss": 0.628,
      "mean_token_accuracy": 0.8040282428264618,
      "num_tokens": 442325792.0,
      "step": 13858
    },
    {
      "entropy": 0.6606711912900209,
      "epoch": 1.2750990506092548,
      "grad_norm": 0.16940262913703918,
      "learning_rate": 5.749992332934646e-05,
      "loss": 0.652,
      "mean_token_accuracy": 0.799287885427475,
      "num_tokens": 442357331.0,
      "step": 13859
    },
    {
      "entropy": 0.6001984924077988,
      "epoch": 1.275191056980696,
      "grad_norm": 0.1576623171567917,
      "learning_rate": 5.749685650320483e-05,
      "loss": 0.5884,
      "mean_token_accuracy": 0.822505671530962,
      "num_tokens": 442389220.0,
      "step": 13860
    },
    {
      "entropy": 0.6381700364872813,
      "epoch": 1.2752830633521373,
      "grad_norm": 0.1608884334564209,
      "learning_rate": 5.7493789677063215e-05,
      "loss": 0.6298,
      "mean_token_accuracy": 0.805727381259203,
      "num_tokens": 442421422.0,
      "step": 13861
    },
    {
      "entropy": 0.6713154036551714,
      "epoch": 1.2753750697235784,
      "grad_norm": 0.1662663072347641,
      "learning_rate": 5.749072285092158e-05,
      "loss": 0.6518,
      "mean_token_accuracy": 0.7999961413443089,
      "num_tokens": 442452683.0,
      "step": 13862
    },
    {
      "entropy": 0.5742783714085817,
      "epoch": 1.2754670760950195,
      "grad_norm": 0.15358808636665344,
      "learning_rate": 5.748765602477996e-05,
      "loss": 0.5729,
      "mean_token_accuracy": 0.8235509432852268,
      "num_tokens": 442484667.0,
      "step": 13863
    },
    {
      "entropy": 0.640785239636898,
      "epoch": 1.2755590824664609,
      "grad_norm": 0.16236808896064758,
      "learning_rate": 5.7484589198638325e-05,
      "loss": 0.6332,
      "mean_token_accuracy": 0.8109557554125786,
      "num_tokens": 442516573.0,
      "step": 13864
    },
    {
      "entropy": 0.5951370066031814,
      "epoch": 1.275651088837902,
      "grad_norm": 0.1589687168598175,
      "learning_rate": 5.748152237249671e-05,
      "loss": 0.5994,
      "mean_token_accuracy": 0.8179879523813725,
      "num_tokens": 442548913.0,
      "step": 13865
    },
    {
      "entropy": 0.5362338447012007,
      "epoch": 1.2757430952093434,
      "grad_norm": 0.14742857217788696,
      "learning_rate": 5.747845554635508e-05,
      "loss": 0.513,
      "mean_token_accuracy": 0.841978169977665,
      "num_tokens": 442581297.0,
      "step": 13866
    },
    {
      "entropy": 0.6002123542129993,
      "epoch": 1.2758351015807845,
      "grad_norm": 0.15513888001441956,
      "learning_rate": 5.747538872021345e-05,
      "loss": 0.5904,
      "mean_token_accuracy": 0.8172590769827366,
      "num_tokens": 442613362.0,
      "step": 13867
    },
    {
      "entropy": 0.604791562538594,
      "epoch": 1.2759271079522256,
      "grad_norm": 0.15436150133609772,
      "learning_rate": 5.7472321894071824e-05,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8183875232934952,
      "num_tokens": 442645382.0,
      "step": 13868
    },
    {
      "entropy": 0.6546464245766401,
      "epoch": 1.276019114323667,
      "grad_norm": 0.16635754704475403,
      "learning_rate": 5.7469255067930205e-05,
      "loss": 0.6547,
      "mean_token_accuracy": 0.8003407903015614,
      "num_tokens": 442677768.0,
      "step": 13869
    },
    {
      "entropy": 0.5626140739768744,
      "epoch": 1.2761111206951081,
      "grad_norm": 0.14636586606502533,
      "learning_rate": 5.746618824178858e-05,
      "loss": 0.534,
      "mean_token_accuracy": 0.8332091085612774,
      "num_tokens": 442709181.0,
      "step": 13870
    },
    {
      "entropy": 0.6373679451644421,
      "epoch": 1.2762031270665495,
      "grad_norm": 0.16426651179790497,
      "learning_rate": 5.746312141564695e-05,
      "loss": 0.624,
      "mean_token_accuracy": 0.8091262653470039,
      "num_tokens": 442741449.0,
      "step": 13871
    },
    {
      "entropy": 0.5476949214935303,
      "epoch": 1.2762951334379906,
      "grad_norm": 0.1590166687965393,
      "learning_rate": 5.746005458950532e-05,
      "loss": 0.5371,
      "mean_token_accuracy": 0.8294711671769619,
      "num_tokens": 442772604.0,
      "step": 13872
    },
    {
      "entropy": 0.6741463355720043,
      "epoch": 1.2763871398094317,
      "grad_norm": 0.15701577067375183,
      "learning_rate": 5.7456987763363704e-05,
      "loss": 0.6677,
      "mean_token_accuracy": 0.7978526018559933,
      "num_tokens": 442805243.0,
      "step": 13873
    },
    {
      "entropy": 0.5980631848797202,
      "epoch": 1.276479146180873,
      "grad_norm": 0.15351064503192902,
      "learning_rate": 5.745392093722207e-05,
      "loss": 0.5944,
      "mean_token_accuracy": 0.8179227560758591,
      "num_tokens": 442837818.0,
      "step": 13874
    },
    {
      "entropy": 0.5846872692927718,
      "epoch": 1.2765711525523142,
      "grad_norm": 0.15152305364608765,
      "learning_rate": 5.7450854111080446e-05,
      "loss": 0.5593,
      "mean_token_accuracy": 0.8291415721178055,
      "num_tokens": 442870294.0,
      "step": 13875
    },
    {
      "entropy": 0.6315066013485193,
      "epoch": 1.2766631589237556,
      "grad_norm": 0.16094501316547394,
      "learning_rate": 5.7447787284938814e-05,
      "loss": 0.622,
      "mean_token_accuracy": 0.8124467618763447,
      "num_tokens": 442901320.0,
      "step": 13876
    },
    {
      "entropy": 0.6440190710127354,
      "epoch": 1.2767551652951967,
      "grad_norm": 0.1623286008834839,
      "learning_rate": 5.7444720458797195e-05,
      "loss": 0.6336,
      "mean_token_accuracy": 0.810612827539444,
      "num_tokens": 442932540.0,
      "step": 13877
    },
    {
      "entropy": 0.5546012930572033,
      "epoch": 1.2768471716666379,
      "grad_norm": 0.1508064866065979,
      "learning_rate": 5.744165363265557e-05,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8314789421856403,
      "num_tokens": 442964669.0,
      "step": 13878
    },
    {
      "entropy": 0.6075981156900525,
      "epoch": 1.2769391780380792,
      "grad_norm": 0.14901448786258698,
      "learning_rate": 5.743858680651394e-05,
      "loss": 0.5917,
      "mean_token_accuracy": 0.81851901486516,
      "num_tokens": 442997142.0,
      "step": 13879
    },
    {
      "entropy": 0.5741992890834808,
      "epoch": 1.2770311844095203,
      "grad_norm": 0.15267673134803772,
      "learning_rate": 5.743551998037232e-05,
      "loss": 0.5546,
      "mean_token_accuracy": 0.8267999924719334,
      "num_tokens": 443029192.0,
      "step": 13880
    },
    {
      "entropy": 0.6102674882858992,
      "epoch": 1.2771231907809617,
      "grad_norm": 0.1511266678571701,
      "learning_rate": 5.7432453154230694e-05,
      "loss": 0.6009,
      "mean_token_accuracy": 0.8175448328256607,
      "num_tokens": 443061583.0,
      "step": 13881
    },
    {
      "entropy": 0.6474477238953114,
      "epoch": 1.2772151971524028,
      "grad_norm": 0.16258618235588074,
      "learning_rate": 5.742938632808906e-05,
      "loss": 0.6404,
      "mean_token_accuracy": 0.8039922937750816,
      "num_tokens": 443093850.0,
      "step": 13882
    },
    {
      "entropy": 0.6355354581028223,
      "epoch": 1.277307203523844,
      "grad_norm": 0.1558583527803421,
      "learning_rate": 5.7426319501947437e-05,
      "loss": 0.6221,
      "mean_token_accuracy": 0.8060208633542061,
      "num_tokens": 443125256.0,
      "step": 13883
    },
    {
      "entropy": 0.6766267828643322,
      "epoch": 1.2773992098952853,
      "grad_norm": 0.1623411774635315,
      "learning_rate": 5.742325267580582e-05,
      "loss": 0.6651,
      "mean_token_accuracy": 0.7953330129384995,
      "num_tokens": 443157218.0,
      "step": 13884
    },
    {
      "entropy": 0.6958759129047394,
      "epoch": 1.2774912162667265,
      "grad_norm": 0.16498631238937378,
      "learning_rate": 5.7420185849664186e-05,
      "loss": 0.7041,
      "mean_token_accuracy": 0.7883907556533813,
      "num_tokens": 443188794.0,
      "step": 13885
    },
    {
      "entropy": 0.6225306894630194,
      "epoch": 1.2775832226381678,
      "grad_norm": 0.15882191061973572,
      "learning_rate": 5.741711902352256e-05,
      "loss": 0.6242,
      "mean_token_accuracy": 0.8108721785247326,
      "num_tokens": 443220729.0,
      "step": 13886
    },
    {
      "entropy": 0.6861286610364914,
      "epoch": 1.277675229009609,
      "grad_norm": 0.15713398158550262,
      "learning_rate": 5.741405219738093e-05,
      "loss": 0.6682,
      "mean_token_accuracy": 0.8001101687550545,
      "num_tokens": 443252989.0,
      "step": 13887
    },
    {
      "entropy": 0.6635510958731174,
      "epoch": 1.27776723538105,
      "grad_norm": 0.16457854211330414,
      "learning_rate": 5.741098537123931e-05,
      "loss": 0.6554,
      "mean_token_accuracy": 0.7990943714976311,
      "num_tokens": 443285065.0,
      "step": 13888
    },
    {
      "entropy": 0.5411069020628929,
      "epoch": 1.2778592417524914,
      "grad_norm": 0.15547390282154083,
      "learning_rate": 5.7407918545097684e-05,
      "loss": 0.5316,
      "mean_token_accuracy": 0.8361757360398769,
      "num_tokens": 443317036.0,
      "step": 13889
    },
    {
      "entropy": 0.6196090411394835,
      "epoch": 1.2779512481239326,
      "grad_norm": 0.16910439729690552,
      "learning_rate": 5.740485171895605e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8167320638895035,
      "num_tokens": 443349292.0,
      "step": 13890
    },
    {
      "entropy": 0.6818442586809397,
      "epoch": 1.278043254495374,
      "grad_norm": 0.16202512383460999,
      "learning_rate": 5.740178489281443e-05,
      "loss": 0.6694,
      "mean_token_accuracy": 0.7934325039386749,
      "num_tokens": 443380620.0,
      "step": 13891
    },
    {
      "entropy": 0.5619233138859272,
      "epoch": 1.278135260866815,
      "grad_norm": 0.14270591735839844,
      "learning_rate": 5.739871806667281e-05,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8353612162172794,
      "num_tokens": 443412739.0,
      "step": 13892
    },
    {
      "entropy": 0.5053449254482985,
      "epoch": 1.2782272672382562,
      "grad_norm": 0.14793437719345093,
      "learning_rate": 5.7395651240531176e-05,
      "loss": 0.4899,
      "mean_token_accuracy": 0.8469527848064899,
      "num_tokens": 443444685.0,
      "step": 13893
    },
    {
      "entropy": 0.601453335955739,
      "epoch": 1.2783192736096975,
      "grad_norm": 0.15477211773395538,
      "learning_rate": 5.739258441438955e-05,
      "loss": 0.6028,
      "mean_token_accuracy": 0.8138456158339977,
      "num_tokens": 443476647.0,
      "step": 13894
    },
    {
      "entropy": 0.6592674730345607,
      "epoch": 1.2784112799811387,
      "grad_norm": 0.15954343974590302,
      "learning_rate": 5.738951758824792e-05,
      "loss": 0.6549,
      "mean_token_accuracy": 0.8022481240332127,
      "num_tokens": 443508313.0,
      "step": 13895
    },
    {
      "entropy": 0.6614724099636078,
      "epoch": 1.27850328635258,
      "grad_norm": 0.1588982492685318,
      "learning_rate": 5.73864507621063e-05,
      "loss": 0.6545,
      "mean_token_accuracy": 0.8011260367929935,
      "num_tokens": 443539467.0,
      "step": 13896
    },
    {
      "entropy": 0.5582099333405495,
      "epoch": 1.2785952927240212,
      "grad_norm": 0.15015427768230438,
      "learning_rate": 5.7383383935964675e-05,
      "loss": 0.5443,
      "mean_token_accuracy": 0.8341596573591232,
      "num_tokens": 443571812.0,
      "step": 13897
    },
    {
      "entropy": 0.6074298359453678,
      "epoch": 1.2786872990954623,
      "grad_norm": 0.15247176587581635,
      "learning_rate": 5.738031710982304e-05,
      "loss": 0.5927,
      "mean_token_accuracy": 0.8189163766801357,
      "num_tokens": 443604281.0,
      "step": 13898
    },
    {
      "entropy": 0.6786485072225332,
      "epoch": 1.2787793054669037,
      "grad_norm": 0.15448060631752014,
      "learning_rate": 5.737725028368142e-05,
      "loss": 0.6625,
      "mean_token_accuracy": 0.7966324761509895,
      "num_tokens": 443636442.0,
      "step": 13899
    },
    {
      "entropy": 0.5783123942092061,
      "epoch": 1.2788713118383448,
      "grad_norm": 0.15984287858009338,
      "learning_rate": 5.73741834575398e-05,
      "loss": 0.559,
      "mean_token_accuracy": 0.827831044793129,
      "num_tokens": 443668414.0,
      "step": 13900
    },
    {
      "entropy": 0.7830788102000952,
      "epoch": 1.2789633182097861,
      "grad_norm": 0.17612740397453308,
      "learning_rate": 5.737111663139817e-05,
      "loss": 0.7916,
      "mean_token_accuracy": 0.7601870708167553,
      "num_tokens": 443699686.0,
      "step": 13901
    },
    {
      "entropy": 0.6530077308416367,
      "epoch": 1.2790553245812273,
      "grad_norm": 0.15881122648715973,
      "learning_rate": 5.736804980525654e-05,
      "loss": 0.6373,
      "mean_token_accuracy": 0.8060609176754951,
      "num_tokens": 443732036.0,
      "step": 13902
    },
    {
      "entropy": 0.6050755344331264,
      "epoch": 1.2791473309526684,
      "grad_norm": 0.1556432992219925,
      "learning_rate": 5.736498297911491e-05,
      "loss": 0.5867,
      "mean_token_accuracy": 0.816082451492548,
      "num_tokens": 443763465.0,
      "step": 13903
    },
    {
      "entropy": 0.5841084159910679,
      "epoch": 1.2792393373241098,
      "grad_norm": 0.1531592607498169,
      "learning_rate": 5.73619161529733e-05,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8198496252298355,
      "num_tokens": 443795719.0,
      "step": 13904
    },
    {
      "entropy": 0.5860727392137051,
      "epoch": 1.279331343695551,
      "grad_norm": 0.15251538157463074,
      "learning_rate": 5.7358849326831665e-05,
      "loss": 0.5779,
      "mean_token_accuracy": 0.8216086067259312,
      "num_tokens": 443827672.0,
      "step": 13905
    },
    {
      "entropy": 0.658170341514051,
      "epoch": 1.2794233500669923,
      "grad_norm": 0.1612427830696106,
      "learning_rate": 5.735578250069003e-05,
      "loss": 0.6446,
      "mean_token_accuracy": 0.8021383732557297,
      "num_tokens": 443859485.0,
      "step": 13906
    },
    {
      "entropy": 0.6303599812090397,
      "epoch": 1.2795153564384334,
      "grad_norm": 0.16068285703659058,
      "learning_rate": 5.735271567454841e-05,
      "loss": 0.6178,
      "mean_token_accuracy": 0.8095703981816769,
      "num_tokens": 443890922.0,
      "step": 13907
    },
    {
      "entropy": 0.6264244494959712,
      "epoch": 1.2796073628098745,
      "grad_norm": 0.15577398240566254,
      "learning_rate": 5.734964884840679e-05,
      "loss": 0.6235,
      "mean_token_accuracy": 0.8102701120078564,
      "num_tokens": 443923087.0,
      "step": 13908
    },
    {
      "entropy": 0.531996110221371,
      "epoch": 1.2796993691813159,
      "grad_norm": 0.1472480297088623,
      "learning_rate": 5.7346582022265164e-05,
      "loss": 0.5189,
      "mean_token_accuracy": 0.838898628950119,
      "num_tokens": 443955058.0,
      "step": 13909
    },
    {
      "entropy": 0.7032715668901801,
      "epoch": 1.279791375552757,
      "grad_norm": 0.16553564369678497,
      "learning_rate": 5.734351519612353e-05,
      "loss": 0.7068,
      "mean_token_accuracy": 0.7826730534434319,
      "num_tokens": 443987285.0,
      "step": 13910
    },
    {
      "entropy": 0.6079462091438472,
      "epoch": 1.2798833819241984,
      "grad_norm": 0.15632589161396027,
      "learning_rate": 5.734044836998191e-05,
      "loss": 0.5871,
      "mean_token_accuracy": 0.8173440843820572,
      "num_tokens": 444018927.0,
      "step": 13911
    },
    {
      "entropy": 0.5597226750105619,
      "epoch": 1.2799753882956395,
      "grad_norm": 0.1509910523891449,
      "learning_rate": 5.733738154384029e-05,
      "loss": 0.5393,
      "mean_token_accuracy": 0.8311402648687363,
      "num_tokens": 444051206.0,
      "step": 13912
    },
    {
      "entropy": 0.5342313107103109,
      "epoch": 1.2800673946670806,
      "grad_norm": 0.1507698893547058,
      "learning_rate": 5.7334314717698656e-05,
      "loss": 0.5228,
      "mean_token_accuracy": 0.8412925116717815,
      "num_tokens": 444083463.0,
      "step": 13913
    },
    {
      "entropy": 0.656307877972722,
      "epoch": 1.280159401038522,
      "grad_norm": 0.168048694729805,
      "learning_rate": 5.733124789155703e-05,
      "loss": 0.642,
      "mean_token_accuracy": 0.8022663742303848,
      "num_tokens": 444115661.0,
      "step": 13914
    },
    {
      "entropy": 0.5275769259314984,
      "epoch": 1.2802514074099631,
      "grad_norm": 0.1569826304912567,
      "learning_rate": 5.732818106541541e-05,
      "loss": 0.5206,
      "mean_token_accuracy": 0.8364079110324383,
      "num_tokens": 444147624.0,
      "step": 13915
    },
    {
      "entropy": 0.6051745638251305,
      "epoch": 1.2803434137814045,
      "grad_norm": 0.15243114531040192,
      "learning_rate": 5.732511423927378e-05,
      "loss": 0.6037,
      "mean_token_accuracy": 0.8143372572958469,
      "num_tokens": 444178960.0,
      "step": 13916
    },
    {
      "entropy": 0.5977047979831696,
      "epoch": 1.2804354201528456,
      "grad_norm": 0.1473105251789093,
      "learning_rate": 5.7322047413132154e-05,
      "loss": 0.5775,
      "mean_token_accuracy": 0.8205301724374294,
      "num_tokens": 444211244.0,
      "step": 13917
    },
    {
      "entropy": 0.5579828973859549,
      "epoch": 1.2805274265242867,
      "grad_norm": 0.15171952545642853,
      "learning_rate": 5.731898058699052e-05,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8330010622739792,
      "num_tokens": 444243245.0,
      "step": 13918
    },
    {
      "entropy": 0.642421749420464,
      "epoch": 1.280619432895728,
      "grad_norm": 0.1569940447807312,
      "learning_rate": 5.7315913760848904e-05,
      "loss": 0.6128,
      "mean_token_accuracy": 0.8115162290632725,
      "num_tokens": 444274867.0,
      "step": 13919
    },
    {
      "entropy": 0.4964998164214194,
      "epoch": 1.2807114392671692,
      "grad_norm": 0.14283116161823273,
      "learning_rate": 5.731284693470728e-05,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8496962077915668,
      "num_tokens": 444307199.0,
      "step": 13920
    },
    {
      "entropy": 0.6392024513334036,
      "epoch": 1.2808034456386106,
      "grad_norm": 0.15604428946971893,
      "learning_rate": 5.7309780108565646e-05,
      "loss": 0.623,
      "mean_token_accuracy": 0.8076278641819954,
      "num_tokens": 444338771.0,
      "step": 13921
    },
    {
      "entropy": 0.6734645124524832,
      "epoch": 1.2808954520100517,
      "grad_norm": 0.16512955725193024,
      "learning_rate": 5.730671328242402e-05,
      "loss": 0.6662,
      "mean_token_accuracy": 0.7932133041322231,
      "num_tokens": 444369366.0,
      "step": 13922
    },
    {
      "entropy": 0.6289970623329282,
      "epoch": 1.2809874583814929,
      "grad_norm": 0.1530550718307495,
      "learning_rate": 5.73036464562824e-05,
      "loss": 0.6201,
      "mean_token_accuracy": 0.8094162195920944,
      "num_tokens": 444402134.0,
      "step": 13923
    },
    {
      "entropy": 0.709039619192481,
      "epoch": 1.2810794647529342,
      "grad_norm": 0.15522493422031403,
      "learning_rate": 5.730057963014077e-05,
      "loss": 0.6963,
      "mean_token_accuracy": 0.7868913486599922,
      "num_tokens": 444434738.0,
      "step": 13924
    },
    {
      "entropy": 0.5966073321178555,
      "epoch": 1.2811714711243753,
      "grad_norm": 0.15839232504367828,
      "learning_rate": 5.7297512803999145e-05,
      "loss": 0.5986,
      "mean_token_accuracy": 0.8194049969315529,
      "num_tokens": 444466638.0,
      "step": 13925
    },
    {
      "entropy": 0.6396216731518507,
      "epoch": 1.2812634774958167,
      "grad_norm": 0.15620547533035278,
      "learning_rate": 5.729444597785751e-05,
      "loss": 0.6228,
      "mean_token_accuracy": 0.8115263357758522,
      "num_tokens": 444499036.0,
      "step": 13926
    },
    {
      "entropy": 0.6450654771178961,
      "epoch": 1.2813554838672578,
      "grad_norm": 0.1647142618894577,
      "learning_rate": 5.7291379151715894e-05,
      "loss": 0.6383,
      "mean_token_accuracy": 0.8096137717366219,
      "num_tokens": 444530788.0,
      "step": 13927
    },
    {
      "entropy": 0.6863249819725752,
      "epoch": 1.281447490238699,
      "grad_norm": 0.16221748292446136,
      "learning_rate": 5.728831232557427e-05,
      "loss": 0.6859,
      "mean_token_accuracy": 0.7953794896602631,
      "num_tokens": 444562985.0,
      "step": 13928
    },
    {
      "entropy": 0.5681913336738944,
      "epoch": 1.2815394966101403,
      "grad_norm": 0.15061919391155243,
      "learning_rate": 5.7285245499432636e-05,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8303997479379177,
      "num_tokens": 444595599.0,
      "step": 13929
    },
    {
      "entropy": 0.5743945240974426,
      "epoch": 1.2816315029815815,
      "grad_norm": 0.16482746601104736,
      "learning_rate": 5.728217867329101e-05,
      "loss": 0.5595,
      "mean_token_accuracy": 0.827164925634861,
      "num_tokens": 444627206.0,
      "step": 13930
    },
    {
      "entropy": 0.5096466802060604,
      "epoch": 1.2817235093530228,
      "grad_norm": 0.15136906504631042,
      "learning_rate": 5.727911184714939e-05,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8418514765799046,
      "num_tokens": 444659255.0,
      "step": 13931
    },
    {
      "entropy": 0.6395183894783258,
      "epoch": 1.281815515724464,
      "grad_norm": 0.16679984331130981,
      "learning_rate": 5.727604502100776e-05,
      "loss": 0.6361,
      "mean_token_accuracy": 0.8083477541804314,
      "num_tokens": 444691098.0,
      "step": 13932
    },
    {
      "entropy": 0.5993913726415485,
      "epoch": 1.281907522095905,
      "grad_norm": 0.1472110152244568,
      "learning_rate": 5.7272978194866135e-05,
      "loss": 0.5826,
      "mean_token_accuracy": 0.8186062835156918,
      "num_tokens": 444723381.0,
      "step": 13933
    },
    {
      "entropy": 0.6620615441352129,
      "epoch": 1.2819995284673464,
      "grad_norm": 0.16367988288402557,
      "learning_rate": 5.72699113687245e-05,
      "loss": 0.65,
      "mean_token_accuracy": 0.7980656400322914,
      "num_tokens": 444754720.0,
      "step": 13934
    },
    {
      "entropy": 0.7607994414865971,
      "epoch": 1.2820915348387876,
      "grad_norm": 0.16631393134593964,
      "learning_rate": 5.7266844542582884e-05,
      "loss": 0.7677,
      "mean_token_accuracy": 0.7708362825214863,
      "num_tokens": 444786889.0,
      "step": 13935
    },
    {
      "entropy": 0.7072200961410999,
      "epoch": 1.282183541210229,
      "grad_norm": 0.16468006372451782,
      "learning_rate": 5.726377771644126e-05,
      "loss": 0.6758,
      "mean_token_accuracy": 0.7955215871334076,
      "num_tokens": 444818092.0,
      "step": 13936
    },
    {
      "entropy": 0.6496148770675063,
      "epoch": 1.28227554758167,
      "grad_norm": 0.1569986790418625,
      "learning_rate": 5.726071089029963e-05,
      "loss": 0.6418,
      "mean_token_accuracy": 0.8037200570106506,
      "num_tokens": 444850142.0,
      "step": 13937
    },
    {
      "entropy": 0.6366099659353495,
      "epoch": 1.2823675539531112,
      "grad_norm": 0.16238898038864136,
      "learning_rate": 5.7257644064158e-05,
      "loss": 0.6339,
      "mean_token_accuracy": 0.8054422065615654,
      "num_tokens": 444882545.0,
      "step": 13938
    },
    {
      "entropy": 0.6071907859295607,
      "epoch": 1.2824595603245525,
      "grad_norm": 0.1558872014284134,
      "learning_rate": 5.725457723801638e-05,
      "loss": 0.595,
      "mean_token_accuracy": 0.8213501311838627,
      "num_tokens": 444914519.0,
      "step": 13939
    },
    {
      "entropy": 0.6177520286291838,
      "epoch": 1.2825515666959937,
      "grad_norm": 0.15828369557857513,
      "learning_rate": 5.725151041187475e-05,
      "loss": 0.6163,
      "mean_token_accuracy": 0.8105834648013115,
      "num_tokens": 444946182.0,
      "step": 13940
    },
    {
      "entropy": 0.6746214497834444,
      "epoch": 1.282643573067435,
      "grad_norm": 0.16179712116718292,
      "learning_rate": 5.7248443585733125e-05,
      "loss": 0.6553,
      "mean_token_accuracy": 0.7999606244266033,
      "num_tokens": 444977911.0,
      "step": 13941
    },
    {
      "entropy": 0.7045242842286825,
      "epoch": 1.2827355794388762,
      "grad_norm": 0.1626707911491394,
      "learning_rate": 5.724537675959151e-05,
      "loss": 0.706,
      "mean_token_accuracy": 0.7827534824609756,
      "num_tokens": 445009113.0,
      "step": 13942
    },
    {
      "entropy": 0.5822938377968967,
      "epoch": 1.2828275858103173,
      "grad_norm": 0.16205139458179474,
      "learning_rate": 5.7242309933449875e-05,
      "loss": 0.5674,
      "mean_token_accuracy": 0.8221582062542439,
      "num_tokens": 445041304.0,
      "step": 13943
    },
    {
      "entropy": 0.65912951156497,
      "epoch": 1.2829195921817587,
      "grad_norm": 0.15650713443756104,
      "learning_rate": 5.723924310730825e-05,
      "loss": 0.649,
      "mean_token_accuracy": 0.8004701137542725,
      "num_tokens": 445073219.0,
      "step": 13944
    },
    {
      "entropy": 0.677761547267437,
      "epoch": 1.2830115985531998,
      "grad_norm": 0.16562122106552124,
      "learning_rate": 5.723617628116662e-05,
      "loss": 0.6763,
      "mean_token_accuracy": 0.7949628047645092,
      "num_tokens": 445105216.0,
      "step": 13945
    },
    {
      "entropy": 0.6676440332084894,
      "epoch": 1.2831036049246412,
      "grad_norm": 0.1578352153301239,
      "learning_rate": 5.7233109455025005e-05,
      "loss": 0.657,
      "mean_token_accuracy": 0.7966528162360191,
      "num_tokens": 445137497.0,
      "step": 13946
    },
    {
      "entropy": 0.6746672447770834,
      "epoch": 1.2831956112960823,
      "grad_norm": 0.15596118569374084,
      "learning_rate": 5.723004262888337e-05,
      "loss": 0.6835,
      "mean_token_accuracy": 0.7942074202001095,
      "num_tokens": 445169248.0,
      "step": 13947
    },
    {
      "entropy": 0.7151540089398623,
      "epoch": 1.2832876176675234,
      "grad_norm": 0.16903771460056305,
      "learning_rate": 5.722697580274175e-05,
      "loss": 0.7209,
      "mean_token_accuracy": 0.7843733876943588,
      "num_tokens": 445200550.0,
      "step": 13948
    },
    {
      "entropy": 0.7138705928809941,
      "epoch": 1.2833796240389648,
      "grad_norm": 0.16670918464660645,
      "learning_rate": 5.7223908976600116e-05,
      "loss": 0.6989,
      "mean_token_accuracy": 0.7900672629475594,
      "num_tokens": 445232608.0,
      "step": 13949
    },
    {
      "entropy": 0.7321032993495464,
      "epoch": 1.283471630410406,
      "grad_norm": 0.15805403888225555,
      "learning_rate": 5.72208421504585e-05,
      "loss": 0.7244,
      "mean_token_accuracy": 0.7813287377357483,
      "num_tokens": 445264878.0,
      "step": 13950
    },
    {
      "entropy": 0.5995832923799753,
      "epoch": 1.2835636367818473,
      "grad_norm": 0.16169840097427368,
      "learning_rate": 5.721777532431687e-05,
      "loss": 0.5761,
      "mean_token_accuracy": 0.8201040625572205,
      "num_tokens": 445296344.0,
      "step": 13951
    },
    {
      "entropy": 0.6074795294553041,
      "epoch": 1.2836556431532884,
      "grad_norm": 0.1568029522895813,
      "learning_rate": 5.721470849817524e-05,
      "loss": 0.5985,
      "mean_token_accuracy": 0.816398274153471,
      "num_tokens": 445328715.0,
      "step": 13952
    },
    {
      "entropy": 0.6095465878024697,
      "epoch": 1.2837476495247295,
      "grad_norm": 0.15441302955150604,
      "learning_rate": 5.7211641672033614e-05,
      "loss": 0.5839,
      "mean_token_accuracy": 0.8216642960906029,
      "num_tokens": 445360768.0,
      "step": 13953
    },
    {
      "entropy": 0.5625454783439636,
      "epoch": 1.2838396558961709,
      "grad_norm": 0.16092853248119354,
      "learning_rate": 5.7208574845891996e-05,
      "loss": 0.5441,
      "mean_token_accuracy": 0.833562683314085,
      "num_tokens": 445392786.0,
      "step": 13954
    },
    {
      "entropy": 0.6417212653905153,
      "epoch": 1.283931662267612,
      "grad_norm": 0.15759584307670593,
      "learning_rate": 5.7205508019750364e-05,
      "loss": 0.6327,
      "mean_token_accuracy": 0.8066905029118061,
      "num_tokens": 445425336.0,
      "step": 13955
    },
    {
      "entropy": 0.6172169037163258,
      "epoch": 1.2840236686390534,
      "grad_norm": 0.1521005630493164,
      "learning_rate": 5.720244119360874e-05,
      "loss": 0.6078,
      "mean_token_accuracy": 0.8140291236341,
      "num_tokens": 445457602.0,
      "step": 13956
    },
    {
      "entropy": 0.6092513846233487,
      "epoch": 1.2841156750104945,
      "grad_norm": 0.1618632674217224,
      "learning_rate": 5.7199374367467106e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8154792934656143,
      "num_tokens": 445489601.0,
      "step": 13957
    },
    {
      "entropy": 0.5709778852760792,
      "epoch": 1.2842076813819356,
      "grad_norm": 0.15427421033382416,
      "learning_rate": 5.719630754132549e-05,
      "loss": 0.5553,
      "mean_token_accuracy": 0.8311301358044147,
      "num_tokens": 445521740.0,
      "step": 13958
    },
    {
      "entropy": 0.6190004323143512,
      "epoch": 1.284299687753377,
      "grad_norm": 0.156508669257164,
      "learning_rate": 5.719324071518386e-05,
      "loss": 0.5954,
      "mean_token_accuracy": 0.8157352060079575,
      "num_tokens": 445552863.0,
      "step": 13959
    },
    {
      "entropy": 0.5364960813894868,
      "epoch": 1.2843916941248181,
      "grad_norm": 0.15338818728923798,
      "learning_rate": 5.719017388904223e-05,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8396273888647556,
      "num_tokens": 445584758.0,
      "step": 13960
    },
    {
      "entropy": 0.6696375943720341,
      "epoch": 1.2844837004962595,
      "grad_norm": 0.16528908908367157,
      "learning_rate": 5.7187107062900605e-05,
      "loss": 0.6646,
      "mean_token_accuracy": 0.7996410578489304,
      "num_tokens": 445616781.0,
      "step": 13961
    },
    {
      "entropy": 0.6345050353556871,
      "epoch": 1.2845757068677006,
      "grad_norm": 0.16502732038497925,
      "learning_rate": 5.7184040236758986e-05,
      "loss": 0.6187,
      "mean_token_accuracy": 0.8116022199392319,
      "num_tokens": 445648439.0,
      "step": 13962
    },
    {
      "entropy": 0.6001757886260748,
      "epoch": 1.2846677132391418,
      "grad_norm": 0.1639164835214615,
      "learning_rate": 5.7180973410617354e-05,
      "loss": 0.6053,
      "mean_token_accuracy": 0.8126375451683998,
      "num_tokens": 445679008.0,
      "step": 13963
    },
    {
      "entropy": 0.6863175481557846,
      "epoch": 1.284759719610583,
      "grad_norm": 0.15739378333091736,
      "learning_rate": 5.717790658447573e-05,
      "loss": 0.6861,
      "mean_token_accuracy": 0.7941100560128689,
      "num_tokens": 445710994.0,
      "step": 13964
    },
    {
      "entropy": 0.6191085958853364,
      "epoch": 1.2848517259820242,
      "grad_norm": 0.15476249158382416,
      "learning_rate": 5.71748397583341e-05,
      "loss": 0.6054,
      "mean_token_accuracy": 0.8146487474441528,
      "num_tokens": 445743545.0,
      "step": 13965
    },
    {
      "entropy": 0.5185008223634213,
      "epoch": 1.2849437323534656,
      "grad_norm": 0.15538392961025238,
      "learning_rate": 5.717177293219248e-05,
      "loss": 0.5088,
      "mean_token_accuracy": 0.8449302427470684,
      "num_tokens": 445774648.0,
      "step": 13966
    },
    {
      "entropy": 0.6260475516319275,
      "epoch": 1.2850357387249067,
      "grad_norm": 0.16241499781608582,
      "learning_rate": 5.716870610605085e-05,
      "loss": 0.626,
      "mean_token_accuracy": 0.8121362775564194,
      "num_tokens": 445806560.0,
      "step": 13967
    },
    {
      "entropy": 0.5738927088677883,
      "epoch": 1.2851277450963479,
      "grad_norm": 0.15127766132354736,
      "learning_rate": 5.716563927990922e-05,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8247968629002571,
      "num_tokens": 445838659.0,
      "step": 13968
    },
    {
      "entropy": 0.6371417529881,
      "epoch": 1.2852197514677892,
      "grad_norm": 0.16202884912490845,
      "learning_rate": 5.7162572453767595e-05,
      "loss": 0.6173,
      "mean_token_accuracy": 0.8088314421474934,
      "num_tokens": 445870394.0,
      "step": 13969
    },
    {
      "entropy": 0.5979816876351833,
      "epoch": 1.2853117578392304,
      "grad_norm": 0.15755139291286469,
      "learning_rate": 5.715950562762598e-05,
      "loss": 0.5848,
      "mean_token_accuracy": 0.8170301914215088,
      "num_tokens": 445902373.0,
      "step": 13970
    },
    {
      "entropy": 0.5916828755289316,
      "epoch": 1.2854037642106717,
      "grad_norm": 0.15523701906204224,
      "learning_rate": 5.7156438801484345e-05,
      "loss": 0.5821,
      "mean_token_accuracy": 0.820569958537817,
      "num_tokens": 445934165.0,
      "step": 13971
    },
    {
      "entropy": 0.6583787081763148,
      "epoch": 1.2854957705821128,
      "grad_norm": 0.16748572885990143,
      "learning_rate": 5.715337197534272e-05,
      "loss": 0.6483,
      "mean_token_accuracy": 0.8023795932531357,
      "num_tokens": 445966135.0,
      "step": 13972
    },
    {
      "entropy": 0.6438766606152058,
      "epoch": 1.285587776953554,
      "grad_norm": 0.1546589732170105,
      "learning_rate": 5.71503051492011e-05,
      "loss": 0.63,
      "mean_token_accuracy": 0.801011212170124,
      "num_tokens": 445998161.0,
      "step": 13973
    },
    {
      "entropy": 0.5614010645076632,
      "epoch": 1.2856797833249953,
      "grad_norm": 0.14944209158420563,
      "learning_rate": 5.714723832305947e-05,
      "loss": 0.5567,
      "mean_token_accuracy": 0.8290481679141521,
      "num_tokens": 446030611.0,
      "step": 13974
    },
    {
      "entropy": 0.5279432525858283,
      "epoch": 1.2857717896964365,
      "grad_norm": 0.14851973950862885,
      "learning_rate": 5.714417149691784e-05,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8390407003462315,
      "num_tokens": 446062445.0,
      "step": 13975
    },
    {
      "entropy": 0.6046153949573636,
      "epoch": 1.2858637960678778,
      "grad_norm": 0.1581742912530899,
      "learning_rate": 5.714110467077621e-05,
      "loss": 0.5917,
      "mean_token_accuracy": 0.8145664781332016,
      "num_tokens": 446094503.0,
      "step": 13976
    },
    {
      "entropy": 0.5919231195002794,
      "epoch": 1.285955802439319,
      "grad_norm": 0.16216111183166504,
      "learning_rate": 5.713803784463459e-05,
      "loss": 0.5703,
      "mean_token_accuracy": 0.8223271928727627,
      "num_tokens": 446125472.0,
      "step": 13977
    },
    {
      "entropy": 0.5395850585773587,
      "epoch": 1.28604780881076,
      "grad_norm": 0.15714263916015625,
      "learning_rate": 5.713497101849297e-05,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8413593582808971,
      "num_tokens": 446157714.0,
      "step": 13978
    },
    {
      "entropy": 0.6488213976845145,
      "epoch": 1.2861398151822014,
      "grad_norm": 0.1615578979253769,
      "learning_rate": 5.7131904192351335e-05,
      "loss": 0.6284,
      "mean_token_accuracy": 0.8090370930731297,
      "num_tokens": 446189405.0,
      "step": 13979
    },
    {
      "entropy": 0.5806027632206678,
      "epoch": 1.2862318215536426,
      "grad_norm": 0.15910804271697998,
      "learning_rate": 5.712883736620971e-05,
      "loss": 0.5622,
      "mean_token_accuracy": 0.8247654847800732,
      "num_tokens": 446221301.0,
      "step": 13980
    },
    {
      "entropy": 0.7497548051178455,
      "epoch": 1.286323827925084,
      "grad_norm": 0.1671525537967682,
      "learning_rate": 5.712577054006809e-05,
      "loss": 0.7327,
      "mean_token_accuracy": 0.7817831709980965,
      "num_tokens": 446253236.0,
      "step": 13981
    },
    {
      "entropy": 0.5812160167843103,
      "epoch": 1.286415834296525,
      "grad_norm": 0.16209182143211365,
      "learning_rate": 5.712270371392646e-05,
      "loss": 0.5846,
      "mean_token_accuracy": 0.820591364055872,
      "num_tokens": 446284788.0,
      "step": 13982
    },
    {
      "entropy": 0.5283577041700482,
      "epoch": 1.2865078406679662,
      "grad_norm": 0.14945925772190094,
      "learning_rate": 5.7119636887784834e-05,
      "loss": 0.5203,
      "mean_token_accuracy": 0.842688761651516,
      "num_tokens": 446317013.0,
      "step": 13983
    },
    {
      "entropy": 0.7596057262271643,
      "epoch": 1.2865998470394076,
      "grad_norm": 0.17152191698551178,
      "learning_rate": 5.71165700616432e-05,
      "loss": 0.7474,
      "mean_token_accuracy": 0.7693666592240334,
      "num_tokens": 446347570.0,
      "step": 13984
    },
    {
      "entropy": 0.6353377364575863,
      "epoch": 1.2866918534108487,
      "grad_norm": 0.16053929924964905,
      "learning_rate": 5.711350323550159e-05,
      "loss": 0.6236,
      "mean_token_accuracy": 0.8085015006363392,
      "num_tokens": 446378883.0,
      "step": 13985
    },
    {
      "entropy": 0.5854429882019758,
      "epoch": 1.28678385978229,
      "grad_norm": 0.16099068522453308,
      "learning_rate": 5.711043640935996e-05,
      "loss": 0.5802,
      "mean_token_accuracy": 0.8197457864880562,
      "num_tokens": 446411584.0,
      "step": 13986
    },
    {
      "entropy": 0.5952069954946637,
      "epoch": 1.2868758661537312,
      "grad_norm": 0.16098667681217194,
      "learning_rate": 5.7107369583218325e-05,
      "loss": 0.5954,
      "mean_token_accuracy": 0.81777523458004,
      "num_tokens": 446442804.0,
      "step": 13987
    },
    {
      "entropy": 0.5749766128137708,
      "epoch": 1.2869678725251723,
      "grad_norm": 0.15745709836483002,
      "learning_rate": 5.71043027570767e-05,
      "loss": 0.5555,
      "mean_token_accuracy": 0.8317190408706665,
      "num_tokens": 446474820.0,
      "step": 13988
    },
    {
      "entropy": 0.6016020681709051,
      "epoch": 1.2870598788966137,
      "grad_norm": 0.15473581850528717,
      "learning_rate": 5.710123593093508e-05,
      "loss": 0.5822,
      "mean_token_accuracy": 0.8200168423354626,
      "num_tokens": 446506876.0,
      "step": 13989
    },
    {
      "entropy": 0.6629216633737087,
      "epoch": 1.2871518852680548,
      "grad_norm": 0.15940505266189575,
      "learning_rate": 5.7098169104793456e-05,
      "loss": 0.6462,
      "mean_token_accuracy": 0.8044280670583248,
      "num_tokens": 446539536.0,
      "step": 13990
    },
    {
      "entropy": 0.5560964588075876,
      "epoch": 1.2872438916394962,
      "grad_norm": 0.16345135867595673,
      "learning_rate": 5.7095102278651824e-05,
      "loss": 0.5452,
      "mean_token_accuracy": 0.833584226667881,
      "num_tokens": 446571752.0,
      "step": 13991
    },
    {
      "entropy": 0.7146218800917268,
      "epoch": 1.2873358980109373,
      "grad_norm": 0.1676354855298996,
      "learning_rate": 5.70920354525102e-05,
      "loss": 0.7042,
      "mean_token_accuracy": 0.7861871793866158,
      "num_tokens": 446602964.0,
      "step": 13992
    },
    {
      "entropy": 0.6588567062281072,
      "epoch": 1.2874279043823784,
      "grad_norm": 0.16895334422588348,
      "learning_rate": 5.708896862636858e-05,
      "loss": 0.6412,
      "mean_token_accuracy": 0.8027874603867531,
      "num_tokens": 446634646.0,
      "step": 13993
    },
    {
      "entropy": 0.5945086269639432,
      "epoch": 1.2875199107538198,
      "grad_norm": 0.16393083333969116,
      "learning_rate": 5.708590180022695e-05,
      "loss": 0.5835,
      "mean_token_accuracy": 0.8205623887479305,
      "num_tokens": 446665821.0,
      "step": 13994
    },
    {
      "entropy": 0.733405087608844,
      "epoch": 1.287611917125261,
      "grad_norm": 0.16414162516593933,
      "learning_rate": 5.708283497408532e-05,
      "loss": 0.7425,
      "mean_token_accuracy": 0.7807683125138283,
      "num_tokens": 446697969.0,
      "step": 13995
    },
    {
      "entropy": 0.7093808744102716,
      "epoch": 1.2877039234967023,
      "grad_norm": 0.16647954285144806,
      "learning_rate": 5.707976814794369e-05,
      "loss": 0.7021,
      "mean_token_accuracy": 0.7899001874029636,
      "num_tokens": 446729028.0,
      "step": 13996
    },
    {
      "entropy": 0.5849340502172709,
      "epoch": 1.2877959298681434,
      "grad_norm": 0.15517130494117737,
      "learning_rate": 5.707670132180207e-05,
      "loss": 0.5806,
      "mean_token_accuracy": 0.8197870701551437,
      "num_tokens": 446761470.0,
      "step": 13997
    },
    {
      "entropy": 0.6507070437073708,
      "epoch": 1.2878879362395845,
      "grad_norm": 0.16066132485866547,
      "learning_rate": 5.7073634495660446e-05,
      "loss": 0.6289,
      "mean_token_accuracy": 0.8048723042011261,
      "num_tokens": 446791687.0,
      "step": 13998
    },
    {
      "entropy": 0.5717760939151049,
      "epoch": 1.2879799426110259,
      "grad_norm": 0.1482459306716919,
      "learning_rate": 5.7070567669518814e-05,
      "loss": 0.565,
      "mean_token_accuracy": 0.8260086141526699,
      "num_tokens": 446823872.0,
      "step": 13999
    },
    {
      "entropy": 0.6567368702962995,
      "epoch": 1.288071948982467,
      "grad_norm": 0.15951859951019287,
      "learning_rate": 5.706750084337719e-05,
      "loss": 0.6409,
      "mean_token_accuracy": 0.8066972754895687,
      "num_tokens": 446855248.0,
      "step": 14000
    },
    {
      "entropy": 0.6097631165757775,
      "epoch": 1.2881639553539084,
      "grad_norm": 0.16435714066028595,
      "learning_rate": 5.706443401723557e-05,
      "loss": 0.5792,
      "mean_token_accuracy": 0.8217272758483887,
      "num_tokens": 446886355.0,
      "step": 14001
    },
    {
      "entropy": 0.6166216898709536,
      "epoch": 1.2882559617253495,
      "grad_norm": 0.15488776564598083,
      "learning_rate": 5.706136719109394e-05,
      "loss": 0.6157,
      "mean_token_accuracy": 0.8152743689715862,
      "num_tokens": 446919044.0,
      "step": 14002
    },
    {
      "entropy": 0.6778901340439916,
      "epoch": 1.2883479680967906,
      "grad_norm": 0.15974344313144684,
      "learning_rate": 5.705830036495231e-05,
      "loss": 0.6667,
      "mean_token_accuracy": 0.7955623008310795,
      "num_tokens": 446950318.0,
      "step": 14003
    },
    {
      "entropy": 0.6221089567989111,
      "epoch": 1.288439974468232,
      "grad_norm": 0.15765292942523956,
      "learning_rate": 5.7055233538810694e-05,
      "loss": 0.611,
      "mean_token_accuracy": 0.814859502017498,
      "num_tokens": 446982602.0,
      "step": 14004
    },
    {
      "entropy": 0.6088612340390682,
      "epoch": 1.2885319808396731,
      "grad_norm": 0.153483584523201,
      "learning_rate": 5.705216671266906e-05,
      "loss": 0.5917,
      "mean_token_accuracy": 0.8177281506359577,
      "num_tokens": 447014716.0,
      "step": 14005
    },
    {
      "entropy": 0.7101770620793104,
      "epoch": 1.2886239872111145,
      "grad_norm": 0.16383078694343567,
      "learning_rate": 5.704909988652744e-05,
      "loss": 0.6972,
      "mean_token_accuracy": 0.7875152043998241,
      "num_tokens": 447047151.0,
      "step": 14006
    },
    {
      "entropy": 0.5544429766014218,
      "epoch": 1.2887159935825556,
      "grad_norm": 0.17319828271865845,
      "learning_rate": 5.7046033060385805e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8319817930459976,
      "num_tokens": 447077931.0,
      "step": 14007
    },
    {
      "entropy": 0.6591421263292432,
      "epoch": 1.2888079999539968,
      "grad_norm": 0.1630212813615799,
      "learning_rate": 5.7042966234244186e-05,
      "loss": 0.6375,
      "mean_token_accuracy": 0.8055839389562607,
      "num_tokens": 447109920.0,
      "step": 14008
    },
    {
      "entropy": 0.46549655171111226,
      "epoch": 1.288900006325438,
      "grad_norm": 0.1425088495016098,
      "learning_rate": 5.703989940810256e-05,
      "loss": 0.4442,
      "mean_token_accuracy": 0.8629516586661339,
      "num_tokens": 447142632.0,
      "step": 14009
    },
    {
      "entropy": 0.6074774116277695,
      "epoch": 1.2889920126968792,
      "grad_norm": 0.1634298712015152,
      "learning_rate": 5.703683258196093e-05,
      "loss": 0.5904,
      "mean_token_accuracy": 0.8183351159095764,
      "num_tokens": 447174937.0,
      "step": 14010
    },
    {
      "entropy": 0.6938739642500877,
      "epoch": 1.2890840190683206,
      "grad_norm": 0.16444720327854156,
      "learning_rate": 5.70337657558193e-05,
      "loss": 0.702,
      "mean_token_accuracy": 0.7862946502864361,
      "num_tokens": 447206995.0,
      "step": 14011
    },
    {
      "entropy": 0.5012623826041818,
      "epoch": 1.2891760254397617,
      "grad_norm": 0.14854520559310913,
      "learning_rate": 5.7030698929677685e-05,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8532154522836208,
      "num_tokens": 447239495.0,
      "step": 14012
    },
    {
      "entropy": 0.634602002799511,
      "epoch": 1.2892680318112029,
      "grad_norm": 0.16519108414649963,
      "learning_rate": 5.702763210353605e-05,
      "loss": 0.6177,
      "mean_token_accuracy": 0.8076952733099461,
      "num_tokens": 447270330.0,
      "step": 14013
    },
    {
      "entropy": 0.6271747406572104,
      "epoch": 1.2893600381826442,
      "grad_norm": 0.1542249172925949,
      "learning_rate": 5.702456527739443e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.8092962987720966,
      "num_tokens": 447301917.0,
      "step": 14014
    },
    {
      "entropy": 0.6139727123081684,
      "epoch": 1.2894520445540854,
      "grad_norm": 0.17236867547035217,
      "learning_rate": 5.7021498451252795e-05,
      "loss": 0.6056,
      "mean_token_accuracy": 0.8137241639196873,
      "num_tokens": 447333270.0,
      "step": 14015
    },
    {
      "entropy": 0.6844496484845877,
      "epoch": 1.2895440509255267,
      "grad_norm": 0.15690098702907562,
      "learning_rate": 5.7018431625111177e-05,
      "loss": 0.6643,
      "mean_token_accuracy": 0.7956405282020569,
      "num_tokens": 447364487.0,
      "step": 14016
    },
    {
      "entropy": 0.5106719359755516,
      "epoch": 1.2896360572969678,
      "grad_norm": 0.15298789739608765,
      "learning_rate": 5.701536479896955e-05,
      "loss": 0.4896,
      "mean_token_accuracy": 0.8453160785138607,
      "num_tokens": 447395985.0,
      "step": 14017
    },
    {
      "entropy": 0.7208800902590156,
      "epoch": 1.289728063668409,
      "grad_norm": 0.17328022420406342,
      "learning_rate": 5.701229797282792e-05,
      "loss": 0.7204,
      "mean_token_accuracy": 0.7797926664352417,
      "num_tokens": 447427954.0,
      "step": 14018
    },
    {
      "entropy": 0.5974624836817384,
      "epoch": 1.2898200700398503,
      "grad_norm": 0.15796814858913422,
      "learning_rate": 5.7009231146686294e-05,
      "loss": 0.595,
      "mean_token_accuracy": 0.8174579404294491,
      "num_tokens": 447459256.0,
      "step": 14019
    },
    {
      "entropy": 0.575535804964602,
      "epoch": 1.2899120764112915,
      "grad_norm": 0.15470397472381592,
      "learning_rate": 5.7006164320544675e-05,
      "loss": 0.5672,
      "mean_token_accuracy": 0.8226194828748703,
      "num_tokens": 447491786.0,
      "step": 14020
    },
    {
      "entropy": 0.6230591777712107,
      "epoch": 1.2900040827827328,
      "grad_norm": 0.16041213274002075,
      "learning_rate": 5.700309749440304e-05,
      "loss": 0.6184,
      "mean_token_accuracy": 0.8154834359884262,
      "num_tokens": 447523283.0,
      "step": 14021
    },
    {
      "entropy": 0.6820690631866455,
      "epoch": 1.290096089154174,
      "grad_norm": 0.16331136226654053,
      "learning_rate": 5.700003066826142e-05,
      "loss": 0.6588,
      "mean_token_accuracy": 0.7973615862429142,
      "num_tokens": 447554495.0,
      "step": 14022
    },
    {
      "entropy": 0.616450612898916,
      "epoch": 1.290188095525615,
      "grad_norm": 0.15609534084796906,
      "learning_rate": 5.6996963842119786e-05,
      "loss": 0.6066,
      "mean_token_accuracy": 0.8113955482840538,
      "num_tokens": 447586977.0,
      "step": 14023
    },
    {
      "entropy": 0.6203049449250102,
      "epoch": 1.2902801018970564,
      "grad_norm": 0.15814615786075592,
      "learning_rate": 5.6993897015978174e-05,
      "loss": 0.6033,
      "mean_token_accuracy": 0.8109190985560417,
      "num_tokens": 447618762.0,
      "step": 14024
    },
    {
      "entropy": 0.5562991872429848,
      "epoch": 1.2903721082684976,
      "grad_norm": 0.15654614567756653,
      "learning_rate": 5.699083018983654e-05,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8284711390733719,
      "num_tokens": 447650071.0,
      "step": 14025
    },
    {
      "entropy": 0.6180775202810764,
      "epoch": 1.290464114639939,
      "grad_norm": 0.16384531557559967,
      "learning_rate": 5.698776336369491e-05,
      "loss": 0.5926,
      "mean_token_accuracy": 0.8194349743425846,
      "num_tokens": 447682380.0,
      "step": 14026
    },
    {
      "entropy": 0.6151044275611639,
      "epoch": 1.29055612101138,
      "grad_norm": 0.1552550047636032,
      "learning_rate": 5.6984696537553284e-05,
      "loss": 0.6084,
      "mean_token_accuracy": 0.8162878453731537,
      "num_tokens": 447713656.0,
      "step": 14027
    },
    {
      "entropy": 0.6558500826358795,
      "epoch": 1.2906481273828212,
      "grad_norm": 0.1668684184551239,
      "learning_rate": 5.6981629711411666e-05,
      "loss": 0.6456,
      "mean_token_accuracy": 0.8004801832139492,
      "num_tokens": 447745542.0,
      "step": 14028
    },
    {
      "entropy": 0.5617868755944073,
      "epoch": 1.2907401337542626,
      "grad_norm": 0.14994406700134277,
      "learning_rate": 5.697856288527004e-05,
      "loss": 0.5592,
      "mean_token_accuracy": 0.827654343098402,
      "num_tokens": 447778046.0,
      "step": 14029
    },
    {
      "entropy": 0.6707556173205376,
      "epoch": 1.2908321401257037,
      "grad_norm": 0.1630813628435135,
      "learning_rate": 5.697549605912841e-05,
      "loss": 0.6699,
      "mean_token_accuracy": 0.7948749735951424,
      "num_tokens": 447810036.0,
      "step": 14030
    },
    {
      "entropy": 0.48306333855725825,
      "epoch": 1.290924146497145,
      "grad_norm": 0.14897659420967102,
      "learning_rate": 5.6972429232986776e-05,
      "loss": 0.4471,
      "mean_token_accuracy": 0.8607721216976643,
      "num_tokens": 447841635.0,
      "step": 14031
    },
    {
      "entropy": 0.5812762696295977,
      "epoch": 1.2910161528685862,
      "grad_norm": 0.15191856026649475,
      "learning_rate": 5.6969362406845164e-05,
      "loss": 0.5791,
      "mean_token_accuracy": 0.8170327693223953,
      "num_tokens": 447873984.0,
      "step": 14032
    },
    {
      "entropy": 0.5871564168483019,
      "epoch": 1.2911081592400273,
      "grad_norm": 0.15987803041934967,
      "learning_rate": 5.696629558070353e-05,
      "loss": 0.5591,
      "mean_token_accuracy": 0.831376489251852,
      "num_tokens": 447906139.0,
      "step": 14033
    },
    {
      "entropy": 0.6109083108603954,
      "epoch": 1.2912001656114687,
      "grad_norm": 0.15518344938755035,
      "learning_rate": 5.696322875456191e-05,
      "loss": 0.5913,
      "mean_token_accuracy": 0.815921276807785,
      "num_tokens": 447937886.0,
      "step": 14034
    },
    {
      "entropy": 0.6980117820203304,
      "epoch": 1.2912921719829098,
      "grad_norm": 0.1584400236606598,
      "learning_rate": 5.696016192842029e-05,
      "loss": 0.6853,
      "mean_token_accuracy": 0.7895571328699589,
      "num_tokens": 447969491.0,
      "step": 14035
    },
    {
      "entropy": 0.5791720254346728,
      "epoch": 1.2913841783543512,
      "grad_norm": 0.15643109381198883,
      "learning_rate": 5.6957095102278656e-05,
      "loss": 0.5842,
      "mean_token_accuracy": 0.8205494657158852,
      "num_tokens": 448001986.0,
      "step": 14036
    },
    {
      "entropy": 0.5982432682067156,
      "epoch": 1.2914761847257923,
      "grad_norm": 0.1671922504901886,
      "learning_rate": 5.695402827613703e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8189477138221264,
      "num_tokens": 448033622.0,
      "step": 14037
    },
    {
      "entropy": 0.483365660533309,
      "epoch": 1.2915681910972334,
      "grad_norm": 0.14973287284374237,
      "learning_rate": 5.69509614499954e-05,
      "loss": 0.4844,
      "mean_token_accuracy": 0.8528380580246449,
      "num_tokens": 448065343.0,
      "step": 14038
    },
    {
      "entropy": 0.648288307711482,
      "epoch": 1.2916601974686748,
      "grad_norm": 0.15763112902641296,
      "learning_rate": 5.694789462385378e-05,
      "loss": 0.6312,
      "mean_token_accuracy": 0.8097008652985096,
      "num_tokens": 448097146.0,
      "step": 14039
    },
    {
      "entropy": 0.6848084889352322,
      "epoch": 1.291752203840116,
      "grad_norm": 0.17653749883174896,
      "learning_rate": 5.6944827797712155e-05,
      "loss": 0.6721,
      "mean_token_accuracy": 0.7992489002645016,
      "num_tokens": 448129914.0,
      "step": 14040
    },
    {
      "entropy": 0.7010174747556448,
      "epoch": 1.2918442102115573,
      "grad_norm": 0.17343966662883759,
      "learning_rate": 5.694176097157052e-05,
      "loss": 0.705,
      "mean_token_accuracy": 0.788276668637991,
      "num_tokens": 448161606.0,
      "step": 14041
    },
    {
      "entropy": 0.5856279917061329,
      "epoch": 1.2919362165829984,
      "grad_norm": 0.1618364006280899,
      "learning_rate": 5.69386941454289e-05,
      "loss": 0.5622,
      "mean_token_accuracy": 0.826987013220787,
      "num_tokens": 448193218.0,
      "step": 14042
    },
    {
      "entropy": 0.6633358001708984,
      "epoch": 1.2920282229544395,
      "grad_norm": 0.1680656373500824,
      "learning_rate": 5.693562731928728e-05,
      "loss": 0.6594,
      "mean_token_accuracy": 0.800275094807148,
      "num_tokens": 448225307.0,
      "step": 14043
    },
    {
      "entropy": 0.7006569169461727,
      "epoch": 1.2921202293258809,
      "grad_norm": 0.16690264642238617,
      "learning_rate": 5.6932560493145646e-05,
      "loss": 0.6997,
      "mean_token_accuracy": 0.791817918419838,
      "num_tokens": 448256679.0,
      "step": 14044
    },
    {
      "entropy": 0.6343168709427118,
      "epoch": 1.292212235697322,
      "grad_norm": 0.15689215064048767,
      "learning_rate": 5.692949366700402e-05,
      "loss": 0.6122,
      "mean_token_accuracy": 0.8124607615172863,
      "num_tokens": 448288404.0,
      "step": 14045
    },
    {
      "entropy": 0.629418296739459,
      "epoch": 1.2923042420687634,
      "grad_norm": 0.1551646739244461,
      "learning_rate": 5.692642684086239e-05,
      "loss": 0.6054,
      "mean_token_accuracy": 0.8143397942185402,
      "num_tokens": 448320796.0,
      "step": 14046
    },
    {
      "entropy": 0.6224557785317302,
      "epoch": 1.2923962484402045,
      "grad_norm": 0.15654326975345612,
      "learning_rate": 5.692336001472077e-05,
      "loss": 0.5935,
      "mean_token_accuracy": 0.8189258426427841,
      "num_tokens": 448353087.0,
      "step": 14047
    },
    {
      "entropy": 0.6442159609869123,
      "epoch": 1.2924882548116456,
      "grad_norm": 0.15034157037734985,
      "learning_rate": 5.6920293188579145e-05,
      "loss": 0.6256,
      "mean_token_accuracy": 0.8104140311479568,
      "num_tokens": 448385392.0,
      "step": 14048
    },
    {
      "entropy": 0.572890181094408,
      "epoch": 1.292580261183087,
      "grad_norm": 0.15265583992004395,
      "learning_rate": 5.691722636243751e-05,
      "loss": 0.5561,
      "mean_token_accuracy": 0.8227478004992008,
      "num_tokens": 448416615.0,
      "step": 14049
    },
    {
      "entropy": 0.603379406966269,
      "epoch": 1.2926722675545281,
      "grad_norm": 0.1586771309375763,
      "learning_rate": 5.691415953629589e-05,
      "loss": 0.5902,
      "mean_token_accuracy": 0.8209464773535728,
      "num_tokens": 448449212.0,
      "step": 14050
    },
    {
      "entropy": 0.6499553807079792,
      "epoch": 1.2927642739259695,
      "grad_norm": 0.1621023416519165,
      "learning_rate": 5.691109271015427e-05,
      "loss": 0.6492,
      "mean_token_accuracy": 0.804377444088459,
      "num_tokens": 448480467.0,
      "step": 14051
    },
    {
      "entropy": 0.58765016682446,
      "epoch": 1.2928562802974106,
      "grad_norm": 0.15311236679553986,
      "learning_rate": 5.690802588401264e-05,
      "loss": 0.5704,
      "mean_token_accuracy": 0.8245320580899715,
      "num_tokens": 448512688.0,
      "step": 14052
    },
    {
      "entropy": 0.6433910904452205,
      "epoch": 1.2929482866688518,
      "grad_norm": 0.16068534553050995,
      "learning_rate": 5.690495905787101e-05,
      "loss": 0.6451,
      "mean_token_accuracy": 0.8060132712125778,
      "num_tokens": 448544374.0,
      "step": 14053
    },
    {
      "entropy": 0.6641067024320364,
      "epoch": 1.293040293040293,
      "grad_norm": 0.16141332685947418,
      "learning_rate": 5.690189223172938e-05,
      "loss": 0.6407,
      "mean_token_accuracy": 0.8015515245497227,
      "num_tokens": 448574912.0,
      "step": 14054
    },
    {
      "entropy": 0.6983122956007719,
      "epoch": 1.2931322994117342,
      "grad_norm": 0.1654028445482254,
      "learning_rate": 5.689882540558776e-05,
      "loss": 0.6811,
      "mean_token_accuracy": 0.7952557057142258,
      "num_tokens": 448607082.0,
      "step": 14055
    },
    {
      "entropy": 0.66097541898489,
      "epoch": 1.2932243057831756,
      "grad_norm": 0.15693804621696472,
      "learning_rate": 5.6895758579446135e-05,
      "loss": 0.6332,
      "mean_token_accuracy": 0.8027726970613003,
      "num_tokens": 448638610.0,
      "step": 14056
    },
    {
      "entropy": 0.5819047586992383,
      "epoch": 1.2933163121546167,
      "grad_norm": 0.1555127650499344,
      "learning_rate": 5.68926917533045e-05,
      "loss": 0.5934,
      "mean_token_accuracy": 0.815970741212368,
      "num_tokens": 448670636.0,
      "step": 14057
    },
    {
      "entropy": 0.7229906804859638,
      "epoch": 1.2934083185260579,
      "grad_norm": 0.15775753557682037,
      "learning_rate": 5.688962492716288e-05,
      "loss": 0.7006,
      "mean_token_accuracy": 0.7858829163014889,
      "num_tokens": 448702983.0,
      "step": 14058
    },
    {
      "entropy": 0.6025979295372963,
      "epoch": 1.2935003248974992,
      "grad_norm": 0.1608542948961258,
      "learning_rate": 5.688655810102126e-05,
      "loss": 0.6028,
      "mean_token_accuracy": 0.81197689473629,
      "num_tokens": 448733793.0,
      "step": 14059
    },
    {
      "entropy": 0.7050354108214378,
      "epoch": 1.2935923312689404,
      "grad_norm": 0.16754406690597534,
      "learning_rate": 5.688349127487963e-05,
      "loss": 0.6911,
      "mean_token_accuracy": 0.7910335548222065,
      "num_tokens": 448765375.0,
      "step": 14060
    },
    {
      "entropy": 0.6641436722129583,
      "epoch": 1.2936843376403817,
      "grad_norm": 0.16714854538440704,
      "learning_rate": 5.6880424448738e-05,
      "loss": 0.6504,
      "mean_token_accuracy": 0.7977238595485687,
      "num_tokens": 448796883.0,
      "step": 14061
    },
    {
      "entropy": 0.5575748579576612,
      "epoch": 1.2937763440118228,
      "grad_norm": 0.15502220392227173,
      "learning_rate": 5.687735762259637e-05,
      "loss": 0.5389,
      "mean_token_accuracy": 0.8297434635460377,
      "num_tokens": 448828476.0,
      "step": 14062
    },
    {
      "entropy": 0.6815229672938585,
      "epoch": 1.293868350383264,
      "grad_norm": 0.15756450593471527,
      "learning_rate": 5.687429079645475e-05,
      "loss": 0.6784,
      "mean_token_accuracy": 0.7926940768957138,
      "num_tokens": 448860903.0,
      "step": 14063
    },
    {
      "entropy": 0.6100843865424395,
      "epoch": 1.2939603567547053,
      "grad_norm": 0.15559224784374237,
      "learning_rate": 5.6871223970313126e-05,
      "loss": 0.5835,
      "mean_token_accuracy": 0.8218072839081287,
      "num_tokens": 448893491.0,
      "step": 14064
    },
    {
      "entropy": 0.6189542328938842,
      "epoch": 1.2940523631261465,
      "grad_norm": 0.16164349019527435,
      "learning_rate": 5.6868157144171494e-05,
      "loss": 0.6045,
      "mean_token_accuracy": 0.8155557066202164,
      "num_tokens": 448925186.0,
      "step": 14065
    },
    {
      "entropy": 0.5764297544956207,
      "epoch": 1.2941443694975878,
      "grad_norm": 0.1612176150083542,
      "learning_rate": 5.686509031802988e-05,
      "loss": 0.5786,
      "mean_token_accuracy": 0.8266394473612309,
      "num_tokens": 448956808.0,
      "step": 14066
    },
    {
      "entropy": 0.5941633246839046,
      "epoch": 1.294236375869029,
      "grad_norm": 0.16545164585113525,
      "learning_rate": 5.686202349188825e-05,
      "loss": 0.5892,
      "mean_token_accuracy": 0.8185088485479355,
      "num_tokens": 448989140.0,
      "step": 14067
    },
    {
      "entropy": 0.7234292114153504,
      "epoch": 1.29432838224047,
      "grad_norm": 0.16196668148040771,
      "learning_rate": 5.6858956665746624e-05,
      "loss": 0.7257,
      "mean_token_accuracy": 0.7807888649404049,
      "num_tokens": 449021416.0,
      "step": 14068
    },
    {
      "entropy": 0.580414017662406,
      "epoch": 1.2944203886119114,
      "grad_norm": 0.15682506561279297,
      "learning_rate": 5.685588983960499e-05,
      "loss": 0.5758,
      "mean_token_accuracy": 0.8244771026074886,
      "num_tokens": 449053322.0,
      "step": 14069
    },
    {
      "entropy": 0.5594294304028153,
      "epoch": 1.2945123949833526,
      "grad_norm": 0.1596907526254654,
      "learning_rate": 5.6852823013463374e-05,
      "loss": 0.5445,
      "mean_token_accuracy": 0.831557996571064,
      "num_tokens": 449085830.0,
      "step": 14070
    },
    {
      "entropy": 0.6598297758027911,
      "epoch": 1.294604401354794,
      "grad_norm": 0.1586473435163498,
      "learning_rate": 5.684975618732175e-05,
      "loss": 0.6463,
      "mean_token_accuracy": 0.8024885691702366,
      "num_tokens": 449117877.0,
      "step": 14071
    },
    {
      "entropy": 0.6655331193469465,
      "epoch": 1.294696407726235,
      "grad_norm": 0.1641789674758911,
      "learning_rate": 5.6846689361180116e-05,
      "loss": 0.67,
      "mean_token_accuracy": 0.7966321669518948,
      "num_tokens": 449150397.0,
      "step": 14072
    },
    {
      "entropy": 0.6540958415716887,
      "epoch": 1.2947884140976762,
      "grad_norm": 0.1489747166633606,
      "learning_rate": 5.684362253503849e-05,
      "loss": 0.6489,
      "mean_token_accuracy": 0.7993064932525158,
      "num_tokens": 449182835.0,
      "step": 14073
    },
    {
      "entropy": 0.6711306758224964,
      "epoch": 1.2948804204691176,
      "grad_norm": 0.16434425115585327,
      "learning_rate": 5.684055570889687e-05,
      "loss": 0.6517,
      "mean_token_accuracy": 0.7964199595153332,
      "num_tokens": 449213812.0,
      "step": 14074
    },
    {
      "entropy": 0.6113068498671055,
      "epoch": 1.2949724268405587,
      "grad_norm": 0.15490293502807617,
      "learning_rate": 5.683748888275524e-05,
      "loss": 0.6003,
      "mean_token_accuracy": 0.8143842630088329,
      "num_tokens": 449245881.0,
      "step": 14075
    },
    {
      "entropy": 0.6176041625440121,
      "epoch": 1.295064433212,
      "grad_norm": 0.15932367742061615,
      "learning_rate": 5.6834422056613615e-05,
      "loss": 0.6212,
      "mean_token_accuracy": 0.8108608350157738,
      "num_tokens": 449278277.0,
      "step": 14076
    },
    {
      "entropy": 0.58239037822932,
      "epoch": 1.2951564395834412,
      "grad_norm": 0.15765854716300964,
      "learning_rate": 5.683135523047198e-05,
      "loss": 0.5618,
      "mean_token_accuracy": 0.8284293077886105,
      "num_tokens": 449310197.0,
      "step": 14077
    },
    {
      "entropy": 0.6309778653085232,
      "epoch": 1.2952484459548823,
      "grad_norm": 0.15861079096794128,
      "learning_rate": 5.6828288404330364e-05,
      "loss": 0.6196,
      "mean_token_accuracy": 0.8132680729031563,
      "num_tokens": 449342888.0,
      "step": 14078
    },
    {
      "entropy": 0.6578690055757761,
      "epoch": 1.2953404523263237,
      "grad_norm": 0.15994690358638763,
      "learning_rate": 5.682522157818874e-05,
      "loss": 0.6535,
      "mean_token_accuracy": 0.7999934181571007,
      "num_tokens": 449374437.0,
      "step": 14079
    },
    {
      "entropy": 0.733477808535099,
      "epoch": 1.2954324586977648,
      "grad_norm": 0.19178566336631775,
      "learning_rate": 5.6822154752047107e-05,
      "loss": 0.7458,
      "mean_token_accuracy": 0.7764397412538528,
      "num_tokens": 449407182.0,
      "step": 14080
    },
    {
      "entropy": 0.5661391830071807,
      "epoch": 1.2955244650692062,
      "grad_norm": 0.15106570720672607,
      "learning_rate": 5.681908792590548e-05,
      "loss": 0.5471,
      "mean_token_accuracy": 0.8324419930577278,
      "num_tokens": 449438989.0,
      "step": 14081
    },
    {
      "entropy": 0.6237515266984701,
      "epoch": 1.2956164714406473,
      "grad_norm": 0.15613976120948792,
      "learning_rate": 5.681602109976386e-05,
      "loss": 0.606,
      "mean_token_accuracy": 0.8152570277452469,
      "num_tokens": 449471346.0,
      "step": 14082
    },
    {
      "entropy": 0.7008398193866014,
      "epoch": 1.2957084778120884,
      "grad_norm": 0.16127519309520721,
      "learning_rate": 5.681295427362223e-05,
      "loss": 0.6839,
      "mean_token_accuracy": 0.789038673043251,
      "num_tokens": 449502842.0,
      "step": 14083
    },
    {
      "entropy": 0.7049426632001996,
      "epoch": 1.2958004841835298,
      "grad_norm": 0.16008538007736206,
      "learning_rate": 5.6809887447480605e-05,
      "loss": 0.7039,
      "mean_token_accuracy": 0.7893760167062283,
      "num_tokens": 449535260.0,
      "step": 14084
    },
    {
      "entropy": 0.6113659460097551,
      "epoch": 1.295892490554971,
      "grad_norm": 0.15325403213500977,
      "learning_rate": 5.680682062133897e-05,
      "loss": 0.5855,
      "mean_token_accuracy": 0.8176307305693626,
      "num_tokens": 449567048.0,
      "step": 14085
    },
    {
      "entropy": 0.6074004825204611,
      "epoch": 1.2959844969264123,
      "grad_norm": 0.15154927968978882,
      "learning_rate": 5.6803753795197354e-05,
      "loss": 0.5989,
      "mean_token_accuracy": 0.8181116804480553,
      "num_tokens": 449599237.0,
      "step": 14086
    },
    {
      "entropy": 0.6355305220931768,
      "epoch": 1.2960765032978534,
      "grad_norm": 0.15137259662151337,
      "learning_rate": 5.680068696905573e-05,
      "loss": 0.6007,
      "mean_token_accuracy": 0.8158092424273491,
      "num_tokens": 449631505.0,
      "step": 14087
    },
    {
      "entropy": 0.6404386050999165,
      "epoch": 1.2961685096692945,
      "grad_norm": 0.15710531175136566,
      "learning_rate": 5.67976201429141e-05,
      "loss": 0.6144,
      "mean_token_accuracy": 0.8128298744559288,
      "num_tokens": 449663940.0,
      "step": 14088
    },
    {
      "entropy": 0.6964611783623695,
      "epoch": 1.2962605160407359,
      "grad_norm": 0.15619881451129913,
      "learning_rate": 5.679455331677247e-05,
      "loss": 0.6821,
      "mean_token_accuracy": 0.7939391620457172,
      "num_tokens": 449695639.0,
      "step": 14089
    },
    {
      "entropy": 0.574175201356411,
      "epoch": 1.296352522412177,
      "grad_norm": 0.15994830429553986,
      "learning_rate": 5.679148649063085e-05,
      "loss": 0.5633,
      "mean_token_accuracy": 0.827039260417223,
      "num_tokens": 449727722.0,
      "step": 14090
    },
    {
      "entropy": 0.5828412147238851,
      "epoch": 1.2964445287836184,
      "grad_norm": 0.15574584901332855,
      "learning_rate": 5.678841966448922e-05,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8332378752529621,
      "num_tokens": 449759403.0,
      "step": 14091
    },
    {
      "entropy": 0.6094749793410301,
      "epoch": 1.2965365351550595,
      "grad_norm": 0.16095614433288574,
      "learning_rate": 5.6785352838347596e-05,
      "loss": 0.6045,
      "mean_token_accuracy": 0.8162420094013214,
      "num_tokens": 449791398.0,
      "step": 14092
    },
    {
      "entropy": 0.600488813361153,
      "epoch": 1.2966285415265006,
      "grad_norm": 0.1525934636592865,
      "learning_rate": 5.6782286012205963e-05,
      "loss": 0.5694,
      "mean_token_accuracy": 0.8231056444346905,
      "num_tokens": 449822726.0,
      "step": 14093
    },
    {
      "entropy": 0.6226524170488119,
      "epoch": 1.296720547897942,
      "grad_norm": 0.16088783740997314,
      "learning_rate": 5.6779219186064345e-05,
      "loss": 0.626,
      "mean_token_accuracy": 0.8098371624946594,
      "num_tokens": 449854663.0,
      "step": 14094
    },
    {
      "entropy": 0.6828446816653013,
      "epoch": 1.2968125542693831,
      "grad_norm": 0.16849280893802643,
      "learning_rate": 5.677615235992272e-05,
      "loss": 0.6687,
      "mean_token_accuracy": 0.7958424687385559,
      "num_tokens": 449886251.0,
      "step": 14095
    },
    {
      "entropy": 0.5238605532795191,
      "epoch": 1.2969045606408245,
      "grad_norm": 0.14240726828575134,
      "learning_rate": 5.677308553378109e-05,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8399419225752354,
      "num_tokens": 449918425.0,
      "step": 14096
    },
    {
      "entropy": 0.676772229373455,
      "epoch": 1.2969965670122656,
      "grad_norm": 0.1636686623096466,
      "learning_rate": 5.677001870763947e-05,
      "loss": 0.6731,
      "mean_token_accuracy": 0.7952029928565025,
      "num_tokens": 449950588.0,
      "step": 14097
    },
    {
      "entropy": 0.5867391061037779,
      "epoch": 1.2970885733837068,
      "grad_norm": 0.15680089592933655,
      "learning_rate": 5.6766951881497843e-05,
      "loss": 0.573,
      "mean_token_accuracy": 0.8227995485067368,
      "num_tokens": 449983070.0,
      "step": 14098
    },
    {
      "entropy": 0.6978129968047142,
      "epoch": 1.297180579755148,
      "grad_norm": 0.16530677676200867,
      "learning_rate": 5.676388505535621e-05,
      "loss": 0.6777,
      "mean_token_accuracy": 0.7931539006531239,
      "num_tokens": 450015208.0,
      "step": 14099
    },
    {
      "entropy": 0.6330585516989231,
      "epoch": 1.2972725861265892,
      "grad_norm": 0.1622617542743683,
      "learning_rate": 5.6760818229214586e-05,
      "loss": 0.6192,
      "mean_token_accuracy": 0.8104122765362263,
      "num_tokens": 450047049.0,
      "step": 14100
    },
    {
      "entropy": 0.6621448807418346,
      "epoch": 1.2973645924980306,
      "grad_norm": 0.16606557369232178,
      "learning_rate": 5.675775140307297e-05,
      "loss": 0.6525,
      "mean_token_accuracy": 0.7987774759531021,
      "num_tokens": 450079409.0,
      "step": 14101
    },
    {
      "entropy": 0.6189676988869905,
      "epoch": 1.2974565988694717,
      "grad_norm": 0.15570539236068726,
      "learning_rate": 5.6754684576931335e-05,
      "loss": 0.6031,
      "mean_token_accuracy": 0.8122152462601662,
      "num_tokens": 450111334.0,
      "step": 14102
    },
    {
      "entropy": 0.6781708258204162,
      "epoch": 1.2975486052409129,
      "grad_norm": 0.16659598052501678,
      "learning_rate": 5.675161775078971e-05,
      "loss": 0.6795,
      "mean_token_accuracy": 0.7911194451153278,
      "num_tokens": 450143004.0,
      "step": 14103
    },
    {
      "entropy": 0.659027922898531,
      "epoch": 1.2976406116123542,
      "grad_norm": 0.1581011563539505,
      "learning_rate": 5.674855092464808e-05,
      "loss": 0.6456,
      "mean_token_accuracy": 0.8069794699549675,
      "num_tokens": 450175500.0,
      "step": 14104
    },
    {
      "entropy": 0.6713562868535519,
      "epoch": 1.2977326179837954,
      "grad_norm": 0.15810315310955048,
      "learning_rate": 5.6745484098506466e-05,
      "loss": 0.6605,
      "mean_token_accuracy": 0.8020285293459892,
      "num_tokens": 450207836.0,
      "step": 14105
    },
    {
      "entropy": 0.6716575790196657,
      "epoch": 1.2978246243552367,
      "grad_norm": 0.16797763109207153,
      "learning_rate": 5.6742417272364834e-05,
      "loss": 0.6473,
      "mean_token_accuracy": 0.8068930320441723,
      "num_tokens": 450239547.0,
      "step": 14106
    },
    {
      "entropy": 0.5737167121842504,
      "epoch": 1.2979166307266778,
      "grad_norm": 0.16067688167095184,
      "learning_rate": 5.67393504462232e-05,
      "loss": 0.5597,
      "mean_token_accuracy": 0.8258233927190304,
      "num_tokens": 450272072.0,
      "step": 14107
    },
    {
      "entropy": 0.6685552410781384,
      "epoch": 1.298008637098119,
      "grad_norm": 0.16255401074886322,
      "learning_rate": 5.6736283620081576e-05,
      "loss": 0.6449,
      "mean_token_accuracy": 0.8020008206367493,
      "num_tokens": 450303660.0,
      "step": 14108
    },
    {
      "entropy": 0.5917443912476301,
      "epoch": 1.2981006434695603,
      "grad_norm": 0.15930522978305817,
      "learning_rate": 5.673321679393996e-05,
      "loss": 0.5885,
      "mean_token_accuracy": 0.8170139044523239,
      "num_tokens": 450336199.0,
      "step": 14109
    },
    {
      "entropy": 0.7293838681653142,
      "epoch": 1.2981926498410015,
      "grad_norm": 0.16745179891586304,
      "learning_rate": 5.673014996779833e-05,
      "loss": 0.729,
      "mean_token_accuracy": 0.7815571911633015,
      "num_tokens": 450368327.0,
      "step": 14110
    },
    {
      "entropy": 0.5730354432016611,
      "epoch": 1.2982846562124428,
      "grad_norm": 0.155316561460495,
      "learning_rate": 5.67270831416567e-05,
      "loss": 0.5699,
      "mean_token_accuracy": 0.8260693028569221,
      "num_tokens": 450400390.0,
      "step": 14111
    },
    {
      "entropy": 0.6720197256654501,
      "epoch": 1.298376662583884,
      "grad_norm": 0.16194011270999908,
      "learning_rate": 5.6724016315515075e-05,
      "loss": 0.6681,
      "mean_token_accuracy": 0.7965727970004082,
      "num_tokens": 450432021.0,
      "step": 14112
    },
    {
      "entropy": 0.7417417764663696,
      "epoch": 1.298468668955325,
      "grad_norm": 0.1648610234260559,
      "learning_rate": 5.6720949489373456e-05,
      "loss": 0.7196,
      "mean_token_accuracy": 0.7855866998434067,
      "num_tokens": 450463736.0,
      "step": 14113
    },
    {
      "entropy": 0.6305225305259228,
      "epoch": 1.2985606753267664,
      "grad_norm": 0.1588418334722519,
      "learning_rate": 5.6717882663231824e-05,
      "loss": 0.6153,
      "mean_token_accuracy": 0.8079644553363323,
      "num_tokens": 450495277.0,
      "step": 14114
    },
    {
      "entropy": 0.5484697264619172,
      "epoch": 1.2986526816982076,
      "grad_norm": 0.14936192333698273,
      "learning_rate": 5.67148158370902e-05,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8376250825822353,
      "num_tokens": 450527801.0,
      "step": 14115
    },
    {
      "entropy": 0.641801668331027,
      "epoch": 1.298744688069649,
      "grad_norm": 0.16357947885990143,
      "learning_rate": 5.671174901094857e-05,
      "loss": 0.63,
      "mean_token_accuracy": 0.8099037297070026,
      "num_tokens": 450559922.0,
      "step": 14116
    },
    {
      "entropy": 0.6723351655527949,
      "epoch": 1.29883669444109,
      "grad_norm": 0.16038425266742706,
      "learning_rate": 5.670868218480695e-05,
      "loss": 0.6502,
      "mean_token_accuracy": 0.802112452685833,
      "num_tokens": 450591023.0,
      "step": 14117
    },
    {
      "entropy": 0.6279600095003843,
      "epoch": 1.2989287008125312,
      "grad_norm": 0.16330771148204803,
      "learning_rate": 5.670561535866532e-05,
      "loss": 0.6174,
      "mean_token_accuracy": 0.8126541115343571,
      "num_tokens": 450623505.0,
      "step": 14118
    },
    {
      "entropy": 0.6868522260338068,
      "epoch": 1.2990207071839726,
      "grad_norm": 0.16103965044021606,
      "learning_rate": 5.670254853252369e-05,
      "loss": 0.6703,
      "mean_token_accuracy": 0.7960306294262409,
      "num_tokens": 450656265.0,
      "step": 14119
    },
    {
      "entropy": 0.6725050881505013,
      "epoch": 1.2991127135554137,
      "grad_norm": 0.16584442555904388,
      "learning_rate": 5.6699481706382065e-05,
      "loss": 0.6656,
      "mean_token_accuracy": 0.8000534810125828,
      "num_tokens": 450687463.0,
      "step": 14120
    },
    {
      "entropy": 0.6818377617746592,
      "epoch": 1.299204719926855,
      "grad_norm": 0.16263628005981445,
      "learning_rate": 5.669641488024045e-05,
      "loss": 0.6839,
      "mean_token_accuracy": 0.7929953783750534,
      "num_tokens": 450719488.0,
      "step": 14121
    },
    {
      "entropy": 0.6108419522643089,
      "epoch": 1.2992967262982962,
      "grad_norm": 0.1586856245994568,
      "learning_rate": 5.6693348054098815e-05,
      "loss": 0.6023,
      "mean_token_accuracy": 0.8143168538808823,
      "num_tokens": 450751166.0,
      "step": 14122
    },
    {
      "entropy": 0.5699164802208543,
      "epoch": 1.2993887326697373,
      "grad_norm": 0.16610506176948547,
      "learning_rate": 5.669028122795719e-05,
      "loss": 0.534,
      "mean_token_accuracy": 0.834653202444315,
      "num_tokens": 450782498.0,
      "step": 14123
    },
    {
      "entropy": 0.6021362021565437,
      "epoch": 1.2994807390411787,
      "grad_norm": 0.16147759556770325,
      "learning_rate": 5.668721440181557e-05,
      "loss": 0.5999,
      "mean_token_accuracy": 0.8167499341070652,
      "num_tokens": 450814484.0,
      "step": 14124
    },
    {
      "entropy": 0.5805344320833683,
      "epoch": 1.2995727454126198,
      "grad_norm": 0.15124104917049408,
      "learning_rate": 5.668414757567394e-05,
      "loss": 0.5677,
      "mean_token_accuracy": 0.8223090544342995,
      "num_tokens": 450846274.0,
      "step": 14125
    },
    {
      "entropy": 0.7293174769729376,
      "epoch": 1.2996647517840612,
      "grad_norm": 0.18005427718162537,
      "learning_rate": 5.668108074953231e-05,
      "loss": 0.7267,
      "mean_token_accuracy": 0.7782646641135216,
      "num_tokens": 450878120.0,
      "step": 14126
    },
    {
      "entropy": 0.5889028385281563,
      "epoch": 1.2997567581555023,
      "grad_norm": 0.15055055916309357,
      "learning_rate": 5.667801392339068e-05,
      "loss": 0.5816,
      "mean_token_accuracy": 0.8196426630020142,
      "num_tokens": 450910726.0,
      "step": 14127
    },
    {
      "entropy": 0.5845724176615477,
      "epoch": 1.2998487645269434,
      "grad_norm": 0.16501590609550476,
      "learning_rate": 5.667494709724906e-05,
      "loss": 0.5795,
      "mean_token_accuracy": 0.8225745260715485,
      "num_tokens": 450943135.0,
      "step": 14128
    },
    {
      "entropy": 0.6190129807218909,
      "epoch": 1.2999407708983848,
      "grad_norm": 0.1634722352027893,
      "learning_rate": 5.667188027110744e-05,
      "loss": 0.614,
      "mean_token_accuracy": 0.8141670152544975,
      "num_tokens": 450975269.0,
      "step": 14129
    },
    {
      "entropy": 0.6195561047643423,
      "epoch": 1.300032777269826,
      "grad_norm": 0.15705546736717224,
      "learning_rate": 5.6668813444965805e-05,
      "loss": 0.6168,
      "mean_token_accuracy": 0.8083833828568459,
      "num_tokens": 451006849.0,
      "step": 14130
    },
    {
      "entropy": 0.6725555267184973,
      "epoch": 1.3001247836412673,
      "grad_norm": 0.15332017838954926,
      "learning_rate": 5.666574661882418e-05,
      "loss": 0.6531,
      "mean_token_accuracy": 0.8008037358522415,
      "num_tokens": 451039201.0,
      "step": 14131
    },
    {
      "entropy": 0.6937914062291384,
      "epoch": 1.3002167900127084,
      "grad_norm": 0.16039317846298218,
      "learning_rate": 5.666267979268256e-05,
      "loss": 0.6819,
      "mean_token_accuracy": 0.7947497516870499,
      "num_tokens": 451071184.0,
      "step": 14132
    },
    {
      "entropy": 0.4841713053174317,
      "epoch": 1.3003087963841495,
      "grad_norm": 0.1450086534023285,
      "learning_rate": 5.665961296654093e-05,
      "loss": 0.4629,
      "mean_token_accuracy": 0.8559462837874889,
      "num_tokens": 451103423.0,
      "step": 14133
    },
    {
      "entropy": 0.5883592739701271,
      "epoch": 1.300400802755591,
      "grad_norm": 0.15572218596935272,
      "learning_rate": 5.6656546140399304e-05,
      "loss": 0.5705,
      "mean_token_accuracy": 0.823624424636364,
      "num_tokens": 451135439.0,
      "step": 14134
    },
    {
      "entropy": 0.628103194758296,
      "epoch": 1.300492809127032,
      "grad_norm": 0.16705146431922913,
      "learning_rate": 5.665347931425767e-05,
      "loss": 0.6127,
      "mean_token_accuracy": 0.8115005381405354,
      "num_tokens": 451166699.0,
      "step": 14135
    },
    {
      "entropy": 0.7239241283386946,
      "epoch": 1.3005848154984734,
      "grad_norm": 0.16525070369243622,
      "learning_rate": 5.665041248811605e-05,
      "loss": 0.708,
      "mean_token_accuracy": 0.782979927957058,
      "num_tokens": 451198831.0,
      "step": 14136
    },
    {
      "entropy": 0.7000889088958502,
      "epoch": 1.3006768218699145,
      "grad_norm": 0.1653728187084198,
      "learning_rate": 5.664734566197443e-05,
      "loss": 0.6931,
      "mean_token_accuracy": 0.7938465178012848,
      "num_tokens": 451231058.0,
      "step": 14137
    },
    {
      "entropy": 0.662241006270051,
      "epoch": 1.3007688282413556,
      "grad_norm": 0.16048675775527954,
      "learning_rate": 5.6644278835832796e-05,
      "loss": 0.6538,
      "mean_token_accuracy": 0.7962746843695641,
      "num_tokens": 451262175.0,
      "step": 14138
    },
    {
      "entropy": 0.58027508482337,
      "epoch": 1.300860834612797,
      "grad_norm": 0.15989738702774048,
      "learning_rate": 5.664121200969117e-05,
      "loss": 0.5587,
      "mean_token_accuracy": 0.8284625671803951,
      "num_tokens": 451293795.0,
      "step": 14139
    },
    {
      "entropy": 0.5825416035950184,
      "epoch": 1.3009528409842381,
      "grad_norm": 0.1585339456796646,
      "learning_rate": 5.663814518354955e-05,
      "loss": 0.5705,
      "mean_token_accuracy": 0.8245270885527134,
      "num_tokens": 451326194.0,
      "step": 14140
    },
    {
      "entropy": 0.5627476722002029,
      "epoch": 1.3010448473556795,
      "grad_norm": 0.14990749955177307,
      "learning_rate": 5.663507835740792e-05,
      "loss": 0.5353,
      "mean_token_accuracy": 0.8329578451812267,
      "num_tokens": 451358019.0,
      "step": 14141
    },
    {
      "entropy": 0.5887848027050495,
      "epoch": 1.3011368537271206,
      "grad_norm": 0.16474469006061554,
      "learning_rate": 5.6632011531266294e-05,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8211576081812382,
      "num_tokens": 451389606.0,
      "step": 14142
    },
    {
      "entropy": 0.6833654511719942,
      "epoch": 1.3012288600985618,
      "grad_norm": 0.16149884462356567,
      "learning_rate": 5.662894470512466e-05,
      "loss": 0.6626,
      "mean_token_accuracy": 0.7976226881146431,
      "num_tokens": 451421262.0,
      "step": 14143
    },
    {
      "entropy": 0.7029880797490478,
      "epoch": 1.3013208664700031,
      "grad_norm": 0.16169105470180511,
      "learning_rate": 5.662587787898305e-05,
      "loss": 0.7002,
      "mean_token_accuracy": 0.7868851833045483,
      "num_tokens": 451452947.0,
      "step": 14144
    },
    {
      "entropy": 0.5861524529755116,
      "epoch": 1.3014128728414442,
      "grad_norm": 0.15630562603473663,
      "learning_rate": 5.662281105284142e-05,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8214069828391075,
      "num_tokens": 451484450.0,
      "step": 14145
    },
    {
      "entropy": 0.6532900333404541,
      "epoch": 1.3015048792128856,
      "grad_norm": 0.16896098852157593,
      "learning_rate": 5.6619744226699786e-05,
      "loss": 0.6483,
      "mean_token_accuracy": 0.8023543134331703,
      "num_tokens": 451515614.0,
      "step": 14146
    },
    {
      "entropy": 0.703533586114645,
      "epoch": 1.3015968855843267,
      "grad_norm": 0.1670503169298172,
      "learning_rate": 5.661667740055816e-05,
      "loss": 0.6912,
      "mean_token_accuracy": 0.7870315387845039,
      "num_tokens": 451546600.0,
      "step": 14147
    },
    {
      "entropy": 0.6408507861196995,
      "epoch": 1.3016888919557679,
      "grad_norm": 0.16626548767089844,
      "learning_rate": 5.661361057441654e-05,
      "loss": 0.6293,
      "mean_token_accuracy": 0.8080746792256832,
      "num_tokens": 451577869.0,
      "step": 14148
    },
    {
      "entropy": 0.6209717988967896,
      "epoch": 1.3017808983272092,
      "grad_norm": 0.1574353575706482,
      "learning_rate": 5.661054374827492e-05,
      "loss": 0.6215,
      "mean_token_accuracy": 0.8100657239556313,
      "num_tokens": 451610276.0,
      "step": 14149
    },
    {
      "entropy": 0.6610043877735734,
      "epoch": 1.3018729046986504,
      "grad_norm": 0.16134098172187805,
      "learning_rate": 5.6607476922133284e-05,
      "loss": 0.6579,
      "mean_token_accuracy": 0.8043201267719269,
      "num_tokens": 451642728.0,
      "step": 14150
    },
    {
      "entropy": 0.5013809725642204,
      "epoch": 1.3019649110700917,
      "grad_norm": 0.14321854710578918,
      "learning_rate": 5.660441009599165e-05,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8481438532471657,
      "num_tokens": 451674135.0,
      "step": 14151
    },
    {
      "entropy": 0.6978876329958439,
      "epoch": 1.3020569174415328,
      "grad_norm": 0.15780675411224365,
      "learning_rate": 5.660134326985004e-05,
      "loss": 0.6965,
      "mean_token_accuracy": 0.789735559374094,
      "num_tokens": 451706469.0,
      "step": 14152
    },
    {
      "entropy": 0.6831526942551136,
      "epoch": 1.302148923812974,
      "grad_norm": 0.16468673944473267,
      "learning_rate": 5.659827644370841e-05,
      "loss": 0.6774,
      "mean_token_accuracy": 0.7946798838675022,
      "num_tokens": 451738080.0,
      "step": 14153
    },
    {
      "entropy": 0.6212801923975348,
      "epoch": 1.3022409301844153,
      "grad_norm": 0.1628989279270172,
      "learning_rate": 5.659520961756678e-05,
      "loss": 0.6101,
      "mean_token_accuracy": 0.8164263516664505,
      "num_tokens": 451770226.0,
      "step": 14154
    },
    {
      "entropy": 0.5694291898980737,
      "epoch": 1.3023329365558565,
      "grad_norm": 0.1523875892162323,
      "learning_rate": 5.6592142791425165e-05,
      "loss": 0.5481,
      "mean_token_accuracy": 0.8275193721055984,
      "num_tokens": 451801100.0,
      "step": 14155
    },
    {
      "entropy": 0.535117874853313,
      "epoch": 1.3024249429272978,
      "grad_norm": 0.15155485272407532,
      "learning_rate": 5.658907596528353e-05,
      "loss": 0.5119,
      "mean_token_accuracy": 0.8407567515969276,
      "num_tokens": 451832452.0,
      "step": 14156
    },
    {
      "entropy": 0.556527242064476,
      "epoch": 1.302516949298739,
      "grad_norm": 0.1486176699399948,
      "learning_rate": 5.658600913914191e-05,
      "loss": 0.5438,
      "mean_token_accuracy": 0.8327492251992226,
      "num_tokens": 451864825.0,
      "step": 14157
    },
    {
      "entropy": 0.6735260314308107,
      "epoch": 1.30260895567018,
      "grad_norm": 0.16119442880153656,
      "learning_rate": 5.6582942313000275e-05,
      "loss": 0.647,
      "mean_token_accuracy": 0.8037379197776318,
      "num_tokens": 451896978.0,
      "step": 14158
    },
    {
      "entropy": 0.6420170795172453,
      "epoch": 1.3027009620416214,
      "grad_norm": 0.15974125266075134,
      "learning_rate": 5.6579875486858656e-05,
      "loss": 0.6274,
      "mean_token_accuracy": 0.8098234608769417,
      "num_tokens": 451928311.0,
      "step": 14159
    },
    {
      "entropy": 0.6258079884573817,
      "epoch": 1.3027929684130626,
      "grad_norm": 0.16415266692638397,
      "learning_rate": 5.657680866071703e-05,
      "loss": 0.6048,
      "mean_token_accuracy": 0.8154684454202652,
      "num_tokens": 451960530.0,
      "step": 14160
    },
    {
      "entropy": 0.6628070212900639,
      "epoch": 1.302884974784504,
      "grad_norm": 0.16879737377166748,
      "learning_rate": 5.65737418345754e-05,
      "loss": 0.6502,
      "mean_token_accuracy": 0.8006644174456596,
      "num_tokens": 451992444.0,
      "step": 14161
    },
    {
      "entropy": 0.6556344516575336,
      "epoch": 1.302976981155945,
      "grad_norm": 0.15669044852256775,
      "learning_rate": 5.6570675008433773e-05,
      "loss": 0.6457,
      "mean_token_accuracy": 0.8016831651329994,
      "num_tokens": 452024728.0,
      "step": 14162
    },
    {
      "entropy": 0.6253615943714976,
      "epoch": 1.3030689875273862,
      "grad_norm": 0.15464112162590027,
      "learning_rate": 5.6567608182292155e-05,
      "loss": 0.6127,
      "mean_token_accuracy": 0.808548666536808,
      "num_tokens": 452056913.0,
      "step": 14163
    },
    {
      "entropy": 0.5492186211049557,
      "epoch": 1.3031609938988276,
      "grad_norm": 0.14868983626365662,
      "learning_rate": 5.656454135615052e-05,
      "loss": 0.5416,
      "mean_token_accuracy": 0.830946259200573,
      "num_tokens": 452089462.0,
      "step": 14164
    },
    {
      "entropy": 0.677087876945734,
      "epoch": 1.3032530002702687,
      "grad_norm": 0.16725902259349823,
      "learning_rate": 5.65614745300089e-05,
      "loss": 0.6848,
      "mean_token_accuracy": 0.7908815704286098,
      "num_tokens": 452120910.0,
      "step": 14165
    },
    {
      "entropy": 0.7726412490010262,
      "epoch": 1.30334500664171,
      "grad_norm": 0.17075291275978088,
      "learning_rate": 5.6558407703867265e-05,
      "loss": 0.7895,
      "mean_token_accuracy": 0.7665707282721996,
      "num_tokens": 452152948.0,
      "step": 14166
    },
    {
      "entropy": 0.6857729209586978,
      "epoch": 1.3034370130131512,
      "grad_norm": 0.17010173201560974,
      "learning_rate": 5.655534087772565e-05,
      "loss": 0.6743,
      "mean_token_accuracy": 0.7998028956353664,
      "num_tokens": 452184876.0,
      "step": 14167
    },
    {
      "entropy": 0.4678496830165386,
      "epoch": 1.3035290193845923,
      "grad_norm": 0.14278192818164825,
      "learning_rate": 5.655227405158402e-05,
      "loss": 0.4366,
      "mean_token_accuracy": 0.8602608889341354,
      "num_tokens": 452216917.0,
      "step": 14168
    },
    {
      "entropy": 0.6749248839914799,
      "epoch": 1.3036210257560337,
      "grad_norm": 0.15994229912757874,
      "learning_rate": 5.654920722544239e-05,
      "loss": 0.6532,
      "mean_token_accuracy": 0.7989485077559948,
      "num_tokens": 452249105.0,
      "step": 14169
    },
    {
      "entropy": 0.5747872367501259,
      "epoch": 1.3037130321274748,
      "grad_norm": 0.15526999533176422,
      "learning_rate": 5.6546140399300764e-05,
      "loss": 0.5781,
      "mean_token_accuracy": 0.8233098089694977,
      "num_tokens": 452280521.0,
      "step": 14170
    },
    {
      "entropy": 0.565888368524611,
      "epoch": 1.3038050384989162,
      "grad_norm": 0.15456268191337585,
      "learning_rate": 5.6543073573159145e-05,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8268180415034294,
      "num_tokens": 452311798.0,
      "step": 14171
    },
    {
      "entropy": 0.4669596664607525,
      "epoch": 1.3038970448703573,
      "grad_norm": 0.16009590029716492,
      "learning_rate": 5.654000674701751e-05,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8587457202374935,
      "num_tokens": 452343878.0,
      "step": 14172
    },
    {
      "entropy": 0.6124655399471521,
      "epoch": 1.3039890512417984,
      "grad_norm": 0.16571716964244843,
      "learning_rate": 5.653693992087589e-05,
      "loss": 0.6018,
      "mean_token_accuracy": 0.8165828734636307,
      "num_tokens": 452375363.0,
      "step": 14173
    },
    {
      "entropy": 0.5967504503205419,
      "epoch": 1.3040810576132398,
      "grad_norm": 0.15122421085834503,
      "learning_rate": 5.6533873094734256e-05,
      "loss": 0.5811,
      "mean_token_accuracy": 0.8210471943020821,
      "num_tokens": 452407711.0,
      "step": 14174
    },
    {
      "entropy": 0.6720668263733387,
      "epoch": 1.304173063984681,
      "grad_norm": 0.16511985659599304,
      "learning_rate": 5.653080626859264e-05,
      "loss": 0.6699,
      "mean_token_accuracy": 0.7966867685317993,
      "num_tokens": 452439787.0,
      "step": 14175
    },
    {
      "entropy": 0.6509403977543116,
      "epoch": 1.3042650703561223,
      "grad_norm": 0.15977199375629425,
      "learning_rate": 5.652773944245101e-05,
      "loss": 0.6402,
      "mean_token_accuracy": 0.8022051118314266,
      "num_tokens": 452471371.0,
      "step": 14176
    },
    {
      "entropy": 0.5676786676049232,
      "epoch": 1.3043570767275634,
      "grad_norm": 0.1637677103281021,
      "learning_rate": 5.652467261630938e-05,
      "loss": 0.5639,
      "mean_token_accuracy": 0.8293512277305126,
      "num_tokens": 452503606.0,
      "step": 14177
    },
    {
      "entropy": 0.6532020596787333,
      "epoch": 1.3044490830990045,
      "grad_norm": 0.1557745337486267,
      "learning_rate": 5.6521605790167754e-05,
      "loss": 0.6195,
      "mean_token_accuracy": 0.8130415193736553,
      "num_tokens": 452535240.0,
      "step": 14178
    },
    {
      "entropy": 0.6321513261646032,
      "epoch": 1.304541089470446,
      "grad_norm": 0.1710381954908371,
      "learning_rate": 5.6518538964026136e-05,
      "loss": 0.6159,
      "mean_token_accuracy": 0.8137585073709488,
      "num_tokens": 452566608.0,
      "step": 14179
    },
    {
      "entropy": 0.5560627086088061,
      "epoch": 1.304633095841887,
      "grad_norm": 0.14683350920677185,
      "learning_rate": 5.6515472137884504e-05,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8331592530012131,
      "num_tokens": 452598751.0,
      "step": 14180
    },
    {
      "entropy": 0.7503755344077945,
      "epoch": 1.3047251022133284,
      "grad_norm": 0.16914047300815582,
      "learning_rate": 5.651240531174288e-05,
      "loss": 0.7352,
      "mean_token_accuracy": 0.7783846110105515,
      "num_tokens": 452630563.0,
      "step": 14181
    },
    {
      "entropy": 0.5051311310380697,
      "epoch": 1.3048171085847695,
      "grad_norm": 0.1503400355577469,
      "learning_rate": 5.6509338485601246e-05,
      "loss": 0.472,
      "mean_token_accuracy": 0.8501496985554695,
      "num_tokens": 452661418.0,
      "step": 14182
    },
    {
      "entropy": 0.5661604646593332,
      "epoch": 1.3049091149562106,
      "grad_norm": 0.15560080111026764,
      "learning_rate": 5.650627165945963e-05,
      "loss": 0.5553,
      "mean_token_accuracy": 0.8306216262280941,
      "num_tokens": 452693772.0,
      "step": 14183
    },
    {
      "entropy": 0.5544636640697718,
      "epoch": 1.305001121327652,
      "grad_norm": 0.14958856999874115,
      "learning_rate": 5.6503204833318e-05,
      "loss": 0.5396,
      "mean_token_accuracy": 0.8322720639407635,
      "num_tokens": 452725522.0,
      "step": 14184
    },
    {
      "entropy": 0.6422067107632756,
      "epoch": 1.3050931276990931,
      "grad_norm": 0.15473192930221558,
      "learning_rate": 5.650013800717637e-05,
      "loss": 0.6322,
      "mean_token_accuracy": 0.8074900470674038,
      "num_tokens": 452758157.0,
      "step": 14185
    },
    {
      "entropy": 0.6548685748130083,
      "epoch": 1.3051851340705345,
      "grad_norm": 0.1675262451171875,
      "learning_rate": 5.649707118103476e-05,
      "loss": 0.6541,
      "mean_token_accuracy": 0.804037194699049,
      "num_tokens": 452789536.0,
      "step": 14186
    },
    {
      "entropy": 0.5317169614136219,
      "epoch": 1.3052771404419756,
      "grad_norm": 0.14818090200424194,
      "learning_rate": 5.6494004354893126e-05,
      "loss": 0.5213,
      "mean_token_accuracy": 0.8370732069015503,
      "num_tokens": 452821891.0,
      "step": 14187
    },
    {
      "entropy": 0.7075565438717604,
      "epoch": 1.3053691468134168,
      "grad_norm": 0.163914754986763,
      "learning_rate": 5.64909375287515e-05,
      "loss": 0.7142,
      "mean_token_accuracy": 0.7839655764400959,
      "num_tokens": 452853856.0,
      "step": 14188
    },
    {
      "entropy": 0.46147291315719485,
      "epoch": 1.3054611531848581,
      "grad_norm": 0.14618393778800964,
      "learning_rate": 5.648787070260987e-05,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8595870807766914,
      "num_tokens": 452886341.0,
      "step": 14189
    },
    {
      "entropy": 0.5367695484310389,
      "epoch": 1.3055531595562992,
      "grad_norm": 0.14136357605457306,
      "learning_rate": 5.648480387646825e-05,
      "loss": 0.5237,
      "mean_token_accuracy": 0.8377616107463837,
      "num_tokens": 452918968.0,
      "step": 14190
    },
    {
      "entropy": 0.617236795835197,
      "epoch": 1.3056451659277406,
      "grad_norm": 0.15540499985218048,
      "learning_rate": 5.6481737050326625e-05,
      "loss": 0.6077,
      "mean_token_accuracy": 0.8134312480688095,
      "num_tokens": 452950632.0,
      "step": 14191
    },
    {
      "entropy": 0.6690814457833767,
      "epoch": 1.3057371722991817,
      "grad_norm": 0.16012518107891083,
      "learning_rate": 5.647867022418499e-05,
      "loss": 0.6528,
      "mean_token_accuracy": 0.7955675236880779,
      "num_tokens": 452983048.0,
      "step": 14192
    },
    {
      "entropy": 0.6013571713119745,
      "epoch": 1.3058291786706229,
      "grad_norm": 0.15393255650997162,
      "learning_rate": 5.647560339804337e-05,
      "loss": 0.5955,
      "mean_token_accuracy": 0.81996975466609,
      "num_tokens": 453015483.0,
      "step": 14193
    },
    {
      "entropy": 0.5652347011491656,
      "epoch": 1.3059211850420642,
      "grad_norm": 0.1506577879190445,
      "learning_rate": 5.647253657190175e-05,
      "loss": 0.5581,
      "mean_token_accuracy": 0.8268205374479294,
      "num_tokens": 453047911.0,
      "step": 14194
    },
    {
      "entropy": 0.642705611884594,
      "epoch": 1.3060131914135054,
      "grad_norm": 0.15598973631858826,
      "learning_rate": 5.6469469745760117e-05,
      "loss": 0.6299,
      "mean_token_accuracy": 0.8049218915402889,
      "num_tokens": 453079297.0,
      "step": 14195
    },
    {
      "entropy": 0.5964294285513461,
      "epoch": 1.3061051977849467,
      "grad_norm": 0.155684694647789,
      "learning_rate": 5.646640291961849e-05,
      "loss": 0.5912,
      "mean_token_accuracy": 0.8193826377391815,
      "num_tokens": 453109882.0,
      "step": 14196
    },
    {
      "entropy": 0.6459777420386672,
      "epoch": 1.3061972041563878,
      "grad_norm": 0.15923330187797546,
      "learning_rate": 5.646333609347686e-05,
      "loss": 0.6447,
      "mean_token_accuracy": 0.8020971976220608,
      "num_tokens": 453141863.0,
      "step": 14197
    },
    {
      "entropy": 0.6359434500336647,
      "epoch": 1.306289210527829,
      "grad_norm": 0.16064627468585968,
      "learning_rate": 5.646026926733524e-05,
      "loss": 0.6305,
      "mean_token_accuracy": 0.8068743757903576,
      "num_tokens": 453173976.0,
      "step": 14198
    },
    {
      "entropy": 0.6606067195534706,
      "epoch": 1.3063812168992703,
      "grad_norm": 0.16578233242034912,
      "learning_rate": 5.6457202441193615e-05,
      "loss": 0.648,
      "mean_token_accuracy": 0.7973568327724934,
      "num_tokens": 453206290.0,
      "step": 14199
    },
    {
      "entropy": 0.6388007197529078,
      "epoch": 1.3064732232707115,
      "grad_norm": 0.15871314704418182,
      "learning_rate": 5.645413561505198e-05,
      "loss": 0.6423,
      "mean_token_accuracy": 0.8056211732327938,
      "num_tokens": 453238793.0,
      "step": 14200
    },
    {
      "entropy": 0.6534602884203196,
      "epoch": 1.3065652296421528,
      "grad_norm": 0.17106540501117706,
      "learning_rate": 5.645106878891036e-05,
      "loss": 0.639,
      "mean_token_accuracy": 0.8036491572856903,
      "num_tokens": 453269641.0,
      "step": 14201
    },
    {
      "entropy": 0.5789098292589188,
      "epoch": 1.306657236013594,
      "grad_norm": 0.15193891525268555,
      "learning_rate": 5.644800196276874e-05,
      "loss": 0.5654,
      "mean_token_accuracy": 0.8238321207463741,
      "num_tokens": 453301523.0,
      "step": 14202
    },
    {
      "entropy": 0.5703686121851206,
      "epoch": 1.306749242385035,
      "grad_norm": 0.16647589206695557,
      "learning_rate": 5.644493513662711e-05,
      "loss": 0.5627,
      "mean_token_accuracy": 0.8304155170917511,
      "num_tokens": 453332993.0,
      "step": 14203
    },
    {
      "entropy": 0.4248194666579366,
      "epoch": 1.3068412487564764,
      "grad_norm": 0.14863815903663635,
      "learning_rate": 5.644186831048548e-05,
      "loss": 0.4123,
      "mean_token_accuracy": 0.870030052959919,
      "num_tokens": 453364774.0,
      "step": 14204
    },
    {
      "entropy": 0.574378527700901,
      "epoch": 1.3069332551279176,
      "grad_norm": 0.1536719799041748,
      "learning_rate": 5.643880148434385e-05,
      "loss": 0.5619,
      "mean_token_accuracy": 0.8240927532315254,
      "num_tokens": 453397426.0,
      "step": 14205
    },
    {
      "entropy": 0.7497223988175392,
      "epoch": 1.307025261499359,
      "grad_norm": 0.1734118014574051,
      "learning_rate": 5.643573465820223e-05,
      "loss": 0.7507,
      "mean_token_accuracy": 0.7759820595383644,
      "num_tokens": 453429509.0,
      "step": 14206
    },
    {
      "entropy": 0.7639605402946472,
      "epoch": 1.3071172678708,
      "grad_norm": 0.16609257459640503,
      "learning_rate": 5.6432667832060606e-05,
      "loss": 0.7435,
      "mean_token_accuracy": 0.7773479335010052,
      "num_tokens": 453461393.0,
      "step": 14207
    },
    {
      "entropy": 0.6672632470726967,
      "epoch": 1.3072092742422412,
      "grad_norm": 0.16230686008930206,
      "learning_rate": 5.6429601005918973e-05,
      "loss": 0.6619,
      "mean_token_accuracy": 0.7997845336794853,
      "num_tokens": 453493977.0,
      "step": 14208
    },
    {
      "entropy": 0.5912244226783514,
      "epoch": 1.3073012806136826,
      "grad_norm": 0.1539868861436844,
      "learning_rate": 5.642653417977735e-05,
      "loss": 0.5748,
      "mean_token_accuracy": 0.8211885541677475,
      "num_tokens": 453525420.0,
      "step": 14209
    },
    {
      "entropy": 0.5713470987975597,
      "epoch": 1.3073932869851237,
      "grad_norm": 0.15395788848400116,
      "learning_rate": 5.642346735363573e-05,
      "loss": 0.549,
      "mean_token_accuracy": 0.8303076960146427,
      "num_tokens": 453557056.0,
      "step": 14210
    },
    {
      "entropy": 0.6017084661871195,
      "epoch": 1.307485293356565,
      "grad_norm": 0.1545892357826233,
      "learning_rate": 5.64204005274941e-05,
      "loss": 0.5885,
      "mean_token_accuracy": 0.8188052400946617,
      "num_tokens": 453588504.0,
      "step": 14211
    },
    {
      "entropy": 0.5432381951250136,
      "epoch": 1.3075772997280062,
      "grad_norm": 0.15406270325183868,
      "learning_rate": 5.641733370135247e-05,
      "loss": 0.5274,
      "mean_token_accuracy": 0.836008582264185,
      "num_tokens": 453620389.0,
      "step": 14212
    },
    {
      "entropy": 0.688975665718317,
      "epoch": 1.3076693060994473,
      "grad_norm": 0.16378186643123627,
      "learning_rate": 5.641426687521084e-05,
      "loss": 0.6809,
      "mean_token_accuracy": 0.7963170371949673,
      "num_tokens": 453652633.0,
      "step": 14213
    },
    {
      "entropy": 0.6382385157048702,
      "epoch": 1.3077613124708887,
      "grad_norm": 0.15718354284763336,
      "learning_rate": 5.641120004906922e-05,
      "loss": 0.625,
      "mean_token_accuracy": 0.8086726553738117,
      "num_tokens": 453684451.0,
      "step": 14214
    },
    {
      "entropy": 0.5251387730240822,
      "epoch": 1.3078533188423298,
      "grad_norm": 0.164112851023674,
      "learning_rate": 5.6408133222927596e-05,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8395457230508327,
      "num_tokens": 453716158.0,
      "step": 14215
    },
    {
      "entropy": 0.6618406779598445,
      "epoch": 1.3079453252137712,
      "grad_norm": 0.15705345571041107,
      "learning_rate": 5.6405066396785964e-05,
      "loss": 0.6345,
      "mean_token_accuracy": 0.8060081675648689,
      "num_tokens": 453748284.0,
      "step": 14216
    },
    {
      "entropy": 0.628295049071312,
      "epoch": 1.3080373315852123,
      "grad_norm": 0.1562836468219757,
      "learning_rate": 5.6401999570644345e-05,
      "loss": 0.6196,
      "mean_token_accuracy": 0.8102926760911942,
      "num_tokens": 453780074.0,
      "step": 14217
    },
    {
      "entropy": 0.6792711541056633,
      "epoch": 1.3081293379566534,
      "grad_norm": 0.15783335268497467,
      "learning_rate": 5.639893274450272e-05,
      "loss": 0.6653,
      "mean_token_accuracy": 0.796589408069849,
      "num_tokens": 453811801.0,
      "step": 14218
    },
    {
      "entropy": 0.8093993719667196,
      "epoch": 1.3082213443280948,
      "grad_norm": 0.1723400056362152,
      "learning_rate": 5.639586591836109e-05,
      "loss": 0.8167,
      "mean_token_accuracy": 0.7546344734728336,
      "num_tokens": 453844064.0,
      "step": 14219
    },
    {
      "entropy": 0.6650958508253098,
      "epoch": 1.308313350699536,
      "grad_norm": 0.1567000150680542,
      "learning_rate": 5.639279909221946e-05,
      "loss": 0.6542,
      "mean_token_accuracy": 0.8011481836438179,
      "num_tokens": 453875418.0,
      "step": 14220
    },
    {
      "entropy": 0.5923005044460297,
      "epoch": 1.3084053570709773,
      "grad_norm": 0.15913188457489014,
      "learning_rate": 5.6389732266077844e-05,
      "loss": 0.5873,
      "mean_token_accuracy": 0.8186895586550236,
      "num_tokens": 453907154.0,
      "step": 14221
    },
    {
      "entropy": 0.5857669822871685,
      "epoch": 1.3084973634424184,
      "grad_norm": 0.15656688809394836,
      "learning_rate": 5.638666543993621e-05,
      "loss": 0.5802,
      "mean_token_accuracy": 0.8225698992609978,
      "num_tokens": 453939433.0,
      "step": 14222
    },
    {
      "entropy": 0.644032683223486,
      "epoch": 1.3085893698138595,
      "grad_norm": 0.16518253087997437,
      "learning_rate": 5.6383598613794586e-05,
      "loss": 0.6391,
      "mean_token_accuracy": 0.8003973327577114,
      "num_tokens": 453971240.0,
      "step": 14223
    },
    {
      "entropy": 0.6180855957791209,
      "epoch": 1.308681376185301,
      "grad_norm": 0.1623259335756302,
      "learning_rate": 5.6380531787652954e-05,
      "loss": 0.6002,
      "mean_token_accuracy": 0.8144916445016861,
      "num_tokens": 454002976.0,
      "step": 14224
    },
    {
      "entropy": 0.6932114735245705,
      "epoch": 1.308773382556742,
      "grad_norm": 0.158138245344162,
      "learning_rate": 5.637746496151134e-05,
      "loss": 0.6763,
      "mean_token_accuracy": 0.7931058220565319,
      "num_tokens": 454034994.0,
      "step": 14225
    },
    {
      "entropy": 0.6705208150669932,
      "epoch": 1.3088653889281834,
      "grad_norm": 0.1595105081796646,
      "learning_rate": 5.637439813536971e-05,
      "loss": 0.6589,
      "mean_token_accuracy": 0.7975498996675014,
      "num_tokens": 454067288.0,
      "step": 14226
    },
    {
      "entropy": 0.6687650703825057,
      "epoch": 1.3089573952996245,
      "grad_norm": 0.16126830875873566,
      "learning_rate": 5.637133130922808e-05,
      "loss": 0.6506,
      "mean_token_accuracy": 0.8006312772631645,
      "num_tokens": 454099542.0,
      "step": 14227
    },
    {
      "entropy": 0.6981386430561543,
      "epoch": 1.3090494016710656,
      "grad_norm": 0.15932871401309967,
      "learning_rate": 5.636826448308645e-05,
      "loss": 0.6924,
      "mean_token_accuracy": 0.7925928197801113,
      "num_tokens": 454132310.0,
      "step": 14228
    },
    {
      "entropy": 0.5631336327642202,
      "epoch": 1.309141408042507,
      "grad_norm": 0.15705250203609467,
      "learning_rate": 5.6365197656944834e-05,
      "loss": 0.5517,
      "mean_token_accuracy": 0.827445175498724,
      "num_tokens": 454163639.0,
      "step": 14229
    },
    {
      "entropy": 0.6061041671782732,
      "epoch": 1.3092334144139481,
      "grad_norm": 0.156753271818161,
      "learning_rate": 5.636213083080321e-05,
      "loss": 0.5952,
      "mean_token_accuracy": 0.8158175684511662,
      "num_tokens": 454195060.0,
      "step": 14230
    },
    {
      "entropy": 0.6575558576732874,
      "epoch": 1.3093254207853895,
      "grad_norm": 0.16130410134792328,
      "learning_rate": 5.635906400466158e-05,
      "loss": 0.6441,
      "mean_token_accuracy": 0.8032171502709389,
      "num_tokens": 454226751.0,
      "step": 14231
    },
    {
      "entropy": 0.6544948085211217,
      "epoch": 1.3094174271568306,
      "grad_norm": 0.1630346179008484,
      "learning_rate": 5.635599717851995e-05,
      "loss": 0.6341,
      "mean_token_accuracy": 0.8063949383795261,
      "num_tokens": 454258660.0,
      "step": 14232
    },
    {
      "entropy": 0.5851624049246311,
      "epoch": 1.3095094335282718,
      "grad_norm": 0.1542077660560608,
      "learning_rate": 5.635293035237833e-05,
      "loss": 0.5706,
      "mean_token_accuracy": 0.8247383870184422,
      "num_tokens": 454290681.0,
      "step": 14233
    },
    {
      "entropy": 0.5482022482901812,
      "epoch": 1.3096014398997131,
      "grad_norm": 0.16690626740455627,
      "learning_rate": 5.63498635262367e-05,
      "loss": 0.5325,
      "mean_token_accuracy": 0.8349030166864395,
      "num_tokens": 454322329.0,
      "step": 14234
    },
    {
      "entropy": 0.6855246936902404,
      "epoch": 1.3096934462711542,
      "grad_norm": 0.1683550477027893,
      "learning_rate": 5.6346796700095075e-05,
      "loss": 0.6608,
      "mean_token_accuracy": 0.797224897891283,
      "num_tokens": 454353886.0,
      "step": 14235
    },
    {
      "entropy": 0.5992552172392607,
      "epoch": 1.3097854526425956,
      "grad_norm": 0.16218139231204987,
      "learning_rate": 5.634372987395344e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.818369422107935,
      "num_tokens": 454386374.0,
      "step": 14236
    },
    {
      "entropy": 0.6698393588885665,
      "epoch": 1.3098774590140367,
      "grad_norm": 0.1561673879623413,
      "learning_rate": 5.6340663047811825e-05,
      "loss": 0.6543,
      "mean_token_accuracy": 0.799030389636755,
      "num_tokens": 454418524.0,
      "step": 14237
    },
    {
      "entropy": 0.5537024871446192,
      "epoch": 1.3099694653854779,
      "grad_norm": 0.15026244521141052,
      "learning_rate": 5.63375962216702e-05,
      "loss": 0.534,
      "mean_token_accuracy": 0.8293924778699875,
      "num_tokens": 454450119.0,
      "step": 14238
    },
    {
      "entropy": 0.655488520860672,
      "epoch": 1.3100614717569192,
      "grad_norm": 0.16205711662769318,
      "learning_rate": 5.633452939552857e-05,
      "loss": 0.6619,
      "mean_token_accuracy": 0.7991825453937054,
      "num_tokens": 454481653.0,
      "step": 14239
    },
    {
      "entropy": 0.6538149258121848,
      "epoch": 1.3101534781283604,
      "grad_norm": 0.15605370700359344,
      "learning_rate": 5.633146256938694e-05,
      "loss": 0.6345,
      "mean_token_accuracy": 0.8050349615514278,
      "num_tokens": 454512932.0,
      "step": 14240
    },
    {
      "entropy": 0.6820078529417515,
      "epoch": 1.3102454844998017,
      "grad_norm": 0.15810523927211761,
      "learning_rate": 5.632839574324532e-05,
      "loss": 0.6695,
      "mean_token_accuracy": 0.795282494276762,
      "num_tokens": 454545079.0,
      "step": 14241
    },
    {
      "entropy": 0.6858618836849928,
      "epoch": 1.3103374908712428,
      "grad_norm": 0.16821661591529846,
      "learning_rate": 5.632532891710369e-05,
      "loss": 0.6804,
      "mean_token_accuracy": 0.7939349748194218,
      "num_tokens": 454576600.0,
      "step": 14242
    },
    {
      "entropy": 0.7309833075851202,
      "epoch": 1.310429497242684,
      "grad_norm": 0.16468605399131775,
      "learning_rate": 5.6322262090962066e-05,
      "loss": 0.7296,
      "mean_token_accuracy": 0.7792327255010605,
      "num_tokens": 454608258.0,
      "step": 14243
    },
    {
      "entropy": 0.5639790454879403,
      "epoch": 1.3105215036141253,
      "grad_norm": 0.14903543889522552,
      "learning_rate": 5.6319195264820434e-05,
      "loss": 0.5614,
      "mean_token_accuracy": 0.8260557986795902,
      "num_tokens": 454640209.0,
      "step": 14244
    },
    {
      "entropy": 0.6907211635261774,
      "epoch": 1.3106135099855665,
      "grad_norm": 0.1641438603401184,
      "learning_rate": 5.6316128438678815e-05,
      "loss": 0.6756,
      "mean_token_accuracy": 0.7909909524023533,
      "num_tokens": 454672020.0,
      "step": 14245
    },
    {
      "entropy": 0.6239675944671035,
      "epoch": 1.3107055163570078,
      "grad_norm": 0.1567392200231552,
      "learning_rate": 5.631306161253719e-05,
      "loss": 0.6149,
      "mean_token_accuracy": 0.8127138987183571,
      "num_tokens": 454703970.0,
      "step": 14246
    },
    {
      "entropy": 0.5533350892364979,
      "epoch": 1.310797522728449,
      "grad_norm": 0.1497531235218048,
      "learning_rate": 5.630999478639556e-05,
      "loss": 0.5491,
      "mean_token_accuracy": 0.8270482607185841,
      "num_tokens": 454736541.0,
      "step": 14247
    },
    {
      "entropy": 0.6456318376585841,
      "epoch": 1.31088952909989,
      "grad_norm": 0.15386641025543213,
      "learning_rate": 5.630692796025394e-05,
      "loss": 0.6212,
      "mean_token_accuracy": 0.8078200146555901,
      "num_tokens": 454768161.0,
      "step": 14248
    },
    {
      "entropy": 0.6246470990590751,
      "epoch": 1.3109815354713314,
      "grad_norm": 0.16012372076511383,
      "learning_rate": 5.6303861134112314e-05,
      "loss": 0.6118,
      "mean_token_accuracy": 0.8133852854371071,
      "num_tokens": 454799413.0,
      "step": 14249
    },
    {
      "entropy": 0.6808357425034046,
      "epoch": 1.3110735418427726,
      "grad_norm": 0.1606132984161377,
      "learning_rate": 5.630079430797068e-05,
      "loss": 0.678,
      "mean_token_accuracy": 0.7916521318256855,
      "num_tokens": 454831651.0,
      "step": 14250
    },
    {
      "entropy": 0.7222789525985718,
      "epoch": 1.311165548214214,
      "grad_norm": 0.1619492471218109,
      "learning_rate": 5.6297727481829056e-05,
      "loss": 0.7178,
      "mean_token_accuracy": 0.7837371863424778,
      "num_tokens": 454864411.0,
      "step": 14251
    },
    {
      "entropy": 0.6852001175284386,
      "epoch": 1.311257554585655,
      "grad_norm": 0.16675494611263275,
      "learning_rate": 5.629466065568744e-05,
      "loss": 0.6859,
      "mean_token_accuracy": 0.7951492555439472,
      "num_tokens": 454896008.0,
      "step": 14252
    },
    {
      "entropy": 0.7531926706433296,
      "epoch": 1.3113495609570962,
      "grad_norm": 0.17276442050933838,
      "learning_rate": 5.6291593829545805e-05,
      "loss": 0.742,
      "mean_token_accuracy": 0.7743937112390995,
      "num_tokens": 454927326.0,
      "step": 14253
    },
    {
      "entropy": 0.662922540679574,
      "epoch": 1.3114415673285376,
      "grad_norm": 0.16223566234111786,
      "learning_rate": 5.628852700340418e-05,
      "loss": 0.6496,
      "mean_token_accuracy": 0.7998629510402679,
      "num_tokens": 454958609.0,
      "step": 14254
    },
    {
      "entropy": 0.6541419364511967,
      "epoch": 1.3115335736999787,
      "grad_norm": 0.15650339424610138,
      "learning_rate": 5.628546017726255e-05,
      "loss": 0.642,
      "mean_token_accuracy": 0.8058395199477673,
      "num_tokens": 454990901.0,
      "step": 14255
    },
    {
      "entropy": 0.6325326543301344,
      "epoch": 1.31162558007142,
      "grad_norm": 0.15754574537277222,
      "learning_rate": 5.628239335112093e-05,
      "loss": 0.6073,
      "mean_token_accuracy": 0.8114927373826504,
      "num_tokens": 455022891.0,
      "step": 14256
    },
    {
      "entropy": 0.650550689548254,
      "epoch": 1.3117175864428612,
      "grad_norm": 0.16125385463237762,
      "learning_rate": 5.6279326524979304e-05,
      "loss": 0.6487,
      "mean_token_accuracy": 0.8056738562881947,
      "num_tokens": 455055192.0,
      "step": 14257
    },
    {
      "entropy": 0.6818867921829224,
      "epoch": 1.3118095928143023,
      "grad_norm": 0.16455182433128357,
      "learning_rate": 5.627625969883767e-05,
      "loss": 0.6734,
      "mean_token_accuracy": 0.7960145212709904,
      "num_tokens": 455087011.0,
      "step": 14258
    },
    {
      "entropy": 0.7285914421081543,
      "epoch": 1.3119015991857437,
      "grad_norm": 0.16498339176177979,
      "learning_rate": 5.6273192872696047e-05,
      "loss": 0.7155,
      "mean_token_accuracy": 0.781865268945694,
      "num_tokens": 455119248.0,
      "step": 14259
    },
    {
      "entropy": 0.5373873533681035,
      "epoch": 1.3119936055571848,
      "grad_norm": 0.1647735983133316,
      "learning_rate": 5.627012604655443e-05,
      "loss": 0.5242,
      "mean_token_accuracy": 0.8384292460978031,
      "num_tokens": 455151026.0,
      "step": 14260
    },
    {
      "entropy": 0.5642903167754412,
      "epoch": 1.3120856119286262,
      "grad_norm": 0.1652287244796753,
      "learning_rate": 5.6267059220412796e-05,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8272133506834507,
      "num_tokens": 455182820.0,
      "step": 14261
    },
    {
      "entropy": 0.6946740560233593,
      "epoch": 1.3121776183000673,
      "grad_norm": 0.16896823048591614,
      "learning_rate": 5.626399239427117e-05,
      "loss": 0.6904,
      "mean_token_accuracy": 0.7879503034055233,
      "num_tokens": 455214521.0,
      "step": 14262
    },
    {
      "entropy": 0.5277412859722972,
      "epoch": 1.3122696246715084,
      "grad_norm": 0.14966405928134918,
      "learning_rate": 5.626092556812954e-05,
      "loss": 0.5125,
      "mean_token_accuracy": 0.8422884047031403,
      "num_tokens": 455247261.0,
      "step": 14263
    },
    {
      "entropy": 0.689933548681438,
      "epoch": 1.3123616310429498,
      "grad_norm": 0.1678270399570465,
      "learning_rate": 5.6257858741987927e-05,
      "loss": 0.6834,
      "mean_token_accuracy": 0.7930788584053516,
      "num_tokens": 455278926.0,
      "step": 14264
    },
    {
      "entropy": 0.6216852022334933,
      "epoch": 1.312453637414391,
      "grad_norm": 0.15657176077365875,
      "learning_rate": 5.6254791915846294e-05,
      "loss": 0.6014,
      "mean_token_accuracy": 0.8195048086345196,
      "num_tokens": 455310984.0,
      "step": 14265
    },
    {
      "entropy": 0.6229301448911428,
      "epoch": 1.3125456437858323,
      "grad_norm": 0.15599793195724487,
      "learning_rate": 5.625172508970466e-05,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8135586827993393,
      "num_tokens": 455343744.0,
      "step": 14266
    },
    {
      "entropy": 0.7245909236371517,
      "epoch": 1.3126376501572734,
      "grad_norm": 0.1562384068965912,
      "learning_rate": 5.624865826356304e-05,
      "loss": 0.7127,
      "mean_token_accuracy": 0.7834360413253307,
      "num_tokens": 455376238.0,
      "step": 14267
    },
    {
      "entropy": 0.6181972781196237,
      "epoch": 1.3127296565287145,
      "grad_norm": 0.15811923146247864,
      "learning_rate": 5.624559143742142e-05,
      "loss": 0.5873,
      "mean_token_accuracy": 0.8180348873138428,
      "num_tokens": 455408283.0,
      "step": 14268
    },
    {
      "entropy": 0.638216448482126,
      "epoch": 1.312821662900156,
      "grad_norm": 0.15359438955783844,
      "learning_rate": 5.624252461127979e-05,
      "loss": 0.611,
      "mean_token_accuracy": 0.8103735409677029,
      "num_tokens": 455440005.0,
      "step": 14269
    },
    {
      "entropy": 0.7065864484757185,
      "epoch": 1.312913669271597,
      "grad_norm": 0.16181014478206635,
      "learning_rate": 5.623945778513816e-05,
      "loss": 0.6787,
      "mean_token_accuracy": 0.7961168810725212,
      "num_tokens": 455471279.0,
      "step": 14270
    },
    {
      "entropy": 0.5998503435403109,
      "epoch": 1.3130056756430384,
      "grad_norm": 0.16414067149162292,
      "learning_rate": 5.623639095899653e-05,
      "loss": 0.5961,
      "mean_token_accuracy": 0.8192868903279305,
      "num_tokens": 455502717.0,
      "step": 14271
    },
    {
      "entropy": 0.6139820870012045,
      "epoch": 1.3130976820144795,
      "grad_norm": 0.1577177792787552,
      "learning_rate": 5.623332413285492e-05,
      "loss": 0.6028,
      "mean_token_accuracy": 0.8177148997783661,
      "num_tokens": 455534310.0,
      "step": 14272
    },
    {
      "entropy": 0.6673392616212368,
      "epoch": 1.3131896883859207,
      "grad_norm": 0.16542144119739532,
      "learning_rate": 5.6230257306713285e-05,
      "loss": 0.6583,
      "mean_token_accuracy": 0.7991878315806389,
      "num_tokens": 455567078.0,
      "step": 14273
    },
    {
      "entropy": 0.6109679462388158,
      "epoch": 1.313281694757362,
      "grad_norm": 0.14988096058368683,
      "learning_rate": 5.622719048057166e-05,
      "loss": 0.6026,
      "mean_token_accuracy": 0.8181577511131763,
      "num_tokens": 455599583.0,
      "step": 14274
    },
    {
      "entropy": 0.5541861988604069,
      "epoch": 1.3133737011288031,
      "grad_norm": 0.15400919318199158,
      "learning_rate": 5.622412365443003e-05,
      "loss": 0.5358,
      "mean_token_accuracy": 0.8313758447766304,
      "num_tokens": 455631376.0,
      "step": 14275
    },
    {
      "entropy": 0.5893823746591806,
      "epoch": 1.3134657075002445,
      "grad_norm": 0.15761935710906982,
      "learning_rate": 5.622105682828841e-05,
      "loss": 0.5533,
      "mean_token_accuracy": 0.8270940817892551,
      "num_tokens": 455663152.0,
      "step": 14276
    },
    {
      "entropy": 0.7214862322434783,
      "epoch": 1.3135577138716856,
      "grad_norm": 0.16710138320922852,
      "learning_rate": 5.6217990002146783e-05,
      "loss": 0.7037,
      "mean_token_accuracy": 0.7860258370637894,
      "num_tokens": 455694133.0,
      "step": 14277
    },
    {
      "entropy": 0.6369915418326855,
      "epoch": 1.3136497202431268,
      "grad_norm": 0.16460563242435455,
      "learning_rate": 5.621492317600515e-05,
      "loss": 0.6354,
      "mean_token_accuracy": 0.8045721128582954,
      "num_tokens": 455726673.0,
      "step": 14278
    },
    {
      "entropy": 0.5620156582444906,
      "epoch": 1.3137417266145681,
      "grad_norm": 0.1597939133644104,
      "learning_rate": 5.621185634986353e-05,
      "loss": 0.552,
      "mean_token_accuracy": 0.8287848830223083,
      "num_tokens": 455758920.0,
      "step": 14279
    },
    {
      "entropy": 0.6581323035061359,
      "epoch": 1.3138337329860093,
      "grad_norm": 0.15593285858631134,
      "learning_rate": 5.620878952372191e-05,
      "loss": 0.6453,
      "mean_token_accuracy": 0.8016141839325428,
      "num_tokens": 455791366.0,
      "step": 14280
    },
    {
      "entropy": 0.5903088105842471,
      "epoch": 1.3139257393574506,
      "grad_norm": 0.15223237872123718,
      "learning_rate": 5.6205722697580275e-05,
      "loss": 0.5748,
      "mean_token_accuracy": 0.8195203766226768,
      "num_tokens": 455823044.0,
      "step": 14281
    },
    {
      "entropy": 0.5320273032411933,
      "epoch": 1.3140177457288917,
      "grad_norm": 0.14569896459579468,
      "learning_rate": 5.620265587143865e-05,
      "loss": 0.5257,
      "mean_token_accuracy": 0.836896326392889,
      "num_tokens": 455854588.0,
      "step": 14282
    },
    {
      "entropy": 0.5800527092069387,
      "epoch": 1.3141097521003329,
      "grad_norm": 0.15453685820102692,
      "learning_rate": 5.619958904529703e-05,
      "loss": 0.5748,
      "mean_token_accuracy": 0.8212558589875698,
      "num_tokens": 455886779.0,
      "step": 14283
    },
    {
      "entropy": 0.5755104180425406,
      "epoch": 1.3142017584717742,
      "grad_norm": 0.15006892383098602,
      "learning_rate": 5.61965222191554e-05,
      "loss": 0.5702,
      "mean_token_accuracy": 0.8273985460400581,
      "num_tokens": 455919076.0,
      "step": 14284
    },
    {
      "entropy": 0.5452785547822714,
      "epoch": 1.3142937648432154,
      "grad_norm": 0.1545899510383606,
      "learning_rate": 5.6193455393013774e-05,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8364806920289993,
      "num_tokens": 455951080.0,
      "step": 14285
    },
    {
      "entropy": 0.6554987076669931,
      "epoch": 1.3143857712146567,
      "grad_norm": 0.16308172047138214,
      "learning_rate": 5.619038856687214e-05,
      "loss": 0.6536,
      "mean_token_accuracy": 0.8012369498610497,
      "num_tokens": 455983339.0,
      "step": 14286
    },
    {
      "entropy": 0.5276092053391039,
      "epoch": 1.3144777775860979,
      "grad_norm": 0.15014037489891052,
      "learning_rate": 5.618732174073052e-05,
      "loss": 0.515,
      "mean_token_accuracy": 0.8380127176642418,
      "num_tokens": 456015530.0,
      "step": 14287
    },
    {
      "entropy": 0.6694241855293512,
      "epoch": 1.314569783957539,
      "grad_norm": 0.15935468673706055,
      "learning_rate": 5.61842549145889e-05,
      "loss": 0.6578,
      "mean_token_accuracy": 0.801991630345583,
      "num_tokens": 456047051.0,
      "step": 14288
    },
    {
      "entropy": 0.559768492821604,
      "epoch": 1.3146617903289803,
      "grad_norm": 0.15483158826828003,
      "learning_rate": 5.6181188088447266e-05,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8315756320953369,
      "num_tokens": 456078577.0,
      "step": 14289
    },
    {
      "entropy": 0.5845174547284842,
      "epoch": 1.3147537967004215,
      "grad_norm": 0.1583409309387207,
      "learning_rate": 5.617812126230564e-05,
      "loss": 0.5746,
      "mean_token_accuracy": 0.8222522735595703,
      "num_tokens": 456111248.0,
      "step": 14290
    },
    {
      "entropy": 0.5779199730604887,
      "epoch": 1.3148458030718628,
      "grad_norm": 0.16361188888549805,
      "learning_rate": 5.617505443616402e-05,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8327660001814365,
      "num_tokens": 456142665.0,
      "step": 14291
    },
    {
      "entropy": 0.5821020947769284,
      "epoch": 1.314937809443304,
      "grad_norm": 0.1632312685251236,
      "learning_rate": 5.617198761002239e-05,
      "loss": 0.5633,
      "mean_token_accuracy": 0.8252839408814907,
      "num_tokens": 456175151.0,
      "step": 14292
    },
    {
      "entropy": 0.6041314173489809,
      "epoch": 1.315029815814745,
      "grad_norm": 0.1544383317232132,
      "learning_rate": 5.6168920783880764e-05,
      "loss": 0.5925,
      "mean_token_accuracy": 0.8190574683248997,
      "num_tokens": 456207151.0,
      "step": 14293
    },
    {
      "entropy": 0.5172856021672487,
      "epoch": 1.3151218221861865,
      "grad_norm": 0.15621575713157654,
      "learning_rate": 5.616585395773913e-05,
      "loss": 0.4989,
      "mean_token_accuracy": 0.84850212931633,
      "num_tokens": 456239857.0,
      "step": 14294
    },
    {
      "entropy": 0.6184176672250032,
      "epoch": 1.3152138285576276,
      "grad_norm": 0.1567763388156891,
      "learning_rate": 5.6162787131597514e-05,
      "loss": 0.6056,
      "mean_token_accuracy": 0.8124345727264881,
      "num_tokens": 456271764.0,
      "step": 14295
    },
    {
      "entropy": 0.6587536102160811,
      "epoch": 1.315305834929069,
      "grad_norm": 0.1670931577682495,
      "learning_rate": 5.615972030545589e-05,
      "loss": 0.6436,
      "mean_token_accuracy": 0.810546163469553,
      "num_tokens": 456303140.0,
      "step": 14296
    },
    {
      "entropy": 0.6190512869507074,
      "epoch": 1.31539784130051,
      "grad_norm": 0.15926295518875122,
      "learning_rate": 5.6156653479314256e-05,
      "loss": 0.603,
      "mean_token_accuracy": 0.8158088251948357,
      "num_tokens": 456335066.0,
      "step": 14297
    },
    {
      "entropy": 0.6048125680536032,
      "epoch": 1.3154898476719512,
      "grad_norm": 0.1559811383485794,
      "learning_rate": 5.615358665317263e-05,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8212872073054314,
      "num_tokens": 456367314.0,
      "step": 14298
    },
    {
      "entropy": 0.6023055436089635,
      "epoch": 1.3155818540433926,
      "grad_norm": 0.15343110263347626,
      "learning_rate": 5.615051982703101e-05,
      "loss": 0.6006,
      "mean_token_accuracy": 0.8142195269465446,
      "num_tokens": 456399447.0,
      "step": 14299
    },
    {
      "entropy": 0.5855016801506281,
      "epoch": 1.3156738604148337,
      "grad_norm": 0.15718570351600647,
      "learning_rate": 5.614745300088938e-05,
      "loss": 0.5714,
      "mean_token_accuracy": 0.821625143289566,
      "num_tokens": 456431842.0,
      "step": 14300
    },
    {
      "entropy": 0.6495391707867384,
      "epoch": 1.315765866786275,
      "grad_norm": 0.1566428542137146,
      "learning_rate": 5.6144386174747755e-05,
      "loss": 0.6463,
      "mean_token_accuracy": 0.8037073500454426,
      "num_tokens": 456464007.0,
      "step": 14301
    },
    {
      "entropy": 0.648623501881957,
      "epoch": 1.3158578731577162,
      "grad_norm": 0.16226403415203094,
      "learning_rate": 5.614131934860612e-05,
      "loss": 0.6121,
      "mean_token_accuracy": 0.810089036822319,
      "num_tokens": 456495514.0,
      "step": 14302
    },
    {
      "entropy": 0.536782099865377,
      "epoch": 1.3159498795291573,
      "grad_norm": 0.15871337056159973,
      "learning_rate": 5.6138252522464504e-05,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8347918540239334,
      "num_tokens": 456527542.0,
      "step": 14303
    },
    {
      "entropy": 0.6388052515685558,
      "epoch": 1.3160418859005987,
      "grad_norm": 0.15867365896701813,
      "learning_rate": 5.613518569632288e-05,
      "loss": 0.6266,
      "mean_token_accuracy": 0.8091149590909481,
      "num_tokens": 456559378.0,
      "step": 14304
    },
    {
      "entropy": 0.5850212443619967,
      "epoch": 1.3161338922720398,
      "grad_norm": 0.17406588792800903,
      "learning_rate": 5.6132118870181246e-05,
      "loss": 0.5815,
      "mean_token_accuracy": 0.8236736208200455,
      "num_tokens": 456590734.0,
      "step": 14305
    },
    {
      "entropy": 0.629714097827673,
      "epoch": 1.3162258986434812,
      "grad_norm": 0.17023196816444397,
      "learning_rate": 5.612905204403962e-05,
      "loss": 0.633,
      "mean_token_accuracy": 0.8090553805232048,
      "num_tokens": 456622625.0,
      "step": 14306
    },
    {
      "entropy": 0.5990107711404562,
      "epoch": 1.3163179050149223,
      "grad_norm": 0.159720778465271,
      "learning_rate": 5.6125985217898e-05,
      "loss": 0.6084,
      "mean_token_accuracy": 0.8152787536382675,
      "num_tokens": 456654813.0,
      "step": 14307
    },
    {
      "entropy": 0.5776107860729098,
      "epoch": 1.3164099113863634,
      "grad_norm": 0.15146517753601074,
      "learning_rate": 5.612291839175638e-05,
      "loss": 0.5545,
      "mean_token_accuracy": 0.8270467706024647,
      "num_tokens": 456686801.0,
      "step": 14308
    },
    {
      "entropy": 0.5821972563862801,
      "epoch": 1.3165019177578048,
      "grad_norm": 0.14831650257110596,
      "learning_rate": 5.6119851565614745e-05,
      "loss": 0.5617,
      "mean_token_accuracy": 0.8255185633897781,
      "num_tokens": 456718430.0,
      "step": 14309
    },
    {
      "entropy": 0.5424175690859556,
      "epoch": 1.316593924129246,
      "grad_norm": 0.15142740309238434,
      "learning_rate": 5.6116784739473126e-05,
      "loss": 0.5168,
      "mean_token_accuracy": 0.8414005935192108,
      "num_tokens": 456750096.0,
      "step": 14310
    },
    {
      "entropy": 0.5882885549217463,
      "epoch": 1.3166859305006873,
      "grad_norm": 0.15798240900039673,
      "learning_rate": 5.61137179133315e-05,
      "loss": 0.5796,
      "mean_token_accuracy": 0.8250640816986561,
      "num_tokens": 456781435.0,
      "step": 14311
    },
    {
      "entropy": 0.7190563622862101,
      "epoch": 1.3167779368721284,
      "grad_norm": 0.1691783368587494,
      "learning_rate": 5.611065108718987e-05,
      "loss": 0.7128,
      "mean_token_accuracy": 0.7878545112907887,
      "num_tokens": 456813365.0,
      "step": 14312
    },
    {
      "entropy": 0.670741556212306,
      "epoch": 1.3168699432435695,
      "grad_norm": 0.15555615723133087,
      "learning_rate": 5.6107584261048244e-05,
      "loss": 0.6559,
      "mean_token_accuracy": 0.7978518046438694,
      "num_tokens": 456845621.0,
      "step": 14313
    },
    {
      "entropy": 0.558360243216157,
      "epoch": 1.316961949615011,
      "grad_norm": 0.15487989783287048,
      "learning_rate": 5.6104517434906625e-05,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8310375064611435,
      "num_tokens": 456877465.0,
      "step": 14314
    },
    {
      "entropy": 0.7007043920457363,
      "epoch": 1.317053955986452,
      "grad_norm": 0.1672583520412445,
      "learning_rate": 5.610145060876499e-05,
      "loss": 0.6936,
      "mean_token_accuracy": 0.7864038348197937,
      "num_tokens": 456908601.0,
      "step": 14315
    },
    {
      "entropy": 0.627567826770246,
      "epoch": 1.3171459623578934,
      "grad_norm": 0.1607467383146286,
      "learning_rate": 5.609838378262337e-05,
      "loss": 0.609,
      "mean_token_accuracy": 0.8149882964789867,
      "num_tokens": 456940704.0,
      "step": 14316
    },
    {
      "entropy": 0.6331809125840664,
      "epoch": 1.3172379687293345,
      "grad_norm": 0.16280362010002136,
      "learning_rate": 5.6095316956481735e-05,
      "loss": 0.6369,
      "mean_token_accuracy": 0.807162020355463,
      "num_tokens": 456972384.0,
      "step": 14317
    },
    {
      "entropy": 0.5950271934270859,
      "epoch": 1.3173299751007757,
      "grad_norm": 0.15642060339450836,
      "learning_rate": 5.609225013034012e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.822408813983202,
      "num_tokens": 457005109.0,
      "step": 14318
    },
    {
      "entropy": 0.6101786457002163,
      "epoch": 1.317421981472217,
      "grad_norm": 0.1566852182149887,
      "learning_rate": 5.608918330419849e-05,
      "loss": 0.5998,
      "mean_token_accuracy": 0.8215819634497166,
      "num_tokens": 457037036.0,
      "step": 14319
    },
    {
      "entropy": 0.7006552927196026,
      "epoch": 1.3175139878436581,
      "grad_norm": 0.16823391616344452,
      "learning_rate": 5.608611647805686e-05,
      "loss": 0.6945,
      "mean_token_accuracy": 0.7896288670599461,
      "num_tokens": 457069085.0,
      "step": 14320
    },
    {
      "entropy": 0.6383642749860883,
      "epoch": 1.3176059942150995,
      "grad_norm": 0.1598946750164032,
      "learning_rate": 5.6083049651915234e-05,
      "loss": 0.6174,
      "mean_token_accuracy": 0.8123338855803013,
      "num_tokens": 457099883.0,
      "step": 14321
    },
    {
      "entropy": 0.6377084190025926,
      "epoch": 1.3176980005865406,
      "grad_norm": 0.15458174049854279,
      "learning_rate": 5.6079982825773615e-05,
      "loss": 0.6237,
      "mean_token_accuracy": 0.8116252832114697,
      "num_tokens": 457131829.0,
      "step": 14322
    },
    {
      "entropy": 0.6212920714169741,
      "epoch": 1.3177900069579818,
      "grad_norm": 0.15234003961086273,
      "learning_rate": 5.607691599963198e-05,
      "loss": 0.6061,
      "mean_token_accuracy": 0.8118273057043552,
      "num_tokens": 457164037.0,
      "step": 14323
    },
    {
      "entropy": 0.5865630889311433,
      "epoch": 1.3178820133294231,
      "grad_norm": 0.15448720753192902,
      "learning_rate": 5.607384917349036e-05,
      "loss": 0.5725,
      "mean_token_accuracy": 0.824253898113966,
      "num_tokens": 457195207.0,
      "step": 14324
    },
    {
      "entropy": 0.5724903428927064,
      "epoch": 1.3179740197008643,
      "grad_norm": 0.15112487971782684,
      "learning_rate": 5.6070782347348726e-05,
      "loss": 0.5558,
      "mean_token_accuracy": 0.8275623433291912,
      "num_tokens": 457226430.0,
      "step": 14325
    },
    {
      "entropy": 0.5536251626908779,
      "epoch": 1.3180660260723056,
      "grad_norm": 0.15124279260635376,
      "learning_rate": 5.606771552120711e-05,
      "loss": 0.5468,
      "mean_token_accuracy": 0.8295178189873695,
      "num_tokens": 457258322.0,
      "step": 14326
    },
    {
      "entropy": 0.5862578060477972,
      "epoch": 1.3181580324437467,
      "grad_norm": 0.15307965874671936,
      "learning_rate": 5.606464869506548e-05,
      "loss": 0.5703,
      "mean_token_accuracy": 0.8216967135667801,
      "num_tokens": 457289943.0,
      "step": 14327
    },
    {
      "entropy": 0.6473401514813304,
      "epoch": 1.3182500388151879,
      "grad_norm": 0.15649108588695526,
      "learning_rate": 5.606158186892385e-05,
      "loss": 0.647,
      "mean_token_accuracy": 0.8027572743594646,
      "num_tokens": 457322678.0,
      "step": 14328
    },
    {
      "entropy": 0.5847535897046328,
      "epoch": 1.3183420451866292,
      "grad_norm": 0.16222789883613586,
      "learning_rate": 5.6058515042782224e-05,
      "loss": 0.5874,
      "mean_token_accuracy": 0.8238011375069618,
      "num_tokens": 457355248.0,
      "step": 14329
    },
    {
      "entropy": 0.5643026009202003,
      "epoch": 1.3184340515580704,
      "grad_norm": 0.1513214260339737,
      "learning_rate": 5.6055448216640606e-05,
      "loss": 0.5654,
      "mean_token_accuracy": 0.8251530192792416,
      "num_tokens": 457387468.0,
      "step": 14330
    },
    {
      "entropy": 0.5780859924852848,
      "epoch": 1.3185260579295117,
      "grad_norm": 0.16005206108093262,
      "learning_rate": 5.6052381390498974e-05,
      "loss": 0.5759,
      "mean_token_accuracy": 0.8217889964580536,
      "num_tokens": 457419638.0,
      "step": 14331
    },
    {
      "entropy": 0.6795655619353056,
      "epoch": 1.3186180643009529,
      "grad_norm": 0.16023504734039307,
      "learning_rate": 5.604931456435735e-05,
      "loss": 0.6755,
      "mean_token_accuracy": 0.791520718485117,
      "num_tokens": 457451675.0,
      "step": 14332
    },
    {
      "entropy": 0.6093870894983411,
      "epoch": 1.318710070672394,
      "grad_norm": 0.1639716476202011,
      "learning_rate": 5.6046247738215716e-05,
      "loss": 0.601,
      "mean_token_accuracy": 0.8174678832292557,
      "num_tokens": 457483248.0,
      "step": 14333
    },
    {
      "entropy": 0.6274670669808984,
      "epoch": 1.3188020770438353,
      "grad_norm": 0.16020803153514862,
      "learning_rate": 5.60431809120741e-05,
      "loss": 0.6251,
      "mean_token_accuracy": 0.8094567060470581,
      "num_tokens": 457515094.0,
      "step": 14334
    },
    {
      "entropy": 0.6269160769879818,
      "epoch": 1.3188940834152765,
      "grad_norm": 0.15699467062950134,
      "learning_rate": 5.604011408593247e-05,
      "loss": 0.6201,
      "mean_token_accuracy": 0.8095708899199963,
      "num_tokens": 457546927.0,
      "step": 14335
    },
    {
      "entropy": 0.6121396068483591,
      "epoch": 1.3189860897867178,
      "grad_norm": 0.15616635978221893,
      "learning_rate": 5.603704725979084e-05,
      "loss": 0.5884,
      "mean_token_accuracy": 0.8172132968902588,
      "num_tokens": 457578935.0,
      "step": 14336
    },
    {
      "entropy": 0.6010236889123917,
      "epoch": 1.319078096158159,
      "grad_norm": 0.15298834443092346,
      "learning_rate": 5.6033980433649215e-05,
      "loss": 0.5712,
      "mean_token_accuracy": 0.8218872435390949,
      "num_tokens": 457610601.0,
      "step": 14337
    },
    {
      "entropy": 0.6475709807127714,
      "epoch": 1.3191701025296,
      "grad_norm": 0.16115571558475494,
      "learning_rate": 5.6030913607507596e-05,
      "loss": 0.6295,
      "mean_token_accuracy": 0.8079609982669353,
      "num_tokens": 457642579.0,
      "step": 14338
    },
    {
      "entropy": 0.5321576856076717,
      "epoch": 1.3192621089010415,
      "grad_norm": 0.15747933089733124,
      "learning_rate": 5.6027846781365964e-05,
      "loss": 0.5257,
      "mean_token_accuracy": 0.83638471737504,
      "num_tokens": 457673602.0,
      "step": 14339
    },
    {
      "entropy": 0.6855260012671351,
      "epoch": 1.3193541152724826,
      "grad_norm": 0.16276206076145172,
      "learning_rate": 5.602477995522434e-05,
      "loss": 0.6926,
      "mean_token_accuracy": 0.7880529426038265,
      "num_tokens": 457705773.0,
      "step": 14340
    },
    {
      "entropy": 0.5950098671019077,
      "epoch": 1.319446121643924,
      "grad_norm": 0.15772190690040588,
      "learning_rate": 5.602171312908272e-05,
      "loss": 0.588,
      "mean_token_accuracy": 0.8195012100040913,
      "num_tokens": 457737288.0,
      "step": 14341
    },
    {
      "entropy": 0.6876471526920795,
      "epoch": 1.319538128015365,
      "grad_norm": 0.162477046251297,
      "learning_rate": 5.601864630294109e-05,
      "loss": 0.6869,
      "mean_token_accuracy": 0.7926183640956879,
      "num_tokens": 457769907.0,
      "step": 14342
    },
    {
      "entropy": 0.5371973365545273,
      "epoch": 1.3196301343868062,
      "grad_norm": 0.14926567673683167,
      "learning_rate": 5.601557947679946e-05,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8380269855260849,
      "num_tokens": 457801929.0,
      "step": 14343
    },
    {
      "entropy": 0.6704232534393668,
      "epoch": 1.3197221407582476,
      "grad_norm": 0.1644313484430313,
      "learning_rate": 5.601251265065783e-05,
      "loss": 0.6517,
      "mean_token_accuracy": 0.7965182214975357,
      "num_tokens": 457833843.0,
      "step": 14344
    },
    {
      "entropy": 0.5881751934066415,
      "epoch": 1.3198141471296887,
      "grad_norm": 0.1612999439239502,
      "learning_rate": 5.600944582451622e-05,
      "loss": 0.5803,
      "mean_token_accuracy": 0.8243576698005199,
      "num_tokens": 457865474.0,
      "step": 14345
    },
    {
      "entropy": 0.5380340181291103,
      "epoch": 1.31990615350113,
      "grad_norm": 0.14329321682453156,
      "learning_rate": 5.600637899837459e-05,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8388099670410156,
      "num_tokens": 457897233.0,
      "step": 14346
    },
    {
      "entropy": 0.6772993407212198,
      "epoch": 1.3199981598725712,
      "grad_norm": 0.15698875486850739,
      "learning_rate": 5.6003312172232955e-05,
      "loss": 0.6631,
      "mean_token_accuracy": 0.8066830523312092,
      "num_tokens": 457929251.0,
      "step": 14347
    },
    {
      "entropy": 0.5466817393898964,
      "epoch": 1.3200901662440123,
      "grad_norm": 0.15791064500808716,
      "learning_rate": 5.600024534609133e-05,
      "loss": 0.5337,
      "mean_token_accuracy": 0.8354505859315395,
      "num_tokens": 457961715.0,
      "step": 14348
    },
    {
      "entropy": 0.7277012839913368,
      "epoch": 1.3201821726154537,
      "grad_norm": 0.16310670971870422,
      "learning_rate": 5.599717851994971e-05,
      "loss": 0.715,
      "mean_token_accuracy": 0.7845185026526451,
      "num_tokens": 457994053.0,
      "step": 14349
    },
    {
      "entropy": 0.5952043309807777,
      "epoch": 1.3202741789868948,
      "grad_norm": 0.15930312871932983,
      "learning_rate": 5.5994111693808085e-05,
      "loss": 0.5818,
      "mean_token_accuracy": 0.8214195892214775,
      "num_tokens": 458025106.0,
      "step": 14350
    },
    {
      "entropy": 0.5934214917942882,
      "epoch": 1.3203661853583362,
      "grad_norm": 0.1527300775051117,
      "learning_rate": 5.599104486766645e-05,
      "loss": 0.5748,
      "mean_token_accuracy": 0.8191104345023632,
      "num_tokens": 458057524.0,
      "step": 14351
    },
    {
      "entropy": 0.6556130181998014,
      "epoch": 1.3204581917297773,
      "grad_norm": 0.1568373739719391,
      "learning_rate": 5.598797804152483e-05,
      "loss": 0.6451,
      "mean_token_accuracy": 0.8055623732507229,
      "num_tokens": 458089507.0,
      "step": 14352
    },
    {
      "entropy": 0.6621993258595467,
      "epoch": 1.3205501981012184,
      "grad_norm": 0.1569075584411621,
      "learning_rate": 5.598491121538321e-05,
      "loss": 0.6426,
      "mean_token_accuracy": 0.8020729795098305,
      "num_tokens": 458122247.0,
      "step": 14353
    },
    {
      "entropy": 0.604265196248889,
      "epoch": 1.3206422044726598,
      "grad_norm": 0.15170902013778687,
      "learning_rate": 5.598184438924158e-05,
      "loss": 0.5925,
      "mean_token_accuracy": 0.8183776140213013,
      "num_tokens": 458153859.0,
      "step": 14354
    },
    {
      "entropy": 0.6565006449818611,
      "epoch": 1.320734210844101,
      "grad_norm": 0.16170887649059296,
      "learning_rate": 5.597877756309995e-05,
      "loss": 0.6383,
      "mean_token_accuracy": 0.8048034720122814,
      "num_tokens": 458186070.0,
      "step": 14355
    },
    {
      "entropy": 0.6184587515890598,
      "epoch": 1.3208262172155423,
      "grad_norm": 0.15416696667671204,
      "learning_rate": 5.597571073695832e-05,
      "loss": 0.5974,
      "mean_token_accuracy": 0.8159245029091835,
      "num_tokens": 458217464.0,
      "step": 14356
    },
    {
      "entropy": 0.6868350375443697,
      "epoch": 1.3209182235869834,
      "grad_norm": 0.16425766050815582,
      "learning_rate": 5.59726439108167e-05,
      "loss": 0.6788,
      "mean_token_accuracy": 0.7911044210195541,
      "num_tokens": 458249592.0,
      "step": 14357
    },
    {
      "entropy": 0.6397701613605022,
      "epoch": 1.3210102299584245,
      "grad_norm": 0.1562775820493698,
      "learning_rate": 5.5969577084675076e-05,
      "loss": 0.6357,
      "mean_token_accuracy": 0.8044923767447472,
      "num_tokens": 458281714.0,
      "step": 14358
    },
    {
      "entropy": 0.4990499969571829,
      "epoch": 1.321102236329866,
      "grad_norm": 0.1582186073064804,
      "learning_rate": 5.5966510258533444e-05,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8488713875412941,
      "num_tokens": 458313474.0,
      "step": 14359
    },
    {
      "entropy": 0.509779118001461,
      "epoch": 1.321194242701307,
      "grad_norm": 0.14967317879199982,
      "learning_rate": 5.596344343239182e-05,
      "loss": 0.4978,
      "mean_token_accuracy": 0.8466059044003487,
      "num_tokens": 458346042.0,
      "step": 14360
    },
    {
      "entropy": 0.6614109072834253,
      "epoch": 1.3212862490727484,
      "grad_norm": 0.16324277222156525,
      "learning_rate": 5.59603766062502e-05,
      "loss": 0.6582,
      "mean_token_accuracy": 0.7987163662910461,
      "num_tokens": 458377473.0,
      "step": 14361
    },
    {
      "entropy": 0.44483199110254645,
      "epoch": 1.3213782554441895,
      "grad_norm": 0.14750346541404724,
      "learning_rate": 5.595730978010857e-05,
      "loss": 0.4344,
      "mean_token_accuracy": 0.8633127473294735,
      "num_tokens": 458409271.0,
      "step": 14362
    },
    {
      "entropy": 0.7304984563961625,
      "epoch": 1.3214702618156307,
      "grad_norm": 0.1769295185804367,
      "learning_rate": 5.595424295396694e-05,
      "loss": 0.7266,
      "mean_token_accuracy": 0.7820142172276974,
      "num_tokens": 458441291.0,
      "step": 14363
    },
    {
      "entropy": 0.5160506838001311,
      "epoch": 1.321562268187072,
      "grad_norm": 0.15762101113796234,
      "learning_rate": 5.595117612782531e-05,
      "loss": 0.5069,
      "mean_token_accuracy": 0.8436752483248711,
      "num_tokens": 458473094.0,
      "step": 14364
    },
    {
      "entropy": 0.5859627351164818,
      "epoch": 1.3216542745585131,
      "grad_norm": 0.15062426030635834,
      "learning_rate": 5.594810930168369e-05,
      "loss": 0.5562,
      "mean_token_accuracy": 0.8263354077935219,
      "num_tokens": 458504764.0,
      "step": 14365
    },
    {
      "entropy": 0.6544387945905328,
      "epoch": 1.3217462809299545,
      "grad_norm": 0.16052736341953278,
      "learning_rate": 5.5945042475542066e-05,
      "loss": 0.6441,
      "mean_token_accuracy": 0.8064819872379303,
      "num_tokens": 458536516.0,
      "step": 14366
    },
    {
      "entropy": 0.6782863419502974,
      "epoch": 1.3218382873013956,
      "grad_norm": 0.16182300448417664,
      "learning_rate": 5.5941975649400434e-05,
      "loss": 0.6779,
      "mean_token_accuracy": 0.7949536442756653,
      "num_tokens": 458569201.0,
      "step": 14367
    },
    {
      "entropy": 0.5809610178694129,
      "epoch": 1.3219302936728368,
      "grad_norm": 0.15497581660747528,
      "learning_rate": 5.5938908823258815e-05,
      "loss": 0.5639,
      "mean_token_accuracy": 0.8270348608493805,
      "num_tokens": 458600999.0,
      "step": 14368
    },
    {
      "entropy": 0.6055756425485015,
      "epoch": 1.3220223000442781,
      "grad_norm": 0.15633367002010345,
      "learning_rate": 5.593584199711719e-05,
      "loss": 0.5813,
      "mean_token_accuracy": 0.8211613297462463,
      "num_tokens": 458632591.0,
      "step": 14369
    },
    {
      "entropy": 0.6289436873048544,
      "epoch": 1.3221143064157193,
      "grad_norm": 0.17039532959461212,
      "learning_rate": 5.593277517097556e-05,
      "loss": 0.6231,
      "mean_token_accuracy": 0.8111468553543091,
      "num_tokens": 458663310.0,
      "step": 14370
    },
    {
      "entropy": 0.6449506673961878,
      "epoch": 1.3222063127871606,
      "grad_norm": 0.15903863310813904,
      "learning_rate": 5.592970834483393e-05,
      "loss": 0.6324,
      "mean_token_accuracy": 0.8069311380386353,
      "num_tokens": 458694505.0,
      "step": 14371
    },
    {
      "entropy": 0.6367842238396406,
      "epoch": 1.3222983191586017,
      "grad_norm": 0.1561354398727417,
      "learning_rate": 5.5926641518692314e-05,
      "loss": 0.6328,
      "mean_token_accuracy": 0.8062745593488216,
      "num_tokens": 458726728.0,
      "step": 14372
    },
    {
      "entropy": 0.6034509055316448,
      "epoch": 1.3223903255300429,
      "grad_norm": 0.15767411887645721,
      "learning_rate": 5.592357469255068e-05,
      "loss": 0.6001,
      "mean_token_accuracy": 0.817828793078661,
      "num_tokens": 458758457.0,
      "step": 14373
    },
    {
      "entropy": 0.5606782357208431,
      "epoch": 1.3224823319014842,
      "grad_norm": 0.1530206948518753,
      "learning_rate": 5.5920507866409056e-05,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8319174125790596,
      "num_tokens": 458790413.0,
      "step": 14374
    },
    {
      "entropy": 0.6146647278219461,
      "epoch": 1.3225743382729254,
      "grad_norm": 0.15808473527431488,
      "learning_rate": 5.5917441040267424e-05,
      "loss": 0.6118,
      "mean_token_accuracy": 0.813124805688858,
      "num_tokens": 458823150.0,
      "step": 14375
    },
    {
      "entropy": 0.6792084854096174,
      "epoch": 1.3226663446443667,
      "grad_norm": 0.16296197474002838,
      "learning_rate": 5.5914374214125806e-05,
      "loss": 0.6632,
      "mean_token_accuracy": 0.7980935834348202,
      "num_tokens": 458854817.0,
      "step": 14376
    },
    {
      "entropy": 0.672207523137331,
      "epoch": 1.3227583510158079,
      "grad_norm": 0.1612107753753662,
      "learning_rate": 5.591130738798418e-05,
      "loss": 0.6686,
      "mean_token_accuracy": 0.7953696884214878,
      "num_tokens": 458886647.0,
      "step": 14377
    },
    {
      "entropy": 0.5280847074463964,
      "epoch": 1.322850357387249,
      "grad_norm": 0.14659735560417175,
      "learning_rate": 5.590824056184255e-05,
      "loss": 0.5228,
      "mean_token_accuracy": 0.8403030894696712,
      "num_tokens": 458918723.0,
      "step": 14378
    },
    {
      "entropy": 0.6814106609672308,
      "epoch": 1.3229423637586903,
      "grad_norm": 0.16366586089134216,
      "learning_rate": 5.590517373570092e-05,
      "loss": 0.6693,
      "mean_token_accuracy": 0.7994055785238743,
      "num_tokens": 458950664.0,
      "step": 14379
    },
    {
      "entropy": 0.566112233325839,
      "epoch": 1.3230343701301315,
      "grad_norm": 0.14978960156440735,
      "learning_rate": 5.5902106909559304e-05,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8328451253473759,
      "num_tokens": 458983351.0,
      "step": 14380
    },
    {
      "entropy": 0.59717614389956,
      "epoch": 1.3231263765015728,
      "grad_norm": 0.15228575468063354,
      "learning_rate": 5.589904008341767e-05,
      "loss": 0.5742,
      "mean_token_accuracy": 0.8222027085721493,
      "num_tokens": 459015076.0,
      "step": 14381
    },
    {
      "entropy": 0.6230165036395192,
      "epoch": 1.323218382873014,
      "grad_norm": 0.15703518688678741,
      "learning_rate": 5.589597325727605e-05,
      "loss": 0.6094,
      "mean_token_accuracy": 0.8156979382038116,
      "num_tokens": 459047328.0,
      "step": 14382
    },
    {
      "entropy": 0.6772934158798307,
      "epoch": 1.323310389244455,
      "grad_norm": 0.16027435660362244,
      "learning_rate": 5.5892906431134415e-05,
      "loss": 0.6651,
      "mean_token_accuracy": 0.7999740764498711,
      "num_tokens": 459079621.0,
      "step": 14383
    },
    {
      "entropy": 0.5133299371227622,
      "epoch": 1.3234023956158965,
      "grad_norm": 0.15298408269882202,
      "learning_rate": 5.58898396049928e-05,
      "loss": 0.507,
      "mean_token_accuracy": 0.8469127751886845,
      "num_tokens": 459111642.0,
      "step": 14384
    },
    {
      "entropy": 0.6170206246897578,
      "epoch": 1.3234944019873376,
      "grad_norm": 0.1559583842754364,
      "learning_rate": 5.588677277885117e-05,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8140238672494888,
      "num_tokens": 459144217.0,
      "step": 14385
    },
    {
      "entropy": 0.5723481103777885,
      "epoch": 1.323586408358779,
      "grad_norm": 0.16489177942276,
      "learning_rate": 5.588370595270954e-05,
      "loss": 0.5429,
      "mean_token_accuracy": 0.829243578016758,
      "num_tokens": 459175601.0,
      "step": 14386
    },
    {
      "entropy": 0.6228690817952156,
      "epoch": 1.32367841473022,
      "grad_norm": 0.16019825637340546,
      "learning_rate": 5.588063912656791e-05,
      "loss": 0.6074,
      "mean_token_accuracy": 0.8148720152676105,
      "num_tokens": 459207650.0,
      "step": 14387
    },
    {
      "entropy": 0.5373925948515534,
      "epoch": 1.3237704211016612,
      "grad_norm": 0.14708183705806732,
      "learning_rate": 5.5877572300426295e-05,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8381722941994667,
      "num_tokens": 459239713.0,
      "step": 14388
    },
    {
      "entropy": 0.6607765341177583,
      "epoch": 1.3238624274731026,
      "grad_norm": 0.1702759712934494,
      "learning_rate": 5.587450547428467e-05,
      "loss": 0.6684,
      "mean_token_accuracy": 0.7986993119120598,
      "num_tokens": 459272481.0,
      "step": 14389
    },
    {
      "entropy": 0.6424854025244713,
      "epoch": 1.3239544338445437,
      "grad_norm": 0.1635957509279251,
      "learning_rate": 5.587143864814304e-05,
      "loss": 0.6442,
      "mean_token_accuracy": 0.8039128705859184,
      "num_tokens": 459304553.0,
      "step": 14390
    },
    {
      "entropy": 0.6172916367650032,
      "epoch": 1.324046440215985,
      "grad_norm": 0.15853829681873322,
      "learning_rate": 5.5868371822001405e-05,
      "loss": 0.62,
      "mean_token_accuracy": 0.8118678331375122,
      "num_tokens": 459336229.0,
      "step": 14391
    },
    {
      "entropy": 0.6250624638050795,
      "epoch": 1.3241384465874262,
      "grad_norm": 0.16248925030231476,
      "learning_rate": 5.586530499585979e-05,
      "loss": 0.6288,
      "mean_token_accuracy": 0.807981576770544,
      "num_tokens": 459367507.0,
      "step": 14392
    },
    {
      "entropy": 0.6397762093693018,
      "epoch": 1.3242304529588673,
      "grad_norm": 0.15870517492294312,
      "learning_rate": 5.586223816971816e-05,
      "loss": 0.6305,
      "mean_token_accuracy": 0.8033646456897259,
      "num_tokens": 459399499.0,
      "step": 14393
    },
    {
      "entropy": 0.633925125002861,
      "epoch": 1.3243224593303087,
      "grad_norm": 0.1602535992860794,
      "learning_rate": 5.5859171343576536e-05,
      "loss": 0.621,
      "mean_token_accuracy": 0.8079922161996365,
      "num_tokens": 459431405.0,
      "step": 14394
    },
    {
      "entropy": 0.6573430318385363,
      "epoch": 1.3244144657017498,
      "grad_norm": 0.16661801934242249,
      "learning_rate": 5.5856104517434904e-05,
      "loss": 0.647,
      "mean_token_accuracy": 0.8039291128516197,
      "num_tokens": 459462255.0,
      "step": 14395
    },
    {
      "entropy": 0.6493573468178511,
      "epoch": 1.3245064720731912,
      "grad_norm": 0.16173872351646423,
      "learning_rate": 5.5853037691293285e-05,
      "loss": 0.6394,
      "mean_token_accuracy": 0.8054585158824921,
      "num_tokens": 459494437.0,
      "step": 14396
    },
    {
      "entropy": 0.6575450375676155,
      "epoch": 1.3245984784446323,
      "grad_norm": 0.1578081101179123,
      "learning_rate": 5.584997086515166e-05,
      "loss": 0.6526,
      "mean_token_accuracy": 0.8044386468827724,
      "num_tokens": 459527095.0,
      "step": 14397
    },
    {
      "entropy": 0.658793032169342,
      "epoch": 1.3246904848160734,
      "grad_norm": 0.16694587469100952,
      "learning_rate": 5.584690403901003e-05,
      "loss": 0.6517,
      "mean_token_accuracy": 0.8013628832995892,
      "num_tokens": 459559621.0,
      "step": 14398
    },
    {
      "entropy": 0.6284939800389111,
      "epoch": 1.3247824911875148,
      "grad_norm": 0.15379516780376434,
      "learning_rate": 5.584383721286841e-05,
      "loss": 0.6107,
      "mean_token_accuracy": 0.8108744211494923,
      "num_tokens": 459591625.0,
      "step": 14399
    },
    {
      "entropy": 0.6941548511385918,
      "epoch": 1.324874497558956,
      "grad_norm": 0.15875262022018433,
      "learning_rate": 5.5840770386726784e-05,
      "loss": 0.681,
      "mean_token_accuracy": 0.7948163598775864,
      "num_tokens": 459623456.0,
      "step": 14400
    },
    {
      "entropy": 0.6646914444863796,
      "epoch": 1.3249665039303973,
      "grad_norm": 0.15719391405582428,
      "learning_rate": 5.583770356058515e-05,
      "loss": 0.6522,
      "mean_token_accuracy": 0.7997282445430756,
      "num_tokens": 459655816.0,
      "step": 14401
    },
    {
      "entropy": 0.5767603134736419,
      "epoch": 1.3250585103018384,
      "grad_norm": 0.1494196206331253,
      "learning_rate": 5.5834636734443526e-05,
      "loss": 0.5546,
      "mean_token_accuracy": 0.8300332985818386,
      "num_tokens": 459687938.0,
      "step": 14402
    },
    {
      "entropy": 0.6681399904191494,
      "epoch": 1.3251505166732795,
      "grad_norm": 0.1607617884874344,
      "learning_rate": 5.583156990830191e-05,
      "loss": 0.6452,
      "mean_token_accuracy": 0.7994741201400757,
      "num_tokens": 459719152.0,
      "step": 14403
    },
    {
      "entropy": 0.6097706696018577,
      "epoch": 1.325242523044721,
      "grad_norm": 0.1551206111907959,
      "learning_rate": 5.5828503082160276e-05,
      "loss": 0.5988,
      "mean_token_accuracy": 0.8187141008675098,
      "num_tokens": 459751198.0,
      "step": 14404
    },
    {
      "entropy": 0.6534766629338264,
      "epoch": 1.325334529416162,
      "grad_norm": 0.15588544309139252,
      "learning_rate": 5.582543625601865e-05,
      "loss": 0.6281,
      "mean_token_accuracy": 0.8058770783245564,
      "num_tokens": 459783579.0,
      "step": 14405
    },
    {
      "entropy": 0.5378757105208933,
      "epoch": 1.3254265357876034,
      "grad_norm": 0.15180708467960358,
      "learning_rate": 5.582236942987702e-05,
      "loss": 0.511,
      "mean_token_accuracy": 0.8402684293687344,
      "num_tokens": 459815526.0,
      "step": 14406
    },
    {
      "entropy": 0.59434406273067,
      "epoch": 1.3255185421590445,
      "grad_norm": 0.15083010494709015,
      "learning_rate": 5.58193026037354e-05,
      "loss": 0.5857,
      "mean_token_accuracy": 0.8195525594055653,
      "num_tokens": 459848062.0,
      "step": 14407
    },
    {
      "entropy": 0.7261941116303205,
      "epoch": 1.3256105485304857,
      "grad_norm": 0.1663103997707367,
      "learning_rate": 5.5816235777593774e-05,
      "loss": 0.7229,
      "mean_token_accuracy": 0.7814601548016071,
      "num_tokens": 459880339.0,
      "step": 14408
    },
    {
      "entropy": 0.5724414121359587,
      "epoch": 1.325702554901927,
      "grad_norm": 0.15267473459243774,
      "learning_rate": 5.581316895145214e-05,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8258856907486916,
      "num_tokens": 459911963.0,
      "step": 14409
    },
    {
      "entropy": 0.626915168017149,
      "epoch": 1.3257945612733681,
      "grad_norm": 0.15889562666416168,
      "learning_rate": 5.581010212531052e-05,
      "loss": 0.6098,
      "mean_token_accuracy": 0.8120537586510181,
      "num_tokens": 459943649.0,
      "step": 14410
    },
    {
      "entropy": 0.6492106774821877,
      "epoch": 1.3258865676448095,
      "grad_norm": 0.16171565651893616,
      "learning_rate": 5.58070352991689e-05,
      "loss": 0.6305,
      "mean_token_accuracy": 0.8099975809454918,
      "num_tokens": 459975734.0,
      "step": 14411
    },
    {
      "entropy": 0.6456405306234956,
      "epoch": 1.3259785740162506,
      "grad_norm": 0.15692168474197388,
      "learning_rate": 5.5803968473027266e-05,
      "loss": 0.6421,
      "mean_token_accuracy": 0.8020241260528564,
      "num_tokens": 460007425.0,
      "step": 14412
    },
    {
      "entropy": 0.617704726755619,
      "epoch": 1.3260705803876918,
      "grad_norm": 0.16464699804782867,
      "learning_rate": 5.580090164688564e-05,
      "loss": 0.6149,
      "mean_token_accuracy": 0.816474799066782,
      "num_tokens": 460039469.0,
      "step": 14413
    },
    {
      "entropy": 0.47528793197125196,
      "epoch": 1.3261625867591331,
      "grad_norm": 0.1561247855424881,
      "learning_rate": 5.579783482074401e-05,
      "loss": 0.4581,
      "mean_token_accuracy": 0.8549025245010853,
      "num_tokens": 460071527.0,
      "step": 14414
    },
    {
      "entropy": 0.5752186542376876,
      "epoch": 1.3262545931305743,
      "grad_norm": 0.15093380212783813,
      "learning_rate": 5.579476799460239e-05,
      "loss": 0.5734,
      "mean_token_accuracy": 0.8211898282170296,
      "num_tokens": 460103050.0,
      "step": 14415
    },
    {
      "entropy": 0.6781531423330307,
      "epoch": 1.3263465995020156,
      "grad_norm": 0.16542522609233856,
      "learning_rate": 5.5791701168460765e-05,
      "loss": 0.6889,
      "mean_token_accuracy": 0.7914833687245846,
      "num_tokens": 460134376.0,
      "step": 14416
    },
    {
      "entropy": 0.6711187530308962,
      "epoch": 1.3264386058734567,
      "grad_norm": 0.16561399400234222,
      "learning_rate": 5.578863434231913e-05,
      "loss": 0.6823,
      "mean_token_accuracy": 0.7939467169344425,
      "num_tokens": 460166902.0,
      "step": 14417
    },
    {
      "entropy": 0.6828630566596985,
      "epoch": 1.3265306122448979,
      "grad_norm": 0.1723017394542694,
      "learning_rate": 5.578556751617751e-05,
      "loss": 0.7019,
      "mean_token_accuracy": 0.7918225638568401,
      "num_tokens": 460198293.0,
      "step": 14418
    },
    {
      "entropy": 0.6424890011548996,
      "epoch": 1.3266226186163392,
      "grad_norm": 0.16127683222293854,
      "learning_rate": 5.578250069003589e-05,
      "loss": 0.6088,
      "mean_token_accuracy": 0.8114796057343483,
      "num_tokens": 460229768.0,
      "step": 14419
    },
    {
      "entropy": 0.7425209414213896,
      "epoch": 1.3267146249877804,
      "grad_norm": 0.1649257093667984,
      "learning_rate": 5.5779433863894256e-05,
      "loss": 0.7327,
      "mean_token_accuracy": 0.7816731035709381,
      "num_tokens": 460261902.0,
      "step": 14420
    },
    {
      "entropy": 0.5408813608810306,
      "epoch": 1.3268066313592217,
      "grad_norm": 0.14734791219234467,
      "learning_rate": 5.577636703775263e-05,
      "loss": 0.5256,
      "mean_token_accuracy": 0.8374324180185795,
      "num_tokens": 460294054.0,
      "step": 14421
    },
    {
      "entropy": 0.6813992308452725,
      "epoch": 1.3268986377306629,
      "grad_norm": 0.1672595590353012,
      "learning_rate": 5.5773300211611e-05,
      "loss": 0.6692,
      "mean_token_accuracy": 0.7974129170179367,
      "num_tokens": 460325563.0,
      "step": 14422
    },
    {
      "entropy": 0.6582678570412099,
      "epoch": 1.326990644102104,
      "grad_norm": 0.16064846515655518,
      "learning_rate": 5.577023338546938e-05,
      "loss": 0.639,
      "mean_token_accuracy": 0.8024744465947151,
      "num_tokens": 460357250.0,
      "step": 14423
    },
    {
      "entropy": 0.6287343688309193,
      "epoch": 1.3270826504735453,
      "grad_norm": 0.15160737931728363,
      "learning_rate": 5.5767166559327755e-05,
      "loss": 0.6028,
      "mean_token_accuracy": 0.8114610202610493,
      "num_tokens": 460389020.0,
      "step": 14424
    },
    {
      "entropy": 0.698663083370775,
      "epoch": 1.3271746568449865,
      "grad_norm": 0.16202153265476227,
      "learning_rate": 5.576409973318612e-05,
      "loss": 0.6751,
      "mean_token_accuracy": 0.7973127290606499,
      "num_tokens": 460420262.0,
      "step": 14425
    },
    {
      "entropy": 0.5492636300623417,
      "epoch": 1.3272666632164278,
      "grad_norm": 0.15656423568725586,
      "learning_rate": 5.57610329070445e-05,
      "loss": 0.5248,
      "mean_token_accuracy": 0.8342520222067833,
      "num_tokens": 460451702.0,
      "step": 14426
    },
    {
      "entropy": 0.6750367991626263,
      "epoch": 1.327358669587869,
      "grad_norm": 0.1628343164920807,
      "learning_rate": 5.575796608090288e-05,
      "loss": 0.6623,
      "mean_token_accuracy": 0.7947282716631889,
      "num_tokens": 460483670.0,
      "step": 14427
    },
    {
      "entropy": 0.6993088088929653,
      "epoch": 1.32745067595931,
      "grad_norm": 0.16079898178577423,
      "learning_rate": 5.5754899254761254e-05,
      "loss": 0.6861,
      "mean_token_accuracy": 0.7942884564399719,
      "num_tokens": 460515689.0,
      "step": 14428
    },
    {
      "entropy": 0.6636185962706804,
      "epoch": 1.3275426823307515,
      "grad_norm": 0.16250838339328766,
      "learning_rate": 5.575183242861962e-05,
      "loss": 0.6581,
      "mean_token_accuracy": 0.8013267479836941,
      "num_tokens": 460547710.0,
      "step": 14429
    },
    {
      "entropy": 0.5961956726387143,
      "epoch": 1.3276346887021926,
      "grad_norm": 0.1525125950574875,
      "learning_rate": 5.5748765602478e-05,
      "loss": 0.5868,
      "mean_token_accuracy": 0.822100818157196,
      "num_tokens": 460580330.0,
      "step": 14430
    },
    {
      "entropy": 0.6171290073543787,
      "epoch": 1.327726695073634,
      "grad_norm": 0.16771191358566284,
      "learning_rate": 5.574569877633638e-05,
      "loss": 0.6104,
      "mean_token_accuracy": 0.812702689319849,
      "num_tokens": 460611284.0,
      "step": 14431
    },
    {
      "entropy": 0.5122120650485158,
      "epoch": 1.327818701445075,
      "grad_norm": 0.1463739275932312,
      "learning_rate": 5.5742631950194745e-05,
      "loss": 0.4882,
      "mean_token_accuracy": 0.84395382553339,
      "num_tokens": 460643003.0,
      "step": 14432
    },
    {
      "entropy": 0.606245880946517,
      "epoch": 1.3279107078165162,
      "grad_norm": 0.16034665703773499,
      "learning_rate": 5.573956512405312e-05,
      "loss": 0.6068,
      "mean_token_accuracy": 0.8176310174167156,
      "num_tokens": 460674499.0,
      "step": 14433
    },
    {
      "entropy": 0.6605525137856603,
      "epoch": 1.3280027141879576,
      "grad_norm": 0.16369497776031494,
      "learning_rate": 5.57364982979115e-05,
      "loss": 0.6528,
      "mean_token_accuracy": 0.8003305457532406,
      "num_tokens": 460706410.0,
      "step": 14434
    },
    {
      "entropy": 0.5766303054988384,
      "epoch": 1.3280947205593987,
      "grad_norm": 0.15774571895599365,
      "learning_rate": 5.573343147176987e-05,
      "loss": 0.5654,
      "mean_token_accuracy": 0.825241357088089,
      "num_tokens": 460737765.0,
      "step": 14435
    },
    {
      "entropy": 0.5641635740175843,
      "epoch": 1.32818672693084,
      "grad_norm": 0.15275906026363373,
      "learning_rate": 5.5730364645628244e-05,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8334121853113174,
      "num_tokens": 460769336.0,
      "step": 14436
    },
    {
      "entropy": 0.6066486574709415,
      "epoch": 1.3282787333022812,
      "grad_norm": 0.15734966099262238,
      "learning_rate": 5.572729781948661e-05,
      "loss": 0.5997,
      "mean_token_accuracy": 0.8189746923744678,
      "num_tokens": 460801056.0,
      "step": 14437
    },
    {
      "entropy": 0.7099963333457708,
      "epoch": 1.3283707396737223,
      "grad_norm": 0.16541391611099243,
      "learning_rate": 5.572423099334499e-05,
      "loss": 0.7002,
      "mean_token_accuracy": 0.7862544395029545,
      "num_tokens": 460832964.0,
      "step": 14438
    },
    {
      "entropy": 0.7050770437344909,
      "epoch": 1.3284627460451637,
      "grad_norm": 0.17275448143482208,
      "learning_rate": 5.572116416720337e-05,
      "loss": 0.6909,
      "mean_token_accuracy": 0.7922174781560898,
      "num_tokens": 460864830.0,
      "step": 14439
    },
    {
      "entropy": 0.6921308394521475,
      "epoch": 1.3285547524166048,
      "grad_norm": 0.16066651046276093,
      "learning_rate": 5.5718097341061736e-05,
      "loss": 0.6867,
      "mean_token_accuracy": 0.7881753630936146,
      "num_tokens": 460896298.0,
      "step": 14440
    },
    {
      "entropy": 0.6267073191702366,
      "epoch": 1.3286467587880462,
      "grad_norm": 0.1631920337677002,
      "learning_rate": 5.571503051492011e-05,
      "loss": 0.6248,
      "mean_token_accuracy": 0.8081871867179871,
      "num_tokens": 460928396.0,
      "step": 14441
    },
    {
      "entropy": 0.5448039937764406,
      "epoch": 1.3287387651594873,
      "grad_norm": 0.15291529893875122,
      "learning_rate": 5.571196368877849e-05,
      "loss": 0.5254,
      "mean_token_accuracy": 0.8400403000414371,
      "num_tokens": 460961164.0,
      "step": 14442
    },
    {
      "entropy": 0.49508774280548096,
      "epoch": 1.3288307715309284,
      "grad_norm": 0.1550692468881607,
      "learning_rate": 5.570889686263686e-05,
      "loss": 0.4773,
      "mean_token_accuracy": 0.8536647036671638,
      "num_tokens": 460993392.0,
      "step": 14443
    },
    {
      "entropy": 0.6137406472116709,
      "epoch": 1.3289227779023698,
      "grad_norm": 0.15946398675441742,
      "learning_rate": 5.5705830036495234e-05,
      "loss": 0.6035,
      "mean_token_accuracy": 0.8140948414802551,
      "num_tokens": 461025310.0,
      "step": 14444
    },
    {
      "entropy": 0.5784679409116507,
      "epoch": 1.329014784273811,
      "grad_norm": 0.1549598127603531,
      "learning_rate": 5.57027632103536e-05,
      "loss": 0.563,
      "mean_token_accuracy": 0.8233480341732502,
      "num_tokens": 461056879.0,
      "step": 14445
    },
    {
      "entropy": 0.6583687299862504,
      "epoch": 1.3291067906452523,
      "grad_norm": 0.16308079659938812,
      "learning_rate": 5.5699696384211984e-05,
      "loss": 0.6408,
      "mean_token_accuracy": 0.8037627674639225,
      "num_tokens": 461088514.0,
      "step": 14446
    },
    {
      "entropy": 0.5869943741708994,
      "epoch": 1.3291987970166934,
      "grad_norm": 0.1546449363231659,
      "learning_rate": 5.569662955807036e-05,
      "loss": 0.5719,
      "mean_token_accuracy": 0.8244005143642426,
      "num_tokens": 461120312.0,
      "step": 14447
    },
    {
      "entropy": 0.6684396136552095,
      "epoch": 1.3292908033881345,
      "grad_norm": 0.1663816124200821,
      "learning_rate": 5.5693562731928726e-05,
      "loss": 0.6557,
      "mean_token_accuracy": 0.8016509413719177,
      "num_tokens": 461152085.0,
      "step": 14448
    },
    {
      "entropy": 0.6190246804617345,
      "epoch": 1.329382809759576,
      "grad_norm": 0.1533937156200409,
      "learning_rate": 5.56904959057871e-05,
      "loss": 0.5893,
      "mean_token_accuracy": 0.8177538476884365,
      "num_tokens": 461183659.0,
      "step": 14449
    },
    {
      "entropy": 0.7259950041770935,
      "epoch": 1.329474816131017,
      "grad_norm": 0.1707986295223236,
      "learning_rate": 5.568742907964548e-05,
      "loss": 0.7187,
      "mean_token_accuracy": 0.7821751162409782,
      "num_tokens": 461214563.0,
      "step": 14450
    },
    {
      "entropy": 0.5722694611176848,
      "epoch": 1.3295668225024584,
      "grad_norm": 0.16510488092899323,
      "learning_rate": 5.568436225350385e-05,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8256399482488632,
      "num_tokens": 461246006.0,
      "step": 14451
    },
    {
      "entropy": 0.682668985798955,
      "epoch": 1.3296588288738995,
      "grad_norm": 0.15967994928359985,
      "learning_rate": 5.5681295427362225e-05,
      "loss": 0.6856,
      "mean_token_accuracy": 0.7885411940515041,
      "num_tokens": 461277601.0,
      "step": 14452
    },
    {
      "entropy": 0.5146767194382846,
      "epoch": 1.3297508352453407,
      "grad_norm": 0.1451113075017929,
      "learning_rate": 5.567822860122059e-05,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8471417762339115,
      "num_tokens": 461309730.0,
      "step": 14453
    },
    {
      "entropy": 0.6888346839696169,
      "epoch": 1.329842841616782,
      "grad_norm": 0.1620924472808838,
      "learning_rate": 5.5675161775078974e-05,
      "loss": 0.6742,
      "mean_token_accuracy": 0.7924071662127972,
      "num_tokens": 461342121.0,
      "step": 14454
    },
    {
      "entropy": 0.619576083496213,
      "epoch": 1.3299348479882231,
      "grad_norm": 0.1688697785139084,
      "learning_rate": 5.567209494893735e-05,
      "loss": 0.6037,
      "mean_token_accuracy": 0.8161693774163723,
      "num_tokens": 461372914.0,
      "step": 14455
    },
    {
      "entropy": 0.5211191773414612,
      "epoch": 1.3300268543596645,
      "grad_norm": 0.1567046344280243,
      "learning_rate": 5.5669028122795717e-05,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8412358090281487,
      "num_tokens": 461404211.0,
      "step": 14456
    },
    {
      "entropy": 0.7144050598144531,
      "epoch": 1.3301188607311056,
      "grad_norm": 0.16884495317935944,
      "learning_rate": 5.566596129665409e-05,
      "loss": 0.7013,
      "mean_token_accuracy": 0.7897385954856873,
      "num_tokens": 461435741.0,
      "step": 14457
    },
    {
      "entropy": 0.7168893674388528,
      "epoch": 1.3302108671025468,
      "grad_norm": 0.15991592407226562,
      "learning_rate": 5.566289447051247e-05,
      "loss": 0.709,
      "mean_token_accuracy": 0.7851334437727928,
      "num_tokens": 461467869.0,
      "step": 14458
    },
    {
      "entropy": 0.7603101190179586,
      "epoch": 1.3303028734739881,
      "grad_norm": 0.16614781320095062,
      "learning_rate": 5.565982764437084e-05,
      "loss": 0.7531,
      "mean_token_accuracy": 0.7684352956712246,
      "num_tokens": 461499981.0,
      "step": 14459
    },
    {
      "entropy": 0.5894179432652891,
      "epoch": 1.3303948798454293,
      "grad_norm": 0.1519666314125061,
      "learning_rate": 5.5656760818229215e-05,
      "loss": 0.5724,
      "mean_token_accuracy": 0.8249103128910065,
      "num_tokens": 461531146.0,
      "step": 14460
    },
    {
      "entropy": 0.5530823022127151,
      "epoch": 1.3304868862168706,
      "grad_norm": 0.1586655229330063,
      "learning_rate": 5.5653693992087597e-05,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8340612575411797,
      "num_tokens": 461563319.0,
      "step": 14461
    },
    {
      "entropy": 0.6339332554489374,
      "epoch": 1.3305788925883117,
      "grad_norm": 0.1618899703025818,
      "learning_rate": 5.5650627165945964e-05,
      "loss": 0.6243,
      "mean_token_accuracy": 0.8109728693962097,
      "num_tokens": 461594923.0,
      "step": 14462
    },
    {
      "entropy": 0.5953389769420028,
      "epoch": 1.3306708989597529,
      "grad_norm": 0.16330543160438538,
      "learning_rate": 5.564756033980434e-05,
      "loss": 0.5926,
      "mean_token_accuracy": 0.8188716545701027,
      "num_tokens": 461627394.0,
      "step": 14463
    },
    {
      "entropy": 0.5861460650339723,
      "epoch": 1.3307629053311942,
      "grad_norm": 0.159479558467865,
      "learning_rate": 5.564449351366271e-05,
      "loss": 0.5752,
      "mean_token_accuracy": 0.8190908171236515,
      "num_tokens": 461659218.0,
      "step": 14464
    },
    {
      "entropy": 0.6727364799007773,
      "epoch": 1.3308549117026354,
      "grad_norm": 0.16203093528747559,
      "learning_rate": 5.5641426687521095e-05,
      "loss": 0.6702,
      "mean_token_accuracy": 0.7954865097999573,
      "num_tokens": 461690454.0,
      "step": 14465
    },
    {
      "entropy": 0.6202947106212378,
      "epoch": 1.3309469180740767,
      "grad_norm": 0.1587403118610382,
      "learning_rate": 5.563835986137946e-05,
      "loss": 0.6128,
      "mean_token_accuracy": 0.8149010948836803,
      "num_tokens": 461722453.0,
      "step": 14466
    },
    {
      "entropy": 0.5164048979058862,
      "epoch": 1.3310389244455179,
      "grad_norm": 0.15850290656089783,
      "learning_rate": 5.563529303523783e-05,
      "loss": 0.49,
      "mean_token_accuracy": 0.8505431041121483,
      "num_tokens": 461754217.0,
      "step": 14467
    },
    {
      "entropy": 0.5119165503419936,
      "epoch": 1.331130930816959,
      "grad_norm": 0.15416565537452698,
      "learning_rate": 5.5632226209096206e-05,
      "loss": 0.5103,
      "mean_token_accuracy": 0.8415441438555717,
      "num_tokens": 461785822.0,
      "step": 14468
    },
    {
      "entropy": 0.5909610288217664,
      "epoch": 1.3312229371884003,
      "grad_norm": 0.15963536500930786,
      "learning_rate": 5.562915938295459e-05,
      "loss": 0.572,
      "mean_token_accuracy": 0.8245884254574776,
      "num_tokens": 461817049.0,
      "step": 14469
    },
    {
      "entropy": 0.661874046549201,
      "epoch": 1.3313149435598415,
      "grad_norm": 0.16014961898326874,
      "learning_rate": 5.562609255681296e-05,
      "loss": 0.657,
      "mean_token_accuracy": 0.7991653829813004,
      "num_tokens": 461849442.0,
      "step": 14470
    },
    {
      "entropy": 0.6673479964956641,
      "epoch": 1.3314069499312828,
      "grad_norm": 0.17080937325954437,
      "learning_rate": 5.562302573067133e-05,
      "loss": 0.6666,
      "mean_token_accuracy": 0.80052350461483,
      "num_tokens": 461881361.0,
      "step": 14471
    },
    {
      "entropy": 0.636769587174058,
      "epoch": 1.331498956302724,
      "grad_norm": 0.16338513791561127,
      "learning_rate": 5.5619958904529704e-05,
      "loss": 0.6128,
      "mean_token_accuracy": 0.8123266994953156,
      "num_tokens": 461913576.0,
      "step": 14472
    },
    {
      "entropy": 0.5188631080091,
      "epoch": 1.331590962674165,
      "grad_norm": 0.1441473811864853,
      "learning_rate": 5.5616892078388086e-05,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8402199223637581,
      "num_tokens": 461945589.0,
      "step": 14473
    },
    {
      "entropy": 0.5488900141790509,
      "epoch": 1.3316829690456065,
      "grad_norm": 0.15053440630435944,
      "learning_rate": 5.5613825252246453e-05,
      "loss": 0.5307,
      "mean_token_accuracy": 0.8358579874038696,
      "num_tokens": 461977728.0,
      "step": 14474
    },
    {
      "entropy": 0.6116762049496174,
      "epoch": 1.3317749754170476,
      "grad_norm": 0.16158118844032288,
      "learning_rate": 5.561075842610483e-05,
      "loss": 0.5945,
      "mean_token_accuracy": 0.8190388828516006,
      "num_tokens": 462008295.0,
      "step": 14475
    },
    {
      "entropy": 0.5926147224381566,
      "epoch": 1.331866981788489,
      "grad_norm": 0.15017668902873993,
      "learning_rate": 5.5607691599963196e-05,
      "loss": 0.5733,
      "mean_token_accuracy": 0.8228863961994648,
      "num_tokens": 462040051.0,
      "step": 14476
    },
    {
      "entropy": 0.6285447813570499,
      "epoch": 1.33195898815993,
      "grad_norm": 0.16286659240722656,
      "learning_rate": 5.560462477382158e-05,
      "loss": 0.6154,
      "mean_token_accuracy": 0.8122457712888718,
      "num_tokens": 462071854.0,
      "step": 14477
    },
    {
      "entropy": 0.673042168840766,
      "epoch": 1.3320509945313712,
      "grad_norm": 0.16430135071277618,
      "learning_rate": 5.560155794767995e-05,
      "loss": 0.6706,
      "mean_token_accuracy": 0.7964512668550014,
      "num_tokens": 462102893.0,
      "step": 14478
    },
    {
      "entropy": 0.7137251403182745,
      "epoch": 1.3321430009028126,
      "grad_norm": 0.16105203330516815,
      "learning_rate": 5.559849112153832e-05,
      "loss": 0.7104,
      "mean_token_accuracy": 0.7857039533555508,
      "num_tokens": 462134872.0,
      "step": 14479
    },
    {
      "entropy": 0.5134280119091272,
      "epoch": 1.3322350072742537,
      "grad_norm": 0.15558157861232758,
      "learning_rate": 5.5595424295396695e-05,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8460186421871185,
      "num_tokens": 462167488.0,
      "step": 14480
    },
    {
      "entropy": 0.6328403512015939,
      "epoch": 1.332327013645695,
      "grad_norm": 0.1560143381357193,
      "learning_rate": 5.5592357469255076e-05,
      "loss": 0.6078,
      "mean_token_accuracy": 0.8139887638390064,
      "num_tokens": 462199186.0,
      "step": 14481
    },
    {
      "entropy": 0.5997415482997894,
      "epoch": 1.3324190200171362,
      "grad_norm": 0.1588798612356186,
      "learning_rate": 5.5589290643113444e-05,
      "loss": 0.5847,
      "mean_token_accuracy": 0.8238187246024609,
      "num_tokens": 462230773.0,
      "step": 14482
    },
    {
      "entropy": 0.6673230361193419,
      "epoch": 1.3325110263885773,
      "grad_norm": 0.15898524224758148,
      "learning_rate": 5.558622381697182e-05,
      "loss": 0.6626,
      "mean_token_accuracy": 0.7983726672828197,
      "num_tokens": 462263106.0,
      "step": 14483
    },
    {
      "entropy": 0.5720591540448368,
      "epoch": 1.3326030327600187,
      "grad_norm": 0.1622304469347,
      "learning_rate": 5.5583156990830186e-05,
      "loss": 0.5555,
      "mean_token_accuracy": 0.8272851929068565,
      "num_tokens": 462295216.0,
      "step": 14484
    },
    {
      "entropy": 0.5955195524729788,
      "epoch": 1.3326950391314598,
      "grad_norm": 0.1541413962841034,
      "learning_rate": 5.558009016468857e-05,
      "loss": 0.5911,
      "mean_token_accuracy": 0.8196911029517651,
      "num_tokens": 462327904.0,
      "step": 14485
    },
    {
      "entropy": 0.6134832929819822,
      "epoch": 1.3327870455029012,
      "grad_norm": 0.16127507388591766,
      "learning_rate": 5.557702333854694e-05,
      "loss": 0.6064,
      "mean_token_accuracy": 0.8137442022562027,
      "num_tokens": 462359619.0,
      "step": 14486
    },
    {
      "entropy": 0.6843208223581314,
      "epoch": 1.3328790518743423,
      "grad_norm": 0.17040786147117615,
      "learning_rate": 5.557395651240531e-05,
      "loss": 0.6765,
      "mean_token_accuracy": 0.7969470173120499,
      "num_tokens": 462391108.0,
      "step": 14487
    },
    {
      "entropy": 0.6187680028378963,
      "epoch": 1.3329710582457834,
      "grad_norm": 0.1600627601146698,
      "learning_rate": 5.5570889686263685e-05,
      "loss": 0.615,
      "mean_token_accuracy": 0.8143176846206188,
      "num_tokens": 462422892.0,
      "step": 14488
    },
    {
      "entropy": 0.6752285901457071,
      "epoch": 1.3330630646172248,
      "grad_norm": 0.15730254352092743,
      "learning_rate": 5.5567822860122066e-05,
      "loss": 0.6727,
      "mean_token_accuracy": 0.7963389903306961,
      "num_tokens": 462454111.0,
      "step": 14489
    },
    {
      "entropy": 0.5532999541610479,
      "epoch": 1.333155070988666,
      "grad_norm": 0.15004472434520721,
      "learning_rate": 5.5564756033980434e-05,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8370266258716583,
      "num_tokens": 462486104.0,
      "step": 14490
    },
    {
      "entropy": 0.6694894647225738,
      "epoch": 1.3332470773601073,
      "grad_norm": 0.158040851354599,
      "learning_rate": 5.556168920783881e-05,
      "loss": 0.6602,
      "mean_token_accuracy": 0.8000914789736271,
      "num_tokens": 462517349.0,
      "step": 14491
    },
    {
      "entropy": 0.6493754778057337,
      "epoch": 1.3333390837315484,
      "grad_norm": 0.1536777913570404,
      "learning_rate": 5.555862238169719e-05,
      "loss": 0.6437,
      "mean_token_accuracy": 0.8007073178887367,
      "num_tokens": 462548858.0,
      "step": 14492
    },
    {
      "entropy": 0.6397463399916887,
      "epoch": 1.3334310901029895,
      "grad_norm": 0.15449485182762146,
      "learning_rate": 5.555555555555556e-05,
      "loss": 0.6195,
      "mean_token_accuracy": 0.8074741996824741,
      "num_tokens": 462580817.0,
      "step": 14493
    },
    {
      "entropy": 0.7836001552641392,
      "epoch": 1.333523096474431,
      "grad_norm": 0.17196066677570343,
      "learning_rate": 5.555248872941393e-05,
      "loss": 0.7719,
      "mean_token_accuracy": 0.7697333246469498,
      "num_tokens": 462612728.0,
      "step": 14494
    },
    {
      "entropy": 0.6858805567026138,
      "epoch": 1.333615102845872,
      "grad_norm": 0.16154010593891144,
      "learning_rate": 5.55494219032723e-05,
      "loss": 0.6779,
      "mean_token_accuracy": 0.7937280237674713,
      "num_tokens": 462644760.0,
      "step": 14495
    },
    {
      "entropy": 0.6095299478620291,
      "epoch": 1.3337071092173134,
      "grad_norm": 0.15858638286590576,
      "learning_rate": 5.554635507713068e-05,
      "loss": 0.602,
      "mean_token_accuracy": 0.8176052868366241,
      "num_tokens": 462676372.0,
      "step": 14496
    },
    {
      "entropy": 0.6711037252098322,
      "epoch": 1.3337991155887545,
      "grad_norm": 0.15971364080905914,
      "learning_rate": 5.554328825098906e-05,
      "loss": 0.6585,
      "mean_token_accuracy": 0.800826471298933,
      "num_tokens": 462708375.0,
      "step": 14497
    },
    {
      "entropy": 0.6848444733768702,
      "epoch": 1.3338911219601957,
      "grad_norm": 0.17365069687366486,
      "learning_rate": 5.5540221424847425e-05,
      "loss": 0.6848,
      "mean_token_accuracy": 0.7917322292923927,
      "num_tokens": 462740233.0,
      "step": 14498
    },
    {
      "entropy": 0.6031951326876879,
      "epoch": 1.333983128331637,
      "grad_norm": 0.16421687602996826,
      "learning_rate": 5.55371545987058e-05,
      "loss": 0.5922,
      "mean_token_accuracy": 0.8190814070403576,
      "num_tokens": 462771448.0,
      "step": 14499
    },
    {
      "entropy": 0.5894454447552562,
      "epoch": 1.3340751347030781,
      "grad_norm": 0.1540152132511139,
      "learning_rate": 5.553408777256418e-05,
      "loss": 0.594,
      "mean_token_accuracy": 0.8181961216032505,
      "num_tokens": 462803574.0,
      "step": 14500
    },
    {
      "entropy": 0.6360881039872766,
      "epoch": 1.3341671410745195,
      "grad_norm": 0.15049052238464355,
      "learning_rate": 5.553102094642255e-05,
      "loss": 0.6126,
      "mean_token_accuracy": 0.8137758485972881,
      "num_tokens": 462835957.0,
      "step": 14501
    },
    {
      "entropy": 0.6290986426174641,
      "epoch": 1.3342591474459606,
      "grad_norm": 0.16181840002536774,
      "learning_rate": 5.552795412028092e-05,
      "loss": 0.6083,
      "mean_token_accuracy": 0.8145762756466866,
      "num_tokens": 462867030.0,
      "step": 14502
    },
    {
      "entropy": 0.6494290381669998,
      "epoch": 1.3343511538174018,
      "grad_norm": 0.15928548574447632,
      "learning_rate": 5.552488729413929e-05,
      "loss": 0.6466,
      "mean_token_accuracy": 0.8031244352459908,
      "num_tokens": 462899279.0,
      "step": 14503
    },
    {
      "entropy": 0.7369049284607172,
      "epoch": 1.3344431601888431,
      "grad_norm": 0.1641530841588974,
      "learning_rate": 5.552182046799768e-05,
      "loss": 0.7149,
      "mean_token_accuracy": 0.782608225941658,
      "num_tokens": 462930603.0,
      "step": 14504
    },
    {
      "entropy": 0.6967377532273531,
      "epoch": 1.3345351665602843,
      "grad_norm": 0.1608887016773224,
      "learning_rate": 5.551875364185605e-05,
      "loss": 0.6844,
      "mean_token_accuracy": 0.7963503561913967,
      "num_tokens": 462962545.0,
      "step": 14505
    },
    {
      "entropy": 0.6717738704755902,
      "epoch": 1.3346271729317256,
      "grad_norm": 0.15919406712055206,
      "learning_rate": 5.5515686815714415e-05,
      "loss": 0.6551,
      "mean_token_accuracy": 0.799633365124464,
      "num_tokens": 462995016.0,
      "step": 14506
    },
    {
      "entropy": 0.641620023176074,
      "epoch": 1.3347191793031667,
      "grad_norm": 0.1620682179927826,
      "learning_rate": 5.551261998957279e-05,
      "loss": 0.623,
      "mean_token_accuracy": 0.8119829371571541,
      "num_tokens": 463026194.0,
      "step": 14507
    },
    {
      "entropy": 0.6757025634869933,
      "epoch": 1.3348111856746079,
      "grad_norm": 0.15948815643787384,
      "learning_rate": 5.550955316343117e-05,
      "loss": 0.6592,
      "mean_token_accuracy": 0.7937206476926804,
      "num_tokens": 463057909.0,
      "step": 14508
    },
    {
      "entropy": 0.6299752052873373,
      "epoch": 1.3349031920460492,
      "grad_norm": 0.16082598268985748,
      "learning_rate": 5.5506486337289546e-05,
      "loss": 0.6163,
      "mean_token_accuracy": 0.8121503740549088,
      "num_tokens": 463089979.0,
      "step": 14509
    },
    {
      "entropy": 0.6402903189882636,
      "epoch": 1.3349951984174904,
      "grad_norm": 0.1633899062871933,
      "learning_rate": 5.5503419511147914e-05,
      "loss": 0.6217,
      "mean_token_accuracy": 0.81116558983922,
      "num_tokens": 463120643.0,
      "step": 14510
    },
    {
      "entropy": 0.6180581916123629,
      "epoch": 1.3350872047889317,
      "grad_norm": 0.15351803600788116,
      "learning_rate": 5.550035268500628e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8081353195011616,
      "num_tokens": 463153033.0,
      "step": 14511
    },
    {
      "entropy": 0.7047320483252406,
      "epoch": 1.3351792111603729,
      "grad_norm": 0.16007296741008759,
      "learning_rate": 5.549728585886467e-05,
      "loss": 0.6994,
      "mean_token_accuracy": 0.7835200503468513,
      "num_tokens": 463184799.0,
      "step": 14512
    },
    {
      "entropy": 0.6117585063911974,
      "epoch": 1.335271217531814,
      "grad_norm": 0.1537911295890808,
      "learning_rate": 5.549421903272304e-05,
      "loss": 0.594,
      "mean_token_accuracy": 0.817797165364027,
      "num_tokens": 463216693.0,
      "step": 14513
    },
    {
      "entropy": 0.6530936295166612,
      "epoch": 1.3353632239032553,
      "grad_norm": 0.1602376103401184,
      "learning_rate": 5.549115220658141e-05,
      "loss": 0.6341,
      "mean_token_accuracy": 0.8057854436337948,
      "num_tokens": 463247809.0,
      "step": 14514
    },
    {
      "entropy": 0.6446726247668266,
      "epoch": 1.3354552302746965,
      "grad_norm": 0.16163085401058197,
      "learning_rate": 5.548808538043978e-05,
      "loss": 0.6315,
      "mean_token_accuracy": 0.8063433468341827,
      "num_tokens": 463279172.0,
      "step": 14515
    },
    {
      "entropy": 0.6729410933330655,
      "epoch": 1.3355472366461378,
      "grad_norm": 0.1781492680311203,
      "learning_rate": 5.548501855429816e-05,
      "loss": 0.6783,
      "mean_token_accuracy": 0.7986119315028191,
      "num_tokens": 463310922.0,
      "step": 14516
    },
    {
      "entropy": 0.6343136439099908,
      "epoch": 1.335639243017579,
      "grad_norm": 0.1680612862110138,
      "learning_rate": 5.5481951728156536e-05,
      "loss": 0.6253,
      "mean_token_accuracy": 0.8129147104918957,
      "num_tokens": 463343263.0,
      "step": 14517
    },
    {
      "entropy": 0.6402264758944511,
      "epoch": 1.33573124938902,
      "grad_norm": 0.16779464483261108,
      "learning_rate": 5.5478884902014904e-05,
      "loss": 0.6459,
      "mean_token_accuracy": 0.8091515935957432,
      "num_tokens": 463374427.0,
      "step": 14518
    },
    {
      "entropy": 0.5483184857293963,
      "epoch": 1.3358232557604615,
      "grad_norm": 0.14806325733661652,
      "learning_rate": 5.547581807587328e-05,
      "loss": 0.5456,
      "mean_token_accuracy": 0.8322589732706547,
      "num_tokens": 463406788.0,
      "step": 14519
    },
    {
      "entropy": 0.6829013302922249,
      "epoch": 1.3359152621319026,
      "grad_norm": 0.16870738565921783,
      "learning_rate": 5.547275124973166e-05,
      "loss": 0.6737,
      "mean_token_accuracy": 0.7906236983835697,
      "num_tokens": 463436750.0,
      "step": 14520
    },
    {
      "entropy": 0.6418870575726032,
      "epoch": 1.336007268503344,
      "grad_norm": 0.15791675448417664,
      "learning_rate": 5.546968442359003e-05,
      "loss": 0.6388,
      "mean_token_accuracy": 0.8005970530211926,
      "num_tokens": 463468146.0,
      "step": 14521
    },
    {
      "entropy": 0.5992852225899696,
      "epoch": 1.336099274874785,
      "grad_norm": 0.157169371843338,
      "learning_rate": 5.54666175974484e-05,
      "loss": 0.5903,
      "mean_token_accuracy": 0.8155548833310604,
      "num_tokens": 463499757.0,
      "step": 14522
    },
    {
      "entropy": 0.6656354637816548,
      "epoch": 1.3361912812462262,
      "grad_norm": 0.1596848964691162,
      "learning_rate": 5.5463550771306784e-05,
      "loss": 0.6571,
      "mean_token_accuracy": 0.7979285418987274,
      "num_tokens": 463531572.0,
      "step": 14523
    },
    {
      "entropy": 0.6332176476716995,
      "epoch": 1.3362832876176676,
      "grad_norm": 0.15477055311203003,
      "learning_rate": 5.546048394516515e-05,
      "loss": 0.6181,
      "mean_token_accuracy": 0.8092262297868729,
      "num_tokens": 463562994.0,
      "step": 14524
    },
    {
      "entropy": 0.6398006700910628,
      "epoch": 1.3363752939891087,
      "grad_norm": 0.15768498182296753,
      "learning_rate": 5.5457417119023527e-05,
      "loss": 0.6197,
      "mean_token_accuracy": 0.808952983468771,
      "num_tokens": 463594147.0,
      "step": 14525
    },
    {
      "entropy": 0.6252258755266666,
      "epoch": 1.33646730036055,
      "grad_norm": 0.15696139633655548,
      "learning_rate": 5.5454350292881894e-05,
      "loss": 0.5962,
      "mean_token_accuracy": 0.8203690126538277,
      "num_tokens": 463624778.0,
      "step": 14526
    },
    {
      "entropy": 0.6572722587734461,
      "epoch": 1.3365593067319912,
      "grad_norm": 0.1586192399263382,
      "learning_rate": 5.5451283466740276e-05,
      "loss": 0.6371,
      "mean_token_accuracy": 0.8048565573990345,
      "num_tokens": 463656439.0,
      "step": 14527
    },
    {
      "entropy": 0.5788278393447399,
      "epoch": 1.3366513131034323,
      "grad_norm": 0.1538158655166626,
      "learning_rate": 5.544821664059865e-05,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8302943371236324,
      "num_tokens": 463687365.0,
      "step": 14528
    },
    {
      "entropy": 0.5391317801550031,
      "epoch": 1.3367433194748737,
      "grad_norm": 0.15243756771087646,
      "learning_rate": 5.544514981445702e-05,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8379557803273201,
      "num_tokens": 463718409.0,
      "step": 14529
    },
    {
      "entropy": 0.5386734805069864,
      "epoch": 1.3368353258463148,
      "grad_norm": 0.15356312692165375,
      "learning_rate": 5.544208298831539e-05,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8402892760932446,
      "num_tokens": 463750080.0,
      "step": 14530
    },
    {
      "entropy": 0.6117658410221338,
      "epoch": 1.3369273322177562,
      "grad_norm": 0.1511985808610916,
      "learning_rate": 5.5439016162173774e-05,
      "loss": 0.5991,
      "mean_token_accuracy": 0.8136431090533733,
      "num_tokens": 463781321.0,
      "step": 14531
    },
    {
      "entropy": 0.5996205396950245,
      "epoch": 1.3370193385891973,
      "grad_norm": 0.1517624706029892,
      "learning_rate": 5.543594933603214e-05,
      "loss": 0.5971,
      "mean_token_accuracy": 0.8153595328330994,
      "num_tokens": 463812882.0,
      "step": 14532
    },
    {
      "entropy": 0.6680297162383795,
      "epoch": 1.3371113449606384,
      "grad_norm": 0.1574474424123764,
      "learning_rate": 5.543288250989052e-05,
      "loss": 0.6467,
      "mean_token_accuracy": 0.8033435344696045,
      "num_tokens": 463845098.0,
      "step": 14533
    },
    {
      "entropy": 0.6684643309563398,
      "epoch": 1.3372033513320798,
      "grad_norm": 0.15636394917964935,
      "learning_rate": 5.5429815683748885e-05,
      "loss": 0.6522,
      "mean_token_accuracy": 0.8011781759560108,
      "num_tokens": 463876258.0,
      "step": 14534
    },
    {
      "entropy": 0.5173950409516692,
      "epoch": 1.337295357703521,
      "grad_norm": 0.15349243581295013,
      "learning_rate": 5.5426748857607266e-05,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8405390381813049,
      "num_tokens": 463908342.0,
      "step": 14535
    },
    {
      "entropy": 0.7038077376782894,
      "epoch": 1.3373873640749623,
      "grad_norm": 0.1577114760875702,
      "learning_rate": 5.542368203146564e-05,
      "loss": 0.6935,
      "mean_token_accuracy": 0.7887147478759289,
      "num_tokens": 463940903.0,
      "step": 14536
    },
    {
      "entropy": 0.499802484177053,
      "epoch": 1.3374793704464034,
      "grad_norm": 0.14055611193180084,
      "learning_rate": 5.542061520532401e-05,
      "loss": 0.5013,
      "mean_token_accuracy": 0.8412703387439251,
      "num_tokens": 463973393.0,
      "step": 14537
    },
    {
      "entropy": 0.5999884186312556,
      "epoch": 1.3375713768178445,
      "grad_norm": 0.15264470875263214,
      "learning_rate": 5.5417548379182383e-05,
      "loss": 0.5972,
      "mean_token_accuracy": 0.8167230114340782,
      "num_tokens": 464005316.0,
      "step": 14538
    },
    {
      "entropy": 0.5919666290283203,
      "epoch": 1.337663383189286,
      "grad_norm": 0.158732071518898,
      "learning_rate": 5.5414481553040765e-05,
      "loss": 0.5811,
      "mean_token_accuracy": 0.8177611492574215,
      "num_tokens": 464037118.0,
      "step": 14539
    },
    {
      "entropy": 0.648607874289155,
      "epoch": 1.337755389560727,
      "grad_norm": 0.15741997957229614,
      "learning_rate": 5.541141472689913e-05,
      "loss": 0.6554,
      "mean_token_accuracy": 0.8007791414856911,
      "num_tokens": 464069384.0,
      "step": 14540
    },
    {
      "entropy": 0.5642108004540205,
      "epoch": 1.3378473959321684,
      "grad_norm": 0.1607840657234192,
      "learning_rate": 5.540834790075751e-05,
      "loss": 0.5618,
      "mean_token_accuracy": 0.8280154317617416,
      "num_tokens": 464099533.0,
      "step": 14541
    },
    {
      "entropy": 0.6239833664149046,
      "epoch": 1.3379394023036095,
      "grad_norm": 0.15763011574745178,
      "learning_rate": 5.5405281074615875e-05,
      "loss": 0.6295,
      "mean_token_accuracy": 0.8074835054576397,
      "num_tokens": 464131699.0,
      "step": 14542
    },
    {
      "entropy": 0.6520706377923489,
      "epoch": 1.3380314086750507,
      "grad_norm": 0.15913838148117065,
      "learning_rate": 5.5402214248474263e-05,
      "loss": 0.66,
      "mean_token_accuracy": 0.8002638109028339,
      "num_tokens": 464163897.0,
      "step": 14543
    },
    {
      "entropy": 0.5036133551038802,
      "epoch": 1.338123415046492,
      "grad_norm": 0.1403285264968872,
      "learning_rate": 5.539914742233263e-05,
      "loss": 0.4967,
      "mean_token_accuracy": 0.8445331081748009,
      "num_tokens": 464196116.0,
      "step": 14544
    },
    {
      "entropy": 0.5818773526698351,
      "epoch": 1.3382154214179331,
      "grad_norm": 0.1541047990322113,
      "learning_rate": 5.5396080596191e-05,
      "loss": 0.573,
      "mean_token_accuracy": 0.8213014826178551,
      "num_tokens": 464228418.0,
      "step": 14545
    },
    {
      "entropy": 0.649690642952919,
      "epoch": 1.3383074277893745,
      "grad_norm": 0.16517971456050873,
      "learning_rate": 5.5393013770049374e-05,
      "loss": 0.644,
      "mean_token_accuracy": 0.8022758550941944,
      "num_tokens": 464259943.0,
      "step": 14546
    },
    {
      "entropy": 0.6043940568342805,
      "epoch": 1.3383994341608156,
      "grad_norm": 0.15949736535549164,
      "learning_rate": 5.5389946943907755e-05,
      "loss": 0.5966,
      "mean_token_accuracy": 0.8135321512818336,
      "num_tokens": 464291394.0,
      "step": 14547
    },
    {
      "entropy": 0.7651250809431076,
      "epoch": 1.3384914405322568,
      "grad_norm": 0.16495446860790253,
      "learning_rate": 5.538688011776613e-05,
      "loss": 0.7792,
      "mean_token_accuracy": 0.7669337466359138,
      "num_tokens": 464323773.0,
      "step": 14548
    },
    {
      "entropy": 0.4834187775850296,
      "epoch": 1.3385834469036981,
      "grad_norm": 0.15762147307395935,
      "learning_rate": 5.53838132916245e-05,
      "loss": 0.4679,
      "mean_token_accuracy": 0.8582226559519768,
      "num_tokens": 464355570.0,
      "step": 14549
    },
    {
      "entropy": 0.6871478129178286,
      "epoch": 1.3386754532751393,
      "grad_norm": 0.16605214774608612,
      "learning_rate": 5.5380746465482866e-05,
      "loss": 0.6633,
      "mean_token_accuracy": 0.7966687977313995,
      "num_tokens": 464387691.0,
      "step": 14550
    },
    {
      "entropy": 0.6444044420495629,
      "epoch": 1.3387674596465806,
      "grad_norm": 0.16704414784908295,
      "learning_rate": 5.5377679639341254e-05,
      "loss": 0.6328,
      "mean_token_accuracy": 0.8077816963195801,
      "num_tokens": 464419991.0,
      "step": 14551
    },
    {
      "entropy": 0.5813427995890379,
      "epoch": 1.3388594660180217,
      "grad_norm": 0.1573123335838318,
      "learning_rate": 5.537461281319962e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.824774406850338,
      "num_tokens": 464452543.0,
      "step": 14552
    },
    {
      "entropy": 0.6403444651514292,
      "epoch": 1.3389514723894629,
      "grad_norm": 0.15794862806797028,
      "learning_rate": 5.5371545987057996e-05,
      "loss": 0.6227,
      "mean_token_accuracy": 0.8112939335405827,
      "num_tokens": 464484397.0,
      "step": 14553
    },
    {
      "entropy": 0.6809296244755387,
      "epoch": 1.3390434787609042,
      "grad_norm": 0.1627720296382904,
      "learning_rate": 5.536847916091638e-05,
      "loss": 0.6477,
      "mean_token_accuracy": 0.8001640625298023,
      "num_tokens": 464516016.0,
      "step": 14554
    },
    {
      "entropy": 0.6830106414854527,
      "epoch": 1.3391354851323454,
      "grad_norm": 0.16273683309555054,
      "learning_rate": 5.5365412334774746e-05,
      "loss": 0.6743,
      "mean_token_accuracy": 0.7985956333577633,
      "num_tokens": 464547617.0,
      "step": 14555
    },
    {
      "entropy": 0.6194204576313496,
      "epoch": 1.3392274915037867,
      "grad_norm": 0.1635269820690155,
      "learning_rate": 5.536234550863312e-05,
      "loss": 0.6067,
      "mean_token_accuracy": 0.8139382600784302,
      "num_tokens": 464578499.0,
      "step": 14556
    },
    {
      "entropy": 0.5807373346760869,
      "epoch": 1.3393194978752279,
      "grad_norm": 0.15309756994247437,
      "learning_rate": 5.535927868249149e-05,
      "loss": 0.5578,
      "mean_token_accuracy": 0.827396247535944,
      "num_tokens": 464610099.0,
      "step": 14557
    },
    {
      "entropy": 0.6380708776414394,
      "epoch": 1.339411504246669,
      "grad_norm": 0.15712442994117737,
      "learning_rate": 5.535621185634987e-05,
      "loss": 0.6111,
      "mean_token_accuracy": 0.8166894093155861,
      "num_tokens": 464641667.0,
      "step": 14558
    },
    {
      "entropy": 0.6045759608969092,
      "epoch": 1.3395035106181103,
      "grad_norm": 0.15448640286922455,
      "learning_rate": 5.5353145030208244e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8178783282637596,
      "num_tokens": 464674156.0,
      "step": 14559
    },
    {
      "entropy": 0.5620197830721736,
      "epoch": 1.3395955169895515,
      "grad_norm": 0.14808781445026398,
      "learning_rate": 5.535007820406661e-05,
      "loss": 0.538,
      "mean_token_accuracy": 0.8321709632873535,
      "num_tokens": 464705811.0,
      "step": 14560
    },
    {
      "entropy": 0.5944844670593739,
      "epoch": 1.3396875233609928,
      "grad_norm": 0.15505889058113098,
      "learning_rate": 5.534701137792499e-05,
      "loss": 0.5755,
      "mean_token_accuracy": 0.8249280117452145,
      "num_tokens": 464738387.0,
      "step": 14561
    },
    {
      "entropy": 0.6739346450194716,
      "epoch": 1.339779529732434,
      "grad_norm": 0.1577652245759964,
      "learning_rate": 5.534394455178337e-05,
      "loss": 0.6572,
      "mean_token_accuracy": 0.800433773547411,
      "num_tokens": 464770047.0,
      "step": 14562
    },
    {
      "entropy": 0.5668320162221789,
      "epoch": 1.339871536103875,
      "grad_norm": 0.14977428317070007,
      "learning_rate": 5.5340877725641736e-05,
      "loss": 0.5567,
      "mean_token_accuracy": 0.8303255848586559,
      "num_tokens": 464802155.0,
      "step": 14563
    },
    {
      "entropy": 0.7234597317874432,
      "epoch": 1.3399635424753165,
      "grad_norm": 0.1631777137517929,
      "learning_rate": 5.533781089950011e-05,
      "loss": 0.72,
      "mean_token_accuracy": 0.7806147299706936,
      "num_tokens": 464833675.0,
      "step": 14564
    },
    {
      "entropy": 0.637717304751277,
      "epoch": 1.3400555488467576,
      "grad_norm": 0.16803409159183502,
      "learning_rate": 5.533474407335848e-05,
      "loss": 0.6409,
      "mean_token_accuracy": 0.8055274188518524,
      "num_tokens": 464865842.0,
      "step": 14565
    },
    {
      "entropy": 0.4560142168775201,
      "epoch": 1.340147555218199,
      "grad_norm": 0.14745397865772247,
      "learning_rate": 5.533167724721686e-05,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8609347566962242,
      "num_tokens": 464898068.0,
      "step": 14566
    },
    {
      "entropy": 0.5780769279226661,
      "epoch": 1.34023956158964,
      "grad_norm": 0.15213695168495178,
      "learning_rate": 5.5328610421075235e-05,
      "loss": 0.563,
      "mean_token_accuracy": 0.8221845664083958,
      "num_tokens": 464929929.0,
      "step": 14567
    },
    {
      "entropy": 0.6855164337903261,
      "epoch": 1.3403315679610812,
      "grad_norm": 0.16909492015838623,
      "learning_rate": 5.53255435949336e-05,
      "loss": 0.6838,
      "mean_token_accuracy": 0.7947349883615971,
      "num_tokens": 464961280.0,
      "step": 14568
    },
    {
      "entropy": 0.5611421596258879,
      "epoch": 1.3404235743325226,
      "grad_norm": 0.14955830574035645,
      "learning_rate": 5.532247676879198e-05,
      "loss": 0.5459,
      "mean_token_accuracy": 0.83179135248065,
      "num_tokens": 464993317.0,
      "step": 14569
    },
    {
      "entropy": 0.6962741855531931,
      "epoch": 1.3405155807039637,
      "grad_norm": 0.1616121083498001,
      "learning_rate": 5.531940994265036e-05,
      "loss": 0.7087,
      "mean_token_accuracy": 0.7803705893456936,
      "num_tokens": 465025456.0,
      "step": 14570
    },
    {
      "entropy": 0.6991517578717321,
      "epoch": 1.340607587075405,
      "grad_norm": 0.16076508164405823,
      "learning_rate": 5.5316343116508727e-05,
      "loss": 0.6894,
      "mean_token_accuracy": 0.7940920256078243,
      "num_tokens": 465056827.0,
      "step": 14571
    },
    {
      "entropy": 0.5618357285857201,
      "epoch": 1.3406995934468462,
      "grad_norm": 0.14988335967063904,
      "learning_rate": 5.53132762903671e-05,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8296152874827385,
      "num_tokens": 465088835.0,
      "step": 14572
    },
    {
      "entropy": 0.6929016630165279,
      "epoch": 1.3407915998182873,
      "grad_norm": 0.16353827714920044,
      "learning_rate": 5.531020946422547e-05,
      "loss": 0.6809,
      "mean_token_accuracy": 0.7925403788685799,
      "num_tokens": 465120382.0,
      "step": 14573
    },
    {
      "entropy": 0.5725655183196068,
      "epoch": 1.3408836061897287,
      "grad_norm": 0.15429267287254333,
      "learning_rate": 5.530714263808385e-05,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8275073990225792,
      "num_tokens": 465152536.0,
      "step": 14574
    },
    {
      "entropy": 0.6500510480254889,
      "epoch": 1.3409756125611698,
      "grad_norm": 0.1566227823495865,
      "learning_rate": 5.5304075811942225e-05,
      "loss": 0.6384,
      "mean_token_accuracy": 0.8033285588026047,
      "num_tokens": 465184554.0,
      "step": 14575
    },
    {
      "entropy": 0.5723518179729581,
      "epoch": 1.3410676189326112,
      "grad_norm": 0.1669560968875885,
      "learning_rate": 5.530100898580059e-05,
      "loss": 0.5508,
      "mean_token_accuracy": 0.826117143034935,
      "num_tokens": 465216161.0,
      "step": 14576
    },
    {
      "entropy": 0.7537697143852711,
      "epoch": 1.3411596253040523,
      "grad_norm": 0.16875948011875153,
      "learning_rate": 5.529794215965897e-05,
      "loss": 0.7477,
      "mean_token_accuracy": 0.7744300253689289,
      "num_tokens": 465248294.0,
      "step": 14577
    },
    {
      "entropy": 0.5734994057565928,
      "epoch": 1.3412516316754934,
      "grad_norm": 0.1489192694425583,
      "learning_rate": 5.529487533351735e-05,
      "loss": 0.5603,
      "mean_token_accuracy": 0.8265109546482563,
      "num_tokens": 465280267.0,
      "step": 14578
    },
    {
      "entropy": 0.5230711624026299,
      "epoch": 1.3413436380469348,
      "grad_norm": 0.15730705857276917,
      "learning_rate": 5.529180850737572e-05,
      "loss": 0.5098,
      "mean_token_accuracy": 0.8412227220833302,
      "num_tokens": 465311671.0,
      "step": 14579
    },
    {
      "entropy": 0.6795086050406098,
      "epoch": 1.341435644418376,
      "grad_norm": 0.16927611827850342,
      "learning_rate": 5.528874168123409e-05,
      "loss": 0.6752,
      "mean_token_accuracy": 0.7963729649782181,
      "num_tokens": 465343585.0,
      "step": 14580
    },
    {
      "entropy": 0.6017687264829874,
      "epoch": 1.3415276507898173,
      "grad_norm": 0.15514343976974487,
      "learning_rate": 5.528567485509246e-05,
      "loss": 0.603,
      "mean_token_accuracy": 0.8183830827474594,
      "num_tokens": 465375728.0,
      "step": 14581
    },
    {
      "entropy": 0.655028460547328,
      "epoch": 1.3416196571612584,
      "grad_norm": 0.1631273329257965,
      "learning_rate": 5.528260802895084e-05,
      "loss": 0.6405,
      "mean_token_accuracy": 0.8073387406766415,
      "num_tokens": 465407939.0,
      "step": 14582
    },
    {
      "entropy": 0.6495299926027656,
      "epoch": 1.3417116635326995,
      "grad_norm": 0.1558968722820282,
      "learning_rate": 5.5279541202809216e-05,
      "loss": 0.6427,
      "mean_token_accuracy": 0.8012167401611805,
      "num_tokens": 465439610.0,
      "step": 14583
    },
    {
      "entropy": 0.6581801846623421,
      "epoch": 1.341803669904141,
      "grad_norm": 0.16964465379714966,
      "learning_rate": 5.527647437666758e-05,
      "loss": 0.6598,
      "mean_token_accuracy": 0.8025665953755379,
      "num_tokens": 465471275.0,
      "step": 14584
    },
    {
      "entropy": 0.5288279447704554,
      "epoch": 1.341895676275582,
      "grad_norm": 0.1486670970916748,
      "learning_rate": 5.527340755052597e-05,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8452484905719757,
      "num_tokens": 465503345.0,
      "step": 14585
    },
    {
      "entropy": 0.6444483492523432,
      "epoch": 1.3419876826470234,
      "grad_norm": 0.16117922961711884,
      "learning_rate": 5.527034072438434e-05,
      "loss": 0.6317,
      "mean_token_accuracy": 0.8068473227322102,
      "num_tokens": 465533965.0,
      "step": 14586
    },
    {
      "entropy": 0.6738056652247906,
      "epoch": 1.3420796890184645,
      "grad_norm": 0.15860223770141602,
      "learning_rate": 5.5267273898242714e-05,
      "loss": 0.666,
      "mean_token_accuracy": 0.7969825305044651,
      "num_tokens": 465565885.0,
      "step": 14587
    },
    {
      "entropy": 0.5808619577437639,
      "epoch": 1.3421716953899057,
      "grad_norm": 0.1490841954946518,
      "learning_rate": 5.526420707210108e-05,
      "loss": 0.5802,
      "mean_token_accuracy": 0.8183219060301781,
      "num_tokens": 465598327.0,
      "step": 14588
    },
    {
      "entropy": 0.7039571180939674,
      "epoch": 1.342263701761347,
      "grad_norm": 0.1641073226928711,
      "learning_rate": 5.5261140245959463e-05,
      "loss": 0.6893,
      "mean_token_accuracy": 0.7881631813943386,
      "num_tokens": 465630383.0,
      "step": 14589
    },
    {
      "entropy": 0.5784109337255359,
      "epoch": 1.3423557081327882,
      "grad_norm": 0.14845310151576996,
      "learning_rate": 5.525807341981784e-05,
      "loss": 0.577,
      "mean_token_accuracy": 0.8213746286928654,
      "num_tokens": 465662672.0,
      "step": 14590
    },
    {
      "entropy": 0.6610774770379066,
      "epoch": 1.3424477145042295,
      "grad_norm": 0.15812331438064575,
      "learning_rate": 5.5255006593676206e-05,
      "loss": 0.6431,
      "mean_token_accuracy": 0.8035270534455776,
      "num_tokens": 465694866.0,
      "step": 14591
    },
    {
      "entropy": 0.7031540609896183,
      "epoch": 1.3425397208756706,
      "grad_norm": 0.16303084790706635,
      "learning_rate": 5.525193976753458e-05,
      "loss": 0.7012,
      "mean_token_accuracy": 0.7863119356334209,
      "num_tokens": 465726810.0,
      "step": 14592
    },
    {
      "entropy": 0.6036721225827932,
      "epoch": 1.3426317272471118,
      "grad_norm": 0.1498565971851349,
      "learning_rate": 5.524887294139296e-05,
      "loss": 0.58,
      "mean_token_accuracy": 0.8225781694054604,
      "num_tokens": 465759255.0,
      "step": 14593
    },
    {
      "entropy": 0.5700885569676757,
      "epoch": 1.3427237336185531,
      "grad_norm": 0.14702996611595154,
      "learning_rate": 5.524580611525133e-05,
      "loss": 0.5548,
      "mean_token_accuracy": 0.8303910158574581,
      "num_tokens": 465791545.0,
      "step": 14594
    },
    {
      "entropy": 0.6444214717485011,
      "epoch": 1.3428157399899943,
      "grad_norm": 0.15962517261505127,
      "learning_rate": 5.5242739289109705e-05,
      "loss": 0.6338,
      "mean_token_accuracy": 0.8071480691432953,
      "num_tokens": 465823275.0,
      "step": 14595
    },
    {
      "entropy": 0.6583511680364609,
      "epoch": 1.3429077463614356,
      "grad_norm": 0.1595609337091446,
      "learning_rate": 5.523967246296807e-05,
      "loss": 0.6511,
      "mean_token_accuracy": 0.8022337518632412,
      "num_tokens": 465854639.0,
      "step": 14596
    },
    {
      "entropy": 0.7242467198520899,
      "epoch": 1.3429997527328768,
      "grad_norm": 0.16471004486083984,
      "learning_rate": 5.5236605636826454e-05,
      "loss": 0.7154,
      "mean_token_accuracy": 0.7864199765026569,
      "num_tokens": 465886432.0,
      "step": 14597
    },
    {
      "entropy": 0.5828711660578847,
      "epoch": 1.3430917591043179,
      "grad_norm": 0.15240991115570068,
      "learning_rate": 5.523353881068483e-05,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8262817896902561,
      "num_tokens": 465919099.0,
      "step": 14598
    },
    {
      "entropy": 0.5874059805646539,
      "epoch": 1.3431837654757592,
      "grad_norm": 0.15550190210342407,
      "learning_rate": 5.5230471984543196e-05,
      "loss": 0.5866,
      "mean_token_accuracy": 0.8234008178114891,
      "num_tokens": 465951767.0,
      "step": 14599
    },
    {
      "entropy": 0.5669521447271109,
      "epoch": 1.3432757718472004,
      "grad_norm": 0.15077407658100128,
      "learning_rate": 5.522740515840157e-05,
      "loss": 0.5416,
      "mean_token_accuracy": 0.8331436887383461,
      "num_tokens": 465983131.0,
      "step": 14600
    },
    {
      "entropy": 0.5999411102384329,
      "epoch": 1.3433677782186417,
      "grad_norm": 0.15848684310913086,
      "learning_rate": 5.522433833225995e-05,
      "loss": 0.5882,
      "mean_token_accuracy": 0.8184205815196037,
      "num_tokens": 466015539.0,
      "step": 14601
    },
    {
      "entropy": 0.5887102670967579,
      "epoch": 1.3434597845900829,
      "grad_norm": 0.15767690539360046,
      "learning_rate": 5.522127150611832e-05,
      "loss": 0.5806,
      "mean_token_accuracy": 0.8177564032375813,
      "num_tokens": 466047418.0,
      "step": 14602
    },
    {
      "entropy": 0.5965588726103306,
      "epoch": 1.343551790961524,
      "grad_norm": 0.14999833703041077,
      "learning_rate": 5.5218204679976695e-05,
      "loss": 0.595,
      "mean_token_accuracy": 0.8186494410037994,
      "num_tokens": 466079322.0,
      "step": 14603
    },
    {
      "entropy": 0.6808634963817894,
      "epoch": 1.3436437973329654,
      "grad_norm": 0.1613369882106781,
      "learning_rate": 5.521513785383506e-05,
      "loss": 0.6839,
      "mean_token_accuracy": 0.796429093927145,
      "num_tokens": 466111521.0,
      "step": 14604
    },
    {
      "entropy": 0.5935218092054129,
      "epoch": 1.3437358037044065,
      "grad_norm": 0.16624251008033752,
      "learning_rate": 5.5212071027693444e-05,
      "loss": 0.5929,
      "mean_token_accuracy": 0.8197607770562172,
      "num_tokens": 466143394.0,
      "step": 14605
    },
    {
      "entropy": 0.6626315386965871,
      "epoch": 1.3438278100758478,
      "grad_norm": 0.15897013247013092,
      "learning_rate": 5.520900420155182e-05,
      "loss": 0.6556,
      "mean_token_accuracy": 0.7966256439685822,
      "num_tokens": 466175798.0,
      "step": 14606
    },
    {
      "entropy": 0.6354603422805667,
      "epoch": 1.343919816447289,
      "grad_norm": 0.15694864094257355,
      "learning_rate": 5.520593737541019e-05,
      "loss": 0.6331,
      "mean_token_accuracy": 0.8072206303477287,
      "num_tokens": 466207846.0,
      "step": 14607
    },
    {
      "entropy": 0.5796185079962015,
      "epoch": 1.34401182281873,
      "grad_norm": 0.15777646005153656,
      "learning_rate": 5.520287054926856e-05,
      "loss": 0.5661,
      "mean_token_accuracy": 0.8275038450956345,
      "num_tokens": 466239518.0,
      "step": 14608
    },
    {
      "entropy": 0.6663821078836918,
      "epoch": 1.3441038291901715,
      "grad_norm": 0.15908990800380707,
      "learning_rate": 5.519980372312694e-05,
      "loss": 0.6553,
      "mean_token_accuracy": 0.8032567165791988,
      "num_tokens": 466271857.0,
      "step": 14609
    },
    {
      "entropy": 0.5747182834893465,
      "epoch": 1.3441958355616126,
      "grad_norm": 0.15297099947929382,
      "learning_rate": 5.519673689698531e-05,
      "loss": 0.5387,
      "mean_token_accuracy": 0.8306274004280567,
      "num_tokens": 466304614.0,
      "step": 14610
    },
    {
      "entropy": 0.6519698761403561,
      "epoch": 1.344287841933054,
      "grad_norm": 0.16176660358905792,
      "learning_rate": 5.5193670070843685e-05,
      "loss": 0.6294,
      "mean_token_accuracy": 0.808183565735817,
      "num_tokens": 466336949.0,
      "step": 14611
    },
    {
      "entropy": 0.6339093614369631,
      "epoch": 1.344379848304495,
      "grad_norm": 0.16077403724193573,
      "learning_rate": 5.519060324470207e-05,
      "loss": 0.6183,
      "mean_token_accuracy": 0.8099652081727982,
      "num_tokens": 466368910.0,
      "step": 14612
    },
    {
      "entropy": 0.6598029872402549,
      "epoch": 1.3444718546759362,
      "grad_norm": 0.1610211730003357,
      "learning_rate": 5.5187536418560435e-05,
      "loss": 0.6522,
      "mean_token_accuracy": 0.8018473945558071,
      "num_tokens": 466401190.0,
      "step": 14613
    },
    {
      "entropy": 0.6495620654895902,
      "epoch": 1.3445638610473776,
      "grad_norm": 0.16144508123397827,
      "learning_rate": 5.518446959241881e-05,
      "loss": 0.6396,
      "mean_token_accuracy": 0.8025612607598305,
      "num_tokens": 466432143.0,
      "step": 14614
    },
    {
      "entropy": 0.6982021071016788,
      "epoch": 1.3446558674188187,
      "grad_norm": 0.16554927825927734,
      "learning_rate": 5.518140276627718e-05,
      "loss": 0.6957,
      "mean_token_accuracy": 0.7885153219103813,
      "num_tokens": 466464371.0,
      "step": 14615
    },
    {
      "entropy": 0.5308394562453032,
      "epoch": 1.34474787379026,
      "grad_norm": 0.1455153524875641,
      "learning_rate": 5.517833594013556e-05,
      "loss": 0.5223,
      "mean_token_accuracy": 0.835718497633934,
      "num_tokens": 466496292.0,
      "step": 14616
    },
    {
      "entropy": 0.6886152401566505,
      "epoch": 1.3448398801617012,
      "grad_norm": 0.16412508487701416,
      "learning_rate": 5.517526911399393e-05,
      "loss": 0.6841,
      "mean_token_accuracy": 0.7919296585023403,
      "num_tokens": 466527783.0,
      "step": 14617
    },
    {
      "entropy": 0.7042386829853058,
      "epoch": 1.3449318865331423,
      "grad_norm": 0.16515062749385834,
      "learning_rate": 5.51722022878523e-05,
      "loss": 0.6904,
      "mean_token_accuracy": 0.785403486341238,
      "num_tokens": 466559099.0,
      "step": 14618
    },
    {
      "entropy": 0.6258791405707598,
      "epoch": 1.3450238929045837,
      "grad_norm": 0.15969984233379364,
      "learning_rate": 5.5169135461710676e-05,
      "loss": 0.6318,
      "mean_token_accuracy": 0.8045515045523643,
      "num_tokens": 466591115.0,
      "step": 14619
    },
    {
      "entropy": 0.5191247840411961,
      "epoch": 1.3451158992760248,
      "grad_norm": 0.15011000633239746,
      "learning_rate": 5.516606863556906e-05,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8426393866539001,
      "num_tokens": 466623021.0,
      "step": 14620
    },
    {
      "entropy": 0.587442715652287,
      "epoch": 1.3452079056474662,
      "grad_norm": 0.1571742445230484,
      "learning_rate": 5.5163001809427425e-05,
      "loss": 0.5973,
      "mean_token_accuracy": 0.8197650536894798,
      "num_tokens": 466654899.0,
      "step": 14621
    },
    {
      "entropy": 0.6160374097526073,
      "epoch": 1.3452999120189073,
      "grad_norm": 0.1626960188150406,
      "learning_rate": 5.51599349832858e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8184729069471359,
      "num_tokens": 466686939.0,
      "step": 14622
    },
    {
      "entropy": 0.7052385210990906,
      "epoch": 1.3453919183903484,
      "grad_norm": 0.15579672157764435,
      "learning_rate": 5.515686815714417e-05,
      "loss": 0.6934,
      "mean_token_accuracy": 0.786295160651207,
      "num_tokens": 466719302.0,
      "step": 14623
    },
    {
      "entropy": 0.64674380607903,
      "epoch": 1.3454839247617898,
      "grad_norm": 0.1560927927494049,
      "learning_rate": 5.5153801331002556e-05,
      "loss": 0.6508,
      "mean_token_accuracy": 0.8010960035026073,
      "num_tokens": 466750836.0,
      "step": 14624
    },
    {
      "entropy": 0.5864535039290786,
      "epoch": 1.345575931133231,
      "grad_norm": 0.15365934371948242,
      "learning_rate": 5.5150734504860924e-05,
      "loss": 0.555,
      "mean_token_accuracy": 0.824067085981369,
      "num_tokens": 466782363.0,
      "step": 14625
    },
    {
      "entropy": 0.5638302089646459,
      "epoch": 1.3456679375046723,
      "grad_norm": 0.16407440602779388,
      "learning_rate": 5.514766767871929e-05,
      "loss": 0.5445,
      "mean_token_accuracy": 0.8347475044429302,
      "num_tokens": 466814163.0,
      "step": 14626
    },
    {
      "entropy": 0.5749542200937867,
      "epoch": 1.3457599438761134,
      "grad_norm": 0.1451949179172516,
      "learning_rate": 5.5144600852577666e-05,
      "loss": 0.5588,
      "mean_token_accuracy": 0.8285631164908409,
      "num_tokens": 466845928.0,
      "step": 14627
    },
    {
      "entropy": 0.6269979570060968,
      "epoch": 1.3458519502475546,
      "grad_norm": 0.16454268991947174,
      "learning_rate": 5.514153402643605e-05,
      "loss": 0.6056,
      "mean_token_accuracy": 0.8139622695744038,
      "num_tokens": 466877283.0,
      "step": 14628
    },
    {
      "entropy": 0.6682307599112391,
      "epoch": 1.345943956618996,
      "grad_norm": 0.16247707605361938,
      "learning_rate": 5.513846720029442e-05,
      "loss": 0.6605,
      "mean_token_accuracy": 0.7981833256781101,
      "num_tokens": 466908619.0,
      "step": 14629
    },
    {
      "entropy": 0.6187205379828811,
      "epoch": 1.346035962990437,
      "grad_norm": 0.1586172878742218,
      "learning_rate": 5.513540037415279e-05,
      "loss": 0.6123,
      "mean_token_accuracy": 0.8150833584368229,
      "num_tokens": 466940465.0,
      "step": 14630
    },
    {
      "entropy": 0.6844214173033834,
      "epoch": 1.3461279693618784,
      "grad_norm": 0.15568162500858307,
      "learning_rate": 5.5132333548011165e-05,
      "loss": 0.6631,
      "mean_token_accuracy": 0.796856228262186,
      "num_tokens": 466972137.0,
      "step": 14631
    },
    {
      "entropy": 0.6361434236168861,
      "epoch": 1.3462199757333195,
      "grad_norm": 0.1586819440126419,
      "learning_rate": 5.5129266721869546e-05,
      "loss": 0.6239,
      "mean_token_accuracy": 0.8076202794909477,
      "num_tokens": 467004218.0,
      "step": 14632
    },
    {
      "entropy": 0.6757839638739824,
      "epoch": 1.3463119821047607,
      "grad_norm": 0.16212399303913116,
      "learning_rate": 5.5126199895727914e-05,
      "loss": 0.6653,
      "mean_token_accuracy": 0.7942702025175095,
      "num_tokens": 467035431.0,
      "step": 14633
    },
    {
      "entropy": 0.637092487886548,
      "epoch": 1.346403988476202,
      "grad_norm": 0.16230563819408417,
      "learning_rate": 5.512313306958629e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.8084519319236279,
      "num_tokens": 467066322.0,
      "step": 14634
    },
    {
      "entropy": 0.6691391821950674,
      "epoch": 1.3464959948476432,
      "grad_norm": 0.16203851997852325,
      "learning_rate": 5.5120066243444657e-05,
      "loss": 0.6603,
      "mean_token_accuracy": 0.7997932769358158,
      "num_tokens": 467098160.0,
      "step": 14635
    },
    {
      "entropy": 0.5866084136068821,
      "epoch": 1.3465880012190845,
      "grad_norm": 0.15489359200000763,
      "learning_rate": 5.511699941730304e-05,
      "loss": 0.5782,
      "mean_token_accuracy": 0.8224041350185871,
      "num_tokens": 467129771.0,
      "step": 14636
    },
    {
      "entropy": 0.6128857228904963,
      "epoch": 1.3466800075905256,
      "grad_norm": 0.1561719924211502,
      "learning_rate": 5.511393259116141e-05,
      "loss": 0.592,
      "mean_token_accuracy": 0.8159217275679111,
      "num_tokens": 467161263.0,
      "step": 14637
    },
    {
      "entropy": 0.5987232085317373,
      "epoch": 1.3467720139619668,
      "grad_norm": 0.15391910076141357,
      "learning_rate": 5.511086576501978e-05,
      "loss": 0.5885,
      "mean_token_accuracy": 0.8197337202727795,
      "num_tokens": 467192736.0,
      "step": 14638
    },
    {
      "entropy": 0.6260922215878963,
      "epoch": 1.3468640203334081,
      "grad_norm": 0.15472418069839478,
      "learning_rate": 5.5107798938878155e-05,
      "loss": 0.621,
      "mean_token_accuracy": 0.8114812932908535,
      "num_tokens": 467224722.0,
      "step": 14639
    },
    {
      "entropy": 0.5830506042111665,
      "epoch": 1.3469560267048493,
      "grad_norm": 0.1546601802110672,
      "learning_rate": 5.5104732112736537e-05,
      "loss": 0.5742,
      "mean_token_accuracy": 0.8234020993113518,
      "num_tokens": 467256660.0,
      "step": 14640
    },
    {
      "entropy": 0.6253827288746834,
      "epoch": 1.3470480330762906,
      "grad_norm": 0.16429705917835236,
      "learning_rate": 5.5101665286594904e-05,
      "loss": 0.6192,
      "mean_token_accuracy": 0.8069103434681892,
      "num_tokens": 467287968.0,
      "step": 14641
    },
    {
      "entropy": 0.614273251965642,
      "epoch": 1.3471400394477318,
      "grad_norm": 0.16000930964946747,
      "learning_rate": 5.509859846045328e-05,
      "loss": 0.6188,
      "mean_token_accuracy": 0.8071829378604889,
      "num_tokens": 467319096.0,
      "step": 14642
    },
    {
      "entropy": 0.6349589941091835,
      "epoch": 1.3472320458191729,
      "grad_norm": 0.15747585892677307,
      "learning_rate": 5.509553163431166e-05,
      "loss": 0.6322,
      "mean_token_accuracy": 0.8106686733663082,
      "num_tokens": 467351534.0,
      "step": 14643
    },
    {
      "entropy": 0.6326035670936108,
      "epoch": 1.3473240521906142,
      "grad_norm": 0.1578516662120819,
      "learning_rate": 5.509246480817003e-05,
      "loss": 0.6258,
      "mean_token_accuracy": 0.8079079128801823,
      "num_tokens": 467383192.0,
      "step": 14644
    },
    {
      "entropy": 0.6963685229420662,
      "epoch": 1.3474160585620554,
      "grad_norm": 0.16342301666736603,
      "learning_rate": 5.50893979820284e-05,
      "loss": 0.6757,
      "mean_token_accuracy": 0.7920359000563622,
      "num_tokens": 467414505.0,
      "step": 14645
    },
    {
      "entropy": 0.6057593822479248,
      "epoch": 1.3475080649334967,
      "grad_norm": 0.15615549683570862,
      "learning_rate": 5.508633115588677e-05,
      "loss": 0.5945,
      "mean_token_accuracy": 0.8163361884653568,
      "num_tokens": 467446341.0,
      "step": 14646
    },
    {
      "entropy": 0.5905669322237372,
      "epoch": 1.3476000713049379,
      "grad_norm": 0.14912478625774384,
      "learning_rate": 5.508326432974515e-05,
      "loss": 0.5866,
      "mean_token_accuracy": 0.8196186311542988,
      "num_tokens": 467478441.0,
      "step": 14647
    },
    {
      "entropy": 0.5478211913723499,
      "epoch": 1.347692077676379,
      "grad_norm": 0.15946662425994873,
      "learning_rate": 5.508019750360353e-05,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8330729641020298,
      "num_tokens": 467510200.0,
      "step": 14648
    },
    {
      "entropy": 0.5853137318044901,
      "epoch": 1.3477840840478204,
      "grad_norm": 0.17327284812927246,
      "learning_rate": 5.5077130677461895e-05,
      "loss": 0.5693,
      "mean_token_accuracy": 0.8241978585720062,
      "num_tokens": 467542130.0,
      "step": 14649
    },
    {
      "entropy": 0.7089829556643963,
      "epoch": 1.3478760904192615,
      "grad_norm": 0.16843582689762115,
      "learning_rate": 5.507406385132027e-05,
      "loss": 0.706,
      "mean_token_accuracy": 0.7846762873232365,
      "num_tokens": 467573209.0,
      "step": 14650
    },
    {
      "entropy": 0.5522478390485048,
      "epoch": 1.3479680967907028,
      "grad_norm": 0.14941349625587463,
      "learning_rate": 5.507099702517865e-05,
      "loss": 0.5334,
      "mean_token_accuracy": 0.8323520012199879,
      "num_tokens": 467604879.0,
      "step": 14651
    },
    {
      "entropy": 0.5973620377480984,
      "epoch": 1.348060103162144,
      "grad_norm": 0.16265559196472168,
      "learning_rate": 5.506793019903702e-05,
      "loss": 0.5821,
      "mean_token_accuracy": 0.8236492276191711,
      "num_tokens": 467637066.0,
      "step": 14652
    },
    {
      "entropy": 0.5732437232509255,
      "epoch": 1.348152109533585,
      "grad_norm": 0.15032778680324554,
      "learning_rate": 5.5064863372895393e-05,
      "loss": 0.5627,
      "mean_token_accuracy": 0.824813649058342,
      "num_tokens": 467669522.0,
      "step": 14653
    },
    {
      "entropy": 0.6221229564398527,
      "epoch": 1.3482441159050265,
      "grad_norm": 0.1596708446741104,
      "learning_rate": 5.506179654675376e-05,
      "loss": 0.614,
      "mean_token_accuracy": 0.8092232123017311,
      "num_tokens": 467701420.0,
      "step": 14654
    },
    {
      "entropy": 0.663469479419291,
      "epoch": 1.3483361222764676,
      "grad_norm": 0.15816128253936768,
      "learning_rate": 5.505872972061214e-05,
      "loss": 0.6477,
      "mean_token_accuracy": 0.8026783168315887,
      "num_tokens": 467733468.0,
      "step": 14655
    },
    {
      "entropy": 0.5608764197677374,
      "epoch": 1.348428128647909,
      "grad_norm": 0.15188798308372498,
      "learning_rate": 5.505566289447052e-05,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8350550308823586,
      "num_tokens": 467765474.0,
      "step": 14656
    },
    {
      "entropy": 0.6777988281100988,
      "epoch": 1.34852013501935,
      "grad_norm": 0.16851459443569183,
      "learning_rate": 5.5052596068328885e-05,
      "loss": 0.6732,
      "mean_token_accuracy": 0.7969768792390823,
      "num_tokens": 467797424.0,
      "step": 14657
    },
    {
      "entropy": 0.5983353182673454,
      "epoch": 1.3486121413907912,
      "grad_norm": 0.15396103262901306,
      "learning_rate": 5.504952924218726e-05,
      "loss": 0.5933,
      "mean_token_accuracy": 0.8171647824347019,
      "num_tokens": 467829844.0,
      "step": 14658
    },
    {
      "entropy": 0.6394079327583313,
      "epoch": 1.3487041477622326,
      "grad_norm": 0.15945762395858765,
      "learning_rate": 5.504646241604564e-05,
      "loss": 0.6295,
      "mean_token_accuracy": 0.8073664903640747,
      "num_tokens": 467862117.0,
      "step": 14659
    },
    {
      "entropy": 0.7406965177506208,
      "epoch": 1.3487961541336737,
      "grad_norm": 0.1578383594751358,
      "learning_rate": 5.504339558990401e-05,
      "loss": 0.7087,
      "mean_token_accuracy": 0.7758539579808712,
      "num_tokens": 467893021.0,
      "step": 14660
    },
    {
      "entropy": 0.6460036020725965,
      "epoch": 1.348888160505115,
      "grad_norm": 0.1575346291065216,
      "learning_rate": 5.5040328763762384e-05,
      "loss": 0.64,
      "mean_token_accuracy": 0.8053660839796066,
      "num_tokens": 467925692.0,
      "step": 14661
    },
    {
      "entropy": 0.618834288790822,
      "epoch": 1.3489801668765562,
      "grad_norm": 0.1537853181362152,
      "learning_rate": 5.503726193762075e-05,
      "loss": 0.607,
      "mean_token_accuracy": 0.8124931342899799,
      "num_tokens": 467957496.0,
      "step": 14662
    },
    {
      "entropy": 0.7773927338421345,
      "epoch": 1.3490721732479973,
      "grad_norm": 0.16289763152599335,
      "learning_rate": 5.503419511147914e-05,
      "loss": 0.757,
      "mean_token_accuracy": 0.768048420548439,
      "num_tokens": 467989357.0,
      "step": 14663
    },
    {
      "entropy": 0.6361457277089357,
      "epoch": 1.3491641796194387,
      "grad_norm": 0.15931326150894165,
      "learning_rate": 5.503112828533751e-05,
      "loss": 0.6223,
      "mean_token_accuracy": 0.8092742711305618,
      "num_tokens": 468020904.0,
      "step": 14664
    },
    {
      "entropy": 0.5969007853418589,
      "epoch": 1.3492561859908798,
      "grad_norm": 0.1502041071653366,
      "learning_rate": 5.5028061459195876e-05,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8240650072693825,
      "num_tokens": 468053420.0,
      "step": 14665
    },
    {
      "entropy": 0.6268738582730293,
      "epoch": 1.3493481923623212,
      "grad_norm": 0.15839803218841553,
      "learning_rate": 5.502499463305425e-05,
      "loss": 0.6221,
      "mean_token_accuracy": 0.810091495513916,
      "num_tokens": 468085955.0,
      "step": 14666
    },
    {
      "entropy": 0.6813087947666645,
      "epoch": 1.3494401987337623,
      "grad_norm": 0.16195067763328552,
      "learning_rate": 5.502192780691263e-05,
      "loss": 0.6874,
      "mean_token_accuracy": 0.7919379882514477,
      "num_tokens": 468118124.0,
      "step": 14667
    },
    {
      "entropy": 0.7073140554130077,
      "epoch": 1.3495322051052034,
      "grad_norm": 0.1591597944498062,
      "learning_rate": 5.5018860980771006e-05,
      "loss": 0.6944,
      "mean_token_accuracy": 0.7895864173769951,
      "num_tokens": 468150042.0,
      "step": 14668
    },
    {
      "entropy": 0.615777206607163,
      "epoch": 1.3496242114766448,
      "grad_norm": 0.15488925576210022,
      "learning_rate": 5.5015794154629374e-05,
      "loss": 0.6016,
      "mean_token_accuracy": 0.811143197119236,
      "num_tokens": 468182040.0,
      "step": 14669
    },
    {
      "entropy": 0.58083526045084,
      "epoch": 1.349716217848086,
      "grad_norm": 0.1539848893880844,
      "learning_rate": 5.501272732848774e-05,
      "loss": 0.5605,
      "mean_token_accuracy": 0.8291188441216946,
      "num_tokens": 468212812.0,
      "step": 14670
    },
    {
      "entropy": 0.6596193872392178,
      "epoch": 1.3498082242195273,
      "grad_norm": 0.16315235197544098,
      "learning_rate": 5.500966050234613e-05,
      "loss": 0.6507,
      "mean_token_accuracy": 0.8037034198641777,
      "num_tokens": 468243633.0,
      "step": 14671
    },
    {
      "entropy": 0.590236664749682,
      "epoch": 1.3499002305909684,
      "grad_norm": 0.15151505172252655,
      "learning_rate": 5.50065936762045e-05,
      "loss": 0.5806,
      "mean_token_accuracy": 0.8211543820798397,
      "num_tokens": 468274968.0,
      "step": 14672
    },
    {
      "entropy": 0.6169228637591004,
      "epoch": 1.3499922369624096,
      "grad_norm": 0.15061479806900024,
      "learning_rate": 5.500352685006287e-05,
      "loss": 0.6123,
      "mean_token_accuracy": 0.8175216764211655,
      "num_tokens": 468307664.0,
      "step": 14673
    },
    {
      "entropy": 0.6300032972358167,
      "epoch": 1.350084243333851,
      "grad_norm": 0.15836650133132935,
      "learning_rate": 5.5000460023921254e-05,
      "loss": 0.6298,
      "mean_token_accuracy": 0.805866926908493,
      "num_tokens": 468339676.0,
      "step": 14674
    },
    {
      "entropy": 0.7475789599120617,
      "epoch": 1.350176249705292,
      "grad_norm": 0.1722394973039627,
      "learning_rate": 5.499739319777962e-05,
      "loss": 0.7485,
      "mean_token_accuracy": 0.7730598263442516,
      "num_tokens": 468372193.0,
      "step": 14675
    },
    {
      "entropy": 0.6198837943375111,
      "epoch": 1.3502682560767334,
      "grad_norm": 0.1580936163663864,
      "learning_rate": 5.4994326371638e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8111650422215462,
      "num_tokens": 468404184.0,
      "step": 14676
    },
    {
      "entropy": 0.5735403327271342,
      "epoch": 1.3503602624481745,
      "grad_norm": 0.1576077789068222,
      "learning_rate": 5.4991259545496365e-05,
      "loss": 0.5674,
      "mean_token_accuracy": 0.8292984403669834,
      "num_tokens": 468436823.0,
      "step": 14677
    },
    {
      "entropy": 0.6618277970701456,
      "epoch": 1.3504522688196157,
      "grad_norm": 0.15672726929187775,
      "learning_rate": 5.4988192719354746e-05,
      "loss": 0.6411,
      "mean_token_accuracy": 0.8020278178155422,
      "num_tokens": 468468841.0,
      "step": 14678
    },
    {
      "entropy": 0.5213112882338464,
      "epoch": 1.350544275191057,
      "grad_norm": 0.14612917602062225,
      "learning_rate": 5.498512589321312e-05,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8435754142701626,
      "num_tokens": 468501263.0,
      "step": 14679
    },
    {
      "entropy": 0.5263933020178229,
      "epoch": 1.3506362815624982,
      "grad_norm": 0.14793480932712555,
      "learning_rate": 5.498205906707149e-05,
      "loss": 0.5093,
      "mean_token_accuracy": 0.8412965312600136,
      "num_tokens": 468532326.0,
      "step": 14680
    },
    {
      "entropy": 0.516806548461318,
      "epoch": 1.3507282879339395,
      "grad_norm": 0.14176443219184875,
      "learning_rate": 5.497899224092986e-05,
      "loss": 0.4952,
      "mean_token_accuracy": 0.8444674909114838,
      "num_tokens": 468564367.0,
      "step": 14681
    },
    {
      "entropy": 0.6871064677834511,
      "epoch": 1.3508202943053806,
      "grad_norm": 0.16106398403644562,
      "learning_rate": 5.4975925414788245e-05,
      "loss": 0.6924,
      "mean_token_accuracy": 0.7916266433894634,
      "num_tokens": 468596809.0,
      "step": 14682
    },
    {
      "entropy": 0.5974480248987675,
      "epoch": 1.3509123006768218,
      "grad_norm": 0.1557808667421341,
      "learning_rate": 5.497285858864661e-05,
      "loss": 0.5865,
      "mean_token_accuracy": 0.8204926960170269,
      "num_tokens": 468628905.0,
      "step": 14683
    },
    {
      "entropy": 0.6203712113201618,
      "epoch": 1.3510043070482631,
      "grad_norm": 0.16421616077423096,
      "learning_rate": 5.496979176250499e-05,
      "loss": 0.6169,
      "mean_token_accuracy": 0.8119299225509167,
      "num_tokens": 468660102.0,
      "step": 14684
    },
    {
      "entropy": 0.6856940649449825,
      "epoch": 1.3510963134197043,
      "grad_norm": 0.16242758929729462,
      "learning_rate": 5.4966724936363355e-05,
      "loss": 0.6805,
      "mean_token_accuracy": 0.795138780027628,
      "num_tokens": 468692673.0,
      "step": 14685
    },
    {
      "entropy": 0.6076365206390619,
      "epoch": 1.3511883197911456,
      "grad_norm": 0.15696954727172852,
      "learning_rate": 5.4963658110221736e-05,
      "loss": 0.5991,
      "mean_token_accuracy": 0.8164095245301723,
      "num_tokens": 468725201.0,
      "step": 14686
    },
    {
      "entropy": 0.6638117656111717,
      "epoch": 1.3512803261625868,
      "grad_norm": 0.1616266518831253,
      "learning_rate": 5.496059128408011e-05,
      "loss": 0.6561,
      "mean_token_accuracy": 0.8007662743330002,
      "num_tokens": 468757735.0,
      "step": 14687
    },
    {
      "entropy": 0.6191193033009768,
      "epoch": 1.3513723325340279,
      "grad_norm": 0.15107330679893494,
      "learning_rate": 5.495752445793848e-05,
      "loss": 0.6079,
      "mean_token_accuracy": 0.8139500357210636,
      "num_tokens": 468790114.0,
      "step": 14688
    },
    {
      "entropy": 0.5608787182718515,
      "epoch": 1.3514643389054692,
      "grad_norm": 0.15100523829460144,
      "learning_rate": 5.4954457631796854e-05,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8287653475999832,
      "num_tokens": 468820835.0,
      "step": 14689
    },
    {
      "entropy": 0.6653584716841578,
      "epoch": 1.3515563452769104,
      "grad_norm": 0.164289653301239,
      "learning_rate": 5.4951390805655235e-05,
      "loss": 0.6437,
      "mean_token_accuracy": 0.8025845997035503,
      "num_tokens": 468852717.0,
      "step": 14690
    },
    {
      "entropy": 0.6280757524073124,
      "epoch": 1.3516483516483517,
      "grad_norm": 0.15710724890232086,
      "learning_rate": 5.49483239795136e-05,
      "loss": 0.6223,
      "mean_token_accuracy": 0.8055935353040695,
      "num_tokens": 468885229.0,
      "step": 14691
    },
    {
      "entropy": 0.5385112585499883,
      "epoch": 1.3517403580197929,
      "grad_norm": 0.14927566051483154,
      "learning_rate": 5.494525715337198e-05,
      "loss": 0.5263,
      "mean_token_accuracy": 0.8376956395804882,
      "num_tokens": 468917702.0,
      "step": 14692
    },
    {
      "entropy": 0.6375886695459485,
      "epoch": 1.351832364391234,
      "grad_norm": 0.16353949904441833,
      "learning_rate": 5.4942190327230345e-05,
      "loss": 0.6318,
      "mean_token_accuracy": 0.8061693869531155,
      "num_tokens": 468949391.0,
      "step": 14693
    },
    {
      "entropy": 0.6976904198527336,
      "epoch": 1.3519243707626754,
      "grad_norm": 0.16168011724948883,
      "learning_rate": 5.493912350108873e-05,
      "loss": 0.6756,
      "mean_token_accuracy": 0.7968547642230988,
      "num_tokens": 468981608.0,
      "step": 14694
    },
    {
      "entropy": 0.593445397913456,
      "epoch": 1.3520163771341165,
      "grad_norm": 0.15312519669532776,
      "learning_rate": 5.49360566749471e-05,
      "loss": 0.586,
      "mean_token_accuracy": 0.816348809748888,
      "num_tokens": 469013679.0,
      "step": 14695
    },
    {
      "entropy": 0.6461929911747575,
      "epoch": 1.3521083835055578,
      "grad_norm": 0.1696833223104477,
      "learning_rate": 5.493298984880547e-05,
      "loss": 0.6275,
      "mean_token_accuracy": 0.8072440847754478,
      "num_tokens": 469044641.0,
      "step": 14696
    },
    {
      "entropy": 0.6005293559283018,
      "epoch": 1.352200389876999,
      "grad_norm": 0.16369770467281342,
      "learning_rate": 5.4929923022663844e-05,
      "loss": 0.577,
      "mean_token_accuracy": 0.8199713341891766,
      "num_tokens": 469075541.0,
      "step": 14697
    },
    {
      "entropy": 0.6205514967441559,
      "epoch": 1.35229239624844,
      "grad_norm": 0.16148220002651215,
      "learning_rate": 5.4926856196522225e-05,
      "loss": 0.5898,
      "mean_token_accuracy": 0.816047914326191,
      "num_tokens": 469106242.0,
      "step": 14698
    },
    {
      "entropy": 0.6542916428297758,
      "epoch": 1.3523844026198815,
      "grad_norm": 0.15750500559806824,
      "learning_rate": 5.492378937038059e-05,
      "loss": 0.666,
      "mean_token_accuracy": 0.7957699969410896,
      "num_tokens": 469138341.0,
      "step": 14699
    },
    {
      "entropy": 0.6494984570890665,
      "epoch": 1.3524764089913226,
      "grad_norm": 0.16120287775993347,
      "learning_rate": 5.492072254423897e-05,
      "loss": 0.6434,
      "mean_token_accuracy": 0.8063666895031929,
      "num_tokens": 469170801.0,
      "step": 14700
    },
    {
      "entropy": 0.6669717533513904,
      "epoch": 1.352568415362764,
      "grad_norm": 0.16027094423770905,
      "learning_rate": 5.4917655718097336e-05,
      "loss": 0.6551,
      "mean_token_accuracy": 0.8013458587229252,
      "num_tokens": 469202775.0,
      "step": 14701
    },
    {
      "entropy": 0.6031773304566741,
      "epoch": 1.352660421734205,
      "grad_norm": 0.16765953600406647,
      "learning_rate": 5.491458889195572e-05,
      "loss": 0.5917,
      "mean_token_accuracy": 0.8199883848428726,
      "num_tokens": 469234106.0,
      "step": 14702
    },
    {
      "entropy": 0.5859543588012457,
      "epoch": 1.3527524281056462,
      "grad_norm": 0.15349110960960388,
      "learning_rate": 5.491152206581409e-05,
      "loss": 0.569,
      "mean_token_accuracy": 0.823901642113924,
      "num_tokens": 469264883.0,
      "step": 14703
    },
    {
      "entropy": 0.6448238855227828,
      "epoch": 1.3528444344770876,
      "grad_norm": 0.16023406386375427,
      "learning_rate": 5.490845523967246e-05,
      "loss": 0.6379,
      "mean_token_accuracy": 0.8066481910645962,
      "num_tokens": 469296350.0,
      "step": 14704
    },
    {
      "entropy": 0.659963137935847,
      "epoch": 1.3529364408485287,
      "grad_norm": 0.1613236665725708,
      "learning_rate": 5.490538841353085e-05,
      "loss": 0.659,
      "mean_token_accuracy": 0.802082184702158,
      "num_tokens": 469328593.0,
      "step": 14705
    },
    {
      "entropy": 0.6111921388655901,
      "epoch": 1.35302844721997,
      "grad_norm": 0.15556840598583221,
      "learning_rate": 5.4902321587389216e-05,
      "loss": 0.5932,
      "mean_token_accuracy": 0.8173818998038769,
      "num_tokens": 469361048.0,
      "step": 14706
    },
    {
      "entropy": 0.5505812987685204,
      "epoch": 1.3531204535914112,
      "grad_norm": 0.1509464830160141,
      "learning_rate": 5.489925476124759e-05,
      "loss": 0.5244,
      "mean_token_accuracy": 0.839062474668026,
      "num_tokens": 469392943.0,
      "step": 14707
    },
    {
      "entropy": 0.6771602295339108,
      "epoch": 1.3532124599628523,
      "grad_norm": 0.16232311725616455,
      "learning_rate": 5.489618793510596e-05,
      "loss": 0.6676,
      "mean_token_accuracy": 0.7989814206957817,
      "num_tokens": 469425262.0,
      "step": 14708
    },
    {
      "entropy": 0.5769048929214478,
      "epoch": 1.3533044663342937,
      "grad_norm": 0.15592406690120697,
      "learning_rate": 5.489312110896434e-05,
      "loss": 0.5603,
      "mean_token_accuracy": 0.8268850073218346,
      "num_tokens": 469457464.0,
      "step": 14709
    },
    {
      "entropy": 0.5922532361000776,
      "epoch": 1.3533964727057348,
      "grad_norm": 0.15798050165176392,
      "learning_rate": 5.4890054282822714e-05,
      "loss": 0.5868,
      "mean_token_accuracy": 0.8198157176375389,
      "num_tokens": 469489552.0,
      "step": 14710
    },
    {
      "entropy": 0.6154823657125235,
      "epoch": 1.3534884790771762,
      "grad_norm": 0.15962103009223938,
      "learning_rate": 5.488698745668108e-05,
      "loss": 0.6045,
      "mean_token_accuracy": 0.8199844770133495,
      "num_tokens": 469521846.0,
      "step": 14711
    },
    {
      "entropy": 0.5014054700732231,
      "epoch": 1.3535804854486173,
      "grad_norm": 0.14511972665786743,
      "learning_rate": 5.488392063053946e-05,
      "loss": 0.4927,
      "mean_token_accuracy": 0.8488524183630943,
      "num_tokens": 469553690.0,
      "step": 14712
    },
    {
      "entropy": 0.549377916380763,
      "epoch": 1.3536724918200584,
      "grad_norm": 0.15441003441810608,
      "learning_rate": 5.488085380439784e-05,
      "loss": 0.5357,
      "mean_token_accuracy": 0.8302528113126755,
      "num_tokens": 469585741.0,
      "step": 14713
    },
    {
      "entropy": 0.5502155553549528,
      "epoch": 1.3537644981914998,
      "grad_norm": 0.148070827126503,
      "learning_rate": 5.4877786978256206e-05,
      "loss": 0.5332,
      "mean_token_accuracy": 0.8340898044407368,
      "num_tokens": 469617642.0,
      "step": 14714
    },
    {
      "entropy": 0.49619755893945694,
      "epoch": 1.353856504562941,
      "grad_norm": 0.14979444444179535,
      "learning_rate": 5.487472015211458e-05,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8519890047609806,
      "num_tokens": 469649676.0,
      "step": 14715
    },
    {
      "entropy": 0.6025059875100851,
      "epoch": 1.3539485109343823,
      "grad_norm": 0.15026655793190002,
      "learning_rate": 5.487165332597295e-05,
      "loss": 0.602,
      "mean_token_accuracy": 0.8144104368984699,
      "num_tokens": 469682134.0,
      "step": 14716
    },
    {
      "entropy": 0.5593804214149714,
      "epoch": 1.3540405173058234,
      "grad_norm": 0.15612006187438965,
      "learning_rate": 5.486858649983133e-05,
      "loss": 0.5504,
      "mean_token_accuracy": 0.8287537582218647,
      "num_tokens": 469714064.0,
      "step": 14717
    },
    {
      "entropy": 0.5862536914646626,
      "epoch": 1.3541325236772646,
      "grad_norm": 0.1626586616039276,
      "learning_rate": 5.4865519673689705e-05,
      "loss": 0.5705,
      "mean_token_accuracy": 0.8272892460227013,
      "num_tokens": 469745837.0,
      "step": 14718
    },
    {
      "entropy": 0.6404256988316774,
      "epoch": 1.354224530048706,
      "grad_norm": 0.1591843217611313,
      "learning_rate": 5.486245284754807e-05,
      "loss": 0.6345,
      "mean_token_accuracy": 0.8042816780507565,
      "num_tokens": 469778143.0,
      "step": 14719
    },
    {
      "entropy": 0.5818128706887364,
      "epoch": 1.354316536420147,
      "grad_norm": 0.15921717882156372,
      "learning_rate": 5.485938602140645e-05,
      "loss": 0.5523,
      "mean_token_accuracy": 0.8297712467610836,
      "num_tokens": 469810518.0,
      "step": 14720
    },
    {
      "entropy": 0.6387194292619824,
      "epoch": 1.3544085427915884,
      "grad_norm": 0.1640874296426773,
      "learning_rate": 5.485631919526483e-05,
      "loss": 0.6246,
      "mean_token_accuracy": 0.8098399341106415,
      "num_tokens": 469842292.0,
      "step": 14721
    },
    {
      "entropy": 0.5555978938937187,
      "epoch": 1.3545005491630295,
      "grad_norm": 0.147792249917984,
      "learning_rate": 5.48532523691232e-05,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8354363478720188,
      "num_tokens": 469874384.0,
      "step": 14722
    },
    {
      "entropy": 0.6060567181557417,
      "epoch": 1.3545925555344707,
      "grad_norm": 0.16105790436267853,
      "learning_rate": 5.485018554298157e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8181341029703617,
      "num_tokens": 469905568.0,
      "step": 14723
    },
    {
      "entropy": 0.5013373577967286,
      "epoch": 1.354684561905912,
      "grad_norm": 0.16173683106899261,
      "learning_rate": 5.484711871683994e-05,
      "loss": 0.4913,
      "mean_token_accuracy": 0.8492575474083424,
      "num_tokens": 469937654.0,
      "step": 14724
    },
    {
      "entropy": 0.6633637314662337,
      "epoch": 1.3547765682773532,
      "grad_norm": 0.165988951921463,
      "learning_rate": 5.484405189069832e-05,
      "loss": 0.6545,
      "mean_token_accuracy": 0.7992563396692276,
      "num_tokens": 469970023.0,
      "step": 14725
    },
    {
      "entropy": 0.7304394487291574,
      "epoch": 1.3548685746487945,
      "grad_norm": 0.1651276797056198,
      "learning_rate": 5.4840985064556695e-05,
      "loss": 0.7232,
      "mean_token_accuracy": 0.7801786847412586,
      "num_tokens": 470001889.0,
      "step": 14726
    },
    {
      "entropy": 0.5683437073603272,
      "epoch": 1.3549605810202356,
      "grad_norm": 0.15148860216140747,
      "learning_rate": 5.483791823841506e-05,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8314051926136017,
      "num_tokens": 470034167.0,
      "step": 14727
    },
    {
      "entropy": 0.7034685644321144,
      "epoch": 1.3550525873916768,
      "grad_norm": 0.15680283308029175,
      "learning_rate": 5.483485141227344e-05,
      "loss": 0.6946,
      "mean_token_accuracy": 0.7885586284101009,
      "num_tokens": 470066324.0,
      "step": 14728
    },
    {
      "entropy": 0.6133805681020021,
      "epoch": 1.3551445937631181,
      "grad_norm": 0.15167081356048584,
      "learning_rate": 5.483178458613182e-05,
      "loss": 0.6015,
      "mean_token_accuracy": 0.8173081912100315,
      "num_tokens": 470098608.0,
      "step": 14729
    },
    {
      "entropy": 0.6086510196328163,
      "epoch": 1.3552366001345593,
      "grad_norm": 0.1616763174533844,
      "learning_rate": 5.482871775999019e-05,
      "loss": 0.5952,
      "mean_token_accuracy": 0.8176645264029503,
      "num_tokens": 470130817.0,
      "step": 14730
    },
    {
      "entropy": 0.5428434135392308,
      "epoch": 1.3553286065060006,
      "grad_norm": 0.15099331736564636,
      "learning_rate": 5.482565093384856e-05,
      "loss": 0.5322,
      "mean_token_accuracy": 0.8380420990288258,
      "num_tokens": 470163477.0,
      "step": 14731
    },
    {
      "entropy": 0.5084349093958735,
      "epoch": 1.3554206128774418,
      "grad_norm": 0.15301960706710815,
      "learning_rate": 5.482258410770693e-05,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8457734473049641,
      "num_tokens": 470196015.0,
      "step": 14732
    },
    {
      "entropy": 0.6496388949453831,
      "epoch": 1.3555126192488829,
      "grad_norm": 0.1612861454486847,
      "learning_rate": 5.481951728156531e-05,
      "loss": 0.6387,
      "mean_token_accuracy": 0.8042855300009251,
      "num_tokens": 470227701.0,
      "step": 14733
    },
    {
      "entropy": 0.6250578677281737,
      "epoch": 1.3556046256203242,
      "grad_norm": 0.16123203933238983,
      "learning_rate": 5.4816450455423686e-05,
      "loss": 0.6071,
      "mean_token_accuracy": 0.8119805864989758,
      "num_tokens": 470258401.0,
      "step": 14734
    },
    {
      "entropy": 0.7368014380335808,
      "epoch": 1.3556966319917654,
      "grad_norm": 0.16977395117282867,
      "learning_rate": 5.4813383629282054e-05,
      "loss": 0.7353,
      "mean_token_accuracy": 0.7772468961775303,
      "num_tokens": 470290645.0,
      "step": 14735
    },
    {
      "entropy": 0.6666177348233759,
      "epoch": 1.3557886383632067,
      "grad_norm": 0.15822561085224152,
      "learning_rate": 5.4810316803140435e-05,
      "loss": 0.6622,
      "mean_token_accuracy": 0.7974517345428467,
      "num_tokens": 470323000.0,
      "step": 14736
    },
    {
      "entropy": 0.5431550154462457,
      "epoch": 1.3558806447346479,
      "grad_norm": 0.1597980409860611,
      "learning_rate": 5.480724997699881e-05,
      "loss": 0.536,
      "mean_token_accuracy": 0.8348720148205757,
      "num_tokens": 470355084.0,
      "step": 14737
    },
    {
      "entropy": 0.5724566685967147,
      "epoch": 1.355972651106089,
      "grad_norm": 0.15829907357692719,
      "learning_rate": 5.480418315085718e-05,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8283042386174202,
      "num_tokens": 470386428.0,
      "step": 14738
    },
    {
      "entropy": 0.7035950515419245,
      "epoch": 1.3560646574775304,
      "grad_norm": 0.15978491306304932,
      "learning_rate": 5.480111632471555e-05,
      "loss": 0.6967,
      "mean_token_accuracy": 0.7916225679218769,
      "num_tokens": 470418446.0,
      "step": 14739
    },
    {
      "entropy": 0.5684420662000775,
      "epoch": 1.3561566638489715,
      "grad_norm": 0.1498793363571167,
      "learning_rate": 5.4798049498573934e-05,
      "loss": 0.53,
      "mean_token_accuracy": 0.8327326066792011,
      "num_tokens": 470448148.0,
      "step": 14740
    },
    {
      "entropy": 0.6875486988574266,
      "epoch": 1.3562486702204128,
      "grad_norm": 0.15793615579605103,
      "learning_rate": 5.47949826724323e-05,
      "loss": 0.676,
      "mean_token_accuracy": 0.7909862399101257,
      "num_tokens": 470480397.0,
      "step": 14741
    },
    {
      "entropy": 0.6170417740941048,
      "epoch": 1.356340676591854,
      "grad_norm": 0.1566699892282486,
      "learning_rate": 5.4791915846290676e-05,
      "loss": 0.6007,
      "mean_token_accuracy": 0.8183836191892624,
      "num_tokens": 470512131.0,
      "step": 14742
    },
    {
      "entropy": 0.6200792752206326,
      "epoch": 1.3564326829632951,
      "grad_norm": 0.15664821863174438,
      "learning_rate": 5.4788849020149044e-05,
      "loss": 0.6058,
      "mean_token_accuracy": 0.8132312558591366,
      "num_tokens": 470544516.0,
      "step": 14743
    },
    {
      "entropy": 0.6042091064155102,
      "epoch": 1.3565246893347365,
      "grad_norm": 0.15700402855873108,
      "learning_rate": 5.478578219400743e-05,
      "loss": 0.5946,
      "mean_token_accuracy": 0.8164425678551197,
      "num_tokens": 470576353.0,
      "step": 14744
    },
    {
      "entropy": 0.5156581560149789,
      "epoch": 1.3566166957061776,
      "grad_norm": 0.15514178574085236,
      "learning_rate": 5.47827153678658e-05,
      "loss": 0.4971,
      "mean_token_accuracy": 0.8448445126414299,
      "num_tokens": 470608326.0,
      "step": 14745
    },
    {
      "entropy": 0.7324144411832094,
      "epoch": 1.356708702077619,
      "grad_norm": 0.16028863191604614,
      "learning_rate": 5.477964854172417e-05,
      "loss": 0.7303,
      "mean_token_accuracy": 0.7812350578606129,
      "num_tokens": 470640905.0,
      "step": 14746
    },
    {
      "entropy": 0.7703166119754314,
      "epoch": 1.35680070844906,
      "grad_norm": 0.16505025327205658,
      "learning_rate": 5.477658171558254e-05,
      "loss": 0.7583,
      "mean_token_accuracy": 0.772431094199419,
      "num_tokens": 470672996.0,
      "step": 14747
    },
    {
      "entropy": 0.5731086498126388,
      "epoch": 1.3568927148205012,
      "grad_norm": 0.14712443947792053,
      "learning_rate": 5.4773514889440924e-05,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8270552903413773,
      "num_tokens": 470705344.0,
      "step": 14748
    },
    {
      "entropy": 0.6807742724195123,
      "epoch": 1.3569847211919426,
      "grad_norm": 0.15706093609333038,
      "learning_rate": 5.47704480632993e-05,
      "loss": 0.6744,
      "mean_token_accuracy": 0.7982071749866009,
      "num_tokens": 470737205.0,
      "step": 14749
    },
    {
      "entropy": 0.6395709551870823,
      "epoch": 1.3570767275633837,
      "grad_norm": 0.15909631550312042,
      "learning_rate": 5.4767381237157666e-05,
      "loss": 0.6217,
      "mean_token_accuracy": 0.8108632080256939,
      "num_tokens": 470769565.0,
      "step": 14750
    },
    {
      "entropy": 0.5977461459115148,
      "epoch": 1.357168733934825,
      "grad_norm": 0.15591953694820404,
      "learning_rate": 5.476431441101604e-05,
      "loss": 0.5785,
      "mean_token_accuracy": 0.8200635276734829,
      "num_tokens": 470801768.0,
      "step": 14751
    },
    {
      "entropy": 0.6532891346141696,
      "epoch": 1.3572607403062662,
      "grad_norm": 0.16403041779994965,
      "learning_rate": 5.476124758487442e-05,
      "loss": 0.6404,
      "mean_token_accuracy": 0.8029187843203545,
      "num_tokens": 470833639.0,
      "step": 14752
    },
    {
      "entropy": 0.7143934806808829,
      "epoch": 1.3573527466777073,
      "grad_norm": 0.16935120522975922,
      "learning_rate": 5.475818075873279e-05,
      "loss": 0.7034,
      "mean_token_accuracy": 0.7874809838831425,
      "num_tokens": 470864610.0,
      "step": 14753
    },
    {
      "entropy": 0.5987176350317895,
      "epoch": 1.3574447530491487,
      "grad_norm": 0.15136030316352844,
      "learning_rate": 5.4755113932591165e-05,
      "loss": 0.5939,
      "mean_token_accuracy": 0.8182382769882679,
      "num_tokens": 470896975.0,
      "step": 14754
    },
    {
      "entropy": 0.6217348836362362,
      "epoch": 1.3575367594205898,
      "grad_norm": 0.161088228225708,
      "learning_rate": 5.475204710644953e-05,
      "loss": 0.6226,
      "mean_token_accuracy": 0.811268113553524,
      "num_tokens": 470928792.0,
      "step": 14755
    },
    {
      "entropy": 0.7452399209141731,
      "epoch": 1.3576287657920312,
      "grad_norm": 0.1662234365940094,
      "learning_rate": 5.4748980280307914e-05,
      "loss": 0.7476,
      "mean_token_accuracy": 0.7735028304159641,
      "num_tokens": 470960955.0,
      "step": 14756
    },
    {
      "entropy": 0.6142778191715479,
      "epoch": 1.3577207721634723,
      "grad_norm": 0.16878865659236908,
      "learning_rate": 5.474591345416629e-05,
      "loss": 0.6035,
      "mean_token_accuracy": 0.8179073706269264,
      "num_tokens": 470992003.0,
      "step": 14757
    },
    {
      "entropy": 0.594173526391387,
      "epoch": 1.3578127785349134,
      "grad_norm": 0.15906153619289398,
      "learning_rate": 5.474284662802466e-05,
      "loss": 0.5798,
      "mean_token_accuracy": 0.8200754597783089,
      "num_tokens": 471024048.0,
      "step": 14758
    },
    {
      "entropy": 0.6447041928768158,
      "epoch": 1.3579047849063548,
      "grad_norm": 0.16550561785697937,
      "learning_rate": 5.473977980188303e-05,
      "loss": 0.6263,
      "mean_token_accuracy": 0.8103642985224724,
      "num_tokens": 471054299.0,
      "step": 14759
    },
    {
      "entropy": 0.5634053777903318,
      "epoch": 1.357996791277796,
      "grad_norm": 0.14741645753383636,
      "learning_rate": 5.473671297574141e-05,
      "loss": 0.5471,
      "mean_token_accuracy": 0.831094279885292,
      "num_tokens": 471087067.0,
      "step": 14760
    },
    {
      "entropy": 0.6746086124330759,
      "epoch": 1.3580887976492373,
      "grad_norm": 0.16668327152729034,
      "learning_rate": 5.473364614959978e-05,
      "loss": 0.6673,
      "mean_token_accuracy": 0.7938984408974648,
      "num_tokens": 471118229.0,
      "step": 14761
    },
    {
      "entropy": 0.5869448278099298,
      "epoch": 1.3581808040206784,
      "grad_norm": 0.15832731127738953,
      "learning_rate": 5.4730579323458155e-05,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8204493075609207,
      "num_tokens": 471150674.0,
      "step": 14762
    },
    {
      "entropy": 0.7792690265923738,
      "epoch": 1.3582728103921196,
      "grad_norm": 0.17224332690238953,
      "learning_rate": 5.472751249731652e-05,
      "loss": 0.7636,
      "mean_token_accuracy": 0.7689434476196766,
      "num_tokens": 471182503.0,
      "step": 14763
    },
    {
      "entropy": 0.5112734008580446,
      "epoch": 1.358364816763561,
      "grad_norm": 0.14958137273788452,
      "learning_rate": 5.4724445671174905e-05,
      "loss": 0.4826,
      "mean_token_accuracy": 0.8485530465841293,
      "num_tokens": 471212967.0,
      "step": 14764
    },
    {
      "entropy": 0.7285653427243233,
      "epoch": 1.358456823135002,
      "grad_norm": 0.16553862392902374,
      "learning_rate": 5.472137884503328e-05,
      "loss": 0.699,
      "mean_token_accuracy": 0.792279876768589,
      "num_tokens": 471244816.0,
      "step": 14765
    },
    {
      "entropy": 0.6202983055263758,
      "epoch": 1.3585488295064434,
      "grad_norm": 0.15813754498958588,
      "learning_rate": 5.471831201889165e-05,
      "loss": 0.6235,
      "mean_token_accuracy": 0.8113929033279419,
      "num_tokens": 471276534.0,
      "step": 14766
    },
    {
      "entropy": 0.465388341806829,
      "epoch": 1.3586408358778845,
      "grad_norm": 0.14026616513729095,
      "learning_rate": 5.471524519275003e-05,
      "loss": 0.4628,
      "mean_token_accuracy": 0.854987908154726,
      "num_tokens": 471309017.0,
      "step": 14767
    },
    {
      "entropy": 0.521997332572937,
      "epoch": 1.3587328422493257,
      "grad_norm": 0.1477198749780655,
      "learning_rate": 5.47121783666084e-05,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8378437981009483,
      "num_tokens": 471341337.0,
      "step": 14768
    },
    {
      "entropy": 0.5170130301266909,
      "epoch": 1.358824848620767,
      "grad_norm": 0.145376518368721,
      "learning_rate": 5.470911154046677e-05,
      "loss": 0.5082,
      "mean_token_accuracy": 0.8419976606965065,
      "num_tokens": 471373479.0,
      "step": 14769
    },
    {
      "entropy": 0.6247686194255948,
      "epoch": 1.3589168549922082,
      "grad_norm": 0.16640080511569977,
      "learning_rate": 5.4706044714325146e-05,
      "loss": 0.6129,
      "mean_token_accuracy": 0.8130148127675056,
      "num_tokens": 471405797.0,
      "step": 14770
    },
    {
      "entropy": 0.552773549221456,
      "epoch": 1.3590088613636495,
      "grad_norm": 0.1570340394973755,
      "learning_rate": 5.470297788818353e-05,
      "loss": 0.5514,
      "mean_token_accuracy": 0.8315332606434822,
      "num_tokens": 471438285.0,
      "step": 14771
    },
    {
      "entropy": 0.547497428022325,
      "epoch": 1.3591008677350906,
      "grad_norm": 0.1574798822402954,
      "learning_rate": 5.4699911062041895e-05,
      "loss": 0.5234,
      "mean_token_accuracy": 0.8379277177155018,
      "num_tokens": 471469337.0,
      "step": 14772
    },
    {
      "entropy": 0.6708313655108213,
      "epoch": 1.3591928741065318,
      "grad_norm": 0.16511660814285278,
      "learning_rate": 5.469684423590027e-05,
      "loss": 0.6705,
      "mean_token_accuracy": 0.7996749505400658,
      "num_tokens": 471501631.0,
      "step": 14773
    },
    {
      "entropy": 0.6148986304178834,
      "epoch": 1.3592848804779731,
      "grad_norm": 0.15127679705619812,
      "learning_rate": 5.469377740975864e-05,
      "loss": 0.6066,
      "mean_token_accuracy": 0.8118594400584698,
      "num_tokens": 471533799.0,
      "step": 14774
    },
    {
      "entropy": 0.5758499521762133,
      "epoch": 1.3593768868494143,
      "grad_norm": 0.15796971321105957,
      "learning_rate": 5.469071058361702e-05,
      "loss": 0.5715,
      "mean_token_accuracy": 0.8251705281436443,
      "num_tokens": 471565641.0,
      "step": 14775
    },
    {
      "entropy": 0.6336061786860228,
      "epoch": 1.3594688932208556,
      "grad_norm": 0.15916535258293152,
      "learning_rate": 5.4687643757475394e-05,
      "loss": 0.6361,
      "mean_token_accuracy": 0.811239380389452,
      "num_tokens": 471597474.0,
      "step": 14776
    },
    {
      "entropy": 0.7412385791540146,
      "epoch": 1.3595608995922968,
      "grad_norm": 0.16261285543441772,
      "learning_rate": 5.468457693133376e-05,
      "loss": 0.7255,
      "mean_token_accuracy": 0.7777695879340172,
      "num_tokens": 471629934.0,
      "step": 14777
    },
    {
      "entropy": 0.7093592849560082,
      "epoch": 1.359652905963738,
      "grad_norm": 0.1647273749113083,
      "learning_rate": 5.4681510105192136e-05,
      "loss": 0.715,
      "mean_token_accuracy": 0.7845073714852333,
      "num_tokens": 471661834.0,
      "step": 14778
    },
    {
      "entropy": 0.5974999479949474,
      "epoch": 1.3597449123351792,
      "grad_norm": 0.15183967351913452,
      "learning_rate": 5.467844327905052e-05,
      "loss": 0.5795,
      "mean_token_accuracy": 0.8205709159374237,
      "num_tokens": 471693789.0,
      "step": 14779
    },
    {
      "entropy": 0.542597041465342,
      "epoch": 1.3598369187066204,
      "grad_norm": 0.15290583670139313,
      "learning_rate": 5.4675376452908886e-05,
      "loss": 0.538,
      "mean_token_accuracy": 0.8350273370742798,
      "num_tokens": 471725799.0,
      "step": 14780
    },
    {
      "entropy": 0.5728659387677908,
      "epoch": 1.3599289250780617,
      "grad_norm": 0.15263420343399048,
      "learning_rate": 5.467230962676726e-05,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8349909111857414,
      "num_tokens": 471758334.0,
      "step": 14781
    },
    {
      "entropy": 0.6081976722925901,
      "epoch": 1.3600209314495029,
      "grad_norm": 0.16279800236225128,
      "learning_rate": 5.466924280062563e-05,
      "loss": 0.5866,
      "mean_token_accuracy": 0.8219171054661274,
      "num_tokens": 471789017.0,
      "step": 14782
    },
    {
      "entropy": 0.5885456884279847,
      "epoch": 1.360112937820944,
      "grad_norm": 0.15556079149246216,
      "learning_rate": 5.4666175974484016e-05,
      "loss": 0.5594,
      "mean_token_accuracy": 0.8278969638049603,
      "num_tokens": 471821084.0,
      "step": 14783
    },
    {
      "entropy": 0.6488203033804893,
      "epoch": 1.3602049441923854,
      "grad_norm": 0.15654350817203522,
      "learning_rate": 5.4663109148342384e-05,
      "loss": 0.6363,
      "mean_token_accuracy": 0.8053701184689999,
      "num_tokens": 471853278.0,
      "step": 14784
    },
    {
      "entropy": 0.605016715824604,
      "epoch": 1.3602969505638265,
      "grad_norm": 0.15789903700351715,
      "learning_rate": 5.466004232220075e-05,
      "loss": 0.6002,
      "mean_token_accuracy": 0.8178737238049507,
      "num_tokens": 471884958.0,
      "step": 14785
    },
    {
      "entropy": 0.5635826587677002,
      "epoch": 1.3603889569352678,
      "grad_norm": 0.15051047503948212,
      "learning_rate": 5.465697549605913e-05,
      "loss": 0.5387,
      "mean_token_accuracy": 0.8319021686911583,
      "num_tokens": 471916883.0,
      "step": 14786
    },
    {
      "entropy": 0.6193576762452722,
      "epoch": 1.360480963306709,
      "grad_norm": 0.16366122663021088,
      "learning_rate": 5.465390866991751e-05,
      "loss": 0.5946,
      "mean_token_accuracy": 0.812518049031496,
      "num_tokens": 471947943.0,
      "step": 14787
    },
    {
      "entropy": 0.5318534029647708,
      "epoch": 1.3605729696781501,
      "grad_norm": 0.21704436838626862,
      "learning_rate": 5.465084184377588e-05,
      "loss": 0.5202,
      "mean_token_accuracy": 0.8357645869255066,
      "num_tokens": 471979881.0,
      "step": 14788
    },
    {
      "entropy": 0.7622074596583843,
      "epoch": 1.3606649760495915,
      "grad_norm": 0.16176946461200714,
      "learning_rate": 5.464777501763425e-05,
      "loss": 0.7569,
      "mean_token_accuracy": 0.7692379616200924,
      "num_tokens": 472012124.0,
      "step": 14789
    },
    {
      "entropy": 0.5973674114793539,
      "epoch": 1.3607569824210326,
      "grad_norm": 0.15791481733322144,
      "learning_rate": 5.464470819149262e-05,
      "loss": 0.5874,
      "mean_token_accuracy": 0.817779790610075,
      "num_tokens": 472044502.0,
      "step": 14790
    },
    {
      "entropy": 0.5729891264345497,
      "epoch": 1.360848988792474,
      "grad_norm": 0.1686742901802063,
      "learning_rate": 5.464164136535101e-05,
      "loss": 0.5665,
      "mean_token_accuracy": 0.8268828168511391,
      "num_tokens": 472076625.0,
      "step": 14791
    },
    {
      "entropy": 0.6226290483027697,
      "epoch": 1.360940995163915,
      "grad_norm": 0.164840430021286,
      "learning_rate": 5.4638574539209375e-05,
      "loss": 0.6194,
      "mean_token_accuracy": 0.8127108477056026,
      "num_tokens": 472108249.0,
      "step": 14792
    },
    {
      "entropy": 0.6482587028294802,
      "epoch": 1.3610330015353562,
      "grad_norm": 0.15939822793006897,
      "learning_rate": 5.463550771306775e-05,
      "loss": 0.6287,
      "mean_token_accuracy": 0.8072500824928284,
      "num_tokens": 472140636.0,
      "step": 14793
    },
    {
      "entropy": 0.6288119489327073,
      "epoch": 1.3611250079067976,
      "grad_norm": 0.15551961958408356,
      "learning_rate": 5.463244088692612e-05,
      "loss": 0.6136,
      "mean_token_accuracy": 0.8118827790021896,
      "num_tokens": 472172858.0,
      "step": 14794
    },
    {
      "entropy": 0.6629222314804792,
      "epoch": 1.3612170142782387,
      "grad_norm": 0.1615617871284485,
      "learning_rate": 5.46293740607845e-05,
      "loss": 0.6464,
      "mean_token_accuracy": 0.801449604332447,
      "num_tokens": 472204814.0,
      "step": 14795
    },
    {
      "entropy": 0.6129262894392014,
      "epoch": 1.36130902064968,
      "grad_norm": 0.16570723056793213,
      "learning_rate": 5.462630723464287e-05,
      "loss": 0.5969,
      "mean_token_accuracy": 0.8152804374694824,
      "num_tokens": 472236072.0,
      "step": 14796
    },
    {
      "entropy": 0.5519015807658434,
      "epoch": 1.3614010270211212,
      "grad_norm": 0.15722595155239105,
      "learning_rate": 5.462324040850124e-05,
      "loss": 0.5375,
      "mean_token_accuracy": 0.8341149128973484,
      "num_tokens": 472267609.0,
      "step": 14797
    },
    {
      "entropy": 0.5828439081087708,
      "epoch": 1.3614930333925623,
      "grad_norm": 0.14858971536159515,
      "learning_rate": 5.462017358235962e-05,
      "loss": 0.5829,
      "mean_token_accuracy": 0.8219590447843075,
      "num_tokens": 472300043.0,
      "step": 14798
    },
    {
      "entropy": 0.6395940762013197,
      "epoch": 1.3615850397640037,
      "grad_norm": 0.16645562648773193,
      "learning_rate": 5.4617106756218e-05,
      "loss": 0.6275,
      "mean_token_accuracy": 0.8070606514811516,
      "num_tokens": 472331345.0,
      "step": 14799
    },
    {
      "entropy": 0.626725347712636,
      "epoch": 1.3616770461354448,
      "grad_norm": 0.15733511745929718,
      "learning_rate": 5.4614039930076365e-05,
      "loss": 0.6224,
      "mean_token_accuracy": 0.8117739297449589,
      "num_tokens": 472363859.0,
      "step": 14800
    },
    {
      "entropy": 0.6577073065564036,
      "epoch": 1.3617690525068862,
      "grad_norm": 0.1578184962272644,
      "learning_rate": 5.461097310393474e-05,
      "loss": 0.6384,
      "mean_token_accuracy": 0.8044055253267288,
      "num_tokens": 472396112.0,
      "step": 14801
    },
    {
      "entropy": 0.6715552322566509,
      "epoch": 1.3618610588783273,
      "grad_norm": 0.16469848155975342,
      "learning_rate": 5.460790627779312e-05,
      "loss": 0.6598,
      "mean_token_accuracy": 0.7971605025231838,
      "num_tokens": 472428106.0,
      "step": 14802
    },
    {
      "entropy": 0.5657972991466522,
      "epoch": 1.3619530652497684,
      "grad_norm": 0.16077539324760437,
      "learning_rate": 5.460483945165149e-05,
      "loss": 0.5463,
      "mean_token_accuracy": 0.8252173103392124,
      "num_tokens": 472459068.0,
      "step": 14803
    },
    {
      "entropy": 0.5970916757360101,
      "epoch": 1.3620450716212098,
      "grad_norm": 0.15882357954978943,
      "learning_rate": 5.4601772625509864e-05,
      "loss": 0.5908,
      "mean_token_accuracy": 0.818274449557066,
      "num_tokens": 472490518.0,
      "step": 14804
    },
    {
      "entropy": 0.5779086872935295,
      "epoch": 1.362137077992651,
      "grad_norm": 0.15716001391410828,
      "learning_rate": 5.459870579936823e-05,
      "loss": 0.5574,
      "mean_token_accuracy": 0.8257689848542213,
      "num_tokens": 472522360.0,
      "step": 14805
    },
    {
      "entropy": 0.7507755625993013,
      "epoch": 1.3622290843640923,
      "grad_norm": 0.1672755777835846,
      "learning_rate": 5.459563897322661e-05,
      "loss": 0.7375,
      "mean_token_accuracy": 0.7823594026267529,
      "num_tokens": 472554796.0,
      "step": 14806
    },
    {
      "entropy": 0.770779263228178,
      "epoch": 1.3623210907355334,
      "grad_norm": 0.17010219395160675,
      "learning_rate": 5.459257214708499e-05,
      "loss": 0.7481,
      "mean_token_accuracy": 0.7710510827600956,
      "num_tokens": 472586942.0,
      "step": 14807
    },
    {
      "entropy": 0.6110842265188694,
      "epoch": 1.3624130971069746,
      "grad_norm": 0.15969768166542053,
      "learning_rate": 5.4589505320943355e-05,
      "loss": 0.613,
      "mean_token_accuracy": 0.8129841051995754,
      "num_tokens": 472618910.0,
      "step": 14808
    },
    {
      "entropy": 0.6914944425225258,
      "epoch": 1.362505103478416,
      "grad_norm": 0.17817965149879456,
      "learning_rate": 5.458643849480173e-05,
      "loss": 0.685,
      "mean_token_accuracy": 0.7891444861888885,
      "num_tokens": 472650725.0,
      "step": 14809
    },
    {
      "entropy": 0.6087187435477972,
      "epoch": 1.362597109849857,
      "grad_norm": 0.15884554386138916,
      "learning_rate": 5.458337166866011e-05,
      "loss": 0.581,
      "mean_token_accuracy": 0.8181317448616028,
      "num_tokens": 472682031.0,
      "step": 14810
    },
    {
      "entropy": 0.6044431896880269,
      "epoch": 1.3626891162212984,
      "grad_norm": 0.15935151278972626,
      "learning_rate": 5.458030484251848e-05,
      "loss": 0.5876,
      "mean_token_accuracy": 0.8204010054469109,
      "num_tokens": 472713298.0,
      "step": 14811
    },
    {
      "entropy": 0.7552844695746899,
      "epoch": 1.3627811225927395,
      "grad_norm": 0.16841834783554077,
      "learning_rate": 5.4577238016376854e-05,
      "loss": 0.7568,
      "mean_token_accuracy": 0.7771924659609795,
      "num_tokens": 472745354.0,
      "step": 14812
    },
    {
      "entropy": 0.5653055720031261,
      "epoch": 1.3628731289641807,
      "grad_norm": 0.1554718017578125,
      "learning_rate": 5.457417119023522e-05,
      "loss": 0.561,
      "mean_token_accuracy": 0.8267281353473663,
      "num_tokens": 472778122.0,
      "step": 14813
    },
    {
      "entropy": 0.7246195208281279,
      "epoch": 1.362965135335622,
      "grad_norm": 0.16463547945022583,
      "learning_rate": 5.45711043640936e-05,
      "loss": 0.726,
      "mean_token_accuracy": 0.7783796004951,
      "num_tokens": 472809744.0,
      "step": 14814
    },
    {
      "entropy": 0.5811671689152718,
      "epoch": 1.3630571417070632,
      "grad_norm": 0.15145893394947052,
      "learning_rate": 5.456803753795198e-05,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8224967643618584,
      "num_tokens": 472842071.0,
      "step": 14815
    },
    {
      "entropy": 0.6244808631017804,
      "epoch": 1.3631491480785045,
      "grad_norm": 0.1602703183889389,
      "learning_rate": 5.4564970711810346e-05,
      "loss": 0.6223,
      "mean_token_accuracy": 0.8101966679096222,
      "num_tokens": 472873986.0,
      "step": 14816
    },
    {
      "entropy": 0.5334235671907663,
      "epoch": 1.3632411544499456,
      "grad_norm": 0.1542893797159195,
      "learning_rate": 5.456190388566872e-05,
      "loss": 0.5121,
      "mean_token_accuracy": 0.8449419178068638,
      "num_tokens": 472905665.0,
      "step": 14817
    },
    {
      "entropy": 0.5651484676636755,
      "epoch": 1.3633331608213868,
      "grad_norm": 0.16227369010448456,
      "learning_rate": 5.45588370595271e-05,
      "loss": 0.5381,
      "mean_token_accuracy": 0.8338333815336227,
      "num_tokens": 472936470.0,
      "step": 14818
    },
    {
      "entropy": 0.5998210329562426,
      "epoch": 1.3634251671928281,
      "grad_norm": 0.155211940407753,
      "learning_rate": 5.455577023338547e-05,
      "loss": 0.5952,
      "mean_token_accuracy": 0.8147284761071205,
      "num_tokens": 472968026.0,
      "step": 14819
    },
    {
      "entropy": 0.6152719920501113,
      "epoch": 1.3635171735642693,
      "grad_norm": 0.15782776474952698,
      "learning_rate": 5.4552703407243844e-05,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8116865493357182,
      "num_tokens": 473000547.0,
      "step": 14820
    },
    {
      "entropy": 0.7474411241710186,
      "epoch": 1.3636091799357106,
      "grad_norm": 0.17025691270828247,
      "learning_rate": 5.454963658110221e-05,
      "loss": 0.7379,
      "mean_token_accuracy": 0.7764107920229435,
      "num_tokens": 473032815.0,
      "step": 14821
    },
    {
      "entropy": 0.5787136591970921,
      "epoch": 1.3637011863071518,
      "grad_norm": 0.15294912457466125,
      "learning_rate": 5.4546569754960594e-05,
      "loss": 0.5665,
      "mean_token_accuracy": 0.8222388364374638,
      "num_tokens": 473065269.0,
      "step": 14822
    },
    {
      "entropy": 0.6405050735920668,
      "epoch": 1.363793192678593,
      "grad_norm": 0.16316409409046173,
      "learning_rate": 5.454350292881897e-05,
      "loss": 0.6268,
      "mean_token_accuracy": 0.8048049882054329,
      "num_tokens": 473097773.0,
      "step": 14823
    },
    {
      "entropy": 0.5936155766248703,
      "epoch": 1.3638851990500342,
      "grad_norm": 0.15177060663700104,
      "learning_rate": 5.4540436102677336e-05,
      "loss": 0.5633,
      "mean_token_accuracy": 0.8276866972446442,
      "num_tokens": 473130288.0,
      "step": 14824
    },
    {
      "entropy": 0.6258907606825233,
      "epoch": 1.3639772054214754,
      "grad_norm": 0.14889633655548096,
      "learning_rate": 5.453736927653571e-05,
      "loss": 0.6129,
      "mean_token_accuracy": 0.8099110014736652,
      "num_tokens": 473162622.0,
      "step": 14825
    },
    {
      "entropy": 0.597690284717828,
      "epoch": 1.3640692117929167,
      "grad_norm": 0.1616983264684677,
      "learning_rate": 5.453430245039409e-05,
      "loss": 0.5762,
      "mean_token_accuracy": 0.824566375464201,
      "num_tokens": 473194761.0,
      "step": 14826
    },
    {
      "entropy": 0.7348250951617956,
      "epoch": 1.3641612181643579,
      "grad_norm": 0.17054860293865204,
      "learning_rate": 5.453123562425247e-05,
      "loss": 0.7243,
      "mean_token_accuracy": 0.7805736996233463,
      "num_tokens": 473225982.0,
      "step": 14827
    },
    {
      "entropy": 0.6035751653835177,
      "epoch": 1.364253224535799,
      "grad_norm": 0.17113707959651947,
      "learning_rate": 5.4528168798110835e-05,
      "loss": 0.5941,
      "mean_token_accuracy": 0.819229606539011,
      "num_tokens": 473257939.0,
      "step": 14828
    },
    {
      "entropy": 0.6307136798277497,
      "epoch": 1.3643452309072404,
      "grad_norm": 0.1566263735294342,
      "learning_rate": 5.4525101971969216e-05,
      "loss": 0.616,
      "mean_token_accuracy": 0.8075053840875626,
      "num_tokens": 473289855.0,
      "step": 14829
    },
    {
      "entropy": 0.659479012247175,
      "epoch": 1.3644372372786815,
      "grad_norm": 0.15671740472316742,
      "learning_rate": 5.452203514582759e-05,
      "loss": 0.6446,
      "mean_token_accuracy": 0.8007677644491196,
      "num_tokens": 473322280.0,
      "step": 14830
    },
    {
      "entropy": 0.6417530979961157,
      "epoch": 1.3645292436501228,
      "grad_norm": 0.16756847500801086,
      "learning_rate": 5.451896831968596e-05,
      "loss": 0.6532,
      "mean_token_accuracy": 0.8031021952629089,
      "num_tokens": 473353051.0,
      "step": 14831
    },
    {
      "entropy": 0.6309918449260294,
      "epoch": 1.364621250021564,
      "grad_norm": 0.16399738192558289,
      "learning_rate": 5.451590149354433e-05,
      "loss": 0.632,
      "mean_token_accuracy": 0.8074105083942413,
      "num_tokens": 473384610.0,
      "step": 14832
    },
    {
      "entropy": 0.721406452357769,
      "epoch": 1.3647132563930051,
      "grad_norm": 0.17403919994831085,
      "learning_rate": 5.4512834667402715e-05,
      "loss": 0.7291,
      "mean_token_accuracy": 0.7814735472202301,
      "num_tokens": 473416081.0,
      "step": 14833
    },
    {
      "entropy": 0.6250427290797234,
      "epoch": 1.3648052627644465,
      "grad_norm": 0.15554864704608917,
      "learning_rate": 5.450976784126108e-05,
      "loss": 0.6195,
      "mean_token_accuracy": 0.8086059018969536,
      "num_tokens": 473447406.0,
      "step": 14834
    },
    {
      "entropy": 0.6158986212685704,
      "epoch": 1.3648972691358876,
      "grad_norm": 0.1541067361831665,
      "learning_rate": 5.450670101511946e-05,
      "loss": 0.59,
      "mean_token_accuracy": 0.8198179826140404,
      "num_tokens": 473478997.0,
      "step": 14835
    },
    {
      "entropy": 0.6639467868953943,
      "epoch": 1.364989275507329,
      "grad_norm": 0.1571158915758133,
      "learning_rate": 5.4503634188977825e-05,
      "loss": 0.6362,
      "mean_token_accuracy": 0.8054708652198315,
      "num_tokens": 473511064.0,
      "step": 14836
    },
    {
      "entropy": 0.5761940656229854,
      "epoch": 1.36508128187877,
      "grad_norm": 0.15345630049705505,
      "learning_rate": 5.4500567362836207e-05,
      "loss": 0.5696,
      "mean_token_accuracy": 0.8216156549751759,
      "num_tokens": 473542875.0,
      "step": 14837
    },
    {
      "entropy": 0.7047865185886621,
      "epoch": 1.3651732882502112,
      "grad_norm": 0.17281481623649597,
      "learning_rate": 5.449750053669458e-05,
      "loss": 0.6987,
      "mean_token_accuracy": 0.7901402041316032,
      "num_tokens": 473573465.0,
      "step": 14838
    },
    {
      "entropy": 0.693654359318316,
      "epoch": 1.3652652946216526,
      "grad_norm": 0.1624569147825241,
      "learning_rate": 5.449443371055295e-05,
      "loss": 0.6653,
      "mean_token_accuracy": 0.7953478805720806,
      "num_tokens": 473605555.0,
      "step": 14839
    },
    {
      "entropy": 0.619381190277636,
      "epoch": 1.3653573009930937,
      "grad_norm": 0.15694975852966309,
      "learning_rate": 5.4491366884411324e-05,
      "loss": 0.6115,
      "mean_token_accuracy": 0.813109777867794,
      "num_tokens": 473637302.0,
      "step": 14840
    },
    {
      "entropy": 0.5310512036085129,
      "epoch": 1.365449307364535,
      "grad_norm": 0.15617065131664276,
      "learning_rate": 5.4488300058269705e-05,
      "loss": 0.5275,
      "mean_token_accuracy": 0.8400905467569828,
      "num_tokens": 473668608.0,
      "step": 14841
    },
    {
      "entropy": 0.7432443480938673,
      "epoch": 1.3655413137359762,
      "grad_norm": 0.16631552577018738,
      "learning_rate": 5.448523323212807e-05,
      "loss": 0.7407,
      "mean_token_accuracy": 0.7785496115684509,
      "num_tokens": 473700699.0,
      "step": 14842
    },
    {
      "entropy": 0.6606606794521213,
      "epoch": 1.3656333201074173,
      "grad_norm": 0.16075436770915985,
      "learning_rate": 5.448216640598645e-05,
      "loss": 0.6683,
      "mean_token_accuracy": 0.7969315014779568,
      "num_tokens": 473731916.0,
      "step": 14843
    },
    {
      "entropy": 0.6127339508384466,
      "epoch": 1.3657253264788587,
      "grad_norm": 0.15022720396518707,
      "learning_rate": 5.4479099579844816e-05,
      "loss": 0.5884,
      "mean_token_accuracy": 0.8194640278816223,
      "num_tokens": 473764348.0,
      "step": 14844
    },
    {
      "entropy": 0.6366215832531452,
      "epoch": 1.3658173328502998,
      "grad_norm": 0.1548628807067871,
      "learning_rate": 5.44760327537032e-05,
      "loss": 0.6273,
      "mean_token_accuracy": 0.8077674508094788,
      "num_tokens": 473797116.0,
      "step": 14845
    },
    {
      "entropy": 0.6105389334261417,
      "epoch": 1.3659093392217412,
      "grad_norm": 0.1574733555316925,
      "learning_rate": 5.447296592756157e-05,
      "loss": 0.6019,
      "mean_token_accuracy": 0.8173138946294785,
      "num_tokens": 473829827.0,
      "step": 14846
    },
    {
      "entropy": 0.5806025005877018,
      "epoch": 1.3660013455931823,
      "grad_norm": 0.15756674110889435,
      "learning_rate": 5.446989910141994e-05,
      "loss": 0.572,
      "mean_token_accuracy": 0.8220369555056095,
      "num_tokens": 473861697.0,
      "step": 14847
    },
    {
      "entropy": 0.6067609824240208,
      "epoch": 1.3660933519646234,
      "grad_norm": 0.15545949339866638,
      "learning_rate": 5.4466832275278314e-05,
      "loss": 0.5947,
      "mean_token_accuracy": 0.8138863630592823,
      "num_tokens": 473894089.0,
      "step": 14848
    },
    {
      "entropy": 0.6446395460516214,
      "epoch": 1.3661853583360648,
      "grad_norm": 0.1576579511165619,
      "learning_rate": 5.4463765449136696e-05,
      "loss": 0.6295,
      "mean_token_accuracy": 0.8119549788534641,
      "num_tokens": 473926044.0,
      "step": 14849
    },
    {
      "entropy": 0.6574149681255221,
      "epoch": 1.366277364707506,
      "grad_norm": 0.16167402267456055,
      "learning_rate": 5.4460698622995063e-05,
      "loss": 0.6584,
      "mean_token_accuracy": 0.8026246316730976,
      "num_tokens": 473957404.0,
      "step": 14850
    },
    {
      "entropy": 0.6689197737723589,
      "epoch": 1.3663693710789473,
      "grad_norm": 0.15919630229473114,
      "learning_rate": 5.445763179685344e-05,
      "loss": 0.67,
      "mean_token_accuracy": 0.7971770726144314,
      "num_tokens": 473989575.0,
      "step": 14851
    },
    {
      "entropy": 0.653816414065659,
      "epoch": 1.3664613774503884,
      "grad_norm": 0.15797017514705658,
      "learning_rate": 5.4454564970711806e-05,
      "loss": 0.6344,
      "mean_token_accuracy": 0.8058624416589737,
      "num_tokens": 474021589.0,
      "step": 14852
    },
    {
      "entropy": 0.5922977151349187,
      "epoch": 1.3665533838218296,
      "grad_norm": 0.15488219261169434,
      "learning_rate": 5.445149814457019e-05,
      "loss": 0.5688,
      "mean_token_accuracy": 0.8212926797568798,
      "num_tokens": 474053245.0,
      "step": 14853
    },
    {
      "entropy": 0.6246970165520906,
      "epoch": 1.366645390193271,
      "grad_norm": 0.15885908901691437,
      "learning_rate": 5.444843131842856e-05,
      "loss": 0.6171,
      "mean_token_accuracy": 0.8122226521372795,
      "num_tokens": 474084616.0,
      "step": 14854
    },
    {
      "entropy": 0.5374174825847149,
      "epoch": 1.366737396564712,
      "grad_norm": 0.15372520685195923,
      "learning_rate": 5.444536449228693e-05,
      "loss": 0.5156,
      "mean_token_accuracy": 0.8418491035699844,
      "num_tokens": 474115928.0,
      "step": 14855
    },
    {
      "entropy": 0.6552369887940586,
      "epoch": 1.3668294029361534,
      "grad_norm": 0.16396315395832062,
      "learning_rate": 5.444229766614531e-05,
      "loss": 0.6333,
      "mean_token_accuracy": 0.8084022663533688,
      "num_tokens": 474147787.0,
      "step": 14856
    },
    {
      "entropy": 0.5074883843772113,
      "epoch": 1.3669214093075945,
      "grad_norm": 0.1615222990512848,
      "learning_rate": 5.4439230840003686e-05,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8457372933626175,
      "num_tokens": 474179073.0,
      "step": 14857
    },
    {
      "entropy": 0.6052651219069958,
      "epoch": 1.3670134156790357,
      "grad_norm": 0.15707868337631226,
      "learning_rate": 5.4436164013862054e-05,
      "loss": 0.5886,
      "mean_token_accuracy": 0.8184614107012749,
      "num_tokens": 474211157.0,
      "step": 14858
    },
    {
      "entropy": 0.5801599398255348,
      "epoch": 1.367105422050477,
      "grad_norm": 0.157980278134346,
      "learning_rate": 5.443309718772043e-05,
      "loss": 0.561,
      "mean_token_accuracy": 0.8277778401970863,
      "num_tokens": 474242850.0,
      "step": 14859
    },
    {
      "entropy": 0.611263089813292,
      "epoch": 1.3671974284219182,
      "grad_norm": 0.16172994673252106,
      "learning_rate": 5.443003036157881e-05,
      "loss": 0.5995,
      "mean_token_accuracy": 0.8185182586312294,
      "num_tokens": 474274621.0,
      "step": 14860
    },
    {
      "entropy": 0.6083276364952326,
      "epoch": 1.3672894347933595,
      "grad_norm": 0.16299006342887878,
      "learning_rate": 5.442696353543718e-05,
      "loss": 0.6001,
      "mean_token_accuracy": 0.8140943571925163,
      "num_tokens": 474305844.0,
      "step": 14861
    },
    {
      "entropy": 0.5073516108095646,
      "epoch": 1.3673814411648006,
      "grad_norm": 0.15028104186058044,
      "learning_rate": 5.442389670929555e-05,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8485007025301456,
      "num_tokens": 474338023.0,
      "step": 14862
    },
    {
      "entropy": 0.7381275799125433,
      "epoch": 1.3674734475362418,
      "grad_norm": 0.17506396770477295,
      "learning_rate": 5.442082988315392e-05,
      "loss": 0.7284,
      "mean_token_accuracy": 0.7773973718285561,
      "num_tokens": 474370101.0,
      "step": 14863
    },
    {
      "entropy": 0.6185234794393182,
      "epoch": 1.3675654539076831,
      "grad_norm": 0.1634891927242279,
      "learning_rate": 5.441776305701231e-05,
      "loss": 0.5837,
      "mean_token_accuracy": 0.8183968737721443,
      "num_tokens": 474401788.0,
      "step": 14864
    },
    {
      "entropy": 0.5785906105302274,
      "epoch": 1.3676574602791243,
      "grad_norm": 0.1476881355047226,
      "learning_rate": 5.4414696230870676e-05,
      "loss": 0.562,
      "mean_token_accuracy": 0.8252273499965668,
      "num_tokens": 474433566.0,
      "step": 14865
    },
    {
      "entropy": 0.5527751832269132,
      "epoch": 1.3677494666505656,
      "grad_norm": 0.14643536508083344,
      "learning_rate": 5.4411629404729044e-05,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8305991739034653,
      "num_tokens": 474466178.0,
      "step": 14866
    },
    {
      "entropy": 0.5947403032332659,
      "epoch": 1.3678414730220068,
      "grad_norm": 0.15524381399154663,
      "learning_rate": 5.440856257858742e-05,
      "loss": 0.5817,
      "mean_token_accuracy": 0.8233547173440456,
      "num_tokens": 474498104.0,
      "step": 14867
    },
    {
      "entropy": 0.7445166651159525,
      "epoch": 1.367933479393448,
      "grad_norm": 0.17035000026226044,
      "learning_rate": 5.44054957524458e-05,
      "loss": 0.7495,
      "mean_token_accuracy": 0.7758444994688034,
      "num_tokens": 474530087.0,
      "step": 14868
    },
    {
      "entropy": 0.5506325522437692,
      "epoch": 1.3680254857648892,
      "grad_norm": 0.1537391096353531,
      "learning_rate": 5.4402428926304175e-05,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8325963169336319,
      "num_tokens": 474562593.0,
      "step": 14869
    },
    {
      "entropy": 0.5940755093470216,
      "epoch": 1.3681174921363304,
      "grad_norm": 0.16118550300598145,
      "learning_rate": 5.439936210016254e-05,
      "loss": 0.5842,
      "mean_token_accuracy": 0.8225292377173901,
      "num_tokens": 474595100.0,
      "step": 14870
    },
    {
      "entropy": 0.6428136462345719,
      "epoch": 1.3682094985077717,
      "grad_norm": 0.1632632613182068,
      "learning_rate": 5.439629527402092e-05,
      "loss": 0.6264,
      "mean_token_accuracy": 0.8114411905407906,
      "num_tokens": 474626358.0,
      "step": 14871
    },
    {
      "entropy": 0.5930825788527727,
      "epoch": 1.3683015048792129,
      "grad_norm": 0.15263473987579346,
      "learning_rate": 5.43932284478793e-05,
      "loss": 0.584,
      "mean_token_accuracy": 0.8228264302015305,
      "num_tokens": 474658131.0,
      "step": 14872
    },
    {
      "entropy": 0.6010708883404732,
      "epoch": 1.368393511250654,
      "grad_norm": 0.15749865770339966,
      "learning_rate": 5.439016162173767e-05,
      "loss": 0.6066,
      "mean_token_accuracy": 0.8175395242869854,
      "num_tokens": 474690201.0,
      "step": 14873
    },
    {
      "entropy": 0.5901040313765407,
      "epoch": 1.3684855176220954,
      "grad_norm": 0.15229637920856476,
      "learning_rate": 5.438709479559604e-05,
      "loss": 0.5704,
      "mean_token_accuracy": 0.8185281231999397,
      "num_tokens": 474722002.0,
      "step": 14874
    },
    {
      "entropy": 0.6739374212920666,
      "epoch": 1.3685775239935365,
      "grad_norm": 0.16616389155387878,
      "learning_rate": 5.438402796945441e-05,
      "loss": 0.6752,
      "mean_token_accuracy": 0.7922751344740391,
      "num_tokens": 474754215.0,
      "step": 14875
    },
    {
      "entropy": 0.5694593265652657,
      "epoch": 1.3686695303649778,
      "grad_norm": 0.16645480692386627,
      "learning_rate": 5.438096114331279e-05,
      "loss": 0.5481,
      "mean_token_accuracy": 0.8357750996947289,
      "num_tokens": 474785884.0,
      "step": 14876
    },
    {
      "entropy": 0.5898637496866286,
      "epoch": 1.368761536736419,
      "grad_norm": 0.15699559450149536,
      "learning_rate": 5.4377894317171165e-05,
      "loss": 0.5778,
      "mean_token_accuracy": 0.8188724368810654,
      "num_tokens": 474817593.0,
      "step": 14877
    },
    {
      "entropy": 0.6675460822880268,
      "epoch": 1.3688535431078601,
      "grad_norm": 0.15829120576381683,
      "learning_rate": 5.437482749102953e-05,
      "loss": 0.6757,
      "mean_token_accuracy": 0.8008308447897434,
      "num_tokens": 474849100.0,
      "step": 14878
    },
    {
      "entropy": 0.5381361320614815,
      "epoch": 1.3689455494793015,
      "grad_norm": 0.14896848797798157,
      "learning_rate": 5.437176066488791e-05,
      "loss": 0.5344,
      "mean_token_accuracy": 0.8368161208927631,
      "num_tokens": 474880532.0,
      "step": 14879
    },
    {
      "entropy": 0.590598470531404,
      "epoch": 1.3690375558507426,
      "grad_norm": 0.15709273517131805,
      "learning_rate": 5.436869383874629e-05,
      "loss": 0.5887,
      "mean_token_accuracy": 0.820056289434433,
      "num_tokens": 474913093.0,
      "step": 14880
    },
    {
      "entropy": 0.5781155792064965,
      "epoch": 1.369129562222184,
      "grad_norm": 0.14672799408435822,
      "learning_rate": 5.436562701260466e-05,
      "loss": 0.5556,
      "mean_token_accuracy": 0.8264653161168098,
      "num_tokens": 474944935.0,
      "step": 14881
    },
    {
      "entropy": 0.6326858196407557,
      "epoch": 1.369221568593625,
      "grad_norm": 0.15443412959575653,
      "learning_rate": 5.436256018646303e-05,
      "loss": 0.6148,
      "mean_token_accuracy": 0.8104564808309078,
      "num_tokens": 474976404.0,
      "step": 14882
    },
    {
      "entropy": 0.6788447946310043,
      "epoch": 1.3693135749650662,
      "grad_norm": 0.16754133999347687,
      "learning_rate": 5.43594933603214e-05,
      "loss": 0.669,
      "mean_token_accuracy": 0.7969121672213078,
      "num_tokens": 475008197.0,
      "step": 14883
    },
    {
      "entropy": 0.574551367200911,
      "epoch": 1.3694055813365076,
      "grad_norm": 0.17190413177013397,
      "learning_rate": 5.435642653417978e-05,
      "loss": 0.5629,
      "mean_token_accuracy": 0.8290866613388062,
      "num_tokens": 475038692.0,
      "step": 14884
    },
    {
      "entropy": 0.628354037180543,
      "epoch": 1.3694975877079487,
      "grad_norm": 0.15953034162521362,
      "learning_rate": 5.4353359708038156e-05,
      "loss": 0.6045,
      "mean_token_accuracy": 0.8165908046066761,
      "num_tokens": 475070950.0,
      "step": 14885
    },
    {
      "entropy": 0.5960626723244786,
      "epoch": 1.36958959407939,
      "grad_norm": 0.1487419307231903,
      "learning_rate": 5.4350292881896524e-05,
      "loss": 0.5904,
      "mean_token_accuracy": 0.8206704929471016,
      "num_tokens": 475103421.0,
      "step": 14886
    },
    {
      "entropy": 0.7228640057146549,
      "epoch": 1.3696816004508312,
      "grad_norm": 0.16579246520996094,
      "learning_rate": 5.4347226055754905e-05,
      "loss": 0.7162,
      "mean_token_accuracy": 0.7832732535898685,
      "num_tokens": 475134998.0,
      "step": 14887
    },
    {
      "entropy": 0.71418615616858,
      "epoch": 1.3697736068222723,
      "grad_norm": 0.1615007519721985,
      "learning_rate": 5.434415922961328e-05,
      "loss": 0.6992,
      "mean_token_accuracy": 0.7893525697290897,
      "num_tokens": 475167166.0,
      "step": 14888
    },
    {
      "entropy": 0.5117521500214934,
      "epoch": 1.3698656131937137,
      "grad_norm": 0.14506462216377258,
      "learning_rate": 5.434109240347165e-05,
      "loss": 0.4968,
      "mean_token_accuracy": 0.8450639732182026,
      "num_tokens": 475198752.0,
      "step": 14889
    },
    {
      "entropy": 0.6173694618046284,
      "epoch": 1.3699576195651548,
      "grad_norm": 0.15831980109214783,
      "learning_rate": 5.433802557733002e-05,
      "loss": 0.5987,
      "mean_token_accuracy": 0.8133906088769436,
      "num_tokens": 475230510.0,
      "step": 14890
    },
    {
      "entropy": 0.6708178520202637,
      "epoch": 1.3700496259365962,
      "grad_norm": 0.15979434549808502,
      "learning_rate": 5.4334958751188404e-05,
      "loss": 0.6646,
      "mean_token_accuracy": 0.7943404279649258,
      "num_tokens": 475262087.0,
      "step": 14891
    },
    {
      "entropy": 0.7762134876102209,
      "epoch": 1.3701416323080373,
      "grad_norm": 0.16818411648273468,
      "learning_rate": 5.433189192504677e-05,
      "loss": 0.7672,
      "mean_token_accuracy": 0.7687661908566952,
      "num_tokens": 475294538.0,
      "step": 14892
    },
    {
      "entropy": 0.6628153277561069,
      "epoch": 1.3702336386794784,
      "grad_norm": 0.15749314427375793,
      "learning_rate": 5.4328825098905146e-05,
      "loss": 0.6591,
      "mean_token_accuracy": 0.797665748745203,
      "num_tokens": 475326268.0,
      "step": 14893
    },
    {
      "entropy": 0.5957813560962677,
      "epoch": 1.3703256450509198,
      "grad_norm": 0.15724897384643555,
      "learning_rate": 5.4325758272763514e-05,
      "loss": 0.5791,
      "mean_token_accuracy": 0.8233277834951878,
      "num_tokens": 475358278.0,
      "step": 14894
    },
    {
      "entropy": 0.657971803098917,
      "epoch": 1.370417651422361,
      "grad_norm": 0.1700311303138733,
      "learning_rate": 5.4322691446621895e-05,
      "loss": 0.6686,
      "mean_token_accuracy": 0.7960115037858486,
      "num_tokens": 475390776.0,
      "step": 14895
    },
    {
      "entropy": 0.5743267592042685,
      "epoch": 1.3705096577938023,
      "grad_norm": 0.15865685045719147,
      "learning_rate": 5.431962462048027e-05,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8271050825715065,
      "num_tokens": 475421162.0,
      "step": 14896
    },
    {
      "entropy": 0.5670245569199324,
      "epoch": 1.3706016641652434,
      "grad_norm": 0.1550329327583313,
      "learning_rate": 5.431655779433864e-05,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8325482569634914,
      "num_tokens": 475452804.0,
      "step": 14897
    },
    {
      "entropy": 0.5114805037155747,
      "epoch": 1.3706936705366846,
      "grad_norm": 0.14961715042591095,
      "learning_rate": 5.431349096819701e-05,
      "loss": 0.5044,
      "mean_token_accuracy": 0.844264879822731,
      "num_tokens": 475485304.0,
      "step": 14898
    },
    {
      "entropy": 0.6827250346541405,
      "epoch": 1.370785676908126,
      "grad_norm": 0.15722383558750153,
      "learning_rate": 5.4310424142055394e-05,
      "loss": 0.6679,
      "mean_token_accuracy": 0.7977500222623348,
      "num_tokens": 475517836.0,
      "step": 14899
    },
    {
      "entropy": 0.6422661524266005,
      "epoch": 1.370877683279567,
      "grad_norm": 0.15578578412532806,
      "learning_rate": 5.430735731591376e-05,
      "loss": 0.6297,
      "mean_token_accuracy": 0.8039793968200684,
      "num_tokens": 475549698.0,
      "step": 14900
    },
    {
      "entropy": 0.6488235425204039,
      "epoch": 1.3709696896510084,
      "grad_norm": 0.15675760805606842,
      "learning_rate": 5.4304290489772137e-05,
      "loss": 0.643,
      "mean_token_accuracy": 0.803089801222086,
      "num_tokens": 475582373.0,
      "step": 14901
    },
    {
      "entropy": 0.6622053664177656,
      "epoch": 1.3710616960224495,
      "grad_norm": 0.159122496843338,
      "learning_rate": 5.4301223663630504e-05,
      "loss": 0.6394,
      "mean_token_accuracy": 0.7978577241301537,
      "num_tokens": 475613703.0,
      "step": 14902
    },
    {
      "entropy": 0.7200161423534155,
      "epoch": 1.3711537023938907,
      "grad_norm": 0.16966088116168976,
      "learning_rate": 5.429815683748889e-05,
      "loss": 0.7087,
      "mean_token_accuracy": 0.7868615873157978,
      "num_tokens": 475645559.0,
      "step": 14903
    },
    {
      "entropy": 0.6032181028276682,
      "epoch": 1.371245708765332,
      "grad_norm": 0.1498129814863205,
      "learning_rate": 5.429509001134726e-05,
      "loss": 0.5933,
      "mean_token_accuracy": 0.8175404220819473,
      "num_tokens": 475677741.0,
      "step": 14904
    },
    {
      "entropy": 0.6834641979075968,
      "epoch": 1.3713377151367732,
      "grad_norm": 0.15917041897773743,
      "learning_rate": 5.429202318520563e-05,
      "loss": 0.6695,
      "mean_token_accuracy": 0.7986400201916695,
      "num_tokens": 475709650.0,
      "step": 14905
    },
    {
      "entropy": 0.6026671696454287,
      "epoch": 1.3714297215082145,
      "grad_norm": 0.16216519474983215,
      "learning_rate": 5.4288956359064e-05,
      "loss": 0.5918,
      "mean_token_accuracy": 0.8184634521603584,
      "num_tokens": 475741558.0,
      "step": 14906
    },
    {
      "entropy": 0.6627792250365019,
      "epoch": 1.3715217278796556,
      "grad_norm": 0.165084570646286,
      "learning_rate": 5.4285889532922384e-05,
      "loss": 0.6593,
      "mean_token_accuracy": 0.7990543693304062,
      "num_tokens": 475773563.0,
      "step": 14907
    },
    {
      "entropy": 0.5404683882370591,
      "epoch": 1.3716137342510968,
      "grad_norm": 0.15504030883312225,
      "learning_rate": 5.428282270678076e-05,
      "loss": 0.5168,
      "mean_token_accuracy": 0.8398977555334568,
      "num_tokens": 475805489.0,
      "step": 14908
    },
    {
      "entropy": 0.6037964820861816,
      "epoch": 1.3717057406225381,
      "grad_norm": 0.16052764654159546,
      "learning_rate": 5.427975588063913e-05,
      "loss": 0.5987,
      "mean_token_accuracy": 0.8144530653953552,
      "num_tokens": 475837563.0,
      "step": 14909
    },
    {
      "entropy": 0.6372378915548325,
      "epoch": 1.3717977469939793,
      "grad_norm": 0.15905533730983734,
      "learning_rate": 5.4276689054497495e-05,
      "loss": 0.6196,
      "mean_token_accuracy": 0.8087698668241501,
      "num_tokens": 475869937.0,
      "step": 14910
    },
    {
      "entropy": 0.6649472638964653,
      "epoch": 1.3718897533654206,
      "grad_norm": 0.15856100618839264,
      "learning_rate": 5.427362222835588e-05,
      "loss": 0.6498,
      "mean_token_accuracy": 0.8033129535615444,
      "num_tokens": 475901742.0,
      "step": 14911
    },
    {
      "entropy": 0.6078518005087972,
      "epoch": 1.3719817597368618,
      "grad_norm": 0.16381117701530457,
      "learning_rate": 5.427055540221425e-05,
      "loss": 0.6184,
      "mean_token_accuracy": 0.8114999309182167,
      "num_tokens": 475932873.0,
      "step": 14912
    },
    {
      "entropy": 0.6511949338018894,
      "epoch": 1.372073766108303,
      "grad_norm": 0.16267450153827667,
      "learning_rate": 5.4267488576072626e-05,
      "loss": 0.6412,
      "mean_token_accuracy": 0.8020018748939037,
      "num_tokens": 475963824.0,
      "step": 14913
    },
    {
      "entropy": 0.5571130737662315,
      "epoch": 1.3721657724797443,
      "grad_norm": 0.1556788831949234,
      "learning_rate": 5.4264421749930993e-05,
      "loss": 0.5397,
      "mean_token_accuracy": 0.8266228996217251,
      "num_tokens": 475995770.0,
      "step": 14914
    },
    {
      "entropy": 0.6134438598528504,
      "epoch": 1.3722577788511854,
      "grad_norm": 0.16849982738494873,
      "learning_rate": 5.4261354923789375e-05,
      "loss": 0.5956,
      "mean_token_accuracy": 0.82152059674263,
      "num_tokens": 476026949.0,
      "step": 14915
    },
    {
      "entropy": 0.6822099164128304,
      "epoch": 1.3723497852226267,
      "grad_norm": 0.1673053801059723,
      "learning_rate": 5.425828809764775e-05,
      "loss": 0.6674,
      "mean_token_accuracy": 0.7990541867911816,
      "num_tokens": 476059103.0,
      "step": 14916
    },
    {
      "entropy": 0.6646579457446933,
      "epoch": 1.3724417915940679,
      "grad_norm": 0.16281060874462128,
      "learning_rate": 5.425522127150612e-05,
      "loss": 0.6595,
      "mean_token_accuracy": 0.8028423897922039,
      "num_tokens": 476090585.0,
      "step": 14917
    },
    {
      "entropy": 0.5738340057432652,
      "epoch": 1.372533797965509,
      "grad_norm": 0.15015539526939392,
      "learning_rate": 5.42521544453645e-05,
      "loss": 0.5575,
      "mean_token_accuracy": 0.8264924101531506,
      "num_tokens": 476121733.0,
      "step": 14918
    },
    {
      "entropy": 0.7694063372910023,
      "epoch": 1.3726258043369504,
      "grad_norm": 0.16549956798553467,
      "learning_rate": 5.4249087619222873e-05,
      "loss": 0.748,
      "mean_token_accuracy": 0.7704296670854092,
      "num_tokens": 476153041.0,
      "step": 14919
    },
    {
      "entropy": 0.5457449201494455,
      "epoch": 1.3727178107083915,
      "grad_norm": 0.147752583026886,
      "learning_rate": 5.424602079308124e-05,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8399834744632244,
      "num_tokens": 476185388.0,
      "step": 14920
    },
    {
      "entropy": 0.6684708949178457,
      "epoch": 1.3728098170798329,
      "grad_norm": 0.15777859091758728,
      "learning_rate": 5.4242953966939616e-05,
      "loss": 0.6414,
      "mean_token_accuracy": 0.8052492886781693,
      "num_tokens": 476217645.0,
      "step": 14921
    },
    {
      "entropy": 0.6173944545444101,
      "epoch": 1.372901823451274,
      "grad_norm": 0.1578381359577179,
      "learning_rate": 5.4239887140798e-05,
      "loss": 0.6022,
      "mean_token_accuracy": 0.8144170418381691,
      "num_tokens": 476249929.0,
      "step": 14922
    },
    {
      "entropy": 0.6773587320931256,
      "epoch": 1.3729938298227151,
      "grad_norm": 0.15703195333480835,
      "learning_rate": 5.4236820314656365e-05,
      "loss": 0.6712,
      "mean_token_accuracy": 0.795068085193634,
      "num_tokens": 476281974.0,
      "step": 14923
    },
    {
      "entropy": 0.6438519507646561,
      "epoch": 1.3730858361941565,
      "grad_norm": 0.15941546857357025,
      "learning_rate": 5.423375348851474e-05,
      "loss": 0.6321,
      "mean_token_accuracy": 0.8076662495732307,
      "num_tokens": 476314345.0,
      "step": 14924
    },
    {
      "entropy": 0.6668091854080558,
      "epoch": 1.3731778425655976,
      "grad_norm": 0.1667299121618271,
      "learning_rate": 5.423068666237311e-05,
      "loss": 0.6684,
      "mean_token_accuracy": 0.7994198277592659,
      "num_tokens": 476346037.0,
      "step": 14925
    },
    {
      "entropy": 0.5572089804336429,
      "epoch": 1.373269848937039,
      "grad_norm": 0.1539291888475418,
      "learning_rate": 5.422761983623149e-05,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8326855711638927,
      "num_tokens": 476377988.0,
      "step": 14926
    },
    {
      "entropy": 0.695832371711731,
      "epoch": 1.37336185530848,
      "grad_norm": 0.16865475475788116,
      "learning_rate": 5.4224553010089864e-05,
      "loss": 0.7064,
      "mean_token_accuracy": 0.784368671476841,
      "num_tokens": 476410078.0,
      "step": 14927
    },
    {
      "entropy": 0.6296723233535886,
      "epoch": 1.3734538616799212,
      "grad_norm": 0.16721509397029877,
      "learning_rate": 5.422148618394823e-05,
      "loss": 0.6169,
      "mean_token_accuracy": 0.8149751052260399,
      "num_tokens": 476441972.0,
      "step": 14928
    },
    {
      "entropy": 0.5915725566446781,
      "epoch": 1.3735458680513626,
      "grad_norm": 0.15250064432621002,
      "learning_rate": 5.4218419357806606e-05,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8235325664281845,
      "num_tokens": 476473781.0,
      "step": 14929
    },
    {
      "entropy": 0.45378031535074115,
      "epoch": 1.3736378744228037,
      "grad_norm": 0.14577716588974,
      "learning_rate": 5.421535253166499e-05,
      "loss": 0.4432,
      "mean_token_accuracy": 0.8630438409745693,
      "num_tokens": 476505986.0,
      "step": 14930
    },
    {
      "entropy": 0.5906660961918533,
      "epoch": 1.373729880794245,
      "grad_norm": 0.16181032359600067,
      "learning_rate": 5.4212285705523356e-05,
      "loss": 0.5751,
      "mean_token_accuracy": 0.8207370638847351,
      "num_tokens": 476537211.0,
      "step": 14931
    },
    {
      "entropy": 0.6621408006176353,
      "epoch": 1.3738218871656862,
      "grad_norm": 0.16396458446979523,
      "learning_rate": 5.420921887938173e-05,
      "loss": 0.6529,
      "mean_token_accuracy": 0.8028239011764526,
      "num_tokens": 476568686.0,
      "step": 14932
    },
    {
      "entropy": 0.6233959570527077,
      "epoch": 1.3739138935371273,
      "grad_norm": 0.16840438544750214,
      "learning_rate": 5.42061520532401e-05,
      "loss": 0.6046,
      "mean_token_accuracy": 0.8161724507808685,
      "num_tokens": 476600851.0,
      "step": 14933
    },
    {
      "entropy": 0.6087331082671881,
      "epoch": 1.3740058999085687,
      "grad_norm": 0.15533605217933655,
      "learning_rate": 5.420308522709848e-05,
      "loss": 0.5965,
      "mean_token_accuracy": 0.8162521980702877,
      "num_tokens": 476631797.0,
      "step": 14934
    },
    {
      "entropy": 0.6721489056944847,
      "epoch": 1.3740979062800098,
      "grad_norm": 0.16184942424297333,
      "learning_rate": 5.4200018400956854e-05,
      "loss": 0.6687,
      "mean_token_accuracy": 0.7977869249880314,
      "num_tokens": 476663533.0,
      "step": 14935
    },
    {
      "entropy": 0.7073728051036596,
      "epoch": 1.3741899126514512,
      "grad_norm": 0.16507865488529205,
      "learning_rate": 5.419695157481522e-05,
      "loss": 0.7066,
      "mean_token_accuracy": 0.7857302390038967,
      "num_tokens": 476695467.0,
      "step": 14936
    },
    {
      "entropy": 0.7006683964282274,
      "epoch": 1.3742819190228923,
      "grad_norm": 0.16022302210330963,
      "learning_rate": 5.41938847486736e-05,
      "loss": 0.6934,
      "mean_token_accuracy": 0.7897386327385902,
      "num_tokens": 476727863.0,
      "step": 14937
    },
    {
      "entropy": 0.6242291200906038,
      "epoch": 1.3743739253943335,
      "grad_norm": 0.15441511571407318,
      "learning_rate": 5.419081792253198e-05,
      "loss": 0.6284,
      "mean_token_accuracy": 0.8076505176723003,
      "num_tokens": 476760260.0,
      "step": 14938
    },
    {
      "entropy": 0.5531758470460773,
      "epoch": 1.3744659317657748,
      "grad_norm": 0.14987966418266296,
      "learning_rate": 5.4187751096390346e-05,
      "loss": 0.5579,
      "mean_token_accuracy": 0.8284235149621964,
      "num_tokens": 476792731.0,
      "step": 14939
    },
    {
      "entropy": 0.6076144333928823,
      "epoch": 1.374557938137216,
      "grad_norm": 0.15852390229701996,
      "learning_rate": 5.418468427024872e-05,
      "loss": 0.5861,
      "mean_token_accuracy": 0.8208057768642902,
      "num_tokens": 476824781.0,
      "step": 14940
    },
    {
      "entropy": 0.6609813831746578,
      "epoch": 1.3746499445086573,
      "grad_norm": 0.15536239743232727,
      "learning_rate": 5.418161744410709e-05,
      "loss": 0.6465,
      "mean_token_accuracy": 0.8005546294152737,
      "num_tokens": 476857184.0,
      "step": 14941
    },
    {
      "entropy": 0.646066527813673,
      "epoch": 1.3747419508800984,
      "grad_norm": 0.15560486912727356,
      "learning_rate": 5.417855061796547e-05,
      "loss": 0.6268,
      "mean_token_accuracy": 0.8072894252836704,
      "num_tokens": 476889563.0,
      "step": 14942
    },
    {
      "entropy": 0.6807376118376851,
      "epoch": 1.3748339572515396,
      "grad_norm": 0.1597168743610382,
      "learning_rate": 5.4175483791823845e-05,
      "loss": 0.6578,
      "mean_token_accuracy": 0.8030672967433929,
      "num_tokens": 476921747.0,
      "step": 14943
    },
    {
      "entropy": 0.6894102860242128,
      "epoch": 1.374925963622981,
      "grad_norm": 0.16127389669418335,
      "learning_rate": 5.417241696568221e-05,
      "loss": 0.6772,
      "mean_token_accuracy": 0.7946926839649677,
      "num_tokens": 476954491.0,
      "step": 14944
    },
    {
      "entropy": 0.6464547924697399,
      "epoch": 1.375017969994422,
      "grad_norm": 0.1649363487958908,
      "learning_rate": 5.416935013954059e-05,
      "loss": 0.6416,
      "mean_token_accuracy": 0.8082542568445206,
      "num_tokens": 476985390.0,
      "step": 14945
    },
    {
      "entropy": 0.5954149379394948,
      "epoch": 1.3751099763658634,
      "grad_norm": 0.1545429676771164,
      "learning_rate": 5.416628331339897e-05,
      "loss": 0.5663,
      "mean_token_accuracy": 0.8221376836299896,
      "num_tokens": 477016846.0,
      "step": 14946
    },
    {
      "entropy": 0.5842277267947793,
      "epoch": 1.3752019827373045,
      "grad_norm": 0.15563949942588806,
      "learning_rate": 5.416321648725734e-05,
      "loss": 0.5596,
      "mean_token_accuracy": 0.8278061673045158,
      "num_tokens": 477048530.0,
      "step": 14947
    },
    {
      "entropy": 0.5850693434476852,
      "epoch": 1.3752939891087457,
      "grad_norm": 0.1570942997932434,
      "learning_rate": 5.416014966111571e-05,
      "loss": 0.5804,
      "mean_token_accuracy": 0.821444496512413,
      "num_tokens": 477080025.0,
      "step": 14948
    },
    {
      "entropy": 0.6061463970690966,
      "epoch": 1.375385995480187,
      "grad_norm": 0.1638065129518509,
      "learning_rate": 5.415708283497409e-05,
      "loss": 0.5979,
      "mean_token_accuracy": 0.81351513043046,
      "num_tokens": 477111858.0,
      "step": 14949
    },
    {
      "entropy": 0.6208301279693842,
      "epoch": 1.3754780018516282,
      "grad_norm": 0.16443195939064026,
      "learning_rate": 5.415401600883247e-05,
      "loss": 0.6033,
      "mean_token_accuracy": 0.8120645023882389,
      "num_tokens": 477143593.0,
      "step": 14950
    },
    {
      "entropy": 0.6375775896012783,
      "epoch": 1.3755700082230695,
      "grad_norm": 0.1540077030658722,
      "learning_rate": 5.4150949182690835e-05,
      "loss": 0.6334,
      "mean_token_accuracy": 0.8031171336770058,
      "num_tokens": 477175832.0,
      "step": 14951
    },
    {
      "entropy": 0.6014215890318155,
      "epoch": 1.3756620145945107,
      "grad_norm": 0.15303002297878265,
      "learning_rate": 5.414788235654921e-05,
      "loss": 0.5811,
      "mean_token_accuracy": 0.8187747597694397,
      "num_tokens": 477207241.0,
      "step": 14952
    },
    {
      "entropy": 0.7780362404882908,
      "epoch": 1.3757540209659518,
      "grad_norm": 0.1707969307899475,
      "learning_rate": 5.414481553040759e-05,
      "loss": 0.7773,
      "mean_token_accuracy": 0.7663059011101723,
      "num_tokens": 477238937.0,
      "step": 14953
    },
    {
      "entropy": 0.5915431915782392,
      "epoch": 1.3758460273373931,
      "grad_norm": 0.15392686426639557,
      "learning_rate": 5.414174870426596e-05,
      "loss": 0.5902,
      "mean_token_accuracy": 0.8207919783890247,
      "num_tokens": 477271088.0,
      "step": 14954
    },
    {
      "entropy": 0.7848700080066919,
      "epoch": 1.3759380337088343,
      "grad_norm": 0.17320013046264648,
      "learning_rate": 5.4138681878124334e-05,
      "loss": 0.7844,
      "mean_token_accuracy": 0.766521118581295,
      "num_tokens": 477302787.0,
      "step": 14955
    },
    {
      "entropy": 0.7249782485887408,
      "epoch": 1.3760300400802756,
      "grad_norm": 0.16258549690246582,
      "learning_rate": 5.41356150519827e-05,
      "loss": 0.7318,
      "mean_token_accuracy": 0.7794199697673321,
      "num_tokens": 477335183.0,
      "step": 14956
    },
    {
      "entropy": 0.6321470141410828,
      "epoch": 1.3761220464517168,
      "grad_norm": 0.15147671103477478,
      "learning_rate": 5.413254822584108e-05,
      "loss": 0.6047,
      "mean_token_accuracy": 0.8132954724133015,
      "num_tokens": 477366516.0,
      "step": 14957
    },
    {
      "entropy": 0.6025239657610655,
      "epoch": 1.376214052823158,
      "grad_norm": 0.15633292496204376,
      "learning_rate": 5.412948139969946e-05,
      "loss": 0.5966,
      "mean_token_accuracy": 0.8181670047342777,
      "num_tokens": 477397979.0,
      "step": 14958
    },
    {
      "entropy": 0.6001942194998264,
      "epoch": 1.3763060591945993,
      "grad_norm": 0.1659117341041565,
      "learning_rate": 5.4126414573557826e-05,
      "loss": 0.5946,
      "mean_token_accuracy": 0.8199860192835331,
      "num_tokens": 477430013.0,
      "step": 14959
    },
    {
      "entropy": 0.5718221617862582,
      "epoch": 1.3763980655660404,
      "grad_norm": 0.16179119050502777,
      "learning_rate": 5.41233477474162e-05,
      "loss": 0.5641,
      "mean_token_accuracy": 0.8249674774706364,
      "num_tokens": 477462311.0,
      "step": 14960
    },
    {
      "entropy": 0.6472466103732586,
      "epoch": 1.3764900719374817,
      "grad_norm": 0.16060514748096466,
      "learning_rate": 5.412028092127458e-05,
      "loss": 0.6372,
      "mean_token_accuracy": 0.8036318495869637,
      "num_tokens": 477493871.0,
      "step": 14961
    },
    {
      "entropy": 0.6296706907451153,
      "epoch": 1.3765820783089229,
      "grad_norm": 0.1598348617553711,
      "learning_rate": 5.411721409513295e-05,
      "loss": 0.6248,
      "mean_token_accuracy": 0.81458830088377,
      "num_tokens": 477526288.0,
      "step": 14962
    },
    {
      "entropy": 0.678931251168251,
      "epoch": 1.376674084680364,
      "grad_norm": 0.16157111525535583,
      "learning_rate": 5.4114147268991324e-05,
      "loss": 0.6474,
      "mean_token_accuracy": 0.8023487366735935,
      "num_tokens": 477557627.0,
      "step": 14963
    },
    {
      "entropy": 0.621091527864337,
      "epoch": 1.3767660910518054,
      "grad_norm": 0.1541248857975006,
      "learning_rate": 5.411108044284969e-05,
      "loss": 0.5998,
      "mean_token_accuracy": 0.8174206390976906,
      "num_tokens": 477589733.0,
      "step": 14964
    },
    {
      "entropy": 0.6390293366275728,
      "epoch": 1.3768580974232465,
      "grad_norm": 0.1569078117609024,
      "learning_rate": 5.410801361670807e-05,
      "loss": 0.6295,
      "mean_token_accuracy": 0.8076532632112503,
      "num_tokens": 477622497.0,
      "step": 14965
    },
    {
      "entropy": 0.6113298204727471,
      "epoch": 1.3769501037946879,
      "grad_norm": 0.1548633873462677,
      "learning_rate": 5.410494679056645e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8199019357562065,
      "num_tokens": 477654995.0,
      "step": 14966
    },
    {
      "entropy": 0.6093850806355476,
      "epoch": 1.377042110166129,
      "grad_norm": 0.15288002789020538,
      "learning_rate": 5.4101879964424816e-05,
      "loss": 0.5954,
      "mean_token_accuracy": 0.8191415220499039,
      "num_tokens": 477687462.0,
      "step": 14967
    },
    {
      "entropy": 0.5133176017552614,
      "epoch": 1.3771341165375701,
      "grad_norm": 0.14790447056293488,
      "learning_rate": 5.409881313828319e-05,
      "loss": 0.481,
      "mean_token_accuracy": 0.8482778705656528,
      "num_tokens": 477719266.0,
      "step": 14968
    },
    {
      "entropy": 0.48965707095339894,
      "epoch": 1.3772261229090115,
      "grad_norm": 0.1466759294271469,
      "learning_rate": 5.409574631214157e-05,
      "loss": 0.4649,
      "mean_token_accuracy": 0.8522424399852753,
      "num_tokens": 477750722.0,
      "step": 14969
    },
    {
      "entropy": 0.5929316906258464,
      "epoch": 1.3773181292804526,
      "grad_norm": 0.15356934070587158,
      "learning_rate": 5.409267948599994e-05,
      "loss": 0.589,
      "mean_token_accuracy": 0.8218838274478912,
      "num_tokens": 477782727.0,
      "step": 14970
    },
    {
      "entropy": 0.5568674430251122,
      "epoch": 1.377410135651894,
      "grad_norm": 0.16415168344974518,
      "learning_rate": 5.4089612659858314e-05,
      "loss": 0.5449,
      "mean_token_accuracy": 0.8335238248109818,
      "num_tokens": 477815187.0,
      "step": 14971
    },
    {
      "entropy": 0.5396550511941314,
      "epoch": 1.377502142023335,
      "grad_norm": 0.1541062742471695,
      "learning_rate": 5.408654583371668e-05,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8348345831036568,
      "num_tokens": 477847273.0,
      "step": 14972
    },
    {
      "entropy": 0.6643192125484347,
      "epoch": 1.3775941483947762,
      "grad_norm": 0.1624174416065216,
      "learning_rate": 5.4083479007575064e-05,
      "loss": 0.6538,
      "mean_token_accuracy": 0.7994563095271587,
      "num_tokens": 477878810.0,
      "step": 14973
    },
    {
      "entropy": 0.6851068306714296,
      "epoch": 1.3776861547662176,
      "grad_norm": 0.1626092493534088,
      "learning_rate": 5.408041218143344e-05,
      "loss": 0.6669,
      "mean_token_accuracy": 0.7998993545770645,
      "num_tokens": 477909328.0,
      "step": 14974
    },
    {
      "entropy": 0.730296466499567,
      "epoch": 1.3777781611376587,
      "grad_norm": 0.16266939043998718,
      "learning_rate": 5.4077345355291806e-05,
      "loss": 0.7234,
      "mean_token_accuracy": 0.7787055633962154,
      "num_tokens": 477941034.0,
      "step": 14975
    },
    {
      "entropy": 0.6126102562993765,
      "epoch": 1.3778701675091,
      "grad_norm": 0.1580500304698944,
      "learning_rate": 5.407427852915018e-05,
      "loss": 0.6014,
      "mean_token_accuracy": 0.8159619830548763,
      "num_tokens": 477972603.0,
      "step": 14976
    },
    {
      "entropy": 0.5875917375087738,
      "epoch": 1.3779621738805412,
      "grad_norm": 0.15936924517154694,
      "learning_rate": 5.407121170300856e-05,
      "loss": 0.5762,
      "mean_token_accuracy": 0.8238633461296558,
      "num_tokens": 478004380.0,
      "step": 14977
    },
    {
      "entropy": 0.569952230900526,
      "epoch": 1.3780541802519823,
      "grad_norm": 0.15177969634532928,
      "learning_rate": 5.406814487686693e-05,
      "loss": 0.5514,
      "mean_token_accuracy": 0.8311657048761845,
      "num_tokens": 478035854.0,
      "step": 14978
    },
    {
      "entropy": 0.6196613516658545,
      "epoch": 1.3781461866234237,
      "grad_norm": 0.1579984575510025,
      "learning_rate": 5.4065078050725305e-05,
      "loss": 0.6029,
      "mean_token_accuracy": 0.8151780180633068,
      "num_tokens": 478067723.0,
      "step": 14979
    },
    {
      "entropy": 0.5257614590227604,
      "epoch": 1.3782381929948648,
      "grad_norm": 0.15479078888893127,
      "learning_rate": 5.4062011224583686e-05,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8354016318917274,
      "num_tokens": 478099179.0,
      "step": 14980
    },
    {
      "entropy": 0.6489925738424063,
      "epoch": 1.3783301993663062,
      "grad_norm": 0.1672946661710739,
      "learning_rate": 5.4058944398442054e-05,
      "loss": 0.6399,
      "mean_token_accuracy": 0.8049986511468887,
      "num_tokens": 478129807.0,
      "step": 14981
    },
    {
      "entropy": 0.6549622789025307,
      "epoch": 1.3784222057377473,
      "grad_norm": 0.16954101622104645,
      "learning_rate": 5.405587757230043e-05,
      "loss": 0.6495,
      "mean_token_accuracy": 0.8000534102320671,
      "num_tokens": 478159966.0,
      "step": 14982
    },
    {
      "entropy": 0.6780518721789122,
      "epoch": 1.3785142121091885,
      "grad_norm": 0.1772109866142273,
      "learning_rate": 5.40528107461588e-05,
      "loss": 0.6808,
      "mean_token_accuracy": 0.7937671355903149,
      "num_tokens": 478191744.0,
      "step": 14983
    },
    {
      "entropy": 0.5963972462341189,
      "epoch": 1.3786062184806298,
      "grad_norm": 0.15773724019527435,
      "learning_rate": 5.4049743920017185e-05,
      "loss": 0.5852,
      "mean_token_accuracy": 0.8246431611478329,
      "num_tokens": 478224273.0,
      "step": 14984
    },
    {
      "entropy": 0.6693075336515903,
      "epoch": 1.378698224852071,
      "grad_norm": 0.1630958467721939,
      "learning_rate": 5.404667709387555e-05,
      "loss": 0.6549,
      "mean_token_accuracy": 0.7969756051898003,
      "num_tokens": 478255443.0,
      "step": 14985
    },
    {
      "entropy": 0.5404051523655653,
      "epoch": 1.3787902312235123,
      "grad_norm": 0.15340377390384674,
      "learning_rate": 5.404361026773392e-05,
      "loss": 0.5341,
      "mean_token_accuracy": 0.8335658945143223,
      "num_tokens": 478287686.0,
      "step": 14986
    },
    {
      "entropy": 0.7285548830404878,
      "epoch": 1.3788822375949534,
      "grad_norm": 0.1629677712917328,
      "learning_rate": 5.4040543441592295e-05,
      "loss": 0.7289,
      "mean_token_accuracy": 0.780209805816412,
      "num_tokens": 478320171.0,
      "step": 14987
    },
    {
      "entropy": 0.6975354366004467,
      "epoch": 1.3789742439663946,
      "grad_norm": 0.1629534810781479,
      "learning_rate": 5.403747661545068e-05,
      "loss": 0.6999,
      "mean_token_accuracy": 0.7840665243566036,
      "num_tokens": 478351492.0,
      "step": 14988
    },
    {
      "entropy": 0.6440494647249579,
      "epoch": 1.379066250337836,
      "grad_norm": 0.16921094059944153,
      "learning_rate": 5.403440978930905e-05,
      "loss": 0.6198,
      "mean_token_accuracy": 0.8098917044699192,
      "num_tokens": 478382558.0,
      "step": 14989
    },
    {
      "entropy": 0.7207122538238764,
      "epoch": 1.379158256709277,
      "grad_norm": 0.16431032121181488,
      "learning_rate": 5.403134296316742e-05,
      "loss": 0.703,
      "mean_token_accuracy": 0.7862970903515816,
      "num_tokens": 478414446.0,
      "step": 14990
    },
    {
      "entropy": 0.6440649125725031,
      "epoch": 1.3792502630807184,
      "grad_norm": 0.1578667163848877,
      "learning_rate": 5.4028276137025794e-05,
      "loss": 0.6279,
      "mean_token_accuracy": 0.8037366457283497,
      "num_tokens": 478445936.0,
      "step": 14991
    },
    {
      "entropy": 0.6503578275442123,
      "epoch": 1.3793422694521595,
      "grad_norm": 0.15502312779426575,
      "learning_rate": 5.4025209310884175e-05,
      "loss": 0.6363,
      "mean_token_accuracy": 0.8062879741191864,
      "num_tokens": 478478317.0,
      "step": 14992
    },
    {
      "entropy": 0.5858661392703652,
      "epoch": 1.3794342758236007,
      "grad_norm": 0.15446995198726654,
      "learning_rate": 5.402214248474254e-05,
      "loss": 0.573,
      "mean_token_accuracy": 0.8200505711138248,
      "num_tokens": 478510020.0,
      "step": 14993
    },
    {
      "entropy": 0.6279553426429629,
      "epoch": 1.379526282195042,
      "grad_norm": 0.15133118629455566,
      "learning_rate": 5.401907565860092e-05,
      "loss": 0.6134,
      "mean_token_accuracy": 0.8086814992129803,
      "num_tokens": 478542107.0,
      "step": 14994
    },
    {
      "entropy": 0.5817398340441287,
      "epoch": 1.3796182885664832,
      "grad_norm": 0.16062569618225098,
      "learning_rate": 5.4016008832459286e-05,
      "loss": 0.5722,
      "mean_token_accuracy": 0.8220590353012085,
      "num_tokens": 478573941.0,
      "step": 14995
    },
    {
      "entropy": 0.7425936358049512,
      "epoch": 1.3797102949379245,
      "grad_norm": 0.1605396866798401,
      "learning_rate": 5.401294200631767e-05,
      "loss": 0.7294,
      "mean_token_accuracy": 0.7794704213738441,
      "num_tokens": 478605992.0,
      "step": 14996
    },
    {
      "entropy": 0.6473066676408052,
      "epoch": 1.3798023013093657,
      "grad_norm": 0.14965051412582397,
      "learning_rate": 5.400987518017604e-05,
      "loss": 0.6247,
      "mean_token_accuracy": 0.8093862757086754,
      "num_tokens": 478638148.0,
      "step": 14997
    },
    {
      "entropy": 0.6247008079662919,
      "epoch": 1.3798943076808068,
      "grad_norm": 0.17262914776802063,
      "learning_rate": 5.400680835403441e-05,
      "loss": 0.6213,
      "mean_token_accuracy": 0.8108941577374935,
      "num_tokens": 478670014.0,
      "step": 14998
    },
    {
      "entropy": 0.5164309190586209,
      "epoch": 1.3799863140522481,
      "grad_norm": 0.14479126036167145,
      "learning_rate": 5.4003741527892784e-05,
      "loss": 0.5137,
      "mean_token_accuracy": 0.8392765298485756,
      "num_tokens": 478702655.0,
      "step": 14999
    },
    {
      "entropy": 0.5359436827711761,
      "epoch": 1.3800783204236893,
      "grad_norm": 0.1496240496635437,
      "learning_rate": 5.4000674701751166e-05,
      "loss": 0.514,
      "mean_token_accuracy": 0.8401033207774162,
      "num_tokens": 478734605.0,
      "step": 15000
    },
    {
      "entropy": 0.6185065116733313,
      "epoch": 1.3801703267951306,
      "grad_norm": 0.16110318899154663,
      "learning_rate": 5.3997607875609534e-05,
      "loss": 0.5838,
      "mean_token_accuracy": 0.8215113878250122,
      "num_tokens": 478766338.0,
      "step": 15001
    },
    {
      "entropy": 0.6143472399562597,
      "epoch": 1.3802623331665718,
      "grad_norm": 0.15979760885238647,
      "learning_rate": 5.399454104946791e-05,
      "loss": 0.6078,
      "mean_token_accuracy": 0.8155271969735622,
      "num_tokens": 478797792.0,
      "step": 15002
    },
    {
      "entropy": 0.5335310762748122,
      "epoch": 1.380354339538013,
      "grad_norm": 0.14508089423179626,
      "learning_rate": 5.3991474223326276e-05,
      "loss": 0.5099,
      "mean_token_accuracy": 0.8414219506084919,
      "num_tokens": 478830153.0,
      "step": 15003
    },
    {
      "entropy": 0.5688100233674049,
      "epoch": 1.3804463459094543,
      "grad_norm": 0.15641526877880096,
      "learning_rate": 5.398840739718466e-05,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8309963792562485,
      "num_tokens": 478862045.0,
      "step": 15004
    },
    {
      "entropy": 0.6460552830249071,
      "epoch": 1.3805383522808954,
      "grad_norm": 0.16073668003082275,
      "learning_rate": 5.398534057104303e-05,
      "loss": 0.6506,
      "mean_token_accuracy": 0.804061770439148,
      "num_tokens": 478893494.0,
      "step": 15005
    },
    {
      "entropy": 0.6307973035145551,
      "epoch": 1.3806303586523367,
      "grad_norm": 0.15409615635871887,
      "learning_rate": 5.39822737449014e-05,
      "loss": 0.631,
      "mean_token_accuracy": 0.8054283149540424,
      "num_tokens": 478925470.0,
      "step": 15006
    },
    {
      "entropy": 0.5753796044737101,
      "epoch": 1.3807223650237779,
      "grad_norm": 0.16327784955501556,
      "learning_rate": 5.3979206918759775e-05,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8262509219348431,
      "num_tokens": 478957524.0,
      "step": 15007
    },
    {
      "entropy": 0.7297482471913099,
      "epoch": 1.380814371395219,
      "grad_norm": 0.1712968945503235,
      "learning_rate": 5.3976140092618156e-05,
      "loss": 0.7142,
      "mean_token_accuracy": 0.7827882654964924,
      "num_tokens": 478989801.0,
      "step": 15008
    },
    {
      "entropy": 0.7003817902877927,
      "epoch": 1.3809063777666604,
      "grad_norm": 0.16429539024829865,
      "learning_rate": 5.3973073266476524e-05,
      "loss": 0.6852,
      "mean_token_accuracy": 0.7929309085011482,
      "num_tokens": 479021656.0,
      "step": 15009
    },
    {
      "entropy": 0.5802193805575371,
      "epoch": 1.3809983841381015,
      "grad_norm": 0.1579047292470932,
      "learning_rate": 5.39700064403349e-05,
      "loss": 0.5737,
      "mean_token_accuracy": 0.8273662999272346,
      "num_tokens": 479053781.0,
      "step": 15010
    },
    {
      "entropy": 0.5672704000025988,
      "epoch": 1.3810903905095429,
      "grad_norm": 0.1511450856924057,
      "learning_rate": 5.396693961419328e-05,
      "loss": 0.545,
      "mean_token_accuracy": 0.8296195417642593,
      "num_tokens": 479085306.0,
      "step": 15011
    },
    {
      "entropy": 0.554361030459404,
      "epoch": 1.381182396880984,
      "grad_norm": 0.150045707821846,
      "learning_rate": 5.396387278805165e-05,
      "loss": 0.5506,
      "mean_token_accuracy": 0.8310933820903301,
      "num_tokens": 479117206.0,
      "step": 15012
    },
    {
      "entropy": 0.6267484975978732,
      "epoch": 1.3812744032524251,
      "grad_norm": 0.1571190506219864,
      "learning_rate": 5.396080596191002e-05,
      "loss": 0.6154,
      "mean_token_accuracy": 0.8129322901368141,
      "num_tokens": 479149645.0,
      "step": 15013
    },
    {
      "entropy": 0.5969376154243946,
      "epoch": 1.3813664096238665,
      "grad_norm": 0.1543402075767517,
      "learning_rate": 5.395773913576839e-05,
      "loss": 0.6014,
      "mean_token_accuracy": 0.8141588792204857,
      "num_tokens": 479181272.0,
      "step": 15014
    },
    {
      "entropy": 0.609663711860776,
      "epoch": 1.3814584159953076,
      "grad_norm": 0.15388818085193634,
      "learning_rate": 5.395467230962677e-05,
      "loss": 0.6014,
      "mean_token_accuracy": 0.8170346617698669,
      "num_tokens": 479213072.0,
      "step": 15015
    },
    {
      "entropy": 0.6136439442634583,
      "epoch": 1.381550422366749,
      "grad_norm": 0.1560717225074768,
      "learning_rate": 5.3951605483485147e-05,
      "loss": 0.5929,
      "mean_token_accuracy": 0.8178392499685287,
      "num_tokens": 479244466.0,
      "step": 15016
    },
    {
      "entropy": 0.49896139558404684,
      "epoch": 1.38164242873819,
      "grad_norm": 0.1522819697856903,
      "learning_rate": 5.3948538657343514e-05,
      "loss": 0.4974,
      "mean_token_accuracy": 0.8466627970337868,
      "num_tokens": 479276776.0,
      "step": 15017
    },
    {
      "entropy": 0.597970062866807,
      "epoch": 1.3817344351096312,
      "grad_norm": 0.15863490104675293,
      "learning_rate": 5.394547183120189e-05,
      "loss": 0.5849,
      "mean_token_accuracy": 0.820741705596447,
      "num_tokens": 479309155.0,
      "step": 15018
    },
    {
      "entropy": 0.6007932219654322,
      "epoch": 1.3818264414810726,
      "grad_norm": 0.16121463477611542,
      "learning_rate": 5.394240500506027e-05,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8201408982276917,
      "num_tokens": 479341191.0,
      "step": 15019
    },
    {
      "entropy": 0.6086448221467435,
      "epoch": 1.3819184478525137,
      "grad_norm": 0.15991990268230438,
      "learning_rate": 5.393933817891864e-05,
      "loss": 0.6013,
      "mean_token_accuracy": 0.8165146633982658,
      "num_tokens": 479373165.0,
      "step": 15020
    },
    {
      "entropy": 0.556672926992178,
      "epoch": 1.382010454223955,
      "grad_norm": 0.16236868500709534,
      "learning_rate": 5.393627135277701e-05,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8284193128347397,
      "num_tokens": 479404632.0,
      "step": 15021
    },
    {
      "entropy": 0.6818356104195118,
      "epoch": 1.3821024605953962,
      "grad_norm": 0.15493476390838623,
      "learning_rate": 5.393320452663538e-05,
      "loss": 0.6838,
      "mean_token_accuracy": 0.7934299074113369,
      "num_tokens": 479436682.0,
      "step": 15022
    },
    {
      "entropy": 0.7117904433980584,
      "epoch": 1.3821944669668373,
      "grad_norm": 0.16159141063690186,
      "learning_rate": 5.393013770049377e-05,
      "loss": 0.6898,
      "mean_token_accuracy": 0.7907328195869923,
      "num_tokens": 479468089.0,
      "step": 15023
    },
    {
      "entropy": 0.6025333912111819,
      "epoch": 1.3822864733382787,
      "grad_norm": 0.1475444883108139,
      "learning_rate": 5.392707087435214e-05,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8216991573572159,
      "num_tokens": 479500618.0,
      "step": 15024
    },
    {
      "entropy": 0.5389242614619434,
      "epoch": 1.3823784797097198,
      "grad_norm": 0.1526334285736084,
      "learning_rate": 5.3924004048210505e-05,
      "loss": 0.5241,
      "mean_token_accuracy": 0.8370568454265594,
      "num_tokens": 479532634.0,
      "step": 15025
    },
    {
      "entropy": 0.640519404783845,
      "epoch": 1.3824704860811612,
      "grad_norm": 0.15652033686637878,
      "learning_rate": 5.392093722206888e-05,
      "loss": 0.6244,
      "mean_token_accuracy": 0.8060200251638889,
      "num_tokens": 479564390.0,
      "step": 15026
    },
    {
      "entropy": 0.520042672753334,
      "epoch": 1.3825624924526023,
      "grad_norm": 0.1502406746149063,
      "learning_rate": 5.391787039592726e-05,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8470633067190647,
      "num_tokens": 479595738.0,
      "step": 15027
    },
    {
      "entropy": 0.5914176166988909,
      "epoch": 1.3826544988240435,
      "grad_norm": 0.15870077908039093,
      "learning_rate": 5.3914803569785636e-05,
      "loss": 0.579,
      "mean_token_accuracy": 0.8244524374604225,
      "num_tokens": 479627215.0,
      "step": 15028
    },
    {
      "entropy": 0.6019380462821573,
      "epoch": 1.3827465051954848,
      "grad_norm": 0.15423409640789032,
      "learning_rate": 5.3911736743644003e-05,
      "loss": 0.5848,
      "mean_token_accuracy": 0.820493221282959,
      "num_tokens": 479659341.0,
      "step": 15029
    },
    {
      "entropy": 0.5767673943191767,
      "epoch": 1.382838511566926,
      "grad_norm": 0.15452925860881805,
      "learning_rate": 5.390866991750237e-05,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8274812810122967,
      "num_tokens": 479690508.0,
      "step": 15030
    },
    {
      "entropy": 0.6906782947480679,
      "epoch": 1.3829305179383673,
      "grad_norm": 0.1595180332660675,
      "learning_rate": 5.390560309136076e-05,
      "loss": 0.6784,
      "mean_token_accuracy": 0.7927804253995419,
      "num_tokens": 479722443.0,
      "step": 15031
    },
    {
      "entropy": 0.5198345426470041,
      "epoch": 1.3830225243098084,
      "grad_norm": 0.14924463629722595,
      "learning_rate": 5.390253626521913e-05,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8407010957598686,
      "num_tokens": 479753839.0,
      "step": 15032
    },
    {
      "entropy": 0.5914210025221109,
      "epoch": 1.3831145306812496,
      "grad_norm": 0.1535811722278595,
      "learning_rate": 5.38994694390775e-05,
      "loss": 0.5769,
      "mean_token_accuracy": 0.8160724528133869,
      "num_tokens": 479785671.0,
      "step": 15033
    },
    {
      "entropy": 0.6792634837329388,
      "epoch": 1.383206537052691,
      "grad_norm": 0.15814119577407837,
      "learning_rate": 5.389640261293587e-05,
      "loss": 0.6668,
      "mean_token_accuracy": 0.7967772521078587,
      "num_tokens": 479817525.0,
      "step": 15034
    },
    {
      "entropy": 0.6095374701544642,
      "epoch": 1.383298543424132,
      "grad_norm": 0.16440792381763458,
      "learning_rate": 5.389333578679425e-05,
      "loss": 0.6119,
      "mean_token_accuracy": 0.8150526769459248,
      "num_tokens": 479849743.0,
      "step": 15035
    },
    {
      "entropy": 0.6934779398143291,
      "epoch": 1.3833905497955734,
      "grad_norm": 0.16450487077236176,
      "learning_rate": 5.3890268960652626e-05,
      "loss": 0.6857,
      "mean_token_accuracy": 0.7909518927335739,
      "num_tokens": 479881226.0,
      "step": 15036
    },
    {
      "entropy": 0.6048901826143265,
      "epoch": 1.3834825561670145,
      "grad_norm": 0.1595529317855835,
      "learning_rate": 5.3887202134510994e-05,
      "loss": 0.5975,
      "mean_token_accuracy": 0.8175789378583431,
      "num_tokens": 479913144.0,
      "step": 15037
    },
    {
      "entropy": 0.6734392959624529,
      "epoch": 1.3835745625384557,
      "grad_norm": 0.16423486173152924,
      "learning_rate": 5.388413530836937e-05,
      "loss": 0.6638,
      "mean_token_accuracy": 0.7979573383927345,
      "num_tokens": 479943898.0,
      "step": 15038
    },
    {
      "entropy": 0.6966054439544678,
      "epoch": 1.383666568909897,
      "grad_norm": 0.17275530099868774,
      "learning_rate": 5.388106848222775e-05,
      "loss": 0.6874,
      "mean_token_accuracy": 0.7860759384930134,
      "num_tokens": 479975412.0,
      "step": 15039
    },
    {
      "entropy": 0.5510879866778851,
      "epoch": 1.3837585752813382,
      "grad_norm": 0.15449151396751404,
      "learning_rate": 5.387800165608612e-05,
      "loss": 0.545,
      "mean_token_accuracy": 0.8321738876402378,
      "num_tokens": 480007317.0,
      "step": 15040
    },
    {
      "entropy": 0.6894300431013107,
      "epoch": 1.3838505816527795,
      "grad_norm": 0.16714809834957123,
      "learning_rate": 5.387493482994449e-05,
      "loss": 0.6936,
      "mean_token_accuracy": 0.7934844568371773,
      "num_tokens": 480039444.0,
      "step": 15041
    },
    {
      "entropy": 0.6194839328527451,
      "epoch": 1.3839425880242207,
      "grad_norm": 0.15879708528518677,
      "learning_rate": 5.3871868003802874e-05,
      "loss": 0.5973,
      "mean_token_accuracy": 0.8161039762198925,
      "num_tokens": 480070342.0,
      "step": 15042
    },
    {
      "entropy": 0.5356742721050978,
      "epoch": 1.3840345943956618,
      "grad_norm": 0.15329709649085999,
      "learning_rate": 5.386880117766124e-05,
      "loss": 0.5181,
      "mean_token_accuracy": 0.8429401405155659,
      "num_tokens": 480102378.0,
      "step": 15043
    },
    {
      "entropy": 0.5813276525586843,
      "epoch": 1.3841266007671031,
      "grad_norm": 0.15270957350730896,
      "learning_rate": 5.3865734351519616e-05,
      "loss": 0.5594,
      "mean_token_accuracy": 0.8226130679249763,
      "num_tokens": 480134436.0,
      "step": 15044
    },
    {
      "entropy": 0.7102247402071953,
      "epoch": 1.3842186071385443,
      "grad_norm": 0.1684463918209076,
      "learning_rate": 5.3862667525377984e-05,
      "loss": 0.7049,
      "mean_token_accuracy": 0.790333416312933,
      "num_tokens": 480167204.0,
      "step": 15045
    },
    {
      "entropy": 0.6989480312913656,
      "epoch": 1.3843106135099856,
      "grad_norm": 0.16092996299266815,
      "learning_rate": 5.3859600699236366e-05,
      "loss": 0.6788,
      "mean_token_accuracy": 0.7888235785067081,
      "num_tokens": 480199185.0,
      "step": 15046
    },
    {
      "entropy": 0.679448900744319,
      "epoch": 1.3844026198814268,
      "grad_norm": 0.16355526447296143,
      "learning_rate": 5.385653387309474e-05,
      "loss": 0.6681,
      "mean_token_accuracy": 0.7963674329221249,
      "num_tokens": 480230850.0,
      "step": 15047
    },
    {
      "entropy": 0.7407621871680021,
      "epoch": 1.384494626252868,
      "grad_norm": 0.16060024499893188,
      "learning_rate": 5.385346704695311e-05,
      "loss": 0.743,
      "mean_token_accuracy": 0.7748262397944927,
      "num_tokens": 480263483.0,
      "step": 15048
    },
    {
      "entropy": 0.6045486368238926,
      "epoch": 1.3845866326243093,
      "grad_norm": 0.15624995529651642,
      "learning_rate": 5.385040022081148e-05,
      "loss": 0.5919,
      "mean_token_accuracy": 0.8159331530332565,
      "num_tokens": 480294544.0,
      "step": 15049
    },
    {
      "entropy": 0.591907620895654,
      "epoch": 1.3846786389957504,
      "grad_norm": 0.15669696033000946,
      "learning_rate": 5.3847333394669864e-05,
      "loss": 0.5803,
      "mean_token_accuracy": 0.823590062558651,
      "num_tokens": 480325973.0,
      "step": 15050
    },
    {
      "entropy": 0.6198105597868562,
      "epoch": 1.3847706453671917,
      "grad_norm": 0.15466852486133575,
      "learning_rate": 5.384426656852823e-05,
      "loss": 0.6045,
      "mean_token_accuracy": 0.8107556663453579,
      "num_tokens": 480358629.0,
      "step": 15051
    },
    {
      "entropy": 0.7567916214466095,
      "epoch": 1.3848626517386329,
      "grad_norm": 0.17016009986400604,
      "learning_rate": 5.384119974238661e-05,
      "loss": 0.7565,
      "mean_token_accuracy": 0.7744597867131233,
      "num_tokens": 480389520.0,
      "step": 15052
    },
    {
      "entropy": 0.6704023256897926,
      "epoch": 1.384954658110074,
      "grad_norm": 0.16340090334415436,
      "learning_rate": 5.3838132916244975e-05,
      "loss": 0.6727,
      "mean_token_accuracy": 0.7950354181230068,
      "num_tokens": 480421623.0,
      "step": 15053
    },
    {
      "entropy": 0.5520228799432516,
      "epoch": 1.3850466644815154,
      "grad_norm": 0.15273256599903107,
      "learning_rate": 5.3835066090103356e-05,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8345955349504948,
      "num_tokens": 480453356.0,
      "step": 15054
    },
    {
      "entropy": 0.5539502897299826,
      "epoch": 1.3851386708529565,
      "grad_norm": 0.1525501161813736,
      "learning_rate": 5.383199926396173e-05,
      "loss": 0.5521,
      "mean_token_accuracy": 0.8282056488096714,
      "num_tokens": 480485818.0,
      "step": 15055
    },
    {
      "entropy": 0.5362869054079056,
      "epoch": 1.3852306772243979,
      "grad_norm": 0.16154702007770538,
      "learning_rate": 5.38289324378201e-05,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8410099074244499,
      "num_tokens": 480517862.0,
      "step": 15056
    },
    {
      "entropy": 0.5650068856775761,
      "epoch": 1.385322683595839,
      "grad_norm": 0.14911340177059174,
      "learning_rate": 5.382586561167847e-05,
      "loss": 0.5519,
      "mean_token_accuracy": 0.832304697483778,
      "num_tokens": 480550041.0,
      "step": 15057
    },
    {
      "entropy": 0.5472307046875358,
      "epoch": 1.3854146899672801,
      "grad_norm": 0.1476265788078308,
      "learning_rate": 5.3822798785536855e-05,
      "loss": 0.537,
      "mean_token_accuracy": 0.8356661386787891,
      "num_tokens": 480582402.0,
      "step": 15058
    },
    {
      "entropy": 0.6952767297625542,
      "epoch": 1.3855066963387215,
      "grad_norm": 0.16822519898414612,
      "learning_rate": 5.381973195939522e-05,
      "loss": 0.6862,
      "mean_token_accuracy": 0.790110357105732,
      "num_tokens": 480613549.0,
      "step": 15059
    },
    {
      "entropy": 0.616052808240056,
      "epoch": 1.3855987027101626,
      "grad_norm": 0.1537257879972458,
      "learning_rate": 5.38166651332536e-05,
      "loss": 0.6017,
      "mean_token_accuracy": 0.8152834922075272,
      "num_tokens": 480645367.0,
      "step": 15060
    },
    {
      "entropy": 0.647583732381463,
      "epoch": 1.385690709081604,
      "grad_norm": 0.16250789165496826,
      "learning_rate": 5.3813598307111965e-05,
      "loss": 0.628,
      "mean_token_accuracy": 0.8050988726317883,
      "num_tokens": 480677435.0,
      "step": 15061
    },
    {
      "entropy": 0.5429914421401918,
      "epoch": 1.385782715453045,
      "grad_norm": 0.14924342930316925,
      "learning_rate": 5.3810531480970346e-05,
      "loss": 0.527,
      "mean_token_accuracy": 0.836400680243969,
      "num_tokens": 480709395.0,
      "step": 15062
    },
    {
      "entropy": 0.6466148225590587,
      "epoch": 1.3858747218244862,
      "grad_norm": 0.16293534636497498,
      "learning_rate": 5.380746465482872e-05,
      "loss": 0.6474,
      "mean_token_accuracy": 0.8038857132196426,
      "num_tokens": 480741268.0,
      "step": 15063
    },
    {
      "entropy": 0.6173511883243918,
      "epoch": 1.3859667281959276,
      "grad_norm": 0.15149274468421936,
      "learning_rate": 5.380439782868709e-05,
      "loss": 0.6018,
      "mean_token_accuracy": 0.8165186457335949,
      "num_tokens": 480772239.0,
      "step": 15064
    },
    {
      "entropy": 0.6305900644510984,
      "epoch": 1.3860587345673687,
      "grad_norm": 0.15732036530971527,
      "learning_rate": 5.3801331002545464e-05,
      "loss": 0.61,
      "mean_token_accuracy": 0.8102799691259861,
      "num_tokens": 480803882.0,
      "step": 15065
    },
    {
      "entropy": 0.7029399387538433,
      "epoch": 1.38615074093881,
      "grad_norm": 0.16512496769428253,
      "learning_rate": 5.3798264176403845e-05,
      "loss": 0.696,
      "mean_token_accuracy": 0.7905776761472225,
      "num_tokens": 480835248.0,
      "step": 15066
    },
    {
      "entropy": 0.6203714590519667,
      "epoch": 1.3862427473102512,
      "grad_norm": 0.16051119565963745,
      "learning_rate": 5.379519735026222e-05,
      "loss": 0.5978,
      "mean_token_accuracy": 0.8130321241915226,
      "num_tokens": 480866765.0,
      "step": 15067
    },
    {
      "entropy": 0.7789949327707291,
      "epoch": 1.3863347536816923,
      "grad_norm": 0.16798104345798492,
      "learning_rate": 5.379213052412059e-05,
      "loss": 0.7699,
      "mean_token_accuracy": 0.7636334635317326,
      "num_tokens": 480898396.0,
      "step": 15068
    },
    {
      "entropy": 0.6634422754868865,
      "epoch": 1.3864267600531337,
      "grad_norm": 0.16066837310791016,
      "learning_rate": 5.3789063697978955e-05,
      "loss": 0.6513,
      "mean_token_accuracy": 0.8016883358359337,
      "num_tokens": 480930310.0,
      "step": 15069
    },
    {
      "entropy": 0.6589894061908126,
      "epoch": 1.3865187664245748,
      "grad_norm": 0.15749482810497284,
      "learning_rate": 5.3785996871837344e-05,
      "loss": 0.6435,
      "mean_token_accuracy": 0.808730062097311,
      "num_tokens": 480962173.0,
      "step": 15070
    },
    {
      "entropy": 0.6499456409364939,
      "epoch": 1.3866107727960162,
      "grad_norm": 0.15956512093544006,
      "learning_rate": 5.378293004569571e-05,
      "loss": 0.6369,
      "mean_token_accuracy": 0.8040928468108177,
      "num_tokens": 480994125.0,
      "step": 15071
    },
    {
      "entropy": 0.6173209380358458,
      "epoch": 1.3867027791674573,
      "grad_norm": 0.157173290848732,
      "learning_rate": 5.3779863219554086e-05,
      "loss": 0.6101,
      "mean_token_accuracy": 0.8168232887983322,
      "num_tokens": 481026163.0,
      "step": 15072
    },
    {
      "entropy": 0.5974016515538096,
      "epoch": 1.3867947855388985,
      "grad_norm": 0.15391099452972412,
      "learning_rate": 5.377679639341247e-05,
      "loss": 0.5808,
      "mean_token_accuracy": 0.8230699002742767,
      "num_tokens": 481058525.0,
      "step": 15073
    },
    {
      "entropy": 0.5525264162570238,
      "epoch": 1.3868867919103398,
      "grad_norm": 0.15278635919094086,
      "learning_rate": 5.3773729567270835e-05,
      "loss": 0.5185,
      "mean_token_accuracy": 0.8416446931660175,
      "num_tokens": 481090849.0,
      "step": 15074
    },
    {
      "entropy": 0.47322733979672194,
      "epoch": 1.386978798281781,
      "grad_norm": 0.14905966818332672,
      "learning_rate": 5.377066274112921e-05,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8562816604971886,
      "num_tokens": 481122987.0,
      "step": 15075
    },
    {
      "entropy": 0.7051634751260281,
      "epoch": 1.3870708046532223,
      "grad_norm": 0.1608947217464447,
      "learning_rate": 5.376759591498758e-05,
      "loss": 0.7001,
      "mean_token_accuracy": 0.7886574752628803,
      "num_tokens": 481155062.0,
      "step": 15076
    },
    {
      "entropy": 0.4882275420241058,
      "epoch": 1.3871628110246634,
      "grad_norm": 0.15413789451122284,
      "learning_rate": 5.376452908884596e-05,
      "loss": 0.4692,
      "mean_token_accuracy": 0.8555795848369598,
      "num_tokens": 481186959.0,
      "step": 15077
    },
    {
      "entropy": 0.6672544647008181,
      "epoch": 1.3872548173961046,
      "grad_norm": 0.1790514588356018,
      "learning_rate": 5.3761462262704334e-05,
      "loss": 0.6686,
      "mean_token_accuracy": 0.7956382408738136,
      "num_tokens": 481219334.0,
      "step": 15078
    },
    {
      "entropy": 0.6058958973735571,
      "epoch": 1.387346823767546,
      "grad_norm": 0.16276580095291138,
      "learning_rate": 5.37583954365627e-05,
      "loss": 0.5945,
      "mean_token_accuracy": 0.817156221717596,
      "num_tokens": 481250840.0,
      "step": 15079
    },
    {
      "entropy": 0.5239001261070371,
      "epoch": 1.387438830138987,
      "grad_norm": 0.14520445466041565,
      "learning_rate": 5.3755328610421077e-05,
      "loss": 0.5197,
      "mean_token_accuracy": 0.8396640680730343,
      "num_tokens": 481282752.0,
      "step": 15080
    },
    {
      "entropy": 0.6593295382335782,
      "epoch": 1.3875308365104284,
      "grad_norm": 0.1590399295091629,
      "learning_rate": 5.375226178427946e-05,
      "loss": 0.6556,
      "mean_token_accuracy": 0.8016006387770176,
      "num_tokens": 481315089.0,
      "step": 15081
    },
    {
      "entropy": 0.721117977052927,
      "epoch": 1.3876228428818695,
      "grad_norm": 0.16020657122135162,
      "learning_rate": 5.3749194958137826e-05,
      "loss": 0.7109,
      "mean_token_accuracy": 0.7825213335454464,
      "num_tokens": 481346701.0,
      "step": 15082
    },
    {
      "entropy": 0.6150806080549955,
      "epoch": 1.3877148492533107,
      "grad_norm": 0.158853679895401,
      "learning_rate": 5.37461281319962e-05,
      "loss": 0.6043,
      "mean_token_accuracy": 0.8113737143576145,
      "num_tokens": 481379084.0,
      "step": 15083
    },
    {
      "entropy": 0.5508286841213703,
      "epoch": 1.387806855624752,
      "grad_norm": 0.1473166048526764,
      "learning_rate": 5.374306130585457e-05,
      "loss": 0.5388,
      "mean_token_accuracy": 0.8340629637241364,
      "num_tokens": 481411446.0,
      "step": 15084
    },
    {
      "entropy": 0.581901459954679,
      "epoch": 1.3878988619961932,
      "grad_norm": 0.1535482406616211,
      "learning_rate": 5.373999447971295e-05,
      "loss": 0.5765,
      "mean_token_accuracy": 0.8207255303859711,
      "num_tokens": 481443709.0,
      "step": 15085
    },
    {
      "entropy": 0.6669548861682415,
      "epoch": 1.3879908683676345,
      "grad_norm": 0.1615408957004547,
      "learning_rate": 5.3736927653571324e-05,
      "loss": 0.6673,
      "mean_token_accuracy": 0.7987936697900295,
      "num_tokens": 481475469.0,
      "step": 15086
    },
    {
      "entropy": 0.5774583630263805,
      "epoch": 1.3880828747390757,
      "grad_norm": 0.1608763039112091,
      "learning_rate": 5.373386082742969e-05,
      "loss": 0.5737,
      "mean_token_accuracy": 0.8248417787253857,
      "num_tokens": 481507401.0,
      "step": 15087
    },
    {
      "entropy": 0.5757575500756502,
      "epoch": 1.3881748811105168,
      "grad_norm": 0.14940407872200012,
      "learning_rate": 5.373079400128807e-05,
      "loss": 0.5561,
      "mean_token_accuracy": 0.826555110514164,
      "num_tokens": 481539647.0,
      "step": 15088
    },
    {
      "entropy": 0.6307274987921119,
      "epoch": 1.3882668874819581,
      "grad_norm": 0.15621069073677063,
      "learning_rate": 5.372772717514645e-05,
      "loss": 0.6121,
      "mean_token_accuracy": 0.8114646151661873,
      "num_tokens": 481571663.0,
      "step": 15089
    },
    {
      "entropy": 0.6546815950423479,
      "epoch": 1.3883588938533993,
      "grad_norm": 0.16003923118114471,
      "learning_rate": 5.3724660349004816e-05,
      "loss": 0.6532,
      "mean_token_accuracy": 0.8029684014618397,
      "num_tokens": 481603338.0,
      "step": 15090
    },
    {
      "entropy": 0.7155684120953083,
      "epoch": 1.3884509002248406,
      "grad_norm": 0.16542235016822815,
      "learning_rate": 5.372159352286319e-05,
      "loss": 0.7006,
      "mean_token_accuracy": 0.7872339375317097,
      "num_tokens": 481634537.0,
      "step": 15091
    },
    {
      "entropy": 0.5675797036383301,
      "epoch": 1.3885429065962818,
      "grad_norm": 0.15181207656860352,
      "learning_rate": 5.371852669672156e-05,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8301781266927719,
      "num_tokens": 481667134.0,
      "step": 15092
    },
    {
      "entropy": 0.5707734236493707,
      "epoch": 1.388634912967723,
      "grad_norm": 0.16145852208137512,
      "learning_rate": 5.371545987057994e-05,
      "loss": 0.5619,
      "mean_token_accuracy": 0.8274017348885536,
      "num_tokens": 481699558.0,
      "step": 15093
    },
    {
      "entropy": 0.5756315132603049,
      "epoch": 1.3887269193391643,
      "grad_norm": 0.1454664021730423,
      "learning_rate": 5.3712393044438315e-05,
      "loss": 0.5754,
      "mean_token_accuracy": 0.8243116699159145,
      "num_tokens": 481731766.0,
      "step": 15094
    },
    {
      "entropy": 0.676866352558136,
      "epoch": 1.3888189257106054,
      "grad_norm": 0.16624489426612854,
      "learning_rate": 5.370932621829668e-05,
      "loss": 0.6622,
      "mean_token_accuracy": 0.7963552065193653,
      "num_tokens": 481763579.0,
      "step": 15095
    },
    {
      "entropy": 0.6834894865751266,
      "epoch": 1.3889109320820467,
      "grad_norm": 0.16377359628677368,
      "learning_rate": 5.370625939215506e-05,
      "loss": 0.6805,
      "mean_token_accuracy": 0.790365893393755,
      "num_tokens": 481795933.0,
      "step": 15096
    },
    {
      "entropy": 0.6428903508931398,
      "epoch": 1.3890029384534879,
      "grad_norm": 0.15732133388519287,
      "learning_rate": 5.370319256601344e-05,
      "loss": 0.6451,
      "mean_token_accuracy": 0.8055059239268303,
      "num_tokens": 481828014.0,
      "step": 15097
    },
    {
      "entropy": 0.5475013272371143,
      "epoch": 1.389094944824929,
      "grad_norm": 0.15568320453166962,
      "learning_rate": 5.370012573987181e-05,
      "loss": 0.5323,
      "mean_token_accuracy": 0.8361629955470562,
      "num_tokens": 481859716.0,
      "step": 15098
    },
    {
      "entropy": 0.6448834575712681,
      "epoch": 1.3891869511963704,
      "grad_norm": 0.15926487743854523,
      "learning_rate": 5.369705891373018e-05,
      "loss": 0.6156,
      "mean_token_accuracy": 0.8070113472640514,
      "num_tokens": 481890423.0,
      "step": 15099
    },
    {
      "entropy": 0.632417744025588,
      "epoch": 1.3892789575678115,
      "grad_norm": 0.16104784607887268,
      "learning_rate": 5.369399208758856e-05,
      "loss": 0.622,
      "mean_token_accuracy": 0.8101359158754349,
      "num_tokens": 481921819.0,
      "step": 15100
    },
    {
      "entropy": 0.6189237516373396,
      "epoch": 1.3893709639392529,
      "grad_norm": 0.16517768800258636,
      "learning_rate": 5.369092526144693e-05,
      "loss": 0.6217,
      "mean_token_accuracy": 0.809347290545702,
      "num_tokens": 481953221.0,
      "step": 15101
    },
    {
      "entropy": 0.6692878734320402,
      "epoch": 1.389462970310694,
      "grad_norm": 0.16486956179141998,
      "learning_rate": 5.3687858435305305e-05,
      "loss": 0.6581,
      "mean_token_accuracy": 0.8012467063963413,
      "num_tokens": 481984899.0,
      "step": 15102
    },
    {
      "entropy": 0.5981682930141687,
      "epoch": 1.3895549766821351,
      "grad_norm": 0.15706056356430054,
      "learning_rate": 5.368479160916367e-05,
      "loss": 0.5808,
      "mean_token_accuracy": 0.8187293596565723,
      "num_tokens": 482016691.0,
      "step": 15103
    },
    {
      "entropy": 0.5772356381639838,
      "epoch": 1.3896469830535765,
      "grad_norm": 0.1499229073524475,
      "learning_rate": 5.368172478302206e-05,
      "loss": 0.5696,
      "mean_token_accuracy": 0.8235937356948853,
      "num_tokens": 482048718.0,
      "step": 15104
    },
    {
      "entropy": 0.5982987419702113,
      "epoch": 1.3897389894250176,
      "grad_norm": 0.15820346772670746,
      "learning_rate": 5.367865795688043e-05,
      "loss": 0.5839,
      "mean_token_accuracy": 0.8204263933002949,
      "num_tokens": 482080185.0,
      "step": 15105
    },
    {
      "entropy": 0.6113948160782456,
      "epoch": 1.389830995796459,
      "grad_norm": 0.15892209112644196,
      "learning_rate": 5.36755911307388e-05,
      "loss": 0.5948,
      "mean_token_accuracy": 0.8170584961771965,
      "num_tokens": 482112484.0,
      "step": 15106
    },
    {
      "entropy": 0.6410810761153698,
      "epoch": 1.3899230021679,
      "grad_norm": 0.15426281094551086,
      "learning_rate": 5.367252430459717e-05,
      "loss": 0.6252,
      "mean_token_accuracy": 0.808081179857254,
      "num_tokens": 482144601.0,
      "step": 15107
    },
    {
      "entropy": 0.6087575359269977,
      "epoch": 1.3900150085393412,
      "grad_norm": 0.1544707864522934,
      "learning_rate": 5.366945747845555e-05,
      "loss": 0.6042,
      "mean_token_accuracy": 0.8129563443362713,
      "num_tokens": 482176475.0,
      "step": 15108
    },
    {
      "entropy": 0.6184211429208517,
      "epoch": 1.3901070149107826,
      "grad_norm": 0.1629578173160553,
      "learning_rate": 5.366639065231393e-05,
      "loss": 0.6067,
      "mean_token_accuracy": 0.8142150230705738,
      "num_tokens": 482209084.0,
      "step": 15109
    },
    {
      "entropy": 0.7095652464777231,
      "epoch": 1.3901990212822237,
      "grad_norm": 0.18161119520664215,
      "learning_rate": 5.3663323826172296e-05,
      "loss": 0.7173,
      "mean_token_accuracy": 0.7839862704277039,
      "num_tokens": 482239812.0,
      "step": 15110
    },
    {
      "entropy": 0.5860241409391165,
      "epoch": 1.390291027653665,
      "grad_norm": 0.15363487601280212,
      "learning_rate": 5.366025700003067e-05,
      "loss": 0.576,
      "mean_token_accuracy": 0.8232356011867523,
      "num_tokens": 482271838.0,
      "step": 15111
    },
    {
      "entropy": 0.6921687722206116,
      "epoch": 1.3903830340251062,
      "grad_norm": 0.16051803529262543,
      "learning_rate": 5.365719017388905e-05,
      "loss": 0.6726,
      "mean_token_accuracy": 0.7954920716583729,
      "num_tokens": 482302868.0,
      "step": 15112
    },
    {
      "entropy": 0.6712545827031136,
      "epoch": 1.3904750403965473,
      "grad_norm": 0.15969017148017883,
      "learning_rate": 5.365412334774742e-05,
      "loss": 0.6667,
      "mean_token_accuracy": 0.7997078150510788,
      "num_tokens": 482334184.0,
      "step": 15113
    },
    {
      "entropy": 0.6661553159356117,
      "epoch": 1.3905670467679887,
      "grad_norm": 0.16148218512535095,
      "learning_rate": 5.3651056521605794e-05,
      "loss": 0.6631,
      "mean_token_accuracy": 0.7963365465402603,
      "num_tokens": 482366070.0,
      "step": 15114
    },
    {
      "entropy": 0.6717393640428782,
      "epoch": 1.3906590531394298,
      "grad_norm": 0.16962258517742157,
      "learning_rate": 5.364798969546416e-05,
      "loss": 0.6521,
      "mean_token_accuracy": 0.8041979223489761,
      "num_tokens": 482398429.0,
      "step": 15115
    },
    {
      "entropy": 0.6186907039955258,
      "epoch": 1.3907510595108712,
      "grad_norm": 0.15797926485538483,
      "learning_rate": 5.3644922869322544e-05,
      "loss": 0.607,
      "mean_token_accuracy": 0.8140568844974041,
      "num_tokens": 482430685.0,
      "step": 15116
    },
    {
      "entropy": 0.686114065349102,
      "epoch": 1.3908430658823123,
      "grad_norm": 0.15757736563682556,
      "learning_rate": 5.364185604318092e-05,
      "loss": 0.6713,
      "mean_token_accuracy": 0.7934623844921589,
      "num_tokens": 482462506.0,
      "step": 15117
    },
    {
      "entropy": 0.5939579354599118,
      "epoch": 1.3909350722537535,
      "grad_norm": 0.14462746679782867,
      "learning_rate": 5.3638789217039286e-05,
      "loss": 0.5603,
      "mean_token_accuracy": 0.8267203457653522,
      "num_tokens": 482494558.0,
      "step": 15118
    },
    {
      "entropy": 0.6170077370479703,
      "epoch": 1.3910270786251948,
      "grad_norm": 0.15392078459262848,
      "learning_rate": 5.363572239089766e-05,
      "loss": 0.6122,
      "mean_token_accuracy": 0.8120029978454113,
      "num_tokens": 482525932.0,
      "step": 15119
    },
    {
      "entropy": 0.6062852498143911,
      "epoch": 1.391119084996636,
      "grad_norm": 0.15510085225105286,
      "learning_rate": 5.363265556475604e-05,
      "loss": 0.5876,
      "mean_token_accuracy": 0.8184941411018372,
      "num_tokens": 482557348.0,
      "step": 15120
    },
    {
      "entropy": 0.6959567293524742,
      "epoch": 1.3912110913680773,
      "grad_norm": 0.1644957810640335,
      "learning_rate": 5.362958873861441e-05,
      "loss": 0.6867,
      "mean_token_accuracy": 0.7905135191977024,
      "num_tokens": 482589251.0,
      "step": 15121
    },
    {
      "entropy": 0.5766162150539458,
      "epoch": 1.3913030977395184,
      "grad_norm": 0.15538816154003143,
      "learning_rate": 5.3626521912472785e-05,
      "loss": 0.5719,
      "mean_token_accuracy": 0.8257271125912666,
      "num_tokens": 482621144.0,
      "step": 15122
    },
    {
      "entropy": 0.6682594511657953,
      "epoch": 1.3913951041109596,
      "grad_norm": 0.16092954576015472,
      "learning_rate": 5.362345508633115e-05,
      "loss": 0.6675,
      "mean_token_accuracy": 0.7980110384523869,
      "num_tokens": 482653298.0,
      "step": 15123
    },
    {
      "entropy": 0.6167380474507809,
      "epoch": 1.391487110482401,
      "grad_norm": 0.16455714404582977,
      "learning_rate": 5.3620388260189534e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.8157143071293831,
      "num_tokens": 482685461.0,
      "step": 15124
    },
    {
      "entropy": 0.5512083247303963,
      "epoch": 1.391579116853842,
      "grad_norm": 0.15424013137817383,
      "learning_rate": 5.361732143404791e-05,
      "loss": 0.5429,
      "mean_token_accuracy": 0.8315113820135593,
      "num_tokens": 482717535.0,
      "step": 15125
    },
    {
      "entropy": 0.7335672844201326,
      "epoch": 1.3916711232252834,
      "grad_norm": 0.1610735058784485,
      "learning_rate": 5.3614254607906276e-05,
      "loss": 0.724,
      "mean_token_accuracy": 0.7780125662684441,
      "num_tokens": 482749175.0,
      "step": 15126
    },
    {
      "entropy": 0.626841944642365,
      "epoch": 1.3917631295967245,
      "grad_norm": 0.15378136932849884,
      "learning_rate": 5.361118778176465e-05,
      "loss": 0.6204,
      "mean_token_accuracy": 0.8092267215251923,
      "num_tokens": 482781418.0,
      "step": 15127
    },
    {
      "entropy": 0.5979949235916138,
      "epoch": 1.3918551359681657,
      "grad_norm": 0.1568886637687683,
      "learning_rate": 5.360812095562303e-05,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8196239434182644,
      "num_tokens": 482813797.0,
      "step": 15128
    },
    {
      "entropy": 0.5593167655169964,
      "epoch": 1.391947142339607,
      "grad_norm": 0.16826483607292175,
      "learning_rate": 5.36050541294814e-05,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8281670399010181,
      "num_tokens": 482846075.0,
      "step": 15129
    },
    {
      "entropy": 0.6620109211653471,
      "epoch": 1.3920391487110482,
      "grad_norm": 0.1604452133178711,
      "learning_rate": 5.3601987303339775e-05,
      "loss": 0.6597,
      "mean_token_accuracy": 0.7980256751179695,
      "num_tokens": 482877916.0,
      "step": 15130
    },
    {
      "entropy": 0.5878921858966351,
      "epoch": 1.3921311550824895,
      "grad_norm": 0.15871334075927734,
      "learning_rate": 5.3598920477198156e-05,
      "loss": 0.573,
      "mean_token_accuracy": 0.8271348588168621,
      "num_tokens": 482910495.0,
      "step": 15131
    },
    {
      "entropy": 0.6243736958131194,
      "epoch": 1.3922231614539307,
      "grad_norm": 0.1612803190946579,
      "learning_rate": 5.3595853651056524e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8140394315123558,
      "num_tokens": 482941988.0,
      "step": 15132
    },
    {
      "entropy": 0.6687428373843431,
      "epoch": 1.3923151678253718,
      "grad_norm": 0.16080094873905182,
      "learning_rate": 5.35927868249149e-05,
      "loss": 0.6583,
      "mean_token_accuracy": 0.8032273538410664,
      "num_tokens": 482974104.0,
      "step": 15133
    },
    {
      "entropy": 0.5875765159726143,
      "epoch": 1.3924071741968131,
      "grad_norm": 0.1543435901403427,
      "learning_rate": 5.358971999877327e-05,
      "loss": 0.576,
      "mean_token_accuracy": 0.8235383108258247,
      "num_tokens": 483005592.0,
      "step": 15134
    },
    {
      "entropy": 0.6880113482475281,
      "epoch": 1.3924991805682543,
      "grad_norm": 0.15819425880908966,
      "learning_rate": 5.358665317263165e-05,
      "loss": 0.6695,
      "mean_token_accuracy": 0.7956010550260544,
      "num_tokens": 483037279.0,
      "step": 15135
    },
    {
      "entropy": 0.7027365975081921,
      "epoch": 1.3925911869396956,
      "grad_norm": 0.16683663427829742,
      "learning_rate": 5.358358634649002e-05,
      "loss": 0.6903,
      "mean_token_accuracy": 0.7966227754950523,
      "num_tokens": 483069485.0,
      "step": 15136
    },
    {
      "entropy": 0.5587956588715315,
      "epoch": 1.3926831933111368,
      "grad_norm": 0.15345001220703125,
      "learning_rate": 5.358051952034839e-05,
      "loss": 0.544,
      "mean_token_accuracy": 0.8347478024661541,
      "num_tokens": 483102113.0,
      "step": 15137
    },
    {
      "entropy": 0.6296444665640593,
      "epoch": 1.392775199682578,
      "grad_norm": 0.15397804975509644,
      "learning_rate": 5.3577452694206765e-05,
      "loss": 0.6119,
      "mean_token_accuracy": 0.8135824576020241,
      "num_tokens": 483134424.0,
      "step": 15138
    },
    {
      "entropy": 0.663333373144269,
      "epoch": 1.3928672060540193,
      "grad_norm": 0.16291698813438416,
      "learning_rate": 5.357438586806515e-05,
      "loss": 0.665,
      "mean_token_accuracy": 0.797654815018177,
      "num_tokens": 483166032.0,
      "step": 15139
    },
    {
      "entropy": 0.5417357259429991,
      "epoch": 1.3929592124254604,
      "grad_norm": 0.1429964154958725,
      "learning_rate": 5.3571319041923515e-05,
      "loss": 0.5299,
      "mean_token_accuracy": 0.8367475718259811,
      "num_tokens": 483198733.0,
      "step": 15140
    },
    {
      "entropy": 0.5986736174672842,
      "epoch": 1.3930512187969017,
      "grad_norm": 0.15980912744998932,
      "learning_rate": 5.356825221578189e-05,
      "loss": 0.5887,
      "mean_token_accuracy": 0.8174174875020981,
      "num_tokens": 483229617.0,
      "step": 15141
    },
    {
      "entropy": 0.6108477395027876,
      "epoch": 1.3931432251683429,
      "grad_norm": 0.15039627254009247,
      "learning_rate": 5.356518538964026e-05,
      "loss": 0.5971,
      "mean_token_accuracy": 0.8175426796078682,
      "num_tokens": 483261544.0,
      "step": 15142
    },
    {
      "entropy": 0.6331802429631352,
      "epoch": 1.393235231539784,
      "grad_norm": 0.1577913761138916,
      "learning_rate": 5.3562118563498645e-05,
      "loss": 0.612,
      "mean_token_accuracy": 0.8132497631013393,
      "num_tokens": 483294043.0,
      "step": 15143
    },
    {
      "entropy": 0.6123892674222589,
      "epoch": 1.3933272379112254,
      "grad_norm": 0.15725749731063843,
      "learning_rate": 5.355905173735701e-05,
      "loss": 0.6069,
      "mean_token_accuracy": 0.8128358013927937,
      "num_tokens": 483326719.0,
      "step": 15144
    },
    {
      "entropy": 0.6221149675548077,
      "epoch": 1.3934192442826665,
      "grad_norm": 0.15716104209423065,
      "learning_rate": 5.355598491121538e-05,
      "loss": 0.6103,
      "mean_token_accuracy": 0.8196910358965397,
      "num_tokens": 483359154.0,
      "step": 15145
    },
    {
      "entropy": 0.7047959910705686,
      "epoch": 1.3935112506541079,
      "grad_norm": 0.16235917806625366,
      "learning_rate": 5.3552918085073756e-05,
      "loss": 0.6914,
      "mean_token_accuracy": 0.7941413559019566,
      "num_tokens": 483391311.0,
      "step": 15146
    },
    {
      "entropy": 0.5327480845153332,
      "epoch": 1.393603257025549,
      "grad_norm": 0.1592390090227127,
      "learning_rate": 5.354985125893214e-05,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8411899879574776,
      "num_tokens": 483423556.0,
      "step": 15147
    },
    {
      "entropy": 0.5458124671131372,
      "epoch": 1.3936952633969901,
      "grad_norm": 0.1587514728307724,
      "learning_rate": 5.354678443279051e-05,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8343716636300087,
      "num_tokens": 483456051.0,
      "step": 15148
    },
    {
      "entropy": 0.6273721754550934,
      "epoch": 1.3937872697684315,
      "grad_norm": 0.15528284013271332,
      "learning_rate": 5.354371760664888e-05,
      "loss": 0.611,
      "mean_token_accuracy": 0.8123288825154305,
      "num_tokens": 483488425.0,
      "step": 15149
    },
    {
      "entropy": 0.6388930976390839,
      "epoch": 1.3938792761398726,
      "grad_norm": 0.1615820825099945,
      "learning_rate": 5.354065078050725e-05,
      "loss": 0.63,
      "mean_token_accuracy": 0.8074002489447594,
      "num_tokens": 483519192.0,
      "step": 15150
    },
    {
      "entropy": 0.5461455336771905,
      "epoch": 1.393971282511314,
      "grad_norm": 0.159457266330719,
      "learning_rate": 5.3537583954365636e-05,
      "loss": 0.5256,
      "mean_token_accuracy": 0.8340172134339809,
      "num_tokens": 483550453.0,
      "step": 15151
    },
    {
      "entropy": 0.6648688688874245,
      "epoch": 1.394063288882755,
      "grad_norm": 0.16021738946437836,
      "learning_rate": 5.3534517128224004e-05,
      "loss": 0.6638,
      "mean_token_accuracy": 0.7980843149125576,
      "num_tokens": 483581984.0,
      "step": 15152
    },
    {
      "entropy": 0.5930600259453058,
      "epoch": 1.3941552952541962,
      "grad_norm": 0.15382273495197296,
      "learning_rate": 5.353145030208238e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8190644793212414,
      "num_tokens": 483614752.0,
      "step": 15153
    },
    {
      "entropy": 0.6720713023096323,
      "epoch": 1.3942473016256376,
      "grad_norm": 0.16558700799942017,
      "learning_rate": 5.3528383475940746e-05,
      "loss": 0.6642,
      "mean_token_accuracy": 0.794465396553278,
      "num_tokens": 483646671.0,
      "step": 15154
    },
    {
      "entropy": 0.5876091551035643,
      "epoch": 1.3943393079970787,
      "grad_norm": 0.15854030847549438,
      "learning_rate": 5.352531664979913e-05,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8225438892841339,
      "num_tokens": 483678609.0,
      "step": 15155
    },
    {
      "entropy": 0.6670588199049234,
      "epoch": 1.39443131436852,
      "grad_norm": 0.16130530834197998,
      "learning_rate": 5.35222498236575e-05,
      "loss": 0.6585,
      "mean_token_accuracy": 0.7997955605387688,
      "num_tokens": 483710703.0,
      "step": 15156
    },
    {
      "entropy": 0.6715096868574619,
      "epoch": 1.3945233207399612,
      "grad_norm": 0.16376253962516785,
      "learning_rate": 5.351918299751587e-05,
      "loss": 0.6762,
      "mean_token_accuracy": 0.7949693687260151,
      "num_tokens": 483742802.0,
      "step": 15157
    },
    {
      "entropy": 0.6999020464718342,
      "epoch": 1.3946153271114023,
      "grad_norm": 0.16731682419776917,
      "learning_rate": 5.3516116171374245e-05,
      "loss": 0.6997,
      "mean_token_accuracy": 0.7870360687375069,
      "num_tokens": 483774900.0,
      "step": 15158
    },
    {
      "entropy": 0.5660539418458939,
      "epoch": 1.3947073334828437,
      "grad_norm": 0.1534574329853058,
      "learning_rate": 5.3513049345232626e-05,
      "loss": 0.5484,
      "mean_token_accuracy": 0.8297363631427288,
      "num_tokens": 483806530.0,
      "step": 15159
    },
    {
      "entropy": 0.5544678214937449,
      "epoch": 1.3947993398542848,
      "grad_norm": 0.1502067893743515,
      "learning_rate": 5.3509982519090994e-05,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8286012783646584,
      "num_tokens": 483839038.0,
      "step": 15160
    },
    {
      "entropy": 0.7234027162194252,
      "epoch": 1.3948913462257262,
      "grad_norm": 0.16950462758541107,
      "learning_rate": 5.350691569294937e-05,
      "loss": 0.7142,
      "mean_token_accuracy": 0.784739226102829,
      "num_tokens": 483870966.0,
      "step": 15161
    },
    {
      "entropy": 0.6277991263195872,
      "epoch": 1.3949833525971673,
      "grad_norm": 0.15764062106609344,
      "learning_rate": 5.350384886680775e-05,
      "loss": 0.6124,
      "mean_token_accuracy": 0.8104136660695076,
      "num_tokens": 483902605.0,
      "step": 15162
    },
    {
      "entropy": 0.53183702705428,
      "epoch": 1.3950753589686085,
      "grad_norm": 0.1528007835149765,
      "learning_rate": 5.350078204066612e-05,
      "loss": 0.5099,
      "mean_token_accuracy": 0.8410982824862003,
      "num_tokens": 483934070.0,
      "step": 15163
    },
    {
      "entropy": 0.682472831569612,
      "epoch": 1.3951673653400498,
      "grad_norm": 0.16072030365467072,
      "learning_rate": 5.349771521452449e-05,
      "loss": 0.6729,
      "mean_token_accuracy": 0.7979100272059441,
      "num_tokens": 483966109.0,
      "step": 15164
    },
    {
      "entropy": 0.6193535849452019,
      "epoch": 1.395259371711491,
      "grad_norm": 0.16889256238937378,
      "learning_rate": 5.349464838838286e-05,
      "loss": 0.6002,
      "mean_token_accuracy": 0.8126201555132866,
      "num_tokens": 483997292.0,
      "step": 15165
    },
    {
      "entropy": 0.6039032626431435,
      "epoch": 1.3953513780829323,
      "grad_norm": 0.15318071842193604,
      "learning_rate": 5.349158156224124e-05,
      "loss": 0.5983,
      "mean_token_accuracy": 0.8174367137253284,
      "num_tokens": 484028846.0,
      "step": 15166
    },
    {
      "entropy": 0.57245186669752,
      "epoch": 1.3954433844543734,
      "grad_norm": 0.1490100622177124,
      "learning_rate": 5.348851473609962e-05,
      "loss": 0.5683,
      "mean_token_accuracy": 0.8279928602278233,
      "num_tokens": 484060822.0,
      "step": 15167
    },
    {
      "entropy": 0.6788538582623005,
      "epoch": 1.3955353908258146,
      "grad_norm": 0.15527845919132233,
      "learning_rate": 5.3485447909957985e-05,
      "loss": 0.6707,
      "mean_token_accuracy": 0.7912475243210793,
      "num_tokens": 484093216.0,
      "step": 15168
    },
    {
      "entropy": 0.6916705071926117,
      "epoch": 1.395627397197256,
      "grad_norm": 0.16403347253799438,
      "learning_rate": 5.348238108381636e-05,
      "loss": 0.6795,
      "mean_token_accuracy": 0.792442187666893,
      "num_tokens": 484125032.0,
      "step": 15169
    },
    {
      "entropy": 0.5529796555638313,
      "epoch": 1.395719403568697,
      "grad_norm": 0.15420491993427277,
      "learning_rate": 5.347931425767474e-05,
      "loss": 0.5389,
      "mean_token_accuracy": 0.8368982337415218,
      "num_tokens": 484157524.0,
      "step": 15170
    },
    {
      "entropy": 0.5821375008672476,
      "epoch": 1.3958114099401384,
      "grad_norm": 0.1546124368906021,
      "learning_rate": 5.347624743153311e-05,
      "loss": 0.5746,
      "mean_token_accuracy": 0.8218651413917542,
      "num_tokens": 484189669.0,
      "step": 15171
    },
    {
      "entropy": 0.5060271229594946,
      "epoch": 1.3959034163115795,
      "grad_norm": 0.14916196465492249,
      "learning_rate": 5.347318060539148e-05,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8454355373978615,
      "num_tokens": 484221799.0,
      "step": 15172
    },
    {
      "entropy": 0.6362815643660724,
      "epoch": 1.3959954226830207,
      "grad_norm": 0.1530921310186386,
      "learning_rate": 5.347011377924985e-05,
      "loss": 0.6203,
      "mean_token_accuracy": 0.8115149028599262,
      "num_tokens": 484253815.0,
      "step": 15173
    },
    {
      "entropy": 0.5966842044144869,
      "epoch": 1.396087429054462,
      "grad_norm": 0.15439969301223755,
      "learning_rate": 5.346704695310823e-05,
      "loss": 0.5789,
      "mean_token_accuracy": 0.8246920295059681,
      "num_tokens": 484285406.0,
      "step": 15174
    },
    {
      "entropy": 0.6085546016693115,
      "epoch": 1.3961794354259032,
      "grad_norm": 0.15819723904132843,
      "learning_rate": 5.346398012696661e-05,
      "loss": 0.5946,
      "mean_token_accuracy": 0.821663610637188,
      "num_tokens": 484317943.0,
      "step": 15175
    },
    {
      "entropy": 0.697729249484837,
      "epoch": 1.3962714417973445,
      "grad_norm": 0.16205623745918274,
      "learning_rate": 5.3460913300824975e-05,
      "loss": 0.6853,
      "mean_token_accuracy": 0.7931268103420734,
      "num_tokens": 484350163.0,
      "step": 15176
    },
    {
      "entropy": 0.5941299023106694,
      "epoch": 1.3963634481687857,
      "grad_norm": 0.15642090141773224,
      "learning_rate": 5.345784647468335e-05,
      "loss": 0.5815,
      "mean_token_accuracy": 0.8201844803988934,
      "num_tokens": 484382426.0,
      "step": 15177
    },
    {
      "entropy": 0.55201296787709,
      "epoch": 1.3964554545402268,
      "grad_norm": 0.15704195201396942,
      "learning_rate": 5.345477964854173e-05,
      "loss": 0.5418,
      "mean_token_accuracy": 0.8316918984055519,
      "num_tokens": 484414413.0,
      "step": 15178
    },
    {
      "entropy": 0.7584404069930315,
      "epoch": 1.3965474609116681,
      "grad_norm": 0.16541312634944916,
      "learning_rate": 5.34517128224001e-05,
      "loss": 0.7494,
      "mean_token_accuracy": 0.7731148228049278,
      "num_tokens": 484445390.0,
      "step": 15179
    },
    {
      "entropy": 0.6860215864144266,
      "epoch": 1.3966394672831093,
      "grad_norm": 0.15675272047519684,
      "learning_rate": 5.3448645996258474e-05,
      "loss": 0.6617,
      "mean_token_accuracy": 0.8014574572443962,
      "num_tokens": 484477862.0,
      "step": 15180
    },
    {
      "entropy": 0.5848916256800294,
      "epoch": 1.3967314736545506,
      "grad_norm": 0.15256206691265106,
      "learning_rate": 5.344557917011684e-05,
      "loss": 0.5767,
      "mean_token_accuracy": 0.8234541155397892,
      "num_tokens": 484509191.0,
      "step": 15181
    },
    {
      "entropy": 0.6282033333554864,
      "epoch": 1.3968234800259918,
      "grad_norm": 0.1594715714454651,
      "learning_rate": 5.344251234397522e-05,
      "loss": 0.6183,
      "mean_token_accuracy": 0.8142357692122459,
      "num_tokens": 484541565.0,
      "step": 15182
    },
    {
      "entropy": 0.6050581801682711,
      "epoch": 1.396915486397433,
      "grad_norm": 0.15986128151416779,
      "learning_rate": 5.34394455178336e-05,
      "loss": 0.6052,
      "mean_token_accuracy": 0.8179474547505379,
      "num_tokens": 484573859.0,
      "step": 15183
    },
    {
      "entropy": 0.6309493817389011,
      "epoch": 1.3970074927688743,
      "grad_norm": 0.15460677444934845,
      "learning_rate": 5.3436378691691965e-05,
      "loss": 0.6255,
      "mean_token_accuracy": 0.8081497699022293,
      "num_tokens": 484605810.0,
      "step": 15184
    },
    {
      "entropy": 0.6876831278204918,
      "epoch": 1.3970994991403154,
      "grad_norm": 0.1792559176683426,
      "learning_rate": 5.343331186555034e-05,
      "loss": 0.6681,
      "mean_token_accuracy": 0.7965166233479977,
      "num_tokens": 484637735.0,
      "step": 15185
    },
    {
      "entropy": 0.6176596563309431,
      "epoch": 1.3971915055117567,
      "grad_norm": 0.15759097039699554,
      "learning_rate": 5.343024503940872e-05,
      "loss": 0.6052,
      "mean_token_accuracy": 0.8133482038974762,
      "num_tokens": 484669656.0,
      "step": 15186
    },
    {
      "entropy": 0.6143154455348849,
      "epoch": 1.3972835118831979,
      "grad_norm": 0.1557808369398117,
      "learning_rate": 5.3427178213267096e-05,
      "loss": 0.6065,
      "mean_token_accuracy": 0.8157669678330421,
      "num_tokens": 484702424.0,
      "step": 15187
    },
    {
      "entropy": 0.5916835647076368,
      "epoch": 1.397375518254639,
      "grad_norm": 0.15013185143470764,
      "learning_rate": 5.3424111387125464e-05,
      "loss": 0.5764,
      "mean_token_accuracy": 0.8240131437778473,
      "num_tokens": 484734809.0,
      "step": 15188
    },
    {
      "entropy": 0.5911649595946074,
      "epoch": 1.3974675246260804,
      "grad_norm": 0.15344995260238647,
      "learning_rate": 5.342104456098383e-05,
      "loss": 0.5681,
      "mean_token_accuracy": 0.8247897066175938,
      "num_tokens": 484766928.0,
      "step": 15189
    },
    {
      "entropy": 0.6822904217988253,
      "epoch": 1.3975595309975215,
      "grad_norm": 0.16478902101516724,
      "learning_rate": 5.341797773484222e-05,
      "loss": 0.6685,
      "mean_token_accuracy": 0.7963370271027088,
      "num_tokens": 484799163.0,
      "step": 15190
    },
    {
      "entropy": 0.5985718015581369,
      "epoch": 1.3976515373689629,
      "grad_norm": 0.1552872657775879,
      "learning_rate": 5.341491090870059e-05,
      "loss": 0.5939,
      "mean_token_accuracy": 0.8170474655926228,
      "num_tokens": 484831495.0,
      "step": 15191
    },
    {
      "entropy": 0.7098405598662794,
      "epoch": 1.397743543740404,
      "grad_norm": 0.1600891649723053,
      "learning_rate": 5.341184408255896e-05,
      "loss": 0.6922,
      "mean_token_accuracy": 0.7931519486010075,
      "num_tokens": 484863770.0,
      "step": 15192
    },
    {
      "entropy": 0.6049203081056476,
      "epoch": 1.3978355501118451,
      "grad_norm": 0.1632710099220276,
      "learning_rate": 5.3408777256417344e-05,
      "loss": 0.5827,
      "mean_token_accuracy": 0.8161060772836208,
      "num_tokens": 484895181.0,
      "step": 15193
    },
    {
      "entropy": 0.6696636416018009,
      "epoch": 1.3979275564832865,
      "grad_norm": 0.15896053612232208,
      "learning_rate": 5.340571043027571e-05,
      "loss": 0.6599,
      "mean_token_accuracy": 0.7996800169348717,
      "num_tokens": 484927422.0,
      "step": 15194
    },
    {
      "entropy": 0.5931581696495414,
      "epoch": 1.3980195628547276,
      "grad_norm": 0.14937488734722137,
      "learning_rate": 5.3402643604134086e-05,
      "loss": 0.5923,
      "mean_token_accuracy": 0.8202966675162315,
      "num_tokens": 484959784.0,
      "step": 15195
    },
    {
      "entropy": 0.561268787831068,
      "epoch": 1.398111569226169,
      "grad_norm": 0.15067073702812195,
      "learning_rate": 5.3399576777992454e-05,
      "loss": 0.552,
      "mean_token_accuracy": 0.8255253657698631,
      "num_tokens": 484991692.0,
      "step": 15196
    },
    {
      "entropy": 0.5587304532527924,
      "epoch": 1.39820357559761,
      "grad_norm": 0.14902344346046448,
      "learning_rate": 5.3396509951850836e-05,
      "loss": 0.5625,
      "mean_token_accuracy": 0.822979960590601,
      "num_tokens": 485023673.0,
      "step": 15197
    },
    {
      "entropy": 0.5788237694650888,
      "epoch": 1.3982955819690512,
      "grad_norm": 0.15042243897914886,
      "learning_rate": 5.339344312570921e-05,
      "loss": 0.5716,
      "mean_token_accuracy": 0.8225456848740578,
      "num_tokens": 485055903.0,
      "step": 15198
    },
    {
      "entropy": 0.6593305841088295,
      "epoch": 1.3983875883404926,
      "grad_norm": 0.1631711721420288,
      "learning_rate": 5.339037629956758e-05,
      "loss": 0.6501,
      "mean_token_accuracy": 0.8011138886213303,
      "num_tokens": 485088030.0,
      "step": 15199
    },
    {
      "entropy": 0.5737909395247698,
      "epoch": 1.3984795947119337,
      "grad_norm": 0.14803124964237213,
      "learning_rate": 5.338730947342595e-05,
      "loss": 0.5659,
      "mean_token_accuracy": 0.8224553130567074,
      "num_tokens": 485120471.0,
      "step": 15200
    },
    {
      "entropy": 0.6076087616384029,
      "epoch": 1.398571601083375,
      "grad_norm": 0.15843670070171356,
      "learning_rate": 5.3384242647284334e-05,
      "loss": 0.6038,
      "mean_token_accuracy": 0.8165304698050022,
      "num_tokens": 485152501.0,
      "step": 15201
    },
    {
      "entropy": 0.6831332361325622,
      "epoch": 1.3986636074548162,
      "grad_norm": 0.1581231951713562,
      "learning_rate": 5.33811758211427e-05,
      "loss": 0.6809,
      "mean_token_accuracy": 0.7957522310316563,
      "num_tokens": 485184794.0,
      "step": 15202
    },
    {
      "entropy": 0.6252107750624418,
      "epoch": 1.3987556138262573,
      "grad_norm": 0.16439013183116913,
      "learning_rate": 5.337810899500108e-05,
      "loss": 0.6143,
      "mean_token_accuracy": 0.8105451688170433,
      "num_tokens": 485216554.0,
      "step": 15203
    },
    {
      "entropy": 0.6075268154963851,
      "epoch": 1.3988476201976987,
      "grad_norm": 0.1556159406900406,
      "learning_rate": 5.3375042168859445e-05,
      "loss": 0.5993,
      "mean_token_accuracy": 0.8188886493444443,
      "num_tokens": 485248734.0,
      "step": 15204
    },
    {
      "entropy": 0.5392001215368509,
      "epoch": 1.3989396265691398,
      "grad_norm": 0.14413069188594818,
      "learning_rate": 5.3371975342717826e-05,
      "loss": 0.5292,
      "mean_token_accuracy": 0.8376503512263298,
      "num_tokens": 485281104.0,
      "step": 15205
    },
    {
      "entropy": 0.6098469896242023,
      "epoch": 1.3990316329405812,
      "grad_norm": 0.15596473217010498,
      "learning_rate": 5.33689085165762e-05,
      "loss": 0.5897,
      "mean_token_accuracy": 0.8203212581574917,
      "num_tokens": 485312373.0,
      "step": 15206
    },
    {
      "entropy": 0.5912623507902026,
      "epoch": 1.3991236393120223,
      "grad_norm": 0.15412019193172455,
      "learning_rate": 5.336584169043457e-05,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8231315352022648,
      "num_tokens": 485344848.0,
      "step": 15207
    },
    {
      "entropy": 0.527862417511642,
      "epoch": 1.3992156456834635,
      "grad_norm": 0.15142126381397247,
      "learning_rate": 5.336277486429294e-05,
      "loss": 0.4994,
      "mean_token_accuracy": 0.8451798632740974,
      "num_tokens": 485376627.0,
      "step": 15208
    },
    {
      "entropy": 0.555587169714272,
      "epoch": 1.3993076520549048,
      "grad_norm": 0.16320593655109406,
      "learning_rate": 5.3359708038151325e-05,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8346041962504387,
      "num_tokens": 485408289.0,
      "step": 15209
    },
    {
      "entropy": 0.6007113326340914,
      "epoch": 1.399399658426346,
      "grad_norm": 0.15475252270698547,
      "learning_rate": 5.335664121200969e-05,
      "loss": 0.5944,
      "mean_token_accuracy": 0.8216967210173607,
      "num_tokens": 485440522.0,
      "step": 15210
    },
    {
      "entropy": 0.59678990021348,
      "epoch": 1.3994916647977873,
      "grad_norm": 0.14872875809669495,
      "learning_rate": 5.335357438586807e-05,
      "loss": 0.5821,
      "mean_token_accuracy": 0.8195930309593678,
      "num_tokens": 485472855.0,
      "step": 15211
    },
    {
      "entropy": 0.6215633004903793,
      "epoch": 1.3995836711692284,
      "grad_norm": 0.1588040590286255,
      "learning_rate": 5.3350507559726435e-05,
      "loss": 0.6131,
      "mean_token_accuracy": 0.8101638369262218,
      "num_tokens": 485505063.0,
      "step": 15212
    },
    {
      "entropy": 0.5540709402412176,
      "epoch": 1.3996756775406696,
      "grad_norm": 0.15621674060821533,
      "learning_rate": 5.3347440733584817e-05,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8296704478561878,
      "num_tokens": 485536799.0,
      "step": 15213
    },
    {
      "entropy": 0.5643017990514636,
      "epoch": 1.399767683912111,
      "grad_norm": 0.15854327380657196,
      "learning_rate": 5.334437390744319e-05,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8296115174889565,
      "num_tokens": 485569105.0,
      "step": 15214
    },
    {
      "entropy": 0.6627131756395102,
      "epoch": 1.399859690283552,
      "grad_norm": 0.15880024433135986,
      "learning_rate": 5.334130708130156e-05,
      "loss": 0.6596,
      "mean_token_accuracy": 0.8003541603684425,
      "num_tokens": 485601112.0,
      "step": 15215
    },
    {
      "entropy": 0.5065802987664938,
      "epoch": 1.3999516966549934,
      "grad_norm": 0.16081617772579193,
      "learning_rate": 5.3338240255159934e-05,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8476205505430698,
      "num_tokens": 485632573.0,
      "step": 15216
    },
    {
      "entropy": 0.6537554487586021,
      "epoch": 1.4000437030264345,
      "grad_norm": 0.16771452128887177,
      "learning_rate": 5.3335173429018315e-05,
      "loss": 0.6457,
      "mean_token_accuracy": 0.8020741268992424,
      "num_tokens": 485664617.0,
      "step": 15217
    },
    {
      "entropy": 0.6495654154568911,
      "epoch": 1.4001357093978757,
      "grad_norm": 0.1575821489095688,
      "learning_rate": 5.333210660287668e-05,
      "loss": 0.6499,
      "mean_token_accuracy": 0.8015691041946411,
      "num_tokens": 485696087.0,
      "step": 15218
    },
    {
      "entropy": 0.6954506144393235,
      "epoch": 1.400227715769317,
      "grad_norm": 0.15388034284114838,
      "learning_rate": 5.332903977673506e-05,
      "loss": 0.6819,
      "mean_token_accuracy": 0.7946239709854126,
      "num_tokens": 485728657.0,
      "step": 15219
    },
    {
      "entropy": 0.6791565306484699,
      "epoch": 1.4003197221407582,
      "grad_norm": 0.15990084409713745,
      "learning_rate": 5.3325972950593426e-05,
      "loss": 0.6592,
      "mean_token_accuracy": 0.8013941757380962,
      "num_tokens": 485760532.0,
      "step": 15220
    },
    {
      "entropy": 0.7030488289892673,
      "epoch": 1.4004117285121995,
      "grad_norm": 0.15555597841739655,
      "learning_rate": 5.332290612445181e-05,
      "loss": 0.6876,
      "mean_token_accuracy": 0.792548306286335,
      "num_tokens": 485792546.0,
      "step": 15221
    },
    {
      "entropy": 0.5610217610374093,
      "epoch": 1.4005037348836407,
      "grad_norm": 0.1551191359758377,
      "learning_rate": 5.331983929831018e-05,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8344278819859028,
      "num_tokens": 485824961.0,
      "step": 15222
    },
    {
      "entropy": 0.6276563964784145,
      "epoch": 1.4005957412550818,
      "grad_norm": 0.1511392593383789,
      "learning_rate": 5.331677247216855e-05,
      "loss": 0.6182,
      "mean_token_accuracy": 0.8104291968047619,
      "num_tokens": 485857097.0,
      "step": 15223
    },
    {
      "entropy": 0.6451674802228808,
      "epoch": 1.4006877476265231,
      "grad_norm": 0.161118283867836,
      "learning_rate": 5.331370564602694e-05,
      "loss": 0.63,
      "mean_token_accuracy": 0.8086323365569115,
      "num_tokens": 485889551.0,
      "step": 15224
    },
    {
      "entropy": 0.5949078947305679,
      "epoch": 1.4007797539979643,
      "grad_norm": 0.1516989916563034,
      "learning_rate": 5.3310638819885306e-05,
      "loss": 0.5809,
      "mean_token_accuracy": 0.8196915052831173,
      "num_tokens": 485921183.0,
      "step": 15225
    },
    {
      "entropy": 0.6014637108892202,
      "epoch": 1.4008717603694056,
      "grad_norm": 0.1590190827846527,
      "learning_rate": 5.330757199374368e-05,
      "loss": 0.5801,
      "mean_token_accuracy": 0.8204270340502262,
      "num_tokens": 485953124.0,
      "step": 15226
    },
    {
      "entropy": 0.6055864011868834,
      "epoch": 1.4009637667408468,
      "grad_norm": 0.153182715177536,
      "learning_rate": 5.330450516760205e-05,
      "loss": 0.5921,
      "mean_token_accuracy": 0.8175683915615082,
      "num_tokens": 485985892.0,
      "step": 15227
    },
    {
      "entropy": 0.6746439440175891,
      "epoch": 1.401055773112288,
      "grad_norm": 0.16355878114700317,
      "learning_rate": 5.330143834146043e-05,
      "loss": 0.6709,
      "mean_token_accuracy": 0.8003303036093712,
      "num_tokens": 486017769.0,
      "step": 15228
    },
    {
      "entropy": 0.6786901112645864,
      "epoch": 1.4011477794837293,
      "grad_norm": 0.16573598980903625,
      "learning_rate": 5.3298371515318804e-05,
      "loss": 0.6694,
      "mean_token_accuracy": 0.7920028530061245,
      "num_tokens": 486049211.0,
      "step": 15229
    },
    {
      "entropy": 0.6334695387631655,
      "epoch": 1.4012397858551704,
      "grad_norm": 0.15296554565429688,
      "learning_rate": 5.329530468917717e-05,
      "loss": 0.6232,
      "mean_token_accuracy": 0.8098510801792145,
      "num_tokens": 486081593.0,
      "step": 15230
    },
    {
      "entropy": 0.6893298709765077,
      "epoch": 1.4013317922266117,
      "grad_norm": 0.1586633026599884,
      "learning_rate": 5.329223786303555e-05,
      "loss": 0.6809,
      "mean_token_accuracy": 0.7918584458529949,
      "num_tokens": 486114197.0,
      "step": 15231
    },
    {
      "entropy": 0.5593343237414956,
      "epoch": 1.4014237985980529,
      "grad_norm": 0.15181468427181244,
      "learning_rate": 5.328917103689393e-05,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8326040469110012,
      "num_tokens": 486146109.0,
      "step": 15232
    },
    {
      "entropy": 0.7007709871977568,
      "epoch": 1.401515804969494,
      "grad_norm": 0.16239044070243835,
      "learning_rate": 5.3286104210752296e-05,
      "loss": 0.6926,
      "mean_token_accuracy": 0.7841852270066738,
      "num_tokens": 486177772.0,
      "step": 15233
    },
    {
      "entropy": 0.5815663831308484,
      "epoch": 1.4016078113409354,
      "grad_norm": 0.15620234608650208,
      "learning_rate": 5.328303738461067e-05,
      "loss": 0.5758,
      "mean_token_accuracy": 0.823066707700491,
      "num_tokens": 486210090.0,
      "step": 15234
    },
    {
      "entropy": 0.6021432112902403,
      "epoch": 1.4016998177123765,
      "grad_norm": 0.16076600551605225,
      "learning_rate": 5.327997055846904e-05,
      "loss": 0.6036,
      "mean_token_accuracy": 0.8169089555740356,
      "num_tokens": 486242110.0,
      "step": 15235
    },
    {
      "entropy": 0.6010150676593184,
      "epoch": 1.4017918240838179,
      "grad_norm": 0.16007660329341888,
      "learning_rate": 5.327690373232742e-05,
      "loss": 0.5865,
      "mean_token_accuracy": 0.8163632825016975,
      "num_tokens": 486273921.0,
      "step": 15236
    },
    {
      "entropy": 0.4796686419285834,
      "epoch": 1.401883830455259,
      "grad_norm": 0.1419937014579773,
      "learning_rate": 5.3273836906185795e-05,
      "loss": 0.4596,
      "mean_token_accuracy": 0.8537754788994789,
      "num_tokens": 486306359.0,
      "step": 15237
    },
    {
      "entropy": 0.5681639146059752,
      "epoch": 1.4019758368267001,
      "grad_norm": 0.15945148468017578,
      "learning_rate": 5.327077008004416e-05,
      "loss": 0.5577,
      "mean_token_accuracy": 0.8271456174552441,
      "num_tokens": 486336789.0,
      "step": 15238
    },
    {
      "entropy": 0.5394621519371867,
      "epoch": 1.4020678431981415,
      "grad_norm": 0.14996901154518127,
      "learning_rate": 5.326770325390254e-05,
      "loss": 0.5248,
      "mean_token_accuracy": 0.8358619809150696,
      "num_tokens": 486368460.0,
      "step": 15239
    },
    {
      "entropy": 0.6015882678329945,
      "epoch": 1.4021598495695826,
      "grad_norm": 0.15540194511413574,
      "learning_rate": 5.326463642776092e-05,
      "loss": 0.5887,
      "mean_token_accuracy": 0.818665899336338,
      "num_tokens": 486400537.0,
      "step": 15240
    },
    {
      "entropy": 0.7016670648008585,
      "epoch": 1.402251855941024,
      "grad_norm": 0.16642850637435913,
      "learning_rate": 5.3261569601619286e-05,
      "loss": 0.6944,
      "mean_token_accuracy": 0.79373649507761,
      "num_tokens": 486432344.0,
      "step": 15241
    },
    {
      "entropy": 0.594503091648221,
      "epoch": 1.402343862312465,
      "grad_norm": 0.16048982739448547,
      "learning_rate": 5.325850277547766e-05,
      "loss": 0.5769,
      "mean_token_accuracy": 0.8236000649631023,
      "num_tokens": 486463964.0,
      "step": 15242
    },
    {
      "entropy": 0.6360301962122321,
      "epoch": 1.4024358686839062,
      "grad_norm": 0.1533174216747284,
      "learning_rate": 5.325543594933603e-05,
      "loss": 0.6341,
      "mean_token_accuracy": 0.8057439289987087,
      "num_tokens": 486496207.0,
      "step": 15243
    },
    {
      "entropy": 0.5756992064416409,
      "epoch": 1.4025278750553476,
      "grad_norm": 0.1603887975215912,
      "learning_rate": 5.325236912319441e-05,
      "loss": 0.5754,
      "mean_token_accuracy": 0.8246300406754017,
      "num_tokens": 486527061.0,
      "step": 15244
    },
    {
      "entropy": 0.5535460205283016,
      "epoch": 1.4026198814267887,
      "grad_norm": 0.1640520840883255,
      "learning_rate": 5.3249302297052785e-05,
      "loss": 0.549,
      "mean_token_accuracy": 0.8320127427577972,
      "num_tokens": 486558928.0,
      "step": 15245
    },
    {
      "entropy": 0.6114097870886326,
      "epoch": 1.40271188779823,
      "grad_norm": 0.16247598826885223,
      "learning_rate": 5.324623547091115e-05,
      "loss": 0.6071,
      "mean_token_accuracy": 0.8134049102663994,
      "num_tokens": 486590640.0,
      "step": 15246
    },
    {
      "entropy": 0.6356755746528506,
      "epoch": 1.4028038941696712,
      "grad_norm": 0.1648918092250824,
      "learning_rate": 5.324316864476953e-05,
      "loss": 0.6233,
      "mean_token_accuracy": 0.8065075054764748,
      "num_tokens": 486622335.0,
      "step": 15247
    },
    {
      "entropy": 0.4993219589814544,
      "epoch": 1.4028959005411124,
      "grad_norm": 0.15179064869880676,
      "learning_rate": 5.324010181862791e-05,
      "loss": 0.4683,
      "mean_token_accuracy": 0.8534139581024647,
      "num_tokens": 486654111.0,
      "step": 15248
    },
    {
      "entropy": 0.7238290384411812,
      "epoch": 1.4029879069125537,
      "grad_norm": 0.16211949288845062,
      "learning_rate": 5.323703499248628e-05,
      "loss": 0.7086,
      "mean_token_accuracy": 0.7844134569168091,
      "num_tokens": 486686462.0,
      "step": 15249
    },
    {
      "entropy": 0.6047627795487642,
      "epoch": 1.4030799132839948,
      "grad_norm": 0.15794150531291962,
      "learning_rate": 5.323396816634465e-05,
      "loss": 0.5955,
      "mean_token_accuracy": 0.8187708333134651,
      "num_tokens": 486717688.0,
      "step": 15250
    },
    {
      "entropy": 0.4594041993841529,
      "epoch": 1.4031719196554362,
      "grad_norm": 0.15585875511169434,
      "learning_rate": 5.323090134020302e-05,
      "loss": 0.4385,
      "mean_token_accuracy": 0.8645736239850521,
      "num_tokens": 486750143.0,
      "step": 15251
    },
    {
      "entropy": 0.5738633936271071,
      "epoch": 1.4032639260268773,
      "grad_norm": 0.16547007858753204,
      "learning_rate": 5.32278345140614e-05,
      "loss": 0.5629,
      "mean_token_accuracy": 0.8271709866821766,
      "num_tokens": 486781830.0,
      "step": 15252
    },
    {
      "entropy": 0.5646119732409716,
      "epoch": 1.4033559323983185,
      "grad_norm": 0.14784854650497437,
      "learning_rate": 5.3224767687919775e-05,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8268708139657974,
      "num_tokens": 486813861.0,
      "step": 15253
    },
    {
      "entropy": 0.6528734341263771,
      "epoch": 1.4034479387697598,
      "grad_norm": 0.16293881833553314,
      "learning_rate": 5.322170086177814e-05,
      "loss": 0.6476,
      "mean_token_accuracy": 0.80480170622468,
      "num_tokens": 486845994.0,
      "step": 15254
    },
    {
      "entropy": 0.64247758500278,
      "epoch": 1.403539945141201,
      "grad_norm": 0.15516386926174164,
      "learning_rate": 5.3218634035636525e-05,
      "loss": 0.6273,
      "mean_token_accuracy": 0.8098942525684834,
      "num_tokens": 486878265.0,
      "step": 15255
    },
    {
      "entropy": 0.6705501372925937,
      "epoch": 1.4036319515126423,
      "grad_norm": 0.1608368307352066,
      "learning_rate": 5.32155672094949e-05,
      "loss": 0.6687,
      "mean_token_accuracy": 0.7994280122220516,
      "num_tokens": 486910126.0,
      "step": 15256
    },
    {
      "entropy": 0.5870199128985405,
      "epoch": 1.4037239578840834,
      "grad_norm": 0.15381260216236115,
      "learning_rate": 5.321250038335327e-05,
      "loss": 0.576,
      "mean_token_accuracy": 0.8207539096474648,
      "num_tokens": 486942611.0,
      "step": 15257
    },
    {
      "entropy": 0.5791694996878505,
      "epoch": 1.4038159642555246,
      "grad_norm": 0.16311593353748322,
      "learning_rate": 5.320943355721164e-05,
      "loss": 0.555,
      "mean_token_accuracy": 0.8277761563658714,
      "num_tokens": 486974240.0,
      "step": 15258
    },
    {
      "entropy": 0.575820900965482,
      "epoch": 1.403907970626966,
      "grad_norm": 0.15530037879943848,
      "learning_rate": 5.320636673107002e-05,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8271275907754898,
      "num_tokens": 487005536.0,
      "step": 15259
    },
    {
      "entropy": 0.6875792443752289,
      "epoch": 1.403999976998407,
      "grad_norm": 0.16182570159435272,
      "learning_rate": 5.320329990492839e-05,
      "loss": 0.6873,
      "mean_token_accuracy": 0.7926139608025551,
      "num_tokens": 487037497.0,
      "step": 15260
    },
    {
      "entropy": 0.6832538787275553,
      "epoch": 1.4040919833698484,
      "grad_norm": 0.16401934623718262,
      "learning_rate": 5.3200233078786766e-05,
      "loss": 0.6633,
      "mean_token_accuracy": 0.7977738454937935,
      "num_tokens": 487069060.0,
      "step": 15261
    },
    {
      "entropy": 0.6577444961294532,
      "epoch": 1.4041839897412896,
      "grad_norm": 0.16453154385089874,
      "learning_rate": 5.3197166252645134e-05,
      "loss": 0.6644,
      "mean_token_accuracy": 0.8009327761828899,
      "num_tokens": 487099872.0,
      "step": 15262
    },
    {
      "entropy": 0.6072616837918758,
      "epoch": 1.4042759961127307,
      "grad_norm": 0.15728069841861725,
      "learning_rate": 5.319409942650352e-05,
      "loss": 0.59,
      "mean_token_accuracy": 0.8159242458641529,
      "num_tokens": 487131190.0,
      "step": 15263
    },
    {
      "entropy": 0.6108940113335848,
      "epoch": 1.404368002484172,
      "grad_norm": 0.15343637764453888,
      "learning_rate": 5.319103260036189e-05,
      "loss": 0.6088,
      "mean_token_accuracy": 0.8160094544291496,
      "num_tokens": 487163510.0,
      "step": 15264
    },
    {
      "entropy": 0.5550868678838015,
      "epoch": 1.4044600088556132,
      "grad_norm": 0.15500247478485107,
      "learning_rate": 5.318796577422026e-05,
      "loss": 0.5435,
      "mean_token_accuracy": 0.829854317009449,
      "num_tokens": 487195579.0,
      "step": 15265
    },
    {
      "entropy": 0.6527521833777428,
      "epoch": 1.4045520152270545,
      "grad_norm": 0.1585613340139389,
      "learning_rate": 5.318489894807863e-05,
      "loss": 0.6561,
      "mean_token_accuracy": 0.8025599457323551,
      "num_tokens": 487227911.0,
      "step": 15266
    },
    {
      "entropy": 0.6035809554159641,
      "epoch": 1.4046440215984957,
      "grad_norm": 0.15534767508506775,
      "learning_rate": 5.3181832121937014e-05,
      "loss": 0.5935,
      "mean_token_accuracy": 0.8172976821660995,
      "num_tokens": 487259489.0,
      "step": 15267
    },
    {
      "entropy": 0.648838876048103,
      "epoch": 1.4047360279699368,
      "grad_norm": 0.15594948828220367,
      "learning_rate": 5.317876529579539e-05,
      "loss": 0.6315,
      "mean_token_accuracy": 0.8072623647749424,
      "num_tokens": 487292205.0,
      "step": 15268
    },
    {
      "entropy": 0.5898030889220536,
      "epoch": 1.4048280343413782,
      "grad_norm": 0.1494666188955307,
      "learning_rate": 5.3175698469653756e-05,
      "loss": 0.5664,
      "mean_token_accuracy": 0.8257778026163578,
      "num_tokens": 487324572.0,
      "step": 15269
    },
    {
      "entropy": 0.624253885820508,
      "epoch": 1.4049200407128193,
      "grad_norm": 0.1556445211172104,
      "learning_rate": 5.317263164351213e-05,
      "loss": 0.6128,
      "mean_token_accuracy": 0.8126649670302868,
      "num_tokens": 487356697.0,
      "step": 15270
    },
    {
      "entropy": 0.5445368764922023,
      "epoch": 1.4050120470842606,
      "grad_norm": 0.14631003141403198,
      "learning_rate": 5.316956481737051e-05,
      "loss": 0.5255,
      "mean_token_accuracy": 0.8400739170610905,
      "num_tokens": 487389156.0,
      "step": 15271
    },
    {
      "entropy": 0.6590197430923581,
      "epoch": 1.4051040534557018,
      "grad_norm": 0.16472162306308746,
      "learning_rate": 5.316649799122888e-05,
      "loss": 0.6557,
      "mean_token_accuracy": 0.8038139827549458,
      "num_tokens": 487420883.0,
      "step": 15272
    },
    {
      "entropy": 0.5184468715451658,
      "epoch": 1.405196059827143,
      "grad_norm": 0.15846216678619385,
      "learning_rate": 5.3163431165087255e-05,
      "loss": 0.4879,
      "mean_token_accuracy": 0.8468754626810551,
      "num_tokens": 487452284.0,
      "step": 15273
    },
    {
      "entropy": 0.5400831438601017,
      "epoch": 1.4052880661985843,
      "grad_norm": 0.15062986314296722,
      "learning_rate": 5.316036433894562e-05,
      "loss": 0.534,
      "mean_token_accuracy": 0.8377424627542496,
      "num_tokens": 487484233.0,
      "step": 15274
    },
    {
      "entropy": 0.6073900181800127,
      "epoch": 1.4053800725700254,
      "grad_norm": 0.16294530034065247,
      "learning_rate": 5.3157297512804004e-05,
      "loss": 0.5951,
      "mean_token_accuracy": 0.8181222081184387,
      "num_tokens": 487516461.0,
      "step": 15275
    },
    {
      "entropy": 0.5915073323994875,
      "epoch": 1.4054720789414668,
      "grad_norm": 0.15056703984737396,
      "learning_rate": 5.315423068666238e-05,
      "loss": 0.5839,
      "mean_token_accuracy": 0.8199011757969856,
      "num_tokens": 487548313.0,
      "step": 15276
    },
    {
      "entropy": 0.5703252851963043,
      "epoch": 1.4055640853129079,
      "grad_norm": 0.15520495176315308,
      "learning_rate": 5.3151163860520747e-05,
      "loss": 0.5572,
      "mean_token_accuracy": 0.8301072642207146,
      "num_tokens": 487580208.0,
      "step": 15277
    },
    {
      "entropy": 0.6348697040230036,
      "epoch": 1.405656091684349,
      "grad_norm": 0.16497385501861572,
      "learning_rate": 5.314809703437912e-05,
      "loss": 0.6372,
      "mean_token_accuracy": 0.8069254755973816,
      "num_tokens": 487611558.0,
      "step": 15278
    },
    {
      "entropy": 0.5036752466112375,
      "epoch": 1.4057480980557904,
      "grad_norm": 0.15566475689411163,
      "learning_rate": 5.31450302082375e-05,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8483621217310429,
      "num_tokens": 487643086.0,
      "step": 15279
    },
    {
      "entropy": 0.6196948057040572,
      "epoch": 1.4058401044272315,
      "grad_norm": 0.16236470639705658,
      "learning_rate": 5.314196338209587e-05,
      "loss": 0.6104,
      "mean_token_accuracy": 0.815090473741293,
      "num_tokens": 487674791.0,
      "step": 15280
    },
    {
      "entropy": 0.721234505996108,
      "epoch": 1.4059321107986729,
      "grad_norm": 0.16236263513565063,
      "learning_rate": 5.3138896555954245e-05,
      "loss": 0.7304,
      "mean_token_accuracy": 0.777460265904665,
      "num_tokens": 487706295.0,
      "step": 15281
    },
    {
      "entropy": 0.7408735733479261,
      "epoch": 1.406024117170114,
      "grad_norm": 0.16345198452472687,
      "learning_rate": 5.313582972981261e-05,
      "loss": 0.73,
      "mean_token_accuracy": 0.7822703868150711,
      "num_tokens": 487738702.0,
      "step": 15282
    },
    {
      "entropy": 0.542714549228549,
      "epoch": 1.4061161235415551,
      "grad_norm": 0.14653120934963226,
      "learning_rate": 5.3132762903670994e-05,
      "loss": 0.5215,
      "mean_token_accuracy": 0.8433949984610081,
      "num_tokens": 487770518.0,
      "step": 15283
    },
    {
      "entropy": 0.688910536468029,
      "epoch": 1.4062081299129965,
      "grad_norm": 0.16220290958881378,
      "learning_rate": 5.312969607752937e-05,
      "loss": 0.6765,
      "mean_token_accuracy": 0.7935415804386139,
      "num_tokens": 487802346.0,
      "step": 15284
    },
    {
      "entropy": 0.564340578392148,
      "epoch": 1.4063001362844376,
      "grad_norm": 0.15278306603431702,
      "learning_rate": 5.312662925138774e-05,
      "loss": 0.5283,
      "mean_token_accuracy": 0.8363953158259392,
      "num_tokens": 487834141.0,
      "step": 15285
    },
    {
      "entropy": 0.7222544541582465,
      "epoch": 1.406392142655879,
      "grad_norm": 0.165878564119339,
      "learning_rate": 5.312356242524612e-05,
      "loss": 0.7072,
      "mean_token_accuracy": 0.786216501146555,
      "num_tokens": 487866222.0,
      "step": 15286
    },
    {
      "entropy": 0.6307913810014725,
      "epoch": 1.40648414902732,
      "grad_norm": 0.15863999724388123,
      "learning_rate": 5.312049559910449e-05,
      "loss": 0.6154,
      "mean_token_accuracy": 0.8127797953784466,
      "num_tokens": 487897954.0,
      "step": 15287
    },
    {
      "entropy": 0.7433861009776592,
      "epoch": 1.4065761553987612,
      "grad_norm": 0.1666620671749115,
      "learning_rate": 5.311742877296286e-05,
      "loss": 0.7216,
      "mean_token_accuracy": 0.7858527153730392,
      "num_tokens": 487930717.0,
      "step": 15288
    },
    {
      "entropy": 0.6327617662027478,
      "epoch": 1.4066681617702026,
      "grad_norm": 0.1576872169971466,
      "learning_rate": 5.3114361946821236e-05,
      "loss": 0.624,
      "mean_token_accuracy": 0.8047334589064121,
      "num_tokens": 487962690.0,
      "step": 15289
    },
    {
      "entropy": 0.5742378709837794,
      "epoch": 1.4067601681416437,
      "grad_norm": 0.1516074389219284,
      "learning_rate": 5.311129512067962e-05,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8299559764564037,
      "num_tokens": 487994703.0,
      "step": 15290
    },
    {
      "entropy": 0.5793960941955447,
      "epoch": 1.406852174513085,
      "grad_norm": 0.16419410705566406,
      "learning_rate": 5.3108228294537985e-05,
      "loss": 0.5763,
      "mean_token_accuracy": 0.8232254646718502,
      "num_tokens": 488026290.0,
      "step": 15291
    },
    {
      "entropy": 0.7588746389374137,
      "epoch": 1.4069441808845262,
      "grad_norm": 0.16524772346019745,
      "learning_rate": 5.310516146839636e-05,
      "loss": 0.7511,
      "mean_token_accuracy": 0.774524874985218,
      "num_tokens": 488058327.0,
      "step": 15292
    },
    {
      "entropy": 0.6104453383013606,
      "epoch": 1.4070361872559674,
      "grad_norm": 0.1536269634962082,
      "learning_rate": 5.310209464225473e-05,
      "loss": 0.6026,
      "mean_token_accuracy": 0.8129221275448799,
      "num_tokens": 488090128.0,
      "step": 15293
    },
    {
      "entropy": 0.6277781678363681,
      "epoch": 1.4071281936274087,
      "grad_norm": 0.15277168154716492,
      "learning_rate": 5.309902781611311e-05,
      "loss": 0.6196,
      "mean_token_accuracy": 0.8112550303339958,
      "num_tokens": 488122475.0,
      "step": 15294
    },
    {
      "entropy": 0.6680510956794024,
      "epoch": 1.4072201999988498,
      "grad_norm": 0.1595848798751831,
      "learning_rate": 5.3095960989971483e-05,
      "loss": 0.6648,
      "mean_token_accuracy": 0.7959019765257835,
      "num_tokens": 488154306.0,
      "step": 15295
    },
    {
      "entropy": 0.5898949457332492,
      "epoch": 1.4073122063702912,
      "grad_norm": 0.1518932431936264,
      "learning_rate": 5.309289416382985e-05,
      "loss": 0.5762,
      "mean_token_accuracy": 0.8247500285506248,
      "num_tokens": 488185585.0,
      "step": 15296
    },
    {
      "entropy": 0.6874928968027234,
      "epoch": 1.4074042127417323,
      "grad_norm": 0.15734204649925232,
      "learning_rate": 5.3089827337688226e-05,
      "loss": 0.6785,
      "mean_token_accuracy": 0.792572844773531,
      "num_tokens": 488217805.0,
      "step": 15297
    },
    {
      "entropy": 0.6445957208052278,
      "epoch": 1.4074962191131735,
      "grad_norm": 0.15577290952205658,
      "learning_rate": 5.308676051154661e-05,
      "loss": 0.6327,
      "mean_token_accuracy": 0.8095278404653072,
      "num_tokens": 488250318.0,
      "step": 15298
    },
    {
      "entropy": 0.6178026394918561,
      "epoch": 1.4075882254846148,
      "grad_norm": 0.15668973326683044,
      "learning_rate": 5.3083693685404975e-05,
      "loss": 0.6068,
      "mean_token_accuracy": 0.8166442401707172,
      "num_tokens": 488282207.0,
      "step": 15299
    },
    {
      "entropy": 0.6447202805429697,
      "epoch": 1.407680231856056,
      "grad_norm": 0.1571938842535019,
      "learning_rate": 5.308062685926335e-05,
      "loss": 0.6368,
      "mean_token_accuracy": 0.8069269582629204,
      "num_tokens": 488313740.0,
      "step": 15300
    },
    {
      "entropy": 0.6182971373200417,
      "epoch": 1.4077722382274973,
      "grad_norm": 0.16172391176223755,
      "learning_rate": 5.307756003312172e-05,
      "loss": 0.6184,
      "mean_token_accuracy": 0.8100530281662941,
      "num_tokens": 488345269.0,
      "step": 15301
    },
    {
      "entropy": 0.5553032474126667,
      "epoch": 1.4078642445989384,
      "grad_norm": 0.15460146963596344,
      "learning_rate": 5.3074493206980106e-05,
      "loss": 0.555,
      "mean_token_accuracy": 0.8295189626514912,
      "num_tokens": 488377570.0,
      "step": 15302
    },
    {
      "entropy": 0.5026899934746325,
      "epoch": 1.4079562509703796,
      "grad_norm": 0.1420556604862213,
      "learning_rate": 5.3071426380838474e-05,
      "loss": 0.4803,
      "mean_token_accuracy": 0.8478940799832344,
      "num_tokens": 488408781.0,
      "step": 15303
    },
    {
      "entropy": 0.6519004860892892,
      "epoch": 1.408048257341821,
      "grad_norm": 0.15612943470478058,
      "learning_rate": 5.306835955469684e-05,
      "loss": 0.6461,
      "mean_token_accuracy": 0.8049517944455147,
      "num_tokens": 488440685.0,
      "step": 15304
    },
    {
      "entropy": 0.6073935232125223,
      "epoch": 1.408140263713262,
      "grad_norm": 0.16148874163627625,
      "learning_rate": 5.3065292728555216e-05,
      "loss": 0.5979,
      "mean_token_accuracy": 0.8160014748573303,
      "num_tokens": 488471631.0,
      "step": 15305
    },
    {
      "entropy": 0.6831851699389517,
      "epoch": 1.4082322700847034,
      "grad_norm": 0.16295728087425232,
      "learning_rate": 5.30622259024136e-05,
      "loss": 0.6741,
      "mean_token_accuracy": 0.7950636856257915,
      "num_tokens": 488503326.0,
      "step": 15306
    },
    {
      "entropy": 0.6265941401943564,
      "epoch": 1.4083242764561446,
      "grad_norm": 0.15983128547668457,
      "learning_rate": 5.305915907627197e-05,
      "loss": 0.6196,
      "mean_token_accuracy": 0.8087420910596848,
      "num_tokens": 488535943.0,
      "step": 15307
    },
    {
      "entropy": 0.621041995473206,
      "epoch": 1.4084162828275857,
      "grad_norm": 0.15993064641952515,
      "learning_rate": 5.305609225013034e-05,
      "loss": 0.6092,
      "mean_token_accuracy": 0.8139655590057373,
      "num_tokens": 488568711.0,
      "step": 15308
    },
    {
      "entropy": 0.6622148994356394,
      "epoch": 1.408508289199027,
      "grad_norm": 0.15588541328907013,
      "learning_rate": 5.305302542398871e-05,
      "loss": 0.6609,
      "mean_token_accuracy": 0.8024056404829025,
      "num_tokens": 488601243.0,
      "step": 15309
    },
    {
      "entropy": 0.5910425465553999,
      "epoch": 1.4086002955704682,
      "grad_norm": 0.15388070046901703,
      "learning_rate": 5.3049958597847096e-05,
      "loss": 0.5748,
      "mean_token_accuracy": 0.8236626982688904,
      "num_tokens": 488633463.0,
      "step": 15310
    },
    {
      "entropy": 0.6075641820207238,
      "epoch": 1.4086923019419095,
      "grad_norm": 0.15845094621181488,
      "learning_rate": 5.3046891771705464e-05,
      "loss": 0.6007,
      "mean_token_accuracy": 0.8149989023804665,
      "num_tokens": 488665476.0,
      "step": 15311
    },
    {
      "entropy": 0.5296744578517973,
      "epoch": 1.4087843083133507,
      "grad_norm": 0.15348409116268158,
      "learning_rate": 5.304382494556384e-05,
      "loss": 0.5096,
      "mean_token_accuracy": 0.8415365926921368,
      "num_tokens": 488697806.0,
      "step": 15312
    },
    {
      "entropy": 0.5861926898360252,
      "epoch": 1.4088763146847918,
      "grad_norm": 0.1525213122367859,
      "learning_rate": 5.304075811942221e-05,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8211039006710052,
      "num_tokens": 488729903.0,
      "step": 15313
    },
    {
      "entropy": 0.6535147521644831,
      "epoch": 1.4089683210562332,
      "grad_norm": 0.1632235050201416,
      "learning_rate": 5.303769129328059e-05,
      "loss": 0.6456,
      "mean_token_accuracy": 0.802250012755394,
      "num_tokens": 488762488.0,
      "step": 15314
    },
    {
      "entropy": 0.5723133534193039,
      "epoch": 1.4090603274276743,
      "grad_norm": 0.1633490025997162,
      "learning_rate": 5.303462446713896e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8226849660277367,
      "num_tokens": 488794248.0,
      "step": 15315
    },
    {
      "entropy": 0.6262085177004337,
      "epoch": 1.4091523337991156,
      "grad_norm": 0.1625182330608368,
      "learning_rate": 5.303155764099733e-05,
      "loss": 0.6062,
      "mean_token_accuracy": 0.8163077794015408,
      "num_tokens": 488826484.0,
      "step": 15316
    },
    {
      "entropy": 0.601121200248599,
      "epoch": 1.4092443401705568,
      "grad_norm": 0.1622476577758789,
      "learning_rate": 5.302849081485571e-05,
      "loss": 0.5928,
      "mean_token_accuracy": 0.8172049894928932,
      "num_tokens": 488858658.0,
      "step": 15317
    },
    {
      "entropy": 0.6480224295519292,
      "epoch": 1.409336346541998,
      "grad_norm": 0.1519586741924286,
      "learning_rate": 5.302542398871409e-05,
      "loss": 0.6371,
      "mean_token_accuracy": 0.8072790205478668,
      "num_tokens": 488890999.0,
      "step": 15318
    },
    {
      "entropy": 0.609926288947463,
      "epoch": 1.4094283529134393,
      "grad_norm": 0.15716920793056488,
      "learning_rate": 5.3022357162572455e-05,
      "loss": 0.595,
      "mean_token_accuracy": 0.8200767710804939,
      "num_tokens": 488922545.0,
      "step": 15319
    },
    {
      "entropy": 0.699685275554657,
      "epoch": 1.4095203592848804,
      "grad_norm": 0.16404686868190765,
      "learning_rate": 5.301929033643083e-05,
      "loss": 0.6834,
      "mean_token_accuracy": 0.7929071001708508,
      "num_tokens": 488954124.0,
      "step": 15320
    },
    {
      "entropy": 0.6062635518610477,
      "epoch": 1.4096123656563218,
      "grad_norm": 0.15761254727840424,
      "learning_rate": 5.301622351028921e-05,
      "loss": 0.6016,
      "mean_token_accuracy": 0.8176194056868553,
      "num_tokens": 488985480.0,
      "step": 15321
    },
    {
      "entropy": 0.5913577154278755,
      "epoch": 1.4097043720277629,
      "grad_norm": 0.14988066256046295,
      "learning_rate": 5.301315668414758e-05,
      "loss": 0.5827,
      "mean_token_accuracy": 0.8189326785504818,
      "num_tokens": 489017505.0,
      "step": 15322
    },
    {
      "entropy": 0.6806764854118228,
      "epoch": 1.409796378399204,
      "grad_norm": 0.16075032949447632,
      "learning_rate": 5.301008985800595e-05,
      "loss": 0.6723,
      "mean_token_accuracy": 0.7979416027665138,
      "num_tokens": 489049207.0,
      "step": 15323
    },
    {
      "entropy": 0.6870513432659209,
      "epoch": 1.4098883847706454,
      "grad_norm": 0.16593563556671143,
      "learning_rate": 5.300702303186432e-05,
      "loss": 0.687,
      "mean_token_accuracy": 0.7888948321342468,
      "num_tokens": 489079910.0,
      "step": 15324
    },
    {
      "entropy": 0.6955335829406977,
      "epoch": 1.4099803911420865,
      "grad_norm": 0.16737547516822815,
      "learning_rate": 5.30039562057227e-05,
      "loss": 0.6809,
      "mean_token_accuracy": 0.7884808778762817,
      "num_tokens": 489110351.0,
      "step": 15325
    },
    {
      "entropy": 0.6076020412147045,
      "epoch": 1.4100723975135279,
      "grad_norm": 0.1563960760831833,
      "learning_rate": 5.300088937958108e-05,
      "loss": 0.6021,
      "mean_token_accuracy": 0.8112610429525375,
      "num_tokens": 489142221.0,
      "step": 15326
    },
    {
      "entropy": 0.5997037943452597,
      "epoch": 1.410164403884969,
      "grad_norm": 0.15424318611621857,
      "learning_rate": 5.2997822553439445e-05,
      "loss": 0.586,
      "mean_token_accuracy": 0.8176491633057594,
      "num_tokens": 489173867.0,
      "step": 15327
    },
    {
      "entropy": 0.5974665060639381,
      "epoch": 1.4102564102564101,
      "grad_norm": 0.15849053859710693,
      "learning_rate": 5.299475572729782e-05,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8197358176112175,
      "num_tokens": 489205784.0,
      "step": 15328
    },
    {
      "entropy": 0.5676660221070051,
      "epoch": 1.4103484166278515,
      "grad_norm": 0.15392442047595978,
      "learning_rate": 5.29916889011562e-05,
      "loss": 0.5496,
      "mean_token_accuracy": 0.8269581906497478,
      "num_tokens": 489238027.0,
      "step": 15329
    },
    {
      "entropy": 0.7298863278701901,
      "epoch": 1.4104404229992926,
      "grad_norm": 0.16154582798480988,
      "learning_rate": 5.298862207501457e-05,
      "loss": 0.7133,
      "mean_token_accuracy": 0.785091008991003,
      "num_tokens": 489270266.0,
      "step": 15330
    },
    {
      "entropy": 0.6711881086230278,
      "epoch": 1.410532429370734,
      "grad_norm": 0.158863827586174,
      "learning_rate": 5.2985555248872944e-05,
      "loss": 0.6531,
      "mean_token_accuracy": 0.8038175851106644,
      "num_tokens": 489302452.0,
      "step": 15331
    },
    {
      "entropy": 0.6507463809102774,
      "epoch": 1.410624435742175,
      "grad_norm": 0.163125678896904,
      "learning_rate": 5.298248842273131e-05,
      "loss": 0.641,
      "mean_token_accuracy": 0.8041867539286613,
      "num_tokens": 489334437.0,
      "step": 15332
    },
    {
      "entropy": 0.6823274083435535,
      "epoch": 1.4107164421136162,
      "grad_norm": 0.16595692932605743,
      "learning_rate": 5.297942159658969e-05,
      "loss": 0.6663,
      "mean_token_accuracy": 0.7955467291176319,
      "num_tokens": 489365699.0,
      "step": 15333
    },
    {
      "entropy": 0.555912759155035,
      "epoch": 1.4108084484850576,
      "grad_norm": 0.15535421669483185,
      "learning_rate": 5.297635477044807e-05,
      "loss": 0.5468,
      "mean_token_accuracy": 0.8342216573655605,
      "num_tokens": 489398073.0,
      "step": 15334
    },
    {
      "entropy": 0.6570857772603631,
      "epoch": 1.4109004548564987,
      "grad_norm": 0.15709732472896576,
      "learning_rate": 5.2973287944306435e-05,
      "loss": 0.6421,
      "mean_token_accuracy": 0.8031359799206257,
      "num_tokens": 489429985.0,
      "step": 15335
    },
    {
      "entropy": 0.6242386549711227,
      "epoch": 1.41099246122794,
      "grad_norm": 0.16070076823234558,
      "learning_rate": 5.297022111816481e-05,
      "loss": 0.6151,
      "mean_token_accuracy": 0.8126376494765282,
      "num_tokens": 489461797.0,
      "step": 15336
    },
    {
      "entropy": 0.6306798774749041,
      "epoch": 1.4110844675993812,
      "grad_norm": 0.16001084446907043,
      "learning_rate": 5.296715429202319e-05,
      "loss": 0.6208,
      "mean_token_accuracy": 0.8085699081420898,
      "num_tokens": 489492627.0,
      "step": 15337
    },
    {
      "entropy": 0.6180215030908585,
      "epoch": 1.4111764739708224,
      "grad_norm": 0.15011432766914368,
      "learning_rate": 5.296408746588156e-05,
      "loss": 0.6056,
      "mean_token_accuracy": 0.8136832006275654,
      "num_tokens": 489524952.0,
      "step": 15338
    },
    {
      "entropy": 0.6698107505217195,
      "epoch": 1.4112684803422637,
      "grad_norm": 0.15524789690971375,
      "learning_rate": 5.2961020639739934e-05,
      "loss": 0.6542,
      "mean_token_accuracy": 0.7996007911860943,
      "num_tokens": 489557061.0,
      "step": 15339
    },
    {
      "entropy": 0.5444951658137143,
      "epoch": 1.4113604867137048,
      "grad_norm": 0.153168186545372,
      "learning_rate": 5.29579538135983e-05,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8387733809649944,
      "num_tokens": 489589564.0,
      "step": 15340
    },
    {
      "entropy": 0.6859938129782677,
      "epoch": 1.4114524930851462,
      "grad_norm": 0.1675758957862854,
      "learning_rate": 5.295488698745668e-05,
      "loss": 0.6714,
      "mean_token_accuracy": 0.7942909970879555,
      "num_tokens": 489621831.0,
      "step": 15341
    },
    {
      "entropy": 0.5286345602944493,
      "epoch": 1.4115444994565873,
      "grad_norm": 0.1444603055715561,
      "learning_rate": 5.295182016131506e-05,
      "loss": 0.5168,
      "mean_token_accuracy": 0.8399224653840065,
      "num_tokens": 489654079.0,
      "step": 15342
    },
    {
      "entropy": 0.6350334053859115,
      "epoch": 1.4116365058280285,
      "grad_norm": 0.15901495516300201,
      "learning_rate": 5.2948753335173426e-05,
      "loss": 0.6264,
      "mean_token_accuracy": 0.8071246109902859,
      "num_tokens": 489686379.0,
      "step": 15343
    },
    {
      "entropy": 0.5742808356881142,
      "epoch": 1.4117285121994698,
      "grad_norm": 0.15422335267066956,
      "learning_rate": 5.2945686509031814e-05,
      "loss": 0.5662,
      "mean_token_accuracy": 0.8268059678375721,
      "num_tokens": 489718118.0,
      "step": 15344
    },
    {
      "entropy": 0.5542657375335693,
      "epoch": 1.411820518570911,
      "grad_norm": 0.15431435406208038,
      "learning_rate": 5.294261968289018e-05,
      "loss": 0.5312,
      "mean_token_accuracy": 0.8338841386139393,
      "num_tokens": 489749813.0,
      "step": 15345
    },
    {
      "entropy": 0.5465168668888509,
      "epoch": 1.4119125249423523,
      "grad_norm": 0.15230534970760345,
      "learning_rate": 5.2939552856748557e-05,
      "loss": 0.5362,
      "mean_token_accuracy": 0.83586236089468,
      "num_tokens": 489782170.0,
      "step": 15346
    },
    {
      "entropy": 0.5897638257592916,
      "epoch": 1.4120045313137934,
      "grad_norm": 0.15708941221237183,
      "learning_rate": 5.2936486030606924e-05,
      "loss": 0.586,
      "mean_token_accuracy": 0.8166618831455708,
      "num_tokens": 489813969.0,
      "step": 15347
    },
    {
      "entropy": 0.6051710322499275,
      "epoch": 1.4120965376852346,
      "grad_norm": 0.1534903198480606,
      "learning_rate": 5.2933419204465306e-05,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8209854885935783,
      "num_tokens": 489845790.0,
      "step": 15348
    },
    {
      "entropy": 0.6455524563789368,
      "epoch": 1.412188544056676,
      "grad_norm": 0.150975301861763,
      "learning_rate": 5.293035237832368e-05,
      "loss": 0.6422,
      "mean_token_accuracy": 0.8029180951416492,
      "num_tokens": 489878554.0,
      "step": 15349
    },
    {
      "entropy": 0.6750368038192391,
      "epoch": 1.412280550428117,
      "grad_norm": 0.15725605189800262,
      "learning_rate": 5.292728555218205e-05,
      "loss": 0.66,
      "mean_token_accuracy": 0.800733856856823,
      "num_tokens": 489910316.0,
      "step": 15350
    },
    {
      "entropy": 0.721755538135767,
      "epoch": 1.4123725567995584,
      "grad_norm": 0.16148732602596283,
      "learning_rate": 5.292421872604042e-05,
      "loss": 0.7034,
      "mean_token_accuracy": 0.7869605123996735,
      "num_tokens": 489942032.0,
      "step": 15351
    },
    {
      "entropy": 0.7731480542570353,
      "epoch": 1.4124645631709996,
      "grad_norm": 0.16494271159172058,
      "learning_rate": 5.2921151899898804e-05,
      "loss": 0.7561,
      "mean_token_accuracy": 0.7749280147254467,
      "num_tokens": 489974026.0,
      "step": 15352
    },
    {
      "entropy": 0.6586611047387123,
      "epoch": 1.4125565695424407,
      "grad_norm": 0.16115830838680267,
      "learning_rate": 5.291808507375717e-05,
      "loss": 0.6435,
      "mean_token_accuracy": 0.8037238605320454,
      "num_tokens": 490006042.0,
      "step": 15353
    },
    {
      "entropy": 0.6122302617877722,
      "epoch": 1.412648575913882,
      "grad_norm": 0.15708020329475403,
      "learning_rate": 5.291501824761555e-05,
      "loss": 0.6101,
      "mean_token_accuracy": 0.811190627515316,
      "num_tokens": 490038205.0,
      "step": 15354
    },
    {
      "entropy": 0.5569055285304785,
      "epoch": 1.4127405822853232,
      "grad_norm": 0.15231379866600037,
      "learning_rate": 5.2911951421473915e-05,
      "loss": 0.5416,
      "mean_token_accuracy": 0.8357484191656113,
      "num_tokens": 490070718.0,
      "step": 15355
    },
    {
      "entropy": 0.6188960820436478,
      "epoch": 1.4128325886567645,
      "grad_norm": 0.15184922516345978,
      "learning_rate": 5.2908884595332296e-05,
      "loss": 0.6101,
      "mean_token_accuracy": 0.8142894133925438,
      "num_tokens": 490103130.0,
      "step": 15356
    },
    {
      "entropy": 0.6196503462269902,
      "epoch": 1.4129245950282057,
      "grad_norm": 0.15573057532310486,
      "learning_rate": 5.290581776919067e-05,
      "loss": 0.6149,
      "mean_token_accuracy": 0.812189344316721,
      "num_tokens": 490135407.0,
      "step": 15357
    },
    {
      "entropy": 0.6315686041489244,
      "epoch": 1.4130166013996468,
      "grad_norm": 0.15835842490196228,
      "learning_rate": 5.290275094304904e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.8122154213488102,
      "num_tokens": 490167441.0,
      "step": 15358
    },
    {
      "entropy": 0.6289777681231499,
      "epoch": 1.4131086077710882,
      "grad_norm": 0.15899847447872162,
      "learning_rate": 5.2899684116907413e-05,
      "loss": 0.6118,
      "mean_token_accuracy": 0.8113442994654179,
      "num_tokens": 490198625.0,
      "step": 15359
    },
    {
      "entropy": 0.7057408522814512,
      "epoch": 1.4132006141425293,
      "grad_norm": 0.16440710425376892,
      "learning_rate": 5.2896617290765795e-05,
      "loss": 0.707,
      "mean_token_accuracy": 0.7894854694604874,
      "num_tokens": 490231110.0,
      "step": 15360
    },
    {
      "entropy": 0.5788752799853683,
      "epoch": 1.4132926205139706,
      "grad_norm": 0.16423757374286652,
      "learning_rate": 5.289355046462416e-05,
      "loss": 0.5631,
      "mean_token_accuracy": 0.8263331390917301,
      "num_tokens": 490262421.0,
      "step": 15361
    },
    {
      "entropy": 0.6621713563799858,
      "epoch": 1.4133846268854118,
      "grad_norm": 0.15367774665355682,
      "learning_rate": 5.289048363848254e-05,
      "loss": 0.6439,
      "mean_token_accuracy": 0.8016893267631531,
      "num_tokens": 490294798.0,
      "step": 15362
    },
    {
      "entropy": 0.6316179595887661,
      "epoch": 1.413476633256853,
      "grad_norm": 0.15551622211933136,
      "learning_rate": 5.2887416812340905e-05,
      "loss": 0.6268,
      "mean_token_accuracy": 0.8114778324961662,
      "num_tokens": 490327034.0,
      "step": 15363
    },
    {
      "entropy": 0.5332371620461345,
      "epoch": 1.4135686396282943,
      "grad_norm": 0.15245792269706726,
      "learning_rate": 5.288434998619929e-05,
      "loss": 0.5185,
      "mean_token_accuracy": 0.8360911644995213,
      "num_tokens": 490358983.0,
      "step": 15364
    },
    {
      "entropy": 0.5637179538607597,
      "epoch": 1.4136606459997354,
      "grad_norm": 0.15103159844875336,
      "learning_rate": 5.288128316005766e-05,
      "loss": 0.5512,
      "mean_token_accuracy": 0.8279215209186077,
      "num_tokens": 490391127.0,
      "step": 15365
    },
    {
      "entropy": 0.5855046808719635,
      "epoch": 1.4137526523711768,
      "grad_norm": 0.1514948159456253,
      "learning_rate": 5.287821633391603e-05,
      "loss": 0.5695,
      "mean_token_accuracy": 0.8239837363362312,
      "num_tokens": 490422910.0,
      "step": 15366
    },
    {
      "entropy": 0.6207795944064856,
      "epoch": 1.4138446587426179,
      "grad_norm": 0.1634671837091446,
      "learning_rate": 5.2875149507774404e-05,
      "loss": 0.599,
      "mean_token_accuracy": 0.8158352002501488,
      "num_tokens": 490454317.0,
      "step": 15367
    },
    {
      "entropy": 0.5639979559928179,
      "epoch": 1.413936665114059,
      "grad_norm": 0.15518425405025482,
      "learning_rate": 5.2872082681632785e-05,
      "loss": 0.5598,
      "mean_token_accuracy": 0.8295666947960854,
      "num_tokens": 490486579.0,
      "step": 15368
    },
    {
      "entropy": 0.6093049459159374,
      "epoch": 1.4140286714855004,
      "grad_norm": 0.16170631349086761,
      "learning_rate": 5.286901585549115e-05,
      "loss": 0.6078,
      "mean_token_accuracy": 0.8123147077858448,
      "num_tokens": 490518258.0,
      "step": 15369
    },
    {
      "entropy": 0.6983311893418431,
      "epoch": 1.4141206778569415,
      "grad_norm": 0.1724918782711029,
      "learning_rate": 5.286594902934953e-05,
      "loss": 0.7119,
      "mean_token_accuracy": 0.7860892228782177,
      "num_tokens": 490550959.0,
      "step": 15370
    },
    {
      "entropy": 0.6538557857275009,
      "epoch": 1.4142126842283829,
      "grad_norm": 0.16840046644210815,
      "learning_rate": 5.2862882203207896e-05,
      "loss": 0.6411,
      "mean_token_accuracy": 0.8009047582745552,
      "num_tokens": 490583362.0,
      "step": 15371
    },
    {
      "entropy": 0.6034431159496307,
      "epoch": 1.414304690599824,
      "grad_norm": 0.15946784615516663,
      "learning_rate": 5.285981537706628e-05,
      "loss": 0.5933,
      "mean_token_accuracy": 0.8178427517414093,
      "num_tokens": 490615381.0,
      "step": 15372
    },
    {
      "entropy": 0.6762716826051474,
      "epoch": 1.4143966969712651,
      "grad_norm": 0.1726544052362442,
      "learning_rate": 5.285674855092465e-05,
      "loss": 0.6769,
      "mean_token_accuracy": 0.7942431345582008,
      "num_tokens": 490647707.0,
      "step": 15373
    },
    {
      "entropy": 0.6479336377233267,
      "epoch": 1.4144887033427065,
      "grad_norm": 0.15547235310077667,
      "learning_rate": 5.285368172478302e-05,
      "loss": 0.635,
      "mean_token_accuracy": 0.806144967675209,
      "num_tokens": 490679215.0,
      "step": 15374
    },
    {
      "entropy": 0.7301639914512634,
      "epoch": 1.4145807097141476,
      "grad_norm": 0.1626797765493393,
      "learning_rate": 5.28506148986414e-05,
      "loss": 0.7315,
      "mean_token_accuracy": 0.7793500758707523,
      "num_tokens": 490710885.0,
      "step": 15375
    },
    {
      "entropy": 0.6598696717992425,
      "epoch": 1.414672716085589,
      "grad_norm": 0.163276806473732,
      "learning_rate": 5.2847548072499776e-05,
      "loss": 0.6446,
      "mean_token_accuracy": 0.8047741614282131,
      "num_tokens": 490742475.0,
      "step": 15376
    },
    {
      "entropy": 0.5603220975026488,
      "epoch": 1.41476472245703,
      "grad_norm": 0.1453525573015213,
      "learning_rate": 5.2844481246358144e-05,
      "loss": 0.5473,
      "mean_token_accuracy": 0.8293982371687889,
      "num_tokens": 490775101.0,
      "step": 15377
    },
    {
      "entropy": 0.684409886598587,
      "epoch": 1.4148567288284712,
      "grad_norm": 0.1616021990776062,
      "learning_rate": 5.284141442021652e-05,
      "loss": 0.6649,
      "mean_token_accuracy": 0.7976300977170467,
      "num_tokens": 490806865.0,
      "step": 15378
    },
    {
      "entropy": 0.5752520803362131,
      "epoch": 1.4149487351999126,
      "grad_norm": 0.15168112516403198,
      "learning_rate": 5.28383475940749e-05,
      "loss": 0.5635,
      "mean_token_accuracy": 0.8273430056869984,
      "num_tokens": 490839165.0,
      "step": 15379
    },
    {
      "entropy": 0.5964994048699737,
      "epoch": 1.415040741571354,
      "grad_norm": 0.14845909178256989,
      "learning_rate": 5.283528076793327e-05,
      "loss": 0.5742,
      "mean_token_accuracy": 0.8203756399452686,
      "num_tokens": 490871052.0,
      "step": 15380
    },
    {
      "entropy": 0.6235447432845831,
      "epoch": 1.415132747942795,
      "grad_norm": 0.15624257922172546,
      "learning_rate": 5.283221394179164e-05,
      "loss": 0.6035,
      "mean_token_accuracy": 0.8134434334933758,
      "num_tokens": 490903250.0,
      "step": 15381
    },
    {
      "entropy": 0.6272587943822145,
      "epoch": 1.4152247543142362,
      "grad_norm": 0.1604537069797516,
      "learning_rate": 5.282914711565001e-05,
      "loss": 0.6167,
      "mean_token_accuracy": 0.8103026151657104,
      "num_tokens": 490935163.0,
      "step": 15382
    },
    {
      "entropy": 0.722938634455204,
      "epoch": 1.4153167606856774,
      "grad_norm": 0.1613156944513321,
      "learning_rate": 5.28260802895084e-05,
      "loss": 0.722,
      "mean_token_accuracy": 0.7838884368538857,
      "num_tokens": 490967366.0,
      "step": 15383
    },
    {
      "entropy": 0.6140503715723753,
      "epoch": 1.4154087670571187,
      "grad_norm": 0.1542789787054062,
      "learning_rate": 5.2823013463366766e-05,
      "loss": 0.6014,
      "mean_token_accuracy": 0.8156932629644871,
      "num_tokens": 490999594.0,
      "step": 15384
    },
    {
      "entropy": 0.60285066626966,
      "epoch": 1.41550077342856,
      "grad_norm": 0.16850167512893677,
      "learning_rate": 5.2819946637225134e-05,
      "loss": 0.6048,
      "mean_token_accuracy": 0.8166209198534489,
      "num_tokens": 491031630.0,
      "step": 15385
    },
    {
      "entropy": 0.6961261089891195,
      "epoch": 1.4155927798000012,
      "grad_norm": 0.15897786617279053,
      "learning_rate": 5.281687981108351e-05,
      "loss": 0.6935,
      "mean_token_accuracy": 0.7873920649290085,
      "num_tokens": 491063815.0,
      "step": 15386
    },
    {
      "entropy": 0.5035765590146184,
      "epoch": 1.4156847861714423,
      "grad_norm": 0.14493784308433533,
      "learning_rate": 5.281381298494189e-05,
      "loss": 0.4883,
      "mean_token_accuracy": 0.8499904498457909,
      "num_tokens": 491096232.0,
      "step": 15387
    },
    {
      "entropy": 0.6789825651794672,
      "epoch": 1.4157767925428835,
      "grad_norm": 0.1622830629348755,
      "learning_rate": 5.2810746158800265e-05,
      "loss": 0.6677,
      "mean_token_accuracy": 0.7941774874925613,
      "num_tokens": 491128243.0,
      "step": 15388
    },
    {
      "entropy": 0.5712355617433786,
      "epoch": 1.4158687989143248,
      "grad_norm": 0.14671632647514343,
      "learning_rate": 5.280767933265863e-05,
      "loss": 0.5654,
      "mean_token_accuracy": 0.8275863640010357,
      "num_tokens": 491160918.0,
      "step": 15389
    },
    {
      "entropy": 0.6189969349652529,
      "epoch": 1.4159608052857662,
      "grad_norm": 0.16060248017311096,
      "learning_rate": 5.280461250651701e-05,
      "loss": 0.6077,
      "mean_token_accuracy": 0.814625009894371,
      "num_tokens": 491193310.0,
      "step": 15390
    },
    {
      "entropy": 0.7265803776681423,
      "epoch": 1.4160528116572073,
      "grad_norm": 0.16709662973880768,
      "learning_rate": 5.280154568037539e-05,
      "loss": 0.7303,
      "mean_token_accuracy": 0.781378623098135,
      "num_tokens": 491225424.0,
      "step": 15391
    },
    {
      "entropy": 0.6109869107604027,
      "epoch": 1.4161448180286484,
      "grad_norm": 0.16356788575649261,
      "learning_rate": 5.2798478854233757e-05,
      "loss": 0.6041,
      "mean_token_accuracy": 0.8140734545886517,
      "num_tokens": 491257778.0,
      "step": 15392
    },
    {
      "entropy": 0.6439702664501965,
      "epoch": 1.4162368244000896,
      "grad_norm": 0.15353479981422424,
      "learning_rate": 5.279541202809213e-05,
      "loss": 0.6313,
      "mean_token_accuracy": 0.8101891092956066,
      "num_tokens": 491289698.0,
      "step": 15393
    },
    {
      "entropy": 0.7032667230814695,
      "epoch": 1.416328830771531,
      "grad_norm": 0.1656343787908554,
      "learning_rate": 5.27923452019505e-05,
      "loss": 0.6991,
      "mean_token_accuracy": 0.78865185379982,
      "num_tokens": 491321366.0,
      "step": 15394
    },
    {
      "entropy": 0.6220672614872456,
      "epoch": 1.4164208371429723,
      "grad_norm": 0.1612556129693985,
      "learning_rate": 5.278927837580888e-05,
      "loss": 0.6004,
      "mean_token_accuracy": 0.8189092837274075,
      "num_tokens": 491353623.0,
      "step": 15395
    },
    {
      "entropy": 0.6414111875928938,
      "epoch": 1.4165128435144134,
      "grad_norm": 0.15375208854675293,
      "learning_rate": 5.2786211549667255e-05,
      "loss": 0.6324,
      "mean_token_accuracy": 0.807960107922554,
      "num_tokens": 491386058.0,
      "step": 15396
    },
    {
      "entropy": 0.6554678222164512,
      "epoch": 1.4166048498858546,
      "grad_norm": 0.1594211906194687,
      "learning_rate": 5.278314472352562e-05,
      "loss": 0.6491,
      "mean_token_accuracy": 0.7984637916088104,
      "num_tokens": 491417511.0,
      "step": 15397
    },
    {
      "entropy": 0.5460104332305491,
      "epoch": 1.4166968562572957,
      "grad_norm": 0.15424196422100067,
      "learning_rate": 5.2780077897384e-05,
      "loss": 0.5312,
      "mean_token_accuracy": 0.8381285667419434,
      "num_tokens": 491449328.0,
      "step": 15398
    },
    {
      "entropy": 0.716932462528348,
      "epoch": 1.416788862628737,
      "grad_norm": 0.16827182471752167,
      "learning_rate": 5.277701107124238e-05,
      "loss": 0.7196,
      "mean_token_accuracy": 0.7831595465540886,
      "num_tokens": 491481105.0,
      "step": 15399
    },
    {
      "entropy": 0.6078282725065947,
      "epoch": 1.4168808690001784,
      "grad_norm": 0.1562427133321762,
      "learning_rate": 5.277394424510075e-05,
      "loss": 0.5983,
      "mean_token_accuracy": 0.8159366883337498,
      "num_tokens": 491513139.0,
      "step": 15400
    },
    {
      "entropy": 0.6720555913634598,
      "epoch": 1.4169728753716195,
      "grad_norm": 0.1541070193052292,
      "learning_rate": 5.277087741895912e-05,
      "loss": 0.6478,
      "mean_token_accuracy": 0.8057103753089905,
      "num_tokens": 491545618.0,
      "step": 15401
    },
    {
      "entropy": 0.5919599728658795,
      "epoch": 1.4170648817430607,
      "grad_norm": 0.15529850125312805,
      "learning_rate": 5.276781059281749e-05,
      "loss": 0.5892,
      "mean_token_accuracy": 0.8204655759036541,
      "num_tokens": 491578194.0,
      "step": 15402
    },
    {
      "entropy": 0.6191137852147222,
      "epoch": 1.4171568881145018,
      "grad_norm": 0.15694795548915863,
      "learning_rate": 5.276474376667587e-05,
      "loss": 0.6079,
      "mean_token_accuracy": 0.8148420415818691,
      "num_tokens": 491610647.0,
      "step": 15403
    },
    {
      "entropy": 0.5277523119002581,
      "epoch": 1.4172488944859432,
      "grad_norm": 0.1465207189321518,
      "learning_rate": 5.2761676940534246e-05,
      "loss": 0.5005,
      "mean_token_accuracy": 0.8476972430944443,
      "num_tokens": 491642191.0,
      "step": 15404
    },
    {
      "entropy": 0.5771569702774286,
      "epoch": 1.4173409008573845,
      "grad_norm": 0.1512133926153183,
      "learning_rate": 5.275861011439261e-05,
      "loss": 0.5574,
      "mean_token_accuracy": 0.8269226625561714,
      "num_tokens": 491674034.0,
      "step": 15405
    },
    {
      "entropy": 0.6940582152456045,
      "epoch": 1.4174329072288256,
      "grad_norm": 0.160121351480484,
      "learning_rate": 5.2755543288250995e-05,
      "loss": 0.6844,
      "mean_token_accuracy": 0.7883287891745567,
      "num_tokens": 491706167.0,
      "step": 15406
    },
    {
      "entropy": 0.5947671169415116,
      "epoch": 1.4175249136002668,
      "grad_norm": 0.15885616838932037,
      "learning_rate": 5.275247646210937e-05,
      "loss": 0.5927,
      "mean_token_accuracy": 0.8186366930603981,
      "num_tokens": 491738734.0,
      "step": 15407
    },
    {
      "entropy": 0.5641571991145611,
      "epoch": 1.417616919971708,
      "grad_norm": 0.15183289349079132,
      "learning_rate": 5.274940963596774e-05,
      "loss": 0.5537,
      "mean_token_accuracy": 0.831402450799942,
      "num_tokens": 491770259.0,
      "step": 15408
    },
    {
      "entropy": 0.5207720836624503,
      "epoch": 1.4177089263431493,
      "grad_norm": 0.14777326583862305,
      "learning_rate": 5.274634280982611e-05,
      "loss": 0.5135,
      "mean_token_accuracy": 0.8452372662723064,
      "num_tokens": 491802880.0,
      "step": 15409
    },
    {
      "entropy": 0.6689429124817252,
      "epoch": 1.4178009327145906,
      "grad_norm": 0.15641269087791443,
      "learning_rate": 5.2743275983684493e-05,
      "loss": 0.6472,
      "mean_token_accuracy": 0.8076067045331001,
      "num_tokens": 491834772.0,
      "step": 15410
    },
    {
      "entropy": 0.7459311205893755,
      "epoch": 1.4178929390860318,
      "grad_norm": 0.16639746725559235,
      "learning_rate": 5.274020915754286e-05,
      "loss": 0.7412,
      "mean_token_accuracy": 0.7801926322281361,
      "num_tokens": 491867167.0,
      "step": 15411
    },
    {
      "entropy": 0.6830470683053136,
      "epoch": 1.417984945457473,
      "grad_norm": 0.16392222046852112,
      "learning_rate": 5.2737142331401236e-05,
      "loss": 0.6797,
      "mean_token_accuracy": 0.7935988306999207,
      "num_tokens": 491899471.0,
      "step": 15412
    },
    {
      "entropy": 0.5613020358141512,
      "epoch": 1.418076951828914,
      "grad_norm": 0.15851061046123505,
      "learning_rate": 5.2734075505259604e-05,
      "loss": 0.5455,
      "mean_token_accuracy": 0.8315790072083473,
      "num_tokens": 491930721.0,
      "step": 15413
    },
    {
      "entropy": 0.5556157864630222,
      "epoch": 1.4181689582003554,
      "grad_norm": 0.15512055158615112,
      "learning_rate": 5.2731008679117985e-05,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8317376039922237,
      "num_tokens": 491962554.0,
      "step": 15414
    },
    {
      "entropy": 0.7100781686604023,
      "epoch": 1.4182609645717967,
      "grad_norm": 0.16485324501991272,
      "learning_rate": 5.272794185297636e-05,
      "loss": 0.7013,
      "mean_token_accuracy": 0.7864921055734158,
      "num_tokens": 491994908.0,
      "step": 15415
    },
    {
      "entropy": 0.7297190260142088,
      "epoch": 1.4183529709432379,
      "grad_norm": 0.16101129353046417,
      "learning_rate": 5.272487502683473e-05,
      "loss": 0.7258,
      "mean_token_accuracy": 0.7794414423406124,
      "num_tokens": 492026684.0,
      "step": 15416
    },
    {
      "entropy": 0.6331856492906809,
      "epoch": 1.418444977314679,
      "grad_norm": 0.15615709125995636,
      "learning_rate": 5.27218082006931e-05,
      "loss": 0.62,
      "mean_token_accuracy": 0.8088640905916691,
      "num_tokens": 492059041.0,
      "step": 15417
    },
    {
      "entropy": 0.6601867210119963,
      "epoch": 1.4185369836861201,
      "grad_norm": 0.15802103281021118,
      "learning_rate": 5.2718741374551484e-05,
      "loss": 0.6551,
      "mean_token_accuracy": 0.7977999337017536,
      "num_tokens": 492091108.0,
      "step": 15418
    },
    {
      "entropy": 0.5997693911194801,
      "epoch": 1.4186289900575615,
      "grad_norm": 0.15536008775234222,
      "learning_rate": 5.271567454840985e-05,
      "loss": 0.5921,
      "mean_token_accuracy": 0.8171800039708614,
      "num_tokens": 492123096.0,
      "step": 15419
    },
    {
      "entropy": 0.5429612486623228,
      "epoch": 1.4187209964290028,
      "grad_norm": 0.14652982354164124,
      "learning_rate": 5.2712607722268226e-05,
      "loss": 0.5427,
      "mean_token_accuracy": 0.8323616795241833,
      "num_tokens": 492155786.0,
      "step": 15420
    },
    {
      "entropy": 0.6155583527870476,
      "epoch": 1.418813002800444,
      "grad_norm": 0.15846028923988342,
      "learning_rate": 5.2709540896126594e-05,
      "loss": 0.603,
      "mean_token_accuracy": 0.8177987486124039,
      "num_tokens": 492187830.0,
      "step": 15421
    },
    {
      "entropy": 0.5501620471477509,
      "epoch": 1.4189050091718851,
      "grad_norm": 0.1529226452112198,
      "learning_rate": 5.270647406998498e-05,
      "loss": 0.5388,
      "mean_token_accuracy": 0.8315199539065361,
      "num_tokens": 492219823.0,
      "step": 15422
    },
    {
      "entropy": 0.6288833515718579,
      "epoch": 1.4189970155433262,
      "grad_norm": 0.1588977873325348,
      "learning_rate": 5.270340724384335e-05,
      "loss": 0.622,
      "mean_token_accuracy": 0.807062666863203,
      "num_tokens": 492251797.0,
      "step": 15423
    },
    {
      "entropy": 0.5172520484775305,
      "epoch": 1.4190890219147676,
      "grad_norm": 0.15158918499946594,
      "learning_rate": 5.270034041770172e-05,
      "loss": 0.5139,
      "mean_token_accuracy": 0.8436535373330116,
      "num_tokens": 492283258.0,
      "step": 15424
    },
    {
      "entropy": 0.604634060524404,
      "epoch": 1.419181028286209,
      "grad_norm": 0.1536514014005661,
      "learning_rate": 5.269727359156009e-05,
      "loss": 0.5965,
      "mean_token_accuracy": 0.8180724121630192,
      "num_tokens": 492314711.0,
      "step": 15425
    },
    {
      "entropy": 0.6216080281883478,
      "epoch": 1.41927303465765,
      "grad_norm": 0.15262900292873383,
      "learning_rate": 5.2694206765418474e-05,
      "loss": 0.6152,
      "mean_token_accuracy": 0.8117267563939095,
      "num_tokens": 492346483.0,
      "step": 15426
    },
    {
      "entropy": 0.5105062294751406,
      "epoch": 1.4193650410290912,
      "grad_norm": 0.1462351679801941,
      "learning_rate": 5.269113993927685e-05,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8482773266732693,
      "num_tokens": 492378036.0,
      "step": 15427
    },
    {
      "entropy": 0.8423711471259594,
      "epoch": 1.4194570474005324,
      "grad_norm": 0.173638716340065,
      "learning_rate": 5.268807311313522e-05,
      "loss": 0.8337,
      "mean_token_accuracy": 0.750945195555687,
      "num_tokens": 492409886.0,
      "step": 15428
    },
    {
      "entropy": 0.44831938948482275,
      "epoch": 1.4195490537719737,
      "grad_norm": 0.13945670425891876,
      "learning_rate": 5.2685006286993585e-05,
      "loss": 0.4276,
      "mean_token_accuracy": 0.866335067898035,
      "num_tokens": 492442043.0,
      "step": 15429
    },
    {
      "entropy": 0.48203158285468817,
      "epoch": 1.419641060143415,
      "grad_norm": 0.14383277297019958,
      "learning_rate": 5.268193946085197e-05,
      "loss": 0.4739,
      "mean_token_accuracy": 0.8563549183309078,
      "num_tokens": 492474282.0,
      "step": 15430
    },
    {
      "entropy": 0.6937399562448263,
      "epoch": 1.4197330665148562,
      "grad_norm": 0.15905322134494781,
      "learning_rate": 5.267887263471034e-05,
      "loss": 0.6948,
      "mean_token_accuracy": 0.7873479053378105,
      "num_tokens": 492506537.0,
      "step": 15431
    },
    {
      "entropy": 0.6339296102523804,
      "epoch": 1.4198250728862973,
      "grad_norm": 0.15568943321704865,
      "learning_rate": 5.2675805808568715e-05,
      "loss": 0.6294,
      "mean_token_accuracy": 0.8074956946074963,
      "num_tokens": 492538276.0,
      "step": 15432
    },
    {
      "entropy": 0.5832704873755574,
      "epoch": 1.4199170792577385,
      "grad_norm": 0.15406584739685059,
      "learning_rate": 5.267273898242708e-05,
      "loss": 0.5553,
      "mean_token_accuracy": 0.831025492399931,
      "num_tokens": 492570706.0,
      "step": 15433
    },
    {
      "entropy": 0.6195313949137926,
      "epoch": 1.4200090856291798,
      "grad_norm": 0.16686712205410004,
      "learning_rate": 5.2669672156285465e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.8157282248139381,
      "num_tokens": 492602126.0,
      "step": 15434
    },
    {
      "entropy": 0.5266499072313309,
      "epoch": 1.4201010920006212,
      "grad_norm": 0.14821334183216095,
      "learning_rate": 5.266660533014384e-05,
      "loss": 0.513,
      "mean_token_accuracy": 0.8402007892727852,
      "num_tokens": 492634705.0,
      "step": 15435
    },
    {
      "entropy": 0.6044804267585278,
      "epoch": 1.4201930983720623,
      "grad_norm": 0.15374092757701874,
      "learning_rate": 5.266353850400221e-05,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8177616931498051,
      "num_tokens": 492667186.0,
      "step": 15436
    },
    {
      "entropy": 0.6372524909675121,
      "epoch": 1.4202851047435034,
      "grad_norm": 0.16174006462097168,
      "learning_rate": 5.266047167786059e-05,
      "loss": 0.6313,
      "mean_token_accuracy": 0.8112174309790134,
      "num_tokens": 492699011.0,
      "step": 15437
    },
    {
      "entropy": 0.7507132096216083,
      "epoch": 1.4203771111149446,
      "grad_norm": 0.16751840710639954,
      "learning_rate": 5.265740485171896e-05,
      "loss": 0.7557,
      "mean_token_accuracy": 0.7739903628826141,
      "num_tokens": 492731199.0,
      "step": 15438
    },
    {
      "entropy": 0.4653239520266652,
      "epoch": 1.420469117486386,
      "grad_norm": 0.13886545598506927,
      "learning_rate": 5.265433802557733e-05,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8616646640002728,
      "num_tokens": 492763453.0,
      "step": 15439
    },
    {
      "entropy": 0.6506461333483458,
      "epoch": 1.4205611238578273,
      "grad_norm": 0.15779469907283783,
      "learning_rate": 5.2651271199435706e-05,
      "loss": 0.6356,
      "mean_token_accuracy": 0.8029118068516254,
      "num_tokens": 492795025.0,
      "step": 15440
    },
    {
      "entropy": 0.5680940989404917,
      "epoch": 1.4206531302292684,
      "grad_norm": 0.16170217096805573,
      "learning_rate": 5.264820437329409e-05,
      "loss": 0.557,
      "mean_token_accuracy": 0.8301823362708092,
      "num_tokens": 492826550.0,
      "step": 15441
    },
    {
      "entropy": 0.6944792978465557,
      "epoch": 1.4207451366007096,
      "grad_norm": 0.1711236536502838,
      "learning_rate": 5.2645137547152455e-05,
      "loss": 0.6833,
      "mean_token_accuracy": 0.7927571907639503,
      "num_tokens": 492858698.0,
      "step": 15442
    },
    {
      "entropy": 0.61983472853899,
      "epoch": 1.4208371429721507,
      "grad_norm": 0.15814639627933502,
      "learning_rate": 5.264207072101083e-05,
      "loss": 0.6051,
      "mean_token_accuracy": 0.8179514966905117,
      "num_tokens": 492890451.0,
      "step": 15443
    },
    {
      "entropy": 0.4787241346202791,
      "epoch": 1.420929149343592,
      "grad_norm": 0.14858637750148773,
      "learning_rate": 5.26390038948692e-05,
      "loss": 0.4688,
      "mean_token_accuracy": 0.8563638180494308,
      "num_tokens": 492921385.0,
      "step": 15444
    },
    {
      "entropy": 0.694378150627017,
      "epoch": 1.4210211557150334,
      "grad_norm": 0.16060350835323334,
      "learning_rate": 5.263593706872758e-05,
      "loss": 0.6698,
      "mean_token_accuracy": 0.7903412878513336,
      "num_tokens": 492953096.0,
      "step": 15445
    },
    {
      "entropy": 0.698095646686852,
      "epoch": 1.4211131620864745,
      "grad_norm": 0.1597421020269394,
      "learning_rate": 5.2632870242585954e-05,
      "loss": 0.6649,
      "mean_token_accuracy": 0.7956738285720348,
      "num_tokens": 492984125.0,
      "step": 15446
    },
    {
      "entropy": 0.6081800907850266,
      "epoch": 1.4212051684579157,
      "grad_norm": 0.16011328995227814,
      "learning_rate": 5.262980341644432e-05,
      "loss": 0.5955,
      "mean_token_accuracy": 0.8165701813995838,
      "num_tokens": 493015451.0,
      "step": 15447
    },
    {
      "entropy": 0.649647668004036,
      "epoch": 1.4212971748293568,
      "grad_norm": 0.16127711534500122,
      "learning_rate": 5.2626736590302696e-05,
      "loss": 0.6314,
      "mean_token_accuracy": 0.8074377626180649,
      "num_tokens": 493047130.0,
      "step": 15448
    },
    {
      "entropy": 0.5577567317523062,
      "epoch": 1.4213891812007982,
      "grad_norm": 0.1492285132408142,
      "learning_rate": 5.262366976416108e-05,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8392284996807575,
      "num_tokens": 493079112.0,
      "step": 15449
    },
    {
      "entropy": 0.648871548473835,
      "epoch": 1.4214811875722395,
      "grad_norm": 0.1591200977563858,
      "learning_rate": 5.2620602938019445e-05,
      "loss": 0.6403,
      "mean_token_accuracy": 0.8053562678396702,
      "num_tokens": 493111527.0,
      "step": 15450
    },
    {
      "entropy": 0.600224070250988,
      "epoch": 1.4215731939436806,
      "grad_norm": 0.1590854823589325,
      "learning_rate": 5.261753611187782e-05,
      "loss": 0.5983,
      "mean_token_accuracy": 0.8155535086989403,
      "num_tokens": 493143138.0,
      "step": 15451
    },
    {
      "entropy": 0.5958604114130139,
      "epoch": 1.4216652003151218,
      "grad_norm": 0.1521848440170288,
      "learning_rate": 5.261446928573619e-05,
      "loss": 0.5845,
      "mean_token_accuracy": 0.823216125369072,
      "num_tokens": 493175399.0,
      "step": 15452
    },
    {
      "entropy": 0.6854305155575275,
      "epoch": 1.421757206686563,
      "grad_norm": 0.16008244454860687,
      "learning_rate": 5.261140245959457e-05,
      "loss": 0.6896,
      "mean_token_accuracy": 0.789950292557478,
      "num_tokens": 493207069.0,
      "step": 15453
    },
    {
      "entropy": 0.5411847243085504,
      "epoch": 1.4218492130580043,
      "grad_norm": 0.14960095286369324,
      "learning_rate": 5.2608335633452944e-05,
      "loss": 0.5184,
      "mean_token_accuracy": 0.8387462384998798,
      "num_tokens": 493239086.0,
      "step": 15454
    },
    {
      "entropy": 0.5305888322182,
      "epoch": 1.4219412194294456,
      "grad_norm": 0.15202143788337708,
      "learning_rate": 5.260526880731131e-05,
      "loss": 0.5235,
      "mean_token_accuracy": 0.8426261954009533,
      "num_tokens": 493271437.0,
      "step": 15455
    },
    {
      "entropy": 0.5631477413699031,
      "epoch": 1.4220332258008868,
      "grad_norm": 0.1438053995370865,
      "learning_rate": 5.2602201981169687e-05,
      "loss": 0.5522,
      "mean_token_accuracy": 0.8324790149927139,
      "num_tokens": 493303973.0,
      "step": 15456
    },
    {
      "entropy": 0.5424318914301693,
      "epoch": 1.422125232172328,
      "grad_norm": 0.14758062362670898,
      "learning_rate": 5.259913515502807e-05,
      "loss": 0.5312,
      "mean_token_accuracy": 0.835383839905262,
      "num_tokens": 493335806.0,
      "step": 15457
    },
    {
      "entropy": 0.549853203818202,
      "epoch": 1.422217238543769,
      "grad_norm": 0.16142234206199646,
      "learning_rate": 5.2596068328886436e-05,
      "loss": 0.5444,
      "mean_token_accuracy": 0.8325000815093517,
      "num_tokens": 493368097.0,
      "step": 15458
    },
    {
      "entropy": 0.649102620780468,
      "epoch": 1.4223092449152104,
      "grad_norm": 0.15795624256134033,
      "learning_rate": 5.259300150274481e-05,
      "loss": 0.6343,
      "mean_token_accuracy": 0.8025830686092377,
      "num_tokens": 493400149.0,
      "step": 15459
    },
    {
      "entropy": 0.5596595872193575,
      "epoch": 1.4224012512866517,
      "grad_norm": 0.1513034999370575,
      "learning_rate": 5.258993467660318e-05,
      "loss": 0.5556,
      "mean_token_accuracy": 0.8309655822813511,
      "num_tokens": 493432720.0,
      "step": 15460
    },
    {
      "entropy": 0.6172481961548328,
      "epoch": 1.4224932576580929,
      "grad_norm": 0.16099978983402252,
      "learning_rate": 5.258686785046156e-05,
      "loss": 0.618,
      "mean_token_accuracy": 0.8095487616956234,
      "num_tokens": 493464568.0,
      "step": 15461
    },
    {
      "entropy": 0.6748708002269268,
      "epoch": 1.422585264029534,
      "grad_norm": 0.16911453008651733,
      "learning_rate": 5.2583801024319934e-05,
      "loss": 0.6641,
      "mean_token_accuracy": 0.7988059557974339,
      "num_tokens": 493495925.0,
      "step": 15462
    },
    {
      "entropy": 0.519480163231492,
      "epoch": 1.4226772704009751,
      "grad_norm": 0.15022024512290955,
      "learning_rate": 5.25807341981783e-05,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8479218445718288,
      "num_tokens": 493528378.0,
      "step": 15463
    },
    {
      "entropy": 0.6237487983889878,
      "epoch": 1.4227692767724165,
      "grad_norm": 0.15082550048828125,
      "learning_rate": 5.257766737203668e-05,
      "loss": 0.6116,
      "mean_token_accuracy": 0.815644171088934,
      "num_tokens": 493560561.0,
      "step": 15464
    },
    {
      "entropy": 0.46898384392261505,
      "epoch": 1.4228612831438578,
      "grad_norm": 0.15188191831111908,
      "learning_rate": 5.257460054589506e-05,
      "loss": 0.4466,
      "mean_token_accuracy": 0.8573509752750397,
      "num_tokens": 493592197.0,
      "step": 15465
    },
    {
      "entropy": 0.713208713568747,
      "epoch": 1.422953289515299,
      "grad_norm": 0.1638454794883728,
      "learning_rate": 5.257153371975343e-05,
      "loss": 0.7064,
      "mean_token_accuracy": 0.7851984575390816,
      "num_tokens": 493624504.0,
      "step": 15466
    },
    {
      "entropy": 0.697346618399024,
      "epoch": 1.4230452958867401,
      "grad_norm": 0.16331255435943604,
      "learning_rate": 5.25684668936118e-05,
      "loss": 0.6924,
      "mean_token_accuracy": 0.7891150116920471,
      "num_tokens": 493656425.0,
      "step": 15467
    },
    {
      "entropy": 0.5275817755609751,
      "epoch": 1.4231373022581812,
      "grad_norm": 0.15567328035831451,
      "learning_rate": 5.256540006747018e-05,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8426481895148754,
      "num_tokens": 493688059.0,
      "step": 15468
    },
    {
      "entropy": 0.5762913944199681,
      "epoch": 1.4232293086296226,
      "grad_norm": 0.15855233371257782,
      "learning_rate": 5.256233324132856e-05,
      "loss": 0.575,
      "mean_token_accuracy": 0.82537791877985,
      "num_tokens": 493719275.0,
      "step": 15469
    },
    {
      "entropy": 0.6176075870171189,
      "epoch": 1.423321315001064,
      "grad_norm": 0.15397588908672333,
      "learning_rate": 5.2559266415186925e-05,
      "loss": 0.6013,
      "mean_token_accuracy": 0.8122598119080067,
      "num_tokens": 493751660.0,
      "step": 15470
    },
    {
      "entropy": 0.4784331116825342,
      "epoch": 1.423413321372505,
      "grad_norm": 0.13852235674858093,
      "learning_rate": 5.25561995890453e-05,
      "loss": 0.459,
      "mean_token_accuracy": 0.856749452650547,
      "num_tokens": 493783984.0,
      "step": 15471
    },
    {
      "entropy": 0.6424979362636805,
      "epoch": 1.4235053277439462,
      "grad_norm": 0.152029350399971,
      "learning_rate": 5.255313276290368e-05,
      "loss": 0.6317,
      "mean_token_accuracy": 0.8069554455578327,
      "num_tokens": 493816161.0,
      "step": 15472
    },
    {
      "entropy": 0.6716794921085238,
      "epoch": 1.4235973341153874,
      "grad_norm": 0.16343562304973602,
      "learning_rate": 5.255006593676205e-05,
      "loss": 0.6525,
      "mean_token_accuracy": 0.8028629831969738,
      "num_tokens": 493848545.0,
      "step": 15473
    },
    {
      "entropy": 0.6697252318263054,
      "epoch": 1.4236893404868287,
      "grad_norm": 0.15321142971515656,
      "learning_rate": 5.2546999110620423e-05,
      "loss": 0.6637,
      "mean_token_accuracy": 0.7984671071171761,
      "num_tokens": 493880626.0,
      "step": 15474
    },
    {
      "entropy": 0.6147329467348754,
      "epoch": 1.42378134685827,
      "grad_norm": 0.1592283993959427,
      "learning_rate": 5.254393228447879e-05,
      "loss": 0.601,
      "mean_token_accuracy": 0.8148143626749516,
      "num_tokens": 493911647.0,
      "step": 15475
    },
    {
      "entropy": 0.5569762038066983,
      "epoch": 1.4238733532297112,
      "grad_norm": 0.1488126814365387,
      "learning_rate": 5.254086545833717e-05,
      "loss": 0.5484,
      "mean_token_accuracy": 0.8332904540002346,
      "num_tokens": 493943376.0,
      "step": 15476
    },
    {
      "entropy": 0.5717708887532353,
      "epoch": 1.4239653596011523,
      "grad_norm": 0.15403100848197937,
      "learning_rate": 5.253779863219555e-05,
      "loss": 0.5622,
      "mean_token_accuracy": 0.8280018717050552,
      "num_tokens": 493974029.0,
      "step": 15477
    },
    {
      "entropy": 0.6091271918267012,
      "epoch": 1.4240573659725935,
      "grad_norm": 0.15442760288715363,
      "learning_rate": 5.2534731806053915e-05,
      "loss": 0.6188,
      "mean_token_accuracy": 0.814203355461359,
      "num_tokens": 494006326.0,
      "step": 15478
    },
    {
      "entropy": 0.6360110584646463,
      "epoch": 1.4241493723440348,
      "grad_norm": 0.16406896710395813,
      "learning_rate": 5.253166497991229e-05,
      "loss": 0.6362,
      "mean_token_accuracy": 0.810696117579937,
      "num_tokens": 494038168.0,
      "step": 15479
    },
    {
      "entropy": 0.5672980449162424,
      "epoch": 1.4242413787154762,
      "grad_norm": 0.15549945831298828,
      "learning_rate": 5.252859815377067e-05,
      "loss": 0.5714,
      "mean_token_accuracy": 0.8292445354163647,
      "num_tokens": 494070215.0,
      "step": 15480
    },
    {
      "entropy": 0.6170927453786135,
      "epoch": 1.4243333850869173,
      "grad_norm": 0.1608782708644867,
      "learning_rate": 5.252553132762904e-05,
      "loss": 0.6275,
      "mean_token_accuracy": 0.8079848699271679,
      "num_tokens": 494102202.0,
      "step": 15481
    },
    {
      "entropy": 0.573135506361723,
      "epoch": 1.4244253914583584,
      "grad_norm": 0.16623716056346893,
      "learning_rate": 5.2522464501487414e-05,
      "loss": 0.5692,
      "mean_token_accuracy": 0.8254134654998779,
      "num_tokens": 494133210.0,
      "step": 15482
    },
    {
      "entropy": 0.6557944156229496,
      "epoch": 1.4245173978297996,
      "grad_norm": 0.15875041484832764,
      "learning_rate": 5.251939767534578e-05,
      "loss": 0.6493,
      "mean_token_accuracy": 0.8003769814968109,
      "num_tokens": 494165759.0,
      "step": 15483
    },
    {
      "entropy": 0.6276382557116449,
      "epoch": 1.424609404201241,
      "grad_norm": 0.15901285409927368,
      "learning_rate": 5.251633084920416e-05,
      "loss": 0.6153,
      "mean_token_accuracy": 0.8093756884336472,
      "num_tokens": 494197560.0,
      "step": 15484
    },
    {
      "entropy": 0.7169501446187496,
      "epoch": 1.4247014105726823,
      "grad_norm": 0.16499772667884827,
      "learning_rate": 5.251326402306254e-05,
      "loss": 0.7031,
      "mean_token_accuracy": 0.7863481752574444,
      "num_tokens": 494229337.0,
      "step": 15485
    },
    {
      "entropy": 0.6216110987588763,
      "epoch": 1.4247934169441234,
      "grad_norm": 0.16021858155727386,
      "learning_rate": 5.2510197196920906e-05,
      "loss": 0.6107,
      "mean_token_accuracy": 0.8133955337107182,
      "num_tokens": 494261613.0,
      "step": 15486
    },
    {
      "entropy": 0.5284197516739368,
      "epoch": 1.4248854233155646,
      "grad_norm": 0.1530025452375412,
      "learning_rate": 5.250713037077928e-05,
      "loss": 0.51,
      "mean_token_accuracy": 0.8415730632841587,
      "num_tokens": 494293937.0,
      "step": 15487
    },
    {
      "entropy": 0.5499851498752832,
      "epoch": 1.4249774296870057,
      "grad_norm": 0.14426003396511078,
      "learning_rate": 5.250406354463766e-05,
      "loss": 0.526,
      "mean_token_accuracy": 0.8377081081271172,
      "num_tokens": 494326019.0,
      "step": 15488
    },
    {
      "entropy": 0.6070480644702911,
      "epoch": 1.425069436058447,
      "grad_norm": 0.15790696442127228,
      "learning_rate": 5.250099671849603e-05,
      "loss": 0.6022,
      "mean_token_accuracy": 0.8204613141715527,
      "num_tokens": 494358250.0,
      "step": 15489
    },
    {
      "entropy": 0.5584985613822937,
      "epoch": 1.4251614424298884,
      "grad_norm": 0.1526651233434677,
      "learning_rate": 5.2497929892354404e-05,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8318184167146683,
      "num_tokens": 494390531.0,
      "step": 15490
    },
    {
      "entropy": 0.7372242789715528,
      "epoch": 1.4252534488013295,
      "grad_norm": 0.1687421053647995,
      "learning_rate": 5.249486306621277e-05,
      "loss": 0.724,
      "mean_token_accuracy": 0.7825111858546734,
      "num_tokens": 494421472.0,
      "step": 15491
    },
    {
      "entropy": 0.6616543922573328,
      "epoch": 1.4253454551727707,
      "grad_norm": 0.15571968257427216,
      "learning_rate": 5.2491796240071154e-05,
      "loss": 0.6458,
      "mean_token_accuracy": 0.8004332557320595,
      "num_tokens": 494453346.0,
      "step": 15492
    },
    {
      "entropy": 0.6287122913636267,
      "epoch": 1.4254374615442118,
      "grad_norm": 0.15762270987033844,
      "learning_rate": 5.248872941392953e-05,
      "loss": 0.6091,
      "mean_token_accuracy": 0.8152336403727531,
      "num_tokens": 494486042.0,
      "step": 15493
    },
    {
      "entropy": 0.5919103398919106,
      "epoch": 1.4255294679156532,
      "grad_norm": 0.16516706347465515,
      "learning_rate": 5.2485662587787896e-05,
      "loss": 0.586,
      "mean_token_accuracy": 0.819897998124361,
      "num_tokens": 494517609.0,
      "step": 15494
    },
    {
      "entropy": 0.6638571973890066,
      "epoch": 1.4256214742870945,
      "grad_norm": 0.15782806277275085,
      "learning_rate": 5.248259576164627e-05,
      "loss": 0.6659,
      "mean_token_accuracy": 0.7983730621635914,
      "num_tokens": 494549792.0,
      "step": 15495
    },
    {
      "entropy": 0.6964529231190681,
      "epoch": 1.4257134806585356,
      "grad_norm": 0.16344279050827026,
      "learning_rate": 5.247952893550465e-05,
      "loss": 0.6949,
      "mean_token_accuracy": 0.7888756692409515,
      "num_tokens": 494582400.0,
      "step": 15496
    },
    {
      "entropy": 0.5617287373170257,
      "epoch": 1.4258054870299768,
      "grad_norm": 0.15550006926059723,
      "learning_rate": 5.247646210936302e-05,
      "loss": 0.5596,
      "mean_token_accuracy": 0.8271918185055256,
      "num_tokens": 494614586.0,
      "step": 15497
    },
    {
      "entropy": 0.5905014872550964,
      "epoch": 1.425897493401418,
      "grad_norm": 0.1555827558040619,
      "learning_rate": 5.2473395283221395e-05,
      "loss": 0.5727,
      "mean_token_accuracy": 0.8205106295645237,
      "num_tokens": 494645689.0,
      "step": 15498
    },
    {
      "entropy": 0.6069281222298741,
      "epoch": 1.4259894997728593,
      "grad_norm": 0.15804670751094818,
      "learning_rate": 5.2470328457079776e-05,
      "loss": 0.5894,
      "mean_token_accuracy": 0.8178148791193962,
      "num_tokens": 494677490.0,
      "step": 15499
    },
    {
      "entropy": 0.5971739124506712,
      "epoch": 1.4260815061443006,
      "grad_norm": 0.15573245286941528,
      "learning_rate": 5.2467261630938144e-05,
      "loss": 0.5913,
      "mean_token_accuracy": 0.8226595297455788,
      "num_tokens": 494708911.0,
      "step": 15500
    },
    {
      "entropy": 0.6146353492513299,
      "epoch": 1.4261735125157418,
      "grad_norm": 0.15323524177074432,
      "learning_rate": 5.246419480479652e-05,
      "loss": 0.595,
      "mean_token_accuracy": 0.8151458539068699,
      "num_tokens": 494740850.0,
      "step": 15501
    },
    {
      "entropy": 0.8214519312605262,
      "epoch": 1.426265518887183,
      "grad_norm": 0.167527437210083,
      "learning_rate": 5.2461127978654886e-05,
      "loss": 0.8156,
      "mean_token_accuracy": 0.7581864222884178,
      "num_tokens": 494772657.0,
      "step": 15502
    },
    {
      "entropy": 0.47018243838101625,
      "epoch": 1.426357525258624,
      "grad_norm": 0.14443692564964294,
      "learning_rate": 5.2458061152513275e-05,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8613918125629425,
      "num_tokens": 494805139.0,
      "step": 15503
    },
    {
      "entropy": 0.5635250881314278,
      "epoch": 1.4264495316300654,
      "grad_norm": 0.14645032584667206,
      "learning_rate": 5.245499432637164e-05,
      "loss": 0.544,
      "mean_token_accuracy": 0.8347110599279404,
      "num_tokens": 494837444.0,
      "step": 15504
    },
    {
      "entropy": 0.6015375219285488,
      "epoch": 1.4265415380015067,
      "grad_norm": 0.1529208868741989,
      "learning_rate": 5.245192750023001e-05,
      "loss": 0.5838,
      "mean_token_accuracy": 0.8183358758687973,
      "num_tokens": 494869442.0,
      "step": 15505
    },
    {
      "entropy": 0.5208538942970335,
      "epoch": 1.4266335443729479,
      "grad_norm": 0.14207255840301514,
      "learning_rate": 5.2448860674088385e-05,
      "loss": 0.5018,
      "mean_token_accuracy": 0.8440082855522633,
      "num_tokens": 494901462.0,
      "step": 15506
    },
    {
      "entropy": 0.5146657004952431,
      "epoch": 1.426725550744389,
      "grad_norm": 0.14714764058589935,
      "learning_rate": 5.2445793847946766e-05,
      "loss": 0.488,
      "mean_token_accuracy": 0.851256463676691,
      "num_tokens": 494933776.0,
      "step": 15507
    },
    {
      "entropy": 0.6827997267246246,
      "epoch": 1.4268175571158301,
      "grad_norm": 0.16255541145801544,
      "learning_rate": 5.244272702180514e-05,
      "loss": 0.6754,
      "mean_token_accuracy": 0.7970612086355686,
      "num_tokens": 494965862.0,
      "step": 15508
    },
    {
      "entropy": 0.598397970199585,
      "epoch": 1.4269095634872715,
      "grad_norm": 0.16387929022312164,
      "learning_rate": 5.243966019566351e-05,
      "loss": 0.577,
      "mean_token_accuracy": 0.8228239975869656,
      "num_tokens": 494997769.0,
      "step": 15509
    },
    {
      "entropy": 0.7195404376834631,
      "epoch": 1.4270015698587128,
      "grad_norm": 0.16005578637123108,
      "learning_rate": 5.2436593369521884e-05,
      "loss": 0.6997,
      "mean_token_accuracy": 0.7870693579316139,
      "num_tokens": 495029547.0,
      "step": 15510
    },
    {
      "entropy": 0.6684610284864902,
      "epoch": 1.427093576230154,
      "grad_norm": 0.16052021086215973,
      "learning_rate": 5.2433526543380265e-05,
      "loss": 0.669,
      "mean_token_accuracy": 0.7977222613990307,
      "num_tokens": 495062030.0,
      "step": 15511
    },
    {
      "entropy": 0.6265624207444489,
      "epoch": 1.4271855826015951,
      "grad_norm": 0.15128636360168457,
      "learning_rate": 5.243045971723863e-05,
      "loss": 0.6157,
      "mean_token_accuracy": 0.8093544580042362,
      "num_tokens": 495094569.0,
      "step": 15512
    },
    {
      "entropy": 0.6803357200697064,
      "epoch": 1.4272775889730362,
      "grad_norm": 0.1681784987449646,
      "learning_rate": 5.242739289109701e-05,
      "loss": 0.6625,
      "mean_token_accuracy": 0.7963150255382061,
      "num_tokens": 495126089.0,
      "step": 15513
    },
    {
      "entropy": 0.6701808422803879,
      "epoch": 1.4273695953444776,
      "grad_norm": 0.16046062111854553,
      "learning_rate": 5.2424326064955375e-05,
      "loss": 0.6603,
      "mean_token_accuracy": 0.799200814217329,
      "num_tokens": 495158219.0,
      "step": 15514
    },
    {
      "entropy": 0.6680217757821083,
      "epoch": 1.427461601715919,
      "grad_norm": 0.16348989307880402,
      "learning_rate": 5.242125923881376e-05,
      "loss": 0.6801,
      "mean_token_accuracy": 0.7964480519294739,
      "num_tokens": 495189907.0,
      "step": 15515
    },
    {
      "entropy": 0.6623592469841242,
      "epoch": 1.42755360808736,
      "grad_norm": 0.16864866018295288,
      "learning_rate": 5.241819241267213e-05,
      "loss": 0.6537,
      "mean_token_accuracy": 0.7979048527777195,
      "num_tokens": 495221784.0,
      "step": 15516
    },
    {
      "entropy": 0.8263909835368395,
      "epoch": 1.4276456144588012,
      "grad_norm": 0.1701730191707611,
      "learning_rate": 5.24151255865305e-05,
      "loss": 0.8148,
      "mean_token_accuracy": 0.7535849697887897,
      "num_tokens": 495253623.0,
      "step": 15517
    },
    {
      "entropy": 0.585370269138366,
      "epoch": 1.4277376208302424,
      "grad_norm": 0.1546233743429184,
      "learning_rate": 5.2412058760388874e-05,
      "loss": 0.578,
      "mean_token_accuracy": 0.8252248540520668,
      "num_tokens": 495285622.0,
      "step": 15518
    },
    {
      "entropy": 0.5743485558778048,
      "epoch": 1.4278296272016837,
      "grad_norm": 0.15352453291416168,
      "learning_rate": 5.2408991934247255e-05,
      "loss": 0.5521,
      "mean_token_accuracy": 0.823016207665205,
      "num_tokens": 495316944.0,
      "step": 15519
    },
    {
      "entropy": 0.6220565922558308,
      "epoch": 1.427921633573125,
      "grad_norm": 0.16064772009849548,
      "learning_rate": 5.240592510810562e-05,
      "loss": 0.6009,
      "mean_token_accuracy": 0.8160230778157711,
      "num_tokens": 495348875.0,
      "step": 15520
    },
    {
      "entropy": 0.6096883825957775,
      "epoch": 1.4280136399445662,
      "grad_norm": 0.1637454479932785,
      "learning_rate": 5.2402858281964e-05,
      "loss": 0.6064,
      "mean_token_accuracy": 0.8164507336914539,
      "num_tokens": 495380341.0,
      "step": 15521
    },
    {
      "entropy": 0.6037893607281148,
      "epoch": 1.4281056463160073,
      "grad_norm": 0.1544988751411438,
      "learning_rate": 5.2399791455822366e-05,
      "loss": 0.5979,
      "mean_token_accuracy": 0.8171330280601978,
      "num_tokens": 495411570.0,
      "step": 15522
    },
    {
      "entropy": 0.7070438042283058,
      "epoch": 1.4281976526874485,
      "grad_norm": 0.16264590620994568,
      "learning_rate": 5.239672462968075e-05,
      "loss": 0.6947,
      "mean_token_accuracy": 0.7846408188343048,
      "num_tokens": 495442615.0,
      "step": 15523
    },
    {
      "entropy": 0.5019555301405489,
      "epoch": 1.4282896590588898,
      "grad_norm": 0.14463888108730316,
      "learning_rate": 5.239365780353912e-05,
      "loss": 0.4979,
      "mean_token_accuracy": 0.847332913428545,
      "num_tokens": 495475007.0,
      "step": 15524
    },
    {
      "entropy": 0.6825884971767664,
      "epoch": 1.4283816654303312,
      "grad_norm": 0.15230657160282135,
      "learning_rate": 5.239059097739749e-05,
      "loss": 0.6743,
      "mean_token_accuracy": 0.7962395064532757,
      "num_tokens": 495507540.0,
      "step": 15525
    },
    {
      "entropy": 0.606282171793282,
      "epoch": 1.4284736718017723,
      "grad_norm": 0.1571529656648636,
      "learning_rate": 5.2387524151255864e-05,
      "loss": 0.5873,
      "mean_token_accuracy": 0.8186900354921818,
      "num_tokens": 495539477.0,
      "step": 15526
    },
    {
      "entropy": 0.7039528526365757,
      "epoch": 1.4285656781732134,
      "grad_norm": 0.1623801440000534,
      "learning_rate": 5.2384457325114246e-05,
      "loss": 0.6939,
      "mean_token_accuracy": 0.7900258265435696,
      "num_tokens": 495571165.0,
      "step": 15527
    },
    {
      "entropy": 0.6743398569524288,
      "epoch": 1.4286576845446546,
      "grad_norm": 0.16377325356006622,
      "learning_rate": 5.2381390498972614e-05,
      "loss": 0.6663,
      "mean_token_accuracy": 0.7967530637979507,
      "num_tokens": 495602161.0,
      "step": 15528
    },
    {
      "entropy": 0.6770967356860638,
      "epoch": 1.428749690916096,
      "grad_norm": 0.15396133065223694,
      "learning_rate": 5.237832367283099e-05,
      "loss": 0.6575,
      "mean_token_accuracy": 0.8009378127753735,
      "num_tokens": 495634753.0,
      "step": 15529
    },
    {
      "entropy": 0.6261424925178289,
      "epoch": 1.4288416972875373,
      "grad_norm": 0.16255968809127808,
      "learning_rate": 5.237525684668937e-05,
      "loss": 0.6151,
      "mean_token_accuracy": 0.8099711462855339,
      "num_tokens": 495666262.0,
      "step": 15530
    },
    {
      "entropy": 0.6696912087500095,
      "epoch": 1.4289337036589784,
      "grad_norm": 0.1588861346244812,
      "learning_rate": 5.237219002054774e-05,
      "loss": 0.6779,
      "mean_token_accuracy": 0.7960153333842754,
      "num_tokens": 495698574.0,
      "step": 15531
    },
    {
      "entropy": 0.4911328498274088,
      "epoch": 1.4290257100304196,
      "grad_norm": 0.14802949130535126,
      "learning_rate": 5.236912319440611e-05,
      "loss": 0.485,
      "mean_token_accuracy": 0.8495148532092571,
      "num_tokens": 495730098.0,
      "step": 15532
    },
    {
      "entropy": 0.5986253754235804,
      "epoch": 1.4291177164018607,
      "grad_norm": 0.15341894328594208,
      "learning_rate": 5.236605636826448e-05,
      "loss": 0.5907,
      "mean_token_accuracy": 0.8162140734493732,
      "num_tokens": 495762532.0,
      "step": 15533
    },
    {
      "entropy": 0.5411570221185684,
      "epoch": 1.429209722773302,
      "grad_norm": 0.16025404632091522,
      "learning_rate": 5.236298954212286e-05,
      "loss": 0.5325,
      "mean_token_accuracy": 0.8350036665797234,
      "num_tokens": 495794148.0,
      "step": 15534
    },
    {
      "entropy": 0.6571971084922552,
      "epoch": 1.4293017291447434,
      "grad_norm": 0.15625780820846558,
      "learning_rate": 5.2359922715981236e-05,
      "loss": 0.6559,
      "mean_token_accuracy": 0.7992537841200829,
      "num_tokens": 495826528.0,
      "step": 15535
    },
    {
      "entropy": 0.600527934730053,
      "epoch": 1.4293937355161845,
      "grad_norm": 0.160993754863739,
      "learning_rate": 5.2356855889839604e-05,
      "loss": 0.591,
      "mean_token_accuracy": 0.8189083933830261,
      "num_tokens": 495858044.0,
      "step": 15536
    },
    {
      "entropy": 0.5849609514698386,
      "epoch": 1.4294857418876257,
      "grad_norm": 0.15181121230125427,
      "learning_rate": 5.235378906369798e-05,
      "loss": 0.573,
      "mean_token_accuracy": 0.8237663693726063,
      "num_tokens": 495890001.0,
      "step": 15537
    },
    {
      "entropy": 0.605909607373178,
      "epoch": 1.4295777482590668,
      "grad_norm": 0.15270961821079254,
      "learning_rate": 5.235072223755636e-05,
      "loss": 0.6066,
      "mean_token_accuracy": 0.8166850097477436,
      "num_tokens": 495922546.0,
      "step": 15538
    },
    {
      "entropy": 0.6771186171099544,
      "epoch": 1.4296697546305082,
      "grad_norm": 0.15986496210098267,
      "learning_rate": 5.234765541141473e-05,
      "loss": 0.6652,
      "mean_token_accuracy": 0.7939507327973843,
      "num_tokens": 495954577.0,
      "step": 15539
    },
    {
      "entropy": 0.6305023268796504,
      "epoch": 1.4297617610019495,
      "grad_norm": 0.15191328525543213,
      "learning_rate": 5.23445885852731e-05,
      "loss": 0.6127,
      "mean_token_accuracy": 0.8116863891482353,
      "num_tokens": 495986673.0,
      "step": 15540
    },
    {
      "entropy": 0.777996364980936,
      "epoch": 1.4298537673733906,
      "grad_norm": 0.16707053780555725,
      "learning_rate": 5.234152175913147e-05,
      "loss": 0.7827,
      "mean_token_accuracy": 0.7644574828445911,
      "num_tokens": 496019169.0,
      "step": 15541
    },
    {
      "entropy": 0.6923096179962158,
      "epoch": 1.4299457737448318,
      "grad_norm": 0.16088423132896423,
      "learning_rate": 5.233845493298986e-05,
      "loss": 0.6893,
      "mean_token_accuracy": 0.7936426214873791,
      "num_tokens": 496051658.0,
      "step": 15542
    },
    {
      "entropy": 0.7082524131983519,
      "epoch": 1.430037780116273,
      "grad_norm": 0.16984882950782776,
      "learning_rate": 5.233538810684823e-05,
      "loss": 0.714,
      "mean_token_accuracy": 0.784333098679781,
      "num_tokens": 496083425.0,
      "step": 15543
    },
    {
      "entropy": 0.5681199717801064,
      "epoch": 1.4301297864877143,
      "grad_norm": 0.14803798496723175,
      "learning_rate": 5.2332321280706595e-05,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8322671949863434,
      "num_tokens": 496116044.0,
      "step": 15544
    },
    {
      "entropy": 0.5837467247620225,
      "epoch": 1.4302217928591556,
      "grad_norm": 0.15712998807430267,
      "learning_rate": 5.232925445456497e-05,
      "loss": 0.567,
      "mean_token_accuracy": 0.824740894138813,
      "num_tokens": 496147641.0,
      "step": 15545
    },
    {
      "entropy": 0.6727136354893446,
      "epoch": 1.4303137992305968,
      "grad_norm": 0.1571642905473709,
      "learning_rate": 5.232618762842335e-05,
      "loss": 0.647,
      "mean_token_accuracy": 0.8051378838717937,
      "num_tokens": 496180300.0,
      "step": 15546
    },
    {
      "entropy": 0.7248712573200464,
      "epoch": 1.430405805602038,
      "grad_norm": 0.16477234661579132,
      "learning_rate": 5.2323120802281725e-05,
      "loss": 0.705,
      "mean_token_accuracy": 0.78655606508255,
      "num_tokens": 496212143.0,
      "step": 15547
    },
    {
      "entropy": 0.6836782619357109,
      "epoch": 1.430497811973479,
      "grad_norm": 0.16371457278728485,
      "learning_rate": 5.232005397614009e-05,
      "loss": 0.6654,
      "mean_token_accuracy": 0.7942860312759876,
      "num_tokens": 496243457.0,
      "step": 15548
    },
    {
      "entropy": 0.48127134214155376,
      "epoch": 1.4305898183449204,
      "grad_norm": 0.15062293410301208,
      "learning_rate": 5.231698714999846e-05,
      "loss": 0.4629,
      "mean_token_accuracy": 0.855895396322012,
      "num_tokens": 496275789.0,
      "step": 15549
    },
    {
      "entropy": 0.5651678089052439,
      "epoch": 1.4306818247163617,
      "grad_norm": 0.15826795995235443,
      "learning_rate": 5.231392032385685e-05,
      "loss": 0.5599,
      "mean_token_accuracy": 0.8280559740960598,
      "num_tokens": 496306537.0,
      "step": 15550
    },
    {
      "entropy": 0.6159114374313504,
      "epoch": 1.4307738310878029,
      "grad_norm": 0.15783841907978058,
      "learning_rate": 5.231085349771522e-05,
      "loss": 0.6029,
      "mean_token_accuracy": 0.8151340819895267,
      "num_tokens": 496338416.0,
      "step": 15551
    },
    {
      "entropy": 0.5583408316597342,
      "epoch": 1.430865837459244,
      "grad_norm": 0.15825775265693665,
      "learning_rate": 5.230778667157359e-05,
      "loss": 0.5597,
      "mean_token_accuracy": 0.8290204852819443,
      "num_tokens": 496370374.0,
      "step": 15552
    },
    {
      "entropy": 0.6395311951637268,
      "epoch": 1.4309578438306851,
      "grad_norm": 0.16246768832206726,
      "learning_rate": 5.230471984543196e-05,
      "loss": 0.6323,
      "mean_token_accuracy": 0.8061164431273937,
      "num_tokens": 496401284.0,
      "step": 15553
    },
    {
      "entropy": 0.5832535242661834,
      "epoch": 1.4310498502021265,
      "grad_norm": 0.1504589468240738,
      "learning_rate": 5.230165301929034e-05,
      "loss": 0.5798,
      "mean_token_accuracy": 0.824747309088707,
      "num_tokens": 496433416.0,
      "step": 15554
    },
    {
      "entropy": 0.5659232120960951,
      "epoch": 1.4311418565735678,
      "grad_norm": 0.15203030407428741,
      "learning_rate": 5.2298586193148716e-05,
      "loss": 0.5603,
      "mean_token_accuracy": 0.826669916510582,
      "num_tokens": 496465003.0,
      "step": 15555
    },
    {
      "entropy": 0.5704447254538536,
      "epoch": 1.431233862945009,
      "grad_norm": 0.1532922387123108,
      "learning_rate": 5.2295519367007084e-05,
      "loss": 0.5575,
      "mean_token_accuracy": 0.8282331898808479,
      "num_tokens": 496496869.0,
      "step": 15556
    },
    {
      "entropy": 0.657305283471942,
      "epoch": 1.4313258693164501,
      "grad_norm": 0.16869838535785675,
      "learning_rate": 5.229245254086546e-05,
      "loss": 0.6411,
      "mean_token_accuracy": 0.8054153509438038,
      "num_tokens": 496528679.0,
      "step": 15557
    },
    {
      "entropy": 0.5186298228800297,
      "epoch": 1.4314178756878913,
      "grad_norm": 0.1503845751285553,
      "learning_rate": 5.228938571472384e-05,
      "loss": 0.5117,
      "mean_token_accuracy": 0.842134028673172,
      "num_tokens": 496561371.0,
      "step": 15558
    },
    {
      "entropy": 0.5710265669040382,
      "epoch": 1.4315098820593326,
      "grad_norm": 0.15567637979984283,
      "learning_rate": 5.228631888858221e-05,
      "loss": 0.5663,
      "mean_token_accuracy": 0.8263734579086304,
      "num_tokens": 496593247.0,
      "step": 15559
    },
    {
      "entropy": 0.6094832462258637,
      "epoch": 1.431601888430774,
      "grad_norm": 0.15532760322093964,
      "learning_rate": 5.228325206244058e-05,
      "loss": 0.5852,
      "mean_token_accuracy": 0.8163229003548622,
      "num_tokens": 496625140.0,
      "step": 15560
    },
    {
      "entropy": 0.6899444796144962,
      "epoch": 1.431693894802215,
      "grad_norm": 0.16899873316287994,
      "learning_rate": 5.2280185236298964e-05,
      "loss": 0.6788,
      "mean_token_accuracy": 0.7939147129654884,
      "num_tokens": 496657183.0,
      "step": 15561
    },
    {
      "entropy": 0.6562434025108814,
      "epoch": 1.4317859011736562,
      "grad_norm": 0.15254877507686615,
      "learning_rate": 5.227711841015733e-05,
      "loss": 0.6392,
      "mean_token_accuracy": 0.8035753108561039,
      "num_tokens": 496688940.0,
      "step": 15562
    },
    {
      "entropy": 0.7038390804082155,
      "epoch": 1.4318779075450974,
      "grad_norm": 0.16772902011871338,
      "learning_rate": 5.2274051584015706e-05,
      "loss": 0.6839,
      "mean_token_accuracy": 0.7914648205041885,
      "num_tokens": 496720573.0,
      "step": 15563
    },
    {
      "entropy": 0.5427443133667111,
      "epoch": 1.4319699139165387,
      "grad_norm": 0.14819267392158508,
      "learning_rate": 5.2270984757874074e-05,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8383729234337807,
      "num_tokens": 496752802.0,
      "step": 15564
    },
    {
      "entropy": 0.652543980628252,
      "epoch": 1.43206192028798,
      "grad_norm": 0.15710948407649994,
      "learning_rate": 5.2267917931732455e-05,
      "loss": 0.6383,
      "mean_token_accuracy": 0.8050408996641636,
      "num_tokens": 496785438.0,
      "step": 15565
    },
    {
      "entropy": 0.676066592335701,
      "epoch": 1.4321539266594212,
      "grad_norm": 0.15932267904281616,
      "learning_rate": 5.226485110559083e-05,
      "loss": 0.6644,
      "mean_token_accuracy": 0.7945845127105713,
      "num_tokens": 496817667.0,
      "step": 15566
    },
    {
      "entropy": 0.7452034838497639,
      "epoch": 1.4322459330308623,
      "grad_norm": 0.16575980186462402,
      "learning_rate": 5.22617842794492e-05,
      "loss": 0.7339,
      "mean_token_accuracy": 0.7787391915917397,
      "num_tokens": 496849550.0,
      "step": 15567
    },
    {
      "entropy": 0.519769661128521,
      "epoch": 1.4323379394023035,
      "grad_norm": 0.15292055904865265,
      "learning_rate": 5.225871745330757e-05,
      "loss": 0.5069,
      "mean_token_accuracy": 0.8452047072350979,
      "num_tokens": 496881515.0,
      "step": 15568
    },
    {
      "entropy": 0.6270600128918886,
      "epoch": 1.4324299457737448,
      "grad_norm": 0.16322365403175354,
      "learning_rate": 5.2255650627165954e-05,
      "loss": 0.6261,
      "mean_token_accuracy": 0.8064805902540684,
      "num_tokens": 496912508.0,
      "step": 15569
    },
    {
      "entropy": 0.6274191737174988,
      "epoch": 1.4325219521451862,
      "grad_norm": 0.1572556495666504,
      "learning_rate": 5.225258380102432e-05,
      "loss": 0.6396,
      "mean_token_accuracy": 0.8053786791861057,
      "num_tokens": 496944195.0,
      "step": 15570
    },
    {
      "entropy": 0.5807233024388552,
      "epoch": 1.4326139585166273,
      "grad_norm": 0.15483389794826508,
      "learning_rate": 5.2249516974882696e-05,
      "loss": 0.5725,
      "mean_token_accuracy": 0.8266893960535526,
      "num_tokens": 496976576.0,
      "step": 15571
    },
    {
      "entropy": 0.7257951013743877,
      "epoch": 1.4327059648880685,
      "grad_norm": 0.1604730784893036,
      "learning_rate": 5.2246450148741064e-05,
      "loss": 0.7181,
      "mean_token_accuracy": 0.781004223972559,
      "num_tokens": 497008424.0,
      "step": 15572
    },
    {
      "entropy": 0.6351621299982071,
      "epoch": 1.4327979712595096,
      "grad_norm": 0.15546129643917084,
      "learning_rate": 5.2243383322599446e-05,
      "loss": 0.6253,
      "mean_token_accuracy": 0.8061861507594585,
      "num_tokens": 497040718.0,
      "step": 15573
    },
    {
      "entropy": 0.5865325201302767,
      "epoch": 1.432889977630951,
      "grad_norm": 0.16476744413375854,
      "learning_rate": 5.224031649645782e-05,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8208578452467918,
      "num_tokens": 497072071.0,
      "step": 15574
    },
    {
      "entropy": 0.5767264193855226,
      "epoch": 1.4329819840023923,
      "grad_norm": 0.15214946866035461,
      "learning_rate": 5.223724967031619e-05,
      "loss": 0.5696,
      "mean_token_accuracy": 0.8223808892071247,
      "num_tokens": 497104472.0,
      "step": 15575
    },
    {
      "entropy": 0.6176936458796263,
      "epoch": 1.4330739903738334,
      "grad_norm": 0.1537388414144516,
      "learning_rate": 5.223418284417456e-05,
      "loss": 0.5937,
      "mean_token_accuracy": 0.8215548098087311,
      "num_tokens": 497136350.0,
      "step": 15576
    },
    {
      "entropy": 0.6339197922497988,
      "epoch": 1.4331659967452746,
      "grad_norm": 0.1575649529695511,
      "learning_rate": 5.2231116018032944e-05,
      "loss": 0.624,
      "mean_token_accuracy": 0.8054828867316246,
      "num_tokens": 497168921.0,
      "step": 15577
    },
    {
      "entropy": 0.6101653166115284,
      "epoch": 1.4332580031167157,
      "grad_norm": 0.15653713047504425,
      "learning_rate": 5.222804919189131e-05,
      "loss": 0.611,
      "mean_token_accuracy": 0.8161866776645184,
      "num_tokens": 497200200.0,
      "step": 15578
    },
    {
      "entropy": 0.4948614537715912,
      "epoch": 1.433350009488157,
      "grad_norm": 0.14620588719844818,
      "learning_rate": 5.222498236574969e-05,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8500591814517975,
      "num_tokens": 497232294.0,
      "step": 15579
    },
    {
      "entropy": 0.5219084955751896,
      "epoch": 1.4334420158595984,
      "grad_norm": 0.14899034798145294,
      "learning_rate": 5.2221915539608055e-05,
      "loss": 0.5225,
      "mean_token_accuracy": 0.8374692350625992,
      "num_tokens": 497264788.0,
      "step": 15580
    },
    {
      "entropy": 0.5367708122357726,
      "epoch": 1.4335340222310395,
      "grad_norm": 0.15524810552597046,
      "learning_rate": 5.2218848713466436e-05,
      "loss": 0.5323,
      "mean_token_accuracy": 0.8382177501916885,
      "num_tokens": 497297070.0,
      "step": 15581
    },
    {
      "entropy": 0.45924649597145617,
      "epoch": 1.4336260286024807,
      "grad_norm": 0.13894911110401154,
      "learning_rate": 5.221578188732481e-05,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8615890145301819,
      "num_tokens": 497328441.0,
      "step": 15582
    },
    {
      "entropy": 0.6116392151452601,
      "epoch": 1.4337180349739218,
      "grad_norm": 0.1574968844652176,
      "learning_rate": 5.221271506118318e-05,
      "loss": 0.6047,
      "mean_token_accuracy": 0.8166877068579197,
      "num_tokens": 497360512.0,
      "step": 15583
    },
    {
      "entropy": 0.6459121108055115,
      "epoch": 1.4338100413453632,
      "grad_norm": 0.16094985604286194,
      "learning_rate": 5.220964823504155e-05,
      "loss": 0.6351,
      "mean_token_accuracy": 0.805361021310091,
      "num_tokens": 497392441.0,
      "step": 15584
    },
    {
      "entropy": 0.6338962360750884,
      "epoch": 1.4339020477168045,
      "grad_norm": 0.1682024598121643,
      "learning_rate": 5.2206581408899935e-05,
      "loss": 0.6226,
      "mean_token_accuracy": 0.8082119151949883,
      "num_tokens": 497423878.0,
      "step": 15585
    },
    {
      "entropy": 0.5431099990382791,
      "epoch": 1.4339940540882457,
      "grad_norm": 0.15079694986343384,
      "learning_rate": 5.220351458275831e-05,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8376238383352757,
      "num_tokens": 497455933.0,
      "step": 15586
    },
    {
      "entropy": 0.7056921441107988,
      "epoch": 1.4340860604596868,
      "grad_norm": 0.1613268256187439,
      "learning_rate": 5.220044775661668e-05,
      "loss": 0.6929,
      "mean_token_accuracy": 0.7879372946918011,
      "num_tokens": 497488316.0,
      "step": 15587
    },
    {
      "entropy": 0.6108889486640692,
      "epoch": 1.434178066831128,
      "grad_norm": 0.14608365297317505,
      "learning_rate": 5.219738093047506e-05,
      "loss": 0.5849,
      "mean_token_accuracy": 0.818357590585947,
      "num_tokens": 497520432.0,
      "step": 15588
    },
    {
      "entropy": 0.6475992687046528,
      "epoch": 1.4342700732025693,
      "grad_norm": 0.15560220181941986,
      "learning_rate": 5.219431410433343e-05,
      "loss": 0.6415,
      "mean_token_accuracy": 0.8053792268037796,
      "num_tokens": 497551737.0,
      "step": 15589
    },
    {
      "entropy": 0.5963029712438583,
      "epoch": 1.4343620795740106,
      "grad_norm": 0.1582372784614563,
      "learning_rate": 5.21912472781918e-05,
      "loss": 0.5769,
      "mean_token_accuracy": 0.8217727690935135,
      "num_tokens": 497583788.0,
      "step": 15590
    },
    {
      "entropy": 0.5835982300341129,
      "epoch": 1.4344540859454518,
      "grad_norm": 0.15706956386566162,
      "learning_rate": 5.2188180452050176e-05,
      "loss": 0.5887,
      "mean_token_accuracy": 0.8231544904410839,
      "num_tokens": 497615606.0,
      "step": 15591
    },
    {
      "entropy": 0.6426843078806996,
      "epoch": 1.434546092316893,
      "grad_norm": 0.16749578714370728,
      "learning_rate": 5.218511362590856e-05,
      "loss": 0.6356,
      "mean_token_accuracy": 0.8057844527065754,
      "num_tokens": 497647559.0,
      "step": 15592
    },
    {
      "entropy": 0.618081521242857,
      "epoch": 1.434638098688334,
      "grad_norm": 0.1683102697134018,
      "learning_rate": 5.2182046799766925e-05,
      "loss": 0.6079,
      "mean_token_accuracy": 0.815703734755516,
      "num_tokens": 497679151.0,
      "step": 15593
    },
    {
      "entropy": 0.685479050502181,
      "epoch": 1.4347301050597754,
      "grad_norm": 0.15824250876903534,
      "learning_rate": 5.21789799736253e-05,
      "loss": 0.6665,
      "mean_token_accuracy": 0.7980564832687378,
      "num_tokens": 497711114.0,
      "step": 15594
    },
    {
      "entropy": 0.676484664902091,
      "epoch": 1.4348221114312167,
      "grad_norm": 0.1559326946735382,
      "learning_rate": 5.217591314748367e-05,
      "loss": 0.6624,
      "mean_token_accuracy": 0.798570055514574,
      "num_tokens": 497743781.0,
      "step": 15595
    },
    {
      "entropy": 0.6954952292144299,
      "epoch": 1.4349141178026579,
      "grad_norm": 0.16477741301059723,
      "learning_rate": 5.217284632134205e-05,
      "loss": 0.6817,
      "mean_token_accuracy": 0.7955702990293503,
      "num_tokens": 497774907.0,
      "step": 15596
    },
    {
      "entropy": 0.5785264726728201,
      "epoch": 1.435006124174099,
      "grad_norm": 0.15809564292430878,
      "learning_rate": 5.2169779495200424e-05,
      "loss": 0.5585,
      "mean_token_accuracy": 0.822939570993185,
      "num_tokens": 497806573.0,
      "step": 15597
    },
    {
      "entropy": 0.5892633004114032,
      "epoch": 1.4350981305455401,
      "grad_norm": 0.14904990792274475,
      "learning_rate": 5.216671266905879e-05,
      "loss": 0.568,
      "mean_token_accuracy": 0.8214247301220894,
      "num_tokens": 497839198.0,
      "step": 15598
    },
    {
      "entropy": 0.620760187972337,
      "epoch": 1.4351901369169815,
      "grad_norm": 0.15674641728401184,
      "learning_rate": 5.2163645842917166e-05,
      "loss": 0.6071,
      "mean_token_accuracy": 0.8132009319961071,
      "num_tokens": 497871163.0,
      "step": 15599
    },
    {
      "entropy": 0.607427385635674,
      "epoch": 1.4352821432884229,
      "grad_norm": 0.15817029774188995,
      "learning_rate": 5.216057901677555e-05,
      "loss": 0.5936,
      "mean_token_accuracy": 0.8227142505347729,
      "num_tokens": 497903281.0,
      "step": 15600
    },
    {
      "entropy": 0.537072230130434,
      "epoch": 1.435374149659864,
      "grad_norm": 0.14723728597164154,
      "learning_rate": 5.2157512190633916e-05,
      "loss": 0.5169,
      "mean_token_accuracy": 0.8365003280341625,
      "num_tokens": 497935696.0,
      "step": 15601
    },
    {
      "entropy": 0.6667176764458418,
      "epoch": 1.4354661560313051,
      "grad_norm": 0.16264256834983826,
      "learning_rate": 5.215444536449229e-05,
      "loss": 0.6649,
      "mean_token_accuracy": 0.7950919978320599,
      "num_tokens": 497966944.0,
      "step": 15602
    },
    {
      "entropy": 0.6629846505820751,
      "epoch": 1.4355581624027463,
      "grad_norm": 0.15894436836242676,
      "learning_rate": 5.215137853835066e-05,
      "loss": 0.6631,
      "mean_token_accuracy": 0.7998712100088596,
      "num_tokens": 497998284.0,
      "step": 15603
    },
    {
      "entropy": 0.6136459661647677,
      "epoch": 1.4356501687741876,
      "grad_norm": 0.1493716537952423,
      "learning_rate": 5.214831171220904e-05,
      "loss": 0.6107,
      "mean_token_accuracy": 0.8129575178027153,
      "num_tokens": 498030498.0,
      "step": 15604
    },
    {
      "entropy": 0.5717803556472063,
      "epoch": 1.435742175145629,
      "grad_norm": 0.16584526002407074,
      "learning_rate": 5.2145244886067414e-05,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8293655663728714,
      "num_tokens": 498063154.0,
      "step": 15605
    },
    {
      "entropy": 0.6127206981182098,
      "epoch": 1.43583418151707,
      "grad_norm": 0.15468193590641022,
      "learning_rate": 5.214217805992578e-05,
      "loss": 0.5892,
      "mean_token_accuracy": 0.818120464682579,
      "num_tokens": 498095202.0,
      "step": 15606
    },
    {
      "entropy": 0.6193830240517855,
      "epoch": 1.4359261878885112,
      "grad_norm": 0.17214757204055786,
      "learning_rate": 5.213911123378416e-05,
      "loss": 0.6115,
      "mean_token_accuracy": 0.8117590509355068,
      "num_tokens": 498126934.0,
      "step": 15607
    },
    {
      "entropy": 0.5804014690220356,
      "epoch": 1.4360181942599524,
      "grad_norm": 0.14645341038703918,
      "learning_rate": 5.213604440764254e-05,
      "loss": 0.565,
      "mean_token_accuracy": 0.8273640125989914,
      "num_tokens": 498159183.0,
      "step": 15608
    },
    {
      "entropy": 0.5925946459174156,
      "epoch": 1.4361102006313937,
      "grad_norm": 0.15128527581691742,
      "learning_rate": 5.2132977581500906e-05,
      "loss": 0.5807,
      "mean_token_accuracy": 0.8240508250892162,
      "num_tokens": 498191646.0,
      "step": 15609
    },
    {
      "entropy": 0.573908006772399,
      "epoch": 1.436202207002835,
      "grad_norm": 0.1495758593082428,
      "learning_rate": 5.212991075535928e-05,
      "loss": 0.552,
      "mean_token_accuracy": 0.8295629508793354,
      "num_tokens": 498224096.0,
      "step": 15610
    },
    {
      "entropy": 0.5113772079348564,
      "epoch": 1.4362942133742762,
      "grad_norm": 0.15001676976680756,
      "learning_rate": 5.212684392921765e-05,
      "loss": 0.4982,
      "mean_token_accuracy": 0.8456969261169434,
      "num_tokens": 498255148.0,
      "step": 15611
    },
    {
      "entropy": 0.7122325096279383,
      "epoch": 1.4363862197457173,
      "grad_norm": 0.1612595170736313,
      "learning_rate": 5.212377710307603e-05,
      "loss": 0.7072,
      "mean_token_accuracy": 0.7859127596020699,
      "num_tokens": 498287600.0,
      "step": 15612
    },
    {
      "entropy": 0.47927866876125336,
      "epoch": 1.4364782261171587,
      "grad_norm": 0.14834220707416534,
      "learning_rate": 5.2120710276934405e-05,
      "loss": 0.4553,
      "mean_token_accuracy": 0.8595594502985477,
      "num_tokens": 498319714.0,
      "step": 15613
    },
    {
      "entropy": 0.6932126814499497,
      "epoch": 1.4365702324885998,
      "grad_norm": 0.15969933569431305,
      "learning_rate": 5.211764345079277e-05,
      "loss": 0.6871,
      "mean_token_accuracy": 0.7911808751523495,
      "num_tokens": 498351686.0,
      "step": 15614
    },
    {
      "entropy": 0.6337010562419891,
      "epoch": 1.4366622388600412,
      "grad_norm": 0.1581137776374817,
      "learning_rate": 5.211457662465115e-05,
      "loss": 0.6262,
      "mean_token_accuracy": 0.8063249848783016,
      "num_tokens": 498383915.0,
      "step": 15615
    },
    {
      "entropy": 0.7065134197473526,
      "epoch": 1.4367542452314823,
      "grad_norm": 0.15847843885421753,
      "learning_rate": 5.211150979850953e-05,
      "loss": 0.6905,
      "mean_token_accuracy": 0.7880505695939064,
      "num_tokens": 498416476.0,
      "step": 15616
    },
    {
      "entropy": 0.6045380509458482,
      "epoch": 1.4368462516029235,
      "grad_norm": 0.15655553340911865,
      "learning_rate": 5.2108442972367896e-05,
      "loss": 0.5989,
      "mean_token_accuracy": 0.8171024285256863,
      "num_tokens": 498448872.0,
      "step": 15617
    },
    {
      "entropy": 0.5760469129309058,
      "epoch": 1.4369382579743648,
      "grad_norm": 0.15749427676200867,
      "learning_rate": 5.210537614622627e-05,
      "loss": 0.56,
      "mean_token_accuracy": 0.8267820104956627,
      "num_tokens": 498481207.0,
      "step": 15618
    },
    {
      "entropy": 0.5986538305878639,
      "epoch": 1.437030264345806,
      "grad_norm": 0.15148203074932098,
      "learning_rate": 5.210230932008465e-05,
      "loss": 0.5913,
      "mean_token_accuracy": 0.8201174661517143,
      "num_tokens": 498513639.0,
      "step": 15619
    },
    {
      "entropy": 0.5622076205909252,
      "epoch": 1.4371222707172473,
      "grad_norm": 0.15906912088394165,
      "learning_rate": 5.209924249394302e-05,
      "loss": 0.5503,
      "mean_token_accuracy": 0.8280416578054428,
      "num_tokens": 498545940.0,
      "step": 15620
    },
    {
      "entropy": 0.5285150874406099,
      "epoch": 1.4372142770886884,
      "grad_norm": 0.16354769468307495,
      "learning_rate": 5.2096175667801395e-05,
      "loss": 0.5276,
      "mean_token_accuracy": 0.8374639749526978,
      "num_tokens": 498578428.0,
      "step": 15621
    },
    {
      "entropy": 0.6004327796399593,
      "epoch": 1.4373062834601296,
      "grad_norm": 0.16100946068763733,
      "learning_rate": 5.209310884165976e-05,
      "loss": 0.5928,
      "mean_token_accuracy": 0.8194555230438709,
      "num_tokens": 498609461.0,
      "step": 15622
    },
    {
      "entropy": 0.6218350359704345,
      "epoch": 1.437398289831571,
      "grad_norm": 0.15006349980831146,
      "learning_rate": 5.209004201551815e-05,
      "loss": 0.6178,
      "mean_token_accuracy": 0.8119430206716061,
      "num_tokens": 498640636.0,
      "step": 15623
    },
    {
      "entropy": 0.5870259872172028,
      "epoch": 1.437490296203012,
      "grad_norm": 0.1511276364326477,
      "learning_rate": 5.208697518937652e-05,
      "loss": 0.5857,
      "mean_token_accuracy": 0.8238219022750854,
      "num_tokens": 498673008.0,
      "step": 15624
    },
    {
      "entropy": 0.6126443799585104,
      "epoch": 1.4375823025744534,
      "grad_norm": 0.15612582862377167,
      "learning_rate": 5.208390836323489e-05,
      "loss": 0.6145,
      "mean_token_accuracy": 0.8123681172728539,
      "num_tokens": 498705159.0,
      "step": 15625
    },
    {
      "entropy": 0.7117244657129049,
      "epoch": 1.4376743089458945,
      "grad_norm": 0.1590295284986496,
      "learning_rate": 5.208084153709326e-05,
      "loss": 0.7139,
      "mean_token_accuracy": 0.781468965113163,
      "num_tokens": 498737275.0,
      "step": 15626
    },
    {
      "entropy": 0.6290026139467955,
      "epoch": 1.4377663153173357,
      "grad_norm": 0.15457434952259064,
      "learning_rate": 5.207777471095164e-05,
      "loss": 0.6208,
      "mean_token_accuracy": 0.808571845293045,
      "num_tokens": 498768828.0,
      "step": 15627
    },
    {
      "entropy": 0.5305772763676941,
      "epoch": 1.437858321688777,
      "grad_norm": 0.1408088654279709,
      "learning_rate": 5.207470788481002e-05,
      "loss": 0.5094,
      "mean_token_accuracy": 0.8436126075685024,
      "num_tokens": 498801596.0,
      "step": 15628
    },
    {
      "entropy": 0.6635271133854985,
      "epoch": 1.4379503280602182,
      "grad_norm": 0.15744806826114655,
      "learning_rate": 5.2071641058668385e-05,
      "loss": 0.6605,
      "mean_token_accuracy": 0.7970437705516815,
      "num_tokens": 498833572.0,
      "step": 15629
    },
    {
      "entropy": 0.6018239539116621,
      "epoch": 1.4380423344316595,
      "grad_norm": 0.15430986881256104,
      "learning_rate": 5.206857423252676e-05,
      "loss": 0.597,
      "mean_token_accuracy": 0.8211268149316311,
      "num_tokens": 498866278.0,
      "step": 15630
    },
    {
      "entropy": 0.5558368414640427,
      "epoch": 1.4381343408031007,
      "grad_norm": 0.15225227177143097,
      "learning_rate": 5.206550740638514e-05,
      "loss": 0.5522,
      "mean_token_accuracy": 0.8282800503075123,
      "num_tokens": 498898735.0,
      "step": 15631
    },
    {
      "entropy": 0.5513075785711408,
      "epoch": 1.4382263471745418,
      "grad_norm": 0.1535380780696869,
      "learning_rate": 5.206244058024351e-05,
      "loss": 0.5432,
      "mean_token_accuracy": 0.8365579321980476,
      "num_tokens": 498930463.0,
      "step": 15632
    },
    {
      "entropy": 0.6824104888364673,
      "epoch": 1.4383183535459831,
      "grad_norm": 0.1611691117286682,
      "learning_rate": 5.2059373754101884e-05,
      "loss": 0.6791,
      "mean_token_accuracy": 0.7958348356187344,
      "num_tokens": 498962215.0,
      "step": 15633
    },
    {
      "entropy": 0.6280222535133362,
      "epoch": 1.4384103599174243,
      "grad_norm": 0.15951068699359894,
      "learning_rate": 5.205630692796025e-05,
      "loss": 0.6021,
      "mean_token_accuracy": 0.8154070153832436,
      "num_tokens": 498994141.0,
      "step": 15634
    },
    {
      "entropy": 0.6276498045772314,
      "epoch": 1.4385023662888656,
      "grad_norm": 0.15399810671806335,
      "learning_rate": 5.205324010181863e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8126625865697861,
      "num_tokens": 499025886.0,
      "step": 15635
    },
    {
      "entropy": 0.6408166605979204,
      "epoch": 1.4385943726603068,
      "grad_norm": 0.1582910567522049,
      "learning_rate": 5.205017327567701e-05,
      "loss": 0.6324,
      "mean_token_accuracy": 0.8077059648931026,
      "num_tokens": 499057712.0,
      "step": 15636
    },
    {
      "entropy": 0.5386843653395772,
      "epoch": 1.438686379031748,
      "grad_norm": 0.15806183218955994,
      "learning_rate": 5.2047106449535376e-05,
      "loss": 0.5293,
      "mean_token_accuracy": 0.8424341753125191,
      "num_tokens": 499089333.0,
      "step": 15637
    },
    {
      "entropy": 0.6753366151824594,
      "epoch": 1.4387783854031893,
      "grad_norm": 0.1607438027858734,
      "learning_rate": 5.204403962339375e-05,
      "loss": 0.6509,
      "mean_token_accuracy": 0.8009168654680252,
      "num_tokens": 499121633.0,
      "step": 15638
    },
    {
      "entropy": 0.48136403877288103,
      "epoch": 1.4388703917746304,
      "grad_norm": 0.1369350552558899,
      "learning_rate": 5.204097279725213e-05,
      "loss": 0.4688,
      "mean_token_accuracy": 0.850258368998766,
      "num_tokens": 499154190.0,
      "step": 15639
    },
    {
      "entropy": 0.6912176217883825,
      "epoch": 1.4389623981460717,
      "grad_norm": 0.15693560242652893,
      "learning_rate": 5.20379059711105e-05,
      "loss": 0.6774,
      "mean_token_accuracy": 0.7921726517379284,
      "num_tokens": 499185704.0,
      "step": 15640
    },
    {
      "entropy": 0.6166213965043426,
      "epoch": 1.4390544045175129,
      "grad_norm": 0.154525026679039,
      "learning_rate": 5.2034839144968874e-05,
      "loss": 0.6149,
      "mean_token_accuracy": 0.8140550032258034,
      "num_tokens": 499217651.0,
      "step": 15641
    },
    {
      "entropy": 0.6142898593097925,
      "epoch": 1.439146410888954,
      "grad_norm": 0.16064471006393433,
      "learning_rate": 5.203177231882724e-05,
      "loss": 0.6046,
      "mean_token_accuracy": 0.8145329914987087,
      "num_tokens": 499249076.0,
      "step": 15642
    },
    {
      "entropy": 0.633676958270371,
      "epoch": 1.4392384172603954,
      "grad_norm": 0.15536576509475708,
      "learning_rate": 5.2028705492685624e-05,
      "loss": 0.6235,
      "mean_token_accuracy": 0.8110143803060055,
      "num_tokens": 499280927.0,
      "step": 15643
    },
    {
      "entropy": 0.5410933922976255,
      "epoch": 1.4393304236318365,
      "grad_norm": 0.1491195410490036,
      "learning_rate": 5.2025638666544e-05,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8372093699872494,
      "num_tokens": 499312750.0,
      "step": 15644
    },
    {
      "entropy": 0.6816350240260363,
      "epoch": 1.4394224300032779,
      "grad_norm": 0.16582629084587097,
      "learning_rate": 5.2022571840402366e-05,
      "loss": 0.6795,
      "mean_token_accuracy": 0.789965994656086,
      "num_tokens": 499344038.0,
      "step": 15645
    },
    {
      "entropy": 0.5513217272236943,
      "epoch": 1.439514436374719,
      "grad_norm": 0.15326489508152008,
      "learning_rate": 5.201950501426074e-05,
      "loss": 0.5319,
      "mean_token_accuracy": 0.8351566046476364,
      "num_tokens": 499376247.0,
      "step": 15646
    },
    {
      "entropy": 0.6722101885825396,
      "epoch": 1.4396064427461601,
      "grad_norm": 0.15858645737171173,
      "learning_rate": 5.201643818811912e-05,
      "loss": 0.6775,
      "mean_token_accuracy": 0.7960621453821659,
      "num_tokens": 499408227.0,
      "step": 15647
    },
    {
      "entropy": 0.6235938398167491,
      "epoch": 1.4396984491176015,
      "grad_norm": 0.15827742218971252,
      "learning_rate": 5.201337136197749e-05,
      "loss": 0.6001,
      "mean_token_accuracy": 0.8170912712812424,
      "num_tokens": 499439677.0,
      "step": 15648
    },
    {
      "entropy": 0.5270544067025185,
      "epoch": 1.4397904554890426,
      "grad_norm": 0.14626461267471313,
      "learning_rate": 5.2010304535835865e-05,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8410940282046795,
      "num_tokens": 499470959.0,
      "step": 15649
    },
    {
      "entropy": 0.660162441432476,
      "epoch": 1.439882461860484,
      "grad_norm": 0.1598125696182251,
      "learning_rate": 5.2007237709694246e-05,
      "loss": 0.6465,
      "mean_token_accuracy": 0.8044916689395905,
      "num_tokens": 499503193.0,
      "step": 15650
    },
    {
      "entropy": 0.6440750639885664,
      "epoch": 1.439974468231925,
      "grad_norm": 0.16225944459438324,
      "learning_rate": 5.2004170883552614e-05,
      "loss": 0.6385,
      "mean_token_accuracy": 0.804326456040144,
      "num_tokens": 499534401.0,
      "step": 15651
    },
    {
      "entropy": 0.633007675409317,
      "epoch": 1.4400664746033662,
      "grad_norm": 0.17360137403011322,
      "learning_rate": 5.200110405741099e-05,
      "loss": 0.6314,
      "mean_token_accuracy": 0.8110081776976585,
      "num_tokens": 499566319.0,
      "step": 15652
    },
    {
      "entropy": 0.5955833252519369,
      "epoch": 1.4401584809748076,
      "grad_norm": 0.15803764760494232,
      "learning_rate": 5.1998037231269357e-05,
      "loss": 0.5875,
      "mean_token_accuracy": 0.819840244948864,
      "num_tokens": 499598138.0,
      "step": 15653
    },
    {
      "entropy": 0.670654121786356,
      "epoch": 1.4402504873462487,
      "grad_norm": 0.1572466492652893,
      "learning_rate": 5.199497040512774e-05,
      "loss": 0.671,
      "mean_token_accuracy": 0.7939980141818523,
      "num_tokens": 499630071.0,
      "step": 15654
    },
    {
      "entropy": 0.6363110542297363,
      "epoch": 1.44034249371769,
      "grad_norm": 0.15920047461986542,
      "learning_rate": 5.199190357898611e-05,
      "loss": 0.6249,
      "mean_token_accuracy": 0.8084778115153313,
      "num_tokens": 499661775.0,
      "step": 15655
    },
    {
      "entropy": 0.6114938799291849,
      "epoch": 1.4404345000891312,
      "grad_norm": 0.16049180924892426,
      "learning_rate": 5.198883675284448e-05,
      "loss": 0.5994,
      "mean_token_accuracy": 0.8150862641632557,
      "num_tokens": 499693434.0,
      "step": 15656
    },
    {
      "entropy": 0.6495343595743179,
      "epoch": 1.4405265064605723,
      "grad_norm": 0.15924540162086487,
      "learning_rate": 5.1985769926702855e-05,
      "loss": 0.6415,
      "mean_token_accuracy": 0.8076490759849548,
      "num_tokens": 499725318.0,
      "step": 15657
    },
    {
      "entropy": 0.6452169604599476,
      "epoch": 1.4406185128320137,
      "grad_norm": 0.15765561163425446,
      "learning_rate": 5.1982703100561237e-05,
      "loss": 0.6363,
      "mean_token_accuracy": 0.8005719967186451,
      "num_tokens": 499757354.0,
      "step": 15658
    },
    {
      "entropy": 0.6893719788640738,
      "epoch": 1.4407105192034548,
      "grad_norm": 0.16859681904315948,
      "learning_rate": 5.1979636274419604e-05,
      "loss": 0.6743,
      "mean_token_accuracy": 0.7941508367657661,
      "num_tokens": 499788946.0,
      "step": 15659
    },
    {
      "entropy": 0.6164061985909939,
      "epoch": 1.4408025255748962,
      "grad_norm": 0.15115857124328613,
      "learning_rate": 5.197656944827798e-05,
      "loss": 0.6049,
      "mean_token_accuracy": 0.811129666864872,
      "num_tokens": 499820997.0,
      "step": 15660
    },
    {
      "entropy": 0.6069641290232539,
      "epoch": 1.4408945319463373,
      "grad_norm": 0.14805473387241364,
      "learning_rate": 5.197350262213635e-05,
      "loss": 0.5849,
      "mean_token_accuracy": 0.8221410661935806,
      "num_tokens": 499853495.0,
      "step": 15661
    },
    {
      "entropy": 0.657615662086755,
      "epoch": 1.4409865383177785,
      "grad_norm": 0.1541614830493927,
      "learning_rate": 5.1970435795994735e-05,
      "loss": 0.6557,
      "mean_token_accuracy": 0.8001722916960716,
      "num_tokens": 499885182.0,
      "step": 15662
    },
    {
      "entropy": 0.6029439140111208,
      "epoch": 1.4410785446892198,
      "grad_norm": 0.15936248004436493,
      "learning_rate": 5.19673689698531e-05,
      "loss": 0.5948,
      "mean_token_accuracy": 0.8187226988375187,
      "num_tokens": 499917659.0,
      "step": 15663
    },
    {
      "entropy": 0.5619012378156185,
      "epoch": 1.441170551060661,
      "grad_norm": 0.15465407073497772,
      "learning_rate": 5.196430214371147e-05,
      "loss": 0.5576,
      "mean_token_accuracy": 0.8304653763771057,
      "num_tokens": 499949549.0,
      "step": 15664
    },
    {
      "entropy": 0.7075056787580252,
      "epoch": 1.4412625574321023,
      "grad_norm": 0.1659955084323883,
      "learning_rate": 5.1961235317569846e-05,
      "loss": 0.7004,
      "mean_token_accuracy": 0.7854507118463516,
      "num_tokens": 499980622.0,
      "step": 15665
    },
    {
      "entropy": 0.7281126379966736,
      "epoch": 1.4413545638035434,
      "grad_norm": 0.16069824993610382,
      "learning_rate": 5.195816849142823e-05,
      "loss": 0.721,
      "mean_token_accuracy": 0.7869280464947224,
      "num_tokens": 500013366.0,
      "step": 15666
    },
    {
      "entropy": 0.5802493337541819,
      "epoch": 1.4414465701749846,
      "grad_norm": 0.15219660103321075,
      "learning_rate": 5.19551016652866e-05,
      "loss": 0.5616,
      "mean_token_accuracy": 0.8281449116766453,
      "num_tokens": 500044940.0,
      "step": 15667
    },
    {
      "entropy": 0.5937228719703853,
      "epoch": 1.441538576546426,
      "grad_norm": 0.15463490784168243,
      "learning_rate": 5.195203483914497e-05,
      "loss": 0.5861,
      "mean_token_accuracy": 0.8226738534867764,
      "num_tokens": 500076614.0,
      "step": 15668
    },
    {
      "entropy": 0.7125667668879032,
      "epoch": 1.441630582917867,
      "grad_norm": 0.17180116474628448,
      "learning_rate": 5.194896801300334e-05,
      "loss": 0.7052,
      "mean_token_accuracy": 0.7857908681035042,
      "num_tokens": 500108073.0,
      "step": 15669
    },
    {
      "entropy": 0.5646152119152248,
      "epoch": 1.4417225892893084,
      "grad_norm": 0.149471715092659,
      "learning_rate": 5.1945901186861726e-05,
      "loss": 0.565,
      "mean_token_accuracy": 0.8272683583199978,
      "num_tokens": 500139999.0,
      "step": 15670
    },
    {
      "entropy": 0.5395527193322778,
      "epoch": 1.4418145956607495,
      "grad_norm": 0.15241573750972748,
      "learning_rate": 5.1942834360720093e-05,
      "loss": 0.531,
      "mean_token_accuracy": 0.8403129503130913,
      "num_tokens": 500171415.0,
      "step": 15671
    },
    {
      "entropy": 0.7055346872657537,
      "epoch": 1.4419066020321907,
      "grad_norm": 0.16150319576263428,
      "learning_rate": 5.193976753457847e-05,
      "loss": 0.6886,
      "mean_token_accuracy": 0.7932313270866871,
      "num_tokens": 500202749.0,
      "step": 15672
    },
    {
      "entropy": 0.506667266599834,
      "epoch": 1.441998608403632,
      "grad_norm": 0.15654256939888,
      "learning_rate": 5.1936700708436836e-05,
      "loss": 0.4978,
      "mean_token_accuracy": 0.8470388166606426,
      "num_tokens": 500234369.0,
      "step": 15673
    },
    {
      "entropy": 0.6589371170848608,
      "epoch": 1.4420906147750732,
      "grad_norm": 0.15969504415988922,
      "learning_rate": 5.193363388229522e-05,
      "loss": 0.6355,
      "mean_token_accuracy": 0.8044284284114838,
      "num_tokens": 500265416.0,
      "step": 15674
    },
    {
      "entropy": 0.621686976402998,
      "epoch": 1.4421826211465145,
      "grad_norm": 0.1575791835784912,
      "learning_rate": 5.193056705615359e-05,
      "loss": 0.5964,
      "mean_token_accuracy": 0.8131720088422298,
      "num_tokens": 500296991.0,
      "step": 15675
    },
    {
      "entropy": 0.5439040521159768,
      "epoch": 1.4422746275179557,
      "grad_norm": 0.15401700139045715,
      "learning_rate": 5.192750023001196e-05,
      "loss": 0.5383,
      "mean_token_accuracy": 0.834080196917057,
      "num_tokens": 500329108.0,
      "step": 15676
    },
    {
      "entropy": 0.5053191110491753,
      "epoch": 1.4423666338893968,
      "grad_norm": 0.14923827350139618,
      "learning_rate": 5.1924433403870335e-05,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8504025638103485,
      "num_tokens": 500360645.0,
      "step": 15677
    },
    {
      "entropy": 0.542908578645438,
      "epoch": 1.4424586402608381,
      "grad_norm": 0.14979174733161926,
      "learning_rate": 5.1921366577728716e-05,
      "loss": 0.5334,
      "mean_token_accuracy": 0.8348880186676979,
      "num_tokens": 500392794.0,
      "step": 15678
    },
    {
      "entropy": 0.5896643046289682,
      "epoch": 1.4425506466322793,
      "grad_norm": 0.15678080916404724,
      "learning_rate": 5.1918299751587084e-05,
      "loss": 0.5722,
      "mean_token_accuracy": 0.8214852064847946,
      "num_tokens": 500424521.0,
      "step": 15679
    },
    {
      "entropy": 0.5667223846539855,
      "epoch": 1.4426426530037206,
      "grad_norm": 0.16164043545722961,
      "learning_rate": 5.191523292544546e-05,
      "loss": 0.5616,
      "mean_token_accuracy": 0.8284224234521389,
      "num_tokens": 500455935.0,
      "step": 15680
    },
    {
      "entropy": 0.7139086090028286,
      "epoch": 1.4427346593751618,
      "grad_norm": 0.16793544590473175,
      "learning_rate": 5.191216609930384e-05,
      "loss": 0.7132,
      "mean_token_accuracy": 0.7814307734370232,
      "num_tokens": 500488341.0,
      "step": 15681
    },
    {
      "entropy": 0.7078898642212152,
      "epoch": 1.442826665746603,
      "grad_norm": 0.1633855104446411,
      "learning_rate": 5.190909927316221e-05,
      "loss": 0.7092,
      "mean_token_accuracy": 0.7862892672419548,
      "num_tokens": 500519595.0,
      "step": 15682
    },
    {
      "entropy": 0.6344558410346508,
      "epoch": 1.4429186721180443,
      "grad_norm": 0.15467305481433868,
      "learning_rate": 5.190603244702058e-05,
      "loss": 0.6032,
      "mean_token_accuracy": 0.813832513988018,
      "num_tokens": 500551701.0,
      "step": 15683
    },
    {
      "entropy": 0.6606058049947023,
      "epoch": 1.4430106784894854,
      "grad_norm": 0.16060973703861237,
      "learning_rate": 5.190296562087895e-05,
      "loss": 0.658,
      "mean_token_accuracy": 0.8024740070104599,
      "num_tokens": 500583365.0,
      "step": 15684
    },
    {
      "entropy": 0.5787179060280323,
      "epoch": 1.4431026848609267,
      "grad_norm": 0.15332470834255219,
      "learning_rate": 5.189989879473733e-05,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8229568190872669,
      "num_tokens": 500614001.0,
      "step": 15685
    },
    {
      "entropy": 0.5590316173620522,
      "epoch": 1.4431946912323679,
      "grad_norm": 0.15578514337539673,
      "learning_rate": 5.1896831968595706e-05,
      "loss": 0.5543,
      "mean_token_accuracy": 0.8331095278263092,
      "num_tokens": 500646504.0,
      "step": 15686
    },
    {
      "entropy": 0.5997512266039848,
      "epoch": 1.443286697603809,
      "grad_norm": 0.1598382592201233,
      "learning_rate": 5.1893765142454074e-05,
      "loss": 0.5933,
      "mean_token_accuracy": 0.8181789591908455,
      "num_tokens": 500678655.0,
      "step": 15687
    },
    {
      "entropy": 0.5268417927436531,
      "epoch": 1.4433787039752504,
      "grad_norm": 0.14876608550548553,
      "learning_rate": 5.189069831631245e-05,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8403434529900551,
      "num_tokens": 500710237.0,
      "step": 15688
    },
    {
      "entropy": 0.627172164618969,
      "epoch": 1.4434707103466915,
      "grad_norm": 0.16137829422950745,
      "learning_rate": 5.188763149017083e-05,
      "loss": 0.6227,
      "mean_token_accuracy": 0.8083162307739258,
      "num_tokens": 500741780.0,
      "step": 15689
    },
    {
      "entropy": 0.5982195353135467,
      "epoch": 1.4435627167181329,
      "grad_norm": 0.15245676040649414,
      "learning_rate": 5.18845646640292e-05,
      "loss": 0.5893,
      "mean_token_accuracy": 0.8184335418045521,
      "num_tokens": 500774063.0,
      "step": 15690
    },
    {
      "entropy": 0.6574294995516539,
      "epoch": 1.443654723089574,
      "grad_norm": 0.15811751782894135,
      "learning_rate": 5.188149783788757e-05,
      "loss": 0.6474,
      "mean_token_accuracy": 0.7984780259430408,
      "num_tokens": 500805380.0,
      "step": 15691
    },
    {
      "entropy": 0.6478750398382545,
      "epoch": 1.4437467294610151,
      "grad_norm": 0.15317970514297485,
      "learning_rate": 5.187843101174594e-05,
      "loss": 0.6328,
      "mean_token_accuracy": 0.8042971678078175,
      "num_tokens": 500836637.0,
      "step": 15692
    },
    {
      "entropy": 0.6450932249426842,
      "epoch": 1.4438387358324565,
      "grad_norm": 0.1551068276166916,
      "learning_rate": 5.187536418560432e-05,
      "loss": 0.6268,
      "mean_token_accuracy": 0.8050121963024139,
      "num_tokens": 500868177.0,
      "step": 15693
    },
    {
      "entropy": 0.67843665368855,
      "epoch": 1.4439307422038976,
      "grad_norm": 0.16133606433868408,
      "learning_rate": 5.18722973594627e-05,
      "loss": 0.6789,
      "mean_token_accuracy": 0.7948955930769444,
      "num_tokens": 500900296.0,
      "step": 15694
    },
    {
      "entropy": 0.5515557583421469,
      "epoch": 1.444022748575339,
      "grad_norm": 0.15010158717632294,
      "learning_rate": 5.1869230533321065e-05,
      "loss": 0.5438,
      "mean_token_accuracy": 0.8340113088488579,
      "num_tokens": 500932086.0,
      "step": 15695
    },
    {
      "entropy": 0.5964016942307353,
      "epoch": 1.44411475494678,
      "grad_norm": 0.15995769202709198,
      "learning_rate": 5.186616370717944e-05,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8195374235510826,
      "num_tokens": 500964354.0,
      "step": 15696
    },
    {
      "entropy": 0.5885719694197178,
      "epoch": 1.4442067613182212,
      "grad_norm": 0.1548122614622116,
      "learning_rate": 5.186309688103782e-05,
      "loss": 0.5878,
      "mean_token_accuracy": 0.8220949470996857,
      "num_tokens": 500996841.0,
      "step": 15697
    },
    {
      "entropy": 0.6225017569959164,
      "epoch": 1.4442987676896626,
      "grad_norm": 0.15708167850971222,
      "learning_rate": 5.186003005489619e-05,
      "loss": 0.6155,
      "mean_token_accuracy": 0.8116728737950325,
      "num_tokens": 501028934.0,
      "step": 15698
    },
    {
      "entropy": 0.6768473535776138,
      "epoch": 1.4443907740611037,
      "grad_norm": 0.16341355443000793,
      "learning_rate": 5.185696322875456e-05,
      "loss": 0.6765,
      "mean_token_accuracy": 0.7937556393444538,
      "num_tokens": 501061252.0,
      "step": 15699
    },
    {
      "entropy": 0.5597397582605481,
      "epoch": 1.444482780432545,
      "grad_norm": 0.15325109660625458,
      "learning_rate": 5.185389640261293e-05,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8297169916331768,
      "num_tokens": 501092981.0,
      "step": 15700
    },
    {
      "entropy": 0.75951929949224,
      "epoch": 1.4445747868039862,
      "grad_norm": 0.17490188777446747,
      "learning_rate": 5.185082957647131e-05,
      "loss": 0.7318,
      "mean_token_accuracy": 0.7780756056308746,
      "num_tokens": 501124795.0,
      "step": 15701
    },
    {
      "entropy": 0.6506090806797147,
      "epoch": 1.4446667931754273,
      "grad_norm": 0.1629764884710312,
      "learning_rate": 5.184776275032969e-05,
      "loss": 0.6462,
      "mean_token_accuracy": 0.7988641783595085,
      "num_tokens": 501155876.0,
      "step": 15702
    },
    {
      "entropy": 0.5752848517149687,
      "epoch": 1.4447587995468687,
      "grad_norm": 0.14675134420394897,
      "learning_rate": 5.1844695924188055e-05,
      "loss": 0.5602,
      "mean_token_accuracy": 0.8260462954640388,
      "num_tokens": 501188510.0,
      "step": 15703
    },
    {
      "entropy": 0.5586128868162632,
      "epoch": 1.4448508059183098,
      "grad_norm": 0.1524234116077423,
      "learning_rate": 5.184162909804643e-05,
      "loss": 0.5503,
      "mean_token_accuracy": 0.8324042819440365,
      "num_tokens": 501220496.0,
      "step": 15704
    },
    {
      "entropy": 0.5406489460729063,
      "epoch": 1.4449428122897512,
      "grad_norm": 0.1487109512090683,
      "learning_rate": 5.183856227190481e-05,
      "loss": 0.532,
      "mean_token_accuracy": 0.8422480821609497,
      "num_tokens": 501252386.0,
      "step": 15705
    },
    {
      "entropy": 0.5842731688171625,
      "epoch": 1.4450348186611923,
      "grad_norm": 0.1572112739086151,
      "learning_rate": 5.1835495445763186e-05,
      "loss": 0.5723,
      "mean_token_accuracy": 0.8251101784408092,
      "num_tokens": 501284567.0,
      "step": 15706
    },
    {
      "entropy": 0.7237949445843697,
      "epoch": 1.4451268250326335,
      "grad_norm": 0.16169941425323486,
      "learning_rate": 5.1832428619621554e-05,
      "loss": 0.7013,
      "mean_token_accuracy": 0.7836882472038269,
      "num_tokens": 501316866.0,
      "step": 15707
    },
    {
      "entropy": 0.7200224567204714,
      "epoch": 1.4452188314040748,
      "grad_norm": 0.1642746478319168,
      "learning_rate": 5.182936179347992e-05,
      "loss": 0.6945,
      "mean_token_accuracy": 0.7917207814753056,
      "num_tokens": 501348033.0,
      "step": 15708
    },
    {
      "entropy": 0.5855368925258517,
      "epoch": 1.445310837775516,
      "grad_norm": 0.1566953808069229,
      "learning_rate": 5.182629496733831e-05,
      "loss": 0.5719,
      "mean_token_accuracy": 0.8275202624499798,
      "num_tokens": 501380340.0,
      "step": 15709
    },
    {
      "entropy": 0.6179522159509361,
      "epoch": 1.4454028441469573,
      "grad_norm": 0.14940103888511658,
      "learning_rate": 5.182322814119668e-05,
      "loss": 0.5961,
      "mean_token_accuracy": 0.8197749815881252,
      "num_tokens": 501411842.0,
      "step": 15710
    },
    {
      "entropy": 0.6687678238376975,
      "epoch": 1.4454948505183984,
      "grad_norm": 0.15976449847221375,
      "learning_rate": 5.182016131505505e-05,
      "loss": 0.674,
      "mean_token_accuracy": 0.7948578894138336,
      "num_tokens": 501443649.0,
      "step": 15711
    },
    {
      "entropy": 0.5607018731534481,
      "epoch": 1.4455868568898396,
      "grad_norm": 0.15983925759792328,
      "learning_rate": 5.1817094488913434e-05,
      "loss": 0.5429,
      "mean_token_accuracy": 0.8292003870010376,
      "num_tokens": 501475240.0,
      "step": 15712
    },
    {
      "entropy": 0.6080004312098026,
      "epoch": 1.445678863261281,
      "grad_norm": 0.15898041427135468,
      "learning_rate": 5.18140276627718e-05,
      "loss": 0.5868,
      "mean_token_accuracy": 0.8196976743638515,
      "num_tokens": 501507069.0,
      "step": 15713
    },
    {
      "entropy": 0.5558412238024175,
      "epoch": 1.445770869632722,
      "grad_norm": 0.1624632030725479,
      "learning_rate": 5.1810960836630176e-05,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8318880386650562,
      "num_tokens": 501538504.0,
      "step": 15714
    },
    {
      "entropy": 0.6510980539023876,
      "epoch": 1.4458628760041634,
      "grad_norm": 0.15881188213825226,
      "learning_rate": 5.1807894010488544e-05,
      "loss": 0.6447,
      "mean_token_accuracy": 0.8032910116016865,
      "num_tokens": 501570094.0,
      "step": 15715
    },
    {
      "entropy": 0.6762349978089333,
      "epoch": 1.4459548823756045,
      "grad_norm": 0.16690441966056824,
      "learning_rate": 5.1804827184346925e-05,
      "loss": 0.6657,
      "mean_token_accuracy": 0.7950247377157211,
      "num_tokens": 501600341.0,
      "step": 15716
    },
    {
      "entropy": 0.6388398418202996,
      "epoch": 1.4460468887470457,
      "grad_norm": 0.15627124905586243,
      "learning_rate": 5.18017603582053e-05,
      "loss": 0.6228,
      "mean_token_accuracy": 0.808622270822525,
      "num_tokens": 501631398.0,
      "step": 15717
    },
    {
      "entropy": 0.7236428689211607,
      "epoch": 1.446138895118487,
      "grad_norm": 0.16691288352012634,
      "learning_rate": 5.179869353206367e-05,
      "loss": 0.7151,
      "mean_token_accuracy": 0.7831502705812454,
      "num_tokens": 501663862.0,
      "step": 15718
    },
    {
      "entropy": 0.5912930369377136,
      "epoch": 1.4462309014899282,
      "grad_norm": 0.15332618355751038,
      "learning_rate": 5.179562670592204e-05,
      "loss": 0.5751,
      "mean_token_accuracy": 0.821050975471735,
      "num_tokens": 501695604.0,
      "step": 15719
    },
    {
      "entropy": 0.6732875965535641,
      "epoch": 1.4463229078613695,
      "grad_norm": 0.16260212659835815,
      "learning_rate": 5.1792559879780424e-05,
      "loss": 0.6658,
      "mean_token_accuracy": 0.7954854369163513,
      "num_tokens": 501727536.0,
      "step": 15720
    },
    {
      "entropy": 0.6463106740266085,
      "epoch": 1.4464149142328107,
      "grad_norm": 0.15724311769008636,
      "learning_rate": 5.178949305363879e-05,
      "loss": 0.6433,
      "mean_token_accuracy": 0.8019617535173893,
      "num_tokens": 501759621.0,
      "step": 15721
    },
    {
      "entropy": 0.5950822131708264,
      "epoch": 1.4465069206042518,
      "grad_norm": 0.15180577337741852,
      "learning_rate": 5.1786426227497167e-05,
      "loss": 0.5786,
      "mean_token_accuracy": 0.819844264537096,
      "num_tokens": 501792092.0,
      "step": 15722
    },
    {
      "entropy": 0.6712682452052832,
      "epoch": 1.4465989269756931,
      "grad_norm": 0.16107553243637085,
      "learning_rate": 5.1783359401355534e-05,
      "loss": 0.6887,
      "mean_token_accuracy": 0.7896644100546837,
      "num_tokens": 501824176.0,
      "step": 15723
    },
    {
      "entropy": 0.724453741684556,
      "epoch": 1.4466909333471343,
      "grad_norm": 0.16721071302890778,
      "learning_rate": 5.1780292575213916e-05,
      "loss": 0.7276,
      "mean_token_accuracy": 0.7834668383002281,
      "num_tokens": 501856283.0,
      "step": 15724
    },
    {
      "entropy": 0.5421556863002479,
      "epoch": 1.4467829397185756,
      "grad_norm": 0.14902029931545258,
      "learning_rate": 5.177722574907229e-05,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8352654315531254,
      "num_tokens": 501887875.0,
      "step": 15725
    },
    {
      "entropy": 0.5834228973835707,
      "epoch": 1.4468749460900168,
      "grad_norm": 0.15465106070041656,
      "learning_rate": 5.177415892293066e-05,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8266318924725056,
      "num_tokens": 501918863.0,
      "step": 15726
    },
    {
      "entropy": 0.6530616693198681,
      "epoch": 1.446966952461458,
      "grad_norm": 0.15362121164798737,
      "learning_rate": 5.177109209678903e-05,
      "loss": 0.6372,
      "mean_token_accuracy": 0.8030346892774105,
      "num_tokens": 501951108.0,
      "step": 15727
    },
    {
      "entropy": 0.5966401076875627,
      "epoch": 1.4470589588328993,
      "grad_norm": 0.15059727430343628,
      "learning_rate": 5.1768025270647414e-05,
      "loss": 0.5795,
      "mean_token_accuracy": 0.8237702995538712,
      "num_tokens": 501983763.0,
      "step": 15728
    },
    {
      "entropy": 0.6208657566457987,
      "epoch": 1.4471509652043404,
      "grad_norm": 0.155916228890419,
      "learning_rate": 5.176495844450578e-05,
      "loss": 0.6081,
      "mean_token_accuracy": 0.8114113733172417,
      "num_tokens": 502015864.0,
      "step": 15729
    },
    {
      "entropy": 0.6423929817974567,
      "epoch": 1.4472429715757817,
      "grad_norm": 0.15912939608097076,
      "learning_rate": 5.176189161836416e-05,
      "loss": 0.6295,
      "mean_token_accuracy": 0.802539199590683,
      "num_tokens": 502046670.0,
      "step": 15730
    },
    {
      "entropy": 0.5989354960620403,
      "epoch": 1.4473349779472229,
      "grad_norm": 0.163077250123024,
      "learning_rate": 5.1758824792222525e-05,
      "loss": 0.5945,
      "mean_token_accuracy": 0.8157868422567844,
      "num_tokens": 502077485.0,
      "step": 15731
    },
    {
      "entropy": 0.5545177264139056,
      "epoch": 1.447426984318664,
      "grad_norm": 0.15527501702308655,
      "learning_rate": 5.1755757966080906e-05,
      "loss": 0.5432,
      "mean_token_accuracy": 0.8328269980847836,
      "num_tokens": 502109715.0,
      "step": 15732
    },
    {
      "entropy": 0.5626667905598879,
      "epoch": 1.4475189906901054,
      "grad_norm": 0.14996790885925293,
      "learning_rate": 5.175269113993928e-05,
      "loss": 0.5395,
      "mean_token_accuracy": 0.8322620950639248,
      "num_tokens": 502141609.0,
      "step": 15733
    },
    {
      "entropy": 0.6319514755159616,
      "epoch": 1.4476109970615465,
      "grad_norm": 0.15822233259677887,
      "learning_rate": 5.174962431379765e-05,
      "loss": 0.6384,
      "mean_token_accuracy": 0.8020589835941792,
      "num_tokens": 502173128.0,
      "step": 15734
    },
    {
      "entropy": 0.6258345600217581,
      "epoch": 1.4477030034329879,
      "grad_norm": 0.15542520582675934,
      "learning_rate": 5.1746557487656023e-05,
      "loss": 0.6209,
      "mean_token_accuracy": 0.8050011806190014,
      "num_tokens": 502205148.0,
      "step": 15735
    },
    {
      "entropy": 0.6134357172995806,
      "epoch": 1.447795009804429,
      "grad_norm": 0.1577172875404358,
      "learning_rate": 5.1743490661514405e-05,
      "loss": 0.5952,
      "mean_token_accuracy": 0.8140515238046646,
      "num_tokens": 502237352.0,
      "step": 15736
    },
    {
      "entropy": 0.631394550204277,
      "epoch": 1.4478870161758701,
      "grad_norm": 0.15193916857242584,
      "learning_rate": 5.174042383537277e-05,
      "loss": 0.6251,
      "mean_token_accuracy": 0.8061434552073479,
      "num_tokens": 502270010.0,
      "step": 15737
    },
    {
      "entropy": 0.6849610470235348,
      "epoch": 1.4479790225473115,
      "grad_norm": 6.822237014770508,
      "learning_rate": 5.173735700923115e-05,
      "loss": 0.6616,
      "mean_token_accuracy": 0.8010885324329138,
      "num_tokens": 502301646.0,
      "step": 15738
    },
    {
      "entropy": 0.7273469157516956,
      "epoch": 1.4480710289187526,
      "grad_norm": 0.16405382752418518,
      "learning_rate": 5.1734290183089515e-05,
      "loss": 0.7187,
      "mean_token_accuracy": 0.781328771263361,
      "num_tokens": 502333893.0,
      "step": 15739
    },
    {
      "entropy": 0.5894296355545521,
      "epoch": 1.448163035290194,
      "grad_norm": 0.1522318422794342,
      "learning_rate": 5.17312233569479e-05,
      "loss": 0.5798,
      "mean_token_accuracy": 0.8220011964440346,
      "num_tokens": 502366228.0,
      "step": 15740
    },
    {
      "entropy": 0.5755199268460274,
      "epoch": 1.448255041661635,
      "grad_norm": 0.15365225076675415,
      "learning_rate": 5.172815653080627e-05,
      "loss": 0.5722,
      "mean_token_accuracy": 0.8244568109512329,
      "num_tokens": 502398269.0,
      "step": 15741
    },
    {
      "entropy": 0.5454120337963104,
      "epoch": 1.4483470480330762,
      "grad_norm": 0.15237192809581757,
      "learning_rate": 5.172508970466464e-05,
      "loss": 0.5325,
      "mean_token_accuracy": 0.8365891836583614,
      "num_tokens": 502429979.0,
      "step": 15742
    },
    {
      "entropy": 0.6565990950912237,
      "epoch": 1.4484390544045176,
      "grad_norm": 0.16289488971233368,
      "learning_rate": 5.172202287852303e-05,
      "loss": 0.6443,
      "mean_token_accuracy": 0.8050572276115417,
      "num_tokens": 502462021.0,
      "step": 15743
    },
    {
      "entropy": 0.6506851240992546,
      "epoch": 1.4485310607759587,
      "grad_norm": 0.16085833311080933,
      "learning_rate": 5.1718956052381395e-05,
      "loss": 0.6327,
      "mean_token_accuracy": 0.8048226423561573,
      "num_tokens": 502494043.0,
      "step": 15744
    },
    {
      "entropy": 0.5594254890456796,
      "epoch": 1.4486230671474,
      "grad_norm": 0.1511635184288025,
      "learning_rate": 5.171588922623976e-05,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8292663656175137,
      "num_tokens": 502526098.0,
      "step": 15745
    },
    {
      "entropy": 0.625136480666697,
      "epoch": 1.4487150735188412,
      "grad_norm": 0.15686990320682526,
      "learning_rate": 5.171282240009814e-05,
      "loss": 0.6184,
      "mean_token_accuracy": 0.8118405416607857,
      "num_tokens": 502557546.0,
      "step": 15746
    },
    {
      "entropy": 0.5279641300439835,
      "epoch": 1.4488070798902823,
      "grad_norm": 0.14943793416023254,
      "learning_rate": 5.170975557395652e-05,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8466445915400982,
      "num_tokens": 502589305.0,
      "step": 15747
    },
    {
      "entropy": 0.6482254769653082,
      "epoch": 1.4488990862617237,
      "grad_norm": 0.15765586495399475,
      "learning_rate": 5.1706688747814894e-05,
      "loss": 0.635,
      "mean_token_accuracy": 0.8047546036541462,
      "num_tokens": 502620693.0,
      "step": 15748
    },
    {
      "entropy": 0.5908115785568953,
      "epoch": 1.4489910926331648,
      "grad_norm": 0.15644146502017975,
      "learning_rate": 5.170362192167326e-05,
      "loss": 0.5772,
      "mean_token_accuracy": 0.823126781731844,
      "num_tokens": 502652308.0,
      "step": 15749
    },
    {
      "entropy": 0.6606712490320206,
      "epoch": 1.4490830990046062,
      "grad_norm": 0.16032136976718903,
      "learning_rate": 5.1700555095531636e-05,
      "loss": 0.6423,
      "mean_token_accuracy": 0.8045300133526325,
      "num_tokens": 502684497.0,
      "step": 15750
    },
    {
      "entropy": 0.6888804286718369,
      "epoch": 1.4491751053760473,
      "grad_norm": 0.16843238472938538,
      "learning_rate": 5.169748826939002e-05,
      "loss": 0.6873,
      "mean_token_accuracy": 0.7921571731567383,
      "num_tokens": 502717057.0,
      "step": 15751
    },
    {
      "entropy": 0.7746430356055498,
      "epoch": 1.4492671117474885,
      "grad_norm": 0.17095544934272766,
      "learning_rate": 5.1694421443248386e-05,
      "loss": 0.7781,
      "mean_token_accuracy": 0.7691900320351124,
      "num_tokens": 502748606.0,
      "step": 15752
    },
    {
      "entropy": 0.5973783135414124,
      "epoch": 1.4493591181189298,
      "grad_norm": 0.15823951363563538,
      "learning_rate": 5.169135461710676e-05,
      "loss": 0.594,
      "mean_token_accuracy": 0.8160071484744549,
      "num_tokens": 502780880.0,
      "step": 15753
    },
    {
      "entropy": 0.6544720437377691,
      "epoch": 1.449451124490371,
      "grad_norm": 0.1647266298532486,
      "learning_rate": 5.168828779096513e-05,
      "loss": 0.6517,
      "mean_token_accuracy": 0.7995687611401081,
      "num_tokens": 502812134.0,
      "step": 15754
    },
    {
      "entropy": 0.6002061339095235,
      "epoch": 1.4495431308618123,
      "grad_norm": 0.15344776213169098,
      "learning_rate": 5.168522096482351e-05,
      "loss": 0.594,
      "mean_token_accuracy": 0.8184713535010815,
      "num_tokens": 502843997.0,
      "step": 15755
    },
    {
      "entropy": 0.5463123098015785,
      "epoch": 1.4496351372332534,
      "grad_norm": 0.1492263227701187,
      "learning_rate": 5.1682154138681884e-05,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8375916630029678,
      "num_tokens": 502875905.0,
      "step": 15756
    },
    {
      "entropy": 0.6721840798854828,
      "epoch": 1.4497271436046946,
      "grad_norm": 0.16254417598247528,
      "learning_rate": 5.167908731254025e-05,
      "loss": 0.6707,
      "mean_token_accuracy": 0.7957402467727661,
      "num_tokens": 502907889.0,
      "step": 15757
    },
    {
      "entropy": 0.6671585515141487,
      "epoch": 1.449819149976136,
      "grad_norm": 0.16031818091869354,
      "learning_rate": 5.167602048639863e-05,
      "loss": 0.6556,
      "mean_token_accuracy": 0.7969964258372784,
      "num_tokens": 502939485.0,
      "step": 15758
    },
    {
      "entropy": 0.6003734888508916,
      "epoch": 1.449911156347577,
      "grad_norm": 0.1489659994840622,
      "learning_rate": 5.167295366025701e-05,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8264736980199814,
      "num_tokens": 502970627.0,
      "step": 15759
    },
    {
      "entropy": 0.5413641696795821,
      "epoch": 1.4500031627190184,
      "grad_norm": 0.14658091962337494,
      "learning_rate": 5.1669886834115376e-05,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8305419497191906,
      "num_tokens": 503002653.0,
      "step": 15760
    },
    {
      "entropy": 0.579532706644386,
      "epoch": 1.4500951690904595,
      "grad_norm": 0.16481827199459076,
      "learning_rate": 5.166682000797375e-05,
      "loss": 0.5578,
      "mean_token_accuracy": 0.827094029635191,
      "num_tokens": 503034954.0,
      "step": 15761
    },
    {
      "entropy": 0.5766190933063626,
      "epoch": 1.4501871754619007,
      "grad_norm": 0.15493981540203094,
      "learning_rate": 5.166375318183212e-05,
      "loss": 0.5588,
      "mean_token_accuracy": 0.8231207765638828,
      "num_tokens": 503066845.0,
      "step": 15762
    },
    {
      "entropy": 0.7131756898015738,
      "epoch": 1.450279181833342,
      "grad_norm": 0.1655331254005432,
      "learning_rate": 5.16606863556905e-05,
      "loss": 0.7013,
      "mean_token_accuracy": 0.789719246327877,
      "num_tokens": 503098222.0,
      "step": 15763
    },
    {
      "entropy": 0.599360017105937,
      "epoch": 1.4503711882047832,
      "grad_norm": 0.1474238634109497,
      "learning_rate": 5.1657619529548875e-05,
      "loss": 0.5924,
      "mean_token_accuracy": 0.8199055157601833,
      "num_tokens": 503130691.0,
      "step": 15764
    },
    {
      "entropy": 0.6825484922155738,
      "epoch": 1.4504631945762245,
      "grad_norm": 0.16421030461788177,
      "learning_rate": 5.165455270340724e-05,
      "loss": 0.6689,
      "mean_token_accuracy": 0.796640183776617,
      "num_tokens": 503163344.0,
      "step": 15765
    },
    {
      "entropy": 0.5503674065694213,
      "epoch": 1.4505552009476657,
      "grad_norm": 0.16059468686580658,
      "learning_rate": 5.165148587726562e-05,
      "loss": 0.5357,
      "mean_token_accuracy": 0.8366094119846821,
      "num_tokens": 503194882.0,
      "step": 15766
    },
    {
      "entropy": 0.6575087942183018,
      "epoch": 1.4506472073191068,
      "grad_norm": 0.1651662141084671,
      "learning_rate": 5.1648419051124e-05,
      "loss": 0.6488,
      "mean_token_accuracy": 0.8007034100592136,
      "num_tokens": 503227200.0,
      "step": 15767
    },
    {
      "entropy": 0.6606774404644966,
      "epoch": 1.4507392136905481,
      "grad_norm": 0.1598503738641739,
      "learning_rate": 5.1645352224982367e-05,
      "loss": 0.6551,
      "mean_token_accuracy": 0.8024585098028183,
      "num_tokens": 503258954.0,
      "step": 15768
    },
    {
      "entropy": 0.5625569429248571,
      "epoch": 1.4508312200619893,
      "grad_norm": 0.15411381423473358,
      "learning_rate": 5.164228539884074e-05,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8294287100434303,
      "num_tokens": 503290776.0,
      "step": 15769
    },
    {
      "entropy": 0.642357999458909,
      "epoch": 1.4509232264334306,
      "grad_norm": 0.1605043113231659,
      "learning_rate": 5.163921857269911e-05,
      "loss": 0.6285,
      "mean_token_accuracy": 0.8102966398000717,
      "num_tokens": 503323349.0,
      "step": 15770
    },
    {
      "entropy": 0.6702718427404761,
      "epoch": 1.4510152328048718,
      "grad_norm": 0.16130128502845764,
      "learning_rate": 5.163615174655749e-05,
      "loss": 0.6576,
      "mean_token_accuracy": 0.7969168797135353,
      "num_tokens": 503354790.0,
      "step": 15771
    },
    {
      "entropy": 0.6732852403074503,
      "epoch": 1.451107239176313,
      "grad_norm": 0.16453315317630768,
      "learning_rate": 5.1633084920415865e-05,
      "loss": 0.6593,
      "mean_token_accuracy": 0.7982374653220177,
      "num_tokens": 503386650.0,
      "step": 15772
    },
    {
      "entropy": 0.6218149904161692,
      "epoch": 1.4511992455477543,
      "grad_norm": 0.16084541380405426,
      "learning_rate": 5.163001809427423e-05,
      "loss": 0.609,
      "mean_token_accuracy": 0.8121970444917679,
      "num_tokens": 503418650.0,
      "step": 15773
    },
    {
      "entropy": 0.7611789051443338,
      "epoch": 1.4512912519191954,
      "grad_norm": 0.16716627776622772,
      "learning_rate": 5.1626951268132614e-05,
      "loss": 0.7671,
      "mean_token_accuracy": 0.7684641852974892,
      "num_tokens": 503450046.0,
      "step": 15774
    },
    {
      "entropy": 0.5615000538527966,
      "epoch": 1.4513832582906367,
      "grad_norm": 0.14921335875988007,
      "learning_rate": 5.162388444199099e-05,
      "loss": 0.55,
      "mean_token_accuracy": 0.8258992470800877,
      "num_tokens": 503482056.0,
      "step": 15775
    },
    {
      "entropy": 0.5529997097328305,
      "epoch": 1.4514752646620779,
      "grad_norm": 0.16251875460147858,
      "learning_rate": 5.162081761584936e-05,
      "loss": 0.5326,
      "mean_token_accuracy": 0.8344514258205891,
      "num_tokens": 503513757.0,
      "step": 15776
    },
    {
      "entropy": 0.551959065720439,
      "epoch": 1.451567271033519,
      "grad_norm": 0.15023227035999298,
      "learning_rate": 5.161775078970773e-05,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8327785842120647,
      "num_tokens": 503544749.0,
      "step": 15777
    },
    {
      "entropy": 0.7472381414845586,
      "epoch": 1.4516592774049604,
      "grad_norm": 0.16714227199554443,
      "learning_rate": 5.161468396356611e-05,
      "loss": 0.7315,
      "mean_token_accuracy": 0.7807230353355408,
      "num_tokens": 503576672.0,
      "step": 15778
    },
    {
      "entropy": 0.5037945955991745,
      "epoch": 1.4517512837764015,
      "grad_norm": 0.145772323012352,
      "learning_rate": 5.161161713742448e-05,
      "loss": 0.4913,
      "mean_token_accuracy": 0.8458196744322777,
      "num_tokens": 503608779.0,
      "step": 15779
    },
    {
      "entropy": 0.7220970178022981,
      "epoch": 1.4518432901478429,
      "grad_norm": 0.1616988629102707,
      "learning_rate": 5.1608550311282856e-05,
      "loss": 0.719,
      "mean_token_accuracy": 0.7842064686119556,
      "num_tokens": 503640970.0,
      "step": 15780
    },
    {
      "entropy": 0.5943058654665947,
      "epoch": 1.451935296519284,
      "grad_norm": 0.15329048037528992,
      "learning_rate": 5.160548348514122e-05,
      "loss": 0.5887,
      "mean_token_accuracy": 0.8155795261263847,
      "num_tokens": 503673443.0,
      "step": 15781
    },
    {
      "entropy": 0.5263654887676239,
      "epoch": 1.4520273028907251,
      "grad_norm": 0.1561894565820694,
      "learning_rate": 5.160241665899961e-05,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8429156020283699,
      "num_tokens": 503705448.0,
      "step": 15782
    },
    {
      "entropy": 0.5677182897925377,
      "epoch": 1.4521193092621665,
      "grad_norm": 0.15173101425170898,
      "learning_rate": 5.159934983285798e-05,
      "loss": 0.5604,
      "mean_token_accuracy": 0.82490274310112,
      "num_tokens": 503737236.0,
      "step": 15783
    },
    {
      "entropy": 0.5604721121490002,
      "epoch": 1.4522113156336076,
      "grad_norm": 0.15027062594890594,
      "learning_rate": 5.159628300671635e-05,
      "loss": 0.547,
      "mean_token_accuracy": 0.8314025439321995,
      "num_tokens": 503769822.0,
      "step": 15784
    },
    {
      "entropy": 0.5153427459299564,
      "epoch": 1.452303322005049,
      "grad_norm": 0.14916911721229553,
      "learning_rate": 5.159321618057472e-05,
      "loss": 0.505,
      "mean_token_accuracy": 0.8414570689201355,
      "num_tokens": 503801988.0,
      "step": 15785
    },
    {
      "entropy": 0.679119520355016,
      "epoch": 1.45239532837649,
      "grad_norm": 0.15985755622386932,
      "learning_rate": 5.15901493544331e-05,
      "loss": 0.671,
      "mean_token_accuracy": 0.7940322794020176,
      "num_tokens": 503834286.0,
      "step": 15786
    },
    {
      "entropy": 0.7503518015146255,
      "epoch": 1.4524873347479312,
      "grad_norm": 0.1691657155752182,
      "learning_rate": 5.158708252829148e-05,
      "loss": 0.7503,
      "mean_token_accuracy": 0.7682212144136429,
      "num_tokens": 503865713.0,
      "step": 15787
    },
    {
      "entropy": 0.6978887263685465,
      "epoch": 1.4525793411193726,
      "grad_norm": 0.15969353914260864,
      "learning_rate": 5.1584015702149846e-05,
      "loss": 0.69,
      "mean_token_accuracy": 0.7887703403830528,
      "num_tokens": 503897925.0,
      "step": 15788
    },
    {
      "entropy": 0.7337517719715834,
      "epoch": 1.4526713474908137,
      "grad_norm": 0.17235659062862396,
      "learning_rate": 5.1580948876008214e-05,
      "loss": 0.7171,
      "mean_token_accuracy": 0.7853335216641426,
      "num_tokens": 503929531.0,
      "step": 15789
    },
    {
      "entropy": 0.6871141269803047,
      "epoch": 1.452763353862255,
      "grad_norm": 0.1625438779592514,
      "learning_rate": 5.15778820498666e-05,
      "loss": 0.6793,
      "mean_token_accuracy": 0.7939862608909607,
      "num_tokens": 503960973.0,
      "step": 15790
    },
    {
      "entropy": 0.5529995488468558,
      "epoch": 1.4528553602336962,
      "grad_norm": 0.15305835008621216,
      "learning_rate": 5.157481522372497e-05,
      "loss": 0.5229,
      "mean_token_accuracy": 0.8352208025753498,
      "num_tokens": 503992012.0,
      "step": 15791
    },
    {
      "entropy": 0.7236314807087183,
      "epoch": 1.4529473666051373,
      "grad_norm": 0.1714264303445816,
      "learning_rate": 5.1571748397583344e-05,
      "loss": 0.7099,
      "mean_token_accuracy": 0.7859515883028507,
      "num_tokens": 504023896.0,
      "step": 15792
    },
    {
      "entropy": 0.5942583531141281,
      "epoch": 1.4530393729765787,
      "grad_norm": 0.1522926539182663,
      "learning_rate": 5.156868157144171e-05,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8237128257751465,
      "num_tokens": 504056055.0,
      "step": 15793
    },
    {
      "entropy": 0.6862996853888035,
      "epoch": 1.4531313793480198,
      "grad_norm": 0.16716104745864868,
      "learning_rate": 5.1565614745300094e-05,
      "loss": 0.6859,
      "mean_token_accuracy": 0.7949174866080284,
      "num_tokens": 504087518.0,
      "step": 15794
    },
    {
      "entropy": 0.5311277466826141,
      "epoch": 1.4532233857194612,
      "grad_norm": 0.14790713787078857,
      "learning_rate": 5.156254791915847e-05,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8377847969532013,
      "num_tokens": 504119132.0,
      "step": 15795
    },
    {
      "entropy": 0.5403058603405952,
      "epoch": 1.4533153920909023,
      "grad_norm": 0.15611770749092102,
      "learning_rate": 5.1559481093016836e-05,
      "loss": 0.5196,
      "mean_token_accuracy": 0.8410667218267918,
      "num_tokens": 504150468.0,
      "step": 15796
    },
    {
      "entropy": 0.7092645540833473,
      "epoch": 1.4534073984623435,
      "grad_norm": 0.1616283506155014,
      "learning_rate": 5.155641426687521e-05,
      "loss": 0.6943,
      "mean_token_accuracy": 0.7882866151630878,
      "num_tokens": 504181860.0,
      "step": 15797
    },
    {
      "entropy": 0.5907447086647153,
      "epoch": 1.4534994048337848,
      "grad_norm": 0.15358573198318481,
      "learning_rate": 5.155334744073359e-05,
      "loss": 0.5678,
      "mean_token_accuracy": 0.8232538625597954,
      "num_tokens": 504213565.0,
      "step": 15798
    },
    {
      "entropy": 0.548805695027113,
      "epoch": 1.453591411205226,
      "grad_norm": 0.15003728866577148,
      "learning_rate": 5.155028061459196e-05,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8341030813753605,
      "num_tokens": 504244611.0,
      "step": 15799
    },
    {
      "entropy": 0.7239287691190839,
      "epoch": 1.4536834175766673,
      "grad_norm": 0.16481515765190125,
      "learning_rate": 5.1547213788450335e-05,
      "loss": 0.707,
      "mean_token_accuracy": 0.7841576151549816,
      "num_tokens": 504275569.0,
      "step": 15800
    },
    {
      "entropy": 0.613507951144129,
      "epoch": 1.4537754239481084,
      "grad_norm": 0.16017155349254608,
      "learning_rate": 5.15441469623087e-05,
      "loss": 0.5967,
      "mean_token_accuracy": 0.8151057697832584,
      "num_tokens": 504307822.0,
      "step": 15801
    },
    {
      "entropy": 0.6481163026764989,
      "epoch": 1.4538674303195496,
      "grad_norm": 0.1700168251991272,
      "learning_rate": 5.1541080136167084e-05,
      "loss": 0.6421,
      "mean_token_accuracy": 0.8070324324071407,
      "num_tokens": 504339014.0,
      "step": 15802
    },
    {
      "entropy": 0.6571158617734909,
      "epoch": 1.453959436690991,
      "grad_norm": 0.16184751689434052,
      "learning_rate": 5.153801331002546e-05,
      "loss": 0.6506,
      "mean_token_accuracy": 0.8003628067672253,
      "num_tokens": 504371409.0,
      "step": 15803
    },
    {
      "entropy": 0.6139889019541442,
      "epoch": 1.454051443062432,
      "grad_norm": 0.1573951691389084,
      "learning_rate": 5.153494648388383e-05,
      "loss": 0.606,
      "mean_token_accuracy": 0.8151005804538727,
      "num_tokens": 504402380.0,
      "step": 15804
    },
    {
      "entropy": 0.6613219324499369,
      "epoch": 1.4541434494338734,
      "grad_norm": 0.1627090871334076,
      "learning_rate": 5.153187965774221e-05,
      "loss": 0.6615,
      "mean_token_accuracy": 0.8010417334735394,
      "num_tokens": 504433769.0,
      "step": 15805
    },
    {
      "entropy": 0.6188844237476587,
      "epoch": 1.4542354558053145,
      "grad_norm": 0.151238352060318,
      "learning_rate": 5.152881283160058e-05,
      "loss": 0.6068,
      "mean_token_accuracy": 0.8115369379520416,
      "num_tokens": 504465772.0,
      "step": 15806
    },
    {
      "entropy": 0.5570871066302061,
      "epoch": 1.4543274621767557,
      "grad_norm": 0.15971580147743225,
      "learning_rate": 5.152574600545895e-05,
      "loss": 0.5322,
      "mean_token_accuracy": 0.8331260941922665,
      "num_tokens": 504497320.0,
      "step": 15807
    },
    {
      "entropy": 0.598905717022717,
      "epoch": 1.454419468548197,
      "grad_norm": 0.16426445543766022,
      "learning_rate": 5.1522679179317325e-05,
      "loss": 0.5851,
      "mean_token_accuracy": 0.8211383894085884,
      "num_tokens": 504528406.0,
      "step": 15808
    },
    {
      "entropy": 0.6246505249291658,
      "epoch": 1.4545114749196382,
      "grad_norm": 0.1550159454345703,
      "learning_rate": 5.151961235317571e-05,
      "loss": 0.6167,
      "mean_token_accuracy": 0.8123116828501225,
      "num_tokens": 504560625.0,
      "step": 15809
    },
    {
      "entropy": 0.6150377001613379,
      "epoch": 1.4546034812910795,
      "grad_norm": 0.16279752552509308,
      "learning_rate": 5.1516545527034075e-05,
      "loss": 0.6056,
      "mean_token_accuracy": 0.8183284513652325,
      "num_tokens": 504591676.0,
      "step": 15810
    },
    {
      "entropy": 0.6116444766521454,
      "epoch": 1.4546954876625207,
      "grad_norm": 0.16338807344436646,
      "learning_rate": 5.151347870089245e-05,
      "loss": 0.5984,
      "mean_token_accuracy": 0.8120118901133537,
      "num_tokens": 504622332.0,
      "step": 15811
    },
    {
      "entropy": 0.5603195666335523,
      "epoch": 1.4547874940339618,
      "grad_norm": 0.1419735848903656,
      "learning_rate": 5.151041187475082e-05,
      "loss": 0.5512,
      "mean_token_accuracy": 0.8283580094575882,
      "num_tokens": 504654544.0,
      "step": 15812
    },
    {
      "entropy": 0.6557175843045115,
      "epoch": 1.4548795004054031,
      "grad_norm": 0.15831224620342255,
      "learning_rate": 5.15073450486092e-05,
      "loss": 0.6427,
      "mean_token_accuracy": 0.8027049489319324,
      "num_tokens": 504686779.0,
      "step": 15813
    },
    {
      "entropy": 0.6289523188024759,
      "epoch": 1.4549715067768443,
      "grad_norm": 0.15706782042980194,
      "learning_rate": 5.150427822246757e-05,
      "loss": 0.6243,
      "mean_token_accuracy": 0.8102222979068756,
      "num_tokens": 504719182.0,
      "step": 15814
    },
    {
      "entropy": 0.687131161801517,
      "epoch": 1.4550635131482856,
      "grad_norm": 0.16247442364692688,
      "learning_rate": 5.150121139632594e-05,
      "loss": 0.6778,
      "mean_token_accuracy": 0.8010516874492168,
      "num_tokens": 504751746.0,
      "step": 15815
    },
    {
      "entropy": 0.6256302110850811,
      "epoch": 1.4551555195197268,
      "grad_norm": 0.16106535494327545,
      "learning_rate": 5.1498144570184316e-05,
      "loss": 0.618,
      "mean_token_accuracy": 0.8099383600056171,
      "num_tokens": 504783489.0,
      "step": 15816
    },
    {
      "entropy": 0.5273675890639424,
      "epoch": 1.455247525891168,
      "grad_norm": 0.14637644588947296,
      "learning_rate": 5.14950777440427e-05,
      "loss": 0.5228,
      "mean_token_accuracy": 0.8386023044586182,
      "num_tokens": 504815674.0,
      "step": 15817
    },
    {
      "entropy": 0.5409243507310748,
      "epoch": 1.4553395322626093,
      "grad_norm": 0.15350952744483948,
      "learning_rate": 5.1492010917901065e-05,
      "loss": 0.5325,
      "mean_token_accuracy": 0.8370373770594597,
      "num_tokens": 504846879.0,
      "step": 15818
    },
    {
      "entropy": 0.6405602409504354,
      "epoch": 1.4554315386340504,
      "grad_norm": 0.15845870971679688,
      "learning_rate": 5.148894409175944e-05,
      "loss": 0.6271,
      "mean_token_accuracy": 0.8025601245462894,
      "num_tokens": 504877760.0,
      "step": 15819
    },
    {
      "entropy": 0.5666216262616217,
      "epoch": 1.4555235450054917,
      "grad_norm": 0.15690977871418,
      "learning_rate": 5.148587726561781e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8303386941552162,
      "num_tokens": 504910294.0,
      "step": 15820
    },
    {
      "entropy": 0.6157129406929016,
      "epoch": 1.4556155513769329,
      "grad_norm": 0.15824569761753082,
      "learning_rate": 5.148281043947619e-05,
      "loss": 0.6096,
      "mean_token_accuracy": 0.809375885874033,
      "num_tokens": 504942376.0,
      "step": 15821
    },
    {
      "entropy": 0.6679643336683512,
      "epoch": 1.455707557748374,
      "grad_norm": 0.15827225148677826,
      "learning_rate": 5.1479743613334564e-05,
      "loss": 0.6508,
      "mean_token_accuracy": 0.7997612990438938,
      "num_tokens": 504973403.0,
      "step": 15822
    },
    {
      "entropy": 0.7346063051372766,
      "epoch": 1.4557995641198154,
      "grad_norm": 0.16567164659500122,
      "learning_rate": 5.147667678719293e-05,
      "loss": 0.7319,
      "mean_token_accuracy": 0.7809737510979176,
      "num_tokens": 505005232.0,
      "step": 15823
    },
    {
      "entropy": 0.461185545893386,
      "epoch": 1.4558915704912565,
      "grad_norm": 0.14391493797302246,
      "learning_rate": 5.1473609961051306e-05,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8584975004196167,
      "num_tokens": 505036827.0,
      "step": 15824
    },
    {
      "entropy": 0.6930051436647773,
      "epoch": 1.4559835768626979,
      "grad_norm": 0.16557596623897552,
      "learning_rate": 5.147054313490969e-05,
      "loss": 0.6841,
      "mean_token_accuracy": 0.7921140864491463,
      "num_tokens": 505068822.0,
      "step": 15825
    },
    {
      "entropy": 0.6910206992179155,
      "epoch": 1.456075583234139,
      "grad_norm": 0.1653420776128769,
      "learning_rate": 5.146747630876806e-05,
      "loss": 0.6644,
      "mean_token_accuracy": 0.7981648109853268,
      "num_tokens": 505100273.0,
      "step": 15826
    },
    {
      "entropy": 0.5796243846416473,
      "epoch": 1.4561675896055801,
      "grad_norm": 0.15690977871418,
      "learning_rate": 5.146440948262643e-05,
      "loss": 0.5768,
      "mean_token_accuracy": 0.8215974122285843,
      "num_tokens": 505132417.0,
      "step": 15827
    },
    {
      "entropy": 0.5493333516642451,
      "epoch": 1.4562595959770215,
      "grad_norm": 0.15755823254585266,
      "learning_rate": 5.14613426564848e-05,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8315065167844296,
      "num_tokens": 505163831.0,
      "step": 15828
    },
    {
      "entropy": 0.5097054736688733,
      "epoch": 1.4563516023484626,
      "grad_norm": 0.14849476516246796,
      "learning_rate": 5.1458275830343186e-05,
      "loss": 0.4893,
      "mean_token_accuracy": 0.8504113145172596,
      "num_tokens": 505195292.0,
      "step": 15829
    },
    {
      "entropy": 0.6903154784813523,
      "epoch": 1.456443608719904,
      "grad_norm": 0.16186966001987457,
      "learning_rate": 5.1455209004201554e-05,
      "loss": 0.6752,
      "mean_token_accuracy": 0.7988710254430771,
      "num_tokens": 505226868.0,
      "step": 15830
    },
    {
      "entropy": 0.6446634046733379,
      "epoch": 1.456535615091345,
      "grad_norm": 0.16838961839675903,
      "learning_rate": 5.145214217805993e-05,
      "loss": 0.6426,
      "mean_token_accuracy": 0.8052138909697533,
      "num_tokens": 505259555.0,
      "step": 15831
    },
    {
      "entropy": 0.5559707204811275,
      "epoch": 1.4566276214627862,
      "grad_norm": 0.16011936962604523,
      "learning_rate": 5.144907535191831e-05,
      "loss": 0.5352,
      "mean_token_accuracy": 0.8348453268408775,
      "num_tokens": 505290638.0,
      "step": 15832
    },
    {
      "entropy": 0.6177871022373438,
      "epoch": 1.4567196278342276,
      "grad_norm": 0.1533050537109375,
      "learning_rate": 5.144600852577668e-05,
      "loss": 0.5966,
      "mean_token_accuracy": 0.8182569816708565,
      "num_tokens": 505322269.0,
      "step": 15833
    },
    {
      "entropy": 0.6733123809099197,
      "epoch": 1.4568116342056687,
      "grad_norm": 0.15860141813755035,
      "learning_rate": 5.144294169963505e-05,
      "loss": 0.6637,
      "mean_token_accuracy": 0.7984003387391567,
      "num_tokens": 505354696.0,
      "step": 15834
    },
    {
      "entropy": 0.6748689655214548,
      "epoch": 1.45690364057711,
      "grad_norm": 0.16394269466400146,
      "learning_rate": 5.143987487349342e-05,
      "loss": 0.6653,
      "mean_token_accuracy": 0.7980838716030121,
      "num_tokens": 505386967.0,
      "step": 15835
    },
    {
      "entropy": 0.5356528637930751,
      "epoch": 1.4569956469485512,
      "grad_norm": 0.146586611866951,
      "learning_rate": 5.14368080473518e-05,
      "loss": 0.5172,
      "mean_token_accuracy": 0.8429821282625198,
      "num_tokens": 505419595.0,
      "step": 15836
    },
    {
      "entropy": 0.557378988713026,
      "epoch": 1.4570876533199923,
      "grad_norm": 0.1470826268196106,
      "learning_rate": 5.1433741221210177e-05,
      "loss": 0.5456,
      "mean_token_accuracy": 0.8295264802873135,
      "num_tokens": 505451244.0,
      "step": 15837
    },
    {
      "entropy": 0.6041176486760378,
      "epoch": 1.4571796596914337,
      "grad_norm": 0.16320112347602844,
      "learning_rate": 5.1430674395068544e-05,
      "loss": 0.5954,
      "mean_token_accuracy": 0.8187270946800709,
      "num_tokens": 505483569.0,
      "step": 15838
    },
    {
      "entropy": 0.5649297889322042,
      "epoch": 1.4572716660628748,
      "grad_norm": 0.1577875316143036,
      "learning_rate": 5.142760756892692e-05,
      "loss": 0.5404,
      "mean_token_accuracy": 0.8319936394691467,
      "num_tokens": 505515815.0,
      "step": 15839
    },
    {
      "entropy": 0.7161605507135391,
      "epoch": 1.4573636724343162,
      "grad_norm": 0.17125515639781952,
      "learning_rate": 5.14245407427853e-05,
      "loss": 0.7061,
      "mean_token_accuracy": 0.783320177346468,
      "num_tokens": 505546492.0,
      "step": 15840
    },
    {
      "entropy": 0.5979152633808553,
      "epoch": 1.4574556788057573,
      "grad_norm": 0.15191560983657837,
      "learning_rate": 5.142147391664367e-05,
      "loss": 0.5779,
      "mean_token_accuracy": 0.8228286132216454,
      "num_tokens": 505578093.0,
      "step": 15841
    },
    {
      "entropy": 0.584631685167551,
      "epoch": 1.4575476851771985,
      "grad_norm": 0.1584203988313675,
      "learning_rate": 5.141840709050204e-05,
      "loss": 0.5798,
      "mean_token_accuracy": 0.8217008970677853,
      "num_tokens": 505609720.0,
      "step": 15842
    },
    {
      "entropy": 0.567213524132967,
      "epoch": 1.4576396915486398,
      "grad_norm": 0.1484849900007248,
      "learning_rate": 5.141534026436041e-05,
      "loss": 0.5458,
      "mean_token_accuracy": 0.831494614481926,
      "num_tokens": 505641665.0,
      "step": 15843
    },
    {
      "entropy": 0.5800243690609932,
      "epoch": 1.457731697920081,
      "grad_norm": 0.15931539237499237,
      "learning_rate": 5.141227343821879e-05,
      "loss": 0.5726,
      "mean_token_accuracy": 0.8252805843949318,
      "num_tokens": 505673680.0,
      "step": 15844
    },
    {
      "entropy": 0.5434837117791176,
      "epoch": 1.4578237042915223,
      "grad_norm": 0.14991740882396698,
      "learning_rate": 5.140920661207717e-05,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8346811942756176,
      "num_tokens": 505705801.0,
      "step": 15845
    },
    {
      "entropy": 0.6168646300211549,
      "epoch": 1.4579157106629634,
      "grad_norm": 0.1568572074174881,
      "learning_rate": 5.1406139785935535e-05,
      "loss": 0.6133,
      "mean_token_accuracy": 0.8119605109095573,
      "num_tokens": 505738202.0,
      "step": 15846
    },
    {
      "entropy": 0.5380668547004461,
      "epoch": 1.4580077170344046,
      "grad_norm": 0.14694033563137054,
      "learning_rate": 5.140307295979391e-05,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8381116054952145,
      "num_tokens": 505770894.0,
      "step": 15847
    },
    {
      "entropy": 0.6736534461379051,
      "epoch": 1.458099723405846,
      "grad_norm": 0.1662304848432541,
      "learning_rate": 5.140000613365229e-05,
      "loss": 0.6759,
      "mean_token_accuracy": 0.79625129327178,
      "num_tokens": 505803356.0,
      "step": 15848
    },
    {
      "entropy": 0.513797047547996,
      "epoch": 1.458191729777287,
      "grad_norm": 0.15537405014038086,
      "learning_rate": 5.139693930751066e-05,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8449638411402702,
      "num_tokens": 505835518.0,
      "step": 15849
    },
    {
      "entropy": 0.5998875610530376,
      "epoch": 1.4582837361487284,
      "grad_norm": 0.1554698497056961,
      "learning_rate": 5.1393872481369033e-05,
      "loss": 0.5917,
      "mean_token_accuracy": 0.8202530555427074,
      "num_tokens": 505868014.0,
      "step": 15850
    },
    {
      "entropy": 0.6695145908743143,
      "epoch": 1.4583757425201695,
      "grad_norm": 0.1600845456123352,
      "learning_rate": 5.13908056552274e-05,
      "loss": 0.6536,
      "mean_token_accuracy": 0.8010216504335403,
      "num_tokens": 505900292.0,
      "step": 15851
    },
    {
      "entropy": 0.6367033161222935,
      "epoch": 1.4584677488916107,
      "grad_norm": 0.15665361285209656,
      "learning_rate": 5.138773882908578e-05,
      "loss": 0.6161,
      "mean_token_accuracy": 0.8059038333594799,
      "num_tokens": 505932041.0,
      "step": 15852
    },
    {
      "entropy": 0.5068095037713647,
      "epoch": 1.458559755263052,
      "grad_norm": 0.13997673988342285,
      "learning_rate": 5.138467200294416e-05,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8457505740225315,
      "num_tokens": 505964229.0,
      "step": 15853
    },
    {
      "entropy": 0.587566529167816,
      "epoch": 1.4586517616344932,
      "grad_norm": 0.15159448981285095,
      "learning_rate": 5.1381605176802525e-05,
      "loss": 0.568,
      "mean_token_accuracy": 0.8278598673641682,
      "num_tokens": 505996465.0,
      "step": 15854
    },
    {
      "entropy": 0.5910827554762363,
      "epoch": 1.4587437680059345,
      "grad_norm": 0.16499090194702148,
      "learning_rate": 5.13785383506609e-05,
      "loss": 0.575,
      "mean_token_accuracy": 0.8262392170727253,
      "num_tokens": 506027602.0,
      "step": 15855
    },
    {
      "entropy": 0.5550304455682635,
      "epoch": 1.4588357743773757,
      "grad_norm": 0.14802969992160797,
      "learning_rate": 5.137547152451928e-05,
      "loss": 0.5267,
      "mean_token_accuracy": 0.8364850953221321,
      "num_tokens": 506059420.0,
      "step": 15856
    },
    {
      "entropy": 0.5645258743315935,
      "epoch": 1.4589277807488168,
      "grad_norm": 0.1559331715106964,
      "learning_rate": 5.137240469837765e-05,
      "loss": 0.5697,
      "mean_token_accuracy": 0.8213145323097706,
      "num_tokens": 506091049.0,
      "step": 15857
    },
    {
      "entropy": 0.6131428759545088,
      "epoch": 1.4590197871202581,
      "grad_norm": 0.14935249090194702,
      "learning_rate": 5.1369337872236024e-05,
      "loss": 0.5935,
      "mean_token_accuracy": 0.8179425336420536,
      "num_tokens": 506123180.0,
      "step": 15858
    },
    {
      "entropy": 0.5099128559231758,
      "epoch": 1.4591117934916993,
      "grad_norm": 0.14392879605293274,
      "learning_rate": 5.136627104609439e-05,
      "loss": 0.4863,
      "mean_token_accuracy": 0.8480139747262001,
      "num_tokens": 506155233.0,
      "step": 15859
    },
    {
      "entropy": 0.5218269471079111,
      "epoch": 1.4592037998631406,
      "grad_norm": 0.1432790607213974,
      "learning_rate": 5.136320421995277e-05,
      "loss": 0.512,
      "mean_token_accuracy": 0.8399169258773327,
      "num_tokens": 506187356.0,
      "step": 15860
    },
    {
      "entropy": 0.5702369371429086,
      "epoch": 1.4592958062345818,
      "grad_norm": 0.14906342327594757,
      "learning_rate": 5.136013739381115e-05,
      "loss": 0.553,
      "mean_token_accuracy": 0.8294149897992611,
      "num_tokens": 506220124.0,
      "step": 15861
    },
    {
      "entropy": 0.5777165957260877,
      "epoch": 1.459387812606023,
      "grad_norm": 0.15072770416736603,
      "learning_rate": 5.1357070567669516e-05,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8280680030584335,
      "num_tokens": 506252434.0,
      "step": 15862
    },
    {
      "entropy": 0.5393446944653988,
      "epoch": 1.4594798189774643,
      "grad_norm": 0.15146751701831818,
      "learning_rate": 5.1354003741527904e-05,
      "loss": 0.5318,
      "mean_token_accuracy": 0.836260337382555,
      "num_tokens": 506283829.0,
      "step": 15863
    },
    {
      "entropy": 0.5661842375993729,
      "epoch": 1.4595718253489054,
      "grad_norm": 0.1539762169122696,
      "learning_rate": 5.135093691538627e-05,
      "loss": 0.5581,
      "mean_token_accuracy": 0.8261567838490009,
      "num_tokens": 506316018.0,
      "step": 15864
    },
    {
      "entropy": 0.5697293244302273,
      "epoch": 1.4596638317203467,
      "grad_norm": 0.15349815785884857,
      "learning_rate": 5.134787008924464e-05,
      "loss": 0.5597,
      "mean_token_accuracy": 0.8317884914577007,
      "num_tokens": 506348425.0,
      "step": 15865
    },
    {
      "entropy": 0.5902610635384917,
      "epoch": 1.4597558380917879,
      "grad_norm": 0.15944841504096985,
      "learning_rate": 5.1344803263103014e-05,
      "loss": 0.5841,
      "mean_token_accuracy": 0.822501502931118,
      "num_tokens": 506380758.0,
      "step": 15866
    },
    {
      "entropy": 0.7174284979701042,
      "epoch": 1.459847844463229,
      "grad_norm": 0.16837261617183685,
      "learning_rate": 5.1341736436961396e-05,
      "loss": 0.7083,
      "mean_token_accuracy": 0.7874757014214993,
      "num_tokens": 506412031.0,
      "step": 15867
    },
    {
      "entropy": 0.6727189049124718,
      "epoch": 1.4599398508346704,
      "grad_norm": 0.16429097950458527,
      "learning_rate": 5.133866961081977e-05,
      "loss": 0.6688,
      "mean_token_accuracy": 0.7965656667947769,
      "num_tokens": 506444636.0,
      "step": 15868
    },
    {
      "entropy": 0.656867902725935,
      "epoch": 1.4600318572061115,
      "grad_norm": 0.16187843680381775,
      "learning_rate": 5.133560278467814e-05,
      "loss": 0.6557,
      "mean_token_accuracy": 0.7988658733665943,
      "num_tokens": 506475958.0,
      "step": 15869
    },
    {
      "entropy": 0.641137408092618,
      "epoch": 1.4601238635775529,
      "grad_norm": 0.1613541543483734,
      "learning_rate": 5.133253595853651e-05,
      "loss": 0.6427,
      "mean_token_accuracy": 0.8019476421177387,
      "num_tokens": 506508415.0,
      "step": 15870
    },
    {
      "entropy": 0.5807381700724363,
      "epoch": 1.460215869948994,
      "grad_norm": 0.15647581219673157,
      "learning_rate": 5.1329469132394894e-05,
      "loss": 0.5721,
      "mean_token_accuracy": 0.8255370333790779,
      "num_tokens": 506540030.0,
      "step": 15871
    },
    {
      "entropy": 0.663539320230484,
      "epoch": 1.4603078763204351,
      "grad_norm": 0.164651557803154,
      "learning_rate": 5.132640230625326e-05,
      "loss": 0.6519,
      "mean_token_accuracy": 0.799929678440094,
      "num_tokens": 506571434.0,
      "step": 15872
    },
    {
      "entropy": 0.5751142017543316,
      "epoch": 1.4603998826918765,
      "grad_norm": 0.15299898386001587,
      "learning_rate": 5.132333548011164e-05,
      "loss": 0.55,
      "mean_token_accuracy": 0.8313946537673473,
      "num_tokens": 506602953.0,
      "step": 15873
    },
    {
      "entropy": 0.755827846005559,
      "epoch": 1.4604918890633176,
      "grad_norm": 0.16542436182498932,
      "learning_rate": 5.1320268653970005e-05,
      "loss": 0.7502,
      "mean_token_accuracy": 0.7736002020537853,
      "num_tokens": 506634743.0,
      "step": 15874
    },
    {
      "entropy": 0.5940661076456308,
      "epoch": 1.460583895434759,
      "grad_norm": 0.15204912424087524,
      "learning_rate": 5.1317201827828386e-05,
      "loss": 0.5777,
      "mean_token_accuracy": 0.8232545331120491,
      "num_tokens": 506665902.0,
      "step": 15875
    },
    {
      "entropy": 0.6186506003141403,
      "epoch": 1.4606759018062,
      "grad_norm": 0.16324126720428467,
      "learning_rate": 5.131413500168676e-05,
      "loss": 0.6142,
      "mean_token_accuracy": 0.809911098331213,
      "num_tokens": 506697861.0,
      "step": 15876
    },
    {
      "entropy": 0.6764592863619328,
      "epoch": 1.4607679081776412,
      "grad_norm": 0.16672943532466888,
      "learning_rate": 5.131106817554513e-05,
      "loss": 0.6734,
      "mean_token_accuracy": 0.8000675216317177,
      "num_tokens": 506728948.0,
      "step": 15877
    },
    {
      "entropy": 0.6605377439409494,
      "epoch": 1.4608599145490826,
      "grad_norm": 0.161664679646492,
      "learning_rate": 5.13080013494035e-05,
      "loss": 0.6499,
      "mean_token_accuracy": 0.8030348867177963,
      "num_tokens": 506760518.0,
      "step": 15878
    },
    {
      "entropy": 0.5585836749523878,
      "epoch": 1.4609519209205237,
      "grad_norm": 0.16382937133312225,
      "learning_rate": 5.1304934523261885e-05,
      "loss": 0.5521,
      "mean_token_accuracy": 0.830992329865694,
      "num_tokens": 506792138.0,
      "step": 15879
    },
    {
      "entropy": 0.6054228637367487,
      "epoch": 1.461043927291965,
      "grad_norm": 0.1555086225271225,
      "learning_rate": 5.130186769712025e-05,
      "loss": 0.5991,
      "mean_token_accuracy": 0.8154040686786175,
      "num_tokens": 506824083.0,
      "step": 15880
    },
    {
      "entropy": 0.5608537802472711,
      "epoch": 1.4611359336634062,
      "grad_norm": 0.14924824237823486,
      "learning_rate": 5.129880087097863e-05,
      "loss": 0.5371,
      "mean_token_accuracy": 0.8295242227613926,
      "num_tokens": 506856310.0,
      "step": 15881
    },
    {
      "entropy": 0.6515063736587763,
      "epoch": 1.4612279400348474,
      "grad_norm": 0.15614445507526398,
      "learning_rate": 5.1295734044836995e-05,
      "loss": 0.6493,
      "mean_token_accuracy": 0.8017243072390556,
      "num_tokens": 506888067.0,
      "step": 15882
    },
    {
      "entropy": 0.6853741817176342,
      "epoch": 1.4613199464062887,
      "grad_norm": 0.16151225566864014,
      "learning_rate": 5.1292667218695376e-05,
      "loss": 0.6769,
      "mean_token_accuracy": 0.7948987483978271,
      "num_tokens": 506919567.0,
      "step": 15883
    },
    {
      "entropy": 0.7005405994132161,
      "epoch": 1.4614119527777298,
      "grad_norm": 0.16857722401618958,
      "learning_rate": 5.128960039255375e-05,
      "loss": 0.6921,
      "mean_token_accuracy": 0.7861473746597767,
      "num_tokens": 506951570.0,
      "step": 15884
    },
    {
      "entropy": 0.5420658821240067,
      "epoch": 1.4615039591491712,
      "grad_norm": 0.1490577757358551,
      "learning_rate": 5.128653356641212e-05,
      "loss": 0.539,
      "mean_token_accuracy": 0.8309638909995556,
      "num_tokens": 506983658.0,
      "step": 15885
    },
    {
      "entropy": 0.6203089444898069,
      "epoch": 1.4615959655206123,
      "grad_norm": 0.15342701971530914,
      "learning_rate": 5.1283466740270494e-05,
      "loss": 0.6133,
      "mean_token_accuracy": 0.8117764964699745,
      "num_tokens": 507015644.0,
      "step": 15886
    },
    {
      "entropy": 0.731168707832694,
      "epoch": 1.4616879718920535,
      "grad_norm": 0.16821454465389252,
      "learning_rate": 5.1280399914128875e-05,
      "loss": 0.721,
      "mean_token_accuracy": 0.782596729695797,
      "num_tokens": 507046634.0,
      "step": 15887
    },
    {
      "entropy": 0.6473845429718494,
      "epoch": 1.4617799782634948,
      "grad_norm": 0.15607880055904388,
      "learning_rate": 5.127733308798724e-05,
      "loss": 0.6374,
      "mean_token_accuracy": 0.8090558983385563,
      "num_tokens": 507079069.0,
      "step": 15888
    },
    {
      "entropy": 0.5996034829877317,
      "epoch": 1.461871984634936,
      "grad_norm": 0.14889118075370789,
      "learning_rate": 5.127426626184562e-05,
      "loss": 0.5842,
      "mean_token_accuracy": 0.8186839483678341,
      "num_tokens": 507111088.0,
      "step": 15889
    },
    {
      "entropy": 0.5632627909071743,
      "epoch": 1.4619639910063773,
      "grad_norm": 0.1545657068490982,
      "learning_rate": 5.1271199435703985e-05,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8321934565901756,
      "num_tokens": 507143856.0,
      "step": 15890
    },
    {
      "entropy": 0.6417528819292784,
      "epoch": 1.4620559973778184,
      "grad_norm": 0.15693001449108124,
      "learning_rate": 5.126813260956237e-05,
      "loss": 0.6268,
      "mean_token_accuracy": 0.8073729947209358,
      "num_tokens": 507176258.0,
      "step": 15891
    },
    {
      "entropy": 0.5787632949650288,
      "epoch": 1.4621480037492596,
      "grad_norm": 0.16075432300567627,
      "learning_rate": 5.126506578342074e-05,
      "loss": 0.5768,
      "mean_token_accuracy": 0.8255566656589508,
      "num_tokens": 507207049.0,
      "step": 15892
    },
    {
      "entropy": 0.5611402369104326,
      "epoch": 1.462240010120701,
      "grad_norm": 0.1523512601852417,
      "learning_rate": 5.126199895727911e-05,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8340384438633919,
      "num_tokens": 507239804.0,
      "step": 15893
    },
    {
      "entropy": 0.5347157400101423,
      "epoch": 1.462332016492142,
      "grad_norm": 0.15350501239299774,
      "learning_rate": 5.125893213113749e-05,
      "loss": 0.5285,
      "mean_token_accuracy": 0.8310201279819012,
      "num_tokens": 507271542.0,
      "step": 15894
    },
    {
      "entropy": 0.64595468621701,
      "epoch": 1.4624240228635834,
      "grad_norm": 0.15967440605163574,
      "learning_rate": 5.1255865304995865e-05,
      "loss": 0.636,
      "mean_token_accuracy": 0.8042149841785431,
      "num_tokens": 507303096.0,
      "step": 15895
    },
    {
      "entropy": 0.5754511058330536,
      "epoch": 1.4625160292350246,
      "grad_norm": 0.1511564552783966,
      "learning_rate": 5.125279847885423e-05,
      "loss": 0.5574,
      "mean_token_accuracy": 0.8291707336902618,
      "num_tokens": 507335864.0,
      "step": 15896
    },
    {
      "entropy": 0.6354533703997731,
      "epoch": 1.4626080356064657,
      "grad_norm": 0.1588154286146164,
      "learning_rate": 5.124973165271261e-05,
      "loss": 0.6214,
      "mean_token_accuracy": 0.8080038875341415,
      "num_tokens": 507368172.0,
      "step": 15897
    },
    {
      "entropy": 0.623896487057209,
      "epoch": 1.462700041977907,
      "grad_norm": 0.1621084362268448,
      "learning_rate": 5.124666482657099e-05,
      "loss": 0.6256,
      "mean_token_accuracy": 0.8121611550450325,
      "num_tokens": 507399474.0,
      "step": 15898
    },
    {
      "entropy": 0.6759926620870829,
      "epoch": 1.4627920483493482,
      "grad_norm": 0.1654987335205078,
      "learning_rate": 5.124359800042936e-05,
      "loss": 0.6626,
      "mean_token_accuracy": 0.7972207553684711,
      "num_tokens": 507431409.0,
      "step": 15899
    },
    {
      "entropy": 0.6005591936409473,
      "epoch": 1.4628840547207895,
      "grad_norm": 0.15742768347263336,
      "learning_rate": 5.124053117428773e-05,
      "loss": 0.5977,
      "mean_token_accuracy": 0.8185376636683941,
      "num_tokens": 507463592.0,
      "step": 15900
    },
    {
      "entropy": 0.623132417909801,
      "epoch": 1.4629760610922307,
      "grad_norm": 0.1573338359594345,
      "learning_rate": 5.12374643481461e-05,
      "loss": 0.6089,
      "mean_token_accuracy": 0.810604490339756,
      "num_tokens": 507495859.0,
      "step": 15901
    },
    {
      "entropy": 0.7386713214218616,
      "epoch": 1.4630680674636718,
      "grad_norm": 0.16260655224323273,
      "learning_rate": 5.123439752200449e-05,
      "loss": 0.7444,
      "mean_token_accuracy": 0.7782795242965221,
      "num_tokens": 507528299.0,
      "step": 15902
    },
    {
      "entropy": 0.5181692084297538,
      "epoch": 1.4631600738351132,
      "grad_norm": 0.15056835114955902,
      "learning_rate": 5.1231330695862856e-05,
      "loss": 0.512,
      "mean_token_accuracy": 0.8441139571368694,
      "num_tokens": 507560390.0,
      "step": 15903
    },
    {
      "entropy": 0.6403921484015882,
      "epoch": 1.4632520802065543,
      "grad_norm": 0.15617486834526062,
      "learning_rate": 5.1228263869721224e-05,
      "loss": 0.621,
      "mean_token_accuracy": 0.8068017065525055,
      "num_tokens": 507591381.0,
      "step": 15904
    },
    {
      "entropy": 0.5861259996891022,
      "epoch": 1.4633440865779956,
      "grad_norm": 0.15513966977596283,
      "learning_rate": 5.12251970435796e-05,
      "loss": 0.5753,
      "mean_token_accuracy": 0.8237701132893562,
      "num_tokens": 507623418.0,
      "step": 15905
    },
    {
      "entropy": 0.5946513283997774,
      "epoch": 1.4634360929494368,
      "grad_norm": 0.159029021859169,
      "learning_rate": 5.122213021743798e-05,
      "loss": 0.5854,
      "mean_token_accuracy": 0.8179615139961243,
      "num_tokens": 507655204.0,
      "step": 15906
    },
    {
      "entropy": 0.5997709846124053,
      "epoch": 1.463528099320878,
      "grad_norm": 0.1523367315530777,
      "learning_rate": 5.1219063391296354e-05,
      "loss": 0.586,
      "mean_token_accuracy": 0.8195608854293823,
      "num_tokens": 507687357.0,
      "step": 15907
    },
    {
      "entropy": 0.47903452813625336,
      "epoch": 1.4636201056923193,
      "grad_norm": 0.14307361841201782,
      "learning_rate": 5.121599656515472e-05,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8509043790400028,
      "num_tokens": 507719899.0,
      "step": 15908
    },
    {
      "entropy": 0.5412639458663762,
      "epoch": 1.4637121120637604,
      "grad_norm": 0.14707843959331512,
      "learning_rate": 5.121292973901309e-05,
      "loss": 0.5256,
      "mean_token_accuracy": 0.8385359011590481,
      "num_tokens": 507752514.0,
      "step": 15909
    },
    {
      "entropy": 0.6435603313148022,
      "epoch": 1.4638041184352018,
      "grad_norm": 0.1646440625190735,
      "learning_rate": 5.120986291287148e-05,
      "loss": 0.641,
      "mean_token_accuracy": 0.8079498670995235,
      "num_tokens": 507784369.0,
      "step": 15910
    },
    {
      "entropy": 0.5567183243110776,
      "epoch": 1.4638961248066429,
      "grad_norm": 0.15615710616111755,
      "learning_rate": 5.1206796086729846e-05,
      "loss": 0.55,
      "mean_token_accuracy": 0.8297693580389023,
      "num_tokens": 507816469.0,
      "step": 15911
    },
    {
      "entropy": 0.6658146195113659,
      "epoch": 1.463988131178084,
      "grad_norm": 0.16090849041938782,
      "learning_rate": 5.120372926058822e-05,
      "loss": 0.659,
      "mean_token_accuracy": 0.7993878871202469,
      "num_tokens": 507848573.0,
      "step": 15912
    },
    {
      "entropy": 0.6205844096839428,
      "epoch": 1.4640801375495254,
      "grad_norm": 0.1543278843164444,
      "learning_rate": 5.120066243444659e-05,
      "loss": 0.6117,
      "mean_token_accuracy": 0.8120178543031216,
      "num_tokens": 507881185.0,
      "step": 15913
    },
    {
      "entropy": 0.6350229803938419,
      "epoch": 1.4641721439209665,
      "grad_norm": 0.16436268389225006,
      "learning_rate": 5.119759560830497e-05,
      "loss": 0.6172,
      "mean_token_accuracy": 0.8106080517172813,
      "num_tokens": 507912764.0,
      "step": 15914
    },
    {
      "entropy": 0.6670551989227533,
      "epoch": 1.4642641502924079,
      "grad_norm": 0.15635280311107635,
      "learning_rate": 5.1194528782163345e-05,
      "loss": 0.6489,
      "mean_token_accuracy": 0.799011580646038,
      "num_tokens": 507944909.0,
      "step": 15915
    },
    {
      "entropy": 0.5817303247749805,
      "epoch": 1.464356156663849,
      "grad_norm": 0.15498532354831696,
      "learning_rate": 5.119146195602171e-05,
      "loss": 0.5697,
      "mean_token_accuracy": 0.8226425684988499,
      "num_tokens": 507976186.0,
      "step": 15916
    },
    {
      "entropy": 0.6852188371121883,
      "epoch": 1.4644481630352901,
      "grad_norm": 0.16456769406795502,
      "learning_rate": 5.118839512988009e-05,
      "loss": 0.6767,
      "mean_token_accuracy": 0.795818917453289,
      "num_tokens": 508008018.0,
      "step": 15917
    },
    {
      "entropy": 0.5696108546108007,
      "epoch": 1.4645401694067315,
      "grad_norm": 0.15294642746448517,
      "learning_rate": 5.118532830373847e-05,
      "loss": 0.5521,
      "mean_token_accuracy": 0.8296087235212326,
      "num_tokens": 508040710.0,
      "step": 15918
    },
    {
      "entropy": 0.6282053710892797,
      "epoch": 1.4646321757781726,
      "grad_norm": 0.15326738357543945,
      "learning_rate": 5.118226147759684e-05,
      "loss": 0.6039,
      "mean_token_accuracy": 0.8118971362709999,
      "num_tokens": 508071379.0,
      "step": 15919
    },
    {
      "entropy": 0.6287098852917552,
      "epoch": 1.464724182149614,
      "grad_norm": 0.1523112952709198,
      "learning_rate": 5.117919465145521e-05,
      "loss": 0.6209,
      "mean_token_accuracy": 0.8126650415360928,
      "num_tokens": 508103694.0,
      "step": 15920
    },
    {
      "entropy": 0.6035135718993843,
      "epoch": 1.464816188521055,
      "grad_norm": 0.15889933705329895,
      "learning_rate": 5.117612782531358e-05,
      "loss": 0.5998,
      "mean_token_accuracy": 0.8184774555265903,
      "num_tokens": 508135561.0,
      "step": 15921
    },
    {
      "entropy": 0.5903111677616835,
      "epoch": 1.4649081948924962,
      "grad_norm": 0.15381459891796112,
      "learning_rate": 5.117306099917196e-05,
      "loss": 0.5742,
      "mean_token_accuracy": 0.8235205970704556,
      "num_tokens": 508166995.0,
      "step": 15922
    },
    {
      "entropy": 0.6698893941938877,
      "epoch": 1.4650002012639376,
      "grad_norm": 0.16782407462596893,
      "learning_rate": 5.1169994173030335e-05,
      "loss": 0.6722,
      "mean_token_accuracy": 0.7961210086941719,
      "num_tokens": 508198963.0,
      "step": 15923
    },
    {
      "entropy": 0.7107939198613167,
      "epoch": 1.4650922076353787,
      "grad_norm": 0.16866366565227509,
      "learning_rate": 5.11669273468887e-05,
      "loss": 0.719,
      "mean_token_accuracy": 0.782434731721878,
      "num_tokens": 508231075.0,
      "step": 15924
    },
    {
      "entropy": 0.5295754526741803,
      "epoch": 1.46518421400682,
      "grad_norm": 0.15860974788665771,
      "learning_rate": 5.1163860520747085e-05,
      "loss": 0.5357,
      "mean_token_accuracy": 0.840716477483511,
      "num_tokens": 508262583.0,
      "step": 15925
    },
    {
      "entropy": 0.6167087499052286,
      "epoch": 1.4652762203782612,
      "grad_norm": 0.15992256999015808,
      "learning_rate": 5.116079369460546e-05,
      "loss": 0.6217,
      "mean_token_accuracy": 0.8117366880178452,
      "num_tokens": 508295351.0,
      "step": 15926
    },
    {
      "entropy": 0.6505383774638176,
      "epoch": 1.4653682267497024,
      "grad_norm": 0.15997202694416046,
      "learning_rate": 5.115772686846383e-05,
      "loss": 0.6452,
      "mean_token_accuracy": 0.8068117648363113,
      "num_tokens": 508327527.0,
      "step": 15927
    },
    {
      "entropy": 0.574298856779933,
      "epoch": 1.4654602331211437,
      "grad_norm": 0.15693914890289307,
      "learning_rate": 5.11546600423222e-05,
      "loss": 0.5711,
      "mean_token_accuracy": 0.825566753745079,
      "num_tokens": 508359472.0,
      "step": 15928
    },
    {
      "entropy": 0.5562572106719017,
      "epoch": 1.4655522394925848,
      "grad_norm": 0.1485726684331894,
      "learning_rate": 5.115159321618058e-05,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8288935907185078,
      "num_tokens": 508390840.0,
      "step": 15929
    },
    {
      "entropy": 0.5935538774356246,
      "epoch": 1.4656442458640262,
      "grad_norm": 0.15970315039157867,
      "learning_rate": 5.114852639003895e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8263090029358864,
      "num_tokens": 508422509.0,
      "step": 15930
    },
    {
      "entropy": 0.6148797497153282,
      "epoch": 1.4657362522354673,
      "grad_norm": 0.15710654854774475,
      "learning_rate": 5.1145459563897326e-05,
      "loss": 0.5917,
      "mean_token_accuracy": 0.8173316419124603,
      "num_tokens": 508453925.0,
      "step": 15931
    },
    {
      "entropy": 0.7273372150957584,
      "epoch": 1.4658282586069085,
      "grad_norm": 0.16064032912254333,
      "learning_rate": 5.1142392737755694e-05,
      "loss": 0.6943,
      "mean_token_accuracy": 0.7874250113964081,
      "num_tokens": 508485732.0,
      "step": 15932
    },
    {
      "entropy": 0.6659738477319479,
      "epoch": 1.4659202649783498,
      "grad_norm": 0.15378761291503906,
      "learning_rate": 5.1139325911614075e-05,
      "loss": 0.6463,
      "mean_token_accuracy": 0.805313166230917,
      "num_tokens": 508518415.0,
      "step": 15933
    },
    {
      "entropy": 0.6446541715413332,
      "epoch": 1.466012271349791,
      "grad_norm": 0.16106048226356506,
      "learning_rate": 5.113625908547245e-05,
      "loss": 0.6432,
      "mean_token_accuracy": 0.8029591739177704,
      "num_tokens": 508549441.0,
      "step": 15934
    },
    {
      "entropy": 0.5932256700471044,
      "epoch": 1.4661042777212323,
      "grad_norm": 0.15495772659778595,
      "learning_rate": 5.113319225933082e-05,
      "loss": 0.5732,
      "mean_token_accuracy": 0.8232782036066055,
      "num_tokens": 508580811.0,
      "step": 15935
    },
    {
      "entropy": 0.6811099937185645,
      "epoch": 1.4661962840926734,
      "grad_norm": 0.16076351702213287,
      "learning_rate": 5.113012543318919e-05,
      "loss": 0.6682,
      "mean_token_accuracy": 0.7932293638586998,
      "num_tokens": 508611685.0,
      "step": 15936
    },
    {
      "entropy": 0.6073485780507326,
      "epoch": 1.4662882904641146,
      "grad_norm": 0.15786060690879822,
      "learning_rate": 5.1127058607047574e-05,
      "loss": 0.6061,
      "mean_token_accuracy": 0.8145323134958744,
      "num_tokens": 508643151.0,
      "step": 15937
    },
    {
      "entropy": 0.6549968738108873,
      "epoch": 1.466380296835556,
      "grad_norm": 0.1625804305076599,
      "learning_rate": 5.112399178090594e-05,
      "loss": 0.6592,
      "mean_token_accuracy": 0.7974734865128994,
      "num_tokens": 508675769.0,
      "step": 15938
    },
    {
      "entropy": 0.570815050508827,
      "epoch": 1.466472303206997,
      "grad_norm": 0.15215975046157837,
      "learning_rate": 5.1120924954764316e-05,
      "loss": 0.5689,
      "mean_token_accuracy": 0.8239960819482803,
      "num_tokens": 508707642.0,
      "step": 15939
    },
    {
      "entropy": 0.6717222342267632,
      "epoch": 1.4665643095784384,
      "grad_norm": 0.1613398790359497,
      "learning_rate": 5.1117858128622684e-05,
      "loss": 0.6625,
      "mean_token_accuracy": 0.7990337871015072,
      "num_tokens": 508739207.0,
      "step": 15940
    },
    {
      "entropy": 0.6019633989781141,
      "epoch": 1.4666563159498796,
      "grad_norm": 0.15876072645187378,
      "learning_rate": 5.111479130248107e-05,
      "loss": 0.5934,
      "mean_token_accuracy": 0.816921204328537,
      "num_tokens": 508770369.0,
      "step": 15941
    },
    {
      "entropy": 0.5599199552088976,
      "epoch": 1.4667483223213207,
      "grad_norm": 0.15733402967453003,
      "learning_rate": 5.111172447633944e-05,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8252818956971169,
      "num_tokens": 508802159.0,
      "step": 15942
    },
    {
      "entropy": 0.5548652671277523,
      "epoch": 1.466840328692762,
      "grad_norm": 0.14915278553962708,
      "learning_rate": 5.110865765019781e-05,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8338733501732349,
      "num_tokens": 508833623.0,
      "step": 15943
    },
    {
      "entropy": 0.5427144030109048,
      "epoch": 1.4669323350642032,
      "grad_norm": 0.14849120378494263,
      "learning_rate": 5.110559082405618e-05,
      "loss": 0.5184,
      "mean_token_accuracy": 0.838168278336525,
      "num_tokens": 508865365.0,
      "step": 15944
    },
    {
      "entropy": 0.6662249490618706,
      "epoch": 1.4670243414356445,
      "grad_norm": 0.15243838727474213,
      "learning_rate": 5.1102523997914564e-05,
      "loss": 0.6476,
      "mean_token_accuracy": 0.8041662648320198,
      "num_tokens": 508896985.0,
      "step": 15945
    },
    {
      "entropy": 0.7024514507502317,
      "epoch": 1.4671163478070857,
      "grad_norm": 0.16752123832702637,
      "learning_rate": 5.109945717177294e-05,
      "loss": 0.6846,
      "mean_token_accuracy": 0.7895078100264072,
      "num_tokens": 508929406.0,
      "step": 15946
    },
    {
      "entropy": 0.562535974662751,
      "epoch": 1.4672083541785268,
      "grad_norm": 0.15031246840953827,
      "learning_rate": 5.1096390345631306e-05,
      "loss": 0.541,
      "mean_token_accuracy": 0.8310008123517036,
      "num_tokens": 508960731.0,
      "step": 15947
    },
    {
      "entropy": 0.5863947980105877,
      "epoch": 1.4673003605499682,
      "grad_norm": 0.16005833446979523,
      "learning_rate": 5.1093323519489674e-05,
      "loss": 0.5736,
      "mean_token_accuracy": 0.8218329772353172,
      "num_tokens": 508992088.0,
      "step": 15948
    },
    {
      "entropy": 0.7189249880611897,
      "epoch": 1.4673923669214093,
      "grad_norm": 0.16434045135974884,
      "learning_rate": 5.109025669334806e-05,
      "loss": 0.7147,
      "mean_token_accuracy": 0.7825240902602673,
      "num_tokens": 509023952.0,
      "step": 15949
    },
    {
      "entropy": 0.6591387633234262,
      "epoch": 1.4674843732928506,
      "grad_norm": 0.15799911320209503,
      "learning_rate": 5.108718986720643e-05,
      "loss": 0.6577,
      "mean_token_accuracy": 0.8013168349862099,
      "num_tokens": 509055416.0,
      "step": 15950
    },
    {
      "entropy": 0.6902538705617189,
      "epoch": 1.4675763796642918,
      "grad_norm": 0.16323047876358032,
      "learning_rate": 5.1084123041064805e-05,
      "loss": 0.6797,
      "mean_token_accuracy": 0.7958275489509106,
      "num_tokens": 509088052.0,
      "step": 15951
    },
    {
      "entropy": 0.5717755653895438,
      "epoch": 1.467668386035733,
      "grad_norm": 0.15130029618740082,
      "learning_rate": 5.108105621492317e-05,
      "loss": 0.5582,
      "mean_token_accuracy": 0.8267852738499641,
      "num_tokens": 509119960.0,
      "step": 15952
    },
    {
      "entropy": 0.586000083014369,
      "epoch": 1.4677603924071743,
      "grad_norm": 0.15720263123512268,
      "learning_rate": 5.1077989388781554e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8192725032567978,
      "num_tokens": 509151889.0,
      "step": 15953
    },
    {
      "entropy": 0.6936807502061129,
      "epoch": 1.4678523987786154,
      "grad_norm": 0.16561627388000488,
      "learning_rate": 5.107492256263993e-05,
      "loss": 0.689,
      "mean_token_accuracy": 0.7905058898031712,
      "num_tokens": 509183722.0,
      "step": 15954
    },
    {
      "entropy": 0.5746378148905933,
      "epoch": 1.4679444051500568,
      "grad_norm": 0.14689065515995026,
      "learning_rate": 5.10718557364983e-05,
      "loss": 0.562,
      "mean_token_accuracy": 0.8312898389995098,
      "num_tokens": 509216434.0,
      "step": 15955
    },
    {
      "entropy": 0.6550695169717073,
      "epoch": 1.4680364115214979,
      "grad_norm": 0.16188980638980865,
      "learning_rate": 5.106878891035668e-05,
      "loss": 0.6547,
      "mean_token_accuracy": 0.7995774522423744,
      "num_tokens": 509248672.0,
      "step": 15956
    },
    {
      "entropy": 0.5881237601861358,
      "epoch": 1.468128417892939,
      "grad_norm": 0.15113408863544464,
      "learning_rate": 5.106572208421505e-05,
      "loss": 0.5812,
      "mean_token_accuracy": 0.8226920031011105,
      "num_tokens": 509280684.0,
      "step": 15957
    },
    {
      "entropy": 0.6656683813780546,
      "epoch": 1.4682204242643804,
      "grad_norm": 0.16587373614311218,
      "learning_rate": 5.106265525807342e-05,
      "loss": 0.6554,
      "mean_token_accuracy": 0.8009638376533985,
      "num_tokens": 509311810.0,
      "step": 15958
    },
    {
      "entropy": 0.6488853311166167,
      "epoch": 1.4683124306358215,
      "grad_norm": 0.15786142647266388,
      "learning_rate": 5.1059588431931795e-05,
      "loss": 0.6447,
      "mean_token_accuracy": 0.8087005540728569,
      "num_tokens": 509343782.0,
      "step": 15959
    },
    {
      "entropy": 0.5177758997306228,
      "epoch": 1.4684044370072629,
      "grad_norm": 0.14818821847438812,
      "learning_rate": 5.105652160579018e-05,
      "loss": 0.4939,
      "mean_token_accuracy": 0.8512648604810238,
      "num_tokens": 509375506.0,
      "step": 15960
    },
    {
      "entropy": 0.5873172865249217,
      "epoch": 1.468496443378704,
      "grad_norm": 0.1477336883544922,
      "learning_rate": 5.1053454779648545e-05,
      "loss": 0.5767,
      "mean_token_accuracy": 0.8233020901679993,
      "num_tokens": 509408084.0,
      "step": 15961
    },
    {
      "entropy": 0.5765517605468631,
      "epoch": 1.4685884497501451,
      "grad_norm": 0.1597055047750473,
      "learning_rate": 5.105038795350692e-05,
      "loss": 0.5592,
      "mean_token_accuracy": 0.8249683901667595,
      "num_tokens": 509439284.0,
      "step": 15962
    },
    {
      "entropy": 0.604067524895072,
      "epoch": 1.4686804561215865,
      "grad_norm": 0.15576283633708954,
      "learning_rate": 5.104732112736529e-05,
      "loss": 0.601,
      "mean_token_accuracy": 0.8150692209601402,
      "num_tokens": 509470735.0,
      "step": 15963
    },
    {
      "entropy": 0.5154793094843626,
      "epoch": 1.4687724624930276,
      "grad_norm": 0.16475552320480347,
      "learning_rate": 5.104425430122367e-05,
      "loss": 0.5113,
      "mean_token_accuracy": 0.84091891720891,
      "num_tokens": 509502542.0,
      "step": 15964
    },
    {
      "entropy": 0.5347779858857393,
      "epoch": 1.468864468864469,
      "grad_norm": 0.15549896657466888,
      "learning_rate": 5.104118747508204e-05,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8442765027284622,
      "num_tokens": 509534595.0,
      "step": 15965
    },
    {
      "entropy": 0.5990712493658066,
      "epoch": 1.46895647523591,
      "grad_norm": 0.16095127165317535,
      "learning_rate": 5.103812064894041e-05,
      "loss": 0.5678,
      "mean_token_accuracy": 0.823313545435667,
      "num_tokens": 509566785.0,
      "step": 15966
    },
    {
      "entropy": 0.6539916750043631,
      "epoch": 1.4690484816073512,
      "grad_norm": 0.15690496563911438,
      "learning_rate": 5.1035053822798786e-05,
      "loss": 0.6394,
      "mean_token_accuracy": 0.8025204688310623,
      "num_tokens": 509599131.0,
      "step": 15967
    },
    {
      "entropy": 0.6282392889261246,
      "epoch": 1.4691404879787926,
      "grad_norm": 0.15658795833587646,
      "learning_rate": 5.103198699665717e-05,
      "loss": 0.6219,
      "mean_token_accuracy": 0.8044080771505833,
      "num_tokens": 509630622.0,
      "step": 15968
    },
    {
      "entropy": 0.6111203841865063,
      "epoch": 1.4692324943502337,
      "grad_norm": 0.16632325947284698,
      "learning_rate": 5.1028920170515535e-05,
      "loss": 0.5945,
      "mean_token_accuracy": 0.8187356255948544,
      "num_tokens": 509662115.0,
      "step": 15969
    },
    {
      "entropy": 0.7127977106720209,
      "epoch": 1.469324500721675,
      "grad_norm": 0.15941132605075836,
      "learning_rate": 5.102585334437391e-05,
      "loss": 0.7112,
      "mean_token_accuracy": 0.7837253920733929,
      "num_tokens": 509694789.0,
      "step": 15970
    },
    {
      "entropy": 0.5910996943712234,
      "epoch": 1.4694165070931162,
      "grad_norm": 0.15133225917816162,
      "learning_rate": 5.102278651823228e-05,
      "loss": 0.5933,
      "mean_token_accuracy": 0.8196336217224598,
      "num_tokens": 509726881.0,
      "step": 15971
    },
    {
      "entropy": 0.616640951950103,
      "epoch": 1.4695085134645574,
      "grad_norm": 0.16347061097621918,
      "learning_rate": 5.101971969209066e-05,
      "loss": 0.6212,
      "mean_token_accuracy": 0.8136050850152969,
      "num_tokens": 509758813.0,
      "step": 15972
    },
    {
      "entropy": 0.7275374457240105,
      "epoch": 1.4696005198359987,
      "grad_norm": 0.16148822009563446,
      "learning_rate": 5.1016652865949034e-05,
      "loss": 0.718,
      "mean_token_accuracy": 0.7791686505079269,
      "num_tokens": 509790672.0,
      "step": 15973
    },
    {
      "entropy": 0.5749047603458166,
      "epoch": 1.4696925262074398,
      "grad_norm": 0.16265623271465302,
      "learning_rate": 5.10135860398074e-05,
      "loss": 0.5601,
      "mean_token_accuracy": 0.8300409391522408,
      "num_tokens": 509822416.0,
      "step": 15974
    },
    {
      "entropy": 0.6293880557641387,
      "epoch": 1.4697845325788812,
      "grad_norm": 0.15264296531677246,
      "learning_rate": 5.1010519213665776e-05,
      "loss": 0.6295,
      "mean_token_accuracy": 0.806852214038372,
      "num_tokens": 509855069.0,
      "step": 15975
    },
    {
      "entropy": 0.6069798413664103,
      "epoch": 1.4698765389503223,
      "grad_norm": 0.1572076827287674,
      "learning_rate": 5.100745238752416e-05,
      "loss": 0.5928,
      "mean_token_accuracy": 0.8209596499800682,
      "num_tokens": 509886500.0,
      "step": 15976
    },
    {
      "entropy": 0.5654487647116184,
      "epoch": 1.4699685453217635,
      "grad_norm": 0.15272414684295654,
      "learning_rate": 5.1004385561382526e-05,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8271587565541267,
      "num_tokens": 509918273.0,
      "step": 15977
    },
    {
      "entropy": 0.7084674704819918,
      "epoch": 1.4700605516932048,
      "grad_norm": 0.17020846903324127,
      "learning_rate": 5.10013187352409e-05,
      "loss": 0.6895,
      "mean_token_accuracy": 0.7890656590461731,
      "num_tokens": 509949272.0,
      "step": 15978
    },
    {
      "entropy": 0.7004134925082326,
      "epoch": 1.470152558064646,
      "grad_norm": 0.16177591681480408,
      "learning_rate": 5.099825190909927e-05,
      "loss": 0.6821,
      "mean_token_accuracy": 0.794464435428381,
      "num_tokens": 509981160.0,
      "step": 15979
    },
    {
      "entropy": 0.5840126178227365,
      "epoch": 1.4702445644360873,
      "grad_norm": 0.15505032241344452,
      "learning_rate": 5.099518508295765e-05,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8232966065406799,
      "num_tokens": 510012983.0,
      "step": 15980
    },
    {
      "entropy": 0.7045270763337612,
      "epoch": 1.4703365708075284,
      "grad_norm": 0.1602136492729187,
      "learning_rate": 5.0992118256816024e-05,
      "loss": 0.6896,
      "mean_token_accuracy": 0.7881079167127609,
      "num_tokens": 510044806.0,
      "step": 15981
    },
    {
      "entropy": 0.6538307592272758,
      "epoch": 1.4704285771789696,
      "grad_norm": 0.1616784632205963,
      "learning_rate": 5.098905143067439e-05,
      "loss": 0.6414,
      "mean_token_accuracy": 0.8033874779939651,
      "num_tokens": 510076583.0,
      "step": 15982
    },
    {
      "entropy": 0.6799112348817289,
      "epoch": 1.470520583550411,
      "grad_norm": 0.16252058744430542,
      "learning_rate": 5.098598460453277e-05,
      "loss": 0.6666,
      "mean_token_accuracy": 0.7959653288125992,
      "num_tokens": 510107989.0,
      "step": 15983
    },
    {
      "entropy": 0.6512375194579363,
      "epoch": 1.470612589921852,
      "grad_norm": 0.16322055459022522,
      "learning_rate": 5.098291777839115e-05,
      "loss": 0.6305,
      "mean_token_accuracy": 0.8005880527198315,
      "num_tokens": 510138816.0,
      "step": 15984
    },
    {
      "entropy": 0.5621978202834725,
      "epoch": 1.4707045962932934,
      "grad_norm": 0.1498989462852478,
      "learning_rate": 5.097985095224952e-05,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8287069238722324,
      "num_tokens": 510170480.0,
      "step": 15985
    },
    {
      "entropy": 0.6055818479508162,
      "epoch": 1.4707966026647346,
      "grad_norm": 0.15334175527095795,
      "learning_rate": 5.097678412610789e-05,
      "loss": 0.6,
      "mean_token_accuracy": 0.8172479085624218,
      "num_tokens": 510202397.0,
      "step": 15986
    },
    {
      "entropy": 0.777574323117733,
      "epoch": 1.4708886090361757,
      "grad_norm": 0.1651112288236618,
      "learning_rate": 5.097371729996627e-05,
      "loss": 0.7735,
      "mean_token_accuracy": 0.7686921395361423,
      "num_tokens": 510234555.0,
      "step": 15987
    },
    {
      "entropy": 0.6397453248500824,
      "epoch": 1.470980615407617,
      "grad_norm": 0.15385708212852478,
      "learning_rate": 5.097065047382465e-05,
      "loss": 0.6284,
      "mean_token_accuracy": 0.8051528781652451,
      "num_tokens": 510266636.0,
      "step": 15988
    },
    {
      "entropy": 0.5391913950443268,
      "epoch": 1.4710726217790582,
      "grad_norm": 0.14930851757526398,
      "learning_rate": 5.0967583647683015e-05,
      "loss": 0.5107,
      "mean_token_accuracy": 0.8419300727546215,
      "num_tokens": 510297455.0,
      "step": 15989
    },
    {
      "entropy": 0.6529263909906149,
      "epoch": 1.4711646281504995,
      "grad_norm": 0.15816810727119446,
      "learning_rate": 5.096451682154139e-05,
      "loss": 0.6453,
      "mean_token_accuracy": 0.806100819259882,
      "num_tokens": 510329315.0,
      "step": 15990
    },
    {
      "entropy": 0.5095427678897977,
      "epoch": 1.4712566345219407,
      "grad_norm": 0.14546532928943634,
      "learning_rate": 5.096144999539977e-05,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8450485207140446,
      "num_tokens": 510361498.0,
      "step": 15991
    },
    {
      "entropy": 0.6180995274335146,
      "epoch": 1.4713486408933818,
      "grad_norm": 0.1589595526456833,
      "learning_rate": 5.095838316925814e-05,
      "loss": 0.6077,
      "mean_token_accuracy": 0.8121523484587669,
      "num_tokens": 510393363.0,
      "step": 15992
    },
    {
      "entropy": 0.5356566249392927,
      "epoch": 1.4714406472648232,
      "grad_norm": 0.146757572889328,
      "learning_rate": 5.095531634311651e-05,
      "loss": 0.5324,
      "mean_token_accuracy": 0.8386358022689819,
      "num_tokens": 510426131.0,
      "step": 15993
    },
    {
      "entropy": 0.7396923992782831,
      "epoch": 1.4715326536362643,
      "grad_norm": 0.17426709830760956,
      "learning_rate": 5.095224951697488e-05,
      "loss": 0.7373,
      "mean_token_accuracy": 0.7747406847774982,
      "num_tokens": 510457240.0,
      "step": 15994
    },
    {
      "entropy": 0.5260018771514297,
      "epoch": 1.4716246600077056,
      "grad_norm": 0.15518806874752045,
      "learning_rate": 5.094918269083326e-05,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8431943282485008,
      "num_tokens": 510488833.0,
      "step": 15995
    },
    {
      "entropy": 0.6848260816186666,
      "epoch": 1.4717166663791468,
      "grad_norm": 0.15965278446674347,
      "learning_rate": 5.094611586469164e-05,
      "loss": 0.678,
      "mean_token_accuracy": 0.7973788529634476,
      "num_tokens": 510521030.0,
      "step": 15996
    },
    {
      "entropy": 0.5671626813709736,
      "epoch": 1.471808672750588,
      "grad_norm": 0.15450580418109894,
      "learning_rate": 5.0943049038550005e-05,
      "loss": 0.5495,
      "mean_token_accuracy": 0.8311156146228313,
      "num_tokens": 510552378.0,
      "step": 15997
    },
    {
      "entropy": 0.5645353021100163,
      "epoch": 1.4719006791220293,
      "grad_norm": 0.15685118734836578,
      "learning_rate": 5.093998221240838e-05,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8306384198367596,
      "num_tokens": 510584725.0,
      "step": 15998
    },
    {
      "entropy": 0.5389210518915206,
      "epoch": 1.4719926854934704,
      "grad_norm": 0.1501077264547348,
      "learning_rate": 5.093691538626676e-05,
      "loss": 0.5252,
      "mean_token_accuracy": 0.834149856120348,
      "num_tokens": 510616401.0,
      "step": 15999
    },
    {
      "entropy": 0.6647173259407282,
      "epoch": 1.4720846918649118,
      "grad_norm": 0.16269034147262573,
      "learning_rate": 5.093384856012513e-05,
      "loss": 0.6523,
      "mean_token_accuracy": 0.7955413870513439,
      "num_tokens": 510648038.0,
      "step": 16000
    },
    {
      "entropy": 0.5823867348954082,
      "epoch": 1.4721766982363529,
      "grad_norm": 0.15429380536079407,
      "learning_rate": 5.0930781733983504e-05,
      "loss": 0.5614,
      "mean_token_accuracy": 0.826870072633028,
      "num_tokens": 510679887.0,
      "step": 16001
    },
    {
      "entropy": 0.7015952728688717,
      "epoch": 1.472268704607794,
      "grad_norm": 0.1657872200012207,
      "learning_rate": 5.092771490784187e-05,
      "loss": 0.6943,
      "mean_token_accuracy": 0.7862445376813412,
      "num_tokens": 510711239.0,
      "step": 16002
    },
    {
      "entropy": 0.7133355792611837,
      "epoch": 1.4723607109792354,
      "grad_norm": 0.16099214553833008,
      "learning_rate": 5.092464808170025e-05,
      "loss": 0.7111,
      "mean_token_accuracy": 0.7855964154005051,
      "num_tokens": 510743573.0,
      "step": 16003
    },
    {
      "entropy": 0.6791190537624061,
      "epoch": 1.4724527173506765,
      "grad_norm": 0.16803361475467682,
      "learning_rate": 5.092158125555863e-05,
      "loss": 0.668,
      "mean_token_accuracy": 0.796227939426899,
      "num_tokens": 510775354.0,
      "step": 16004
    },
    {
      "entropy": 0.6214567627757788,
      "epoch": 1.4725447237221179,
      "grad_norm": 0.15450800955295563,
      "learning_rate": 5.0918514429416995e-05,
      "loss": 0.612,
      "mean_token_accuracy": 0.8100589960813522,
      "num_tokens": 510808113.0,
      "step": 16005
    },
    {
      "entropy": 0.46473312890157104,
      "epoch": 1.472636730093559,
      "grad_norm": 0.15197256207466125,
      "learning_rate": 5.091544760327537e-05,
      "loss": 0.4514,
      "mean_token_accuracy": 0.8591911457479,
      "num_tokens": 510839780.0,
      "step": 16006
    },
    {
      "entropy": 0.5536595657467842,
      "epoch": 1.4727287364650001,
      "grad_norm": 0.1482667326927185,
      "learning_rate": 5.091238077713375e-05,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8375285528600216,
      "num_tokens": 510871863.0,
      "step": 16007
    },
    {
      "entropy": 0.6084594111889601,
      "epoch": 1.4728207428364415,
      "grad_norm": 0.1580580174922943,
      "learning_rate": 5.090931395099212e-05,
      "loss": 0.5867,
      "mean_token_accuracy": 0.8172558955848217,
      "num_tokens": 510903624.0,
      "step": 16008
    },
    {
      "entropy": 0.5670912554487586,
      "epoch": 1.4729127492078826,
      "grad_norm": 0.15518389642238617,
      "learning_rate": 5.0906247124850494e-05,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8283307924866676,
      "num_tokens": 510936198.0,
      "step": 16009
    },
    {
      "entropy": 0.5904342709109187,
      "epoch": 1.473004755579324,
      "grad_norm": 0.15387026965618134,
      "learning_rate": 5.090318029870886e-05,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8237414881587029,
      "num_tokens": 510968892.0,
      "step": 16010
    },
    {
      "entropy": 0.7020953642204404,
      "epoch": 1.473096761950765,
      "grad_norm": 0.16638270020484924,
      "learning_rate": 5.090011347256724e-05,
      "loss": 0.7099,
      "mean_token_accuracy": 0.786855548620224,
      "num_tokens": 511000742.0,
      "step": 16011
    },
    {
      "entropy": 0.6071746572852135,
      "epoch": 1.4731887683222062,
      "grad_norm": 0.15898069739341736,
      "learning_rate": 5.089704664642562e-05,
      "loss": 0.5951,
      "mean_token_accuracy": 0.8167122006416321,
      "num_tokens": 511032802.0,
      "step": 16012
    },
    {
      "entropy": 0.6249652290716767,
      "epoch": 1.4732807746936476,
      "grad_norm": 0.16602225601673126,
      "learning_rate": 5.0893979820283986e-05,
      "loss": 0.6118,
      "mean_token_accuracy": 0.8130382895469666,
      "num_tokens": 511064356.0,
      "step": 16013
    },
    {
      "entropy": 0.5452067730948329,
      "epoch": 1.4733727810650887,
      "grad_norm": 0.15065303444862366,
      "learning_rate": 5.089091299414236e-05,
      "loss": 0.5347,
      "mean_token_accuracy": 0.8352636098861694,
      "num_tokens": 511096328.0,
      "step": 16014
    },
    {
      "entropy": 0.6050297697074711,
      "epoch": 1.47346478743653,
      "grad_norm": 0.1606193333864212,
      "learning_rate": 5.088784616800074e-05,
      "loss": 0.6049,
      "mean_token_accuracy": 0.8129940144717693,
      "num_tokens": 511128555.0,
      "step": 16015
    },
    {
      "entropy": 0.5676583878230304,
      "epoch": 1.4735567938079712,
      "grad_norm": 0.15260358154773712,
      "learning_rate": 5.088477934185911e-05,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8301826417446136,
      "num_tokens": 511160926.0,
      "step": 16016
    },
    {
      "entropy": 0.5831018257886171,
      "epoch": 1.4736488001794124,
      "grad_norm": 0.15344996750354767,
      "learning_rate": 5.0881712515717484e-05,
      "loss": 0.5814,
      "mean_token_accuracy": 0.8259548880159855,
      "num_tokens": 511193018.0,
      "step": 16017
    },
    {
      "entropy": 0.47388966800644994,
      "epoch": 1.4737408065508537,
      "grad_norm": 0.15264129638671875,
      "learning_rate": 5.0878645689575866e-05,
      "loss": 0.468,
      "mean_token_accuracy": 0.8559945970773697,
      "num_tokens": 511225709.0,
      "step": 16018
    },
    {
      "entropy": 0.6918405573815107,
      "epoch": 1.4738328129222948,
      "grad_norm": 0.16826236248016357,
      "learning_rate": 5.0875578863434234e-05,
      "loss": 0.7002,
      "mean_token_accuracy": 0.786638006567955,
      "num_tokens": 511258062.0,
      "step": 16019
    },
    {
      "entropy": 0.6150410380214453,
      "epoch": 1.4739248192937362,
      "grad_norm": 0.15197192132472992,
      "learning_rate": 5.087251203729261e-05,
      "loss": 0.6001,
      "mean_token_accuracy": 0.8132493235170841,
      "num_tokens": 511289591.0,
      "step": 16020
    },
    {
      "entropy": 0.6398514490574598,
      "epoch": 1.4740168256651773,
      "grad_norm": 0.16062603890895844,
      "learning_rate": 5.0869445211150976e-05,
      "loss": 0.6389,
      "mean_token_accuracy": 0.8034036569297314,
      "num_tokens": 511321570.0,
      "step": 16021
    },
    {
      "entropy": 0.6207520607858896,
      "epoch": 1.4741088320366185,
      "grad_norm": 0.14865200221538544,
      "learning_rate": 5.0866378385009364e-05,
      "loss": 0.607,
      "mean_token_accuracy": 0.8098250441253185,
      "num_tokens": 511353746.0,
      "step": 16022
    },
    {
      "entropy": 0.5678999265655875,
      "epoch": 1.4742008384080598,
      "grad_norm": 0.15562909841537476,
      "learning_rate": 5.086331155886773e-05,
      "loss": 0.5612,
      "mean_token_accuracy": 0.8262464180588722,
      "num_tokens": 511385298.0,
      "step": 16023
    },
    {
      "entropy": 0.6963678542524576,
      "epoch": 1.474292844779501,
      "grad_norm": 0.16092480719089508,
      "learning_rate": 5.08602447327261e-05,
      "loss": 0.6704,
      "mean_token_accuracy": 0.7997297570109367,
      "num_tokens": 511417598.0,
      "step": 16024
    },
    {
      "entropy": 0.49863080494105816,
      "epoch": 1.4743848511509423,
      "grad_norm": 0.1501077562570572,
      "learning_rate": 5.0857177906584475e-05,
      "loss": 0.4932,
      "mean_token_accuracy": 0.8490837179124355,
      "num_tokens": 511449758.0,
      "step": 16025
    },
    {
      "entropy": 0.6100638774223626,
      "epoch": 1.4744768575223834,
      "grad_norm": 0.15779484808444977,
      "learning_rate": 5.0854111080442856e-05,
      "loss": 0.5868,
      "mean_token_accuracy": 0.8198613300919533,
      "num_tokens": 511481517.0,
      "step": 16026
    },
    {
      "entropy": 0.6425882475450635,
      "epoch": 1.4745688638938246,
      "grad_norm": 0.1600872278213501,
      "learning_rate": 5.085104425430123e-05,
      "loss": 0.6165,
      "mean_token_accuracy": 0.8086658120155334,
      "num_tokens": 511513038.0,
      "step": 16027
    },
    {
      "entropy": 0.5638876943849027,
      "epoch": 1.474660870265266,
      "grad_norm": 0.14394772052764893,
      "learning_rate": 5.08479774281596e-05,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8352477364242077,
      "num_tokens": 511544797.0,
      "step": 16028
    },
    {
      "entropy": 0.5585115645080805,
      "epoch": 1.474752876636707,
      "grad_norm": 0.15672047436237335,
      "learning_rate": 5.084491060201797e-05,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8278173692524433,
      "num_tokens": 511576930.0,
      "step": 16029
    },
    {
      "entropy": 0.5890337750315666,
      "epoch": 1.4748448830081484,
      "grad_norm": 0.1583382487297058,
      "learning_rate": 5.0841843775876355e-05,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8203676752746105,
      "num_tokens": 511608743.0,
      "step": 16030
    },
    {
      "entropy": 0.5282049048691988,
      "epoch": 1.4749368893795896,
      "grad_norm": 0.14596010744571686,
      "learning_rate": 5.083877694973472e-05,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8455988317728043,
      "num_tokens": 511641278.0,
      "step": 16031
    },
    {
      "entropy": 0.5050257137045264,
      "epoch": 1.4750288957510307,
      "grad_norm": 0.14993366599082947,
      "learning_rate": 5.08357101235931e-05,
      "loss": 0.4926,
      "mean_token_accuracy": 0.844811350107193,
      "num_tokens": 511673170.0,
      "step": 16032
    },
    {
      "entropy": 0.654677658341825,
      "epoch": 1.475120902122472,
      "grad_norm": 0.16142508387565613,
      "learning_rate": 5.0832643297451465e-05,
      "loss": 0.6561,
      "mean_token_accuracy": 0.8024044185876846,
      "num_tokens": 511705375.0,
      "step": 16033
    },
    {
      "entropy": 0.5171549543738365,
      "epoch": 1.4752129084939132,
      "grad_norm": 0.15596584975719452,
      "learning_rate": 5.0829576471309847e-05,
      "loss": 0.512,
      "mean_token_accuracy": 0.8416088931262493,
      "num_tokens": 511737342.0,
      "step": 16034
    },
    {
      "entropy": 0.7229429110884666,
      "epoch": 1.4753049148653545,
      "grad_norm": 0.16646341979503632,
      "learning_rate": 5.082650964516822e-05,
      "loss": 0.7233,
      "mean_token_accuracy": 0.7824760861694813,
      "num_tokens": 511769254.0,
      "step": 16035
    },
    {
      "entropy": 0.6285952590405941,
      "epoch": 1.4753969212367957,
      "grad_norm": 0.16415785253047943,
      "learning_rate": 5.082344281902659e-05,
      "loss": 0.6305,
      "mean_token_accuracy": 0.812701515853405,
      "num_tokens": 511801186.0,
      "step": 16036
    },
    {
      "entropy": 0.6511784130707383,
      "epoch": 1.4754889276082368,
      "grad_norm": 0.16272473335266113,
      "learning_rate": 5.0820375992884964e-05,
      "loss": 0.6345,
      "mean_token_accuracy": 0.8063649721443653,
      "num_tokens": 511833141.0,
      "step": 16037
    },
    {
      "entropy": 0.5512428961228579,
      "epoch": 1.4755809339796782,
      "grad_norm": 0.15701092779636383,
      "learning_rate": 5.0817309166743345e-05,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8345290571451187,
      "num_tokens": 511864640.0,
      "step": 16038
    },
    {
      "entropy": 0.5422277469187975,
      "epoch": 1.4756729403511193,
      "grad_norm": 0.15512889623641968,
      "learning_rate": 5.081424234060171e-05,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8341970667243004,
      "num_tokens": 511896822.0,
      "step": 16039
    },
    {
      "entropy": 0.6271087545901537,
      "epoch": 1.4757649467225606,
      "grad_norm": 0.15699146687984467,
      "learning_rate": 5.081117551446009e-05,
      "loss": 0.6194,
      "mean_token_accuracy": 0.8096061125397682,
      "num_tokens": 511929429.0,
      "step": 16040
    },
    {
      "entropy": 0.5825403584167361,
      "epoch": 1.4758569530940018,
      "grad_norm": 0.14764384925365448,
      "learning_rate": 5.0808108688318456e-05,
      "loss": 0.5576,
      "mean_token_accuracy": 0.8274793215095997,
      "num_tokens": 511961680.0,
      "step": 16041
    },
    {
      "entropy": 0.5429397313855588,
      "epoch": 1.475948959465443,
      "grad_norm": 0.1588088721036911,
      "learning_rate": 5.080504186217684e-05,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8343306295573711,
      "num_tokens": 511993587.0,
      "step": 16042
    },
    {
      "entropy": 0.6923926752060652,
      "epoch": 1.4760409658368843,
      "grad_norm": 0.16473114490509033,
      "learning_rate": 5.080197503603521e-05,
      "loss": 0.6882,
      "mean_token_accuracy": 0.7864477150142193,
      "num_tokens": 512025187.0,
      "step": 16043
    },
    {
      "entropy": 0.6878744028508663,
      "epoch": 1.4761329722083254,
      "grad_norm": 0.16651350259780884,
      "learning_rate": 5.079890820989358e-05,
      "loss": 0.6781,
      "mean_token_accuracy": 0.7937735691666603,
      "num_tokens": 512056849.0,
      "step": 16044
    },
    {
      "entropy": 0.596764157526195,
      "epoch": 1.4762249785797668,
      "grad_norm": 0.15020789206027985,
      "learning_rate": 5.0795841383751954e-05,
      "loss": 0.5808,
      "mean_token_accuracy": 0.824017234146595,
      "num_tokens": 512089610.0,
      "step": 16045
    },
    {
      "entropy": 0.6227611498907208,
      "epoch": 1.476316984951208,
      "grad_norm": 0.16003867983818054,
      "learning_rate": 5.0792774557610336e-05,
      "loss": 0.6113,
      "mean_token_accuracy": 0.8089326582849026,
      "num_tokens": 512121094.0,
      "step": 16046
    },
    {
      "entropy": 0.6221178816631436,
      "epoch": 1.476408991322649,
      "grad_norm": 0.16766490042209625,
      "learning_rate": 5.0789707731468703e-05,
      "loss": 0.6161,
      "mean_token_accuracy": 0.8108029216527939,
      "num_tokens": 512152750.0,
      "step": 16047
    },
    {
      "entropy": 0.6618838608264923,
      "epoch": 1.4765009976940904,
      "grad_norm": 0.1642068773508072,
      "learning_rate": 5.078664090532708e-05,
      "loss": 0.6663,
      "mean_token_accuracy": 0.8000991269946098,
      "num_tokens": 512184684.0,
      "step": 16048
    },
    {
      "entropy": 0.6490077897906303,
      "epoch": 1.4765930040655315,
      "grad_norm": 0.1613558977842331,
      "learning_rate": 5.078357407918546e-05,
      "loss": 0.6423,
      "mean_token_accuracy": 0.7987562902271748,
      "num_tokens": 512216742.0,
      "step": 16049
    },
    {
      "entropy": 0.6791489291936159,
      "epoch": 1.4766850104369729,
      "grad_norm": 0.15518581867218018,
      "learning_rate": 5.078050725304383e-05,
      "loss": 0.6723,
      "mean_token_accuracy": 0.7968743406236172,
      "num_tokens": 512248880.0,
      "step": 16050
    },
    {
      "entropy": 0.5941023956984282,
      "epoch": 1.476777016808414,
      "grad_norm": 0.16363653540611267,
      "learning_rate": 5.07774404269022e-05,
      "loss": 0.592,
      "mean_token_accuracy": 0.8207066580653191,
      "num_tokens": 512280632.0,
      "step": 16051
    },
    {
      "entropy": 0.572066280990839,
      "epoch": 1.4768690231798551,
      "grad_norm": 0.15602664649486542,
      "learning_rate": 5.077437360076057e-05,
      "loss": 0.5566,
      "mean_token_accuracy": 0.8286199532449245,
      "num_tokens": 512311971.0,
      "step": 16052
    },
    {
      "entropy": 0.6454997677356005,
      "epoch": 1.4769610295512965,
      "grad_norm": 0.1588660180568695,
      "learning_rate": 5.077130677461895e-05,
      "loss": 0.6317,
      "mean_token_accuracy": 0.8055740147829056,
      "num_tokens": 512343501.0,
      "step": 16053
    },
    {
      "entropy": 0.48423856403678656,
      "epoch": 1.4770530359227376,
      "grad_norm": 0.14541518688201904,
      "learning_rate": 5.0768239948477326e-05,
      "loss": 0.4725,
      "mean_token_accuracy": 0.8547562509775162,
      "num_tokens": 512375561.0,
      "step": 16054
    },
    {
      "entropy": 0.5898207849822938,
      "epoch": 1.477145042294179,
      "grad_norm": 0.150970458984375,
      "learning_rate": 5.0765173122335694e-05,
      "loss": 0.5872,
      "mean_token_accuracy": 0.823562853038311,
      "num_tokens": 512408157.0,
      "step": 16055
    },
    {
      "entropy": 0.7079749908298254,
      "epoch": 1.4772370486656201,
      "grad_norm": 0.16034115850925446,
      "learning_rate": 5.076210629619407e-05,
      "loss": 0.6855,
      "mean_token_accuracy": 0.7925847694277763,
      "num_tokens": 512439815.0,
      "step": 16056
    },
    {
      "entropy": 0.5237994166091084,
      "epoch": 1.4773290550370612,
      "grad_norm": 0.1529180109500885,
      "learning_rate": 5.075903947005245e-05,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8371656984090805,
      "num_tokens": 512472084.0,
      "step": 16057
    },
    {
      "entropy": 0.46568831615149975,
      "epoch": 1.4774210614085026,
      "grad_norm": 0.14589619636535645,
      "learning_rate": 5.075597264391082e-05,
      "loss": 0.4462,
      "mean_token_accuracy": 0.8580259084701538,
      "num_tokens": 512503273.0,
      "step": 16058
    },
    {
      "entropy": 0.5404984978958964,
      "epoch": 1.4775130677799437,
      "grad_norm": 0.15153701603412628,
      "learning_rate": 5.075290581776919e-05,
      "loss": 0.518,
      "mean_token_accuracy": 0.8395073115825653,
      "num_tokens": 512534853.0,
      "step": 16059
    },
    {
      "entropy": 0.6163366592954844,
      "epoch": 1.477605074151385,
      "grad_norm": 0.15242986381053925,
      "learning_rate": 5.074983899162756e-05,
      "loss": 0.5988,
      "mean_token_accuracy": 0.8160647936165333,
      "num_tokens": 512566802.0,
      "step": 16060
    },
    {
      "entropy": 0.6118268673308194,
      "epoch": 1.4776970805228262,
      "grad_norm": 0.1565944403409958,
      "learning_rate": 5.074677216548595e-05,
      "loss": 0.5913,
      "mean_token_accuracy": 0.8207729794085026,
      "num_tokens": 512598256.0,
      "step": 16061
    },
    {
      "entropy": 0.5668819807469845,
      "epoch": 1.4777890868942674,
      "grad_norm": 0.15182992815971375,
      "learning_rate": 5.0743705339344316e-05,
      "loss": 0.5519,
      "mean_token_accuracy": 0.8271508291363716,
      "num_tokens": 512630580.0,
      "step": 16062
    },
    {
      "entropy": 0.5890184305608273,
      "epoch": 1.4778810932657087,
      "grad_norm": 0.16358181834220886,
      "learning_rate": 5.0740638513202684e-05,
      "loss": 0.5727,
      "mean_token_accuracy": 0.8246911019086838,
      "num_tokens": 512661820.0,
      "step": 16063
    },
    {
      "entropy": 0.5608248431235552,
      "epoch": 1.4779730996371498,
      "grad_norm": 0.1584094911813736,
      "learning_rate": 5.073757168706106e-05,
      "loss": 0.5535,
      "mean_token_accuracy": 0.8348924070596695,
      "num_tokens": 512693032.0,
      "step": 16064
    },
    {
      "entropy": 0.6138962786644697,
      "epoch": 1.4780651060085912,
      "grad_norm": 0.16663378477096558,
      "learning_rate": 5.073450486091944e-05,
      "loss": 0.6135,
      "mean_token_accuracy": 0.8158909045159817,
      "num_tokens": 512724965.0,
      "step": 16065
    },
    {
      "entropy": 0.5944944024085999,
      "epoch": 1.4781571123800323,
      "grad_norm": 0.15987341105937958,
      "learning_rate": 5.0731438034777815e-05,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8195530250668526,
      "num_tokens": 512757326.0,
      "step": 16066
    },
    {
      "entropy": 0.6159391161054373,
      "epoch": 1.4782491187514735,
      "grad_norm": 0.1619202047586441,
      "learning_rate": 5.072837120863618e-05,
      "loss": 0.595,
      "mean_token_accuracy": 0.8175023347139359,
      "num_tokens": 512789001.0,
      "step": 16067
    },
    {
      "entropy": 0.6353000123053789,
      "epoch": 1.4783411251229148,
      "grad_norm": 0.16055789589881897,
      "learning_rate": 5.072530438249455e-05,
      "loss": 0.6242,
      "mean_token_accuracy": 0.8119189999997616,
      "num_tokens": 512820589.0,
      "step": 16068
    },
    {
      "entropy": 0.6006145514547825,
      "epoch": 1.478433131494356,
      "grad_norm": 0.15889111161231995,
      "learning_rate": 5.072223755635294e-05,
      "loss": 0.6075,
      "mean_token_accuracy": 0.8136765323579311,
      "num_tokens": 512853128.0,
      "step": 16069
    },
    {
      "entropy": 0.613546334207058,
      "epoch": 1.4785251378657973,
      "grad_norm": 0.15619328618049622,
      "learning_rate": 5.071917073021131e-05,
      "loss": 0.6004,
      "mean_token_accuracy": 0.8172740526497364,
      "num_tokens": 512884786.0,
      "step": 16070
    },
    {
      "entropy": 0.6926618032157421,
      "epoch": 1.4786171442372384,
      "grad_norm": 0.16801555454730988,
      "learning_rate": 5.071610390406968e-05,
      "loss": 0.6948,
      "mean_token_accuracy": 0.7904590852558613,
      "num_tokens": 512916550.0,
      "step": 16071
    },
    {
      "entropy": 0.49046480376273394,
      "epoch": 1.4787091506086796,
      "grad_norm": 0.14621412754058838,
      "learning_rate": 5.071303707792805e-05,
      "loss": 0.4737,
      "mean_token_accuracy": 0.8563446514308453,
      "num_tokens": 512949318.0,
      "step": 16072
    },
    {
      "entropy": 0.6099044242873788,
      "epoch": 1.478801156980121,
      "grad_norm": 0.16106480360031128,
      "learning_rate": 5.070997025178643e-05,
      "loss": 0.605,
      "mean_token_accuracy": 0.8161625601351261,
      "num_tokens": 512981050.0,
      "step": 16073
    },
    {
      "entropy": 0.5383069100789726,
      "epoch": 1.478893163351562,
      "grad_norm": 0.15463410317897797,
      "learning_rate": 5.0706903425644805e-05,
      "loss": 0.5286,
      "mean_token_accuracy": 0.8356892541050911,
      "num_tokens": 513012371.0,
      "step": 16074
    },
    {
      "entropy": 0.6671279259026051,
      "epoch": 1.4789851697230034,
      "grad_norm": 0.1571977734565735,
      "learning_rate": 5.070383659950317e-05,
      "loss": 0.6528,
      "mean_token_accuracy": 0.7963910885155201,
      "num_tokens": 513043748.0,
      "step": 16075
    },
    {
      "entropy": 0.654227240011096,
      "epoch": 1.4790771760944446,
      "grad_norm": 0.16773264110088348,
      "learning_rate": 5.0700769773361555e-05,
      "loss": 0.6259,
      "mean_token_accuracy": 0.8098270036280155,
      "num_tokens": 513075984.0,
      "step": 16076
    },
    {
      "entropy": 0.5857298420742154,
      "epoch": 1.4791691824658857,
      "grad_norm": 0.15290652215480804,
      "learning_rate": 5.069770294721993e-05,
      "loss": 0.5703,
      "mean_token_accuracy": 0.8263112381100655,
      "num_tokens": 513107580.0,
      "step": 16077
    },
    {
      "entropy": 0.6203469838947058,
      "epoch": 1.479261188837327,
      "grad_norm": 0.16007563471794128,
      "learning_rate": 5.06946361210783e-05,
      "loss": 0.624,
      "mean_token_accuracy": 0.8103689961135387,
      "num_tokens": 513139109.0,
      "step": 16078
    },
    {
      "entropy": 0.5672541549429297,
      "epoch": 1.4793531952087682,
      "grad_norm": 0.15374596416950226,
      "learning_rate": 5.069156929493667e-05,
      "loss": 0.5612,
      "mean_token_accuracy": 0.8273997530341148,
      "num_tokens": 513170928.0,
      "step": 16079
    },
    {
      "entropy": 0.6822249256074429,
      "epoch": 1.4794452015802095,
      "grad_norm": 0.16600903868675232,
      "learning_rate": 5.068850246879505e-05,
      "loss": 0.652,
      "mean_token_accuracy": 0.7997731976211071,
      "num_tokens": 513202426.0,
      "step": 16080
    },
    {
      "entropy": 0.5636249668896198,
      "epoch": 1.4795372079516507,
      "grad_norm": 0.15861636400222778,
      "learning_rate": 5.068543564265342e-05,
      "loss": 0.5713,
      "mean_token_accuracy": 0.8236238770186901,
      "num_tokens": 513234991.0,
      "step": 16081
    },
    {
      "entropy": 0.6402336098253727,
      "epoch": 1.4796292143230918,
      "grad_norm": 0.16340944170951843,
      "learning_rate": 5.0682368816511796e-05,
      "loss": 0.6429,
      "mean_token_accuracy": 0.8004857450723648,
      "num_tokens": 513265687.0,
      "step": 16082
    },
    {
      "entropy": 0.5682840747758746,
      "epoch": 1.4797212206945332,
      "grad_norm": 0.14946527779102325,
      "learning_rate": 5.0679301990370164e-05,
      "loss": 0.5506,
      "mean_token_accuracy": 0.826967116445303,
      "num_tokens": 513297752.0,
      "step": 16083
    },
    {
      "entropy": 0.586499304510653,
      "epoch": 1.4798132270659743,
      "grad_norm": 0.1685948222875595,
      "learning_rate": 5.0676235164228545e-05,
      "loss": 0.6007,
      "mean_token_accuracy": 0.8170319125056267,
      "num_tokens": 513329848.0,
      "step": 16084
    },
    {
      "entropy": 0.6395126283168793,
      "epoch": 1.4799052334374156,
      "grad_norm": 0.15803097188472748,
      "learning_rate": 5.067316833808692e-05,
      "loss": 0.6242,
      "mean_token_accuracy": 0.8127102628350258,
      "num_tokens": 513361682.0,
      "step": 16085
    },
    {
      "entropy": 0.567735911346972,
      "epoch": 1.4799972398088568,
      "grad_norm": 0.15744444727897644,
      "learning_rate": 5.067010151194529e-05,
      "loss": 0.5592,
      "mean_token_accuracy": 0.83447440341115,
      "num_tokens": 513393634.0,
      "step": 16086
    },
    {
      "entropy": 0.7063456792384386,
      "epoch": 1.480089246180298,
      "grad_norm": 0.1604876071214676,
      "learning_rate": 5.066703468580366e-05,
      "loss": 0.7014,
      "mean_token_accuracy": 0.7845691964030266,
      "num_tokens": 513425599.0,
      "step": 16087
    },
    {
      "entropy": 0.6159630622714758,
      "epoch": 1.4801812525517393,
      "grad_norm": 0.15456119179725647,
      "learning_rate": 5.0663967859662044e-05,
      "loss": 0.5994,
      "mean_token_accuracy": 0.8173937983810902,
      "num_tokens": 513458210.0,
      "step": 16088
    },
    {
      "entropy": 0.6553422892466187,
      "epoch": 1.4802732589231804,
      "grad_norm": 0.16407077014446259,
      "learning_rate": 5.066090103352041e-05,
      "loss": 0.643,
      "mean_token_accuracy": 0.8058932386338711,
      "num_tokens": 513490103.0,
      "step": 16089
    },
    {
      "entropy": 0.646432138979435,
      "epoch": 1.4803652652946218,
      "grad_norm": 0.14927512407302856,
      "learning_rate": 5.0657834207378786e-05,
      "loss": 0.625,
      "mean_token_accuracy": 0.8092367462813854,
      "num_tokens": 513522174.0,
      "step": 16090
    },
    {
      "entropy": 0.6345349624752998,
      "epoch": 1.480457271666063,
      "grad_norm": 0.16453760862350464,
      "learning_rate": 5.0654767381237154e-05,
      "loss": 0.6142,
      "mean_token_accuracy": 0.815247505903244,
      "num_tokens": 513553995.0,
      "step": 16091
    },
    {
      "entropy": 0.5793380802497268,
      "epoch": 1.480549278037504,
      "grad_norm": 0.1510246843099594,
      "learning_rate": 5.0651700555095535e-05,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8243974186480045,
      "num_tokens": 513585900.0,
      "step": 16092
    },
    {
      "entropy": 0.4880145248025656,
      "epoch": 1.4806412844089454,
      "grad_norm": 0.1437702775001526,
      "learning_rate": 5.064863372895391e-05,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8516700118780136,
      "num_tokens": 513618145.0,
      "step": 16093
    },
    {
      "entropy": 0.574979405850172,
      "epoch": 1.4807332907803865,
      "grad_norm": 0.15494520962238312,
      "learning_rate": 5.064556690281228e-05,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8226847648620605,
      "num_tokens": 513650230.0,
      "step": 16094
    },
    {
      "entropy": 0.5639351047575474,
      "epoch": 1.4808252971518279,
      "grad_norm": 0.1517832726240158,
      "learning_rate": 5.064250007667065e-05,
      "loss": 0.5425,
      "mean_token_accuracy": 0.8292075134813786,
      "num_tokens": 513681364.0,
      "step": 16095
    },
    {
      "entropy": 0.7478372976183891,
      "epoch": 1.480917303523269,
      "grad_norm": 0.17402683198451996,
      "learning_rate": 5.0639433250529034e-05,
      "loss": 0.7435,
      "mean_token_accuracy": 0.7748860009014606,
      "num_tokens": 513713151.0,
      "step": 16096
    },
    {
      "entropy": 0.6142439264804125,
      "epoch": 1.4810093098947101,
      "grad_norm": 0.1598377525806427,
      "learning_rate": 5.06363664243874e-05,
      "loss": 0.6095,
      "mean_token_accuracy": 0.8159720785915852,
      "num_tokens": 513744803.0,
      "step": 16097
    },
    {
      "entropy": 0.64082121104002,
      "epoch": 1.4811013162661515,
      "grad_norm": 0.16152651607990265,
      "learning_rate": 5.0633299598245777e-05,
      "loss": 0.6225,
      "mean_token_accuracy": 0.8093561641871929,
      "num_tokens": 513776034.0,
      "step": 16098
    },
    {
      "entropy": 0.6537200431339443,
      "epoch": 1.4811933226375926,
      "grad_norm": 0.1545664668083191,
      "learning_rate": 5.0630232772104144e-05,
      "loss": 0.6368,
      "mean_token_accuracy": 0.807172030210495,
      "num_tokens": 513808727.0,
      "step": 16099
    },
    {
      "entropy": 0.6443739980459213,
      "epoch": 1.481285329009034,
      "grad_norm": 0.1583775281906128,
      "learning_rate": 5.0627165945962526e-05,
      "loss": 0.6304,
      "mean_token_accuracy": 0.8022038266062737,
      "num_tokens": 513840452.0,
      "step": 16100
    },
    {
      "entropy": 0.594266970641911,
      "epoch": 1.4813773353804751,
      "grad_norm": 0.15524567663669586,
      "learning_rate": 5.06240991198209e-05,
      "loss": 0.5885,
      "mean_token_accuracy": 0.8197060935199261,
      "num_tokens": 513871150.0,
      "step": 16101
    },
    {
      "entropy": 0.6418464705348015,
      "epoch": 1.4814693417519162,
      "grad_norm": 0.1570194959640503,
      "learning_rate": 5.062103229367927e-05,
      "loss": 0.6309,
      "mean_token_accuracy": 0.8035012446343899,
      "num_tokens": 513903219.0,
      "step": 16102
    },
    {
      "entropy": 0.5540559543296695,
      "epoch": 1.4815613481233576,
      "grad_norm": 0.15669570863246918,
      "learning_rate": 5.061796546753764e-05,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8353727236390114,
      "num_tokens": 513934865.0,
      "step": 16103
    },
    {
      "entropy": 0.5573987108655274,
      "epoch": 1.4816533544947987,
      "grad_norm": 0.15103064477443695,
      "learning_rate": 5.0614898641396024e-05,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8356275744736195,
      "num_tokens": 513966724.0,
      "step": 16104
    },
    {
      "entropy": 0.6696040173992515,
      "epoch": 1.48174536086624,
      "grad_norm": 0.1575339287519455,
      "learning_rate": 5.06118318152544e-05,
      "loss": 0.6576,
      "mean_token_accuracy": 0.8005973957479,
      "num_tokens": 513999272.0,
      "step": 16105
    },
    {
      "entropy": 0.665422422811389,
      "epoch": 1.4818373672376812,
      "grad_norm": 0.1557517647743225,
      "learning_rate": 5.060876498911277e-05,
      "loss": 0.6524,
      "mean_token_accuracy": 0.8025294058024883,
      "num_tokens": 514031326.0,
      "step": 16106
    },
    {
      "entropy": 0.6697114491835237,
      "epoch": 1.4819293736091224,
      "grad_norm": 0.16216912865638733,
      "learning_rate": 5.060569816297115e-05,
      "loss": 0.6726,
      "mean_token_accuracy": 0.7947506010532379,
      "num_tokens": 514063355.0,
      "step": 16107
    },
    {
      "entropy": 0.5655484087765217,
      "epoch": 1.4820213799805637,
      "grad_norm": 0.15626157820224762,
      "learning_rate": 5.060263133682952e-05,
      "loss": 0.5562,
      "mean_token_accuracy": 0.8281633034348488,
      "num_tokens": 514095585.0,
      "step": 16108
    },
    {
      "entropy": 0.5921599715948105,
      "epoch": 1.4821133863520048,
      "grad_norm": 0.1583063006401062,
      "learning_rate": 5.059956451068789e-05,
      "loss": 0.5948,
      "mean_token_accuracy": 0.8228722363710403,
      "num_tokens": 514127767.0,
      "step": 16109
    },
    {
      "entropy": 0.6713088639080524,
      "epoch": 1.4822053927234462,
      "grad_norm": 0.16202442348003387,
      "learning_rate": 5.0596497684546266e-05,
      "loss": 0.6527,
      "mean_token_accuracy": 0.8000332713127136,
      "num_tokens": 514160043.0,
      "step": 16110
    },
    {
      "entropy": 0.6105905640870333,
      "epoch": 1.4822973990948873,
      "grad_norm": 0.158506378531456,
      "learning_rate": 5.059343085840465e-05,
      "loss": 0.6092,
      "mean_token_accuracy": 0.819517232477665,
      "num_tokens": 514191486.0,
      "step": 16111
    },
    {
      "entropy": 0.6265278905630112,
      "epoch": 1.4823894054663285,
      "grad_norm": 0.15529701113700867,
      "learning_rate": 5.0590364032263015e-05,
      "loss": 0.6148,
      "mean_token_accuracy": 0.8117177039384842,
      "num_tokens": 514223853.0,
      "step": 16112
    },
    {
      "entropy": 0.6359787220135331,
      "epoch": 1.4824814118377698,
      "grad_norm": 0.15860623121261597,
      "learning_rate": 5.058729720612139e-05,
      "loss": 0.6409,
      "mean_token_accuracy": 0.8086076229810715,
      "num_tokens": 514256263.0,
      "step": 16113
    },
    {
      "entropy": 0.4724173923023045,
      "epoch": 1.482573418209211,
      "grad_norm": 0.14564283192157745,
      "learning_rate": 5.058423037997976e-05,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8599707297980785,
      "num_tokens": 514288795.0,
      "step": 16114
    },
    {
      "entropy": 0.5131559614092112,
      "epoch": 1.4826654245806523,
      "grad_norm": 0.13600747287273407,
      "learning_rate": 5.058116355383814e-05,
      "loss": 0.4963,
      "mean_token_accuracy": 0.843426987528801,
      "num_tokens": 514321218.0,
      "step": 16115
    },
    {
      "entropy": 0.7264975570142269,
      "epoch": 1.4827574309520934,
      "grad_norm": 0.16710901260375977,
      "learning_rate": 5.0578096727696513e-05,
      "loss": 0.7057,
      "mean_token_accuracy": 0.7854963541030884,
      "num_tokens": 514353126.0,
      "step": 16116
    },
    {
      "entropy": 0.4747876198962331,
      "epoch": 1.4828494373235346,
      "grad_norm": 0.14183902740478516,
      "learning_rate": 5.057502990155488e-05,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8578038401901722,
      "num_tokens": 514385573.0,
      "step": 16117
    },
    {
      "entropy": 0.5750042553991079,
      "epoch": 1.482941443694976,
      "grad_norm": 0.1595827341079712,
      "learning_rate": 5.0571963075413256e-05,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8298050351440907,
      "num_tokens": 514417460.0,
      "step": 16118
    },
    {
      "entropy": 0.5422269124537706,
      "epoch": 1.483033450066417,
      "grad_norm": 0.15363985300064087,
      "learning_rate": 5.056889624927164e-05,
      "loss": 0.523,
      "mean_token_accuracy": 0.8354398384690285,
      "num_tokens": 514450012.0,
      "step": 16119
    },
    {
      "entropy": 0.6880043838173151,
      "epoch": 1.4831254564378584,
      "grad_norm": 0.16214703023433685,
      "learning_rate": 5.0565829423130005e-05,
      "loss": 0.6698,
      "mean_token_accuracy": 0.7945517972111702,
      "num_tokens": 514481179.0,
      "step": 16120
    },
    {
      "entropy": 0.5331525607034564,
      "epoch": 1.4832174628092996,
      "grad_norm": 0.15025027096271515,
      "learning_rate": 5.056276259698838e-05,
      "loss": 0.522,
      "mean_token_accuracy": 0.8373595289885998,
      "num_tokens": 514513053.0,
      "step": 16121
    },
    {
      "entropy": 0.569239349104464,
      "epoch": 1.4833094691807407,
      "grad_norm": 0.16097700595855713,
      "learning_rate": 5.055969577084675e-05,
      "loss": 0.5684,
      "mean_token_accuracy": 0.8265933655202389,
      "num_tokens": 514545796.0,
      "step": 16122
    },
    {
      "entropy": 0.6979657262563705,
      "epoch": 1.483401475552182,
      "grad_norm": 0.16313734650611877,
      "learning_rate": 5.055662894470513e-05,
      "loss": 0.6775,
      "mean_token_accuracy": 0.7935514077544212,
      "num_tokens": 514577534.0,
      "step": 16123
    },
    {
      "entropy": 0.5375326536595821,
      "epoch": 1.4834934819236232,
      "grad_norm": 0.1482933908700943,
      "learning_rate": 5.0553562118563504e-05,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8366243578493595,
      "num_tokens": 514609492.0,
      "step": 16124
    },
    {
      "entropy": 0.7253273725509644,
      "epoch": 1.4835854882950645,
      "grad_norm": 0.15407992899417877,
      "learning_rate": 5.055049529242187e-05,
      "loss": 0.6964,
      "mean_token_accuracy": 0.7874709405004978,
      "num_tokens": 514642004.0,
      "step": 16125
    },
    {
      "entropy": 0.5613313112407923,
      "epoch": 1.4836774946665057,
      "grad_norm": 0.15290305018424988,
      "learning_rate": 5.0547428466280246e-05,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8315282054245472,
      "num_tokens": 514674334.0,
      "step": 16126
    },
    {
      "entropy": 0.6906211413443089,
      "epoch": 1.4837695010379468,
      "grad_norm": 0.15603038668632507,
      "learning_rate": 5.054436164013863e-05,
      "loss": 0.68,
      "mean_token_accuracy": 0.7923731133341789,
      "num_tokens": 514706819.0,
      "step": 16127
    },
    {
      "entropy": 0.6027239579707384,
      "epoch": 1.4838615074093882,
      "grad_norm": 0.15506714582443237,
      "learning_rate": 5.0541294813996996e-05,
      "loss": 0.6035,
      "mean_token_accuracy": 0.8174517750740051,
      "num_tokens": 514738967.0,
      "step": 16128
    },
    {
      "entropy": 0.6013673692941666,
      "epoch": 1.4839535137808293,
      "grad_norm": 0.15954133868217468,
      "learning_rate": 5.053822798785537e-05,
      "loss": 0.5886,
      "mean_token_accuracy": 0.8164821155369282,
      "num_tokens": 514771030.0,
      "step": 16129
    },
    {
      "entropy": 0.5153239173814654,
      "epoch": 1.4840455201522706,
      "grad_norm": 0.14495468139648438,
      "learning_rate": 5.053516116171374e-05,
      "loss": 0.5068,
      "mean_token_accuracy": 0.8444551005959511,
      "num_tokens": 514803513.0,
      "step": 16130
    },
    {
      "entropy": 0.5069654192775488,
      "epoch": 1.4841375265237118,
      "grad_norm": 0.14992740750312805,
      "learning_rate": 5.053209433557212e-05,
      "loss": 0.4924,
      "mean_token_accuracy": 0.847723513841629,
      "num_tokens": 514834803.0,
      "step": 16131
    },
    {
      "entropy": 0.45381412003189325,
      "epoch": 1.484229532895153,
      "grad_norm": 0.1542329043149948,
      "learning_rate": 5.0529027509430494e-05,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8620407544076443,
      "num_tokens": 514867003.0,
      "step": 16132
    },
    {
      "entropy": 0.63799644773826,
      "epoch": 1.4843215392665943,
      "grad_norm": 0.16344577074050903,
      "learning_rate": 5.052596068328886e-05,
      "loss": 0.6396,
      "mean_token_accuracy": 0.8067412190139294,
      "num_tokens": 514898830.0,
      "step": 16133
    },
    {
      "entropy": 0.5556032340973616,
      "epoch": 1.4844135456380354,
      "grad_norm": 0.1548805981874466,
      "learning_rate": 5.052289385714724e-05,
      "loss": 0.5346,
      "mean_token_accuracy": 0.8326989449560642,
      "num_tokens": 514930509.0,
      "step": 16134
    },
    {
      "entropy": 0.5827672723680735,
      "epoch": 1.4845055520094768,
      "grad_norm": 0.15320909023284912,
      "learning_rate": 5.051982703100562e-05,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8250007554888725,
      "num_tokens": 514962309.0,
      "step": 16135
    },
    {
      "entropy": 0.5217318385839462,
      "epoch": 1.484597558380918,
      "grad_norm": 0.16112864017486572,
      "learning_rate": 5.0516760204863986e-05,
      "loss": 0.5106,
      "mean_token_accuracy": 0.8409978114068508,
      "num_tokens": 514993623.0,
      "step": 16136
    },
    {
      "entropy": 0.663711866363883,
      "epoch": 1.484689564752359,
      "grad_norm": 0.15744903683662415,
      "learning_rate": 5.051369337872236e-05,
      "loss": 0.6524,
      "mean_token_accuracy": 0.7975933998823166,
      "num_tokens": 515024965.0,
      "step": 16137
    },
    {
      "entropy": 0.6690626181662083,
      "epoch": 1.4847815711238004,
      "grad_norm": 0.163033127784729,
      "learning_rate": 5.051062655258074e-05,
      "loss": 0.6488,
      "mean_token_accuracy": 0.8003667369484901,
      "num_tokens": 515056084.0,
      "step": 16138
    },
    {
      "entropy": 0.5848624054342508,
      "epoch": 1.4848735774952415,
      "grad_norm": 0.1478581726551056,
      "learning_rate": 5.050755972643911e-05,
      "loss": 0.5807,
      "mean_token_accuracy": 0.8216748684644699,
      "num_tokens": 515088765.0,
      "step": 16139
    },
    {
      "entropy": 0.6521319542080164,
      "epoch": 1.4849655838666829,
      "grad_norm": 0.16436459124088287,
      "learning_rate": 5.0504492900297485e-05,
      "loss": 0.6264,
      "mean_token_accuracy": 0.8054182268679142,
      "num_tokens": 515119932.0,
      "step": 16140
    },
    {
      "entropy": 0.651344433426857,
      "epoch": 1.485057590238124,
      "grad_norm": 0.16235384345054626,
      "learning_rate": 5.050142607415585e-05,
      "loss": 0.6353,
      "mean_token_accuracy": 0.8068019039928913,
      "num_tokens": 515151349.0,
      "step": 16141
    },
    {
      "entropy": 0.6783323492854834,
      "epoch": 1.4851495966095651,
      "grad_norm": 0.16976462304592133,
      "learning_rate": 5.049835924801424e-05,
      "loss": 0.6748,
      "mean_token_accuracy": 0.7938030026853085,
      "num_tokens": 515183704.0,
      "step": 16142
    },
    {
      "entropy": 0.6321980310603976,
      "epoch": 1.4852416029810065,
      "grad_norm": 0.16879208385944366,
      "learning_rate": 5.049529242187261e-05,
      "loss": 0.6431,
      "mean_token_accuracy": 0.8025574795901775,
      "num_tokens": 515215257.0,
      "step": 16143
    },
    {
      "entropy": 0.5282552279531956,
      "epoch": 1.4853336093524476,
      "grad_norm": 0.15760943293571472,
      "learning_rate": 5.0492225595730976e-05,
      "loss": 0.5221,
      "mean_token_accuracy": 0.8409321159124374,
      "num_tokens": 515247424.0,
      "step": 16144
    },
    {
      "entropy": 0.5414022160694003,
      "epoch": 1.485425615723889,
      "grad_norm": 0.15054218471050262,
      "learning_rate": 5.048915876958935e-05,
      "loss": 0.5129,
      "mean_token_accuracy": 0.840088602155447,
      "num_tokens": 515278644.0,
      "step": 16145
    },
    {
      "entropy": 0.6644223099574447,
      "epoch": 1.4855176220953301,
      "grad_norm": 0.15677358210086823,
      "learning_rate": 5.048609194344773e-05,
      "loss": 0.6505,
      "mean_token_accuracy": 0.8037404678761959,
      "num_tokens": 515310630.0,
      "step": 16146
    },
    {
      "entropy": 0.5607881862670183,
      "epoch": 1.4856096284667712,
      "grad_norm": 0.15437407791614532,
      "learning_rate": 5.048302511730611e-05,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8308925442397594,
      "num_tokens": 515341927.0,
      "step": 16147
    },
    {
      "entropy": 0.6984633198007941,
      "epoch": 1.4857016348382126,
      "grad_norm": 0.16157874464988708,
      "learning_rate": 5.0479958291164475e-05,
      "loss": 0.6882,
      "mean_token_accuracy": 0.7890633381903172,
      "num_tokens": 515373536.0,
      "step": 16148
    },
    {
      "entropy": 0.6298819789662957,
      "epoch": 1.4857936412096537,
      "grad_norm": 0.1545710265636444,
      "learning_rate": 5.047689146502285e-05,
      "loss": 0.6224,
      "mean_token_accuracy": 0.8105611950159073,
      "num_tokens": 515405760.0,
      "step": 16149
    },
    {
      "entropy": 0.6912771333009005,
      "epoch": 1.485885647581095,
      "grad_norm": 0.16032227873802185,
      "learning_rate": 5.047382463888123e-05,
      "loss": 0.6845,
      "mean_token_accuracy": 0.7926223427057266,
      "num_tokens": 515437757.0,
      "step": 16150
    },
    {
      "entropy": 0.6461029937490821,
      "epoch": 1.4859776539525362,
      "grad_norm": 0.15858885645866394,
      "learning_rate": 5.04707578127396e-05,
      "loss": 0.6412,
      "mean_token_accuracy": 0.8052701838314533,
      "num_tokens": 515469137.0,
      "step": 16151
    },
    {
      "entropy": 0.6759928520768881,
      "epoch": 1.4860696603239774,
      "grad_norm": 0.15619628131389618,
      "learning_rate": 5.0467690986597974e-05,
      "loss": 0.6768,
      "mean_token_accuracy": 0.7951079793274403,
      "num_tokens": 515501088.0,
      "step": 16152
    },
    {
      "entropy": 0.504689269233495,
      "epoch": 1.4861616666954187,
      "grad_norm": 0.14898568391799927,
      "learning_rate": 5.046462416045634e-05,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8436406292021275,
      "num_tokens": 515533114.0,
      "step": 16153
    },
    {
      "entropy": 0.6980325281620026,
      "epoch": 1.4862536730668598,
      "grad_norm": 0.16157715022563934,
      "learning_rate": 5.046155733431472e-05,
      "loss": 0.6915,
      "mean_token_accuracy": 0.7959745563566685,
      "num_tokens": 515565238.0,
      "step": 16154
    },
    {
      "entropy": 0.6103603346273303,
      "epoch": 1.4863456794383012,
      "grad_norm": 0.1498032659292221,
      "learning_rate": 5.04584905081731e-05,
      "loss": 0.5973,
      "mean_token_accuracy": 0.8163856491446495,
      "num_tokens": 515596952.0,
      "step": 16155
    },
    {
      "entropy": 0.5625821547582746,
      "epoch": 1.4864376858097423,
      "grad_norm": 0.15348049998283386,
      "learning_rate": 5.0455423682031465e-05,
      "loss": 0.5486,
      "mean_token_accuracy": 0.8322275727987289,
      "num_tokens": 515629458.0,
      "step": 16156
    },
    {
      "entropy": 0.6314675882458687,
      "epoch": 1.4865296921811835,
      "grad_norm": 0.15552875399589539,
      "learning_rate": 5.045235685588984e-05,
      "loss": 0.6158,
      "mean_token_accuracy": 0.8091889843344688,
      "num_tokens": 515661199.0,
      "step": 16157
    },
    {
      "entropy": 0.568878729827702,
      "epoch": 1.4866216985526248,
      "grad_norm": 0.1478244513273239,
      "learning_rate": 5.044929002974822e-05,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8316179811954498,
      "num_tokens": 515693144.0,
      "step": 16158
    },
    {
      "entropy": 0.5028918907046318,
      "epoch": 1.486713704924066,
      "grad_norm": 0.1519918590784073,
      "learning_rate": 5.044622320360659e-05,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8438693396747112,
      "num_tokens": 515725494.0,
      "step": 16159
    },
    {
      "entropy": 0.6699254345148802,
      "epoch": 1.4868057112955073,
      "grad_norm": 0.1577645093202591,
      "learning_rate": 5.0443156377464964e-05,
      "loss": 0.6561,
      "mean_token_accuracy": 0.7990474179387093,
      "num_tokens": 515757202.0,
      "step": 16160
    },
    {
      "entropy": 0.594770897179842,
      "epoch": 1.4868977176669484,
      "grad_norm": 0.16052685678005219,
      "learning_rate": 5.044008955132333e-05,
      "loss": 0.5777,
      "mean_token_accuracy": 0.8207250274717808,
      "num_tokens": 515789636.0,
      "step": 16161
    },
    {
      "entropy": 0.6639395523816347,
      "epoch": 1.4869897240383896,
      "grad_norm": 0.16436892747879028,
      "learning_rate": 5.043702272518171e-05,
      "loss": 0.6514,
      "mean_token_accuracy": 0.8028261438012123,
      "num_tokens": 515821934.0,
      "step": 16162
    },
    {
      "entropy": 0.6504925210028887,
      "epoch": 1.487081730409831,
      "grad_norm": 0.15590491890907288,
      "learning_rate": 5.043395589904009e-05,
      "loss": 0.6384,
      "mean_token_accuracy": 0.8030591905117035,
      "num_tokens": 515853317.0,
      "step": 16163
    },
    {
      "entropy": 0.5420741476118565,
      "epoch": 1.487173736781272,
      "grad_norm": 0.1477051079273224,
      "learning_rate": 5.0430889072898456e-05,
      "loss": 0.5291,
      "mean_token_accuracy": 0.8321202322840691,
      "num_tokens": 515885611.0,
      "step": 16164
    },
    {
      "entropy": 0.6376443617045879,
      "epoch": 1.4872657431527134,
      "grad_norm": 0.16186349093914032,
      "learning_rate": 5.042782224675683e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.8090373538434505,
      "num_tokens": 515917165.0,
      "step": 16165
    },
    {
      "entropy": 0.6239594584330916,
      "epoch": 1.4873577495241546,
      "grad_norm": 0.1616056114435196,
      "learning_rate": 5.042475542061521e-05,
      "loss": 0.6126,
      "mean_token_accuracy": 0.814458429813385,
      "num_tokens": 515949313.0,
      "step": 16166
    },
    {
      "entropy": 0.6772261392325163,
      "epoch": 1.4874497558955957,
      "grad_norm": 0.16193445026874542,
      "learning_rate": 5.042168859447358e-05,
      "loss": 0.6746,
      "mean_token_accuracy": 0.7909603416919708,
      "num_tokens": 515981250.0,
      "step": 16167
    },
    {
      "entropy": 0.5914163505658507,
      "epoch": 1.487541762267037,
      "grad_norm": 0.15303660929203033,
      "learning_rate": 5.0418621768331954e-05,
      "loss": 0.5844,
      "mean_token_accuracy": 0.8256238102912903,
      "num_tokens": 516013644.0,
      "step": 16168
    },
    {
      "entropy": 0.5289221371058375,
      "epoch": 1.4876337686384782,
      "grad_norm": 0.1485247164964676,
      "learning_rate": 5.0415554942190336e-05,
      "loss": 0.522,
      "mean_token_accuracy": 0.8381413817405701,
      "num_tokens": 516045866.0,
      "step": 16169
    },
    {
      "entropy": 0.5059668552130461,
      "epoch": 1.4877257750099195,
      "grad_norm": 0.14245207607746124,
      "learning_rate": 5.0412488116048704e-05,
      "loss": 0.4988,
      "mean_token_accuracy": 0.8413607887923717,
      "num_tokens": 516077828.0,
      "step": 16170
    },
    {
      "entropy": 0.6357972025871277,
      "epoch": 1.4878177813813607,
      "grad_norm": 0.15392959117889404,
      "learning_rate": 5.040942128990708e-05,
      "loss": 0.6153,
      "mean_token_accuracy": 0.8100596182048321,
      "num_tokens": 516109757.0,
      "step": 16171
    },
    {
      "entropy": 0.6183966211974621,
      "epoch": 1.4879097877528018,
      "grad_norm": 0.1553707718849182,
      "learning_rate": 5.0406354463765446e-05,
      "loss": 0.6248,
      "mean_token_accuracy": 0.8101348504424095,
      "num_tokens": 516141125.0,
      "step": 16172
    },
    {
      "entropy": 0.5847351681441069,
      "epoch": 1.4880017941242432,
      "grad_norm": 0.15185175836086273,
      "learning_rate": 5.040328763762383e-05,
      "loss": 0.5875,
      "mean_token_accuracy": 0.816451296210289,
      "num_tokens": 516173677.0,
      "step": 16173
    },
    {
      "entropy": 0.681676147505641,
      "epoch": 1.4880938004956843,
      "grad_norm": 0.16568303108215332,
      "learning_rate": 5.04002208114822e-05,
      "loss": 0.6726,
      "mean_token_accuracy": 0.7934736907482147,
      "num_tokens": 516206156.0,
      "step": 16174
    },
    {
      "entropy": 0.6181088797748089,
      "epoch": 1.4881858068671256,
      "grad_norm": 0.15614765882492065,
      "learning_rate": 5.039715398534057e-05,
      "loss": 0.6118,
      "mean_token_accuracy": 0.8110475279390812,
      "num_tokens": 516237770.0,
      "step": 16175
    },
    {
      "entropy": 0.5438322387635708,
      "epoch": 1.4882778132385668,
      "grad_norm": 0.1508798450231552,
      "learning_rate": 5.0394087159198945e-05,
      "loss": 0.5416,
      "mean_token_accuracy": 0.8341372944414616,
      "num_tokens": 516270063.0,
      "step": 16176
    },
    {
      "entropy": 0.643830394372344,
      "epoch": 1.488369819610008,
      "grad_norm": 0.1582023948431015,
      "learning_rate": 5.0391020333057326e-05,
      "loss": 0.6389,
      "mean_token_accuracy": 0.8080593571066856,
      "num_tokens": 516301747.0,
      "step": 16177
    },
    {
      "entropy": 0.5574062736704946,
      "epoch": 1.4884618259814493,
      "grad_norm": 0.15211018919944763,
      "learning_rate": 5.0387953506915694e-05,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8344499990344048,
      "num_tokens": 516333660.0,
      "step": 16178
    },
    {
      "entropy": 0.5904898722656071,
      "epoch": 1.4885538323528904,
      "grad_norm": 0.15323907136917114,
      "learning_rate": 5.038488668077407e-05,
      "loss": 0.5839,
      "mean_token_accuracy": 0.8223464265465736,
      "num_tokens": 516365596.0,
      "step": 16179
    },
    {
      "entropy": 0.5993799120187759,
      "epoch": 1.4886458387243318,
      "grad_norm": 0.15258412063121796,
      "learning_rate": 5.038181985463244e-05,
      "loss": 0.5839,
      "mean_token_accuracy": 0.8164182901382446,
      "num_tokens": 516397198.0,
      "step": 16180
    },
    {
      "entropy": 0.5192406345158815,
      "epoch": 1.488737845095773,
      "grad_norm": 0.15115542709827423,
      "learning_rate": 5.0378753028490825e-05,
      "loss": 0.5082,
      "mean_token_accuracy": 0.8458072543144226,
      "num_tokens": 516429709.0,
      "step": 16181
    },
    {
      "entropy": 0.716437540948391,
      "epoch": 1.488829851467214,
      "grad_norm": 0.16603529453277588,
      "learning_rate": 5.037568620234919e-05,
      "loss": 0.7063,
      "mean_token_accuracy": 0.7881963066756725,
      "num_tokens": 516461582.0,
      "step": 16182
    },
    {
      "entropy": 0.5132129611447453,
      "epoch": 1.4889218578386554,
      "grad_norm": 0.15071532130241394,
      "learning_rate": 5.037261937620756e-05,
      "loss": 0.5081,
      "mean_token_accuracy": 0.8436971828341484,
      "num_tokens": 516494331.0,
      "step": 16183
    },
    {
      "entropy": 0.6382366195321083,
      "epoch": 1.4890138642100965,
      "grad_norm": 0.15908168256282806,
      "learning_rate": 5.0369552550065935e-05,
      "loss": 0.6264,
      "mean_token_accuracy": 0.8076206408441067,
      "num_tokens": 516526454.0,
      "step": 16184
    },
    {
      "entropy": 0.511110863648355,
      "epoch": 1.4891058705815379,
      "grad_norm": 0.14743740856647491,
      "learning_rate": 5.036648572392432e-05,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8418456092476845,
      "num_tokens": 516558785.0,
      "step": 16185
    },
    {
      "entropy": 0.5114666824229062,
      "epoch": 1.489197876952979,
      "grad_norm": 0.14277523756027222,
      "learning_rate": 5.036341889778269e-05,
      "loss": 0.4905,
      "mean_token_accuracy": 0.847908366471529,
      "num_tokens": 516591303.0,
      "step": 16186
    },
    {
      "entropy": 0.660923894494772,
      "epoch": 1.4892898833244201,
      "grad_norm": 0.1560402363538742,
      "learning_rate": 5.036035207164106e-05,
      "loss": 0.644,
      "mean_token_accuracy": 0.8040119595825672,
      "num_tokens": 516624071.0,
      "step": 16187
    },
    {
      "entropy": 0.583776568993926,
      "epoch": 1.4893818896958615,
      "grad_norm": 0.14883436262607574,
      "learning_rate": 5.035728524549943e-05,
      "loss": 0.5763,
      "mean_token_accuracy": 0.8224257603287697,
      "num_tokens": 516656624.0,
      "step": 16188
    },
    {
      "entropy": 0.6286794152110815,
      "epoch": 1.4894738960673026,
      "grad_norm": 0.16101589798927307,
      "learning_rate": 5.0354218419357815e-05,
      "loss": 0.6288,
      "mean_token_accuracy": 0.8088016398251057,
      "num_tokens": 516688660.0,
      "step": 16189
    },
    {
      "entropy": 0.6226568929851055,
      "epoch": 1.489565902438744,
      "grad_norm": 0.15869447588920593,
      "learning_rate": 5.035115159321618e-05,
      "loss": 0.6277,
      "mean_token_accuracy": 0.8103545121848583,
      "num_tokens": 516720843.0,
      "step": 16190
    },
    {
      "entropy": 0.668093878775835,
      "epoch": 1.4896579088101851,
      "grad_norm": 0.1597071886062622,
      "learning_rate": 5.034808476707456e-05,
      "loss": 0.6483,
      "mean_token_accuracy": 0.8020178489387035,
      "num_tokens": 516752278.0,
      "step": 16191
    },
    {
      "entropy": 0.6617699451744556,
      "epoch": 1.4897499151816262,
      "grad_norm": 0.17632855474948883,
      "learning_rate": 5.0345017940932926e-05,
      "loss": 0.654,
      "mean_token_accuracy": 0.7999658323824406,
      "num_tokens": 516782768.0,
      "step": 16192
    },
    {
      "entropy": 0.7430152744054794,
      "epoch": 1.4898419215530676,
      "grad_norm": 0.16582229733467102,
      "learning_rate": 5.034195111479131e-05,
      "loss": 0.7434,
      "mean_token_accuracy": 0.7779799364507198,
      "num_tokens": 516815069.0,
      "step": 16193
    },
    {
      "entropy": 0.6442425604909658,
      "epoch": 1.4899339279245087,
      "grad_norm": 0.1622278243303299,
      "learning_rate": 5.033888428864968e-05,
      "loss": 0.6312,
      "mean_token_accuracy": 0.8079820945858955,
      "num_tokens": 516846199.0,
      "step": 16194
    },
    {
      "entropy": 0.5834477385506034,
      "epoch": 1.49002593429595,
      "grad_norm": 0.16128501296043396,
      "learning_rate": 5.033581746250805e-05,
      "loss": 0.5809,
      "mean_token_accuracy": 0.8216868527233601,
      "num_tokens": 516877538.0,
      "step": 16195
    },
    {
      "entropy": 0.7191119976341724,
      "epoch": 1.4901179406673912,
      "grad_norm": 0.17409244179725647,
      "learning_rate": 5.0332750636366424e-05,
      "loss": 0.7147,
      "mean_token_accuracy": 0.784320455044508,
      "num_tokens": 516908854.0,
      "step": 16196
    },
    {
      "entropy": 0.6749866846948862,
      "epoch": 1.4902099470388324,
      "grad_norm": 0.15948505699634552,
      "learning_rate": 5.0329683810224806e-05,
      "loss": 0.6592,
      "mean_token_accuracy": 0.7978093847632408,
      "num_tokens": 516940753.0,
      "step": 16197
    },
    {
      "entropy": 0.48049821984022856,
      "epoch": 1.4903019534102737,
      "grad_norm": 0.14689011871814728,
      "learning_rate": 5.0326616984083174e-05,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8557126969099045,
      "num_tokens": 516972807.0,
      "step": 16198
    },
    {
      "entropy": 0.7072955891489983,
      "epoch": 1.4903939597817148,
      "grad_norm": 0.16508467495441437,
      "learning_rate": 5.032355015794155e-05,
      "loss": 0.6998,
      "mean_token_accuracy": 0.7876431234180927,
      "num_tokens": 517004638.0,
      "step": 16199
    },
    {
      "entropy": 0.6687352247536182,
      "epoch": 1.4904859661531562,
      "grad_norm": 0.1596597582101822,
      "learning_rate": 5.032048333179993e-05,
      "loss": 0.6516,
      "mean_token_accuracy": 0.798374842852354,
      "num_tokens": 517036197.0,
      "step": 16200
    },
    {
      "entropy": 0.6305385045707226,
      "epoch": 1.4905779725245973,
      "grad_norm": 0.15415501594543457,
      "learning_rate": 5.03174165056583e-05,
      "loss": 0.6189,
      "mean_token_accuracy": 0.8112533874809742,
      "num_tokens": 517068463.0,
      "step": 16201
    },
    {
      "entropy": 0.5591384554281831,
      "epoch": 1.4906699788960385,
      "grad_norm": 0.15475155413150787,
      "learning_rate": 5.031434967951667e-05,
      "loss": 0.546,
      "mean_token_accuracy": 0.8294692933559418,
      "num_tokens": 517100330.0,
      "step": 16202
    },
    {
      "entropy": 0.5256894817575812,
      "epoch": 1.4907619852674798,
      "grad_norm": 0.14159433543682098,
      "learning_rate": 5.031128285337504e-05,
      "loss": 0.5009,
      "mean_token_accuracy": 0.841724306344986,
      "num_tokens": 517132170.0,
      "step": 16203
    },
    {
      "entropy": 0.6267874510958791,
      "epoch": 1.490853991638921,
      "grad_norm": 0.15058933198451996,
      "learning_rate": 5.030821602723342e-05,
      "loss": 0.6184,
      "mean_token_accuracy": 0.8140408210456371,
      "num_tokens": 517164755.0,
      "step": 16204
    },
    {
      "entropy": 0.579365624114871,
      "epoch": 1.4909459980103623,
      "grad_norm": 0.16063319146633148,
      "learning_rate": 5.0305149201091796e-05,
      "loss": 0.5656,
      "mean_token_accuracy": 0.8236632533371449,
      "num_tokens": 517196384.0,
      "step": 16205
    },
    {
      "entropy": 0.6985575519502163,
      "epoch": 1.4910380043818035,
      "grad_norm": 0.15527985990047455,
      "learning_rate": 5.0302082374950164e-05,
      "loss": 0.6854,
      "mean_token_accuracy": 0.7928141392767429,
      "num_tokens": 517228623.0,
      "step": 16206
    },
    {
      "entropy": 0.6757974363863468,
      "epoch": 1.4911300107532446,
      "grad_norm": 0.16801927983760834,
      "learning_rate": 5.029901554880854e-05,
      "loss": 0.665,
      "mean_token_accuracy": 0.7975552342832088,
      "num_tokens": 517259649.0,
      "step": 16207
    },
    {
      "entropy": 0.6597025282680988,
      "epoch": 1.491222017124686,
      "grad_norm": 0.15670695900917053,
      "learning_rate": 5.029594872266692e-05,
      "loss": 0.6489,
      "mean_token_accuracy": 0.8032318651676178,
      "num_tokens": 517291889.0,
      "step": 16208
    },
    {
      "entropy": 0.5262983068823814,
      "epoch": 1.491314023496127,
      "grad_norm": 0.15092158317565918,
      "learning_rate": 5.029288189652529e-05,
      "loss": 0.5058,
      "mean_token_accuracy": 0.8446837514638901,
      "num_tokens": 517323858.0,
      "step": 16209
    },
    {
      "entropy": 0.5551241170614958,
      "epoch": 1.4914060298675684,
      "grad_norm": 0.14824295043945312,
      "learning_rate": 5.028981507038366e-05,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8328935690224171,
      "num_tokens": 517355360.0,
      "step": 16210
    },
    {
      "entropy": 0.6143254609778523,
      "epoch": 1.4914980362390096,
      "grad_norm": 0.1584097295999527,
      "learning_rate": 5.028674824424203e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8118981644511223,
      "num_tokens": 517387495.0,
      "step": 16211
    },
    {
      "entropy": 0.5740426676347852,
      "epoch": 1.4915900426104507,
      "grad_norm": 0.15490856766700745,
      "learning_rate": 5.028368141810041e-05,
      "loss": 0.58,
      "mean_token_accuracy": 0.822069339454174,
      "num_tokens": 517419409.0,
      "step": 16212
    },
    {
      "entropy": 0.6302547957748175,
      "epoch": 1.491682048981892,
      "grad_norm": 0.16100923717021942,
      "learning_rate": 5.0280614591958787e-05,
      "loss": 0.6328,
      "mean_token_accuracy": 0.8078658916056156,
      "num_tokens": 517451654.0,
      "step": 16213
    },
    {
      "entropy": 0.6400513807311654,
      "epoch": 1.4917740553533332,
      "grad_norm": 0.17049647867679596,
      "learning_rate": 5.0277547765817154e-05,
      "loss": 0.6339,
      "mean_token_accuracy": 0.8085396513342857,
      "num_tokens": 517483578.0,
      "step": 16214
    },
    {
      "entropy": 0.7730420865118504,
      "epoch": 1.4918660617247745,
      "grad_norm": 0.16443131864070892,
      "learning_rate": 5.027448093967553e-05,
      "loss": 0.7678,
      "mean_token_accuracy": 0.7672378085553646,
      "num_tokens": 517514637.0,
      "step": 16215
    },
    {
      "entropy": 0.7165831420570612,
      "epoch": 1.4919580680962157,
      "grad_norm": 0.16320393979549408,
      "learning_rate": 5.027141411353391e-05,
      "loss": 0.7169,
      "mean_token_accuracy": 0.7826709933578968,
      "num_tokens": 517545822.0,
      "step": 16216
    },
    {
      "entropy": 0.5395801570266485,
      "epoch": 1.4920500744676568,
      "grad_norm": 0.1501472294330597,
      "learning_rate": 5.026834728739228e-05,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8343892022967339,
      "num_tokens": 517577169.0,
      "step": 16217
    },
    {
      "entropy": 0.6960512660443783,
      "epoch": 1.4921420808390982,
      "grad_norm": 0.16544978320598602,
      "learning_rate": 5.026528046125065e-05,
      "loss": 0.6915,
      "mean_token_accuracy": 0.7911573015153408,
      "num_tokens": 517608972.0,
      "step": 16218
    },
    {
      "entropy": 0.5514113034587353,
      "epoch": 1.4922340872105393,
      "grad_norm": 0.1607915163040161,
      "learning_rate": 5.026221363510902e-05,
      "loss": 0.5342,
      "mean_token_accuracy": 0.835975207388401,
      "num_tokens": 517641217.0,
      "step": 16219
    },
    {
      "entropy": 0.6700161760672927,
      "epoch": 1.4923260935819807,
      "grad_norm": 0.16003455221652985,
      "learning_rate": 5.02591468089674e-05,
      "loss": 0.6557,
      "mean_token_accuracy": 0.8014818392693996,
      "num_tokens": 517672617.0,
      "step": 16220
    },
    {
      "entropy": 0.6305635059252381,
      "epoch": 1.4924180999534218,
      "grad_norm": 0.15501533448696136,
      "learning_rate": 5.025607998282578e-05,
      "loss": 0.6238,
      "mean_token_accuracy": 0.8103632815182209,
      "num_tokens": 517704910.0,
      "step": 16221
    },
    {
      "entropy": 0.6667200250085443,
      "epoch": 1.492510106324863,
      "grad_norm": 0.15654930472373962,
      "learning_rate": 5.0253013156684145e-05,
      "loss": 0.6583,
      "mean_token_accuracy": 0.8035890199244022,
      "num_tokens": 517736923.0,
      "step": 16222
    },
    {
      "entropy": 0.523686470463872,
      "epoch": 1.4926021126963043,
      "grad_norm": 0.1476583629846573,
      "learning_rate": 5.024994633054252e-05,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8396972119808197,
      "num_tokens": 517768987.0,
      "step": 16223
    },
    {
      "entropy": 0.5617240089923143,
      "epoch": 1.4926941190677454,
      "grad_norm": 0.15598176419734955,
      "learning_rate": 5.02468795044009e-05,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8285439349710941,
      "num_tokens": 517801066.0,
      "step": 16224
    },
    {
      "entropy": 0.6435711402446032,
      "epoch": 1.4927861254391868,
      "grad_norm": 0.15794095396995544,
      "learning_rate": 5.0243812678259276e-05,
      "loss": 0.6249,
      "mean_token_accuracy": 0.8087073788046837,
      "num_tokens": 517832711.0,
      "step": 16225
    },
    {
      "entropy": 0.6148434830829501,
      "epoch": 1.492878131810628,
      "grad_norm": 0.15649619698524475,
      "learning_rate": 5.024074585211764e-05,
      "loss": 0.5869,
      "mean_token_accuracy": 0.8193994648754597,
      "num_tokens": 517863746.0,
      "step": 16226
    },
    {
      "entropy": 0.6909322030842304,
      "epoch": 1.492970138182069,
      "grad_norm": 0.16089767217636108,
      "learning_rate": 5.023767902597601e-05,
      "loss": 0.6903,
      "mean_token_accuracy": 0.7903914414346218,
      "num_tokens": 517895868.0,
      "step": 16227
    },
    {
      "entropy": 0.7207761565223336,
      "epoch": 1.4930621445535104,
      "grad_norm": 0.16763731837272644,
      "learning_rate": 5.02346121998344e-05,
      "loss": 0.7162,
      "mean_token_accuracy": 0.7842054590582848,
      "num_tokens": 517927799.0,
      "step": 16228
    },
    {
      "entropy": 0.6912538465112448,
      "epoch": 1.4931541509249515,
      "grad_norm": 0.15581730008125305,
      "learning_rate": 5.023154537369277e-05,
      "loss": 0.6828,
      "mean_token_accuracy": 0.7918225862085819,
      "num_tokens": 517960220.0,
      "step": 16229
    },
    {
      "entropy": 0.619774853810668,
      "epoch": 1.4932461572963929,
      "grad_norm": 0.15562470257282257,
      "learning_rate": 5.022847854755114e-05,
      "loss": 0.6159,
      "mean_token_accuracy": 0.8117432594299316,
      "num_tokens": 517992145.0,
      "step": 16230
    },
    {
      "entropy": 0.5850679716095328,
      "epoch": 1.493338163667834,
      "grad_norm": 0.15341129899024963,
      "learning_rate": 5.0225411721409523e-05,
      "loss": 0.5697,
      "mean_token_accuracy": 0.8284372054040432,
      "num_tokens": 518023208.0,
      "step": 16231
    },
    {
      "entropy": 0.5949359722435474,
      "epoch": 1.4934301700392751,
      "grad_norm": 0.15149223804473877,
      "learning_rate": 5.022234489526789e-05,
      "loss": 0.5922,
      "mean_token_accuracy": 0.816733930259943,
      "num_tokens": 518055054.0,
      "step": 16232
    },
    {
      "entropy": 0.5239692516624928,
      "epoch": 1.4935221764107165,
      "grad_norm": 0.1473686546087265,
      "learning_rate": 5.0219278069126266e-05,
      "loss": 0.5179,
      "mean_token_accuracy": 0.8386280499398708,
      "num_tokens": 518086703.0,
      "step": 16233
    },
    {
      "entropy": 0.7086114110425115,
      "epoch": 1.4936141827821576,
      "grad_norm": 0.16411474347114563,
      "learning_rate": 5.0216211242984634e-05,
      "loss": 0.6963,
      "mean_token_accuracy": 0.7902999557554722,
      "num_tokens": 518118855.0,
      "step": 16234
    },
    {
      "entropy": 0.5286179650574923,
      "epoch": 1.493706189153599,
      "grad_norm": 0.15164542198181152,
      "learning_rate": 5.0213144416843015e-05,
      "loss": 0.5262,
      "mean_token_accuracy": 0.8418199233710766,
      "num_tokens": 518151287.0,
      "step": 16235
    },
    {
      "entropy": 0.6586718643084168,
      "epoch": 1.4937981955250401,
      "grad_norm": 0.1588474065065384,
      "learning_rate": 5.021007759070139e-05,
      "loss": 0.6506,
      "mean_token_accuracy": 0.7987498305737972,
      "num_tokens": 518183749.0,
      "step": 16236
    },
    {
      "entropy": 0.6399428145959973,
      "epoch": 1.4938902018964813,
      "grad_norm": 0.16235369443893433,
      "learning_rate": 5.020701076455976e-05,
      "loss": 0.6396,
      "mean_token_accuracy": 0.8048094697296619,
      "num_tokens": 518215728.0,
      "step": 16237
    },
    {
      "entropy": 0.7146992776542902,
      "epoch": 1.4939822082679226,
      "grad_norm": 0.16230709850788116,
      "learning_rate": 5.020394393841813e-05,
      "loss": 0.7039,
      "mean_token_accuracy": 0.7873551733791828,
      "num_tokens": 518247055.0,
      "step": 16238
    },
    {
      "entropy": 0.5411536935716867,
      "epoch": 1.4940742146393637,
      "grad_norm": 0.15520139038562775,
      "learning_rate": 5.0200877112276514e-05,
      "loss": 0.519,
      "mean_token_accuracy": 0.8393915519118309,
      "num_tokens": 518279087.0,
      "step": 16239
    },
    {
      "entropy": 0.6777658704668283,
      "epoch": 1.494166221010805,
      "grad_norm": 0.15693913400173187,
      "learning_rate": 5.019781028613488e-05,
      "loss": 0.6622,
      "mean_token_accuracy": 0.7987099848687649,
      "num_tokens": 518310939.0,
      "step": 16240
    },
    {
      "entropy": 0.6366395317018032,
      "epoch": 1.4942582273822462,
      "grad_norm": 0.1679409146308899,
      "learning_rate": 5.0194743459993256e-05,
      "loss": 0.6257,
      "mean_token_accuracy": 0.8059302642941475,
      "num_tokens": 518342628.0,
      "step": 16241
    },
    {
      "entropy": 0.5663635875098407,
      "epoch": 1.4943502337536874,
      "grad_norm": 0.1545296013355255,
      "learning_rate": 5.0191676633851624e-05,
      "loss": 0.5522,
      "mean_token_accuracy": 0.8295724764466286,
      "num_tokens": 518374646.0,
      "step": 16242
    },
    {
      "entropy": 0.6709370221942663,
      "epoch": 1.4944422401251287,
      "grad_norm": 0.16449667513370514,
      "learning_rate": 5.0188609807710006e-05,
      "loss": 0.6517,
      "mean_token_accuracy": 0.8048068434000015,
      "num_tokens": 518406141.0,
      "step": 16243
    },
    {
      "entropy": 0.6975988987833261,
      "epoch": 1.4945342464965699,
      "grad_norm": 0.1572275608778,
      "learning_rate": 5.018554298156838e-05,
      "loss": 0.6854,
      "mean_token_accuracy": 0.7909569926559925,
      "num_tokens": 518438478.0,
      "step": 16244
    },
    {
      "entropy": 0.6308197705075145,
      "epoch": 1.4946262528680112,
      "grad_norm": 0.16099265217781067,
      "learning_rate": 5.018247615542675e-05,
      "loss": 0.6258,
      "mean_token_accuracy": 0.8036006689071655,
      "num_tokens": 518470342.0,
      "step": 16245
    },
    {
      "entropy": 0.6574935615062714,
      "epoch": 1.4947182592394523,
      "grad_norm": 0.15585312247276306,
      "learning_rate": 5.017940932928512e-05,
      "loss": 0.6413,
      "mean_token_accuracy": 0.8023420870304108,
      "num_tokens": 518501757.0,
      "step": 16246
    },
    {
      "entropy": 0.639295494183898,
      "epoch": 1.4948102656108935,
      "grad_norm": 0.15535880625247955,
      "learning_rate": 5.0176342503143504e-05,
      "loss": 0.622,
      "mean_token_accuracy": 0.8071105442941189,
      "num_tokens": 518534049.0,
      "step": 16247
    },
    {
      "entropy": 0.46663554082624614,
      "epoch": 1.4949022719823348,
      "grad_norm": 0.1412316858768463,
      "learning_rate": 5.017327567700187e-05,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8564235530793667,
      "num_tokens": 518566453.0,
      "step": 16248
    },
    {
      "entropy": 0.599495098926127,
      "epoch": 1.494994278353776,
      "grad_norm": 0.1685599684715271,
      "learning_rate": 5.017020885086025e-05,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8176647499203682,
      "num_tokens": 518598061.0,
      "step": 16249
    },
    {
      "entropy": 0.616158977150917,
      "epoch": 1.4950862847252173,
      "grad_norm": 0.1615329384803772,
      "learning_rate": 5.0167142024718615e-05,
      "loss": 0.6028,
      "mean_token_accuracy": 0.8132088854908943,
      "num_tokens": 518629091.0,
      "step": 16250
    },
    {
      "entropy": 0.4919720459729433,
      "epoch": 1.4951782910966585,
      "grad_norm": 0.14817050099372864,
      "learning_rate": 5.0164075198576996e-05,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8527227118611336,
      "num_tokens": 518661403.0,
      "step": 16251
    },
    {
      "entropy": 0.6043423064984381,
      "epoch": 1.4952702974680996,
      "grad_norm": 0.16005395352840424,
      "learning_rate": 5.016100837243537e-05,
      "loss": 0.6003,
      "mean_token_accuracy": 0.8127457089722157,
      "num_tokens": 518692172.0,
      "step": 16252
    },
    {
      "entropy": 0.6280464921146631,
      "epoch": 1.495362303839541,
      "grad_norm": 0.17052911221981049,
      "learning_rate": 5.015794154629374e-05,
      "loss": 0.6208,
      "mean_token_accuracy": 0.8090980686247349,
      "num_tokens": 518723036.0,
      "step": 16253
    },
    {
      "entropy": 0.6939353700727224,
      "epoch": 1.495454310210982,
      "grad_norm": 0.15936297178268433,
      "learning_rate": 5.015487472015211e-05,
      "loss": 0.6847,
      "mean_token_accuracy": 0.7899843640625477,
      "num_tokens": 518755233.0,
      "step": 16254
    },
    {
      "entropy": 0.6905739326030016,
      "epoch": 1.4955463165824234,
      "grad_norm": 0.1595885157585144,
      "learning_rate": 5.0151807894010495e-05,
      "loss": 0.6835,
      "mean_token_accuracy": 0.7974032126367092,
      "num_tokens": 518786851.0,
      "step": 16255
    },
    {
      "entropy": 0.625312490388751,
      "epoch": 1.4956383229538646,
      "grad_norm": 0.15671837329864502,
      "learning_rate": 5.014874106786886e-05,
      "loss": 0.619,
      "mean_token_accuracy": 0.809685330837965,
      "num_tokens": 518818640.0,
      "step": 16256
    },
    {
      "entropy": 0.5898359157145023,
      "epoch": 1.4957303293253057,
      "grad_norm": 0.14982067048549652,
      "learning_rate": 5.014567424172724e-05,
      "loss": 0.5713,
      "mean_token_accuracy": 0.8243403024971485,
      "num_tokens": 518850772.0,
      "step": 16257
    },
    {
      "entropy": 0.5571118844673038,
      "epoch": 1.495822335696747,
      "grad_norm": 0.15748028457164764,
      "learning_rate": 5.0142607415585605e-05,
      "loss": 0.5505,
      "mean_token_accuracy": 0.8300941213965416,
      "num_tokens": 518882458.0,
      "step": 16258
    },
    {
      "entropy": 0.6064098887145519,
      "epoch": 1.4959143420681882,
      "grad_norm": 0.1506923884153366,
      "learning_rate": 5.0139540589443986e-05,
      "loss": 0.6032,
      "mean_token_accuracy": 0.816335704177618,
      "num_tokens": 518914809.0,
      "step": 16259
    },
    {
      "entropy": 0.547521352302283,
      "epoch": 1.4960063484396295,
      "grad_norm": 0.15263281762599945,
      "learning_rate": 5.013647376330236e-05,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8371458575129509,
      "num_tokens": 518946450.0,
      "step": 16260
    },
    {
      "entropy": 0.6611374458298087,
      "epoch": 1.4960983548110707,
      "grad_norm": 0.159267395734787,
      "learning_rate": 5.013340693716073e-05,
      "loss": 0.652,
      "mean_token_accuracy": 0.8050085715949535,
      "num_tokens": 518978680.0,
      "step": 16261
    },
    {
      "entropy": 0.6134055140428245,
      "epoch": 1.4961903611825118,
      "grad_norm": 0.15438470244407654,
      "learning_rate": 5.013034011101912e-05,
      "loss": 0.6059,
      "mean_token_accuracy": 0.8178128302097321,
      "num_tokens": 519010455.0,
      "step": 16262
    },
    {
      "entropy": 0.5612787203863263,
      "epoch": 1.4962823675539532,
      "grad_norm": 0.1520514041185379,
      "learning_rate": 5.0127273284877485e-05,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8288862220942974,
      "num_tokens": 519042787.0,
      "step": 16263
    },
    {
      "entropy": 0.5864141946658492,
      "epoch": 1.4963743739253943,
      "grad_norm": 0.1524561643600464,
      "learning_rate": 5.012420645873585e-05,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8255137093365192,
      "num_tokens": 519074974.0,
      "step": 16264
    },
    {
      "entropy": 0.6102110482752323,
      "epoch": 1.4964663802968357,
      "grad_norm": 0.1577424854040146,
      "learning_rate": 5.012113963259423e-05,
      "loss": 0.6146,
      "mean_token_accuracy": 0.812428917735815,
      "num_tokens": 519107605.0,
      "step": 16265
    },
    {
      "entropy": 0.5195951480418444,
      "epoch": 1.4965583866682768,
      "grad_norm": 0.14931051433086395,
      "learning_rate": 5.011807280645261e-05,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8431267626583576,
      "num_tokens": 519139126.0,
      "step": 16266
    },
    {
      "entropy": 0.6170388199388981,
      "epoch": 1.496650393039718,
      "grad_norm": 0.15290363132953644,
      "learning_rate": 5.0115005980310984e-05,
      "loss": 0.6051,
      "mean_token_accuracy": 0.8104843385517597,
      "num_tokens": 519171060.0,
      "step": 16267
    },
    {
      "entropy": 0.6563302830327302,
      "epoch": 1.4967423994111593,
      "grad_norm": 0.1564388871192932,
      "learning_rate": 5.011193915416935e-05,
      "loss": 0.6546,
      "mean_token_accuracy": 0.8045634962618351,
      "num_tokens": 519203300.0,
      "step": 16268
    },
    {
      "entropy": 0.6190138328820467,
      "epoch": 1.4968344057826004,
      "grad_norm": 0.15102502703666687,
      "learning_rate": 5.0108872328027726e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.8125868141651154,
      "num_tokens": 519235457.0,
      "step": 16269
    },
    {
      "entropy": 0.7008902989327908,
      "epoch": 1.4969264121540418,
      "grad_norm": 0.1620507538318634,
      "learning_rate": 5.010580550188611e-05,
      "loss": 0.7044,
      "mean_token_accuracy": 0.7907343283295631,
      "num_tokens": 519266873.0,
      "step": 16270
    },
    {
      "entropy": 0.7567003332078457,
      "epoch": 1.497018418525483,
      "grad_norm": 0.16551336646080017,
      "learning_rate": 5.0102738675744475e-05,
      "loss": 0.7543,
      "mean_token_accuracy": 0.7736301831901073,
      "num_tokens": 519298839.0,
      "step": 16271
    },
    {
      "entropy": 0.6382582876831293,
      "epoch": 1.497110424896924,
      "grad_norm": 0.1514878273010254,
      "learning_rate": 5.009967184960285e-05,
      "loss": 0.6132,
      "mean_token_accuracy": 0.8107943423092365,
      "num_tokens": 519331518.0,
      "step": 16272
    },
    {
      "entropy": 0.6850503664463758,
      "epoch": 1.4972024312683654,
      "grad_norm": 0.15273965895175934,
      "learning_rate": 5.009660502346122e-05,
      "loss": 0.6644,
      "mean_token_accuracy": 0.7956764698028564,
      "num_tokens": 519363821.0,
      "step": 16273
    },
    {
      "entropy": 0.763950552791357,
      "epoch": 1.4972944376398065,
      "grad_norm": 0.16970515251159668,
      "learning_rate": 5.00935381973196e-05,
      "loss": 0.7539,
      "mean_token_accuracy": 0.7691413089632988,
      "num_tokens": 519395189.0,
      "step": 16274
    },
    {
      "entropy": 0.6198550779372454,
      "epoch": 1.4973864440112479,
      "grad_norm": 0.15250352025032043,
      "learning_rate": 5.0090471371177974e-05,
      "loss": 0.5967,
      "mean_token_accuracy": 0.8169227838516235,
      "num_tokens": 519427129.0,
      "step": 16275
    },
    {
      "entropy": 0.6428937967866659,
      "epoch": 1.497478450382689,
      "grad_norm": 0.16409069299697876,
      "learning_rate": 5.008740454503634e-05,
      "loss": 0.6325,
      "mean_token_accuracy": 0.807512179017067,
      "num_tokens": 519459082.0,
      "step": 16276
    },
    {
      "entropy": 0.5642026886343956,
      "epoch": 1.4975704567541301,
      "grad_norm": 0.15475204586982727,
      "learning_rate": 5.0084337718894717e-05,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8373178988695145,
      "num_tokens": 519490525.0,
      "step": 16277
    },
    {
      "entropy": 0.5719184060581028,
      "epoch": 1.4976624631255715,
      "grad_norm": 0.16776274144649506,
      "learning_rate": 5.00812708927531e-05,
      "loss": 0.5752,
      "mean_token_accuracy": 0.8233360983431339,
      "num_tokens": 519523009.0,
      "step": 16278
    },
    {
      "entropy": 0.676095805130899,
      "epoch": 1.4977544694970126,
      "grad_norm": 0.15834613144397736,
      "learning_rate": 5.0078204066611466e-05,
      "loss": 0.6678,
      "mean_token_accuracy": 0.7940887585282326,
      "num_tokens": 519555097.0,
      "step": 16279
    },
    {
      "entropy": 0.5639263298362494,
      "epoch": 1.497846475868454,
      "grad_norm": 0.15447494387626648,
      "learning_rate": 5.007513724046984e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8275305554270744,
      "num_tokens": 519587440.0,
      "step": 16280
    },
    {
      "entropy": 0.6441043578088284,
      "epoch": 1.4979384822398951,
      "grad_norm": 0.16203239560127258,
      "learning_rate": 5.007207041432821e-05,
      "loss": 0.6279,
      "mean_token_accuracy": 0.8061078116297722,
      "num_tokens": 519619383.0,
      "step": 16281
    },
    {
      "entropy": 0.6773220864124596,
      "epoch": 1.4980304886113363,
      "grad_norm": 0.15434102714061737,
      "learning_rate": 5.006900358818659e-05,
      "loss": 0.6528,
      "mean_token_accuracy": 0.80196463316679,
      "num_tokens": 519651013.0,
      "step": 16282
    },
    {
      "entropy": 0.5783246653154492,
      "epoch": 1.4981224949827776,
      "grad_norm": 0.15285392105579376,
      "learning_rate": 5.0065936762044964e-05,
      "loss": 0.5657,
      "mean_token_accuracy": 0.823984831571579,
      "num_tokens": 519682517.0,
      "step": 16283
    },
    {
      "entropy": 0.6785071063786745,
      "epoch": 1.4982145013542187,
      "grad_norm": 0.16220736503601074,
      "learning_rate": 5.006286993590333e-05,
      "loss": 0.6664,
      "mean_token_accuracy": 0.792412132024765,
      "num_tokens": 519713827.0,
      "step": 16284
    },
    {
      "entropy": 0.6664417143911123,
      "epoch": 1.49830650772566,
      "grad_norm": 0.15951085090637207,
      "learning_rate": 5.005980310976171e-05,
      "loss": 0.6447,
      "mean_token_accuracy": 0.8002356477081776,
      "num_tokens": 519744835.0,
      "step": 16285
    },
    {
      "entropy": 0.6308552001137286,
      "epoch": 1.4983985140971012,
      "grad_norm": 0.15804430842399597,
      "learning_rate": 5.005673628362009e-05,
      "loss": 0.6268,
      "mean_token_accuracy": 0.8100051768124104,
      "num_tokens": 519777473.0,
      "step": 16286
    },
    {
      "entropy": 0.6033377489075065,
      "epoch": 1.4984905204685424,
      "grad_norm": 0.1540280133485794,
      "learning_rate": 5.0053669457478456e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8178284578025341,
      "num_tokens": 519809544.0,
      "step": 16287
    },
    {
      "entropy": 0.6175946281291544,
      "epoch": 1.4985825268399837,
      "grad_norm": 0.15588587522506714,
      "learning_rate": 5.005060263133683e-05,
      "loss": 0.6122,
      "mean_token_accuracy": 0.8149074427783489,
      "num_tokens": 519841607.0,
      "step": 16288
    },
    {
      "entropy": 0.46615503961220384,
      "epoch": 1.4986745332114249,
      "grad_norm": 0.14136573672294617,
      "learning_rate": 5.00475358051952e-05,
      "loss": 0.4541,
      "mean_token_accuracy": 0.8593435361981392,
      "num_tokens": 519873770.0,
      "step": 16289
    },
    {
      "entropy": 0.6025034720078111,
      "epoch": 1.4987665395828662,
      "grad_norm": 0.15841516852378845,
      "learning_rate": 5.004446897905358e-05,
      "loss": 0.6028,
      "mean_token_accuracy": 0.8158068805932999,
      "num_tokens": 519906181.0,
      "step": 16290
    },
    {
      "entropy": 0.594915421679616,
      "epoch": 1.4988585459543073,
      "grad_norm": 0.14835573732852936,
      "learning_rate": 5.0041402152911955e-05,
      "loss": 0.5832,
      "mean_token_accuracy": 0.8199011199176311,
      "num_tokens": 519937964.0,
      "step": 16291
    },
    {
      "entropy": 0.6991227306425571,
      "epoch": 1.4989505523257485,
      "grad_norm": 0.16599014401435852,
      "learning_rate": 5.003833532677032e-05,
      "loss": 0.696,
      "mean_token_accuracy": 0.7892464362084866,
      "num_tokens": 519969638.0,
      "step": 16292
    },
    {
      "entropy": 0.5880889277905226,
      "epoch": 1.4990425586971898,
      "grad_norm": 0.1504155397415161,
      "learning_rate": 5.0035268500628704e-05,
      "loss": 0.5887,
      "mean_token_accuracy": 0.8231081552803516,
      "num_tokens": 520001884.0,
      "step": 16293
    },
    {
      "entropy": 0.5361072532832623,
      "epoch": 1.499134565068631,
      "grad_norm": 0.15097536146640778,
      "learning_rate": 5.003220167448708e-05,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8379601091146469,
      "num_tokens": 520033946.0,
      "step": 16294
    },
    {
      "entropy": 0.6614172365516424,
      "epoch": 1.4992265714400723,
      "grad_norm": 0.16643927991390228,
      "learning_rate": 5.002913484834545e-05,
      "loss": 0.6605,
      "mean_token_accuracy": 0.80240473523736,
      "num_tokens": 520065922.0,
      "step": 16295
    },
    {
      "entropy": 0.5925206979736686,
      "epoch": 1.4993185778115135,
      "grad_norm": 0.15476907789707184,
      "learning_rate": 5.002606802220382e-05,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8212395459413528,
      "num_tokens": 520097996.0,
      "step": 16296
    },
    {
      "entropy": 0.6498627765104175,
      "epoch": 1.4994105841829546,
      "grad_norm": 0.15443289279937744,
      "learning_rate": 5.00230011960622e-05,
      "loss": 0.6299,
      "mean_token_accuracy": 0.8038502521812916,
      "num_tokens": 520129303.0,
      "step": 16297
    },
    {
      "entropy": 0.4557531587779522,
      "epoch": 1.499502590554396,
      "grad_norm": 0.15332429111003876,
      "learning_rate": 5.001993436992057e-05,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8641344383358955,
      "num_tokens": 520161275.0,
      "step": 16298
    },
    {
      "entropy": 0.5607723286375403,
      "epoch": 1.499594596925837,
      "grad_norm": 0.1524176001548767,
      "learning_rate": 5.0016867543778945e-05,
      "loss": 0.5475,
      "mean_token_accuracy": 0.8290630728006363,
      "num_tokens": 520192837.0,
      "step": 16299
    },
    {
      "entropy": 0.5931360349059105,
      "epoch": 1.4996866032972784,
      "grad_norm": 0.15756237506866455,
      "learning_rate": 5.001380071763731e-05,
      "loss": 0.5791,
      "mean_token_accuracy": 0.8204130157828331,
      "num_tokens": 520225216.0,
      "step": 16300
    },
    {
      "entropy": 0.5759957628324628,
      "epoch": 1.4997786096687196,
      "grad_norm": 0.1571849137544632,
      "learning_rate": 5.00107338914957e-05,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8281769454479218,
      "num_tokens": 520257001.0,
      "step": 16301
    },
    {
      "entropy": 0.6319784559309483,
      "epoch": 1.4998706160401607,
      "grad_norm": 0.15942931175231934,
      "learning_rate": 5.000766706535407e-05,
      "loss": 0.6305,
      "mean_token_accuracy": 0.8121825158596039,
      "num_tokens": 520288551.0,
      "step": 16302
    },
    {
      "entropy": 0.7025346700102091,
      "epoch": 1.499962622411602,
      "grad_norm": 0.16002918779850006,
      "learning_rate": 5.000460023921244e-05,
      "loss": 0.6866,
      "mean_token_accuracy": 0.7883297726511955,
      "num_tokens": 520320679.0,
      "step": 16303
    },
    {
      "entropy": 0.7363336849957705,
      "epoch": 1.5000546287830432,
      "grad_norm": 0.16545289754867554,
      "learning_rate": 5.000153341307081e-05,
      "loss": 0.7331,
      "mean_token_accuracy": 0.7801949456334114,
      "num_tokens": 520352257.0,
      "step": 16304
    },
    {
      "entropy": 0.6765636149793863,
      "epoch": 1.5001466351544845,
      "grad_norm": 0.1534818708896637,
      "learning_rate": 4.9998466586929186e-05,
      "loss": 0.6516,
      "mean_token_accuracy": 0.7969535477459431,
      "num_tokens": 520383855.0,
      "step": 16305
    },
    {
      "entropy": 0.742636626586318,
      "epoch": 1.5002386415259257,
      "grad_norm": 0.16392290592193604,
      "learning_rate": 4.999539976078757e-05,
      "loss": 0.7373,
      "mean_token_accuracy": 0.7769818790256977,
      "num_tokens": 520415442.0,
      "step": 16306
    },
    {
      "entropy": 0.7794741466641426,
      "epoch": 1.5003306478973668,
      "grad_norm": 0.16315583884716034,
      "learning_rate": 4.9992332934645936e-05,
      "loss": 0.7657,
      "mean_token_accuracy": 0.7698907405138016,
      "num_tokens": 520447252.0,
      "step": 16307
    },
    {
      "entropy": 0.6145577533170581,
      "epoch": 1.5004226542688082,
      "grad_norm": 0.15769991278648376,
      "learning_rate": 4.998926610850431e-05,
      "loss": 0.6001,
      "mean_token_accuracy": 0.8164447583258152,
      "num_tokens": 520478994.0,
      "step": 16308
    },
    {
      "entropy": 0.6835497571155429,
      "epoch": 1.5005146606402493,
      "grad_norm": 0.16329145431518555,
      "learning_rate": 4.9986199282362685e-05,
      "loss": 0.6788,
      "mean_token_accuracy": 0.7974919304251671,
      "num_tokens": 520510478.0,
      "step": 16309
    },
    {
      "entropy": 0.6171260643750429,
      "epoch": 1.5006066670116907,
      "grad_norm": 0.1596706211566925,
      "learning_rate": 4.998313245622106e-05,
      "loss": 0.6146,
      "mean_token_accuracy": 0.8141146786510944,
      "num_tokens": 520542807.0,
      "step": 16310
    },
    {
      "entropy": 0.6589476228691638,
      "epoch": 1.5006986733831318,
      "grad_norm": 0.16131478548049927,
      "learning_rate": 4.9980065630079434e-05,
      "loss": 0.653,
      "mean_token_accuracy": 0.8049104660749435,
      "num_tokens": 520575290.0,
      "step": 16311
    },
    {
      "entropy": 0.5669786669313908,
      "epoch": 1.500790679754573,
      "grad_norm": 0.146406352519989,
      "learning_rate": 4.997699880393781e-05,
      "loss": 0.5589,
      "mean_token_accuracy": 0.8267541155219078,
      "num_tokens": 520606831.0,
      "step": 16312
    },
    {
      "entropy": 0.600586992688477,
      "epoch": 1.5008826861260143,
      "grad_norm": 0.1711856722831726,
      "learning_rate": 4.997393197779618e-05,
      "loss": 0.587,
      "mean_token_accuracy": 0.8176977038383484,
      "num_tokens": 520638430.0,
      "step": 16313
    },
    {
      "entropy": 0.6144288573414087,
      "epoch": 1.5009746924974554,
      "grad_norm": 0.15748196840286255,
      "learning_rate": 4.997086515165456e-05,
      "loss": 0.6118,
      "mean_token_accuracy": 0.8124257400631905,
      "num_tokens": 520670138.0,
      "step": 16314
    },
    {
      "entropy": 0.48265715688467026,
      "epoch": 1.5010666988688968,
      "grad_norm": 0.15043480694293976,
      "learning_rate": 4.9967798325512926e-05,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8531522415578365,
      "num_tokens": 520702297.0,
      "step": 16315
    },
    {
      "entropy": 0.5167260449379683,
      "epoch": 1.501158705240338,
      "grad_norm": 0.15788580477237701,
      "learning_rate": 4.99647314993713e-05,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8390432707965374,
      "num_tokens": 520734685.0,
      "step": 16316
    },
    {
      "entropy": 0.5884448327124119,
      "epoch": 1.501250711611779,
      "grad_norm": 0.1548207551240921,
      "learning_rate": 4.9961664673229675e-05,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8166767843067646,
      "num_tokens": 520766454.0,
      "step": 16317
    },
    {
      "entropy": 0.565355908125639,
      "epoch": 1.5013427179832204,
      "grad_norm": 0.15548041462898254,
      "learning_rate": 4.995859784708805e-05,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8294635340571404,
      "num_tokens": 520798453.0,
      "step": 16318
    },
    {
      "entropy": 0.5132227446883917,
      "epoch": 1.5014347243546615,
      "grad_norm": 0.14797142148017883,
      "learning_rate": 4.9955531020946425e-05,
      "loss": 0.4909,
      "mean_token_accuracy": 0.8494327664375305,
      "num_tokens": 520830098.0,
      "step": 16319
    },
    {
      "entropy": 0.6637027766555548,
      "epoch": 1.5015267307261029,
      "grad_norm": 0.16541461646556854,
      "learning_rate": 4.99524641948048e-05,
      "loss": 0.663,
      "mean_token_accuracy": 0.8005969114601612,
      "num_tokens": 520862294.0,
      "step": 16320
    },
    {
      "entropy": 0.6606342559680343,
      "epoch": 1.501618737097544,
      "grad_norm": 0.16691948473453522,
      "learning_rate": 4.9949397368663174e-05,
      "loss": 0.6624,
      "mean_token_accuracy": 0.8005809187889099,
      "num_tokens": 520893889.0,
      "step": 16321
    },
    {
      "entropy": 0.6170861590653658,
      "epoch": 1.5017107434689851,
      "grad_norm": 0.15392425656318665,
      "learning_rate": 4.994633054252155e-05,
      "loss": 0.6103,
      "mean_token_accuracy": 0.814464621245861,
      "num_tokens": 520925994.0,
      "step": 16322
    },
    {
      "entropy": 0.5764527982100844,
      "epoch": 1.5018027498404265,
      "grad_norm": 0.15187707543373108,
      "learning_rate": 4.994326371637992e-05,
      "loss": 0.551,
      "mean_token_accuracy": 0.8243389204144478,
      "num_tokens": 520957275.0,
      "step": 16323
    },
    {
      "entropy": 0.7157519850879908,
      "epoch": 1.5018947562118676,
      "grad_norm": 0.16653192043304443,
      "learning_rate": 4.994019689023829e-05,
      "loss": 0.6938,
      "mean_token_accuracy": 0.7902208007872105,
      "num_tokens": 520988439.0,
      "step": 16324
    },
    {
      "entropy": 0.6026729568839073,
      "epoch": 1.501986762583309,
      "grad_norm": 0.16110284626483917,
      "learning_rate": 4.993713006409667e-05,
      "loss": 0.5945,
      "mean_token_accuracy": 0.8186162486672401,
      "num_tokens": 521020039.0,
      "step": 16325
    },
    {
      "entropy": 0.6274428977631032,
      "epoch": 1.5020787689547501,
      "grad_norm": 0.16076160967350006,
      "learning_rate": 4.993406323795504e-05,
      "loss": 0.6216,
      "mean_token_accuracy": 0.8089384809136391,
      "num_tokens": 521051699.0,
      "step": 16326
    },
    {
      "entropy": 0.61060088314116,
      "epoch": 1.5021707753261913,
      "grad_norm": 0.15861888229846954,
      "learning_rate": 4.993099641181342e-05,
      "loss": 0.593,
      "mean_token_accuracy": 0.8178570047020912,
      "num_tokens": 521083574.0,
      "step": 16327
    },
    {
      "entropy": 0.6468413043767214,
      "epoch": 1.5022627816976326,
      "grad_norm": 0.1576913446187973,
      "learning_rate": 4.992792958567179e-05,
      "loss": 0.6402,
      "mean_token_accuracy": 0.8004915416240692,
      "num_tokens": 521115558.0,
      "step": 16328
    },
    {
      "entropy": 0.5241647940129042,
      "epoch": 1.5023547880690737,
      "grad_norm": 0.15202602744102478,
      "learning_rate": 4.9924862759530164e-05,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8410617336630821,
      "num_tokens": 521147925.0,
      "step": 16329
    },
    {
      "entropy": 0.5466732131317258,
      "epoch": 1.502446794440515,
      "grad_norm": 0.1539016216993332,
      "learning_rate": 4.992179593338854e-05,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8350077867507935,
      "num_tokens": 521179424.0,
      "step": 16330
    },
    {
      "entropy": 0.6362898498773575,
      "epoch": 1.5025388008119562,
      "grad_norm": 0.15439626574516296,
      "learning_rate": 4.9918729107246914e-05,
      "loss": 0.6034,
      "mean_token_accuracy": 0.8098659478127956,
      "num_tokens": 521210426.0,
      "step": 16331
    },
    {
      "entropy": 0.7003332730382681,
      "epoch": 1.5026308071833974,
      "grad_norm": 0.1629142016172409,
      "learning_rate": 4.991566228110529e-05,
      "loss": 0.6813,
      "mean_token_accuracy": 0.7931044325232506,
      "num_tokens": 521242715.0,
      "step": 16332
    },
    {
      "entropy": 0.672012778930366,
      "epoch": 1.5027228135548387,
      "grad_norm": 0.1583200842142105,
      "learning_rate": 4.991259545496366e-05,
      "loss": 0.6507,
      "mean_token_accuracy": 0.8012231215834618,
      "num_tokens": 521274533.0,
      "step": 16333
    },
    {
      "entropy": 0.554248783737421,
      "epoch": 1.5028148199262799,
      "grad_norm": 0.15057647228240967,
      "learning_rate": 4.990952862882203e-05,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8337250016629696,
      "num_tokens": 521305606.0,
      "step": 16334
    },
    {
      "entropy": 0.6919771060347557,
      "epoch": 1.5029068262977212,
      "grad_norm": 0.16147315502166748,
      "learning_rate": 4.990646180268041e-05,
      "loss": 0.6915,
      "mean_token_accuracy": 0.7914533466100693,
      "num_tokens": 521337507.0,
      "step": 16335
    },
    {
      "entropy": 0.590173838660121,
      "epoch": 1.5029988326691623,
      "grad_norm": 0.14875911176204681,
      "learning_rate": 4.990339497653878e-05,
      "loss": 0.5806,
      "mean_token_accuracy": 0.8264310210943222,
      "num_tokens": 521369417.0,
      "step": 16336
    },
    {
      "entropy": 0.5330926249735057,
      "epoch": 1.5030908390406035,
      "grad_norm": 0.15171056985855103,
      "learning_rate": 4.9900328150397155e-05,
      "loss": 0.5141,
      "mean_token_accuracy": 0.8418653458356857,
      "num_tokens": 521401946.0,
      "step": 16337
    },
    {
      "entropy": 0.6180866523645818,
      "epoch": 1.5031828454120448,
      "grad_norm": 0.1616845428943634,
      "learning_rate": 4.989726132425553e-05,
      "loss": 0.617,
      "mean_token_accuracy": 0.8095943331718445,
      "num_tokens": 521432986.0,
      "step": 16338
    },
    {
      "entropy": 0.6826612427830696,
      "epoch": 1.503274851783486,
      "grad_norm": 0.16339337825775146,
      "learning_rate": 4.9894194498113904e-05,
      "loss": 0.6915,
      "mean_token_accuracy": 0.7894563935697079,
      "num_tokens": 521464525.0,
      "step": 16339
    },
    {
      "entropy": 0.6250942945480347,
      "epoch": 1.5033668581549273,
      "grad_norm": 0.16019637882709503,
      "learning_rate": 4.989112767197228e-05,
      "loss": 0.6057,
      "mean_token_accuracy": 0.8118110448122025,
      "num_tokens": 521495012.0,
      "step": 16340
    },
    {
      "entropy": 0.5357760721817613,
      "epoch": 1.5034588645263685,
      "grad_norm": 0.15489044785499573,
      "learning_rate": 4.988806084583065e-05,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8377189375460148,
      "num_tokens": 521527564.0,
      "step": 16341
    },
    {
      "entropy": 0.6096489988267422,
      "epoch": 1.5035508708978096,
      "grad_norm": 0.15040992200374603,
      "learning_rate": 4.988499401968902e-05,
      "loss": 0.5964,
      "mean_token_accuracy": 0.8184109851717949,
      "num_tokens": 521559572.0,
      "step": 16342
    },
    {
      "entropy": 0.6828893842175603,
      "epoch": 1.503642877269251,
      "grad_norm": 0.16986243426799774,
      "learning_rate": 4.98819271935474e-05,
      "loss": 0.6717,
      "mean_token_accuracy": 0.7996262274682522,
      "num_tokens": 521591156.0,
      "step": 16343
    },
    {
      "entropy": 0.605214532930404,
      "epoch": 1.503734883640692,
      "grad_norm": 0.15742000937461853,
      "learning_rate": 4.987886036740577e-05,
      "loss": 0.6004,
      "mean_token_accuracy": 0.8149600327014923,
      "num_tokens": 521623391.0,
      "step": 16344
    },
    {
      "entropy": 0.6123142298310995,
      "epoch": 1.5038268900121334,
      "grad_norm": 0.16112332046031952,
      "learning_rate": 4.987579354126415e-05,
      "loss": 0.6084,
      "mean_token_accuracy": 0.814785324037075,
      "num_tokens": 521656023.0,
      "step": 16345
    },
    {
      "entropy": 0.5889260936528444,
      "epoch": 1.5039188963835746,
      "grad_norm": 0.15480844676494598,
      "learning_rate": 4.987272671512252e-05,
      "loss": 0.5748,
      "mean_token_accuracy": 0.8216128274798393,
      "num_tokens": 521687864.0,
      "step": 16346
    },
    {
      "entropy": 0.590854100883007,
      "epoch": 1.5040109027550157,
      "grad_norm": 0.15622805058956146,
      "learning_rate": 4.9869659888980894e-05,
      "loss": 0.5831,
      "mean_token_accuracy": 0.8246207609772682,
      "num_tokens": 521719707.0,
      "step": 16347
    },
    {
      "entropy": 0.6823996622115374,
      "epoch": 1.504102909126457,
      "grad_norm": 0.16466018557548523,
      "learning_rate": 4.986659306283927e-05,
      "loss": 0.6786,
      "mean_token_accuracy": 0.7959307320415974,
      "num_tokens": 521751275.0,
      "step": 16348
    },
    {
      "entropy": 0.6200321642681956,
      "epoch": 1.5041949154978982,
      "grad_norm": 0.1539524793624878,
      "learning_rate": 4.9863526236697644e-05,
      "loss": 0.6007,
      "mean_token_accuracy": 0.8165811561048031,
      "num_tokens": 521782991.0,
      "step": 16349
    },
    {
      "entropy": 0.5988291455432773,
      "epoch": 1.5042869218693395,
      "grad_norm": 0.16265030205249786,
      "learning_rate": 4.986045941055602e-05,
      "loss": 0.5876,
      "mean_token_accuracy": 0.8198855109512806,
      "num_tokens": 521815238.0,
      "step": 16350
    },
    {
      "entropy": 0.6606251541525126,
      "epoch": 1.5043789282407807,
      "grad_norm": 0.15285958349704742,
      "learning_rate": 4.985739258441439e-05,
      "loss": 0.6444,
      "mean_token_accuracy": 0.7984039708971977,
      "num_tokens": 521847061.0,
      "step": 16351
    },
    {
      "entropy": 0.6123846657574177,
      "epoch": 1.5044709346122218,
      "grad_norm": 0.15396662056446075,
      "learning_rate": 4.985432575827277e-05,
      "loss": 0.5983,
      "mean_token_accuracy": 0.8138068281114101,
      "num_tokens": 521879125.0,
      "step": 16352
    },
    {
      "entropy": 0.6343252882361412,
      "epoch": 1.5045629409836632,
      "grad_norm": 0.16085012257099152,
      "learning_rate": 4.985125893213114e-05,
      "loss": 0.6325,
      "mean_token_accuracy": 0.8045223876833916,
      "num_tokens": 521911053.0,
      "step": 16353
    },
    {
      "entropy": 0.516839895863086,
      "epoch": 1.5046549473551043,
      "grad_norm": 0.1567508429288864,
      "learning_rate": 4.984819210598952e-05,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8465340286493301,
      "num_tokens": 521943078.0,
      "step": 16354
    },
    {
      "entropy": 0.5136163460556418,
      "epoch": 1.5047469537265457,
      "grad_norm": 0.14551512897014618,
      "learning_rate": 4.9845125279847885e-05,
      "loss": 0.505,
      "mean_token_accuracy": 0.8469688259065151,
      "num_tokens": 521975243.0,
      "step": 16355
    },
    {
      "entropy": 0.6883100252598524,
      "epoch": 1.5048389600979868,
      "grad_norm": 0.1614481657743454,
      "learning_rate": 4.9842058453706266e-05,
      "loss": 0.6854,
      "mean_token_accuracy": 0.7953351065516472,
      "num_tokens": 522007336.0,
      "step": 16356
    },
    {
      "entropy": 0.5746427476406097,
      "epoch": 1.504930966469428,
      "grad_norm": 0.15419307351112366,
      "learning_rate": 4.9838991627564634e-05,
      "loss": 0.564,
      "mean_token_accuracy": 0.8265024647116661,
      "num_tokens": 522039816.0,
      "step": 16357
    },
    {
      "entropy": 0.7247334457933903,
      "epoch": 1.5050229728408693,
      "grad_norm": 0.15982313454151154,
      "learning_rate": 4.983592480142301e-05,
      "loss": 0.7269,
      "mean_token_accuracy": 0.7811844870448112,
      "num_tokens": 522071975.0,
      "step": 16358
    },
    {
      "entropy": 0.49009907338768244,
      "epoch": 1.5051149792123104,
      "grad_norm": 0.14270493388175964,
      "learning_rate": 4.9832857975281383e-05,
      "loss": 0.4815,
      "mean_token_accuracy": 0.8486810252070427,
      "num_tokens": 522104271.0,
      "step": 16359
    },
    {
      "entropy": 0.6923829847946763,
      "epoch": 1.5052069855837518,
      "grad_norm": 0.1629944145679474,
      "learning_rate": 4.982979114913976e-05,
      "loss": 0.6843,
      "mean_token_accuracy": 0.7909551374614239,
      "num_tokens": 522135201.0,
      "step": 16360
    },
    {
      "entropy": 0.5499452576041222,
      "epoch": 1.505298991955193,
      "grad_norm": 0.14778843522071838,
      "learning_rate": 4.982672432299813e-05,
      "loss": 0.5478,
      "mean_token_accuracy": 0.8332520090043545,
      "num_tokens": 522167777.0,
      "step": 16361
    },
    {
      "entropy": 0.5414037951268256,
      "epoch": 1.505390998326634,
      "grad_norm": 0.1517229825258255,
      "learning_rate": 4.982365749685651e-05,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8390632644295692,
      "num_tokens": 522200545.0,
      "step": 16362
    },
    {
      "entropy": 0.651836359873414,
      "epoch": 1.5054830046980754,
      "grad_norm": 0.15506939589977264,
      "learning_rate": 4.9820590670714875e-05,
      "loss": 0.6445,
      "mean_token_accuracy": 0.8046170324087143,
      "num_tokens": 522232010.0,
      "step": 16363
    },
    {
      "entropy": 0.6348950816318393,
      "epoch": 1.5055750110695165,
      "grad_norm": 0.15664079785346985,
      "learning_rate": 4.981752384457326e-05,
      "loss": 0.6209,
      "mean_token_accuracy": 0.8099624514579773,
      "num_tokens": 522264207.0,
      "step": 16364
    },
    {
      "entropy": 0.6229902543127537,
      "epoch": 1.5056670174409579,
      "grad_norm": 0.15486881136894226,
      "learning_rate": 4.9814457018431625e-05,
      "loss": 0.608,
      "mean_token_accuracy": 0.8143616318702698,
      "num_tokens": 522296310.0,
      "step": 16365
    },
    {
      "entropy": 0.661310900002718,
      "epoch": 1.505759023812399,
      "grad_norm": 0.1638251692056656,
      "learning_rate": 4.9811390192290006e-05,
      "loss": 0.6589,
      "mean_token_accuracy": 0.7962377034127712,
      "num_tokens": 522327130.0,
      "step": 16366
    },
    {
      "entropy": 0.6002229247242212,
      "epoch": 1.5058510301838401,
      "grad_norm": 0.15928366780281067,
      "learning_rate": 4.9808323366148374e-05,
      "loss": 0.5898,
      "mean_token_accuracy": 0.8161758594214916,
      "num_tokens": 522358537.0,
      "step": 16367
    },
    {
      "entropy": 0.68096605502069,
      "epoch": 1.5059430365552815,
      "grad_norm": 0.15672406554222107,
      "learning_rate": 4.980525654000675e-05,
      "loss": 0.6735,
      "mean_token_accuracy": 0.793179627507925,
      "num_tokens": 522390928.0,
      "step": 16368
    },
    {
      "entropy": 0.6116173304617405,
      "epoch": 1.5060350429267226,
      "grad_norm": 0.1609281450510025,
      "learning_rate": 4.980218971386512e-05,
      "loss": 0.5947,
      "mean_token_accuracy": 0.8193437233567238,
      "num_tokens": 522423117.0,
      "step": 16369
    },
    {
      "entropy": 0.6446241661906242,
      "epoch": 1.506127049298164,
      "grad_norm": 0.15380996465682983,
      "learning_rate": 4.97991228877235e-05,
      "loss": 0.6399,
      "mean_token_accuracy": 0.8038244545459747,
      "num_tokens": 522455328.0,
      "step": 16370
    },
    {
      "entropy": 0.664314977824688,
      "epoch": 1.5062190556696051,
      "grad_norm": 0.16039705276489258,
      "learning_rate": 4.979605606158187e-05,
      "loss": 0.6615,
      "mean_token_accuracy": 0.7959950007498264,
      "num_tokens": 522486662.0,
      "step": 16371
    },
    {
      "entropy": 0.5053537769708782,
      "epoch": 1.5063110620410463,
      "grad_norm": 0.14630873501300812,
      "learning_rate": 4.979298923544025e-05,
      "loss": 0.478,
      "mean_token_accuracy": 0.8469624146819115,
      "num_tokens": 522518179.0,
      "step": 16372
    },
    {
      "entropy": 0.6810359749943018,
      "epoch": 1.5064030684124876,
      "grad_norm": 0.1652657389640808,
      "learning_rate": 4.9789922409298615e-05,
      "loss": 0.6677,
      "mean_token_accuracy": 0.797539509832859,
      "num_tokens": 522548889.0,
      "step": 16373
    },
    {
      "entropy": 0.511934501118958,
      "epoch": 1.5064950747839287,
      "grad_norm": 0.14738023281097412,
      "learning_rate": 4.9786855583156996e-05,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8443538323044777,
      "num_tokens": 522580461.0,
      "step": 16374
    },
    {
      "entropy": 0.6109574665315449,
      "epoch": 1.50658708115537,
      "grad_norm": 0.15068316459655762,
      "learning_rate": 4.9783788757015364e-05,
      "loss": 0.5988,
      "mean_token_accuracy": 0.8178425133228302,
      "num_tokens": 522612839.0,
      "step": 16375
    },
    {
      "entropy": 0.6887601399794221,
      "epoch": 1.5066790875268112,
      "grad_norm": 0.17140604555606842,
      "learning_rate": 4.978072193087374e-05,
      "loss": 0.6941,
      "mean_token_accuracy": 0.7933692522346973,
      "num_tokens": 522644274.0,
      "step": 16376
    },
    {
      "entropy": 0.6364605091512203,
      "epoch": 1.5067710938982524,
      "grad_norm": 0.16204366087913513,
      "learning_rate": 4.9777655104732114e-05,
      "loss": 0.6191,
      "mean_token_accuracy": 0.8093516826629639,
      "num_tokens": 522675687.0,
      "step": 16377
    },
    {
      "entropy": 0.7081547882407904,
      "epoch": 1.5068631002696937,
      "grad_norm": 0.168548122048378,
      "learning_rate": 4.977458827859049e-05,
      "loss": 0.7062,
      "mean_token_accuracy": 0.7828826904296875,
      "num_tokens": 522707312.0,
      "step": 16378
    },
    {
      "entropy": 0.6621488619130105,
      "epoch": 1.5069551066411349,
      "grad_norm": 0.15656279027462006,
      "learning_rate": 4.977152145244886e-05,
      "loss": 0.6608,
      "mean_token_accuracy": 0.796688124537468,
      "num_tokens": 522739333.0,
      "step": 16379
    },
    {
      "entropy": 0.6713601090013981,
      "epoch": 1.5070471130125762,
      "grad_norm": 0.15973584353923798,
      "learning_rate": 4.976845462630724e-05,
      "loss": 0.6631,
      "mean_token_accuracy": 0.797757014632225,
      "num_tokens": 522771194.0,
      "step": 16380
    },
    {
      "entropy": 0.6710949223488569,
      "epoch": 1.5071391193840173,
      "grad_norm": 0.16041572391986847,
      "learning_rate": 4.976538780016561e-05,
      "loss": 0.6688,
      "mean_token_accuracy": 0.7955072075128555,
      "num_tokens": 522803190.0,
      "step": 16381
    },
    {
      "entropy": 0.7700161170214415,
      "epoch": 1.5072311257554585,
      "grad_norm": 0.16035889089107513,
      "learning_rate": 4.976232097402399e-05,
      "loss": 0.7542,
      "mean_token_accuracy": 0.7684964910149574,
      "num_tokens": 522835372.0,
      "step": 16382
    },
    {
      "entropy": 0.6049975529313087,
      "epoch": 1.5073231321268998,
      "grad_norm": 0.15574412047863007,
      "learning_rate": 4.975925414788236e-05,
      "loss": 0.5979,
      "mean_token_accuracy": 0.815814733505249,
      "num_tokens": 522867552.0,
      "step": 16383
    },
    {
      "entropy": 0.6840643305331469,
      "epoch": 1.507415138498341,
      "grad_norm": 0.16440121829509735,
      "learning_rate": 4.975618732174073e-05,
      "loss": 0.656,
      "mean_token_accuracy": 0.7950108200311661,
      "num_tokens": 522898991.0,
      "step": 16384
    },
    {
      "entropy": 0.6332362778484821,
      "epoch": 1.5075071448697823,
      "grad_norm": 0.15274281799793243,
      "learning_rate": 4.975312049559911e-05,
      "loss": 0.6264,
      "mean_token_accuracy": 0.8071577586233616,
      "num_tokens": 522930679.0,
      "step": 16385
    },
    {
      "entropy": 0.610038090031594,
      "epoch": 1.5075991512412235,
      "grad_norm": 0.15372918546199799,
      "learning_rate": 4.975005366945748e-05,
      "loss": 0.5969,
      "mean_token_accuracy": 0.8149901032447815,
      "num_tokens": 522962916.0,
      "step": 16386
    },
    {
      "entropy": 0.5279651284217834,
      "epoch": 1.5076911576126646,
      "grad_norm": 0.15311846137046814,
      "learning_rate": 4.974698684331586e-05,
      "loss": 0.5098,
      "mean_token_accuracy": 0.8416672646999359,
      "num_tokens": 522994928.0,
      "step": 16387
    },
    {
      "entropy": 0.5125464517623186,
      "epoch": 1.507783163984106,
      "grad_norm": 0.15288028120994568,
      "learning_rate": 4.974392001717423e-05,
      "loss": 0.519,
      "mean_token_accuracy": 0.8394517377018929,
      "num_tokens": 523025942.0,
      "step": 16388
    },
    {
      "entropy": 0.5405121315270662,
      "epoch": 1.507875170355547,
      "grad_norm": 0.15855994820594788,
      "learning_rate": 4.97408531910326e-05,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8373795635998249,
      "num_tokens": 523058253.0,
      "step": 16389
    },
    {
      "entropy": 0.6258140671998262,
      "epoch": 1.5079671767269884,
      "grad_norm": 0.15728384256362915,
      "learning_rate": 4.973778636489098e-05,
      "loss": 0.6175,
      "mean_token_accuracy": 0.8078401833772659,
      "num_tokens": 523090047.0,
      "step": 16390
    },
    {
      "entropy": 0.6542252227663994,
      "epoch": 1.5080591830984296,
      "grad_norm": 0.16102662682533264,
      "learning_rate": 4.973471953874935e-05,
      "loss": 0.6486,
      "mean_token_accuracy": 0.8007468394935131,
      "num_tokens": 523121073.0,
      "step": 16391
    },
    {
      "entropy": 0.6405909806489944,
      "epoch": 1.5081511894698707,
      "grad_norm": 0.15657593309879303,
      "learning_rate": 4.9731652712607726e-05,
      "loss": 0.6269,
      "mean_token_accuracy": 0.8057692758738995,
      "num_tokens": 523152974.0,
      "step": 16392
    },
    {
      "entropy": 0.5038892542943358,
      "epoch": 1.508243195841312,
      "grad_norm": 0.14662683010101318,
      "learning_rate": 4.97285858864661e-05,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8463908359408379,
      "num_tokens": 523185051.0,
      "step": 16393
    },
    {
      "entropy": 0.5880353301763535,
      "epoch": 1.5083352022127532,
      "grad_norm": 0.15613889694213867,
      "learning_rate": 4.972551906032447e-05,
      "loss": 0.5732,
      "mean_token_accuracy": 0.8173594735562801,
      "num_tokens": 523216556.0,
      "step": 16394
    },
    {
      "entropy": 0.6148339966312051,
      "epoch": 1.5084272085841945,
      "grad_norm": 0.15645168721675873,
      "learning_rate": 4.972245223418285e-05,
      "loss": 0.6187,
      "mean_token_accuracy": 0.8116189315915108,
      "num_tokens": 523249176.0,
      "step": 16395
    },
    {
      "entropy": 0.5655514486134052,
      "epoch": 1.5085192149556357,
      "grad_norm": 0.1485845446586609,
      "learning_rate": 4.971938540804122e-05,
      "loss": 0.5433,
      "mean_token_accuracy": 0.829513181000948,
      "num_tokens": 523280897.0,
      "step": 16396
    },
    {
      "entropy": 0.5891941310837865,
      "epoch": 1.5086112213270768,
      "grad_norm": 0.14833763241767883,
      "learning_rate": 4.971631858189959e-05,
      "loss": 0.5753,
      "mean_token_accuracy": 0.8239930272102356,
      "num_tokens": 523313137.0,
      "step": 16397
    },
    {
      "entropy": 0.5971573889255524,
      "epoch": 1.5087032276985182,
      "grad_norm": 0.15337781608104706,
      "learning_rate": 4.971325175575797e-05,
      "loss": 0.5947,
      "mean_token_accuracy": 0.8214845359325409,
      "num_tokens": 523344664.0,
      "step": 16398
    },
    {
      "entropy": 0.5971156619489193,
      "epoch": 1.5087952340699593,
      "grad_norm": 0.15257154405117035,
      "learning_rate": 4.971018492961634e-05,
      "loss": 0.5762,
      "mean_token_accuracy": 0.8205722942948341,
      "num_tokens": 523376542.0,
      "step": 16399
    },
    {
      "entropy": 0.517319900682196,
      "epoch": 1.5088872404414007,
      "grad_norm": 0.14357848465442657,
      "learning_rate": 4.970711810347472e-05,
      "loss": 0.5109,
      "mean_token_accuracy": 0.8417353853583336,
      "num_tokens": 523408788.0,
      "step": 16400
    },
    {
      "entropy": 0.5123237743973732,
      "epoch": 1.5089792468128418,
      "grad_norm": 0.15082401037216187,
      "learning_rate": 4.970405127733309e-05,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8432066068053246,
      "num_tokens": 523441246.0,
      "step": 16401
    },
    {
      "entropy": 0.6093061715364456,
      "epoch": 1.509071253184283,
      "grad_norm": 0.15399259328842163,
      "learning_rate": 4.970098445119146e-05,
      "loss": 0.6133,
      "mean_token_accuracy": 0.8137886039912701,
      "num_tokens": 523472672.0,
      "step": 16402
    },
    {
      "entropy": 0.6982007715851068,
      "epoch": 1.5091632595557243,
      "grad_norm": 0.16020990908145905,
      "learning_rate": 4.969791762504984e-05,
      "loss": 0.6899,
      "mean_token_accuracy": 0.7922412753105164,
      "num_tokens": 523503641.0,
      "step": 16403
    },
    {
      "entropy": 0.591031963005662,
      "epoch": 1.5092552659271654,
      "grad_norm": 0.14901219308376312,
      "learning_rate": 4.969485079890821e-05,
      "loss": 0.5757,
      "mean_token_accuracy": 0.8229443095624447,
      "num_tokens": 523534947.0,
      "step": 16404
    },
    {
      "entropy": 0.545577097684145,
      "epoch": 1.5093472722986068,
      "grad_norm": 0.16748638451099396,
      "learning_rate": 4.969178397276659e-05,
      "loss": 0.5249,
      "mean_token_accuracy": 0.8406426757574081,
      "num_tokens": 523566579.0,
      "step": 16405
    },
    {
      "entropy": 0.6087081488221884,
      "epoch": 1.509439278670048,
      "grad_norm": 0.1503390371799469,
      "learning_rate": 4.968871714662496e-05,
      "loss": 0.5973,
      "mean_token_accuracy": 0.8175829611718655,
      "num_tokens": 523598336.0,
      "step": 16406
    },
    {
      "entropy": 0.4978989977389574,
      "epoch": 1.509531285041489,
      "grad_norm": 0.14492081105709076,
      "learning_rate": 4.968565032048333e-05,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8511750139296055,
      "num_tokens": 523630510.0,
      "step": 16407
    },
    {
      "entropy": 0.6004605125635862,
      "epoch": 1.5096232914129304,
      "grad_norm": 0.1493738442659378,
      "learning_rate": 4.968258349434171e-05,
      "loss": 0.5775,
      "mean_token_accuracy": 0.8216243498027325,
      "num_tokens": 523662080.0,
      "step": 16408
    },
    {
      "entropy": 0.7106331698596478,
      "epoch": 1.5097152977843715,
      "grad_norm": 0.16376349329948425,
      "learning_rate": 4.967951666820008e-05,
      "loss": 0.7077,
      "mean_token_accuracy": 0.786718025803566,
      "num_tokens": 523694266.0,
      "step": 16409
    },
    {
      "entropy": 0.5739306006580591,
      "epoch": 1.5098073041558129,
      "grad_norm": 0.15150421857833862,
      "learning_rate": 4.9676449842058457e-05,
      "loss": 0.5576,
      "mean_token_accuracy": 0.8294266238808632,
      "num_tokens": 523726330.0,
      "step": 16410
    },
    {
      "entropy": 0.7084923181682825,
      "epoch": 1.509899310527254,
      "grad_norm": 0.16280624270439148,
      "learning_rate": 4.967338301591683e-05,
      "loss": 0.7189,
      "mean_token_accuracy": 0.7853424660861492,
      "num_tokens": 523758352.0,
      "step": 16411
    },
    {
      "entropy": 0.5280152820050716,
      "epoch": 1.5099913168986951,
      "grad_norm": 0.15145239233970642,
      "learning_rate": 4.9670316189775206e-05,
      "loss": 0.5154,
      "mean_token_accuracy": 0.8397051580250263,
      "num_tokens": 523790263.0,
      "step": 16412
    },
    {
      "entropy": 0.5779776112176478,
      "epoch": 1.5100833232701365,
      "grad_norm": 0.15939028561115265,
      "learning_rate": 4.966724936363358e-05,
      "loss": 0.5658,
      "mean_token_accuracy": 0.825232345610857,
      "num_tokens": 523821311.0,
      "step": 16413
    },
    {
      "entropy": 0.5398863302543759,
      "epoch": 1.5101753296415776,
      "grad_norm": 0.1521347463130951,
      "learning_rate": 4.9664182537491955e-05,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8375434540212154,
      "num_tokens": 523853022.0,
      "step": 16414
    },
    {
      "entropy": 0.5419476022943854,
      "epoch": 1.510267336013019,
      "grad_norm": 0.1546827256679535,
      "learning_rate": 4.966111571135032e-05,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8361411802470684,
      "num_tokens": 523885266.0,
      "step": 16415
    },
    {
      "entropy": 0.5780805023387074,
      "epoch": 1.5103593423844601,
      "grad_norm": 0.15738394856452942,
      "learning_rate": 4.9658048885208704e-05,
      "loss": 0.5705,
      "mean_token_accuracy": 0.824682742357254,
      "num_tokens": 523917729.0,
      "step": 16416
    },
    {
      "entropy": 0.6219223467633128,
      "epoch": 1.5104513487559013,
      "grad_norm": 0.16160579025745392,
      "learning_rate": 4.965498205906707e-05,
      "loss": 0.6144,
      "mean_token_accuracy": 0.8146139942109585,
      "num_tokens": 523949057.0,
      "step": 16417
    },
    {
      "entropy": 0.5480977077968419,
      "epoch": 1.5105433551273426,
      "grad_norm": 0.15414302051067352,
      "learning_rate": 4.965191523292545e-05,
      "loss": 0.537,
      "mean_token_accuracy": 0.8333761915564537,
      "num_tokens": 523981597.0,
      "step": 16418
    },
    {
      "entropy": 0.5258810771629214,
      "epoch": 1.5106353614987837,
      "grad_norm": 0.14664778113365173,
      "learning_rate": 4.964884840678382e-05,
      "loss": 0.504,
      "mean_token_accuracy": 0.8419066853821278,
      "num_tokens": 524013112.0,
      "step": 16419
    },
    {
      "entropy": 0.6719522355124354,
      "epoch": 1.510727367870225,
      "grad_norm": 0.16304834187030792,
      "learning_rate": 4.9645781580642196e-05,
      "loss": 0.6549,
      "mean_token_accuracy": 0.7993730939924717,
      "num_tokens": 524044973.0,
      "step": 16420
    },
    {
      "entropy": 0.5954354740679264,
      "epoch": 1.5108193742416662,
      "grad_norm": 0.1569843590259552,
      "learning_rate": 4.964271475450057e-05,
      "loss": 0.5772,
      "mean_token_accuracy": 0.8242672644555569,
      "num_tokens": 524077054.0,
      "step": 16421
    },
    {
      "entropy": 0.6046548951417208,
      "epoch": 1.5109113806131074,
      "grad_norm": 0.15922699868679047,
      "learning_rate": 4.9639647928358946e-05,
      "loss": 0.5885,
      "mean_token_accuracy": 0.8159687407314777,
      "num_tokens": 524109480.0,
      "step": 16422
    },
    {
      "entropy": 0.6498640216886997,
      "epoch": 1.5110033869845487,
      "grad_norm": 0.15532690286636353,
      "learning_rate": 4.9636581102217313e-05,
      "loss": 0.6489,
      "mean_token_accuracy": 0.8051257468760014,
      "num_tokens": 524142063.0,
      "step": 16423
    },
    {
      "entropy": 0.622819465585053,
      "epoch": 1.5110953933559899,
      "grad_norm": 0.15316249430179596,
      "learning_rate": 4.9633514276075695e-05,
      "loss": 0.6117,
      "mean_token_accuracy": 0.8122252151370049,
      "num_tokens": 524174182.0,
      "step": 16424
    },
    {
      "entropy": 0.5577126899734139,
      "epoch": 1.5111873997274312,
      "grad_norm": 0.15662942826747894,
      "learning_rate": 4.963044744993406e-05,
      "loss": 0.524,
      "mean_token_accuracy": 0.837782695889473,
      "num_tokens": 524204952.0,
      "step": 16425
    },
    {
      "entropy": 0.6629860866814852,
      "epoch": 1.5112794060988723,
      "grad_norm": 0.16207347810268402,
      "learning_rate": 4.9627380623792444e-05,
      "loss": 0.6608,
      "mean_token_accuracy": 0.7958904057741165,
      "num_tokens": 524236821.0,
      "step": 16426
    },
    {
      "entropy": 0.5328289857134223,
      "epoch": 1.5113714124703135,
      "grad_norm": 0.1489851474761963,
      "learning_rate": 4.962431379765081e-05,
      "loss": 0.5201,
      "mean_token_accuracy": 0.840586643666029,
      "num_tokens": 524268858.0,
      "step": 16427
    },
    {
      "entropy": 0.5826719817705452,
      "epoch": 1.5114634188417548,
      "grad_norm": 0.15626239776611328,
      "learning_rate": 4.962124697150919e-05,
      "loss": 0.584,
      "mean_token_accuracy": 0.8219802901148796,
      "num_tokens": 524300971.0,
      "step": 16428
    },
    {
      "entropy": 0.6717667207121849,
      "epoch": 1.511555425213196,
      "grad_norm": 0.16686366498470306,
      "learning_rate": 4.961818014536756e-05,
      "loss": 0.6627,
      "mean_token_accuracy": 0.7982704825699329,
      "num_tokens": 524332777.0,
      "step": 16429
    },
    {
      "entropy": 0.5774929886683822,
      "epoch": 1.5116474315846373,
      "grad_norm": 0.15422774851322174,
      "learning_rate": 4.9615113319225936e-05,
      "loss": 0.5729,
      "mean_token_accuracy": 0.8236156664788723,
      "num_tokens": 524364716.0,
      "step": 16430
    },
    {
      "entropy": 0.6304360684007406,
      "epoch": 1.5117394379560785,
      "grad_norm": 0.16395984590053558,
      "learning_rate": 4.961204649308431e-05,
      "loss": 0.614,
      "mean_token_accuracy": 0.8093161284923553,
      "num_tokens": 524396659.0,
      "step": 16431
    },
    {
      "entropy": 0.6699433475732803,
      "epoch": 1.5118314443275196,
      "grad_norm": 0.1693120300769806,
      "learning_rate": 4.9608979666942685e-05,
      "loss": 0.6663,
      "mean_token_accuracy": 0.7983846440911293,
      "num_tokens": 524428454.0,
      "step": 16432
    },
    {
      "entropy": 0.5826223567128181,
      "epoch": 1.511923450698961,
      "grad_norm": 0.15558652579784393,
      "learning_rate": 4.960591284080105e-05,
      "loss": 0.5603,
      "mean_token_accuracy": 0.8275687769055367,
      "num_tokens": 524460250.0,
      "step": 16433
    },
    {
      "entropy": 0.5489384457468987,
      "epoch": 1.512015457070402,
      "grad_norm": 0.14771316945552826,
      "learning_rate": 4.9602846014659435e-05,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8315367288887501,
      "num_tokens": 524492594.0,
      "step": 16434
    },
    {
      "entropy": 0.6260805353522301,
      "epoch": 1.5121074634418434,
      "grad_norm": 0.1642848253250122,
      "learning_rate": 4.95997791885178e-05,
      "loss": 0.6222,
      "mean_token_accuracy": 0.8115999847650528,
      "num_tokens": 524524555.0,
      "step": 16435
    },
    {
      "entropy": 0.7571729626506567,
      "epoch": 1.5121994698132846,
      "grad_norm": 0.16269752383232117,
      "learning_rate": 4.959671236237618e-05,
      "loss": 0.7539,
      "mean_token_accuracy": 0.7712948396801949,
      "num_tokens": 524556433.0,
      "step": 16436
    },
    {
      "entropy": 0.7192706279456615,
      "epoch": 1.5122914761847257,
      "grad_norm": 0.15937457978725433,
      "learning_rate": 4.959364553623455e-05,
      "loss": 0.7095,
      "mean_token_accuracy": 0.7877982668578625,
      "num_tokens": 524588948.0,
      "step": 16437
    },
    {
      "entropy": 0.4913281223271042,
      "epoch": 1.512383482556167,
      "grad_norm": 0.14792819321155548,
      "learning_rate": 4.9590578710092926e-05,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8479973636567593,
      "num_tokens": 524621108.0,
      "step": 16438
    },
    {
      "entropy": 0.5349740805104375,
      "epoch": 1.5124754889276082,
      "grad_norm": 0.14718899130821228,
      "learning_rate": 4.95875118839513e-05,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8406924791634083,
      "num_tokens": 524653376.0,
      "step": 16439
    },
    {
      "entropy": 0.6239903196692467,
      "epoch": 1.5125674952990495,
      "grad_norm": 0.15908290445804596,
      "learning_rate": 4.9584445057809676e-05,
      "loss": 0.621,
      "mean_token_accuracy": 0.8116497881710529,
      "num_tokens": 524685555.0,
      "step": 16440
    },
    {
      "entropy": 0.7222330272197723,
      "epoch": 1.5126595016704907,
      "grad_norm": 0.16524969041347504,
      "learning_rate": 4.9581378231668044e-05,
      "loss": 0.7113,
      "mean_token_accuracy": 0.7840781509876251,
      "num_tokens": 524717221.0,
      "step": 16441
    },
    {
      "entropy": 0.6140110176056623,
      "epoch": 1.5127515080419318,
      "grad_norm": 0.15647491812705994,
      "learning_rate": 4.9578311405526425e-05,
      "loss": 0.6044,
      "mean_token_accuracy": 0.8132853843271732,
      "num_tokens": 524749038.0,
      "step": 16442
    },
    {
      "entropy": 0.5666463132947683,
      "epoch": 1.5128435144133732,
      "grad_norm": 0.15520429611206055,
      "learning_rate": 4.95752445793848e-05,
      "loss": 0.5656,
      "mean_token_accuracy": 0.8280054554343224,
      "num_tokens": 524780631.0,
      "step": 16443
    },
    {
      "entropy": 0.6728631742298603,
      "epoch": 1.5129355207848143,
      "grad_norm": 0.16151422262191772,
      "learning_rate": 4.957217775324317e-05,
      "loss": 0.6575,
      "mean_token_accuracy": 0.7990815080702305,
      "num_tokens": 524811756.0,
      "step": 16444
    },
    {
      "entropy": 0.6598709244281054,
      "epoch": 1.5130275271562557,
      "grad_norm": 0.16053742170333862,
      "learning_rate": 4.956911092710155e-05,
      "loss": 0.6541,
      "mean_token_accuracy": 0.8016269989311695,
      "num_tokens": 524843585.0,
      "step": 16445
    },
    {
      "entropy": 0.6791190514340997,
      "epoch": 1.5131195335276968,
      "grad_norm": 0.15702283382415771,
      "learning_rate": 4.956604410095992e-05,
      "loss": 0.6779,
      "mean_token_accuracy": 0.7922293283045292,
      "num_tokens": 524876303.0,
      "step": 16446
    },
    {
      "entropy": 0.4932596944272518,
      "epoch": 1.513211539899138,
      "grad_norm": 0.15114019811153412,
      "learning_rate": 4.95629772748183e-05,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8466220013797283,
      "num_tokens": 524908585.0,
      "step": 16447
    },
    {
      "entropy": 0.5039003826677799,
      "epoch": 1.5133035462705793,
      "grad_norm": 0.14042514562606812,
      "learning_rate": 4.9559910448676666e-05,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8456263318657875,
      "num_tokens": 524940294.0,
      "step": 16448
    },
    {
      "entropy": 0.5184528848621994,
      "epoch": 1.5133955526420204,
      "grad_norm": 0.15162396430969238,
      "learning_rate": 4.955684362253504e-05,
      "loss": 0.5068,
      "mean_token_accuracy": 0.8409995548427105,
      "num_tokens": 524972616.0,
      "step": 16449
    },
    {
      "entropy": 0.7013223059475422,
      "epoch": 1.5134875590134618,
      "grad_norm": 0.16470085084438324,
      "learning_rate": 4.9553776796393415e-05,
      "loss": 0.6924,
      "mean_token_accuracy": 0.7875389643013477,
      "num_tokens": 525003609.0,
      "step": 16450
    },
    {
      "entropy": 0.6029633302241564,
      "epoch": 1.513579565384903,
      "grad_norm": 0.15726707875728607,
      "learning_rate": 4.955070997025179e-05,
      "loss": 0.5962,
      "mean_token_accuracy": 0.819722656160593,
      "num_tokens": 525035474.0,
      "step": 16451
    },
    {
      "entropy": 0.6104047354310751,
      "epoch": 1.513671571756344,
      "grad_norm": 0.15838263928890228,
      "learning_rate": 4.9547643144110165e-05,
      "loss": 0.6028,
      "mean_token_accuracy": 0.8127826973795891,
      "num_tokens": 525067020.0,
      "step": 16452
    },
    {
      "entropy": 0.5645928531885147,
      "epoch": 1.5137635781277854,
      "grad_norm": 0.14933742582798004,
      "learning_rate": 4.954457631796854e-05,
      "loss": 0.561,
      "mean_token_accuracy": 0.8317053876817226,
      "num_tokens": 525098820.0,
      "step": 16453
    },
    {
      "entropy": 0.5712949051521719,
      "epoch": 1.5138555844992265,
      "grad_norm": 0.15545330941677094,
      "learning_rate": 4.954150949182691e-05,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8312475197017193,
      "num_tokens": 525128777.0,
      "step": 16454
    },
    {
      "entropy": 0.6536805341020226,
      "epoch": 1.5139475908706679,
      "grad_norm": 0.16069568693637848,
      "learning_rate": 4.953844266568529e-05,
      "loss": 0.6279,
      "mean_token_accuracy": 0.8076161704957485,
      "num_tokens": 525160754.0,
      "step": 16455
    },
    {
      "entropy": 0.5275046667084098,
      "epoch": 1.514039597242109,
      "grad_norm": 0.14476178586483002,
      "learning_rate": 4.9535375839543656e-05,
      "loss": 0.508,
      "mean_token_accuracy": 0.8432426564395428,
      "num_tokens": 525192733.0,
      "step": 16456
    },
    {
      "entropy": 0.7434333618730307,
      "epoch": 1.5141316036135501,
      "grad_norm": 0.16461771726608276,
      "learning_rate": 4.953230901340203e-05,
      "loss": 0.7252,
      "mean_token_accuracy": 0.7770531587302685,
      "num_tokens": 525224620.0,
      "step": 16457
    },
    {
      "entropy": 0.6483962088823318,
      "epoch": 1.5142236099849915,
      "grad_norm": 0.1556069254875183,
      "learning_rate": 4.9529242187260406e-05,
      "loss": 0.6293,
      "mean_token_accuracy": 0.8005147762596607,
      "num_tokens": 525256273.0,
      "step": 16458
    },
    {
      "entropy": 0.6143429661169648,
      "epoch": 1.5143156163564326,
      "grad_norm": 0.15655596554279327,
      "learning_rate": 4.952617536111878e-05,
      "loss": 0.6031,
      "mean_token_accuracy": 0.8129107505083084,
      "num_tokens": 525288052.0,
      "step": 16459
    },
    {
      "entropy": 0.67038544267416,
      "epoch": 1.514407622727874,
      "grad_norm": 0.15766234695911407,
      "learning_rate": 4.9523108534977155e-05,
      "loss": 0.6529,
      "mean_token_accuracy": 0.8027070835232735,
      "num_tokens": 525319923.0,
      "step": 16460
    },
    {
      "entropy": 0.7235256303101778,
      "epoch": 1.5144996290993151,
      "grad_norm": 0.16536614298820496,
      "learning_rate": 4.952004170883553e-05,
      "loss": 0.7062,
      "mean_token_accuracy": 0.7866819463670254,
      "num_tokens": 525352257.0,
      "step": 16461
    },
    {
      "entropy": 0.5620538834482431,
      "epoch": 1.5145916354707563,
      "grad_norm": 0.15222403407096863,
      "learning_rate": 4.95169748826939e-05,
      "loss": 0.5445,
      "mean_token_accuracy": 0.8307767435908318,
      "num_tokens": 525384187.0,
      "step": 16462
    },
    {
      "entropy": 0.6782028134912252,
      "epoch": 1.5146836418421976,
      "grad_norm": 0.1667155921459198,
      "learning_rate": 4.951390805655228e-05,
      "loss": 0.6723,
      "mean_token_accuracy": 0.7967058420181274,
      "num_tokens": 525416222.0,
      "step": 16463
    },
    {
      "entropy": 0.639498533681035,
      "epoch": 1.5147756482136387,
      "grad_norm": 0.16197898983955383,
      "learning_rate": 4.951084123041065e-05,
      "loss": 0.6426,
      "mean_token_accuracy": 0.7986823059618473,
      "num_tokens": 525448375.0,
      "step": 16464
    },
    {
      "entropy": 0.6570445140823722,
      "epoch": 1.51486765458508,
      "grad_norm": 0.16215206682682037,
      "learning_rate": 4.950777440426903e-05,
      "loss": 0.6525,
      "mean_token_accuracy": 0.8037302605807781,
      "num_tokens": 525480558.0,
      "step": 16465
    },
    {
      "entropy": 0.6517907027155161,
      "epoch": 1.5149596609565212,
      "grad_norm": 0.15421505272388458,
      "learning_rate": 4.9504707578127396e-05,
      "loss": 0.6368,
      "mean_token_accuracy": 0.8029937148094177,
      "num_tokens": 525512386.0,
      "step": 16466
    },
    {
      "entropy": 0.5723423117306083,
      "epoch": 1.5150516673279624,
      "grad_norm": 0.15331323444843292,
      "learning_rate": 4.950164075198577e-05,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8244944699108601,
      "num_tokens": 525544816.0,
      "step": 16467
    },
    {
      "entropy": 0.6338437842205167,
      "epoch": 1.5151436736994037,
      "grad_norm": 0.1566825956106186,
      "learning_rate": 4.9498573925844145e-05,
      "loss": 0.6176,
      "mean_token_accuracy": 0.8110004477202892,
      "num_tokens": 525576478.0,
      "step": 16468
    },
    {
      "entropy": 0.5427188980393112,
      "epoch": 1.5152356800708449,
      "grad_norm": 0.149854376912117,
      "learning_rate": 4.949550709970252e-05,
      "loss": 0.5215,
      "mean_token_accuracy": 0.8351171612739563,
      "num_tokens": 525608359.0,
      "step": 16469
    },
    {
      "entropy": 0.6220479449257255,
      "epoch": 1.5153276864422862,
      "grad_norm": 0.16002795100212097,
      "learning_rate": 4.9492440273560895e-05,
      "loss": 0.6133,
      "mean_token_accuracy": 0.8130283914506435,
      "num_tokens": 525640851.0,
      "step": 16470
    },
    {
      "entropy": 0.6081933062523603,
      "epoch": 1.5154196928137273,
      "grad_norm": 0.1538315713405609,
      "learning_rate": 4.948937344741927e-05,
      "loss": 0.5973,
      "mean_token_accuracy": 0.81698989123106,
      "num_tokens": 525673383.0,
      "step": 16471
    },
    {
      "entropy": 0.6157899349927902,
      "epoch": 1.5155116991851685,
      "grad_norm": 0.15421636402606964,
      "learning_rate": 4.948630662127764e-05,
      "loss": 0.6082,
      "mean_token_accuracy": 0.8154948949813843,
      "num_tokens": 525705604.0,
      "step": 16472
    },
    {
      "entropy": 0.6078154128044844,
      "epoch": 1.5156037055566098,
      "grad_norm": 0.1547004133462906,
      "learning_rate": 4.948323979513602e-05,
      "loss": 0.6043,
      "mean_token_accuracy": 0.8163070604205132,
      "num_tokens": 525738146.0,
      "step": 16473
    },
    {
      "entropy": 0.6415057983249426,
      "epoch": 1.515695711928051,
      "grad_norm": 0.15960153937339783,
      "learning_rate": 4.948017296899439e-05,
      "loss": 0.6154,
      "mean_token_accuracy": 0.8079086095094681,
      "num_tokens": 525769219.0,
      "step": 16474
    },
    {
      "entropy": 0.6246890034526587,
      "epoch": 1.5157877182994923,
      "grad_norm": 0.15542006492614746,
      "learning_rate": 4.947710614285276e-05,
      "loss": 0.6119,
      "mean_token_accuracy": 0.8149034529924393,
      "num_tokens": 525801685.0,
      "step": 16475
    },
    {
      "entropy": 0.6004705000668764,
      "epoch": 1.5158797246709335,
      "grad_norm": 0.15760137140750885,
      "learning_rate": 4.947403931671114e-05,
      "loss": 0.6004,
      "mean_token_accuracy": 0.8125746473670006,
      "num_tokens": 525833891.0,
      "step": 16476
    },
    {
      "entropy": 0.6676219450309873,
      "epoch": 1.5159717310423746,
      "grad_norm": 0.16147467494010925,
      "learning_rate": 4.947097249056951e-05,
      "loss": 0.659,
      "mean_token_accuracy": 0.7982079386711121,
      "num_tokens": 525866034.0,
      "step": 16477
    },
    {
      "entropy": 0.5834331046789885,
      "epoch": 1.516063737413816,
      "grad_norm": 0.15604177117347717,
      "learning_rate": 4.9467905664427885e-05,
      "loss": 0.5658,
      "mean_token_accuracy": 0.8239332027733326,
      "num_tokens": 525898125.0,
      "step": 16478
    },
    {
      "entropy": 0.6017081514000893,
      "epoch": 1.516155743785257,
      "grad_norm": 0.15752726793289185,
      "learning_rate": 4.946483883828626e-05,
      "loss": 0.6033,
      "mean_token_accuracy": 0.817605447024107,
      "num_tokens": 525930574.0,
      "step": 16479
    },
    {
      "entropy": 0.6032543028704822,
      "epoch": 1.5162477501566984,
      "grad_norm": 0.1641794890165329,
      "learning_rate": 4.9461772012144634e-05,
      "loss": 0.594,
      "mean_token_accuracy": 0.8188711926341057,
      "num_tokens": 525962030.0,
      "step": 16480
    },
    {
      "entropy": 0.6389053678140044,
      "epoch": 1.5163397565281396,
      "grad_norm": 0.15319757163524628,
      "learning_rate": 4.945870518600301e-05,
      "loss": 0.6313,
      "mean_token_accuracy": 0.8079757280647755,
      "num_tokens": 525993885.0,
      "step": 16481
    },
    {
      "entropy": 0.6958352876827121,
      "epoch": 1.5164317628995807,
      "grad_norm": 0.16024774312973022,
      "learning_rate": 4.9455638359861384e-05,
      "loss": 0.7023,
      "mean_token_accuracy": 0.7901992723345757,
      "num_tokens": 526025835.0,
      "step": 16482
    },
    {
      "entropy": 0.5646014865487814,
      "epoch": 1.516523769271022,
      "grad_norm": 0.14759472012519836,
      "learning_rate": 4.945257153371975e-05,
      "loss": 0.5562,
      "mean_token_accuracy": 0.8324916139245033,
      "num_tokens": 526058084.0,
      "step": 16483
    },
    {
      "entropy": 0.6095221359282732,
      "epoch": 1.5166157756424632,
      "grad_norm": 0.15157414972782135,
      "learning_rate": 4.944950470757813e-05,
      "loss": 0.5962,
      "mean_token_accuracy": 0.8169955350458622,
      "num_tokens": 526090353.0,
      "step": 16484
    },
    {
      "entropy": 0.5696427589282393,
      "epoch": 1.5167077820139045,
      "grad_norm": 0.1455613374710083,
      "learning_rate": 4.94464378814365e-05,
      "loss": 0.5618,
      "mean_token_accuracy": 0.8260600753128529,
      "num_tokens": 526122445.0,
      "step": 16485
    },
    {
      "entropy": 0.6575075089931488,
      "epoch": 1.5167997883853457,
      "grad_norm": 0.16038312017917633,
      "learning_rate": 4.944337105529488e-05,
      "loss": 0.6417,
      "mean_token_accuracy": 0.8066590689122677,
      "num_tokens": 526154435.0,
      "step": 16486
    },
    {
      "entropy": 0.6031435709446669,
      "epoch": 1.5168917947567868,
      "grad_norm": 0.15057259798049927,
      "learning_rate": 4.944030422915325e-05,
      "loss": 0.5917,
      "mean_token_accuracy": 0.820949450135231,
      "num_tokens": 526186181.0,
      "step": 16487
    },
    {
      "entropy": 0.639779519289732,
      "epoch": 1.5169838011282282,
      "grad_norm": 0.15485210716724396,
      "learning_rate": 4.9437237403011625e-05,
      "loss": 0.6315,
      "mean_token_accuracy": 0.8066113628447056,
      "num_tokens": 526218782.0,
      "step": 16488
    },
    {
      "entropy": 0.6153357289731503,
      "epoch": 1.5170758074996693,
      "grad_norm": 0.1610313504934311,
      "learning_rate": 4.943417057687e-05,
      "loss": 0.592,
      "mean_token_accuracy": 0.8149732239544392,
      "num_tokens": 526250578.0,
      "step": 16489
    },
    {
      "entropy": 0.554743617773056,
      "epoch": 1.5171678138711107,
      "grad_norm": 0.1508634388446808,
      "learning_rate": 4.9431103750728374e-05,
      "loss": 0.5566,
      "mean_token_accuracy": 0.8308373503386974,
      "num_tokens": 526283035.0,
      "step": 16490
    },
    {
      "entropy": 0.5944658294320107,
      "epoch": 1.5172598202425518,
      "grad_norm": 0.15711693465709686,
      "learning_rate": 4.942803692458675e-05,
      "loss": 0.5787,
      "mean_token_accuracy": 0.8220564164221287,
      "num_tokens": 526314677.0,
      "step": 16491
    },
    {
      "entropy": 0.5870146593078971,
      "epoch": 1.517351826613993,
      "grad_norm": 0.15250276029109955,
      "learning_rate": 4.9424970098445123e-05,
      "loss": 0.5748,
      "mean_token_accuracy": 0.8259579874575138,
      "num_tokens": 526346767.0,
      "step": 16492
    },
    {
      "entropy": 0.5796989724040031,
      "epoch": 1.5174438329854343,
      "grad_norm": 0.16843488812446594,
      "learning_rate": 4.942190327230349e-05,
      "loss": 0.5688,
      "mean_token_accuracy": 0.8245409727096558,
      "num_tokens": 526378764.0,
      "step": 16493
    },
    {
      "entropy": 0.5876847933977842,
      "epoch": 1.5175358393568754,
      "grad_norm": 0.15294773876667023,
      "learning_rate": 4.941883644616187e-05,
      "loss": 0.5704,
      "mean_token_accuracy": 0.8238460868597031,
      "num_tokens": 526410205.0,
      "step": 16494
    },
    {
      "entropy": 0.6412427425384521,
      "epoch": 1.5176278457283168,
      "grad_norm": 0.1624068021774292,
      "learning_rate": 4.941576962002024e-05,
      "loss": 0.6362,
      "mean_token_accuracy": 0.7990022338926792,
      "num_tokens": 526441256.0,
      "step": 16495
    },
    {
      "entropy": 0.6499526435509324,
      "epoch": 1.517719852099758,
      "grad_norm": 0.1535618007183075,
      "learning_rate": 4.9412702793878615e-05,
      "loss": 0.6341,
      "mean_token_accuracy": 0.8071523793041706,
      "num_tokens": 526473516.0,
      "step": 16496
    },
    {
      "entropy": 0.668819684535265,
      "epoch": 1.517811858471199,
      "grad_norm": 0.1597776859998703,
      "learning_rate": 4.940963596773699e-05,
      "loss": 0.6659,
      "mean_token_accuracy": 0.7940640263259411,
      "num_tokens": 526505358.0,
      "step": 16497
    },
    {
      "entropy": 0.6549066565930843,
      "epoch": 1.5179038648426404,
      "grad_norm": 0.1545163094997406,
      "learning_rate": 4.9406569141595365e-05,
      "loss": 0.6381,
      "mean_token_accuracy": 0.8050271198153496,
      "num_tokens": 526537216.0,
      "step": 16498
    },
    {
      "entropy": 0.6294102212414145,
      "epoch": 1.5179958712140815,
      "grad_norm": 0.15729683637619019,
      "learning_rate": 4.940350231545374e-05,
      "loss": 0.6172,
      "mean_token_accuracy": 0.8100723884999752,
      "num_tokens": 526569627.0,
      "step": 16499
    },
    {
      "entropy": 0.6597276590764523,
      "epoch": 1.5180878775855229,
      "grad_norm": 0.16044820845127106,
      "learning_rate": 4.9400435489312114e-05,
      "loss": 0.6487,
      "mean_token_accuracy": 0.7988115325570107,
      "num_tokens": 526601513.0,
      "step": 16500
    },
    {
      "entropy": 0.5734455659985542,
      "epoch": 1.518179883956964,
      "grad_norm": 0.15273302793502808,
      "learning_rate": 4.939736866317048e-05,
      "loss": 0.5617,
      "mean_token_accuracy": 0.8258678093552589,
      "num_tokens": 526633189.0,
      "step": 16501
    },
    {
      "entropy": 0.6405914947390556,
      "epoch": 1.5182718903284051,
      "grad_norm": 0.15933077037334442,
      "learning_rate": 4.939430183702886e-05,
      "loss": 0.6243,
      "mean_token_accuracy": 0.8056851550936699,
      "num_tokens": 526665691.0,
      "step": 16502
    },
    {
      "entropy": 0.5916606062091887,
      "epoch": 1.5183638966998465,
      "grad_norm": 0.15563228726387024,
      "learning_rate": 4.939123501088724e-05,
      "loss": 0.5799,
      "mean_token_accuracy": 0.8214592151343822,
      "num_tokens": 526696878.0,
      "step": 16503
    },
    {
      "entropy": 0.6873699352145195,
      "epoch": 1.5184559030712876,
      "grad_norm": 0.16203981637954712,
      "learning_rate": 4.9388168184745606e-05,
      "loss": 0.6832,
      "mean_token_accuracy": 0.7922581359744072,
      "num_tokens": 526728060.0,
      "step": 16504
    },
    {
      "entropy": 0.6111687030643225,
      "epoch": 1.518547909442729,
      "grad_norm": 0.1542670726776123,
      "learning_rate": 4.938510135860399e-05,
      "loss": 0.6059,
      "mean_token_accuracy": 0.8175611644983292,
      "num_tokens": 526760144.0,
      "step": 16505
    },
    {
      "entropy": 0.5933074299246073,
      "epoch": 1.5186399158141701,
      "grad_norm": 0.15630070865154266,
      "learning_rate": 4.9382034532462355e-05,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8170476853847504,
      "num_tokens": 526792574.0,
      "step": 16506
    },
    {
      "entropy": 0.5767204165458679,
      "epoch": 1.5187319221856113,
      "grad_norm": 0.16168953478336334,
      "learning_rate": 4.9378967706320736e-05,
      "loss": 0.5597,
      "mean_token_accuracy": 0.8281556181609631,
      "num_tokens": 526824682.0,
      "step": 16507
    },
    {
      "entropy": 0.6288373824208975,
      "epoch": 1.5188239285570526,
      "grad_norm": 0.15820500254631042,
      "learning_rate": 4.9375900880179104e-05,
      "loss": 0.6318,
      "mean_token_accuracy": 0.8097222074866295,
      "num_tokens": 526856870.0,
      "step": 16508
    },
    {
      "entropy": 0.5519189797341824,
      "epoch": 1.5189159349284937,
      "grad_norm": 0.14823651313781738,
      "learning_rate": 4.937283405403748e-05,
      "loss": 0.5424,
      "mean_token_accuracy": 0.8297552205622196,
      "num_tokens": 526889602.0,
      "step": 16509
    },
    {
      "entropy": 0.6102744620293379,
      "epoch": 1.519007941299935,
      "grad_norm": 0.15959563851356506,
      "learning_rate": 4.9369767227895854e-05,
      "loss": 0.6095,
      "mean_token_accuracy": 0.8115913383662701,
      "num_tokens": 526921356.0,
      "step": 16510
    },
    {
      "entropy": 0.5638866927474737,
      "epoch": 1.5190999476713762,
      "grad_norm": 0.1516043096780777,
      "learning_rate": 4.936670040175423e-05,
      "loss": 0.5522,
      "mean_token_accuracy": 0.8300323970615864,
      "num_tokens": 526953162.0,
      "step": 16511
    },
    {
      "entropy": 0.6111222207546234,
      "epoch": 1.5191919540428174,
      "grad_norm": 0.16150639951229095,
      "learning_rate": 4.93636335756126e-05,
      "loss": 0.6079,
      "mean_token_accuracy": 0.8178420290350914,
      "num_tokens": 526985340.0,
      "step": 16512
    },
    {
      "entropy": 0.6490694154053926,
      "epoch": 1.5192839604142587,
      "grad_norm": 0.15604516863822937,
      "learning_rate": 4.936056674947098e-05,
      "loss": 0.6309,
      "mean_token_accuracy": 0.8071588352322578,
      "num_tokens": 527016919.0,
      "step": 16513
    },
    {
      "entropy": 0.6364948842674494,
      "epoch": 1.5193759667856999,
      "grad_norm": 0.15791484713554382,
      "learning_rate": 4.9357499923329345e-05,
      "loss": 0.6099,
      "mean_token_accuracy": 0.8159605413675308,
      "num_tokens": 527048454.0,
      "step": 16514
    },
    {
      "entropy": 0.6354039087891579,
      "epoch": 1.5194679731571412,
      "grad_norm": 0.1551956981420517,
      "learning_rate": 4.935443309718773e-05,
      "loss": 0.6296,
      "mean_token_accuracy": 0.806107223033905,
      "num_tokens": 527080847.0,
      "step": 16515
    },
    {
      "entropy": 0.5803968235850334,
      "epoch": 1.5195599795285823,
      "grad_norm": 0.15678228437900543,
      "learning_rate": 4.9351366271046095e-05,
      "loss": 0.5665,
      "mean_token_accuracy": 0.8279697448015213,
      "num_tokens": 527112488.0,
      "step": 16516
    },
    {
      "entropy": 0.6011387561447918,
      "epoch": 1.5196519859000235,
      "grad_norm": 0.14999113976955414,
      "learning_rate": 4.934829944490447e-05,
      "loss": 0.5815,
      "mean_token_accuracy": 0.8210340104997158,
      "num_tokens": 527144773.0,
      "step": 16517
    },
    {
      "entropy": 0.5905690779909492,
      "epoch": 1.5197439922714648,
      "grad_norm": 0.15185026824474335,
      "learning_rate": 4.9345232618762844e-05,
      "loss": 0.5734,
      "mean_token_accuracy": 0.8214596100151539,
      "num_tokens": 527176704.0,
      "step": 16518
    },
    {
      "entropy": 0.6608882676810026,
      "epoch": 1.519835998642906,
      "grad_norm": 0.15917174518108368,
      "learning_rate": 4.934216579262122e-05,
      "loss": 0.6557,
      "mean_token_accuracy": 0.7969521172344685,
      "num_tokens": 527208985.0,
      "step": 16519
    },
    {
      "entropy": 0.5636718245223165,
      "epoch": 1.5199280050143473,
      "grad_norm": 0.15091224014759064,
      "learning_rate": 4.933909896647959e-05,
      "loss": 0.5567,
      "mean_token_accuracy": 0.8283116854727268,
      "num_tokens": 527241190.0,
      "step": 16520
    },
    {
      "entropy": 0.7258144728839397,
      "epoch": 1.5200200113857885,
      "grad_norm": 0.16192159056663513,
      "learning_rate": 4.933603214033797e-05,
      "loss": 0.7149,
      "mean_token_accuracy": 0.7816574722528458,
      "num_tokens": 527272925.0,
      "step": 16521
    },
    {
      "entropy": 0.590102399000898,
      "epoch": 1.5201120177572296,
      "grad_norm": 0.15885156393051147,
      "learning_rate": 4.9332965314196336e-05,
      "loss": 0.5748,
      "mean_token_accuracy": 0.8254778981208801,
      "num_tokens": 527304847.0,
      "step": 16522
    },
    {
      "entropy": 0.6517719505354762,
      "epoch": 1.520204024128671,
      "grad_norm": 0.1566925346851349,
      "learning_rate": 4.932989848805472e-05,
      "loss": 0.6423,
      "mean_token_accuracy": 0.8048259131610394,
      "num_tokens": 527337117.0,
      "step": 16523
    },
    {
      "entropy": 0.6827370543032885,
      "epoch": 1.520296030500112,
      "grad_norm": 0.1584622710943222,
      "learning_rate": 4.9326831661913085e-05,
      "loss": 0.6691,
      "mean_token_accuracy": 0.7924227491021156,
      "num_tokens": 527368428.0,
      "step": 16524
    },
    {
      "entropy": 0.5113977817818522,
      "epoch": 1.5203880368715534,
      "grad_norm": 0.15039043128490448,
      "learning_rate": 4.9323764835771466e-05,
      "loss": 0.5093,
      "mean_token_accuracy": 0.8450251668691635,
      "num_tokens": 527401087.0,
      "step": 16525
    },
    {
      "entropy": 0.5791545603424311,
      "epoch": 1.5204800432429946,
      "grad_norm": 0.15915343165397644,
      "learning_rate": 4.9320698009629834e-05,
      "loss": 0.579,
      "mean_token_accuracy": 0.8227903284132481,
      "num_tokens": 527432516.0,
      "step": 16526
    },
    {
      "entropy": 0.519720271229744,
      "epoch": 1.5205720496144357,
      "grad_norm": 0.14825798571109772,
      "learning_rate": 4.931763118348821e-05,
      "loss": 0.511,
      "mean_token_accuracy": 0.8395257070660591,
      "num_tokens": 527464328.0,
      "step": 16527
    },
    {
      "entropy": 0.602475443854928,
      "epoch": 1.520664055985877,
      "grad_norm": 0.15442149341106415,
      "learning_rate": 4.9314564357346584e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.815783329308033,
      "num_tokens": 527496464.0,
      "step": 16528
    },
    {
      "entropy": 0.5320751229301095,
      "epoch": 1.5207560623573182,
      "grad_norm": 0.1525323987007141,
      "learning_rate": 4.931149753120496e-05,
      "loss": 0.5083,
      "mean_token_accuracy": 0.8422673754394054,
      "num_tokens": 527528645.0,
      "step": 16529
    },
    {
      "entropy": 0.6703555407002568,
      "epoch": 1.5208480687287596,
      "grad_norm": 0.17915183305740356,
      "learning_rate": 4.930843070506333e-05,
      "loss": 0.6485,
      "mean_token_accuracy": 0.8006265982985497,
      "num_tokens": 527560341.0,
      "step": 16530
    },
    {
      "entropy": 0.6943517131730914,
      "epoch": 1.5209400751002007,
      "grad_norm": 0.16760139167308807,
      "learning_rate": 4.930536387892171e-05,
      "loss": 0.6865,
      "mean_token_accuracy": 0.7878704071044922,
      "num_tokens": 527591749.0,
      "step": 16531
    },
    {
      "entropy": 0.6104346653446555,
      "epoch": 1.5210320814716418,
      "grad_norm": 0.15855997800827026,
      "learning_rate": 4.9302297052780075e-05,
      "loss": 0.6066,
      "mean_token_accuracy": 0.8122986294329166,
      "num_tokens": 527623075.0,
      "step": 16532
    },
    {
      "entropy": 0.6118032787926495,
      "epoch": 1.5211240878430832,
      "grad_norm": 0.16004154086112976,
      "learning_rate": 4.929923022663846e-05,
      "loss": 0.5885,
      "mean_token_accuracy": 0.8145452216267586,
      "num_tokens": 527654649.0,
      "step": 16533
    },
    {
      "entropy": 0.5914721470326185,
      "epoch": 1.5212160942145243,
      "grad_norm": 0.15794338285923004,
      "learning_rate": 4.929616340049683e-05,
      "loss": 0.5706,
      "mean_token_accuracy": 0.8253653571009636,
      "num_tokens": 527686762.0,
      "step": 16534
    },
    {
      "entropy": 0.6896715462207794,
      "epoch": 1.5213081005859657,
      "grad_norm": 0.1615317165851593,
      "learning_rate": 4.92930965743552e-05,
      "loss": 0.6781,
      "mean_token_accuracy": 0.7948140278458595,
      "num_tokens": 527718131.0,
      "step": 16535
    },
    {
      "entropy": 0.6503585055470467,
      "epoch": 1.5214001069574068,
      "grad_norm": 0.15542463958263397,
      "learning_rate": 4.929002974821358e-05,
      "loss": 0.632,
      "mean_token_accuracy": 0.8066948093473911,
      "num_tokens": 527750318.0,
      "step": 16536
    },
    {
      "entropy": 0.6306429039686918,
      "epoch": 1.521492113328848,
      "grad_norm": 0.15433400869369507,
      "learning_rate": 4.928696292207195e-05,
      "loss": 0.6271,
      "mean_token_accuracy": 0.81059854850173,
      "num_tokens": 527782563.0,
      "step": 16537
    },
    {
      "entropy": 0.6573727671056986,
      "epoch": 1.5215841197002893,
      "grad_norm": 0.15777279436588287,
      "learning_rate": 4.928389609593032e-05,
      "loss": 0.6601,
      "mean_token_accuracy": 0.8001940585672855,
      "num_tokens": 527814863.0,
      "step": 16538
    },
    {
      "entropy": 0.6637552976608276,
      "epoch": 1.5216761260717304,
      "grad_norm": 0.15510553121566772,
      "learning_rate": 4.92808292697887e-05,
      "loss": 0.6471,
      "mean_token_accuracy": 0.7981804050505161,
      "num_tokens": 527847056.0,
      "step": 16539
    },
    {
      "entropy": 0.6122811138629913,
      "epoch": 1.5217681324431718,
      "grad_norm": 0.16687612235546112,
      "learning_rate": 4.927776244364707e-05,
      "loss": 0.6111,
      "mean_token_accuracy": 0.8121418245136738,
      "num_tokens": 527879461.0,
      "step": 16540
    },
    {
      "entropy": 0.6998784057796001,
      "epoch": 1.521860138814613,
      "grad_norm": 0.16340163350105286,
      "learning_rate": 4.927469561750545e-05,
      "loss": 0.6887,
      "mean_token_accuracy": 0.7905695289373398,
      "num_tokens": 527910800.0,
      "step": 16541
    },
    {
      "entropy": 0.6108220871537924,
      "epoch": 1.521952145186054,
      "grad_norm": 0.16150616109371185,
      "learning_rate": 4.927162879136382e-05,
      "loss": 0.6117,
      "mean_token_accuracy": 0.8108691312372684,
      "num_tokens": 527942796.0,
      "step": 16542
    },
    {
      "entropy": 0.5878836251795292,
      "epoch": 1.5220441515574954,
      "grad_norm": 0.15422342717647552,
      "learning_rate": 4.926856196522219e-05,
      "loss": 0.5907,
      "mean_token_accuracy": 0.8189480639994144,
      "num_tokens": 527974818.0,
      "step": 16543
    },
    {
      "entropy": 0.6859790831804276,
      "epoch": 1.5221361579289365,
      "grad_norm": 0.15682680904865265,
      "learning_rate": 4.926549513908057e-05,
      "loss": 0.6784,
      "mean_token_accuracy": 0.7907069995999336,
      "num_tokens": 528007119.0,
      "step": 16544
    },
    {
      "entropy": 0.6116712703369558,
      "epoch": 1.5222281643003779,
      "grad_norm": 0.15353785455226898,
      "learning_rate": 4.926242831293894e-05,
      "loss": 0.595,
      "mean_token_accuracy": 0.8167304880917072,
      "num_tokens": 528039157.0,
      "step": 16545
    },
    {
      "entropy": 0.6319145746529102,
      "epoch": 1.522320170671819,
      "grad_norm": 0.15747153759002686,
      "learning_rate": 4.925936148679732e-05,
      "loss": 0.615,
      "mean_token_accuracy": 0.8133190609514713,
      "num_tokens": 528071396.0,
      "step": 16546
    },
    {
      "entropy": 0.6290771933272481,
      "epoch": 1.5224121770432602,
      "grad_norm": 0.15793509781360626,
      "learning_rate": 4.925629466065569e-05,
      "loss": 0.606,
      "mean_token_accuracy": 0.8151703663170338,
      "num_tokens": 528102758.0,
      "step": 16547
    },
    {
      "entropy": 0.6603141278028488,
      "epoch": 1.5225041834147015,
      "grad_norm": 0.1583685427904129,
      "learning_rate": 4.925322783451406e-05,
      "loss": 0.6521,
      "mean_token_accuracy": 0.8015799857676029,
      "num_tokens": 528134471.0,
      "step": 16548
    },
    {
      "entropy": 0.588864678516984,
      "epoch": 1.5225961897861426,
      "grad_norm": 0.15699289739131927,
      "learning_rate": 4.925016100837244e-05,
      "loss": 0.5791,
      "mean_token_accuracy": 0.8237920664250851,
      "num_tokens": 528166497.0,
      "step": 16549
    },
    {
      "entropy": 0.5526434995699674,
      "epoch": 1.522688196157584,
      "grad_norm": 0.15717977285385132,
      "learning_rate": 4.924709418223081e-05,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8314349316060543,
      "num_tokens": 528198127.0,
      "step": 16550
    },
    {
      "entropy": 0.606762109324336,
      "epoch": 1.5227802025290251,
      "grad_norm": 0.15662892162799835,
      "learning_rate": 4.924402735608919e-05,
      "loss": 0.5994,
      "mean_token_accuracy": 0.8177457451820374,
      "num_tokens": 528230238.0,
      "step": 16551
    },
    {
      "entropy": 0.6239919355139136,
      "epoch": 1.5228722089004663,
      "grad_norm": 0.16194488108158112,
      "learning_rate": 4.924096052994756e-05,
      "loss": 0.6339,
      "mean_token_accuracy": 0.8072803989052773,
      "num_tokens": 528262340.0,
      "step": 16552
    },
    {
      "entropy": 0.7145175114274025,
      "epoch": 1.5229642152719076,
      "grad_norm": 0.1660831719636917,
      "learning_rate": 4.923789370380593e-05,
      "loss": 0.7075,
      "mean_token_accuracy": 0.7821329906582832,
      "num_tokens": 528294295.0,
      "step": 16553
    },
    {
      "entropy": 0.5717959078028798,
      "epoch": 1.5230562216433488,
      "grad_norm": 0.15196159482002258,
      "learning_rate": 4.923482687766431e-05,
      "loss": 0.561,
      "mean_token_accuracy": 0.8271680437028408,
      "num_tokens": 528326423.0,
      "step": 16554
    },
    {
      "entropy": 0.524331412743777,
      "epoch": 1.52314822801479,
      "grad_norm": 0.14516593515872955,
      "learning_rate": 4.923176005152268e-05,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8443445339798927,
      "num_tokens": 528358739.0,
      "step": 16555
    },
    {
      "entropy": 0.6307591330260038,
      "epoch": 1.5232402343862312,
      "grad_norm": 0.16015437245368958,
      "learning_rate": 4.9228693225381053e-05,
      "loss": 0.6273,
      "mean_token_accuracy": 0.8101574704051018,
      "num_tokens": 528391175.0,
      "step": 16556
    },
    {
      "entropy": 0.5772523740306497,
      "epoch": 1.5233322407576724,
      "grad_norm": 0.15431365370750427,
      "learning_rate": 4.922562639923943e-05,
      "loss": 0.5625,
      "mean_token_accuracy": 0.8281646445393562,
      "num_tokens": 528423337.0,
      "step": 16557
    },
    {
      "entropy": 0.6808778271079063,
      "epoch": 1.5234242471291137,
      "grad_norm": 0.17241564393043518,
      "learning_rate": 4.92225595730978e-05,
      "loss": 0.6698,
      "mean_token_accuracy": 0.7968861535191536,
      "num_tokens": 528455221.0,
      "step": 16558
    },
    {
      "entropy": 0.5450560487806797,
      "epoch": 1.5235162535005549,
      "grad_norm": 0.15030166506767273,
      "learning_rate": 4.921949274695618e-05,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8405566476285458,
      "num_tokens": 528487484.0,
      "step": 16559
    },
    {
      "entropy": 0.6865616403520107,
      "epoch": 1.5236082598719962,
      "grad_norm": 0.16116127371788025,
      "learning_rate": 4.921642592081455e-05,
      "loss": 0.6643,
      "mean_token_accuracy": 0.796602800488472,
      "num_tokens": 528518917.0,
      "step": 16560
    },
    {
      "entropy": 0.5408937442116439,
      "epoch": 1.5237002662434374,
      "grad_norm": 0.15165290236473083,
      "learning_rate": 4.921335909467292e-05,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8389518596231937,
      "num_tokens": 528550229.0,
      "step": 16561
    },
    {
      "entropy": 0.644461995922029,
      "epoch": 1.5237922726148785,
      "grad_norm": 0.14969828724861145,
      "learning_rate": 4.92102922685313e-05,
      "loss": 0.6377,
      "mean_token_accuracy": 0.7995346933603287,
      "num_tokens": 528582633.0,
      "step": 16562
    },
    {
      "entropy": 0.5679704286158085,
      "epoch": 1.5238842789863198,
      "grad_norm": 0.15901388227939606,
      "learning_rate": 4.920722544238967e-05,
      "loss": 0.5629,
      "mean_token_accuracy": 0.8283427208662033,
      "num_tokens": 528614373.0,
      "step": 16563
    },
    {
      "entropy": 0.5892972191795707,
      "epoch": 1.523976285357761,
      "grad_norm": 0.15560925006866455,
      "learning_rate": 4.9204158616248044e-05,
      "loss": 0.581,
      "mean_token_accuracy": 0.8195765875279903,
      "num_tokens": 528645963.0,
      "step": 16564
    },
    {
      "entropy": 0.5624236464500427,
      "epoch": 1.5240682917292023,
      "grad_norm": 0.15102098882198334,
      "learning_rate": 4.9201091790106425e-05,
      "loss": 0.5551,
      "mean_token_accuracy": 0.8298157230019569,
      "num_tokens": 528678173.0,
      "step": 16565
    },
    {
      "entropy": 0.612639999948442,
      "epoch": 1.5241602981006435,
      "grad_norm": 0.15589025616645813,
      "learning_rate": 4.919802496396479e-05,
      "loss": 0.6015,
      "mean_token_accuracy": 0.8153745718300343,
      "num_tokens": 528710494.0,
      "step": 16566
    },
    {
      "entropy": 0.6232336442917585,
      "epoch": 1.5242523044720846,
      "grad_norm": 0.15908372402191162,
      "learning_rate": 4.9194958137823175e-05,
      "loss": 0.6233,
      "mean_token_accuracy": 0.8082961142063141,
      "num_tokens": 528742392.0,
      "step": 16567
    },
    {
      "entropy": 0.6169014275074005,
      "epoch": 1.524344310843526,
      "grad_norm": 0.15811142325401306,
      "learning_rate": 4.919189131168154e-05,
      "loss": 0.6065,
      "mean_token_accuracy": 0.8173240981996059,
      "num_tokens": 528774949.0,
      "step": 16568
    },
    {
      "entropy": 0.526100117713213,
      "epoch": 1.524436317214967,
      "grad_norm": 0.16424019634723663,
      "learning_rate": 4.918882448553992e-05,
      "loss": 0.5109,
      "mean_token_accuracy": 0.8400375694036484,
      "num_tokens": 528806660.0,
      "step": 16569
    },
    {
      "entropy": 0.5294187162071466,
      "epoch": 1.5245283235864084,
      "grad_norm": 0.1463291198015213,
      "learning_rate": 4.918575765939829e-05,
      "loss": 0.5177,
      "mean_token_accuracy": 0.8383526429533958,
      "num_tokens": 528839310.0,
      "step": 16570
    },
    {
      "entropy": 0.6393635347485542,
      "epoch": 1.5246203299578496,
      "grad_norm": 0.1526903510093689,
      "learning_rate": 4.9182690833256666e-05,
      "loss": 0.6245,
      "mean_token_accuracy": 0.8077404014766216,
      "num_tokens": 528871434.0,
      "step": 16571
    },
    {
      "entropy": 0.6031456300988793,
      "epoch": 1.5247123363292907,
      "grad_norm": 0.15266330540180206,
      "learning_rate": 4.917962400711504e-05,
      "loss": 0.589,
      "mean_token_accuracy": 0.8175104185938835,
      "num_tokens": 528903746.0,
      "step": 16572
    },
    {
      "entropy": 0.5792659316211939,
      "epoch": 1.524804342700732,
      "grad_norm": 0.15548986196517944,
      "learning_rate": 4.9176557180973416e-05,
      "loss": 0.5624,
      "mean_token_accuracy": 0.8290232084691525,
      "num_tokens": 528935230.0,
      "step": 16573
    },
    {
      "entropy": 0.6898874957114458,
      "epoch": 1.5248963490721732,
      "grad_norm": 0.16112476587295532,
      "learning_rate": 4.9173490354831784e-05,
      "loss": 0.6753,
      "mean_token_accuracy": 0.7951638847589493,
      "num_tokens": 528967379.0,
      "step": 16574
    },
    {
      "entropy": 0.6889080181717873,
      "epoch": 1.5249883554436146,
      "grad_norm": 0.15681138634681702,
      "learning_rate": 4.9170423528690165e-05,
      "loss": 0.667,
      "mean_token_accuracy": 0.7975130975246429,
      "num_tokens": 528999404.0,
      "step": 16575
    },
    {
      "entropy": 0.5869499207474291,
      "epoch": 1.5250803618150557,
      "grad_norm": 0.14922139048576355,
      "learning_rate": 4.916735670254853e-05,
      "loss": 0.5622,
      "mean_token_accuracy": 0.8262416310608387,
      "num_tokens": 529030824.0,
      "step": 16576
    },
    {
      "entropy": 0.5506965280510485,
      "epoch": 1.5251723681864968,
      "grad_norm": 0.15112529695034027,
      "learning_rate": 4.916428987640691e-05,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8355726525187492,
      "num_tokens": 529062783.0,
      "step": 16577
    },
    {
      "entropy": 0.6901394687592983,
      "epoch": 1.5252643745579382,
      "grad_norm": 0.16617204248905182,
      "learning_rate": 4.916122305026528e-05,
      "loss": 0.6917,
      "mean_token_accuracy": 0.7934467270970345,
      "num_tokens": 529093587.0,
      "step": 16578
    },
    {
      "entropy": 0.6473276736214757,
      "epoch": 1.5253563809293793,
      "grad_norm": 0.15974482893943787,
      "learning_rate": 4.915815622412366e-05,
      "loss": 0.6356,
      "mean_token_accuracy": 0.807346198707819,
      "num_tokens": 529125007.0,
      "step": 16579
    },
    {
      "entropy": 0.4659056290984154,
      "epoch": 1.5254483873008207,
      "grad_norm": 0.1578729748725891,
      "learning_rate": 4.915508939798203e-05,
      "loss": 0.4623,
      "mean_token_accuracy": 0.8563388884067535,
      "num_tokens": 529156838.0,
      "step": 16580
    },
    {
      "entropy": 0.5947121786884964,
      "epoch": 1.5255403936722618,
      "grad_norm": 0.1543368697166443,
      "learning_rate": 4.9152022571840406e-05,
      "loss": 0.5818,
      "mean_token_accuracy": 0.8212741687893867,
      "num_tokens": 529188454.0,
      "step": 16581
    },
    {
      "entropy": 0.5673114461824298,
      "epoch": 1.525632400043703,
      "grad_norm": 0.15198026597499847,
      "learning_rate": 4.9148955745698774e-05,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8263893611729145,
      "num_tokens": 529220684.0,
      "step": 16582
    },
    {
      "entropy": 0.5929160974919796,
      "epoch": 1.5257244064151443,
      "grad_norm": 0.1609189361333847,
      "learning_rate": 4.9145888919557155e-05,
      "loss": 0.5897,
      "mean_token_accuracy": 0.8157070502638817,
      "num_tokens": 529251993.0,
      "step": 16583
    },
    {
      "entropy": 0.5091171823441982,
      "epoch": 1.5258164127865854,
      "grad_norm": 0.14754609763622284,
      "learning_rate": 4.914282209341552e-05,
      "loss": 0.4981,
      "mean_token_accuracy": 0.8479176387190819,
      "num_tokens": 529284452.0,
      "step": 16584
    },
    {
      "entropy": 0.6757408604025841,
      "epoch": 1.5259084191580268,
      "grad_norm": 0.15858076512813568,
      "learning_rate": 4.9139755267273905e-05,
      "loss": 0.6774,
      "mean_token_accuracy": 0.7941181845963001,
      "num_tokens": 529316627.0,
      "step": 16585
    },
    {
      "entropy": 0.5404332033358514,
      "epoch": 1.526000425529468,
      "grad_norm": 0.1495986133813858,
      "learning_rate": 4.913668844113227e-05,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8422566838562489,
      "num_tokens": 529348322.0,
      "step": 16586
    },
    {
      "entropy": 0.5095608532428741,
      "epoch": 1.526092431900909,
      "grad_norm": 0.1473742425441742,
      "learning_rate": 4.913362161499065e-05,
      "loss": 0.4987,
      "mean_token_accuracy": 0.8432375974953175,
      "num_tokens": 529380259.0,
      "step": 16587
    },
    {
      "entropy": 0.45354621787555516,
      "epoch": 1.5261844382723504,
      "grad_norm": 0.14529697597026825,
      "learning_rate": 4.913055478884902e-05,
      "loss": 0.4478,
      "mean_token_accuracy": 0.8640005961060524,
      "num_tokens": 529413025.0,
      "step": 16588
    },
    {
      "entropy": 0.5412250645458698,
      "epoch": 1.5262764446437915,
      "grad_norm": 0.1614299863576889,
      "learning_rate": 4.9127487962707397e-05,
      "loss": 0.545,
      "mean_token_accuracy": 0.8334466964006424,
      "num_tokens": 529443685.0,
      "step": 16589
    },
    {
      "entropy": 0.5903942380100489,
      "epoch": 1.5263684510152329,
      "grad_norm": 0.1568058282136917,
      "learning_rate": 4.912442113656577e-05,
      "loss": 0.577,
      "mean_token_accuracy": 0.8241964988410473,
      "num_tokens": 529474622.0,
      "step": 16590
    },
    {
      "entropy": 0.6595178376883268,
      "epoch": 1.526460457386674,
      "grad_norm": 0.1597389578819275,
      "learning_rate": 4.9121354310424146e-05,
      "loss": 0.6588,
      "mean_token_accuracy": 0.8016118034720421,
      "num_tokens": 529506336.0,
      "step": 16591
    },
    {
      "entropy": 0.7054630909115076,
      "epoch": 1.5265524637581152,
      "grad_norm": 0.16514761745929718,
      "learning_rate": 4.9118287484282514e-05,
      "loss": 0.7018,
      "mean_token_accuracy": 0.7915130108594894,
      "num_tokens": 529538541.0,
      "step": 16592
    },
    {
      "entropy": 0.5638193534687161,
      "epoch": 1.5266444701295565,
      "grad_norm": 0.14898581802845,
      "learning_rate": 4.9115220658140895e-05,
      "loss": 0.5444,
      "mean_token_accuracy": 0.827970627695322,
      "num_tokens": 529570891.0,
      "step": 16593
    },
    {
      "entropy": 0.6225284785032272,
      "epoch": 1.5267364765009976,
      "grad_norm": 0.16008099913597107,
      "learning_rate": 4.911215383199926e-05,
      "loss": 0.606,
      "mean_token_accuracy": 0.8100022375583649,
      "num_tokens": 529602213.0,
      "step": 16594
    },
    {
      "entropy": 0.5306655284948647,
      "epoch": 1.526828482872439,
      "grad_norm": 0.15633714199066162,
      "learning_rate": 4.910908700585764e-05,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8388569131493568,
      "num_tokens": 529633989.0,
      "step": 16595
    },
    {
      "entropy": 0.603593435138464,
      "epoch": 1.5269204892438801,
      "grad_norm": 0.15270943939685822,
      "learning_rate": 4.910602017971602e-05,
      "loss": 0.6063,
      "mean_token_accuracy": 0.8170741610229015,
      "num_tokens": 529666110.0,
      "step": 16596
    },
    {
      "entropy": 0.6868840977549553,
      "epoch": 1.5270124956153213,
      "grad_norm": 0.16824166476726532,
      "learning_rate": 4.910295335357439e-05,
      "loss": 0.6922,
      "mean_token_accuracy": 0.7938742414116859,
      "num_tokens": 529698038.0,
      "step": 16597
    },
    {
      "entropy": 0.5856648944318295,
      "epoch": 1.5271045019867626,
      "grad_norm": 0.15261483192443848,
      "learning_rate": 4.909988652743276e-05,
      "loss": 0.5737,
      "mean_token_accuracy": 0.8236256800591946,
      "num_tokens": 529729828.0,
      "step": 16598
    },
    {
      "entropy": 0.644649799913168,
      "epoch": 1.5271965083582038,
      "grad_norm": 0.15888775885105133,
      "learning_rate": 4.9096819701291136e-05,
      "loss": 0.6211,
      "mean_token_accuracy": 0.8073699735105038,
      "num_tokens": 529761417.0,
      "step": 16599
    },
    {
      "entropy": 0.5689111491665244,
      "epoch": 1.527288514729645,
      "grad_norm": 0.14542926847934723,
      "learning_rate": 4.909375287514951e-05,
      "loss": 0.5472,
      "mean_token_accuracy": 0.8312892466783524,
      "num_tokens": 529793640.0,
      "step": 16600
    },
    {
      "entropy": 0.5176237700507045,
      "epoch": 1.5273805211010862,
      "grad_norm": 0.14641934633255005,
      "learning_rate": 4.9090686049007886e-05,
      "loss": 0.5013,
      "mean_token_accuracy": 0.8357004597783089,
      "num_tokens": 529824560.0,
      "step": 16601
    },
    {
      "entropy": 0.644124430604279,
      "epoch": 1.5274725274725274,
      "grad_norm": 0.15319009125232697,
      "learning_rate": 4.908761922286626e-05,
      "loss": 0.627,
      "mean_token_accuracy": 0.8102684356272221,
      "num_tokens": 529855412.0,
      "step": 16602
    },
    {
      "entropy": 0.7319613313302398,
      "epoch": 1.5275645338439687,
      "grad_norm": 0.16399610042572021,
      "learning_rate": 4.908455239672463e-05,
      "loss": 0.7188,
      "mean_token_accuracy": 0.7779453359544277,
      "num_tokens": 529887686.0,
      "step": 16603
    },
    {
      "entropy": 0.62395397759974,
      "epoch": 1.5276565402154099,
      "grad_norm": 0.16541001200675964,
      "learning_rate": 4.908148557058301e-05,
      "loss": 0.6192,
      "mean_token_accuracy": 0.8100747317075729,
      "num_tokens": 529919059.0,
      "step": 16604
    },
    {
      "entropy": 0.5781202497892082,
      "epoch": 1.5277485465868512,
      "grad_norm": 0.1536174714565277,
      "learning_rate": 4.907841874444138e-05,
      "loss": 0.5653,
      "mean_token_accuracy": 0.8220346979796886,
      "num_tokens": 529950750.0,
      "step": 16605
    },
    {
      "entropy": 0.5716502359136939,
      "epoch": 1.5278405529582924,
      "grad_norm": 0.15323975682258606,
      "learning_rate": 4.907535191829976e-05,
      "loss": 0.5746,
      "mean_token_accuracy": 0.8226268775761127,
      "num_tokens": 529982428.0,
      "step": 16606
    },
    {
      "entropy": 0.6897704391740263,
      "epoch": 1.5279325593297335,
      "grad_norm": 0.15610261261463165,
      "learning_rate": 4.9072285092158127e-05,
      "loss": 0.6828,
      "mean_token_accuracy": 0.7934139929711819,
      "num_tokens": 530014315.0,
      "step": 16607
    },
    {
      "entropy": 0.6152018308639526,
      "epoch": 1.5280245657011748,
      "grad_norm": 0.15403933823108673,
      "learning_rate": 4.90692182660165e-05,
      "loss": 0.6103,
      "mean_token_accuracy": 0.8147405944764614,
      "num_tokens": 530046892.0,
      "step": 16608
    },
    {
      "entropy": 0.5324181290343404,
      "epoch": 1.528116572072616,
      "grad_norm": 0.15273317694664001,
      "learning_rate": 4.9066151439874876e-05,
      "loss": 0.5144,
      "mean_token_accuracy": 0.8394965305924416,
      "num_tokens": 530078475.0,
      "step": 16609
    },
    {
      "entropy": 0.5057840719819069,
      "epoch": 1.5282085784440573,
      "grad_norm": 0.15578241646289825,
      "learning_rate": 4.906308461373325e-05,
      "loss": 0.4946,
      "mean_token_accuracy": 0.846567265689373,
      "num_tokens": 530110707.0,
      "step": 16610
    },
    {
      "entropy": 0.5856322487816215,
      "epoch": 1.5283005848154985,
      "grad_norm": 0.14726465940475464,
      "learning_rate": 4.9060017787591625e-05,
      "loss": 0.569,
      "mean_token_accuracy": 0.8261259160935879,
      "num_tokens": 530142964.0,
      "step": 16611
    },
    {
      "entropy": 0.6255930550396442,
      "epoch": 1.5283925911869396,
      "grad_norm": 0.15723253786563873,
      "learning_rate": 4.905695096145e-05,
      "loss": 0.6151,
      "mean_token_accuracy": 0.8109627738595009,
      "num_tokens": 530175230.0,
      "step": 16612
    },
    {
      "entropy": 0.604949189350009,
      "epoch": 1.528484597558381,
      "grad_norm": 0.1615719348192215,
      "learning_rate": 4.905388413530837e-05,
      "loss": 0.5969,
      "mean_token_accuracy": 0.8174238204956055,
      "num_tokens": 530206448.0,
      "step": 16613
    },
    {
      "entropy": 0.6014221142977476,
      "epoch": 1.528576603929822,
      "grad_norm": 0.15043573081493378,
      "learning_rate": 4.905081730916675e-05,
      "loss": 0.5906,
      "mean_token_accuracy": 0.8200630024075508,
      "num_tokens": 530238650.0,
      "step": 16614
    },
    {
      "entropy": 0.6016643699258566,
      "epoch": 1.5286686103012634,
      "grad_norm": 0.1589881032705307,
      "learning_rate": 4.904775048302512e-05,
      "loss": 0.5905,
      "mean_token_accuracy": 0.8186891749501228,
      "num_tokens": 530270416.0,
      "step": 16615
    },
    {
      "entropy": 0.4937232434749603,
      "epoch": 1.5287606166727046,
      "grad_norm": 0.14372655749320984,
      "learning_rate": 4.904468365688349e-05,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8525296449661255,
      "num_tokens": 530302836.0,
      "step": 16616
    },
    {
      "entropy": 0.5985220316797495,
      "epoch": 1.5288526230441457,
      "grad_norm": 0.15659675002098083,
      "learning_rate": 4.9041616830741866e-05,
      "loss": 0.5797,
      "mean_token_accuracy": 0.8201946951448917,
      "num_tokens": 530335217.0,
      "step": 16617
    },
    {
      "entropy": 0.6861875392496586,
      "epoch": 1.528944629415587,
      "grad_norm": 0.15662506222724915,
      "learning_rate": 4.903855000460024e-05,
      "loss": 0.6782,
      "mean_token_accuracy": 0.7947066947817802,
      "num_tokens": 530367957.0,
      "step": 16618
    },
    {
      "entropy": 0.5980888083577156,
      "epoch": 1.5290366357870282,
      "grad_norm": 0.15480278432369232,
      "learning_rate": 4.9035483178458616e-05,
      "loss": 0.5963,
      "mean_token_accuracy": 0.818034715950489,
      "num_tokens": 530400480.0,
      "step": 16619
    },
    {
      "entropy": 0.668041281402111,
      "epoch": 1.5291286421584696,
      "grad_norm": 0.15949440002441406,
      "learning_rate": 4.903241635231699e-05,
      "loss": 0.6507,
      "mean_token_accuracy": 0.802837073802948,
      "num_tokens": 530432510.0,
      "step": 16620
    },
    {
      "entropy": 0.6033818144351244,
      "epoch": 1.5292206485299107,
      "grad_norm": 0.1541086733341217,
      "learning_rate": 4.902934952617536e-05,
      "loss": 0.608,
      "mean_token_accuracy": 0.8121641837060452,
      "num_tokens": 530464001.0,
      "step": 16621
    },
    {
      "entropy": 0.5126804020255804,
      "epoch": 1.5293126549013518,
      "grad_norm": 0.15112438797950745,
      "learning_rate": 4.902628270003374e-05,
      "loss": 0.4948,
      "mean_token_accuracy": 0.8479503877460957,
      "num_tokens": 530496253.0,
      "step": 16622
    },
    {
      "entropy": 0.5884058908559382,
      "epoch": 1.5294046612727932,
      "grad_norm": 0.15069401264190674,
      "learning_rate": 4.902321587389211e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8224900551140308,
      "num_tokens": 530528039.0,
      "step": 16623
    },
    {
      "entropy": 0.5845776156056672,
      "epoch": 1.5294966676442343,
      "grad_norm": 0.15633410215377808,
      "learning_rate": 4.902014904775048e-05,
      "loss": 0.5853,
      "mean_token_accuracy": 0.8193132691085339,
      "num_tokens": 530560312.0,
      "step": 16624
    },
    {
      "entropy": 0.48532241769135,
      "epoch": 1.5295886740156757,
      "grad_norm": 0.1441277712583542,
      "learning_rate": 4.9017082221608863e-05,
      "loss": 0.4682,
      "mean_token_accuracy": 0.85352573543787,
      "num_tokens": 530592718.0,
      "step": 16625
    },
    {
      "entropy": 0.6081004061270505,
      "epoch": 1.5296806803871168,
      "grad_norm": 0.15527519583702087,
      "learning_rate": 4.901401539546723e-05,
      "loss": 0.603,
      "mean_token_accuracy": 0.8155422136187553,
      "num_tokens": 530624830.0,
      "step": 16626
    },
    {
      "entropy": 0.6745463516563177,
      "epoch": 1.529772686758558,
      "grad_norm": 0.16330915689468384,
      "learning_rate": 4.901094856932561e-05,
      "loss": 0.658,
      "mean_token_accuracy": 0.799688208848238,
      "num_tokens": 530656131.0,
      "step": 16627
    },
    {
      "entropy": 0.6008928194642067,
      "epoch": 1.5298646931299993,
      "grad_norm": 0.1591942459344864,
      "learning_rate": 4.900788174318398e-05,
      "loss": 0.5996,
      "mean_token_accuracy": 0.8157485239207745,
      "num_tokens": 530687278.0,
      "step": 16628
    },
    {
      "entropy": 0.6307365484535694,
      "epoch": 1.5299566995014404,
      "grad_norm": 0.15327520668506622,
      "learning_rate": 4.9004814917042355e-05,
      "loss": 0.6174,
      "mean_token_accuracy": 0.8103295341134071,
      "num_tokens": 530718702.0,
      "step": 16629
    },
    {
      "entropy": 0.5370337897911668,
      "epoch": 1.5300487058728818,
      "grad_norm": 0.15016305446624756,
      "learning_rate": 4.900174809090073e-05,
      "loss": 0.5332,
      "mean_token_accuracy": 0.8382038585841656,
      "num_tokens": 530750448.0,
      "step": 16630
    },
    {
      "entropy": 0.5938581619411707,
      "epoch": 1.530140712244323,
      "grad_norm": 0.16662907600402832,
      "learning_rate": 4.8998681264759105e-05,
      "loss": 0.5897,
      "mean_token_accuracy": 0.8194583989679813,
      "num_tokens": 530782243.0,
      "step": 16631
    },
    {
      "entropy": 0.571426962967962,
      "epoch": 1.530232718615764,
      "grad_norm": 0.17081645131111145,
      "learning_rate": 4.899561443861748e-05,
      "loss": 0.5692,
      "mean_token_accuracy": 0.8307923413813114,
      "num_tokens": 530814823.0,
      "step": 16632
    },
    {
      "entropy": 0.6717910822480917,
      "epoch": 1.5303247249872054,
      "grad_norm": 0.15765047073364258,
      "learning_rate": 4.8992547612475854e-05,
      "loss": 0.6747,
      "mean_token_accuracy": 0.7959094867110252,
      "num_tokens": 530846899.0,
      "step": 16633
    },
    {
      "entropy": 0.6502723768353462,
      "epoch": 1.5304167313586465,
      "grad_norm": 0.15815700590610504,
      "learning_rate": 4.898948078633422e-05,
      "loss": 0.6516,
      "mean_token_accuracy": 0.8041570149362087,
      "num_tokens": 530879434.0,
      "step": 16634
    },
    {
      "entropy": 0.4505192628130317,
      "epoch": 1.5305087377300879,
      "grad_norm": 0.1403140425682068,
      "learning_rate": 4.89864139601926e-05,
      "loss": 0.44,
      "mean_token_accuracy": 0.8615367710590363,
      "num_tokens": 530911551.0,
      "step": 16635
    },
    {
      "entropy": 0.6352963699027896,
      "epoch": 1.530600744101529,
      "grad_norm": 0.15829618275165558,
      "learning_rate": 4.898334713405097e-05,
      "loss": 0.6332,
      "mean_token_accuracy": 0.8088048920035362,
      "num_tokens": 530943807.0,
      "step": 16636
    },
    {
      "entropy": 0.5906238593161106,
      "epoch": 1.5306927504729702,
      "grad_norm": 0.1438986212015152,
      "learning_rate": 4.8980280307909346e-05,
      "loss": 0.5694,
      "mean_token_accuracy": 0.8233708068728447,
      "num_tokens": 530975916.0,
      "step": 16637
    },
    {
      "entropy": 0.6652483213692904,
      "epoch": 1.5307847568444115,
      "grad_norm": 0.15785947442054749,
      "learning_rate": 4.897721348176772e-05,
      "loss": 0.6429,
      "mean_token_accuracy": 0.8034623637795448,
      "num_tokens": 531007047.0,
      "step": 16638
    },
    {
      "entropy": 0.5629390422254801,
      "epoch": 1.5308767632158526,
      "grad_norm": 0.15487270057201385,
      "learning_rate": 4.8974146655626095e-05,
      "loss": 0.5426,
      "mean_token_accuracy": 0.8363204263150692,
      "num_tokens": 531039257.0,
      "step": 16639
    },
    {
      "entropy": 0.6189919952303171,
      "epoch": 1.530968769587294,
      "grad_norm": 0.14965684711933136,
      "learning_rate": 4.897107982948447e-05,
      "loss": 0.6112,
      "mean_token_accuracy": 0.8084173500537872,
      "num_tokens": 531071936.0,
      "step": 16640
    },
    {
      "entropy": 0.5446044262498617,
      "epoch": 1.5310607759587351,
      "grad_norm": 0.14827556908130646,
      "learning_rate": 4.8968013003342844e-05,
      "loss": 0.5478,
      "mean_token_accuracy": 0.8316228464245796,
      "num_tokens": 531104225.0,
      "step": 16641
    },
    {
      "entropy": 0.6171448053792119,
      "epoch": 1.5311527823301763,
      "grad_norm": 0.15235379338264465,
      "learning_rate": 4.896494617720121e-05,
      "loss": 0.6109,
      "mean_token_accuracy": 0.8100639134645462,
      "num_tokens": 531136022.0,
      "step": 16642
    },
    {
      "entropy": 0.5934058157727122,
      "epoch": 1.5312447887016176,
      "grad_norm": 0.15000678598880768,
      "learning_rate": 4.8961879351059594e-05,
      "loss": 0.5691,
      "mean_token_accuracy": 0.8280476368963718,
      "num_tokens": 531167851.0,
      "step": 16643
    },
    {
      "entropy": 0.5667247008532286,
      "epoch": 1.5313367950730588,
      "grad_norm": 0.14305377006530762,
      "learning_rate": 4.895881252491796e-05,
      "loss": 0.5535,
      "mean_token_accuracy": 0.830923356115818,
      "num_tokens": 531200279.0,
      "step": 16644
    },
    {
      "entropy": 0.5792864188551903,
      "epoch": 1.5314288014445,
      "grad_norm": 0.15312492847442627,
      "learning_rate": 4.895574569877634e-05,
      "loss": 0.5692,
      "mean_token_accuracy": 0.822537861764431,
      "num_tokens": 531232616.0,
      "step": 16645
    },
    {
      "entropy": 0.6841884963214397,
      "epoch": 1.5315208078159412,
      "grad_norm": 0.1733655035495758,
      "learning_rate": 4.895267887263471e-05,
      "loss": 0.667,
      "mean_token_accuracy": 0.7927428185939789,
      "num_tokens": 531264140.0,
      "step": 16646
    },
    {
      "entropy": 0.6781822219491005,
      "epoch": 1.5316128141873824,
      "grad_norm": 0.16188964247703552,
      "learning_rate": 4.8949612046493085e-05,
      "loss": 0.6728,
      "mean_token_accuracy": 0.7969134040176868,
      "num_tokens": 531296711.0,
      "step": 16647
    },
    {
      "entropy": 0.5893618473783135,
      "epoch": 1.5317048205588237,
      "grad_norm": 0.1588018834590912,
      "learning_rate": 4.894654522035146e-05,
      "loss": 0.5893,
      "mean_token_accuracy": 0.8220020085573196,
      "num_tokens": 531328903.0,
      "step": 16648
    },
    {
      "entropy": 0.5759569443762302,
      "epoch": 1.5317968269302649,
      "grad_norm": 0.16322802007198334,
      "learning_rate": 4.8943478394209835e-05,
      "loss": 0.5717,
      "mean_token_accuracy": 0.8268704675137997,
      "num_tokens": 531361431.0,
      "step": 16649
    },
    {
      "entropy": 0.5955304643139243,
      "epoch": 1.5318888333017062,
      "grad_norm": 0.14875562489032745,
      "learning_rate": 4.894041156806821e-05,
      "loss": 0.5827,
      "mean_token_accuracy": 0.824214905500412,
      "num_tokens": 531393479.0,
      "step": 16650
    },
    {
      "entropy": 0.6018773261457682,
      "epoch": 1.5319808396731474,
      "grad_norm": 0.15670734643936157,
      "learning_rate": 4.8937344741926584e-05,
      "loss": 0.5879,
      "mean_token_accuracy": 0.8177613392472267,
      "num_tokens": 531425426.0,
      "step": 16651
    },
    {
      "entropy": 0.6039286162704229,
      "epoch": 1.5320728460445885,
      "grad_norm": 0.1597594916820526,
      "learning_rate": 4.893427791578495e-05,
      "loss": 0.5901,
      "mean_token_accuracy": 0.8204764910042286,
      "num_tokens": 531457820.0,
      "step": 16652
    },
    {
      "entropy": 0.5656763203442097,
      "epoch": 1.5321648524160298,
      "grad_norm": 0.15611256659030914,
      "learning_rate": 4.893121108964333e-05,
      "loss": 0.5402,
      "mean_token_accuracy": 0.8303453549742699,
      "num_tokens": 531490335.0,
      "step": 16653
    },
    {
      "entropy": 0.6991248494014144,
      "epoch": 1.5322568587874712,
      "grad_norm": 0.16586142778396606,
      "learning_rate": 4.89281442635017e-05,
      "loss": 0.7031,
      "mean_token_accuracy": 0.7848452515900135,
      "num_tokens": 531522302.0,
      "step": 16654
    },
    {
      "entropy": 0.6162538155913353,
      "epoch": 1.5323488651589123,
      "grad_norm": 0.1586848497390747,
      "learning_rate": 4.8925077437360076e-05,
      "loss": 0.6013,
      "mean_token_accuracy": 0.8168446645140648,
      "num_tokens": 531554333.0,
      "step": 16655
    },
    {
      "entropy": 0.5896953158080578,
      "epoch": 1.5324408715303535,
      "grad_norm": 0.14847610890865326,
      "learning_rate": 4.892201061121846e-05,
      "loss": 0.5732,
      "mean_token_accuracy": 0.819385688751936,
      "num_tokens": 531586617.0,
      "step": 16656
    },
    {
      "entropy": 0.6233134772628546,
      "epoch": 1.5325328779017946,
      "grad_norm": 0.15272121131420135,
      "learning_rate": 4.8918943785076825e-05,
      "loss": 0.6164,
      "mean_token_accuracy": 0.8107400983572006,
      "num_tokens": 531618765.0,
      "step": 16657
    },
    {
      "entropy": 0.7047692267224193,
      "epoch": 1.532624884273236,
      "grad_norm": 0.16521471738815308,
      "learning_rate": 4.89158769589352e-05,
      "loss": 0.7093,
      "mean_token_accuracy": 0.7870591543614864,
      "num_tokens": 531651510.0,
      "step": 16658
    },
    {
      "entropy": 0.5621075159870088,
      "epoch": 1.5327168906446773,
      "grad_norm": 0.14508448541164398,
      "learning_rate": 4.8912810132793574e-05,
      "loss": 0.5468,
      "mean_token_accuracy": 0.834070585668087,
      "num_tokens": 531683940.0,
      "step": 16659
    },
    {
      "entropy": 0.6266862582415342,
      "epoch": 1.5328088970161184,
      "grad_norm": 0.15110842883586884,
      "learning_rate": 4.890974330665195e-05,
      "loss": 0.6078,
      "mean_token_accuracy": 0.8122256807982922,
      "num_tokens": 531716322.0,
      "step": 16660
    },
    {
      "entropy": 0.6013573613017797,
      "epoch": 1.5329009033875596,
      "grad_norm": 0.15289324522018433,
      "learning_rate": 4.8906676480510324e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8174616396427155,
      "num_tokens": 531748499.0,
      "step": 16661
    },
    {
      "entropy": 0.5612768856808543,
      "epoch": 1.5329929097590007,
      "grad_norm": 0.15635135769844055,
      "learning_rate": 4.89036096543687e-05,
      "loss": 0.5432,
      "mean_token_accuracy": 0.8331555053591728,
      "num_tokens": 531780458.0,
      "step": 16662
    },
    {
      "entropy": 0.5733124697580934,
      "epoch": 1.533084916130442,
      "grad_norm": 0.1501568853855133,
      "learning_rate": 4.8900542828227066e-05,
      "loss": 0.5502,
      "mean_token_accuracy": 0.8309986591339111,
      "num_tokens": 531812415.0,
      "step": 16663
    },
    {
      "entropy": 0.5382702429778874,
      "epoch": 1.5331769225018834,
      "grad_norm": 0.14644856750965118,
      "learning_rate": 4.889747600208545e-05,
      "loss": 0.5107,
      "mean_token_accuracy": 0.8424866162240505,
      "num_tokens": 531844613.0,
      "step": 16664
    },
    {
      "entropy": 0.6322224866598845,
      "epoch": 1.5332689288733246,
      "grad_norm": 0.16455194354057312,
      "learning_rate": 4.8894409175943816e-05,
      "loss": 0.6348,
      "mean_token_accuracy": 0.809073805809021,
      "num_tokens": 531877001.0,
      "step": 16665
    },
    {
      "entropy": 0.597591548692435,
      "epoch": 1.5333609352447657,
      "grad_norm": 0.1564944088459015,
      "learning_rate": 4.88913423498022e-05,
      "loss": 0.5841,
      "mean_token_accuracy": 0.8216569572687149,
      "num_tokens": 531909031.0,
      "step": 16666
    },
    {
      "entropy": 0.6313079334795475,
      "epoch": 1.5334529416162068,
      "grad_norm": 0.16050103306770325,
      "learning_rate": 4.8888275523660565e-05,
      "loss": 0.6311,
      "mean_token_accuracy": 0.8061057962477207,
      "num_tokens": 531941089.0,
      "step": 16667
    },
    {
      "entropy": 0.6122173722833395,
      "epoch": 1.5335449479876482,
      "grad_norm": 0.15885518491268158,
      "learning_rate": 4.888520869751894e-05,
      "loss": 0.6013,
      "mean_token_accuracy": 0.8170132488012314,
      "num_tokens": 531972462.0,
      "step": 16668
    },
    {
      "entropy": 0.6524430452845991,
      "epoch": 1.5336369543590895,
      "grad_norm": 0.16495376825332642,
      "learning_rate": 4.8882141871377314e-05,
      "loss": 0.6499,
      "mean_token_accuracy": 0.802241075783968,
      "num_tokens": 532005230.0,
      "step": 16669
    },
    {
      "entropy": 0.5785846607759595,
      "epoch": 1.5337289607305307,
      "grad_norm": 0.15307657420635223,
      "learning_rate": 4.887907504523569e-05,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8265798054635525,
      "num_tokens": 532037158.0,
      "step": 16670
    },
    {
      "entropy": 0.5381657555699348,
      "epoch": 1.5338209671019718,
      "grad_norm": 0.15508578717708588,
      "learning_rate": 4.8876008219094063e-05,
      "loss": 0.5284,
      "mean_token_accuracy": 0.8350585140287876,
      "num_tokens": 532069482.0,
      "step": 16671
    },
    {
      "entropy": 0.6474285088479519,
      "epoch": 1.533912973473413,
      "grad_norm": 0.16052773594856262,
      "learning_rate": 4.887294139295244e-05,
      "loss": 0.6345,
      "mean_token_accuracy": 0.8036474213004112,
      "num_tokens": 532102017.0,
      "step": 16672
    },
    {
      "entropy": 0.7008578618988395,
      "epoch": 1.5340049798448543,
      "grad_norm": 0.16175587475299835,
      "learning_rate": 4.8869874566810806e-05,
      "loss": 0.7033,
      "mean_token_accuracy": 0.7831700555980206,
      "num_tokens": 532133986.0,
      "step": 16673
    },
    {
      "entropy": 0.4605826176702976,
      "epoch": 1.5340969862162956,
      "grad_norm": 0.14482422173023224,
      "learning_rate": 4.886680774066919e-05,
      "loss": 0.4437,
      "mean_token_accuracy": 0.8619328364729881,
      "num_tokens": 532166616.0,
      "step": 16674
    },
    {
      "entropy": 0.7657330799847841,
      "epoch": 1.5341889925877368,
      "grad_norm": 0.16430462896823883,
      "learning_rate": 4.8863740914527555e-05,
      "loss": 0.7591,
      "mean_token_accuracy": 0.7735075354576111,
      "num_tokens": 532198797.0,
      "step": 16675
    },
    {
      "entropy": 0.6293457429856062,
      "epoch": 1.534280998959178,
      "grad_norm": 0.15480011701583862,
      "learning_rate": 4.886067408838593e-05,
      "loss": 0.6214,
      "mean_token_accuracy": 0.8089748881757259,
      "num_tokens": 532231307.0,
      "step": 16676
    },
    {
      "entropy": 0.5405803984031081,
      "epoch": 1.534373005330619,
      "grad_norm": 0.14856606721878052,
      "learning_rate": 4.8857607262244305e-05,
      "loss": 0.511,
      "mean_token_accuracy": 0.8402869552373886,
      "num_tokens": 532262154.0,
      "step": 16677
    },
    {
      "entropy": 0.7071920104790479,
      "epoch": 1.5344650117020604,
      "grad_norm": 0.15633615851402283,
      "learning_rate": 4.885454043610268e-05,
      "loss": 0.6942,
      "mean_token_accuracy": 0.7889284677803516,
      "num_tokens": 532294608.0,
      "step": 16678
    },
    {
      "entropy": 0.6945059932768345,
      "epoch": 1.5345570180735018,
      "grad_norm": 0.1605415642261505,
      "learning_rate": 4.8851473609961054e-05,
      "loss": 0.6951,
      "mean_token_accuracy": 0.7898929677903652,
      "num_tokens": 532326057.0,
      "step": 16679
    },
    {
      "entropy": 0.7223389558494091,
      "epoch": 1.5346490244449429,
      "grad_norm": 0.1634630411863327,
      "learning_rate": 4.884840678381943e-05,
      "loss": 0.733,
      "mean_token_accuracy": 0.7790688686072826,
      "num_tokens": 532357923.0,
      "step": 16680
    },
    {
      "entropy": 0.46645402535796165,
      "epoch": 1.534741030816384,
      "grad_norm": 0.1376510113477707,
      "learning_rate": 4.8845339957677796e-05,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8559658005833626,
      "num_tokens": 532389556.0,
      "step": 16681
    },
    {
      "entropy": 0.6717223324812949,
      "epoch": 1.5348330371878252,
      "grad_norm": 0.15410004556179047,
      "learning_rate": 4.884227313153618e-05,
      "loss": 0.6548,
      "mean_token_accuracy": 0.8058447502553463,
      "num_tokens": 532421678.0,
      "step": 16682
    },
    {
      "entropy": 0.5924117853865027,
      "epoch": 1.5349250435592665,
      "grad_norm": 0.1518900841474533,
      "learning_rate": 4.8839206305394546e-05,
      "loss": 0.5801,
      "mean_token_accuracy": 0.82110995054245,
      "num_tokens": 532453986.0,
      "step": 16683
    },
    {
      "entropy": 0.676208607852459,
      "epoch": 1.5350170499307079,
      "grad_norm": 0.1563485562801361,
      "learning_rate": 4.883613947925293e-05,
      "loss": 0.6787,
      "mean_token_accuracy": 0.7952880300581455,
      "num_tokens": 532486062.0,
      "step": 16684
    },
    {
      "entropy": 0.6767987366765738,
      "epoch": 1.535109056302149,
      "grad_norm": 0.15635822713375092,
      "learning_rate": 4.8833072653111295e-05,
      "loss": 0.6757,
      "mean_token_accuracy": 0.7924316674470901,
      "num_tokens": 532518670.0,
      "step": 16685
    },
    {
      "entropy": 0.5538074132055044,
      "epoch": 1.5352010626735901,
      "grad_norm": 0.14764775335788727,
      "learning_rate": 4.883000582696967e-05,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8351359702646732,
      "num_tokens": 532550954.0,
      "step": 16686
    },
    {
      "entropy": 0.6425099708139896,
      "epoch": 1.5352930690450313,
      "grad_norm": 0.16071702539920807,
      "learning_rate": 4.882693900082805e-05,
      "loss": 0.622,
      "mean_token_accuracy": 0.8084884323179722,
      "num_tokens": 532582043.0,
      "step": 16687
    },
    {
      "entropy": 0.5302258357405663,
      "epoch": 1.5353850754164726,
      "grad_norm": 0.15345506370067596,
      "learning_rate": 4.882387217468642e-05,
      "loss": 0.5124,
      "mean_token_accuracy": 0.8372346200048923,
      "num_tokens": 532613927.0,
      "step": 16688
    },
    {
      "entropy": 0.6772467140108347,
      "epoch": 1.535477081787914,
      "grad_norm": 0.15625083446502686,
      "learning_rate": 4.8820805348544793e-05,
      "loss": 0.6655,
      "mean_token_accuracy": 0.7971406728029251,
      "num_tokens": 532646396.0,
      "step": 16689
    },
    {
      "entropy": 0.6148962508887053,
      "epoch": 1.5355690881593551,
      "grad_norm": 0.15106025338172913,
      "learning_rate": 4.881773852240317e-05,
      "loss": 0.5937,
      "mean_token_accuracy": 0.8168198093771935,
      "num_tokens": 532678924.0,
      "step": 16690
    },
    {
      "entropy": 0.5803386624902487,
      "epoch": 1.5356610945307962,
      "grad_norm": 0.14902809262275696,
      "learning_rate": 4.881467169626154e-05,
      "loss": 0.5751,
      "mean_token_accuracy": 0.8243416324257851,
      "num_tokens": 532711337.0,
      "step": 16691
    },
    {
      "entropy": 0.6010224847123027,
      "epoch": 1.5357531009022374,
      "grad_norm": 0.15132151544094086,
      "learning_rate": 4.881160487011992e-05,
      "loss": 0.5911,
      "mean_token_accuracy": 0.8199749365448952,
      "num_tokens": 532743578.0,
      "step": 16692
    },
    {
      "entropy": 0.6474008355289698,
      "epoch": 1.5358451072736787,
      "grad_norm": 0.1754487007856369,
      "learning_rate": 4.880853804397829e-05,
      "loss": 0.6362,
      "mean_token_accuracy": 0.8057382479310036,
      "num_tokens": 532775010.0,
      "step": 16693
    },
    {
      "entropy": 0.574708984233439,
      "epoch": 1.53593711364512,
      "grad_norm": 0.1543949693441391,
      "learning_rate": 4.880547121783666e-05,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8291515968739986,
      "num_tokens": 532807097.0,
      "step": 16694
    },
    {
      "entropy": 0.6019217129796743,
      "epoch": 1.5360291200165612,
      "grad_norm": 0.15820766985416412,
      "learning_rate": 4.880240439169504e-05,
      "loss": 0.5953,
      "mean_token_accuracy": 0.8193766176700592,
      "num_tokens": 532838922.0,
      "step": 16695
    },
    {
      "entropy": 0.6895260820165277,
      "epoch": 1.5361211263880024,
      "grad_norm": 0.16718529164791107,
      "learning_rate": 4.879933756555341e-05,
      "loss": 0.6924,
      "mean_token_accuracy": 0.7891257666051388,
      "num_tokens": 532870486.0,
      "step": 16696
    },
    {
      "entropy": 0.6488480968400836,
      "epoch": 1.5362131327594435,
      "grad_norm": 0.16100850701332092,
      "learning_rate": 4.8796270739411784e-05,
      "loss": 0.6403,
      "mean_token_accuracy": 0.804622620344162,
      "num_tokens": 532903254.0,
      "step": 16697
    },
    {
      "entropy": 0.5933830412104726,
      "epoch": 1.5363051391308848,
      "grad_norm": 0.15581002831459045,
      "learning_rate": 4.879320391327016e-05,
      "loss": 0.5833,
      "mean_token_accuracy": 0.8233758509159088,
      "num_tokens": 532935690.0,
      "step": 16698
    },
    {
      "entropy": 0.6568577094003558,
      "epoch": 1.5363971455023262,
      "grad_norm": 0.15718239545822144,
      "learning_rate": 4.879013708712853e-05,
      "loss": 0.6472,
      "mean_token_accuracy": 0.8060807026922703,
      "num_tokens": 532968274.0,
      "step": 16699
    },
    {
      "entropy": 0.5641084890812635,
      "epoch": 1.5364891518737673,
      "grad_norm": 0.15360265970230103,
      "learning_rate": 4.878707026098691e-05,
      "loss": 0.5598,
      "mean_token_accuracy": 0.8234237991273403,
      "num_tokens": 533000220.0,
      "step": 16700
    },
    {
      "entropy": 0.6382030192762613,
      "epoch": 1.5365811582452085,
      "grad_norm": 0.15776170790195465,
      "learning_rate": 4.878400343484528e-05,
      "loss": 0.6207,
      "mean_token_accuracy": 0.8119431473314762,
      "num_tokens": 533032682.0,
      "step": 16701
    },
    {
      "entropy": 0.7056819330900908,
      "epoch": 1.5366731646166496,
      "grad_norm": 0.1644447296857834,
      "learning_rate": 4.878093660870365e-05,
      "loss": 0.6977,
      "mean_token_accuracy": 0.7899991311132908,
      "num_tokens": 533063590.0,
      "step": 16702
    },
    {
      "entropy": 0.6081531997770071,
      "epoch": 1.536765170988091,
      "grad_norm": 0.1549336314201355,
      "learning_rate": 4.877786978256203e-05,
      "loss": 0.5841,
      "mean_token_accuracy": 0.8195102326571941,
      "num_tokens": 533094575.0,
      "step": 16703
    },
    {
      "entropy": 0.6477363612502813,
      "epoch": 1.5368571773595323,
      "grad_norm": 0.16168421506881714,
      "learning_rate": 4.87748029564204e-05,
      "loss": 0.6558,
      "mean_token_accuracy": 0.7980470918118954,
      "num_tokens": 533125609.0,
      "step": 16704
    },
    {
      "entropy": 0.5999901574105024,
      "epoch": 1.5369491837309734,
      "grad_norm": 0.1497228443622589,
      "learning_rate": 4.877173613027878e-05,
      "loss": 0.5776,
      "mean_token_accuracy": 0.8217574320733547,
      "num_tokens": 533158284.0,
      "step": 16705
    },
    {
      "entropy": 0.6490564085543156,
      "epoch": 1.5370411901024146,
      "grad_norm": 0.1534540355205536,
      "learning_rate": 4.876866930413715e-05,
      "loss": 0.6381,
      "mean_token_accuracy": 0.8056481555104256,
      "num_tokens": 533189792.0,
      "step": 16706
    },
    {
      "entropy": 0.6902514845132828,
      "epoch": 1.5371331964738557,
      "grad_norm": 0.16632488369941711,
      "learning_rate": 4.8765602477995524e-05,
      "loss": 0.6745,
      "mean_token_accuracy": 0.799822848290205,
      "num_tokens": 533220814.0,
      "step": 16707
    },
    {
      "entropy": 0.7360730767250061,
      "epoch": 1.537225202845297,
      "grad_norm": 0.16446572542190552,
      "learning_rate": 4.87625356518539e-05,
      "loss": 0.727,
      "mean_token_accuracy": 0.7802900560200214,
      "num_tokens": 533253058.0,
      "step": 16708
    },
    {
      "entropy": 0.6738643236458302,
      "epoch": 1.5373172092167384,
      "grad_norm": 0.15477055311203003,
      "learning_rate": 4.875946882571227e-05,
      "loss": 0.6546,
      "mean_token_accuracy": 0.797528114169836,
      "num_tokens": 533284899.0,
      "step": 16709
    },
    {
      "entropy": 0.5609199749305844,
      "epoch": 1.5374092155881796,
      "grad_norm": 0.14616334438323975,
      "learning_rate": 4.875640199957065e-05,
      "loss": 0.5424,
      "mean_token_accuracy": 0.8359145857393742,
      "num_tokens": 533317048.0,
      "step": 16710
    },
    {
      "entropy": 0.5417928625829518,
      "epoch": 1.5375012219596207,
      "grad_norm": 0.14650525152683258,
      "learning_rate": 4.875333517342902e-05,
      "loss": 0.5274,
      "mean_token_accuracy": 0.839940071105957,
      "num_tokens": 533349241.0,
      "step": 16711
    },
    {
      "entropy": 0.615096696652472,
      "epoch": 1.5375932283310618,
      "grad_norm": 0.15478812158107758,
      "learning_rate": 4.875026834728739e-05,
      "loss": 0.5935,
      "mean_token_accuracy": 0.8155270926654339,
      "num_tokens": 533380691.0,
      "step": 16712
    },
    {
      "entropy": 0.7191495532169938,
      "epoch": 1.5376852347025032,
      "grad_norm": 0.1601143479347229,
      "learning_rate": 4.874720152114577e-05,
      "loss": 0.711,
      "mean_token_accuracy": 0.7853939719498158,
      "num_tokens": 533413186.0,
      "step": 16713
    },
    {
      "entropy": 0.6345072835683823,
      "epoch": 1.5377772410739445,
      "grad_norm": 0.1556129902601242,
      "learning_rate": 4.874413469500414e-05,
      "loss": 0.628,
      "mean_token_accuracy": 0.8037090674042702,
      "num_tokens": 533445160.0,
      "step": 16714
    },
    {
      "entropy": 0.6747263558208942,
      "epoch": 1.5378692474453857,
      "grad_norm": 0.15680554509162903,
      "learning_rate": 4.8741067868862514e-05,
      "loss": 0.6688,
      "mean_token_accuracy": 0.7963961884379387,
      "num_tokens": 533476861.0,
      "step": 16715
    },
    {
      "entropy": 0.6093288145493716,
      "epoch": 1.5379612538168268,
      "grad_norm": 0.149540513753891,
      "learning_rate": 4.873800104272089e-05,
      "loss": 0.5985,
      "mean_token_accuracy": 0.8194749616086483,
      "num_tokens": 533509272.0,
      "step": 16716
    },
    {
      "entropy": 0.5796257946640253,
      "epoch": 1.538053260188268,
      "grad_norm": 0.15482284128665924,
      "learning_rate": 4.873493421657926e-05,
      "loss": 0.5554,
      "mean_token_accuracy": 0.826549094170332,
      "num_tokens": 533540658.0,
      "step": 16717
    },
    {
      "entropy": 0.6808862667530775,
      "epoch": 1.5381452665597093,
      "grad_norm": 0.16316193342208862,
      "learning_rate": 4.873186739043764e-05,
      "loss": 0.6675,
      "mean_token_accuracy": 0.794792402535677,
      "num_tokens": 533572309.0,
      "step": 16718
    },
    {
      "entropy": 0.5673627685755491,
      "epoch": 1.5382372729311506,
      "grad_norm": 0.16731782257556915,
      "learning_rate": 4.872880056429601e-05,
      "loss": 0.5585,
      "mean_token_accuracy": 0.8272229135036469,
      "num_tokens": 533602810.0,
      "step": 16719
    },
    {
      "entropy": 0.5157920867204666,
      "epoch": 1.5383292793025918,
      "grad_norm": 0.1466541886329651,
      "learning_rate": 4.872573373815439e-05,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8423049002885818,
      "num_tokens": 533634672.0,
      "step": 16720
    },
    {
      "entropy": 0.5895738992840052,
      "epoch": 1.538421285674033,
      "grad_norm": 0.16288624703884125,
      "learning_rate": 4.872266691201276e-05,
      "loss": 0.5882,
      "mean_token_accuracy": 0.8188086226582527,
      "num_tokens": 533665814.0,
      "step": 16721
    },
    {
      "entropy": 0.5781047027558088,
      "epoch": 1.538513292045474,
      "grad_norm": 0.15623857080936432,
      "learning_rate": 4.8719600085871137e-05,
      "loss": 0.5727,
      "mean_token_accuracy": 0.8201611861586571,
      "num_tokens": 533696586.0,
      "step": 16722
    },
    {
      "entropy": 0.6294315166305751,
      "epoch": 1.5386052984169154,
      "grad_norm": 0.17334984242916107,
      "learning_rate": 4.8716533259729504e-05,
      "loss": 0.6277,
      "mean_token_accuracy": 0.8061946593225002,
      "num_tokens": 533728350.0,
      "step": 16723
    },
    {
      "entropy": 0.7119524963200092,
      "epoch": 1.5386973047883568,
      "grad_norm": 0.16230995953083038,
      "learning_rate": 4.8713466433587886e-05,
      "loss": 0.7051,
      "mean_token_accuracy": 0.7872883789241314,
      "num_tokens": 533760645.0,
      "step": 16724
    },
    {
      "entropy": 0.4734776224941015,
      "epoch": 1.538789311159798,
      "grad_norm": 0.1463349461555481,
      "learning_rate": 4.8710399607446254e-05,
      "loss": 0.4694,
      "mean_token_accuracy": 0.85218720510602,
      "num_tokens": 533792283.0,
      "step": 16725
    },
    {
      "entropy": 0.6344854477792978,
      "epoch": 1.538881317531239,
      "grad_norm": 0.15783479809761047,
      "learning_rate": 4.8707332781304635e-05,
      "loss": 0.631,
      "mean_token_accuracy": 0.8068316839635372,
      "num_tokens": 533824598.0,
      "step": 16726
    },
    {
      "entropy": 0.6974272429943085,
      "epoch": 1.5389733239026802,
      "grad_norm": 0.1640259474515915,
      "learning_rate": 4.8704265955163e-05,
      "loss": 0.6743,
      "mean_token_accuracy": 0.7946178428828716,
      "num_tokens": 533855802.0,
      "step": 16727
    },
    {
      "entropy": 0.6250097402371466,
      "epoch": 1.5390653302741215,
      "grad_norm": 0.1628042310476303,
      "learning_rate": 4.870119912902138e-05,
      "loss": 0.6088,
      "mean_token_accuracy": 0.8119150027632713,
      "num_tokens": 533887592.0,
      "step": 16728
    },
    {
      "entropy": 0.6422786023467779,
      "epoch": 1.5391573366455629,
      "grad_norm": 0.16321785748004913,
      "learning_rate": 4.869813230287975e-05,
      "loss": 0.6333,
      "mean_token_accuracy": 0.8084430173039436,
      "num_tokens": 533918566.0,
      "step": 16729
    },
    {
      "entropy": 0.575001603923738,
      "epoch": 1.539249343017004,
      "grad_norm": 0.15427012741565704,
      "learning_rate": 4.869506547673813e-05,
      "loss": 0.5658,
      "mean_token_accuracy": 0.8239607661962509,
      "num_tokens": 533950905.0,
      "step": 16730
    },
    {
      "entropy": 0.5906529063358903,
      "epoch": 1.5393413493884451,
      "grad_norm": 0.15089957416057587,
      "learning_rate": 4.86919986505965e-05,
      "loss": 0.5802,
      "mean_token_accuracy": 0.8218013234436512,
      "num_tokens": 533982869.0,
      "step": 16731
    },
    {
      "entropy": 0.6174094621092081,
      "epoch": 1.5394333557598863,
      "grad_norm": 0.15292319655418396,
      "learning_rate": 4.8688931824454876e-05,
      "loss": 0.6042,
      "mean_token_accuracy": 0.8154423534870148,
      "num_tokens": 534015193.0,
      "step": 16732
    },
    {
      "entropy": 0.5825012158602476,
      "epoch": 1.5395253621313276,
      "grad_norm": 0.15807582437992096,
      "learning_rate": 4.8685864998313244e-05,
      "loss": 0.569,
      "mean_token_accuracy": 0.8221127986907959,
      "num_tokens": 534045589.0,
      "step": 16733
    },
    {
      "entropy": 0.5902309473603964,
      "epoch": 1.539617368502769,
      "grad_norm": 0.15253396332263947,
      "learning_rate": 4.8682798172171626e-05,
      "loss": 0.5855,
      "mean_token_accuracy": 0.8182416744530201,
      "num_tokens": 534077600.0,
      "step": 16734
    },
    {
      "entropy": 0.6256960267201066,
      "epoch": 1.5397093748742101,
      "grad_norm": 0.15555043518543243,
      "learning_rate": 4.8679731346029993e-05,
      "loss": 0.6136,
      "mean_token_accuracy": 0.8140337914228439,
      "num_tokens": 534109516.0,
      "step": 16735
    },
    {
      "entropy": 0.5505384718999267,
      "epoch": 1.5398013812456512,
      "grad_norm": 0.15151281654834747,
      "learning_rate": 4.867666451988837e-05,
      "loss": 0.5472,
      "mean_token_accuracy": 0.8348408490419388,
      "num_tokens": 534141994.0,
      "step": 16736
    },
    {
      "entropy": 0.7445137146860361,
      "epoch": 1.5398933876170924,
      "grad_norm": 0.16529931128025055,
      "learning_rate": 4.867359769374674e-05,
      "loss": 0.7275,
      "mean_token_accuracy": 0.7729001827538013,
      "num_tokens": 534173481.0,
      "step": 16737
    },
    {
      "entropy": 0.6346455421298742,
      "epoch": 1.5399853939885337,
      "grad_norm": 0.1555176079273224,
      "learning_rate": 4.867053086760512e-05,
      "loss": 0.639,
      "mean_token_accuracy": 0.8053391575813293,
      "num_tokens": 534205448.0,
      "step": 16738
    },
    {
      "entropy": 0.6570374118164182,
      "epoch": 1.540077400359975,
      "grad_norm": 0.15335869789123535,
      "learning_rate": 4.866746404146349e-05,
      "loss": 0.6419,
      "mean_token_accuracy": 0.8038530685007572,
      "num_tokens": 534237123.0,
      "step": 16739
    },
    {
      "entropy": 0.5655712448060513,
      "epoch": 1.5401694067314162,
      "grad_norm": 0.15329095721244812,
      "learning_rate": 4.866439721532187e-05,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8274016156792641,
      "num_tokens": 534269015.0,
      "step": 16740
    },
    {
      "entropy": 0.6264520976692438,
      "epoch": 1.5402614131028574,
      "grad_norm": 0.14915122091770172,
      "learning_rate": 4.8661330389180235e-05,
      "loss": 0.6105,
      "mean_token_accuracy": 0.8122941292822361,
      "num_tokens": 534301100.0,
      "step": 16741
    },
    {
      "entropy": 0.6121608717367053,
      "epoch": 1.5403534194742985,
      "grad_norm": 0.15226009488105774,
      "learning_rate": 4.8658263563038616e-05,
      "loss": 0.5913,
      "mean_token_accuracy": 0.8156581409275532,
      "num_tokens": 534332973.0,
      "step": 16742
    },
    {
      "entropy": 0.619570467621088,
      "epoch": 1.5404454258457398,
      "grad_norm": 0.151201531291008,
      "learning_rate": 4.8655196736896984e-05,
      "loss": 0.6076,
      "mean_token_accuracy": 0.8173029385507107,
      "num_tokens": 534365137.0,
      "step": 16743
    },
    {
      "entropy": 0.6122921188361943,
      "epoch": 1.5405374322171812,
      "grad_norm": 0.15922486782073975,
      "learning_rate": 4.8652129910755365e-05,
      "loss": 0.6077,
      "mean_token_accuracy": 0.8164732679724693,
      "num_tokens": 534397650.0,
      "step": 16744
    },
    {
      "entropy": 0.6475800219923258,
      "epoch": 1.5406294385886223,
      "grad_norm": 0.15864650905132294,
      "learning_rate": 4.864906308461373e-05,
      "loss": 0.6587,
      "mean_token_accuracy": 0.7987233772873878,
      "num_tokens": 534429625.0,
      "step": 16745
    },
    {
      "entropy": 0.570534098893404,
      "epoch": 1.5407214449600635,
      "grad_norm": 0.15436910092830658,
      "learning_rate": 4.864599625847211e-05,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8278074264526367,
      "num_tokens": 534461402.0,
      "step": 16746
    },
    {
      "entropy": 0.5727490540593863,
      "epoch": 1.5408134513315046,
      "grad_norm": 0.15910908579826355,
      "learning_rate": 4.864292943233049e-05,
      "loss": 0.5556,
      "mean_token_accuracy": 0.829381138086319,
      "num_tokens": 534493327.0,
      "step": 16747
    },
    {
      "entropy": 0.48341842368245125,
      "epoch": 1.540905457702946,
      "grad_norm": 0.14316074550151825,
      "learning_rate": 4.863986260618886e-05,
      "loss": 0.4635,
      "mean_token_accuracy": 0.8571001030504704,
      "num_tokens": 534525902.0,
      "step": 16748
    },
    {
      "entropy": 0.6708500161767006,
      "epoch": 1.5409974640743873,
      "grad_norm": 0.1678410768508911,
      "learning_rate": 4.863679578004723e-05,
      "loss": 0.6559,
      "mean_token_accuracy": 0.7989755123853683,
      "num_tokens": 534556841.0,
      "step": 16749
    },
    {
      "entropy": 0.6399373356252909,
      "epoch": 1.5410894704458284,
      "grad_norm": 0.1599031388759613,
      "learning_rate": 4.8633728953905606e-05,
      "loss": 0.6413,
      "mean_token_accuracy": 0.8024808391928673,
      "num_tokens": 534588377.0,
      "step": 16750
    },
    {
      "entropy": 0.6321983300149441,
      "epoch": 1.5411814768172696,
      "grad_norm": 0.15907342731952667,
      "learning_rate": 4.863066212776398e-05,
      "loss": 0.6346,
      "mean_token_accuracy": 0.806833740323782,
      "num_tokens": 534620896.0,
      "step": 16751
    },
    {
      "entropy": 0.5740562891587615,
      "epoch": 1.5412734831887107,
      "grad_norm": 0.15766964852809906,
      "learning_rate": 4.8627595301622356e-05,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8258088268339634,
      "num_tokens": 534652787.0,
      "step": 16752
    },
    {
      "entropy": 0.6591412723064423,
      "epoch": 1.541365489560152,
      "grad_norm": 0.15944482386112213,
      "learning_rate": 4.862452847548073e-05,
      "loss": 0.6512,
      "mean_token_accuracy": 0.801820881664753,
      "num_tokens": 534685054.0,
      "step": 16753
    },
    {
      "entropy": 0.6574253551661968,
      "epoch": 1.5414574959315934,
      "grad_norm": 0.16448861360549927,
      "learning_rate": 4.86214616493391e-05,
      "loss": 0.6382,
      "mean_token_accuracy": 0.809797577559948,
      "num_tokens": 534716679.0,
      "step": 16754
    },
    {
      "entropy": 0.6455120192840695,
      "epoch": 1.5415495023030346,
      "grad_norm": 0.15502282977104187,
      "learning_rate": 4.861839482319748e-05,
      "loss": 0.6404,
      "mean_token_accuracy": 0.8043806180357933,
      "num_tokens": 534749039.0,
      "step": 16755
    },
    {
      "entropy": 0.5970958294346929,
      "epoch": 1.5416415086744757,
      "grad_norm": 0.15370602905750275,
      "learning_rate": 4.861532799705585e-05,
      "loss": 0.5879,
      "mean_token_accuracy": 0.81724464148283,
      "num_tokens": 534780416.0,
      "step": 16756
    },
    {
      "entropy": 0.6060113133862615,
      "epoch": 1.5417335150459168,
      "grad_norm": 0.15851424634456635,
      "learning_rate": 4.861226117091422e-05,
      "loss": 0.5917,
      "mean_token_accuracy": 0.8172442354261875,
      "num_tokens": 534812621.0,
      "step": 16757
    },
    {
      "entropy": 0.5955435205250978,
      "epoch": 1.5418255214173582,
      "grad_norm": 0.1869523823261261,
      "learning_rate": 4.86091943447726e-05,
      "loss": 0.5726,
      "mean_token_accuracy": 0.8225115612149239,
      "num_tokens": 534844118.0,
      "step": 16758
    },
    {
      "entropy": 0.6174914967268705,
      "epoch": 1.5419175277887995,
      "grad_norm": 0.15322615206241608,
      "learning_rate": 4.860612751863097e-05,
      "loss": 0.6166,
      "mean_token_accuracy": 0.813122846186161,
      "num_tokens": 534876459.0,
      "step": 16759
    },
    {
      "entropy": 0.6921174339950085,
      "epoch": 1.5420095341602407,
      "grad_norm": 0.156356543302536,
      "learning_rate": 4.8603060692489346e-05,
      "loss": 0.682,
      "mean_token_accuracy": 0.7913378924131393,
      "num_tokens": 534908453.0,
      "step": 16760
    },
    {
      "entropy": 0.6870741229504347,
      "epoch": 1.5421015405316818,
      "grad_norm": 0.16409918665885925,
      "learning_rate": 4.859999386634772e-05,
      "loss": 0.6862,
      "mean_token_accuracy": 0.792221937328577,
      "num_tokens": 534939918.0,
      "step": 16761
    },
    {
      "entropy": 0.568550038151443,
      "epoch": 1.542193546903123,
      "grad_norm": 0.15143725275993347,
      "learning_rate": 4.859692704020609e-05,
      "loss": 0.5742,
      "mean_token_accuracy": 0.8296327218413353,
      "num_tokens": 534972049.0,
      "step": 16762
    },
    {
      "entropy": 0.6584862973541021,
      "epoch": 1.5422855532745643,
      "grad_norm": 0.16120454668998718,
      "learning_rate": 4.859386021406447e-05,
      "loss": 0.653,
      "mean_token_accuracy": 0.7998167760670185,
      "num_tokens": 535003606.0,
      "step": 16763
    },
    {
      "entropy": 0.7736897096037865,
      "epoch": 1.5423775596460056,
      "grad_norm": 0.16607722640037537,
      "learning_rate": 4.859079338792284e-05,
      "loss": 0.7632,
      "mean_token_accuracy": 0.769744511693716,
      "num_tokens": 535035728.0,
      "step": 16764
    },
    {
      "entropy": 0.6544139590114355,
      "epoch": 1.5424695660174468,
      "grad_norm": 0.16111204028129578,
      "learning_rate": 4.858772656178122e-05,
      "loss": 0.6325,
      "mean_token_accuracy": 0.806391429156065,
      "num_tokens": 535067307.0,
      "step": 16765
    },
    {
      "entropy": 0.6113275098614395,
      "epoch": 1.542561572388888,
      "grad_norm": 0.15343402326107025,
      "learning_rate": 4.858465973563959e-05,
      "loss": 0.5988,
      "mean_token_accuracy": 0.8149918913841248,
      "num_tokens": 535099265.0,
      "step": 16766
    },
    {
      "entropy": 0.6337373126298189,
      "epoch": 1.542653578760329,
      "grad_norm": 0.15868373215198517,
      "learning_rate": 4.858159290949796e-05,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8131214939057827,
      "num_tokens": 535131490.0,
      "step": 16767
    },
    {
      "entropy": 0.6554018557071686,
      "epoch": 1.5427455851317704,
      "grad_norm": 0.16215084493160248,
      "learning_rate": 4.8578526083356336e-05,
      "loss": 0.6329,
      "mean_token_accuracy": 0.8051623702049255,
      "num_tokens": 535162486.0,
      "step": 16768
    },
    {
      "entropy": 0.6926757954061031,
      "epoch": 1.5428375915032118,
      "grad_norm": 0.16013532876968384,
      "learning_rate": 4.857545925721471e-05,
      "loss": 0.6626,
      "mean_token_accuracy": 0.7960186004638672,
      "num_tokens": 535194100.0,
      "step": 16769
    },
    {
      "entropy": 0.5855578081682324,
      "epoch": 1.542929597874653,
      "grad_norm": 0.15347842872142792,
      "learning_rate": 4.8572392431073086e-05,
      "loss": 0.5766,
      "mean_token_accuracy": 0.8226808346807957,
      "num_tokens": 535226360.0,
      "step": 16770
    },
    {
      "entropy": 0.6494536907412112,
      "epoch": 1.543021604246094,
      "grad_norm": 0.16376857459545135,
      "learning_rate": 4.856932560493146e-05,
      "loss": 0.6279,
      "mean_token_accuracy": 0.8073702119290829,
      "num_tokens": 535258325.0,
      "step": 16771
    },
    {
      "entropy": 0.7570474799722433,
      "epoch": 1.5431136106175352,
      "grad_norm": 0.1651514768600464,
      "learning_rate": 4.856625877878983e-05,
      "loss": 0.742,
      "mean_token_accuracy": 0.7769633196294308,
      "num_tokens": 535290845.0,
      "step": 16772
    },
    {
      "entropy": 0.5959161892533302,
      "epoch": 1.5432056169889765,
      "grad_norm": 0.14994193613529205,
      "learning_rate": 4.856319195264821e-05,
      "loss": 0.5887,
      "mean_token_accuracy": 0.821856252849102,
      "num_tokens": 535323437.0,
      "step": 16773
    },
    {
      "entropy": 0.6113578816875815,
      "epoch": 1.5432976233604179,
      "grad_norm": 0.16380999982357025,
      "learning_rate": 4.856012512650658e-05,
      "loss": 0.5973,
      "mean_token_accuracy": 0.8132656514644623,
      "num_tokens": 535354712.0,
      "step": 16774
    },
    {
      "entropy": 0.5230695977807045,
      "epoch": 1.543389629731859,
      "grad_norm": 0.15371191501617432,
      "learning_rate": 4.855705830036495e-05,
      "loss": 0.5125,
      "mean_token_accuracy": 0.8388974033296108,
      "num_tokens": 535386941.0,
      "step": 16775
    },
    {
      "entropy": 0.668524419888854,
      "epoch": 1.5434816361033001,
      "grad_norm": 0.1604067087173462,
      "learning_rate": 4.855399147422333e-05,
      "loss": 0.6543,
      "mean_token_accuracy": 0.8053231835365295,
      "num_tokens": 535419307.0,
      "step": 16776
    },
    {
      "entropy": 0.5782194435596466,
      "epoch": 1.5435736424747413,
      "grad_norm": 0.15529532730579376,
      "learning_rate": 4.85509246480817e-05,
      "loss": 0.575,
      "mean_token_accuracy": 0.8218185231089592,
      "num_tokens": 535450255.0,
      "step": 16777
    },
    {
      "entropy": 0.5416859146207571,
      "epoch": 1.5436656488461826,
      "grad_norm": 0.15018592774868011,
      "learning_rate": 4.8547857821940076e-05,
      "loss": 0.5365,
      "mean_token_accuracy": 0.8344463482499123,
      "num_tokens": 535482273.0,
      "step": 16778
    },
    {
      "entropy": 0.6087373988702893,
      "epoch": 1.543757655217624,
      "grad_norm": 0.16249412298202515,
      "learning_rate": 4.854479099579845e-05,
      "loss": 0.613,
      "mean_token_accuracy": 0.815627995878458,
      "num_tokens": 535514705.0,
      "step": 16779
    },
    {
      "entropy": 0.6630628034472466,
      "epoch": 1.5438496615890651,
      "grad_norm": 0.1633557677268982,
      "learning_rate": 4.8541724169656825e-05,
      "loss": 0.658,
      "mean_token_accuracy": 0.8046206384897232,
      "num_tokens": 535547134.0,
      "step": 16780
    },
    {
      "entropy": 0.579534194432199,
      "epoch": 1.5439416679605062,
      "grad_norm": 0.1537521928548813,
      "learning_rate": 4.85386573435152e-05,
      "loss": 0.5711,
      "mean_token_accuracy": 0.8231966383755207,
      "num_tokens": 535579452.0,
      "step": 16781
    },
    {
      "entropy": 0.6800615414977074,
      "epoch": 1.5440336743319474,
      "grad_norm": 0.16268180310726166,
      "learning_rate": 4.8535590517373575e-05,
      "loss": 0.6703,
      "mean_token_accuracy": 0.7985139712691307,
      "num_tokens": 535610904.0,
      "step": 16782
    },
    {
      "entropy": 0.5972783258184791,
      "epoch": 1.5441256807033887,
      "grad_norm": 0.15644283592700958,
      "learning_rate": 4.853252369123194e-05,
      "loss": 0.5843,
      "mean_token_accuracy": 0.8209734037518501,
      "num_tokens": 535642694.0,
      "step": 16783
    },
    {
      "entropy": 0.5527463648468256,
      "epoch": 1.54421768707483,
      "grad_norm": 0.14751186966896057,
      "learning_rate": 4.8529456865090324e-05,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8332393877208233,
      "num_tokens": 535674504.0,
      "step": 16784
    },
    {
      "entropy": 0.5368259819224477,
      "epoch": 1.5443096934462712,
      "grad_norm": 0.1487436592578888,
      "learning_rate": 4.852639003894869e-05,
      "loss": 0.527,
      "mean_token_accuracy": 0.837278887629509,
      "num_tokens": 535706477.0,
      "step": 16785
    },
    {
      "entropy": 0.5151106994599104,
      "epoch": 1.5444016998177124,
      "grad_norm": 0.14716751873493195,
      "learning_rate": 4.852332321280707e-05,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8404967151582241,
      "num_tokens": 535738550.0,
      "step": 16786
    },
    {
      "entropy": 0.6262980215251446,
      "epoch": 1.5444937061891535,
      "grad_norm": 0.15129999816417694,
      "learning_rate": 4.852025638666544e-05,
      "loss": 0.6063,
      "mean_token_accuracy": 0.8136518523097038,
      "num_tokens": 535770615.0,
      "step": 16787
    },
    {
      "entropy": 0.6476047448813915,
      "epoch": 1.5445857125605948,
      "grad_norm": 0.15740659832954407,
      "learning_rate": 4.8517189560523816e-05,
      "loss": 0.6542,
      "mean_token_accuracy": 0.8000140152871609,
      "num_tokens": 535802883.0,
      "step": 16788
    },
    {
      "entropy": 0.693029260262847,
      "epoch": 1.5446777189320362,
      "grad_norm": 0.16197864711284637,
      "learning_rate": 4.851412273438219e-05,
      "loss": 0.6736,
      "mean_token_accuracy": 0.7999381944537163,
      "num_tokens": 535834647.0,
      "step": 16789
    },
    {
      "entropy": 0.5495074018836021,
      "epoch": 1.5447697253034773,
      "grad_norm": 0.15485426783561707,
      "learning_rate": 4.8511055908240565e-05,
      "loss": 0.5345,
      "mean_token_accuracy": 0.8352390825748444,
      "num_tokens": 535866485.0,
      "step": 16790
    },
    {
      "entropy": 0.659773469902575,
      "epoch": 1.5448617316749185,
      "grad_norm": 0.15280675888061523,
      "learning_rate": 4.850798908209894e-05,
      "loss": 0.6511,
      "mean_token_accuracy": 0.8005794398486614,
      "num_tokens": 535898347.0,
      "step": 16791
    },
    {
      "entropy": 0.666874710470438,
      "epoch": 1.5449537380463596,
      "grad_norm": 0.16204412281513214,
      "learning_rate": 4.8504922255957314e-05,
      "loss": 0.6543,
      "mean_token_accuracy": 0.8058662749826908,
      "num_tokens": 535930422.0,
      "step": 16792
    },
    {
      "entropy": 0.5994910802692175,
      "epoch": 1.545045744417801,
      "grad_norm": 0.1627541333436966,
      "learning_rate": 4.850185542981568e-05,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8161672838032246,
      "num_tokens": 535962154.0,
      "step": 16793
    },
    {
      "entropy": 0.6316513903439045,
      "epoch": 1.5451377507892423,
      "grad_norm": 0.15755629539489746,
      "learning_rate": 4.8498788603674064e-05,
      "loss": 0.6156,
      "mean_token_accuracy": 0.8101803883910179,
      "num_tokens": 535994098.0,
      "step": 16794
    },
    {
      "entropy": 0.6117746368981898,
      "epoch": 1.5452297571606834,
      "grad_norm": 0.15590347349643707,
      "learning_rate": 4.849572177753243e-05,
      "loss": 0.5967,
      "mean_token_accuracy": 0.8173241354525089,
      "num_tokens": 536026866.0,
      "step": 16795
    },
    {
      "entropy": 0.5135190784931183,
      "epoch": 1.5453217635321246,
      "grad_norm": 0.14487324655056,
      "learning_rate": 4.8492654951390806e-05,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8499728664755821,
      "num_tokens": 536058972.0,
      "step": 16796
    },
    {
      "entropy": 0.6321863452903926,
      "epoch": 1.5454137699035657,
      "grad_norm": 0.15465782582759857,
      "learning_rate": 4.848958812524918e-05,
      "loss": 0.6236,
      "mean_token_accuracy": 0.8088383264839649,
      "num_tokens": 536090976.0,
      "step": 16797
    },
    {
      "entropy": 0.6003154087811708,
      "epoch": 1.545505776275007,
      "grad_norm": 0.15336744487285614,
      "learning_rate": 4.8486521299107556e-05,
      "loss": 0.59,
      "mean_token_accuracy": 0.8177503906190395,
      "num_tokens": 536123282.0,
      "step": 16798
    },
    {
      "entropy": 0.5297454334795475,
      "epoch": 1.5455977826464484,
      "grad_norm": 0.15395021438598633,
      "learning_rate": 4.848345447296593e-05,
      "loss": 0.5232,
      "mean_token_accuracy": 0.8381331861019135,
      "num_tokens": 536155230.0,
      "step": 16799
    },
    {
      "entropy": 0.5782006774097681,
      "epoch": 1.5456897890178896,
      "grad_norm": 0.149706169962883,
      "learning_rate": 4.8480387646824305e-05,
      "loss": 0.5745,
      "mean_token_accuracy": 0.8222797550261021,
      "num_tokens": 536187490.0,
      "step": 16800
    },
    {
      "entropy": 0.6139357052743435,
      "epoch": 1.5457817953893307,
      "grad_norm": 0.1474168449640274,
      "learning_rate": 4.847732082068267e-05,
      "loss": 0.6108,
      "mean_token_accuracy": 0.814196340739727,
      "num_tokens": 536220217.0,
      "step": 16801
    },
    {
      "entropy": 0.5689513124525547,
      "epoch": 1.5458738017607718,
      "grad_norm": 0.15521526336669922,
      "learning_rate": 4.8474253994541054e-05,
      "loss": 0.5597,
      "mean_token_accuracy": 0.8255514837801456,
      "num_tokens": 536251597.0,
      "step": 16802
    },
    {
      "entropy": 0.6071568606421351,
      "epoch": 1.5459658081322132,
      "grad_norm": 0.1487666666507721,
      "learning_rate": 4.847118716839942e-05,
      "loss": 0.5957,
      "mean_token_accuracy": 0.8193725198507309,
      "num_tokens": 536284108.0,
      "step": 16803
    },
    {
      "entropy": 0.6812950782477856,
      "epoch": 1.5460578145036545,
      "grad_norm": 0.16095878183841705,
      "learning_rate": 4.8468120342257803e-05,
      "loss": 0.6717,
      "mean_token_accuracy": 0.7978796251118183,
      "num_tokens": 536316187.0,
      "step": 16804
    },
    {
      "entropy": 0.6744360933080316,
      "epoch": 1.5461498208750957,
      "grad_norm": 0.15495403110980988,
      "learning_rate": 4.846505351611617e-05,
      "loss": 0.665,
      "mean_token_accuracy": 0.798248540610075,
      "num_tokens": 536348107.0,
      "step": 16805
    },
    {
      "entropy": 0.5662409961223602,
      "epoch": 1.5462418272465368,
      "grad_norm": 0.15348820388317108,
      "learning_rate": 4.8461986689974546e-05,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8253890052437782,
      "num_tokens": 536380033.0,
      "step": 16806
    },
    {
      "entropy": 0.6291577517986298,
      "epoch": 1.546333833617978,
      "grad_norm": 0.1551407277584076,
      "learning_rate": 4.845891986383292e-05,
      "loss": 0.6182,
      "mean_token_accuracy": 0.8102274909615517,
      "num_tokens": 536412033.0,
      "step": 16807
    },
    {
      "entropy": 0.5692737996578217,
      "epoch": 1.5464258399894193,
      "grad_norm": 0.15206874907016754,
      "learning_rate": 4.8455853037691295e-05,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8296345248818398,
      "num_tokens": 536443335.0,
      "step": 16808
    },
    {
      "entropy": 0.7248868513852358,
      "epoch": 1.5465178463608606,
      "grad_norm": 0.16511540114879608,
      "learning_rate": 4.845278621154967e-05,
      "loss": 0.7133,
      "mean_token_accuracy": 0.7804566472768784,
      "num_tokens": 536475569.0,
      "step": 16809
    },
    {
      "entropy": 0.7204847084358335,
      "epoch": 1.5466098527323018,
      "grad_norm": 0.16061857342720032,
      "learning_rate": 4.8449719385408045e-05,
      "loss": 0.7032,
      "mean_token_accuracy": 0.7890807837247849,
      "num_tokens": 536507365.0,
      "step": 16810
    },
    {
      "entropy": 0.5822309916839004,
      "epoch": 1.546701859103743,
      "grad_norm": 0.1537160724401474,
      "learning_rate": 4.844665255926642e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.8235328383743763,
      "num_tokens": 536540094.0,
      "step": 16811
    },
    {
      "entropy": 0.6538620665669441,
      "epoch": 1.546793865475184,
      "grad_norm": 0.16697680950164795,
      "learning_rate": 4.8443585733124794e-05,
      "loss": 0.6313,
      "mean_token_accuracy": 0.8034873567521572,
      "num_tokens": 536571702.0,
      "step": 16812
    },
    {
      "entropy": 0.6470622420310974,
      "epoch": 1.5468858718466254,
      "grad_norm": 0.1601923108100891,
      "learning_rate": 4.844051890698317e-05,
      "loss": 0.6465,
      "mean_token_accuracy": 0.802482184022665,
      "num_tokens": 536603933.0,
      "step": 16813
    },
    {
      "entropy": 0.6622630106285214,
      "epoch": 1.5469778782180668,
      "grad_norm": 0.15955324470996857,
      "learning_rate": 4.8437452080841536e-05,
      "loss": 0.6562,
      "mean_token_accuracy": 0.8015442490577698,
      "num_tokens": 536636219.0,
      "step": 16814
    },
    {
      "entropy": 0.6948969289660454,
      "epoch": 1.547069884589508,
      "grad_norm": 0.1548943668603897,
      "learning_rate": 4.843438525469992e-05,
      "loss": 0.6878,
      "mean_token_accuracy": 0.7912449762225151,
      "num_tokens": 536668356.0,
      "step": 16815
    },
    {
      "entropy": 0.5460700010880828,
      "epoch": 1.547161890960949,
      "grad_norm": 0.15257692337036133,
      "learning_rate": 4.8431318428558286e-05,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8320209197700024,
      "num_tokens": 536700742.0,
      "step": 16816
    },
    {
      "entropy": 0.6482355492189527,
      "epoch": 1.5472538973323902,
      "grad_norm": 0.16417697072029114,
      "learning_rate": 4.842825160241666e-05,
      "loss": 0.6511,
      "mean_token_accuracy": 0.8034920580685139,
      "num_tokens": 536732967.0,
      "step": 16817
    },
    {
      "entropy": 0.5323441196233034,
      "epoch": 1.5473459037038315,
      "grad_norm": 0.14783106744289398,
      "learning_rate": 4.8425184776275035e-05,
      "loss": 0.522,
      "mean_token_accuracy": 0.8392868898808956,
      "num_tokens": 536765488.0,
      "step": 16818
    },
    {
      "entropy": 0.6858398457989097,
      "epoch": 1.5474379100752729,
      "grad_norm": 0.16312627494335175,
      "learning_rate": 4.842211795013341e-05,
      "loss": 0.6728,
      "mean_token_accuracy": 0.7969440594315529,
      "num_tokens": 536797202.0,
      "step": 16819
    },
    {
      "entropy": 0.5851907962933183,
      "epoch": 1.547529916446714,
      "grad_norm": 0.16368696093559265,
      "learning_rate": 4.8419051123991784e-05,
      "loss": 0.5695,
      "mean_token_accuracy": 0.8242566585540771,
      "num_tokens": 536829516.0,
      "step": 16820
    },
    {
      "entropy": 0.6237015817314386,
      "epoch": 1.5476219228181551,
      "grad_norm": 0.15786288678646088,
      "learning_rate": 4.841598429785016e-05,
      "loss": 0.6155,
      "mean_token_accuracy": 0.8143219761550426,
      "num_tokens": 536861593.0,
      "step": 16821
    },
    {
      "entropy": 0.5894962102174759,
      "epoch": 1.5477139291895963,
      "grad_norm": 0.14935220777988434,
      "learning_rate": 4.841291747170853e-05,
      "loss": 0.5708,
      "mean_token_accuracy": 0.8265173360705376,
      "num_tokens": 536893627.0,
      "step": 16822
    },
    {
      "entropy": 0.6574267810210586,
      "epoch": 1.5478059355610376,
      "grad_norm": 0.16323791444301605,
      "learning_rate": 4.840985064556691e-05,
      "loss": 0.6441,
      "mean_token_accuracy": 0.8042137064039707,
      "num_tokens": 536924841.0,
      "step": 16823
    },
    {
      "entropy": 0.6363742873072624,
      "epoch": 1.547897941932479,
      "grad_norm": 0.15786823630332947,
      "learning_rate": 4.8406783819425276e-05,
      "loss": 0.6304,
      "mean_token_accuracy": 0.8093893490731716,
      "num_tokens": 536957321.0,
      "step": 16824
    },
    {
      "entropy": 0.6214093843009323,
      "epoch": 1.5479899483039201,
      "grad_norm": 0.1562998741865158,
      "learning_rate": 4.840371699328366e-05,
      "loss": 0.6116,
      "mean_token_accuracy": 0.8112667798995972,
      "num_tokens": 536989657.0,
      "step": 16825
    },
    {
      "entropy": 0.659534128382802,
      "epoch": 1.5480819546753612,
      "grad_norm": 0.1533999741077423,
      "learning_rate": 4.8400650167142025e-05,
      "loss": 0.6451,
      "mean_token_accuracy": 0.802195530384779,
      "num_tokens": 537022123.0,
      "step": 16826
    },
    {
      "entropy": 0.6165014812722802,
      "epoch": 1.5481739610468024,
      "grad_norm": 0.1666567623615265,
      "learning_rate": 4.83975833410004e-05,
      "loss": 0.5996,
      "mean_token_accuracy": 0.8140166886150837,
      "num_tokens": 537053259.0,
      "step": 16827
    },
    {
      "entropy": 0.6214941572397947,
      "epoch": 1.5482659674182437,
      "grad_norm": 0.16542531549930573,
      "learning_rate": 4.8394516514858775e-05,
      "loss": 0.6334,
      "mean_token_accuracy": 0.8087045438587666,
      "num_tokens": 537085344.0,
      "step": 16828
    },
    {
      "entropy": 0.6386388307437301,
      "epoch": 1.548357973789685,
      "grad_norm": 0.15848198533058167,
      "learning_rate": 4.839144968871715e-05,
      "loss": 0.6428,
      "mean_token_accuracy": 0.8024182692170143,
      "num_tokens": 537117509.0,
      "step": 16829
    },
    {
      "entropy": 0.5661832606419921,
      "epoch": 1.5484499801611262,
      "grad_norm": 0.15416598320007324,
      "learning_rate": 4.8388382862575524e-05,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8246203698217869,
      "num_tokens": 537150103.0,
      "step": 16830
    },
    {
      "entropy": 0.635637553408742,
      "epoch": 1.5485419865325674,
      "grad_norm": 0.153608039021492,
      "learning_rate": 4.83853160364339e-05,
      "loss": 0.6284,
      "mean_token_accuracy": 0.8115742914378643,
      "num_tokens": 537182081.0,
      "step": 16831
    },
    {
      "entropy": 0.5923128193244338,
      "epoch": 1.5486339929040085,
      "grad_norm": 0.1605375111103058,
      "learning_rate": 4.8382249210292266e-05,
      "loss": 0.581,
      "mean_token_accuracy": 0.8235440477728844,
      "num_tokens": 537213468.0,
      "step": 16832
    },
    {
      "entropy": 0.6715577617287636,
      "epoch": 1.5487259992754498,
      "grad_norm": 0.15939553081989288,
      "learning_rate": 4.837918238415065e-05,
      "loss": 0.667,
      "mean_token_accuracy": 0.7942696325480938,
      "num_tokens": 537245398.0,
      "step": 16833
    },
    {
      "entropy": 0.6362502556294203,
      "epoch": 1.5488180056468912,
      "grad_norm": 0.1630130410194397,
      "learning_rate": 4.8376115558009016e-05,
      "loss": 0.6323,
      "mean_token_accuracy": 0.8085350506007671,
      "num_tokens": 537276995.0,
      "step": 16834
    },
    {
      "entropy": 0.5130029688589275,
      "epoch": 1.5489100120183323,
      "grad_norm": 0.1423085629940033,
      "learning_rate": 4.837304873186739e-05,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8467982448637486,
      "num_tokens": 537308996.0,
      "step": 16835
    },
    {
      "entropy": 0.6140719540417194,
      "epoch": 1.5490020183897735,
      "grad_norm": 0.1584002673625946,
      "learning_rate": 4.8369981905725765e-05,
      "loss": 0.6021,
      "mean_token_accuracy": 0.8156212493777275,
      "num_tokens": 537340943.0,
      "step": 16836
    },
    {
      "entropy": 0.5803517419844866,
      "epoch": 1.5490940247612146,
      "grad_norm": 0.1527986377477646,
      "learning_rate": 4.836691507958414e-05,
      "loss": 0.5634,
      "mean_token_accuracy": 0.8270045183598995,
      "num_tokens": 537373270.0,
      "step": 16837
    },
    {
      "entropy": 0.6494263876229525,
      "epoch": 1.549186031132656,
      "grad_norm": 0.17034024000167847,
      "learning_rate": 4.8363848253442514e-05,
      "loss": 0.6467,
      "mean_token_accuracy": 0.7990933582186699,
      "num_tokens": 537405178.0,
      "step": 16838
    },
    {
      "entropy": 0.6618201122619212,
      "epoch": 1.5492780375040973,
      "grad_norm": 0.1617821455001831,
      "learning_rate": 4.836078142730089e-05,
      "loss": 0.6438,
      "mean_token_accuracy": 0.8030790574848652,
      "num_tokens": 537436920.0,
      "step": 16839
    },
    {
      "entropy": 0.6023155692964792,
      "epoch": 1.5493700438755384,
      "grad_norm": 0.15267875790596008,
      "learning_rate": 4.8357714601159264e-05,
      "loss": 0.5925,
      "mean_token_accuracy": 0.8177425861358643,
      "num_tokens": 537469539.0,
      "step": 16840
    },
    {
      "entropy": 0.7327085975557566,
      "epoch": 1.5494620502469796,
      "grad_norm": 0.17069244384765625,
      "learning_rate": 4.835464777501764e-05,
      "loss": 0.7239,
      "mean_token_accuracy": 0.7773448936641216,
      "num_tokens": 537500817.0,
      "step": 16841
    },
    {
      "entropy": 0.55307746399194,
      "epoch": 1.5495540566184207,
      "grad_norm": 0.1471729725599289,
      "learning_rate": 4.835158094887601e-05,
      "loss": 0.5277,
      "mean_token_accuracy": 0.8329284563660622,
      "num_tokens": 537532831.0,
      "step": 16842
    },
    {
      "entropy": 0.6247746497392654,
      "epoch": 1.549646062989862,
      "grad_norm": 0.15509004890918732,
      "learning_rate": 4.834851412273438e-05,
      "loss": 0.609,
      "mean_token_accuracy": 0.8119172230362892,
      "num_tokens": 537564639.0,
      "step": 16843
    },
    {
      "entropy": 0.6811153255403042,
      "epoch": 1.5497380693613034,
      "grad_norm": 0.15862908959388733,
      "learning_rate": 4.834544729659276e-05,
      "loss": 0.6664,
      "mean_token_accuracy": 0.7922000773251057,
      "num_tokens": 537596478.0,
      "step": 16844
    },
    {
      "entropy": 0.68638267647475,
      "epoch": 1.5498300757327446,
      "grad_norm": 0.15888994932174683,
      "learning_rate": 4.834238047045113e-05,
      "loss": 0.6723,
      "mean_token_accuracy": 0.7937124781310558,
      "num_tokens": 537628731.0,
      "step": 16845
    },
    {
      "entropy": 0.5645057898946106,
      "epoch": 1.5499220821041857,
      "grad_norm": 0.16197660565376282,
      "learning_rate": 4.833931364430951e-05,
      "loss": 0.56,
      "mean_token_accuracy": 0.8227940462529659,
      "num_tokens": 537660244.0,
      "step": 16846
    },
    {
      "entropy": 0.6301847184076905,
      "epoch": 1.5500140884756268,
      "grad_norm": 0.15198445320129395,
      "learning_rate": 4.833624681816788e-05,
      "loss": 0.6138,
      "mean_token_accuracy": 0.8146571293473244,
      "num_tokens": 537692777.0,
      "step": 16847
    },
    {
      "entropy": 0.6606928333640099,
      "epoch": 1.5501060948470682,
      "grad_norm": 0.16163310408592224,
      "learning_rate": 4.8333179992026254e-05,
      "loss": 0.6539,
      "mean_token_accuracy": 0.8012738488614559,
      "num_tokens": 537725481.0,
      "step": 16848
    },
    {
      "entropy": 0.5277027487754822,
      "epoch": 1.5501981012185095,
      "grad_norm": 0.14655032753944397,
      "learning_rate": 4.833011316588463e-05,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8391853906214237,
      "num_tokens": 537758249.0,
      "step": 16849
    },
    {
      "entropy": 0.5911285746842623,
      "epoch": 1.5502901075899507,
      "grad_norm": 0.15247659385204315,
      "learning_rate": 4.8327046339743e-05,
      "loss": 0.583,
      "mean_token_accuracy": 0.8206344582140446,
      "num_tokens": 537790756.0,
      "step": 16850
    },
    {
      "entropy": 0.5259783510118723,
      "epoch": 1.5503821139613918,
      "grad_norm": 0.15374867618083954,
      "learning_rate": 4.832397951360138e-05,
      "loss": 0.5057,
      "mean_token_accuracy": 0.8434191979467869,
      "num_tokens": 537823208.0,
      "step": 16851
    },
    {
      "entropy": 0.6595614496618509,
      "epoch": 1.550474120332833,
      "grad_norm": 0.15791191160678864,
      "learning_rate": 4.832091268745975e-05,
      "loss": 0.6583,
      "mean_token_accuracy": 0.8002524189651012,
      "num_tokens": 537855835.0,
      "step": 16852
    },
    {
      "entropy": 0.665057210251689,
      "epoch": 1.5505661267042743,
      "grad_norm": 0.15904657542705536,
      "learning_rate": 4.831784586131812e-05,
      "loss": 0.6529,
      "mean_token_accuracy": 0.8033314421772957,
      "num_tokens": 537888057.0,
      "step": 16853
    },
    {
      "entropy": 0.5799215100705624,
      "epoch": 1.5506581330757157,
      "grad_norm": 0.1494247168302536,
      "learning_rate": 4.83147790351765e-05,
      "loss": 0.5776,
      "mean_token_accuracy": 0.8224682733416557,
      "num_tokens": 537920275.0,
      "step": 16854
    },
    {
      "entropy": 0.6199096124619246,
      "epoch": 1.5507501394471568,
      "grad_norm": 0.15554822981357574,
      "learning_rate": 4.831171220903487e-05,
      "loss": 0.6052,
      "mean_token_accuracy": 0.8109720721840858,
      "num_tokens": 537952596.0,
      "step": 16855
    },
    {
      "entropy": 0.7496019285172224,
      "epoch": 1.550842145818598,
      "grad_norm": 0.16332003474235535,
      "learning_rate": 4.8308645382893244e-05,
      "loss": 0.7452,
      "mean_token_accuracy": 0.7720877192914486,
      "num_tokens": 537985107.0,
      "step": 16856
    },
    {
      "entropy": 0.6790919303894043,
      "epoch": 1.550934152190039,
      "grad_norm": 0.1698521226644516,
      "learning_rate": 4.830557855675162e-05,
      "loss": 0.6667,
      "mean_token_accuracy": 0.7941385358572006,
      "num_tokens": 538015761.0,
      "step": 16857
    },
    {
      "entropy": 0.5770970177836716,
      "epoch": 1.5510261585614804,
      "grad_norm": 0.15687572956085205,
      "learning_rate": 4.8302511730609994e-05,
      "loss": 0.5589,
      "mean_token_accuracy": 0.8297050558030605,
      "num_tokens": 538047574.0,
      "step": 16858
    },
    {
      "entropy": 0.6022099000401795,
      "epoch": 1.5511181649329218,
      "grad_norm": 0.15588989853858948,
      "learning_rate": 4.829944490446837e-05,
      "loss": 0.5929,
      "mean_token_accuracy": 0.8206201791763306,
      "num_tokens": 538080114.0,
      "step": 16859
    },
    {
      "entropy": 0.5093274209648371,
      "epoch": 1.551210171304363,
      "grad_norm": 0.14278340339660645,
      "learning_rate": 4.829637807832674e-05,
      "loss": 0.4888,
      "mean_token_accuracy": 0.8461382240056992,
      "num_tokens": 538112683.0,
      "step": 16860
    },
    {
      "entropy": 0.6227012183517218,
      "epoch": 1.551302177675804,
      "grad_norm": 0.15579666197299957,
      "learning_rate": 4.829331125218511e-05,
      "loss": 0.6079,
      "mean_token_accuracy": 0.8117585331201553,
      "num_tokens": 538144857.0,
      "step": 16861
    },
    {
      "entropy": 0.6680541117675602,
      "epoch": 1.5513941840472452,
      "grad_norm": 0.15733572840690613,
      "learning_rate": 4.829024442604349e-05,
      "loss": 0.6668,
      "mean_token_accuracy": 0.7947811782360077,
      "num_tokens": 538177519.0,
      "step": 16862
    },
    {
      "entropy": 0.5317955417558551,
      "epoch": 1.5514861904186865,
      "grad_norm": 0.14927519857883453,
      "learning_rate": 4.828717759990186e-05,
      "loss": 0.5198,
      "mean_token_accuracy": 0.8369297347962856,
      "num_tokens": 538209512.0,
      "step": 16863
    },
    {
      "entropy": 0.6333469729870558,
      "epoch": 1.5515781967901279,
      "grad_norm": 0.15648780763149261,
      "learning_rate": 4.828411077376024e-05,
      "loss": 0.6146,
      "mean_token_accuracy": 0.8129834085702896,
      "num_tokens": 538241564.0,
      "step": 16864
    },
    {
      "entropy": 0.5147939566522837,
      "epoch": 1.551670203161569,
      "grad_norm": 0.14319969713687897,
      "learning_rate": 4.828104394761861e-05,
      "loss": 0.4835,
      "mean_token_accuracy": 0.850363627076149,
      "num_tokens": 538273664.0,
      "step": 16865
    },
    {
      "entropy": 0.6922069825232029,
      "epoch": 1.5517622095330101,
      "grad_norm": 0.16067704558372498,
      "learning_rate": 4.8277977121476984e-05,
      "loss": 0.6786,
      "mean_token_accuracy": 0.7914809435606003,
      "num_tokens": 538304916.0,
      "step": 16866
    },
    {
      "entropy": 0.5656511094421148,
      "epoch": 1.5518542159044513,
      "grad_norm": 0.1516963094472885,
      "learning_rate": 4.827491029533536e-05,
      "loss": 0.5627,
      "mean_token_accuracy": 0.8297001719474792,
      "num_tokens": 538337460.0,
      "step": 16867
    },
    {
      "entropy": 0.49595236126333475,
      "epoch": 1.5519462222758926,
      "grad_norm": 0.14871612191200256,
      "learning_rate": 4.8271843469193733e-05,
      "loss": 0.4838,
      "mean_token_accuracy": 0.850448239594698,
      "num_tokens": 538369868.0,
      "step": 16868
    },
    {
      "entropy": 0.6157837454229593,
      "epoch": 1.552038228647334,
      "grad_norm": 0.161015123128891,
      "learning_rate": 4.826877664305211e-05,
      "loss": 0.5993,
      "mean_token_accuracy": 0.815193697810173,
      "num_tokens": 538401197.0,
      "step": 16869
    },
    {
      "entropy": 0.5946946078911424,
      "epoch": 1.5521302350187751,
      "grad_norm": 0.16067636013031006,
      "learning_rate": 4.826570981691048e-05,
      "loss": 0.585,
      "mean_token_accuracy": 0.816437978297472,
      "num_tokens": 538432851.0,
      "step": 16870
    },
    {
      "entropy": 0.5837146863341331,
      "epoch": 1.5522222413902163,
      "grad_norm": 0.1454564481973648,
      "learning_rate": 4.826264299076886e-05,
      "loss": 0.5764,
      "mean_token_accuracy": 0.8245887495577335,
      "num_tokens": 538464803.0,
      "step": 16871
    },
    {
      "entropy": 0.6548323947936296,
      "epoch": 1.5523142477616574,
      "grad_norm": 0.1658470183610916,
      "learning_rate": 4.825957616462723e-05,
      "loss": 0.6396,
      "mean_token_accuracy": 0.8048544749617577,
      "num_tokens": 538496892.0,
      "step": 16872
    },
    {
      "entropy": 0.587608315050602,
      "epoch": 1.5524062541330987,
      "grad_norm": 0.15833733975887299,
      "learning_rate": 4.825650933848561e-05,
      "loss": 0.5915,
      "mean_token_accuracy": 0.8162438347935677,
      "num_tokens": 538529172.0,
      "step": 16873
    },
    {
      "entropy": 0.6673035342246294,
      "epoch": 1.55249826050454,
      "grad_norm": 0.16323408484458923,
      "learning_rate": 4.8253442512343975e-05,
      "loss": 0.6496,
      "mean_token_accuracy": 0.8021915182471275,
      "num_tokens": 538560140.0,
      "step": 16874
    },
    {
      "entropy": 0.6460228161886334,
      "epoch": 1.5525902668759812,
      "grad_norm": 0.15921229124069214,
      "learning_rate": 4.8250375686202356e-05,
      "loss": 0.641,
      "mean_token_accuracy": 0.8079546578228474,
      "num_tokens": 538591286.0,
      "step": 16875
    },
    {
      "entropy": 0.6631326526403427,
      "epoch": 1.5526822732474224,
      "grad_norm": 0.17378142476081848,
      "learning_rate": 4.8247308860060724e-05,
      "loss": 0.6521,
      "mean_token_accuracy": 0.8020143769681454,
      "num_tokens": 538623570.0,
      "step": 16876
    },
    {
      "entropy": 0.5972441078629345,
      "epoch": 1.5527742796188635,
      "grad_norm": 0.15339308977127075,
      "learning_rate": 4.82442420339191e-05,
      "loss": 0.5803,
      "mean_token_accuracy": 0.8196616768836975,
      "num_tokens": 538655977.0,
      "step": 16877
    },
    {
      "entropy": 0.6572736166417599,
      "epoch": 1.5528662859903049,
      "grad_norm": 0.1566147357225418,
      "learning_rate": 4.824117520777747e-05,
      "loss": 0.6516,
      "mean_token_accuracy": 0.8005131185054779,
      "num_tokens": 538688383.0,
      "step": 16878
    },
    {
      "entropy": 0.5371394669637084,
      "epoch": 1.5529582923617462,
      "grad_norm": 0.15801699459552765,
      "learning_rate": 4.823810838163585e-05,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8385567553341389,
      "num_tokens": 538720900.0,
      "step": 16879
    },
    {
      "entropy": 0.6157824853435159,
      "epoch": 1.5530502987331873,
      "grad_norm": 0.15522253513336182,
      "learning_rate": 4.823504155549422e-05,
      "loss": 0.6068,
      "mean_token_accuracy": 0.8157833889126778,
      "num_tokens": 538752786.0,
      "step": 16880
    },
    {
      "entropy": 0.5932908244431019,
      "epoch": 1.5531423051046285,
      "grad_norm": 0.14812801778316498,
      "learning_rate": 4.82319747293526e-05,
      "loss": 0.5836,
      "mean_token_accuracy": 0.818306241184473,
      "num_tokens": 538784599.0,
      "step": 16881
    },
    {
      "entropy": 0.5153076080605388,
      "epoch": 1.5532343114760696,
      "grad_norm": 0.14479859173297882,
      "learning_rate": 4.8228907903210965e-05,
      "loss": 0.5099,
      "mean_token_accuracy": 0.8421944230794907,
      "num_tokens": 538816082.0,
      "step": 16882
    },
    {
      "entropy": 0.6365571646019816,
      "epoch": 1.553326317847511,
      "grad_norm": 0.15688970685005188,
      "learning_rate": 4.8225841077069346e-05,
      "loss": 0.621,
      "mean_token_accuracy": 0.8081857189536095,
      "num_tokens": 538847870.0,
      "step": 16883
    },
    {
      "entropy": 0.6378952022641897,
      "epoch": 1.5534183242189523,
      "grad_norm": 0.1616443544626236,
      "learning_rate": 4.8222774250927714e-05,
      "loss": 0.6292,
      "mean_token_accuracy": 0.8052861802279949,
      "num_tokens": 538879474.0,
      "step": 16884
    },
    {
      "entropy": 0.6003139261156321,
      "epoch": 1.5535103305903935,
      "grad_norm": 0.15850259363651276,
      "learning_rate": 4.8219707424786096e-05,
      "loss": 0.5897,
      "mean_token_accuracy": 0.8155240938067436,
      "num_tokens": 538911094.0,
      "step": 16885
    },
    {
      "entropy": 0.6961224945262074,
      "epoch": 1.5536023369618346,
      "grad_norm": 0.1626812368631363,
      "learning_rate": 4.8216640598644464e-05,
      "loss": 0.683,
      "mean_token_accuracy": 0.7927363216876984,
      "num_tokens": 538942948.0,
      "step": 16886
    },
    {
      "entropy": 0.6794290794059634,
      "epoch": 1.5536943433332757,
      "grad_norm": 0.157316654920578,
      "learning_rate": 4.821357377250284e-05,
      "loss": 0.676,
      "mean_token_accuracy": 0.7909890450537205,
      "num_tokens": 538975459.0,
      "step": 16887
    },
    {
      "entropy": 0.6777465976774693,
      "epoch": 1.553786349704717,
      "grad_norm": 0.1606663167476654,
      "learning_rate": 4.821050694636121e-05,
      "loss": 0.6749,
      "mean_token_accuracy": 0.796750120818615,
      "num_tokens": 539007306.0,
      "step": 16888
    },
    {
      "entropy": 0.6841382198035717,
      "epoch": 1.5538783560761584,
      "grad_norm": 0.1585787683725357,
      "learning_rate": 4.820744012021959e-05,
      "loss": 0.6727,
      "mean_token_accuracy": 0.7939280122518539,
      "num_tokens": 539039435.0,
      "step": 16889
    },
    {
      "entropy": 0.6011803415603936,
      "epoch": 1.5539703624475996,
      "grad_norm": 0.15528878569602966,
      "learning_rate": 4.820437329407796e-05,
      "loss": 0.5933,
      "mean_token_accuracy": 0.8194537311792374,
      "num_tokens": 539071578.0,
      "step": 16890
    },
    {
      "entropy": 0.5937186991795897,
      "epoch": 1.5540623688190407,
      "grad_norm": 0.15199430286884308,
      "learning_rate": 4.820130646793634e-05,
      "loss": 0.5979,
      "mean_token_accuracy": 0.8156440556049347,
      "num_tokens": 539103965.0,
      "step": 16891
    },
    {
      "entropy": 0.6380165619775653,
      "epoch": 1.5541543751904818,
      "grad_norm": 0.1688891351222992,
      "learning_rate": 4.8198239641794705e-05,
      "loss": 0.6317,
      "mean_token_accuracy": 0.810836311429739,
      "num_tokens": 539134771.0,
      "step": 16892
    },
    {
      "entropy": 0.6020946158096194,
      "epoch": 1.5542463815619232,
      "grad_norm": 0.15554018318653107,
      "learning_rate": 4.8195172815653086e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8196677640080452,
      "num_tokens": 539166749.0,
      "step": 16893
    },
    {
      "entropy": 0.5701466351747513,
      "epoch": 1.5543383879333645,
      "grad_norm": 0.15199102461338043,
      "learning_rate": 4.8192105989511454e-05,
      "loss": 0.568,
      "mean_token_accuracy": 0.8284638747572899,
      "num_tokens": 539198020.0,
      "step": 16894
    },
    {
      "entropy": 0.5933902207762003,
      "epoch": 1.5544303943048057,
      "grad_norm": 0.15352103114128113,
      "learning_rate": 4.818903916336983e-05,
      "loss": 0.5723,
      "mean_token_accuracy": 0.8232559971511364,
      "num_tokens": 539229675.0,
      "step": 16895
    },
    {
      "entropy": 0.5734577081166208,
      "epoch": 1.5545224006762468,
      "grad_norm": 0.14768488705158234,
      "learning_rate": 4.81859723372282e-05,
      "loss": 0.54,
      "mean_token_accuracy": 0.8320793621242046,
      "num_tokens": 539260872.0,
      "step": 16896
    },
    {
      "entropy": 0.5772641450166702,
      "epoch": 1.554614407047688,
      "grad_norm": 0.1469101905822754,
      "learning_rate": 4.818290551108658e-05,
      "loss": 0.5654,
      "mean_token_accuracy": 0.8235838674008846,
      "num_tokens": 539293225.0,
      "step": 16897
    },
    {
      "entropy": 0.6103631462901831,
      "epoch": 1.5547064134191293,
      "grad_norm": 0.16125811636447906,
      "learning_rate": 4.817983868494495e-05,
      "loss": 0.5976,
      "mean_token_accuracy": 0.8157299757003784,
      "num_tokens": 539325559.0,
      "step": 16898
    },
    {
      "entropy": 0.5137110520154238,
      "epoch": 1.5547984197905707,
      "grad_norm": 0.1434779167175293,
      "learning_rate": 4.817677185880333e-05,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8475112654268742,
      "num_tokens": 539357911.0,
      "step": 16899
    },
    {
      "entropy": 0.644060842692852,
      "epoch": 1.5548904261620118,
      "grad_norm": 0.16057027876377106,
      "learning_rate": 4.81737050326617e-05,
      "loss": 0.6401,
      "mean_token_accuracy": 0.8049417585134506,
      "num_tokens": 539389908.0,
      "step": 16900
    },
    {
      "entropy": 0.5917244292795658,
      "epoch": 1.554982432533453,
      "grad_norm": 0.16030116379261017,
      "learning_rate": 4.8170638206520076e-05,
      "loss": 0.5811,
      "mean_token_accuracy": 0.8206113837659359,
      "num_tokens": 539422147.0,
      "step": 16901
    },
    {
      "entropy": 0.5694777946919203,
      "epoch": 1.555074438904894,
      "grad_norm": 0.15259237587451935,
      "learning_rate": 4.816757138037845e-05,
      "loss": 0.5723,
      "mean_token_accuracy": 0.8224456310272217,
      "num_tokens": 539454351.0,
      "step": 16902
    },
    {
      "entropy": 0.5939163770526648,
      "epoch": 1.5551664452763354,
      "grad_norm": 0.15310390293598175,
      "learning_rate": 4.816450455423682e-05,
      "loss": 0.5864,
      "mean_token_accuracy": 0.823145117610693,
      "num_tokens": 539486955.0,
      "step": 16903
    },
    {
      "entropy": 0.6289293505251408,
      "epoch": 1.5552584516477768,
      "grad_norm": 0.16152192652225494,
      "learning_rate": 4.81614377280952e-05,
      "loss": 0.6264,
      "mean_token_accuracy": 0.8069568909704685,
      "num_tokens": 539519125.0,
      "step": 16904
    },
    {
      "entropy": 0.5800776481628418,
      "epoch": 1.555350458019218,
      "grad_norm": 0.14782051742076874,
      "learning_rate": 4.815837090195357e-05,
      "loss": 0.5698,
      "mean_token_accuracy": 0.822857242077589,
      "num_tokens": 539551802.0,
      "step": 16905
    },
    {
      "entropy": 0.6908879624679685,
      "epoch": 1.555442464390659,
      "grad_norm": 0.16188687086105347,
      "learning_rate": 4.815530407581195e-05,
      "loss": 0.672,
      "mean_token_accuracy": 0.792552325874567,
      "num_tokens": 539582597.0,
      "step": 16906
    },
    {
      "entropy": 0.5923152072355151,
      "epoch": 1.5555344707621002,
      "grad_norm": 0.15687160193920135,
      "learning_rate": 4.815223724967032e-05,
      "loss": 0.5854,
      "mean_token_accuracy": 0.8218693397939205,
      "num_tokens": 539614784.0,
      "step": 16907
    },
    {
      "entropy": 0.6976561695337296,
      "epoch": 1.5556264771335415,
      "grad_norm": 0.16640256345272064,
      "learning_rate": 4.814917042352869e-05,
      "loss": 0.685,
      "mean_token_accuracy": 0.7876560688018799,
      "num_tokens": 539644926.0,
      "step": 16908
    },
    {
      "entropy": 0.5172362178564072,
      "epoch": 1.5557184835049829,
      "grad_norm": 0.1484460085630417,
      "learning_rate": 4.814610359738707e-05,
      "loss": 0.5099,
      "mean_token_accuracy": 0.8421775661408901,
      "num_tokens": 539677694.0,
      "step": 16909
    },
    {
      "entropy": 0.5533578526228666,
      "epoch": 1.555810489876424,
      "grad_norm": 0.150816410779953,
      "learning_rate": 4.814303677124544e-05,
      "loss": 0.5394,
      "mean_token_accuracy": 0.8317542970180511,
      "num_tokens": 539710199.0,
      "step": 16910
    },
    {
      "entropy": 0.683261239901185,
      "epoch": 1.5559024962478651,
      "grad_norm": 0.16506719589233398,
      "learning_rate": 4.8139969945103816e-05,
      "loss": 0.6764,
      "mean_token_accuracy": 0.7966607995331287,
      "num_tokens": 539741791.0,
      "step": 16911
    },
    {
      "entropy": 0.5576768405735493,
      "epoch": 1.5559945026193063,
      "grad_norm": 0.15128439664840698,
      "learning_rate": 4.813690311896219e-05,
      "loss": 0.5404,
      "mean_token_accuracy": 0.8331075869500637,
      "num_tokens": 539773566.0,
      "step": 16912
    },
    {
      "entropy": 0.5435215379111469,
      "epoch": 1.5560865089907476,
      "grad_norm": 0.15370842814445496,
      "learning_rate": 4.813383629282056e-05,
      "loss": 0.5432,
      "mean_token_accuracy": 0.8378221988677979,
      "num_tokens": 539806003.0,
      "step": 16913
    },
    {
      "entropy": 0.6220966535620391,
      "epoch": 1.556178515362189,
      "grad_norm": 0.16302277147769928,
      "learning_rate": 4.813076946667894e-05,
      "loss": 0.6153,
      "mean_token_accuracy": 0.8109432868659496,
      "num_tokens": 539837685.0,
      "step": 16914
    },
    {
      "entropy": 0.6529379952698946,
      "epoch": 1.5562705217336301,
      "grad_norm": 0.15670105814933777,
      "learning_rate": 4.812770264053731e-05,
      "loss": 0.6545,
      "mean_token_accuracy": 0.7983526065945625,
      "num_tokens": 539870305.0,
      "step": 16915
    },
    {
      "entropy": 0.5159652750007808,
      "epoch": 1.5563625281050713,
      "grad_norm": 0.14363963901996613,
      "learning_rate": 4.812463581439568e-05,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8464633896946907,
      "num_tokens": 539902675.0,
      "step": 16916
    },
    {
      "entropy": 0.5712733371183276,
      "epoch": 1.5564545344765124,
      "grad_norm": 0.16101042926311493,
      "learning_rate": 4.812156898825406e-05,
      "loss": 0.5671,
      "mean_token_accuracy": 0.8284684903919697,
      "num_tokens": 539935443.0,
      "step": 16917
    },
    {
      "entropy": 0.5768107771873474,
      "epoch": 1.5565465408479537,
      "grad_norm": 0.15416984260082245,
      "learning_rate": 4.811850216211243e-05,
      "loss": 0.5579,
      "mean_token_accuracy": 0.8293735012412071,
      "num_tokens": 539967719.0,
      "step": 16918
    },
    {
      "entropy": 0.4392927587032318,
      "epoch": 1.556638547219395,
      "grad_norm": 0.1417098194360733,
      "learning_rate": 4.8115435335970807e-05,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8655703291296959,
      "num_tokens": 539999869.0,
      "step": 16919
    },
    {
      "entropy": 0.6477336958050728,
      "epoch": 1.5567305535908362,
      "grad_norm": 0.15602360665798187,
      "learning_rate": 4.811236850982918e-05,
      "loss": 0.6448,
      "mean_token_accuracy": 0.80296990275383,
      "num_tokens": 540032303.0,
      "step": 16920
    },
    {
      "entropy": 0.7245255894958973,
      "epoch": 1.5568225599622774,
      "grad_norm": 0.16118276119232178,
      "learning_rate": 4.810930168368755e-05,
      "loss": 0.7072,
      "mean_token_accuracy": 0.7828141115605831,
      "num_tokens": 540063862.0,
      "step": 16921
    },
    {
      "entropy": 0.5576057899743319,
      "epoch": 1.5569145663337185,
      "grad_norm": 0.15240895748138428,
      "learning_rate": 4.810623485754593e-05,
      "loss": 0.5525,
      "mean_token_accuracy": 0.8283314779400826,
      "num_tokens": 540095514.0,
      "step": 16922
    },
    {
      "entropy": 0.6399878868833184,
      "epoch": 1.5570065727051599,
      "grad_norm": 0.15514957904815674,
      "learning_rate": 4.81031680314043e-05,
      "loss": 0.6305,
      "mean_token_accuracy": 0.8065376318991184,
      "num_tokens": 540127205.0,
      "step": 16923
    },
    {
      "entropy": 0.6305287480354309,
      "epoch": 1.5570985790766012,
      "grad_norm": 0.16388455033302307,
      "learning_rate": 4.810010120526268e-05,
      "loss": 0.6237,
      "mean_token_accuracy": 0.8058769255876541,
      "num_tokens": 540158497.0,
      "step": 16924
    },
    {
      "entropy": 0.6161862015724182,
      "epoch": 1.5571905854480423,
      "grad_norm": 0.15283627808094025,
      "learning_rate": 4.809703437912105e-05,
      "loss": 0.608,
      "mean_token_accuracy": 0.81242411211133,
      "num_tokens": 540190900.0,
      "step": 16925
    },
    {
      "entropy": 0.5845544226467609,
      "epoch": 1.5572825918194835,
      "grad_norm": 0.14517754316329956,
      "learning_rate": 4.809396755297942e-05,
      "loss": 0.5676,
      "mean_token_accuracy": 0.821904432028532,
      "num_tokens": 540222854.0,
      "step": 16926
    },
    {
      "entropy": 0.7421532925218344,
      "epoch": 1.5573745981909246,
      "grad_norm": 0.16291582584381104,
      "learning_rate": 4.80909007268378e-05,
      "loss": 0.7342,
      "mean_token_accuracy": 0.7801032476127148,
      "num_tokens": 540254646.0,
      "step": 16927
    },
    {
      "entropy": 0.6176619837060571,
      "epoch": 1.557466604562366,
      "grad_norm": 0.15616144239902496,
      "learning_rate": 4.808783390069617e-05,
      "loss": 0.6099,
      "mean_token_accuracy": 0.8158208467066288,
      "num_tokens": 540286942.0,
      "step": 16928
    },
    {
      "entropy": 0.5366966966539621,
      "epoch": 1.5575586109338073,
      "grad_norm": 0.14572343230247498,
      "learning_rate": 4.8084767074554546e-05,
      "loss": 0.5255,
      "mean_token_accuracy": 0.8340786509215832,
      "num_tokens": 540319279.0,
      "step": 16929
    },
    {
      "entropy": 0.6315982006490231,
      "epoch": 1.5576506173052485,
      "grad_norm": 0.1543375551700592,
      "learning_rate": 4.808170024841292e-05,
      "loss": 0.6161,
      "mean_token_accuracy": 0.8091884180903435,
      "num_tokens": 540350992.0,
      "step": 16930
    },
    {
      "entropy": 0.5859789568930864,
      "epoch": 1.5577426236766896,
      "grad_norm": 0.15722833573818207,
      "learning_rate": 4.8078633422271296e-05,
      "loss": 0.5803,
      "mean_token_accuracy": 0.8213418498635292,
      "num_tokens": 540381803.0,
      "step": 16931
    },
    {
      "entropy": 0.5567561071366072,
      "epoch": 1.5578346300481307,
      "grad_norm": 0.14514385163784027,
      "learning_rate": 4.807556659612967e-05,
      "loss": 0.5426,
      "mean_token_accuracy": 0.8284836448729038,
      "num_tokens": 540413633.0,
      "step": 16932
    },
    {
      "entropy": 0.6554358433932066,
      "epoch": 1.557926636419572,
      "grad_norm": 0.15729032456874847,
      "learning_rate": 4.8072499769988045e-05,
      "loss": 0.6401,
      "mean_token_accuracy": 0.8068714514374733,
      "num_tokens": 540445098.0,
      "step": 16933
    },
    {
      "entropy": 0.6073678452521563,
      "epoch": 1.5580186427910134,
      "grad_norm": 0.1554742157459259,
      "learning_rate": 4.806943294384641e-05,
      "loss": 0.5904,
      "mean_token_accuracy": 0.8158774487674236,
      "num_tokens": 540476862.0,
      "step": 16934
    },
    {
      "entropy": 0.5435816338285804,
      "epoch": 1.5581106491624546,
      "grad_norm": 0.1504540592432022,
      "learning_rate": 4.8066366117704794e-05,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8334810920059681,
      "num_tokens": 540508517.0,
      "step": 16935
    },
    {
      "entropy": 0.5715521760284901,
      "epoch": 1.5582026555338957,
      "grad_norm": 0.1476559042930603,
      "learning_rate": 4.806329929156316e-05,
      "loss": 0.5646,
      "mean_token_accuracy": 0.8313102200627327,
      "num_tokens": 540540311.0,
      "step": 16936
    },
    {
      "entropy": 0.5730502677615732,
      "epoch": 1.5582946619053368,
      "grad_norm": 0.15444420278072357,
      "learning_rate": 4.806023246542154e-05,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8278621397912502,
      "num_tokens": 540571982.0,
      "step": 16937
    },
    {
      "entropy": 0.6735983975231647,
      "epoch": 1.5583866682767782,
      "grad_norm": 0.16258183121681213,
      "learning_rate": 4.805716563927991e-05,
      "loss": 0.6714,
      "mean_token_accuracy": 0.79395791888237,
      "num_tokens": 540603824.0,
      "step": 16938
    },
    {
      "entropy": 0.7214549779891968,
      "epoch": 1.5584786746482195,
      "grad_norm": 0.16686661541461945,
      "learning_rate": 4.8054098813138286e-05,
      "loss": 0.7238,
      "mean_token_accuracy": 0.7847379371523857,
      "num_tokens": 540635345.0,
      "step": 16939
    },
    {
      "entropy": 0.4968668036162853,
      "epoch": 1.5585706810196607,
      "grad_norm": 0.16049762070178986,
      "learning_rate": 4.805103198699666e-05,
      "loss": 0.4948,
      "mean_token_accuracy": 0.851196400821209,
      "num_tokens": 540667325.0,
      "step": 16940
    },
    {
      "entropy": 0.7226730631664395,
      "epoch": 1.5586626873911018,
      "grad_norm": 0.17221318185329437,
      "learning_rate": 4.8047965160855035e-05,
      "loss": 0.7072,
      "mean_token_accuracy": 0.7866709157824516,
      "num_tokens": 540698096.0,
      "step": 16941
    },
    {
      "entropy": 0.6244318764656782,
      "epoch": 1.558754693762543,
      "grad_norm": 0.15884089469909668,
      "learning_rate": 4.80448983347134e-05,
      "loss": 0.6028,
      "mean_token_accuracy": 0.8156474232673645,
      "num_tokens": 540729446.0,
      "step": 16942
    },
    {
      "entropy": 0.5782786216586828,
      "epoch": 1.5588467001339843,
      "grad_norm": 0.17593148350715637,
      "learning_rate": 4.8041831508571785e-05,
      "loss": 0.5928,
      "mean_token_accuracy": 0.8197923004627228,
      "num_tokens": 540761615.0,
      "step": 16943
    },
    {
      "entropy": 0.5410898909904063,
      "epoch": 1.5589387065054257,
      "grad_norm": 0.1517399400472641,
      "learning_rate": 4.803876468243015e-05,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8363790065050125,
      "num_tokens": 540793137.0,
      "step": 16944
    },
    {
      "entropy": 0.7166554220020771,
      "epoch": 1.5590307128768668,
      "grad_norm": 0.16189883649349213,
      "learning_rate": 4.8035697856288534e-05,
      "loss": 0.6998,
      "mean_token_accuracy": 0.7879813574254513,
      "num_tokens": 540825389.0,
      "step": 16945
    },
    {
      "entropy": 0.6693460047245026,
      "epoch": 1.559122719248308,
      "grad_norm": 0.16171319782733917,
      "learning_rate": 4.80326310301469e-05,
      "loss": 0.647,
      "mean_token_accuracy": 0.80380579829216,
      "num_tokens": 540857045.0,
      "step": 16946
    },
    {
      "entropy": 0.5597273670136929,
      "epoch": 1.559214725619749,
      "grad_norm": 0.15388505160808563,
      "learning_rate": 4.8029564204005276e-05,
      "loss": 0.5402,
      "mean_token_accuracy": 0.8296188153326511,
      "num_tokens": 540888790.0,
      "step": 16947
    },
    {
      "entropy": 0.682404775172472,
      "epoch": 1.5593067319911904,
      "grad_norm": 0.1653805524110794,
      "learning_rate": 4.802649737786365e-05,
      "loss": 0.6542,
      "mean_token_accuracy": 0.8015361241996288,
      "num_tokens": 540920812.0,
      "step": 16948
    },
    {
      "entropy": 0.6760833505541086,
      "epoch": 1.5593987383626318,
      "grad_norm": 0.16029949486255646,
      "learning_rate": 4.8023430551722026e-05,
      "loss": 0.6642,
      "mean_token_accuracy": 0.796030081808567,
      "num_tokens": 540952572.0,
      "step": 16949
    },
    {
      "entropy": 0.6154549084603786,
      "epoch": 1.559490744734073,
      "grad_norm": 0.16042305529117584,
      "learning_rate": 4.80203637255804e-05,
      "loss": 0.6127,
      "mean_token_accuracy": 0.8149422891438007,
      "num_tokens": 540984236.0,
      "step": 16950
    },
    {
      "entropy": 0.48595955641940236,
      "epoch": 1.559582751105514,
      "grad_norm": 0.14668792486190796,
      "learning_rate": 4.8017296899438775e-05,
      "loss": 0.4736,
      "mean_token_accuracy": 0.8543429225683212,
      "num_tokens": 541016939.0,
      "step": 16951
    },
    {
      "entropy": 0.6991585027426481,
      "epoch": 1.5596747574769552,
      "grad_norm": 0.165376216173172,
      "learning_rate": 4.801423007329714e-05,
      "loss": 0.6814,
      "mean_token_accuracy": 0.790705855935812,
      "num_tokens": 541047993.0,
      "step": 16952
    },
    {
      "entropy": 0.549565676599741,
      "epoch": 1.5597667638483965,
      "grad_norm": 0.1449407935142517,
      "learning_rate": 4.8011163247155524e-05,
      "loss": 0.5337,
      "mean_token_accuracy": 0.837771050632,
      "num_tokens": 541080253.0,
      "step": 16953
    },
    {
      "entropy": 0.6389753497205675,
      "epoch": 1.5598587702198379,
      "grad_norm": 0.15576206147670746,
      "learning_rate": 4.800809642101389e-05,
      "loss": 0.6325,
      "mean_token_accuracy": 0.8051219433546066,
      "num_tokens": 541112498.0,
      "step": 16954
    },
    {
      "entropy": 0.6639365516602993,
      "epoch": 1.559950776591279,
      "grad_norm": 0.1622207760810852,
      "learning_rate": 4.800502959487227e-05,
      "loss": 0.6612,
      "mean_token_accuracy": 0.8035777173936367,
      "num_tokens": 541144493.0,
      "step": 16955
    },
    {
      "entropy": 0.6297060237266123,
      "epoch": 1.5600427829627201,
      "grad_norm": 0.1541309505701065,
      "learning_rate": 4.800196276873064e-05,
      "loss": 0.6077,
      "mean_token_accuracy": 0.8136062957346439,
      "num_tokens": 541175191.0,
      "step": 16956
    },
    {
      "entropy": 0.5609583184123039,
      "epoch": 1.5601347893341613,
      "grad_norm": 0.14880666136741638,
      "learning_rate": 4.7998895942589016e-05,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8302407339215279,
      "num_tokens": 541207703.0,
      "step": 16957
    },
    {
      "entropy": 0.581012349575758,
      "epoch": 1.5602267957056026,
      "grad_norm": 0.15402856469154358,
      "learning_rate": 4.799582911644739e-05,
      "loss": 0.5657,
      "mean_token_accuracy": 0.8239794261753559,
      "num_tokens": 541239840.0,
      "step": 16958
    },
    {
      "entropy": 0.5565020591020584,
      "epoch": 1.560318802077044,
      "grad_norm": 0.15275241434574127,
      "learning_rate": 4.7992762290305765e-05,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8284539058804512,
      "num_tokens": 541272221.0,
      "step": 16959
    },
    {
      "entropy": 0.5828147940337658,
      "epoch": 1.5604108084484851,
      "grad_norm": 0.15590135753154755,
      "learning_rate": 4.798969546416413e-05,
      "loss": 0.5837,
      "mean_token_accuracy": 0.8209983073174953,
      "num_tokens": 541304229.0,
      "step": 16960
    },
    {
      "entropy": 0.7314867321401834,
      "epoch": 1.5605028148199263,
      "grad_norm": 0.16686542332172394,
      "learning_rate": 4.7986628638022515e-05,
      "loss": 0.7215,
      "mean_token_accuracy": 0.7835241705179214,
      "num_tokens": 541336119.0,
      "step": 16961
    },
    {
      "entropy": 0.6683014901354909,
      "epoch": 1.5605948211913674,
      "grad_norm": 0.1696888655424118,
      "learning_rate": 4.798356181188089e-05,
      "loss": 0.6713,
      "mean_token_accuracy": 0.796862430870533,
      "num_tokens": 541367947.0,
      "step": 16962
    },
    {
      "entropy": 0.6943355482071638,
      "epoch": 1.5606868275628087,
      "grad_norm": 0.15739108622074127,
      "learning_rate": 4.798049498573926e-05,
      "loss": 0.6763,
      "mean_token_accuracy": 0.7984101176261902,
      "num_tokens": 541399927.0,
      "step": 16963
    },
    {
      "entropy": 0.622891865670681,
      "epoch": 1.56077883393425,
      "grad_norm": 0.15337619185447693,
      "learning_rate": 4.797742815959764e-05,
      "loss": 0.6063,
      "mean_token_accuracy": 0.8108342550694942,
      "num_tokens": 541431831.0,
      "step": 16964
    },
    {
      "entropy": 0.550036771222949,
      "epoch": 1.5608708403056912,
      "grad_norm": 0.15335556864738464,
      "learning_rate": 4.7974361333456006e-05,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8326299637556076,
      "num_tokens": 541463714.0,
      "step": 16965
    },
    {
      "entropy": 0.5102520538493991,
      "epoch": 1.5609628466771324,
      "grad_norm": 0.15397238731384277,
      "learning_rate": 4.797129450731439e-05,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8441371731460094,
      "num_tokens": 541495110.0,
      "step": 16966
    },
    {
      "entropy": 0.5877146958373487,
      "epoch": 1.5610548530485735,
      "grad_norm": 0.14610733091831207,
      "learning_rate": 4.7968227681172756e-05,
      "loss": 0.5715,
      "mean_token_accuracy": 0.8237082660198212,
      "num_tokens": 541526802.0,
      "step": 16967
    },
    {
      "entropy": 0.600098142400384,
      "epoch": 1.5611468594200149,
      "grad_norm": 0.15208053588867188,
      "learning_rate": 4.796516085503113e-05,
      "loss": 0.5734,
      "mean_token_accuracy": 0.8191959485411644,
      "num_tokens": 541558742.0,
      "step": 16968
    },
    {
      "entropy": 0.7320534968748689,
      "epoch": 1.5612388657914562,
      "grad_norm": 0.1617090106010437,
      "learning_rate": 4.7962094028889505e-05,
      "loss": 0.7264,
      "mean_token_accuracy": 0.7787245362997055,
      "num_tokens": 541590898.0,
      "step": 16969
    },
    {
      "entropy": 0.46664432249963284,
      "epoch": 1.5613308721628973,
      "grad_norm": 0.13871192932128906,
      "learning_rate": 4.795902720274788e-05,
      "loss": 0.4584,
      "mean_token_accuracy": 0.8588743694126606,
      "num_tokens": 541623445.0,
      "step": 16970
    },
    {
      "entropy": 0.6012039298657328,
      "epoch": 1.5614228785343385,
      "grad_norm": 0.15433016419410706,
      "learning_rate": 4.7955960376606254e-05,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8173097483813763,
      "num_tokens": 541655486.0,
      "step": 16971
    },
    {
      "entropy": 0.6338136065751314,
      "epoch": 1.5615148849057796,
      "grad_norm": 0.16309301555156708,
      "learning_rate": 4.795289355046463e-05,
      "loss": 0.6166,
      "mean_token_accuracy": 0.810226958245039,
      "num_tokens": 541686984.0,
      "step": 16972
    },
    {
      "entropy": 0.6225014161318541,
      "epoch": 1.561606891277221,
      "grad_norm": 0.16101492941379547,
      "learning_rate": 4.7949826724323e-05,
      "loss": 0.6186,
      "mean_token_accuracy": 0.8130158558487892,
      "num_tokens": 541718875.0,
      "step": 16973
    },
    {
      "entropy": 0.6279719267040491,
      "epoch": 1.5616988976486623,
      "grad_norm": 0.1593695729970932,
      "learning_rate": 4.794675989818138e-05,
      "loss": 0.62,
      "mean_token_accuracy": 0.8084156364202499,
      "num_tokens": 541750349.0,
      "step": 16974
    },
    {
      "entropy": 0.6658581420779228,
      "epoch": 1.5617909040201035,
      "grad_norm": 0.15521983802318573,
      "learning_rate": 4.7943693072039746e-05,
      "loss": 0.6523,
      "mean_token_accuracy": 0.7981556877493858,
      "num_tokens": 541782328.0,
      "step": 16975
    },
    {
      "entropy": 0.6502876970916986,
      "epoch": 1.5618829103915446,
      "grad_norm": 0.15751022100448608,
      "learning_rate": 4.794062624589812e-05,
      "loss": 0.6473,
      "mean_token_accuracy": 0.8018810115754604,
      "num_tokens": 541813172.0,
      "step": 16976
    },
    {
      "entropy": 0.5806296057999134,
      "epoch": 1.5619749167629857,
      "grad_norm": 0.15908274054527283,
      "learning_rate": 4.7937559419756495e-05,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8263547606766224,
      "num_tokens": 541844985.0,
      "step": 16977
    },
    {
      "entropy": 0.6598385199904442,
      "epoch": 1.562066923134427,
      "grad_norm": 0.1556321382522583,
      "learning_rate": 4.793449259361487e-05,
      "loss": 0.6533,
      "mean_token_accuracy": 0.799910344183445,
      "num_tokens": 541876798.0,
      "step": 16978
    },
    {
      "entropy": 0.645426869392395,
      "epoch": 1.5621589295058684,
      "grad_norm": 0.15634100139141083,
      "learning_rate": 4.7931425767473245e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.8099141977727413,
      "num_tokens": 541908612.0,
      "step": 16979
    },
    {
      "entropy": 0.6404429078102112,
      "epoch": 1.5622509358773096,
      "grad_norm": 0.15531601011753082,
      "learning_rate": 4.792835894133162e-05,
      "loss": 0.615,
      "mean_token_accuracy": 0.8133598528802395,
      "num_tokens": 541940552.0,
      "step": 16980
    },
    {
      "entropy": 0.6277308650314808,
      "epoch": 1.5623429422487507,
      "grad_norm": 0.1581084281206131,
      "learning_rate": 4.792529211518999e-05,
      "loss": 0.6152,
      "mean_token_accuracy": 0.815336748957634,
      "num_tokens": 541973272.0,
      "step": 16981
    },
    {
      "entropy": 0.722930601797998,
      "epoch": 1.5624349486201918,
      "grad_norm": 0.16661110520362854,
      "learning_rate": 4.792222528904837e-05,
      "loss": 0.7184,
      "mean_token_accuracy": 0.7831715829670429,
      "num_tokens": 542005637.0,
      "step": 16982
    },
    {
      "entropy": 0.5350319519639015,
      "epoch": 1.5625269549916332,
      "grad_norm": 0.1550159454345703,
      "learning_rate": 4.7919158462906737e-05,
      "loss": 0.5207,
      "mean_token_accuracy": 0.8383871912956238,
      "num_tokens": 542038069.0,
      "step": 16983
    },
    {
      "entropy": 0.6964630875736475,
      "epoch": 1.5626189613630745,
      "grad_norm": 0.1629531979560852,
      "learning_rate": 4.791609163676512e-05,
      "loss": 0.7077,
      "mean_token_accuracy": 0.7824721969664097,
      "num_tokens": 542070455.0,
      "step": 16984
    },
    {
      "entropy": 0.626460294239223,
      "epoch": 1.5627109677345157,
      "grad_norm": 0.1540650874376297,
      "learning_rate": 4.7913024810623486e-05,
      "loss": 0.6172,
      "mean_token_accuracy": 0.8114516846835613,
      "num_tokens": 542102522.0,
      "step": 16985
    },
    {
      "entropy": 0.5070397555828094,
      "epoch": 1.5628029741059568,
      "grad_norm": 0.14546893537044525,
      "learning_rate": 4.790995798448186e-05,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8461498618125916,
      "num_tokens": 542134836.0,
      "step": 16986
    },
    {
      "entropy": 0.6814198354259133,
      "epoch": 1.562894980477398,
      "grad_norm": 0.16291625797748566,
      "learning_rate": 4.7906891158340235e-05,
      "loss": 0.6965,
      "mean_token_accuracy": 0.7919612899422646,
      "num_tokens": 542166939.0,
      "step": 16987
    },
    {
      "entropy": 0.6150445519015193,
      "epoch": 1.5629869868488393,
      "grad_norm": 0.1602942794561386,
      "learning_rate": 4.790382433219861e-05,
      "loss": 0.5989,
      "mean_token_accuracy": 0.8199897482991219,
      "num_tokens": 542198415.0,
      "step": 16988
    },
    {
      "entropy": 0.49378224182873964,
      "epoch": 1.5630789932202807,
      "grad_norm": 0.14925973117351532,
      "learning_rate": 4.7900757506056984e-05,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8471593856811523,
      "num_tokens": 542230706.0,
      "step": 16989
    },
    {
      "entropy": 0.5920986942946911,
      "epoch": 1.5631709995917218,
      "grad_norm": 0.1583532691001892,
      "learning_rate": 4.789769067991536e-05,
      "loss": 0.592,
      "mean_token_accuracy": 0.8177226409316063,
      "num_tokens": 542263162.0,
      "step": 16990
    },
    {
      "entropy": 0.5624576006084681,
      "epoch": 1.563263005963163,
      "grad_norm": 0.14883510768413544,
      "learning_rate": 4.7894623853773734e-05,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8251732774078846,
      "num_tokens": 542295633.0,
      "step": 16991
    },
    {
      "entropy": 0.5157265020534396,
      "epoch": 1.563355012334604,
      "grad_norm": 0.14795391261577606,
      "learning_rate": 4.789155702763211e-05,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8474596925079823,
      "num_tokens": 542328401.0,
      "step": 16992
    },
    {
      "entropy": 0.6274944320321083,
      "epoch": 1.5634470187060454,
      "grad_norm": 0.15427552163600922,
      "learning_rate": 4.788849020149048e-05,
      "loss": 0.616,
      "mean_token_accuracy": 0.8102590516209602,
      "num_tokens": 542360076.0,
      "step": 16993
    },
    {
      "entropy": 0.6042216308414936,
      "epoch": 1.5635390250774868,
      "grad_norm": 0.15263696014881134,
      "learning_rate": 4.788542337534885e-05,
      "loss": 0.5962,
      "mean_token_accuracy": 0.8141313940286636,
      "num_tokens": 542392423.0,
      "step": 16994
    },
    {
      "entropy": 0.6706614550203085,
      "epoch": 1.563631031448928,
      "grad_norm": 0.1623212695121765,
      "learning_rate": 4.788235654920723e-05,
      "loss": 0.6468,
      "mean_token_accuracy": 0.8013029098510742,
      "num_tokens": 542424035.0,
      "step": 16995
    },
    {
      "entropy": 0.7031190879642963,
      "epoch": 1.563723037820369,
      "grad_norm": 0.1602984517812729,
      "learning_rate": 4.78792897230656e-05,
      "loss": 0.6755,
      "mean_token_accuracy": 0.7916977405548096,
      "num_tokens": 542455622.0,
      "step": 16996
    },
    {
      "entropy": 0.5951339267194271,
      "epoch": 1.5638150441918102,
      "grad_norm": 0.15071704983711243,
      "learning_rate": 4.7876222896923975e-05,
      "loss": 0.5722,
      "mean_token_accuracy": 0.8223372027277946,
      "num_tokens": 542487412.0,
      "step": 16997
    },
    {
      "entropy": 0.6249010134488344,
      "epoch": 1.5639070505632515,
      "grad_norm": 0.15618260204792023,
      "learning_rate": 4.787315607078235e-05,
      "loss": 0.5942,
      "mean_token_accuracy": 0.8157787248492241,
      "num_tokens": 542517946.0,
      "step": 16998
    },
    {
      "entropy": 0.6377061195671558,
      "epoch": 1.5639990569346929,
      "grad_norm": 0.16675986349582672,
      "learning_rate": 4.7870089244640724e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.8117336519062519,
      "num_tokens": 542549834.0,
      "step": 16999
    },
    {
      "entropy": 0.669497162103653,
      "epoch": 1.564091063306134,
      "grad_norm": 0.15944138169288635,
      "learning_rate": 4.78670224184991e-05,
      "loss": 0.664,
      "mean_token_accuracy": 0.8002081029117107,
      "num_tokens": 542581929.0,
      "step": 17000
    },
    {
      "entropy": 0.7193190269172192,
      "epoch": 1.5641830696775751,
      "grad_norm": 0.16822193562984467,
      "learning_rate": 4.7863955592357473e-05,
      "loss": 0.7204,
      "mean_token_accuracy": 0.7852589078247547,
      "num_tokens": 542613174.0,
      "step": 17001
    },
    {
      "entropy": 0.5954258413985372,
      "epoch": 1.5642750760490163,
      "grad_norm": 0.15022358298301697,
      "learning_rate": 4.786088876621584e-05,
      "loss": 0.5701,
      "mean_token_accuracy": 0.8250363506376743,
      "num_tokens": 542644764.0,
      "step": 17002
    },
    {
      "entropy": 0.6625263411551714,
      "epoch": 1.5643670824204576,
      "grad_norm": 0.15676648914813995,
      "learning_rate": 4.785782194007422e-05,
      "loss": 0.6624,
      "mean_token_accuracy": 0.7957218959927559,
      "num_tokens": 542677247.0,
      "step": 17003
    },
    {
      "entropy": 0.4962779739871621,
      "epoch": 1.564459088791899,
      "grad_norm": 0.15532571077346802,
      "learning_rate": 4.785475511393259e-05,
      "loss": 0.4927,
      "mean_token_accuracy": 0.8456964194774628,
      "num_tokens": 542709766.0,
      "step": 17004
    },
    {
      "entropy": 0.4775468045845628,
      "epoch": 1.5645510951633401,
      "grad_norm": 0.149307519197464,
      "learning_rate": 4.785168828779097e-05,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8530677258968353,
      "num_tokens": 542741549.0,
      "step": 17005
    },
    {
      "entropy": 0.5978935975581408,
      "epoch": 1.5646431015347813,
      "grad_norm": 0.15256242454051971,
      "learning_rate": 4.784862146164934e-05,
      "loss": 0.588,
      "mean_token_accuracy": 0.824051670730114,
      "num_tokens": 542773653.0,
      "step": 17006
    },
    {
      "entropy": 0.6075341114774346,
      "epoch": 1.5647351079062224,
      "grad_norm": 0.1495390385389328,
      "learning_rate": 4.7845554635507715e-05,
      "loss": 0.5945,
      "mean_token_accuracy": 0.8193510808050632,
      "num_tokens": 542805722.0,
      "step": 17007
    },
    {
      "entropy": 0.6746673453599215,
      "epoch": 1.5648271142776637,
      "grad_norm": 0.15582381188869476,
      "learning_rate": 4.784248780936609e-05,
      "loss": 0.6758,
      "mean_token_accuracy": 0.7934990860521793,
      "num_tokens": 542837797.0,
      "step": 17008
    },
    {
      "entropy": 0.6504545770585537,
      "epoch": 1.564919120649105,
      "grad_norm": 0.15548722445964813,
      "learning_rate": 4.7839420983224464e-05,
      "loss": 0.6247,
      "mean_token_accuracy": 0.8078318573534489,
      "num_tokens": 542869909.0,
      "step": 17009
    },
    {
      "entropy": 0.5836726240813732,
      "epoch": 1.5650111270205462,
      "grad_norm": 0.15209412574768066,
      "learning_rate": 4.783635415708284e-05,
      "loss": 0.5599,
      "mean_token_accuracy": 0.8283923529088497,
      "num_tokens": 542902184.0,
      "step": 17010
    },
    {
      "entropy": 0.5809586234390736,
      "epoch": 1.5651031333919874,
      "grad_norm": 0.15975305438041687,
      "learning_rate": 4.783328733094121e-05,
      "loss": 0.5611,
      "mean_token_accuracy": 0.8297773376107216,
      "num_tokens": 542934481.0,
      "step": 17011
    },
    {
      "entropy": 0.565169139765203,
      "epoch": 1.5651951397634285,
      "grad_norm": 0.16575868427753448,
      "learning_rate": 4.783022050479958e-05,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8298337869346142,
      "num_tokens": 542966291.0,
      "step": 17012
    },
    {
      "entropy": 0.721586199477315,
      "epoch": 1.5652871461348699,
      "grad_norm": 0.16532060503959656,
      "learning_rate": 4.782715367865796e-05,
      "loss": 0.7069,
      "mean_token_accuracy": 0.7878416888415813,
      "num_tokens": 542998498.0,
      "step": 17013
    },
    {
      "entropy": 0.6262413393706083,
      "epoch": 1.5653791525063112,
      "grad_norm": 0.16247566044330597,
      "learning_rate": 4.782408685251633e-05,
      "loss": 0.6206,
      "mean_token_accuracy": 0.8095951750874519,
      "num_tokens": 543030850.0,
      "step": 17014
    },
    {
      "entropy": 0.5607587872073054,
      "epoch": 1.5654711588777523,
      "grad_norm": 0.14717522263526917,
      "learning_rate": 4.7821020026374705e-05,
      "loss": 0.5579,
      "mean_token_accuracy": 0.829169575124979,
      "num_tokens": 543063581.0,
      "step": 17015
    },
    {
      "entropy": 0.5890572592616081,
      "epoch": 1.5655631652491935,
      "grad_norm": 0.1534767895936966,
      "learning_rate": 4.781795320023308e-05,
      "loss": 0.5828,
      "mean_token_accuracy": 0.8198216035962105,
      "num_tokens": 543096291.0,
      "step": 17016
    },
    {
      "entropy": 0.5412318669259548,
      "epoch": 1.5656551716206346,
      "grad_norm": 0.15437878668308258,
      "learning_rate": 4.7814886374091454e-05,
      "loss": 0.5277,
      "mean_token_accuracy": 0.8371837064623833,
      "num_tokens": 543128418.0,
      "step": 17017
    },
    {
      "entropy": 0.5960258580744267,
      "epoch": 1.565747177992076,
      "grad_norm": 0.1529259979724884,
      "learning_rate": 4.781181954794983e-05,
      "loss": 0.6015,
      "mean_token_accuracy": 0.8181772157549858,
      "num_tokens": 543160835.0,
      "step": 17018
    },
    {
      "entropy": 0.6553479442372918,
      "epoch": 1.5658391843635173,
      "grad_norm": 0.16070431470870972,
      "learning_rate": 4.7808752721808204e-05,
      "loss": 0.6498,
      "mean_token_accuracy": 0.8022882863879204,
      "num_tokens": 543193219.0,
      "step": 17019
    },
    {
      "entropy": 0.7654071999713778,
      "epoch": 1.5659311907349585,
      "grad_norm": 0.16260379552841187,
      "learning_rate": 4.780568589566657e-05,
      "loss": 0.7515,
      "mean_token_accuracy": 0.7736910432577133,
      "num_tokens": 543225259.0,
      "step": 17020
    },
    {
      "entropy": 0.5304662957787514,
      "epoch": 1.5660231971063996,
      "grad_norm": 0.14731715619564056,
      "learning_rate": 4.780261906952495e-05,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8404916189610958,
      "num_tokens": 543257250.0,
      "step": 17021
    },
    {
      "entropy": 0.6008216990157962,
      "epoch": 1.5661152034778407,
      "grad_norm": 0.16076481342315674,
      "learning_rate": 4.779955224338333e-05,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8227437175810337,
      "num_tokens": 543288771.0,
      "step": 17022
    },
    {
      "entropy": 0.579690427519381,
      "epoch": 1.566207209849282,
      "grad_norm": 0.15964359045028687,
      "learning_rate": 4.7796485417241695e-05,
      "loss": 0.555,
      "mean_token_accuracy": 0.8272137753665447,
      "num_tokens": 543320094.0,
      "step": 17023
    },
    {
      "entropy": 0.6087274262681603,
      "epoch": 1.5662992162207234,
      "grad_norm": 0.1530449390411377,
      "learning_rate": 4.779341859110008e-05,
      "loss": 0.6051,
      "mean_token_accuracy": 0.8156892918050289,
      "num_tokens": 543352088.0,
      "step": 17024
    },
    {
      "entropy": 0.5386966029182076,
      "epoch": 1.5663912225921646,
      "grad_norm": 0.1542174518108368,
      "learning_rate": 4.7790351764958445e-05,
      "loss": 0.5323,
      "mean_token_accuracy": 0.8345778658986092,
      "num_tokens": 543383756.0,
      "step": 17025
    },
    {
      "entropy": 0.6097683720290661,
      "epoch": 1.5664832289636057,
      "grad_norm": 0.15140946209430695,
      "learning_rate": 4.7787284938816826e-05,
      "loss": 0.5859,
      "mean_token_accuracy": 0.8168801888823509,
      "num_tokens": 543415221.0,
      "step": 17026
    },
    {
      "entropy": 0.5895569622516632,
      "epoch": 1.5665752353350468,
      "grad_norm": 0.15191328525543213,
      "learning_rate": 4.7784218112675194e-05,
      "loss": 0.5868,
      "mean_token_accuracy": 0.8209693133831024,
      "num_tokens": 543446578.0,
      "step": 17027
    },
    {
      "entropy": 0.6817703377455473,
      "epoch": 1.5666672417064882,
      "grad_norm": 0.1670692265033722,
      "learning_rate": 4.778115128653357e-05,
      "loss": 0.6705,
      "mean_token_accuracy": 0.7963819541037083,
      "num_tokens": 543478263.0,
      "step": 17028
    },
    {
      "entropy": 0.6455013900995255,
      "epoch": 1.5667592480779295,
      "grad_norm": 0.16559046506881714,
      "learning_rate": 4.777808446039194e-05,
      "loss": 0.6439,
      "mean_token_accuracy": 0.8077379800379276,
      "num_tokens": 543509967.0,
      "step": 17029
    },
    {
      "entropy": 0.5873005664907396,
      "epoch": 1.5668512544493707,
      "grad_norm": 0.15703757107257843,
      "learning_rate": 4.777501763425032e-05,
      "loss": 0.5782,
      "mean_token_accuracy": 0.824066124856472,
      "num_tokens": 543541313.0,
      "step": 17030
    },
    {
      "entropy": 0.5912674013525248,
      "epoch": 1.5669432608208118,
      "grad_norm": 0.15697601437568665,
      "learning_rate": 4.777195080810869e-05,
      "loss": 0.5824,
      "mean_token_accuracy": 0.823423232883215,
      "num_tokens": 543571505.0,
      "step": 17031
    },
    {
      "entropy": 0.6619510781019926,
      "epoch": 1.567035267192253,
      "grad_norm": 0.1507483869791031,
      "learning_rate": 4.776888398196707e-05,
      "loss": 0.6448,
      "mean_token_accuracy": 0.7990536987781525,
      "num_tokens": 543603515.0,
      "step": 17032
    },
    {
      "entropy": 0.5472749136388302,
      "epoch": 1.5671272735636943,
      "grad_norm": 0.15933842957019806,
      "learning_rate": 4.7765817155825435e-05,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8355779238045216,
      "num_tokens": 543635291.0,
      "step": 17033
    },
    {
      "entropy": 0.6578534878790379,
      "epoch": 1.5672192799351357,
      "grad_norm": 0.15744724869728088,
      "learning_rate": 4.7762750329683817e-05,
      "loss": 0.6281,
      "mean_token_accuracy": 0.8057134374976158,
      "num_tokens": 543666639.0,
      "step": 17034
    },
    {
      "entropy": 0.6124729868024588,
      "epoch": 1.5673112863065768,
      "grad_norm": 0.1520490199327469,
      "learning_rate": 4.7759683503542184e-05,
      "loss": 0.5979,
      "mean_token_accuracy": 0.8174434676766396,
      "num_tokens": 543698569.0,
      "step": 17035
    },
    {
      "entropy": 0.5887286551296711,
      "epoch": 1.567403292678018,
      "grad_norm": 0.15287558734416962,
      "learning_rate": 4.775661667740056e-05,
      "loss": 0.578,
      "mean_token_accuracy": 0.8253246806561947,
      "num_tokens": 543730814.0,
      "step": 17036
    },
    {
      "entropy": 0.5987151293084025,
      "epoch": 1.567495299049459,
      "grad_norm": 0.15486162900924683,
      "learning_rate": 4.7753549851258934e-05,
      "loss": 0.5865,
      "mean_token_accuracy": 0.8198989927768707,
      "num_tokens": 543762546.0,
      "step": 17037
    },
    {
      "entropy": 0.6265807589516044,
      "epoch": 1.5675873054209004,
      "grad_norm": 0.15311671793460846,
      "learning_rate": 4.775048302511731e-05,
      "loss": 0.6202,
      "mean_token_accuracy": 0.8075467981398106,
      "num_tokens": 543794739.0,
      "step": 17038
    },
    {
      "entropy": 0.5972684435546398,
      "epoch": 1.5676793117923418,
      "grad_norm": 0.1524234414100647,
      "learning_rate": 4.774741619897568e-05,
      "loss": 0.6032,
      "mean_token_accuracy": 0.8178052417933941,
      "num_tokens": 543826964.0,
      "step": 17039
    },
    {
      "entropy": 0.49733811570331454,
      "epoch": 1.567771318163783,
      "grad_norm": 0.1434365063905716,
      "learning_rate": 4.774434937283406e-05,
      "loss": 0.467,
      "mean_token_accuracy": 0.8545226939022541,
      "num_tokens": 543858984.0,
      "step": 17040
    },
    {
      "entropy": 0.6069827349856496,
      "epoch": 1.567863324535224,
      "grad_norm": 0.15335485339164734,
      "learning_rate": 4.7741282546692425e-05,
      "loss": 0.599,
      "mean_token_accuracy": 0.8172951489686966,
      "num_tokens": 543891326.0,
      "step": 17041
    },
    {
      "entropy": 0.6304009594023228,
      "epoch": 1.5679553309066652,
      "grad_norm": 0.16362281143665314,
      "learning_rate": 4.773821572055081e-05,
      "loss": 0.6222,
      "mean_token_accuracy": 0.8100421391427517,
      "num_tokens": 543922734.0,
      "step": 17042
    },
    {
      "entropy": 0.5180795225314796,
      "epoch": 1.5680473372781065,
      "grad_norm": 0.1540052592754364,
      "learning_rate": 4.7735148894409175e-05,
      "loss": 0.4954,
      "mean_token_accuracy": 0.8464696519076824,
      "num_tokens": 543954413.0,
      "step": 17043
    },
    {
      "entropy": 0.6636233131866902,
      "epoch": 1.5681393436495479,
      "grad_norm": 0.1615520715713501,
      "learning_rate": 4.7732082068267556e-05,
      "loss": 0.6566,
      "mean_token_accuracy": 0.7985086143016815,
      "num_tokens": 543986453.0,
      "step": 17044
    },
    {
      "entropy": 0.6630700193345547,
      "epoch": 1.568231350020989,
      "grad_norm": 0.15877342224121094,
      "learning_rate": 4.7729015242125924e-05,
      "loss": 0.6543,
      "mean_token_accuracy": 0.8032457008957863,
      "num_tokens": 544018044.0,
      "step": 17045
    },
    {
      "entropy": 0.5882303724065423,
      "epoch": 1.5683233563924301,
      "grad_norm": 0.16157369315624237,
      "learning_rate": 4.77259484159843e-05,
      "loss": 0.5753,
      "mean_token_accuracy": 0.8225559741258621,
      "num_tokens": 544049640.0,
      "step": 17046
    },
    {
      "entropy": 0.7433116286993027,
      "epoch": 1.5684153627638713,
      "grad_norm": 0.16388778388500214,
      "learning_rate": 4.772288158984267e-05,
      "loss": 0.7308,
      "mean_token_accuracy": 0.7812152579426765,
      "num_tokens": 544081802.0,
      "step": 17047
    },
    {
      "entropy": 0.72552938060835,
      "epoch": 1.5685073691353126,
      "grad_norm": 0.16084825992584229,
      "learning_rate": 4.771981476370105e-05,
      "loss": 0.7154,
      "mean_token_accuracy": 0.7854840755462646,
      "num_tokens": 544113116.0,
      "step": 17048
    },
    {
      "entropy": 0.5545881688594818,
      "epoch": 1.568599375506754,
      "grad_norm": 0.1459731012582779,
      "learning_rate": 4.771674793755942e-05,
      "loss": 0.5443,
      "mean_token_accuracy": 0.8343359492719173,
      "num_tokens": 544145534.0,
      "step": 17049
    },
    {
      "entropy": 0.5364762172102928,
      "epoch": 1.5686913818781951,
      "grad_norm": 0.1505458652973175,
      "learning_rate": 4.77136811114178e-05,
      "loss": 0.5291,
      "mean_token_accuracy": 0.836921151727438,
      "num_tokens": 544177994.0,
      "step": 17050
    },
    {
      "entropy": 0.6091045513749123,
      "epoch": 1.5687833882496363,
      "grad_norm": 0.15646333992481232,
      "learning_rate": 4.7710614285276165e-05,
      "loss": 0.5885,
      "mean_token_accuracy": 0.8220726437866688,
      "num_tokens": 544210096.0,
      "step": 17051
    },
    {
      "entropy": 0.6990275401622057,
      "epoch": 1.5688753946210774,
      "grad_norm": 0.16058897972106934,
      "learning_rate": 4.770754745913455e-05,
      "loss": 0.6809,
      "mean_token_accuracy": 0.7915055565536022,
      "num_tokens": 544241792.0,
      "step": 17052
    },
    {
      "entropy": 0.6697375494986773,
      "epoch": 1.5689674009925187,
      "grad_norm": 0.16125927865505219,
      "learning_rate": 4.770448063299292e-05,
      "loss": 0.6645,
      "mean_token_accuracy": 0.7994320876896381,
      "num_tokens": 544274560.0,
      "step": 17053
    },
    {
      "entropy": 0.47486250940710306,
      "epoch": 1.56905940736396,
      "grad_norm": 0.15650543570518494,
      "learning_rate": 4.770141380685129e-05,
      "loss": 0.4692,
      "mean_token_accuracy": 0.8573379255831242,
      "num_tokens": 544306689.0,
      "step": 17054
    },
    {
      "entropy": 0.6008210508152843,
      "epoch": 1.5691514137354012,
      "grad_norm": 0.15010619163513184,
      "learning_rate": 4.769834698070967e-05,
      "loss": 0.5935,
      "mean_token_accuracy": 0.8163955174386501,
      "num_tokens": 544338313.0,
      "step": 17055
    },
    {
      "entropy": 0.6364187449216843,
      "epoch": 1.5692434201068424,
      "grad_norm": 0.16218982636928558,
      "learning_rate": 4.769528015456804e-05,
      "loss": 0.6343,
      "mean_token_accuracy": 0.8031776584684849,
      "num_tokens": 544369127.0,
      "step": 17056
    },
    {
      "entropy": 0.6545947678387165,
      "epoch": 1.5693354264782835,
      "grad_norm": 0.15920504927635193,
      "learning_rate": 4.769221332842641e-05,
      "loss": 0.6547,
      "mean_token_accuracy": 0.7996243461966515,
      "num_tokens": 544400918.0,
      "step": 17057
    },
    {
      "entropy": 0.5763465669006109,
      "epoch": 1.5694274328497249,
      "grad_norm": 0.1508476883172989,
      "learning_rate": 4.768914650228479e-05,
      "loss": 0.5729,
      "mean_token_accuracy": 0.8237517289817333,
      "num_tokens": 544433229.0,
      "step": 17058
    },
    {
      "entropy": 0.6027202978730202,
      "epoch": 1.5695194392211662,
      "grad_norm": 0.15337108075618744,
      "learning_rate": 4.768607967614316e-05,
      "loss": 0.598,
      "mean_token_accuracy": 0.8144224993884563,
      "num_tokens": 544465237.0,
      "step": 17059
    },
    {
      "entropy": 0.7470695730298758,
      "epoch": 1.5696114455926073,
      "grad_norm": 0.1608259230852127,
      "learning_rate": 4.768301285000154e-05,
      "loss": 0.7369,
      "mean_token_accuracy": 0.7776651158928871,
      "num_tokens": 544497724.0,
      "step": 17060
    },
    {
      "entropy": 0.6502459850162268,
      "epoch": 1.5697034519640485,
      "grad_norm": 0.15671376883983612,
      "learning_rate": 4.767994602385991e-05,
      "loss": 0.6346,
      "mean_token_accuracy": 0.8048661164939404,
      "num_tokens": 544530086.0,
      "step": 17061
    },
    {
      "entropy": 0.6546431500464678,
      "epoch": 1.5697954583354896,
      "grad_norm": 0.15613514184951782,
      "learning_rate": 4.767687919771828e-05,
      "loss": 0.6474,
      "mean_token_accuracy": 0.8004521243274212,
      "num_tokens": 544561617.0,
      "step": 17062
    },
    {
      "entropy": 0.4691129978746176,
      "epoch": 1.569887464706931,
      "grad_norm": 0.14863869547843933,
      "learning_rate": 4.767381237157666e-05,
      "loss": 0.4473,
      "mean_token_accuracy": 0.8591318912804127,
      "num_tokens": 544592647.0,
      "step": 17063
    },
    {
      "entropy": 0.5340330554172397,
      "epoch": 1.5699794710783723,
      "grad_norm": 0.15491510927677155,
      "learning_rate": 4.767074554543503e-05,
      "loss": 0.5185,
      "mean_token_accuracy": 0.8383833840489388,
      "num_tokens": 544624956.0,
      "step": 17064
    },
    {
      "entropy": 0.619185596704483,
      "epoch": 1.5700714774498135,
      "grad_norm": 0.1562587022781372,
      "learning_rate": 4.766767871929341e-05,
      "loss": 0.6048,
      "mean_token_accuracy": 0.8170270062983036,
      "num_tokens": 544656865.0,
      "step": 17065
    },
    {
      "entropy": 0.709619328379631,
      "epoch": 1.5701634838212546,
      "grad_norm": 0.16173239052295685,
      "learning_rate": 4.766461189315178e-05,
      "loss": 0.6957,
      "mean_token_accuracy": 0.7882242240011692,
      "num_tokens": 544688564.0,
      "step": 17066
    },
    {
      "entropy": 0.6631835773587227,
      "epoch": 1.5702554901926957,
      "grad_norm": 0.15956370532512665,
      "learning_rate": 4.766154506701015e-05,
      "loss": 0.653,
      "mean_token_accuracy": 0.8010526522994041,
      "num_tokens": 544720593.0,
      "step": 17067
    },
    {
      "entropy": 0.5851899180561304,
      "epoch": 1.570347496564137,
      "grad_norm": 0.16183722019195557,
      "learning_rate": 4.765847824086853e-05,
      "loss": 0.5644,
      "mean_token_accuracy": 0.8259475231170654,
      "num_tokens": 544751905.0,
      "step": 17068
    },
    {
      "entropy": 0.60802898183465,
      "epoch": 1.5704395029355784,
      "grad_norm": 0.16194725036621094,
      "learning_rate": 4.76554114147269e-05,
      "loss": 0.5935,
      "mean_token_accuracy": 0.8209335692226887,
      "num_tokens": 544783359.0,
      "step": 17069
    },
    {
      "entropy": 0.6571177737787366,
      "epoch": 1.5705315093070196,
      "grad_norm": 0.16198286414146423,
      "learning_rate": 4.765234458858528e-05,
      "loss": 0.6342,
      "mean_token_accuracy": 0.8031006641685963,
      "num_tokens": 544814400.0,
      "step": 17070
    },
    {
      "entropy": 0.655156257096678,
      "epoch": 1.5706235156784607,
      "grad_norm": 0.1568504422903061,
      "learning_rate": 4.764927776244365e-05,
      "loss": 0.6381,
      "mean_token_accuracy": 0.8026408813893795,
      "num_tokens": 544846609.0,
      "step": 17071
    },
    {
      "entropy": 0.5006589060649276,
      "epoch": 1.5707155220499018,
      "grad_norm": 0.14083755016326904,
      "learning_rate": 4.764621093630202e-05,
      "loss": 0.4982,
      "mean_token_accuracy": 0.8492497503757477,
      "num_tokens": 544879276.0,
      "step": 17072
    },
    {
      "entropy": 0.5683826431632042,
      "epoch": 1.5708075284213432,
      "grad_norm": 0.14694392681121826,
      "learning_rate": 4.76431441101604e-05,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8289941102266312,
      "num_tokens": 544911315.0,
      "step": 17073
    },
    {
      "entropy": 0.5125743225216866,
      "epoch": 1.5708995347927845,
      "grad_norm": 0.14717547595500946,
      "learning_rate": 4.764007728401877e-05,
      "loss": 0.5,
      "mean_token_accuracy": 0.8469490222632885,
      "num_tokens": 544943241.0,
      "step": 17074
    },
    {
      "entropy": 0.6339569594711065,
      "epoch": 1.5709915411642257,
      "grad_norm": 0.1658685952425003,
      "learning_rate": 4.763701045787714e-05,
      "loss": 0.6161,
      "mean_token_accuracy": 0.8132749125361443,
      "num_tokens": 544975286.0,
      "step": 17075
    },
    {
      "entropy": 0.6263838987797499,
      "epoch": 1.5710835475356668,
      "grad_norm": 0.15720151364803314,
      "learning_rate": 4.763394363173552e-05,
      "loss": 0.625,
      "mean_token_accuracy": 0.8087953440845013,
      "num_tokens": 545007779.0,
      "step": 17076
    },
    {
      "entropy": 0.5789707372896373,
      "epoch": 1.571175553907108,
      "grad_norm": 0.15265007317066193,
      "learning_rate": 4.763087680559389e-05,
      "loss": 0.5581,
      "mean_token_accuracy": 0.8311333991587162,
      "num_tokens": 545040251.0,
      "step": 17077
    },
    {
      "entropy": 0.5426044464111328,
      "epoch": 1.5712675602785493,
      "grad_norm": 0.1543177217245102,
      "learning_rate": 4.762780997945227e-05,
      "loss": 0.5212,
      "mean_token_accuracy": 0.83657056838274,
      "num_tokens": 545071973.0,
      "step": 17078
    },
    {
      "entropy": 0.6680910605937243,
      "epoch": 1.5713595666499907,
      "grad_norm": 0.1604710966348648,
      "learning_rate": 4.762474315331064e-05,
      "loss": 0.6725,
      "mean_token_accuracy": 0.7962633706629276,
      "num_tokens": 545104413.0,
      "step": 17079
    },
    {
      "entropy": 0.6203606948256493,
      "epoch": 1.5714515730214318,
      "grad_norm": 0.1653319150209427,
      "learning_rate": 4.762167632716901e-05,
      "loss": 0.5984,
      "mean_token_accuracy": 0.8152971267700195,
      "num_tokens": 545136122.0,
      "step": 17080
    },
    {
      "entropy": 0.5632173186168075,
      "epoch": 1.571543579392873,
      "grad_norm": 0.14851900935173035,
      "learning_rate": 4.761860950102739e-05,
      "loss": 0.5547,
      "mean_token_accuracy": 0.8283562287688255,
      "num_tokens": 545167858.0,
      "step": 17081
    },
    {
      "entropy": 0.5014135595411062,
      "epoch": 1.571635585764314,
      "grad_norm": 0.1436472088098526,
      "learning_rate": 4.761554267488576e-05,
      "loss": 0.4885,
      "mean_token_accuracy": 0.8495801165699959,
      "num_tokens": 545200121.0,
      "step": 17082
    },
    {
      "entropy": 0.5717768194153905,
      "epoch": 1.5717275921357554,
      "grad_norm": 0.1531398892402649,
      "learning_rate": 4.7612475848744134e-05,
      "loss": 0.5597,
      "mean_token_accuracy": 0.8286494761705399,
      "num_tokens": 545232339.0,
      "step": 17083
    },
    {
      "entropy": 0.5102578671649098,
      "epoch": 1.5718195985071968,
      "grad_norm": 0.1564328819513321,
      "learning_rate": 4.7609409022602515e-05,
      "loss": 0.5004,
      "mean_token_accuracy": 0.8456403166055679,
      "num_tokens": 545264364.0,
      "step": 17084
    },
    {
      "entropy": 0.654880790039897,
      "epoch": 1.571911604878638,
      "grad_norm": 0.1565561592578888,
      "learning_rate": 4.760634219646088e-05,
      "loss": 0.6502,
      "mean_token_accuracy": 0.8037426248192787,
      "num_tokens": 545296466.0,
      "step": 17085
    },
    {
      "entropy": 0.6690162420272827,
      "epoch": 1.572003611250079,
      "grad_norm": 0.16899418830871582,
      "learning_rate": 4.7603275370319264e-05,
      "loss": 0.661,
      "mean_token_accuracy": 0.8031307868659496,
      "num_tokens": 545328506.0,
      "step": 17086
    },
    {
      "entropy": 0.5389540451578796,
      "epoch": 1.5720956176215202,
      "grad_norm": 0.155915766954422,
      "learning_rate": 4.760020854417763e-05,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8416682258248329,
      "num_tokens": 545359444.0,
      "step": 17087
    },
    {
      "entropy": 0.5753929819911718,
      "epoch": 1.5721876239929615,
      "grad_norm": 0.1570620983839035,
      "learning_rate": 4.759714171803601e-05,
      "loss": 0.5793,
      "mean_token_accuracy": 0.821269690990448,
      "num_tokens": 545391103.0,
      "step": 17088
    },
    {
      "entropy": 0.5160142416134477,
      "epoch": 1.5722796303644029,
      "grad_norm": 0.1457356959581375,
      "learning_rate": 4.759407489189438e-05,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8420724347233772,
      "num_tokens": 545423452.0,
      "step": 17089
    },
    {
      "entropy": 0.6172914300113916,
      "epoch": 1.572371636735844,
      "grad_norm": 0.15551188588142395,
      "learning_rate": 4.7591008065752756e-05,
      "loss": 0.6105,
      "mean_token_accuracy": 0.8129637539386749,
      "num_tokens": 545455460.0,
      "step": 17090
    },
    {
      "entropy": 0.603679177351296,
      "epoch": 1.5724636431072851,
      "grad_norm": 0.1561899185180664,
      "learning_rate": 4.758794123961113e-05,
      "loss": 0.5946,
      "mean_token_accuracy": 0.8115699253976345,
      "num_tokens": 545486234.0,
      "step": 17091
    },
    {
      "entropy": 0.5408529071137309,
      "epoch": 1.5725556494787263,
      "grad_norm": 0.15327061712741852,
      "learning_rate": 4.7584874413469505e-05,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8363208621740341,
      "num_tokens": 545517900.0,
      "step": 17092
    },
    {
      "entropy": 0.6276458036154509,
      "epoch": 1.5726476558501676,
      "grad_norm": 0.1590700000524521,
      "learning_rate": 4.758180758732787e-05,
      "loss": 0.618,
      "mean_token_accuracy": 0.8124554082751274,
      "num_tokens": 545549067.0,
      "step": 17093
    },
    {
      "entropy": 0.509167367592454,
      "epoch": 1.572739662221609,
      "grad_norm": 0.14708589017391205,
      "learning_rate": 4.7578740761186255e-05,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8472808822989464,
      "num_tokens": 545581163.0,
      "step": 17094
    },
    {
      "entropy": 0.6319268373772502,
      "epoch": 1.5728316685930501,
      "grad_norm": 0.15812306106090546,
      "learning_rate": 4.757567393504462e-05,
      "loss": 0.6287,
      "mean_token_accuracy": 0.8060627542436123,
      "num_tokens": 545612552.0,
      "step": 17095
    },
    {
      "entropy": 0.6041099969297647,
      "epoch": 1.5729236749644913,
      "grad_norm": 0.15494133532047272,
      "learning_rate": 4.7572607108903e-05,
      "loss": 0.5898,
      "mean_token_accuracy": 0.820923175662756,
      "num_tokens": 545644494.0,
      "step": 17096
    },
    {
      "entropy": 0.5237900093197823,
      "epoch": 1.5730156813359324,
      "grad_norm": 0.14564310014247894,
      "learning_rate": 4.756954028276137e-05,
      "loss": 0.5027,
      "mean_token_accuracy": 0.841711338609457,
      "num_tokens": 545675814.0,
      "step": 17097
    },
    {
      "entropy": 0.6178318690508604,
      "epoch": 1.5731076877073737,
      "grad_norm": 0.15228913724422455,
      "learning_rate": 4.7566473456619747e-05,
      "loss": 0.6101,
      "mean_token_accuracy": 0.8111914843320847,
      "num_tokens": 545708050.0,
      "step": 17098
    },
    {
      "entropy": 0.6479746326804161,
      "epoch": 1.573199694078815,
      "grad_norm": 0.15779678523540497,
      "learning_rate": 4.756340663047812e-05,
      "loss": 0.6371,
      "mean_token_accuracy": 0.8060195744037628,
      "num_tokens": 545739872.0,
      "step": 17099
    },
    {
      "entropy": 0.6444749366492033,
      "epoch": 1.5732917004502562,
      "grad_norm": 0.17961321771144867,
      "learning_rate": 4.7560339804336496e-05,
      "loss": 0.6241,
      "mean_token_accuracy": 0.8090741336345673,
      "num_tokens": 545772448.0,
      "step": 17100
    },
    {
      "entropy": 0.6248152144253254,
      "epoch": 1.5733837068216974,
      "grad_norm": 0.15987500548362732,
      "learning_rate": 4.7557272978194864e-05,
      "loss": 0.6046,
      "mean_token_accuracy": 0.8167676962912083,
      "num_tokens": 545803828.0,
      "step": 17101
    },
    {
      "entropy": 0.6877743294462562,
      "epoch": 1.5734757131931385,
      "grad_norm": 0.16150091588497162,
      "learning_rate": 4.7554206152053245e-05,
      "loss": 0.6805,
      "mean_token_accuracy": 0.7935284040868282,
      "num_tokens": 545835755.0,
      "step": 17102
    },
    {
      "entropy": 0.6629192624241114,
      "epoch": 1.5735677195645799,
      "grad_norm": 0.15586590766906738,
      "learning_rate": 4.755113932591161e-05,
      "loss": 0.6622,
      "mean_token_accuracy": 0.7975038774311543,
      "num_tokens": 545867592.0,
      "step": 17103
    },
    {
      "entropy": 0.5814272155985236,
      "epoch": 1.5736597259360212,
      "grad_norm": 0.1511196494102478,
      "learning_rate": 4.7548072499769994e-05,
      "loss": 0.5578,
      "mean_token_accuracy": 0.8260821141302586,
      "num_tokens": 545899747.0,
      "step": 17104
    },
    {
      "entropy": 0.7063285019248724,
      "epoch": 1.5737517323074623,
      "grad_norm": 0.16428451240062714,
      "learning_rate": 4.754500567362836e-05,
      "loss": 0.6991,
      "mean_token_accuracy": 0.786528542637825,
      "num_tokens": 545930975.0,
      "step": 17105
    },
    {
      "entropy": 0.6229520477354527,
      "epoch": 1.5738437386789035,
      "grad_norm": 0.15764926373958588,
      "learning_rate": 4.754193884748674e-05,
      "loss": 0.6121,
      "mean_token_accuracy": 0.8083864115178585,
      "num_tokens": 545962480.0,
      "step": 17106
    },
    {
      "entropy": 0.5826554312370718,
      "epoch": 1.5739357450503446,
      "grad_norm": 0.1586071401834488,
      "learning_rate": 4.753887202134511e-05,
      "loss": 0.5753,
      "mean_token_accuracy": 0.8202768377959728,
      "num_tokens": 545994577.0,
      "step": 17107
    },
    {
      "entropy": 0.6410672385245562,
      "epoch": 1.574027751421786,
      "grad_norm": 0.16307470202445984,
      "learning_rate": 4.7535805195203486e-05,
      "loss": 0.6427,
      "mean_token_accuracy": 0.8016727045178413,
      "num_tokens": 546026056.0,
      "step": 17108
    },
    {
      "entropy": 0.6969601698219776,
      "epoch": 1.5741197577932273,
      "grad_norm": 0.16386087238788605,
      "learning_rate": 4.753273836906186e-05,
      "loss": 0.7019,
      "mean_token_accuracy": 0.7843081764876842,
      "num_tokens": 546058081.0,
      "step": 17109
    },
    {
      "entropy": 0.5139843337237835,
      "epoch": 1.5742117641646685,
      "grad_norm": 0.14895592629909515,
      "learning_rate": 4.7529671542920236e-05,
      "loss": 0.4876,
      "mean_token_accuracy": 0.846751406788826,
      "num_tokens": 546089547.0,
      "step": 17110
    },
    {
      "entropy": 0.6969523672014475,
      "epoch": 1.5743037705361096,
      "grad_norm": 0.1627657264471054,
      "learning_rate": 4.7526604716778603e-05,
      "loss": 0.6972,
      "mean_token_accuracy": 0.7893077507615089,
      "num_tokens": 546120952.0,
      "step": 17111
    },
    {
      "entropy": 0.6910208035260439,
      "epoch": 1.5743957769075507,
      "grad_norm": 0.16408134996891022,
      "learning_rate": 4.7523537890636985e-05,
      "loss": 0.681,
      "mean_token_accuracy": 0.7934880070388317,
      "num_tokens": 546152099.0,
      "step": 17112
    },
    {
      "entropy": 0.5624227896332741,
      "epoch": 1.574487783278992,
      "grad_norm": 0.15064875781536102,
      "learning_rate": 4.752047106449536e-05,
      "loss": 0.5471,
      "mean_token_accuracy": 0.8320666700601578,
      "num_tokens": 546184182.0,
      "step": 17113
    },
    {
      "entropy": 0.615127719938755,
      "epoch": 1.5745797896504334,
      "grad_norm": 0.15761588513851166,
      "learning_rate": 4.751740423835373e-05,
      "loss": 0.6016,
      "mean_token_accuracy": 0.8121427446603775,
      "num_tokens": 546215329.0,
      "step": 17114
    },
    {
      "entropy": 0.6255456323269755,
      "epoch": 1.5746717960218746,
      "grad_norm": 0.15207123756408691,
      "learning_rate": 4.751433741221211e-05,
      "loss": 0.6195,
      "mean_token_accuracy": 0.8100421465933323,
      "num_tokens": 546247955.0,
      "step": 17115
    },
    {
      "entropy": 0.723395636305213,
      "epoch": 1.5747638023933157,
      "grad_norm": 0.1663830578327179,
      "learning_rate": 4.751127058607048e-05,
      "loss": 0.713,
      "mean_token_accuracy": 0.7841677963733673,
      "num_tokens": 546279583.0,
      "step": 17116
    },
    {
      "entropy": 0.6454102452844381,
      "epoch": 1.5748558087647568,
      "grad_norm": 0.1631133258342743,
      "learning_rate": 4.750820375992885e-05,
      "loss": 0.6416,
      "mean_token_accuracy": 0.8027440533041954,
      "num_tokens": 546310684.0,
      "step": 17117
    },
    {
      "entropy": 0.7152496790513396,
      "epoch": 1.5749478151361982,
      "grad_norm": 0.16357095539569855,
      "learning_rate": 4.7505136933787226e-05,
      "loss": 0.7083,
      "mean_token_accuracy": 0.7824101597070694,
      "num_tokens": 546342870.0,
      "step": 17118
    },
    {
      "entropy": 0.5960429329425097,
      "epoch": 1.5750398215076395,
      "grad_norm": 0.14695000648498535,
      "learning_rate": 4.75020701076456e-05,
      "loss": 0.5829,
      "mean_token_accuracy": 0.8197418339550495,
      "num_tokens": 546375036.0,
      "step": 17119
    },
    {
      "entropy": 0.5916581777855754,
      "epoch": 1.5751318278790807,
      "grad_norm": 0.16663038730621338,
      "learning_rate": 4.7499003281503975e-05,
      "loss": 0.569,
      "mean_token_accuracy": 0.8242411613464355,
      "num_tokens": 546406622.0,
      "step": 17120
    },
    {
      "entropy": 0.5478697381913662,
      "epoch": 1.5752238342505218,
      "grad_norm": 0.14903169870376587,
      "learning_rate": 4.749593645536235e-05,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8323812261223793,
      "num_tokens": 546439123.0,
      "step": 17121
    },
    {
      "entropy": 0.5684241498820484,
      "epoch": 1.575315840621963,
      "grad_norm": 0.1481020152568817,
      "learning_rate": 4.749286962922072e-05,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8317357748746872,
      "num_tokens": 546471583.0,
      "step": 17122
    },
    {
      "entropy": 0.5620642993599176,
      "epoch": 1.5754078469934043,
      "grad_norm": 0.15488769114017487,
      "learning_rate": 4.74898028030791e-05,
      "loss": 0.5377,
      "mean_token_accuracy": 0.8342850133776665,
      "num_tokens": 546503691.0,
      "step": 17123
    },
    {
      "entropy": 0.6130684372037649,
      "epoch": 1.5754998533648457,
      "grad_norm": 0.15430942177772522,
      "learning_rate": 4.748673597693747e-05,
      "loss": 0.6043,
      "mean_token_accuracy": 0.8138044662773609,
      "num_tokens": 546535988.0,
      "step": 17124
    },
    {
      "entropy": 0.6371645508334041,
      "epoch": 1.5755918597362868,
      "grad_norm": 0.15901121497154236,
      "learning_rate": 4.748366915079585e-05,
      "loss": 0.6323,
      "mean_token_accuracy": 0.8043836653232574,
      "num_tokens": 546567680.0,
      "step": 17125
    },
    {
      "entropy": 0.6395663507282734,
      "epoch": 1.575683866107728,
      "grad_norm": 0.16660597920417786,
      "learning_rate": 4.7480602324654216e-05,
      "loss": 0.6242,
      "mean_token_accuracy": 0.8065934367477894,
      "num_tokens": 546598922.0,
      "step": 17126
    },
    {
      "entropy": 0.5458716377615929,
      "epoch": 1.575775872479169,
      "grad_norm": 0.15074436366558075,
      "learning_rate": 4.747753549851259e-05,
      "loss": 0.5386,
      "mean_token_accuracy": 0.8318939916789532,
      "num_tokens": 546631347.0,
      "step": 17127
    },
    {
      "entropy": 0.6656217984855175,
      "epoch": 1.5758678788506104,
      "grad_norm": 0.15996916592121124,
      "learning_rate": 4.7474468672370966e-05,
      "loss": 0.6616,
      "mean_token_accuracy": 0.7990351617336273,
      "num_tokens": 546664115.0,
      "step": 17128
    },
    {
      "entropy": 0.6162683330476284,
      "epoch": 1.5759598852220518,
      "grad_norm": 0.15947633981704712,
      "learning_rate": 4.747140184622934e-05,
      "loss": 0.6108,
      "mean_token_accuracy": 0.8073750212788582,
      "num_tokens": 546696030.0,
      "step": 17129
    },
    {
      "entropy": 0.6040589979384094,
      "epoch": 1.576051891593493,
      "grad_norm": 0.15149088203907013,
      "learning_rate": 4.7468335020087715e-05,
      "loss": 0.5952,
      "mean_token_accuracy": 0.8228621520102024,
      "num_tokens": 546728726.0,
      "step": 17130
    },
    {
      "entropy": 0.7244839705526829,
      "epoch": 1.576143897964934,
      "grad_norm": 0.16809038817882538,
      "learning_rate": 4.746526819394609e-05,
      "loss": 0.7098,
      "mean_token_accuracy": 0.7846744693815708,
      "num_tokens": 546760856.0,
      "step": 17131
    },
    {
      "entropy": 0.6297617377713323,
      "epoch": 1.5762359043363752,
      "grad_norm": 0.1624630242586136,
      "learning_rate": 4.746220136780446e-05,
      "loss": 0.6257,
      "mean_token_accuracy": 0.8107068799436092,
      "num_tokens": 546793355.0,
      "step": 17132
    },
    {
      "entropy": 0.7105263322591782,
      "epoch": 1.5763279107078165,
      "grad_norm": 0.1677854061126709,
      "learning_rate": 4.745913454166284e-05,
      "loss": 0.7049,
      "mean_token_accuracy": 0.788682121783495,
      "num_tokens": 546825099.0,
      "step": 17133
    },
    {
      "entropy": 0.6238522529602051,
      "epoch": 1.5764199170792579,
      "grad_norm": 0.1577150970697403,
      "learning_rate": 4.745606771552121e-05,
      "loss": 0.5995,
      "mean_token_accuracy": 0.8191373944282532,
      "num_tokens": 546857072.0,
      "step": 17134
    },
    {
      "entropy": 0.657599838450551,
      "epoch": 1.576511923450699,
      "grad_norm": 0.15531764924526215,
      "learning_rate": 4.745300088937958e-05,
      "loss": 0.6459,
      "mean_token_accuracy": 0.8001746833324432,
      "num_tokens": 546889490.0,
      "step": 17135
    },
    {
      "entropy": 0.6646223776042461,
      "epoch": 1.5766039298221401,
      "grad_norm": 0.15934330224990845,
      "learning_rate": 4.7449934063237956e-05,
      "loss": 0.6508,
      "mean_token_accuracy": 0.7989486679434776,
      "num_tokens": 546921365.0,
      "step": 17136
    },
    {
      "entropy": 0.6132154110819101,
      "epoch": 1.5766959361935813,
      "grad_norm": 0.1485031545162201,
      "learning_rate": 4.744686723709633e-05,
      "loss": 0.594,
      "mean_token_accuracy": 0.8168659768998623,
      "num_tokens": 546953067.0,
      "step": 17137
    },
    {
      "entropy": 0.6437358930706978,
      "epoch": 1.5767879425650226,
      "grad_norm": 0.15645530819892883,
      "learning_rate": 4.7443800410954705e-05,
      "loss": 0.6167,
      "mean_token_accuracy": 0.8105561956763268,
      "num_tokens": 546984811.0,
      "step": 17138
    },
    {
      "entropy": 0.5596108175814152,
      "epoch": 1.576879948936464,
      "grad_norm": 0.1556711494922638,
      "learning_rate": 4.744073358481308e-05,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8305409885942936,
      "num_tokens": 547016159.0,
      "step": 17139
    },
    {
      "entropy": 0.6053741620853543,
      "epoch": 1.5769719553079051,
      "grad_norm": 0.15213218331336975,
      "learning_rate": 4.743766675867145e-05,
      "loss": 0.5971,
      "mean_token_accuracy": 0.81947061419487,
      "num_tokens": 547048830.0,
      "step": 17140
    },
    {
      "entropy": 0.5854282267391682,
      "epoch": 1.5770639616793463,
      "grad_norm": 0.14731226861476898,
      "learning_rate": 4.743459993252983e-05,
      "loss": 0.5803,
      "mean_token_accuracy": 0.8242211155593395,
      "num_tokens": 547081389.0,
      "step": 17141
    },
    {
      "entropy": 0.6071947030723095,
      "epoch": 1.5771559680507874,
      "grad_norm": 0.15666824579238892,
      "learning_rate": 4.74315331063882e-05,
      "loss": 0.6006,
      "mean_token_accuracy": 0.8183872476220131,
      "num_tokens": 547113432.0,
      "step": 17142
    },
    {
      "entropy": 0.7045609299093485,
      "epoch": 1.5772479744222287,
      "grad_norm": 0.16219770908355713,
      "learning_rate": 4.742846628024657e-05,
      "loss": 0.7128,
      "mean_token_accuracy": 0.7847355082631111,
      "num_tokens": 547145662.0,
      "step": 17143
    },
    {
      "entropy": 0.4713833020068705,
      "epoch": 1.57733998079367,
      "grad_norm": 0.143607959151268,
      "learning_rate": 4.742539945410495e-05,
      "loss": 0.4591,
      "mean_token_accuracy": 0.8562598414719105,
      "num_tokens": 547178024.0,
      "step": 17144
    },
    {
      "entropy": 0.5937671745195985,
      "epoch": 1.5774319871651112,
      "grad_norm": 0.15549446642398834,
      "learning_rate": 4.742233262796332e-05,
      "loss": 0.5841,
      "mean_token_accuracy": 0.8188236691057682,
      "num_tokens": 547209752.0,
      "step": 17145
    },
    {
      "entropy": 0.6428636275231838,
      "epoch": 1.5775239935365524,
      "grad_norm": 0.16153615713119507,
      "learning_rate": 4.74192658018217e-05,
      "loss": 0.6342,
      "mean_token_accuracy": 0.8055179379880428,
      "num_tokens": 547241353.0,
      "step": 17146
    },
    {
      "entropy": 0.6588586140424013,
      "epoch": 1.5776159999079935,
      "grad_norm": 0.15854981541633606,
      "learning_rate": 4.741619897568007e-05,
      "loss": 0.6413,
      "mean_token_accuracy": 0.8041065558791161,
      "num_tokens": 547274061.0,
      "step": 17147
    },
    {
      "entropy": 0.5682955328375101,
      "epoch": 1.5777080062794349,
      "grad_norm": 0.1520422101020813,
      "learning_rate": 4.7413132149538445e-05,
      "loss": 0.5561,
      "mean_token_accuracy": 0.8312624096870422,
      "num_tokens": 547306259.0,
      "step": 17148
    },
    {
      "entropy": 0.54153798148036,
      "epoch": 1.5778000126508762,
      "grad_norm": 0.14950114488601685,
      "learning_rate": 4.741006532339682e-05,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8333747908473015,
      "num_tokens": 547338759.0,
      "step": 17149
    },
    {
      "entropy": 0.49537351494655013,
      "epoch": 1.5778920190223173,
      "grad_norm": 0.14384178817272186,
      "learning_rate": 4.7406998497255194e-05,
      "loss": 0.4896,
      "mean_token_accuracy": 0.8487379252910614,
      "num_tokens": 547370904.0,
      "step": 17150
    },
    {
      "entropy": 0.5401685833930969,
      "epoch": 1.5779840253937585,
      "grad_norm": 0.14881017804145813,
      "learning_rate": 4.740393167111357e-05,
      "loss": 0.5402,
      "mean_token_accuracy": 0.8317680545151234,
      "num_tokens": 547402621.0,
      "step": 17151
    },
    {
      "entropy": 0.5984025890938938,
      "epoch": 1.5780760317651996,
      "grad_norm": 0.15065890550613403,
      "learning_rate": 4.7400864844971944e-05,
      "loss": 0.59,
      "mean_token_accuracy": 0.8203104734420776,
      "num_tokens": 547434943.0,
      "step": 17152
    },
    {
      "entropy": 0.5580399259924889,
      "epoch": 1.578168038136641,
      "grad_norm": 0.14974620938301086,
      "learning_rate": 4.739779801883031e-05,
      "loss": 0.556,
      "mean_token_accuracy": 0.8293487913906574,
      "num_tokens": 547467419.0,
      "step": 17153
    },
    {
      "entropy": 0.630718357861042,
      "epoch": 1.5782600445080823,
      "grad_norm": 0.15323880314826965,
      "learning_rate": 4.739473119268869e-05,
      "loss": 0.6108,
      "mean_token_accuracy": 0.8103530779480934,
      "num_tokens": 547499585.0,
      "step": 17154
    },
    {
      "entropy": 0.6614488158375025,
      "epoch": 1.5783520508795235,
      "grad_norm": 0.16321280598640442,
      "learning_rate": 4.739166436654706e-05,
      "loss": 0.6624,
      "mean_token_accuracy": 0.7977663241326809,
      "num_tokens": 547531126.0,
      "step": 17155
    },
    {
      "entropy": 0.6489768177270889,
      "epoch": 1.5784440572509646,
      "grad_norm": 0.16101479530334473,
      "learning_rate": 4.7388597540405435e-05,
      "loss": 0.633,
      "mean_token_accuracy": 0.8040363974869251,
      "num_tokens": 547563164.0,
      "step": 17156
    },
    {
      "entropy": 0.6577582210302353,
      "epoch": 1.5785360636224057,
      "grad_norm": 0.15880100429058075,
      "learning_rate": 4.738553071426381e-05,
      "loss": 0.6411,
      "mean_token_accuracy": 0.8018363825976849,
      "num_tokens": 547594901.0,
      "step": 17157
    },
    {
      "entropy": 0.6596464719623327,
      "epoch": 1.578628069993847,
      "grad_norm": 0.15843699872493744,
      "learning_rate": 4.7382463888122185e-05,
      "loss": 0.6581,
      "mean_token_accuracy": 0.7982022538781166,
      "num_tokens": 547626095.0,
      "step": 17158
    },
    {
      "entropy": 0.5723990676924586,
      "epoch": 1.5787200763652884,
      "grad_norm": 0.1494695395231247,
      "learning_rate": 4.737939706198056e-05,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8273310028016567,
      "num_tokens": 547658107.0,
      "step": 17159
    },
    {
      "entropy": 0.6564028924331069,
      "epoch": 1.5788120827367296,
      "grad_norm": 0.1580173224210739,
      "learning_rate": 4.7376330235838934e-05,
      "loss": 0.6458,
      "mean_token_accuracy": 0.8005405738949776,
      "num_tokens": 547689799.0,
      "step": 17160
    },
    {
      "entropy": 0.6085729952901602,
      "epoch": 1.5789040891081707,
      "grad_norm": 0.14836642146110535,
      "learning_rate": 4.73732634096973e-05,
      "loss": 0.5952,
      "mean_token_accuracy": 0.8165943436324596,
      "num_tokens": 547721588.0,
      "step": 17161
    },
    {
      "entropy": 0.5701492689549923,
      "epoch": 1.5789960954796118,
      "grad_norm": 0.1538352519273758,
      "learning_rate": 4.737019658355568e-05,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8267039880156517,
      "num_tokens": 547753153.0,
      "step": 17162
    },
    {
      "entropy": 0.5848608501255512,
      "epoch": 1.5790881018510532,
      "grad_norm": 0.15873835980892181,
      "learning_rate": 4.736712975741405e-05,
      "loss": 0.5661,
      "mean_token_accuracy": 0.8253401480615139,
      "num_tokens": 547783651.0,
      "step": 17163
    },
    {
      "entropy": 0.5906747616827488,
      "epoch": 1.5791801082224945,
      "grad_norm": 0.1640232354402542,
      "learning_rate": 4.736406293127243e-05,
      "loss": 0.5813,
      "mean_token_accuracy": 0.8200059570372105,
      "num_tokens": 547814849.0,
      "step": 17164
    },
    {
      "entropy": 0.6057431194931269,
      "epoch": 1.5792721145939357,
      "grad_norm": 0.15611110627651215,
      "learning_rate": 4.73609961051308e-05,
      "loss": 0.6056,
      "mean_token_accuracy": 0.8122803494334221,
      "num_tokens": 547846299.0,
      "step": 17165
    },
    {
      "entropy": 0.6182456184178591,
      "epoch": 1.5793641209653768,
      "grad_norm": 0.15764938294887543,
      "learning_rate": 4.7357929278989175e-05,
      "loss": 0.6123,
      "mean_token_accuracy": 0.8128765150904655,
      "num_tokens": 547877697.0,
      "step": 17166
    },
    {
      "entropy": 0.5953467292711139,
      "epoch": 1.579456127336818,
      "grad_norm": 0.15376873314380646,
      "learning_rate": 4.735486245284755e-05,
      "loss": 0.5625,
      "mean_token_accuracy": 0.8207242786884308,
      "num_tokens": 547909092.0,
      "step": 17167
    },
    {
      "entropy": 0.5989406243897974,
      "epoch": 1.5795481337082593,
      "grad_norm": 0.15396560728549957,
      "learning_rate": 4.7351795626705924e-05,
      "loss": 0.5717,
      "mean_token_accuracy": 0.8217873647809029,
      "num_tokens": 547940610.0,
      "step": 17168
    },
    {
      "entropy": 0.5952049698680639,
      "epoch": 1.5796401400797007,
      "grad_norm": 0.15488436818122864,
      "learning_rate": 4.73487288005643e-05,
      "loss": 0.5917,
      "mean_token_accuracy": 0.8217909149825573,
      "num_tokens": 547972575.0,
      "step": 17169
    },
    {
      "entropy": 0.526267021894455,
      "epoch": 1.5797321464511418,
      "grad_norm": 0.14714106917381287,
      "learning_rate": 4.7345661974422674e-05,
      "loss": 0.5111,
      "mean_token_accuracy": 0.8425286822021008,
      "num_tokens": 548004931.0,
      "step": 17170
    },
    {
      "entropy": 0.6795961055904627,
      "epoch": 1.579824152822583,
      "grad_norm": 0.15890300273895264,
      "learning_rate": 4.734259514828104e-05,
      "loss": 0.6593,
      "mean_token_accuracy": 0.7999304607510567,
      "num_tokens": 548036906.0,
      "step": 17171
    },
    {
      "entropy": 0.5752454306930304,
      "epoch": 1.579916159194024,
      "grad_norm": 0.15026254951953888,
      "learning_rate": 4.733952832213942e-05,
      "loss": 0.5668,
      "mean_token_accuracy": 0.8245389275252819,
      "num_tokens": 548069083.0,
      "step": 17172
    },
    {
      "entropy": 0.5686573304701596,
      "epoch": 1.5800081655654654,
      "grad_norm": 0.15659630298614502,
      "learning_rate": 4.733646149599779e-05,
      "loss": 0.5581,
      "mean_token_accuracy": 0.8326665349304676,
      "num_tokens": 548101380.0,
      "step": 17173
    },
    {
      "entropy": 0.5690696174278855,
      "epoch": 1.5801001719369068,
      "grad_norm": 0.14662180840969086,
      "learning_rate": 4.7333394669856166e-05,
      "loss": 0.5655,
      "mean_token_accuracy": 0.8287525624036789,
      "num_tokens": 548133864.0,
      "step": 17174
    },
    {
      "entropy": 0.5737315970472991,
      "epoch": 1.580192178308348,
      "grad_norm": 0.14445848762989044,
      "learning_rate": 4.733032784371455e-05,
      "loss": 0.5604,
      "mean_token_accuracy": 0.827739767730236,
      "num_tokens": 548166126.0,
      "step": 17175
    },
    {
      "entropy": 0.5982474503107369,
      "epoch": 1.580284184679789,
      "grad_norm": 0.14973962306976318,
      "learning_rate": 4.7327261017572915e-05,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8219047598540783,
      "num_tokens": 548198418.0,
      "step": 17176
    },
    {
      "entropy": 0.6219963077455759,
      "epoch": 1.5803761910512302,
      "grad_norm": 0.16056334972381592,
      "learning_rate": 4.732419419143129e-05,
      "loss": 0.6168,
      "mean_token_accuracy": 0.8102629408240318,
      "num_tokens": 548229843.0,
      "step": 17177
    },
    {
      "entropy": 0.6014267378486693,
      "epoch": 1.5804681974226715,
      "grad_norm": 0.1562049686908722,
      "learning_rate": 4.7321127365289664e-05,
      "loss": 0.5837,
      "mean_token_accuracy": 0.8174160495400429,
      "num_tokens": 548260748.0,
      "step": 17178
    },
    {
      "entropy": 0.5998056884855032,
      "epoch": 1.5805602037941129,
      "grad_norm": 0.15876592695713043,
      "learning_rate": 4.731806053914804e-05,
      "loss": 0.5896,
      "mean_token_accuracy": 0.8192646540701389,
      "num_tokens": 548292850.0,
      "step": 17179
    },
    {
      "entropy": 0.5370111586526036,
      "epoch": 1.580652210165554,
      "grad_norm": 0.1499624103307724,
      "learning_rate": 4.7314993713006413e-05,
      "loss": 0.5279,
      "mean_token_accuracy": 0.8389746248722076,
      "num_tokens": 548324799.0,
      "step": 17180
    },
    {
      "entropy": 0.6718806065618992,
      "epoch": 1.5807442165369952,
      "grad_norm": 0.15971916913986206,
      "learning_rate": 4.731192688686479e-05,
      "loss": 0.6414,
      "mean_token_accuracy": 0.8033573962748051,
      "num_tokens": 548356513.0,
      "step": 17181
    },
    {
      "entropy": 0.5204277336597443,
      "epoch": 1.5808362229084363,
      "grad_norm": 0.14639122784137726,
      "learning_rate": 4.7308860060723156e-05,
      "loss": 0.5108,
      "mean_token_accuracy": 0.8418228477239609,
      "num_tokens": 548388715.0,
      "step": 17182
    },
    {
      "entropy": 0.537041618488729,
      "epoch": 1.5809282292798776,
      "grad_norm": 0.15005461871623993,
      "learning_rate": 4.730579323458154e-05,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8382411226630211,
      "num_tokens": 548419652.0,
      "step": 17183
    },
    {
      "entropy": 0.6002835687249899,
      "epoch": 1.581020235651319,
      "grad_norm": 0.15915514528751373,
      "learning_rate": 4.7302726408439905e-05,
      "loss": 0.5876,
      "mean_token_accuracy": 0.8202528357505798,
      "num_tokens": 548451395.0,
      "step": 17184
    },
    {
      "entropy": 0.7428519893437624,
      "epoch": 1.5811122420227601,
      "grad_norm": 0.16262929141521454,
      "learning_rate": 4.729965958229829e-05,
      "loss": 0.7349,
      "mean_token_accuracy": 0.7777785696089268,
      "num_tokens": 548483688.0,
      "step": 17185
    },
    {
      "entropy": 0.44899845961481333,
      "epoch": 1.5812042483942013,
      "grad_norm": 0.14255528151988983,
      "learning_rate": 4.7296592756156655e-05,
      "loss": 0.4363,
      "mean_token_accuracy": 0.862346526235342,
      "num_tokens": 548515350.0,
      "step": 17186
    },
    {
      "entropy": 0.5480848886072636,
      "epoch": 1.5812962547656424,
      "grad_norm": 0.14695818722248077,
      "learning_rate": 4.729352593001503e-05,
      "loss": 0.552,
      "mean_token_accuracy": 0.827464085072279,
      "num_tokens": 548547718.0,
      "step": 17187
    },
    {
      "entropy": 0.627758314833045,
      "epoch": 1.5813882611370838,
      "grad_norm": 0.15370965003967285,
      "learning_rate": 4.7290459103873404e-05,
      "loss": 0.6252,
      "mean_token_accuracy": 0.8086415752768517,
      "num_tokens": 548579618.0,
      "step": 17188
    },
    {
      "entropy": 0.6671025697141886,
      "epoch": 1.581480267508525,
      "grad_norm": 0.1543537676334381,
      "learning_rate": 4.728739227773178e-05,
      "loss": 0.667,
      "mean_token_accuracy": 0.7984829545021057,
      "num_tokens": 548611618.0,
      "step": 17189
    },
    {
      "entropy": 0.6315167341381311,
      "epoch": 1.5815722738799662,
      "grad_norm": 0.1594032347202301,
      "learning_rate": 4.728432545159015e-05,
      "loss": 0.6205,
      "mean_token_accuracy": 0.8026236295700073,
      "num_tokens": 548643851.0,
      "step": 17190
    },
    {
      "entropy": 0.6095159146934748,
      "epoch": 1.5816642802514074,
      "grad_norm": 0.15292364358901978,
      "learning_rate": 4.728125862544853e-05,
      "loss": 0.6033,
      "mean_token_accuracy": 0.8167394548654556,
      "num_tokens": 548675627.0,
      "step": 17191
    },
    {
      "entropy": 0.6323738279752433,
      "epoch": 1.5817562866228485,
      "grad_norm": 0.15809431672096252,
      "learning_rate": 4.7278191799306896e-05,
      "loss": 0.6149,
      "mean_token_accuracy": 0.8100833147764206,
      "num_tokens": 548707688.0,
      "step": 17192
    },
    {
      "entropy": 0.544045967515558,
      "epoch": 1.5818482929942899,
      "grad_norm": 0.1472608745098114,
      "learning_rate": 4.727512497316528e-05,
      "loss": 0.5217,
      "mean_token_accuracy": 0.8393196240067482,
      "num_tokens": 548739314.0,
      "step": 17193
    },
    {
      "entropy": 0.5705924723297358,
      "epoch": 1.5819402993657312,
      "grad_norm": 0.15503531694412231,
      "learning_rate": 4.7272058147023645e-05,
      "loss": 0.5418,
      "mean_token_accuracy": 0.8325949274003506,
      "num_tokens": 548771059.0,
      "step": 17194
    },
    {
      "entropy": 0.6420497884973884,
      "epoch": 1.5820323057371724,
      "grad_norm": 0.15600767731666565,
      "learning_rate": 4.726899132088202e-05,
      "loss": 0.6343,
      "mean_token_accuracy": 0.802443478256464,
      "num_tokens": 548802681.0,
      "step": 17195
    },
    {
      "entropy": 0.6875431779772043,
      "epoch": 1.5821243121086135,
      "grad_norm": 0.16178005933761597,
      "learning_rate": 4.7265924494740394e-05,
      "loss": 0.6823,
      "mean_token_accuracy": 0.7918109744787216,
      "num_tokens": 548834761.0,
      "step": 17196
    },
    {
      "entropy": 0.6364830294623971,
      "epoch": 1.5822163184800546,
      "grad_norm": 0.15671581029891968,
      "learning_rate": 4.726285766859877e-05,
      "loss": 0.6201,
      "mean_token_accuracy": 0.8069938533008099,
      "num_tokens": 548866852.0,
      "step": 17197
    },
    {
      "entropy": 0.5081106210127473,
      "epoch": 1.582308324851496,
      "grad_norm": 0.14226911962032318,
      "learning_rate": 4.7259790842457144e-05,
      "loss": 0.4924,
      "mean_token_accuracy": 0.847493764013052,
      "num_tokens": 548898757.0,
      "step": 17198
    },
    {
      "entropy": 0.6233567483723164,
      "epoch": 1.5824003312229373,
      "grad_norm": 0.1547093391418457,
      "learning_rate": 4.725672401631552e-05,
      "loss": 0.6005,
      "mean_token_accuracy": 0.8154957294464111,
      "num_tokens": 548931170.0,
      "step": 17199
    },
    {
      "entropy": 0.6385440360754728,
      "epoch": 1.5824923375943785,
      "grad_norm": 0.15307117998600006,
      "learning_rate": 4.7253657190173886e-05,
      "loss": 0.6287,
      "mean_token_accuracy": 0.8073010146617889,
      "num_tokens": 548962587.0,
      "step": 17200
    },
    {
      "entropy": 0.540330246090889,
      "epoch": 1.5825843439658196,
      "grad_norm": 0.15837565064430237,
      "learning_rate": 4.725059036403227e-05,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8364821821451187,
      "num_tokens": 548993264.0,
      "step": 17201
    },
    {
      "entropy": 0.5848226491361856,
      "epoch": 1.5826763503372607,
      "grad_norm": 0.1589815616607666,
      "learning_rate": 4.7247523537890635e-05,
      "loss": 0.5699,
      "mean_token_accuracy": 0.8249616324901581,
      "num_tokens": 549024897.0,
      "step": 17202
    },
    {
      "entropy": 0.6029954515397549,
      "epoch": 1.582768356708702,
      "grad_norm": 0.15680259466171265,
      "learning_rate": 4.724445671174901e-05,
      "loss": 0.589,
      "mean_token_accuracy": 0.8176603429019451,
      "num_tokens": 549056989.0,
      "step": 17203
    },
    {
      "entropy": 0.6277605094946921,
      "epoch": 1.5828603630801434,
      "grad_norm": 0.159760519862175,
      "learning_rate": 4.7241389885607385e-05,
      "loss": 0.6115,
      "mean_token_accuracy": 0.8137124888598919,
      "num_tokens": 549089193.0,
      "step": 17204
    },
    {
      "entropy": 0.6538015846163034,
      "epoch": 1.5829523694515846,
      "grad_norm": 0.1674542874097824,
      "learning_rate": 4.723832305946576e-05,
      "loss": 0.6541,
      "mean_token_accuracy": 0.8011259734630585,
      "num_tokens": 549120935.0,
      "step": 17205
    },
    {
      "entropy": 0.5382857890799642,
      "epoch": 1.5830443758230257,
      "grad_norm": 0.1626414954662323,
      "learning_rate": 4.723525623332414e-05,
      "loss": 0.5331,
      "mean_token_accuracy": 0.8359395228326321,
      "num_tokens": 549152692.0,
      "step": 17206
    },
    {
      "entropy": 0.750809321179986,
      "epoch": 1.5831363821944668,
      "grad_norm": 0.16542673110961914,
      "learning_rate": 4.723218940718251e-05,
      "loss": 0.7461,
      "mean_token_accuracy": 0.7740796953439713,
      "num_tokens": 549184378.0,
      "step": 17207
    },
    {
      "entropy": 0.5714062452316284,
      "epoch": 1.5832283885659082,
      "grad_norm": 0.16111581027507782,
      "learning_rate": 4.722912258104088e-05,
      "loss": 0.5713,
      "mean_token_accuracy": 0.8240353949368,
      "num_tokens": 549214879.0,
      "step": 17208
    },
    {
      "entropy": 0.57699572481215,
      "epoch": 1.5833203949373496,
      "grad_norm": 0.15801195800304413,
      "learning_rate": 4.722605575489926e-05,
      "loss": 0.5763,
      "mean_token_accuracy": 0.8221567794680595,
      "num_tokens": 549247040.0,
      "step": 17209
    },
    {
      "entropy": 0.5350518662016839,
      "epoch": 1.5834124013087907,
      "grad_norm": 0.15347115695476532,
      "learning_rate": 4.722298892875763e-05,
      "loss": 0.5324,
      "mean_token_accuracy": 0.839922621846199,
      "num_tokens": 549279232.0,
      "step": 17210
    },
    {
      "entropy": 0.5055956924334168,
      "epoch": 1.5835044076802318,
      "grad_norm": 0.14051887392997742,
      "learning_rate": 4.721992210261601e-05,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8463623523712158,
      "num_tokens": 549311942.0,
      "step": 17211
    },
    {
      "entropy": 0.662161267362535,
      "epoch": 1.583596414051673,
      "grad_norm": 0.15621881186962128,
      "learning_rate": 4.721685527647438e-05,
      "loss": 0.6438,
      "mean_token_accuracy": 0.8042758777737617,
      "num_tokens": 549344050.0,
      "step": 17212
    },
    {
      "entropy": 0.5788511577993631,
      "epoch": 1.5836884204231143,
      "grad_norm": 0.15280002355575562,
      "learning_rate": 4.721378845033275e-05,
      "loss": 0.563,
      "mean_token_accuracy": 0.8275604248046875,
      "num_tokens": 549375551.0,
      "step": 17213
    },
    {
      "entropy": 0.68495640065521,
      "epoch": 1.5837804267945557,
      "grad_norm": 0.15471190214157104,
      "learning_rate": 4.721072162419113e-05,
      "loss": 0.6647,
      "mean_token_accuracy": 0.7960507795214653,
      "num_tokens": 549407191.0,
      "step": 17214
    },
    {
      "entropy": 0.6706165550276637,
      "epoch": 1.5838724331659968,
      "grad_norm": 0.15801195800304413,
      "learning_rate": 4.72076547980495e-05,
      "loss": 0.663,
      "mean_token_accuracy": 0.7962225787341595,
      "num_tokens": 549439416.0,
      "step": 17215
    },
    {
      "entropy": 0.556857792660594,
      "epoch": 1.583964439537438,
      "grad_norm": 0.1522272378206253,
      "learning_rate": 4.7204587971907874e-05,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8313604146242142,
      "num_tokens": 549471748.0,
      "step": 17216
    },
    {
      "entropy": 0.7490406259894371,
      "epoch": 1.584056445908879,
      "grad_norm": 0.16492301225662231,
      "learning_rate": 4.720152114576625e-05,
      "loss": 0.7449,
      "mean_token_accuracy": 0.7768629677593708,
      "num_tokens": 549503903.0,
      "step": 17217
    },
    {
      "entropy": 0.6866011787205935,
      "epoch": 1.5841484522803204,
      "grad_norm": 0.15574538707733154,
      "learning_rate": 4.719845431962462e-05,
      "loss": 0.6667,
      "mean_token_accuracy": 0.7956314161419868,
      "num_tokens": 549535788.0,
      "step": 17218
    },
    {
      "entropy": 0.5577317951247096,
      "epoch": 1.5842404586517618,
      "grad_norm": 0.14592969417572021,
      "learning_rate": 4.7195387493483e-05,
      "loss": 0.5372,
      "mean_token_accuracy": 0.8388761952519417,
      "num_tokens": 549568488.0,
      "step": 17219
    },
    {
      "entropy": 0.6205584993585944,
      "epoch": 1.584332465023203,
      "grad_norm": 0.1500846892595291,
      "learning_rate": 4.719232066734137e-05,
      "loss": 0.6085,
      "mean_token_accuracy": 0.8141405694186687,
      "num_tokens": 549600925.0,
      "step": 17220
    },
    {
      "entropy": 0.5884916903451085,
      "epoch": 1.584424471394644,
      "grad_norm": 0.16235704720020294,
      "learning_rate": 4.718925384119974e-05,
      "loss": 0.5844,
      "mean_token_accuracy": 0.8219817355275154,
      "num_tokens": 549632533.0,
      "step": 17221
    },
    {
      "entropy": 0.6136264503002167,
      "epoch": 1.5845164777660852,
      "grad_norm": 0.1491910219192505,
      "learning_rate": 4.718618701505812e-05,
      "loss": 0.6092,
      "mean_token_accuracy": 0.8097474686801434,
      "num_tokens": 549664818.0,
      "step": 17222
    },
    {
      "entropy": 0.5960055589675903,
      "epoch": 1.5846084841375265,
      "grad_norm": 0.15174320340156555,
      "learning_rate": 4.718312018891649e-05,
      "loss": 0.596,
      "mean_token_accuracy": 0.8157007545232773,
      "num_tokens": 549696720.0,
      "step": 17223
    },
    {
      "entropy": 0.5920317489653826,
      "epoch": 1.5847004905089679,
      "grad_norm": 0.15170149505138397,
      "learning_rate": 4.718005336277487e-05,
      "loss": 0.5762,
      "mean_token_accuracy": 0.8183524012565613,
      "num_tokens": 549727704.0,
      "step": 17224
    },
    {
      "entropy": 0.5446124775335193,
      "epoch": 1.584792496880409,
      "grad_norm": 0.15731076896190643,
      "learning_rate": 4.717698653663324e-05,
      "loss": 0.5324,
      "mean_token_accuracy": 0.8348651938140392,
      "num_tokens": 549759447.0,
      "step": 17225
    },
    {
      "entropy": 0.5376856317743659,
      "epoch": 1.5848845032518502,
      "grad_norm": 0.152279332280159,
      "learning_rate": 4.717391971049161e-05,
      "loss": 0.5273,
      "mean_token_accuracy": 0.836134422570467,
      "num_tokens": 549791612.0,
      "step": 17226
    },
    {
      "entropy": 0.6301244162023067,
      "epoch": 1.5849765096232913,
      "grad_norm": 0.17057976126670837,
      "learning_rate": 4.717085288434999e-05,
      "loss": 0.6273,
      "mean_token_accuracy": 0.8050364479422569,
      "num_tokens": 549822785.0,
      "step": 17227
    },
    {
      "entropy": 0.6994412222411484,
      "epoch": 1.5850685159947326,
      "grad_norm": 0.16319745779037476,
      "learning_rate": 4.716778605820836e-05,
      "loss": 0.6918,
      "mean_token_accuracy": 0.7905564978718758,
      "num_tokens": 549855303.0,
      "step": 17228
    },
    {
      "entropy": 0.6840738642495126,
      "epoch": 1.585160522366174,
      "grad_norm": 0.157151460647583,
      "learning_rate": 4.716471923206674e-05,
      "loss": 0.6713,
      "mean_token_accuracy": 0.7970346733927727,
      "num_tokens": 549887571.0,
      "step": 17229
    },
    {
      "entropy": 0.5831845793873072,
      "epoch": 1.5852525287376151,
      "grad_norm": 0.14794357120990753,
      "learning_rate": 4.716165240592511e-05,
      "loss": 0.569,
      "mean_token_accuracy": 0.8206489980220795,
      "num_tokens": 549919736.0,
      "step": 17230
    },
    {
      "entropy": 0.6409462294541299,
      "epoch": 1.5853445351090563,
      "grad_norm": 0.15290318429470062,
      "learning_rate": 4.715858557978348e-05,
      "loss": 0.6346,
      "mean_token_accuracy": 0.804053321480751,
      "num_tokens": 549952376.0,
      "step": 17231
    },
    {
      "entropy": 0.4909524228423834,
      "epoch": 1.5854365414804974,
      "grad_norm": 0.1469520926475525,
      "learning_rate": 4.715551875364186e-05,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8531882651150227,
      "num_tokens": 549984572.0,
      "step": 17232
    },
    {
      "entropy": 0.6139780003577471,
      "epoch": 1.5855285478519388,
      "grad_norm": 0.15363387763500214,
      "learning_rate": 4.715245192750023e-05,
      "loss": 0.6019,
      "mean_token_accuracy": 0.8144874311983585,
      "num_tokens": 550017058.0,
      "step": 17233
    },
    {
      "entropy": 0.6234558783471584,
      "epoch": 1.58562055422338,
      "grad_norm": 0.15524861216545105,
      "learning_rate": 4.7149385101358604e-05,
      "loss": 0.6067,
      "mean_token_accuracy": 0.8150551915168762,
      "num_tokens": 550048602.0,
      "step": 17234
    },
    {
      "entropy": 0.6459231832996011,
      "epoch": 1.5857125605948212,
      "grad_norm": 0.15365895628929138,
      "learning_rate": 4.7146318275216985e-05,
      "loss": 0.6474,
      "mean_token_accuracy": 0.8008827194571495,
      "num_tokens": 550080403.0,
      "step": 17235
    },
    {
      "entropy": 0.5162313152104616,
      "epoch": 1.5858045669662624,
      "grad_norm": 0.14895856380462646,
      "learning_rate": 4.714325144907535e-05,
      "loss": 0.5196,
      "mean_token_accuracy": 0.8407998271286488,
      "num_tokens": 550113104.0,
      "step": 17236
    },
    {
      "entropy": 0.662294426932931,
      "epoch": 1.5858965733377035,
      "grad_norm": 0.15560385584831238,
      "learning_rate": 4.714018462293373e-05,
      "loss": 0.6608,
      "mean_token_accuracy": 0.802034754306078,
      "num_tokens": 550145418.0,
      "step": 17237
    },
    {
      "entropy": 0.5885314326733351,
      "epoch": 1.5859885797091449,
      "grad_norm": 0.14879687130451202,
      "learning_rate": 4.71371177967921e-05,
      "loss": 0.5741,
      "mean_token_accuracy": 0.8207731693983078,
      "num_tokens": 550178022.0,
      "step": 17238
    },
    {
      "entropy": 0.6515207188203931,
      "epoch": 1.5860805860805862,
      "grad_norm": 0.16127261519432068,
      "learning_rate": 4.713405097065048e-05,
      "loss": 0.64,
      "mean_token_accuracy": 0.8023936897516251,
      "num_tokens": 550209608.0,
      "step": 17239
    },
    {
      "entropy": 0.5392730655148625,
      "epoch": 1.5861725924520274,
      "grad_norm": 0.15322589874267578,
      "learning_rate": 4.713098414450885e-05,
      "loss": 0.5168,
      "mean_token_accuracy": 0.8409888558089733,
      "num_tokens": 550241747.0,
      "step": 17240
    },
    {
      "entropy": 0.5609770156443119,
      "epoch": 1.5862645988234685,
      "grad_norm": 0.1529446244239807,
      "learning_rate": 4.7127917318367226e-05,
      "loss": 0.5508,
      "mean_token_accuracy": 0.830734845250845,
      "num_tokens": 550274122.0,
      "step": 17241
    },
    {
      "entropy": 0.60217041708529,
      "epoch": 1.5863566051949096,
      "grad_norm": 0.15639027953147888,
      "learning_rate": 4.7124850492225594e-05,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8168587200343609,
      "num_tokens": 550306273.0,
      "step": 17242
    },
    {
      "entropy": 0.5604018177837133,
      "epoch": 1.586448611566351,
      "grad_norm": 0.15163147449493408,
      "learning_rate": 4.7121783666083976e-05,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8304190561175346,
      "num_tokens": 550338200.0,
      "step": 17243
    },
    {
      "entropy": 0.6630261484533548,
      "epoch": 1.5865406179377923,
      "grad_norm": 0.15654942393302917,
      "learning_rate": 4.7118716839942343e-05,
      "loss": 0.6525,
      "mean_token_accuracy": 0.7990017384290695,
      "num_tokens": 550370383.0,
      "step": 17244
    },
    {
      "entropy": 0.5996436954010278,
      "epoch": 1.5866326243092335,
      "grad_norm": 0.16125279664993286,
      "learning_rate": 4.7115650013800725e-05,
      "loss": 0.5975,
      "mean_token_accuracy": 0.8155657351016998,
      "num_tokens": 550402763.0,
      "step": 17245
    },
    {
      "entropy": 0.6178810866549611,
      "epoch": 1.5867246306806746,
      "grad_norm": 0.1627112627029419,
      "learning_rate": 4.711258318765909e-05,
      "loss": 0.599,
      "mean_token_accuracy": 0.8176997192203999,
      "num_tokens": 550433626.0,
      "step": 17246
    },
    {
      "entropy": 0.5638852615375072,
      "epoch": 1.5868166370521157,
      "grad_norm": 0.1578766256570816,
      "learning_rate": 4.710951636151747e-05,
      "loss": 0.5575,
      "mean_token_accuracy": 0.828500758856535,
      "num_tokens": 550465345.0,
      "step": 17247
    },
    {
      "entropy": 0.5962979970499873,
      "epoch": 1.586908643423557,
      "grad_norm": 0.16113600134849548,
      "learning_rate": 4.710644953537584e-05,
      "loss": 0.5809,
      "mean_token_accuracy": 0.8197517022490501,
      "num_tokens": 550496392.0,
      "step": 17248
    },
    {
      "entropy": 0.7058214694261551,
      "epoch": 1.5870006497949984,
      "grad_norm": 0.1652499884366989,
      "learning_rate": 4.710338270923422e-05,
      "loss": 0.6989,
      "mean_token_accuracy": 0.7899340093135834,
      "num_tokens": 550528034.0,
      "step": 17249
    },
    {
      "entropy": 0.5887055220082402,
      "epoch": 1.5870926561664396,
      "grad_norm": 0.15287108719348907,
      "learning_rate": 4.710031588309259e-05,
      "loss": 0.584,
      "mean_token_accuracy": 0.8192290142178535,
      "num_tokens": 550560285.0,
      "step": 17250
    },
    {
      "entropy": 0.5896537527441978,
      "epoch": 1.5871846625378807,
      "grad_norm": 0.16028954088687897,
      "learning_rate": 4.7097249056950966e-05,
      "loss": 0.593,
      "mean_token_accuracy": 0.8258451409637928,
      "num_tokens": 550592249.0,
      "step": 17251
    },
    {
      "entropy": 0.6136754211038351,
      "epoch": 1.5872766689093218,
      "grad_norm": 0.15454840660095215,
      "learning_rate": 4.7094182230809334e-05,
      "loss": 0.6172,
      "mean_token_accuracy": 0.8082726746797562,
      "num_tokens": 550623884.0,
      "step": 17252
    },
    {
      "entropy": 0.6450157668441534,
      "epoch": 1.5873686752807632,
      "grad_norm": 0.16006159782409668,
      "learning_rate": 4.7091115404667715e-05,
      "loss": 0.6423,
      "mean_token_accuracy": 0.8069131039083004,
      "num_tokens": 550655693.0,
      "step": 17253
    },
    {
      "entropy": 0.5651371823623776,
      "epoch": 1.5874606816522046,
      "grad_norm": 0.14800256490707397,
      "learning_rate": 4.708804857852608e-05,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8275998160243034,
      "num_tokens": 550688244.0,
      "step": 17254
    },
    {
      "entropy": 0.5283901263028383,
      "epoch": 1.5875526880236457,
      "grad_norm": 0.154303640127182,
      "learning_rate": 4.708498175238446e-05,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8440536260604858,
      "num_tokens": 550720055.0,
      "step": 17255
    },
    {
      "entropy": 0.6371227307245135,
      "epoch": 1.5876446943950868,
      "grad_norm": 0.16008360683918,
      "learning_rate": 4.708191492624283e-05,
      "loss": 0.6194,
      "mean_token_accuracy": 0.807068832218647,
      "num_tokens": 550752170.0,
      "step": 17256
    },
    {
      "entropy": 0.6753180585801601,
      "epoch": 1.587736700766528,
      "grad_norm": 0.15965324640274048,
      "learning_rate": 4.707884810010121e-05,
      "loss": 0.6735,
      "mean_token_accuracy": 0.7977302446961403,
      "num_tokens": 550784818.0,
      "step": 17257
    },
    {
      "entropy": 0.6486703054979444,
      "epoch": 1.5878287071379693,
      "grad_norm": 0.16003692150115967,
      "learning_rate": 4.707578127395958e-05,
      "loss": 0.6402,
      "mean_token_accuracy": 0.8083135820925236,
      "num_tokens": 550816056.0,
      "step": 17258
    },
    {
      "entropy": 0.5509741995483637,
      "epoch": 1.5879207135094107,
      "grad_norm": 0.14647430181503296,
      "learning_rate": 4.7072714447817956e-05,
      "loss": 0.5287,
      "mean_token_accuracy": 0.836028553545475,
      "num_tokens": 550848316.0,
      "step": 17259
    },
    {
      "entropy": 0.7313811052590609,
      "epoch": 1.5880127198808518,
      "grad_norm": 0.16274859011173248,
      "learning_rate": 4.7069647621676324e-05,
      "loss": 0.7253,
      "mean_token_accuracy": 0.7762185335159302,
      "num_tokens": 550880734.0,
      "step": 17260
    },
    {
      "entropy": 0.5992793855257332,
      "epoch": 1.588104726252293,
      "grad_norm": 0.15357574820518494,
      "learning_rate": 4.7066580795534706e-05,
      "loss": 0.5879,
      "mean_token_accuracy": 0.8191413395106792,
      "num_tokens": 550912854.0,
      "step": 17261
    },
    {
      "entropy": 0.6363813206553459,
      "epoch": 1.588196732623734,
      "grad_norm": 0.15664787590503693,
      "learning_rate": 4.7063513969393074e-05,
      "loss": 0.6254,
      "mean_token_accuracy": 0.8087003640830517,
      "num_tokens": 550944585.0,
      "step": 17262
    },
    {
      "entropy": 0.6962259151041508,
      "epoch": 1.5882887389951754,
      "grad_norm": 0.16010276973247528,
      "learning_rate": 4.706044714325145e-05,
      "loss": 0.6912,
      "mean_token_accuracy": 0.7898617014288902,
      "num_tokens": 550975915.0,
      "step": 17263
    },
    {
      "entropy": 0.5616347584873438,
      "epoch": 1.5883807453666168,
      "grad_norm": 0.15379948914051056,
      "learning_rate": 4.705738031710982e-05,
      "loss": 0.5315,
      "mean_token_accuracy": 0.8361615017056465,
      "num_tokens": 551007659.0,
      "step": 17264
    },
    {
      "entropy": 0.6712458934634924,
      "epoch": 1.588472751738058,
      "grad_norm": 0.16437719762325287,
      "learning_rate": 4.70543134909682e-05,
      "loss": 0.6584,
      "mean_token_accuracy": 0.801846444606781,
      "num_tokens": 551039203.0,
      "step": 17265
    },
    {
      "entropy": 0.6177470460534096,
      "epoch": 1.588564758109499,
      "grad_norm": 0.1483398675918579,
      "learning_rate": 4.705124666482658e-05,
      "loss": 0.5927,
      "mean_token_accuracy": 0.817857164889574,
      "num_tokens": 551071335.0,
      "step": 17266
    },
    {
      "entropy": 0.6138487420976162,
      "epoch": 1.5886567644809402,
      "grad_norm": 0.1558712124824524,
      "learning_rate": 4.704817983868495e-05,
      "loss": 0.5949,
      "mean_token_accuracy": 0.8167527131736279,
      "num_tokens": 551102875.0,
      "step": 17267
    },
    {
      "entropy": 0.604390730150044,
      "epoch": 1.5887487708523815,
      "grad_norm": 0.1586112231016159,
      "learning_rate": 4.704511301254332e-05,
      "loss": 0.5931,
      "mean_token_accuracy": 0.8178552910685539,
      "num_tokens": 551135225.0,
      "step": 17268
    },
    {
      "entropy": 0.588624058291316,
      "epoch": 1.5888407772238229,
      "grad_norm": 0.15183985233306885,
      "learning_rate": 4.7042046186401696e-05,
      "loss": 0.5839,
      "mean_token_accuracy": 0.8200203776359558,
      "num_tokens": 551166900.0,
      "step": 17269
    },
    {
      "entropy": 0.5790469534695148,
      "epoch": 1.588932783595264,
      "grad_norm": 0.14955031871795654,
      "learning_rate": 4.703897936026007e-05,
      "loss": 0.5686,
      "mean_token_accuracy": 0.8252292796969414,
      "num_tokens": 551198917.0,
      "step": 17270
    },
    {
      "entropy": 0.580485531128943,
      "epoch": 1.5890247899667052,
      "grad_norm": 0.1463015377521515,
      "learning_rate": 4.7035912534118445e-05,
      "loss": 0.5686,
      "mean_token_accuracy": 0.8223413228988647,
      "num_tokens": 551231354.0,
      "step": 17271
    },
    {
      "entropy": 0.5469827931374311,
      "epoch": 1.5891167963381463,
      "grad_norm": 0.1477474719285965,
      "learning_rate": 4.703284570797682e-05,
      "loss": 0.5311,
      "mean_token_accuracy": 0.8386106230318546,
      "num_tokens": 551264087.0,
      "step": 17272
    },
    {
      "entropy": 0.7104151956737041,
      "epoch": 1.5892088027095876,
      "grad_norm": 0.16337423026561737,
      "learning_rate": 4.702977888183519e-05,
      "loss": 0.694,
      "mean_token_accuracy": 0.7885334976017475,
      "num_tokens": 551295986.0,
      "step": 17273
    },
    {
      "entropy": 0.6901795826852322,
      "epoch": 1.589300809081029,
      "grad_norm": 0.16533319652080536,
      "learning_rate": 4.702671205569357e-05,
      "loss": 0.6938,
      "mean_token_accuracy": 0.7883811704814434,
      "num_tokens": 551328448.0,
      "step": 17274
    },
    {
      "entropy": 0.751216234639287,
      "epoch": 1.5893928154524701,
      "grad_norm": 0.16602833569049835,
      "learning_rate": 4.702364522955194e-05,
      "loss": 0.7472,
      "mean_token_accuracy": 0.7709835171699524,
      "num_tokens": 551360527.0,
      "step": 17275
    },
    {
      "entropy": 0.6028240267187357,
      "epoch": 1.5894848218239113,
      "grad_norm": 0.1609295755624771,
      "learning_rate": 4.702057840341031e-05,
      "loss": 0.5971,
      "mean_token_accuracy": 0.8172751106321812,
      "num_tokens": 551391432.0,
      "step": 17276
    },
    {
      "entropy": 0.6351351868361235,
      "epoch": 1.5895768281953524,
      "grad_norm": 0.15911827981472015,
      "learning_rate": 4.7017511577268686e-05,
      "loss": 0.6369,
      "mean_token_accuracy": 0.8095726035535336,
      "num_tokens": 551423454.0,
      "step": 17277
    },
    {
      "entropy": 0.5703206807374954,
      "epoch": 1.5896688345667938,
      "grad_norm": 0.1537005752325058,
      "learning_rate": 4.701444475112706e-05,
      "loss": 0.5721,
      "mean_token_accuracy": 0.8251103088259697,
      "num_tokens": 551454950.0,
      "step": 17278
    },
    {
      "entropy": 0.5044965464621782,
      "epoch": 1.589760840938235,
      "grad_norm": 0.1505553275346756,
      "learning_rate": 4.7011377924985436e-05,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8477720208466053,
      "num_tokens": 551486902.0,
      "step": 17279
    },
    {
      "entropy": 0.6661854749545455,
      "epoch": 1.5898528473096762,
      "grad_norm": 0.1586749106645584,
      "learning_rate": 4.700831109884381e-05,
      "loss": 0.6549,
      "mean_token_accuracy": 0.7995399720966816,
      "num_tokens": 551518699.0,
      "step": 17280
    },
    {
      "entropy": 0.5203749854117632,
      "epoch": 1.5899448536811174,
      "grad_norm": 0.15002048015594482,
      "learning_rate": 4.700524427270218e-05,
      "loss": 0.5,
      "mean_token_accuracy": 0.8449756875634193,
      "num_tokens": 551549873.0,
      "step": 17281
    },
    {
      "entropy": 0.6114677125588059,
      "epoch": 1.5900368600525585,
      "grad_norm": 0.1618233025074005,
      "learning_rate": 4.700217744656056e-05,
      "loss": 0.6043,
      "mean_token_accuracy": 0.813486073166132,
      "num_tokens": 551581673.0,
      "step": 17282
    },
    {
      "entropy": 0.5824228087440133,
      "epoch": 1.5901288664239999,
      "grad_norm": 0.16011349856853485,
      "learning_rate": 4.699911062041893e-05,
      "loss": 0.568,
      "mean_token_accuracy": 0.8278033696115017,
      "num_tokens": 551613710.0,
      "step": 17283
    },
    {
      "entropy": 0.6371847558766603,
      "epoch": 1.5902208727954412,
      "grad_norm": 0.1632065773010254,
      "learning_rate": 4.699604379427731e-05,
      "loss": 0.6344,
      "mean_token_accuracy": 0.805861659348011,
      "num_tokens": 551645656.0,
      "step": 17284
    },
    {
      "entropy": 0.6152529455721378,
      "epoch": 1.5903128791668824,
      "grad_norm": 0.1633043736219406,
      "learning_rate": 4.699297696813568e-05,
      "loss": 0.6075,
      "mean_token_accuracy": 0.8151079639792442,
      "num_tokens": 551677591.0,
      "step": 17285
    },
    {
      "entropy": 0.6570737045258284,
      "epoch": 1.5904048855383235,
      "grad_norm": 0.1527007818222046,
      "learning_rate": 4.698991014199405e-05,
      "loss": 0.6484,
      "mean_token_accuracy": 0.803346149623394,
      "num_tokens": 551709179.0,
      "step": 17286
    },
    {
      "entropy": 0.4886588007211685,
      "epoch": 1.5904968919097646,
      "grad_norm": 0.14598815143108368,
      "learning_rate": 4.6986843315852426e-05,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8524528741836548,
      "num_tokens": 551740623.0,
      "step": 17287
    },
    {
      "entropy": 0.49873096588999033,
      "epoch": 1.590588898281206,
      "grad_norm": 0.14676067233085632,
      "learning_rate": 4.69837764897108e-05,
      "loss": 0.481,
      "mean_token_accuracy": 0.8532067537307739,
      "num_tokens": 551772938.0,
      "step": 17288
    },
    {
      "entropy": 0.5998070668429136,
      "epoch": 1.5906809046526473,
      "grad_norm": 0.1517818123102188,
      "learning_rate": 4.6980709663569175e-05,
      "loss": 0.5838,
      "mean_token_accuracy": 0.8179526589810848,
      "num_tokens": 551804963.0,
      "step": 17289
    },
    {
      "entropy": 0.5808638706803322,
      "epoch": 1.5907729110240885,
      "grad_norm": 0.15512380003929138,
      "learning_rate": 4.697764283742755e-05,
      "loss": 0.583,
      "mean_token_accuracy": 0.8209508210420609,
      "num_tokens": 551836946.0,
      "step": 17290
    },
    {
      "entropy": 0.6079311668872833,
      "epoch": 1.5908649173955296,
      "grad_norm": 0.15314830839633942,
      "learning_rate": 4.697457601128592e-05,
      "loss": 0.602,
      "mean_token_accuracy": 0.8154666312038898,
      "num_tokens": 551868358.0,
      "step": 17291
    },
    {
      "entropy": 0.5560766793787479,
      "epoch": 1.5909569237669707,
      "grad_norm": 0.15331916511058807,
      "learning_rate": 4.69715091851443e-05,
      "loss": 0.5437,
      "mean_token_accuracy": 0.830791499465704,
      "num_tokens": 551900031.0,
      "step": 17292
    },
    {
      "entropy": 0.6265639625489712,
      "epoch": 1.591048930138412,
      "grad_norm": 0.15612289309501648,
      "learning_rate": 4.696844235900267e-05,
      "loss": 0.6122,
      "mean_token_accuracy": 0.8095423169434071,
      "num_tokens": 551932528.0,
      "step": 17293
    },
    {
      "entropy": 0.5418840693309903,
      "epoch": 1.5911409365098534,
      "grad_norm": 0.15064223110675812,
      "learning_rate": 4.696537553286104e-05,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8343319781124592,
      "num_tokens": 551963866.0,
      "step": 17294
    },
    {
      "entropy": 0.6492280755192041,
      "epoch": 1.5912329428812946,
      "grad_norm": 0.15549112856388092,
      "learning_rate": 4.6962308706719417e-05,
      "loss": 0.644,
      "mean_token_accuracy": 0.8050264157354832,
      "num_tokens": 551996448.0,
      "step": 17295
    },
    {
      "entropy": 0.6018771585077047,
      "epoch": 1.5913249492527357,
      "grad_norm": 0.15851108729839325,
      "learning_rate": 4.695924188057779e-05,
      "loss": 0.6007,
      "mean_token_accuracy": 0.8155762664973736,
      "num_tokens": 552028188.0,
      "step": 17296
    },
    {
      "entropy": 0.6163225779309869,
      "epoch": 1.5914169556241768,
      "grad_norm": 0.15719611942768097,
      "learning_rate": 4.6956175054436166e-05,
      "loss": 0.6019,
      "mean_token_accuracy": 0.8149137236177921,
      "num_tokens": 552059390.0,
      "step": 17297
    },
    {
      "entropy": 0.6732851676642895,
      "epoch": 1.5915089619956182,
      "grad_norm": 0.16010160744190216,
      "learning_rate": 4.695310822829454e-05,
      "loss": 0.6668,
      "mean_token_accuracy": 0.7957071289420128,
      "num_tokens": 552091079.0,
      "step": 17298
    },
    {
      "entropy": 0.616686801193282,
      "epoch": 1.5916009683670596,
      "grad_norm": 0.15954215824604034,
      "learning_rate": 4.6950041402152915e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8136047460138798,
      "num_tokens": 552123222.0,
      "step": 17299
    },
    {
      "entropy": 0.6276923306286335,
      "epoch": 1.5916929747385007,
      "grad_norm": 0.16188260912895203,
      "learning_rate": 4.694697457601129e-05,
      "loss": 0.6182,
      "mean_token_accuracy": 0.8074701987206936,
      "num_tokens": 552155045.0,
      "step": 17300
    },
    {
      "entropy": 0.5704368501901627,
      "epoch": 1.5917849811099418,
      "grad_norm": 0.14957940578460693,
      "learning_rate": 4.6943907749869664e-05,
      "loss": 0.5647,
      "mean_token_accuracy": 0.8258426897227764,
      "num_tokens": 552187813.0,
      "step": 17301
    },
    {
      "entropy": 0.7001447593793273,
      "epoch": 1.591876987481383,
      "grad_norm": 0.16497576236724854,
      "learning_rate": 4.694084092372803e-05,
      "loss": 0.6904,
      "mean_token_accuracy": 0.7913804724812508,
      "num_tokens": 552219810.0,
      "step": 17302
    },
    {
      "entropy": 0.5588367627933621,
      "epoch": 1.5919689938528243,
      "grad_norm": 0.15012028813362122,
      "learning_rate": 4.6937774097586414e-05,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8337580375373363,
      "num_tokens": 552252080.0,
      "step": 17303
    },
    {
      "entropy": 0.6399114951491356,
      "epoch": 1.5920610002242657,
      "grad_norm": 0.1560550183057785,
      "learning_rate": 4.693470727144478e-05,
      "loss": 0.6255,
      "mean_token_accuracy": 0.8058739490807056,
      "num_tokens": 552283127.0,
      "step": 17304
    },
    {
      "entropy": 0.5859512630850077,
      "epoch": 1.5921530065957068,
      "grad_norm": 0.14936354756355286,
      "learning_rate": 4.693164044530316e-05,
      "loss": 0.5684,
      "mean_token_accuracy": 0.824503980576992,
      "num_tokens": 552315258.0,
      "step": 17305
    },
    {
      "entropy": 0.5989923607558012,
      "epoch": 1.592245012967148,
      "grad_norm": 0.15759685635566711,
      "learning_rate": 4.692857361916153e-05,
      "loss": 0.5874,
      "mean_token_accuracy": 0.8223132155835629,
      "num_tokens": 552346886.0,
      "step": 17306
    },
    {
      "entropy": 0.5741000380367041,
      "epoch": 1.592337019338589,
      "grad_norm": 0.15495355427265167,
      "learning_rate": 4.6925506793019906e-05,
      "loss": 0.5658,
      "mean_token_accuracy": 0.8271766789257526,
      "num_tokens": 552378727.0,
      "step": 17307
    },
    {
      "entropy": 0.6923017147928476,
      "epoch": 1.5924290257100304,
      "grad_norm": 0.1670246720314026,
      "learning_rate": 4.692243996687828e-05,
      "loss": 0.6859,
      "mean_token_accuracy": 0.7910178601741791,
      "num_tokens": 552409715.0,
      "step": 17308
    },
    {
      "entropy": 0.6254730094224215,
      "epoch": 1.5925210320814718,
      "grad_norm": 0.15375085175037384,
      "learning_rate": 4.6919373140736655e-05,
      "loss": 0.6215,
      "mean_token_accuracy": 0.808133639395237,
      "num_tokens": 552441787.0,
      "step": 17309
    },
    {
      "entropy": 0.6858981437981129,
      "epoch": 1.592613038452913,
      "grad_norm": 0.1632765829563141,
      "learning_rate": 4.691630631459503e-05,
      "loss": 0.6779,
      "mean_token_accuracy": 0.7927988432347775,
      "num_tokens": 552473278.0,
      "step": 17310
    },
    {
      "entropy": 0.5817203763872385,
      "epoch": 1.592705044824354,
      "grad_norm": 0.14824102818965912,
      "learning_rate": 4.6913239488453404e-05,
      "loss": 0.567,
      "mean_token_accuracy": 0.8246191516518593,
      "num_tokens": 552505354.0,
      "step": 17311
    },
    {
      "entropy": 0.6088007632642984,
      "epoch": 1.5927970511957952,
      "grad_norm": 0.15529367327690125,
      "learning_rate": 4.691017266231177e-05,
      "loss": 0.5912,
      "mean_token_accuracy": 0.8190714605152607,
      "num_tokens": 552537065.0,
      "step": 17312
    },
    {
      "entropy": 0.6852185260504484,
      "epoch": 1.5928890575672365,
      "grad_norm": 0.15820324420928955,
      "learning_rate": 4.6907105836170153e-05,
      "loss": 0.6942,
      "mean_token_accuracy": 0.7883853204548359,
      "num_tokens": 552568280.0,
      "step": 17313
    },
    {
      "entropy": 0.5749502219259739,
      "epoch": 1.5929810639386779,
      "grad_norm": 0.1499762386083603,
      "learning_rate": 4.690403901002852e-05,
      "loss": 0.5467,
      "mean_token_accuracy": 0.8310773149132729,
      "num_tokens": 552600419.0,
      "step": 17314
    },
    {
      "entropy": 0.6743513187393546,
      "epoch": 1.593073070310119,
      "grad_norm": 0.1563340425491333,
      "learning_rate": 4.6900972183886896e-05,
      "loss": 0.6497,
      "mean_token_accuracy": 0.7994319051504135,
      "num_tokens": 552632796.0,
      "step": 17315
    },
    {
      "entropy": 0.5741426581516862,
      "epoch": 1.5931650766815602,
      "grad_norm": 0.1558190882205963,
      "learning_rate": 4.689790535774527e-05,
      "loss": 0.5691,
      "mean_token_accuracy": 0.8245677091181278,
      "num_tokens": 552664823.0,
      "step": 17316
    },
    {
      "entropy": 0.6764082461595535,
      "epoch": 1.5932570830530013,
      "grad_norm": 0.17044314742088318,
      "learning_rate": 4.6894838531603645e-05,
      "loss": 0.6768,
      "mean_token_accuracy": 0.7917434945702553,
      "num_tokens": 552696898.0,
      "step": 17317
    },
    {
      "entropy": 0.6795915495604277,
      "epoch": 1.5933490894244426,
      "grad_norm": 0.16298967599868774,
      "learning_rate": 4.689177170546202e-05,
      "loss": 0.6683,
      "mean_token_accuracy": 0.7950000874698162,
      "num_tokens": 552729213.0,
      "step": 17318
    },
    {
      "entropy": 0.6709051392972469,
      "epoch": 1.593441095795884,
      "grad_norm": 0.16085682809352875,
      "learning_rate": 4.6888704879320395e-05,
      "loss": 0.6649,
      "mean_token_accuracy": 0.7983705624938011,
      "num_tokens": 552760466.0,
      "step": 17319
    },
    {
      "entropy": 0.581931403838098,
      "epoch": 1.5935331021673251,
      "grad_norm": 0.15808609127998352,
      "learning_rate": 4.688563805317876e-05,
      "loss": 0.5572,
      "mean_token_accuracy": 0.828574176877737,
      "num_tokens": 552791719.0,
      "step": 17320
    },
    {
      "entropy": 0.660325912758708,
      "epoch": 1.5936251085387663,
      "grad_norm": 0.1593647301197052,
      "learning_rate": 4.6882571227037144e-05,
      "loss": 0.6534,
      "mean_token_accuracy": 0.7968180701136589,
      "num_tokens": 552823449.0,
      "step": 17321
    },
    {
      "entropy": 0.5641504023224115,
      "epoch": 1.5937171149102074,
      "grad_norm": 0.15422755479812622,
      "learning_rate": 4.687950440089551e-05,
      "loss": 0.5558,
      "mean_token_accuracy": 0.8300426341593266,
      "num_tokens": 552853614.0,
      "step": 17322
    },
    {
      "entropy": 0.7159838154911995,
      "epoch": 1.5938091212816488,
      "grad_norm": 0.16461056470870972,
      "learning_rate": 4.6876437574753886e-05,
      "loss": 0.7008,
      "mean_token_accuracy": 0.7864412404596806,
      "num_tokens": 552884959.0,
      "step": 17323
    },
    {
      "entropy": 0.6513334535993636,
      "epoch": 1.59390112765309,
      "grad_norm": 0.15887275338172913,
      "learning_rate": 4.687337074861226e-05,
      "loss": 0.6298,
      "mean_token_accuracy": 0.8059676587581635,
      "num_tokens": 552916197.0,
      "step": 17324
    },
    {
      "entropy": 0.6393599659204483,
      "epoch": 1.5939931340245312,
      "grad_norm": 0.1566588282585144,
      "learning_rate": 4.6870303922470636e-05,
      "loss": 0.6327,
      "mean_token_accuracy": 0.8097217865288258,
      "num_tokens": 552948584.0,
      "step": 17325
    },
    {
      "entropy": 0.611744724214077,
      "epoch": 1.5940851403959724,
      "grad_norm": 0.14927400648593903,
      "learning_rate": 4.686723709632901e-05,
      "loss": 0.593,
      "mean_token_accuracy": 0.818660344928503,
      "num_tokens": 552980321.0,
      "step": 17326
    },
    {
      "entropy": 0.7191246598958969,
      "epoch": 1.5941771467674135,
      "grad_norm": 0.16616950929164886,
      "learning_rate": 4.6864170270187385e-05,
      "loss": 0.7172,
      "mean_token_accuracy": 0.7857485897839069,
      "num_tokens": 553012385.0,
      "step": 17327
    },
    {
      "entropy": 0.5546397129073739,
      "epoch": 1.5942691531388549,
      "grad_norm": 0.16020987927913666,
      "learning_rate": 4.686110344404576e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.832588866353035,
      "num_tokens": 553044338.0,
      "step": 17328
    },
    {
      "entropy": 0.487130175344646,
      "epoch": 1.5943611595102962,
      "grad_norm": 0.14285194873809814,
      "learning_rate": 4.6858036617904134e-05,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8568311333656311,
      "num_tokens": 553076504.0,
      "step": 17329
    },
    {
      "entropy": 0.6549659371376038,
      "epoch": 1.5944531658817374,
      "grad_norm": 0.15642578899860382,
      "learning_rate": 4.685496979176251e-05,
      "loss": 0.6513,
      "mean_token_accuracy": 0.803568858653307,
      "num_tokens": 553108711.0,
      "step": 17330
    },
    {
      "entropy": 0.6695391032844782,
      "epoch": 1.5945451722531785,
      "grad_norm": 0.15802448987960815,
      "learning_rate": 4.6851902965620884e-05,
      "loss": 0.6552,
      "mean_token_accuracy": 0.797603253275156,
      "num_tokens": 553140342.0,
      "step": 17331
    },
    {
      "entropy": 0.5902714319527149,
      "epoch": 1.5946371786246196,
      "grad_norm": 0.1537695825099945,
      "learning_rate": 4.684883613947926e-05,
      "loss": 0.5871,
      "mean_token_accuracy": 0.8228741064667702,
      "num_tokens": 553172132.0,
      "step": 17332
    },
    {
      "entropy": 0.5869333776645362,
      "epoch": 1.594729184996061,
      "grad_norm": 0.15351589024066925,
      "learning_rate": 4.6845769313337626e-05,
      "loss": 0.5677,
      "mean_token_accuracy": 0.8212643191218376,
      "num_tokens": 553203499.0,
      "step": 17333
    },
    {
      "entropy": 0.5854882188141346,
      "epoch": 1.5948211913675023,
      "grad_norm": 0.15343625843524933,
      "learning_rate": 4.684270248719601e-05,
      "loss": 0.5824,
      "mean_token_accuracy": 0.817091565579176,
      "num_tokens": 553235270.0,
      "step": 17334
    },
    {
      "entropy": 0.6043896060436964,
      "epoch": 1.5949131977389435,
      "grad_norm": 0.15523186326026917,
      "learning_rate": 4.6839635661054375e-05,
      "loss": 0.5945,
      "mean_token_accuracy": 0.8197827972471714,
      "num_tokens": 553266699.0,
      "step": 17335
    },
    {
      "entropy": 0.557685773819685,
      "epoch": 1.5950052041103846,
      "grad_norm": 0.14470359683036804,
      "learning_rate": 4.683656883491275e-05,
      "loss": 0.544,
      "mean_token_accuracy": 0.8319027572870255,
      "num_tokens": 553299128.0,
      "step": 17336
    },
    {
      "entropy": 0.6527090482413769,
      "epoch": 1.5950972104818257,
      "grad_norm": 0.16196905076503754,
      "learning_rate": 4.6833502008771125e-05,
      "loss": 0.6396,
      "mean_token_accuracy": 0.8024529851973057,
      "num_tokens": 553330388.0,
      "step": 17337
    },
    {
      "entropy": 0.6705875135958195,
      "epoch": 1.595189216853267,
      "grad_norm": 0.16555774211883545,
      "learning_rate": 4.68304351826295e-05,
      "loss": 0.6608,
      "mean_token_accuracy": 0.7986540831625462,
      "num_tokens": 553362745.0,
      "step": 17338
    },
    {
      "entropy": 0.6184958070516586,
      "epoch": 1.5952812232247084,
      "grad_norm": 0.15874934196472168,
      "learning_rate": 4.6827368356487874e-05,
      "loss": 0.6159,
      "mean_token_accuracy": 0.8150686696171761,
      "num_tokens": 553394795.0,
      "step": 17339
    },
    {
      "entropy": 0.6948220878839493,
      "epoch": 1.5953732295961496,
      "grad_norm": 0.1616954505443573,
      "learning_rate": 4.682430153034625e-05,
      "loss": 0.6859,
      "mean_token_accuracy": 0.7942549139261246,
      "num_tokens": 553426882.0,
      "step": 17340
    },
    {
      "entropy": 0.5420425441116095,
      "epoch": 1.5954652359675907,
      "grad_norm": 0.15503790974617004,
      "learning_rate": 4.6821234704204616e-05,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8360644392669201,
      "num_tokens": 553459213.0,
      "step": 17341
    },
    {
      "entropy": 0.7408944517374039,
      "epoch": 1.5955572423390318,
      "grad_norm": 0.1665981411933899,
      "learning_rate": 4.6818167878063e-05,
      "loss": 0.7229,
      "mean_token_accuracy": 0.7790061384439468,
      "num_tokens": 553490463.0,
      "step": 17342
    },
    {
      "entropy": 0.5049485606141388,
      "epoch": 1.5956492487104732,
      "grad_norm": 0.15327543020248413,
      "learning_rate": 4.6815101051921366e-05,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8464972339570522,
      "num_tokens": 553521488.0,
      "step": 17343
    },
    {
      "entropy": 0.5328005936462432,
      "epoch": 1.5957412550819146,
      "grad_norm": 0.1496233195066452,
      "learning_rate": 4.681203422577975e-05,
      "loss": 0.5223,
      "mean_token_accuracy": 0.8394538052380085,
      "num_tokens": 553553927.0,
      "step": 17344
    },
    {
      "entropy": 0.5977550968527794,
      "epoch": 1.5958332614533557,
      "grad_norm": 0.16349059343338013,
      "learning_rate": 4.6808967399638115e-05,
      "loss": 0.5819,
      "mean_token_accuracy": 0.8209315948188305,
      "num_tokens": 553586615.0,
      "step": 17345
    },
    {
      "entropy": 0.6393994428217411,
      "epoch": 1.5959252678247968,
      "grad_norm": 0.15965870022773743,
      "learning_rate": 4.680590057349649e-05,
      "loss": 0.6258,
      "mean_token_accuracy": 0.8081825524568558,
      "num_tokens": 553618814.0,
      "step": 17346
    },
    {
      "entropy": 0.6184785850346088,
      "epoch": 1.596017274196238,
      "grad_norm": 0.1573093682527542,
      "learning_rate": 4.6802833747354864e-05,
      "loss": 0.6141,
      "mean_token_accuracy": 0.8113559223711491,
      "num_tokens": 553650149.0,
      "step": 17347
    },
    {
      "entropy": 0.43015290703624487,
      "epoch": 1.5961092805676793,
      "grad_norm": 0.14686349034309387,
      "learning_rate": 4.679976692121324e-05,
      "loss": 0.4103,
      "mean_token_accuracy": 0.870822586119175,
      "num_tokens": 553681860.0,
      "step": 17348
    },
    {
      "entropy": 0.7295316904783249,
      "epoch": 1.5962012869391207,
      "grad_norm": 0.16114556789398193,
      "learning_rate": 4.6796700095071614e-05,
      "loss": 0.716,
      "mean_token_accuracy": 0.7789799384772778,
      "num_tokens": 553714275.0,
      "step": 17349
    },
    {
      "entropy": 0.6360704991966486,
      "epoch": 1.5962932933105618,
      "grad_norm": 0.15841488540172577,
      "learning_rate": 4.679363326892999e-05,
      "loss": 0.6193,
      "mean_token_accuracy": 0.8087552934885025,
      "num_tokens": 553745966.0,
      "step": 17350
    },
    {
      "entropy": 0.6363892648369074,
      "epoch": 1.596385299682003,
      "grad_norm": 0.16327886283397675,
      "learning_rate": 4.6790566442788356e-05,
      "loss": 0.6091,
      "mean_token_accuracy": 0.8091661147773266,
      "num_tokens": 553777762.0,
      "step": 17351
    },
    {
      "entropy": 0.5341871734708548,
      "epoch": 1.596477306053444,
      "grad_norm": 0.15065963566303253,
      "learning_rate": 4.678749961664674e-05,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8372554630041122,
      "num_tokens": 553810198.0,
      "step": 17352
    },
    {
      "entropy": 0.6156328627839684,
      "epoch": 1.5965693124248854,
      "grad_norm": 0.15573406219482422,
      "learning_rate": 4.6784432790505105e-05,
      "loss": 0.6083,
      "mean_token_accuracy": 0.8146674558520317,
      "num_tokens": 553842346.0,
      "step": 17353
    },
    {
      "entropy": 0.6341654844582081,
      "epoch": 1.5966613187963268,
      "grad_norm": 0.15608395636081696,
      "learning_rate": 4.678136596436348e-05,
      "loss": 0.6245,
      "mean_token_accuracy": 0.8111165016889572,
      "num_tokens": 553873561.0,
      "step": 17354
    },
    {
      "entropy": 0.6129954988136888,
      "epoch": 1.596753325167768,
      "grad_norm": 0.15097665786743164,
      "learning_rate": 4.6778299138221855e-05,
      "loss": 0.6004,
      "mean_token_accuracy": 0.8188612051308155,
      "num_tokens": 553904985.0,
      "step": 17355
    },
    {
      "entropy": 0.6525896023958921,
      "epoch": 1.596845331539209,
      "grad_norm": 0.1562524139881134,
      "learning_rate": 4.677523231208023e-05,
      "loss": 0.6434,
      "mean_token_accuracy": 0.8014295995235443,
      "num_tokens": 553937577.0,
      "step": 17356
    },
    {
      "entropy": 0.6953767659142613,
      "epoch": 1.5969373379106502,
      "grad_norm": 0.15795747935771942,
      "learning_rate": 4.6772165485938604e-05,
      "loss": 0.6895,
      "mean_token_accuracy": 0.7938313968479633,
      "num_tokens": 553969693.0,
      "step": 17357
    },
    {
      "entropy": 0.5596385090611875,
      "epoch": 1.5970293442820915,
      "grad_norm": 0.15870392322540283,
      "learning_rate": 4.676909865979698e-05,
      "loss": 0.5563,
      "mean_token_accuracy": 0.827978502959013,
      "num_tokens": 554001551.0,
      "step": 17358
    },
    {
      "entropy": 0.6633436139672995,
      "epoch": 1.597121350653533,
      "grad_norm": 0.16186195611953735,
      "learning_rate": 4.676603183365535e-05,
      "loss": 0.6528,
      "mean_token_accuracy": 0.8003555461764336,
      "num_tokens": 554032801.0,
      "step": 17359
    },
    {
      "entropy": 0.780376885086298,
      "epoch": 1.597213357024974,
      "grad_norm": 0.1656826287508011,
      "learning_rate": 4.676296500751373e-05,
      "loss": 0.7813,
      "mean_token_accuracy": 0.7677910290658474,
      "num_tokens": 554065140.0,
      "step": 17360
    },
    {
      "entropy": 0.7011492177844048,
      "epoch": 1.5973053633964152,
      "grad_norm": 0.15487182140350342,
      "learning_rate": 4.67598981813721e-05,
      "loss": 0.695,
      "mean_token_accuracy": 0.7892970778048038,
      "num_tokens": 554097692.0,
      "step": 17361
    },
    {
      "entropy": 0.6676167491823435,
      "epoch": 1.5973973697678563,
      "grad_norm": 0.15522554516792297,
      "learning_rate": 4.675683135523047e-05,
      "loss": 0.6607,
      "mean_token_accuracy": 0.7987304702401161,
      "num_tokens": 554130373.0,
      "step": 17362
    },
    {
      "entropy": 0.5643365134019405,
      "epoch": 1.5974893761392976,
      "grad_norm": 0.14554692804813385,
      "learning_rate": 4.675376452908885e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8322933465242386,
      "num_tokens": 554162634.0,
      "step": 17363
    },
    {
      "entropy": 0.6483741868287325,
      "epoch": 1.597581382510739,
      "grad_norm": 0.15696200728416443,
      "learning_rate": 4.675069770294722e-05,
      "loss": 0.6391,
      "mean_token_accuracy": 0.8084573708474636,
      "num_tokens": 554194041.0,
      "step": 17364
    },
    {
      "entropy": 0.5763349067419767,
      "epoch": 1.5976733888821801,
      "grad_norm": 0.1469612717628479,
      "learning_rate": 4.67476308768056e-05,
      "loss": 0.5591,
      "mean_token_accuracy": 0.8250813260674477,
      "num_tokens": 554225780.0,
      "step": 17365
    },
    {
      "entropy": 0.5674100443720818,
      "epoch": 1.5977653952536213,
      "grad_norm": 0.15638798475265503,
      "learning_rate": 4.674456405066397e-05,
      "loss": 0.5438,
      "mean_token_accuracy": 0.8301015198230743,
      "num_tokens": 554256131.0,
      "step": 17366
    },
    {
      "entropy": 0.6860989220440388,
      "epoch": 1.5978574016250624,
      "grad_norm": 0.16196322441101074,
      "learning_rate": 4.6741497224522344e-05,
      "loss": 0.6788,
      "mean_token_accuracy": 0.7921440713107586,
      "num_tokens": 554288015.0,
      "step": 17367
    },
    {
      "entropy": 0.6732331775128841,
      "epoch": 1.5979494079965038,
      "grad_norm": 0.17033794522285461,
      "learning_rate": 4.673843039838072e-05,
      "loss": 0.6798,
      "mean_token_accuracy": 0.7956615798175335,
      "num_tokens": 554319012.0,
      "step": 17368
    },
    {
      "entropy": 0.5934430938214064,
      "epoch": 1.5980414143679451,
      "grad_norm": 0.15524382889270782,
      "learning_rate": 4.673536357223909e-05,
      "loss": 0.5841,
      "mean_token_accuracy": 0.8233371302485466,
      "num_tokens": 554351544.0,
      "step": 17369
    },
    {
      "entropy": 0.6707937149330974,
      "epoch": 1.5981334207393862,
      "grad_norm": 0.1583922654390335,
      "learning_rate": 4.673229674609747e-05,
      "loss": 0.6659,
      "mean_token_accuracy": 0.8008548952639103,
      "num_tokens": 554383622.0,
      "step": 17370
    },
    {
      "entropy": 0.6520941406488419,
      "epoch": 1.5982254271108274,
      "grad_norm": 0.15357105433940887,
      "learning_rate": 4.672922991995584e-05,
      "loss": 0.6488,
      "mean_token_accuracy": 0.8021370805799961,
      "num_tokens": 554415317.0,
      "step": 17371
    },
    {
      "entropy": 0.5562971495091915,
      "epoch": 1.5983174334822685,
      "grad_norm": 0.1588454246520996,
      "learning_rate": 4.672616309381421e-05,
      "loss": 0.5349,
      "mean_token_accuracy": 0.8327007703483105,
      "num_tokens": 554446324.0,
      "step": 17372
    },
    {
      "entropy": 0.6313788965344429,
      "epoch": 1.5984094398537099,
      "grad_norm": 0.16209134459495544,
      "learning_rate": 4.672309626767259e-05,
      "loss": 0.6202,
      "mean_token_accuracy": 0.8076026923954487,
      "num_tokens": 554476323.0,
      "step": 17373
    },
    {
      "entropy": 0.5935082565993071,
      "epoch": 1.5985014462251512,
      "grad_norm": 0.15915243327617645,
      "learning_rate": 4.672002944153096e-05,
      "loss": 0.5846,
      "mean_token_accuracy": 0.8233541436493397,
      "num_tokens": 554508427.0,
      "step": 17374
    },
    {
      "entropy": 0.6498674266040325,
      "epoch": 1.5985934525965924,
      "grad_norm": 0.15779276192188263,
      "learning_rate": 4.6716962615389334e-05,
      "loss": 0.6366,
      "mean_token_accuracy": 0.8041076436638832,
      "num_tokens": 554540292.0,
      "step": 17375
    },
    {
      "entropy": 0.5639319894835353,
      "epoch": 1.5986854589680335,
      "grad_norm": 0.15314066410064697,
      "learning_rate": 4.671389578924771e-05,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8346424959599972,
      "num_tokens": 554572377.0,
      "step": 17376
    },
    {
      "entropy": 0.618813281878829,
      "epoch": 1.5987774653394746,
      "grad_norm": 0.15886341035366058,
      "learning_rate": 4.6710828963106083e-05,
      "loss": 0.6068,
      "mean_token_accuracy": 0.815531849861145,
      "num_tokens": 554603688.0,
      "step": 17377
    },
    {
      "entropy": 0.5774128381162882,
      "epoch": 1.598869471710916,
      "grad_norm": 0.15727902948856354,
      "learning_rate": 4.670776213696446e-05,
      "loss": 0.5587,
      "mean_token_accuracy": 0.8294234946370125,
      "num_tokens": 554636078.0,
      "step": 17378
    },
    {
      "entropy": 0.5373574197292328,
      "epoch": 1.5989614780823573,
      "grad_norm": 0.14827962219715118,
      "learning_rate": 4.670469531082283e-05,
      "loss": 0.528,
      "mean_token_accuracy": 0.8370965234935284,
      "num_tokens": 554668685.0,
      "step": 17379
    },
    {
      "entropy": 0.64363818988204,
      "epoch": 1.5990534844537985,
      "grad_norm": 0.16230174899101257,
      "learning_rate": 4.67016284846812e-05,
      "loss": 0.6223,
      "mean_token_accuracy": 0.8092714622616768,
      "num_tokens": 554700383.0,
      "step": 17380
    },
    {
      "entropy": 0.7508567459881306,
      "epoch": 1.5991454908252396,
      "grad_norm": 0.16914485394954681,
      "learning_rate": 4.669856165853958e-05,
      "loss": 0.7493,
      "mean_token_accuracy": 0.7754619270563126,
      "num_tokens": 554732560.0,
      "step": 17381
    },
    {
      "entropy": 0.5424686577171087,
      "epoch": 1.5992374971966807,
      "grad_norm": 0.14770469069480896,
      "learning_rate": 4.669549483239795e-05,
      "loss": 0.522,
      "mean_token_accuracy": 0.8400966711342335,
      "num_tokens": 554764693.0,
      "step": 17382
    },
    {
      "entropy": 0.7045532492920756,
      "epoch": 1.599329503568122,
      "grad_norm": 0.16288089752197266,
      "learning_rate": 4.6692428006256325e-05,
      "loss": 0.7071,
      "mean_token_accuracy": 0.7877545692026615,
      "num_tokens": 554797224.0,
      "step": 17383
    },
    {
      "entropy": 0.6958820540457964,
      "epoch": 1.5994215099395634,
      "grad_norm": 0.1637406051158905,
      "learning_rate": 4.66893611801147e-05,
      "loss": 0.6838,
      "mean_token_accuracy": 0.7900810539722443,
      "num_tokens": 554827910.0,
      "step": 17384
    },
    {
      "entropy": 0.5279269618913531,
      "epoch": 1.5995135163110046,
      "grad_norm": 0.14282412827014923,
      "learning_rate": 4.6686294353973074e-05,
      "loss": 0.5077,
      "mean_token_accuracy": 0.8403127677738667,
      "num_tokens": 554860342.0,
      "step": 17385
    },
    {
      "entropy": 0.7331554889678955,
      "epoch": 1.5996055226824457,
      "grad_norm": 0.1638190597295761,
      "learning_rate": 4.668322752783145e-05,
      "loss": 0.7274,
      "mean_token_accuracy": 0.7794185020029545,
      "num_tokens": 554892189.0,
      "step": 17386
    },
    {
      "entropy": 0.5877863112837076,
      "epoch": 1.5996975290538868,
      "grad_norm": 0.1552312672138214,
      "learning_rate": 4.668016070168982e-05,
      "loss": 0.5702,
      "mean_token_accuracy": 0.8224539496004581,
      "num_tokens": 554924567.0,
      "step": 17387
    },
    {
      "entropy": 0.5077908518724144,
      "epoch": 1.5997895354253282,
      "grad_norm": 0.14254461228847504,
      "learning_rate": 4.66770938755482e-05,
      "loss": 0.4992,
      "mean_token_accuracy": 0.8436728082597256,
      "num_tokens": 554956943.0,
      "step": 17388
    },
    {
      "entropy": 0.7204472813755274,
      "epoch": 1.5998815417967696,
      "grad_norm": 0.1686946004629135,
      "learning_rate": 4.667402704940657e-05,
      "loss": 0.7245,
      "mean_token_accuracy": 0.7798423245549202,
      "num_tokens": 554988852.0,
      "step": 17389
    },
    {
      "entropy": 0.6117822043597698,
      "epoch": 1.5999735481682107,
      "grad_norm": 0.16404588520526886,
      "learning_rate": 4.667096022326495e-05,
      "loss": 0.6078,
      "mean_token_accuracy": 0.8146913275122643,
      "num_tokens": 555020504.0,
      "step": 17390
    },
    {
      "entropy": 0.5573426019400358,
      "epoch": 1.6000655545396518,
      "grad_norm": 0.15084004402160645,
      "learning_rate": 4.666789339712332e-05,
      "loss": 0.5504,
      "mean_token_accuracy": 0.8275575339794159,
      "num_tokens": 555053002.0,
      "step": 17391
    },
    {
      "entropy": 0.5700408294796944,
      "epoch": 1.600157560911093,
      "grad_norm": 0.15526887774467468,
      "learning_rate": 4.6664826570981696e-05,
      "loss": 0.56,
      "mean_token_accuracy": 0.8289894945919514,
      "num_tokens": 555085030.0,
      "step": 17392
    },
    {
      "entropy": 0.574810535646975,
      "epoch": 1.6002495672825343,
      "grad_norm": 0.15959812700748444,
      "learning_rate": 4.6661759744840064e-05,
      "loss": 0.5694,
      "mean_token_accuracy": 0.8226217404007912,
      "num_tokens": 555117011.0,
      "step": 17393
    },
    {
      "entropy": 0.6111467089504004,
      "epoch": 1.6003415736539757,
      "grad_norm": 0.15752367675304413,
      "learning_rate": 4.6658692918698446e-05,
      "loss": 0.6094,
      "mean_token_accuracy": 0.814623199403286,
      "num_tokens": 555149486.0,
      "step": 17394
    },
    {
      "entropy": 0.687653646338731,
      "epoch": 1.6004335800254168,
      "grad_norm": 0.16518686711788177,
      "learning_rate": 4.6655626092556814e-05,
      "loss": 0.6808,
      "mean_token_accuracy": 0.793267510831356,
      "num_tokens": 555181298.0,
      "step": 17395
    },
    {
      "entropy": 0.5651208069175482,
      "epoch": 1.600525586396858,
      "grad_norm": 0.15643475949764252,
      "learning_rate": 4.665255926641519e-05,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8288239501416683,
      "num_tokens": 555213473.0,
      "step": 17396
    },
    {
      "entropy": 0.7351449392735958,
      "epoch": 1.600617592768299,
      "grad_norm": 0.16548268496990204,
      "learning_rate": 4.664949244027356e-05,
      "loss": 0.7213,
      "mean_token_accuracy": 0.7808121927082539,
      "num_tokens": 555245434.0,
      "step": 17397
    },
    {
      "entropy": 0.6827635113149881,
      "epoch": 1.6007095991397404,
      "grad_norm": 0.15653863549232483,
      "learning_rate": 4.664642561413194e-05,
      "loss": 0.6667,
      "mean_token_accuracy": 0.7907703928649426,
      "num_tokens": 555276428.0,
      "step": 17398
    },
    {
      "entropy": 0.5464648026973009,
      "epoch": 1.6008016055111818,
      "grad_norm": 0.1542312651872635,
      "learning_rate": 4.664335878799031e-05,
      "loss": 0.5275,
      "mean_token_accuracy": 0.8388038799166679,
      "num_tokens": 555307873.0,
      "step": 17399
    },
    {
      "entropy": 0.6879167277365923,
      "epoch": 1.600893611882623,
      "grad_norm": 0.16209474205970764,
      "learning_rate": 4.664029196184869e-05,
      "loss": 0.6733,
      "mean_token_accuracy": 0.7940997704863548,
      "num_tokens": 555339876.0,
      "step": 17400
    },
    {
      "entropy": 0.5960098472423851,
      "epoch": 1.600985618254064,
      "grad_norm": 0.1486070603132248,
      "learning_rate": 4.6637225135707055e-05,
      "loss": 0.575,
      "mean_token_accuracy": 0.8237071521580219,
      "num_tokens": 555372232.0,
      "step": 17401
    },
    {
      "entropy": 0.6319793239235878,
      "epoch": 1.6010776246255052,
      "grad_norm": 0.1531968116760254,
      "learning_rate": 4.6634158309565436e-05,
      "loss": 0.6241,
      "mean_token_accuracy": 0.8111091777682304,
      "num_tokens": 555404620.0,
      "step": 17402
    },
    {
      "entropy": 0.5653302357532084,
      "epoch": 1.6011696309969465,
      "grad_norm": 0.15044227242469788,
      "learning_rate": 4.6631091483423804e-05,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8299350999295712,
      "num_tokens": 555435939.0,
      "step": 17403
    },
    {
      "entropy": 0.6437962444033474,
      "epoch": 1.601261637368388,
      "grad_norm": 0.15396080911159515,
      "learning_rate": 4.6628024657282185e-05,
      "loss": 0.6372,
      "mean_token_accuracy": 0.8018505610525608,
      "num_tokens": 555467979.0,
      "step": 17404
    },
    {
      "entropy": 0.5361565034836531,
      "epoch": 1.601353643739829,
      "grad_norm": 0.14957186579704285,
      "learning_rate": 4.662495783114055e-05,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8383483812212944,
      "num_tokens": 555500676.0,
      "step": 17405
    },
    {
      "entropy": 0.7581098861992359,
      "epoch": 1.6014456501112702,
      "grad_norm": 0.1657324731349945,
      "learning_rate": 4.662189100499893e-05,
      "loss": 0.7392,
      "mean_token_accuracy": 0.777847845107317,
      "num_tokens": 555532431.0,
      "step": 17406
    },
    {
      "entropy": 0.6278737708926201,
      "epoch": 1.6015376564827113,
      "grad_norm": 0.1585872620344162,
      "learning_rate": 4.66188241788573e-05,
      "loss": 0.6191,
      "mean_token_accuracy": 0.8102063834667206,
      "num_tokens": 555564412.0,
      "step": 17407
    },
    {
      "entropy": 0.636494911275804,
      "epoch": 1.6016296628541526,
      "grad_norm": 0.1604262888431549,
      "learning_rate": 4.661575735271568e-05,
      "loss": 0.6282,
      "mean_token_accuracy": 0.8080724701285362,
      "num_tokens": 555596166.0,
      "step": 17408
    },
    {
      "entropy": 0.6151054911315441,
      "epoch": 1.601721669225594,
      "grad_norm": 0.15621431171894073,
      "learning_rate": 4.661269052657405e-05,
      "loss": 0.6132,
      "mean_token_accuracy": 0.8134359121322632,
      "num_tokens": 555628039.0,
      "step": 17409
    },
    {
      "entropy": 0.6555610438808799,
      "epoch": 1.6018136755970351,
      "grad_norm": 0.15894602239131927,
      "learning_rate": 4.6609623700432427e-05,
      "loss": 0.6387,
      "mean_token_accuracy": 0.8045738376677036,
      "num_tokens": 555659615.0,
      "step": 17410
    },
    {
      "entropy": 0.5372354043647647,
      "epoch": 1.6019056819684763,
      "grad_norm": 0.1523684561252594,
      "learning_rate": 4.6606556874290794e-05,
      "loss": 0.5284,
      "mean_token_accuracy": 0.8377204164862633,
      "num_tokens": 555691238.0,
      "step": 17411
    },
    {
      "entropy": 0.5883991168811917,
      "epoch": 1.6019976883399174,
      "grad_norm": 0.1539577841758728,
      "learning_rate": 4.6603490048149176e-05,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8257290758192539,
      "num_tokens": 555723393.0,
      "step": 17412
    },
    {
      "entropy": 0.5568088572472334,
      "epoch": 1.6020896947113588,
      "grad_norm": 0.15090765058994293,
      "learning_rate": 4.6600423222007544e-05,
      "loss": 0.5622,
      "mean_token_accuracy": 0.8300878033041954,
      "num_tokens": 555755900.0,
      "step": 17413
    },
    {
      "entropy": 0.6037592738866806,
      "epoch": 1.6021817010828001,
      "grad_norm": 0.15200965106487274,
      "learning_rate": 4.659735639586592e-05,
      "loss": 0.5988,
      "mean_token_accuracy": 0.8176746927201748,
      "num_tokens": 555788369.0,
      "step": 17414
    },
    {
      "entropy": 0.600107479840517,
      "epoch": 1.6022737074542412,
      "grad_norm": 0.15975329279899597,
      "learning_rate": 4.659428956972429e-05,
      "loss": 0.5777,
      "mean_token_accuracy": 0.8211971260607243,
      "num_tokens": 555819532.0,
      "step": 17415
    },
    {
      "entropy": 0.5849685196299106,
      "epoch": 1.6023657138256824,
      "grad_norm": 0.15122264623641968,
      "learning_rate": 4.659122274358267e-05,
      "loss": 0.5638,
      "mean_token_accuracy": 0.8272525891661644,
      "num_tokens": 555851182.0,
      "step": 17416
    },
    {
      "entropy": 0.4435387868434191,
      "epoch": 1.6024577201971235,
      "grad_norm": 0.14218206703662872,
      "learning_rate": 4.658815591744104e-05,
      "loss": 0.4348,
      "mean_token_accuracy": 0.8657281138002872,
      "num_tokens": 555882924.0,
      "step": 17417
    },
    {
      "entropy": 0.6084568463265896,
      "epoch": 1.6025497265685649,
      "grad_norm": 0.1567726731300354,
      "learning_rate": 4.658508909129942e-05,
      "loss": 0.5953,
      "mean_token_accuracy": 0.8159128166735172,
      "num_tokens": 555915427.0,
      "step": 17418
    },
    {
      "entropy": 0.615587342530489,
      "epoch": 1.6026417329400062,
      "grad_norm": 0.1627284437417984,
      "learning_rate": 4.658202226515779e-05,
      "loss": 0.6012,
      "mean_token_accuracy": 0.8148558735847473,
      "num_tokens": 555947486.0,
      "step": 17419
    },
    {
      "entropy": 0.6105477046221495,
      "epoch": 1.6027337393114474,
      "grad_norm": 0.16122952103614807,
      "learning_rate": 4.6578955439016166e-05,
      "loss": 0.5987,
      "mean_token_accuracy": 0.8131947480142117,
      "num_tokens": 555979498.0,
      "step": 17420
    },
    {
      "entropy": 0.671414190903306,
      "epoch": 1.6028257456828885,
      "grad_norm": 0.1613941490650177,
      "learning_rate": 4.657588861287454e-05,
      "loss": 0.674,
      "mean_token_accuracy": 0.7943513430655003,
      "num_tokens": 556011013.0,
      "step": 17421
    },
    {
      "entropy": 0.5290704686194658,
      "epoch": 1.6029177520543296,
      "grad_norm": 0.14482399821281433,
      "learning_rate": 4.657282178673291e-05,
      "loss": 0.5223,
      "mean_token_accuracy": 0.8403506018221378,
      "num_tokens": 556043300.0,
      "step": 17422
    },
    {
      "entropy": 0.6486639939248562,
      "epoch": 1.603009758425771,
      "grad_norm": 0.161046102643013,
      "learning_rate": 4.656975496059129e-05,
      "loss": 0.6463,
      "mean_token_accuracy": 0.8067154176533222,
      "num_tokens": 556075763.0,
      "step": 17423
    },
    {
      "entropy": 0.5796013288199902,
      "epoch": 1.6031017647972123,
      "grad_norm": 0.15155676007270813,
      "learning_rate": 4.656668813444966e-05,
      "loss": 0.5653,
      "mean_token_accuracy": 0.8221333436667919,
      "num_tokens": 556107692.0,
      "step": 17424
    },
    {
      "entropy": 0.6057600523345172,
      "epoch": 1.6031937711686535,
      "grad_norm": 0.15587694942951202,
      "learning_rate": 4.656362130830804e-05,
      "loss": 0.5859,
      "mean_token_accuracy": 0.8177054822444916,
      "num_tokens": 556138835.0,
      "step": 17425
    },
    {
      "entropy": 0.5848910305649042,
      "epoch": 1.6032857775400946,
      "grad_norm": 0.15539799630641937,
      "learning_rate": 4.656055448216641e-05,
      "loss": 0.5743,
      "mean_token_accuracy": 0.8194634467363358,
      "num_tokens": 556170980.0,
      "step": 17426
    },
    {
      "entropy": 0.649871401488781,
      "epoch": 1.6033777839115357,
      "grad_norm": 0.1569100171327591,
      "learning_rate": 4.655748765602478e-05,
      "loss": 0.63,
      "mean_token_accuracy": 0.8115618340671062,
      "num_tokens": 556203145.0,
      "step": 17427
    },
    {
      "entropy": 0.5483626648783684,
      "epoch": 1.603469790282977,
      "grad_norm": 0.15711534023284912,
      "learning_rate": 4.6554420829883157e-05,
      "loss": 0.5329,
      "mean_token_accuracy": 0.8366206660866737,
      "num_tokens": 556235913.0,
      "step": 17428
    },
    {
      "entropy": 0.5581258721649647,
      "epoch": 1.6035617966544184,
      "grad_norm": 0.15202729403972626,
      "learning_rate": 4.655135400374153e-05,
      "loss": 0.5449,
      "mean_token_accuracy": 0.829258669167757,
      "num_tokens": 556268393.0,
      "step": 17429
    },
    {
      "entropy": 0.6661475114524364,
      "epoch": 1.6036538030258596,
      "grad_norm": 0.16303324699401855,
      "learning_rate": 4.6548287177599906e-05,
      "loss": 0.6618,
      "mean_token_accuracy": 0.7948164455592632,
      "num_tokens": 556300280.0,
      "step": 17430
    },
    {
      "entropy": 0.6824828116223216,
      "epoch": 1.6037458093973007,
      "grad_norm": 0.1570524275302887,
      "learning_rate": 4.654522035145828e-05,
      "loss": 0.6653,
      "mean_token_accuracy": 0.7967737689614296,
      "num_tokens": 556332365.0,
      "step": 17431
    },
    {
      "entropy": 0.547926114872098,
      "epoch": 1.6038378157687418,
      "grad_norm": 0.14455169439315796,
      "learning_rate": 4.654215352531665e-05,
      "loss": 0.528,
      "mean_token_accuracy": 0.8394347801804543,
      "num_tokens": 556364090.0,
      "step": 17432
    },
    {
      "entropy": 0.6645328244194388,
      "epoch": 1.6039298221401832,
      "grad_norm": 0.15387599170207977,
      "learning_rate": 4.653908669917503e-05,
      "loss": 0.6574,
      "mean_token_accuracy": 0.8025646843016148,
      "num_tokens": 556396556.0,
      "step": 17433
    },
    {
      "entropy": 0.6274789907038212,
      "epoch": 1.6040218285116246,
      "grad_norm": 0.15589235723018646,
      "learning_rate": 4.65360198730334e-05,
      "loss": 0.6099,
      "mean_token_accuracy": 0.8096941858530045,
      "num_tokens": 556428306.0,
      "step": 17434
    },
    {
      "entropy": 0.5404412157367915,
      "epoch": 1.6041138348830657,
      "grad_norm": 0.1478751003742218,
      "learning_rate": 4.653295304689177e-05,
      "loss": 0.5303,
      "mean_token_accuracy": 0.834822703152895,
      "num_tokens": 556460326.0,
      "step": 17435
    },
    {
      "entropy": 0.6228210553526878,
      "epoch": 1.6042058412545068,
      "grad_norm": 0.15058965981006622,
      "learning_rate": 4.652988622075015e-05,
      "loss": 0.6167,
      "mean_token_accuracy": 0.8127839714288712,
      "num_tokens": 556492759.0,
      "step": 17436
    },
    {
      "entropy": 0.794783828780055,
      "epoch": 1.604297847625948,
      "grad_norm": 0.1716737300157547,
      "learning_rate": 4.652681939460852e-05,
      "loss": 0.8048,
      "mean_token_accuracy": 0.761423759162426,
      "num_tokens": 556524508.0,
      "step": 17437
    },
    {
      "entropy": 0.5969148147851229,
      "epoch": 1.6043898539973893,
      "grad_norm": 0.156156986951828,
      "learning_rate": 4.6523752568466896e-05,
      "loss": 0.5901,
      "mean_token_accuracy": 0.819124486297369,
      "num_tokens": 556556956.0,
      "step": 17438
    },
    {
      "entropy": 0.5971282087266445,
      "epoch": 1.6044818603688307,
      "grad_norm": 0.15142226219177246,
      "learning_rate": 4.652068574232527e-05,
      "loss": 0.5901,
      "mean_token_accuracy": 0.8184718079864979,
      "num_tokens": 556588667.0,
      "step": 17439
    },
    {
      "entropy": 0.5793345812708139,
      "epoch": 1.6045738667402718,
      "grad_norm": 0.15572476387023926,
      "learning_rate": 4.651761891618364e-05,
      "loss": 0.5684,
      "mean_token_accuracy": 0.8198249414563179,
      "num_tokens": 556619972.0,
      "step": 17440
    },
    {
      "entropy": 0.5413028057664633,
      "epoch": 1.604665873111713,
      "grad_norm": 0.15477487444877625,
      "learning_rate": 4.651455209004202e-05,
      "loss": 0.5334,
      "mean_token_accuracy": 0.8383471481502056,
      "num_tokens": 556651696.0,
      "step": 17441
    },
    {
      "entropy": 0.7071652319282293,
      "epoch": 1.604757879483154,
      "grad_norm": 0.1640888899564743,
      "learning_rate": 4.651148526390039e-05,
      "loss": 0.7047,
      "mean_token_accuracy": 0.787962406873703,
      "num_tokens": 556682889.0,
      "step": 17442
    },
    {
      "entropy": 0.6674027424305677,
      "epoch": 1.6048498858545954,
      "grad_norm": 0.1584145724773407,
      "learning_rate": 4.650841843775877e-05,
      "loss": 0.6538,
      "mean_token_accuracy": 0.7966838851571083,
      "num_tokens": 556715103.0,
      "step": 17443
    },
    {
      "entropy": 0.640647891908884,
      "epoch": 1.6049418922260368,
      "grad_norm": 0.1618005782365799,
      "learning_rate": 4.650535161161714e-05,
      "loss": 0.6439,
      "mean_token_accuracy": 0.8096354007720947,
      "num_tokens": 556747437.0,
      "step": 17444
    },
    {
      "entropy": 0.6254130452871323,
      "epoch": 1.605033898597478,
      "grad_norm": 0.16394010186195374,
      "learning_rate": 4.650228478547551e-05,
      "loss": 0.6094,
      "mean_token_accuracy": 0.8171681985259056,
      "num_tokens": 556779934.0,
      "step": 17445
    },
    {
      "entropy": 0.5843552174046636,
      "epoch": 1.605125904968919,
      "grad_norm": 0.15193124115467072,
      "learning_rate": 4.649921795933389e-05,
      "loss": 0.5693,
      "mean_token_accuracy": 0.8259908817708492,
      "num_tokens": 556811580.0,
      "step": 17446
    },
    {
      "entropy": 0.5829885406419635,
      "epoch": 1.6052179113403602,
      "grad_norm": 0.15165944397449493,
      "learning_rate": 4.649615113319226e-05,
      "loss": 0.5796,
      "mean_token_accuracy": 0.8211876265704632,
      "num_tokens": 556843789.0,
      "step": 17447
    },
    {
      "entropy": 0.6426444966346025,
      "epoch": 1.6053099177118015,
      "grad_norm": 0.15892678499221802,
      "learning_rate": 4.6493084307050636e-05,
      "loss": 0.6264,
      "mean_token_accuracy": 0.804452657699585,
      "num_tokens": 556874785.0,
      "step": 17448
    },
    {
      "entropy": 0.5129110808484256,
      "epoch": 1.605401924083243,
      "grad_norm": 0.1461961567401886,
      "learning_rate": 4.649001748090901e-05,
      "loss": 0.5013,
      "mean_token_accuracy": 0.8482327535748482,
      "num_tokens": 556906913.0,
      "step": 17449
    },
    {
      "entropy": 0.5772253791801631,
      "epoch": 1.605493930454684,
      "grad_norm": 0.14994966983795166,
      "learning_rate": 4.6486950654767385e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8233412057161331,
      "num_tokens": 556938653.0,
      "step": 17450
    },
    {
      "entropy": 0.49010048247873783,
      "epoch": 1.6055859368261252,
      "grad_norm": 0.1524031162261963,
      "learning_rate": 4.648388382862576e-05,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8537222184240818,
      "num_tokens": 556970658.0,
      "step": 17451
    },
    {
      "entropy": 0.6341253221035004,
      "epoch": 1.6056779431975663,
      "grad_norm": 0.15570513904094696,
      "learning_rate": 4.6480817002484135e-05,
      "loss": 0.6264,
      "mean_token_accuracy": 0.8096006810665131,
      "num_tokens": 557003270.0,
      "step": 17452
    },
    {
      "entropy": 0.713491638423875,
      "epoch": 1.6057699495690076,
      "grad_norm": 0.15812475979328156,
      "learning_rate": 4.64777501763425e-05,
      "loss": 0.7016,
      "mean_token_accuracy": 0.791439950466156,
      "num_tokens": 557035752.0,
      "step": 17453
    },
    {
      "entropy": 0.6111651225946844,
      "epoch": 1.605861955940449,
      "grad_norm": 0.1522684246301651,
      "learning_rate": 4.6474683350200884e-05,
      "loss": 0.5878,
      "mean_token_accuracy": 0.8188711628317833,
      "num_tokens": 557067110.0,
      "step": 17454
    },
    {
      "entropy": 0.6213798928074539,
      "epoch": 1.6059539623118901,
      "grad_norm": 0.16955551505088806,
      "learning_rate": 4.647161652405925e-05,
      "loss": 0.6119,
      "mean_token_accuracy": 0.8159720115363598,
      "num_tokens": 557098807.0,
      "step": 17455
    },
    {
      "entropy": 0.6594769228249788,
      "epoch": 1.6060459686833313,
      "grad_norm": 0.1582072526216507,
      "learning_rate": 4.6468549697917626e-05,
      "loss": 0.6481,
      "mean_token_accuracy": 0.8025893457233906,
      "num_tokens": 557130450.0,
      "step": 17456
    },
    {
      "entropy": 0.6449019815772772,
      "epoch": 1.6061379750547724,
      "grad_norm": 0.1489245593547821,
      "learning_rate": 4.6465482871776e-05,
      "loss": 0.6317,
      "mean_token_accuracy": 0.8076993264257908,
      "num_tokens": 557162969.0,
      "step": 17457
    },
    {
      "entropy": 0.6178886978887022,
      "epoch": 1.6062299814262138,
      "grad_norm": 0.15414166450500488,
      "learning_rate": 4.6462416045634376e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8163226507604122,
      "num_tokens": 557195401.0,
      "step": 17458
    },
    {
      "entropy": 0.6194975851103663,
      "epoch": 1.6063219877976551,
      "grad_norm": 0.1522343009710312,
      "learning_rate": 4.645934921949275e-05,
      "loss": 0.6007,
      "mean_token_accuracy": 0.8134279623627663,
      "num_tokens": 557226972.0,
      "step": 17459
    },
    {
      "entropy": 0.4852743800729513,
      "epoch": 1.6064139941690962,
      "grad_norm": 0.14357870817184448,
      "learning_rate": 4.6456282393351125e-05,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8504672273993492,
      "num_tokens": 557258702.0,
      "step": 17460
    },
    {
      "entropy": 0.6051835818216205,
      "epoch": 1.6065060005405374,
      "grad_norm": 0.1505623459815979,
      "learning_rate": 4.645321556720949e-05,
      "loss": 0.5896,
      "mean_token_accuracy": 0.8188775107264519,
      "num_tokens": 557291202.0,
      "step": 17461
    },
    {
      "entropy": 0.594799080863595,
      "epoch": 1.6065980069119785,
      "grad_norm": 0.16199535131454468,
      "learning_rate": 4.6450148741067874e-05,
      "loss": 0.5766,
      "mean_token_accuracy": 0.8216538652777672,
      "num_tokens": 557323402.0,
      "step": 17462
    },
    {
      "entropy": 0.5275127431377769,
      "epoch": 1.6066900132834199,
      "grad_norm": 0.145639106631279,
      "learning_rate": 4.644708191492624e-05,
      "loss": 0.5135,
      "mean_token_accuracy": 0.8415839038789272,
      "num_tokens": 557354715.0,
      "step": 17463
    },
    {
      "entropy": 0.6785784550011158,
      "epoch": 1.6067820196548612,
      "grad_norm": 0.15886768698692322,
      "learning_rate": 4.6444015088784624e-05,
      "loss": 0.6675,
      "mean_token_accuracy": 0.8000557720661163,
      "num_tokens": 557387379.0,
      "step": 17464
    },
    {
      "entropy": 0.5964267458766699,
      "epoch": 1.6068740260263024,
      "grad_norm": 0.14834807813167572,
      "learning_rate": 4.644094826264299e-05,
      "loss": 0.5833,
      "mean_token_accuracy": 0.8205602653324604,
      "num_tokens": 557419108.0,
      "step": 17465
    },
    {
      "entropy": 0.5100677702575922,
      "epoch": 1.6069660323977435,
      "grad_norm": 0.15710999071598053,
      "learning_rate": 4.6437881436501366e-05,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8463062942028046,
      "num_tokens": 557450018.0,
      "step": 17466
    },
    {
      "entropy": 0.5690752109512687,
      "epoch": 1.6070580387691846,
      "grad_norm": 0.14697393774986267,
      "learning_rate": 4.643481461035974e-05,
      "loss": 0.556,
      "mean_token_accuracy": 0.8303008563816547,
      "num_tokens": 557482204.0,
      "step": 17467
    },
    {
      "entropy": 0.544894224498421,
      "epoch": 1.607150045140626,
      "grad_norm": 0.14759261906147003,
      "learning_rate": 4.6431747784218115e-05,
      "loss": 0.5393,
      "mean_token_accuracy": 0.8361821658909321,
      "num_tokens": 557514722.0,
      "step": 17468
    },
    {
      "entropy": 0.5918578300625086,
      "epoch": 1.6072420515120673,
      "grad_norm": 0.15623390674591064,
      "learning_rate": 4.642868095807649e-05,
      "loss": 0.5709,
      "mean_token_accuracy": 0.8220916427671909,
      "num_tokens": 557546382.0,
      "step": 17469
    },
    {
      "entropy": 0.5870149056427181,
      "epoch": 1.6073340578835085,
      "grad_norm": 0.1534956395626068,
      "learning_rate": 4.6425614131934865e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.828751876950264,
      "num_tokens": 557578705.0,
      "step": 17470
    },
    {
      "entropy": 0.6323249647393823,
      "epoch": 1.6074260642549496,
      "grad_norm": 0.16029039025306702,
      "learning_rate": 4.642254730579323e-05,
      "loss": 0.6231,
      "mean_token_accuracy": 0.8105973787605762,
      "num_tokens": 557610963.0,
      "step": 17471
    },
    {
      "entropy": 0.6484651435166597,
      "epoch": 1.6075180706263907,
      "grad_norm": 0.15953095257282257,
      "learning_rate": 4.6419480479651614e-05,
      "loss": 0.6371,
      "mean_token_accuracy": 0.8070896156132221,
      "num_tokens": 557642855.0,
      "step": 17472
    },
    {
      "entropy": 0.5650432172697037,
      "epoch": 1.607610076997832,
      "grad_norm": 0.15001989901065826,
      "learning_rate": 4.641641365350998e-05,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8315557613968849,
      "num_tokens": 557675464.0,
      "step": 17473
    },
    {
      "entropy": 0.5806641345843673,
      "epoch": 1.6077020833692734,
      "grad_norm": 0.15465860068798065,
      "learning_rate": 4.6413346827368357e-05,
      "loss": 0.5817,
      "mean_token_accuracy": 0.8198429979383945,
      "num_tokens": 557707627.0,
      "step": 17474
    },
    {
      "entropy": 0.6561975292861462,
      "epoch": 1.6077940897407146,
      "grad_norm": 0.15805912017822266,
      "learning_rate": 4.641028000122673e-05,
      "loss": 0.6467,
      "mean_token_accuracy": 0.8038895055651665,
      "num_tokens": 557739229.0,
      "step": 17475
    },
    {
      "entropy": 0.6511763129383326,
      "epoch": 1.6078860961121557,
      "grad_norm": 0.16354900598526,
      "learning_rate": 4.6407213175085106e-05,
      "loss": 0.6515,
      "mean_token_accuracy": 0.8043708875775337,
      "num_tokens": 557771416.0,
      "step": 17476
    },
    {
      "entropy": 0.5430748742073774,
      "epoch": 1.6079781024835968,
      "grad_norm": 0.15606656670570374,
      "learning_rate": 4.640414634894348e-05,
      "loss": 0.533,
      "mean_token_accuracy": 0.8323608376085758,
      "num_tokens": 557802754.0,
      "step": 17477
    },
    {
      "entropy": 0.6347108390182257,
      "epoch": 1.6080701088550382,
      "grad_norm": 0.16149269044399261,
      "learning_rate": 4.6401079522801855e-05,
      "loss": 0.6279,
      "mean_token_accuracy": 0.805280726402998,
      "num_tokens": 557834857.0,
      "step": 17478
    },
    {
      "entropy": 0.6641124337911606,
      "epoch": 1.6081621152264796,
      "grad_norm": 0.15981656312942505,
      "learning_rate": 4.639801269666023e-05,
      "loss": 0.665,
      "mean_token_accuracy": 0.7923991121351719,
      "num_tokens": 557867300.0,
      "step": 17479
    },
    {
      "entropy": 0.570964653044939,
      "epoch": 1.6082541215979207,
      "grad_norm": 0.16054502129554749,
      "learning_rate": 4.6394945870518604e-05,
      "loss": 0.5596,
      "mean_token_accuracy": 0.8284892551600933,
      "num_tokens": 557899852.0,
      "step": 17480
    },
    {
      "entropy": 0.5716898217797279,
      "epoch": 1.6083461279693618,
      "grad_norm": 0.15082310140132904,
      "learning_rate": 4.639187904437698e-05,
      "loss": 0.5784,
      "mean_token_accuracy": 0.823974747210741,
      "num_tokens": 557931320.0,
      "step": 17481
    },
    {
      "entropy": 0.6574204191565514,
      "epoch": 1.608438134340803,
      "grad_norm": 0.16181470453739166,
      "learning_rate": 4.638881221823535e-05,
      "loss": 0.6462,
      "mean_token_accuracy": 0.8020723685622215,
      "num_tokens": 557963432.0,
      "step": 17482
    },
    {
      "entropy": 0.6329269777052104,
      "epoch": 1.6085301407122443,
      "grad_norm": 0.16010206937789917,
      "learning_rate": 4.638574539209373e-05,
      "loss": 0.6102,
      "mean_token_accuracy": 0.8127921484410763,
      "num_tokens": 557995333.0,
      "step": 17483
    },
    {
      "entropy": 0.6005680896341801,
      "epoch": 1.6086221470836857,
      "grad_norm": 0.15104639530181885,
      "learning_rate": 4.6382678565952096e-05,
      "loss": 0.5885,
      "mean_token_accuracy": 0.8195523098111153,
      "num_tokens": 558027822.0,
      "step": 17484
    },
    {
      "entropy": 0.6869800835847855,
      "epoch": 1.6087141534551268,
      "grad_norm": 0.16249732673168182,
      "learning_rate": 4.637961173981048e-05,
      "loss": 0.6765,
      "mean_token_accuracy": 0.797555647790432,
      "num_tokens": 558059801.0,
      "step": 17485
    },
    {
      "entropy": 0.6552161718718708,
      "epoch": 1.608806159826568,
      "grad_norm": 0.16029763221740723,
      "learning_rate": 4.6376544913668846e-05,
      "loss": 0.6307,
      "mean_token_accuracy": 0.8060556575655937,
      "num_tokens": 558091076.0,
      "step": 17486
    },
    {
      "entropy": 0.5967641426250339,
      "epoch": 1.608898166198009,
      "grad_norm": 0.14960557222366333,
      "learning_rate": 4.637347808752722e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8201534636318684,
      "num_tokens": 558123439.0,
      "step": 17487
    },
    {
      "entropy": 0.64583929348737,
      "epoch": 1.6089901725694504,
      "grad_norm": 0.15884147584438324,
      "learning_rate": 4.6370411261385595e-05,
      "loss": 0.6449,
      "mean_token_accuracy": 0.8003845550119877,
      "num_tokens": 558155749.0,
      "step": 17488
    },
    {
      "entropy": 0.6280965600162745,
      "epoch": 1.6090821789408918,
      "grad_norm": 0.15392577648162842,
      "learning_rate": 4.636734443524397e-05,
      "loss": 0.6241,
      "mean_token_accuracy": 0.8081072941422462,
      "num_tokens": 558188123.0,
      "step": 17489
    },
    {
      "entropy": 0.5786479241214693,
      "epoch": 1.609174185312333,
      "grad_norm": 0.1558489203453064,
      "learning_rate": 4.6364277609102344e-05,
      "loss": 0.5528,
      "mean_token_accuracy": 0.8262701481580734,
      "num_tokens": 558220891.0,
      "step": 17490
    },
    {
      "entropy": 0.6588387470692396,
      "epoch": 1.609266191683774,
      "grad_norm": 0.15778835117816925,
      "learning_rate": 4.636121078296072e-05,
      "loss": 0.6416,
      "mean_token_accuracy": 0.8016809858381748,
      "num_tokens": 558252311.0,
      "step": 17491
    },
    {
      "entropy": 0.6599852899089456,
      "epoch": 1.6093581980552152,
      "grad_norm": 0.1567077338695526,
      "learning_rate": 4.635814395681909e-05,
      "loss": 0.65,
      "mean_token_accuracy": 0.7979404702782631,
      "num_tokens": 558284525.0,
      "step": 17492
    },
    {
      "entropy": 0.627837423235178,
      "epoch": 1.6094502044266565,
      "grad_norm": 0.1553865522146225,
      "learning_rate": 4.635507713067747e-05,
      "loss": 0.6285,
      "mean_token_accuracy": 0.8082710951566696,
      "num_tokens": 558317193.0,
      "step": 17493
    },
    {
      "entropy": 0.5506362249143422,
      "epoch": 1.609542210798098,
      "grad_norm": 0.1479157656431198,
      "learning_rate": 4.6352010304535836e-05,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8364238031208515,
      "num_tokens": 558349754.0,
      "step": 17494
    },
    {
      "entropy": 0.54957598564215,
      "epoch": 1.609634217169539,
      "grad_norm": 0.15316104888916016,
      "learning_rate": 4.634894347839421e-05,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8345811888575554,
      "num_tokens": 558381893.0,
      "step": 17495
    },
    {
      "entropy": 0.5910650547593832,
      "epoch": 1.6097262235409802,
      "grad_norm": 0.15560100972652435,
      "learning_rate": 4.6345876652252585e-05,
      "loss": 0.5786,
      "mean_token_accuracy": 0.8236086666584015,
      "num_tokens": 558414523.0,
      "step": 17496
    },
    {
      "entropy": 0.7003950793296099,
      "epoch": 1.6098182299124213,
      "grad_norm": 0.16313432157039642,
      "learning_rate": 4.634280982611096e-05,
      "loss": 0.6938,
      "mean_token_accuracy": 0.7890109829604626,
      "num_tokens": 558446653.0,
      "step": 17497
    },
    {
      "entropy": 0.571780126541853,
      "epoch": 1.6099102362838627,
      "grad_norm": 0.15879391133785248,
      "learning_rate": 4.6339742999969335e-05,
      "loss": 0.5665,
      "mean_token_accuracy": 0.8275970481336117,
      "num_tokens": 558478720.0,
      "step": 17498
    },
    {
      "entropy": 0.5606716237962246,
      "epoch": 1.610002242655304,
      "grad_norm": 0.14485715329647064,
      "learning_rate": 4.633667617382771e-05,
      "loss": 0.55,
      "mean_token_accuracy": 0.8315306752920151,
      "num_tokens": 558511256.0,
      "step": 17499
    },
    {
      "entropy": 0.6848841430619359,
      "epoch": 1.6100942490267451,
      "grad_norm": 0.16378365457057953,
      "learning_rate": 4.633360934768608e-05,
      "loss": 0.665,
      "mean_token_accuracy": 0.7970220297574997,
      "num_tokens": 558542717.0,
      "step": 17500
    },
    {
      "entropy": 0.5679123289883137,
      "epoch": 1.6101862553981863,
      "grad_norm": 0.15699468553066254,
      "learning_rate": 4.633054252154446e-05,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8280131369829178,
      "num_tokens": 558574663.0,
      "step": 17501
    },
    {
      "entropy": 0.5965096019208431,
      "epoch": 1.6102782617696274,
      "grad_norm": 0.15032950043678284,
      "learning_rate": 4.6327475695402826e-05,
      "loss": 0.5834,
      "mean_token_accuracy": 0.8223349899053574,
      "num_tokens": 558606389.0,
      "step": 17502
    },
    {
      "entropy": 0.6647724485956132,
      "epoch": 1.6103702681410688,
      "grad_norm": 0.16016049683094025,
      "learning_rate": 4.632440886926121e-05,
      "loss": 0.6556,
      "mean_token_accuracy": 0.802052766084671,
      "num_tokens": 558637709.0,
      "step": 17503
    },
    {
      "entropy": 0.5832825712859631,
      "epoch": 1.6104622745125101,
      "grad_norm": 0.15975067019462585,
      "learning_rate": 4.6321342043119576e-05,
      "loss": 0.5789,
      "mean_token_accuracy": 0.8264950290322304,
      "num_tokens": 558669702.0,
      "step": 17504
    },
    {
      "entropy": 0.6406064331531525,
      "epoch": 1.6105542808839513,
      "grad_norm": 0.1603943109512329,
      "learning_rate": 4.631827521697795e-05,
      "loss": 0.6402,
      "mean_token_accuracy": 0.8029492869973183,
      "num_tokens": 558701497.0,
      "step": 17505
    },
    {
      "entropy": 0.5204739505425096,
      "epoch": 1.6106462872553924,
      "grad_norm": 0.14946255087852478,
      "learning_rate": 4.6315208390836325e-05,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8453235700726509,
      "num_tokens": 558733770.0,
      "step": 17506
    },
    {
      "entropy": 0.6744629293680191,
      "epoch": 1.6107382936268335,
      "grad_norm": 0.15603071451187134,
      "learning_rate": 4.63121415646947e-05,
      "loss": 0.6739,
      "mean_token_accuracy": 0.7966865748167038,
      "num_tokens": 558766358.0,
      "step": 17507
    },
    {
      "entropy": 0.6333722285926342,
      "epoch": 1.6108302999982749,
      "grad_norm": 0.15192203223705292,
      "learning_rate": 4.6309074738553074e-05,
      "loss": 0.6272,
      "mean_token_accuracy": 0.8078503459692001,
      "num_tokens": 558798415.0,
      "step": 17508
    },
    {
      "entropy": 0.6967080589383841,
      "epoch": 1.6109223063697162,
      "grad_norm": 0.16515624523162842,
      "learning_rate": 4.630600791241145e-05,
      "loss": 0.6948,
      "mean_token_accuracy": 0.7908778339624405,
      "num_tokens": 558830294.0,
      "step": 17509
    },
    {
      "entropy": 0.5663769692182541,
      "epoch": 1.6110143127411574,
      "grad_norm": 0.1505759358406067,
      "learning_rate": 4.6302941086269823e-05,
      "loss": 0.5594,
      "mean_token_accuracy": 0.8287228792905807,
      "num_tokens": 558862458.0,
      "step": 17510
    },
    {
      "entropy": 0.6406414052471519,
      "epoch": 1.6111063191125985,
      "grad_norm": 0.15868613123893738,
      "learning_rate": 4.62998742601282e-05,
      "loss": 0.6253,
      "mean_token_accuracy": 0.8061326742172241,
      "num_tokens": 558893970.0,
      "step": 17511
    },
    {
      "entropy": 0.6959962649270892,
      "epoch": 1.6111983254840396,
      "grad_norm": 0.16180814802646637,
      "learning_rate": 4.629680743398657e-05,
      "loss": 0.6741,
      "mean_token_accuracy": 0.792423639446497,
      "num_tokens": 558925553.0,
      "step": 17512
    },
    {
      "entropy": 0.602536041289568,
      "epoch": 1.611290331855481,
      "grad_norm": 0.16217245161533356,
      "learning_rate": 4.629374060784494e-05,
      "loss": 0.5798,
      "mean_token_accuracy": 0.8197517581284046,
      "num_tokens": 558957351.0,
      "step": 17513
    },
    {
      "entropy": 0.56290627643466,
      "epoch": 1.6113823382269223,
      "grad_norm": 0.14968125522136688,
      "learning_rate": 4.629067378170332e-05,
      "loss": 0.5453,
      "mean_token_accuracy": 0.8277059756219387,
      "num_tokens": 558989498.0,
      "step": 17514
    },
    {
      "entropy": 0.6476096296682954,
      "epoch": 1.6114743445983635,
      "grad_norm": 0.1561359465122223,
      "learning_rate": 4.628760695556169e-05,
      "loss": 0.6229,
      "mean_token_accuracy": 0.8094584159553051,
      "num_tokens": 559021965.0,
      "step": 17515
    },
    {
      "entropy": 0.5732286870479584,
      "epoch": 1.6115663509698046,
      "grad_norm": 0.15795989334583282,
      "learning_rate": 4.6284540129420065e-05,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8275236450135708,
      "num_tokens": 559053334.0,
      "step": 17516
    },
    {
      "entropy": 0.6330113164149225,
      "epoch": 1.6116583573412457,
      "grad_norm": 0.15311695635318756,
      "learning_rate": 4.628147330327844e-05,
      "loss": 0.6158,
      "mean_token_accuracy": 0.814426563680172,
      "num_tokens": 559085888.0,
      "step": 17517
    },
    {
      "entropy": 0.6371185760945082,
      "epoch": 1.611750363712687,
      "grad_norm": 0.15635225176811218,
      "learning_rate": 4.6278406477136814e-05,
      "loss": 0.6252,
      "mean_token_accuracy": 0.8065694458782673,
      "num_tokens": 559117766.0,
      "step": 17518
    },
    {
      "entropy": 0.642926805652678,
      "epoch": 1.6118423700841285,
      "grad_norm": 0.15734875202178955,
      "learning_rate": 4.627533965099519e-05,
      "loss": 0.6292,
      "mean_token_accuracy": 0.8071915991604328,
      "num_tokens": 559149944.0,
      "step": 17519
    },
    {
      "entropy": 0.5854745237156749,
      "epoch": 1.6119343764555696,
      "grad_norm": 0.15573275089263916,
      "learning_rate": 4.627227282485356e-05,
      "loss": 0.5753,
      "mean_token_accuracy": 0.8240702636539936,
      "num_tokens": 559182401.0,
      "step": 17520
    },
    {
      "entropy": 0.5835944339632988,
      "epoch": 1.6120263828270107,
      "grad_norm": 0.15242759883403778,
      "learning_rate": 4.626920599871193e-05,
      "loss": 0.5743,
      "mean_token_accuracy": 0.8243948146700859,
      "num_tokens": 559214809.0,
      "step": 17521
    },
    {
      "entropy": 0.7852814784273505,
      "epoch": 1.6121183891984519,
      "grad_norm": 0.17484928667545319,
      "learning_rate": 4.626613917257031e-05,
      "loss": 0.7844,
      "mean_token_accuracy": 0.764711894094944,
      "num_tokens": 559247139.0,
      "step": 17522
    },
    {
      "entropy": 0.6511913239955902,
      "epoch": 1.6122103955698932,
      "grad_norm": 0.15851172804832458,
      "learning_rate": 4.626307234642868e-05,
      "loss": 0.6461,
      "mean_token_accuracy": 0.8029165416955948,
      "num_tokens": 559279428.0,
      "step": 17523
    },
    {
      "entropy": 0.6687136236578226,
      "epoch": 1.6123024019413346,
      "grad_norm": 0.1636350452899933,
      "learning_rate": 4.626000552028706e-05,
      "loss": 0.6773,
      "mean_token_accuracy": 0.7959966324269772,
      "num_tokens": 559311246.0,
      "step": 17524
    },
    {
      "entropy": 0.6399038732051849,
      "epoch": 1.6123944083127757,
      "grad_norm": 0.15441475808620453,
      "learning_rate": 4.625693869414543e-05,
      "loss": 0.6282,
      "mean_token_accuracy": 0.8047981895506382,
      "num_tokens": 559343264.0,
      "step": 17525
    },
    {
      "entropy": 0.49871095921844244,
      "epoch": 1.6124864146842168,
      "grad_norm": 0.1494656801223755,
      "learning_rate": 4.6253871868003804e-05,
      "loss": 0.4912,
      "mean_token_accuracy": 0.8469885289669037,
      "num_tokens": 559375220.0,
      "step": 17526
    },
    {
      "entropy": 0.5865662824362516,
      "epoch": 1.612578421055658,
      "grad_norm": 0.1505713164806366,
      "learning_rate": 4.625080504186218e-05,
      "loss": 0.5854,
      "mean_token_accuracy": 0.8202890902757645,
      "num_tokens": 559406819.0,
      "step": 17527
    },
    {
      "entropy": 0.6230405988171697,
      "epoch": 1.6126704274270993,
      "grad_norm": 0.15504907071590424,
      "learning_rate": 4.6247738215720554e-05,
      "loss": 0.6161,
      "mean_token_accuracy": 0.8137092143297195,
      "num_tokens": 559438570.0,
      "step": 17528
    },
    {
      "entropy": 0.5751097360625863,
      "epoch": 1.6127624337985407,
      "grad_norm": 0.15065233409404755,
      "learning_rate": 4.624467138957893e-05,
      "loss": 0.5604,
      "mean_token_accuracy": 0.829809233546257,
      "num_tokens": 559470047.0,
      "step": 17529
    },
    {
      "entropy": 0.6681636441498995,
      "epoch": 1.6128544401699818,
      "grad_norm": 0.15925830602645874,
      "learning_rate": 4.62416045634373e-05,
      "loss": 0.646,
      "mean_token_accuracy": 0.8018548339605331,
      "num_tokens": 559500627.0,
      "step": 17530
    },
    {
      "entropy": 0.6163792461156845,
      "epoch": 1.612946446541423,
      "grad_norm": 0.14873066544532776,
      "learning_rate": 4.623853773729567e-05,
      "loss": 0.5965,
      "mean_token_accuracy": 0.8167514726519585,
      "num_tokens": 559532476.0,
      "step": 17531
    },
    {
      "entropy": 0.5722589248325676,
      "epoch": 1.613038452912864,
      "grad_norm": 0.14424502849578857,
      "learning_rate": 4.623547091115405e-05,
      "loss": 0.5548,
      "mean_token_accuracy": 0.8320575319230556,
      "num_tokens": 559565037.0,
      "step": 17532
    },
    {
      "entropy": 0.5407416615635157,
      "epoch": 1.6131304592843054,
      "grad_norm": 0.15339703857898712,
      "learning_rate": 4.623240408501242e-05,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8348691761493683,
      "num_tokens": 559596718.0,
      "step": 17533
    },
    {
      "entropy": 0.6590034235268831,
      "epoch": 1.6132224656557468,
      "grad_norm": 0.1638828068971634,
      "learning_rate": 4.6229337258870795e-05,
      "loss": 0.6667,
      "mean_token_accuracy": 0.7961879447102547,
      "num_tokens": 559628251.0,
      "step": 17534
    },
    {
      "entropy": 0.5395421581342816,
      "epoch": 1.613314472027188,
      "grad_norm": 0.14690423011779785,
      "learning_rate": 4.622627043272917e-05,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8437882922589779,
      "num_tokens": 559660347.0,
      "step": 17535
    },
    {
      "entropy": 0.5165337547659874,
      "epoch": 1.613406478398629,
      "grad_norm": 0.15198113024234772,
      "learning_rate": 4.6223203606587544e-05,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8411539196968079,
      "num_tokens": 559692398.0,
      "step": 17536
    },
    {
      "entropy": 0.5724102621898055,
      "epoch": 1.6134984847700702,
      "grad_norm": 0.1555643528699875,
      "learning_rate": 4.622013678044592e-05,
      "loss": 0.567,
      "mean_token_accuracy": 0.8323008380830288,
      "num_tokens": 559723563.0,
      "step": 17537
    },
    {
      "entropy": 0.5906764818355441,
      "epoch": 1.6135904911415115,
      "grad_norm": 0.1553087830543518,
      "learning_rate": 4.621706995430429e-05,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8197499439120293,
      "num_tokens": 559755979.0,
      "step": 17538
    },
    {
      "entropy": 0.7005741950124502,
      "epoch": 1.613682497512953,
      "grad_norm": 0.16842523217201233,
      "learning_rate": 4.621400312816266e-05,
      "loss": 0.6995,
      "mean_token_accuracy": 0.7883813716471195,
      "num_tokens": 559787426.0,
      "step": 17539
    },
    {
      "entropy": 0.5791436564177275,
      "epoch": 1.613774503884394,
      "grad_norm": 0.1532442718744278,
      "learning_rate": 4.621093630202104e-05,
      "loss": 0.5627,
      "mean_token_accuracy": 0.8273614645004272,
      "num_tokens": 559819535.0,
      "step": 17540
    },
    {
      "entropy": 0.5937895425595343,
      "epoch": 1.6138665102558352,
      "grad_norm": 0.1536843180656433,
      "learning_rate": 4.620786947587942e-05,
      "loss": 0.5773,
      "mean_token_accuracy": 0.8232925608754158,
      "num_tokens": 559852055.0,
      "step": 17541
    },
    {
      "entropy": 0.5984996967017651,
      "epoch": 1.6139585166272763,
      "grad_norm": 0.1574965864419937,
      "learning_rate": 4.6204802649737785e-05,
      "loss": 0.5922,
      "mean_token_accuracy": 0.815970778465271,
      "num_tokens": 559882994.0,
      "step": 17542
    },
    {
      "entropy": 0.5832142056897283,
      "epoch": 1.6140505229987177,
      "grad_norm": 0.15386104583740234,
      "learning_rate": 4.6201735823596167e-05,
      "loss": 0.5708,
      "mean_token_accuracy": 0.8273943103849888,
      "num_tokens": 559915295.0,
      "step": 17543
    },
    {
      "entropy": 0.6761104678735137,
      "epoch": 1.614142529370159,
      "grad_norm": 0.16535182297229767,
      "learning_rate": 4.6198668997454534e-05,
      "loss": 0.6878,
      "mean_token_accuracy": 0.7939346432685852,
      "num_tokens": 559947045.0,
      "step": 17544
    },
    {
      "entropy": 0.5938074104487896,
      "epoch": 1.6142345357416001,
      "grad_norm": 0.1628013402223587,
      "learning_rate": 4.6195602171312916e-05,
      "loss": 0.5787,
      "mean_token_accuracy": 0.8198057785630226,
      "num_tokens": 559977687.0,
      "step": 17545
    },
    {
      "entropy": 0.6433861255645752,
      "epoch": 1.6143265421130413,
      "grad_norm": 0.16068409383296967,
      "learning_rate": 4.6192535345171284e-05,
      "loss": 0.6205,
      "mean_token_accuracy": 0.8118277490139008,
      "num_tokens": 560008987.0,
      "step": 17546
    },
    {
      "entropy": 0.6115557886660099,
      "epoch": 1.6144185484844824,
      "grad_norm": 0.155338317155838,
      "learning_rate": 4.618946851902966e-05,
      "loss": 0.5961,
      "mean_token_accuracy": 0.8145620562136173,
      "num_tokens": 560041151.0,
      "step": 17547
    },
    {
      "entropy": 0.7340655876323581,
      "epoch": 1.6145105548559238,
      "grad_norm": 0.16184554994106293,
      "learning_rate": 4.618640169288803e-05,
      "loss": 0.7171,
      "mean_token_accuracy": 0.7796087376773357,
      "num_tokens": 560072312.0,
      "step": 17548
    },
    {
      "entropy": 0.6333396341651678,
      "epoch": 1.6146025612273651,
      "grad_norm": 0.15979379415512085,
      "learning_rate": 4.618333486674641e-05,
      "loss": 0.6188,
      "mean_token_accuracy": 0.8083180114626884,
      "num_tokens": 560104434.0,
      "step": 17549
    },
    {
      "entropy": 0.6328769726678729,
      "epoch": 1.6146945675988063,
      "grad_norm": 0.1582156866788864,
      "learning_rate": 4.618026804060478e-05,
      "loss": 0.639,
      "mean_token_accuracy": 0.8062199279665947,
      "num_tokens": 560136573.0,
      "step": 17550
    },
    {
      "entropy": 0.6108391201123595,
      "epoch": 1.6147865739702474,
      "grad_norm": 0.15996290743350983,
      "learning_rate": 4.617720121446316e-05,
      "loss": 0.6061,
      "mean_token_accuracy": 0.8102548606693745,
      "num_tokens": 560168142.0,
      "step": 17551
    },
    {
      "entropy": 0.5168550200760365,
      "epoch": 1.6148785803416885,
      "grad_norm": 0.14924079179763794,
      "learning_rate": 4.6174134388321525e-05,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8403526283800602,
      "num_tokens": 560199257.0,
      "step": 17552
    },
    {
      "entropy": 0.5562458019703627,
      "epoch": 1.6149705867131299,
      "grad_norm": 0.15705187618732452,
      "learning_rate": 4.6171067562179906e-05,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8309539593756199,
      "num_tokens": 560230952.0,
      "step": 17553
    },
    {
      "entropy": 0.4623796483501792,
      "epoch": 1.6150625930845712,
      "grad_norm": 0.14573007822036743,
      "learning_rate": 4.6168000736038274e-05,
      "loss": 0.4599,
      "mean_token_accuracy": 0.8576932102441788,
      "num_tokens": 560262671.0,
      "step": 17554
    },
    {
      "entropy": 0.5588565263897181,
      "epoch": 1.6151545994560124,
      "grad_norm": 0.1428855061531067,
      "learning_rate": 4.616493390989665e-05,
      "loss": 0.5496,
      "mean_token_accuracy": 0.8286826647818089,
      "num_tokens": 560294825.0,
      "step": 17555
    },
    {
      "entropy": 0.619672411121428,
      "epoch": 1.6152466058274535,
      "grad_norm": 0.16010800004005432,
      "learning_rate": 4.6161867083755023e-05,
      "loss": 0.6094,
      "mean_token_accuracy": 0.8123138584196568,
      "num_tokens": 560327474.0,
      "step": 17556
    },
    {
      "entropy": 0.5660004392266273,
      "epoch": 1.6153386121988946,
      "grad_norm": 0.15044234693050385,
      "learning_rate": 4.61588002576134e-05,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8279813416302204,
      "num_tokens": 560359874.0,
      "step": 17557
    },
    {
      "entropy": 0.6078404518775642,
      "epoch": 1.615430618570336,
      "grad_norm": 0.15183894336223602,
      "learning_rate": 4.615573343147177e-05,
      "loss": 0.5994,
      "mean_token_accuracy": 0.8161401264369488,
      "num_tokens": 560391871.0,
      "step": 17558
    },
    {
      "entropy": 0.6131289228796959,
      "epoch": 1.6155226249417773,
      "grad_norm": 0.15372410416603088,
      "learning_rate": 4.615266660533015e-05,
      "loss": 0.6097,
      "mean_token_accuracy": 0.81446623057127,
      "num_tokens": 560424013.0,
      "step": 17559
    },
    {
      "entropy": 0.6297242064028978,
      "epoch": 1.6156146313132185,
      "grad_norm": 0.15494205057621002,
      "learning_rate": 4.6149599779188515e-05,
      "loss": 0.6316,
      "mean_token_accuracy": 0.8048682697117329,
      "num_tokens": 560456679.0,
      "step": 17560
    },
    {
      "entropy": 0.6294058440253139,
      "epoch": 1.6157066376846596,
      "grad_norm": 0.1588345170021057,
      "learning_rate": 4.61465329530469e-05,
      "loss": 0.6239,
      "mean_token_accuracy": 0.8112617135047913,
      "num_tokens": 560488620.0,
      "step": 17561
    },
    {
      "entropy": 0.5660280000884086,
      "epoch": 1.6157986440561007,
      "grad_norm": 0.14621137082576752,
      "learning_rate": 4.6143466126905265e-05,
      "loss": 0.5487,
      "mean_token_accuracy": 0.8296401500701904,
      "num_tokens": 560521171.0,
      "step": 17562
    },
    {
      "entropy": 0.516694406978786,
      "epoch": 1.615890650427542,
      "grad_norm": 0.1508360207080841,
      "learning_rate": 4.6140399300763646e-05,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8464846163988113,
      "num_tokens": 560553077.0,
      "step": 17563
    },
    {
      "entropy": 0.6467911638319492,
      "epoch": 1.6159826567989835,
      "grad_norm": 0.15985579788684845,
      "learning_rate": 4.6137332474622014e-05,
      "loss": 0.6349,
      "mean_token_accuracy": 0.8036250323057175,
      "num_tokens": 560584817.0,
      "step": 17564
    },
    {
      "entropy": 0.4996953271329403,
      "epoch": 1.6160746631704246,
      "grad_norm": 0.14402393996715546,
      "learning_rate": 4.613426564848039e-05,
      "loss": 0.4904,
      "mean_token_accuracy": 0.8475804179906845,
      "num_tokens": 560617274.0,
      "step": 17565
    },
    {
      "entropy": 0.6571931298822165,
      "epoch": 1.6161666695418657,
      "grad_norm": 0.1533932089805603,
      "learning_rate": 4.613119882233876e-05,
      "loss": 0.6474,
      "mean_token_accuracy": 0.8021744452416897,
      "num_tokens": 560649484.0,
      "step": 17566
    },
    {
      "entropy": 0.6723258420825005,
      "epoch": 1.6162586759133069,
      "grad_norm": 0.16087652742862701,
      "learning_rate": 4.612813199619714e-05,
      "loss": 0.6573,
      "mean_token_accuracy": 0.7960666008293629,
      "num_tokens": 560680604.0,
      "step": 17567
    },
    {
      "entropy": 0.616443837992847,
      "epoch": 1.6163506822847482,
      "grad_norm": 0.15916845202445984,
      "learning_rate": 4.612506517005551e-05,
      "loss": 0.6013,
      "mean_token_accuracy": 0.8145846463739872,
      "num_tokens": 560711678.0,
      "step": 17568
    },
    {
      "entropy": 0.5209729003254324,
      "epoch": 1.6164426886561896,
      "grad_norm": 0.14694073796272278,
      "learning_rate": 4.612199834391389e-05,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8431878350675106,
      "num_tokens": 560743503.0,
      "step": 17569
    },
    {
      "entropy": 0.6472979346290231,
      "epoch": 1.6165346950276307,
      "grad_norm": 0.1594913750886917,
      "learning_rate": 4.6118931517772255e-05,
      "loss": 0.6439,
      "mean_token_accuracy": 0.8047049529850483,
      "num_tokens": 560776208.0,
      "step": 17570
    },
    {
      "entropy": 0.48892561439424753,
      "epoch": 1.6166267013990718,
      "grad_norm": 0.1523415744304657,
      "learning_rate": 4.6115864691630636e-05,
      "loss": 0.4878,
      "mean_token_accuracy": 0.848516371101141,
      "num_tokens": 560808627.0,
      "step": 17571
    },
    {
      "entropy": 0.6196350324898958,
      "epoch": 1.616718707770513,
      "grad_norm": 0.1534002274274826,
      "learning_rate": 4.611279786548901e-05,
      "loss": 0.6135,
      "mean_token_accuracy": 0.8101006522774696,
      "num_tokens": 560840391.0,
      "step": 17572
    },
    {
      "entropy": 0.6543221492320299,
      "epoch": 1.6168107141419543,
      "grad_norm": 0.16506808996200562,
      "learning_rate": 4.610973103934738e-05,
      "loss": 0.6339,
      "mean_token_accuracy": 0.807551771402359,
      "num_tokens": 560871649.0,
      "step": 17573
    },
    {
      "entropy": 0.6129415049217641,
      "epoch": 1.6169027205133957,
      "grad_norm": 0.15747514367103577,
      "learning_rate": 4.610666421320576e-05,
      "loss": 0.6035,
      "mean_token_accuracy": 0.8164429739117622,
      "num_tokens": 560903205.0,
      "step": 17574
    },
    {
      "entropy": 0.549281011801213,
      "epoch": 1.6169947268848368,
      "grad_norm": 0.15251287817955017,
      "learning_rate": 4.610359738706413e-05,
      "loss": 0.5489,
      "mean_token_accuracy": 0.8277245536446571,
      "num_tokens": 560933924.0,
      "step": 17575
    },
    {
      "entropy": 0.6633269097656012,
      "epoch": 1.617086733256278,
      "grad_norm": 0.16066034138202667,
      "learning_rate": 4.61005305609225e-05,
      "loss": 0.655,
      "mean_token_accuracy": 0.7988852821290493,
      "num_tokens": 560965437.0,
      "step": 17576
    },
    {
      "entropy": 0.5526090683415532,
      "epoch": 1.617178739627719,
      "grad_norm": 0.1491652876138687,
      "learning_rate": 4.609746373478088e-05,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8376220352947712,
      "num_tokens": 560997784.0,
      "step": 17577
    },
    {
      "entropy": 0.5539158014580607,
      "epoch": 1.6172707459991604,
      "grad_norm": 0.15159624814987183,
      "learning_rate": 4.609439690863925e-05,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8342948108911514,
      "num_tokens": 561029169.0,
      "step": 17578
    },
    {
      "entropy": 0.6069529314991087,
      "epoch": 1.6173627523706018,
      "grad_norm": 0.16062481701374054,
      "learning_rate": 4.609133008249763e-05,
      "loss": 0.5896,
      "mean_token_accuracy": 0.8173455409705639,
      "num_tokens": 561060791.0,
      "step": 17579
    },
    {
      "entropy": 0.6642577284947038,
      "epoch": 1.617454758742043,
      "grad_norm": 0.1575773060321808,
      "learning_rate": 4.6088263256356e-05,
      "loss": 0.6559,
      "mean_token_accuracy": 0.7998754531145096,
      "num_tokens": 561093142.0,
      "step": 17580
    },
    {
      "entropy": 0.5812005717307329,
      "epoch": 1.617546765113484,
      "grad_norm": 0.1505051702260971,
      "learning_rate": 4.608519643021437e-05,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8282398916780949,
      "num_tokens": 561125013.0,
      "step": 17581
    },
    {
      "entropy": 0.5637873148079962,
      "epoch": 1.6176387714849252,
      "grad_norm": 0.15347619354724884,
      "learning_rate": 4.608212960407275e-05,
      "loss": 0.5528,
      "mean_token_accuracy": 0.8269647732377052,
      "num_tokens": 561157418.0,
      "step": 17582
    },
    {
      "entropy": 0.6494222842156887,
      "epoch": 1.6177307778563665,
      "grad_norm": 0.15746520459651947,
      "learning_rate": 4.607906277793112e-05,
      "loss": 0.6349,
      "mean_token_accuracy": 0.8062295950949192,
      "num_tokens": 561189051.0,
      "step": 17583
    },
    {
      "entropy": 0.6028131376951933,
      "epoch": 1.617822784227808,
      "grad_norm": 0.15292824804782867,
      "learning_rate": 4.60759959517895e-05,
      "loss": 0.5968,
      "mean_token_accuracy": 0.8147169053554535,
      "num_tokens": 561221649.0,
      "step": 17584
    },
    {
      "entropy": 0.6506537240929902,
      "epoch": 1.617914790599249,
      "grad_norm": 0.16673143208026886,
      "learning_rate": 4.607292912564787e-05,
      "loss": 0.6414,
      "mean_token_accuracy": 0.8023666515946388,
      "num_tokens": 561252672.0,
      "step": 17585
    },
    {
      "entropy": 0.6916785892099142,
      "epoch": 1.6180067969706902,
      "grad_norm": 0.16218392550945282,
      "learning_rate": 4.606986229950624e-05,
      "loss": 0.6671,
      "mean_token_accuracy": 0.7937545962631702,
      "num_tokens": 561285195.0,
      "step": 17586
    },
    {
      "entropy": 0.6736481450498104,
      "epoch": 1.6180988033421313,
      "grad_norm": 0.15767613053321838,
      "learning_rate": 4.606679547336462e-05,
      "loss": 0.6641,
      "mean_token_accuracy": 0.7994914725422859,
      "num_tokens": 561317413.0,
      "step": 17587
    },
    {
      "entropy": 0.6550032328814268,
      "epoch": 1.6181908097135727,
      "grad_norm": 0.15718142688274384,
      "learning_rate": 4.606372864722299e-05,
      "loss": 0.6478,
      "mean_token_accuracy": 0.8019910790026188,
      "num_tokens": 561349494.0,
      "step": 17588
    },
    {
      "entropy": 0.5313377492129803,
      "epoch": 1.618282816085014,
      "grad_norm": 0.14495448768138885,
      "learning_rate": 4.6060661821081366e-05,
      "loss": 0.511,
      "mean_token_accuracy": 0.8389703705906868,
      "num_tokens": 561381790.0,
      "step": 17589
    },
    {
      "entropy": 0.5037620728835464,
      "epoch": 1.6183748224564551,
      "grad_norm": 0.15554504096508026,
      "learning_rate": 4.605759499493974e-05,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8493625335395336,
      "num_tokens": 561413462.0,
      "step": 17590
    },
    {
      "entropy": 0.5507787549868226,
      "epoch": 1.6184668288278963,
      "grad_norm": 0.15222232043743134,
      "learning_rate": 4.605452816879811e-05,
      "loss": 0.5367,
      "mean_token_accuracy": 0.8365486152470112,
      "num_tokens": 561446152.0,
      "step": 17591
    },
    {
      "entropy": 0.5749750584363937,
      "epoch": 1.6185588351993374,
      "grad_norm": 0.15011101961135864,
      "learning_rate": 4.605146134265649e-05,
      "loss": 0.5503,
      "mean_token_accuracy": 0.8287843428552151,
      "num_tokens": 561477940.0,
      "step": 17592
    },
    {
      "entropy": 0.6505255699157715,
      "epoch": 1.6186508415707788,
      "grad_norm": 0.16157115995883942,
      "learning_rate": 4.604839451651486e-05,
      "loss": 0.6412,
      "mean_token_accuracy": 0.8059894219040871,
      "num_tokens": 561510468.0,
      "step": 17593
    },
    {
      "entropy": 0.6456192545592785,
      "epoch": 1.6187428479422201,
      "grad_norm": 0.154238760471344,
      "learning_rate": 4.604532769037323e-05,
      "loss": 0.6383,
      "mean_token_accuracy": 0.8023125752806664,
      "num_tokens": 561543171.0,
      "step": 17594
    },
    {
      "entropy": 0.5722297951579094,
      "epoch": 1.6188348543136613,
      "grad_norm": 0.15453898906707764,
      "learning_rate": 4.604226086423161e-05,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8229694291949272,
      "num_tokens": 561575853.0,
      "step": 17595
    },
    {
      "entropy": 0.6720445640385151,
      "epoch": 1.6189268606851024,
      "grad_norm": 0.16358286142349243,
      "learning_rate": 4.603919403808998e-05,
      "loss": 0.6697,
      "mean_token_accuracy": 0.7953943312168121,
      "num_tokens": 561607859.0,
      "step": 17596
    },
    {
      "entropy": 0.6354157142341137,
      "epoch": 1.6190188670565435,
      "grad_norm": 0.16973194479942322,
      "learning_rate": 4.603612721194836e-05,
      "loss": 0.6275,
      "mean_token_accuracy": 0.8117069564759731,
      "num_tokens": 561639658.0,
      "step": 17597
    },
    {
      "entropy": 0.5109672551043332,
      "epoch": 1.6191108734279849,
      "grad_norm": 0.1462024599313736,
      "learning_rate": 4.603306038580673e-05,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8404867351055145,
      "num_tokens": 561671387.0,
      "step": 17598
    },
    {
      "entropy": 0.6660945136100054,
      "epoch": 1.6192028797994262,
      "grad_norm": 0.15767288208007812,
      "learning_rate": 4.60299935596651e-05,
      "loss": 0.6582,
      "mean_token_accuracy": 0.8014330118894577,
      "num_tokens": 561702733.0,
      "step": 17599
    },
    {
      "entropy": 0.606036240234971,
      "epoch": 1.6192948861708674,
      "grad_norm": 0.15241001546382904,
      "learning_rate": 4.602692673352348e-05,
      "loss": 0.5971,
      "mean_token_accuracy": 0.8179849162697792,
      "num_tokens": 561735301.0,
      "step": 17600
    },
    {
      "entropy": 0.5860166517086327,
      "epoch": 1.6193868925423085,
      "grad_norm": 0.15821757912635803,
      "learning_rate": 4.6023859907381855e-05,
      "loss": 0.574,
      "mean_token_accuracy": 0.8221433982253075,
      "num_tokens": 561767260.0,
      "step": 17601
    },
    {
      "entropy": 0.6528866793960333,
      "epoch": 1.6194788989137496,
      "grad_norm": 0.16138729453086853,
      "learning_rate": 4.602079308124022e-05,
      "loss": 0.6405,
      "mean_token_accuracy": 0.8040642142295837,
      "num_tokens": 561798965.0,
      "step": 17602
    },
    {
      "entropy": 0.556471161544323,
      "epoch": 1.619570905285191,
      "grad_norm": 0.1551724672317505,
      "learning_rate": 4.6017726255098605e-05,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8338987156748772,
      "num_tokens": 561830385.0,
      "step": 17603
    },
    {
      "entropy": 0.6039825426414609,
      "epoch": 1.6196629116566323,
      "grad_norm": 0.14735235273838043,
      "learning_rate": 4.601465942895697e-05,
      "loss": 0.5845,
      "mean_token_accuracy": 0.8233994580805302,
      "num_tokens": 561862986.0,
      "step": 17604
    },
    {
      "entropy": 0.734599694609642,
      "epoch": 1.6197549180280735,
      "grad_norm": 0.15798205137252808,
      "learning_rate": 4.6011592602815354e-05,
      "loss": 0.7201,
      "mean_token_accuracy": 0.7794390842318535,
      "num_tokens": 561894650.0,
      "step": 17605
    },
    {
      "entropy": 0.6530949743464589,
      "epoch": 1.6198469243995146,
      "grad_norm": 0.1568678617477417,
      "learning_rate": 4.600852577667372e-05,
      "loss": 0.6315,
      "mean_token_accuracy": 0.8061380349099636,
      "num_tokens": 561926260.0,
      "step": 17606
    },
    {
      "entropy": 0.565856110304594,
      "epoch": 1.6199389307709557,
      "grad_norm": 0.14764532446861267,
      "learning_rate": 4.6005458950532097e-05,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8310247771441936,
      "num_tokens": 561958537.0,
      "step": 17607
    },
    {
      "entropy": 0.5814054068177938,
      "epoch": 1.620030937142397,
      "grad_norm": 0.14922721683979034,
      "learning_rate": 4.600239212439047e-05,
      "loss": 0.5799,
      "mean_token_accuracy": 0.8215662576258183,
      "num_tokens": 561991164.0,
      "step": 17608
    },
    {
      "entropy": 0.6080864961259067,
      "epoch": 1.6201229435138385,
      "grad_norm": 0.15481363236904144,
      "learning_rate": 4.5999325298248846e-05,
      "loss": 0.5949,
      "mean_token_accuracy": 0.8193825148046017,
      "num_tokens": 562023539.0,
      "step": 17609
    },
    {
      "entropy": 0.636718014255166,
      "epoch": 1.6202149498852796,
      "grad_norm": 0.1534159779548645,
      "learning_rate": 4.599625847210722e-05,
      "loss": 0.624,
      "mean_token_accuracy": 0.8051497526466846,
      "num_tokens": 562055842.0,
      "step": 17610
    },
    {
      "entropy": 0.48408601246774197,
      "epoch": 1.6203069562567207,
      "grad_norm": 0.147276371717453,
      "learning_rate": 4.5993191645965595e-05,
      "loss": 0.4771,
      "mean_token_accuracy": 0.8511279448866844,
      "num_tokens": 562088248.0,
      "step": 17611
    },
    {
      "entropy": 0.5890598781406879,
      "epoch": 1.6203989626281619,
      "grad_norm": 0.1577337384223938,
      "learning_rate": 4.599012481982396e-05,
      "loss": 0.5908,
      "mean_token_accuracy": 0.8204598911106586,
      "num_tokens": 562120262.0,
      "step": 17612
    },
    {
      "entropy": 0.5190798277035356,
      "epoch": 1.6204909689996032,
      "grad_norm": 0.14648346602916718,
      "learning_rate": 4.5987057993682344e-05,
      "loss": 0.5021,
      "mean_token_accuracy": 0.8443814069032669,
      "num_tokens": 562152309.0,
      "step": 17613
    },
    {
      "entropy": 0.7214558459818363,
      "epoch": 1.6205829753710446,
      "grad_norm": 0.16092762351036072,
      "learning_rate": 4.598399116754071e-05,
      "loss": 0.6984,
      "mean_token_accuracy": 0.7895839102566242,
      "num_tokens": 562183729.0,
      "step": 17614
    },
    {
      "entropy": 0.6119940355420113,
      "epoch": 1.6206749817424857,
      "grad_norm": 0.1560182124376297,
      "learning_rate": 4.598092434139909e-05,
      "loss": 0.6063,
      "mean_token_accuracy": 0.8138740658760071,
      "num_tokens": 562215368.0,
      "step": 17615
    },
    {
      "entropy": 0.6686999714002013,
      "epoch": 1.6207669881139268,
      "grad_norm": 0.15862387418746948,
      "learning_rate": 4.597785751525746e-05,
      "loss": 0.6645,
      "mean_token_accuracy": 0.7990272007882595,
      "num_tokens": 562247355.0,
      "step": 17616
    },
    {
      "entropy": 0.6806930070742965,
      "epoch": 1.620858994485368,
      "grad_norm": 0.1579895317554474,
      "learning_rate": 4.5974790689115836e-05,
      "loss": 0.6755,
      "mean_token_accuracy": 0.7966034449636936,
      "num_tokens": 562279318.0,
      "step": 17617
    },
    {
      "entropy": 0.6203882489353418,
      "epoch": 1.6209510008568093,
      "grad_norm": 0.15378892421722412,
      "learning_rate": 4.597172386297421e-05,
      "loss": 0.6048,
      "mean_token_accuracy": 0.8163637071847916,
      "num_tokens": 562311782.0,
      "step": 17618
    },
    {
      "entropy": 0.5995144944172353,
      "epoch": 1.6210430072282507,
      "grad_norm": 0.14346514642238617,
      "learning_rate": 4.5968657036832586e-05,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8177515864372253,
      "num_tokens": 562344550.0,
      "step": 17619
    },
    {
      "entropy": 0.4216906900983304,
      "epoch": 1.6211350135996918,
      "grad_norm": 0.1465955227613449,
      "learning_rate": 4.5965590210690953e-05,
      "loss": 0.3955,
      "mean_token_accuracy": 0.8781150951981544,
      "num_tokens": 562376763.0,
      "step": 17620
    },
    {
      "entropy": 0.5718246870674193,
      "epoch": 1.621227019971133,
      "grad_norm": 0.14753830432891846,
      "learning_rate": 4.5962523384549335e-05,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8238744921982288,
      "num_tokens": 562409510.0,
      "step": 17621
    },
    {
      "entropy": 0.6356112314388156,
      "epoch": 1.621319026342574,
      "grad_norm": 0.15597529709339142,
      "learning_rate": 4.59594565584077e-05,
      "loss": 0.6321,
      "mean_token_accuracy": 0.8092902824282646,
      "num_tokens": 562441741.0,
      "step": 17622
    },
    {
      "entropy": 0.6693520341068506,
      "epoch": 1.6214110327140154,
      "grad_norm": 0.16611190140247345,
      "learning_rate": 4.5956389732266084e-05,
      "loss": 0.6539,
      "mean_token_accuracy": 0.8031826429069042,
      "num_tokens": 562472949.0,
      "step": 17623
    },
    {
      "entropy": 0.6738807894289494,
      "epoch": 1.6215030390854568,
      "grad_norm": 0.16146741807460785,
      "learning_rate": 4.595332290612445e-05,
      "loss": 0.6677,
      "mean_token_accuracy": 0.8007539846003056,
      "num_tokens": 562505152.0,
      "step": 17624
    },
    {
      "entropy": 0.7371621187776327,
      "epoch": 1.621595045456898,
      "grad_norm": 0.16176508367061615,
      "learning_rate": 4.595025607998283e-05,
      "loss": 0.7374,
      "mean_token_accuracy": 0.7783225066959858,
      "num_tokens": 562537144.0,
      "step": 17625
    },
    {
      "entropy": 0.5657526049762964,
      "epoch": 1.621687051828339,
      "grad_norm": 0.14107589423656464,
      "learning_rate": 4.59471892538412e-05,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8278683163225651,
      "num_tokens": 562569683.0,
      "step": 17626
    },
    {
      "entropy": 0.6343997558578849,
      "epoch": 1.6217790581997802,
      "grad_norm": 0.15867847204208374,
      "learning_rate": 4.5944122427699576e-05,
      "loss": 0.6222,
      "mean_token_accuracy": 0.8073475621640682,
      "num_tokens": 562600490.0,
      "step": 17627
    },
    {
      "entropy": 0.5486204959452152,
      "epoch": 1.6218710645712215,
      "grad_norm": 0.14858098328113556,
      "learning_rate": 4.594105560155795e-05,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8274767510592937,
      "num_tokens": 562632432.0,
      "step": 17628
    },
    {
      "entropy": 0.5834578052163124,
      "epoch": 1.621963070942663,
      "grad_norm": 0.15793238580226898,
      "learning_rate": 4.5937988775416325e-05,
      "loss": 0.5757,
      "mean_token_accuracy": 0.8229964040219784,
      "num_tokens": 562664306.0,
      "step": 17629
    },
    {
      "entropy": 0.6678530313074589,
      "epoch": 1.622055077314104,
      "grad_norm": 0.1588418185710907,
      "learning_rate": 4.593492194927469e-05,
      "loss": 0.6711,
      "mean_token_accuracy": 0.7991420105099678,
      "num_tokens": 562696870.0,
      "step": 17630
    },
    {
      "entropy": 0.5696177398785949,
      "epoch": 1.6221470836855452,
      "grad_norm": 0.14766645431518555,
      "learning_rate": 4.5931855123133075e-05,
      "loss": 0.5574,
      "mean_token_accuracy": 0.8256295435130596,
      "num_tokens": 562729116.0,
      "step": 17631
    },
    {
      "entropy": 0.6826026327908039,
      "epoch": 1.6222390900569863,
      "grad_norm": 0.15987809002399445,
      "learning_rate": 4.592878829699145e-05,
      "loss": 0.6704,
      "mean_token_accuracy": 0.7951645739376545,
      "num_tokens": 562761740.0,
      "step": 17632
    },
    {
      "entropy": 0.5739548783749342,
      "epoch": 1.6223310964284277,
      "grad_norm": 0.15333665907382965,
      "learning_rate": 4.592572147084982e-05,
      "loss": 0.5515,
      "mean_token_accuracy": 0.828120481222868,
      "num_tokens": 562793105.0,
      "step": 17633
    },
    {
      "entropy": 0.6175371780991554,
      "epoch": 1.622423102799869,
      "grad_norm": 0.15530331432819366,
      "learning_rate": 4.59226546447082e-05,
      "loss": 0.6008,
      "mean_token_accuracy": 0.8153719939291477,
      "num_tokens": 562825052.0,
      "step": 17634
    },
    {
      "entropy": 0.6928404392674565,
      "epoch": 1.6225151091713101,
      "grad_norm": 0.16511346399784088,
      "learning_rate": 4.5919587818566566e-05,
      "loss": 0.6854,
      "mean_token_accuracy": 0.7915300652384758,
      "num_tokens": 562856837.0,
      "step": 17635
    },
    {
      "entropy": 0.6345274336636066,
      "epoch": 1.6226071155427513,
      "grad_norm": 0.14918546378612518,
      "learning_rate": 4.591652099242494e-05,
      "loss": 0.6114,
      "mean_token_accuracy": 0.8101187087595463,
      "num_tokens": 562889091.0,
      "step": 17636
    },
    {
      "entropy": 0.5618543687742203,
      "epoch": 1.6226991219141924,
      "grad_norm": 0.15746080875396729,
      "learning_rate": 4.5913454166283316e-05,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8304241858422756,
      "num_tokens": 562920806.0,
      "step": 17637
    },
    {
      "entropy": 0.7598466835916042,
      "epoch": 1.6227911282856338,
      "grad_norm": 0.16220489144325256,
      "learning_rate": 4.591038734014169e-05,
      "loss": 0.7461,
      "mean_token_accuracy": 0.7708836682140827,
      "num_tokens": 562952771.0,
      "step": 17638
    },
    {
      "entropy": 0.5612823837436736,
      "epoch": 1.6228831346570751,
      "grad_norm": 0.15604929625988007,
      "learning_rate": 4.5907320514000065e-05,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8331983610987663,
      "num_tokens": 562984754.0,
      "step": 17639
    },
    {
      "entropy": 0.6585187949240208,
      "epoch": 1.6229751410285163,
      "grad_norm": 0.15400755405426025,
      "learning_rate": 4.590425368785844e-05,
      "loss": 0.6515,
      "mean_token_accuracy": 0.8013216219842434,
      "num_tokens": 563017399.0,
      "step": 17640
    },
    {
      "entropy": 0.6228405460715294,
      "epoch": 1.6230671473999574,
      "grad_norm": 0.15416784584522247,
      "learning_rate": 4.590118686171681e-05,
      "loss": 0.6194,
      "mean_token_accuracy": 0.8087447732686996,
      "num_tokens": 563049116.0,
      "step": 17641
    },
    {
      "entropy": 0.6771945431828499,
      "epoch": 1.6231591537713985,
      "grad_norm": 0.1589067280292511,
      "learning_rate": 4.589812003557519e-05,
      "loss": 0.672,
      "mean_token_accuracy": 0.7943817935883999,
      "num_tokens": 563081217.0,
      "step": 17642
    },
    {
      "entropy": 0.5143127464689314,
      "epoch": 1.6232511601428399,
      "grad_norm": 0.1548255831003189,
      "learning_rate": 4.589505320943356e-05,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8421720489859581,
      "num_tokens": 563113530.0,
      "step": 17643
    },
    {
      "entropy": 0.7074330057948828,
      "epoch": 1.6233431665142812,
      "grad_norm": 0.16268198192119598,
      "learning_rate": 4.589198638329194e-05,
      "loss": 0.7051,
      "mean_token_accuracy": 0.7862961553037167,
      "num_tokens": 563145349.0,
      "step": 17644
    },
    {
      "entropy": 0.5230715044308454,
      "epoch": 1.6234351728857224,
      "grad_norm": 0.1401727795600891,
      "learning_rate": 4.5888919557150306e-05,
      "loss": 0.5077,
      "mean_token_accuracy": 0.8483936935663223,
      "num_tokens": 563177515.0,
      "step": 17645
    },
    {
      "entropy": 0.5833548940718174,
      "epoch": 1.6235271792571635,
      "grad_norm": 0.15289530158042908,
      "learning_rate": 4.588585273100868e-05,
      "loss": 0.5813,
      "mean_token_accuracy": 0.8203197978436947,
      "num_tokens": 563209667.0,
      "step": 17646
    },
    {
      "entropy": 0.5497688283212483,
      "epoch": 1.6236191856286046,
      "grad_norm": 0.14406396448612213,
      "learning_rate": 4.5882785904867055e-05,
      "loss": 0.535,
      "mean_token_accuracy": 0.8331570066511631,
      "num_tokens": 563242074.0,
      "step": 17647
    },
    {
      "entropy": 0.5713912667706609,
      "epoch": 1.623711192000046,
      "grad_norm": 0.15423175692558289,
      "learning_rate": 4.587971907872543e-05,
      "loss": 0.5745,
      "mean_token_accuracy": 0.8247703798115253,
      "num_tokens": 563274479.0,
      "step": 17648
    },
    {
      "entropy": 0.5874504493549466,
      "epoch": 1.6238031983714873,
      "grad_norm": 0.15666832029819489,
      "learning_rate": 4.5876652252583805e-05,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8285096809267998,
      "num_tokens": 563306775.0,
      "step": 17649
    },
    {
      "entropy": 0.5956499818712473,
      "epoch": 1.6238952047429285,
      "grad_norm": 0.15832985937595367,
      "learning_rate": 4.587358542644218e-05,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8175435066223145,
      "num_tokens": 563338290.0,
      "step": 17650
    },
    {
      "entropy": 0.5543025992810726,
      "epoch": 1.6239872111143696,
      "grad_norm": 0.15460754930973053,
      "learning_rate": 4.587051860030055e-05,
      "loss": 0.535,
      "mean_token_accuracy": 0.8324840255081654,
      "num_tokens": 563370523.0,
      "step": 17651
    },
    {
      "entropy": 0.5629763398319483,
      "epoch": 1.6240792174858107,
      "grad_norm": 0.15317487716674805,
      "learning_rate": 4.586745177415893e-05,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8274426460266113,
      "num_tokens": 563402539.0,
      "step": 17652
    },
    {
      "entropy": 0.7582437358796597,
      "epoch": 1.624171223857252,
      "grad_norm": 0.16917933523654938,
      "learning_rate": 4.5864384948017296e-05,
      "loss": 0.7475,
      "mean_token_accuracy": 0.7756477780640125,
      "num_tokens": 563433848.0,
      "step": 17653
    },
    {
      "entropy": 0.5388970449566841,
      "epoch": 1.6242632302286935,
      "grad_norm": 0.157659113407135,
      "learning_rate": 4.586131812187567e-05,
      "loss": 0.5265,
      "mean_token_accuracy": 0.8393608406186104,
      "num_tokens": 563465859.0,
      "step": 17654
    },
    {
      "entropy": 0.5339662004262209,
      "epoch": 1.6243552366001346,
      "grad_norm": 0.152385413646698,
      "learning_rate": 4.5858251295734046e-05,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8386702388525009,
      "num_tokens": 563498173.0,
      "step": 17655
    },
    {
      "entropy": 0.5243678651750088,
      "epoch": 1.6244472429715757,
      "grad_norm": 0.14253799617290497,
      "learning_rate": 4.585518446959242e-05,
      "loss": 0.5053,
      "mean_token_accuracy": 0.8406244739890099,
      "num_tokens": 563530163.0,
      "step": 17656
    },
    {
      "entropy": 0.5919871018268168,
      "epoch": 1.6245392493430169,
      "grad_norm": 0.15301775932312012,
      "learning_rate": 4.5852117643450795e-05,
      "loss": 0.5779,
      "mean_token_accuracy": 0.8225881569087505,
      "num_tokens": 563562759.0,
      "step": 17657
    },
    {
      "entropy": 0.6050185495987535,
      "epoch": 1.6246312557144582,
      "grad_norm": 0.15466831624507904,
      "learning_rate": 4.584905081730917e-05,
      "loss": 0.5972,
      "mean_token_accuracy": 0.8203265927731991,
      "num_tokens": 563594729.0,
      "step": 17658
    },
    {
      "entropy": 0.5794617682695389,
      "epoch": 1.6247232620858996,
      "grad_norm": 0.15113410353660583,
      "learning_rate": 4.584598399116754e-05,
      "loss": 0.5741,
      "mean_token_accuracy": 0.8217495009303093,
      "num_tokens": 563626021.0,
      "step": 17659
    },
    {
      "entropy": 0.6139328358694911,
      "epoch": 1.6248152684573407,
      "grad_norm": 0.15298758447170258,
      "learning_rate": 4.584291716502592e-05,
      "loss": 0.5882,
      "mean_token_accuracy": 0.8163884580135345,
      "num_tokens": 563658209.0,
      "step": 17660
    },
    {
      "entropy": 0.6089153904467821,
      "epoch": 1.6249072748287818,
      "grad_norm": 0.1578344702720642,
      "learning_rate": 4.583985033888429e-05,
      "loss": 0.6116,
      "mean_token_accuracy": 0.8106761090457439,
      "num_tokens": 563690320.0,
      "step": 17661
    },
    {
      "entropy": 0.6173359211534262,
      "epoch": 1.624999281200223,
      "grad_norm": 0.1549990177154541,
      "learning_rate": 4.583678351274266e-05,
      "loss": 0.615,
      "mean_token_accuracy": 0.8095545545220375,
      "num_tokens": 563722577.0,
      "step": 17662
    },
    {
      "entropy": 0.5849922075867653,
      "epoch": 1.6250912875716643,
      "grad_norm": 0.1489308476448059,
      "learning_rate": 4.583371668660104e-05,
      "loss": 0.5798,
      "mean_token_accuracy": 0.8206691108644009,
      "num_tokens": 563754372.0,
      "step": 17663
    },
    {
      "entropy": 0.6270868205465376,
      "epoch": 1.6251832939431057,
      "grad_norm": 0.1558988243341446,
      "learning_rate": 4.583064986045941e-05,
      "loss": 0.6259,
      "mean_token_accuracy": 0.8089127019047737,
      "num_tokens": 563785937.0,
      "step": 17664
    },
    {
      "entropy": 0.5787683655507863,
      "epoch": 1.6252753003145468,
      "grad_norm": 0.15214379131793976,
      "learning_rate": 4.582758303431779e-05,
      "loss": 0.5795,
      "mean_token_accuracy": 0.821601752191782,
      "num_tokens": 563817520.0,
      "step": 17665
    },
    {
      "entropy": 0.5457033955026418,
      "epoch": 1.625367306685988,
      "grad_norm": 0.14614084362983704,
      "learning_rate": 4.582451620817616e-05,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8339305743575096,
      "num_tokens": 563849321.0,
      "step": 17666
    },
    {
      "entropy": 0.5933140907436609,
      "epoch": 1.625459313057429,
      "grad_norm": 0.15046155452728271,
      "learning_rate": 4.5821449382034535e-05,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8206638917326927,
      "num_tokens": 563881769.0,
      "step": 17667
    },
    {
      "entropy": 0.5900199189782143,
      "epoch": 1.6255513194288704,
      "grad_norm": 0.15145346522331238,
      "learning_rate": 4.581838255589291e-05,
      "loss": 0.5737,
      "mean_token_accuracy": 0.8235291875898838,
      "num_tokens": 563914052.0,
      "step": 17668
    },
    {
      "entropy": 0.5507936752401292,
      "epoch": 1.6256433258003118,
      "grad_norm": 0.15655308961868286,
      "learning_rate": 4.5815315729751284e-05,
      "loss": 0.5397,
      "mean_token_accuracy": 0.8350730389356613,
      "num_tokens": 563946155.0,
      "step": 17669
    },
    {
      "entropy": 0.6542985234409571,
      "epoch": 1.625735332171753,
      "grad_norm": 0.16398899257183075,
      "learning_rate": 4.581224890360966e-05,
      "loss": 0.6396,
      "mean_token_accuracy": 0.8033582009375095,
      "num_tokens": 563977806.0,
      "step": 17670
    },
    {
      "entropy": 0.5916280625388026,
      "epoch": 1.625827338543194,
      "grad_norm": 0.15359823405742645,
      "learning_rate": 4.580918207746803e-05,
      "loss": 0.5892,
      "mean_token_accuracy": 0.820200502872467,
      "num_tokens": 564009964.0,
      "step": 17671
    },
    {
      "entropy": 0.6262435596436262,
      "epoch": 1.6259193449146352,
      "grad_norm": 0.15996752679347992,
      "learning_rate": 4.58061152513264e-05,
      "loss": 0.6077,
      "mean_token_accuracy": 0.8130194246768951,
      "num_tokens": 564042156.0,
      "step": 17672
    },
    {
      "entropy": 0.6193608674220741,
      "epoch": 1.6260113512860765,
      "grad_norm": 0.15443067252635956,
      "learning_rate": 4.580304842518478e-05,
      "loss": 0.597,
      "mean_token_accuracy": 0.8184283785521984,
      "num_tokens": 564073897.0,
      "step": 17673
    },
    {
      "entropy": 0.5564071331173182,
      "epoch": 1.626103357657518,
      "grad_norm": 0.1510579138994217,
      "learning_rate": 4.579998159904315e-05,
      "loss": 0.5425,
      "mean_token_accuracy": 0.8313402868807316,
      "num_tokens": 564106372.0,
      "step": 17674
    },
    {
      "entropy": 0.6169271692633629,
      "epoch": 1.626195364028959,
      "grad_norm": 0.16139031946659088,
      "learning_rate": 4.5796914772901525e-05,
      "loss": 0.6015,
      "mean_token_accuracy": 0.8176755867898464,
      "num_tokens": 564138333.0,
      "step": 17675
    },
    {
      "entropy": 0.7070359364151955,
      "epoch": 1.6262873704004002,
      "grad_norm": 0.16985167562961578,
      "learning_rate": 4.57938479467599e-05,
      "loss": 0.6957,
      "mean_token_accuracy": 0.786908321082592,
      "num_tokens": 564170435.0,
      "step": 17676
    },
    {
      "entropy": 0.5347176259383559,
      "epoch": 1.6263793767718413,
      "grad_norm": 0.14944179356098175,
      "learning_rate": 4.5790781120618274e-05,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8376003839075565,
      "num_tokens": 564202318.0,
      "step": 17677
    },
    {
      "entropy": 0.605217857286334,
      "epoch": 1.6264713831432827,
      "grad_norm": 0.15487545728683472,
      "learning_rate": 4.578771429447665e-05,
      "loss": 0.6015,
      "mean_token_accuracy": 0.8167965821921825,
      "num_tokens": 564234940.0,
      "step": 17678
    },
    {
      "entropy": 0.7166769374161959,
      "epoch": 1.626563389514724,
      "grad_norm": 0.16216430068016052,
      "learning_rate": 4.5784647468335024e-05,
      "loss": 0.7105,
      "mean_token_accuracy": 0.779553335160017,
      "num_tokens": 564266584.0,
      "step": 17679
    },
    {
      "entropy": 0.47070576529949903,
      "epoch": 1.6266553958861651,
      "grad_norm": 0.14424802362918854,
      "learning_rate": 4.578158064219339e-05,
      "loss": 0.4502,
      "mean_token_accuracy": 0.857589490711689,
      "num_tokens": 564298679.0,
      "step": 17680
    },
    {
      "entropy": 0.6333659775555134,
      "epoch": 1.6267474022576063,
      "grad_norm": 0.15668438374996185,
      "learning_rate": 4.577851381605177e-05,
      "loss": 0.6362,
      "mean_token_accuracy": 0.8059347569942474,
      "num_tokens": 564330051.0,
      "step": 17681
    },
    {
      "entropy": 0.5756493741646409,
      "epoch": 1.6268394086290474,
      "grad_norm": 0.15599721670150757,
      "learning_rate": 4.577544698991014e-05,
      "loss": 0.5761,
      "mean_token_accuracy": 0.8214837796986103,
      "num_tokens": 564362263.0,
      "step": 17682
    },
    {
      "entropy": 0.5726744495332241,
      "epoch": 1.6269314150004888,
      "grad_norm": 0.1535663902759552,
      "learning_rate": 4.577238016376852e-05,
      "loss": 0.5627,
      "mean_token_accuracy": 0.8299490809440613,
      "num_tokens": 564394219.0,
      "step": 17683
    },
    {
      "entropy": 0.5064241923391819,
      "epoch": 1.6270234213719301,
      "grad_norm": 0.14625124633312225,
      "learning_rate": 4.576931333762689e-05,
      "loss": 0.4951,
      "mean_token_accuracy": 0.8482900224626064,
      "num_tokens": 564426012.0,
      "step": 17684
    },
    {
      "entropy": 0.6130437031388283,
      "epoch": 1.6271154277433713,
      "grad_norm": 0.16093900799751282,
      "learning_rate": 4.5766246511485265e-05,
      "loss": 0.6071,
      "mean_token_accuracy": 0.8158838450908661,
      "num_tokens": 564456374.0,
      "step": 17685
    },
    {
      "entropy": 0.6166143855080009,
      "epoch": 1.6272074341148124,
      "grad_norm": 0.1607324182987213,
      "learning_rate": 4.576317968534364e-05,
      "loss": 0.6189,
      "mean_token_accuracy": 0.8111753389239311,
      "num_tokens": 564488374.0,
      "step": 17686
    },
    {
      "entropy": 0.6506353747099638,
      "epoch": 1.6272994404862535,
      "grad_norm": 0.15715819597244263,
      "learning_rate": 4.5760112859202014e-05,
      "loss": 0.6476,
      "mean_token_accuracy": 0.7984244041144848,
      "num_tokens": 564520505.0,
      "step": 17687
    },
    {
      "entropy": 0.5635434444993734,
      "epoch": 1.6273914468576949,
      "grad_norm": 0.15274780988693237,
      "learning_rate": 4.575704603306039e-05,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8271752446889877,
      "num_tokens": 564552589.0,
      "step": 17688
    },
    {
      "entropy": 0.5832740366458893,
      "epoch": 1.6274834532291362,
      "grad_norm": 0.16407296061515808,
      "learning_rate": 4.5753979206918763e-05,
      "loss": 0.5719,
      "mean_token_accuracy": 0.8264477849006653,
      "num_tokens": 564584696.0,
      "step": 17689
    },
    {
      "entropy": 0.6481256382539868,
      "epoch": 1.6275754596005774,
      "grad_norm": 0.15632404386997223,
      "learning_rate": 4.575091238077713e-05,
      "loss": 0.6417,
      "mean_token_accuracy": 0.8031194284558296,
      "num_tokens": 564616987.0,
      "step": 17690
    },
    {
      "entropy": 0.6063327994197607,
      "epoch": 1.6276674659720185,
      "grad_norm": 0.16098788380622864,
      "learning_rate": 4.574784555463551e-05,
      "loss": 0.5882,
      "mean_token_accuracy": 0.8193506486713886,
      "num_tokens": 564648922.0,
      "step": 17691
    },
    {
      "entropy": 0.5847358296159655,
      "epoch": 1.6277594723434596,
      "grad_norm": 0.1517340987920761,
      "learning_rate": 4.574477872849388e-05,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8252099007368088,
      "num_tokens": 564681408.0,
      "step": 17692
    },
    {
      "entropy": 0.7409079503268003,
      "epoch": 1.627851478714901,
      "grad_norm": 0.1572525054216385,
      "learning_rate": 4.5741711902352255e-05,
      "loss": 0.7297,
      "mean_token_accuracy": 0.777902714908123,
      "num_tokens": 564713588.0,
      "step": 17693
    },
    {
      "entropy": 0.6312433984130621,
      "epoch": 1.6279434850863423,
      "grad_norm": 0.1538809835910797,
      "learning_rate": 4.573864507621064e-05,
      "loss": 0.6214,
      "mean_token_accuracy": 0.8086367398500443,
      "num_tokens": 564745482.0,
      "step": 17694
    },
    {
      "entropy": 0.6812398340553045,
      "epoch": 1.6280354914577835,
      "grad_norm": 0.1586005836725235,
      "learning_rate": 4.5735578250069005e-05,
      "loss": 0.6615,
      "mean_token_accuracy": 0.7988447733223438,
      "num_tokens": 564777616.0,
      "step": 17695
    },
    {
      "entropy": 0.5311529925093055,
      "epoch": 1.6281274978292246,
      "grad_norm": 0.1403936743736267,
      "learning_rate": 4.573251142392738e-05,
      "loss": 0.504,
      "mean_token_accuracy": 0.8482943177223206,
      "num_tokens": 564810310.0,
      "step": 17696
    },
    {
      "entropy": 0.6180878113955259,
      "epoch": 1.6282195042006657,
      "grad_norm": 0.15237338840961456,
      "learning_rate": 4.5729444597785754e-05,
      "loss": 0.6007,
      "mean_token_accuracy": 0.8187608607113361,
      "num_tokens": 564841263.0,
      "step": 17697
    },
    {
      "entropy": 0.5615589146036655,
      "epoch": 1.628311510572107,
      "grad_norm": 0.16213995218276978,
      "learning_rate": 4.572637777164413e-05,
      "loss": 0.5543,
      "mean_token_accuracy": 0.8287797272205353,
      "num_tokens": 564873057.0,
      "step": 17698
    },
    {
      "entropy": 0.6471722200512886,
      "epoch": 1.6284035169435485,
      "grad_norm": 0.15311653912067413,
      "learning_rate": 4.57233109455025e-05,
      "loss": 0.6295,
      "mean_token_accuracy": 0.8057138882577419,
      "num_tokens": 564905121.0,
      "step": 17699
    },
    {
      "entropy": 0.5540884835645556,
      "epoch": 1.6284955233149896,
      "grad_norm": 0.14830787479877472,
      "learning_rate": 4.572024411936088e-05,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8350439742207527,
      "num_tokens": 564936884.0,
      "step": 17700
    },
    {
      "entropy": 0.6768482476472855,
      "epoch": 1.6285875296864307,
      "grad_norm": 0.1582619547843933,
      "learning_rate": 4.5717177293219246e-05,
      "loss": 0.6746,
      "mean_token_accuracy": 0.7932287380099297,
      "num_tokens": 564968748.0,
      "step": 17701
    },
    {
      "entropy": 0.7076269425451756,
      "epoch": 1.6286795360578719,
      "grad_norm": 0.16132910549640656,
      "learning_rate": 4.571411046707763e-05,
      "loss": 0.7001,
      "mean_token_accuracy": 0.7886176593601704,
      "num_tokens": 565000618.0,
      "step": 17702
    },
    {
      "entropy": 0.5836967825889587,
      "epoch": 1.6287715424293132,
      "grad_norm": 0.15399134159088135,
      "learning_rate": 4.5711043640935995e-05,
      "loss": 0.5657,
      "mean_token_accuracy": 0.8226247243583202,
      "num_tokens": 565032398.0,
      "step": 17703
    },
    {
      "entropy": 0.6351749580353498,
      "epoch": 1.6288635488007546,
      "grad_norm": 0.15603388845920563,
      "learning_rate": 4.5707976814794376e-05,
      "loss": 0.6218,
      "mean_token_accuracy": 0.8100393749773502,
      "num_tokens": 565064511.0,
      "step": 17704
    },
    {
      "entropy": 0.6225511310622096,
      "epoch": 1.6289555551721957,
      "grad_norm": 0.16173316538333893,
      "learning_rate": 4.5704909988652744e-05,
      "loss": 0.6313,
      "mean_token_accuracy": 0.8136365339159966,
      "num_tokens": 565096778.0,
      "step": 17705
    },
    {
      "entropy": 0.5992487650364637,
      "epoch": 1.6290475615436368,
      "grad_norm": 0.16168226301670074,
      "learning_rate": 4.570184316251112e-05,
      "loss": 0.5924,
      "mean_token_accuracy": 0.8155617453157902,
      "num_tokens": 565126812.0,
      "step": 17706
    },
    {
      "entropy": 0.5940585099160671,
      "epoch": 1.629139567915078,
      "grad_norm": 0.16000282764434814,
      "learning_rate": 4.5698776336369494e-05,
      "loss": 0.5944,
      "mean_token_accuracy": 0.8176786601543427,
      "num_tokens": 565158856.0,
      "step": 17707
    },
    {
      "entropy": 0.6836568070575595,
      "epoch": 1.6292315742865193,
      "grad_norm": 0.16237396001815796,
      "learning_rate": 4.569570951022787e-05,
      "loss": 0.6785,
      "mean_token_accuracy": 0.7921840101480484,
      "num_tokens": 565190651.0,
      "step": 17708
    },
    {
      "entropy": 0.6022575665265322,
      "epoch": 1.6293235806579607,
      "grad_norm": 0.15437814593315125,
      "learning_rate": 4.569264268408624e-05,
      "loss": 0.5994,
      "mean_token_accuracy": 0.8191561102867126,
      "num_tokens": 565223064.0,
      "step": 17709
    },
    {
      "entropy": 0.5276249293237925,
      "epoch": 1.6294155870294018,
      "grad_norm": 0.15720774233341217,
      "learning_rate": 4.568957585794462e-05,
      "loss": 0.5263,
      "mean_token_accuracy": 0.8356506489217281,
      "num_tokens": 565254059.0,
      "step": 17710
    },
    {
      "entropy": 0.671818945556879,
      "epoch": 1.629507593400843,
      "grad_norm": 0.16234326362609863,
      "learning_rate": 4.5686509031802985e-05,
      "loss": 0.6701,
      "mean_token_accuracy": 0.7976867184042931,
      "num_tokens": 565285686.0,
      "step": 17711
    },
    {
      "entropy": 0.612807109951973,
      "epoch": 1.629599599772284,
      "grad_norm": 0.15408039093017578,
      "learning_rate": 4.568344220566137e-05,
      "loss": 0.6172,
      "mean_token_accuracy": 0.809896606951952,
      "num_tokens": 565317786.0,
      "step": 17712
    },
    {
      "entropy": 0.48488541692495346,
      "epoch": 1.6296916061437254,
      "grad_norm": 0.14827130734920502,
      "learning_rate": 4.5680375379519735e-05,
      "loss": 0.4857,
      "mean_token_accuracy": 0.8533112816512585,
      "num_tokens": 565349544.0,
      "step": 17713
    },
    {
      "entropy": 0.653469929471612,
      "epoch": 1.6297836125151668,
      "grad_norm": 0.15538588166236877,
      "learning_rate": 4.567730855337811e-05,
      "loss": 0.653,
      "mean_token_accuracy": 0.7992041371762753,
      "num_tokens": 565382120.0,
      "step": 17714
    },
    {
      "entropy": 0.5729878321290016,
      "epoch": 1.629875618886608,
      "grad_norm": 0.16125242412090302,
      "learning_rate": 4.5674241727236484e-05,
      "loss": 0.5629,
      "mean_token_accuracy": 0.8291018269956112,
      "num_tokens": 565413519.0,
      "step": 17715
    },
    {
      "entropy": 0.6886250358074903,
      "epoch": 1.629967625258049,
      "grad_norm": 0.16019503772258759,
      "learning_rate": 4.567117490109486e-05,
      "loss": 0.6704,
      "mean_token_accuracy": 0.796610277146101,
      "num_tokens": 565444543.0,
      "step": 17716
    },
    {
      "entropy": 0.6031002411618829,
      "epoch": 1.6300596316294902,
      "grad_norm": 0.14923211932182312,
      "learning_rate": 4.566810807495323e-05,
      "loss": 0.5862,
      "mean_token_accuracy": 0.8231797032058239,
      "num_tokens": 565476289.0,
      "step": 17717
    },
    {
      "entropy": 0.5754358367994428,
      "epoch": 1.6301516380009315,
      "grad_norm": 0.15482982993125916,
      "learning_rate": 4.566504124881161e-05,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8268493488430977,
      "num_tokens": 565507799.0,
      "step": 17718
    },
    {
      "entropy": 0.5692971227690578,
      "epoch": 1.630243644372373,
      "grad_norm": 0.14836271107196808,
      "learning_rate": 4.5661974422669976e-05,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8297423124313354,
      "num_tokens": 565539580.0,
      "step": 17719
    },
    {
      "entropy": 0.5956183569505811,
      "epoch": 1.630335650743814,
      "grad_norm": 0.1582699865102768,
      "learning_rate": 4.565890759652836e-05,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8220105320215225,
      "num_tokens": 565572016.0,
      "step": 17720
    },
    {
      "entropy": 0.5412567900493741,
      "epoch": 1.6304276571152552,
      "grad_norm": 0.1479692906141281,
      "learning_rate": 4.5655840770386725e-05,
      "loss": 0.5277,
      "mean_token_accuracy": 0.8388317413628101,
      "num_tokens": 565603264.0,
      "step": 17721
    },
    {
      "entropy": 0.5945216882973909,
      "epoch": 1.6305196634866963,
      "grad_norm": 0.15065455436706543,
      "learning_rate": 4.56527739442451e-05,
      "loss": 0.5845,
      "mean_token_accuracy": 0.818368561565876,
      "num_tokens": 565635073.0,
      "step": 17722
    },
    {
      "entropy": 0.6134762968868017,
      "epoch": 1.6306116698581377,
      "grad_norm": 0.15282376110553741,
      "learning_rate": 4.564970711810348e-05,
      "loss": 0.5974,
      "mean_token_accuracy": 0.8170627802610397,
      "num_tokens": 565667357.0,
      "step": 17723
    },
    {
      "entropy": 0.6700414111837745,
      "epoch": 1.630703676229579,
      "grad_norm": 0.1538347601890564,
      "learning_rate": 4.564664029196185e-05,
      "loss": 0.6642,
      "mean_token_accuracy": 0.7994994819164276,
      "num_tokens": 565699500.0,
      "step": 17724
    },
    {
      "entropy": 0.6450831443071365,
      "epoch": 1.6307956826010201,
      "grad_norm": 0.16546469926834106,
      "learning_rate": 4.564357346582023e-05,
      "loss": 0.6381,
      "mean_token_accuracy": 0.8045657984912395,
      "num_tokens": 565730926.0,
      "step": 17725
    },
    {
      "entropy": 0.5716448193415999,
      "epoch": 1.6308876889724613,
      "grad_norm": 0.14779238402843475,
      "learning_rate": 4.56405066396786e-05,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8291682749986649,
      "num_tokens": 565763081.0,
      "step": 17726
    },
    {
      "entropy": 0.6585580445826054,
      "epoch": 1.6309796953439024,
      "grad_norm": 0.15961384773254395,
      "learning_rate": 4.563743981353697e-05,
      "loss": 0.67,
      "mean_token_accuracy": 0.7979151457548141,
      "num_tokens": 565794718.0,
      "step": 17727
    },
    {
      "entropy": 0.6761747673153877,
      "epoch": 1.6310717017153438,
      "grad_norm": 0.15753890573978424,
      "learning_rate": 4.563437298739535e-05,
      "loss": 0.6627,
      "mean_token_accuracy": 0.7964846268296242,
      "num_tokens": 565826664.0,
      "step": 17728
    },
    {
      "entropy": 0.5550961189437658,
      "epoch": 1.6311637080867851,
      "grad_norm": 0.14966675639152527,
      "learning_rate": 4.563130616125372e-05,
      "loss": 0.5458,
      "mean_token_accuracy": 0.8340779170393944,
      "num_tokens": 565859194.0,
      "step": 17729
    },
    {
      "entropy": 0.6015951046720147,
      "epoch": 1.6312557144582263,
      "grad_norm": 0.1591639667749405,
      "learning_rate": 4.56282393351121e-05,
      "loss": 0.5836,
      "mean_token_accuracy": 0.8235682547092438,
      "num_tokens": 565890880.0,
      "step": 17730
    },
    {
      "entropy": 0.4926480804570019,
      "epoch": 1.6313477208296674,
      "grad_norm": 0.1467362344264984,
      "learning_rate": 4.562517250897047e-05,
      "loss": 0.4788,
      "mean_token_accuracy": 0.8549846485257149,
      "num_tokens": 565922975.0,
      "step": 17731
    },
    {
      "entropy": 0.6417694613337517,
      "epoch": 1.6314397272011085,
      "grad_norm": 0.15493734180927277,
      "learning_rate": 4.562210568282884e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.8064266182482243,
      "num_tokens": 565954753.0,
      "step": 17732
    },
    {
      "entropy": 0.648925231769681,
      "epoch": 1.6315317335725499,
      "grad_norm": 0.15883740782737732,
      "learning_rate": 4.561903885668722e-05,
      "loss": 0.6446,
      "mean_token_accuracy": 0.8050539940595627,
      "num_tokens": 565986798.0,
      "step": 17733
    },
    {
      "entropy": 0.5831764377653599,
      "epoch": 1.6316237399439912,
      "grad_norm": 0.15652675926685333,
      "learning_rate": 4.561597203054559e-05,
      "loss": 0.555,
      "mean_token_accuracy": 0.8266134336590767,
      "num_tokens": 566018036.0,
      "step": 17734
    },
    {
      "entropy": 0.7192401140928268,
      "epoch": 1.6317157463154324,
      "grad_norm": 0.1622152477502823,
      "learning_rate": 4.561290520440396e-05,
      "loss": 0.7082,
      "mean_token_accuracy": 0.7846149913966656,
      "num_tokens": 566049749.0,
      "step": 17735
    },
    {
      "entropy": 0.6267928490415215,
      "epoch": 1.6318077526868735,
      "grad_norm": 0.15183961391448975,
      "learning_rate": 4.560983837826234e-05,
      "loss": 0.6083,
      "mean_token_accuracy": 0.8138791806995869,
      "num_tokens": 566081903.0,
      "step": 17736
    },
    {
      "entropy": 0.5274764220230281,
      "epoch": 1.6318997590583146,
      "grad_norm": 0.1455797255039215,
      "learning_rate": 4.560677155212071e-05,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8448412604629993,
      "num_tokens": 566114098.0,
      "step": 17737
    },
    {
      "entropy": 0.5295306760817766,
      "epoch": 1.631991765429756,
      "grad_norm": 0.14911915361881256,
      "learning_rate": 4.560370472597909e-05,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8436954542994499,
      "num_tokens": 566146433.0,
      "step": 17738
    },
    {
      "entropy": 0.6305457600392401,
      "epoch": 1.6320837718011973,
      "grad_norm": 0.1556706726551056,
      "learning_rate": 4.560063789983746e-05,
      "loss": 0.6236,
      "mean_token_accuracy": 0.8063069321215153,
      "num_tokens": 566178428.0,
      "step": 17739
    },
    {
      "entropy": 0.5847582826390862,
      "epoch": 1.6321757781726385,
      "grad_norm": 0.15947698056697845,
      "learning_rate": 4.559757107369583e-05,
      "loss": 0.5851,
      "mean_token_accuracy": 0.8222775645554066,
      "num_tokens": 566210143.0,
      "step": 17740
    },
    {
      "entropy": 0.6140529420226812,
      "epoch": 1.6322677845440796,
      "grad_norm": 0.1602889448404312,
      "learning_rate": 4.559450424755421e-05,
      "loss": 0.6028,
      "mean_token_accuracy": 0.8164141103625298,
      "num_tokens": 566242357.0,
      "step": 17741
    },
    {
      "entropy": 0.535666425479576,
      "epoch": 1.6323597909155207,
      "grad_norm": 0.14916430413722992,
      "learning_rate": 4.559143742141258e-05,
      "loss": 0.5202,
      "mean_token_accuracy": 0.8408259302377701,
      "num_tokens": 566274336.0,
      "step": 17742
    },
    {
      "entropy": 0.7594916671514511,
      "epoch": 1.632451797286962,
      "grad_norm": 0.17634864151477814,
      "learning_rate": 4.558837059527096e-05,
      "loss": 0.7517,
      "mean_token_accuracy": 0.7716458365321159,
      "num_tokens": 566306457.0,
      "step": 17743
    },
    {
      "entropy": 0.5743121183477342,
      "epoch": 1.6325438036584035,
      "grad_norm": 0.15907134115695953,
      "learning_rate": 4.558530376912933e-05,
      "loss": 0.5674,
      "mean_token_accuracy": 0.8300248980522156,
      "num_tokens": 566338946.0,
      "step": 17744
    },
    {
      "entropy": 0.6439201859757304,
      "epoch": 1.6326358100298446,
      "grad_norm": 0.15811960399150848,
      "learning_rate": 4.55822369429877e-05,
      "loss": 0.628,
      "mean_token_accuracy": 0.8089150711894035,
      "num_tokens": 566371592.0,
      "step": 17745
    },
    {
      "entropy": 0.6546051483601332,
      "epoch": 1.6327278164012857,
      "grad_norm": 0.16370141506195068,
      "learning_rate": 4.557917011684608e-05,
      "loss": 0.6493,
      "mean_token_accuracy": 0.8030550368130207,
      "num_tokens": 566403740.0,
      "step": 17746
    },
    {
      "entropy": 0.6200022241100669,
      "epoch": 1.6328198227727269,
      "grad_norm": 0.15597736835479736,
      "learning_rate": 4.557610329070445e-05,
      "loss": 0.6111,
      "mean_token_accuracy": 0.8135559037327766,
      "num_tokens": 566436089.0,
      "step": 17747
    },
    {
      "entropy": 0.6810074942186475,
      "epoch": 1.6329118291441682,
      "grad_norm": 0.1574532389640808,
      "learning_rate": 4.557303646456283e-05,
      "loss": 0.6711,
      "mean_token_accuracy": 0.7955805994570255,
      "num_tokens": 566468593.0,
      "step": 17748
    },
    {
      "entropy": 0.5316601051017642,
      "epoch": 1.6330038355156096,
      "grad_norm": 0.1518038511276245,
      "learning_rate": 4.55699696384212e-05,
      "loss": 0.5158,
      "mean_token_accuracy": 0.8421497568488121,
      "num_tokens": 566500576.0,
      "step": 17749
    },
    {
      "entropy": 0.5956124579533935,
      "epoch": 1.6330958418870507,
      "grad_norm": 0.16343475878238678,
      "learning_rate": 4.556690281227957e-05,
      "loss": 0.584,
      "mean_token_accuracy": 0.8223761208355427,
      "num_tokens": 566532674.0,
      "step": 17750
    },
    {
      "entropy": 0.7170812077820301,
      "epoch": 1.6331878482584918,
      "grad_norm": 0.16642442345619202,
      "learning_rate": 4.556383598613795e-05,
      "loss": 0.7021,
      "mean_token_accuracy": 0.7880812361836433,
      "num_tokens": 566564201.0,
      "step": 17751
    },
    {
      "entropy": 0.5196863091550767,
      "epoch": 1.633279854629933,
      "grad_norm": 0.1535627394914627,
      "learning_rate": 4.556076915999632e-05,
      "loss": 0.5132,
      "mean_token_accuracy": 0.839704904705286,
      "num_tokens": 566596055.0,
      "step": 17752
    },
    {
      "entropy": 0.5785280428826809,
      "epoch": 1.6333718610013743,
      "grad_norm": 0.15565554797649384,
      "learning_rate": 4.5557702333854693e-05,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8290782272815704,
      "num_tokens": 566627846.0,
      "step": 17753
    },
    {
      "entropy": 0.5529928728938103,
      "epoch": 1.6334638673728157,
      "grad_norm": 0.1608409881591797,
      "learning_rate": 4.5554635507713075e-05,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8292484097182751,
      "num_tokens": 566659896.0,
      "step": 17754
    },
    {
      "entropy": 0.6225582845509052,
      "epoch": 1.6335558737442568,
      "grad_norm": 0.15333408117294312,
      "learning_rate": 4.555156868157144e-05,
      "loss": 0.6048,
      "mean_token_accuracy": 0.813753817230463,
      "num_tokens": 566691710.0,
      "step": 17755
    },
    {
      "entropy": 0.5635590571910143,
      "epoch": 1.633647880115698,
      "grad_norm": 0.15239489078521729,
      "learning_rate": 4.554850185542982e-05,
      "loss": 0.5583,
      "mean_token_accuracy": 0.8286555670201778,
      "num_tokens": 566724129.0,
      "step": 17756
    },
    {
      "entropy": 0.6980666443705559,
      "epoch": 1.633739886487139,
      "grad_norm": 0.1670314073562622,
      "learning_rate": 4.554543502928819e-05,
      "loss": 0.6782,
      "mean_token_accuracy": 0.7916963435709476,
      "num_tokens": 566755490.0,
      "step": 17757
    },
    {
      "entropy": 0.5939077353104949,
      "epoch": 1.6338318928585804,
      "grad_norm": 0.1570827066898346,
      "learning_rate": 4.554236820314657e-05,
      "loss": 0.5816,
      "mean_token_accuracy": 0.8218638598918915,
      "num_tokens": 566787598.0,
      "step": 17758
    },
    {
      "entropy": 0.5785277206450701,
      "epoch": 1.6339238992300218,
      "grad_norm": 0.15895113348960876,
      "learning_rate": 4.553930137700494e-05,
      "loss": 0.5706,
      "mean_token_accuracy": 0.8244913257658482,
      "num_tokens": 566820098.0,
      "step": 17759
    },
    {
      "entropy": 0.584912552498281,
      "epoch": 1.634015905601463,
      "grad_norm": 0.14861831068992615,
      "learning_rate": 4.5536234550863316e-05,
      "loss": 0.5778,
      "mean_token_accuracy": 0.8220036700367928,
      "num_tokens": 566852724.0,
      "step": 17760
    },
    {
      "entropy": 0.687725119292736,
      "epoch": 1.634107911972904,
      "grad_norm": 0.15898382663726807,
      "learning_rate": 4.5533167724721684e-05,
      "loss": 0.6819,
      "mean_token_accuracy": 0.7896619103848934,
      "num_tokens": 566885072.0,
      "step": 17761
    },
    {
      "entropy": 0.5425646319054067,
      "epoch": 1.6341999183443452,
      "grad_norm": 0.15013667941093445,
      "learning_rate": 4.5530100898580065e-05,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8365427739918232,
      "num_tokens": 566917188.0,
      "step": 17762
    },
    {
      "entropy": 0.5123956776224077,
      "epoch": 1.6342919247157865,
      "grad_norm": 0.14569169282913208,
      "learning_rate": 4.552703407243843e-05,
      "loss": 0.5001,
      "mean_token_accuracy": 0.843796893954277,
      "num_tokens": 566949230.0,
      "step": 17763
    },
    {
      "entropy": 0.5749507937580347,
      "epoch": 1.634383931087228,
      "grad_norm": 0.15043513476848602,
      "learning_rate": 4.5523967246296815e-05,
      "loss": 0.5672,
      "mean_token_accuracy": 0.8262970931828022,
      "num_tokens": 566980988.0,
      "step": 17764
    },
    {
      "entropy": 0.7231700774282217,
      "epoch": 1.634475937458669,
      "grad_norm": 0.16470196843147278,
      "learning_rate": 4.552090042015518e-05,
      "loss": 0.709,
      "mean_token_accuracy": 0.7795716561377048,
      "num_tokens": 567012916.0,
      "step": 17765
    },
    {
      "entropy": 0.5887714186683297,
      "epoch": 1.6345679438301102,
      "grad_norm": 0.1531851589679718,
      "learning_rate": 4.551783359401356e-05,
      "loss": 0.5819,
      "mean_token_accuracy": 0.8224165327847004,
      "num_tokens": 567044953.0,
      "step": 17766
    },
    {
      "entropy": 0.6107837450690567,
      "epoch": 1.6346599502015513,
      "grad_norm": 0.15057827532291412,
      "learning_rate": 4.551476676787193e-05,
      "loss": 0.6155,
      "mean_token_accuracy": 0.8106069751083851,
      "num_tokens": 567077721.0,
      "step": 17767
    },
    {
      "entropy": 0.5020563248544931,
      "epoch": 1.6347519565729927,
      "grad_norm": 0.14149387180805206,
      "learning_rate": 4.5511699941730306e-05,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8488568961620331,
      "num_tokens": 567109958.0,
      "step": 17768
    },
    {
      "entropy": 0.7185599394142628,
      "epoch": 1.634843962944434,
      "grad_norm": 0.15944695472717285,
      "learning_rate": 4.550863311558868e-05,
      "loss": 0.7119,
      "mean_token_accuracy": 0.7874947786331177,
      "num_tokens": 567142262.0,
      "step": 17769
    },
    {
      "entropy": 0.6828294582664967,
      "epoch": 1.6349359693158751,
      "grad_norm": 0.1568102240562439,
      "learning_rate": 4.5505566289447056e-05,
      "loss": 0.6824,
      "mean_token_accuracy": 0.795043334364891,
      "num_tokens": 567173378.0,
      "step": 17770
    },
    {
      "entropy": 0.629958507604897,
      "epoch": 1.6350279756873163,
      "grad_norm": 0.15457698702812195,
      "learning_rate": 4.5502499463305424e-05,
      "loss": 0.6229,
      "mean_token_accuracy": 0.8091951198875904,
      "num_tokens": 567205446.0,
      "step": 17771
    },
    {
      "entropy": 0.6139678210020065,
      "epoch": 1.6351199820587574,
      "grad_norm": 0.15319553017616272,
      "learning_rate": 4.5499432637163805e-05,
      "loss": 0.6083,
      "mean_token_accuracy": 0.8119939491152763,
      "num_tokens": 567237422.0,
      "step": 17772
    },
    {
      "entropy": 0.6735959891229868,
      "epoch": 1.6352119884301988,
      "grad_norm": 0.1627088487148285,
      "learning_rate": 4.549636581102217e-05,
      "loss": 0.6612,
      "mean_token_accuracy": 0.7985762022435665,
      "num_tokens": 567268927.0,
      "step": 17773
    },
    {
      "entropy": 0.5734220016747713,
      "epoch": 1.6353039948016401,
      "grad_norm": 0.14508207142353058,
      "learning_rate": 4.549329898488055e-05,
      "loss": 0.5596,
      "mean_token_accuracy": 0.8303388208150864,
      "num_tokens": 567301469.0,
      "step": 17774
    },
    {
      "entropy": 0.5848006047308445,
      "epoch": 1.6353960011730813,
      "grad_norm": 0.15034236013889313,
      "learning_rate": 4.549023215873892e-05,
      "loss": 0.5757,
      "mean_token_accuracy": 0.8234538920223713,
      "num_tokens": 567333282.0,
      "step": 17775
    },
    {
      "entropy": 0.5486111296340823,
      "epoch": 1.6354880075445224,
      "grad_norm": 0.14663521945476532,
      "learning_rate": 4.54871653325973e-05,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8331131748855114,
      "num_tokens": 567365422.0,
      "step": 17776
    },
    {
      "entropy": 0.646320978179574,
      "epoch": 1.6355800139159635,
      "grad_norm": 0.1558404415845871,
      "learning_rate": 4.548409850645567e-05,
      "loss": 0.6218,
      "mean_token_accuracy": 0.8082056306302547,
      "num_tokens": 567396893.0,
      "step": 17777
    },
    {
      "entropy": 0.4919106010347605,
      "epoch": 1.6356720202874049,
      "grad_norm": 0.14776845276355743,
      "learning_rate": 4.5481031680314046e-05,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8509463965892792,
      "num_tokens": 567428206.0,
      "step": 17778
    },
    {
      "entropy": 0.5801529586315155,
      "epoch": 1.6357640266588462,
      "grad_norm": 0.15442918241024017,
      "learning_rate": 4.5477964854172414e-05,
      "loss": 0.5751,
      "mean_token_accuracy": 0.823800578713417,
      "num_tokens": 567460289.0,
      "step": 17779
    },
    {
      "entropy": 0.5617355015128851,
      "epoch": 1.6358560330302874,
      "grad_norm": 0.15574878454208374,
      "learning_rate": 4.5474898028030795e-05,
      "loss": 0.5543,
      "mean_token_accuracy": 0.8294081166386604,
      "num_tokens": 567492282.0,
      "step": 17780
    },
    {
      "entropy": 0.5961917079985142,
      "epoch": 1.6359480394017285,
      "grad_norm": 0.15985161066055298,
      "learning_rate": 4.547183120188916e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8194462768733501,
      "num_tokens": 567524101.0,
      "step": 17781
    },
    {
      "entropy": 0.5674320794641972,
      "epoch": 1.6360400457731696,
      "grad_norm": 0.1484011858701706,
      "learning_rate": 4.546876437574754e-05,
      "loss": 0.5517,
      "mean_token_accuracy": 0.8255032263696194,
      "num_tokens": 567556177.0,
      "step": 17782
    },
    {
      "entropy": 0.7372874859720469,
      "epoch": 1.636132052144611,
      "grad_norm": 0.16614125669002533,
      "learning_rate": 4.546569754960591e-05,
      "loss": 0.7191,
      "mean_token_accuracy": 0.7824387066066265,
      "num_tokens": 567588094.0,
      "step": 17783
    },
    {
      "entropy": 0.598331113345921,
      "epoch": 1.6362240585160523,
      "grad_norm": 0.16421473026275635,
      "learning_rate": 4.546263072346429e-05,
      "loss": 0.591,
      "mean_token_accuracy": 0.8194397427141666,
      "num_tokens": 567619201.0,
      "step": 17784
    },
    {
      "entropy": 0.5843673180788755,
      "epoch": 1.6363160648874935,
      "grad_norm": 0.15450742840766907,
      "learning_rate": 4.545956389732267e-05,
      "loss": 0.5804,
      "mean_token_accuracy": 0.8210508450865746,
      "num_tokens": 567650819.0,
      "step": 17785
    },
    {
      "entropy": 0.7707011066377163,
      "epoch": 1.6364080712589346,
      "grad_norm": 0.1670563966035843,
      "learning_rate": 4.5456497071181036e-05,
      "loss": 0.7771,
      "mean_token_accuracy": 0.766554232686758,
      "num_tokens": 567683131.0,
      "step": 17786
    },
    {
      "entropy": 0.6320126038044691,
      "epoch": 1.6365000776303757,
      "grad_norm": 0.1572142243385315,
      "learning_rate": 4.545343024503941e-05,
      "loss": 0.6277,
      "mean_token_accuracy": 0.8070419616997242,
      "num_tokens": 567715297.0,
      "step": 17787
    },
    {
      "entropy": 0.4892138526774943,
      "epoch": 1.636592084001817,
      "grad_norm": 0.1437394767999649,
      "learning_rate": 4.5450363418897786e-05,
      "loss": 0.4784,
      "mean_token_accuracy": 0.8516903892159462,
      "num_tokens": 567747476.0,
      "step": 17788
    },
    {
      "entropy": 0.6410316377878189,
      "epoch": 1.6366840903732585,
      "grad_norm": 0.15734504163265228,
      "learning_rate": 4.544729659275616e-05,
      "loss": 0.6329,
      "mean_token_accuracy": 0.8092267215251923,
      "num_tokens": 567779951.0,
      "step": 17789
    },
    {
      "entropy": 0.668855395168066,
      "epoch": 1.6367760967446996,
      "grad_norm": 0.16092686355113983,
      "learning_rate": 4.5444229766614535e-05,
      "loss": 0.6578,
      "mean_token_accuracy": 0.7966791428625584,
      "num_tokens": 567810955.0,
      "step": 17790
    },
    {
      "entropy": 0.6583823319524527,
      "epoch": 1.6368681031161407,
      "grad_norm": 0.17409773170948029,
      "learning_rate": 4.544116294047291e-05,
      "loss": 0.6514,
      "mean_token_accuracy": 0.7995495274662971,
      "num_tokens": 567843100.0,
      "step": 17791
    },
    {
      "entropy": 0.5311616696417332,
      "epoch": 1.6369601094875819,
      "grad_norm": 0.1450931578874588,
      "learning_rate": 4.543809611433128e-05,
      "loss": 0.5276,
      "mean_token_accuracy": 0.8366958871483803,
      "num_tokens": 567875360.0,
      "step": 17792
    },
    {
      "entropy": 0.5969677940011024,
      "epoch": 1.6370521158590232,
      "grad_norm": 0.15007616579532623,
      "learning_rate": 4.543502928818966e-05,
      "loss": 0.596,
      "mean_token_accuracy": 0.8150299750268459,
      "num_tokens": 567907423.0,
      "step": 17793
    },
    {
      "entropy": 0.7279502786695957,
      "epoch": 1.6371441222304646,
      "grad_norm": 0.1708497703075409,
      "learning_rate": 4.543196246204803e-05,
      "loss": 0.7273,
      "mean_token_accuracy": 0.7838058955967426,
      "num_tokens": 567940165.0,
      "step": 17794
    },
    {
      "entropy": 0.7117631137371063,
      "epoch": 1.6372361286019057,
      "grad_norm": 0.16621950268745422,
      "learning_rate": 4.54288956359064e-05,
      "loss": 0.707,
      "mean_token_accuracy": 0.7858130820095539,
      "num_tokens": 567972103.0,
      "step": 17795
    },
    {
      "entropy": 0.6202612635679543,
      "epoch": 1.6373281349733468,
      "grad_norm": 0.1523170918226242,
      "learning_rate": 4.5425828809764776e-05,
      "loss": 0.5964,
      "mean_token_accuracy": 0.8166312947869301,
      "num_tokens": 568004127.0,
      "step": 17796
    },
    {
      "entropy": 0.6033721873536706,
      "epoch": 1.637420141344788,
      "grad_norm": 0.15248997509479523,
      "learning_rate": 4.542276198362315e-05,
      "loss": 0.6,
      "mean_token_accuracy": 0.8169702738523483,
      "num_tokens": 568036476.0,
      "step": 17797
    },
    {
      "entropy": 0.5653615780174732,
      "epoch": 1.6375121477162293,
      "grad_norm": 0.1578492820262909,
      "learning_rate": 4.5419695157481525e-05,
      "loss": 0.5518,
      "mean_token_accuracy": 0.833521343767643,
      "num_tokens": 568068024.0,
      "step": 17798
    },
    {
      "entropy": 0.6415158901363611,
      "epoch": 1.6376041540876707,
      "grad_norm": 0.15248458087444305,
      "learning_rate": 4.54166283313399e-05,
      "loss": 0.6269,
      "mean_token_accuracy": 0.8031349182128906,
      "num_tokens": 568099948.0,
      "step": 17799
    },
    {
      "entropy": 0.5837619639933109,
      "epoch": 1.6376961604591118,
      "grad_norm": 0.15485629439353943,
      "learning_rate": 4.541356150519827e-05,
      "loss": 0.5727,
      "mean_token_accuracy": 0.8231355957686901,
      "num_tokens": 568131491.0,
      "step": 17800
    },
    {
      "entropy": 0.5904533956199884,
      "epoch": 1.637788166830553,
      "grad_norm": 0.1583954095840454,
      "learning_rate": 4.541049467905665e-05,
      "loss": 0.5753,
      "mean_token_accuracy": 0.8200223185122013,
      "num_tokens": 568162599.0,
      "step": 17801
    },
    {
      "entropy": 0.6071220356971025,
      "epoch": 1.637880173201994,
      "grad_norm": 0.16112275421619415,
      "learning_rate": 4.540742785291502e-05,
      "loss": 0.6038,
      "mean_token_accuracy": 0.8183063715696335,
      "num_tokens": 568194758.0,
      "step": 17802
    },
    {
      "entropy": 0.6299661677330732,
      "epoch": 1.6379721795734354,
      "grad_norm": 0.15681621432304382,
      "learning_rate": 4.54043610267734e-05,
      "loss": 0.5991,
      "mean_token_accuracy": 0.8142127729952335,
      "num_tokens": 568226340.0,
      "step": 17803
    },
    {
      "entropy": 0.4692814676091075,
      "epoch": 1.6380641859448768,
      "grad_norm": 0.15052294731140137,
      "learning_rate": 4.5401294200631767e-05,
      "loss": 0.4547,
      "mean_token_accuracy": 0.85740976780653,
      "num_tokens": 568258215.0,
      "step": 17804
    },
    {
      "entropy": 0.5598073117434978,
      "epoch": 1.638156192316318,
      "grad_norm": 0.15504179894924164,
      "learning_rate": 4.539822737449014e-05,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8310186676681042,
      "num_tokens": 568290026.0,
      "step": 17805
    },
    {
      "entropy": 0.7014594171196222,
      "epoch": 1.638248198687759,
      "grad_norm": 0.16531766951084137,
      "learning_rate": 4.5395160548348516e-05,
      "loss": 0.6906,
      "mean_token_accuracy": 0.786021288484335,
      "num_tokens": 568322044.0,
      "step": 17806
    },
    {
      "entropy": 0.7125425543636084,
      "epoch": 1.6383402050592002,
      "grad_norm": 0.16337864100933075,
      "learning_rate": 4.539209372220689e-05,
      "loss": 0.7012,
      "mean_token_accuracy": 0.7905019298195839,
      "num_tokens": 568354260.0,
      "step": 17807
    },
    {
      "entropy": 0.6542597953230143,
      "epoch": 1.6384322114306415,
      "grad_norm": 0.154709130525589,
      "learning_rate": 4.5389026896065265e-05,
      "loss": 0.6446,
      "mean_token_accuracy": 0.8022446371614933,
      "num_tokens": 568386302.0,
      "step": 17808
    },
    {
      "entropy": 0.5882760342210531,
      "epoch": 1.638524217802083,
      "grad_norm": 0.15233469009399414,
      "learning_rate": 4.538596006992364e-05,
      "loss": 0.5767,
      "mean_token_accuracy": 0.8221522271633148,
      "num_tokens": 568418537.0,
      "step": 17809
    },
    {
      "entropy": 0.49031928507611156,
      "epoch": 1.638616224173524,
      "grad_norm": 0.1427900791168213,
      "learning_rate": 4.538289324378201e-05,
      "loss": 0.4658,
      "mean_token_accuracy": 0.8562346138060093,
      "num_tokens": 568450904.0,
      "step": 17810
    },
    {
      "entropy": 0.6796510871499777,
      "epoch": 1.6387082305449652,
      "grad_norm": 0.16130782663822174,
      "learning_rate": 4.537982641764039e-05,
      "loss": 0.6813,
      "mean_token_accuracy": 0.794017605483532,
      "num_tokens": 568482876.0,
      "step": 17811
    },
    {
      "entropy": 0.5977785484865308,
      "epoch": 1.6388002369164063,
      "grad_norm": 0.15171095728874207,
      "learning_rate": 4.537675959149876e-05,
      "loss": 0.5903,
      "mean_token_accuracy": 0.8177107200026512,
      "num_tokens": 568515625.0,
      "step": 17812
    },
    {
      "entropy": 0.6103422180749476,
      "epoch": 1.6388922432878477,
      "grad_norm": 0.15356676280498505,
      "learning_rate": 4.537369276535713e-05,
      "loss": 0.6031,
      "mean_token_accuracy": 0.8145244494080544,
      "num_tokens": 568546917.0,
      "step": 17813
    },
    {
      "entropy": 0.5820826329290867,
      "epoch": 1.638984249659289,
      "grad_norm": 0.15770204365253448,
      "learning_rate": 4.5370625939215506e-05,
      "loss": 0.5778,
      "mean_token_accuracy": 0.8190198726952076,
      "num_tokens": 568579283.0,
      "step": 17814
    },
    {
      "entropy": 0.5505415452644229,
      "epoch": 1.6390762560307301,
      "grad_norm": 0.15301935374736786,
      "learning_rate": 4.536755911307388e-05,
      "loss": 0.5439,
      "mean_token_accuracy": 0.832745473831892,
      "num_tokens": 568611031.0,
      "step": 17815
    },
    {
      "entropy": 0.5719368513673544,
      "epoch": 1.6391682624021713,
      "grad_norm": 0.1585230529308319,
      "learning_rate": 4.5364492286932256e-05,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8311421684920788,
      "num_tokens": 568643322.0,
      "step": 17816
    },
    {
      "entropy": 0.6187512120231986,
      "epoch": 1.6392602687736124,
      "grad_norm": 0.15910299122333527,
      "learning_rate": 4.536142546079063e-05,
      "loss": 0.6021,
      "mean_token_accuracy": 0.8157400265336037,
      "num_tokens": 568675441.0,
      "step": 17817
    },
    {
      "entropy": 0.6333004301413894,
      "epoch": 1.6393522751450538,
      "grad_norm": 0.15305499732494354,
      "learning_rate": 4.5358358634649005e-05,
      "loss": 0.619,
      "mean_token_accuracy": 0.8128488138318062,
      "num_tokens": 568707313.0,
      "step": 17818
    },
    {
      "entropy": 0.5162450522184372,
      "epoch": 1.6394442815164951,
      "grad_norm": 0.1450931280851364,
      "learning_rate": 4.535529180850738e-05,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8418653048574924,
      "num_tokens": 568739242.0,
      "step": 17819
    },
    {
      "entropy": 0.7408787291496992,
      "epoch": 1.6395362878879363,
      "grad_norm": 0.16380265355110168,
      "learning_rate": 4.5352224982365754e-05,
      "loss": 0.7387,
      "mean_token_accuracy": 0.7739404253661633,
      "num_tokens": 568771017.0,
      "step": 17820
    },
    {
      "entropy": 0.5299119893461466,
      "epoch": 1.6396282942593774,
      "grad_norm": 0.15229614078998566,
      "learning_rate": 4.534915815622412e-05,
      "loss": 0.5124,
      "mean_token_accuracy": 0.8417514786124229,
      "num_tokens": 568802992.0,
      "step": 17821
    },
    {
      "entropy": 0.6376312682405114,
      "epoch": 1.6397203006308185,
      "grad_norm": 0.16279523074626923,
      "learning_rate": 4.5346091330082503e-05,
      "loss": 0.6172,
      "mean_token_accuracy": 0.8071298561990261,
      "num_tokens": 568834411.0,
      "step": 17822
    },
    {
      "entropy": 0.6111845523118973,
      "epoch": 1.6398123070022599,
      "grad_norm": 0.15742990374565125,
      "learning_rate": 4.534302450394087e-05,
      "loss": 0.6169,
      "mean_token_accuracy": 0.8119388967752457,
      "num_tokens": 568866368.0,
      "step": 17823
    },
    {
      "entropy": 0.57645224314183,
      "epoch": 1.6399043133737012,
      "grad_norm": 0.1588544100522995,
      "learning_rate": 4.533995767779925e-05,
      "loss": 0.5502,
      "mean_token_accuracy": 0.8278718329966068,
      "num_tokens": 568897603.0,
      "step": 17824
    },
    {
      "entropy": 0.5441555827856064,
      "epoch": 1.6399963197451424,
      "grad_norm": 0.1524433195590973,
      "learning_rate": 4.533689085165762e-05,
      "loss": 0.537,
      "mean_token_accuracy": 0.835609495639801,
      "num_tokens": 568929917.0,
      "step": 17825
    },
    {
      "entropy": 0.5886726975440979,
      "epoch": 1.6400883261165835,
      "grad_norm": 0.155929297208786,
      "learning_rate": 4.5333824025515995e-05,
      "loss": 0.5762,
      "mean_token_accuracy": 0.8229660168290138,
      "num_tokens": 568961817.0,
      "step": 17826
    },
    {
      "entropy": 0.6999181918799877,
      "epoch": 1.6401803324880246,
      "grad_norm": 0.1632905900478363,
      "learning_rate": 4.533075719937437e-05,
      "loss": 0.691,
      "mean_token_accuracy": 0.7900977730751038,
      "num_tokens": 568993651.0,
      "step": 17827
    },
    {
      "entropy": 0.6108215004205704,
      "epoch": 1.640272338859466,
      "grad_norm": 0.1589670181274414,
      "learning_rate": 4.5327690373232745e-05,
      "loss": 0.6105,
      "mean_token_accuracy": 0.8115341700613499,
      "num_tokens": 569025487.0,
      "step": 17828
    },
    {
      "entropy": 0.579405982978642,
      "epoch": 1.6403643452309074,
      "grad_norm": 0.1589975655078888,
      "learning_rate": 4.532462354709112e-05,
      "loss": 0.5722,
      "mean_token_accuracy": 0.8273649215698242,
      "num_tokens": 569057457.0,
      "step": 17829
    },
    {
      "entropy": 0.6391524914652109,
      "epoch": 1.6404563516023485,
      "grad_norm": 0.15762671828269958,
      "learning_rate": 4.5321556720949494e-05,
      "loss": 0.6325,
      "mean_token_accuracy": 0.804540291428566,
      "num_tokens": 569088490.0,
      "step": 17830
    },
    {
      "entropy": 0.5500223636627197,
      "epoch": 1.6405483579737896,
      "grad_norm": 0.1521042287349701,
      "learning_rate": 4.531848989480786e-05,
      "loss": 0.5432,
      "mean_token_accuracy": 0.8297235742211342,
      "num_tokens": 569120317.0,
      "step": 17831
    },
    {
      "entropy": 0.7172684241086245,
      "epoch": 1.6406403643452308,
      "grad_norm": 0.16500726342201233,
      "learning_rate": 4.531542306866624e-05,
      "loss": 0.7201,
      "mean_token_accuracy": 0.782511867582798,
      "num_tokens": 569152400.0,
      "step": 17832
    },
    {
      "entropy": 0.5404115132987499,
      "epoch": 1.640732370716672,
      "grad_norm": 0.1470848023891449,
      "learning_rate": 4.531235624252461e-05,
      "loss": 0.5339,
      "mean_token_accuracy": 0.8372330330312252,
      "num_tokens": 569185144.0,
      "step": 17833
    },
    {
      "entropy": 0.5863529406487942,
      "epoch": 1.6408243770881135,
      "grad_norm": 0.15621547400951385,
      "learning_rate": 4.5309289416382986e-05,
      "loss": 0.5711,
      "mean_token_accuracy": 0.8220230303704739,
      "num_tokens": 569216629.0,
      "step": 17834
    },
    {
      "entropy": 0.5420011412352324,
      "epoch": 1.6409163834595546,
      "grad_norm": 0.14734740555286407,
      "learning_rate": 4.530622259024136e-05,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8315859287977219,
      "num_tokens": 569249036.0,
      "step": 17835
    },
    {
      "entropy": 0.6040893206372857,
      "epoch": 1.6410083898309957,
      "grad_norm": 0.1513812392950058,
      "learning_rate": 4.5303155764099735e-05,
      "loss": 0.5885,
      "mean_token_accuracy": 0.819236259907484,
      "num_tokens": 569280919.0,
      "step": 17836
    },
    {
      "entropy": 0.5479511525481939,
      "epoch": 1.6411003962024369,
      "grad_norm": 0.14783820509910583,
      "learning_rate": 4.530008893795811e-05,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8362606950104237,
      "num_tokens": 569313155.0,
      "step": 17837
    },
    {
      "entropy": 0.6034442530944943,
      "epoch": 1.6411924025738782,
      "grad_norm": 0.15632598102092743,
      "learning_rate": 4.5297022111816484e-05,
      "loss": 0.5767,
      "mean_token_accuracy": 0.820350144058466,
      "num_tokens": 569345433.0,
      "step": 17838
    },
    {
      "entropy": 0.6146118110045791,
      "epoch": 1.6412844089453196,
      "grad_norm": 0.16255930066108704,
      "learning_rate": 4.529395528567485e-05,
      "loss": 0.6082,
      "mean_token_accuracy": 0.8106978014111519,
      "num_tokens": 569376901.0,
      "step": 17839
    },
    {
      "entropy": 0.6487459354102612,
      "epoch": 1.6413764153167607,
      "grad_norm": 0.15785326063632965,
      "learning_rate": 4.5290888459533234e-05,
      "loss": 0.6283,
      "mean_token_accuracy": 0.8028433509171009,
      "num_tokens": 569408524.0,
      "step": 17840
    },
    {
      "entropy": 0.5624745720997453,
      "epoch": 1.6414684216882018,
      "grad_norm": 0.1524367481470108,
      "learning_rate": 4.52878216333916e-05,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8303419835865498,
      "num_tokens": 569440302.0,
      "step": 17841
    },
    {
      "entropy": 0.590660153888166,
      "epoch": 1.641560428059643,
      "grad_norm": 0.15239939093589783,
      "learning_rate": 4.5284754807249976e-05,
      "loss": 0.5632,
      "mean_token_accuracy": 0.8239805288612843,
      "num_tokens": 569471818.0,
      "step": 17842
    },
    {
      "entropy": 0.4961615726351738,
      "epoch": 1.6416524344310843,
      "grad_norm": 0.13864128291606903,
      "learning_rate": 4.528168798110835e-05,
      "loss": 0.4895,
      "mean_token_accuracy": 0.8483363799750805,
      "num_tokens": 569503695.0,
      "step": 17843
    },
    {
      "entropy": 0.6468328088521957,
      "epoch": 1.6417444408025257,
      "grad_norm": 0.15608271956443787,
      "learning_rate": 4.5278621154966725e-05,
      "loss": 0.6374,
      "mean_token_accuracy": 0.8044871389865875,
      "num_tokens": 569535125.0,
      "step": 17844
    },
    {
      "entropy": 0.5911794654093683,
      "epoch": 1.6418364471739668,
      "grad_norm": 0.15716880559921265,
      "learning_rate": 4.527555432882511e-05,
      "loss": 0.5812,
      "mean_token_accuracy": 0.8219478502869606,
      "num_tokens": 569566577.0,
      "step": 17845
    },
    {
      "entropy": 0.5477539673447609,
      "epoch": 1.641928453545408,
      "grad_norm": 0.15161767601966858,
      "learning_rate": 4.5272487502683475e-05,
      "loss": 0.5395,
      "mean_token_accuracy": 0.8342555500566959,
      "num_tokens": 569597941.0,
      "step": 17846
    },
    {
      "entropy": 0.4751307424157858,
      "epoch": 1.642020459916849,
      "grad_norm": 0.14001671969890594,
      "learning_rate": 4.526942067654185e-05,
      "loss": 0.4566,
      "mean_token_accuracy": 0.8582270108163357,
      "num_tokens": 569630430.0,
      "step": 17847
    },
    {
      "entropy": 0.5175348808988929,
      "epoch": 1.6421124662882904,
      "grad_norm": 0.14671681821346283,
      "learning_rate": 4.5266353850400224e-05,
      "loss": 0.5021,
      "mean_token_accuracy": 0.8423979058861732,
      "num_tokens": 569661939.0,
      "step": 17848
    },
    {
      "entropy": 0.5502931252121925,
      "epoch": 1.6422044726597318,
      "grad_norm": 0.14742767810821533,
      "learning_rate": 4.52632870242586e-05,
      "loss": 0.5393,
      "mean_token_accuracy": 0.8329269140958786,
      "num_tokens": 569694311.0,
      "step": 17849
    },
    {
      "entropy": 0.5084139192476869,
      "epoch": 1.642296479031173,
      "grad_norm": 0.1484837383031845,
      "learning_rate": 4.526022019811697e-05,
      "loss": 0.4872,
      "mean_token_accuracy": 0.8484311029314995,
      "num_tokens": 569725853.0,
      "step": 17850
    },
    {
      "entropy": 0.6740380022674799,
      "epoch": 1.642388485402614,
      "grad_norm": 0.1637452244758606,
      "learning_rate": 4.525715337197535e-05,
      "loss": 0.6692,
      "mean_token_accuracy": 0.7984800226986408,
      "num_tokens": 569757801.0,
      "step": 17851
    },
    {
      "entropy": 0.6690152175724506,
      "epoch": 1.6424804917740552,
      "grad_norm": 0.15527324378490448,
      "learning_rate": 4.5254086545833716e-05,
      "loss": 0.6757,
      "mean_token_accuracy": 0.795094758272171,
      "num_tokens": 569789559.0,
      "step": 17852
    },
    {
      "entropy": 0.5720125497318804,
      "epoch": 1.6425724981454966,
      "grad_norm": 0.1494637131690979,
      "learning_rate": 4.52510197196921e-05,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8272472806274891,
      "num_tokens": 569821973.0,
      "step": 17853
    },
    {
      "entropy": 0.626431580632925,
      "epoch": 1.642664504516938,
      "grad_norm": 0.1522340476512909,
      "learning_rate": 4.5247952893550465e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.8099761120975018,
      "num_tokens": 569853883.0,
      "step": 17854
    },
    {
      "entropy": 0.6042951699346304,
      "epoch": 1.642756510888379,
      "grad_norm": 0.15866093337535858,
      "learning_rate": 4.524488606740884e-05,
      "loss": 0.5963,
      "mean_token_accuracy": 0.8134225308895111,
      "num_tokens": 569884725.0,
      "step": 17855
    },
    {
      "entropy": 0.5216146684251726,
      "epoch": 1.6428485172598202,
      "grad_norm": 0.14628179371356964,
      "learning_rate": 4.5241819241267214e-05,
      "loss": 0.5058,
      "mean_token_accuracy": 0.8434362709522247,
      "num_tokens": 569916965.0,
      "step": 17856
    },
    {
      "entropy": 0.6360859367996454,
      "epoch": 1.6429405236312613,
      "grad_norm": 0.1597154140472412,
      "learning_rate": 4.523875241512559e-05,
      "loss": 0.6129,
      "mean_token_accuracy": 0.8092762306332588,
      "num_tokens": 569948783.0,
      "step": 17857
    },
    {
      "entropy": 0.5888946745544672,
      "epoch": 1.6430325300027027,
      "grad_norm": 0.1523541361093521,
      "learning_rate": 4.5235685588983964e-05,
      "loss": 0.5762,
      "mean_token_accuracy": 0.8231318481266499,
      "num_tokens": 569980620.0,
      "step": 17858
    },
    {
      "entropy": 0.613241333514452,
      "epoch": 1.643124536374144,
      "grad_norm": 0.15950223803520203,
      "learning_rate": 4.523261876284234e-05,
      "loss": 0.6047,
      "mean_token_accuracy": 0.8135683573782444,
      "num_tokens": 570013070.0,
      "step": 17859
    },
    {
      "entropy": 0.678469005972147,
      "epoch": 1.6432165427455852,
      "grad_norm": 0.16752728819847107,
      "learning_rate": 4.5229551936700706e-05,
      "loss": 0.6792,
      "mean_token_accuracy": 0.7908682227134705,
      "num_tokens": 570044203.0,
      "step": 17860
    },
    {
      "entropy": 0.6572464182972908,
      "epoch": 1.6433085491170263,
      "grad_norm": 0.16700014472007751,
      "learning_rate": 4.522648511055909e-05,
      "loss": 0.6562,
      "mean_token_accuracy": 0.8023353219032288,
      "num_tokens": 570075724.0,
      "step": 17861
    },
    {
      "entropy": 0.5964886359870434,
      "epoch": 1.6434005554884674,
      "grad_norm": 0.15237076580524445,
      "learning_rate": 4.5223418284417455e-05,
      "loss": 0.5912,
      "mean_token_accuracy": 0.820186834782362,
      "num_tokens": 570107280.0,
      "step": 17862
    },
    {
      "entropy": 0.7524480279535055,
      "epoch": 1.6434925618599088,
      "grad_norm": 0.16402818262577057,
      "learning_rate": 4.522035145827584e-05,
      "loss": 0.7443,
      "mean_token_accuracy": 0.7759027741849422,
      "num_tokens": 570138789.0,
      "step": 17863
    },
    {
      "entropy": 0.5850910814478993,
      "epoch": 1.6435845682313501,
      "grad_norm": 0.15049324929714203,
      "learning_rate": 4.5217284632134205e-05,
      "loss": 0.567,
      "mean_token_accuracy": 0.8246501572430134,
      "num_tokens": 570171265.0,
      "step": 17864
    },
    {
      "entropy": 0.6356695769354701,
      "epoch": 1.6436765746027913,
      "grad_norm": 0.15345360338687897,
      "learning_rate": 4.521421780599258e-05,
      "loss": 0.6197,
      "mean_token_accuracy": 0.8107804171741009,
      "num_tokens": 570202730.0,
      "step": 17865
    },
    {
      "entropy": 0.6925992295145988,
      "epoch": 1.6437685809742324,
      "grad_norm": 0.16316944360733032,
      "learning_rate": 4.5211150979850954e-05,
      "loss": 0.6686,
      "mean_token_accuracy": 0.8007317297160625,
      "num_tokens": 570233936.0,
      "step": 17866
    },
    {
      "entropy": 0.6398725565522909,
      "epoch": 1.6438605873456735,
      "grad_norm": 0.16377510130405426,
      "learning_rate": 4.520808415370933e-05,
      "loss": 0.6439,
      "mean_token_accuracy": 0.80451450496912,
      "num_tokens": 570266501.0,
      "step": 17867
    },
    {
      "entropy": 0.5215266700834036,
      "epoch": 1.6439525937171149,
      "grad_norm": 0.1456378698348999,
      "learning_rate": 4.52050173275677e-05,
      "loss": 0.5093,
      "mean_token_accuracy": 0.8425530530512333,
      "num_tokens": 570298486.0,
      "step": 17868
    },
    {
      "entropy": 0.5178817724809051,
      "epoch": 1.6440446000885562,
      "grad_norm": 0.1510242372751236,
      "learning_rate": 4.520195050142608e-05,
      "loss": 0.5125,
      "mean_token_accuracy": 0.8411091677844524,
      "num_tokens": 570330822.0,
      "step": 17869
    },
    {
      "entropy": 0.5674204477109015,
      "epoch": 1.6441366064599974,
      "grad_norm": 0.153060644865036,
      "learning_rate": 4.5198883675284446e-05,
      "loss": 0.5473,
      "mean_token_accuracy": 0.8281413652002811,
      "num_tokens": 570362151.0,
      "step": 17870
    },
    {
      "entropy": 0.6105600595474243,
      "epoch": 1.6442286128314385,
      "grad_norm": 0.14976969361305237,
      "learning_rate": 4.519581684914283e-05,
      "loss": 0.6192,
      "mean_token_accuracy": 0.809936810284853,
      "num_tokens": 570394567.0,
      "step": 17871
    },
    {
      "entropy": 0.6190236741676927,
      "epoch": 1.6443206192028796,
      "grad_norm": 0.1542583703994751,
      "learning_rate": 4.5192750023001195e-05,
      "loss": 0.6221,
      "mean_token_accuracy": 0.8077370896935463,
      "num_tokens": 570425696.0,
      "step": 17872
    },
    {
      "entropy": 0.6506138797849417,
      "epoch": 1.644412625574321,
      "grad_norm": 0.15763388574123383,
      "learning_rate": 4.518968319685957e-05,
      "loss": 0.6489,
      "mean_token_accuracy": 0.8017415106296539,
      "num_tokens": 570457736.0,
      "step": 17873
    },
    {
      "entropy": 0.6072725486010313,
      "epoch": 1.6445046319457624,
      "grad_norm": 0.1508648693561554,
      "learning_rate": 4.5186616370717944e-05,
      "loss": 0.6046,
      "mean_token_accuracy": 0.8137628808617592,
      "num_tokens": 570489729.0,
      "step": 17874
    },
    {
      "entropy": 0.5283399447798729,
      "epoch": 1.6445966383172035,
      "grad_norm": 0.14823231101036072,
      "learning_rate": 4.518354954457632e-05,
      "loss": 0.5088,
      "mean_token_accuracy": 0.840742152184248,
      "num_tokens": 570521136.0,
      "step": 17875
    },
    {
      "entropy": 0.6786608342081308,
      "epoch": 1.6446886446886446,
      "grad_norm": 0.1568453162908554,
      "learning_rate": 4.5180482718434694e-05,
      "loss": 0.6582,
      "mean_token_accuracy": 0.7972679324448109,
      "num_tokens": 570553069.0,
      "step": 17876
    },
    {
      "entropy": 0.49262914061546326,
      "epoch": 1.6447806510600858,
      "grad_norm": 0.14923225343227386,
      "learning_rate": 4.517741589229307e-05,
      "loss": 0.4799,
      "mean_token_accuracy": 0.8505535013973713,
      "num_tokens": 570584588.0,
      "step": 17877
    },
    {
      "entropy": 0.6844069212675095,
      "epoch": 1.644872657431527,
      "grad_norm": 0.15951630473136902,
      "learning_rate": 4.517434906615144e-05,
      "loss": 0.6941,
      "mean_token_accuracy": 0.790986955165863,
      "num_tokens": 570615981.0,
      "step": 17878
    },
    {
      "entropy": 0.6833724174648523,
      "epoch": 1.6449646638029685,
      "grad_norm": 0.1626068502664566,
      "learning_rate": 4.517128224000982e-05,
      "loss": 0.6734,
      "mean_token_accuracy": 0.7942856103181839,
      "num_tokens": 570647159.0,
      "step": 17879
    },
    {
      "entropy": 0.6514050243422389,
      "epoch": 1.6450566701744096,
      "grad_norm": 0.1631929725408554,
      "learning_rate": 4.516821541386819e-05,
      "loss": 0.6406,
      "mean_token_accuracy": 0.8007881231606007,
      "num_tokens": 570678839.0,
      "step": 17880
    },
    {
      "entropy": 0.6607302958145738,
      "epoch": 1.6451486765458507,
      "grad_norm": 0.15502241253852844,
      "learning_rate": 4.516514858772656e-05,
      "loss": 0.6516,
      "mean_token_accuracy": 0.7995791174471378,
      "num_tokens": 570711495.0,
      "step": 17881
    },
    {
      "entropy": 0.5910527147352695,
      "epoch": 1.6452406829172919,
      "grad_norm": 0.1578790694475174,
      "learning_rate": 4.516208176158494e-05,
      "loss": 0.5829,
      "mean_token_accuracy": 0.8215505629777908,
      "num_tokens": 570743714.0,
      "step": 17882
    },
    {
      "entropy": 0.7116758320480585,
      "epoch": 1.6453326892887332,
      "grad_norm": 0.15590164065361023,
      "learning_rate": 4.515901493544331e-05,
      "loss": 0.6957,
      "mean_token_accuracy": 0.7865889817476273,
      "num_tokens": 570775734.0,
      "step": 17883
    },
    {
      "entropy": 0.6088232882320881,
      "epoch": 1.6454246956601746,
      "grad_norm": 0.15390849113464355,
      "learning_rate": 4.515594810930169e-05,
      "loss": 0.5983,
      "mean_token_accuracy": 0.8193558864295483,
      "num_tokens": 570808169.0,
      "step": 17884
    },
    {
      "entropy": 0.5560739925131202,
      "epoch": 1.6455167020316157,
      "grad_norm": 0.16144108772277832,
      "learning_rate": 4.515288128316006e-05,
      "loss": 0.5512,
      "mean_token_accuracy": 0.8289951346814632,
      "num_tokens": 570840535.0,
      "step": 17885
    },
    {
      "entropy": 0.6786738922819495,
      "epoch": 1.6456087084030568,
      "grad_norm": 0.16610777378082275,
      "learning_rate": 4.5149814457018433e-05,
      "loss": 0.673,
      "mean_token_accuracy": 0.7946116290986538,
      "num_tokens": 570872366.0,
      "step": 17886
    },
    {
      "entropy": 0.641319178044796,
      "epoch": 1.645700714774498,
      "grad_norm": 0.1575978547334671,
      "learning_rate": 4.514674763087681e-05,
      "loss": 0.633,
      "mean_token_accuracy": 0.8057567365467548,
      "num_tokens": 570904208.0,
      "step": 17887
    },
    {
      "entropy": 0.5947358813136816,
      "epoch": 1.6457927211459393,
      "grad_norm": 0.1530696153640747,
      "learning_rate": 4.514368080473518e-05,
      "loss": 0.5813,
      "mean_token_accuracy": 0.8204573839902878,
      "num_tokens": 570935419.0,
      "step": 17888
    },
    {
      "entropy": 0.6447816854342818,
      "epoch": 1.6458847275173807,
      "grad_norm": 0.1556701809167862,
      "learning_rate": 4.514061397859356e-05,
      "loss": 0.6287,
      "mean_token_accuracy": 0.807439174503088,
      "num_tokens": 570967183.0,
      "step": 17889
    },
    {
      "entropy": 0.6124552777037024,
      "epoch": 1.6459767338888218,
      "grad_norm": 0.16085317730903625,
      "learning_rate": 4.513754715245193e-05,
      "loss": 0.6011,
      "mean_token_accuracy": 0.8110159374773502,
      "num_tokens": 570998332.0,
      "step": 17890
    },
    {
      "entropy": 0.5917114848271012,
      "epoch": 1.646068740260263,
      "grad_norm": 0.1518530696630478,
      "learning_rate": 4.51344803263103e-05,
      "loss": 0.5828,
      "mean_token_accuracy": 0.8218094892799854,
      "num_tokens": 571029794.0,
      "step": 17891
    },
    {
      "entropy": 0.7195227518677711,
      "epoch": 1.646160746631704,
      "grad_norm": 0.16349229216575623,
      "learning_rate": 4.513141350016868e-05,
      "loss": 0.7122,
      "mean_token_accuracy": 0.7854994833469391,
      "num_tokens": 571062171.0,
      "step": 17892
    },
    {
      "entropy": 0.5756879635155201,
      "epoch": 1.6462527530031454,
      "grad_norm": 0.1509205400943756,
      "learning_rate": 4.512834667402705e-05,
      "loss": 0.5618,
      "mean_token_accuracy": 0.8208430856466293,
      "num_tokens": 571094606.0,
      "step": 17893
    },
    {
      "entropy": 0.6696104230359197,
      "epoch": 1.6463447593745868,
      "grad_norm": 0.16188673675060272,
      "learning_rate": 4.5125279847885424e-05,
      "loss": 0.6788,
      "mean_token_accuracy": 0.795146569609642,
      "num_tokens": 571126431.0,
      "step": 17894
    },
    {
      "entropy": 0.6603629682213068,
      "epoch": 1.646436765746028,
      "grad_norm": 0.1564675271511078,
      "learning_rate": 4.51222130217438e-05,
      "loss": 0.652,
      "mean_token_accuracy": 0.7997394390404224,
      "num_tokens": 571158056.0,
      "step": 17895
    },
    {
      "entropy": 0.5971706607379019,
      "epoch": 1.646528772117469,
      "grad_norm": 0.15542419254779816,
      "learning_rate": 4.511914619560217e-05,
      "loss": 0.5839,
      "mean_token_accuracy": 0.8215497806668282,
      "num_tokens": 571190031.0,
      "step": 17896
    },
    {
      "entropy": 0.5941147543489933,
      "epoch": 1.6466207784889102,
      "grad_norm": 0.14954262971878052,
      "learning_rate": 4.511607936946055e-05,
      "loss": 0.5864,
      "mean_token_accuracy": 0.8192750066518784,
      "num_tokens": 571222688.0,
      "step": 17897
    },
    {
      "entropy": 0.6410716800019145,
      "epoch": 1.6467127848603516,
      "grad_norm": 0.15856793522834778,
      "learning_rate": 4.511301254331892e-05,
      "loss": 0.6307,
      "mean_token_accuracy": 0.8082916997373104,
      "num_tokens": 571254389.0,
      "step": 17898
    },
    {
      "entropy": 0.6114595499821007,
      "epoch": 1.646804791231793,
      "grad_norm": 0.16453418135643005,
      "learning_rate": 4.510994571717729e-05,
      "loss": 0.6079,
      "mean_token_accuracy": 0.8102443255484104,
      "num_tokens": 571286459.0,
      "step": 17899
    },
    {
      "entropy": 0.5522905979305506,
      "epoch": 1.646896797603234,
      "grad_norm": 0.1583118885755539,
      "learning_rate": 4.510687889103567e-05,
      "loss": 0.5458,
      "mean_token_accuracy": 0.8360114693641663,
      "num_tokens": 571318680.0,
      "step": 17900
    },
    {
      "entropy": 0.5936418073251843,
      "epoch": 1.6469888039746752,
      "grad_norm": 0.15510088205337524,
      "learning_rate": 4.510381206489404e-05,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8192081451416016,
      "num_tokens": 571351239.0,
      "step": 17901
    },
    {
      "entropy": 0.4967019958421588,
      "epoch": 1.6470808103461163,
      "grad_norm": 0.1585046350955963,
      "learning_rate": 4.5100745238752414e-05,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8483894541859627,
      "num_tokens": 571383400.0,
      "step": 17902
    },
    {
      "entropy": 0.6673358958214521,
      "epoch": 1.6471728167175577,
      "grad_norm": 0.1619929075241089,
      "learning_rate": 4.509767841261079e-05,
      "loss": 0.6442,
      "mean_token_accuracy": 0.8062969371676445,
      "num_tokens": 571415841.0,
      "step": 17903
    },
    {
      "entropy": 0.6109105497598648,
      "epoch": 1.647264823088999,
      "grad_norm": 0.1560165137052536,
      "learning_rate": 4.5094611586469164e-05,
      "loss": 0.6026,
      "mean_token_accuracy": 0.8165610507130623,
      "num_tokens": 571448167.0,
      "step": 17904
    },
    {
      "entropy": 0.4995802892372012,
      "epoch": 1.6473568294604402,
      "grad_norm": 0.1397603154182434,
      "learning_rate": 4.509154476032754e-05,
      "loss": 0.4885,
      "mean_token_accuracy": 0.8481671884655952,
      "num_tokens": 571479994.0,
      "step": 17905
    },
    {
      "entropy": 0.6155916098505259,
      "epoch": 1.6474488358318813,
      "grad_norm": 0.15606985986232758,
      "learning_rate": 4.508847793418591e-05,
      "loss": 0.6037,
      "mean_token_accuracy": 0.8133266642689705,
      "num_tokens": 571512360.0,
      "step": 17906
    },
    {
      "entropy": 0.6120213335379958,
      "epoch": 1.6475408422033224,
      "grad_norm": 0.15245914459228516,
      "learning_rate": 4.508541110804429e-05,
      "loss": 0.5949,
      "mean_token_accuracy": 0.8133569471538067,
      "num_tokens": 571543938.0,
      "step": 17907
    },
    {
      "entropy": 0.631780203897506,
      "epoch": 1.6476328485747638,
      "grad_norm": 0.15848328173160553,
      "learning_rate": 4.508234428190266e-05,
      "loss": 0.6176,
      "mean_token_accuracy": 0.8094696961343288,
      "num_tokens": 571576087.0,
      "step": 17908
    },
    {
      "entropy": 0.6255066953599453,
      "epoch": 1.6477248549462051,
      "grad_norm": 0.1560261845588684,
      "learning_rate": 4.507927745576104e-05,
      "loss": 0.6282,
      "mean_token_accuracy": 0.8028248511254787,
      "num_tokens": 571608611.0,
      "step": 17909
    },
    {
      "entropy": 0.48978146398440003,
      "epoch": 1.6478168613176463,
      "grad_norm": 0.14830604195594788,
      "learning_rate": 4.507621062961941e-05,
      "loss": 0.4908,
      "mean_token_accuracy": 0.8465211801230907,
      "num_tokens": 571640635.0,
      "step": 17910
    },
    {
      "entropy": 0.65287702716887,
      "epoch": 1.6479088676890874,
      "grad_norm": 0.1571444571018219,
      "learning_rate": 4.5073143803477786e-05,
      "loss": 0.6482,
      "mean_token_accuracy": 0.8002186082303524,
      "num_tokens": 571673079.0,
      "step": 17911
    },
    {
      "entropy": 0.6622057678177953,
      "epoch": 1.6480008740605285,
      "grad_norm": 0.16660015285015106,
      "learning_rate": 4.5070076977336154e-05,
      "loss": 0.6461,
      "mean_token_accuracy": 0.8034160286188126,
      "num_tokens": 571704488.0,
      "step": 17912
    },
    {
      "entropy": 0.5586019195616245,
      "epoch": 1.6480928804319699,
      "grad_norm": 0.15475672483444214,
      "learning_rate": 4.5067010151194535e-05,
      "loss": 0.5375,
      "mean_token_accuracy": 0.8297803103923798,
      "num_tokens": 571736376.0,
      "step": 17913
    },
    {
      "entropy": 0.7445345167070627,
      "epoch": 1.6481848868034112,
      "grad_norm": 0.1623125821352005,
      "learning_rate": 4.50639433250529e-05,
      "loss": 0.7381,
      "mean_token_accuracy": 0.7769242003560066,
      "num_tokens": 571768283.0,
      "step": 17914
    },
    {
      "entropy": 0.5208107400685549,
      "epoch": 1.6482768931748524,
      "grad_norm": 0.14317314326763153,
      "learning_rate": 4.506087649891128e-05,
      "loss": 0.508,
      "mean_token_accuracy": 0.8404495604336262,
      "num_tokens": 571800524.0,
      "step": 17915
    },
    {
      "entropy": 0.49805828323587775,
      "epoch": 1.6483688995462935,
      "grad_norm": 0.14978651702404022,
      "learning_rate": 4.505780967276965e-05,
      "loss": 0.4815,
      "mean_token_accuracy": 0.8501144275069237,
      "num_tokens": 571832047.0,
      "step": 17916
    },
    {
      "entropy": 0.5721400054171681,
      "epoch": 1.6484609059177346,
      "grad_norm": 0.14761470258235931,
      "learning_rate": 4.505474284662803e-05,
      "loss": 0.548,
      "mean_token_accuracy": 0.8338770307600498,
      "num_tokens": 571864220.0,
      "step": 17917
    },
    {
      "entropy": 0.6279718838632107,
      "epoch": 1.648552912289176,
      "grad_norm": 0.1509236842393875,
      "learning_rate": 4.50516760204864e-05,
      "loss": 0.6038,
      "mean_token_accuracy": 0.8157816715538502,
      "num_tokens": 571896978.0,
      "step": 17918
    },
    {
      "entropy": 0.581975007429719,
      "epoch": 1.6486449186606174,
      "grad_norm": 0.1525518149137497,
      "learning_rate": 4.5048609194344777e-05,
      "loss": 0.5652,
      "mean_token_accuracy": 0.8235763870179653,
      "num_tokens": 571928617.0,
      "step": 17919
    },
    {
      "entropy": 0.5830812137573957,
      "epoch": 1.6487369250320585,
      "grad_norm": 0.1506119817495346,
      "learning_rate": 4.5045542368203144e-05,
      "loss": 0.5655,
      "mean_token_accuracy": 0.8245438188314438,
      "num_tokens": 571960981.0,
      "step": 17920
    },
    {
      "entropy": 0.5028396220877767,
      "epoch": 1.6488289314034996,
      "grad_norm": 0.14063243567943573,
      "learning_rate": 4.5042475542061526e-05,
      "loss": 0.4938,
      "mean_token_accuracy": 0.8468637764453888,
      "num_tokens": 571993453.0,
      "step": 17921
    },
    {
      "entropy": 0.5146072553470731,
      "epoch": 1.6489209377749408,
      "grad_norm": 0.1486063301563263,
      "learning_rate": 4.5039408715919894e-05,
      "loss": 0.5119,
      "mean_token_accuracy": 0.8356643579900265,
      "num_tokens": 572026038.0,
      "step": 17922
    },
    {
      "entropy": 0.5511841382831335,
      "epoch": 1.649012944146382,
      "grad_norm": 0.1509329229593277,
      "learning_rate": 4.5036341889778275e-05,
      "loss": 0.5513,
      "mean_token_accuracy": 0.8304036222398281,
      "num_tokens": 572058544.0,
      "step": 17923
    },
    {
      "entropy": 0.46818089368753135,
      "epoch": 1.6491049505178235,
      "grad_norm": 0.14353784918785095,
      "learning_rate": 4.503327506363664e-05,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8568013869225979,
      "num_tokens": 572090582.0,
      "step": 17924
    },
    {
      "entropy": 0.5947971818968654,
      "epoch": 1.6491969568892646,
      "grad_norm": 0.14959776401519775,
      "learning_rate": 4.503020823749502e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.8256024196743965,
      "num_tokens": 572121888.0,
      "step": 17925
    },
    {
      "entropy": 0.5987301953136921,
      "epoch": 1.6492889632607057,
      "grad_norm": 0.15666739642620087,
      "learning_rate": 4.502714141135339e-05,
      "loss": 0.5819,
      "mean_token_accuracy": 0.8246219456195831,
      "num_tokens": 572153517.0,
      "step": 17926
    },
    {
      "entropy": 0.5847441842779517,
      "epoch": 1.6493809696321469,
      "grad_norm": 0.1492239236831665,
      "learning_rate": 4.502407458521177e-05,
      "loss": 0.5706,
      "mean_token_accuracy": 0.8269481249153614,
      "num_tokens": 572185846.0,
      "step": 17927
    },
    {
      "entropy": 0.6503035770729184,
      "epoch": 1.6494729760035882,
      "grad_norm": 0.16670559346675873,
      "learning_rate": 4.502100775907014e-05,
      "loss": 0.6354,
      "mean_token_accuracy": 0.8093233443796635,
      "num_tokens": 572217680.0,
      "step": 17928
    },
    {
      "entropy": 0.6403976092115045,
      "epoch": 1.6495649823750296,
      "grad_norm": 0.15450188517570496,
      "learning_rate": 4.5017940932928516e-05,
      "loss": 0.6283,
      "mean_token_accuracy": 0.8067164346575737,
      "num_tokens": 572249520.0,
      "step": 17929
    },
    {
      "entropy": 0.5971850426867604,
      "epoch": 1.6496569887464707,
      "grad_norm": 0.15912771224975586,
      "learning_rate": 4.5014874106786884e-05,
      "loss": 0.6055,
      "mean_token_accuracy": 0.8185030780732632,
      "num_tokens": 572281082.0,
      "step": 17930
    },
    {
      "entropy": 0.6538698058575392,
      "epoch": 1.6497489951179118,
      "grad_norm": 0.15481065213680267,
      "learning_rate": 4.5011807280645266e-05,
      "loss": 0.6541,
      "mean_token_accuracy": 0.8004724904894829,
      "num_tokens": 572313449.0,
      "step": 17931
    },
    {
      "entropy": 0.5828907378017902,
      "epoch": 1.649841001489353,
      "grad_norm": 0.14938010275363922,
      "learning_rate": 4.5008740454503633e-05,
      "loss": 0.5816,
      "mean_token_accuracy": 0.8223481848835945,
      "num_tokens": 572345568.0,
      "step": 17932
    },
    {
      "entropy": 0.5496855080127716,
      "epoch": 1.6499330078607943,
      "grad_norm": 0.15349604189395905,
      "learning_rate": 4.500567362836201e-05,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8314433209598064,
      "num_tokens": 572377665.0,
      "step": 17933
    },
    {
      "entropy": 0.5748599916696548,
      "epoch": 1.6500250142322357,
      "grad_norm": 0.154911071062088,
      "learning_rate": 4.500260680222038e-05,
      "loss": 0.5749,
      "mean_token_accuracy": 0.8239085003733635,
      "num_tokens": 572409960.0,
      "step": 17934
    },
    {
      "entropy": 0.6304971016943455,
      "epoch": 1.6501170206036768,
      "grad_norm": 0.15601757168769836,
      "learning_rate": 4.499953997607876e-05,
      "loss": 0.6221,
      "mean_token_accuracy": 0.8088325783610344,
      "num_tokens": 572441679.0,
      "step": 17935
    },
    {
      "entropy": 0.5123607669956982,
      "epoch": 1.650209026975118,
      "grad_norm": 0.15075141191482544,
      "learning_rate": 4.499647314993713e-05,
      "loss": 0.5018,
      "mean_token_accuracy": 0.8423372954130173,
      "num_tokens": 572473240.0,
      "step": 17936
    },
    {
      "entropy": 0.6465067807585001,
      "epoch": 1.650301033346559,
      "grad_norm": 0.16164717078208923,
      "learning_rate": 4.499340632379551e-05,
      "loss": 0.6392,
      "mean_token_accuracy": 0.8033463545143604,
      "num_tokens": 572505083.0,
      "step": 17937
    },
    {
      "entropy": 0.5703146941959858,
      "epoch": 1.6503930397180004,
      "grad_norm": 0.15419892966747284,
      "learning_rate": 4.499033949765388e-05,
      "loss": 0.5679,
      "mean_token_accuracy": 0.826228853315115,
      "num_tokens": 572536792.0,
      "step": 17938
    },
    {
      "entropy": 0.7191582471132278,
      "epoch": 1.6504850460894418,
      "grad_norm": 0.15894915163516998,
      "learning_rate": 4.4987272671512256e-05,
      "loss": 0.7115,
      "mean_token_accuracy": 0.7806838750839233,
      "num_tokens": 572568956.0,
      "step": 17939
    },
    {
      "entropy": 0.4958818058948964,
      "epoch": 1.650577052460883,
      "grad_norm": 0.14697006344795227,
      "learning_rate": 4.498420584537063e-05,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8538658693432808,
      "num_tokens": 572599993.0,
      "step": 17940
    },
    {
      "entropy": 0.6889557642862201,
      "epoch": 1.650669058832324,
      "grad_norm": 0.15561328828334808,
      "learning_rate": 4.4981139019229e-05,
      "loss": 0.6882,
      "mean_token_accuracy": 0.7915941849350929,
      "num_tokens": 572631960.0,
      "step": 17941
    },
    {
      "entropy": 0.6502674510702491,
      "epoch": 1.6507610652037652,
      "grad_norm": 0.15920181572437286,
      "learning_rate": 4.497807219308738e-05,
      "loss": 0.6541,
      "mean_token_accuracy": 0.8002465143799782,
      "num_tokens": 572663368.0,
      "step": 17942
    },
    {
      "entropy": 0.5875687478110194,
      "epoch": 1.6508530715752066,
      "grad_norm": 0.15246620774269104,
      "learning_rate": 4.497500536694575e-05,
      "loss": 0.5662,
      "mean_token_accuracy": 0.823261272162199,
      "num_tokens": 572694860.0,
      "step": 17943
    },
    {
      "entropy": 0.6048404760658741,
      "epoch": 1.650945077946648,
      "grad_norm": 0.1531287431716919,
      "learning_rate": 4.497193854080413e-05,
      "loss": 0.5822,
      "mean_token_accuracy": 0.8198568969964981,
      "num_tokens": 572726550.0,
      "step": 17944
    },
    {
      "entropy": 0.6385020408779383,
      "epoch": 1.651037084318089,
      "grad_norm": 0.15276934206485748,
      "learning_rate": 4.49688717146625e-05,
      "loss": 0.6132,
      "mean_token_accuracy": 0.8068753108382225,
      "num_tokens": 572758313.0,
      "step": 17945
    },
    {
      "entropy": 0.7262556087225676,
      "epoch": 1.6511290906895302,
      "grad_norm": 0.16104170680046082,
      "learning_rate": 4.496580488852087e-05,
      "loss": 0.7222,
      "mean_token_accuracy": 0.783306036144495,
      "num_tokens": 572790636.0,
      "step": 17946
    },
    {
      "entropy": 0.6293928623199463,
      "epoch": 1.6512210970609713,
      "grad_norm": 0.16203029453754425,
      "learning_rate": 4.4962738062379246e-05,
      "loss": 0.6387,
      "mean_token_accuracy": 0.8080032132565975,
      "num_tokens": 572822861.0,
      "step": 17947
    },
    {
      "entropy": 0.6423506774008274,
      "epoch": 1.6513131034324127,
      "grad_norm": 0.1585145890712738,
      "learning_rate": 4.495967123623762e-05,
      "loss": 0.6233,
      "mean_token_accuracy": 0.8082794919610023,
      "num_tokens": 572853993.0,
      "step": 17948
    },
    {
      "entropy": 0.45637337025254965,
      "epoch": 1.651405109803854,
      "grad_norm": 0.13575130701065063,
      "learning_rate": 4.4956604410095996e-05,
      "loss": 0.4446,
      "mean_token_accuracy": 0.8612548671662807,
      "num_tokens": 572886308.0,
      "step": 17949
    },
    {
      "entropy": 0.5935531794093549,
      "epoch": 1.6514971161752952,
      "grad_norm": 0.152655690908432,
      "learning_rate": 4.495353758395437e-05,
      "loss": 0.5846,
      "mean_token_accuracy": 0.8237752541899681,
      "num_tokens": 572918568.0,
      "step": 17950
    },
    {
      "entropy": 0.4706328413449228,
      "epoch": 1.6515891225467363,
      "grad_norm": 0.14571931958198547,
      "learning_rate": 4.495047075781274e-05,
      "loss": 0.4479,
      "mean_token_accuracy": 0.8588506691157818,
      "num_tokens": 572950529.0,
      "step": 17951
    },
    {
      "entropy": 0.6005186447873712,
      "epoch": 1.6516811289181774,
      "grad_norm": 0.14590874314308167,
      "learning_rate": 4.494740393167112e-05,
      "loss": 0.5843,
      "mean_token_accuracy": 0.8197913430631161,
      "num_tokens": 572983228.0,
      "step": 17952
    },
    {
      "entropy": 0.622830780223012,
      "epoch": 1.6517731352896188,
      "grad_norm": 0.14746056497097015,
      "learning_rate": 4.494433710552949e-05,
      "loss": 0.6136,
      "mean_token_accuracy": 0.8137329034507275,
      "num_tokens": 573015774.0,
      "step": 17953
    },
    {
      "entropy": 0.512634998653084,
      "epoch": 1.6518651416610601,
      "grad_norm": 0.1500527411699295,
      "learning_rate": 4.494127027938786e-05,
      "loss": 0.4941,
      "mean_token_accuracy": 0.8432161100208759,
      "num_tokens": 573046582.0,
      "step": 17954
    },
    {
      "entropy": 0.5970009570010006,
      "epoch": 1.6519571480325013,
      "grad_norm": 0.15342681109905243,
      "learning_rate": 4.493820345324624e-05,
      "loss": 0.5771,
      "mean_token_accuracy": 0.8214581981301308,
      "num_tokens": 573078751.0,
      "step": 17955
    },
    {
      "entropy": 0.5701214801520109,
      "epoch": 1.6520491544039424,
      "grad_norm": 0.15433737635612488,
      "learning_rate": 4.493513662710461e-05,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8285452164709568,
      "num_tokens": 573110583.0,
      "step": 17956
    },
    {
      "entropy": 0.7367727514356375,
      "epoch": 1.6521411607753835,
      "grad_norm": 0.16393452882766724,
      "learning_rate": 4.4932069800962986e-05,
      "loss": 0.7159,
      "mean_token_accuracy": 0.7831017561256886,
      "num_tokens": 573141709.0,
      "step": 17957
    },
    {
      "entropy": 0.6155882049351931,
      "epoch": 1.6522331671468249,
      "grad_norm": 0.14887912571430206,
      "learning_rate": 4.492900297482136e-05,
      "loss": 0.6003,
      "mean_token_accuracy": 0.8135052993893623,
      "num_tokens": 573173840.0,
      "step": 17958
    },
    {
      "entropy": 0.597988598048687,
      "epoch": 1.6523251735182662,
      "grad_norm": 0.15433254837989807,
      "learning_rate": 4.492593614867973e-05,
      "loss": 0.6017,
      "mean_token_accuracy": 0.8182274103164673,
      "num_tokens": 573204642.0,
      "step": 17959
    },
    {
      "entropy": 0.5814739125780761,
      "epoch": 1.6524171798897074,
      "grad_norm": 0.15689541399478912,
      "learning_rate": 4.492286932253811e-05,
      "loss": 0.5753,
      "mean_token_accuracy": 0.8237165659666061,
      "num_tokens": 573236727.0,
      "step": 17960
    },
    {
      "entropy": 0.6298194415867329,
      "epoch": 1.6525091862611485,
      "grad_norm": 0.15979748964309692,
      "learning_rate": 4.491980249639648e-05,
      "loss": 0.6152,
      "mean_token_accuracy": 0.8102431073784828,
      "num_tokens": 573268024.0,
      "step": 17961
    },
    {
      "entropy": 0.6467712987214327,
      "epoch": 1.6526011926325896,
      "grad_norm": 0.16535049676895142,
      "learning_rate": 4.491673567025485e-05,
      "loss": 0.6349,
      "mean_token_accuracy": 0.8037067502737045,
      "num_tokens": 573298851.0,
      "step": 17962
    },
    {
      "entropy": 0.6995880715548992,
      "epoch": 1.652693199004031,
      "grad_norm": 0.167426198720932,
      "learning_rate": 4.491366884411323e-05,
      "loss": 0.7048,
      "mean_token_accuracy": 0.7859979942440987,
      "num_tokens": 573330164.0,
      "step": 17963
    },
    {
      "entropy": 0.6035536425188184,
      "epoch": 1.6527852053754724,
      "grad_norm": 0.1630251407623291,
      "learning_rate": 4.49106020179716e-05,
      "loss": 0.5979,
      "mean_token_accuracy": 0.8176639042794704,
      "num_tokens": 573362683.0,
      "step": 17964
    },
    {
      "entropy": 0.6391331730410457,
      "epoch": 1.6528772117469135,
      "grad_norm": 0.15878717601299286,
      "learning_rate": 4.4907535191829976e-05,
      "loss": 0.629,
      "mean_token_accuracy": 0.8106228411197662,
      "num_tokens": 573394621.0,
      "step": 17965
    },
    {
      "entropy": 0.6681746151298285,
      "epoch": 1.6529692181183546,
      "grad_norm": 0.15667904913425446,
      "learning_rate": 4.490446836568835e-05,
      "loss": 0.6539,
      "mean_token_accuracy": 0.7964588664472103,
      "num_tokens": 573426122.0,
      "step": 17966
    },
    {
      "entropy": 0.5075780584011227,
      "epoch": 1.6530612244897958,
      "grad_norm": 0.14547409117221832,
      "learning_rate": 4.4901401539546726e-05,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8474978134036064,
      "num_tokens": 573458104.0,
      "step": 17967
    },
    {
      "entropy": 0.6265882831066847,
      "epoch": 1.653153230861237,
      "grad_norm": 0.15780679881572723,
      "learning_rate": 4.48983347134051e-05,
      "loss": 0.6219,
      "mean_token_accuracy": 0.8102605380117893,
      "num_tokens": 573489965.0,
      "step": 17968
    },
    {
      "entropy": 0.6086218990385532,
      "epoch": 1.6532452372326785,
      "grad_norm": 0.16261234879493713,
      "learning_rate": 4.4895267887263475e-05,
      "loss": 0.5917,
      "mean_token_accuracy": 0.8172046914696693,
      "num_tokens": 573521836.0,
      "step": 17969
    },
    {
      "entropy": 0.6241317950189114,
      "epoch": 1.6533372436041196,
      "grad_norm": 0.15095190703868866,
      "learning_rate": 4.489220106112185e-05,
      "loss": 0.6129,
      "mean_token_accuracy": 0.8123645037412643,
      "num_tokens": 573553593.0,
      "step": 17970
    },
    {
      "entropy": 0.6366913188248873,
      "epoch": 1.6534292499755607,
      "grad_norm": 0.15910027921199799,
      "learning_rate": 4.4889134234980224e-05,
      "loss": 0.6401,
      "mean_token_accuracy": 0.8030433356761932,
      "num_tokens": 573585476.0,
      "step": 17971
    },
    {
      "entropy": 0.5618000440299511,
      "epoch": 1.6535212563470019,
      "grad_norm": 0.15674936771392822,
      "learning_rate": 4.488606740883859e-05,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8255190886557102,
      "num_tokens": 573617703.0,
      "step": 17972
    },
    {
      "entropy": 0.6311264140531421,
      "epoch": 1.6536132627184432,
      "grad_norm": 0.1619318127632141,
      "learning_rate": 4.4883000582696974e-05,
      "loss": 0.6249,
      "mean_token_accuracy": 0.8115505166351795,
      "num_tokens": 573649560.0,
      "step": 17973
    },
    {
      "entropy": 0.5248315057251602,
      "epoch": 1.6537052690898846,
      "grad_norm": 0.16045624017715454,
      "learning_rate": 4.487993375655534e-05,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8437677621841431,
      "num_tokens": 573682081.0,
      "step": 17974
    },
    {
      "entropy": 0.5993173439055681,
      "epoch": 1.6537972754613257,
      "grad_norm": 0.16328375041484833,
      "learning_rate": 4.4876866930413716e-05,
      "loss": 0.6072,
      "mean_token_accuracy": 0.8186302036046982,
      "num_tokens": 573713881.0,
      "step": 17975
    },
    {
      "entropy": 0.6748819453641772,
      "epoch": 1.6538892818327668,
      "grad_norm": 0.15824632346630096,
      "learning_rate": 4.487380010427209e-05,
      "loss": 0.6639,
      "mean_token_accuracy": 0.7962042838335037,
      "num_tokens": 573746024.0,
      "step": 17976
    },
    {
      "entropy": 0.6088535785675049,
      "epoch": 1.653981288204208,
      "grad_norm": 0.16125556826591492,
      "learning_rate": 4.4870733278130465e-05,
      "loss": 0.5977,
      "mean_token_accuracy": 0.8158589638769627,
      "num_tokens": 573777351.0,
      "step": 17977
    },
    {
      "entropy": 0.5909535065293312,
      "epoch": 1.6540732945756493,
      "grad_norm": 0.15132007002830505,
      "learning_rate": 4.486766645198884e-05,
      "loss": 0.5807,
      "mean_token_accuracy": 0.8228648267686367,
      "num_tokens": 573809745.0,
      "step": 17978
    },
    {
      "entropy": 0.581357330083847,
      "epoch": 1.6541653009470907,
      "grad_norm": 0.15499013662338257,
      "learning_rate": 4.4864599625847215e-05,
      "loss": 0.5672,
      "mean_token_accuracy": 0.8310367651283741,
      "num_tokens": 573841563.0,
      "step": 17979
    },
    {
      "entropy": 0.7285208804532886,
      "epoch": 1.6542573073185318,
      "grad_norm": 0.1683715432882309,
      "learning_rate": 4.486153279970558e-05,
      "loss": 0.7155,
      "mean_token_accuracy": 0.7849111631512642,
      "num_tokens": 573873490.0,
      "step": 17980
    },
    {
      "entropy": 0.4626779002137482,
      "epoch": 1.654349313689973,
      "grad_norm": 0.14014169573783875,
      "learning_rate": 4.4858465973563964e-05,
      "loss": 0.4582,
      "mean_token_accuracy": 0.8576792739331722,
      "num_tokens": 573905613.0,
      "step": 17981
    },
    {
      "entropy": 0.6553596686571836,
      "epoch": 1.654441320061414,
      "grad_norm": 0.15550842881202698,
      "learning_rate": 4.485539914742233e-05,
      "loss": 0.6417,
      "mean_token_accuracy": 0.8036219291388988,
      "num_tokens": 573938037.0,
      "step": 17982
    },
    {
      "entropy": 0.6081982031464577,
      "epoch": 1.6545333264328554,
      "grad_norm": 0.15108229219913483,
      "learning_rate": 4.485233232128071e-05,
      "loss": 0.6016,
      "mean_token_accuracy": 0.8181911408901215,
      "num_tokens": 573970154.0,
      "step": 17983
    },
    {
      "entropy": 0.605467364192009,
      "epoch": 1.6546253328042968,
      "grad_norm": 0.15354520082473755,
      "learning_rate": 4.484926549513908e-05,
      "loss": 0.5917,
      "mean_token_accuracy": 0.8165123499929905,
      "num_tokens": 574002195.0,
      "step": 17984
    },
    {
      "entropy": 0.6613468565046787,
      "epoch": 1.654717339175738,
      "grad_norm": 0.16145452857017517,
      "learning_rate": 4.4846198668997456e-05,
      "loss": 0.6553,
      "mean_token_accuracy": 0.8029578104615211,
      "num_tokens": 574034224.0,
      "step": 17985
    },
    {
      "entropy": 0.6409602127969265,
      "epoch": 1.654809345547179,
      "grad_norm": 0.15660220384597778,
      "learning_rate": 4.484313184285583e-05,
      "loss": 0.6218,
      "mean_token_accuracy": 0.8044566921889782,
      "num_tokens": 574066142.0,
      "step": 17986
    },
    {
      "entropy": 0.5840492751449347,
      "epoch": 1.6549013519186202,
      "grad_norm": 0.14870376884937286,
      "learning_rate": 4.4840065016714205e-05,
      "loss": 0.5767,
      "mean_token_accuracy": 0.8179136738181114,
      "num_tokens": 574098151.0,
      "step": 17987
    },
    {
      "entropy": 0.6231877710670233,
      "epoch": 1.6549933582900616,
      "grad_norm": 0.15683306753635406,
      "learning_rate": 4.483699819057258e-05,
      "loss": 0.6147,
      "mean_token_accuracy": 0.8153428956866264,
      "num_tokens": 574130523.0,
      "step": 17988
    },
    {
      "entropy": 0.5288860891014338,
      "epoch": 1.655085364661503,
      "grad_norm": 0.14937537908554077,
      "learning_rate": 4.4833931364430954e-05,
      "loss": 0.5164,
      "mean_token_accuracy": 0.8417195864021778,
      "num_tokens": 574163291.0,
      "step": 17989
    },
    {
      "entropy": 0.5939929466694593,
      "epoch": 1.655177371032944,
      "grad_norm": 0.1521151065826416,
      "learning_rate": 4.483086453828932e-05,
      "loss": 0.5771,
      "mean_token_accuracy": 0.8204876147210598,
      "num_tokens": 574195151.0,
      "step": 17990
    },
    {
      "entropy": 0.6126599768176675,
      "epoch": 1.6552693774043852,
      "grad_norm": 0.1526506096124649,
      "learning_rate": 4.4827797712147704e-05,
      "loss": 0.592,
      "mean_token_accuracy": 0.8215601891279221,
      "num_tokens": 574226660.0,
      "step": 17991
    },
    {
      "entropy": 0.5961508238688111,
      "epoch": 1.6553613837758263,
      "grad_norm": 0.15564264357089996,
      "learning_rate": 4.482473088600607e-05,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8171588219702244,
      "num_tokens": 574257997.0,
      "step": 17992
    },
    {
      "entropy": 0.6786542721092701,
      "epoch": 1.6554533901472677,
      "grad_norm": 0.1550474762916565,
      "learning_rate": 4.4821664059864446e-05,
      "loss": 0.6674,
      "mean_token_accuracy": 0.7968449853360653,
      "num_tokens": 574290290.0,
      "step": 17993
    },
    {
      "entropy": 0.5556579548865557,
      "epoch": 1.655545396518709,
      "grad_norm": 0.14843519032001495,
      "learning_rate": 4.481859723372282e-05,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8333571441471577,
      "num_tokens": 574322392.0,
      "step": 17994
    },
    {
      "entropy": 0.5264284815639257,
      "epoch": 1.6556374028901502,
      "grad_norm": 0.1470731645822525,
      "learning_rate": 4.4815530407581196e-05,
      "loss": 0.5123,
      "mean_token_accuracy": 0.8408543653786182,
      "num_tokens": 574355050.0,
      "step": 17995
    },
    {
      "entropy": 0.6309391483664513,
      "epoch": 1.6557294092615913,
      "grad_norm": 0.155859112739563,
      "learning_rate": 4.481246358143957e-05,
      "loss": 0.6304,
      "mean_token_accuracy": 0.8053272739052773,
      "num_tokens": 574386641.0,
      "step": 17996
    },
    {
      "entropy": 0.7171065807342529,
      "epoch": 1.6558214156330324,
      "grad_norm": 0.163505420088768,
      "learning_rate": 4.4809396755297945e-05,
      "loss": 0.724,
      "mean_token_accuracy": 0.7797792814671993,
      "num_tokens": 574418253.0,
      "step": 17997
    },
    {
      "entropy": 0.5853305216878653,
      "epoch": 1.6559134220044738,
      "grad_norm": 0.15366053581237793,
      "learning_rate": 4.480632992915632e-05,
      "loss": 0.5766,
      "mean_token_accuracy": 0.8179108984768391,
      "num_tokens": 574450409.0,
      "step": 17998
    },
    {
      "entropy": 0.6691236579790711,
      "epoch": 1.6560054283759151,
      "grad_norm": 0.1576545238494873,
      "learning_rate": 4.4803263103014694e-05,
      "loss": 0.6632,
      "mean_token_accuracy": 0.7964361943304539,
      "num_tokens": 574482604.0,
      "step": 17999
    },
    {
      "entropy": 0.5664147809147835,
      "epoch": 1.6560974347473563,
      "grad_norm": 0.15642951428890228,
      "learning_rate": 4.480019627687307e-05,
      "loss": 0.5638,
      "mean_token_accuracy": 0.8253801055252552,
      "num_tokens": 574514554.0,
      "step": 18000
    },
    {
      "entropy": 0.5926323318853974,
      "epoch": 1.6561894411187974,
      "grad_norm": 0.1539933681488037,
      "learning_rate": 4.479712945073144e-05,
      "loss": 0.5852,
      "mean_token_accuracy": 0.8163289576768875,
      "num_tokens": 574546637.0,
      "step": 18001
    },
    {
      "entropy": 0.7828766703605652,
      "epoch": 1.6562814474902385,
      "grad_norm": 0.1716596931219101,
      "learning_rate": 4.479406262458982e-05,
      "loss": 0.7859,
      "mean_token_accuracy": 0.7655859664082527,
      "num_tokens": 574577724.0,
      "step": 18002
    },
    {
      "entropy": 0.6722813546657562,
      "epoch": 1.65637345386168,
      "grad_norm": 0.15629327297210693,
      "learning_rate": 4.4790995798448186e-05,
      "loss": 0.6741,
      "mean_token_accuracy": 0.7948299646377563,
      "num_tokens": 574609710.0,
      "step": 18003
    },
    {
      "entropy": 0.676087224856019,
      "epoch": 1.6564654602331212,
      "grad_norm": 0.16491396725177765,
      "learning_rate": 4.478792897230657e-05,
      "loss": 0.675,
      "mean_token_accuracy": 0.7926600277423859,
      "num_tokens": 574641085.0,
      "step": 18004
    },
    {
      "entropy": 0.5122663537040353,
      "epoch": 1.6565574666045624,
      "grad_norm": 0.14870163798332214,
      "learning_rate": 4.4784862146164935e-05,
      "loss": 0.5082,
      "mean_token_accuracy": 0.8416198119521141,
      "num_tokens": 574673431.0,
      "step": 18005
    },
    {
      "entropy": 0.6767975110560656,
      "epoch": 1.6566494729760035,
      "grad_norm": 0.15628117322921753,
      "learning_rate": 4.478179532002331e-05,
      "loss": 0.6624,
      "mean_token_accuracy": 0.8020676411688328,
      "num_tokens": 574705898.0,
      "step": 18006
    },
    {
      "entropy": 0.565249071456492,
      "epoch": 1.6567414793474446,
      "grad_norm": 0.15390987694263458,
      "learning_rate": 4.4778728493881685e-05,
      "loss": 0.5618,
      "mean_token_accuracy": 0.8299192301928997,
      "num_tokens": 574738098.0,
      "step": 18007
    },
    {
      "entropy": 0.6122033959254622,
      "epoch": 1.656833485718886,
      "grad_norm": 0.15633758902549744,
      "learning_rate": 4.477566166774006e-05,
      "loss": 0.6154,
      "mean_token_accuracy": 0.8182802796363831,
      "num_tokens": 574769950.0,
      "step": 18008
    },
    {
      "entropy": 0.5728526180610061,
      "epoch": 1.6569254920903274,
      "grad_norm": 0.15684911608695984,
      "learning_rate": 4.4772594841598434e-05,
      "loss": 0.5548,
      "mean_token_accuracy": 0.8282267674803734,
      "num_tokens": 574802535.0,
      "step": 18009
    },
    {
      "entropy": 0.6422329153865576,
      "epoch": 1.6570174984617685,
      "grad_norm": 0.1569713056087494,
      "learning_rate": 4.476952801545681e-05,
      "loss": 0.6266,
      "mean_token_accuracy": 0.8092595972120762,
      "num_tokens": 574835122.0,
      "step": 18010
    },
    {
      "entropy": 0.6445924863219261,
      "epoch": 1.6571095048332096,
      "grad_norm": 0.16070111095905304,
      "learning_rate": 4.4766461189315176e-05,
      "loss": 0.6193,
      "mean_token_accuracy": 0.8103103786706924,
      "num_tokens": 574866178.0,
      "step": 18011
    },
    {
      "entropy": 0.7145569063723087,
      "epoch": 1.6572015112046508,
      "grad_norm": 0.16655321419239044,
      "learning_rate": 4.476339436317356e-05,
      "loss": 0.6988,
      "mean_token_accuracy": 0.7878029011189938,
      "num_tokens": 574897727.0,
      "step": 18012
    },
    {
      "entropy": 0.5398356644436717,
      "epoch": 1.6572935175760921,
      "grad_norm": 0.1539754569530487,
      "learning_rate": 4.4760327537031926e-05,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8408563993871212,
      "num_tokens": 574929938.0,
      "step": 18013
    },
    {
      "entropy": 0.6541794445365667,
      "epoch": 1.6573855239475335,
      "grad_norm": 0.1646013855934143,
      "learning_rate": 4.47572607108903e-05,
      "loss": 0.638,
      "mean_token_accuracy": 0.806171890348196,
      "num_tokens": 574961686.0,
      "step": 18014
    },
    {
      "entropy": 0.5441033514216542,
      "epoch": 1.6574775303189746,
      "grad_norm": 0.1459563821554184,
      "learning_rate": 4.4754193884748675e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8302698917686939,
      "num_tokens": 574994454.0,
      "step": 18015
    },
    {
      "entropy": 0.6258156914263964,
      "epoch": 1.6575695366904157,
      "grad_norm": 0.15591910481452942,
      "learning_rate": 4.475112705860705e-05,
      "loss": 0.6234,
      "mean_token_accuracy": 0.8127529136836529,
      "num_tokens": 575026808.0,
      "step": 18016
    },
    {
      "entropy": 0.6165930917486548,
      "epoch": 1.6576615430618569,
      "grad_norm": 0.15962758660316467,
      "learning_rate": 4.4748060232465424e-05,
      "loss": 0.5938,
      "mean_token_accuracy": 0.815124474465847,
      "num_tokens": 575058717.0,
      "step": 18017
    },
    {
      "entropy": 0.6671857684850693,
      "epoch": 1.6577535494332982,
      "grad_norm": 0.15718966722488403,
      "learning_rate": 4.47449934063238e-05,
      "loss": 0.6679,
      "mean_token_accuracy": 0.7953507862985134,
      "num_tokens": 575090590.0,
      "step": 18018
    },
    {
      "entropy": 0.6622865591198206,
      "epoch": 1.6578455558047396,
      "grad_norm": 0.15619085729122162,
      "learning_rate": 4.474192658018217e-05,
      "loss": 0.6489,
      "mean_token_accuracy": 0.8016753010451794,
      "num_tokens": 575122336.0,
      "step": 18019
    },
    {
      "entropy": 0.7343048304319382,
      "epoch": 1.6579375621761807,
      "grad_norm": 0.15941095352172852,
      "learning_rate": 4.473885975404055e-05,
      "loss": 0.723,
      "mean_token_accuracy": 0.7816201969981194,
      "num_tokens": 575154328.0,
      "step": 18020
    },
    {
      "entropy": 0.5882188137620687,
      "epoch": 1.6580295685476218,
      "grad_norm": 0.15025359392166138,
      "learning_rate": 4.4735792927898916e-05,
      "loss": 0.5815,
      "mean_token_accuracy": 0.8175830505788326,
      "num_tokens": 575186576.0,
      "step": 18021
    },
    {
      "entropy": 0.47512665623798966,
      "epoch": 1.658121574919063,
      "grad_norm": 0.14296835660934448,
      "learning_rate": 4.473272610175729e-05,
      "loss": 0.4564,
      "mean_token_accuracy": 0.8563417308032513,
      "num_tokens": 575218766.0,
      "step": 18022
    },
    {
      "entropy": 0.6392529550939798,
      "epoch": 1.6582135812905043,
      "grad_norm": 0.15784098207950592,
      "learning_rate": 4.4729659275615665e-05,
      "loss": 0.6208,
      "mean_token_accuracy": 0.8089197389781475,
      "num_tokens": 575250615.0,
      "step": 18023
    },
    {
      "entropy": 0.5476682130247355,
      "epoch": 1.6583055876619457,
      "grad_norm": 0.15785594284534454,
      "learning_rate": 4.472659244947404e-05,
      "loss": 0.5456,
      "mean_token_accuracy": 0.8344421908259392,
      "num_tokens": 575283050.0,
      "step": 18024
    },
    {
      "entropy": 0.6292442437261343,
      "epoch": 1.6583975940333868,
      "grad_norm": 0.16285869479179382,
      "learning_rate": 4.4723525623332415e-05,
      "loss": 0.6335,
      "mean_token_accuracy": 0.8117702640593052,
      "num_tokens": 575314701.0,
      "step": 18025
    },
    {
      "entropy": 0.5733131300657988,
      "epoch": 1.658489600404828,
      "grad_norm": 0.1561255306005478,
      "learning_rate": 4.472045879719079e-05,
      "loss": 0.5558,
      "mean_token_accuracy": 0.8313980139791965,
      "num_tokens": 575346735.0,
      "step": 18026
    },
    {
      "entropy": 0.5387018509209156,
      "epoch": 1.658581606776269,
      "grad_norm": 0.15365168452262878,
      "learning_rate": 4.4717391971049164e-05,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8339833952486515,
      "num_tokens": 575378204.0,
      "step": 18027
    },
    {
      "entropy": 0.6024173479527235,
      "epoch": 1.6586736131477104,
      "grad_norm": 0.15534967184066772,
      "learning_rate": 4.471432514490754e-05,
      "loss": 0.6196,
      "mean_token_accuracy": 0.810537800192833,
      "num_tokens": 575410452.0,
      "step": 18028
    },
    {
      "entropy": 0.4977486040443182,
      "epoch": 1.6587656195191518,
      "grad_norm": 0.14928849041461945,
      "learning_rate": 4.471125831876591e-05,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8493347801268101,
      "num_tokens": 575442223.0,
      "step": 18029
    },
    {
      "entropy": 0.5582638066262007,
      "epoch": 1.658857625890593,
      "grad_norm": 0.15241718292236328,
      "learning_rate": 4.470819149262429e-05,
      "loss": 0.5489,
      "mean_token_accuracy": 0.828496091067791,
      "num_tokens": 575474085.0,
      "step": 18030
    },
    {
      "entropy": 0.6555322324857116,
      "epoch": 1.658949632262034,
      "grad_norm": 0.16787773370742798,
      "learning_rate": 4.470512466648266e-05,
      "loss": 0.6601,
      "mean_token_accuracy": 0.8010609857738018,
      "num_tokens": 575506385.0,
      "step": 18031
    },
    {
      "entropy": 0.5325894011184573,
      "epoch": 1.6590416386334752,
      "grad_norm": 0.15075361728668213,
      "learning_rate": 4.470205784034103e-05,
      "loss": 0.5284,
      "mean_token_accuracy": 0.8350830264389515,
      "num_tokens": 575538367.0,
      "step": 18032
    },
    {
      "entropy": 0.5542085357010365,
      "epoch": 1.6591336450049166,
      "grad_norm": 0.15934138000011444,
      "learning_rate": 4.469899101419941e-05,
      "loss": 0.54,
      "mean_token_accuracy": 0.8295574113726616,
      "num_tokens": 575570192.0,
      "step": 18033
    },
    {
      "entropy": 0.6016766121610999,
      "epoch": 1.659225651376358,
      "grad_norm": 0.1560574173927307,
      "learning_rate": 4.469592418805778e-05,
      "loss": 0.5963,
      "mean_token_accuracy": 0.8211071714758873,
      "num_tokens": 575602070.0,
      "step": 18034
    },
    {
      "entropy": 0.5408251266926527,
      "epoch": 1.659317657747799,
      "grad_norm": 0.14993391931056976,
      "learning_rate": 4.4692857361916154e-05,
      "loss": 0.5268,
      "mean_token_accuracy": 0.8376109413802624,
      "num_tokens": 575634684.0,
      "step": 18035
    },
    {
      "entropy": 0.6589452493935823,
      "epoch": 1.6594096641192402,
      "grad_norm": 0.15621958673000336,
      "learning_rate": 4.468979053577453e-05,
      "loss": 0.653,
      "mean_token_accuracy": 0.8001441285014153,
      "num_tokens": 575667003.0,
      "step": 18036
    },
    {
      "entropy": 0.593667758628726,
      "epoch": 1.6595016704906813,
      "grad_norm": 0.15137933194637299,
      "learning_rate": 4.4686723709632904e-05,
      "loss": 0.5759,
      "mean_token_accuracy": 0.8210833743214607,
      "num_tokens": 575698351.0,
      "step": 18037
    },
    {
      "entropy": 0.5403475612401962,
      "epoch": 1.6595936768621227,
      "grad_norm": 0.14760014414787292,
      "learning_rate": 4.468365688349128e-05,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8395303301513195,
      "num_tokens": 575730810.0,
      "step": 18038
    },
    {
      "entropy": 0.5782657535746694,
      "epoch": 1.659685683233564,
      "grad_norm": 0.15070763230323792,
      "learning_rate": 4.468059005734965e-05,
      "loss": 0.5744,
      "mean_token_accuracy": 0.8257276378571987,
      "num_tokens": 575762846.0,
      "step": 18039
    },
    {
      "entropy": 0.639416029676795,
      "epoch": 1.6597776896050052,
      "grad_norm": 0.15322303771972656,
      "learning_rate": 4.467752323120802e-05,
      "loss": 0.6166,
      "mean_token_accuracy": 0.8104389645159245,
      "num_tokens": 575794700.0,
      "step": 18040
    },
    {
      "entropy": 0.6674619149416685,
      "epoch": 1.6598696959764463,
      "grad_norm": 0.15668199956417084,
      "learning_rate": 4.46744564050664e-05,
      "loss": 0.6525,
      "mean_token_accuracy": 0.7993850521743298,
      "num_tokens": 575826837.0,
      "step": 18041
    },
    {
      "entropy": 0.6311318539083004,
      "epoch": 1.6599617023478876,
      "grad_norm": 0.14969155192375183,
      "learning_rate": 4.467138957892477e-05,
      "loss": 0.6201,
      "mean_token_accuracy": 0.8112843558192253,
      "num_tokens": 575859162.0,
      "step": 18042
    },
    {
      "entropy": 0.6716200290247798,
      "epoch": 1.6600537087193288,
      "grad_norm": 0.16001221537590027,
      "learning_rate": 4.466832275278315e-05,
      "loss": 0.6738,
      "mean_token_accuracy": 0.7948304712772369,
      "num_tokens": 575890905.0,
      "step": 18043
    },
    {
      "entropy": 0.6349484175443649,
      "epoch": 1.6601457150907701,
      "grad_norm": 0.1560164839029312,
      "learning_rate": 4.466525592664152e-05,
      "loss": 0.631,
      "mean_token_accuracy": 0.8057285770773888,
      "num_tokens": 575921864.0,
      "step": 18044
    },
    {
      "entropy": 0.6369542814791203,
      "epoch": 1.6602377214622113,
      "grad_norm": 0.15723901987075806,
      "learning_rate": 4.4662189100499894e-05,
      "loss": 0.6348,
      "mean_token_accuracy": 0.8078317530453205,
      "num_tokens": 575954041.0,
      "step": 18045
    },
    {
      "entropy": 0.5407876409590244,
      "epoch": 1.6603297278336524,
      "grad_norm": 0.1443590223789215,
      "learning_rate": 4.465912227435827e-05,
      "loss": 0.5305,
      "mean_token_accuracy": 0.835625272244215,
      "num_tokens": 575986308.0,
      "step": 18046
    },
    {
      "entropy": 0.6788274301216006,
      "epoch": 1.6604217342050938,
      "grad_norm": 0.1562638282775879,
      "learning_rate": 4.465605544821664e-05,
      "loss": 0.6687,
      "mean_token_accuracy": 0.7958102412521839,
      "num_tokens": 576018690.0,
      "step": 18047
    },
    {
      "entropy": 0.6338363294489682,
      "epoch": 1.660513740576535,
      "grad_norm": 0.15765050053596497,
      "learning_rate": 4.465298862207502e-05,
      "loss": 0.6311,
      "mean_token_accuracy": 0.8089170195162296,
      "num_tokens": 576051106.0,
      "step": 18048
    },
    {
      "entropy": 0.6894770171493292,
      "epoch": 1.6606057469479762,
      "grad_norm": 0.1649559736251831,
      "learning_rate": 4.464992179593339e-05,
      "loss": 0.6822,
      "mean_token_accuracy": 0.7903286702930927,
      "num_tokens": 576082495.0,
      "step": 18049
    },
    {
      "entropy": 0.6317182458005846,
      "epoch": 1.6606977533194174,
      "grad_norm": 0.1515890210866928,
      "learning_rate": 4.464685496979176e-05,
      "loss": 0.618,
      "mean_token_accuracy": 0.8131836988031864,
      "num_tokens": 576114649.0,
      "step": 18050
    },
    {
      "entropy": 0.6365022845566273,
      "epoch": 1.6607897596908585,
      "grad_norm": 0.15688291192054749,
      "learning_rate": 4.464378814365014e-05,
      "loss": 0.6118,
      "mean_token_accuracy": 0.809945821762085,
      "num_tokens": 576146766.0,
      "step": 18051
    },
    {
      "entropy": 0.6464200252667069,
      "epoch": 1.6608817660622999,
      "grad_norm": 0.16660159826278687,
      "learning_rate": 4.464072131750851e-05,
      "loss": 0.6404,
      "mean_token_accuracy": 0.8083170279860497,
      "num_tokens": 576178015.0,
      "step": 18052
    },
    {
      "entropy": 0.59038209868595,
      "epoch": 1.660973772433741,
      "grad_norm": 0.1621173769235611,
      "learning_rate": 4.4637654491366884e-05,
      "loss": 0.5828,
      "mean_token_accuracy": 0.8256188929080963,
      "num_tokens": 576210390.0,
      "step": 18053
    },
    {
      "entropy": 0.581020737066865,
      "epoch": 1.6610657788051824,
      "grad_norm": 0.15313515067100525,
      "learning_rate": 4.463458766522526e-05,
      "loss": 0.5768,
      "mean_token_accuracy": 0.8257301039993763,
      "num_tokens": 576242332.0,
      "step": 18054
    },
    {
      "entropy": 0.6416329499334097,
      "epoch": 1.6611577851766235,
      "grad_norm": 0.15685246884822845,
      "learning_rate": 4.4631520839083634e-05,
      "loss": 0.6281,
      "mean_token_accuracy": 0.8077561296522617,
      "num_tokens": 576274307.0,
      "step": 18055
    },
    {
      "entropy": 0.6184592470526695,
      "epoch": 1.6612497915480646,
      "grad_norm": 0.15535937249660492,
      "learning_rate": 4.462845401294201e-05,
      "loss": 0.618,
      "mean_token_accuracy": 0.8101210854947567,
      "num_tokens": 576306892.0,
      "step": 18056
    },
    {
      "entropy": 0.6015987703576684,
      "epoch": 1.661341797919506,
      "grad_norm": 0.14796172082424164,
      "learning_rate": 4.462538718680038e-05,
      "loss": 0.5894,
      "mean_token_accuracy": 0.8206216432154179,
      "num_tokens": 576339660.0,
      "step": 18057
    },
    {
      "entropy": 0.7084956783801317,
      "epoch": 1.6614338042909471,
      "grad_norm": 0.15969395637512207,
      "learning_rate": 4.462232036065875e-05,
      "loss": 0.6986,
      "mean_token_accuracy": 0.7900053635239601,
      "num_tokens": 576371996.0,
      "step": 18058
    },
    {
      "entropy": 0.6259436700493097,
      "epoch": 1.6615258106623885,
      "grad_norm": 0.16362084448337555,
      "learning_rate": 4.461925353451713e-05,
      "loss": 0.6215,
      "mean_token_accuracy": 0.809430293738842,
      "num_tokens": 576403825.0,
      "step": 18059
    },
    {
      "entropy": 0.6428906721994281,
      "epoch": 1.6616178170338296,
      "grad_norm": 0.1586197167634964,
      "learning_rate": 4.461618670837551e-05,
      "loss": 0.6324,
      "mean_token_accuracy": 0.8080931305885315,
      "num_tokens": 576435185.0,
      "step": 18060
    },
    {
      "entropy": 0.6050984053872526,
      "epoch": 1.6617098234052707,
      "grad_norm": 0.15473002195358276,
      "learning_rate": 4.4613119882233875e-05,
      "loss": 0.5762,
      "mean_token_accuracy": 0.82209886983037,
      "num_tokens": 576466683.0,
      "step": 18061
    },
    {
      "entropy": 0.6777467969805002,
      "epoch": 1.661801829776712,
      "grad_norm": 0.15761470794677734,
      "learning_rate": 4.4610053056092256e-05,
      "loss": 0.6684,
      "mean_token_accuracy": 0.792513981461525,
      "num_tokens": 576498468.0,
      "step": 18062
    },
    {
      "entropy": 0.6535167992115021,
      "epoch": 1.6618938361481532,
      "grad_norm": 0.15369409322738647,
      "learning_rate": 4.4606986229950624e-05,
      "loss": 0.66,
      "mean_token_accuracy": 0.7947251163423061,
      "num_tokens": 576530587.0,
      "step": 18063
    },
    {
      "entropy": 0.5358627373352647,
      "epoch": 1.6619858425195946,
      "grad_norm": 0.14965586364269257,
      "learning_rate": 4.4603919403809006e-05,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8341746106743813,
      "num_tokens": 576563207.0,
      "step": 18064
    },
    {
      "entropy": 0.8105085473507643,
      "epoch": 1.6620778488910357,
      "grad_norm": 0.16916175186634064,
      "learning_rate": 4.4600852577667373e-05,
      "loss": 0.8092,
      "mean_token_accuracy": 0.7569985836744308,
      "num_tokens": 576594342.0,
      "step": 18065
    },
    {
      "entropy": 0.5040823016315699,
      "epoch": 1.6621698552624768,
      "grad_norm": 0.14731627702713013,
      "learning_rate": 4.459778575152575e-05,
      "loss": 0.4913,
      "mean_token_accuracy": 0.8472427725791931,
      "num_tokens": 576626040.0,
      "step": 18066
    },
    {
      "entropy": 0.5588035881519318,
      "epoch": 1.6622618616339182,
      "grad_norm": 0.15176689624786377,
      "learning_rate": 4.459471892538412e-05,
      "loss": 0.5404,
      "mean_token_accuracy": 0.8343801163136959,
      "num_tokens": 576657867.0,
      "step": 18067
    },
    {
      "entropy": 0.5697964075952768,
      "epoch": 1.6623538680053593,
      "grad_norm": 0.15220338106155396,
      "learning_rate": 4.45916520992425e-05,
      "loss": 0.5593,
      "mean_token_accuracy": 0.8273047469556332,
      "num_tokens": 576689604.0,
      "step": 18068
    },
    {
      "entropy": 0.6615147697739303,
      "epoch": 1.6624458743768007,
      "grad_norm": 0.1555754542350769,
      "learning_rate": 4.458858527310087e-05,
      "loss": 0.6606,
      "mean_token_accuracy": 0.8003087639808655,
      "num_tokens": 576721897.0,
      "step": 18069
    },
    {
      "entropy": 0.6551344562321901,
      "epoch": 1.6625378807482418,
      "grad_norm": 0.15769724547863007,
      "learning_rate": 4.458551844695925e-05,
      "loss": 0.6389,
      "mean_token_accuracy": 0.8010283149778843,
      "num_tokens": 576752916.0,
      "step": 18070
    },
    {
      "entropy": 0.7021841937676072,
      "epoch": 1.662629887119683,
      "grad_norm": 0.1589888036251068,
      "learning_rate": 4.4582451620817615e-05,
      "loss": 0.6861,
      "mean_token_accuracy": 0.7886137627065182,
      "num_tokens": 576784246.0,
      "step": 18071
    },
    {
      "entropy": 0.6296205641701818,
      "epoch": 1.6627218934911243,
      "grad_norm": 0.15507464110851288,
      "learning_rate": 4.4579384794675996e-05,
      "loss": 0.6217,
      "mean_token_accuracy": 0.8047246560454369,
      "num_tokens": 576816066.0,
      "step": 18072
    },
    {
      "entropy": 0.624412689357996,
      "epoch": 1.6628138998625654,
      "grad_norm": 0.16192443668842316,
      "learning_rate": 4.4576317968534364e-05,
      "loss": 0.618,
      "mean_token_accuracy": 0.8092009052634239,
      "num_tokens": 576847775.0,
      "step": 18073
    },
    {
      "entropy": 0.6891696527600288,
      "epoch": 1.6629059062340068,
      "grad_norm": 0.15934064984321594,
      "learning_rate": 4.457325114239274e-05,
      "loss": 0.6863,
      "mean_token_accuracy": 0.7920280583202839,
      "num_tokens": 576879474.0,
      "step": 18074
    },
    {
      "entropy": 0.47494963370263577,
      "epoch": 1.662997912605448,
      "grad_norm": 0.14858165383338928,
      "learning_rate": 4.457018431625111e-05,
      "loss": 0.4631,
      "mean_token_accuracy": 0.8558255955576897,
      "num_tokens": 576912242.0,
      "step": 18075
    },
    {
      "entropy": 0.6773597132414579,
      "epoch": 1.663089918976889,
      "grad_norm": 0.15910103917121887,
      "learning_rate": 4.456711749010949e-05,
      "loss": 0.6668,
      "mean_token_accuracy": 0.7930590733885765,
      "num_tokens": 576943918.0,
      "step": 18076
    },
    {
      "entropy": 0.6217924081720412,
      "epoch": 1.6631819253483304,
      "grad_norm": 0.1417572945356369,
      "learning_rate": 4.456405066396786e-05,
      "loss": 0.6041,
      "mean_token_accuracy": 0.8149918057024479,
      "num_tokens": 576976318.0,
      "step": 18077
    },
    {
      "entropy": 0.5203657187521458,
      "epoch": 1.6632739317197716,
      "grad_norm": 0.15324446558952332,
      "learning_rate": 4.456098383782624e-05,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8414462804794312,
      "num_tokens": 577008414.0,
      "step": 18078
    },
    {
      "entropy": 0.7582183182239532,
      "epoch": 1.663365938091213,
      "grad_norm": 0.16705526411533356,
      "learning_rate": 4.4557917011684605e-05,
      "loss": 0.7498,
      "mean_token_accuracy": 0.7728880867362022,
      "num_tokens": 577040757.0,
      "step": 18079
    },
    {
      "entropy": 0.6663035154342651,
      "epoch": 1.663457944462654,
      "grad_norm": 0.1651802361011505,
      "learning_rate": 4.4554850185542986e-05,
      "loss": 0.6423,
      "mean_token_accuracy": 0.8007744699716568,
      "num_tokens": 577073045.0,
      "step": 18080
    },
    {
      "entropy": 0.5788248546887189,
      "epoch": 1.6635499508340952,
      "grad_norm": 0.14677536487579346,
      "learning_rate": 4.4551783359401354e-05,
      "loss": 0.5581,
      "mean_token_accuracy": 0.8291273899376392,
      "num_tokens": 577105382.0,
      "step": 18081
    },
    {
      "entropy": 0.6929194377735257,
      "epoch": 1.6636419572055365,
      "grad_norm": 0.16172794997692108,
      "learning_rate": 4.454871653325973e-05,
      "loss": 0.6831,
      "mean_token_accuracy": 0.7931446358561516,
      "num_tokens": 577137709.0,
      "step": 18082
    },
    {
      "entropy": 0.5968289412558079,
      "epoch": 1.6637339635769777,
      "grad_norm": 0.15605302155017853,
      "learning_rate": 4.4545649707118104e-05,
      "loss": 0.5833,
      "mean_token_accuracy": 0.8195335678756237,
      "num_tokens": 577170417.0,
      "step": 18083
    },
    {
      "entropy": 0.6125463917851448,
      "epoch": 1.663825969948419,
      "grad_norm": 0.14706481993198395,
      "learning_rate": 4.454258288097648e-05,
      "loss": 0.5961,
      "mean_token_accuracy": 0.8173158839344978,
      "num_tokens": 577202185.0,
      "step": 18084
    },
    {
      "entropy": 0.645797586068511,
      "epoch": 1.6639179763198602,
      "grad_norm": 0.15359027683734894,
      "learning_rate": 4.453951605483485e-05,
      "loss": 0.6296,
      "mean_token_accuracy": 0.805704690515995,
      "num_tokens": 577234684.0,
      "step": 18085
    },
    {
      "entropy": 0.635418669320643,
      "epoch": 1.6640099826913013,
      "grad_norm": 0.1586684286594391,
      "learning_rate": 4.453644922869323e-05,
      "loss": 0.6229,
      "mean_token_accuracy": 0.8088104762136936,
      "num_tokens": 577265614.0,
      "step": 18086
    },
    {
      "entropy": 0.6293468838557601,
      "epoch": 1.6641019890627426,
      "grad_norm": 0.1624356359243393,
      "learning_rate": 4.45333824025516e-05,
      "loss": 0.6272,
      "mean_token_accuracy": 0.8087403103709221,
      "num_tokens": 577297495.0,
      "step": 18087
    },
    {
      "entropy": 0.5674877078272402,
      "epoch": 1.6641939954341838,
      "grad_norm": 0.14954060316085815,
      "learning_rate": 4.453031557640998e-05,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8275160752236843,
      "num_tokens": 577329556.0,
      "step": 18088
    },
    {
      "entropy": 0.549138767644763,
      "epoch": 1.6642860018056251,
      "grad_norm": 0.1459449976682663,
      "learning_rate": 4.452724875026835e-05,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8272685520350933,
      "num_tokens": 577361436.0,
      "step": 18089
    },
    {
      "entropy": 0.6353329587727785,
      "epoch": 1.6643780081770663,
      "grad_norm": 0.16101127862930298,
      "learning_rate": 4.4524181924126726e-05,
      "loss": 0.6329,
      "mean_token_accuracy": 0.805918488651514,
      "num_tokens": 577393509.0,
      "step": 18090
    },
    {
      "entropy": 0.6423663087189198,
      "epoch": 1.6644700145485074,
      "grad_norm": 0.1620064526796341,
      "learning_rate": 4.45211150979851e-05,
      "loss": 0.6458,
      "mean_token_accuracy": 0.803744163364172,
      "num_tokens": 577425078.0,
      "step": 18091
    },
    {
      "entropy": 0.7446543592959642,
      "epoch": 1.6645620209199488,
      "grad_norm": 0.1655103862285614,
      "learning_rate": 4.451804827184347e-05,
      "loss": 0.7546,
      "mean_token_accuracy": 0.7709299512207508,
      "num_tokens": 577457713.0,
      "step": 18092
    },
    {
      "entropy": 0.5488154962658882,
      "epoch": 1.66465402729139,
      "grad_norm": 0.15062011778354645,
      "learning_rate": 4.451498144570185e-05,
      "loss": 0.5495,
      "mean_token_accuracy": 0.828723669052124,
      "num_tokens": 577490120.0,
      "step": 18093
    },
    {
      "entropy": 0.5786855267360806,
      "epoch": 1.6647460336628312,
      "grad_norm": 0.1497105211019516,
      "learning_rate": 4.451191461956022e-05,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8205403201282024,
      "num_tokens": 577521951.0,
      "step": 18094
    },
    {
      "entropy": 0.5943891108036041,
      "epoch": 1.6648380400342724,
      "grad_norm": 0.1482742428779602,
      "learning_rate": 4.450884779341859e-05,
      "loss": 0.5797,
      "mean_token_accuracy": 0.8235910944640636,
      "num_tokens": 577554421.0,
      "step": 18095
    },
    {
      "entropy": 0.6590743232518435,
      "epoch": 1.6649300464057135,
      "grad_norm": 0.15776947140693665,
      "learning_rate": 4.450578096727697e-05,
      "loss": 0.639,
      "mean_token_accuracy": 0.8031991831958294,
      "num_tokens": 577586355.0,
      "step": 18096
    },
    {
      "entropy": 0.5885818435344845,
      "epoch": 1.6650220527771549,
      "grad_norm": 0.15594454109668732,
      "learning_rate": 4.450271414113534e-05,
      "loss": 0.564,
      "mean_token_accuracy": 0.8255924917757511,
      "num_tokens": 577617539.0,
      "step": 18097
    },
    {
      "entropy": 0.6469578687101603,
      "epoch": 1.665114059148596,
      "grad_norm": 0.16511377692222595,
      "learning_rate": 4.4499647314993716e-05,
      "loss": 0.6363,
      "mean_token_accuracy": 0.8060152977705002,
      "num_tokens": 577648481.0,
      "step": 18098
    },
    {
      "entropy": 0.576120575889945,
      "epoch": 1.6652060655200374,
      "grad_norm": 0.15868811309337616,
      "learning_rate": 4.449658048885209e-05,
      "loss": 0.5543,
      "mean_token_accuracy": 0.828580129891634,
      "num_tokens": 577679896.0,
      "step": 18099
    },
    {
      "entropy": 0.5471557476557791,
      "epoch": 1.6652980718914785,
      "grad_norm": 0.14702175557613373,
      "learning_rate": 4.449351366271046e-05,
      "loss": 0.528,
      "mean_token_accuracy": 0.8355026841163635,
      "num_tokens": 577712196.0,
      "step": 18100
    },
    {
      "entropy": 0.6738431006669998,
      "epoch": 1.6653900782629196,
      "grad_norm": 0.15502560138702393,
      "learning_rate": 4.449044683656884e-05,
      "loss": 0.6615,
      "mean_token_accuracy": 0.7954094856977463,
      "num_tokens": 577744442.0,
      "step": 18101
    },
    {
      "entropy": 0.5924012567847967,
      "epoch": 1.665482084634361,
      "grad_norm": 0.1584709882736206,
      "learning_rate": 4.448738001042721e-05,
      "loss": 0.5908,
      "mean_token_accuracy": 0.8177027031779289,
      "num_tokens": 577775994.0,
      "step": 18102
    },
    {
      "entropy": 0.5393248219043016,
      "epoch": 1.6655740910058021,
      "grad_norm": 0.15077567100524902,
      "learning_rate": 4.448431318428559e-05,
      "loss": 0.5268,
      "mean_token_accuracy": 0.8375218026340008,
      "num_tokens": 577807736.0,
      "step": 18103
    },
    {
      "entropy": 0.5710276253521442,
      "epoch": 1.6656660973772435,
      "grad_norm": 0.15304119884967804,
      "learning_rate": 4.448124635814396e-05,
      "loss": 0.5594,
      "mean_token_accuracy": 0.8265897668898106,
      "num_tokens": 577839753.0,
      "step": 18104
    },
    {
      "entropy": 0.6308994358405471,
      "epoch": 1.6657581037486846,
      "grad_norm": 0.15975964069366455,
      "learning_rate": 4.447817953200233e-05,
      "loss": 0.6148,
      "mean_token_accuracy": 0.8062648996710777,
      "num_tokens": 577871216.0,
      "step": 18105
    },
    {
      "entropy": 0.6061580181121826,
      "epoch": 1.6658501101201257,
      "grad_norm": 0.15545488893985748,
      "learning_rate": 4.447511270586071e-05,
      "loss": 0.5953,
      "mean_token_accuracy": 0.8147757835686207,
      "num_tokens": 577901830.0,
      "step": 18106
    },
    {
      "entropy": 0.6364169307053089,
      "epoch": 1.665942116491567,
      "grad_norm": 0.16477392613887787,
      "learning_rate": 4.447204587971908e-05,
      "loss": 0.6312,
      "mean_token_accuracy": 0.8100823014974594,
      "num_tokens": 577934191.0,
      "step": 18107
    },
    {
      "entropy": 0.6537208277732134,
      "epoch": 1.6660341228630082,
      "grad_norm": 0.161940336227417,
      "learning_rate": 4.4468979053577456e-05,
      "loss": 0.6445,
      "mean_token_accuracy": 0.8033558838069439,
      "num_tokens": 577965471.0,
      "step": 18108
    },
    {
      "entropy": 0.6455942466855049,
      "epoch": 1.6661261292344496,
      "grad_norm": 0.1636590212583542,
      "learning_rate": 4.446591222743583e-05,
      "loss": 0.6445,
      "mean_token_accuracy": 0.8049628026783466,
      "num_tokens": 577996535.0,
      "step": 18109
    },
    {
      "entropy": 0.6286803688853979,
      "epoch": 1.6662181356058907,
      "grad_norm": 0.15461444854736328,
      "learning_rate": 4.44628454012942e-05,
      "loss": 0.6286,
      "mean_token_accuracy": 0.8087188154459,
      "num_tokens": 578028520.0,
      "step": 18110
    },
    {
      "entropy": 0.6219283835962415,
      "epoch": 1.6663101419773318,
      "grad_norm": 0.15693344175815582,
      "learning_rate": 4.445977857515258e-05,
      "loss": 0.6124,
      "mean_token_accuracy": 0.811643410474062,
      "num_tokens": 578060932.0,
      "step": 18111
    },
    {
      "entropy": 0.5503249280154705,
      "epoch": 1.6664021483487732,
      "grad_norm": 0.14652983844280243,
      "learning_rate": 4.445671174901095e-05,
      "loss": 0.5302,
      "mean_token_accuracy": 0.8322583511471748,
      "num_tokens": 578092633.0,
      "step": 18112
    },
    {
      "entropy": 0.5998499859124422,
      "epoch": 1.6664941547202143,
      "grad_norm": 0.15826961398124695,
      "learning_rate": 4.445364492286932e-05,
      "loss": 0.5997,
      "mean_token_accuracy": 0.815842404961586,
      "num_tokens": 578124631.0,
      "step": 18113
    },
    {
      "entropy": 0.5785106457769871,
      "epoch": 1.6665861610916557,
      "grad_norm": 0.1560920923948288,
      "learning_rate": 4.44505780967277e-05,
      "loss": 0.5665,
      "mean_token_accuracy": 0.8253421187400818,
      "num_tokens": 578156545.0,
      "step": 18114
    },
    {
      "entropy": 0.657822210341692,
      "epoch": 1.6666781674630968,
      "grad_norm": 0.15836170315742493,
      "learning_rate": 4.444751127058607e-05,
      "loss": 0.6524,
      "mean_token_accuracy": 0.8046371079981327,
      "num_tokens": 578187741.0,
      "step": 18115
    },
    {
      "entropy": 0.5403626812621951,
      "epoch": 1.666770173834538,
      "grad_norm": 0.1483544111251831,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8344581462442875,
      "num_tokens": 578219545.0,
      "step": 18116
    },
    {
      "entropy": 0.6916290754452348,
      "epoch": 1.6668621802059793,
      "grad_norm": 0.15794293582439423,
      "learning_rate": 4.444137761830282e-05,
      "loss": 0.676,
      "mean_token_accuracy": 0.7941446378827095,
      "num_tokens": 578251212.0,
      "step": 18117
    },
    {
      "entropy": 0.6080209091305733,
      "epoch": 1.6669541865774204,
      "grad_norm": 0.15454481542110443,
      "learning_rate": 4.443831079216119e-05,
      "loss": 0.6001,
      "mean_token_accuracy": 0.8159416802227497,
      "num_tokens": 578282881.0,
      "step": 18118
    },
    {
      "entropy": 0.6529159937053919,
      "epoch": 1.6670461929488618,
      "grad_norm": 0.15454067289829254,
      "learning_rate": 4.443524396601957e-05,
      "loss": 0.6385,
      "mean_token_accuracy": 0.8025971613824368,
      "num_tokens": 578314823.0,
      "step": 18119
    },
    {
      "entropy": 0.6256624180823565,
      "epoch": 1.667138199320303,
      "grad_norm": 0.15087811648845673,
      "learning_rate": 4.4432177139877945e-05,
      "loss": 0.6076,
      "mean_token_accuracy": 0.8081984259188175,
      "num_tokens": 578346269.0,
      "step": 18120
    },
    {
      "entropy": 0.6432315008714795,
      "epoch": 1.667230205691744,
      "grad_norm": 0.15455397963523865,
      "learning_rate": 4.442911031373631e-05,
      "loss": 0.6244,
      "mean_token_accuracy": 0.8097775168716908,
      "num_tokens": 578378110.0,
      "step": 18121
    },
    {
      "entropy": 0.6592534841038287,
      "epoch": 1.6673222120631854,
      "grad_norm": 0.17277754843235016,
      "learning_rate": 4.4426043487594694e-05,
      "loss": 0.6483,
      "mean_token_accuracy": 0.8057345449924469,
      "num_tokens": 578409617.0,
      "step": 18122
    },
    {
      "entropy": 0.5327677791938186,
      "epoch": 1.6674142184346266,
      "grad_norm": 0.14881767332553864,
      "learning_rate": 4.442297666145306e-05,
      "loss": 0.5052,
      "mean_token_accuracy": 0.8422058187425137,
      "num_tokens": 578441305.0,
      "step": 18123
    },
    {
      "entropy": 0.7107112612575293,
      "epoch": 1.667506224806068,
      "grad_norm": 0.16556581854820251,
      "learning_rate": 4.4419909835311444e-05,
      "loss": 0.689,
      "mean_token_accuracy": 0.7862027511000633,
      "num_tokens": 578473056.0,
      "step": 18124
    },
    {
      "entropy": 0.5910538733005524,
      "epoch": 1.667598231177509,
      "grad_norm": 0.15353074669837952,
      "learning_rate": 4.441684300916981e-05,
      "loss": 0.5949,
      "mean_token_accuracy": 0.8182858824729919,
      "num_tokens": 578504182.0,
      "step": 18125
    },
    {
      "entropy": 0.7040506424382329,
      "epoch": 1.6676902375489502,
      "grad_norm": 0.1648697853088379,
      "learning_rate": 4.4413776183028186e-05,
      "loss": 0.7016,
      "mean_token_accuracy": 0.7835803776979446,
      "num_tokens": 578535013.0,
      "step": 18126
    },
    {
      "entropy": 0.48364877328276634,
      "epoch": 1.6677822439203915,
      "grad_norm": 0.14766520261764526,
      "learning_rate": 4.441070935688656e-05,
      "loss": 0.4701,
      "mean_token_accuracy": 0.8531778007745743,
      "num_tokens": 578567176.0,
      "step": 18127
    },
    {
      "entropy": 0.5603998443111777,
      "epoch": 1.6678742502918327,
      "grad_norm": 0.14873658120632172,
      "learning_rate": 4.4407642530744936e-05,
      "loss": 0.5545,
      "mean_token_accuracy": 0.8285655789077282,
      "num_tokens": 578599180.0,
      "step": 18128
    },
    {
      "entropy": 0.709986605681479,
      "epoch": 1.667966256663274,
      "grad_norm": 0.16415710747241974,
      "learning_rate": 4.440457570460331e-05,
      "loss": 0.6994,
      "mean_token_accuracy": 0.7871778719127178,
      "num_tokens": 578630770.0,
      "step": 18129
    },
    {
      "entropy": 0.6075197383761406,
      "epoch": 1.6680582630347152,
      "grad_norm": 0.1502065658569336,
      "learning_rate": 4.4401508878461685e-05,
      "loss": 0.6035,
      "mean_token_accuracy": 0.8143816106021404,
      "num_tokens": 578663284.0,
      "step": 18130
    },
    {
      "entropy": 0.606867897324264,
      "epoch": 1.6681502694061563,
      "grad_norm": 0.15098769962787628,
      "learning_rate": 4.439844205232005e-05,
      "loss": 0.5966,
      "mean_token_accuracy": 0.8233256116509438,
      "num_tokens": 578695381.0,
      "step": 18131
    },
    {
      "entropy": 0.6262895371764898,
      "epoch": 1.6682422757775976,
      "grad_norm": 0.153891921043396,
      "learning_rate": 4.4395375226178434e-05,
      "loss": 0.6275,
      "mean_token_accuracy": 0.8085298612713814,
      "num_tokens": 578727435.0,
      "step": 18132
    },
    {
      "entropy": 0.6152610073331743,
      "epoch": 1.6683342821490388,
      "grad_norm": 0.159868523478508,
      "learning_rate": 4.43923084000368e-05,
      "loss": 0.6126,
      "mean_token_accuracy": 0.8105676174163818,
      "num_tokens": 578759227.0,
      "step": 18133
    },
    {
      "entropy": 0.6172127472236753,
      "epoch": 1.6684262885204801,
      "grad_norm": 0.1568783074617386,
      "learning_rate": 4.438924157389518e-05,
      "loss": 0.6015,
      "mean_token_accuracy": 0.8161834441125393,
      "num_tokens": 578791419.0,
      "step": 18134
    },
    {
      "entropy": 0.6388647239655256,
      "epoch": 1.6685182948919213,
      "grad_norm": 0.15400925278663635,
      "learning_rate": 4.438617474775355e-05,
      "loss": 0.6288,
      "mean_token_accuracy": 0.8039311058819294,
      "num_tokens": 578823570.0,
      "step": 18135
    },
    {
      "entropy": 0.6438169358298182,
      "epoch": 1.6686103012633624,
      "grad_norm": 0.16099874675273895,
      "learning_rate": 4.4383107921611926e-05,
      "loss": 0.6452,
      "mean_token_accuracy": 0.803695023059845,
      "num_tokens": 578856336.0,
      "step": 18136
    },
    {
      "entropy": 0.6055660881102085,
      "epoch": 1.6687023076348038,
      "grad_norm": 0.15763571858406067,
      "learning_rate": 4.43800410954703e-05,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8154997304081917,
      "num_tokens": 578887982.0,
      "step": 18137
    },
    {
      "entropy": 0.4689360617194325,
      "epoch": 1.668794314006245,
      "grad_norm": 0.14584365487098694,
      "learning_rate": 4.4376974269328675e-05,
      "loss": 0.4503,
      "mean_token_accuracy": 0.8579132333397865,
      "num_tokens": 578919754.0,
      "step": 18138
    },
    {
      "entropy": 0.6055549141019583,
      "epoch": 1.6688863203776862,
      "grad_norm": 0.15443217754364014,
      "learning_rate": 4.437390744318704e-05,
      "loss": 0.5961,
      "mean_token_accuracy": 0.8207067400217056,
      "num_tokens": 578951947.0,
      "step": 18139
    },
    {
      "entropy": 0.5538912629708648,
      "epoch": 1.6689783267491274,
      "grad_norm": 0.14908693730831146,
      "learning_rate": 4.4370840617045425e-05,
      "loss": 0.5477,
      "mean_token_accuracy": 0.8346233442425728,
      "num_tokens": 578983477.0,
      "step": 18140
    },
    {
      "entropy": 0.6502531692385674,
      "epoch": 1.6690703331205685,
      "grad_norm": 0.16029676795005798,
      "learning_rate": 4.436777379090379e-05,
      "loss": 0.6469,
      "mean_token_accuracy": 0.800416387617588,
      "num_tokens": 579015155.0,
      "step": 18141
    },
    {
      "entropy": 0.6004307097755373,
      "epoch": 1.6691623394920099,
      "grad_norm": 0.16278813779354095,
      "learning_rate": 4.4364706964762174e-05,
      "loss": 0.5913,
      "mean_token_accuracy": 0.817766971886158,
      "num_tokens": 579046481.0,
      "step": 18142
    },
    {
      "entropy": 0.5538263600319624,
      "epoch": 1.669254345863451,
      "grad_norm": 0.14893732964992523,
      "learning_rate": 4.436164013862054e-05,
      "loss": 0.5398,
      "mean_token_accuracy": 0.83177300542593,
      "num_tokens": 579079204.0,
      "step": 18143
    },
    {
      "entropy": 0.5738144107162952,
      "epoch": 1.6693463522348924,
      "grad_norm": 0.15234892070293427,
      "learning_rate": 4.4358573312478916e-05,
      "loss": 0.5665,
      "mean_token_accuracy": 0.8262820281088352,
      "num_tokens": 579111278.0,
      "step": 18144
    },
    {
      "entropy": 0.64921710267663,
      "epoch": 1.6694383586063335,
      "grad_norm": 0.1583639234304428,
      "learning_rate": 4.435550648633729e-05,
      "loss": 0.6429,
      "mean_token_accuracy": 0.8040217906236649,
      "num_tokens": 579142799.0,
      "step": 18145
    },
    {
      "entropy": 0.5605803392827511,
      "epoch": 1.6695303649777746,
      "grad_norm": 0.15677690505981445,
      "learning_rate": 4.4352439660195666e-05,
      "loss": 0.5617,
      "mean_token_accuracy": 0.8217168860137463,
      "num_tokens": 579174301.0,
      "step": 18146
    },
    {
      "entropy": 0.45065529691055417,
      "epoch": 1.669622371349216,
      "grad_norm": 0.1422102153301239,
      "learning_rate": 4.434937283405404e-05,
      "loss": 0.429,
      "mean_token_accuracy": 0.8668825104832649,
      "num_tokens": 579206546.0,
      "step": 18147
    },
    {
      "entropy": 0.5312519897706807,
      "epoch": 1.6697143777206571,
      "grad_norm": 0.154485285282135,
      "learning_rate": 4.4346306007912415e-05,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8405753187835217,
      "num_tokens": 579238855.0,
      "step": 18148
    },
    {
      "entropy": 0.7315480634570122,
      "epoch": 1.6698063840920985,
      "grad_norm": 0.16272255778312683,
      "learning_rate": 4.434323918177078e-05,
      "loss": 0.7322,
      "mean_token_accuracy": 0.7760665714740753,
      "num_tokens": 579271048.0,
      "step": 18149
    },
    {
      "entropy": 0.6605300009250641,
      "epoch": 1.6698983904635396,
      "grad_norm": 0.15959668159484863,
      "learning_rate": 4.4340172355629164e-05,
      "loss": 0.6445,
      "mean_token_accuracy": 0.8047159239649773,
      "num_tokens": 579303013.0,
      "step": 18150
    },
    {
      "entropy": 0.5857319196220487,
      "epoch": 1.6699903968349807,
      "grad_norm": 0.1500224471092224,
      "learning_rate": 4.433710552948754e-05,
      "loss": 0.5718,
      "mean_token_accuracy": 0.8207821659743786,
      "num_tokens": 579335388.0,
      "step": 18151
    },
    {
      "entropy": 0.5609546285122633,
      "epoch": 1.670082403206422,
      "grad_norm": 0.14816004037857056,
      "learning_rate": 4.433403870334591e-05,
      "loss": 0.5494,
      "mean_token_accuracy": 0.8289247862994671,
      "num_tokens": 579367708.0,
      "step": 18152
    },
    {
      "entropy": 0.6107095871120691,
      "epoch": 1.6701744095778632,
      "grad_norm": 0.15753932297229767,
      "learning_rate": 4.433097187720429e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8184393271803856,
      "num_tokens": 579399826.0,
      "step": 18153
    },
    {
      "entropy": 0.573121593799442,
      "epoch": 1.6702664159493046,
      "grad_norm": 0.15018294751644135,
      "learning_rate": 4.4327905051062656e-05,
      "loss": 0.5545,
      "mean_token_accuracy": 0.827273566275835,
      "num_tokens": 579432304.0,
      "step": 18154
    },
    {
      "entropy": 0.6720980815589428,
      "epoch": 1.6703584223207457,
      "grad_norm": 0.16012856364250183,
      "learning_rate": 4.432483822492103e-05,
      "loss": 0.6523,
      "mean_token_accuracy": 0.8044956363737583,
      "num_tokens": 579463273.0,
      "step": 18155
    },
    {
      "entropy": 0.5838097613304853,
      "epoch": 1.6704504286921869,
      "grad_norm": 0.15854519605636597,
      "learning_rate": 4.4321771398779405e-05,
      "loss": 0.5788,
      "mean_token_accuracy": 0.8223811946809292,
      "num_tokens": 579494950.0,
      "step": 18156
    },
    {
      "entropy": 0.6763705778867006,
      "epoch": 1.6705424350636282,
      "grad_norm": 0.15640713274478912,
      "learning_rate": 4.431870457263778e-05,
      "loss": 0.6692,
      "mean_token_accuracy": 0.795939601957798,
      "num_tokens": 579526716.0,
      "step": 18157
    },
    {
      "entropy": 0.6668069083243608,
      "epoch": 1.6706344414350693,
      "grad_norm": 0.1662401407957077,
      "learning_rate": 4.4315637746496155e-05,
      "loss": 0.6588,
      "mean_token_accuracy": 0.7992674000561237,
      "num_tokens": 579557517.0,
      "step": 18158
    },
    {
      "entropy": 0.6225216332823038,
      "epoch": 1.6707264478065107,
      "grad_norm": 0.1515757292509079,
      "learning_rate": 4.431257092035453e-05,
      "loss": 0.6123,
      "mean_token_accuracy": 0.8077355176210403,
      "num_tokens": 579589712.0,
      "step": 18159
    },
    {
      "entropy": 0.6177056459710002,
      "epoch": 1.6708184541779518,
      "grad_norm": 0.1538446992635727,
      "learning_rate": 4.43095040942129e-05,
      "loss": 0.6178,
      "mean_token_accuracy": 0.8117533065378666,
      "num_tokens": 579621924.0,
      "step": 18160
    },
    {
      "entropy": 0.4675157438032329,
      "epoch": 1.670910460549393,
      "grad_norm": 0.14244891703128815,
      "learning_rate": 4.430643726807128e-05,
      "loss": 0.4464,
      "mean_token_accuracy": 0.8604089841246605,
      "num_tokens": 579654321.0,
      "step": 18161
    },
    {
      "entropy": 0.6136738173663616,
      "epoch": 1.6710024669208343,
      "grad_norm": 0.16191346943378448,
      "learning_rate": 4.4303370441929646e-05,
      "loss": 0.617,
      "mean_token_accuracy": 0.811640877276659,
      "num_tokens": 579686487.0,
      "step": 18162
    },
    {
      "entropy": 0.5570085868239403,
      "epoch": 1.6710944732922755,
      "grad_norm": 0.16172261536121368,
      "learning_rate": 4.430030361578803e-05,
      "loss": 0.5375,
      "mean_token_accuracy": 0.8346844650804996,
      "num_tokens": 579718233.0,
      "step": 18163
    },
    {
      "entropy": 0.4997693384066224,
      "epoch": 1.6711864796637168,
      "grad_norm": 0.14287161827087402,
      "learning_rate": 4.4297236789646396e-05,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8512484356760979,
      "num_tokens": 579750281.0,
      "step": 18164
    },
    {
      "entropy": 0.6568299010396004,
      "epoch": 1.671278486035158,
      "grad_norm": 0.15553025901317596,
      "learning_rate": 4.429416996350477e-05,
      "loss": 0.6426,
      "mean_token_accuracy": 0.8033244535326958,
      "num_tokens": 579782656.0,
      "step": 18165
    },
    {
      "entropy": 0.6849106717854738,
      "epoch": 1.671370492406599,
      "grad_norm": 0.1587713360786438,
      "learning_rate": 4.4291103137363145e-05,
      "loss": 0.6848,
      "mean_token_accuracy": 0.7916042283177376,
      "num_tokens": 579814543.0,
      "step": 18166
    },
    {
      "entropy": 0.6232204819098115,
      "epoch": 1.6714624987780404,
      "grad_norm": 0.15416614711284637,
      "learning_rate": 4.428803631122152e-05,
      "loss": 0.6093,
      "mean_token_accuracy": 0.8151821978390217,
      "num_tokens": 579846463.0,
      "step": 18167
    },
    {
      "entropy": 0.6244965754449368,
      "epoch": 1.6715545051494816,
      "grad_norm": 0.16342145204544067,
      "learning_rate": 4.4284969485079894e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.8069352470338345,
      "num_tokens": 579878753.0,
      "step": 18168
    },
    {
      "entropy": 0.48759941570460796,
      "epoch": 1.671646511520923,
      "grad_norm": 0.1460922509431839,
      "learning_rate": 4.428190265893827e-05,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8515124246478081,
      "num_tokens": 579910275.0,
      "step": 18169
    },
    {
      "entropy": 0.6374774109572172,
      "epoch": 1.671738517892364,
      "grad_norm": 0.1613079458475113,
      "learning_rate": 4.427883583279664e-05,
      "loss": 0.6242,
      "mean_token_accuracy": 0.8101133890450001,
      "num_tokens": 579941890.0,
      "step": 18170
    },
    {
      "entropy": 0.5232750354334712,
      "epoch": 1.6718305242638052,
      "grad_norm": 0.16037452220916748,
      "learning_rate": 4.427576900665502e-05,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8415237255394459,
      "num_tokens": 579974243.0,
      "step": 18171
    },
    {
      "entropy": 0.471366825979203,
      "epoch": 1.6719225306352465,
      "grad_norm": 0.1493290364742279,
      "learning_rate": 4.4272702180513386e-05,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8585121892392635,
      "num_tokens": 580006439.0,
      "step": 18172
    },
    {
      "entropy": 0.6371158761903644,
      "epoch": 1.6720145370066877,
      "grad_norm": 0.15698064863681793,
      "learning_rate": 4.426963535437176e-05,
      "loss": 0.6342,
      "mean_token_accuracy": 0.8073808141052723,
      "num_tokens": 580038683.0,
      "step": 18173
    },
    {
      "entropy": 0.6166638117283583,
      "epoch": 1.672106543378129,
      "grad_norm": 0.15500260889530182,
      "learning_rate": 4.4266568528230135e-05,
      "loss": 0.5961,
      "mean_token_accuracy": 0.8164867274463177,
      "num_tokens": 580070872.0,
      "step": 18174
    },
    {
      "entropy": 0.6803553216159344,
      "epoch": 1.6721985497495702,
      "grad_norm": 0.16041484475135803,
      "learning_rate": 4.426350170208851e-05,
      "loss": 0.6742,
      "mean_token_accuracy": 0.7950804270803928,
      "num_tokens": 580102971.0,
      "step": 18175
    },
    {
      "entropy": 0.6060168789699674,
      "epoch": 1.6722905561210113,
      "grad_norm": 0.15346312522888184,
      "learning_rate": 4.4260434875946885e-05,
      "loss": 0.6046,
      "mean_token_accuracy": 0.814266886562109,
      "num_tokens": 580134857.0,
      "step": 18176
    },
    {
      "entropy": 0.5616825092583895,
      "epoch": 1.6723825624924527,
      "grad_norm": 0.14765937626361847,
      "learning_rate": 4.425736804980526e-05,
      "loss": 0.5437,
      "mean_token_accuracy": 0.8341588526964188,
      "num_tokens": 580167353.0,
      "step": 18177
    },
    {
      "entropy": 0.6513635911978781,
      "epoch": 1.6724745688638938,
      "grad_norm": 0.1583004891872406,
      "learning_rate": 4.425430122366363e-05,
      "loss": 0.6431,
      "mean_token_accuracy": 0.8079972229897976,
      "num_tokens": 580199681.0,
      "step": 18178
    },
    {
      "entropy": 0.583634982816875,
      "epoch": 1.6725665752353351,
      "grad_norm": 0.15362851321697235,
      "learning_rate": 4.425123439752201e-05,
      "loss": 0.5737,
      "mean_token_accuracy": 0.8233050182461739,
      "num_tokens": 580231897.0,
      "step": 18179
    },
    {
      "entropy": 0.5702108107507229,
      "epoch": 1.6726585816067763,
      "grad_norm": 0.15982036292552948,
      "learning_rate": 4.4248167571380377e-05,
      "loss": 0.5674,
      "mean_token_accuracy": 0.8281018622219563,
      "num_tokens": 580263485.0,
      "step": 18180
    },
    {
      "entropy": 0.5763865690678358,
      "epoch": 1.6727505879782174,
      "grad_norm": 0.16630299389362335,
      "learning_rate": 4.424510074523875e-05,
      "loss": 0.5683,
      "mean_token_accuracy": 0.8253846280276775,
      "num_tokens": 580295581.0,
      "step": 18181
    },
    {
      "entropy": 0.5878303016070276,
      "epoch": 1.6728425943496588,
      "grad_norm": 0.15722189843654633,
      "learning_rate": 4.424203391909713e-05,
      "loss": 0.5753,
      "mean_token_accuracy": 0.8245633170008659,
      "num_tokens": 580326863.0,
      "step": 18182
    },
    {
      "entropy": 0.4951922604814172,
      "epoch": 1.6729346007211,
      "grad_norm": 0.14852531254291534,
      "learning_rate": 4.42389670929555e-05,
      "loss": 0.488,
      "mean_token_accuracy": 0.8480178862810135,
      "num_tokens": 580358919.0,
      "step": 18183
    },
    {
      "entropy": 0.6806748514063656,
      "epoch": 1.6730266070925413,
      "grad_norm": 0.15835314989089966,
      "learning_rate": 4.423590026681388e-05,
      "loss": 0.6612,
      "mean_token_accuracy": 0.7967010177671909,
      "num_tokens": 580391039.0,
      "step": 18184
    },
    {
      "entropy": 0.5952403098344803,
      "epoch": 1.6731186134639824,
      "grad_norm": 0.15211568772792816,
      "learning_rate": 4.423283344067225e-05,
      "loss": 0.5878,
      "mean_token_accuracy": 0.8224467188119888,
      "num_tokens": 580423018.0,
      "step": 18185
    },
    {
      "entropy": 0.5220026094466448,
      "epoch": 1.6732106198354235,
      "grad_norm": 0.14973480999469757,
      "learning_rate": 4.4229766614530624e-05,
      "loss": 0.503,
      "mean_token_accuracy": 0.842437244951725,
      "num_tokens": 580454712.0,
      "step": 18186
    },
    {
      "entropy": 0.6029074285179377,
      "epoch": 1.6733026262068649,
      "grad_norm": 0.15044492483139038,
      "learning_rate": 4.4226699788389e-05,
      "loss": 0.5952,
      "mean_token_accuracy": 0.8193572908639908,
      "num_tokens": 580487309.0,
      "step": 18187
    },
    {
      "entropy": 0.5889033218845725,
      "epoch": 1.673394632578306,
      "grad_norm": 0.15108926594257355,
      "learning_rate": 4.4223632962247374e-05,
      "loss": 0.5831,
      "mean_token_accuracy": 0.8194884099066257,
      "num_tokens": 580518751.0,
      "step": 18188
    },
    {
      "entropy": 0.6053304215893149,
      "epoch": 1.6734866389497474,
      "grad_norm": 0.1552760750055313,
      "learning_rate": 4.422056613610575e-05,
      "loss": 0.5852,
      "mean_token_accuracy": 0.8171234466135502,
      "num_tokens": 580551212.0,
      "step": 18189
    },
    {
      "entropy": 0.5988119710236788,
      "epoch": 1.6735786453211885,
      "grad_norm": 0.1599041074514389,
      "learning_rate": 4.421749930996412e-05,
      "loss": 0.5829,
      "mean_token_accuracy": 0.8211552910506725,
      "num_tokens": 580582565.0,
      "step": 18190
    },
    {
      "entropy": 0.6185165154747665,
      "epoch": 1.6736706516926296,
      "grad_norm": 0.15956220030784607,
      "learning_rate": 4.421443248382249e-05,
      "loss": 0.5947,
      "mean_token_accuracy": 0.81314392760396,
      "num_tokens": 580614110.0,
      "step": 18191
    },
    {
      "entropy": 0.5151458019390702,
      "epoch": 1.673762658064071,
      "grad_norm": 0.15009784698486328,
      "learning_rate": 4.421136565768087e-05,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8464117497205734,
      "num_tokens": 580646851.0,
      "step": 18192
    },
    {
      "entropy": 0.6883588656783104,
      "epoch": 1.6738546644355121,
      "grad_norm": 0.15961213409900665,
      "learning_rate": 4.420829883153924e-05,
      "loss": 0.6635,
      "mean_token_accuracy": 0.7966697253286839,
      "num_tokens": 580679057.0,
      "step": 18193
    },
    {
      "entropy": 0.6632695142179728,
      "epoch": 1.6739466708069535,
      "grad_norm": 0.15960584580898285,
      "learning_rate": 4.4205232005397615e-05,
      "loss": 0.6456,
      "mean_token_accuracy": 0.8038932457566261,
      "num_tokens": 580710771.0,
      "step": 18194
    },
    {
      "entropy": 0.6230181250721216,
      "epoch": 1.6740386771783946,
      "grad_norm": 0.1506483554840088,
      "learning_rate": 4.420216517925599e-05,
      "loss": 0.6036,
      "mean_token_accuracy": 0.8173487223684788,
      "num_tokens": 580742912.0,
      "step": 18195
    },
    {
      "entropy": 0.554826807230711,
      "epoch": 1.6741306835498357,
      "grad_norm": 0.15061554312705994,
      "learning_rate": 4.4199098353114364e-05,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8312606550753117,
      "num_tokens": 580775125.0,
      "step": 18196
    },
    {
      "entropy": 0.5544152287766337,
      "epoch": 1.674222689921277,
      "grad_norm": 0.15091414749622345,
      "learning_rate": 4.419603152697274e-05,
      "loss": 0.5364,
      "mean_token_accuracy": 0.832951758056879,
      "num_tokens": 580807280.0,
      "step": 18197
    },
    {
      "entropy": 0.6434188364073634,
      "epoch": 1.6743146962927182,
      "grad_norm": 0.16453512012958527,
      "learning_rate": 4.4192964700831113e-05,
      "loss": 0.6379,
      "mean_token_accuracy": 0.8042763359844685,
      "num_tokens": 580838657.0,
      "step": 18198
    },
    {
      "entropy": 0.6425942899659276,
      "epoch": 1.6744067026641596,
      "grad_norm": 0.16151206195354462,
      "learning_rate": 4.418989787468948e-05,
      "loss": 0.6324,
      "mean_token_accuracy": 0.8032471500337124,
      "num_tokens": 580869601.0,
      "step": 18199
    },
    {
      "entropy": 0.605420783162117,
      "epoch": 1.6744987090356007,
      "grad_norm": 0.15634967386722565,
      "learning_rate": 4.418683104854786e-05,
      "loss": 0.586,
      "mean_token_accuracy": 0.8189966902136803,
      "num_tokens": 580900917.0,
      "step": 18200
    },
    {
      "entropy": 0.6986872628331184,
      "epoch": 1.6745907154070419,
      "grad_norm": 0.15841235220432281,
      "learning_rate": 4.418376422240623e-05,
      "loss": 0.6907,
      "mean_token_accuracy": 0.7906171418726444,
      "num_tokens": 580933613.0,
      "step": 18201
    },
    {
      "entropy": 0.6897006407380104,
      "epoch": 1.6746827217784832,
      "grad_norm": 0.16785375773906708,
      "learning_rate": 4.418069739626461e-05,
      "loss": 0.701,
      "mean_token_accuracy": 0.7891482077538967,
      "num_tokens": 580966055.0,
      "step": 18202
    },
    {
      "entropy": 0.5375266466289759,
      "epoch": 1.6747747281499243,
      "grad_norm": 0.14754201471805573,
      "learning_rate": 4.417763057012298e-05,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8357976675033569,
      "num_tokens": 580998627.0,
      "step": 18203
    },
    {
      "entropy": 0.40490588638931513,
      "epoch": 1.6748667345213657,
      "grad_norm": 0.14097586274147034,
      "learning_rate": 4.4174563743981355e-05,
      "loss": 0.3824,
      "mean_token_accuracy": 0.8795833177864552,
      "num_tokens": 581029495.0,
      "step": 18204
    },
    {
      "entropy": 0.6059495098888874,
      "epoch": 1.6749587408928068,
      "grad_norm": 0.15165777504444122,
      "learning_rate": 4.417149691783973e-05,
      "loss": 0.6035,
      "mean_token_accuracy": 0.815094955265522,
      "num_tokens": 581061862.0,
      "step": 18205
    },
    {
      "entropy": 0.6582096554338932,
      "epoch": 1.675050747264248,
      "grad_norm": 0.16481277346611023,
      "learning_rate": 4.4168430091698104e-05,
      "loss": 0.6576,
      "mean_token_accuracy": 0.7983002848923206,
      "num_tokens": 581093775.0,
      "step": 18206
    },
    {
      "entropy": 0.5253190035000443,
      "epoch": 1.6751427536356893,
      "grad_norm": 0.15070638060569763,
      "learning_rate": 4.416536326555648e-05,
      "loss": 0.5143,
      "mean_token_accuracy": 0.8385881148278713,
      "num_tokens": 581126098.0,
      "step": 18207
    },
    {
      "entropy": 0.5503243622370064,
      "epoch": 1.6752347600071305,
      "grad_norm": 0.15482188761234283,
      "learning_rate": 4.416229643941485e-05,
      "loss": 0.5377,
      "mean_token_accuracy": 0.8332651369273663,
      "num_tokens": 581157761.0,
      "step": 18208
    },
    {
      "entropy": 0.5429349550977349,
      "epoch": 1.6753267663785718,
      "grad_norm": 0.14931701123714447,
      "learning_rate": 4.415922961327322e-05,
      "loss": 0.539,
      "mean_token_accuracy": 0.8336086757481098,
      "num_tokens": 581189782.0,
      "step": 18209
    },
    {
      "entropy": 0.5997992167249322,
      "epoch": 1.675418772750013,
      "grad_norm": 0.15225978195667267,
      "learning_rate": 4.41561627871316e-05,
      "loss": 0.5747,
      "mean_token_accuracy": 0.8216887712478638,
      "num_tokens": 581221395.0,
      "step": 18210
    },
    {
      "entropy": 0.6008401094004512,
      "epoch": 1.675510779121454,
      "grad_norm": 0.16043967008590698,
      "learning_rate": 4.415309596098998e-05,
      "loss": 0.5836,
      "mean_token_accuracy": 0.8186898194253445,
      "num_tokens": 581253890.0,
      "step": 18211
    },
    {
      "entropy": 0.5906290467828512,
      "epoch": 1.6756027854928954,
      "grad_norm": 0.14755429327487946,
      "learning_rate": 4.4150029134848345e-05,
      "loss": 0.5835,
      "mean_token_accuracy": 0.8206791803240776,
      "num_tokens": 581286576.0,
      "step": 18212
    },
    {
      "entropy": 0.5058987280353904,
      "epoch": 1.6756947918643366,
      "grad_norm": 0.1479005515575409,
      "learning_rate": 4.4146962308706726e-05,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8490747176110744,
      "num_tokens": 581318059.0,
      "step": 18213
    },
    {
      "entropy": 0.5466127500403672,
      "epoch": 1.675786798235778,
      "grad_norm": 0.14468596875667572,
      "learning_rate": 4.4143895482565094e-05,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8346060961484909,
      "num_tokens": 581350566.0,
      "step": 18214
    },
    {
      "entropy": 0.5110579477623105,
      "epoch": 1.675878804607219,
      "grad_norm": 0.14605078101158142,
      "learning_rate": 4.414082865642347e-05,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8482770398259163,
      "num_tokens": 581382484.0,
      "step": 18215
    },
    {
      "entropy": 0.5763339409604669,
      "epoch": 1.6759708109786602,
      "grad_norm": 0.15175580978393555,
      "learning_rate": 4.4137761830281844e-05,
      "loss": 0.5691,
      "mean_token_accuracy": 0.8268751464784145,
      "num_tokens": 581415117.0,
      "step": 18216
    },
    {
      "entropy": 0.6340409778058529,
      "epoch": 1.6760628173501015,
      "grad_norm": 0.1634058654308319,
      "learning_rate": 4.413469500414022e-05,
      "loss": 0.6204,
      "mean_token_accuracy": 0.8069071173667908,
      "num_tokens": 581446172.0,
      "step": 18217
    },
    {
      "entropy": 0.6474154982715845,
      "epoch": 1.6761548237215427,
      "grad_norm": 0.14937950670719147,
      "learning_rate": 4.413162817799859e-05,
      "loss": 0.6298,
      "mean_token_accuracy": 0.8094747178256512,
      "num_tokens": 581478557.0,
      "step": 18218
    },
    {
      "entropy": 0.6058244127780199,
      "epoch": 1.676246830092984,
      "grad_norm": 0.15718653798103333,
      "learning_rate": 4.412856135185697e-05,
      "loss": 0.5953,
      "mean_token_accuracy": 0.8183009289205074,
      "num_tokens": 581510428.0,
      "step": 18219
    },
    {
      "entropy": 0.6915724240243435,
      "epoch": 1.6763388364644252,
      "grad_norm": 0.16599993407726288,
      "learning_rate": 4.4125494525715335e-05,
      "loss": 0.6872,
      "mean_token_accuracy": 0.7930403687059879,
      "num_tokens": 581542061.0,
      "step": 18220
    },
    {
      "entropy": 0.572188210207969,
      "epoch": 1.6764308428358663,
      "grad_norm": 0.155237078666687,
      "learning_rate": 4.412242769957372e-05,
      "loss": 0.5725,
      "mean_token_accuracy": 0.8256289623677731,
      "num_tokens": 581574829.0,
      "step": 18221
    },
    {
      "entropy": 0.5710374731570482,
      "epoch": 1.6765228492073077,
      "grad_norm": 0.15414534509181976,
      "learning_rate": 4.4119360873432085e-05,
      "loss": 0.5605,
      "mean_token_accuracy": 0.8264559097588062,
      "num_tokens": 581607166.0,
      "step": 18222
    },
    {
      "entropy": 0.6250860840082169,
      "epoch": 1.6766148555787488,
      "grad_norm": 0.16327515244483948,
      "learning_rate": 4.4116294047290466e-05,
      "loss": 0.6088,
      "mean_token_accuracy": 0.8102397210896015,
      "num_tokens": 581638406.0,
      "step": 18223
    },
    {
      "entropy": 0.513842255808413,
      "epoch": 1.6767068619501901,
      "grad_norm": 0.14268723130226135,
      "learning_rate": 4.4113227221148834e-05,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8431757725775242,
      "num_tokens": 581670502.0,
      "step": 18224
    },
    {
      "entropy": 0.6172311017289758,
      "epoch": 1.6767988683216313,
      "grad_norm": 0.15031762421131134,
      "learning_rate": 4.411016039500721e-05,
      "loss": 0.6115,
      "mean_token_accuracy": 0.8119894452393055,
      "num_tokens": 581702237.0,
      "step": 18225
    },
    {
      "entropy": 0.5362009601667523,
      "epoch": 1.6768908746930724,
      "grad_norm": 0.14803528785705566,
      "learning_rate": 4.410709356886558e-05,
      "loss": 0.5265,
      "mean_token_accuracy": 0.8387880437076092,
      "num_tokens": 581734570.0,
      "step": 18226
    },
    {
      "entropy": 0.6103075686842203,
      "epoch": 1.6769828810645138,
      "grad_norm": 0.1582091897726059,
      "learning_rate": 4.410402674272396e-05,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8180103749036789,
      "num_tokens": 581765952.0,
      "step": 18227
    },
    {
      "entropy": 0.5525245242752135,
      "epoch": 1.677074887435955,
      "grad_norm": 0.14610189199447632,
      "learning_rate": 4.410095991658233e-05,
      "loss": 0.5315,
      "mean_token_accuracy": 0.8379650004208088,
      "num_tokens": 581797857.0,
      "step": 18228
    },
    {
      "entropy": 0.5021331403404474,
      "epoch": 1.6771668938073963,
      "grad_norm": 0.15175575017929077,
      "learning_rate": 4.409789309044071e-05,
      "loss": 0.5037,
      "mean_token_accuracy": 0.8478409647941589,
      "num_tokens": 581830333.0,
      "step": 18229
    },
    {
      "entropy": 0.6293753972277045,
      "epoch": 1.6772589001788374,
      "grad_norm": 0.15594549477100372,
      "learning_rate": 4.4094826264299075e-05,
      "loss": 0.6189,
      "mean_token_accuracy": 0.8129195161163807,
      "num_tokens": 581862641.0,
      "step": 18230
    },
    {
      "entropy": 0.5735914115794003,
      "epoch": 1.6773509065502785,
      "grad_norm": 0.14909344911575317,
      "learning_rate": 4.4091759438157457e-05,
      "loss": 0.5617,
      "mean_token_accuracy": 0.8304219953715801,
      "num_tokens": 581895195.0,
      "step": 18231
    },
    {
      "entropy": 0.7004291713237762,
      "epoch": 1.6774429129217199,
      "grad_norm": 0.16271574795246124,
      "learning_rate": 4.4088692612015824e-05,
      "loss": 0.706,
      "mean_token_accuracy": 0.7906401492655277,
      "num_tokens": 581926468.0,
      "step": 18232
    },
    {
      "entropy": 0.6742025595158339,
      "epoch": 1.677534919293161,
      "grad_norm": 0.15661388635635376,
      "learning_rate": 4.40856257858742e-05,
      "loss": 0.6545,
      "mean_token_accuracy": 0.7983842827379704,
      "num_tokens": 581958870.0,
      "step": 18233
    },
    {
      "entropy": 0.5266702296212316,
      "epoch": 1.6776269256646024,
      "grad_norm": 0.15392431616783142,
      "learning_rate": 4.4082558959732574e-05,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8393741473555565,
      "num_tokens": 581991115.0,
      "step": 18234
    },
    {
      "entropy": 0.5188803048804402,
      "epoch": 1.6777189320360435,
      "grad_norm": 0.14643065631389618,
      "learning_rate": 4.407949213359095e-05,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8441047295928001,
      "num_tokens": 582023446.0,
      "step": 18235
    },
    {
      "entropy": 0.5215158956125379,
      "epoch": 1.6778109384074846,
      "grad_norm": 0.1434207260608673,
      "learning_rate": 4.407642530744932e-05,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8412204384803772,
      "num_tokens": 582055039.0,
      "step": 18236
    },
    {
      "entropy": 0.501891546882689,
      "epoch": 1.677902944778926,
      "grad_norm": 0.14708556234836578,
      "learning_rate": 4.40733584813077e-05,
      "loss": 0.477,
      "mean_token_accuracy": 0.8478039763867855,
      "num_tokens": 582086752.0,
      "step": 18237
    },
    {
      "entropy": 0.6592328734695911,
      "epoch": 1.6779949511503671,
      "grad_norm": 0.16025936603546143,
      "learning_rate": 4.4070291655166065e-05,
      "loss": 0.6595,
      "mean_token_accuracy": 0.7975529581308365,
      "num_tokens": 582119031.0,
      "step": 18238
    },
    {
      "entropy": 0.5608880678191781,
      "epoch": 1.6780869575218085,
      "grad_norm": 0.15069596469402313,
      "learning_rate": 4.406722482902445e-05,
      "loss": 0.5424,
      "mean_token_accuracy": 0.833646796643734,
      "num_tokens": 582151017.0,
      "step": 18239
    },
    {
      "entropy": 0.5964005645364523,
      "epoch": 1.6781789638932496,
      "grad_norm": 0.1512700468301773,
      "learning_rate": 4.4064158002882815e-05,
      "loss": 0.5765,
      "mean_token_accuracy": 0.8212694525718689,
      "num_tokens": 582183246.0,
      "step": 18240
    },
    {
      "entropy": 0.5924228690564632,
      "epoch": 1.6782709702646907,
      "grad_norm": 0.15854409337043762,
      "learning_rate": 4.406109117674119e-05,
      "loss": 0.5933,
      "mean_token_accuracy": 0.8170623481273651,
      "num_tokens": 582214881.0,
      "step": 18241
    },
    {
      "entropy": 0.6295134359970689,
      "epoch": 1.678362976636132,
      "grad_norm": 0.15197345614433289,
      "learning_rate": 4.405802435059957e-05,
      "loss": 0.6091,
      "mean_token_accuracy": 0.8156432807445526,
      "num_tokens": 582246566.0,
      "step": 18242
    },
    {
      "entropy": 0.6685749851167202,
      "epoch": 1.6784549830075732,
      "grad_norm": 0.15860028564929962,
      "learning_rate": 4.405495752445794e-05,
      "loss": 0.6583,
      "mean_token_accuracy": 0.8005097024142742,
      "num_tokens": 582278381.0,
      "step": 18243
    },
    {
      "entropy": 0.6234974712133408,
      "epoch": 1.6785469893790146,
      "grad_norm": 0.15887963771820068,
      "learning_rate": 4.405189069831632e-05,
      "loss": 0.6178,
      "mean_token_accuracy": 0.8130819089710712,
      "num_tokens": 582310579.0,
      "step": 18244
    },
    {
      "entropy": 0.7034777905791998,
      "epoch": 1.6786389957504557,
      "grad_norm": 0.1627415269613266,
      "learning_rate": 4.404882387217469e-05,
      "loss": 0.7028,
      "mean_token_accuracy": 0.7879655659198761,
      "num_tokens": 582341298.0,
      "step": 18245
    },
    {
      "entropy": 0.5992383304983377,
      "epoch": 1.6787310021218969,
      "grad_norm": 0.15483564138412476,
      "learning_rate": 4.404575704603306e-05,
      "loss": 0.5936,
      "mean_token_accuracy": 0.8178233727812767,
      "num_tokens": 582373487.0,
      "step": 18246
    },
    {
      "entropy": 0.6919922986999154,
      "epoch": 1.6788230084933382,
      "grad_norm": 0.15689237415790558,
      "learning_rate": 4.404269021989144e-05,
      "loss": 0.6721,
      "mean_token_accuracy": 0.7898262403905392,
      "num_tokens": 582405942.0,
      "step": 18247
    },
    {
      "entropy": 0.5466506127268076,
      "epoch": 1.6789150148647793,
      "grad_norm": 0.15589193999767303,
      "learning_rate": 4.403962339374981e-05,
      "loss": 0.5536,
      "mean_token_accuracy": 0.829013254493475,
      "num_tokens": 582438161.0,
      "step": 18248
    },
    {
      "entropy": 0.5835589193738997,
      "epoch": 1.6790070212362207,
      "grad_norm": 0.15281552076339722,
      "learning_rate": 4.4036556567608187e-05,
      "loss": 0.5723,
      "mean_token_accuracy": 0.8235187567770481,
      "num_tokens": 582470178.0,
      "step": 18249
    },
    {
      "entropy": 0.5593003570102155,
      "epoch": 1.6790990276076618,
      "grad_norm": 0.14092876017093658,
      "learning_rate": 4.403348974146656e-05,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8326295018196106,
      "num_tokens": 582502930.0,
      "step": 18250
    },
    {
      "entropy": 0.4998738616704941,
      "epoch": 1.679191033979103,
      "grad_norm": 0.14269669353961945,
      "learning_rate": 4.403042291532493e-05,
      "loss": 0.4901,
      "mean_token_accuracy": 0.8505158387124538,
      "num_tokens": 582534917.0,
      "step": 18251
    },
    {
      "entropy": 0.5924201812595129,
      "epoch": 1.6792830403505443,
      "grad_norm": 0.15922801196575165,
      "learning_rate": 4.402735608918331e-05,
      "loss": 0.5853,
      "mean_token_accuracy": 0.8208322897553444,
      "num_tokens": 582566086.0,
      "step": 18252
    },
    {
      "entropy": 0.7182062873616815,
      "epoch": 1.6793750467219855,
      "grad_norm": 0.16807107627391815,
      "learning_rate": 4.402428926304168e-05,
      "loss": 0.7225,
      "mean_token_accuracy": 0.7784571535885334,
      "num_tokens": 582597973.0,
      "step": 18253
    },
    {
      "entropy": 0.5472306162118912,
      "epoch": 1.6794670530934268,
      "grad_norm": 0.15369348227977753,
      "learning_rate": 4.402122243690005e-05,
      "loss": 0.5312,
      "mean_token_accuracy": 0.8339691385626793,
      "num_tokens": 582629669.0,
      "step": 18254
    },
    {
      "entropy": 0.6304202880710363,
      "epoch": 1.679559059464868,
      "grad_norm": 0.15553520619869232,
      "learning_rate": 4.401815561075843e-05,
      "loss": 0.6283,
      "mean_token_accuracy": 0.8045927062630653,
      "num_tokens": 582661614.0,
      "step": 18255
    },
    {
      "entropy": 0.6498059537261724,
      "epoch": 1.679651065836309,
      "grad_norm": 0.15957172214984894,
      "learning_rate": 4.40150887846168e-05,
      "loss": 0.6456,
      "mean_token_accuracy": 0.8022035583853722,
      "num_tokens": 582693710.0,
      "step": 18256
    },
    {
      "entropy": 0.7272587167099118,
      "epoch": 1.6797430722077504,
      "grad_norm": 0.16406698524951935,
      "learning_rate": 4.401202195847518e-05,
      "loss": 0.7045,
      "mean_token_accuracy": 0.7843319997191429,
      "num_tokens": 582725133.0,
      "step": 18257
    },
    {
      "entropy": 0.6008610380813479,
      "epoch": 1.6798350785791916,
      "grad_norm": 0.15474717319011688,
      "learning_rate": 4.400895513233355e-05,
      "loss": 0.595,
      "mean_token_accuracy": 0.8178685531020164,
      "num_tokens": 582757600.0,
      "step": 18258
    },
    {
      "entropy": 0.5705265374854207,
      "epoch": 1.679927084950633,
      "grad_norm": 0.15138965845108032,
      "learning_rate": 4.400588830619192e-05,
      "loss": 0.554,
      "mean_token_accuracy": 0.8272294886410236,
      "num_tokens": 582789961.0,
      "step": 18259
    },
    {
      "entropy": 0.5935626104474068,
      "epoch": 1.680019091322074,
      "grad_norm": 0.15297219157218933,
      "learning_rate": 4.40028214800503e-05,
      "loss": 0.5752,
      "mean_token_accuracy": 0.8238450698554516,
      "num_tokens": 582822122.0,
      "step": 18260
    },
    {
      "entropy": 0.5925865964964032,
      "epoch": 1.6801110976935152,
      "grad_norm": 0.1632986217737198,
      "learning_rate": 4.399975465390867e-05,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8201166987419128,
      "num_tokens": 582853897.0,
      "step": 18261
    },
    {
      "entropy": 0.6315636523067951,
      "epoch": 1.6802031040649565,
      "grad_norm": 0.1592751294374466,
      "learning_rate": 4.399668782776705e-05,
      "loss": 0.6338,
      "mean_token_accuracy": 0.8091042824089527,
      "num_tokens": 582886484.0,
      "step": 18262
    },
    {
      "entropy": 0.4936336176469922,
      "epoch": 1.6802951104363977,
      "grad_norm": 0.1443060338497162,
      "learning_rate": 4.399362100162542e-05,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8511234819889069,
      "num_tokens": 582918701.0,
      "step": 18263
    },
    {
      "entropy": 0.5895609520375729,
      "epoch": 1.680387116807839,
      "grad_norm": 0.15615904331207275,
      "learning_rate": 4.399055417548379e-05,
      "loss": 0.5708,
      "mean_token_accuracy": 0.8219513334333897,
      "num_tokens": 582950691.0,
      "step": 18264
    },
    {
      "entropy": 0.6326336562633514,
      "epoch": 1.6804791231792802,
      "grad_norm": 0.163932204246521,
      "learning_rate": 4.398748734934217e-05,
      "loss": 0.6066,
      "mean_token_accuracy": 0.8116262890398502,
      "num_tokens": 582981608.0,
      "step": 18265
    },
    {
      "entropy": 0.5117055848240852,
      "epoch": 1.6805711295507213,
      "grad_norm": 0.15072286128997803,
      "learning_rate": 4.398442052320054e-05,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8463119566440582,
      "num_tokens": 583013146.0,
      "step": 18266
    },
    {
      "entropy": 0.6691371183842421,
      "epoch": 1.6806631359221627,
      "grad_norm": 0.16155201196670532,
      "learning_rate": 4.398135369705892e-05,
      "loss": 0.6666,
      "mean_token_accuracy": 0.8008532002568245,
      "num_tokens": 583045286.0,
      "step": 18267
    },
    {
      "entropy": 0.6412241719663143,
      "epoch": 1.6807551422936038,
      "grad_norm": 0.1563841849565506,
      "learning_rate": 4.397828687091729e-05,
      "loss": 0.6338,
      "mean_token_accuracy": 0.8072704076766968,
      "num_tokens": 583077327.0,
      "step": 18268
    },
    {
      "entropy": 0.5691340127959847,
      "epoch": 1.6808471486650451,
      "grad_norm": 0.14912576973438263,
      "learning_rate": 4.397522004477566e-05,
      "loss": 0.5574,
      "mean_token_accuracy": 0.826173447072506,
      "num_tokens": 583109343.0,
      "step": 18269
    },
    {
      "entropy": 0.6575247570872307,
      "epoch": 1.6809391550364863,
      "grad_norm": 0.16124844551086426,
      "learning_rate": 4.397215321863404e-05,
      "loss": 0.6523,
      "mean_token_accuracy": 0.8031696751713753,
      "num_tokens": 583140725.0,
      "step": 18270
    },
    {
      "entropy": 0.5577293802052736,
      "epoch": 1.6810311614079274,
      "grad_norm": 0.14417271316051483,
      "learning_rate": 4.396908639249241e-05,
      "loss": 0.5445,
      "mean_token_accuracy": 0.8310305289924145,
      "num_tokens": 583172957.0,
      "step": 18271
    },
    {
      "entropy": 0.5321153635159135,
      "epoch": 1.6811231677793688,
      "grad_norm": 0.14985491335391998,
      "learning_rate": 4.396601956635078e-05,
      "loss": 0.5197,
      "mean_token_accuracy": 0.8387012109160423,
      "num_tokens": 583205401.0,
      "step": 18272
    },
    {
      "entropy": 0.49671847536228597,
      "epoch": 1.68121517415081,
      "grad_norm": 0.13970769941806793,
      "learning_rate": 4.3962952740209165e-05,
      "loss": 0.4761,
      "mean_token_accuracy": 0.849360428750515,
      "num_tokens": 583237387.0,
      "step": 18273
    },
    {
      "entropy": 0.6378919025883079,
      "epoch": 1.6813071805222513,
      "grad_norm": 0.15777648985385895,
      "learning_rate": 4.395988591406753e-05,
      "loss": 0.6401,
      "mean_token_accuracy": 0.8037179857492447,
      "num_tokens": 583269344.0,
      "step": 18274
    },
    {
      "entropy": 0.6251575611531734,
      "epoch": 1.6813991868936924,
      "grad_norm": 0.15621036291122437,
      "learning_rate": 4.395681908792591e-05,
      "loss": 0.5934,
      "mean_token_accuracy": 0.812733855098486,
      "num_tokens": 583299685.0,
      "step": 18275
    },
    {
      "entropy": 0.6423271540552378,
      "epoch": 1.6814911932651335,
      "grad_norm": 0.15845851600170135,
      "learning_rate": 4.395375226178428e-05,
      "loss": 0.627,
      "mean_token_accuracy": 0.8057023473083973,
      "num_tokens": 583331522.0,
      "step": 18276
    },
    {
      "entropy": 0.4602011423557997,
      "epoch": 1.6815831996365749,
      "grad_norm": 0.13917593657970428,
      "learning_rate": 4.3950685435642656e-05,
      "loss": 0.44,
      "mean_token_accuracy": 0.8661895841360092,
      "num_tokens": 583363951.0,
      "step": 18277
    },
    {
      "entropy": 0.5954477172344923,
      "epoch": 1.681675206008016,
      "grad_norm": 0.15405257046222687,
      "learning_rate": 4.394761860950103e-05,
      "loss": 0.5842,
      "mean_token_accuracy": 0.8210797682404518,
      "num_tokens": 583395890.0,
      "step": 18278
    },
    {
      "entropy": 0.5745592792518437,
      "epoch": 1.6817672123794574,
      "grad_norm": 0.15132783353328705,
      "learning_rate": 4.3944551783359406e-05,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8288263753056526,
      "num_tokens": 583427553.0,
      "step": 18279
    },
    {
      "entropy": 0.7170029766857624,
      "epoch": 1.6818592187508985,
      "grad_norm": 0.16537030041217804,
      "learning_rate": 4.3941484957217774e-05,
      "loss": 0.7087,
      "mean_token_accuracy": 0.7860399782657623,
      "num_tokens": 583458324.0,
      "step": 18280
    },
    {
      "entropy": 0.7062690667808056,
      "epoch": 1.6819512251223396,
      "grad_norm": 0.16503433883190155,
      "learning_rate": 4.3938418131076155e-05,
      "loss": 0.6989,
      "mean_token_accuracy": 0.7873138710856438,
      "num_tokens": 583489508.0,
      "step": 18281
    },
    {
      "entropy": 0.5471284883096814,
      "epoch": 1.682043231493781,
      "grad_norm": 0.15403759479522705,
      "learning_rate": 4.393535130493452e-05,
      "loss": 0.5417,
      "mean_token_accuracy": 0.8336870186030865,
      "num_tokens": 583521631.0,
      "step": 18282
    },
    {
      "entropy": 0.5740578379482031,
      "epoch": 1.6821352378652221,
      "grad_norm": 0.14987805485725403,
      "learning_rate": 4.3932284478792904e-05,
      "loss": 0.57,
      "mean_token_accuracy": 0.8232158832252026,
      "num_tokens": 583553713.0,
      "step": 18283
    },
    {
      "entropy": 0.7685333974659443,
      "epoch": 1.6822272442366635,
      "grad_norm": 0.17120720446109772,
      "learning_rate": 4.392921765265127e-05,
      "loss": 0.7601,
      "mean_token_accuracy": 0.7722236588597298,
      "num_tokens": 583584895.0,
      "step": 18284
    },
    {
      "entropy": 0.7526569627225399,
      "epoch": 1.6823192506081046,
      "grad_norm": 0.17047350108623505,
      "learning_rate": 4.392615082650965e-05,
      "loss": 0.7476,
      "mean_token_accuracy": 0.7726166360080242,
      "num_tokens": 583617060.0,
      "step": 18285
    },
    {
      "entropy": 0.6042210701853037,
      "epoch": 1.6824112569795457,
      "grad_norm": 0.15026697516441345,
      "learning_rate": 4.392308400036802e-05,
      "loss": 0.5954,
      "mean_token_accuracy": 0.8188386857509613,
      "num_tokens": 583649664.0,
      "step": 18286
    },
    {
      "entropy": 0.6391297755762935,
      "epoch": 1.682503263350987,
      "grad_norm": 0.1619480550289154,
      "learning_rate": 4.3920017174226396e-05,
      "loss": 0.6228,
      "mean_token_accuracy": 0.8074298053979874,
      "num_tokens": 583680745.0,
      "step": 18287
    },
    {
      "entropy": 0.6092376918531954,
      "epoch": 1.6825952697224282,
      "grad_norm": 0.15016788244247437,
      "learning_rate": 4.391695034808477e-05,
      "loss": 0.5892,
      "mean_token_accuracy": 0.8163487426936626,
      "num_tokens": 583712949.0,
      "step": 18288
    },
    {
      "entropy": 0.5476008728146553,
      "epoch": 1.6826872760938696,
      "grad_norm": 0.15065431594848633,
      "learning_rate": 4.3913883521943145e-05,
      "loss": 0.54,
      "mean_token_accuracy": 0.8318971358239651,
      "num_tokens": 583745295.0,
      "step": 18289
    },
    {
      "entropy": 0.6112810960039496,
      "epoch": 1.6827792824653107,
      "grad_norm": 0.1484098732471466,
      "learning_rate": 4.391081669580151e-05,
      "loss": 0.6057,
      "mean_token_accuracy": 0.8183838315308094,
      "num_tokens": 583777084.0,
      "step": 18290
    },
    {
      "entropy": 0.6011072508990765,
      "epoch": 1.6828712888367519,
      "grad_norm": 0.1544007509946823,
      "learning_rate": 4.3907749869659895e-05,
      "loss": 0.5857,
      "mean_token_accuracy": 0.819329958409071,
      "num_tokens": 583808916.0,
      "step": 18291
    },
    {
      "entropy": 0.6321235205978155,
      "epoch": 1.6829632952081932,
      "grad_norm": 0.15460488200187683,
      "learning_rate": 4.390468304351826e-05,
      "loss": 0.6194,
      "mean_token_accuracy": 0.8112559802830219,
      "num_tokens": 583841193.0,
      "step": 18292
    },
    {
      "entropy": 0.6367239933460951,
      "epoch": 1.6830553015796343,
      "grad_norm": 0.16074064373970032,
      "learning_rate": 4.390161621737664e-05,
      "loss": 0.6447,
      "mean_token_accuracy": 0.8064451105892658,
      "num_tokens": 583872760.0,
      "step": 18293
    },
    {
      "entropy": 0.5119185708463192,
      "epoch": 1.6831473079510757,
      "grad_norm": 0.14744208753108978,
      "learning_rate": 4.389854939123501e-05,
      "loss": 0.4914,
      "mean_token_accuracy": 0.8482880294322968,
      "num_tokens": 583905121.0,
      "step": 18294
    },
    {
      "entropy": 0.6375892637297511,
      "epoch": 1.6832393143225168,
      "grad_norm": 0.15220829844474792,
      "learning_rate": 4.3895482565093387e-05,
      "loss": 0.6229,
      "mean_token_accuracy": 0.8085787445306778,
      "num_tokens": 583937538.0,
      "step": 18295
    },
    {
      "entropy": 0.6199583364650607,
      "epoch": 1.683331320693958,
      "grad_norm": 0.15610116720199585,
      "learning_rate": 4.389241573895176e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8138264194130898,
      "num_tokens": 583970129.0,
      "step": 18296
    },
    {
      "entropy": 0.4992179302498698,
      "epoch": 1.6834233270653993,
      "grad_norm": 0.1439947485923767,
      "learning_rate": 4.3889348912810136e-05,
      "loss": 0.4912,
      "mean_token_accuracy": 0.8487752377986908,
      "num_tokens": 584002543.0,
      "step": 18297
    },
    {
      "entropy": 0.5793168600648642,
      "epoch": 1.6835153334368405,
      "grad_norm": 0.15667669475078583,
      "learning_rate": 4.3886282086668504e-05,
      "loss": 0.557,
      "mean_token_accuracy": 0.8265281170606613,
      "num_tokens": 584035110.0,
      "step": 18298
    },
    {
      "entropy": 0.601556277833879,
      "epoch": 1.6836073398082818,
      "grad_norm": 0.15349264442920685,
      "learning_rate": 4.3883215260526885e-05,
      "loss": 0.5964,
      "mean_token_accuracy": 0.8164326027035713,
      "num_tokens": 584067013.0,
      "step": 18299
    },
    {
      "entropy": 0.6413939548656344,
      "epoch": 1.683699346179723,
      "grad_norm": 0.16009339690208435,
      "learning_rate": 4.388014843438525e-05,
      "loss": 0.6312,
      "mean_token_accuracy": 0.8061821013689041,
      "num_tokens": 584098519.0,
      "step": 18300
    },
    {
      "entropy": 0.5757073601707816,
      "epoch": 1.683791352551164,
      "grad_norm": 0.15152613818645477,
      "learning_rate": 4.387708160824363e-05,
      "loss": 0.5688,
      "mean_token_accuracy": 0.8283414803445339,
      "num_tokens": 584130559.0,
      "step": 18301
    },
    {
      "entropy": 0.5974721587263048,
      "epoch": 1.6838833589226054,
      "grad_norm": 0.15097957849502563,
      "learning_rate": 4.3874014782102e-05,
      "loss": 0.5949,
      "mean_token_accuracy": 0.8194234818220139,
      "num_tokens": 584163054.0,
      "step": 18302
    },
    {
      "entropy": 0.5210330635309219,
      "epoch": 1.6839753652940466,
      "grad_norm": 0.1497763842344284,
      "learning_rate": 4.387094795596038e-05,
      "loss": 0.4987,
      "mean_token_accuracy": 0.8437016233801842,
      "num_tokens": 584194344.0,
      "step": 18303
    },
    {
      "entropy": 0.5612441888079047,
      "epoch": 1.684067371665488,
      "grad_norm": 0.1476142257452011,
      "learning_rate": 4.386788112981876e-05,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8332802392542362,
      "num_tokens": 584226590.0,
      "step": 18304
    },
    {
      "entropy": 0.519160927971825,
      "epoch": 1.684159378036929,
      "grad_norm": 0.15169471502304077,
      "learning_rate": 4.3864814303677126e-05,
      "loss": 0.5103,
      "mean_token_accuracy": 0.844407070428133,
      "num_tokens": 584258681.0,
      "step": 18305
    },
    {
      "entropy": 0.5941620543599129,
      "epoch": 1.6842513844083702,
      "grad_norm": 0.15446804463863373,
      "learning_rate": 4.38617474775355e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.818816177546978,
      "num_tokens": 584291448.0,
      "step": 18306
    },
    {
      "entropy": 0.5471654934808612,
      "epoch": 1.6843433907798115,
      "grad_norm": 0.14885641634464264,
      "learning_rate": 4.3858680651393876e-05,
      "loss": 0.5376,
      "mean_token_accuracy": 0.832906685769558,
      "num_tokens": 584323486.0,
      "step": 18307
    },
    {
      "entropy": 0.5983629170805216,
      "epoch": 1.6844353971512527,
      "grad_norm": 0.16028451919555664,
      "learning_rate": 4.385561382525225e-05,
      "loss": 0.5954,
      "mean_token_accuracy": 0.8162489272654057,
      "num_tokens": 584355382.0,
      "step": 18308
    },
    {
      "entropy": 0.5222775600850582,
      "epoch": 1.684527403522694,
      "grad_norm": 0.15474744141101837,
      "learning_rate": 4.3852546999110625e-05,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8413292728364468,
      "num_tokens": 584386640.0,
      "step": 18309
    },
    {
      "entropy": 0.6889529530890286,
      "epoch": 1.6846194098941352,
      "grad_norm": 0.16113592684268951,
      "learning_rate": 4.3849480172969e-05,
      "loss": 0.6862,
      "mean_token_accuracy": 0.7900679111480713,
      "num_tokens": 584418227.0,
      "step": 18310
    },
    {
      "entropy": 0.5714503861963749,
      "epoch": 1.6847114162655763,
      "grad_norm": 0.1657312512397766,
      "learning_rate": 4.384641334682737e-05,
      "loss": 0.5644,
      "mean_token_accuracy": 0.825586561113596,
      "num_tokens": 584449869.0,
      "step": 18311
    },
    {
      "entropy": 0.5139881763607264,
      "epoch": 1.6848034226370177,
      "grad_norm": 0.14523214101791382,
      "learning_rate": 4.384334652068575e-05,
      "loss": 0.4988,
      "mean_token_accuracy": 0.8422465771436691,
      "num_tokens": 584482000.0,
      "step": 18312
    },
    {
      "entropy": 0.655033765360713,
      "epoch": 1.6848954290084588,
      "grad_norm": 0.15457908809185028,
      "learning_rate": 4.384027969454412e-05,
      "loss": 0.6398,
      "mean_token_accuracy": 0.8062746934592724,
      "num_tokens": 584514504.0,
      "step": 18313
    },
    {
      "entropy": 0.6518444218672812,
      "epoch": 1.6849874353799001,
      "grad_norm": 0.15509840846061707,
      "learning_rate": 4.383721286840249e-05,
      "loss": 0.6404,
      "mean_token_accuracy": 0.8062356896698475,
      "num_tokens": 584546557.0,
      "step": 18314
    },
    {
      "entropy": 0.6006749346852303,
      "epoch": 1.6850794417513413,
      "grad_norm": 0.15732136368751526,
      "learning_rate": 4.3834146042260866e-05,
      "loss": 0.5907,
      "mean_token_accuracy": 0.8184856958687305,
      "num_tokens": 584578411.0,
      "step": 18315
    },
    {
      "entropy": 0.6034524233546108,
      "epoch": 1.6851714481227824,
      "grad_norm": 0.15009263157844543,
      "learning_rate": 4.383107921611924e-05,
      "loss": 0.5807,
      "mean_token_accuracy": 0.825228612869978,
      "num_tokens": 584610412.0,
      "step": 18316
    },
    {
      "entropy": 0.5757454675622284,
      "epoch": 1.6852634544942238,
      "grad_norm": 0.15307874977588654,
      "learning_rate": 4.3828012389977615e-05,
      "loss": 0.5686,
      "mean_token_accuracy": 0.8266474008560181,
      "num_tokens": 584642490.0,
      "step": 18317
    },
    {
      "entropy": 0.6078398479148746,
      "epoch": 1.685355460865665,
      "grad_norm": 0.1531205177307129,
      "learning_rate": 4.382494556383599e-05,
      "loss": 0.5937,
      "mean_token_accuracy": 0.8183726631104946,
      "num_tokens": 584673979.0,
      "step": 18318
    },
    {
      "entropy": 0.7107502594590187,
      "epoch": 1.6854474672371063,
      "grad_norm": 0.1601436585187912,
      "learning_rate": 4.382187873769436e-05,
      "loss": 0.6913,
      "mean_token_accuracy": 0.7891173101961613,
      "num_tokens": 584706361.0,
      "step": 18319
    },
    {
      "entropy": 0.6528557203710079,
      "epoch": 1.6855394736085474,
      "grad_norm": 0.16151650249958038,
      "learning_rate": 4.381881191155274e-05,
      "loss": 0.6606,
      "mean_token_accuracy": 0.8015039823949337,
      "num_tokens": 584737782.0,
      "step": 18320
    },
    {
      "entropy": 0.6800206545740366,
      "epoch": 1.6856314799799885,
      "grad_norm": 0.16777303814888,
      "learning_rate": 4.381574508541111e-05,
      "loss": 0.6835,
      "mean_token_accuracy": 0.7934298887848854,
      "num_tokens": 584769020.0,
      "step": 18321
    },
    {
      "entropy": 0.5079681696370244,
      "epoch": 1.6857234863514299,
      "grad_norm": 0.14851714670658112,
      "learning_rate": 4.381267825926949e-05,
      "loss": 0.5063,
      "mean_token_accuracy": 0.8447441384196281,
      "num_tokens": 584801011.0,
      "step": 18322
    },
    {
      "entropy": 0.57472974807024,
      "epoch": 1.685815492722871,
      "grad_norm": 0.15923964977264404,
      "learning_rate": 4.3809611433127856e-05,
      "loss": 0.5824,
      "mean_token_accuracy": 0.8190155662596226,
      "num_tokens": 584833020.0,
      "step": 18323
    },
    {
      "entropy": 0.6127247249241918,
      "epoch": 1.6859074990943124,
      "grad_norm": 0.14891518652439117,
      "learning_rate": 4.380654460698623e-05,
      "loss": 0.5918,
      "mean_token_accuracy": 0.8211951889097691,
      "num_tokens": 584864372.0,
      "step": 18324
    },
    {
      "entropy": 0.73013586550951,
      "epoch": 1.6859995054657535,
      "grad_norm": 0.16220633685588837,
      "learning_rate": 4.3803477780844606e-05,
      "loss": 0.7198,
      "mean_token_accuracy": 0.785042367875576,
      "num_tokens": 584896390.0,
      "step": 18325
    },
    {
      "entropy": 0.575688480399549,
      "epoch": 1.6860915118371946,
      "grad_norm": 0.15480881929397583,
      "learning_rate": 4.380041095470298e-05,
      "loss": 0.5652,
      "mean_token_accuracy": 0.8273265101015568,
      "num_tokens": 584928208.0,
      "step": 18326
    },
    {
      "entropy": 0.7508062049746513,
      "epoch": 1.686183518208636,
      "grad_norm": 0.16643026471138,
      "learning_rate": 4.3797344128561355e-05,
      "loss": 0.7464,
      "mean_token_accuracy": 0.774582851678133,
      "num_tokens": 584959847.0,
      "step": 18327
    },
    {
      "entropy": 0.5757310874760151,
      "epoch": 1.6862755245800771,
      "grad_norm": 0.14998385310173035,
      "learning_rate": 4.379427730241973e-05,
      "loss": 0.5734,
      "mean_token_accuracy": 0.8239204026758671,
      "num_tokens": 584992177.0,
      "step": 18328
    },
    {
      "entropy": 0.6297884713858366,
      "epoch": 1.6863675309515185,
      "grad_norm": 0.1513999104499817,
      "learning_rate": 4.37912104762781e-05,
      "loss": 0.6201,
      "mean_token_accuracy": 0.8114747516810894,
      "num_tokens": 585024380.0,
      "step": 18329
    },
    {
      "entropy": 0.5453423755243421,
      "epoch": 1.6864595373229596,
      "grad_norm": 0.15467792749404907,
      "learning_rate": 4.378814365013648e-05,
      "loss": 0.5316,
      "mean_token_accuracy": 0.8363192118704319,
      "num_tokens": 585055853.0,
      "step": 18330
    },
    {
      "entropy": 0.6046038493514061,
      "epoch": 1.6865515436944007,
      "grad_norm": 0.15358959138393402,
      "learning_rate": 4.378507682399485e-05,
      "loss": 0.5952,
      "mean_token_accuracy": 0.8151990100741386,
      "num_tokens": 585087725.0,
      "step": 18331
    },
    {
      "entropy": 0.6944432631134987,
      "epoch": 1.686643550065842,
      "grad_norm": 0.1537497490644455,
      "learning_rate": 4.378200999785322e-05,
      "loss": 0.6799,
      "mean_token_accuracy": 0.7913361378014088,
      "num_tokens": 585120238.0,
      "step": 18332
    },
    {
      "entropy": 0.5395734841004014,
      "epoch": 1.6867355564372832,
      "grad_norm": 0.15236221253871918,
      "learning_rate": 4.37789431717116e-05,
      "loss": 0.5252,
      "mean_token_accuracy": 0.8356862999498844,
      "num_tokens": 585152592.0,
      "step": 18333
    },
    {
      "entropy": 0.49862429965287447,
      "epoch": 1.6868275628087246,
      "grad_norm": 0.15070192515850067,
      "learning_rate": 4.377587634556997e-05,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8463908843696117,
      "num_tokens": 585184624.0,
      "step": 18334
    },
    {
      "entropy": 0.5457268785685301,
      "epoch": 1.6869195691801657,
      "grad_norm": 0.14741499722003937,
      "learning_rate": 4.3772809519428345e-05,
      "loss": 0.5426,
      "mean_token_accuracy": 0.8321007415652275,
      "num_tokens": 585216695.0,
      "step": 18335
    },
    {
      "entropy": 0.6530294492840767,
      "epoch": 1.6870115755516069,
      "grad_norm": 0.17048956453800201,
      "learning_rate": 4.376974269328672e-05,
      "loss": 0.6519,
      "mean_token_accuracy": 0.8013546727597713,
      "num_tokens": 585248382.0,
      "step": 18336
    },
    {
      "entropy": 0.5888009136542678,
      "epoch": 1.6871035819230482,
      "grad_norm": 0.15109354257583618,
      "learning_rate": 4.3766675867145095e-05,
      "loss": 0.5701,
      "mean_token_accuracy": 0.82380136474967,
      "num_tokens": 585280212.0,
      "step": 18337
    },
    {
      "entropy": 0.5153891621157527,
      "epoch": 1.6871955882944893,
      "grad_norm": 0.15275326371192932,
      "learning_rate": 4.376360904100347e-05,
      "loss": 0.5109,
      "mean_token_accuracy": 0.8431210219860077,
      "num_tokens": 585312314.0,
      "step": 18338
    },
    {
      "entropy": 0.6511192582547665,
      "epoch": 1.6872875946659307,
      "grad_norm": 0.1569906622171402,
      "learning_rate": 4.3760542214861844e-05,
      "loss": 0.6342,
      "mean_token_accuracy": 0.8046337887644768,
      "num_tokens": 585344445.0,
      "step": 18339
    },
    {
      "entropy": 0.5337929464876652,
      "epoch": 1.6873796010373718,
      "grad_norm": 0.15637393295764923,
      "learning_rate": 4.375747538872021e-05,
      "loss": 0.521,
      "mean_token_accuracy": 0.8350762277841568,
      "num_tokens": 585376171.0,
      "step": 18340
    },
    {
      "entropy": 0.5759244533255696,
      "epoch": 1.687471607408813,
      "grad_norm": 0.148925319314003,
      "learning_rate": 4.375440856257859e-05,
      "loss": 0.5659,
      "mean_token_accuracy": 0.8275878950953484,
      "num_tokens": 585407631.0,
      "step": 18341
    },
    {
      "entropy": 0.5986175499856472,
      "epoch": 1.6875636137802543,
      "grad_norm": 0.15303681790828705,
      "learning_rate": 4.375134173643696e-05,
      "loss": 0.5896,
      "mean_token_accuracy": 0.8202645145356655,
      "num_tokens": 585439228.0,
      "step": 18342
    },
    {
      "entropy": 0.5354221109300852,
      "epoch": 1.6876556201516955,
      "grad_norm": 0.15531586110591888,
      "learning_rate": 4.374827491029534e-05,
      "loss": 0.5105,
      "mean_token_accuracy": 0.8396165519952774,
      "num_tokens": 585470570.0,
      "step": 18343
    },
    {
      "entropy": 0.6285147909075022,
      "epoch": 1.6877476265231368,
      "grad_norm": 0.152794748544693,
      "learning_rate": 4.374520808415371e-05,
      "loss": 0.6165,
      "mean_token_accuracy": 0.8101099766790867,
      "num_tokens": 585502679.0,
      "step": 18344
    },
    {
      "entropy": 0.6263280604034662,
      "epoch": 1.687839632894578,
      "grad_norm": 0.16394193470478058,
      "learning_rate": 4.3742141258012085e-05,
      "loss": 0.6187,
      "mean_token_accuracy": 0.81076904758811,
      "num_tokens": 585535092.0,
      "step": 18345
    },
    {
      "entropy": 0.6694249017164111,
      "epoch": 1.687931639266019,
      "grad_norm": 0.1593048870563507,
      "learning_rate": 4.373907443187046e-05,
      "loss": 0.6575,
      "mean_token_accuracy": 0.7962511368095875,
      "num_tokens": 585566734.0,
      "step": 18346
    },
    {
      "entropy": 0.5362485023215413,
      "epoch": 1.6880236456374604,
      "grad_norm": 0.14829587936401367,
      "learning_rate": 4.3736007605728834e-05,
      "loss": 0.5241,
      "mean_token_accuracy": 0.84070248529315,
      "num_tokens": 585598223.0,
      "step": 18347
    },
    {
      "entropy": 0.5792909442679957,
      "epoch": 1.6881156520089016,
      "grad_norm": 0.15421223640441895,
      "learning_rate": 4.373294077958721e-05,
      "loss": 0.5729,
      "mean_token_accuracy": 0.8251766301691532,
      "num_tokens": 585630064.0,
      "step": 18348
    },
    {
      "entropy": 0.6168117495253682,
      "epoch": 1.688207658380343,
      "grad_norm": 0.15360327064990997,
      "learning_rate": 4.3729873953445584e-05,
      "loss": 0.6101,
      "mean_token_accuracy": 0.8112287819385529,
      "num_tokens": 585661954.0,
      "step": 18349
    },
    {
      "entropy": 0.6474901409819722,
      "epoch": 1.688299664751784,
      "grad_norm": 0.16032522916793823,
      "learning_rate": 4.372680712730395e-05,
      "loss": 0.6363,
      "mean_token_accuracy": 0.8018609620630741,
      "num_tokens": 585693967.0,
      "step": 18350
    },
    {
      "entropy": 0.6707997247576714,
      "epoch": 1.6883916711232252,
      "grad_norm": 0.1575835794210434,
      "learning_rate": 4.372374030116233e-05,
      "loss": 0.6726,
      "mean_token_accuracy": 0.7935571111738682,
      "num_tokens": 585725761.0,
      "step": 18351
    },
    {
      "entropy": 0.5980529058724642,
      "epoch": 1.6884836774946665,
      "grad_norm": 0.15206734836101532,
      "learning_rate": 4.37206734750207e-05,
      "loss": 0.5932,
      "mean_token_accuracy": 0.8165548592805862,
      "num_tokens": 585758039.0,
      "step": 18352
    },
    {
      "entropy": 0.5833911467343569,
      "epoch": 1.6885756838661077,
      "grad_norm": 0.15189722180366516,
      "learning_rate": 4.3717606648879075e-05,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8216124214231968,
      "num_tokens": 585790639.0,
      "step": 18353
    },
    {
      "entropy": 0.6696184687316418,
      "epoch": 1.688667690237549,
      "grad_norm": 0.15731579065322876,
      "learning_rate": 4.371453982273745e-05,
      "loss": 0.6607,
      "mean_token_accuracy": 0.7992384769022465,
      "num_tokens": 585822096.0,
      "step": 18354
    },
    {
      "entropy": 0.5738721955567598,
      "epoch": 1.6887596966089902,
      "grad_norm": 0.15357784926891327,
      "learning_rate": 4.3711472996595825e-05,
      "loss": 0.5601,
      "mean_token_accuracy": 0.8294500485062599,
      "num_tokens": 585853621.0,
      "step": 18355
    },
    {
      "entropy": 0.5377235533669591,
      "epoch": 1.6888517029804313,
      "grad_norm": 0.15369875729084015,
      "learning_rate": 4.37084061704542e-05,
      "loss": 0.5345,
      "mean_token_accuracy": 0.8379846885800362,
      "num_tokens": 585885594.0,
      "step": 18356
    },
    {
      "entropy": 0.4944965159520507,
      "epoch": 1.6889437093518727,
      "grad_norm": 0.14146895706653595,
      "learning_rate": 4.3705339344312574e-05,
      "loss": 0.4788,
      "mean_token_accuracy": 0.8510550335049629,
      "num_tokens": 585917800.0,
      "step": 18357
    },
    {
      "entropy": 0.6811487227678299,
      "epoch": 1.6890357157233138,
      "grad_norm": 0.15921780467033386,
      "learning_rate": 4.370227251817094e-05,
      "loss": 0.6776,
      "mean_token_accuracy": 0.7943317741155624,
      "num_tokens": 585949928.0,
      "step": 18358
    },
    {
      "entropy": 0.5781120108440518,
      "epoch": 1.6891277220947551,
      "grad_norm": 0.15565474331378937,
      "learning_rate": 4.369920569202932e-05,
      "loss": 0.5666,
      "mean_token_accuracy": 0.8235251791775227,
      "num_tokens": 585982081.0,
      "step": 18359
    },
    {
      "entropy": 0.6697069802321494,
      "epoch": 1.6892197284661963,
      "grad_norm": 0.1592462658882141,
      "learning_rate": 4.369613886588769e-05,
      "loss": 0.6657,
      "mean_token_accuracy": 0.8005416989326477,
      "num_tokens": 586014245.0,
      "step": 18360
    },
    {
      "entropy": 0.5617673862725496,
      "epoch": 1.6893117348376374,
      "grad_norm": 0.15338559448719025,
      "learning_rate": 4.3693072039746066e-05,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8243122324347496,
      "num_tokens": 586045598.0,
      "step": 18361
    },
    {
      "entropy": 0.5660423915833235,
      "epoch": 1.6894037412090788,
      "grad_norm": 0.1470344215631485,
      "learning_rate": 4.369000521360444e-05,
      "loss": 0.5513,
      "mean_token_accuracy": 0.8299991637468338,
      "num_tokens": 586077983.0,
      "step": 18362
    },
    {
      "entropy": 0.66982202231884,
      "epoch": 1.68949574758052,
      "grad_norm": 0.16106751561164856,
      "learning_rate": 4.3686938387462815e-05,
      "loss": 0.6542,
      "mean_token_accuracy": 0.7939707301557064,
      "num_tokens": 586109307.0,
      "step": 18363
    },
    {
      "entropy": 0.680996872484684,
      "epoch": 1.6895877539519613,
      "grad_norm": 0.1608162671327591,
      "learning_rate": 4.3683871561321197e-05,
      "loss": 0.6858,
      "mean_token_accuracy": 0.7931499406695366,
      "num_tokens": 586141620.0,
      "step": 18364
    },
    {
      "entropy": 0.6363926809281111,
      "epoch": 1.6896797603234024,
      "grad_norm": 0.1552644670009613,
      "learning_rate": 4.3680804735179564e-05,
      "loss": 0.6183,
      "mean_token_accuracy": 0.8072918355464935,
      "num_tokens": 586174162.0,
      "step": 18365
    },
    {
      "entropy": 0.6401008823886514,
      "epoch": 1.6897717666948435,
      "grad_norm": 0.15618202090263367,
      "learning_rate": 4.367773790903794e-05,
      "loss": 0.6311,
      "mean_token_accuracy": 0.8076570741832256,
      "num_tokens": 586205483.0,
      "step": 18366
    },
    {
      "entropy": 0.5848010275512934,
      "epoch": 1.6898637730662849,
      "grad_norm": 0.150425985455513,
      "learning_rate": 4.3674671082896314e-05,
      "loss": 0.5733,
      "mean_token_accuracy": 0.8251743763685226,
      "num_tokens": 586236673.0,
      "step": 18367
    },
    {
      "entropy": 0.496879227925092,
      "epoch": 1.689955779437726,
      "grad_norm": 0.1619945466518402,
      "learning_rate": 4.367160425675469e-05,
      "loss": 0.4811,
      "mean_token_accuracy": 0.8517354018986225,
      "num_tokens": 586268266.0,
      "step": 18368
    },
    {
      "entropy": 0.5724665804300457,
      "epoch": 1.6900477858091674,
      "grad_norm": 0.15558943152427673,
      "learning_rate": 4.366853743061306e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8291406109929085,
      "num_tokens": 586300633.0,
      "step": 18369
    },
    {
      "entropy": 0.6353645231574774,
      "epoch": 1.6901397921806085,
      "grad_norm": 0.1578582525253296,
      "learning_rate": 4.366547060447144e-05,
      "loss": 0.6297,
      "mean_token_accuracy": 0.8067265041172504,
      "num_tokens": 586332626.0,
      "step": 18370
    },
    {
      "entropy": 0.4869861928746104,
      "epoch": 1.6902317985520496,
      "grad_norm": 0.14623942971229553,
      "learning_rate": 4.3662403778329806e-05,
      "loss": 0.4738,
      "mean_token_accuracy": 0.8531530350446701,
      "num_tokens": 586364340.0,
      "step": 18371
    },
    {
      "entropy": 0.5804362567141652,
      "epoch": 1.690323804923491,
      "grad_norm": 0.1545773446559906,
      "learning_rate": 4.365933695218819e-05,
      "loss": 0.57,
      "mean_token_accuracy": 0.8247597478330135,
      "num_tokens": 586395488.0,
      "step": 18372
    },
    {
      "entropy": 0.5574918519705534,
      "epoch": 1.6904158112949321,
      "grad_norm": 0.14538265764713287,
      "learning_rate": 4.3656270126046555e-05,
      "loss": 0.5567,
      "mean_token_accuracy": 0.8286291025578976,
      "num_tokens": 586427889.0,
      "step": 18373
    },
    {
      "entropy": 0.5713267344981432,
      "epoch": 1.6905078176663735,
      "grad_norm": 0.14691567420959473,
      "learning_rate": 4.365320329990493e-05,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8276946470141411,
      "num_tokens": 586459954.0,
      "step": 18374
    },
    {
      "entropy": 0.5531822505872697,
      "epoch": 1.6905998240378146,
      "grad_norm": 0.15175823867321014,
      "learning_rate": 4.3650136473763304e-05,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8330871909856796,
      "num_tokens": 586492058.0,
      "step": 18375
    },
    {
      "entropy": 0.5494305107276887,
      "epoch": 1.6906918304092557,
      "grad_norm": 0.14425188302993774,
      "learning_rate": 4.364706964762168e-05,
      "loss": 0.532,
      "mean_token_accuracy": 0.8362694159150124,
      "num_tokens": 586524435.0,
      "step": 18376
    },
    {
      "entropy": 0.5848115272819996,
      "epoch": 1.690783836780697,
      "grad_norm": 0.15116959810256958,
      "learning_rate": 4.3644002821480053e-05,
      "loss": 0.5606,
      "mean_token_accuracy": 0.8232622034847736,
      "num_tokens": 586556452.0,
      "step": 18377
    },
    {
      "entropy": 0.5582094769924879,
      "epoch": 1.6908758431521382,
      "grad_norm": 0.15399520099163055,
      "learning_rate": 4.364093599533843e-05,
      "loss": 0.5626,
      "mean_token_accuracy": 0.8304580450057983,
      "num_tokens": 586588782.0,
      "step": 18378
    },
    {
      "entropy": 0.5841601341962814,
      "epoch": 1.6909678495235796,
      "grad_norm": 0.14688391983509064,
      "learning_rate": 4.3637869169196796e-05,
      "loss": 0.5652,
      "mean_token_accuracy": 0.8248622193932533,
      "num_tokens": 586621012.0,
      "step": 18379
    },
    {
      "entropy": 0.578871856443584,
      "epoch": 1.6910598558950207,
      "grad_norm": 0.15142886340618134,
      "learning_rate": 4.363480234305518e-05,
      "loss": 0.581,
      "mean_token_accuracy": 0.8292490802705288,
      "num_tokens": 586652767.0,
      "step": 18380
    },
    {
      "entropy": 0.6507391575723886,
      "epoch": 1.6911518622664619,
      "grad_norm": 0.1602127104997635,
      "learning_rate": 4.3631735516913545e-05,
      "loss": 0.6424,
      "mean_token_accuracy": 0.8013669028878212,
      "num_tokens": 586684230.0,
      "step": 18381
    },
    {
      "entropy": 0.6583095034584403,
      "epoch": 1.6912438686379032,
      "grad_norm": 0.16246303915977478,
      "learning_rate": 4.362866869077193e-05,
      "loss": 0.6612,
      "mean_token_accuracy": 0.80109428986907,
      "num_tokens": 586716885.0,
      "step": 18382
    },
    {
      "entropy": 0.5815250128507614,
      "epoch": 1.6913358750093443,
      "grad_norm": 0.15033847093582153,
      "learning_rate": 4.3625601864630295e-05,
      "loss": 0.5787,
      "mean_token_accuracy": 0.8255843371152878,
      "num_tokens": 586748999.0,
      "step": 18383
    },
    {
      "entropy": 0.6294097285717726,
      "epoch": 1.6914278813807857,
      "grad_norm": 0.15490396320819855,
      "learning_rate": 4.362253503848867e-05,
      "loss": 0.6225,
      "mean_token_accuracy": 0.808561310172081,
      "num_tokens": 586781767.0,
      "step": 18384
    },
    {
      "entropy": 0.5642022807151079,
      "epoch": 1.6915198877522268,
      "grad_norm": 0.1569616049528122,
      "learning_rate": 4.3619468212347044e-05,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8355674035847187,
      "num_tokens": 586813555.0,
      "step": 18385
    },
    {
      "entropy": 0.6040282100439072,
      "epoch": 1.691611894123668,
      "grad_norm": 0.1614800989627838,
      "learning_rate": 4.361640138620542e-05,
      "loss": 0.5945,
      "mean_token_accuracy": 0.8197048678994179,
      "num_tokens": 586845150.0,
      "step": 18386
    },
    {
      "entropy": 0.6229828465729952,
      "epoch": 1.6917039004951093,
      "grad_norm": 0.1567278653383255,
      "learning_rate": 4.361333456006379e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.8172919228672981,
      "num_tokens": 586876907.0,
      "step": 18387
    },
    {
      "entropy": 0.6115531902760267,
      "epoch": 1.6917959068665505,
      "grad_norm": 0.15673989057540894,
      "learning_rate": 4.361026773392217e-05,
      "loss": 0.6034,
      "mean_token_accuracy": 0.8194367326796055,
      "num_tokens": 586909229.0,
      "step": 18388
    },
    {
      "entropy": 0.5472872843965888,
      "epoch": 1.6918879132379918,
      "grad_norm": 0.14979049563407898,
      "learning_rate": 4.3607200907780536e-05,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8339706994593143,
      "num_tokens": 586941446.0,
      "step": 18389
    },
    {
      "entropy": 0.7082349676638842,
      "epoch": 1.691979919609433,
      "grad_norm": 0.16012506186962128,
      "learning_rate": 4.360413408163892e-05,
      "loss": 0.7063,
      "mean_token_accuracy": 0.7861600704491138,
      "num_tokens": 586973604.0,
      "step": 18390
    },
    {
      "entropy": 0.6296662129461765,
      "epoch": 1.692071925980874,
      "grad_norm": 0.15730193257331848,
      "learning_rate": 4.3601067255497285e-05,
      "loss": 0.6178,
      "mean_token_accuracy": 0.8103850409388542,
      "num_tokens": 587005464.0,
      "step": 18391
    },
    {
      "entropy": 0.6306332405656576,
      "epoch": 1.6921639323523154,
      "grad_norm": 0.1607389897108078,
      "learning_rate": 4.359800042935566e-05,
      "loss": 0.6241,
      "mean_token_accuracy": 0.8103515096008778,
      "num_tokens": 587037521.0,
      "step": 18392
    },
    {
      "entropy": 0.5434005912393332,
      "epoch": 1.6922559387237566,
      "grad_norm": 0.14828315377235413,
      "learning_rate": 4.3594933603214034e-05,
      "loss": 0.5344,
      "mean_token_accuracy": 0.8358854427933693,
      "num_tokens": 587069160.0,
      "step": 18393
    },
    {
      "entropy": 0.6049639340490103,
      "epoch": 1.692347945095198,
      "grad_norm": 0.15310348570346832,
      "learning_rate": 4.359186677707241e-05,
      "loss": 0.5945,
      "mean_token_accuracy": 0.8156154528260231,
      "num_tokens": 587101252.0,
      "step": 18394
    },
    {
      "entropy": 0.665145848877728,
      "epoch": 1.692439951466639,
      "grad_norm": 0.15921828150749207,
      "learning_rate": 4.3588799950930784e-05,
      "loss": 0.6527,
      "mean_token_accuracy": 0.7980097457766533,
      "num_tokens": 587133459.0,
      "step": 18395
    },
    {
      "entropy": 0.583815343445167,
      "epoch": 1.6925319578380802,
      "grad_norm": 0.15307298302650452,
      "learning_rate": 4.358573312478916e-05,
      "loss": 0.5659,
      "mean_token_accuracy": 0.8233412951231003,
      "num_tokens": 587164253.0,
      "step": 18396
    },
    {
      "entropy": 0.6639682874083519,
      "epoch": 1.6926239642095215,
      "grad_norm": 0.15856249630451202,
      "learning_rate": 4.358266629864753e-05,
      "loss": 0.6661,
      "mean_token_accuracy": 0.7969589456915855,
      "num_tokens": 587196662.0,
      "step": 18397
    },
    {
      "entropy": 0.5776577852666378,
      "epoch": 1.6927159705809627,
      "grad_norm": 0.15907840430736542,
      "learning_rate": 4.357959947250591e-05,
      "loss": 0.5736,
      "mean_token_accuracy": 0.8266055695712566,
      "num_tokens": 587229112.0,
      "step": 18398
    },
    {
      "entropy": 0.5593933714553714,
      "epoch": 1.692807976952404,
      "grad_norm": 0.1581360548734665,
      "learning_rate": 4.357653264636428e-05,
      "loss": 0.545,
      "mean_token_accuracy": 0.8325214572250843,
      "num_tokens": 587261364.0,
      "step": 18399
    },
    {
      "entropy": 0.49413512740284204,
      "epoch": 1.6928999833238452,
      "grad_norm": 0.14407430589199066,
      "learning_rate": 4.357346582022265e-05,
      "loss": 0.4722,
      "mean_token_accuracy": 0.8545210957527161,
      "num_tokens": 587293072.0,
      "step": 18400
    },
    {
      "entropy": 0.6291030403226614,
      "epoch": 1.6929919896952863,
      "grad_norm": 0.15613949298858643,
      "learning_rate": 4.357039899408103e-05,
      "loss": 0.6223,
      "mean_token_accuracy": 0.8134948536753654,
      "num_tokens": 587325119.0,
      "step": 18401
    },
    {
      "entropy": 0.6347861588001251,
      "epoch": 1.6930839960667277,
      "grad_norm": 0.15487341582775116,
      "learning_rate": 4.35673321679394e-05,
      "loss": 0.6253,
      "mean_token_accuracy": 0.8049218915402889,
      "num_tokens": 587356689.0,
      "step": 18402
    },
    {
      "entropy": 0.5674461582675576,
      "epoch": 1.6931760024381688,
      "grad_norm": 0.14322669804096222,
      "learning_rate": 4.356426534179778e-05,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8324043303728104,
      "num_tokens": 587389138.0,
      "step": 18403
    },
    {
      "entropy": 0.634026349755004,
      "epoch": 1.6932680088096101,
      "grad_norm": 0.15816208720207214,
      "learning_rate": 4.356119851565615e-05,
      "loss": 0.6158,
      "mean_token_accuracy": 0.809504572302103,
      "num_tokens": 587420010.0,
      "step": 18404
    },
    {
      "entropy": 0.5994433192536235,
      "epoch": 1.6933600151810513,
      "grad_norm": 0.15612775087356567,
      "learning_rate": 4.355813168951452e-05,
      "loss": 0.5719,
      "mean_token_accuracy": 0.8234921470284462,
      "num_tokens": 587451166.0,
      "step": 18405
    },
    {
      "entropy": 0.6443960107862949,
      "epoch": 1.6934520215524924,
      "grad_norm": 0.16582418978214264,
      "learning_rate": 4.35550648633729e-05,
      "loss": 0.6347,
      "mean_token_accuracy": 0.805833961814642,
      "num_tokens": 587481521.0,
      "step": 18406
    },
    {
      "entropy": 0.5641657039523125,
      "epoch": 1.6935440279239338,
      "grad_norm": 0.15143322944641113,
      "learning_rate": 4.355199803723127e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8289021477103233,
      "num_tokens": 587513942.0,
      "step": 18407
    },
    {
      "entropy": 0.6654325239360332,
      "epoch": 1.693636034295375,
      "grad_norm": 0.15921102464199066,
      "learning_rate": 4.354893121108965e-05,
      "loss": 0.6434,
      "mean_token_accuracy": 0.8031970635056496,
      "num_tokens": 587546213.0,
      "step": 18408
    },
    {
      "entropy": 0.5989088404458016,
      "epoch": 1.6937280406668163,
      "grad_norm": 0.15234315395355225,
      "learning_rate": 4.354586438494802e-05,
      "loss": 0.5846,
      "mean_token_accuracy": 0.8217759728431702,
      "num_tokens": 587578226.0,
      "step": 18409
    },
    {
      "entropy": 0.6387266777455807,
      "epoch": 1.6938200470382574,
      "grad_norm": 0.15533053874969482,
      "learning_rate": 4.354279755880639e-05,
      "loss": 0.6396,
      "mean_token_accuracy": 0.8051913902163506,
      "num_tokens": 587610350.0,
      "step": 18410
    },
    {
      "entropy": 0.6493906239047647,
      "epoch": 1.6939120534096985,
      "grad_norm": 0.15942718088626862,
      "learning_rate": 4.353973073266477e-05,
      "loss": 0.6509,
      "mean_token_accuracy": 0.8063803054392338,
      "num_tokens": 587642239.0,
      "step": 18411
    },
    {
      "entropy": 0.6510259546339512,
      "epoch": 1.6940040597811399,
      "grad_norm": 0.15967486798763275,
      "learning_rate": 4.353666390652314e-05,
      "loss": 0.6297,
      "mean_token_accuracy": 0.8074197545647621,
      "num_tokens": 587674552.0,
      "step": 18412
    },
    {
      "entropy": 0.6084071034565568,
      "epoch": 1.694096066152581,
      "grad_norm": 0.1570311039686203,
      "learning_rate": 4.3533597080381514e-05,
      "loss": 0.6117,
      "mean_token_accuracy": 0.8118571825325489,
      "num_tokens": 587706790.0,
      "step": 18413
    },
    {
      "entropy": 0.5552931614220142,
      "epoch": 1.6941880725240224,
      "grad_norm": 0.15197858214378357,
      "learning_rate": 4.353053025423989e-05,
      "loss": 0.5469,
      "mean_token_accuracy": 0.8278703987598419,
      "num_tokens": 587738014.0,
      "step": 18414
    },
    {
      "entropy": 0.6030499832704663,
      "epoch": 1.6942800788954635,
      "grad_norm": 0.15615518391132355,
      "learning_rate": 4.352746342809826e-05,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8196456097066402,
      "num_tokens": 587769437.0,
      "step": 18415
    },
    {
      "entropy": 0.5783545486629009,
      "epoch": 1.6943720852669046,
      "grad_norm": 0.1625809222459793,
      "learning_rate": 4.352439660195664e-05,
      "loss": 0.5554,
      "mean_token_accuracy": 0.8267388604581356,
      "num_tokens": 587799835.0,
      "step": 18416
    },
    {
      "entropy": 0.6507317870855331,
      "epoch": 1.694464091638346,
      "grad_norm": 0.15909388661384583,
      "learning_rate": 4.352132977581501e-05,
      "loss": 0.6493,
      "mean_token_accuracy": 0.8021978214383125,
      "num_tokens": 587832299.0,
      "step": 18417
    },
    {
      "entropy": 0.5031209867447615,
      "epoch": 1.6945560980097871,
      "grad_norm": 0.13889966905117035,
      "learning_rate": 4.351826294967338e-05,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8457944914698601,
      "num_tokens": 587864423.0,
      "step": 18418
    },
    {
      "entropy": 0.6173863764852285,
      "epoch": 1.6946481043812285,
      "grad_norm": 0.15141154825687408,
      "learning_rate": 4.351519612353176e-05,
      "loss": 0.6187,
      "mean_token_accuracy": 0.8073650225996971,
      "num_tokens": 587896802.0,
      "step": 18419
    },
    {
      "entropy": 0.44303925335407257,
      "epoch": 1.6947401107526696,
      "grad_norm": 0.14652001857757568,
      "learning_rate": 4.351212929739013e-05,
      "loss": 0.4466,
      "mean_token_accuracy": 0.8649734854698181,
      "num_tokens": 587928255.0,
      "step": 18420
    },
    {
      "entropy": 0.5416290983557701,
      "epoch": 1.6948321171241107,
      "grad_norm": 0.14984728395938873,
      "learning_rate": 4.3509062471248504e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.833978071808815,
      "num_tokens": 587959962.0,
      "step": 18421
    },
    {
      "entropy": 0.5696462905034423,
      "epoch": 1.694924123495552,
      "grad_norm": 0.14707182347774506,
      "learning_rate": 4.350599564510688e-05,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8340601660311222,
      "num_tokens": 587992100.0,
      "step": 18422
    },
    {
      "entropy": 0.5483913887292147,
      "epoch": 1.6950161298669932,
      "grad_norm": 0.15575909614562988,
      "learning_rate": 4.350292881896525e-05,
      "loss": 0.5416,
      "mean_token_accuracy": 0.8320976532995701,
      "num_tokens": 588024077.0,
      "step": 18423
    },
    {
      "entropy": 0.6273685740306973,
      "epoch": 1.6951081362384346,
      "grad_norm": 0.1589784175157547,
      "learning_rate": 4.349986199282363e-05,
      "loss": 0.6111,
      "mean_token_accuracy": 0.8145561069250107,
      "num_tokens": 588056000.0,
      "step": 18424
    },
    {
      "entropy": 0.5952739184722304,
      "epoch": 1.6952001426098757,
      "grad_norm": 0.15679332613945007,
      "learning_rate": 4.3496795166682e-05,
      "loss": 0.5789,
      "mean_token_accuracy": 0.8267504870891571,
      "num_tokens": 588087335.0,
      "step": 18425
    },
    {
      "entropy": 0.6791492914780974,
      "epoch": 1.6952921489813169,
      "grad_norm": 0.15606367588043213,
      "learning_rate": 4.349372834054038e-05,
      "loss": 0.664,
      "mean_token_accuracy": 0.7978039272129536,
      "num_tokens": 588119575.0,
      "step": 18426
    },
    {
      "entropy": 0.6291680498979986,
      "epoch": 1.6953841553527582,
      "grad_norm": 0.16094909608364105,
      "learning_rate": 4.349066151439875e-05,
      "loss": 0.6105,
      "mean_token_accuracy": 0.8154231309890747,
      "num_tokens": 588150914.0,
      "step": 18427
    },
    {
      "entropy": 0.590786324813962,
      "epoch": 1.6954761617241993,
      "grad_norm": 0.15915368497371674,
      "learning_rate": 4.3487594688257127e-05,
      "loss": 0.5816,
      "mean_token_accuracy": 0.8226877637207508,
      "num_tokens": 588183001.0,
      "step": 18428
    },
    {
      "entropy": 0.6157570602372289,
      "epoch": 1.6955681680956407,
      "grad_norm": 0.15760819613933563,
      "learning_rate": 4.34845278621155e-05,
      "loss": 0.5992,
      "mean_token_accuracy": 0.8173147737979889,
      "num_tokens": 588214526.0,
      "step": 18429
    },
    {
      "entropy": 0.6460991464555264,
      "epoch": 1.6956601744670818,
      "grad_norm": 0.15788830816745758,
      "learning_rate": 4.3481461035973876e-05,
      "loss": 0.6354,
      "mean_token_accuracy": 0.8054445497691631,
      "num_tokens": 588246473.0,
      "step": 18430
    },
    {
      "entropy": 0.6587066855281591,
      "epoch": 1.695752180838523,
      "grad_norm": 0.16106612980365753,
      "learning_rate": 4.3478394209832244e-05,
      "loss": 0.6411,
      "mean_token_accuracy": 0.8019272051751614,
      "num_tokens": 588278261.0,
      "step": 18431
    },
    {
      "entropy": 0.6025052312761545,
      "epoch": 1.6958441872099643,
      "grad_norm": 0.15042267739772797,
      "learning_rate": 4.3475327383690625e-05,
      "loss": 0.5895,
      "mean_token_accuracy": 0.8174622021615505,
      "num_tokens": 588310163.0,
      "step": 18432
    },
    {
      "entropy": 0.5577194560319185,
      "epoch": 1.6959361935814055,
      "grad_norm": 0.15206314623355865,
      "learning_rate": 4.347226055754899e-05,
      "loss": 0.5472,
      "mean_token_accuracy": 0.8299205787479877,
      "num_tokens": 588341288.0,
      "step": 18433
    },
    {
      "entropy": 0.5707679300103337,
      "epoch": 1.6960281999528468,
      "grad_norm": 0.14676176011562347,
      "learning_rate": 4.346919373140737e-05,
      "loss": 0.5639,
      "mean_token_accuracy": 0.8256403282284737,
      "num_tokens": 588373420.0,
      "step": 18434
    },
    {
      "entropy": 0.6274117734283209,
      "epoch": 1.696120206324288,
      "grad_norm": 0.15926070511341095,
      "learning_rate": 4.346612690526574e-05,
      "loss": 0.6214,
      "mean_token_accuracy": 0.8111732937395573,
      "num_tokens": 588405744.0,
      "step": 18435
    },
    {
      "entropy": 0.5405623577535152,
      "epoch": 1.696212212695729,
      "grad_norm": 0.14515653252601624,
      "learning_rate": 4.346306007912412e-05,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8366645462810993,
      "num_tokens": 588438265.0,
      "step": 18436
    },
    {
      "entropy": 0.551134985871613,
      "epoch": 1.6963042190671704,
      "grad_norm": 0.15556693077087402,
      "learning_rate": 4.345999325298249e-05,
      "loss": 0.5407,
      "mean_token_accuracy": 0.8332723490893841,
      "num_tokens": 588470176.0,
      "step": 18437
    },
    {
      "entropy": 0.4811762971803546,
      "epoch": 1.6963962254386116,
      "grad_norm": 0.14058619737625122,
      "learning_rate": 4.3456926426840866e-05,
      "loss": 0.4694,
      "mean_token_accuracy": 0.8543487451970577,
      "num_tokens": 588502342.0,
      "step": 18438
    },
    {
      "entropy": 0.6076651480980217,
      "epoch": 1.696488231810053,
      "grad_norm": 0.15628540515899658,
      "learning_rate": 4.3453859600699234e-05,
      "loss": 0.5962,
      "mean_token_accuracy": 0.817248985171318,
      "num_tokens": 588534227.0,
      "step": 18439
    },
    {
      "entropy": 0.6729256473481655,
      "epoch": 1.696580238181494,
      "grad_norm": 0.16274741291999817,
      "learning_rate": 4.3450792774557616e-05,
      "loss": 0.6751,
      "mean_token_accuracy": 0.792446568608284,
      "num_tokens": 588565986.0,
      "step": 18440
    },
    {
      "entropy": 0.5433006568346173,
      "epoch": 1.6966722445529352,
      "grad_norm": 0.14443205296993256,
      "learning_rate": 4.3447725948415983e-05,
      "loss": 0.529,
      "mean_token_accuracy": 0.8366485685110092,
      "num_tokens": 588597588.0,
      "step": 18441
    },
    {
      "entropy": 0.6173856118693948,
      "epoch": 1.6967642509243765,
      "grad_norm": 0.16082148253917694,
      "learning_rate": 4.3444659122274365e-05,
      "loss": 0.6021,
      "mean_token_accuracy": 0.8146566562354565,
      "num_tokens": 588628709.0,
      "step": 18442
    },
    {
      "entropy": 0.629576669074595,
      "epoch": 1.6968562572958177,
      "grad_norm": 0.15479624271392822,
      "learning_rate": 4.344159229613273e-05,
      "loss": 0.6136,
      "mean_token_accuracy": 0.8085192106664181,
      "num_tokens": 588660565.0,
      "step": 18443
    },
    {
      "entropy": 0.502192092128098,
      "epoch": 1.696948263667259,
      "grad_norm": 0.14398640394210815,
      "learning_rate": 4.343852546999111e-05,
      "loss": 0.4856,
      "mean_token_accuracy": 0.8491214364767075,
      "num_tokens": 588692799.0,
      "step": 18444
    },
    {
      "entropy": 0.6572241298854351,
      "epoch": 1.6970402700387002,
      "grad_norm": 0.16123105585575104,
      "learning_rate": 4.343545864384948e-05,
      "loss": 0.6574,
      "mean_token_accuracy": 0.7972839809954166,
      "num_tokens": 588725153.0,
      "step": 18445
    },
    {
      "entropy": 0.590052873827517,
      "epoch": 1.6971322764101413,
      "grad_norm": 0.15180185437202454,
      "learning_rate": 4.343239181770786e-05,
      "loss": 0.591,
      "mean_token_accuracy": 0.8169620893895626,
      "num_tokens": 588757570.0,
      "step": 18446
    },
    {
      "entropy": 0.5136648397892714,
      "epoch": 1.6972242827815827,
      "grad_norm": 0.14383892714977264,
      "learning_rate": 4.342932499156623e-05,
      "loss": 0.5123,
      "mean_token_accuracy": 0.8443947546184063,
      "num_tokens": 588789870.0,
      "step": 18447
    },
    {
      "entropy": 0.683208012022078,
      "epoch": 1.6973162891530238,
      "grad_norm": 0.15918804705142975,
      "learning_rate": 4.3426258165424606e-05,
      "loss": 0.6698,
      "mean_token_accuracy": 0.7991132214665413,
      "num_tokens": 588821814.0,
      "step": 18448
    },
    {
      "entropy": 0.623269010335207,
      "epoch": 1.6974082955244651,
      "grad_norm": 0.1526372879743576,
      "learning_rate": 4.3423191339282974e-05,
      "loss": 0.5995,
      "mean_token_accuracy": 0.8165196403861046,
      "num_tokens": 588853871.0,
      "step": 18449
    },
    {
      "entropy": 0.6831203615292907,
      "epoch": 1.6975003018959063,
      "grad_norm": 0.15776199102401733,
      "learning_rate": 4.3420124513141355e-05,
      "loss": 0.6799,
      "mean_token_accuracy": 0.78752401471138,
      "num_tokens": 588886384.0,
      "step": 18450
    },
    {
      "entropy": 0.4655967294238508,
      "epoch": 1.6975923082673474,
      "grad_norm": 0.14306609332561493,
      "learning_rate": 4.341705768699972e-05,
      "loss": 0.456,
      "mean_token_accuracy": 0.856754370033741,
      "num_tokens": 588917993.0,
      "step": 18451
    },
    {
      "entropy": 0.6142483185976744,
      "epoch": 1.6976843146387888,
      "grad_norm": 0.15191037952899933,
      "learning_rate": 4.34139908608581e-05,
      "loss": 0.5968,
      "mean_token_accuracy": 0.8158519119024277,
      "num_tokens": 588949415.0,
      "step": 18452
    },
    {
      "entropy": 0.5904209775617346,
      "epoch": 1.69777632101023,
      "grad_norm": 0.1499367356300354,
      "learning_rate": 4.341092403471647e-05,
      "loss": 0.5852,
      "mean_token_accuracy": 0.8194957785308361,
      "num_tokens": 588980962.0,
      "step": 18453
    },
    {
      "entropy": 0.5185675658285618,
      "epoch": 1.6978683273816713,
      "grad_norm": 0.14244352281093597,
      "learning_rate": 4.340785720857485e-05,
      "loss": 0.5123,
      "mean_token_accuracy": 0.8414500094950199,
      "num_tokens": 589013364.0,
      "step": 18454
    },
    {
      "entropy": 0.5396894440054893,
      "epoch": 1.6979603337531124,
      "grad_norm": 0.1522633582353592,
      "learning_rate": 4.340479038243322e-05,
      "loss": 0.5366,
      "mean_token_accuracy": 0.834348302334547,
      "num_tokens": 589045310.0,
      "step": 18455
    },
    {
      "entropy": 0.7208814360201359,
      "epoch": 1.6980523401245535,
      "grad_norm": 0.1626274585723877,
      "learning_rate": 4.3401723556291596e-05,
      "loss": 0.725,
      "mean_token_accuracy": 0.7811965495347977,
      "num_tokens": 589077242.0,
      "step": 18456
    },
    {
      "entropy": 0.6658765077590942,
      "epoch": 1.6981443464959949,
      "grad_norm": 0.15944470465183258,
      "learning_rate": 4.339865673014997e-05,
      "loss": 0.6679,
      "mean_token_accuracy": 0.7981298752129078,
      "num_tokens": 589109764.0,
      "step": 18457
    },
    {
      "entropy": 0.5910831466317177,
      "epoch": 1.698236352867436,
      "grad_norm": 0.15589796006679535,
      "learning_rate": 4.3395589904008346e-05,
      "loss": 0.5946,
      "mean_token_accuracy": 0.8152587041258812,
      "num_tokens": 589141998.0,
      "step": 18458
    },
    {
      "entropy": 0.7856415286660194,
      "epoch": 1.6983283592388774,
      "grad_norm": 0.1704552322626114,
      "learning_rate": 4.339252307786672e-05,
      "loss": 0.7815,
      "mean_token_accuracy": 0.7604516632854939,
      "num_tokens": 589173909.0,
      "step": 18459
    },
    {
      "entropy": 0.5775080737657845,
      "epoch": 1.6984203656103185,
      "grad_norm": 0.14872293174266815,
      "learning_rate": 4.338945625172509e-05,
      "loss": 0.5624,
      "mean_token_accuracy": 0.8252491168677807,
      "num_tokens": 589205833.0,
      "step": 18460
    },
    {
      "entropy": 0.6166323041543365,
      "epoch": 1.6985123719817596,
      "grad_norm": 0.1499153971672058,
      "learning_rate": 4.338638942558347e-05,
      "loss": 0.5957,
      "mean_token_accuracy": 0.8157620690762997,
      "num_tokens": 589238249.0,
      "step": 18461
    },
    {
      "entropy": 0.5793084008619189,
      "epoch": 1.698604378353201,
      "grad_norm": 0.15395447611808777,
      "learning_rate": 4.338332259944184e-05,
      "loss": 0.5704,
      "mean_token_accuracy": 0.8275145143270493,
      "num_tokens": 589270484.0,
      "step": 18462
    },
    {
      "entropy": 0.6144809797406197,
      "epoch": 1.6986963847246421,
      "grad_norm": 0.15577644109725952,
      "learning_rate": 4.338025577330022e-05,
      "loss": 0.6048,
      "mean_token_accuracy": 0.8158523738384247,
      "num_tokens": 589302553.0,
      "step": 18463
    },
    {
      "entropy": 0.5932198560331017,
      "epoch": 1.6987883910960835,
      "grad_norm": 0.1531713753938675,
      "learning_rate": 4.337718894715859e-05,
      "loss": 0.5786,
      "mean_token_accuracy": 0.8204208500683308,
      "num_tokens": 589334300.0,
      "step": 18464
    },
    {
      "entropy": 0.6586618069559336,
      "epoch": 1.6988803974675246,
      "grad_norm": 0.15669222176074982,
      "learning_rate": 4.337412212101696e-05,
      "loss": 0.6476,
      "mean_token_accuracy": 0.8018350712954998,
      "num_tokens": 589365997.0,
      "step": 18465
    },
    {
      "entropy": 0.6517608175054193,
      "epoch": 1.6989724038389658,
      "grad_norm": 0.1600392609834671,
      "learning_rate": 4.3371055294875336e-05,
      "loss": 0.6469,
      "mean_token_accuracy": 0.8020252510905266,
      "num_tokens": 589397511.0,
      "step": 18466
    },
    {
      "entropy": 0.6071168072521687,
      "epoch": 1.699064410210407,
      "grad_norm": 0.1472749412059784,
      "learning_rate": 4.336798846873371e-05,
      "loss": 0.5959,
      "mean_token_accuracy": 0.8172474801540375,
      "num_tokens": 589429990.0,
      "step": 18467
    },
    {
      "entropy": 0.7253922801464796,
      "epoch": 1.6991564165818482,
      "grad_norm": 0.16044436395168304,
      "learning_rate": 4.3364921642592085e-05,
      "loss": 0.7244,
      "mean_token_accuracy": 0.7867164611816406,
      "num_tokens": 589461234.0,
      "step": 18468
    },
    {
      "entropy": 0.5768087208271027,
      "epoch": 1.6992484229532896,
      "grad_norm": 0.15431222319602966,
      "learning_rate": 4.336185481645046e-05,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8307097479701042,
      "num_tokens": 589493514.0,
      "step": 18469
    },
    {
      "entropy": 0.638455281034112,
      "epoch": 1.6993404293247307,
      "grad_norm": 0.15965895354747772,
      "learning_rate": 4.335878799030883e-05,
      "loss": 0.6266,
      "mean_token_accuracy": 0.8069493807852268,
      "num_tokens": 589525623.0,
      "step": 18470
    },
    {
      "entropy": 0.6780338007956743,
      "epoch": 1.6994324356961719,
      "grad_norm": 0.16333338618278503,
      "learning_rate": 4.335572116416721e-05,
      "loss": 0.6785,
      "mean_token_accuracy": 0.7919697128236294,
      "num_tokens": 589557904.0,
      "step": 18471
    },
    {
      "entropy": 0.6887124311178923,
      "epoch": 1.6995244420676132,
      "grad_norm": 0.16525916755199432,
      "learning_rate": 4.335265433802558e-05,
      "loss": 0.6933,
      "mean_token_accuracy": 0.7910476885735989,
      "num_tokens": 589589887.0,
      "step": 18472
    },
    {
      "entropy": 0.6245485232211649,
      "epoch": 1.6996164484390544,
      "grad_norm": 0.15462790429592133,
      "learning_rate": 4.334958751188395e-05,
      "loss": 0.6145,
      "mean_token_accuracy": 0.8132783249020576,
      "num_tokens": 589622274.0,
      "step": 18473
    },
    {
      "entropy": 0.6184563506394625,
      "epoch": 1.6997084548104957,
      "grad_norm": 0.15475334227085114,
      "learning_rate": 4.3346520685742326e-05,
      "loss": 0.61,
      "mean_token_accuracy": 0.812042985111475,
      "num_tokens": 589654107.0,
      "step": 18474
    },
    {
      "entropy": 0.6878101825714111,
      "epoch": 1.6998004611819368,
      "grad_norm": 0.15852995216846466,
      "learning_rate": 4.33434538596007e-05,
      "loss": 0.6744,
      "mean_token_accuracy": 0.7951664179563522,
      "num_tokens": 589686215.0,
      "step": 18475
    },
    {
      "entropy": 0.6582869682461023,
      "epoch": 1.699892467553378,
      "grad_norm": 0.15731529891490936,
      "learning_rate": 4.3340387033459076e-05,
      "loss": 0.6536,
      "mean_token_accuracy": 0.8001772835850716,
      "num_tokens": 589717710.0,
      "step": 18476
    },
    {
      "entropy": 0.5706868525594473,
      "epoch": 1.6999844739248193,
      "grad_norm": 0.14754796028137207,
      "learning_rate": 4.333732020731745e-05,
      "loss": 0.5541,
      "mean_token_accuracy": 0.8308042287826538,
      "num_tokens": 589750409.0,
      "step": 18477
    },
    {
      "entropy": 0.6379914712160826,
      "epoch": 1.7000764802962605,
      "grad_norm": 0.15372717380523682,
      "learning_rate": 4.333425338117582e-05,
      "loss": 0.6241,
      "mean_token_accuracy": 0.804260291159153,
      "num_tokens": 589781917.0,
      "step": 18478
    },
    {
      "entropy": 0.6341215381398797,
      "epoch": 1.7001684866677018,
      "grad_norm": 0.15756331384181976,
      "learning_rate": 4.33311865550342e-05,
      "loss": 0.6134,
      "mean_token_accuracy": 0.8069992773234844,
      "num_tokens": 589813198.0,
      "step": 18479
    },
    {
      "entropy": 0.5614200662821531,
      "epoch": 1.700260493039143,
      "grad_norm": 0.15407519042491913,
      "learning_rate": 4.332811972889257e-05,
      "loss": 0.5622,
      "mean_token_accuracy": 0.8283643573522568,
      "num_tokens": 589845158.0,
      "step": 18480
    },
    {
      "entropy": 0.6028850544244051,
      "epoch": 1.700352499410584,
      "grad_norm": 0.15604844689369202,
      "learning_rate": 4.332505290275094e-05,
      "loss": 0.5884,
      "mean_token_accuracy": 0.8164391592144966,
      "num_tokens": 589877303.0,
      "step": 18481
    },
    {
      "entropy": 0.6392002254724503,
      "epoch": 1.7004445057820254,
      "grad_norm": 0.15965236723423004,
      "learning_rate": 4.332198607660932e-05,
      "loss": 0.6267,
      "mean_token_accuracy": 0.8092700354754925,
      "num_tokens": 589909289.0,
      "step": 18482
    },
    {
      "entropy": 0.6364929955452681,
      "epoch": 1.7005365121534666,
      "grad_norm": 0.16164182126522064,
      "learning_rate": 4.331891925046769e-05,
      "loss": 0.613,
      "mean_token_accuracy": 0.8101599924266338,
      "num_tokens": 589941337.0,
      "step": 18483
    },
    {
      "entropy": 0.5279473338741809,
      "epoch": 1.700628518524908,
      "grad_norm": 0.14758892357349396,
      "learning_rate": 4.3315852424326066e-05,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8392584100365639,
      "num_tokens": 589973971.0,
      "step": 18484
    },
    {
      "entropy": 0.6356919407844543,
      "epoch": 1.700720524896349,
      "grad_norm": 0.15838530659675598,
      "learning_rate": 4.331278559818444e-05,
      "loss": 0.6311,
      "mean_token_accuracy": 0.8047512322664261,
      "num_tokens": 590005731.0,
      "step": 18485
    },
    {
      "entropy": 0.681212093681097,
      "epoch": 1.7008125312677902,
      "grad_norm": 0.16079983115196228,
      "learning_rate": 4.3309718772042815e-05,
      "loss": 0.6751,
      "mean_token_accuracy": 0.7954626455903053,
      "num_tokens": 590037587.0,
      "step": 18486
    },
    {
      "entropy": 0.5407469645142555,
      "epoch": 1.7009045376392316,
      "grad_norm": 0.1517559289932251,
      "learning_rate": 4.330665194590119e-05,
      "loss": 0.5283,
      "mean_token_accuracy": 0.8336001299321651,
      "num_tokens": 590068738.0,
      "step": 18487
    },
    {
      "entropy": 0.6657405588775873,
      "epoch": 1.7009965440106727,
      "grad_norm": 0.16403748095035553,
      "learning_rate": 4.3303585119759565e-05,
      "loss": 0.6587,
      "mean_token_accuracy": 0.8011924102902412,
      "num_tokens": 590100202.0,
      "step": 18488
    },
    {
      "entropy": 0.6906909700483084,
      "epoch": 1.701088550382114,
      "grad_norm": 0.15829278528690338,
      "learning_rate": 4.330051829361794e-05,
      "loss": 0.6838,
      "mean_token_accuracy": 0.7914650626480579,
      "num_tokens": 590131483.0,
      "step": 18489
    },
    {
      "entropy": 0.6088766641914845,
      "epoch": 1.7011805567535552,
      "grad_norm": 0.15714426338672638,
      "learning_rate": 4.3297451467476314e-05,
      "loss": 0.5956,
      "mean_token_accuracy": 0.8116152621805668,
      "num_tokens": 590163380.0,
      "step": 18490
    },
    {
      "entropy": 0.5469550052657723,
      "epoch": 1.7012725631249963,
      "grad_norm": 0.1439819186925888,
      "learning_rate": 4.329438464133468e-05,
      "loss": 0.5394,
      "mean_token_accuracy": 0.8336788415908813,
      "num_tokens": 590195212.0,
      "step": 18491
    },
    {
      "entropy": 0.5461481306701899,
      "epoch": 1.7013645694964377,
      "grad_norm": 0.14712782204151154,
      "learning_rate": 4.329131781519306e-05,
      "loss": 0.5328,
      "mean_token_accuracy": 0.8320152461528778,
      "num_tokens": 590227220.0,
      "step": 18492
    },
    {
      "entropy": 0.49039215222001076,
      "epoch": 1.7014565758678788,
      "grad_norm": 0.14523953199386597,
      "learning_rate": 4.328825098905143e-05,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8504128083586693,
      "num_tokens": 590259784.0,
      "step": 18493
    },
    {
      "entropy": 0.6964020980522037,
      "epoch": 1.7015485822393202,
      "grad_norm": 0.16092947125434875,
      "learning_rate": 4.3285184162909806e-05,
      "loss": 0.6959,
      "mean_token_accuracy": 0.7900769673287868,
      "num_tokens": 590291167.0,
      "step": 18494
    },
    {
      "entropy": 0.4358631824143231,
      "epoch": 1.7016405886107613,
      "grad_norm": 0.1462910771369934,
      "learning_rate": 4.328211733676818e-05,
      "loss": 0.4186,
      "mean_token_accuracy": 0.8646043837070465,
      "num_tokens": 590322883.0,
      "step": 18495
    },
    {
      "entropy": 0.7313349470496178,
      "epoch": 1.7017325949822024,
      "grad_norm": 0.1667155921459198,
      "learning_rate": 4.3279050510626555e-05,
      "loss": 0.7177,
      "mean_token_accuracy": 0.783086534589529,
      "num_tokens": 590354932.0,
      "step": 18496
    },
    {
      "entropy": 0.64105917327106,
      "epoch": 1.7018246013536438,
      "grad_norm": 0.16041232645511627,
      "learning_rate": 4.327598368448493e-05,
      "loss": 0.6491,
      "mean_token_accuracy": 0.8036639243364334,
      "num_tokens": 590386648.0,
      "step": 18497
    },
    {
      "entropy": 0.5991866383701563,
      "epoch": 1.701916607725085,
      "grad_norm": 0.15559524297714233,
      "learning_rate": 4.3272916858343304e-05,
      "loss": 0.5859,
      "mean_token_accuracy": 0.819328211247921,
      "num_tokens": 590418371.0,
      "step": 18498
    },
    {
      "entropy": 0.6858785524964333,
      "epoch": 1.7020086140965263,
      "grad_norm": 0.16218741238117218,
      "learning_rate": 4.326985003220167e-05,
      "loss": 0.6785,
      "mean_token_accuracy": 0.7939840219914913,
      "num_tokens": 590450188.0,
      "step": 18499
    },
    {
      "entropy": 0.5796874864026904,
      "epoch": 1.7021006204679674,
      "grad_norm": 0.1464882493019104,
      "learning_rate": 4.3266783206060054e-05,
      "loss": 0.5677,
      "mean_token_accuracy": 0.8262629620730877,
      "num_tokens": 590482241.0,
      "step": 18500
    },
    {
      "entropy": 0.5671443548053503,
      "epoch": 1.7021926268394085,
      "grad_norm": 0.15536896884441376,
      "learning_rate": 4.326371637991842e-05,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8279390782117844,
      "num_tokens": 590514385.0,
      "step": 18501
    },
    {
      "entropy": 0.5730184782296419,
      "epoch": 1.7022846332108499,
      "grad_norm": 0.15870970487594604,
      "learning_rate": 4.32606495537768e-05,
      "loss": 0.5638,
      "mean_token_accuracy": 0.8219228871166706,
      "num_tokens": 590546378.0,
      "step": 18502
    },
    {
      "entropy": 0.6489887591451406,
      "epoch": 1.7023766395822912,
      "grad_norm": 0.1569584459066391,
      "learning_rate": 4.325758272763517e-05,
      "loss": 0.6452,
      "mean_token_accuracy": 0.8012672290205956,
      "num_tokens": 590578664.0,
      "step": 18503
    },
    {
      "entropy": 0.5594515046104789,
      "epoch": 1.7024686459537324,
      "grad_norm": 0.15053047239780426,
      "learning_rate": 4.3254515901493546e-05,
      "loss": 0.5446,
      "mean_token_accuracy": 0.829731147736311,
      "num_tokens": 590610428.0,
      "step": 18504
    },
    {
      "entropy": 0.7234377525746822,
      "epoch": 1.7025606523251735,
      "grad_norm": 0.16239503026008606,
      "learning_rate": 4.325144907535192e-05,
      "loss": 0.7091,
      "mean_token_accuracy": 0.7848448753356934,
      "num_tokens": 590642041.0,
      "step": 18505
    },
    {
      "entropy": 0.5662931110709906,
      "epoch": 1.7026526586966146,
      "grad_norm": 0.14759430289268494,
      "learning_rate": 4.3248382249210295e-05,
      "loss": 0.5588,
      "mean_token_accuracy": 0.8273534849286079,
      "num_tokens": 590674141.0,
      "step": 18506
    },
    {
      "entropy": 0.5711950706318021,
      "epoch": 1.702744665068056,
      "grad_norm": 0.15647603571414948,
      "learning_rate": 4.324531542306867e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.826300822198391,
      "num_tokens": 590705900.0,
      "step": 18507
    },
    {
      "entropy": 0.5398175492882729,
      "epoch": 1.7028366714394974,
      "grad_norm": 0.14288891851902008,
      "learning_rate": 4.3242248596927044e-05,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8375294730067253,
      "num_tokens": 590738267.0,
      "step": 18508
    },
    {
      "entropy": 0.5621651108376682,
      "epoch": 1.7029286778109385,
      "grad_norm": 0.15868909657001495,
      "learning_rate": 4.323918177078541e-05,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8260793909430504,
      "num_tokens": 590769928.0,
      "step": 18509
    },
    {
      "entropy": 0.5454700705595315,
      "epoch": 1.7030206841823796,
      "grad_norm": 0.14401759207248688,
      "learning_rate": 4.3236114944643793e-05,
      "loss": 0.539,
      "mean_token_accuracy": 0.838415551930666,
      "num_tokens": 590802675.0,
      "step": 18510
    },
    {
      "entropy": 0.5467212228104472,
      "epoch": 1.7031126905538208,
      "grad_norm": 0.14751212298870087,
      "learning_rate": 4.323304811850216e-05,
      "loss": 0.5473,
      "mean_token_accuracy": 0.8322051391005516,
      "num_tokens": 590834709.0,
      "step": 18511
    },
    {
      "entropy": 0.5860138991847634,
      "epoch": 1.703204696925262,
      "grad_norm": 0.1499340981245041,
      "learning_rate": 4.3229981292360536e-05,
      "loss": 0.5777,
      "mean_token_accuracy": 0.8223005011677742,
      "num_tokens": 590867039.0,
      "step": 18512
    },
    {
      "entropy": 0.5613792622461915,
      "epoch": 1.7032967032967035,
      "grad_norm": 0.15714824199676514,
      "learning_rate": 4.322691446621891e-05,
      "loss": 0.5426,
      "mean_token_accuracy": 0.8308545649051666,
      "num_tokens": 590898182.0,
      "step": 18513
    },
    {
      "entropy": 0.5822851471602917,
      "epoch": 1.7033887096681446,
      "grad_norm": 0.14910495281219482,
      "learning_rate": 4.3223847640077285e-05,
      "loss": 0.5747,
      "mean_token_accuracy": 0.8249287605285645,
      "num_tokens": 590930888.0,
      "step": 18514
    },
    {
      "entropy": 0.6130818948149681,
      "epoch": 1.7034807160395857,
      "grad_norm": 0.16012457013130188,
      "learning_rate": 4.322078081393566e-05,
      "loss": 0.6216,
      "mean_token_accuracy": 0.8068731129169464,
      "num_tokens": 590963007.0,
      "step": 18515
    },
    {
      "entropy": 0.530451353173703,
      "epoch": 1.7035727224110269,
      "grad_norm": 0.14995579421520233,
      "learning_rate": 4.3217713987794035e-05,
      "loss": 0.5185,
      "mean_token_accuracy": 0.8406858816742897,
      "num_tokens": 590994967.0,
      "step": 18516
    },
    {
      "entropy": 0.6245351452380419,
      "epoch": 1.7036647287824682,
      "grad_norm": 0.15816013514995575,
      "learning_rate": 4.321464716165241e-05,
      "loss": 0.6153,
      "mean_token_accuracy": 0.8113754652440548,
      "num_tokens": 591026933.0,
      "step": 18517
    },
    {
      "entropy": 0.5433363942429423,
      "epoch": 1.7037567351539096,
      "grad_norm": 0.14948679506778717,
      "learning_rate": 4.3211580335510784e-05,
      "loss": 0.5479,
      "mean_token_accuracy": 0.8320051729679108,
      "num_tokens": 591058983.0,
      "step": 18518
    },
    {
      "entropy": 0.5966420788317919,
      "epoch": 1.7038487415253507,
      "grad_norm": 0.1531527191400528,
      "learning_rate": 4.320851350936916e-05,
      "loss": 0.5763,
      "mean_token_accuracy": 0.8193149492144585,
      "num_tokens": 591091229.0,
      "step": 18519
    },
    {
      "entropy": 0.5295281279832125,
      "epoch": 1.7039407478967918,
      "grad_norm": 0.14936797320842743,
      "learning_rate": 4.3205446683227526e-05,
      "loss": 0.523,
      "mean_token_accuracy": 0.8337665274739265,
      "num_tokens": 591123276.0,
      "step": 18520
    },
    {
      "entropy": 0.5903604025952518,
      "epoch": 1.704032754268233,
      "grad_norm": 0.15583258867263794,
      "learning_rate": 4.320237985708591e-05,
      "loss": 0.5921,
      "mean_token_accuracy": 0.8189320452511311,
      "num_tokens": 591154560.0,
      "step": 18521
    },
    {
      "entropy": 0.7213636115193367,
      "epoch": 1.7041247606396743,
      "grad_norm": 0.1667565554380417,
      "learning_rate": 4.3199313030944276e-05,
      "loss": 0.72,
      "mean_token_accuracy": 0.7807835899293423,
      "num_tokens": 591186232.0,
      "step": 18522
    },
    {
      "entropy": 0.7120594773441553,
      "epoch": 1.7042167670111157,
      "grad_norm": 0.15809576213359833,
      "learning_rate": 4.319624620480266e-05,
      "loss": 0.6957,
      "mean_token_accuracy": 0.7890820242464542,
      "num_tokens": 591217502.0,
      "step": 18523
    },
    {
      "entropy": 0.708829689770937,
      "epoch": 1.7043087733825568,
      "grad_norm": 0.1639758050441742,
      "learning_rate": 4.3193179378661025e-05,
      "loss": 0.6945,
      "mean_token_accuracy": 0.7900526821613312,
      "num_tokens": 591249123.0,
      "step": 18524
    },
    {
      "entropy": 0.6492002829909325,
      "epoch": 1.704400779753998,
      "grad_norm": 0.1579829901456833,
      "learning_rate": 4.31901125525194e-05,
      "loss": 0.6259,
      "mean_token_accuracy": 0.8046450577676296,
      "num_tokens": 591280714.0,
      "step": 18525
    },
    {
      "entropy": 0.5680166287347674,
      "epoch": 1.704492786125439,
      "grad_norm": 0.15612870454788208,
      "learning_rate": 4.3187045726377774e-05,
      "loss": 0.553,
      "mean_token_accuracy": 0.8312224298715591,
      "num_tokens": 591313343.0,
      "step": 18526
    },
    {
      "entropy": 0.6190493442118168,
      "epoch": 1.7045847924968804,
      "grad_norm": 0.15407058596611023,
      "learning_rate": 4.318397890023615e-05,
      "loss": 0.6017,
      "mean_token_accuracy": 0.8162629418075085,
      "num_tokens": 591345453.0,
      "step": 18527
    },
    {
      "entropy": 0.5598542522639036,
      "epoch": 1.7046767988683218,
      "grad_norm": 0.15266302227973938,
      "learning_rate": 4.3180912074094524e-05,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8359952345490456,
      "num_tokens": 591376874.0,
      "step": 18528
    },
    {
      "entropy": 0.5848038140684366,
      "epoch": 1.704768805239763,
      "grad_norm": 0.1520274132490158,
      "learning_rate": 4.31778452479529e-05,
      "loss": 0.5535,
      "mean_token_accuracy": 0.8224668204784393,
      "num_tokens": 591408147.0,
      "step": 18529
    },
    {
      "entropy": 0.5580922700464725,
      "epoch": 1.704860811611204,
      "grad_norm": 0.1547228842973709,
      "learning_rate": 4.3174778421811266e-05,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8352249450981617,
      "num_tokens": 591440199.0,
      "step": 18530
    },
    {
      "entropy": 0.6370693109929562,
      "epoch": 1.7049528179826452,
      "grad_norm": 0.1520196497440338,
      "learning_rate": 4.317171159566965e-05,
      "loss": 0.6237,
      "mean_token_accuracy": 0.8100930489599705,
      "num_tokens": 591472464.0,
      "step": 18531
    },
    {
      "entropy": 0.5008453289046884,
      "epoch": 1.7050448243540866,
      "grad_norm": 0.14922459423542023,
      "learning_rate": 4.3168644769528015e-05,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8463876210153103,
      "num_tokens": 591503821.0,
      "step": 18532
    },
    {
      "entropy": 0.5618497999384999,
      "epoch": 1.705136830725528,
      "grad_norm": 0.15066206455230713,
      "learning_rate": 4.316557794338639e-05,
      "loss": 0.5508,
      "mean_token_accuracy": 0.831241600215435,
      "num_tokens": 591535596.0,
      "step": 18533
    },
    {
      "entropy": 0.7131576370447874,
      "epoch": 1.705228837096969,
      "grad_norm": 0.1682099550962448,
      "learning_rate": 4.3162511117244765e-05,
      "loss": 0.7145,
      "mean_token_accuracy": 0.7884935773909092,
      "num_tokens": 591567909.0,
      "step": 18534
    },
    {
      "entropy": 0.4142937930300832,
      "epoch": 1.7053208434684102,
      "grad_norm": 0.13256002962589264,
      "learning_rate": 4.315944429110314e-05,
      "loss": 0.4018,
      "mean_token_accuracy": 0.8726906776428223,
      "num_tokens": 591599959.0,
      "step": 18535
    },
    {
      "entropy": 0.5971131324768066,
      "epoch": 1.7054128498398513,
      "grad_norm": 0.14954769611358643,
      "learning_rate": 4.3156377464961514e-05,
      "loss": 0.585,
      "mean_token_accuracy": 0.822620328515768,
      "num_tokens": 591632107.0,
      "step": 18536
    },
    {
      "entropy": 0.5633002854883671,
      "epoch": 1.7055048562112927,
      "grad_norm": 0.14648787677288055,
      "learning_rate": 4.315331063881989e-05,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8296898007392883,
      "num_tokens": 591664580.0,
      "step": 18537
    },
    {
      "entropy": 0.5550855710171163,
      "epoch": 1.705596862582734,
      "grad_norm": 0.15242180228233337,
      "learning_rate": 4.3150243812678256e-05,
      "loss": 0.5578,
      "mean_token_accuracy": 0.829811304807663,
      "num_tokens": 591696004.0,
      "step": 18538
    },
    {
      "entropy": 0.48382814414799213,
      "epoch": 1.7056888689541752,
      "grad_norm": 0.1502729058265686,
      "learning_rate": 4.314717698653664e-05,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8477377593517303,
      "num_tokens": 591728420.0,
      "step": 18539
    },
    {
      "entropy": 0.6592730600386858,
      "epoch": 1.7057808753256163,
      "grad_norm": 0.15873222053050995,
      "learning_rate": 4.3144110160395006e-05,
      "loss": 0.6412,
      "mean_token_accuracy": 0.8022210597991943,
      "num_tokens": 591759970.0,
      "step": 18540
    },
    {
      "entropy": 0.5987363150343299,
      "epoch": 1.7058728816970574,
      "grad_norm": 0.15537257492542267,
      "learning_rate": 4.314104333425338e-05,
      "loss": 0.5984,
      "mean_token_accuracy": 0.8212200589478016,
      "num_tokens": 591791996.0,
      "step": 18541
    },
    {
      "entropy": 0.5783218508586287,
      "epoch": 1.7059648880684988,
      "grad_norm": 0.15505388379096985,
      "learning_rate": 4.3137976508111755e-05,
      "loss": 0.5688,
      "mean_token_accuracy": 0.8221551179885864,
      "num_tokens": 591824289.0,
      "step": 18542
    },
    {
      "entropy": 0.6156395561993122,
      "epoch": 1.7060568944399401,
      "grad_norm": 0.15842337906360626,
      "learning_rate": 4.313490968197013e-05,
      "loss": 0.6093,
      "mean_token_accuracy": 0.8107838518917561,
      "num_tokens": 591856514.0,
      "step": 18543
    },
    {
      "entropy": 0.6749777123332024,
      "epoch": 1.7061489008113813,
      "grad_norm": 0.15818434953689575,
      "learning_rate": 4.3131842855828504e-05,
      "loss": 0.6576,
      "mean_token_accuracy": 0.7979209497570992,
      "num_tokens": 591887806.0,
      "step": 18544
    },
    {
      "entropy": 0.5888166769873351,
      "epoch": 1.7062409071828224,
      "grad_norm": 0.15084560215473175,
      "learning_rate": 4.312877602968688e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.8239969871938229,
      "num_tokens": 591919931.0,
      "step": 18545
    },
    {
      "entropy": 0.5617733858525753,
      "epoch": 1.7063329135542635,
      "grad_norm": 0.14966511726379395,
      "learning_rate": 4.3125709203545254e-05,
      "loss": 0.548,
      "mean_token_accuracy": 0.8305093683302402,
      "num_tokens": 591952233.0,
      "step": 18546
    },
    {
      "entropy": 0.5599923804402351,
      "epoch": 1.7064249199257049,
      "grad_norm": 0.14818412065505981,
      "learning_rate": 4.312264237740363e-05,
      "loss": 0.541,
      "mean_token_accuracy": 0.8355659246444702,
      "num_tokens": 591983314.0,
      "step": 18547
    },
    {
      "entropy": 0.6374699422158301,
      "epoch": 1.7065169262971462,
      "grad_norm": 0.1612899899482727,
      "learning_rate": 4.3119575551262e-05,
      "loss": 0.628,
      "mean_token_accuracy": 0.8097050562500954,
      "num_tokens": 592015544.0,
      "step": 18548
    },
    {
      "entropy": 0.6006647888571024,
      "epoch": 1.7066089326685874,
      "grad_norm": 0.1582821011543274,
      "learning_rate": 4.311650872512038e-05,
      "loss": 0.5771,
      "mean_token_accuracy": 0.8220110759139061,
      "num_tokens": 592047642.0,
      "step": 18549
    },
    {
      "entropy": 0.5248645825777203,
      "epoch": 1.7067009390400285,
      "grad_norm": 0.14855974912643433,
      "learning_rate": 4.311344189897875e-05,
      "loss": 0.5204,
      "mean_token_accuracy": 0.8418196700513363,
      "num_tokens": 592080250.0,
      "step": 18550
    },
    {
      "entropy": 0.6450114534236491,
      "epoch": 1.7067929454114696,
      "grad_norm": 0.1582130789756775,
      "learning_rate": 4.311037507283712e-05,
      "loss": 0.6448,
      "mean_token_accuracy": 0.8034885413944721,
      "num_tokens": 592111238.0,
      "step": 18551
    },
    {
      "entropy": 0.7020458355545998,
      "epoch": 1.706884951782911,
      "grad_norm": 0.16211311519145966,
      "learning_rate": 4.31073082466955e-05,
      "loss": 0.6859,
      "mean_token_accuracy": 0.7898649163544178,
      "num_tokens": 592142553.0,
      "step": 18552
    },
    {
      "entropy": 0.5793888978660107,
      "epoch": 1.7069769581543524,
      "grad_norm": 0.15434877574443817,
      "learning_rate": 4.310424142055387e-05,
      "loss": 0.5602,
      "mean_token_accuracy": 0.8236661963164806,
      "num_tokens": 592174056.0,
      "step": 18553
    },
    {
      "entropy": 0.58451340906322,
      "epoch": 1.7070689645257935,
      "grad_norm": 0.1539301723241806,
      "learning_rate": 4.3101174594412244e-05,
      "loss": 0.5772,
      "mean_token_accuracy": 0.8195346221327782,
      "num_tokens": 592205068.0,
      "step": 18554
    },
    {
      "entropy": 0.69652945920825,
      "epoch": 1.7071609708972346,
      "grad_norm": 0.16723862290382385,
      "learning_rate": 4.309810776827062e-05,
      "loss": 0.6971,
      "mean_token_accuracy": 0.7918095737695694,
      "num_tokens": 592236168.0,
      "step": 18555
    },
    {
      "entropy": 0.5754241235554218,
      "epoch": 1.7072529772686758,
      "grad_norm": 0.15240873396396637,
      "learning_rate": 4.309504094212899e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8272681012749672,
      "num_tokens": 592268360.0,
      "step": 18556
    },
    {
      "entropy": 0.6015109252184629,
      "epoch": 1.707344983640117,
      "grad_norm": 0.15102991461753845,
      "learning_rate": 4.309197411598737e-05,
      "loss": 0.5936,
      "mean_token_accuracy": 0.8202612325549126,
      "num_tokens": 592299660.0,
      "step": 18557
    },
    {
      "entropy": 0.5768546620383859,
      "epoch": 1.7074369900115585,
      "grad_norm": 0.1598718911409378,
      "learning_rate": 4.308890728984574e-05,
      "loss": 0.5798,
      "mean_token_accuracy": 0.8238951005041599,
      "num_tokens": 592330517.0,
      "step": 18558
    },
    {
      "entropy": 0.5037902663461864,
      "epoch": 1.7075289963829996,
      "grad_norm": 0.14830507338047028,
      "learning_rate": 4.308584046370411e-05,
      "loss": 0.4961,
      "mean_token_accuracy": 0.846704974770546,
      "num_tokens": 592361971.0,
      "step": 18559
    },
    {
      "entropy": 0.5903837792575359,
      "epoch": 1.7076210027544407,
      "grad_norm": 0.14629578590393066,
      "learning_rate": 4.308277363756249e-05,
      "loss": 0.5706,
      "mean_token_accuracy": 0.8229061178863049,
      "num_tokens": 592393805.0,
      "step": 18560
    },
    {
      "entropy": 0.633722297847271,
      "epoch": 1.7077130091258819,
      "grad_norm": 0.15794843435287476,
      "learning_rate": 4.307970681142086e-05,
      "loss": 0.6215,
      "mean_token_accuracy": 0.811347309499979,
      "num_tokens": 592426521.0,
      "step": 18561
    },
    {
      "entropy": 0.6479030381888151,
      "epoch": 1.7078050154973232,
      "grad_norm": 0.15832464396953583,
      "learning_rate": 4.307663998527924e-05,
      "loss": 0.6331,
      "mean_token_accuracy": 0.8091930150985718,
      "num_tokens": 592458876.0,
      "step": 18562
    },
    {
      "entropy": 0.6662773601710796,
      "epoch": 1.7078970218687646,
      "grad_norm": 0.16332194209098816,
      "learning_rate": 4.307357315913761e-05,
      "loss": 0.6533,
      "mean_token_accuracy": 0.8018324822187424,
      "num_tokens": 592490637.0,
      "step": 18563
    },
    {
      "entropy": 0.6846255362033844,
      "epoch": 1.7079890282402057,
      "grad_norm": 0.15844498574733734,
      "learning_rate": 4.3070506332995984e-05,
      "loss": 0.6755,
      "mean_token_accuracy": 0.7939007878303528,
      "num_tokens": 592521786.0,
      "step": 18564
    },
    {
      "entropy": 0.7393733412027359,
      "epoch": 1.7080810346116468,
      "grad_norm": 0.1707116961479187,
      "learning_rate": 4.306743950685436e-05,
      "loss": 0.727,
      "mean_token_accuracy": 0.7763060368597507,
      "num_tokens": 592553757.0,
      "step": 18565
    },
    {
      "entropy": 0.5549713429063559,
      "epoch": 1.708173040983088,
      "grad_norm": 0.1549854278564453,
      "learning_rate": 4.306437268071273e-05,
      "loss": 0.5467,
      "mean_token_accuracy": 0.8343307189643383,
      "num_tokens": 592586525.0,
      "step": 18566
    },
    {
      "entropy": 0.6231631338596344,
      "epoch": 1.7082650473545293,
      "grad_norm": 0.1567842960357666,
      "learning_rate": 4.306130585457111e-05,
      "loss": 0.6134,
      "mean_token_accuracy": 0.8145304098725319,
      "num_tokens": 592618654.0,
      "step": 18567
    },
    {
      "entropy": 0.6252487003803253,
      "epoch": 1.7083570537259707,
      "grad_norm": 0.15938034653663635,
      "learning_rate": 4.305823902842948e-05,
      "loss": 0.6323,
      "mean_token_accuracy": 0.8086613230407238,
      "num_tokens": 592649842.0,
      "step": 18568
    },
    {
      "entropy": 0.5011957446113229,
      "epoch": 1.7084490600974118,
      "grad_norm": 0.14815951883792877,
      "learning_rate": 4.305517220228785e-05,
      "loss": 0.4829,
      "mean_token_accuracy": 0.8473300486803055,
      "num_tokens": 592681593.0,
      "step": 18569
    },
    {
      "entropy": 0.4925826136022806,
      "epoch": 1.708541066468853,
      "grad_norm": 0.14513693749904633,
      "learning_rate": 4.305210537614623e-05,
      "loss": 0.479,
      "mean_token_accuracy": 0.8486065566539764,
      "num_tokens": 592713466.0,
      "step": 18570
    },
    {
      "entropy": 0.7085178522393107,
      "epoch": 1.708633072840294,
      "grad_norm": 0.15916158258914948,
      "learning_rate": 4.30490385500046e-05,
      "loss": 0.6897,
      "mean_token_accuracy": 0.7886235006153584,
      "num_tokens": 592745279.0,
      "step": 18571
    },
    {
      "entropy": 0.6440632548183203,
      "epoch": 1.7087250792117354,
      "grad_norm": 0.15706005692481995,
      "learning_rate": 4.3045971723862974e-05,
      "loss": 0.637,
      "mean_token_accuracy": 0.8048165589570999,
      "num_tokens": 592777319.0,
      "step": 18572
    },
    {
      "entropy": 0.6897039427421987,
      "epoch": 1.7088170855831768,
      "grad_norm": 0.15625564754009247,
      "learning_rate": 4.304290489772135e-05,
      "loss": 0.6856,
      "mean_token_accuracy": 0.791066188365221,
      "num_tokens": 592810087.0,
      "step": 18573
    },
    {
      "entropy": 0.6522763948887587,
      "epoch": 1.708909091954618,
      "grad_norm": 0.15691262483596802,
      "learning_rate": 4.3039838071579723e-05,
      "loss": 0.6509,
      "mean_token_accuracy": 0.8042047470808029,
      "num_tokens": 592842327.0,
      "step": 18574
    },
    {
      "entropy": 0.6279069883748889,
      "epoch": 1.709001098326059,
      "grad_norm": 0.15541912615299225,
      "learning_rate": 4.30367712454381e-05,
      "loss": 0.6219,
      "mean_token_accuracy": 0.813308160752058,
      "num_tokens": 592874534.0,
      "step": 18575
    },
    {
      "entropy": 0.6152468807995319,
      "epoch": 1.7090931046975002,
      "grad_norm": 0.16162872314453125,
      "learning_rate": 4.303370441929647e-05,
      "loss": 0.6045,
      "mean_token_accuracy": 0.8143512047827244,
      "num_tokens": 592906681.0,
      "step": 18576
    },
    {
      "entropy": 0.6538624819368124,
      "epoch": 1.7091851110689416,
      "grad_norm": 0.1542944312095642,
      "learning_rate": 4.303063759315485e-05,
      "loss": 0.6375,
      "mean_token_accuracy": 0.8057322911918163,
      "num_tokens": 592939207.0,
      "step": 18577
    },
    {
      "entropy": 0.615428427234292,
      "epoch": 1.709277117440383,
      "grad_norm": 0.16048195958137512,
      "learning_rate": 4.302757076701322e-05,
      "loss": 0.6027,
      "mean_token_accuracy": 0.8116350285708904,
      "num_tokens": 592970275.0,
      "step": 18578
    },
    {
      "entropy": 0.6322948934976012,
      "epoch": 1.709369123811824,
      "grad_norm": 0.15746335685253143,
      "learning_rate": 4.30245039408716e-05,
      "loss": 0.6176,
      "mean_token_accuracy": 0.808678399771452,
      "num_tokens": 593001857.0,
      "step": 18579
    },
    {
      "entropy": 0.6194499339908361,
      "epoch": 1.7094611301832652,
      "grad_norm": 0.15217235684394836,
      "learning_rate": 4.3021437114729965e-05,
      "loss": 0.6166,
      "mean_token_accuracy": 0.8140109330415726,
      "num_tokens": 593034469.0,
      "step": 18580
    },
    {
      "entropy": 0.5614497177302837,
      "epoch": 1.7095531365547063,
      "grad_norm": 0.14820003509521484,
      "learning_rate": 4.3018370288588346e-05,
      "loss": 0.5363,
      "mean_token_accuracy": 0.8315073251724243,
      "num_tokens": 593065844.0,
      "step": 18581
    },
    {
      "entropy": 0.6497903075069189,
      "epoch": 1.7096451429261477,
      "grad_norm": 0.16075065732002258,
      "learning_rate": 4.3015303462446714e-05,
      "loss": 0.6429,
      "mean_token_accuracy": 0.7988933958113194,
      "num_tokens": 593096811.0,
      "step": 18582
    },
    {
      "entropy": 0.6299812868237495,
      "epoch": 1.709737149297589,
      "grad_norm": 0.15604311227798462,
      "learning_rate": 4.3012236636305095e-05,
      "loss": 0.625,
      "mean_token_accuracy": 0.8116488493978977,
      "num_tokens": 593128507.0,
      "step": 18583
    },
    {
      "entropy": 0.5864133748691529,
      "epoch": 1.7098291556690302,
      "grad_norm": 0.1484430879354477,
      "learning_rate": 4.300916981016346e-05,
      "loss": 0.5813,
      "mean_token_accuracy": 0.8229323551058769,
      "num_tokens": 593160909.0,
      "step": 18584
    },
    {
      "entropy": 0.6677423277869821,
      "epoch": 1.7099211620404713,
      "grad_norm": 0.16705970466136932,
      "learning_rate": 4.300610298402184e-05,
      "loss": 0.6765,
      "mean_token_accuracy": 0.7987125739455223,
      "num_tokens": 593192183.0,
      "step": 18585
    },
    {
      "entropy": 0.5304914563894272,
      "epoch": 1.7100131684119124,
      "grad_norm": 0.15982158482074738,
      "learning_rate": 4.300303615788021e-05,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8385265953838825,
      "num_tokens": 593223772.0,
      "step": 18586
    },
    {
      "entropy": 0.5992724755778909,
      "epoch": 1.7101051747833538,
      "grad_norm": 0.1599227637052536,
      "learning_rate": 4.299996933173859e-05,
      "loss": 0.5811,
      "mean_token_accuracy": 0.8200402855873108,
      "num_tokens": 593255681.0,
      "step": 18587
    },
    {
      "entropy": 0.6527821887284517,
      "epoch": 1.7101971811547951,
      "grad_norm": 0.15756507217884064,
      "learning_rate": 4.299690250559696e-05,
      "loss": 0.6332,
      "mean_token_accuracy": 0.8050932921469212,
      "num_tokens": 593286406.0,
      "step": 18588
    },
    {
      "entropy": 0.5930569169577211,
      "epoch": 1.7102891875262363,
      "grad_norm": 0.15308596193790436,
      "learning_rate": 4.2993835679455336e-05,
      "loss": 0.5641,
      "mean_token_accuracy": 0.8223659098148346,
      "num_tokens": 593318250.0,
      "step": 18589
    },
    {
      "entropy": 0.5506706405431032,
      "epoch": 1.7103811938976774,
      "grad_norm": 0.14833520352840424,
      "learning_rate": 4.2990768853313704e-05,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8341872617602348,
      "num_tokens": 593349816.0,
      "step": 18590
    },
    {
      "entropy": 0.5928919622674584,
      "epoch": 1.7104732002691185,
      "grad_norm": 0.15335068106651306,
      "learning_rate": 4.2987702027172086e-05,
      "loss": 0.5865,
      "mean_token_accuracy": 0.8166923820972443,
      "num_tokens": 593381277.0,
      "step": 18591
    },
    {
      "entropy": 0.49029590329155326,
      "epoch": 1.7105652066405599,
      "grad_norm": 0.14688675105571747,
      "learning_rate": 4.2984635201030454e-05,
      "loss": 0.4822,
      "mean_token_accuracy": 0.8560506999492645,
      "num_tokens": 593413512.0,
      "step": 18592
    },
    {
      "entropy": 0.6387984026223421,
      "epoch": 1.7106572130120012,
      "grad_norm": 0.16397725045681,
      "learning_rate": 4.298156837488883e-05,
      "loss": 0.6319,
      "mean_token_accuracy": 0.8070162534713745,
      "num_tokens": 593445530.0,
      "step": 18593
    },
    {
      "entropy": 0.6541102021001279,
      "epoch": 1.7107492193834424,
      "grad_norm": 0.1578264683485031,
      "learning_rate": 4.29785015487472e-05,
      "loss": 0.6464,
      "mean_token_accuracy": 0.8030143827199936,
      "num_tokens": 593477889.0,
      "step": 18594
    },
    {
      "entropy": 0.5018319650553167,
      "epoch": 1.7108412257548835,
      "grad_norm": 0.1416640728712082,
      "learning_rate": 4.297543472260558e-05,
      "loss": 0.4869,
      "mean_token_accuracy": 0.8476288914680481,
      "num_tokens": 593510569.0,
      "step": 18595
    },
    {
      "entropy": 0.631610905751586,
      "epoch": 1.7109332321263246,
      "grad_norm": 0.15584667026996613,
      "learning_rate": 4.297236789646395e-05,
      "loss": 0.6024,
      "mean_token_accuracy": 0.8128025718033314,
      "num_tokens": 593542322.0,
      "step": 18596
    },
    {
      "entropy": 0.5832160264253616,
      "epoch": 1.711025238497766,
      "grad_norm": 0.15273034572601318,
      "learning_rate": 4.296930107032233e-05,
      "loss": 0.5721,
      "mean_token_accuracy": 0.8198876865208149,
      "num_tokens": 593574122.0,
      "step": 18597
    },
    {
      "entropy": 0.5637242272496223,
      "epoch": 1.7111172448692074,
      "grad_norm": 0.14748162031173706,
      "learning_rate": 4.2966234244180695e-05,
      "loss": 0.5567,
      "mean_token_accuracy": 0.8259805962443352,
      "num_tokens": 593606320.0,
      "step": 18598
    },
    {
      "entropy": 0.666810660623014,
      "epoch": 1.7112092512406485,
      "grad_norm": 0.16063693165779114,
      "learning_rate": 4.2963167418039076e-05,
      "loss": 0.6521,
      "mean_token_accuracy": 0.8008303232491016,
      "num_tokens": 593638379.0,
      "step": 18599
    },
    {
      "entropy": 0.596760205924511,
      "epoch": 1.7113012576120896,
      "grad_norm": 0.14878243207931519,
      "learning_rate": 4.2960100591897444e-05,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8193047717213631,
      "num_tokens": 593670276.0,
      "step": 18600
    },
    {
      "entropy": 0.562294140458107,
      "epoch": 1.7113932639835308,
      "grad_norm": 0.15063297748565674,
      "learning_rate": 4.295703376575582e-05,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8325030505657196,
      "num_tokens": 593701462.0,
      "step": 18601
    },
    {
      "entropy": 0.6728185210376978,
      "epoch": 1.711485270354972,
      "grad_norm": 0.1588989943265915,
      "learning_rate": 4.295396693961419e-05,
      "loss": 0.6753,
      "mean_token_accuracy": 0.7972433753311634,
      "num_tokens": 593733182.0,
      "step": 18602
    },
    {
      "entropy": 0.6529406195040792,
      "epoch": 1.7115772767264135,
      "grad_norm": 0.15190042555332184,
      "learning_rate": 4.295090011347257e-05,
      "loss": 0.6521,
      "mean_token_accuracy": 0.8072912730276585,
      "num_tokens": 593765621.0,
      "step": 18603
    },
    {
      "entropy": 0.6220085043460131,
      "epoch": 1.7116692830978546,
      "grad_norm": 0.15609024465084076,
      "learning_rate": 4.294783328733094e-05,
      "loss": 0.6239,
      "mean_token_accuracy": 0.8052593804895878,
      "num_tokens": 593797482.0,
      "step": 18604
    },
    {
      "entropy": 0.5075786628294736,
      "epoch": 1.7117612894692957,
      "grad_norm": 0.14733818173408508,
      "learning_rate": 4.294476646118932e-05,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8422782979905605,
      "num_tokens": 593829111.0,
      "step": 18605
    },
    {
      "entropy": 0.6895252270624042,
      "epoch": 1.7118532958407369,
      "grad_norm": 0.15700210630893707,
      "learning_rate": 4.294169963504769e-05,
      "loss": 0.6679,
      "mean_token_accuracy": 0.7902493998408318,
      "num_tokens": 593860978.0,
      "step": 18606
    },
    {
      "entropy": 0.6157594919204712,
      "epoch": 1.7119453022121782,
      "grad_norm": 0.16611996293067932,
      "learning_rate": 4.2938632808906066e-05,
      "loss": 0.5903,
      "mean_token_accuracy": 0.814691673964262,
      "num_tokens": 593891871.0,
      "step": 18607
    },
    {
      "entropy": 0.68453673645854,
      "epoch": 1.7120373085836196,
      "grad_norm": 0.16279520094394684,
      "learning_rate": 4.293556598276444e-05,
      "loss": 0.6804,
      "mean_token_accuracy": 0.7932364083826542,
      "num_tokens": 593923946.0,
      "step": 18608
    },
    {
      "entropy": 0.5814753184095025,
      "epoch": 1.7121293149550607,
      "grad_norm": 0.1506289541721344,
      "learning_rate": 4.2932499156622816e-05,
      "loss": 0.5743,
      "mean_token_accuracy": 0.822074081748724,
      "num_tokens": 593956095.0,
      "step": 18609
    },
    {
      "entropy": 0.5616172282025218,
      "epoch": 1.7122213213265018,
      "grad_norm": 0.14972500503063202,
      "learning_rate": 4.292943233048119e-05,
      "loss": 0.5481,
      "mean_token_accuracy": 0.832381084561348,
      "num_tokens": 593988275.0,
      "step": 18610
    },
    {
      "entropy": 0.570147093385458,
      "epoch": 1.712313327697943,
      "grad_norm": 0.15751641988754272,
      "learning_rate": 4.292636550433956e-05,
      "loss": 0.5616,
      "mean_token_accuracy": 0.8278515003621578,
      "num_tokens": 594020614.0,
      "step": 18611
    },
    {
      "entropy": 0.5935282539576292,
      "epoch": 1.7124053340693843,
      "grad_norm": 0.15189911425113678,
      "learning_rate": 4.292329867819794e-05,
      "loss": 0.5834,
      "mean_token_accuracy": 0.8229424469172955,
      "num_tokens": 594052298.0,
      "step": 18612
    },
    {
      "entropy": 0.5835447451099753,
      "epoch": 1.7124973404408257,
      "grad_norm": 0.15195365250110626,
      "learning_rate": 4.292023185205631e-05,
      "loss": 0.5761,
      "mean_token_accuracy": 0.8222760073840618,
      "num_tokens": 594084523.0,
      "step": 18613
    },
    {
      "entropy": 0.7331805238500237,
      "epoch": 1.7125893468122668,
      "grad_norm": 0.16579905152320862,
      "learning_rate": 4.291716502591468e-05,
      "loss": 0.7166,
      "mean_token_accuracy": 0.7795041352510452,
      "num_tokens": 594115623.0,
      "step": 18614
    },
    {
      "entropy": 0.5468313274905086,
      "epoch": 1.712681353183708,
      "grad_norm": 0.14878909289836884,
      "learning_rate": 4.291409819977306e-05,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8350576981902122,
      "num_tokens": 594147600.0,
      "step": 18615
    },
    {
      "entropy": 0.7110353969037533,
      "epoch": 1.712773359555149,
      "grad_norm": 0.1619841605424881,
      "learning_rate": 4.291103137363143e-05,
      "loss": 0.7104,
      "mean_token_accuracy": 0.7852285355329514,
      "num_tokens": 594179142.0,
      "step": 18616
    },
    {
      "entropy": 0.61651511490345,
      "epoch": 1.7128653659265904,
      "grad_norm": 0.15254083275794983,
      "learning_rate": 4.2907964547489806e-05,
      "loss": 0.5939,
      "mean_token_accuracy": 0.8169547393918037,
      "num_tokens": 594210485.0,
      "step": 18617
    },
    {
      "entropy": 0.6919538602232933,
      "epoch": 1.7129573722980318,
      "grad_norm": 0.16229814291000366,
      "learning_rate": 4.290489772134818e-05,
      "loss": 0.6876,
      "mean_token_accuracy": 0.786827340722084,
      "num_tokens": 594241941.0,
      "step": 18618
    },
    {
      "entropy": 0.6709566805511713,
      "epoch": 1.713049378669473,
      "grad_norm": 0.1598464846611023,
      "learning_rate": 4.290183089520655e-05,
      "loss": 0.6487,
      "mean_token_accuracy": 0.8000471144914627,
      "num_tokens": 594272569.0,
      "step": 18619
    },
    {
      "entropy": 0.5408893255516887,
      "epoch": 1.713141385040914,
      "grad_norm": 0.15308158099651337,
      "learning_rate": 4.289876406906493e-05,
      "loss": 0.5356,
      "mean_token_accuracy": 0.8355008065700531,
      "num_tokens": 594305046.0,
      "step": 18620
    },
    {
      "entropy": 0.6683670114725828,
      "epoch": 1.7132333914123552,
      "grad_norm": 0.16318552196025848,
      "learning_rate": 4.28956972429233e-05,
      "loss": 0.6521,
      "mean_token_accuracy": 0.8008392192423344,
      "num_tokens": 594336831.0,
      "step": 18621
    },
    {
      "entropy": 0.46764848148450255,
      "epoch": 1.7133253977837966,
      "grad_norm": 0.15609854459762573,
      "learning_rate": 4.289263041678168e-05,
      "loss": 0.4468,
      "mean_token_accuracy": 0.8610377386212349,
      "num_tokens": 594368962.0,
      "step": 18622
    },
    {
      "entropy": 0.6295841059181839,
      "epoch": 1.713417404155238,
      "grad_norm": 0.1586647927761078,
      "learning_rate": 4.288956359064005e-05,
      "loss": 0.6056,
      "mean_token_accuracy": 0.8124643191695213,
      "num_tokens": 594400518.0,
      "step": 18623
    },
    {
      "entropy": 0.5500833373516798,
      "epoch": 1.713509410526679,
      "grad_norm": 0.14682820439338684,
      "learning_rate": 4.288649676449842e-05,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8373046107590199,
      "num_tokens": 594432895.0,
      "step": 18624
    },
    {
      "entropy": 0.6413569077849388,
      "epoch": 1.7136014168981202,
      "grad_norm": 0.16072705388069153,
      "learning_rate": 4.2883429938356797e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8155490085482597,
      "num_tokens": 594464480.0,
      "step": 18625
    },
    {
      "entropy": 0.6710848454385996,
      "epoch": 1.7136934232695613,
      "grad_norm": 0.16332195699214935,
      "learning_rate": 4.288036311221517e-05,
      "loss": 0.6663,
      "mean_token_accuracy": 0.8010462671518326,
      "num_tokens": 594496453.0,
      "step": 18626
    },
    {
      "entropy": 0.5826425906270742,
      "epoch": 1.7137854296410027,
      "grad_norm": 0.14992521703243256,
      "learning_rate": 4.2877296286073546e-05,
      "loss": 0.5871,
      "mean_token_accuracy": 0.8200895600020885,
      "num_tokens": 594528682.0,
      "step": 18627
    },
    {
      "entropy": 0.6211324818432331,
      "epoch": 1.713877436012444,
      "grad_norm": 0.15361258387565613,
      "learning_rate": 4.287422945993192e-05,
      "loss": 0.6037,
      "mean_token_accuracy": 0.817153062671423,
      "num_tokens": 594560805.0,
      "step": 18628
    },
    {
      "entropy": 0.6321671027690172,
      "epoch": 1.7139694423838852,
      "grad_norm": 0.15566207468509674,
      "learning_rate": 4.287116263379029e-05,
      "loss": 0.6229,
      "mean_token_accuracy": 0.8080545626580715,
      "num_tokens": 594592340.0,
      "step": 18629
    },
    {
      "entropy": 0.5528152836486697,
      "epoch": 1.7140614487553263,
      "grad_norm": 0.1499110758304596,
      "learning_rate": 4.286809580764867e-05,
      "loss": 0.5536,
      "mean_token_accuracy": 0.8273102194070816,
      "num_tokens": 594624589.0,
      "step": 18630
    },
    {
      "entropy": 0.6087893852964044,
      "epoch": 1.7141534551267674,
      "grad_norm": 0.15170277655124664,
      "learning_rate": 4.286502898150704e-05,
      "loss": 0.5982,
      "mean_token_accuracy": 0.8183591514825821,
      "num_tokens": 594655824.0,
      "step": 18631
    },
    {
      "entropy": 0.587302245432511,
      "epoch": 1.7142454614982088,
      "grad_norm": 0.1605362743139267,
      "learning_rate": 4.286196215536541e-05,
      "loss": 0.5764,
      "mean_token_accuracy": 0.8215187527239323,
      "num_tokens": 594687884.0,
      "step": 18632
    },
    {
      "entropy": 0.537959910929203,
      "epoch": 1.7143374678696501,
      "grad_norm": 0.15043719112873077,
      "learning_rate": 4.285889532922379e-05,
      "loss": 0.5308,
      "mean_token_accuracy": 0.835751436650753,
      "num_tokens": 594720081.0,
      "step": 18633
    },
    {
      "entropy": 0.6116235014051199,
      "epoch": 1.7144294742410913,
      "grad_norm": 0.15955215692520142,
      "learning_rate": 4.285582850308216e-05,
      "loss": 0.5972,
      "mean_token_accuracy": 0.8211684934794903,
      "num_tokens": 594751942.0,
      "step": 18634
    },
    {
      "entropy": 0.7175960522145033,
      "epoch": 1.7145214806125324,
      "grad_norm": 0.16655300557613373,
      "learning_rate": 4.2852761676940536e-05,
      "loss": 0.7368,
      "mean_token_accuracy": 0.7800209634006023,
      "num_tokens": 594783868.0,
      "step": 18635
    },
    {
      "entropy": 0.606422659009695,
      "epoch": 1.7146134869839735,
      "grad_norm": 0.15336491167545319,
      "learning_rate": 4.284969485079891e-05,
      "loss": 0.5992,
      "mean_token_accuracy": 0.8157277181744576,
      "num_tokens": 594816083.0,
      "step": 18636
    },
    {
      "entropy": 0.6685738554224372,
      "epoch": 1.714705493355415,
      "grad_norm": 0.1609172374010086,
      "learning_rate": 4.284662802465728e-05,
      "loss": 0.6721,
      "mean_token_accuracy": 0.7927752137184143,
      "num_tokens": 594848175.0,
      "step": 18637
    },
    {
      "entropy": 0.6685944106429815,
      "epoch": 1.7147974997268562,
      "grad_norm": 0.16624344885349274,
      "learning_rate": 4.284356119851566e-05,
      "loss": 0.6515,
      "mean_token_accuracy": 0.8017987832427025,
      "num_tokens": 594880075.0,
      "step": 18638
    },
    {
      "entropy": 0.5971534131094813,
      "epoch": 1.7148895060982974,
      "grad_norm": 0.1487657129764557,
      "learning_rate": 4.2840494372374035e-05,
      "loss": 0.5845,
      "mean_token_accuracy": 0.8193048760294914,
      "num_tokens": 594911938.0,
      "step": 18639
    },
    {
      "entropy": 0.5805832715705037,
      "epoch": 1.7149815124697385,
      "grad_norm": 0.15105368196964264,
      "learning_rate": 4.28374275462324e-05,
      "loss": 0.5716,
      "mean_token_accuracy": 0.8234638199210167,
      "num_tokens": 594943426.0,
      "step": 18640
    },
    {
      "entropy": 0.6018551215529442,
      "epoch": 1.7150735188411796,
      "grad_norm": 0.15217280387878418,
      "learning_rate": 4.2834360720090784e-05,
      "loss": 0.579,
      "mean_token_accuracy": 0.818292673677206,
      "num_tokens": 594974938.0,
      "step": 18641
    },
    {
      "entropy": 0.6836679792031646,
      "epoch": 1.715165525212621,
      "grad_norm": 0.1605721116065979,
      "learning_rate": 4.283129389394915e-05,
      "loss": 0.6661,
      "mean_token_accuracy": 0.7966345958411694,
      "num_tokens": 595007187.0,
      "step": 18642
    },
    {
      "entropy": 0.5974672455340624,
      "epoch": 1.7152575315840624,
      "grad_norm": 0.16262133419513702,
      "learning_rate": 4.2828227067807533e-05,
      "loss": 0.5845,
      "mean_token_accuracy": 0.8212598040699959,
      "num_tokens": 595038366.0,
      "step": 18643
    },
    {
      "entropy": 0.6909939255565405,
      "epoch": 1.7153495379555035,
      "grad_norm": 0.16196103394031525,
      "learning_rate": 4.28251602416659e-05,
      "loss": 0.677,
      "mean_token_accuracy": 0.7926809377968311,
      "num_tokens": 595069835.0,
      "step": 18644
    },
    {
      "entropy": 0.5188715439289808,
      "epoch": 1.7154415443269446,
      "grad_norm": 0.14794078469276428,
      "learning_rate": 4.2822093415524276e-05,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8420902639627457,
      "num_tokens": 595102301.0,
      "step": 18645
    },
    {
      "entropy": 0.6405024935957044,
      "epoch": 1.7155335506983858,
      "grad_norm": 0.15343928337097168,
      "learning_rate": 4.281902658938265e-05,
      "loss": 0.608,
      "mean_token_accuracy": 0.8167093768715858,
      "num_tokens": 595132997.0,
      "step": 18646
    },
    {
      "entropy": 0.501690368168056,
      "epoch": 1.7156255570698271,
      "grad_norm": 0.14215484261512756,
      "learning_rate": 4.2815959763241025e-05,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8451055586338043,
      "num_tokens": 595165411.0,
      "step": 18647
    },
    {
      "entropy": 0.4829440275207162,
      "epoch": 1.7157175634412685,
      "grad_norm": 0.14246854186058044,
      "learning_rate": 4.28128929370994e-05,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8537286631762981,
      "num_tokens": 595197853.0,
      "step": 18648
    },
    {
      "entropy": 0.6662945561110973,
      "epoch": 1.7158095698127096,
      "grad_norm": 0.16099239885807037,
      "learning_rate": 4.2809826110957775e-05,
      "loss": 0.6641,
      "mean_token_accuracy": 0.7974800728261471,
      "num_tokens": 595229980.0,
      "step": 18649
    },
    {
      "entropy": 0.6099058310501277,
      "epoch": 1.7159015761841507,
      "grad_norm": 0.15894447267055511,
      "learning_rate": 4.280675928481614e-05,
      "loss": 0.6111,
      "mean_token_accuracy": 0.8184676431119442,
      "num_tokens": 595261830.0,
      "step": 18650
    },
    {
      "entropy": 0.5588706042617559,
      "epoch": 1.7159935825555919,
      "grad_norm": 0.15712472796440125,
      "learning_rate": 4.2803692458674524e-05,
      "loss": 0.5395,
      "mean_token_accuracy": 0.8314060941338539,
      "num_tokens": 595293696.0,
      "step": 18651
    },
    {
      "entropy": 0.683025011792779,
      "epoch": 1.7160855889270332,
      "grad_norm": 0.15997976064682007,
      "learning_rate": 4.280062563253289e-05,
      "loss": 0.6872,
      "mean_token_accuracy": 0.7906922847032547,
      "num_tokens": 595326162.0,
      "step": 18652
    },
    {
      "entropy": 0.6947181578725576,
      "epoch": 1.7161775952984746,
      "grad_norm": 0.1613796502351761,
      "learning_rate": 4.2797558806391266e-05,
      "loss": 0.6891,
      "mean_token_accuracy": 0.7924043126404285,
      "num_tokens": 595357004.0,
      "step": 18653
    },
    {
      "entropy": 0.5700676357373595,
      "epoch": 1.7162696016699157,
      "grad_norm": 0.148630291223526,
      "learning_rate": 4.279449198024964e-05,
      "loss": 0.5702,
      "mean_token_accuracy": 0.8256397545337677,
      "num_tokens": 595388579.0,
      "step": 18654
    },
    {
      "entropy": 0.6656178552657366,
      "epoch": 1.7163616080413568,
      "grad_norm": 0.1622547209262848,
      "learning_rate": 4.2791425154108016e-05,
      "loss": 0.6654,
      "mean_token_accuracy": 0.8001205325126648,
      "num_tokens": 595419746.0,
      "step": 18655
    },
    {
      "entropy": 0.6174825886264443,
      "epoch": 1.716453614412798,
      "grad_norm": 0.15413378179073334,
      "learning_rate": 4.278835832796639e-05,
      "loss": 0.6082,
      "mean_token_accuracy": 0.8127110153436661,
      "num_tokens": 595451784.0,
      "step": 18656
    },
    {
      "entropy": 0.5718026841059327,
      "epoch": 1.7165456207842393,
      "grad_norm": 0.14940981566905975,
      "learning_rate": 4.2785291501824765e-05,
      "loss": 0.5622,
      "mean_token_accuracy": 0.8287006728351116,
      "num_tokens": 595483294.0,
      "step": 18657
    },
    {
      "entropy": 0.5723532382398844,
      "epoch": 1.7166376271556807,
      "grad_norm": 0.15458610653877258,
      "learning_rate": 4.278222467568313e-05,
      "loss": 0.562,
      "mean_token_accuracy": 0.8214116618037224,
      "num_tokens": 595514232.0,
      "step": 18658
    },
    {
      "entropy": 0.49709557369351387,
      "epoch": 1.7167296335271218,
      "grad_norm": 0.14761139452457428,
      "learning_rate": 4.2779157849541514e-05,
      "loss": 0.4866,
      "mean_token_accuracy": 0.8493692278862,
      "num_tokens": 595546988.0,
      "step": 18659
    },
    {
      "entropy": 0.5694780219346285,
      "epoch": 1.716821639898563,
      "grad_norm": 0.147380068898201,
      "learning_rate": 4.277609102339988e-05,
      "loss": 0.5547,
      "mean_token_accuracy": 0.8280248716473579,
      "num_tokens": 595579042.0,
      "step": 18660
    },
    {
      "entropy": 0.6194128915667534,
      "epoch": 1.716913646270004,
      "grad_norm": 0.1535666584968567,
      "learning_rate": 4.277302419725826e-05,
      "loss": 0.5982,
      "mean_token_accuracy": 0.8149301409721375,
      "num_tokens": 595610802.0,
      "step": 18661
    },
    {
      "entropy": 0.5508185001090169,
      "epoch": 1.7170056526414454,
      "grad_norm": 0.15540023148059845,
      "learning_rate": 4.276995737111663e-05,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8313575424253941,
      "num_tokens": 595643212.0,
      "step": 18662
    },
    {
      "entropy": 0.6080079302191734,
      "epoch": 1.7170976590128868,
      "grad_norm": 0.14971823990345,
      "learning_rate": 4.2766890544975006e-05,
      "loss": 0.6026,
      "mean_token_accuracy": 0.8134148418903351,
      "num_tokens": 595675745.0,
      "step": 18663
    },
    {
      "entropy": 0.6094846557825804,
      "epoch": 1.717189665384328,
      "grad_norm": 0.15146668255329132,
      "learning_rate": 4.276382371883338e-05,
      "loss": 0.5954,
      "mean_token_accuracy": 0.820320438593626,
      "num_tokens": 595708339.0,
      "step": 18664
    },
    {
      "entropy": 0.674252447206527,
      "epoch": 1.717281671755769,
      "grad_norm": 0.16335538029670715,
      "learning_rate": 4.2760756892691755e-05,
      "loss": 0.6609,
      "mean_token_accuracy": 0.7977587506175041,
      "num_tokens": 595740735.0,
      "step": 18665
    },
    {
      "entropy": 0.6921152547001839,
      "epoch": 1.7173736781272102,
      "grad_norm": 0.15722788870334625,
      "learning_rate": 4.275769006655013e-05,
      "loss": 0.6817,
      "mean_token_accuracy": 0.793195515871048,
      "num_tokens": 595772998.0,
      "step": 18666
    },
    {
      "entropy": 0.6602446343749762,
      "epoch": 1.7174656844986516,
      "grad_norm": 0.15565544366836548,
      "learning_rate": 4.2754623240408505e-05,
      "loss": 0.6498,
      "mean_token_accuracy": 0.7975087910890579,
      "num_tokens": 595805182.0,
      "step": 18667
    },
    {
      "entropy": 0.6754110646434128,
      "epoch": 1.717557690870093,
      "grad_norm": 0.15900208055973053,
      "learning_rate": 4.275155641426687e-05,
      "loss": 0.6643,
      "mean_token_accuracy": 0.7953316867351532,
      "num_tokens": 595837204.0,
      "step": 18668
    },
    {
      "entropy": 0.59120699390769,
      "epoch": 1.717649697241534,
      "grad_norm": 0.16181035339832306,
      "learning_rate": 4.2748489588125254e-05,
      "loss": 0.569,
      "mean_token_accuracy": 0.8232941254973412,
      "num_tokens": 595869308.0,
      "step": 18669
    },
    {
      "entropy": 0.636240791529417,
      "epoch": 1.7177417036129752,
      "grad_norm": 0.15989570319652557,
      "learning_rate": 4.274542276198363e-05,
      "loss": 0.6224,
      "mean_token_accuracy": 0.8137090392410755,
      "num_tokens": 595901436.0,
      "step": 18670
    },
    {
      "entropy": 0.6276546004228294,
      "epoch": 1.7178337099844163,
      "grad_norm": 0.15501829981803894,
      "learning_rate": 4.2742355935841997e-05,
      "loss": 0.6259,
      "mean_token_accuracy": 0.8085947409272194,
      "num_tokens": 595934008.0,
      "step": 18671
    },
    {
      "entropy": 0.6944107040762901,
      "epoch": 1.7179257163558577,
      "grad_norm": 0.1676190197467804,
      "learning_rate": 4.273928910970038e-05,
      "loss": 0.69,
      "mean_token_accuracy": 0.7906716801226139,
      "num_tokens": 595965603.0,
      "step": 18672
    },
    {
      "entropy": 0.6287892977707088,
      "epoch": 1.718017722727299,
      "grad_norm": 0.15748797357082367,
      "learning_rate": 4.2736222283558746e-05,
      "loss": 0.6302,
      "mean_token_accuracy": 0.8080521449446678,
      "num_tokens": 595997687.0,
      "step": 18673
    },
    {
      "entropy": 0.5571961207315326,
      "epoch": 1.7181097290987402,
      "grad_norm": 0.14529171586036682,
      "learning_rate": 4.273315545741712e-05,
      "loss": 0.5578,
      "mean_token_accuracy": 0.8319132477045059,
      "num_tokens": 596029778.0,
      "step": 18674
    },
    {
      "entropy": 0.6466646594926715,
      "epoch": 1.7182017354701813,
      "grad_norm": 0.15844261646270752,
      "learning_rate": 4.2730088631275495e-05,
      "loss": 0.6254,
      "mean_token_accuracy": 0.8074357956647873,
      "num_tokens": 596060705.0,
      "step": 18675
    },
    {
      "entropy": 0.5728799626231194,
      "epoch": 1.7182937418416224,
      "grad_norm": 0.15702159702777863,
      "learning_rate": 4.272702180513387e-05,
      "loss": 0.5637,
      "mean_token_accuracy": 0.824549924582243,
      "num_tokens": 596091572.0,
      "step": 18676
    },
    {
      "entropy": 0.6132001150399446,
      "epoch": 1.7183857482130638,
      "grad_norm": 0.15710942447185516,
      "learning_rate": 4.2723954978992244e-05,
      "loss": 0.6039,
      "mean_token_accuracy": 0.8140703104436398,
      "num_tokens": 596123179.0,
      "step": 18677
    },
    {
      "entropy": 0.6213192781433463,
      "epoch": 1.7184777545845051,
      "grad_norm": 0.155583918094635,
      "learning_rate": 4.272088815285062e-05,
      "loss": 0.608,
      "mean_token_accuracy": 0.8151810839772224,
      "num_tokens": 596155472.0,
      "step": 18678
    },
    {
      "entropy": 0.6261079029645771,
      "epoch": 1.7185697609559463,
      "grad_norm": 0.15381696820259094,
      "learning_rate": 4.271782132670899e-05,
      "loss": 0.6112,
      "mean_token_accuracy": 0.813203439116478,
      "num_tokens": 596186947.0,
      "step": 18679
    },
    {
      "entropy": 0.6504339454695582,
      "epoch": 1.7186617673273874,
      "grad_norm": 0.16396348178386688,
      "learning_rate": 4.271475450056737e-05,
      "loss": 0.6513,
      "mean_token_accuracy": 0.8006839230656624,
      "num_tokens": 596219715.0,
      "step": 18680
    },
    {
      "entropy": 0.5987766664475203,
      "epoch": 1.7187537736988285,
      "grad_norm": 0.1576036810874939,
      "learning_rate": 4.2711687674425736e-05,
      "loss": 0.597,
      "mean_token_accuracy": 0.8170289061963558,
      "num_tokens": 596252173.0,
      "step": 18681
    },
    {
      "entropy": 0.5959054529666901,
      "epoch": 1.71884578007027,
      "grad_norm": 0.15041406452655792,
      "learning_rate": 4.270862084828412e-05,
      "loss": 0.5854,
      "mean_token_accuracy": 0.8174444995820522,
      "num_tokens": 596283926.0,
      "step": 18682
    },
    {
      "entropy": 0.6035791873000562,
      "epoch": 1.7189377864417112,
      "grad_norm": 0.16185343265533447,
      "learning_rate": 4.2705554022142485e-05,
      "loss": 0.5966,
      "mean_token_accuracy": 0.8171377889811993,
      "num_tokens": 596315813.0,
      "step": 18683
    },
    {
      "entropy": 0.6358019234612584,
      "epoch": 1.7190297928131524,
      "grad_norm": 0.1549578160047531,
      "learning_rate": 4.270248719600086e-05,
      "loss": 0.6288,
      "mean_token_accuracy": 0.8073740936815739,
      "num_tokens": 596347942.0,
      "step": 18684
    },
    {
      "entropy": 0.6220850721001625,
      "epoch": 1.7191217991845935,
      "grad_norm": 0.15170170366764069,
      "learning_rate": 4.2699420369859235e-05,
      "loss": 0.605,
      "mean_token_accuracy": 0.8104937300086021,
      "num_tokens": 596380053.0,
      "step": 18685
    },
    {
      "entropy": 0.5669726165942848,
      "epoch": 1.7192138055560346,
      "grad_norm": 0.15453410148620605,
      "learning_rate": 4.269635354371761e-05,
      "loss": 0.5445,
      "mean_token_accuracy": 0.829756636172533,
      "num_tokens": 596411631.0,
      "step": 18686
    },
    {
      "entropy": 0.6173996482975781,
      "epoch": 1.719305811927476,
      "grad_norm": 0.16165941953659058,
      "learning_rate": 4.2693286717575984e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.814340453594923,
      "num_tokens": 596444038.0,
      "step": 18687
    },
    {
      "entropy": 0.7150531634688377,
      "epoch": 1.7193978182989174,
      "grad_norm": 0.16188156604766846,
      "learning_rate": 4.269021989143436e-05,
      "loss": 0.7014,
      "mean_token_accuracy": 0.7841871343553066,
      "num_tokens": 596475350.0,
      "step": 18688
    },
    {
      "entropy": 0.6886276043951511,
      "epoch": 1.7194898246703585,
      "grad_norm": 0.16041795909404755,
      "learning_rate": 4.2687153065292727e-05,
      "loss": 0.6883,
      "mean_token_accuracy": 0.7906437404453754,
      "num_tokens": 596506937.0,
      "step": 18689
    },
    {
      "entropy": 0.6031605498865247,
      "epoch": 1.7195818310417996,
      "grad_norm": 0.15770085155963898,
      "learning_rate": 4.268408623915111e-05,
      "loss": 0.6004,
      "mean_token_accuracy": 0.8175084143877029,
      "num_tokens": 596539138.0,
      "step": 18690
    },
    {
      "entropy": 0.6308001633733511,
      "epoch": 1.7196738374132408,
      "grad_norm": 0.15663954615592957,
      "learning_rate": 4.2681019413009476e-05,
      "loss": 0.6114,
      "mean_token_accuracy": 0.8096433319151402,
      "num_tokens": 596570959.0,
      "step": 18691
    },
    {
      "entropy": 0.7292324248701334,
      "epoch": 1.7197658437846821,
      "grad_norm": 0.169825941324234,
      "learning_rate": 4.267795258686785e-05,
      "loss": 0.7242,
      "mean_token_accuracy": 0.7836735025048256,
      "num_tokens": 596602632.0,
      "step": 18692
    },
    {
      "entropy": 0.614569615572691,
      "epoch": 1.7198578501561235,
      "grad_norm": 0.1573660522699356,
      "learning_rate": 4.2674885760726225e-05,
      "loss": 0.6003,
      "mean_token_accuracy": 0.8146829083561897,
      "num_tokens": 596634634.0,
      "step": 18693
    },
    {
      "entropy": 0.631322493776679,
      "epoch": 1.7199498565275646,
      "grad_norm": 0.15685708820819855,
      "learning_rate": 4.26718189345846e-05,
      "loss": 0.6178,
      "mean_token_accuracy": 0.813079085201025,
      "num_tokens": 596666933.0,
      "step": 18694
    },
    {
      "entropy": 0.6036824528127909,
      "epoch": 1.7200418628990057,
      "grad_norm": 0.152161106467247,
      "learning_rate": 4.2668752108442974e-05,
      "loss": 0.6067,
      "mean_token_accuracy": 0.8151000142097473,
      "num_tokens": 596698742.0,
      "step": 18695
    },
    {
      "entropy": 0.5698661007918417,
      "epoch": 1.7201338692704469,
      "grad_norm": 0.1624162346124649,
      "learning_rate": 4.266568528230135e-05,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8303275071084499,
      "num_tokens": 596730713.0,
      "step": 18696
    },
    {
      "entropy": 0.6620407924056053,
      "epoch": 1.7202258756418882,
      "grad_norm": 0.1646869033575058,
      "learning_rate": 4.266261845615972e-05,
      "loss": 0.6508,
      "mean_token_accuracy": 0.8043395429849625,
      "num_tokens": 596763170.0,
      "step": 18697
    },
    {
      "entropy": 0.6176734436303377,
      "epoch": 1.7203178820133296,
      "grad_norm": 0.15773169696331024,
      "learning_rate": 4.26595516300181e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.8107342720031738,
      "num_tokens": 596794832.0,
      "step": 18698
    },
    {
      "entropy": 0.6664251442998648,
      "epoch": 1.7204098883847707,
      "grad_norm": 0.15685078501701355,
      "learning_rate": 4.265648480387647e-05,
      "loss": 0.6634,
      "mean_token_accuracy": 0.7951051853597164,
      "num_tokens": 596826689.0,
      "step": 18699
    },
    {
      "entropy": 0.711011603474617,
      "epoch": 1.7205018947562118,
      "grad_norm": 0.165338933467865,
      "learning_rate": 4.265341797773484e-05,
      "loss": 0.7095,
      "mean_token_accuracy": 0.7856147848069668,
      "num_tokens": 596857974.0,
      "step": 18700
    },
    {
      "entropy": 0.7118798196315765,
      "epoch": 1.720593901127653,
      "grad_norm": 0.16702161729335785,
      "learning_rate": 4.265035115159322e-05,
      "loss": 0.7039,
      "mean_token_accuracy": 0.7851848155260086,
      "num_tokens": 596889723.0,
      "step": 18701
    },
    {
      "entropy": 0.5661494475789368,
      "epoch": 1.7206859074990943,
      "grad_norm": 0.15642192959785461,
      "learning_rate": 4.264728432545159e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8276655748486519,
      "num_tokens": 596922170.0,
      "step": 18702
    },
    {
      "entropy": 0.59000409161672,
      "epoch": 1.7207779138705357,
      "grad_norm": 0.16000910103321075,
      "learning_rate": 4.264421749930997e-05,
      "loss": 0.581,
      "mean_token_accuracy": 0.8202684335410595,
      "num_tokens": 596953406.0,
      "step": 18703
    },
    {
      "entropy": 0.4910478317178786,
      "epoch": 1.7208699202419768,
      "grad_norm": 0.1414054036140442,
      "learning_rate": 4.264115067316834e-05,
      "loss": 0.4614,
      "mean_token_accuracy": 0.854062981903553,
      "num_tokens": 596985182.0,
      "step": 18704
    },
    {
      "entropy": 0.49088639626279473,
      "epoch": 1.720961926613418,
      "grad_norm": 0.14936785399913788,
      "learning_rate": 4.2638083847026714e-05,
      "loss": 0.4755,
      "mean_token_accuracy": 0.850143875926733,
      "num_tokens": 597016562.0,
      "step": 18705
    },
    {
      "entropy": 0.6131110563874245,
      "epoch": 1.721053932984859,
      "grad_norm": 0.14996004104614258,
      "learning_rate": 4.263501702088509e-05,
      "loss": 0.6035,
      "mean_token_accuracy": 0.8123840354382992,
      "num_tokens": 597049018.0,
      "step": 18706
    },
    {
      "entropy": 0.5263089369982481,
      "epoch": 1.7211459393563004,
      "grad_norm": 0.14629362523555756,
      "learning_rate": 4.2631950194743463e-05,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8434848301112652,
      "num_tokens": 597081250.0,
      "step": 18707
    },
    {
      "entropy": 0.615070603787899,
      "epoch": 1.7212379457277418,
      "grad_norm": 0.16337715089321136,
      "learning_rate": 4.262888336860184e-05,
      "loss": 0.5948,
      "mean_token_accuracy": 0.8175633363425732,
      "num_tokens": 597113067.0,
      "step": 18708
    },
    {
      "entropy": 0.6993865789845586,
      "epoch": 1.721329952099183,
      "grad_norm": 0.16392530500888824,
      "learning_rate": 4.262581654246021e-05,
      "loss": 0.686,
      "mean_token_accuracy": 0.7866712287068367,
      "num_tokens": 597144708.0,
      "step": 18709
    },
    {
      "entropy": 0.4690217664465308,
      "epoch": 1.721421958470624,
      "grad_norm": 0.1412772238254547,
      "learning_rate": 4.262274971631858e-05,
      "loss": 0.4665,
      "mean_token_accuracy": 0.8539780490100384,
      "num_tokens": 597176969.0,
      "step": 18710
    },
    {
      "entropy": 0.49257620982825756,
      "epoch": 1.7215139648420652,
      "grad_norm": 0.14427898824214935,
      "learning_rate": 4.261968289017696e-05,
      "loss": 0.4724,
      "mean_token_accuracy": 0.8518822826445103,
      "num_tokens": 597209217.0,
      "step": 18711
    },
    {
      "entropy": 0.7002028506249189,
      "epoch": 1.7216059712135066,
      "grad_norm": 0.1613532453775406,
      "learning_rate": 4.261661606403533e-05,
      "loss": 0.6934,
      "mean_token_accuracy": 0.791912630200386,
      "num_tokens": 597241533.0,
      "step": 18712
    },
    {
      "entropy": 0.624772546812892,
      "epoch": 1.721697977584948,
      "grad_norm": 0.15747645497322083,
      "learning_rate": 4.2613549237893705e-05,
      "loss": 0.6081,
      "mean_token_accuracy": 0.813452884554863,
      "num_tokens": 597273727.0,
      "step": 18713
    },
    {
      "entropy": 0.624244986101985,
      "epoch": 1.721789983956389,
      "grad_norm": 0.16179518401622772,
      "learning_rate": 4.261048241175208e-05,
      "loss": 0.6109,
      "mean_token_accuracy": 0.8125009387731552,
      "num_tokens": 597303909.0,
      "step": 18714
    },
    {
      "entropy": 0.6967216841876507,
      "epoch": 1.7218819903278302,
      "grad_norm": 0.16150091588497162,
      "learning_rate": 4.2607415585610454e-05,
      "loss": 0.6837,
      "mean_token_accuracy": 0.7919182516634464,
      "num_tokens": 597336077.0,
      "step": 18715
    },
    {
      "entropy": 0.6474388316273689,
      "epoch": 1.7219739966992713,
      "grad_norm": 0.16787803173065186,
      "learning_rate": 4.260434875946883e-05,
      "loss": 0.6495,
      "mean_token_accuracy": 0.8039448596537113,
      "num_tokens": 597367353.0,
      "step": 18716
    },
    {
      "entropy": 0.650826022028923,
      "epoch": 1.7220660030707127,
      "grad_norm": 0.1590779721736908,
      "learning_rate": 4.26012819333272e-05,
      "loss": 0.6419,
      "mean_token_accuracy": 0.8000667542219162,
      "num_tokens": 597399375.0,
      "step": 18717
    },
    {
      "entropy": 0.495645085349679,
      "epoch": 1.722158009442154,
      "grad_norm": 0.14875507354736328,
      "learning_rate": 4.259821510718557e-05,
      "loss": 0.4866,
      "mean_token_accuracy": 0.8497687205672264,
      "num_tokens": 597431580.0,
      "step": 18718
    },
    {
      "entropy": 0.5838420062791556,
      "epoch": 1.7222500158135952,
      "grad_norm": 0.16324028372764587,
      "learning_rate": 4.259514828104395e-05,
      "loss": 0.5801,
      "mean_token_accuracy": 0.8188368193805218,
      "num_tokens": 597463220.0,
      "step": 18719
    },
    {
      "entropy": 0.5132117597386241,
      "epoch": 1.7223420221850363,
      "grad_norm": 0.15814091265201569,
      "learning_rate": 4.259208145490232e-05,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8407192565500736,
      "num_tokens": 597494109.0,
      "step": 18720
    },
    {
      "entropy": 0.6789046116173267,
      "epoch": 1.7224340285564774,
      "grad_norm": 0.16268327832221985,
      "learning_rate": 4.2589014628760695e-05,
      "loss": 0.6812,
      "mean_token_accuracy": 0.7956169210374355,
      "num_tokens": 597526019.0,
      "step": 18721
    },
    {
      "entropy": 0.7133170869201422,
      "epoch": 1.7225260349279188,
      "grad_norm": 0.16561196744441986,
      "learning_rate": 4.258594780261907e-05,
      "loss": 0.7228,
      "mean_token_accuracy": 0.7819140255451202,
      "num_tokens": 597558274.0,
      "step": 18722
    },
    {
      "entropy": 0.4181837602518499,
      "epoch": 1.7226180412993601,
      "grad_norm": 0.1374886929988861,
      "learning_rate": 4.2582880976477444e-05,
      "loss": 0.4046,
      "mean_token_accuracy": 0.8764131627976894,
      "num_tokens": 597590723.0,
      "step": 18723
    },
    {
      "entropy": 0.5985846053808928,
      "epoch": 1.7227100476708013,
      "grad_norm": 0.158891960978508,
      "learning_rate": 4.257981415033582e-05,
      "loss": 0.5809,
      "mean_token_accuracy": 0.8216747418045998,
      "num_tokens": 597622515.0,
      "step": 18724
    },
    {
      "entropy": 0.5627135913819075,
      "epoch": 1.7228020540422424,
      "grad_norm": 0.14694567024707794,
      "learning_rate": 4.2576747324194194e-05,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8317748531699181,
      "num_tokens": 597654778.0,
      "step": 18725
    },
    {
      "entropy": 0.5326122101396322,
      "epoch": 1.7228940604136835,
      "grad_norm": 0.16076074540615082,
      "learning_rate": 4.257368049805257e-05,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8386765196919441,
      "num_tokens": 597687208.0,
      "step": 18726
    },
    {
      "entropy": 0.5935116456821561,
      "epoch": 1.722986066785125,
      "grad_norm": 0.15175855159759521,
      "learning_rate": 4.257061367191094e-05,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8250058926641941,
      "num_tokens": 597718994.0,
      "step": 18727
    },
    {
      "entropy": 0.575743816094473,
      "epoch": 1.7230780731565662,
      "grad_norm": 0.1446414291858673,
      "learning_rate": 4.256754684576931e-05,
      "loss": 0.5471,
      "mean_token_accuracy": 0.8286002315580845,
      "num_tokens": 597751595.0,
      "step": 18728
    },
    {
      "entropy": 0.6278365598991513,
      "epoch": 1.7231700795280074,
      "grad_norm": 0.15558558702468872,
      "learning_rate": 4.256448001962769e-05,
      "loss": 0.6292,
      "mean_token_accuracy": 0.8084704130887985,
      "num_tokens": 597783857.0,
      "step": 18729
    },
    {
      "entropy": 0.6372435628436506,
      "epoch": 1.7232620858994485,
      "grad_norm": 0.1581469178199768,
      "learning_rate": 4.256141319348607e-05,
      "loss": 0.6275,
      "mean_token_accuracy": 0.8081312030553818,
      "num_tokens": 597816010.0,
      "step": 18730
    },
    {
      "entropy": 0.5910394042730331,
      "epoch": 1.7233540922708896,
      "grad_norm": 0.15322600305080414,
      "learning_rate": 4.2558346367344435e-05,
      "loss": 0.5673,
      "mean_token_accuracy": 0.8249485641717911,
      "num_tokens": 597847975.0,
      "step": 18731
    },
    {
      "entropy": 0.6688684672117233,
      "epoch": 1.723446098642331,
      "grad_norm": 0.15576985478401184,
      "learning_rate": 4.2555279541202816e-05,
      "loss": 0.6569,
      "mean_token_accuracy": 0.7974669151008129,
      "num_tokens": 597880613.0,
      "step": 18732
    },
    {
      "entropy": 0.46648225747048855,
      "epoch": 1.7235381050137724,
      "grad_norm": 0.13949987292289734,
      "learning_rate": 4.2552212715061184e-05,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8620974346995354,
      "num_tokens": 597913049.0,
      "step": 18733
    },
    {
      "entropy": 0.5572111615911126,
      "epoch": 1.7236301113852135,
      "grad_norm": 0.15273499488830566,
      "learning_rate": 4.254914588891956e-05,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8354408778250217,
      "num_tokens": 597945369.0,
      "step": 18734
    },
    {
      "entropy": 0.6473877690732479,
      "epoch": 1.7237221177566546,
      "grad_norm": 0.162742480635643,
      "learning_rate": 4.254607906277793e-05,
      "loss": 0.6446,
      "mean_token_accuracy": 0.8082365840673447,
      "num_tokens": 597977048.0,
      "step": 18735
    },
    {
      "entropy": 0.5464080623351038,
      "epoch": 1.7238141241280958,
      "grad_norm": 0.1476314514875412,
      "learning_rate": 4.254301223663631e-05,
      "loss": 0.5345,
      "mean_token_accuracy": 0.8306864500045776,
      "num_tokens": 598008701.0,
      "step": 18736
    },
    {
      "entropy": 0.6263816431164742,
      "epoch": 1.7239061304995371,
      "grad_norm": 0.15517431497573853,
      "learning_rate": 4.253994541049468e-05,
      "loss": 0.6152,
      "mean_token_accuracy": 0.8169272020459175,
      "num_tokens": 598041469.0,
      "step": 18737
    },
    {
      "entropy": 0.594895962625742,
      "epoch": 1.7239981368709785,
      "grad_norm": 0.15397091209888458,
      "learning_rate": 4.253687858435306e-05,
      "loss": 0.5876,
      "mean_token_accuracy": 0.8178054988384247,
      "num_tokens": 598073500.0,
      "step": 18738
    },
    {
      "entropy": 0.6562718711793423,
      "epoch": 1.7240901432424196,
      "grad_norm": 0.15401093661785126,
      "learning_rate": 4.2533811758211425e-05,
      "loss": 0.6575,
      "mean_token_accuracy": 0.799934484064579,
      "num_tokens": 598105719.0,
      "step": 18739
    },
    {
      "entropy": 0.6309219091199338,
      "epoch": 1.7241821496138607,
      "grad_norm": 0.15146812796592712,
      "learning_rate": 4.2530744932069807e-05,
      "loss": 0.6137,
      "mean_token_accuracy": 0.810460738837719,
      "num_tokens": 598137831.0,
      "step": 18740
    },
    {
      "entropy": 0.5972519954666495,
      "epoch": 1.7242741559853019,
      "grad_norm": 0.1565370112657547,
      "learning_rate": 4.2527678105928174e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8168376870453358,
      "num_tokens": 598168826.0,
      "step": 18741
    },
    {
      "entropy": 0.6347688110545278,
      "epoch": 1.7243661623567432,
      "grad_norm": 0.15974849462509155,
      "learning_rate": 4.2524611279786556e-05,
      "loss": 0.6394,
      "mean_token_accuracy": 0.8053505085408688,
      "num_tokens": 598199776.0,
      "step": 18742
    },
    {
      "entropy": 0.6417576093226671,
      "epoch": 1.7244581687281846,
      "grad_norm": 0.1535370945930481,
      "learning_rate": 4.2521544453644924e-05,
      "loss": 0.6292,
      "mean_token_accuracy": 0.8074252381920815,
      "num_tokens": 598232192.0,
      "step": 18743
    },
    {
      "entropy": 0.5634278343059123,
      "epoch": 1.7245501750996257,
      "grad_norm": 0.1406482309103012,
      "learning_rate": 4.25184776275033e-05,
      "loss": 0.549,
      "mean_token_accuracy": 0.8299289755523205,
      "num_tokens": 598264451.0,
      "step": 18744
    },
    {
      "entropy": 0.6123904511332512,
      "epoch": 1.7246421814710668,
      "grad_norm": 0.16151271760463715,
      "learning_rate": 4.251541080136167e-05,
      "loss": 0.6081,
      "mean_token_accuracy": 0.8128625191748142,
      "num_tokens": 598296931.0,
      "step": 18745
    },
    {
      "entropy": 0.6412615654990077,
      "epoch": 1.724734187842508,
      "grad_norm": 0.1558612883090973,
      "learning_rate": 4.251234397522005e-05,
      "loss": 0.6443,
      "mean_token_accuracy": 0.8058784939348698,
      "num_tokens": 598329018.0,
      "step": 18746
    },
    {
      "entropy": 0.7409670874476433,
      "epoch": 1.7248261942139493,
      "grad_norm": 0.16784979403018951,
      "learning_rate": 4.250927714907842e-05,
      "loss": 0.7358,
      "mean_token_accuracy": 0.7795825228095055,
      "num_tokens": 598360473.0,
      "step": 18747
    },
    {
      "entropy": 0.6342871226370335,
      "epoch": 1.7249182005853907,
      "grad_norm": 0.15103591978549957,
      "learning_rate": 4.25062103229368e-05,
      "loss": 0.6249,
      "mean_token_accuracy": 0.8091719597578049,
      "num_tokens": 598393241.0,
      "step": 18748
    },
    {
      "entropy": 0.657616812735796,
      "epoch": 1.7250102069568318,
      "grad_norm": 0.16119785606861115,
      "learning_rate": 4.2503143496795165e-05,
      "loss": 0.6398,
      "mean_token_accuracy": 0.8063726462423801,
      "num_tokens": 598425366.0,
      "step": 18749
    },
    {
      "entropy": 0.6470424719154835,
      "epoch": 1.725102213328273,
      "grad_norm": 0.15436339378356934,
      "learning_rate": 4.2500076670653546e-05,
      "loss": 0.64,
      "mean_token_accuracy": 0.8073941431939602,
      "num_tokens": 598457946.0,
      "step": 18750
    },
    {
      "entropy": 0.6543567646294832,
      "epoch": 1.725194219699714,
      "grad_norm": 0.1587587147951126,
      "learning_rate": 4.2497009844511914e-05,
      "loss": 0.6381,
      "mean_token_accuracy": 0.8027055338025093,
      "num_tokens": 598489925.0,
      "step": 18751
    },
    {
      "entropy": 0.6821699580177665,
      "epoch": 1.7252862260711554,
      "grad_norm": 0.1601942479610443,
      "learning_rate": 4.249394301837029e-05,
      "loss": 0.672,
      "mean_token_accuracy": 0.7943791747093201,
      "num_tokens": 598522216.0,
      "step": 18752
    },
    {
      "entropy": 0.7177654129918665,
      "epoch": 1.7253782324425968,
      "grad_norm": 0.15927040576934814,
      "learning_rate": 4.2490876192228663e-05,
      "loss": 0.7127,
      "mean_token_accuracy": 0.7837658897042274,
      "num_tokens": 598554899.0,
      "step": 18753
    },
    {
      "entropy": 0.6901990734040737,
      "epoch": 1.725470238814038,
      "grad_norm": 0.16157905757427216,
      "learning_rate": 4.248780936608704e-05,
      "loss": 0.6801,
      "mean_token_accuracy": 0.7909192144870758,
      "num_tokens": 598586770.0,
      "step": 18754
    },
    {
      "entropy": 0.5529921911656857,
      "epoch": 1.725562245185479,
      "grad_norm": 0.15242141485214233,
      "learning_rate": 4.248474253994541e-05,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8297051675617695,
      "num_tokens": 598618755.0,
      "step": 18755
    },
    {
      "entropy": 0.6692357277497649,
      "epoch": 1.7256542515569202,
      "grad_norm": 0.1602068990468979,
      "learning_rate": 4.248167571380379e-05,
      "loss": 0.6649,
      "mean_token_accuracy": 0.7971430271863937,
      "num_tokens": 598650669.0,
      "step": 18756
    },
    {
      "entropy": 0.5985615812242031,
      "epoch": 1.7257462579283616,
      "grad_norm": 0.160661980509758,
      "learning_rate": 4.2478608887662155e-05,
      "loss": 0.5745,
      "mean_token_accuracy": 0.8214727565646172,
      "num_tokens": 598682614.0,
      "step": 18757
    },
    {
      "entropy": 0.54968227352947,
      "epoch": 1.725838264299803,
      "grad_norm": 0.15941452980041504,
      "learning_rate": 4.247554206152054e-05,
      "loss": 0.5403,
      "mean_token_accuracy": 0.8318698853254318,
      "num_tokens": 598714065.0,
      "step": 18758
    },
    {
      "entropy": 0.5473819710314274,
      "epoch": 1.725930270671244,
      "grad_norm": 0.15845274925231934,
      "learning_rate": 4.2472475235378905e-05,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8392169997096062,
      "num_tokens": 598746790.0,
      "step": 18759
    },
    {
      "entropy": 0.5581188881769776,
      "epoch": 1.7260222770426852,
      "grad_norm": 0.15249744057655334,
      "learning_rate": 4.246940840923728e-05,
      "loss": 0.5499,
      "mean_token_accuracy": 0.828160535544157,
      "num_tokens": 598778376.0,
      "step": 18760
    },
    {
      "entropy": 0.6558958664536476,
      "epoch": 1.7261142834141263,
      "grad_norm": 0.15948475897312164,
      "learning_rate": 4.246634158309566e-05,
      "loss": 0.6495,
      "mean_token_accuracy": 0.8012000732123852,
      "num_tokens": 598810833.0,
      "step": 18761
    },
    {
      "entropy": 0.6048575583845377,
      "epoch": 1.7262062897855677,
      "grad_norm": 0.14889660477638245,
      "learning_rate": 4.246327475695403e-05,
      "loss": 0.5872,
      "mean_token_accuracy": 0.8191363997757435,
      "num_tokens": 598843049.0,
      "step": 18762
    },
    {
      "entropy": 0.4974434822797775,
      "epoch": 1.726298296157009,
      "grad_norm": 0.14991061389446259,
      "learning_rate": 4.246020793081241e-05,
      "loss": 0.4994,
      "mean_token_accuracy": 0.84335857629776,
      "num_tokens": 598874595.0,
      "step": 18763
    },
    {
      "entropy": 0.6949564013630152,
      "epoch": 1.7263903025284502,
      "grad_norm": 0.1624458283185959,
      "learning_rate": 4.245714110467078e-05,
      "loss": 0.687,
      "mean_token_accuracy": 0.7918921299278736,
      "num_tokens": 598906033.0,
      "step": 18764
    },
    {
      "entropy": 0.5929019497707486,
      "epoch": 1.7264823088998913,
      "grad_norm": 0.1479976326227188,
      "learning_rate": 4.245407427852915e-05,
      "loss": 0.5904,
      "mean_token_accuracy": 0.8197917826473713,
      "num_tokens": 598938739.0,
      "step": 18765
    },
    {
      "entropy": 0.5651146108284593,
      "epoch": 1.7265743152713324,
      "grad_norm": 0.1504720002412796,
      "learning_rate": 4.245100745238753e-05,
      "loss": 0.5586,
      "mean_token_accuracy": 0.8283776082098484,
      "num_tokens": 598970817.0,
      "step": 18766
    },
    {
      "entropy": 0.5779312485828996,
      "epoch": 1.7266663216427738,
      "grad_norm": 0.14650224149227142,
      "learning_rate": 4.24479406262459e-05,
      "loss": 0.5585,
      "mean_token_accuracy": 0.8270390368998051,
      "num_tokens": 599002599.0,
      "step": 18767
    },
    {
      "entropy": 0.5601243739947677,
      "epoch": 1.7267583280142151,
      "grad_norm": 0.14976902306079865,
      "learning_rate": 4.2444873800104276e-05,
      "loss": 0.5336,
      "mean_token_accuracy": 0.8320377059280872,
      "num_tokens": 599033663.0,
      "step": 18768
    },
    {
      "entropy": 0.679733463563025,
      "epoch": 1.7268503343856563,
      "grad_norm": 0.16414335370063782,
      "learning_rate": 4.244180697396265e-05,
      "loss": 0.6689,
      "mean_token_accuracy": 0.7925274595618248,
      "num_tokens": 599065463.0,
      "step": 18769
    },
    {
      "entropy": 0.5479773096740246,
      "epoch": 1.7269423407570974,
      "grad_norm": 0.14677606523036957,
      "learning_rate": 4.243874014782102e-05,
      "loss": 0.5455,
      "mean_token_accuracy": 0.8349798433482647,
      "num_tokens": 599098003.0,
      "step": 18770
    },
    {
      "entropy": 0.693467739969492,
      "epoch": 1.7270343471285385,
      "grad_norm": 0.15963691473007202,
      "learning_rate": 4.24356733216794e-05,
      "loss": 0.6851,
      "mean_token_accuracy": 0.7899991013109684,
      "num_tokens": 599130037.0,
      "step": 18771
    },
    {
      "entropy": 0.6616175044327974,
      "epoch": 1.72712635349998,
      "grad_norm": 0.16159631311893463,
      "learning_rate": 4.243260649553777e-05,
      "loss": 0.6512,
      "mean_token_accuracy": 0.8013336099684238,
      "num_tokens": 599162258.0,
      "step": 18772
    },
    {
      "entropy": 0.615893267095089,
      "epoch": 1.7272183598714212,
      "grad_norm": 0.15686753392219543,
      "learning_rate": 4.242953966939614e-05,
      "loss": 0.6038,
      "mean_token_accuracy": 0.8189969919621944,
      "num_tokens": 599194273.0,
      "step": 18773
    },
    {
      "entropy": 0.6651415126398206,
      "epoch": 1.7273103662428624,
      "grad_norm": 0.16077570617198944,
      "learning_rate": 4.242647284325452e-05,
      "loss": 0.6501,
      "mean_token_accuracy": 0.7995653636753559,
      "num_tokens": 599225670.0,
      "step": 18774
    },
    {
      "entropy": 0.6271183360368013,
      "epoch": 1.7274023726143035,
      "grad_norm": 0.1598881036043167,
      "learning_rate": 4.242340601711289e-05,
      "loss": 0.6051,
      "mean_token_accuracy": 0.8154629208147526,
      "num_tokens": 599257307.0,
      "step": 18775
    },
    {
      "entropy": 0.6080810585990548,
      "epoch": 1.7274943789857446,
      "grad_norm": 0.15997864305973053,
      "learning_rate": 4.242033919097127e-05,
      "loss": 0.5981,
      "mean_token_accuracy": 0.8169115148484707,
      "num_tokens": 599289521.0,
      "step": 18776
    },
    {
      "entropy": 0.5952079957351089,
      "epoch": 1.727586385357186,
      "grad_norm": 0.15387499332427979,
      "learning_rate": 4.241727236482964e-05,
      "loss": 0.5988,
      "mean_token_accuracy": 0.818506695330143,
      "num_tokens": 599321993.0,
      "step": 18777
    },
    {
      "entropy": 0.623122988268733,
      "epoch": 1.7276783917286274,
      "grad_norm": 0.156886026263237,
      "learning_rate": 4.241420553868801e-05,
      "loss": 0.6141,
      "mean_token_accuracy": 0.8131698742508888,
      "num_tokens": 599353736.0,
      "step": 18778
    },
    {
      "entropy": 0.6794359320774674,
      "epoch": 1.7277703981000685,
      "grad_norm": 0.15802282094955444,
      "learning_rate": 4.241113871254639e-05,
      "loss": 0.6639,
      "mean_token_accuracy": 0.7988394759595394,
      "num_tokens": 599385649.0,
      "step": 18779
    },
    {
      "entropy": 0.6694005150347948,
      "epoch": 1.7278624044715096,
      "grad_norm": 0.15857116878032684,
      "learning_rate": 4.240807188640476e-05,
      "loss": 0.6578,
      "mean_token_accuracy": 0.8001030944287777,
      "num_tokens": 599417712.0,
      "step": 18780
    },
    {
      "entropy": 0.661597391590476,
      "epoch": 1.7279544108429508,
      "grad_norm": 0.1574275940656662,
      "learning_rate": 4.240500506026313e-05,
      "loss": 0.6411,
      "mean_token_accuracy": 0.8063443228602409,
      "num_tokens": 599450126.0,
      "step": 18781
    },
    {
      "entropy": 0.5822187438607216,
      "epoch": 1.7280464172143921,
      "grad_norm": 0.1536860316991806,
      "learning_rate": 4.240193823412151e-05,
      "loss": 0.5609,
      "mean_token_accuracy": 0.8262606933712959,
      "num_tokens": 599482215.0,
      "step": 18782
    },
    {
      "entropy": 0.5817486559972167,
      "epoch": 1.7281384235858335,
      "grad_norm": 0.1586386114358902,
      "learning_rate": 4.239887140797988e-05,
      "loss": 0.574,
      "mean_token_accuracy": 0.8223308026790619,
      "num_tokens": 599513764.0,
      "step": 18783
    },
    {
      "entropy": 0.6221984601579607,
      "epoch": 1.7282304299572746,
      "grad_norm": 0.1550411880016327,
      "learning_rate": 4.239580458183826e-05,
      "loss": 0.6109,
      "mean_token_accuracy": 0.811048973351717,
      "num_tokens": 599545895.0,
      "step": 18784
    },
    {
      "entropy": 0.6015756474807858,
      "epoch": 1.7283224363287157,
      "grad_norm": 0.15273797512054443,
      "learning_rate": 4.239273775569663e-05,
      "loss": 0.5721,
      "mean_token_accuracy": 0.8226415403187275,
      "num_tokens": 599577714.0,
      "step": 18785
    },
    {
      "entropy": 0.7537926509976387,
      "epoch": 1.7284144427001569,
      "grad_norm": 0.17365260422229767,
      "learning_rate": 4.2389670929555006e-05,
      "loss": 0.7476,
      "mean_token_accuracy": 0.7722075991332531,
      "num_tokens": 599608379.0,
      "step": 18786
    },
    {
      "entropy": 0.6269450727850199,
      "epoch": 1.7285064490715982,
      "grad_norm": 0.16070181131362915,
      "learning_rate": 4.238660410341338e-05,
      "loss": 0.6181,
      "mean_token_accuracy": 0.8083692416548729,
      "num_tokens": 599640333.0,
      "step": 18787
    },
    {
      "entropy": 0.5047663524746895,
      "epoch": 1.7285984554430396,
      "grad_norm": 0.14568640291690826,
      "learning_rate": 4.238353727727175e-05,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8529488481581211,
      "num_tokens": 599672424.0,
      "step": 18788
    },
    {
      "entropy": 0.6349676502868533,
      "epoch": 1.7286904618144807,
      "grad_norm": 0.1591658890247345,
      "learning_rate": 4.238047045113013e-05,
      "loss": 0.6181,
      "mean_token_accuracy": 0.8076958768069744,
      "num_tokens": 599704047.0,
      "step": 18789
    },
    {
      "entropy": 0.5822779834270477,
      "epoch": 1.7287824681859219,
      "grad_norm": 0.15315236151218414,
      "learning_rate": 4.23774036249885e-05,
      "loss": 0.5726,
      "mean_token_accuracy": 0.8232446350157261,
      "num_tokens": 599736318.0,
      "step": 18790
    },
    {
      "entropy": 0.582124033709988,
      "epoch": 1.728874474557363,
      "grad_norm": 0.1508396565914154,
      "learning_rate": 4.237433679884687e-05,
      "loss": 0.5698,
      "mean_token_accuracy": 0.8197614476084709,
      "num_tokens": 599768026.0,
      "step": 18791
    },
    {
      "entropy": 0.6367468531243503,
      "epoch": 1.7289664809288043,
      "grad_norm": 0.1558639258146286,
      "learning_rate": 4.2371269972705254e-05,
      "loss": 0.6416,
      "mean_token_accuracy": 0.8034106567502022,
      "num_tokens": 599800042.0,
      "step": 18792
    },
    {
      "entropy": 0.6351031251251698,
      "epoch": 1.7290584873002457,
      "grad_norm": 0.15734784305095673,
      "learning_rate": 4.236820314656362e-05,
      "loss": 0.6261,
      "mean_token_accuracy": 0.8129230849444866,
      "num_tokens": 599832213.0,
      "step": 18793
    },
    {
      "entropy": 0.5960308983922005,
      "epoch": 1.7291504936716868,
      "grad_norm": 0.15157945454120636,
      "learning_rate": 4.2365136320422e-05,
      "loss": 0.5939,
      "mean_token_accuracy": 0.818046797066927,
      "num_tokens": 599864366.0,
      "step": 18794
    },
    {
      "entropy": 0.5812858939170837,
      "epoch": 1.729242500043128,
      "grad_norm": 0.15876588225364685,
      "learning_rate": 4.236206949428037e-05,
      "loss": 0.5686,
      "mean_token_accuracy": 0.8247823491692543,
      "num_tokens": 599895873.0,
      "step": 18795
    },
    {
      "entropy": 0.5533134769648314,
      "epoch": 1.729334506414569,
      "grad_norm": 0.15641437470912933,
      "learning_rate": 4.2359002668138746e-05,
      "loss": 0.5446,
      "mean_token_accuracy": 0.8307269476354122,
      "num_tokens": 599927431.0,
      "step": 18796
    },
    {
      "entropy": 0.6334799565374851,
      "epoch": 1.7294265127860105,
      "grad_norm": 0.16206558048725128,
      "learning_rate": 4.235593584199712e-05,
      "loss": 0.6178,
      "mean_token_accuracy": 0.8121458478271961,
      "num_tokens": 599958718.0,
      "step": 18797
    },
    {
      "entropy": 0.6300189513713121,
      "epoch": 1.7295185191574518,
      "grad_norm": 0.1646416336297989,
      "learning_rate": 4.2352869015855495e-05,
      "loss": 0.62,
      "mean_token_accuracy": 0.8109044693410397,
      "num_tokens": 599990531.0,
      "step": 18798
    },
    {
      "entropy": 0.46968519128859043,
      "epoch": 1.729610525528893,
      "grad_norm": 0.14543063938617706,
      "learning_rate": 4.234980218971386e-05,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8573682047426701,
      "num_tokens": 600022428.0,
      "step": 18799
    },
    {
      "entropy": 0.564839200116694,
      "epoch": 1.729702531900334,
      "grad_norm": 0.14778441190719604,
      "learning_rate": 4.2346735363572245e-05,
      "loss": 0.5485,
      "mean_token_accuracy": 0.832365520298481,
      "num_tokens": 600054417.0,
      "step": 18800
    },
    {
      "entropy": 0.515252117998898,
      "epoch": 1.7297945382717752,
      "grad_norm": 0.14351879060268402,
      "learning_rate": 4.234366853743061e-05,
      "loss": 0.4971,
      "mean_token_accuracy": 0.8489923141896725,
      "num_tokens": 600086826.0,
      "step": 18801
    },
    {
      "entropy": 0.5061147680971771,
      "epoch": 1.7298865446432166,
      "grad_norm": 0.1420290768146515,
      "learning_rate": 4.2340601711288994e-05,
      "loss": 0.4893,
      "mean_token_accuracy": 0.8514286018908024,
      "num_tokens": 600118800.0,
      "step": 18802
    },
    {
      "entropy": 0.6806196887046099,
      "epoch": 1.729978551014658,
      "grad_norm": 0.1607835441827774,
      "learning_rate": 4.233753488514736e-05,
      "loss": 0.6733,
      "mean_token_accuracy": 0.7955358028411865,
      "num_tokens": 600150113.0,
      "step": 18803
    },
    {
      "entropy": 0.6866794563829899,
      "epoch": 1.730070557386099,
      "grad_norm": 0.16148711740970612,
      "learning_rate": 4.2334468059005737e-05,
      "loss": 0.6766,
      "mean_token_accuracy": 0.7931140065193176,
      "num_tokens": 600181287.0,
      "step": 18804
    },
    {
      "entropy": 0.596106730401516,
      "epoch": 1.7301625637575402,
      "grad_norm": 0.15146811306476593,
      "learning_rate": 4.233140123286411e-05,
      "loss": 0.5888,
      "mean_token_accuracy": 0.8166471272706985,
      "num_tokens": 600212719.0,
      "step": 18805
    },
    {
      "entropy": 0.5693533215671778,
      "epoch": 1.7302545701289813,
      "grad_norm": 0.14708170294761658,
      "learning_rate": 4.2328334406722486e-05,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8235963061451912,
      "num_tokens": 600245018.0,
      "step": 18806
    },
    {
      "entropy": 0.47382653784006834,
      "epoch": 1.7303465765004227,
      "grad_norm": 0.13705091178417206,
      "learning_rate": 4.232526758058086e-05,
      "loss": 0.4619,
      "mean_token_accuracy": 0.8547112308442593,
      "num_tokens": 600277273.0,
      "step": 18807
    },
    {
      "entropy": 0.6951630841940641,
      "epoch": 1.730438582871864,
      "grad_norm": 0.1647549569606781,
      "learning_rate": 4.2322200754439235e-05,
      "loss": 0.702,
      "mean_token_accuracy": 0.7885302193462849,
      "num_tokens": 600309180.0,
      "step": 18808
    },
    {
      "entropy": 0.6380895767360926,
      "epoch": 1.7305305892433052,
      "grad_norm": 0.15875877439975739,
      "learning_rate": 4.23191339282976e-05,
      "loss": 0.6442,
      "mean_token_accuracy": 0.8053009733557701,
      "num_tokens": 600340807.0,
      "step": 18809
    },
    {
      "entropy": 0.5360943221021444,
      "epoch": 1.7306225956147463,
      "grad_norm": 0.14113643765449524,
      "learning_rate": 4.2316067102155984e-05,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8359577693045139,
      "num_tokens": 600373231.0,
      "step": 18810
    },
    {
      "entropy": 0.6769628711044788,
      "epoch": 1.7307146019861874,
      "grad_norm": 0.16313698887825012,
      "learning_rate": 4.231300027601435e-05,
      "loss": 0.6818,
      "mean_token_accuracy": 0.7902455404400826,
      "num_tokens": 600404906.0,
      "step": 18811
    },
    {
      "entropy": 0.6063888259232044,
      "epoch": 1.7308066083576288,
      "grad_norm": 0.16020488739013672,
      "learning_rate": 4.230993344987273e-05,
      "loss": 0.6019,
      "mean_token_accuracy": 0.8160061724483967,
      "num_tokens": 600437240.0,
      "step": 18812
    },
    {
      "entropy": 0.5742949629202485,
      "epoch": 1.7308986147290701,
      "grad_norm": 0.15683917701244354,
      "learning_rate": 4.23068666237311e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.821500588208437,
      "num_tokens": 600468947.0,
      "step": 18813
    },
    {
      "entropy": 0.6194980833679438,
      "epoch": 1.7309906211005113,
      "grad_norm": 0.15883822739124298,
      "learning_rate": 4.2303799797589476e-05,
      "loss": 0.6062,
      "mean_token_accuracy": 0.8121887221932411,
      "num_tokens": 600499943.0,
      "step": 18814
    },
    {
      "entropy": 0.5960036450996995,
      "epoch": 1.7310826274719524,
      "grad_norm": 0.16421131789684296,
      "learning_rate": 4.230073297144785e-05,
      "loss": 0.5842,
      "mean_token_accuracy": 0.8200581669807434,
      "num_tokens": 600531164.0,
      "step": 18815
    },
    {
      "entropy": 0.513653053669259,
      "epoch": 1.7311746338433935,
      "grad_norm": 0.1518278270959854,
      "learning_rate": 4.2297666145306226e-05,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8438251167535782,
      "num_tokens": 600562837.0,
      "step": 18816
    },
    {
      "entropy": 0.7738536037504673,
      "epoch": 1.731266640214835,
      "grad_norm": 0.16748926043510437,
      "learning_rate": 4.2294599319164593e-05,
      "loss": 0.7682,
      "mean_token_accuracy": 0.7657725140452385,
      "num_tokens": 600594683.0,
      "step": 18817
    },
    {
      "entropy": 0.5224501956254244,
      "epoch": 1.7313586465862763,
      "grad_norm": 0.1545124500989914,
      "learning_rate": 4.2291532493022975e-05,
      "loss": 0.5025,
      "mean_token_accuracy": 0.844904538244009,
      "num_tokens": 600625915.0,
      "step": 18818
    },
    {
      "entropy": 0.6562407519668341,
      "epoch": 1.7314506529577174,
      "grad_norm": 0.1539071798324585,
      "learning_rate": 4.228846566688134e-05,
      "loss": 0.6363,
      "mean_token_accuracy": 0.8058958724141121,
      "num_tokens": 600657265.0,
      "step": 18819
    },
    {
      "entropy": 0.6433726865798235,
      "epoch": 1.7315426593291585,
      "grad_norm": 0.152007594704628,
      "learning_rate": 4.228539884073972e-05,
      "loss": 0.6384,
      "mean_token_accuracy": 0.8056704364717007,
      "num_tokens": 600689928.0,
      "step": 18820
    },
    {
      "entropy": 0.7189935250207782,
      "epoch": 1.7316346657005997,
      "grad_norm": 0.1593324840068817,
      "learning_rate": 4.22823320145981e-05,
      "loss": 0.7049,
      "mean_token_accuracy": 0.7856765165925026,
      "num_tokens": 600722275.0,
      "step": 18821
    },
    {
      "entropy": 0.6335627362132072,
      "epoch": 1.731726672072041,
      "grad_norm": 0.16715285181999207,
      "learning_rate": 4.227926518845647e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.811591625213623,
      "num_tokens": 600753228.0,
      "step": 18822
    },
    {
      "entropy": 0.5457864617928863,
      "epoch": 1.7318186784434824,
      "grad_norm": 0.1512870341539383,
      "learning_rate": 4.227619836231485e-05,
      "loss": 0.5344,
      "mean_token_accuracy": 0.8353350684046745,
      "num_tokens": 600784274.0,
      "step": 18823
    },
    {
      "entropy": 0.6725518237799406,
      "epoch": 1.7319106848149235,
      "grad_norm": 0.15734772384166718,
      "learning_rate": 4.2273131536173216e-05,
      "loss": 0.6509,
      "mean_token_accuracy": 0.7989542298018932,
      "num_tokens": 600815499.0,
      "step": 18824
    },
    {
      "entropy": 0.5770807689987123,
      "epoch": 1.7320026911863646,
      "grad_norm": 0.15240955352783203,
      "learning_rate": 4.227006471003159e-05,
      "loss": 0.5588,
      "mean_token_accuracy": 0.829025711864233,
      "num_tokens": 600848000.0,
      "step": 18825
    },
    {
      "entropy": 0.690018666908145,
      "epoch": 1.7320946975578058,
      "grad_norm": 0.15894228219985962,
      "learning_rate": 4.2266997883889965e-05,
      "loss": 0.678,
      "mean_token_accuracy": 0.7938508912920952,
      "num_tokens": 600880151.0,
      "step": 18826
    },
    {
      "entropy": 0.7294075936079025,
      "epoch": 1.7321867039292471,
      "grad_norm": 0.1637323945760727,
      "learning_rate": 4.226393105774834e-05,
      "loss": 0.7144,
      "mean_token_accuracy": 0.7831262908875942,
      "num_tokens": 600912209.0,
      "step": 18827
    },
    {
      "entropy": 0.591367750428617,
      "epoch": 1.7322787103006885,
      "grad_norm": 0.1463409662246704,
      "learning_rate": 4.2260864231606715e-05,
      "loss": 0.5771,
      "mean_token_accuracy": 0.8218822069466114,
      "num_tokens": 600944597.0,
      "step": 18828
    },
    {
      "entropy": 0.7060023061931133,
      "epoch": 1.7323707166721296,
      "grad_norm": 0.16159196197986603,
      "learning_rate": 4.225779740546509e-05,
      "loss": 0.7188,
      "mean_token_accuracy": 0.7856814973056316,
      "num_tokens": 600976391.0,
      "step": 18829
    },
    {
      "entropy": 0.5186847154982388,
      "epoch": 1.7324627230435707,
      "grad_norm": 0.152225062251091,
      "learning_rate": 4.225473057932346e-05,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8386551253497601,
      "num_tokens": 601008569.0,
      "step": 18830
    },
    {
      "entropy": 0.6575809791684151,
      "epoch": 1.7325547294150119,
      "grad_norm": 0.15377332270145416,
      "learning_rate": 4.225166375318184e-05,
      "loss": 0.6446,
      "mean_token_accuracy": 0.8002496212720871,
      "num_tokens": 601040339.0,
      "step": 18831
    },
    {
      "entropy": 0.5754322931170464,
      "epoch": 1.7326467357864532,
      "grad_norm": 0.15795183181762695,
      "learning_rate": 4.2248596927040206e-05,
      "loss": 0.5664,
      "mean_token_accuracy": 0.8241823613643646,
      "num_tokens": 601072781.0,
      "step": 18832
    },
    {
      "entropy": 0.5287236217409372,
      "epoch": 1.7327387421578946,
      "grad_norm": 0.14449913799762726,
      "learning_rate": 4.224553010089858e-05,
      "loss": 0.5144,
      "mean_token_accuracy": 0.8428319282829762,
      "num_tokens": 601105182.0,
      "step": 18833
    },
    {
      "entropy": 0.6252371165901423,
      "epoch": 1.7328307485293357,
      "grad_norm": 0.15823695063591003,
      "learning_rate": 4.2242463274756956e-05,
      "loss": 0.6245,
      "mean_token_accuracy": 0.8111359626054764,
      "num_tokens": 601136800.0,
      "step": 18834
    },
    {
      "entropy": 0.5448673078790307,
      "epoch": 1.7329227549007769,
      "grad_norm": 0.1522957682609558,
      "learning_rate": 4.223939644861533e-05,
      "loss": 0.5408,
      "mean_token_accuracy": 0.8355144411325455,
      "num_tokens": 601169344.0,
      "step": 18835
    },
    {
      "entropy": 0.6158992163836956,
      "epoch": 1.733014761272218,
      "grad_norm": 0.1479862630367279,
      "learning_rate": 4.2236329622473705e-05,
      "loss": 0.601,
      "mean_token_accuracy": 0.8167539983987808,
      "num_tokens": 601201311.0,
      "step": 18836
    },
    {
      "entropy": 0.5963720791041851,
      "epoch": 1.7331067676436593,
      "grad_norm": 0.14986364543437958,
      "learning_rate": 4.223326279633208e-05,
      "loss": 0.5843,
      "mean_token_accuracy": 0.8171462714672089,
      "num_tokens": 601233171.0,
      "step": 18837
    },
    {
      "entropy": 0.6768324133008718,
      "epoch": 1.7331987740151007,
      "grad_norm": 0.1640443652868271,
      "learning_rate": 4.223019597019045e-05,
      "loss": 0.6796,
      "mean_token_accuracy": 0.7978736534714699,
      "num_tokens": 601264848.0,
      "step": 18838
    },
    {
      "entropy": 0.6192285679280758,
      "epoch": 1.7332907803865418,
      "grad_norm": 0.15402181446552277,
      "learning_rate": 4.222712914404883e-05,
      "loss": 0.6124,
      "mean_token_accuracy": 0.8155877962708473,
      "num_tokens": 601297312.0,
      "step": 18839
    },
    {
      "entropy": 0.5477281750645489,
      "epoch": 1.733382786757983,
      "grad_norm": 0.15337061882019043,
      "learning_rate": 4.22240623179072e-05,
      "loss": 0.5279,
      "mean_token_accuracy": 0.8343467004597187,
      "num_tokens": 601329427.0,
      "step": 18840
    },
    {
      "entropy": 0.600143427029252,
      "epoch": 1.733474793129424,
      "grad_norm": 0.14826035499572754,
      "learning_rate": 4.222099549176558e-05,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8217121623456478,
      "num_tokens": 601361299.0,
      "step": 18841
    },
    {
      "entropy": 0.6249179681763053,
      "epoch": 1.7335667995008655,
      "grad_norm": 0.15571412444114685,
      "learning_rate": 4.2217928665623946e-05,
      "loss": 0.6143,
      "mean_token_accuracy": 0.8130206689238548,
      "num_tokens": 601393223.0,
      "step": 18842
    },
    {
      "entropy": 0.5260046161711216,
      "epoch": 1.7336588058723068,
      "grad_norm": 0.15144580602645874,
      "learning_rate": 4.221486183948232e-05,
      "loss": 0.5145,
      "mean_token_accuracy": 0.8404380455613136,
      "num_tokens": 601425535.0,
      "step": 18843
    },
    {
      "entropy": 0.5398557111620903,
      "epoch": 1.733750812243748,
      "grad_norm": 0.15351174771785736,
      "learning_rate": 4.2211795013340695e-05,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8372367955744267,
      "num_tokens": 601457313.0,
      "step": 18844
    },
    {
      "entropy": 0.5842023603618145,
      "epoch": 1.733842818615189,
      "grad_norm": 0.15692946314811707,
      "learning_rate": 4.220872818719907e-05,
      "loss": 0.5754,
      "mean_token_accuracy": 0.8262364007532597,
      "num_tokens": 601489557.0,
      "step": 18845
    },
    {
      "entropy": 0.6216004956513643,
      "epoch": 1.7339348249866302,
      "grad_norm": 0.16226781904697418,
      "learning_rate": 4.2205661361057445e-05,
      "loss": 0.6149,
      "mean_token_accuracy": 0.810600757598877,
      "num_tokens": 601521349.0,
      "step": 18846
    },
    {
      "entropy": 0.6085889972746372,
      "epoch": 1.7340268313580716,
      "grad_norm": 0.16301056742668152,
      "learning_rate": 4.220259453491582e-05,
      "loss": 0.5969,
      "mean_token_accuracy": 0.8190958648920059,
      "num_tokens": 601552687.0,
      "step": 18847
    },
    {
      "entropy": 0.6058426832314581,
      "epoch": 1.734118837729513,
      "grad_norm": 0.15172304213047028,
      "learning_rate": 4.219952770877419e-05,
      "loss": 0.6054,
      "mean_token_accuracy": 0.8181876577436924,
      "num_tokens": 601585035.0,
      "step": 18848
    },
    {
      "entropy": 0.5715881902724504,
      "epoch": 1.734210844100954,
      "grad_norm": 0.1610708087682724,
      "learning_rate": 4.219646088263257e-05,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8271909281611443,
      "num_tokens": 601616306.0,
      "step": 18849
    },
    {
      "entropy": 0.7041213689371943,
      "epoch": 1.7343028504723952,
      "grad_norm": 0.15907356142997742,
      "learning_rate": 4.2193394056490936e-05,
      "loss": 0.6906,
      "mean_token_accuracy": 0.7874523140490055,
      "num_tokens": 601648710.0,
      "step": 18850
    },
    {
      "entropy": 0.6288966201245785,
      "epoch": 1.7343948568438363,
      "grad_norm": 0.1610017567873001,
      "learning_rate": 4.219032723034931e-05,
      "loss": 0.6147,
      "mean_token_accuracy": 0.8127221763134003,
      "num_tokens": 601679943.0,
      "step": 18851
    },
    {
      "entropy": 0.6250236150808632,
      "epoch": 1.7344868632152777,
      "grad_norm": 0.15652073919773102,
      "learning_rate": 4.218726040420769e-05,
      "loss": 0.6188,
      "mean_token_accuracy": 0.8115817271173,
      "num_tokens": 601711959.0,
      "step": 18852
    },
    {
      "entropy": 0.6316497900988907,
      "epoch": 1.734578869586719,
      "grad_norm": 0.16139435768127441,
      "learning_rate": 4.218419357806606e-05,
      "loss": 0.621,
      "mean_token_accuracy": 0.8117442391812801,
      "num_tokens": 601743630.0,
      "step": 18853
    },
    {
      "entropy": 0.5757109932601452,
      "epoch": 1.7346708759581602,
      "grad_norm": 0.1494656354188919,
      "learning_rate": 4.2181126751924435e-05,
      "loss": 0.5646,
      "mean_token_accuracy": 0.8244369216263294,
      "num_tokens": 601775386.0,
      "step": 18854
    },
    {
      "entropy": 0.5221302926074713,
      "epoch": 1.7347628823296013,
      "grad_norm": 0.15078341960906982,
      "learning_rate": 4.217805992578281e-05,
      "loss": 0.5082,
      "mean_token_accuracy": 0.8395588807761669,
      "num_tokens": 601806486.0,
      "step": 18855
    },
    {
      "entropy": 0.6509780958294868,
      "epoch": 1.7348548887010424,
      "grad_norm": 0.15953227877616882,
      "learning_rate": 4.2174993099641184e-05,
      "loss": 0.644,
      "mean_token_accuracy": 0.8000358827412128,
      "num_tokens": 601838202.0,
      "step": 18856
    },
    {
      "entropy": 0.5751794492825866,
      "epoch": 1.7349468950724838,
      "grad_norm": 0.15216852724552155,
      "learning_rate": 4.217192627349956e-05,
      "loss": 0.557,
      "mean_token_accuracy": 0.8292422331869602,
      "num_tokens": 601869596.0,
      "step": 18857
    },
    {
      "entropy": 0.571254150941968,
      "epoch": 1.7350389014439251,
      "grad_norm": 0.1560012400150299,
      "learning_rate": 4.2168859447357934e-05,
      "loss": 0.5504,
      "mean_token_accuracy": 0.828889399766922,
      "num_tokens": 601901051.0,
      "step": 18858
    },
    {
      "entropy": 0.6430804673582315,
      "epoch": 1.7351309078153663,
      "grad_norm": 0.15443967282772064,
      "learning_rate": 4.21657926212163e-05,
      "loss": 0.633,
      "mean_token_accuracy": 0.80467863753438,
      "num_tokens": 601933657.0,
      "step": 18859
    },
    {
      "entropy": 0.643119104206562,
      "epoch": 1.7352229141868074,
      "grad_norm": 0.1569230556488037,
      "learning_rate": 4.216272579507468e-05,
      "loss": 0.6227,
      "mean_token_accuracy": 0.810212142765522,
      "num_tokens": 601965241.0,
      "step": 18860
    },
    {
      "entropy": 0.5816749054938555,
      "epoch": 1.7353149205582485,
      "grad_norm": 0.15156017243862152,
      "learning_rate": 4.215965896893305e-05,
      "loss": 0.5835,
      "mean_token_accuracy": 0.8203502483665943,
      "num_tokens": 601995990.0,
      "step": 18861
    },
    {
      "entropy": 0.5258151907473803,
      "epoch": 1.73540692692969,
      "grad_norm": 0.14581863582134247,
      "learning_rate": 4.215659214279143e-05,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8386580608785152,
      "num_tokens": 602027666.0,
      "step": 18862
    },
    {
      "entropy": 0.5250915782526135,
      "epoch": 1.7354989333011313,
      "grad_norm": 0.15249617397785187,
      "learning_rate": 4.21535253166498e-05,
      "loss": 0.5172,
      "mean_token_accuracy": 0.8470356576144695,
      "num_tokens": 602059752.0,
      "step": 18863
    },
    {
      "entropy": 0.5581158492714167,
      "epoch": 1.7355909396725724,
      "grad_norm": 0.1469288170337677,
      "learning_rate": 4.2150458490508175e-05,
      "loss": 0.5425,
      "mean_token_accuracy": 0.8314113020896912,
      "num_tokens": 602090976.0,
      "step": 18864
    },
    {
      "entropy": 0.5244380170479417,
      "epoch": 1.7356829460440135,
      "grad_norm": 0.1467086523771286,
      "learning_rate": 4.214739166436655e-05,
      "loss": 0.513,
      "mean_token_accuracy": 0.8363667167723179,
      "num_tokens": 602122681.0,
      "step": 18865
    },
    {
      "entropy": 0.599687248468399,
      "epoch": 1.7357749524154547,
      "grad_norm": 0.15224295854568481,
      "learning_rate": 4.2144324838224924e-05,
      "loss": 0.5868,
      "mean_token_accuracy": 0.8189199641346931,
      "num_tokens": 602154615.0,
      "step": 18866
    },
    {
      "entropy": 0.647417213767767,
      "epoch": 1.735866958786896,
      "grad_norm": 0.15864112973213196,
      "learning_rate": 4.21412580120833e-05,
      "loss": 0.6423,
      "mean_token_accuracy": 0.805709533393383,
      "num_tokens": 602187265.0,
      "step": 18867
    },
    {
      "entropy": 0.5534328040666878,
      "epoch": 1.7359589651583374,
      "grad_norm": 0.15043975412845612,
      "learning_rate": 4.213819118594167e-05,
      "loss": 0.5481,
      "mean_token_accuracy": 0.832963764667511,
      "num_tokens": 602219674.0,
      "step": 18868
    },
    {
      "entropy": 0.6590967606753111,
      "epoch": 1.7360509715297785,
      "grad_norm": 0.15713021159172058,
      "learning_rate": 4.213512435980004e-05,
      "loss": 0.6547,
      "mean_token_accuracy": 0.8013758286833763,
      "num_tokens": 602251597.0,
      "step": 18869
    },
    {
      "entropy": 0.6536568030714989,
      "epoch": 1.7361429779012196,
      "grad_norm": 0.1608688086271286,
      "learning_rate": 4.213205753365842e-05,
      "loss": 0.6485,
      "mean_token_accuracy": 0.7998201623558998,
      "num_tokens": 602283610.0,
      "step": 18870
    },
    {
      "entropy": 0.6331498119980097,
      "epoch": 1.7362349842726608,
      "grad_norm": 0.16176320612430573,
      "learning_rate": 4.212899070751679e-05,
      "loss": 0.6157,
      "mean_token_accuracy": 0.8120024725794792,
      "num_tokens": 602315272.0,
      "step": 18871
    },
    {
      "entropy": 0.6087036468088627,
      "epoch": 1.7363269906441021,
      "grad_norm": 0.15887761116027832,
      "learning_rate": 4.2125923881375165e-05,
      "loss": 0.5971,
      "mean_token_accuracy": 0.8147946372628212,
      "num_tokens": 602346719.0,
      "step": 18872
    },
    {
      "entropy": 0.6253848467022181,
      "epoch": 1.7364189970155435,
      "grad_norm": 0.165482759475708,
      "learning_rate": 4.212285705523354e-05,
      "loss": 0.642,
      "mean_token_accuracy": 0.8048324026167393,
      "num_tokens": 602378537.0,
      "step": 18873
    },
    {
      "entropy": 0.5817783484235406,
      "epoch": 1.7365110033869846,
      "grad_norm": 0.14789873361587524,
      "learning_rate": 4.2119790229091914e-05,
      "loss": 0.578,
      "mean_token_accuracy": 0.8202953860163689,
      "num_tokens": 602410907.0,
      "step": 18874
    },
    {
      "entropy": 0.6321502476930618,
      "epoch": 1.7366030097584257,
      "grad_norm": 0.1536732167005539,
      "learning_rate": 4.211672340295029e-05,
      "loss": 0.6132,
      "mean_token_accuracy": 0.8120230734348297,
      "num_tokens": 602443445.0,
      "step": 18875
    },
    {
      "entropy": 0.6257749013602734,
      "epoch": 1.7366950161298669,
      "grad_norm": 0.16017131507396698,
      "learning_rate": 4.2113656576808664e-05,
      "loss": 0.6034,
      "mean_token_accuracy": 0.8109328337013721,
      "num_tokens": 602474900.0,
      "step": 18876
    },
    {
      "entropy": 0.5569409187883139,
      "epoch": 1.7367870225013082,
      "grad_norm": 0.14630621671676636,
      "learning_rate": 4.211058975066703e-05,
      "loss": 0.552,
      "mean_token_accuracy": 0.8277061134576797,
      "num_tokens": 602506899.0,
      "step": 18877
    },
    {
      "entropy": 0.5006356341764331,
      "epoch": 1.7368790288727496,
      "grad_norm": 0.147138774394989,
      "learning_rate": 4.210752292452541e-05,
      "loss": 0.4982,
      "mean_token_accuracy": 0.8446715548634529,
      "num_tokens": 602538813.0,
      "step": 18878
    },
    {
      "entropy": 0.6310586091130972,
      "epoch": 1.7369710352441907,
      "grad_norm": 0.15487313270568848,
      "learning_rate": 4.210445609838378e-05,
      "loss": 0.6252,
      "mean_token_accuracy": 0.8094345517456532,
      "num_tokens": 602571072.0,
      "step": 18879
    },
    {
      "entropy": 0.5893867956474423,
      "epoch": 1.7370630416156319,
      "grad_norm": 0.15112337470054626,
      "learning_rate": 4.2101389272242156e-05,
      "loss": 0.578,
      "mean_token_accuracy": 0.8203949518501759,
      "num_tokens": 602602871.0,
      "step": 18880
    },
    {
      "entropy": 0.5963109717704356,
      "epoch": 1.737155047987073,
      "grad_norm": 0.15401121973991394,
      "learning_rate": 4.209832244610053e-05,
      "loss": 0.595,
      "mean_token_accuracy": 0.8198564536869526,
      "num_tokens": 602634387.0,
      "step": 18881
    },
    {
      "entropy": 0.602221891283989,
      "epoch": 1.7372470543585143,
      "grad_norm": 0.15127603709697723,
      "learning_rate": 4.2095255619958905e-05,
      "loss": 0.6034,
      "mean_token_accuracy": 0.8133271709084511,
      "num_tokens": 602666663.0,
      "step": 18882
    },
    {
      "entropy": 0.6946284677833319,
      "epoch": 1.7373390607299557,
      "grad_norm": 0.16235585510730743,
      "learning_rate": 4.2092188793817286e-05,
      "loss": 0.6906,
      "mean_token_accuracy": 0.7912711948156357,
      "num_tokens": 602698531.0,
      "step": 18883
    },
    {
      "entropy": 0.5961194280534983,
      "epoch": 1.7374310671013968,
      "grad_norm": 0.14784476161003113,
      "learning_rate": 4.2089121967675654e-05,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8205495998263359,
      "num_tokens": 602730805.0,
      "step": 18884
    },
    {
      "entropy": 0.5996664594858885,
      "epoch": 1.737523073472838,
      "grad_norm": 0.1537388563156128,
      "learning_rate": 4.208605514153403e-05,
      "loss": 0.5906,
      "mean_token_accuracy": 0.8199531510472298,
      "num_tokens": 602763020.0,
      "step": 18885
    },
    {
      "entropy": 0.5744144814088941,
      "epoch": 1.737615079844279,
      "grad_norm": 0.14854629337787628,
      "learning_rate": 4.2082988315392403e-05,
      "loss": 0.5668,
      "mean_token_accuracy": 0.8273636139929295,
      "num_tokens": 602795528.0,
      "step": 18886
    },
    {
      "entropy": 0.7127011083066463,
      "epoch": 1.7377070862157205,
      "grad_norm": 0.16563060879707336,
      "learning_rate": 4.207992148925078e-05,
      "loss": 0.7041,
      "mean_token_accuracy": 0.7854208126664162,
      "num_tokens": 602825952.0,
      "step": 18887
    },
    {
      "entropy": 0.5925232050940394,
      "epoch": 1.7377990925871618,
      "grad_norm": 0.15045945346355438,
      "learning_rate": 4.207685466310915e-05,
      "loss": 0.5782,
      "mean_token_accuracy": 0.8210625611245632,
      "num_tokens": 602858141.0,
      "step": 18888
    },
    {
      "entropy": 0.6101887738332152,
      "epoch": 1.737891098958603,
      "grad_norm": 0.1640608012676239,
      "learning_rate": 4.207378783696753e-05,
      "loss": 0.5901,
      "mean_token_accuracy": 0.8210658840835094,
      "num_tokens": 602890611.0,
      "step": 18889
    },
    {
      "entropy": 0.5008365162648261,
      "epoch": 1.737983105330044,
      "grad_norm": 0.14313824474811554,
      "learning_rate": 4.2070721010825895e-05,
      "loss": 0.4839,
      "mean_token_accuracy": 0.8492686748504639,
      "num_tokens": 602922723.0,
      "step": 18890
    },
    {
      "entropy": 0.5712152421474457,
      "epoch": 1.7380751117014852,
      "grad_norm": 0.15606938302516937,
      "learning_rate": 4.206765418468428e-05,
      "loss": 0.5503,
      "mean_token_accuracy": 0.8290042467415333,
      "num_tokens": 602954773.0,
      "step": 18891
    },
    {
      "entropy": 0.6576368920505047,
      "epoch": 1.7381671180729266,
      "grad_norm": 0.15816627442836761,
      "learning_rate": 4.2064587358542645e-05,
      "loss": 0.6503,
      "mean_token_accuracy": 0.8011680468916893,
      "num_tokens": 602986425.0,
      "step": 18892
    },
    {
      "entropy": 0.4641693951562047,
      "epoch": 1.738259124444368,
      "grad_norm": 0.141921728849411,
      "learning_rate": 4.206152053240102e-05,
      "loss": 0.4459,
      "mean_token_accuracy": 0.859856054186821,
      "num_tokens": 603017466.0,
      "step": 18893
    },
    {
      "entropy": 0.5612538158893585,
      "epoch": 1.738351130815809,
      "grad_norm": 0.14839288592338562,
      "learning_rate": 4.2058453706259394e-05,
      "loss": 0.5423,
      "mean_token_accuracy": 0.8276170454919338,
      "num_tokens": 603049301.0,
      "step": 18894
    },
    {
      "entropy": 0.6064086891710758,
      "epoch": 1.7384431371872502,
      "grad_norm": 0.15560919046401978,
      "learning_rate": 4.205538688011777e-05,
      "loss": 0.6045,
      "mean_token_accuracy": 0.8128041103482246,
      "num_tokens": 603081137.0,
      "step": 18895
    },
    {
      "entropy": 0.5461269086226821,
      "epoch": 1.7385351435586913,
      "grad_norm": 0.15073098242282867,
      "learning_rate": 4.205232005397614e-05,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8340305984020233,
      "num_tokens": 603113453.0,
      "step": 18896
    },
    {
      "entropy": 0.7081242576241493,
      "epoch": 1.7386271499301327,
      "grad_norm": 0.16369253396987915,
      "learning_rate": 4.204925322783452e-05,
      "loss": 0.7036,
      "mean_token_accuracy": 0.7852742373943329,
      "num_tokens": 603145028.0,
      "step": 18897
    },
    {
      "entropy": 0.609211590141058,
      "epoch": 1.738719156301574,
      "grad_norm": 0.15458989143371582,
      "learning_rate": 4.2046186401692886e-05,
      "loss": 0.593,
      "mean_token_accuracy": 0.8170433081686497,
      "num_tokens": 603176317.0,
      "step": 18898
    },
    {
      "entropy": 0.49538624472916126,
      "epoch": 1.7388111626730152,
      "grad_norm": 0.1530454009771347,
      "learning_rate": 4.204311957555127e-05,
      "loss": 0.5,
      "mean_token_accuracy": 0.846094224601984,
      "num_tokens": 603208047.0,
      "step": 18899
    },
    {
      "entropy": 0.6196106178686023,
      "epoch": 1.7389031690444563,
      "grad_norm": 0.15907855331897736,
      "learning_rate": 4.2040052749409635e-05,
      "loss": 0.6287,
      "mean_token_accuracy": 0.8115137405693531,
      "num_tokens": 603240430.0,
      "step": 18900
    },
    {
      "entropy": 0.6308850254863501,
      "epoch": 1.7389951754158974,
      "grad_norm": 0.1563754677772522,
      "learning_rate": 4.2036985923268016e-05,
      "loss": 0.6195,
      "mean_token_accuracy": 0.8042851686477661,
      "num_tokens": 603272652.0,
      "step": 18901
    },
    {
      "entropy": 0.6087502194568515,
      "epoch": 1.7390871817873388,
      "grad_norm": 0.15338096022605896,
      "learning_rate": 4.2033919097126384e-05,
      "loss": 0.6015,
      "mean_token_accuracy": 0.8163457401096821,
      "num_tokens": 603303621.0,
      "step": 18902
    },
    {
      "entropy": 0.7143909968435764,
      "epoch": 1.7391791881587801,
      "grad_norm": 0.1605740189552307,
      "learning_rate": 4.203085227098476e-05,
      "loss": 0.715,
      "mean_token_accuracy": 0.7846557460725307,
      "num_tokens": 603335437.0,
      "step": 18903
    },
    {
      "entropy": 0.5925226546823978,
      "epoch": 1.7392711945302213,
      "grad_norm": 0.1534019112586975,
      "learning_rate": 4.2027785444843134e-05,
      "loss": 0.5734,
      "mean_token_accuracy": 0.8206611536443233,
      "num_tokens": 603367180.0,
      "step": 18904
    },
    {
      "entropy": 0.5215223131235689,
      "epoch": 1.7393632009016624,
      "grad_norm": 0.14816054701805115,
      "learning_rate": 4.202471861870151e-05,
      "loss": 0.5121,
      "mean_token_accuracy": 0.8411221392452717,
      "num_tokens": 603399684.0,
      "step": 18905
    },
    {
      "entropy": 0.6887851785868406,
      "epoch": 1.7394552072731035,
      "grad_norm": 0.16171516478061676,
      "learning_rate": 4.202165179255988e-05,
      "loss": 0.6826,
      "mean_token_accuracy": 0.7913898266851902,
      "num_tokens": 603431541.0,
      "step": 18906
    },
    {
      "entropy": 0.6426271572709084,
      "epoch": 1.739547213644545,
      "grad_norm": 0.15506397187709808,
      "learning_rate": 4.201858496641826e-05,
      "loss": 0.6261,
      "mean_token_accuracy": 0.8085445687174797,
      "num_tokens": 603463332.0,
      "step": 18907
    },
    {
      "entropy": 0.6447505569085479,
      "epoch": 1.7396392200159863,
      "grad_norm": 0.16273276507854462,
      "learning_rate": 4.2015518140276625e-05,
      "loss": 0.6456,
      "mean_token_accuracy": 0.8026386983692646,
      "num_tokens": 603495876.0,
      "step": 18908
    },
    {
      "entropy": 0.669253334403038,
      "epoch": 1.7397312263874274,
      "grad_norm": 0.15868471562862396,
      "learning_rate": 4.201245131413501e-05,
      "loss": 0.6615,
      "mean_token_accuracy": 0.7984006777405739,
      "num_tokens": 603528177.0,
      "step": 18909
    },
    {
      "entropy": 0.6171634029597044,
      "epoch": 1.7398232327588685,
      "grad_norm": 0.1561916470527649,
      "learning_rate": 4.2009384487993375e-05,
      "loss": 0.5946,
      "mean_token_accuracy": 0.8156193122267723,
      "num_tokens": 603560245.0,
      "step": 18910
    },
    {
      "entropy": 0.6791915483772755,
      "epoch": 1.7399152391303097,
      "grad_norm": 0.1583314687013626,
      "learning_rate": 4.200631766185175e-05,
      "loss": 0.6717,
      "mean_token_accuracy": 0.7983976602554321,
      "num_tokens": 603592691.0,
      "step": 18911
    },
    {
      "entropy": 0.6724912356585264,
      "epoch": 1.740007245501751,
      "grad_norm": 0.15541060268878937,
      "learning_rate": 4.2003250835710124e-05,
      "loss": 0.6539,
      "mean_token_accuracy": 0.802518717944622,
      "num_tokens": 603623858.0,
      "step": 18912
    },
    {
      "entropy": 0.6329523157328367,
      "epoch": 1.7400992518731924,
      "grad_norm": 0.1528375893831253,
      "learning_rate": 4.20001840095685e-05,
      "loss": 0.6329,
      "mean_token_accuracy": 0.8077016547322273,
      "num_tokens": 603655870.0,
      "step": 18913
    },
    {
      "entropy": 0.6062539517879486,
      "epoch": 1.7401912582446335,
      "grad_norm": 0.15229398012161255,
      "learning_rate": 4.199711718342687e-05,
      "loss": 0.587,
      "mean_token_accuracy": 0.8167297840118408,
      "num_tokens": 603688211.0,
      "step": 18914
    },
    {
      "entropy": 0.6621478535234928,
      "epoch": 1.7402832646160746,
      "grad_norm": 0.15567171573638916,
      "learning_rate": 4.199405035728525e-05,
      "loss": 0.6589,
      "mean_token_accuracy": 0.7971150726079941,
      "num_tokens": 603720497.0,
      "step": 18915
    },
    {
      "entropy": 0.5712015656754375,
      "epoch": 1.7403752709875158,
      "grad_norm": 0.1504756361246109,
      "learning_rate": 4.199098353114362e-05,
      "loss": 0.5653,
      "mean_token_accuracy": 0.8273103013634682,
      "num_tokens": 603752831.0,
      "step": 18916
    },
    {
      "entropy": 0.5895668398588896,
      "epoch": 1.7404672773589571,
      "grad_norm": 0.15379905700683594,
      "learning_rate": 4.1987916705002e-05,
      "loss": 0.5699,
      "mean_token_accuracy": 0.8244770616292953,
      "num_tokens": 603785206.0,
      "step": 18917
    },
    {
      "entropy": 0.5740209054201841,
      "epoch": 1.7405592837303985,
      "grad_norm": 0.1564187854528427,
      "learning_rate": 4.198484987886037e-05,
      "loss": 0.5709,
      "mean_token_accuracy": 0.825866412371397,
      "num_tokens": 603817028.0,
      "step": 18918
    },
    {
      "entropy": 0.5276920644100755,
      "epoch": 1.7406512901018396,
      "grad_norm": 0.15401098132133484,
      "learning_rate": 4.198178305271874e-05,
      "loss": 0.5142,
      "mean_token_accuracy": 0.841917872428894,
      "num_tokens": 603848198.0,
      "step": 18919
    },
    {
      "entropy": 0.6282826503738761,
      "epoch": 1.7407432964732807,
      "grad_norm": 0.1544046700000763,
      "learning_rate": 4.197871622657712e-05,
      "loss": 0.6082,
      "mean_token_accuracy": 0.812689658254385,
      "num_tokens": 603880109.0,
      "step": 18920
    },
    {
      "entropy": 0.6120225246995687,
      "epoch": 1.7408353028447219,
      "grad_norm": 0.1548732966184616,
      "learning_rate": 4.197564940043549e-05,
      "loss": 0.6215,
      "mean_token_accuracy": 0.8086276985704899,
      "num_tokens": 603912572.0,
      "step": 18921
    },
    {
      "entropy": 0.5306780557148159,
      "epoch": 1.7409273092161632,
      "grad_norm": 0.1525736004114151,
      "learning_rate": 4.197258257429387e-05,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8347797058522701,
      "num_tokens": 603944603.0,
      "step": 18922
    },
    {
      "entropy": 0.6339456830173731,
      "epoch": 1.7410193155876046,
      "grad_norm": 0.15526054799556732,
      "learning_rate": 4.196951574815224e-05,
      "loss": 0.6421,
      "mean_token_accuracy": 0.8039596788585186,
      "num_tokens": 603976385.0,
      "step": 18923
    },
    {
      "entropy": 0.7197327930480242,
      "epoch": 1.7411113219590457,
      "grad_norm": 0.16099011898040771,
      "learning_rate": 4.196644892201061e-05,
      "loss": 0.7054,
      "mean_token_accuracy": 0.7864749431610107,
      "num_tokens": 604008515.0,
      "step": 18924
    },
    {
      "entropy": 0.7163858329877257,
      "epoch": 1.7412033283304869,
      "grad_norm": 0.16648083925247192,
      "learning_rate": 4.196338209586899e-05,
      "loss": 0.6983,
      "mean_token_accuracy": 0.7877045050263405,
      "num_tokens": 604040031.0,
      "step": 18925
    },
    {
      "entropy": 0.5353901381604373,
      "epoch": 1.741295334701928,
      "grad_norm": 0.15741322934627533,
      "learning_rate": 4.196031526972736e-05,
      "loss": 0.5275,
      "mean_token_accuracy": 0.8410791791975498,
      "num_tokens": 604071907.0,
      "step": 18926
    },
    {
      "entropy": 0.6750953793525696,
      "epoch": 1.7413873410733693,
      "grad_norm": 0.161361962556839,
      "learning_rate": 4.195724844358574e-05,
      "loss": 0.6626,
      "mean_token_accuracy": 0.7992555797100067,
      "num_tokens": 604103036.0,
      "step": 18927
    },
    {
      "entropy": 0.6755589637905359,
      "epoch": 1.7414793474448107,
      "grad_norm": 0.16254237294197083,
      "learning_rate": 4.195418161744411e-05,
      "loss": 0.6535,
      "mean_token_accuracy": 0.8001275770366192,
      "num_tokens": 604135440.0,
      "step": 18928
    },
    {
      "entropy": 0.6965082548558712,
      "epoch": 1.7415713538162518,
      "grad_norm": 0.15700379014015198,
      "learning_rate": 4.195111479130248e-05,
      "loss": 0.6797,
      "mean_token_accuracy": 0.7927725128829479,
      "num_tokens": 604167292.0,
      "step": 18929
    },
    {
      "entropy": 0.5902611250057817,
      "epoch": 1.741663360187693,
      "grad_norm": 0.15976764261722565,
      "learning_rate": 4.194804796516086e-05,
      "loss": 0.5819,
      "mean_token_accuracy": 0.8223399482667446,
      "num_tokens": 604198832.0,
      "step": 18930
    },
    {
      "entropy": 0.5604426423087716,
      "epoch": 1.741755366559134,
      "grad_norm": 0.15224149823188782,
      "learning_rate": 4.194498113901923e-05,
      "loss": 0.5509,
      "mean_token_accuracy": 0.8281377851963043,
      "num_tokens": 604230765.0,
      "step": 18931
    },
    {
      "entropy": 0.591717672534287,
      "epoch": 1.7418473729305755,
      "grad_norm": 0.1504969447851181,
      "learning_rate": 4.19419143128776e-05,
      "loss": 0.5918,
      "mean_token_accuracy": 0.8162132985889912,
      "num_tokens": 604263353.0,
      "step": 18932
    },
    {
      "entropy": 0.5695671066641808,
      "epoch": 1.7419393793020168,
      "grad_norm": 0.14945626258850098,
      "learning_rate": 4.193884748673598e-05,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8256500698626041,
      "num_tokens": 604295507.0,
      "step": 18933
    },
    {
      "entropy": 0.663484388962388,
      "epoch": 1.742031385673458,
      "grad_norm": 0.16288040578365326,
      "learning_rate": 4.193578066059435e-05,
      "loss": 0.6507,
      "mean_token_accuracy": 0.8026250004768372,
      "num_tokens": 604327096.0,
      "step": 18934
    },
    {
      "entropy": 0.6392512619495392,
      "epoch": 1.742123392044899,
      "grad_norm": 0.1641092747449875,
      "learning_rate": 4.193271383445273e-05,
      "loss": 0.6183,
      "mean_token_accuracy": 0.8089606203138828,
      "num_tokens": 604358298.0,
      "step": 18935
    },
    {
      "entropy": 0.6274412889033556,
      "epoch": 1.7422153984163402,
      "grad_norm": 0.15229634940624237,
      "learning_rate": 4.19296470083111e-05,
      "loss": 0.6132,
      "mean_token_accuracy": 0.8127674572169781,
      "num_tokens": 604390145.0,
      "step": 18936
    },
    {
      "entropy": 0.5380861759185791,
      "epoch": 1.7423074047877816,
      "grad_norm": 0.14959117770195007,
      "learning_rate": 4.192658018216947e-05,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8331530839204788,
      "num_tokens": 604422616.0,
      "step": 18937
    },
    {
      "entropy": 0.6234356202185154,
      "epoch": 1.742399411159223,
      "grad_norm": 0.14894291758537292,
      "learning_rate": 4.192351335602785e-05,
      "loss": 0.6074,
      "mean_token_accuracy": 0.8118157275021076,
      "num_tokens": 604454948.0,
      "step": 18938
    },
    {
      "entropy": 0.6808871328830719,
      "epoch": 1.742491417530664,
      "grad_norm": 0.1601387858390808,
      "learning_rate": 4.192044652988622e-05,
      "loss": 0.6896,
      "mean_token_accuracy": 0.7918186411261559,
      "num_tokens": 604486960.0,
      "step": 18939
    },
    {
      "entropy": 0.6535183601081371,
      "epoch": 1.7425834239021052,
      "grad_norm": 0.1578759104013443,
      "learning_rate": 4.1917379703744594e-05,
      "loss": 0.6491,
      "mean_token_accuracy": 0.8007809743285179,
      "num_tokens": 604518524.0,
      "step": 18940
    },
    {
      "entropy": 0.5517001478001475,
      "epoch": 1.7426754302735463,
      "grad_norm": 0.15003737807273865,
      "learning_rate": 4.191431287760297e-05,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8344559036195278,
      "num_tokens": 604549517.0,
      "step": 18941
    },
    {
      "entropy": 0.581650972366333,
      "epoch": 1.7427674366449877,
      "grad_norm": 0.15148989856243134,
      "learning_rate": 4.191124605146134e-05,
      "loss": 0.5672,
      "mean_token_accuracy": 0.8276172839105129,
      "num_tokens": 604581597.0,
      "step": 18942
    },
    {
      "entropy": 0.644284755922854,
      "epoch": 1.742859443016429,
      "grad_norm": 0.1540708988904953,
      "learning_rate": 4.1908179225319724e-05,
      "loss": 0.6477,
      "mean_token_accuracy": 0.8054324015974998,
      "num_tokens": 604613557.0,
      "step": 18943
    },
    {
      "entropy": 0.6026373477652669,
      "epoch": 1.7429514493878702,
      "grad_norm": 0.1572548747062683,
      "learning_rate": 4.190511239917809e-05,
      "loss": 0.5846,
      "mean_token_accuracy": 0.8147625289857388,
      "num_tokens": 604644359.0,
      "step": 18944
    },
    {
      "entropy": 0.5785134369507432,
      "epoch": 1.7430434557593113,
      "grad_norm": 0.1561329960823059,
      "learning_rate": 4.190204557303647e-05,
      "loss": 0.5746,
      "mean_token_accuracy": 0.8251364380121231,
      "num_tokens": 604675196.0,
      "step": 18945
    },
    {
      "entropy": 0.6166622741147876,
      "epoch": 1.7431354621307524,
      "grad_norm": 0.15410089492797852,
      "learning_rate": 4.189897874689484e-05,
      "loss": 0.614,
      "mean_token_accuracy": 0.8138334713876247,
      "num_tokens": 604706350.0,
      "step": 18946
    },
    {
      "entropy": 0.5866459757089615,
      "epoch": 1.7432274685021938,
      "grad_norm": 0.1540946662425995,
      "learning_rate": 4.1895911920753216e-05,
      "loss": 0.5757,
      "mean_token_accuracy": 0.8188332952558994,
      "num_tokens": 604737478.0,
      "step": 18947
    },
    {
      "entropy": 0.603767222724855,
      "epoch": 1.7433194748736351,
      "grad_norm": 0.15813425183296204,
      "learning_rate": 4.189284509461159e-05,
      "loss": 0.5972,
      "mean_token_accuracy": 0.8163457252085209,
      "num_tokens": 604769566.0,
      "step": 18948
    },
    {
      "entropy": 0.6269696624949574,
      "epoch": 1.7434114812450763,
      "grad_norm": 0.15795360505580902,
      "learning_rate": 4.1889778268469966e-05,
      "loss": 0.6118,
      "mean_token_accuracy": 0.8166306056082249,
      "num_tokens": 604801365.0,
      "step": 18949
    },
    {
      "entropy": 0.6080723945051432,
      "epoch": 1.7435034876165174,
      "grad_norm": 0.1571371704339981,
      "learning_rate": 4.1886711442328333e-05,
      "loss": 0.6033,
      "mean_token_accuracy": 0.8096595332026482,
      "num_tokens": 604833310.0,
      "step": 18950
    },
    {
      "entropy": 0.637987706810236,
      "epoch": 1.7435954939879585,
      "grad_norm": 0.1572650969028473,
      "learning_rate": 4.1883644616186715e-05,
      "loss": 0.6299,
      "mean_token_accuracy": 0.8054482415318489,
      "num_tokens": 604864915.0,
      "step": 18951
    },
    {
      "entropy": 0.6016921950504184,
      "epoch": 1.7436875003594,
      "grad_norm": 0.1545862853527069,
      "learning_rate": 4.188057779004508e-05,
      "loss": 0.5816,
      "mean_token_accuracy": 0.819539126008749,
      "num_tokens": 604896720.0,
      "step": 18952
    },
    {
      "entropy": 0.6674605347216129,
      "epoch": 1.7437795067308413,
      "grad_norm": 0.1585761159658432,
      "learning_rate": 4.187751096390346e-05,
      "loss": 0.6589,
      "mean_token_accuracy": 0.8027708120644093,
      "num_tokens": 604928488.0,
      "step": 18953
    },
    {
      "entropy": 0.48735043220221996,
      "epoch": 1.7438715131022824,
      "grad_norm": 0.14084532856941223,
      "learning_rate": 4.187444413776183e-05,
      "loss": 0.4802,
      "mean_token_accuracy": 0.8461776971817017,
      "num_tokens": 604960626.0,
      "step": 18954
    },
    {
      "entropy": 0.5935924891382456,
      "epoch": 1.7439635194737235,
      "grad_norm": 0.15700243413448334,
      "learning_rate": 4.187137731162021e-05,
      "loss": 0.5805,
      "mean_token_accuracy": 0.8198273330926895,
      "num_tokens": 604992958.0,
      "step": 18955
    },
    {
      "entropy": 0.6535293643828481,
      "epoch": 1.7440555258451647,
      "grad_norm": 0.1581263691186905,
      "learning_rate": 4.186831048547858e-05,
      "loss": 0.6386,
      "mean_token_accuracy": 0.8048110119998455,
      "num_tokens": 605024256.0,
      "step": 18956
    },
    {
      "entropy": 0.6159811234101653,
      "epoch": 1.744147532216606,
      "grad_norm": 0.15611188113689423,
      "learning_rate": 4.1865243659336956e-05,
      "loss": 0.6042,
      "mean_token_accuracy": 0.8169330470263958,
      "num_tokens": 605055494.0,
      "step": 18957
    },
    {
      "entropy": 0.5287050567567348,
      "epoch": 1.7442395385880474,
      "grad_norm": 0.1477816104888916,
      "learning_rate": 4.1862176833195324e-05,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8377485573291779,
      "num_tokens": 605086997.0,
      "step": 18958
    },
    {
      "entropy": 0.6454189093783498,
      "epoch": 1.7443315449594885,
      "grad_norm": 0.1555437296628952,
      "learning_rate": 4.1859110007053705e-05,
      "loss": 0.6174,
      "mean_token_accuracy": 0.8076643869280815,
      "num_tokens": 605118306.0,
      "step": 18959
    },
    {
      "entropy": 0.6116811903193593,
      "epoch": 1.7444235513309296,
      "grad_norm": 0.1573735773563385,
      "learning_rate": 4.185604318091207e-05,
      "loss": 0.5937,
      "mean_token_accuracy": 0.8159789927303791,
      "num_tokens": 605150495.0,
      "step": 18960
    },
    {
      "entropy": 0.6028266255743802,
      "epoch": 1.7445155577023708,
      "grad_norm": 0.1509263813495636,
      "learning_rate": 4.1852976354770455e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8207990638911724,
      "num_tokens": 605182881.0,
      "step": 18961
    },
    {
      "entropy": 0.5606627967208624,
      "epoch": 1.7446075640738121,
      "grad_norm": 0.15049923956394196,
      "learning_rate": 4.184990952862882e-05,
      "loss": 0.5437,
      "mean_token_accuracy": 0.8325780853629112,
      "num_tokens": 605214758.0,
      "step": 18962
    },
    {
      "entropy": 0.5549530182033777,
      "epoch": 1.7446995704452535,
      "grad_norm": 0.14853809773921967,
      "learning_rate": 4.18468427024872e-05,
      "loss": 0.5372,
      "mean_token_accuracy": 0.837111834436655,
      "num_tokens": 605246231.0,
      "step": 18963
    },
    {
      "entropy": 0.5689058932475746,
      "epoch": 1.7447915768166946,
      "grad_norm": 0.15137608349323273,
      "learning_rate": 4.184377587634557e-05,
      "loss": 0.5628,
      "mean_token_accuracy": 0.8220884129405022,
      "num_tokens": 605278099.0,
      "step": 18964
    },
    {
      "entropy": 0.6871358882635832,
      "epoch": 1.7448835831881357,
      "grad_norm": 0.16764889657497406,
      "learning_rate": 4.1840709050203946e-05,
      "loss": 0.683,
      "mean_token_accuracy": 0.7940003089606762,
      "num_tokens": 605310040.0,
      "step": 18965
    },
    {
      "entropy": 0.5799547750502825,
      "epoch": 1.7449755895595769,
      "grad_norm": 0.1503172665834427,
      "learning_rate": 4.183764222406232e-05,
      "loss": 0.5803,
      "mean_token_accuracy": 0.8235917203128338,
      "num_tokens": 605342530.0,
      "step": 18966
    },
    {
      "entropy": 0.5624538278207183,
      "epoch": 1.7450675959310182,
      "grad_norm": 0.1551269292831421,
      "learning_rate": 4.1834575397920696e-05,
      "loss": 0.5444,
      "mean_token_accuracy": 0.8301258608698845,
      "num_tokens": 605374581.0,
      "step": 18967
    },
    {
      "entropy": 0.6331241601146758,
      "epoch": 1.7451596023024596,
      "grad_norm": 0.15536271035671234,
      "learning_rate": 4.1831508571779064e-05,
      "loss": 0.6388,
      "mean_token_accuracy": 0.8059673123061657,
      "num_tokens": 605406388.0,
      "step": 18968
    },
    {
      "entropy": 0.7180468160659075,
      "epoch": 1.7452516086739007,
      "grad_norm": 0.16607239842414856,
      "learning_rate": 4.1828441745637445e-05,
      "loss": 0.7285,
      "mean_token_accuracy": 0.7817682214081287,
      "num_tokens": 605439002.0,
      "step": 18969
    },
    {
      "entropy": 0.6011262219399214,
      "epoch": 1.7453436150453419,
      "grad_norm": 0.15607422590255737,
      "learning_rate": 4.182537491949581e-05,
      "loss": 0.5895,
      "mean_token_accuracy": 0.8162476979196072,
      "num_tokens": 605471283.0,
      "step": 18970
    },
    {
      "entropy": 0.5957333920523524,
      "epoch": 1.745435621416783,
      "grad_norm": 0.1581902652978897,
      "learning_rate": 4.182230809335419e-05,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8197089917957783,
      "num_tokens": 605502993.0,
      "step": 18971
    },
    {
      "entropy": 0.6450443919748068,
      "epoch": 1.7455276277882243,
      "grad_norm": 0.15700238943099976,
      "learning_rate": 4.181924126721256e-05,
      "loss": 0.6369,
      "mean_token_accuracy": 0.8049737773835659,
      "num_tokens": 605534437.0,
      "step": 18972
    },
    {
      "entropy": 0.5163764441385865,
      "epoch": 1.7456196341596657,
      "grad_norm": 0.14451168477535248,
      "learning_rate": 4.181617444107094e-05,
      "loss": 0.5114,
      "mean_token_accuracy": 0.8440160006284714,
      "num_tokens": 605566647.0,
      "step": 18973
    },
    {
      "entropy": 0.6139185521751642,
      "epoch": 1.7457116405311068,
      "grad_norm": 0.15109451115131378,
      "learning_rate": 4.181310761492931e-05,
      "loss": 0.6023,
      "mean_token_accuracy": 0.8087961226701736,
      "num_tokens": 605598780.0,
      "step": 18974
    },
    {
      "entropy": 0.6793694561347365,
      "epoch": 1.745803646902548,
      "grad_norm": 0.16305597126483917,
      "learning_rate": 4.1810040788787686e-05,
      "loss": 0.6672,
      "mean_token_accuracy": 0.7965548448264599,
      "num_tokens": 605629987.0,
      "step": 18975
    },
    {
      "entropy": 0.6333385407924652,
      "epoch": 1.745895653273989,
      "grad_norm": 0.1607937067747116,
      "learning_rate": 4.180697396264606e-05,
      "loss": 0.6372,
      "mean_token_accuracy": 0.8073692508041859,
      "num_tokens": 605662432.0,
      "step": 18976
    },
    {
      "entropy": 0.4991969335824251,
      "epoch": 1.7459876596454305,
      "grad_norm": 0.15178313851356506,
      "learning_rate": 4.1803907136504435e-05,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8462244719266891,
      "num_tokens": 605694705.0,
      "step": 18977
    },
    {
      "entropy": 0.5606359678786248,
      "epoch": 1.7460796660168718,
      "grad_norm": 0.1464255154132843,
      "learning_rate": 4.180084031036281e-05,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8269402086734772,
      "num_tokens": 605727238.0,
      "step": 18978
    },
    {
      "entropy": 0.514404300134629,
      "epoch": 1.746171672388313,
      "grad_norm": 0.15298017859458923,
      "learning_rate": 4.179777348422118e-05,
      "loss": 0.4879,
      "mean_token_accuracy": 0.8482009619474411,
      "num_tokens": 605759062.0,
      "step": 18979
    },
    {
      "entropy": 0.5667544202879071,
      "epoch": 1.746263678759754,
      "grad_norm": 0.15224696695804596,
      "learning_rate": 4.179470665807956e-05,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8273454010486603,
      "num_tokens": 605790726.0,
      "step": 18980
    },
    {
      "entropy": 0.6791427042335272,
      "epoch": 1.7463556851311952,
      "grad_norm": 0.15961354970932007,
      "learning_rate": 4.179163983193793e-05,
      "loss": 0.6627,
      "mean_token_accuracy": 0.7967328540980816,
      "num_tokens": 605822927.0,
      "step": 18981
    },
    {
      "entropy": 0.6231837207451463,
      "epoch": 1.7464476915026366,
      "grad_norm": 0.15477657318115234,
      "learning_rate": 4.178857300579631e-05,
      "loss": 0.612,
      "mean_token_accuracy": 0.8093395456671715,
      "num_tokens": 605854571.0,
      "step": 18982
    },
    {
      "entropy": 0.7074177749454975,
      "epoch": 1.746539697874078,
      "grad_norm": 0.16220296919345856,
      "learning_rate": 4.1785506179654676e-05,
      "loss": 0.6971,
      "mean_token_accuracy": 0.7870027311146259,
      "num_tokens": 605885851.0,
      "step": 18983
    },
    {
      "entropy": 0.5888018812984228,
      "epoch": 1.746631704245519,
      "grad_norm": 0.15055780112743378,
      "learning_rate": 4.178243935351305e-05,
      "loss": 0.5823,
      "mean_token_accuracy": 0.824279174208641,
      "num_tokens": 605918124.0,
      "step": 18984
    },
    {
      "entropy": 0.5618829606100917,
      "epoch": 1.7467237106169602,
      "grad_norm": 0.14938855171203613,
      "learning_rate": 4.1779372527371426e-05,
      "loss": 0.5444,
      "mean_token_accuracy": 0.8287164904177189,
      "num_tokens": 605950070.0,
      "step": 18985
    },
    {
      "entropy": 0.6712947878986597,
      "epoch": 1.7468157169884013,
      "grad_norm": 0.15892605483531952,
      "learning_rate": 4.17763057012298e-05,
      "loss": 0.6636,
      "mean_token_accuracy": 0.796335980296135,
      "num_tokens": 605981804.0,
      "step": 18986
    },
    {
      "entropy": 0.5982852419838309,
      "epoch": 1.7469077233598427,
      "grad_norm": 0.15163947641849518,
      "learning_rate": 4.1773238875088175e-05,
      "loss": 0.5887,
      "mean_token_accuracy": 0.8168790340423584,
      "num_tokens": 606013932.0,
      "step": 18987
    },
    {
      "entropy": 0.6287400112487376,
      "epoch": 1.746999729731284,
      "grad_norm": 0.1533065140247345,
      "learning_rate": 4.177017204894655e-05,
      "loss": 0.6017,
      "mean_token_accuracy": 0.8098687119781971,
      "num_tokens": 606045329.0,
      "step": 18988
    },
    {
      "entropy": 0.7039467515423894,
      "epoch": 1.7470917361027252,
      "grad_norm": 0.1582724153995514,
      "learning_rate": 4.176710522280492e-05,
      "loss": 0.6832,
      "mean_token_accuracy": 0.7925003208220005,
      "num_tokens": 606076858.0,
      "step": 18989
    },
    {
      "entropy": 0.6050849466118962,
      "epoch": 1.7471837424741663,
      "grad_norm": 0.15288561582565308,
      "learning_rate": 4.17640383966633e-05,
      "loss": 0.5883,
      "mean_token_accuracy": 0.8198932409286499,
      "num_tokens": 606109517.0,
      "step": 18990
    },
    {
      "entropy": 0.5477679315954447,
      "epoch": 1.7472757488456074,
      "grad_norm": 0.1610974818468094,
      "learning_rate": 4.176097157052167e-05,
      "loss": 0.5555,
      "mean_token_accuracy": 0.8306951597332954,
      "num_tokens": 606141715.0,
      "step": 18991
    },
    {
      "entropy": 0.5150000443682075,
      "epoch": 1.7473677552170488,
      "grad_norm": 0.1462128907442093,
      "learning_rate": 4.175790474438004e-05,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8465107120573521,
      "num_tokens": 606173556.0,
      "step": 18992
    },
    {
      "entropy": 0.5654730640817434,
      "epoch": 1.7474597615884901,
      "grad_norm": 0.14675763249397278,
      "learning_rate": 4.1754837918238416e-05,
      "loss": 0.5494,
      "mean_token_accuracy": 0.830366313457489,
      "num_tokens": 606205760.0,
      "step": 18993
    },
    {
      "entropy": 0.6204254208132625,
      "epoch": 1.7475517679599313,
      "grad_norm": 0.15745767951011658,
      "learning_rate": 4.175177109209679e-05,
      "loss": 0.6198,
      "mean_token_accuracy": 0.8118895180523396,
      "num_tokens": 606237843.0,
      "step": 18994
    },
    {
      "entropy": 0.568571004550904,
      "epoch": 1.7476437743313724,
      "grad_norm": 0.15180884301662445,
      "learning_rate": 4.1748704265955165e-05,
      "loss": 0.5487,
      "mean_token_accuracy": 0.8318855576217175,
      "num_tokens": 606269623.0,
      "step": 18995
    },
    {
      "entropy": 0.5725711663253605,
      "epoch": 1.7477357807028135,
      "grad_norm": 0.15693944692611694,
      "learning_rate": 4.174563743981354e-05,
      "loss": 0.5627,
      "mean_token_accuracy": 0.825993001461029,
      "num_tokens": 606301565.0,
      "step": 18996
    },
    {
      "entropy": 0.6974701303988695,
      "epoch": 1.747827787074255,
      "grad_norm": 0.16501709818840027,
      "learning_rate": 4.174257061367191e-05,
      "loss": 0.6956,
      "mean_token_accuracy": 0.7875528112053871,
      "num_tokens": 606332767.0,
      "step": 18997
    },
    {
      "entropy": 0.5227490486577153,
      "epoch": 1.7479197934456963,
      "grad_norm": 0.14821331202983856,
      "learning_rate": 4.173950378753029e-05,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8400495536625385,
      "num_tokens": 606364149.0,
      "step": 18998
    },
    {
      "entropy": 0.5593122662976384,
      "epoch": 1.7480117998171374,
      "grad_norm": 0.15746040642261505,
      "learning_rate": 4.173643696138866e-05,
      "loss": 0.5656,
      "mean_token_accuracy": 0.8293017745018005,
      "num_tokens": 606395821.0,
      "step": 18999
    },
    {
      "entropy": 0.5308979842811823,
      "epoch": 1.7481038061885785,
      "grad_norm": 0.1433177888393402,
      "learning_rate": 4.173337013524703e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8314257636666298,
      "num_tokens": 606428140.0,
      "step": 19000
    },
    {
      "entropy": 0.5064555164426565,
      "epoch": 1.7481958125600197,
      "grad_norm": 0.15682223439216614,
      "learning_rate": 4.1730303309105407e-05,
      "loss": 0.494,
      "mean_token_accuracy": 0.8479796797037125,
      "num_tokens": 606460796.0,
      "step": 19001
    },
    {
      "entropy": 0.7496210597455502,
      "epoch": 1.748287818931461,
      "grad_norm": 0.16281184554100037,
      "learning_rate": 4.172723648296378e-05,
      "loss": 0.7375,
      "mean_token_accuracy": 0.77632200345397,
      "num_tokens": 606492336.0,
      "step": 19002
    },
    {
      "entropy": 0.5848562056198716,
      "epoch": 1.7483798253029024,
      "grad_norm": 0.1534808725118637,
      "learning_rate": 4.1724169656822156e-05,
      "loss": 0.5778,
      "mean_token_accuracy": 0.8240968137979507,
      "num_tokens": 606525047.0,
      "step": 19003
    },
    {
      "entropy": 0.6329195639118552,
      "epoch": 1.7484718316743435,
      "grad_norm": 0.15835855901241302,
      "learning_rate": 4.172110283068053e-05,
      "loss": 0.6122,
      "mean_token_accuracy": 0.8137977607548237,
      "num_tokens": 606556956.0,
      "step": 19004
    },
    {
      "entropy": 0.7202121298760176,
      "epoch": 1.7485638380457846,
      "grad_norm": 0.1570628434419632,
      "learning_rate": 4.1718036004538905e-05,
      "loss": 0.706,
      "mean_token_accuracy": 0.7868960537016392,
      "num_tokens": 606589343.0,
      "step": 19005
    },
    {
      "entropy": 0.7022245340049267,
      "epoch": 1.7486558444172258,
      "grad_norm": 0.16115370392799377,
      "learning_rate": 4.171496917839728e-05,
      "loss": 0.6814,
      "mean_token_accuracy": 0.7911697626113892,
      "num_tokens": 606621241.0,
      "step": 19006
    },
    {
      "entropy": 0.6369646470993757,
      "epoch": 1.7487478507886671,
      "grad_norm": 0.1591259092092514,
      "learning_rate": 4.1711902352255654e-05,
      "loss": 0.622,
      "mean_token_accuracy": 0.8114693686366081,
      "num_tokens": 606652503.0,
      "step": 19007
    },
    {
      "entropy": 0.5740658175200224,
      "epoch": 1.7488398571601085,
      "grad_norm": 0.15258148312568665,
      "learning_rate": 4.170883552611403e-05,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8282914087176323,
      "num_tokens": 606684230.0,
      "step": 19008
    },
    {
      "entropy": 0.6268722582608461,
      "epoch": 1.7489318635315496,
      "grad_norm": 0.15249574184417725,
      "learning_rate": 4.1705768699972404e-05,
      "loss": 0.6108,
      "mean_token_accuracy": 0.8101908192038536,
      "num_tokens": 606716887.0,
      "step": 19009
    },
    {
      "entropy": 0.5436038998886943,
      "epoch": 1.7490238699029907,
      "grad_norm": 0.14623169600963593,
      "learning_rate": 4.170270187383077e-05,
      "loss": 0.5316,
      "mean_token_accuracy": 0.8332532234489918,
      "num_tokens": 606749017.0,
      "step": 19010
    },
    {
      "entropy": 0.6167960520833731,
      "epoch": 1.7491158762744319,
      "grad_norm": 0.1540830135345459,
      "learning_rate": 4.169963504768915e-05,
      "loss": 0.5921,
      "mean_token_accuracy": 0.8188757449388504,
      "num_tokens": 606780680.0,
      "step": 19011
    },
    {
      "entropy": 0.5941995307803154,
      "epoch": 1.7492078826458732,
      "grad_norm": 0.15129563212394714,
      "learning_rate": 4.169656822154752e-05,
      "loss": 0.5823,
      "mean_token_accuracy": 0.8214666359126568,
      "num_tokens": 606813172.0,
      "step": 19012
    },
    {
      "entropy": 0.5838430752046406,
      "epoch": 1.7492998890173146,
      "grad_norm": 0.15049365162849426,
      "learning_rate": 4.1693501395405896e-05,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8287052512168884,
      "num_tokens": 606845275.0,
      "step": 19013
    },
    {
      "entropy": 0.6982726817950606,
      "epoch": 1.7493918953887557,
      "grad_norm": 0.15394531190395355,
      "learning_rate": 4.169043456926427e-05,
      "loss": 0.6861,
      "mean_token_accuracy": 0.787609051913023,
      "num_tokens": 606877697.0,
      "step": 19014
    },
    {
      "entropy": 0.5787779670208693,
      "epoch": 1.7494839017601969,
      "grad_norm": 0.15511462092399597,
      "learning_rate": 4.1687367743122645e-05,
      "loss": 0.571,
      "mean_token_accuracy": 0.820647094398737,
      "num_tokens": 606909769.0,
      "step": 19015
    },
    {
      "entropy": 0.5180239547044039,
      "epoch": 1.749575908131638,
      "grad_norm": 0.14194633066654205,
      "learning_rate": 4.168430091698102e-05,
      "loss": 0.5054,
      "mean_token_accuracy": 0.844327300786972,
      "num_tokens": 606942238.0,
      "step": 19016
    },
    {
      "entropy": 0.5624481271952391,
      "epoch": 1.7496679145030793,
      "grad_norm": 0.152133971452713,
      "learning_rate": 4.1681234090839394e-05,
      "loss": 0.5505,
      "mean_token_accuracy": 0.8274163492023945,
      "num_tokens": 606974313.0,
      "step": 19017
    },
    {
      "entropy": 0.6466563753783703,
      "epoch": 1.7497599208745207,
      "grad_norm": 0.15899744629859924,
      "learning_rate": 4.167816726469776e-05,
      "loss": 0.6393,
      "mean_token_accuracy": 0.8013211339712143,
      "num_tokens": 607006180.0,
      "step": 19018
    },
    {
      "entropy": 0.5906625334173441,
      "epoch": 1.7498519272459618,
      "grad_norm": 0.14971429109573364,
      "learning_rate": 4.1675100438556143e-05,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8173555098474026,
      "num_tokens": 607038102.0,
      "step": 19019
    },
    {
      "entropy": 0.6697855070233345,
      "epoch": 1.749943933617403,
      "grad_norm": 0.15961290895938873,
      "learning_rate": 4.167203361241451e-05,
      "loss": 0.6658,
      "mean_token_accuracy": 0.7976555861532688,
      "num_tokens": 607070152.0,
      "step": 19020
    },
    {
      "entropy": 0.5052434606477618,
      "epoch": 1.750035939988844,
      "grad_norm": 0.15170276165008545,
      "learning_rate": 4.166896678627289e-05,
      "loss": 0.5029,
      "mean_token_accuracy": 0.8432533480226994,
      "num_tokens": 607102238.0,
      "step": 19021
    },
    {
      "entropy": 0.6473940084688365,
      "epoch": 1.7501279463602855,
      "grad_norm": 0.15709279477596283,
      "learning_rate": 4.166589996013126e-05,
      "loss": 0.6501,
      "mean_token_accuracy": 0.8013940118253231,
      "num_tokens": 607134527.0,
      "step": 19022
    },
    {
      "entropy": 0.5662986333481967,
      "epoch": 1.7502199527317268,
      "grad_norm": 0.15454289317131042,
      "learning_rate": 4.1662833133989635e-05,
      "loss": 0.5602,
      "mean_token_accuracy": 0.8265665173530579,
      "num_tokens": 607166073.0,
      "step": 19023
    },
    {
      "entropy": 0.5401400940027088,
      "epoch": 1.750311959103168,
      "grad_norm": 0.14884904026985168,
      "learning_rate": 4.165976630784801e-05,
      "loss": 0.526,
      "mean_token_accuracy": 0.8375366292893887,
      "num_tokens": 607198841.0,
      "step": 19024
    },
    {
      "entropy": 0.6351240370422602,
      "epoch": 1.750403965474609,
      "grad_norm": 0.16125145554542542,
      "learning_rate": 4.1656699481706385e-05,
      "loss": 0.6425,
      "mean_token_accuracy": 0.7999243214726448,
      "num_tokens": 607229816.0,
      "step": 19025
    },
    {
      "entropy": 0.5591802932322025,
      "epoch": 1.7504959718460502,
      "grad_norm": 0.14939168095588684,
      "learning_rate": 4.165363265556476e-05,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8287230916321278,
      "num_tokens": 607262167.0,
      "step": 19026
    },
    {
      "entropy": 0.6228501368314028,
      "epoch": 1.7505879782174916,
      "grad_norm": 0.15157249569892883,
      "learning_rate": 4.1650565829423134e-05,
      "loss": 0.622,
      "mean_token_accuracy": 0.8105882033705711,
      "num_tokens": 607294699.0,
      "step": 19027
    },
    {
      "entropy": 0.5607379684224725,
      "epoch": 1.750679984588933,
      "grad_norm": 0.16264155507087708,
      "learning_rate": 4.16474990032815e-05,
      "loss": 0.5447,
      "mean_token_accuracy": 0.8315287679433823,
      "num_tokens": 607326010.0,
      "step": 19028
    },
    {
      "entropy": 0.6501550637185574,
      "epoch": 1.750771990960374,
      "grad_norm": 0.1591395139694214,
      "learning_rate": 4.164443217713988e-05,
      "loss": 0.647,
      "mean_token_accuracy": 0.8026022911071777,
      "num_tokens": 607358147.0,
      "step": 19029
    },
    {
      "entropy": 0.5299756322056055,
      "epoch": 1.7508639973318152,
      "grad_norm": 0.14866812527179718,
      "learning_rate": 4.164136535099825e-05,
      "loss": 0.5186,
      "mean_token_accuracy": 0.839774027466774,
      "num_tokens": 607390039.0,
      "step": 19030
    },
    {
      "entropy": 0.6576817687600851,
      "epoch": 1.7509560037032563,
      "grad_norm": 0.15996263921260834,
      "learning_rate": 4.1638298524856626e-05,
      "loss": 0.6335,
      "mean_token_accuracy": 0.8047953173518181,
      "num_tokens": 607421316.0,
      "step": 19031
    },
    {
      "entropy": 0.4743795543909073,
      "epoch": 1.7510480100746977,
      "grad_norm": 0.15001723170280457,
      "learning_rate": 4.1635231698715e-05,
      "loss": 0.4587,
      "mean_token_accuracy": 0.8564139120280743,
      "num_tokens": 607452787.0,
      "step": 19032
    },
    {
      "entropy": 0.5854383483529091,
      "epoch": 1.751140016446139,
      "grad_norm": 0.15378813445568085,
      "learning_rate": 4.1632164872573375e-05,
      "loss": 0.5741,
      "mean_token_accuracy": 0.8247830234467983,
      "num_tokens": 607485105.0,
      "step": 19033
    },
    {
      "entropy": 0.6467857006937265,
      "epoch": 1.7512320228175802,
      "grad_norm": 0.15633746981620789,
      "learning_rate": 4.162909804643175e-05,
      "loss": 0.6336,
      "mean_token_accuracy": 0.8081753067672253,
      "num_tokens": 607516883.0,
      "step": 19034
    },
    {
      "entropy": 0.6381613612174988,
      "epoch": 1.7513240291890213,
      "grad_norm": 0.1546400934457779,
      "learning_rate": 4.1626031220290124e-05,
      "loss": 0.6335,
      "mean_token_accuracy": 0.8047912493348122,
      "num_tokens": 607548923.0,
      "step": 19035
    },
    {
      "entropy": 0.6949703246355057,
      "epoch": 1.7514160355604624,
      "grad_norm": 0.1602914184331894,
      "learning_rate": 4.16229643941485e-05,
      "loss": 0.6862,
      "mean_token_accuracy": 0.7954008653759956,
      "num_tokens": 607581156.0,
      "step": 19036
    },
    {
      "entropy": 0.6447350643575191,
      "epoch": 1.7515080419319038,
      "grad_norm": 0.15221348404884338,
      "learning_rate": 4.1619897568006874e-05,
      "loss": 0.6362,
      "mean_token_accuracy": 0.8073132261633873,
      "num_tokens": 607613551.0,
      "step": 19037
    },
    {
      "entropy": 0.6742587722837925,
      "epoch": 1.7516000483033451,
      "grad_norm": 0.15804649889469147,
      "learning_rate": 4.161683074186525e-05,
      "loss": 0.6645,
      "mean_token_accuracy": 0.7975421659648418,
      "num_tokens": 607644521.0,
      "step": 19038
    },
    {
      "entropy": 0.6451102048158646,
      "epoch": 1.7516920546747863,
      "grad_norm": 0.15510056912899017,
      "learning_rate": 4.1613763915723616e-05,
      "loss": 0.6409,
      "mean_token_accuracy": 0.8031417280435562,
      "num_tokens": 607676938.0,
      "step": 19039
    },
    {
      "entropy": 0.6302368380129337,
      "epoch": 1.7517840610462274,
      "grad_norm": 0.15532104671001434,
      "learning_rate": 4.1610697089582e-05,
      "loss": 0.6244,
      "mean_token_accuracy": 0.812759630382061,
      "num_tokens": 607709501.0,
      "step": 19040
    },
    {
      "entropy": 0.5529199186712503,
      "epoch": 1.7518760674176685,
      "grad_norm": 0.14736391603946686,
      "learning_rate": 4.1607630263440365e-05,
      "loss": 0.5365,
      "mean_token_accuracy": 0.8352396450936794,
      "num_tokens": 607741655.0,
      "step": 19041
    },
    {
      "entropy": 0.528076047077775,
      "epoch": 1.75196807378911,
      "grad_norm": 0.14286521077156067,
      "learning_rate": 4.160456343729875e-05,
      "loss": 0.5177,
      "mean_token_accuracy": 0.8399285636842251,
      "num_tokens": 607774123.0,
      "step": 19042
    },
    {
      "entropy": 0.5507827736437321,
      "epoch": 1.7520600801605513,
      "grad_norm": 0.15712861716747284,
      "learning_rate": 4.1601496611157115e-05,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8346259109675884,
      "num_tokens": 607806504.0,
      "step": 19043
    },
    {
      "entropy": 0.6202923581004143,
      "epoch": 1.7521520865319924,
      "grad_norm": 0.1585107147693634,
      "learning_rate": 4.159842978501549e-05,
      "loss": 0.6212,
      "mean_token_accuracy": 0.8103036731481552,
      "num_tokens": 607839040.0,
      "step": 19044
    },
    {
      "entropy": 0.6353509873151779,
      "epoch": 1.7522440929034335,
      "grad_norm": 0.16116423904895782,
      "learning_rate": 4.1595362958873864e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.8056905977427959,
      "num_tokens": 607870242.0,
      "step": 19045
    },
    {
      "entropy": 0.612562651745975,
      "epoch": 1.7523360992748747,
      "grad_norm": 0.15170685946941376,
      "learning_rate": 4.159229613273224e-05,
      "loss": 0.6124,
      "mean_token_accuracy": 0.8106423728168011,
      "num_tokens": 607902931.0,
      "step": 19046
    },
    {
      "entropy": 0.5965875843539834,
      "epoch": 1.752428105646316,
      "grad_norm": 0.15377937257289886,
      "learning_rate": 4.158922930659061e-05,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8181613571941853,
      "num_tokens": 607934142.0,
      "step": 19047
    },
    {
      "entropy": 0.5523910727351904,
      "epoch": 1.7525201120177574,
      "grad_norm": 0.15178915858268738,
      "learning_rate": 4.158616248044899e-05,
      "loss": 0.5443,
      "mean_token_accuracy": 0.8288293033838272,
      "num_tokens": 607966261.0,
      "step": 19048
    },
    {
      "entropy": 0.6389783988706768,
      "epoch": 1.7526121183891985,
      "grad_norm": 0.1556919813156128,
      "learning_rate": 4.1583095654307356e-05,
      "loss": 0.6406,
      "mean_token_accuracy": 0.8077142015099525,
      "num_tokens": 607998984.0,
      "step": 19049
    },
    {
      "entropy": 0.6319818166084588,
      "epoch": 1.7527041247606396,
      "grad_norm": 0.1569332629442215,
      "learning_rate": 4.158002882816574e-05,
      "loss": 0.621,
      "mean_token_accuracy": 0.8084758743643761,
      "num_tokens": 608030955.0,
      "step": 19050
    },
    {
      "entropy": 0.5832287105731666,
      "epoch": 1.7527961311320808,
      "grad_norm": 0.14836373925209045,
      "learning_rate": 4.1576962002024105e-05,
      "loss": 0.574,
      "mean_token_accuracy": 0.8203871250152588,
      "num_tokens": 608062182.0,
      "step": 19051
    },
    {
      "entropy": 0.6304134577512741,
      "epoch": 1.7528881375035221,
      "grad_norm": 0.16048355400562286,
      "learning_rate": 4.157389517588248e-05,
      "loss": 0.6175,
      "mean_token_accuracy": 0.811127133667469,
      "num_tokens": 608093377.0,
      "step": 19052
    },
    {
      "entropy": 0.5272427201271057,
      "epoch": 1.7529801438749635,
      "grad_norm": 0.1489383727312088,
      "learning_rate": 4.1570828349740854e-05,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8416360840201378,
      "num_tokens": 608125055.0,
      "step": 19053
    },
    {
      "entropy": 0.7491339333355427,
      "epoch": 1.7530721502464046,
      "grad_norm": 0.16751542687416077,
      "learning_rate": 4.156776152359923e-05,
      "loss": 0.7355,
      "mean_token_accuracy": 0.7727388702332973,
      "num_tokens": 608157134.0,
      "step": 19054
    },
    {
      "entropy": 0.6041224850341678,
      "epoch": 1.7531641566178457,
      "grad_norm": 0.1580776572227478,
      "learning_rate": 4.1564694697457604e-05,
      "loss": 0.5938,
      "mean_token_accuracy": 0.8179506585001945,
      "num_tokens": 608188177.0,
      "step": 19055
    },
    {
      "entropy": 0.5655973316170275,
      "epoch": 1.7532561629892869,
      "grad_norm": 0.14708811044692993,
      "learning_rate": 4.156162787131598e-05,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8316726759076118,
      "num_tokens": 608220518.0,
      "step": 19056
    },
    {
      "entropy": 0.622043730225414,
      "epoch": 1.7533481693607282,
      "grad_norm": 0.15081241726875305,
      "learning_rate": 4.1558561045174346e-05,
      "loss": 0.5949,
      "mean_token_accuracy": 0.8144946694374084,
      "num_tokens": 608252503.0,
      "step": 19057
    },
    {
      "entropy": 0.6325775468721986,
      "epoch": 1.7534401757321696,
      "grad_norm": 0.15453381836414337,
      "learning_rate": 4.155549421903273e-05,
      "loss": 0.6309,
      "mean_token_accuracy": 0.8064327538013458,
      "num_tokens": 608283862.0,
      "step": 19058
    },
    {
      "entropy": 0.5448739184066653,
      "epoch": 1.7535321821036107,
      "grad_norm": 0.14313529431819916,
      "learning_rate": 4.1552427392891095e-05,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8357763960957527,
      "num_tokens": 608315749.0,
      "step": 19059
    },
    {
      "entropy": 0.5695199863985181,
      "epoch": 1.7536241884750519,
      "grad_norm": 0.1566399484872818,
      "learning_rate": 4.154936056674947e-05,
      "loss": 0.5565,
      "mean_token_accuracy": 0.829145722091198,
      "num_tokens": 608347906.0,
      "step": 19060
    },
    {
      "entropy": 0.5934568154625595,
      "epoch": 1.753716194846493,
      "grad_norm": 0.15148305892944336,
      "learning_rate": 4.1546293740607845e-05,
      "loss": 0.5883,
      "mean_token_accuracy": 0.8225543424487114,
      "num_tokens": 608380467.0,
      "step": 19061
    },
    {
      "entropy": 0.6078901048749685,
      "epoch": 1.7538082012179343,
      "grad_norm": 0.15531061589717865,
      "learning_rate": 4.154322691446622e-05,
      "loss": 0.6012,
      "mean_token_accuracy": 0.8179597742855549,
      "num_tokens": 608412673.0,
      "step": 19062
    },
    {
      "entropy": 0.6017694985494018,
      "epoch": 1.7539002075893757,
      "grad_norm": 0.1538974493741989,
      "learning_rate": 4.1540160088324594e-05,
      "loss": 0.5836,
      "mean_token_accuracy": 0.8170096948742867,
      "num_tokens": 608443780.0,
      "step": 19063
    },
    {
      "entropy": 0.5750778615474701,
      "epoch": 1.7539922139608168,
      "grad_norm": 0.15747439861297607,
      "learning_rate": 4.153709326218297e-05,
      "loss": 0.5693,
      "mean_token_accuracy": 0.8267878964543343,
      "num_tokens": 608475669.0,
      "step": 19064
    },
    {
      "entropy": 0.6040342776104808,
      "epoch": 1.754084220332258,
      "grad_norm": 0.15350428223609924,
      "learning_rate": 4.153402643604134e-05,
      "loss": 0.6032,
      "mean_token_accuracy": 0.8174277655780315,
      "num_tokens": 608507693.0,
      "step": 19065
    },
    {
      "entropy": 0.4791451683267951,
      "epoch": 1.754176226703699,
      "grad_norm": 0.13961996138095856,
      "learning_rate": 4.153095960989972e-05,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8524732142686844,
      "num_tokens": 608539783.0,
      "step": 19066
    },
    {
      "entropy": 0.6088655833154917,
      "epoch": 1.7542682330751405,
      "grad_norm": 0.15448957681655884,
      "learning_rate": 4.152789278375809e-05,
      "loss": 0.59,
      "mean_token_accuracy": 0.8161894418299198,
      "num_tokens": 608572252.0,
      "step": 19067
    },
    {
      "entropy": 0.5401041463483125,
      "epoch": 1.7543602394465818,
      "grad_norm": 0.14954496920108795,
      "learning_rate": 4.152482595761647e-05,
      "loss": 0.5345,
      "mean_token_accuracy": 0.8346280194818974,
      "num_tokens": 608604368.0,
      "step": 19068
    },
    {
      "entropy": 0.5948046557605267,
      "epoch": 1.754452245818023,
      "grad_norm": 0.14984913170337677,
      "learning_rate": 4.152175913147484e-05,
      "loss": 0.5743,
      "mean_token_accuracy": 0.8206764832139015,
      "num_tokens": 608635795.0,
      "step": 19069
    },
    {
      "entropy": 0.5253113526850939,
      "epoch": 1.754544252189464,
      "grad_norm": 0.15125419199466705,
      "learning_rate": 4.151869230533321e-05,
      "loss": 0.5091,
      "mean_token_accuracy": 0.840433731675148,
      "num_tokens": 608667080.0,
      "step": 19070
    },
    {
      "entropy": 0.5303541999310255,
      "epoch": 1.7546362585609052,
      "grad_norm": 0.14883345365524292,
      "learning_rate": 4.151562547919159e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8389666303992271,
      "num_tokens": 608698708.0,
      "step": 19071
    },
    {
      "entropy": 0.6534818513318896,
      "epoch": 1.7547282649323466,
      "grad_norm": 0.1564939320087433,
      "learning_rate": 4.151255865304996e-05,
      "loss": 0.6556,
      "mean_token_accuracy": 0.8010306432843208,
      "num_tokens": 608730365.0,
      "step": 19072
    },
    {
      "entropy": 0.6190655999816954,
      "epoch": 1.754820271303788,
      "grad_norm": 0.17038288712501526,
      "learning_rate": 4.1509491826908334e-05,
      "loss": 0.6259,
      "mean_token_accuracy": 0.806782279163599,
      "num_tokens": 608762254.0,
      "step": 19073
    },
    {
      "entropy": 0.7227852176874876,
      "epoch": 1.754912277675229,
      "grad_norm": 0.16649657487869263,
      "learning_rate": 4.150642500076671e-05,
      "loss": 0.7156,
      "mean_token_accuracy": 0.7862221784889698,
      "num_tokens": 608794110.0,
      "step": 19074
    },
    {
      "entropy": 0.623510617762804,
      "epoch": 1.7550042840466702,
      "grad_norm": 0.16247422993183136,
      "learning_rate": 4.150335817462508e-05,
      "loss": 0.6154,
      "mean_token_accuracy": 0.8114158622920513,
      "num_tokens": 608825418.0,
      "step": 19075
    },
    {
      "entropy": 0.5731516107916832,
      "epoch": 1.7550962904181113,
      "grad_norm": 0.1508314162492752,
      "learning_rate": 4.150029134848346e-05,
      "loss": 0.5633,
      "mean_token_accuracy": 0.8290759399533272,
      "num_tokens": 608857592.0,
      "step": 19076
    },
    {
      "entropy": 0.7101188097149134,
      "epoch": 1.7551882967895527,
      "grad_norm": 0.1568392962217331,
      "learning_rate": 4.149722452234183e-05,
      "loss": 0.6911,
      "mean_token_accuracy": 0.7889789827167988,
      "num_tokens": 608889895.0,
      "step": 19077
    },
    {
      "entropy": 0.711393772624433,
      "epoch": 1.755280303160994,
      "grad_norm": 0.1604827344417572,
      "learning_rate": 4.14941576962002e-05,
      "loss": 0.692,
      "mean_token_accuracy": 0.7904000356793404,
      "num_tokens": 608920863.0,
      "step": 19078
    },
    {
      "entropy": 0.5447638342157006,
      "epoch": 1.7553723095324352,
      "grad_norm": 0.1461867243051529,
      "learning_rate": 4.149109087005858e-05,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8340710327029228,
      "num_tokens": 608952735.0,
      "step": 19079
    },
    {
      "entropy": 0.6904314253479242,
      "epoch": 1.7554643159038763,
      "grad_norm": 0.1619957685470581,
      "learning_rate": 4.148802404391695e-05,
      "loss": 0.6864,
      "mean_token_accuracy": 0.7936356514692307,
      "num_tokens": 608985084.0,
      "step": 19080
    },
    {
      "entropy": 0.5793271586298943,
      "epoch": 1.7555563222753174,
      "grad_norm": 0.14886049926280975,
      "learning_rate": 4.148495721777533e-05,
      "loss": 0.5681,
      "mean_token_accuracy": 0.824138455092907,
      "num_tokens": 609017077.0,
      "step": 19081
    },
    {
      "entropy": 0.5702716838568449,
      "epoch": 1.7556483286467588,
      "grad_norm": 0.1495383232831955,
      "learning_rate": 4.14818903916337e-05,
      "loss": 0.5572,
      "mean_token_accuracy": 0.8296073824167252,
      "num_tokens": 609049119.0,
      "step": 19082
    },
    {
      "entropy": 0.6730485912412405,
      "epoch": 1.7557403350182001,
      "grad_norm": 0.15848074853420258,
      "learning_rate": 4.1478823565492073e-05,
      "loss": 0.6536,
      "mean_token_accuracy": 0.8017488643527031,
      "num_tokens": 609080650.0,
      "step": 19083
    },
    {
      "entropy": 0.48264908976852894,
      "epoch": 1.7558323413896413,
      "grad_norm": 0.1443280428647995,
      "learning_rate": 4.147575673935045e-05,
      "loss": 0.4705,
      "mean_token_accuracy": 0.8544499315321445,
      "num_tokens": 609112745.0,
      "step": 19084
    },
    {
      "entropy": 0.6633457564748824,
      "epoch": 1.7559243477610824,
      "grad_norm": 0.16130311787128448,
      "learning_rate": 4.147268991320882e-05,
      "loss": 0.6565,
      "mean_token_accuracy": 0.805428046733141,
      "num_tokens": 609144051.0,
      "step": 19085
    },
    {
      "entropy": 0.6554378848522902,
      "epoch": 1.7560163541325235,
      "grad_norm": 0.15971790254116058,
      "learning_rate": 4.14696230870672e-05,
      "loss": 0.653,
      "mean_token_accuracy": 0.7993462309241295,
      "num_tokens": 609176545.0,
      "step": 19086
    },
    {
      "entropy": 0.622024105861783,
      "epoch": 1.756108360503965,
      "grad_norm": 0.15592943131923676,
      "learning_rate": 4.146655626092557e-05,
      "loss": 0.6181,
      "mean_token_accuracy": 0.8139617629349232,
      "num_tokens": 609208832.0,
      "step": 19087
    },
    {
      "entropy": 0.526950185187161,
      "epoch": 1.7562003668754063,
      "grad_norm": 0.14660440385341644,
      "learning_rate": 4.146348943478394e-05,
      "loss": 0.5248,
      "mean_token_accuracy": 0.8394265398383141,
      "num_tokens": 609240833.0,
      "step": 19088
    },
    {
      "entropy": 0.5801294092088938,
      "epoch": 1.7562923732468474,
      "grad_norm": 0.16484199464321136,
      "learning_rate": 4.146042260864232e-05,
      "loss": 0.5695,
      "mean_token_accuracy": 0.8193403445184231,
      "num_tokens": 609272735.0,
      "step": 19089
    },
    {
      "entropy": 0.6048416667617857,
      "epoch": 1.7563843796182885,
      "grad_norm": 0.1551641970872879,
      "learning_rate": 4.145735578250069e-05,
      "loss": 0.5937,
      "mean_token_accuracy": 0.8157837949693203,
      "num_tokens": 609305153.0,
      "step": 19090
    },
    {
      "entropy": 0.522768447175622,
      "epoch": 1.7564763859897297,
      "grad_norm": 0.16019658744335175,
      "learning_rate": 4.1454288956359064e-05,
      "loss": 0.515,
      "mean_token_accuracy": 0.8449134901165962,
      "num_tokens": 609337497.0,
      "step": 19091
    },
    {
      "entropy": 0.6596762798726559,
      "epoch": 1.756568392361171,
      "grad_norm": 0.15850839018821716,
      "learning_rate": 4.145122213021744e-05,
      "loss": 0.6546,
      "mean_token_accuracy": 0.8014524467289448,
      "num_tokens": 609369846.0,
      "step": 19092
    },
    {
      "entropy": 0.6601245608180761,
      "epoch": 1.7566603987326124,
      "grad_norm": 0.1620195209980011,
      "learning_rate": 4.144815530407581e-05,
      "loss": 0.6694,
      "mean_token_accuracy": 0.7955327667295933,
      "num_tokens": 609402246.0,
      "step": 19093
    },
    {
      "entropy": 0.601070936769247,
      "epoch": 1.7567524051040535,
      "grad_norm": 0.15716052055358887,
      "learning_rate": 4.144508847793419e-05,
      "loss": 0.5854,
      "mean_token_accuracy": 0.8205757178366184,
      "num_tokens": 609434357.0,
      "step": 19094
    },
    {
      "entropy": 0.6158333644270897,
      "epoch": 1.7568444114754946,
      "grad_norm": 0.15335120260715485,
      "learning_rate": 4.144202165179256e-05,
      "loss": 0.6042,
      "mean_token_accuracy": 0.8113305456936359,
      "num_tokens": 609466710.0,
      "step": 19095
    },
    {
      "entropy": 0.5334361339919269,
      "epoch": 1.7569364178469358,
      "grad_norm": 0.1452058106660843,
      "learning_rate": 4.143895482565094e-05,
      "loss": 0.5118,
      "mean_token_accuracy": 0.842510387301445,
      "num_tokens": 609498738.0,
      "step": 19096
    },
    {
      "entropy": 0.6153035778552294,
      "epoch": 1.7570284242183771,
      "grad_norm": 0.16144175827503204,
      "learning_rate": 4.143588799950931e-05,
      "loss": 0.6041,
      "mean_token_accuracy": 0.8130035847425461,
      "num_tokens": 609530049.0,
      "step": 19097
    },
    {
      "entropy": 0.5972603810951114,
      "epoch": 1.7571204305898185,
      "grad_norm": 0.15520331263542175,
      "learning_rate": 4.1432821173367686e-05,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8181374408304691,
      "num_tokens": 609562110.0,
      "step": 19098
    },
    {
      "entropy": 0.6671513337641954,
      "epoch": 1.7572124369612596,
      "grad_norm": 0.16137298941612244,
      "learning_rate": 4.1429754347226054e-05,
      "loss": 0.6542,
      "mean_token_accuracy": 0.8027463257312775,
      "num_tokens": 609593743.0,
      "step": 19099
    },
    {
      "entropy": 0.601836696267128,
      "epoch": 1.7573044433327007,
      "grad_norm": 0.15616187453269958,
      "learning_rate": 4.1426687521084436e-05,
      "loss": 0.6043,
      "mean_token_accuracy": 0.8150919005274773,
      "num_tokens": 609626343.0,
      "step": 19100
    },
    {
      "entropy": 0.6370095405727625,
      "epoch": 1.7573964497041419,
      "grad_norm": 0.15717121958732605,
      "learning_rate": 4.1423620694942804e-05,
      "loss": 0.6331,
      "mean_token_accuracy": 0.8100361563265324,
      "num_tokens": 609658731.0,
      "step": 19101
    },
    {
      "entropy": 0.6161282602697611,
      "epoch": 1.7574884560755832,
      "grad_norm": 0.15423640608787537,
      "learning_rate": 4.1420553868801185e-05,
      "loss": 0.6012,
      "mean_token_accuracy": 0.8137118779122829,
      "num_tokens": 609689341.0,
      "step": 19102
    },
    {
      "entropy": 0.6135658631101251,
      "epoch": 1.7575804624470246,
      "grad_norm": 0.1560726761817932,
      "learning_rate": 4.141748704265955e-05,
      "loss": 0.5905,
      "mean_token_accuracy": 0.8149869330227375,
      "num_tokens": 609721312.0,
      "step": 19103
    },
    {
      "entropy": 0.5984882973134518,
      "epoch": 1.7576724688184657,
      "grad_norm": 0.14566336572170258,
      "learning_rate": 4.141442021651793e-05,
      "loss": 0.5729,
      "mean_token_accuracy": 0.818773228675127,
      "num_tokens": 609753434.0,
      "step": 19104
    },
    {
      "entropy": 0.5799127644859254,
      "epoch": 1.7577644751899069,
      "grad_norm": 0.15437600016593933,
      "learning_rate": 4.14113533903763e-05,
      "loss": 0.5751,
      "mean_token_accuracy": 0.8196482993662357,
      "num_tokens": 609784900.0,
      "step": 19105
    },
    {
      "entropy": 0.5185920773074031,
      "epoch": 1.757856481561348,
      "grad_norm": 0.15107308328151703,
      "learning_rate": 4.140828656423468e-05,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8431973867118359,
      "num_tokens": 609816621.0,
      "step": 19106
    },
    {
      "entropy": 0.6307043414562941,
      "epoch": 1.7579484879327893,
      "grad_norm": 0.15512540936470032,
      "learning_rate": 4.140521973809305e-05,
      "loss": 0.609,
      "mean_token_accuracy": 0.8079760037362576,
      "num_tokens": 609848651.0,
      "step": 19107
    },
    {
      "entropy": 0.6008984446525574,
      "epoch": 1.7580404943042307,
      "grad_norm": 0.1487671434879303,
      "learning_rate": 4.1402152911951426e-05,
      "loss": 0.5914,
      "mean_token_accuracy": 0.8153907433152199,
      "num_tokens": 609880578.0,
      "step": 19108
    },
    {
      "entropy": 0.5886555202305317,
      "epoch": 1.7581325006756718,
      "grad_norm": 0.14759822189807892,
      "learning_rate": 4.1399086085809794e-05,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8226158879697323,
      "num_tokens": 609912331.0,
      "step": 19109
    },
    {
      "entropy": 0.5601716786623001,
      "epoch": 1.758224507047113,
      "grad_norm": 0.14243362843990326,
      "learning_rate": 4.1396019259668175e-05,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8299310132861137,
      "num_tokens": 609944422.0,
      "step": 19110
    },
    {
      "entropy": 0.6577439475804567,
      "epoch": 1.758316513418554,
      "grad_norm": 0.15354831516742706,
      "learning_rate": 4.139295243352654e-05,
      "loss": 0.659,
      "mean_token_accuracy": 0.8024037964642048,
      "num_tokens": 609976879.0,
      "step": 19111
    },
    {
      "entropy": 0.6564547251909971,
      "epoch": 1.7584085197899955,
      "grad_norm": 0.16008038818836212,
      "learning_rate": 4.138988560738492e-05,
      "loss": 0.6548,
      "mean_token_accuracy": 0.8004736974835396,
      "num_tokens": 610008904.0,
      "step": 19112
    },
    {
      "entropy": 0.5072426935657859,
      "epoch": 1.7585005261614368,
      "grad_norm": 0.14572015404701233,
      "learning_rate": 4.138681878124329e-05,
      "loss": 0.4928,
      "mean_token_accuracy": 0.8454993963241577,
      "num_tokens": 610040131.0,
      "step": 19113
    },
    {
      "entropy": 0.6068824231624603,
      "epoch": 1.758592532532878,
      "grad_norm": 0.1556551456451416,
      "learning_rate": 4.138375195510167e-05,
      "loss": 0.6032,
      "mean_token_accuracy": 0.817047480493784,
      "num_tokens": 610071578.0,
      "step": 19114
    },
    {
      "entropy": 0.6587246139533818,
      "epoch": 1.758684538904319,
      "grad_norm": 0.16079913079738617,
      "learning_rate": 4.138068512896004e-05,
      "loss": 0.6528,
      "mean_token_accuracy": 0.8032409846782684,
      "num_tokens": 610103527.0,
      "step": 19115
    },
    {
      "entropy": 0.6036436017602682,
      "epoch": 1.7587765452757602,
      "grad_norm": 0.15284408628940582,
      "learning_rate": 4.1377618302818417e-05,
      "loss": 0.607,
      "mean_token_accuracy": 0.8146251291036606,
      "num_tokens": 610135628.0,
      "step": 19116
    },
    {
      "entropy": 0.5563490940257907,
      "epoch": 1.7588685516472016,
      "grad_norm": 0.15412887930870056,
      "learning_rate": 4.1374551476676784e-05,
      "loss": 0.5447,
      "mean_token_accuracy": 0.8305287137627602,
      "num_tokens": 610167888.0,
      "step": 19117
    },
    {
      "entropy": 0.5825919383205473,
      "epoch": 1.758960558018643,
      "grad_norm": 0.15810543298721313,
      "learning_rate": 4.1371484650535166e-05,
      "loss": 0.5812,
      "mean_token_accuracy": 0.8220688849687576,
      "num_tokens": 610200272.0,
      "step": 19118
    },
    {
      "entropy": 0.576705064624548,
      "epoch": 1.759052564390084,
      "grad_norm": 0.15357929468154907,
      "learning_rate": 4.1368417824393534e-05,
      "loss": 0.5689,
      "mean_token_accuracy": 0.8217475935816765,
      "num_tokens": 610232623.0,
      "step": 19119
    },
    {
      "entropy": 0.5791017096489668,
      "epoch": 1.7591445707615252,
      "grad_norm": 0.14828351140022278,
      "learning_rate": 4.136535099825191e-05,
      "loss": 0.566,
      "mean_token_accuracy": 0.8279194720089436,
      "num_tokens": 610265104.0,
      "step": 19120
    },
    {
      "entropy": 0.6044644778594375,
      "epoch": 1.7592365771329663,
      "grad_norm": 0.15613466501235962,
      "learning_rate": 4.136228417211028e-05,
      "loss": 0.5882,
      "mean_token_accuracy": 0.8198056295514107,
      "num_tokens": 610297100.0,
      "step": 19121
    },
    {
      "entropy": 0.721256572753191,
      "epoch": 1.7593285835044077,
      "grad_norm": 0.16369281709194183,
      "learning_rate": 4.135921734596866e-05,
      "loss": 0.7041,
      "mean_token_accuracy": 0.7830605767667294,
      "num_tokens": 610328490.0,
      "step": 19122
    },
    {
      "entropy": 0.6548959668725729,
      "epoch": 1.759420589875849,
      "grad_norm": 0.15565697848796844,
      "learning_rate": 4.135615051982703e-05,
      "loss": 0.6532,
      "mean_token_accuracy": 0.8032598085701466,
      "num_tokens": 610360439.0,
      "step": 19123
    },
    {
      "entropy": 0.6011337926611304,
      "epoch": 1.7595125962472902,
      "grad_norm": 0.1546880304813385,
      "learning_rate": 4.135308369368541e-05,
      "loss": 0.588,
      "mean_token_accuracy": 0.8158554844558239,
      "num_tokens": 610391966.0,
      "step": 19124
    },
    {
      "entropy": 0.5406528036110103,
      "epoch": 1.7596046026187313,
      "grad_norm": 0.14561140537261963,
      "learning_rate": 4.135001686754378e-05,
      "loss": 0.524,
      "mean_token_accuracy": 0.8430426307022572,
      "num_tokens": 610424446.0,
      "step": 19125
    },
    {
      "entropy": 0.6071658153086901,
      "epoch": 1.7596966089901724,
      "grad_norm": 0.14887595176696777,
      "learning_rate": 4.1346950041402156e-05,
      "loss": 0.596,
      "mean_token_accuracy": 0.8194962181150913,
      "num_tokens": 610456299.0,
      "step": 19126
    },
    {
      "entropy": 0.5380248948931694,
      "epoch": 1.7597886153616138,
      "grad_norm": 0.14551471173763275,
      "learning_rate": 4.134388321526053e-05,
      "loss": 0.5253,
      "mean_token_accuracy": 0.8357577584683895,
      "num_tokens": 610488094.0,
      "step": 19127
    },
    {
      "entropy": 0.6367363166064024,
      "epoch": 1.7598806217330552,
      "grad_norm": 0.15298675000667572,
      "learning_rate": 4.1340816389118906e-05,
      "loss": 0.6232,
      "mean_token_accuracy": 0.8093093410134315,
      "num_tokens": 610520633.0,
      "step": 19128
    },
    {
      "entropy": 0.5690097929909825,
      "epoch": 1.7599726281044963,
      "grad_norm": 0.15206223726272583,
      "learning_rate": 4.133774956297728e-05,
      "loss": 0.5561,
      "mean_token_accuracy": 0.8286900036036968,
      "num_tokens": 610552991.0,
      "step": 19129
    },
    {
      "entropy": 0.7178013678640127,
      "epoch": 1.7600646344759374,
      "grad_norm": 0.16528339684009552,
      "learning_rate": 4.133468273683565e-05,
      "loss": 0.6994,
      "mean_token_accuracy": 0.7880042158067226,
      "num_tokens": 610584376.0,
      "step": 19130
    },
    {
      "entropy": 0.650084376335144,
      "epoch": 1.7601566408473786,
      "grad_norm": 0.1579996645450592,
      "learning_rate": 4.133161591069403e-05,
      "loss": 0.6522,
      "mean_token_accuracy": 0.8028866574168205,
      "num_tokens": 610616704.0,
      "step": 19131
    },
    {
      "entropy": 0.6379134124144912,
      "epoch": 1.76024864721882,
      "grad_norm": 0.15945346653461456,
      "learning_rate": 4.13285490845524e-05,
      "loss": 0.618,
      "mean_token_accuracy": 0.8110391348600388,
      "num_tokens": 610648132.0,
      "step": 19132
    },
    {
      "entropy": 0.537617146037519,
      "epoch": 1.7603406535902613,
      "grad_norm": 0.15009084343910217,
      "learning_rate": 4.132548225841077e-05,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8418445140123367,
      "num_tokens": 610680182.0,
      "step": 19133
    },
    {
      "entropy": 0.61588220205158,
      "epoch": 1.7604326599617024,
      "grad_norm": 0.1523984670639038,
      "learning_rate": 4.132241543226915e-05,
      "loss": 0.6083,
      "mean_token_accuracy": 0.8104044198989868,
      "num_tokens": 610712278.0,
      "step": 19134
    },
    {
      "entropy": 0.5208516214042902,
      "epoch": 1.7605246663331435,
      "grad_norm": 0.14899803698062897,
      "learning_rate": 4.131934860612752e-05,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8427783250808716,
      "num_tokens": 610743617.0,
      "step": 19135
    },
    {
      "entropy": 0.6813563741743565,
      "epoch": 1.7606166727045847,
      "grad_norm": 0.16153071820735931,
      "learning_rate": 4.1316281779985896e-05,
      "loss": 0.6714,
      "mean_token_accuracy": 0.796801570802927,
      "num_tokens": 610775773.0,
      "step": 19136
    },
    {
      "entropy": 0.5313835171982646,
      "epoch": 1.760708679076026,
      "grad_norm": 0.15499576926231384,
      "learning_rate": 4.131321495384427e-05,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8403392471373081,
      "num_tokens": 610807502.0,
      "step": 19137
    },
    {
      "entropy": 0.62526033539325,
      "epoch": 1.7608006854474674,
      "grad_norm": 0.15747323632240295,
      "learning_rate": 4.131014812770264e-05,
      "loss": 0.6187,
      "mean_token_accuracy": 0.8074978925287724,
      "num_tokens": 610839499.0,
      "step": 19138
    },
    {
      "entropy": 0.5931365694850683,
      "epoch": 1.7608926918189085,
      "grad_norm": 0.1554892659187317,
      "learning_rate": 4.130708130156102e-05,
      "loss": 0.5918,
      "mean_token_accuracy": 0.8209166415035725,
      "num_tokens": 610871674.0,
      "step": 19139
    },
    {
      "entropy": 0.5839841030538082,
      "epoch": 1.7609846981903496,
      "grad_norm": 0.15105965733528137,
      "learning_rate": 4.130401447541939e-05,
      "loss": 0.5692,
      "mean_token_accuracy": 0.8211916871368885,
      "num_tokens": 610903808.0,
      "step": 19140
    },
    {
      "entropy": 0.6160612776875496,
      "epoch": 1.7610767045617908,
      "grad_norm": 0.1596289575099945,
      "learning_rate": 4.130094764927777e-05,
      "loss": 0.61,
      "mean_token_accuracy": 0.8175734356045723,
      "num_tokens": 610935056.0,
      "step": 19141
    },
    {
      "entropy": 0.6466558650135994,
      "epoch": 1.7611687109332321,
      "grad_norm": 0.15791499614715576,
      "learning_rate": 4.129788082313614e-05,
      "loss": 0.6324,
      "mean_token_accuracy": 0.8074085041880608,
      "num_tokens": 610967450.0,
      "step": 19142
    },
    {
      "entropy": 0.6611157041043043,
      "epoch": 1.7612607173046735,
      "grad_norm": 0.16155603528022766,
      "learning_rate": 4.129481399699451e-05,
      "loss": 0.6611,
      "mean_token_accuracy": 0.7990453168749809,
      "num_tokens": 611000074.0,
      "step": 19143
    },
    {
      "entropy": 0.5196818104013801,
      "epoch": 1.7613527236761146,
      "grad_norm": 0.1498953253030777,
      "learning_rate": 4.1291747170852886e-05,
      "loss": 0.5069,
      "mean_token_accuracy": 0.84070635586977,
      "num_tokens": 611032098.0,
      "step": 19144
    },
    {
      "entropy": 0.6706633046269417,
      "epoch": 1.7614447300475558,
      "grad_norm": 0.16466142237186432,
      "learning_rate": 4.128868034471126e-05,
      "loss": 0.6697,
      "mean_token_accuracy": 0.8002796247601509,
      "num_tokens": 611064263.0,
      "step": 19145
    },
    {
      "entropy": 0.5621722596697509,
      "epoch": 1.7615367364189969,
      "grad_norm": 0.14897605776786804,
      "learning_rate": 4.1285613518569636e-05,
      "loss": 0.5587,
      "mean_token_accuracy": 0.8312661498785019,
      "num_tokens": 611096690.0,
      "step": 19146
    },
    {
      "entropy": 0.6426506259012967,
      "epoch": 1.7616287427904382,
      "grad_norm": 0.1611117571592331,
      "learning_rate": 4.128254669242801e-05,
      "loss": 0.6391,
      "mean_token_accuracy": 0.8061103820800781,
      "num_tokens": 611127730.0,
      "step": 19147
    },
    {
      "entropy": 0.5806536916643381,
      "epoch": 1.7617207491618796,
      "grad_norm": 0.15119972825050354,
      "learning_rate": 4.127947986628638e-05,
      "loss": 0.5796,
      "mean_token_accuracy": 0.8237042687833309,
      "num_tokens": 611159595.0,
      "step": 19148
    },
    {
      "entropy": 0.591741381213069,
      "epoch": 1.7618127555333207,
      "grad_norm": 0.1577240526676178,
      "learning_rate": 4.127641304014476e-05,
      "loss": 0.5716,
      "mean_token_accuracy": 0.8227860853075981,
      "num_tokens": 611190609.0,
      "step": 19149
    },
    {
      "entropy": 0.6027887407690287,
      "epoch": 1.7619047619047619,
      "grad_norm": 0.15385758876800537,
      "learning_rate": 4.127334621400313e-05,
      "loss": 0.6027,
      "mean_token_accuracy": 0.8175178840756416,
      "num_tokens": 611222797.0,
      "step": 19150
    },
    {
      "entropy": 0.6025541350245476,
      "epoch": 1.761996768276203,
      "grad_norm": 0.15234768390655518,
      "learning_rate": 4.12702793878615e-05,
      "loss": 0.5991,
      "mean_token_accuracy": 0.8174135722219944,
      "num_tokens": 611255132.0,
      "step": 19151
    },
    {
      "entropy": 0.5911812502890825,
      "epoch": 1.7620887746476444,
      "grad_norm": 0.15163956582546234,
      "learning_rate": 4.126721256171988e-05,
      "loss": 0.5831,
      "mean_token_accuracy": 0.8194975256919861,
      "num_tokens": 611286648.0,
      "step": 19152
    },
    {
      "entropy": 0.5764160882681608,
      "epoch": 1.7621807810190857,
      "grad_norm": 0.151494562625885,
      "learning_rate": 4.126414573557825e-05,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8279771506786346,
      "num_tokens": 611317995.0,
      "step": 19153
    },
    {
      "entropy": 0.6929999887943268,
      "epoch": 1.7622727873905268,
      "grad_norm": 0.1633455753326416,
      "learning_rate": 4.1261078909436626e-05,
      "loss": 0.6983,
      "mean_token_accuracy": 0.7877116836607456,
      "num_tokens": 611350457.0,
      "step": 19154
    },
    {
      "entropy": 0.5559319402091205,
      "epoch": 1.762364793761968,
      "grad_norm": 0.15277917683124542,
      "learning_rate": 4.1258012083295e-05,
      "loss": 0.5456,
      "mean_token_accuracy": 0.8307423740625381,
      "num_tokens": 611382833.0,
      "step": 19155
    },
    {
      "entropy": 0.5561770722270012,
      "epoch": 1.762456800133409,
      "grad_norm": 0.1462421864271164,
      "learning_rate": 4.125494525715337e-05,
      "loss": 0.5423,
      "mean_token_accuracy": 0.8337363414466381,
      "num_tokens": 611415043.0,
      "step": 19156
    },
    {
      "entropy": 0.5710400706157088,
      "epoch": 1.7625488065048505,
      "grad_norm": 0.15664449334144592,
      "learning_rate": 4.125187843101175e-05,
      "loss": 0.5699,
      "mean_token_accuracy": 0.8238530643284321,
      "num_tokens": 611446898.0,
      "step": 19157
    },
    {
      "entropy": 0.5984462536871433,
      "epoch": 1.7626408128762918,
      "grad_norm": 0.1554669290781021,
      "learning_rate": 4.1248811604870125e-05,
      "loss": 0.5716,
      "mean_token_accuracy": 0.8199938870966434,
      "num_tokens": 611478302.0,
      "step": 19158
    },
    {
      "entropy": 0.5083274962380528,
      "epoch": 1.762732819247733,
      "grad_norm": 0.14244532585144043,
      "learning_rate": 4.124574477872849e-05,
      "loss": 0.4999,
      "mean_token_accuracy": 0.8436707444489002,
      "num_tokens": 611509951.0,
      "step": 19159
    },
    {
      "entropy": 0.6781518496572971,
      "epoch": 1.762824825619174,
      "grad_norm": 0.15557977557182312,
      "learning_rate": 4.1242677952586874e-05,
      "loss": 0.6607,
      "mean_token_accuracy": 0.8021545708179474,
      "num_tokens": 611542128.0,
      "step": 19160
    },
    {
      "entropy": 0.48110597394406796,
      "epoch": 1.7629168319906152,
      "grad_norm": 0.139712855219841,
      "learning_rate": 4.123961112644524e-05,
      "loss": 0.46,
      "mean_token_accuracy": 0.8531520627439022,
      "num_tokens": 611573975.0,
      "step": 19161
    },
    {
      "entropy": 0.5327544333413243,
      "epoch": 1.7630088383620566,
      "grad_norm": 0.16122864186763763,
      "learning_rate": 4.123654430030362e-05,
      "loss": 0.5189,
      "mean_token_accuracy": 0.8387136571109295,
      "num_tokens": 611606434.0,
      "step": 19162
    },
    {
      "entropy": 0.5933559006080031,
      "epoch": 1.763100844733498,
      "grad_norm": 0.155628964304924,
      "learning_rate": 4.123347747416199e-05,
      "loss": 0.5904,
      "mean_token_accuracy": 0.8183843456208706,
      "num_tokens": 611638242.0,
      "step": 19163
    },
    {
      "entropy": 0.5283039603382349,
      "epoch": 1.763192851104939,
      "grad_norm": 0.14898823201656342,
      "learning_rate": 4.1230410648020366e-05,
      "loss": 0.518,
      "mean_token_accuracy": 0.8410684317350388,
      "num_tokens": 611670062.0,
      "step": 19164
    },
    {
      "entropy": 0.7304014395922422,
      "epoch": 1.7632848574763802,
      "grad_norm": 0.16979648172855377,
      "learning_rate": 4.122734382187874e-05,
      "loss": 0.7243,
      "mean_token_accuracy": 0.7828702330589294,
      "num_tokens": 611701719.0,
      "step": 19165
    },
    {
      "entropy": 0.604650741443038,
      "epoch": 1.7633768638478213,
      "grad_norm": 0.15274475514888763,
      "learning_rate": 4.1224276995737115e-05,
      "loss": 0.5955,
      "mean_token_accuracy": 0.817396555095911,
      "num_tokens": 611733632.0,
      "step": 19166
    },
    {
      "entropy": 0.6098520741797984,
      "epoch": 1.7634688702192627,
      "grad_norm": 0.15884236991405487,
      "learning_rate": 4.122121016959549e-05,
      "loss": 0.5995,
      "mean_token_accuracy": 0.8175870291888714,
      "num_tokens": 611765329.0,
      "step": 19167
    },
    {
      "entropy": 0.6251417957246304,
      "epoch": 1.763560876590704,
      "grad_norm": 0.15510234236717224,
      "learning_rate": 4.1218143343453864e-05,
      "loss": 0.6284,
      "mean_token_accuracy": 0.808320801705122,
      "num_tokens": 611797748.0,
      "step": 19168
    },
    {
      "entropy": 0.7005877625197172,
      "epoch": 1.7636528829621452,
      "grad_norm": 0.16209222376346588,
      "learning_rate": 4.121507651731223e-05,
      "loss": 0.6989,
      "mean_token_accuracy": 0.7895579002797604,
      "num_tokens": 611829433.0,
      "step": 19169
    },
    {
      "entropy": 0.6205676356330514,
      "epoch": 1.7637448893335863,
      "grad_norm": 0.14991696178913116,
      "learning_rate": 4.1212009691170614e-05,
      "loss": 0.6122,
      "mean_token_accuracy": 0.8132313303649426,
      "num_tokens": 611861937.0,
      "step": 19170
    },
    {
      "entropy": 0.6652755523100495,
      "epoch": 1.7638368957050274,
      "grad_norm": 0.16145539283752441,
      "learning_rate": 4.120894286502898e-05,
      "loss": 0.6542,
      "mean_token_accuracy": 0.7996858283877373,
      "num_tokens": 611892973.0,
      "step": 19171
    },
    {
      "entropy": 0.5394933559000492,
      "epoch": 1.7639289020764688,
      "grad_norm": 0.1472652107477188,
      "learning_rate": 4.1205876038887356e-05,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8372183330357075,
      "num_tokens": 611924814.0,
      "step": 19172
    },
    {
      "entropy": 0.6549736689776182,
      "epoch": 1.7640209084479102,
      "grad_norm": 0.1589881181716919,
      "learning_rate": 4.120280921274573e-05,
      "loss": 0.6367,
      "mean_token_accuracy": 0.8052956163883209,
      "num_tokens": 611957531.0,
      "step": 19173
    },
    {
      "entropy": 0.6556270066648722,
      "epoch": 1.7641129148193513,
      "grad_norm": 0.15771381556987762,
      "learning_rate": 4.1199742386604105e-05,
      "loss": 0.6518,
      "mean_token_accuracy": 0.7993382550776005,
      "num_tokens": 611988917.0,
      "step": 19174
    },
    {
      "entropy": 0.6275620926171541,
      "epoch": 1.7642049211907924,
      "grad_norm": 0.15537023544311523,
      "learning_rate": 4.119667556046248e-05,
      "loss": 0.6241,
      "mean_token_accuracy": 0.8095750324428082,
      "num_tokens": 612020796.0,
      "step": 19175
    },
    {
      "entropy": 0.6944689210504293,
      "epoch": 1.7642969275622336,
      "grad_norm": 0.15271122753620148,
      "learning_rate": 4.1193608734320855e-05,
      "loss": 0.6717,
      "mean_token_accuracy": 0.7928572632372379,
      "num_tokens": 612052590.0,
      "step": 19176
    },
    {
      "entropy": 0.6964299697428942,
      "epoch": 1.764388933933675,
      "grad_norm": 0.16089005768299103,
      "learning_rate": 4.119054190817922e-05,
      "loss": 0.6854,
      "mean_token_accuracy": 0.7954432144761086,
      "num_tokens": 612084446.0,
      "step": 19177
    },
    {
      "entropy": 0.5587370633147657,
      "epoch": 1.7644809403051163,
      "grad_norm": 0.14824846386909485,
      "learning_rate": 4.1187475082037604e-05,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8334801718592644,
      "num_tokens": 612116426.0,
      "step": 19178
    },
    {
      "entropy": 0.6977652478963137,
      "epoch": 1.7645729466765574,
      "grad_norm": 0.16406656801700592,
      "learning_rate": 4.118440825589597e-05,
      "loss": 0.6846,
      "mean_token_accuracy": 0.7915931046009064,
      "num_tokens": 612147913.0,
      "step": 19179
    },
    {
      "entropy": 0.6451562200672925,
      "epoch": 1.7646649530479985,
      "grad_norm": 0.15872453153133392,
      "learning_rate": 4.1181341429754347e-05,
      "loss": 0.6288,
      "mean_token_accuracy": 0.8062588609755039,
      "num_tokens": 612179839.0,
      "step": 19180
    },
    {
      "entropy": 0.6426993263885379,
      "epoch": 1.7647569594194397,
      "grad_norm": 0.15637674927711487,
      "learning_rate": 4.117827460361272e-05,
      "loss": 0.6378,
      "mean_token_accuracy": 0.808086697012186,
      "num_tokens": 612212264.0,
      "step": 19181
    },
    {
      "entropy": 0.5319612640887499,
      "epoch": 1.764848965790881,
      "grad_norm": 0.1424800604581833,
      "learning_rate": 4.1175207777471096e-05,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8377645537257195,
      "num_tokens": 612244681.0,
      "step": 19182
    },
    {
      "entropy": 0.5440970230847597,
      "epoch": 1.7649409721623224,
      "grad_norm": 0.15067867934703827,
      "learning_rate": 4.117214095132947e-05,
      "loss": 0.5479,
      "mean_token_accuracy": 0.8287228532135487,
      "num_tokens": 612277404.0,
      "step": 19183
    },
    {
      "entropy": 0.5893942681141198,
      "epoch": 1.7650329785337635,
      "grad_norm": 0.1621224582195282,
      "learning_rate": 4.1169074125187845e-05,
      "loss": 0.5894,
      "mean_token_accuracy": 0.8239320516586304,
      "num_tokens": 612309607.0,
      "step": 19184
    },
    {
      "entropy": 0.6771736536175013,
      "epoch": 1.7651249849052046,
      "grad_norm": 0.16178399324417114,
      "learning_rate": 4.116600729904622e-05,
      "loss": 0.6662,
      "mean_token_accuracy": 0.7966158427298069,
      "num_tokens": 612341298.0,
      "step": 19185
    },
    {
      "entropy": 0.5988347232341766,
      "epoch": 1.7652169912766458,
      "grad_norm": 0.16297326982021332,
      "learning_rate": 4.1162940472904594e-05,
      "loss": 0.5813,
      "mean_token_accuracy": 0.8219203278422356,
      "num_tokens": 612372701.0,
      "step": 19186
    },
    {
      "entropy": 0.5650125369429588,
      "epoch": 1.7653089976480871,
      "grad_norm": 0.15358398854732513,
      "learning_rate": 4.115987364676297e-05,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8312968090176582,
      "num_tokens": 612403928.0,
      "step": 19187
    },
    {
      "entropy": 0.58233388280496,
      "epoch": 1.7654010040195285,
      "grad_norm": 0.1493379771709442,
      "learning_rate": 4.1156806820621344e-05,
      "loss": 0.5747,
      "mean_token_accuracy": 0.823033019900322,
      "num_tokens": 612436290.0,
      "step": 19188
    },
    {
      "entropy": 0.6060700691305101,
      "epoch": 1.7654930103909696,
      "grad_norm": 0.15038272738456726,
      "learning_rate": 4.115373999447972e-05,
      "loss": 0.5936,
      "mean_token_accuracy": 0.8185318410396576,
      "num_tokens": 612467971.0,
      "step": 19189
    },
    {
      "entropy": 0.584547583013773,
      "epoch": 1.7655850167624108,
      "grad_norm": 0.15051482617855072,
      "learning_rate": 4.1150673168338086e-05,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8226565420627594,
      "num_tokens": 612500011.0,
      "step": 19190
    },
    {
      "entropy": 0.49093507416546345,
      "epoch": 1.7656770231338519,
      "grad_norm": 0.14031092822551727,
      "learning_rate": 4.114760634219647e-05,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8537493459880352,
      "num_tokens": 612532254.0,
      "step": 19191
    },
    {
      "entropy": 0.5708145136013627,
      "epoch": 1.7657690295052932,
      "grad_norm": 0.15831811726093292,
      "learning_rate": 4.1144539516054836e-05,
      "loss": 0.5617,
      "mean_token_accuracy": 0.8269103057682514,
      "num_tokens": 612564535.0,
      "step": 19192
    },
    {
      "entropy": 0.6815436836332083,
      "epoch": 1.7658610358767346,
      "grad_norm": 0.1636464148759842,
      "learning_rate": 4.114147268991321e-05,
      "loss": 0.6864,
      "mean_token_accuracy": 0.7955589704215527,
      "num_tokens": 612596635.0,
      "step": 19193
    },
    {
      "entropy": 0.6643506921827793,
      "epoch": 1.7659530422481757,
      "grad_norm": 0.16068604588508606,
      "learning_rate": 4.1138405863771585e-05,
      "loss": 0.6512,
      "mean_token_accuracy": 0.7999295145273209,
      "num_tokens": 612628661.0,
      "step": 19194
    },
    {
      "entropy": 0.597972322255373,
      "epoch": 1.7660450486196169,
      "grad_norm": 0.15599748492240906,
      "learning_rate": 4.113533903762996e-05,
      "loss": 0.5886,
      "mean_token_accuracy": 0.8145229443907738,
      "num_tokens": 612660618.0,
      "step": 19195
    },
    {
      "entropy": 0.7041512168943882,
      "epoch": 1.766137054991058,
      "grad_norm": 0.16372215747833252,
      "learning_rate": 4.1132272211488334e-05,
      "loss": 0.6944,
      "mean_token_accuracy": 0.786935456097126,
      "num_tokens": 612691357.0,
      "step": 19196
    },
    {
      "entropy": 0.6569698192179203,
      "epoch": 1.7662290613624994,
      "grad_norm": 0.15615881979465485,
      "learning_rate": 4.112920538534671e-05,
      "loss": 0.645,
      "mean_token_accuracy": 0.8027772307395935,
      "num_tokens": 612723057.0,
      "step": 19197
    },
    {
      "entropy": 0.6653582733124495,
      "epoch": 1.7663210677339407,
      "grad_norm": 0.1547240912914276,
      "learning_rate": 4.112613855920508e-05,
      "loss": 0.6474,
      "mean_token_accuracy": 0.7987943142652512,
      "num_tokens": 612754944.0,
      "step": 19198
    },
    {
      "entropy": 0.6584567204117775,
      "epoch": 1.7664130741053818,
      "grad_norm": 0.15373334288597107,
      "learning_rate": 4.112307173306346e-05,
      "loss": 0.6436,
      "mean_token_accuracy": 0.8049040734767914,
      "num_tokens": 612787240.0,
      "step": 19199
    },
    {
      "entropy": 0.5772416233085096,
      "epoch": 1.766505080476823,
      "grad_norm": 0.15283511579036713,
      "learning_rate": 4.1120004906921826e-05,
      "loss": 0.5713,
      "mean_token_accuracy": 0.8257986009120941,
      "num_tokens": 612818988.0,
      "step": 19200
    },
    {
      "entropy": 0.7396116517484188,
      "epoch": 1.766597086848264,
      "grad_norm": 0.16587196290493011,
      "learning_rate": 4.111693808078021e-05,
      "loss": 0.7277,
      "mean_token_accuracy": 0.7741243503987789,
      "num_tokens": 612849846.0,
      "step": 19201
    },
    {
      "entropy": 0.5312547739595175,
      "epoch": 1.7666890932197055,
      "grad_norm": 0.15243522822856903,
      "learning_rate": 4.1113871254638575e-05,
      "loss": 0.5179,
      "mean_token_accuracy": 0.8395699672400951,
      "num_tokens": 612881759.0,
      "step": 19202
    },
    {
      "entropy": 0.4868663512170315,
      "epoch": 1.7667810995911468,
      "grad_norm": 0.13997788727283478,
      "learning_rate": 4.111080442849695e-05,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8539500683546066,
      "num_tokens": 612913415.0,
      "step": 19203
    },
    {
      "entropy": 0.6038715168833733,
      "epoch": 1.766873105962588,
      "grad_norm": 0.15790677070617676,
      "learning_rate": 4.1107737602355325e-05,
      "loss": 0.5975,
      "mean_token_accuracy": 0.8130933865904808,
      "num_tokens": 612944945.0,
      "step": 19204
    },
    {
      "entropy": 0.5236852336674929,
      "epoch": 1.766965112334029,
      "grad_norm": 0.15390534698963165,
      "learning_rate": 4.11046707762137e-05,
      "loss": 0.5082,
      "mean_token_accuracy": 0.8422041162848473,
      "num_tokens": 612976537.0,
      "step": 19205
    },
    {
      "entropy": 0.6147444120142609,
      "epoch": 1.7670571187054702,
      "grad_norm": 0.15584279596805573,
      "learning_rate": 4.1101603950072074e-05,
      "loss": 0.5962,
      "mean_token_accuracy": 0.8153346106410027,
      "num_tokens": 613008252.0,
      "step": 19206
    },
    {
      "entropy": 0.5830676017794758,
      "epoch": 1.7671491250769116,
      "grad_norm": 0.15242095291614532,
      "learning_rate": 4.109853712393045e-05,
      "loss": 0.5798,
      "mean_token_accuracy": 0.8180710226297379,
      "num_tokens": 613040012.0,
      "step": 19207
    },
    {
      "entropy": 0.6660137120634317,
      "epoch": 1.767241131448353,
      "grad_norm": 0.1616414487361908,
      "learning_rate": 4.1095470297788816e-05,
      "loss": 0.6556,
      "mean_token_accuracy": 0.80381665751338,
      "num_tokens": 613071908.0,
      "step": 19208
    },
    {
      "entropy": 0.647943839430809,
      "epoch": 1.767333137819794,
      "grad_norm": 0.15778318047523499,
      "learning_rate": 4.10924034716472e-05,
      "loss": 0.6326,
      "mean_token_accuracy": 0.8028069585561752,
      "num_tokens": 613102781.0,
      "step": 19209
    },
    {
      "entropy": 0.6613576505333185,
      "epoch": 1.7674251441912352,
      "grad_norm": 0.16009946167469025,
      "learning_rate": 4.1089336645505566e-05,
      "loss": 0.6391,
      "mean_token_accuracy": 0.801283709704876,
      "num_tokens": 613134445.0,
      "step": 19210
    },
    {
      "entropy": 0.6342017762362957,
      "epoch": 1.7675171505626763,
      "grad_norm": 0.15748904645442963,
      "learning_rate": 4.108626981936394e-05,
      "loss": 0.6295,
      "mean_token_accuracy": 0.8056006841361523,
      "num_tokens": 613166363.0,
      "step": 19211
    },
    {
      "entropy": 0.7319600200280547,
      "epoch": 1.7676091569341177,
      "grad_norm": 0.1599220484495163,
      "learning_rate": 4.1083202993222315e-05,
      "loss": 0.7105,
      "mean_token_accuracy": 0.7827688157558441,
      "num_tokens": 613198819.0,
      "step": 19212
    },
    {
      "entropy": 0.6649634018540382,
      "epoch": 1.767701163305559,
      "grad_norm": 0.16222696006298065,
      "learning_rate": 4.108013616708069e-05,
      "loss": 0.6606,
      "mean_token_accuracy": 0.7985853254795074,
      "num_tokens": 613230261.0,
      "step": 19213
    },
    {
      "entropy": 0.6927696652710438,
      "epoch": 1.7677931696770002,
      "grad_norm": 0.15994097292423248,
      "learning_rate": 4.1077069340939064e-05,
      "loss": 0.7,
      "mean_token_accuracy": 0.7913739122450352,
      "num_tokens": 613262226.0,
      "step": 19214
    },
    {
      "entropy": 0.6640940671786666,
      "epoch": 1.7678851760484413,
      "grad_norm": 0.15710189938545227,
      "learning_rate": 4.107400251479744e-05,
      "loss": 0.6612,
      "mean_token_accuracy": 0.8003752678632736,
      "num_tokens": 613294662.0,
      "step": 19215
    },
    {
      "entropy": 0.624059647321701,
      "epoch": 1.7679771824198824,
      "grad_norm": 0.16316917538642883,
      "learning_rate": 4.107093568865581e-05,
      "loss": 0.6204,
      "mean_token_accuracy": 0.8065422773361206,
      "num_tokens": 613326624.0,
      "step": 19216
    },
    {
      "entropy": 0.4977406756952405,
      "epoch": 1.7680691887913238,
      "grad_norm": 0.15112490952014923,
      "learning_rate": 4.106786886251419e-05,
      "loss": 0.4771,
      "mean_token_accuracy": 0.8521203212440014,
      "num_tokens": 613358756.0,
      "step": 19217
    },
    {
      "entropy": 0.5051762899383903,
      "epoch": 1.7681611951627652,
      "grad_norm": 0.14859512448310852,
      "learning_rate": 4.106480203637256e-05,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8457594811916351,
      "num_tokens": 613391265.0,
      "step": 19218
    },
    {
      "entropy": 0.5600707130506635,
      "epoch": 1.7682532015342063,
      "grad_norm": 0.14491383731365204,
      "learning_rate": 4.106173521023093e-05,
      "loss": 0.5493,
      "mean_token_accuracy": 0.834351684898138,
      "num_tokens": 613423720.0,
      "step": 19219
    },
    {
      "entropy": 0.6166912317276001,
      "epoch": 1.7683452079056474,
      "grad_norm": 0.15454013645648956,
      "learning_rate": 4.105866838408931e-05,
      "loss": 0.5981,
      "mean_token_accuracy": 0.8165423311293125,
      "num_tokens": 613455496.0,
      "step": 19220
    },
    {
      "entropy": 0.516917790286243,
      "epoch": 1.7684372142770886,
      "grad_norm": 0.1455754190683365,
      "learning_rate": 4.105560155794768e-05,
      "loss": 0.4941,
      "mean_token_accuracy": 0.848742064088583,
      "num_tokens": 613488264.0,
      "step": 19221
    },
    {
      "entropy": 0.5617925338447094,
      "epoch": 1.76852922064853,
      "grad_norm": 0.14800000190734863,
      "learning_rate": 4.105253473180606e-05,
      "loss": 0.554,
      "mean_token_accuracy": 0.8294927999377251,
      "num_tokens": 613520521.0,
      "step": 19222
    },
    {
      "entropy": 0.7467650659382343,
      "epoch": 1.7686212270199713,
      "grad_norm": 0.16128678619861603,
      "learning_rate": 4.104946790566443e-05,
      "loss": 0.7356,
      "mean_token_accuracy": 0.7772736214101315,
      "num_tokens": 613552825.0,
      "step": 19223
    },
    {
      "entropy": 0.5852450388483703,
      "epoch": 1.7687132333914124,
      "grad_norm": 0.15685437619686127,
      "learning_rate": 4.1046401079522804e-05,
      "loss": 0.571,
      "mean_token_accuracy": 0.8226759620010853,
      "num_tokens": 613584682.0,
      "step": 19224
    },
    {
      "entropy": 0.5018087504431605,
      "epoch": 1.7688052397628535,
      "grad_norm": 0.15158390998840332,
      "learning_rate": 4.104333425338118e-05,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8491310216486454,
      "num_tokens": 613617349.0,
      "step": 19225
    },
    {
      "entropy": 0.5127664878964424,
      "epoch": 1.7688972461342947,
      "grad_norm": 0.1463548243045807,
      "learning_rate": 4.104026742723955e-05,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8431496061384678,
      "num_tokens": 613649663.0,
      "step": 19226
    },
    {
      "entropy": 0.5489731756970286,
      "epoch": 1.768989252505736,
      "grad_norm": 0.14355657994747162,
      "learning_rate": 4.103720060109793e-05,
      "loss": 0.5449,
      "mean_token_accuracy": 0.8319962956011295,
      "num_tokens": 613681828.0,
      "step": 19227
    },
    {
      "entropy": 0.48932841094210744,
      "epoch": 1.7690812588771774,
      "grad_norm": 0.14068526029586792,
      "learning_rate": 4.10341337749563e-05,
      "loss": 0.4802,
      "mean_token_accuracy": 0.8530670627951622,
      "num_tokens": 613713843.0,
      "step": 19228
    },
    {
      "entropy": 0.5752556985244155,
      "epoch": 1.7691732652486185,
      "grad_norm": 0.15058720111846924,
      "learning_rate": 4.103106694881467e-05,
      "loss": 0.5593,
      "mean_token_accuracy": 0.8237934745848179,
      "num_tokens": 613745986.0,
      "step": 19229
    },
    {
      "entropy": 0.5746751371771097,
      "epoch": 1.7692652716200596,
      "grad_norm": 0.14790652692317963,
      "learning_rate": 4.102800012267305e-05,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8285114616155624,
      "num_tokens": 613778202.0,
      "step": 19230
    },
    {
      "entropy": 0.5853431662544608,
      "epoch": 1.7693572779915008,
      "grad_norm": 0.1517871618270874,
      "learning_rate": 4.102493329653142e-05,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8270026929676533,
      "num_tokens": 613810264.0,
      "step": 19231
    },
    {
      "entropy": 0.6373228766024113,
      "epoch": 1.7694492843629421,
      "grad_norm": 0.15460379421710968,
      "learning_rate": 4.1021866470389794e-05,
      "loss": 0.6217,
      "mean_token_accuracy": 0.8092655427753925,
      "num_tokens": 613841964.0,
      "step": 19232
    },
    {
      "entropy": 0.6522016916424036,
      "epoch": 1.7695412907343835,
      "grad_norm": 0.1603693664073944,
      "learning_rate": 4.101879964424817e-05,
      "loss": 0.64,
      "mean_token_accuracy": 0.8038547597825527,
      "num_tokens": 613873785.0,
      "step": 19233
    },
    {
      "entropy": 0.5505627859383821,
      "epoch": 1.7696332971058246,
      "grad_norm": 0.1435912698507309,
      "learning_rate": 4.1015732818106544e-05,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8352420218288898,
      "num_tokens": 613906256.0,
      "step": 19234
    },
    {
      "entropy": 0.5797744551673532,
      "epoch": 1.7697253034772658,
      "grad_norm": 0.14988000690937042,
      "learning_rate": 4.101266599196492e-05,
      "loss": 0.5716,
      "mean_token_accuracy": 0.8252931162714958,
      "num_tokens": 613939024.0,
      "step": 19235
    },
    {
      "entropy": 0.5812291651964188,
      "epoch": 1.7698173098487069,
      "grad_norm": 0.14814947545528412,
      "learning_rate": 4.100959916582329e-05,
      "loss": 0.5683,
      "mean_token_accuracy": 0.8275147676467896,
      "num_tokens": 613970332.0,
      "step": 19236
    },
    {
      "entropy": 0.5363809112459421,
      "epoch": 1.7699093162201482,
      "grad_norm": 0.14813244342803955,
      "learning_rate": 4.100653233968166e-05,
      "loss": 0.526,
      "mean_token_accuracy": 0.8355163224041462,
      "num_tokens": 614001764.0,
      "step": 19237
    },
    {
      "entropy": 0.5935838592704386,
      "epoch": 1.7700013225915896,
      "grad_norm": 0.1473783552646637,
      "learning_rate": 4.100346551354004e-05,
      "loss": 0.5821,
      "mean_token_accuracy": 0.8239375874400139,
      "num_tokens": 614033926.0,
      "step": 19238
    },
    {
      "entropy": 0.6300501395016909,
      "epoch": 1.7700933289630307,
      "grad_norm": 0.1585536152124405,
      "learning_rate": 4.100039868739841e-05,
      "loss": 0.6207,
      "mean_token_accuracy": 0.8087001033127308,
      "num_tokens": 614065725.0,
      "step": 19239
    },
    {
      "entropy": 0.6292230729013681,
      "epoch": 1.7701853353344719,
      "grad_norm": 0.15163928270339966,
      "learning_rate": 4.0997331861256785e-05,
      "loss": 0.6312,
      "mean_token_accuracy": 0.805448018014431,
      "num_tokens": 614097995.0,
      "step": 19240
    },
    {
      "entropy": 0.4761274876073003,
      "epoch": 1.770277341705913,
      "grad_norm": 0.15190772712230682,
      "learning_rate": 4.099426503511516e-05,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8489442989230156,
      "num_tokens": 614130255.0,
      "step": 19241
    },
    {
      "entropy": 0.596309594810009,
      "epoch": 1.7703693480773544,
      "grad_norm": 0.15453693270683289,
      "learning_rate": 4.0991198208973534e-05,
      "loss": 0.593,
      "mean_token_accuracy": 0.8171074539422989,
      "num_tokens": 614162275.0,
      "step": 19242
    },
    {
      "entropy": 0.6062624510377645,
      "epoch": 1.7704613544487957,
      "grad_norm": 0.15042123198509216,
      "learning_rate": 4.098813138283191e-05,
      "loss": 0.5973,
      "mean_token_accuracy": 0.8160581775009632,
      "num_tokens": 614194336.0,
      "step": 19243
    },
    {
      "entropy": 0.6170411026105285,
      "epoch": 1.7705533608202368,
      "grad_norm": 0.16306810081005096,
      "learning_rate": 4.098506455669028e-05,
      "loss": 0.6112,
      "mean_token_accuracy": 0.8114476650953293,
      "num_tokens": 614226131.0,
      "step": 19244
    },
    {
      "entropy": 0.6715835174545646,
      "epoch": 1.770645367191678,
      "grad_norm": 0.15403054654598236,
      "learning_rate": 4.098199773054866e-05,
      "loss": 0.6677,
      "mean_token_accuracy": 0.7957676947116852,
      "num_tokens": 614258484.0,
      "step": 19245
    },
    {
      "entropy": 0.5017677322030067,
      "epoch": 1.770737373563119,
      "grad_norm": 0.14014193415641785,
      "learning_rate": 4.097893090440703e-05,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8447137512266636,
      "num_tokens": 614290851.0,
      "step": 19246
    },
    {
      "entropy": 0.4246373677160591,
      "epoch": 1.7708293799345605,
      "grad_norm": 0.14434726536273956,
      "learning_rate": 4.09758640782654e-05,
      "loss": 0.3993,
      "mean_token_accuracy": 0.8747029043734074,
      "num_tokens": 614322371.0,
      "step": 19247
    },
    {
      "entropy": 0.5987805724143982,
      "epoch": 1.7709213863060018,
      "grad_norm": 0.1544717699289322,
      "learning_rate": 4.097279725212378e-05,
      "loss": 0.5768,
      "mean_token_accuracy": 0.8221237808465958,
      "num_tokens": 614354238.0,
      "step": 19248
    },
    {
      "entropy": 0.5826413612812757,
      "epoch": 1.771013392677443,
      "grad_norm": 0.15378357470035553,
      "learning_rate": 4.0969730425982157e-05,
      "loss": 0.5695,
      "mean_token_accuracy": 0.8256261087954044,
      "num_tokens": 614385917.0,
      "step": 19249
    },
    {
      "entropy": 0.6686650961637497,
      "epoch": 1.771105399048884,
      "grad_norm": 0.1550590842962265,
      "learning_rate": 4.0966663599840524e-05,
      "loss": 0.653,
      "mean_token_accuracy": 0.8013434410095215,
      "num_tokens": 614417508.0,
      "step": 19250
    },
    {
      "entropy": 0.6400720837991685,
      "epoch": 1.7711974054203252,
      "grad_norm": 0.15004406869411469,
      "learning_rate": 4.0963596773698906e-05,
      "loss": 0.6088,
      "mean_token_accuracy": 0.8123033046722412,
      "num_tokens": 614449529.0,
      "step": 19251
    },
    {
      "entropy": 0.6340414471924305,
      "epoch": 1.7712894117917666,
      "grad_norm": 0.15262524783611298,
      "learning_rate": 4.0960529947557274e-05,
      "loss": 0.6214,
      "mean_token_accuracy": 0.8110571391880512,
      "num_tokens": 614481984.0,
      "step": 19252
    },
    {
      "entropy": 0.5904355207458138,
      "epoch": 1.771381418163208,
      "grad_norm": 0.16100113093852997,
      "learning_rate": 4.095746312141565e-05,
      "loss": 0.5659,
      "mean_token_accuracy": 0.825705710798502,
      "num_tokens": 614513449.0,
      "step": 19253
    },
    {
      "entropy": 0.6967086745426059,
      "epoch": 1.771473424534649,
      "grad_norm": 0.15873633325099945,
      "learning_rate": 4.095439629527402e-05,
      "loss": 0.6905,
      "mean_token_accuracy": 0.785109456628561,
      "num_tokens": 614545420.0,
      "step": 19254
    },
    {
      "entropy": 0.6226256396621466,
      "epoch": 1.7715654309060902,
      "grad_norm": 0.15477943420410156,
      "learning_rate": 4.09513294691324e-05,
      "loss": 0.6,
      "mean_token_accuracy": 0.8157532624900341,
      "num_tokens": 614577819.0,
      "step": 19255
    },
    {
      "entropy": 0.5425547789782286,
      "epoch": 1.7716574372775313,
      "grad_norm": 0.14749523997306824,
      "learning_rate": 4.094826264299077e-05,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8344622328877449,
      "num_tokens": 614609444.0,
      "step": 19256
    },
    {
      "entropy": 0.710570496506989,
      "epoch": 1.7717494436489727,
      "grad_norm": 0.16250912845134735,
      "learning_rate": 4.094519581684915e-05,
      "loss": 0.7042,
      "mean_token_accuracy": 0.7885507829487324,
      "num_tokens": 614640826.0,
      "step": 19257
    },
    {
      "entropy": 0.633452967274934,
      "epoch": 1.771841450020414,
      "grad_norm": 0.1599225401878357,
      "learning_rate": 4.0942128990707515e-05,
      "loss": 0.6224,
      "mean_token_accuracy": 0.8084648102521896,
      "num_tokens": 614672846.0,
      "step": 19258
    },
    {
      "entropy": 0.6455473778769374,
      "epoch": 1.7719334563918552,
      "grad_norm": 0.16361160576343536,
      "learning_rate": 4.0939062164565896e-05,
      "loss": 0.646,
      "mean_token_accuracy": 0.8020756095647812,
      "num_tokens": 614703929.0,
      "step": 19259
    },
    {
      "entropy": 0.6825797092169523,
      "epoch": 1.7720254627632963,
      "grad_norm": 0.15818513929843903,
      "learning_rate": 4.0935995338424264e-05,
      "loss": 0.6783,
      "mean_token_accuracy": 0.7932658270001411,
      "num_tokens": 614736058.0,
      "step": 19260
    },
    {
      "entropy": 0.7082426119595766,
      "epoch": 1.7721174691347374,
      "grad_norm": 0.16779285669326782,
      "learning_rate": 4.0932928512282646e-05,
      "loss": 0.7017,
      "mean_token_accuracy": 0.786484457552433,
      "num_tokens": 614767105.0,
      "step": 19261
    },
    {
      "entropy": 0.6212917976081371,
      "epoch": 1.7722094755061788,
      "grad_norm": 0.15517282485961914,
      "learning_rate": 4.0929861686141013e-05,
      "loss": 0.6115,
      "mean_token_accuracy": 0.8127920888364315,
      "num_tokens": 614799274.0,
      "step": 19262
    },
    {
      "entropy": 0.5589230298064649,
      "epoch": 1.7723014818776202,
      "grad_norm": 0.15417544543743134,
      "learning_rate": 4.092679485999939e-05,
      "loss": 0.5436,
      "mean_token_accuracy": 0.830602154135704,
      "num_tokens": 614831196.0,
      "step": 19263
    },
    {
      "entropy": 0.49504078226163983,
      "epoch": 1.7723934882490613,
      "grad_norm": 0.1544865220785141,
      "learning_rate": 4.092372803385776e-05,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8472822308540344,
      "num_tokens": 614863231.0,
      "step": 19264
    },
    {
      "entropy": 0.6075216885656118,
      "epoch": 1.7724854946205024,
      "grad_norm": 0.15804414451122284,
      "learning_rate": 4.092066120771614e-05,
      "loss": 0.5983,
      "mean_token_accuracy": 0.8130542524158955,
      "num_tokens": 614895833.0,
      "step": 19265
    },
    {
      "entropy": 0.6398003753274679,
      "epoch": 1.7725775009919436,
      "grad_norm": 0.15683691203594208,
      "learning_rate": 4.091759438157451e-05,
      "loss": 0.6361,
      "mean_token_accuracy": 0.8049778863787651,
      "num_tokens": 614927827.0,
      "step": 19266
    },
    {
      "entropy": 0.7305775918066502,
      "epoch": 1.772669507363385,
      "grad_norm": 0.1701153665781021,
      "learning_rate": 4.091452755543289e-05,
      "loss": 0.7386,
      "mean_token_accuracy": 0.7758681252598763,
      "num_tokens": 614958937.0,
      "step": 19267
    },
    {
      "entropy": 0.5097410895396024,
      "epoch": 1.7727615137348263,
      "grad_norm": 0.1509188413619995,
      "learning_rate": 4.0911460729291255e-05,
      "loss": 0.51,
      "mean_token_accuracy": 0.8446149341762066,
      "num_tokens": 614991560.0,
      "step": 19268
    },
    {
      "entropy": 0.6032358296215534,
      "epoch": 1.7728535201062674,
      "grad_norm": 0.1636064052581787,
      "learning_rate": 4.0908393903149636e-05,
      "loss": 0.5999,
      "mean_token_accuracy": 0.8192426785826683,
      "num_tokens": 615023448.0,
      "step": 19269
    },
    {
      "entropy": 0.5235235430300236,
      "epoch": 1.7729455264777085,
      "grad_norm": 0.14119073748588562,
      "learning_rate": 4.0905327077008004e-05,
      "loss": 0.508,
      "mean_token_accuracy": 0.8439401425421238,
      "num_tokens": 615055997.0,
      "step": 19270
    },
    {
      "entropy": 0.5731991920620203,
      "epoch": 1.7730375328491497,
      "grad_norm": 0.14923176169395447,
      "learning_rate": 4.090226025086638e-05,
      "loss": 0.5567,
      "mean_token_accuracy": 0.8263181671500206,
      "num_tokens": 615087816.0,
      "step": 19271
    },
    {
      "entropy": 0.49045774806290865,
      "epoch": 1.773129539220591,
      "grad_norm": 0.14367108047008514,
      "learning_rate": 4.089919342472475e-05,
      "loss": 0.4749,
      "mean_token_accuracy": 0.8542164228856564,
      "num_tokens": 615120373.0,
      "step": 19272
    },
    {
      "entropy": 0.5956164430826902,
      "epoch": 1.7732215455920324,
      "grad_norm": 0.15529562532901764,
      "learning_rate": 4.089612659858313e-05,
      "loss": 0.5891,
      "mean_token_accuracy": 0.8156051635742188,
      "num_tokens": 615151927.0,
      "step": 19273
    },
    {
      "entropy": 0.5603784453123808,
      "epoch": 1.7733135519634735,
      "grad_norm": 0.15219490230083466,
      "learning_rate": 4.08930597724415e-05,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8329316899180412,
      "num_tokens": 615184216.0,
      "step": 19274
    },
    {
      "entropy": 0.5484124924987555,
      "epoch": 1.7734055583349146,
      "grad_norm": 0.14718136191368103,
      "learning_rate": 4.088999294629988e-05,
      "loss": 0.527,
      "mean_token_accuracy": 0.8342149741947651,
      "num_tokens": 615216037.0,
      "step": 19275
    },
    {
      "entropy": 0.660651545971632,
      "epoch": 1.7734975647063558,
      "grad_norm": 0.1653081476688385,
      "learning_rate": 4.0886926120158245e-05,
      "loss": 0.6487,
      "mean_token_accuracy": 0.8028227351605892,
      "num_tokens": 615247507.0,
      "step": 19276
    },
    {
      "entropy": 0.6255127815529704,
      "epoch": 1.7735895710777971,
      "grad_norm": 0.15169158577919006,
      "learning_rate": 4.0883859294016626e-05,
      "loss": 0.6027,
      "mean_token_accuracy": 0.8113902360200882,
      "num_tokens": 615279509.0,
      "step": 19277
    },
    {
      "entropy": 0.5887383385561407,
      "epoch": 1.7736815774492385,
      "grad_norm": 0.15193606913089752,
      "learning_rate": 4.0880792467874994e-05,
      "loss": 0.5781,
      "mean_token_accuracy": 0.824335440993309,
      "num_tokens": 615310971.0,
      "step": 19278
    },
    {
      "entropy": 0.6035459199920297,
      "epoch": 1.7737735838206796,
      "grad_norm": 0.15156826376914978,
      "learning_rate": 4.087772564173337e-05,
      "loss": 0.5961,
      "mean_token_accuracy": 0.8173638172447681,
      "num_tokens": 615343494.0,
      "step": 19279
    },
    {
      "entropy": 0.619488675147295,
      "epoch": 1.7738655901921208,
      "grad_norm": 0.16087111830711365,
      "learning_rate": 4.087465881559175e-05,
      "loss": 0.6177,
      "mean_token_accuracy": 0.8118380978703499,
      "num_tokens": 615375180.0,
      "step": 19280
    },
    {
      "entropy": 0.6269407663494349,
      "epoch": 1.773957596563562,
      "grad_norm": 0.1552698165178299,
      "learning_rate": 4.087159198945012e-05,
      "loss": 0.6231,
      "mean_token_accuracy": 0.8139290437102318,
      "num_tokens": 615407773.0,
      "step": 19281
    },
    {
      "entropy": 0.5185138415545225,
      "epoch": 1.7740496029350032,
      "grad_norm": 0.14731311798095703,
      "learning_rate": 4.08685251633085e-05,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8404700830578804,
      "num_tokens": 615439194.0,
      "step": 19282
    },
    {
      "entropy": 0.6771367508918047,
      "epoch": 1.7741416093064446,
      "grad_norm": 0.15802522003650665,
      "learning_rate": 4.086545833716687e-05,
      "loss": 0.6509,
      "mean_token_accuracy": 0.8002302087843418,
      "num_tokens": 615470891.0,
      "step": 19283
    },
    {
      "entropy": 0.5685902638360858,
      "epoch": 1.7742336156778857,
      "grad_norm": 0.1526743620634079,
      "learning_rate": 4.086239151102524e-05,
      "loss": 0.5713,
      "mean_token_accuracy": 0.825993325561285,
      "num_tokens": 615503342.0,
      "step": 19284
    },
    {
      "entropy": 0.585833746008575,
      "epoch": 1.7743256220493269,
      "grad_norm": 0.15102869272232056,
      "learning_rate": 4.085932468488362e-05,
      "loss": 0.5818,
      "mean_token_accuracy": 0.825806200504303,
      "num_tokens": 615535245.0,
      "step": 19285
    },
    {
      "entropy": 0.6255216631107032,
      "epoch": 1.774417628420768,
      "grad_norm": 0.1607740819454193,
      "learning_rate": 4.085625785874199e-05,
      "loss": 0.6254,
      "mean_token_accuracy": 0.8081391975283623,
      "num_tokens": 615566518.0,
      "step": 19286
    },
    {
      "entropy": 0.5412426525726914,
      "epoch": 1.7745096347922094,
      "grad_norm": 0.15104465186595917,
      "learning_rate": 4.0853191032600366e-05,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8346518836915493,
      "num_tokens": 615598732.0,
      "step": 19287
    },
    {
      "entropy": 0.5839775055646896,
      "epoch": 1.7746016411636507,
      "grad_norm": 0.1533668041229248,
      "learning_rate": 4.085012420645874e-05,
      "loss": 0.5823,
      "mean_token_accuracy": 0.8228689394891262,
      "num_tokens": 615631266.0,
      "step": 19288
    },
    {
      "entropy": 0.609574830159545,
      "epoch": 1.7746936475350918,
      "grad_norm": 0.15623514354228973,
      "learning_rate": 4.084705738031711e-05,
      "loss": 0.5895,
      "mean_token_accuracy": 0.815965261310339,
      "num_tokens": 615662791.0,
      "step": 19289
    },
    {
      "entropy": 0.6494690962135792,
      "epoch": 1.774785653906533,
      "grad_norm": 0.16264905035495758,
      "learning_rate": 4.084399055417549e-05,
      "loss": 0.6499,
      "mean_token_accuracy": 0.8026752769947052,
      "num_tokens": 615693914.0,
      "step": 19290
    },
    {
      "entropy": 0.5625622607767582,
      "epoch": 1.7748776602779741,
      "grad_norm": 0.14848116040229797,
      "learning_rate": 4.084092372803386e-05,
      "loss": 0.5547,
      "mean_token_accuracy": 0.826312318444252,
      "num_tokens": 615726476.0,
      "step": 19291
    },
    {
      "entropy": 0.5699272509664297,
      "epoch": 1.7749696666494155,
      "grad_norm": 0.15227021276950836,
      "learning_rate": 4.083785690189223e-05,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8305531665682793,
      "num_tokens": 615758303.0,
      "step": 19292
    },
    {
      "entropy": 0.6337460298091173,
      "epoch": 1.7750616730208568,
      "grad_norm": 0.15277022123336792,
      "learning_rate": 4.083479007575061e-05,
      "loss": 0.6397,
      "mean_token_accuracy": 0.8003792315721512,
      "num_tokens": 615790755.0,
      "step": 19293
    },
    {
      "entropy": 0.6643698289990425,
      "epoch": 1.775153679392298,
      "grad_norm": 0.15862002968788147,
      "learning_rate": 4.083172324960898e-05,
      "loss": 0.6574,
      "mean_token_accuracy": 0.8009845204651356,
      "num_tokens": 615822332.0,
      "step": 19294
    },
    {
      "entropy": 0.7452677078545094,
      "epoch": 1.775245685763739,
      "grad_norm": 0.16541458666324615,
      "learning_rate": 4.0828656423467356e-05,
      "loss": 0.7265,
      "mean_token_accuracy": 0.7855896316468716,
      "num_tokens": 615854092.0,
      "step": 19295
    },
    {
      "entropy": 0.624284740537405,
      "epoch": 1.7753376921351802,
      "grad_norm": 0.16162770986557007,
      "learning_rate": 4.082558959732573e-05,
      "loss": 0.6185,
      "mean_token_accuracy": 0.8137671276926994,
      "num_tokens": 615885805.0,
      "step": 19296
    },
    {
      "entropy": 0.5455516236834228,
      "epoch": 1.7754296985066216,
      "grad_norm": 0.1482667773962021,
      "learning_rate": 4.08225227711841e-05,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8378506377339363,
      "num_tokens": 615917564.0,
      "step": 19297
    },
    {
      "entropy": 0.6088163126260042,
      "epoch": 1.775521704878063,
      "grad_norm": 0.1567237675189972,
      "learning_rate": 4.081945594504248e-05,
      "loss": 0.605,
      "mean_token_accuracy": 0.8130179606378078,
      "num_tokens": 615949236.0,
      "step": 19298
    },
    {
      "entropy": 0.5478186905384064,
      "epoch": 1.775613711249504,
      "grad_norm": 0.1543591320514679,
      "learning_rate": 4.081638911890085e-05,
      "loss": 0.546,
      "mean_token_accuracy": 0.8337034806609154,
      "num_tokens": 615981213.0,
      "step": 19299
    },
    {
      "entropy": 0.5278043425641954,
      "epoch": 1.7757057176209452,
      "grad_norm": 0.14809182286262512,
      "learning_rate": 4.081332229275922e-05,
      "loss": 0.496,
      "mean_token_accuracy": 0.8447104841470718,
      "num_tokens": 616012965.0,
      "step": 19300
    },
    {
      "entropy": 0.6221133787184954,
      "epoch": 1.7757977239923863,
      "grad_norm": 0.1546306163072586,
      "learning_rate": 4.08102554666176e-05,
      "loss": 0.6189,
      "mean_token_accuracy": 0.8135264329612255,
      "num_tokens": 616045535.0,
      "step": 19301
    },
    {
      "entropy": 0.6187393609434366,
      "epoch": 1.7758897303638277,
      "grad_norm": 0.1484777331352234,
      "learning_rate": 4.080718864047597e-05,
      "loss": 0.6116,
      "mean_token_accuracy": 0.811452180147171,
      "num_tokens": 616077808.0,
      "step": 19302
    },
    {
      "entropy": 0.6929229144006968,
      "epoch": 1.775981736735269,
      "grad_norm": 0.15995344519615173,
      "learning_rate": 4.080412181433435e-05,
      "loss": 0.681,
      "mean_token_accuracy": 0.7888068668544292,
      "num_tokens": 616109409.0,
      "step": 19303
    },
    {
      "entropy": 0.545766388066113,
      "epoch": 1.7760737431067102,
      "grad_norm": 0.14933086931705475,
      "learning_rate": 4.080105498819272e-05,
      "loss": 0.5278,
      "mean_token_accuracy": 0.8378793336451054,
      "num_tokens": 616141844.0,
      "step": 19304
    },
    {
      "entropy": 0.5980487503111362,
      "epoch": 1.7761657494781513,
      "grad_norm": 0.14993059635162354,
      "learning_rate": 4.0797988162051096e-05,
      "loss": 0.5757,
      "mean_token_accuracy": 0.8222357295453548,
      "num_tokens": 616173568.0,
      "step": 19305
    },
    {
      "entropy": 0.5136332903057337,
      "epoch": 1.7762577558495924,
      "grad_norm": 0.14777567982673645,
      "learning_rate": 4.079492133590947e-05,
      "loss": 0.5006,
      "mean_token_accuracy": 0.8444352000951767,
      "num_tokens": 616205702.0,
      "step": 19306
    },
    {
      "entropy": 0.6392662795260549,
      "epoch": 1.7763497622210338,
      "grad_norm": 0.15649789571762085,
      "learning_rate": 4.079185450976784e-05,
      "loss": 0.6349,
      "mean_token_accuracy": 0.8079506605863571,
      "num_tokens": 616238470.0,
      "step": 19307
    },
    {
      "entropy": 0.6273494130000472,
      "epoch": 1.7764417685924752,
      "grad_norm": 0.1616516262292862,
      "learning_rate": 4.078878768362622e-05,
      "loss": 0.6177,
      "mean_token_accuracy": 0.8102927692234516,
      "num_tokens": 616270197.0,
      "step": 19308
    },
    {
      "entropy": 0.6099435752257705,
      "epoch": 1.7765337749639163,
      "grad_norm": 0.1524585336446762,
      "learning_rate": 4.0785720857484595e-05,
      "loss": 0.5922,
      "mean_token_accuracy": 0.8137170858681202,
      "num_tokens": 616302254.0,
      "step": 19309
    },
    {
      "entropy": 0.5407339856028557,
      "epoch": 1.7766257813353574,
      "grad_norm": 0.14841821789741516,
      "learning_rate": 4.078265403134296e-05,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8366318345069885,
      "num_tokens": 616334366.0,
      "step": 19310
    },
    {
      "entropy": 0.5834013689309359,
      "epoch": 1.7767177877067986,
      "grad_norm": 0.15524983406066895,
      "learning_rate": 4.0779587205201344e-05,
      "loss": 0.5756,
      "mean_token_accuracy": 0.8232226744294167,
      "num_tokens": 616366136.0,
      "step": 19311
    },
    {
      "entropy": 0.6673464421182871,
      "epoch": 1.77680979407824,
      "grad_norm": 0.15686681866645813,
      "learning_rate": 4.077652037905971e-05,
      "loss": 0.6597,
      "mean_token_accuracy": 0.7954033799469471,
      "num_tokens": 616398127.0,
      "step": 19312
    },
    {
      "entropy": 0.44426987832412124,
      "epoch": 1.7769018004496813,
      "grad_norm": 0.14466191828250885,
      "learning_rate": 4.0773453552918087e-05,
      "loss": 0.4319,
      "mean_token_accuracy": 0.8663232326507568,
      "num_tokens": 616430416.0,
      "step": 19313
    },
    {
      "entropy": 0.6860130317509174,
      "epoch": 1.7769938068211224,
      "grad_norm": 0.15881353616714478,
      "learning_rate": 4.077038672677646e-05,
      "loss": 0.6858,
      "mean_token_accuracy": 0.7923532873392105,
      "num_tokens": 616462223.0,
      "step": 19314
    },
    {
      "entropy": 0.6296082548797131,
      "epoch": 1.7770858131925635,
      "grad_norm": 0.15853191912174225,
      "learning_rate": 4.0767319900634836e-05,
      "loss": 0.6212,
      "mean_token_accuracy": 0.8081512823700905,
      "num_tokens": 616494341.0,
      "step": 19315
    },
    {
      "entropy": 0.6351584941148758,
      "epoch": 1.7771778195640047,
      "grad_norm": 0.15808849036693573,
      "learning_rate": 4.076425307449321e-05,
      "loss": 0.62,
      "mean_token_accuracy": 0.8071064464747906,
      "num_tokens": 616526470.0,
      "step": 19316
    },
    {
      "entropy": 0.5904481131583452,
      "epoch": 1.777269825935446,
      "grad_norm": 0.15194171667099,
      "learning_rate": 4.0761186248351585e-05,
      "loss": 0.5771,
      "mean_token_accuracy": 0.8238503150641918,
      "num_tokens": 616558592.0,
      "step": 19317
    },
    {
      "entropy": 0.7337514329701662,
      "epoch": 1.7773618323068874,
      "grad_norm": 0.1618199497461319,
      "learning_rate": 4.075811942220995e-05,
      "loss": 0.7139,
      "mean_token_accuracy": 0.7785102352499962,
      "num_tokens": 616590083.0,
      "step": 19318
    },
    {
      "entropy": 0.6197800068184733,
      "epoch": 1.7774538386783285,
      "grad_norm": 0.15631461143493652,
      "learning_rate": 4.0755052596068334e-05,
      "loss": 0.6149,
      "mean_token_accuracy": 0.8122651875019073,
      "num_tokens": 616621278.0,
      "step": 19319
    },
    {
      "entropy": 0.5078245913609862,
      "epoch": 1.7775458450497696,
      "grad_norm": 0.1474807858467102,
      "learning_rate": 4.07519857699267e-05,
      "loss": 0.4982,
      "mean_token_accuracy": 0.8467894904315472,
      "num_tokens": 616653164.0,
      "step": 19320
    },
    {
      "entropy": 0.6432785131037235,
      "epoch": 1.7776378514212108,
      "grad_norm": 0.16109523177146912,
      "learning_rate": 4.0748918943785084e-05,
      "loss": 0.6241,
      "mean_token_accuracy": 0.8046672530472279,
      "num_tokens": 616683366.0,
      "step": 19321
    },
    {
      "entropy": 0.6697529461234808,
      "epoch": 1.7777298577926521,
      "grad_norm": 0.1572074443101883,
      "learning_rate": 4.074585211764345e-05,
      "loss": 0.6658,
      "mean_token_accuracy": 0.7957393936812878,
      "num_tokens": 616715918.0,
      "step": 19322
    },
    {
      "entropy": 0.6454856898635626,
      "epoch": 1.7778218641640935,
      "grad_norm": 0.15970979630947113,
      "learning_rate": 4.0742785291501826e-05,
      "loss": 0.6374,
      "mean_token_accuracy": 0.8054381087422371,
      "num_tokens": 616747042.0,
      "step": 19323
    },
    {
      "entropy": 0.638367410749197,
      "epoch": 1.7779138705355346,
      "grad_norm": 0.16065678000450134,
      "learning_rate": 4.07397184653602e-05,
      "loss": 0.6293,
      "mean_token_accuracy": 0.8069113381206989,
      "num_tokens": 616779135.0,
      "step": 19324
    },
    {
      "entropy": 0.6500659338198602,
      "epoch": 1.7780058769069758,
      "grad_norm": 0.15854527056217194,
      "learning_rate": 4.0736651639218576e-05,
      "loss": 0.638,
      "mean_token_accuracy": 0.8032433874905109,
      "num_tokens": 616811031.0,
      "step": 19325
    },
    {
      "entropy": 0.6215934120118618,
      "epoch": 1.778097883278417,
      "grad_norm": 0.15377932786941528,
      "learning_rate": 4.073358481307695e-05,
      "loss": 0.5922,
      "mean_token_accuracy": 0.8162139728665352,
      "num_tokens": 616842770.0,
      "step": 19326
    },
    {
      "entropy": 0.5223806970752776,
      "epoch": 1.7781898896498582,
      "grad_norm": 0.15566223859786987,
      "learning_rate": 4.0730517986935325e-05,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8404981195926666,
      "num_tokens": 616873936.0,
      "step": 19327
    },
    {
      "entropy": 0.6313034947961569,
      "epoch": 1.7782818960212996,
      "grad_norm": 0.15638358891010284,
      "learning_rate": 4.072745116079369e-05,
      "loss": 0.6373,
      "mean_token_accuracy": 0.8020717464387417,
      "num_tokens": 616906287.0,
      "step": 19328
    },
    {
      "entropy": 0.6370447464287281,
      "epoch": 1.7783739023927407,
      "grad_norm": 0.16179031133651733,
      "learning_rate": 4.0724384334652074e-05,
      "loss": 0.6344,
      "mean_token_accuracy": 0.8037056028842926,
      "num_tokens": 616938147.0,
      "step": 19329
    },
    {
      "entropy": 0.6516362596303225,
      "epoch": 1.7784659087641819,
      "grad_norm": 0.15952858328819275,
      "learning_rate": 4.072131750851044e-05,
      "loss": 0.655,
      "mean_token_accuracy": 0.8018849156796932,
      "num_tokens": 616970556.0,
      "step": 19330
    },
    {
      "entropy": 0.6782269757241011,
      "epoch": 1.778557915135623,
      "grad_norm": 0.1564025580883026,
      "learning_rate": 4.071825068236882e-05,
      "loss": 0.6644,
      "mean_token_accuracy": 0.7937888540327549,
      "num_tokens": 617002619.0,
      "step": 19331
    },
    {
      "entropy": 0.631675710901618,
      "epoch": 1.7786499215070644,
      "grad_norm": 0.15775534510612488,
      "learning_rate": 4.071518385622719e-05,
      "loss": 0.6254,
      "mean_token_accuracy": 0.8074002712965012,
      "num_tokens": 617033772.0,
      "step": 19332
    },
    {
      "entropy": 0.5331694539636374,
      "epoch": 1.7787419278785057,
      "grad_norm": 0.1436474472284317,
      "learning_rate": 4.0712117030085566e-05,
      "loss": 0.5276,
      "mean_token_accuracy": 0.8375514596700668,
      "num_tokens": 617065606.0,
      "step": 19333
    },
    {
      "entropy": 0.7136117201298475,
      "epoch": 1.7788339342499468,
      "grad_norm": 0.16373257339000702,
      "learning_rate": 4.070905020394394e-05,
      "loss": 0.7194,
      "mean_token_accuracy": 0.7842353731393814,
      "num_tokens": 617097729.0,
      "step": 19334
    },
    {
      "entropy": 0.5815692525357008,
      "epoch": 1.778925940621388,
      "grad_norm": 0.15552423894405365,
      "learning_rate": 4.0705983377802315e-05,
      "loss": 0.5667,
      "mean_token_accuracy": 0.8229001238942146,
      "num_tokens": 617129352.0,
      "step": 19335
    },
    {
      "entropy": 0.587492504157126,
      "epoch": 1.7790179469928291,
      "grad_norm": 0.15586695075035095,
      "learning_rate": 4.070291655166068e-05,
      "loss": 0.5647,
      "mean_token_accuracy": 0.8265712521970272,
      "num_tokens": 617161644.0,
      "step": 19336
    },
    {
      "entropy": 0.643276390619576,
      "epoch": 1.7791099533642705,
      "grad_norm": 0.1603449285030365,
      "learning_rate": 4.0699849725519065e-05,
      "loss": 0.6129,
      "mean_token_accuracy": 0.8126229085028172,
      "num_tokens": 617193216.0,
      "step": 19337
    },
    {
      "entropy": 0.6343389246612787,
      "epoch": 1.7792019597357118,
      "grad_norm": 0.1558314710855484,
      "learning_rate": 4.069678289937743e-05,
      "loss": 0.617,
      "mean_token_accuracy": 0.8108134791254997,
      "num_tokens": 617224980.0,
      "step": 19338
    },
    {
      "entropy": 0.5976569037884474,
      "epoch": 1.779293966107153,
      "grad_norm": 0.15776415169239044,
      "learning_rate": 4.069371607323581e-05,
      "loss": 0.5856,
      "mean_token_accuracy": 0.8179762363433838,
      "num_tokens": 617257031.0,
      "step": 19339
    },
    {
      "entropy": 0.5003809914924204,
      "epoch": 1.779385972478594,
      "grad_norm": 0.1455221176147461,
      "learning_rate": 4.069064924709419e-05,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8493923246860504,
      "num_tokens": 617288928.0,
      "step": 19340
    },
    {
      "entropy": 0.5690196843352169,
      "epoch": 1.7794779788500352,
      "grad_norm": 0.1515074372291565,
      "learning_rate": 4.0687582420952556e-05,
      "loss": 0.5592,
      "mean_token_accuracy": 0.8297321312129498,
      "num_tokens": 617320113.0,
      "step": 19341
    },
    {
      "entropy": 0.6324479170143604,
      "epoch": 1.7795699852214766,
      "grad_norm": 0.15281574428081512,
      "learning_rate": 4.068451559481094e-05,
      "loss": 0.6225,
      "mean_token_accuracy": 0.8094739802181721,
      "num_tokens": 617351845.0,
      "step": 19342
    },
    {
      "entropy": 0.5554801551625133,
      "epoch": 1.779661991592918,
      "grad_norm": 0.15016625821590424,
      "learning_rate": 4.0681448768669306e-05,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8354038149118423,
      "num_tokens": 617383730.0,
      "step": 19343
    },
    {
      "entropy": 0.635894276201725,
      "epoch": 1.779753997964359,
      "grad_norm": 0.15888488292694092,
      "learning_rate": 4.067838194252768e-05,
      "loss": 0.6307,
      "mean_token_accuracy": 0.8084451667964458,
      "num_tokens": 617416033.0,
      "step": 19344
    },
    {
      "entropy": 0.6799620920792222,
      "epoch": 1.7798460043358002,
      "grad_norm": 0.1587878167629242,
      "learning_rate": 4.0675315116386055e-05,
      "loss": 0.6704,
      "mean_token_accuracy": 0.7963862381875515,
      "num_tokens": 617447657.0,
      "step": 19345
    },
    {
      "entropy": 0.5603113023098558,
      "epoch": 1.7799380107072413,
      "grad_norm": 0.14817586541175842,
      "learning_rate": 4.067224829024443e-05,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8308286294341087,
      "num_tokens": 617479524.0,
      "step": 19346
    },
    {
      "entropy": 0.5369715974666178,
      "epoch": 1.7800300170786827,
      "grad_norm": 0.1478525549173355,
      "learning_rate": 4.0669181464102804e-05,
      "loss": 0.5312,
      "mean_token_accuracy": 0.834588099271059,
      "num_tokens": 617511695.0,
      "step": 19347
    },
    {
      "entropy": 0.6180419381707907,
      "epoch": 1.780122023450124,
      "grad_norm": 0.15077345073223114,
      "learning_rate": 4.066611463796118e-05,
      "loss": 0.5882,
      "mean_token_accuracy": 0.818224836140871,
      "num_tokens": 617543557.0,
      "step": 19348
    },
    {
      "entropy": 0.6186595819890499,
      "epoch": 1.7802140298215652,
      "grad_norm": 0.15373262763023376,
      "learning_rate": 4.066304781181955e-05,
      "loss": 0.6076,
      "mean_token_accuracy": 0.8132040500640869,
      "num_tokens": 617575942.0,
      "step": 19349
    },
    {
      "entropy": 0.5279638799838722,
      "epoch": 1.7803060361930063,
      "grad_norm": 0.1528480499982834,
      "learning_rate": 4.065998098567793e-05,
      "loss": 0.526,
      "mean_token_accuracy": 0.838721938431263,
      "num_tokens": 617607537.0,
      "step": 19350
    },
    {
      "entropy": 0.6154088322073221,
      "epoch": 1.7803980425644474,
      "grad_norm": 0.15164780616760254,
      "learning_rate": 4.0656914159536296e-05,
      "loss": 0.6044,
      "mean_token_accuracy": 0.8128847554326057,
      "num_tokens": 617639928.0,
      "step": 19351
    },
    {
      "entropy": 0.6038249433040619,
      "epoch": 1.7804900489358888,
      "grad_norm": 0.15399812161922455,
      "learning_rate": 4.065384733339467e-05,
      "loss": 0.5926,
      "mean_token_accuracy": 0.8174360729753971,
      "num_tokens": 617671778.0,
      "step": 19352
    },
    {
      "entropy": 0.5853242613375187,
      "epoch": 1.7805820553073302,
      "grad_norm": 0.1500568389892578,
      "learning_rate": 4.0650780507253045e-05,
      "loss": 0.5789,
      "mean_token_accuracy": 0.8206394240260124,
      "num_tokens": 617703845.0,
      "step": 19353
    },
    {
      "entropy": 0.6297962516546249,
      "epoch": 1.7806740616787713,
      "grad_norm": 0.15431979298591614,
      "learning_rate": 4.064771368111142e-05,
      "loss": 0.6312,
      "mean_token_accuracy": 0.8075109086930752,
      "num_tokens": 617736397.0,
      "step": 19354
    },
    {
      "entropy": 0.6937397010624409,
      "epoch": 1.7807660680502124,
      "grad_norm": 0.1568097025156021,
      "learning_rate": 4.0644646854969795e-05,
      "loss": 0.6919,
      "mean_token_accuracy": 0.7891132533550262,
      "num_tokens": 617769008.0,
      "step": 19355
    },
    {
      "entropy": 0.67493562027812,
      "epoch": 1.7808580744216536,
      "grad_norm": 0.1605442762374878,
      "learning_rate": 4.064158002882817e-05,
      "loss": 0.6648,
      "mean_token_accuracy": 0.797492478042841,
      "num_tokens": 617799663.0,
      "step": 19356
    },
    {
      "entropy": 0.6183915287256241,
      "epoch": 1.780950080793095,
      "grad_norm": 0.15030832588672638,
      "learning_rate": 4.063851320268654e-05,
      "loss": 0.607,
      "mean_token_accuracy": 0.8125099167227745,
      "num_tokens": 617831760.0,
      "step": 19357
    },
    {
      "entropy": 0.6094676898792386,
      "epoch": 1.7810420871645363,
      "grad_norm": 0.1561734676361084,
      "learning_rate": 4.063544637654492e-05,
      "loss": 0.5932,
      "mean_token_accuracy": 0.8194180801510811,
      "num_tokens": 617863805.0,
      "step": 19358
    },
    {
      "entropy": 0.6629252154380083,
      "epoch": 1.7811340935359774,
      "grad_norm": 0.1561354398727417,
      "learning_rate": 4.0632379550403286e-05,
      "loss": 0.6493,
      "mean_token_accuracy": 0.7991216741502285,
      "num_tokens": 617895562.0,
      "step": 19359
    },
    {
      "entropy": 0.6147411679849029,
      "epoch": 1.7812260999074185,
      "grad_norm": 0.15059995651245117,
      "learning_rate": 4.062931272426166e-05,
      "loss": 0.5818,
      "mean_token_accuracy": 0.8195209391415119,
      "num_tokens": 617927430.0,
      "step": 19360
    },
    {
      "entropy": 0.5835388572886586,
      "epoch": 1.7813181062788597,
      "grad_norm": 0.14790676534175873,
      "learning_rate": 4.0626245898120036e-05,
      "loss": 0.5633,
      "mean_token_accuracy": 0.8263020552694798,
      "num_tokens": 617959687.0,
      "step": 19361
    },
    {
      "entropy": 0.6468011643737555,
      "epoch": 1.781410112650301,
      "grad_norm": 0.15608282387256622,
      "learning_rate": 4.062317907197841e-05,
      "loss": 0.6438,
      "mean_token_accuracy": 0.804775282740593,
      "num_tokens": 617992455.0,
      "step": 19362
    },
    {
      "entropy": 0.647794621065259,
      "epoch": 1.7815021190217424,
      "grad_norm": 0.1622731238603592,
      "learning_rate": 4.0620112245836785e-05,
      "loss": 0.6417,
      "mean_token_accuracy": 0.806134320795536,
      "num_tokens": 618024266.0,
      "step": 19363
    },
    {
      "entropy": 0.6220854688435793,
      "epoch": 1.7815941253931835,
      "grad_norm": 0.15826494991779327,
      "learning_rate": 4.061704541969516e-05,
      "loss": 0.6206,
      "mean_token_accuracy": 0.8083418384194374,
      "num_tokens": 618056343.0,
      "step": 19364
    },
    {
      "entropy": 0.5409425459802151,
      "epoch": 1.7816861317646246,
      "grad_norm": 0.14872236549854279,
      "learning_rate": 4.0613978593553534e-05,
      "loss": 0.5328,
      "mean_token_accuracy": 0.8356994688510895,
      "num_tokens": 618088478.0,
      "step": 19365
    },
    {
      "entropy": 0.543692484498024,
      "epoch": 1.7817781381360658,
      "grad_norm": 0.15091903507709503,
      "learning_rate": 4.061091176741191e-05,
      "loss": 0.543,
      "mean_token_accuracy": 0.8296348340809345,
      "num_tokens": 618120414.0,
      "step": 19366
    },
    {
      "entropy": 0.528448099270463,
      "epoch": 1.7818701445075071,
      "grad_norm": 0.15045605599880219,
      "learning_rate": 4.060784494127028e-05,
      "loss": 0.5241,
      "mean_token_accuracy": 0.8334177024662495,
      "num_tokens": 618152928.0,
      "step": 19367
    },
    {
      "entropy": 0.620751716196537,
      "epoch": 1.7819621508789485,
      "grad_norm": 0.15160001814365387,
      "learning_rate": 4.060477811512866e-05,
      "loss": 0.6035,
      "mean_token_accuracy": 0.8122178092598915,
      "num_tokens": 618184733.0,
      "step": 19368
    },
    {
      "entropy": 0.6267432840541005,
      "epoch": 1.7820541572503896,
      "grad_norm": 0.16207276284694672,
      "learning_rate": 4.0601711288987026e-05,
      "loss": 0.6147,
      "mean_token_accuracy": 0.8094052895903587,
      "num_tokens": 618216183.0,
      "step": 19369
    },
    {
      "entropy": 0.5082770697772503,
      "epoch": 1.7821461636218308,
      "grad_norm": 0.15529531240463257,
      "learning_rate": 4.05986444628454e-05,
      "loss": 0.487,
      "mean_token_accuracy": 0.8470663614571095,
      "num_tokens": 618247762.0,
      "step": 19370
    },
    {
      "entropy": 0.6007422059774399,
      "epoch": 1.782238169993272,
      "grad_norm": 0.1583109200000763,
      "learning_rate": 4.059557763670378e-05,
      "loss": 0.6017,
      "mean_token_accuracy": 0.8198062516748905,
      "num_tokens": 618279427.0,
      "step": 19371
    },
    {
      "entropy": 0.5361032169312239,
      "epoch": 1.7823301763647132,
      "grad_norm": 0.15163756906986237,
      "learning_rate": 4.059251081056215e-05,
      "loss": 0.525,
      "mean_token_accuracy": 0.8421176895499229,
      "num_tokens": 618311672.0,
      "step": 19372
    },
    {
      "entropy": 0.5361483814194798,
      "epoch": 1.7824221827361546,
      "grad_norm": 0.1476820856332779,
      "learning_rate": 4.0589443984420525e-05,
      "loss": 0.5145,
      "mean_token_accuracy": 0.8394832126796246,
      "num_tokens": 618342919.0,
      "step": 19373
    },
    {
      "entropy": 0.707514557056129,
      "epoch": 1.7825141891075957,
      "grad_norm": 0.15625424683094025,
      "learning_rate": 4.05863771582789e-05,
      "loss": 0.6919,
      "mean_token_accuracy": 0.787882924079895,
      "num_tokens": 618374855.0,
      "step": 19374
    },
    {
      "entropy": 0.6972712492570281,
      "epoch": 1.7826061954790369,
      "grad_norm": 0.16503480076789856,
      "learning_rate": 4.0583310332137274e-05,
      "loss": 0.668,
      "mean_token_accuracy": 0.7964018024504185,
      "num_tokens": 618406455.0,
      "step": 19375
    },
    {
      "entropy": 0.5212913099676371,
      "epoch": 1.782698201850478,
      "grad_norm": 0.1457734853029251,
      "learning_rate": 4.058024350599565e-05,
      "loss": 0.5083,
      "mean_token_accuracy": 0.8411864899098873,
      "num_tokens": 618438408.0,
      "step": 19376
    },
    {
      "entropy": 0.5714408550411463,
      "epoch": 1.7827902082219194,
      "grad_norm": 0.15314331650733948,
      "learning_rate": 4.057717667985402e-05,
      "loss": 0.5624,
      "mean_token_accuracy": 0.8265142701566219,
      "num_tokens": 618470159.0,
      "step": 19377
    },
    {
      "entropy": 0.5738061561714858,
      "epoch": 1.7828822145933607,
      "grad_norm": 0.14882177114486694,
      "learning_rate": 4.057410985371239e-05,
      "loss": 0.5761,
      "mean_token_accuracy": 0.8243330158293247,
      "num_tokens": 618502359.0,
      "step": 19378
    },
    {
      "entropy": 0.6283177779987454,
      "epoch": 1.7829742209648018,
      "grad_norm": 0.1610935479402542,
      "learning_rate": 4.057104302757077e-05,
      "loss": 0.6046,
      "mean_token_accuracy": 0.814644068479538,
      "num_tokens": 618533719.0,
      "step": 19379
    },
    {
      "entropy": 0.7496877796947956,
      "epoch": 1.783066227336243,
      "grad_norm": 0.16520072519779205,
      "learning_rate": 4.056797620142914e-05,
      "loss": 0.7355,
      "mean_token_accuracy": 0.7797323726117611,
      "num_tokens": 618565467.0,
      "step": 19380
    },
    {
      "entropy": 0.6190337631851435,
      "epoch": 1.7831582337076841,
      "grad_norm": 0.15276969969272614,
      "learning_rate": 4.056490937528752e-05,
      "loss": 0.6022,
      "mean_token_accuracy": 0.8119705803692341,
      "num_tokens": 618596972.0,
      "step": 19381
    },
    {
      "entropy": 0.5684828320518136,
      "epoch": 1.7832502400791255,
      "grad_norm": 0.14756153523921967,
      "learning_rate": 4.056184254914589e-05,
      "loss": 0.556,
      "mean_token_accuracy": 0.8264755681157112,
      "num_tokens": 618629376.0,
      "step": 19382
    },
    {
      "entropy": 0.6380664948374033,
      "epoch": 1.7833422464505668,
      "grad_norm": 0.15820902585983276,
      "learning_rate": 4.0558775723004264e-05,
      "loss": 0.6327,
      "mean_token_accuracy": 0.8064534850418568,
      "num_tokens": 618661090.0,
      "step": 19383
    },
    {
      "entropy": 0.5097991665825248,
      "epoch": 1.783434252822008,
      "grad_norm": 0.1527581512928009,
      "learning_rate": 4.055570889686264e-05,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8441308923065662,
      "num_tokens": 618692979.0,
      "step": 19384
    },
    {
      "entropy": 0.6060345582664013,
      "epoch": 1.783526259193449,
      "grad_norm": 0.16170981526374817,
      "learning_rate": 4.0552642070721014e-05,
      "loss": 0.6101,
      "mean_token_accuracy": 0.8177350200712681,
      "num_tokens": 618724574.0,
      "step": 19385
    },
    {
      "entropy": 0.6303611323237419,
      "epoch": 1.7836182655648902,
      "grad_norm": 0.15853962302207947,
      "learning_rate": 4.054957524457939e-05,
      "loss": 0.6334,
      "mean_token_accuracy": 0.8031510226428509,
      "num_tokens": 618756709.0,
      "step": 19386
    },
    {
      "entropy": 0.648733826354146,
      "epoch": 1.7837102719363316,
      "grad_norm": 0.16017091274261475,
      "learning_rate": 4.054650841843776e-05,
      "loss": 0.665,
      "mean_token_accuracy": 0.7952555231750011,
      "num_tokens": 618788816.0,
      "step": 19387
    },
    {
      "entropy": 0.543773652985692,
      "epoch": 1.783802278307773,
      "grad_norm": 0.15019646286964417,
      "learning_rate": 4.054344159229613e-05,
      "loss": 0.5381,
      "mean_token_accuracy": 0.836064662784338,
      "num_tokens": 618820824.0,
      "step": 19388
    },
    {
      "entropy": 0.5842665638774633,
      "epoch": 1.783894284679214,
      "grad_norm": 0.15700194239616394,
      "learning_rate": 4.054037476615451e-05,
      "loss": 0.5744,
      "mean_token_accuracy": 0.823285385966301,
      "num_tokens": 618852004.0,
      "step": 19389
    },
    {
      "entropy": 0.598788027651608,
      "epoch": 1.7839862910506552,
      "grad_norm": 0.14958555996418,
      "learning_rate": 4.053730794001288e-05,
      "loss": 0.5897,
      "mean_token_accuracy": 0.8183458149433136,
      "num_tokens": 618884417.0,
      "step": 19390
    },
    {
      "entropy": 0.49426087853498757,
      "epoch": 1.7840782974220963,
      "grad_norm": 0.14236240088939667,
      "learning_rate": 4.0534241113871255e-05,
      "loss": 0.4812,
      "mean_token_accuracy": 0.8514648377895355,
      "num_tokens": 618917144.0,
      "step": 19391
    },
    {
      "entropy": 0.5508274976164103,
      "epoch": 1.7841703037935377,
      "grad_norm": 0.14616477489471436,
      "learning_rate": 4.053117428772963e-05,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8355760462582111,
      "num_tokens": 618949655.0,
      "step": 19392
    },
    {
      "entropy": 0.5753605514764786,
      "epoch": 1.784262310164979,
      "grad_norm": 0.15197882056236267,
      "learning_rate": 4.0528107461588004e-05,
      "loss": 0.57,
      "mean_token_accuracy": 0.8252900466322899,
      "num_tokens": 618981461.0,
      "step": 19393
    },
    {
      "entropy": 0.5665397886186838,
      "epoch": 1.7843543165364202,
      "grad_norm": 0.15048739314079285,
      "learning_rate": 4.052504063544638e-05,
      "loss": 0.5556,
      "mean_token_accuracy": 0.8297578543424606,
      "num_tokens": 619013814.0,
      "step": 19394
    },
    {
      "entropy": 0.5520739909261465,
      "epoch": 1.7844463229078613,
      "grad_norm": 0.15293216705322266,
      "learning_rate": 4.0521973809304753e-05,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8332019932568073,
      "num_tokens": 619045666.0,
      "step": 19395
    },
    {
      "entropy": 0.6229084357619286,
      "epoch": 1.7845383292793024,
      "grad_norm": 0.15209431946277618,
      "learning_rate": 4.051890698316312e-05,
      "loss": 0.6109,
      "mean_token_accuracy": 0.8144810907542706,
      "num_tokens": 619078042.0,
      "step": 19396
    },
    {
      "entropy": 0.6142837405204773,
      "epoch": 1.7846303356507438,
      "grad_norm": 0.16216455399990082,
      "learning_rate": 4.05158401570215e-05,
      "loss": 0.587,
      "mean_token_accuracy": 0.8203213550150394,
      "num_tokens": 619109756.0,
      "step": 19397
    },
    {
      "entropy": 0.5564159005880356,
      "epoch": 1.7847223420221852,
      "grad_norm": 0.14579521119594574,
      "learning_rate": 4.051277333087987e-05,
      "loss": 0.5423,
      "mean_token_accuracy": 0.8315212763845921,
      "num_tokens": 619142079.0,
      "step": 19398
    },
    {
      "entropy": 0.5709600588306785,
      "epoch": 1.7848143483936263,
      "grad_norm": 0.14711928367614746,
      "learning_rate": 4.0509706504738245e-05,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8276379965245724,
      "num_tokens": 619174425.0,
      "step": 19399
    },
    {
      "entropy": 0.7492733784019947,
      "epoch": 1.7849063547650674,
      "grad_norm": 0.17185470461845398,
      "learning_rate": 4.050663967859662e-05,
      "loss": 0.748,
      "mean_token_accuracy": 0.7723137997090816,
      "num_tokens": 619205605.0,
      "step": 19400
    },
    {
      "entropy": 0.577132023870945,
      "epoch": 1.7849983611365086,
      "grad_norm": 0.15134809911251068,
      "learning_rate": 4.0503572852454995e-05,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8260027579963207,
      "num_tokens": 619238002.0,
      "step": 19401
    },
    {
      "entropy": 0.47995578730478883,
      "epoch": 1.78509036750795,
      "grad_norm": 0.14022444188594818,
      "learning_rate": 4.0500506026313376e-05,
      "loss": 0.462,
      "mean_token_accuracy": 0.8533686175942421,
      "num_tokens": 619269465.0,
      "step": 19402
    },
    {
      "entropy": 0.6682004574686289,
      "epoch": 1.7851823738793913,
      "grad_norm": 0.16693517565727234,
      "learning_rate": 4.0497439200171744e-05,
      "loss": 0.6704,
      "mean_token_accuracy": 0.7987548857927322,
      "num_tokens": 619301078.0,
      "step": 19403
    },
    {
      "entropy": 0.6005545370280743,
      "epoch": 1.7852743802508324,
      "grad_norm": 0.15814149379730225,
      "learning_rate": 4.049437237403012e-05,
      "loss": 0.6028,
      "mean_token_accuracy": 0.814648300409317,
      "num_tokens": 619333043.0,
      "step": 19404
    },
    {
      "entropy": 0.5741257909685373,
      "epoch": 1.7853663866222735,
      "grad_norm": 0.15613967180252075,
      "learning_rate": 4.049130554788849e-05,
      "loss": 0.5612,
      "mean_token_accuracy": 0.8214654587209225,
      "num_tokens": 619364463.0,
      "step": 19405
    },
    {
      "entropy": 0.5739250713959336,
      "epoch": 1.7854583929937147,
      "grad_norm": 0.15090397000312805,
      "learning_rate": 4.048823872174687e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8263272009789944,
      "num_tokens": 619396218.0,
      "step": 19406
    },
    {
      "entropy": 0.5120488358661532,
      "epoch": 1.785550399365156,
      "grad_norm": 0.14493465423583984,
      "learning_rate": 4.048517189560524e-05,
      "loss": 0.506,
      "mean_token_accuracy": 0.8418328128755093,
      "num_tokens": 619428239.0,
      "step": 19407
    },
    {
      "entropy": 0.5322709772735834,
      "epoch": 1.7856424057365974,
      "grad_norm": 0.15113817155361176,
      "learning_rate": 4.048210506946362e-05,
      "loss": 0.5179,
      "mean_token_accuracy": 0.8399789296090603,
      "num_tokens": 619459954.0,
      "step": 19408
    },
    {
      "entropy": 0.6743805445730686,
      "epoch": 1.7857344121080385,
      "grad_norm": 0.1569666862487793,
      "learning_rate": 4.0479038243321985e-05,
      "loss": 0.671,
      "mean_token_accuracy": 0.7981225475668907,
      "num_tokens": 619492185.0,
      "step": 19409
    },
    {
      "entropy": 0.531962369568646,
      "epoch": 1.7858264184794796,
      "grad_norm": 0.15929952263832092,
      "learning_rate": 4.0475971417180366e-05,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8383386805653572,
      "num_tokens": 619524149.0,
      "step": 19410
    },
    {
      "entropy": 0.6625535599887371,
      "epoch": 1.7859184248509208,
      "grad_norm": 0.1573571115732193,
      "learning_rate": 4.0472904591038734e-05,
      "loss": 0.6504,
      "mean_token_accuracy": 0.8031328432261944,
      "num_tokens": 619555627.0,
      "step": 19411
    },
    {
      "entropy": 0.6629178933799267,
      "epoch": 1.7860104312223621,
      "grad_norm": 0.16129058599472046,
      "learning_rate": 4.046983776489711e-05,
      "loss": 0.6684,
      "mean_token_accuracy": 0.7979425080120564,
      "num_tokens": 619587299.0,
      "step": 19412
    },
    {
      "entropy": 0.6864123269915581,
      "epoch": 1.7861024375938035,
      "grad_norm": 0.15851804614067078,
      "learning_rate": 4.0466770938755484e-05,
      "loss": 0.6777,
      "mean_token_accuracy": 0.792851272970438,
      "num_tokens": 619618929.0,
      "step": 19413
    },
    {
      "entropy": 0.6282221078872681,
      "epoch": 1.7861944439652446,
      "grad_norm": 0.1528451144695282,
      "learning_rate": 4.046370411261386e-05,
      "loss": 0.5998,
      "mean_token_accuracy": 0.8127072900533676,
      "num_tokens": 619651007.0,
      "step": 19414
    },
    {
      "entropy": 0.5854148026555777,
      "epoch": 1.7862864503366858,
      "grad_norm": 0.15663400292396545,
      "learning_rate": 4.046063728647223e-05,
      "loss": 0.5723,
      "mean_token_accuracy": 0.8208932168781757,
      "num_tokens": 619682513.0,
      "step": 19415
    },
    {
      "entropy": 0.3992458516731858,
      "epoch": 1.786378456708127,
      "grad_norm": 0.13477690517902374,
      "learning_rate": 4.045757046033061e-05,
      "loss": 0.3858,
      "mean_token_accuracy": 0.8793476484715939,
      "num_tokens": 619714873.0,
      "step": 19416
    },
    {
      "entropy": 0.6450210059992969,
      "epoch": 1.7864704630795682,
      "grad_norm": 0.15605829656124115,
      "learning_rate": 4.0454503634188975e-05,
      "loss": 0.6276,
      "mean_token_accuracy": 0.8075947500765324,
      "num_tokens": 619747044.0,
      "step": 19417
    },
    {
      "entropy": 0.6335801854729652,
      "epoch": 1.7865624694510096,
      "grad_norm": 0.15109550952911377,
      "learning_rate": 4.045143680804736e-05,
      "loss": 0.6158,
      "mean_token_accuracy": 0.8121349848806858,
      "num_tokens": 619778686.0,
      "step": 19418
    },
    {
      "entropy": 0.6066373633220792,
      "epoch": 1.7866544758224507,
      "grad_norm": 0.16070930659770966,
      "learning_rate": 4.0448369981905725e-05,
      "loss": 0.5948,
      "mean_token_accuracy": 0.8206999897956848,
      "num_tokens": 619809796.0,
      "step": 19419
    },
    {
      "entropy": 0.5440232641994953,
      "epoch": 1.7867464821938919,
      "grad_norm": 0.14702089130878448,
      "learning_rate": 4.04453031557641e-05,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8285122960805893,
      "num_tokens": 619841525.0,
      "step": 19420
    },
    {
      "entropy": 0.6043472848832607,
      "epoch": 1.786838488565333,
      "grad_norm": 0.15190161764621735,
      "learning_rate": 4.0442236329622474e-05,
      "loss": 0.5816,
      "mean_token_accuracy": 0.8183502219617367,
      "num_tokens": 619873437.0,
      "step": 19421
    },
    {
      "entropy": 0.5114789633080363,
      "epoch": 1.7869304949367744,
      "grad_norm": 0.14652663469314575,
      "learning_rate": 4.043916950348085e-05,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8429297842085361,
      "num_tokens": 619905018.0,
      "step": 19422
    },
    {
      "entropy": 0.6450016871094704,
      "epoch": 1.7870225013082157,
      "grad_norm": 0.15896062552928925,
      "learning_rate": 4.043610267733922e-05,
      "loss": 0.6266,
      "mean_token_accuracy": 0.8029185906052589,
      "num_tokens": 619936164.0,
      "step": 19423
    },
    {
      "entropy": 0.6891979295760393,
      "epoch": 1.7871145076796568,
      "grad_norm": 0.1604168862104416,
      "learning_rate": 4.04330358511976e-05,
      "loss": 0.6801,
      "mean_token_accuracy": 0.7922062911093235,
      "num_tokens": 619968349.0,
      "step": 19424
    },
    {
      "entropy": 0.5167165640741587,
      "epoch": 1.787206514051098,
      "grad_norm": 0.14883540570735931,
      "learning_rate": 4.042996902505597e-05,
      "loss": 0.501,
      "mean_token_accuracy": 0.8426244556903839,
      "num_tokens": 620000025.0,
      "step": 19425
    },
    {
      "entropy": 0.48757585580460727,
      "epoch": 1.7872985204225391,
      "grad_norm": 0.1435290277004242,
      "learning_rate": 4.042690219891435e-05,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8546939678490162,
      "num_tokens": 620032128.0,
      "step": 19426
    },
    {
      "entropy": 0.5371782323345542,
      "epoch": 1.7873905267939805,
      "grad_norm": 0.15197806060314178,
      "learning_rate": 4.0423835372772715e-05,
      "loss": 0.5247,
      "mean_token_accuracy": 0.8404375985264778,
      "num_tokens": 620064248.0,
      "step": 19427
    },
    {
      "entropy": 0.5311647886410356,
      "epoch": 1.7874825331654218,
      "grad_norm": 0.15105675160884857,
      "learning_rate": 4.0420768546631096e-05,
      "loss": 0.521,
      "mean_token_accuracy": 0.8354666605591774,
      "num_tokens": 620096338.0,
      "step": 19428
    },
    {
      "entropy": 0.6032483708113432,
      "epoch": 1.787574539536863,
      "grad_norm": 0.15484698116779327,
      "learning_rate": 4.0417701720489464e-05,
      "loss": 0.6048,
      "mean_token_accuracy": 0.8111373409628868,
      "num_tokens": 620127596.0,
      "step": 19429
    },
    {
      "entropy": 0.5518551748245955,
      "epoch": 1.787666545908304,
      "grad_norm": 0.1504993438720703,
      "learning_rate": 4.041463489434784e-05,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8269849680364132,
      "num_tokens": 620160345.0,
      "step": 19430
    },
    {
      "entropy": 0.6067683380097151,
      "epoch": 1.7877585522797452,
      "grad_norm": 0.15027935802936554,
      "learning_rate": 4.041156806820622e-05,
      "loss": 0.5905,
      "mean_token_accuracy": 0.8205276317894459,
      "num_tokens": 620191899.0,
      "step": 19431
    },
    {
      "entropy": 0.6244331784546375,
      "epoch": 1.7878505586511866,
      "grad_norm": 0.16088362038135529,
      "learning_rate": 4.040850124206459e-05,
      "loss": 0.6267,
      "mean_token_accuracy": 0.8095851764082909,
      "num_tokens": 620223595.0,
      "step": 19432
    },
    {
      "entropy": 0.6005472457036376,
      "epoch": 1.787942565022628,
      "grad_norm": 0.1564405858516693,
      "learning_rate": 4.040543441592296e-05,
      "loss": 0.5825,
      "mean_token_accuracy": 0.8191997744143009,
      "num_tokens": 620254635.0,
      "step": 19433
    },
    {
      "entropy": 0.560071736574173,
      "epoch": 1.788034571394069,
      "grad_norm": 0.15496431291103363,
      "learning_rate": 4.040236758978134e-05,
      "loss": 0.5447,
      "mean_token_accuracy": 0.828532338142395,
      "num_tokens": 620285766.0,
      "step": 19434
    },
    {
      "entropy": 0.595000920817256,
      "epoch": 1.7881265777655102,
      "grad_norm": 0.1553345024585724,
      "learning_rate": 4.039930076363971e-05,
      "loss": 0.585,
      "mean_token_accuracy": 0.82387974858284,
      "num_tokens": 620317256.0,
      "step": 19435
    },
    {
      "entropy": 0.6789679229259491,
      "epoch": 1.7882185841369513,
      "grad_norm": 0.1621452122926712,
      "learning_rate": 4.039623393749809e-05,
      "loss": 0.6708,
      "mean_token_accuracy": 0.794887438416481,
      "num_tokens": 620349571.0,
      "step": 19436
    },
    {
      "entropy": 0.674022862687707,
      "epoch": 1.7883105905083927,
      "grad_norm": 0.16251768171787262,
      "learning_rate": 4.039316711135646e-05,
      "loss": 0.6537,
      "mean_token_accuracy": 0.7995859645307064,
      "num_tokens": 620381931.0,
      "step": 19437
    },
    {
      "entropy": 0.697617806494236,
      "epoch": 1.788402596879834,
      "grad_norm": 0.16446952521800995,
      "learning_rate": 4.039010028521483e-05,
      "loss": 0.7037,
      "mean_token_accuracy": 0.7869775369763374,
      "num_tokens": 620413298.0,
      "step": 19438
    },
    {
      "entropy": 0.7043021097779274,
      "epoch": 1.7884946032512752,
      "grad_norm": 0.16088329255580902,
      "learning_rate": 4.038703345907321e-05,
      "loss": 0.7001,
      "mean_token_accuracy": 0.7867378629744053,
      "num_tokens": 620444805.0,
      "step": 19439
    },
    {
      "entropy": 0.6240355595946312,
      "epoch": 1.7885866096227163,
      "grad_norm": 0.1574057936668396,
      "learning_rate": 4.038396663293158e-05,
      "loss": 0.6049,
      "mean_token_accuracy": 0.8117480538785458,
      "num_tokens": 620475810.0,
      "step": 19440
    },
    {
      "entropy": 0.6785184685140848,
      "epoch": 1.7886786159941575,
      "grad_norm": 0.15467917919158936,
      "learning_rate": 4.038089980678996e-05,
      "loss": 0.6727,
      "mean_token_accuracy": 0.7915291823446751,
      "num_tokens": 620508105.0,
      "step": 19441
    },
    {
      "entropy": 0.5988051351159811,
      "epoch": 1.7887706223655988,
      "grad_norm": 0.1556602418422699,
      "learning_rate": 4.037783298064833e-05,
      "loss": 0.5883,
      "mean_token_accuracy": 0.818275410681963,
      "num_tokens": 620540492.0,
      "step": 19442
    },
    {
      "entropy": 0.5828497326001525,
      "epoch": 1.7888626287370402,
      "grad_norm": 0.15281589329242706,
      "learning_rate": 4.03747661545067e-05,
      "loss": 0.5616,
      "mean_token_accuracy": 0.8220208659768105,
      "num_tokens": 620571803.0,
      "step": 19443
    },
    {
      "entropy": 0.522799483500421,
      "epoch": 1.7889546351084813,
      "grad_norm": 0.15659697353839874,
      "learning_rate": 4.037169932836508e-05,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8438123613595963,
      "num_tokens": 620602917.0,
      "step": 19444
    },
    {
      "entropy": 0.5886427350342274,
      "epoch": 1.7890466414799224,
      "grad_norm": 0.15230081975460052,
      "learning_rate": 4.036863250222345e-05,
      "loss": 0.5867,
      "mean_token_accuracy": 0.8190001659095287,
      "num_tokens": 620635145.0,
      "step": 19445
    },
    {
      "entropy": 0.6434482727199793,
      "epoch": 1.7891386478513636,
      "grad_norm": 0.15473097562789917,
      "learning_rate": 4.0365565676081827e-05,
      "loss": 0.6291,
      "mean_token_accuracy": 0.8087826855480671,
      "num_tokens": 620666688.0,
      "step": 19446
    },
    {
      "entropy": 0.5686134733259678,
      "epoch": 1.789230654222805,
      "grad_norm": 0.14524951577186584,
      "learning_rate": 4.03624988499402e-05,
      "loss": 0.5593,
      "mean_token_accuracy": 0.8275332041084766,
      "num_tokens": 620699081.0,
      "step": 19447
    },
    {
      "entropy": 0.6237299358472228,
      "epoch": 1.7893226605942463,
      "grad_norm": 0.15425504744052887,
      "learning_rate": 4.035943202379857e-05,
      "loss": 0.6071,
      "mean_token_accuracy": 0.8142830282449722,
      "num_tokens": 620730970.0,
      "step": 19448
    },
    {
      "entropy": 0.6431021429598331,
      "epoch": 1.7894146669656874,
      "grad_norm": 0.1586689054965973,
      "learning_rate": 4.035636519765695e-05,
      "loss": 0.6381,
      "mean_token_accuracy": 0.8067659884691238,
      "num_tokens": 620762704.0,
      "step": 19449
    },
    {
      "entropy": 0.619950957596302,
      "epoch": 1.7895066733371285,
      "grad_norm": 0.16406364738941193,
      "learning_rate": 4.035329837151532e-05,
      "loss": 0.6051,
      "mean_token_accuracy": 0.8155817538499832,
      "num_tokens": 620795397.0,
      "step": 19450
    },
    {
      "entropy": 0.6036900691688061,
      "epoch": 1.7895986797085697,
      "grad_norm": 0.15737776458263397,
      "learning_rate": 4.035023154537369e-05,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8184571117162704,
      "num_tokens": 620826848.0,
      "step": 19451
    },
    {
      "entropy": 0.6006881799548864,
      "epoch": 1.789690686080011,
      "grad_norm": 0.15701015293598175,
      "learning_rate": 4.034716471923207e-05,
      "loss": 0.5845,
      "mean_token_accuracy": 0.8209408447146416,
      "num_tokens": 620858945.0,
      "step": 19452
    },
    {
      "entropy": 0.6248437976464629,
      "epoch": 1.7897826924514524,
      "grad_norm": 0.15242330729961395,
      "learning_rate": 4.034409789309044e-05,
      "loss": 0.6144,
      "mean_token_accuracy": 0.8103888779878616,
      "num_tokens": 620891712.0,
      "step": 19453
    },
    {
      "entropy": 0.5925332987681031,
      "epoch": 1.7898746988228935,
      "grad_norm": 0.14907628297805786,
      "learning_rate": 4.034103106694882e-05,
      "loss": 0.5603,
      "mean_token_accuracy": 0.8250542394816875,
      "num_tokens": 620923294.0,
      "step": 19454
    },
    {
      "entropy": 0.5393299893476069,
      "epoch": 1.7899667051943347,
      "grad_norm": 0.1519044190645218,
      "learning_rate": 4.033796424080719e-05,
      "loss": 0.536,
      "mean_token_accuracy": 0.8358015194535255,
      "num_tokens": 620955252.0,
      "step": 19455
    },
    {
      "entropy": 0.6372549964580685,
      "epoch": 1.7900587115657758,
      "grad_norm": 0.15757311880588531,
      "learning_rate": 4.033489741466556e-05,
      "loss": 0.6245,
      "mean_token_accuracy": 0.8090530224144459,
      "num_tokens": 620986903.0,
      "step": 19456
    },
    {
      "entropy": 0.5225949126761407,
      "epoch": 1.7901507179372171,
      "grad_norm": 0.14893995225429535,
      "learning_rate": 4.033183058852394e-05,
      "loss": 0.5162,
      "mean_token_accuracy": 0.839904673397541,
      "num_tokens": 621019092.0,
      "step": 19457
    },
    {
      "entropy": 0.6061562481336296,
      "epoch": 1.7902427243086585,
      "grad_norm": 0.1559334397315979,
      "learning_rate": 4.032876376238231e-05,
      "loss": 0.6036,
      "mean_token_accuracy": 0.8192960172891617,
      "num_tokens": 621050187.0,
      "step": 19458
    },
    {
      "entropy": 0.7035151515156031,
      "epoch": 1.7903347306800996,
      "grad_norm": 0.15746861696243286,
      "learning_rate": 4.0325696936240683e-05,
      "loss": 0.7062,
      "mean_token_accuracy": 0.7857277728617191,
      "num_tokens": 621082181.0,
      "step": 19459
    },
    {
      "entropy": 0.6163928527384996,
      "epoch": 1.7904267370515408,
      "grad_norm": 0.15396158397197723,
      "learning_rate": 4.032263011009906e-05,
      "loss": 0.6031,
      "mean_token_accuracy": 0.8099206686019897,
      "num_tokens": 621114457.0,
      "step": 19460
    },
    {
      "entropy": 0.6267971880733967,
      "epoch": 1.790518743422982,
      "grad_norm": 0.15260356664657593,
      "learning_rate": 4.031956328395743e-05,
      "loss": 0.6255,
      "mean_token_accuracy": 0.8056301921606064,
      "num_tokens": 621147225.0,
      "step": 19461
    },
    {
      "entropy": 0.5720449201762676,
      "epoch": 1.7906107497944233,
      "grad_norm": 0.14951016008853912,
      "learning_rate": 4.0316496457815814e-05,
      "loss": 0.5651,
      "mean_token_accuracy": 0.8263776265084743,
      "num_tokens": 621179288.0,
      "step": 19462
    },
    {
      "entropy": 0.5738233011215925,
      "epoch": 1.7907027561658646,
      "grad_norm": 0.15415675938129425,
      "learning_rate": 4.031342963167418e-05,
      "loss": 0.5797,
      "mean_token_accuracy": 0.820650115609169,
      "num_tokens": 621210832.0,
      "step": 19463
    },
    {
      "entropy": 0.5898313606157899,
      "epoch": 1.7907947625373057,
      "grad_norm": 0.15342725813388824,
      "learning_rate": 4.031036280553256e-05,
      "loss": 0.5833,
      "mean_token_accuracy": 0.8224771805107594,
      "num_tokens": 621243154.0,
      "step": 19464
    },
    {
      "entropy": 0.5830861236900091,
      "epoch": 1.7908867689087469,
      "grad_norm": 0.1590854823589325,
      "learning_rate": 4.030729597939093e-05,
      "loss": 0.5759,
      "mean_token_accuracy": 0.824543721973896,
      "num_tokens": 621275068.0,
      "step": 19465
    },
    {
      "entropy": 0.456204011105001,
      "epoch": 1.790978775280188,
      "grad_norm": 0.13714174926280975,
      "learning_rate": 4.0304229153249306e-05,
      "loss": 0.4443,
      "mean_token_accuracy": 0.8598401136696339,
      "num_tokens": 621307319.0,
      "step": 19466
    },
    {
      "entropy": 0.63907206710428,
      "epoch": 1.7910707816516294,
      "grad_norm": 0.15715263783931732,
      "learning_rate": 4.030116232710768e-05,
      "loss": 0.6235,
      "mean_token_accuracy": 0.808080580085516,
      "num_tokens": 621338827.0,
      "step": 19467
    },
    {
      "entropy": 0.6238085869699717,
      "epoch": 1.7911627880230707,
      "grad_norm": 0.15826797485351562,
      "learning_rate": 4.0298095500966055e-05,
      "loss": 0.6224,
      "mean_token_accuracy": 0.8103451505303383,
      "num_tokens": 621371063.0,
      "step": 19468
    },
    {
      "entropy": 0.5610912637785077,
      "epoch": 1.7912547943945119,
      "grad_norm": 0.1571831852197647,
      "learning_rate": 4.029502867482442e-05,
      "loss": 0.5478,
      "mean_token_accuracy": 0.8277348130941391,
      "num_tokens": 621402979.0,
      "step": 19469
    },
    {
      "entropy": 0.651059371419251,
      "epoch": 1.791346800765953,
      "grad_norm": 0.15669922530651093,
      "learning_rate": 4.0291961848682805e-05,
      "loss": 0.6286,
      "mean_token_accuracy": 0.804577324539423,
      "num_tokens": 621434424.0,
      "step": 19470
    },
    {
      "entropy": 0.6182642299681902,
      "epoch": 1.7914388071373941,
      "grad_norm": 0.15247875452041626,
      "learning_rate": 4.028889502254117e-05,
      "loss": 0.6095,
      "mean_token_accuracy": 0.8146678060293198,
      "num_tokens": 621467192.0,
      "step": 19471
    },
    {
      "entropy": 0.6936570070683956,
      "epoch": 1.7915308135088355,
      "grad_norm": 0.1541554033756256,
      "learning_rate": 4.028582819639955e-05,
      "loss": 0.6757,
      "mean_token_accuracy": 0.7946087121963501,
      "num_tokens": 621499480.0,
      "step": 19472
    },
    {
      "entropy": 0.5593642294406891,
      "epoch": 1.7916228198802768,
      "grad_norm": 0.15181730687618256,
      "learning_rate": 4.028276137025792e-05,
      "loss": 0.5554,
      "mean_token_accuracy": 0.8281313218176365,
      "num_tokens": 621531268.0,
      "step": 19473
    },
    {
      "entropy": 0.5841260384768248,
      "epoch": 1.791714826251718,
      "grad_norm": 0.15769830346107483,
      "learning_rate": 4.0279694544116296e-05,
      "loss": 0.5766,
      "mean_token_accuracy": 0.8223591521382332,
      "num_tokens": 621562718.0,
      "step": 19474
    },
    {
      "entropy": 0.654976848512888,
      "epoch": 1.791806832623159,
      "grad_norm": 0.1520797312259674,
      "learning_rate": 4.027662771797467e-05,
      "loss": 0.6465,
      "mean_token_accuracy": 0.7994986772537231,
      "num_tokens": 621594610.0,
      "step": 19475
    },
    {
      "entropy": 0.6221233308315277,
      "epoch": 1.7918988389946002,
      "grad_norm": 0.15475428104400635,
      "learning_rate": 4.0273560891833046e-05,
      "loss": 0.6141,
      "mean_token_accuracy": 0.8068985976278782,
      "num_tokens": 621626280.0,
      "step": 19476
    },
    {
      "entropy": 0.7037905361503363,
      "epoch": 1.7919908453660416,
      "grad_norm": 0.15949377417564392,
      "learning_rate": 4.0270494065691414e-05,
      "loss": 0.7068,
      "mean_token_accuracy": 0.785997923463583,
      "num_tokens": 621658170.0,
      "step": 19477
    },
    {
      "entropy": 0.6075007813051343,
      "epoch": 1.792082851737483,
      "grad_norm": 0.15365634858608246,
      "learning_rate": 4.0267427239549795e-05,
      "loss": 0.6079,
      "mean_token_accuracy": 0.8105350323021412,
      "num_tokens": 621690037.0,
      "step": 19478
    },
    {
      "entropy": 0.6726132519543171,
      "epoch": 1.792174858108924,
      "grad_norm": 0.16063259541988373,
      "learning_rate": 4.026436041340816e-05,
      "loss": 0.662,
      "mean_token_accuracy": 0.8006707206368446,
      "num_tokens": 621722520.0,
      "step": 19479
    },
    {
      "entropy": 0.45116962864995,
      "epoch": 1.7922668644803652,
      "grad_norm": 0.13770973682403564,
      "learning_rate": 4.026129358726654e-05,
      "loss": 0.4348,
      "mean_token_accuracy": 0.8645273633301258,
      "num_tokens": 621755288.0,
      "step": 19480
    },
    {
      "entropy": 0.5853784345090389,
      "epoch": 1.7923588708518063,
      "grad_norm": 0.15004003047943115,
      "learning_rate": 4.025822676112491e-05,
      "loss": 0.5872,
      "mean_token_accuracy": 0.8180258795619011,
      "num_tokens": 621787098.0,
      "step": 19481
    },
    {
      "entropy": 0.6093548089265823,
      "epoch": 1.7924508772232477,
      "grad_norm": 0.15602749586105347,
      "learning_rate": 4.025515993498329e-05,
      "loss": 0.5991,
      "mean_token_accuracy": 0.8112314529716969,
      "num_tokens": 621818411.0,
      "step": 19482
    },
    {
      "entropy": 0.6275549060665071,
      "epoch": 1.792542883594689,
      "grad_norm": 0.15257933735847473,
      "learning_rate": 4.025209310884166e-05,
      "loss": 0.6194,
      "mean_token_accuracy": 0.8092596456408501,
      "num_tokens": 621850237.0,
      "step": 19483
    },
    {
      "entropy": 0.6022745054215193,
      "epoch": 1.7926348899661302,
      "grad_norm": 0.15321297943592072,
      "learning_rate": 4.0249026282700036e-05,
      "loss": 0.5908,
      "mean_token_accuracy": 0.8193902000784874,
      "num_tokens": 621882311.0,
      "step": 19484
    },
    {
      "entropy": 0.60987897682935,
      "epoch": 1.7927268963375713,
      "grad_norm": 0.1660332977771759,
      "learning_rate": 4.024595945655841e-05,
      "loss": 0.6009,
      "mean_token_accuracy": 0.8153146579861641,
      "num_tokens": 621914278.0,
      "step": 19485
    },
    {
      "entropy": 0.5744119007140398,
      "epoch": 1.7928189027090125,
      "grad_norm": 0.14763037860393524,
      "learning_rate": 4.0242892630416785e-05,
      "loss": 0.567,
      "mean_token_accuracy": 0.8268242925405502,
      "num_tokens": 621946272.0,
      "step": 19486
    },
    {
      "entropy": 0.6051635248586535,
      "epoch": 1.7929109090804538,
      "grad_norm": 0.15212321281433105,
      "learning_rate": 4.023982580427515e-05,
      "loss": 0.5943,
      "mean_token_accuracy": 0.8149620220065117,
      "num_tokens": 621978428.0,
      "step": 19487
    },
    {
      "entropy": 0.637501978315413,
      "epoch": 1.7930029154518952,
      "grad_norm": 0.16326875984668732,
      "learning_rate": 4.0236758978133535e-05,
      "loss": 0.6314,
      "mean_token_accuracy": 0.8086782470345497,
      "num_tokens": 622010484.0,
      "step": 19488
    },
    {
      "entropy": 0.6102305371314287,
      "epoch": 1.7930949218233363,
      "grad_norm": 0.1519545614719391,
      "learning_rate": 4.02336921519919e-05,
      "loss": 0.5992,
      "mean_token_accuracy": 0.8130411729216576,
      "num_tokens": 622042898.0,
      "step": 19489
    },
    {
      "entropy": 0.5158922951668501,
      "epoch": 1.7931869281947774,
      "grad_norm": 0.15026021003723145,
      "learning_rate": 4.023062532585028e-05,
      "loss": 0.5004,
      "mean_token_accuracy": 0.8450391963124275,
      "num_tokens": 622074792.0,
      "step": 19490
    },
    {
      "entropy": 0.6019369077403098,
      "epoch": 1.7932789345662186,
      "grad_norm": 0.16132742166519165,
      "learning_rate": 4.022755849970865e-05,
      "loss": 0.593,
      "mean_token_accuracy": 0.8187772147357464,
      "num_tokens": 622106491.0,
      "step": 19491
    },
    {
      "entropy": 0.6293381508439779,
      "epoch": 1.79337094093766,
      "grad_norm": 0.15546965599060059,
      "learning_rate": 4.0224491673567027e-05,
      "loss": 0.6194,
      "mean_token_accuracy": 0.8079463206231594,
      "num_tokens": 622138610.0,
      "step": 19492
    },
    {
      "entropy": 0.5480932351201773,
      "epoch": 1.7934629473091013,
      "grad_norm": 0.14958727359771729,
      "learning_rate": 4.02214248474254e-05,
      "loss": 0.5418,
      "mean_token_accuracy": 0.831038661301136,
      "num_tokens": 622170251.0,
      "step": 19493
    },
    {
      "entropy": 0.6709445975720882,
      "epoch": 1.7935549536805424,
      "grad_norm": 0.16016080975532532,
      "learning_rate": 4.0218358021283776e-05,
      "loss": 0.6562,
      "mean_token_accuracy": 0.8002257235348225,
      "num_tokens": 622202525.0,
      "step": 19494
    },
    {
      "entropy": 0.48444786481559277,
      "epoch": 1.7936469600519835,
      "grad_norm": 0.15457649528980255,
      "learning_rate": 4.021529119514215e-05,
      "loss": 0.4769,
      "mean_token_accuracy": 0.8529354967176914,
      "num_tokens": 622234412.0,
      "step": 19495
    },
    {
      "entropy": 0.6157173756510019,
      "epoch": 1.7937389664234247,
      "grad_norm": 0.15662658214569092,
      "learning_rate": 4.0212224369000525e-05,
      "loss": 0.6058,
      "mean_token_accuracy": 0.8149858303368092,
      "num_tokens": 622266748.0,
      "step": 19496
    },
    {
      "entropy": 0.5890843509696424,
      "epoch": 1.793830972794866,
      "grad_norm": 0.14860734343528748,
      "learning_rate": 4.02091575428589e-05,
      "loss": 0.5609,
      "mean_token_accuracy": 0.8242240399122238,
      "num_tokens": 622297816.0,
      "step": 19497
    },
    {
      "entropy": 0.6507026311010122,
      "epoch": 1.7939229791663074,
      "grad_norm": 0.15712712705135345,
      "learning_rate": 4.020609071671727e-05,
      "loss": 0.6357,
      "mean_token_accuracy": 0.8015968538820744,
      "num_tokens": 622329528.0,
      "step": 19498
    },
    {
      "entropy": 0.6665265751071274,
      "epoch": 1.7940149855377485,
      "grad_norm": 0.15882140398025513,
      "learning_rate": 4.020302389057565e-05,
      "loss": 0.6576,
      "mean_token_accuracy": 0.7970383204519749,
      "num_tokens": 622361047.0,
      "step": 19499
    },
    {
      "entropy": 0.6103649884462357,
      "epoch": 1.7941069919091897,
      "grad_norm": 0.15559156239032745,
      "learning_rate": 4.019995706443402e-05,
      "loss": 0.5955,
      "mean_token_accuracy": 0.816752701997757,
      "num_tokens": 622392858.0,
      "step": 19500
    },
    {
      "entropy": 0.5051949890330434,
      "epoch": 1.7941989982806308,
      "grad_norm": 0.1499263495206833,
      "learning_rate": 4.01968902382924e-05,
      "loss": 0.4949,
      "mean_token_accuracy": 0.851163387298584,
      "num_tokens": 622425356.0,
      "step": 19501
    },
    {
      "entropy": 0.5630968566983938,
      "epoch": 1.7942910046520721,
      "grad_norm": 0.15049003064632416,
      "learning_rate": 4.0193823412150766e-05,
      "loss": 0.5582,
      "mean_token_accuracy": 0.8247440978884697,
      "num_tokens": 622457476.0,
      "step": 19502
    },
    {
      "entropy": 0.5867312354966998,
      "epoch": 1.7943830110235135,
      "grad_norm": 0.14196181297302246,
      "learning_rate": 4.019075658600914e-05,
      "loss": 0.5751,
      "mean_token_accuracy": 0.8218698091804981,
      "num_tokens": 622489891.0,
      "step": 19503
    },
    {
      "entropy": 0.6096700420603156,
      "epoch": 1.7944750173949546,
      "grad_norm": 0.15045206248760223,
      "learning_rate": 4.0187689759867515e-05,
      "loss": 0.5925,
      "mean_token_accuracy": 0.8213077858090401,
      "num_tokens": 622522608.0,
      "step": 19504
    },
    {
      "entropy": 0.5824045017361641,
      "epoch": 1.7945670237663958,
      "grad_norm": 0.15423256158828735,
      "learning_rate": 4.018462293372589e-05,
      "loss": 0.5764,
      "mean_token_accuracy": 0.8228941820561886,
      "num_tokens": 622555027.0,
      "step": 19505
    },
    {
      "entropy": 0.6184666287153959,
      "epoch": 1.794659030137837,
      "grad_norm": 0.15644297003746033,
      "learning_rate": 4.0181556107584265e-05,
      "loss": 0.6205,
      "mean_token_accuracy": 0.8115994557738304,
      "num_tokens": 622586602.0,
      "step": 19506
    },
    {
      "entropy": 0.45212111016735435,
      "epoch": 1.7947510365092783,
      "grad_norm": 0.1459340751171112,
      "learning_rate": 4.017848928144264e-05,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8608143404126167,
      "num_tokens": 622618653.0,
      "step": 19507
    },
    {
      "entropy": 0.579104833304882,
      "epoch": 1.7948430428807196,
      "grad_norm": 0.15869146585464478,
      "learning_rate": 4.017542245530101e-05,
      "loss": 0.5724,
      "mean_token_accuracy": 0.8214460611343384,
      "num_tokens": 622649552.0,
      "step": 19508
    },
    {
      "entropy": 0.529763001948595,
      "epoch": 1.7949350492521607,
      "grad_norm": 0.14588546752929688,
      "learning_rate": 4.017235562915939e-05,
      "loss": 0.5293,
      "mean_token_accuracy": 0.8373063169419765,
      "num_tokens": 622682256.0,
      "step": 19509
    },
    {
      "entropy": 0.5169894807040691,
      "epoch": 1.7950270556236019,
      "grad_norm": 0.14754422008991241,
      "learning_rate": 4.0169288803017757e-05,
      "loss": 0.5232,
      "mean_token_accuracy": 0.8412900120019913,
      "num_tokens": 622713255.0,
      "step": 19510
    },
    {
      "entropy": 0.5480415411293507,
      "epoch": 1.795119061995043,
      "grad_norm": 0.14842350780963898,
      "learning_rate": 4.016622197687613e-05,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8358223177492619,
      "num_tokens": 622745818.0,
      "step": 19511
    },
    {
      "entropy": 0.5596487345173955,
      "epoch": 1.7952110683664844,
      "grad_norm": 0.14585058391094208,
      "learning_rate": 4.0163155150734506e-05,
      "loss": 0.5377,
      "mean_token_accuracy": 0.829110499471426,
      "num_tokens": 622777651.0,
      "step": 19512
    },
    {
      "entropy": 0.6515575461089611,
      "epoch": 1.7953030747379257,
      "grad_norm": 0.16086195409297943,
      "learning_rate": 4.016008832459288e-05,
      "loss": 0.6491,
      "mean_token_accuracy": 0.802736297249794,
      "num_tokens": 622809940.0,
      "step": 19513
    },
    {
      "entropy": 0.6578889284282923,
      "epoch": 1.7953950811093669,
      "grad_norm": 0.1594780683517456,
      "learning_rate": 4.0157021498451255e-05,
      "loss": 0.6612,
      "mean_token_accuracy": 0.7950654067099094,
      "num_tokens": 622841811.0,
      "step": 19514
    },
    {
      "entropy": 0.6419019494205713,
      "epoch": 1.795487087480808,
      "grad_norm": 0.15245220065116882,
      "learning_rate": 4.015395467230963e-05,
      "loss": 0.6348,
      "mean_token_accuracy": 0.8078682720661163,
      "num_tokens": 622873732.0,
      "step": 19515
    },
    {
      "entropy": 0.59763473412022,
      "epoch": 1.7955790938522491,
      "grad_norm": 0.1517390012741089,
      "learning_rate": 4.0150887846168e-05,
      "loss": 0.5898,
      "mean_token_accuracy": 0.818672701716423,
      "num_tokens": 622906072.0,
      "step": 19516
    },
    {
      "entropy": 0.5901925228536129,
      "epoch": 1.7956711002236905,
      "grad_norm": 0.1500285416841507,
      "learning_rate": 4.014782102002638e-05,
      "loss": 0.5842,
      "mean_token_accuracy": 0.8205366358160973,
      "num_tokens": 622938378.0,
      "step": 19517
    },
    {
      "entropy": 0.5954364156350493,
      "epoch": 1.7957631065951318,
      "grad_norm": 0.15823689103126526,
      "learning_rate": 4.014475419388475e-05,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8188992068171501,
      "num_tokens": 622969327.0,
      "step": 19518
    },
    {
      "entropy": 0.6648181863129139,
      "epoch": 1.795855112966573,
      "grad_norm": 0.1549546718597412,
      "learning_rate": 4.014168736774312e-05,
      "loss": 0.6513,
      "mean_token_accuracy": 0.7998059950768948,
      "num_tokens": 623000670.0,
      "step": 19519
    },
    {
      "entropy": 0.6245649319607764,
      "epoch": 1.795947119338014,
      "grad_norm": 0.15492267906665802,
      "learning_rate": 4.0138620541601496e-05,
      "loss": 0.6147,
      "mean_token_accuracy": 0.8171831928193569,
      "num_tokens": 623032306.0,
      "step": 19520
    },
    {
      "entropy": 0.6766307447105646,
      "epoch": 1.7960391257094552,
      "grad_norm": 0.15711866319179535,
      "learning_rate": 4.013555371545987e-05,
      "loss": 0.6626,
      "mean_token_accuracy": 0.7981351427733898,
      "num_tokens": 623064283.0,
      "step": 19521
    },
    {
      "entropy": 0.70171676017344,
      "epoch": 1.7961311320808966,
      "grad_norm": 0.15468916296958923,
      "learning_rate": 4.0132486889318246e-05,
      "loss": 0.6917,
      "mean_token_accuracy": 0.7913065813481808,
      "num_tokens": 623096388.0,
      "step": 19522
    },
    {
      "entropy": 0.5462312661111355,
      "epoch": 1.796223138452338,
      "grad_norm": 0.1558401882648468,
      "learning_rate": 4.012942006317662e-05,
      "loss": 0.5322,
      "mean_token_accuracy": 0.8364636972546577,
      "num_tokens": 623128162.0,
      "step": 19523
    },
    {
      "entropy": 0.5841459771618247,
      "epoch": 1.796315144823779,
      "grad_norm": 0.1566859483718872,
      "learning_rate": 4.0126353237034995e-05,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8203765116631985,
      "num_tokens": 623159224.0,
      "step": 19524
    },
    {
      "entropy": 0.5498985182493925,
      "epoch": 1.7964071511952202,
      "grad_norm": 0.15756972134113312,
      "learning_rate": 4.012328641089337e-05,
      "loss": 0.535,
      "mean_token_accuracy": 0.8370789773762226,
      "num_tokens": 623191102.0,
      "step": 19525
    },
    {
      "entropy": 0.689766176044941,
      "epoch": 1.7964991575666613,
      "grad_norm": 0.15963596105575562,
      "learning_rate": 4.0120219584751744e-05,
      "loss": 0.6628,
      "mean_token_accuracy": 0.7972214110195637,
      "num_tokens": 623223114.0,
      "step": 19526
    },
    {
      "entropy": 0.6486291522160172,
      "epoch": 1.7965911639381027,
      "grad_norm": 0.1528448909521103,
      "learning_rate": 4.011715275861012e-05,
      "loss": 0.632,
      "mean_token_accuracy": 0.8104747124016285,
      "num_tokens": 623255025.0,
      "step": 19527
    },
    {
      "entropy": 0.595500179566443,
      "epoch": 1.796683170309544,
      "grad_norm": 0.15483549237251282,
      "learning_rate": 4.0114085932468493e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8192808777093887,
      "num_tokens": 623286534.0,
      "step": 19528
    },
    {
      "entropy": 0.5576311042532325,
      "epoch": 1.7967751766809852,
      "grad_norm": 0.15372776985168457,
      "learning_rate": 4.011101910632686e-05,
      "loss": 0.5552,
      "mean_token_accuracy": 0.8292244598269463,
      "num_tokens": 623318249.0,
      "step": 19529
    },
    {
      "entropy": 0.5604456826113164,
      "epoch": 1.7968671830524263,
      "grad_norm": 0.15377432107925415,
      "learning_rate": 4.010795228018524e-05,
      "loss": 0.5473,
      "mean_token_accuracy": 0.8331066332757473,
      "num_tokens": 623349773.0,
      "step": 19530
    },
    {
      "entropy": 0.6162936883047223,
      "epoch": 1.7969591894238675,
      "grad_norm": 0.1522544026374817,
      "learning_rate": 4.010488545404361e-05,
      "loss": 0.6038,
      "mean_token_accuracy": 0.813172847032547,
      "num_tokens": 623381195.0,
      "step": 19531
    },
    {
      "entropy": 0.5568767786026001,
      "epoch": 1.7970511957953088,
      "grad_norm": 0.1435166597366333,
      "learning_rate": 4.0101818627901985e-05,
      "loss": 0.5428,
      "mean_token_accuracy": 0.834366500377655,
      "num_tokens": 623413716.0,
      "step": 19532
    },
    {
      "entropy": 0.591919285710901,
      "epoch": 1.7971432021667502,
      "grad_norm": 0.15734624862670898,
      "learning_rate": 4.009875180176036e-05,
      "loss": 0.5703,
      "mean_token_accuracy": 0.8255306705832481,
      "num_tokens": 623444861.0,
      "step": 19533
    },
    {
      "entropy": 0.5895804343745112,
      "epoch": 1.7972352085381913,
      "grad_norm": 0.1531468778848648,
      "learning_rate": 4.0095684975618735e-05,
      "loss": 0.5756,
      "mean_token_accuracy": 0.8253632225096226,
      "num_tokens": 623477479.0,
      "step": 19534
    },
    {
      "entropy": 0.6466648187488317,
      "epoch": 1.7973272149096324,
      "grad_norm": 0.15367832779884338,
      "learning_rate": 4.009261814947711e-05,
      "loss": 0.6472,
      "mean_token_accuracy": 0.7987400069832802,
      "num_tokens": 623510167.0,
      "step": 19535
    },
    {
      "entropy": 0.6649334635585546,
      "epoch": 1.7974192212810736,
      "grad_norm": 0.15452130138874054,
      "learning_rate": 4.0089551323335484e-05,
      "loss": 0.6501,
      "mean_token_accuracy": 0.8033564761281013,
      "num_tokens": 623542217.0,
      "step": 19536
    },
    {
      "entropy": 0.6664229780435562,
      "epoch": 1.797511227652515,
      "grad_norm": 0.15730144083499908,
      "learning_rate": 4.008648449719385e-05,
      "loss": 0.6584,
      "mean_token_accuracy": 0.7985906712710857,
      "num_tokens": 623574532.0,
      "step": 19537
    },
    {
      "entropy": 0.6119575090706348,
      "epoch": 1.7976032340239563,
      "grad_norm": 0.15441080927848816,
      "learning_rate": 4.008341767105223e-05,
      "loss": 0.6039,
      "mean_token_accuracy": 0.8160217180848122,
      "num_tokens": 623606069.0,
      "step": 19538
    },
    {
      "entropy": 0.5962635865435004,
      "epoch": 1.7976952403953974,
      "grad_norm": 0.15536567568778992,
      "learning_rate": 4.00803508449106e-05,
      "loss": 0.599,
      "mean_token_accuracy": 0.8134813271462917,
      "num_tokens": 623637779.0,
      "step": 19539
    },
    {
      "entropy": 0.6530426423996687,
      "epoch": 1.7977872467668385,
      "grad_norm": 0.1716586947441101,
      "learning_rate": 4.0077284018768976e-05,
      "loss": 0.6513,
      "mean_token_accuracy": 0.8025504164397717,
      "num_tokens": 623669989.0,
      "step": 19540
    },
    {
      "entropy": 0.5111966263502836,
      "epoch": 1.7978792531382797,
      "grad_norm": 0.15230777859687805,
      "learning_rate": 4.007421719262735e-05,
      "loss": 0.507,
      "mean_token_accuracy": 0.8405850790441036,
      "num_tokens": 623702265.0,
      "step": 19541
    },
    {
      "entropy": 0.5346921239979565,
      "epoch": 1.797971259509721,
      "grad_norm": 0.15105454623699188,
      "learning_rate": 4.0071150366485725e-05,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8371927738189697,
      "num_tokens": 623734247.0,
      "step": 19542
    },
    {
      "entropy": 0.58688735216856,
      "epoch": 1.7980632658811624,
      "grad_norm": 0.15754690766334534,
      "learning_rate": 4.00680835403441e-05,
      "loss": 0.5861,
      "mean_token_accuracy": 0.8205020651221275,
      "num_tokens": 623765404.0,
      "step": 19543
    },
    {
      "entropy": 0.6710375840775669,
      "epoch": 1.7981552722526035,
      "grad_norm": 0.15963256359100342,
      "learning_rate": 4.0065016714202474e-05,
      "loss": 0.6607,
      "mean_token_accuracy": 0.8012924641370773,
      "num_tokens": 623797351.0,
      "step": 19544
    },
    {
      "entropy": 0.5998558579012752,
      "epoch": 1.7982472786240447,
      "grad_norm": 0.1592186689376831,
      "learning_rate": 4.006194988806085e-05,
      "loss": 0.5989,
      "mean_token_accuracy": 0.816078033298254,
      "num_tokens": 623827714.0,
      "step": 19545
    },
    {
      "entropy": 0.6383824832737446,
      "epoch": 1.7983392849954858,
      "grad_norm": 0.1519862413406372,
      "learning_rate": 4.0058883061919224e-05,
      "loss": 0.639,
      "mean_token_accuracy": 0.8062646389007568,
      "num_tokens": 623859977.0,
      "step": 19546
    },
    {
      "entropy": 0.7073239823803306,
      "epoch": 1.7984312913669271,
      "grad_norm": 0.16713890433311462,
      "learning_rate": 4.005581623577759e-05,
      "loss": 0.7106,
      "mean_token_accuracy": 0.7843986079096794,
      "num_tokens": 623891142.0,
      "step": 19547
    },
    {
      "entropy": 0.5872738808393478,
      "epoch": 1.7985232977383685,
      "grad_norm": 0.15411150455474854,
      "learning_rate": 4.005274940963597e-05,
      "loss": 0.5572,
      "mean_token_accuracy": 0.8269380554556847,
      "num_tokens": 623922334.0,
      "step": 19548
    },
    {
      "entropy": 0.5297822570428252,
      "epoch": 1.7986153041098096,
      "grad_norm": 0.14701960980892181,
      "learning_rate": 4.004968258349434e-05,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8387752585113049,
      "num_tokens": 623954740.0,
      "step": 19549
    },
    {
      "entropy": 0.6181171170901507,
      "epoch": 1.7987073104812508,
      "grad_norm": 0.15522141754627228,
      "learning_rate": 4.0046615757352715e-05,
      "loss": 0.598,
      "mean_token_accuracy": 0.8132783956825733,
      "num_tokens": 623985854.0,
      "step": 19550
    },
    {
      "entropy": 0.6166265909560025,
      "epoch": 1.798799316852692,
      "grad_norm": 0.15387164056301117,
      "learning_rate": 4.004354893121109e-05,
      "loss": 0.6192,
      "mean_token_accuracy": 0.8128038868308067,
      "num_tokens": 624016820.0,
      "step": 19551
    },
    {
      "entropy": 0.6101682335138321,
      "epoch": 1.7988913232241333,
      "grad_norm": 0.15374116599559784,
      "learning_rate": 4.0040482105069465e-05,
      "loss": 0.6069,
      "mean_token_accuracy": 0.8151930719614029,
      "num_tokens": 624048772.0,
      "step": 19552
    },
    {
      "entropy": 0.5922936163842678,
      "epoch": 1.7989833295955746,
      "grad_norm": 0.1467006504535675,
      "learning_rate": 4.003741527892784e-05,
      "loss": 0.5714,
      "mean_token_accuracy": 0.8240376338362694,
      "num_tokens": 624080407.0,
      "step": 19553
    },
    {
      "entropy": 0.6641650348901749,
      "epoch": 1.7990753359670157,
      "grad_norm": 0.15162305533885956,
      "learning_rate": 4.0034348452786214e-05,
      "loss": 0.6419,
      "mean_token_accuracy": 0.8025067076086998,
      "num_tokens": 624112567.0,
      "step": 19554
    },
    {
      "entropy": 0.5860179262235761,
      "epoch": 1.7991673423384569,
      "grad_norm": 0.15514029562473297,
      "learning_rate": 4.003128162664459e-05,
      "loss": 0.5782,
      "mean_token_accuracy": 0.8266749270260334,
      "num_tokens": 624144430.0,
      "step": 19555
    },
    {
      "entropy": 0.6759459618479013,
      "epoch": 1.799259348709898,
      "grad_norm": 0.16276268661022186,
      "learning_rate": 4.002821480050296e-05,
      "loss": 0.6571,
      "mean_token_accuracy": 0.7985553219914436,
      "num_tokens": 624176776.0,
      "step": 19556
    },
    {
      "entropy": 0.6247066939249635,
      "epoch": 1.7993513550813394,
      "grad_norm": 0.15223991870880127,
      "learning_rate": 4.002514797436134e-05,
      "loss": 0.6101,
      "mean_token_accuracy": 0.8117533102631569,
      "num_tokens": 624209072.0,
      "step": 19557
    },
    {
      "entropy": 0.6241557616740465,
      "epoch": 1.7994433614527807,
      "grad_norm": 0.1531524360179901,
      "learning_rate": 4.0022081148219706e-05,
      "loss": 0.6207,
      "mean_token_accuracy": 0.8108379766345024,
      "num_tokens": 624241070.0,
      "step": 19558
    },
    {
      "entropy": 0.6138408733531833,
      "epoch": 1.7995353678242219,
      "grad_norm": 0.1580529808998108,
      "learning_rate": 4.001901432207809e-05,
      "loss": 0.6009,
      "mean_token_accuracy": 0.8103660680353642,
      "num_tokens": 624272367.0,
      "step": 19559
    },
    {
      "entropy": 0.6487562209367752,
      "epoch": 1.799627374195663,
      "grad_norm": 0.1605759710073471,
      "learning_rate": 4.0015947495936455e-05,
      "loss": 0.6468,
      "mean_token_accuracy": 0.8023499175906181,
      "num_tokens": 624303978.0,
      "step": 19560
    },
    {
      "entropy": 0.6602863389998674,
      "epoch": 1.7997193805671041,
      "grad_norm": 0.15995705127716064,
      "learning_rate": 4.0012880669794837e-05,
      "loss": 0.6393,
      "mean_token_accuracy": 0.8037018924951553,
      "num_tokens": 624335034.0,
      "step": 19561
    },
    {
      "entropy": 0.5130288684740663,
      "epoch": 1.7998113869385455,
      "grad_norm": 0.14730097353458405,
      "learning_rate": 4.0009813843653204e-05,
      "loss": 0.509,
      "mean_token_accuracy": 0.840821273624897,
      "num_tokens": 624366800.0,
      "step": 19562
    },
    {
      "entropy": 0.6663315868936479,
      "epoch": 1.7999033933099868,
      "grad_norm": 0.16402751207351685,
      "learning_rate": 4.000674701751158e-05,
      "loss": 0.6535,
      "mean_token_accuracy": 0.8024434149265289,
      "num_tokens": 624398316.0,
      "step": 19563
    },
    {
      "entropy": 0.47702168207615614,
      "epoch": 1.799995399681428,
      "grad_norm": 0.14356392621994019,
      "learning_rate": 4.0003680191369954e-05,
      "loss": 0.463,
      "mean_token_accuracy": 0.8531092144548893,
      "num_tokens": 624429714.0,
      "step": 19564
    },
    {
      "entropy": 0.7499855384230614,
      "epoch": 1.800087406052869,
      "grad_norm": 0.16554418206214905,
      "learning_rate": 4.000061336522833e-05,
      "loss": 0.7454,
      "mean_token_accuracy": 0.7726123705506325,
      "num_tokens": 624461759.0,
      "step": 19565
    },
    {
      "entropy": 0.5974156931042671,
      "epoch": 1.8001794124243102,
      "grad_norm": 0.15859773755073547,
      "learning_rate": 3.99975465390867e-05,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8180508054792881,
      "num_tokens": 624493704.0,
      "step": 19566
    },
    {
      "entropy": 0.6127894958481193,
      "epoch": 1.8002714187957516,
      "grad_norm": 0.15334713459014893,
      "learning_rate": 3.999447971294508e-05,
      "loss": 0.597,
      "mean_token_accuracy": 0.8160706274211407,
      "num_tokens": 624525523.0,
      "step": 19567
    },
    {
      "entropy": 0.6276132240891457,
      "epoch": 1.800363425167193,
      "grad_norm": 0.15133659541606903,
      "learning_rate": 3.9991412886803446e-05,
      "loss": 0.6202,
      "mean_token_accuracy": 0.8110663443803787,
      "num_tokens": 624557413.0,
      "step": 19568
    },
    {
      "entropy": 0.5487578206229955,
      "epoch": 1.800455431538634,
      "grad_norm": 0.1510002613067627,
      "learning_rate": 3.998834606066183e-05,
      "loss": 0.5197,
      "mean_token_accuracy": 0.8350774087011814,
      "num_tokens": 624588654.0,
      "step": 19569
    },
    {
      "entropy": 0.6931815007701516,
      "epoch": 1.8005474379100752,
      "grad_norm": 0.16227412223815918,
      "learning_rate": 3.9985279234520195e-05,
      "loss": 0.6853,
      "mean_token_accuracy": 0.788811270147562,
      "num_tokens": 624619964.0,
      "step": 19570
    },
    {
      "entropy": 0.5450196284800768,
      "epoch": 1.8006394442815163,
      "grad_norm": 0.15176525712013245,
      "learning_rate": 3.998221240837857e-05,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8356047160923481,
      "num_tokens": 624652292.0,
      "step": 19571
    },
    {
      "entropy": 0.663514968007803,
      "epoch": 1.8007314506529577,
      "grad_norm": 0.15826372802257538,
      "learning_rate": 3.9979145582236944e-05,
      "loss": 0.6491,
      "mean_token_accuracy": 0.8020415343344212,
      "num_tokens": 624684101.0,
      "step": 19572
    },
    {
      "entropy": 0.597786265425384,
      "epoch": 1.800823457024399,
      "grad_norm": 0.16290423274040222,
      "learning_rate": 3.997607875609532e-05,
      "loss": 0.5823,
      "mean_token_accuracy": 0.8193294778466225,
      "num_tokens": 624715986.0,
      "step": 19573
    },
    {
      "entropy": 0.6754504726268351,
      "epoch": 1.8009154633958402,
      "grad_norm": 0.15745694935321808,
      "learning_rate": 3.9973011929953693e-05,
      "loss": 0.668,
      "mean_token_accuracy": 0.7902017086744308,
      "num_tokens": 624748071.0,
      "step": 19574
    },
    {
      "entropy": 0.6874187253415585,
      "epoch": 1.8010074697672813,
      "grad_norm": 0.16365402936935425,
      "learning_rate": 3.996994510381207e-05,
      "loss": 0.6828,
      "mean_token_accuracy": 0.79768330976367,
      "num_tokens": 624779732.0,
      "step": 19575
    },
    {
      "entropy": 0.5298658832907677,
      "epoch": 1.8010994761387225,
      "grad_norm": 0.1547670215368271,
      "learning_rate": 3.9966878277670436e-05,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8373215571045876,
      "num_tokens": 624811001.0,
      "step": 19576
    },
    {
      "entropy": 0.5993571230210364,
      "epoch": 1.8011914825101638,
      "grad_norm": 0.14948680996894836,
      "learning_rate": 3.996381145152882e-05,
      "loss": 0.5988,
      "mean_token_accuracy": 0.8172869496047497,
      "num_tokens": 624843270.0,
      "step": 19577
    },
    {
      "entropy": 0.6802291015628725,
      "epoch": 1.8012834888816052,
      "grad_norm": 0.16141566634178162,
      "learning_rate": 3.9960744625387185e-05,
      "loss": 0.6676,
      "mean_token_accuracy": 0.7989247776567936,
      "num_tokens": 624874691.0,
      "step": 19578
    },
    {
      "entropy": 0.6192069500684738,
      "epoch": 1.8013754952530463,
      "grad_norm": 0.15775270760059357,
      "learning_rate": 3.995767779924556e-05,
      "loss": 0.6237,
      "mean_token_accuracy": 0.8089586496353149,
      "num_tokens": 624906806.0,
      "step": 19579
    },
    {
      "entropy": 0.5718890884891152,
      "epoch": 1.8014675016244874,
      "grad_norm": 0.15939141809940338,
      "learning_rate": 3.9954610973103935e-05,
      "loss": 0.5737,
      "mean_token_accuracy": 0.8279653638601303,
      "num_tokens": 624938077.0,
      "step": 19580
    },
    {
      "entropy": 0.6071142051368952,
      "epoch": 1.8015595079959286,
      "grad_norm": 0.16014541685581207,
      "learning_rate": 3.995154414696231e-05,
      "loss": 0.6099,
      "mean_token_accuracy": 0.8143065832555294,
      "num_tokens": 624969156.0,
      "step": 19581
    },
    {
      "entropy": 0.5630092835053802,
      "epoch": 1.80165151436737,
      "grad_norm": 0.14819765090942383,
      "learning_rate": 3.9948477320820684e-05,
      "loss": 0.5535,
      "mean_token_accuracy": 0.8330673314630985,
      "num_tokens": 625001180.0,
      "step": 19582
    },
    {
      "entropy": 0.7158947214484215,
      "epoch": 1.8017435207388113,
      "grad_norm": 0.16667088866233826,
      "learning_rate": 3.994541049467906e-05,
      "loss": 0.7183,
      "mean_token_accuracy": 0.7828992754220963,
      "num_tokens": 625031754.0,
      "step": 19583
    },
    {
      "entropy": 0.6095191389322281,
      "epoch": 1.8018355271102524,
      "grad_norm": 0.1529616415500641,
      "learning_rate": 3.994234366853743e-05,
      "loss": 0.5988,
      "mean_token_accuracy": 0.8185408152639866,
      "num_tokens": 625063777.0,
      "step": 19584
    },
    {
      "entropy": 0.672790015116334,
      "epoch": 1.8019275334816935,
      "grad_norm": 0.15958769619464874,
      "learning_rate": 3.993927684239581e-05,
      "loss": 0.6715,
      "mean_token_accuracy": 0.7940299957990646,
      "num_tokens": 625095369.0,
      "step": 19585
    },
    {
      "entropy": 0.6065562763251364,
      "epoch": 1.8020195398531347,
      "grad_norm": 0.15777720510959625,
      "learning_rate": 3.993621001625418e-05,
      "loss": 0.5866,
      "mean_token_accuracy": 0.8172905184328556,
      "num_tokens": 625126679.0,
      "step": 19586
    },
    {
      "entropy": 0.5015456438995898,
      "epoch": 1.802111546224576,
      "grad_norm": 0.1461149901151657,
      "learning_rate": 3.993314319011256e-05,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8493653498589993,
      "num_tokens": 625158668.0,
      "step": 19587
    },
    {
      "entropy": 0.539461100474,
      "epoch": 1.8022035525960174,
      "grad_norm": 0.14631956815719604,
      "learning_rate": 3.993007636397093e-05,
      "loss": 0.5218,
      "mean_token_accuracy": 0.8375903964042664,
      "num_tokens": 625190881.0,
      "step": 19588
    },
    {
      "entropy": 0.7147765159606934,
      "epoch": 1.8022955589674585,
      "grad_norm": 0.15791048109531403,
      "learning_rate": 3.99270095378293e-05,
      "loss": 0.6943,
      "mean_token_accuracy": 0.7893975786864758,
      "num_tokens": 625222819.0,
      "step": 19589
    },
    {
      "entropy": 0.6514865234494209,
      "epoch": 1.8023875653388997,
      "grad_norm": 0.16120189428329468,
      "learning_rate": 3.992394271168768e-05,
      "loss": 0.6406,
      "mean_token_accuracy": 0.8067074492573738,
      "num_tokens": 625254805.0,
      "step": 19590
    },
    {
      "entropy": 0.6615978926420212,
      "epoch": 1.8024795717103408,
      "grad_norm": 0.15905599296092987,
      "learning_rate": 3.992087588554605e-05,
      "loss": 0.6593,
      "mean_token_accuracy": 0.7991819120943546,
      "num_tokens": 625286812.0,
      "step": 19591
    },
    {
      "entropy": 0.5544223207980394,
      "epoch": 1.8025715780817821,
      "grad_norm": 0.14551375806331635,
      "learning_rate": 3.9917809059404423e-05,
      "loss": 0.533,
      "mean_token_accuracy": 0.8314010575413704,
      "num_tokens": 625319142.0,
      "step": 19592
    },
    {
      "entropy": 0.5182607774622738,
      "epoch": 1.8026635844532235,
      "grad_norm": 0.14131920039653778,
      "learning_rate": 3.99147422332628e-05,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8453205302357674,
      "num_tokens": 625351444.0,
      "step": 19593
    },
    {
      "entropy": 0.632507411763072,
      "epoch": 1.8027555908246646,
      "grad_norm": 0.15023396909236908,
      "learning_rate": 3.991167540712117e-05,
      "loss": 0.6263,
      "mean_token_accuracy": 0.8094142265617847,
      "num_tokens": 625383708.0,
      "step": 19594
    },
    {
      "entropy": 0.5880174990743399,
      "epoch": 1.8028475971961058,
      "grad_norm": 0.15325795114040375,
      "learning_rate": 3.990860858097955e-05,
      "loss": 0.5674,
      "mean_token_accuracy": 0.8239697925746441,
      "num_tokens": 625415330.0,
      "step": 19595
    },
    {
      "entropy": 0.6272688116878271,
      "epoch": 1.802939603567547,
      "grad_norm": 0.15488339960575104,
      "learning_rate": 3.990554175483792e-05,
      "loss": 0.6221,
      "mean_token_accuracy": 0.8081093318760395,
      "num_tokens": 625446731.0,
      "step": 19596
    },
    {
      "entropy": 0.5183239961043,
      "epoch": 1.8030316099389883,
      "grad_norm": 0.14913924038410187,
      "learning_rate": 3.990247492869629e-05,
      "loss": 0.5026,
      "mean_token_accuracy": 0.8456869311630726,
      "num_tokens": 625479263.0,
      "step": 19597
    },
    {
      "entropy": 0.6424841748084873,
      "epoch": 1.8031236163104296,
      "grad_norm": 0.15796853601932526,
      "learning_rate": 3.989940810255467e-05,
      "loss": 0.619,
      "mean_token_accuracy": 0.8080770559608936,
      "num_tokens": 625511116.0,
      "step": 19598
    },
    {
      "entropy": 0.675168564543128,
      "epoch": 1.8032156226818707,
      "grad_norm": 0.1622203290462494,
      "learning_rate": 3.989634127641304e-05,
      "loss": 0.6589,
      "mean_token_accuracy": 0.8006605543196201,
      "num_tokens": 625543081.0,
      "step": 19599
    },
    {
      "entropy": 0.7074645236134529,
      "epoch": 1.8033076290533119,
      "grad_norm": 0.1601330041885376,
      "learning_rate": 3.989327445027142e-05,
      "loss": 0.7161,
      "mean_token_accuracy": 0.7812425307929516,
      "num_tokens": 625575717.0,
      "step": 19600
    },
    {
      "entropy": 0.6560005284845829,
      "epoch": 1.803399635424753,
      "grad_norm": 0.16087397933006287,
      "learning_rate": 3.989020762412979e-05,
      "loss": 0.6545,
      "mean_token_accuracy": 0.8009996563196182,
      "num_tokens": 625607462.0,
      "step": 19601
    },
    {
      "entropy": 0.5757854152470827,
      "epoch": 1.8034916417961944,
      "grad_norm": 0.14905717968940735,
      "learning_rate": 3.988714079798816e-05,
      "loss": 0.5636,
      "mean_token_accuracy": 0.8219144977629185,
      "num_tokens": 625639008.0,
      "step": 19602
    },
    {
      "entropy": 0.5570674929767847,
      "epoch": 1.8035836481676357,
      "grad_norm": 0.14927050471305847,
      "learning_rate": 3.988407397184654e-05,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8310859613120556,
      "num_tokens": 625670771.0,
      "step": 19603
    },
    {
      "entropy": 0.554690039716661,
      "epoch": 1.8036756545390769,
      "grad_norm": 0.15286876261234283,
      "learning_rate": 3.988100714570491e-05,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8317662812769413,
      "num_tokens": 625702562.0,
      "step": 19604
    },
    {
      "entropy": 0.47013003495521843,
      "epoch": 1.803767660910518,
      "grad_norm": 0.14660879969596863,
      "learning_rate": 3.987794031956329e-05,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8628227412700653,
      "num_tokens": 625734809.0,
      "step": 19605
    },
    {
      "entropy": 0.6892109513282776,
      "epoch": 1.8038596672819591,
      "grad_norm": 0.16299954056739807,
      "learning_rate": 3.987487349342166e-05,
      "loss": 0.6911,
      "mean_token_accuracy": 0.7929520793259144,
      "num_tokens": 625767171.0,
      "step": 19606
    },
    {
      "entropy": 0.6643305495381355,
      "epoch": 1.8039516736534005,
      "grad_norm": 0.157846137881279,
      "learning_rate": 3.987180666728003e-05,
      "loss": 0.6567,
      "mean_token_accuracy": 0.8015414103865623,
      "num_tokens": 625799638.0,
      "step": 19607
    },
    {
      "entropy": 0.699280871078372,
      "epoch": 1.8040436800248418,
      "grad_norm": 0.15580326318740845,
      "learning_rate": 3.986873984113841e-05,
      "loss": 0.6862,
      "mean_token_accuracy": 0.7942078970372677,
      "num_tokens": 625831957.0,
      "step": 19608
    },
    {
      "entropy": 0.6413244251161814,
      "epoch": 1.804135686396283,
      "grad_norm": 0.15481851994991302,
      "learning_rate": 3.986567301499678e-05,
      "loss": 0.6287,
      "mean_token_accuracy": 0.8063286989927292,
      "num_tokens": 625863944.0,
      "step": 19609
    },
    {
      "entropy": 0.6760950293391943,
      "epoch": 1.804227692767724,
      "grad_norm": 0.1623523086309433,
      "learning_rate": 3.9862606188855154e-05,
      "loss": 0.6549,
      "mean_token_accuracy": 0.8019720576703548,
      "num_tokens": 625896081.0,
      "step": 19610
    },
    {
      "entropy": 0.7566351378336549,
      "epoch": 1.8043196991391652,
      "grad_norm": 0.1630450338125229,
      "learning_rate": 3.985953936271353e-05,
      "loss": 0.7552,
      "mean_token_accuracy": 0.7727772071957588,
      "num_tokens": 625928004.0,
      "step": 19611
    },
    {
      "entropy": 0.6146023385226727,
      "epoch": 1.8044117055106066,
      "grad_norm": 0.14944912493228912,
      "learning_rate": 3.98564725365719e-05,
      "loss": 0.5992,
      "mean_token_accuracy": 0.8137473464012146,
      "num_tokens": 625960559.0,
      "step": 19612
    },
    {
      "entropy": 0.6159031689167023,
      "epoch": 1.804503711882048,
      "grad_norm": 0.15453773736953735,
      "learning_rate": 3.985340571043028e-05,
      "loss": 0.6068,
      "mean_token_accuracy": 0.812160175293684,
      "num_tokens": 625992321.0,
      "step": 19613
    },
    {
      "entropy": 0.4776503201574087,
      "epoch": 1.804595718253489,
      "grad_norm": 0.1461823433637619,
      "learning_rate": 3.985033888428865e-05,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8493527173995972,
      "num_tokens": 626025089.0,
      "step": 19614
    },
    {
      "entropy": 0.673839008435607,
      "epoch": 1.8046877246249302,
      "grad_norm": 0.16290000081062317,
      "learning_rate": 3.984727205814703e-05,
      "loss": 0.6706,
      "mean_token_accuracy": 0.796706885099411,
      "num_tokens": 626057043.0,
      "step": 19615
    },
    {
      "entropy": 0.6074454747140408,
      "epoch": 1.8047797309963713,
      "grad_norm": 0.15484432876110077,
      "learning_rate": 3.98442052320054e-05,
      "loss": 0.5974,
      "mean_token_accuracy": 0.8138963133096695,
      "num_tokens": 626088755.0,
      "step": 19616
    },
    {
      "entropy": 0.5097125424072146,
      "epoch": 1.8048717373678127,
      "grad_norm": 0.14751915633678436,
      "learning_rate": 3.9841138405863776e-05,
      "loss": 0.5066,
      "mean_token_accuracy": 0.8435538895428181,
      "num_tokens": 626121004.0,
      "step": 19617
    },
    {
      "entropy": 0.6826008586212993,
      "epoch": 1.804963743739254,
      "grad_norm": 0.16046683490276337,
      "learning_rate": 3.9838071579722144e-05,
      "loss": 0.686,
      "mean_token_accuracy": 0.7932809069752693,
      "num_tokens": 626153349.0,
      "step": 19618
    },
    {
      "entropy": 0.6916448511183262,
      "epoch": 1.8050557501106952,
      "grad_norm": 0.15529237687587738,
      "learning_rate": 3.9835004753580525e-05,
      "loss": 0.6732,
      "mean_token_accuracy": 0.7955942638218403,
      "num_tokens": 626185337.0,
      "step": 19619
    },
    {
      "entropy": 0.5981101393699646,
      "epoch": 1.8051477564821363,
      "grad_norm": 0.15385101735591888,
      "learning_rate": 3.983193792743889e-05,
      "loss": 0.5984,
      "mean_token_accuracy": 0.8177425116300583,
      "num_tokens": 626216669.0,
      "step": 19620
    },
    {
      "entropy": 0.6912703607231379,
      "epoch": 1.8052397628535775,
      "grad_norm": 0.16498428583145142,
      "learning_rate": 3.9828871101297275e-05,
      "loss": 0.6716,
      "mean_token_accuracy": 0.7921780645847321,
      "num_tokens": 626248476.0,
      "step": 19621
    },
    {
      "entropy": 0.633050964679569,
      "epoch": 1.8053317692250188,
      "grad_norm": 0.16211800277233124,
      "learning_rate": 3.982580427515564e-05,
      "loss": 0.6224,
      "mean_token_accuracy": 0.8066808134317398,
      "num_tokens": 626280809.0,
      "step": 19622
    },
    {
      "entropy": 0.6701497733592987,
      "epoch": 1.8054237755964602,
      "grad_norm": 0.16478273272514343,
      "learning_rate": 3.982273744901402e-05,
      "loss": 0.6447,
      "mean_token_accuracy": 0.80359211191535,
      "num_tokens": 626313179.0,
      "step": 19623
    },
    {
      "entropy": 0.6635673306882381,
      "epoch": 1.8055157819679013,
      "grad_norm": 0.15644513070583344,
      "learning_rate": 3.981967062287239e-05,
      "loss": 0.6601,
      "mean_token_accuracy": 0.802894689142704,
      "num_tokens": 626345849.0,
      "step": 19624
    },
    {
      "entropy": 0.5750508429482579,
      "epoch": 1.8056077883393424,
      "grad_norm": 0.15299062430858612,
      "learning_rate": 3.9816603796730767e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.8222697116434574,
      "num_tokens": 626378057.0,
      "step": 19625
    },
    {
      "entropy": 0.5805551311932504,
      "epoch": 1.8056997947107836,
      "grad_norm": 0.15652501583099365,
      "learning_rate": 3.981353697058914e-05,
      "loss": 0.5596,
      "mean_token_accuracy": 0.8287164457142353,
      "num_tokens": 626409944.0,
      "step": 19626
    },
    {
      "entropy": 0.6467918232083321,
      "epoch": 1.805791801082225,
      "grad_norm": 0.16451458632946014,
      "learning_rate": 3.9810470144447516e-05,
      "loss": 0.6506,
      "mean_token_accuracy": 0.7995905987918377,
      "num_tokens": 626441249.0,
      "step": 19627
    },
    {
      "entropy": 0.460691939573735,
      "epoch": 1.8058838074536663,
      "grad_norm": 0.14370431005954742,
      "learning_rate": 3.9807403318305884e-05,
      "loss": 0.4421,
      "mean_token_accuracy": 0.8603102825582027,
      "num_tokens": 626473433.0,
      "step": 19628
    },
    {
      "entropy": 0.5399512257426977,
      "epoch": 1.8059758138251074,
      "grad_norm": 0.15525315701961517,
      "learning_rate": 3.9804336492164265e-05,
      "loss": 0.5381,
      "mean_token_accuracy": 0.8307579793035984,
      "num_tokens": 626504897.0,
      "step": 19629
    },
    {
      "entropy": 0.581140864174813,
      "epoch": 1.8060678201965485,
      "grad_norm": 0.15888144075870514,
      "learning_rate": 3.980126966602263e-05,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8241128996014595,
      "num_tokens": 626537076.0,
      "step": 19630
    },
    {
      "entropy": 0.5243297945708036,
      "epoch": 1.8061598265679897,
      "grad_norm": 0.14551176130771637,
      "learning_rate": 3.979820283988101e-05,
      "loss": 0.5197,
      "mean_token_accuracy": 0.8345689959824085,
      "num_tokens": 626569153.0,
      "step": 19631
    },
    {
      "entropy": 0.603745736181736,
      "epoch": 1.806251832939431,
      "grad_norm": 0.1571008712053299,
      "learning_rate": 3.979513601373938e-05,
      "loss": 0.6019,
      "mean_token_accuracy": 0.8146941997110844,
      "num_tokens": 626601042.0,
      "step": 19632
    },
    {
      "entropy": 0.5433600628748536,
      "epoch": 1.8063438393108724,
      "grad_norm": 0.14887331426143646,
      "learning_rate": 3.979206918759776e-05,
      "loss": 0.5326,
      "mean_token_accuracy": 0.8355321288108826,
      "num_tokens": 626632672.0,
      "step": 19633
    },
    {
      "entropy": 0.5994568336755037,
      "epoch": 1.8064358456823135,
      "grad_norm": 0.15959696471691132,
      "learning_rate": 3.978900236145613e-05,
      "loss": 0.5876,
      "mean_token_accuracy": 0.8154003582894802,
      "num_tokens": 626664012.0,
      "step": 19634
    },
    {
      "entropy": 0.6147602105047554,
      "epoch": 1.8065278520537547,
      "grad_norm": 0.14678119122982025,
      "learning_rate": 3.9785935535314506e-05,
      "loss": 0.599,
      "mean_token_accuracy": 0.8151426501572132,
      "num_tokens": 626696341.0,
      "step": 19635
    },
    {
      "entropy": 0.470109979622066,
      "epoch": 1.8066198584251958,
      "grad_norm": 0.1412798911333084,
      "learning_rate": 3.9782868709172874e-05,
      "loss": 0.4635,
      "mean_token_accuracy": 0.8576240949332714,
      "num_tokens": 626727955.0,
      "step": 19636
    },
    {
      "entropy": 0.6685242485255003,
      "epoch": 1.8067118647966371,
      "grad_norm": 0.1619720458984375,
      "learning_rate": 3.9779801883031256e-05,
      "loss": 0.6628,
      "mean_token_accuracy": 0.7984933480620384,
      "num_tokens": 626759699.0,
      "step": 19637
    },
    {
      "entropy": 0.6162727857008576,
      "epoch": 1.8068038711680785,
      "grad_norm": 0.15333864092826843,
      "learning_rate": 3.9776735056889623e-05,
      "loss": 0.6099,
      "mean_token_accuracy": 0.8137576058506966,
      "num_tokens": 626791760.0,
      "step": 19638
    },
    {
      "entropy": 0.6832552012056112,
      "epoch": 1.8068958775395196,
      "grad_norm": 0.1612907201051712,
      "learning_rate": 3.9773668230748e-05,
      "loss": 0.6746,
      "mean_token_accuracy": 0.7943813353776932,
      "num_tokens": 626824030.0,
      "step": 19639
    },
    {
      "entropy": 0.5086513636633754,
      "epoch": 1.8069878839109608,
      "grad_norm": 0.14268864691257477,
      "learning_rate": 3.977060140460637e-05,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8466499224305153,
      "num_tokens": 626855996.0,
      "step": 19640
    },
    {
      "entropy": 0.6137937810271978,
      "epoch": 1.807079890282402,
      "grad_norm": 0.15318015217781067,
      "learning_rate": 3.976753457846475e-05,
      "loss": 0.6039,
      "mean_token_accuracy": 0.8176166340708733,
      "num_tokens": 626887841.0,
      "step": 19641
    },
    {
      "entropy": 0.5871914685703814,
      "epoch": 1.8071718966538433,
      "grad_norm": 0.143169566988945,
      "learning_rate": 3.976446775232312e-05,
      "loss": 0.5639,
      "mean_token_accuracy": 0.8269250355660915,
      "num_tokens": 626920247.0,
      "step": 19642
    },
    {
      "entropy": 0.6841760510578752,
      "epoch": 1.8072639030252846,
      "grad_norm": 0.15626397728919983,
      "learning_rate": 3.97614009261815e-05,
      "loss": 0.6603,
      "mean_token_accuracy": 0.7970505580306053,
      "num_tokens": 626953015.0,
      "step": 19643
    },
    {
      "entropy": 0.552998268045485,
      "epoch": 1.8073559093967257,
      "grad_norm": 0.1438778191804886,
      "learning_rate": 3.975833410003987e-05,
      "loss": 0.533,
      "mean_token_accuracy": 0.8315803520381451,
      "num_tokens": 626984966.0,
      "step": 19644
    },
    {
      "entropy": 0.6776526905596256,
      "epoch": 1.8074479157681669,
      "grad_norm": 0.1627083420753479,
      "learning_rate": 3.9755267273898246e-05,
      "loss": 0.6657,
      "mean_token_accuracy": 0.7970243021845818,
      "num_tokens": 627016847.0,
      "step": 19645
    },
    {
      "entropy": 0.6516958400607109,
      "epoch": 1.807539922139608,
      "grad_norm": 0.16326472163200378,
      "learning_rate": 3.975220044775662e-05,
      "loss": 0.6408,
      "mean_token_accuracy": 0.8019921071827412,
      "num_tokens": 627049011.0,
      "step": 19646
    },
    {
      "entropy": 0.6041748151183128,
      "epoch": 1.8076319285110494,
      "grad_norm": 0.15012909471988678,
      "learning_rate": 3.9749133621614995e-05,
      "loss": 0.5896,
      "mean_token_accuracy": 0.8176882304251194,
      "num_tokens": 627081217.0,
      "step": 19647
    },
    {
      "entropy": 0.6770211942493916,
      "epoch": 1.8077239348824907,
      "grad_norm": 0.16324415802955627,
      "learning_rate": 3.974606679547337e-05,
      "loss": 0.6738,
      "mean_token_accuracy": 0.7923110611736774,
      "num_tokens": 627113378.0,
      "step": 19648
    },
    {
      "entropy": 0.5906978934071958,
      "epoch": 1.8078159412539319,
      "grad_norm": 0.14981310069561005,
      "learning_rate": 3.974299996933174e-05,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8213430196046829,
      "num_tokens": 627145361.0,
      "step": 19649
    },
    {
      "entropy": 0.62324183806777,
      "epoch": 1.807907947625373,
      "grad_norm": 0.16011789441108704,
      "learning_rate": 3.973993314319012e-05,
      "loss": 0.6225,
      "mean_token_accuracy": 0.8065179772675037,
      "num_tokens": 627176645.0,
      "step": 19650
    },
    {
      "entropy": 0.5927658644504845,
      "epoch": 1.8079999539968141,
      "grad_norm": 0.15172302722930908,
      "learning_rate": 3.973686631704849e-05,
      "loss": 0.5878,
      "mean_token_accuracy": 0.8214004784822464,
      "num_tokens": 627209043.0,
      "step": 19651
    },
    {
      "entropy": 0.5895733935758471,
      "epoch": 1.8080919603682555,
      "grad_norm": 0.15902145206928253,
      "learning_rate": 3.973379949090686e-05,
      "loss": 0.5806,
      "mean_token_accuracy": 0.8187830932438374,
      "num_tokens": 627240813.0,
      "step": 19652
    },
    {
      "entropy": 0.5731200408190489,
      "epoch": 1.8081839667396968,
      "grad_norm": 0.1529735028743744,
      "learning_rate": 3.9730732664765236e-05,
      "loss": 0.5786,
      "mean_token_accuracy": 0.820721298456192,
      "num_tokens": 627272905.0,
      "step": 19653
    },
    {
      "entropy": 0.6665705479681492,
      "epoch": 1.808275973111138,
      "grad_norm": 0.15835808217525482,
      "learning_rate": 3.972766583862361e-05,
      "loss": 0.6675,
      "mean_token_accuracy": 0.7926051616668701,
      "num_tokens": 627305240.0,
      "step": 19654
    },
    {
      "entropy": 0.6534702572971582,
      "epoch": 1.808367979482579,
      "grad_norm": 0.1633821427822113,
      "learning_rate": 3.9724599012481986e-05,
      "loss": 0.6558,
      "mean_token_accuracy": 0.7964672073721886,
      "num_tokens": 627336937.0,
      "step": 19655
    },
    {
      "entropy": 0.6467355517670512,
      "epoch": 1.8084599858540202,
      "grad_norm": 0.1664830893278122,
      "learning_rate": 3.972153218634036e-05,
      "loss": 0.6409,
      "mean_token_accuracy": 0.8023283258080482,
      "num_tokens": 627368682.0,
      "step": 19656
    },
    {
      "entropy": 0.6302555203437805,
      "epoch": 1.8085519922254616,
      "grad_norm": 0.15854424238204956,
      "learning_rate": 3.971846536019873e-05,
      "loss": 0.6286,
      "mean_token_accuracy": 0.8084743730723858,
      "num_tokens": 627400318.0,
      "step": 19657
    },
    {
      "entropy": 0.6116820406168699,
      "epoch": 1.808643998596903,
      "grad_norm": 0.15676969289779663,
      "learning_rate": 3.971539853405711e-05,
      "loss": 0.607,
      "mean_token_accuracy": 0.8154205717146397,
      "num_tokens": 627432448.0,
      "step": 19658
    },
    {
      "entropy": 0.696822096593678,
      "epoch": 1.808736004968344,
      "grad_norm": 0.1656191200017929,
      "learning_rate": 3.971233170791548e-05,
      "loss": 0.6959,
      "mean_token_accuracy": 0.7916296049952507,
      "num_tokens": 627464248.0,
      "step": 19659
    },
    {
      "entropy": 0.5494493255391717,
      "epoch": 1.8088280113397852,
      "grad_norm": 0.14777375757694244,
      "learning_rate": 3.970926488177386e-05,
      "loss": 0.546,
      "mean_token_accuracy": 0.8331186957657337,
      "num_tokens": 627496277.0,
      "step": 19660
    },
    {
      "entropy": 0.6279708351939917,
      "epoch": 1.8089200177112263,
      "grad_norm": 0.15496326982975006,
      "learning_rate": 3.970619805563223e-05,
      "loss": 0.6122,
      "mean_token_accuracy": 0.8077601827681065,
      "num_tokens": 627527541.0,
      "step": 19661
    },
    {
      "entropy": 0.6444310508668423,
      "epoch": 1.8090120240826677,
      "grad_norm": 0.15678004920482635,
      "learning_rate": 3.97031312294906e-05,
      "loss": 0.624,
      "mean_token_accuracy": 0.8115927986800671,
      "num_tokens": 627559815.0,
      "step": 19662
    },
    {
      "entropy": 0.5789706772193313,
      "epoch": 1.809104030454109,
      "grad_norm": 0.14405949413776398,
      "learning_rate": 3.9700064403348976e-05,
      "loss": 0.573,
      "mean_token_accuracy": 0.8221257142722607,
      "num_tokens": 627591893.0,
      "step": 19663
    },
    {
      "entropy": 0.6454068794846535,
      "epoch": 1.8091960368255502,
      "grad_norm": 0.15753352642059326,
      "learning_rate": 3.969699757720735e-05,
      "loss": 0.631,
      "mean_token_accuracy": 0.8060212172567844,
      "num_tokens": 627623968.0,
      "step": 19664
    },
    {
      "entropy": 0.623670669272542,
      "epoch": 1.8092880431969913,
      "grad_norm": 0.15532860159873962,
      "learning_rate": 3.9693930751065725e-05,
      "loss": 0.6061,
      "mean_token_accuracy": 0.8128023259341717,
      "num_tokens": 627655687.0,
      "step": 19665
    },
    {
      "entropy": 0.5377589035779238,
      "epoch": 1.8093800495684325,
      "grad_norm": 0.14832980930805206,
      "learning_rate": 3.96908639249241e-05,
      "loss": 0.5249,
      "mean_token_accuracy": 0.8374062739312649,
      "num_tokens": 627687716.0,
      "step": 19666
    },
    {
      "entropy": 0.5742120947688818,
      "epoch": 1.8094720559398738,
      "grad_norm": 0.14446763694286346,
      "learning_rate": 3.968779709878247e-05,
      "loss": 0.561,
      "mean_token_accuracy": 0.8299418799579144,
      "num_tokens": 627720478.0,
      "step": 19667
    },
    {
      "entropy": 0.5798861859366298,
      "epoch": 1.8095640623113152,
      "grad_norm": 0.15320245921611786,
      "learning_rate": 3.968473027264085e-05,
      "loss": 0.565,
      "mean_token_accuracy": 0.8230891264975071,
      "num_tokens": 627752908.0,
      "step": 19668
    },
    {
      "entropy": 0.6220960305072367,
      "epoch": 1.8096560686827563,
      "grad_norm": 0.15212859213352203,
      "learning_rate": 3.968166344649922e-05,
      "loss": 0.6055,
      "mean_token_accuracy": 0.814737219363451,
      "num_tokens": 627784963.0,
      "step": 19669
    },
    {
      "entropy": 0.63857277110219,
      "epoch": 1.8097480750541974,
      "grad_norm": 0.1516975313425064,
      "learning_rate": 3.967859662035759e-05,
      "loss": 0.6289,
      "mean_token_accuracy": 0.8030969649553299,
      "num_tokens": 627817300.0,
      "step": 19670
    },
    {
      "entropy": 0.6172918742522597,
      "epoch": 1.8098400814256386,
      "grad_norm": 0.152446448802948,
      "learning_rate": 3.9675529794215966e-05,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8175718896090984,
      "num_tokens": 627849008.0,
      "step": 19671
    },
    {
      "entropy": 0.6132394429296255,
      "epoch": 1.80993208779708,
      "grad_norm": 0.1516912281513214,
      "learning_rate": 3.967246296807434e-05,
      "loss": 0.6051,
      "mean_token_accuracy": 0.8129455409944057,
      "num_tokens": 627880924.0,
      "step": 19672
    },
    {
      "entropy": 0.6194743940141052,
      "epoch": 1.8100240941685213,
      "grad_norm": 0.1512969434261322,
      "learning_rate": 3.9669396141932716e-05,
      "loss": 0.614,
      "mean_token_accuracy": 0.8104345314204693,
      "num_tokens": 627912690.0,
      "step": 19673
    },
    {
      "entropy": 0.5985826491378248,
      "epoch": 1.8101161005399624,
      "grad_norm": 0.15292873978614807,
      "learning_rate": 3.966632931579109e-05,
      "loss": 0.5933,
      "mean_token_accuracy": 0.816997341811657,
      "num_tokens": 627945455.0,
      "step": 19674
    },
    {
      "entropy": 0.7187135852873325,
      "epoch": 1.8102081069114035,
      "grad_norm": 0.15833356976509094,
      "learning_rate": 3.966326248964946e-05,
      "loss": 0.7083,
      "mean_token_accuracy": 0.7811440266668797,
      "num_tokens": 627976602.0,
      "step": 19675
    },
    {
      "entropy": 0.6594322845339775,
      "epoch": 1.8103001132828447,
      "grad_norm": 0.16023971140384674,
      "learning_rate": 3.966019566350784e-05,
      "loss": 0.6532,
      "mean_token_accuracy": 0.8020622804760933,
      "num_tokens": 628008627.0,
      "step": 19676
    },
    {
      "entropy": 0.6178475208580494,
      "epoch": 1.810392119654286,
      "grad_norm": 0.15351584553718567,
      "learning_rate": 3.9657128837366214e-05,
      "loss": 0.6096,
      "mean_token_accuracy": 0.809376772493124,
      "num_tokens": 628040807.0,
      "step": 19677
    },
    {
      "entropy": 0.6504745595157146,
      "epoch": 1.8104841260257274,
      "grad_norm": 0.16227437555789948,
      "learning_rate": 3.965406201122458e-05,
      "loss": 0.6466,
      "mean_token_accuracy": 0.8042128272354603,
      "num_tokens": 628072879.0,
      "step": 19678
    },
    {
      "entropy": 0.5982335391454399,
      "epoch": 1.8105761323971685,
      "grad_norm": 0.15333744883537292,
      "learning_rate": 3.9650995185082964e-05,
      "loss": 0.5827,
      "mean_token_accuracy": 0.8210091292858124,
      "num_tokens": 628104424.0,
      "step": 19679
    },
    {
      "entropy": 0.6618705056607723,
      "epoch": 1.8106681387686097,
      "grad_norm": 0.1643526554107666,
      "learning_rate": 3.964792835894133e-05,
      "loss": 0.6438,
      "mean_token_accuracy": 0.8042960986495018,
      "num_tokens": 628135695.0,
      "step": 19680
    },
    {
      "entropy": 0.6380316540598869,
      "epoch": 1.8107601451400508,
      "grad_norm": 0.1640622913837433,
      "learning_rate": 3.964486153279971e-05,
      "loss": 0.6227,
      "mean_token_accuracy": 0.8093059286475182,
      "num_tokens": 628167939.0,
      "step": 19681
    },
    {
      "entropy": 0.5122465454041958,
      "epoch": 1.8108521515114921,
      "grad_norm": 0.134424090385437,
      "learning_rate": 3.964179470665808e-05,
      "loss": 0.5055,
      "mean_token_accuracy": 0.8434580378234386,
      "num_tokens": 628200171.0,
      "step": 19682
    },
    {
      "entropy": 0.6119568422436714,
      "epoch": 1.8109441578829335,
      "grad_norm": 0.1563064306974411,
      "learning_rate": 3.9638727880516455e-05,
      "loss": 0.5984,
      "mean_token_accuracy": 0.8166946992278099,
      "num_tokens": 628231675.0,
      "step": 19683
    },
    {
      "entropy": 0.6241343263536692,
      "epoch": 1.8110361642543746,
      "grad_norm": 0.1553719937801361,
      "learning_rate": 3.963566105437483e-05,
      "loss": 0.6147,
      "mean_token_accuracy": 0.8125998117029667,
      "num_tokens": 628263687.0,
      "step": 19684
    },
    {
      "entropy": 0.6335924118757248,
      "epoch": 1.8111281706258158,
      "grad_norm": 0.15284977853298187,
      "learning_rate": 3.9632594228233205e-05,
      "loss": 0.6301,
      "mean_token_accuracy": 0.8085798099637032,
      "num_tokens": 628296373.0,
      "step": 19685
    },
    {
      "entropy": 0.6521509476006031,
      "epoch": 1.811220176997257,
      "grad_norm": 0.15602418780326843,
      "learning_rate": 3.962952740209158e-05,
      "loss": 0.653,
      "mean_token_accuracy": 0.7963877134025097,
      "num_tokens": 628327894.0,
      "step": 19686
    },
    {
      "entropy": 0.5320874815806746,
      "epoch": 1.8113121833686983,
      "grad_norm": 0.1467352658510208,
      "learning_rate": 3.9626460575949954e-05,
      "loss": 0.5111,
      "mean_token_accuracy": 0.840484581887722,
      "num_tokens": 628358614.0,
      "step": 19687
    },
    {
      "entropy": 0.5231002988293767,
      "epoch": 1.8114041897401396,
      "grad_norm": 0.14479777216911316,
      "learning_rate": 3.962339374980832e-05,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8395036980509758,
      "num_tokens": 628390463.0,
      "step": 19688
    },
    {
      "entropy": 0.5493338275700808,
      "epoch": 1.8114961961115807,
      "grad_norm": 0.1434902548789978,
      "learning_rate": 3.96203269236667e-05,
      "loss": 0.5294,
      "mean_token_accuracy": 0.836127232760191,
      "num_tokens": 628422591.0,
      "step": 19689
    },
    {
      "entropy": 0.6085146218538284,
      "epoch": 1.8115882024830219,
      "grad_norm": 0.15082666277885437,
      "learning_rate": 3.961726009752507e-05,
      "loss": 0.5981,
      "mean_token_accuracy": 0.8151587657630444,
      "num_tokens": 628455122.0,
      "step": 19690
    },
    {
      "entropy": 0.5778056536801159,
      "epoch": 1.811680208854463,
      "grad_norm": 0.1453019082546234,
      "learning_rate": 3.9614193271383446e-05,
      "loss": 0.5668,
      "mean_token_accuracy": 0.8242072835564613,
      "num_tokens": 628487139.0,
      "step": 19691
    },
    {
      "entropy": 0.6580625409260392,
      "epoch": 1.8117722152259044,
      "grad_norm": 0.16247721016407013,
      "learning_rate": 3.961112644524182e-05,
      "loss": 0.6388,
      "mean_token_accuracy": 0.8030034564435482,
      "num_tokens": 628518783.0,
      "step": 19692
    },
    {
      "entropy": 0.584586514160037,
      "epoch": 1.8118642215973457,
      "grad_norm": 0.15264011919498444,
      "learning_rate": 3.9608059619100195e-05,
      "loss": 0.5711,
      "mean_token_accuracy": 0.8234907798469067,
      "num_tokens": 628550814.0,
      "step": 19693
    },
    {
      "entropy": 0.602263362146914,
      "epoch": 1.8119562279687869,
      "grad_norm": 0.15396302938461304,
      "learning_rate": 3.960499279295857e-05,
      "loss": 0.6012,
      "mean_token_accuracy": 0.8162215650081635,
      "num_tokens": 628582617.0,
      "step": 19694
    },
    {
      "entropy": 0.5671120677143335,
      "epoch": 1.812048234340228,
      "grad_norm": 0.1575791984796524,
      "learning_rate": 3.9601925966816944e-05,
      "loss": 0.5567,
      "mean_token_accuracy": 0.82882134988904,
      "num_tokens": 628614397.0,
      "step": 19695
    },
    {
      "entropy": 0.5681473631411791,
      "epoch": 1.8121402407116691,
      "grad_norm": 0.14778997004032135,
      "learning_rate": 3.959885914067531e-05,
      "loss": 0.5589,
      "mean_token_accuracy": 0.8255801610648632,
      "num_tokens": 628646714.0,
      "step": 19696
    },
    {
      "entropy": 0.5264808405190706,
      "epoch": 1.8122322470831105,
      "grad_norm": 0.1436534970998764,
      "learning_rate": 3.9595792314533694e-05,
      "loss": 0.528,
      "mean_token_accuracy": 0.8375953026115894,
      "num_tokens": 628679131.0,
      "step": 19697
    },
    {
      "entropy": 0.6305881338194013,
      "epoch": 1.8123242534545518,
      "grad_norm": 0.15311592817306519,
      "learning_rate": 3.959272548839206e-05,
      "loss": 0.6264,
      "mean_token_accuracy": 0.8080295026302338,
      "num_tokens": 628711667.0,
      "step": 19698
    },
    {
      "entropy": 0.6151694301515818,
      "epoch": 1.812416259825993,
      "grad_norm": 0.15470287203788757,
      "learning_rate": 3.9589658662250436e-05,
      "loss": 0.6164,
      "mean_token_accuracy": 0.8079346679151058,
      "num_tokens": 628743723.0,
      "step": 19699
    },
    {
      "entropy": 0.5169369033537805,
      "epoch": 1.812508266197434,
      "grad_norm": 0.14927425980567932,
      "learning_rate": 3.958659183610881e-05,
      "loss": 0.5146,
      "mean_token_accuracy": 0.8410098776221275,
      "num_tokens": 628775829.0,
      "step": 19700
    },
    {
      "entropy": 0.568115446716547,
      "epoch": 1.8126002725688752,
      "grad_norm": 0.1524483859539032,
      "learning_rate": 3.9583525009967186e-05,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8268774636089802,
      "num_tokens": 628807382.0,
      "step": 19701
    },
    {
      "entropy": 0.5744534619152546,
      "epoch": 1.8126922789403166,
      "grad_norm": 0.15048383176326752,
      "learning_rate": 3.958045818382556e-05,
      "loss": 0.5709,
      "mean_token_accuracy": 0.8240157328546047,
      "num_tokens": 628839324.0,
      "step": 19702
    },
    {
      "entropy": 0.6627124436199665,
      "epoch": 1.812784285311758,
      "grad_norm": 0.17060625553131104,
      "learning_rate": 3.9577391357683935e-05,
      "loss": 0.6608,
      "mean_token_accuracy": 0.7947409972548485,
      "num_tokens": 628869846.0,
      "step": 19703
    },
    {
      "entropy": 0.4791695661842823,
      "epoch": 1.812876291683199,
      "grad_norm": 0.13919039070606232,
      "learning_rate": 3.957432453154231e-05,
      "loss": 0.4589,
      "mean_token_accuracy": 0.8562414981424809,
      "num_tokens": 628902185.0,
      "step": 19704
    },
    {
      "entropy": 0.537745650857687,
      "epoch": 1.8129682980546402,
      "grad_norm": 0.15471796691417694,
      "learning_rate": 3.9571257705400684e-05,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8378068171441555,
      "num_tokens": 628933750.0,
      "step": 19705
    },
    {
      "entropy": 0.6150593236088753,
      "epoch": 1.8130603044260813,
      "grad_norm": 0.15215405821800232,
      "learning_rate": 3.956819087925906e-05,
      "loss": 0.6095,
      "mean_token_accuracy": 0.8120644278824329,
      "num_tokens": 628964707.0,
      "step": 19706
    },
    {
      "entropy": 0.6210300559177995,
      "epoch": 1.8131523107975227,
      "grad_norm": 0.1572948545217514,
      "learning_rate": 3.9565124053117433e-05,
      "loss": 0.6184,
      "mean_token_accuracy": 0.8110714256763458,
      "num_tokens": 628996534.0,
      "step": 19707
    },
    {
      "entropy": 0.6500789020210505,
      "epoch": 1.813244317168964,
      "grad_norm": 0.15416067838668823,
      "learning_rate": 3.956205722697581e-05,
      "loss": 0.6414,
      "mean_token_accuracy": 0.8051158599555492,
      "num_tokens": 629028998.0,
      "step": 19708
    },
    {
      "entropy": 0.6662484016269445,
      "epoch": 1.8133363235404052,
      "grad_norm": 0.16312384605407715,
      "learning_rate": 3.9558990400834176e-05,
      "loss": 0.6477,
      "mean_token_accuracy": 0.8077716417610645,
      "num_tokens": 629060438.0,
      "step": 19709
    },
    {
      "entropy": 0.6253886353224516,
      "epoch": 1.8134283299118463,
      "grad_norm": 0.15393348038196564,
      "learning_rate": 3.955592357469256e-05,
      "loss": 0.6234,
      "mean_token_accuracy": 0.8142906613647938,
      "num_tokens": 629092716.0,
      "step": 19710
    },
    {
      "entropy": 0.5715093603357673,
      "epoch": 1.8135203362832875,
      "grad_norm": 0.14729921519756317,
      "learning_rate": 3.9552856748550925e-05,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8304898962378502,
      "num_tokens": 629124955.0,
      "step": 19711
    },
    {
      "entropy": 0.743172274902463,
      "epoch": 1.8136123426547288,
      "grad_norm": 0.16798511147499084,
      "learning_rate": 3.95497899224093e-05,
      "loss": 0.7437,
      "mean_token_accuracy": 0.7750237509608269,
      "num_tokens": 629156623.0,
      "step": 19712
    },
    {
      "entropy": 0.6193050723522902,
      "epoch": 1.8137043490261702,
      "grad_norm": 0.16288545727729797,
      "learning_rate": 3.9546723096267675e-05,
      "loss": 0.6231,
      "mean_token_accuracy": 0.8078775815665722,
      "num_tokens": 629188171.0,
      "step": 19713
    },
    {
      "entropy": 0.5938343936577439,
      "epoch": 1.8137963553976113,
      "grad_norm": 0.15190906822681427,
      "learning_rate": 3.954365627012605e-05,
      "loss": 0.5792,
      "mean_token_accuracy": 0.8249057233333588,
      "num_tokens": 629219911.0,
      "step": 19714
    },
    {
      "entropy": 0.6478706244379282,
      "epoch": 1.8138883617690524,
      "grad_norm": 0.1607835590839386,
      "learning_rate": 3.9540589443984424e-05,
      "loss": 0.6411,
      "mean_token_accuracy": 0.8029736280441284,
      "num_tokens": 629251226.0,
      "step": 19715
    },
    {
      "entropy": 0.6149553749710321,
      "epoch": 1.8139803681404936,
      "grad_norm": 0.1489783674478531,
      "learning_rate": 3.95375226178428e-05,
      "loss": 0.5998,
      "mean_token_accuracy": 0.8143852315843105,
      "num_tokens": 629283187.0,
      "step": 19716
    },
    {
      "entropy": 0.5488078556954861,
      "epoch": 1.814072374511935,
      "grad_norm": 0.14112313091754913,
      "learning_rate": 3.9534455791701166e-05,
      "loss": 0.5344,
      "mean_token_accuracy": 0.8337725512683392,
      "num_tokens": 629315304.0,
      "step": 19717
    },
    {
      "entropy": 0.6854302790015936,
      "epoch": 1.8141643808833763,
      "grad_norm": 0.15913434326648712,
      "learning_rate": 3.953138896555955e-05,
      "loss": 0.6794,
      "mean_token_accuracy": 0.7930089384317398,
      "num_tokens": 629347271.0,
      "step": 19718
    },
    {
      "entropy": 0.5553973792120814,
      "epoch": 1.8142563872548174,
      "grad_norm": 0.1545877456665039,
      "learning_rate": 3.9528322139417916e-05,
      "loss": 0.5409,
      "mean_token_accuracy": 0.831506285816431,
      "num_tokens": 629379016.0,
      "step": 19719
    },
    {
      "entropy": 0.5504013886675239,
      "epoch": 1.8143483936262585,
      "grad_norm": 0.15012450516223907,
      "learning_rate": 3.95252553132763e-05,
      "loss": 0.543,
      "mean_token_accuracy": 0.8336714841425419,
      "num_tokens": 629410994.0,
      "step": 19720
    },
    {
      "entropy": 0.38484102254733443,
      "epoch": 1.8144403999976997,
      "grad_norm": 0.13792476058006287,
      "learning_rate": 3.9522188487134665e-05,
      "loss": 0.3584,
      "mean_token_accuracy": 0.8816006779670715,
      "num_tokens": 629442511.0,
      "step": 19721
    },
    {
      "entropy": 0.5453576277941465,
      "epoch": 1.814532406369141,
      "grad_norm": 0.15709304809570312,
      "learning_rate": 3.951912166099304e-05,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8345802277326584,
      "num_tokens": 629473485.0,
      "step": 19722
    },
    {
      "entropy": 0.6637547928839922,
      "epoch": 1.8146244127405824,
      "grad_norm": 0.16025881469249725,
      "learning_rate": 3.9516054834851414e-05,
      "loss": 0.6468,
      "mean_token_accuracy": 0.8003109507262707,
      "num_tokens": 629504058.0,
      "step": 19723
    },
    {
      "entropy": 0.6238271147012711,
      "epoch": 1.8147164191120235,
      "grad_norm": 0.15629273653030396,
      "learning_rate": 3.951298800870979e-05,
      "loss": 0.6141,
      "mean_token_accuracy": 0.8134229220449924,
      "num_tokens": 629536494.0,
      "step": 19724
    },
    {
      "entropy": 0.5855920650064945,
      "epoch": 1.8148084254834647,
      "grad_norm": 0.14767485857009888,
      "learning_rate": 3.9509921182568164e-05,
      "loss": 0.5697,
      "mean_token_accuracy": 0.8275228329002857,
      "num_tokens": 629568618.0,
      "step": 19725
    },
    {
      "entropy": 0.7276223711669445,
      "epoch": 1.8149004318549058,
      "grad_norm": 0.15991097688674927,
      "learning_rate": 3.950685435642654e-05,
      "loss": 0.715,
      "mean_token_accuracy": 0.7796902917325497,
      "num_tokens": 629600316.0,
      "step": 19726
    },
    {
      "entropy": 0.5978738889098167,
      "epoch": 1.8149924382263471,
      "grad_norm": 0.15801508724689484,
      "learning_rate": 3.9503787530284906e-05,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8232935443520546,
      "num_tokens": 629632126.0,
      "step": 19727
    },
    {
      "entropy": 0.5875893570482731,
      "epoch": 1.8150844445977885,
      "grad_norm": 0.154197558760643,
      "learning_rate": 3.950072070414329e-05,
      "loss": 0.5736,
      "mean_token_accuracy": 0.8197552785277367,
      "num_tokens": 629662380.0,
      "step": 19728
    },
    {
      "entropy": 0.5951430397108197,
      "epoch": 1.8151764509692296,
      "grad_norm": 0.1492380052804947,
      "learning_rate": 3.9497653878001655e-05,
      "loss": 0.5789,
      "mean_token_accuracy": 0.8202965930104256,
      "num_tokens": 629695033.0,
      "step": 19729
    },
    {
      "entropy": 0.5739066116511822,
      "epoch": 1.8152684573406708,
      "grad_norm": 0.15680357813835144,
      "learning_rate": 3.949458705186003e-05,
      "loss": 0.5616,
      "mean_token_accuracy": 0.8281660825014114,
      "num_tokens": 629727520.0,
      "step": 19730
    },
    {
      "entropy": 0.5444769635796547,
      "epoch": 1.815360463712112,
      "grad_norm": 0.14970168471336365,
      "learning_rate": 3.9491520225718405e-05,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8353056125342846,
      "num_tokens": 629759324.0,
      "step": 19731
    },
    {
      "entropy": 0.5757051100954413,
      "epoch": 1.8154524700835533,
      "grad_norm": 0.15515469014644623,
      "learning_rate": 3.948845339957678e-05,
      "loss": 0.5653,
      "mean_token_accuracy": 0.8217968456447124,
      "num_tokens": 629791138.0,
      "step": 19732
    },
    {
      "entropy": 0.5892382645979524,
      "epoch": 1.8155444764549946,
      "grad_norm": 0.1542702615261078,
      "learning_rate": 3.9485386573435154e-05,
      "loss": 0.5825,
      "mean_token_accuracy": 0.8159847743809223,
      "num_tokens": 629823209.0,
      "step": 19733
    },
    {
      "entropy": 0.6118314862251282,
      "epoch": 1.8156364828264357,
      "grad_norm": 0.15408022701740265,
      "learning_rate": 3.948231974729353e-05,
      "loss": 0.6063,
      "mean_token_accuracy": 0.8118037320673466,
      "num_tokens": 629855803.0,
      "step": 19734
    },
    {
      "entropy": 0.6086890585720539,
      "epoch": 1.8157284891978769,
      "grad_norm": 0.1643252670764923,
      "learning_rate": 3.9479252921151896e-05,
      "loss": 0.614,
      "mean_token_accuracy": 0.8124655373394489,
      "num_tokens": 629887460.0,
      "step": 19735
    },
    {
      "entropy": 0.602963694371283,
      "epoch": 1.815820495569318,
      "grad_norm": 0.1584659367799759,
      "learning_rate": 3.947618609501028e-05,
      "loss": 0.5847,
      "mean_token_accuracy": 0.8210466951131821,
      "num_tokens": 629918422.0,
      "step": 19736
    },
    {
      "entropy": 0.6641138703562319,
      "epoch": 1.8159125019407594,
      "grad_norm": 0.16033942997455597,
      "learning_rate": 3.947311926886865e-05,
      "loss": 0.6699,
      "mean_token_accuracy": 0.798770010471344,
      "num_tokens": 629949515.0,
      "step": 19737
    },
    {
      "entropy": 0.6168871927075088,
      "epoch": 1.8160045083122007,
      "grad_norm": 0.1516742706298828,
      "learning_rate": 3.947005244272702e-05,
      "loss": 0.6101,
      "mean_token_accuracy": 0.8125150613486767,
      "num_tokens": 629981157.0,
      "step": 19738
    },
    {
      "entropy": 0.6687012854963541,
      "epoch": 1.8160965146836419,
      "grad_norm": 0.16222846508026123,
      "learning_rate": 3.94669856165854e-05,
      "loss": 0.6647,
      "mean_token_accuracy": 0.7968453094363213,
      "num_tokens": 630013150.0,
      "step": 19739
    },
    {
      "entropy": 0.5480058584362268,
      "epoch": 1.816188521055083,
      "grad_norm": 0.14838817715644836,
      "learning_rate": 3.946391879044377e-05,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8347251154482365,
      "num_tokens": 630045324.0,
      "step": 19740
    },
    {
      "entropy": 0.5431709391996264,
      "epoch": 1.8162805274265241,
      "grad_norm": 0.15010856091976166,
      "learning_rate": 3.946085196430215e-05,
      "loss": 0.5325,
      "mean_token_accuracy": 0.8378129303455353,
      "num_tokens": 630076834.0,
      "step": 19741
    },
    {
      "entropy": 0.5808139871805906,
      "epoch": 1.8163725337979655,
      "grad_norm": 0.15243010222911835,
      "learning_rate": 3.945778513816052e-05,
      "loss": 0.5716,
      "mean_token_accuracy": 0.8200706988573074,
      "num_tokens": 630108566.0,
      "step": 19742
    },
    {
      "entropy": 0.6152374763041735,
      "epoch": 1.8164645401694068,
      "grad_norm": 0.15277312695980072,
      "learning_rate": 3.9454718312018894e-05,
      "loss": 0.6006,
      "mean_token_accuracy": 0.8143699280917645,
      "num_tokens": 630140320.0,
      "step": 19743
    },
    {
      "entropy": 0.6205722843296826,
      "epoch": 1.816556546540848,
      "grad_norm": 0.1509697437286377,
      "learning_rate": 3.945165148587727e-05,
      "loss": 0.6102,
      "mean_token_accuracy": 0.8117179200053215,
      "num_tokens": 630172071.0,
      "step": 19744
    },
    {
      "entropy": 0.5409172819927335,
      "epoch": 1.816648552912289,
      "grad_norm": 0.14332666993141174,
      "learning_rate": 3.944858465973564e-05,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8365112394094467,
      "num_tokens": 630204368.0,
      "step": 19745
    },
    {
      "entropy": 0.5638316185213625,
      "epoch": 1.8167405592837302,
      "grad_norm": 0.14540734887123108,
      "learning_rate": 3.944551783359402e-05,
      "loss": 0.5519,
      "mean_token_accuracy": 0.8320231214165688,
      "num_tokens": 630236355.0,
      "step": 19746
    },
    {
      "entropy": 0.5649634597357363,
      "epoch": 1.8168325656551716,
      "grad_norm": 0.14680038392543793,
      "learning_rate": 3.944245100745239e-05,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8306943662464619,
      "num_tokens": 630268486.0,
      "step": 19747
    },
    {
      "entropy": 0.5316953994333744,
      "epoch": 1.816924572026613,
      "grad_norm": 0.15098562836647034,
      "learning_rate": 3.943938418131076e-05,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8316545262932777,
      "num_tokens": 630300300.0,
      "step": 19748
    },
    {
      "entropy": 0.6183298509567976,
      "epoch": 1.817016578398054,
      "grad_norm": 0.1558936983346939,
      "learning_rate": 3.943631735516914e-05,
      "loss": 0.6026,
      "mean_token_accuracy": 0.817649345844984,
      "num_tokens": 630332326.0,
      "step": 19749
    },
    {
      "entropy": 0.5852311942726374,
      "epoch": 1.8171085847694952,
      "grad_norm": 0.15285801887512207,
      "learning_rate": 3.943325052902751e-05,
      "loss": 0.577,
      "mean_token_accuracy": 0.8229865878820419,
      "num_tokens": 630363830.0,
      "step": 19750
    },
    {
      "entropy": 0.6152221150696278,
      "epoch": 1.8172005911409363,
      "grad_norm": 0.15042981505393982,
      "learning_rate": 3.9430183702885884e-05,
      "loss": 0.6083,
      "mean_token_accuracy": 0.809850849211216,
      "num_tokens": 630395727.0,
      "step": 19751
    },
    {
      "entropy": 0.5435171769931912,
      "epoch": 1.8172925975123777,
      "grad_norm": 0.1446521133184433,
      "learning_rate": 3.942711687674426e-05,
      "loss": 0.5213,
      "mean_token_accuracy": 0.8404418900609016,
      "num_tokens": 630428439.0,
      "step": 19752
    },
    {
      "entropy": 0.47208412177860737,
      "epoch": 1.817384603883819,
      "grad_norm": 0.1403018832206726,
      "learning_rate": 3.942405005060263e-05,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8593733347952366,
      "num_tokens": 630460478.0,
      "step": 19753
    },
    {
      "entropy": 0.569549985229969,
      "epoch": 1.8174766102552602,
      "grad_norm": 0.14403308928012848,
      "learning_rate": 3.942098322446101e-05,
      "loss": 0.5512,
      "mean_token_accuracy": 0.8279048651456833,
      "num_tokens": 630492394.0,
      "step": 19754
    },
    {
      "entropy": 0.5783231938257813,
      "epoch": 1.8175686166267013,
      "grad_norm": 0.15060964226722717,
      "learning_rate": 3.941791639831938e-05,
      "loss": 0.5734,
      "mean_token_accuracy": 0.8227127008140087,
      "num_tokens": 630524286.0,
      "step": 19755
    },
    {
      "entropy": 0.5931335184723139,
      "epoch": 1.8176606229981425,
      "grad_norm": 0.15663570165634155,
      "learning_rate": 3.941484957217775e-05,
      "loss": 0.5816,
      "mean_token_accuracy": 0.820449236780405,
      "num_tokens": 630556586.0,
      "step": 19756
    },
    {
      "entropy": 0.6532607916742563,
      "epoch": 1.8177526293695838,
      "grad_norm": 0.1629168689250946,
      "learning_rate": 3.941178274603613e-05,
      "loss": 0.6403,
      "mean_token_accuracy": 0.8014426343142986,
      "num_tokens": 630588746.0,
      "step": 19757
    },
    {
      "entropy": 0.5997466575354338,
      "epoch": 1.8178446357410252,
      "grad_norm": 0.15911872684955597,
      "learning_rate": 3.94087159198945e-05,
      "loss": 0.5918,
      "mean_token_accuracy": 0.8203820213675499,
      "num_tokens": 630620413.0,
      "step": 19758
    },
    {
      "entropy": 0.5997819118201733,
      "epoch": 1.8179366421124663,
      "grad_norm": 0.15112343430519104,
      "learning_rate": 3.9405649093752874e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8167109377682209,
      "num_tokens": 630652403.0,
      "step": 19759
    },
    {
      "entropy": 0.5674559427425265,
      "epoch": 1.8180286484839074,
      "grad_norm": 0.14801065623760223,
      "learning_rate": 3.940258226761125e-05,
      "loss": 0.555,
      "mean_token_accuracy": 0.8282895050942898,
      "num_tokens": 630684101.0,
      "step": 19760
    },
    {
      "entropy": 0.580199096351862,
      "epoch": 1.8181206548553486,
      "grad_norm": 0.15430352091789246,
      "learning_rate": 3.9399515441469624e-05,
      "loss": 0.5733,
      "mean_token_accuracy": 0.8203308507800102,
      "num_tokens": 630715754.0,
      "step": 19761
    },
    {
      "entropy": 0.5491148428991437,
      "epoch": 1.81821266122679,
      "grad_norm": 0.14829117059707642,
      "learning_rate": 3.9396448615328e-05,
      "loss": 0.5394,
      "mean_token_accuracy": 0.8327841088175774,
      "num_tokens": 630747858.0,
      "step": 19762
    },
    {
      "entropy": 0.6168736033141613,
      "epoch": 1.8183046675982313,
      "grad_norm": 0.15814171731472015,
      "learning_rate": 3.939338178918637e-05,
      "loss": 0.6042,
      "mean_token_accuracy": 0.8158666491508484,
      "num_tokens": 630779570.0,
      "step": 19763
    },
    {
      "entropy": 0.6315469890832901,
      "epoch": 1.8183966739696724,
      "grad_norm": 0.1555919349193573,
      "learning_rate": 3.939031496304475e-05,
      "loss": 0.6204,
      "mean_token_accuracy": 0.810073509812355,
      "num_tokens": 630810464.0,
      "step": 19764
    },
    {
      "entropy": 0.5470310715027153,
      "epoch": 1.8184886803411136,
      "grad_norm": 0.1535697728395462,
      "learning_rate": 3.938724813690312e-05,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8377785198390484,
      "num_tokens": 630842572.0,
      "step": 19765
    },
    {
      "entropy": 0.6067739520221949,
      "epoch": 1.8185806867125547,
      "grad_norm": 0.15352113544940948,
      "learning_rate": 3.938418131076149e-05,
      "loss": 0.6048,
      "mean_token_accuracy": 0.8154333755373955,
      "num_tokens": 630874905.0,
      "step": 19766
    },
    {
      "entropy": 0.6752205993980169,
      "epoch": 1.818672693083996,
      "grad_norm": 0.1548513025045395,
      "learning_rate": 3.938111448461987e-05,
      "loss": 0.6697,
      "mean_token_accuracy": 0.7947340160608292,
      "num_tokens": 630907061.0,
      "step": 19767
    },
    {
      "entropy": 0.5739436661824584,
      "epoch": 1.8187646994554374,
      "grad_norm": 0.15076084434986115,
      "learning_rate": 3.9378047658478246e-05,
      "loss": 0.5598,
      "mean_token_accuracy": 0.8275816291570663,
      "num_tokens": 630939073.0,
      "step": 19768
    },
    {
      "entropy": 0.5608623493462801,
      "epoch": 1.8188567058268785,
      "grad_norm": 0.1490159034729004,
      "learning_rate": 3.9374980832336614e-05,
      "loss": 0.543,
      "mean_token_accuracy": 0.8323273360729218,
      "num_tokens": 630971379.0,
      "step": 19769
    },
    {
      "entropy": 0.6533097699284554,
      "epoch": 1.8189487121983197,
      "grad_norm": 0.16116264462471008,
      "learning_rate": 3.9371914006194996e-05,
      "loss": 0.6426,
      "mean_token_accuracy": 0.8067579381167889,
      "num_tokens": 631003032.0,
      "step": 19770
    },
    {
      "entropy": 0.5474647684022784,
      "epoch": 1.8190407185697608,
      "grad_norm": 0.15620918571949005,
      "learning_rate": 3.9368847180053363e-05,
      "loss": 0.5443,
      "mean_token_accuracy": 0.8272419199347496,
      "num_tokens": 631034794.0,
      "step": 19771
    },
    {
      "entropy": 0.560555418021977,
      "epoch": 1.8191327249412022,
      "grad_norm": 0.14557693898677826,
      "learning_rate": 3.936578035391174e-05,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8287065513432026,
      "num_tokens": 631066936.0,
      "step": 19772
    },
    {
      "entropy": 0.6116153243929148,
      "epoch": 1.8192247313126435,
      "grad_norm": 0.16047804057598114,
      "learning_rate": 3.936271352777011e-05,
      "loss": 0.6068,
      "mean_token_accuracy": 0.811907272785902,
      "num_tokens": 631098780.0,
      "step": 19773
    },
    {
      "entropy": 0.6443048571236432,
      "epoch": 1.8193167376840846,
      "grad_norm": 0.1587250679731369,
      "learning_rate": 3.935964670162849e-05,
      "loss": 0.6263,
      "mean_token_accuracy": 0.8076195232570171,
      "num_tokens": 631130628.0,
      "step": 19774
    },
    {
      "entropy": 0.6270075421780348,
      "epoch": 1.8194087440555258,
      "grad_norm": 0.16180972754955292,
      "learning_rate": 3.935657987548686e-05,
      "loss": 0.614,
      "mean_token_accuracy": 0.8099572733044624,
      "num_tokens": 631162275.0,
      "step": 19775
    },
    {
      "entropy": 0.5304785813204944,
      "epoch": 1.819500750426967,
      "grad_norm": 0.14700011909008026,
      "learning_rate": 3.935351304934524e-05,
      "loss": 0.5127,
      "mean_token_accuracy": 0.8400330506265163,
      "num_tokens": 631194198.0,
      "step": 19776
    },
    {
      "entropy": 0.6074763834476471,
      "epoch": 1.8195927567984083,
      "grad_norm": 0.1535351574420929,
      "learning_rate": 3.9350446223203605e-05,
      "loss": 0.5932,
      "mean_token_accuracy": 0.8197964131832123,
      "num_tokens": 631226043.0,
      "step": 19777
    },
    {
      "entropy": 0.5268052583560348,
      "epoch": 1.8196847631698496,
      "grad_norm": 0.14781434834003448,
      "learning_rate": 3.9347379397061986e-05,
      "loss": 0.5106,
      "mean_token_accuracy": 0.8403132744133472,
      "num_tokens": 631257631.0,
      "step": 19778
    },
    {
      "entropy": 0.6667416524142027,
      "epoch": 1.8197767695412908,
      "grad_norm": 0.15836089849472046,
      "learning_rate": 3.9344312570920354e-05,
      "loss": 0.6471,
      "mean_token_accuracy": 0.8061173856258392,
      "num_tokens": 631289346.0,
      "step": 19779
    },
    {
      "entropy": 0.574977919459343,
      "epoch": 1.8198687759127319,
      "grad_norm": 0.15306390821933746,
      "learning_rate": 3.9341245744778735e-05,
      "loss": 0.5588,
      "mean_token_accuracy": 0.8264318965375423,
      "num_tokens": 631320516.0,
      "step": 19780
    },
    {
      "entropy": 0.5773590225726366,
      "epoch": 1.819960782284173,
      "grad_norm": 0.15007390081882477,
      "learning_rate": 3.93381789186371e-05,
      "loss": 0.5695,
      "mean_token_accuracy": 0.8242329098284245,
      "num_tokens": 631352782.0,
      "step": 19781
    },
    {
      "entropy": 0.6126144509762526,
      "epoch": 1.8200527886556144,
      "grad_norm": 0.15867426991462708,
      "learning_rate": 3.933511209249548e-05,
      "loss": 0.6146,
      "mean_token_accuracy": 0.8126066662371159,
      "num_tokens": 631385098.0,
      "step": 19782
    },
    {
      "entropy": 0.589085889980197,
      "epoch": 1.8201447950270557,
      "grad_norm": 0.15244127810001373,
      "learning_rate": 3.933204526635385e-05,
      "loss": 0.5928,
      "mean_token_accuracy": 0.8139771334826946,
      "num_tokens": 631417038.0,
      "step": 19783
    },
    {
      "entropy": 0.4934038622304797,
      "epoch": 1.8202368013984969,
      "grad_norm": 0.14001718163490295,
      "learning_rate": 3.932897844021223e-05,
      "loss": 0.4778,
      "mean_token_accuracy": 0.8489920943975449,
      "num_tokens": 631449448.0,
      "step": 19784
    },
    {
      "entropy": 0.5253515001386404,
      "epoch": 1.820328807769938,
      "grad_norm": 0.14452306926250458,
      "learning_rate": 3.93259116140706e-05,
      "loss": 0.5144,
      "mean_token_accuracy": 0.8372367024421692,
      "num_tokens": 631480997.0,
      "step": 19785
    },
    {
      "entropy": 0.61878811288625,
      "epoch": 1.8204208141413791,
      "grad_norm": 0.15377084910869598,
      "learning_rate": 3.9322844787928976e-05,
      "loss": 0.597,
      "mean_token_accuracy": 0.8119247332215309,
      "num_tokens": 631511712.0,
      "step": 19786
    },
    {
      "entropy": 0.6565455701202154,
      "epoch": 1.8205128205128205,
      "grad_norm": 0.16044799983501434,
      "learning_rate": 3.9319777961787344e-05,
      "loss": 0.649,
      "mean_token_accuracy": 0.8013779297471046,
      "num_tokens": 631542451.0,
      "step": 19787
    },
    {
      "entropy": 0.6352641293779016,
      "epoch": 1.8206048268842618,
      "grad_norm": 0.15890547633171082,
      "learning_rate": 3.9316711135645726e-05,
      "loss": 0.6258,
      "mean_token_accuracy": 0.8098976537585258,
      "num_tokens": 631574553.0,
      "step": 19788
    },
    {
      "entropy": 0.701197418384254,
      "epoch": 1.820696833255703,
      "grad_norm": 0.16583281755447388,
      "learning_rate": 3.9313644309504094e-05,
      "loss": 0.7025,
      "mean_token_accuracy": 0.7880775481462479,
      "num_tokens": 631606521.0,
      "step": 19789
    },
    {
      "entropy": 0.530003952793777,
      "epoch": 1.820788839627144,
      "grad_norm": 0.14862646162509918,
      "learning_rate": 3.931057748336247e-05,
      "loss": 0.5255,
      "mean_token_accuracy": 0.8393947668373585,
      "num_tokens": 631638293.0,
      "step": 19790
    },
    {
      "entropy": 0.4715066347271204,
      "epoch": 1.8208808459985852,
      "grad_norm": 0.14435617625713348,
      "learning_rate": 3.930751065722084e-05,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8567493371665478,
      "num_tokens": 631670461.0,
      "step": 19791
    },
    {
      "entropy": 0.5731705725193024,
      "epoch": 1.8209728523700266,
      "grad_norm": 0.15881192684173584,
      "learning_rate": 3.930444383107922e-05,
      "loss": 0.572,
      "mean_token_accuracy": 0.8248885497450829,
      "num_tokens": 631703206.0,
      "step": 19792
    },
    {
      "entropy": 0.46784820035099983,
      "epoch": 1.821064858741468,
      "grad_norm": 0.1445855349302292,
      "learning_rate": 3.930137700493759e-05,
      "loss": 0.4612,
      "mean_token_accuracy": 0.8563790954649448,
      "num_tokens": 631735436.0,
      "step": 19793
    },
    {
      "entropy": 0.5838362611830235,
      "epoch": 1.821156865112909,
      "grad_norm": 0.15169905126094818,
      "learning_rate": 3.929831017879597e-05,
      "loss": 0.5718,
      "mean_token_accuracy": 0.8215962648391724,
      "num_tokens": 631766920.0,
      "step": 19794
    },
    {
      "entropy": 0.60914128087461,
      "epoch": 1.8212488714843502,
      "grad_norm": 0.1730475276708603,
      "learning_rate": 3.9295243352654335e-05,
      "loss": 0.62,
      "mean_token_accuracy": 0.814936388283968,
      "num_tokens": 631798568.0,
      "step": 19795
    },
    {
      "entropy": 0.5722020035609603,
      "epoch": 1.8213408778557914,
      "grad_norm": 0.15139129757881165,
      "learning_rate": 3.9292176526512716e-05,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8271509930491447,
      "num_tokens": 631830495.0,
      "step": 19796
    },
    {
      "entropy": 0.6473117806017399,
      "epoch": 1.8214328842272327,
      "grad_norm": 0.15577568113803864,
      "learning_rate": 3.9289109700371084e-05,
      "loss": 0.6309,
      "mean_token_accuracy": 0.8047800175845623,
      "num_tokens": 631862447.0,
      "step": 19797
    },
    {
      "entropy": 0.5784150836989284,
      "epoch": 1.821524890598674,
      "grad_norm": 0.1552949994802475,
      "learning_rate": 3.928604287422946e-05,
      "loss": 0.5697,
      "mean_token_accuracy": 0.8262351118028164,
      "num_tokens": 631894617.0,
      "step": 19798
    },
    {
      "entropy": 0.6174054937437177,
      "epoch": 1.8216168969701152,
      "grad_norm": 0.15573450922966003,
      "learning_rate": 3.928297604808784e-05,
      "loss": 0.5906,
      "mean_token_accuracy": 0.814210083335638,
      "num_tokens": 631925631.0,
      "step": 19799
    },
    {
      "entropy": 0.5745542729273438,
      "epoch": 1.8217089033415563,
      "grad_norm": 0.15601183474063873,
      "learning_rate": 3.927990922194621e-05,
      "loss": 0.5661,
      "mean_token_accuracy": 0.8295669369399548,
      "num_tokens": 631956711.0,
      "step": 19800
    },
    {
      "entropy": 0.6619389485567808,
      "epoch": 1.8218009097129975,
      "grad_norm": 0.15327604115009308,
      "learning_rate": 3.927684239580459e-05,
      "loss": 0.6436,
      "mean_token_accuracy": 0.8031568117439747,
      "num_tokens": 631988774.0,
      "step": 19801
    },
    {
      "entropy": 0.7016433756798506,
      "epoch": 1.8218929160844388,
      "grad_norm": 0.16466665267944336,
      "learning_rate": 3.927377556966296e-05,
      "loss": 0.6874,
      "mean_token_accuracy": 0.7918103374540806,
      "num_tokens": 632019393.0,
      "step": 19802
    },
    {
      "entropy": 0.6588362650945783,
      "epoch": 1.8219849224558802,
      "grad_norm": 0.16073262691497803,
      "learning_rate": 3.927070874352133e-05,
      "loss": 0.6484,
      "mean_token_accuracy": 0.8022724241018295,
      "num_tokens": 632051668.0,
      "step": 19803
    },
    {
      "entropy": 0.6217780746519566,
      "epoch": 1.8220769288273213,
      "grad_norm": 0.1528976410627365,
      "learning_rate": 3.9267641917379706e-05,
      "loss": 0.609,
      "mean_token_accuracy": 0.8114331029355526,
      "num_tokens": 632084060.0,
      "step": 19804
    },
    {
      "entropy": 0.6427281219512224,
      "epoch": 1.8221689351987624,
      "grad_norm": 0.15237164497375488,
      "learning_rate": 3.926457509123808e-05,
      "loss": 0.6288,
      "mean_token_accuracy": 0.8050901293754578,
      "num_tokens": 632116304.0,
      "step": 19805
    },
    {
      "entropy": 0.5720177507027984,
      "epoch": 1.8222609415702036,
      "grad_norm": 0.14661270380020142,
      "learning_rate": 3.9261508265096456e-05,
      "loss": 0.5592,
      "mean_token_accuracy": 0.8296266570687294,
      "num_tokens": 632148635.0,
      "step": 19806
    },
    {
      "entropy": 0.506156996358186,
      "epoch": 1.822352947941645,
      "grad_norm": 0.14379213750362396,
      "learning_rate": 3.925844143895483e-05,
      "loss": 0.4915,
      "mean_token_accuracy": 0.8497302941977978,
      "num_tokens": 632180883.0,
      "step": 19807
    },
    {
      "entropy": 0.6199362650513649,
      "epoch": 1.8224449543130863,
      "grad_norm": 0.15158961713314056,
      "learning_rate": 3.92553746128132e-05,
      "loss": 0.6001,
      "mean_token_accuracy": 0.8126920983195305,
      "num_tokens": 632213000.0,
      "step": 19808
    },
    {
      "entropy": 0.598214359022677,
      "epoch": 1.8225369606845274,
      "grad_norm": 0.15871447324752808,
      "learning_rate": 3.925230778667158e-05,
      "loss": 0.5861,
      "mean_token_accuracy": 0.82323332503438,
      "num_tokens": 632244640.0,
      "step": 19809
    },
    {
      "entropy": 0.567473204806447,
      "epoch": 1.8226289670559686,
      "grad_norm": 0.14873924851417542,
      "learning_rate": 3.924924096052995e-05,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8330237045884132,
      "num_tokens": 632276671.0,
      "step": 19810
    },
    {
      "entropy": 0.5912908473983407,
      "epoch": 1.8227209734274097,
      "grad_norm": 0.15278007090091705,
      "learning_rate": 3.924617413438832e-05,
      "loss": 0.5843,
      "mean_token_accuracy": 0.8162257708609104,
      "num_tokens": 632308685.0,
      "step": 19811
    },
    {
      "entropy": 0.7012377306818962,
      "epoch": 1.822812979798851,
      "grad_norm": 0.15983538329601288,
      "learning_rate": 3.92431073082467e-05,
      "loss": 0.6768,
      "mean_token_accuracy": 0.7911415100097656,
      "num_tokens": 632339355.0,
      "step": 19812
    },
    {
      "entropy": 0.5798200163990259,
      "epoch": 1.8229049861702924,
      "grad_norm": 0.14966019988059998,
      "learning_rate": 3.924004048210507e-05,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8243272118270397,
      "num_tokens": 632370937.0,
      "step": 19813
    },
    {
      "entropy": 0.4719609811436385,
      "epoch": 1.8229969925417335,
      "grad_norm": 0.14259831607341766,
      "learning_rate": 3.9236973655963446e-05,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8610514625906944,
      "num_tokens": 632403502.0,
      "step": 19814
    },
    {
      "entropy": 0.5994810126721859,
      "epoch": 1.8230889989131747,
      "grad_norm": 0.14957883954048157,
      "learning_rate": 3.923390682982182e-05,
      "loss": 0.588,
      "mean_token_accuracy": 0.8223004266619682,
      "num_tokens": 632435741.0,
      "step": 19815
    },
    {
      "entropy": 0.5781439477577806,
      "epoch": 1.8231810052846158,
      "grad_norm": 0.15745790302753448,
      "learning_rate": 3.923084000368019e-05,
      "loss": 0.5824,
      "mean_token_accuracy": 0.8170263506472111,
      "num_tokens": 632467736.0,
      "step": 19816
    },
    {
      "entropy": 0.6174198295921087,
      "epoch": 1.8232730116560572,
      "grad_norm": 0.1556810885667801,
      "learning_rate": 3.922777317753857e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8078222721815109,
      "num_tokens": 632499199.0,
      "step": 19817
    },
    {
      "entropy": 0.6296527278609574,
      "epoch": 1.8233650180274985,
      "grad_norm": 0.16015799343585968,
      "learning_rate": 3.922470635139694e-05,
      "loss": 0.6197,
      "mean_token_accuracy": 0.8084740750491619,
      "num_tokens": 632531129.0,
      "step": 19818
    },
    {
      "entropy": 0.58924017008394,
      "epoch": 1.8234570243989396,
      "grad_norm": 0.14960461854934692,
      "learning_rate": 3.922163952525531e-05,
      "loss": 0.5871,
      "mean_token_accuracy": 0.8211496733129025,
      "num_tokens": 632563532.0,
      "step": 19819
    },
    {
      "entropy": 0.6038908092305064,
      "epoch": 1.8235490307703808,
      "grad_norm": 0.1537250131368637,
      "learning_rate": 3.921857269911369e-05,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8182860240340233,
      "num_tokens": 632595677.0,
      "step": 19820
    },
    {
      "entropy": 0.59375792555511,
      "epoch": 1.823641037141822,
      "grad_norm": 0.15674975514411926,
      "learning_rate": 3.921550587297206e-05,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8174643814563751,
      "num_tokens": 632627766.0,
      "step": 19821
    },
    {
      "entropy": 0.506486302241683,
      "epoch": 1.8237330435132633,
      "grad_norm": 0.14141269028186798,
      "learning_rate": 3.9212439046830437e-05,
      "loss": 0.4904,
      "mean_token_accuracy": 0.8436726219952106,
      "num_tokens": 632660132.0,
      "step": 19822
    },
    {
      "entropy": 0.6376088354736567,
      "epoch": 1.8238250498847046,
      "grad_norm": 0.15830039978027344,
      "learning_rate": 3.920937222068881e-05,
      "loss": 0.6343,
      "mean_token_accuracy": 0.8063193149864674,
      "num_tokens": 632691651.0,
      "step": 19823
    },
    {
      "entropy": 0.61360246501863,
      "epoch": 1.8239170562561458,
      "grad_norm": 0.15192550420761108,
      "learning_rate": 3.9206305394547186e-05,
      "loss": 0.6097,
      "mean_token_accuracy": 0.8136050328612328,
      "num_tokens": 632723613.0,
      "step": 19824
    },
    {
      "entropy": 0.6629470661282539,
      "epoch": 1.8240090626275869,
      "grad_norm": 0.1548573076725006,
      "learning_rate": 3.920323856840556e-05,
      "loss": 0.6553,
      "mean_token_accuracy": 0.8009595163166523,
      "num_tokens": 632756306.0,
      "step": 19825
    },
    {
      "entropy": 0.5650609638541937,
      "epoch": 1.824101068999028,
      "grad_norm": 0.14665338397026062,
      "learning_rate": 3.920017174226393e-05,
      "loss": 0.5446,
      "mean_token_accuracy": 0.8297947756946087,
      "num_tokens": 632788666.0,
      "step": 19826
    },
    {
      "entropy": 0.557115700095892,
      "epoch": 1.8241930753704694,
      "grad_norm": 0.1457880288362503,
      "learning_rate": 3.919710491612231e-05,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8338144607841969,
      "num_tokens": 632820955.0,
      "step": 19827
    },
    {
      "entropy": 0.5669169519096613,
      "epoch": 1.8242850817419107,
      "grad_norm": 0.14779885113239288,
      "learning_rate": 3.9194038089980684e-05,
      "loss": 0.5572,
      "mean_token_accuracy": 0.8285987675189972,
      "num_tokens": 632853167.0,
      "step": 19828
    },
    {
      "entropy": 0.6293983124196529,
      "epoch": 1.8243770881133519,
      "grad_norm": 0.15935036540031433,
      "learning_rate": 3.919097126383905e-05,
      "loss": 0.6318,
      "mean_token_accuracy": 0.8040303289890289,
      "num_tokens": 632884859.0,
      "step": 19829
    },
    {
      "entropy": 0.6517956340685487,
      "epoch": 1.824469094484793,
      "grad_norm": 0.1603582799434662,
      "learning_rate": 3.9187904437697434e-05,
      "loss": 0.644,
      "mean_token_accuracy": 0.8011641651391983,
      "num_tokens": 632916202.0,
      "step": 19830
    },
    {
      "entropy": 0.6104820491746068,
      "epoch": 1.8245611008562341,
      "grad_norm": 0.15735076367855072,
      "learning_rate": 3.91848376115558e-05,
      "loss": 0.5928,
      "mean_token_accuracy": 0.8172875791788101,
      "num_tokens": 632947415.0,
      "step": 19831
    },
    {
      "entropy": 0.6371979210525751,
      "epoch": 1.8246531072276755,
      "grad_norm": 0.15566714107990265,
      "learning_rate": 3.9181770785414176e-05,
      "loss": 0.6271,
      "mean_token_accuracy": 0.8140976056456566,
      "num_tokens": 632979115.0,
      "step": 19832
    },
    {
      "entropy": 0.6685224305838346,
      "epoch": 1.8247451135991168,
      "grad_norm": 0.15916772186756134,
      "learning_rate": 3.917870395927255e-05,
      "loss": 0.6624,
      "mean_token_accuracy": 0.7986115403473377,
      "num_tokens": 633011537.0,
      "step": 19833
    },
    {
      "entropy": 0.5839961264282465,
      "epoch": 1.824837119970558,
      "grad_norm": 0.14988869428634644,
      "learning_rate": 3.9175637133130926e-05,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8234565556049347,
      "num_tokens": 633042839.0,
      "step": 19834
    },
    {
      "entropy": 0.6499148011207581,
      "epoch": 1.824929126341999,
      "grad_norm": 0.1549239456653595,
      "learning_rate": 3.91725703069893e-05,
      "loss": 0.6386,
      "mean_token_accuracy": 0.8013083562254906,
      "num_tokens": 633074744.0,
      "step": 19835
    },
    {
      "entropy": 0.6393122347071767,
      "epoch": 1.8250211327134402,
      "grad_norm": 0.1617501825094223,
      "learning_rate": 3.9169503480847675e-05,
      "loss": 0.6362,
      "mean_token_accuracy": 0.8077269829809666,
      "num_tokens": 633107052.0,
      "step": 19836
    },
    {
      "entropy": 0.6474182698875666,
      "epoch": 1.8251131390848816,
      "grad_norm": 0.15906934440135956,
      "learning_rate": 3.916643665470604e-05,
      "loss": 0.6476,
      "mean_token_accuracy": 0.8009101487696171,
      "num_tokens": 633138122.0,
      "step": 19837
    },
    {
      "entropy": 0.6330184899270535,
      "epoch": 1.825205145456323,
      "grad_norm": 0.15777632594108582,
      "learning_rate": 3.9163369828564424e-05,
      "loss": 0.6269,
      "mean_token_accuracy": 0.8062306046485901,
      "num_tokens": 633169744.0,
      "step": 19838
    },
    {
      "entropy": 0.6686136908829212,
      "epoch": 1.825297151827764,
      "grad_norm": 0.16348421573638916,
      "learning_rate": 3.916030300242279e-05,
      "loss": 0.6523,
      "mean_token_accuracy": 0.8003730215132236,
      "num_tokens": 633200687.0,
      "step": 19839
    },
    {
      "entropy": 0.5414921282790601,
      "epoch": 1.8253891581992052,
      "grad_norm": 0.14158686995506287,
      "learning_rate": 3.9157236176281173e-05,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8332053273916245,
      "num_tokens": 633232966.0,
      "step": 19840
    },
    {
      "entropy": 0.6112580085173249,
      "epoch": 1.8254811645706464,
      "grad_norm": 0.15773580968379974,
      "learning_rate": 3.915416935013954e-05,
      "loss": 0.5977,
      "mean_token_accuracy": 0.8161465860903263,
      "num_tokens": 633264688.0,
      "step": 19841
    },
    {
      "entropy": 0.5816117357462645,
      "epoch": 1.8255731709420877,
      "grad_norm": 0.15458640456199646,
      "learning_rate": 3.9151102523997916e-05,
      "loss": 0.5746,
      "mean_token_accuracy": 0.8242774568498135,
      "num_tokens": 633296042.0,
      "step": 19842
    },
    {
      "entropy": 0.47842342872172594,
      "epoch": 1.825665177313529,
      "grad_norm": 0.14482048153877258,
      "learning_rate": 3.914803569785629e-05,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8497943133115768,
      "num_tokens": 633328600.0,
      "step": 19843
    },
    {
      "entropy": 0.6627462888136506,
      "epoch": 1.8257571836849702,
      "grad_norm": 0.15639446675777435,
      "learning_rate": 3.9144968871714665e-05,
      "loss": 0.6601,
      "mean_token_accuracy": 0.7956443354487419,
      "num_tokens": 633361289.0,
      "step": 19844
    },
    {
      "entropy": 0.5582944545894861,
      "epoch": 1.8258491900564113,
      "grad_norm": 0.14855732023715973,
      "learning_rate": 3.914190204557304e-05,
      "loss": 0.5437,
      "mean_token_accuracy": 0.8315523155033588,
      "num_tokens": 633393149.0,
      "step": 19845
    },
    {
      "entropy": 0.5536662675440311,
      "epoch": 1.8259411964278525,
      "grad_norm": 0.15319974720478058,
      "learning_rate": 3.9138835219431415e-05,
      "loss": 0.5535,
      "mean_token_accuracy": 0.8322511613368988,
      "num_tokens": 633424952.0,
      "step": 19846
    },
    {
      "entropy": 0.6775252223014832,
      "epoch": 1.8260332027992938,
      "grad_norm": 0.16197389364242554,
      "learning_rate": 3.913576839328978e-05,
      "loss": 0.6696,
      "mean_token_accuracy": 0.8009588681161404,
      "num_tokens": 633456610.0,
      "step": 19847
    },
    {
      "entropy": 0.6456598532386124,
      "epoch": 1.8261252091707352,
      "grad_norm": 0.15763802826404572,
      "learning_rate": 3.9132701567148164e-05,
      "loss": 0.6406,
      "mean_token_accuracy": 0.8091340810060501,
      "num_tokens": 633488602.0,
      "step": 19848
    },
    {
      "entropy": 0.6756261307746172,
      "epoch": 1.8262172155421763,
      "grad_norm": 0.16841869056224823,
      "learning_rate": 3.912963474100653e-05,
      "loss": 0.6775,
      "mean_token_accuracy": 0.7944507002830505,
      "num_tokens": 633520924.0,
      "step": 19849
    },
    {
      "entropy": 0.7124388664960861,
      "epoch": 1.8263092219136174,
      "grad_norm": 0.1615920215845108,
      "learning_rate": 3.9126567914864906e-05,
      "loss": 0.6965,
      "mean_token_accuracy": 0.7873638793826103,
      "num_tokens": 633552724.0,
      "step": 19850
    },
    {
      "entropy": 0.5491883372887969,
      "epoch": 1.8264012282850586,
      "grad_norm": 0.14688290655612946,
      "learning_rate": 3.912350108872328e-05,
      "loss": 0.5367,
      "mean_token_accuracy": 0.8352403528988361,
      "num_tokens": 633584658.0,
      "step": 19851
    },
    {
      "entropy": 0.5974816682282835,
      "epoch": 1.8264932346565,
      "grad_norm": 0.14902400970458984,
      "learning_rate": 3.9120434262581656e-05,
      "loss": 0.5773,
      "mean_token_accuracy": 0.8263432309031487,
      "num_tokens": 633616682.0,
      "step": 19852
    },
    {
      "entropy": 0.6016256213188171,
      "epoch": 1.8265852410279413,
      "grad_norm": 0.15757644176483154,
      "learning_rate": 3.911736743644003e-05,
      "loss": 0.5903,
      "mean_token_accuracy": 0.8187605068087578,
      "num_tokens": 633648727.0,
      "step": 19853
    },
    {
      "entropy": 0.6486489996314049,
      "epoch": 1.8266772473993824,
      "grad_norm": 0.15394090116024017,
      "learning_rate": 3.9114300610298405e-05,
      "loss": 0.6391,
      "mean_token_accuracy": 0.8033328279852867,
      "num_tokens": 633680878.0,
      "step": 19854
    },
    {
      "entropy": 0.6494254022836685,
      "epoch": 1.8267692537708236,
      "grad_norm": 0.15659460425376892,
      "learning_rate": 3.911123378415677e-05,
      "loss": 0.644,
      "mean_token_accuracy": 0.8040993884205818,
      "num_tokens": 633712924.0,
      "step": 19855
    },
    {
      "entropy": 0.5586623037233949,
      "epoch": 1.8268612601422647,
      "grad_norm": 0.1448449194431305,
      "learning_rate": 3.9108166958015154e-05,
      "loss": 0.5323,
      "mean_token_accuracy": 0.8294123783707619,
      "num_tokens": 633743970.0,
      "step": 19856
    },
    {
      "entropy": 0.5615970042999834,
      "epoch": 1.826953266513706,
      "grad_norm": 0.14378778636455536,
      "learning_rate": 3.910510013187352e-05,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8288321010768414,
      "num_tokens": 633776341.0,
      "step": 19857
    },
    {
      "entropy": 0.5624927114695311,
      "epoch": 1.8270452728851474,
      "grad_norm": 0.1492282599210739,
      "learning_rate": 3.91020333057319e-05,
      "loss": 0.54,
      "mean_token_accuracy": 0.8291319757699966,
      "num_tokens": 633807851.0,
      "step": 19858
    },
    {
      "entropy": 0.6143679097294807,
      "epoch": 1.8271372792565885,
      "grad_norm": 0.15611213445663452,
      "learning_rate": 3.909896647959028e-05,
      "loss": 0.6082,
      "mean_token_accuracy": 0.8114490322768688,
      "num_tokens": 633839649.0,
      "step": 19859
    },
    {
      "entropy": 0.6479475721716881,
      "epoch": 1.8272292856280297,
      "grad_norm": 0.16031794250011444,
      "learning_rate": 3.9095899653448646e-05,
      "loss": 0.6419,
      "mean_token_accuracy": 0.8035708703100681,
      "num_tokens": 633871085.0,
      "step": 19860
    },
    {
      "entropy": 0.7030995711684227,
      "epoch": 1.8273212919994708,
      "grad_norm": 0.16671623289585114,
      "learning_rate": 3.909283282730703e-05,
      "loss": 0.6941,
      "mean_token_accuracy": 0.7903513014316559,
      "num_tokens": 633901865.0,
      "step": 19861
    },
    {
      "entropy": 0.5566661767661572,
      "epoch": 1.8274132983709122,
      "grad_norm": 0.15358774363994598,
      "learning_rate": 3.9089766001165395e-05,
      "loss": 0.553,
      "mean_token_accuracy": 0.8272085748612881,
      "num_tokens": 633933503.0,
      "step": 19862
    },
    {
      "entropy": 0.5136841328348964,
      "epoch": 1.8275053047423535,
      "grad_norm": 0.14496691524982452,
      "learning_rate": 3.908669917502377e-05,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8413980454206467,
      "num_tokens": 633965299.0,
      "step": 19863
    },
    {
      "entropy": 0.5686427191831172,
      "epoch": 1.8275973111137946,
      "grad_norm": 0.14838187396526337,
      "learning_rate": 3.9083632348882145e-05,
      "loss": 0.5606,
      "mean_token_accuracy": 0.8250141404569149,
      "num_tokens": 633997463.0,
      "step": 19864
    },
    {
      "entropy": 0.6434519793838263,
      "epoch": 1.8276893174852358,
      "grad_norm": 0.15733298659324646,
      "learning_rate": 3.908056552274052e-05,
      "loss": 0.658,
      "mean_token_accuracy": 0.8011999577283859,
      "num_tokens": 634029737.0,
      "step": 19865
    },
    {
      "entropy": 0.5492911636829376,
      "epoch": 1.827781323856677,
      "grad_norm": 0.1483149379491806,
      "learning_rate": 3.9077498696598894e-05,
      "loss": 0.543,
      "mean_token_accuracy": 0.8321733139455318,
      "num_tokens": 634061806.0,
      "step": 19866
    },
    {
      "entropy": 0.5113993743434548,
      "epoch": 1.8278733302281183,
      "grad_norm": 0.15125855803489685,
      "learning_rate": 3.907443187045727e-05,
      "loss": 0.5035,
      "mean_token_accuracy": 0.8402430824935436,
      "num_tokens": 634092813.0,
      "step": 19867
    },
    {
      "entropy": 0.4981024111621082,
      "epoch": 1.8279653365995596,
      "grad_norm": 0.13994039595127106,
      "learning_rate": 3.9071365044315636e-05,
      "loss": 0.4839,
      "mean_token_accuracy": 0.848338283598423,
      "num_tokens": 634125226.0,
      "step": 19868
    },
    {
      "entropy": 0.5927440710365772,
      "epoch": 1.8280573429710008,
      "grad_norm": 0.1574888527393341,
      "learning_rate": 3.906829821817402e-05,
      "loss": 0.6008,
      "mean_token_accuracy": 0.8144616708159447,
      "num_tokens": 634157741.0,
      "step": 19869
    },
    {
      "entropy": 0.6052021030336618,
      "epoch": 1.8281493493424419,
      "grad_norm": 0.1574500948190689,
      "learning_rate": 3.9065231392032386e-05,
      "loss": 0.5961,
      "mean_token_accuracy": 0.8132549785077572,
      "num_tokens": 634189471.0,
      "step": 19870
    },
    {
      "entropy": 0.636259245686233,
      "epoch": 1.828241355713883,
      "grad_norm": 0.15421156585216522,
      "learning_rate": 3.906216456589076e-05,
      "loss": 0.6264,
      "mean_token_accuracy": 0.8061500899493694,
      "num_tokens": 634221817.0,
      "step": 19871
    },
    {
      "entropy": 0.6591760646551847,
      "epoch": 1.8283333620853244,
      "grad_norm": 0.15373198688030243,
      "learning_rate": 3.9059097739749135e-05,
      "loss": 0.65,
      "mean_token_accuracy": 0.8036130182445049,
      "num_tokens": 634253928.0,
      "step": 19872
    },
    {
      "entropy": 0.6070649232715368,
      "epoch": 1.8284253684567657,
      "grad_norm": 0.147611603140831,
      "learning_rate": 3.905603091360751e-05,
      "loss": 0.5856,
      "mean_token_accuracy": 0.8204986341297626,
      "num_tokens": 634286196.0,
      "step": 19873
    },
    {
      "entropy": 0.7054956052452326,
      "epoch": 1.8285173748282069,
      "grad_norm": 0.16134116053581238,
      "learning_rate": 3.9052964087465884e-05,
      "loss": 0.6893,
      "mean_token_accuracy": 0.7880976721644402,
      "num_tokens": 634317911.0,
      "step": 19874
    },
    {
      "entropy": 0.4933500229381025,
      "epoch": 1.828609381199648,
      "grad_norm": 0.14188970625400543,
      "learning_rate": 3.904989726132426e-05,
      "loss": 0.4836,
      "mean_token_accuracy": 0.8523126766085625,
      "num_tokens": 634349856.0,
      "step": 19875
    },
    {
      "entropy": 0.6059444788843393,
      "epoch": 1.8287013875710891,
      "grad_norm": 0.1579563021659851,
      "learning_rate": 3.904683043518263e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8124179765582085,
      "num_tokens": 634381506.0,
      "step": 19876
    },
    {
      "entropy": 0.5511695668101311,
      "epoch": 1.8287933939425305,
      "grad_norm": 0.15095946192741394,
      "learning_rate": 3.904376360904101e-05,
      "loss": 0.5481,
      "mean_token_accuracy": 0.8327987343072891,
      "num_tokens": 634413193.0,
      "step": 19877
    },
    {
      "entropy": 0.572994421236217,
      "epoch": 1.8288854003139718,
      "grad_norm": 0.14503099024295807,
      "learning_rate": 3.9040696782899376e-05,
      "loss": 0.5554,
      "mean_token_accuracy": 0.826188113540411,
      "num_tokens": 634445337.0,
      "step": 19878
    },
    {
      "entropy": 0.620670079253614,
      "epoch": 1.828977406685413,
      "grad_norm": 0.1527947634458542,
      "learning_rate": 3.903762995675775e-05,
      "loss": 0.6081,
      "mean_token_accuracy": 0.8122021853923798,
      "num_tokens": 634476641.0,
      "step": 19879
    },
    {
      "entropy": 0.7121052243746817,
      "epoch": 1.829069413056854,
      "grad_norm": 0.1664702594280243,
      "learning_rate": 3.9034563130616125e-05,
      "loss": 0.7028,
      "mean_token_accuracy": 0.7835039906203747,
      "num_tokens": 634507421.0,
      "step": 19880
    },
    {
      "entropy": 0.6365646850317717,
      "epoch": 1.8291614194282952,
      "grad_norm": 0.15492604672908783,
      "learning_rate": 3.90314963044745e-05,
      "loss": 0.6268,
      "mean_token_accuracy": 0.8068000115454197,
      "num_tokens": 634539894.0,
      "step": 19881
    },
    {
      "entropy": 0.49801459163427353,
      "epoch": 1.8292534257997366,
      "grad_norm": 0.14262603223323822,
      "learning_rate": 3.9028429478332875e-05,
      "loss": 0.4796,
      "mean_token_accuracy": 0.849945243448019,
      "num_tokens": 634571605.0,
      "step": 19882
    },
    {
      "entropy": 0.6985286045819521,
      "epoch": 1.829345432171178,
      "grad_norm": 0.16070663928985596,
      "learning_rate": 3.902536265219125e-05,
      "loss": 0.6765,
      "mean_token_accuracy": 0.7940074093639851,
      "num_tokens": 634603173.0,
      "step": 19883
    },
    {
      "entropy": 0.5424622590653598,
      "epoch": 1.829437438542619,
      "grad_norm": 0.14792658388614655,
      "learning_rate": 3.9022295826049624e-05,
      "loss": 0.5395,
      "mean_token_accuracy": 0.8313979208469391,
      "num_tokens": 634635084.0,
      "step": 19884
    },
    {
      "entropy": 0.5323533955961466,
      "epoch": 1.8295294449140602,
      "grad_norm": 0.14811761677265167,
      "learning_rate": 3.9019228999908e-05,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8366335853934288,
      "num_tokens": 634667181.0,
      "step": 19885
    },
    {
      "entropy": 0.4924714853987098,
      "epoch": 1.8296214512855014,
      "grad_norm": 0.14804711937904358,
      "learning_rate": 3.9016162173766367e-05,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8447805382311344,
      "num_tokens": 634698835.0,
      "step": 19886
    },
    {
      "entropy": 0.5339353019371629,
      "epoch": 1.8297134576569427,
      "grad_norm": 0.15032202005386353,
      "learning_rate": 3.901309534762475e-05,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8306933417916298,
      "num_tokens": 634730921.0,
      "step": 19887
    },
    {
      "entropy": 0.6655202452093363,
      "epoch": 1.829805464028384,
      "grad_norm": 0.16846531629562378,
      "learning_rate": 3.9010028521483116e-05,
      "loss": 0.6635,
      "mean_token_accuracy": 0.7950512394309044,
      "num_tokens": 634762258.0,
      "step": 19888
    },
    {
      "entropy": 0.6981137059628963,
      "epoch": 1.8298974703998252,
      "grad_norm": 0.16780465841293335,
      "learning_rate": 3.900696169534149e-05,
      "loss": 0.6881,
      "mean_token_accuracy": 0.7906225547194481,
      "num_tokens": 634794517.0,
      "step": 19889
    },
    {
      "entropy": 0.5242338189855218,
      "epoch": 1.8299894767712663,
      "grad_norm": 0.148575097322464,
      "learning_rate": 3.900389486919987e-05,
      "loss": 0.5113,
      "mean_token_accuracy": 0.840966984629631,
      "num_tokens": 634827001.0,
      "step": 19890
    },
    {
      "entropy": 0.6884240098297596,
      "epoch": 1.8300814831427077,
      "grad_norm": 0.16116617619991302,
      "learning_rate": 3.900082804305824e-05,
      "loss": 0.6889,
      "mean_token_accuracy": 0.7962048463523388,
      "num_tokens": 634858829.0,
      "step": 19891
    },
    {
      "entropy": 0.6852407548576593,
      "epoch": 1.8301734895141488,
      "grad_norm": 0.15616671741008759,
      "learning_rate": 3.8997761216916614e-05,
      "loss": 0.6761,
      "mean_token_accuracy": 0.794133760035038,
      "num_tokens": 634891142.0,
      "step": 19892
    },
    {
      "entropy": 0.5297086043283343,
      "epoch": 1.8302654958855902,
      "grad_norm": 0.14439569413661957,
      "learning_rate": 3.899469439077499e-05,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8385345488786697,
      "num_tokens": 634923857.0,
      "step": 19893
    },
    {
      "entropy": 0.7422567848116159,
      "epoch": 1.8303575022570313,
      "grad_norm": 0.16355633735656738,
      "learning_rate": 3.8991627564633364e-05,
      "loss": 0.7218,
      "mean_token_accuracy": 0.7823065780103207,
      "num_tokens": 634955701.0,
      "step": 19894
    },
    {
      "entropy": 0.5537389335222542,
      "epoch": 1.8304495086284724,
      "grad_norm": 0.1456032693386078,
      "learning_rate": 3.898856073849174e-05,
      "loss": 0.5429,
      "mean_token_accuracy": 0.8303555808961391,
      "num_tokens": 634987434.0,
      "step": 19895
    },
    {
      "entropy": 0.6102211959660053,
      "epoch": 1.8305415149999138,
      "grad_norm": 0.1483270674943924,
      "learning_rate": 3.898549391235011e-05,
      "loss": 0.595,
      "mean_token_accuracy": 0.8165906667709351,
      "num_tokens": 635019688.0,
      "step": 19896
    },
    {
      "entropy": 0.5811237078160048,
      "epoch": 1.830633521371355,
      "grad_norm": 0.152780681848526,
      "learning_rate": 3.898242708620848e-05,
      "loss": 0.5756,
      "mean_token_accuracy": 0.8199742361903191,
      "num_tokens": 635051255.0,
      "step": 19897
    },
    {
      "entropy": 0.6259737852960825,
      "epoch": 1.8307255277427963,
      "grad_norm": 0.1507588028907776,
      "learning_rate": 3.897936026006686e-05,
      "loss": 0.6069,
      "mean_token_accuracy": 0.8112322576344013,
      "num_tokens": 635083078.0,
      "step": 19898
    },
    {
      "entropy": 0.5214025187306106,
      "epoch": 1.8308175341142374,
      "grad_norm": 0.1449107676744461,
      "learning_rate": 3.897629343392523e-05,
      "loss": 0.497,
      "mean_token_accuracy": 0.8441835530102253,
      "num_tokens": 635114341.0,
      "step": 19899
    },
    {
      "entropy": 0.5563333239406347,
      "epoch": 1.8309095404856786,
      "grad_norm": 0.15792666375637054,
      "learning_rate": 3.897322660778361e-05,
      "loss": 0.5636,
      "mean_token_accuracy": 0.8272029533982277,
      "num_tokens": 635145798.0,
      "step": 19900
    },
    {
      "entropy": 0.6774523369967937,
      "epoch": 1.83100154685712,
      "grad_norm": 0.15626291930675507,
      "learning_rate": 3.897015978164198e-05,
      "loss": 0.6702,
      "mean_token_accuracy": 0.7950319424271584,
      "num_tokens": 635178198.0,
      "step": 19901
    },
    {
      "entropy": 0.5485848388634622,
      "epoch": 1.831093553228561,
      "grad_norm": 0.146332249045372,
      "learning_rate": 3.8967092955500354e-05,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8371820971369743,
      "num_tokens": 635210642.0,
      "step": 19902
    },
    {
      "entropy": 0.5278429575264454,
      "epoch": 1.8311855596000024,
      "grad_norm": 0.15099318325519562,
      "learning_rate": 3.896402612935873e-05,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8422920852899551,
      "num_tokens": 635241040.0,
      "step": 19903
    },
    {
      "entropy": 0.6375988777726889,
      "epoch": 1.8312775659714435,
      "grad_norm": 0.15642240643501282,
      "learning_rate": 3.8960959303217103e-05,
      "loss": 0.6274,
      "mean_token_accuracy": 0.8069791831076145,
      "num_tokens": 635272307.0,
      "step": 19904
    },
    {
      "entropy": 0.5606421157717705,
      "epoch": 1.8313695723428847,
      "grad_norm": 0.1502746045589447,
      "learning_rate": 3.895789247707548e-05,
      "loss": 0.5543,
      "mean_token_accuracy": 0.8299155794084072,
      "num_tokens": 635304228.0,
      "step": 19905
    },
    {
      "entropy": 0.5330839795060456,
      "epoch": 1.831461578714326,
      "grad_norm": 0.15060746669769287,
      "learning_rate": 3.895482565093385e-05,
      "loss": 0.5145,
      "mean_token_accuracy": 0.8386052921414375,
      "num_tokens": 635336163.0,
      "step": 19906
    },
    {
      "entropy": 0.6470966637134552,
      "epoch": 1.8315535850857672,
      "grad_norm": 0.1558593511581421,
      "learning_rate": 3.895175882479222e-05,
      "loss": 0.6197,
      "mean_token_accuracy": 0.8114648759365082,
      "num_tokens": 635367815.0,
      "step": 19907
    },
    {
      "entropy": 0.5857968553900719,
      "epoch": 1.8316455914572085,
      "grad_norm": 0.15508174896240234,
      "learning_rate": 3.89486919986506e-05,
      "loss": 0.5702,
      "mean_token_accuracy": 0.8255869187414646,
      "num_tokens": 635399867.0,
      "step": 19908
    },
    {
      "entropy": 0.6263441108167171,
      "epoch": 1.8317375978286496,
      "grad_norm": 0.15409457683563232,
      "learning_rate": 3.894562517250897e-05,
      "loss": 0.62,
      "mean_token_accuracy": 0.8082178458571434,
      "num_tokens": 635432242.0,
      "step": 19909
    },
    {
      "entropy": 0.6011911612004042,
      "epoch": 1.8318296042000908,
      "grad_norm": 0.15030334889888763,
      "learning_rate": 3.8942558346367345e-05,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8181871436536312,
      "num_tokens": 635463630.0,
      "step": 19910
    },
    {
      "entropy": 0.6362036969512701,
      "epoch": 1.8319216105715321,
      "grad_norm": 0.16115646064281464,
      "learning_rate": 3.893949152022572e-05,
      "loss": 0.6196,
      "mean_token_accuracy": 0.8088785633444786,
      "num_tokens": 635495291.0,
      "step": 19911
    },
    {
      "entropy": 0.43092378904111683,
      "epoch": 1.8320136169429733,
      "grad_norm": 0.13816939294338226,
      "learning_rate": 3.8936424694084094e-05,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8669088929891586,
      "num_tokens": 635528059.0,
      "step": 19912
    },
    {
      "entropy": 0.6273954138159752,
      "epoch": 1.8321056233144146,
      "grad_norm": 0.15417280793190002,
      "learning_rate": 3.893335786794247e-05,
      "loss": 0.6221,
      "mean_token_accuracy": 0.8105021975934505,
      "num_tokens": 635560399.0,
      "step": 19913
    },
    {
      "entropy": 0.5455795400775969,
      "epoch": 1.8321976296858558,
      "grad_norm": 0.14414386451244354,
      "learning_rate": 3.893029104180084e-05,
      "loss": 0.538,
      "mean_token_accuracy": 0.8359489440917969,
      "num_tokens": 635593167.0,
      "step": 19914
    },
    {
      "entropy": 0.6564499959349632,
      "epoch": 1.8322896360572969,
      "grad_norm": 0.1516200751066208,
      "learning_rate": 3.892722421565921e-05,
      "loss": 0.6419,
      "mean_token_accuracy": 0.8061893880367279,
      "num_tokens": 635625428.0,
      "step": 19915
    },
    {
      "entropy": 0.5806982866488397,
      "epoch": 1.8323816424287382,
      "grad_norm": 0.1594536453485489,
      "learning_rate": 3.892415738951759e-05,
      "loss": 0.568,
      "mean_token_accuracy": 0.8259316757321358,
      "num_tokens": 635657146.0,
      "step": 19916
    },
    {
      "entropy": 0.5817151898518205,
      "epoch": 1.8324736488001794,
      "grad_norm": 0.1556638479232788,
      "learning_rate": 3.892109056337596e-05,
      "loss": 0.573,
      "mean_token_accuracy": 0.824923150241375,
      "num_tokens": 635689283.0,
      "step": 19917
    },
    {
      "entropy": 0.5731867151334882,
      "epoch": 1.8325656551716207,
      "grad_norm": 0.15018215775489807,
      "learning_rate": 3.8918023737234335e-05,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8280707262456417,
      "num_tokens": 635720546.0,
      "step": 19918
    },
    {
      "entropy": 0.5322421346791089,
      "epoch": 1.8326576615430619,
      "grad_norm": 0.1497965008020401,
      "learning_rate": 3.891495691109271e-05,
      "loss": 0.5294,
      "mean_token_accuracy": 0.8386739380657673,
      "num_tokens": 635752328.0,
      "step": 19919
    },
    {
      "entropy": 0.6055553089827299,
      "epoch": 1.832749667914503,
      "grad_norm": 0.16074508428573608,
      "learning_rate": 3.8911890084951084e-05,
      "loss": 0.5969,
      "mean_token_accuracy": 0.8149608820676804,
      "num_tokens": 635783911.0,
      "step": 19920
    },
    {
      "entropy": 0.524290207773447,
      "epoch": 1.8328416742859444,
      "grad_norm": 0.14358218014240265,
      "learning_rate": 3.8908823258809466e-05,
      "loss": 0.5103,
      "mean_token_accuracy": 0.8386290036141872,
      "num_tokens": 635815756.0,
      "step": 19921
    },
    {
      "entropy": 0.6050272230058908,
      "epoch": 1.8329336806573855,
      "grad_norm": 0.1498749852180481,
      "learning_rate": 3.8905756432667834e-05,
      "loss": 0.5987,
      "mean_token_accuracy": 0.8171642012894154,
      "num_tokens": 635848056.0,
      "step": 19922
    },
    {
      "entropy": 0.5768037436064333,
      "epoch": 1.8330256870288268,
      "grad_norm": 0.14696268737316132,
      "learning_rate": 3.890268960652621e-05,
      "loss": 0.5629,
      "mean_token_accuracy": 0.8263318277895451,
      "num_tokens": 635880384.0,
      "step": 19923
    },
    {
      "entropy": 0.5013795564882457,
      "epoch": 1.833117693400268,
      "grad_norm": 0.14353983104228973,
      "learning_rate": 3.889962278038458e-05,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8460786752402782,
      "num_tokens": 635912774.0,
      "step": 19924
    },
    {
      "entropy": 0.7038442641496658,
      "epoch": 1.8332096997717091,
      "grad_norm": 0.1603655070066452,
      "learning_rate": 3.889655595424296e-05,
      "loss": 0.7075,
      "mean_token_accuracy": 0.7843307256698608,
      "num_tokens": 635945123.0,
      "step": 19925
    },
    {
      "entropy": 0.6244057454168797,
      "epoch": 1.8333017061431505,
      "grad_norm": 0.1572864055633545,
      "learning_rate": 3.889348912810133e-05,
      "loss": 0.6118,
      "mean_token_accuracy": 0.8114470429718494,
      "num_tokens": 635976759.0,
      "step": 19926
    },
    {
      "entropy": 0.530144490301609,
      "epoch": 1.8333937125145916,
      "grad_norm": 0.1476181298494339,
      "learning_rate": 3.889042230195971e-05,
      "loss": 0.5145,
      "mean_token_accuracy": 0.8369509242475033,
      "num_tokens": 636008200.0,
      "step": 19927
    },
    {
      "entropy": 0.7696761153638363,
      "epoch": 1.833485718886033,
      "grad_norm": 0.16552557051181793,
      "learning_rate": 3.8887355475818075e-05,
      "loss": 0.76,
      "mean_token_accuracy": 0.7722381912171841,
      "num_tokens": 636039828.0,
      "step": 19928
    },
    {
      "entropy": 0.5965676391497254,
      "epoch": 1.833577725257474,
      "grad_norm": 0.15509961545467377,
      "learning_rate": 3.8884288649676456e-05,
      "loss": 0.602,
      "mean_token_accuracy": 0.8134764544665813,
      "num_tokens": 636071163.0,
      "step": 19929
    },
    {
      "entropy": 0.5062624104321003,
      "epoch": 1.8336697316289152,
      "grad_norm": 0.1485663652420044,
      "learning_rate": 3.8881221823534824e-05,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8493199124932289,
      "num_tokens": 636102932.0,
      "step": 19930
    },
    {
      "entropy": 0.6634527230635285,
      "epoch": 1.8337617380003566,
      "grad_norm": 0.15717777609825134,
      "learning_rate": 3.88781549973932e-05,
      "loss": 0.6487,
      "mean_token_accuracy": 0.8009742870926857,
      "num_tokens": 636134194.0,
      "step": 19931
    },
    {
      "entropy": 0.5992861613631248,
      "epoch": 1.8338537443717977,
      "grad_norm": 0.15653099119663239,
      "learning_rate": 3.887508817125157e-05,
      "loss": 0.5932,
      "mean_token_accuracy": 0.8201137073338032,
      "num_tokens": 636165787.0,
      "step": 19932
    },
    {
      "entropy": 0.5619413610547781,
      "epoch": 1.833945750743239,
      "grad_norm": 0.14748035371303558,
      "learning_rate": 3.887202134510995e-05,
      "loss": 0.5491,
      "mean_token_accuracy": 0.8323351554572582,
      "num_tokens": 636198090.0,
      "step": 19933
    },
    {
      "entropy": 0.5921482006087899,
      "epoch": 1.8340377571146802,
      "grad_norm": 0.15092812478542328,
      "learning_rate": 3.886895451896832e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.8148861527442932,
      "num_tokens": 636230212.0,
      "step": 19934
    },
    {
      "entropy": 0.6949867811053991,
      "epoch": 1.8341297634861213,
      "grad_norm": 0.16532468795776367,
      "learning_rate": 3.88658876928267e-05,
      "loss": 0.6946,
      "mean_token_accuracy": 0.7918732538819313,
      "num_tokens": 636262306.0,
      "step": 19935
    },
    {
      "entropy": 0.5365142934024334,
      "epoch": 1.8342217698575627,
      "grad_norm": 0.15413019061088562,
      "learning_rate": 3.8862820866685065e-05,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8383426889777184,
      "num_tokens": 636293918.0,
      "step": 19936
    },
    {
      "entropy": 0.7241945769637823,
      "epoch": 1.8343137762290038,
      "grad_norm": 0.16206510365009308,
      "learning_rate": 3.8859754040543447e-05,
      "loss": 0.7081,
      "mean_token_accuracy": 0.7848675511777401,
      "num_tokens": 636324904.0,
      "step": 19937
    },
    {
      "entropy": 0.5981542076915503,
      "epoch": 1.8344057826004452,
      "grad_norm": 0.15417726337909698,
      "learning_rate": 3.8856687214401814e-05,
      "loss": 0.5806,
      "mean_token_accuracy": 0.8212191164493561,
      "num_tokens": 636356567.0,
      "step": 19938
    },
    {
      "entropy": 0.5821943753398955,
      "epoch": 1.8344977889718863,
      "grad_norm": 0.16265147924423218,
      "learning_rate": 3.885362038826019e-05,
      "loss": 0.5668,
      "mean_token_accuracy": 0.8273236155509949,
      "num_tokens": 636388623.0,
      "step": 19939
    },
    {
      "entropy": 0.6705471482127905,
      "epoch": 1.8345897953433274,
      "grad_norm": 0.15928290784358978,
      "learning_rate": 3.8850553562118564e-05,
      "loss": 0.6659,
      "mean_token_accuracy": 0.8002046123147011,
      "num_tokens": 636420494.0,
      "step": 19940
    },
    {
      "entropy": 0.5860496051609516,
      "epoch": 1.8346818017147688,
      "grad_norm": 0.15391896665096283,
      "learning_rate": 3.884748673597694e-05,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8278233595192432,
      "num_tokens": 636452792.0,
      "step": 19941
    },
    {
      "entropy": 0.658867159858346,
      "epoch": 1.83477380808621,
      "grad_norm": 0.15264850854873657,
      "learning_rate": 3.884441990983531e-05,
      "loss": 0.6373,
      "mean_token_accuracy": 0.7997621223330498,
      "num_tokens": 636485196.0,
      "step": 19942
    },
    {
      "entropy": 0.5532033052295446,
      "epoch": 1.8348658144576513,
      "grad_norm": 0.15188267827033997,
      "learning_rate": 3.884135308369369e-05,
      "loss": 0.5352,
      "mean_token_accuracy": 0.8320564478635788,
      "num_tokens": 636517505.0,
      "step": 19943
    },
    {
      "entropy": 0.640749295707792,
      "epoch": 1.8349578208290924,
      "grad_norm": 0.1569356769323349,
      "learning_rate": 3.883828625755206e-05,
      "loss": 0.6204,
      "mean_token_accuracy": 0.8109818957746029,
      "num_tokens": 636549402.0,
      "step": 19944
    },
    {
      "entropy": 0.5781880440190434,
      "epoch": 1.8350498272005336,
      "grad_norm": 0.14844264090061188,
      "learning_rate": 3.883521943141044e-05,
      "loss": 0.5546,
      "mean_token_accuracy": 0.8271820172667503,
      "num_tokens": 636581285.0,
      "step": 19945
    },
    {
      "entropy": 0.6867942325770855,
      "epoch": 1.835141833571975,
      "grad_norm": 0.16522197425365448,
      "learning_rate": 3.8832152605268805e-05,
      "loss": 0.6715,
      "mean_token_accuracy": 0.79599953815341,
      "num_tokens": 636611936.0,
      "step": 19946
    },
    {
      "entropy": 0.4972334997728467,
      "epoch": 1.835233839943416,
      "grad_norm": 0.14738602936267853,
      "learning_rate": 3.8829085779127186e-05,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8480436615645885,
      "num_tokens": 636643615.0,
      "step": 19947
    },
    {
      "entropy": 0.49734822753816843,
      "epoch": 1.8353258463148574,
      "grad_norm": 0.1446876972913742,
      "learning_rate": 3.8826018952985554e-05,
      "loss": 0.4961,
      "mean_token_accuracy": 0.8476453311741352,
      "num_tokens": 636676355.0,
      "step": 19948
    },
    {
      "entropy": 0.6029567485675216,
      "epoch": 1.8354178526862985,
      "grad_norm": 0.15714165568351746,
      "learning_rate": 3.882295212684393e-05,
      "loss": 0.5956,
      "mean_token_accuracy": 0.8154157884418964,
      "num_tokens": 636707682.0,
      "step": 19949
    },
    {
      "entropy": 0.6645949527155608,
      "epoch": 1.8355098590577397,
      "grad_norm": 0.1569555699825287,
      "learning_rate": 3.881988530070231e-05,
      "loss": 0.6578,
      "mean_token_accuracy": 0.8006493113934994,
      "num_tokens": 636740295.0,
      "step": 19950
    },
    {
      "entropy": 0.5662939269095659,
      "epoch": 1.835601865429181,
      "grad_norm": 0.152163565158844,
      "learning_rate": 3.881681847456068e-05,
      "loss": 0.5619,
      "mean_token_accuracy": 0.8252880200743675,
      "num_tokens": 636771614.0,
      "step": 19951
    },
    {
      "entropy": 0.720666790381074,
      "epoch": 1.8356938718006222,
      "grad_norm": 0.16050662100315094,
      "learning_rate": 3.881375164841905e-05,
      "loss": 0.723,
      "mean_token_accuracy": 0.7776171565055847,
      "num_tokens": 636803592.0,
      "step": 19952
    },
    {
      "entropy": 0.537034903652966,
      "epoch": 1.8357858781720635,
      "grad_norm": 0.15000700950622559,
      "learning_rate": 3.881068482227743e-05,
      "loss": 0.5302,
      "mean_token_accuracy": 0.8372335098683834,
      "num_tokens": 636835420.0,
      "step": 19953
    },
    {
      "entropy": 0.5321205202490091,
      "epoch": 1.8358778845435046,
      "grad_norm": 0.14884401857852936,
      "learning_rate": 3.88076179961358e-05,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8395555950701237,
      "num_tokens": 636866872.0,
      "step": 19954
    },
    {
      "entropy": 0.6467593414708972,
      "epoch": 1.8359698909149458,
      "grad_norm": 0.155996635556221,
      "learning_rate": 3.880455116999418e-05,
      "loss": 0.6294,
      "mean_token_accuracy": 0.8080744594335556,
      "num_tokens": 636899121.0,
      "step": 19955
    },
    {
      "entropy": 0.5809578653424978,
      "epoch": 1.8360618972863871,
      "grad_norm": 0.14877894520759583,
      "learning_rate": 3.880148434385255e-05,
      "loss": 0.5791,
      "mean_token_accuracy": 0.8206145688891411,
      "num_tokens": 636931826.0,
      "step": 19956
    },
    {
      "entropy": 0.584531806409359,
      "epoch": 1.8361539036578283,
      "grad_norm": 0.16399161517620087,
      "learning_rate": 3.879841751771092e-05,
      "loss": 0.5846,
      "mean_token_accuracy": 0.8205420635640621,
      "num_tokens": 636963690.0,
      "step": 19957
    },
    {
      "entropy": 0.737165791913867,
      "epoch": 1.8362459100292696,
      "grad_norm": 0.16391362249851227,
      "learning_rate": 3.87953506915693e-05,
      "loss": 0.728,
      "mean_token_accuracy": 0.7789775840938091,
      "num_tokens": 636996002.0,
      "step": 19958
    },
    {
      "entropy": 0.5468871053308249,
      "epoch": 1.8363379164007108,
      "grad_norm": 0.1532467156648636,
      "learning_rate": 3.879228386542767e-05,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8337397538125515,
      "num_tokens": 637027902.0,
      "step": 19959
    },
    {
      "entropy": 0.5514256693422794,
      "epoch": 1.836429922772152,
      "grad_norm": 0.14317433536052704,
      "learning_rate": 3.878921703928605e-05,
      "loss": 0.5427,
      "mean_token_accuracy": 0.8301882036030293,
      "num_tokens": 637060038.0,
      "step": 19960
    },
    {
      "entropy": 0.5842607859522104,
      "epoch": 1.8365219291435932,
      "grad_norm": 0.15345118939876556,
      "learning_rate": 3.878615021314442e-05,
      "loss": 0.5675,
      "mean_token_accuracy": 0.821783360093832,
      "num_tokens": 637090981.0,
      "step": 19961
    },
    {
      "entropy": 0.5854823109693825,
      "epoch": 1.8366139355150344,
      "grad_norm": 0.15327025949954987,
      "learning_rate": 3.878308338700279e-05,
      "loss": 0.5754,
      "mean_token_accuracy": 0.8227276839315891,
      "num_tokens": 637122762.0,
      "step": 19962
    },
    {
      "entropy": 0.6582710798829794,
      "epoch": 1.8367059418864757,
      "grad_norm": 0.16491541266441345,
      "learning_rate": 3.878001656086117e-05,
      "loss": 0.645,
      "mean_token_accuracy": 0.8076563365757465,
      "num_tokens": 637154558.0,
      "step": 19963
    },
    {
      "entropy": 0.5142359044402838,
      "epoch": 1.8367979482579169,
      "grad_norm": 0.14866963028907776,
      "learning_rate": 3.877694973471954e-05,
      "loss": 0.4995,
      "mean_token_accuracy": 0.8427528627216816,
      "num_tokens": 637186185.0,
      "step": 19964
    },
    {
      "entropy": 0.7405131850391626,
      "epoch": 1.836889954629358,
      "grad_norm": 0.1655406802892685,
      "learning_rate": 3.8773882908577916e-05,
      "loss": 0.718,
      "mean_token_accuracy": 0.7778990864753723,
      "num_tokens": 637217796.0,
      "step": 19965
    },
    {
      "entropy": 0.6547384429723024,
      "epoch": 1.8369819610007994,
      "grad_norm": 0.15849150717258453,
      "learning_rate": 3.877081608243629e-05,
      "loss": 0.6507,
      "mean_token_accuracy": 0.8002000637352467,
      "num_tokens": 637249551.0,
      "step": 19966
    },
    {
      "entropy": 0.6092281108722091,
      "epoch": 1.8370739673722405,
      "grad_norm": 0.15124788880348206,
      "learning_rate": 3.876774925629466e-05,
      "loss": 0.5943,
      "mean_token_accuracy": 0.8173546679317951,
      "num_tokens": 637282319.0,
      "step": 19967
    },
    {
      "entropy": 0.7372710891067982,
      "epoch": 1.8371659737436818,
      "grad_norm": 0.16100476682186127,
      "learning_rate": 3.876468243015304e-05,
      "loss": 0.7245,
      "mean_token_accuracy": 0.7770933620631695,
      "num_tokens": 637313899.0,
      "step": 19968
    },
    {
      "entropy": 0.7127416152507067,
      "epoch": 1.837257980115123,
      "grad_norm": 0.16429732739925385,
      "learning_rate": 3.876161560401141e-05,
      "loss": 0.6916,
      "mean_token_accuracy": 0.7903903014957905,
      "num_tokens": 637345679.0,
      "step": 19969
    },
    {
      "entropy": 0.6560510261915624,
      "epoch": 1.8373499864865641,
      "grad_norm": 0.15468113124370575,
      "learning_rate": 3.875854877786978e-05,
      "loss": 0.6478,
      "mean_token_accuracy": 0.8030033186078072,
      "num_tokens": 637378355.0,
      "step": 19970
    },
    {
      "entropy": 0.5092473179101944,
      "epoch": 1.8374419928580055,
      "grad_norm": 0.14198949933052063,
      "learning_rate": 3.875548195172816e-05,
      "loss": 0.497,
      "mean_token_accuracy": 0.8430809825658798,
      "num_tokens": 637410310.0,
      "step": 19971
    },
    {
      "entropy": 0.609218793688342,
      "epoch": 1.8375339992294466,
      "grad_norm": 0.14809486269950867,
      "learning_rate": 3.875241512558653e-05,
      "loss": 0.5993,
      "mean_token_accuracy": 0.815590787678957,
      "num_tokens": 637442409.0,
      "step": 19972
    },
    {
      "entropy": 0.6340678576380014,
      "epoch": 1.837626005600888,
      "grad_norm": 0.15688544511795044,
      "learning_rate": 3.874934829944491e-05,
      "loss": 0.6307,
      "mean_token_accuracy": 0.8080394044518471,
      "num_tokens": 637474376.0,
      "step": 19973
    },
    {
      "entropy": 0.6284843320026994,
      "epoch": 1.837718011972329,
      "grad_norm": 0.15765517950057983,
      "learning_rate": 3.874628147330328e-05,
      "loss": 0.6161,
      "mean_token_accuracy": 0.8119017891585827,
      "num_tokens": 637505928.0,
      "step": 19974
    },
    {
      "entropy": 0.6259815394878387,
      "epoch": 1.8378100183437702,
      "grad_norm": 0.15726393461227417,
      "learning_rate": 3.874321464716165e-05,
      "loss": 0.616,
      "mean_token_accuracy": 0.8102291040122509,
      "num_tokens": 637538235.0,
      "step": 19975
    },
    {
      "entropy": 0.559429288841784,
      "epoch": 1.8379020247152116,
      "grad_norm": 0.15151427686214447,
      "learning_rate": 3.874014782102003e-05,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8354174047708511,
      "num_tokens": 637570450.0,
      "step": 19976
    },
    {
      "entropy": 0.6835477715358138,
      "epoch": 1.8379940310866527,
      "grad_norm": 0.1606181263923645,
      "learning_rate": 3.87370809948784e-05,
      "loss": 0.6759,
      "mean_token_accuracy": 0.7942614741623402,
      "num_tokens": 637602302.0,
      "step": 19977
    },
    {
      "entropy": 0.42145349248312414,
      "epoch": 1.838086037458094,
      "grad_norm": 0.1415565013885498,
      "learning_rate": 3.873401416873677e-05,
      "loss": 0.4124,
      "mean_token_accuracy": 0.869242038577795,
      "num_tokens": 637634462.0,
      "step": 19978
    },
    {
      "entropy": 0.6490520536899567,
      "epoch": 1.8381780438295352,
      "grad_norm": 0.153804212808609,
      "learning_rate": 3.873094734259515e-05,
      "loss": 0.644,
      "mean_token_accuracy": 0.8040615394711494,
      "num_tokens": 637666091.0,
      "step": 19979
    },
    {
      "entropy": 0.5510455425828695,
      "epoch": 1.8382700502009763,
      "grad_norm": 0.148776113986969,
      "learning_rate": 3.872788051645352e-05,
      "loss": 0.5545,
      "mean_token_accuracy": 0.8309744000434875,
      "num_tokens": 637698086.0,
      "step": 19980
    },
    {
      "entropy": 0.4866012129932642,
      "epoch": 1.8383620565724177,
      "grad_norm": 0.14091703295707703,
      "learning_rate": 3.8724813690311904e-05,
      "loss": 0.473,
      "mean_token_accuracy": 0.8540927655994892,
      "num_tokens": 637730658.0,
      "step": 19981
    },
    {
      "entropy": 0.5477015478536487,
      "epoch": 1.8384540629438588,
      "grad_norm": 0.160078227519989,
      "learning_rate": 3.872174686417027e-05,
      "loss": 0.5488,
      "mean_token_accuracy": 0.8300369828939438,
      "num_tokens": 637762539.0,
      "step": 19982
    },
    {
      "entropy": 0.6882401928305626,
      "epoch": 1.8385460693153002,
      "grad_norm": 0.16227707266807556,
      "learning_rate": 3.8718680038028646e-05,
      "loss": 0.6893,
      "mean_token_accuracy": 0.792876023799181,
      "num_tokens": 637793453.0,
      "step": 19983
    },
    {
      "entropy": 0.6363444493617862,
      "epoch": 1.8386380756867413,
      "grad_norm": 0.1643504798412323,
      "learning_rate": 3.871561321188702e-05,
      "loss": 0.6383,
      "mean_token_accuracy": 0.8034000806510448,
      "num_tokens": 637825100.0,
      "step": 19984
    },
    {
      "entropy": 0.5296730250120163,
      "epoch": 1.8387300820581824,
      "grad_norm": 0.15059718489646912,
      "learning_rate": 3.8712546385745396e-05,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8413482755422592,
      "num_tokens": 637857580.0,
      "step": 19985
    },
    {
      "entropy": 0.7560716001316905,
      "epoch": 1.8388220884296238,
      "grad_norm": 0.16091720759868622,
      "learning_rate": 3.870947955960377e-05,
      "loss": 0.749,
      "mean_token_accuracy": 0.7720684632658958,
      "num_tokens": 637889849.0,
      "step": 19986
    },
    {
      "entropy": 0.7214313913136721,
      "epoch": 1.838914094801065,
      "grad_norm": 0.16007304191589355,
      "learning_rate": 3.8706412733462145e-05,
      "loss": 0.7118,
      "mean_token_accuracy": 0.782272681593895,
      "num_tokens": 637922158.0,
      "step": 19987
    },
    {
      "entropy": 0.5436559724621475,
      "epoch": 1.8390061011725063,
      "grad_norm": 0.14495135843753815,
      "learning_rate": 3.870334590732051e-05,
      "loss": 0.5252,
      "mean_token_accuracy": 0.8371280170977116,
      "num_tokens": 637954252.0,
      "step": 19988
    },
    {
      "entropy": 0.5792759377509356,
      "epoch": 1.8390981075439474,
      "grad_norm": 0.14335022866725922,
      "learning_rate": 3.8700279081178894e-05,
      "loss": 0.5703,
      "mean_token_accuracy": 0.8218590319156647,
      "num_tokens": 637986311.0,
      "step": 19989
    },
    {
      "entropy": 0.6402243226766586,
      "epoch": 1.8391901139153886,
      "grad_norm": 0.15451766550540924,
      "learning_rate": 3.869721225503726e-05,
      "loss": 0.6307,
      "mean_token_accuracy": 0.8091142028570175,
      "num_tokens": 638018966.0,
      "step": 19990
    },
    {
      "entropy": 0.651440929621458,
      "epoch": 1.83928212028683,
      "grad_norm": 0.15067271888256073,
      "learning_rate": 3.869414542889564e-05,
      "loss": 0.6331,
      "mean_token_accuracy": 0.8071469441056252,
      "num_tokens": 638051333.0,
      "step": 19991
    },
    {
      "entropy": 0.6781598478555679,
      "epoch": 1.839374126658271,
      "grad_norm": 0.1557474285364151,
      "learning_rate": 3.869107860275401e-05,
      "loss": 0.6693,
      "mean_token_accuracy": 0.7970371320843697,
      "num_tokens": 638082800.0,
      "step": 19992
    },
    {
      "entropy": 0.4963422955479473,
      "epoch": 1.8394661330297124,
      "grad_norm": 0.1458248496055603,
      "learning_rate": 3.8688011776612386e-05,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8497533798217773,
      "num_tokens": 638115106.0,
      "step": 19993
    },
    {
      "entropy": 0.6620564442127943,
      "epoch": 1.8395581394011535,
      "grad_norm": 0.15638858079910278,
      "learning_rate": 3.868494495047076e-05,
      "loss": 0.6375,
      "mean_token_accuracy": 0.8026158958673477,
      "num_tokens": 638146448.0,
      "step": 19994
    },
    {
      "entropy": 0.5679172065574676,
      "epoch": 1.8396501457725947,
      "grad_norm": 0.14979702234268188,
      "learning_rate": 3.8681878124329135e-05,
      "loss": 0.5521,
      "mean_token_accuracy": 0.827313806861639,
      "num_tokens": 638178308.0,
      "step": 19995
    },
    {
      "entropy": 0.6565613187849522,
      "epoch": 1.839742152144036,
      "grad_norm": 0.15566855669021606,
      "learning_rate": 3.86788112981875e-05,
      "loss": 0.6328,
      "mean_token_accuracy": 0.805752482265234,
      "num_tokens": 638209336.0,
      "step": 19996
    },
    {
      "entropy": 0.5285169780254364,
      "epoch": 1.8398341585154772,
      "grad_norm": 0.15008683502674103,
      "learning_rate": 3.8675744472045885e-05,
      "loss": 0.5085,
      "mean_token_accuracy": 0.8410523347556591,
      "num_tokens": 638241028.0,
      "step": 19997
    },
    {
      "entropy": 0.5408989945426583,
      "epoch": 1.8399261648869185,
      "grad_norm": 0.1581496000289917,
      "learning_rate": 3.867267764590425e-05,
      "loss": 0.5196,
      "mean_token_accuracy": 0.840527381747961,
      "num_tokens": 638273014.0,
      "step": 19998
    },
    {
      "entropy": 0.6497137639671564,
      "epoch": 1.8400181712583596,
      "grad_norm": 0.15760906040668488,
      "learning_rate": 3.866961081976263e-05,
      "loss": 0.6389,
      "mean_token_accuracy": 0.8065091297030449,
      "num_tokens": 638303475.0,
      "step": 19999
    },
    {
      "entropy": 0.6321191396564245,
      "epoch": 1.8401101776298008,
      "grad_norm": 0.15653882920742035,
      "learning_rate": 3.8666543993621e-05,
      "loss": 0.6142,
      "mean_token_accuracy": 0.8131419867277145,
      "num_tokens": 638334947.0,
      "step": 20000
    },
    {
      "entropy": 0.48379048332571983,
      "epoch": 1.8402021840012421,
      "grad_norm": 0.14875201880931854,
      "learning_rate": 3.8663477167479377e-05,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8522448278963566,
      "num_tokens": 638366727.0,
      "step": 20001
    },
    {
      "entropy": 0.6968495287001133,
      "epoch": 1.8402941903726833,
      "grad_norm": 0.16744917631149292,
      "learning_rate": 3.866041034133775e-05,
      "loss": 0.7013,
      "mean_token_accuracy": 0.7913552299141884,
      "num_tokens": 638398171.0,
      "step": 20002
    },
    {
      "entropy": 0.64611541852355,
      "epoch": 1.8403861967441246,
      "grad_norm": 0.15518105030059814,
      "learning_rate": 3.8657343515196126e-05,
      "loss": 0.6438,
      "mean_token_accuracy": 0.8029381930828094,
      "num_tokens": 638430761.0,
      "step": 20003
    },
    {
      "entropy": 0.7728232480585575,
      "epoch": 1.8404782031155658,
      "grad_norm": 0.1687997579574585,
      "learning_rate": 3.86542766890545e-05,
      "loss": 0.7696,
      "mean_token_accuracy": 0.766920231282711,
      "num_tokens": 638462493.0,
      "step": 20004
    },
    {
      "entropy": 0.5837087528780103,
      "epoch": 1.840570209487007,
      "grad_norm": 0.14651432633399963,
      "learning_rate": 3.8651209862912875e-05,
      "loss": 0.569,
      "mean_token_accuracy": 0.828155867755413,
      "num_tokens": 638494675.0,
      "step": 20005
    },
    {
      "entropy": 0.5941619104705751,
      "epoch": 1.8406622158584482,
      "grad_norm": 0.15293650329113007,
      "learning_rate": 3.864814303677124e-05,
      "loss": 0.5811,
      "mean_token_accuracy": 0.8213626965880394,
      "num_tokens": 638526337.0,
      "step": 20006
    },
    {
      "entropy": 0.6806563083082438,
      "epoch": 1.8407542222298894,
      "grad_norm": 0.1601821929216385,
      "learning_rate": 3.8645076210629624e-05,
      "loss": 0.6848,
      "mean_token_accuracy": 0.7916236557066441,
      "num_tokens": 638558426.0,
      "step": 20007
    },
    {
      "entropy": 0.6210633460432291,
      "epoch": 1.8408462286013307,
      "grad_norm": 0.15988267958164215,
      "learning_rate": 3.864200938448799e-05,
      "loss": 0.602,
      "mean_token_accuracy": 0.80959602445364,
      "num_tokens": 638590310.0,
      "step": 20008
    },
    {
      "entropy": 0.699476346373558,
      "epoch": 1.8409382349727719,
      "grad_norm": 0.1647486537694931,
      "learning_rate": 3.863894255834637e-05,
      "loss": 0.6969,
      "mean_token_accuracy": 0.7862222529947758,
      "num_tokens": 638621094.0,
      "step": 20009
    },
    {
      "entropy": 0.6244696751236916,
      "epoch": 1.841030241344213,
      "grad_norm": 0.15394794940948486,
      "learning_rate": 3.863587573220474e-05,
      "loss": 0.6202,
      "mean_token_accuracy": 0.807159136980772,
      "num_tokens": 638653203.0,
      "step": 20010
    },
    {
      "entropy": 0.5189372887834907,
      "epoch": 1.8411222477156544,
      "grad_norm": 0.1471371054649353,
      "learning_rate": 3.8632808906063116e-05,
      "loss": 0.5046,
      "mean_token_accuracy": 0.8424654304981232,
      "num_tokens": 638685505.0,
      "step": 20011
    },
    {
      "entropy": 0.7761774947866797,
      "epoch": 1.8412142540870955,
      "grad_norm": 0.17095163464546204,
      "learning_rate": 3.862974207992149e-05,
      "loss": 0.7664,
      "mean_token_accuracy": 0.769897323101759,
      "num_tokens": 638717362.0,
      "step": 20012
    },
    {
      "entropy": 0.6533992709591985,
      "epoch": 1.8413062604585368,
      "grad_norm": 0.14722077548503876,
      "learning_rate": 3.8626675253779866e-05,
      "loss": 0.6392,
      "mean_token_accuracy": 0.8013523630797863,
      "num_tokens": 638749658.0,
      "step": 20013
    },
    {
      "entropy": 0.6146052367985249,
      "epoch": 1.841398266829978,
      "grad_norm": 0.1574859917163849,
      "learning_rate": 3.862360842763824e-05,
      "loss": 0.6091,
      "mean_token_accuracy": 0.8118554204702377,
      "num_tokens": 638781085.0,
      "step": 20014
    },
    {
      "entropy": 0.61720778234303,
      "epoch": 1.8414902732014191,
      "grad_norm": 0.15157118439674377,
      "learning_rate": 3.8620541601496615e-05,
      "loss": 0.6163,
      "mean_token_accuracy": 0.8078468032181263,
      "num_tokens": 638813077.0,
      "step": 20015
    },
    {
      "entropy": 0.5553380884230137,
      "epoch": 1.8415822795728605,
      "grad_norm": 0.1482563614845276,
      "learning_rate": 3.861747477535499e-05,
      "loss": 0.5429,
      "mean_token_accuracy": 0.8306521885097027,
      "num_tokens": 638845347.0,
      "step": 20016
    },
    {
      "entropy": 0.5707292621955276,
      "epoch": 1.8416742859443016,
      "grad_norm": 0.1567341536283493,
      "learning_rate": 3.861440794921336e-05,
      "loss": 0.5546,
      "mean_token_accuracy": 0.8286782428622246,
      "num_tokens": 638877934.0,
      "step": 20017
    },
    {
      "entropy": 0.5515505305957049,
      "epoch": 1.841766292315743,
      "grad_norm": 0.1489046812057495,
      "learning_rate": 3.861134112307174e-05,
      "loss": 0.5416,
      "mean_token_accuracy": 0.8327934183180332,
      "num_tokens": 638910334.0,
      "step": 20018
    },
    {
      "entropy": 0.6089257514104247,
      "epoch": 1.841858298687184,
      "grad_norm": 0.16190588474273682,
      "learning_rate": 3.860827429693011e-05,
      "loss": 0.6014,
      "mean_token_accuracy": 0.8167511261999607,
      "num_tokens": 638941589.0,
      "step": 20019
    },
    {
      "entropy": 0.5832000952214003,
      "epoch": 1.8419503050586252,
      "grad_norm": 0.15199920535087585,
      "learning_rate": 3.860520747078849e-05,
      "loss": 0.5618,
      "mean_token_accuracy": 0.821981891989708,
      "num_tokens": 638972985.0,
      "step": 20020
    },
    {
      "entropy": 0.4821748062968254,
      "epoch": 1.8420423114300666,
      "grad_norm": 0.14635302126407623,
      "learning_rate": 3.8602140644646856e-05,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8511394001543522,
      "num_tokens": 639004524.0,
      "step": 20021
    },
    {
      "entropy": 0.5064049111679196,
      "epoch": 1.8421343178015077,
      "grad_norm": 0.14813901484012604,
      "learning_rate": 3.859907381850523e-05,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8483937717974186,
      "num_tokens": 639036553.0,
      "step": 20022
    },
    {
      "entropy": 0.5877478336915374,
      "epoch": 1.842226324172949,
      "grad_norm": 0.14988185465335846,
      "learning_rate": 3.8596006992363605e-05,
      "loss": 0.5733,
      "mean_token_accuracy": 0.8232003562152386,
      "num_tokens": 639069134.0,
      "step": 20023
    },
    {
      "entropy": 0.482781101251021,
      "epoch": 1.8423183305443902,
      "grad_norm": 0.1404532492160797,
      "learning_rate": 3.859294016622198e-05,
      "loss": 0.4842,
      "mean_token_accuracy": 0.8497029431164265,
      "num_tokens": 639101395.0,
      "step": 20024
    },
    {
      "entropy": 0.6311209108680487,
      "epoch": 1.8424103369158313,
      "grad_norm": 0.16064172983169556,
      "learning_rate": 3.8589873340080355e-05,
      "loss": 0.6201,
      "mean_token_accuracy": 0.8087693266570568,
      "num_tokens": 639133383.0,
      "step": 20025
    },
    {
      "entropy": 0.5869985967874527,
      "epoch": 1.8425023432872727,
      "grad_norm": 0.14873796701431274,
      "learning_rate": 3.858680651393873e-05,
      "loss": 0.5725,
      "mean_token_accuracy": 0.825566366314888,
      "num_tokens": 639165392.0,
      "step": 20026
    },
    {
      "entropy": 0.7062993366271257,
      "epoch": 1.8425943496587138,
      "grad_norm": 0.163631871342659,
      "learning_rate": 3.85837396877971e-05,
      "loss": 0.6914,
      "mean_token_accuracy": 0.7879646606743336,
      "num_tokens": 639196086.0,
      "step": 20027
    },
    {
      "entropy": 0.6606154013425112,
      "epoch": 1.8426863560301552,
      "grad_norm": 0.16266104578971863,
      "learning_rate": 3.858067286165548e-05,
      "loss": 0.6409,
      "mean_token_accuracy": 0.8053916096687317,
      "num_tokens": 639227973.0,
      "step": 20028
    },
    {
      "entropy": 0.6983091291040182,
      "epoch": 1.8427783624015963,
      "grad_norm": 0.16629427671432495,
      "learning_rate": 3.8577606035513846e-05,
      "loss": 0.693,
      "mean_token_accuracy": 0.790419451892376,
      "num_tokens": 639259444.0,
      "step": 20029
    },
    {
      "entropy": 0.5579508617520332,
      "epoch": 1.8428703687730374,
      "grad_norm": 0.15379974246025085,
      "learning_rate": 3.857453920937222e-05,
      "loss": 0.5523,
      "mean_token_accuracy": 0.8313245922327042,
      "num_tokens": 639291434.0,
      "step": 20030
    },
    {
      "entropy": 0.4906453341245651,
      "epoch": 1.8429623751444788,
      "grad_norm": 0.14774630963802338,
      "learning_rate": 3.8571472383230596e-05,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8524918109178543,
      "num_tokens": 639323650.0,
      "step": 20031
    },
    {
      "entropy": 0.6741529684513807,
      "epoch": 1.84305438151592,
      "grad_norm": 0.15533782541751862,
      "learning_rate": 3.856840555708897e-05,
      "loss": 0.6656,
      "mean_token_accuracy": 0.7956737913191319,
      "num_tokens": 639355071.0,
      "step": 20032
    },
    {
      "entropy": 0.6453537419438362,
      "epoch": 1.8431463878873613,
      "grad_norm": 0.1521233469247818,
      "learning_rate": 3.8565338730947345e-05,
      "loss": 0.6396,
      "mean_token_accuracy": 0.8039830029010773,
      "num_tokens": 639387397.0,
      "step": 20033
    },
    {
      "entropy": 0.5744583886116743,
      "epoch": 1.8432383942588024,
      "grad_norm": 0.15146265923976898,
      "learning_rate": 3.856227190480572e-05,
      "loss": 0.5695,
      "mean_token_accuracy": 0.824590664356947,
      "num_tokens": 639419050.0,
      "step": 20034
    },
    {
      "entropy": 0.6854988895356655,
      "epoch": 1.8433304006302436,
      "grad_norm": 0.16213887929916382,
      "learning_rate": 3.855920507866409e-05,
      "loss": 0.6904,
      "mean_token_accuracy": 0.791279923170805,
      "num_tokens": 639451818.0,
      "step": 20035
    },
    {
      "entropy": 0.6433630920946598,
      "epoch": 1.843422407001685,
      "grad_norm": 0.16251565515995026,
      "learning_rate": 3.855613825252247e-05,
      "loss": 0.6359,
      "mean_token_accuracy": 0.8060469031333923,
      "num_tokens": 639482816.0,
      "step": 20036
    },
    {
      "entropy": 0.707586208358407,
      "epoch": 1.843514413373126,
      "grad_norm": 0.16038694977760315,
      "learning_rate": 3.855307142638084e-05,
      "loss": 0.7047,
      "mean_token_accuracy": 0.7843170613050461,
      "num_tokens": 639514379.0,
      "step": 20037
    },
    {
      "entropy": 0.6235167910344899,
      "epoch": 1.8436064197445674,
      "grad_norm": 0.16071201860904694,
      "learning_rate": 3.855000460023921e-05,
      "loss": 0.6079,
      "mean_token_accuracy": 0.8113139756023884,
      "num_tokens": 639545365.0,
      "step": 20038
    },
    {
      "entropy": 0.5877449214458466,
      "epoch": 1.8436984261160085,
      "grad_norm": 0.1505109816789627,
      "learning_rate": 3.8546937774097586e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.82020428404212,
      "num_tokens": 639576822.0,
      "step": 20039
    },
    {
      "entropy": 0.5849296096712351,
      "epoch": 1.8437904324874497,
      "grad_norm": 0.14978206157684326,
      "learning_rate": 3.854387094795596e-05,
      "loss": 0.5638,
      "mean_token_accuracy": 0.8259280137717724,
      "num_tokens": 639608730.0,
      "step": 20040
    },
    {
      "entropy": 0.6383691262453794,
      "epoch": 1.843882438858891,
      "grad_norm": 0.16087374091148376,
      "learning_rate": 3.8540804121814335e-05,
      "loss": 0.6398,
      "mean_token_accuracy": 0.8041788525879383,
      "num_tokens": 639640109.0,
      "step": 20041
    },
    {
      "entropy": 0.6040534749627113,
      "epoch": 1.8439744452303322,
      "grad_norm": 0.15963807702064514,
      "learning_rate": 3.853773729567271e-05,
      "loss": 0.5785,
      "mean_token_accuracy": 0.8191026113927364,
      "num_tokens": 639670752.0,
      "step": 20042
    },
    {
      "entropy": 0.540946401655674,
      "epoch": 1.8440664516017735,
      "grad_norm": 0.15517312288284302,
      "learning_rate": 3.8534670469531085e-05,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8367282971739769,
      "num_tokens": 639702695.0,
      "step": 20043
    },
    {
      "entropy": 0.5787505432963371,
      "epoch": 1.8441584579732146,
      "grad_norm": 0.1497597098350525,
      "learning_rate": 3.853160364338946e-05,
      "loss": 0.5589,
      "mean_token_accuracy": 0.8243384808301926,
      "num_tokens": 639735082.0,
      "step": 20044
    },
    {
      "entropy": 0.6307552601210773,
      "epoch": 1.8442504643446558,
      "grad_norm": 0.15550002455711365,
      "learning_rate": 3.8528536817247834e-05,
      "loss": 0.6281,
      "mean_token_accuracy": 0.8090693056583405,
      "num_tokens": 639766963.0,
      "step": 20045
    },
    {
      "entropy": 0.638868615962565,
      "epoch": 1.8443424707160971,
      "grad_norm": 0.15421392023563385,
      "learning_rate": 3.852546999110621e-05,
      "loss": 0.6297,
      "mean_token_accuracy": 0.8092930279672146,
      "num_tokens": 639798773.0,
      "step": 20046
    },
    {
      "entropy": 0.5824055131524801,
      "epoch": 1.8444344770875383,
      "grad_norm": 0.15159225463867188,
      "learning_rate": 3.852240316496458e-05,
      "loss": 0.5666,
      "mean_token_accuracy": 0.8252305127680302,
      "num_tokens": 639831234.0,
      "step": 20047
    },
    {
      "entropy": 0.5471841115504503,
      "epoch": 1.8445264834589796,
      "grad_norm": 0.15581117570400238,
      "learning_rate": 3.851933633882295e-05,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8324789740145206,
      "num_tokens": 639863059.0,
      "step": 20048
    },
    {
      "entropy": 0.574886791408062,
      "epoch": 1.8446184898304208,
      "grad_norm": 0.15252411365509033,
      "learning_rate": 3.851626951268133e-05,
      "loss": 0.5433,
      "mean_token_accuracy": 0.8306805565953255,
      "num_tokens": 639894534.0,
      "step": 20049
    },
    {
      "entropy": 0.5742353666573763,
      "epoch": 1.844710496201862,
      "grad_norm": 0.14503611624240875,
      "learning_rate": 3.85132026865397e-05,
      "loss": 0.5651,
      "mean_token_accuracy": 0.8257688470184803,
      "num_tokens": 639927233.0,
      "step": 20050
    },
    {
      "entropy": 0.6208107881247997,
      "epoch": 1.8448025025733032,
      "grad_norm": 0.14891129732131958,
      "learning_rate": 3.8510135860398075e-05,
      "loss": 0.6012,
      "mean_token_accuracy": 0.8136306293308735,
      "num_tokens": 639958613.0,
      "step": 20051
    },
    {
      "entropy": 0.6673312969505787,
      "epoch": 1.8448945089447444,
      "grad_norm": 0.158777117729187,
      "learning_rate": 3.850706903425645e-05,
      "loss": 0.6469,
      "mean_token_accuracy": 0.8022806718945503,
      "num_tokens": 639990428.0,
      "step": 20052
    },
    {
      "entropy": 0.6503367051482201,
      "epoch": 1.8449865153161857,
      "grad_norm": 0.1549687683582306,
      "learning_rate": 3.8504002208114824e-05,
      "loss": 0.6387,
      "mean_token_accuracy": 0.8037982322275639,
      "num_tokens": 640022118.0,
      "step": 20053
    },
    {
      "entropy": 0.6687499899417162,
      "epoch": 1.8450785216876269,
      "grad_norm": 0.15978844463825226,
      "learning_rate": 3.85009353819732e-05,
      "loss": 0.6528,
      "mean_token_accuracy": 0.8025717586278915,
      "num_tokens": 640054079.0,
      "step": 20054
    },
    {
      "entropy": 0.7516948617994785,
      "epoch": 1.845170528059068,
      "grad_norm": 0.1609044373035431,
      "learning_rate": 3.8497868555831574e-05,
      "loss": 0.7517,
      "mean_token_accuracy": 0.7706064246594906,
      "num_tokens": 640086730.0,
      "step": 20055
    },
    {
      "entropy": 0.58563750423491,
      "epoch": 1.8452625344305094,
      "grad_norm": 0.14792010188102722,
      "learning_rate": 3.849480172968994e-05,
      "loss": 0.5737,
      "mean_token_accuracy": 0.8229446560144424,
      "num_tokens": 640118789.0,
      "step": 20056
    },
    {
      "entropy": 0.5185917129274458,
      "epoch": 1.8453545408019505,
      "grad_norm": 0.14684200286865234,
      "learning_rate": 3.849173490354832e-05,
      "loss": 0.4962,
      "mean_token_accuracy": 0.8429522849619389,
      "num_tokens": 640150618.0,
      "step": 20057
    },
    {
      "entropy": 0.6809114795178175,
      "epoch": 1.8454465471733918,
      "grad_norm": 0.1586151272058487,
      "learning_rate": 3.848866807740669e-05,
      "loss": 0.6829,
      "mean_token_accuracy": 0.7915721647441387,
      "num_tokens": 640183177.0,
      "step": 20058
    },
    {
      "entropy": 0.5132728796452284,
      "epoch": 1.845538553544833,
      "grad_norm": 0.14284110069274902,
      "learning_rate": 3.8485601251265065e-05,
      "loss": 0.506,
      "mean_token_accuracy": 0.8413734436035156,
      "num_tokens": 640215545.0,
      "step": 20059
    },
    {
      "entropy": 0.6633385848253965,
      "epoch": 1.8456305599162741,
      "grad_norm": 0.15769325196743011,
      "learning_rate": 3.848253442512344e-05,
      "loss": 0.6552,
      "mean_token_accuracy": 0.7982728704810143,
      "num_tokens": 640247244.0,
      "step": 20060
    },
    {
      "entropy": 0.6161631364375353,
      "epoch": 1.8457225662877155,
      "grad_norm": 0.14471128582954407,
      "learning_rate": 3.8479467598981815e-05,
      "loss": 0.6062,
      "mean_token_accuracy": 0.8095591403543949,
      "num_tokens": 640279796.0,
      "step": 20061
    },
    {
      "entropy": 0.6525340690277517,
      "epoch": 1.8458145726591566,
      "grad_norm": 0.1509903073310852,
      "learning_rate": 3.847640077284019e-05,
      "loss": 0.641,
      "mean_token_accuracy": 0.802724365144968,
      "num_tokens": 640311966.0,
      "step": 20062
    },
    {
      "entropy": 0.5676870511379093,
      "epoch": 1.845906579030598,
      "grad_norm": 0.15715205669403076,
      "learning_rate": 3.8473333946698564e-05,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8283251263201237,
      "num_tokens": 640343328.0,
      "step": 20063
    },
    {
      "entropy": 0.5189276011660695,
      "epoch": 1.845998585402039,
      "grad_norm": 0.14518466591835022,
      "learning_rate": 3.847026712055694e-05,
      "loss": 0.4995,
      "mean_token_accuracy": 0.8447122164070606,
      "num_tokens": 640375219.0,
      "step": 20064
    },
    {
      "entropy": 0.6441161865368485,
      "epoch": 1.8460905917734802,
      "grad_norm": 0.1550215780735016,
      "learning_rate": 3.846720029441531e-05,
      "loss": 0.6419,
      "mean_token_accuracy": 0.803137868642807,
      "num_tokens": 640407470.0,
      "step": 20065
    },
    {
      "entropy": 0.481475530192256,
      "epoch": 1.8461825981449216,
      "grad_norm": 0.14668934047222137,
      "learning_rate": 3.846413346827368e-05,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8544245027005672,
      "num_tokens": 640439758.0,
      "step": 20066
    },
    {
      "entropy": 0.5900414986535907,
      "epoch": 1.8462746045163627,
      "grad_norm": 0.15287159383296967,
      "learning_rate": 3.846106664213206e-05,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8258500397205353,
      "num_tokens": 640471452.0,
      "step": 20067
    },
    {
      "entropy": 0.5445210272446275,
      "epoch": 1.846366610887804,
      "grad_norm": 0.15253916382789612,
      "learning_rate": 3.845799981599043e-05,
      "loss": 0.5325,
      "mean_token_accuracy": 0.8332512974739075,
      "num_tokens": 640503785.0,
      "step": 20068
    },
    {
      "entropy": 0.6024077208712697,
      "epoch": 1.8464586172592452,
      "grad_norm": 0.15702074766159058,
      "learning_rate": 3.8454932989848805e-05,
      "loss": 0.5989,
      "mean_token_accuracy": 0.8183034248650074,
      "num_tokens": 640535830.0,
      "step": 20069
    },
    {
      "entropy": 0.6287025567144156,
      "epoch": 1.8465506236306863,
      "grad_norm": 0.15301819145679474,
      "learning_rate": 3.845186616370718e-05,
      "loss": 0.623,
      "mean_token_accuracy": 0.8067066967487335,
      "num_tokens": 640566927.0,
      "step": 20070
    },
    {
      "entropy": 0.667429456487298,
      "epoch": 1.8466426300021277,
      "grad_norm": 0.16161677241325378,
      "learning_rate": 3.8448799337565554e-05,
      "loss": 0.6671,
      "mean_token_accuracy": 0.7949389405548573,
      "num_tokens": 640599083.0,
      "step": 20071
    },
    {
      "entropy": 0.6502394611015916,
      "epoch": 1.8467346363735688,
      "grad_norm": 0.15628035366535187,
      "learning_rate": 3.844573251142393e-05,
      "loss": 0.6494,
      "mean_token_accuracy": 0.8040188550949097,
      "num_tokens": 640631603.0,
      "step": 20072
    },
    {
      "entropy": 0.5886916946619749,
      "epoch": 1.8468266427450102,
      "grad_norm": 0.15738070011138916,
      "learning_rate": 3.8442665685282304e-05,
      "loss": 0.5859,
      "mean_token_accuracy": 0.8238221108913422,
      "num_tokens": 640662799.0,
      "step": 20073
    },
    {
      "entropy": 0.5129391849040985,
      "epoch": 1.8469186491164513,
      "grad_norm": 0.1567527949810028,
      "learning_rate": 3.843959885914068e-05,
      "loss": 0.491,
      "mean_token_accuracy": 0.8423788845539093,
      "num_tokens": 640694309.0,
      "step": 20074
    },
    {
      "entropy": 0.6113803246989846,
      "epoch": 1.8470106554878924,
      "grad_norm": 0.15300814807415009,
      "learning_rate": 3.843653203299905e-05,
      "loss": 0.6143,
      "mean_token_accuracy": 0.8147330991923809,
      "num_tokens": 640726805.0,
      "step": 20075
    },
    {
      "entropy": 0.543318668846041,
      "epoch": 1.8471026618593338,
      "grad_norm": 0.15636985003948212,
      "learning_rate": 3.843346520685743e-05,
      "loss": 0.5237,
      "mean_token_accuracy": 0.8388210572302341,
      "num_tokens": 640758457.0,
      "step": 20076
    },
    {
      "entropy": 0.6641465779393911,
      "epoch": 1.847194668230775,
      "grad_norm": 0.15319180488586426,
      "learning_rate": 3.8430398380715796e-05,
      "loss": 0.6525,
      "mean_token_accuracy": 0.7983454801142216,
      "num_tokens": 640790116.0,
      "step": 20077
    },
    {
      "entropy": 0.5714216362684965,
      "epoch": 1.8472866746022163,
      "grad_norm": 0.15205179154872894,
      "learning_rate": 3.842733155457418e-05,
      "loss": 0.565,
      "mean_token_accuracy": 0.827721256762743,
      "num_tokens": 640822054.0,
      "step": 20078
    },
    {
      "entropy": 0.5823385319672525,
      "epoch": 1.8473786809736574,
      "grad_norm": 0.1533077508211136,
      "learning_rate": 3.8424264728432545e-05,
      "loss": 0.5688,
      "mean_token_accuracy": 0.8272978439927101,
      "num_tokens": 640853954.0,
      "step": 20079
    },
    {
      "entropy": 0.58760298602283,
      "epoch": 1.8474706873450986,
      "grad_norm": 0.14955148100852966,
      "learning_rate": 3.8421197902290926e-05,
      "loss": 0.574,
      "mean_token_accuracy": 0.8252568133175373,
      "num_tokens": 640885669.0,
      "step": 20080
    },
    {
      "entropy": 0.6845572479069233,
      "epoch": 1.84756269371654,
      "grad_norm": 0.16520892083644867,
      "learning_rate": 3.8418131076149294e-05,
      "loss": 0.6693,
      "mean_token_accuracy": 0.792245402932167,
      "num_tokens": 640916946.0,
      "step": 20081
    },
    {
      "entropy": 0.6465343348681927,
      "epoch": 1.847654700087981,
      "grad_norm": 0.15773896872997284,
      "learning_rate": 3.841506425000767e-05,
      "loss": 0.6272,
      "mean_token_accuracy": 0.8095633238554001,
      "num_tokens": 640948586.0,
      "step": 20082
    },
    {
      "entropy": 0.6248660404235125,
      "epoch": 1.8477467064594224,
      "grad_norm": 0.15362152457237244,
      "learning_rate": 3.8411997423866043e-05,
      "loss": 0.6118,
      "mean_token_accuracy": 0.8109485693275928,
      "num_tokens": 640980824.0,
      "step": 20083
    },
    {
      "entropy": 0.5825074454769492,
      "epoch": 1.8478387128308635,
      "grad_norm": 0.1549200862646103,
      "learning_rate": 3.840893059772442e-05,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8285205475986004,
      "num_tokens": 641010841.0,
      "step": 20084
    },
    {
      "entropy": 0.5585773903876543,
      "epoch": 1.8479307192023047,
      "grad_norm": 0.14915403723716736,
      "learning_rate": 3.840586377158279e-05,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8363346345722675,
      "num_tokens": 641042538.0,
      "step": 20085
    },
    {
      "entropy": 0.6050785724073648,
      "epoch": 1.848022725573746,
      "grad_norm": 0.15522843599319458,
      "learning_rate": 3.840279694544117e-05,
      "loss": 0.5969,
      "mean_token_accuracy": 0.8184297941625118,
      "num_tokens": 641074323.0,
      "step": 20086
    },
    {
      "entropy": 0.5607195813208818,
      "epoch": 1.8481147319451872,
      "grad_norm": 0.1488194763660431,
      "learning_rate": 3.8399730119299535e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8285407721996307,
      "num_tokens": 641106392.0,
      "step": 20087
    },
    {
      "entropy": 0.6801619492471218,
      "epoch": 1.8482067383166285,
      "grad_norm": 0.16791808605194092,
      "learning_rate": 3.839666329315792e-05,
      "loss": 0.6749,
      "mean_token_accuracy": 0.7973649650812149,
      "num_tokens": 641137957.0,
      "step": 20088
    },
    {
      "entropy": 0.5675851227715611,
      "epoch": 1.8482987446880697,
      "grad_norm": 0.15577921271324158,
      "learning_rate": 3.8393596467016285e-05,
      "loss": 0.5555,
      "mean_token_accuracy": 0.8289405256509781,
      "num_tokens": 641170165.0,
      "step": 20089
    },
    {
      "entropy": 0.5510773230344057,
      "epoch": 1.8483907510595108,
      "grad_norm": 0.15117038786411285,
      "learning_rate": 3.839052964087466e-05,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8341333158314228,
      "num_tokens": 641201371.0,
      "step": 20090
    },
    {
      "entropy": 0.6605174876749516,
      "epoch": 1.8484827574309521,
      "grad_norm": 0.15378673374652863,
      "learning_rate": 3.8387462814733034e-05,
      "loss": 0.6584,
      "mean_token_accuracy": 0.8009090200066566,
      "num_tokens": 641234038.0,
      "step": 20091
    },
    {
      "entropy": 0.48556422675028443,
      "epoch": 1.8485747638023933,
      "grad_norm": 0.14866995811462402,
      "learning_rate": 3.838439598859141e-05,
      "loss": 0.4631,
      "mean_token_accuracy": 0.8561174981296062,
      "num_tokens": 641265987.0,
      "step": 20092
    },
    {
      "entropy": 0.5854393243789673,
      "epoch": 1.8486667701738346,
      "grad_norm": 0.15453235805034637,
      "learning_rate": 3.838132916244978e-05,
      "loss": 0.5856,
      "mean_token_accuracy": 0.8195261768996716,
      "num_tokens": 641297934.0,
      "step": 20093
    },
    {
      "entropy": 0.6282234722748399,
      "epoch": 1.8487587765452758,
      "grad_norm": 0.15584510564804077,
      "learning_rate": 3.837826233630816e-05,
      "loss": 0.6271,
      "mean_token_accuracy": 0.810032032430172,
      "num_tokens": 641329837.0,
      "step": 20094
    },
    {
      "entropy": 0.6874830583110452,
      "epoch": 1.848850782916717,
      "grad_norm": 0.1622319221496582,
      "learning_rate": 3.8375195510166526e-05,
      "loss": 0.676,
      "mean_token_accuracy": 0.7943408824503422,
      "num_tokens": 641361289.0,
      "step": 20095
    },
    {
      "entropy": 0.548552704975009,
      "epoch": 1.8489427892881583,
      "grad_norm": 0.14832183718681335,
      "learning_rate": 3.837212868402491e-05,
      "loss": 0.534,
      "mean_token_accuracy": 0.8368270061910152,
      "num_tokens": 641393087.0,
      "step": 20096
    },
    {
      "entropy": 0.6663108896464109,
      "epoch": 1.8490347956595994,
      "grad_norm": 0.16505196690559387,
      "learning_rate": 3.8369061857883275e-05,
      "loss": 0.6618,
      "mean_token_accuracy": 0.7962261401116848,
      "num_tokens": 641424541.0,
      "step": 20097
    },
    {
      "entropy": 0.5329313576221466,
      "epoch": 1.8491268020310407,
      "grad_norm": 0.14129464328289032,
      "learning_rate": 3.836599503174165e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8398328609764576,
      "num_tokens": 641456863.0,
      "step": 20098
    },
    {
      "entropy": 0.5587940325494856,
      "epoch": 1.8492188084024819,
      "grad_norm": 0.14687485992908478,
      "learning_rate": 3.8362928205600024e-05,
      "loss": 0.5528,
      "mean_token_accuracy": 0.8294937610626221,
      "num_tokens": 641488777.0,
      "step": 20099
    },
    {
      "entropy": 0.603771198540926,
      "epoch": 1.849310814773923,
      "grad_norm": 0.15344461798667908,
      "learning_rate": 3.83598613794584e-05,
      "loss": 0.5922,
      "mean_token_accuracy": 0.8162858672440052,
      "num_tokens": 641520494.0,
      "step": 20100
    },
    {
      "entropy": 0.6636497573927045,
      "epoch": 1.8494028211453644,
      "grad_norm": 0.15592466294765472,
      "learning_rate": 3.8356794553316774e-05,
      "loss": 0.6567,
      "mean_token_accuracy": 0.7962500415742397,
      "num_tokens": 641552492.0,
      "step": 20101
    },
    {
      "entropy": 0.6527704056352377,
      "epoch": 1.8494948275168055,
      "grad_norm": 0.16189642250537872,
      "learning_rate": 3.835372772717515e-05,
      "loss": 0.6442,
      "mean_token_accuracy": 0.7973109222948551,
      "num_tokens": 641584028.0,
      "step": 20102
    },
    {
      "entropy": 0.6791033074259758,
      "epoch": 1.8495868338882469,
      "grad_norm": 0.16819602251052856,
      "learning_rate": 3.835066090103352e-05,
      "loss": 0.6705,
      "mean_token_accuracy": 0.7982393465936184,
      "num_tokens": 641614887.0,
      "step": 20103
    },
    {
      "entropy": 0.556054150685668,
      "epoch": 1.849678840259688,
      "grad_norm": 0.1634623408317566,
      "learning_rate": 3.83475940748919e-05,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8318787813186646,
      "num_tokens": 641646676.0,
      "step": 20104
    },
    {
      "entropy": 0.6233669836074114,
      "epoch": 1.8497708466311291,
      "grad_norm": 0.15262217819690704,
      "learning_rate": 3.834452724875027e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.8111688643693924,
      "num_tokens": 641679262.0,
      "step": 20105
    },
    {
      "entropy": 0.5372018800117075,
      "epoch": 1.8498628530025705,
      "grad_norm": 0.1462676227092743,
      "learning_rate": 3.834146042260865e-05,
      "loss": 0.5262,
      "mean_token_accuracy": 0.8372966088354588,
      "num_tokens": 641711318.0,
      "step": 20106
    },
    {
      "entropy": 0.6299753380008042,
      "epoch": 1.8499548593740116,
      "grad_norm": 0.15159940719604492,
      "learning_rate": 3.833839359646702e-05,
      "loss": 0.6191,
      "mean_token_accuracy": 0.8137801960110664,
      "num_tokens": 641743731.0,
      "step": 20107
    },
    {
      "entropy": 0.7089261058717966,
      "epoch": 1.850046865745453,
      "grad_norm": 0.15676048398017883,
      "learning_rate": 3.833532677032539e-05,
      "loss": 0.6924,
      "mean_token_accuracy": 0.7871595434844494,
      "num_tokens": 641775285.0,
      "step": 20108
    },
    {
      "entropy": 0.6745587084442377,
      "epoch": 1.850138872116894,
      "grad_norm": 0.16059044003486633,
      "learning_rate": 3.833225994418377e-05,
      "loss": 0.6803,
      "mean_token_accuracy": 0.7917057015001774,
      "num_tokens": 641807420.0,
      "step": 20109
    },
    {
      "entropy": 0.5919946478679776,
      "epoch": 1.8502308784883352,
      "grad_norm": 0.15327763557434082,
      "learning_rate": 3.832919311804214e-05,
      "loss": 0.5846,
      "mean_token_accuracy": 0.8177672699093819,
      "num_tokens": 641838911.0,
      "step": 20110
    },
    {
      "entropy": 0.6356851998716593,
      "epoch": 1.8503228848597766,
      "grad_norm": 0.16158141195774078,
      "learning_rate": 3.832612629190051e-05,
      "loss": 0.6326,
      "mean_token_accuracy": 0.8053868785500526,
      "num_tokens": 641871247.0,
      "step": 20111
    },
    {
      "entropy": 0.5056082927621901,
      "epoch": 1.8504148912312177,
      "grad_norm": 0.14697028696537018,
      "learning_rate": 3.832305946575889e-05,
      "loss": 0.4967,
      "mean_token_accuracy": 0.845918532460928,
      "num_tokens": 641903704.0,
      "step": 20112
    },
    {
      "entropy": 0.5749184507876635,
      "epoch": 1.850506897602659,
      "grad_norm": 0.15567509829998016,
      "learning_rate": 3.831999263961726e-05,
      "loss": 0.5694,
      "mean_token_accuracy": 0.826347392052412,
      "num_tokens": 641935445.0,
      "step": 20113
    },
    {
      "entropy": 0.49936466477811337,
      "epoch": 1.8505989039741002,
      "grad_norm": 0.1531064808368683,
      "learning_rate": 3.831692581347564e-05,
      "loss": 0.499,
      "mean_token_accuracy": 0.8432282209396362,
      "num_tokens": 641967235.0,
      "step": 20114
    },
    {
      "entropy": 0.6161254327744246,
      "epoch": 1.8506909103455413,
      "grad_norm": 0.1524621993303299,
      "learning_rate": 3.831385898733401e-05,
      "loss": 0.6046,
      "mean_token_accuracy": 0.8141258619725704,
      "num_tokens": 641999650.0,
      "step": 20115
    },
    {
      "entropy": 0.6216013869270682,
      "epoch": 1.8507829167169827,
      "grad_norm": 0.15235473215579987,
      "learning_rate": 3.831079216119238e-05,
      "loss": 0.6048,
      "mean_token_accuracy": 0.8154021948575974,
      "num_tokens": 642031773.0,
      "step": 20116
    },
    {
      "entropy": 0.6826407387852669,
      "epoch": 1.8508749230884238,
      "grad_norm": 0.1647946834564209,
      "learning_rate": 3.830772533505076e-05,
      "loss": 0.6748,
      "mean_token_accuracy": 0.7909054644405842,
      "num_tokens": 642063084.0,
      "step": 20117
    },
    {
      "entropy": 0.6363363824784756,
      "epoch": 1.8509669294598652,
      "grad_norm": 0.1586863398551941,
      "learning_rate": 3.830465850890913e-05,
      "loss": 0.6386,
      "mean_token_accuracy": 0.8077666833996773,
      "num_tokens": 642094956.0,
      "step": 20118
    },
    {
      "entropy": 0.6965777669101954,
      "epoch": 1.8510589358313063,
      "grad_norm": 0.16494451463222504,
      "learning_rate": 3.8301591682767504e-05,
      "loss": 0.6731,
      "mean_token_accuracy": 0.7952885031700134,
      "num_tokens": 642126602.0,
      "step": 20119
    },
    {
      "entropy": 0.6649625077843666,
      "epoch": 1.8511509422027475,
      "grad_norm": 0.15876929461956024,
      "learning_rate": 3.829852485662588e-05,
      "loss": 0.6627,
      "mean_token_accuracy": 0.796858761459589,
      "num_tokens": 642159034.0,
      "step": 20120
    },
    {
      "entropy": 0.5917795095592737,
      "epoch": 1.8512429485741888,
      "grad_norm": 0.1565108448266983,
      "learning_rate": 3.829545803048425e-05,
      "loss": 0.5761,
      "mean_token_accuracy": 0.8206711933016777,
      "num_tokens": 642190613.0,
      "step": 20121
    },
    {
      "entropy": 0.5543080763891339,
      "epoch": 1.85133495494563,
      "grad_norm": 0.1434730440378189,
      "learning_rate": 3.829239120434263e-05,
      "loss": 0.5518,
      "mean_token_accuracy": 0.8292067535221577,
      "num_tokens": 642223290.0,
      "step": 20122
    },
    {
      "entropy": 0.5594840822741389,
      "epoch": 1.8514269613170713,
      "grad_norm": 0.14533527195453644,
      "learning_rate": 3.8289324378201e-05,
      "loss": 0.541,
      "mean_token_accuracy": 0.8356471210718155,
      "num_tokens": 642255205.0,
      "step": 20123
    },
    {
      "entropy": 0.6274147173389792,
      "epoch": 1.8515189676885124,
      "grad_norm": 0.1540180891752243,
      "learning_rate": 3.828625755205938e-05,
      "loss": 0.6103,
      "mean_token_accuracy": 0.8131873607635498,
      "num_tokens": 642287963.0,
      "step": 20124
    },
    {
      "entropy": 0.5859491426963359,
      "epoch": 1.8516109740599536,
      "grad_norm": 0.15644831955432892,
      "learning_rate": 3.828319072591775e-05,
      "loss": 0.5825,
      "mean_token_accuracy": 0.8193375170230865,
      "num_tokens": 642319870.0,
      "step": 20125
    },
    {
      "entropy": 0.6591705027967691,
      "epoch": 1.851702980431395,
      "grad_norm": 0.15142549574375153,
      "learning_rate": 3.828012389977612e-05,
      "loss": 0.6578,
      "mean_token_accuracy": 0.7969162613153458,
      "num_tokens": 642351811.0,
      "step": 20126
    },
    {
      "entropy": 0.5345064932480454,
      "epoch": 1.851794986802836,
      "grad_norm": 0.14041244983673096,
      "learning_rate": 3.82770570736345e-05,
      "loss": 0.5154,
      "mean_token_accuracy": 0.8384746015071869,
      "num_tokens": 642383809.0,
      "step": 20127
    },
    {
      "entropy": 0.6327606067061424,
      "epoch": 1.8518869931742774,
      "grad_norm": 0.15839456021785736,
      "learning_rate": 3.827399024749287e-05,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8105944283306599,
      "num_tokens": 642415128.0,
      "step": 20128
    },
    {
      "entropy": 0.6338675990700722,
      "epoch": 1.8519789995457185,
      "grad_norm": 0.1552390307188034,
      "learning_rate": 3.827092342135124e-05,
      "loss": 0.6294,
      "mean_token_accuracy": 0.8087133355438709,
      "num_tokens": 642445906.0,
      "step": 20129
    },
    {
      "entropy": 0.6021791324019432,
      "epoch": 1.8520710059171597,
      "grad_norm": 0.15313184261322021,
      "learning_rate": 3.826785659520962e-05,
      "loss": 0.5885,
      "mean_token_accuracy": 0.8167801983654499,
      "num_tokens": 642478134.0,
      "step": 20130
    },
    {
      "entropy": 0.6300938464701176,
      "epoch": 1.852163012288601,
      "grad_norm": 0.1607762575149536,
      "learning_rate": 3.826478976906799e-05,
      "loss": 0.639,
      "mean_token_accuracy": 0.8077557124197483,
      "num_tokens": 642509252.0,
      "step": 20131
    },
    {
      "entropy": 0.6383300274610519,
      "epoch": 1.8522550186600422,
      "grad_norm": 0.16055725514888763,
      "learning_rate": 3.826172294292637e-05,
      "loss": 0.6324,
      "mean_token_accuracy": 0.806234672665596,
      "num_tokens": 642541141.0,
      "step": 20132
    },
    {
      "entropy": 0.623403238132596,
      "epoch": 1.8523470250314835,
      "grad_norm": 0.15467093884944916,
      "learning_rate": 3.825865611678474e-05,
      "loss": 0.6178,
      "mean_token_accuracy": 0.8093832246959209,
      "num_tokens": 642573099.0,
      "step": 20133
    },
    {
      "entropy": 0.49085687659680843,
      "epoch": 1.8524390314029247,
      "grad_norm": 0.14813201129436493,
      "learning_rate": 3.8255589290643117e-05,
      "loss": 0.4775,
      "mean_token_accuracy": 0.851643793284893,
      "num_tokens": 642605190.0,
      "step": 20134
    },
    {
      "entropy": 0.5495682479813695,
      "epoch": 1.8525310377743658,
      "grad_norm": 0.1513737589120865,
      "learning_rate": 3.825252246450149e-05,
      "loss": 0.5418,
      "mean_token_accuracy": 0.8337811380624771,
      "num_tokens": 642637958.0,
      "step": 20135
    },
    {
      "entropy": 0.5686314832419157,
      "epoch": 1.8526230441458071,
      "grad_norm": 0.15813478827476501,
      "learning_rate": 3.8249455638359866e-05,
      "loss": 0.5677,
      "mean_token_accuracy": 0.8265688493847847,
      "num_tokens": 642669466.0,
      "step": 20136
    },
    {
      "entropy": 0.6740667335689068,
      "epoch": 1.8527150505172483,
      "grad_norm": 0.15247038006782532,
      "learning_rate": 3.8246388812218234e-05,
      "loss": 0.654,
      "mean_token_accuracy": 0.7980121187865734,
      "num_tokens": 642701591.0,
      "step": 20137
    },
    {
      "entropy": 0.6928689545020461,
      "epoch": 1.8528070568886896,
      "grad_norm": 0.15841704607009888,
      "learning_rate": 3.8243321986076615e-05,
      "loss": 0.6821,
      "mean_token_accuracy": 0.7887511514127254,
      "num_tokens": 642733706.0,
      "step": 20138
    },
    {
      "entropy": 0.5669419057667255,
      "epoch": 1.8528990632601308,
      "grad_norm": 0.15023601055145264,
      "learning_rate": 3.824025515993498e-05,
      "loss": 0.5544,
      "mean_token_accuracy": 0.832451481372118,
      "num_tokens": 642765156.0,
      "step": 20139
    },
    {
      "entropy": 0.6447481699287891,
      "epoch": 1.852991069631572,
      "grad_norm": 0.15353333950042725,
      "learning_rate": 3.8237188333793364e-05,
      "loss": 0.6183,
      "mean_token_accuracy": 0.8061834163963795,
      "num_tokens": 642796372.0,
      "step": 20140
    },
    {
      "entropy": 0.6456287875771523,
      "epoch": 1.8530830760030133,
      "grad_norm": 0.16346195340156555,
      "learning_rate": 3.823412150765173e-05,
      "loss": 0.6404,
      "mean_token_accuracy": 0.8030321523547173,
      "num_tokens": 642827871.0,
      "step": 20141
    },
    {
      "entropy": 0.4827807806432247,
      "epoch": 1.8531750823744544,
      "grad_norm": 0.13687318563461304,
      "learning_rate": 3.823105468151011e-05,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8515429273247719,
      "num_tokens": 642860449.0,
      "step": 20142
    },
    {
      "entropy": 0.4758247225545347,
      "epoch": 1.8532670887458957,
      "grad_norm": 0.14305585622787476,
      "learning_rate": 3.822798785536848e-05,
      "loss": 0.4735,
      "mean_token_accuracy": 0.8544538654386997,
      "num_tokens": 642891955.0,
      "step": 20143
    },
    {
      "entropy": 0.5979050993919373,
      "epoch": 1.8533590951173369,
      "grad_norm": 0.15296417474746704,
      "learning_rate": 3.8224921029226856e-05,
      "loss": 0.5848,
      "mean_token_accuracy": 0.8209497071802616,
      "num_tokens": 642924366.0,
      "step": 20144
    },
    {
      "entropy": 0.5904347752220929,
      "epoch": 1.853451101488778,
      "grad_norm": 0.15604791045188904,
      "learning_rate": 3.822185420308523e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8211163394153118,
      "num_tokens": 642955713.0,
      "step": 20145
    },
    {
      "entropy": 0.5493530072271824,
      "epoch": 1.8535431078602194,
      "grad_norm": 0.14751175045967102,
      "learning_rate": 3.8218787376943606e-05,
      "loss": 0.5481,
      "mean_token_accuracy": 0.830479085445404,
      "num_tokens": 642988086.0,
      "step": 20146
    },
    {
      "entropy": 0.5254292306490242,
      "epoch": 1.8536351142316605,
      "grad_norm": 0.14860959351062775,
      "learning_rate": 3.8215720550801973e-05,
      "loss": 0.5114,
      "mean_token_accuracy": 0.8416835442185402,
      "num_tokens": 643019591.0,
      "step": 20147
    },
    {
      "entropy": 0.5208263311069459,
      "epoch": 1.8537271206031019,
      "grad_norm": 0.14888465404510498,
      "learning_rate": 3.8212653724660355e-05,
      "loss": 0.5042,
      "mean_token_accuracy": 0.8429956808686256,
      "num_tokens": 643051058.0,
      "step": 20148
    },
    {
      "entropy": 0.592694922350347,
      "epoch": 1.853819126974543,
      "grad_norm": 0.16000133752822876,
      "learning_rate": 3.820958689851872e-05,
      "loss": 0.5838,
      "mean_token_accuracy": 0.817302867770195,
      "num_tokens": 643081960.0,
      "step": 20149
    },
    {
      "entropy": 0.5742978085763752,
      "epoch": 1.8539111333459841,
      "grad_norm": 0.14845137298107147,
      "learning_rate": 3.82065200723771e-05,
      "loss": 0.5824,
      "mean_token_accuracy": 0.8172314576804638,
      "num_tokens": 643114698.0,
      "step": 20150
    },
    {
      "entropy": 0.7366017904132605,
      "epoch": 1.8540031397174255,
      "grad_norm": 0.16017501056194305,
      "learning_rate": 3.820345324623547e-05,
      "loss": 0.726,
      "mean_token_accuracy": 0.7788531742990017,
      "num_tokens": 643146647.0,
      "step": 20151
    },
    {
      "entropy": 0.6066845376044512,
      "epoch": 1.8540951460888666,
      "grad_norm": 0.15344713628292084,
      "learning_rate": 3.820038642009385e-05,
      "loss": 0.5854,
      "mean_token_accuracy": 0.8169460631906986,
      "num_tokens": 643177663.0,
      "step": 20152
    },
    {
      "entropy": 0.6050769994035363,
      "epoch": 1.854187152460308,
      "grad_norm": 0.16155187785625458,
      "learning_rate": 3.819731959395222e-05,
      "loss": 0.6015,
      "mean_token_accuracy": 0.8150060400366783,
      "num_tokens": 643208858.0,
      "step": 20153
    },
    {
      "entropy": 0.6270965822041035,
      "epoch": 1.854279158831749,
      "grad_norm": 0.15508270263671875,
      "learning_rate": 3.8194252767810596e-05,
      "loss": 0.6185,
      "mean_token_accuracy": 0.8088742010295391,
      "num_tokens": 643241360.0,
      "step": 20154
    },
    {
      "entropy": 0.5924027995206416,
      "epoch": 1.8543711652031902,
      "grad_norm": 0.15835346281528473,
      "learning_rate": 3.8191185941668964e-05,
      "loss": 0.5868,
      "mean_token_accuracy": 0.8175483085215092,
      "num_tokens": 643273328.0,
      "step": 20155
    },
    {
      "entropy": 0.5784739404916763,
      "epoch": 1.8544631715746316,
      "grad_norm": 0.1474360078573227,
      "learning_rate": 3.8188119115527345e-05,
      "loss": 0.568,
      "mean_token_accuracy": 0.8252287209033966,
      "num_tokens": 643304956.0,
      "step": 20156
    },
    {
      "entropy": 0.5892106450628489,
      "epoch": 1.8545551779460727,
      "grad_norm": 0.15271909534931183,
      "learning_rate": 3.818505228938571e-05,
      "loss": 0.5806,
      "mean_token_accuracy": 0.8207590393722057,
      "num_tokens": 643337362.0,
      "step": 20157
    },
    {
      "entropy": 0.610480927862227,
      "epoch": 1.854647184317514,
      "grad_norm": 0.15252019464969635,
      "learning_rate": 3.818198546324409e-05,
      "loss": 0.599,
      "mean_token_accuracy": 0.8197630159556866,
      "num_tokens": 643368820.0,
      "step": 20158
    },
    {
      "entropy": 0.572196583263576,
      "epoch": 1.8547391906889552,
      "grad_norm": 0.14287716150283813,
      "learning_rate": 3.817891863710246e-05,
      "loss": 0.5597,
      "mean_token_accuracy": 0.8243460766971111,
      "num_tokens": 643400630.0,
      "step": 20159
    },
    {
      "entropy": 0.5960161425173283,
      "epoch": 1.8548311970603963,
      "grad_norm": 0.15561988949775696,
      "learning_rate": 3.817585181096084e-05,
      "loss": 0.5805,
      "mean_token_accuracy": 0.8210885487496853,
      "num_tokens": 643431799.0,
      "step": 20160
    },
    {
      "entropy": 0.5909542944282293,
      "epoch": 1.8549232034318377,
      "grad_norm": 0.1628892868757248,
      "learning_rate": 3.817278498481921e-05,
      "loss": 0.5896,
      "mean_token_accuracy": 0.8181793764233589,
      "num_tokens": 643463689.0,
      "step": 20161
    },
    {
      "entropy": 0.5546099301427603,
      "epoch": 1.8550152098032788,
      "grad_norm": 0.14265261590480804,
      "learning_rate": 3.8169718158677586e-05,
      "loss": 0.5472,
      "mean_token_accuracy": 0.8301013074815273,
      "num_tokens": 643496377.0,
      "step": 20162
    },
    {
      "entropy": 0.5881641646847129,
      "epoch": 1.8551072161747202,
      "grad_norm": 0.15643279254436493,
      "learning_rate": 3.8166651332535954e-05,
      "loss": 0.5825,
      "mean_token_accuracy": 0.8213860802352428,
      "num_tokens": 643527609.0,
      "step": 20163
    },
    {
      "entropy": 0.5862077670171857,
      "epoch": 1.8551992225461613,
      "grad_norm": 0.1497085690498352,
      "learning_rate": 3.8163584506394336e-05,
      "loss": 0.5835,
      "mean_token_accuracy": 0.8143374882638454,
      "num_tokens": 643559411.0,
      "step": 20164
    },
    {
      "entropy": 0.6521482160314918,
      "epoch": 1.8552912289176025,
      "grad_norm": 0.15702185034751892,
      "learning_rate": 3.816051768025271e-05,
      "loss": 0.656,
      "mean_token_accuracy": 0.7999181114137173,
      "num_tokens": 643590664.0,
      "step": 20165
    },
    {
      "entropy": 0.5738735143095255,
      "epoch": 1.8553832352890438,
      "grad_norm": 0.14700061082839966,
      "learning_rate": 3.8157450854111085e-05,
      "loss": 0.5608,
      "mean_token_accuracy": 0.8261691182851791,
      "num_tokens": 643622797.0,
      "step": 20166
    },
    {
      "entropy": 0.5596347160171717,
      "epoch": 1.855475241660485,
      "grad_norm": 0.148319274187088,
      "learning_rate": 3.815438402796946e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8285940513014793,
      "num_tokens": 643655084.0,
      "step": 20167
    },
    {
      "entropy": 0.6538048684597015,
      "epoch": 1.8555672480319263,
      "grad_norm": 0.15982499718666077,
      "learning_rate": 3.815131720182783e-05,
      "loss": 0.6358,
      "mean_token_accuracy": 0.8041831739246845,
      "num_tokens": 643686724.0,
      "step": 20168
    },
    {
      "entropy": 0.5267978655174375,
      "epoch": 1.8556592544033674,
      "grad_norm": 0.14465680718421936,
      "learning_rate": 3.814825037568621e-05,
      "loss": 0.521,
      "mean_token_accuracy": 0.8382110558450222,
      "num_tokens": 643718845.0,
      "step": 20169
    },
    {
      "entropy": 0.5836936086416245,
      "epoch": 1.8557512607748086,
      "grad_norm": 0.15149220824241638,
      "learning_rate": 3.814518354954458e-05,
      "loss": 0.5657,
      "mean_token_accuracy": 0.8258719444274902,
      "num_tokens": 643751115.0,
      "step": 20170
    },
    {
      "entropy": 0.639935964718461,
      "epoch": 1.85584326714625,
      "grad_norm": 0.15503588318824768,
      "learning_rate": 3.814211672340295e-05,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8075538612902164,
      "num_tokens": 643782716.0,
      "step": 20171
    },
    {
      "entropy": 0.6049822364002466,
      "epoch": 1.855935273517691,
      "grad_norm": 0.15240418910980225,
      "learning_rate": 3.8139049897261326e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8171542510390282,
      "num_tokens": 643815011.0,
      "step": 20172
    },
    {
      "entropy": 0.6382066365331411,
      "epoch": 1.8560272798891324,
      "grad_norm": 0.15895354747772217,
      "learning_rate": 3.81359830711197e-05,
      "loss": 0.6323,
      "mean_token_accuracy": 0.8078216090798378,
      "num_tokens": 643847321.0,
      "step": 20173
    },
    {
      "entropy": 0.5806713029742241,
      "epoch": 1.8561192862605735,
      "grad_norm": 0.15066863596439362,
      "learning_rate": 3.8132916244978075e-05,
      "loss": 0.5638,
      "mean_token_accuracy": 0.8235518336296082,
      "num_tokens": 643878613.0,
      "step": 20174
    },
    {
      "entropy": 0.5831570085138083,
      "epoch": 1.8562112926320147,
      "grad_norm": 0.15324629843235016,
      "learning_rate": 3.812984941883645e-05,
      "loss": 0.579,
      "mean_token_accuracy": 0.8207078017294407,
      "num_tokens": 643911083.0,
      "step": 20175
    },
    {
      "entropy": 0.5468734013848007,
      "epoch": 1.856303299003456,
      "grad_norm": 0.150834321975708,
      "learning_rate": 3.812678259269482e-05,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8337765671312809,
      "num_tokens": 643942666.0,
      "step": 20176
    },
    {
      "entropy": 0.5278650792315602,
      "epoch": 1.8563953053748972,
      "grad_norm": 0.144990473985672,
      "learning_rate": 3.81237157665532e-05,
      "loss": 0.5127,
      "mean_token_accuracy": 0.8407311588525772,
      "num_tokens": 643974906.0,
      "step": 20177
    },
    {
      "entropy": 0.6255029533058405,
      "epoch": 1.8564873117463385,
      "grad_norm": 0.1533399075269699,
      "learning_rate": 3.812064894041157e-05,
      "loss": 0.624,
      "mean_token_accuracy": 0.8125024251639843,
      "num_tokens": 644007487.0,
      "step": 20178
    },
    {
      "entropy": 0.6198471691459417,
      "epoch": 1.8565793181177797,
      "grad_norm": 0.15745669603347778,
      "learning_rate": 3.811758211426994e-05,
      "loss": 0.624,
      "mean_token_accuracy": 0.8082638159394264,
      "num_tokens": 644039968.0,
      "step": 20179
    },
    {
      "entropy": 0.5873255394399166,
      "epoch": 1.8566713244892208,
      "grad_norm": 0.14616993069648743,
      "learning_rate": 3.8114515288128316e-05,
      "loss": 0.5768,
      "mean_token_accuracy": 0.8201756440103054,
      "num_tokens": 644072428.0,
      "step": 20180
    },
    {
      "entropy": 0.6213527768850327,
      "epoch": 1.8567633308606621,
      "grad_norm": 0.15586458146572113,
      "learning_rate": 3.811144846198669e-05,
      "loss": 0.6181,
      "mean_token_accuracy": 0.8129736632108688,
      "num_tokens": 644104851.0,
      "step": 20181
    },
    {
      "entropy": 0.49847467872314155,
      "epoch": 1.8568553372321033,
      "grad_norm": 0.14329512417316437,
      "learning_rate": 3.8108381635845066e-05,
      "loss": 0.4824,
      "mean_token_accuracy": 0.8481305055320263,
      "num_tokens": 644136546.0,
      "step": 20182
    },
    {
      "entropy": 0.6370272599160671,
      "epoch": 1.8569473436035446,
      "grad_norm": 0.1550871580839157,
      "learning_rate": 3.810531480970344e-05,
      "loss": 0.6391,
      "mean_token_accuracy": 0.8100720718502998,
      "num_tokens": 644168457.0,
      "step": 20183
    },
    {
      "entropy": 0.5816693119704723,
      "epoch": 1.8570393499749858,
      "grad_norm": 0.15793421864509583,
      "learning_rate": 3.8102247983561815e-05,
      "loss": 0.5671,
      "mean_token_accuracy": 0.8236556090414524,
      "num_tokens": 644199380.0,
      "step": 20184
    },
    {
      "entropy": 0.6254759691655636,
      "epoch": 1.857131356346427,
      "grad_norm": 0.15736596286296844,
      "learning_rate": 3.809918115742019e-05,
      "loss": 0.6068,
      "mean_token_accuracy": 0.8117800578474998,
      "num_tokens": 644230574.0,
      "step": 20185
    },
    {
      "entropy": 0.555052244104445,
      "epoch": 1.8572233627178683,
      "grad_norm": 0.1507463902235031,
      "learning_rate": 3.809611433127856e-05,
      "loss": 0.5683,
      "mean_token_accuracy": 0.8255070559680462,
      "num_tokens": 644261984.0,
      "step": 20186
    },
    {
      "entropy": 0.5114808119833469,
      "epoch": 1.8573153690893094,
      "grad_norm": 0.14507544040679932,
      "learning_rate": 3.809304750513694e-05,
      "loss": 0.5001,
      "mean_token_accuracy": 0.8444084599614143,
      "num_tokens": 644293994.0,
      "step": 20187
    },
    {
      "entropy": 0.7252793367952108,
      "epoch": 1.8574073754607507,
      "grad_norm": 0.16865509748458862,
      "learning_rate": 3.808998067899531e-05,
      "loss": 0.7462,
      "mean_token_accuracy": 0.7764151729643345,
      "num_tokens": 644325460.0,
      "step": 20188
    },
    {
      "entropy": 0.5888302968814969,
      "epoch": 1.8574993818321919,
      "grad_norm": 0.15637542307376862,
      "learning_rate": 3.808691385285368e-05,
      "loss": 0.5771,
      "mean_token_accuracy": 0.8244288489222527,
      "num_tokens": 644356618.0,
      "step": 20189
    },
    {
      "entropy": 0.7313155476003885,
      "epoch": 1.857591388203633,
      "grad_norm": 0.16373324394226074,
      "learning_rate": 3.8083847026712056e-05,
      "loss": 0.7162,
      "mean_token_accuracy": 0.7810565829277039,
      "num_tokens": 644387919.0,
      "step": 20190
    },
    {
      "entropy": 0.6348447855561972,
      "epoch": 1.8576833945750744,
      "grad_norm": 0.16101489961147308,
      "learning_rate": 3.808078020057043e-05,
      "loss": 0.6186,
      "mean_token_accuracy": 0.8079969137907028,
      "num_tokens": 644418999.0,
      "step": 20191
    },
    {
      "entropy": 0.5922735594213009,
      "epoch": 1.8577754009465155,
      "grad_norm": 0.15146638453006744,
      "learning_rate": 3.8077713374428805e-05,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8200131356716156,
      "num_tokens": 644451460.0,
      "step": 20192
    },
    {
      "entropy": 0.6229637283831835,
      "epoch": 1.8578674073179569,
      "grad_norm": 0.1517084389925003,
      "learning_rate": 3.807464654828718e-05,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8065945245325565,
      "num_tokens": 644483455.0,
      "step": 20193
    },
    {
      "entropy": 0.6324127987027168,
      "epoch": 1.857959413689398,
      "grad_norm": 0.16079886257648468,
      "learning_rate": 3.8071579722145555e-05,
      "loss": 0.6163,
      "mean_token_accuracy": 0.8115124627947807,
      "num_tokens": 644514704.0,
      "step": 20194
    },
    {
      "entropy": 0.605857552960515,
      "epoch": 1.8580514200608391,
      "grad_norm": 0.1611052006483078,
      "learning_rate": 3.806851289600393e-05,
      "loss": 0.5823,
      "mean_token_accuracy": 0.8186449855566025,
      "num_tokens": 644546071.0,
      "step": 20195
    },
    {
      "entropy": 0.5878856724593788,
      "epoch": 1.8581434264322805,
      "grad_norm": 0.15166962146759033,
      "learning_rate": 3.8065446069862304e-05,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8216238580644131,
      "num_tokens": 644578062.0,
      "step": 20196
    },
    {
      "entropy": 0.6371051301248372,
      "epoch": 1.8582354328037216,
      "grad_norm": 0.1546907275915146,
      "learning_rate": 3.806237924372067e-05,
      "loss": 0.6264,
      "mean_token_accuracy": 0.8091386891901493,
      "num_tokens": 644609971.0,
      "step": 20197
    },
    {
      "entropy": 0.6979553550481796,
      "epoch": 1.858327439175163,
      "grad_norm": 0.1613386869430542,
      "learning_rate": 3.805931241757905e-05,
      "loss": 0.6891,
      "mean_token_accuracy": 0.7887177728116512,
      "num_tokens": 644641866.0,
      "step": 20198
    },
    {
      "entropy": 0.6000586245208979,
      "epoch": 1.858419445546604,
      "grad_norm": 0.1518559455871582,
      "learning_rate": 3.805624559143742e-05,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8215757757425308,
      "num_tokens": 644674540.0,
      "step": 20199
    },
    {
      "entropy": 0.7027628421783447,
      "epoch": 1.8585114519180452,
      "grad_norm": 0.16469117999076843,
      "learning_rate": 3.80531787652958e-05,
      "loss": 0.703,
      "mean_token_accuracy": 0.7903471626341343,
      "num_tokens": 644705826.0,
      "step": 20200
    },
    {
      "entropy": 0.6682853177189827,
      "epoch": 1.8586034582894866,
      "grad_norm": 0.15727585554122925,
      "learning_rate": 3.805011193915417e-05,
      "loss": 0.6712,
      "mean_token_accuracy": 0.7949102595448494,
      "num_tokens": 644738200.0,
      "step": 20201
    },
    {
      "entropy": 0.7122803293168545,
      "epoch": 1.8586954646609277,
      "grad_norm": 0.1666327267885208,
      "learning_rate": 3.8047045113012545e-05,
      "loss": 0.7074,
      "mean_token_accuracy": 0.786298606544733,
      "num_tokens": 644769543.0,
      "step": 20202
    },
    {
      "entropy": 0.5248357485979795,
      "epoch": 1.858787471032369,
      "grad_norm": 0.14800043404102325,
      "learning_rate": 3.804397828687092e-05,
      "loss": 0.5222,
      "mean_token_accuracy": 0.8390410616993904,
      "num_tokens": 644802065.0,
      "step": 20203
    },
    {
      "entropy": 0.5965329553000629,
      "epoch": 1.8588794774038102,
      "grad_norm": 0.15865622460842133,
      "learning_rate": 3.8040911460729294e-05,
      "loss": 0.5848,
      "mean_token_accuracy": 0.8190498314797878,
      "num_tokens": 644833961.0,
      "step": 20204
    },
    {
      "entropy": 0.6611807020381093,
      "epoch": 1.8589714837752513,
      "grad_norm": 0.15140828490257263,
      "learning_rate": 3.803784463458767e-05,
      "loss": 0.6438,
      "mean_token_accuracy": 0.8037238605320454,
      "num_tokens": 644866142.0,
      "step": 20205
    },
    {
      "entropy": 0.6315493693109602,
      "epoch": 1.8590634901466927,
      "grad_norm": 0.15681375563144684,
      "learning_rate": 3.8034777808446044e-05,
      "loss": 0.6174,
      "mean_token_accuracy": 0.8114238865673542,
      "num_tokens": 644898296.0,
      "step": 20206
    },
    {
      "entropy": 0.5514801922254264,
      "epoch": 1.8591554965181338,
      "grad_norm": 0.15277330577373505,
      "learning_rate": 3.803171098230441e-05,
      "loss": 0.5317,
      "mean_token_accuracy": 0.8331925980746746,
      "num_tokens": 644929781.0,
      "step": 20207
    },
    {
      "entropy": 0.6815092326141894,
      "epoch": 1.8592475028895752,
      "grad_norm": 0.16153867542743683,
      "learning_rate": 3.802864415616279e-05,
      "loss": 0.6693,
      "mean_token_accuracy": 0.7933277413249016,
      "num_tokens": 644961914.0,
      "step": 20208
    },
    {
      "entropy": 0.6086123478598893,
      "epoch": 1.8593395092610163,
      "grad_norm": 0.1595204621553421,
      "learning_rate": 3.802557733002116e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.818330492824316,
      "num_tokens": 644993010.0,
      "step": 20209
    },
    {
      "entropy": 0.5847871480509639,
      "epoch": 1.8594315156324575,
      "grad_norm": 0.15228207409381866,
      "learning_rate": 3.8022510503879536e-05,
      "loss": 0.581,
      "mean_token_accuracy": 0.8198279030621052,
      "num_tokens": 645024691.0,
      "step": 20210
    },
    {
      "entropy": 0.6468964051455259,
      "epoch": 1.8595235220038988,
      "grad_norm": 0.15652546286582947,
      "learning_rate": 3.801944367773791e-05,
      "loss": 0.6376,
      "mean_token_accuracy": 0.8060960322618484,
      "num_tokens": 645056900.0,
      "step": 20211
    },
    {
      "entropy": 0.6322236796841025,
      "epoch": 1.85961552837534,
      "grad_norm": 0.16328980028629303,
      "learning_rate": 3.8016376851596285e-05,
      "loss": 0.6269,
      "mean_token_accuracy": 0.806219644844532,
      "num_tokens": 645088152.0,
      "step": 20212
    },
    {
      "entropy": 0.5940929856151342,
      "epoch": 1.8597075347467813,
      "grad_norm": 0.1501302719116211,
      "learning_rate": 3.801331002545466e-05,
      "loss": 0.5844,
      "mean_token_accuracy": 0.8205597102642059,
      "num_tokens": 645120430.0,
      "step": 20213
    },
    {
      "entropy": 0.6258723102509975,
      "epoch": 1.8597995411182224,
      "grad_norm": 0.15844371914863586,
      "learning_rate": 3.8010243199313034e-05,
      "loss": 0.6085,
      "mean_token_accuracy": 0.8155757933855057,
      "num_tokens": 645152951.0,
      "step": 20214
    },
    {
      "entropy": 0.5167998420074582,
      "epoch": 1.8598915474896636,
      "grad_norm": 0.14283926784992218,
      "learning_rate": 3.80071763731714e-05,
      "loss": 0.5069,
      "mean_token_accuracy": 0.8435018621385098,
      "num_tokens": 645185581.0,
      "step": 20215
    },
    {
      "entropy": 0.6207781576085836,
      "epoch": 1.859983553861105,
      "grad_norm": 0.15141962468624115,
      "learning_rate": 3.8004109547029783e-05,
      "loss": 0.6099,
      "mean_token_accuracy": 0.8192229904234409,
      "num_tokens": 645218067.0,
      "step": 20216
    },
    {
      "entropy": 0.5385414347983897,
      "epoch": 1.860075560232546,
      "grad_norm": 0.14527647197246552,
      "learning_rate": 3.800104272088815e-05,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8335303924977779,
      "num_tokens": 645250544.0,
      "step": 20217
    },
    {
      "entropy": 0.6443868158385158,
      "epoch": 1.8601675666039874,
      "grad_norm": 0.1564304232597351,
      "learning_rate": 3.7997975894746526e-05,
      "loss": 0.6314,
      "mean_token_accuracy": 0.8057489842176437,
      "num_tokens": 645281830.0,
      "step": 20218
    },
    {
      "entropy": 0.5901691946201026,
      "epoch": 1.8602595729754285,
      "grad_norm": 0.14882604777812958,
      "learning_rate": 3.79949090686049e-05,
      "loss": 0.5812,
      "mean_token_accuracy": 0.8222624845802784,
      "num_tokens": 645314363.0,
      "step": 20219
    },
    {
      "entropy": 0.6935639623552561,
      "epoch": 1.8603515793468697,
      "grad_norm": 0.15529774129390717,
      "learning_rate": 3.7991842242463275e-05,
      "loss": 0.6866,
      "mean_token_accuracy": 0.7927219942212105,
      "num_tokens": 645347093.0,
      "step": 20220
    },
    {
      "entropy": 0.5999538563191891,
      "epoch": 1.860443585718311,
      "grad_norm": 0.15091365575790405,
      "learning_rate": 3.798877541632165e-05,
      "loss": 0.5771,
      "mean_token_accuracy": 0.8212966844439507,
      "num_tokens": 645378300.0,
      "step": 20221
    },
    {
      "entropy": 0.5345150413922966,
      "epoch": 1.8605355920897522,
      "grad_norm": 0.15177005529403687,
      "learning_rate": 3.7985708590180025e-05,
      "loss": 0.5207,
      "mean_token_accuracy": 0.839471310377121,
      "num_tokens": 645409864.0,
      "step": 20222
    },
    {
      "entropy": 0.5405553141608834,
      "epoch": 1.8606275984611935,
      "grad_norm": 0.14782263338565826,
      "learning_rate": 3.798264176403839e-05,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8332877084612846,
      "num_tokens": 645442017.0,
      "step": 20223
    },
    {
      "entropy": 0.6185010238550603,
      "epoch": 1.8607196048326347,
      "grad_norm": 0.15256857872009277,
      "learning_rate": 3.7979574937896774e-05,
      "loss": 0.6049,
      "mean_token_accuracy": 0.8102580159902573,
      "num_tokens": 645474346.0,
      "step": 20224
    },
    {
      "entropy": 0.6185669489204884,
      "epoch": 1.8608116112040758,
      "grad_norm": 0.15296731889247894,
      "learning_rate": 3.797650811175515e-05,
      "loss": 0.6084,
      "mean_token_accuracy": 0.8082662411034107,
      "num_tokens": 645506298.0,
      "step": 20225
    },
    {
      "entropy": 0.651063296943903,
      "epoch": 1.8609036175755171,
      "grad_norm": 0.152785986661911,
      "learning_rate": 3.797344128561352e-05,
      "loss": 0.6424,
      "mean_token_accuracy": 0.8015863858163357,
      "num_tokens": 645538764.0,
      "step": 20226
    },
    {
      "entropy": 0.6298990743234754,
      "epoch": 1.8609956239469583,
      "grad_norm": 0.15484680235385895,
      "learning_rate": 3.79703744594719e-05,
      "loss": 0.618,
      "mean_token_accuracy": 0.8121781200170517,
      "num_tokens": 645570450.0,
      "step": 20227
    },
    {
      "entropy": 0.5715260356664658,
      "epoch": 1.8610876303183996,
      "grad_norm": 0.14961791038513184,
      "learning_rate": 3.7967307633330266e-05,
      "loss": 0.5547,
      "mean_token_accuracy": 0.8293020315468311,
      "num_tokens": 645602546.0,
      "step": 20228
    },
    {
      "entropy": 0.5587880173698068,
      "epoch": 1.8611796366898408,
      "grad_norm": 0.15061630308628082,
      "learning_rate": 3.796424080718865e-05,
      "loss": 0.5429,
      "mean_token_accuracy": 0.830343808978796,
      "num_tokens": 645634456.0,
      "step": 20229
    },
    {
      "entropy": 0.4921937435865402,
      "epoch": 1.861271643061282,
      "grad_norm": 0.1467963308095932,
      "learning_rate": 3.7961173981047015e-05,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8494103662669659,
      "num_tokens": 645666192.0,
      "step": 20230
    },
    {
      "entropy": 0.5011443821713328,
      "epoch": 1.8613636494327233,
      "grad_norm": 0.14566612243652344,
      "learning_rate": 3.795810715490539e-05,
      "loss": 0.491,
      "mean_token_accuracy": 0.8477202616631985,
      "num_tokens": 645698174.0,
      "step": 20231
    },
    {
      "entropy": 0.5392362074926496,
      "epoch": 1.8614556558041644,
      "grad_norm": 0.14587754011154175,
      "learning_rate": 3.7955040328763764e-05,
      "loss": 0.528,
      "mean_token_accuracy": 0.8363351784646511,
      "num_tokens": 645730406.0,
      "step": 20232
    },
    {
      "entropy": 0.5661570429801941,
      "epoch": 1.8615476621756057,
      "grad_norm": 0.15715239942073822,
      "learning_rate": 3.795197350262214e-05,
      "loss": 0.5602,
      "mean_token_accuracy": 0.8298424556851387,
      "num_tokens": 645763174.0,
      "step": 20233
    },
    {
      "entropy": 0.5110070314258337,
      "epoch": 1.8616396685470469,
      "grad_norm": 0.1449805200099945,
      "learning_rate": 3.7948906676480514e-05,
      "loss": 0.5038,
      "mean_token_accuracy": 0.8464826606214046,
      "num_tokens": 645795942.0,
      "step": 20234
    },
    {
      "entropy": 0.6580328904092312,
      "epoch": 1.861731674918488,
      "grad_norm": 0.1569705605506897,
      "learning_rate": 3.794583985033889e-05,
      "loss": 0.654,
      "mean_token_accuracy": 0.7987189516425133,
      "num_tokens": 645827925.0,
      "step": 20235
    },
    {
      "entropy": 0.5492188271600753,
      "epoch": 1.8618236812899294,
      "grad_norm": 0.145441934466362,
      "learning_rate": 3.7942773024197256e-05,
      "loss": 0.5365,
      "mean_token_accuracy": 0.8323618322610855,
      "num_tokens": 645859635.0,
      "step": 20236
    },
    {
      "entropy": 0.6197147471830249,
      "epoch": 1.8619156876613705,
      "grad_norm": 0.15288394689559937,
      "learning_rate": 3.793970619805564e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8079299367964268,
      "num_tokens": 645891552.0,
      "step": 20237
    },
    {
      "entropy": 0.5866594221442938,
      "epoch": 1.8620076940328119,
      "grad_norm": 0.15569336712360382,
      "learning_rate": 3.7936639371914005e-05,
      "loss": 0.572,
      "mean_token_accuracy": 0.8277784660458565,
      "num_tokens": 645922795.0,
      "step": 20238
    },
    {
      "entropy": 0.6416776878759265,
      "epoch": 1.862099700404253,
      "grad_norm": 0.1527707278728485,
      "learning_rate": 3.793357254577238e-05,
      "loss": 0.6197,
      "mean_token_accuracy": 0.808171235024929,
      "num_tokens": 645954693.0,
      "step": 20239
    },
    {
      "entropy": 0.6075128316879272,
      "epoch": 1.8621917067756941,
      "grad_norm": 0.1501515507698059,
      "learning_rate": 3.7930505719630755e-05,
      "loss": 0.5982,
      "mean_token_accuracy": 0.8142769485712051,
      "num_tokens": 645986651.0,
      "step": 20240
    },
    {
      "entropy": 0.5770551422610879,
      "epoch": 1.8622837131471355,
      "grad_norm": 0.15230470895767212,
      "learning_rate": 3.792743889348913e-05,
      "loss": 0.566,
      "mean_token_accuracy": 0.8247948475182056,
      "num_tokens": 646018141.0,
      "step": 20241
    },
    {
      "entropy": 0.6433315631002188,
      "epoch": 1.8623757195185766,
      "grad_norm": 0.15834422409534454,
      "learning_rate": 3.7924372067347504e-05,
      "loss": 0.6286,
      "mean_token_accuracy": 0.8086157441139221,
      "num_tokens": 646050082.0,
      "step": 20242
    },
    {
      "entropy": 0.6325675984844565,
      "epoch": 1.862467725890018,
      "grad_norm": 0.15082231163978577,
      "learning_rate": 3.792130524120588e-05,
      "loss": 0.6203,
      "mean_token_accuracy": 0.8102931529283524,
      "num_tokens": 646082190.0,
      "step": 20243
    },
    {
      "entropy": 0.4941807631403208,
      "epoch": 1.862559732261459,
      "grad_norm": 0.14447155594825745,
      "learning_rate": 3.791823841506425e-05,
      "loss": 0.4839,
      "mean_token_accuracy": 0.8521003499627113,
      "num_tokens": 646114154.0,
      "step": 20244
    },
    {
      "entropy": 0.5711313467472792,
      "epoch": 1.8626517386329002,
      "grad_norm": 0.15275993943214417,
      "learning_rate": 3.791517158892263e-05,
      "loss": 0.5696,
      "mean_token_accuracy": 0.8205891586840153,
      "num_tokens": 646145809.0,
      "step": 20245
    },
    {
      "entropy": 0.5685192905366421,
      "epoch": 1.8627437450043416,
      "grad_norm": 0.14746713638305664,
      "learning_rate": 3.7912104762780996e-05,
      "loss": 0.5561,
      "mean_token_accuracy": 0.8269962854683399,
      "num_tokens": 646177989.0,
      "step": 20246
    },
    {
      "entropy": 0.6809146422892809,
      "epoch": 1.8628357513757827,
      "grad_norm": 0.1582629233598709,
      "learning_rate": 3.790903793663938e-05,
      "loss": 0.6627,
      "mean_token_accuracy": 0.7977814190089703,
      "num_tokens": 646209609.0,
      "step": 20247
    },
    {
      "entropy": 0.5340476045385003,
      "epoch": 1.862927757747224,
      "grad_norm": 0.14824619889259338,
      "learning_rate": 3.7905971110497745e-05,
      "loss": 0.528,
      "mean_token_accuracy": 0.835956584662199,
      "num_tokens": 646240932.0,
      "step": 20248
    },
    {
      "entropy": 0.6005916716530919,
      "epoch": 1.8630197641186652,
      "grad_norm": 0.15623407065868378,
      "learning_rate": 3.790290428435612e-05,
      "loss": 0.5965,
      "mean_token_accuracy": 0.815921388566494,
      "num_tokens": 646273341.0,
      "step": 20249
    },
    {
      "entropy": 0.5542439334094524,
      "epoch": 1.8631117704901063,
      "grad_norm": 0.15468217432498932,
      "learning_rate": 3.7899837458214494e-05,
      "loss": 0.5367,
      "mean_token_accuracy": 0.8332127220928669,
      "num_tokens": 646305284.0,
      "step": 20250
    },
    {
      "entropy": 0.6356377629563212,
      "epoch": 1.8632037768615477,
      "grad_norm": 0.15608367323875427,
      "learning_rate": 3.789677063207287e-05,
      "loss": 0.6177,
      "mean_token_accuracy": 0.8080092258751392,
      "num_tokens": 646336681.0,
      "step": 20251
    },
    {
      "entropy": 0.6113678868860006,
      "epoch": 1.8632957832329888,
      "grad_norm": 0.14713306725025177,
      "learning_rate": 3.7893703805931244e-05,
      "loss": 0.5921,
      "mean_token_accuracy": 0.8191657140851021,
      "num_tokens": 646368918.0,
      "step": 20252
    },
    {
      "entropy": 0.6074873739853501,
      "epoch": 1.8633877896044302,
      "grad_norm": 0.14712001383304596,
      "learning_rate": 3.789063697978962e-05,
      "loss": 0.6051,
      "mean_token_accuracy": 0.8168438971042633,
      "num_tokens": 646401348.0,
      "step": 20253
    },
    {
      "entropy": 0.6623212695121765,
      "epoch": 1.8634797959758713,
      "grad_norm": 0.15555505454540253,
      "learning_rate": 3.7887570153647986e-05,
      "loss": 0.6559,
      "mean_token_accuracy": 0.7977797761559486,
      "num_tokens": 646434099.0,
      "step": 20254
    },
    {
      "entropy": 0.5215495261363685,
      "epoch": 1.8635718023473125,
      "grad_norm": 0.14527864754199982,
      "learning_rate": 3.788450332750637e-05,
      "loss": 0.5138,
      "mean_token_accuracy": 0.8426460847258568,
      "num_tokens": 646466184.0,
      "step": 20255
    },
    {
      "entropy": 0.6398505177348852,
      "epoch": 1.8636638087187538,
      "grad_norm": 0.15369053184986115,
      "learning_rate": 3.788143650136474e-05,
      "loss": 0.6392,
      "mean_token_accuracy": 0.8046781122684479,
      "num_tokens": 646498613.0,
      "step": 20256
    },
    {
      "entropy": 0.6029981626197696,
      "epoch": 1.863755815090195,
      "grad_norm": 0.15560349822044373,
      "learning_rate": 3.787836967522311e-05,
      "loss": 0.5904,
      "mean_token_accuracy": 0.8190376609563828,
      "num_tokens": 646530283.0,
      "step": 20257
    },
    {
      "entropy": 0.6486782114952803,
      "epoch": 1.8638478214616363,
      "grad_norm": 0.15681372582912445,
      "learning_rate": 3.787530284908149e-05,
      "loss": 0.6307,
      "mean_token_accuracy": 0.8099173009395599,
      "num_tokens": 646561976.0,
      "step": 20258
    },
    {
      "entropy": 0.4943433655425906,
      "epoch": 1.8639398278330774,
      "grad_norm": 0.15138697624206543,
      "learning_rate": 3.787223602293986e-05,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8470326028764248,
      "num_tokens": 646593325.0,
      "step": 20259
    },
    {
      "entropy": 0.6257907161489129,
      "epoch": 1.8640318342045186,
      "grad_norm": 0.1572730392217636,
      "learning_rate": 3.786916919679824e-05,
      "loss": 0.6179,
      "mean_token_accuracy": 0.8099597916007042,
      "num_tokens": 646624796.0,
      "step": 20260
    },
    {
      "entropy": 0.5474969877395779,
      "epoch": 1.86412384057596,
      "grad_norm": 0.14498218894004822,
      "learning_rate": 3.786610237065661e-05,
      "loss": 0.5285,
      "mean_token_accuracy": 0.8376217745244503,
      "num_tokens": 646656215.0,
      "step": 20261
    },
    {
      "entropy": 0.6555479122325778,
      "epoch": 1.864215846947401,
      "grad_norm": 0.16302649676799774,
      "learning_rate": 3.786303554451498e-05,
      "loss": 0.645,
      "mean_token_accuracy": 0.8050543777644634,
      "num_tokens": 646688012.0,
      "step": 20262
    },
    {
      "entropy": 0.6083686016499996,
      "epoch": 1.8643078533188424,
      "grad_norm": 0.15425902605056763,
      "learning_rate": 3.785996871837336e-05,
      "loss": 0.5977,
      "mean_token_accuracy": 0.819736909121275,
      "num_tokens": 646720051.0,
      "step": 20263
    },
    {
      "entropy": 0.581524197012186,
      "epoch": 1.8643998596902835,
      "grad_norm": 0.15265578031539917,
      "learning_rate": 3.785690189223173e-05,
      "loss": 0.5708,
      "mean_token_accuracy": 0.8242971114814281,
      "num_tokens": 646752016.0,
      "step": 20264
    },
    {
      "entropy": 0.5931144962087274,
      "epoch": 1.8644918660617247,
      "grad_norm": 0.15153630077838898,
      "learning_rate": 3.785383506609011e-05,
      "loss": 0.5827,
      "mean_token_accuracy": 0.8224877640604973,
      "num_tokens": 646783764.0,
      "step": 20265
    },
    {
      "entropy": 0.6327254511415958,
      "epoch": 1.864583872433166,
      "grad_norm": 0.15576379001140594,
      "learning_rate": 3.785076823994848e-05,
      "loss": 0.6338,
      "mean_token_accuracy": 0.805065780878067,
      "num_tokens": 646816288.0,
      "step": 20266
    },
    {
      "entropy": 0.5737252850085497,
      "epoch": 1.8646758788046072,
      "grad_norm": 0.14750415086746216,
      "learning_rate": 3.784770141380685e-05,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8251942284405231,
      "num_tokens": 646849054.0,
      "step": 20267
    },
    {
      "entropy": 0.5886333901435137,
      "epoch": 1.8647678851760485,
      "grad_norm": 0.16020312905311584,
      "learning_rate": 3.784463458766523e-05,
      "loss": 0.5837,
      "mean_token_accuracy": 0.8216510005295277,
      "num_tokens": 646880537.0,
      "step": 20268
    },
    {
      "entropy": 0.622876176610589,
      "epoch": 1.8648598915474897,
      "grad_norm": 0.15742383897304535,
      "learning_rate": 3.78415677615236e-05,
      "loss": 0.6202,
      "mean_token_accuracy": 0.8128629550337791,
      "num_tokens": 646912051.0,
      "step": 20269
    },
    {
      "entropy": 0.5844016252085567,
      "epoch": 1.8649518979189308,
      "grad_norm": 0.15926016867160797,
      "learning_rate": 3.7838500935381974e-05,
      "loss": 0.5814,
      "mean_token_accuracy": 0.8234496042132378,
      "num_tokens": 646943671.0,
      "step": 20270
    },
    {
      "entropy": 0.5370414825156331,
      "epoch": 1.8650439042903721,
      "grad_norm": 0.1427209973335266,
      "learning_rate": 3.783543410924035e-05,
      "loss": 0.5248,
      "mean_token_accuracy": 0.8406646251678467,
      "num_tokens": 646975784.0,
      "step": 20271
    },
    {
      "entropy": 0.784557699225843,
      "epoch": 1.8651359106618133,
      "grad_norm": 0.16776598989963531,
      "learning_rate": 3.783236728309872e-05,
      "loss": 0.7843,
      "mean_token_accuracy": 0.7638371773064137,
      "num_tokens": 647007657.0,
      "step": 20272
    },
    {
      "entropy": 0.6117871711030602,
      "epoch": 1.8652279170332546,
      "grad_norm": 0.15370634198188782,
      "learning_rate": 3.78293004569571e-05,
      "loss": 0.6076,
      "mean_token_accuracy": 0.8154137022793293,
      "num_tokens": 647039804.0,
      "step": 20273
    },
    {
      "entropy": 0.5981995170004666,
      "epoch": 1.8653199234046958,
      "grad_norm": 0.15203747153282166,
      "learning_rate": 3.782623363081547e-05,
      "loss": 0.5866,
      "mean_token_accuracy": 0.82015335932374,
      "num_tokens": 647071919.0,
      "step": 20274
    },
    {
      "entropy": 0.5727057547774166,
      "epoch": 1.865411929776137,
      "grad_norm": 0.15208253264427185,
      "learning_rate": 3.782316680467384e-05,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8255399577319622,
      "num_tokens": 647104068.0,
      "step": 20275
    },
    {
      "entropy": 0.5016544638201594,
      "epoch": 1.8655039361475783,
      "grad_norm": 0.14715895056724548,
      "learning_rate": 3.782009997853222e-05,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8442498929798603,
      "num_tokens": 647136158.0,
      "step": 20276
    },
    {
      "entropy": 0.5040190741419792,
      "epoch": 1.8655959425190194,
      "grad_norm": 0.15116716921329498,
      "learning_rate": 3.781703315239059e-05,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8494429849088192,
      "num_tokens": 647168600.0,
      "step": 20277
    },
    {
      "entropy": 0.5341287571936846,
      "epoch": 1.8656879488904607,
      "grad_norm": 0.14642980694770813,
      "learning_rate": 3.7813966326248964e-05,
      "loss": 0.5254,
      "mean_token_accuracy": 0.838009238243103,
      "num_tokens": 647201267.0,
      "step": 20278
    },
    {
      "entropy": 0.5536263789981604,
      "epoch": 1.8657799552619019,
      "grad_norm": 0.14925885200500488,
      "learning_rate": 3.781089950010734e-05,
      "loss": 0.5418,
      "mean_token_accuracy": 0.8311180621385574,
      "num_tokens": 647233565.0,
      "step": 20279
    },
    {
      "entropy": 0.5869045425206423,
      "epoch": 1.865871961633343,
      "grad_norm": 0.14868156611919403,
      "learning_rate": 3.7807832673965713e-05,
      "loss": 0.5721,
      "mean_token_accuracy": 0.8246258907020092,
      "num_tokens": 647265714.0,
      "step": 20280
    },
    {
      "entropy": 0.4696218236349523,
      "epoch": 1.8659639680047844,
      "grad_norm": 0.14237754046916962,
      "learning_rate": 3.780476584782409e-05,
      "loss": 0.4473,
      "mean_token_accuracy": 0.8615485951304436,
      "num_tokens": 647297762.0,
      "step": 20281
    },
    {
      "entropy": 0.5157052148133516,
      "epoch": 1.8660559743762255,
      "grad_norm": 0.13987189531326294,
      "learning_rate": 3.780169902168246e-05,
      "loss": 0.5069,
      "mean_token_accuracy": 0.8437850251793861,
      "num_tokens": 647330137.0,
      "step": 20282
    },
    {
      "entropy": 0.6486897836439312,
      "epoch": 1.8661479807476669,
      "grad_norm": 0.15603405237197876,
      "learning_rate": 3.779863219554083e-05,
      "loss": 0.6356,
      "mean_token_accuracy": 0.8036926537752151,
      "num_tokens": 647362019.0,
      "step": 20283
    },
    {
      "entropy": 0.6083865016698837,
      "epoch": 1.866239987119108,
      "grad_norm": 0.15216578543186188,
      "learning_rate": 3.779556536939921e-05,
      "loss": 0.5956,
      "mean_token_accuracy": 0.8182044886052608,
      "num_tokens": 647394275.0,
      "step": 20284
    },
    {
      "entropy": 0.6304816771298647,
      "epoch": 1.8663319934905491,
      "grad_norm": 0.15865975618362427,
      "learning_rate": 3.779249854325758e-05,
      "loss": 0.6182,
      "mean_token_accuracy": 0.8047781363129616,
      "num_tokens": 647425928.0,
      "step": 20285
    },
    {
      "entropy": 0.6534892302006483,
      "epoch": 1.8664239998619905,
      "grad_norm": 0.1622002124786377,
      "learning_rate": 3.778943171711596e-05,
      "loss": 0.6414,
      "mean_token_accuracy": 0.8052781522274017,
      "num_tokens": 647457078.0,
      "step": 20286
    },
    {
      "entropy": 0.5652960762381554,
      "epoch": 1.8665160062334316,
      "grad_norm": 0.14523190259933472,
      "learning_rate": 3.7786364890974336e-05,
      "loss": 0.5463,
      "mean_token_accuracy": 0.8342987075448036,
      "num_tokens": 647489586.0,
      "step": 20287
    },
    {
      "entropy": 0.48390603717416525,
      "epoch": 1.866608012604873,
      "grad_norm": 0.14725853502750397,
      "learning_rate": 3.7783298064832704e-05,
      "loss": 0.4738,
      "mean_token_accuracy": 0.8521615080535412,
      "num_tokens": 647521305.0,
      "step": 20288
    },
    {
      "entropy": 0.5178457908332348,
      "epoch": 1.866700018976314,
      "grad_norm": 0.14097563922405243,
      "learning_rate": 3.7780231238691085e-05,
      "loss": 0.5066,
      "mean_token_accuracy": 0.840353824198246,
      "num_tokens": 647553218.0,
      "step": 20289
    },
    {
      "entropy": 0.5697726290673018,
      "epoch": 1.8667920253477552,
      "grad_norm": 0.14763273298740387,
      "learning_rate": 3.777716441254945e-05,
      "loss": 0.5599,
      "mean_token_accuracy": 0.8334154263138771,
      "num_tokens": 647585562.0,
      "step": 20290
    },
    {
      "entropy": 0.6296047223731875,
      "epoch": 1.8668840317191966,
      "grad_norm": 0.16171382367610931,
      "learning_rate": 3.777409758640783e-05,
      "loss": 0.6172,
      "mean_token_accuracy": 0.811497375369072,
      "num_tokens": 647617899.0,
      "step": 20291
    },
    {
      "entropy": 0.5194513760507107,
      "epoch": 1.8669760380906377,
      "grad_norm": 0.14728017151355743,
      "learning_rate": 3.77710307602662e-05,
      "loss": 0.5149,
      "mean_token_accuracy": 0.8395626470446587,
      "num_tokens": 647649164.0,
      "step": 20292
    },
    {
      "entropy": 0.5906245689839125,
      "epoch": 1.867068044462079,
      "grad_norm": 0.1484692245721817,
      "learning_rate": 3.776796393412458e-05,
      "loss": 0.5763,
      "mean_token_accuracy": 0.8249156326055527,
      "num_tokens": 647681097.0,
      "step": 20293
    },
    {
      "entropy": 0.544868178665638,
      "epoch": 1.8671600508335202,
      "grad_norm": 0.14869827032089233,
      "learning_rate": 3.776489710798295e-05,
      "loss": 0.54,
      "mean_token_accuracy": 0.8396576680243015,
      "num_tokens": 647713396.0,
      "step": 20294
    },
    {
      "entropy": 0.6378877852112055,
      "epoch": 1.8672520572049613,
      "grad_norm": 0.15964724123477936,
      "learning_rate": 3.7761830281841326e-05,
      "loss": 0.6332,
      "mean_token_accuracy": 0.8072122633457184,
      "num_tokens": 647745415.0,
      "step": 20295
    },
    {
      "entropy": 0.5943547878414392,
      "epoch": 1.8673440635764027,
      "grad_norm": 0.15255357325077057,
      "learning_rate": 3.7758763455699694e-05,
      "loss": 0.5944,
      "mean_token_accuracy": 0.8193650133907795,
      "num_tokens": 647778120.0,
      "step": 20296
    },
    {
      "entropy": 0.8063027765601873,
      "epoch": 1.8674360699478438,
      "grad_norm": 0.1700294315814972,
      "learning_rate": 3.7755696629558076e-05,
      "loss": 0.7876,
      "mean_token_accuracy": 0.767342459410429,
      "num_tokens": 647810888.0,
      "step": 20297
    },
    {
      "entropy": 0.5859038224443793,
      "epoch": 1.8675280763192852,
      "grad_norm": 0.14889730513095856,
      "learning_rate": 3.7752629803416444e-05,
      "loss": 0.5693,
      "mean_token_accuracy": 0.8259221538901329,
      "num_tokens": 647842649.0,
      "step": 20298
    },
    {
      "entropy": 0.6019263043999672,
      "epoch": 1.8676200826907263,
      "grad_norm": 0.15576234459877014,
      "learning_rate": 3.7749562977274825e-05,
      "loss": 0.579,
      "mean_token_accuracy": 0.8175527900457382,
      "num_tokens": 647873724.0,
      "step": 20299
    },
    {
      "entropy": 0.6271724421530962,
      "epoch": 1.8677120890621675,
      "grad_norm": 0.15647239983081818,
      "learning_rate": 3.774649615113319e-05,
      "loss": 0.6357,
      "mean_token_accuracy": 0.8054239973425865,
      "num_tokens": 647905969.0,
      "step": 20300
    },
    {
      "entropy": 0.6194916032254696,
      "epoch": 1.8678040954336088,
      "grad_norm": 0.15713845193386078,
      "learning_rate": 3.774342932499157e-05,
      "loss": 0.6109,
      "mean_token_accuracy": 0.8118647560477257,
      "num_tokens": 647936997.0,
      "step": 20301
    },
    {
      "entropy": 0.49057427421212196,
      "epoch": 1.86789610180505,
      "grad_norm": 0.1479576677083969,
      "learning_rate": 3.774036249884994e-05,
      "loss": 0.4831,
      "mean_token_accuracy": 0.8501548357307911,
      "num_tokens": 647969287.0,
      "step": 20302
    },
    {
      "entropy": 0.5500474269501865,
      "epoch": 1.8679881081764913,
      "grad_norm": 0.1512787640094757,
      "learning_rate": 3.773729567270832e-05,
      "loss": 0.543,
      "mean_token_accuracy": 0.8331368006765842,
      "num_tokens": 648001554.0,
      "step": 20303
    },
    {
      "entropy": 0.6174688674509525,
      "epoch": 1.8680801145479324,
      "grad_norm": 0.1548098623752594,
      "learning_rate": 3.773422884656669e-05,
      "loss": 0.609,
      "mean_token_accuracy": 0.8179096803069115,
      "num_tokens": 648033598.0,
      "step": 20304
    },
    {
      "entropy": 0.5514998212456703,
      "epoch": 1.8681721209193736,
      "grad_norm": 0.1515912115573883,
      "learning_rate": 3.7731162020425066e-05,
      "loss": 0.5442,
      "mean_token_accuracy": 0.83470843359828,
      "num_tokens": 648065855.0,
      "step": 20305
    },
    {
      "entropy": 0.6612456850707531,
      "epoch": 1.868264127290815,
      "grad_norm": 0.16457542777061462,
      "learning_rate": 3.7728095194283434e-05,
      "loss": 0.6607,
      "mean_token_accuracy": 0.7976459786295891,
      "num_tokens": 648098310.0,
      "step": 20306
    },
    {
      "entropy": 0.4917286478448659,
      "epoch": 1.868356133662256,
      "grad_norm": 0.1423846036195755,
      "learning_rate": 3.7725028368141815e-05,
      "loss": 0.4806,
      "mean_token_accuracy": 0.850574005395174,
      "num_tokens": 648131078.0,
      "step": 20307
    },
    {
      "entropy": 0.6292599532753229,
      "epoch": 1.8684481400336974,
      "grad_norm": 0.1545303612947464,
      "learning_rate": 3.772196154200018e-05,
      "loss": 0.6254,
      "mean_token_accuracy": 0.8109968081116676,
      "num_tokens": 648162930.0,
      "step": 20308
    },
    {
      "entropy": 0.6473775543272495,
      "epoch": 1.8685401464051385,
      "grad_norm": 0.15519940853118896,
      "learning_rate": 3.771889471585856e-05,
      "loss": 0.635,
      "mean_token_accuracy": 0.8056828826665878,
      "num_tokens": 648194936.0,
      "step": 20309
    },
    {
      "entropy": 0.6426515765488148,
      "epoch": 1.8686321527765797,
      "grad_norm": 0.15276561677455902,
      "learning_rate": 3.771582788971693e-05,
      "loss": 0.6336,
      "mean_token_accuracy": 0.8048768304288387,
      "num_tokens": 648226870.0,
      "step": 20310
    },
    {
      "entropy": 0.6922327326610684,
      "epoch": 1.868724159148021,
      "grad_norm": 0.1576477289199829,
      "learning_rate": 3.771276106357531e-05,
      "loss": 0.6797,
      "mean_token_accuracy": 0.7935947477817535,
      "num_tokens": 648258803.0,
      "step": 20311
    },
    {
      "entropy": 0.6435654642991722,
      "epoch": 1.8688161655194622,
      "grad_norm": 0.15383493900299072,
      "learning_rate": 3.770969423743368e-05,
      "loss": 0.6246,
      "mean_token_accuracy": 0.8079394772648811,
      "num_tokens": 648291268.0,
      "step": 20312
    },
    {
      "entropy": 0.6236674562096596,
      "epoch": 1.8689081718909035,
      "grad_norm": 0.16026154160499573,
      "learning_rate": 3.7706627411292057e-05,
      "loss": 0.6149,
      "mean_token_accuracy": 0.812560647726059,
      "num_tokens": 648322781.0,
      "step": 20313
    },
    {
      "entropy": 0.7447362579405308,
      "epoch": 1.8690001782623447,
      "grad_norm": 0.1682957410812378,
      "learning_rate": 3.7703560585150424e-05,
      "loss": 0.7258,
      "mean_token_accuracy": 0.781266164034605,
      "num_tokens": 648353822.0,
      "step": 20314
    },
    {
      "entropy": 0.5959836775436997,
      "epoch": 1.8690921846337858,
      "grad_norm": 0.16029833257198334,
      "learning_rate": 3.7700493759008806e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8215759135782719,
      "num_tokens": 648385210.0,
      "step": 20315
    },
    {
      "entropy": 0.7347798943519592,
      "epoch": 1.8691841910052271,
      "grad_norm": 0.16252703964710236,
      "learning_rate": 3.769742693286718e-05,
      "loss": 0.7278,
      "mean_token_accuracy": 0.7741533890366554,
      "num_tokens": 648417626.0,
      "step": 20316
    },
    {
      "entropy": 0.6349491216242313,
      "epoch": 1.8692761973766683,
      "grad_norm": 0.16186261177062988,
      "learning_rate": 3.769436010672555e-05,
      "loss": 0.6296,
      "mean_token_accuracy": 0.8074848614633083,
      "num_tokens": 648448871.0,
      "step": 20317
    },
    {
      "entropy": 0.6241475897841156,
      "epoch": 1.8693682037481096,
      "grad_norm": 0.15938769280910492,
      "learning_rate": 3.769129328058393e-05,
      "loss": 0.6151,
      "mean_token_accuracy": 0.8106968067586422,
      "num_tokens": 648480603.0,
      "step": 20318
    },
    {
      "entropy": 0.51923543587327,
      "epoch": 1.8694602101195508,
      "grad_norm": 0.1506064087152481,
      "learning_rate": 3.76882264544423e-05,
      "loss": 0.5042,
      "mean_token_accuracy": 0.8442118428647518,
      "num_tokens": 648512631.0,
      "step": 20319
    },
    {
      "entropy": 0.6255223206244409,
      "epoch": 1.869552216490992,
      "grad_norm": 0.15500755608081818,
      "learning_rate": 3.768515962830068e-05,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8114504665136337,
      "num_tokens": 648544813.0,
      "step": 20320
    },
    {
      "entropy": 0.5270634947810322,
      "epoch": 1.8696442228624333,
      "grad_norm": 0.14166511595249176,
      "learning_rate": 3.768209280215905e-05,
      "loss": 0.519,
      "mean_token_accuracy": 0.8398987129330635,
      "num_tokens": 648576476.0,
      "step": 20321
    },
    {
      "entropy": 0.6625029090791941,
      "epoch": 1.8697362292338744,
      "grad_norm": 0.15811687707901,
      "learning_rate": 3.767902597601742e-05,
      "loss": 0.6623,
      "mean_token_accuracy": 0.7995595298707485,
      "num_tokens": 648607795.0,
      "step": 20322
    },
    {
      "entropy": 0.5139205874875188,
      "epoch": 1.8698282356053157,
      "grad_norm": 0.14801622927188873,
      "learning_rate": 3.7675959149875796e-05,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8482631482183933,
      "num_tokens": 648639587.0,
      "step": 20323
    },
    {
      "entropy": 0.6283845957368612,
      "epoch": 1.8699202419767569,
      "grad_norm": 0.15354406833648682,
      "learning_rate": 3.767289232373417e-05,
      "loss": 0.6059,
      "mean_token_accuracy": 0.8118734434247017,
      "num_tokens": 648670035.0,
      "step": 20324
    },
    {
      "entropy": 0.615868154913187,
      "epoch": 1.870012248348198,
      "grad_norm": 0.16277965903282166,
      "learning_rate": 3.7669825497592545e-05,
      "loss": 0.6037,
      "mean_token_accuracy": 0.8147959820926189,
      "num_tokens": 648701006.0,
      "step": 20325
    },
    {
      "entropy": 0.47828571451827884,
      "epoch": 1.8701042547196394,
      "grad_norm": 0.13970893621444702,
      "learning_rate": 3.766675867145092e-05,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8547809682786465,
      "num_tokens": 648732988.0,
      "step": 20326
    },
    {
      "entropy": 0.6143448613584042,
      "epoch": 1.8701962610910805,
      "grad_norm": 0.1555081605911255,
      "learning_rate": 3.766369184530929e-05,
      "loss": 0.6082,
      "mean_token_accuracy": 0.8132609874010086,
      "num_tokens": 648764939.0,
      "step": 20327
    },
    {
      "entropy": 0.6527563920244575,
      "epoch": 1.8702882674625219,
      "grad_norm": 0.1594577580690384,
      "learning_rate": 3.766062501916767e-05,
      "loss": 0.6525,
      "mean_token_accuracy": 0.800682820379734,
      "num_tokens": 648796374.0,
      "step": 20328
    },
    {
      "entropy": 0.5852813124656677,
      "epoch": 1.870380273833963,
      "grad_norm": 0.1497090756893158,
      "learning_rate": 3.765755819302604e-05,
      "loss": 0.5832,
      "mean_token_accuracy": 0.8181551210582256,
      "num_tokens": 648827939.0,
      "step": 20329
    },
    {
      "entropy": 0.3839591699652374,
      "epoch": 1.8704722802054041,
      "grad_norm": 0.13552606105804443,
      "learning_rate": 3.765449136688441e-05,
      "loss": 0.3728,
      "mean_token_accuracy": 0.8816293179988861,
      "num_tokens": 648859262.0,
      "step": 20330
    },
    {
      "entropy": 0.6403313409537077,
      "epoch": 1.8705642865768455,
      "grad_norm": 0.1552695482969284,
      "learning_rate": 3.7651424540742787e-05,
      "loss": 0.6213,
      "mean_token_accuracy": 0.8114034160971642,
      "num_tokens": 648891507.0,
      "step": 20331
    },
    {
      "entropy": 0.6599928038194776,
      "epoch": 1.8706562929482866,
      "grad_norm": 0.1617809236049652,
      "learning_rate": 3.764835771460116e-05,
      "loss": 0.6517,
      "mean_token_accuracy": 0.8026236966252327,
      "num_tokens": 648923686.0,
      "step": 20332
    },
    {
      "entropy": 0.630007435567677,
      "epoch": 1.870748299319728,
      "grad_norm": 0.16119159758090973,
      "learning_rate": 3.7645290888459536e-05,
      "loss": 0.6303,
      "mean_token_accuracy": 0.8049723766744137,
      "num_tokens": 648955509.0,
      "step": 20333
    },
    {
      "entropy": 0.7270469088107347,
      "epoch": 1.870840305691169,
      "grad_norm": 0.16466432809829712,
      "learning_rate": 3.764222406231791e-05,
      "loss": 0.7182,
      "mean_token_accuracy": 0.7808149643242359,
      "num_tokens": 648986758.0,
      "step": 20334
    },
    {
      "entropy": 0.5166259557008743,
      "epoch": 1.8709323120626102,
      "grad_norm": 0.15282359719276428,
      "learning_rate": 3.763915723617628e-05,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8401177451014519,
      "num_tokens": 649018982.0,
      "step": 20335
    },
    {
      "entropy": 0.70300183724612,
      "epoch": 1.8710243184340516,
      "grad_norm": 0.16350367665290833,
      "learning_rate": 3.763609041003466e-05,
      "loss": 0.6944,
      "mean_token_accuracy": 0.7874953337013721,
      "num_tokens": 649050542.0,
      "step": 20336
    },
    {
      "entropy": 0.5552638713270426,
      "epoch": 1.8711163248054927,
      "grad_norm": 0.14821107685565948,
      "learning_rate": 3.763302358389303e-05,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8298650421202183,
      "num_tokens": 649082199.0,
      "step": 20337
    },
    {
      "entropy": 0.5577905466780066,
      "epoch": 1.871208331176934,
      "grad_norm": 0.1491255909204483,
      "learning_rate": 3.76299567577514e-05,
      "loss": 0.5594,
      "mean_token_accuracy": 0.8276870138943195,
      "num_tokens": 649114804.0,
      "step": 20338
    },
    {
      "entropy": 0.5374587094411254,
      "epoch": 1.8713003375483752,
      "grad_norm": 0.14685344696044922,
      "learning_rate": 3.762688993160978e-05,
      "loss": 0.526,
      "mean_token_accuracy": 0.837272122502327,
      "num_tokens": 649147313.0,
      "step": 20339
    },
    {
      "entropy": 0.6607564762234688,
      "epoch": 1.8713923439198163,
      "grad_norm": 0.16036559641361237,
      "learning_rate": 3.762382310546815e-05,
      "loss": 0.6573,
      "mean_token_accuracy": 0.802940770983696,
      "num_tokens": 649179251.0,
      "step": 20340
    },
    {
      "entropy": 0.6638198215514421,
      "epoch": 1.8714843502912577,
      "grad_norm": 0.1589352935552597,
      "learning_rate": 3.7620756279326526e-05,
      "loss": 0.6481,
      "mean_token_accuracy": 0.8031161427497864,
      "num_tokens": 649210562.0,
      "step": 20341
    },
    {
      "entropy": 0.657429775223136,
      "epoch": 1.8715763566626988,
      "grad_norm": 0.15710869431495667,
      "learning_rate": 3.76176894531849e-05,
      "loss": 0.6583,
      "mean_token_accuracy": 0.7950557172298431,
      "num_tokens": 649242884.0,
      "step": 20342
    },
    {
      "entropy": 0.6301285363733768,
      "epoch": 1.8716683630341402,
      "grad_norm": 0.1554500013589859,
      "learning_rate": 3.7614622627043276e-05,
      "loss": 0.6083,
      "mean_token_accuracy": 0.8102195747196674,
      "num_tokens": 649274466.0,
      "step": 20343
    },
    {
      "entropy": 0.5438408078625798,
      "epoch": 1.8717603694055813,
      "grad_norm": 0.1494918018579483,
      "learning_rate": 3.761155580090165e-05,
      "loss": 0.5341,
      "mean_token_accuracy": 0.8330769017338753,
      "num_tokens": 649306167.0,
      "step": 20344
    },
    {
      "entropy": 0.6347728529945016,
      "epoch": 1.8718523757770225,
      "grad_norm": 0.15302245318889618,
      "learning_rate": 3.760848897476002e-05,
      "loss": 0.6331,
      "mean_token_accuracy": 0.8049486130475998,
      "num_tokens": 649338141.0,
      "step": 20345
    },
    {
      "entropy": 0.5900555052794516,
      "epoch": 1.8719443821484638,
      "grad_norm": 0.14715631306171417,
      "learning_rate": 3.76054221486184e-05,
      "loss": 0.5699,
      "mean_token_accuracy": 0.8241811916232109,
      "num_tokens": 649369802.0,
      "step": 20346
    },
    {
      "entropy": 0.6400969121605158,
      "epoch": 1.872036388519905,
      "grad_norm": 0.15473461151123047,
      "learning_rate": 3.7602355322476774e-05,
      "loss": 0.6374,
      "mean_token_accuracy": 0.8055545836687088,
      "num_tokens": 649401912.0,
      "step": 20347
    },
    {
      "entropy": 0.5167219266295433,
      "epoch": 1.8721283948913463,
      "grad_norm": 0.1473911553621292,
      "learning_rate": 3.759928849633514e-05,
      "loss": 0.4992,
      "mean_token_accuracy": 0.8434219509363174,
      "num_tokens": 649433058.0,
      "step": 20348
    },
    {
      "entropy": 0.545892670750618,
      "epoch": 1.8722204012627874,
      "grad_norm": 0.15493984520435333,
      "learning_rate": 3.7596221670193523e-05,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8340590856969357,
      "num_tokens": 649465419.0,
      "step": 20349
    },
    {
      "entropy": 0.6906795804388821,
      "epoch": 1.8723124076342286,
      "grad_norm": 0.15621165931224823,
      "learning_rate": 3.759315484405189e-05,
      "loss": 0.6881,
      "mean_token_accuracy": 0.7885004878044128,
      "num_tokens": 649497872.0,
      "step": 20350
    },
    {
      "entropy": 0.471608467400074,
      "epoch": 1.87240441400567,
      "grad_norm": 0.15323898196220398,
      "learning_rate": 3.7590088017910266e-05,
      "loss": 0.4553,
      "mean_token_accuracy": 0.8586773574352264,
      "num_tokens": 649529890.0,
      "step": 20351
    },
    {
      "entropy": 0.5810965364798903,
      "epoch": 1.8724964203771113,
      "grad_norm": 0.1542435735464096,
      "learning_rate": 3.758702119176864e-05,
      "loss": 0.5715,
      "mean_token_accuracy": 0.8254704959690571,
      "num_tokens": 649561735.0,
      "step": 20352
    },
    {
      "entropy": 0.6388971810229123,
      "epoch": 1.8725884267485524,
      "grad_norm": 0.16505131125450134,
      "learning_rate": 3.7583954365627015e-05,
      "loss": 0.6346,
      "mean_token_accuracy": 0.8085073418915272,
      "num_tokens": 649593924.0,
      "step": 20353
    },
    {
      "entropy": 0.6468667723238468,
      "epoch": 1.8726804331199935,
      "grad_norm": 0.15681084990501404,
      "learning_rate": 3.758088753948539e-05,
      "loss": 0.6277,
      "mean_token_accuracy": 0.8079922646284103,
      "num_tokens": 649625883.0,
      "step": 20354
    },
    {
      "entropy": 0.4788303133100271,
      "epoch": 1.8727724394914347,
      "grad_norm": 0.14052191376686096,
      "learning_rate": 3.7577820713343765e-05,
      "loss": 0.4549,
      "mean_token_accuracy": 0.8605225272476673,
      "num_tokens": 649657577.0,
      "step": 20355
    },
    {
      "entropy": 0.6544817015528679,
      "epoch": 1.872864445862876,
      "grad_norm": 0.16408218443393707,
      "learning_rate": 3.757475388720213e-05,
      "loss": 0.6311,
      "mean_token_accuracy": 0.8075063116848469,
      "num_tokens": 649689807.0,
      "step": 20356
    },
    {
      "entropy": 0.5782691109925508,
      "epoch": 1.8729564522343174,
      "grad_norm": 0.1551266312599182,
      "learning_rate": 3.7571687061060514e-05,
      "loss": 0.5662,
      "mean_token_accuracy": 0.829191830009222,
      "num_tokens": 649721563.0,
      "step": 20357
    },
    {
      "entropy": 0.6030059594195336,
      "epoch": 1.8730484586057585,
      "grad_norm": 0.15814675390720367,
      "learning_rate": 3.756862023491888e-05,
      "loss": 0.6023,
      "mean_token_accuracy": 0.8141042292118073,
      "num_tokens": 649753336.0,
      "step": 20358
    },
    {
      "entropy": 0.5903878305107355,
      "epoch": 1.8731404649771997,
      "grad_norm": 0.15391215682029724,
      "learning_rate": 3.756555340877726e-05,
      "loss": 0.5717,
      "mean_token_accuracy": 0.82069506123662,
      "num_tokens": 649784667.0,
      "step": 20359
    },
    {
      "entropy": 0.6197503197472543,
      "epoch": 1.8732324713486408,
      "grad_norm": 0.15232840180397034,
      "learning_rate": 3.756248658263563e-05,
      "loss": 0.5994,
      "mean_token_accuracy": 0.8158120810985565,
      "num_tokens": 649816699.0,
      "step": 20360
    },
    {
      "entropy": 0.5796319227665663,
      "epoch": 1.8733244777200821,
      "grad_norm": 0.1467931568622589,
      "learning_rate": 3.7559419756494006e-05,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8277257457375526,
      "num_tokens": 649848749.0,
      "step": 20361
    },
    {
      "entropy": 0.6909361183643341,
      "epoch": 1.8734164840915235,
      "grad_norm": 0.1629001796245575,
      "learning_rate": 3.755635293035238e-05,
      "loss": 0.6866,
      "mean_token_accuracy": 0.7891135178506374,
      "num_tokens": 649880617.0,
      "step": 20362
    },
    {
      "entropy": 0.5950709655880928,
      "epoch": 1.8735084904629646,
      "grad_norm": 0.1557748168706894,
      "learning_rate": 3.7553286104210755e-05,
      "loss": 0.5881,
      "mean_token_accuracy": 0.819590050727129,
      "num_tokens": 649913236.0,
      "step": 20363
    },
    {
      "entropy": 0.7166619654744864,
      "epoch": 1.8736004968344058,
      "grad_norm": 0.162390798330307,
      "learning_rate": 3.755021927806913e-05,
      "loss": 0.7132,
      "mean_token_accuracy": 0.7831631153821945,
      "num_tokens": 649945387.0,
      "step": 20364
    },
    {
      "entropy": 0.5488041620701551,
      "epoch": 1.873692503205847,
      "grad_norm": 0.15180733799934387,
      "learning_rate": 3.7547152451927504e-05,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8291279897093773,
      "num_tokens": 649977397.0,
      "step": 20365
    },
    {
      "entropy": 0.5921749006956816,
      "epoch": 1.8737845095772883,
      "grad_norm": 0.15218143165111542,
      "learning_rate": 3.754408562578587e-05,
      "loss": 0.5834,
      "mean_token_accuracy": 0.8204645328223705,
      "num_tokens": 650009505.0,
      "step": 20366
    },
    {
      "entropy": 0.6577141284942627,
      "epoch": 1.8738765159487296,
      "grad_norm": 0.1677580326795578,
      "learning_rate": 3.7541018799644254e-05,
      "loss": 0.6558,
      "mean_token_accuracy": 0.7967012077569962,
      "num_tokens": 650040444.0,
      "step": 20367
    },
    {
      "entropy": 0.5725044412538409,
      "epoch": 1.8739685223201707,
      "grad_norm": 0.15016211569309235,
      "learning_rate": 3.753795197350262e-05,
      "loss": 0.5754,
      "mean_token_accuracy": 0.8207443058490753,
      "num_tokens": 650071507.0,
      "step": 20368
    },
    {
      "entropy": 0.6766606047749519,
      "epoch": 1.8740605286916119,
      "grad_norm": 0.15891297161579132,
      "learning_rate": 3.7534885147360996e-05,
      "loss": 0.6691,
      "mean_token_accuracy": 0.7941320948302746,
      "num_tokens": 650103990.0,
      "step": 20369
    },
    {
      "entropy": 0.5884555950760841,
      "epoch": 1.874152535063053,
      "grad_norm": 0.16276805102825165,
      "learning_rate": 3.753181832121937e-05,
      "loss": 0.5826,
      "mean_token_accuracy": 0.8219552785158157,
      "num_tokens": 650135243.0,
      "step": 20370
    },
    {
      "entropy": 0.62554153916426,
      "epoch": 1.8742445414344944,
      "grad_norm": 0.15456685423851013,
      "learning_rate": 3.7528751495077745e-05,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8071788810193539,
      "num_tokens": 650167875.0,
      "step": 20371
    },
    {
      "entropy": 0.597050353884697,
      "epoch": 1.8743365478059357,
      "grad_norm": 0.15204130113124847,
      "learning_rate": 3.752568466893612e-05,
      "loss": 0.5761,
      "mean_token_accuracy": 0.8225988671183586,
      "num_tokens": 650200112.0,
      "step": 20372
    },
    {
      "entropy": 0.561307605355978,
      "epoch": 1.8744285541773769,
      "grad_norm": 0.15098202228546143,
      "learning_rate": 3.7522617842794495e-05,
      "loss": 0.5468,
      "mean_token_accuracy": 0.8303666263818741,
      "num_tokens": 650232032.0,
      "step": 20373
    },
    {
      "entropy": 0.6115573327988386,
      "epoch": 1.874520560548818,
      "grad_norm": 0.14729700982570648,
      "learning_rate": 3.751955101665286e-05,
      "loss": 0.592,
      "mean_token_accuracy": 0.8169667050242424,
      "num_tokens": 650264496.0,
      "step": 20374
    },
    {
      "entropy": 0.6824539639055729,
      "epoch": 1.8746125669202591,
      "grad_norm": 0.15506114065647125,
      "learning_rate": 3.7516484190511244e-05,
      "loss": 0.6759,
      "mean_token_accuracy": 0.7933532148599625,
      "num_tokens": 650297063.0,
      "step": 20375
    },
    {
      "entropy": 0.5611634589731693,
      "epoch": 1.8747045732917005,
      "grad_norm": 0.1548604965209961,
      "learning_rate": 3.751341736436961e-05,
      "loss": 0.5547,
      "mean_token_accuracy": 0.8299667574465275,
      "num_tokens": 650329476.0,
      "step": 20376
    },
    {
      "entropy": 0.6652100021019578,
      "epoch": 1.8747965796631418,
      "grad_norm": 0.15925933420658112,
      "learning_rate": 3.7510350538227987e-05,
      "loss": 0.6489,
      "mean_token_accuracy": 0.7963706292212009,
      "num_tokens": 650361216.0,
      "step": 20377
    },
    {
      "entropy": 0.642666669562459,
      "epoch": 1.874888586034583,
      "grad_norm": 0.16292038559913635,
      "learning_rate": 3.750728371208637e-05,
      "loss": 0.6368,
      "mean_token_accuracy": 0.8060494959354401,
      "num_tokens": 650393104.0,
      "step": 20378
    },
    {
      "entropy": 0.5764529104344547,
      "epoch": 1.874980592406024,
      "grad_norm": 0.15200193226337433,
      "learning_rate": 3.7504216885944736e-05,
      "loss": 0.5782,
      "mean_token_accuracy": 0.8221861086785793,
      "num_tokens": 650424632.0,
      "step": 20379
    },
    {
      "entropy": 0.6831139903515577,
      "epoch": 1.8750725987774652,
      "grad_norm": 0.16547225415706635,
      "learning_rate": 3.750115005980312e-05,
      "loss": 0.6571,
      "mean_token_accuracy": 0.7964161448180676,
      "num_tokens": 650456028.0,
      "step": 20380
    },
    {
      "entropy": 0.6274551497772336,
      "epoch": 1.8751646051489066,
      "grad_norm": 0.15084737539291382,
      "learning_rate": 3.7498083233661485e-05,
      "loss": 0.6142,
      "mean_token_accuracy": 0.8088030777871609,
      "num_tokens": 650487940.0,
      "step": 20381
    },
    {
      "entropy": 0.6494339369237423,
      "epoch": 1.875256611520348,
      "grad_norm": 0.15484492480754852,
      "learning_rate": 3.749501640751986e-05,
      "loss": 0.6339,
      "mean_token_accuracy": 0.8068530932068825,
      "num_tokens": 650520137.0,
      "step": 20382
    },
    {
      "entropy": 0.4879958052188158,
      "epoch": 1.875348617891789,
      "grad_norm": 0.14090724289417267,
      "learning_rate": 3.7491949581378234e-05,
      "loss": 0.473,
      "mean_token_accuracy": 0.8547460287809372,
      "num_tokens": 650552481.0,
      "step": 20383
    },
    {
      "entropy": 0.591544290073216,
      "epoch": 1.8754406242632302,
      "grad_norm": 0.1524932086467743,
      "learning_rate": 3.748888275523661e-05,
      "loss": 0.5887,
      "mean_token_accuracy": 0.8186671808362007,
      "num_tokens": 650584683.0,
      "step": 20384
    },
    {
      "entropy": 0.5113988583907485,
      "epoch": 1.8755326306346713,
      "grad_norm": 0.1467595249414444,
      "learning_rate": 3.7485815929094984e-05,
      "loss": 0.4989,
      "mean_token_accuracy": 0.8442582152783871,
      "num_tokens": 650617167.0,
      "step": 20385
    },
    {
      "entropy": 0.6536541022360325,
      "epoch": 1.8756246370061127,
      "grad_norm": 0.15598855912685394,
      "learning_rate": 3.748274910295336e-05,
      "loss": 0.631,
      "mean_token_accuracy": 0.8079562224447727,
      "num_tokens": 650649376.0,
      "step": 20386
    },
    {
      "entropy": 0.5755582619458437,
      "epoch": 1.875716643377554,
      "grad_norm": 0.1522272378206253,
      "learning_rate": 3.7479682276811726e-05,
      "loss": 0.5699,
      "mean_token_accuracy": 0.823286022990942,
      "num_tokens": 650681179.0,
      "step": 20387
    },
    {
      "entropy": 0.5585754858329892,
      "epoch": 1.8758086497489952,
      "grad_norm": 0.15498991310596466,
      "learning_rate": 3.747661545067011e-05,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8316647596657276,
      "num_tokens": 650713082.0,
      "step": 20388
    },
    {
      "entropy": 0.6285503040999174,
      "epoch": 1.8759006561204363,
      "grad_norm": 0.1629282832145691,
      "learning_rate": 3.7473548624528476e-05,
      "loss": 0.6152,
      "mean_token_accuracy": 0.8050252310931683,
      "num_tokens": 650744323.0,
      "step": 20389
    },
    {
      "entropy": 0.5595280658453703,
      "epoch": 1.8759926624918775,
      "grad_norm": 0.14696860313415527,
      "learning_rate": 3.747048179838685e-05,
      "loss": 0.5444,
      "mean_token_accuracy": 0.830630861222744,
      "num_tokens": 650776201.0,
      "step": 20390
    },
    {
      "entropy": 0.6514391675591469,
      "epoch": 1.8760846688633188,
      "grad_norm": 0.16085588932037354,
      "learning_rate": 3.7467414972245225e-05,
      "loss": 0.6401,
      "mean_token_accuracy": 0.8027487881481647,
      "num_tokens": 650807607.0,
      "step": 20391
    },
    {
      "entropy": 0.6412530038505793,
      "epoch": 1.8761766752347602,
      "grad_norm": 0.15129293501377106,
      "learning_rate": 3.74643481461036e-05,
      "loss": 0.6364,
      "mean_token_accuracy": 0.8041367642581463,
      "num_tokens": 650840246.0,
      "step": 20392
    },
    {
      "entropy": 0.6351837832480669,
      "epoch": 1.8762686816062013,
      "grad_norm": 0.1510544866323471,
      "learning_rate": 3.7461281319961974e-05,
      "loss": 0.6253,
      "mean_token_accuracy": 0.8067305870354176,
      "num_tokens": 650872788.0,
      "step": 20393
    },
    {
      "entropy": 0.5286234617233276,
      "epoch": 1.8763606879776424,
      "grad_norm": 0.14436407387256622,
      "learning_rate": 3.745821449382035e-05,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8448861911892891,
      "num_tokens": 650904562.0,
      "step": 20394
    },
    {
      "entropy": 0.562138743698597,
      "epoch": 1.8764526943490836,
      "grad_norm": 0.1500539481639862,
      "learning_rate": 3.745514766767872e-05,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8284884132444859,
      "num_tokens": 650936378.0,
      "step": 20395
    },
    {
      "entropy": 0.5583719033747911,
      "epoch": 1.876544700720525,
      "grad_norm": 0.14698946475982666,
      "learning_rate": 3.74520808415371e-05,
      "loss": 0.5493,
      "mean_token_accuracy": 0.8266566656529903,
      "num_tokens": 650968228.0,
      "step": 20396
    },
    {
      "entropy": 0.5782865267246962,
      "epoch": 1.8766367070919663,
      "grad_norm": 0.14777891337871552,
      "learning_rate": 3.7449014015395466e-05,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8264338932931423,
      "num_tokens": 651000339.0,
      "step": 20397
    },
    {
      "entropy": 0.7111164554953575,
      "epoch": 1.8767287134634074,
      "grad_norm": 0.16664792597293854,
      "learning_rate": 3.744594718925384e-05,
      "loss": 0.7108,
      "mean_token_accuracy": 0.7844281792640686,
      "num_tokens": 651032160.0,
      "step": 20398
    },
    {
      "entropy": 0.5988824097439647,
      "epoch": 1.8768207198348485,
      "grad_norm": 0.15780000388622284,
      "learning_rate": 3.7442880363112215e-05,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8205706290900707,
      "num_tokens": 651063829.0,
      "step": 20399
    },
    {
      "entropy": 0.5801811460405588,
      "epoch": 1.8769127262062897,
      "grad_norm": 0.1492239236831665,
      "learning_rate": 3.743981353697059e-05,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8240360952913761,
      "num_tokens": 651096063.0,
      "step": 20400
    },
    {
      "entropy": 0.5475428926292807,
      "epoch": 1.877004732577731,
      "grad_norm": 0.14548006653785706,
      "learning_rate": 3.7436746710828965e-05,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8321353681385517,
      "num_tokens": 651128336.0,
      "step": 20401
    },
    {
      "entropy": 0.6919078063219786,
      "epoch": 1.8770967389491724,
      "grad_norm": 0.16013756394386292,
      "learning_rate": 3.743367988468734e-05,
      "loss": 0.6925,
      "mean_token_accuracy": 0.7895747348666191,
      "num_tokens": 651160832.0,
      "step": 20402
    },
    {
      "entropy": 0.6008948013186455,
      "epoch": 1.8771887453206135,
      "grad_norm": 0.15780571103096008,
      "learning_rate": 3.7430613058545714e-05,
      "loss": 0.5977,
      "mean_token_accuracy": 0.8202020116150379,
      "num_tokens": 651192633.0,
      "step": 20403
    },
    {
      "entropy": 0.6031614579260349,
      "epoch": 1.8772807516920547,
      "grad_norm": 0.15230219066143036,
      "learning_rate": 3.742754623240409e-05,
      "loss": 0.6012,
      "mean_token_accuracy": 0.8166682049632072,
      "num_tokens": 651224097.0,
      "step": 20404
    },
    {
      "entropy": 0.6345590362325311,
      "epoch": 1.8773727580634958,
      "grad_norm": 0.15736564993858337,
      "learning_rate": 3.7424479406262456e-05,
      "loss": 0.6295,
      "mean_token_accuracy": 0.8074197508394718,
      "num_tokens": 651256002.0,
      "step": 20405
    },
    {
      "entropy": 0.5459285862743855,
      "epoch": 1.8774647644349372,
      "grad_norm": 0.1513427346944809,
      "learning_rate": 3.742141258012084e-05,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8344030492007732,
      "num_tokens": 651288158.0,
      "step": 20406
    },
    {
      "entropy": 0.63471017498523,
      "epoch": 1.8775567708063785,
      "grad_norm": 0.15537866950035095,
      "learning_rate": 3.7418345753979206e-05,
      "loss": 0.631,
      "mean_token_accuracy": 0.8109998367726803,
      "num_tokens": 651319995.0,
      "step": 20407
    },
    {
      "entropy": 0.7146575069054961,
      "epoch": 1.8776487771778196,
      "grad_norm": 0.1693895310163498,
      "learning_rate": 3.741527892783758e-05,
      "loss": 0.7008,
      "mean_token_accuracy": 0.7854869365692139,
      "num_tokens": 651351931.0,
      "step": 20408
    },
    {
      "entropy": 0.4759451234713197,
      "epoch": 1.8777407835492608,
      "grad_norm": 0.14191670715808868,
      "learning_rate": 3.741221210169596e-05,
      "loss": 0.469,
      "mean_token_accuracy": 0.8577514812350273,
      "num_tokens": 651384069.0,
      "step": 20409
    },
    {
      "entropy": 0.6790676107630134,
      "epoch": 1.877832789920702,
      "grad_norm": 0.15672476589679718,
      "learning_rate": 3.740914527555433e-05,
      "loss": 0.6723,
      "mean_token_accuracy": 0.797436784952879,
      "num_tokens": 651415372.0,
      "step": 20410
    },
    {
      "entropy": 0.6435285042971373,
      "epoch": 1.8779247962921433,
      "grad_norm": 0.15583251416683197,
      "learning_rate": 3.7406078449412704e-05,
      "loss": 0.6255,
      "mean_token_accuracy": 0.806840181350708,
      "num_tokens": 651447201.0,
      "step": 20411
    },
    {
      "entropy": 0.5403334591537714,
      "epoch": 1.8780168026635846,
      "grad_norm": 0.14788533747196198,
      "learning_rate": 3.740301162327108e-05,
      "loss": 0.5214,
      "mean_token_accuracy": 0.8379901573061943,
      "num_tokens": 651478769.0,
      "step": 20412
    },
    {
      "entropy": 0.7045591324567795,
      "epoch": 1.8781088090350258,
      "grad_norm": 0.1656073033809662,
      "learning_rate": 3.7399944797129453e-05,
      "loss": 0.697,
      "mean_token_accuracy": 0.7885672338306904,
      "num_tokens": 651510532.0,
      "step": 20413
    },
    {
      "entropy": 0.6139884199947119,
      "epoch": 1.8782008154064669,
      "grad_norm": 0.15217046439647675,
      "learning_rate": 3.739687797098783e-05,
      "loss": 0.6099,
      "mean_token_accuracy": 0.810219332575798,
      "num_tokens": 651542684.0,
      "step": 20414
    },
    {
      "entropy": 0.6836214736104012,
      "epoch": 1.878292821777908,
      "grad_norm": 0.16192656755447388,
      "learning_rate": 3.73938111448462e-05,
      "loss": 0.6661,
      "mean_token_accuracy": 0.7932244949042797,
      "num_tokens": 651574578.0,
      "step": 20415
    },
    {
      "entropy": 0.5654850928112864,
      "epoch": 1.8783848281493494,
      "grad_norm": 0.14486312866210938,
      "learning_rate": 3.739074431870457e-05,
      "loss": 0.5456,
      "mean_token_accuracy": 0.8332360908389091,
      "num_tokens": 651606672.0,
      "step": 20416
    },
    {
      "entropy": 0.648032745346427,
      "epoch": 1.8784768345207907,
      "grad_norm": 0.1489066630601883,
      "learning_rate": 3.738767749256295e-05,
      "loss": 0.6229,
      "mean_token_accuracy": 0.806297592818737,
      "num_tokens": 651638701.0,
      "step": 20417
    },
    {
      "entropy": 0.7229907922446728,
      "epoch": 1.8785688408922319,
      "grad_norm": 0.15668648481369019,
      "learning_rate": 3.738461066642132e-05,
      "loss": 0.6909,
      "mean_token_accuracy": 0.7918341495096684,
      "num_tokens": 651670265.0,
      "step": 20418
    },
    {
      "entropy": 0.6361892875283957,
      "epoch": 1.878660847263673,
      "grad_norm": 0.1473292112350464,
      "learning_rate": 3.73815438402797e-05,
      "loss": 0.6155,
      "mean_token_accuracy": 0.8085280582308769,
      "num_tokens": 651702425.0,
      "step": 20419
    },
    {
      "entropy": 0.6766858492046595,
      "epoch": 1.8787528536351141,
      "grad_norm": 0.16658459603786469,
      "learning_rate": 3.737847701413807e-05,
      "loss": 0.6568,
      "mean_token_accuracy": 0.7990319319069386,
      "num_tokens": 651732793.0,
      "step": 20420
    },
    {
      "entropy": 0.6211369968950748,
      "epoch": 1.8788448600065555,
      "grad_norm": 0.1503210812807083,
      "learning_rate": 3.7375410187996444e-05,
      "loss": 0.6079,
      "mean_token_accuracy": 0.8132143244147301,
      "num_tokens": 651765021.0,
      "step": 20421
    },
    {
      "entropy": 0.5417829172220081,
      "epoch": 1.8789368663779968,
      "grad_norm": 0.1473347693681717,
      "learning_rate": 3.737234336185482e-05,
      "loss": 0.5328,
      "mean_token_accuracy": 0.8356819674372673,
      "num_tokens": 651796289.0,
      "step": 20422
    },
    {
      "entropy": 0.398537565022707,
      "epoch": 1.879028872749438,
      "grad_norm": 0.129128560423851,
      "learning_rate": 3.736927653571319e-05,
      "loss": 0.384,
      "mean_token_accuracy": 0.8776400871574879,
      "num_tokens": 651827966.0,
      "step": 20423
    },
    {
      "entropy": 0.5876167807728052,
      "epoch": 1.879120879120879,
      "grad_norm": 0.152541846036911,
      "learning_rate": 3.736620970957157e-05,
      "loss": 0.5949,
      "mean_token_accuracy": 0.8161150887608528,
      "num_tokens": 651860353.0,
      "step": 20424
    },
    {
      "entropy": 0.6474423818290234,
      "epoch": 1.8792128854923202,
      "grad_norm": 0.16082623600959778,
      "learning_rate": 3.736314288342994e-05,
      "loss": 0.6439,
      "mean_token_accuracy": 0.7971453107893467,
      "num_tokens": 651891422.0,
      "step": 20425
    },
    {
      "entropy": 0.5700470646843314,
      "epoch": 1.8793048918637616,
      "grad_norm": 0.15131287276744843,
      "learning_rate": 3.736007605728831e-05,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8247003108263016,
      "num_tokens": 651923091.0,
      "step": 20426
    },
    {
      "entropy": 0.550682857632637,
      "epoch": 1.879396898235203,
      "grad_norm": 0.152308851480484,
      "learning_rate": 3.735700923114669e-05,
      "loss": 0.5587,
      "mean_token_accuracy": 0.8268433809280396,
      "num_tokens": 651955158.0,
      "step": 20427
    },
    {
      "entropy": 0.6012499807402492,
      "epoch": 1.879488904606644,
      "grad_norm": 0.16148211061954498,
      "learning_rate": 3.735394240500506e-05,
      "loss": 0.6016,
      "mean_token_accuracy": 0.8135002180933952,
      "num_tokens": 651986870.0,
      "step": 20428
    },
    {
      "entropy": 0.504382009152323,
      "epoch": 1.8795809109780852,
      "grad_norm": 0.1507500410079956,
      "learning_rate": 3.7350875578863434e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8444369807839394,
      "num_tokens": 652019638.0,
      "step": 20429
    },
    {
      "entropy": 0.5910447612404823,
      "epoch": 1.8796729173495264,
      "grad_norm": 0.15260113775730133,
      "learning_rate": 3.734780875272181e-05,
      "loss": 0.5818,
      "mean_token_accuracy": 0.8205129392445087,
      "num_tokens": 652051588.0,
      "step": 20430
    },
    {
      "entropy": 0.6465833324473351,
      "epoch": 1.8797649237209677,
      "grad_norm": 0.15735460817813873,
      "learning_rate": 3.7344741926580184e-05,
      "loss": 0.6383,
      "mean_token_accuracy": 0.8053355701267719,
      "num_tokens": 652083326.0,
      "step": 20431
    },
    {
      "entropy": 0.6073672166094184,
      "epoch": 1.879856930092409,
      "grad_norm": 0.1531253457069397,
      "learning_rate": 3.734167510043856e-05,
      "loss": 0.5983,
      "mean_token_accuracy": 0.8157905898988247,
      "num_tokens": 652115423.0,
      "step": 20432
    },
    {
      "entropy": 0.6540533974766731,
      "epoch": 1.8799489364638502,
      "grad_norm": 0.15718448162078857,
      "learning_rate": 3.733860827429693e-05,
      "loss": 0.6441,
      "mean_token_accuracy": 0.8025729432702065,
      "num_tokens": 652147402.0,
      "step": 20433
    },
    {
      "entropy": 0.620308228302747,
      "epoch": 1.8800409428352913,
      "grad_norm": 0.1538448929786682,
      "learning_rate": 3.73355414481553e-05,
      "loss": 0.6149,
      "mean_token_accuracy": 0.8092070892453194,
      "num_tokens": 652179340.0,
      "step": 20434
    },
    {
      "entropy": 0.6764511931687593,
      "epoch": 1.8801329492067325,
      "grad_norm": 0.16101180016994476,
      "learning_rate": 3.733247462201368e-05,
      "loss": 0.6635,
      "mean_token_accuracy": 0.7965671829879284,
      "num_tokens": 652210737.0,
      "step": 20435
    },
    {
      "entropy": 0.5727704521268606,
      "epoch": 1.8802249555781738,
      "grad_norm": 0.14690545201301575,
      "learning_rate": 3.732940779587205e-05,
      "loss": 0.564,
      "mean_token_accuracy": 0.826698325574398,
      "num_tokens": 652243158.0,
      "step": 20436
    },
    {
      "entropy": 0.6364552024751902,
      "epoch": 1.8803169619496152,
      "grad_norm": 0.15851859748363495,
      "learning_rate": 3.7326340969730425e-05,
      "loss": 0.6249,
      "mean_token_accuracy": 0.8128317818045616,
      "num_tokens": 652274824.0,
      "step": 20437
    },
    {
      "entropy": 0.5364432772621512,
      "epoch": 1.8804089683210563,
      "grad_norm": 0.14337827265262604,
      "learning_rate": 3.7323274143588806e-05,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8364838100969791,
      "num_tokens": 652306642.0,
      "step": 20438
    },
    {
      "entropy": 0.6011263988912106,
      "epoch": 1.8805009746924974,
      "grad_norm": 0.15730087459087372,
      "learning_rate": 3.7320207317447174e-05,
      "loss": 0.5832,
      "mean_token_accuracy": 0.8187817484140396,
      "num_tokens": 652337883.0,
      "step": 20439
    },
    {
      "entropy": 0.7053113970905542,
      "epoch": 1.8805929810639386,
      "grad_norm": 0.1611095666885376,
      "learning_rate": 3.7317140491305555e-05,
      "loss": 0.698,
      "mean_token_accuracy": 0.7856894060969353,
      "num_tokens": 652370334.0,
      "step": 20440
    },
    {
      "entropy": 0.5902689406648278,
      "epoch": 1.88068498743538,
      "grad_norm": 0.1517544835805893,
      "learning_rate": 3.731407366516392e-05,
      "loss": 0.5736,
      "mean_token_accuracy": 0.820686612278223,
      "num_tokens": 652401915.0,
      "step": 20441
    },
    {
      "entropy": 0.7567095002159476,
      "epoch": 1.8807769938068213,
      "grad_norm": 0.16339601576328278,
      "learning_rate": 3.73110068390223e-05,
      "loss": 0.7491,
      "mean_token_accuracy": 0.7764552794396877,
      "num_tokens": 652433755.0,
      "step": 20442
    },
    {
      "entropy": 0.599516352172941,
      "epoch": 1.8808690001782624,
      "grad_norm": 0.15037858486175537,
      "learning_rate": 3.730794001288067e-05,
      "loss": 0.5847,
      "mean_token_accuracy": 0.8206273540854454,
      "num_tokens": 652464892.0,
      "step": 20443
    },
    {
      "entropy": 0.5397110478952527,
      "epoch": 1.8809610065497036,
      "grad_norm": 0.1461339294910431,
      "learning_rate": 3.730487318673905e-05,
      "loss": 0.5279,
      "mean_token_accuracy": 0.8353141136467457,
      "num_tokens": 652497340.0,
      "step": 20444
    },
    {
      "entropy": 0.5996588296256959,
      "epoch": 1.8810530129211447,
      "grad_norm": 0.15627048909664154,
      "learning_rate": 3.730180636059742e-05,
      "loss": 0.58,
      "mean_token_accuracy": 0.8208685405552387,
      "num_tokens": 652529426.0,
      "step": 20445
    },
    {
      "entropy": 0.5131772984750569,
      "epoch": 1.881145019292586,
      "grad_norm": 0.14789648354053497,
      "learning_rate": 3.7298739534455797e-05,
      "loss": 0.4989,
      "mean_token_accuracy": 0.8414336256682873,
      "num_tokens": 652561154.0,
      "step": 20446
    },
    {
      "entropy": 0.517800766043365,
      "epoch": 1.8812370256640274,
      "grad_norm": 0.14582054316997528,
      "learning_rate": 3.7295672708314164e-05,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8433656208217144,
      "num_tokens": 652592774.0,
      "step": 20447
    },
    {
      "entropy": 0.5190470251254737,
      "epoch": 1.8813290320354685,
      "grad_norm": 0.15433372557163239,
      "learning_rate": 3.7292605882172546e-05,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8402065858244896,
      "num_tokens": 652624863.0,
      "step": 20448
    },
    {
      "entropy": 0.6621332764625549,
      "epoch": 1.8814210384069097,
      "grad_norm": 0.1581066995859146,
      "learning_rate": 3.7289539056030914e-05,
      "loss": 0.6649,
      "mean_token_accuracy": 0.7945593222975731,
      "num_tokens": 652657206.0,
      "step": 20449
    },
    {
      "entropy": 0.5861445451155305,
      "epoch": 1.8815130447783508,
      "grad_norm": 0.1530643105506897,
      "learning_rate": 3.728647222988929e-05,
      "loss": 0.5797,
      "mean_token_accuracy": 0.8211613558232784,
      "num_tokens": 652689250.0,
      "step": 20450
    },
    {
      "entropy": 0.6673767319880426,
      "epoch": 1.8816050511497922,
      "grad_norm": 0.1558447629213333,
      "learning_rate": 3.728340540374766e-05,
      "loss": 0.6524,
      "mean_token_accuracy": 0.8024502769112587,
      "num_tokens": 652721072.0,
      "step": 20451
    },
    {
      "entropy": 0.6738601746037602,
      "epoch": 1.8816970575212335,
      "grad_norm": 0.15788327157497406,
      "learning_rate": 3.728033857760604e-05,
      "loss": 0.6692,
      "mean_token_accuracy": 0.7968647293746471,
      "num_tokens": 652753191.0,
      "step": 20452
    },
    {
      "entropy": 0.5640986964572221,
      "epoch": 1.8817890638926746,
      "grad_norm": 0.1498832255601883,
      "learning_rate": 3.727727175146441e-05,
      "loss": 0.5619,
      "mean_token_accuracy": 0.8271548338234425,
      "num_tokens": 652785061.0,
      "step": 20453
    },
    {
      "entropy": 0.5340516651049256,
      "epoch": 1.8818810702641158,
      "grad_norm": 0.14415615797042847,
      "learning_rate": 3.727420492532279e-05,
      "loss": 0.5195,
      "mean_token_accuracy": 0.8390040658414364,
      "num_tokens": 652817476.0,
      "step": 20454
    },
    {
      "entropy": 0.6219555214047432,
      "epoch": 1.881973076635557,
      "grad_norm": 0.15428289771080017,
      "learning_rate": 3.7271138099181155e-05,
      "loss": 0.6082,
      "mean_token_accuracy": 0.808771688491106,
      "num_tokens": 652848969.0,
      "step": 20455
    },
    {
      "entropy": 0.6004019882529974,
      "epoch": 1.8820650830069983,
      "grad_norm": 0.1520041823387146,
      "learning_rate": 3.7268071273039536e-05,
      "loss": 0.5878,
      "mean_token_accuracy": 0.820387028157711,
      "num_tokens": 652880815.0,
      "step": 20456
    },
    {
      "entropy": 0.6638024244457483,
      "epoch": 1.8821570893784396,
      "grad_norm": 0.15198981761932373,
      "learning_rate": 3.7265004446897904e-05,
      "loss": 0.6549,
      "mean_token_accuracy": 0.7993559911847115,
      "num_tokens": 652913103.0,
      "step": 20457
    },
    {
      "entropy": 0.5866254363209009,
      "epoch": 1.8822490957498808,
      "grad_norm": 0.15642604231834412,
      "learning_rate": 3.726193762075628e-05,
      "loss": 0.5813,
      "mean_token_accuracy": 0.821444358676672,
      "num_tokens": 652944213.0,
      "step": 20458
    },
    {
      "entropy": 0.6805712208151817,
      "epoch": 1.8823411021213219,
      "grad_norm": 0.1643347442150116,
      "learning_rate": 3.7258870794614653e-05,
      "loss": 0.67,
      "mean_token_accuracy": 0.7937572300434113,
      "num_tokens": 652976232.0,
      "step": 20459
    },
    {
      "entropy": 0.6250235480256379,
      "epoch": 1.882433108492763,
      "grad_norm": 0.1539846956729889,
      "learning_rate": 3.725580396847303e-05,
      "loss": 0.6166,
      "mean_token_accuracy": 0.8109009079635143,
      "num_tokens": 653008723.0,
      "step": 20460
    },
    {
      "entropy": 0.6015912462025881,
      "epoch": 1.8825251148642044,
      "grad_norm": 0.1553526669740677,
      "learning_rate": 3.72527371423314e-05,
      "loss": 0.6014,
      "mean_token_accuracy": 0.815595980733633,
      "num_tokens": 653040281.0,
      "step": 20461
    },
    {
      "entropy": 0.5701332376338542,
      "epoch": 1.8826171212356457,
      "grad_norm": 0.14101001620292664,
      "learning_rate": 3.724967031618978e-05,
      "loss": 0.5538,
      "mean_token_accuracy": 0.8290249705314636,
      "num_tokens": 653072318.0,
      "step": 20462
    },
    {
      "entropy": 0.6114246230572462,
      "epoch": 1.8827091276070869,
      "grad_norm": 0.15925225615501404,
      "learning_rate": 3.724660349004815e-05,
      "loss": 0.6078,
      "mean_token_accuracy": 0.8148197866976261,
      "num_tokens": 653103945.0,
      "step": 20463
    },
    {
      "entropy": 0.6496864855289459,
      "epoch": 1.882801133978528,
      "grad_norm": 0.1632925271987915,
      "learning_rate": 3.724353666390653e-05,
      "loss": 0.6342,
      "mean_token_accuracy": 0.8064769878983498,
      "num_tokens": 653135597.0,
      "step": 20464
    },
    {
      "entropy": 0.5273113148286939,
      "epoch": 1.8828931403499691,
      "grad_norm": 0.1464960277080536,
      "learning_rate": 3.7240469837764895e-05,
      "loss": 0.5106,
      "mean_token_accuracy": 0.840367991477251,
      "num_tokens": 653167751.0,
      "step": 20465
    },
    {
      "entropy": 0.5031407373026013,
      "epoch": 1.8829851467214105,
      "grad_norm": 0.1458062082529068,
      "learning_rate": 3.7237403011623276e-05,
      "loss": 0.4904,
      "mean_token_accuracy": 0.8480575308203697,
      "num_tokens": 653200292.0,
      "step": 20466
    },
    {
      "entropy": 0.49445758666843176,
      "epoch": 1.8830771530928518,
      "grad_norm": 0.14746254682540894,
      "learning_rate": 3.7234336185481644e-05,
      "loss": 0.4792,
      "mean_token_accuracy": 0.8474000915884972,
      "num_tokens": 653232311.0,
      "step": 20467
    },
    {
      "entropy": 0.5882583409547806,
      "epoch": 1.883169159464293,
      "grad_norm": 0.15392711758613586,
      "learning_rate": 3.723126935934002e-05,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8223315328359604,
      "num_tokens": 653263700.0,
      "step": 20468
    },
    {
      "entropy": 0.5660631819628179,
      "epoch": 1.883261165835734,
      "grad_norm": 0.1420300453901291,
      "learning_rate": 3.72282025331984e-05,
      "loss": 0.5536,
      "mean_token_accuracy": 0.828773744404316,
      "num_tokens": 653296009.0,
      "step": 20469
    },
    {
      "entropy": 0.6667218329384923,
      "epoch": 1.8833531722071752,
      "grad_norm": 0.15561464428901672,
      "learning_rate": 3.722513570705677e-05,
      "loss": 0.6623,
      "mean_token_accuracy": 0.8018937222659588,
      "num_tokens": 653328204.0,
      "step": 20470
    },
    {
      "entropy": 0.5170070230960846,
      "epoch": 1.8834451785786166,
      "grad_norm": 0.14692236483097076,
      "learning_rate": 3.722206888091514e-05,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8451326936483383,
      "num_tokens": 653360536.0,
      "step": 20471
    },
    {
      "entropy": 0.5498584350571036,
      "epoch": 1.883537184950058,
      "grad_norm": 0.14924567937850952,
      "learning_rate": 3.721900205477352e-05,
      "loss": 0.5367,
      "mean_token_accuracy": 0.8356159403920174,
      "num_tokens": 653392413.0,
      "step": 20472
    },
    {
      "entropy": 0.603764409199357,
      "epoch": 1.883629191321499,
      "grad_norm": 0.147426038980484,
      "learning_rate": 3.721593522863189e-05,
      "loss": 0.5891,
      "mean_token_accuracy": 0.8171705305576324,
      "num_tokens": 653424478.0,
      "step": 20473
    },
    {
      "entropy": 0.6513115940615535,
      "epoch": 1.8837211976929402,
      "grad_norm": 0.16280677914619446,
      "learning_rate": 3.7212868402490266e-05,
      "loss": 0.6461,
      "mean_token_accuracy": 0.7976936511695385,
      "num_tokens": 653456279.0,
      "step": 20474
    },
    {
      "entropy": 0.6143420580774546,
      "epoch": 1.8838132040643814,
      "grad_norm": 0.1573072224855423,
      "learning_rate": 3.720980157634864e-05,
      "loss": 0.6069,
      "mean_token_accuracy": 0.8182207122445107,
      "num_tokens": 653488585.0,
      "step": 20475
    },
    {
      "entropy": 0.6049470212310553,
      "epoch": 1.8839052104358227,
      "grad_norm": 0.15194042026996613,
      "learning_rate": 3.720673475020701e-05,
      "loss": 0.5833,
      "mean_token_accuracy": 0.8173979446291924,
      "num_tokens": 653519871.0,
      "step": 20476
    },
    {
      "entropy": 0.5460547776892781,
      "epoch": 1.883997216807264,
      "grad_norm": 0.1515328288078308,
      "learning_rate": 3.720366792406539e-05,
      "loss": 0.5302,
      "mean_token_accuracy": 0.836488600820303,
      "num_tokens": 653552042.0,
      "step": 20477
    },
    {
      "entropy": 0.6563340723514557,
      "epoch": 1.8840892231787052,
      "grad_norm": 0.15589654445648193,
      "learning_rate": 3.720060109792376e-05,
      "loss": 0.6437,
      "mean_token_accuracy": 0.8011636883020401,
      "num_tokens": 653583389.0,
      "step": 20478
    },
    {
      "entropy": 0.6649831850081682,
      "epoch": 1.8841812295501463,
      "grad_norm": 0.15901030600070953,
      "learning_rate": 3.719753427178214e-05,
      "loss": 0.666,
      "mean_token_accuracy": 0.7964760512113571,
      "num_tokens": 653615084.0,
      "step": 20479
    },
    {
      "entropy": 0.6197346113622189,
      "epoch": 1.8842732359215875,
      "grad_norm": 0.15624664723873138,
      "learning_rate": 3.719446744564051e-05,
      "loss": 0.6282,
      "mean_token_accuracy": 0.8111806362867355,
      "num_tokens": 653647052.0,
      "step": 20480
    },
    {
      "entropy": 0.6258018286898732,
      "epoch": 1.8843652422930288,
      "grad_norm": 0.15988212823867798,
      "learning_rate": 3.719140061949888e-05,
      "loss": 0.6179,
      "mean_token_accuracy": 0.8120639882981777,
      "num_tokens": 653679401.0,
      "step": 20481
    },
    {
      "entropy": 0.5440200716257095,
      "epoch": 1.8844572486644702,
      "grad_norm": 0.15295901894569397,
      "learning_rate": 3.718833379335726e-05,
      "loss": 0.5252,
      "mean_token_accuracy": 0.8405452333390713,
      "num_tokens": 653710689.0,
      "step": 20482
    },
    {
      "entropy": 0.5629430669359863,
      "epoch": 1.8845492550359113,
      "grad_norm": 0.15203119814395905,
      "learning_rate": 3.718526696721563e-05,
      "loss": 0.5594,
      "mean_token_accuracy": 0.8314422853291035,
      "num_tokens": 653743148.0,
      "step": 20483
    },
    {
      "entropy": 0.5469514955766499,
      "epoch": 1.8846412614073524,
      "grad_norm": 0.15377727150917053,
      "learning_rate": 3.7182200141074006e-05,
      "loss": 0.536,
      "mean_token_accuracy": 0.8332965783774853,
      "num_tokens": 653775436.0,
      "step": 20484
    },
    {
      "entropy": 0.6086756782606244,
      "epoch": 1.8847332677787936,
      "grad_norm": 0.15364159643650055,
      "learning_rate": 3.717913331493238e-05,
      "loss": 0.6002,
      "mean_token_accuracy": 0.817359060049057,
      "num_tokens": 653807744.0,
      "step": 20485
    },
    {
      "entropy": 0.6918705757707357,
      "epoch": 1.884825274150235,
      "grad_norm": 0.1600377857685089,
      "learning_rate": 3.717606648879075e-05,
      "loss": 0.6978,
      "mean_token_accuracy": 0.7858597747981548,
      "num_tokens": 653839319.0,
      "step": 20486
    },
    {
      "entropy": 0.6228171768598258,
      "epoch": 1.8849172805216763,
      "grad_norm": 0.14921988546848297,
      "learning_rate": 3.717299966264913e-05,
      "loss": 0.6125,
      "mean_token_accuracy": 0.8133954592049122,
      "num_tokens": 653871747.0,
      "step": 20487
    },
    {
      "entropy": 0.5835735471919179,
      "epoch": 1.8850092868931174,
      "grad_norm": 0.150941401720047,
      "learning_rate": 3.71699328365075e-05,
      "loss": 0.5791,
      "mean_token_accuracy": 0.8225232474505901,
      "num_tokens": 653904198.0,
      "step": 20488
    },
    {
      "entropy": 0.6404789220541716,
      "epoch": 1.8851012932645586,
      "grad_norm": 0.16444896161556244,
      "learning_rate": 3.716686601036587e-05,
      "loss": 0.6408,
      "mean_token_accuracy": 0.8059851378202438,
      "num_tokens": 653935523.0,
      "step": 20489
    },
    {
      "entropy": 0.5783604849129915,
      "epoch": 1.8851932996359997,
      "grad_norm": 0.14942364394664764,
      "learning_rate": 3.716379918422425e-05,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8284849189221859,
      "num_tokens": 653967737.0,
      "step": 20490
    },
    {
      "entropy": 0.6474160673096776,
      "epoch": 1.885285306007441,
      "grad_norm": 0.16128745675086975,
      "learning_rate": 3.716073235808262e-05,
      "loss": 0.6353,
      "mean_token_accuracy": 0.802955437451601,
      "num_tokens": 653999618.0,
      "step": 20491
    },
    {
      "entropy": 0.6572075448930264,
      "epoch": 1.8853773123788824,
      "grad_norm": 0.1541927009820938,
      "learning_rate": 3.7157665531940996e-05,
      "loss": 0.641,
      "mean_token_accuracy": 0.8046676330268383,
      "num_tokens": 654031438.0,
      "step": 20492
    },
    {
      "entropy": 0.5928140366449952,
      "epoch": 1.8854693187503235,
      "grad_norm": 0.15593041479587555,
      "learning_rate": 3.715459870579937e-05,
      "loss": 0.5773,
      "mean_token_accuracy": 0.8192313872277737,
      "num_tokens": 654062703.0,
      "step": 20493
    },
    {
      "entropy": 0.6335612591356039,
      "epoch": 1.8855613251217647,
      "grad_norm": 0.15678469836711884,
      "learning_rate": 3.715153187965774e-05,
      "loss": 0.6145,
      "mean_token_accuracy": 0.8105778284370899,
      "num_tokens": 654095069.0,
      "step": 20494
    },
    {
      "entropy": 0.5786684132181108,
      "epoch": 1.8856533314932058,
      "grad_norm": 0.15067648887634277,
      "learning_rate": 3.714846505351612e-05,
      "loss": 0.5639,
      "mean_token_accuracy": 0.8227180168032646,
      "num_tokens": 654126607.0,
      "step": 20495
    },
    {
      "entropy": 0.6262596137821674,
      "epoch": 1.8857453378646472,
      "grad_norm": 0.151550754904747,
      "learning_rate": 3.714539822737449e-05,
      "loss": 0.6106,
      "mean_token_accuracy": 0.8124787174165249,
      "num_tokens": 654158609.0,
      "step": 20496
    },
    {
      "entropy": 0.6478900760412216,
      "epoch": 1.8858373442360885,
      "grad_norm": 0.1539705991744995,
      "learning_rate": 3.714233140123286e-05,
      "loss": 0.6377,
      "mean_token_accuracy": 0.8047105222940445,
      "num_tokens": 654190947.0,
      "step": 20497
    },
    {
      "entropy": 0.6324740555137396,
      "epoch": 1.8859293506075296,
      "grad_norm": 0.15252406895160675,
      "learning_rate": 3.713926457509124e-05,
      "loss": 0.6254,
      "mean_token_accuracy": 0.8108946867287159,
      "num_tokens": 654223475.0,
      "step": 20498
    },
    {
      "entropy": 0.43455940391868353,
      "epoch": 1.8860213569789708,
      "grad_norm": 0.14133936166763306,
      "learning_rate": 3.713619774894961e-05,
      "loss": 0.4223,
      "mean_token_accuracy": 0.8642280213534832,
      "num_tokens": 654254850.0,
      "step": 20499
    },
    {
      "entropy": 0.5932803191244602,
      "epoch": 1.886113363350412,
      "grad_norm": 0.1520935595035553,
      "learning_rate": 3.7133130922807994e-05,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8192647993564606,
      "num_tokens": 654286833.0,
      "step": 20500
    },
    {
      "entropy": 0.6899553500115871,
      "epoch": 1.8862053697218533,
      "grad_norm": 0.1585550755262375,
      "learning_rate": 3.713006409666636e-05,
      "loss": 0.6722,
      "mean_token_accuracy": 0.7968708872795105,
      "num_tokens": 654318727.0,
      "step": 20501
    },
    {
      "entropy": 0.6348686460405588,
      "epoch": 1.8862973760932946,
      "grad_norm": 0.1608988642692566,
      "learning_rate": 3.7126997270524736e-05,
      "loss": 0.6303,
      "mean_token_accuracy": 0.8065533488988876,
      "num_tokens": 654350990.0,
      "step": 20502
    },
    {
      "entropy": 0.5747887268662453,
      "epoch": 1.8863893824647358,
      "grad_norm": 0.15275874733924866,
      "learning_rate": 3.712393044438311e-05,
      "loss": 0.5587,
      "mean_token_accuracy": 0.8236473053693771,
      "num_tokens": 654382743.0,
      "step": 20503
    },
    {
      "entropy": 0.6808136254549026,
      "epoch": 1.8864813888361769,
      "grad_norm": 0.15645244717597961,
      "learning_rate": 3.7120863618241485e-05,
      "loss": 0.6749,
      "mean_token_accuracy": 0.796025887131691,
      "num_tokens": 654414842.0,
      "step": 20504
    },
    {
      "entropy": 0.5765939150005579,
      "epoch": 1.886573395207618,
      "grad_norm": 0.15791554749011993,
      "learning_rate": 3.711779679209986e-05,
      "loss": 0.5612,
      "mean_token_accuracy": 0.8245592154562473,
      "num_tokens": 654445761.0,
      "step": 20505
    },
    {
      "entropy": 0.6751337945461273,
      "epoch": 1.8866654015790594,
      "grad_norm": 0.15831191837787628,
      "learning_rate": 3.7114729965958235e-05,
      "loss": 0.6684,
      "mean_token_accuracy": 0.7978342324495316,
      "num_tokens": 654478059.0,
      "step": 20506
    },
    {
      "entropy": 0.5910321185365319,
      "epoch": 1.8867574079505007,
      "grad_norm": 0.1551073044538498,
      "learning_rate": 3.71116631398166e-05,
      "loss": 0.5873,
      "mean_token_accuracy": 0.8204599618911743,
      "num_tokens": 654510195.0,
      "step": 20507
    },
    {
      "entropy": 0.5694260150194168,
      "epoch": 1.8868494143219419,
      "grad_norm": 0.1595672369003296,
      "learning_rate": 3.7108596313674984e-05,
      "loss": 0.5659,
      "mean_token_accuracy": 0.8223972246050835,
      "num_tokens": 654541009.0,
      "step": 20508
    },
    {
      "entropy": 0.588766188826412,
      "epoch": 1.886941420693383,
      "grad_norm": 0.14638744294643402,
      "learning_rate": 3.710552948753335e-05,
      "loss": 0.5718,
      "mean_token_accuracy": 0.8225950822234154,
      "num_tokens": 654573217.0,
      "step": 20509
    },
    {
      "entropy": 0.6144774658605456,
      "epoch": 1.8870334270648241,
      "grad_norm": 0.1592666208744049,
      "learning_rate": 3.7102462661391727e-05,
      "loss": 0.5976,
      "mean_token_accuracy": 0.8150769136846066,
      "num_tokens": 654605483.0,
      "step": 20510
    },
    {
      "entropy": 0.7354085147380829,
      "epoch": 1.8871254334362655,
      "grad_norm": 0.16518855094909668,
      "learning_rate": 3.70993958352501e-05,
      "loss": 0.7349,
      "mean_token_accuracy": 0.7797732725739479,
      "num_tokens": 654637447.0,
      "step": 20511
    },
    {
      "entropy": 0.5939891580492258,
      "epoch": 1.8872174398077068,
      "grad_norm": 0.1570463329553604,
      "learning_rate": 3.7096329009108476e-05,
      "loss": 0.581,
      "mean_token_accuracy": 0.8184110783040524,
      "num_tokens": 654668230.0,
      "step": 20512
    },
    {
      "entropy": 0.6080569103360176,
      "epoch": 1.887309446179148,
      "grad_norm": 0.15211409330368042,
      "learning_rate": 3.709326218296685e-05,
      "loss": 0.5961,
      "mean_token_accuracy": 0.8157993853092194,
      "num_tokens": 654700382.0,
      "step": 20513
    },
    {
      "entropy": 0.632908757776022,
      "epoch": 1.887401452550589,
      "grad_norm": 0.15848353505134583,
      "learning_rate": 3.7090195356825225e-05,
      "loss": 0.6367,
      "mean_token_accuracy": 0.8096827380359173,
      "num_tokens": 654732091.0,
      "step": 20514
    },
    {
      "entropy": 0.612649243324995,
      "epoch": 1.8874934589220302,
      "grad_norm": 0.1566445678472519,
      "learning_rate": 3.708712853068359e-05,
      "loss": 0.5959,
      "mean_token_accuracy": 0.8167964853346348,
      "num_tokens": 654763901.0,
      "step": 20515
    },
    {
      "entropy": 0.6467593070119619,
      "epoch": 1.8875854652934716,
      "grad_norm": 0.1536368876695633,
      "learning_rate": 3.7084061704541974e-05,
      "loss": 0.6299,
      "mean_token_accuracy": 0.8075779639184475,
      "num_tokens": 654796589.0,
      "step": 20516
    },
    {
      "entropy": 0.5530796777456999,
      "epoch": 1.887677471664913,
      "grad_norm": 0.15212541818618774,
      "learning_rate": 3.708099487840034e-05,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8298387266695499,
      "num_tokens": 654829012.0,
      "step": 20517
    },
    {
      "entropy": 0.5809855968691409,
      "epoch": 1.887769478036354,
      "grad_norm": 0.15344808995723724,
      "learning_rate": 3.707792805225872e-05,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8222090937197208,
      "num_tokens": 654861213.0,
      "step": 20518
    },
    {
      "entropy": 0.6045116689056158,
      "epoch": 1.8878614844077952,
      "grad_norm": 0.15137962996959686,
      "learning_rate": 3.707486122611709e-05,
      "loss": 0.5925,
      "mean_token_accuracy": 0.8196960538625717,
      "num_tokens": 654893902.0,
      "step": 20519
    },
    {
      "entropy": 0.6252966299653053,
      "epoch": 1.8879534907792364,
      "grad_norm": 0.15530413389205933,
      "learning_rate": 3.7071794399975466e-05,
      "loss": 0.6194,
      "mean_token_accuracy": 0.8094313964247704,
      "num_tokens": 654925510.0,
      "step": 20520
    },
    {
      "entropy": 0.5490043172612786,
      "epoch": 1.8880454971506777,
      "grad_norm": 0.14738355576992035,
      "learning_rate": 3.706872757383384e-05,
      "loss": 0.5339,
      "mean_token_accuracy": 0.8387524448335171,
      "num_tokens": 654957783.0,
      "step": 20521
    },
    {
      "entropy": 0.557597697712481,
      "epoch": 1.888137503522119,
      "grad_norm": 0.14915235340595245,
      "learning_rate": 3.7065660747692216e-05,
      "loss": 0.5505,
      "mean_token_accuracy": 0.830311119556427,
      "num_tokens": 654990040.0,
      "step": 20522
    },
    {
      "entropy": 0.613866014406085,
      "epoch": 1.8882295098935602,
      "grad_norm": 0.150363489985466,
      "learning_rate": 3.706259392155059e-05,
      "loss": 0.6117,
      "mean_token_accuracy": 0.8106575831770897,
      "num_tokens": 655022482.0,
      "step": 20523
    },
    {
      "entropy": 0.686033833771944,
      "epoch": 1.8883215162650013,
      "grad_norm": 0.16041375696659088,
      "learning_rate": 3.7059527095408965e-05,
      "loss": 0.6681,
      "mean_token_accuracy": 0.7947268933057785,
      "num_tokens": 655053940.0,
      "step": 20524
    },
    {
      "entropy": 0.6606724709272385,
      "epoch": 1.8884135226364425,
      "grad_norm": 0.15142908692359924,
      "learning_rate": 3.705646026926733e-05,
      "loss": 0.6443,
      "mean_token_accuracy": 0.8036284260451794,
      "num_tokens": 655086094.0,
      "step": 20525
    },
    {
      "entropy": 0.5706528034061193,
      "epoch": 1.8885055290078838,
      "grad_norm": 0.14914140105247498,
      "learning_rate": 3.7053393443125714e-05,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8249669931828976,
      "num_tokens": 655118273.0,
      "step": 20526
    },
    {
      "entropy": 0.7065205574035645,
      "epoch": 1.8885975353793252,
      "grad_norm": 0.15727928280830383,
      "learning_rate": 3.705032661698408e-05,
      "loss": 0.7042,
      "mean_token_accuracy": 0.788977786898613,
      "num_tokens": 655150807.0,
      "step": 20527
    },
    {
      "entropy": 0.5143995345570147,
      "epoch": 1.8886895417507663,
      "grad_norm": 0.1504741609096527,
      "learning_rate": 3.704725979084246e-05,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8472022414207458,
      "num_tokens": 655183110.0,
      "step": 20528
    },
    {
      "entropy": 0.515281087718904,
      "epoch": 1.8887815481222074,
      "grad_norm": 0.14576351642608643,
      "learning_rate": 3.704419296470083e-05,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8465125858783722,
      "num_tokens": 655215456.0,
      "step": 20529
    },
    {
      "entropy": 0.680703729391098,
      "epoch": 1.8888735544936486,
      "grad_norm": 0.16101273894309998,
      "learning_rate": 3.7041126138559206e-05,
      "loss": 0.6909,
      "mean_token_accuracy": 0.7891299314796925,
      "num_tokens": 655248042.0,
      "step": 20530
    },
    {
      "entropy": 0.545241423882544,
      "epoch": 1.88896556086509,
      "grad_norm": 0.1439598947763443,
      "learning_rate": 3.703805931241758e-05,
      "loss": 0.5338,
      "mean_token_accuracy": 0.832953292876482,
      "num_tokens": 655279703.0,
      "step": 20531
    },
    {
      "entropy": 0.585349137429148,
      "epoch": 1.8890575672365313,
      "grad_norm": 0.15374755859375,
      "learning_rate": 3.7034992486275955e-05,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8228131085634232,
      "num_tokens": 655311523.0,
      "step": 20532
    },
    {
      "entropy": 0.5472720619291067,
      "epoch": 1.8891495736079724,
      "grad_norm": 0.1524309366941452,
      "learning_rate": 3.703192566013433e-05,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8374283649027348,
      "num_tokens": 655343716.0,
      "step": 20533
    },
    {
      "entropy": 0.6518481359817088,
      "epoch": 1.8892415799794136,
      "grad_norm": 0.15802785754203796,
      "learning_rate": 3.7028858833992705e-05,
      "loss": 0.6484,
      "mean_token_accuracy": 0.8087150566279888,
      "num_tokens": 655375363.0,
      "step": 20534
    },
    {
      "entropy": 0.6493612863123417,
      "epoch": 1.8893335863508547,
      "grad_norm": 0.1578502058982849,
      "learning_rate": 3.702579200785108e-05,
      "loss": 0.6361,
      "mean_token_accuracy": 0.8039544969797134,
      "num_tokens": 655407483.0,
      "step": 20535
    },
    {
      "entropy": 0.5573835503309965,
      "epoch": 1.889425592722296,
      "grad_norm": 0.15346018970012665,
      "learning_rate": 3.702272518170945e-05,
      "loss": 0.5497,
      "mean_token_accuracy": 0.8323032781481743,
      "num_tokens": 655438461.0,
      "step": 20536
    },
    {
      "entropy": 0.7437872178852558,
      "epoch": 1.8895175990937374,
      "grad_norm": 0.16417664289474487,
      "learning_rate": 3.701965835556783e-05,
      "loss": 0.7355,
      "mean_token_accuracy": 0.7782368361949921,
      "num_tokens": 655470171.0,
      "step": 20537
    },
    {
      "entropy": 0.5537210237234831,
      "epoch": 1.8896096054651785,
      "grad_norm": 0.14670181274414062,
      "learning_rate": 3.7016591529426196e-05,
      "loss": 0.5504,
      "mean_token_accuracy": 0.8307632468640804,
      "num_tokens": 655502610.0,
      "step": 20538
    },
    {
      "entropy": 0.636171318590641,
      "epoch": 1.8897016118366197,
      "grad_norm": 0.16136504709720612,
      "learning_rate": 3.701352470328458e-05,
      "loss": 0.6176,
      "mean_token_accuracy": 0.8118271939456463,
      "num_tokens": 655534978.0,
      "step": 20539
    },
    {
      "entropy": 0.6096402939874679,
      "epoch": 1.8897936182080608,
      "grad_norm": 0.1495138555765152,
      "learning_rate": 3.7010457877142946e-05,
      "loss": 0.5769,
      "mean_token_accuracy": 0.8198346793651581,
      "num_tokens": 655565995.0,
      "step": 20540
    },
    {
      "entropy": 0.6086553074419498,
      "epoch": 1.8898856245795022,
      "grad_norm": 0.1551857590675354,
      "learning_rate": 3.700739105100132e-05,
      "loss": 0.6012,
      "mean_token_accuracy": 0.8121207132935524,
      "num_tokens": 655597587.0,
      "step": 20541
    },
    {
      "entropy": 0.6268911277875304,
      "epoch": 1.8899776309509435,
      "grad_norm": 0.15220439434051514,
      "learning_rate": 3.7004324224859695e-05,
      "loss": 0.6186,
      "mean_token_accuracy": 0.8141062743961811,
      "num_tokens": 655629743.0,
      "step": 20542
    },
    {
      "entropy": 0.584410959854722,
      "epoch": 1.8900696373223846,
      "grad_norm": 0.1519598364830017,
      "learning_rate": 3.700125739871807e-05,
      "loss": 0.5638,
      "mean_token_accuracy": 0.823399979621172,
      "num_tokens": 655662140.0,
      "step": 20543
    },
    {
      "entropy": 0.6980295851826668,
      "epoch": 1.8901616436938258,
      "grad_norm": 0.16341710090637207,
      "learning_rate": 3.6998190572576444e-05,
      "loss": 0.6828,
      "mean_token_accuracy": 0.7904841639101505,
      "num_tokens": 655693134.0,
      "step": 20544
    },
    {
      "entropy": 0.5503797912970185,
      "epoch": 1.890253650065267,
      "grad_norm": 0.14742852747440338,
      "learning_rate": 3.699512374643482e-05,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8371250182390213,
      "num_tokens": 655724767.0,
      "step": 20545
    },
    {
      "entropy": 0.6343416552990675,
      "epoch": 1.8903456564367083,
      "grad_norm": 0.16473232209682465,
      "learning_rate": 3.699205692029319e-05,
      "loss": 0.6243,
      "mean_token_accuracy": 0.8074632808566093,
      "num_tokens": 655756418.0,
      "step": 20546
    },
    {
      "entropy": 0.5098768984898925,
      "epoch": 1.8904376628081496,
      "grad_norm": 0.15081113576889038,
      "learning_rate": 3.698899009415157e-05,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8431964591145515,
      "num_tokens": 655788618.0,
      "step": 20547
    },
    {
      "entropy": 0.7207122556865215,
      "epoch": 1.8905296691795908,
      "grad_norm": 0.1666545867919922,
      "learning_rate": 3.6985923268009936e-05,
      "loss": 0.7073,
      "mean_token_accuracy": 0.7838005125522614,
      "num_tokens": 655819967.0,
      "step": 20548
    },
    {
      "entropy": 0.6161044873297215,
      "epoch": 1.8906216755510319,
      "grad_norm": 0.15187232196331024,
      "learning_rate": 3.698285644186831e-05,
      "loss": 0.6153,
      "mean_token_accuracy": 0.8112354055047035,
      "num_tokens": 655852717.0,
      "step": 20549
    },
    {
      "entropy": 0.5344935627654195,
      "epoch": 1.890713681922473,
      "grad_norm": 0.14482593536376953,
      "learning_rate": 3.6979789615726685e-05,
      "loss": 0.5227,
      "mean_token_accuracy": 0.840659387409687,
      "num_tokens": 655885291.0,
      "step": 20550
    },
    {
      "entropy": 0.5808232207782567,
      "epoch": 1.8908056882939144,
      "grad_norm": 0.15611383318901062,
      "learning_rate": 3.697672278958506e-05,
      "loss": 0.5841,
      "mean_token_accuracy": 0.8207594938576221,
      "num_tokens": 655917661.0,
      "step": 20551
    },
    {
      "entropy": 0.5286053698509932,
      "epoch": 1.8908976946653557,
      "grad_norm": 0.1527092456817627,
      "learning_rate": 3.6973655963443435e-05,
      "loss": 0.5195,
      "mean_token_accuracy": 0.8388797678053379,
      "num_tokens": 655949652.0,
      "step": 20552
    },
    {
      "entropy": 0.6651683803647757,
      "epoch": 1.8909897010367969,
      "grad_norm": 0.1619216501712799,
      "learning_rate": 3.697058913730181e-05,
      "loss": 0.6746,
      "mean_token_accuracy": 0.7966857254505157,
      "num_tokens": 655981461.0,
      "step": 20553
    },
    {
      "entropy": 0.6388550121337175,
      "epoch": 1.891081707408238,
      "grad_norm": 0.15701739490032196,
      "learning_rate": 3.696752231116018e-05,
      "loss": 0.6105,
      "mean_token_accuracy": 0.8105076923966408,
      "num_tokens": 656012587.0,
      "step": 20554
    },
    {
      "entropy": 0.6970151290297508,
      "epoch": 1.8911737137796791,
      "grad_norm": 0.16227854788303375,
      "learning_rate": 3.696445548501856e-05,
      "loss": 0.695,
      "mean_token_accuracy": 0.7859952822327614,
      "num_tokens": 656044502.0,
      "step": 20555
    },
    {
      "entropy": 0.5552813149988651,
      "epoch": 1.8912657201511205,
      "grad_norm": 0.15002764761447906,
      "learning_rate": 3.6961388658876926e-05,
      "loss": 0.547,
      "mean_token_accuracy": 0.830603051930666,
      "num_tokens": 656076408.0,
      "step": 20556
    },
    {
      "entropy": 0.5829848237335682,
      "epoch": 1.8913577265225618,
      "grad_norm": 0.14458639919757843,
      "learning_rate": 3.69583218327353e-05,
      "loss": 0.5659,
      "mean_token_accuracy": 0.827257800847292,
      "num_tokens": 656108748.0,
      "step": 20557
    },
    {
      "entropy": 0.6459936685860157,
      "epoch": 1.891449732894003,
      "grad_norm": 0.1554967164993286,
      "learning_rate": 3.6955255006593676e-05,
      "loss": 0.6375,
      "mean_token_accuracy": 0.802279457449913,
      "num_tokens": 656140981.0,
      "step": 20558
    },
    {
      "entropy": 0.6256263228133321,
      "epoch": 1.891541739265444,
      "grad_norm": 0.15121714770793915,
      "learning_rate": 3.695218818045205e-05,
      "loss": 0.617,
      "mean_token_accuracy": 0.8090228289365768,
      "num_tokens": 656172682.0,
      "step": 20559
    },
    {
      "entropy": 0.5123894447460771,
      "epoch": 1.8916337456368852,
      "grad_norm": 0.14846445620059967,
      "learning_rate": 3.694912135431043e-05,
      "loss": 0.5016,
      "mean_token_accuracy": 0.8449485972523689,
      "num_tokens": 656204730.0,
      "step": 20560
    },
    {
      "entropy": 0.5296467309817672,
      "epoch": 1.8917257520083266,
      "grad_norm": 0.14786465466022491,
      "learning_rate": 3.69460545281688e-05,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8400680050253868,
      "num_tokens": 656236907.0,
      "step": 20561
    },
    {
      "entropy": 0.6141238873824477,
      "epoch": 1.891817758379768,
      "grad_norm": 0.15010122954845428,
      "learning_rate": 3.6942987702027174e-05,
      "loss": 0.5948,
      "mean_token_accuracy": 0.8180873654782772,
      "num_tokens": 656268888.0,
      "step": 20562
    },
    {
      "entropy": 0.6551941689103842,
      "epoch": 1.891909764751209,
      "grad_norm": 0.15796145796775818,
      "learning_rate": 3.693992087588555e-05,
      "loss": 0.654,
      "mean_token_accuracy": 0.8008983507752419,
      "num_tokens": 656300916.0,
      "step": 20563
    },
    {
      "entropy": 0.6782443355768919,
      "epoch": 1.8920017711226502,
      "grad_norm": 0.16168242692947388,
      "learning_rate": 3.6936854049743924e-05,
      "loss": 0.6815,
      "mean_token_accuracy": 0.791256669908762,
      "num_tokens": 656332899.0,
      "step": 20564
    },
    {
      "entropy": 0.6071002362295985,
      "epoch": 1.8920937774940914,
      "grad_norm": 0.15578888356685638,
      "learning_rate": 3.69337872236023e-05,
      "loss": 0.6021,
      "mean_token_accuracy": 0.8148874416947365,
      "num_tokens": 656363806.0,
      "step": 20565
    },
    {
      "entropy": 0.5697994576767087,
      "epoch": 1.8921857838655327,
      "grad_norm": 0.14795033633708954,
      "learning_rate": 3.693072039746067e-05,
      "loss": 0.5595,
      "mean_token_accuracy": 0.826535128057003,
      "num_tokens": 656395744.0,
      "step": 20566
    },
    {
      "entropy": 0.5643708677962422,
      "epoch": 1.892277790236974,
      "grad_norm": 0.14621320366859436,
      "learning_rate": 3.692765357131904e-05,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8275788985192776,
      "num_tokens": 656428454.0,
      "step": 20567
    },
    {
      "entropy": 0.6581464819610119,
      "epoch": 1.8923697966084152,
      "grad_norm": 0.16235461831092834,
      "learning_rate": 3.692458674517742e-05,
      "loss": 0.6515,
      "mean_token_accuracy": 0.8011713065207005,
      "num_tokens": 656460337.0,
      "step": 20568
    },
    {
      "entropy": 0.6121187657117844,
      "epoch": 1.8924618029798563,
      "grad_norm": 0.15615114569664001,
      "learning_rate": 3.692151991903579e-05,
      "loss": 0.5974,
      "mean_token_accuracy": 0.8187981732189655,
      "num_tokens": 656492204.0,
      "step": 20569
    },
    {
      "entropy": 0.6109717506915331,
      "epoch": 1.8925538093512975,
      "grad_norm": 0.1524396389722824,
      "learning_rate": 3.6918453092894165e-05,
      "loss": 0.5991,
      "mean_token_accuracy": 0.8171047456562519,
      "num_tokens": 656524253.0,
      "step": 20570
    },
    {
      "entropy": 0.6340795801952481,
      "epoch": 1.8926458157227388,
      "grad_norm": 0.1596037894487381,
      "learning_rate": 3.691538626675254e-05,
      "loss": 0.6299,
      "mean_token_accuracy": 0.8047255426645279,
      "num_tokens": 656556346.0,
      "step": 20571
    },
    {
      "entropy": 0.5656970990821719,
      "epoch": 1.8927378220941802,
      "grad_norm": 0.15096861124038696,
      "learning_rate": 3.6912319440610914e-05,
      "loss": 0.5519,
      "mean_token_accuracy": 0.8275048062205315,
      "num_tokens": 656587984.0,
      "step": 20572
    },
    {
      "entropy": 0.6876035369932652,
      "epoch": 1.8928298284656213,
      "grad_norm": 0.1639631688594818,
      "learning_rate": 3.690925261446929e-05,
      "loss": 0.6848,
      "mean_token_accuracy": 0.7903395034372807,
      "num_tokens": 656619000.0,
      "step": 20573
    },
    {
      "entropy": 0.583177737891674,
      "epoch": 1.8929218348370624,
      "grad_norm": 0.15072110295295715,
      "learning_rate": 3.690618578832766e-05,
      "loss": 0.5711,
      "mean_token_accuracy": 0.8227859288454056,
      "num_tokens": 656650599.0,
      "step": 20574
    },
    {
      "entropy": 0.6707805544137955,
      "epoch": 1.8930138412085036,
      "grad_norm": 0.16698114573955536,
      "learning_rate": 3.690311896218603e-05,
      "loss": 0.6798,
      "mean_token_accuracy": 0.7917990274727345,
      "num_tokens": 656682284.0,
      "step": 20575
    },
    {
      "entropy": 0.7006223136559129,
      "epoch": 1.893105847579945,
      "grad_norm": 0.16054505109786987,
      "learning_rate": 3.690005213604441e-05,
      "loss": 0.7012,
      "mean_token_accuracy": 0.7866602428257465,
      "num_tokens": 656714659.0,
      "step": 20576
    },
    {
      "entropy": 0.6693970318883657,
      "epoch": 1.8931978539513863,
      "grad_norm": 0.1616026759147644,
      "learning_rate": 3.689698530990278e-05,
      "loss": 0.6729,
      "mean_token_accuracy": 0.7968200147151947,
      "num_tokens": 656746485.0,
      "step": 20577
    },
    {
      "entropy": 0.6278382688760757,
      "epoch": 1.8932898603228274,
      "grad_norm": 0.1522424966096878,
      "learning_rate": 3.6893918483761155e-05,
      "loss": 0.6204,
      "mean_token_accuracy": 0.8076294139027596,
      "num_tokens": 656777983.0,
      "step": 20578
    },
    {
      "entropy": 0.5912092979997396,
      "epoch": 1.8933818666942686,
      "grad_norm": 0.15247154235839844,
      "learning_rate": 3.689085165761953e-05,
      "loss": 0.5846,
      "mean_token_accuracy": 0.8206098042428493,
      "num_tokens": 656808951.0,
      "step": 20579
    },
    {
      "entropy": 0.49591861851513386,
      "epoch": 1.8934738730657097,
      "grad_norm": 0.14736899733543396,
      "learning_rate": 3.6887784831477904e-05,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8478221148252487,
      "num_tokens": 656840084.0,
      "step": 20580
    },
    {
      "entropy": 0.7196319214999676,
      "epoch": 1.893565879437151,
      "grad_norm": 0.16546989977359772,
      "learning_rate": 3.688471800533628e-05,
      "loss": 0.714,
      "mean_token_accuracy": 0.7835902832448483,
      "num_tokens": 656872058.0,
      "step": 20581
    },
    {
      "entropy": 0.6099151293747127,
      "epoch": 1.8936578858085924,
      "grad_norm": 0.15838827192783356,
      "learning_rate": 3.6881651179194654e-05,
      "loss": 0.607,
      "mean_token_accuracy": 0.8128664791584015,
      "num_tokens": 656903606.0,
      "step": 20582
    },
    {
      "entropy": 0.5752838682383299,
      "epoch": 1.8937498921800335,
      "grad_norm": 0.148825004696846,
      "learning_rate": 3.687858435305303e-05,
      "loss": 0.5586,
      "mean_token_accuracy": 0.8235389664769173,
      "num_tokens": 656934897.0,
      "step": 20583
    },
    {
      "entropy": 0.5780243966728449,
      "epoch": 1.8938418985514747,
      "grad_norm": 0.14996124804019928,
      "learning_rate": 3.68755175269114e-05,
      "loss": 0.5632,
      "mean_token_accuracy": 0.8308132067322731,
      "num_tokens": 656967173.0,
      "step": 20584
    },
    {
      "entropy": 0.5963309621438384,
      "epoch": 1.8939339049229158,
      "grad_norm": 0.15173660218715668,
      "learning_rate": 3.687245070076977e-05,
      "loss": 0.5769,
      "mean_token_accuracy": 0.8246921747922897,
      "num_tokens": 656998495.0,
      "step": 20585
    },
    {
      "entropy": 0.6602252200245857,
      "epoch": 1.8940259112943572,
      "grad_norm": 0.16061730682849884,
      "learning_rate": 3.686938387462815e-05,
      "loss": 0.6501,
      "mean_token_accuracy": 0.8017287403345108,
      "num_tokens": 657030451.0,
      "step": 20586
    },
    {
      "entropy": 0.5130877615883946,
      "epoch": 1.8941179176657985,
      "grad_norm": 0.13933368027210236,
      "learning_rate": 3.686631704848652e-05,
      "loss": 0.4945,
      "mean_token_accuracy": 0.8465742506086826,
      "num_tokens": 657063219.0,
      "step": 20587
    },
    {
      "entropy": 0.6361262695863843,
      "epoch": 1.8942099240372396,
      "grad_norm": 0.15245182812213898,
      "learning_rate": 3.6863250222344895e-05,
      "loss": 0.628,
      "mean_token_accuracy": 0.8063956201076508,
      "num_tokens": 657094530.0,
      "step": 20588
    },
    {
      "entropy": 0.5706221326254308,
      "epoch": 1.8943019304086808,
      "grad_norm": 0.1492668241262436,
      "learning_rate": 3.686018339620327e-05,
      "loss": 0.5448,
      "mean_token_accuracy": 0.834033876657486,
      "num_tokens": 657126785.0,
      "step": 20589
    },
    {
      "entropy": 0.6963326651602983,
      "epoch": 1.894393936780122,
      "grad_norm": 0.16347919404506683,
      "learning_rate": 3.6857116570061644e-05,
      "loss": 0.6623,
      "mean_token_accuracy": 0.7940581366419792,
      "num_tokens": 657157655.0,
      "step": 20590
    },
    {
      "entropy": 0.5741553902626038,
      "epoch": 1.8944859431515633,
      "grad_norm": 0.1545308381319046,
      "learning_rate": 3.685404974392002e-05,
      "loss": 0.5638,
      "mean_token_accuracy": 0.8254738599061966,
      "num_tokens": 657189629.0,
      "step": 20591
    },
    {
      "entropy": 0.6736860051751137,
      "epoch": 1.8945779495230046,
      "grad_norm": 0.16031275689601898,
      "learning_rate": 3.6850982917778393e-05,
      "loss": 0.6534,
      "mean_token_accuracy": 0.8016556650400162,
      "num_tokens": 657221237.0,
      "step": 20592
    },
    {
      "entropy": 0.6437602946534753,
      "epoch": 1.8946699558944458,
      "grad_norm": 0.15528683364391327,
      "learning_rate": 3.684791609163677e-05,
      "loss": 0.6474,
      "mean_token_accuracy": 0.8039398528635502,
      "num_tokens": 657253514.0,
      "step": 20593
    },
    {
      "entropy": 0.6758951134979725,
      "epoch": 1.894761962265887,
      "grad_norm": 0.15382060408592224,
      "learning_rate": 3.684484926549514e-05,
      "loss": 0.6641,
      "mean_token_accuracy": 0.7963695637881756,
      "num_tokens": 657285254.0,
      "step": 20594
    },
    {
      "entropy": 0.6460079196840525,
      "epoch": 1.894853968637328,
      "grad_norm": 0.1521385908126831,
      "learning_rate": 3.684178243935352e-05,
      "loss": 0.6323,
      "mean_token_accuracy": 0.8051769770681858,
      "num_tokens": 657317187.0,
      "step": 20595
    },
    {
      "entropy": 0.646613223478198,
      "epoch": 1.8949459750087694,
      "grad_norm": 0.16119582951068878,
      "learning_rate": 3.6838715613211885e-05,
      "loss": 0.6415,
      "mean_token_accuracy": 0.8043827675282955,
      "num_tokens": 657349084.0,
      "step": 20596
    },
    {
      "entropy": 0.5168444467708468,
      "epoch": 1.8950379813802107,
      "grad_norm": 0.14573803544044495,
      "learning_rate": 3.683564878707027e-05,
      "loss": 0.5079,
      "mean_token_accuracy": 0.840923860669136,
      "num_tokens": 657380615.0,
      "step": 20597
    },
    {
      "entropy": 0.607164403423667,
      "epoch": 1.8951299877516519,
      "grad_norm": 0.1514754593372345,
      "learning_rate": 3.6832581960928635e-05,
      "loss": 0.5913,
      "mean_token_accuracy": 0.818777821958065,
      "num_tokens": 657412900.0,
      "step": 20598
    },
    {
      "entropy": 0.6406329041346908,
      "epoch": 1.895221994123093,
      "grad_norm": 0.1571802794933319,
      "learning_rate": 3.6829515134787016e-05,
      "loss": 0.6213,
      "mean_token_accuracy": 0.808980543166399,
      "num_tokens": 657444621.0,
      "step": 20599
    },
    {
      "entropy": 0.5714195156469941,
      "epoch": 1.8953140004945341,
      "grad_norm": 0.14803491532802582,
      "learning_rate": 3.6826448308645384e-05,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8300492390990257,
      "num_tokens": 657477366.0,
      "step": 20600
    },
    {
      "entropy": 0.47695068968459964,
      "epoch": 1.8954060068659755,
      "grad_norm": 0.13785916566848755,
      "learning_rate": 3.682338148250376e-05,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8560302332043648,
      "num_tokens": 657509384.0,
      "step": 20601
    },
    {
      "entropy": 0.6795201189815998,
      "epoch": 1.8954980132374168,
      "grad_norm": 0.16310173273086548,
      "learning_rate": 3.682031465636213e-05,
      "loss": 0.6607,
      "mean_token_accuracy": 0.7947343587875366,
      "num_tokens": 657540202.0,
      "step": 20602
    },
    {
      "entropy": 0.6253451053053141,
      "epoch": 1.895590019608858,
      "grad_norm": 0.1570916622877121,
      "learning_rate": 3.681724783022051e-05,
      "loss": 0.6246,
      "mean_token_accuracy": 0.8087362572550774,
      "num_tokens": 657572543.0,
      "step": 20603
    },
    {
      "entropy": 0.5279541574418545,
      "epoch": 1.8956820259802991,
      "grad_norm": 0.15283572673797607,
      "learning_rate": 3.681418100407888e-05,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8370394334197044,
      "num_tokens": 657603565.0,
      "step": 20604
    },
    {
      "entropy": 0.6149057308211923,
      "epoch": 1.8957740323517402,
      "grad_norm": 0.1563723236322403,
      "learning_rate": 3.681111417793726e-05,
      "loss": 0.615,
      "mean_token_accuracy": 0.8127635978162289,
      "num_tokens": 657635013.0,
      "step": 20605
    },
    {
      "entropy": 0.6933577638119459,
      "epoch": 1.8958660387231816,
      "grad_norm": 0.15890918672084808,
      "learning_rate": 3.6808047351795625e-05,
      "loss": 0.6954,
      "mean_token_accuracy": 0.7885665073990822,
      "num_tokens": 657667405.0,
      "step": 20606
    },
    {
      "entropy": 0.6789960525929928,
      "epoch": 1.895958045094623,
      "grad_norm": 0.15897642076015472,
      "learning_rate": 3.6804980525654006e-05,
      "loss": 0.6707,
      "mean_token_accuracy": 0.793330293148756,
      "num_tokens": 657698561.0,
      "step": 20607
    },
    {
      "entropy": 0.6451615514233708,
      "epoch": 1.896050051466064,
      "grad_norm": 0.1544840931892395,
      "learning_rate": 3.6801913699512374e-05,
      "loss": 0.6455,
      "mean_token_accuracy": 0.8045890629291534,
      "num_tokens": 657730999.0,
      "step": 20608
    },
    {
      "entropy": 0.6248176600784063,
      "epoch": 1.8961420578375052,
      "grad_norm": 0.15655291080474854,
      "learning_rate": 3.679884687337075e-05,
      "loss": 0.6212,
      "mean_token_accuracy": 0.8098314441740513,
      "num_tokens": 657762120.0,
      "step": 20609
    },
    {
      "entropy": 0.6178407622501254,
      "epoch": 1.8962340642089464,
      "grad_norm": 0.15537922084331512,
      "learning_rate": 3.6795780047229124e-05,
      "loss": 0.6002,
      "mean_token_accuracy": 0.8109054043889046,
      "num_tokens": 657793630.0,
      "step": 20610
    },
    {
      "entropy": 0.6164693366736174,
      "epoch": 1.8963260705803877,
      "grad_norm": 0.15611545741558075,
      "learning_rate": 3.67927132210875e-05,
      "loss": 0.6105,
      "mean_token_accuracy": 0.8168047666549683,
      "num_tokens": 657825750.0,
      "step": 20611
    },
    {
      "entropy": 0.6507544796913862,
      "epoch": 1.896418076951829,
      "grad_norm": 0.1511814445257187,
      "learning_rate": 3.678964639494587e-05,
      "loss": 0.6369,
      "mean_token_accuracy": 0.806624710559845,
      "num_tokens": 657857890.0,
      "step": 20612
    },
    {
      "entropy": 0.5377644123509526,
      "epoch": 1.8965100833232702,
      "grad_norm": 0.1495581865310669,
      "learning_rate": 3.678657956880425e-05,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8358509875833988,
      "num_tokens": 657889872.0,
      "step": 20613
    },
    {
      "entropy": 0.5904602017253637,
      "epoch": 1.8966020896947113,
      "grad_norm": 0.15002769231796265,
      "learning_rate": 3.6783512742662615e-05,
      "loss": 0.5828,
      "mean_token_accuracy": 0.8191173449158669,
      "num_tokens": 657922219.0,
      "step": 20614
    },
    {
      "entropy": 0.5773279285058379,
      "epoch": 1.8966940960661525,
      "grad_norm": 0.15364256501197815,
      "learning_rate": 3.6780445916521e-05,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8235106132924557,
      "num_tokens": 657954041.0,
      "step": 20615
    },
    {
      "entropy": 0.6349945981055498,
      "epoch": 1.8967861024375938,
      "grad_norm": 0.15599481761455536,
      "learning_rate": 3.6777379090379365e-05,
      "loss": 0.634,
      "mean_token_accuracy": 0.8014467358589172,
      "num_tokens": 657986315.0,
      "step": 20616
    },
    {
      "entropy": 0.7391075454652309,
      "epoch": 1.8968781088090352,
      "grad_norm": 0.16679231822490692,
      "learning_rate": 3.677431226423774e-05,
      "loss": 0.7366,
      "mean_token_accuracy": 0.7783462777733803,
      "num_tokens": 658018697.0,
      "step": 20617
    },
    {
      "entropy": 0.6503420835360885,
      "epoch": 1.8969701151804763,
      "grad_norm": 0.1608278900384903,
      "learning_rate": 3.6771245438096114e-05,
      "loss": 0.6415,
      "mean_token_accuracy": 0.8048616088926792,
      "num_tokens": 658050559.0,
      "step": 20618
    },
    {
      "entropy": 0.578287404961884,
      "epoch": 1.8970621215519174,
      "grad_norm": 0.14966504275798798,
      "learning_rate": 3.676817861195449e-05,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8240439109504223,
      "num_tokens": 658082644.0,
      "step": 20619
    },
    {
      "entropy": 0.6681525073945522,
      "epoch": 1.8971541279233586,
      "grad_norm": 0.16607666015625,
      "learning_rate": 3.676511178581286e-05,
      "loss": 0.6673,
      "mean_token_accuracy": 0.7989933677017689,
      "num_tokens": 658114867.0,
      "step": 20620
    },
    {
      "entropy": 0.5930913500487804,
      "epoch": 1.8972461342948,
      "grad_norm": 0.15856556594371796,
      "learning_rate": 3.676204495967124e-05,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8218420594930649,
      "num_tokens": 658146881.0,
      "step": 20621
    },
    {
      "entropy": 0.5789475180208683,
      "epoch": 1.8973381406662413,
      "grad_norm": 0.15114183723926544,
      "learning_rate": 3.675897813352961e-05,
      "loss": 0.5736,
      "mean_token_accuracy": 0.825593538582325,
      "num_tokens": 658178340.0,
      "step": 20622
    },
    {
      "entropy": 0.6767120314761996,
      "epoch": 1.8974301470376824,
      "grad_norm": 0.1576332300901413,
      "learning_rate": 3.675591130738799e-05,
      "loss": 0.6627,
      "mean_token_accuracy": 0.7985725253820419,
      "num_tokens": 658209889.0,
      "step": 20623
    },
    {
      "entropy": 0.575576912611723,
      "epoch": 1.8975221534091236,
      "grad_norm": 0.15028797090053558,
      "learning_rate": 3.675284448124636e-05,
      "loss": 0.5591,
      "mean_token_accuracy": 0.8325939141213894,
      "num_tokens": 658241942.0,
      "step": 20624
    },
    {
      "entropy": 0.5594172850251198,
      "epoch": 1.8976141597805647,
      "grad_norm": 0.14477330446243286,
      "learning_rate": 3.6749777655104736e-05,
      "loss": 0.5509,
      "mean_token_accuracy": 0.8261824399232864,
      "num_tokens": 658274316.0,
      "step": 20625
    },
    {
      "entropy": 0.5962465479969978,
      "epoch": 1.897706166152006,
      "grad_norm": 0.15351933240890503,
      "learning_rate": 3.674671082896311e-05,
      "loss": 0.5896,
      "mean_token_accuracy": 0.8208458460867405,
      "num_tokens": 658306685.0,
      "step": 20626
    },
    {
      "entropy": 0.6353097376413643,
      "epoch": 1.8977981725234474,
      "grad_norm": 0.16134168207645416,
      "learning_rate": 3.674364400282148e-05,
      "loss": 0.6141,
      "mean_token_accuracy": 0.8126231133937836,
      "num_tokens": 658338236.0,
      "step": 20627
    },
    {
      "entropy": 0.6240457827225327,
      "epoch": 1.8978901788948885,
      "grad_norm": 0.15106473863124847,
      "learning_rate": 3.674057717667986e-05,
      "loss": 0.6126,
      "mean_token_accuracy": 0.8119592554867268,
      "num_tokens": 658370509.0,
      "step": 20628
    },
    {
      "entropy": 0.6466499920934439,
      "epoch": 1.8979821852663297,
      "grad_norm": 0.15942838788032532,
      "learning_rate": 3.673751035053823e-05,
      "loss": 0.6384,
      "mean_token_accuracy": 0.801893662661314,
      "num_tokens": 658402172.0,
      "step": 20629
    },
    {
      "entropy": 0.7202057559043169,
      "epoch": 1.8980741916377708,
      "grad_norm": 0.16505067050457,
      "learning_rate": 3.67344435243966e-05,
      "loss": 0.7082,
      "mean_token_accuracy": 0.7858003489673138,
      "num_tokens": 658434046.0,
      "step": 20630
    },
    {
      "entropy": 0.7071021441370249,
      "epoch": 1.8981661980092122,
      "grad_norm": 0.16132567822933197,
      "learning_rate": 3.673137669825498e-05,
      "loss": 0.6951,
      "mean_token_accuracy": 0.7866274416446686,
      "num_tokens": 658466214.0,
      "step": 20631
    },
    {
      "entropy": 0.5914899539202452,
      "epoch": 1.8982582043806535,
      "grad_norm": 0.15282490849494934,
      "learning_rate": 3.672830987211335e-05,
      "loss": 0.5786,
      "mean_token_accuracy": 0.8216210529208183,
      "num_tokens": 658497957.0,
      "step": 20632
    },
    {
      "entropy": 0.552270632237196,
      "epoch": 1.8983502107520946,
      "grad_norm": 0.14646406471729279,
      "learning_rate": 3.672524304597173e-05,
      "loss": 0.5453,
      "mean_token_accuracy": 0.8319859579205513,
      "num_tokens": 658529445.0,
      "step": 20633
    },
    {
      "entropy": 0.5761431911960244,
      "epoch": 1.8984422171235358,
      "grad_norm": 0.1440998613834381,
      "learning_rate": 3.67221762198301e-05,
      "loss": 0.5618,
      "mean_token_accuracy": 0.8255277015268803,
      "num_tokens": 658561243.0,
      "step": 20634
    },
    {
      "entropy": 0.5229187640361488,
      "epoch": 1.898534223494977,
      "grad_norm": 0.1503681093454361,
      "learning_rate": 3.671910939368847e-05,
      "loss": 0.5143,
      "mean_token_accuracy": 0.8406769633293152,
      "num_tokens": 658593168.0,
      "step": 20635
    },
    {
      "entropy": 0.560537688434124,
      "epoch": 1.8986262298664183,
      "grad_norm": 0.15103007853031158,
      "learning_rate": 3.671604256754685e-05,
      "loss": 0.55,
      "mean_token_accuracy": 0.8296951465308666,
      "num_tokens": 658625790.0,
      "step": 20636
    },
    {
      "entropy": 0.6073828423395753,
      "epoch": 1.8987182362378596,
      "grad_norm": 0.1569230854511261,
      "learning_rate": 3.671297574140522e-05,
      "loss": 0.6034,
      "mean_token_accuracy": 0.8149823360145092,
      "num_tokens": 658658007.0,
      "step": 20637
    },
    {
      "entropy": 0.5959390513598919,
      "epoch": 1.8988102426093008,
      "grad_norm": 0.14875437319278717,
      "learning_rate": 3.670990891526359e-05,
      "loss": 0.5782,
      "mean_token_accuracy": 0.8198520988225937,
      "num_tokens": 658689909.0,
      "step": 20638
    },
    {
      "entropy": 0.7056759772822261,
      "epoch": 1.898902248980742,
      "grad_norm": 0.15649311244487762,
      "learning_rate": 3.670684208912197e-05,
      "loss": 0.6953,
      "mean_token_accuracy": 0.7897915989160538,
      "num_tokens": 658722406.0,
      "step": 20639
    },
    {
      "entropy": 0.575776819139719,
      "epoch": 1.898994255352183,
      "grad_norm": 0.15198540687561035,
      "learning_rate": 3.670377526298034e-05,
      "loss": 0.5521,
      "mean_token_accuracy": 0.8265110477805138,
      "num_tokens": 658754146.0,
      "step": 20640
    },
    {
      "entropy": 0.4802215229719877,
      "epoch": 1.8990862617236244,
      "grad_norm": 0.1373167783021927,
      "learning_rate": 3.670070843683872e-05,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8545407466590405,
      "num_tokens": 658786639.0,
      "step": 20641
    },
    {
      "entropy": 0.533205577172339,
      "epoch": 1.8991782680950657,
      "grad_norm": 0.1490245759487152,
      "learning_rate": 3.669764161069709e-05,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8347936235368252,
      "num_tokens": 658818334.0,
      "step": 20642
    },
    {
      "entropy": 0.6347825471311808,
      "epoch": 1.8992702744665069,
      "grad_norm": 0.15952257812023163,
      "learning_rate": 3.6694574784555467e-05,
      "loss": 0.6382,
      "mean_token_accuracy": 0.8038492314517498,
      "num_tokens": 658850002.0,
      "step": 20643
    },
    {
      "entropy": 0.6089100139215589,
      "epoch": 1.899362280837948,
      "grad_norm": 0.15627069771289825,
      "learning_rate": 3.669150795841384e-05,
      "loss": 0.6053,
      "mean_token_accuracy": 0.8135153502225876,
      "num_tokens": 658882328.0,
      "step": 20644
    },
    {
      "entropy": 0.5355602246709168,
      "epoch": 1.8994542872093891,
      "grad_norm": 0.15103387832641602,
      "learning_rate": 3.668844113227221e-05,
      "loss": 0.52,
      "mean_token_accuracy": 0.8355474658310413,
      "num_tokens": 658914865.0,
      "step": 20645
    },
    {
      "entropy": 0.5561314518563449,
      "epoch": 1.8995462935808305,
      "grad_norm": 0.14610819518566132,
      "learning_rate": 3.668537430613059e-05,
      "loss": 0.5397,
      "mean_token_accuracy": 0.8347977101802826,
      "num_tokens": 658946694.0,
      "step": 20646
    },
    {
      "entropy": 0.5952125675976276,
      "epoch": 1.8996382999522718,
      "grad_norm": 0.14731496572494507,
      "learning_rate": 3.668230747998896e-05,
      "loss": 0.5895,
      "mean_token_accuracy": 0.816066849976778,
      "num_tokens": 658978833.0,
      "step": 20647
    },
    {
      "entropy": 0.4543672427535057,
      "epoch": 1.899730306323713,
      "grad_norm": 0.14360125362873077,
      "learning_rate": 3.667924065384733e-05,
      "loss": 0.4454,
      "mean_token_accuracy": 0.8643513806164265,
      "num_tokens": 659009599.0,
      "step": 20648
    },
    {
      "entropy": 0.5461054155603051,
      "epoch": 1.8998223126951541,
      "grad_norm": 0.15150175988674164,
      "learning_rate": 3.667617382770571e-05,
      "loss": 0.5306,
      "mean_token_accuracy": 0.8399771563708782,
      "num_tokens": 659041158.0,
      "step": 20649
    },
    {
      "entropy": 0.5253457040525973,
      "epoch": 1.8999143190665952,
      "grad_norm": 0.141582652926445,
      "learning_rate": 3.667310700156408e-05,
      "loss": 0.524,
      "mean_token_accuracy": 0.8375244103372097,
      "num_tokens": 659073842.0,
      "step": 20650
    },
    {
      "entropy": 0.6037341915071011,
      "epoch": 1.9000063254380366,
      "grad_norm": 0.1536523848772049,
      "learning_rate": 3.667004017542246e-05,
      "loss": 0.5947,
      "mean_token_accuracy": 0.8180810660123825,
      "num_tokens": 659105508.0,
      "step": 20651
    },
    {
      "entropy": 0.5112867299467325,
      "epoch": 1.900098331809478,
      "grad_norm": 0.14839085936546326,
      "learning_rate": 3.666697334928083e-05,
      "loss": 0.4913,
      "mean_token_accuracy": 0.8466464318335056,
      "num_tokens": 659137845.0,
      "step": 20652
    },
    {
      "entropy": 0.6926206517964602,
      "epoch": 1.900190338180919,
      "grad_norm": 0.16569411754608154,
      "learning_rate": 3.6663906523139206e-05,
      "loss": 0.6848,
      "mean_token_accuracy": 0.7913794480264187,
      "num_tokens": 659168924.0,
      "step": 20653
    },
    {
      "entropy": 0.6296077384613454,
      "epoch": 1.9002823445523602,
      "grad_norm": 0.1521598994731903,
      "learning_rate": 3.666083969699758e-05,
      "loss": 0.6206,
      "mean_token_accuracy": 0.8109196238219738,
      "num_tokens": 659201361.0,
      "step": 20654
    },
    {
      "entropy": 0.6904521435499191,
      "epoch": 1.9003743509238014,
      "grad_norm": 0.1586548089981079,
      "learning_rate": 3.6657772870855956e-05,
      "loss": 0.6882,
      "mean_token_accuracy": 0.7914330959320068,
      "num_tokens": 659233394.0,
      "step": 20655
    },
    {
      "entropy": 0.6947105955332518,
      "epoch": 1.9004663572952427,
      "grad_norm": 0.15980742871761322,
      "learning_rate": 3.6654706044714323e-05,
      "loss": 0.696,
      "mean_token_accuracy": 0.7883718870580196,
      "num_tokens": 659265499.0,
      "step": 20656
    },
    {
      "entropy": 0.6438021454960108,
      "epoch": 1.900558363666684,
      "grad_norm": 0.16080279648303986,
      "learning_rate": 3.6651639218572705e-05,
      "loss": 0.631,
      "mean_token_accuracy": 0.8088885433971882,
      "num_tokens": 659297476.0,
      "step": 20657
    },
    {
      "entropy": 0.6416181586682796,
      "epoch": 1.9006503700381252,
      "grad_norm": 0.15341439843177795,
      "learning_rate": 3.664857239243107e-05,
      "loss": 0.6382,
      "mean_token_accuracy": 0.8005954697728157,
      "num_tokens": 659329534.0,
      "step": 20658
    },
    {
      "entropy": 0.6329493634402752,
      "epoch": 1.9007423764095663,
      "grad_norm": 0.15505559742450714,
      "learning_rate": 3.6645505566289454e-05,
      "loss": 0.6218,
      "mean_token_accuracy": 0.8111184276640415,
      "num_tokens": 659361390.0,
      "step": 20659
    },
    {
      "entropy": 0.6603962760418653,
      "epoch": 1.9008343827810075,
      "grad_norm": 0.16291923820972443,
      "learning_rate": 3.664243874014782e-05,
      "loss": 0.6379,
      "mean_token_accuracy": 0.8045843243598938,
      "num_tokens": 659393147.0,
      "step": 20660
    },
    {
      "entropy": 0.5856500398367643,
      "epoch": 1.9009263891524488,
      "grad_norm": 0.15249978005886078,
      "learning_rate": 3.66393719140062e-05,
      "loss": 0.5769,
      "mean_token_accuracy": 0.8229150772094727,
      "num_tokens": 659424900.0,
      "step": 20661
    },
    {
      "entropy": 0.5535024385899305,
      "epoch": 1.9010183955238902,
      "grad_norm": 0.14492355287075043,
      "learning_rate": 3.663630508786457e-05,
      "loss": 0.5343,
      "mean_token_accuracy": 0.83536496758461,
      "num_tokens": 659456666.0,
      "step": 20662
    },
    {
      "entropy": 0.556367427110672,
      "epoch": 1.9011104018953313,
      "grad_norm": 0.14870116114616394,
      "learning_rate": 3.6633238261722946e-05,
      "loss": 0.5427,
      "mean_token_accuracy": 0.8285738080739975,
      "num_tokens": 659488345.0,
      "step": 20663
    },
    {
      "entropy": 0.5273566003888845,
      "epoch": 1.9012024082667724,
      "grad_norm": 0.14718790352344513,
      "learning_rate": 3.663017143558132e-05,
      "loss": 0.514,
      "mean_token_accuracy": 0.8402637839317322,
      "num_tokens": 659520455.0,
      "step": 20664
    },
    {
      "entropy": 0.5807982143014669,
      "epoch": 1.9012944146382136,
      "grad_norm": 0.15069274604320526,
      "learning_rate": 3.6627104609439695e-05,
      "loss": 0.568,
      "mean_token_accuracy": 0.8238515108823776,
      "num_tokens": 659552114.0,
      "step": 20665
    },
    {
      "entropy": 0.613302806392312,
      "epoch": 1.901386421009655,
      "grad_norm": 0.16272146999835968,
      "learning_rate": 3.662403778329806e-05,
      "loss": 0.6196,
      "mean_token_accuracy": 0.8095227256417274,
      "num_tokens": 659583765.0,
      "step": 20666
    },
    {
      "entropy": 0.6213557980954647,
      "epoch": 1.9014784273810963,
      "grad_norm": 0.1603720635175705,
      "learning_rate": 3.6620970957156445e-05,
      "loss": 0.6149,
      "mean_token_accuracy": 0.8123243302106857,
      "num_tokens": 659614498.0,
      "step": 20667
    },
    {
      "entropy": 0.6320752631872892,
      "epoch": 1.9015704337525374,
      "grad_norm": 0.15560515224933624,
      "learning_rate": 3.661790413101481e-05,
      "loss": 0.6282,
      "mean_token_accuracy": 0.8079050406813622,
      "num_tokens": 659646269.0,
      "step": 20668
    },
    {
      "entropy": 0.5490120146423578,
      "epoch": 1.9016624401239786,
      "grad_norm": 0.14983250200748444,
      "learning_rate": 3.661483730487319e-05,
      "loss": 0.556,
      "mean_token_accuracy": 0.8296320885419846,
      "num_tokens": 659678240.0,
      "step": 20669
    },
    {
      "entropy": 0.5876326505094767,
      "epoch": 1.9017544464954197,
      "grad_norm": 0.15471157431602478,
      "learning_rate": 3.661177047873156e-05,
      "loss": 0.5833,
      "mean_token_accuracy": 0.8233029730618,
      "num_tokens": 659710428.0,
      "step": 20670
    },
    {
      "entropy": 0.5677379835397005,
      "epoch": 1.901846452866861,
      "grad_norm": 0.15011858940124512,
      "learning_rate": 3.6608703652589936e-05,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8231402672827244,
      "num_tokens": 659742246.0,
      "step": 20671
    },
    {
      "entropy": 0.7030194029211998,
      "epoch": 1.9019384592383024,
      "grad_norm": 0.16938209533691406,
      "learning_rate": 3.660563682644831e-05,
      "loss": 0.6928,
      "mean_token_accuracy": 0.7926535904407501,
      "num_tokens": 659773551.0,
      "step": 20672
    },
    {
      "entropy": 0.5123208449222147,
      "epoch": 1.9020304656097435,
      "grad_norm": 0.1427135020494461,
      "learning_rate": 3.6602570000306686e-05,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8480687513947487,
      "num_tokens": 659805590.0,
      "step": 20673
    },
    {
      "entropy": 0.6011566957458854,
      "epoch": 1.9021224719811847,
      "grad_norm": 0.15752023458480835,
      "learning_rate": 3.6599503174165054e-05,
      "loss": 0.5837,
      "mean_token_accuracy": 0.821788415312767,
      "num_tokens": 659837900.0,
      "step": 20674
    },
    {
      "entropy": 0.5692737204954028,
      "epoch": 1.9022144783526258,
      "grad_norm": 0.14706464111804962,
      "learning_rate": 3.6596436348023435e-05,
      "loss": 0.5608,
      "mean_token_accuracy": 0.828201524913311,
      "num_tokens": 659870229.0,
      "step": 20675
    },
    {
      "entropy": 0.5886263551656157,
      "epoch": 1.9023064847240672,
      "grad_norm": 0.1512494832277298,
      "learning_rate": 3.65933695218818e-05,
      "loss": 0.575,
      "mean_token_accuracy": 0.8222513757646084,
      "num_tokens": 659902362.0,
      "step": 20676
    },
    {
      "entropy": 0.567823420278728,
      "epoch": 1.9023984910955085,
      "grad_norm": 0.14927491545677185,
      "learning_rate": 3.659030269574018e-05,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8273273073136806,
      "num_tokens": 659934327.0,
      "step": 20677
    },
    {
      "entropy": 0.7373795006424189,
      "epoch": 1.9024904974669496,
      "grad_norm": 0.16123835742473602,
      "learning_rate": 3.658723586959855e-05,
      "loss": 0.7378,
      "mean_token_accuracy": 0.7768840603530407,
      "num_tokens": 659966837.0,
      "step": 20678
    },
    {
      "entropy": 0.6424840446561575,
      "epoch": 1.9025825038383908,
      "grad_norm": 0.15028497576713562,
      "learning_rate": 3.658416904345693e-05,
      "loss": 0.6268,
      "mean_token_accuracy": 0.8120489045977592,
      "num_tokens": 659998898.0,
      "step": 20679
    },
    {
      "entropy": 0.6448180838488042,
      "epoch": 1.902674510209832,
      "grad_norm": 0.1590096652507782,
      "learning_rate": 3.65811022173153e-05,
      "loss": 0.6354,
      "mean_token_accuracy": 0.8051482029259205,
      "num_tokens": 660030417.0,
      "step": 20680
    },
    {
      "entropy": 0.4547608639113605,
      "epoch": 1.9027665165812733,
      "grad_norm": 0.13626274466514587,
      "learning_rate": 3.6578035391173676e-05,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8614227995276451,
      "num_tokens": 660062812.0,
      "step": 20681
    },
    {
      "entropy": 0.65650365781039,
      "epoch": 1.9028585229527146,
      "grad_norm": 0.15867707133293152,
      "learning_rate": 3.657496856503205e-05,
      "loss": 0.6451,
      "mean_token_accuracy": 0.8053614757955074,
      "num_tokens": 660094564.0,
      "step": 20682
    },
    {
      "entropy": 0.6235310668125749,
      "epoch": 1.9029505293241558,
      "grad_norm": 0.15345096588134766,
      "learning_rate": 3.6571901738890425e-05,
      "loss": 0.6156,
      "mean_token_accuracy": 0.8084639385342598,
      "num_tokens": 660126881.0,
      "step": 20683
    },
    {
      "entropy": 0.6717018079943955,
      "epoch": 1.903042535695597,
      "grad_norm": 0.1627083718776703,
      "learning_rate": 3.65688349127488e-05,
      "loss": 0.6624,
      "mean_token_accuracy": 0.794872235506773,
      "num_tokens": 660158329.0,
      "step": 20684
    },
    {
      "entropy": 0.5326972482725978,
      "epoch": 1.903134542067038,
      "grad_norm": 0.147065207362175,
      "learning_rate": 3.6565768086607175e-05,
      "loss": 0.5135,
      "mean_token_accuracy": 0.8399796448647976,
      "num_tokens": 660190852.0,
      "step": 20685
    },
    {
      "entropy": 0.5514343548566103,
      "epoch": 1.9032265484384794,
      "grad_norm": 0.14574158191680908,
      "learning_rate": 3.656270126046555e-05,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8368417657911777,
      "num_tokens": 660222203.0,
      "step": 20686
    },
    {
      "entropy": 0.5378538286313415,
      "epoch": 1.9033185548099207,
      "grad_norm": 0.1456189602613449,
      "learning_rate": 3.655963443432392e-05,
      "loss": 0.5141,
      "mean_token_accuracy": 0.8356646969914436,
      "num_tokens": 660253681.0,
      "step": 20687
    },
    {
      "entropy": 0.5578897651284933,
      "epoch": 1.9034105611813619,
      "grad_norm": 0.1484718918800354,
      "learning_rate": 3.65565676081823e-05,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8314306475222111,
      "num_tokens": 660285659.0,
      "step": 20688
    },
    {
      "entropy": 0.5828864593058825,
      "epoch": 1.903502567552803,
      "grad_norm": 0.15906623005867004,
      "learning_rate": 3.6553500782040666e-05,
      "loss": 0.568,
      "mean_token_accuracy": 0.8206850253045559,
      "num_tokens": 660317511.0,
      "step": 20689
    },
    {
      "entropy": 0.6080256353598088,
      "epoch": 1.9035945739242441,
      "grad_norm": 0.15274332463741302,
      "learning_rate": 3.655043395589904e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.816921878606081,
      "num_tokens": 660349698.0,
      "step": 20690
    },
    {
      "entropy": 0.5122081632725894,
      "epoch": 1.9036865802956855,
      "grad_norm": 0.14502263069152832,
      "learning_rate": 3.6547367129757416e-05,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8421926125884056,
      "num_tokens": 660381061.0,
      "step": 20691
    },
    {
      "entropy": 0.5970144339371473,
      "epoch": 1.9037785866671268,
      "grad_norm": 0.17512939870357513,
      "learning_rate": 3.654430030361579e-05,
      "loss": 0.5987,
      "mean_token_accuracy": 0.8181188777089119,
      "num_tokens": 660413480.0,
      "step": 20692
    },
    {
      "entropy": 0.5831502266228199,
      "epoch": 1.903870593038568,
      "grad_norm": 0.15681038796901703,
      "learning_rate": 3.6541233477474165e-05,
      "loss": 0.574,
      "mean_token_accuracy": 0.8249235078692436,
      "num_tokens": 660445458.0,
      "step": 20693
    },
    {
      "entropy": 0.7357718329876661,
      "epoch": 1.9039625994100091,
      "grad_norm": 0.16524624824523926,
      "learning_rate": 3.653816665133254e-05,
      "loss": 0.738,
      "mean_token_accuracy": 0.783083364367485,
      "num_tokens": 660478092.0,
      "step": 20694
    },
    {
      "entropy": 0.7157076932489872,
      "epoch": 1.9040546057814502,
      "grad_norm": 0.16300219297409058,
      "learning_rate": 3.653509982519091e-05,
      "loss": 0.7067,
      "mean_token_accuracy": 0.7843960709869862,
      "num_tokens": 660509401.0,
      "step": 20695
    },
    {
      "entropy": 0.5687329545617104,
      "epoch": 1.9041466121528916,
      "grad_norm": 0.1488109529018402,
      "learning_rate": 3.653203299904929e-05,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8292624317109585,
      "num_tokens": 660540033.0,
      "step": 20696
    },
    {
      "entropy": 0.5617192341014743,
      "epoch": 1.904238618524333,
      "grad_norm": 0.15662935376167297,
      "learning_rate": 3.652896617290766e-05,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8276923820376396,
      "num_tokens": 660571630.0,
      "step": 20697
    },
    {
      "entropy": 0.6351551841944456,
      "epoch": 1.904330624895774,
      "grad_norm": 0.1556495875120163,
      "learning_rate": 3.652589934676603e-05,
      "loss": 0.625,
      "mean_token_accuracy": 0.8096549957990646,
      "num_tokens": 660604145.0,
      "step": 20698
    },
    {
      "entropy": 0.6621757531538606,
      "epoch": 1.9044226312672152,
      "grad_norm": 0.16031892597675323,
      "learning_rate": 3.6522832520624406e-05,
      "loss": 0.6479,
      "mean_token_accuracy": 0.8027179129421711,
      "num_tokens": 660636180.0,
      "step": 20699
    },
    {
      "entropy": 0.6600587135180831,
      "epoch": 1.9045146376386564,
      "grad_norm": 0.1523679494857788,
      "learning_rate": 3.651976569448278e-05,
      "loss": 0.6424,
      "mean_token_accuracy": 0.8064951971173286,
      "num_tokens": 660668407.0,
      "step": 20700
    },
    {
      "entropy": 0.6146310660988092,
      "epoch": 1.9046066440100977,
      "grad_norm": 0.15669173002243042,
      "learning_rate": 3.6516698868341155e-05,
      "loss": 0.6008,
      "mean_token_accuracy": 0.8150413557887077,
      "num_tokens": 660699864.0,
      "step": 20701
    },
    {
      "entropy": 0.5399672295898199,
      "epoch": 1.904698650381539,
      "grad_norm": 0.1481713503599167,
      "learning_rate": 3.651363204219953e-05,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8374328389763832,
      "num_tokens": 660732057.0,
      "step": 20702
    },
    {
      "entropy": 0.5755283539183438,
      "epoch": 1.9047906567529802,
      "grad_norm": 0.14631886780261993,
      "learning_rate": 3.6510565216057905e-05,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8290458396077156,
      "num_tokens": 660764178.0,
      "step": 20703
    },
    {
      "entropy": 0.6221381165087223,
      "epoch": 1.9048826631244213,
      "grad_norm": 0.15118475258350372,
      "learning_rate": 3.650749838991628e-05,
      "loss": 0.6065,
      "mean_token_accuracy": 0.8119752965867519,
      "num_tokens": 660796464.0,
      "step": 20704
    },
    {
      "entropy": 0.5736597869545221,
      "epoch": 1.9049746694958625,
      "grad_norm": 0.15819095075130463,
      "learning_rate": 3.650443156377465e-05,
      "loss": 0.5569,
      "mean_token_accuracy": 0.8227775059640408,
      "num_tokens": 660827767.0,
      "step": 20705
    },
    {
      "entropy": 0.5441023744642735,
      "epoch": 1.9050666758673038,
      "grad_norm": 0.14776651561260223,
      "learning_rate": 3.650136473763303e-05,
      "loss": 0.5236,
      "mean_token_accuracy": 0.8359231539070606,
      "num_tokens": 660859680.0,
      "step": 20706
    },
    {
      "entropy": 0.6618937011808157,
      "epoch": 1.9051586822387452,
      "grad_norm": 0.15246707201004028,
      "learning_rate": 3.6498297911491397e-05,
      "loss": 0.6544,
      "mean_token_accuracy": 0.7974699027836323,
      "num_tokens": 660891641.0,
      "step": 20707
    },
    {
      "entropy": 0.6600274667143822,
      "epoch": 1.9052506886101863,
      "grad_norm": 0.16135403513908386,
      "learning_rate": 3.649523108534977e-05,
      "loss": 0.6522,
      "mean_token_accuracy": 0.7970578446984291,
      "num_tokens": 660922984.0,
      "step": 20708
    },
    {
      "entropy": 0.585596589371562,
      "epoch": 1.9053426949816274,
      "grad_norm": 0.14864306151866913,
      "learning_rate": 3.6492164259208146e-05,
      "loss": 0.5666,
      "mean_token_accuracy": 0.8249100297689438,
      "num_tokens": 660954802.0,
      "step": 20709
    },
    {
      "entropy": 0.5879478100687265,
      "epoch": 1.9054347013530686,
      "grad_norm": 0.15778301656246185,
      "learning_rate": 3.648909743306652e-05,
      "loss": 0.5818,
      "mean_token_accuracy": 0.8213910236954689,
      "num_tokens": 660986468.0,
      "step": 20710
    },
    {
      "entropy": 0.6742158085107803,
      "epoch": 1.90552670772451,
      "grad_norm": 0.16355456411838531,
      "learning_rate": 3.6486030606924895e-05,
      "loss": 0.6598,
      "mean_token_accuracy": 0.7993843294680119,
      "num_tokens": 661017441.0,
      "step": 20711
    },
    {
      "entropy": 0.5672491025179625,
      "epoch": 1.9056187140959513,
      "grad_norm": 0.15535174310207367,
      "learning_rate": 3.648296378078327e-05,
      "loss": 0.5473,
      "mean_token_accuracy": 0.8297860957682133,
      "num_tokens": 661049375.0,
      "step": 20712
    },
    {
      "entropy": 0.5742563407402486,
      "epoch": 1.9057107204673924,
      "grad_norm": 0.16072697937488556,
      "learning_rate": 3.6479896954641644e-05,
      "loss": 0.5642,
      "mean_token_accuracy": 0.827045138925314,
      "num_tokens": 661079922.0,
      "step": 20713
    },
    {
      "entropy": 0.6488204495981336,
      "epoch": 1.9058027268388336,
      "grad_norm": 0.1551344096660614,
      "learning_rate": 3.647683012850002e-05,
      "loss": 0.6368,
      "mean_token_accuracy": 0.8042083159089088,
      "num_tokens": 661111918.0,
      "step": 20714
    },
    {
      "entropy": 0.568367863073945,
      "epoch": 1.9058947332102747,
      "grad_norm": 0.1577272266149521,
      "learning_rate": 3.6473763302358394e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8232818469405174,
      "num_tokens": 661143322.0,
      "step": 20715
    },
    {
      "entropy": 0.5031783208251,
      "epoch": 1.905986739581716,
      "grad_norm": 0.14773331582546234,
      "learning_rate": 3.647069647621676e-05,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8454015552997589,
      "num_tokens": 661175374.0,
      "step": 20716
    },
    {
      "entropy": 0.5823043123818934,
      "epoch": 1.9060787459531574,
      "grad_norm": 0.15332449972629547,
      "learning_rate": 3.646762965007514e-05,
      "loss": 0.5749,
      "mean_token_accuracy": 0.8279526680707932,
      "num_tokens": 661207634.0,
      "step": 20717
    },
    {
      "entropy": 0.4969023400917649,
      "epoch": 1.9061707523245985,
      "grad_norm": 0.14469441771507263,
      "learning_rate": 3.646456282393351e-05,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8481209240853786,
      "num_tokens": 661239648.0,
      "step": 20718
    },
    {
      "entropy": 0.6210073558613658,
      "epoch": 1.9062627586960397,
      "grad_norm": 0.15263481438159943,
      "learning_rate": 3.646149599779189e-05,
      "loss": 0.6115,
      "mean_token_accuracy": 0.8128578290343285,
      "num_tokens": 661271219.0,
      "step": 20719
    },
    {
      "entropy": 0.5953887607902288,
      "epoch": 1.9063547650674808,
      "grad_norm": 0.1461346596479416,
      "learning_rate": 3.645842917165026e-05,
      "loss": 0.5725,
      "mean_token_accuracy": 0.8232212737202644,
      "num_tokens": 661303602.0,
      "step": 20720
    },
    {
      "entropy": 0.6475308649241924,
      "epoch": 1.9064467714389222,
      "grad_norm": 0.1603531539440155,
      "learning_rate": 3.6455362345508635e-05,
      "loss": 0.6344,
      "mean_token_accuracy": 0.8023786656558514,
      "num_tokens": 661335561.0,
      "step": 20721
    },
    {
      "entropy": 0.6683901026844978,
      "epoch": 1.9065387778103635,
      "grad_norm": 0.16788670420646667,
      "learning_rate": 3.645229551936701e-05,
      "loss": 0.6683,
      "mean_token_accuracy": 0.7985051572322845,
      "num_tokens": 661367861.0,
      "step": 20722
    },
    {
      "entropy": 0.6816346067935228,
      "epoch": 1.9066307841818046,
      "grad_norm": 0.15687115490436554,
      "learning_rate": 3.6449228693225384e-05,
      "loss": 0.6701,
      "mean_token_accuracy": 0.796792957931757,
      "num_tokens": 661399979.0,
      "step": 20723
    },
    {
      "entropy": 0.5239997655153275,
      "epoch": 1.9067227905532458,
      "grad_norm": 0.14564134180545807,
      "learning_rate": 3.644616186708376e-05,
      "loss": 0.5191,
      "mean_token_accuracy": 0.8368933163583279,
      "num_tokens": 661432653.0,
      "step": 20724
    },
    {
      "entropy": 0.5996464090421796,
      "epoch": 1.906814796924687,
      "grad_norm": 0.15523388981819153,
      "learning_rate": 3.6443095040942133e-05,
      "loss": 0.589,
      "mean_token_accuracy": 0.8182047940790653,
      "num_tokens": 661464787.0,
      "step": 20725
    },
    {
      "entropy": 0.5948739424347878,
      "epoch": 1.9069068032961283,
      "grad_norm": 0.15587550401687622,
      "learning_rate": 3.64400282148005e-05,
      "loss": 0.5973,
      "mean_token_accuracy": 0.8165260441601276,
      "num_tokens": 661496868.0,
      "step": 20726
    },
    {
      "entropy": 0.5703207310289145,
      "epoch": 1.9069988096675696,
      "grad_norm": 0.1431388407945633,
      "learning_rate": 3.643696138865888e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8238419480621815,
      "num_tokens": 661529238.0,
      "step": 20727
    },
    {
      "entropy": 0.5147719085216522,
      "epoch": 1.9070908160390108,
      "grad_norm": 0.15045678615570068,
      "learning_rate": 3.643389456251725e-05,
      "loss": 0.502,
      "mean_token_accuracy": 0.8451012633740902,
      "num_tokens": 661561422.0,
      "step": 20728
    },
    {
      "entropy": 0.5446175318211317,
      "epoch": 1.907182822410452,
      "grad_norm": 0.14929194748401642,
      "learning_rate": 3.6430827736375625e-05,
      "loss": 0.5308,
      "mean_token_accuracy": 0.834351722151041,
      "num_tokens": 661593421.0,
      "step": 20729
    },
    {
      "entropy": 0.47889124508947134,
      "epoch": 1.907274828781893,
      "grad_norm": 0.14531297981739044,
      "learning_rate": 3.6427760910234e-05,
      "loss": 0.4729,
      "mean_token_accuracy": 0.8543171398341656,
      "num_tokens": 661625559.0,
      "step": 20730
    },
    {
      "entropy": 0.6025033770129085,
      "epoch": 1.9073668351533344,
      "grad_norm": 0.15299227833747864,
      "learning_rate": 3.6424694084092375e-05,
      "loss": 0.5882,
      "mean_token_accuracy": 0.8181763701140881,
      "num_tokens": 661657351.0,
      "step": 20731
    },
    {
      "entropy": 0.6100777573883533,
      "epoch": 1.9074588415247757,
      "grad_norm": 0.15540307760238647,
      "learning_rate": 3.642162725795075e-05,
      "loss": 0.5988,
      "mean_token_accuracy": 0.8095560297369957,
      "num_tokens": 661688501.0,
      "step": 20732
    },
    {
      "entropy": 0.5986000839620829,
      "epoch": 1.9075508478962169,
      "grad_norm": 0.15175263583660126,
      "learning_rate": 3.6418560431809124e-05,
      "loss": 0.5938,
      "mean_token_accuracy": 0.8164215572178364,
      "num_tokens": 661721124.0,
      "step": 20733
    },
    {
      "entropy": 0.57364715449512,
      "epoch": 1.907642854267658,
      "grad_norm": 0.1578178107738495,
      "learning_rate": 3.641549360566749e-05,
      "loss": 0.557,
      "mean_token_accuracy": 0.8273843042552471,
      "num_tokens": 661752833.0,
      "step": 20734
    },
    {
      "entropy": 0.4702920289710164,
      "epoch": 1.9077348606390991,
      "grad_norm": 0.14278081059455872,
      "learning_rate": 3.641242677952587e-05,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8547913134098053,
      "num_tokens": 661784978.0,
      "step": 20735
    },
    {
      "entropy": 0.679161760956049,
      "epoch": 1.9078268670105405,
      "grad_norm": 0.16216504573822021,
      "learning_rate": 3.640935995338424e-05,
      "loss": 0.6842,
      "mean_token_accuracy": 0.790106438100338,
      "num_tokens": 661816224.0,
      "step": 20736
    },
    {
      "entropy": 0.6430536415427923,
      "epoch": 1.9079188733819819,
      "grad_norm": 0.15668848156929016,
      "learning_rate": 3.6406293127242616e-05,
      "loss": 0.6362,
      "mean_token_accuracy": 0.8036785870790482,
      "num_tokens": 661848286.0,
      "step": 20737
    },
    {
      "entropy": 0.606288536451757,
      "epoch": 1.908010879753423,
      "grad_norm": 0.15735051035881042,
      "learning_rate": 3.640322630110099e-05,
      "loss": 0.6019,
      "mean_token_accuracy": 0.8199121132493019,
      "num_tokens": 661880750.0,
      "step": 20738
    },
    {
      "entropy": 0.7157243546098471,
      "epoch": 1.9081028861248641,
      "grad_norm": 0.16286148130893707,
      "learning_rate": 3.6400159474959365e-05,
      "loss": 0.7196,
      "mean_token_accuracy": 0.7786234505474567,
      "num_tokens": 661912077.0,
      "step": 20739
    },
    {
      "entropy": 0.695396775379777,
      "epoch": 1.9081948924963053,
      "grad_norm": 0.16151022911071777,
      "learning_rate": 3.639709264881774e-05,
      "loss": 0.6776,
      "mean_token_accuracy": 0.7912869788706303,
      "num_tokens": 661943464.0,
      "step": 20740
    },
    {
      "entropy": 0.6156779676675797,
      "epoch": 1.9082868988677466,
      "grad_norm": 0.1562749594449997,
      "learning_rate": 3.6394025822676114e-05,
      "loss": 0.603,
      "mean_token_accuracy": 0.8156570717692375,
      "num_tokens": 661975487.0,
      "step": 20741
    },
    {
      "entropy": 0.6582894325256348,
      "epoch": 1.908378905239188,
      "grad_norm": 0.15494908392429352,
      "learning_rate": 3.639095899653448e-05,
      "loss": 0.6388,
      "mean_token_accuracy": 0.8047960363328457,
      "num_tokens": 662007656.0,
      "step": 20742
    },
    {
      "entropy": 0.6093619363382459,
      "epoch": 1.908470911610629,
      "grad_norm": 0.15366263687610626,
      "learning_rate": 3.6387892170392864e-05,
      "loss": 0.5919,
      "mean_token_accuracy": 0.817861270159483,
      "num_tokens": 662040062.0,
      "step": 20743
    },
    {
      "entropy": 0.636895002797246,
      "epoch": 1.9085629179820702,
      "grad_norm": 0.15458227694034576,
      "learning_rate": 3.638482534425124e-05,
      "loss": 0.6161,
      "mean_token_accuracy": 0.808959349989891,
      "num_tokens": 662072271.0,
      "step": 20744
    },
    {
      "entropy": 0.6152244871482253,
      "epoch": 1.9086549243535114,
      "grad_norm": 0.1513753980398178,
      "learning_rate": 3.638175851810961e-05,
      "loss": 0.5944,
      "mean_token_accuracy": 0.8181249238550663,
      "num_tokens": 662104470.0,
      "step": 20745
    },
    {
      "entropy": 0.49500284250825644,
      "epoch": 1.9087469307249527,
      "grad_norm": 0.1434074193239212,
      "learning_rate": 3.637869169196799e-05,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8502294421195984,
      "num_tokens": 662135840.0,
      "step": 20746
    },
    {
      "entropy": 0.759597834199667,
      "epoch": 1.908838937096394,
      "grad_norm": 0.1609869748353958,
      "learning_rate": 3.6375624865826355e-05,
      "loss": 0.7479,
      "mean_token_accuracy": 0.7716591507196426,
      "num_tokens": 662167728.0,
      "step": 20747
    },
    {
      "entropy": 0.6643054932355881,
      "epoch": 1.9089309434678352,
      "grad_norm": 0.15835660696029663,
      "learning_rate": 3.637255803968474e-05,
      "loss": 0.6583,
      "mean_token_accuracy": 0.8018292337656021,
      "num_tokens": 662200125.0,
      "step": 20748
    },
    {
      "entropy": 0.5873083891347051,
      "epoch": 1.9090229498392763,
      "grad_norm": 0.15578241646289825,
      "learning_rate": 3.6369491213543105e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8213223665952682,
      "num_tokens": 662232099.0,
      "step": 20749
    },
    {
      "entropy": 0.7538289036601782,
      "epoch": 1.9091149562107175,
      "grad_norm": 0.17036183178424835,
      "learning_rate": 3.636642438740148e-05,
      "loss": 0.7402,
      "mean_token_accuracy": 0.7745286747813225,
      "num_tokens": 662263440.0,
      "step": 20750
    },
    {
      "entropy": 0.5630215471610427,
      "epoch": 1.9092069625821588,
      "grad_norm": 0.14950576424598694,
      "learning_rate": 3.6363357561259854e-05,
      "loss": 0.549,
      "mean_token_accuracy": 0.8290915824472904,
      "num_tokens": 662295358.0,
      "step": 20751
    },
    {
      "entropy": 0.6553228739649057,
      "epoch": 1.9092989689536002,
      "grad_norm": 0.16551573574543,
      "learning_rate": 3.636029073511823e-05,
      "loss": 0.65,
      "mean_token_accuracy": 0.7993959076702595,
      "num_tokens": 662326713.0,
      "step": 20752
    },
    {
      "entropy": 0.6528066676110029,
      "epoch": 1.9093909753250413,
      "grad_norm": 0.15152651071548462,
      "learning_rate": 3.63572239089766e-05,
      "loss": 0.6414,
      "mean_token_accuracy": 0.8025261722505093,
      "num_tokens": 662358974.0,
      "step": 20753
    },
    {
      "entropy": 0.6554160080850124,
      "epoch": 1.9094829816964825,
      "grad_norm": 0.16068878769874573,
      "learning_rate": 3.635415708283498e-05,
      "loss": 0.6513,
      "mean_token_accuracy": 0.7990160770714283,
      "num_tokens": 662391016.0,
      "step": 20754
    },
    {
      "entropy": 0.5852353116497397,
      "epoch": 1.9095749880679236,
      "grad_norm": 0.15413948893547058,
      "learning_rate": 3.6351090256693346e-05,
      "loss": 0.5605,
      "mean_token_accuracy": 0.8287800699472427,
      "num_tokens": 662423478.0,
      "step": 20755
    },
    {
      "entropy": 0.631497198715806,
      "epoch": 1.909666994439365,
      "grad_norm": 0.15745662152767181,
      "learning_rate": 3.634802343055173e-05,
      "loss": 0.6363,
      "mean_token_accuracy": 0.8084511607885361,
      "num_tokens": 662455603.0,
      "step": 20756
    },
    {
      "entropy": 0.5869254134595394,
      "epoch": 1.9097590008108063,
      "grad_norm": 0.1502452939748764,
      "learning_rate": 3.6344956604410095e-05,
      "loss": 0.5837,
      "mean_token_accuracy": 0.8194271512329578,
      "num_tokens": 662487830.0,
      "step": 20757
    },
    {
      "entropy": 0.5715118392836303,
      "epoch": 1.9098510071822474,
      "grad_norm": 0.14711102843284607,
      "learning_rate": 3.634188977826847e-05,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8313761316239834,
      "num_tokens": 662520351.0,
      "step": 20758
    },
    {
      "entropy": 0.6356975547969341,
      "epoch": 1.9099430135536886,
      "grad_norm": 0.16045771539211273,
      "learning_rate": 3.6338822952126844e-05,
      "loss": 0.631,
      "mean_token_accuracy": 0.809759434312582,
      "num_tokens": 662552836.0,
      "step": 20759
    },
    {
      "entropy": 0.4110408639535308,
      "epoch": 1.9100350199251297,
      "grad_norm": 0.13886524736881256,
      "learning_rate": 3.633575612598522e-05,
      "loss": 0.3933,
      "mean_token_accuracy": 0.8754980638623238,
      "num_tokens": 662585111.0,
      "step": 20760
    },
    {
      "entropy": 0.5493309972807765,
      "epoch": 1.910127026296571,
      "grad_norm": 0.14520613849163055,
      "learning_rate": 3.6332689299843594e-05,
      "loss": 0.5331,
      "mean_token_accuracy": 0.8327051252126694,
      "num_tokens": 662616716.0,
      "step": 20761
    },
    {
      "entropy": 0.6639711037278175,
      "epoch": 1.9102190326680124,
      "grad_norm": 0.1614563912153244,
      "learning_rate": 3.632962247370197e-05,
      "loss": 0.6553,
      "mean_token_accuracy": 0.8018145374953747,
      "num_tokens": 662649103.0,
      "step": 20762
    },
    {
      "entropy": 0.5523843737319112,
      "epoch": 1.9103110390394535,
      "grad_norm": 0.14725236594676971,
      "learning_rate": 3.632655564756034e-05,
      "loss": 0.5362,
      "mean_token_accuracy": 0.835797481238842,
      "num_tokens": 662680856.0,
      "step": 20763
    },
    {
      "entropy": 0.5479537267237902,
      "epoch": 1.9104030454108947,
      "grad_norm": 0.1438441276550293,
      "learning_rate": 3.632348882141872e-05,
      "loss": 0.5316,
      "mean_token_accuracy": 0.8361682780086994,
      "num_tokens": 662713181.0,
      "step": 20764
    },
    {
      "entropy": 0.5904450742527843,
      "epoch": 1.9104950517823358,
      "grad_norm": 0.1518143117427826,
      "learning_rate": 3.6320421995277085e-05,
      "loss": 0.5831,
      "mean_token_accuracy": 0.8168175183236599,
      "num_tokens": 662744733.0,
      "step": 20765
    },
    {
      "entropy": 0.6305539160966873,
      "epoch": 1.9105870581537772,
      "grad_norm": 0.15854214131832123,
      "learning_rate": 3.631735516913547e-05,
      "loss": 0.63,
      "mean_token_accuracy": 0.8072301410138607,
      "num_tokens": 662776864.0,
      "step": 20766
    },
    {
      "entropy": 0.6112588811665773,
      "epoch": 1.9106790645252185,
      "grad_norm": 0.15522409975528717,
      "learning_rate": 3.6314288342993835e-05,
      "loss": 0.6013,
      "mean_token_accuracy": 0.8153468072414398,
      "num_tokens": 662808977.0,
      "step": 20767
    },
    {
      "entropy": 0.5238819057121873,
      "epoch": 1.9107710708966597,
      "grad_norm": 0.14569780230522156,
      "learning_rate": 3.631122151685221e-05,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8412045128643513,
      "num_tokens": 662840498.0,
      "step": 20768
    },
    {
      "entropy": 0.5942734535783529,
      "epoch": 1.9108630772681008,
      "grad_norm": 0.15391118824481964,
      "learning_rate": 3.6308154690710584e-05,
      "loss": 0.5896,
      "mean_token_accuracy": 0.8198480606079102,
      "num_tokens": 662872846.0,
      "step": 20769
    },
    {
      "entropy": 0.65629150159657,
      "epoch": 1.910955083639542,
      "grad_norm": 0.15864552557468414,
      "learning_rate": 3.630508786456896e-05,
      "loss": 0.6536,
      "mean_token_accuracy": 0.7977937720716,
      "num_tokens": 662904500.0,
      "step": 20770
    },
    {
      "entropy": 0.5237622810527682,
      "epoch": 1.9110470900109833,
      "grad_norm": 0.1445273458957672,
      "learning_rate": 3.630202103842733e-05,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8421185873448849,
      "num_tokens": 662937196.0,
      "step": 20771
    },
    {
      "entropy": 0.5261124595999718,
      "epoch": 1.9111390963824246,
      "grad_norm": 0.14491596817970276,
      "learning_rate": 3.629895421228571e-05,
      "loss": 0.5228,
      "mean_token_accuracy": 0.835280392318964,
      "num_tokens": 662968546.0,
      "step": 20772
    },
    {
      "entropy": 0.5673379618674517,
      "epoch": 1.9112311027538658,
      "grad_norm": 0.14730232954025269,
      "learning_rate": 3.6295887386144076e-05,
      "loss": 0.5522,
      "mean_token_accuracy": 0.8298962935805321,
      "num_tokens": 663001015.0,
      "step": 20773
    },
    {
      "entropy": 0.6336454371921718,
      "epoch": 1.911323109125307,
      "grad_norm": 0.15564782917499542,
      "learning_rate": 3.629282056000246e-05,
      "loss": 0.6128,
      "mean_token_accuracy": 0.815350204706192,
      "num_tokens": 663032264.0,
      "step": 20774
    },
    {
      "entropy": 0.6032504104077816,
      "epoch": 1.911415115496748,
      "grad_norm": 0.15602374076843262,
      "learning_rate": 3.628975373386083e-05,
      "loss": 0.5897,
      "mean_token_accuracy": 0.8159161098301411,
      "num_tokens": 663064353.0,
      "step": 20775
    },
    {
      "entropy": 0.609914667904377,
      "epoch": 1.9115071218681894,
      "grad_norm": 0.15022800862789154,
      "learning_rate": 3.62866869077192e-05,
      "loss": 0.5905,
      "mean_token_accuracy": 0.8177641965448856,
      "num_tokens": 663096324.0,
      "step": 20776
    },
    {
      "entropy": 0.5802504159510136,
      "epoch": 1.9115991282396307,
      "grad_norm": 0.1510176956653595,
      "learning_rate": 3.628362008157758e-05,
      "loss": 0.5593,
      "mean_token_accuracy": 0.827518105506897,
      "num_tokens": 663128465.0,
      "step": 20777
    },
    {
      "entropy": 0.6368914386257529,
      "epoch": 1.9116911346110719,
      "grad_norm": 0.15132759511470795,
      "learning_rate": 3.628055325543595e-05,
      "loss": 0.6239,
      "mean_token_accuracy": 0.807159561663866,
      "num_tokens": 663160863.0,
      "step": 20778
    },
    {
      "entropy": 0.6001093257218599,
      "epoch": 1.911783140982513,
      "grad_norm": 0.1466401219367981,
      "learning_rate": 3.627748642929433e-05,
      "loss": 0.5843,
      "mean_token_accuracy": 0.817955557256937,
      "num_tokens": 663193463.0,
      "step": 20779
    },
    {
      "entropy": 0.6670964257791638,
      "epoch": 1.9118751473539541,
      "grad_norm": 0.15641635656356812,
      "learning_rate": 3.62744196031527e-05,
      "loss": 0.6571,
      "mean_token_accuracy": 0.7982362359762192,
      "num_tokens": 663225577.0,
      "step": 20780
    },
    {
      "entropy": 0.6197549924254417,
      "epoch": 1.9119671537253955,
      "grad_norm": 0.15688279271125793,
      "learning_rate": 3.627135277701107e-05,
      "loss": 0.6125,
      "mean_token_accuracy": 0.8123319298028946,
      "num_tokens": 663257573.0,
      "step": 20781
    },
    {
      "entropy": 0.5833896473050117,
      "epoch": 1.9120591600968369,
      "grad_norm": 0.15014530718326569,
      "learning_rate": 3.626828595086945e-05,
      "loss": 0.5748,
      "mean_token_accuracy": 0.8205228261649609,
      "num_tokens": 663289059.0,
      "step": 20782
    },
    {
      "entropy": 0.6474110689014196,
      "epoch": 1.912151166468278,
      "grad_norm": 0.16308628022670746,
      "learning_rate": 3.626521912472782e-05,
      "loss": 0.6333,
      "mean_token_accuracy": 0.8017594888806343,
      "num_tokens": 663320204.0,
      "step": 20783
    },
    {
      "entropy": 0.5932575864717364,
      "epoch": 1.9122431728397191,
      "grad_norm": 0.15068233013153076,
      "learning_rate": 3.62621522985862e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8229363709688187,
      "num_tokens": 663352639.0,
      "step": 20784
    },
    {
      "entropy": 0.6903835218399763,
      "epoch": 1.9123351792111603,
      "grad_norm": 0.16816960275173187,
      "learning_rate": 3.625908547244457e-05,
      "loss": 0.6943,
      "mean_token_accuracy": 0.7897027991712093,
      "num_tokens": 663384415.0,
      "step": 20785
    },
    {
      "entropy": 0.594719251152128,
      "epoch": 1.9124271855826016,
      "grad_norm": 0.15275117754936218,
      "learning_rate": 3.625601864630294e-05,
      "loss": 0.5736,
      "mean_token_accuracy": 0.821839552372694,
      "num_tokens": 663416266.0,
      "step": 20786
    },
    {
      "entropy": 0.5895698331296444,
      "epoch": 1.912519191954043,
      "grad_norm": 0.15319792926311493,
      "learning_rate": 3.625295182016132e-05,
      "loss": 0.5737,
      "mean_token_accuracy": 0.822042178362608,
      "num_tokens": 663447961.0,
      "step": 20787
    },
    {
      "entropy": 0.521204493008554,
      "epoch": 1.912611198325484,
      "grad_norm": 0.14164239168167114,
      "learning_rate": 3.624988499401969e-05,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8408844992518425,
      "num_tokens": 663479993.0,
      "step": 20788
    },
    {
      "entropy": 0.6037734467536211,
      "epoch": 1.9127032046969252,
      "grad_norm": 0.15483978390693665,
      "learning_rate": 3.6246818167878063e-05,
      "loss": 0.5895,
      "mean_token_accuracy": 0.8196497820317745,
      "num_tokens": 663511183.0,
      "step": 20789
    },
    {
      "entropy": 0.6406950573436916,
      "epoch": 1.9127952110683664,
      "grad_norm": 0.1583806574344635,
      "learning_rate": 3.624375134173644e-05,
      "loss": 0.6418,
      "mean_token_accuracy": 0.8063582144677639,
      "num_tokens": 663542808.0,
      "step": 20790
    },
    {
      "entropy": 0.6850486109033227,
      "epoch": 1.9128872174398077,
      "grad_norm": 0.16031908988952637,
      "learning_rate": 3.624068451559481e-05,
      "loss": 0.6798,
      "mean_token_accuracy": 0.7930860035121441,
      "num_tokens": 663574177.0,
      "step": 20791
    },
    {
      "entropy": 0.5772062949836254,
      "epoch": 1.912979223811249,
      "grad_norm": 0.15402601659297943,
      "learning_rate": 3.623761768945319e-05,
      "loss": 0.5751,
      "mean_token_accuracy": 0.8243962563574314,
      "num_tokens": 663605924.0,
      "step": 20792
    },
    {
      "entropy": 0.5439597126096487,
      "epoch": 1.9130712301826902,
      "grad_norm": 0.15370769798755646,
      "learning_rate": 3.623455086331156e-05,
      "loss": 0.5356,
      "mean_token_accuracy": 0.8324616551399231,
      "num_tokens": 663637444.0,
      "step": 20793
    },
    {
      "entropy": 0.6763143185526133,
      "epoch": 1.9131632365541313,
      "grad_norm": 0.17023012042045593,
      "learning_rate": 3.623148403716993e-05,
      "loss": 0.6715,
      "mean_token_accuracy": 0.7943612970411777,
      "num_tokens": 663667419.0,
      "step": 20794
    },
    {
      "entropy": 0.5223344508558512,
      "epoch": 1.9132552429255725,
      "grad_norm": 0.14668338000774384,
      "learning_rate": 3.622841721102831e-05,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8423492349684238,
      "num_tokens": 663699441.0,
      "step": 20795
    },
    {
      "entropy": 0.6157304476946592,
      "epoch": 1.9133472492970138,
      "grad_norm": 0.14966455101966858,
      "learning_rate": 3.622535038488668e-05,
      "loss": 0.6194,
      "mean_token_accuracy": 0.8133971281349659,
      "num_tokens": 663731188.0,
      "step": 20796
    },
    {
      "entropy": 0.5668303892016411,
      "epoch": 1.9134392556684552,
      "grad_norm": 0.14693887531757355,
      "learning_rate": 3.6222283558745054e-05,
      "loss": 0.5575,
      "mean_token_accuracy": 0.826265424489975,
      "num_tokens": 663762840.0,
      "step": 20797
    },
    {
      "entropy": 0.5965250339359045,
      "epoch": 1.9135312620398963,
      "grad_norm": 0.14798180758953094,
      "learning_rate": 3.621921673260343e-05,
      "loss": 0.5805,
      "mean_token_accuracy": 0.825855229049921,
      "num_tokens": 663795544.0,
      "step": 20798
    },
    {
      "entropy": 0.5564404963515699,
      "epoch": 1.9136232684113375,
      "grad_norm": 0.14517885446548462,
      "learning_rate": 3.62161499064618e-05,
      "loss": 0.5324,
      "mean_token_accuracy": 0.8387274108827114,
      "num_tokens": 663828312.0,
      "step": 20799
    },
    {
      "entropy": 0.6099802125245333,
      "epoch": 1.9137152747827786,
      "grad_norm": 0.15677228569984436,
      "learning_rate": 3.621308308032018e-05,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8184162862598896,
      "num_tokens": 663860124.0,
      "step": 20800
    },
    {
      "entropy": 0.5392262386158109,
      "epoch": 1.91380728115422,
      "grad_norm": 0.14580905437469482,
      "learning_rate": 3.621001625417855e-05,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8389629386365414,
      "num_tokens": 663892047.0,
      "step": 20801
    },
    {
      "entropy": 0.6115954276174307,
      "epoch": 1.9138992875256613,
      "grad_norm": 0.1548423320055008,
      "learning_rate": 3.620694942803692e-05,
      "loss": 0.5977,
      "mean_token_accuracy": 0.8150379583239555,
      "num_tokens": 663924044.0,
      "step": 20802
    },
    {
      "entropy": 0.4735211683437228,
      "epoch": 1.9139912938971024,
      "grad_norm": 0.14209584891796112,
      "learning_rate": 3.62038826018953e-05,
      "loss": 0.4588,
      "mean_token_accuracy": 0.8585540987551212,
      "num_tokens": 663956103.0,
      "step": 20803
    },
    {
      "entropy": 0.5419161915779114,
      "epoch": 1.9140833002685436,
      "grad_norm": 0.1463947296142578,
      "learning_rate": 3.6200815775753676e-05,
      "loss": 0.5352,
      "mean_token_accuracy": 0.837164118885994,
      "num_tokens": 663988379.0,
      "step": 20804
    },
    {
      "entropy": 0.4378137132152915,
      "epoch": 1.9141753066399847,
      "grad_norm": 0.137457937002182,
      "learning_rate": 3.619774894961205e-05,
      "loss": 0.4144,
      "mean_token_accuracy": 0.867561336606741,
      "num_tokens": 664020657.0,
      "step": 20805
    },
    {
      "entropy": 0.5800305800512433,
      "epoch": 1.914267313011426,
      "grad_norm": 0.14977627992630005,
      "learning_rate": 3.6194682123470426e-05,
      "loss": 0.5618,
      "mean_token_accuracy": 0.827593021094799,
      "num_tokens": 664052519.0,
      "step": 20806
    },
    {
      "entropy": 0.5357607088517398,
      "epoch": 1.9143593193828674,
      "grad_norm": 0.14797943830490112,
      "learning_rate": 3.6191615297328794e-05,
      "loss": 0.5224,
      "mean_token_accuracy": 0.839631162583828,
      "num_tokens": 664084436.0,
      "step": 20807
    },
    {
      "entropy": 0.5903576985001564,
      "epoch": 1.9144513257543085,
      "grad_norm": 0.1583079844713211,
      "learning_rate": 3.6188548471187175e-05,
      "loss": 0.5845,
      "mean_token_accuracy": 0.8209549598395824,
      "num_tokens": 664116565.0,
      "step": 20808
    },
    {
      "entropy": 0.5648666471242905,
      "epoch": 1.9145433321257497,
      "grad_norm": 0.15385295450687408,
      "learning_rate": 3.618548164504554e-05,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8279174938797951,
      "num_tokens": 664148599.0,
      "step": 20809
    },
    {
      "entropy": 0.484548585023731,
      "epoch": 1.9146353384971908,
      "grad_norm": 0.14743275940418243,
      "learning_rate": 3.618241481890392e-05,
      "loss": 0.4758,
      "mean_token_accuracy": 0.8479429706931114,
      "num_tokens": 664180255.0,
      "step": 20810
    },
    {
      "entropy": 0.584478129632771,
      "epoch": 1.9147273448686322,
      "grad_norm": 0.15095554292201996,
      "learning_rate": 3.617934799276229e-05,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8245970271527767,
      "num_tokens": 664212051.0,
      "step": 20811
    },
    {
      "entropy": 0.5211787805892527,
      "epoch": 1.9148193512400735,
      "grad_norm": 0.15079954266548157,
      "learning_rate": 3.617628116662067e-05,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8404637090861797,
      "num_tokens": 664244062.0,
      "step": 20812
    },
    {
      "entropy": 0.555401898920536,
      "epoch": 1.9149113576115147,
      "grad_norm": 0.14987237751483917,
      "learning_rate": 3.617321434047904e-05,
      "loss": 0.5469,
      "mean_token_accuracy": 0.8284975849092007,
      "num_tokens": 664276051.0,
      "step": 20813
    },
    {
      "entropy": 0.6222527623176575,
      "epoch": 1.9150033639829558,
      "grad_norm": 0.16163809597492218,
      "learning_rate": 3.6170147514337416e-05,
      "loss": 0.6089,
      "mean_token_accuracy": 0.8151163496077061,
      "num_tokens": 664307301.0,
      "step": 20814
    },
    {
      "entropy": 0.5762091763317585,
      "epoch": 1.915095370354397,
      "grad_norm": 0.14883089065551758,
      "learning_rate": 3.6167080688195784e-05,
      "loss": 0.5684,
      "mean_token_accuracy": 0.8247692845761776,
      "num_tokens": 664339557.0,
      "step": 20815
    },
    {
      "entropy": 0.5654682647436857,
      "epoch": 1.9151873767258383,
      "grad_norm": 0.1499491184949875,
      "learning_rate": 3.6164013862054165e-05,
      "loss": 0.5528,
      "mean_token_accuracy": 0.8234684988856316,
      "num_tokens": 664372061.0,
      "step": 20816
    },
    {
      "entropy": 0.6172684691846371,
      "epoch": 1.9152793830972796,
      "grad_norm": 0.1523122489452362,
      "learning_rate": 3.616094703591253e-05,
      "loss": 0.606,
      "mean_token_accuracy": 0.8130214065313339,
      "num_tokens": 664403520.0,
      "step": 20817
    },
    {
      "entropy": 0.5584589205682278,
      "epoch": 1.9153713894687208,
      "grad_norm": 0.1490156054496765,
      "learning_rate": 3.615788020977091e-05,
      "loss": 0.5417,
      "mean_token_accuracy": 0.8310836590826511,
      "num_tokens": 664435226.0,
      "step": 20818
    },
    {
      "entropy": 0.5167499617673457,
      "epoch": 1.915463395840162,
      "grad_norm": 0.14620670676231384,
      "learning_rate": 3.615481338362928e-05,
      "loss": 0.5148,
      "mean_token_accuracy": 0.839331604540348,
      "num_tokens": 664467692.0,
      "step": 20819
    },
    {
      "entropy": 0.5594063401222229,
      "epoch": 1.915555402211603,
      "grad_norm": 0.164520725607872,
      "learning_rate": 3.615174655748766e-05,
      "loss": 0.5491,
      "mean_token_accuracy": 0.8370840772986412,
      "num_tokens": 664499005.0,
      "step": 20820
    },
    {
      "entropy": 0.5210023587569594,
      "epoch": 1.9156474085830444,
      "grad_norm": 0.1468697190284729,
      "learning_rate": 3.614867973134603e-05,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8410505689680576,
      "num_tokens": 664531236.0,
      "step": 20821
    },
    {
      "entropy": 0.6471519060432911,
      "epoch": 1.9157394149544857,
      "grad_norm": 0.15362541377544403,
      "learning_rate": 3.6145612905204407e-05,
      "loss": 0.6431,
      "mean_token_accuracy": 0.7996760830283165,
      "num_tokens": 664563767.0,
      "step": 20822
    },
    {
      "entropy": 0.6118771303445101,
      "epoch": 1.9158314213259269,
      "grad_norm": 0.15682508051395416,
      "learning_rate": 3.614254607906278e-05,
      "loss": 0.6066,
      "mean_token_accuracy": 0.8146712332963943,
      "num_tokens": 664595630.0,
      "step": 20823
    },
    {
      "entropy": 0.5934860212728381,
      "epoch": 1.915923427697368,
      "grad_norm": 0.15247978270053864,
      "learning_rate": 3.6139479252921156e-05,
      "loss": 0.5954,
      "mean_token_accuracy": 0.8160737603902817,
      "num_tokens": 664627935.0,
      "step": 20824
    },
    {
      "entropy": 0.5994682405143976,
      "epoch": 1.9160154340688091,
      "grad_norm": 0.15167376399040222,
      "learning_rate": 3.6136412426779524e-05,
      "loss": 0.5942,
      "mean_token_accuracy": 0.8168532848358154,
      "num_tokens": 664660088.0,
      "step": 20825
    },
    {
      "entropy": 0.5822318345308304,
      "epoch": 1.9161074404402505,
      "grad_norm": 0.15277962386608124,
      "learning_rate": 3.6133345600637905e-05,
      "loss": 0.5723,
      "mean_token_accuracy": 0.825227040797472,
      "num_tokens": 664691779.0,
      "step": 20826
    },
    {
      "entropy": 0.6235121181234717,
      "epoch": 1.9161994468116919,
      "grad_norm": 0.15359549224376678,
      "learning_rate": 3.613027877449627e-05,
      "loss": 0.621,
      "mean_token_accuracy": 0.8126099407672882,
      "num_tokens": 664724008.0,
      "step": 20827
    },
    {
      "entropy": 0.6732888985425234,
      "epoch": 1.916291453183133,
      "grad_norm": 0.15210644900798798,
      "learning_rate": 3.612721194835465e-05,
      "loss": 0.6675,
      "mean_token_accuracy": 0.7988726012408733,
      "num_tokens": 664755756.0,
      "step": 20828
    },
    {
      "entropy": 0.5079360804520547,
      "epoch": 1.9163834595545741,
      "grad_norm": 0.14227916300296783,
      "learning_rate": 3.612414512221302e-05,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8498639203608036,
      "num_tokens": 664787779.0,
      "step": 20829
    },
    {
      "entropy": 0.5886639263480902,
      "epoch": 1.9164754659260153,
      "grad_norm": 0.16169053316116333,
      "learning_rate": 3.61210782960714e-05,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8246727995574474,
      "num_tokens": 664818329.0,
      "step": 20830
    },
    {
      "entropy": 0.6927664789836854,
      "epoch": 1.9165674722974566,
      "grad_norm": 0.15622802078723907,
      "learning_rate": 3.611801146992977e-05,
      "loss": 0.676,
      "mean_token_accuracy": 0.7936620712280273,
      "num_tokens": 664850878.0,
      "step": 20831
    },
    {
      "entropy": 0.638501139357686,
      "epoch": 1.916659478668898,
      "grad_norm": 0.16090695559978485,
      "learning_rate": 3.6114944643788146e-05,
      "loss": 0.6144,
      "mean_token_accuracy": 0.8079972229897976,
      "num_tokens": 664881880.0,
      "step": 20832
    },
    {
      "entropy": 0.5365561787039042,
      "epoch": 1.916751485040339,
      "grad_norm": 0.15388169884681702,
      "learning_rate": 3.6111877817646514e-05,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8421596437692642,
      "num_tokens": 664913684.0,
      "step": 20833
    },
    {
      "entropy": 0.6457265531644225,
      "epoch": 1.9168434914117802,
      "grad_norm": 0.15416528284549713,
      "learning_rate": 3.6108810991504896e-05,
      "loss": 0.6274,
      "mean_token_accuracy": 0.8041187301278114,
      "num_tokens": 664945404.0,
      "step": 20834
    },
    {
      "entropy": 0.5504526002332568,
      "epoch": 1.9169354977832214,
      "grad_norm": 0.1464005708694458,
      "learning_rate": 3.610574416536327e-05,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8346189707517624,
      "num_tokens": 664977770.0,
      "step": 20835
    },
    {
      "entropy": 0.6300287414342165,
      "epoch": 1.9170275041546627,
      "grad_norm": 0.15640173852443695,
      "learning_rate": 3.610267733922164e-05,
      "loss": 0.6031,
      "mean_token_accuracy": 0.814479649066925,
      "num_tokens": 665009727.0,
      "step": 20836
    },
    {
      "entropy": 0.5597592638805509,
      "epoch": 1.917119510526104,
      "grad_norm": 0.15616652369499207,
      "learning_rate": 3.609961051308002e-05,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8316142037510872,
      "num_tokens": 665040634.0,
      "step": 20837
    },
    {
      "entropy": 0.6425102506764233,
      "epoch": 1.9172115168975452,
      "grad_norm": 0.15233981609344482,
      "learning_rate": 3.609654368693839e-05,
      "loss": 0.636,
      "mean_token_accuracy": 0.8078685253858566,
      "num_tokens": 665073130.0,
      "step": 20838
    },
    {
      "entropy": 0.6230375533923507,
      "epoch": 1.9173035232689863,
      "grad_norm": 0.15628179907798767,
      "learning_rate": 3.609347686079677e-05,
      "loss": 0.6229,
      "mean_token_accuracy": 0.8081874884665012,
      "num_tokens": 665105331.0,
      "step": 20839
    },
    {
      "entropy": 0.5988241732120514,
      "epoch": 1.9173955296404275,
      "grad_norm": 0.15163782238960266,
      "learning_rate": 3.609041003465514e-05,
      "loss": 0.5778,
      "mean_token_accuracy": 0.822765376418829,
      "num_tokens": 665136461.0,
      "step": 20840
    },
    {
      "entropy": 0.6710358895361423,
      "epoch": 1.9174875360118688,
      "grad_norm": 0.1551017314195633,
      "learning_rate": 3.608734320851351e-05,
      "loss": 0.6604,
      "mean_token_accuracy": 0.7994733825325966,
      "num_tokens": 665168602.0,
      "step": 20841
    },
    {
      "entropy": 0.5572140263393521,
      "epoch": 1.9175795423833102,
      "grad_norm": 0.1527174413204193,
      "learning_rate": 3.6084276382371886e-05,
      "loss": 0.5437,
      "mean_token_accuracy": 0.8339335955679417,
      "num_tokens": 665200407.0,
      "step": 20842
    },
    {
      "entropy": 0.600748055614531,
      "epoch": 1.9176715487547513,
      "grad_norm": 0.15388481318950653,
      "learning_rate": 3.608120955623026e-05,
      "loss": 0.5945,
      "mean_token_accuracy": 0.819882545620203,
      "num_tokens": 665232675.0,
      "step": 20843
    },
    {
      "entropy": 0.6247436366975307,
      "epoch": 1.9177635551261925,
      "grad_norm": 0.15674136579036713,
      "learning_rate": 3.6078142730088635e-05,
      "loss": 0.6329,
      "mean_token_accuracy": 0.8038428388535976,
      "num_tokens": 665264676.0,
      "step": 20844
    },
    {
      "entropy": 0.5453500698786229,
      "epoch": 1.9178555614976336,
      "grad_norm": 0.15194694697856903,
      "learning_rate": 3.607507590394701e-05,
      "loss": 0.5341,
      "mean_token_accuracy": 0.8337279595434666,
      "num_tokens": 665295736.0,
      "step": 20845
    },
    {
      "entropy": 0.5042762961238623,
      "epoch": 1.917947567869075,
      "grad_norm": 0.14374405145645142,
      "learning_rate": 3.607200907780538e-05,
      "loss": 0.5029,
      "mean_token_accuracy": 0.8447401113808155,
      "num_tokens": 665327989.0,
      "step": 20846
    },
    {
      "entropy": 0.6180787440389395,
      "epoch": 1.9180395742405163,
      "grad_norm": 0.15728306770324707,
      "learning_rate": 3.606894225166376e-05,
      "loss": 0.6106,
      "mean_token_accuracy": 0.8096323385834694,
      "num_tokens": 665359532.0,
      "step": 20847
    },
    {
      "entropy": 0.578242214396596,
      "epoch": 1.9181315806119574,
      "grad_norm": 0.15148413181304932,
      "learning_rate": 3.606587542552213e-05,
      "loss": 0.573,
      "mean_token_accuracy": 0.8257750682532787,
      "num_tokens": 665391619.0,
      "step": 20848
    },
    {
      "entropy": 0.5979071240872145,
      "epoch": 1.9182235869833986,
      "grad_norm": 0.15478476881980896,
      "learning_rate": 3.60628085993805e-05,
      "loss": 0.5984,
      "mean_token_accuracy": 0.8118546344339848,
      "num_tokens": 665424363.0,
      "step": 20849
    },
    {
      "entropy": 0.6030234852805734,
      "epoch": 1.9183155933548397,
      "grad_norm": 0.1589181274175644,
      "learning_rate": 3.6059741773238876e-05,
      "loss": 0.5926,
      "mean_token_accuracy": 0.8154468163847923,
      "num_tokens": 665456210.0,
      "step": 20850
    },
    {
      "entropy": 0.5822359211742878,
      "epoch": 1.918407599726281,
      "grad_norm": 0.15150132775306702,
      "learning_rate": 3.605667494709725e-05,
      "loss": 0.5795,
      "mean_token_accuracy": 0.8200784213840961,
      "num_tokens": 665488615.0,
      "step": 20851
    },
    {
      "entropy": 0.6769286561757326,
      "epoch": 1.9184996060977224,
      "grad_norm": 0.1618957370519638,
      "learning_rate": 3.6053608120955626e-05,
      "loss": 0.6647,
      "mean_token_accuracy": 0.7985028922557831,
      "num_tokens": 665520953.0,
      "step": 20852
    },
    {
      "entropy": 0.538587773218751,
      "epoch": 1.9185916124691635,
      "grad_norm": 0.1470111459493637,
      "learning_rate": 3.6050541294814e-05,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8358744382858276,
      "num_tokens": 665552913.0,
      "step": 20853
    },
    {
      "entropy": 0.6245228033512831,
      "epoch": 1.9186836188406047,
      "grad_norm": 0.1559908241033554,
      "learning_rate": 3.604747446867237e-05,
      "loss": 0.6176,
      "mean_token_accuracy": 0.8083319701254368,
      "num_tokens": 665585542.0,
      "step": 20854
    },
    {
      "entropy": 0.6549681276082993,
      "epoch": 1.9187756252120458,
      "grad_norm": 0.15899206697940826,
      "learning_rate": 3.604440764253075e-05,
      "loss": 0.6512,
      "mean_token_accuracy": 0.7954369001090527,
      "num_tokens": 665617693.0,
      "step": 20855
    },
    {
      "entropy": 0.6017654575407505,
      "epoch": 1.9188676315834872,
      "grad_norm": 0.15171384811401367,
      "learning_rate": 3.604134081638912e-05,
      "loss": 0.5976,
      "mean_token_accuracy": 0.8151166662573814,
      "num_tokens": 665649547.0,
      "step": 20856
    },
    {
      "entropy": 0.5798635389655828,
      "epoch": 1.9189596379549285,
      "grad_norm": 0.14947816729545593,
      "learning_rate": 3.603827399024749e-05,
      "loss": 0.5703,
      "mean_token_accuracy": 0.8234479315578938,
      "num_tokens": 665681816.0,
      "step": 20857
    },
    {
      "entropy": 0.6676889760419726,
      "epoch": 1.9190516443263697,
      "grad_norm": 0.16418248414993286,
      "learning_rate": 3.603520716410587e-05,
      "loss": 0.6554,
      "mean_token_accuracy": 0.8023300431668758,
      "num_tokens": 665713904.0,
      "step": 20858
    },
    {
      "entropy": 0.5741744562983513,
      "epoch": 1.9191436506978108,
      "grad_norm": 0.1576007455587387,
      "learning_rate": 3.603214033796424e-05,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8273701220750809,
      "num_tokens": 665745598.0,
      "step": 20859
    },
    {
      "entropy": 0.5817370871081948,
      "epoch": 1.919235657069252,
      "grad_norm": 0.14885838329792023,
      "learning_rate": 3.6029073511822616e-05,
      "loss": 0.5802,
      "mean_token_accuracy": 0.8251231983304024,
      "num_tokens": 665777508.0,
      "step": 20860
    },
    {
      "entropy": 0.576451925560832,
      "epoch": 1.9193276634406933,
      "grad_norm": 0.1474623680114746,
      "learning_rate": 3.602600668568099e-05,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8310559429228306,
      "num_tokens": 665809495.0,
      "step": 20861
    },
    {
      "entropy": 0.6478750435635448,
      "epoch": 1.9194196698121346,
      "grad_norm": 0.15215981006622314,
      "learning_rate": 3.602293985953936e-05,
      "loss": 0.6304,
      "mean_token_accuracy": 0.8103627189993858,
      "num_tokens": 665842263.0,
      "step": 20862
    },
    {
      "entropy": 0.6290976698510349,
      "epoch": 1.9195116761835758,
      "grad_norm": 0.15466485917568207,
      "learning_rate": 3.601987303339774e-05,
      "loss": 0.6105,
      "mean_token_accuracy": 0.813022967427969,
      "num_tokens": 665873869.0,
      "step": 20863
    },
    {
      "entropy": 0.5891646640375257,
      "epoch": 1.919603682555017,
      "grad_norm": 0.15686115622520447,
      "learning_rate": 3.601680620725611e-05,
      "loss": 0.5775,
      "mean_token_accuracy": 0.8197301216423512,
      "num_tokens": 665905636.0,
      "step": 20864
    },
    {
      "entropy": 0.5423522510100156,
      "epoch": 1.919695688926458,
      "grad_norm": 0.1490565538406372,
      "learning_rate": 3.601373938111449e-05,
      "loss": 0.5388,
      "mean_token_accuracy": 0.8327115513384342,
      "num_tokens": 665937832.0,
      "step": 20865
    },
    {
      "entropy": 0.6567023270763457,
      "epoch": 1.9197876952978994,
      "grad_norm": 0.15999604761600494,
      "learning_rate": 3.6010672554972864e-05,
      "loss": 0.6456,
      "mean_token_accuracy": 0.805592492222786,
      "num_tokens": 665968380.0,
      "step": 20866
    },
    {
      "entropy": 0.585979051887989,
      "epoch": 1.9198797016693407,
      "grad_norm": 0.15454179048538208,
      "learning_rate": 3.600760572883123e-05,
      "loss": 0.5711,
      "mean_token_accuracy": 0.8222269974648952,
      "num_tokens": 665999862.0,
      "step": 20867
    },
    {
      "entropy": 0.6508873570710421,
      "epoch": 1.9199717080407819,
      "grad_norm": 0.15713833272457123,
      "learning_rate": 3.600453890268961e-05,
      "loss": 0.6417,
      "mean_token_accuracy": 0.8056985847651958,
      "num_tokens": 666031563.0,
      "step": 20868
    },
    {
      "entropy": 0.6214977204799652,
      "epoch": 1.920063714412223,
      "grad_norm": 0.15949735045433044,
      "learning_rate": 3.600147207654798e-05,
      "loss": 0.609,
      "mean_token_accuracy": 0.8099955096840858,
      "num_tokens": 666062886.0,
      "step": 20869
    },
    {
      "entropy": 0.701680576428771,
      "epoch": 1.9201557207836641,
      "grad_norm": 0.16555646061897278,
      "learning_rate": 3.5998405250406356e-05,
      "loss": 0.6782,
      "mean_token_accuracy": 0.7933143600821495,
      "num_tokens": 666093776.0,
      "step": 20870
    },
    {
      "entropy": 0.646352575160563,
      "epoch": 1.9202477271551055,
      "grad_norm": 0.15862281620502472,
      "learning_rate": 3.599533842426473e-05,
      "loss": 0.623,
      "mean_token_accuracy": 0.8053168579936028,
      "num_tokens": 666124999.0,
      "step": 20871
    },
    {
      "entropy": 0.6756944507360458,
      "epoch": 1.9203397335265469,
      "grad_norm": 0.1566648632287979,
      "learning_rate": 3.5992271598123105e-05,
      "loss": 0.6631,
      "mean_token_accuracy": 0.7989233322441578,
      "num_tokens": 666157348.0,
      "step": 20872
    },
    {
      "entropy": 0.5201394855976105,
      "epoch": 1.920431739897988,
      "grad_norm": 0.14971667528152466,
      "learning_rate": 3.598920477198148e-05,
      "loss": 0.5101,
      "mean_token_accuracy": 0.838461022824049,
      "num_tokens": 666189595.0,
      "step": 20873
    },
    {
      "entropy": 0.5600634915754199,
      "epoch": 1.9205237462694291,
      "grad_norm": 0.15093764662742615,
      "learning_rate": 3.5986137945839854e-05,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8268038593232632,
      "num_tokens": 666221572.0,
      "step": 20874
    },
    {
      "entropy": 0.7005637977272272,
      "epoch": 1.9206157526408703,
      "grad_norm": 0.16756968200206757,
      "learning_rate": 3.598307111969822e-05,
      "loss": 0.6973,
      "mean_token_accuracy": 0.7876463048160076,
      "num_tokens": 666253569.0,
      "step": 20875
    },
    {
      "entropy": 0.5839880704879761,
      "epoch": 1.9207077590123116,
      "grad_norm": 0.15065129101276398,
      "learning_rate": 3.5980004293556604e-05,
      "loss": 0.5935,
      "mean_token_accuracy": 0.8226796351373196,
      "num_tokens": 666286069.0,
      "step": 20876
    },
    {
      "entropy": 0.47559571266174316,
      "epoch": 1.920799765383753,
      "grad_norm": 0.14464011788368225,
      "learning_rate": 3.597693746741497e-05,
      "loss": 0.4654,
      "mean_token_accuracy": 0.854029692709446,
      "num_tokens": 666318035.0,
      "step": 20877
    },
    {
      "entropy": 0.6650223601609468,
      "epoch": 1.920891771755194,
      "grad_norm": 0.1571066975593567,
      "learning_rate": 3.5973870641273346e-05,
      "loss": 0.6714,
      "mean_token_accuracy": 0.7970026694238186,
      "num_tokens": 666350692.0,
      "step": 20878
    },
    {
      "entropy": 0.5471287686377764,
      "epoch": 1.9209837781266352,
      "grad_norm": 0.1530984342098236,
      "learning_rate": 3.597080381513172e-05,
      "loss": 0.5569,
      "mean_token_accuracy": 0.8265293501317501,
      "num_tokens": 666382832.0,
      "step": 20879
    },
    {
      "entropy": 0.6977034322917461,
      "epoch": 1.9210757844980764,
      "grad_norm": 0.15585088729858398,
      "learning_rate": 3.5967736988990095e-05,
      "loss": 0.6927,
      "mean_token_accuracy": 0.7870481722056866,
      "num_tokens": 666415005.0,
      "step": 20880
    },
    {
      "entropy": 0.6706019099801779,
      "epoch": 1.9211677908695177,
      "grad_norm": 0.1578443944454193,
      "learning_rate": 3.596467016284847e-05,
      "loss": 0.657,
      "mean_token_accuracy": 0.8017650209367275,
      "num_tokens": 666447161.0,
      "step": 20881
    },
    {
      "entropy": 0.703808257356286,
      "epoch": 1.921259797240959,
      "grad_norm": 0.16943170130252838,
      "learning_rate": 3.5961603336706845e-05,
      "loss": 0.7072,
      "mean_token_accuracy": 0.7887771762907505,
      "num_tokens": 666478671.0,
      "step": 20882
    },
    {
      "entropy": 0.6725028920918703,
      "epoch": 1.9213518036124002,
      "grad_norm": 0.15717479586601257,
      "learning_rate": 3.595853651056522e-05,
      "loss": 0.651,
      "mean_token_accuracy": 0.8014010861515999,
      "num_tokens": 666510994.0,
      "step": 20883
    },
    {
      "entropy": 0.622589779086411,
      "epoch": 1.9214438099838413,
      "grad_norm": 0.14812105894088745,
      "learning_rate": 3.5955469684423594e-05,
      "loss": 0.6103,
      "mean_token_accuracy": 0.8159330002963543,
      "num_tokens": 666543353.0,
      "step": 20884
    },
    {
      "entropy": 0.5718463053926826,
      "epoch": 1.9215358163552825,
      "grad_norm": 0.14987429976463318,
      "learning_rate": 3.595240285828196e-05,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8242249637842178,
      "num_tokens": 666575407.0,
      "step": 20885
    },
    {
      "entropy": 0.6526988614350557,
      "epoch": 1.9216278227267238,
      "grad_norm": 0.15738238394260406,
      "learning_rate": 3.594933603214034e-05,
      "loss": 0.6325,
      "mean_token_accuracy": 0.8063186667859554,
      "num_tokens": 666606899.0,
      "step": 20886
    },
    {
      "entropy": 0.6425107177346945,
      "epoch": 1.9217198290981652,
      "grad_norm": 0.1545429229736328,
      "learning_rate": 3.594626920599871e-05,
      "loss": 0.619,
      "mean_token_accuracy": 0.8093093484640121,
      "num_tokens": 666639138.0,
      "step": 20887
    },
    {
      "entropy": 0.5956533951684833,
      "epoch": 1.9218118354696063,
      "grad_norm": 0.14720606803894043,
      "learning_rate": 3.5943202379857086e-05,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8253291882574558,
      "num_tokens": 666671505.0,
      "step": 20888
    },
    {
      "entropy": 0.6075662281364202,
      "epoch": 1.9219038418410475,
      "grad_norm": 0.16209588944911957,
      "learning_rate": 3.594013555371546e-05,
      "loss": 0.5895,
      "mean_token_accuracy": 0.8220278918743134,
      "num_tokens": 666703758.0,
      "step": 20889
    },
    {
      "entropy": 0.4659623019397259,
      "epoch": 1.9219958482124886,
      "grad_norm": 0.15066459774971008,
      "learning_rate": 3.5937068727573835e-05,
      "loss": 0.4603,
      "mean_token_accuracy": 0.8581283800303936,
      "num_tokens": 666734692.0,
      "step": 20890
    },
    {
      "entropy": 0.5871344078332186,
      "epoch": 1.92208785458393,
      "grad_norm": 0.15264055132865906,
      "learning_rate": 3.593400190143221e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8169372230768204,
      "num_tokens": 666766319.0,
      "step": 20891
    },
    {
      "entropy": 0.6739100404083729,
      "epoch": 1.9221798609553713,
      "grad_norm": 0.1580684930086136,
      "learning_rate": 3.5930935075290584e-05,
      "loss": 0.6594,
      "mean_token_accuracy": 0.7982731461524963,
      "num_tokens": 666798237.0,
      "step": 20892
    },
    {
      "entropy": 0.6678210105746984,
      "epoch": 1.9222718673268124,
      "grad_norm": 0.1536102443933487,
      "learning_rate": 3.592786824914895e-05,
      "loss": 0.6505,
      "mean_token_accuracy": 0.7989983074367046,
      "num_tokens": 666830221.0,
      "step": 20893
    },
    {
      "entropy": 0.5842302022501826,
      "epoch": 1.9223638736982536,
      "grad_norm": 0.15998192131519318,
      "learning_rate": 3.5924801423007334e-05,
      "loss": 0.5797,
      "mean_token_accuracy": 0.8195296823978424,
      "num_tokens": 666862447.0,
      "step": 20894
    },
    {
      "entropy": 0.5858440669253469,
      "epoch": 1.9224558800696947,
      "grad_norm": 0.15449324250221252,
      "learning_rate": 3.59217345968657e-05,
      "loss": 0.5926,
      "mean_token_accuracy": 0.8229421116411686,
      "num_tokens": 666895215.0,
      "step": 20895
    },
    {
      "entropy": 0.6280314419418573,
      "epoch": 1.922547886441136,
      "grad_norm": 0.15748374164104462,
      "learning_rate": 3.5918667770724076e-05,
      "loss": 0.62,
      "mean_token_accuracy": 0.8094335459172726,
      "num_tokens": 666927055.0,
      "step": 20896
    },
    {
      "entropy": 0.6990397907793522,
      "epoch": 1.9226398928125774,
      "grad_norm": 0.15774236619472504,
      "learning_rate": 3.591560094458246e-05,
      "loss": 0.6765,
      "mean_token_accuracy": 0.7927207462489605,
      "num_tokens": 666958981.0,
      "step": 20897
    },
    {
      "entropy": 0.6445838818326592,
      "epoch": 1.9227318991840185,
      "grad_norm": 0.15921321511268616,
      "learning_rate": 3.5912534118440826e-05,
      "loss": 0.6323,
      "mean_token_accuracy": 0.8064374886453152,
      "num_tokens": 666990697.0,
      "step": 20898
    },
    {
      "entropy": 0.6628125999122858,
      "epoch": 1.9228239055554597,
      "grad_norm": 0.15936179459095,
      "learning_rate": 3.590946729229921e-05,
      "loss": 0.649,
      "mean_token_accuracy": 0.7988079488277435,
      "num_tokens": 667022200.0,
      "step": 20899
    },
    {
      "entropy": 0.6734493877738714,
      "epoch": 1.9229159119269008,
      "grad_norm": 0.16290967166423798,
      "learning_rate": 3.5906400466157575e-05,
      "loss": 0.6801,
      "mean_token_accuracy": 0.7949211597442627,
      "num_tokens": 667053386.0,
      "step": 20900
    },
    {
      "entropy": 0.6269977418705821,
      "epoch": 1.9230079182983422,
      "grad_norm": 0.15300053358078003,
      "learning_rate": 3.590333364001595e-05,
      "loss": 0.6334,
      "mean_token_accuracy": 0.8107634894549847,
      "num_tokens": 667086149.0,
      "step": 20901
    },
    {
      "entropy": 0.5727738267742097,
      "epoch": 1.9230999246697835,
      "grad_norm": 0.1487153023481369,
      "learning_rate": 3.5900266813874324e-05,
      "loss": 0.5675,
      "mean_token_accuracy": 0.8244356028735638,
      "num_tokens": 667117996.0,
      "step": 20902
    },
    {
      "entropy": 0.5952338026836514,
      "epoch": 1.9231919310412247,
      "grad_norm": 0.15904977917671204,
      "learning_rate": 3.58971999877327e-05,
      "loss": 0.5973,
      "mean_token_accuracy": 0.8179921805858612,
      "num_tokens": 667150246.0,
      "step": 20903
    },
    {
      "entropy": 0.6168419113382697,
      "epoch": 1.9232839374126658,
      "grad_norm": 0.15635590255260468,
      "learning_rate": 3.5894133161591073e-05,
      "loss": 0.6099,
      "mean_token_accuracy": 0.8119124099612236,
      "num_tokens": 667181851.0,
      "step": 20904
    },
    {
      "entropy": 0.5076748463325202,
      "epoch": 1.923375943784107,
      "grad_norm": 0.14171244204044342,
      "learning_rate": 3.589106633544945e-05,
      "loss": 0.4994,
      "mean_token_accuracy": 0.8411156088113785,
      "num_tokens": 667213919.0,
      "step": 20905
    },
    {
      "entropy": 0.5002440977841616,
      "epoch": 1.9234679501555483,
      "grad_norm": 0.14275656640529633,
      "learning_rate": 3.5887999509307816e-05,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8465732522308826,
      "num_tokens": 667246022.0,
      "step": 20906
    },
    {
      "entropy": 0.6730241272598505,
      "epoch": 1.9235599565269896,
      "grad_norm": 0.15601135790348053,
      "learning_rate": 3.58849326831662e-05,
      "loss": 0.6587,
      "mean_token_accuracy": 0.8012006618082523,
      "num_tokens": 667278126.0,
      "step": 20907
    },
    {
      "entropy": 0.5011402876116335,
      "epoch": 1.9236519628984308,
      "grad_norm": 0.14363782107830048,
      "learning_rate": 3.5881865857024565e-05,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8500814698636532,
      "num_tokens": 667310541.0,
      "step": 20908
    },
    {
      "entropy": 0.667830603197217,
      "epoch": 1.923743969269872,
      "grad_norm": 0.16175542771816254,
      "learning_rate": 3.587879903088294e-05,
      "loss": 0.6494,
      "mean_token_accuracy": 0.8015452697873116,
      "num_tokens": 667342267.0,
      "step": 20909
    },
    {
      "entropy": 0.648101401515305,
      "epoch": 1.923835975641313,
      "grad_norm": 0.15505538880825043,
      "learning_rate": 3.5875732204741315e-05,
      "loss": 0.6448,
      "mean_token_accuracy": 0.8027565106749535,
      "num_tokens": 667374168.0,
      "step": 20910
    },
    {
      "entropy": 0.5710237799212337,
      "epoch": 1.9239279820127544,
      "grad_norm": 0.15160386264324188,
      "learning_rate": 3.587266537859969e-05,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8250172026455402,
      "num_tokens": 667406597.0,
      "step": 20911
    },
    {
      "entropy": 0.7045861538499594,
      "epoch": 1.9240199883841957,
      "grad_norm": 0.15752585232257843,
      "learning_rate": 3.5869598552458064e-05,
      "loss": 0.6861,
      "mean_token_accuracy": 0.7885650396347046,
      "num_tokens": 667439106.0,
      "step": 20912
    },
    {
      "entropy": 0.6074042599648237,
      "epoch": 1.9241119947556369,
      "grad_norm": 0.15751880407333374,
      "learning_rate": 3.586653172631644e-05,
      "loss": 0.5921,
      "mean_token_accuracy": 0.8181850388646126,
      "num_tokens": 667470972.0,
      "step": 20913
    },
    {
      "entropy": 0.6396322385407984,
      "epoch": 1.924204001127078,
      "grad_norm": 0.1526961773633957,
      "learning_rate": 3.5863464900174806e-05,
      "loss": 0.618,
      "mean_token_accuracy": 0.8087462410330772,
      "num_tokens": 667502484.0,
      "step": 20914
    },
    {
      "entropy": 0.5959991817362607,
      "epoch": 1.9242960074985191,
      "grad_norm": 0.15263204276561737,
      "learning_rate": 3.586039807403319e-05,
      "loss": 0.5863,
      "mean_token_accuracy": 0.8204733431339264,
      "num_tokens": 667534107.0,
      "step": 20915
    },
    {
      "entropy": 0.6767136389389634,
      "epoch": 1.9243880138699605,
      "grad_norm": 0.15726175904273987,
      "learning_rate": 3.5857331247891556e-05,
      "loss": 0.664,
      "mean_token_accuracy": 0.7928234823048115,
      "num_tokens": 667565753.0,
      "step": 20916
    },
    {
      "entropy": 0.6532901898026466,
      "epoch": 1.9244800202414019,
      "grad_norm": 0.15818138420581818,
      "learning_rate": 3.585426442174993e-05,
      "loss": 0.6438,
      "mean_token_accuracy": 0.8008339703083038,
      "num_tokens": 667597185.0,
      "step": 20917
    },
    {
      "entropy": 0.6425359230488539,
      "epoch": 1.924572026612843,
      "grad_norm": 0.1542520672082901,
      "learning_rate": 3.5851197595608305e-05,
      "loss": 0.6383,
      "mean_token_accuracy": 0.8090091347694397,
      "num_tokens": 667629107.0,
      "step": 20918
    },
    {
      "entropy": 0.592678714543581,
      "epoch": 1.9246640329842841,
      "grad_norm": 0.1478932648897171,
      "learning_rate": 3.584813076946668e-05,
      "loss": 0.5804,
      "mean_token_accuracy": 0.8226908519864082,
      "num_tokens": 667661196.0,
      "step": 20919
    },
    {
      "entropy": 0.6986059546470642,
      "epoch": 1.9247560393557253,
      "grad_norm": 0.16861440241336823,
      "learning_rate": 3.5845063943325054e-05,
      "loss": 0.6842,
      "mean_token_accuracy": 0.7890169583261013,
      "num_tokens": 667692454.0,
      "step": 20920
    },
    {
      "entropy": 0.579173588193953,
      "epoch": 1.9248480457271666,
      "grad_norm": 0.15622739493846893,
      "learning_rate": 3.584199711718343e-05,
      "loss": 0.5662,
      "mean_token_accuracy": 0.8272637315094471,
      "num_tokens": 667723846.0,
      "step": 20921
    },
    {
      "entropy": 0.548562178388238,
      "epoch": 1.924940052098608,
      "grad_norm": 0.14404547214508057,
      "learning_rate": 3.58389302910418e-05,
      "loss": 0.5345,
      "mean_token_accuracy": 0.832494780421257,
      "num_tokens": 667756445.0,
      "step": 20922
    },
    {
      "entropy": 0.5426093898713589,
      "epoch": 1.925032058470049,
      "grad_norm": 0.14286789298057556,
      "learning_rate": 3.583586346490018e-05,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8359294943511486,
      "num_tokens": 667788611.0,
      "step": 20923
    },
    {
      "entropy": 0.5999702559784055,
      "epoch": 1.9251240648414902,
      "grad_norm": 0.1525174379348755,
      "learning_rate": 3.5832796638758546e-05,
      "loss": 0.5936,
      "mean_token_accuracy": 0.8174733705818653,
      "num_tokens": 667820513.0,
      "step": 20924
    },
    {
      "entropy": 0.6312863547354937,
      "epoch": 1.9252160712129314,
      "grad_norm": 0.15094804763793945,
      "learning_rate": 3.582972981261693e-05,
      "loss": 0.6144,
      "mean_token_accuracy": 0.8087395280599594,
      "num_tokens": 667852129.0,
      "step": 20925
    },
    {
      "entropy": 0.5922365104779601,
      "epoch": 1.9253080775843727,
      "grad_norm": 0.1547476351261139,
      "learning_rate": 3.58266629864753e-05,
      "loss": 0.5908,
      "mean_token_accuracy": 0.8203539699316025,
      "num_tokens": 667883708.0,
      "step": 20926
    },
    {
      "entropy": 0.6311000445857644,
      "epoch": 1.925400083955814,
      "grad_norm": 0.15261204540729523,
      "learning_rate": 3.582359616033367e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8088265210390091,
      "num_tokens": 667916249.0,
      "step": 20927
    },
    {
      "entropy": 0.6137969316914678,
      "epoch": 1.9254920903272552,
      "grad_norm": 0.15407289564609528,
      "learning_rate": 3.582052933419205e-05,
      "loss": 0.611,
      "mean_token_accuracy": 0.8162822052836418,
      "num_tokens": 667948291.0,
      "step": 20928
    },
    {
      "entropy": 0.6282985042780638,
      "epoch": 1.9255840966986963,
      "grad_norm": 0.16169489920139313,
      "learning_rate": 3.581746250805042e-05,
      "loss": 0.616,
      "mean_token_accuracy": 0.8133943639695644,
      "num_tokens": 667980720.0,
      "step": 20929
    },
    {
      "entropy": 0.6556427292525768,
      "epoch": 1.9256761030701375,
      "grad_norm": 0.15999507904052734,
      "learning_rate": 3.5814395681908794e-05,
      "loss": 0.6565,
      "mean_token_accuracy": 0.7991676367819309,
      "num_tokens": 668012894.0,
      "step": 20930
    },
    {
      "entropy": 0.5694716284051538,
      "epoch": 1.9257681094415788,
      "grad_norm": 0.14631590247154236,
      "learning_rate": 3.581132885576717e-05,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8260504603385925,
      "num_tokens": 668044446.0,
      "step": 20931
    },
    {
      "entropy": 0.7130844723433256,
      "epoch": 1.9258601158130202,
      "grad_norm": 0.16044923663139343,
      "learning_rate": 3.580826202962554e-05,
      "loss": 0.7127,
      "mean_token_accuracy": 0.7807388454675674,
      "num_tokens": 668076283.0,
      "step": 20932
    },
    {
      "entropy": 0.5320655610412359,
      "epoch": 1.9259521221844613,
      "grad_norm": 0.15244226157665253,
      "learning_rate": 3.580519520348392e-05,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8396819122135639,
      "num_tokens": 668107772.0,
      "step": 20933
    },
    {
      "entropy": 0.638291759416461,
      "epoch": 1.9260441285559025,
      "grad_norm": 0.1554151475429535,
      "learning_rate": 3.580212837734229e-05,
      "loss": 0.636,
      "mean_token_accuracy": 0.8064165040850639,
      "num_tokens": 668138795.0,
      "step": 20934
    },
    {
      "entropy": 0.7369159404188395,
      "epoch": 1.9261361349273436,
      "grad_norm": 0.1615416705608368,
      "learning_rate": 3.579906155120066e-05,
      "loss": 0.7254,
      "mean_token_accuracy": 0.781586017459631,
      "num_tokens": 668170804.0,
      "step": 20935
    },
    {
      "entropy": 0.5749469473958015,
      "epoch": 1.926228141298785,
      "grad_norm": 0.15085378289222717,
      "learning_rate": 3.579599472505904e-05,
      "loss": 0.5693,
      "mean_token_accuracy": 0.8231405504047871,
      "num_tokens": 668203214.0,
      "step": 20936
    },
    {
      "entropy": 0.5382599979639053,
      "epoch": 1.9263201476702263,
      "grad_norm": 0.1445733606815338,
      "learning_rate": 3.579292789891741e-05,
      "loss": 0.5276,
      "mean_token_accuracy": 0.84034363925457,
      "num_tokens": 668235749.0,
      "step": 20937
    },
    {
      "entropy": 0.6975209843367338,
      "epoch": 1.9264121540416674,
      "grad_norm": 0.1604008674621582,
      "learning_rate": 3.5789861072775784e-05,
      "loss": 0.6779,
      "mean_token_accuracy": 0.789719682186842,
      "num_tokens": 668268051.0,
      "step": 20938
    },
    {
      "entropy": 0.5501402299851179,
      "epoch": 1.9265041604131086,
      "grad_norm": 0.15154552459716797,
      "learning_rate": 3.578679424663416e-05,
      "loss": 0.5401,
      "mean_token_accuracy": 0.8298440128564835,
      "num_tokens": 668298940.0,
      "step": 20939
    },
    {
      "entropy": 0.6444477736949921,
      "epoch": 1.9265961667845497,
      "grad_norm": 0.15440380573272705,
      "learning_rate": 3.5783727420492534e-05,
      "loss": 0.6212,
      "mean_token_accuracy": 0.8031317517161369,
      "num_tokens": 668329713.0,
      "step": 20940
    },
    {
      "entropy": 0.5395638970658183,
      "epoch": 1.926688173155991,
      "grad_norm": 0.14603714644908905,
      "learning_rate": 3.578066059435091e-05,
      "loss": 0.536,
      "mean_token_accuracy": 0.8345241881906986,
      "num_tokens": 668362107.0,
      "step": 20941
    },
    {
      "entropy": 0.6290880478918552,
      "epoch": 1.9267801795274324,
      "grad_norm": 0.1612931340932846,
      "learning_rate": 3.577759376820928e-05,
      "loss": 0.6115,
      "mean_token_accuracy": 0.8095795325934887,
      "num_tokens": 668394185.0,
      "step": 20942
    },
    {
      "entropy": 0.4618234517984092,
      "epoch": 1.9268721858988735,
      "grad_norm": 0.13895907998085022,
      "learning_rate": 3.577452694206766e-05,
      "loss": 0.4423,
      "mean_token_accuracy": 0.8636475019156933,
      "num_tokens": 668426456.0,
      "step": 20943
    },
    {
      "entropy": 0.6416669823229313,
      "epoch": 1.9269641922703147,
      "grad_norm": 0.15270878374576569,
      "learning_rate": 3.577146011592603e-05,
      "loss": 0.6268,
      "mean_token_accuracy": 0.808589518070221,
      "num_tokens": 668458632.0,
      "step": 20944
    },
    {
      "entropy": 0.5720983911305666,
      "epoch": 1.9270561986417558,
      "grad_norm": 0.1476099044084549,
      "learning_rate": 3.57683932897844e-05,
      "loss": 0.5716,
      "mean_token_accuracy": 0.8232794143259525,
      "num_tokens": 668490646.0,
      "step": 20945
    },
    {
      "entropy": 0.5877476995810866,
      "epoch": 1.9271482050131972,
      "grad_norm": 0.15047381818294525,
      "learning_rate": 3.576532646364278e-05,
      "loss": 0.5783,
      "mean_token_accuracy": 0.822748064994812,
      "num_tokens": 668522351.0,
      "step": 20946
    },
    {
      "entropy": 0.5607993877492845,
      "epoch": 1.9272402113846385,
      "grad_norm": 0.15281568467617035,
      "learning_rate": 3.576225963750115e-05,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8265345357358456,
      "num_tokens": 668554580.0,
      "step": 20947
    },
    {
      "entropy": 0.5928238686174154,
      "epoch": 1.9273322177560797,
      "grad_norm": 0.15214328467845917,
      "learning_rate": 3.5759192811359524e-05,
      "loss": 0.5986,
      "mean_token_accuracy": 0.8150172755122185,
      "num_tokens": 668587050.0,
      "step": 20948
    },
    {
      "entropy": 0.6097109792754054,
      "epoch": 1.9274242241275208,
      "grad_norm": 0.1547391563653946,
      "learning_rate": 3.57561259852179e-05,
      "loss": 0.5919,
      "mean_token_accuracy": 0.8141755424439907,
      "num_tokens": 668618621.0,
      "step": 20949
    },
    {
      "entropy": 0.642817635089159,
      "epoch": 1.927516230498962,
      "grad_norm": 0.15532290935516357,
      "learning_rate": 3.575305915907627e-05,
      "loss": 0.6155,
      "mean_token_accuracy": 0.8067924305796623,
      "num_tokens": 668649647.0,
      "step": 20950
    },
    {
      "entropy": 0.682373333722353,
      "epoch": 1.9276082368704033,
      "grad_norm": 0.1616903394460678,
      "learning_rate": 3.574999233293465e-05,
      "loss": 0.6736,
      "mean_token_accuracy": 0.7905936427414417,
      "num_tokens": 668681713.0,
      "step": 20951
    },
    {
      "entropy": 0.5752063419204205,
      "epoch": 1.9277002432418446,
      "grad_norm": 0.1484963744878769,
      "learning_rate": 3.574692550679302e-05,
      "loss": 0.57,
      "mean_token_accuracy": 0.8224292993545532,
      "num_tokens": 668713946.0,
      "step": 20952
    },
    {
      "entropy": 0.6355324732139707,
      "epoch": 1.9277922496132858,
      "grad_norm": 0.15481281280517578,
      "learning_rate": 3.574385868065139e-05,
      "loss": 0.6346,
      "mean_token_accuracy": 0.806499257683754,
      "num_tokens": 668745661.0,
      "step": 20953
    },
    {
      "entropy": 0.6137036490254104,
      "epoch": 1.927884255984727,
      "grad_norm": 0.15671919286251068,
      "learning_rate": 3.574079185450977e-05,
      "loss": 0.5996,
      "mean_token_accuracy": 0.8181240521371365,
      "num_tokens": 668777398.0,
      "step": 20954
    },
    {
      "entropy": 0.6307436740025878,
      "epoch": 1.927976262356168,
      "grad_norm": 0.1546757072210312,
      "learning_rate": 3.573772502836814e-05,
      "loss": 0.6237,
      "mean_token_accuracy": 0.8091492354869843,
      "num_tokens": 668809487.0,
      "step": 20955
    },
    {
      "entropy": 0.5078455833718181,
      "epoch": 1.9280682687276094,
      "grad_norm": 0.1432161033153534,
      "learning_rate": 3.5734658202226514e-05,
      "loss": 0.5005,
      "mean_token_accuracy": 0.8459009230136871,
      "num_tokens": 668841518.0,
      "step": 20956
    },
    {
      "entropy": 0.47899413947016,
      "epoch": 1.9281602750990507,
      "grad_norm": 0.14537973701953888,
      "learning_rate": 3.5731591376084896e-05,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8544695600867271,
      "num_tokens": 668873672.0,
      "step": 20957
    },
    {
      "entropy": 0.6451385039836168,
      "epoch": 1.9282522814704919,
      "grad_norm": 0.1594550907611847,
      "learning_rate": 3.5728524549943264e-05,
      "loss": 0.638,
      "mean_token_accuracy": 0.803968995809555,
      "num_tokens": 668905891.0,
      "step": 20958
    },
    {
      "entropy": 0.5558997076004744,
      "epoch": 1.928344287841933,
      "grad_norm": 0.15085576474666595,
      "learning_rate": 3.5725457723801645e-05,
      "loss": 0.5505,
      "mean_token_accuracy": 0.8320667296648026,
      "num_tokens": 668937967.0,
      "step": 20959
    },
    {
      "entropy": 0.5247505805455148,
      "epoch": 1.9284362942133741,
      "grad_norm": 0.14740976691246033,
      "learning_rate": 3.572239089766001e-05,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8428799659013748,
      "num_tokens": 668969689.0,
      "step": 20960
    },
    {
      "entropy": 0.5859443871304393,
      "epoch": 1.9285283005848155,
      "grad_norm": 0.1503281146287918,
      "learning_rate": 3.571932407151839e-05,
      "loss": 0.5824,
      "mean_token_accuracy": 0.8245326615869999,
      "num_tokens": 669002084.0,
      "step": 20961
    },
    {
      "entropy": 0.6157779823988676,
      "epoch": 1.9286203069562569,
      "grad_norm": 0.15577495098114014,
      "learning_rate": 3.571625724537676e-05,
      "loss": 0.6113,
      "mean_token_accuracy": 0.8132405653595924,
      "num_tokens": 669034458.0,
      "step": 20962
    },
    {
      "entropy": 0.5376713895238936,
      "epoch": 1.928712313327698,
      "grad_norm": 0.14746952056884766,
      "learning_rate": 3.571319041923514e-05,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8426657766103745,
      "num_tokens": 669066106.0,
      "step": 20963
    },
    {
      "entropy": 0.5652703456580639,
      "epoch": 1.9288043196991391,
      "grad_norm": 0.15245884656906128,
      "learning_rate": 3.571012359309351e-05,
      "loss": 0.5561,
      "mean_token_accuracy": 0.8286477215588093,
      "num_tokens": 669098377.0,
      "step": 20964
    },
    {
      "entropy": 0.587549252435565,
      "epoch": 1.9288963260705803,
      "grad_norm": 0.1475956290960312,
      "learning_rate": 3.5707056766951886e-05,
      "loss": 0.5657,
      "mean_token_accuracy": 0.8236222080886364,
      "num_tokens": 669129891.0,
      "step": 20965
    },
    {
      "entropy": 0.5275663137435913,
      "epoch": 1.9289883324420216,
      "grad_norm": 0.14714473485946655,
      "learning_rate": 3.5703989940810254e-05,
      "loss": 0.5127,
      "mean_token_accuracy": 0.8385657593607903,
      "num_tokens": 669161983.0,
      "step": 20966
    },
    {
      "entropy": 0.6289474666118622,
      "epoch": 1.929080338813463,
      "grad_norm": 0.15138402581214905,
      "learning_rate": 3.5700923114668636e-05,
      "loss": 0.6088,
      "mean_token_accuracy": 0.8125463053584099,
      "num_tokens": 669193406.0,
      "step": 20967
    },
    {
      "entropy": 0.5558250332251191,
      "epoch": 1.929172345184904,
      "grad_norm": 0.1466681808233261,
      "learning_rate": 3.5697856288527003e-05,
      "loss": 0.546,
      "mean_token_accuracy": 0.8317023329436779,
      "num_tokens": 669225729.0,
      "step": 20968
    },
    {
      "entropy": 0.5249992478638887,
      "epoch": 1.9292643515563452,
      "grad_norm": 0.14425458014011383,
      "learning_rate": 3.569478946238538e-05,
      "loss": 0.507,
      "mean_token_accuracy": 0.8417224399745464,
      "num_tokens": 669257424.0,
      "step": 20969
    },
    {
      "entropy": 0.5410570073872805,
      "epoch": 1.9293563579277864,
      "grad_norm": 0.14298953115940094,
      "learning_rate": 3.569172263624375e-05,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8332069739699364,
      "num_tokens": 669289629.0,
      "step": 20970
    },
    {
      "entropy": 0.6018887339159846,
      "epoch": 1.9294483642992277,
      "grad_norm": 0.15356746315956116,
      "learning_rate": 3.568865581010213e-05,
      "loss": 0.5926,
      "mean_token_accuracy": 0.8187834918498993,
      "num_tokens": 669321371.0,
      "step": 20971
    },
    {
      "entropy": 0.6107905870303512,
      "epoch": 1.929540370670669,
      "grad_norm": 0.15499047935009003,
      "learning_rate": 3.56855889839605e-05,
      "loss": 0.5983,
      "mean_token_accuracy": 0.8151769824326038,
      "num_tokens": 669353512.0,
      "step": 20972
    },
    {
      "entropy": 0.5251995422877371,
      "epoch": 1.9296323770421102,
      "grad_norm": 0.14435520768165588,
      "learning_rate": 3.568252215781888e-05,
      "loss": 0.5316,
      "mean_token_accuracy": 0.8367696776986122,
      "num_tokens": 669386022.0,
      "step": 20973
    },
    {
      "entropy": 0.55700902082026,
      "epoch": 1.9297243834135513,
      "grad_norm": 0.15254397690296173,
      "learning_rate": 3.5679455331677245e-05,
      "loss": 0.5413,
      "mean_token_accuracy": 0.8291926346719265,
      "num_tokens": 669417874.0,
      "step": 20974
    },
    {
      "entropy": 0.5676609538495541,
      "epoch": 1.9298163897849925,
      "grad_norm": 0.14636902511119843,
      "learning_rate": 3.5676388505535626e-05,
      "loss": 0.5581,
      "mean_token_accuracy": 0.8319153226912022,
      "num_tokens": 669450441.0,
      "step": 20975
    },
    {
      "entropy": 0.7338172253221273,
      "epoch": 1.9299083961564338,
      "grad_norm": 0.17106960713863373,
      "learning_rate": 3.5673321679393994e-05,
      "loss": 0.7307,
      "mean_token_accuracy": 0.7780205644667149,
      "num_tokens": 669481277.0,
      "step": 20976
    },
    {
      "entropy": 0.7154701892286539,
      "epoch": 1.9300004025278752,
      "grad_norm": 0.1675308495759964,
      "learning_rate": 3.567025485325237e-05,
      "loss": 0.714,
      "mean_token_accuracy": 0.7858095727860928,
      "num_tokens": 669513299.0,
      "step": 20977
    },
    {
      "entropy": 0.7292438428848982,
      "epoch": 1.9300924088993163,
      "grad_norm": 0.16732482612133026,
      "learning_rate": 3.566718802711074e-05,
      "loss": 0.741,
      "mean_token_accuracy": 0.7757912315428257,
      "num_tokens": 669544473.0,
      "step": 20978
    },
    {
      "entropy": 0.6367388060316443,
      "epoch": 1.9301844152707575,
      "grad_norm": 0.16199848055839539,
      "learning_rate": 3.566412120096912e-05,
      "loss": 0.646,
      "mean_token_accuracy": 0.801589023321867,
      "num_tokens": 669575771.0,
      "step": 20979
    },
    {
      "entropy": 0.5799961276352406,
      "epoch": 1.9302764216421986,
      "grad_norm": 0.1538671851158142,
      "learning_rate": 3.566105437482749e-05,
      "loss": 0.5712,
      "mean_token_accuracy": 0.8216382600367069,
      "num_tokens": 669607941.0,
      "step": 20980
    },
    {
      "entropy": 0.4660914884880185,
      "epoch": 1.93036842801364,
      "grad_norm": 0.13884887099266052,
      "learning_rate": 3.565798754868587e-05,
      "loss": 0.4603,
      "mean_token_accuracy": 0.8555714935064316,
      "num_tokens": 669640653.0,
      "step": 20981
    },
    {
      "entropy": 0.5163142981473356,
      "epoch": 1.9304604343850813,
      "grad_norm": 0.14280563592910767,
      "learning_rate": 3.5654920722544235e-05,
      "loss": 0.5077,
      "mean_token_accuracy": 0.8412291668355465,
      "num_tokens": 669672716.0,
      "step": 20982
    },
    {
      "entropy": 0.5799162834882736,
      "epoch": 1.9305524407565224,
      "grad_norm": 0.15739953517913818,
      "learning_rate": 3.5651853896402616e-05,
      "loss": 0.5908,
      "mean_token_accuracy": 0.8184514530003071,
      "num_tokens": 669704366.0,
      "step": 20983
    },
    {
      "entropy": 0.6266928762197495,
      "epoch": 1.9306444471279636,
      "grad_norm": 0.1550017148256302,
      "learning_rate": 3.5648787070260984e-05,
      "loss": 0.6128,
      "mean_token_accuracy": 0.8072903603315353,
      "num_tokens": 669736100.0,
      "step": 20984
    },
    {
      "entropy": 0.5576075012795627,
      "epoch": 1.9307364534994047,
      "grad_norm": 0.1518106907606125,
      "learning_rate": 3.5645720244119366e-05,
      "loss": 0.5562,
      "mean_token_accuracy": 0.8310026861727238,
      "num_tokens": 669768868.0,
      "step": 20985
    },
    {
      "entropy": 0.5733818840235472,
      "epoch": 1.930828459870846,
      "grad_norm": 0.15476973354816437,
      "learning_rate": 3.5642653417977734e-05,
      "loss": 0.571,
      "mean_token_accuracy": 0.8244299776852131,
      "num_tokens": 669800510.0,
      "step": 20986
    },
    {
      "entropy": 0.5854878034442663,
      "epoch": 1.9309204662422874,
      "grad_norm": 0.15328796207904816,
      "learning_rate": 3.563958659183611e-05,
      "loss": 0.5667,
      "mean_token_accuracy": 0.8239156194031239,
      "num_tokens": 669832578.0,
      "step": 20987
    },
    {
      "entropy": 0.6770383045077324,
      "epoch": 1.9310124726137285,
      "grad_norm": 0.15309540927410126,
      "learning_rate": 3.563651976569449e-05,
      "loss": 0.6586,
      "mean_token_accuracy": 0.7987090386450291,
      "num_tokens": 669864145.0,
      "step": 20988
    },
    {
      "entropy": 0.4956454678904265,
      "epoch": 1.9311044789851697,
      "grad_norm": 0.1432906687259674,
      "learning_rate": 3.563345293955286e-05,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8496037498116493,
      "num_tokens": 669896150.0,
      "step": 20989
    },
    {
      "entropy": 0.6068426901474595,
      "epoch": 1.9311964853566108,
      "grad_norm": 0.14900290966033936,
      "learning_rate": 3.563038611341123e-05,
      "loss": 0.5838,
      "mean_token_accuracy": 0.8217085711658001,
      "num_tokens": 669928271.0,
      "step": 20990
    },
    {
      "entropy": 0.5315742427483201,
      "epoch": 1.9312884917280522,
      "grad_norm": 0.14841563999652863,
      "learning_rate": 3.562731928726961e-05,
      "loss": 0.5111,
      "mean_token_accuracy": 0.842099167406559,
      "num_tokens": 669959474.0,
      "step": 20991
    },
    {
      "entropy": 0.6512664090842009,
      "epoch": 1.9313804980994935,
      "grad_norm": 0.1543082594871521,
      "learning_rate": 3.562425246112798e-05,
      "loss": 0.6468,
      "mean_token_accuracy": 0.7987298592925072,
      "num_tokens": 669991840.0,
      "step": 20992
    },
    {
      "entropy": 0.5543630225583911,
      "epoch": 1.9314725044709347,
      "grad_norm": 0.14220446348190308,
      "learning_rate": 3.5621185634986356e-05,
      "loss": 0.5294,
      "mean_token_accuracy": 0.8322172239422798,
      "num_tokens": 670023708.0,
      "step": 20993
    },
    {
      "entropy": 0.5434652417898178,
      "epoch": 1.9315645108423758,
      "grad_norm": 0.1520615965127945,
      "learning_rate": 3.561811880884473e-05,
      "loss": 0.5327,
      "mean_token_accuracy": 0.8368837386369705,
      "num_tokens": 670055649.0,
      "step": 20994
    },
    {
      "entropy": 0.5048559131100774,
      "epoch": 1.931656517213817,
      "grad_norm": 0.14028844237327576,
      "learning_rate": 3.56150519827031e-05,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8447946980595589,
      "num_tokens": 670087596.0,
      "step": 20995
    },
    {
      "entropy": 0.49495579255744815,
      "epoch": 1.9317485235852583,
      "grad_norm": 0.1499784141778946,
      "learning_rate": 3.561198515656148e-05,
      "loss": 0.4771,
      "mean_token_accuracy": 0.8524036966264248,
      "num_tokens": 670120351.0,
      "step": 20996
    },
    {
      "entropy": 0.679379278793931,
      "epoch": 1.9318405299566996,
      "grad_norm": 0.1596830189228058,
      "learning_rate": 3.560891833041985e-05,
      "loss": 0.6605,
      "mean_token_accuracy": 0.794262208044529,
      "num_tokens": 670152292.0,
      "step": 20997
    },
    {
      "entropy": 0.6148742716759443,
      "epoch": 1.9319325363281408,
      "grad_norm": 0.15246787667274475,
      "learning_rate": 3.560585150427822e-05,
      "loss": 0.6099,
      "mean_token_accuracy": 0.8130801133811474,
      "num_tokens": 670185060.0,
      "step": 20998
    },
    {
      "entropy": 0.6036155726760626,
      "epoch": 1.932024542699582,
      "grad_norm": 0.15156981348991394,
      "learning_rate": 3.56027846781366e-05,
      "loss": 0.5963,
      "mean_token_accuracy": 0.8136178776621819,
      "num_tokens": 670216998.0,
      "step": 20999
    },
    {
      "entropy": 0.5985369477421045,
      "epoch": 1.932116549071023,
      "grad_norm": 0.1520117074251175,
      "learning_rate": 3.559971785199497e-05,
      "loss": 0.5795,
      "mean_token_accuracy": 0.8196519948542118,
      "num_tokens": 670248377.0,
      "step": 21000
    },
    {
      "entropy": 0.5509498775936663,
      "epoch": 1.9322085554424644,
      "grad_norm": 0.1415025293827057,
      "learning_rate": 3.5596651025853346e-05,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8350675962865353,
      "num_tokens": 670280722.0,
      "step": 21001
    },
    {
      "entropy": 0.594791192561388,
      "epoch": 1.9323005618139057,
      "grad_norm": 0.15966001152992249,
      "learning_rate": 3.559358419971172e-05,
      "loss": 0.5892,
      "mean_token_accuracy": 0.817914891988039,
      "num_tokens": 670311990.0,
      "step": 21002
    },
    {
      "entropy": 0.6274745669215918,
      "epoch": 1.9323925681853469,
      "grad_norm": 0.15553617477416992,
      "learning_rate": 3.5590517373570096e-05,
      "loss": 0.6281,
      "mean_token_accuracy": 0.8084848932921886,
      "num_tokens": 670344322.0,
      "step": 21003
    },
    {
      "entropy": 0.5447042975574732,
      "epoch": 1.932484574556788,
      "grad_norm": 0.15063922107219696,
      "learning_rate": 3.558745054742847e-05,
      "loss": 0.5365,
      "mean_token_accuracy": 0.8301139026880264,
      "num_tokens": 670376054.0,
      "step": 21004
    },
    {
      "entropy": 0.7164693493396044,
      "epoch": 1.9325765809282291,
      "grad_norm": 0.16180744767189026,
      "learning_rate": 3.558438372128684e-05,
      "loss": 0.7138,
      "mean_token_accuracy": 0.7833103910088539,
      "num_tokens": 670408194.0,
      "step": 21005
    },
    {
      "entropy": 0.6696190573275089,
      "epoch": 1.9326685872996705,
      "grad_norm": 0.15589003264904022,
      "learning_rate": 3.558131689514522e-05,
      "loss": 0.6593,
      "mean_token_accuracy": 0.7968557924032211,
      "num_tokens": 670439157.0,
      "step": 21006
    },
    {
      "entropy": 0.6086272895336151,
      "epoch": 1.9327605936711119,
      "grad_norm": 0.1530265212059021,
      "learning_rate": 3.557825006900359e-05,
      "loss": 0.6078,
      "mean_token_accuracy": 0.8166218772530556,
      "num_tokens": 670471925.0,
      "step": 21007
    },
    {
      "entropy": 0.5504309497773647,
      "epoch": 1.932852600042553,
      "grad_norm": 0.1511172652244568,
      "learning_rate": 3.557518324286196e-05,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8351453468203545,
      "num_tokens": 670503775.0,
      "step": 21008
    },
    {
      "entropy": 0.73553048633039,
      "epoch": 1.9329446064139941,
      "grad_norm": 0.16675704717636108,
      "learning_rate": 3.557211641672034e-05,
      "loss": 0.7319,
      "mean_token_accuracy": 0.7767983488738537,
      "num_tokens": 670536246.0,
      "step": 21009
    },
    {
      "entropy": 0.6317906668409705,
      "epoch": 1.9330366127854353,
      "grad_norm": 0.15880967676639557,
      "learning_rate": 3.556904959057871e-05,
      "loss": 0.6301,
      "mean_token_accuracy": 0.8098153062164783,
      "num_tokens": 670567925.0,
      "step": 21010
    },
    {
      "entropy": 0.6552189625799656,
      "epoch": 1.9331286191568766,
      "grad_norm": 0.157531276345253,
      "learning_rate": 3.5565982764437086e-05,
      "loss": 0.6507,
      "mean_token_accuracy": 0.8032572343945503,
      "num_tokens": 670600046.0,
      "step": 21011
    },
    {
      "entropy": 0.6389741552993655,
      "epoch": 1.933220625528318,
      "grad_norm": 0.15603069961071014,
      "learning_rate": 3.556291593829546e-05,
      "loss": 0.6213,
      "mean_token_accuracy": 0.8063843622803688,
      "num_tokens": 670631509.0,
      "step": 21012
    },
    {
      "entropy": 0.6103151328861713,
      "epoch": 1.933312631899759,
      "grad_norm": 0.15107743442058563,
      "learning_rate": 3.555984911215383e-05,
      "loss": 0.6058,
      "mean_token_accuracy": 0.8185202926397324,
      "num_tokens": 670663252.0,
      "step": 21013
    },
    {
      "entropy": 0.46238038409501314,
      "epoch": 1.9334046382712002,
      "grad_norm": 0.13482458889484406,
      "learning_rate": 3.555678228601221e-05,
      "loss": 0.4501,
      "mean_token_accuracy": 0.8565055951476097,
      "num_tokens": 670695601.0,
      "step": 21014
    },
    {
      "entropy": 0.5911774653941393,
      "epoch": 1.9334966446426414,
      "grad_norm": 0.15190550684928894,
      "learning_rate": 3.555371545987058e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.8225640989840031,
      "num_tokens": 670727524.0,
      "step": 21015
    },
    {
      "entropy": 0.5862418878823519,
      "epoch": 1.9335886510140827,
      "grad_norm": 0.14824794232845306,
      "learning_rate": 3.555064863372895e-05,
      "loss": 0.5806,
      "mean_token_accuracy": 0.8187780193984509,
      "num_tokens": 670759445.0,
      "step": 21016
    },
    {
      "entropy": 0.5760104444343597,
      "epoch": 1.933680657385524,
      "grad_norm": 0.1488177627325058,
      "learning_rate": 3.554758180758733e-05,
      "loss": 0.5504,
      "mean_token_accuracy": 0.8320136331021786,
      "num_tokens": 670791529.0,
      "step": 21017
    },
    {
      "entropy": 0.6074622869491577,
      "epoch": 1.9337726637569652,
      "grad_norm": 0.1558322161436081,
      "learning_rate": 3.55445149814457e-05,
      "loss": 0.5973,
      "mean_token_accuracy": 0.8192459531128407,
      "num_tokens": 670823590.0,
      "step": 21018
    },
    {
      "entropy": 0.6034395524766296,
      "epoch": 1.9338646701284063,
      "grad_norm": 0.15149499475955963,
      "learning_rate": 3.554144815530408e-05,
      "loss": 0.5823,
      "mean_token_accuracy": 0.8216255158185959,
      "num_tokens": 670855392.0,
      "step": 21019
    },
    {
      "entropy": 0.5511016622185707,
      "epoch": 1.9339566764998475,
      "grad_norm": 0.14694558084011078,
      "learning_rate": 3.553838132916245e-05,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8332310281693935,
      "num_tokens": 670887545.0,
      "step": 21020
    },
    {
      "entropy": 0.7073823679238558,
      "epoch": 1.9340486828712888,
      "grad_norm": 0.16330783069133759,
      "learning_rate": 3.5535314503020826e-05,
      "loss": 0.6869,
      "mean_token_accuracy": 0.7925877310335636,
      "num_tokens": 670919830.0,
      "step": 21021
    },
    {
      "entropy": 0.5590445827692747,
      "epoch": 1.9341406892427302,
      "grad_norm": 0.1483929306268692,
      "learning_rate": 3.55322476768792e-05,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8311042562127113,
      "num_tokens": 670951478.0,
      "step": 21022
    },
    {
      "entropy": 0.5660791276022792,
      "epoch": 1.9342326956141713,
      "grad_norm": 0.15462014079093933,
      "learning_rate": 3.5529180850737575e-05,
      "loss": 0.5681,
      "mean_token_accuracy": 0.8216983750462532,
      "num_tokens": 670983660.0,
      "step": 21023
    },
    {
      "entropy": 0.6186610395088792,
      "epoch": 1.9343247019856125,
      "grad_norm": 0.154026061296463,
      "learning_rate": 3.552611402459595e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.8163434639573097,
      "num_tokens": 671015278.0,
      "step": 21024
    },
    {
      "entropy": 0.6029504362959415,
      "epoch": 1.9344167083570536,
      "grad_norm": 0.15419615805149078,
      "learning_rate": 3.5523047198454324e-05,
      "loss": 0.5977,
      "mean_token_accuracy": 0.8165684752166271,
      "num_tokens": 671047926.0,
      "step": 21025
    },
    {
      "entropy": 0.6904262118041515,
      "epoch": 1.934508714728495,
      "grad_norm": 0.158382847905159,
      "learning_rate": 3.551998037231269e-05,
      "loss": 0.682,
      "mean_token_accuracy": 0.7924633733928204,
      "num_tokens": 671079577.0,
      "step": 21026
    },
    {
      "entropy": 0.5819196933880448,
      "epoch": 1.9346007210999363,
      "grad_norm": 0.1489962488412857,
      "learning_rate": 3.5516913546171074e-05,
      "loss": 0.5814,
      "mean_token_accuracy": 0.820697408169508,
      "num_tokens": 671111934.0,
      "step": 21027
    },
    {
      "entropy": 0.6817225459963083,
      "epoch": 1.9346927274713774,
      "grad_norm": 0.16290228068828583,
      "learning_rate": 3.551384672002944e-05,
      "loss": 0.68,
      "mean_token_accuracy": 0.7920538447797298,
      "num_tokens": 671143138.0,
      "step": 21028
    },
    {
      "entropy": 0.6105472203344107,
      "epoch": 1.9347847338428186,
      "grad_norm": 0.15334975719451904,
      "learning_rate": 3.5510779893887816e-05,
      "loss": 0.6025,
      "mean_token_accuracy": 0.8145626150071621,
      "num_tokens": 671175345.0,
      "step": 21029
    },
    {
      "entropy": 0.5759483375586569,
      "epoch": 1.9348767402142597,
      "grad_norm": 0.14780615270137787,
      "learning_rate": 3.550771306774619e-05,
      "loss": 0.5554,
      "mean_token_accuracy": 0.826749574393034,
      "num_tokens": 671206886.0,
      "step": 21030
    },
    {
      "entropy": 0.5810622721910477,
      "epoch": 1.934968746585701,
      "grad_norm": 0.15748079121112823,
      "learning_rate": 3.5504646241604566e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8249508552253246,
      "num_tokens": 671238617.0,
      "step": 21031
    },
    {
      "entropy": 0.6750452211126685,
      "epoch": 1.9350607529571424,
      "grad_norm": 0.16046632826328278,
      "learning_rate": 3.550157941546294e-05,
      "loss": 0.6742,
      "mean_token_accuracy": 0.7945746630430222,
      "num_tokens": 671270012.0,
      "step": 21032
    },
    {
      "entropy": 0.6634003091603518,
      "epoch": 1.9351527593285835,
      "grad_norm": 0.16093768179416656,
      "learning_rate": 3.5498512589321315e-05,
      "loss": 0.6657,
      "mean_token_accuracy": 0.7985420860350132,
      "num_tokens": 671301345.0,
      "step": 21033
    },
    {
      "entropy": 0.5370614770799875,
      "epoch": 1.9352447657000247,
      "grad_norm": 0.1417294442653656,
      "learning_rate": 3.549544576317968e-05,
      "loss": 0.5156,
      "mean_token_accuracy": 0.8378510288894176,
      "num_tokens": 671333395.0,
      "step": 21034
    },
    {
      "entropy": 0.7045867927372456,
      "epoch": 1.9353367720714658,
      "grad_norm": 0.15832467377185822,
      "learning_rate": 3.5492378937038064e-05,
      "loss": 0.6819,
      "mean_token_accuracy": 0.790394451469183,
      "num_tokens": 671365951.0,
      "step": 21035
    },
    {
      "entropy": 0.6366962790489197,
      "epoch": 1.9354287784429072,
      "grad_norm": 0.1576095074415207,
      "learning_rate": 3.548931211089643e-05,
      "loss": 0.6245,
      "mean_token_accuracy": 0.8072497509419918,
      "num_tokens": 671397342.0,
      "step": 21036
    },
    {
      "entropy": 0.5739589140284806,
      "epoch": 1.9355207848143485,
      "grad_norm": 0.14997798204421997,
      "learning_rate": 3.548624528475481e-05,
      "loss": 0.5634,
      "mean_token_accuracy": 0.8281136825680733,
      "num_tokens": 671429233.0,
      "step": 21037
    },
    {
      "entropy": 0.5745242740958929,
      "epoch": 1.9356127911857897,
      "grad_norm": 0.14722371101379395,
      "learning_rate": 3.548317845861318e-05,
      "loss": 0.5663,
      "mean_token_accuracy": 0.8280689716339111,
      "num_tokens": 671460961.0,
      "step": 21038
    },
    {
      "entropy": 0.5961984610185027,
      "epoch": 1.9357047975572308,
      "grad_norm": 0.16228623688220978,
      "learning_rate": 3.5480111632471556e-05,
      "loss": 0.5863,
      "mean_token_accuracy": 0.8189680986106396,
      "num_tokens": 671492889.0,
      "step": 21039
    },
    {
      "entropy": 0.5314139234833419,
      "epoch": 1.935796803928672,
      "grad_norm": 0.14354626834392548,
      "learning_rate": 3.547704480632993e-05,
      "loss": 0.508,
      "mean_token_accuracy": 0.8398086726665497,
      "num_tokens": 671524859.0,
      "step": 21040
    },
    {
      "entropy": 0.529807454906404,
      "epoch": 1.9358888103001133,
      "grad_norm": 0.1434502750635147,
      "learning_rate": 3.5473977980188305e-05,
      "loss": 0.5185,
      "mean_token_accuracy": 0.8394778706133366,
      "num_tokens": 671556344.0,
      "step": 21041
    },
    {
      "entropy": 0.5909506431780756,
      "epoch": 1.9359808166715546,
      "grad_norm": 0.1496492475271225,
      "learning_rate": 3.547091115404668e-05,
      "loss": 0.5773,
      "mean_token_accuracy": 0.8230418153107166,
      "num_tokens": 671588394.0,
      "step": 21042
    },
    {
      "entropy": 0.4864596966654062,
      "epoch": 1.9360728230429958,
      "grad_norm": 0.14517204463481903,
      "learning_rate": 3.5467844327905055e-05,
      "loss": 0.4648,
      "mean_token_accuracy": 0.8509184941649437,
      "num_tokens": 671620032.0,
      "step": 21043
    },
    {
      "entropy": 0.6890198104083538,
      "epoch": 1.936164829414437,
      "grad_norm": 0.157669335603714,
      "learning_rate": 3.546477750176342e-05,
      "loss": 0.6739,
      "mean_token_accuracy": 0.7937834896147251,
      "num_tokens": 671652338.0,
      "step": 21044
    },
    {
      "entropy": 0.671354933641851,
      "epoch": 1.936256835785878,
      "grad_norm": 0.16258160769939423,
      "learning_rate": 3.5461710675621804e-05,
      "loss": 0.6679,
      "mean_token_accuracy": 0.7967691496014595,
      "num_tokens": 671683954.0,
      "step": 21045
    },
    {
      "entropy": 0.5552721489220858,
      "epoch": 1.9363488421573194,
      "grad_norm": 0.14946427941322327,
      "learning_rate": 3.545864384948017e-05,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8289774879813194,
      "num_tokens": 671714950.0,
      "step": 21046
    },
    {
      "entropy": 0.7620738744735718,
      "epoch": 1.9364408485287607,
      "grad_norm": 0.16499429941177368,
      "learning_rate": 3.5455577023338546e-05,
      "loss": 0.7715,
      "mean_token_accuracy": 0.7695983648300171,
      "num_tokens": 671747224.0,
      "step": 21047
    },
    {
      "entropy": 0.6368988025933504,
      "epoch": 1.9365328549002019,
      "grad_norm": 0.16705043613910675,
      "learning_rate": 3.545251019719693e-05,
      "loss": 0.6359,
      "mean_token_accuracy": 0.8042364977300167,
      "num_tokens": 671777133.0,
      "step": 21048
    },
    {
      "entropy": 0.6488419063389301,
      "epoch": 1.936624861271643,
      "grad_norm": 0.1574087142944336,
      "learning_rate": 3.5449443371055296e-05,
      "loss": 0.6474,
      "mean_token_accuracy": 0.8001207672059536,
      "num_tokens": 671809115.0,
      "step": 21049
    },
    {
      "entropy": 0.6558872209861875,
      "epoch": 1.9367168676430842,
      "grad_norm": 0.16187308728694916,
      "learning_rate": 3.544637654491367e-05,
      "loss": 0.6506,
      "mean_token_accuracy": 0.8019889071583748,
      "num_tokens": 671840317.0,
      "step": 21050
    },
    {
      "entropy": 0.7279415279626846,
      "epoch": 1.9368088740145255,
      "grad_norm": 0.16366557776927948,
      "learning_rate": 3.5443309718772045e-05,
      "loss": 0.7219,
      "mean_token_accuracy": 0.7809479162096977,
      "num_tokens": 671871621.0,
      "step": 21051
    },
    {
      "entropy": 0.6148368921130896,
      "epoch": 1.9369008803859669,
      "grad_norm": 0.1532166600227356,
      "learning_rate": 3.544024289263042e-05,
      "loss": 0.5924,
      "mean_token_accuracy": 0.8181763961911201,
      "num_tokens": 671903460.0,
      "step": 21052
    },
    {
      "entropy": 0.5535617601126432,
      "epoch": 1.936992886757408,
      "grad_norm": 0.14910238981246948,
      "learning_rate": 3.5437176066488794e-05,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8353792205452919,
      "num_tokens": 671934795.0,
      "step": 21053
    },
    {
      "entropy": 0.5725520867854357,
      "epoch": 1.9370848931288491,
      "grad_norm": 0.14689470827579498,
      "learning_rate": 3.543410924034717e-05,
      "loss": 0.5494,
      "mean_token_accuracy": 0.8293032348155975,
      "num_tokens": 671966721.0,
      "step": 21054
    },
    {
      "entropy": 0.5601379035506397,
      "epoch": 1.9371768995002903,
      "grad_norm": 0.1475437432527542,
      "learning_rate": 3.543104241420554e-05,
      "loss": 0.5512,
      "mean_token_accuracy": 0.8300545625388622,
      "num_tokens": 671998592.0,
      "step": 21055
    },
    {
      "entropy": 0.5567045072093606,
      "epoch": 1.9372689058717316,
      "grad_norm": 0.144245445728302,
      "learning_rate": 3.542797558806392e-05,
      "loss": 0.5424,
      "mean_token_accuracy": 0.8314091451466084,
      "num_tokens": 672030987.0,
      "step": 21056
    },
    {
      "entropy": 0.5480721769854426,
      "epoch": 1.937360912243173,
      "grad_norm": 0.15121717751026154,
      "learning_rate": 3.5424908761922286e-05,
      "loss": 0.543,
      "mean_token_accuracy": 0.8334245085716248,
      "num_tokens": 672063551.0,
      "step": 21057
    },
    {
      "entropy": 0.5679698344320059,
      "epoch": 1.937452918614614,
      "grad_norm": 0.14954236149787903,
      "learning_rate": 3.542184193578067e-05,
      "loss": 0.5556,
      "mean_token_accuracy": 0.8290955014526844,
      "num_tokens": 672095831.0,
      "step": 21058
    },
    {
      "entropy": 0.5903918091207743,
      "epoch": 1.9375449249860552,
      "grad_norm": 0.15566781163215637,
      "learning_rate": 3.5418775109639035e-05,
      "loss": 0.5814,
      "mean_token_accuracy": 0.8216296173632145,
      "num_tokens": 672127472.0,
      "step": 21059
    },
    {
      "entropy": 0.680519123794511,
      "epoch": 1.9376369313574964,
      "grad_norm": 0.15862521529197693,
      "learning_rate": 3.541570828349741e-05,
      "loss": 0.6667,
      "mean_token_accuracy": 0.7958555594086647,
      "num_tokens": 672159097.0,
      "step": 21060
    },
    {
      "entropy": 0.6613015364855528,
      "epoch": 1.9377289377289377,
      "grad_norm": 0.16799317300319672,
      "learning_rate": 3.5412641457355785e-05,
      "loss": 0.6458,
      "mean_token_accuracy": 0.8024346195161343,
      "num_tokens": 672190145.0,
      "step": 21061
    },
    {
      "entropy": 0.5561405452899635,
      "epoch": 1.937820944100379,
      "grad_norm": 0.15074095129966736,
      "learning_rate": 3.540957463121416e-05,
      "loss": 0.5578,
      "mean_token_accuracy": 0.8277603089809418,
      "num_tokens": 672222586.0,
      "step": 21062
    },
    {
      "entropy": 0.5803225804120302,
      "epoch": 1.9379129504718202,
      "grad_norm": 0.1520179659128189,
      "learning_rate": 3.5406507805072534e-05,
      "loss": 0.5644,
      "mean_token_accuracy": 0.8246145322918892,
      "num_tokens": 672254577.0,
      "step": 21063
    },
    {
      "entropy": 0.5606180168688297,
      "epoch": 1.9380049568432614,
      "grad_norm": 0.1575687676668167,
      "learning_rate": 3.540344097893091e-05,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8288042657077312,
      "num_tokens": 672286171.0,
      "step": 21064
    },
    {
      "entropy": 0.6245539095252752,
      "epoch": 1.9380969632147025,
      "grad_norm": 0.15356481075286865,
      "learning_rate": 3.5400374152789276e-05,
      "loss": 0.6112,
      "mean_token_accuracy": 0.8114136680960655,
      "num_tokens": 672318388.0,
      "step": 21065
    },
    {
      "entropy": 0.5305067133158445,
      "epoch": 1.9381889695861438,
      "grad_norm": 0.144336998462677,
      "learning_rate": 3.539730732664766e-05,
      "loss": 0.5315,
      "mean_token_accuracy": 0.8378524743020535,
      "num_tokens": 672350583.0,
      "step": 21066
    },
    {
      "entropy": 0.6018648066092283,
      "epoch": 1.9382809759575852,
      "grad_norm": 0.15473908185958862,
      "learning_rate": 3.5394240500506026e-05,
      "loss": 0.5844,
      "mean_token_accuracy": 0.8172145187854767,
      "num_tokens": 672381882.0,
      "step": 21067
    },
    {
      "entropy": 0.638340562582016,
      "epoch": 1.9383729823290263,
      "grad_norm": 0.1571953147649765,
      "learning_rate": 3.53911736743644e-05,
      "loss": 0.622,
      "mean_token_accuracy": 0.8136678971350193,
      "num_tokens": 672414455.0,
      "step": 21068
    },
    {
      "entropy": 0.5311969169415534,
      "epoch": 1.9384649887004675,
      "grad_norm": 0.14701789617538452,
      "learning_rate": 3.5388106848222775e-05,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8391000330448151,
      "num_tokens": 672446552.0,
      "step": 21069
    },
    {
      "entropy": 0.5454166643321514,
      "epoch": 1.9385569950719086,
      "grad_norm": 0.14627376198768616,
      "learning_rate": 3.538504002208115e-05,
      "loss": 0.5306,
      "mean_token_accuracy": 0.8345199935138226,
      "num_tokens": 672478943.0,
      "step": 21070
    },
    {
      "entropy": 0.54716657102108,
      "epoch": 1.93864900144335,
      "grad_norm": 0.1509687602519989,
      "learning_rate": 3.5381973195939524e-05,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8332277834415436,
      "num_tokens": 672511377.0,
      "step": 21071
    },
    {
      "entropy": 0.5992094203829765,
      "epoch": 1.9387410078147913,
      "grad_norm": 0.16002914309501648,
      "learning_rate": 3.53789063697979e-05,
      "loss": 0.6,
      "mean_token_accuracy": 0.8158404864370823,
      "num_tokens": 672543725.0,
      "step": 21072
    },
    {
      "entropy": 0.585513043217361,
      "epoch": 1.9388330141862324,
      "grad_norm": 0.14997336268424988,
      "learning_rate": 3.537583954365627e-05,
      "loss": 0.5764,
      "mean_token_accuracy": 0.8218410164117813,
      "num_tokens": 672576302.0,
      "step": 21073
    },
    {
      "entropy": 0.5962495459243655,
      "epoch": 1.9389250205576736,
      "grad_norm": 0.15920546650886536,
      "learning_rate": 3.537277271751465e-05,
      "loss": 0.5932,
      "mean_token_accuracy": 0.8138962276279926,
      "num_tokens": 672608182.0,
      "step": 21074
    },
    {
      "entropy": 0.6396163199096918,
      "epoch": 1.9390170269291147,
      "grad_norm": 0.15572337806224823,
      "learning_rate": 3.5369705891373016e-05,
      "loss": 0.6332,
      "mean_token_accuracy": 0.8037468530237675,
      "num_tokens": 672640420.0,
      "step": 21075
    },
    {
      "entropy": 0.5418031474109739,
      "epoch": 1.939109033300556,
      "grad_norm": 0.15634150803089142,
      "learning_rate": 3.536663906523139e-05,
      "loss": 0.524,
      "mean_token_accuracy": 0.834706474095583,
      "num_tokens": 672671491.0,
      "step": 21076
    },
    {
      "entropy": 0.7140256725251675,
      "epoch": 1.9392010396719974,
      "grad_norm": 0.1688416302204132,
      "learning_rate": 3.5363572239089765e-05,
      "loss": 0.7208,
      "mean_token_accuracy": 0.783384557813406,
      "num_tokens": 672703522.0,
      "step": 21077
    },
    {
      "entropy": 0.6379261612892151,
      "epoch": 1.9392930460434386,
      "grad_norm": 0.15758955478668213,
      "learning_rate": 3.536050541294814e-05,
      "loss": 0.6345,
      "mean_token_accuracy": 0.8029408305883408,
      "num_tokens": 672734902.0,
      "step": 21078
    },
    {
      "entropy": 0.518437503837049,
      "epoch": 1.9393850524148797,
      "grad_norm": 0.14304105937480927,
      "learning_rate": 3.535743858680652e-05,
      "loss": 0.5048,
      "mean_token_accuracy": 0.8418732061982155,
      "num_tokens": 672767376.0,
      "step": 21079
    },
    {
      "entropy": 0.4834328384604305,
      "epoch": 1.9394770587863208,
      "grad_norm": 0.14102478325366974,
      "learning_rate": 3.535437176066489e-05,
      "loss": 0.4824,
      "mean_token_accuracy": 0.8468277379870415,
      "num_tokens": 672799095.0,
      "step": 21080
    },
    {
      "entropy": 0.653508416377008,
      "epoch": 1.9395690651577622,
      "grad_norm": 0.15854927897453308,
      "learning_rate": 3.5351304934523264e-05,
      "loss": 0.6451,
      "mean_token_accuracy": 0.8027828857302666,
      "num_tokens": 672831089.0,
      "step": 21081
    },
    {
      "entropy": 0.5784225314855576,
      "epoch": 1.9396610715292035,
      "grad_norm": 0.1548815816640854,
      "learning_rate": 3.534823810838164e-05,
      "loss": 0.5683,
      "mean_token_accuracy": 0.8212367035448551,
      "num_tokens": 672861978.0,
      "step": 21082
    },
    {
      "entropy": 0.5670578982681036,
      "epoch": 1.9397530779006447,
      "grad_norm": 0.15146560966968536,
      "learning_rate": 3.534517128224001e-05,
      "loss": 0.5505,
      "mean_token_accuracy": 0.8315583541989326,
      "num_tokens": 672894314.0,
      "step": 21083
    },
    {
      "entropy": 0.5948806665837765,
      "epoch": 1.9398450842720858,
      "grad_norm": 0.15256443619728088,
      "learning_rate": 3.534210445609839e-05,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8179754763841629,
      "num_tokens": 672926286.0,
      "step": 21084
    },
    {
      "entropy": 0.5539268665015697,
      "epoch": 1.939937090643527,
      "grad_norm": 0.153995543718338,
      "learning_rate": 3.533903762995676e-05,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8308608718216419,
      "num_tokens": 672958701.0,
      "step": 21085
    },
    {
      "entropy": 0.6816992964595556,
      "epoch": 1.9400290970149683,
      "grad_norm": 0.15622027218341827,
      "learning_rate": 3.533597080381513e-05,
      "loss": 0.6777,
      "mean_token_accuracy": 0.7930000200867653,
      "num_tokens": 672991466.0,
      "step": 21086
    },
    {
      "entropy": 0.7099622040987015,
      "epoch": 1.9401211033864096,
      "grad_norm": 0.17000322043895721,
      "learning_rate": 3.533290397767351e-05,
      "loss": 0.7143,
      "mean_token_accuracy": 0.7874784842133522,
      "num_tokens": 673023407.0,
      "step": 21087
    },
    {
      "entropy": 0.6169787200633436,
      "epoch": 1.9402131097578508,
      "grad_norm": 0.1517297476530075,
      "learning_rate": 3.532983715153188e-05,
      "loss": 0.6085,
      "mean_token_accuracy": 0.8138560764491558,
      "num_tokens": 673055993.0,
      "step": 21088
    },
    {
      "entropy": 0.7156086452305317,
      "epoch": 1.940305116129292,
      "grad_norm": 0.16535848379135132,
      "learning_rate": 3.5326770325390254e-05,
      "loss": 0.6995,
      "mean_token_accuracy": 0.7916831448674202,
      "num_tokens": 673087795.0,
      "step": 21089
    },
    {
      "entropy": 0.5574422515928745,
      "epoch": 1.940397122500733,
      "grad_norm": 0.14718714356422424,
      "learning_rate": 3.532370349924863e-05,
      "loss": 0.5417,
      "mean_token_accuracy": 0.8302916660904884,
      "num_tokens": 673119677.0,
      "step": 21090
    },
    {
      "entropy": 0.5159404468722641,
      "epoch": 1.9404891288721744,
      "grad_norm": 0.14048920571804047,
      "learning_rate": 3.5320636673107004e-05,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8433087207376957,
      "num_tokens": 673152115.0,
      "step": 21091
    },
    {
      "entropy": 0.4722126917913556,
      "epoch": 1.9405811352436158,
      "grad_norm": 0.1459016501903534,
      "learning_rate": 3.531756984696538e-05,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8512912504374981,
      "num_tokens": 673183049.0,
      "step": 21092
    },
    {
      "entropy": 0.6596486642956734,
      "epoch": 1.9406731416150569,
      "grad_norm": 0.15640948712825775,
      "learning_rate": 3.531450302082375e-05,
      "loss": 0.645,
      "mean_token_accuracy": 0.8003394193947315,
      "num_tokens": 673214700.0,
      "step": 21093
    },
    {
      "entropy": 0.5302407231647521,
      "epoch": 1.940765147986498,
      "grad_norm": 0.14168362319469452,
      "learning_rate": 3.531143619468212e-05,
      "loss": 0.5141,
      "mean_token_accuracy": 0.84062360227108,
      "num_tokens": 673247265.0,
      "step": 21094
    },
    {
      "entropy": 0.6055217115208507,
      "epoch": 1.9408571543579392,
      "grad_norm": 0.1535567045211792,
      "learning_rate": 3.53083693685405e-05,
      "loss": 0.5792,
      "mean_token_accuracy": 0.8179774545133114,
      "num_tokens": 673278515.0,
      "step": 21095
    },
    {
      "entropy": 0.5196006121113896,
      "epoch": 1.9409491607293805,
      "grad_norm": 0.14893615245819092,
      "learning_rate": 3.530530254239887e-05,
      "loss": 0.493,
      "mean_token_accuracy": 0.8483791202306747,
      "num_tokens": 673310261.0,
      "step": 21096
    },
    {
      "entropy": 0.5973775889724493,
      "epoch": 1.9410411671008219,
      "grad_norm": 0.15082325041294098,
      "learning_rate": 3.5302235716257245e-05,
      "loss": 0.5746,
      "mean_token_accuracy": 0.8194984421133995,
      "num_tokens": 673342549.0,
      "step": 21097
    },
    {
      "entropy": 0.6213447228074074,
      "epoch": 1.941133173472263,
      "grad_norm": 0.15786127746105194,
      "learning_rate": 3.529916889011562e-05,
      "loss": 0.6115,
      "mean_token_accuracy": 0.813865166157484,
      "num_tokens": 673374008.0,
      "step": 21098
    },
    {
      "entropy": 0.57998260948807,
      "epoch": 1.9412251798437041,
      "grad_norm": 0.15159597992897034,
      "learning_rate": 3.5296102063973994e-05,
      "loss": 0.5671,
      "mean_token_accuracy": 0.8225993178784847,
      "num_tokens": 673406094.0,
      "step": 21099
    },
    {
      "entropy": 0.6760702207684517,
      "epoch": 1.9413171862151453,
      "grad_norm": 0.16838495433330536,
      "learning_rate": 3.529303523783237e-05,
      "loss": 0.665,
      "mean_token_accuracy": 0.7956082485616207,
      "num_tokens": 673437434.0,
      "step": 21100
    },
    {
      "entropy": 0.648907845839858,
      "epoch": 1.9414091925865866,
      "grad_norm": 0.1617743819952011,
      "learning_rate": 3.5289968411690743e-05,
      "loss": 0.6405,
      "mean_token_accuracy": 0.8007350526750088,
      "num_tokens": 673468547.0,
      "step": 21101
    },
    {
      "entropy": 0.6757451016455889,
      "epoch": 1.941501198958028,
      "grad_norm": 0.1610538810491562,
      "learning_rate": 3.528690158554912e-05,
      "loss": 0.6689,
      "mean_token_accuracy": 0.7901255562901497,
      "num_tokens": 673499992.0,
      "step": 21102
    },
    {
      "entropy": 0.620975551661104,
      "epoch": 1.941593205329469,
      "grad_norm": 0.15631914138793945,
      "learning_rate": 3.528383475940749e-05,
      "loss": 0.6111,
      "mean_token_accuracy": 0.8123730570077896,
      "num_tokens": 673532051.0,
      "step": 21103
    },
    {
      "entropy": 0.6451004287227988,
      "epoch": 1.9416852117009102,
      "grad_norm": 0.16381248831748962,
      "learning_rate": 3.528076793326586e-05,
      "loss": 0.6585,
      "mean_token_accuracy": 0.7962328419089317,
      "num_tokens": 673564166.0,
      "step": 21104
    },
    {
      "entropy": 0.5340946279466152,
      "epoch": 1.9417772180723514,
      "grad_norm": 0.15023936331272125,
      "learning_rate": 3.527770110712424e-05,
      "loss": 0.534,
      "mean_token_accuracy": 0.8312482796609402,
      "num_tokens": 673596222.0,
      "step": 21105
    },
    {
      "entropy": 0.6827433342114091,
      "epoch": 1.9418692244437927,
      "grad_norm": 0.16161347925662994,
      "learning_rate": 3.527463428098261e-05,
      "loss": 0.6828,
      "mean_token_accuracy": 0.7910895757377148,
      "num_tokens": 673627963.0,
      "step": 21106
    },
    {
      "entropy": 0.7399024683982134,
      "epoch": 1.941961230815234,
      "grad_norm": 0.17195776104927063,
      "learning_rate": 3.5271567454840985e-05,
      "loss": 0.7395,
      "mean_token_accuracy": 0.7784049585461617,
      "num_tokens": 673658147.0,
      "step": 21107
    },
    {
      "entropy": 0.6156222857534885,
      "epoch": 1.9420532371866752,
      "grad_norm": 0.1529000699520111,
      "learning_rate": 3.526850062869936e-05,
      "loss": 0.6095,
      "mean_token_accuracy": 0.811502356082201,
      "num_tokens": 673690253.0,
      "step": 21108
    },
    {
      "entropy": 0.5939344807993621,
      "epoch": 1.9421452435581164,
      "grad_norm": 0.15004222095012665,
      "learning_rate": 3.5265433802557734e-05,
      "loss": 0.5871,
      "mean_token_accuracy": 0.819267313927412,
      "num_tokens": 673722461.0,
      "step": 21109
    },
    {
      "entropy": 0.6097060653846711,
      "epoch": 1.9422372499295575,
      "grad_norm": 0.15976926684379578,
      "learning_rate": 3.526236697641611e-05,
      "loss": 0.5967,
      "mean_token_accuracy": 0.8169668726623058,
      "num_tokens": 673752882.0,
      "step": 21110
    },
    {
      "entropy": 0.6457788739353418,
      "epoch": 1.9423292563009988,
      "grad_norm": 0.1520669013261795,
      "learning_rate": 3.525930015027448e-05,
      "loss": 0.638,
      "mean_token_accuracy": 0.8037924133241177,
      "num_tokens": 673785100.0,
      "step": 21111
    },
    {
      "entropy": 0.5910355187952518,
      "epoch": 1.9424212626724402,
      "grad_norm": 0.15229327976703644,
      "learning_rate": 3.525623332413286e-05,
      "loss": 0.584,
      "mean_token_accuracy": 0.8140804953873158,
      "num_tokens": 673816202.0,
      "step": 21112
    },
    {
      "entropy": 0.5944309905171394,
      "epoch": 1.9425132690438813,
      "grad_norm": 0.15165483951568604,
      "learning_rate": 3.525316649799123e-05,
      "loss": 0.5786,
      "mean_token_accuracy": 0.821059089154005,
      "num_tokens": 673848064.0,
      "step": 21113
    },
    {
      "entropy": 0.6718060686253011,
      "epoch": 1.9426052754153225,
      "grad_norm": 0.1640038639307022,
      "learning_rate": 3.525009967184961e-05,
      "loss": 0.6722,
      "mean_token_accuracy": 0.7942634336650372,
      "num_tokens": 673879726.0,
      "step": 21114
    },
    {
      "entropy": 0.5386967118829489,
      "epoch": 1.9426972817867636,
      "grad_norm": 0.14361542463302612,
      "learning_rate": 3.5247032845707975e-05,
      "loss": 0.5271,
      "mean_token_accuracy": 0.838067252188921,
      "num_tokens": 673912401.0,
      "step": 21115
    },
    {
      "entropy": 0.5571354227140546,
      "epoch": 1.942789288158205,
      "grad_norm": 0.15065163373947144,
      "learning_rate": 3.5243966019566356e-05,
      "loss": 0.5519,
      "mean_token_accuracy": 0.8272191472351551,
      "num_tokens": 673944437.0,
      "step": 21116
    },
    {
      "entropy": 0.6177350906655192,
      "epoch": 1.9428812945296463,
      "grad_norm": 0.15302814543247223,
      "learning_rate": 3.5240899193424724e-05,
      "loss": 0.5949,
      "mean_token_accuracy": 0.8181536681950092,
      "num_tokens": 673976526.0,
      "step": 21117
    },
    {
      "entropy": 0.7474772818386555,
      "epoch": 1.9429733009010874,
      "grad_norm": 0.161213219165802,
      "learning_rate": 3.5237832367283106e-05,
      "loss": 0.7436,
      "mean_token_accuracy": 0.774564404040575,
      "num_tokens": 674008695.0,
      "step": 21118
    },
    {
      "entropy": 0.7545532565563917,
      "epoch": 1.9430653072725286,
      "grad_norm": 0.16054843366146088,
      "learning_rate": 3.5234765541141474e-05,
      "loss": 0.7553,
      "mean_token_accuracy": 0.7738386280834675,
      "num_tokens": 674040143.0,
      "step": 21119
    },
    {
      "entropy": 0.6749900411814451,
      "epoch": 1.9431573136439697,
      "grad_norm": 0.15499627590179443,
      "learning_rate": 3.523169871499985e-05,
      "loss": 0.6682,
      "mean_token_accuracy": 0.798163678497076,
      "num_tokens": 674072354.0,
      "step": 21120
    },
    {
      "entropy": 0.666409807279706,
      "epoch": 1.943249320015411,
      "grad_norm": 0.1540798544883728,
      "learning_rate": 3.522863188885822e-05,
      "loss": 0.6567,
      "mean_token_accuracy": 0.7996412105858326,
      "num_tokens": 674104746.0,
      "step": 21121
    },
    {
      "entropy": 0.5442041754722595,
      "epoch": 1.9433413263868524,
      "grad_norm": 0.15611013770103455,
      "learning_rate": 3.52255650627166e-05,
      "loss": 0.5279,
      "mean_token_accuracy": 0.8323401845991611,
      "num_tokens": 674136472.0,
      "step": 21122
    },
    {
      "entropy": 0.5975530222058296,
      "epoch": 1.9434333327582936,
      "grad_norm": 0.15374962985515594,
      "learning_rate": 3.522249823657497e-05,
      "loss": 0.6013,
      "mean_token_accuracy": 0.8156718872487545,
      "num_tokens": 674168686.0,
      "step": 21123
    },
    {
      "entropy": 0.6092519536614418,
      "epoch": 1.9435253391297347,
      "grad_norm": 0.1519843488931656,
      "learning_rate": 3.521943141043335e-05,
      "loss": 0.6059,
      "mean_token_accuracy": 0.8122321106493473,
      "num_tokens": 674201013.0,
      "step": 21124
    },
    {
      "entropy": 0.5147844322491437,
      "epoch": 1.9436173455011758,
      "grad_norm": 0.1464810073375702,
      "learning_rate": 3.5216364584291715e-05,
      "loss": 0.511,
      "mean_token_accuracy": 0.845318466424942,
      "num_tokens": 674232593.0,
      "step": 21125
    },
    {
      "entropy": 0.4749322235584259,
      "epoch": 1.9437093518726172,
      "grad_norm": 0.14321060478687286,
      "learning_rate": 3.5213297758150096e-05,
      "loss": 0.4568,
      "mean_token_accuracy": 0.8564694561064243,
      "num_tokens": 674264866.0,
      "step": 21126
    },
    {
      "entropy": 0.5724223684519529,
      "epoch": 1.9438013582440585,
      "grad_norm": 0.15480278432369232,
      "learning_rate": 3.5210230932008464e-05,
      "loss": 0.5726,
      "mean_token_accuracy": 0.8279247060418129,
      "num_tokens": 674296783.0,
      "step": 21127
    },
    {
      "entropy": 0.6496510626748204,
      "epoch": 1.9438933646154997,
      "grad_norm": 0.15441520512104034,
      "learning_rate": 3.520716410586684e-05,
      "loss": 0.6256,
      "mean_token_accuracy": 0.8053148649632931,
      "num_tokens": 674329047.0,
      "step": 21128
    },
    {
      "entropy": 0.553263945505023,
      "epoch": 1.9439853709869408,
      "grad_norm": 0.1483747810125351,
      "learning_rate": 3.520409727972521e-05,
      "loss": 0.5411,
      "mean_token_accuracy": 0.830707985907793,
      "num_tokens": 674360550.0,
      "step": 21129
    },
    {
      "entropy": 0.6625782288610935,
      "epoch": 1.944077377358382,
      "grad_norm": 0.15963780879974365,
      "learning_rate": 3.520103045358359e-05,
      "loss": 0.6597,
      "mean_token_accuracy": 0.8017937056720257,
      "num_tokens": 674391124.0,
      "step": 21130
    },
    {
      "entropy": 0.536981095559895,
      "epoch": 1.9441693837298233,
      "grad_norm": 0.14362163841724396,
      "learning_rate": 3.519796362744196e-05,
      "loss": 0.5207,
      "mean_token_accuracy": 0.8377651497721672,
      "num_tokens": 674423541.0,
      "step": 21131
    },
    {
      "entropy": 0.5893666096962988,
      "epoch": 1.9442613901012646,
      "grad_norm": 0.15447543561458588,
      "learning_rate": 3.519489680130034e-05,
      "loss": 0.569,
      "mean_token_accuracy": 0.8202257938683033,
      "num_tokens": 674454811.0,
      "step": 21132
    },
    {
      "entropy": 0.5051316637545824,
      "epoch": 1.9443533964727058,
      "grad_norm": 0.14048807322978973,
      "learning_rate": 3.5191829975158705e-05,
      "loss": 0.4869,
      "mean_token_accuracy": 0.8467642851173878,
      "num_tokens": 674486811.0,
      "step": 21133
    },
    {
      "entropy": 0.6028444338589907,
      "epoch": 1.944445402844147,
      "grad_norm": 0.1495034545660019,
      "learning_rate": 3.5188763149017087e-05,
      "loss": 0.6031,
      "mean_token_accuracy": 0.8159864619374275,
      "num_tokens": 674518967.0,
      "step": 21134
    },
    {
      "entropy": 0.442231684923172,
      "epoch": 1.944537409215588,
      "grad_norm": 0.14409038424491882,
      "learning_rate": 3.5185696322875454e-05,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8652712255716324,
      "num_tokens": 674551006.0,
      "step": 21135
    },
    {
      "entropy": 0.5563779277727008,
      "epoch": 1.9446294155870294,
      "grad_norm": 0.1503056436777115,
      "learning_rate": 3.518262949673383e-05,
      "loss": 0.5423,
      "mean_token_accuracy": 0.8321196250617504,
      "num_tokens": 674583214.0,
      "step": 21136
    },
    {
      "entropy": 0.6214941469952464,
      "epoch": 1.9447214219584708,
      "grad_norm": 0.1600808948278427,
      "learning_rate": 3.5179562670592204e-05,
      "loss": 0.6062,
      "mean_token_accuracy": 0.8131311237812042,
      "num_tokens": 674615176.0,
      "step": 21137
    },
    {
      "entropy": 0.5862725721672177,
      "epoch": 1.9448134283299119,
      "grad_norm": 0.15820875763893127,
      "learning_rate": 3.517649584445058e-05,
      "loss": 0.5693,
      "mean_token_accuracy": 0.8246641382575035,
      "num_tokens": 674647774.0,
      "step": 21138
    },
    {
      "entropy": 0.6366270165890455,
      "epoch": 1.944905434701353,
      "grad_norm": 0.15203215181827545,
      "learning_rate": 3.517342901830895e-05,
      "loss": 0.6275,
      "mean_token_accuracy": 0.8050635941326618,
      "num_tokens": 674680031.0,
      "step": 21139
    },
    {
      "entropy": 0.5892486702650785,
      "epoch": 1.9449974410727942,
      "grad_norm": 0.15298454463481903,
      "learning_rate": 3.517036219216733e-05,
      "loss": 0.5874,
      "mean_token_accuracy": 0.8213777877390385,
      "num_tokens": 674712222.0,
      "step": 21140
    },
    {
      "entropy": 0.563159310258925,
      "epoch": 1.9450894474442355,
      "grad_norm": 0.15539130568504333,
      "learning_rate": 3.51672953660257e-05,
      "loss": 0.5421,
      "mean_token_accuracy": 0.8300041519105434,
      "num_tokens": 674743299.0,
      "step": 21141
    },
    {
      "entropy": 0.5649460600689054,
      "epoch": 1.9451814538156769,
      "grad_norm": 0.14780119061470032,
      "learning_rate": 3.516422853988408e-05,
      "loss": 0.5449,
      "mean_token_accuracy": 0.8303318992257118,
      "num_tokens": 674774666.0,
      "step": 21142
    },
    {
      "entropy": 0.6922942008823156,
      "epoch": 1.945273460187118,
      "grad_norm": 0.1567469835281372,
      "learning_rate": 3.516116171374245e-05,
      "loss": 0.6669,
      "mean_token_accuracy": 0.7952334955334663,
      "num_tokens": 674806662.0,
      "step": 21143
    },
    {
      "entropy": 0.6413060482591391,
      "epoch": 1.9453654665585591,
      "grad_norm": 0.15748682618141174,
      "learning_rate": 3.5158094887600826e-05,
      "loss": 0.631,
      "mean_token_accuracy": 0.8018411807715893,
      "num_tokens": 674838877.0,
      "step": 21144
    },
    {
      "entropy": 0.49851447995752096,
      "epoch": 1.9454574729300003,
      "grad_norm": 0.1481986939907074,
      "learning_rate": 3.51550280614592e-05,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8468871302902699,
      "num_tokens": 674870974.0,
      "step": 21145
    },
    {
      "entropy": 0.5872944509610534,
      "epoch": 1.9455494793014416,
      "grad_norm": 0.14673027396202087,
      "learning_rate": 3.515196123531757e-05,
      "loss": 0.5797,
      "mean_token_accuracy": 0.8217665702104568,
      "num_tokens": 674903484.0,
      "step": 21146
    },
    {
      "entropy": 0.6022896114736795,
      "epoch": 1.945641485672883,
      "grad_norm": 0.15218332409858704,
      "learning_rate": 3.514889440917595e-05,
      "loss": 0.6034,
      "mean_token_accuracy": 0.8126121014356613,
      "num_tokens": 674935903.0,
      "step": 21147
    },
    {
      "entropy": 0.49689415376633406,
      "epoch": 1.945733492044324,
      "grad_norm": 0.14688187837600708,
      "learning_rate": 3.514582758303432e-05,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8453241661190987,
      "num_tokens": 674968242.0,
      "step": 21148
    },
    {
      "entropy": 0.6492391750216484,
      "epoch": 1.9458254984157652,
      "grad_norm": 0.15737171471118927,
      "learning_rate": 3.514276075689269e-05,
      "loss": 0.6349,
      "mean_token_accuracy": 0.8067057467997074,
      "num_tokens": 675000701.0,
      "step": 21149
    },
    {
      "entropy": 0.5093914568424225,
      "epoch": 1.9459175047872064,
      "grad_norm": 0.15608787536621094,
      "learning_rate": 3.513969393075107e-05,
      "loss": 0.4932,
      "mean_token_accuracy": 0.8409525901079178,
      "num_tokens": 675032245.0,
      "step": 21150
    },
    {
      "entropy": 0.7050650212913752,
      "epoch": 1.9460095111586477,
      "grad_norm": 0.15808387100696564,
      "learning_rate": 3.513662710460944e-05,
      "loss": 0.6954,
      "mean_token_accuracy": 0.7871873341500759,
      "num_tokens": 675064734.0,
      "step": 21151
    },
    {
      "entropy": 0.5956526603549719,
      "epoch": 1.946101517530089,
      "grad_norm": 0.1496989130973816,
      "learning_rate": 3.5133560278467817e-05,
      "loss": 0.5783,
      "mean_token_accuracy": 0.821366623044014,
      "num_tokens": 675096099.0,
      "step": 21152
    },
    {
      "entropy": 0.7196011878550053,
      "epoch": 1.9461935239015302,
      "grad_norm": 0.16480940580368042,
      "learning_rate": 3.513049345232619e-05,
      "loss": 0.7083,
      "mean_token_accuracy": 0.787197720259428,
      "num_tokens": 675127861.0,
      "step": 21153
    },
    {
      "entropy": 0.6184890419244766,
      "epoch": 1.9462855302729714,
      "grad_norm": 0.16054661571979523,
      "learning_rate": 3.512742662618456e-05,
      "loss": 0.5961,
      "mean_token_accuracy": 0.8150029331445694,
      "num_tokens": 675159240.0,
      "step": 21154
    },
    {
      "entropy": 0.5856940457597375,
      "epoch": 1.9463775366444125,
      "grad_norm": 0.15374989807605743,
      "learning_rate": 3.512435980004294e-05,
      "loss": 0.577,
      "mean_token_accuracy": 0.8259494677186012,
      "num_tokens": 675190632.0,
      "step": 21155
    },
    {
      "entropy": 0.6671400079503655,
      "epoch": 1.9464695430158538,
      "grad_norm": 0.15164706110954285,
      "learning_rate": 3.512129297390131e-05,
      "loss": 0.6596,
      "mean_token_accuracy": 0.7945193685591221,
      "num_tokens": 675222892.0,
      "step": 21156
    },
    {
      "entropy": 0.46430599177256227,
      "epoch": 1.9465615493872952,
      "grad_norm": 0.1323835700750351,
      "learning_rate": 3.511822614775968e-05,
      "loss": 0.455,
      "mean_token_accuracy": 0.8569858260452747,
      "num_tokens": 675255660.0,
      "step": 21157
    },
    {
      "entropy": 0.741547167301178,
      "epoch": 1.9466535557587363,
      "grad_norm": 0.1702246069908142,
      "learning_rate": 3.511515932161806e-05,
      "loss": 0.7411,
      "mean_token_accuracy": 0.7773579470813274,
      "num_tokens": 675287620.0,
      "step": 21158
    },
    {
      "entropy": 0.4852101863361895,
      "epoch": 1.9467455621301775,
      "grad_norm": 0.14732863008975983,
      "learning_rate": 3.511209249547643e-05,
      "loss": 0.4692,
      "mean_token_accuracy": 0.8549586869776249,
      "num_tokens": 675319457.0,
      "step": 21159
    },
    {
      "entropy": 0.5897373594343662,
      "epoch": 1.9468375685016186,
      "grad_norm": 0.15191657841205597,
      "learning_rate": 3.510902566933481e-05,
      "loss": 0.5846,
      "mean_token_accuracy": 0.822922557592392,
      "num_tokens": 675351111.0,
      "step": 21160
    },
    {
      "entropy": 0.62624466791749,
      "epoch": 1.94692957487306,
      "grad_norm": 0.15380829572677612,
      "learning_rate": 3.510595884319318e-05,
      "loss": 0.6131,
      "mean_token_accuracy": 0.8113961927592754,
      "num_tokens": 675382583.0,
      "step": 21161
    },
    {
      "entropy": 0.5578221217729151,
      "epoch": 1.9470215812445013,
      "grad_norm": 0.14792168140411377,
      "learning_rate": 3.5102892017051556e-05,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8269552029669285,
      "num_tokens": 675414554.0,
      "step": 21162
    },
    {
      "entropy": 0.5837977407500148,
      "epoch": 1.9471135876159424,
      "grad_norm": 0.14814646542072296,
      "learning_rate": 3.509982519090993e-05,
      "loss": 0.5752,
      "mean_token_accuracy": 0.8226541541516781,
      "num_tokens": 675447043.0,
      "step": 21163
    },
    {
      "entropy": 0.673289836384356,
      "epoch": 1.9472055939873836,
      "grad_norm": 0.16611330211162567,
      "learning_rate": 3.50967583647683e-05,
      "loss": 0.671,
      "mean_token_accuracy": 0.7966173961758614,
      "num_tokens": 675478492.0,
      "step": 21164
    },
    {
      "entropy": 0.542772464454174,
      "epoch": 1.9472976003588247,
      "grad_norm": 0.14890050888061523,
      "learning_rate": 3.509369153862668e-05,
      "loss": 0.5372,
      "mean_token_accuracy": 0.8326513916254044,
      "num_tokens": 675510850.0,
      "step": 21165
    },
    {
      "entropy": 0.6253093536943197,
      "epoch": 1.947389606730266,
      "grad_norm": 0.1569082885980606,
      "learning_rate": 3.509062471248505e-05,
      "loss": 0.6164,
      "mean_token_accuracy": 0.8106536455452442,
      "num_tokens": 675542804.0,
      "step": 21166
    },
    {
      "entropy": 0.6433787336573005,
      "epoch": 1.9474816131017074,
      "grad_norm": 0.1537175178527832,
      "learning_rate": 3.508755788634342e-05,
      "loss": 0.6453,
      "mean_token_accuracy": 0.8027309887111187,
      "num_tokens": 675575428.0,
      "step": 21167
    },
    {
      "entropy": 0.5890108235180378,
      "epoch": 1.9475736194731486,
      "grad_norm": 0.1486189216375351,
      "learning_rate": 3.50844910602018e-05,
      "loss": 0.5726,
      "mean_token_accuracy": 0.8213592544198036,
      "num_tokens": 675607209.0,
      "step": 21168
    },
    {
      "entropy": 0.5913401013240218,
      "epoch": 1.9476656258445897,
      "grad_norm": 0.159102663397789,
      "learning_rate": 3.508142423406017e-05,
      "loss": 0.5893,
      "mean_token_accuracy": 0.8160383626818657,
      "num_tokens": 675638561.0,
      "step": 21169
    },
    {
      "entropy": 0.59797528013587,
      "epoch": 1.9477576322160308,
      "grad_norm": 0.149477019906044,
      "learning_rate": 3.507835740791855e-05,
      "loss": 0.5836,
      "mean_token_accuracy": 0.8220556564629078,
      "num_tokens": 675670821.0,
      "step": 21170
    },
    {
      "entropy": 0.49406406469643116,
      "epoch": 1.9478496385874722,
      "grad_norm": 0.1428799033164978,
      "learning_rate": 3.507529058177692e-05,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8494922630488873,
      "num_tokens": 675702860.0,
      "step": 21171
    },
    {
      "entropy": 0.611619686242193,
      "epoch": 1.9479416449589135,
      "grad_norm": 0.15272106230258942,
      "learning_rate": 3.5072223755635296e-05,
      "loss": 0.6026,
      "mean_token_accuracy": 0.8139804601669312,
      "num_tokens": 675734867.0,
      "step": 21172
    },
    {
      "entropy": 0.537699393928051,
      "epoch": 1.9480336513303547,
      "grad_norm": 0.1464865803718567,
      "learning_rate": 3.506915692949367e-05,
      "loss": 0.5312,
      "mean_token_accuracy": 0.8372976519167423,
      "num_tokens": 675766388.0,
      "step": 21173
    },
    {
      "entropy": 0.6376993958838284,
      "epoch": 1.9481256577017958,
      "grad_norm": 0.1539304256439209,
      "learning_rate": 3.5066090103352045e-05,
      "loss": 0.6206,
      "mean_token_accuracy": 0.8109608814120293,
      "num_tokens": 675797650.0,
      "step": 21174
    },
    {
      "entropy": 0.529043355025351,
      "epoch": 1.948217664073237,
      "grad_norm": 0.15068072080612183,
      "learning_rate": 3.506302327721041e-05,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8395100347697735,
      "num_tokens": 675829460.0,
      "step": 21175
    },
    {
      "entropy": 0.5135677275247872,
      "epoch": 1.9483096704446783,
      "grad_norm": 0.14589446783065796,
      "learning_rate": 3.5059956451068795e-05,
      "loss": 0.4866,
      "mean_token_accuracy": 0.852164451032877,
      "num_tokens": 675861833.0,
      "step": 21176
    },
    {
      "entropy": 0.5956229725852609,
      "epoch": 1.9484016768161196,
      "grad_norm": 0.14699962735176086,
      "learning_rate": 3.505688962492716e-05,
      "loss": 0.5842,
      "mean_token_accuracy": 0.8198562748730183,
      "num_tokens": 675893931.0,
      "step": 21177
    },
    {
      "entropy": 0.5675919177010655,
      "epoch": 1.9484936831875608,
      "grad_norm": 0.14698660373687744,
      "learning_rate": 3.5053822798785544e-05,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8343793004751205,
      "num_tokens": 675926467.0,
      "step": 21178
    },
    {
      "entropy": 0.472188837826252,
      "epoch": 1.948585689559002,
      "grad_norm": 0.14754116535186768,
      "learning_rate": 3.505075597264391e-05,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8517677113413811,
      "num_tokens": 675957945.0,
      "step": 21179
    },
    {
      "entropy": 0.4716261774301529,
      "epoch": 1.948677695930443,
      "grad_norm": 0.138109490275383,
      "learning_rate": 3.5047689146502286e-05,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8530156388878822,
      "num_tokens": 675989895.0,
      "step": 21180
    },
    {
      "entropy": 0.542541990056634,
      "epoch": 1.9487697023018844,
      "grad_norm": 0.14492376148700714,
      "learning_rate": 3.504462232036066e-05,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8333529345691204,
      "num_tokens": 676021516.0,
      "step": 21181
    },
    {
      "entropy": 0.6487901639193296,
      "epoch": 1.9488617086733258,
      "grad_norm": 0.15455259382724762,
      "learning_rate": 3.5041555494219036e-05,
      "loss": 0.6422,
      "mean_token_accuracy": 0.8106396980583668,
      "num_tokens": 676053821.0,
      "step": 21182
    },
    {
      "entropy": 0.6137834191322327,
      "epoch": 1.9489537150447669,
      "grad_norm": 0.15304484963417053,
      "learning_rate": 3.503848866807741e-05,
      "loss": 0.5938,
      "mean_token_accuracy": 0.8150101453065872,
      "num_tokens": 676085131.0,
      "step": 21183
    },
    {
      "entropy": 0.5170519235543907,
      "epoch": 1.949045721416208,
      "grad_norm": 0.1537398248910904,
      "learning_rate": 3.5035421841935785e-05,
      "loss": 0.5101,
      "mean_token_accuracy": 0.8428502380847931,
      "num_tokens": 676116727.0,
      "step": 21184
    },
    {
      "entropy": 0.6346636340022087,
      "epoch": 1.9491377277876492,
      "grad_norm": 0.16223998367786407,
      "learning_rate": 3.503235501579415e-05,
      "loss": 0.6329,
      "mean_token_accuracy": 0.8109869584441185,
      "num_tokens": 676148906.0,
      "step": 21185
    },
    {
      "entropy": 0.7100310921669006,
      "epoch": 1.9492297341590905,
      "grad_norm": 0.16653968393802643,
      "learning_rate": 3.5029288189652534e-05,
      "loss": 0.722,
      "mean_token_accuracy": 0.7839757800102234,
      "num_tokens": 676180647.0,
      "step": 21186
    },
    {
      "entropy": 0.6078567039221525,
      "epoch": 1.9493217405305319,
      "grad_norm": 0.15427471697330475,
      "learning_rate": 3.50262213635109e-05,
      "loss": 0.6039,
      "mean_token_accuracy": 0.81739367172122,
      "num_tokens": 676212490.0,
      "step": 21187
    },
    {
      "entropy": 0.5414308458566666,
      "epoch": 1.949413746901973,
      "grad_norm": 0.14683759212493896,
      "learning_rate": 3.502315453736928e-05,
      "loss": 0.5371,
      "mean_token_accuracy": 0.8359721899032593,
      "num_tokens": 676244336.0,
      "step": 21188
    },
    {
      "entropy": 0.6083430470898747,
      "epoch": 1.9495057532734141,
      "grad_norm": 0.15204177796840668,
      "learning_rate": 3.502008771122765e-05,
      "loss": 0.6025,
      "mean_token_accuracy": 0.8184302039444447,
      "num_tokens": 676277006.0,
      "step": 21189
    },
    {
      "entropy": 0.5111573999747634,
      "epoch": 1.9495977596448553,
      "grad_norm": 0.14269046485424042,
      "learning_rate": 3.5017020885086026e-05,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8475249223411083,
      "num_tokens": 676309323.0,
      "step": 21190
    },
    {
      "entropy": 0.6324782650917768,
      "epoch": 1.9496897660162966,
      "grad_norm": 0.15132418274879456,
      "learning_rate": 3.50139540589444e-05,
      "loss": 0.6252,
      "mean_token_accuracy": 0.8084431663155556,
      "num_tokens": 676341938.0,
      "step": 21191
    },
    {
      "entropy": 0.5739279799163342,
      "epoch": 1.949781772387738,
      "grad_norm": 0.15161633491516113,
      "learning_rate": 3.5010887232802775e-05,
      "loss": 0.557,
      "mean_token_accuracy": 0.8249202258884907,
      "num_tokens": 676373692.0,
      "step": 21192
    },
    {
      "entropy": 0.6871925899758935,
      "epoch": 1.949873778759179,
      "grad_norm": 0.16307055950164795,
      "learning_rate": 3.500782040666114e-05,
      "loss": 0.6751,
      "mean_token_accuracy": 0.7938744761049747,
      "num_tokens": 676405537.0,
      "step": 21193
    },
    {
      "entropy": 0.5077749695628881,
      "epoch": 1.9499657851306202,
      "grad_norm": 0.14549468457698822,
      "learning_rate": 3.5004753580519525e-05,
      "loss": 0.4946,
      "mean_token_accuracy": 0.8444218561053276,
      "num_tokens": 676437931.0,
      "step": 21194
    },
    {
      "entropy": 0.6611486207693815,
      "epoch": 1.9500577915020614,
      "grad_norm": 0.15591122210025787,
      "learning_rate": 3.500168675437789e-05,
      "loss": 0.6599,
      "mean_token_accuracy": 0.7998425029218197,
      "num_tokens": 676470251.0,
      "step": 21195
    },
    {
      "entropy": 0.5718307960778475,
      "epoch": 1.9501497978735027,
      "grad_norm": 0.14708861708641052,
      "learning_rate": 3.499861992823627e-05,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8277875781059265,
      "num_tokens": 676502897.0,
      "step": 21196
    },
    {
      "entropy": 0.6186590623110533,
      "epoch": 1.950241804244944,
      "grad_norm": 0.1577519178390503,
      "learning_rate": 3.499555310209464e-05,
      "loss": 0.6154,
      "mean_token_accuracy": 0.8105944879353046,
      "num_tokens": 676534196.0,
      "step": 21197
    },
    {
      "entropy": 0.632647879421711,
      "epoch": 1.9503338106163852,
      "grad_norm": 0.15539008378982544,
      "learning_rate": 3.4992486275953017e-05,
      "loss": 0.6224,
      "mean_token_accuracy": 0.8094610832631588,
      "num_tokens": 676566579.0,
      "step": 21198
    },
    {
      "entropy": 0.5532479304820299,
      "epoch": 1.9504258169878264,
      "grad_norm": 0.14991913735866547,
      "learning_rate": 3.498941944981139e-05,
      "loss": 0.5452,
      "mean_token_accuracy": 0.833648607134819,
      "num_tokens": 676598114.0,
      "step": 21199
    },
    {
      "entropy": 0.5584917804226279,
      "epoch": 1.9505178233592675,
      "grad_norm": 0.15014797449111938,
      "learning_rate": 3.4986352623669766e-05,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8279544673860073,
      "num_tokens": 676629643.0,
      "step": 21200
    },
    {
      "entropy": 0.6265275795012712,
      "epoch": 1.9506098297307088,
      "grad_norm": 0.15130209922790527,
      "learning_rate": 3.498328579752814e-05,
      "loss": 0.6165,
      "mean_token_accuracy": 0.8089497685432434,
      "num_tokens": 676661528.0,
      "step": 21201
    },
    {
      "entropy": 0.549820520915091,
      "epoch": 1.9507018361021502,
      "grad_norm": 0.14475257694721222,
      "learning_rate": 3.4980218971386515e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8376443162560463,
      "num_tokens": 676692815.0,
      "step": 21202
    },
    {
      "entropy": 0.5840404834598303,
      "epoch": 1.9507938424735913,
      "grad_norm": 0.1532512605190277,
      "learning_rate": 3.497715214524489e-05,
      "loss": 0.5778,
      "mean_token_accuracy": 0.8236419335007668,
      "num_tokens": 676725168.0,
      "step": 21203
    },
    {
      "entropy": 0.5563944801688194,
      "epoch": 1.9508858488450325,
      "grad_norm": 0.1493322253227234,
      "learning_rate": 3.4974085319103264e-05,
      "loss": 0.5508,
      "mean_token_accuracy": 0.8297654017806053,
      "num_tokens": 676756979.0,
      "step": 21204
    },
    {
      "entropy": 0.6069528949446976,
      "epoch": 1.9509778552164736,
      "grad_norm": 0.15925318002700806,
      "learning_rate": 3.497101849296164e-05,
      "loss": 0.579,
      "mean_token_accuracy": 0.8168859295547009,
      "num_tokens": 676787399.0,
      "step": 21205
    },
    {
      "entropy": 0.5991376873571426,
      "epoch": 1.951069861587915,
      "grad_norm": 0.15710243582725525,
      "learning_rate": 3.496795166682001e-05,
      "loss": 0.5856,
      "mean_token_accuracy": 0.8212700001895428,
      "num_tokens": 676819063.0,
      "step": 21206
    },
    {
      "entropy": 0.5743237130809575,
      "epoch": 1.9511618679593563,
      "grad_norm": 0.14911198616027832,
      "learning_rate": 3.496488484067839e-05,
      "loss": 0.5599,
      "mean_token_accuracy": 0.8241652883589268,
      "num_tokens": 676851310.0,
      "step": 21207
    },
    {
      "entropy": 0.5117264855653048,
      "epoch": 1.9512538743307974,
      "grad_norm": 0.13848719000816345,
      "learning_rate": 3.4961818014536756e-05,
      "loss": 0.498,
      "mean_token_accuracy": 0.8423976413905621,
      "num_tokens": 676883358.0,
      "step": 21208
    },
    {
      "entropy": 0.5458907303400338,
      "epoch": 1.9513458807022386,
      "grad_norm": 0.1522660106420517,
      "learning_rate": 3.495875118839513e-05,
      "loss": 0.5325,
      "mean_token_accuracy": 0.8359706439077854,
      "num_tokens": 676915446.0,
      "step": 21209
    },
    {
      "entropy": 0.48553877836093307,
      "epoch": 1.9514378870736797,
      "grad_norm": 0.14364679157733917,
      "learning_rate": 3.4955684362253506e-05,
      "loss": 0.474,
      "mean_token_accuracy": 0.8484277911484241,
      "num_tokens": 676946404.0,
      "step": 21210
    },
    {
      "entropy": 0.5175825217738748,
      "epoch": 1.951529893445121,
      "grad_norm": 0.1450655460357666,
      "learning_rate": 3.495261753611188e-05,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8413665629923344,
      "num_tokens": 676978812.0,
      "step": 21211
    },
    {
      "entropy": 0.5747079723514616,
      "epoch": 1.9516218998165624,
      "grad_norm": 0.1451064944267273,
      "learning_rate": 3.4949550709970255e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8281086832284927,
      "num_tokens": 677011545.0,
      "step": 21212
    },
    {
      "entropy": 0.6787541005760431,
      "epoch": 1.9517139061880036,
      "grad_norm": 0.16250883042812347,
      "learning_rate": 3.494648388382863e-05,
      "loss": 0.6734,
      "mean_token_accuracy": 0.7946202792227268,
      "num_tokens": 677042376.0,
      "step": 21213
    },
    {
      "entropy": 0.5537306685000658,
      "epoch": 1.9518059125594447,
      "grad_norm": 0.15378151834011078,
      "learning_rate": 3.4943417057687e-05,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8348008915781975,
      "num_tokens": 677074104.0,
      "step": 21214
    },
    {
      "entropy": 0.5585094783455133,
      "epoch": 1.9518979189308858,
      "grad_norm": 0.15257161855697632,
      "learning_rate": 3.494035023154538e-05,
      "loss": 0.546,
      "mean_token_accuracy": 0.8297864012420177,
      "num_tokens": 677105894.0,
      "step": 21215
    },
    {
      "entropy": 0.6775063620880246,
      "epoch": 1.9519899253023272,
      "grad_norm": 0.15650996565818787,
      "learning_rate": 3.493728340540375e-05,
      "loss": 0.6692,
      "mean_token_accuracy": 0.7961771599948406,
      "num_tokens": 677138481.0,
      "step": 21216
    },
    {
      "entropy": 0.626062486320734,
      "epoch": 1.9520819316737685,
      "grad_norm": 0.15326909720897675,
      "learning_rate": 3.493421657926212e-05,
      "loss": 0.6041,
      "mean_token_accuracy": 0.8138752728700638,
      "num_tokens": 677169502.0,
      "step": 21217
    },
    {
      "entropy": 0.5695986300706863,
      "epoch": 1.9521739380452097,
      "grad_norm": 0.14536736905574799,
      "learning_rate": 3.4931149753120496e-05,
      "loss": 0.5517,
      "mean_token_accuracy": 0.8281738720834255,
      "num_tokens": 677201955.0,
      "step": 21218
    },
    {
      "entropy": 0.5839372994378209,
      "epoch": 1.9522659444166508,
      "grad_norm": 0.15014348924160004,
      "learning_rate": 3.492808292697887e-05,
      "loss": 0.5672,
      "mean_token_accuracy": 0.8245483264327049,
      "num_tokens": 677233910.0,
      "step": 21219
    },
    {
      "entropy": 0.6591052152216434,
      "epoch": 1.952357950788092,
      "grad_norm": 0.1705188751220703,
      "learning_rate": 3.4925016100837245e-05,
      "loss": 0.654,
      "mean_token_accuracy": 0.7977685779333115,
      "num_tokens": 677266351.0,
      "step": 21220
    },
    {
      "entropy": 0.5695772133767605,
      "epoch": 1.9524499571595333,
      "grad_norm": 0.1543361395597458,
      "learning_rate": 3.492194927469562e-05,
      "loss": 0.5667,
      "mean_token_accuracy": 0.8257440812885761,
      "num_tokens": 677298591.0,
      "step": 21221
    },
    {
      "entropy": 0.6473557855933905,
      "epoch": 1.9525419635309746,
      "grad_norm": 0.1707785427570343,
      "learning_rate": 3.4918882448553995e-05,
      "loss": 0.6361,
      "mean_token_accuracy": 0.8045564591884613,
      "num_tokens": 677329363.0,
      "step": 21222
    },
    {
      "entropy": 0.5951556768268347,
      "epoch": 1.9526339699024158,
      "grad_norm": 0.1568307876586914,
      "learning_rate": 3.491581562241237e-05,
      "loss": 0.5892,
      "mean_token_accuracy": 0.8159884549677372,
      "num_tokens": 677360400.0,
      "step": 21223
    },
    {
      "entropy": 0.5599502557888627,
      "epoch": 1.952725976273857,
      "grad_norm": 0.15161392092704773,
      "learning_rate": 3.491274879627074e-05,
      "loss": 0.543,
      "mean_token_accuracy": 0.8337754271924496,
      "num_tokens": 677392231.0,
      "step": 21224
    },
    {
      "entropy": 0.6476523838937283,
      "epoch": 1.952817982645298,
      "grad_norm": 0.15757815539836884,
      "learning_rate": 3.490968197012912e-05,
      "loss": 0.6188,
      "mean_token_accuracy": 0.8074588440358639,
      "num_tokens": 677424071.0,
      "step": 21225
    },
    {
      "entropy": 0.6206804579123855,
      "epoch": 1.9529099890167394,
      "grad_norm": 0.15337297320365906,
      "learning_rate": 3.4906615143987486e-05,
      "loss": 0.6191,
      "mean_token_accuracy": 0.8105106018483639,
      "num_tokens": 677456260.0,
      "step": 21226
    },
    {
      "entropy": 0.6940821809694171,
      "epoch": 1.9530019953881808,
      "grad_norm": 0.17024868726730347,
      "learning_rate": 3.490354831784586e-05,
      "loss": 0.6932,
      "mean_token_accuracy": 0.7883101031184196,
      "num_tokens": 677487717.0,
      "step": 21227
    },
    {
      "entropy": 0.48105247411876917,
      "epoch": 1.953094001759622,
      "grad_norm": 0.15024365484714508,
      "learning_rate": 3.4900481491704236e-05,
      "loss": 0.4773,
      "mean_token_accuracy": 0.8507700935006142,
      "num_tokens": 677519780.0,
      "step": 21228
    },
    {
      "entropy": 0.48652282916009426,
      "epoch": 1.953186008131063,
      "grad_norm": 0.14300437271595,
      "learning_rate": 3.489741466556261e-05,
      "loss": 0.4782,
      "mean_token_accuracy": 0.849780723452568,
      "num_tokens": 677551788.0,
      "step": 21229
    },
    {
      "entropy": 0.6724676822777838,
      "epoch": 1.9532780145025042,
      "grad_norm": 0.1620689034461975,
      "learning_rate": 3.4894347839420985e-05,
      "loss": 0.6704,
      "mean_token_accuracy": 0.8030777834355831,
      "num_tokens": 677583606.0,
      "step": 21230
    },
    {
      "entropy": 0.5905240904539824,
      "epoch": 1.9533700208739455,
      "grad_norm": 0.1523929238319397,
      "learning_rate": 3.489128101327936e-05,
      "loss": 0.5862,
      "mean_token_accuracy": 0.8204774968326092,
      "num_tokens": 677615593.0,
      "step": 21231
    },
    {
      "entropy": 0.583328727632761,
      "epoch": 1.9534620272453869,
      "grad_norm": 0.14860589802265167,
      "learning_rate": 3.4888214187137734e-05,
      "loss": 0.563,
      "mean_token_accuracy": 0.8273405395448208,
      "num_tokens": 677647623.0,
      "step": 21232
    },
    {
      "entropy": 0.6145699378103018,
      "epoch": 1.953554033616828,
      "grad_norm": 0.1575554609298706,
      "learning_rate": 3.488514736099611e-05,
      "loss": 0.6022,
      "mean_token_accuracy": 0.8141400031745434,
      "num_tokens": 677678994.0,
      "step": 21233
    },
    {
      "entropy": 0.6067977864295244,
      "epoch": 1.9536460399882691,
      "grad_norm": 0.15616698563098907,
      "learning_rate": 3.4882080534854483e-05,
      "loss": 0.6117,
      "mean_token_accuracy": 0.8104995712637901,
      "num_tokens": 677710886.0,
      "step": 21234
    },
    {
      "entropy": 0.6834138967096806,
      "epoch": 1.9537380463597103,
      "grad_norm": 0.16220968961715698,
      "learning_rate": 3.487901370871285e-05,
      "loss": 0.6654,
      "mean_token_accuracy": 0.7973011396825314,
      "num_tokens": 677742606.0,
      "step": 21235
    },
    {
      "entropy": 0.5547855091281235,
      "epoch": 1.9538300527311516,
      "grad_norm": 0.1493922919034958,
      "learning_rate": 3.487594688257123e-05,
      "loss": 0.546,
      "mean_token_accuracy": 0.8308116160333157,
      "num_tokens": 677775004.0,
      "step": 21236
    },
    {
      "entropy": 0.5371629110304639,
      "epoch": 1.953922059102593,
      "grad_norm": 0.15950892865657806,
      "learning_rate": 3.48728800564296e-05,
      "loss": 0.5389,
      "mean_token_accuracy": 0.8359327428042889,
      "num_tokens": 677807065.0,
      "step": 21237
    },
    {
      "entropy": 0.6824213331565261,
      "epoch": 1.9540140654740341,
      "grad_norm": 0.16428346931934357,
      "learning_rate": 3.486981323028798e-05,
      "loss": 0.6687,
      "mean_token_accuracy": 0.7963108047842979,
      "num_tokens": 677838832.0,
      "step": 21238
    },
    {
      "entropy": 0.5519502102397382,
      "epoch": 1.9541060718454752,
      "grad_norm": 0.14784544706344604,
      "learning_rate": 3.486674640414635e-05,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8371313288807869,
      "num_tokens": 677870816.0,
      "step": 21239
    },
    {
      "entropy": 0.6797834057360888,
      "epoch": 1.9541980782169164,
      "grad_norm": 0.1655811071395874,
      "learning_rate": 3.4863679578004725e-05,
      "loss": 0.6722,
      "mean_token_accuracy": 0.7952373884618282,
      "num_tokens": 677901833.0,
      "step": 21240
    },
    {
      "entropy": 0.6036544693633914,
      "epoch": 1.9542900845883577,
      "grad_norm": 0.15236428380012512,
      "learning_rate": 3.48606127518631e-05,
      "loss": 0.5823,
      "mean_token_accuracy": 0.8254096731543541,
      "num_tokens": 677933682.0,
      "step": 21241
    },
    {
      "entropy": 0.5484328456223011,
      "epoch": 1.954382090959799,
      "grad_norm": 0.14355501532554626,
      "learning_rate": 3.4857545925721474e-05,
      "loss": 0.532,
      "mean_token_accuracy": 0.8326819688081741,
      "num_tokens": 677966450.0,
      "step": 21242
    },
    {
      "entropy": 0.590557225048542,
      "epoch": 1.9544740973312402,
      "grad_norm": 0.14971302449703217,
      "learning_rate": 3.485447909957985e-05,
      "loss": 0.5856,
      "mean_token_accuracy": 0.8249207995831966,
      "num_tokens": 677998939.0,
      "step": 21243
    },
    {
      "entropy": 0.5929172728210688,
      "epoch": 1.9545661037026814,
      "grad_norm": 0.1550942063331604,
      "learning_rate": 3.485141227343822e-05,
      "loss": 0.561,
      "mean_token_accuracy": 0.8244663737714291,
      "num_tokens": 678030307.0,
      "step": 21244
    },
    {
      "entropy": 0.560783464461565,
      "epoch": 1.9546581100741225,
      "grad_norm": 0.1509559601545334,
      "learning_rate": 3.484834544729659e-05,
      "loss": 0.5488,
      "mean_token_accuracy": 0.8292842656373978,
      "num_tokens": 678061986.0,
      "step": 21245
    },
    {
      "entropy": 0.605756614357233,
      "epoch": 1.9547501164455638,
      "grad_norm": 0.15265458822250366,
      "learning_rate": 3.484527862115497e-05,
      "loss": 0.6052,
      "mean_token_accuracy": 0.8135075345635414,
      "num_tokens": 678093459.0,
      "step": 21246
    },
    {
      "entropy": 0.6086089685559273,
      "epoch": 1.9548421228170052,
      "grad_norm": 0.1578754037618637,
      "learning_rate": 3.484221179501334e-05,
      "loss": 0.6043,
      "mean_token_accuracy": 0.8119109869003296,
      "num_tokens": 678125868.0,
      "step": 21247
    },
    {
      "entropy": 0.5708864433690906,
      "epoch": 1.9549341291884463,
      "grad_norm": 0.14788368344306946,
      "learning_rate": 3.4839144968871715e-05,
      "loss": 0.5594,
      "mean_token_accuracy": 0.8272939510643482,
      "num_tokens": 678158124.0,
      "step": 21248
    },
    {
      "entropy": 0.5573158585466444,
      "epoch": 1.9550261355598875,
      "grad_norm": 0.1528993546962738,
      "learning_rate": 3.483607814273009e-05,
      "loss": 0.5556,
      "mean_token_accuracy": 0.8338116891682148,
      "num_tokens": 678189891.0,
      "step": 21249
    },
    {
      "entropy": 0.6562768314033747,
      "epoch": 1.9551181419313286,
      "grad_norm": 0.161867156624794,
      "learning_rate": 3.4833011316588464e-05,
      "loss": 0.6538,
      "mean_token_accuracy": 0.8050266169011593,
      "num_tokens": 678220953.0,
      "step": 21250
    },
    {
      "entropy": 0.6099745621904731,
      "epoch": 1.95521014830277,
      "grad_norm": 0.15474963188171387,
      "learning_rate": 3.482994449044684e-05,
      "loss": 0.5947,
      "mean_token_accuracy": 0.8168131150305271,
      "num_tokens": 678253024.0,
      "step": 21251
    },
    {
      "entropy": 0.5209232196211815,
      "epoch": 1.9553021546742113,
      "grad_norm": 0.1522378921508789,
      "learning_rate": 3.4826877664305214e-05,
      "loss": 0.5131,
      "mean_token_accuracy": 0.83986846357584,
      "num_tokens": 678284126.0,
      "step": 21252
    },
    {
      "entropy": 0.6224127486348152,
      "epoch": 1.9553941610456524,
      "grad_norm": 0.16225677728652954,
      "learning_rate": 3.482381083816358e-05,
      "loss": 0.6079,
      "mean_token_accuracy": 0.8106058351695538,
      "num_tokens": 678315248.0,
      "step": 21253
    },
    {
      "entropy": 0.599237774964422,
      "epoch": 1.9554861674170936,
      "grad_norm": 0.15027746558189392,
      "learning_rate": 3.482074401202196e-05,
      "loss": 0.5879,
      "mean_token_accuracy": 0.8193577267229557,
      "num_tokens": 678347052.0,
      "step": 21254
    },
    {
      "entropy": 0.6926392009481788,
      "epoch": 1.9555781737885347,
      "grad_norm": 0.1610957384109497,
      "learning_rate": 3.481767718588033e-05,
      "loss": 0.6856,
      "mean_token_accuracy": 0.7888110317289829,
      "num_tokens": 678378254.0,
      "step": 21255
    },
    {
      "entropy": 0.6342610856518149,
      "epoch": 1.955670180159976,
      "grad_norm": 0.1547403484582901,
      "learning_rate": 3.4814610359738705e-05,
      "loss": 0.6312,
      "mean_token_accuracy": 0.806495189666748,
      "num_tokens": 678410289.0,
      "step": 21256
    },
    {
      "entropy": 0.63203333504498,
      "epoch": 1.9557621865314174,
      "grad_norm": 0.15730856359004974,
      "learning_rate": 3.481154353359708e-05,
      "loss": 0.6168,
      "mean_token_accuracy": 0.8091539405286312,
      "num_tokens": 678441339.0,
      "step": 21257
    },
    {
      "entropy": 0.631457045674324,
      "epoch": 1.9558541929028586,
      "grad_norm": 0.14786316454410553,
      "learning_rate": 3.4808476707455455e-05,
      "loss": 0.6145,
      "mean_token_accuracy": 0.8114655166864395,
      "num_tokens": 678473572.0,
      "step": 21258
    },
    {
      "entropy": 0.520228173583746,
      "epoch": 1.9559461992742997,
      "grad_norm": 0.14399346709251404,
      "learning_rate": 3.480540988131383e-05,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8420242443680763,
      "num_tokens": 678506323.0,
      "step": 21259
    },
    {
      "entropy": 0.6194788403809071,
      "epoch": 1.9560382056457408,
      "grad_norm": 0.15868139266967773,
      "learning_rate": 3.4802343055172204e-05,
      "loss": 0.6208,
      "mean_token_accuracy": 0.8127582482993603,
      "num_tokens": 678538311.0,
      "step": 21260
    },
    {
      "entropy": 0.5707658205647022,
      "epoch": 1.9561302120171822,
      "grad_norm": 0.14733117818832397,
      "learning_rate": 3.479927622903057e-05,
      "loss": 0.5632,
      "mean_token_accuracy": 0.8247598335146904,
      "num_tokens": 678570640.0,
      "step": 21261
    },
    {
      "entropy": 0.7160279713571072,
      "epoch": 1.9562222183886235,
      "grad_norm": 0.1650717705488205,
      "learning_rate": 3.479620940288895e-05,
      "loss": 0.7066,
      "mean_token_accuracy": 0.7844678349792957,
      "num_tokens": 678602261.0,
      "step": 21262
    },
    {
      "entropy": 0.6107361875474453,
      "epoch": 1.9563142247600647,
      "grad_norm": 0.15341658890247345,
      "learning_rate": 3.479314257674733e-05,
      "loss": 0.592,
      "mean_token_accuracy": 0.8189074657857418,
      "num_tokens": 678634659.0,
      "step": 21263
    },
    {
      "entropy": 0.5976233799010515,
      "epoch": 1.9564062311315058,
      "grad_norm": 0.1562172919511795,
      "learning_rate": 3.47900757506057e-05,
      "loss": 0.5805,
      "mean_token_accuracy": 0.8204693458974361,
      "num_tokens": 678667244.0,
      "step": 21264
    },
    {
      "entropy": 0.5614777500741184,
      "epoch": 1.956498237502947,
      "grad_norm": 0.14617934823036194,
      "learning_rate": 3.478700892446408e-05,
      "loss": 0.5418,
      "mean_token_accuracy": 0.8307629264891148,
      "num_tokens": 678699265.0,
      "step": 21265
    },
    {
      "entropy": 0.6613208279013634,
      "epoch": 1.9565902438743883,
      "grad_norm": 0.1574053317308426,
      "learning_rate": 3.4783942098322445e-05,
      "loss": 0.6551,
      "mean_token_accuracy": 0.8010853752493858,
      "num_tokens": 678731811.0,
      "step": 21266
    },
    {
      "entropy": 0.6598533852957189,
      "epoch": 1.9566822502458296,
      "grad_norm": 0.15685585141181946,
      "learning_rate": 3.4780875272180827e-05,
      "loss": 0.6533,
      "mean_token_accuracy": 0.8000084683299065,
      "num_tokens": 678764088.0,
      "step": 21267
    },
    {
      "entropy": 0.5065051196143031,
      "epoch": 1.9567742566172708,
      "grad_norm": 0.1507766842842102,
      "learning_rate": 3.4777808446039194e-05,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8465737886726856,
      "num_tokens": 678796012.0,
      "step": 21268
    },
    {
      "entropy": 0.6727520525455475,
      "epoch": 1.956866262988712,
      "grad_norm": 0.15700970590114594,
      "learning_rate": 3.477474161989757e-05,
      "loss": 0.667,
      "mean_token_accuracy": 0.7968422770500183,
      "num_tokens": 678828104.0,
      "step": 21269
    },
    {
      "entropy": 0.5759826311841607,
      "epoch": 1.956958269360153,
      "grad_norm": 0.14591501653194427,
      "learning_rate": 3.4771674793755944e-05,
      "loss": 0.573,
      "mean_token_accuracy": 0.8264354318380356,
      "num_tokens": 678860240.0,
      "step": 21270
    },
    {
      "entropy": 0.5514581594616175,
      "epoch": 1.9570502757315944,
      "grad_norm": 0.15243393182754517,
      "learning_rate": 3.476860796761432e-05,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8282387107610703,
      "num_tokens": 678892593.0,
      "step": 21271
    },
    {
      "entropy": 0.5130588728934526,
      "epoch": 1.9571422821030358,
      "grad_norm": 0.1460019052028656,
      "learning_rate": 3.476554114147269e-05,
      "loss": 0.509,
      "mean_token_accuracy": 0.8418891616165638,
      "num_tokens": 678924515.0,
      "step": 21272
    },
    {
      "entropy": 0.6407651118934155,
      "epoch": 1.957234288474477,
      "grad_norm": 0.15766236186027527,
      "learning_rate": 3.476247431533107e-05,
      "loss": 0.631,
      "mean_token_accuracy": 0.8046779185533524,
      "num_tokens": 678956985.0,
      "step": 21273
    },
    {
      "entropy": 0.5458691781386733,
      "epoch": 1.957326294845918,
      "grad_norm": 0.1486053317785263,
      "learning_rate": 3.4759407489189436e-05,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8331080786883831,
      "num_tokens": 678988288.0,
      "step": 21274
    },
    {
      "entropy": 0.6098520131781697,
      "epoch": 1.9574183012173592,
      "grad_norm": 0.14971628785133362,
      "learning_rate": 3.475634066304782e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8160428181290627,
      "num_tokens": 679019974.0,
      "step": 21275
    },
    {
      "entropy": 0.5567984082736075,
      "epoch": 1.9575103075888005,
      "grad_norm": 0.14632569253444672,
      "learning_rate": 3.4753273836906185e-05,
      "loss": 0.5418,
      "mean_token_accuracy": 0.8343732096254826,
      "num_tokens": 679052325.0,
      "step": 21276
    },
    {
      "entropy": 0.6465688436292112,
      "epoch": 1.9576023139602419,
      "grad_norm": 0.161598339676857,
      "learning_rate": 3.475020701076456e-05,
      "loss": 0.6388,
      "mean_token_accuracy": 0.8045131713151932,
      "num_tokens": 679084401.0,
      "step": 21277
    },
    {
      "entropy": 0.5679879738017917,
      "epoch": 1.957694320331683,
      "grad_norm": 0.14986251294612885,
      "learning_rate": 3.4747140184622934e-05,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8298144713044167,
      "num_tokens": 679116402.0,
      "step": 21278
    },
    {
      "entropy": 0.6180616759229451,
      "epoch": 1.9577863267031241,
      "grad_norm": 0.14767557382583618,
      "learning_rate": 3.474407335848131e-05,
      "loss": 0.5961,
      "mean_token_accuracy": 0.8162252753973007,
      "num_tokens": 679148815.0,
      "step": 21279
    },
    {
      "entropy": 0.6311952299438417,
      "epoch": 1.9578783330745653,
      "grad_norm": 0.15780040621757507,
      "learning_rate": 3.4741006532339683e-05,
      "loss": 0.632,
      "mean_token_accuracy": 0.8092598989605904,
      "num_tokens": 679181129.0,
      "step": 21280
    },
    {
      "entropy": 0.5653028702363372,
      "epoch": 1.9579703394460066,
      "grad_norm": 0.14662279188632965,
      "learning_rate": 3.473793970619806e-05,
      "loss": 0.5643,
      "mean_token_accuracy": 0.823939610272646,
      "num_tokens": 679213436.0,
      "step": 21281
    },
    {
      "entropy": 0.6156429853290319,
      "epoch": 1.958062345817448,
      "grad_norm": 0.15307755768299103,
      "learning_rate": 3.473487288005643e-05,
      "loss": 0.6025,
      "mean_token_accuracy": 0.8165868036448956,
      "num_tokens": 679245218.0,
      "step": 21282
    },
    {
      "entropy": 0.5706377467140555,
      "epoch": 1.9581543521888891,
      "grad_norm": 0.1494438350200653,
      "learning_rate": 3.473180605391481e-05,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8259727470576763,
      "num_tokens": 679277228.0,
      "step": 21283
    },
    {
      "entropy": 0.6668741051107645,
      "epoch": 1.9582463585603302,
      "grad_norm": 0.16291803121566772,
      "learning_rate": 3.4728739227773175e-05,
      "loss": 0.6558,
      "mean_token_accuracy": 0.7985824942588806,
      "num_tokens": 679308476.0,
      "step": 21284
    },
    {
      "entropy": 0.6714517697691917,
      "epoch": 1.9583383649317714,
      "grad_norm": 0.15900953114032745,
      "learning_rate": 3.472567240163156e-05,
      "loss": 0.6539,
      "mean_token_accuracy": 0.8031100630760193,
      "num_tokens": 679339584.0,
      "step": 21285
    },
    {
      "entropy": 0.6384329907596111,
      "epoch": 1.9584303713032127,
      "grad_norm": 0.151958167552948,
      "learning_rate": 3.4722605575489925e-05,
      "loss": 0.6297,
      "mean_token_accuracy": 0.8059686683118343,
      "num_tokens": 679372336.0,
      "step": 21286
    },
    {
      "entropy": 0.5481289438903332,
      "epoch": 1.958522377674654,
      "grad_norm": 0.14974835515022278,
      "learning_rate": 3.47195387493483e-05,
      "loss": 0.5404,
      "mean_token_accuracy": 0.834539633244276,
      "num_tokens": 679404599.0,
      "step": 21287
    },
    {
      "entropy": 0.5909217689186335,
      "epoch": 1.9586143840460952,
      "grad_norm": 0.14868390560150146,
      "learning_rate": 3.4716471923206674e-05,
      "loss": 0.577,
      "mean_token_accuracy": 0.8206322863698006,
      "num_tokens": 679436759.0,
      "step": 21288
    },
    {
      "entropy": 0.609249516390264,
      "epoch": 1.9587063904175364,
      "grad_norm": 0.15404346585273743,
      "learning_rate": 3.471340509706505e-05,
      "loss": 0.6025,
      "mean_token_accuracy": 0.8157104402780533,
      "num_tokens": 679468781.0,
      "step": 21289
    },
    {
      "entropy": 0.6374495094642043,
      "epoch": 1.9587983967889775,
      "grad_norm": 0.16641782224178314,
      "learning_rate": 3.471033827092342e-05,
      "loss": 0.6371,
      "mean_token_accuracy": 0.8062924146652222,
      "num_tokens": 679500475.0,
      "step": 21290
    },
    {
      "entropy": 0.5360291213728487,
      "epoch": 1.9588904031604188,
      "grad_norm": 0.15239493548870087,
      "learning_rate": 3.47072714447818e-05,
      "loss": 0.5333,
      "mean_token_accuracy": 0.8373053222894669,
      "num_tokens": 679532425.0,
      "step": 21291
    },
    {
      "entropy": 0.6491107679903507,
      "epoch": 1.9589824095318602,
      "grad_norm": 0.15571239590644836,
      "learning_rate": 3.470420461864017e-05,
      "loss": 0.6373,
      "mean_token_accuracy": 0.8059737496078014,
      "num_tokens": 679564713.0,
      "step": 21292
    },
    {
      "entropy": 0.6373206097632647,
      "epoch": 1.9590744159033013,
      "grad_norm": 0.15556320548057556,
      "learning_rate": 3.470113779249855e-05,
      "loss": 0.6357,
      "mean_token_accuracy": 0.8084164969623089,
      "num_tokens": 679597103.0,
      "step": 21293
    },
    {
      "entropy": 0.5706448014825583,
      "epoch": 1.9591664222747425,
      "grad_norm": 0.1499788612127304,
      "learning_rate": 3.469807096635692e-05,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8260926902294159,
      "num_tokens": 679629223.0,
      "step": 21294
    },
    {
      "entropy": 0.48942989110946655,
      "epoch": 1.9592584286461836,
      "grad_norm": 0.13804039359092712,
      "learning_rate": 3.469500414021529e-05,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8513868227601051,
      "num_tokens": 679661720.0,
      "step": 21295
    },
    {
      "entropy": 0.5125891137868166,
      "epoch": 1.959350435017625,
      "grad_norm": 0.14112327992916107,
      "learning_rate": 3.469193731407367e-05,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8415563590824604,
      "num_tokens": 679693758.0,
      "step": 21296
    },
    {
      "entropy": 0.6224912516772747,
      "epoch": 1.9594424413890663,
      "grad_norm": 0.15435956418514252,
      "learning_rate": 3.468887048793204e-05,
      "loss": 0.6032,
      "mean_token_accuracy": 0.8141206800937653,
      "num_tokens": 679725330.0,
      "step": 21297
    },
    {
      "entropy": 0.48347552958875895,
      "epoch": 1.9595344477605074,
      "grad_norm": 0.14763052761554718,
      "learning_rate": 3.468580366179042e-05,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8520222194492817,
      "num_tokens": 679757724.0,
      "step": 21298
    },
    {
      "entropy": 0.7244624998420477,
      "epoch": 1.9596264541319486,
      "grad_norm": 0.16685761511325836,
      "learning_rate": 3.468273683564879e-05,
      "loss": 0.7152,
      "mean_token_accuracy": 0.7818176485598087,
      "num_tokens": 679788773.0,
      "step": 21299
    },
    {
      "entropy": 0.5612758104689419,
      "epoch": 1.9597184605033897,
      "grad_norm": 0.1495288759469986,
      "learning_rate": 3.467967000950716e-05,
      "loss": 0.554,
      "mean_token_accuracy": 0.8305947035551071,
      "num_tokens": 679820716.0,
      "step": 21300
    },
    {
      "entropy": 0.5787014588713646,
      "epoch": 1.959810466874831,
      "grad_norm": 0.14948225021362305,
      "learning_rate": 3.467660318336554e-05,
      "loss": 0.5745,
      "mean_token_accuracy": 0.8244138918817043,
      "num_tokens": 679852481.0,
      "step": 21301
    },
    {
      "entropy": 0.5785278105176985,
      "epoch": 1.9599024732462724,
      "grad_norm": 0.1547030359506607,
      "learning_rate": 3.467353635722391e-05,
      "loss": 0.5683,
      "mean_token_accuracy": 0.8264549262821674,
      "num_tokens": 679885246.0,
      "step": 21302
    },
    {
      "entropy": 0.5095519004389644,
      "epoch": 1.9599944796177136,
      "grad_norm": 0.14539030194282532,
      "learning_rate": 3.467046953108229e-05,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8439927734434605,
      "num_tokens": 679916611.0,
      "step": 21303
    },
    {
      "entropy": 0.6593188587576151,
      "epoch": 1.9600864859891547,
      "grad_norm": 0.16093704104423523,
      "learning_rate": 3.466740270494066e-05,
      "loss": 0.66,
      "mean_token_accuracy": 0.8015533797442913,
      "num_tokens": 679947456.0,
      "step": 21304
    },
    {
      "entropy": 0.6076072193682194,
      "epoch": 1.9601784923605958,
      "grad_norm": 0.1548081934452057,
      "learning_rate": 3.466433587879903e-05,
      "loss": 0.595,
      "mean_token_accuracy": 0.8164942227303982,
      "num_tokens": 679980082.0,
      "step": 21305
    },
    {
      "entropy": 0.676135765388608,
      "epoch": 1.9602704987320372,
      "grad_norm": 0.1597762405872345,
      "learning_rate": 3.466126905265741e-05,
      "loss": 0.6691,
      "mean_token_accuracy": 0.7991343252360821,
      "num_tokens": 680012657.0,
      "step": 21306
    },
    {
      "entropy": 0.6089157704263926,
      "epoch": 1.9603625051034785,
      "grad_norm": 0.15165120363235474,
      "learning_rate": 3.465820222651578e-05,
      "loss": 0.6053,
      "mean_token_accuracy": 0.8157533593475819,
      "num_tokens": 680044599.0,
      "step": 21307
    },
    {
      "entropy": 0.6537646632641554,
      "epoch": 1.9604545114749197,
      "grad_norm": 0.1486111432313919,
      "learning_rate": 3.465513540037415e-05,
      "loss": 0.6313,
      "mean_token_accuracy": 0.8040067479014397,
      "num_tokens": 680076749.0,
      "step": 21308
    },
    {
      "entropy": 0.6516920812427998,
      "epoch": 1.9605465178463608,
      "grad_norm": 0.15980876982212067,
      "learning_rate": 3.465206857423253e-05,
      "loss": 0.6391,
      "mean_token_accuracy": 0.8042341805994511,
      "num_tokens": 680107483.0,
      "step": 21309
    },
    {
      "entropy": 0.6394985038787127,
      "epoch": 1.960638524217802,
      "grad_norm": 0.15821999311447144,
      "learning_rate": 3.46490017480909e-05,
      "loss": 0.6277,
      "mean_token_accuracy": 0.8074805438518524,
      "num_tokens": 680140042.0,
      "step": 21310
    },
    {
      "entropy": 0.6821497092023492,
      "epoch": 1.9607305305892433,
      "grad_norm": 0.16409088671207428,
      "learning_rate": 3.464593492194928e-05,
      "loss": 0.6698,
      "mean_token_accuracy": 0.7963251620531082,
      "num_tokens": 680170654.0,
      "step": 21311
    },
    {
      "entropy": 0.6324073197320104,
      "epoch": 1.9608225369606846,
      "grad_norm": 0.15644466876983643,
      "learning_rate": 3.464286809580765e-05,
      "loss": 0.6153,
      "mean_token_accuracy": 0.8095633536577225,
      "num_tokens": 680201790.0,
      "step": 21312
    },
    {
      "entropy": 0.6134825311601162,
      "epoch": 1.9609145433321258,
      "grad_norm": 0.15608546137809753,
      "learning_rate": 3.463980126966602e-05,
      "loss": 0.6015,
      "mean_token_accuracy": 0.8143600486218929,
      "num_tokens": 680232847.0,
      "step": 21313
    },
    {
      "entropy": 0.5719862459227443,
      "epoch": 1.961006549703567,
      "grad_norm": 0.1449672281742096,
      "learning_rate": 3.46367344435244e-05,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8295288160443306,
      "num_tokens": 680264533.0,
      "step": 21314
    },
    {
      "entropy": 0.543301360681653,
      "epoch": 1.961098556075008,
      "grad_norm": 0.15336935222148895,
      "learning_rate": 3.463366761738277e-05,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8345499411225319,
      "num_tokens": 680295847.0,
      "step": 21315
    },
    {
      "entropy": 0.7133592907339334,
      "epoch": 1.9611905624464494,
      "grad_norm": 0.16167475283145905,
      "learning_rate": 3.4630600791241144e-05,
      "loss": 0.7101,
      "mean_token_accuracy": 0.7846650294959545,
      "num_tokens": 680327538.0,
      "step": 21316
    },
    {
      "entropy": 0.680513403378427,
      "epoch": 1.9612825688178908,
      "grad_norm": 0.1578439623117447,
      "learning_rate": 3.462753396509952e-05,
      "loss": 0.6715,
      "mean_token_accuracy": 0.7940363027155399,
      "num_tokens": 680359659.0,
      "step": 21317
    },
    {
      "entropy": 0.649445541203022,
      "epoch": 1.961374575189332,
      "grad_norm": 0.15572932362556458,
      "learning_rate": 3.462446713895789e-05,
      "loss": 0.6457,
      "mean_token_accuracy": 0.7993712648749352,
      "num_tokens": 680391139.0,
      "step": 21318
    },
    {
      "entropy": 0.5090573215857148,
      "epoch": 1.961466581560773,
      "grad_norm": 0.14616280794143677,
      "learning_rate": 3.462140031281627e-05,
      "loss": 0.5113,
      "mean_token_accuracy": 0.8411387205123901,
      "num_tokens": 680422824.0,
      "step": 21319
    },
    {
      "entropy": 0.5139555558562279,
      "epoch": 1.9615585879322142,
      "grad_norm": 0.1421659141778946,
      "learning_rate": 3.461833348667464e-05,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8441520854830742,
      "num_tokens": 680455243.0,
      "step": 21320
    },
    {
      "entropy": 0.605599083006382,
      "epoch": 1.9616505943036555,
      "grad_norm": 0.1522017866373062,
      "learning_rate": 3.461526666053301e-05,
      "loss": 0.596,
      "mean_token_accuracy": 0.8144602812826633,
      "num_tokens": 680486850.0,
      "step": 21321
    },
    {
      "entropy": 0.6040174718946218,
      "epoch": 1.9617426006750969,
      "grad_norm": 0.1535668969154358,
      "learning_rate": 3.461219983439139e-05,
      "loss": 0.5944,
      "mean_token_accuracy": 0.8189722411334515,
      "num_tokens": 680518958.0,
      "step": 21322
    },
    {
      "entropy": 0.5993740763515234,
      "epoch": 1.961834607046538,
      "grad_norm": 0.1532396376132965,
      "learning_rate": 3.4609133008249766e-05,
      "loss": 0.6041,
      "mean_token_accuracy": 0.8109113425016403,
      "num_tokens": 680550822.0,
      "step": 21323
    },
    {
      "entropy": 0.6341364393010736,
      "epoch": 1.9619266134179791,
      "grad_norm": 0.15152688324451447,
      "learning_rate": 3.460606618210814e-05,
      "loss": 0.6181,
      "mean_token_accuracy": 0.8130984269082546,
      "num_tokens": 680583196.0,
      "step": 21324
    },
    {
      "entropy": 0.6444807928055525,
      "epoch": 1.9620186197894203,
      "grad_norm": 0.15271049737930298,
      "learning_rate": 3.4602999355966515e-05,
      "loss": 0.6365,
      "mean_token_accuracy": 0.8064321465790272,
      "num_tokens": 680615562.0,
      "step": 21325
    },
    {
      "entropy": 0.5193238351494074,
      "epoch": 1.9621106261608616,
      "grad_norm": 0.1390783041715622,
      "learning_rate": 3.459993252982488e-05,
      "loss": 0.5077,
      "mean_token_accuracy": 0.8433996886014938,
      "num_tokens": 680647437.0,
      "step": 21326
    },
    {
      "entropy": 0.63511932361871,
      "epoch": 1.962202632532303,
      "grad_norm": 0.15739434957504272,
      "learning_rate": 3.4596865703683265e-05,
      "loss": 0.6321,
      "mean_token_accuracy": 0.8105347640812397,
      "num_tokens": 680678649.0,
      "step": 21327
    },
    {
      "entropy": 0.5955421254038811,
      "epoch": 1.9622946389037441,
      "grad_norm": 0.14818967878818512,
      "learning_rate": 3.459379887754163e-05,
      "loss": 0.5819,
      "mean_token_accuracy": 0.8219969980418682,
      "num_tokens": 680710482.0,
      "step": 21328
    },
    {
      "entropy": 0.5817691185511649,
      "epoch": 1.9623866452751852,
      "grad_norm": 0.1549515575170517,
      "learning_rate": 3.459073205140001e-05,
      "loss": 0.5801,
      "mean_token_accuracy": 0.8231864236295223,
      "num_tokens": 680742727.0,
      "step": 21329
    },
    {
      "entropy": 0.5611398126929998,
      "epoch": 1.9624786516466264,
      "grad_norm": 0.1478196233510971,
      "learning_rate": 3.458766522525838e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8293365016579628,
      "num_tokens": 680774740.0,
      "step": 21330
    },
    {
      "entropy": 0.560447359457612,
      "epoch": 1.9625706580180677,
      "grad_norm": 0.14866481721401215,
      "learning_rate": 3.4584598399116757e-05,
      "loss": 0.5525,
      "mean_token_accuracy": 0.8268001228570938,
      "num_tokens": 680806265.0,
      "step": 21331
    },
    {
      "entropy": 0.6059398362413049,
      "epoch": 1.962662664389509,
      "grad_norm": 0.15114794671535492,
      "learning_rate": 3.458153157297513e-05,
      "loss": 0.5931,
      "mean_token_accuracy": 0.8204022608697414,
      "num_tokens": 680838917.0,
      "step": 21332
    },
    {
      "entropy": 0.5281822150573134,
      "epoch": 1.9627546707609502,
      "grad_norm": 0.14811071753501892,
      "learning_rate": 3.4578464746833506e-05,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8436845727264881,
      "num_tokens": 680871122.0,
      "step": 21333
    },
    {
      "entropy": 0.6108919596299529,
      "epoch": 1.9628466771323914,
      "grad_norm": 0.15862321853637695,
      "learning_rate": 3.4575397920691874e-05,
      "loss": 0.5929,
      "mean_token_accuracy": 0.8177039884030819,
      "num_tokens": 680902506.0,
      "step": 21334
    },
    {
      "entropy": 0.5826131412759423,
      "epoch": 1.9629386835038325,
      "grad_norm": 0.15619121491909027,
      "learning_rate": 3.4572331094550255e-05,
      "loss": 0.5733,
      "mean_token_accuracy": 0.8213238567113876,
      "num_tokens": 680933615.0,
      "step": 21335
    },
    {
      "entropy": 0.5711074601858854,
      "epoch": 1.9630306898752738,
      "grad_norm": 0.15092122554779053,
      "learning_rate": 3.456926426840862e-05,
      "loss": 0.5646,
      "mean_token_accuracy": 0.8257797807455063,
      "num_tokens": 680965499.0,
      "step": 21336
    },
    {
      "entropy": 0.6008878704160452,
      "epoch": 1.9631226962467152,
      "grad_norm": 0.1546105444431305,
      "learning_rate": 3.4566197442267e-05,
      "loss": 0.581,
      "mean_token_accuracy": 0.8200973235070705,
      "num_tokens": 680997478.0,
      "step": 21337
    },
    {
      "entropy": 0.6683178171515465,
      "epoch": 1.9632147026181563,
      "grad_norm": 0.1526942104101181,
      "learning_rate": 3.456313061612537e-05,
      "loss": 0.6606,
      "mean_token_accuracy": 0.7975511141121387,
      "num_tokens": 681029722.0,
      "step": 21338
    },
    {
      "entropy": 0.615163080394268,
      "epoch": 1.9633067089895975,
      "grad_norm": 0.1529007852077484,
      "learning_rate": 3.456006378998375e-05,
      "loss": 0.6027,
      "mean_token_accuracy": 0.8141313940286636,
      "num_tokens": 681061736.0,
      "step": 21339
    },
    {
      "entropy": 0.6213803123682737,
      "epoch": 1.9633987153610386,
      "grad_norm": 0.1533537358045578,
      "learning_rate": 3.455699696384212e-05,
      "loss": 0.6241,
      "mean_token_accuracy": 0.8107949793338776,
      "num_tokens": 681093620.0,
      "step": 21340
    },
    {
      "entropy": 0.6208348143845797,
      "epoch": 1.96349072173248,
      "grad_norm": 0.15056495368480682,
      "learning_rate": 3.4553930137700496e-05,
      "loss": 0.6099,
      "mean_token_accuracy": 0.8158344700932503,
      "num_tokens": 681126183.0,
      "step": 21341
    },
    {
      "entropy": 0.5893088094890118,
      "epoch": 1.9635827281039213,
      "grad_norm": 0.14804711937904358,
      "learning_rate": 3.455086331155887e-05,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8224853090941906,
      "num_tokens": 681158090.0,
      "step": 21342
    },
    {
      "entropy": 0.5170803861692548,
      "epoch": 1.9636747344753624,
      "grad_norm": 0.1451568752527237,
      "learning_rate": 3.4547796485417246e-05,
      "loss": 0.52,
      "mean_token_accuracy": 0.839951429516077,
      "num_tokens": 681190379.0,
      "step": 21343
    },
    {
      "entropy": 0.5670495652593672,
      "epoch": 1.9637667408468036,
      "grad_norm": 0.1448916792869568,
      "learning_rate": 3.4544729659275613e-05,
      "loss": 0.5605,
      "mean_token_accuracy": 0.8253439217805862,
      "num_tokens": 681222885.0,
      "step": 21344
    },
    {
      "entropy": 0.5857197940349579,
      "epoch": 1.9638587472182447,
      "grad_norm": 0.15507422387599945,
      "learning_rate": 3.4541662833133995e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8239479511976242,
      "num_tokens": 681254945.0,
      "step": 21345
    },
    {
      "entropy": 0.46034914068877697,
      "epoch": 1.963950753589686,
      "grad_norm": 0.1397617757320404,
      "learning_rate": 3.453859600699236e-05,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8638803549110889,
      "num_tokens": 681286447.0,
      "step": 21346
    },
    {
      "entropy": 0.5911430362612009,
      "epoch": 1.9640427599611274,
      "grad_norm": 0.15803444385528564,
      "learning_rate": 3.453552918085074e-05,
      "loss": 0.5674,
      "mean_token_accuracy": 0.8239190205931664,
      "num_tokens": 681318840.0,
      "step": 21347
    },
    {
      "entropy": 0.6496984362602234,
      "epoch": 1.9641347663325686,
      "grad_norm": 0.15744134783744812,
      "learning_rate": 3.453246235470911e-05,
      "loss": 0.6258,
      "mean_token_accuracy": 0.8067683428525925,
      "num_tokens": 681350341.0,
      "step": 21348
    },
    {
      "entropy": 0.5910794101655483,
      "epoch": 1.9642267727040097,
      "grad_norm": 0.14841869473457336,
      "learning_rate": 3.452939552856749e-05,
      "loss": 0.5763,
      "mean_token_accuracy": 0.8208978027105331,
      "num_tokens": 681382488.0,
      "step": 21349
    },
    {
      "entropy": 0.523166986182332,
      "epoch": 1.9643187790754508,
      "grad_norm": 0.14202262461185455,
      "learning_rate": 3.452632870242586e-05,
      "loss": 0.5131,
      "mean_token_accuracy": 0.8414236158132553,
      "num_tokens": 681414708.0,
      "step": 21350
    },
    {
      "entropy": 0.6269142404198647,
      "epoch": 1.9644107854468922,
      "grad_norm": 0.15734006464481354,
      "learning_rate": 3.4523261876284236e-05,
      "loss": 0.6231,
      "mean_token_accuracy": 0.8095087967813015,
      "num_tokens": 681445694.0,
      "step": 21351
    },
    {
      "entropy": 0.5600879918783903,
      "epoch": 1.9645027918183335,
      "grad_norm": 0.15250332653522491,
      "learning_rate": 3.4520195050142604e-05,
      "loss": 0.5473,
      "mean_token_accuracy": 0.8284119479358196,
      "num_tokens": 681477749.0,
      "step": 21352
    },
    {
      "entropy": 0.6088374871760607,
      "epoch": 1.9645947981897747,
      "grad_norm": 0.1522459238767624,
      "learning_rate": 3.4517128224000985e-05,
      "loss": 0.6041,
      "mean_token_accuracy": 0.8149850331246853,
      "num_tokens": 681508686.0,
      "step": 21353
    },
    {
      "entropy": 0.5721741877496243,
      "epoch": 1.9646868045612158,
      "grad_norm": 0.15807528793811798,
      "learning_rate": 3.451406139785936e-05,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8224803991615772,
      "num_tokens": 681539972.0,
      "step": 21354
    },
    {
      "entropy": 0.6042760517448187,
      "epoch": 1.964778810932657,
      "grad_norm": 0.15928784012794495,
      "learning_rate": 3.451099457171773e-05,
      "loss": 0.5983,
      "mean_token_accuracy": 0.819581676274538,
      "num_tokens": 681571972.0,
      "step": 21355
    },
    {
      "entropy": 0.6632423615083098,
      "epoch": 1.9648708173040983,
      "grad_norm": 0.1606673002243042,
      "learning_rate": 3.450792774557611e-05,
      "loss": 0.6555,
      "mean_token_accuracy": 0.7998734898865223,
      "num_tokens": 681603847.0,
      "step": 21356
    },
    {
      "entropy": 0.4825551363173872,
      "epoch": 1.9649628236755396,
      "grad_norm": 0.14063936471939087,
      "learning_rate": 3.450486091943448e-05,
      "loss": 0.4718,
      "mean_token_accuracy": 0.8523786328732967,
      "num_tokens": 681636348.0,
      "step": 21357
    },
    {
      "entropy": 0.4706544578075409,
      "epoch": 1.9650548300469808,
      "grad_norm": 0.13893046975135803,
      "learning_rate": 3.450179409329286e-05,
      "loss": 0.4637,
      "mean_token_accuracy": 0.8565869964659214,
      "num_tokens": 681669069.0,
      "step": 21358
    },
    {
      "entropy": 0.5611624009907246,
      "epoch": 1.965146836418422,
      "grad_norm": 0.1531236618757248,
      "learning_rate": 3.4498727267151226e-05,
      "loss": 0.5588,
      "mean_token_accuracy": 0.8256942220032215,
      "num_tokens": 681700363.0,
      "step": 21359
    },
    {
      "entropy": 0.6286447513848543,
      "epoch": 1.965238842789863,
      "grad_norm": 0.15789838135242462,
      "learning_rate": 3.44956604410096e-05,
      "loss": 0.6222,
      "mean_token_accuracy": 0.8140952214598656,
      "num_tokens": 681732720.0,
      "step": 21360
    },
    {
      "entropy": 0.5804448078852147,
      "epoch": 1.9653308491613044,
      "grad_norm": 0.1491953432559967,
      "learning_rate": 3.4492593614867976e-05,
      "loss": 0.5686,
      "mean_token_accuracy": 0.827552855014801,
      "num_tokens": 681764456.0,
      "step": 21361
    },
    {
      "entropy": 0.5420804088935256,
      "epoch": 1.9654228555327458,
      "grad_norm": 0.14461985230445862,
      "learning_rate": 3.448952678872635e-05,
      "loss": 0.5292,
      "mean_token_accuracy": 0.836988303810358,
      "num_tokens": 681797126.0,
      "step": 21362
    },
    {
      "entropy": 0.6146933985874057,
      "epoch": 1.965514861904187,
      "grad_norm": 0.1561742126941681,
      "learning_rate": 3.4486459962584725e-05,
      "loss": 0.6091,
      "mean_token_accuracy": 0.8100478984415531,
      "num_tokens": 681828524.0,
      "step": 21363
    },
    {
      "entropy": 0.6818519607186317,
      "epoch": 1.965606868275628,
      "grad_norm": 0.16190151870250702,
      "learning_rate": 3.44833931364431e-05,
      "loss": 0.6786,
      "mean_token_accuracy": 0.7963639572262764,
      "num_tokens": 681860331.0,
      "step": 21364
    },
    {
      "entropy": 0.5675434209406376,
      "epoch": 1.9656988746470692,
      "grad_norm": 0.15287736058235168,
      "learning_rate": 3.448032631030147e-05,
      "loss": 0.5481,
      "mean_token_accuracy": 0.8299851976335049,
      "num_tokens": 681891967.0,
      "step": 21365
    },
    {
      "entropy": 0.5778332771733403,
      "epoch": 1.9657908810185105,
      "grad_norm": 0.15438541769981384,
      "learning_rate": 3.447725948415985e-05,
      "loss": 0.5653,
      "mean_token_accuracy": 0.822092954069376,
      "num_tokens": 681923635.0,
      "step": 21366
    },
    {
      "entropy": 0.5539723876863718,
      "epoch": 1.9658828873899519,
      "grad_norm": 0.1448909044265747,
      "learning_rate": 3.447419265801822e-05,
      "loss": 0.5386,
      "mean_token_accuracy": 0.8299163170158863,
      "num_tokens": 681955586.0,
      "step": 21367
    },
    {
      "entropy": 0.7175794765353203,
      "epoch": 1.965974893761393,
      "grad_norm": 0.16305571794509888,
      "learning_rate": 3.447112583187659e-05,
      "loss": 0.7128,
      "mean_token_accuracy": 0.7821268066763878,
      "num_tokens": 681987667.0,
      "step": 21368
    },
    {
      "entropy": 0.6467436701059341,
      "epoch": 1.9660669001328341,
      "grad_norm": 0.15374204516410828,
      "learning_rate": 3.4468059005734966e-05,
      "loss": 0.6327,
      "mean_token_accuracy": 0.8082457222044468,
      "num_tokens": 682019501.0,
      "step": 21369
    },
    {
      "entropy": 0.6034330800175667,
      "epoch": 1.9661589065042753,
      "grad_norm": 0.15455403923988342,
      "learning_rate": 3.446499217959334e-05,
      "loss": 0.5755,
      "mean_token_accuracy": 0.820607103407383,
      "num_tokens": 682050801.0,
      "step": 21370
    },
    {
      "entropy": 0.6339986026287079,
      "epoch": 1.9662509128757166,
      "grad_norm": 0.1568659245967865,
      "learning_rate": 3.4461925353451715e-05,
      "loss": 0.6314,
      "mean_token_accuracy": 0.8036801442503929,
      "num_tokens": 682082589.0,
      "step": 21371
    },
    {
      "entropy": 0.6121579203754663,
      "epoch": 1.966342919247158,
      "grad_norm": 0.15067189931869507,
      "learning_rate": 3.445885852731009e-05,
      "loss": 0.5959,
      "mean_token_accuracy": 0.8127848990261555,
      "num_tokens": 682114644.0,
      "step": 21372
    },
    {
      "entropy": 0.5865286802873015,
      "epoch": 1.9664349256185991,
      "grad_norm": 0.1485477238893509,
      "learning_rate": 3.445579170116846e-05,
      "loss": 0.5722,
      "mean_token_accuracy": 0.8231410048902035,
      "num_tokens": 682146732.0,
      "step": 21373
    },
    {
      "entropy": 0.5870288405567408,
      "epoch": 1.9665269319900403,
      "grad_norm": 0.14946693181991577,
      "learning_rate": 3.445272487502684e-05,
      "loss": 0.5715,
      "mean_token_accuracy": 0.8225001096725464,
      "num_tokens": 682177916.0,
      "step": 21374
    },
    {
      "entropy": 0.5876409010961652,
      "epoch": 1.9666189383614814,
      "grad_norm": 0.15029080212116241,
      "learning_rate": 3.444965804888521e-05,
      "loss": 0.5808,
      "mean_token_accuracy": 0.8230478540062904,
      "num_tokens": 682210036.0,
      "step": 21375
    },
    {
      "entropy": 0.5097418269142509,
      "epoch": 1.9667109447329227,
      "grad_norm": 0.14845702052116394,
      "learning_rate": 3.444659122274358e-05,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8434588946402073,
      "num_tokens": 682242375.0,
      "step": 21376
    },
    {
      "entropy": 0.47766614775173366,
      "epoch": 1.966802951104364,
      "grad_norm": 0.1422491818666458,
      "learning_rate": 3.4443524396601956e-05,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8551233857870102,
      "num_tokens": 682274402.0,
      "step": 21377
    },
    {
      "entropy": 0.5949374139308929,
      "epoch": 1.9668949574758052,
      "grad_norm": 0.15084853768348694,
      "learning_rate": 3.444045757046033e-05,
      "loss": 0.5788,
      "mean_token_accuracy": 0.8247289657592773,
      "num_tokens": 682306046.0,
      "step": 21378
    },
    {
      "entropy": 0.6562820840626955,
      "epoch": 1.9669869638472464,
      "grad_norm": 0.15773680806159973,
      "learning_rate": 3.4437390744318706e-05,
      "loss": 0.6552,
      "mean_token_accuracy": 0.8019951395690441,
      "num_tokens": 682337551.0,
      "step": 21379
    },
    {
      "entropy": 0.5650584008544683,
      "epoch": 1.9670789702186875,
      "grad_norm": 0.14844854176044464,
      "learning_rate": 3.443432391817708e-05,
      "loss": 0.5492,
      "mean_token_accuracy": 0.829235851764679,
      "num_tokens": 682369571.0,
      "step": 21380
    },
    {
      "entropy": 0.6184273194521666,
      "epoch": 1.9671709765901289,
      "grad_norm": 0.15808387100696564,
      "learning_rate": 3.443125709203545e-05,
      "loss": 0.6119,
      "mean_token_accuracy": 0.8136148788034916,
      "num_tokens": 682400096.0,
      "step": 21381
    },
    {
      "entropy": 0.5580392898991704,
      "epoch": 1.9672629829615702,
      "grad_norm": 0.14857202768325806,
      "learning_rate": 3.442819026589383e-05,
      "loss": 0.5388,
      "mean_token_accuracy": 0.8335470110177994,
      "num_tokens": 682431340.0,
      "step": 21382
    },
    {
      "entropy": 0.6230897661298513,
      "epoch": 1.9673549893330113,
      "grad_norm": 0.15631352365016937,
      "learning_rate": 3.44251234397522e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.8090680204331875,
      "num_tokens": 682463009.0,
      "step": 21383
    },
    {
      "entropy": 0.6006581578403711,
      "epoch": 1.9674469957044525,
      "grad_norm": 0.15539997816085815,
      "learning_rate": 3.442205661361058e-05,
      "loss": 0.5981,
      "mean_token_accuracy": 0.8176520466804504,
      "num_tokens": 682495191.0,
      "step": 21384
    },
    {
      "entropy": 0.5669614616781473,
      "epoch": 1.9675390020758936,
      "grad_norm": 0.15764622390270233,
      "learning_rate": 3.4418989787468954e-05,
      "loss": 0.5629,
      "mean_token_accuracy": 0.825372438877821,
      "num_tokens": 682527108.0,
      "step": 21385
    },
    {
      "entropy": 0.49716425058431923,
      "epoch": 1.967631008447335,
      "grad_norm": 0.1374935358762741,
      "learning_rate": 3.441592296132732e-05,
      "loss": 0.486,
      "mean_token_accuracy": 0.8478306792676449,
      "num_tokens": 682559575.0,
      "step": 21386
    },
    {
      "entropy": 0.6569050531834364,
      "epoch": 1.9677230148187763,
      "grad_norm": 0.15916824340820312,
      "learning_rate": 3.44128561351857e-05,
      "loss": 0.6503,
      "mean_token_accuracy": 0.8017963692545891,
      "num_tokens": 682592053.0,
      "step": 21387
    },
    {
      "entropy": 0.5195838613435626,
      "epoch": 1.9678150211902175,
      "grad_norm": 0.1477191299200058,
      "learning_rate": 3.440978930904407e-05,
      "loss": 0.5125,
      "mean_token_accuracy": 0.8407791368663311,
      "num_tokens": 682624543.0,
      "step": 21388
    },
    {
      "entropy": 0.6630308008752763,
      "epoch": 1.9679070275616586,
      "grad_norm": 0.1501186192035675,
      "learning_rate": 3.4406722482902445e-05,
      "loss": 0.6425,
      "mean_token_accuracy": 0.8080282174050808,
      "num_tokens": 682657001.0,
      "step": 21389
    },
    {
      "entropy": 0.6182821244001389,
      "epoch": 1.9679990339330997,
      "grad_norm": 0.15415498614311218,
      "learning_rate": 3.440365565676082e-05,
      "loss": 0.6201,
      "mean_token_accuracy": 0.8132022470235825,
      "num_tokens": 682689769.0,
      "step": 21390
    },
    {
      "entropy": 0.696206022053957,
      "epoch": 1.968091040304541,
      "grad_norm": 0.16417478024959564,
      "learning_rate": 3.4400588830619195e-05,
      "loss": 0.6778,
      "mean_token_accuracy": 0.7963638752698898,
      "num_tokens": 682721314.0,
      "step": 21391
    },
    {
      "entropy": 0.6139987716451287,
      "epoch": 1.9681830466759824,
      "grad_norm": 0.1567133665084839,
      "learning_rate": 3.439752200447757e-05,
      "loss": 0.6046,
      "mean_token_accuracy": 0.8163300231099129,
      "num_tokens": 682753575.0,
      "step": 21392
    },
    {
      "entropy": 0.7849078522995114,
      "epoch": 1.9682750530474236,
      "grad_norm": 0.16484278440475464,
      "learning_rate": 3.4394455178335944e-05,
      "loss": 0.7651,
      "mean_token_accuracy": 0.7691316232085228,
      "num_tokens": 682785797.0,
      "step": 21393
    },
    {
      "entropy": 0.6398603161796927,
      "epoch": 1.9683670594188647,
      "grad_norm": 0.15703143179416656,
      "learning_rate": 3.439138835219431e-05,
      "loss": 0.6293,
      "mean_token_accuracy": 0.8044772520661354,
      "num_tokens": 682817791.0,
      "step": 21394
    },
    {
      "entropy": 0.5292166583240032,
      "epoch": 1.9684590657903058,
      "grad_norm": 0.14518386125564575,
      "learning_rate": 3.438832152605269e-05,
      "loss": 0.524,
      "mean_token_accuracy": 0.8403446897864342,
      "num_tokens": 682849519.0,
      "step": 21395
    },
    {
      "entropy": 0.6247552074491978,
      "epoch": 1.9685510721617472,
      "grad_norm": 0.15092115104198456,
      "learning_rate": 3.438525469991106e-05,
      "loss": 0.6103,
      "mean_token_accuracy": 0.8150033578276634,
      "num_tokens": 682881507.0,
      "step": 21396
    },
    {
      "entropy": 0.7264052554965019,
      "epoch": 1.9686430785331885,
      "grad_norm": 0.1691385954618454,
      "learning_rate": 3.4382187873769436e-05,
      "loss": 0.7207,
      "mean_token_accuracy": 0.780366413295269,
      "num_tokens": 682912899.0,
      "step": 21397
    },
    {
      "entropy": 0.5089506083168089,
      "epoch": 1.9687350849046297,
      "grad_norm": 0.14153747260570526,
      "learning_rate": 3.437912104762781e-05,
      "loss": 0.4988,
      "mean_token_accuracy": 0.8455067649483681,
      "num_tokens": 682945372.0,
      "step": 21398
    },
    {
      "entropy": 0.5406015440821648,
      "epoch": 1.9688270912760708,
      "grad_norm": 0.14724381268024445,
      "learning_rate": 3.4376054221486185e-05,
      "loss": 0.5345,
      "mean_token_accuracy": 0.8342194631695747,
      "num_tokens": 682977649.0,
      "step": 21399
    },
    {
      "entropy": 0.5391217388678342,
      "epoch": 1.968919097647512,
      "grad_norm": 0.14239661395549774,
      "learning_rate": 3.437298739534456e-05,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8336374647915363,
      "num_tokens": 683009243.0,
      "step": 21400
    },
    {
      "entropy": 0.5763911362737417,
      "epoch": 1.9690111040189533,
      "grad_norm": 0.15171010792255402,
      "learning_rate": 3.4369920569202934e-05,
      "loss": 0.5702,
      "mean_token_accuracy": 0.8228059485554695,
      "num_tokens": 683041040.0,
      "step": 21401
    },
    {
      "entropy": 0.5524709858000278,
      "epoch": 1.9691031103903947,
      "grad_norm": 0.14801393449306488,
      "learning_rate": 3.436685374306131e-05,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8335354961454868,
      "num_tokens": 683073027.0,
      "step": 21402
    },
    {
      "entropy": 0.5018703788518906,
      "epoch": 1.9691951167618358,
      "grad_norm": 0.14049360156059265,
      "learning_rate": 3.4363786916919684e-05,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8480123989284039,
      "num_tokens": 683105666.0,
      "step": 21403
    },
    {
      "entropy": 0.628142761066556,
      "epoch": 1.969287123133277,
      "grad_norm": 0.15579034388065338,
      "learning_rate": 3.436072009077805e-05,
      "loss": 0.6037,
      "mean_token_accuracy": 0.8143339529633522,
      "num_tokens": 683136407.0,
      "step": 21404
    },
    {
      "entropy": 0.5890848147682846,
      "epoch": 1.969379129504718,
      "grad_norm": 0.15255151689052582,
      "learning_rate": 3.435765326463643e-05,
      "loss": 0.5853,
      "mean_token_accuracy": 0.8179485648870468,
      "num_tokens": 683168344.0,
      "step": 21405
    },
    {
      "entropy": 0.523764044046402,
      "epoch": 1.9694711358761594,
      "grad_norm": 0.1397446095943451,
      "learning_rate": 3.43545864384948e-05,
      "loss": 0.5119,
      "mean_token_accuracy": 0.8412925973534584,
      "num_tokens": 683200905.0,
      "step": 21406
    },
    {
      "entropy": 0.6558090252801776,
      "epoch": 1.9695631422476008,
      "grad_norm": 0.1516379415988922,
      "learning_rate": 3.4351519612353176e-05,
      "loss": 0.6506,
      "mean_token_accuracy": 0.8009985238313675,
      "num_tokens": 683233043.0,
      "step": 21407
    },
    {
      "entropy": 0.5633107707835734,
      "epoch": 1.969655148619042,
      "grad_norm": 0.15191525220870972,
      "learning_rate": 3.434845278621155e-05,
      "loss": 0.5544,
      "mean_token_accuracy": 0.8295608870685101,
      "num_tokens": 683264918.0,
      "step": 21408
    },
    {
      "entropy": 0.7106415033340454,
      "epoch": 1.969747154990483,
      "grad_norm": 0.16156697273254395,
      "learning_rate": 3.4345385960069925e-05,
      "loss": 0.7177,
      "mean_token_accuracy": 0.7824672870337963,
      "num_tokens": 683297338.0,
      "step": 21409
    },
    {
      "entropy": 0.7261647013947368,
      "epoch": 1.9698391613619242,
      "grad_norm": 0.16473956406116486,
      "learning_rate": 3.43423191339283e-05,
      "loss": 0.7152,
      "mean_token_accuracy": 0.7824422940611839,
      "num_tokens": 683328460.0,
      "step": 21410
    },
    {
      "entropy": 0.6401586122810841,
      "epoch": 1.9699311677333655,
      "grad_norm": 0.15278072655200958,
      "learning_rate": 3.4339252307786674e-05,
      "loss": 0.6282,
      "mean_token_accuracy": 0.8036460839211941,
      "num_tokens": 683360510.0,
      "step": 21411
    },
    {
      "entropy": 0.5895881373435259,
      "epoch": 1.9700231741048069,
      "grad_norm": 0.15190927684307098,
      "learning_rate": 3.433618548164504e-05,
      "loss": 0.5989,
      "mean_token_accuracy": 0.818682212382555,
      "num_tokens": 683392011.0,
      "step": 21412
    },
    {
      "entropy": 0.6137185674160719,
      "epoch": 1.970115180476248,
      "grad_norm": 0.15487445890903473,
      "learning_rate": 3.4333118655503423e-05,
      "loss": 0.6057,
      "mean_token_accuracy": 0.8141805306077003,
      "num_tokens": 683424066.0,
      "step": 21413
    },
    {
      "entropy": 0.5728686396032572,
      "epoch": 1.9702071868476891,
      "grad_norm": 0.15764054656028748,
      "learning_rate": 3.43300518293618e-05,
      "loss": 0.5702,
      "mean_token_accuracy": 0.8238586559891701,
      "num_tokens": 683455754.0,
      "step": 21414
    },
    {
      "entropy": 0.6097668446600437,
      "epoch": 1.9702991932191303,
      "grad_norm": 0.15507781505584717,
      "learning_rate": 3.4326985003220166e-05,
      "loss": 0.6026,
      "mean_token_accuracy": 0.8137948624789715,
      "num_tokens": 683486720.0,
      "step": 21415
    },
    {
      "entropy": 0.5318193025887012,
      "epoch": 1.9703911995905716,
      "grad_norm": 0.14574219286441803,
      "learning_rate": 3.432391817707855e-05,
      "loss": 0.5194,
      "mean_token_accuracy": 0.836984746158123,
      "num_tokens": 683518652.0,
      "step": 21416
    },
    {
      "entropy": 0.5497919674962759,
      "epoch": 1.970483205962013,
      "grad_norm": 0.14910760521888733,
      "learning_rate": 3.4320851350936915e-05,
      "loss": 0.5441,
      "mean_token_accuracy": 0.83118586987257,
      "num_tokens": 683551420.0,
      "step": 21417
    },
    {
      "entropy": 0.6441788831725717,
      "epoch": 1.9705752123334541,
      "grad_norm": 0.1571865975856781,
      "learning_rate": 3.43177845247953e-05,
      "loss": 0.6418,
      "mean_token_accuracy": 0.8050787821412086,
      "num_tokens": 683583738.0,
      "step": 21418
    },
    {
      "entropy": 0.663470359519124,
      "epoch": 1.9706672187048953,
      "grad_norm": 0.1586083024740219,
      "learning_rate": 3.4314717698653665e-05,
      "loss": 0.6562,
      "mean_token_accuracy": 0.7969927489757538,
      "num_tokens": 683615523.0,
      "step": 21419
    },
    {
      "entropy": 0.5623294413089752,
      "epoch": 1.9707592250763364,
      "grad_norm": 0.14872239530086517,
      "learning_rate": 3.431165087251204e-05,
      "loss": 0.5503,
      "mean_token_accuracy": 0.8291430249810219,
      "num_tokens": 683647245.0,
      "step": 21420
    },
    {
      "entropy": 0.619787291623652,
      "epoch": 1.9708512314477777,
      "grad_norm": 0.1590592861175537,
      "learning_rate": 3.4308584046370414e-05,
      "loss": 0.6165,
      "mean_token_accuracy": 0.8115459308028221,
      "num_tokens": 683679540.0,
      "step": 21421
    },
    {
      "entropy": 0.6372888563200831,
      "epoch": 1.970943237819219,
      "grad_norm": 0.16018228232860565,
      "learning_rate": 3.430551722022879e-05,
      "loss": 0.6356,
      "mean_token_accuracy": 0.8050528839230537,
      "num_tokens": 683711669.0,
      "step": 21422
    },
    {
      "entropy": 0.6113044181838632,
      "epoch": 1.9710352441906602,
      "grad_norm": 0.15173576772212982,
      "learning_rate": 3.430245039408716e-05,
      "loss": 0.5882,
      "mean_token_accuracy": 0.8208930268883705,
      "num_tokens": 683744017.0,
      "step": 21423
    },
    {
      "entropy": 0.6252741161733866,
      "epoch": 1.9711272505621014,
      "grad_norm": 0.14856979250907898,
      "learning_rate": 3.429938356794554e-05,
      "loss": 0.6017,
      "mean_token_accuracy": 0.813473392277956,
      "num_tokens": 683775749.0,
      "step": 21424
    },
    {
      "entropy": 0.5924314968287945,
      "epoch": 1.9712192569335425,
      "grad_norm": 0.15314660966396332,
      "learning_rate": 3.4296316741803906e-05,
      "loss": 0.5705,
      "mean_token_accuracy": 0.8231101594865322,
      "num_tokens": 683806446.0,
      "step": 21425
    },
    {
      "entropy": 0.6036200849339366,
      "epoch": 1.9713112633049839,
      "grad_norm": 0.15114718675613403,
      "learning_rate": 3.429324991566229e-05,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8239055052399635,
      "num_tokens": 683838430.0,
      "step": 21426
    },
    {
      "entropy": 0.7704555299133062,
      "epoch": 1.9714032696764252,
      "grad_norm": 0.16827833652496338,
      "learning_rate": 3.4290183089520655e-05,
      "loss": 0.7564,
      "mean_token_accuracy": 0.770058162510395,
      "num_tokens": 683870113.0,
      "step": 21427
    },
    {
      "entropy": 0.5379094574600458,
      "epoch": 1.9714952760478663,
      "grad_norm": 0.14701834321022034,
      "learning_rate": 3.428711626337903e-05,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8314428776502609,
      "num_tokens": 683902269.0,
      "step": 21428
    },
    {
      "entropy": 0.6084657143801451,
      "epoch": 1.9715872824193075,
      "grad_norm": 0.15546192228794098,
      "learning_rate": 3.4284049437237404e-05,
      "loss": 0.5874,
      "mean_token_accuracy": 0.8167893476784229,
      "num_tokens": 683933758.0,
      "step": 21429
    },
    {
      "entropy": 0.5665991571731865,
      "epoch": 1.9716792887907486,
      "grad_norm": 0.14716596901416779,
      "learning_rate": 3.428098261109578e-05,
      "loss": 0.5632,
      "mean_token_accuracy": 0.8251269124448299,
      "num_tokens": 683966148.0,
      "step": 21430
    },
    {
      "entropy": 0.6635493636131287,
      "epoch": 1.97177129516219,
      "grad_norm": 0.1603752225637436,
      "learning_rate": 3.4277915784954154e-05,
      "loss": 0.6442,
      "mean_token_accuracy": 0.803862065076828,
      "num_tokens": 683997385.0,
      "step": 21431
    },
    {
      "entropy": 0.6191359106451273,
      "epoch": 1.9718633015336313,
      "grad_norm": 0.15537938475608826,
      "learning_rate": 3.427484895881253e-05,
      "loss": 0.6123,
      "mean_token_accuracy": 0.8134285770356655,
      "num_tokens": 684028480.0,
      "step": 21432
    },
    {
      "entropy": 0.5575876105576754,
      "epoch": 1.9719553079050725,
      "grad_norm": 0.14889322221279144,
      "learning_rate": 3.4271782132670896e-05,
      "loss": 0.5299,
      "mean_token_accuracy": 0.8347550332546234,
      "num_tokens": 684060249.0,
      "step": 21433
    },
    {
      "entropy": 0.4755310812033713,
      "epoch": 1.9720473142765136,
      "grad_norm": 0.14476646482944489,
      "learning_rate": 3.426871530652928e-05,
      "loss": 0.4609,
      "mean_token_accuracy": 0.8541630543768406,
      "num_tokens": 684092082.0,
      "step": 21434
    },
    {
      "entropy": 0.6148709934204817,
      "epoch": 1.9721393206479547,
      "grad_norm": 0.15935547649860382,
      "learning_rate": 3.4265648480387645e-05,
      "loss": 0.601,
      "mean_token_accuracy": 0.8175532855093479,
      "num_tokens": 684123474.0,
      "step": 21435
    },
    {
      "entropy": 0.5256874412298203,
      "epoch": 1.972231327019396,
      "grad_norm": 0.14762838184833527,
      "learning_rate": 3.426258165424602e-05,
      "loss": 0.5123,
      "mean_token_accuracy": 0.8372753895819187,
      "num_tokens": 684155366.0,
      "step": 21436
    },
    {
      "entropy": 0.4019956816919148,
      "epoch": 1.9723233333908374,
      "grad_norm": 0.1376553326845169,
      "learning_rate": 3.4259514828104395e-05,
      "loss": 0.3882,
      "mean_token_accuracy": 0.8775302544236183,
      "num_tokens": 684187699.0,
      "step": 21437
    },
    {
      "entropy": 0.636507575167343,
      "epoch": 1.9724153397622786,
      "grad_norm": 0.16043657064437866,
      "learning_rate": 3.425644800196277e-05,
      "loss": 0.633,
      "mean_token_accuracy": 0.8075433112680912,
      "num_tokens": 684218530.0,
      "step": 21438
    },
    {
      "entropy": 0.5969919352792203,
      "epoch": 1.9725073461337197,
      "grad_norm": 0.14872904121875763,
      "learning_rate": 3.4253381175821144e-05,
      "loss": 0.5911,
      "mean_token_accuracy": 0.8170578926801682,
      "num_tokens": 684250984.0,
      "step": 21439
    },
    {
      "entropy": 0.537143609020859,
      "epoch": 1.9725993525051608,
      "grad_norm": 0.15040276944637299,
      "learning_rate": 3.425031434967952e-05,
      "loss": 0.5334,
      "mean_token_accuracy": 0.8353098407387733,
      "num_tokens": 684282977.0,
      "step": 21440
    },
    {
      "entropy": 0.5556581420823932,
      "epoch": 1.9726913588766022,
      "grad_norm": 0.15150289237499237,
      "learning_rate": 3.4247247523537886e-05,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8252881281077862,
      "num_tokens": 684315030.0,
      "step": 21441
    },
    {
      "entropy": 0.5979395918548107,
      "epoch": 1.9727833652480435,
      "grad_norm": 0.1503228396177292,
      "learning_rate": 3.424418069739627e-05,
      "loss": 0.5924,
      "mean_token_accuracy": 0.8185307532548904,
      "num_tokens": 684347574.0,
      "step": 21442
    },
    {
      "entropy": 0.6895224787294865,
      "epoch": 1.9728753716194847,
      "grad_norm": 0.15701378881931305,
      "learning_rate": 3.4241113871254636e-05,
      "loss": 0.6888,
      "mean_token_accuracy": 0.7914458028972149,
      "num_tokens": 684379529.0,
      "step": 21443
    },
    {
      "entropy": 0.661044342443347,
      "epoch": 1.9729673779909258,
      "grad_norm": 0.15715184807777405,
      "learning_rate": 3.423804704511302e-05,
      "loss": 0.6517,
      "mean_token_accuracy": 0.7979865036904812,
      "num_tokens": 684411912.0,
      "step": 21444
    },
    {
      "entropy": 0.5268629882484674,
      "epoch": 1.973059384362367,
      "grad_norm": 0.14724105596542358,
      "learning_rate": 3.423498021897139e-05,
      "loss": 0.5204,
      "mean_token_accuracy": 0.83918521925807,
      "num_tokens": 684444301.0,
      "step": 21445
    },
    {
      "entropy": 0.530481830239296,
      "epoch": 1.9731513907338083,
      "grad_norm": 0.14606991410255432,
      "learning_rate": 3.423191339282976e-05,
      "loss": 0.5137,
      "mean_token_accuracy": 0.84301633015275,
      "num_tokens": 684476301.0,
      "step": 21446
    },
    {
      "entropy": 0.6506156735122204,
      "epoch": 1.9732433971052497,
      "grad_norm": 0.1604631394147873,
      "learning_rate": 3.422884656668814e-05,
      "loss": 0.6328,
      "mean_token_accuracy": 0.8049527704715729,
      "num_tokens": 684508088.0,
      "step": 21447
    },
    {
      "entropy": 0.7059943024069071,
      "epoch": 1.9733354034766908,
      "grad_norm": 0.16334645450115204,
      "learning_rate": 3.422577974054651e-05,
      "loss": 0.6863,
      "mean_token_accuracy": 0.7904835231602192,
      "num_tokens": 684539339.0,
      "step": 21448
    },
    {
      "entropy": 0.6202019285410643,
      "epoch": 1.973427409848132,
      "grad_norm": 0.1528160721063614,
      "learning_rate": 3.4222712914404884e-05,
      "loss": 0.6154,
      "mean_token_accuracy": 0.8126885108649731,
      "num_tokens": 684570828.0,
      "step": 21449
    },
    {
      "entropy": 0.6597518157213926,
      "epoch": 1.973519416219573,
      "grad_norm": 0.1591959297657013,
      "learning_rate": 3.421964608826326e-05,
      "loss": 0.6545,
      "mean_token_accuracy": 0.8021917194128036,
      "num_tokens": 684602901.0,
      "step": 21450
    },
    {
      "entropy": 0.6724513340741396,
      "epoch": 1.9736114225910144,
      "grad_norm": 0.1599515676498413,
      "learning_rate": 3.421657926212163e-05,
      "loss": 0.6635,
      "mean_token_accuracy": 0.7979596108198166,
      "num_tokens": 684635200.0,
      "step": 21451
    },
    {
      "entropy": 0.6633821707218885,
      "epoch": 1.9737034289624558,
      "grad_norm": 0.1577792763710022,
      "learning_rate": 3.421351243598001e-05,
      "loss": 0.6572,
      "mean_token_accuracy": 0.7982345074415207,
      "num_tokens": 684667042.0,
      "step": 21452
    },
    {
      "entropy": 0.6324004484340549,
      "epoch": 1.973795435333897,
      "grad_norm": 0.15697382390499115,
      "learning_rate": 3.421044560983838e-05,
      "loss": 0.6095,
      "mean_token_accuracy": 0.8114731721580029,
      "num_tokens": 684698099.0,
      "step": 21453
    },
    {
      "entropy": 0.6934664370492101,
      "epoch": 1.973887441705338,
      "grad_norm": 0.1628764569759369,
      "learning_rate": 3.420737878369675e-05,
      "loss": 0.709,
      "mean_token_accuracy": 0.7902499809861183,
      "num_tokens": 684729929.0,
      "step": 21454
    },
    {
      "entropy": 0.6516886465251446,
      "epoch": 1.9739794480767792,
      "grad_norm": 0.1647402048110962,
      "learning_rate": 3.420431195755513e-05,
      "loss": 0.6408,
      "mean_token_accuracy": 0.801408227533102,
      "num_tokens": 684761092.0,
      "step": 21455
    },
    {
      "entropy": 0.5911430490668863,
      "epoch": 1.9740714544482205,
      "grad_norm": 0.14874491095542908,
      "learning_rate": 3.42012451314135e-05,
      "loss": 0.5878,
      "mean_token_accuracy": 0.8171895779669285,
      "num_tokens": 684793239.0,
      "step": 21456
    },
    {
      "entropy": 0.7104524467140436,
      "epoch": 1.9741634608196619,
      "grad_norm": 0.1649232804775238,
      "learning_rate": 3.4198178305271874e-05,
      "loss": 0.7064,
      "mean_token_accuracy": 0.7871144413948059,
      "num_tokens": 684824423.0,
      "step": 21457
    },
    {
      "entropy": 0.6184809301048517,
      "epoch": 1.974255467191103,
      "grad_norm": 0.15808239579200745,
      "learning_rate": 3.419511147913025e-05,
      "loss": 0.6202,
      "mean_token_accuracy": 0.8070211745798588,
      "num_tokens": 684856240.0,
      "step": 21458
    },
    {
      "entropy": 0.7260202690958977,
      "epoch": 1.9743474735625441,
      "grad_norm": 0.16122473776340485,
      "learning_rate": 3.419204465298862e-05,
      "loss": 0.6961,
      "mean_token_accuracy": 0.785622276365757,
      "num_tokens": 684887347.0,
      "step": 21459
    },
    {
      "entropy": 0.6274260990321636,
      "epoch": 1.9744394799339853,
      "grad_norm": 0.15224699676036835,
      "learning_rate": 3.4188977826847e-05,
      "loss": 0.6126,
      "mean_token_accuracy": 0.8093602061271667,
      "num_tokens": 684918798.0,
      "step": 21460
    },
    {
      "entropy": 0.5422701810020953,
      "epoch": 1.9745314863054266,
      "grad_norm": 0.14964930713176727,
      "learning_rate": 3.418591100070537e-05,
      "loss": 0.5389,
      "mean_token_accuracy": 0.8318924568593502,
      "num_tokens": 684950209.0,
      "step": 21461
    },
    {
      "entropy": 0.533922410570085,
      "epoch": 1.974623492676868,
      "grad_norm": 0.15221822261810303,
      "learning_rate": 3.418284417456375e-05,
      "loss": 0.5292,
      "mean_token_accuracy": 0.8386295661330223,
      "num_tokens": 684982311.0,
      "step": 21462
    },
    {
      "entropy": 0.6028763633221388,
      "epoch": 1.9747154990483091,
      "grad_norm": 0.15538588166236877,
      "learning_rate": 3.417977734842212e-05,
      "loss": 0.5946,
      "mean_token_accuracy": 0.8178098127245903,
      "num_tokens": 685014811.0,
      "step": 21463
    },
    {
      "entropy": 0.7243824601173401,
      "epoch": 1.9748075054197503,
      "grad_norm": 0.16282619535923004,
      "learning_rate": 3.417671052228049e-05,
      "loss": 0.7028,
      "mean_token_accuracy": 0.7878199145197868,
      "num_tokens": 685045969.0,
      "step": 21464
    },
    {
      "entropy": 0.6954742046073079,
      "epoch": 1.9748995117911914,
      "grad_norm": 0.16118976473808289,
      "learning_rate": 3.417364369613887e-05,
      "loss": 0.6666,
      "mean_token_accuracy": 0.7944438718259335,
      "num_tokens": 685076997.0,
      "step": 21465
    },
    {
      "entropy": 0.6848119571805,
      "epoch": 1.9749915181626327,
      "grad_norm": 0.1590157300233841,
      "learning_rate": 3.417057686999724e-05,
      "loss": 0.686,
      "mean_token_accuracy": 0.7932498455047607,
      "num_tokens": 685108705.0,
      "step": 21466
    },
    {
      "entropy": 0.6461494071409106,
      "epoch": 1.975083524534074,
      "grad_norm": 0.1514071375131607,
      "learning_rate": 3.4167510043855614e-05,
      "loss": 0.62,
      "mean_token_accuracy": 0.8084757216274738,
      "num_tokens": 685141035.0,
      "step": 21467
    },
    {
      "entropy": 0.6306236321106553,
      "epoch": 1.9751755309055152,
      "grad_norm": 0.15200179815292358,
      "learning_rate": 3.416444321771399e-05,
      "loss": 0.6083,
      "mean_token_accuracy": 0.8121722899377346,
      "num_tokens": 685173009.0,
      "step": 21468
    },
    {
      "entropy": 0.49566339142620564,
      "epoch": 1.9752675372769564,
      "grad_norm": 0.14095085859298706,
      "learning_rate": 3.416137639157236e-05,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8485664539039135,
      "num_tokens": 685205623.0,
      "step": 21469
    },
    {
      "entropy": 0.6087652468122542,
      "epoch": 1.9753595436483975,
      "grad_norm": 0.16513505578041077,
      "learning_rate": 3.415830956543074e-05,
      "loss": 0.6064,
      "mean_token_accuracy": 0.8123718574643135,
      "num_tokens": 685237251.0,
      "step": 21470
    },
    {
      "entropy": 0.6451842971146107,
      "epoch": 1.9754515500198389,
      "grad_norm": 0.15827101469039917,
      "learning_rate": 3.415524273928911e-05,
      "loss": 0.6337,
      "mean_token_accuracy": 0.8039448969066143,
      "num_tokens": 685268569.0,
      "step": 21471
    },
    {
      "entropy": 0.81525669246912,
      "epoch": 1.9755435563912802,
      "grad_norm": 0.17719434201717377,
      "learning_rate": 3.415217591314748e-05,
      "loss": 0.8131,
      "mean_token_accuracy": 0.7558621726930141,
      "num_tokens": 685299837.0,
      "step": 21472
    },
    {
      "entropy": 0.6441381014883518,
      "epoch": 1.9756355627627213,
      "grad_norm": 0.1557210385799408,
      "learning_rate": 3.414910908700586e-05,
      "loss": 0.6354,
      "mean_token_accuracy": 0.8053754456341267,
      "num_tokens": 685331528.0,
      "step": 21473
    },
    {
      "entropy": 0.5530406143516302,
      "epoch": 1.9757275691341625,
      "grad_norm": 0.1579904556274414,
      "learning_rate": 3.414604226086423e-05,
      "loss": 0.5402,
      "mean_token_accuracy": 0.8309353888034821,
      "num_tokens": 685362264.0,
      "step": 21474
    },
    {
      "entropy": 0.6118048983626068,
      "epoch": 1.9758195755056036,
      "grad_norm": 0.15499895811080933,
      "learning_rate": 3.4142975434722604e-05,
      "loss": 0.6004,
      "mean_token_accuracy": 0.815332368016243,
      "num_tokens": 685393777.0,
      "step": 21475
    },
    {
      "entropy": 0.6054152557626367,
      "epoch": 1.975911581877045,
      "grad_norm": 0.15262916684150696,
      "learning_rate": 3.4139908608580986e-05,
      "loss": 0.5948,
      "mean_token_accuracy": 0.817244429141283,
      "num_tokens": 685426023.0,
      "step": 21476
    },
    {
      "entropy": 0.612221862655133,
      "epoch": 1.9760035882484863,
      "grad_norm": 0.1554730385541916,
      "learning_rate": 3.4136841782439353e-05,
      "loss": 0.6182,
      "mean_token_accuracy": 0.8132283799350262,
      "num_tokens": 685457774.0,
      "step": 21477
    },
    {
      "entropy": 0.5510639250278473,
      "epoch": 1.9760955946199275,
      "grad_norm": 0.14529447257518768,
      "learning_rate": 3.4133774956297735e-05,
      "loss": 0.5404,
      "mean_token_accuracy": 0.8334782645106316,
      "num_tokens": 685490366.0,
      "step": 21478
    },
    {
      "entropy": 0.5817605257034302,
      "epoch": 1.9761876009913686,
      "grad_norm": 0.15216189622879028,
      "learning_rate": 3.41307081301561e-05,
      "loss": 0.5816,
      "mean_token_accuracy": 0.8218121789395809,
      "num_tokens": 685522188.0,
      "step": 21479
    },
    {
      "entropy": 0.5495714927092195,
      "epoch": 1.9762796073628097,
      "grad_norm": 0.1535530537366867,
      "learning_rate": 3.412764130401448e-05,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8290602490305901,
      "num_tokens": 685554306.0,
      "step": 21480
    },
    {
      "entropy": 0.6377737009897828,
      "epoch": 1.976371613734251,
      "grad_norm": 0.15304884314537048,
      "learning_rate": 3.412457447787285e-05,
      "loss": 0.6319,
      "mean_token_accuracy": 0.8063328452408314,
      "num_tokens": 685586323.0,
      "step": 21481
    },
    {
      "entropy": 0.6181062357500196,
      "epoch": 1.9764636201056924,
      "grad_norm": 0.15567468106746674,
      "learning_rate": 3.412150765173123e-05,
      "loss": 0.6173,
      "mean_token_accuracy": 0.8105673119425774,
      "num_tokens": 685617640.0,
      "step": 21482
    },
    {
      "entropy": 0.5497356969863176,
      "epoch": 1.9765556264771336,
      "grad_norm": 0.14799633622169495,
      "learning_rate": 3.41184408255896e-05,
      "loss": 0.5316,
      "mean_token_accuracy": 0.8333574086427689,
      "num_tokens": 685649674.0,
      "step": 21483
    },
    {
      "entropy": 0.5169323584996164,
      "epoch": 1.9766476328485747,
      "grad_norm": 0.13725125789642334,
      "learning_rate": 3.4115373999447976e-05,
      "loss": 0.5106,
      "mean_token_accuracy": 0.8417587988078594,
      "num_tokens": 685682010.0,
      "step": 21484
    },
    {
      "entropy": 0.628311138600111,
      "epoch": 1.9767396392200158,
      "grad_norm": 0.16062438488006592,
      "learning_rate": 3.4112307173306344e-05,
      "loss": 0.6177,
      "mean_token_accuracy": 0.8119990490376949,
      "num_tokens": 685714266.0,
      "step": 21485
    },
    {
      "entropy": 0.4883853674400598,
      "epoch": 1.9768316455914572,
      "grad_norm": 0.14434432983398438,
      "learning_rate": 3.4109240347164725e-05,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8506960906088352,
      "num_tokens": 685746109.0,
      "step": 21486
    },
    {
      "entropy": 0.5804145038127899,
      "epoch": 1.9769236519628985,
      "grad_norm": 0.14945392310619354,
      "learning_rate": 3.410617352102309e-05,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8214714340865612,
      "num_tokens": 685778382.0,
      "step": 21487
    },
    {
      "entropy": 0.5433510094881058,
      "epoch": 1.9770156583343397,
      "grad_norm": 0.14864309132099152,
      "learning_rate": 3.410310669488147e-05,
      "loss": 0.526,
      "mean_token_accuracy": 0.8360878191888332,
      "num_tokens": 685810264.0,
      "step": 21488
    },
    {
      "entropy": 0.6205571787431836,
      "epoch": 1.9771076647057808,
      "grad_norm": 0.15090788900852203,
      "learning_rate": 3.410003986873984e-05,
      "loss": 0.5964,
      "mean_token_accuracy": 0.8148259222507477,
      "num_tokens": 685842912.0,
      "step": 21489
    },
    {
      "entropy": 0.5167519534006715,
      "epoch": 1.977199671077222,
      "grad_norm": 0.14258718490600586,
      "learning_rate": 3.409697304259822e-05,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8431419879198074,
      "num_tokens": 685874482.0,
      "step": 21490
    },
    {
      "entropy": 0.6386698577553034,
      "epoch": 1.9772916774486633,
      "grad_norm": 0.15147005021572113,
      "learning_rate": 3.409390621645659e-05,
      "loss": 0.6204,
      "mean_token_accuracy": 0.8099080361425877,
      "num_tokens": 685906878.0,
      "step": 21491
    },
    {
      "entropy": 0.6865945588797331,
      "epoch": 1.9773836838201047,
      "grad_norm": 0.15694372355937958,
      "learning_rate": 3.4090839390314966e-05,
      "loss": 0.6654,
      "mean_token_accuracy": 0.7957076020538807,
      "num_tokens": 685938735.0,
      "step": 21492
    },
    {
      "entropy": 0.6011396278627217,
      "epoch": 1.9774756901915458,
      "grad_norm": 0.15504860877990723,
      "learning_rate": 3.4087772564173334e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8233883194625378,
      "num_tokens": 685970974.0,
      "step": 21493
    },
    {
      "entropy": 0.5582927335053682,
      "epoch": 1.977567696562987,
      "grad_norm": 0.15838909149169922,
      "learning_rate": 3.4084705738031716e-05,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8314084969460964,
      "num_tokens": 686002017.0,
      "step": 21494
    },
    {
      "entropy": 0.5547796729952097,
      "epoch": 1.977659702934428,
      "grad_norm": 0.15038420259952545,
      "learning_rate": 3.4081638911890084e-05,
      "loss": 0.5426,
      "mean_token_accuracy": 0.8329500742256641,
      "num_tokens": 686033527.0,
      "step": 21495
    },
    {
      "entropy": 0.5355968866497278,
      "epoch": 1.9777517093058694,
      "grad_norm": 0.14215081930160522,
      "learning_rate": 3.407857208574846e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8360245898365974,
      "num_tokens": 686065576.0,
      "step": 21496
    },
    {
      "entropy": 0.6079967757686973,
      "epoch": 1.9778437156773108,
      "grad_norm": 0.15157043933868408,
      "learning_rate": 3.407550525960683e-05,
      "loss": 0.5939,
      "mean_token_accuracy": 0.8160593956708908,
      "num_tokens": 686097626.0,
      "step": 21497
    },
    {
      "entropy": 0.6355075808241963,
      "epoch": 1.977935722048752,
      "grad_norm": 0.155294269323349,
      "learning_rate": 3.407243843346521e-05,
      "loss": 0.6254,
      "mean_token_accuracy": 0.8051317222416401,
      "num_tokens": 686129238.0,
      "step": 21498
    },
    {
      "entropy": 0.6721584163606167,
      "epoch": 1.978027728420193,
      "grad_norm": 0.16050034761428833,
      "learning_rate": 3.406937160732358e-05,
      "loss": 0.6653,
      "mean_token_accuracy": 0.7959812171757221,
      "num_tokens": 686161316.0,
      "step": 21499
    },
    {
      "entropy": 0.573052872903645,
      "epoch": 1.9781197347916342,
      "grad_norm": 0.1454995721578598,
      "learning_rate": 3.406630478118196e-05,
      "loss": 0.5598,
      "mean_token_accuracy": 0.8256929516792297,
      "num_tokens": 686193580.0,
      "step": 21500
    },
    {
      "entropy": 0.5951954792253673,
      "epoch": 1.9782117411630755,
      "grad_norm": 0.16368550062179565,
      "learning_rate": 3.4063237955040325e-05,
      "loss": 0.5897,
      "mean_token_accuracy": 0.8208403997123241,
      "num_tokens": 686224923.0,
      "step": 21501
    },
    {
      "entropy": 0.5983079196885228,
      "epoch": 1.9783037475345169,
      "grad_norm": 0.1537751704454422,
      "learning_rate": 3.4060171128898706e-05,
      "loss": 0.5864,
      "mean_token_accuracy": 0.8176681138575077,
      "num_tokens": 686256876.0,
      "step": 21502
    },
    {
      "entropy": 0.5539435464888811,
      "epoch": 1.978395753905958,
      "grad_norm": 0.1521485596895218,
      "learning_rate": 3.4057104302757074e-05,
      "loss": 0.5452,
      "mean_token_accuracy": 0.8327867388725281,
      "num_tokens": 686288855.0,
      "step": 21503
    },
    {
      "entropy": 0.6980605870485306,
      "epoch": 1.9784877602773991,
      "grad_norm": 0.15677133202552795,
      "learning_rate": 3.4054037476615455e-05,
      "loss": 0.7,
      "mean_token_accuracy": 0.7867856845259666,
      "num_tokens": 686321317.0,
      "step": 21504
    },
    {
      "entropy": 0.5819638781249523,
      "epoch": 1.9785797666488403,
      "grad_norm": 0.150906503200531,
      "learning_rate": 3.405097065047382e-05,
      "loss": 0.5606,
      "mean_token_accuracy": 0.8228229209780693,
      "num_tokens": 686353114.0,
      "step": 21505
    },
    {
      "entropy": 0.5085404934361577,
      "epoch": 1.9786717730202816,
      "grad_norm": 0.14526382088661194,
      "learning_rate": 3.40479038243322e-05,
      "loss": 0.5039,
      "mean_token_accuracy": 0.845080491155386,
      "num_tokens": 686384946.0,
      "step": 21506
    },
    {
      "entropy": 0.575903850607574,
      "epoch": 1.978763779391723,
      "grad_norm": 0.15042744576931,
      "learning_rate": 3.404483699819058e-05,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8259587474167347,
      "num_tokens": 686416611.0,
      "step": 21507
    },
    {
      "entropy": 0.6738393139094114,
      "epoch": 1.9788557857631641,
      "grad_norm": 0.1621755212545395,
      "learning_rate": 3.404177017204895e-05,
      "loss": 0.6709,
      "mean_token_accuracy": 0.7954128980636597,
      "num_tokens": 686448316.0,
      "step": 21508
    },
    {
      "entropy": 0.5759008610621095,
      "epoch": 1.9789477921346053,
      "grad_norm": 0.15160968899726868,
      "learning_rate": 3.403870334590732e-05,
      "loss": 0.5655,
      "mean_token_accuracy": 0.8218555264174938,
      "num_tokens": 686480394.0,
      "step": 21509
    },
    {
      "entropy": 0.5467234868556261,
      "epoch": 1.9790397985060464,
      "grad_norm": 0.14952786266803741,
      "learning_rate": 3.4035636519765696e-05,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8323003388941288,
      "num_tokens": 686512373.0,
      "step": 21510
    },
    {
      "entropy": 0.6450721677392721,
      "epoch": 1.9791318048774877,
      "grad_norm": 0.1545727699995041,
      "learning_rate": 3.403256969362407e-05,
      "loss": 0.6438,
      "mean_token_accuracy": 0.7980695441365242,
      "num_tokens": 686544381.0,
      "step": 21511
    },
    {
      "entropy": 0.5437479978427291,
      "epoch": 1.979223811248929,
      "grad_norm": 0.14772140979766846,
      "learning_rate": 3.4029502867482446e-05,
      "loss": 0.5281,
      "mean_token_accuracy": 0.834006030112505,
      "num_tokens": 686576224.0,
      "step": 21512
    },
    {
      "entropy": 0.6454015299677849,
      "epoch": 1.9793158176203702,
      "grad_norm": 0.1532926708459854,
      "learning_rate": 3.402643604134082e-05,
      "loss": 0.6361,
      "mean_token_accuracy": 0.8071601390838623,
      "num_tokens": 686608215.0,
      "step": 21513
    },
    {
      "entropy": 0.5124355256557465,
      "epoch": 1.9794078239918114,
      "grad_norm": 0.15072204172611237,
      "learning_rate": 3.402336921519919e-05,
      "loss": 0.5149,
      "mean_token_accuracy": 0.8389008194208145,
      "num_tokens": 686640283.0,
      "step": 21514
    },
    {
      "entropy": 0.5519167389720678,
      "epoch": 1.9794998303632525,
      "grad_norm": 0.14866478741168976,
      "learning_rate": 3.402030238905757e-05,
      "loss": 0.5358,
      "mean_token_accuracy": 0.8309659250080585,
      "num_tokens": 686672104.0,
      "step": 21515
    },
    {
      "entropy": 0.6032457109540701,
      "epoch": 1.9795918367346939,
      "grad_norm": 0.1553364396095276,
      "learning_rate": 3.401723556291594e-05,
      "loss": 0.6047,
      "mean_token_accuracy": 0.8145595565438271,
      "num_tokens": 686704199.0,
      "step": 21516
    },
    {
      "entropy": 0.6021036256570369,
      "epoch": 1.9796838431061352,
      "grad_norm": 0.15096870064735413,
      "learning_rate": 3.401416873677431e-05,
      "loss": 0.6006,
      "mean_token_accuracy": 0.815658200532198,
      "num_tokens": 686735487.0,
      "step": 21517
    },
    {
      "entropy": 0.5227678455412388,
      "epoch": 1.9797758494775763,
      "grad_norm": 0.14845575392246246,
      "learning_rate": 3.401110191063269e-05,
      "loss": 0.5072,
      "mean_token_accuracy": 0.8425244353711605,
      "num_tokens": 686768092.0,
      "step": 21518
    },
    {
      "entropy": 0.6619526464492083,
      "epoch": 1.9798678558490175,
      "grad_norm": 0.15951813757419586,
      "learning_rate": 3.400803508449106e-05,
      "loss": 0.6557,
      "mean_token_accuracy": 0.8010454103350639,
      "num_tokens": 686800163.0,
      "step": 21519
    },
    {
      "entropy": 0.6093063494190574,
      "epoch": 1.9799598622204586,
      "grad_norm": 0.15651480853557587,
      "learning_rate": 3.4004968258349436e-05,
      "loss": 0.6018,
      "mean_token_accuracy": 0.8158976882696152,
      "num_tokens": 686832451.0,
      "step": 21520
    },
    {
      "entropy": 0.680887321010232,
      "epoch": 1.9800518685919,
      "grad_norm": 0.1619807630777359,
      "learning_rate": 3.400190143220781e-05,
      "loss": 0.6705,
      "mean_token_accuracy": 0.7969597466289997,
      "num_tokens": 686865064.0,
      "step": 21521
    },
    {
      "entropy": 0.6612665299326181,
      "epoch": 1.9801438749633413,
      "grad_norm": 0.15712182223796844,
      "learning_rate": 3.3998834606066185e-05,
      "loss": 0.6503,
      "mean_token_accuracy": 0.8010455146431923,
      "num_tokens": 686897250.0,
      "step": 21522
    },
    {
      "entropy": 0.6273528570309281,
      "epoch": 1.9802358813347825,
      "grad_norm": 0.15080121159553528,
      "learning_rate": 3.399576777992456e-05,
      "loss": 0.617,
      "mean_token_accuracy": 0.8085630647838116,
      "num_tokens": 686929193.0,
      "step": 21523
    },
    {
      "entropy": 0.6673552580177784,
      "epoch": 1.9803278877062236,
      "grad_norm": 0.15354032814502716,
      "learning_rate": 3.399270095378293e-05,
      "loss": 0.6485,
      "mean_token_accuracy": 0.7976608164608479,
      "num_tokens": 686961049.0,
      "step": 21524
    },
    {
      "entropy": 0.5515714343637228,
      "epoch": 1.9804198940776647,
      "grad_norm": 0.14315006136894226,
      "learning_rate": 3.398963412764131e-05,
      "loss": 0.5243,
      "mean_token_accuracy": 0.840230654925108,
      "num_tokens": 686992727.0,
      "step": 21525
    },
    {
      "entropy": 0.5994752231054008,
      "epoch": 1.980511900449106,
      "grad_norm": 0.15401817858219147,
      "learning_rate": 3.398656730149968e-05,
      "loss": 0.5764,
      "mean_token_accuracy": 0.8187536522746086,
      "num_tokens": 687024782.0,
      "step": 21526
    },
    {
      "entropy": 0.5996298585087061,
      "epoch": 1.9806039068205474,
      "grad_norm": 0.14796216785907745,
      "learning_rate": 3.398350047535805e-05,
      "loss": 0.5997,
      "mean_token_accuracy": 0.8153866082429886,
      "num_tokens": 687057194.0,
      "step": 21527
    },
    {
      "entropy": 0.5730878296308219,
      "epoch": 1.9806959131919886,
      "grad_norm": 0.15332931280136108,
      "learning_rate": 3.3980433649216427e-05,
      "loss": 0.5709,
      "mean_token_accuracy": 0.8262561038136482,
      "num_tokens": 687089295.0,
      "step": 21528
    },
    {
      "entropy": 0.5438798626419157,
      "epoch": 1.9807879195634297,
      "grad_norm": 0.1501573920249939,
      "learning_rate": 3.39773668230748e-05,
      "loss": 0.5293,
      "mean_token_accuracy": 0.8344556950032711,
      "num_tokens": 687121483.0,
      "step": 21529
    },
    {
      "entropy": 0.6443743966519833,
      "epoch": 1.9808799259348708,
      "grad_norm": 0.15814463794231415,
      "learning_rate": 3.3974299996933176e-05,
      "loss": 0.6323,
      "mean_token_accuracy": 0.805807750672102,
      "num_tokens": 687153789.0,
      "step": 21530
    },
    {
      "entropy": 0.4709599195048213,
      "epoch": 1.9809719323063122,
      "grad_norm": 0.14645789563655853,
      "learning_rate": 3.397123317079155e-05,
      "loss": 0.4566,
      "mean_token_accuracy": 0.8577055558562279,
      "num_tokens": 687186168.0,
      "step": 21531
    },
    {
      "entropy": 0.5115546183660626,
      "epoch": 1.9810639386777535,
      "grad_norm": 0.14533913135528564,
      "learning_rate": 3.396816634464992e-05,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8381548374891281,
      "num_tokens": 687218732.0,
      "step": 21532
    },
    {
      "entropy": 0.5991612672805786,
      "epoch": 1.9811559450491947,
      "grad_norm": 0.1561659574508667,
      "learning_rate": 3.39650995185083e-05,
      "loss": 0.5943,
      "mean_token_accuracy": 0.8169876709580421,
      "num_tokens": 687250532.0,
      "step": 21533
    },
    {
      "entropy": 0.6244137678295374,
      "epoch": 1.9812479514206358,
      "grad_norm": 0.15580038726329803,
      "learning_rate": 3.396203269236667e-05,
      "loss": 0.6207,
      "mean_token_accuracy": 0.8101564981043339,
      "num_tokens": 687282132.0,
      "step": 21534
    },
    {
      "entropy": 0.5416905395686626,
      "epoch": 1.981339957792077,
      "grad_norm": 0.15244272351264954,
      "learning_rate": 3.395896586622504e-05,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8304644413292408,
      "num_tokens": 687313502.0,
      "step": 21535
    },
    {
      "entropy": 0.4373997875954956,
      "epoch": 1.9814319641635183,
      "grad_norm": 0.13429971039295197,
      "learning_rate": 3.3955899040083424e-05,
      "loss": 0.425,
      "mean_token_accuracy": 0.8652217090129852,
      "num_tokens": 687345831.0,
      "step": 21536
    },
    {
      "entropy": 0.5907880926970392,
      "epoch": 1.9815239705349597,
      "grad_norm": 0.15270453691482544,
      "learning_rate": 3.395283221394179e-05,
      "loss": 0.5892,
      "mean_token_accuracy": 0.8203561455011368,
      "num_tokens": 687377581.0,
      "step": 21537
    },
    {
      "entropy": 0.5020901318639517,
      "epoch": 1.9816159769064008,
      "grad_norm": 0.1385432630777359,
      "learning_rate": 3.394976538780017e-05,
      "loss": 0.4998,
      "mean_token_accuracy": 0.8419103659689426,
      "num_tokens": 687409683.0,
      "step": 21538
    },
    {
      "entropy": 0.6375784110277891,
      "epoch": 1.981707983277842,
      "grad_norm": 0.15487627685070038,
      "learning_rate": 3.394669856165854e-05,
      "loss": 0.619,
      "mean_token_accuracy": 0.8130454458296299,
      "num_tokens": 687441829.0,
      "step": 21539
    },
    {
      "entropy": 0.5601273539941758,
      "epoch": 1.981799989649283,
      "grad_norm": 0.1475592851638794,
      "learning_rate": 3.3943631735516916e-05,
      "loss": 0.5506,
      "mean_token_accuracy": 0.8303366228938103,
      "num_tokens": 687473905.0,
      "step": 21540
    },
    {
      "entropy": 0.6067452523857355,
      "epoch": 1.9818919960207244,
      "grad_norm": 0.14932624995708466,
      "learning_rate": 3.394056490937529e-05,
      "loss": 0.5918,
      "mean_token_accuracy": 0.8152428604662418,
      "num_tokens": 687506421.0,
      "step": 21541
    },
    {
      "entropy": 0.656351663172245,
      "epoch": 1.9819840023921658,
      "grad_norm": 0.1547098606824875,
      "learning_rate": 3.3937498083233665e-05,
      "loss": 0.6431,
      "mean_token_accuracy": 0.8031336925923824,
      "num_tokens": 687538858.0,
      "step": 21542
    },
    {
      "entropy": 0.5865275617688894,
      "epoch": 1.982076008763607,
      "grad_norm": 0.15085817873477936,
      "learning_rate": 3.393443125709204e-05,
      "loss": 0.5732,
      "mean_token_accuracy": 0.8260928578674793,
      "num_tokens": 687569949.0,
      "step": 21543
    },
    {
      "entropy": 0.5647258441895247,
      "epoch": 1.982168015135048,
      "grad_norm": 0.14943087100982666,
      "learning_rate": 3.3931364430950414e-05,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8311621882021427,
      "num_tokens": 687600759.0,
      "step": 21544
    },
    {
      "entropy": 0.6731502693146467,
      "epoch": 1.9822600215064892,
      "grad_norm": 0.15685850381851196,
      "learning_rate": 3.392829760480878e-05,
      "loss": 0.6583,
      "mean_token_accuracy": 0.8009099178016186,
      "num_tokens": 687632939.0,
      "step": 21545
    },
    {
      "entropy": 0.6163015654310584,
      "epoch": 1.9823520278779305,
      "grad_norm": 0.14433985948562622,
      "learning_rate": 3.3925230778667163e-05,
      "loss": 0.5958,
      "mean_token_accuracy": 0.8172035627067089,
      "num_tokens": 687664918.0,
      "step": 21546
    },
    {
      "entropy": 0.598819263279438,
      "epoch": 1.9824440342493719,
      "grad_norm": 0.153034970164299,
      "learning_rate": 3.392216395252553e-05,
      "loss": 0.5797,
      "mean_token_accuracy": 0.8217371553182602,
      "num_tokens": 687696316.0,
      "step": 21547
    },
    {
      "entropy": 0.6429418316110969,
      "epoch": 1.982536040620813,
      "grad_norm": 0.15573394298553467,
      "learning_rate": 3.3919097126383906e-05,
      "loss": 0.6302,
      "mean_token_accuracy": 0.8082378283143044,
      "num_tokens": 687727179.0,
      "step": 21548
    },
    {
      "entropy": 0.6416869517415762,
      "epoch": 1.9826280469922541,
      "grad_norm": 0.15500234067440033,
      "learning_rate": 3.391603030024228e-05,
      "loss": 0.6394,
      "mean_token_accuracy": 0.8039083182811737,
      "num_tokens": 687759095.0,
      "step": 21549
    },
    {
      "entropy": 0.6731972484849393,
      "epoch": 1.9827200533636953,
      "grad_norm": 0.1586930900812149,
      "learning_rate": 3.3912963474100655e-05,
      "loss": 0.6614,
      "mean_token_accuracy": 0.7987270876765251,
      "num_tokens": 687791113.0,
      "step": 21550
    },
    {
      "entropy": 0.6750628687441349,
      "epoch": 1.9828120597351366,
      "grad_norm": 0.164741650223732,
      "learning_rate": 3.390989664795903e-05,
      "loss": 0.6831,
      "mean_token_accuracy": 0.7940387390553951,
      "num_tokens": 687823279.0,
      "step": 21551
    },
    {
      "entropy": 0.5831459464970976,
      "epoch": 1.982904066106578,
      "grad_norm": 0.153578981757164,
      "learning_rate": 3.3906829821817405e-05,
      "loss": 0.5675,
      "mean_token_accuracy": 0.8222032375633717,
      "num_tokens": 687855426.0,
      "step": 21552
    },
    {
      "entropy": 0.4887913642451167,
      "epoch": 1.9829960724780191,
      "grad_norm": 0.14099721610546112,
      "learning_rate": 3.390376299567577e-05,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8520812913775444,
      "num_tokens": 687887646.0,
      "step": 21553
    },
    {
      "entropy": 0.646588146686554,
      "epoch": 1.9830880788494603,
      "grad_norm": 0.15902946889400482,
      "learning_rate": 3.3900696169534154e-05,
      "loss": 0.6372,
      "mean_token_accuracy": 0.8076288960874081,
      "num_tokens": 687919546.0,
      "step": 21554
    },
    {
      "entropy": 0.5862221627030522,
      "epoch": 1.9831800852209014,
      "grad_norm": 0.1487301141023636,
      "learning_rate": 3.389762934339252e-05,
      "loss": 0.5829,
      "mean_token_accuracy": 0.8257103636860847,
      "num_tokens": 687951328.0,
      "step": 21555
    },
    {
      "entropy": 0.5705960094928741,
      "epoch": 1.9832720915923427,
      "grad_norm": 0.15919798612594604,
      "learning_rate": 3.3894562517250896e-05,
      "loss": 0.566,
      "mean_token_accuracy": 0.8273290582001209,
      "num_tokens": 687983687.0,
      "step": 21556
    },
    {
      "entropy": 0.6105326265096664,
      "epoch": 1.983364097963784,
      "grad_norm": 0.16069327294826508,
      "learning_rate": 3.389149569110927e-05,
      "loss": 0.6099,
      "mean_token_accuracy": 0.8114370666444302,
      "num_tokens": 688015923.0,
      "step": 21557
    },
    {
      "entropy": 0.7029568888247013,
      "epoch": 1.9834561043352252,
      "grad_norm": 0.16609472036361694,
      "learning_rate": 3.3888428864967646e-05,
      "loss": 0.7003,
      "mean_token_accuracy": 0.7866453714668751,
      "num_tokens": 688046479.0,
      "step": 21558
    },
    {
      "entropy": 0.49721645121462643,
      "epoch": 1.9835481107066664,
      "grad_norm": 0.14500726759433746,
      "learning_rate": 3.388536203882602e-05,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8452561870217323,
      "num_tokens": 688078913.0,
      "step": 21559
    },
    {
      "entropy": 0.6732189487665892,
      "epoch": 1.9836401170781075,
      "grad_norm": 0.15699583292007446,
      "learning_rate": 3.3882295212684395e-05,
      "loss": 0.6806,
      "mean_token_accuracy": 0.7926576547324657,
      "num_tokens": 688111214.0,
      "step": 21560
    },
    {
      "entropy": 0.5305870436131954,
      "epoch": 1.9837321234495489,
      "grad_norm": 0.1476907730102539,
      "learning_rate": 3.387922838654276e-05,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8373527601361275,
      "num_tokens": 688143571.0,
      "step": 21561
    },
    {
      "entropy": 0.5076725073158741,
      "epoch": 1.9838241298209902,
      "grad_norm": 0.14461855590343475,
      "learning_rate": 3.3876161560401144e-05,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8429368920624256,
      "num_tokens": 688175691.0,
      "step": 21562
    },
    {
      "entropy": 0.6313925990834832,
      "epoch": 1.9839161361924313,
      "grad_norm": 0.15588980913162231,
      "learning_rate": 3.387309473425951e-05,
      "loss": 0.6198,
      "mean_token_accuracy": 0.8080747798085213,
      "num_tokens": 688207090.0,
      "step": 21563
    },
    {
      "entropy": 0.5483947806060314,
      "epoch": 1.9840081425638725,
      "grad_norm": 0.15127038955688477,
      "learning_rate": 3.3870027908117894e-05,
      "loss": 0.5355,
      "mean_token_accuracy": 0.8369836211204529,
      "num_tokens": 688239373.0,
      "step": 21564
    },
    {
      "entropy": 0.5462242132052779,
      "epoch": 1.9841001489353136,
      "grad_norm": 0.15016888082027435,
      "learning_rate": 3.386696108197626e-05,
      "loss": 0.541,
      "mean_token_accuracy": 0.8273280039429665,
      "num_tokens": 688270565.0,
      "step": 21565
    },
    {
      "entropy": 0.6334265256300569,
      "epoch": 1.984192155306755,
      "grad_norm": 0.16235016286373138,
      "learning_rate": 3.3863894255834636e-05,
      "loss": 0.6093,
      "mean_token_accuracy": 0.8103154115378857,
      "num_tokens": 688301993.0,
      "step": 21566
    },
    {
      "entropy": 0.4993603052571416,
      "epoch": 1.9842841616781963,
      "grad_norm": 0.1401498168706894,
      "learning_rate": 3.386082742969302e-05,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8455562330782413,
      "num_tokens": 688334415.0,
      "step": 21567
    },
    {
      "entropy": 0.5698496755212545,
      "epoch": 1.9843761680496375,
      "grad_norm": 0.1513710469007492,
      "learning_rate": 3.3857760603551385e-05,
      "loss": 0.5393,
      "mean_token_accuracy": 0.8289843127131462,
      "num_tokens": 688365952.0,
      "step": 21568
    },
    {
      "entropy": 0.6737358178943396,
      "epoch": 1.9844681744210786,
      "grad_norm": 0.1576448529958725,
      "learning_rate": 3.385469377740976e-05,
      "loss": 0.6589,
      "mean_token_accuracy": 0.8019576333463192,
      "num_tokens": 688397433.0,
      "step": 21569
    },
    {
      "entropy": 0.6578704137355089,
      "epoch": 1.9845601807925197,
      "grad_norm": 0.15464374423027039,
      "learning_rate": 3.3851626951268135e-05,
      "loss": 0.6491,
      "mean_token_accuracy": 0.7998494543135166,
      "num_tokens": 688429345.0,
      "step": 21570
    },
    {
      "entropy": 0.4870771619025618,
      "epoch": 1.984652187163961,
      "grad_norm": 0.14496774971485138,
      "learning_rate": 3.384856012512651e-05,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8520874567329884,
      "num_tokens": 688460824.0,
      "step": 21571
    },
    {
      "entropy": 0.6508522871881723,
      "epoch": 1.9847441935354024,
      "grad_norm": 0.16067087650299072,
      "learning_rate": 3.3845493298984884e-05,
      "loss": 0.6468,
      "mean_token_accuracy": 0.8000280633568764,
      "num_tokens": 688491895.0,
      "step": 21572
    },
    {
      "entropy": 0.5198565414175391,
      "epoch": 1.9848361999068436,
      "grad_norm": 0.1402553766965866,
      "learning_rate": 3.384242647284326e-05,
      "loss": 0.5016,
      "mean_token_accuracy": 0.8435371369123459,
      "num_tokens": 688524088.0,
      "step": 21573
    },
    {
      "entropy": 0.5349479578435421,
      "epoch": 1.9849282062782847,
      "grad_norm": 0.14995944499969482,
      "learning_rate": 3.3839359646701627e-05,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8361690565943718,
      "num_tokens": 688556697.0,
      "step": 21574
    },
    {
      "entropy": 0.5680330265313387,
      "epoch": 1.9850202126497258,
      "grad_norm": 0.14508600533008575,
      "learning_rate": 3.383629282056001e-05,
      "loss": 0.5553,
      "mean_token_accuracy": 0.828470017760992,
      "num_tokens": 688588915.0,
      "step": 21575
    },
    {
      "entropy": 0.608504893258214,
      "epoch": 1.9851122190211672,
      "grad_norm": 0.148283451795578,
      "learning_rate": 3.3833225994418376e-05,
      "loss": 0.5905,
      "mean_token_accuracy": 0.8180308863520622,
      "num_tokens": 688621073.0,
      "step": 21576
    },
    {
      "entropy": 0.5267011551186442,
      "epoch": 1.9852042253926085,
      "grad_norm": 0.14633049070835114,
      "learning_rate": 3.383015916827675e-05,
      "loss": 0.5063,
      "mean_token_accuracy": 0.8422659747302532,
      "num_tokens": 688653047.0,
      "step": 21577
    },
    {
      "entropy": 0.500938355922699,
      "epoch": 1.9852962317640497,
      "grad_norm": 0.13887189328670502,
      "learning_rate": 3.3827092342135125e-05,
      "loss": 0.4887,
      "mean_token_accuracy": 0.8481066897511482,
      "num_tokens": 688685536.0,
      "step": 21578
    },
    {
      "entropy": 0.5487195216119289,
      "epoch": 1.9853882381354908,
      "grad_norm": 0.14617666602134705,
      "learning_rate": 3.38240255159935e-05,
      "loss": 0.5336,
      "mean_token_accuracy": 0.8377516195178032,
      "num_tokens": 688717680.0,
      "step": 21579
    },
    {
      "entropy": 0.634046315215528,
      "epoch": 1.985480244506932,
      "grad_norm": 0.1600792407989502,
      "learning_rate": 3.3820958689851874e-05,
      "loss": 0.6309,
      "mean_token_accuracy": 0.8063063435256481,
      "num_tokens": 688749604.0,
      "step": 21580
    },
    {
      "entropy": 0.544994942843914,
      "epoch": 1.9855722508783733,
      "grad_norm": 0.15017250180244446,
      "learning_rate": 3.381789186371025e-05,
      "loss": 0.5407,
      "mean_token_accuracy": 0.8314413800835609,
      "num_tokens": 688781730.0,
      "step": 21581
    },
    {
      "entropy": 0.6328804604709148,
      "epoch": 1.9856642572498147,
      "grad_norm": 0.16394184529781342,
      "learning_rate": 3.3814825037568624e-05,
      "loss": 0.6259,
      "mean_token_accuracy": 0.8081472367048264,
      "num_tokens": 688812414.0,
      "step": 21582
    },
    {
      "entropy": 0.5978228244930506,
      "epoch": 1.9857562636212558,
      "grad_norm": 0.15060105919837952,
      "learning_rate": 3.3811758211427e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8177066445350647,
      "num_tokens": 688844445.0,
      "step": 21583
    },
    {
      "entropy": 0.604547007009387,
      "epoch": 1.985848269992697,
      "grad_norm": 0.15580631792545319,
      "learning_rate": 3.3808691385285366e-05,
      "loss": 0.5942,
      "mean_token_accuracy": 0.8168554604053497,
      "num_tokens": 688875902.0,
      "step": 21584
    },
    {
      "entropy": 0.638314398471266,
      "epoch": 1.985940276364138,
      "grad_norm": 0.15266893804073334,
      "learning_rate": 3.380562455914375e-05,
      "loss": 0.6279,
      "mean_token_accuracy": 0.805937185883522,
      "num_tokens": 688908282.0,
      "step": 21585
    },
    {
      "entropy": 0.5303623951040208,
      "epoch": 1.9860322827355794,
      "grad_norm": 0.15391331911087036,
      "learning_rate": 3.3802557733002115e-05,
      "loss": 0.5271,
      "mean_token_accuracy": 0.837949100881815,
      "num_tokens": 688941048.0,
      "step": 21586
    },
    {
      "entropy": 0.5242257257923484,
      "epoch": 1.9861242891070208,
      "grad_norm": 0.14601916074752808,
      "learning_rate": 3.379949090686049e-05,
      "loss": 0.514,
      "mean_token_accuracy": 0.8375656828284264,
      "num_tokens": 688972433.0,
      "step": 21587
    },
    {
      "entropy": 0.6621140744537115,
      "epoch": 1.986216295478462,
      "grad_norm": 0.15686437487602234,
      "learning_rate": 3.3796424080718865e-05,
      "loss": 0.6517,
      "mean_token_accuracy": 0.799594271928072,
      "num_tokens": 689004540.0,
      "step": 21588
    },
    {
      "entropy": 0.4996066512539983,
      "epoch": 1.986308301849903,
      "grad_norm": 0.14332221448421478,
      "learning_rate": 3.379335725457724e-05,
      "loss": 0.4899,
      "mean_token_accuracy": 0.847476240247488,
      "num_tokens": 689036608.0,
      "step": 21589
    },
    {
      "entropy": 0.6162477135658264,
      "epoch": 1.9864003082213442,
      "grad_norm": 0.15757353603839874,
      "learning_rate": 3.3790290428435614e-05,
      "loss": 0.6105,
      "mean_token_accuracy": 0.8142591193318367,
      "num_tokens": 689069237.0,
      "step": 21590
    },
    {
      "entropy": 0.5310793574899435,
      "epoch": 1.9864923145927855,
      "grad_norm": 0.1508644074201584,
      "learning_rate": 3.378722360229399e-05,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8380997329950333,
      "num_tokens": 689099648.0,
      "step": 21591
    },
    {
      "entropy": 0.609276638366282,
      "epoch": 1.9865843209642269,
      "grad_norm": 0.15119397640228271,
      "learning_rate": 3.3784156776152357e-05,
      "loss": 0.5957,
      "mean_token_accuracy": 0.8162188604474068,
      "num_tokens": 689132149.0,
      "step": 21592
    },
    {
      "entropy": 0.6188142346218228,
      "epoch": 1.986676327335668,
      "grad_norm": 0.15430721640586853,
      "learning_rate": 3.378108995001074e-05,
      "loss": 0.6078,
      "mean_token_accuracy": 0.8106958381831646,
      "num_tokens": 689164395.0,
      "step": 21593
    },
    {
      "entropy": 0.6301027042791247,
      "epoch": 1.9867683337071091,
      "grad_norm": 0.15406742691993713,
      "learning_rate": 3.3778023123869106e-05,
      "loss": 0.6161,
      "mean_token_accuracy": 0.8122220672667027,
      "num_tokens": 689195864.0,
      "step": 21594
    },
    {
      "entropy": 0.5935304742306471,
      "epoch": 1.9868603400785503,
      "grad_norm": 0.15276207029819489,
      "learning_rate": 3.377495629772748e-05,
      "loss": 0.5873,
      "mean_token_accuracy": 0.8199099265038967,
      "num_tokens": 689227626.0,
      "step": 21595
    },
    {
      "entropy": 0.5857824701815844,
      "epoch": 1.9869523464499916,
      "grad_norm": 0.15449494123458862,
      "learning_rate": 3.3771889471585855e-05,
      "loss": 0.5835,
      "mean_token_accuracy": 0.8191321231424809,
      "num_tokens": 689258972.0,
      "step": 21596
    },
    {
      "entropy": 0.668841416016221,
      "epoch": 1.987044352821433,
      "grad_norm": 0.16187681257724762,
      "learning_rate": 3.376882264544423e-05,
      "loss": 0.6624,
      "mean_token_accuracy": 0.7935694642364979,
      "num_tokens": 689291529.0,
      "step": 21597
    },
    {
      "entropy": 0.6723744664341211,
      "epoch": 1.9871363591928741,
      "grad_norm": 0.16065320372581482,
      "learning_rate": 3.376575581930261e-05,
      "loss": 0.6808,
      "mean_token_accuracy": 0.7935713715851307,
      "num_tokens": 689323884.0,
      "step": 21598
    },
    {
      "entropy": 0.4650551830418408,
      "epoch": 1.9872283655643153,
      "grad_norm": 0.14186760783195496,
      "learning_rate": 3.376268899316098e-05,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8575037755072117,
      "num_tokens": 689356428.0,
      "step": 21599
    },
    {
      "entropy": 0.6936111580580473,
      "epoch": 1.9873203719357564,
      "grad_norm": 0.16536974906921387,
      "learning_rate": 3.3759622167019354e-05,
      "loss": 0.6904,
      "mean_token_accuracy": 0.7907715179026127,
      "num_tokens": 689387142.0,
      "step": 21600
    },
    {
      "entropy": 0.610959367826581,
      "epoch": 1.9874123783071977,
      "grad_norm": 0.15702536702156067,
      "learning_rate": 3.375655534087773e-05,
      "loss": 0.6062,
      "mean_token_accuracy": 0.8123453594744205,
      "num_tokens": 689419199.0,
      "step": 21601
    },
    {
      "entropy": 0.5676766550168395,
      "epoch": 1.987504384678639,
      "grad_norm": 0.14914439618587494,
      "learning_rate": 3.37534885147361e-05,
      "loss": 0.5631,
      "mean_token_accuracy": 0.8258108794689178,
      "num_tokens": 689451334.0,
      "step": 21602
    },
    {
      "entropy": 0.5732491500675678,
      "epoch": 1.9875963910500802,
      "grad_norm": 0.14727170765399933,
      "learning_rate": 3.375042168859448e-05,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8206682130694389,
      "num_tokens": 689482468.0,
      "step": 21603
    },
    {
      "entropy": 0.4916043384000659,
      "epoch": 1.9876883974215214,
      "grad_norm": 0.14498493075370789,
      "learning_rate": 3.374735486245285e-05,
      "loss": 0.4944,
      "mean_token_accuracy": 0.8485073335468769,
      "num_tokens": 689515160.0,
      "step": 21604
    },
    {
      "entropy": 0.6103957556188107,
      "epoch": 1.9877804037929625,
      "grad_norm": 0.15276366472244263,
      "learning_rate": 3.374428803631122e-05,
      "loss": 0.5915,
      "mean_token_accuracy": 0.813869059085846,
      "num_tokens": 689547050.0,
      "step": 21605
    },
    {
      "entropy": 0.49915789207443595,
      "epoch": 1.9878724101644039,
      "grad_norm": 0.14544421434402466,
      "learning_rate": 3.37412212101696e-05,
      "loss": 0.4882,
      "mean_token_accuracy": 0.8504155054688454,
      "num_tokens": 689579039.0,
      "step": 21606
    },
    {
      "entropy": 0.6462370604276657,
      "epoch": 1.9879644165358452,
      "grad_norm": 0.15446244180202484,
      "learning_rate": 3.373815438402797e-05,
      "loss": 0.6478,
      "mean_token_accuracy": 0.8022974915802479,
      "num_tokens": 689611430.0,
      "step": 21607
    },
    {
      "entropy": 0.6298318151384592,
      "epoch": 1.9880564229072863,
      "grad_norm": 0.15467168390750885,
      "learning_rate": 3.3735087557886344e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.812696073204279,
      "num_tokens": 689643319.0,
      "step": 21608
    },
    {
      "entropy": 0.5856251809746027,
      "epoch": 1.9881484292787275,
      "grad_norm": 0.1513008326292038,
      "learning_rate": 3.373202073174472e-05,
      "loss": 0.5758,
      "mean_token_accuracy": 0.826822504401207,
      "num_tokens": 689675355.0,
      "step": 21609
    },
    {
      "entropy": 0.7356169885024428,
      "epoch": 1.9882404356501686,
      "grad_norm": 0.16568268835544586,
      "learning_rate": 3.3728953905603093e-05,
      "loss": 0.7184,
      "mean_token_accuracy": 0.7828909382224083,
      "num_tokens": 689706734.0,
      "step": 21610
    },
    {
      "entropy": 0.7295725680887699,
      "epoch": 1.98833244202161,
      "grad_norm": 0.16898401081562042,
      "learning_rate": 3.372588707946147e-05,
      "loss": 0.7131,
      "mean_token_accuracy": 0.7817611321806908,
      "num_tokens": 689737991.0,
      "step": 21611
    },
    {
      "entropy": 0.6310762669891119,
      "epoch": 1.9884244483930513,
      "grad_norm": 0.1538771092891693,
      "learning_rate": 3.372282025331984e-05,
      "loss": 0.6225,
      "mean_token_accuracy": 0.8100895285606384,
      "num_tokens": 689770004.0,
      "step": 21612
    },
    {
      "entropy": 0.5622490923851728,
      "epoch": 1.9885164547644925,
      "grad_norm": 0.15072380006313324,
      "learning_rate": 3.371975342717821e-05,
      "loss": 0.5436,
      "mean_token_accuracy": 0.8279052898287773,
      "num_tokens": 689801699.0,
      "step": 21613
    },
    {
      "entropy": 0.6663888450711966,
      "epoch": 1.9886084611359336,
      "grad_norm": 0.1582961231470108,
      "learning_rate": 3.371668660103659e-05,
      "loss": 0.6396,
      "mean_token_accuracy": 0.8042829521000385,
      "num_tokens": 689832509.0,
      "step": 21614
    },
    {
      "entropy": 0.5687052551656961,
      "epoch": 1.9887004675073747,
      "grad_norm": 0.14525172114372253,
      "learning_rate": 3.371361977489496e-05,
      "loss": 0.5496,
      "mean_token_accuracy": 0.8318118825554848,
      "num_tokens": 689864393.0,
      "step": 21615
    },
    {
      "entropy": 0.653488028794527,
      "epoch": 1.988792473878816,
      "grad_norm": 0.16533468663692474,
      "learning_rate": 3.3710552948753335e-05,
      "loss": 0.6523,
      "mean_token_accuracy": 0.8021936863660812,
      "num_tokens": 689895067.0,
      "step": 21616
    },
    {
      "entropy": 0.60154590010643,
      "epoch": 1.9888844802502574,
      "grad_norm": 0.15773870050907135,
      "learning_rate": 3.370748612261171e-05,
      "loss": 0.5902,
      "mean_token_accuracy": 0.820705134421587,
      "num_tokens": 689926712.0,
      "step": 21617
    },
    {
      "entropy": 0.7187583697959781,
      "epoch": 1.9889764866216986,
      "grad_norm": 0.163474440574646,
      "learning_rate": 3.3704419296470084e-05,
      "loss": 0.7171,
      "mean_token_accuracy": 0.7822774536907673,
      "num_tokens": 689959430.0,
      "step": 21618
    },
    {
      "entropy": 0.5947198267094791,
      "epoch": 1.9890684929931397,
      "grad_norm": 0.15003955364227295,
      "learning_rate": 3.370135247032846e-05,
      "loss": 0.5829,
      "mean_token_accuracy": 0.8194454871118069,
      "num_tokens": 689991527.0,
      "step": 21619
    },
    {
      "entropy": 0.555798601359129,
      "epoch": 1.9891604993645808,
      "grad_norm": 0.15008214116096497,
      "learning_rate": 3.369828564418683e-05,
      "loss": 0.5533,
      "mean_token_accuracy": 0.8303950987756252,
      "num_tokens": 690023704.0,
      "step": 21620
    },
    {
      "entropy": 0.5551923734601587,
      "epoch": 1.9892525057360222,
      "grad_norm": 0.14696241915225983,
      "learning_rate": 3.36952188180452e-05,
      "loss": 0.5455,
      "mean_token_accuracy": 0.831668246537447,
      "num_tokens": 690055911.0,
      "step": 21621
    },
    {
      "entropy": 0.4951747157610953,
      "epoch": 1.9893445121074635,
      "grad_norm": 0.1414617896080017,
      "learning_rate": 3.369215199190358e-05,
      "loss": 0.4912,
      "mean_token_accuracy": 0.8441031910479069,
      "num_tokens": 690087865.0,
      "step": 21622
    },
    {
      "entropy": 0.5667181005701423,
      "epoch": 1.9894365184789047,
      "grad_norm": 0.153811514377594,
      "learning_rate": 3.368908516576195e-05,
      "loss": 0.5601,
      "mean_token_accuracy": 0.8251244910061359,
      "num_tokens": 690120179.0,
      "step": 21623
    },
    {
      "entropy": 0.6125516919419169,
      "epoch": 1.9895285248503458,
      "grad_norm": 0.15444634854793549,
      "learning_rate": 3.368601833962033e-05,
      "loss": 0.6064,
      "mean_token_accuracy": 0.8170103281736374,
      "num_tokens": 690151201.0,
      "step": 21624
    },
    {
      "entropy": 0.6505196923390031,
      "epoch": 1.989620531221787,
      "grad_norm": 0.15553905069828033,
      "learning_rate": 3.36829515134787e-05,
      "loss": 0.6502,
      "mean_token_accuracy": 0.8020145036280155,
      "num_tokens": 690182829.0,
      "step": 21625
    },
    {
      "entropy": 0.455946359783411,
      "epoch": 1.9897125375932283,
      "grad_norm": 0.1422642469406128,
      "learning_rate": 3.3679884687337074e-05,
      "loss": 0.441,
      "mean_token_accuracy": 0.8627371490001678,
      "num_tokens": 690215081.0,
      "step": 21626
    },
    {
      "entropy": 0.6004615088459104,
      "epoch": 1.9898045439646697,
      "grad_norm": 0.14370925724506378,
      "learning_rate": 3.367681786119545e-05,
      "loss": 0.5939,
      "mean_token_accuracy": 0.8200891055166721,
      "num_tokens": 690247684.0,
      "step": 21627
    },
    {
      "entropy": 0.6758940285071731,
      "epoch": 1.9898965503361108,
      "grad_norm": 0.15671321749687195,
      "learning_rate": 3.3673751035053824e-05,
      "loss": 0.6503,
      "mean_token_accuracy": 0.7974932231009007,
      "num_tokens": 690279369.0,
      "step": 21628
    },
    {
      "entropy": 0.6112273279577494,
      "epoch": 1.989988556707552,
      "grad_norm": 0.15959270298480988,
      "learning_rate": 3.36706842089122e-05,
      "loss": 0.6037,
      "mean_token_accuracy": 0.8128701895475388,
      "num_tokens": 690311146.0,
      "step": 21629
    },
    {
      "entropy": 0.529143454041332,
      "epoch": 1.990080563078993,
      "grad_norm": 0.14176522195339203,
      "learning_rate": 3.366761738277057e-05,
      "loss": 0.5254,
      "mean_token_accuracy": 0.8416041843593121,
      "num_tokens": 690343111.0,
      "step": 21630
    },
    {
      "entropy": 0.6418647989630699,
      "epoch": 1.9901725694504344,
      "grad_norm": 0.1590053141117096,
      "learning_rate": 3.366455055662895e-05,
      "loss": 0.629,
      "mean_token_accuracy": 0.8041764162480831,
      "num_tokens": 690374808.0,
      "step": 21631
    },
    {
      "entropy": 0.5604632603935897,
      "epoch": 1.9902645758218758,
      "grad_norm": 0.14866997301578522,
      "learning_rate": 3.366148373048732e-05,
      "loss": 0.5612,
      "mean_token_accuracy": 0.8307192213833332,
      "num_tokens": 690406846.0,
      "step": 21632
    },
    {
      "entropy": 0.6512761935591698,
      "epoch": 1.990356582193317,
      "grad_norm": 0.16187219321727753,
      "learning_rate": 3.36584169043457e-05,
      "loss": 0.6384,
      "mean_token_accuracy": 0.8047578558325768,
      "num_tokens": 690437784.0,
      "step": 21633
    },
    {
      "entropy": 0.6460897629149258,
      "epoch": 1.990448588564758,
      "grad_norm": 0.15901637077331543,
      "learning_rate": 3.3655350078204065e-05,
      "loss": 0.6401,
      "mean_token_accuracy": 0.801513496786356,
      "num_tokens": 690469794.0,
      "step": 21634
    },
    {
      "entropy": 0.6408135499805212,
      "epoch": 1.9905405949361992,
      "grad_norm": 0.15640439093112946,
      "learning_rate": 3.3652283252062446e-05,
      "loss": 0.6363,
      "mean_token_accuracy": 0.8022364526987076,
      "num_tokens": 690502110.0,
      "step": 21635
    },
    {
      "entropy": 0.5774107873439789,
      "epoch": 1.9906326013076405,
      "grad_norm": 0.15458838641643524,
      "learning_rate": 3.3649216425920814e-05,
      "loss": 0.5744,
      "mean_token_accuracy": 0.8232334330677986,
      "num_tokens": 690533493.0,
      "step": 21636
    },
    {
      "entropy": 0.5104016605764627,
      "epoch": 1.9907246076790819,
      "grad_norm": 0.14743506908416748,
      "learning_rate": 3.364614959977919e-05,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8476687632501125,
      "num_tokens": 690564435.0,
      "step": 21637
    },
    {
      "entropy": 0.5313616953790188,
      "epoch": 1.990816614050523,
      "grad_norm": 0.1477167010307312,
      "learning_rate": 3.364308277363756e-05,
      "loss": 0.5294,
      "mean_token_accuracy": 0.833250317722559,
      "num_tokens": 690596354.0,
      "step": 21638
    },
    {
      "entropy": 0.6224566944874823,
      "epoch": 1.9909086204219641,
      "grad_norm": 0.15934686362743378,
      "learning_rate": 3.364001594749594e-05,
      "loss": 0.6106,
      "mean_token_accuracy": 0.815697368234396,
      "num_tokens": 690627641.0,
      "step": 21639
    },
    {
      "entropy": 0.6002692522015423,
      "epoch": 1.9910006267934053,
      "grad_norm": 0.14896894991397858,
      "learning_rate": 3.363694912135431e-05,
      "loss": 0.5919,
      "mean_token_accuracy": 0.8220861665904522,
      "num_tokens": 690659988.0,
      "step": 21640
    },
    {
      "entropy": 0.7069310676306486,
      "epoch": 1.9910926331648466,
      "grad_norm": 0.16403430700302124,
      "learning_rate": 3.363388229521269e-05,
      "loss": 0.684,
      "mean_token_accuracy": 0.7875049225986004,
      "num_tokens": 690691238.0,
      "step": 21641
    },
    {
      "entropy": 0.6021167756989598,
      "epoch": 1.991184639536288,
      "grad_norm": 0.16154751181602478,
      "learning_rate": 3.363081546907106e-05,
      "loss": 0.5996,
      "mean_token_accuracy": 0.8150535114109516,
      "num_tokens": 690722837.0,
      "step": 21642
    },
    {
      "entropy": 0.6118837334215641,
      "epoch": 1.9912766459077291,
      "grad_norm": 0.15402303636074066,
      "learning_rate": 3.3627748642929437e-05,
      "loss": 0.6056,
      "mean_token_accuracy": 0.8140365742146969,
      "num_tokens": 690755387.0,
      "step": 21643
    },
    {
      "entropy": 0.552430528216064,
      "epoch": 1.9913686522791703,
      "grad_norm": 0.14796102046966553,
      "learning_rate": 3.3624681816787804e-05,
      "loss": 0.5463,
      "mean_token_accuracy": 0.833213072270155,
      "num_tokens": 690787147.0,
      "step": 21644
    },
    {
      "entropy": 0.6604297980666161,
      "epoch": 1.9914606586506114,
      "grad_norm": 0.16006699204444885,
      "learning_rate": 3.3621614990646186e-05,
      "loss": 0.6431,
      "mean_token_accuracy": 0.8042919673025608,
      "num_tokens": 690819244.0,
      "step": 21645
    },
    {
      "entropy": 0.5961462520062923,
      "epoch": 1.9915526650220527,
      "grad_norm": 0.15272156894207,
      "learning_rate": 3.3618548164504554e-05,
      "loss": 0.5802,
      "mean_token_accuracy": 0.8189735859632492,
      "num_tokens": 690850102.0,
      "step": 21646
    },
    {
      "entropy": 0.6628443310037255,
      "epoch": 1.991644671393494,
      "grad_norm": 0.16032536327838898,
      "learning_rate": 3.361548133836293e-05,
      "loss": 0.66,
      "mean_token_accuracy": 0.799105953425169,
      "num_tokens": 690881598.0,
      "step": 21647
    },
    {
      "entropy": 0.6801207000389695,
      "epoch": 1.9917366777649352,
      "grad_norm": 0.1544838547706604,
      "learning_rate": 3.36124145122213e-05,
      "loss": 0.6561,
      "mean_token_accuracy": 0.7978668324649334,
      "num_tokens": 690913420.0,
      "step": 21648
    },
    {
      "entropy": 0.506829159334302,
      "epoch": 1.9918286841363764,
      "grad_norm": 0.1416950523853302,
      "learning_rate": 3.360934768607968e-05,
      "loss": 0.4933,
      "mean_token_accuracy": 0.8466536402702332,
      "num_tokens": 690945938.0,
      "step": 21649
    },
    {
      "entropy": 0.6540624760091305,
      "epoch": 1.9919206905078175,
      "grad_norm": 0.15121367573738098,
      "learning_rate": 3.360628085993805e-05,
      "loss": 0.6362,
      "mean_token_accuracy": 0.8007034100592136,
      "num_tokens": 690977727.0,
      "step": 21650
    },
    {
      "entropy": 0.6010622885078192,
      "epoch": 1.9920126968792589,
      "grad_norm": 0.15297171473503113,
      "learning_rate": 3.360321403379643e-05,
      "loss": 0.5977,
      "mean_token_accuracy": 0.816847700625658,
      "num_tokens": 691009489.0,
      "step": 21651
    },
    {
      "entropy": 0.6070909034460783,
      "epoch": 1.9921047032507002,
      "grad_norm": 0.15421058237552643,
      "learning_rate": 3.3600147207654795e-05,
      "loss": 0.5964,
      "mean_token_accuracy": 0.8170707523822784,
      "num_tokens": 691041079.0,
      "step": 21652
    },
    {
      "entropy": 0.518475916236639,
      "epoch": 1.9921967096221413,
      "grad_norm": 0.14305023849010468,
      "learning_rate": 3.3597080381513176e-05,
      "loss": 0.494,
      "mean_token_accuracy": 0.8490061052143574,
      "num_tokens": 691073251.0,
      "step": 21653
    },
    {
      "entropy": 0.6099980799481273,
      "epoch": 1.9922887159935825,
      "grad_norm": 0.15375573933124542,
      "learning_rate": 3.3594013555371544e-05,
      "loss": 0.6073,
      "mean_token_accuracy": 0.8162229508161545,
      "num_tokens": 691105145.0,
      "step": 21654
    },
    {
      "entropy": 0.6103209145367146,
      "epoch": 1.9923807223650236,
      "grad_norm": 0.15255668759346008,
      "learning_rate": 3.359094672922992e-05,
      "loss": 0.6076,
      "mean_token_accuracy": 0.8117245435714722,
      "num_tokens": 691136903.0,
      "step": 21655
    },
    {
      "entropy": 0.6257262546569109,
      "epoch": 1.992472728736465,
      "grad_norm": 0.15107157826423645,
      "learning_rate": 3.3587879903088293e-05,
      "loss": 0.6207,
      "mean_token_accuracy": 0.812111146748066,
      "num_tokens": 691169461.0,
      "step": 21656
    },
    {
      "entropy": 0.6095619844272733,
      "epoch": 1.9925647351079063,
      "grad_norm": 0.15555110573768616,
      "learning_rate": 3.358481307694667e-05,
      "loss": 0.5998,
      "mean_token_accuracy": 0.8134259767830372,
      "num_tokens": 691201529.0,
      "step": 21657
    },
    {
      "entropy": 0.6704163141548634,
      "epoch": 1.9926567414793475,
      "grad_norm": 0.1554785966873169,
      "learning_rate": 3.358174625080505e-05,
      "loss": 0.642,
      "mean_token_accuracy": 0.8021101467311382,
      "num_tokens": 691233055.0,
      "step": 21658
    },
    {
      "entropy": 0.5315172199625522,
      "epoch": 1.9927487478507886,
      "grad_norm": 0.15103593468666077,
      "learning_rate": 3.357867942466342e-05,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8373115547001362,
      "num_tokens": 691264828.0,
      "step": 21659
    },
    {
      "entropy": 0.6389571949839592,
      "epoch": 1.9928407542222297,
      "grad_norm": 0.15286219120025635,
      "learning_rate": 3.357561259852179e-05,
      "loss": 0.635,
      "mean_token_accuracy": 0.8023375310003757,
      "num_tokens": 691297136.0,
      "step": 21660
    },
    {
      "entropy": 0.6988184824585915,
      "epoch": 1.992932760593671,
      "grad_norm": 0.15842704474925995,
      "learning_rate": 3.357254577238017e-05,
      "loss": 0.6916,
      "mean_token_accuracy": 0.7898135557770729,
      "num_tokens": 691329629.0,
      "step": 21661
    },
    {
      "entropy": 0.538586163893342,
      "epoch": 1.9930247669651124,
      "grad_norm": 0.1477169692516327,
      "learning_rate": 3.356947894623854e-05,
      "loss": 0.5332,
      "mean_token_accuracy": 0.8342948816716671,
      "num_tokens": 691361850.0,
      "step": 21662
    },
    {
      "entropy": 0.582433239556849,
      "epoch": 1.9931167733365536,
      "grad_norm": 0.15479552745819092,
      "learning_rate": 3.3566412120096916e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8278555311262608,
      "num_tokens": 691393553.0,
      "step": 21663
    },
    {
      "entropy": 0.613061455078423,
      "epoch": 1.9932087797079947,
      "grad_norm": 0.15271686017513275,
      "learning_rate": 3.356334529395529e-05,
      "loss": 0.6174,
      "mean_token_accuracy": 0.8126940503716469,
      "num_tokens": 691425323.0,
      "step": 21664
    },
    {
      "entropy": 0.7646865155547857,
      "epoch": 1.9933007860794358,
      "grad_norm": 0.1699734479188919,
      "learning_rate": 3.356027846781366e-05,
      "loss": 0.7621,
      "mean_token_accuracy": 0.7689073383808136,
      "num_tokens": 691457634.0,
      "step": 21665
    },
    {
      "entropy": 0.6010234728455544,
      "epoch": 1.9933927924508772,
      "grad_norm": 0.15267178416252136,
      "learning_rate": 3.355721164167204e-05,
      "loss": 0.5932,
      "mean_token_accuracy": 0.8183418735861778,
      "num_tokens": 691488900.0,
      "step": 21666
    },
    {
      "entropy": 0.5423195287585258,
      "epoch": 1.9934847988223185,
      "grad_norm": 0.15010161697864532,
      "learning_rate": 3.355414481553041e-05,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8388812355697155,
      "num_tokens": 691519008.0,
      "step": 21667
    },
    {
      "entropy": 0.6909904275089502,
      "epoch": 1.9935768051937597,
      "grad_norm": 0.15879178047180176,
      "learning_rate": 3.355107798938878e-05,
      "loss": 0.689,
      "mean_token_accuracy": 0.7914682142436504,
      "num_tokens": 691551418.0,
      "step": 21668
    },
    {
      "entropy": 0.7018077643588185,
      "epoch": 1.9936688115652008,
      "grad_norm": 0.16131344437599182,
      "learning_rate": 3.354801116324716e-05,
      "loss": 0.6872,
      "mean_token_accuracy": 0.7926029525697231,
      "num_tokens": 691583532.0,
      "step": 21669
    },
    {
      "entropy": 0.5501084493007511,
      "epoch": 1.993760817936642,
      "grad_norm": 0.1525713950395584,
      "learning_rate": 3.354494433710553e-05,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8350924104452133,
      "num_tokens": 691614135.0,
      "step": 21670
    },
    {
      "entropy": 0.5729947360232472,
      "epoch": 1.9938528243080833,
      "grad_norm": 0.14396178722381592,
      "learning_rate": 3.3541877510963906e-05,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8320415914058685,
      "num_tokens": 691646579.0,
      "step": 21671
    },
    {
      "entropy": 0.6656774077564478,
      "epoch": 1.9939448306795247,
      "grad_norm": 0.16058586537837982,
      "learning_rate": 3.353881068482228e-05,
      "loss": 0.6659,
      "mean_token_accuracy": 0.7973045445978642,
      "num_tokens": 691678331.0,
      "step": 21672
    },
    {
      "entropy": 0.5353384325280786,
      "epoch": 1.9940368370509658,
      "grad_norm": 0.14365091919898987,
      "learning_rate": 3.353574385868065e-05,
      "loss": 0.5109,
      "mean_token_accuracy": 0.8405176065862179,
      "num_tokens": 691710191.0,
      "step": 21673
    },
    {
      "entropy": 0.5856874026358128,
      "epoch": 1.994128843422407,
      "grad_norm": 0.1520976573228836,
      "learning_rate": 3.353267703253903e-05,
      "loss": 0.5913,
      "mean_token_accuracy": 0.8187075108289719,
      "num_tokens": 691741819.0,
      "step": 21674
    },
    {
      "entropy": 0.6022312520071864,
      "epoch": 1.994220849793848,
      "grad_norm": 0.15967218577861786,
      "learning_rate": 3.35296102063974e-05,
      "loss": 0.6002,
      "mean_token_accuracy": 0.8150907307863235,
      "num_tokens": 691773789.0,
      "step": 21675
    },
    {
      "entropy": 0.685103690251708,
      "epoch": 1.9943128561652894,
      "grad_norm": 0.15458431839942932,
      "learning_rate": 3.352654338025577e-05,
      "loss": 0.6762,
      "mean_token_accuracy": 0.7907955460250378,
      "num_tokens": 691806437.0,
      "step": 21676
    },
    {
      "entropy": 0.6473441701382399,
      "epoch": 1.9944048625367308,
      "grad_norm": 0.1609283983707428,
      "learning_rate": 3.352347655411415e-05,
      "loss": 0.6475,
      "mean_token_accuracy": 0.8044667281210423,
      "num_tokens": 691837407.0,
      "step": 21677
    },
    {
      "entropy": 0.6831025928258896,
      "epoch": 1.994496868908172,
      "grad_norm": 0.16212397813796997,
      "learning_rate": 3.352040972797252e-05,
      "loss": 0.6917,
      "mean_token_accuracy": 0.7886796034872532,
      "num_tokens": 691868921.0,
      "step": 21678
    },
    {
      "entropy": 0.5612849676981568,
      "epoch": 1.994588875279613,
      "grad_norm": 0.14853407442569733,
      "learning_rate": 3.35173429018309e-05,
      "loss": 0.5503,
      "mean_token_accuracy": 0.8283043764531612,
      "num_tokens": 691901287.0,
      "step": 21679
    },
    {
      "entropy": 0.6119423070922494,
      "epoch": 1.9946808816510542,
      "grad_norm": 0.1558416485786438,
      "learning_rate": 3.351427607568927e-05,
      "loss": 0.5861,
      "mean_token_accuracy": 0.821190282702446,
      "num_tokens": 691932746.0,
      "step": 21680
    },
    {
      "entropy": 0.5341982804238796,
      "epoch": 1.9947728880224955,
      "grad_norm": 0.14635711908340454,
      "learning_rate": 3.351120924954764e-05,
      "loss": 0.5085,
      "mean_token_accuracy": 0.841052919626236,
      "num_tokens": 691964419.0,
      "step": 21681
    },
    {
      "entropy": 0.7124876640737057,
      "epoch": 1.9948648943939369,
      "grad_norm": 0.15694893896579742,
      "learning_rate": 3.350814242340602e-05,
      "loss": 0.7111,
      "mean_token_accuracy": 0.7791461423039436,
      "num_tokens": 691996696.0,
      "step": 21682
    },
    {
      "entropy": 0.6443781778216362,
      "epoch": 1.994956900765378,
      "grad_norm": 0.1568862646818161,
      "learning_rate": 3.350507559726439e-05,
      "loss": 0.6274,
      "mean_token_accuracy": 0.8050798028707504,
      "num_tokens": 692028014.0,
      "step": 21683
    },
    {
      "entropy": 0.5948461480438709,
      "epoch": 1.9950489071368191,
      "grad_norm": 0.1502501219511032,
      "learning_rate": 3.350200877112277e-05,
      "loss": 0.5871,
      "mean_token_accuracy": 0.8156019113957882,
      "num_tokens": 692060308.0,
      "step": 21684
    },
    {
      "entropy": 0.6629776302725077,
      "epoch": 1.9951409135082603,
      "grad_norm": 0.16202928125858307,
      "learning_rate": 3.349894194498114e-05,
      "loss": 0.6589,
      "mean_token_accuracy": 0.8007161132991314,
      "num_tokens": 692092110.0,
      "step": 21685
    },
    {
      "entropy": 0.6384144965559244,
      "epoch": 1.9952329198797016,
      "grad_norm": 0.15843601524829865,
      "learning_rate": 3.349587511883951e-05,
      "loss": 0.6395,
      "mean_token_accuracy": 0.8061299994587898,
      "num_tokens": 692123598.0,
      "step": 21686
    },
    {
      "entropy": 0.6408043019473553,
      "epoch": 1.995324926251143,
      "grad_norm": 0.15466155111789703,
      "learning_rate": 3.349280829269789e-05,
      "loss": 0.6326,
      "mean_token_accuracy": 0.8081676959991455,
      "num_tokens": 692156208.0,
      "step": 21687
    },
    {
      "entropy": 0.6575449146330357,
      "epoch": 1.9954169326225841,
      "grad_norm": 0.15777446329593658,
      "learning_rate": 3.348974146655626e-05,
      "loss": 0.6534,
      "mean_token_accuracy": 0.7960749231278896,
      "num_tokens": 692188070.0,
      "step": 21688
    },
    {
      "entropy": 0.6143891047686338,
      "epoch": 1.9955089389940253,
      "grad_norm": 0.15123817324638367,
      "learning_rate": 3.3486674640414636e-05,
      "loss": 0.5935,
      "mean_token_accuracy": 0.8175670616328716,
      "num_tokens": 692220588.0,
      "step": 21689
    },
    {
      "entropy": 0.6239649690687656,
      "epoch": 1.9956009453654664,
      "grad_norm": 0.15529756247997284,
      "learning_rate": 3.348360781427301e-05,
      "loss": 0.6179,
      "mean_token_accuracy": 0.8130023814737797,
      "num_tokens": 692252431.0,
      "step": 21690
    },
    {
      "entropy": 0.6239341823384166,
      "epoch": 1.9956929517369077,
      "grad_norm": 0.15588752925395966,
      "learning_rate": 3.3480540988131386e-05,
      "loss": 0.6096,
      "mean_token_accuracy": 0.8149559497833252,
      "num_tokens": 692284872.0,
      "step": 21691
    },
    {
      "entropy": 0.6449278518557549,
      "epoch": 1.995784958108349,
      "grad_norm": 0.16121719777584076,
      "learning_rate": 3.347747416198976e-05,
      "loss": 0.6428,
      "mean_token_accuracy": 0.8078878633677959,
      "num_tokens": 692316751.0,
      "step": 21692
    },
    {
      "entropy": 0.5629924796521664,
      "epoch": 1.9958769644797902,
      "grad_norm": 0.15052838623523712,
      "learning_rate": 3.3474407335848135e-05,
      "loss": 0.5621,
      "mean_token_accuracy": 0.827317338436842,
      "num_tokens": 692348526.0,
      "step": 21693
    },
    {
      "entropy": 0.6594132268801332,
      "epoch": 1.9959689708512314,
      "grad_norm": 0.15771430730819702,
      "learning_rate": 3.34713405097065e-05,
      "loss": 0.6589,
      "mean_token_accuracy": 0.8016374297440052,
      "num_tokens": 692380854.0,
      "step": 21694
    },
    {
      "entropy": 0.6413106489926577,
      "epoch": 1.9960609772226725,
      "grad_norm": 0.165681391954422,
      "learning_rate": 3.3468273683564884e-05,
      "loss": 0.6418,
      "mean_token_accuracy": 0.8005856163799763,
      "num_tokens": 692411905.0,
      "step": 21695
    },
    {
      "entropy": 0.6011346112936735,
      "epoch": 1.9961529835941139,
      "grad_norm": 0.15692993998527527,
      "learning_rate": 3.346520685742325e-05,
      "loss": 0.5943,
      "mean_token_accuracy": 0.8202184215188026,
      "num_tokens": 692444634.0,
      "step": 21696
    },
    {
      "entropy": 0.6842082431539893,
      "epoch": 1.9962449899655552,
      "grad_norm": 0.15861977636814117,
      "learning_rate": 3.346214003128163e-05,
      "loss": 0.6815,
      "mean_token_accuracy": 0.7932886891067028,
      "num_tokens": 692477121.0,
      "step": 21697
    },
    {
      "entropy": 0.5973452869802713,
      "epoch": 1.9963369963369964,
      "grad_norm": 0.1557682454586029,
      "learning_rate": 3.345907320514e-05,
      "loss": 0.586,
      "mean_token_accuracy": 0.818416066467762,
      "num_tokens": 692508465.0,
      "step": 21698
    },
    {
      "entropy": 0.7040898371487856,
      "epoch": 1.9964290027084375,
      "grad_norm": 0.16112545132637024,
      "learning_rate": 3.3456006378998376e-05,
      "loss": 0.6987,
      "mean_token_accuracy": 0.7884187549352646,
      "num_tokens": 692540150.0,
      "step": 21699
    },
    {
      "entropy": 0.5481988303363323,
      "epoch": 1.9965210090798786,
      "grad_norm": 0.15046179294586182,
      "learning_rate": 3.345293955285675e-05,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8333075642585754,
      "num_tokens": 692572304.0,
      "step": 21700
    },
    {
      "entropy": 0.675577949732542,
      "epoch": 1.99661301545132,
      "grad_norm": 0.15812954306602478,
      "learning_rate": 3.3449872726715125e-05,
      "loss": 0.668,
      "mean_token_accuracy": 0.798475194722414,
      "num_tokens": 692604416.0,
      "step": 21701
    },
    {
      "entropy": 0.7184382472187281,
      "epoch": 1.9967050218227613,
      "grad_norm": 0.16136065125465393,
      "learning_rate": 3.34468059005735e-05,
      "loss": 0.7175,
      "mean_token_accuracy": 0.7821364738047123,
      "num_tokens": 692636522.0,
      "step": 21702
    },
    {
      "entropy": 0.6650229543447495,
      "epoch": 1.9967970281942025,
      "grad_norm": 0.15516020357608795,
      "learning_rate": 3.3443739074431875e-05,
      "loss": 0.6511,
      "mean_token_accuracy": 0.7980161644518375,
      "num_tokens": 692668274.0,
      "step": 21703
    },
    {
      "entropy": 0.6015467159450054,
      "epoch": 1.9968890345656436,
      "grad_norm": 0.151646688580513,
      "learning_rate": 3.344067224829024e-05,
      "loss": 0.5755,
      "mean_token_accuracy": 0.8207748904824257,
      "num_tokens": 692700532.0,
      "step": 21704
    },
    {
      "entropy": 0.5884369453415275,
      "epoch": 1.9969810409370847,
      "grad_norm": 0.15198391675949097,
      "learning_rate": 3.3437605422148624e-05,
      "loss": 0.579,
      "mean_token_accuracy": 0.8187618590891361,
      "num_tokens": 692732326.0,
      "step": 21705
    },
    {
      "entropy": 0.574674766510725,
      "epoch": 1.997073047308526,
      "grad_norm": 0.14947175979614258,
      "learning_rate": 3.343453859600699e-05,
      "loss": 0.5563,
      "mean_token_accuracy": 0.822684571146965,
      "num_tokens": 692763384.0,
      "step": 21706
    },
    {
      "entropy": 0.5233308924362063,
      "epoch": 1.9971650536799674,
      "grad_norm": 0.14536452293395996,
      "learning_rate": 3.3431471769865367e-05,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8410404548048973,
      "num_tokens": 692795266.0,
      "step": 21707
    },
    {
      "entropy": 0.6585268154740334,
      "epoch": 1.9972570600514086,
      "grad_norm": 0.16209228336811066,
      "learning_rate": 3.342840494372374e-05,
      "loss": 0.6607,
      "mean_token_accuracy": 0.7940808683633804,
      "num_tokens": 692826073.0,
      "step": 21708
    },
    {
      "entropy": 0.556627961806953,
      "epoch": 1.9973490664228497,
      "grad_norm": 0.15085932612419128,
      "learning_rate": 3.3425338117582116e-05,
      "loss": 0.5495,
      "mean_token_accuracy": 0.8290518783032894,
      "num_tokens": 692857821.0,
      "step": 21709
    },
    {
      "entropy": 0.5510479933582246,
      "epoch": 1.9974410727942908,
      "grad_norm": 0.14469483494758606,
      "learning_rate": 3.342227129144049e-05,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8381562493741512,
      "num_tokens": 692889049.0,
      "step": 21710
    },
    {
      "entropy": 0.5500463456846774,
      "epoch": 1.9975330791657322,
      "grad_norm": 0.15418295562267303,
      "learning_rate": 3.3419204465298865e-05,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8362708725035191,
      "num_tokens": 692921106.0,
      "step": 21711
    },
    {
      "entropy": 0.6780353356152773,
      "epoch": 1.9976250855371736,
      "grad_norm": 0.1576218158006668,
      "learning_rate": 3.341613763915723e-05,
      "loss": 0.6628,
      "mean_token_accuracy": 0.7978283800184727,
      "num_tokens": 692952631.0,
      "step": 21712
    },
    {
      "entropy": 0.6052171946503222,
      "epoch": 1.9977170919086147,
      "grad_norm": 0.14850260317325592,
      "learning_rate": 3.3413070813015614e-05,
      "loss": 0.5815,
      "mean_token_accuracy": 0.8207036852836609,
      "num_tokens": 692985105.0,
      "step": 21713
    },
    {
      "entropy": 0.6552666830830276,
      "epoch": 1.9978090982800558,
      "grad_norm": 0.15985383093357086,
      "learning_rate": 3.341000398687398e-05,
      "loss": 0.6486,
      "mean_token_accuracy": 0.8018363416194916,
      "num_tokens": 693017271.0,
      "step": 21714
    },
    {
      "entropy": 0.622800531797111,
      "epoch": 1.997901104651497,
      "grad_norm": 0.1515122503042221,
      "learning_rate": 3.340693716073236e-05,
      "loss": 0.6146,
      "mean_token_accuracy": 0.8148924261331558,
      "num_tokens": 693049386.0,
      "step": 21715
    },
    {
      "entropy": 0.5310559184290469,
      "epoch": 1.9979931110229383,
      "grad_norm": 0.1487005054950714,
      "learning_rate": 3.340387033459073e-05,
      "loss": 0.5247,
      "mean_token_accuracy": 0.8374759443104267,
      "num_tokens": 693081470.0,
      "step": 21716
    },
    {
      "entropy": 0.6329658608883619,
      "epoch": 1.9980851173943797,
      "grad_norm": 0.1516135036945343,
      "learning_rate": 3.3400803508449106e-05,
      "loss": 0.6315,
      "mean_token_accuracy": 0.8071682564914227,
      "num_tokens": 693113888.0,
      "step": 21717
    },
    {
      "entropy": 0.7291287872940302,
      "epoch": 1.9981771237658208,
      "grad_norm": 0.16389015316963196,
      "learning_rate": 3.339773668230748e-05,
      "loss": 0.7274,
      "mean_token_accuracy": 0.7779957689344883,
      "num_tokens": 693145676.0,
      "step": 21718
    },
    {
      "entropy": 0.6804240774363279,
      "epoch": 1.998269130137262,
      "grad_norm": 0.15961642563343048,
      "learning_rate": 3.3394669856165856e-05,
      "loss": 0.6637,
      "mean_token_accuracy": 0.7980127409100533,
      "num_tokens": 693177293.0,
      "step": 21719
    },
    {
      "entropy": 0.5843066480010748,
      "epoch": 1.998361136508703,
      "grad_norm": 0.1555844098329544,
      "learning_rate": 3.339160303002423e-05,
      "loss": 0.5778,
      "mean_token_accuracy": 0.822947546839714,
      "num_tokens": 693209123.0,
      "step": 21720
    },
    {
      "entropy": 0.5528737148270011,
      "epoch": 1.9984531428801444,
      "grad_norm": 0.1438666582107544,
      "learning_rate": 3.3388536203882605e-05,
      "loss": 0.5399,
      "mean_token_accuracy": 0.8320121504366398,
      "num_tokens": 693241243.0,
      "step": 21721
    },
    {
      "entropy": 0.6134390160441399,
      "epoch": 1.9985451492515858,
      "grad_norm": 0.15036261081695557,
      "learning_rate": 3.338546937774098e-05,
      "loss": 0.5999,
      "mean_token_accuracy": 0.814147811383009,
      "num_tokens": 693272911.0,
      "step": 21722
    },
    {
      "entropy": 0.5854553310200572,
      "epoch": 1.998637155623027,
      "grad_norm": 0.15304525196552277,
      "learning_rate": 3.3382402551599354e-05,
      "loss": 0.5706,
      "mean_token_accuracy": 0.8237321116030216,
      "num_tokens": 693304589.0,
      "step": 21723
    },
    {
      "entropy": 0.574672095477581,
      "epoch": 1.998729161994468,
      "grad_norm": 0.15353548526763916,
      "learning_rate": 3.337933572545773e-05,
      "loss": 0.5676,
      "mean_token_accuracy": 0.825680635869503,
      "num_tokens": 693336055.0,
      "step": 21724
    },
    {
      "entropy": 0.5569671746343374,
      "epoch": 1.9988211683659092,
      "grad_norm": 0.14230790734291077,
      "learning_rate": 3.33762688993161e-05,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8344012387096882,
      "num_tokens": 693368498.0,
      "step": 21725
    },
    {
      "entropy": 0.6752220382913947,
      "epoch": 1.9989131747373505,
      "grad_norm": 0.159891739487648,
      "learning_rate": 3.337320207317448e-05,
      "loss": 0.6651,
      "mean_token_accuracy": 0.7971754968166351,
      "num_tokens": 693399913.0,
      "step": 21726
    },
    {
      "entropy": 0.5297769294120371,
      "epoch": 1.9990051811087919,
      "grad_norm": 0.14461997151374817,
      "learning_rate": 3.3370135247032846e-05,
      "loss": 0.5115,
      "mean_token_accuracy": 0.84383250400424,
      "num_tokens": 693432387.0,
      "step": 21727
    },
    {
      "entropy": 0.5263453936204314,
      "epoch": 1.999097187480233,
      "grad_norm": 0.14113466441631317,
      "learning_rate": 3.336706842089122e-05,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8407589644193649,
      "num_tokens": 693464219.0,
      "step": 21728
    },
    {
      "entropy": 0.5410955250263214,
      "epoch": 1.9991891938516742,
      "grad_norm": 0.15571404993534088,
      "learning_rate": 3.3364001594749595e-05,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8399414643645287,
      "num_tokens": 693496109.0,
      "step": 21729
    },
    {
      "entropy": 0.633310591802001,
      "epoch": 1.9992812002231153,
      "grad_norm": 0.15777933597564697,
      "learning_rate": 3.336093476860797e-05,
      "loss": 0.6302,
      "mean_token_accuracy": 0.8081243261694908,
      "num_tokens": 693528411.0,
      "step": 21730
    },
    {
      "entropy": 0.6724159698933363,
      "epoch": 1.9993732065945566,
      "grad_norm": 0.1548951417207718,
      "learning_rate": 3.3357867942466345e-05,
      "loss": 0.6582,
      "mean_token_accuracy": 0.7935272492468357,
      "num_tokens": 693559872.0,
      "step": 21731
    },
    {
      "entropy": 0.6179715637117624,
      "epoch": 1.999465212965998,
      "grad_norm": 0.14721715450286865,
      "learning_rate": 3.335480111632472e-05,
      "loss": 0.6042,
      "mean_token_accuracy": 0.8184398151934147,
      "num_tokens": 693592423.0,
      "step": 21732
    },
    {
      "entropy": 0.6529521839693189,
      "epoch": 1.9995572193374391,
      "grad_norm": 0.16253310441970825,
      "learning_rate": 3.335173429018309e-05,
      "loss": 0.6375,
      "mean_token_accuracy": 0.8051816150546074,
      "num_tokens": 693623537.0,
      "step": 21733
    },
    {
      "entropy": 0.6867043245583773,
      "epoch": 1.9996492257088803,
      "grad_norm": 0.17611323297023773,
      "learning_rate": 3.334866746404147e-05,
      "loss": 0.6919,
      "mean_token_accuracy": 0.7911158353090286,
      "num_tokens": 693654748.0,
      "step": 21734
    },
    {
      "entropy": 0.6086466033011675,
      "epoch": 1.9997412320803214,
      "grad_norm": 0.15780940651893616,
      "learning_rate": 3.3345600637899836e-05,
      "loss": 0.6026,
      "mean_token_accuracy": 0.8209362775087357,
      "num_tokens": 693686376.0,
      "step": 21735
    },
    {
      "entropy": 0.5476636542007327,
      "epoch": 1.9998332384517628,
      "grad_norm": 0.1485259085893631,
      "learning_rate": 3.334253381175821e-05,
      "loss": 0.5456,
      "mean_token_accuracy": 0.8318690359592438,
      "num_tokens": 693718668.0,
      "step": 21736
    },
    {
      "entropy": 0.6743322946131229,
      "epoch": 1.999925244823204,
      "grad_norm": 0.16111235320568085,
      "learning_rate": 3.3339466985616586e-05,
      "loss": 0.6602,
      "mean_token_accuracy": 0.7955457232892513,
      "num_tokens": 693750389.0,
      "step": 21737
    },
    {
      "entropy": 0.6930741415574,
      "epoch": 2.0,
      "grad_norm": 0.17795565724372864,
      "learning_rate": 3.333640015947496e-05,
      "loss": 0.6822,
      "mean_token_accuracy": 0.79568828527744,
      "num_tokens": 693775524.0,
      "step": 21738
    },
    {
      "entropy": 0.4882159475237131,
      "epoch": 2.000092006371441,
      "grad_norm": 0.12783204019069672,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.422,
      "mean_token_accuracy": 0.868078451603651,
      "num_tokens": 693807747.0,
      "step": 21739
    },
    {
      "entropy": 0.6371120791882277,
      "epoch": 2.0001840127428823,
      "grad_norm": 0.14760734140872955,
      "learning_rate": 3.333026650719171e-05,
      "loss": 0.5602,
      "mean_token_accuracy": 0.8248070366680622,
      "num_tokens": 693839569.0,
      "step": 21740
    },
    {
      "entropy": 0.6112168282270432,
      "epoch": 2.000276019114324,
      "grad_norm": 0.1428893357515335,
      "learning_rate": 3.332719968105008e-05,
      "loss": 0.5468,
      "mean_token_accuracy": 0.8306959830224514,
      "num_tokens": 693870904.0,
      "step": 21741
    },
    {
      "entropy": 0.5775752756744623,
      "epoch": 2.000368025485765,
      "grad_norm": 0.14138899743556976,
      "learning_rate": 3.332413285490846e-05,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8388686738908291,
      "num_tokens": 693903506.0,
      "step": 21742
    },
    {
      "entropy": 0.5249866996891797,
      "epoch": 2.000460031857206,
      "grad_norm": 0.13714388012886047,
      "learning_rate": 3.332106602876683e-05,
      "loss": 0.4559,
      "mean_token_accuracy": 0.853901345282793,
      "num_tokens": 693934769.0,
      "step": 21743
    },
    {
      "entropy": 0.5766644878312945,
      "epoch": 2.0005520382286472,
      "grad_norm": 0.1496879905462265,
      "learning_rate": 3.331799920262521e-05,
      "loss": 0.5331,
      "mean_token_accuracy": 0.8347258530557156,
      "num_tokens": 693966670.0,
      "step": 21744
    },
    {
      "entropy": 0.5035528279840946,
      "epoch": 2.0006440446000884,
      "grad_norm": 0.14157654345035553,
      "learning_rate": 3.3314932376483576e-05,
      "loss": 0.4683,
      "mean_token_accuracy": 0.849264495074749,
      "num_tokens": 693998489.0,
      "step": 21745
    },
    {
      "entropy": 0.5656139068305492,
      "epoch": 2.00073605097153,
      "grad_norm": 0.14867889881134033,
      "learning_rate": 3.331186555034195e-05,
      "loss": 0.5223,
      "mean_token_accuracy": 0.833044420927763,
      "num_tokens": 694030461.0,
      "step": 21746
    },
    {
      "entropy": 0.5640584453940392,
      "epoch": 2.000828057342971,
      "grad_norm": 0.1486293524503708,
      "learning_rate": 3.3308798724200325e-05,
      "loss": 0.5232,
      "mean_token_accuracy": 0.8312431052327156,
      "num_tokens": 694062001.0,
      "step": 21747
    },
    {
      "entropy": 0.5130076967179775,
      "epoch": 2.0009200637144122,
      "grad_norm": 0.1492113173007965,
      "learning_rate": 3.33057318980587e-05,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8419783972203732,
      "num_tokens": 694094226.0,
      "step": 21748
    },
    {
      "entropy": 0.5221845470368862,
      "epoch": 2.0010120700858534,
      "grad_norm": 0.1500317007303238,
      "learning_rate": 3.3302665071917075e-05,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8450457006692886,
      "num_tokens": 694126764.0,
      "step": 21749
    },
    {
      "entropy": 0.47715382650494576,
      "epoch": 2.0011040764572945,
      "grad_norm": 0.14956553280353546,
      "learning_rate": 3.329959824577545e-05,
      "loss": 0.4606,
      "mean_token_accuracy": 0.8505205437541008,
      "num_tokens": 694157669.0,
      "step": 21750
    },
    {
      "entropy": 0.6195081602782011,
      "epoch": 2.001196082828736,
      "grad_norm": 0.15864986181259155,
      "learning_rate": 3.3296531419633824e-05,
      "loss": 0.5902,
      "mean_token_accuracy": 0.8167824111878872,
      "num_tokens": 694189366.0,
      "step": 21751
    },
    {
      "entropy": 0.5580797176808119,
      "epoch": 2.001288089200177,
      "grad_norm": 0.15758010745048523,
      "learning_rate": 3.32934645934922e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8319868892431259,
      "num_tokens": 694221630.0,
      "step": 21752
    },
    {
      "entropy": 0.4994313893839717,
      "epoch": 2.0013800955716183,
      "grad_norm": 0.1562284231185913,
      "learning_rate": 3.329039776735057e-05,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8461102619767189,
      "num_tokens": 694253879.0,
      "step": 21753
    },
    {
      "entropy": 0.5027283830568194,
      "epoch": 2.0014721019430595,
      "grad_norm": 0.14998085796833038,
      "learning_rate": 3.328733094120894e-05,
      "loss": 0.4913,
      "mean_token_accuracy": 0.845512393862009,
      "num_tokens": 694286463.0,
      "step": 21754
    },
    {
      "entropy": 0.4455055904109031,
      "epoch": 2.0015641083145006,
      "grad_norm": 0.14973184466362,
      "learning_rate": 3.328426411506732e-05,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8612036556005478,
      "num_tokens": 694318367.0,
      "step": 21755
    },
    {
      "entropy": 0.5444792024791241,
      "epoch": 2.001656114685942,
      "grad_norm": 0.15793803334236145,
      "learning_rate": 3.328119728892569e-05,
      "loss": 0.5316,
      "mean_token_accuracy": 0.8307078704237938,
      "num_tokens": 694350619.0,
      "step": 21756
    },
    {
      "entropy": 0.5145568400621414,
      "epoch": 2.0017481210573833,
      "grad_norm": 0.15404245257377625,
      "learning_rate": 3.327813046278407e-05,
      "loss": 0.5027,
      "mean_token_accuracy": 0.8385154940187931,
      "num_tokens": 694382608.0,
      "step": 21757
    },
    {
      "entropy": 0.5149971209466457,
      "epoch": 2.0018401274288244,
      "grad_norm": 0.15515607595443726,
      "learning_rate": 3.327506363664244e-05,
      "loss": 0.5044,
      "mean_token_accuracy": 0.8429304957389832,
      "num_tokens": 694415049.0,
      "step": 21758
    },
    {
      "entropy": 0.5047724656760693,
      "epoch": 2.0019321338002656,
      "grad_norm": 0.1600174456834793,
      "learning_rate": 3.3271996810500814e-05,
      "loss": 0.4843,
      "mean_token_accuracy": 0.845643911510706,
      "num_tokens": 694446885.0,
      "step": 21759
    },
    {
      "entropy": 0.5291582299396396,
      "epoch": 2.0020241401717067,
      "grad_norm": 0.1612808257341385,
      "learning_rate": 3.326892998435919e-05,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8370389007031918,
      "num_tokens": 694477829.0,
      "step": 21760
    },
    {
      "entropy": 0.5076766419224441,
      "epoch": 2.0021161465431483,
      "grad_norm": 0.15354423224925995,
      "learning_rate": 3.3265863158217564e-05,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8496737480163574,
      "num_tokens": 694509588.0,
      "step": 21761
    },
    {
      "entropy": 0.5342027265578508,
      "epoch": 2.0022081529145894,
      "grad_norm": 0.15935592353343964,
      "learning_rate": 3.326279633207594e-05,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8373753502964973,
      "num_tokens": 694541477.0,
      "step": 21762
    },
    {
      "entropy": 0.4707281356677413,
      "epoch": 2.0023001592860306,
      "grad_norm": 0.15249145030975342,
      "learning_rate": 3.325972950593431e-05,
      "loss": 0.4495,
      "mean_token_accuracy": 0.8551596663892269,
      "num_tokens": 694573601.0,
      "step": 21763
    },
    {
      "entropy": 0.5525526441633701,
      "epoch": 2.0023921656574717,
      "grad_norm": 0.16468949615955353,
      "learning_rate": 3.325666267979268e-05,
      "loss": 0.5394,
      "mean_token_accuracy": 0.83228949457407,
      "num_tokens": 694605854.0,
      "step": 21764
    },
    {
      "entropy": 0.4684470994397998,
      "epoch": 2.002484172028913,
      "grad_norm": 0.1554773598909378,
      "learning_rate": 3.325359585365106e-05,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8546244725584984,
      "num_tokens": 694638183.0,
      "step": 21765
    },
    {
      "entropy": 0.43567474093288183,
      "epoch": 2.0025761784003544,
      "grad_norm": 0.14562000334262848,
      "learning_rate": 3.325052902750943e-05,
      "loss": 0.4057,
      "mean_token_accuracy": 0.8677710741758347,
      "num_tokens": 694670885.0,
      "step": 21766
    },
    {
      "entropy": 0.5511590670794249,
      "epoch": 2.0026681847717955,
      "grad_norm": 0.16647884249687195,
      "learning_rate": 3.3247462201367805e-05,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8284072130918503,
      "num_tokens": 694702742.0,
      "step": 21767
    },
    {
      "entropy": 0.6082041133195162,
      "epoch": 2.0027601911432367,
      "grad_norm": 0.1643332690000534,
      "learning_rate": 3.324439537522618e-05,
      "loss": 0.5822,
      "mean_token_accuracy": 0.8170853219926357,
      "num_tokens": 694733645.0,
      "step": 21768
    },
    {
      "entropy": 0.5696702264249325,
      "epoch": 2.002852197514678,
      "grad_norm": 0.15864819288253784,
      "learning_rate": 3.3241328549084554e-05,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8233158588409424,
      "num_tokens": 694765288.0,
      "step": 21769
    },
    {
      "entropy": 0.4718777914531529,
      "epoch": 2.002944203886119,
      "grad_norm": 0.14768224954605103,
      "learning_rate": 3.323826172294293e-05,
      "loss": 0.4426,
      "mean_token_accuracy": 0.8603029623627663,
      "num_tokens": 694797708.0,
      "step": 21770
    },
    {
      "entropy": 0.5463611036539078,
      "epoch": 2.0030362102575605,
      "grad_norm": 0.153720885515213,
      "learning_rate": 3.32351948968013e-05,
      "loss": 0.5216,
      "mean_token_accuracy": 0.8331090398132801,
      "num_tokens": 694829382.0,
      "step": 21771
    },
    {
      "entropy": 0.5386064453050494,
      "epoch": 2.0031282166290016,
      "grad_norm": 0.1521752029657364,
      "learning_rate": 3.323212807065967e-05,
      "loss": 0.5107,
      "mean_token_accuracy": 0.8366850540041924,
      "num_tokens": 694861826.0,
      "step": 21772
    },
    {
      "entropy": 0.4450367148965597,
      "epoch": 2.003220223000443,
      "grad_norm": 0.14477549493312836,
      "learning_rate": 3.322906124451805e-05,
      "loss": 0.4271,
      "mean_token_accuracy": 0.8657913357019424,
      "num_tokens": 694894136.0,
      "step": 21773
    },
    {
      "entropy": 0.5866299904882908,
      "epoch": 2.003312229371884,
      "grad_norm": 0.16663230955600739,
      "learning_rate": 3.322599441837642e-05,
      "loss": 0.554,
      "mean_token_accuracy": 0.8281954191625118,
      "num_tokens": 694925130.0,
      "step": 21774
    },
    {
      "entropy": 0.4509043574798852,
      "epoch": 2.003404235743325,
      "grad_norm": 0.1462337076663971,
      "learning_rate": 3.3222927592234795e-05,
      "loss": 0.4303,
      "mean_token_accuracy": 0.8627828732132912,
      "num_tokens": 694957379.0,
      "step": 21775
    },
    {
      "entropy": 0.5189543720334768,
      "epoch": 2.0034962421147666,
      "grad_norm": 0.15347503125667572,
      "learning_rate": 3.321986076609317e-05,
      "loss": 0.4877,
      "mean_token_accuracy": 0.8422541730105877,
      "num_tokens": 694989191.0,
      "step": 21776
    },
    {
      "entropy": 0.47695045080035925,
      "epoch": 2.0035882484862078,
      "grad_norm": 0.15114927291870117,
      "learning_rate": 3.3216793939951544e-05,
      "loss": 0.4481,
      "mean_token_accuracy": 0.8602707125246525,
      "num_tokens": 695021588.0,
      "step": 21777
    },
    {
      "entropy": 0.5272474735975266,
      "epoch": 2.003680254857649,
      "grad_norm": 0.16188840568065643,
      "learning_rate": 3.321372711380992e-05,
      "loss": 0.52,
      "mean_token_accuracy": 0.833332784473896,
      "num_tokens": 695054104.0,
      "step": 21778
    },
    {
      "entropy": 0.516860885778442,
      "epoch": 2.00377226122909,
      "grad_norm": 0.15032705664634705,
      "learning_rate": 3.3210660287668294e-05,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8426769264042377,
      "num_tokens": 695086583.0,
      "step": 21779
    },
    {
      "entropy": 0.457421887665987,
      "epoch": 2.003864267600531,
      "grad_norm": 0.1465654969215393,
      "learning_rate": 3.320759346152667e-05,
      "loss": 0.4368,
      "mean_token_accuracy": 0.8598159104585648,
      "num_tokens": 695118133.0,
      "step": 21780
    },
    {
      "entropy": 0.5303119625896215,
      "epoch": 2.0039562739719727,
      "grad_norm": 0.1599239856004715,
      "learning_rate": 3.320452663538504e-05,
      "loss": 0.507,
      "mean_token_accuracy": 0.8377167694270611,
      "num_tokens": 695149096.0,
      "step": 21781
    },
    {
      "entropy": 0.5678583942353725,
      "epoch": 2.004048280343414,
      "grad_norm": 0.1557624638080597,
      "learning_rate": 3.320145980924342e-05,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8262496441602707,
      "num_tokens": 695181521.0,
      "step": 21782
    },
    {
      "entropy": 0.43308206577785313,
      "epoch": 2.004140286714855,
      "grad_norm": 0.14638152718544006,
      "learning_rate": 3.319839298310179e-05,
      "loss": 0.4186,
      "mean_token_accuracy": 0.8679939806461334,
      "num_tokens": 695213876.0,
      "step": 21783
    },
    {
      "entropy": 0.5031978958286345,
      "epoch": 2.004232293086296,
      "grad_norm": 0.15356332063674927,
      "learning_rate": 3.319532615696017e-05,
      "loss": 0.4863,
      "mean_token_accuracy": 0.8431280702352524,
      "num_tokens": 695245749.0,
      "step": 21784
    },
    {
      "entropy": 0.5282784011214972,
      "epoch": 2.0043242994577373,
      "grad_norm": 0.15452390909194946,
      "learning_rate": 3.3192259330818535e-05,
      "loss": 0.5053,
      "mean_token_accuracy": 0.8389118611812592,
      "num_tokens": 695277836.0,
      "step": 21785
    },
    {
      "entropy": 0.522991606965661,
      "epoch": 2.004416305829179,
      "grad_norm": 0.16025280952453613,
      "learning_rate": 3.3189192504676916e-05,
      "loss": 0.5113,
      "mean_token_accuracy": 0.8380349203944206,
      "num_tokens": 695310032.0,
      "step": 21786
    },
    {
      "entropy": 0.5110913054086268,
      "epoch": 2.00450831220062,
      "grad_norm": 0.1521005779504776,
      "learning_rate": 3.3186125678535284e-05,
      "loss": 0.4982,
      "mean_token_accuracy": 0.84282111749053,
      "num_tokens": 695342103.0,
      "step": 21787
    },
    {
      "entropy": 0.5521968451794237,
      "epoch": 2.004600318572061,
      "grad_norm": 0.16146785020828247,
      "learning_rate": 3.318305885239366e-05,
      "loss": 0.527,
      "mean_token_accuracy": 0.8321195058524609,
      "num_tokens": 695373415.0,
      "step": 21788
    },
    {
      "entropy": 0.49325892701745033,
      "epoch": 2.0046923249435022,
      "grad_norm": 0.15066784620285034,
      "learning_rate": 3.3179992026252033e-05,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8478438407182693,
      "num_tokens": 695405587.0,
      "step": 21789
    },
    {
      "entropy": 0.48143270052969456,
      "epoch": 2.0047843313149434,
      "grad_norm": 0.15818528831005096,
      "learning_rate": 3.317692520011041e-05,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8494676649570465,
      "num_tokens": 695436777.0,
      "step": 21790
    },
    {
      "entropy": 0.43631448689848185,
      "epoch": 2.004876337686385,
      "grad_norm": 0.14061501622200012,
      "learning_rate": 3.317385837396878e-05,
      "loss": 0.4164,
      "mean_token_accuracy": 0.8670011013746262,
      "num_tokens": 695469204.0,
      "step": 21791
    },
    {
      "entropy": 0.5834782496094704,
      "epoch": 2.004968344057826,
      "grad_norm": 0.15840859711170197,
      "learning_rate": 3.317079154782716e-05,
      "loss": 0.5573,
      "mean_token_accuracy": 0.823523472994566,
      "num_tokens": 695501308.0,
      "step": 21792
    },
    {
      "entropy": 0.53267777338624,
      "epoch": 2.0050603504292672,
      "grad_norm": 0.15327313542366028,
      "learning_rate": 3.3167724721685525e-05,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8398448713123798,
      "num_tokens": 695533098.0,
      "step": 21793
    },
    {
      "entropy": 0.46262737456709146,
      "epoch": 2.0051523568007084,
      "grad_norm": 0.14989931881427765,
      "learning_rate": 3.316465789554391e-05,
      "loss": 0.4396,
      "mean_token_accuracy": 0.8571852296590805,
      "num_tokens": 695564887.0,
      "step": 21794
    },
    {
      "entropy": 0.6183658735826612,
      "epoch": 2.0052443631721495,
      "grad_norm": 0.16930879652500153,
      "learning_rate": 3.3161591069402275e-05,
      "loss": 0.5845,
      "mean_token_accuracy": 0.8198942616581917,
      "num_tokens": 695596722.0,
      "step": 21795
    },
    {
      "entropy": 0.5273523882497102,
      "epoch": 2.005336369543591,
      "grad_norm": 0.15260601043701172,
      "learning_rate": 3.315852424326065e-05,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8443194329738617,
      "num_tokens": 695628531.0,
      "step": 21796
    },
    {
      "entropy": 0.5898381285369396,
      "epoch": 2.005428375915032,
      "grad_norm": 0.16195225715637207,
      "learning_rate": 3.3155457417119024e-05,
      "loss": 0.5538,
      "mean_token_accuracy": 0.8274027854204178,
      "num_tokens": 695660422.0,
      "step": 21797
    },
    {
      "entropy": 0.4698146730661392,
      "epoch": 2.0055203822864733,
      "grad_norm": 0.14967232942581177,
      "learning_rate": 3.31523905909774e-05,
      "loss": 0.4398,
      "mean_token_accuracy": 0.8561186529695988,
      "num_tokens": 695692173.0,
      "step": 21798
    },
    {
      "entropy": 0.5328671354800463,
      "epoch": 2.0056123886579145,
      "grad_norm": 0.157274529337883,
      "learning_rate": 3.314932376483577e-05,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8347681015729904,
      "num_tokens": 695724748.0,
      "step": 21799
    },
    {
      "entropy": 0.5632595201022923,
      "epoch": 2.0057043950293556,
      "grad_norm": 0.16370341181755066,
      "learning_rate": 3.314625693869415e-05,
      "loss": 0.5426,
      "mean_token_accuracy": 0.8284100368618965,
      "num_tokens": 695756486.0,
      "step": 21800
    },
    {
      "entropy": 0.609288364648819,
      "epoch": 2.005796401400797,
      "grad_norm": 0.16969409584999084,
      "learning_rate": 3.314319011255252e-05,
      "loss": 0.5732,
      "mean_token_accuracy": 0.8213017731904984,
      "num_tokens": 695788205.0,
      "step": 21801
    },
    {
      "entropy": 0.5425587212666869,
      "epoch": 2.0058884077722383,
      "grad_norm": 0.15575848519802094,
      "learning_rate": 3.31401232864109e-05,
      "loss": 0.5111,
      "mean_token_accuracy": 0.8342747241258621,
      "num_tokens": 695820408.0,
      "step": 21802
    },
    {
      "entropy": 0.5530605753883719,
      "epoch": 2.0059804141436794,
      "grad_norm": 0.17091013491153717,
      "learning_rate": 3.3137056460269265e-05,
      "loss": 0.5455,
      "mean_token_accuracy": 0.8283788934350014,
      "num_tokens": 695852376.0,
      "step": 21803
    },
    {
      "entropy": 0.5839621694758534,
      "epoch": 2.0060724205151206,
      "grad_norm": 0.16118855774402618,
      "learning_rate": 3.3133989634127646e-05,
      "loss": 0.5693,
      "mean_token_accuracy": 0.8218055106699467,
      "num_tokens": 695884271.0,
      "step": 21804
    },
    {
      "entropy": 0.45125711150467396,
      "epoch": 2.0061644268865617,
      "grad_norm": 0.14685557782649994,
      "learning_rate": 3.3130922807986014e-05,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8597032204270363,
      "num_tokens": 695917039.0,
      "step": 21805
    },
    {
      "entropy": 0.522493613883853,
      "epoch": 2.0062564332580033,
      "grad_norm": 0.16277502477169037,
      "learning_rate": 3.312785598184439e-05,
      "loss": 0.5109,
      "mean_token_accuracy": 0.8404820114374161,
      "num_tokens": 695948349.0,
      "step": 21806
    },
    {
      "entropy": 0.5554415192455053,
      "epoch": 2.0063484396294444,
      "grad_norm": 0.1633075326681137,
      "learning_rate": 3.3124789155702764e-05,
      "loss": 0.5377,
      "mean_token_accuracy": 0.8289578594267368,
      "num_tokens": 695980458.0,
      "step": 21807
    },
    {
      "entropy": 0.6439156383275986,
      "epoch": 2.0064404460008856,
      "grad_norm": 0.1709793508052826,
      "learning_rate": 3.312172232956114e-05,
      "loss": 0.6321,
      "mean_token_accuracy": 0.8004859760403633,
      "num_tokens": 696012069.0,
      "step": 21808
    },
    {
      "entropy": 0.4415588262490928,
      "epoch": 2.0065324523723267,
      "grad_norm": 0.1483413428068161,
      "learning_rate": 3.311865550341951e-05,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8630135506391525,
      "num_tokens": 696044763.0,
      "step": 21809
    },
    {
      "entropy": 0.4597538774833083,
      "epoch": 2.006624458743768,
      "grad_norm": 0.15144884586334229,
      "learning_rate": 3.311558867727789e-05,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8613018468022346,
      "num_tokens": 696076852.0,
      "step": 21810
    },
    {
      "entropy": 0.5529990736395121,
      "epoch": 2.0067164651152094,
      "grad_norm": 0.1621587574481964,
      "learning_rate": 3.311252185113626e-05,
      "loss": 0.5401,
      "mean_token_accuracy": 0.8312576301395893,
      "num_tokens": 696108787.0,
      "step": 21811
    },
    {
      "entropy": 0.5830287048593163,
      "epoch": 2.0068084714866505,
      "grad_norm": 0.16427771747112274,
      "learning_rate": 3.310945502499464e-05,
      "loss": 0.567,
      "mean_token_accuracy": 0.8210530541837215,
      "num_tokens": 696140875.0,
      "step": 21812
    },
    {
      "entropy": 0.5347011210396886,
      "epoch": 2.0069004778580917,
      "grad_norm": 0.15750350058078766,
      "learning_rate": 3.310638819885301e-05,
      "loss": 0.5131,
      "mean_token_accuracy": 0.8358508758246899,
      "num_tokens": 696173003.0,
      "step": 21813
    },
    {
      "entropy": 0.501180182211101,
      "epoch": 2.006992484229533,
      "grad_norm": 0.15853048861026764,
      "learning_rate": 3.310332137271138e-05,
      "loss": 0.4999,
      "mean_token_accuracy": 0.8423241563141346,
      "num_tokens": 696205238.0,
      "step": 21814
    },
    {
      "entropy": 0.6615647468715906,
      "epoch": 2.007084490600974,
      "grad_norm": 0.167110413312912,
      "learning_rate": 3.310025454656976e-05,
      "loss": 0.6235,
      "mean_token_accuracy": 0.8057871945202351,
      "num_tokens": 696237467.0,
      "step": 21815
    },
    {
      "entropy": 0.5101162893697619,
      "epoch": 2.0071764969724155,
      "grad_norm": 0.1502256840467453,
      "learning_rate": 3.309718772042813e-05,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8463798239827156,
      "num_tokens": 696269854.0,
      "step": 21816
    },
    {
      "entropy": 0.5409860350191593,
      "epoch": 2.0072685033438566,
      "grad_norm": 0.15814168751239777,
      "learning_rate": 3.309412089428651e-05,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8382706642150879,
      "num_tokens": 696300520.0,
      "step": 21817
    },
    {
      "entropy": 0.558727715164423,
      "epoch": 2.007360509715298,
      "grad_norm": 0.15900707244873047,
      "learning_rate": 3.309105406814488e-05,
      "loss": 0.5426,
      "mean_token_accuracy": 0.8280394598841667,
      "num_tokens": 696332404.0,
      "step": 21818
    },
    {
      "entropy": 0.6189981549978256,
      "epoch": 2.007452516086739,
      "grad_norm": 0.16536059975624084,
      "learning_rate": 3.308798724200325e-05,
      "loss": 0.6003,
      "mean_token_accuracy": 0.8144607543945312,
      "num_tokens": 696365140.0,
      "step": 21819
    },
    {
      "entropy": 0.5760944746434689,
      "epoch": 2.00754452245818,
      "grad_norm": 0.16956789791584015,
      "learning_rate": 3.308492041586163e-05,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8238836228847504,
      "num_tokens": 696396052.0,
      "step": 21820
    },
    {
      "entropy": 0.6156325116753578,
      "epoch": 2.0076365288296216,
      "grad_norm": 0.16366249322891235,
      "learning_rate": 3.308185358972e-05,
      "loss": 0.5775,
      "mean_token_accuracy": 0.8176647461950779,
      "num_tokens": 696428207.0,
      "step": 21821
    },
    {
      "entropy": 0.5036723138764501,
      "epoch": 2.0077285352010628,
      "grad_norm": 0.15452256798744202,
      "learning_rate": 3.3078786763578376e-05,
      "loss": 0.4792,
      "mean_token_accuracy": 0.8465657979249954,
      "num_tokens": 696459840.0,
      "step": 21822
    },
    {
      "entropy": 0.5672266846522689,
      "epoch": 2.007820541572504,
      "grad_norm": 0.1621285080909729,
      "learning_rate": 3.307571993743675e-05,
      "loss": 0.5536,
      "mean_token_accuracy": 0.8273275569081306,
      "num_tokens": 696492035.0,
      "step": 21823
    },
    {
      "entropy": 0.5581860933452845,
      "epoch": 2.007912547943945,
      "grad_norm": 0.16540290415287018,
      "learning_rate": 3.307265311129512e-05,
      "loss": 0.538,
      "mean_token_accuracy": 0.8277989700436592,
      "num_tokens": 696523956.0,
      "step": 21824
    },
    {
      "entropy": 0.3969306359067559,
      "epoch": 2.008004554315386,
      "grad_norm": 0.14133143424987793,
      "learning_rate": 3.30695862851535e-05,
      "loss": 0.3855,
      "mean_token_accuracy": 0.8772581666707993,
      "num_tokens": 696556648.0,
      "step": 21825
    },
    {
      "entropy": 0.44613261334598064,
      "epoch": 2.0080965606868277,
      "grad_norm": 0.14981898665428162,
      "learning_rate": 3.306651945901187e-05,
      "loss": 0.4303,
      "mean_token_accuracy": 0.8610177263617516,
      "num_tokens": 696588034.0,
      "step": 21826
    },
    {
      "entropy": 0.4991959121543914,
      "epoch": 2.008188567058269,
      "grad_norm": 0.15485164523124695,
      "learning_rate": 3.306345263287024e-05,
      "loss": 0.4847,
      "mean_token_accuracy": 0.8451760523021221,
      "num_tokens": 696619674.0,
      "step": 21827
    },
    {
      "entropy": 0.47776742931455374,
      "epoch": 2.00828057342971,
      "grad_norm": 0.1539156436920166,
      "learning_rate": 3.306038580672862e-05,
      "loss": 0.455,
      "mean_token_accuracy": 0.8535896092653275,
      "num_tokens": 696651497.0,
      "step": 21828
    },
    {
      "entropy": 0.521033403230831,
      "epoch": 2.008372579801151,
      "grad_norm": 0.15609389543533325,
      "learning_rate": 3.305731898058699e-05,
      "loss": 0.5053,
      "mean_token_accuracy": 0.8403207175433636,
      "num_tokens": 696683496.0,
      "step": 21829
    },
    {
      "entropy": 0.5739969555288553,
      "epoch": 2.0084645861725923,
      "grad_norm": 0.16667190194129944,
      "learning_rate": 3.305425215444537e-05,
      "loss": 0.559,
      "mean_token_accuracy": 0.8259198069572449,
      "num_tokens": 696715067.0,
      "step": 21830
    },
    {
      "entropy": 0.45680842571891844,
      "epoch": 2.008556592544034,
      "grad_norm": 0.1429274082183838,
      "learning_rate": 3.305118532830374e-05,
      "loss": 0.4372,
      "mean_token_accuracy": 0.8592293597757816,
      "num_tokens": 696747181.0,
      "step": 21831
    },
    {
      "entropy": 0.47445189859718084,
      "epoch": 2.008648598915475,
      "grad_norm": 0.1516052782535553,
      "learning_rate": 3.304811850216211e-05,
      "loss": 0.4647,
      "mean_token_accuracy": 0.8518657498061657,
      "num_tokens": 696779266.0,
      "step": 21832
    },
    {
      "entropy": 0.4643415929749608,
      "epoch": 2.008740605286916,
      "grad_norm": 0.15627960860729218,
      "learning_rate": 3.304505167602049e-05,
      "loss": 0.4454,
      "mean_token_accuracy": 0.8578142002224922,
      "num_tokens": 696811242.0,
      "step": 21833
    },
    {
      "entropy": 0.5215511731803417,
      "epoch": 2.0088326116583572,
      "grad_norm": 0.1651768833398819,
      "learning_rate": 3.304198484987886e-05,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8345527090132236,
      "num_tokens": 696842204.0,
      "step": 21834
    },
    {
      "entropy": 0.4817959205247462,
      "epoch": 2.0089246180297984,
      "grad_norm": 0.1523837149143219,
      "learning_rate": 3.303891802373723e-05,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8491253666579723,
      "num_tokens": 696874008.0,
      "step": 21835
    },
    {
      "entropy": 0.3841371424496174,
      "epoch": 2.00901662440124,
      "grad_norm": 0.14170320332050323,
      "learning_rate": 3.303585119759561e-05,
      "loss": 0.364,
      "mean_token_accuracy": 0.8806306608021259,
      "num_tokens": 696905743.0,
      "step": 21836
    },
    {
      "entropy": 0.5127774372231215,
      "epoch": 2.009108630772681,
      "grad_norm": 0.15094974637031555,
      "learning_rate": 3.303278437145398e-05,
      "loss": 0.4916,
      "mean_token_accuracy": 0.842717107385397,
      "num_tokens": 696938155.0,
      "step": 21837
    },
    {
      "entropy": 0.49978801468387246,
      "epoch": 2.0092006371441222,
      "grad_norm": 0.15185612440109253,
      "learning_rate": 3.302971754531236e-05,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8505255095660686,
      "num_tokens": 696970077.0,
      "step": 21838
    },
    {
      "entropy": 0.5174274891614914,
      "epoch": 2.0092926435155634,
      "grad_norm": 0.15413852035999298,
      "learning_rate": 3.302665071917073e-05,
      "loss": 0.492,
      "mean_token_accuracy": 0.8443313427269459,
      "num_tokens": 697002030.0,
      "step": 21839
    },
    {
      "entropy": 0.4496588371694088,
      "epoch": 2.0093846498870045,
      "grad_norm": 0.1484578400850296,
      "learning_rate": 3.30235838930291e-05,
      "loss": 0.4179,
      "mean_token_accuracy": 0.865664217621088,
      "num_tokens": 697033957.0,
      "step": 21840
    },
    {
      "entropy": 0.5577251110225916,
      "epoch": 2.009476656258446,
      "grad_norm": 0.16113512217998505,
      "learning_rate": 3.302051706688748e-05,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8308097459375858,
      "num_tokens": 697065190.0,
      "step": 21841
    },
    {
      "entropy": 0.5748082427307963,
      "epoch": 2.009568662629887,
      "grad_norm": 0.15625382959842682,
      "learning_rate": 3.3017450240745856e-05,
      "loss": 0.549,
      "mean_token_accuracy": 0.8268111646175385,
      "num_tokens": 697097327.0,
      "step": 21842
    },
    {
      "entropy": 0.5343830568017438,
      "epoch": 2.0096606690013283,
      "grad_norm": 0.15683425962924957,
      "learning_rate": 3.301438341460423e-05,
      "loss": 0.5145,
      "mean_token_accuracy": 0.8405954167246819,
      "num_tokens": 697128282.0,
      "step": 21843
    },
    {
      "entropy": 0.5355354137718678,
      "epoch": 2.0097526753727695,
      "grad_norm": 0.15123149752616882,
      "learning_rate": 3.3011316588462605e-05,
      "loss": 0.5095,
      "mean_token_accuracy": 0.839091494679451,
      "num_tokens": 697160612.0,
      "step": 21844
    },
    {
      "entropy": 0.4467084249481559,
      "epoch": 2.0098446817442106,
      "grad_norm": 0.14296995103359222,
      "learning_rate": 3.300824976232097e-05,
      "loss": 0.4251,
      "mean_token_accuracy": 0.8653739765286446,
      "num_tokens": 697193027.0,
      "step": 21845
    },
    {
      "entropy": 0.5819707033224404,
      "epoch": 2.009936688115652,
      "grad_norm": 0.16959720849990845,
      "learning_rate": 3.3005182936179354e-05,
      "loss": 0.5657,
      "mean_token_accuracy": 0.8227125592529774,
      "num_tokens": 697225395.0,
      "step": 21846
    },
    {
      "entropy": 0.4756511216983199,
      "epoch": 2.0100286944870933,
      "grad_norm": 0.1616821140050888,
      "learning_rate": 3.300211611003772e-05,
      "loss": 0.458,
      "mean_token_accuracy": 0.8554836176335812,
      "num_tokens": 697256670.0,
      "step": 21847
    },
    {
      "entropy": 0.5331211630254984,
      "epoch": 2.0101207008585344,
      "grad_norm": 0.16080306470394135,
      "learning_rate": 3.29990492838961e-05,
      "loss": 0.5283,
      "mean_token_accuracy": 0.8341566696763039,
      "num_tokens": 697288783.0,
      "step": 21848
    },
    {
      "entropy": 0.5066978642717004,
      "epoch": 2.0102127072299756,
      "grad_norm": 0.15785419940948486,
      "learning_rate": 3.299598245775447e-05,
      "loss": 0.4873,
      "mean_token_accuracy": 0.847230575978756,
      "num_tokens": 697320626.0,
      "step": 21849
    },
    {
      "entropy": 0.5645683952607214,
      "epoch": 2.0103047136014167,
      "grad_norm": 0.16489414870738983,
      "learning_rate": 3.2992915631612846e-05,
      "loss": 0.5553,
      "mean_token_accuracy": 0.8254582472145557,
      "num_tokens": 697352839.0,
      "step": 21850
    },
    {
      "entropy": 0.4467596001923084,
      "epoch": 2.0103967199728583,
      "grad_norm": 0.14971086382865906,
      "learning_rate": 3.298984880547122e-05,
      "loss": 0.4334,
      "mean_token_accuracy": 0.8591008856892586,
      "num_tokens": 697384992.0,
      "step": 21851
    },
    {
      "entropy": 0.508360467851162,
      "epoch": 2.0104887263442994,
      "grad_norm": 0.15666940808296204,
      "learning_rate": 3.2986781979329596e-05,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8423918038606644,
      "num_tokens": 697417081.0,
      "step": 21852
    },
    {
      "entropy": 0.4496310940012336,
      "epoch": 2.0105807327157406,
      "grad_norm": 0.14733023941516876,
      "learning_rate": 3.2983715153187963e-05,
      "loss": 0.4326,
      "mean_token_accuracy": 0.8677720203995705,
      "num_tokens": 697448843.0,
      "step": 21853
    },
    {
      "entropy": 0.5523027740418911,
      "epoch": 2.0106727390871817,
      "grad_norm": 0.15867486596107483,
      "learning_rate": 3.2980648327046345e-05,
      "loss": 0.5345,
      "mean_token_accuracy": 0.8312877230346203,
      "num_tokens": 697480527.0,
      "step": 21854
    },
    {
      "entropy": 0.44779312051832676,
      "epoch": 2.010764745458623,
      "grad_norm": 0.15471968054771423,
      "learning_rate": 3.297758150090471e-05,
      "loss": 0.4241,
      "mean_token_accuracy": 0.8659369684755802,
      "num_tokens": 697511540.0,
      "step": 21855
    },
    {
      "entropy": 0.5619842456653714,
      "epoch": 2.0108567518300644,
      "grad_norm": 0.15967847406864166,
      "learning_rate": 3.297451467476309e-05,
      "loss": 0.5413,
      "mean_token_accuracy": 0.8292883932590485,
      "num_tokens": 697544043.0,
      "step": 21856
    },
    {
      "entropy": 0.5746847689151764,
      "epoch": 2.0109487582015055,
      "grad_norm": 0.1639101803302765,
      "learning_rate": 3.297144784862146e-05,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8232330307364464,
      "num_tokens": 697576004.0,
      "step": 21857
    },
    {
      "entropy": 0.5225209170021117,
      "epoch": 2.0110407645729467,
      "grad_norm": 0.15843093395233154,
      "learning_rate": 3.296838102247984e-05,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8428549580276012,
      "num_tokens": 697607161.0,
      "step": 21858
    },
    {
      "entropy": 0.4496118454262614,
      "epoch": 2.011132770944388,
      "grad_norm": 0.14438413083553314,
      "learning_rate": 3.296531419633821e-05,
      "loss": 0.4205,
      "mean_token_accuracy": 0.865742851048708,
      "num_tokens": 697639894.0,
      "step": 21859
    },
    {
      "entropy": 0.4742437992244959,
      "epoch": 2.011224777315829,
      "grad_norm": 0.15444812178611755,
      "learning_rate": 3.2962247370196586e-05,
      "loss": 0.464,
      "mean_token_accuracy": 0.8524683453142643,
      "num_tokens": 697672362.0,
      "step": 21860
    },
    {
      "entropy": 0.41852539870887995,
      "epoch": 2.0113167836872705,
      "grad_norm": 0.1429002583026886,
      "learning_rate": 3.295918054405496e-05,
      "loss": 0.3982,
      "mean_token_accuracy": 0.8735326677560806,
      "num_tokens": 697704070.0,
      "step": 21861
    },
    {
      "entropy": 0.5655536763370037,
      "epoch": 2.0114087900587116,
      "grad_norm": 0.16264119744300842,
      "learning_rate": 3.2956113717913335e-05,
      "loss": 0.5419,
      "mean_token_accuracy": 0.8290996737778187,
      "num_tokens": 697736319.0,
      "step": 21862
    },
    {
      "entropy": 0.5511676631867886,
      "epoch": 2.011500796430153,
      "grad_norm": 0.1616804301738739,
      "learning_rate": 3.29530468917717e-05,
      "loss": 0.5123,
      "mean_token_accuracy": 0.8347729444503784,
      "num_tokens": 697767397.0,
      "step": 21863
    },
    {
      "entropy": 0.4978199221659452,
      "epoch": 2.011592802801594,
      "grad_norm": 0.15868128836154938,
      "learning_rate": 3.2949980065630085e-05,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8491143360733986,
      "num_tokens": 697799409.0,
      "step": 21864
    },
    {
      "entropy": 0.4300439106300473,
      "epoch": 2.011684809173035,
      "grad_norm": 0.14910291135311127,
      "learning_rate": 3.294691323948845e-05,
      "loss": 0.4165,
      "mean_token_accuracy": 0.8659100048244,
      "num_tokens": 697832076.0,
      "step": 21865
    },
    {
      "entropy": 0.4631103063002229,
      "epoch": 2.0117768155444766,
      "grad_norm": 0.15466144680976868,
      "learning_rate": 3.294384641334683e-05,
      "loss": 0.456,
      "mean_token_accuracy": 0.8492441214621067,
      "num_tokens": 697864481.0,
      "step": 21866
    },
    {
      "entropy": 0.47835151804611087,
      "epoch": 2.0118688219159178,
      "grad_norm": 0.1487058401107788,
      "learning_rate": 3.29407795872052e-05,
      "loss": 0.4621,
      "mean_token_accuracy": 0.8521464318037033,
      "num_tokens": 697896682.0,
      "step": 21867
    },
    {
      "entropy": 0.5509091075509787,
      "epoch": 2.011960828287359,
      "grad_norm": 0.16296228766441345,
      "learning_rate": 3.2937712761063576e-05,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8314329981803894,
      "num_tokens": 697928310.0,
      "step": 21868
    },
    {
      "entropy": 0.39969071513041854,
      "epoch": 2.0120528346588,
      "grad_norm": 0.14784666895866394,
      "learning_rate": 3.293464593492195e-05,
      "loss": 0.3618,
      "mean_token_accuracy": 0.8869266733527184,
      "num_tokens": 697960268.0,
      "step": 21869
    },
    {
      "entropy": 0.4655665857717395,
      "epoch": 2.012144841030241,
      "grad_norm": 0.14980705082416534,
      "learning_rate": 3.2931579108780326e-05,
      "loss": 0.4473,
      "mean_token_accuracy": 0.856971975415945,
      "num_tokens": 697992817.0,
      "step": 21870
    },
    {
      "entropy": 0.5658940281718969,
      "epoch": 2.0122368474016827,
      "grad_norm": 0.16405704617500305,
      "learning_rate": 3.2928512282638694e-05,
      "loss": 0.548,
      "mean_token_accuracy": 0.8210257589817047,
      "num_tokens": 698024347.0,
      "step": 21871
    },
    {
      "entropy": 0.4384296366479248,
      "epoch": 2.012328853773124,
      "grad_norm": 0.15036937594413757,
      "learning_rate": 3.2925445456497075e-05,
      "loss": 0.4226,
      "mean_token_accuracy": 0.8655794523656368,
      "num_tokens": 698056250.0,
      "step": 21872
    },
    {
      "entropy": 0.44609422609210014,
      "epoch": 2.012420860144565,
      "grad_norm": 0.15157048404216766,
      "learning_rate": 3.292237863035545e-05,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8605640046298504,
      "num_tokens": 698087580.0,
      "step": 21873
    },
    {
      "entropy": 0.4891770612448454,
      "epoch": 2.012512866516006,
      "grad_norm": 0.1546977162361145,
      "learning_rate": 3.291931180421382e-05,
      "loss": 0.4705,
      "mean_token_accuracy": 0.848074309527874,
      "num_tokens": 698119342.0,
      "step": 21874
    },
    {
      "entropy": 0.5329309720546007,
      "epoch": 2.0126048728874473,
      "grad_norm": 0.1617574691772461,
      "learning_rate": 3.29162449780722e-05,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8355042263865471,
      "num_tokens": 698151590.0,
      "step": 21875
    },
    {
      "entropy": 0.5580606963485479,
      "epoch": 2.012696879258889,
      "grad_norm": 0.16603437066078186,
      "learning_rate": 3.291317815193057e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.82935905829072,
      "num_tokens": 698183934.0,
      "step": 21876
    },
    {
      "entropy": 0.458352071698755,
      "epoch": 2.01278888563033,
      "grad_norm": 0.1508628875017166,
      "learning_rate": 3.291011132578895e-05,
      "loss": 0.4286,
      "mean_token_accuracy": 0.8612022511661053,
      "num_tokens": 698215861.0,
      "step": 21877
    },
    {
      "entropy": 0.46661644661799073,
      "epoch": 2.012880892001771,
      "grad_norm": 0.1540936976671219,
      "learning_rate": 3.2907044499647316e-05,
      "loss": 0.4536,
      "mean_token_accuracy": 0.8567921482026577,
      "num_tokens": 698247840.0,
      "step": 21878
    },
    {
      "entropy": 0.5903770793229342,
      "epoch": 2.0129728983732122,
      "grad_norm": 0.16370974481105804,
      "learning_rate": 3.290397767350569e-05,
      "loss": 0.5655,
      "mean_token_accuracy": 0.8184118643403053,
      "num_tokens": 698280280.0,
      "step": 21879
    },
    {
      "entropy": 0.5370842578122392,
      "epoch": 2.0130649047446534,
      "grad_norm": 0.15803320705890656,
      "learning_rate": 3.2900910847364065e-05,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8332810550928116,
      "num_tokens": 698312042.0,
      "step": 21880
    },
    {
      "entropy": 0.5643099872395396,
      "epoch": 2.013156911116095,
      "grad_norm": 0.16372957825660706,
      "learning_rate": 3.289784402122244e-05,
      "loss": 0.5353,
      "mean_token_accuracy": 0.8313204124569893,
      "num_tokens": 698342711.0,
      "step": 21881
    },
    {
      "entropy": 0.5688073094934225,
      "epoch": 2.013248917487536,
      "grad_norm": 0.16891604661941528,
      "learning_rate": 3.2894777195080815e-05,
      "loss": 0.5433,
      "mean_token_accuracy": 0.8263989090919495,
      "num_tokens": 698374031.0,
      "step": 21882
    },
    {
      "entropy": 0.4705533110536635,
      "epoch": 2.0133409238589772,
      "grad_norm": 0.14948664605617523,
      "learning_rate": 3.289171036893919e-05,
      "loss": 0.4535,
      "mean_token_accuracy": 0.854842122644186,
      "num_tokens": 698406470.0,
      "step": 21883
    },
    {
      "entropy": 0.44470289978198707,
      "epoch": 2.0134329302304184,
      "grad_norm": 0.14981485903263092,
      "learning_rate": 3.288864354279756e-05,
      "loss": 0.4264,
      "mean_token_accuracy": 0.8629787042737007,
      "num_tokens": 698438474.0,
      "step": 21884
    },
    {
      "entropy": 0.5309566715732217,
      "epoch": 2.0135249366018595,
      "grad_norm": 0.1631772220134735,
      "learning_rate": 3.288557671665594e-05,
      "loss": 0.5194,
      "mean_token_accuracy": 0.8357312381267548,
      "num_tokens": 698469537.0,
      "step": 21885
    },
    {
      "entropy": 0.5229780450463295,
      "epoch": 2.013616942973301,
      "grad_norm": 0.15328028798103333,
      "learning_rate": 3.2882509890514306e-05,
      "loss": 0.5006,
      "mean_token_accuracy": 0.8417806476354599,
      "num_tokens": 698502305.0,
      "step": 21886
    },
    {
      "entropy": 0.5412062909454107,
      "epoch": 2.013708949344742,
      "grad_norm": 0.16136807203292847,
      "learning_rate": 3.287944306437268e-05,
      "loss": 0.5236,
      "mean_token_accuracy": 0.8331842236220837,
      "num_tokens": 698534232.0,
      "step": 21887
    },
    {
      "entropy": 0.5374256148934364,
      "epoch": 2.0138009557161833,
      "grad_norm": 0.16007864475250244,
      "learning_rate": 3.2876376238231056e-05,
      "loss": 0.5204,
      "mean_token_accuracy": 0.8335499726235867,
      "num_tokens": 698565747.0,
      "step": 21888
    },
    {
      "entropy": 0.5185939874500036,
      "epoch": 2.0138929620876245,
      "grad_norm": 0.15811018645763397,
      "learning_rate": 3.287330941208943e-05,
      "loss": 0.5235,
      "mean_token_accuracy": 0.8331909030675888,
      "num_tokens": 698598414.0,
      "step": 21889
    },
    {
      "entropy": 0.5111702755093575,
      "epoch": 2.0139849684590656,
      "grad_norm": 0.15784823894500732,
      "learning_rate": 3.2870242585947805e-05,
      "loss": 0.4845,
      "mean_token_accuracy": 0.8475894220173359,
      "num_tokens": 698629131.0,
      "step": 21890
    },
    {
      "entropy": 0.5949757173657417,
      "epoch": 2.014076974830507,
      "grad_norm": 0.16588887572288513,
      "learning_rate": 3.286717575980618e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8165398575365543,
      "num_tokens": 698661498.0,
      "step": 21891
    },
    {
      "entropy": 0.6052792016416788,
      "epoch": 2.0141689812019483,
      "grad_norm": 0.17085577547550201,
      "learning_rate": 3.286410893366455e-05,
      "loss": 0.5861,
      "mean_token_accuracy": 0.8135531023144722,
      "num_tokens": 698692773.0,
      "step": 21892
    },
    {
      "entropy": 0.4608775991946459,
      "epoch": 2.0142609875733894,
      "grad_norm": 0.14848127961158752,
      "learning_rate": 3.286104210752293e-05,
      "loss": 0.4494,
      "mean_token_accuracy": 0.8564968481659889,
      "num_tokens": 698724769.0,
      "step": 21893
    },
    {
      "entropy": 0.5008926764130592,
      "epoch": 2.0143529939448306,
      "grad_norm": 0.16037236154079437,
      "learning_rate": 3.28579752813813e-05,
      "loss": 0.4802,
      "mean_token_accuracy": 0.8465452566742897,
      "num_tokens": 698756742.0,
      "step": 21894
    },
    {
      "entropy": 0.45501542976126075,
      "epoch": 2.0144450003162717,
      "grad_norm": 0.14542827010154724,
      "learning_rate": 3.285490845523967e-05,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8581891618669033,
      "num_tokens": 698789418.0,
      "step": 21895
    },
    {
      "entropy": 0.5895303599536419,
      "epoch": 2.0145370066877133,
      "grad_norm": 0.16348877549171448,
      "learning_rate": 3.2851841629098046e-05,
      "loss": 0.5697,
      "mean_token_accuracy": 0.8196900859475136,
      "num_tokens": 698821606.0,
      "step": 21896
    },
    {
      "entropy": 0.5426561273634434,
      "epoch": 2.0146290130591544,
      "grad_norm": 0.1575843244791031,
      "learning_rate": 3.284877480295642e-05,
      "loss": 0.5247,
      "mean_token_accuracy": 0.8335892520844936,
      "num_tokens": 698853517.0,
      "step": 21897
    },
    {
      "entropy": 0.4924143732059747,
      "epoch": 2.0147210194305956,
      "grad_norm": 0.16013482213020325,
      "learning_rate": 3.2845707976814795e-05,
      "loss": 0.479,
      "mean_token_accuracy": 0.8457790650427341,
      "num_tokens": 698885593.0,
      "step": 21898
    },
    {
      "entropy": 0.5313021428883076,
      "epoch": 2.0148130258020367,
      "grad_norm": 0.15856529772281647,
      "learning_rate": 3.284264115067317e-05,
      "loss": 0.5179,
      "mean_token_accuracy": 0.8359119221568108,
      "num_tokens": 698917758.0,
      "step": 21899
    },
    {
      "entropy": 0.5434017777442932,
      "epoch": 2.014905032173478,
      "grad_norm": 0.16266342997550964,
      "learning_rate": 3.283957432453154e-05,
      "loss": 0.5256,
      "mean_token_accuracy": 0.8333657793700695,
      "num_tokens": 698950004.0,
      "step": 21900
    },
    {
      "entropy": 0.48279383033514023,
      "epoch": 2.0149970385449194,
      "grad_norm": 0.1554631143808365,
      "learning_rate": 3.283650749838992e-05,
      "loss": 0.4602,
      "mean_token_accuracy": 0.8545421585440636,
      "num_tokens": 698981692.0,
      "step": 21901
    },
    {
      "entropy": 0.54949417617172,
      "epoch": 2.0150890449163605,
      "grad_norm": 0.1586754024028778,
      "learning_rate": 3.2833440672248294e-05,
      "loss": 0.5324,
      "mean_token_accuracy": 0.8333414867520332,
      "num_tokens": 699014074.0,
      "step": 21902
    },
    {
      "entropy": 0.5474624177441001,
      "epoch": 2.0151810512878017,
      "grad_norm": 0.16282311081886292,
      "learning_rate": 3.283037384610667e-05,
      "loss": 0.5263,
      "mean_token_accuracy": 0.8367469944059849,
      "num_tokens": 699045444.0,
      "step": 21903
    },
    {
      "entropy": 0.4985347967594862,
      "epoch": 2.015273057659243,
      "grad_norm": 0.1561776101589203,
      "learning_rate": 3.282730701996504e-05,
      "loss": 0.4721,
      "mean_token_accuracy": 0.8484913259744644,
      "num_tokens": 699076390.0,
      "step": 21904
    },
    {
      "entropy": 0.5172533951699734,
      "epoch": 2.015365064030684,
      "grad_norm": 0.16220824420452118,
      "learning_rate": 3.282424019382341e-05,
      "loss": 0.4969,
      "mean_token_accuracy": 0.8405642621219158,
      "num_tokens": 699106962.0,
      "step": 21905
    },
    {
      "entropy": 0.5631919931620359,
      "epoch": 2.0154570704021255,
      "grad_norm": 0.16093340516090393,
      "learning_rate": 3.282117336768179e-05,
      "loss": 0.5333,
      "mean_token_accuracy": 0.8289896883070469,
      "num_tokens": 699138486.0,
      "step": 21906
    },
    {
      "entropy": 0.5220021056011319,
      "epoch": 2.0155490767735667,
      "grad_norm": 0.16057080030441284,
      "learning_rate": 3.281810654154016e-05,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8390703462064266,
      "num_tokens": 699170660.0,
      "step": 21907
    },
    {
      "entropy": 0.6116512133739889,
      "epoch": 2.015641083145008,
      "grad_norm": 0.16695289313793182,
      "learning_rate": 3.2815039715398535e-05,
      "loss": 0.6003,
      "mean_token_accuracy": 0.8122610151767731,
      "num_tokens": 699202935.0,
      "step": 21908
    },
    {
      "entropy": 0.44648472033441067,
      "epoch": 2.015733089516449,
      "grad_norm": 0.1521284580230713,
      "learning_rate": 3.281197288925691e-05,
      "loss": 0.4331,
      "mean_token_accuracy": 0.8623563162982464,
      "num_tokens": 699235008.0,
      "step": 21909
    },
    {
      "entropy": 0.4794784407131374,
      "epoch": 2.01582509588789,
      "grad_norm": 0.15546701848506927,
      "learning_rate": 3.2808906063115284e-05,
      "loss": 0.4606,
      "mean_token_accuracy": 0.8529544621706009,
      "num_tokens": 699267095.0,
      "step": 21910
    },
    {
      "entropy": 0.5074496567249298,
      "epoch": 2.0159171022593316,
      "grad_norm": 0.15727479755878448,
      "learning_rate": 3.280583923697366e-05,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8419616669416428,
      "num_tokens": 699299316.0,
      "step": 21911
    },
    {
      "entropy": 0.4831321663223207,
      "epoch": 2.0160091086307728,
      "grad_norm": 0.15280267596244812,
      "learning_rate": 3.2802772410832034e-05,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8493027947843075,
      "num_tokens": 699331561.0,
      "step": 21912
    },
    {
      "entropy": 0.5274554211646318,
      "epoch": 2.016101115002214,
      "grad_norm": 0.16199690103530884,
      "learning_rate": 3.27997055846904e-05,
      "loss": 0.516,
      "mean_token_accuracy": 0.8376963250339031,
      "num_tokens": 699364068.0,
      "step": 21913
    },
    {
      "entropy": 0.4901702804490924,
      "epoch": 2.016193121373655,
      "grad_norm": 0.15402649343013763,
      "learning_rate": 3.279663875854878e-05,
      "loss": 0.4703,
      "mean_token_accuracy": 0.851044125854969,
      "num_tokens": 699395623.0,
      "step": 21914
    },
    {
      "entropy": 0.5160581730306149,
      "epoch": 2.016285127745096,
      "grad_norm": 0.15821383893489838,
      "learning_rate": 3.279357193240715e-05,
      "loss": 0.4977,
      "mean_token_accuracy": 0.8436126038432121,
      "num_tokens": 699428391.0,
      "step": 21915
    },
    {
      "entropy": 0.4971650307998061,
      "epoch": 2.0163771341165377,
      "grad_norm": 0.1521729826927185,
      "learning_rate": 3.2790505106265526e-05,
      "loss": 0.4719,
      "mean_token_accuracy": 0.8475264422595501,
      "num_tokens": 699460695.0,
      "step": 21916
    },
    {
      "entropy": 0.5761311212554574,
      "epoch": 2.016469140487979,
      "grad_norm": 0.15996456146240234,
      "learning_rate": 3.27874382801239e-05,
      "loss": 0.5571,
      "mean_token_accuracy": 0.8282892853021622,
      "num_tokens": 699493190.0,
      "step": 21917
    },
    {
      "entropy": 0.5708549693226814,
      "epoch": 2.01656114685942,
      "grad_norm": 0.1645536720752716,
      "learning_rate": 3.2784371453982275e-05,
      "loss": 0.568,
      "mean_token_accuracy": 0.8188611604273319,
      "num_tokens": 699524858.0,
      "step": 21918
    },
    {
      "entropy": 0.5495565999299288,
      "epoch": 2.016653153230861,
      "grad_norm": 0.1602034717798233,
      "learning_rate": 3.278130462784065e-05,
      "loss": 0.533,
      "mean_token_accuracy": 0.8325967825949192,
      "num_tokens": 699557233.0,
      "step": 21919
    },
    {
      "entropy": 0.6600127015262842,
      "epoch": 2.0167451596023023,
      "grad_norm": 0.17491848766803741,
      "learning_rate": 3.2778237801699024e-05,
      "loss": 0.6548,
      "mean_token_accuracy": 0.7970418408513069,
      "num_tokens": 699588898.0,
      "step": 21920
    },
    {
      "entropy": 0.5790367629379034,
      "epoch": 2.016837165973744,
      "grad_norm": 0.16461654007434845,
      "learning_rate": 3.27751709755574e-05,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8215351887047291,
      "num_tokens": 699621046.0,
      "step": 21921
    },
    {
      "entropy": 0.5397985456511378,
      "epoch": 2.016929172345185,
      "grad_norm": 0.15843547880649567,
      "learning_rate": 3.2772104149415773e-05,
      "loss": 0.519,
      "mean_token_accuracy": 0.8372411988675594,
      "num_tokens": 699653443.0,
      "step": 21922
    },
    {
      "entropy": 0.45133344642817974,
      "epoch": 2.017021178716626,
      "grad_norm": 0.15069296956062317,
      "learning_rate": 3.276903732327414e-05,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8584954366087914,
      "num_tokens": 699685967.0,
      "step": 21923
    },
    {
      "entropy": 0.5683680716902018,
      "epoch": 2.0171131850880673,
      "grad_norm": 0.16981211304664612,
      "learning_rate": 3.276597049713252e-05,
      "loss": 0.5599,
      "mean_token_accuracy": 0.8189392127096653,
      "num_tokens": 699717712.0,
      "step": 21924
    },
    {
      "entropy": 0.4816339900717139,
      "epoch": 2.0172051914595084,
      "grad_norm": 0.1547597050666809,
      "learning_rate": 3.276290367099089e-05,
      "loss": 0.4651,
      "mean_token_accuracy": 0.8522070609033108,
      "num_tokens": 699749946.0,
      "step": 21925
    },
    {
      "entropy": 0.5196881215088069,
      "epoch": 2.01729719783095,
      "grad_norm": 0.157619446516037,
      "learning_rate": 3.2759836844849265e-05,
      "loss": 0.5137,
      "mean_token_accuracy": 0.8385352976620197,
      "num_tokens": 699782055.0,
      "step": 21926
    },
    {
      "entropy": 0.5369373559951782,
      "epoch": 2.017389204202391,
      "grad_norm": 0.15796974301338196,
      "learning_rate": 3.275677001870764e-05,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8333349004387856,
      "num_tokens": 699814545.0,
      "step": 21927
    },
    {
      "entropy": 0.486878065392375,
      "epoch": 2.0174812105738322,
      "grad_norm": 0.15405823290348053,
      "learning_rate": 3.2753703192566015e-05,
      "loss": 0.4621,
      "mean_token_accuracy": 0.852936502546072,
      "num_tokens": 699846682.0,
      "step": 21928
    },
    {
      "entropy": 0.5687541468068957,
      "epoch": 2.0175732169452734,
      "grad_norm": 0.1649332344532013,
      "learning_rate": 3.275063636642439e-05,
      "loss": 0.5581,
      "mean_token_accuracy": 0.8221318982541561,
      "num_tokens": 699878921.0,
      "step": 21929
    },
    {
      "entropy": 0.46103146811947227,
      "epoch": 2.0176652233167145,
      "grad_norm": 0.14837875962257385,
      "learning_rate": 3.2747569540282764e-05,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8578555323183537,
      "num_tokens": 699910391.0,
      "step": 21930
    },
    {
      "entropy": 0.4897504411637783,
      "epoch": 2.017757229688156,
      "grad_norm": 0.15282028913497925,
      "learning_rate": 3.274450271414113e-05,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8569714687764645,
      "num_tokens": 699942804.0,
      "step": 21931
    },
    {
      "entropy": 0.5843404848128557,
      "epoch": 2.017849236059597,
      "grad_norm": 0.16548296809196472,
      "learning_rate": 3.274143588799951e-05,
      "loss": 0.5527,
      "mean_token_accuracy": 0.824277188628912,
      "num_tokens": 699974874.0,
      "step": 21932
    },
    {
      "entropy": 0.5040371062932536,
      "epoch": 2.0179412424310383,
      "grad_norm": 0.1572290062904358,
      "learning_rate": 3.273836906185789e-05,
      "loss": 0.4815,
      "mean_token_accuracy": 0.8436537832021713,
      "num_tokens": 700006387.0,
      "step": 21933
    },
    {
      "entropy": 0.4619030279573053,
      "epoch": 2.0180332488024795,
      "grad_norm": 0.15418261289596558,
      "learning_rate": 3.2735302235716256e-05,
      "loss": 0.434,
      "mean_token_accuracy": 0.8614646196365356,
      "num_tokens": 700038140.0,
      "step": 21934
    },
    {
      "entropy": 0.44249306712299585,
      "epoch": 2.0181252551739206,
      "grad_norm": 0.14940626919269562,
      "learning_rate": 3.273223540957464e-05,
      "loss": 0.4169,
      "mean_token_accuracy": 0.8684713877737522,
      "num_tokens": 700069602.0,
      "step": 21935
    },
    {
      "entropy": 0.4737016437575221,
      "epoch": 2.018217261545362,
      "grad_norm": 0.15539097785949707,
      "learning_rate": 3.2729168583433005e-05,
      "loss": 0.4548,
      "mean_token_accuracy": 0.854354664683342,
      "num_tokens": 700101462.0,
      "step": 21936
    },
    {
      "entropy": 0.49774451553821564,
      "epoch": 2.0183092679168033,
      "grad_norm": 0.157947838306427,
      "learning_rate": 3.2726101757291386e-05,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8434373959898949,
      "num_tokens": 700133565.0,
      "step": 21937
    },
    {
      "entropy": 0.5814980398863554,
      "epoch": 2.0184012742882445,
      "grad_norm": 0.16309361159801483,
      "learning_rate": 3.2723034931149754e-05,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8255112357437611,
      "num_tokens": 700164989.0,
      "step": 21938
    },
    {
      "entropy": 0.497169591486454,
      "epoch": 2.0184932806596856,
      "grad_norm": 0.16414868831634521,
      "learning_rate": 3.271996810500813e-05,
      "loss": 0.4933,
      "mean_token_accuracy": 0.8409882970154285,
      "num_tokens": 700196785.0,
      "step": 21939
    },
    {
      "entropy": 0.5560973193496466,
      "epoch": 2.0185852870311267,
      "grad_norm": 0.16106250882148743,
      "learning_rate": 3.2716901278866504e-05,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8265803307294846,
      "num_tokens": 700229030.0,
      "step": 21940
    },
    {
      "entropy": 0.557966710999608,
      "epoch": 2.0186772934025683,
      "grad_norm": 0.16201846301555634,
      "learning_rate": 3.271383445272488e-05,
      "loss": 0.5421,
      "mean_token_accuracy": 0.8268611654639244,
      "num_tokens": 700260633.0,
      "step": 21941
    },
    {
      "entropy": 0.5187567798420787,
      "epoch": 2.0187692997740094,
      "grad_norm": 0.15694975852966309,
      "learning_rate": 3.271076762658325e-05,
      "loss": 0.5067,
      "mean_token_accuracy": 0.8405571952462196,
      "num_tokens": 700292794.0,
      "step": 21942
    },
    {
      "entropy": 0.5764286993071437,
      "epoch": 2.0188613061454506,
      "grad_norm": 0.16642150282859802,
      "learning_rate": 3.270770080044163e-05,
      "loss": 0.5689,
      "mean_token_accuracy": 0.8214360065758228,
      "num_tokens": 700324003.0,
      "step": 21943
    },
    {
      "entropy": 0.5304053630679846,
      "epoch": 2.0189533125168917,
      "grad_norm": 0.1609564572572708,
      "learning_rate": 3.2704633974299995e-05,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8323237486183643,
      "num_tokens": 700356176.0,
      "step": 21944
    },
    {
      "entropy": 0.5724618481472135,
      "epoch": 2.019045318888333,
      "grad_norm": 0.16343700885772705,
      "learning_rate": 3.270156714815838e-05,
      "loss": 0.5625,
      "mean_token_accuracy": 0.8221262767910957,
      "num_tokens": 700387873.0,
      "step": 21945
    },
    {
      "entropy": 0.506565916351974,
      "epoch": 2.0191373252597744,
      "grad_norm": 0.15572041273117065,
      "learning_rate": 3.2698500322016745e-05,
      "loss": 0.4984,
      "mean_token_accuracy": 0.839690912514925,
      "num_tokens": 700420058.0,
      "step": 21946
    },
    {
      "entropy": 0.4909514840692282,
      "epoch": 2.0192293316312155,
      "grad_norm": 0.16250190138816833,
      "learning_rate": 3.269543349587512e-05,
      "loss": 0.4884,
      "mean_token_accuracy": 0.8450535945594311,
      "num_tokens": 700451663.0,
      "step": 21947
    },
    {
      "entropy": 0.4598231837153435,
      "epoch": 2.0193213380026567,
      "grad_norm": 0.14808429777622223,
      "learning_rate": 3.2692366669733494e-05,
      "loss": 0.4471,
      "mean_token_accuracy": 0.8541004359722137,
      "num_tokens": 700483914.0,
      "step": 21948
    },
    {
      "entropy": 0.6135139930993319,
      "epoch": 2.019413344374098,
      "grad_norm": 0.16609297692775726,
      "learning_rate": 3.268929984359187e-05,
      "loss": 0.5918,
      "mean_token_accuracy": 0.8172815032303333,
      "num_tokens": 700516680.0,
      "step": 21949
    },
    {
      "entropy": 0.5068761520087719,
      "epoch": 2.019505350745539,
      "grad_norm": 0.15150268375873566,
      "learning_rate": 3.268623301745024e-05,
      "loss": 0.4932,
      "mean_token_accuracy": 0.8442077673971653,
      "num_tokens": 700549052.0,
      "step": 21950
    },
    {
      "entropy": 0.5543286520987749,
      "epoch": 2.0195973571169805,
      "grad_norm": 0.16443315148353577,
      "learning_rate": 3.268316619130862e-05,
      "loss": 0.5268,
      "mean_token_accuracy": 0.8318720944225788,
      "num_tokens": 700581055.0,
      "step": 21951
    },
    {
      "entropy": 0.49254854628816247,
      "epoch": 2.0196893634884217,
      "grad_norm": 0.15831932425498962,
      "learning_rate": 3.2680099365166986e-05,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8487430699169636,
      "num_tokens": 700611556.0,
      "step": 21952
    },
    {
      "entropy": 0.5984737798571587,
      "epoch": 2.019781369859863,
      "grad_norm": 0.16587600111961365,
      "learning_rate": 3.267703253902537e-05,
      "loss": 0.5854,
      "mean_token_accuracy": 0.8165303729474545,
      "num_tokens": 700643737.0,
      "step": 21953
    },
    {
      "entropy": 0.5030574407428503,
      "epoch": 2.019873376231304,
      "grad_norm": 0.15843212604522705,
      "learning_rate": 3.2673965712883735e-05,
      "loss": 0.4821,
      "mean_token_accuracy": 0.8468196131289005,
      "num_tokens": 700675565.0,
      "step": 21954
    },
    {
      "entropy": 0.5845669638365507,
      "epoch": 2.019965382602745,
      "grad_norm": 0.16915126144886017,
      "learning_rate": 3.267089888674211e-05,
      "loss": 0.559,
      "mean_token_accuracy": 0.822652630507946,
      "num_tokens": 700706629.0,
      "step": 21955
    },
    {
      "entropy": 0.42907777475193143,
      "epoch": 2.0200573889741866,
      "grad_norm": 0.1534421145915985,
      "learning_rate": 3.2667832060600484e-05,
      "loss": 0.4135,
      "mean_token_accuracy": 0.8658148720860481,
      "num_tokens": 700737716.0,
      "step": 21956
    },
    {
      "entropy": 0.5309738367795944,
      "epoch": 2.0201493953456278,
      "grad_norm": 0.16191579401493073,
      "learning_rate": 3.266476523445886e-05,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8393290191888809,
      "num_tokens": 700769413.0,
      "step": 21957
    },
    {
      "entropy": 0.5229270318523049,
      "epoch": 2.020241401717069,
      "grad_norm": 0.1590062528848648,
      "learning_rate": 3.2661698408317234e-05,
      "loss": 0.5042,
      "mean_token_accuracy": 0.8402824401855469,
      "num_tokens": 700800908.0,
      "step": 21958
    },
    {
      "entropy": 0.5400259317830205,
      "epoch": 2.02033340808851,
      "grad_norm": 0.16027195751667023,
      "learning_rate": 3.265863158217561e-05,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8356324695050716,
      "num_tokens": 700832875.0,
      "step": 21959
    },
    {
      "entropy": 0.491018476895988,
      "epoch": 2.020425414459951,
      "grad_norm": 0.1526038497686386,
      "learning_rate": 3.2655564756033976e-05,
      "loss": 0.4706,
      "mean_token_accuracy": 0.8481470011174679,
      "num_tokens": 700865360.0,
      "step": 21960
    },
    {
      "entropy": 0.41165861673653126,
      "epoch": 2.0205174208313927,
      "grad_norm": 0.14734657108783722,
      "learning_rate": 3.265249792989236e-05,
      "loss": 0.3947,
      "mean_token_accuracy": 0.8732619062066078,
      "num_tokens": 700897193.0,
      "step": 21961
    },
    {
      "entropy": 0.5196669902652502,
      "epoch": 2.020609427202834,
      "grad_norm": 0.16814465820789337,
      "learning_rate": 3.2649431103750725e-05,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8406430147588253,
      "num_tokens": 700928650.0,
      "step": 21962
    },
    {
      "entropy": 0.5703305187635124,
      "epoch": 2.020701433574275,
      "grad_norm": 0.16274073719978333,
      "learning_rate": 3.264636427760911e-05,
      "loss": 0.5575,
      "mean_token_accuracy": 0.8244023248553276,
      "num_tokens": 700960582.0,
      "step": 21963
    },
    {
      "entropy": 0.5174929751083255,
      "epoch": 2.020793439945716,
      "grad_norm": 0.15595410764217377,
      "learning_rate": 3.264329745146748e-05,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8412653058767319,
      "num_tokens": 700992744.0,
      "step": 21964
    },
    {
      "entropy": 0.529096064157784,
      "epoch": 2.0208854463171573,
      "grad_norm": 0.1572636216878891,
      "learning_rate": 3.264023062532585e-05,
      "loss": 0.5265,
      "mean_token_accuracy": 0.8328158147633076,
      "num_tokens": 701024839.0,
      "step": 21965
    },
    {
      "entropy": 0.46355781331658363,
      "epoch": 2.020977452688599,
      "grad_norm": 0.14897520840168,
      "learning_rate": 3.263716379918423e-05,
      "loss": 0.4405,
      "mean_token_accuracy": 0.8605387732386589,
      "num_tokens": 701056940.0,
      "step": 21966
    },
    {
      "entropy": 0.5515089640393853,
      "epoch": 2.02106945906004,
      "grad_norm": 0.1561877280473709,
      "learning_rate": 3.26340969730426e-05,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8292223587632179,
      "num_tokens": 701089445.0,
      "step": 21967
    },
    {
      "entropy": 0.5210096659138799,
      "epoch": 2.021161465431481,
      "grad_norm": 0.1593599021434784,
      "learning_rate": 3.263103014690097e-05,
      "loss": 0.4948,
      "mean_token_accuracy": 0.8408520296216011,
      "num_tokens": 701121672.0,
      "step": 21968
    },
    {
      "entropy": 0.4324651067145169,
      "epoch": 2.0212534718029223,
      "grad_norm": 0.15414097905158997,
      "learning_rate": 3.262796332075935e-05,
      "loss": 0.4181,
      "mean_token_accuracy": 0.8645920492708683,
      "num_tokens": 701153979.0,
      "step": 21969
    },
    {
      "entropy": 0.5154617382213473,
      "epoch": 2.0213454781743634,
      "grad_norm": 0.15896767377853394,
      "learning_rate": 3.262489649461772e-05,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8461446799337864,
      "num_tokens": 701185790.0,
      "step": 21970
    },
    {
      "entropy": 0.611383177805692,
      "epoch": 2.021437484545805,
      "grad_norm": 0.1678403615951538,
      "learning_rate": 3.26218296684761e-05,
      "loss": 0.5942,
      "mean_token_accuracy": 0.8174673989415169,
      "num_tokens": 701217572.0,
      "step": 21971
    },
    {
      "entropy": 0.6632238812744617,
      "epoch": 2.021529490917246,
      "grad_norm": 0.17845642566680908,
      "learning_rate": 3.261876284233447e-05,
      "loss": 0.6537,
      "mean_token_accuracy": 0.7971382290124893,
      "num_tokens": 701249480.0,
      "step": 21972
    },
    {
      "entropy": 0.5216924538835883,
      "epoch": 2.0216214972886872,
      "grad_norm": 0.15217730402946472,
      "learning_rate": 3.261569601619284e-05,
      "loss": 0.4984,
      "mean_token_accuracy": 0.839003711938858,
      "num_tokens": 701281504.0,
      "step": 21973
    },
    {
      "entropy": 0.6138704447075725,
      "epoch": 2.0217135036601284,
      "grad_norm": 0.1674530804157257,
      "learning_rate": 3.261262919005122e-05,
      "loss": 0.5809,
      "mean_token_accuracy": 0.81784313544631,
      "num_tokens": 701312496.0,
      "step": 21974
    },
    {
      "entropy": 0.5838758703321218,
      "epoch": 2.0218055100315695,
      "grad_norm": 0.1659039407968521,
      "learning_rate": 3.260956236390959e-05,
      "loss": 0.5609,
      "mean_token_accuracy": 0.8190540969371796,
      "num_tokens": 701344119.0,
      "step": 21975
    },
    {
      "entropy": 0.593066981062293,
      "epoch": 2.021897516403011,
      "grad_norm": 0.16505713760852814,
      "learning_rate": 3.2606495537767964e-05,
      "loss": 0.5729,
      "mean_token_accuracy": 0.81995664909482,
      "num_tokens": 701376438.0,
      "step": 21976
    },
    {
      "entropy": 0.5352795054204762,
      "epoch": 2.021989522774452,
      "grad_norm": 0.15696759521961212,
      "learning_rate": 3.260342871162634e-05,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8364660814404488,
      "num_tokens": 701408797.0,
      "step": 21977
    },
    {
      "entropy": 0.584213400259614,
      "epoch": 2.0220815291458933,
      "grad_norm": 0.16420771181583405,
      "learning_rate": 3.260036188548471e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8241406045854092,
      "num_tokens": 701440607.0,
      "step": 21978
    },
    {
      "entropy": 0.41333821625448763,
      "epoch": 2.0221735355173345,
      "grad_norm": 0.14823244512081146,
      "learning_rate": 3.259729505934309e-05,
      "loss": 0.3907,
      "mean_token_accuracy": 0.8756167851388454,
      "num_tokens": 701473164.0,
      "step": 21979
    },
    {
      "entropy": 0.6014729011803865,
      "epoch": 2.0222655418887756,
      "grad_norm": 0.16748958826065063,
      "learning_rate": 3.259422823320146e-05,
      "loss": 0.5882,
      "mean_token_accuracy": 0.8163076899945736,
      "num_tokens": 701505298.0,
      "step": 21980
    },
    {
      "entropy": 0.5222173077054322,
      "epoch": 2.022357548260217,
      "grad_norm": 0.1610279232263565,
      "learning_rate": 3.259116140705984e-05,
      "loss": 0.5107,
      "mean_token_accuracy": 0.8403001762926579,
      "num_tokens": 701536964.0,
      "step": 21981
    },
    {
      "entropy": 0.4631408238783479,
      "epoch": 2.0224495546316583,
      "grad_norm": 0.1497647911310196,
      "learning_rate": 3.258809458091821e-05,
      "loss": 0.451,
      "mean_token_accuracy": 0.8530904911458492,
      "num_tokens": 701568589.0,
      "step": 21982
    },
    {
      "entropy": 0.44898985186591744,
      "epoch": 2.0225415610030995,
      "grad_norm": 0.1552184671163559,
      "learning_rate": 3.258502775477658e-05,
      "loss": 0.443,
      "mean_token_accuracy": 0.8557519018650055,
      "num_tokens": 701600354.0,
      "step": 21983
    },
    {
      "entropy": 0.5048545440658927,
      "epoch": 2.0226335673745406,
      "grad_norm": 0.16229650378227234,
      "learning_rate": 3.258196092863496e-05,
      "loss": 0.4968,
      "mean_token_accuracy": 0.8409890681505203,
      "num_tokens": 701632138.0,
      "step": 21984
    },
    {
      "entropy": 0.5373260495252907,
      "epoch": 2.0227255737459817,
      "grad_norm": 0.16289155185222626,
      "learning_rate": 3.257889410249333e-05,
      "loss": 0.5232,
      "mean_token_accuracy": 0.8332479782402515,
      "num_tokens": 701664003.0,
      "step": 21985
    },
    {
      "entropy": 0.5387150114402175,
      "epoch": 2.0228175801174233,
      "grad_norm": 0.1618725061416626,
      "learning_rate": 3.2575827276351703e-05,
      "loss": 0.5206,
      "mean_token_accuracy": 0.8385115265846252,
      "num_tokens": 701696481.0,
      "step": 21986
    },
    {
      "entropy": 0.5463891262188554,
      "epoch": 2.0229095864888644,
      "grad_norm": 0.16290122270584106,
      "learning_rate": 3.257276045021008e-05,
      "loss": 0.544,
      "mean_token_accuracy": 0.8281944952905178,
      "num_tokens": 701727933.0,
      "step": 21987
    },
    {
      "entropy": 0.5884536756202579,
      "epoch": 2.0230015928603056,
      "grad_norm": 0.16735747456550598,
      "learning_rate": 3.256969362406845e-05,
      "loss": 0.5695,
      "mean_token_accuracy": 0.8224771469831467,
      "num_tokens": 701759851.0,
      "step": 21988
    },
    {
      "entropy": 0.57626480050385,
      "epoch": 2.0230935992317467,
      "grad_norm": 0.1566448211669922,
      "learning_rate": 3.256662679792683e-05,
      "loss": 0.5535,
      "mean_token_accuracy": 0.8264148756861687,
      "num_tokens": 701792523.0,
      "step": 21989
    },
    {
      "entropy": 0.48718042485415936,
      "epoch": 2.023185605603188,
      "grad_norm": 0.154062420129776,
      "learning_rate": 3.25635599717852e-05,
      "loss": 0.4571,
      "mean_token_accuracy": 0.8526941761374474,
      "num_tokens": 701824192.0,
      "step": 21990
    },
    {
      "entropy": 0.560528252273798,
      "epoch": 2.0232776119746294,
      "grad_norm": 0.16116732358932495,
      "learning_rate": 3.256049314564357e-05,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8302668668329716,
      "num_tokens": 701856510.0,
      "step": 21991
    },
    {
      "entropy": 0.5235779788345098,
      "epoch": 2.0233696183460705,
      "grad_norm": 0.16407784819602966,
      "learning_rate": 3.255742631950195e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8380858525633812,
      "num_tokens": 701888499.0,
      "step": 21992
    },
    {
      "entropy": 0.5508610103279352,
      "epoch": 2.0234616247175117,
      "grad_norm": 0.17013518512248993,
      "learning_rate": 3.255435949336032e-05,
      "loss": 0.5453,
      "mean_token_accuracy": 0.8286716602742672,
      "num_tokens": 701920425.0,
      "step": 21993
    },
    {
      "entropy": 0.569616537541151,
      "epoch": 2.023553631088953,
      "grad_norm": 0.16194981336593628,
      "learning_rate": 3.2551292667218694e-05,
      "loss": 0.5433,
      "mean_token_accuracy": 0.8285768777132034,
      "num_tokens": 701952182.0,
      "step": 21994
    },
    {
      "entropy": 0.5704574529081583,
      "epoch": 2.023645637460394,
      "grad_norm": 0.16532038152217865,
      "learning_rate": 3.2548225841077075e-05,
      "loss": 0.568,
      "mean_token_accuracy": 0.8212864361703396,
      "num_tokens": 701984692.0,
      "step": 21995
    },
    {
      "entropy": 0.5053611807525158,
      "epoch": 2.0237376438318355,
      "grad_norm": 0.15290240943431854,
      "learning_rate": 3.254515901493544e-05,
      "loss": 0.4853,
      "mean_token_accuracy": 0.844261147081852,
      "num_tokens": 702016679.0,
      "step": 21996
    },
    {
      "entropy": 0.5060626203194261,
      "epoch": 2.0238296502032767,
      "grad_norm": 0.16277176141738892,
      "learning_rate": 3.2542092188793825e-05,
      "loss": 0.4979,
      "mean_token_accuracy": 0.8420111313462257,
      "num_tokens": 702048745.0,
      "step": 21997
    },
    {
      "entropy": 0.49060916155576706,
      "epoch": 2.023921656574718,
      "grad_norm": 0.15621024370193481,
      "learning_rate": 3.253902536265219e-05,
      "loss": 0.4566,
      "mean_token_accuracy": 0.8544529303908348,
      "num_tokens": 702079813.0,
      "step": 21998
    },
    {
      "entropy": 0.5375341530889273,
      "epoch": 2.024013662946159,
      "grad_norm": 0.15730109810829163,
      "learning_rate": 3.253595853651057e-05,
      "loss": 0.523,
      "mean_token_accuracy": 0.8333170115947723,
      "num_tokens": 702112150.0,
      "step": 21999
    },
    {
      "entropy": 0.4365040569100529,
      "epoch": 2.0241056693176,
      "grad_norm": 0.15050597488880157,
      "learning_rate": 3.253289171036894e-05,
      "loss": 0.4107,
      "mean_token_accuracy": 0.8666605427861214,
      "num_tokens": 702143687.0,
      "step": 22000
    },
    {
      "entropy": 0.7153935162350535,
      "epoch": 2.0241976756890416,
      "grad_norm": 0.17135122418403625,
      "learning_rate": 3.2529824884227316e-05,
      "loss": 0.6882,
      "mean_token_accuracy": 0.7891278974711895,
      "num_tokens": 702175875.0,
      "step": 22001
    },
    {
      "entropy": 0.4263616157695651,
      "epoch": 2.0242896820604828,
      "grad_norm": 0.14174550771713257,
      "learning_rate": 3.252675805808569e-05,
      "loss": 0.4151,
      "mean_token_accuracy": 0.867340374737978,
      "num_tokens": 702208200.0,
      "step": 22002
    },
    {
      "entropy": 0.5301129864528775,
      "epoch": 2.024381688431924,
      "grad_norm": 0.16096194088459015,
      "learning_rate": 3.2523691231944066e-05,
      "loss": 0.5141,
      "mean_token_accuracy": 0.8333553746342659,
      "num_tokens": 702239122.0,
      "step": 22003
    },
    {
      "entropy": 0.5254532461985946,
      "epoch": 2.024473694803365,
      "grad_norm": 0.1649269014596939,
      "learning_rate": 3.2520624405802434e-05,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8403330743312836,
      "num_tokens": 702270414.0,
      "step": 22004
    },
    {
      "entropy": 0.5820904988795519,
      "epoch": 2.024565701174806,
      "grad_norm": 0.1642695516347885,
      "learning_rate": 3.2517557579660815e-05,
      "loss": 0.5558,
      "mean_token_accuracy": 0.8229217827320099,
      "num_tokens": 702301811.0,
      "step": 22005
    },
    {
      "entropy": 0.44959017261862755,
      "epoch": 2.0246577075462477,
      "grad_norm": 0.14688584208488464,
      "learning_rate": 3.251449075351918e-05,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8609480671584606,
      "num_tokens": 702334075.0,
      "step": 22006
    },
    {
      "entropy": 0.5210191714577377,
      "epoch": 2.024749713917689,
      "grad_norm": 0.16258648037910461,
      "learning_rate": 3.251142392737756e-05,
      "loss": 0.4926,
      "mean_token_accuracy": 0.846203938126564,
      "num_tokens": 702365069.0,
      "step": 22007
    },
    {
      "entropy": 0.4956805454567075,
      "epoch": 2.02484172028913,
      "grad_norm": 0.16023555397987366,
      "learning_rate": 3.250835710123593e-05,
      "loss": 0.4863,
      "mean_token_accuracy": 0.8439951911568642,
      "num_tokens": 702397278.0,
      "step": 22008
    },
    {
      "entropy": 0.41435869387350976,
      "epoch": 2.024933726660571,
      "grad_norm": 0.14371919631958008,
      "learning_rate": 3.250529027509431e-05,
      "loss": 0.398,
      "mean_token_accuracy": 0.8705567084252834,
      "num_tokens": 702429666.0,
      "step": 22009
    },
    {
      "entropy": 0.4778852283488959,
      "epoch": 2.0250257330320123,
      "grad_norm": 0.15448041260242462,
      "learning_rate": 3.250222344895268e-05,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8528942912817001,
      "num_tokens": 702461274.0,
      "step": 22010
    },
    {
      "entropy": 0.5880482322536409,
      "epoch": 2.025117739403454,
      "grad_norm": 0.16226473450660706,
      "learning_rate": 3.2499156622811056e-05,
      "loss": 0.5704,
      "mean_token_accuracy": 0.8224683254957199,
      "num_tokens": 702493673.0,
      "step": 22011
    },
    {
      "entropy": 0.5255766334012151,
      "epoch": 2.025209745774895,
      "grad_norm": 0.159864142537117,
      "learning_rate": 3.2496089796669424e-05,
      "loss": 0.5144,
      "mean_token_accuracy": 0.8365638963878155,
      "num_tokens": 702526210.0,
      "step": 22012
    },
    {
      "entropy": 0.499677000567317,
      "epoch": 2.025301752146336,
      "grad_norm": 0.15821723639965057,
      "learning_rate": 3.2493022970527805e-05,
      "loss": 0.4826,
      "mean_token_accuracy": 0.8467301651835442,
      "num_tokens": 702558224.0,
      "step": 22013
    },
    {
      "entropy": 0.5368503928184509,
      "epoch": 2.0253937585177773,
      "grad_norm": 0.16319520771503448,
      "learning_rate": 3.248995614438617e-05,
      "loss": 0.5306,
      "mean_token_accuracy": 0.830591581761837,
      "num_tokens": 702590342.0,
      "step": 22014
    },
    {
      "entropy": 0.5625534616410732,
      "epoch": 2.0254857648892184,
      "grad_norm": 0.169070765376091,
      "learning_rate": 3.248688931824455e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8257043957710266,
      "num_tokens": 702620958.0,
      "step": 22015
    },
    {
      "entropy": 0.47026585042476654,
      "epoch": 2.02557777126066,
      "grad_norm": 0.1561085283756256,
      "learning_rate": 3.248382249210292e-05,
      "loss": 0.4636,
      "mean_token_accuracy": 0.8474183715879917,
      "num_tokens": 702653311.0,
      "step": 22016
    },
    {
      "entropy": 0.507026681676507,
      "epoch": 2.025669777632101,
      "grad_norm": 0.15198597311973572,
      "learning_rate": 3.24807556659613e-05,
      "loss": 0.4981,
      "mean_token_accuracy": 0.8422052711248398,
      "num_tokens": 702685875.0,
      "step": 22017
    },
    {
      "entropy": 0.5144296227954328,
      "epoch": 2.0257617840035422,
      "grad_norm": 0.1582697033882141,
      "learning_rate": 3.247768883981967e-05,
      "loss": 0.508,
      "mean_token_accuracy": 0.8347934931516647,
      "num_tokens": 702718379.0,
      "step": 22018
    },
    {
      "entropy": 0.5237482679076493,
      "epoch": 2.0258537903749834,
      "grad_norm": 0.1578574925661087,
      "learning_rate": 3.2474622013678047e-05,
      "loss": 0.502,
      "mean_token_accuracy": 0.841435145586729,
      "num_tokens": 702751138.0,
      "step": 22019
    },
    {
      "entropy": 0.5513110971078277,
      "epoch": 2.0259457967464245,
      "grad_norm": 0.1692739725112915,
      "learning_rate": 3.2471555187536414e-05,
      "loss": 0.539,
      "mean_token_accuracy": 0.8301381729543209,
      "num_tokens": 702783115.0,
      "step": 22020
    },
    {
      "entropy": 0.5395059511065483,
      "epoch": 2.026037803117866,
      "grad_norm": 0.16456401348114014,
      "learning_rate": 3.2468488361394796e-05,
      "loss": 0.5248,
      "mean_token_accuracy": 0.8346368409693241,
      "num_tokens": 702814856.0,
      "step": 22021
    },
    {
      "entropy": 0.5362414829432964,
      "epoch": 2.026129809489307,
      "grad_norm": 0.16146332025527954,
      "learning_rate": 3.2465421535253164e-05,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8365998305380344,
      "num_tokens": 702847336.0,
      "step": 22022
    },
    {
      "entropy": 0.5792262498289347,
      "epoch": 2.0262218158607483,
      "grad_norm": 0.1692376583814621,
      "learning_rate": 3.2462354709111545e-05,
      "loss": 0.5692,
      "mean_token_accuracy": 0.8219232857227325,
      "num_tokens": 702879101.0,
      "step": 22023
    },
    {
      "entropy": 0.49713869765400887,
      "epoch": 2.0263138222321895,
      "grad_norm": 0.15371713042259216,
      "learning_rate": 3.245928788296992e-05,
      "loss": 0.4797,
      "mean_token_accuracy": 0.8451791629195213,
      "num_tokens": 702911147.0,
      "step": 22024
    },
    {
      "entropy": 0.5527558871544898,
      "epoch": 2.0264058286036306,
      "grad_norm": 0.16027826070785522,
      "learning_rate": 3.245622105682829e-05,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8339207135140896,
      "num_tokens": 702942890.0,
      "step": 22025
    },
    {
      "entropy": 0.5151871903799474,
      "epoch": 2.026497834975072,
      "grad_norm": 0.1620984971523285,
      "learning_rate": 3.245315423068667e-05,
      "loss": 0.5027,
      "mean_token_accuracy": 0.8408127054572105,
      "num_tokens": 702975156.0,
      "step": 22026
    },
    {
      "entropy": 0.5075470288284123,
      "epoch": 2.0265898413465133,
      "grad_norm": 0.15459664165973663,
      "learning_rate": 3.245008740454504e-05,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8503748401999474,
      "num_tokens": 703007091.0,
      "step": 22027
    },
    {
      "entropy": 0.4494113167747855,
      "epoch": 2.0266818477179545,
      "grad_norm": 0.14935345947742462,
      "learning_rate": 3.244702057840341e-05,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8589430190622807,
      "num_tokens": 703038841.0,
      "step": 22028
    },
    {
      "entropy": 0.6244312934577465,
      "epoch": 2.0267738540893956,
      "grad_norm": 0.1659003496170044,
      "learning_rate": 3.2443953752261786e-05,
      "loss": 0.6124,
      "mean_token_accuracy": 0.8090309165418148,
      "num_tokens": 703070931.0,
      "step": 22029
    },
    {
      "entropy": 0.4640434496104717,
      "epoch": 2.0268658604608367,
      "grad_norm": 0.15956364572048187,
      "learning_rate": 3.244088692612016e-05,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8597678951919079,
      "num_tokens": 703102441.0,
      "step": 22030
    },
    {
      "entropy": 0.4477474233135581,
      "epoch": 2.0269578668322783,
      "grad_norm": 0.14844171702861786,
      "learning_rate": 3.2437820099978536e-05,
      "loss": 0.4128,
      "mean_token_accuracy": 0.8628518022596836,
      "num_tokens": 703133305.0,
      "step": 22031
    },
    {
      "entropy": 0.5020966343581676,
      "epoch": 2.0270498732037194,
      "grad_norm": 0.16364829242229462,
      "learning_rate": 3.243475327383691e-05,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8440631069242954,
      "num_tokens": 703164628.0,
      "step": 22032
    },
    {
      "entropy": 0.43556978600099683,
      "epoch": 2.0271418795751606,
      "grad_norm": 0.14855855703353882,
      "learning_rate": 3.243168644769528e-05,
      "loss": 0.4172,
      "mean_token_accuracy": 0.8698387891054153,
      "num_tokens": 703196514.0,
      "step": 22033
    },
    {
      "entropy": 0.6213747775182128,
      "epoch": 2.0272338859466017,
      "grad_norm": 0.17315585911273956,
      "learning_rate": 3.242861962155366e-05,
      "loss": 0.6128,
      "mean_token_accuracy": 0.8112050890922546,
      "num_tokens": 703228418.0,
      "step": 22034
    },
    {
      "entropy": 0.5266063800081611,
      "epoch": 2.027325892318043,
      "grad_norm": 0.15834809839725494,
      "learning_rate": 3.242555279541203e-05,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8402326330542564,
      "num_tokens": 703260020.0,
      "step": 22035
    },
    {
      "entropy": 0.5194541085511446,
      "epoch": 2.0274178986894844,
      "grad_norm": 0.15743786096572876,
      "learning_rate": 3.24224859692704e-05,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8456260226666927,
      "num_tokens": 703292395.0,
      "step": 22036
    },
    {
      "entropy": 0.4846963668242097,
      "epoch": 2.0275099050609255,
      "grad_norm": 0.15997208654880524,
      "learning_rate": 3.241941914312878e-05,
      "loss": 0.4764,
      "mean_token_accuracy": 0.8464978076517582,
      "num_tokens": 703323515.0,
      "step": 22037
    },
    {
      "entropy": 0.5062412898987532,
      "epoch": 2.0276019114323667,
      "grad_norm": 0.1636219024658203,
      "learning_rate": 3.241635231698715e-05,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8409523367881775,
      "num_tokens": 703355459.0,
      "step": 22038
    },
    {
      "entropy": 0.6303065177053213,
      "epoch": 2.027693917803808,
      "grad_norm": 0.1712161749601364,
      "learning_rate": 3.2413285490845526e-05,
      "loss": 0.6141,
      "mean_token_accuracy": 0.8034769631922245,
      "num_tokens": 703387201.0,
      "step": 22039
    },
    {
      "entropy": 0.5411776416003704,
      "epoch": 2.027785924175249,
      "grad_norm": 0.1707560271024704,
      "learning_rate": 3.24102186647039e-05,
      "loss": 0.5382,
      "mean_token_accuracy": 0.82597865909338,
      "num_tokens": 703418725.0,
      "step": 22040
    },
    {
      "entropy": 0.6043683364987373,
      "epoch": 2.0278779305466905,
      "grad_norm": 0.16711312532424927,
      "learning_rate": 3.2407151838562275e-05,
      "loss": 0.5914,
      "mean_token_accuracy": 0.8102822601795197,
      "num_tokens": 703451173.0,
      "step": 22041
    },
    {
      "entropy": 0.44118691701442003,
      "epoch": 2.0279699369181317,
      "grad_norm": 0.14614826440811157,
      "learning_rate": 3.240408501242065e-05,
      "loss": 0.4179,
      "mean_token_accuracy": 0.8660100139677525,
      "num_tokens": 703483215.0,
      "step": 22042
    },
    {
      "entropy": 0.5965199153870344,
      "epoch": 2.028061943289573,
      "grad_norm": 0.17428718507289886,
      "learning_rate": 3.240101818627902e-05,
      "loss": 0.5863,
      "mean_token_accuracy": 0.8140998743474483,
      "num_tokens": 703513626.0,
      "step": 22043
    },
    {
      "entropy": 0.5077769691124558,
      "epoch": 2.028153949661014,
      "grad_norm": 0.16407297551631927,
      "learning_rate": 3.23979513601374e-05,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8430775217711926,
      "num_tokens": 703545167.0,
      "step": 22044
    },
    {
      "entropy": 0.5050848741084337,
      "epoch": 2.028245956032455,
      "grad_norm": 0.1622973382472992,
      "learning_rate": 3.239488453399577e-05,
      "loss": 0.4927,
      "mean_token_accuracy": 0.8405608013272285,
      "num_tokens": 703576685.0,
      "step": 22045
    },
    {
      "entropy": 0.5638315537944436,
      "epoch": 2.0283379624038966,
      "grad_norm": 0.1673995703458786,
      "learning_rate": 3.239181770785414e-05,
      "loss": 0.538,
      "mean_token_accuracy": 0.8308652080595493,
      "num_tokens": 703607995.0,
      "step": 22046
    },
    {
      "entropy": 0.46580384485423565,
      "epoch": 2.0284299687753378,
      "grad_norm": 0.154900461435318,
      "learning_rate": 3.2388750881712516e-05,
      "loss": 0.4504,
      "mean_token_accuracy": 0.8570082746446133,
      "num_tokens": 703639426.0,
      "step": 22047
    },
    {
      "entropy": 0.5760490605607629,
      "epoch": 2.028521975146779,
      "grad_norm": 0.16251355409622192,
      "learning_rate": 3.238568405557089e-05,
      "loss": 0.5605,
      "mean_token_accuracy": 0.8225262388586998,
      "num_tokens": 703671143.0,
      "step": 22048
    },
    {
      "entropy": 0.44086504355072975,
      "epoch": 2.02861398151822,
      "grad_norm": 0.1418934017419815,
      "learning_rate": 3.2382617229429266e-05,
      "loss": 0.4227,
      "mean_token_accuracy": 0.8658895529806614,
      "num_tokens": 703703579.0,
      "step": 22049
    },
    {
      "entropy": 0.4436280746012926,
      "epoch": 2.028705987889661,
      "grad_norm": 0.14546416699886322,
      "learning_rate": 3.237955040328764e-05,
      "loss": 0.4238,
      "mean_token_accuracy": 0.8618060685694218,
      "num_tokens": 703736188.0,
      "step": 22050
    },
    {
      "entropy": 0.5651517538353801,
      "epoch": 2.0287979942611027,
      "grad_norm": 0.16295911371707916,
      "learning_rate": 3.237648357714601e-05,
      "loss": 0.5403,
      "mean_token_accuracy": 0.829198744148016,
      "num_tokens": 703767911.0,
      "step": 22051
    },
    {
      "entropy": 0.5158468475565314,
      "epoch": 2.028890000632544,
      "grad_norm": 0.15849843621253967,
      "learning_rate": 3.237341675100439e-05,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8425956554710865,
      "num_tokens": 703799396.0,
      "step": 22052
    },
    {
      "entropy": 0.42263993062078953,
      "epoch": 2.028982007003985,
      "grad_norm": 0.14479000866413116,
      "learning_rate": 3.237034992486276e-05,
      "loss": 0.3987,
      "mean_token_accuracy": 0.8739247098565102,
      "num_tokens": 703830923.0,
      "step": 22053
    },
    {
      "entropy": 0.5561156710609794,
      "epoch": 2.029074013375426,
      "grad_norm": 0.16634060442447662,
      "learning_rate": 3.236728309872113e-05,
      "loss": 0.5388,
      "mean_token_accuracy": 0.8288820274174213,
      "num_tokens": 703861850.0,
      "step": 22054
    },
    {
      "entropy": 0.47402170300483704,
      "epoch": 2.0291660197468673,
      "grad_norm": 0.15619467198848724,
      "learning_rate": 3.2364216272579513e-05,
      "loss": 0.4568,
      "mean_token_accuracy": 0.8522373512387276,
      "num_tokens": 703893959.0,
      "step": 22055
    },
    {
      "entropy": 0.5360939605161548,
      "epoch": 2.029258026118309,
      "grad_norm": 0.16113710403442383,
      "learning_rate": 3.236114944643788e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8413666486740112,
      "num_tokens": 703925340.0,
      "step": 22056
    },
    {
      "entropy": 0.5237952722236514,
      "epoch": 2.02935003248975,
      "grad_norm": 0.15716148912906647,
      "learning_rate": 3.235808262029626e-05,
      "loss": 0.5,
      "mean_token_accuracy": 0.8446027263998985,
      "num_tokens": 703956828.0,
      "step": 22057
    },
    {
      "entropy": 0.48900532349944115,
      "epoch": 2.029442038861191,
      "grad_norm": 0.15453606843948364,
      "learning_rate": 3.235501579415463e-05,
      "loss": 0.4683,
      "mean_token_accuracy": 0.8496895618736744,
      "num_tokens": 703989169.0,
      "step": 22058
    },
    {
      "entropy": 0.5459619294852018,
      "epoch": 2.0295340452326323,
      "grad_norm": 0.16171030700206757,
      "learning_rate": 3.2351948968013005e-05,
      "loss": 0.534,
      "mean_token_accuracy": 0.8297032304108143,
      "num_tokens": 704020734.0,
      "step": 22059
    },
    {
      "entropy": 0.5599976144731045,
      "epoch": 2.0296260516040734,
      "grad_norm": 0.1674814373254776,
      "learning_rate": 3.234888214187138e-05,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8248513825237751,
      "num_tokens": 704052145.0,
      "step": 22060
    },
    {
      "entropy": 0.4790110820904374,
      "epoch": 2.029718057975515,
      "grad_norm": 0.16075019538402557,
      "learning_rate": 3.2345815315729755e-05,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8541739732027054,
      "num_tokens": 704084449.0,
      "step": 22061
    },
    {
      "entropy": 0.5694238413125277,
      "epoch": 2.029810064346956,
      "grad_norm": 0.1690646857023239,
      "learning_rate": 3.234274848958813e-05,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8208187073469162,
      "num_tokens": 704116046.0,
      "step": 22062
    },
    {
      "entropy": 0.43473338335752487,
      "epoch": 2.0299020707183972,
      "grad_norm": 0.14139512181282043,
      "learning_rate": 3.2339681663446504e-05,
      "loss": 0.4189,
      "mean_token_accuracy": 0.8670500852167606,
      "num_tokens": 704148370.0,
      "step": 22063
    },
    {
      "entropy": 0.5657348148524761,
      "epoch": 2.0299940770898384,
      "grad_norm": 0.16405728459358215,
      "learning_rate": 3.233661483730487e-05,
      "loss": 0.5444,
      "mean_token_accuracy": 0.8281078673899174,
      "num_tokens": 704180472.0,
      "step": 22064
    },
    {
      "entropy": 0.45036303414963186,
      "epoch": 2.0300860834612795,
      "grad_norm": 0.14755132794380188,
      "learning_rate": 3.233354801116325e-05,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8618712052702904,
      "num_tokens": 704212332.0,
      "step": 22065
    },
    {
      "entropy": 0.5319523978978395,
      "epoch": 2.030178089832721,
      "grad_norm": 0.16560307145118713,
      "learning_rate": 3.233048118502162e-05,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8352485299110413,
      "num_tokens": 704243777.0,
      "step": 22066
    },
    {
      "entropy": 0.566919295117259,
      "epoch": 2.030270096204162,
      "grad_norm": 0.16792082786560059,
      "learning_rate": 3.2327414358879996e-05,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8303654156625271,
      "num_tokens": 704275519.0,
      "step": 22067
    },
    {
      "entropy": 0.48065097630023956,
      "epoch": 2.0303621025756033,
      "grad_norm": 0.16037611663341522,
      "learning_rate": 3.232434753273837e-05,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8498300835490227,
      "num_tokens": 704307414.0,
      "step": 22068
    },
    {
      "entropy": 0.5184540636837482,
      "epoch": 2.0304541089470445,
      "grad_norm": 0.16017234325408936,
      "learning_rate": 3.2321280706596745e-05,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8407026380300522,
      "num_tokens": 704338779.0,
      "step": 22069
    },
    {
      "entropy": 0.5068904724903405,
      "epoch": 2.0305461153184856,
      "grad_norm": 0.1603797823190689,
      "learning_rate": 3.231821388045512e-05,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8433790281414986,
      "num_tokens": 704370763.0,
      "step": 22070
    },
    {
      "entropy": 0.543709859251976,
      "epoch": 2.030638121689927,
      "grad_norm": 0.15557587146759033,
      "learning_rate": 3.2315147054313494e-05,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8368074484169483,
      "num_tokens": 704403148.0,
      "step": 22071
    },
    {
      "entropy": 0.6128828469663858,
      "epoch": 2.0307301280613683,
      "grad_norm": 0.17078566551208496,
      "learning_rate": 3.231208022817186e-05,
      "loss": 0.6093,
      "mean_token_accuracy": 0.8102190047502518,
      "num_tokens": 704434678.0,
      "step": 22072
    },
    {
      "entropy": 0.4866829104721546,
      "epoch": 2.0308221344328095,
      "grad_norm": 0.15078432857990265,
      "learning_rate": 3.2309013402030244e-05,
      "loss": 0.4651,
      "mean_token_accuracy": 0.8534174785017967,
      "num_tokens": 704466914.0,
      "step": 22073
    },
    {
      "entropy": 0.4502102294936776,
      "epoch": 2.0309141408042506,
      "grad_norm": 0.15309010446071625,
      "learning_rate": 3.230594657588861e-05,
      "loss": 0.4415,
      "mean_token_accuracy": 0.8572622239589691,
      "num_tokens": 704498879.0,
      "step": 22074
    },
    {
      "entropy": 0.46333723969291896,
      "epoch": 2.0310061471756917,
      "grad_norm": 0.15026924014091492,
      "learning_rate": 3.2302879749746986e-05,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8584891892969608,
      "num_tokens": 704531283.0,
      "step": 22075
    },
    {
      "entropy": 0.44108018558472395,
      "epoch": 2.0310981535471333,
      "grad_norm": 0.1467704325914383,
      "learning_rate": 3.229981292360536e-05,
      "loss": 0.417,
      "mean_token_accuracy": 0.8656436130404472,
      "num_tokens": 704563716.0,
      "step": 22076
    },
    {
      "entropy": 0.5178970387205482,
      "epoch": 2.0311901599185744,
      "grad_norm": 0.16240434348583221,
      "learning_rate": 3.2296746097463735e-05,
      "loss": 0.4971,
      "mean_token_accuracy": 0.8406716249883175,
      "num_tokens": 704596023.0,
      "step": 22077
    },
    {
      "entropy": 0.40587834804318845,
      "epoch": 2.0312821662900156,
      "grad_norm": 0.1459009349346161,
      "learning_rate": 3.229367927132211e-05,
      "loss": 0.3847,
      "mean_token_accuracy": 0.8755914978682995,
      "num_tokens": 704628169.0,
      "step": 22078
    },
    {
      "entropy": 0.5034570274874568,
      "epoch": 2.0313741726614567,
      "grad_norm": 0.16243329644203186,
      "learning_rate": 3.2290612445180485e-05,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8456258997321129,
      "num_tokens": 704659488.0,
      "step": 22079
    },
    {
      "entropy": 0.49809200852178037,
      "epoch": 2.031466179032898,
      "grad_norm": 0.15888778865337372,
      "learning_rate": 3.228754561903885e-05,
      "loss": 0.4828,
      "mean_token_accuracy": 0.8466382212936878,
      "num_tokens": 704691197.0,
      "step": 22080
    },
    {
      "entropy": 0.47455921955406666,
      "epoch": 2.0315581854043394,
      "grad_norm": 0.16561776399612427,
      "learning_rate": 3.2284478792897234e-05,
      "loss": 0.4606,
      "mean_token_accuracy": 0.8532640226185322,
      "num_tokens": 704722148.0,
      "step": 22081
    },
    {
      "entropy": 0.5329192047938704,
      "epoch": 2.0316501917757805,
      "grad_norm": 0.1623637080192566,
      "learning_rate": 3.22814119667556e-05,
      "loss": 0.508,
      "mean_token_accuracy": 0.8402612954378128,
      "num_tokens": 704753618.0,
      "step": 22082
    },
    {
      "entropy": 0.5759749570861459,
      "epoch": 2.0317421981472217,
      "grad_norm": 0.164170503616333,
      "learning_rate": 3.227834514061398e-05,
      "loss": 0.5587,
      "mean_token_accuracy": 0.8238427340984344,
      "num_tokens": 704785979.0,
      "step": 22083
    },
    {
      "entropy": 0.6201305538415909,
      "epoch": 2.031834204518663,
      "grad_norm": 0.16908138990402222,
      "learning_rate": 3.227527831447235e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.809641145169735,
      "num_tokens": 704818217.0,
      "step": 22084
    },
    {
      "entropy": 0.5219004340469837,
      "epoch": 2.031926210890104,
      "grad_norm": 0.15956838428974152,
      "learning_rate": 3.2272211488330726e-05,
      "loss": 0.516,
      "mean_token_accuracy": 0.8379022143781185,
      "num_tokens": 704849845.0,
      "step": 22085
    },
    {
      "entropy": 0.5154461031779647,
      "epoch": 2.0320182172615455,
      "grad_norm": 0.16187700629234314,
      "learning_rate": 3.226914466218911e-05,
      "loss": 0.5083,
      "mean_token_accuracy": 0.8384222015738487,
      "num_tokens": 704881769.0,
      "step": 22086
    },
    {
      "entropy": 0.6430290825664997,
      "epoch": 2.0321102236329867,
      "grad_norm": 0.17813676595687866,
      "learning_rate": 3.2266077836047475e-05,
      "loss": 0.6437,
      "mean_token_accuracy": 0.8011419400572777,
      "num_tokens": 704913860.0,
      "step": 22087
    },
    {
      "entropy": 0.4888028968125582,
      "epoch": 2.032202230004428,
      "grad_norm": 0.15977774560451508,
      "learning_rate": 3.226301100990585e-05,
      "loss": 0.4805,
      "mean_token_accuracy": 0.8457198962569237,
      "num_tokens": 704944998.0,
      "step": 22088
    },
    {
      "entropy": 0.5570105155929923,
      "epoch": 2.032294236375869,
      "grad_norm": 0.1638607531785965,
      "learning_rate": 3.2259944183764224e-05,
      "loss": 0.5475,
      "mean_token_accuracy": 0.8310661241412163,
      "num_tokens": 704976860.0,
      "step": 22089
    },
    {
      "entropy": 0.5839614812284708,
      "epoch": 2.03238624274731,
      "grad_norm": 0.16121870279312134,
      "learning_rate": 3.22568773576226e-05,
      "loss": 0.5635,
      "mean_token_accuracy": 0.8240274526178837,
      "num_tokens": 705008889.0,
      "step": 22090
    },
    {
      "entropy": 0.5808981694281101,
      "epoch": 2.0324782491187516,
      "grad_norm": 0.16805803775787354,
      "learning_rate": 3.2253810531480974e-05,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8215088620781898,
      "num_tokens": 705041017.0,
      "step": 22091
    },
    {
      "entropy": 0.6206187237985432,
      "epoch": 2.0325702554901928,
      "grad_norm": 0.17075183987617493,
      "learning_rate": 3.225074370533935e-05,
      "loss": 0.5981,
      "mean_token_accuracy": 0.8170602954924107,
      "num_tokens": 705072917.0,
      "step": 22092
    },
    {
      "entropy": 0.47157166339457035,
      "epoch": 2.032662261861634,
      "grad_norm": 0.15682323276996613,
      "learning_rate": 3.2247676879197716e-05,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8510980047285557,
      "num_tokens": 705104437.0,
      "step": 22093
    },
    {
      "entropy": 0.4436232422012836,
      "epoch": 2.032754268233075,
      "grad_norm": 0.14062561094760895,
      "learning_rate": 3.22446100530561e-05,
      "loss": 0.4371,
      "mean_token_accuracy": 0.8608074449002743,
      "num_tokens": 705137095.0,
      "step": 22094
    },
    {
      "entropy": 0.5106499651446939,
      "epoch": 2.032846274604516,
      "grad_norm": 0.15905851125717163,
      "learning_rate": 3.2241543226914466e-05,
      "loss": 0.5,
      "mean_token_accuracy": 0.8404437899589539,
      "num_tokens": 705168506.0,
      "step": 22095
    },
    {
      "entropy": 0.5917447954416275,
      "epoch": 2.0329382809759577,
      "grad_norm": 0.1671927273273468,
      "learning_rate": 3.223847640077284e-05,
      "loss": 0.5769,
      "mean_token_accuracy": 0.8174739740788937,
      "num_tokens": 705199634.0,
      "step": 22096
    },
    {
      "entropy": 0.5071500167250633,
      "epoch": 2.033030287347399,
      "grad_norm": 0.16300466656684875,
      "learning_rate": 3.2235409574631215e-05,
      "loss": 0.4799,
      "mean_token_accuracy": 0.8433669395744801,
      "num_tokens": 705230445.0,
      "step": 22097
    },
    {
      "entropy": 0.542832164093852,
      "epoch": 2.03312229371884,
      "grad_norm": 0.16194400191307068,
      "learning_rate": 3.223234274848959e-05,
      "loss": 0.5196,
      "mean_token_accuracy": 0.8373375199735165,
      "num_tokens": 705262203.0,
      "step": 22098
    },
    {
      "entropy": 0.457228479674086,
      "epoch": 2.033214300090281,
      "grad_norm": 0.14982712268829346,
      "learning_rate": 3.2229275922347964e-05,
      "loss": 0.4402,
      "mean_token_accuracy": 0.8557300791144371,
      "num_tokens": 705293888.0,
      "step": 22099
    },
    {
      "entropy": 0.49612031131982803,
      "epoch": 2.0333063064617223,
      "grad_norm": 0.15630775690078735,
      "learning_rate": 3.222620909620634e-05,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8508412912487984,
      "num_tokens": 705325070.0,
      "step": 22100
    },
    {
      "entropy": 0.5881086923182011,
      "epoch": 2.033398312833164,
      "grad_norm": 0.16614285111427307,
      "learning_rate": 3.2223142270064713e-05,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8226639330387115,
      "num_tokens": 705356642.0,
      "step": 22101
    },
    {
      "entropy": 0.4277556552551687,
      "epoch": 2.033490319204605,
      "grad_norm": 0.15163718163967133,
      "learning_rate": 3.222007544392309e-05,
      "loss": 0.424,
      "mean_token_accuracy": 0.867221400141716,
      "num_tokens": 705387895.0,
      "step": 22102
    },
    {
      "entropy": 0.5648830607533455,
      "epoch": 2.033582325576046,
      "grad_norm": 0.16209359467029572,
      "learning_rate": 3.2217008617781456e-05,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8278464041650295,
      "num_tokens": 705419831.0,
      "step": 22103
    },
    {
      "entropy": 0.498621073551476,
      "epoch": 2.0336743319474873,
      "grad_norm": 0.15689051151275635,
      "learning_rate": 3.221394179163984e-05,
      "loss": 0.4818,
      "mean_token_accuracy": 0.8497644700109959,
      "num_tokens": 705452227.0,
      "step": 22104
    },
    {
      "entropy": 0.5707465671002865,
      "epoch": 2.0337663383189284,
      "grad_norm": 0.15921595692634583,
      "learning_rate": 3.2210874965498205e-05,
      "loss": 0.5483,
      "mean_token_accuracy": 0.8285224065184593,
      "num_tokens": 705484601.0,
      "step": 22105
    },
    {
      "entropy": 0.5661639086902142,
      "epoch": 2.03385834469037,
      "grad_norm": 0.1647530049085617,
      "learning_rate": 3.220780813935658e-05,
      "loss": 0.554,
      "mean_token_accuracy": 0.8260684534907341,
      "num_tokens": 705516461.0,
      "step": 22106
    },
    {
      "entropy": 0.6335455141961575,
      "epoch": 2.033950351061811,
      "grad_norm": 0.16622702777385712,
      "learning_rate": 3.2204741313214955e-05,
      "loss": 0.6089,
      "mean_token_accuracy": 0.8105546273291111,
      "num_tokens": 705548143.0,
      "step": 22107
    },
    {
      "entropy": 0.5745503180660307,
      "epoch": 2.0340423574332522,
      "grad_norm": 0.1614062339067459,
      "learning_rate": 3.220167448707333e-05,
      "loss": 0.5568,
      "mean_token_accuracy": 0.8270040154457092,
      "num_tokens": 705580352.0,
      "step": 22108
    },
    {
      "entropy": 0.5447928495705128,
      "epoch": 2.0341343638046934,
      "grad_norm": 0.16386552155017853,
      "learning_rate": 3.2198607660931704e-05,
      "loss": 0.5243,
      "mean_token_accuracy": 0.8333318494260311,
      "num_tokens": 705612566.0,
      "step": 22109
    },
    {
      "entropy": 0.6790245492011309,
      "epoch": 2.0342263701761345,
      "grad_norm": 0.17660392820835114,
      "learning_rate": 3.219554083479008e-05,
      "loss": 0.6533,
      "mean_token_accuracy": 0.7964717634022236,
      "num_tokens": 705644021.0,
      "step": 22110
    },
    {
      "entropy": 0.48223420325666666,
      "epoch": 2.034318376547576,
      "grad_norm": 0.15663135051727295,
      "learning_rate": 3.2192474008648446e-05,
      "loss": 0.467,
      "mean_token_accuracy": 0.8533414974808693,
      "num_tokens": 705676747.0,
      "step": 22111
    },
    {
      "entropy": 0.5757378106936812,
      "epoch": 2.034410382919017,
      "grad_norm": 0.1617913693189621,
      "learning_rate": 3.218940718250683e-05,
      "loss": 0.5602,
      "mean_token_accuracy": 0.8268785588443279,
      "num_tokens": 705709156.0,
      "step": 22112
    },
    {
      "entropy": 0.49230913165956736,
      "epoch": 2.0345023892904583,
      "grad_norm": 0.16047754883766174,
      "learning_rate": 3.2186340356365196e-05,
      "loss": 0.4731,
      "mean_token_accuracy": 0.8486907072365284,
      "num_tokens": 705740202.0,
      "step": 22113
    },
    {
      "entropy": 0.5216591190546751,
      "epoch": 2.0345943956618995,
      "grad_norm": 0.15690891444683075,
      "learning_rate": 3.218327353022357e-05,
      "loss": 0.503,
      "mean_token_accuracy": 0.8390974253416061,
      "num_tokens": 705772072.0,
      "step": 22114
    },
    {
      "entropy": 0.5170350968837738,
      "epoch": 2.0346864020333406,
      "grad_norm": 0.1504867821931839,
      "learning_rate": 3.2180206704081945e-05,
      "loss": 0.497,
      "mean_token_accuracy": 0.840747743844986,
      "num_tokens": 705804271.0,
      "step": 22115
    },
    {
      "entropy": 0.6251952666789293,
      "epoch": 2.034778408404782,
      "grad_norm": 0.17081445455551147,
      "learning_rate": 3.217713987794032e-05,
      "loss": 0.619,
      "mean_token_accuracy": 0.8055742606520653,
      "num_tokens": 705836607.0,
      "step": 22116
    },
    {
      "entropy": 0.5340581983327866,
      "epoch": 2.0348704147762233,
      "grad_norm": 0.16196505725383759,
      "learning_rate": 3.21740730517987e-05,
      "loss": 0.5208,
      "mean_token_accuracy": 0.8385929241776466,
      "num_tokens": 705868427.0,
      "step": 22117
    },
    {
      "entropy": 0.48539130855351686,
      "epoch": 2.0349624211476645,
      "grad_norm": 0.15457932651042938,
      "learning_rate": 3.217100622565707e-05,
      "loss": 0.4819,
      "mean_token_accuracy": 0.8489863127470016,
      "num_tokens": 705901195.0,
      "step": 22118
    },
    {
      "entropy": 0.5440027583390474,
      "epoch": 2.0350544275191056,
      "grad_norm": 0.16129538416862488,
      "learning_rate": 3.2167939399515444e-05,
      "loss": 0.5268,
      "mean_token_accuracy": 0.8334593437612057,
      "num_tokens": 705933449.0,
      "step": 22119
    },
    {
      "entropy": 0.4775239937007427,
      "epoch": 2.0351464338905467,
      "grad_norm": 0.15360093116760254,
      "learning_rate": 3.216487257337382e-05,
      "loss": 0.447,
      "mean_token_accuracy": 0.851346917450428,
      "num_tokens": 705964355.0,
      "step": 22120
    },
    {
      "entropy": 0.5890714288689196,
      "epoch": 2.0352384402619883,
      "grad_norm": 0.17024366557598114,
      "learning_rate": 3.216180574723219e-05,
      "loss": 0.5726,
      "mean_token_accuracy": 0.8173647858202457,
      "num_tokens": 705996335.0,
      "step": 22121
    },
    {
      "entropy": 0.5229213191196322,
      "epoch": 2.0353304466334294,
      "grad_norm": 0.16546796262264252,
      "learning_rate": 3.215873892109057e-05,
      "loss": 0.5048,
      "mean_token_accuracy": 0.8392862416803837,
      "num_tokens": 706028339.0,
      "step": 22122
    },
    {
      "entropy": 0.5064074425026774,
      "epoch": 2.0354224530048706,
      "grad_norm": 0.15765008330345154,
      "learning_rate": 3.215567209494894e-05,
      "loss": 0.4882,
      "mean_token_accuracy": 0.844114113599062,
      "num_tokens": 706060119.0,
      "step": 22123
    },
    {
      "entropy": 0.5334240915253758,
      "epoch": 2.0355144593763117,
      "grad_norm": 0.16224811971187592,
      "learning_rate": 3.215260526880731e-05,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8301789462566376,
      "num_tokens": 706091661.0,
      "step": 22124
    },
    {
      "entropy": 0.453858049120754,
      "epoch": 2.035606465747753,
      "grad_norm": 0.15337303280830383,
      "learning_rate": 3.214953844266569e-05,
      "loss": 0.4291,
      "mean_token_accuracy": 0.8665644787251949,
      "num_tokens": 706122561.0,
      "step": 22125
    },
    {
      "entropy": 0.531898039393127,
      "epoch": 2.0356984721191944,
      "grad_norm": 0.16199366748332977,
      "learning_rate": 3.214647161652406e-05,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8367622680962086,
      "num_tokens": 706154143.0,
      "step": 22126
    },
    {
      "entropy": 0.5168326948769391,
      "epoch": 2.0357904784906355,
      "grad_norm": 0.15451842546463013,
      "learning_rate": 3.2143404790382434e-05,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8406234048306942,
      "num_tokens": 706186562.0,
      "step": 22127
    },
    {
      "entropy": 0.5284803602844477,
      "epoch": 2.0358824848620767,
      "grad_norm": 0.16293196380138397,
      "learning_rate": 3.214033796424081e-05,
      "loss": 0.5111,
      "mean_token_accuracy": 0.8389191068708897,
      "num_tokens": 706218439.0,
      "step": 22128
    },
    {
      "entropy": 0.5732646386604756,
      "epoch": 2.035974491233518,
      "grad_norm": 0.16725367307662964,
      "learning_rate": 3.213727113809918e-05,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8274198584258556,
      "num_tokens": 706250719.0,
      "step": 22129
    },
    {
      "entropy": 0.4831363223493099,
      "epoch": 2.036066497604959,
      "grad_norm": 0.16012731194496155,
      "learning_rate": 3.213420431195756e-05,
      "loss": 0.4731,
      "mean_token_accuracy": 0.8498907648026943,
      "num_tokens": 706282413.0,
      "step": 22130
    },
    {
      "entropy": 0.47237844951450825,
      "epoch": 2.0361585039764005,
      "grad_norm": 0.15677420794963837,
      "learning_rate": 3.213113748581593e-05,
      "loss": 0.4418,
      "mean_token_accuracy": 0.8524912856519222,
      "num_tokens": 706314330.0,
      "step": 22131
    },
    {
      "entropy": 0.4549933895468712,
      "epoch": 2.0362505103478417,
      "grad_norm": 0.15296770632266998,
      "learning_rate": 3.21280706596743e-05,
      "loss": 0.4418,
      "mean_token_accuracy": 0.8620594330132008,
      "num_tokens": 706346602.0,
      "step": 22132
    },
    {
      "entropy": 0.5018701832741499,
      "epoch": 2.036342516719283,
      "grad_norm": 0.1614002138376236,
      "learning_rate": 3.212500383353268e-05,
      "loss": 0.4877,
      "mean_token_accuracy": 0.8449745997786522,
      "num_tokens": 706378009.0,
      "step": 22133
    },
    {
      "entropy": 0.5423955302685499,
      "epoch": 2.036434523090724,
      "grad_norm": 0.17176416516304016,
      "learning_rate": 3.212193700739105e-05,
      "loss": 0.5247,
      "mean_token_accuracy": 0.8304418586194515,
      "num_tokens": 706408125.0,
      "step": 22134
    },
    {
      "entropy": 0.5070161232724786,
      "epoch": 2.036526529462165,
      "grad_norm": 0.16392935812473297,
      "learning_rate": 3.2118870181249424e-05,
      "loss": 0.4933,
      "mean_token_accuracy": 0.8372266441583633,
      "num_tokens": 706440119.0,
      "step": 22135
    },
    {
      "entropy": 0.4755680151283741,
      "epoch": 2.0366185358336066,
      "grad_norm": 0.15584000945091248,
      "learning_rate": 3.21158033551078e-05,
      "loss": 0.466,
      "mean_token_accuracy": 0.8515702933073044,
      "num_tokens": 706472245.0,
      "step": 22136
    },
    {
      "entropy": 0.6204019449651241,
      "epoch": 2.0367105422050478,
      "grad_norm": 0.16880109906196594,
      "learning_rate": 3.2112736528966174e-05,
      "loss": 0.6084,
      "mean_token_accuracy": 0.8091697134077549,
      "num_tokens": 706504324.0,
      "step": 22137
    },
    {
      "entropy": 0.4995537819340825,
      "epoch": 2.036802548576489,
      "grad_norm": 0.15788136422634125,
      "learning_rate": 3.210966970282455e-05,
      "loss": 0.4831,
      "mean_token_accuracy": 0.8455257974565029,
      "num_tokens": 706536136.0,
      "step": 22138
    },
    {
      "entropy": 0.49242488108575344,
      "epoch": 2.03689455494793,
      "grad_norm": 0.1557427942752838,
      "learning_rate": 3.210660287668292e-05,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8480879031121731,
      "num_tokens": 706567711.0,
      "step": 22139
    },
    {
      "entropy": 0.543588362634182,
      "epoch": 2.036986561319371,
      "grad_norm": 0.1629531979560852,
      "learning_rate": 3.210353605054129e-05,
      "loss": 0.5319,
      "mean_token_accuracy": 0.8334037363529205,
      "num_tokens": 706599778.0,
      "step": 22140
    },
    {
      "entropy": 0.44524615863338113,
      "epoch": 2.0370785676908127,
      "grad_norm": 0.15363869071006775,
      "learning_rate": 3.210046922439967e-05,
      "loss": 0.4325,
      "mean_token_accuracy": 0.8641257584095001,
      "num_tokens": 706631690.0,
      "step": 22141
    },
    {
      "entropy": 0.4695296948775649,
      "epoch": 2.037170574062254,
      "grad_norm": 0.15121494233608246,
      "learning_rate": 3.209740239825804e-05,
      "loss": 0.4564,
      "mean_token_accuracy": 0.8546473830938339,
      "num_tokens": 706663536.0,
      "step": 22142
    },
    {
      "entropy": 0.556075875647366,
      "epoch": 2.037262580433695,
      "grad_norm": 0.16292908787727356,
      "learning_rate": 3.209433557211642e-05,
      "loss": 0.5456,
      "mean_token_accuracy": 0.8262713626027107,
      "num_tokens": 706696058.0,
      "step": 22143
    },
    {
      "entropy": 0.589357363525778,
      "epoch": 2.037354586805136,
      "grad_norm": 0.16523903608322144,
      "learning_rate": 3.209126874597479e-05,
      "loss": 0.5838,
      "mean_token_accuracy": 0.8168237693607807,
      "num_tokens": 706727945.0,
      "step": 22144
    },
    {
      "entropy": 0.6282406598329544,
      "epoch": 2.0374465931765773,
      "grad_norm": 0.16767437756061554,
      "learning_rate": 3.2088201919833164e-05,
      "loss": 0.6121,
      "mean_token_accuracy": 0.8077214173972607,
      "num_tokens": 706760643.0,
      "step": 22145
    },
    {
      "entropy": 0.5101437941193581,
      "epoch": 2.037538599548019,
      "grad_norm": 0.16206534206867218,
      "learning_rate": 3.2085135093691545e-05,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8437574282288551,
      "num_tokens": 706792735.0,
      "step": 22146
    },
    {
      "entropy": 0.44808442145586014,
      "epoch": 2.03763060591946,
      "grad_norm": 0.14515787363052368,
      "learning_rate": 3.208206826754991e-05,
      "loss": 0.426,
      "mean_token_accuracy": 0.8608957342803478,
      "num_tokens": 706824149.0,
      "step": 22147
    },
    {
      "entropy": 0.5488352628890425,
      "epoch": 2.037722612290901,
      "grad_norm": 0.1608985811471939,
      "learning_rate": 3.207900144140829e-05,
      "loss": 0.536,
      "mean_token_accuracy": 0.8264982961118221,
      "num_tokens": 706855743.0,
      "step": 22148
    },
    {
      "entropy": 0.5430042995139956,
      "epoch": 2.0378146186623423,
      "grad_norm": 0.16526088118553162,
      "learning_rate": 3.207593461526666e-05,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8329445756971836,
      "num_tokens": 706887053.0,
      "step": 22149
    },
    {
      "entropy": 0.480395830469206,
      "epoch": 2.0379066250337834,
      "grad_norm": 0.16098950803279877,
      "learning_rate": 3.207286778912504e-05,
      "loss": 0.4701,
      "mean_token_accuracy": 0.8498251289129257,
      "num_tokens": 706918124.0,
      "step": 22150
    },
    {
      "entropy": 0.4601527373306453,
      "epoch": 2.037998631405225,
      "grad_norm": 0.14815670251846313,
      "learning_rate": 3.206980096298341e-05,
      "loss": 0.4386,
      "mean_token_accuracy": 0.861958920955658,
      "num_tokens": 706950056.0,
      "step": 22151
    },
    {
      "entropy": 0.5825763950124383,
      "epoch": 2.038090637776666,
      "grad_norm": 0.1642276793718338,
      "learning_rate": 3.2066734136841787e-05,
      "loss": 0.5492,
      "mean_token_accuracy": 0.828833807259798,
      "num_tokens": 706981990.0,
      "step": 22152
    },
    {
      "entropy": 0.5256058052182198,
      "epoch": 2.0381826441481072,
      "grad_norm": 0.16185955703258514,
      "learning_rate": 3.2063667310700154e-05,
      "loss": 0.5127,
      "mean_token_accuracy": 0.8394837267696857,
      "num_tokens": 707013817.0,
      "step": 22153
    },
    {
      "entropy": 0.5053190473699942,
      "epoch": 2.0382746505195484,
      "grad_norm": 0.15581932663917542,
      "learning_rate": 3.2060600484558536e-05,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8472164794802666,
      "num_tokens": 707045643.0,
      "step": 22154
    },
    {
      "entropy": 0.43845764081925154,
      "epoch": 2.0383666568909895,
      "grad_norm": 0.14814649522304535,
      "learning_rate": 3.2057533658416904e-05,
      "loss": 0.4171,
      "mean_token_accuracy": 0.8697951473295689,
      "num_tokens": 707077514.0,
      "step": 22155
    },
    {
      "entropy": 0.4921122081577778,
      "epoch": 2.038458663262431,
      "grad_norm": 0.15534210205078125,
      "learning_rate": 3.205446683227528e-05,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8503806367516518,
      "num_tokens": 707108379.0,
      "step": 22156
    },
    {
      "entropy": 0.587845154106617,
      "epoch": 2.038550669633872,
      "grad_norm": 0.16798743605613708,
      "learning_rate": 3.205140000613365e-05,
      "loss": 0.5692,
      "mean_token_accuracy": 0.8201811350882053,
      "num_tokens": 707140442.0,
      "step": 22157
    },
    {
      "entropy": 0.4892706610262394,
      "epoch": 2.0386426760053133,
      "grad_norm": 0.15921266376972198,
      "learning_rate": 3.204833317999203e-05,
      "loss": 0.4758,
      "mean_token_accuracy": 0.8474405892193317,
      "num_tokens": 707172387.0,
      "step": 22158
    },
    {
      "entropy": 0.5381603129208088,
      "epoch": 2.0387346823767545,
      "grad_norm": 0.15882717072963715,
      "learning_rate": 3.20452663538504e-05,
      "loss": 0.5192,
      "mean_token_accuracy": 0.8391801826655865,
      "num_tokens": 707204860.0,
      "step": 22159
    },
    {
      "entropy": 0.5665100738406181,
      "epoch": 2.0388266887481956,
      "grad_norm": 0.16565977036952972,
      "learning_rate": 3.204219952770878e-05,
      "loss": 0.5484,
      "mean_token_accuracy": 0.8258659318089485,
      "num_tokens": 707236843.0,
      "step": 22160
    },
    {
      "entropy": 0.5063433796167374,
      "epoch": 2.038918695119637,
      "grad_norm": 0.15984958410263062,
      "learning_rate": 3.203913270156715e-05,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8436550945043564,
      "num_tokens": 707269297.0,
      "step": 22161
    },
    {
      "entropy": 0.5282808467745781,
      "epoch": 2.0390107014910783,
      "grad_norm": 0.16357412934303284,
      "learning_rate": 3.2036065875425526e-05,
      "loss": 0.5197,
      "mean_token_accuracy": 0.8312512263655663,
      "num_tokens": 707301108.0,
      "step": 22162
    },
    {
      "entropy": 0.5217448845505714,
      "epoch": 2.0391027078625195,
      "grad_norm": 0.16391049325466156,
      "learning_rate": 3.2032999049283894e-05,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8328217379748821,
      "num_tokens": 707333249.0,
      "step": 22163
    },
    {
      "entropy": 0.5731842741370201,
      "epoch": 2.0391947142339606,
      "grad_norm": 0.16632971167564392,
      "learning_rate": 3.2029932223142276e-05,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8243963308632374,
      "num_tokens": 707364812.0,
      "step": 22164
    },
    {
      "entropy": 0.5635503018274903,
      "epoch": 2.0392867206054017,
      "grad_norm": 0.16201236844062805,
      "learning_rate": 3.2026865397000643e-05,
      "loss": 0.5549,
      "mean_token_accuracy": 0.825439278036356,
      "num_tokens": 707397340.0,
      "step": 22165
    },
    {
      "entropy": 0.43249684665352106,
      "epoch": 2.0393787269768433,
      "grad_norm": 0.15125539898872375,
      "learning_rate": 3.202379857085902e-05,
      "loss": 0.4151,
      "mean_token_accuracy": 0.8663834147155285,
      "num_tokens": 707429496.0,
      "step": 22166
    },
    {
      "entropy": 0.5809471067041159,
      "epoch": 2.0394707333482844,
      "grad_norm": 0.1636117547750473,
      "learning_rate": 3.202073174471739e-05,
      "loss": 0.5684,
      "mean_token_accuracy": 0.8165078945457935,
      "num_tokens": 707461973.0,
      "step": 22167
    },
    {
      "entropy": 0.5365654290653765,
      "epoch": 2.0395627397197256,
      "grad_norm": 0.16524562239646912,
      "learning_rate": 3.201766491857577e-05,
      "loss": 0.5253,
      "mean_token_accuracy": 0.8330828174948692,
      "num_tokens": 707493237.0,
      "step": 22168
    },
    {
      "entropy": 0.5280088586732745,
      "epoch": 2.0396547460911667,
      "grad_norm": 0.1574297398328781,
      "learning_rate": 3.201459809243414e-05,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8424331322312355,
      "num_tokens": 707525475.0,
      "step": 22169
    },
    {
      "entropy": 0.5837240018881857,
      "epoch": 2.039746752462608,
      "grad_norm": 0.16718226671218872,
      "learning_rate": 3.201153126629252e-05,
      "loss": 0.5688,
      "mean_token_accuracy": 0.8219083547592163,
      "num_tokens": 707557032.0,
      "step": 22170
    },
    {
      "entropy": 0.6224393667653203,
      "epoch": 2.0398387588340494,
      "grad_norm": 0.1675873100757599,
      "learning_rate": 3.2008464440150885e-05,
      "loss": 0.6124,
      "mean_token_accuracy": 0.8089926429092884,
      "num_tokens": 707588944.0,
      "step": 22171
    },
    {
      "entropy": 0.538417843170464,
      "epoch": 2.0399307652054905,
      "grad_norm": 0.1658662110567093,
      "learning_rate": 3.2005397614009266e-05,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8294005431234837,
      "num_tokens": 707620598.0,
      "step": 22172
    },
    {
      "entropy": 0.4485391518101096,
      "epoch": 2.0400227715769317,
      "grad_norm": 0.15084806084632874,
      "learning_rate": 3.2002330787867634e-05,
      "loss": 0.4286,
      "mean_token_accuracy": 0.8653685823082924,
      "num_tokens": 707652399.0,
      "step": 22173
    },
    {
      "entropy": 0.5664929216727614,
      "epoch": 2.040114777948373,
      "grad_norm": 0.16592544317245483,
      "learning_rate": 3.199926396172601e-05,
      "loss": 0.5463,
      "mean_token_accuracy": 0.8278035558760166,
      "num_tokens": 707684298.0,
      "step": 22174
    },
    {
      "entropy": 0.5631989203393459,
      "epoch": 2.040206784319814,
      "grad_norm": 0.1638406217098236,
      "learning_rate": 3.199619713558438e-05,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8250185661017895,
      "num_tokens": 707716065.0,
      "step": 22175
    },
    {
      "entropy": 0.4485444901511073,
      "epoch": 2.0402987906912555,
      "grad_norm": 0.1499614268541336,
      "learning_rate": 3.199313030944276e-05,
      "loss": 0.4396,
      "mean_token_accuracy": 0.8574726991355419,
      "num_tokens": 707748544.0,
      "step": 22176
    },
    {
      "entropy": 0.5850379317998886,
      "epoch": 2.0403907970626967,
      "grad_norm": 0.16595551371574402,
      "learning_rate": 3.199006348330114e-05,
      "loss": 0.5752,
      "mean_token_accuracy": 0.8178225867450237,
      "num_tokens": 707780714.0,
      "step": 22177
    },
    {
      "entropy": 0.5627095457166433,
      "epoch": 2.040482803434138,
      "grad_norm": 0.15847739577293396,
      "learning_rate": 3.198699665715951e-05,
      "loss": 0.5403,
      "mean_token_accuracy": 0.8310899585485458,
      "num_tokens": 707812843.0,
      "step": 22178
    },
    {
      "entropy": 0.5563127901405096,
      "epoch": 2.040574809805579,
      "grad_norm": 0.16180691123008728,
      "learning_rate": 3.198392983101788e-05,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8319970928132534,
      "num_tokens": 707844481.0,
      "step": 22179
    },
    {
      "entropy": 0.5721140224486589,
      "epoch": 2.04066681617702,
      "grad_norm": 0.16682536900043488,
      "learning_rate": 3.1980863004876256e-05,
      "loss": 0.5472,
      "mean_token_accuracy": 0.8249685056507587,
      "num_tokens": 707876224.0,
      "step": 22180
    },
    {
      "entropy": 0.6103148432448506,
      "epoch": 2.0407588225484616,
      "grad_norm": 0.16552715003490448,
      "learning_rate": 3.197779617873463e-05,
      "loss": 0.5906,
      "mean_token_accuracy": 0.814331229776144,
      "num_tokens": 707908610.0,
      "step": 22181
    },
    {
      "entropy": 0.6012690095230937,
      "epoch": 2.0408508289199028,
      "grad_norm": 0.16591954231262207,
      "learning_rate": 3.1974729352593006e-05,
      "loss": 0.5853,
      "mean_token_accuracy": 0.8135493919253349,
      "num_tokens": 707940666.0,
      "step": 22182
    },
    {
      "entropy": 0.5090197641402483,
      "epoch": 2.040942835291344,
      "grad_norm": 0.15951049327850342,
      "learning_rate": 3.197166252645138e-05,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8471256792545319,
      "num_tokens": 707972500.0,
      "step": 22183
    },
    {
      "entropy": 0.5115538910031319,
      "epoch": 2.041034841662785,
      "grad_norm": 0.1608581840991974,
      "learning_rate": 3.196859570030975e-05,
      "loss": 0.4786,
      "mean_token_accuracy": 0.8452767319977283,
      "num_tokens": 708002769.0,
      "step": 22184
    },
    {
      "entropy": 0.45862232707440853,
      "epoch": 2.041126848034226,
      "grad_norm": 0.15511326491832733,
      "learning_rate": 3.196552887416813e-05,
      "loss": 0.4449,
      "mean_token_accuracy": 0.8550747223198414,
      "num_tokens": 708034330.0,
      "step": 22185
    },
    {
      "entropy": 0.4885637201368809,
      "epoch": 2.0412188544056677,
      "grad_norm": 0.15452250838279724,
      "learning_rate": 3.19624620480265e-05,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8505696542561054,
      "num_tokens": 708066296.0,
      "step": 22186
    },
    {
      "entropy": 0.54736462328583,
      "epoch": 2.041310860777109,
      "grad_norm": 0.16433101892471313,
      "learning_rate": 3.195939522188487e-05,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8327346593141556,
      "num_tokens": 708097598.0,
      "step": 22187
    },
    {
      "entropy": 0.4921016830485314,
      "epoch": 2.04140286714855,
      "grad_norm": 0.16158674657344818,
      "learning_rate": 3.195632839574325e-05,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8515445813536644,
      "num_tokens": 708129519.0,
      "step": 22188
    },
    {
      "entropy": 0.6243175733834505,
      "epoch": 2.041494873519991,
      "grad_norm": 0.17337678372859955,
      "learning_rate": 3.195326156960162e-05,
      "loss": 0.6076,
      "mean_token_accuracy": 0.8095897436141968,
      "num_tokens": 708161182.0,
      "step": 22189
    },
    {
      "entropy": 0.5552289607003331,
      "epoch": 2.0415868798914323,
      "grad_norm": 0.16595503687858582,
      "learning_rate": 3.1950194743459996e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8282041661441326,
      "num_tokens": 708192661.0,
      "step": 22190
    },
    {
      "entropy": 0.5124632064253092,
      "epoch": 2.041678886262874,
      "grad_norm": 0.16341054439544678,
      "learning_rate": 3.194712791731837e-05,
      "loss": 0.5005,
      "mean_token_accuracy": 0.841601449996233,
      "num_tokens": 708223887.0,
      "step": 22191
    },
    {
      "entropy": 0.4026493774726987,
      "epoch": 2.041770892634315,
      "grad_norm": 0.14780457317829132,
      "learning_rate": 3.194406109117674e-05,
      "loss": 0.3872,
      "mean_token_accuracy": 0.87605020403862,
      "num_tokens": 708256385.0,
      "step": 22192
    },
    {
      "entropy": 0.5048577445559204,
      "epoch": 2.041862899005756,
      "grad_norm": 0.16854970157146454,
      "learning_rate": 3.194099426503512e-05,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8398290388286114,
      "num_tokens": 708287755.0,
      "step": 22193
    },
    {
      "entropy": 0.48911193665117025,
      "epoch": 2.0419549053771973,
      "grad_norm": 0.15258610248565674,
      "learning_rate": 3.193792743889349e-05,
      "loss": 0.4807,
      "mean_token_accuracy": 0.8471770770847797,
      "num_tokens": 708320101.0,
      "step": 22194
    },
    {
      "entropy": 0.5607772911898792,
      "epoch": 2.0420469117486384,
      "grad_norm": 0.15776793658733368,
      "learning_rate": 3.193486061275186e-05,
      "loss": 0.5423,
      "mean_token_accuracy": 0.8303309641778469,
      "num_tokens": 708352869.0,
      "step": 22195
    },
    {
      "entropy": 0.5388634819537401,
      "epoch": 2.04213891812008,
      "grad_norm": 0.16815996170043945,
      "learning_rate": 3.193179378661024e-05,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8344650827348232,
      "num_tokens": 708383500.0,
      "step": 22196
    },
    {
      "entropy": 0.48870265763252974,
      "epoch": 2.042230924491521,
      "grad_norm": 0.1540353149175644,
      "learning_rate": 3.192872696046861e-05,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8500804528594017,
      "num_tokens": 708415504.0,
      "step": 22197
    },
    {
      "entropy": 0.5261145066469908,
      "epoch": 2.0423229308629622,
      "grad_norm": 0.1624937653541565,
      "learning_rate": 3.1925660134326986e-05,
      "loss": 0.51,
      "mean_token_accuracy": 0.834819708019495,
      "num_tokens": 708447708.0,
      "step": 22198
    },
    {
      "entropy": 0.5502403993159533,
      "epoch": 2.0424149372344034,
      "grad_norm": 0.16149556636810303,
      "learning_rate": 3.192259330818536e-05,
      "loss": 0.538,
      "mean_token_accuracy": 0.832944069057703,
      "num_tokens": 708479512.0,
      "step": 22199
    },
    {
      "entropy": 0.4772357288748026,
      "epoch": 2.0425069436058445,
      "grad_norm": 0.1522989720106125,
      "learning_rate": 3.191952648204373e-05,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8507708832621574,
      "num_tokens": 708510561.0,
      "step": 22200
    },
    {
      "entropy": 0.5876251310110092,
      "epoch": 2.042598949977286,
      "grad_norm": 0.1685734987258911,
      "learning_rate": 3.191645965590211e-05,
      "loss": 0.5821,
      "mean_token_accuracy": 0.8154572024941444,
      "num_tokens": 708541993.0,
      "step": 22201
    },
    {
      "entropy": 0.5328443329781294,
      "epoch": 2.042690956348727,
      "grad_norm": 0.16073426604270935,
      "learning_rate": 3.191339282976048e-05,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8413706757128239,
      "num_tokens": 708573897.0,
      "step": 22202
    },
    {
      "entropy": 0.4387510959059,
      "epoch": 2.0427829627201683,
      "grad_norm": 0.15148837864398956,
      "learning_rate": 3.191032600361886e-05,
      "loss": 0.4279,
      "mean_token_accuracy": 0.8616329021751881,
      "num_tokens": 708606511.0,
      "step": 22203
    },
    {
      "entropy": 0.5877676913514733,
      "epoch": 2.0428749690916095,
      "grad_norm": 0.16806165874004364,
      "learning_rate": 3.190725917747723e-05,
      "loss": 0.5702,
      "mean_token_accuracy": 0.8196521773934364,
      "num_tokens": 708638801.0,
      "step": 22204
    },
    {
      "entropy": 0.5084701348096132,
      "epoch": 2.0429669754630506,
      "grad_norm": 0.16032859683036804,
      "learning_rate": 3.19041923513356e-05,
      "loss": 0.4944,
      "mean_token_accuracy": 0.8428276292979717,
      "num_tokens": 708671422.0,
      "step": 22205
    },
    {
      "entropy": 0.4833346717059612,
      "epoch": 2.043058981834492,
      "grad_norm": 0.15819990634918213,
      "learning_rate": 3.190112552519398e-05,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8513051718473434,
      "num_tokens": 708702593.0,
      "step": 22206
    },
    {
      "entropy": 0.5286937523633242,
      "epoch": 2.0431509882059333,
      "grad_norm": 0.16086506843566895,
      "learning_rate": 3.189805869905235e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8346605263650417,
      "num_tokens": 708734872.0,
      "step": 22207
    },
    {
      "entropy": 0.5474350107833743,
      "epoch": 2.0432429945773745,
      "grad_norm": 0.16610436141490936,
      "learning_rate": 3.1894991872910726e-05,
      "loss": 0.5301,
      "mean_token_accuracy": 0.8314799144864082,
      "num_tokens": 708766364.0,
      "step": 22208
    },
    {
      "entropy": 0.5185435054590926,
      "epoch": 2.0433350009488156,
      "grad_norm": 0.1590786576271057,
      "learning_rate": 3.18919250467691e-05,
      "loss": 0.5127,
      "mean_token_accuracy": 0.8398447223007679,
      "num_tokens": 708798712.0,
      "step": 22209
    },
    {
      "entropy": 0.4650513483211398,
      "epoch": 2.0434270073202567,
      "grad_norm": 0.15556694567203522,
      "learning_rate": 3.1888858220627475e-05,
      "loss": 0.461,
      "mean_token_accuracy": 0.8513540849089622,
      "num_tokens": 708830818.0,
      "step": 22210
    },
    {
      "entropy": 0.5666366219520569,
      "epoch": 2.0435190136916983,
      "grad_norm": 0.17013579607009888,
      "learning_rate": 3.188579139448585e-05,
      "loss": 0.5514,
      "mean_token_accuracy": 0.8244754076004028,
      "num_tokens": 708862295.0,
      "step": 22211
    },
    {
      "entropy": 0.5451010055840015,
      "epoch": 2.0436110200631394,
      "grad_norm": 0.1709476262331009,
      "learning_rate": 3.1882724568344225e-05,
      "loss": 0.5329,
      "mean_token_accuracy": 0.8287398852407932,
      "num_tokens": 708893263.0,
      "step": 22212
    },
    {
      "entropy": 0.4971720790490508,
      "epoch": 2.0437030264345806,
      "grad_norm": 0.15853410959243774,
      "learning_rate": 3.187965774220259e-05,
      "loss": 0.4839,
      "mean_token_accuracy": 0.8487725853919983,
      "num_tokens": 708926031.0,
      "step": 22213
    },
    {
      "entropy": 0.6194578632712364,
      "epoch": 2.0437950328060217,
      "grad_norm": 0.17468281090259552,
      "learning_rate": 3.1876590916060974e-05,
      "loss": 0.6189,
      "mean_token_accuracy": 0.8066319674253464,
      "num_tokens": 708958556.0,
      "step": 22214
    },
    {
      "entropy": 0.47709257528185844,
      "epoch": 2.043887039177463,
      "grad_norm": 0.15826216340065002,
      "learning_rate": 3.187352408991934e-05,
      "loss": 0.4647,
      "mean_token_accuracy": 0.8527910485863686,
      "num_tokens": 708990783.0,
      "step": 22215
    },
    {
      "entropy": 0.5374299006070942,
      "epoch": 2.0439790455489044,
      "grad_norm": 0.15609566867351532,
      "learning_rate": 3.1870457263777717e-05,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8350120075047016,
      "num_tokens": 709022782.0,
      "step": 22216
    },
    {
      "entropy": 0.5873697753995657,
      "epoch": 2.0440710519203455,
      "grad_norm": 0.1697993278503418,
      "learning_rate": 3.186739043763609e-05,
      "loss": 0.5602,
      "mean_token_accuracy": 0.8191472813487053,
      "num_tokens": 709054161.0,
      "step": 22217
    },
    {
      "entropy": 0.46202646056190133,
      "epoch": 2.0441630582917867,
      "grad_norm": 0.1506252884864807,
      "learning_rate": 3.1864323611494466e-05,
      "loss": 0.4402,
      "mean_token_accuracy": 0.860959816724062,
      "num_tokens": 709086234.0,
      "step": 22218
    },
    {
      "entropy": 0.49504910968244076,
      "epoch": 2.044255064663228,
      "grad_norm": 0.14923806488513947,
      "learning_rate": 3.186125678535284e-05,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8474649451673031,
      "num_tokens": 709118666.0,
      "step": 22219
    },
    {
      "entropy": 0.48453621100634336,
      "epoch": 2.044347071034669,
      "grad_norm": 0.1578630805015564,
      "learning_rate": 3.1858189959211215e-05,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8477958142757416,
      "num_tokens": 709149731.0,
      "step": 22220
    },
    {
      "entropy": 0.5522200558334589,
      "epoch": 2.0444390774061105,
      "grad_norm": 0.16409263014793396,
      "learning_rate": 3.185512313306959e-05,
      "loss": 0.5306,
      "mean_token_accuracy": 0.8338982090353966,
      "num_tokens": 709181184.0,
      "step": 22221
    },
    {
      "entropy": 0.5531588066369295,
      "epoch": 2.0445310837775517,
      "grad_norm": 0.16466441750526428,
      "learning_rate": 3.1852056306927964e-05,
      "loss": 0.5158,
      "mean_token_accuracy": 0.8341558836400509,
      "num_tokens": 709212407.0,
      "step": 22222
    },
    {
      "entropy": 0.583701394032687,
      "epoch": 2.044623090148993,
      "grad_norm": 0.16830186545848846,
      "learning_rate": 3.184898948078633e-05,
      "loss": 0.5841,
      "mean_token_accuracy": 0.8157384470105171,
      "num_tokens": 709244288.0,
      "step": 22223
    },
    {
      "entropy": 0.5259466243442148,
      "epoch": 2.044715096520434,
      "grad_norm": 0.15523575246334076,
      "learning_rate": 3.1845922654644714e-05,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8410928957164288,
      "num_tokens": 709276910.0,
      "step": 22224
    },
    {
      "entropy": 0.5616196258924901,
      "epoch": 2.044807102891875,
      "grad_norm": 0.16289077699184418,
      "learning_rate": 3.184285582850308e-05,
      "loss": 0.5486,
      "mean_token_accuracy": 0.824367094784975,
      "num_tokens": 709308306.0,
      "step": 22225
    },
    {
      "entropy": 0.48710297886282206,
      "epoch": 2.0448991092633166,
      "grad_norm": 0.1593596637248993,
      "learning_rate": 3.1839789002361456e-05,
      "loss": 0.4812,
      "mean_token_accuracy": 0.8460418358445168,
      "num_tokens": 709340208.0,
      "step": 22226
    },
    {
      "entropy": 0.45811999775469303,
      "epoch": 2.0449911156347578,
      "grad_norm": 0.15407045185565948,
      "learning_rate": 3.183672217621983e-05,
      "loss": 0.4568,
      "mean_token_accuracy": 0.8560984022915363,
      "num_tokens": 709371731.0,
      "step": 22227
    },
    {
      "entropy": 0.510600745677948,
      "epoch": 2.045083122006199,
      "grad_norm": 0.16376352310180664,
      "learning_rate": 3.1833655350078206e-05,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8403412587940693,
      "num_tokens": 709403169.0,
      "step": 22228
    },
    {
      "entropy": 0.6058903532102704,
      "epoch": 2.04517512837764,
      "grad_norm": 0.1676870733499527,
      "learning_rate": 3.183058852393658e-05,
      "loss": 0.5869,
      "mean_token_accuracy": 0.8109122700989246,
      "num_tokens": 709434388.0,
      "step": 22229
    },
    {
      "entropy": 0.5134441475383937,
      "epoch": 2.045267134749081,
      "grad_norm": 0.15960560739040375,
      "learning_rate": 3.1827521697794955e-05,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8401676677167416,
      "num_tokens": 709466638.0,
      "step": 22230
    },
    {
      "entropy": 0.5457140686921775,
      "epoch": 2.0453591411205228,
      "grad_norm": 0.16172367334365845,
      "learning_rate": 3.182445487165332e-05,
      "loss": 0.5277,
      "mean_token_accuracy": 0.8305648043751717,
      "num_tokens": 709498355.0,
      "step": 22231
    },
    {
      "entropy": 0.5046292888000607,
      "epoch": 2.045451147491964,
      "grad_norm": 0.15963555872440338,
      "learning_rate": 3.1821388045511704e-05,
      "loss": 0.4856,
      "mean_token_accuracy": 0.8449274636805058,
      "num_tokens": 709529735.0,
      "step": 22232
    },
    {
      "entropy": 0.49913851264864206,
      "epoch": 2.045543153863405,
      "grad_norm": 0.15545670688152313,
      "learning_rate": 3.181832121937007e-05,
      "loss": 0.4898,
      "mean_token_accuracy": 0.8430145978927612,
      "num_tokens": 709561754.0,
      "step": 22233
    },
    {
      "entropy": 0.5462800837121904,
      "epoch": 2.045635160234846,
      "grad_norm": 0.164767324924469,
      "learning_rate": 3.181525439322845e-05,
      "loss": 0.528,
      "mean_token_accuracy": 0.8366005085408688,
      "num_tokens": 709593432.0,
      "step": 22234
    },
    {
      "entropy": 0.5685871299356222,
      "epoch": 2.0457271666062873,
      "grad_norm": 0.16620467603206635,
      "learning_rate": 3.181218756708682e-05,
      "loss": 0.5598,
      "mean_token_accuracy": 0.8252325430512428,
      "num_tokens": 709625424.0,
      "step": 22235
    },
    {
      "entropy": 0.5835146782919765,
      "epoch": 2.045819172977729,
      "grad_norm": 0.17155449092388153,
      "learning_rate": 3.1809120740945196e-05,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8235369734466076,
      "num_tokens": 709657309.0,
      "step": 22236
    },
    {
      "entropy": 0.5666239708662033,
      "epoch": 2.04591117934917,
      "grad_norm": 0.16603752970695496,
      "learning_rate": 3.180605391480357e-05,
      "loss": 0.5535,
      "mean_token_accuracy": 0.8226149864494801,
      "num_tokens": 709689235.0,
      "step": 22237
    },
    {
      "entropy": 0.5746387569233775,
      "epoch": 2.046003185720611,
      "grad_norm": 0.1664511114358902,
      "learning_rate": 3.1802987088661945e-05,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8280641622841358,
      "num_tokens": 709720574.0,
      "step": 22238
    },
    {
      "entropy": 0.5426457761786878,
      "epoch": 2.0460951920920523,
      "grad_norm": 0.1698056012392044,
      "learning_rate": 3.179992026252032e-05,
      "loss": 0.535,
      "mean_token_accuracy": 0.8299527764320374,
      "num_tokens": 709752301.0,
      "step": 22239
    },
    {
      "entropy": 0.49552370980381966,
      "epoch": 2.0461871984634934,
      "grad_norm": 0.1615206003189087,
      "learning_rate": 3.1796853436378695e-05,
      "loss": 0.4914,
      "mean_token_accuracy": 0.8434023782610893,
      "num_tokens": 709783393.0,
      "step": 22240
    },
    {
      "entropy": 0.49771497026085854,
      "epoch": 2.046279204834935,
      "grad_norm": 0.15957927703857422,
      "learning_rate": 3.179378661023707e-05,
      "loss": 0.482,
      "mean_token_accuracy": 0.8475357182323933,
      "num_tokens": 709815694.0,
      "step": 22241
    },
    {
      "entropy": 0.6138714868575335,
      "epoch": 2.046371211206376,
      "grad_norm": 0.17092275619506836,
      "learning_rate": 3.1790719784095444e-05,
      "loss": 0.604,
      "mean_token_accuracy": 0.8103879652917385,
      "num_tokens": 709847720.0,
      "step": 22242
    },
    {
      "entropy": 0.6253672391176224,
      "epoch": 2.0464632175778172,
      "grad_norm": 0.18248598277568817,
      "learning_rate": 3.178765295795382e-05,
      "loss": 0.6314,
      "mean_token_accuracy": 0.8071197792887688,
      "num_tokens": 709878628.0,
      "step": 22243
    },
    {
      "entropy": 0.5648043733090162,
      "epoch": 2.0465552239492584,
      "grad_norm": 0.1664409637451172,
      "learning_rate": 3.1784586131812186e-05,
      "loss": 0.5457,
      "mean_token_accuracy": 0.827485166490078,
      "num_tokens": 709909951.0,
      "step": 22244
    },
    {
      "entropy": 0.5888278353959322,
      "epoch": 2.0466472303206995,
      "grad_norm": 0.16544140875339508,
      "learning_rate": 3.178151930567057e-05,
      "loss": 0.5631,
      "mean_token_accuracy": 0.8199226558208466,
      "num_tokens": 709941840.0,
      "step": 22245
    },
    {
      "entropy": 0.5317843626253307,
      "epoch": 2.046739236692141,
      "grad_norm": 0.16136445105075836,
      "learning_rate": 3.1778452479528936e-05,
      "loss": 0.5137,
      "mean_token_accuracy": 0.8368305787444115,
      "num_tokens": 709973744.0,
      "step": 22246
    },
    {
      "entropy": 0.44675566628575325,
      "epoch": 2.046831243063582,
      "grad_norm": 0.15191791951656342,
      "learning_rate": 3.177538565338731e-05,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8613245934247971,
      "num_tokens": 710006333.0,
      "step": 22247
    },
    {
      "entropy": 0.6149794533848763,
      "epoch": 2.0469232494350234,
      "grad_norm": 0.16860587894916534,
      "learning_rate": 3.1772318827245685e-05,
      "loss": 0.598,
      "mean_token_accuracy": 0.8107277266681194,
      "num_tokens": 710038247.0,
      "step": 22248
    },
    {
      "entropy": 0.45677989441901445,
      "epoch": 2.0470152558064645,
      "grad_norm": 0.15362675487995148,
      "learning_rate": 3.176925200110406e-05,
      "loss": 0.4372,
      "mean_token_accuracy": 0.8607606738805771,
      "num_tokens": 710070591.0,
      "step": 22249
    },
    {
      "entropy": 0.541802741587162,
      "epoch": 2.0471072621779056,
      "grad_norm": 0.15734359622001648,
      "learning_rate": 3.1766185174962434e-05,
      "loss": 0.515,
      "mean_token_accuracy": 0.8354449234902859,
      "num_tokens": 710102322.0,
      "step": 22250
    },
    {
      "entropy": 0.48258778639137745,
      "epoch": 2.047199268549347,
      "grad_norm": 0.15493342280387878,
      "learning_rate": 3.176311834882081e-05,
      "loss": 0.4641,
      "mean_token_accuracy": 0.8529878407716751,
      "num_tokens": 710132997.0,
      "step": 22251
    },
    {
      "entropy": 0.4897809661924839,
      "epoch": 2.0472912749207883,
      "grad_norm": 0.1546148657798767,
      "learning_rate": 3.176005152267918e-05,
      "loss": 0.478,
      "mean_token_accuracy": 0.8478518910706043,
      "num_tokens": 710164235.0,
      "step": 22252
    },
    {
      "entropy": 0.6083122696727514,
      "epoch": 2.0473832812922295,
      "grad_norm": 0.16558252274990082,
      "learning_rate": 3.175698469653756e-05,
      "loss": 0.5733,
      "mean_token_accuracy": 0.8147445321083069,
      "num_tokens": 710196202.0,
      "step": 22253
    },
    {
      "entropy": 0.5217137429863214,
      "epoch": 2.0474752876636706,
      "grad_norm": 0.16595926880836487,
      "learning_rate": 3.1753917870395926e-05,
      "loss": 0.5224,
      "mean_token_accuracy": 0.835174236446619,
      "num_tokens": 710227908.0,
      "step": 22254
    },
    {
      "entropy": 0.5938577251508832,
      "epoch": 2.0475672940351117,
      "grad_norm": 0.17013731598854065,
      "learning_rate": 3.17508510442543e-05,
      "loss": 0.5862,
      "mean_token_accuracy": 0.8146470636129379,
      "num_tokens": 710258916.0,
      "step": 22255
    },
    {
      "entropy": 0.49933182867243886,
      "epoch": 2.0476593004065533,
      "grad_norm": 0.15321741998195648,
      "learning_rate": 3.1747784218112675e-05,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8472789376974106,
      "num_tokens": 710291469.0,
      "step": 22256
    },
    {
      "entropy": 0.5745809487998486,
      "epoch": 2.0477513067779944,
      "grad_norm": 0.17473195493221283,
      "learning_rate": 3.174471739197105e-05,
      "loss": 0.5601,
      "mean_token_accuracy": 0.8249859809875488,
      "num_tokens": 710323378.0,
      "step": 22257
    },
    {
      "entropy": 0.4612096194177866,
      "epoch": 2.0478433131494356,
      "grad_norm": 0.1551489233970642,
      "learning_rate": 3.1741650565829425e-05,
      "loss": 0.4475,
      "mean_token_accuracy": 0.8545063957571983,
      "num_tokens": 710355711.0,
      "step": 22258
    },
    {
      "entropy": 0.46514591481536627,
      "epoch": 2.0479353195208767,
      "grad_norm": 0.1565844863653183,
      "learning_rate": 3.17385837396878e-05,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8558737374842167,
      "num_tokens": 710387478.0,
      "step": 22259
    },
    {
      "entropy": 0.5523453373461962,
      "epoch": 2.048027325892318,
      "grad_norm": 0.1712452471256256,
      "learning_rate": 3.173551691354617e-05,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8292102813720703,
      "num_tokens": 710418608.0,
      "step": 22260
    },
    {
      "entropy": 0.5124376285821199,
      "epoch": 2.0481193322637594,
      "grad_norm": 0.15877746045589447,
      "learning_rate": 3.173245008740455e-05,
      "loss": 0.5093,
      "mean_token_accuracy": 0.8403433561325073,
      "num_tokens": 710450966.0,
      "step": 22261
    },
    {
      "entropy": 0.378218119032681,
      "epoch": 2.0482113386352006,
      "grad_norm": 0.15270262956619263,
      "learning_rate": 3.1729383261262916e-05,
      "loss": 0.3669,
      "mean_token_accuracy": 0.8818269707262516,
      "num_tokens": 710482644.0,
      "step": 22262
    },
    {
      "entropy": 0.49639786314219236,
      "epoch": 2.0483033450066417,
      "grad_norm": 0.16220217943191528,
      "learning_rate": 3.17263164351213e-05,
      "loss": 0.4831,
      "mean_token_accuracy": 0.8467799350619316,
      "num_tokens": 710514443.0,
      "step": 22263
    },
    {
      "entropy": 0.5626691300421953,
      "epoch": 2.048395351378083,
      "grad_norm": 0.16743947565555573,
      "learning_rate": 3.1723249608979666e-05,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8263086043298244,
      "num_tokens": 710546041.0,
      "step": 22264
    },
    {
      "entropy": 0.5101697323843837,
      "epoch": 2.048487357749524,
      "grad_norm": 0.1652420163154602,
      "learning_rate": 3.172018278283804e-05,
      "loss": 0.499,
      "mean_token_accuracy": 0.8395526595413685,
      "num_tokens": 710576861.0,
      "step": 22265
    },
    {
      "entropy": 0.6149094384163618,
      "epoch": 2.0485793641209655,
      "grad_norm": 0.1711847186088562,
      "learning_rate": 3.1717115956696415e-05,
      "loss": 0.608,
      "mean_token_accuracy": 0.8098741732537746,
      "num_tokens": 710608779.0,
      "step": 22266
    },
    {
      "entropy": 0.4721071347594261,
      "epoch": 2.0486713704924067,
      "grad_norm": 0.1586814969778061,
      "learning_rate": 3.171404913055479e-05,
      "loss": 0.4552,
      "mean_token_accuracy": 0.8568858616054058,
      "num_tokens": 710640779.0,
      "step": 22267
    },
    {
      "entropy": 0.46182132698595524,
      "epoch": 2.048763376863848,
      "grad_norm": 0.1538359373807907,
      "learning_rate": 3.1710982304413164e-05,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8590615279972553,
      "num_tokens": 710673091.0,
      "step": 22268
    },
    {
      "entropy": 0.5679783839732409,
      "epoch": 2.048855383235289,
      "grad_norm": 0.1608426868915558,
      "learning_rate": 3.170791547827154e-05,
      "loss": 0.5496,
      "mean_token_accuracy": 0.8277580887079239,
      "num_tokens": 710704662.0,
      "step": 22269
    },
    {
      "entropy": 0.46820810995996,
      "epoch": 2.04894738960673,
      "grad_norm": 0.15091559290885925,
      "learning_rate": 3.1704848652129914e-05,
      "loss": 0.45,
      "mean_token_accuracy": 0.8579231016337872,
      "num_tokens": 710736553.0,
      "step": 22270
    },
    {
      "entropy": 0.4771493664011359,
      "epoch": 2.0490393959781716,
      "grad_norm": 0.15236487984657288,
      "learning_rate": 3.170178182598829e-05,
      "loss": 0.4588,
      "mean_token_accuracy": 0.8540616743266582,
      "num_tokens": 710768244.0,
      "step": 22271
    },
    {
      "entropy": 0.5763185434043407,
      "epoch": 2.0491314023496128,
      "grad_norm": 0.16507534682750702,
      "learning_rate": 3.169871499984666e-05,
      "loss": 0.5482,
      "mean_token_accuracy": 0.825927123427391,
      "num_tokens": 710799506.0,
      "step": 22272
    },
    {
      "entropy": 0.44168843887746334,
      "epoch": 2.049223408721054,
      "grad_norm": 0.15127356350421906,
      "learning_rate": 3.169564817370503e-05,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8644929304718971,
      "num_tokens": 710831939.0,
      "step": 22273
    },
    {
      "entropy": 0.4949581753462553,
      "epoch": 2.049315415092495,
      "grad_norm": 0.15782445669174194,
      "learning_rate": 3.169258134756341e-05,
      "loss": 0.482,
      "mean_token_accuracy": 0.8484559692442417,
      "num_tokens": 710864003.0,
      "step": 22274
    },
    {
      "entropy": 0.5420243265107274,
      "epoch": 2.049407421463936,
      "grad_norm": 0.16120050847530365,
      "learning_rate": 3.168951452142178e-05,
      "loss": 0.5402,
      "mean_token_accuracy": 0.8326166532933712,
      "num_tokens": 710895912.0,
      "step": 22275
    },
    {
      "entropy": 0.5107416762039065,
      "epoch": 2.0494994278353778,
      "grad_norm": 0.15638601779937744,
      "learning_rate": 3.1686447695280155e-05,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8409953229129314,
      "num_tokens": 710927835.0,
      "step": 22276
    },
    {
      "entropy": 0.4634348629042506,
      "epoch": 2.049591434206819,
      "grad_norm": 0.152720108628273,
      "learning_rate": 3.168338086913853e-05,
      "loss": 0.452,
      "mean_token_accuracy": 0.8546582646667957,
      "num_tokens": 710960100.0,
      "step": 22277
    },
    {
      "entropy": 0.5225980626419187,
      "epoch": 2.04968344057826,
      "grad_norm": 0.15674379467964172,
      "learning_rate": 3.1680314042996904e-05,
      "loss": 0.4927,
      "mean_token_accuracy": 0.8408317305147648,
      "num_tokens": 710991479.0,
      "step": 22278
    },
    {
      "entropy": 0.5462953047826886,
      "epoch": 2.049775446949701,
      "grad_norm": 0.1618843376636505,
      "learning_rate": 3.167724721685528e-05,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8341438621282578,
      "num_tokens": 711022795.0,
      "step": 22279
    },
    {
      "entropy": 0.5255573689937592,
      "epoch": 2.0498674533211423,
      "grad_norm": 0.16602924466133118,
      "learning_rate": 3.167418039071365e-05,
      "loss": 0.5069,
      "mean_token_accuracy": 0.8389909751713276,
      "num_tokens": 711054501.0,
      "step": 22280
    },
    {
      "entropy": 0.49766627326607704,
      "epoch": 2.049959459692584,
      "grad_norm": 0.16017282009124756,
      "learning_rate": 3.167111356457203e-05,
      "loss": 0.4819,
      "mean_token_accuracy": 0.8444238491356373,
      "num_tokens": 711086507.0,
      "step": 22281
    },
    {
      "entropy": 0.5688009411096573,
      "epoch": 2.050051466064025,
      "grad_norm": 0.16125838458538055,
      "learning_rate": 3.16680467384304e-05,
      "loss": 0.5463,
      "mean_token_accuracy": 0.8265541568398476,
      "num_tokens": 711118149.0,
      "step": 22282
    },
    {
      "entropy": 0.4965203898027539,
      "epoch": 2.050143472435466,
      "grad_norm": 0.15975289046764374,
      "learning_rate": 3.166497991228877e-05,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8486460000276566,
      "num_tokens": 711150020.0,
      "step": 22283
    },
    {
      "entropy": 0.47037561796605587,
      "epoch": 2.0502354788069073,
      "grad_norm": 0.15020105242729187,
      "learning_rate": 3.166191308614715e-05,
      "loss": 0.449,
      "mean_token_accuracy": 0.8549406081438065,
      "num_tokens": 711182425.0,
      "step": 22284
    },
    {
      "entropy": 0.5087796547450125,
      "epoch": 2.0503274851783484,
      "grad_norm": 0.16116270422935486,
      "learning_rate": 3.165884626000552e-05,
      "loss": 0.5145,
      "mean_token_accuracy": 0.8422965370118618,
      "num_tokens": 711213831.0,
      "step": 22285
    },
    {
      "entropy": 0.5224776081740856,
      "epoch": 2.05041949154979,
      "grad_norm": 0.16110599040985107,
      "learning_rate": 3.1655779433863894e-05,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8400426991283894,
      "num_tokens": 711246534.0,
      "step": 22286
    },
    {
      "entropy": 0.45310133416205645,
      "epoch": 2.050511497921231,
      "grad_norm": 0.14958839118480682,
      "learning_rate": 3.165271260772227e-05,
      "loss": 0.438,
      "mean_token_accuracy": 0.8587298169732094,
      "num_tokens": 711279152.0,
      "step": 22287
    },
    {
      "entropy": 0.536817442625761,
      "epoch": 2.0506035042926722,
      "grad_norm": 0.16040430963039398,
      "learning_rate": 3.1649645781580644e-05,
      "loss": 0.52,
      "mean_token_accuracy": 0.8379361480474472,
      "num_tokens": 711311356.0,
      "step": 22288
    },
    {
      "entropy": 0.522719400934875,
      "epoch": 2.0506955106641134,
      "grad_norm": 0.1641213595867157,
      "learning_rate": 3.164657895543902e-05,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8385762944817543,
      "num_tokens": 711342947.0,
      "step": 22289
    },
    {
      "entropy": 0.6005381802096963,
      "epoch": 2.0507875170355545,
      "grad_norm": 0.17304053902626038,
      "learning_rate": 3.164351212929739e-05,
      "loss": 0.5894,
      "mean_token_accuracy": 0.8131558410823345,
      "num_tokens": 711375105.0,
      "step": 22290
    },
    {
      "entropy": 0.5086303474381566,
      "epoch": 2.050879523406996,
      "grad_norm": 0.1633520871400833,
      "learning_rate": 3.164044530315576e-05,
      "loss": 0.4962,
      "mean_token_accuracy": 0.8435581810772419,
      "num_tokens": 711407729.0,
      "step": 22291
    },
    {
      "entropy": 0.5030010659247637,
      "epoch": 2.050971529778437,
      "grad_norm": 0.15969251096248627,
      "learning_rate": 3.163737847701414e-05,
      "loss": 0.4872,
      "mean_token_accuracy": 0.845580093562603,
      "num_tokens": 711438623.0,
      "step": 22292
    },
    {
      "entropy": 0.6143827363848686,
      "epoch": 2.0510635361498784,
      "grad_norm": 0.17163696885108948,
      "learning_rate": 3.163431165087251e-05,
      "loss": 0.6,
      "mean_token_accuracy": 0.8079275116324425,
      "num_tokens": 711471104.0,
      "step": 22293
    },
    {
      "entropy": 0.4742765184491873,
      "epoch": 2.0511555425213195,
      "grad_norm": 0.15868854522705078,
      "learning_rate": 3.1631244824730885e-05,
      "loss": 0.4731,
      "mean_token_accuracy": 0.8488551266491413,
      "num_tokens": 711503279.0,
      "step": 22294
    },
    {
      "entropy": 0.5636255387216806,
      "epoch": 2.0512475488927606,
      "grad_norm": 0.1596178114414215,
      "learning_rate": 3.162817799858926e-05,
      "loss": 0.5403,
      "mean_token_accuracy": 0.8271634429693222,
      "num_tokens": 711534767.0,
      "step": 22295
    },
    {
      "entropy": 0.4809952531941235,
      "epoch": 2.051339555264202,
      "grad_norm": 0.15068498253822327,
      "learning_rate": 3.1625111172447634e-05,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8531006164848804,
      "num_tokens": 711566433.0,
      "step": 22296
    },
    {
      "entropy": 0.4478222345933318,
      "epoch": 2.0514315616356433,
      "grad_norm": 0.14522668719291687,
      "learning_rate": 3.162204434630601e-05,
      "loss": 0.4274,
      "mean_token_accuracy": 0.8621592372655869,
      "num_tokens": 711598667.0,
      "step": 22297
    },
    {
      "entropy": 0.44296251609921455,
      "epoch": 2.0515235680070845,
      "grad_norm": 0.1496846228837967,
      "learning_rate": 3.1618977520164383e-05,
      "loss": 0.4317,
      "mean_token_accuracy": 0.8617262281477451,
      "num_tokens": 711631139.0,
      "step": 22298
    },
    {
      "entropy": 0.5011398531496525,
      "epoch": 2.0516155743785256,
      "grad_norm": 0.1570652425289154,
      "learning_rate": 3.161591069402276e-05,
      "loss": 0.48,
      "mean_token_accuracy": 0.8433900289237499,
      "num_tokens": 711661931.0,
      "step": 22299
    },
    {
      "entropy": 0.5323756886646152,
      "epoch": 2.0517075807499667,
      "grad_norm": 0.1634872853755951,
      "learning_rate": 3.161284386788113e-05,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8332018852233887,
      "num_tokens": 711693135.0,
      "step": 22300
    },
    {
      "entropy": 0.4950173795223236,
      "epoch": 2.0517995871214083,
      "grad_norm": 0.15384873747825623,
      "learning_rate": 3.160977704173951e-05,
      "loss": 0.4832,
      "mean_token_accuracy": 0.849608413875103,
      "num_tokens": 711725368.0,
      "step": 22301
    },
    {
      "entropy": 0.4926331308670342,
      "epoch": 2.0518915934928494,
      "grad_norm": 0.15786530077457428,
      "learning_rate": 3.160671021559788e-05,
      "loss": 0.491,
      "mean_token_accuracy": 0.8450788408517838,
      "num_tokens": 711757769.0,
      "step": 22302
    },
    {
      "entropy": 0.607310039922595,
      "epoch": 2.0519835998642906,
      "grad_norm": 0.17147782444953918,
      "learning_rate": 3.160364338945626e-05,
      "loss": 0.5918,
      "mean_token_accuracy": 0.8123137317597866,
      "num_tokens": 711789543.0,
      "step": 22303
    },
    {
      "entropy": 0.5290580419823527,
      "epoch": 2.0520756062357317,
      "grad_norm": 0.16156558692455292,
      "learning_rate": 3.1600576563314625e-05,
      "loss": 0.5088,
      "mean_token_accuracy": 0.8342626839876175,
      "num_tokens": 711821063.0,
      "step": 22304
    },
    {
      "entropy": 0.5672480908688158,
      "epoch": 2.052167612607173,
      "grad_norm": 0.16115295886993408,
      "learning_rate": 3.1597509737173006e-05,
      "loss": 0.5438,
      "mean_token_accuracy": 0.8294797763228416,
      "num_tokens": 711853186.0,
      "step": 22305
    },
    {
      "entropy": 0.5928765516728163,
      "epoch": 2.0522596189786144,
      "grad_norm": 0.1653222143650055,
      "learning_rate": 3.1594442911031374e-05,
      "loss": 0.5747,
      "mean_token_accuracy": 0.817200567573309,
      "num_tokens": 711885857.0,
      "step": 22306
    },
    {
      "entropy": 0.45726071763783693,
      "epoch": 2.0523516253500556,
      "grad_norm": 0.15694838762283325,
      "learning_rate": 3.159137608488975e-05,
      "loss": 0.4423,
      "mean_token_accuracy": 0.8587237671017647,
      "num_tokens": 711917823.0,
      "step": 22307
    },
    {
      "entropy": 0.5671656765043736,
      "epoch": 2.0524436317214967,
      "grad_norm": 0.16517989337444305,
      "learning_rate": 3.158830925874812e-05,
      "loss": 0.5424,
      "mean_token_accuracy": 0.8294703178107738,
      "num_tokens": 711949583.0,
      "step": 22308
    },
    {
      "entropy": 0.7206710372120142,
      "epoch": 2.052535638092938,
      "grad_norm": 0.1846262663602829,
      "learning_rate": 3.15852424326065e-05,
      "loss": 0.7071,
      "mean_token_accuracy": 0.7820555008947849,
      "num_tokens": 711980235.0,
      "step": 22309
    },
    {
      "entropy": 0.5267563834786415,
      "epoch": 2.052627644464379,
      "grad_norm": 0.16183239221572876,
      "learning_rate": 3.158217560646487e-05,
      "loss": 0.5179,
      "mean_token_accuracy": 0.8352445214986801,
      "num_tokens": 712012995.0,
      "step": 22310
    },
    {
      "entropy": 0.5862956522032619,
      "epoch": 2.0527196508358205,
      "grad_norm": 0.16359464824199677,
      "learning_rate": 3.157910878032325e-05,
      "loss": 0.576,
      "mean_token_accuracy": 0.8207691535353661,
      "num_tokens": 712045234.0,
      "step": 22311
    },
    {
      "entropy": 0.5126848667860031,
      "epoch": 2.0528116572072617,
      "grad_norm": 0.16067172586917877,
      "learning_rate": 3.1576041954181615e-05,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8406009711325169,
      "num_tokens": 712077247.0,
      "step": 22312
    },
    {
      "entropy": 0.5373583976179361,
      "epoch": 2.052903663578703,
      "grad_norm": 0.1630997359752655,
      "learning_rate": 3.1572975128039996e-05,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8339725881814957,
      "num_tokens": 712109513.0,
      "step": 22313
    },
    {
      "entropy": 0.5197868533432484,
      "epoch": 2.052995669950144,
      "grad_norm": 0.1610727310180664,
      "learning_rate": 3.1569908301898364e-05,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8369336538016796,
      "num_tokens": 712140697.0,
      "step": 22314
    },
    {
      "entropy": 0.513088465668261,
      "epoch": 2.053087676321585,
      "grad_norm": 0.16151848435401917,
      "learning_rate": 3.156684147575674e-05,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8405912108719349,
      "num_tokens": 712171754.0,
      "step": 22315
    },
    {
      "entropy": 0.4552550255320966,
      "epoch": 2.0531796826930266,
      "grad_norm": 0.15031741559505463,
      "learning_rate": 3.1563774649615114e-05,
      "loss": 0.4394,
      "mean_token_accuracy": 0.8596092872321606,
      "num_tokens": 712204271.0,
      "step": 22316
    },
    {
      "entropy": 0.4911436401307583,
      "epoch": 2.0532716890644678,
      "grad_norm": 0.15942201018333435,
      "learning_rate": 3.156070782347349e-05,
      "loss": 0.4731,
      "mean_token_accuracy": 0.8471367843449116,
      "num_tokens": 712235921.0,
      "step": 22317
    },
    {
      "entropy": 0.49578865244984627,
      "epoch": 2.053363695435909,
      "grad_norm": 0.15811485052108765,
      "learning_rate": 3.155764099733186e-05,
      "loss": 0.4785,
      "mean_token_accuracy": 0.8475617915391922,
      "num_tokens": 712268346.0,
      "step": 22318
    },
    {
      "entropy": 0.5398330790922046,
      "epoch": 2.05345570180735,
      "grad_norm": 0.1559225171804428,
      "learning_rate": 3.155457417119024e-05,
      "loss": 0.525,
      "mean_token_accuracy": 0.8380568027496338,
      "num_tokens": 712300114.0,
      "step": 22319
    },
    {
      "entropy": 0.4730836683884263,
      "epoch": 2.053547708178791,
      "grad_norm": 0.15761798620224,
      "learning_rate": 3.1551507345048605e-05,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8554722741246223,
      "num_tokens": 712331746.0,
      "step": 22320
    },
    {
      "entropy": 0.5107381641864777,
      "epoch": 2.0536397145502328,
      "grad_norm": 0.15892736613750458,
      "learning_rate": 3.154844051890699e-05,
      "loss": 0.496,
      "mean_token_accuracy": 0.8405498638749123,
      "num_tokens": 712363818.0,
      "step": 22321
    },
    {
      "entropy": 0.5322193559259176,
      "epoch": 2.053731720921674,
      "grad_norm": 0.16281524300575256,
      "learning_rate": 3.1545373692765355e-05,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8364743925631046,
      "num_tokens": 712396280.0,
      "step": 22322
    },
    {
      "entropy": 0.6851741299033165,
      "epoch": 2.053823727293115,
      "grad_norm": 0.18293538689613342,
      "learning_rate": 3.1542306866623736e-05,
      "loss": 0.6666,
      "mean_token_accuracy": 0.7903807125985622,
      "num_tokens": 712426777.0,
      "step": 22323
    },
    {
      "entropy": 0.5223727915436029,
      "epoch": 2.053915733664556,
      "grad_norm": 0.1626758724451065,
      "learning_rate": 3.1539240040482104e-05,
      "loss": 0.5081,
      "mean_token_accuracy": 0.8389591351151466,
      "num_tokens": 712458743.0,
      "step": 22324
    },
    {
      "entropy": 0.5952751412987709,
      "epoch": 2.0540077400359973,
      "grad_norm": 0.16225068271160126,
      "learning_rate": 3.153617321434048e-05,
      "loss": 0.5733,
      "mean_token_accuracy": 0.8199800178408623,
      "num_tokens": 712491096.0,
      "step": 22325
    },
    {
      "entropy": 0.5596644207835197,
      "epoch": 2.054099746407439,
      "grad_norm": 0.1685764491558075,
      "learning_rate": 3.153310638819885e-05,
      "loss": 0.5455,
      "mean_token_accuracy": 0.8272288478910923,
      "num_tokens": 712522985.0,
      "step": 22326
    },
    {
      "entropy": 0.41446660552173853,
      "epoch": 2.05419175277888,
      "grad_norm": 0.1543838381767273,
      "learning_rate": 3.153003956205723e-05,
      "loss": 0.4004,
      "mean_token_accuracy": 0.8717163912951946,
      "num_tokens": 712554863.0,
      "step": 22327
    },
    {
      "entropy": 0.4979524053633213,
      "epoch": 2.054283759150321,
      "grad_norm": 0.15688107907772064,
      "learning_rate": 3.15269727359156e-05,
      "loss": 0.4846,
      "mean_token_accuracy": 0.845314547419548,
      "num_tokens": 712586423.0,
      "step": 22328
    },
    {
      "entropy": 0.5027128728106618,
      "epoch": 2.0543757655217623,
      "grad_norm": 0.15571987628936768,
      "learning_rate": 3.152390590977398e-05,
      "loss": 0.4842,
      "mean_token_accuracy": 0.8475375659763813,
      "num_tokens": 712618633.0,
      "step": 22329
    },
    {
      "entropy": 0.5714610749855638,
      "epoch": 2.0544677718932034,
      "grad_norm": 0.16602808237075806,
      "learning_rate": 3.152083908363235e-05,
      "loss": 0.5525,
      "mean_token_accuracy": 0.8251377679407597,
      "num_tokens": 712650250.0,
      "step": 22330
    },
    {
      "entropy": 0.5174584444612265,
      "epoch": 2.054559778264645,
      "grad_norm": 0.16230623424053192,
      "learning_rate": 3.1517772257490726e-05,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8402541242539883,
      "num_tokens": 712681772.0,
      "step": 22331
    },
    {
      "entropy": 0.4975507794879377,
      "epoch": 2.054651784636086,
      "grad_norm": 0.15283501148223877,
      "learning_rate": 3.15147054313491e-05,
      "loss": 0.4898,
      "mean_token_accuracy": 0.8465993590652943,
      "num_tokens": 712714403.0,
      "step": 22332
    },
    {
      "entropy": 0.5185057520866394,
      "epoch": 2.0547437910075272,
      "grad_norm": 0.1575172245502472,
      "learning_rate": 3.151163860520747e-05,
      "loss": 0.5101,
      "mean_token_accuracy": 0.83865150436759,
      "num_tokens": 712745139.0,
      "step": 22333
    },
    {
      "entropy": 0.46766557544469833,
      "epoch": 2.0548357973789684,
      "grad_norm": 0.15869934856891632,
      "learning_rate": 3.150857177906585e-05,
      "loss": 0.4619,
      "mean_token_accuracy": 0.8519315756857395,
      "num_tokens": 712777132.0,
      "step": 22334
    },
    {
      "entropy": 0.46320054261013865,
      "epoch": 2.0549278037504095,
      "grad_norm": 0.15200389921665192,
      "learning_rate": 3.150550495292422e-05,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8527575097978115,
      "num_tokens": 712808852.0,
      "step": 22335
    },
    {
      "entropy": 0.5411114390008152,
      "epoch": 2.055019810121851,
      "grad_norm": 0.15988102555274963,
      "learning_rate": 3.150243812678259e-05,
      "loss": 0.5311,
      "mean_token_accuracy": 0.8351697437465191,
      "num_tokens": 712841313.0,
      "step": 22336
    },
    {
      "entropy": 0.43186100106686354,
      "epoch": 2.055111816493292,
      "grad_norm": 0.1482296735048294,
      "learning_rate": 3.149937130064097e-05,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8689571432769299,
      "num_tokens": 712873522.0,
      "step": 22337
    },
    {
      "entropy": 0.6103125121444464,
      "epoch": 2.0552038228647334,
      "grad_norm": 0.17366889119148254,
      "learning_rate": 3.149630447449934e-05,
      "loss": 0.6,
      "mean_token_accuracy": 0.8122609332203865,
      "num_tokens": 712905037.0,
      "step": 22338
    },
    {
      "entropy": 0.4546457324177027,
      "epoch": 2.0552958292361745,
      "grad_norm": 0.15631404519081116,
      "learning_rate": 3.149323764835772e-05,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8558334335684776,
      "num_tokens": 712937221.0,
      "step": 22339
    },
    {
      "entropy": 0.5181230008602142,
      "epoch": 2.0553878356076156,
      "grad_norm": 0.15683603286743164,
      "learning_rate": 3.149017082221609e-05,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8392732851207256,
      "num_tokens": 712969782.0,
      "step": 22340
    },
    {
      "entropy": 0.5373398186638951,
      "epoch": 2.055479841979057,
      "grad_norm": 0.16866837441921234,
      "learning_rate": 3.1487103996074466e-05,
      "loss": 0.5135,
      "mean_token_accuracy": 0.8344261907041073,
      "num_tokens": 713000567.0,
      "step": 22341
    },
    {
      "entropy": 0.4588765911757946,
      "epoch": 2.0555718483504983,
      "grad_norm": 0.15486645698547363,
      "learning_rate": 3.148403716993284e-05,
      "loss": 0.4521,
      "mean_token_accuracy": 0.856891680508852,
      "num_tokens": 713032605.0,
      "step": 22342
    },
    {
      "entropy": 0.49338513729162514,
      "epoch": 2.0556638547219395,
      "grad_norm": 0.16272540390491486,
      "learning_rate": 3.148097034379121e-05,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8446051739156246,
      "num_tokens": 713064649.0,
      "step": 22343
    },
    {
      "entropy": 0.5514032458886504,
      "epoch": 2.0557558610933806,
      "grad_norm": 0.16026878356933594,
      "learning_rate": 3.147790351764959e-05,
      "loss": 0.5328,
      "mean_token_accuracy": 0.8310211300849915,
      "num_tokens": 713096473.0,
      "step": 22344
    },
    {
      "entropy": 0.5738812573254108,
      "epoch": 2.0558478674648217,
      "grad_norm": 0.17650341987609863,
      "learning_rate": 3.147483669150796e-05,
      "loss": 0.5568,
      "mean_token_accuracy": 0.8252611346542835,
      "num_tokens": 713127498.0,
      "step": 22345
    },
    {
      "entropy": 0.4484291863627732,
      "epoch": 2.0559398738362633,
      "grad_norm": 0.15070228278636932,
      "learning_rate": 3.147176986536633e-05,
      "loss": 0.4341,
      "mean_token_accuracy": 0.8624399155378342,
      "num_tokens": 713160103.0,
      "step": 22346
    },
    {
      "entropy": 0.4542163070291281,
      "epoch": 2.0560318802077044,
      "grad_norm": 0.14889565110206604,
      "learning_rate": 3.146870303922471e-05,
      "loss": 0.4373,
      "mean_token_accuracy": 0.8599953874945641,
      "num_tokens": 713191713.0,
      "step": 22347
    },
    {
      "entropy": 0.49682241026312113,
      "epoch": 2.0561238865791456,
      "grad_norm": 0.15244349837303162,
      "learning_rate": 3.146563621308308e-05,
      "loss": 0.479,
      "mean_token_accuracy": 0.8472513742744923,
      "num_tokens": 713223713.0,
      "step": 22348
    },
    {
      "entropy": 0.5222070589661598,
      "epoch": 2.0562158929505867,
      "grad_norm": 0.1538969725370407,
      "learning_rate": 3.1462569386941457e-05,
      "loss": 0.498,
      "mean_token_accuracy": 0.8405949622392654,
      "num_tokens": 713255756.0,
      "step": 22349
    },
    {
      "entropy": 0.5455788196995854,
      "epoch": 2.056307899322028,
      "grad_norm": 0.169391468167305,
      "learning_rate": 3.145950256079983e-05,
      "loss": 0.5179,
      "mean_token_accuracy": 0.8342102840542793,
      "num_tokens": 713287312.0,
      "step": 22350
    },
    {
      "entropy": 0.48613745998591185,
      "epoch": 2.0563999056934694,
      "grad_norm": 0.157512828707695,
      "learning_rate": 3.14564357346582e-05,
      "loss": 0.4705,
      "mean_token_accuracy": 0.8498138561844826,
      "num_tokens": 713319537.0,
      "step": 22351
    },
    {
      "entropy": 0.5177052654325962,
      "epoch": 2.0564919120649106,
      "grad_norm": 0.15993933379650116,
      "learning_rate": 3.145336890851658e-05,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8420102372765541,
      "num_tokens": 713351072.0,
      "step": 22352
    },
    {
      "entropy": 0.6028684079647064,
      "epoch": 2.0565839184363517,
      "grad_norm": 0.16620591282844543,
      "learning_rate": 3.145030208237495e-05,
      "loss": 0.581,
      "mean_token_accuracy": 0.8134290426969528,
      "num_tokens": 713382690.0,
      "step": 22353
    },
    {
      "entropy": 0.45637505664490163,
      "epoch": 2.056675924807793,
      "grad_norm": 0.1504899561405182,
      "learning_rate": 3.144723525623332e-05,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8594904281198978,
      "num_tokens": 713414510.0,
      "step": 22354
    },
    {
      "entropy": 0.500659616664052,
      "epoch": 2.056767931179234,
      "grad_norm": 0.15518701076507568,
      "learning_rate": 3.14441684300917e-05,
      "loss": 0.4828,
      "mean_token_accuracy": 0.846913255751133,
      "num_tokens": 713446458.0,
      "step": 22355
    },
    {
      "entropy": 0.5877839513123035,
      "epoch": 2.0568599375506755,
      "grad_norm": 0.17328305542469025,
      "learning_rate": 3.144110160395007e-05,
      "loss": 0.5808,
      "mean_token_accuracy": 0.8173444867134094,
      "num_tokens": 713478338.0,
      "step": 22356
    },
    {
      "entropy": 0.5483754528686404,
      "epoch": 2.0569519439221167,
      "grad_norm": 0.16237527132034302,
      "learning_rate": 3.143803477780845e-05,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8335336409509182,
      "num_tokens": 713509651.0,
      "step": 22357
    },
    {
      "entropy": 0.5019070953130722,
      "epoch": 2.057043950293558,
      "grad_norm": 0.16204094886779785,
      "learning_rate": 3.143496795166682e-05,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8443160988390446,
      "num_tokens": 713541734.0,
      "step": 22358
    },
    {
      "entropy": 0.46500387880951166,
      "epoch": 2.057135956664999,
      "grad_norm": 0.15180405974388123,
      "learning_rate": 3.143190112552519e-05,
      "loss": 0.4515,
      "mean_token_accuracy": 0.8572991266846657,
      "num_tokens": 713573518.0,
      "step": 22359
    },
    {
      "entropy": 0.5873374529182911,
      "epoch": 2.05722796303644,
      "grad_norm": 0.16695812344551086,
      "learning_rate": 3.142883429938357e-05,
      "loss": 0.5857,
      "mean_token_accuracy": 0.8184551820158958,
      "num_tokens": 713606010.0,
      "step": 22360
    },
    {
      "entropy": 0.5083217564970255,
      "epoch": 2.0573199694078816,
      "grad_norm": 0.15355022251605988,
      "learning_rate": 3.1425767473241946e-05,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8453508540987968,
      "num_tokens": 713638557.0,
      "step": 22361
    },
    {
      "entropy": 0.47844765335321426,
      "epoch": 2.0574119757793228,
      "grad_norm": 0.15820565819740295,
      "learning_rate": 3.142270064710032e-05,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8490441180765629,
      "num_tokens": 713670559.0,
      "step": 22362
    },
    {
      "entropy": 0.47527928883209825,
      "epoch": 2.057503982150764,
      "grad_norm": 0.1540517657995224,
      "learning_rate": 3.1419633820958695e-05,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8538907505571842,
      "num_tokens": 713701852.0,
      "step": 22363
    },
    {
      "entropy": 0.5140536250546575,
      "epoch": 2.057595988522205,
      "grad_norm": 0.15687260031700134,
      "learning_rate": 3.141656699481706e-05,
      "loss": 0.5022,
      "mean_token_accuracy": 0.840433094650507,
      "num_tokens": 713734088.0,
      "step": 22364
    },
    {
      "entropy": 0.6574119739234447,
      "epoch": 2.057687994893646,
      "grad_norm": 0.17171311378479004,
      "learning_rate": 3.1413500168675444e-05,
      "loss": 0.6412,
      "mean_token_accuracy": 0.7955421283841133,
      "num_tokens": 713765222.0,
      "step": 22365
    },
    {
      "entropy": 0.4744435250759125,
      "epoch": 2.0577800012650878,
      "grad_norm": 0.15272632241249084,
      "learning_rate": 3.141043334253381e-05,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8532989025115967,
      "num_tokens": 713797125.0,
      "step": 22366
    },
    {
      "entropy": 0.4988256115466356,
      "epoch": 2.057872007636529,
      "grad_norm": 0.15915460884571075,
      "learning_rate": 3.140736651639219e-05,
      "loss": 0.4663,
      "mean_token_accuracy": 0.849220409989357,
      "num_tokens": 713829164.0,
      "step": 22367
    },
    {
      "entropy": 0.5270381700247526,
      "epoch": 2.05796401400797,
      "grad_norm": 0.1575097143650055,
      "learning_rate": 3.140429969025056e-05,
      "loss": 0.5114,
      "mean_token_accuracy": 0.8362776450812817,
      "num_tokens": 713861697.0,
      "step": 22368
    },
    {
      "entropy": 0.5356270978227258,
      "epoch": 2.058056020379411,
      "grad_norm": 0.16127923130989075,
      "learning_rate": 3.1401232864108936e-05,
      "loss": 0.5118,
      "mean_token_accuracy": 0.8376453183591366,
      "num_tokens": 713892875.0,
      "step": 22369
    },
    {
      "entropy": 0.5333032067865133,
      "epoch": 2.0581480267508523,
      "grad_norm": 0.17086222767829895,
      "learning_rate": 3.139816603796731e-05,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8357493467628956,
      "num_tokens": 713923423.0,
      "step": 22370
    },
    {
      "entropy": 0.3764507779851556,
      "epoch": 2.058240033122294,
      "grad_norm": 0.14419543743133545,
      "learning_rate": 3.1395099211825685e-05,
      "loss": 0.3635,
      "mean_token_accuracy": 0.8839059583842754,
      "num_tokens": 713955331.0,
      "step": 22371
    },
    {
      "entropy": 0.5193637292832136,
      "epoch": 2.058332039493735,
      "grad_norm": 0.16019634902477264,
      "learning_rate": 3.139203238568405e-05,
      "loss": 0.5004,
      "mean_token_accuracy": 0.8419878892600536,
      "num_tokens": 713987662.0,
      "step": 22372
    },
    {
      "entropy": 0.5678440686315298,
      "epoch": 2.058424045865176,
      "grad_norm": 0.15997929871082306,
      "learning_rate": 3.1388965559542435e-05,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8258152566850185,
      "num_tokens": 714020298.0,
      "step": 22373
    },
    {
      "entropy": 0.5265815947204828,
      "epoch": 2.0585160522366173,
      "grad_norm": 0.15669120848178864,
      "learning_rate": 3.13858987334008e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8376092799007893,
      "num_tokens": 714052899.0,
      "step": 22374
    },
    {
      "entropy": 0.5300065148621798,
      "epoch": 2.0586080586080584,
      "grad_norm": 0.16268572211265564,
      "learning_rate": 3.138283190725918e-05,
      "loss": 0.5127,
      "mean_token_accuracy": 0.8374750576913357,
      "num_tokens": 714084901.0,
      "step": 22375
    },
    {
      "entropy": 0.5443704430945218,
      "epoch": 2.0587000649795,
      "grad_norm": 0.16564734280109406,
      "learning_rate": 3.137976508111755e-05,
      "loss": 0.5347,
      "mean_token_accuracy": 0.8318536318838596,
      "num_tokens": 714116918.0,
      "step": 22376
    },
    {
      "entropy": 0.4627838407177478,
      "epoch": 2.058792071350941,
      "grad_norm": 0.14871680736541748,
      "learning_rate": 3.1376698254975926e-05,
      "loss": 0.4466,
      "mean_token_accuracy": 0.85694245621562,
      "num_tokens": 714149144.0,
      "step": 22377
    },
    {
      "entropy": 0.49953779205679893,
      "epoch": 2.0588840777223822,
      "grad_norm": 0.1678881049156189,
      "learning_rate": 3.13736314288343e-05,
      "loss": 0.4915,
      "mean_token_accuracy": 0.8440989442169666,
      "num_tokens": 714180575.0,
      "step": 22378
    },
    {
      "entropy": 0.5299974363297224,
      "epoch": 2.0589760840938234,
      "grad_norm": 0.16159985959529877,
      "learning_rate": 3.1370564602692676e-05,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8349658995866776,
      "num_tokens": 714212822.0,
      "step": 22379
    },
    {
      "entropy": 0.49265719950199127,
      "epoch": 2.0590680904652645,
      "grad_norm": 0.15463268756866455,
      "learning_rate": 3.1367497776551044e-05,
      "loss": 0.4724,
      "mean_token_accuracy": 0.8503524288535118,
      "num_tokens": 714245166.0,
      "step": 22380
    },
    {
      "entropy": 0.538017145358026,
      "epoch": 2.059160096836706,
      "grad_norm": 0.1606171727180481,
      "learning_rate": 3.1364430950409425e-05,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8328192718327045,
      "num_tokens": 714277147.0,
      "step": 22381
    },
    {
      "entropy": 0.5392560102045536,
      "epoch": 2.0592521032081472,
      "grad_norm": 0.1621427983045578,
      "learning_rate": 3.136136412426779e-05,
      "loss": 0.524,
      "mean_token_accuracy": 0.8344292864203453,
      "num_tokens": 714308516.0,
      "step": 22382
    },
    {
      "entropy": 0.5404833722859621,
      "epoch": 2.0593441095795884,
      "grad_norm": 0.1666218638420105,
      "learning_rate": 3.1358297298126174e-05,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8362071812152863,
      "num_tokens": 714340344.0,
      "step": 22383
    },
    {
      "entropy": 0.5435812715440989,
      "epoch": 2.0594361159510295,
      "grad_norm": 0.16198846697807312,
      "learning_rate": 3.135523047198454e-05,
      "loss": 0.5356,
      "mean_token_accuracy": 0.8285306952893734,
      "num_tokens": 714372368.0,
      "step": 22384
    },
    {
      "entropy": 0.44714626856148243,
      "epoch": 2.0595281223224706,
      "grad_norm": 0.1523459553718567,
      "learning_rate": 3.135216364584292e-05,
      "loss": 0.4281,
      "mean_token_accuracy": 0.8608819916844368,
      "num_tokens": 714404888.0,
      "step": 22385
    },
    {
      "entropy": 0.5761854005977511,
      "epoch": 2.059620128693912,
      "grad_norm": 0.17228224873542786,
      "learning_rate": 3.134909681970129e-05,
      "loss": 0.5763,
      "mean_token_accuracy": 0.8204493932425976,
      "num_tokens": 714437274.0,
      "step": 22386
    },
    {
      "entropy": 0.5322483824566007,
      "epoch": 2.0597121350653533,
      "grad_norm": 0.15930204093456268,
      "learning_rate": 3.1346029993559666e-05,
      "loss": 0.5163,
      "mean_token_accuracy": 0.84053073823452,
      "num_tokens": 714469893.0,
      "step": 22387
    },
    {
      "entropy": 0.6048946604132652,
      "epoch": 2.0598041414367945,
      "grad_norm": 0.17062988877296448,
      "learning_rate": 3.134296316741804e-05,
      "loss": 0.6021,
      "mean_token_accuracy": 0.8157206326723099,
      "num_tokens": 714501830.0,
      "step": 22388
    },
    {
      "entropy": 0.616201488301158,
      "epoch": 2.0598961478082356,
      "grad_norm": 0.1724708080291748,
      "learning_rate": 3.1339896341276415e-05,
      "loss": 0.595,
      "mean_token_accuracy": 0.8111053109169006,
      "num_tokens": 714533689.0,
      "step": 22389
    },
    {
      "entropy": 0.4713795827701688,
      "epoch": 2.0599881541796767,
      "grad_norm": 0.15778283774852753,
      "learning_rate": 3.133682951513479e-05,
      "loss": 0.4599,
      "mean_token_accuracy": 0.8542724885046482,
      "num_tokens": 714566149.0,
      "step": 22390
    },
    {
      "entropy": 0.46176266111433506,
      "epoch": 2.0600801605511183,
      "grad_norm": 0.15562663972377777,
      "learning_rate": 3.1333762688993165e-05,
      "loss": 0.4501,
      "mean_token_accuracy": 0.8565462380647659,
      "num_tokens": 714598582.0,
      "step": 22391
    },
    {
      "entropy": 0.47603205032646656,
      "epoch": 2.0601721669225594,
      "grad_norm": 0.16039521992206573,
      "learning_rate": 3.133069586285154e-05,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8514722771942616,
      "num_tokens": 714631031.0,
      "step": 22392
    },
    {
      "entropy": 0.4965623782481998,
      "epoch": 2.0602641732940006,
      "grad_norm": 0.15368667244911194,
      "learning_rate": 3.132762903670991e-05,
      "loss": 0.4856,
      "mean_token_accuracy": 0.8464513421058655,
      "num_tokens": 714663147.0,
      "step": 22393
    },
    {
      "entropy": 0.42174591962248087,
      "epoch": 2.0603561796654417,
      "grad_norm": 0.15311302244663239,
      "learning_rate": 3.132456221056829e-05,
      "loss": 0.3992,
      "mean_token_accuracy": 0.8706614561378956,
      "num_tokens": 714694795.0,
      "step": 22394
    },
    {
      "entropy": 0.5514352219179273,
      "epoch": 2.060448186036883,
      "grad_norm": 0.16321536898612976,
      "learning_rate": 3.1321495384426657e-05,
      "loss": 0.5332,
      "mean_token_accuracy": 0.8286481238901615,
      "num_tokens": 714726772.0,
      "step": 22395
    },
    {
      "entropy": 0.463690223172307,
      "epoch": 2.0605401924083244,
      "grad_norm": 0.14942805469036102,
      "learning_rate": 3.131842855828503e-05,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8577322103083134,
      "num_tokens": 714759121.0,
      "step": 22396
    },
    {
      "entropy": 0.4775157989934087,
      "epoch": 2.0606321987797656,
      "grad_norm": 0.15647388994693756,
      "learning_rate": 3.1315361732143406e-05,
      "loss": 0.466,
      "mean_token_accuracy": 0.851597860455513,
      "num_tokens": 714791148.0,
      "step": 22397
    },
    {
      "entropy": 0.539890231564641,
      "epoch": 2.0607242051512067,
      "grad_norm": 0.16301660239696503,
      "learning_rate": 3.131229490600178e-05,
      "loss": 0.5145,
      "mean_token_accuracy": 0.8379457108676434,
      "num_tokens": 714822240.0,
      "step": 22398
    },
    {
      "entropy": 0.4976618867367506,
      "epoch": 2.060816211522648,
      "grad_norm": 0.15466903150081635,
      "learning_rate": 3.1309228079860155e-05,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8465052023530006,
      "num_tokens": 714853830.0,
      "step": 22399
    },
    {
      "entropy": 0.564931477420032,
      "epoch": 2.060908217894089,
      "grad_norm": 0.1680838018655777,
      "learning_rate": 3.130616125371853e-05,
      "loss": 0.5513,
      "mean_token_accuracy": 0.824127547442913,
      "num_tokens": 714885231.0,
      "step": 22400
    },
    {
      "entropy": 0.4699125150218606,
      "epoch": 2.0610002242655305,
      "grad_norm": 0.15676212310791016,
      "learning_rate": 3.1303094427576904e-05,
      "loss": 0.4623,
      "mean_token_accuracy": 0.8514115549623966,
      "num_tokens": 714916241.0,
      "step": 22401
    },
    {
      "entropy": 0.4332041274756193,
      "epoch": 2.0610922306369717,
      "grad_norm": 0.15040378272533417,
      "learning_rate": 3.130002760143528e-05,
      "loss": 0.4156,
      "mean_token_accuracy": 0.8670358844101429,
      "num_tokens": 714948381.0,
      "step": 22402
    },
    {
      "entropy": 0.48392294999212027,
      "epoch": 2.061184237008413,
      "grad_norm": 0.15592658519744873,
      "learning_rate": 3.129696077529365e-05,
      "loss": 0.4622,
      "mean_token_accuracy": 0.8520380519330502,
      "num_tokens": 714979719.0,
      "step": 22403
    },
    {
      "entropy": 0.46191028971225023,
      "epoch": 2.061276243379854,
      "grad_norm": 0.16202935576438904,
      "learning_rate": 3.129389394915203e-05,
      "loss": 0.4561,
      "mean_token_accuracy": 0.8580752834677696,
      "num_tokens": 715011399.0,
      "step": 22404
    },
    {
      "entropy": 0.5041541596874595,
      "epoch": 2.061368249751295,
      "grad_norm": 0.1531900316476822,
      "learning_rate": 3.1290827123010396e-05,
      "loss": 0.4979,
      "mean_token_accuracy": 0.8439702577888966,
      "num_tokens": 715043876.0,
      "step": 22405
    },
    {
      "entropy": 0.6533028418198228,
      "epoch": 2.0614602561227366,
      "grad_norm": 0.17110490798950195,
      "learning_rate": 3.128776029686877e-05,
      "loss": 0.6325,
      "mean_token_accuracy": 0.8022549673914909,
      "num_tokens": 715075727.0,
      "step": 22406
    },
    {
      "entropy": 0.5354153821244836,
      "epoch": 2.061552262494178,
      "grad_norm": 0.16457177698612213,
      "learning_rate": 3.1284693470727145e-05,
      "loss": 0.5235,
      "mean_token_accuracy": 0.8354162462055683,
      "num_tokens": 715106328.0,
      "step": 22407
    },
    {
      "entropy": 0.4898986164480448,
      "epoch": 2.061644268865619,
      "grad_norm": 0.15566405653953552,
      "learning_rate": 3.128162664458552e-05,
      "loss": 0.4699,
      "mean_token_accuracy": 0.8475886397063732,
      "num_tokens": 715138015.0,
      "step": 22408
    },
    {
      "entropy": 0.6066852752119303,
      "epoch": 2.06173627523706,
      "grad_norm": 0.16813671588897705,
      "learning_rate": 3.1278559818443895e-05,
      "loss": 0.5906,
      "mean_token_accuracy": 0.8165213577449322,
      "num_tokens": 715170580.0,
      "step": 22409
    },
    {
      "entropy": 0.47097474336624146,
      "epoch": 2.061828281608501,
      "grad_norm": 0.15913936495780945,
      "learning_rate": 3.127549299230227e-05,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8525538519024849,
      "num_tokens": 715201974.0,
      "step": 22410
    },
    {
      "entropy": 0.5520768128335476,
      "epoch": 2.0619202879799428,
      "grad_norm": 0.16325248777866364,
      "learning_rate": 3.127242616616064e-05,
      "loss": 0.5271,
      "mean_token_accuracy": 0.8305477872490883,
      "num_tokens": 715233348.0,
      "step": 22411
    },
    {
      "entropy": 0.5772763779386878,
      "epoch": 2.062012294351384,
      "grad_norm": 0.16650116443634033,
      "learning_rate": 3.126935934001902e-05,
      "loss": 0.5562,
      "mean_token_accuracy": 0.8254686109721661,
      "num_tokens": 715264715.0,
      "step": 22412
    },
    {
      "entropy": 0.5242034252732992,
      "epoch": 2.062104300722825,
      "grad_norm": 0.16425809264183044,
      "learning_rate": 3.1266292513877387e-05,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8422914184629917,
      "num_tokens": 715296165.0,
      "step": 22413
    },
    {
      "entropy": 0.5704994415864348,
      "epoch": 2.062196307094266,
      "grad_norm": 0.16543048620224,
      "learning_rate": 3.126322568773576e-05,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8272164762020111,
      "num_tokens": 715328231.0,
      "step": 22414
    },
    {
      "entropy": 0.4723540502600372,
      "epoch": 2.0622883134657073,
      "grad_norm": 0.14768525958061218,
      "learning_rate": 3.1260158861594136e-05,
      "loss": 0.4472,
      "mean_token_accuracy": 0.85738505423069,
      "num_tokens": 715360205.0,
      "step": 22415
    },
    {
      "entropy": 0.6029770318418741,
      "epoch": 2.062380319837149,
      "grad_norm": 0.16959869861602783,
      "learning_rate": 3.125709203545251e-05,
      "loss": 0.5809,
      "mean_token_accuracy": 0.8157648220658302,
      "num_tokens": 715391204.0,
      "step": 22416
    },
    {
      "entropy": 0.5301998853683472,
      "epoch": 2.06247232620859,
      "grad_norm": 0.15594902634620667,
      "learning_rate": 3.1254025209310885e-05,
      "loss": 0.5138,
      "mean_token_accuracy": 0.8379315696656704,
      "num_tokens": 715423338.0,
      "step": 22417
    },
    {
      "entropy": 0.4423100329004228,
      "epoch": 2.062564332580031,
      "grad_norm": 0.1494518667459488,
      "learning_rate": 3.125095838316926e-05,
      "loss": 0.4249,
      "mean_token_accuracy": 0.863927137106657,
      "num_tokens": 715455126.0,
      "step": 22418
    },
    {
      "entropy": 0.4877754799090326,
      "epoch": 2.0626563389514723,
      "grad_norm": 0.15600061416625977,
      "learning_rate": 3.124789155702763e-05,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8522830531001091,
      "num_tokens": 715487428.0,
      "step": 22419
    },
    {
      "entropy": 0.5350195169448853,
      "epoch": 2.0627483453229134,
      "grad_norm": 0.17236202955245972,
      "learning_rate": 3.124482473088601e-05,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8391848541796207,
      "num_tokens": 715519231.0,
      "step": 22420
    },
    {
      "entropy": 0.49223579885438085,
      "epoch": 2.062840351694355,
      "grad_norm": 0.1567666083574295,
      "learning_rate": 3.1241757904744384e-05,
      "loss": 0.4752,
      "mean_token_accuracy": 0.8483807183802128,
      "num_tokens": 715551186.0,
      "step": 22421
    },
    {
      "entropy": 0.6320508811622858,
      "epoch": 2.062932358065796,
      "grad_norm": 0.17262163758277893,
      "learning_rate": 3.123869107860276e-05,
      "loss": 0.6119,
      "mean_token_accuracy": 0.8107421845197678,
      "num_tokens": 715582470.0,
      "step": 22422
    },
    {
      "entropy": 0.46372546534985304,
      "epoch": 2.0630243644372372,
      "grad_norm": 0.15769639611244202,
      "learning_rate": 3.123562425246113e-05,
      "loss": 0.4502,
      "mean_token_accuracy": 0.85587327927351,
      "num_tokens": 715614279.0,
      "step": 22423
    },
    {
      "entropy": 0.532630623318255,
      "epoch": 2.0631163708086784,
      "grad_norm": 0.1646023541688919,
      "learning_rate": 3.12325574263195e-05,
      "loss": 0.5252,
      "mean_token_accuracy": 0.8330168798565865,
      "num_tokens": 715646157.0,
      "step": 22424
    },
    {
      "entropy": 0.6710794754326344,
      "epoch": 2.0632083771801195,
      "grad_norm": 0.1748478263616562,
      "learning_rate": 3.122949060017788e-05,
      "loss": 0.6624,
      "mean_token_accuracy": 0.7922975271940231,
      "num_tokens": 715678380.0,
      "step": 22425
    },
    {
      "entropy": 0.5451125428080559,
      "epoch": 2.063300383551561,
      "grad_norm": 0.1687472015619278,
      "learning_rate": 3.122642377403625e-05,
      "loss": 0.5375,
      "mean_token_accuracy": 0.8308032378554344,
      "num_tokens": 715710502.0,
      "step": 22426
    },
    {
      "entropy": 0.5512479241006076,
      "epoch": 2.0633923899230022,
      "grad_norm": 0.16555604338645935,
      "learning_rate": 3.1223356947894625e-05,
      "loss": 0.5347,
      "mean_token_accuracy": 0.8323218747973442,
      "num_tokens": 715742529.0,
      "step": 22427
    },
    {
      "entropy": 0.5359931606799364,
      "epoch": 2.0634843962944434,
      "grad_norm": 0.16607928276062012,
      "learning_rate": 3.1220290121753e-05,
      "loss": 0.5217,
      "mean_token_accuracy": 0.8346541300415993,
      "num_tokens": 715774856.0,
      "step": 22428
    },
    {
      "entropy": 0.551836259663105,
      "epoch": 2.0635764026658845,
      "grad_norm": 0.16568945348262787,
      "learning_rate": 3.1217223295611374e-05,
      "loss": 0.5497,
      "mean_token_accuracy": 0.8247610218822956,
      "num_tokens": 715806882.0,
      "step": 22429
    },
    {
      "entropy": 0.5384911228902638,
      "epoch": 2.0636684090373256,
      "grad_norm": 0.16016843914985657,
      "learning_rate": 3.121415646946975e-05,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8327582515776157,
      "num_tokens": 715839479.0,
      "step": 22430
    },
    {
      "entropy": 0.5757613591849804,
      "epoch": 2.063760415408767,
      "grad_norm": 0.16266348958015442,
      "learning_rate": 3.1211089643328123e-05,
      "loss": 0.5514,
      "mean_token_accuracy": 0.8279937878251076,
      "num_tokens": 715872106.0,
      "step": 22431
    },
    {
      "entropy": 0.49116116669029,
      "epoch": 2.0638524217802083,
      "grad_norm": 0.1549210250377655,
      "learning_rate": 3.120802281718649e-05,
      "loss": 0.4853,
      "mean_token_accuracy": 0.846547219902277,
      "num_tokens": 715904341.0,
      "step": 22432
    },
    {
      "entropy": 0.5252683805301785,
      "epoch": 2.0639444281516495,
      "grad_norm": 0.16295385360717773,
      "learning_rate": 3.120495599104487e-05,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8351657204329967,
      "num_tokens": 715936308.0,
      "step": 22433
    },
    {
      "entropy": 0.4757167221978307,
      "epoch": 2.0640364345230906,
      "grad_norm": 0.15418560802936554,
      "learning_rate": 3.120188916490324e-05,
      "loss": 0.465,
      "mean_token_accuracy": 0.8514380380511284,
      "num_tokens": 715967716.0,
      "step": 22434
    },
    {
      "entropy": 0.5958537571132183,
      "epoch": 2.0641284408945317,
      "grad_norm": 0.16933032870292664,
      "learning_rate": 3.1198822338761615e-05,
      "loss": 0.5795,
      "mean_token_accuracy": 0.8142009750008583,
      "num_tokens": 715999537.0,
      "step": 22435
    },
    {
      "entropy": 0.4005924202501774,
      "epoch": 2.0642204472659733,
      "grad_norm": 0.1485927700996399,
      "learning_rate": 3.119575551261999e-05,
      "loss": 0.3851,
      "mean_token_accuracy": 0.876174371689558,
      "num_tokens": 716032136.0,
      "step": 22436
    },
    {
      "entropy": 0.6353486534208059,
      "epoch": 2.0643124536374144,
      "grad_norm": 0.17552897334098816,
      "learning_rate": 3.1192688686478365e-05,
      "loss": 0.6326,
      "mean_token_accuracy": 0.801180075854063,
      "num_tokens": 716064515.0,
      "step": 22437
    },
    {
      "entropy": 0.5009051105007529,
      "epoch": 2.0644044600088556,
      "grad_norm": 0.1502675563097,
      "learning_rate": 3.118962186033674e-05,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8520999886095524,
      "num_tokens": 716097159.0,
      "step": 22438
    },
    {
      "entropy": 0.5069295112043619,
      "epoch": 2.0644964663802967,
      "grad_norm": 0.15520353615283966,
      "learning_rate": 3.1186555034195114e-05,
      "loss": 0.4992,
      "mean_token_accuracy": 0.8453173525631428,
      "num_tokens": 716129707.0,
      "step": 22439
    },
    {
      "entropy": 0.5127569725736976,
      "epoch": 2.064588472751738,
      "grad_norm": 0.15212012827396393,
      "learning_rate": 3.118348820805348e-05,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8439176343381405,
      "num_tokens": 716161034.0,
      "step": 22440
    },
    {
      "entropy": 0.5786544680595398,
      "epoch": 2.0646804791231794,
      "grad_norm": 0.1655971109867096,
      "learning_rate": 3.118042138191186e-05,
      "loss": 0.5509,
      "mean_token_accuracy": 0.8250050432980061,
      "num_tokens": 716192973.0,
      "step": 22441
    },
    {
      "entropy": 0.517729289829731,
      "epoch": 2.0647724854946206,
      "grad_norm": 0.16353149712085724,
      "learning_rate": 3.117735455577023e-05,
      "loss": 0.494,
      "mean_token_accuracy": 0.8395944572985172,
      "num_tokens": 716224173.0,
      "step": 22442
    },
    {
      "entropy": 0.5472362842410803,
      "epoch": 2.0648644918660617,
      "grad_norm": 0.16626796126365662,
      "learning_rate": 3.117428772962861e-05,
      "loss": 0.5333,
      "mean_token_accuracy": 0.8306654207408428,
      "num_tokens": 716255074.0,
      "step": 22443
    },
    {
      "entropy": 0.586895908229053,
      "epoch": 2.064956498237503,
      "grad_norm": 0.16734614968299866,
      "learning_rate": 3.117122090348698e-05,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8173216767609119,
      "num_tokens": 716286450.0,
      "step": 22444
    },
    {
      "entropy": 0.6168718761764467,
      "epoch": 2.065048504608944,
      "grad_norm": 0.16966307163238525,
      "learning_rate": 3.1168154077345355e-05,
      "loss": 0.6046,
      "mean_token_accuracy": 0.813806738704443,
      "num_tokens": 716318156.0,
      "step": 22445
    },
    {
      "entropy": 0.3942455002106726,
      "epoch": 2.0651405109803855,
      "grad_norm": 0.15066441893577576,
      "learning_rate": 3.116508725120373e-05,
      "loss": 0.3695,
      "mean_token_accuracy": 0.879535049200058,
      "num_tokens": 716349694.0,
      "step": 22446
    },
    {
      "entropy": 0.5600652685388923,
      "epoch": 2.0652325173518267,
      "grad_norm": 0.16673022508621216,
      "learning_rate": 3.1162020425062104e-05,
      "loss": 0.5489,
      "mean_token_accuracy": 0.8239256739616394,
      "num_tokens": 716382105.0,
      "step": 22447
    },
    {
      "entropy": 0.5762794241309166,
      "epoch": 2.065324523723268,
      "grad_norm": 0.16761209070682526,
      "learning_rate": 3.115895359892048e-05,
      "loss": 0.5708,
      "mean_token_accuracy": 0.8216730579733849,
      "num_tokens": 716414123.0,
      "step": 22448
    },
    {
      "entropy": 0.46910219732671976,
      "epoch": 2.065416530094709,
      "grad_norm": 0.15513013303279877,
      "learning_rate": 3.1155886772778854e-05,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8526709824800491,
      "num_tokens": 716445469.0,
      "step": 22449
    },
    {
      "entropy": 0.5360219292342663,
      "epoch": 2.06550853646615,
      "grad_norm": 0.16732706129550934,
      "learning_rate": 3.115281994663722e-05,
      "loss": 0.5215,
      "mean_token_accuracy": 0.8376056961715221,
      "num_tokens": 716476758.0,
      "step": 22450
    },
    {
      "entropy": 0.5410167938098311,
      "epoch": 2.0656005428375916,
      "grad_norm": 0.16385437548160553,
      "learning_rate": 3.11497531204956e-05,
      "loss": 0.5358,
      "mean_token_accuracy": 0.8321725986897945,
      "num_tokens": 716508802.0,
      "step": 22451
    },
    {
      "entropy": 0.5382761433720589,
      "epoch": 2.065692549209033,
      "grad_norm": 0.166752889752388,
      "learning_rate": 3.114668629435398e-05,
      "loss": 0.5377,
      "mean_token_accuracy": 0.8333863392472267,
      "num_tokens": 716540086.0,
      "step": 22452
    },
    {
      "entropy": 0.5282694026827812,
      "epoch": 2.065784555580474,
      "grad_norm": 0.16046343743801117,
      "learning_rate": 3.1143619468212345e-05,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8386684730648994,
      "num_tokens": 716572163.0,
      "step": 22453
    },
    {
      "entropy": 0.5223869381006807,
      "epoch": 2.065876561951915,
      "grad_norm": 0.16164399683475494,
      "learning_rate": 3.114055264207073e-05,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8424283526837826,
      "num_tokens": 716603951.0,
      "step": 22454
    },
    {
      "entropy": 0.5571662718430161,
      "epoch": 2.065968568323356,
      "grad_norm": 0.16532984375953674,
      "learning_rate": 3.1137485815929095e-05,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8316881284117699,
      "num_tokens": 716635170.0,
      "step": 22455
    },
    {
      "entropy": 0.5594559544697404,
      "epoch": 2.0660605746947978,
      "grad_norm": 0.16906294226646423,
      "learning_rate": 3.1134418989787476e-05,
      "loss": 0.5592,
      "mean_token_accuracy": 0.8213167674839497,
      "num_tokens": 716666760.0,
      "step": 22456
    },
    {
      "entropy": 0.6186505607329309,
      "epoch": 2.066152581066239,
      "grad_norm": 0.17043675482273102,
      "learning_rate": 3.1131352163645844e-05,
      "loss": 0.6004,
      "mean_token_accuracy": 0.8117261789739132,
      "num_tokens": 716698769.0,
      "step": 22457
    },
    {
      "entropy": 0.4464999372139573,
      "epoch": 2.06624458743768,
      "grad_norm": 0.14978939294815063,
      "learning_rate": 3.112828533750422e-05,
      "loss": 0.4254,
      "mean_token_accuracy": 0.8637266829609871,
      "num_tokens": 716730616.0,
      "step": 22458
    },
    {
      "entropy": 0.4716324945911765,
      "epoch": 2.066336593809121,
      "grad_norm": 0.15137144923210144,
      "learning_rate": 3.112521851136259e-05,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8571335934102535,
      "num_tokens": 716762696.0,
      "step": 22459
    },
    {
      "entropy": 0.5493847243487835,
      "epoch": 2.0664286001805623,
      "grad_norm": 0.15760335326194763,
      "learning_rate": 3.112215168522097e-05,
      "loss": 0.5301,
      "mean_token_accuracy": 0.8329585418105125,
      "num_tokens": 716795062.0,
      "step": 22460
    },
    {
      "entropy": 0.5917806699872017,
      "epoch": 2.066520606552004,
      "grad_norm": 0.16709668934345245,
      "learning_rate": 3.111908485907934e-05,
      "loss": 0.5773,
      "mean_token_accuracy": 0.8219651840627193,
      "num_tokens": 716827012.0,
      "step": 22461
    },
    {
      "entropy": 0.48587752506136894,
      "epoch": 2.066612612923445,
      "grad_norm": 0.14999912679195404,
      "learning_rate": 3.111601803293772e-05,
      "loss": 0.4648,
      "mean_token_accuracy": 0.850292693823576,
      "num_tokens": 716858998.0,
      "step": 22462
    },
    {
      "entropy": 0.6029048915952444,
      "epoch": 2.066704619294886,
      "grad_norm": 0.1604674607515335,
      "learning_rate": 3.1112951206796085e-05,
      "loss": 0.5791,
      "mean_token_accuracy": 0.8167681694030762,
      "num_tokens": 716891305.0,
      "step": 22463
    },
    {
      "entropy": 0.5323959300294518,
      "epoch": 2.0667966256663273,
      "grad_norm": 0.15953059494495392,
      "learning_rate": 3.1109884380654467e-05,
      "loss": 0.516,
      "mean_token_accuracy": 0.8351245671510696,
      "num_tokens": 716924073.0,
      "step": 22464
    },
    {
      "entropy": 0.5686063803732395,
      "epoch": 2.0668886320377684,
      "grad_norm": 0.16538891196250916,
      "learning_rate": 3.1106817554512834e-05,
      "loss": 0.5372,
      "mean_token_accuracy": 0.82938352227211,
      "num_tokens": 716955065.0,
      "step": 22465
    },
    {
      "entropy": 0.5199263854883611,
      "epoch": 2.06698063840921,
      "grad_norm": 0.16411985456943512,
      "learning_rate": 3.110375072837121e-05,
      "loss": 0.5061,
      "mean_token_accuracy": 0.8388587161898613,
      "num_tokens": 716986848.0,
      "step": 22466
    },
    {
      "entropy": 0.546479613520205,
      "epoch": 2.067072644780651,
      "grad_norm": 0.170152485370636,
      "learning_rate": 3.1100683902229584e-05,
      "loss": 0.5317,
      "mean_token_accuracy": 0.8315850682556629,
      "num_tokens": 717018762.0,
      "step": 22467
    },
    {
      "entropy": 0.5418340917676687,
      "epoch": 2.0671646511520922,
      "grad_norm": 0.1670083850622177,
      "learning_rate": 3.109761707608796e-05,
      "loss": 0.528,
      "mean_token_accuracy": 0.8298650272190571,
      "num_tokens": 717050178.0,
      "step": 22468
    },
    {
      "entropy": 0.4321117168292403,
      "epoch": 2.0672566575235334,
      "grad_norm": 0.15199756622314453,
      "learning_rate": 3.109455024994633e-05,
      "loss": 0.4252,
      "mean_token_accuracy": 0.8651282526552677,
      "num_tokens": 717081510.0,
      "step": 22469
    },
    {
      "entropy": 0.5142446858808398,
      "epoch": 2.0673486638949745,
      "grad_norm": 0.16154780983924866,
      "learning_rate": 3.109148342380471e-05,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8410725072026253,
      "num_tokens": 717113951.0,
      "step": 22470
    },
    {
      "entropy": 0.5839595561847091,
      "epoch": 2.067440670266416,
      "grad_norm": 0.16555598378181458,
      "learning_rate": 3.1088416597663076e-05,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8214712254703045,
      "num_tokens": 717145682.0,
      "step": 22471
    },
    {
      "entropy": 0.598362653516233,
      "epoch": 2.0675326766378572,
      "grad_norm": 0.17111924290657043,
      "learning_rate": 3.108534977152146e-05,
      "loss": 0.5882,
      "mean_token_accuracy": 0.8164033032953739,
      "num_tokens": 717176896.0,
      "step": 22472
    },
    {
      "entropy": 0.5165843789000064,
      "epoch": 2.0676246830092984,
      "grad_norm": 0.16060616075992584,
      "learning_rate": 3.1082282945379825e-05,
      "loss": 0.502,
      "mean_token_accuracy": 0.8396940380334854,
      "num_tokens": 717208519.0,
      "step": 22473
    },
    {
      "entropy": 0.49385131895542145,
      "epoch": 2.0677166893807395,
      "grad_norm": 0.16013973951339722,
      "learning_rate": 3.10792161192382e-05,
      "loss": 0.4856,
      "mean_token_accuracy": 0.8466161638498306,
      "num_tokens": 717240522.0,
      "step": 22474
    },
    {
      "entropy": 0.5284152792301029,
      "epoch": 2.0678086957521806,
      "grad_norm": 0.1694606989622116,
      "learning_rate": 3.1076149293096574e-05,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8332427479326725,
      "num_tokens": 717272660.0,
      "step": 22475
    },
    {
      "entropy": 0.5280456398613751,
      "epoch": 2.067900702123622,
      "grad_norm": 0.16228750348091125,
      "learning_rate": 3.107308246695495e-05,
      "loss": 0.5228,
      "mean_token_accuracy": 0.8351113796234131,
      "num_tokens": 717304733.0,
      "step": 22476
    },
    {
      "entropy": 0.4900655159726739,
      "epoch": 2.0679927084950633,
      "grad_norm": 0.14869913458824158,
      "learning_rate": 3.1070015640813323e-05,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8494806662201881,
      "num_tokens": 717337330.0,
      "step": 22477
    },
    {
      "entropy": 0.5890136391390115,
      "epoch": 2.0680847148665045,
      "grad_norm": 0.17480535805225372,
      "learning_rate": 3.10669488146717e-05,
      "loss": 0.588,
      "mean_token_accuracy": 0.8223715424537659,
      "num_tokens": 717368575.0,
      "step": 22478
    },
    {
      "entropy": 0.5428223749622703,
      "epoch": 2.0681767212379456,
      "grad_norm": 0.16297097504138947,
      "learning_rate": 3.1063881988530066e-05,
      "loss": 0.5207,
      "mean_token_accuracy": 0.8337548226118088,
      "num_tokens": 717400243.0,
      "step": 22479
    },
    {
      "entropy": 0.4358354303985834,
      "epoch": 2.0682687276093867,
      "grad_norm": 0.15011857450008392,
      "learning_rate": 3.106081516238845e-05,
      "loss": 0.4234,
      "mean_token_accuracy": 0.8639388494193554,
      "num_tokens": 717432598.0,
      "step": 22480
    },
    {
      "entropy": 0.5394505299627781,
      "epoch": 2.0683607339808283,
      "grad_norm": 0.1673029214143753,
      "learning_rate": 3.1057748336246815e-05,
      "loss": 0.5308,
      "mean_token_accuracy": 0.831534132361412,
      "num_tokens": 717464101.0,
      "step": 22481
    },
    {
      "entropy": 0.48384127859026194,
      "epoch": 2.0684527403522694,
      "grad_norm": 0.1600276678800583,
      "learning_rate": 3.10546815101052e-05,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8501194678246975,
      "num_tokens": 717496122.0,
      "step": 22482
    },
    {
      "entropy": 0.5895882314071059,
      "epoch": 2.0685447467237106,
      "grad_norm": 0.1616256684064865,
      "learning_rate": 3.105161468396357e-05,
      "loss": 0.5733,
      "mean_token_accuracy": 0.8189368173480034,
      "num_tokens": 717528354.0,
      "step": 22483
    },
    {
      "entropy": 0.3987171412445605,
      "epoch": 2.0686367530951517,
      "grad_norm": 0.1442752182483673,
      "learning_rate": 3.104854785782194e-05,
      "loss": 0.3811,
      "mean_token_accuracy": 0.8769692145287991,
      "num_tokens": 717560712.0,
      "step": 22484
    },
    {
      "entropy": 0.5843355171382427,
      "epoch": 2.068728759466593,
      "grad_norm": 0.16552190482616425,
      "learning_rate": 3.104548103168032e-05,
      "loss": 0.5734,
      "mean_token_accuracy": 0.8190691694617271,
      "num_tokens": 717593095.0,
      "step": 22485
    },
    {
      "entropy": 0.5701667293906212,
      "epoch": 2.0688207658380344,
      "grad_norm": 0.16811543703079224,
      "learning_rate": 3.104241420553869e-05,
      "loss": 0.5509,
      "mean_token_accuracy": 0.8266247548162937,
      "num_tokens": 717625204.0,
      "step": 22486
    },
    {
      "entropy": 0.5900323418900371,
      "epoch": 2.0689127722094756,
      "grad_norm": 0.1709301471710205,
      "learning_rate": 3.103934737939706e-05,
      "loss": 0.5654,
      "mean_token_accuracy": 0.8208187073469162,
      "num_tokens": 717656229.0,
      "step": 22487
    },
    {
      "entropy": 0.5735253896564245,
      "epoch": 2.0690047785809167,
      "grad_norm": 0.1689765751361847,
      "learning_rate": 3.103628055325544e-05,
      "loss": 0.571,
      "mean_token_accuracy": 0.8206015080213547,
      "num_tokens": 717687881.0,
      "step": 22488
    },
    {
      "entropy": 0.49999058712273836,
      "epoch": 2.069096784952358,
      "grad_norm": 0.16199040412902832,
      "learning_rate": 3.103321372711381e-05,
      "loss": 0.4785,
      "mean_token_accuracy": 0.8448176011443138,
      "num_tokens": 717718402.0,
      "step": 22489
    },
    {
      "entropy": 0.5848490400239825,
      "epoch": 2.069188791323799,
      "grad_norm": 0.1653970181941986,
      "learning_rate": 3.103014690097219e-05,
      "loss": 0.5605,
      "mean_token_accuracy": 0.8270043805241585,
      "num_tokens": 717749669.0,
      "step": 22490
    },
    {
      "entropy": 0.4365455647930503,
      "epoch": 2.0692807976952405,
      "grad_norm": 0.1509149968624115,
      "learning_rate": 3.102708007483056e-05,
      "loss": 0.4145,
      "mean_token_accuracy": 0.8685203455388546,
      "num_tokens": 717782127.0,
      "step": 22491
    },
    {
      "entropy": 0.48995241522789,
      "epoch": 2.0693728040666817,
      "grad_norm": 0.1505817174911499,
      "learning_rate": 3.102401324868893e-05,
      "loss": 0.4749,
      "mean_token_accuracy": 0.8454425111413002,
      "num_tokens": 717814142.0,
      "step": 22492
    },
    {
      "entropy": 0.5432414263486862,
      "epoch": 2.069464810438123,
      "grad_norm": 0.15883338451385498,
      "learning_rate": 3.102094642254731e-05,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8344897218048573,
      "num_tokens": 717845631.0,
      "step": 22493
    },
    {
      "entropy": 0.5272060660645366,
      "epoch": 2.069556816809564,
      "grad_norm": 0.15883173048496246,
      "learning_rate": 3.101787959640568e-05,
      "loss": 0.5046,
      "mean_token_accuracy": 0.8416201025247574,
      "num_tokens": 717877701.0,
      "step": 22494
    },
    {
      "entropy": 0.5643887128680944,
      "epoch": 2.069648823181005,
      "grad_norm": 0.16618569195270538,
      "learning_rate": 3.1014812770264053e-05,
      "loss": 0.5458,
      "mean_token_accuracy": 0.8311862722039223,
      "num_tokens": 717909768.0,
      "step": 22495
    },
    {
      "entropy": 0.6261502122506499,
      "epoch": 2.0697408295524466,
      "grad_norm": 0.17506273090839386,
      "learning_rate": 3.101174594412243e-05,
      "loss": 0.6154,
      "mean_token_accuracy": 0.8058929294347763,
      "num_tokens": 717941271.0,
      "step": 22496
    },
    {
      "entropy": 0.5153543595224619,
      "epoch": 2.069832835923888,
      "grad_norm": 0.16307462751865387,
      "learning_rate": 3.10086791179808e-05,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8428578749299049,
      "num_tokens": 717973065.0,
      "step": 22497
    },
    {
      "entropy": 0.4860964473336935,
      "epoch": 2.069924842295329,
      "grad_norm": 0.15724128484725952,
      "learning_rate": 3.100561229183918e-05,
      "loss": 0.4799,
      "mean_token_accuracy": 0.8471849076449871,
      "num_tokens": 718005833.0,
      "step": 22498
    },
    {
      "entropy": 0.5496584116481245,
      "epoch": 2.07001684866677,
      "grad_norm": 0.16152030229568481,
      "learning_rate": 3.100254546569755e-05,
      "loss": 0.5331,
      "mean_token_accuracy": 0.8286403119564056,
      "num_tokens": 718037737.0,
      "step": 22499
    },
    {
      "entropy": 0.5758309923112392,
      "epoch": 2.070108855038211,
      "grad_norm": 0.16685467958450317,
      "learning_rate": 3.099947863955593e-05,
      "loss": 0.5582,
      "mean_token_accuracy": 0.8217999041080475,
      "num_tokens": 718069742.0,
      "step": 22500
    },
    {
      "entropy": 0.6300980467349291,
      "epoch": 2.0702008614096528,
      "grad_norm": 0.17819921672344208,
      "learning_rate": 3.09964118134143e-05,
      "loss": 0.6208,
      "mean_token_accuracy": 0.8024071604013443,
      "num_tokens": 718101920.0,
      "step": 22501
    },
    {
      "entropy": 0.4480118490755558,
      "epoch": 2.070292867781094,
      "grad_norm": 0.15738900005817413,
      "learning_rate": 3.099334498727267e-05,
      "loss": 0.4342,
      "mean_token_accuracy": 0.8601987808942795,
      "num_tokens": 718133607.0,
      "step": 22502
    },
    {
      "entropy": 0.4615822425112128,
      "epoch": 2.070384874152535,
      "grad_norm": 0.15276533365249634,
      "learning_rate": 3.099027816113105e-05,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8593147583305836,
      "num_tokens": 718164955.0,
      "step": 22503
    },
    {
      "entropy": 0.5609776983037591,
      "epoch": 2.070476880523976,
      "grad_norm": 0.16820675134658813,
      "learning_rate": 3.098721133498942e-05,
      "loss": 0.5423,
      "mean_token_accuracy": 0.8299048617482185,
      "num_tokens": 718196612.0,
      "step": 22504
    },
    {
      "entropy": 0.5248153731226921,
      "epoch": 2.0705688868954173,
      "grad_norm": 0.1595345139503479,
      "learning_rate": 3.098414450884779e-05,
      "loss": 0.5109,
      "mean_token_accuracy": 0.8364371843636036,
      "num_tokens": 718228798.0,
      "step": 22505
    },
    {
      "entropy": 0.5986238839104772,
      "epoch": 2.070660893266859,
      "grad_norm": 0.1714630126953125,
      "learning_rate": 3.098107768270617e-05,
      "loss": 0.5871,
      "mean_token_accuracy": 0.8130847997963428,
      "num_tokens": 718260767.0,
      "step": 22506
    },
    {
      "entropy": 0.437597269192338,
      "epoch": 2.0707528996383,
      "grad_norm": 0.1548762172460556,
      "learning_rate": 3.097801085656454e-05,
      "loss": 0.4275,
      "mean_token_accuracy": 0.8650286644697189,
      "num_tokens": 718291952.0,
      "step": 22507
    },
    {
      "entropy": 0.5291469970252365,
      "epoch": 2.070844906009741,
      "grad_norm": 0.1716245859861374,
      "learning_rate": 3.097494403042292e-05,
      "loss": 0.516,
      "mean_token_accuracy": 0.8372568488121033,
      "num_tokens": 718323493.0,
      "step": 22508
    },
    {
      "entropy": 0.4732934460043907,
      "epoch": 2.0709369123811823,
      "grad_norm": 0.15529221296310425,
      "learning_rate": 3.097187720428129e-05,
      "loss": 0.4676,
      "mean_token_accuracy": 0.8509267903864384,
      "num_tokens": 718355370.0,
      "step": 22509
    },
    {
      "entropy": 0.5190760064870119,
      "epoch": 2.0710289187526234,
      "grad_norm": 0.16269822418689728,
      "learning_rate": 3.096881037813966e-05,
      "loss": 0.5066,
      "mean_token_accuracy": 0.8401665985584259,
      "num_tokens": 718387786.0,
      "step": 22510
    },
    {
      "entropy": 0.5523649908136576,
      "epoch": 2.071120925124065,
      "grad_norm": 0.16242070496082306,
      "learning_rate": 3.096574355199804e-05,
      "loss": 0.5503,
      "mean_token_accuracy": 0.8291402012109756,
      "num_tokens": 718420554.0,
      "step": 22511
    },
    {
      "entropy": 0.61929683201015,
      "epoch": 2.071212931495506,
      "grad_norm": 0.17526674270629883,
      "learning_rate": 3.0962676725856416e-05,
      "loss": 0.6141,
      "mean_token_accuracy": 0.808822151273489,
      "num_tokens": 718452655.0,
      "step": 22512
    },
    {
      "entropy": 0.6273379642516375,
      "epoch": 2.0713049378669472,
      "grad_norm": 0.17049720883369446,
      "learning_rate": 3.0959609899714784e-05,
      "loss": 0.6112,
      "mean_token_accuracy": 0.8092010729014874,
      "num_tokens": 718484870.0,
      "step": 22513
    },
    {
      "entropy": 0.5088786708656698,
      "epoch": 2.0713969442383884,
      "grad_norm": 0.15580791234970093,
      "learning_rate": 3.0956543073573165e-05,
      "loss": 0.4904,
      "mean_token_accuracy": 0.8454614542424679,
      "num_tokens": 718516951.0,
      "step": 22514
    },
    {
      "entropy": 0.452512483112514,
      "epoch": 2.0714889506098295,
      "grad_norm": 0.15371684730052948,
      "learning_rate": 3.095347624743153e-05,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8617212250828743,
      "num_tokens": 718549014.0,
      "step": 22515
    },
    {
      "entropy": 0.5258276101667434,
      "epoch": 2.071580956981271,
      "grad_norm": 0.1569945365190506,
      "learning_rate": 3.0950409421289914e-05,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8409387692809105,
      "num_tokens": 718580944.0,
      "step": 22516
    },
    {
      "entropy": 0.572980772703886,
      "epoch": 2.0716729633527122,
      "grad_norm": 0.16629423201084137,
      "learning_rate": 3.094734259514828e-05,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8263727277517319,
      "num_tokens": 718612227.0,
      "step": 22517
    },
    {
      "entropy": 0.656570291146636,
      "epoch": 2.0717649697241534,
      "grad_norm": 0.17531004548072815,
      "learning_rate": 3.094427576900666e-05,
      "loss": 0.6237,
      "mean_token_accuracy": 0.8016046658158302,
      "num_tokens": 718643716.0,
      "step": 22518
    },
    {
      "entropy": 0.55153464153409,
      "epoch": 2.0718569760955945,
      "grad_norm": 0.16713915765285492,
      "learning_rate": 3.094120894286503e-05,
      "loss": 0.5386,
      "mean_token_accuracy": 0.8286857530474663,
      "num_tokens": 718674911.0,
      "step": 22519
    },
    {
      "entropy": 0.4894107412546873,
      "epoch": 2.0719489824670356,
      "grad_norm": 0.155349999666214,
      "learning_rate": 3.0938142116723406e-05,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8487933278083801,
      "num_tokens": 718706549.0,
      "step": 22520
    },
    {
      "entropy": 0.53153245896101,
      "epoch": 2.072040988838477,
      "grad_norm": 0.1574290543794632,
      "learning_rate": 3.093507529058178e-05,
      "loss": 0.5186,
      "mean_token_accuracy": 0.8356811180710793,
      "num_tokens": 718738437.0,
      "step": 22521
    },
    {
      "entropy": 0.5922348275780678,
      "epoch": 2.0721329952099183,
      "grad_norm": 0.16618651151657104,
      "learning_rate": 3.0932008464440155e-05,
      "loss": 0.5677,
      "mean_token_accuracy": 0.8215662278234959,
      "num_tokens": 718770473.0,
      "step": 22522
    },
    {
      "entropy": 0.47837487049400806,
      "epoch": 2.0722250015813595,
      "grad_norm": 0.157260000705719,
      "learning_rate": 3.092894163829852e-05,
      "loss": 0.4523,
      "mean_token_accuracy": 0.8555097654461861,
      "num_tokens": 718802291.0,
      "step": 22523
    },
    {
      "entropy": 0.4976952946744859,
      "epoch": 2.0723170079528006,
      "grad_norm": 0.15119726955890656,
      "learning_rate": 3.0925874812156905e-05,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8489757291972637,
      "num_tokens": 718834287.0,
      "step": 22524
    },
    {
      "entropy": 0.4965582387521863,
      "epoch": 2.0724090143242417,
      "grad_norm": 0.15735697746276855,
      "learning_rate": 3.092280798601527e-05,
      "loss": 0.4641,
      "mean_token_accuracy": 0.8526610843837261,
      "num_tokens": 718865991.0,
      "step": 22525
    },
    {
      "entropy": 0.5044826925732195,
      "epoch": 2.0725010206956833,
      "grad_norm": 0.1608879417181015,
      "learning_rate": 3.091974115987365e-05,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8410166949033737,
      "num_tokens": 718898251.0,
      "step": 22526
    },
    {
      "entropy": 0.5668950788676739,
      "epoch": 2.0725930270671244,
      "grad_norm": 0.17147494852542877,
      "learning_rate": 3.091667433373202e-05,
      "loss": 0.5641,
      "mean_token_accuracy": 0.823049996048212,
      "num_tokens": 718930207.0,
      "step": 22527
    },
    {
      "entropy": 0.530939313583076,
      "epoch": 2.0726850334385656,
      "grad_norm": 0.16229519248008728,
      "learning_rate": 3.0913607507590397e-05,
      "loss": 0.514,
      "mean_token_accuracy": 0.8373416773974895,
      "num_tokens": 718961725.0,
      "step": 22528
    },
    {
      "entropy": 0.4584774738177657,
      "epoch": 2.0727770398100067,
      "grad_norm": 0.15550482273101807,
      "learning_rate": 3.091054068144877e-05,
      "loss": 0.4378,
      "mean_token_accuracy": 0.8574099577963352,
      "num_tokens": 718994225.0,
      "step": 22529
    },
    {
      "entropy": 0.5272893439978361,
      "epoch": 2.072869046181448,
      "grad_norm": 0.16496379673480988,
      "learning_rate": 3.0907473855307146e-05,
      "loss": 0.5088,
      "mean_token_accuracy": 0.8346210047602654,
      "num_tokens": 719026017.0,
      "step": 22530
    },
    {
      "entropy": 0.44138250476680696,
      "epoch": 2.0729610525528894,
      "grad_norm": 0.14939987659454346,
      "learning_rate": 3.0904407029165514e-05,
      "loss": 0.419,
      "mean_token_accuracy": 0.8668557740747929,
      "num_tokens": 719057625.0,
      "step": 22531
    },
    {
      "entropy": 0.5541886482387781,
      "epoch": 2.0730530589243306,
      "grad_norm": 0.15959087014198303,
      "learning_rate": 3.0901340203023895e-05,
      "loss": 0.5484,
      "mean_token_accuracy": 0.8277806490659714,
      "num_tokens": 719089978.0,
      "step": 22532
    },
    {
      "entropy": 0.4733476359397173,
      "epoch": 2.0731450652957717,
      "grad_norm": 0.15646299719810486,
      "learning_rate": 3.089827337688226e-05,
      "loss": 0.4701,
      "mean_token_accuracy": 0.8494286574423313,
      "num_tokens": 719122155.0,
      "step": 22533
    },
    {
      "entropy": 0.6280725114047527,
      "epoch": 2.073237071667213,
      "grad_norm": 0.16958780586719513,
      "learning_rate": 3.089520655074064e-05,
      "loss": 0.6227,
      "mean_token_accuracy": 0.8039645813405514,
      "num_tokens": 719154467.0,
      "step": 22534
    },
    {
      "entropy": 0.6322058103978634,
      "epoch": 2.073329078038654,
      "grad_norm": 0.1691163182258606,
      "learning_rate": 3.089213972459901e-05,
      "loss": 0.6134,
      "mean_token_accuracy": 0.8081145659089088,
      "num_tokens": 719186993.0,
      "step": 22535
    },
    {
      "entropy": 0.5669971024617553,
      "epoch": 2.0734210844100955,
      "grad_norm": 0.16824273765087128,
      "learning_rate": 3.088907289845739e-05,
      "loss": 0.543,
      "mean_token_accuracy": 0.8271967433393002,
      "num_tokens": 719218894.0,
      "step": 22536
    },
    {
      "entropy": 0.4245100733824074,
      "epoch": 2.0735130907815367,
      "grad_norm": 0.1477014273405075,
      "learning_rate": 3.088600607231576e-05,
      "loss": 0.4095,
      "mean_token_accuracy": 0.8681325316429138,
      "num_tokens": 719251533.0,
      "step": 22537
    },
    {
      "entropy": 0.4722290802747011,
      "epoch": 2.073605097152978,
      "grad_norm": 0.1517152041196823,
      "learning_rate": 3.0882939246174136e-05,
      "loss": 0.4533,
      "mean_token_accuracy": 0.854509100317955,
      "num_tokens": 719283826.0,
      "step": 22538
    },
    {
      "entropy": 0.5429545028600842,
      "epoch": 2.073697103524419,
      "grad_norm": 0.17009182274341583,
      "learning_rate": 3.0879872420032504e-05,
      "loss": 0.5306,
      "mean_token_accuracy": 0.8359270095825195,
      "num_tokens": 719316232.0,
      "step": 22539
    },
    {
      "entropy": 0.4358648834750056,
      "epoch": 2.07378910989586,
      "grad_norm": 0.15690799057483673,
      "learning_rate": 3.0876805593890886e-05,
      "loss": 0.4256,
      "mean_token_accuracy": 0.8621373698115349,
      "num_tokens": 719348275.0,
      "step": 22540
    },
    {
      "entropy": 0.436158403288573,
      "epoch": 2.0738811162673016,
      "grad_norm": 0.15620774030685425,
      "learning_rate": 3.0873738767749253e-05,
      "loss": 0.4238,
      "mean_token_accuracy": 0.8623386211693287,
      "num_tokens": 719380261.0,
      "step": 22541
    },
    {
      "entropy": 0.5600282177329063,
      "epoch": 2.073973122638743,
      "grad_norm": 0.16308076679706573,
      "learning_rate": 3.0870671941607635e-05,
      "loss": 0.544,
      "mean_token_accuracy": 0.8287603929638863,
      "num_tokens": 719412426.0,
      "step": 22542
    },
    {
      "entropy": 0.49678672943264246,
      "epoch": 2.074065129010184,
      "grad_norm": 0.15605904161930084,
      "learning_rate": 3.086760511546601e-05,
      "loss": 0.4815,
      "mean_token_accuracy": 0.8449519723653793,
      "num_tokens": 719444202.0,
      "step": 22543
    },
    {
      "entropy": 0.45340087078511715,
      "epoch": 2.074157135381625,
      "grad_norm": 0.15449485182762146,
      "learning_rate": 3.086453828932438e-05,
      "loss": 0.4423,
      "mean_token_accuracy": 0.8567115440964699,
      "num_tokens": 719476342.0,
      "step": 22544
    },
    {
      "entropy": 0.4043725626543164,
      "epoch": 2.074249141753066,
      "grad_norm": 0.1444244384765625,
      "learning_rate": 3.086147146318276e-05,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8785021789371967,
      "num_tokens": 719506992.0,
      "step": 22545
    },
    {
      "entropy": 0.5464800205081701,
      "epoch": 2.0743411481245078,
      "grad_norm": 0.16348131000995636,
      "learning_rate": 3.085840463704113e-05,
      "loss": 0.5224,
      "mean_token_accuracy": 0.836534708738327,
      "num_tokens": 719538750.0,
      "step": 22546
    },
    {
      "entropy": 0.5504935719072819,
      "epoch": 2.074433154495949,
      "grad_norm": 0.16544227302074432,
      "learning_rate": 3.08553378108995e-05,
      "loss": 0.528,
      "mean_token_accuracy": 0.8326019532978535,
      "num_tokens": 719570260.0,
      "step": 22547
    },
    {
      "entropy": 0.594648003578186,
      "epoch": 2.07452516086739,
      "grad_norm": 0.1658642590045929,
      "learning_rate": 3.0852270984757876e-05,
      "loss": 0.5839,
      "mean_token_accuracy": 0.8184863887727261,
      "num_tokens": 719602573.0,
      "step": 22548
    },
    {
      "entropy": 0.5553689356893301,
      "epoch": 2.074617167238831,
      "grad_norm": 0.16967372596263885,
      "learning_rate": 3.084920415861625e-05,
      "loss": 0.5495,
      "mean_token_accuracy": 0.8286796770989895,
      "num_tokens": 719633889.0,
      "step": 22549
    },
    {
      "entropy": 0.5578251518309116,
      "epoch": 2.0747091736102723,
      "grad_norm": 0.16687436401844025,
      "learning_rate": 3.0846137332474625e-05,
      "loss": 0.5454,
      "mean_token_accuracy": 0.829148668795824,
      "num_tokens": 719666439.0,
      "step": 22550
    },
    {
      "entropy": 0.5057460945099592,
      "epoch": 2.074801179981714,
      "grad_norm": 0.16621002554893494,
      "learning_rate": 3.0843070506333e-05,
      "loss": 0.5024,
      "mean_token_accuracy": 0.840907346457243,
      "num_tokens": 719698004.0,
      "step": 22551
    },
    {
      "entropy": 0.5010785087943077,
      "epoch": 2.074893186353155,
      "grad_norm": 0.161811962723732,
      "learning_rate": 3.084000368019137e-05,
      "loss": 0.4951,
      "mean_token_accuracy": 0.845108263194561,
      "num_tokens": 719730329.0,
      "step": 22552
    },
    {
      "entropy": 0.5685391016304493,
      "epoch": 2.074985192724596,
      "grad_norm": 0.16631820797920227,
      "learning_rate": 3.083693685404975e-05,
      "loss": 0.5627,
      "mean_token_accuracy": 0.8251144029200077,
      "num_tokens": 719762573.0,
      "step": 22553
    },
    {
      "entropy": 0.5267471056431532,
      "epoch": 2.0750771990960373,
      "grad_norm": 0.16545510292053223,
      "learning_rate": 3.083387002790812e-05,
      "loss": 0.5054,
      "mean_token_accuracy": 0.8386750258505344,
      "num_tokens": 719795015.0,
      "step": 22554
    },
    {
      "entropy": 0.5679234396666288,
      "epoch": 2.0751692054674784,
      "grad_norm": 0.16714540123939514,
      "learning_rate": 3.083080320176649e-05,
      "loss": 0.5578,
      "mean_token_accuracy": 0.8273204974830151,
      "num_tokens": 719827286.0,
      "step": 22555
    },
    {
      "entropy": 0.5935678454115987,
      "epoch": 2.07526121183892,
      "grad_norm": 0.16623583436012268,
      "learning_rate": 3.0827736375624866e-05,
      "loss": 0.5583,
      "mean_token_accuracy": 0.8237094059586525,
      "num_tokens": 719859543.0,
      "step": 22556
    },
    {
      "entropy": 0.5299446526914835,
      "epoch": 2.075353218210361,
      "grad_norm": 0.15902763605117798,
      "learning_rate": 3.082466954948324e-05,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8349569737911224,
      "num_tokens": 719891238.0,
      "step": 22557
    },
    {
      "entropy": 0.45986591232940555,
      "epoch": 2.0754452245818023,
      "grad_norm": 0.15679100155830383,
      "learning_rate": 3.0821602723341616e-05,
      "loss": 0.4359,
      "mean_token_accuracy": 0.8618350178003311,
      "num_tokens": 719923351.0,
      "step": 22558
    },
    {
      "entropy": 0.5146314166486263,
      "epoch": 2.0755372309532434,
      "grad_norm": 0.15602745115756989,
      "learning_rate": 3.081853589719999e-05,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8418434336781502,
      "num_tokens": 719955244.0,
      "step": 22559
    },
    {
      "entropy": 0.5436616563238204,
      "epoch": 2.0756292373246845,
      "grad_norm": 0.16312162578105927,
      "learning_rate": 3.0815469071058365e-05,
      "loss": 0.5265,
      "mean_token_accuracy": 0.832718800753355,
      "num_tokens": 719987061.0,
      "step": 22560
    },
    {
      "entropy": 0.5123800367582589,
      "epoch": 2.075721243696126,
      "grad_norm": 0.1685447245836258,
      "learning_rate": 3.081240224491674e-05,
      "loss": 0.4945,
      "mean_token_accuracy": 0.8403621837496758,
      "num_tokens": 720018462.0,
      "step": 22561
    },
    {
      "entropy": 0.5181693257763982,
      "epoch": 2.0758132500675672,
      "grad_norm": 0.16140618920326233,
      "learning_rate": 3.080933541877511e-05,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8391697742044926,
      "num_tokens": 720049823.0,
      "step": 22562
    },
    {
      "entropy": 0.5568770859390497,
      "epoch": 2.0759052564390084,
      "grad_norm": 0.16340476274490356,
      "learning_rate": 3.080626859263349e-05,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8328065797686577,
      "num_tokens": 720081545.0,
      "step": 22563
    },
    {
      "entropy": 0.4919820027425885,
      "epoch": 2.0759972628104495,
      "grad_norm": 0.15959589183330536,
      "learning_rate": 3.080320176649186e-05,
      "loss": 0.4823,
      "mean_token_accuracy": 0.848013024777174,
      "num_tokens": 720113886.0,
      "step": 22564
    },
    {
      "entropy": 0.4787563467398286,
      "epoch": 2.0760892691818906,
      "grad_norm": 0.1581002175807953,
      "learning_rate": 3.080013494035023e-05,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8548755459487438,
      "num_tokens": 720145538.0,
      "step": 22565
    },
    {
      "entropy": 0.5870666615664959,
      "epoch": 2.076181275553332,
      "grad_norm": 0.16676996648311615,
      "learning_rate": 3.0797068114208606e-05,
      "loss": 0.5698,
      "mean_token_accuracy": 0.8229273520410061,
      "num_tokens": 720177311.0,
      "step": 22566
    },
    {
      "entropy": 0.4627276798710227,
      "epoch": 2.0762732819247733,
      "grad_norm": 0.15415510535240173,
      "learning_rate": 3.079400128806698e-05,
      "loss": 0.4466,
      "mean_token_accuracy": 0.8571675643324852,
      "num_tokens": 720209673.0,
      "step": 22567
    },
    {
      "entropy": 0.5576327573508024,
      "epoch": 2.0763652882962145,
      "grad_norm": 0.15984414517879486,
      "learning_rate": 3.0790934461925355e-05,
      "loss": 0.538,
      "mean_token_accuracy": 0.8280765786767006,
      "num_tokens": 720241061.0,
      "step": 22568
    },
    {
      "entropy": 0.4806949719786644,
      "epoch": 2.0764572946676556,
      "grad_norm": 0.15968336164951324,
      "learning_rate": 3.078786763578373e-05,
      "loss": 0.4604,
      "mean_token_accuracy": 0.8499222733080387,
      "num_tokens": 720272598.0,
      "step": 22569
    },
    {
      "entropy": 0.48153159907087684,
      "epoch": 2.0765493010390967,
      "grad_norm": 0.16090616583824158,
      "learning_rate": 3.07848008096421e-05,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8494507968425751,
      "num_tokens": 720303933.0,
      "step": 22570
    },
    {
      "entropy": 0.5023899609223008,
      "epoch": 2.0766413074105383,
      "grad_norm": 0.15855565667152405,
      "learning_rate": 3.078173398350048e-05,
      "loss": 0.4806,
      "mean_token_accuracy": 0.8449181020259857,
      "num_tokens": 720335889.0,
      "step": 22571
    },
    {
      "entropy": 0.5656631216406822,
      "epoch": 2.0767333137819795,
      "grad_norm": 0.16095690429210663,
      "learning_rate": 3.077866715735885e-05,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8260751999914646,
      "num_tokens": 720368038.0,
      "step": 22572
    },
    {
      "entropy": 0.5480859586969018,
      "epoch": 2.0768253201534206,
      "grad_norm": 0.16360196471214294,
      "learning_rate": 3.077560033121722e-05,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8264410495758057,
      "num_tokens": 720400150.0,
      "step": 22573
    },
    {
      "entropy": 0.5535930935293436,
      "epoch": 2.0769173265248617,
      "grad_norm": 0.16531498730182648,
      "learning_rate": 3.07725335050756e-05,
      "loss": 0.544,
      "mean_token_accuracy": 0.8260465115308762,
      "num_tokens": 720432625.0,
      "step": 22574
    },
    {
      "entropy": 0.4977374644950032,
      "epoch": 2.077009332896303,
      "grad_norm": 0.16220466792583466,
      "learning_rate": 3.076946667893397e-05,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8414218612015247,
      "num_tokens": 720464313.0,
      "step": 22575
    },
    {
      "entropy": 0.4972231015563011,
      "epoch": 2.0771013392677444,
      "grad_norm": 0.15682946145534515,
      "learning_rate": 3.076639985279235e-05,
      "loss": 0.482,
      "mean_token_accuracy": 0.848536778241396,
      "num_tokens": 720496995.0,
      "step": 22576
    },
    {
      "entropy": 0.5579053526744246,
      "epoch": 2.0771933456391856,
      "grad_norm": 0.1641456037759781,
      "learning_rate": 3.076333302665072e-05,
      "loss": 0.5456,
      "mean_token_accuracy": 0.8283017054200172,
      "num_tokens": 720529333.0,
      "step": 22577
    },
    {
      "entropy": 0.41892953030765057,
      "epoch": 2.0772853520106267,
      "grad_norm": 0.1520862579345703,
      "learning_rate": 3.0760266200509095e-05,
      "loss": 0.4112,
      "mean_token_accuracy": 0.868813619017601,
      "num_tokens": 720561218.0,
      "step": 22578
    },
    {
      "entropy": 0.42933069728314877,
      "epoch": 2.077377358382068,
      "grad_norm": 0.15384574234485626,
      "learning_rate": 3.075719937436747e-05,
      "loss": 0.4105,
      "mean_token_accuracy": 0.8715993836522102,
      "num_tokens": 720593000.0,
      "step": 22579
    },
    {
      "entropy": 0.3838691655546427,
      "epoch": 2.077469364753509,
      "grad_norm": 0.14378710091114044,
      "learning_rate": 3.0754132548225844e-05,
      "loss": 0.3628,
      "mean_token_accuracy": 0.8809254318475723,
      "num_tokens": 720625389.0,
      "step": 22580
    },
    {
      "entropy": 0.4898495078086853,
      "epoch": 2.0775613711249505,
      "grad_norm": 0.15989993512630463,
      "learning_rate": 3.075106572208422e-05,
      "loss": 0.4725,
      "mean_token_accuracy": 0.8447154834866524,
      "num_tokens": 720657072.0,
      "step": 22581
    },
    {
      "entropy": 0.4956659357994795,
      "epoch": 2.0776533774963917,
      "grad_norm": 0.1620018482208252,
      "learning_rate": 3.0747998895942594e-05,
      "loss": 0.4811,
      "mean_token_accuracy": 0.8489988222718239,
      "num_tokens": 720689096.0,
      "step": 22582
    },
    {
      "entropy": 0.5219261867459863,
      "epoch": 2.077745383867833,
      "grad_norm": 0.16011540591716766,
      "learning_rate": 3.074493206980096e-05,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8410855457186699,
      "num_tokens": 720721346.0,
      "step": 22583
    },
    {
      "entropy": 0.5880298260599375,
      "epoch": 2.077837390239274,
      "grad_norm": 0.16656021773815155,
      "learning_rate": 3.074186524365934e-05,
      "loss": 0.5598,
      "mean_token_accuracy": 0.8235405832529068,
      "num_tokens": 720752426.0,
      "step": 22584
    },
    {
      "entropy": 0.4441067073494196,
      "epoch": 2.077929396610715,
      "grad_norm": 0.15104345977306366,
      "learning_rate": 3.073879841751771e-05,
      "loss": 0.4297,
      "mean_token_accuracy": 0.8629213310778141,
      "num_tokens": 720784182.0,
      "step": 22585
    },
    {
      "entropy": 0.5278945676982403,
      "epoch": 2.0780214029821567,
      "grad_norm": 0.16231532394886017,
      "learning_rate": 3.0735731591376085e-05,
      "loss": 0.515,
      "mean_token_accuracy": 0.8401350975036621,
      "num_tokens": 720816340.0,
      "step": 22586
    },
    {
      "entropy": 0.5051824199035764,
      "epoch": 2.078113409353598,
      "grad_norm": 0.15923957526683807,
      "learning_rate": 3.073266476523446e-05,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8424668461084366,
      "num_tokens": 720848112.0,
      "step": 22587
    },
    {
      "entropy": 0.4494704070966691,
      "epoch": 2.078205415725039,
      "grad_norm": 0.14671386778354645,
      "learning_rate": 3.0729597939092835e-05,
      "loss": 0.434,
      "mean_token_accuracy": 0.862687099725008,
      "num_tokens": 720880576.0,
      "step": 22588
    },
    {
      "entropy": 0.4987973254173994,
      "epoch": 2.07829742209648,
      "grad_norm": 0.15894679725170135,
      "learning_rate": 3.072653111295121e-05,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8456215560436249,
      "num_tokens": 720912597.0,
      "step": 22589
    },
    {
      "entropy": 0.616643431596458,
      "epoch": 2.078389428467921,
      "grad_norm": 0.1687268167734146,
      "learning_rate": 3.0723464286809584e-05,
      "loss": 0.6011,
      "mean_token_accuracy": 0.8104203715920448,
      "num_tokens": 720945060.0,
      "step": 22590
    },
    {
      "entropy": 0.48653932474553585,
      "epoch": 2.0784814348393628,
      "grad_norm": 0.15622884035110474,
      "learning_rate": 3.072039746066795e-05,
      "loss": 0.4586,
      "mean_token_accuracy": 0.8537812940776348,
      "num_tokens": 720975960.0,
      "step": 22591
    },
    {
      "entropy": 0.54939493839629,
      "epoch": 2.078573441210804,
      "grad_norm": 0.16171397268772125,
      "learning_rate": 3.071733063452633e-05,
      "loss": 0.537,
      "mean_token_accuracy": 0.829080119729042,
      "num_tokens": 721007846.0,
      "step": 22592
    },
    {
      "entropy": 0.5244200103916228,
      "epoch": 2.078665447582245,
      "grad_norm": 0.16348187625408173,
      "learning_rate": 3.07142638083847e-05,
      "loss": 0.5148,
      "mean_token_accuracy": 0.8381061851978302,
      "num_tokens": 721039726.0,
      "step": 22593
    },
    {
      "entropy": 0.5173857798799872,
      "epoch": 2.078757453953686,
      "grad_norm": 0.1629278063774109,
      "learning_rate": 3.0711196982243076e-05,
      "loss": 0.4928,
      "mean_token_accuracy": 0.841568935662508,
      "num_tokens": 721070941.0,
      "step": 22594
    },
    {
      "entropy": 0.4965411841403693,
      "epoch": 2.0788494603251273,
      "grad_norm": 0.15612883865833282,
      "learning_rate": 3.070813015610145e-05,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8459314033389091,
      "num_tokens": 721103109.0,
      "step": 22595
    },
    {
      "entropy": 0.5276910979300737,
      "epoch": 2.078941466696569,
      "grad_norm": 0.16135689616203308,
      "learning_rate": 3.0705063329959825e-05,
      "loss": 0.5101,
      "mean_token_accuracy": 0.8379104360938072,
      "num_tokens": 721134758.0,
      "step": 22596
    },
    {
      "entropy": 0.5196482138708234,
      "epoch": 2.07903347306801,
      "grad_norm": 0.15476740896701813,
      "learning_rate": 3.07019965038182e-05,
      "loss": 0.5038,
      "mean_token_accuracy": 0.8414584696292877,
      "num_tokens": 721166896.0,
      "step": 22597
    },
    {
      "entropy": 0.5542828552424908,
      "epoch": 2.079125479439451,
      "grad_norm": 0.1622430682182312,
      "learning_rate": 3.0698929677676574e-05,
      "loss": 0.5358,
      "mean_token_accuracy": 0.8318337351083755,
      "num_tokens": 721199403.0,
      "step": 22598
    },
    {
      "entropy": 0.5511038843542337,
      "epoch": 2.0792174858108923,
      "grad_norm": 0.1629953384399414,
      "learning_rate": 3.069586285153494e-05,
      "loss": 0.5481,
      "mean_token_accuracy": 0.8249555565416813,
      "num_tokens": 721231592.0,
      "step": 22599
    },
    {
      "entropy": 0.4982480965554714,
      "epoch": 2.0793094921823334,
      "grad_norm": 0.15905970335006714,
      "learning_rate": 3.0692796025393324e-05,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8475290983915329,
      "num_tokens": 721262834.0,
      "step": 22600
    },
    {
      "entropy": 0.6363533269613981,
      "epoch": 2.079401498553775,
      "grad_norm": 0.1666005402803421,
      "learning_rate": 3.068972919925169e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8083412237465382,
      "num_tokens": 721294669.0,
      "step": 22601
    },
    {
      "entropy": 0.6211478682234883,
      "epoch": 2.079493504925216,
      "grad_norm": 0.1710897535085678,
      "learning_rate": 3.068666237311007e-05,
      "loss": 0.5932,
      "mean_token_accuracy": 0.8133144490420818,
      "num_tokens": 721327300.0,
      "step": 22602
    },
    {
      "entropy": 0.3932115286588669,
      "epoch": 2.0795855112966573,
      "grad_norm": 0.1415194571018219,
      "learning_rate": 3.068359554696844e-05,
      "loss": 0.3739,
      "mean_token_accuracy": 0.8786620646715164,
      "num_tokens": 721359502.0,
      "step": 22603
    },
    {
      "entropy": 0.5089999279007316,
      "epoch": 2.0796775176680984,
      "grad_norm": 0.16443093121051788,
      "learning_rate": 3.0680528720826816e-05,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8438733145594597,
      "num_tokens": 721391144.0,
      "step": 22604
    },
    {
      "entropy": 0.5083377556875348,
      "epoch": 2.0797695240395395,
      "grad_norm": 0.1608549952507019,
      "learning_rate": 3.06774618946852e-05,
      "loss": 0.5069,
      "mean_token_accuracy": 0.8422052189707756,
      "num_tokens": 721422843.0,
      "step": 22605
    },
    {
      "entropy": 0.5706372410058975,
      "epoch": 2.079861530410981,
      "grad_norm": 0.16559307277202606,
      "learning_rate": 3.0674395068543565e-05,
      "loss": 0.5536,
      "mean_token_accuracy": 0.8279781751334667,
      "num_tokens": 721455030.0,
      "step": 22606
    },
    {
      "entropy": 0.5690503586083651,
      "epoch": 2.0799535367824222,
      "grad_norm": 0.16744834184646606,
      "learning_rate": 3.067132824240194e-05,
      "loss": 0.55,
      "mean_token_accuracy": 0.824751615524292,
      "num_tokens": 721486574.0,
      "step": 22607
    },
    {
      "entropy": 0.5074295625090599,
      "epoch": 2.0800455431538634,
      "grad_norm": 0.15577276051044464,
      "learning_rate": 3.0668261416260314e-05,
      "loss": 0.4974,
      "mean_token_accuracy": 0.8416871689260006,
      "num_tokens": 721518943.0,
      "step": 22608
    },
    {
      "entropy": 0.5435142358765006,
      "epoch": 2.0801375495253045,
      "grad_norm": 0.15836459398269653,
      "learning_rate": 3.066519459011869e-05,
      "loss": 0.5213,
      "mean_token_accuracy": 0.8371380940079689,
      "num_tokens": 721551185.0,
      "step": 22609
    },
    {
      "entropy": 0.519513338804245,
      "epoch": 2.0802295558967456,
      "grad_norm": 0.15635977685451508,
      "learning_rate": 3.0662127763977063e-05,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8394112177193165,
      "num_tokens": 721583538.0,
      "step": 22610
    },
    {
      "entropy": 0.5454235412180424,
      "epoch": 2.080321562268187,
      "grad_norm": 0.16060446202754974,
      "learning_rate": 3.065906093783544e-05,
      "loss": 0.5329,
      "mean_token_accuracy": 0.8299616612493992,
      "num_tokens": 721615968.0,
      "step": 22611
    },
    {
      "entropy": 0.5396291008219123,
      "epoch": 2.0804135686396283,
      "grad_norm": 0.16797882318496704,
      "learning_rate": 3.0655994111693806e-05,
      "loss": 0.5242,
      "mean_token_accuracy": 0.833745714277029,
      "num_tokens": 721647499.0,
      "step": 22612
    },
    {
      "entropy": 0.5181564278900623,
      "epoch": 2.0805055750110695,
      "grad_norm": 0.16154004633426666,
      "learning_rate": 3.065292728555219e-05,
      "loss": 0.511,
      "mean_token_accuracy": 0.8381890207529068,
      "num_tokens": 721679812.0,
      "step": 22613
    },
    {
      "entropy": 0.42860220512375236,
      "epoch": 2.0805975813825106,
      "grad_norm": 0.14986734092235565,
      "learning_rate": 3.0649860459410555e-05,
      "loss": 0.4126,
      "mean_token_accuracy": 0.867897093296051,
      "num_tokens": 721711676.0,
      "step": 22614
    },
    {
      "entropy": 0.5080153634771705,
      "epoch": 2.0806895877539517,
      "grad_norm": 0.15795816481113434,
      "learning_rate": 3.064679363326893e-05,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8425615131855011,
      "num_tokens": 721743555.0,
      "step": 22615
    },
    {
      "entropy": 0.5384909035637975,
      "epoch": 2.0807815941253933,
      "grad_norm": 0.16105018556118011,
      "learning_rate": 3.0643726807127305e-05,
      "loss": 0.5208,
      "mean_token_accuracy": 0.8338423930108547,
      "num_tokens": 721775328.0,
      "step": 22616
    },
    {
      "entropy": 0.4728115666657686,
      "epoch": 2.0808736004968345,
      "grad_norm": 0.15375925600528717,
      "learning_rate": 3.064065998098568e-05,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8522279374301434,
      "num_tokens": 721807058.0,
      "step": 22617
    },
    {
      "entropy": 0.5039577260613441,
      "epoch": 2.0809656068682756,
      "grad_norm": 0.16406436264514923,
      "learning_rate": 3.0637593154844054e-05,
      "loss": 0.4995,
      "mean_token_accuracy": 0.8363539017736912,
      "num_tokens": 721839100.0,
      "step": 22618
    },
    {
      "entropy": 0.6069460427388549,
      "epoch": 2.0810576132397167,
      "grad_norm": 0.16717177629470825,
      "learning_rate": 3.063452632870243e-05,
      "loss": 0.5884,
      "mean_token_accuracy": 0.8149127513170242,
      "num_tokens": 721871142.0,
      "step": 22619
    },
    {
      "entropy": 0.5723409745842218,
      "epoch": 2.081149619611158,
      "grad_norm": 0.1679980605840683,
      "learning_rate": 3.06314595025608e-05,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8244184292852879,
      "num_tokens": 721902477.0,
      "step": 22620
    },
    {
      "entropy": 0.4813967514783144,
      "epoch": 2.0812416259825994,
      "grad_norm": 0.1568002849817276,
      "learning_rate": 3.062839267641918e-05,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8522522896528244,
      "num_tokens": 721934990.0,
      "step": 22621
    },
    {
      "entropy": 0.4812219990417361,
      "epoch": 2.0813336323540406,
      "grad_norm": 0.15969696640968323,
      "learning_rate": 3.0625325850277546e-05,
      "loss": 0.4764,
      "mean_token_accuracy": 0.8532866835594177,
      "num_tokens": 721966487.0,
      "step": 22622
    },
    {
      "entropy": 0.5934542305767536,
      "epoch": 2.0814256387254817,
      "grad_norm": 0.16744619607925415,
      "learning_rate": 3.062225902413593e-05,
      "loss": 0.5666,
      "mean_token_accuracy": 0.8219959400594234,
      "num_tokens": 721997313.0,
      "step": 22623
    },
    {
      "entropy": 0.5566432978957891,
      "epoch": 2.081517645096923,
      "grad_norm": 0.1668982058763504,
      "learning_rate": 3.0619192197994295e-05,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8305371999740601,
      "num_tokens": 722028664.0,
      "step": 22624
    },
    {
      "entropy": 0.6495349267497659,
      "epoch": 2.081609651468364,
      "grad_norm": 0.1672535538673401,
      "learning_rate": 3.061612537185267e-05,
      "loss": 0.621,
      "mean_token_accuracy": 0.8040989451110363,
      "num_tokens": 722060551.0,
      "step": 22625
    },
    {
      "entropy": 0.5764811635017395,
      "epoch": 2.0817016578398055,
      "grad_norm": 0.1670469045639038,
      "learning_rate": 3.0613058545711044e-05,
      "loss": 0.5562,
      "mean_token_accuracy": 0.8233490623533726,
      "num_tokens": 722091870.0,
      "step": 22626
    },
    {
      "entropy": 0.5117785464972258,
      "epoch": 2.0817936642112467,
      "grad_norm": 0.16676920652389526,
      "learning_rate": 3.060999171956942e-05,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8426318541169167,
      "num_tokens": 722123683.0,
      "step": 22627
    },
    {
      "entropy": 0.5314843724481761,
      "epoch": 2.081885670582688,
      "grad_norm": 0.16275812685489655,
      "learning_rate": 3.0606924893427794e-05,
      "loss": 0.5186,
      "mean_token_accuracy": 0.8369144350290298,
      "num_tokens": 722155295.0,
      "step": 22628
    },
    {
      "entropy": 0.4837530176155269,
      "epoch": 2.081977676954129,
      "grad_norm": 0.1564638316631317,
      "learning_rate": 3.060385806728617e-05,
      "loss": 0.4765,
      "mean_token_accuracy": 0.8530063144862652,
      "num_tokens": 722187161.0,
      "step": 22629
    },
    {
      "entropy": 0.481517571490258,
      "epoch": 2.08206968332557,
      "grad_norm": 0.15443682670593262,
      "learning_rate": 3.0600791241144536e-05,
      "loss": 0.4684,
      "mean_token_accuracy": 0.847659271210432,
      "num_tokens": 722219648.0,
      "step": 22630
    },
    {
      "entropy": 0.5376769285649061,
      "epoch": 2.0821616896970117,
      "grad_norm": 0.1591399759054184,
      "learning_rate": 3.059772441500292e-05,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8358337879180908,
      "num_tokens": 722252216.0,
      "step": 22631
    },
    {
      "entropy": 0.515537329018116,
      "epoch": 2.082253696068453,
      "grad_norm": 0.16367511451244354,
      "learning_rate": 3.0594657588861285e-05,
      "loss": 0.5082,
      "mean_token_accuracy": 0.8394016064703465,
      "num_tokens": 722284020.0,
      "step": 22632
    },
    {
      "entropy": 0.5573376677930355,
      "epoch": 2.082345702439894,
      "grad_norm": 0.1605912744998932,
      "learning_rate": 3.059159076271966e-05,
      "loss": 0.5463,
      "mean_token_accuracy": 0.8272853940725327,
      "num_tokens": 722316383.0,
      "step": 22633
    },
    {
      "entropy": 0.4605006882920861,
      "epoch": 2.082437708811335,
      "grad_norm": 0.15049533545970917,
      "learning_rate": 3.058852393657804e-05,
      "loss": 0.4372,
      "mean_token_accuracy": 0.8572398945689201,
      "num_tokens": 722347786.0,
      "step": 22634
    },
    {
      "entropy": 0.6075412668287754,
      "epoch": 2.082529715182776,
      "grad_norm": 0.17069770395755768,
      "learning_rate": 3.058545711043641e-05,
      "loss": 0.5941,
      "mean_token_accuracy": 0.8121353462338448,
      "num_tokens": 722379802.0,
      "step": 22635
    },
    {
      "entropy": 0.5210925135761499,
      "epoch": 2.0826217215542178,
      "grad_norm": 0.1582501232624054,
      "learning_rate": 3.058239028429479e-05,
      "loss": 0.5114,
      "mean_token_accuracy": 0.844231266528368,
      "num_tokens": 722411861.0,
      "step": 22636
    },
    {
      "entropy": 0.4666674882173538,
      "epoch": 2.082713727925659,
      "grad_norm": 0.1527312994003296,
      "learning_rate": 3.057932345815316e-05,
      "loss": 0.4472,
      "mean_token_accuracy": 0.8565014824271202,
      "num_tokens": 722443697.0,
      "step": 22637
    },
    {
      "entropy": 0.4882065085694194,
      "epoch": 2.0828057342971,
      "grad_norm": 0.15531671047210693,
      "learning_rate": 3.057625663201153e-05,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8490068130195141,
      "num_tokens": 722475549.0,
      "step": 22638
    },
    {
      "entropy": 0.5030930847860873,
      "epoch": 2.082897740668541,
      "grad_norm": 0.15955471992492676,
      "learning_rate": 3.057318980586991e-05,
      "loss": 0.4865,
      "mean_token_accuracy": 0.846695076674223,
      "num_tokens": 722507721.0,
      "step": 22639
    },
    {
      "entropy": 0.507505807094276,
      "epoch": 2.0829897470399823,
      "grad_norm": 0.15381258726119995,
      "learning_rate": 3.057012297972828e-05,
      "loss": 0.4949,
      "mean_token_accuracy": 0.8433988727629185,
      "num_tokens": 722540489.0,
      "step": 22640
    },
    {
      "entropy": 0.49395220330916345,
      "epoch": 2.083081753411424,
      "grad_norm": 0.15151914954185486,
      "learning_rate": 3.056705615358666e-05,
      "loss": 0.4641,
      "mean_token_accuracy": 0.8506293259561062,
      "num_tokens": 722572356.0,
      "step": 22641
    },
    {
      "entropy": 0.45433490350842476,
      "epoch": 2.083173759782865,
      "grad_norm": 0.15457969903945923,
      "learning_rate": 3.056398932744503e-05,
      "loss": 0.4373,
      "mean_token_accuracy": 0.86109584197402,
      "num_tokens": 722603649.0,
      "step": 22642
    },
    {
      "entropy": 0.4837009513285011,
      "epoch": 2.083265766154306,
      "grad_norm": 0.15746985375881195,
      "learning_rate": 3.05609225013034e-05,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8524166606366634,
      "num_tokens": 722634209.0,
      "step": 22643
    },
    {
      "entropy": 0.4753837934695184,
      "epoch": 2.0833577725257473,
      "grad_norm": 0.15583892166614532,
      "learning_rate": 3.055785567516178e-05,
      "loss": 0.471,
      "mean_token_accuracy": 0.8500385880470276,
      "num_tokens": 722666616.0,
      "step": 22644
    },
    {
      "entropy": 0.4421670166775584,
      "epoch": 2.0834497788971884,
      "grad_norm": 0.14972972869873047,
      "learning_rate": 3.055478884902015e-05,
      "loss": 0.4278,
      "mean_token_accuracy": 0.8612549230456352,
      "num_tokens": 722699312.0,
      "step": 22645
    },
    {
      "entropy": 0.5609650230035186,
      "epoch": 2.08354178526863,
      "grad_norm": 0.1677136868238449,
      "learning_rate": 3.0551722022878524e-05,
      "loss": 0.546,
      "mean_token_accuracy": 0.8258668892085552,
      "num_tokens": 722731441.0,
      "step": 22646
    },
    {
      "entropy": 0.5044717928394675,
      "epoch": 2.083633791640071,
      "grad_norm": 0.15212129056453705,
      "learning_rate": 3.05486551967369e-05,
      "loss": 0.4773,
      "mean_token_accuracy": 0.849085621535778,
      "num_tokens": 722762927.0,
      "step": 22647
    },
    {
      "entropy": 0.4301251005381346,
      "epoch": 2.0837257980115123,
      "grad_norm": 0.1491972953081131,
      "learning_rate": 3.054558837059527e-05,
      "loss": 0.4167,
      "mean_token_accuracy": 0.8638844192028046,
      "num_tokens": 722795132.0,
      "step": 22648
    },
    {
      "entropy": 0.5323199518024921,
      "epoch": 2.0838178043829534,
      "grad_norm": 0.16089734435081482,
      "learning_rate": 3.054252154445365e-05,
      "loss": 0.5291,
      "mean_token_accuracy": 0.8308222815394402,
      "num_tokens": 722826930.0,
      "step": 22649
    },
    {
      "entropy": 0.4560493705794215,
      "epoch": 2.0839098107543945,
      "grad_norm": 0.15002304315567017,
      "learning_rate": 3.053945471831202e-05,
      "loss": 0.4432,
      "mean_token_accuracy": 0.8560800403356552,
      "num_tokens": 722859124.0,
      "step": 22650
    },
    {
      "entropy": 0.5233243987895548,
      "epoch": 2.084001817125836,
      "grad_norm": 0.16531811654567719,
      "learning_rate": 3.053638789217039e-05,
      "loss": 0.5127,
      "mean_token_accuracy": 0.8377574123442173,
      "num_tokens": 722890538.0,
      "step": 22651
    },
    {
      "entropy": 0.5687402095645666,
      "epoch": 2.0840938234972772,
      "grad_norm": 0.16639727354049683,
      "learning_rate": 3.053332106602877e-05,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8233859017491341,
      "num_tokens": 722923032.0,
      "step": 22652
    },
    {
      "entropy": 0.5283712544478476,
      "epoch": 2.0841858298687184,
      "grad_norm": 0.16707593202590942,
      "learning_rate": 3.053025423988714e-05,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8378100469708443,
      "num_tokens": 722955198.0,
      "step": 22653
    },
    {
      "entropy": 0.576538348570466,
      "epoch": 2.0842778362401595,
      "grad_norm": 0.16311787068843842,
      "learning_rate": 3.0527187413745514e-05,
      "loss": 0.572,
      "mean_token_accuracy": 0.8229531347751617,
      "num_tokens": 722987756.0,
      "step": 22654
    },
    {
      "entropy": 0.5351309925317764,
      "epoch": 2.0843698426116006,
      "grad_norm": 0.166739821434021,
      "learning_rate": 3.052412058760389e-05,
      "loss": 0.5143,
      "mean_token_accuracy": 0.8380737528204918,
      "num_tokens": 723018897.0,
      "step": 22655
    },
    {
      "entropy": 0.5629710517823696,
      "epoch": 2.084461848983042,
      "grad_norm": 0.17259447276592255,
      "learning_rate": 3.052105376146226e-05,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8205326721072197,
      "num_tokens": 723049724.0,
      "step": 22656
    },
    {
      "entropy": 0.4622397917555645,
      "epoch": 2.0845538553544833,
      "grad_norm": 0.14931432902812958,
      "learning_rate": 3.051798693532064e-05,
      "loss": 0.4509,
      "mean_token_accuracy": 0.85354895144701,
      "num_tokens": 723081543.0,
      "step": 22657
    },
    {
      "entropy": 0.534039399586618,
      "epoch": 2.0846458617259245,
      "grad_norm": 0.16407673060894012,
      "learning_rate": 3.0514920109179013e-05,
      "loss": 0.5208,
      "mean_token_accuracy": 0.8356180377304554,
      "num_tokens": 723112655.0,
      "step": 22658
    },
    {
      "entropy": 0.5201686955988407,
      "epoch": 2.0847378680973656,
      "grad_norm": 0.16637717187404633,
      "learning_rate": 3.0511853283037384e-05,
      "loss": 0.5089,
      "mean_token_accuracy": 0.836250189691782,
      "num_tokens": 723144176.0,
      "step": 22659
    },
    {
      "entropy": 0.4760120874270797,
      "epoch": 2.0848298744688067,
      "grad_norm": 0.15267518162727356,
      "learning_rate": 3.0508786456895762e-05,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8523350283503532,
      "num_tokens": 723176419.0,
      "step": 22660
    },
    {
      "entropy": 0.47486637718975544,
      "epoch": 2.0849218808402483,
      "grad_norm": 0.1592400223016739,
      "learning_rate": 3.0505719630754133e-05,
      "loss": 0.4646,
      "mean_token_accuracy": 0.8530729860067368,
      "num_tokens": 723208620.0,
      "step": 22661
    },
    {
      "entropy": 0.40702193742617965,
      "epoch": 2.0850138872116895,
      "grad_norm": 0.13903111219406128,
      "learning_rate": 3.0502652804612508e-05,
      "loss": 0.3934,
      "mean_token_accuracy": 0.8733656480908394,
      "num_tokens": 723241083.0,
      "step": 22662
    },
    {
      "entropy": 0.4997535403817892,
      "epoch": 2.0851058935831306,
      "grad_norm": 0.15692393481731415,
      "learning_rate": 3.049958597847088e-05,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8423359207808971,
      "num_tokens": 723272941.0,
      "step": 22663
    },
    {
      "entropy": 0.4762855074368417,
      "epoch": 2.0851978999545717,
      "grad_norm": 0.15924972295761108,
      "learning_rate": 3.0496519152329257e-05,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8516641370952129,
      "num_tokens": 723304777.0,
      "step": 22664
    },
    {
      "entropy": 0.5134716285392642,
      "epoch": 2.085289906326013,
      "grad_norm": 0.15754380822181702,
      "learning_rate": 3.0493452326187632e-05,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8430336266756058,
      "num_tokens": 723337073.0,
      "step": 22665
    },
    {
      "entropy": 0.43394848029129207,
      "epoch": 2.0853819126974544,
      "grad_norm": 0.14966024458408356,
      "learning_rate": 3.0490385500046003e-05,
      "loss": 0.4201,
      "mean_token_accuracy": 0.8655837960541248,
      "num_tokens": 723368735.0,
      "step": 22666
    },
    {
      "entropy": 0.509617100469768,
      "epoch": 2.0854739190688956,
      "grad_norm": 0.15594637393951416,
      "learning_rate": 3.048731867390438e-05,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8446916118264198,
      "num_tokens": 723401067.0,
      "step": 22667
    },
    {
      "entropy": 0.624544421210885,
      "epoch": 2.0855659254403367,
      "grad_norm": 0.1696287989616394,
      "learning_rate": 3.0484251847762752e-05,
      "loss": 0.6153,
      "mean_token_accuracy": 0.8060802295804024,
      "num_tokens": 723433267.0,
      "step": 22668
    },
    {
      "entropy": 0.5269936583936214,
      "epoch": 2.085657931811778,
      "grad_norm": 0.16296808421611786,
      "learning_rate": 3.0481185021621127e-05,
      "loss": 0.5075,
      "mean_token_accuracy": 0.8372278325259686,
      "num_tokens": 723464601.0,
      "step": 22669
    },
    {
      "entropy": 0.5009394725784659,
      "epoch": 2.085749938183219,
      "grad_norm": 0.1535264402627945,
      "learning_rate": 3.0478118195479498e-05,
      "loss": 0.484,
      "mean_token_accuracy": 0.8477104678750038,
      "num_tokens": 723496823.0,
      "step": 22670
    },
    {
      "entropy": 0.5335398847237229,
      "epoch": 2.0858419445546605,
      "grad_norm": 0.15759040415287018,
      "learning_rate": 3.0475051369337876e-05,
      "loss": 0.501,
      "mean_token_accuracy": 0.8381397873163223,
      "num_tokens": 723528489.0,
      "step": 22671
    },
    {
      "entropy": 0.5660706292837858,
      "epoch": 2.0859339509261017,
      "grad_norm": 0.1691991239786148,
      "learning_rate": 3.0471984543196248e-05,
      "loss": 0.5497,
      "mean_token_accuracy": 0.8232786618173122,
      "num_tokens": 723559657.0,
      "step": 22672
    },
    {
      "entropy": 0.4977656463161111,
      "epoch": 2.086025957297543,
      "grad_norm": 0.15800558030605316,
      "learning_rate": 3.0468917717054622e-05,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8439127020537853,
      "num_tokens": 723592205.0,
      "step": 22673
    },
    {
      "entropy": 0.5732260956428945,
      "epoch": 2.086117963668984,
      "grad_norm": 0.1671362966299057,
      "learning_rate": 3.0465850890912993e-05,
      "loss": 0.5525,
      "mean_token_accuracy": 0.8245277591049671,
      "num_tokens": 723624360.0,
      "step": 22674
    },
    {
      "entropy": 0.5961032034829259,
      "epoch": 2.086209970040425,
      "grad_norm": 0.17145542800426483,
      "learning_rate": 3.046278406477137e-05,
      "loss": 0.5748,
      "mean_token_accuracy": 0.8170061632990837,
      "num_tokens": 723656141.0,
      "step": 22675
    },
    {
      "entropy": 0.5445612776093185,
      "epoch": 2.0863019764118667,
      "grad_norm": 0.16849204897880554,
      "learning_rate": 3.0459717238629743e-05,
      "loss": 0.5367,
      "mean_token_accuracy": 0.8312562592327595,
      "num_tokens": 723687520.0,
      "step": 22676
    },
    {
      "entropy": 0.49948121048510075,
      "epoch": 2.086393982783308,
      "grad_norm": 0.16481544077396393,
      "learning_rate": 3.045665041248812e-05,
      "loss": 0.4839,
      "mean_token_accuracy": 0.843768410384655,
      "num_tokens": 723718958.0,
      "step": 22677
    },
    {
      "entropy": 0.4459152175113559,
      "epoch": 2.086485989154749,
      "grad_norm": 0.15248799324035645,
      "learning_rate": 3.045358358634649e-05,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8605087660253048,
      "num_tokens": 723750890.0,
      "step": 22678
    },
    {
      "entropy": 0.45227269921451807,
      "epoch": 2.08657799552619,
      "grad_norm": 0.15139180421829224,
      "learning_rate": 3.0450516760204867e-05,
      "loss": 0.4364,
      "mean_token_accuracy": 0.8615966364741325,
      "num_tokens": 723783382.0,
      "step": 22679
    },
    {
      "entropy": 0.5545263476669788,
      "epoch": 2.086670001897631,
      "grad_norm": 0.16521403193473816,
      "learning_rate": 3.0447449934063238e-05,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8266307339072227,
      "num_tokens": 723814960.0,
      "step": 22680
    },
    {
      "entropy": 0.4401194825768471,
      "epoch": 2.0867620082690728,
      "grad_norm": 0.15268176794052124,
      "learning_rate": 3.0444383107921616e-05,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8604314029216766,
      "num_tokens": 723846913.0,
      "step": 22681
    },
    {
      "entropy": 0.5002221604809165,
      "epoch": 2.086854014640514,
      "grad_norm": 0.1645459681749344,
      "learning_rate": 3.0441316281779987e-05,
      "loss": 0.4884,
      "mean_token_accuracy": 0.8439496010541916,
      "num_tokens": 723878907.0,
      "step": 22682
    },
    {
      "entropy": 0.42754787066951394,
      "epoch": 2.086946021011955,
      "grad_norm": 0.15736572444438934,
      "learning_rate": 3.0438249455638362e-05,
      "loss": 0.4162,
      "mean_token_accuracy": 0.8662350997328758,
      "num_tokens": 723911145.0,
      "step": 22683
    },
    {
      "entropy": 0.44708637334406376,
      "epoch": 2.087038027383396,
      "grad_norm": 0.14667744934558868,
      "learning_rate": 3.0435182629496733e-05,
      "loss": 0.4277,
      "mean_token_accuracy": 0.8632644005119801,
      "num_tokens": 723943674.0,
      "step": 22684
    },
    {
      "entropy": 0.5404684850946069,
      "epoch": 2.0871300337548373,
      "grad_norm": 0.16135652363300323,
      "learning_rate": 3.043211580335511e-05,
      "loss": 0.5321,
      "mean_token_accuracy": 0.831237580627203,
      "num_tokens": 723975293.0,
      "step": 22685
    },
    {
      "entropy": 0.5420705555006862,
      "epoch": 2.087222040126279,
      "grad_norm": 0.15903827548027039,
      "learning_rate": 3.0429048977213482e-05,
      "loss": 0.5228,
      "mean_token_accuracy": 0.831231027841568,
      "num_tokens": 724007088.0,
      "step": 22686
    },
    {
      "entropy": 0.5017264834605157,
      "epoch": 2.08731404649772,
      "grad_norm": 0.16742907464504242,
      "learning_rate": 3.0425982151071857e-05,
      "loss": 0.4881,
      "mean_token_accuracy": 0.8439029157161713,
      "num_tokens": 724039227.0,
      "step": 22687
    },
    {
      "entropy": 0.5596952829509974,
      "epoch": 2.087406052869161,
      "grad_norm": 0.16802679002285004,
      "learning_rate": 3.042291532493023e-05,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8283824995160103,
      "num_tokens": 724070673.0,
      "step": 22688
    },
    {
      "entropy": 0.5393780088052154,
      "epoch": 2.0874980592406023,
      "grad_norm": 0.16689707338809967,
      "learning_rate": 3.0419848498788606e-05,
      "loss": 0.5221,
      "mean_token_accuracy": 0.8369407989084721,
      "num_tokens": 724103220.0,
      "step": 22689
    },
    {
      "entropy": 0.5420981366187334,
      "epoch": 2.0875900656120434,
      "grad_norm": 0.16236340999603271,
      "learning_rate": 3.0416781672646978e-05,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8343356624245644,
      "num_tokens": 724134603.0,
      "step": 22690
    },
    {
      "entropy": 0.4911290151067078,
      "epoch": 2.087682071983485,
      "grad_norm": 0.15404446423053741,
      "learning_rate": 3.0413714846505352e-05,
      "loss": 0.4682,
      "mean_token_accuracy": 0.8532623834908009,
      "num_tokens": 724167070.0,
      "step": 22691
    },
    {
      "entropy": 0.4733845670707524,
      "epoch": 2.087774078354926,
      "grad_norm": 0.15243186056613922,
      "learning_rate": 3.0410648020363724e-05,
      "loss": 0.4559,
      "mean_token_accuracy": 0.8543476834893227,
      "num_tokens": 724199356.0,
      "step": 22692
    },
    {
      "entropy": 0.5780572611838579,
      "epoch": 2.0878660847263673,
      "grad_norm": 0.16774141788482666,
      "learning_rate": 3.04075811942221e-05,
      "loss": 0.5649,
      "mean_token_accuracy": 0.8217285424470901,
      "num_tokens": 724230441.0,
      "step": 22693
    },
    {
      "entropy": 0.5605301221366972,
      "epoch": 2.0879580910978084,
      "grad_norm": 0.16461580991744995,
      "learning_rate": 3.0404514368080473e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8303635381162167,
      "num_tokens": 724261818.0,
      "step": 22694
    },
    {
      "entropy": 0.5251771421171725,
      "epoch": 2.0880500974692495,
      "grad_norm": 0.15527576208114624,
      "learning_rate": 3.0401447541938847e-05,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8376518599689007,
      "num_tokens": 724293672.0,
      "step": 22695
    },
    {
      "entropy": 0.5809805728495121,
      "epoch": 2.088142103840691,
      "grad_norm": 0.16780243813991547,
      "learning_rate": 3.0398380715797226e-05,
      "loss": 0.5618,
      "mean_token_accuracy": 0.8219581842422485,
      "num_tokens": 724325528.0,
      "step": 22696
    },
    {
      "entropy": 0.5409061787649989,
      "epoch": 2.0882341102121322,
      "grad_norm": 0.16479003429412842,
      "learning_rate": 3.0395313889655597e-05,
      "loss": 0.5249,
      "mean_token_accuracy": 0.8378648050129414,
      "num_tokens": 724358009.0,
      "step": 22697
    },
    {
      "entropy": 0.6085918601602316,
      "epoch": 2.0883261165835734,
      "grad_norm": 0.1707119643688202,
      "learning_rate": 3.0392247063513975e-05,
      "loss": 0.5924,
      "mean_token_accuracy": 0.812815960496664,
      "num_tokens": 724389855.0,
      "step": 22698
    },
    {
      "entropy": 0.5035363670904189,
      "epoch": 2.0884181229550145,
      "grad_norm": 0.16543586552143097,
      "learning_rate": 3.0389180237372346e-05,
      "loss": 0.4927,
      "mean_token_accuracy": 0.8451118506491184,
      "num_tokens": 724421828.0,
      "step": 22699
    },
    {
      "entropy": 0.596744691953063,
      "epoch": 2.0885101293264556,
      "grad_norm": 0.1729719489812851,
      "learning_rate": 3.038611341123072e-05,
      "loss": 0.5876,
      "mean_token_accuracy": 0.8135011605918407,
      "num_tokens": 724453850.0,
      "step": 22700
    },
    {
      "entropy": 0.5468511413782835,
      "epoch": 2.088602135697897,
      "grad_norm": 0.1624235212802887,
      "learning_rate": 3.0383046585089092e-05,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8339385204017162,
      "num_tokens": 724485733.0,
      "step": 22701
    },
    {
      "entropy": 0.5210913696791977,
      "epoch": 2.0886941420693383,
      "grad_norm": 0.16847124695777893,
      "learning_rate": 3.037997975894747e-05,
      "loss": 0.4974,
      "mean_token_accuracy": 0.8420913331210613,
      "num_tokens": 724516984.0,
      "step": 22702
    },
    {
      "entropy": 0.4958121506497264,
      "epoch": 2.0887861484407795,
      "grad_norm": 0.15785051882266998,
      "learning_rate": 3.037691293280584e-05,
      "loss": 0.4842,
      "mean_token_accuracy": 0.844660434871912,
      "num_tokens": 724548723.0,
      "step": 22703
    },
    {
      "entropy": 0.4584088136907667,
      "epoch": 2.0888781548122206,
      "grad_norm": 0.15118694305419922,
      "learning_rate": 3.0373846106664216e-05,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8595810867846012,
      "num_tokens": 724581491.0,
      "step": 22704
    },
    {
      "entropy": 0.4624634359497577,
      "epoch": 2.0889701611836617,
      "grad_norm": 0.15451878309249878,
      "learning_rate": 3.0370779280522587e-05,
      "loss": 0.4607,
      "mean_token_accuracy": 0.8532929234206676,
      "num_tokens": 724613968.0,
      "step": 22705
    },
    {
      "entropy": 0.4605192095041275,
      "epoch": 2.0890621675551033,
      "grad_norm": 0.16404372453689575,
      "learning_rate": 3.0367712454380965e-05,
      "loss": 0.4318,
      "mean_token_accuracy": 0.8580794371664524,
      "num_tokens": 724645861.0,
      "step": 22706
    },
    {
      "entropy": 0.4665903365239501,
      "epoch": 2.0891541739265445,
      "grad_norm": 0.1572660207748413,
      "learning_rate": 3.0364645628239336e-05,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8580810762941837,
      "num_tokens": 724677040.0,
      "step": 22707
    },
    {
      "entropy": 0.5015923976898193,
      "epoch": 2.0892461802979856,
      "grad_norm": 0.1589721441268921,
      "learning_rate": 3.036157880209771e-05,
      "loss": 0.5005,
      "mean_token_accuracy": 0.839436337351799,
      "num_tokens": 724709230.0,
      "step": 22708
    },
    {
      "entropy": 0.5184937084559351,
      "epoch": 2.0893381866694267,
      "grad_norm": 0.15914104878902435,
      "learning_rate": 3.0358511975956082e-05,
      "loss": 0.514,
      "mean_token_accuracy": 0.8380148224532604,
      "num_tokens": 724741348.0,
      "step": 22709
    },
    {
      "entropy": 0.5347001999616623,
      "epoch": 2.089430193040868,
      "grad_norm": 0.16608285903930664,
      "learning_rate": 3.035544514981446e-05,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8366141431033611,
      "num_tokens": 724773742.0,
      "step": 22710
    },
    {
      "entropy": 0.5222738794982433,
      "epoch": 2.0895221994123094,
      "grad_norm": 0.16598473489284515,
      "learning_rate": 3.035237832367283e-05,
      "loss": 0.5016,
      "mean_token_accuracy": 0.8449919819831848,
      "num_tokens": 724805844.0,
      "step": 22711
    },
    {
      "entropy": 0.4746474423445761,
      "epoch": 2.0896142057837506,
      "grad_norm": 0.15768009424209595,
      "learning_rate": 3.0349311497531206e-05,
      "loss": 0.4662,
      "mean_token_accuracy": 0.8525005169212818,
      "num_tokens": 724838114.0,
      "step": 22712
    },
    {
      "entropy": 0.5671631435398012,
      "epoch": 2.0897062121551917,
      "grad_norm": 0.1714625060558319,
      "learning_rate": 3.0346244671389578e-05,
      "loss": 0.5472,
      "mean_token_accuracy": 0.8267799951136112,
      "num_tokens": 724869234.0,
      "step": 22713
    },
    {
      "entropy": 0.5177236888557673,
      "epoch": 2.089798218526633,
      "grad_norm": 0.16259124875068665,
      "learning_rate": 3.0343177845247956e-05,
      "loss": 0.5073,
      "mean_token_accuracy": 0.8426484130322933,
      "num_tokens": 724901644.0,
      "step": 22714
    },
    {
      "entropy": 0.560981348156929,
      "epoch": 2.089890224898074,
      "grad_norm": 0.16765642166137695,
      "learning_rate": 3.0340111019106327e-05,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8238176442682743,
      "num_tokens": 724933092.0,
      "step": 22715
    },
    {
      "entropy": 0.6301073096692562,
      "epoch": 2.0899822312695155,
      "grad_norm": 0.17237482964992523,
      "learning_rate": 3.03370441929647e-05,
      "loss": 0.613,
      "mean_token_accuracy": 0.8091348595917225,
      "num_tokens": 724965120.0,
      "step": 22716
    },
    {
      "entropy": 0.6148909702897072,
      "epoch": 2.0900742376409567,
      "grad_norm": 0.17390578985214233,
      "learning_rate": 3.0333977366823073e-05,
      "loss": 0.602,
      "mean_token_accuracy": 0.8097187504172325,
      "num_tokens": 724996858.0,
      "step": 22717
    },
    {
      "entropy": 0.4264287131372839,
      "epoch": 2.090166244012398,
      "grad_norm": 0.14364002645015717,
      "learning_rate": 3.033091054068145e-05,
      "loss": 0.4079,
      "mean_token_accuracy": 0.8705874048173428,
      "num_tokens": 725029107.0,
      "step": 22718
    },
    {
      "entropy": 0.5315373986959457,
      "epoch": 2.090258250383839,
      "grad_norm": 0.16393113136291504,
      "learning_rate": 3.0327843714539822e-05,
      "loss": 0.5063,
      "mean_token_accuracy": 0.8403581716120243,
      "num_tokens": 725061695.0,
      "step": 22719
    },
    {
      "entropy": 0.5655374191701412,
      "epoch": 2.09035025675528,
      "grad_norm": 0.16963298618793488,
      "learning_rate": 3.03247768883982e-05,
      "loss": 0.5521,
      "mean_token_accuracy": 0.8285065591335297,
      "num_tokens": 725094134.0,
      "step": 22720
    },
    {
      "entropy": 0.4662358155474067,
      "epoch": 2.0904422631267217,
      "grad_norm": 0.1525449901819229,
      "learning_rate": 3.032171006225657e-05,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8546404577791691,
      "num_tokens": 725125976.0,
      "step": 22721
    },
    {
      "entropy": 0.4414218133315444,
      "epoch": 2.090534269498163,
      "grad_norm": 0.15478484332561493,
      "learning_rate": 3.0318643236114946e-05,
      "loss": 0.4244,
      "mean_token_accuracy": 0.8627953492105007,
      "num_tokens": 725157717.0,
      "step": 22722
    },
    {
      "entropy": 0.5869180760346353,
      "epoch": 2.090626275869604,
      "grad_norm": 0.16949094831943512,
      "learning_rate": 3.0315576409973317e-05,
      "loss": 0.566,
      "mean_token_accuracy": 0.821231197565794,
      "num_tokens": 725189525.0,
      "step": 22723
    },
    {
      "entropy": 0.5408402262255549,
      "epoch": 2.090718282241045,
      "grad_norm": 0.16120101511478424,
      "learning_rate": 3.0312509583831695e-05,
      "loss": 0.5184,
      "mean_token_accuracy": 0.8361307792365551,
      "num_tokens": 725221163.0,
      "step": 22724
    },
    {
      "entropy": 0.48542527575045824,
      "epoch": 2.090810288612486,
      "grad_norm": 0.15979181230068207,
      "learning_rate": 3.0309442757690067e-05,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8511014617979527,
      "num_tokens": 725252686.0,
      "step": 22725
    },
    {
      "entropy": 0.521264654584229,
      "epoch": 2.0909022949839278,
      "grad_norm": 0.159068301320076,
      "learning_rate": 3.030637593154844e-05,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8414465188980103,
      "num_tokens": 725284633.0,
      "step": 22726
    },
    {
      "entropy": 0.4642672936897725,
      "epoch": 2.090994301355369,
      "grad_norm": 0.1500619649887085,
      "learning_rate": 3.030330910540682e-05,
      "loss": 0.4467,
      "mean_token_accuracy": 0.8582112863659859,
      "num_tokens": 725317116.0,
      "step": 22727
    },
    {
      "entropy": 0.5879319198429585,
      "epoch": 2.09108630772681,
      "grad_norm": 0.17416249215602875,
      "learning_rate": 3.030024227926519e-05,
      "loss": 0.5791,
      "mean_token_accuracy": 0.8159047178924084,
      "num_tokens": 725348649.0,
      "step": 22728
    },
    {
      "entropy": 0.5167377647012472,
      "epoch": 2.091178314098251,
      "grad_norm": 0.15719400346279144,
      "learning_rate": 3.0297175453123565e-05,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8417800031602383,
      "num_tokens": 725381024.0,
      "step": 22729
    },
    {
      "entropy": 0.41242874693125486,
      "epoch": 2.0912703204696923,
      "grad_norm": 0.14976267516613007,
      "learning_rate": 3.0294108626981936e-05,
      "loss": 0.391,
      "mean_token_accuracy": 0.8726533092558384,
      "num_tokens": 725412873.0,
      "step": 22730
    },
    {
      "entropy": 0.518389193341136,
      "epoch": 2.091362326841134,
      "grad_norm": 0.16177359223365784,
      "learning_rate": 3.0291041800840314e-05,
      "loss": 0.507,
      "mean_token_accuracy": 0.8383554108440876,
      "num_tokens": 725445236.0,
      "step": 22731
    },
    {
      "entropy": 0.5130906822159886,
      "epoch": 2.091454333212575,
      "grad_norm": 0.160519078373909,
      "learning_rate": 3.0287974974698686e-05,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8405829332768917,
      "num_tokens": 725476550.0,
      "step": 22732
    },
    {
      "entropy": 0.53745769103989,
      "epoch": 2.091546339584016,
      "grad_norm": 0.1611226499080658,
      "learning_rate": 3.028490814855706e-05,
      "loss": 0.5358,
      "mean_token_accuracy": 0.8328100182116032,
      "num_tokens": 725508436.0,
      "step": 22733
    },
    {
      "entropy": 0.529321288689971,
      "epoch": 2.0916383459554573,
      "grad_norm": 0.1584537774324417,
      "learning_rate": 3.028184132241543e-05,
      "loss": 0.5119,
      "mean_token_accuracy": 0.8381274789571762,
      "num_tokens": 725540139.0,
      "step": 22734
    },
    {
      "entropy": 0.45231388695538044,
      "epoch": 2.0917303523268984,
      "grad_norm": 0.14934398233890533,
      "learning_rate": 3.027877449627381e-05,
      "loss": 0.4392,
      "mean_token_accuracy": 0.8610898852348328,
      "num_tokens": 725572117.0,
      "step": 22735
    },
    {
      "entropy": 0.4205099418759346,
      "epoch": 2.09182235869834,
      "grad_norm": 0.1551571488380432,
      "learning_rate": 3.027570767013218e-05,
      "loss": 0.393,
      "mean_token_accuracy": 0.8730558678507805,
      "num_tokens": 725603756.0,
      "step": 22736
    },
    {
      "entropy": 0.4747325964272022,
      "epoch": 2.091914365069781,
      "grad_norm": 0.150995135307312,
      "learning_rate": 3.027264084399056e-05,
      "loss": 0.4632,
      "mean_token_accuracy": 0.850848950445652,
      "num_tokens": 725635939.0,
      "step": 22737
    },
    {
      "entropy": 0.5614481642842293,
      "epoch": 2.0920063714412223,
      "grad_norm": 0.16935980319976807,
      "learning_rate": 3.0269574017848927e-05,
      "loss": 0.5423,
      "mean_token_accuracy": 0.8315841369330883,
      "num_tokens": 725667512.0,
      "step": 22738
    },
    {
      "entropy": 0.5068080467171967,
      "epoch": 2.0920983778126634,
      "grad_norm": 0.16277417540550232,
      "learning_rate": 3.0266507191707305e-05,
      "loss": 0.4944,
      "mean_token_accuracy": 0.8447112366557121,
      "num_tokens": 725699613.0,
      "step": 22739
    },
    {
      "entropy": 0.48291080724447966,
      "epoch": 2.0921903841841045,
      "grad_norm": 0.15553458034992218,
      "learning_rate": 3.0263440365565676e-05,
      "loss": 0.478,
      "mean_token_accuracy": 0.8487420603632927,
      "num_tokens": 725732381.0,
      "step": 22740
    },
    {
      "entropy": 0.5593564370647073,
      "epoch": 2.092282390555546,
      "grad_norm": 0.1666669100522995,
      "learning_rate": 3.0260373539424054e-05,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8270956464111805,
      "num_tokens": 725763602.0,
      "step": 22741
    },
    {
      "entropy": 0.5113493781536818,
      "epoch": 2.0923743969269872,
      "grad_norm": 0.16114111244678497,
      "learning_rate": 3.0257306713282425e-05,
      "loss": 0.501,
      "mean_token_accuracy": 0.8404207564890385,
      "num_tokens": 725795933.0,
      "step": 22742
    },
    {
      "entropy": 0.539191784337163,
      "epoch": 2.0924664032984284,
      "grad_norm": 0.16552530229091644,
      "learning_rate": 3.02542398871408e-05,
      "loss": 0.5307,
      "mean_token_accuracy": 0.8319472409784794,
      "num_tokens": 725827544.0,
      "step": 22743
    },
    {
      "entropy": 0.4505713451653719,
      "epoch": 2.0925584096698695,
      "grad_norm": 0.15477056801319122,
      "learning_rate": 3.025117306099917e-05,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8609679937362671,
      "num_tokens": 725858510.0,
      "step": 22744
    },
    {
      "entropy": 0.4868224994279444,
      "epoch": 2.0926504160413106,
      "grad_norm": 0.16038039326667786,
      "learning_rate": 3.024810623485755e-05,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8476778417825699,
      "num_tokens": 725890231.0,
      "step": 22745
    },
    {
      "entropy": 0.49844879657030106,
      "epoch": 2.092742422412752,
      "grad_norm": 0.16109995543956757,
      "learning_rate": 3.024503940871592e-05,
      "loss": 0.4882,
      "mean_token_accuracy": 0.8424847684800625,
      "num_tokens": 725921561.0,
      "step": 22746
    },
    {
      "entropy": 0.5597710404545069,
      "epoch": 2.0928344287841933,
      "grad_norm": 0.17541047930717468,
      "learning_rate": 3.0241972582574295e-05,
      "loss": 0.5463,
      "mean_token_accuracy": 0.828353326767683,
      "num_tokens": 725952248.0,
      "step": 22747
    },
    {
      "entropy": 0.5031318422406912,
      "epoch": 2.0929264351556345,
      "grad_norm": 0.1620461493730545,
      "learning_rate": 3.0238905756432667e-05,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8437480330467224,
      "num_tokens": 725983734.0,
      "step": 22748
    },
    {
      "entropy": 0.4455717750824988,
      "epoch": 2.0930184415270756,
      "grad_norm": 0.15120598673820496,
      "learning_rate": 3.0235838930291045e-05,
      "loss": 0.4186,
      "mean_token_accuracy": 0.8665408715605736,
      "num_tokens": 726015607.0,
      "step": 22749
    },
    {
      "entropy": 0.4590162462554872,
      "epoch": 2.0931104478985167,
      "grad_norm": 0.1534135341644287,
      "learning_rate": 3.0232772104149416e-05,
      "loss": 0.4403,
      "mean_token_accuracy": 0.8571567684412003,
      "num_tokens": 726047439.0,
      "step": 22750
    },
    {
      "entropy": 0.5447625732049346,
      "epoch": 2.0932024542699583,
      "grad_norm": 0.16849061846733093,
      "learning_rate": 3.022970527800779e-05,
      "loss": 0.5403,
      "mean_token_accuracy": 0.8319552205502987,
      "num_tokens": 726079153.0,
      "step": 22751
    },
    {
      "entropy": 0.528928394895047,
      "epoch": 2.0932944606413995,
      "grad_norm": 0.15994083881378174,
      "learning_rate": 3.0226638451866162e-05,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8355951905250549,
      "num_tokens": 726110992.0,
      "step": 22752
    },
    {
      "entropy": 0.4619696056470275,
      "epoch": 2.0933864670128406,
      "grad_norm": 0.15334078669548035,
      "learning_rate": 3.022357162572454e-05,
      "loss": 0.4414,
      "mean_token_accuracy": 0.85572225227952,
      "num_tokens": 726143089.0,
      "step": 22753
    },
    {
      "entropy": 0.5737354850862175,
      "epoch": 2.0934784733842817,
      "grad_norm": 0.16592971980571747,
      "learning_rate": 3.022050479958291e-05,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8241360671818256,
      "num_tokens": 726174548.0,
      "step": 22754
    },
    {
      "entropy": 0.5864809481427073,
      "epoch": 2.093570479755723,
      "grad_norm": 0.16963554918766022,
      "learning_rate": 3.0217437973441286e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8219408839941025,
      "num_tokens": 726206226.0,
      "step": 22755
    },
    {
      "entropy": 0.5342822279781103,
      "epoch": 2.0936624861271644,
      "grad_norm": 0.16115470230579376,
      "learning_rate": 3.0214371147299664e-05,
      "loss": 0.5196,
      "mean_token_accuracy": 0.8345535546541214,
      "num_tokens": 726238741.0,
      "step": 22756
    },
    {
      "entropy": 0.5567580058705062,
      "epoch": 2.0937544924986056,
      "grad_norm": 0.1572975516319275,
      "learning_rate": 3.0211304321158035e-05,
      "loss": 0.5333,
      "mean_token_accuracy": 0.8325914293527603,
      "num_tokens": 726271496.0,
      "step": 22757
    },
    {
      "entropy": 0.4254615264944732,
      "epoch": 2.0938464988700467,
      "grad_norm": 0.15295329689979553,
      "learning_rate": 3.0208237495016413e-05,
      "loss": 0.4163,
      "mean_token_accuracy": 0.8693451657891273,
      "num_tokens": 726303310.0,
      "step": 22758
    },
    {
      "entropy": 0.50486752204597,
      "epoch": 2.093938505241488,
      "grad_norm": 0.15493202209472656,
      "learning_rate": 3.0205170668874784e-05,
      "loss": 0.4898,
      "mean_token_accuracy": 0.8454213924705982,
      "num_tokens": 726335277.0,
      "step": 22759
    },
    {
      "entropy": 0.5534613048657775,
      "epoch": 2.094030511612929,
      "grad_norm": 0.16842637956142426,
      "learning_rate": 3.020210384273316e-05,
      "loss": 0.5443,
      "mean_token_accuracy": 0.8295769542455673,
      "num_tokens": 726366624.0,
      "step": 22760
    },
    {
      "entropy": 0.47043783869594336,
      "epoch": 2.0941225179843705,
      "grad_norm": 0.1613083779811859,
      "learning_rate": 3.019903701659153e-05,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8515091426670551,
      "num_tokens": 726398628.0,
      "step": 22761
    },
    {
      "entropy": 0.50244758464396,
      "epoch": 2.0942145243558117,
      "grad_norm": 0.15617568790912628,
      "learning_rate": 3.0195970190449908e-05,
      "loss": 0.4785,
      "mean_token_accuracy": 0.848698191344738,
      "num_tokens": 726430962.0,
      "step": 22762
    },
    {
      "entropy": 0.5401722257956862,
      "epoch": 2.094306530727253,
      "grad_norm": 0.16417647898197174,
      "learning_rate": 3.019290336430828e-05,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8347247429192066,
      "num_tokens": 726462960.0,
      "step": 22763
    },
    {
      "entropy": 0.6406240109354258,
      "epoch": 2.094398537098694,
      "grad_norm": 0.1755877286195755,
      "learning_rate": 3.0189836538166654e-05,
      "loss": 0.6376,
      "mean_token_accuracy": 0.8020535260438919,
      "num_tokens": 726494381.0,
      "step": 22764
    },
    {
      "entropy": 0.44704968109726906,
      "epoch": 2.094490543470135,
      "grad_norm": 0.1539335995912552,
      "learning_rate": 3.0186769712025025e-05,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8620743937790394,
      "num_tokens": 726526607.0,
      "step": 22765
    },
    {
      "entropy": 0.5395822275895625,
      "epoch": 2.0945825498415767,
      "grad_norm": 0.1666349172592163,
      "learning_rate": 3.0183702885883403e-05,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8349475376307964,
      "num_tokens": 726558232.0,
      "step": 22766
    },
    {
      "entropy": 0.48422474414110184,
      "epoch": 2.094674556213018,
      "grad_norm": 0.15357479453086853,
      "learning_rate": 3.0180636059741775e-05,
      "loss": 0.4651,
      "mean_token_accuracy": 0.8518160693347454,
      "num_tokens": 726590674.0,
      "step": 22767
    },
    {
      "entropy": 0.5012301858514547,
      "epoch": 2.094766562584459,
      "grad_norm": 0.1613989621400833,
      "learning_rate": 3.017756923360015e-05,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8457377478480339,
      "num_tokens": 726623269.0,
      "step": 22768
    },
    {
      "entropy": 0.589996799826622,
      "epoch": 2.0948585689559,
      "grad_norm": 0.16999351978302002,
      "learning_rate": 3.017450240745852e-05,
      "loss": 0.5725,
      "mean_token_accuracy": 0.8208954818546772,
      "num_tokens": 726655393.0,
      "step": 22769
    },
    {
      "entropy": 0.6042465064674616,
      "epoch": 2.094950575327341,
      "grad_norm": 0.17603683471679688,
      "learning_rate": 3.01714355813169e-05,
      "loss": 0.5951,
      "mean_token_accuracy": 0.8127878345549107,
      "num_tokens": 726686755.0,
      "step": 22770
    },
    {
      "entropy": 0.46417560148984194,
      "epoch": 2.0950425816987828,
      "grad_norm": 0.15829864144325256,
      "learning_rate": 3.016836875517527e-05,
      "loss": 0.4436,
      "mean_token_accuracy": 0.8598888702690601,
      "num_tokens": 726718614.0,
      "step": 22771
    },
    {
      "entropy": 0.6223050132393837,
      "epoch": 2.095134588070224,
      "grad_norm": 0.17337824404239655,
      "learning_rate": 3.0165301929033645e-05,
      "loss": 0.6051,
      "mean_token_accuracy": 0.8094929866492748,
      "num_tokens": 726749441.0,
      "step": 22772
    },
    {
      "entropy": 0.5850254502147436,
      "epoch": 2.095226594441665,
      "grad_norm": 0.1692347228527069,
      "learning_rate": 3.0162235102892016e-05,
      "loss": 0.5751,
      "mean_token_accuracy": 0.8180831670761108,
      "num_tokens": 726780840.0,
      "step": 22773
    },
    {
      "entropy": 0.6068180073052645,
      "epoch": 2.095318600813106,
      "grad_norm": 0.16515642404556274,
      "learning_rate": 3.0159168276750394e-05,
      "loss": 0.5875,
      "mean_token_accuracy": 0.815558809787035,
      "num_tokens": 726813088.0,
      "step": 22774
    },
    {
      "entropy": 0.6274923104792833,
      "epoch": 2.0954106071845473,
      "grad_norm": 0.17787893116474152,
      "learning_rate": 3.0156101450608765e-05,
      "loss": 0.607,
      "mean_token_accuracy": 0.8094659708440304,
      "num_tokens": 726844706.0,
      "step": 22775
    },
    {
      "entropy": 0.4711465938016772,
      "epoch": 2.095502613555989,
      "grad_norm": 0.15222205221652985,
      "learning_rate": 3.015303462446714e-05,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8496418930590153,
      "num_tokens": 726877306.0,
      "step": 22776
    },
    {
      "entropy": 0.44463398354128003,
      "epoch": 2.09559461992743,
      "grad_norm": 0.15467120707035065,
      "learning_rate": 3.014996779832551e-05,
      "loss": 0.4261,
      "mean_token_accuracy": 0.8611973039805889,
      "num_tokens": 726909831.0,
      "step": 22777
    },
    {
      "entropy": 0.4704787125810981,
      "epoch": 2.095686626298871,
      "grad_norm": 0.15672875940799713,
      "learning_rate": 3.014690097218389e-05,
      "loss": 0.448,
      "mean_token_accuracy": 0.8540919236838818,
      "num_tokens": 726940590.0,
      "step": 22778
    },
    {
      "entropy": 0.46362220495939255,
      "epoch": 2.0957786326703123,
      "grad_norm": 0.1572316735982895,
      "learning_rate": 3.014383414604226e-05,
      "loss": 0.447,
      "mean_token_accuracy": 0.8563845269382,
      "num_tokens": 726972582.0,
      "step": 22779
    },
    {
      "entropy": 0.5298021491616964,
      "epoch": 2.0958706390417534,
      "grad_norm": 0.16024187207221985,
      "learning_rate": 3.014076731990064e-05,
      "loss": 0.503,
      "mean_token_accuracy": 0.8418864570558071,
      "num_tokens": 727003965.0,
      "step": 22780
    },
    {
      "entropy": 0.4920317977666855,
      "epoch": 2.095962645413195,
      "grad_norm": 0.15635794401168823,
      "learning_rate": 3.013770049375901e-05,
      "loss": 0.4821,
      "mean_token_accuracy": 0.8433872610330582,
      "num_tokens": 727035572.0,
      "step": 22781
    },
    {
      "entropy": 0.49790505692362785,
      "epoch": 2.096054651784636,
      "grad_norm": 0.15548066794872284,
      "learning_rate": 3.0134633667617384e-05,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8467058725655079,
      "num_tokens": 727067940.0,
      "step": 22782
    },
    {
      "entropy": 0.4875115496106446,
      "epoch": 2.0961466581560773,
      "grad_norm": 0.15644219517707825,
      "learning_rate": 3.0131566841475755e-05,
      "loss": 0.4721,
      "mean_token_accuracy": 0.850966565310955,
      "num_tokens": 727099966.0,
      "step": 22783
    },
    {
      "entropy": 0.5449781557545066,
      "epoch": 2.0962386645275184,
      "grad_norm": 0.1670367419719696,
      "learning_rate": 3.0128500015334134e-05,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8313714377582073,
      "num_tokens": 727131041.0,
      "step": 22784
    },
    {
      "entropy": 0.5021190978586674,
      "epoch": 2.0963306708989595,
      "grad_norm": 0.16181211173534393,
      "learning_rate": 3.0125433189192505e-05,
      "loss": 0.4899,
      "mean_token_accuracy": 0.8414382077753544,
      "num_tokens": 727162966.0,
      "step": 22785
    },
    {
      "entropy": 0.46850763889960945,
      "epoch": 2.096422677270401,
      "grad_norm": 0.15296275913715363,
      "learning_rate": 3.012236636305088e-05,
      "loss": 0.4535,
      "mean_token_accuracy": 0.8551390804350376,
      "num_tokens": 727195170.0,
      "step": 22786
    },
    {
      "entropy": 0.4730679616332054,
      "epoch": 2.0965146836418422,
      "grad_norm": 0.16145136952400208,
      "learning_rate": 3.0119299536909257e-05,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8490660637617111,
      "num_tokens": 727226695.0,
      "step": 22787
    },
    {
      "entropy": 0.51262859813869,
      "epoch": 2.0966066900132834,
      "grad_norm": 0.16932551562786102,
      "learning_rate": 3.011623271076763e-05,
      "loss": 0.5075,
      "mean_token_accuracy": 0.8381906747817993,
      "num_tokens": 727258534.0,
      "step": 22788
    },
    {
      "entropy": 0.47644015960395336,
      "epoch": 2.0966986963847245,
      "grad_norm": 0.16580671072006226,
      "learning_rate": 3.0113165884626003e-05,
      "loss": 0.4839,
      "mean_token_accuracy": 0.8493456952273846,
      "num_tokens": 727290784.0,
      "step": 22789
    },
    {
      "entropy": 0.578924173489213,
      "epoch": 2.0967907027561656,
      "grad_norm": 0.1758306324481964,
      "learning_rate": 3.0110099058484375e-05,
      "loss": 0.5591,
      "mean_token_accuracy": 0.8232092298567295,
      "num_tokens": 727321887.0,
      "step": 22790
    },
    {
      "entropy": 0.6648473348468542,
      "epoch": 2.096882709127607,
      "grad_norm": 0.1717311590909958,
      "learning_rate": 3.0107032232342753e-05,
      "loss": 0.6523,
      "mean_token_accuracy": 0.7961195595562458,
      "num_tokens": 727354242.0,
      "step": 22791
    },
    {
      "entropy": 0.548109445720911,
      "epoch": 2.0969747154990483,
      "grad_norm": 0.1634720116853714,
      "learning_rate": 3.0103965406201124e-05,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8298429176211357,
      "num_tokens": 727385988.0,
      "step": 22792
    },
    {
      "entropy": 0.5945315128192306,
      "epoch": 2.0970667218704895,
      "grad_norm": 0.17371328175067902,
      "learning_rate": 3.01008985800595e-05,
      "loss": 0.5817,
      "mean_token_accuracy": 0.8194877095520496,
      "num_tokens": 727417626.0,
      "step": 22793
    },
    {
      "entropy": 0.47049490828067064,
      "epoch": 2.0971587282419306,
      "grad_norm": 0.14861030876636505,
      "learning_rate": 3.009783175391787e-05,
      "loss": 0.453,
      "mean_token_accuracy": 0.855347715318203,
      "num_tokens": 727450091.0,
      "step": 22794
    },
    {
      "entropy": 0.560358103364706,
      "epoch": 2.0972507346133717,
      "grad_norm": 0.1655351221561432,
      "learning_rate": 3.0094764927776248e-05,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8277631215751171,
      "num_tokens": 727481334.0,
      "step": 22795
    },
    {
      "entropy": 0.6172696426510811,
      "epoch": 2.0973427409848133,
      "grad_norm": 0.17293430864810944,
      "learning_rate": 3.009169810163462e-05,
      "loss": 0.6156,
      "mean_token_accuracy": 0.8068746514618397,
      "num_tokens": 727513496.0,
      "step": 22796
    },
    {
      "entropy": 0.4970203940756619,
      "epoch": 2.0974347473562545,
      "grad_norm": 0.15856671333312988,
      "learning_rate": 3.0088631275492997e-05,
      "loss": 0.4824,
      "mean_token_accuracy": 0.847106758505106,
      "num_tokens": 727545717.0,
      "step": 22797
    },
    {
      "entropy": 0.5318001331761479,
      "epoch": 2.0975267537276956,
      "grad_norm": 0.1596866250038147,
      "learning_rate": 3.0085564449351365e-05,
      "loss": 0.5099,
      "mean_token_accuracy": 0.8393194302916527,
      "num_tokens": 727577509.0,
      "step": 22798
    },
    {
      "entropy": 0.4618429634720087,
      "epoch": 2.0976187600991367,
      "grad_norm": 0.15561291575431824,
      "learning_rate": 3.0082497623209743e-05,
      "loss": 0.4556,
      "mean_token_accuracy": 0.8560185171663761,
      "num_tokens": 727609070.0,
      "step": 22799
    },
    {
      "entropy": 0.5727133527398109,
      "epoch": 2.097710766470578,
      "grad_norm": 0.1602167934179306,
      "learning_rate": 3.0079430797068114e-05,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8245518878102303,
      "num_tokens": 727641712.0,
      "step": 22800
    },
    {
      "entropy": 0.4449690734036267,
      "epoch": 2.0978027728420194,
      "grad_norm": 0.15932397544384003,
      "learning_rate": 3.0076363970926492e-05,
      "loss": 0.4245,
      "mean_token_accuracy": 0.8633051104843616,
      "num_tokens": 727673715.0,
      "step": 22801
    },
    {
      "entropy": 0.45571704767644405,
      "epoch": 2.0978947792134606,
      "grad_norm": 0.14886850118637085,
      "learning_rate": 3.0073297144784864e-05,
      "loss": 0.4287,
      "mean_token_accuracy": 0.8634243905544281,
      "num_tokens": 727706060.0,
      "step": 22802
    },
    {
      "entropy": 0.5007363455370069,
      "epoch": 2.0979867855849017,
      "grad_norm": 0.15520542860031128,
      "learning_rate": 3.0070230318643238e-05,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8435437828302383,
      "num_tokens": 727738175.0,
      "step": 22803
    },
    {
      "entropy": 0.4742095749825239,
      "epoch": 2.098078791956343,
      "grad_norm": 0.1555846631526947,
      "learning_rate": 3.006716349250161e-05,
      "loss": 0.4559,
      "mean_token_accuracy": 0.8575369194149971,
      "num_tokens": 727770365.0,
      "step": 22804
    },
    {
      "entropy": 0.44678105087950826,
      "epoch": 2.098170798327784,
      "grad_norm": 0.15242819488048553,
      "learning_rate": 3.0064096666359988e-05,
      "loss": 0.429,
      "mean_token_accuracy": 0.8632726483047009,
      "num_tokens": 727802870.0,
      "step": 22805
    },
    {
      "entropy": 0.4327684957534075,
      "epoch": 2.0982628046992255,
      "grad_norm": 0.14625921845436096,
      "learning_rate": 3.006102984021836e-05,
      "loss": 0.4055,
      "mean_token_accuracy": 0.8698519058525562,
      "num_tokens": 727834879.0,
      "step": 22806
    },
    {
      "entropy": 0.46984450588934124,
      "epoch": 2.0983548110706667,
      "grad_norm": 0.1549854427576065,
      "learning_rate": 3.0057963014076733e-05,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8562501780688763,
      "num_tokens": 727867563.0,
      "step": 22807
    },
    {
      "entropy": 0.49481016118079424,
      "epoch": 2.098446817442108,
      "grad_norm": 0.16619615256786346,
      "learning_rate": 3.0054896187935105e-05,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8417747989296913,
      "num_tokens": 727898478.0,
      "step": 22808
    },
    {
      "entropy": 0.39294508192688227,
      "epoch": 2.098538823813549,
      "grad_norm": 0.1414063721895218,
      "learning_rate": 3.0051829361793483e-05,
      "loss": 0.3777,
      "mean_token_accuracy": 0.8813646398484707,
      "num_tokens": 727931157.0,
      "step": 22809
    },
    {
      "entropy": 0.6158612677827477,
      "epoch": 2.0986308301849905,
      "grad_norm": 0.18231932818889618,
      "learning_rate": 3.0048762535651854e-05,
      "loss": 0.5864,
      "mean_token_accuracy": 0.8135704882442951,
      "num_tokens": 727962135.0,
      "step": 22810
    },
    {
      "entropy": 0.5317228604108095,
      "epoch": 2.0987228365564317,
      "grad_norm": 0.16558150947093964,
      "learning_rate": 3.004569570951023e-05,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8388153053820133,
      "num_tokens": 727992537.0,
      "step": 22811
    },
    {
      "entropy": 0.5488553047180176,
      "epoch": 2.098814842927873,
      "grad_norm": 0.16718445718288422,
      "learning_rate": 3.00426288833686e-05,
      "loss": 0.5232,
      "mean_token_accuracy": 0.833520133048296,
      "num_tokens": 728024084.0,
      "step": 22812
    },
    {
      "entropy": 0.6145112374797463,
      "epoch": 2.098906849299314,
      "grad_norm": 0.17228540778160095,
      "learning_rate": 3.0039562057226978e-05,
      "loss": 0.6052,
      "mean_token_accuracy": 0.8125412352383137,
      "num_tokens": 728056491.0,
      "step": 22813
    },
    {
      "entropy": 0.5063165100291371,
      "epoch": 2.098998855670755,
      "grad_norm": 0.16039539873600006,
      "learning_rate": 3.003649523108535e-05,
      "loss": 0.4961,
      "mean_token_accuracy": 0.8436312302947044,
      "num_tokens": 728088185.0,
      "step": 22814
    },
    {
      "entropy": 0.5500717260874808,
      "epoch": 2.0990908620421966,
      "grad_norm": 0.16108383238315582,
      "learning_rate": 3.0033428404943724e-05,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8333496265113354,
      "num_tokens": 728120342.0,
      "step": 22815
    },
    {
      "entropy": 0.5204677525907755,
      "epoch": 2.0991828684136378,
      "grad_norm": 0.16094450652599335,
      "learning_rate": 3.0030361578802095e-05,
      "loss": 0.4935,
      "mean_token_accuracy": 0.8427911512553692,
      "num_tokens": 728152533.0,
      "step": 22816
    },
    {
      "entropy": 0.4650015039369464,
      "epoch": 2.099274874785079,
      "grad_norm": 0.1546032726764679,
      "learning_rate": 3.0027294752660473e-05,
      "loss": 0.438,
      "mean_token_accuracy": 0.8596246726810932,
      "num_tokens": 728184503.0,
      "step": 22817
    },
    {
      "entropy": 0.49511182121932507,
      "epoch": 2.09936688115652,
      "grad_norm": 0.1584109365940094,
      "learning_rate": 3.002422792651885e-05,
      "loss": 0.479,
      "mean_token_accuracy": 0.8483413681387901,
      "num_tokens": 728216723.0,
      "step": 22818
    },
    {
      "entropy": 0.6279781409539282,
      "epoch": 2.099458887527961,
      "grad_norm": 0.1706785261631012,
      "learning_rate": 3.0021161100377222e-05,
      "loss": 0.6109,
      "mean_token_accuracy": 0.807793878018856,
      "num_tokens": 728248765.0,
      "step": 22819
    },
    {
      "entropy": 0.4804857103154063,
      "epoch": 2.0995508938994027,
      "grad_norm": 0.148964062333107,
      "learning_rate": 3.0018094274235597e-05,
      "loss": 0.4681,
      "mean_token_accuracy": 0.8488687202334404,
      "num_tokens": 728281352.0,
      "step": 22820
    },
    {
      "entropy": 0.5694323647767305,
      "epoch": 2.099642900270844,
      "grad_norm": 0.1686541587114334,
      "learning_rate": 3.001502744809397e-05,
      "loss": 0.5592,
      "mean_token_accuracy": 0.8274555690586567,
      "num_tokens": 728313455.0,
      "step": 22821
    },
    {
      "entropy": 0.6334991417825222,
      "epoch": 2.099734906642285,
      "grad_norm": 0.17108802497386932,
      "learning_rate": 3.0011960621952346e-05,
      "loss": 0.6092,
      "mean_token_accuracy": 0.80689911916852,
      "num_tokens": 728345371.0,
      "step": 22822
    },
    {
      "entropy": 0.5221232855692506,
      "epoch": 2.099826913013726,
      "grad_norm": 0.16138915717601776,
      "learning_rate": 3.0008893795810718e-05,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8376987427473068,
      "num_tokens": 728377001.0,
      "step": 22823
    },
    {
      "entropy": 0.5222785137593746,
      "epoch": 2.0999189193851673,
      "grad_norm": 0.1662609875202179,
      "learning_rate": 3.0005826969669092e-05,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8320596851408482,
      "num_tokens": 728408350.0,
      "step": 22824
    },
    {
      "entropy": 0.5065051377750933,
      "epoch": 2.100010925756609,
      "grad_norm": 0.16608069837093353,
      "learning_rate": 3.0002760143527464e-05,
      "loss": 0.4974,
      "mean_token_accuracy": 0.8427172377705574,
      "num_tokens": 728439557.0,
      "step": 22825
    },
    {
      "entropy": 0.4974612542428076,
      "epoch": 2.10010293212805,
      "grad_norm": 0.15730850398540497,
      "learning_rate": 2.999969331738584e-05,
      "loss": 0.4752,
      "mean_token_accuracy": 0.8472037464380264,
      "num_tokens": 728471256.0,
      "step": 22826
    },
    {
      "entropy": 0.5751926386728883,
      "epoch": 2.100194938499491,
      "grad_norm": 0.16830860078334808,
      "learning_rate": 2.9996626491244213e-05,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8214105926454067,
      "num_tokens": 728502211.0,
      "step": 22827
    },
    {
      "entropy": 0.4838026473298669,
      "epoch": 2.1002869448709323,
      "grad_norm": 0.15604358911514282,
      "learning_rate": 2.9993559665102588e-05,
      "loss": 0.4732,
      "mean_token_accuracy": 0.849450409412384,
      "num_tokens": 728534790.0,
      "step": 22828
    },
    {
      "entropy": 0.48582740873098373,
      "epoch": 2.1003789512423734,
      "grad_norm": 0.16317136585712433,
      "learning_rate": 2.999049283896096e-05,
      "loss": 0.4822,
      "mean_token_accuracy": 0.8443153090775013,
      "num_tokens": 728566754.0,
      "step": 22829
    },
    {
      "entropy": 0.5974267534911633,
      "epoch": 2.100470957613815,
      "grad_norm": 0.17140896618366241,
      "learning_rate": 2.9987426012819337e-05,
      "loss": 0.5918,
      "mean_token_accuracy": 0.8154514133930206,
      "num_tokens": 728599499.0,
      "step": 22830
    },
    {
      "entropy": 0.5533296698704362,
      "epoch": 2.100562963985256,
      "grad_norm": 0.16653503477573395,
      "learning_rate": 2.9984359186677708e-05,
      "loss": 0.5404,
      "mean_token_accuracy": 0.8274101875722408,
      "num_tokens": 728631179.0,
      "step": 22831
    },
    {
      "entropy": 0.4638023469597101,
      "epoch": 2.1006549703566972,
      "grad_norm": 0.15607069432735443,
      "learning_rate": 2.9981292360536083e-05,
      "loss": 0.454,
      "mean_token_accuracy": 0.8551149964332581,
      "num_tokens": 728663141.0,
      "step": 22832
    },
    {
      "entropy": 0.4655702302698046,
      "epoch": 2.1007469767281384,
      "grad_norm": 0.15590143203735352,
      "learning_rate": 2.9978225534394454e-05,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8521529883146286,
      "num_tokens": 728695227.0,
      "step": 22833
    },
    {
      "entropy": 0.5298623144626617,
      "epoch": 2.1008389830995795,
      "grad_norm": 0.16601575911045074,
      "learning_rate": 2.9975158708252832e-05,
      "loss": 0.5156,
      "mean_token_accuracy": 0.8350227847695351,
      "num_tokens": 728727261.0,
      "step": 22834
    },
    {
      "entropy": 0.4704971518367529,
      "epoch": 2.100930989471021,
      "grad_norm": 0.15946736931800842,
      "learning_rate": 2.9972091882111203e-05,
      "loss": 0.4625,
      "mean_token_accuracy": 0.8512221351265907,
      "num_tokens": 728759592.0,
      "step": 22835
    },
    {
      "entropy": 0.5346943121403456,
      "epoch": 2.101022995842462,
      "grad_norm": 0.15744750201702118,
      "learning_rate": 2.9969025055969578e-05,
      "loss": 0.523,
      "mean_token_accuracy": 0.8342969082295895,
      "num_tokens": 728792188.0,
      "step": 22836
    },
    {
      "entropy": 0.5896954778581858,
      "epoch": 2.1011150022139033,
      "grad_norm": 0.17210949957370758,
      "learning_rate": 2.996595822982795e-05,
      "loss": 0.5626,
      "mean_token_accuracy": 0.8224927671253681,
      "num_tokens": 728823358.0,
      "step": 22837
    },
    {
      "entropy": 0.5397261781617999,
      "epoch": 2.1012070085853445,
      "grad_norm": 0.16620057821273804,
      "learning_rate": 2.9962891403686327e-05,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8363254815340042,
      "num_tokens": 728855035.0,
      "step": 22838
    },
    {
      "entropy": 0.5936598759144545,
      "epoch": 2.1012990149567856,
      "grad_norm": 0.17262640595436096,
      "learning_rate": 2.99598245775447e-05,
      "loss": 0.573,
      "mean_token_accuracy": 0.8180497363209724,
      "num_tokens": 728886991.0,
      "step": 22839
    },
    {
      "entropy": 0.49348919186741114,
      "epoch": 2.101391021328227,
      "grad_norm": 0.15869322419166565,
      "learning_rate": 2.9956757751403077e-05,
      "loss": 0.4719,
      "mean_token_accuracy": 0.8485983461141586,
      "num_tokens": 728918907.0,
      "step": 22840
    },
    {
      "entropy": 0.5100497901439667,
      "epoch": 2.1014830276996683,
      "grad_norm": 0.16399694979190826,
      "learning_rate": 2.9953690925261448e-05,
      "loss": 0.497,
      "mean_token_accuracy": 0.8389025777578354,
      "num_tokens": 728951023.0,
      "step": 22841
    },
    {
      "entropy": 0.45325865503400564,
      "epoch": 2.1015750340711095,
      "grad_norm": 0.1569947451353073,
      "learning_rate": 2.9950624099119822e-05,
      "loss": 0.444,
      "mean_token_accuracy": 0.8587599620223045,
      "num_tokens": 728982152.0,
      "step": 22842
    },
    {
      "entropy": 0.5412753624841571,
      "epoch": 2.1016670404425506,
      "grad_norm": 0.16337072849273682,
      "learning_rate": 2.9947557272978194e-05,
      "loss": 0.5284,
      "mean_token_accuracy": 0.8317077048122883,
      "num_tokens": 729014311.0,
      "step": 22843
    },
    {
      "entropy": 0.4316125132609159,
      "epoch": 2.1017590468139917,
      "grad_norm": 0.15597988665103912,
      "learning_rate": 2.9944490446836572e-05,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8672707453370094,
      "num_tokens": 729045938.0,
      "step": 22844
    },
    {
      "entropy": 0.5372802317142487,
      "epoch": 2.1018510531854333,
      "grad_norm": 0.1635502576828003,
      "learning_rate": 2.9941423620694943e-05,
      "loss": 0.5316,
      "mean_token_accuracy": 0.8314673081040382,
      "num_tokens": 729077827.0,
      "step": 22845
    },
    {
      "entropy": 0.5122205326333642,
      "epoch": 2.1019430595568744,
      "grad_norm": 0.16167622804641724,
      "learning_rate": 2.9938356794553318e-05,
      "loss": 0.505,
      "mean_token_accuracy": 0.8362935483455658,
      "num_tokens": 729110305.0,
      "step": 22846
    },
    {
      "entropy": 0.5353304017335176,
      "epoch": 2.1020350659283156,
      "grad_norm": 0.16455799341201782,
      "learning_rate": 2.993528996841169e-05,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8372282832860947,
      "num_tokens": 729141678.0,
      "step": 22847
    },
    {
      "entropy": 0.45566285215318203,
      "epoch": 2.1021270722997567,
      "grad_norm": 0.1585477590560913,
      "learning_rate": 2.9932223142270067e-05,
      "loss": 0.4554,
      "mean_token_accuracy": 0.8568815216422081,
      "num_tokens": 729173791.0,
      "step": 22848
    },
    {
      "entropy": 0.49195193126797676,
      "epoch": 2.102219078671198,
      "grad_norm": 0.1633092164993286,
      "learning_rate": 2.992915631612844e-05,
      "loss": 0.4726,
      "mean_token_accuracy": 0.848917756229639,
      "num_tokens": 729205430.0,
      "step": 22849
    },
    {
      "entropy": 0.47965950751677155,
      "epoch": 2.1023110850426394,
      "grad_norm": 0.15432915091514587,
      "learning_rate": 2.9926089489986813e-05,
      "loss": 0.4631,
      "mean_token_accuracy": 0.851491667330265,
      "num_tokens": 729237509.0,
      "step": 22850
    },
    {
      "entropy": 0.49265166698023677,
      "epoch": 2.1024030914140805,
      "grad_norm": 0.1565762162208557,
      "learning_rate": 2.992302266384519e-05,
      "loss": 0.4802,
      "mean_token_accuracy": 0.8457052409648895,
      "num_tokens": 729269855.0,
      "step": 22851
    },
    {
      "entropy": 0.48917699418962,
      "epoch": 2.1024950977855217,
      "grad_norm": 0.16246937215328217,
      "learning_rate": 2.9919955837703562e-05,
      "loss": 0.4664,
      "mean_token_accuracy": 0.8488801680505276,
      "num_tokens": 729301076.0,
      "step": 22852
    },
    {
      "entropy": 0.527216924354434,
      "epoch": 2.102587104156963,
      "grad_norm": 0.15922175347805023,
      "learning_rate": 2.9916889011561937e-05,
      "loss": 0.5083,
      "mean_token_accuracy": 0.8391950912773609,
      "num_tokens": 729333056.0,
      "step": 22853
    },
    {
      "entropy": 0.5669365245848894,
      "epoch": 2.102679110528404,
      "grad_norm": 0.17427030205726624,
      "learning_rate": 2.9913822185420308e-05,
      "loss": 0.5584,
      "mean_token_accuracy": 0.82309764996171,
      "num_tokens": 729363362.0,
      "step": 22854
    },
    {
      "entropy": 0.5298520047217607,
      "epoch": 2.1027711168998455,
      "grad_norm": 0.15626566112041473,
      "learning_rate": 2.9910755359278686e-05,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8362511359155178,
      "num_tokens": 729395541.0,
      "step": 22855
    },
    {
      "entropy": 0.5752904070541263,
      "epoch": 2.1028631232712867,
      "grad_norm": 0.16800116002559662,
      "learning_rate": 2.9907688533137057e-05,
      "loss": 0.5465,
      "mean_token_accuracy": 0.824533648788929,
      "num_tokens": 729426754.0,
      "step": 22856
    },
    {
      "entropy": 0.5977001572027802,
      "epoch": 2.102955129642728,
      "grad_norm": 0.16965532302856445,
      "learning_rate": 2.9904621706995435e-05,
      "loss": 0.5828,
      "mean_token_accuracy": 0.8162357434630394,
      "num_tokens": 729458489.0,
      "step": 22857
    },
    {
      "entropy": 0.4936964015942067,
      "epoch": 2.103047136014169,
      "grad_norm": 0.1542668342590332,
      "learning_rate": 2.9901554880853803e-05,
      "loss": 0.4822,
      "mean_token_accuracy": 0.8447039946913719,
      "num_tokens": 729490035.0,
      "step": 22858
    },
    {
      "entropy": 0.4712307648733258,
      "epoch": 2.10313914238561,
      "grad_norm": 0.1536760777235031,
      "learning_rate": 2.989848805471218e-05,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8548024967312813,
      "num_tokens": 729522089.0,
      "step": 22859
    },
    {
      "entropy": 0.6110823089256883,
      "epoch": 2.1032311487570516,
      "grad_norm": 0.16808770596981049,
      "learning_rate": 2.9895421228570553e-05,
      "loss": 0.5966,
      "mean_token_accuracy": 0.8133853822946548,
      "num_tokens": 729554332.0,
      "step": 22860
    },
    {
      "entropy": 0.4866133201867342,
      "epoch": 2.1033231551284928,
      "grad_norm": 0.15304158627986908,
      "learning_rate": 2.989235440242893e-05,
      "loss": 0.473,
      "mean_token_accuracy": 0.8485248573124409,
      "num_tokens": 729586308.0,
      "step": 22861
    },
    {
      "entropy": 0.5418924633413553,
      "epoch": 2.103415161499934,
      "grad_norm": 0.16754111647605896,
      "learning_rate": 2.9889287576287302e-05,
      "loss": 0.5349,
      "mean_token_accuracy": 0.8297210298478603,
      "num_tokens": 729618349.0,
      "step": 22862
    },
    {
      "entropy": 0.5736800734885037,
      "epoch": 2.103507167871375,
      "grad_norm": 0.16623151302337646,
      "learning_rate": 2.9886220750145676e-05,
      "loss": 0.556,
      "mean_token_accuracy": 0.8218614719808102,
      "num_tokens": 729650746.0,
      "step": 22863
    },
    {
      "entropy": 0.45393372513353825,
      "epoch": 2.103599174242816,
      "grad_norm": 0.15761950612068176,
      "learning_rate": 2.9883153924004048e-05,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8557172566652298,
      "num_tokens": 729682758.0,
      "step": 22864
    },
    {
      "entropy": 0.549815071746707,
      "epoch": 2.1036911806142577,
      "grad_norm": 0.16562722623348236,
      "learning_rate": 2.9880087097862426e-05,
      "loss": 0.5301,
      "mean_token_accuracy": 0.8315704725682735,
      "num_tokens": 729714617.0,
      "step": 22865
    },
    {
      "entropy": 0.5394445117563009,
      "epoch": 2.103783186985699,
      "grad_norm": 0.16210080683231354,
      "learning_rate": 2.9877020271720797e-05,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8313228189945221,
      "num_tokens": 729747210.0,
      "step": 22866
    },
    {
      "entropy": 0.5157202437985688,
      "epoch": 2.10387519335714,
      "grad_norm": 0.15865521132946014,
      "learning_rate": 2.987395344557917e-05,
      "loss": 0.4876,
      "mean_token_accuracy": 0.8440041244029999,
      "num_tokens": 729778974.0,
      "step": 22867
    },
    {
      "entropy": 0.4579641930758953,
      "epoch": 2.103967199728581,
      "grad_norm": 0.15040484070777893,
      "learning_rate": 2.9870886619437543e-05,
      "loss": 0.4362,
      "mean_token_accuracy": 0.860031746327877,
      "num_tokens": 729811099.0,
      "step": 22868
    },
    {
      "entropy": 0.5789814125746489,
      "epoch": 2.1040592061000223,
      "grad_norm": 0.16972683370113373,
      "learning_rate": 2.986781979329592e-05,
      "loss": 0.5667,
      "mean_token_accuracy": 0.8232209533452988,
      "num_tokens": 729842730.0,
      "step": 22869
    },
    {
      "entropy": 0.5889539821073413,
      "epoch": 2.104151212471464,
      "grad_norm": 0.16386981308460236,
      "learning_rate": 2.9864752967154292e-05,
      "loss": 0.5753,
      "mean_token_accuracy": 0.8216349668800831,
      "num_tokens": 729874626.0,
      "step": 22870
    },
    {
      "entropy": 0.5510677620768547,
      "epoch": 2.104243218842905,
      "grad_norm": 0.1643325239419937,
      "learning_rate": 2.9861686141012667e-05,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8330447860062122,
      "num_tokens": 729906966.0,
      "step": 22871
    },
    {
      "entropy": 0.5842300616204739,
      "epoch": 2.104335225214346,
      "grad_norm": 0.1647004932165146,
      "learning_rate": 2.9858619314871038e-05,
      "loss": 0.5585,
      "mean_token_accuracy": 0.8239427842199802,
      "num_tokens": 729938943.0,
      "step": 22872
    },
    {
      "entropy": 0.5612283004447818,
      "epoch": 2.1044272315857873,
      "grad_norm": 0.16832219064235687,
      "learning_rate": 2.9855552488729416e-05,
      "loss": 0.5453,
      "mean_token_accuracy": 0.8263469897210598,
      "num_tokens": 729971481.0,
      "step": 22873
    },
    {
      "entropy": 0.5423293048515916,
      "epoch": 2.1045192379572284,
      "grad_norm": 0.16181659698486328,
      "learning_rate": 2.9852485662587787e-05,
      "loss": 0.5055,
      "mean_token_accuracy": 0.8403125815093517,
      "num_tokens": 730002860.0,
      "step": 22874
    },
    {
      "entropy": 0.5188505714759231,
      "epoch": 2.10461124432867,
      "grad_norm": 0.16159506142139435,
      "learning_rate": 2.9849418836446162e-05,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8377290964126587,
      "num_tokens": 730034465.0,
      "step": 22875
    },
    {
      "entropy": 0.5793077982962132,
      "epoch": 2.104703250700111,
      "grad_norm": 0.16542725265026093,
      "learning_rate": 2.9846352010304533e-05,
      "loss": 0.5659,
      "mean_token_accuracy": 0.8198866061866283,
      "num_tokens": 730066113.0,
      "step": 22876
    },
    {
      "entropy": 0.5694359689950943,
      "epoch": 2.1047952570715522,
      "grad_norm": 0.16624785959720612,
      "learning_rate": 2.984328518416291e-05,
      "loss": 0.5582,
      "mean_token_accuracy": 0.8231168501079082,
      "num_tokens": 730097313.0,
      "step": 22877
    },
    {
      "entropy": 0.44566704519093037,
      "epoch": 2.1048872634429934,
      "grad_norm": 0.15504074096679688,
      "learning_rate": 2.984021835802129e-05,
      "loss": 0.4285,
      "mean_token_accuracy": 0.8638571910560131,
      "num_tokens": 730129288.0,
      "step": 22878
    },
    {
      "entropy": 0.5485628694295883,
      "epoch": 2.1049792698144345,
      "grad_norm": 0.16160695254802704,
      "learning_rate": 2.983715153187966e-05,
      "loss": 0.5414,
      "mean_token_accuracy": 0.8293294161558151,
      "num_tokens": 730161808.0,
      "step": 22879
    },
    {
      "entropy": 0.49130330327898264,
      "epoch": 2.105071276185876,
      "grad_norm": 0.15856575965881348,
      "learning_rate": 2.9834084705738035e-05,
      "loss": 0.4857,
      "mean_token_accuracy": 0.8507229276001453,
      "num_tokens": 730194330.0,
      "step": 22880
    },
    {
      "entropy": 0.47883031796664,
      "epoch": 2.105163282557317,
      "grad_norm": 0.15816588699817657,
      "learning_rate": 2.9831017879596407e-05,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8525837659835815,
      "num_tokens": 730225912.0,
      "step": 22881
    },
    {
      "entropy": 0.4281740728765726,
      "epoch": 2.1052552889287584,
      "grad_norm": 0.15217314660549164,
      "learning_rate": 2.9827951053454785e-05,
      "loss": 0.4205,
      "mean_token_accuracy": 0.8666140586137772,
      "num_tokens": 730257726.0,
      "step": 22882
    },
    {
      "entropy": 0.46893076598644257,
      "epoch": 2.1053472953001995,
      "grad_norm": 0.149758979678154,
      "learning_rate": 2.9824884227313156e-05,
      "loss": 0.4475,
      "mean_token_accuracy": 0.8581460490822792,
      "num_tokens": 730289822.0,
      "step": 22883
    },
    {
      "entropy": 0.5008244011551142,
      "epoch": 2.1054393016716406,
      "grad_norm": 0.14784234762191772,
      "learning_rate": 2.982181740117153e-05,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8479166626930237,
      "num_tokens": 730322160.0,
      "step": 22884
    },
    {
      "entropy": 0.6074592294171453,
      "epoch": 2.105531308043082,
      "grad_norm": 0.17180633544921875,
      "learning_rate": 2.9818750575029902e-05,
      "loss": 0.5964,
      "mean_token_accuracy": 0.8146873861551285,
      "num_tokens": 730354137.0,
      "step": 22885
    },
    {
      "entropy": 0.5678798891603947,
      "epoch": 2.1056233144145233,
      "grad_norm": 0.16660411655902863,
      "learning_rate": 2.981568374888828e-05,
      "loss": 0.5586,
      "mean_token_accuracy": 0.8244721405208111,
      "num_tokens": 730385570.0,
      "step": 22886
    },
    {
      "entropy": 0.49751992896199226,
      "epoch": 2.1057153207859645,
      "grad_norm": 0.1526215821504593,
      "learning_rate": 2.981261692274665e-05,
      "loss": 0.473,
      "mean_token_accuracy": 0.8476595655083656,
      "num_tokens": 730417508.0,
      "step": 22887
    },
    {
      "entropy": 0.5604720022529364,
      "epoch": 2.1058073271574056,
      "grad_norm": 0.16528770327568054,
      "learning_rate": 2.9809550096605026e-05,
      "loss": 0.5434,
      "mean_token_accuracy": 0.828963078558445,
      "num_tokens": 730449505.0,
      "step": 22888
    },
    {
      "entropy": 0.4957893551327288,
      "epoch": 2.1058993335288467,
      "grad_norm": 0.16318276524543762,
      "learning_rate": 2.9806483270463397e-05,
      "loss": 0.4992,
      "mean_token_accuracy": 0.8435115218162537,
      "num_tokens": 730481482.0,
      "step": 22889
    },
    {
      "entropy": 0.49000347778201103,
      "epoch": 2.1059913399002883,
      "grad_norm": 0.1558503955602646,
      "learning_rate": 2.9803416444321775e-05,
      "loss": 0.4618,
      "mean_token_accuracy": 0.8492548801004887,
      "num_tokens": 730512934.0,
      "step": 22890
    },
    {
      "entropy": 0.5482386611402035,
      "epoch": 2.1060833462717294,
      "grad_norm": 0.16304998099803925,
      "learning_rate": 2.9800349618180146e-05,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8306249901652336,
      "num_tokens": 730544740.0,
      "step": 22891
    },
    {
      "entropy": 0.5030636070296168,
      "epoch": 2.1061753526431706,
      "grad_norm": 0.16405686736106873,
      "learning_rate": 2.979728279203852e-05,
      "loss": 0.4914,
      "mean_token_accuracy": 0.8458584249019623,
      "num_tokens": 730576960.0,
      "step": 22892
    },
    {
      "entropy": 0.6541821081191301,
      "epoch": 2.1062673590146117,
      "grad_norm": 0.17696484923362732,
      "learning_rate": 2.9794215965896892e-05,
      "loss": 0.6297,
      "mean_token_accuracy": 0.8024156205356121,
      "num_tokens": 730608557.0,
      "step": 22893
    },
    {
      "entropy": 0.5632916344329715,
      "epoch": 2.106359365386053,
      "grad_norm": 0.16898813843727112,
      "learning_rate": 2.979114913975527e-05,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8263812437653542,
      "num_tokens": 730640521.0,
      "step": 22894
    },
    {
      "entropy": 0.48309554532170296,
      "epoch": 2.1064513717574944,
      "grad_norm": 0.1601453274488449,
      "learning_rate": 2.978808231361364e-05,
      "loss": 0.4694,
      "mean_token_accuracy": 0.8473669588565826,
      "num_tokens": 730672481.0,
      "step": 22895
    },
    {
      "entropy": 0.4408524986356497,
      "epoch": 2.1065433781289356,
      "grad_norm": 0.1548348218202591,
      "learning_rate": 2.9785015487472016e-05,
      "loss": 0.4275,
      "mean_token_accuracy": 0.867288526147604,
      "num_tokens": 730704305.0,
      "step": 22896
    },
    {
      "entropy": 0.5166933527216315,
      "epoch": 2.1066353845003767,
      "grad_norm": 0.15699289739131927,
      "learning_rate": 2.9781948661330387e-05,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8428776487708092,
      "num_tokens": 730736426.0,
      "step": 22897
    },
    {
      "entropy": 0.5544506469741464,
      "epoch": 2.106727390871818,
      "grad_norm": 0.1655285805463791,
      "learning_rate": 2.9778881835188765e-05,
      "loss": 0.5397,
      "mean_token_accuracy": 0.828845776617527,
      "num_tokens": 730768205.0,
      "step": 22898
    },
    {
      "entropy": 0.5833972729742527,
      "epoch": 2.106819397243259,
      "grad_norm": 0.17202189564704895,
      "learning_rate": 2.9775815009047137e-05,
      "loss": 0.5616,
      "mean_token_accuracy": 0.8203870058059692,
      "num_tokens": 730799132.0,
      "step": 22899
    },
    {
      "entropy": 0.5040646390989423,
      "epoch": 2.1069114036147005,
      "grad_norm": 0.15365025401115417,
      "learning_rate": 2.9772748182905515e-05,
      "loss": 0.4845,
      "mean_token_accuracy": 0.8477162234485149,
      "num_tokens": 730831834.0,
      "step": 22900
    },
    {
      "entropy": 0.5259168716147542,
      "epoch": 2.1070034099861417,
      "grad_norm": 0.1695622205734253,
      "learning_rate": 2.9769681356763886e-05,
      "loss": 0.5162,
      "mean_token_accuracy": 0.837074764072895,
      "num_tokens": 730864183.0,
      "step": 22901
    },
    {
      "entropy": 0.4893719684332609,
      "epoch": 2.107095416357583,
      "grad_norm": 0.16108983755111694,
      "learning_rate": 2.976661453062226e-05,
      "loss": 0.4665,
      "mean_token_accuracy": 0.8507754355669022,
      "num_tokens": 730895385.0,
      "step": 22902
    },
    {
      "entropy": 0.4435705281794071,
      "epoch": 2.107187422729024,
      "grad_norm": 0.14767614006996155,
      "learning_rate": 2.9763547704480632e-05,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8596671223640442,
      "num_tokens": 730927064.0,
      "step": 22903
    },
    {
      "entropy": 0.48702552961185575,
      "epoch": 2.107279429100465,
      "grad_norm": 0.15799614787101746,
      "learning_rate": 2.976048087833901e-05,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8489444181323051,
      "num_tokens": 730959233.0,
      "step": 22904
    },
    {
      "entropy": 0.476301426300779,
      "epoch": 2.1073714354719066,
      "grad_norm": 0.15075257420539856,
      "learning_rate": 2.975741405219738e-05,
      "loss": 0.4634,
      "mean_token_accuracy": 0.8520895056426525,
      "num_tokens": 730990902.0,
      "step": 22905
    },
    {
      "entropy": 0.6479711718857288,
      "epoch": 2.1074634418433478,
      "grad_norm": 0.17930708825588226,
      "learning_rate": 2.9754347226055756e-05,
      "loss": 0.6248,
      "mean_token_accuracy": 0.8045649155974388,
      "num_tokens": 731021962.0,
      "step": 22906
    },
    {
      "entropy": 0.5331138037145138,
      "epoch": 2.107555448214789,
      "grad_norm": 0.16488751769065857,
      "learning_rate": 2.9751280399914127e-05,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8338759616017342,
      "num_tokens": 731054503.0,
      "step": 22907
    },
    {
      "entropy": 0.5236131011042744,
      "epoch": 2.10764745458623,
      "grad_norm": 0.15885446965694427,
      "learning_rate": 2.9748213573772505e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8353474549949169,
      "num_tokens": 731086831.0,
      "step": 22908
    },
    {
      "entropy": 0.5332627440802753,
      "epoch": 2.107739460957671,
      "grad_norm": 0.16411341726779938,
      "learning_rate": 2.974514674763088e-05,
      "loss": 0.516,
      "mean_token_accuracy": 0.8367175161838531,
      "num_tokens": 731118611.0,
      "step": 22909
    },
    {
      "entropy": 0.4725149106234312,
      "epoch": 2.1078314673291128,
      "grad_norm": 0.1582396924495697,
      "learning_rate": 2.974207992148925e-05,
      "loss": 0.4686,
      "mean_token_accuracy": 0.8521245568990707,
      "num_tokens": 731151125.0,
      "step": 22910
    },
    {
      "entropy": 0.4722012598067522,
      "epoch": 2.107923473700554,
      "grad_norm": 0.15430587530136108,
      "learning_rate": 2.973901309534763e-05,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8537267670035362,
      "num_tokens": 731183828.0,
      "step": 22911
    },
    {
      "entropy": 0.5978923179209232,
      "epoch": 2.108015480071995,
      "grad_norm": 0.1722187101840973,
      "learning_rate": 2.9735946269206e-05,
      "loss": 0.5778,
      "mean_token_accuracy": 0.8173761516809464,
      "num_tokens": 731215545.0,
      "step": 22912
    },
    {
      "entropy": 0.5543837826699018,
      "epoch": 2.108107486443436,
      "grad_norm": 0.17256195843219757,
      "learning_rate": 2.9732879443064375e-05,
      "loss": 0.5494,
      "mean_token_accuracy": 0.8297157697379589,
      "num_tokens": 731246160.0,
      "step": 22913
    },
    {
      "entropy": 0.4331742827780545,
      "epoch": 2.1081994928148773,
      "grad_norm": 0.15077735483646393,
      "learning_rate": 2.9729812616922746e-05,
      "loss": 0.4225,
      "mean_token_accuracy": 0.8665119670331478,
      "num_tokens": 731278928.0,
      "step": 22914
    },
    {
      "entropy": 0.4621338974684477,
      "epoch": 2.108291499186319,
      "grad_norm": 0.15944808721542358,
      "learning_rate": 2.9726745790781124e-05,
      "loss": 0.4512,
      "mean_token_accuracy": 0.855555098503828,
      "num_tokens": 731310509.0,
      "step": 22915
    },
    {
      "entropy": 0.4833188112825155,
      "epoch": 2.10838350555776,
      "grad_norm": 0.15706463158130646,
      "learning_rate": 2.9723678964639496e-05,
      "loss": 0.4718,
      "mean_token_accuracy": 0.8504791595041752,
      "num_tokens": 731342341.0,
      "step": 22916
    },
    {
      "entropy": 0.5511418161913753,
      "epoch": 2.108475511929201,
      "grad_norm": 0.16523756086826324,
      "learning_rate": 2.9720612138497874e-05,
      "loss": 0.5319,
      "mean_token_accuracy": 0.8317261934280396,
      "num_tokens": 731374271.0,
      "step": 22917
    },
    {
      "entropy": 0.5405501844361424,
      "epoch": 2.1085675183006423,
      "grad_norm": 0.1574670672416687,
      "learning_rate": 2.971754531235624e-05,
      "loss": 0.5235,
      "mean_token_accuracy": 0.8302207328379154,
      "num_tokens": 731406753.0,
      "step": 22918
    },
    {
      "entropy": 0.6354760825634003,
      "epoch": 2.1086595246720834,
      "grad_norm": 0.1749548614025116,
      "learning_rate": 2.971447848621462e-05,
      "loss": 0.6138,
      "mean_token_accuracy": 0.8052751310169697,
      "num_tokens": 731438422.0,
      "step": 22919
    },
    {
      "entropy": 0.5334871453233063,
      "epoch": 2.108751531043525,
      "grad_norm": 0.1625843495130539,
      "learning_rate": 2.971141166007299e-05,
      "loss": 0.5184,
      "mean_token_accuracy": 0.8361893929541111,
      "num_tokens": 731470921.0,
      "step": 22920
    },
    {
      "entropy": 0.5262632090598345,
      "epoch": 2.108843537414966,
      "grad_norm": 0.1633625477552414,
      "learning_rate": 2.970834483393137e-05,
      "loss": 0.5263,
      "mean_token_accuracy": 0.8365722522139549,
      "num_tokens": 731503119.0,
      "step": 22921
    },
    {
      "entropy": 0.39742915728129447,
      "epoch": 2.1089355437864072,
      "grad_norm": 0.14609302580356598,
      "learning_rate": 2.970527800778974e-05,
      "loss": 0.387,
      "mean_token_accuracy": 0.8746142461895943,
      "num_tokens": 731534769.0,
      "step": 22922
    },
    {
      "entropy": 0.49015349498949945,
      "epoch": 2.1090275501578484,
      "grad_norm": 0.15681031346321106,
      "learning_rate": 2.9702211181648115e-05,
      "loss": 0.4719,
      "mean_token_accuracy": 0.848303098231554,
      "num_tokens": 731566325.0,
      "step": 22923
    },
    {
      "entropy": 0.5059064961969852,
      "epoch": 2.1091195565292895,
      "grad_norm": 0.1630149632692337,
      "learning_rate": 2.9699144355506486e-05,
      "loss": 0.4936,
      "mean_token_accuracy": 0.8438134901225567,
      "num_tokens": 731599015.0,
      "step": 22924
    },
    {
      "entropy": 0.547536700963974,
      "epoch": 2.109211562900731,
      "grad_norm": 0.1656537503004074,
      "learning_rate": 2.9696077529364864e-05,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8305396288633347,
      "num_tokens": 731630957.0,
      "step": 22925
    },
    {
      "entropy": 0.5365731520578265,
      "epoch": 2.109303569272172,
      "grad_norm": 0.16717694699764252,
      "learning_rate": 2.9693010703223235e-05,
      "loss": 0.5197,
      "mean_token_accuracy": 0.8357802219688892,
      "num_tokens": 731662049.0,
      "step": 22926
    },
    {
      "entropy": 0.6050206981599331,
      "epoch": 2.1093955756436134,
      "grad_norm": 0.17099842429161072,
      "learning_rate": 2.968994387708161e-05,
      "loss": 0.5915,
      "mean_token_accuracy": 0.8155133686959743,
      "num_tokens": 731694401.0,
      "step": 22927
    },
    {
      "entropy": 0.521555726416409,
      "epoch": 2.1094875820150545,
      "grad_norm": 0.16959869861602783,
      "learning_rate": 2.968687705093998e-05,
      "loss": 0.5144,
      "mean_token_accuracy": 0.8383474983274937,
      "num_tokens": 731726613.0,
      "step": 22928
    },
    {
      "entropy": 0.5484418720006943,
      "epoch": 2.1095795883864956,
      "grad_norm": 0.16369295120239258,
      "learning_rate": 2.968381022479836e-05,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8307933285832405,
      "num_tokens": 731758540.0,
      "step": 22929
    },
    {
      "entropy": 0.5907754525542259,
      "epoch": 2.109671594757937,
      "grad_norm": 0.16359303891658783,
      "learning_rate": 2.968074339865673e-05,
      "loss": 0.5775,
      "mean_token_accuracy": 0.8192752115428448,
      "num_tokens": 731791174.0,
      "step": 22930
    },
    {
      "entropy": 0.4288144023157656,
      "epoch": 2.1097636011293783,
      "grad_norm": 0.14849381148815155,
      "learning_rate": 2.9677676572515105e-05,
      "loss": 0.4124,
      "mean_token_accuracy": 0.8676838018000126,
      "num_tokens": 731823244.0,
      "step": 22931
    },
    {
      "entropy": 0.5415521850809455,
      "epoch": 2.1098556075008195,
      "grad_norm": 0.16122324764728546,
      "learning_rate": 2.9674609746373476e-05,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8371446207165718,
      "num_tokens": 731855828.0,
      "step": 22932
    },
    {
      "entropy": 0.47413517627865076,
      "epoch": 2.1099476138722606,
      "grad_norm": 0.15510398149490356,
      "learning_rate": 2.9671542920231854e-05,
      "loss": 0.461,
      "mean_token_accuracy": 0.8544934280216694,
      "num_tokens": 731888581.0,
      "step": 22933
    },
    {
      "entropy": 0.5937707331031561,
      "epoch": 2.1100396202437017,
      "grad_norm": 0.1693655550479889,
      "learning_rate": 2.9668476094090226e-05,
      "loss": 0.5947,
      "mean_token_accuracy": 0.8152953162789345,
      "num_tokens": 731921158.0,
      "step": 22934
    },
    {
      "entropy": 0.6005253428593278,
      "epoch": 2.1101316266151433,
      "grad_norm": 0.17862749099731445,
      "learning_rate": 2.96654092679486e-05,
      "loss": 0.5872,
      "mean_token_accuracy": 0.8133177570998669,
      "num_tokens": 731953024.0,
      "step": 22935
    },
    {
      "entropy": 0.5021905628964305,
      "epoch": 2.1102236329865844,
      "grad_norm": 0.15831521153450012,
      "learning_rate": 2.966234244180697e-05,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8471653684973717,
      "num_tokens": 731984412.0,
      "step": 22936
    },
    {
      "entropy": 0.48382079927250743,
      "epoch": 2.1103156393580256,
      "grad_norm": 0.15938976407051086,
      "learning_rate": 2.965927561566535e-05,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8473847508430481,
      "num_tokens": 732016656.0,
      "step": 22937
    },
    {
      "entropy": 0.5863032438792288,
      "epoch": 2.1104076457294667,
      "grad_norm": 0.16499435901641846,
      "learning_rate": 2.965620878952372e-05,
      "loss": 0.5767,
      "mean_token_accuracy": 0.8193227127194405,
      "num_tokens": 732049062.0,
      "step": 22938
    },
    {
      "entropy": 0.5061192028224468,
      "epoch": 2.110499652100908,
      "grad_norm": 0.16198214888572693,
      "learning_rate": 2.96531419633821e-05,
      "loss": 0.4828,
      "mean_token_accuracy": 0.8458501882851124,
      "num_tokens": 732080393.0,
      "step": 22939
    },
    {
      "entropy": 0.46058296551927924,
      "epoch": 2.1105916584723494,
      "grad_norm": 0.15302543342113495,
      "learning_rate": 2.9650075137240474e-05,
      "loss": 0.4478,
      "mean_token_accuracy": 0.8576982393860817,
      "num_tokens": 732112846.0,
      "step": 22940
    },
    {
      "entropy": 0.4652694296091795,
      "epoch": 2.1106836648437906,
      "grad_norm": 0.1543063074350357,
      "learning_rate": 2.9647008311098845e-05,
      "loss": 0.4506,
      "mean_token_accuracy": 0.8564631454646587,
      "num_tokens": 732144608.0,
      "step": 22941
    },
    {
      "entropy": 0.4170175679028034,
      "epoch": 2.1107756712152317,
      "grad_norm": 0.1509009748697281,
      "learning_rate": 2.9643941484957223e-05,
      "loss": 0.4087,
      "mean_token_accuracy": 0.8694568201899529,
      "num_tokens": 732176436.0,
      "step": 22942
    },
    {
      "entropy": 0.6431980319321156,
      "epoch": 2.110867677586673,
      "grad_norm": 0.17376184463500977,
      "learning_rate": 2.9640874658815594e-05,
      "loss": 0.6333,
      "mean_token_accuracy": 0.7992641180753708,
      "num_tokens": 732208713.0,
      "step": 22943
    },
    {
      "entropy": 0.5146316606551409,
      "epoch": 2.110959683958114,
      "grad_norm": 0.16562321782112122,
      "learning_rate": 2.963780783267397e-05,
      "loss": 0.5067,
      "mean_token_accuracy": 0.8395199552178383,
      "num_tokens": 732240929.0,
      "step": 22944
    },
    {
      "entropy": 0.48471576906740665,
      "epoch": 2.1110516903295555,
      "grad_norm": 0.15856876969337463,
      "learning_rate": 2.963474100653234e-05,
      "loss": 0.4835,
      "mean_token_accuracy": 0.8460978753864765,
      "num_tokens": 732273247.0,
      "step": 22945
    },
    {
      "entropy": 0.5404588686069474,
      "epoch": 2.1111436967009967,
      "grad_norm": 0.16007627546787262,
      "learning_rate": 2.9631674180390718e-05,
      "loss": 0.5215,
      "mean_token_accuracy": 0.8348355405032635,
      "num_tokens": 732304862.0,
      "step": 22946
    },
    {
      "entropy": 0.473876534961164,
      "epoch": 2.111235703072438,
      "grad_norm": 0.15395426750183105,
      "learning_rate": 2.962860735424909e-05,
      "loss": 0.4587,
      "mean_token_accuracy": 0.8532060012221336,
      "num_tokens": 732337430.0,
      "step": 22947
    },
    {
      "entropy": 0.5026230048388243,
      "epoch": 2.111327709443879,
      "grad_norm": 0.15888401865959167,
      "learning_rate": 2.9625540528107464e-05,
      "loss": 0.5006,
      "mean_token_accuracy": 0.8399346098303795,
      "num_tokens": 732369834.0,
      "step": 22948
    },
    {
      "entropy": 0.4898689640685916,
      "epoch": 2.11141971581532,
      "grad_norm": 0.15931253135204315,
      "learning_rate": 2.9622473701965835e-05,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8496352396905422,
      "num_tokens": 732401449.0,
      "step": 22949
    },
    {
      "entropy": 0.4265461312606931,
      "epoch": 2.1115117221867616,
      "grad_norm": 0.16635920107364655,
      "learning_rate": 2.9619406875824213e-05,
      "loss": 0.3962,
      "mean_token_accuracy": 0.8713601939380169,
      "num_tokens": 732432397.0,
      "step": 22950
    },
    {
      "entropy": 0.5253127161413431,
      "epoch": 2.1116037285582028,
      "grad_norm": 0.15897585451602936,
      "learning_rate": 2.9616340049682584e-05,
      "loss": 0.5143,
      "mean_token_accuracy": 0.8376057296991348,
      "num_tokens": 732464850.0,
      "step": 22951
    },
    {
      "entropy": 0.4850304527208209,
      "epoch": 2.111695734929644,
      "grad_norm": 0.15218375623226166,
      "learning_rate": 2.961327322354096e-05,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8499058969318867,
      "num_tokens": 732497000.0,
      "step": 22952
    },
    {
      "entropy": 0.5367869399487972,
      "epoch": 2.111787741301085,
      "grad_norm": 0.15957021713256836,
      "learning_rate": 2.961020639739933e-05,
      "loss": 0.519,
      "mean_token_accuracy": 0.8363280892372131,
      "num_tokens": 732528522.0,
      "step": 22953
    },
    {
      "entropy": 0.5166562590748072,
      "epoch": 2.111879747672526,
      "grad_norm": 0.15652406215667725,
      "learning_rate": 2.960713957125771e-05,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8416936621069908,
      "num_tokens": 732560759.0,
      "step": 22954
    },
    {
      "entropy": 0.4371368680149317,
      "epoch": 2.1119717540439678,
      "grad_norm": 0.14884580671787262,
      "learning_rate": 2.960407274511608e-05,
      "loss": 0.4167,
      "mean_token_accuracy": 0.8674926310777664,
      "num_tokens": 732592508.0,
      "step": 22955
    },
    {
      "entropy": 0.4468614892102778,
      "epoch": 2.112063760415409,
      "grad_norm": 0.14773398637771606,
      "learning_rate": 2.9601005918974454e-05,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8612879030406475,
      "num_tokens": 732624399.0,
      "step": 22956
    },
    {
      "entropy": 0.46540382062084973,
      "epoch": 2.11215576678685,
      "grad_norm": 0.1595790982246399,
      "learning_rate": 2.9597939092832826e-05,
      "loss": 0.4501,
      "mean_token_accuracy": 0.8573299869894981,
      "num_tokens": 732656219.0,
      "step": 22957
    },
    {
      "entropy": 0.4672017516568303,
      "epoch": 2.112247773158291,
      "grad_norm": 0.15688647329807281,
      "learning_rate": 2.9594872266691204e-05,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8587750010192394,
      "num_tokens": 732687237.0,
      "step": 22958
    },
    {
      "entropy": 0.5098302522674203,
      "epoch": 2.1123397795297323,
      "grad_norm": 0.1623287796974182,
      "learning_rate": 2.9591805440549575e-05,
      "loss": 0.505,
      "mean_token_accuracy": 0.8442295305430889,
      "num_tokens": 732717832.0,
      "step": 22959
    },
    {
      "entropy": 0.5751422410830855,
      "epoch": 2.112431785901174,
      "grad_norm": 0.17392219603061676,
      "learning_rate": 2.9588738614407953e-05,
      "loss": 0.5555,
      "mean_token_accuracy": 0.8264601975679398,
      "num_tokens": 732749476.0,
      "step": 22960
    },
    {
      "entropy": 0.4758309405297041,
      "epoch": 2.112523792272615,
      "grad_norm": 0.15474268794059753,
      "learning_rate": 2.9585671788266324e-05,
      "loss": 0.4537,
      "mean_token_accuracy": 0.8527866713702679,
      "num_tokens": 732781302.0,
      "step": 22961
    },
    {
      "entropy": 0.5791139714419842,
      "epoch": 2.112615798644056,
      "grad_norm": 0.1674172580242157,
      "learning_rate": 2.95826049621247e-05,
      "loss": 0.5567,
      "mean_token_accuracy": 0.8257323503494263,
      "num_tokens": 732813136.0,
      "step": 22962
    },
    {
      "entropy": 0.44055252079851925,
      "epoch": 2.1127078050154973,
      "grad_norm": 0.14758315682411194,
      "learning_rate": 2.957953813598307e-05,
      "loss": 0.4195,
      "mean_token_accuracy": 0.8674245104193687,
      "num_tokens": 732844960.0,
      "step": 22963
    },
    {
      "entropy": 0.5722184772603214,
      "epoch": 2.1127998113869384,
      "grad_norm": 0.16086113452911377,
      "learning_rate": 2.9576471309841448e-05,
      "loss": 0.548,
      "mean_token_accuracy": 0.8234597258269787,
      "num_tokens": 732877005.0,
      "step": 22964
    },
    {
      "entropy": 0.5165124982595444,
      "epoch": 2.11289181775838,
      "grad_norm": 0.16527393460273743,
      "learning_rate": 2.957340448369982e-05,
      "loss": 0.4923,
      "mean_token_accuracy": 0.8434286154806614,
      "num_tokens": 732908899.0,
      "step": 22965
    },
    {
      "entropy": 0.5845688059926033,
      "epoch": 2.112983824129821,
      "grad_norm": 0.16903910040855408,
      "learning_rate": 2.9570337657558194e-05,
      "loss": 0.5711,
      "mean_token_accuracy": 0.8192284479737282,
      "num_tokens": 732940873.0,
      "step": 22966
    },
    {
      "entropy": 0.5474888570606709,
      "epoch": 2.1130758305012622,
      "grad_norm": 0.16402818262577057,
      "learning_rate": 2.9567270831416565e-05,
      "loss": 0.5404,
      "mean_token_accuracy": 0.8284761309623718,
      "num_tokens": 732973258.0,
      "step": 22967
    },
    {
      "entropy": 0.5825364533811808,
      "epoch": 2.1131678368727034,
      "grad_norm": 0.17307820916175842,
      "learning_rate": 2.9564204005274943e-05,
      "loss": 0.5781,
      "mean_token_accuracy": 0.8168543428182602,
      "num_tokens": 733004321.0,
      "step": 22968
    },
    {
      "entropy": 0.44983753794804215,
      "epoch": 2.1132598432441445,
      "grad_norm": 0.15065181255340576,
      "learning_rate": 2.9561137179133315e-05,
      "loss": 0.4281,
      "mean_token_accuracy": 0.8635742552578449,
      "num_tokens": 733035890.0,
      "step": 22969
    },
    {
      "entropy": 0.5358957909047604,
      "epoch": 2.113351849615586,
      "grad_norm": 0.1626710295677185,
      "learning_rate": 2.955807035299169e-05,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8373383991420269,
      "num_tokens": 733068139.0,
      "step": 22970
    },
    {
      "entropy": 0.5237115048803389,
      "epoch": 2.113443855987027,
      "grad_norm": 0.16055744886398315,
      "learning_rate": 2.9555003526850067e-05,
      "loss": 0.5184,
      "mean_token_accuracy": 0.8326804675161839,
      "num_tokens": 733099734.0,
      "step": 22971
    },
    {
      "entropy": 0.5466271489858627,
      "epoch": 2.1135358623584684,
      "grad_norm": 0.16336095333099365,
      "learning_rate": 2.955193670070844e-05,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8304910138249397,
      "num_tokens": 733132007.0,
      "step": 22972
    },
    {
      "entropy": 0.5371264778077602,
      "epoch": 2.1136278687299095,
      "grad_norm": 0.16189563274383545,
      "learning_rate": 2.9548869874566813e-05,
      "loss": 0.5288,
      "mean_token_accuracy": 0.832602497190237,
      "num_tokens": 733163849.0,
      "step": 22973
    },
    {
      "entropy": 0.43162166653200984,
      "epoch": 2.1137198751013506,
      "grad_norm": 0.14894969761371613,
      "learning_rate": 2.9545803048425184e-05,
      "loss": 0.4129,
      "mean_token_accuracy": 0.8681604228913784,
      "num_tokens": 733195924.0,
      "step": 22974
    },
    {
      "entropy": 0.596395755186677,
      "epoch": 2.113811881472792,
      "grad_norm": 0.16398313641548157,
      "learning_rate": 2.9542736222283562e-05,
      "loss": 0.5807,
      "mean_token_accuracy": 0.8160440661013126,
      "num_tokens": 733227778.0,
      "step": 22975
    },
    {
      "entropy": 0.5451859571039677,
      "epoch": 2.1139038878442333,
      "grad_norm": 0.16372452676296234,
      "learning_rate": 2.9539669396141934e-05,
      "loss": 0.5367,
      "mean_token_accuracy": 0.8292380571365356,
      "num_tokens": 733260025.0,
      "step": 22976
    },
    {
      "entropy": 0.49705908773466945,
      "epoch": 2.1139958942156745,
      "grad_norm": 0.15636087954044342,
      "learning_rate": 2.9536602570000312e-05,
      "loss": 0.4862,
      "mean_token_accuracy": 0.8494794853031635,
      "num_tokens": 733291766.0,
      "step": 22977
    },
    {
      "entropy": 0.5729357455857098,
      "epoch": 2.1140879005871156,
      "grad_norm": 0.16622354090213776,
      "learning_rate": 2.953353574385868e-05,
      "loss": 0.5571,
      "mean_token_accuracy": 0.8229785859584808,
      "num_tokens": 733323120.0,
      "step": 22978
    },
    {
      "entropy": 0.5913463681936264,
      "epoch": 2.1141799069585567,
      "grad_norm": 0.16647778451442719,
      "learning_rate": 2.9530468917717058e-05,
      "loss": 0.5635,
      "mean_token_accuracy": 0.8207880072295666,
      "num_tokens": 733355148.0,
      "step": 22979
    },
    {
      "entropy": 0.5197525732219219,
      "epoch": 2.1142719133299983,
      "grad_norm": 0.1576247215270996,
      "learning_rate": 2.952740209157543e-05,
      "loss": 0.5001,
      "mean_token_accuracy": 0.8421337343752384,
      "num_tokens": 733387629.0,
      "step": 22980
    },
    {
      "entropy": 0.4866061247885227,
      "epoch": 2.1143639197014394,
      "grad_norm": 0.15138459205627441,
      "learning_rate": 2.9524335265433807e-05,
      "loss": 0.4641,
      "mean_token_accuracy": 0.8534373678267002,
      "num_tokens": 733419280.0,
      "step": 22981
    },
    {
      "entropy": 0.45902549382299185,
      "epoch": 2.1144559260728806,
      "grad_norm": 0.15692384541034698,
      "learning_rate": 2.9521268439292178e-05,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8552880734205246,
      "num_tokens": 733450990.0,
      "step": 22982
    },
    {
      "entropy": 0.4976999806240201,
      "epoch": 2.1145479324443217,
      "grad_norm": 0.15750384330749512,
      "learning_rate": 2.9518201613150553e-05,
      "loss": 0.4747,
      "mean_token_accuracy": 0.84702393040061,
      "num_tokens": 733482410.0,
      "step": 22983
    },
    {
      "entropy": 0.5833371449261904,
      "epoch": 2.114639938815763,
      "grad_norm": 0.17145170271396637,
      "learning_rate": 2.9515134787008924e-05,
      "loss": 0.5789,
      "mean_token_accuracy": 0.8201572299003601,
      "num_tokens": 733514651.0,
      "step": 22984
    },
    {
      "entropy": 0.4701555445790291,
      "epoch": 2.1147319451872044,
      "grad_norm": 0.1596551388502121,
      "learning_rate": 2.9512067960867302e-05,
      "loss": 0.4516,
      "mean_token_accuracy": 0.8554685339331627,
      "num_tokens": 733546304.0,
      "step": 22985
    },
    {
      "entropy": 0.4587017558515072,
      "epoch": 2.1148239515586456,
      "grad_norm": 0.1545988768339157,
      "learning_rate": 2.9509001134725673e-05,
      "loss": 0.4347,
      "mean_token_accuracy": 0.8612982109189034,
      "num_tokens": 733577604.0,
      "step": 22986
    },
    {
      "entropy": 0.45231487043201923,
      "epoch": 2.1149159579300867,
      "grad_norm": 0.1534222662448883,
      "learning_rate": 2.9505934308584048e-05,
      "loss": 0.4462,
      "mean_token_accuracy": 0.8568314276635647,
      "num_tokens": 733609970.0,
      "step": 22987
    },
    {
      "entropy": 0.5386650075670332,
      "epoch": 2.115007964301528,
      "grad_norm": 0.15849736332893372,
      "learning_rate": 2.950286748244242e-05,
      "loss": 0.5301,
      "mean_token_accuracy": 0.8336610458791256,
      "num_tokens": 733642356.0,
      "step": 22988
    },
    {
      "entropy": 0.5897831153124571,
      "epoch": 2.115099970672969,
      "grad_norm": 0.1715707778930664,
      "learning_rate": 2.9499800656300797e-05,
      "loss": 0.5975,
      "mean_token_accuracy": 0.8144503310322762,
      "num_tokens": 733674188.0,
      "step": 22989
    },
    {
      "entropy": 0.50966956326738,
      "epoch": 2.1151919770444105,
      "grad_norm": 0.16328264772891998,
      "learning_rate": 2.949673383015917e-05,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8437913618981838,
      "num_tokens": 733706212.0,
      "step": 22990
    },
    {
      "entropy": 0.5148949045687914,
      "epoch": 2.1152839834158517,
      "grad_norm": 0.1668267399072647,
      "learning_rate": 2.9493667004017543e-05,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8385557271540165,
      "num_tokens": 733738149.0,
      "step": 22991
    },
    {
      "entropy": 0.5611559904646128,
      "epoch": 2.115375989787293,
      "grad_norm": 0.16689181327819824,
      "learning_rate": 2.9490600177875915e-05,
      "loss": 0.5577,
      "mean_token_accuracy": 0.8253151476383209,
      "num_tokens": 733769913.0,
      "step": 22992
    },
    {
      "entropy": 0.5238634292036295,
      "epoch": 2.115467996158734,
      "grad_norm": 0.16698522865772247,
      "learning_rate": 2.9487533351734293e-05,
      "loss": 0.519,
      "mean_token_accuracy": 0.8327458947896957,
      "num_tokens": 733802068.0,
      "step": 22993
    },
    {
      "entropy": 0.5479928019922227,
      "epoch": 2.115560002530175,
      "grad_norm": 0.1601426899433136,
      "learning_rate": 2.9484466525592664e-05,
      "loss": 0.5341,
      "mean_token_accuracy": 0.8331029824912548,
      "num_tokens": 733834218.0,
      "step": 22994
    },
    {
      "entropy": 0.47201912850141525,
      "epoch": 2.1156520089016166,
      "grad_norm": 0.16040806472301483,
      "learning_rate": 2.948139969945104e-05,
      "loss": 0.462,
      "mean_token_accuracy": 0.8514688722789288,
      "num_tokens": 733865527.0,
      "step": 22995
    },
    {
      "entropy": 0.5030275373719633,
      "epoch": 2.1157440152730578,
      "grad_norm": 0.16429398953914642,
      "learning_rate": 2.947833287330941e-05,
      "loss": 0.4921,
      "mean_token_accuracy": 0.844839196652174,
      "num_tokens": 733897531.0,
      "step": 22996
    },
    {
      "entropy": 0.5904878191649914,
      "epoch": 2.115836021644499,
      "grad_norm": 0.1738712191581726,
      "learning_rate": 2.9475266047167788e-05,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8173486776649952,
      "num_tokens": 733928683.0,
      "step": 22997
    },
    {
      "entropy": 0.46417500264942646,
      "epoch": 2.11592802801594,
      "grad_norm": 0.15016086399555206,
      "learning_rate": 2.947219922102616e-05,
      "loss": 0.45,
      "mean_token_accuracy": 0.8585388362407684,
      "num_tokens": 733960537.0,
      "step": 22998
    },
    {
      "entropy": 0.5711429324001074,
      "epoch": 2.116020034387381,
      "grad_norm": 0.17055797576904297,
      "learning_rate": 2.9469132394884537e-05,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8248819969594479,
      "num_tokens": 733991304.0,
      "step": 22999
    },
    {
      "entropy": 0.4867242481559515,
      "epoch": 2.1161120407588228,
      "grad_norm": 0.15362833440303802,
      "learning_rate": 2.9466065568742912e-05,
      "loss": 0.476,
      "mean_token_accuracy": 0.8511790372431278,
      "num_tokens": 734023435.0,
      "step": 23000
    },
    {
      "entropy": 0.5475149899721146,
      "epoch": 2.116204047130264,
      "grad_norm": 0.1657239943742752,
      "learning_rate": 2.9462998742601283e-05,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8348961882293224,
      "num_tokens": 734055222.0,
      "step": 23001
    },
    {
      "entropy": 0.5073411948978901,
      "epoch": 2.116296053501705,
      "grad_norm": 0.1613200157880783,
      "learning_rate": 2.945993191645966e-05,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8474737256765366,
      "num_tokens": 734087407.0,
      "step": 23002
    },
    {
      "entropy": 0.4744142016861588,
      "epoch": 2.116388059873146,
      "grad_norm": 0.15311206877231598,
      "learning_rate": 2.9456865090318032e-05,
      "loss": 0.4561,
      "mean_token_accuracy": 0.8539328873157501,
      "num_tokens": 734120044.0,
      "step": 23003
    },
    {
      "entropy": 0.5718567054718733,
      "epoch": 2.1164800662445873,
      "grad_norm": 0.16708368062973022,
      "learning_rate": 2.9453798264176407e-05,
      "loss": 0.549,
      "mean_token_accuracy": 0.8289631009101868,
      "num_tokens": 734151674.0,
      "step": 23004
    },
    {
      "entropy": 0.5376359988003969,
      "epoch": 2.116572072616029,
      "grad_norm": 0.16085897386074066,
      "learning_rate": 2.9450731438034778e-05,
      "loss": 0.5198,
      "mean_token_accuracy": 0.8383161649107933,
      "num_tokens": 734183431.0,
      "step": 23005
    },
    {
      "entropy": 0.4791726302355528,
      "epoch": 2.11666407898747,
      "grad_norm": 0.15086856484413147,
      "learning_rate": 2.9447664611893156e-05,
      "loss": 0.4649,
      "mean_token_accuracy": 0.8497005179524422,
      "num_tokens": 734215721.0,
      "step": 23006
    },
    {
      "entropy": 0.5074436943978071,
      "epoch": 2.116756085358911,
      "grad_norm": 0.15674111247062683,
      "learning_rate": 2.9444597785751527e-05,
      "loss": 0.4845,
      "mean_token_accuracy": 0.8455780111253262,
      "num_tokens": 734248016.0,
      "step": 23007
    },
    {
      "entropy": 0.40993557358160615,
      "epoch": 2.1168480917303523,
      "grad_norm": 0.14563167095184326,
      "learning_rate": 2.9441530959609902e-05,
      "loss": 0.3867,
      "mean_token_accuracy": 0.8709473572671413,
      "num_tokens": 734279642.0,
      "step": 23008
    },
    {
      "entropy": 0.45695146452635527,
      "epoch": 2.1169400981017934,
      "grad_norm": 0.1528284102678299,
      "learning_rate": 2.9438464133468273e-05,
      "loss": 0.438,
      "mean_token_accuracy": 0.8591111153364182,
      "num_tokens": 734311293.0,
      "step": 23009
    },
    {
      "entropy": 0.5417654728516936,
      "epoch": 2.117032104473235,
      "grad_norm": 0.16300764679908752,
      "learning_rate": 2.943539730732665e-05,
      "loss": 0.5277,
      "mean_token_accuracy": 0.8323596641421318,
      "num_tokens": 734343452.0,
      "step": 23010
    },
    {
      "entropy": 0.46153734903782606,
      "epoch": 2.117124110844676,
      "grad_norm": 0.15311802923679352,
      "learning_rate": 2.9432330481185023e-05,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8546367697417736,
      "num_tokens": 734375692.0,
      "step": 23011
    },
    {
      "entropy": 0.5005263593047857,
      "epoch": 2.1172161172161172,
      "grad_norm": 0.1604296863079071,
      "learning_rate": 2.9429263655043397e-05,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8435325883328915,
      "num_tokens": 734407969.0,
      "step": 23012
    },
    {
      "entropy": 0.37510207667946815,
      "epoch": 2.1173081235875584,
      "grad_norm": 0.14223834872245789,
      "learning_rate": 2.942619682890177e-05,
      "loss": 0.3604,
      "mean_token_accuracy": 0.8827738612890244,
      "num_tokens": 734440170.0,
      "step": 23013
    },
    {
      "entropy": 0.5248089581727982,
      "epoch": 2.1174001299589995,
      "grad_norm": 0.16018320620059967,
      "learning_rate": 2.9423130002760147e-05,
      "loss": 0.516,
      "mean_token_accuracy": 0.8350024372339249,
      "num_tokens": 734472938.0,
      "step": 23014
    },
    {
      "entropy": 0.5442456882447004,
      "epoch": 2.117492136330441,
      "grad_norm": 0.16131961345672607,
      "learning_rate": 2.9420063176618518e-05,
      "loss": 0.5468,
      "mean_token_accuracy": 0.8305110521614552,
      "num_tokens": 734504641.0,
      "step": 23015
    },
    {
      "entropy": 0.44221052853390574,
      "epoch": 2.117584142701882,
      "grad_norm": 0.14877717196941376,
      "learning_rate": 2.9416996350476893e-05,
      "loss": 0.4247,
      "mean_token_accuracy": 0.8639558963477612,
      "num_tokens": 734537203.0,
      "step": 23016
    },
    {
      "entropy": 0.5430961977690458,
      "epoch": 2.1176761490733234,
      "grad_norm": 0.16490846872329712,
      "learning_rate": 2.9413929524335264e-05,
      "loss": 0.5349,
      "mean_token_accuracy": 0.8320132941007614,
      "num_tokens": 734569311.0,
      "step": 23017
    },
    {
      "entropy": 0.5425811652094126,
      "epoch": 2.1177681554447645,
      "grad_norm": 0.1663912981748581,
      "learning_rate": 2.9410862698193642e-05,
      "loss": 0.532,
      "mean_token_accuracy": 0.8308934569358826,
      "num_tokens": 734600738.0,
      "step": 23018
    },
    {
      "entropy": 0.45792086934670806,
      "epoch": 2.1178601618162056,
      "grad_norm": 0.15161532163619995,
      "learning_rate": 2.9407795872052013e-05,
      "loss": 0.4462,
      "mean_token_accuracy": 0.8599408641457558,
      "num_tokens": 734632707.0,
      "step": 23019
    },
    {
      "entropy": 0.5229132324457169,
      "epoch": 2.117952168187647,
      "grad_norm": 0.16369308531284332,
      "learning_rate": 2.940472904591039e-05,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8398680947721004,
      "num_tokens": 734664205.0,
      "step": 23020
    },
    {
      "entropy": 0.5877560395747423,
      "epoch": 2.1180441745590883,
      "grad_norm": 0.17134368419647217,
      "learning_rate": 2.9401662219768762e-05,
      "loss": 0.5701,
      "mean_token_accuracy": 0.819709800183773,
      "num_tokens": 734695618.0,
      "step": 23021
    },
    {
      "entropy": 0.44748622365295887,
      "epoch": 2.1181361809305295,
      "grad_norm": 0.15812332928180695,
      "learning_rate": 2.9398595393627137e-05,
      "loss": 0.4403,
      "mean_token_accuracy": 0.8559484779834747,
      "num_tokens": 734726970.0,
      "step": 23022
    },
    {
      "entropy": 0.5558976419270039,
      "epoch": 2.1182281873019706,
      "grad_norm": 0.16792817413806915,
      "learning_rate": 2.9395528567485508e-05,
      "loss": 0.5429,
      "mean_token_accuracy": 0.8281878307461739,
      "num_tokens": 734758874.0,
      "step": 23023
    },
    {
      "entropy": 0.5103856297209859,
      "epoch": 2.1183201936734117,
      "grad_norm": 0.15834486484527588,
      "learning_rate": 2.9392461741343886e-05,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8440880700945854,
      "num_tokens": 734790788.0,
      "step": 23024
    },
    {
      "entropy": 0.5265940055251122,
      "epoch": 2.1184122000448533,
      "grad_norm": 0.16825734078884125,
      "learning_rate": 2.9389394915202258e-05,
      "loss": 0.5305,
      "mean_token_accuracy": 0.835325762629509,
      "num_tokens": 734823044.0,
      "step": 23025
    },
    {
      "entropy": 0.549311101436615,
      "epoch": 2.1185042064162944,
      "grad_norm": 0.16421177983283997,
      "learning_rate": 2.9386328089060632e-05,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8305770084261894,
      "num_tokens": 734854867.0,
      "step": 23026
    },
    {
      "entropy": 0.5865439549088478,
      "epoch": 2.1185962127877356,
      "grad_norm": 0.16925105452537537,
      "learning_rate": 2.9383261262919003e-05,
      "loss": 0.5709,
      "mean_token_accuracy": 0.8176722936332226,
      "num_tokens": 734886792.0,
      "step": 23027
    },
    {
      "entropy": 0.41712150909006596,
      "epoch": 2.1186882191591767,
      "grad_norm": 0.1454523354768753,
      "learning_rate": 2.938019443677738e-05,
      "loss": 0.3936,
      "mean_token_accuracy": 0.8716934695839882,
      "num_tokens": 734919014.0,
      "step": 23028
    },
    {
      "entropy": 0.5758349541574717,
      "epoch": 2.118780225530618,
      "grad_norm": 0.16556179523468018,
      "learning_rate": 2.9377127610635753e-05,
      "loss": 0.5609,
      "mean_token_accuracy": 0.8229437209665775,
      "num_tokens": 734951077.0,
      "step": 23029
    },
    {
      "entropy": 0.5812749285250902,
      "epoch": 2.1188722319020594,
      "grad_norm": 0.17255935072898865,
      "learning_rate": 2.9374060784494127e-05,
      "loss": 0.5717,
      "mean_token_accuracy": 0.8209838941693306,
      "num_tokens": 734983158.0,
      "step": 23030
    },
    {
      "entropy": 0.525487856939435,
      "epoch": 2.1189642382735006,
      "grad_norm": 0.15979699790477753,
      "learning_rate": 2.9370993958352505e-05,
      "loss": 0.5008,
      "mean_token_accuracy": 0.8393903598189354,
      "num_tokens": 735015634.0,
      "step": 23031
    },
    {
      "entropy": 0.5785303711891174,
      "epoch": 2.1190562446449417,
      "grad_norm": 0.1714034229516983,
      "learning_rate": 2.9367927132210877e-05,
      "loss": 0.567,
      "mean_token_accuracy": 0.8223626092076302,
      "num_tokens": 735047389.0,
      "step": 23032
    },
    {
      "entropy": 0.5841526193544269,
      "epoch": 2.119148251016383,
      "grad_norm": 0.16327854990959167,
      "learning_rate": 2.936486030606925e-05,
      "loss": 0.5668,
      "mean_token_accuracy": 0.8209844753146172,
      "num_tokens": 735080056.0,
      "step": 23033
    },
    {
      "entropy": 0.5000892924144864,
      "epoch": 2.119240257387824,
      "grad_norm": 0.15553095936775208,
      "learning_rate": 2.9361793479927623e-05,
      "loss": 0.4736,
      "mean_token_accuracy": 0.848675861954689,
      "num_tokens": 735112064.0,
      "step": 23034
    },
    {
      "entropy": 0.5507258819416165,
      "epoch": 2.1193322637592655,
      "grad_norm": 0.16133643686771393,
      "learning_rate": 2.9358726653786e-05,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8321410864591599,
      "num_tokens": 735144217.0,
      "step": 23035
    },
    {
      "entropy": 0.41917406767606735,
      "epoch": 2.1194242701307067,
      "grad_norm": 0.14266803860664368,
      "learning_rate": 2.9355659827644372e-05,
      "loss": 0.3938,
      "mean_token_accuracy": 0.8741470873355865,
      "num_tokens": 735176958.0,
      "step": 23036
    },
    {
      "entropy": 0.5852385452017188,
      "epoch": 2.119516276502148,
      "grad_norm": 0.1655765175819397,
      "learning_rate": 2.935259300150275e-05,
      "loss": 0.5677,
      "mean_token_accuracy": 0.8222079016268253,
      "num_tokens": 735209667.0,
      "step": 23037
    },
    {
      "entropy": 0.5375344874337316,
      "epoch": 2.119608282873589,
      "grad_norm": 0.16670112311840057,
      "learning_rate": 2.9349526175361118e-05,
      "loss": 0.523,
      "mean_token_accuracy": 0.8348597884178162,
      "num_tokens": 735240929.0,
      "step": 23038
    },
    {
      "entropy": 0.6200554110109806,
      "epoch": 2.11970028924503,
      "grad_norm": 0.17110662162303925,
      "learning_rate": 2.9346459349219496e-05,
      "loss": 0.6168,
      "mean_token_accuracy": 0.8099636361002922,
      "num_tokens": 735273172.0,
      "step": 23039
    },
    {
      "entropy": 0.5430023679509759,
      "epoch": 2.1197922956164716,
      "grad_norm": 0.1595020443201065,
      "learning_rate": 2.9343392523077867e-05,
      "loss": 0.5107,
      "mean_token_accuracy": 0.8364286310970783,
      "num_tokens": 735304944.0,
      "step": 23040
    },
    {
      "entropy": 0.5836303299292922,
      "epoch": 2.119884301987913,
      "grad_norm": 0.16468366980552673,
      "learning_rate": 2.9340325696936245e-05,
      "loss": 0.5663,
      "mean_token_accuracy": 0.8188655488193035,
      "num_tokens": 735336711.0,
      "step": 23041
    },
    {
      "entropy": 0.42961128847673535,
      "epoch": 2.119976308359354,
      "grad_norm": 0.14939260482788086,
      "learning_rate": 2.9337258870794616e-05,
      "loss": 0.4142,
      "mean_token_accuracy": 0.8664553463459015,
      "num_tokens": 735368981.0,
      "step": 23042
    },
    {
      "entropy": 0.5908795222640038,
      "epoch": 2.120068314730795,
      "grad_norm": 0.17588165402412415,
      "learning_rate": 2.933419204465299e-05,
      "loss": 0.5879,
      "mean_token_accuracy": 0.8161779008805752,
      "num_tokens": 735400718.0,
      "step": 23043
    },
    {
      "entropy": 0.5369321536272764,
      "epoch": 2.120160321102236,
      "grad_norm": 0.168581023812294,
      "learning_rate": 2.9331125218511362e-05,
      "loss": 0.5249,
      "mean_token_accuracy": 0.8328633829951286,
      "num_tokens": 735431721.0,
      "step": 23044
    },
    {
      "entropy": 0.40781629690900445,
      "epoch": 2.1202523274736778,
      "grad_norm": 0.1472863107919693,
      "learning_rate": 2.932805839236974e-05,
      "loss": 0.3889,
      "mean_token_accuracy": 0.8753341995179653,
      "num_tokens": 735464234.0,
      "step": 23045
    },
    {
      "entropy": 0.5305438935756683,
      "epoch": 2.120344333845119,
      "grad_norm": 0.16366370022296906,
      "learning_rate": 2.932499156622811e-05,
      "loss": 0.5261,
      "mean_token_accuracy": 0.8315334729850292,
      "num_tokens": 735496048.0,
      "step": 23046
    },
    {
      "entropy": 0.5730534344911575,
      "epoch": 2.12043634021656,
      "grad_norm": 0.16899526119232178,
      "learning_rate": 2.9321924740086486e-05,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8210141696035862,
      "num_tokens": 735527868.0,
      "step": 23047
    },
    {
      "entropy": 0.49132755445316434,
      "epoch": 2.120528346588001,
      "grad_norm": 0.15587320923805237,
      "learning_rate": 2.9318857913944858e-05,
      "loss": 0.4753,
      "mean_token_accuracy": 0.8517851568758488,
      "num_tokens": 735559943.0,
      "step": 23048
    },
    {
      "entropy": 0.587442131829448,
      "epoch": 2.1206203529594423,
      "grad_norm": 0.1654042750597,
      "learning_rate": 2.9315791087803236e-05,
      "loss": 0.5799,
      "mean_token_accuracy": 0.8185969218611717,
      "num_tokens": 735592319.0,
      "step": 23049
    },
    {
      "entropy": 0.5434135906398296,
      "epoch": 2.120712359330884,
      "grad_norm": 0.16625817120075226,
      "learning_rate": 2.9312724261661607e-05,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8338644430041313,
      "num_tokens": 735624466.0,
      "step": 23050
    },
    {
      "entropy": 0.4662847789004445,
      "epoch": 2.120804365702325,
      "grad_norm": 0.1578216403722763,
      "learning_rate": 2.930965743551998e-05,
      "loss": 0.453,
      "mean_token_accuracy": 0.8554643541574478,
      "num_tokens": 735656222.0,
      "step": 23051
    },
    {
      "entropy": 0.5618917597457767,
      "epoch": 2.120896372073766,
      "grad_norm": 0.17065247893333435,
      "learning_rate": 2.9306590609378353e-05,
      "loss": 0.5569,
      "mean_token_accuracy": 0.8252481110394001,
      "num_tokens": 735687945.0,
      "step": 23052
    },
    {
      "entropy": 0.515143716474995,
      "epoch": 2.1209883784452073,
      "grad_norm": 0.16408826410770416,
      "learning_rate": 2.930352378323673e-05,
      "loss": 0.4943,
      "mean_token_accuracy": 0.8414772227406502,
      "num_tokens": 735719498.0,
      "step": 23053
    },
    {
      "entropy": 0.4916507937014103,
      "epoch": 2.1210803848166484,
      "grad_norm": 0.1587669998407364,
      "learning_rate": 2.9300456957095102e-05,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8470639064908028,
      "num_tokens": 735751554.0,
      "step": 23054
    },
    {
      "entropy": 0.6124073192477226,
      "epoch": 2.12117239118809,
      "grad_norm": 0.1728622317314148,
      "learning_rate": 2.9297390130953477e-05,
      "loss": 0.6085,
      "mean_token_accuracy": 0.8075217753648758,
      "num_tokens": 735783344.0,
      "step": 23055
    },
    {
      "entropy": 0.520302087534219,
      "epoch": 2.121264397559531,
      "grad_norm": 0.15913887321949005,
      "learning_rate": 2.9294323304811848e-05,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8428735174238682,
      "num_tokens": 735815019.0,
      "step": 23056
    },
    {
      "entropy": 0.44701469130814075,
      "epoch": 2.1213564039309722,
      "grad_norm": 0.1521366983652115,
      "learning_rate": 2.9291256478670226e-05,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8615834973752499,
      "num_tokens": 735846898.0,
      "step": 23057
    },
    {
      "entropy": 0.535870186984539,
      "epoch": 2.1214484103024134,
      "grad_norm": 0.15958908200263977,
      "learning_rate": 2.9288189652528597e-05,
      "loss": 0.5176,
      "mean_token_accuracy": 0.8351478241384029,
      "num_tokens": 735879237.0,
      "step": 23058
    },
    {
      "entropy": 0.4117091950029135,
      "epoch": 2.1215404166738545,
      "grad_norm": 0.1489616483449936,
      "learning_rate": 2.9285122826386975e-05,
      "loss": 0.3988,
      "mean_token_accuracy": 0.8725189529359341,
      "num_tokens": 735911257.0,
      "step": 23059
    },
    {
      "entropy": 0.5570105761289597,
      "epoch": 2.121632423045296,
      "grad_norm": 0.16827310621738434,
      "learning_rate": 2.9282056000245343e-05,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8289528787136078,
      "num_tokens": 735941366.0,
      "step": 23060
    },
    {
      "entropy": 0.5008105663582683,
      "epoch": 2.1217244294167372,
      "grad_norm": 0.15933308005332947,
      "learning_rate": 2.927898917410372e-05,
      "loss": 0.4887,
      "mean_token_accuracy": 0.8480590842664242,
      "num_tokens": 735973772.0,
      "step": 23061
    },
    {
      "entropy": 0.5754896672442555,
      "epoch": 2.1218164357881784,
      "grad_norm": 0.16805700957775116,
      "learning_rate": 2.92759223479621e-05,
      "loss": 0.5538,
      "mean_token_accuracy": 0.8242790251970291,
      "num_tokens": 736006374.0,
      "step": 23062
    },
    {
      "entropy": 0.49993452057242393,
      "epoch": 2.1219084421596195,
      "grad_norm": 0.16231563687324524,
      "learning_rate": 2.927285552182047e-05,
      "loss": 0.4833,
      "mean_token_accuracy": 0.8442856669425964,
      "num_tokens": 736038575.0,
      "step": 23063
    },
    {
      "entropy": 0.5907791238278151,
      "epoch": 2.1220004485310606,
      "grad_norm": 0.1703556478023529,
      "learning_rate": 2.9269788695678845e-05,
      "loss": 0.5679,
      "mean_token_accuracy": 0.8202580474317074,
      "num_tokens": 736069995.0,
      "step": 23064
    },
    {
      "entropy": 0.5237251101061702,
      "epoch": 2.122092454902502,
      "grad_norm": 0.16546590626239777,
      "learning_rate": 2.9266721869537216e-05,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8463820554316044,
      "num_tokens": 736101383.0,
      "step": 23065
    },
    {
      "entropy": 0.4811182804405689,
      "epoch": 2.1221844612739433,
      "grad_norm": 0.15421755611896515,
      "learning_rate": 2.9263655043395594e-05,
      "loss": 0.4644,
      "mean_token_accuracy": 0.8497050441801548,
      "num_tokens": 736133030.0,
      "step": 23066
    },
    {
      "entropy": 0.5699380896985531,
      "epoch": 2.1222764676453845,
      "grad_norm": 0.16516618430614471,
      "learning_rate": 2.9260588217253966e-05,
      "loss": 0.5544,
      "mean_token_accuracy": 0.8257325515151024,
      "num_tokens": 736165128.0,
      "step": 23067
    },
    {
      "entropy": 0.5359950135461986,
      "epoch": 2.1223684740168256,
      "grad_norm": 0.15846490859985352,
      "learning_rate": 2.925752139111234e-05,
      "loss": 0.5172,
      "mean_token_accuracy": 0.8367463871836662,
      "num_tokens": 736197313.0,
      "step": 23068
    },
    {
      "entropy": 0.5572112007066607,
      "epoch": 2.1224604803882667,
      "grad_norm": 0.16274087131023407,
      "learning_rate": 2.925445456497071e-05,
      "loss": 0.5475,
      "mean_token_accuracy": 0.8263343460857868,
      "num_tokens": 736229351.0,
      "step": 23069
    },
    {
      "entropy": 0.6201326660811901,
      "epoch": 2.1225524867597083,
      "grad_norm": 0.17621977627277374,
      "learning_rate": 2.925138773882909e-05,
      "loss": 0.6136,
      "mean_token_accuracy": 0.8084195777773857,
      "num_tokens": 736261298.0,
      "step": 23070
    },
    {
      "entropy": 0.4857295239344239,
      "epoch": 2.1226444931311494,
      "grad_norm": 0.16110165417194366,
      "learning_rate": 2.924832091268746e-05,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8488671705126762,
      "num_tokens": 736292887.0,
      "step": 23071
    },
    {
      "entropy": 0.5270316076930612,
      "epoch": 2.1227364995025906,
      "grad_norm": 0.1647619903087616,
      "learning_rate": 2.9245254086545836e-05,
      "loss": 0.5255,
      "mean_token_accuracy": 0.8341190405189991,
      "num_tokens": 736325114.0,
      "step": 23072
    },
    {
      "entropy": 0.510208903811872,
      "epoch": 2.1228285058740317,
      "grad_norm": 0.1649819165468216,
      "learning_rate": 2.9242187260404207e-05,
      "loss": 0.4987,
      "mean_token_accuracy": 0.8408929221332073,
      "num_tokens": 736356077.0,
      "step": 23073
    },
    {
      "entropy": 0.5287308720871806,
      "epoch": 2.122920512245473,
      "grad_norm": 0.16291484236717224,
      "learning_rate": 2.9239120434262585e-05,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8417009823024273,
      "num_tokens": 736387339.0,
      "step": 23074
    },
    {
      "entropy": 0.4330770382657647,
      "epoch": 2.1230125186169144,
      "grad_norm": 0.15034671127796173,
      "learning_rate": 2.9236053608120956e-05,
      "loss": 0.4238,
      "mean_token_accuracy": 0.863451074808836,
      "num_tokens": 736419224.0,
      "step": 23075
    },
    {
      "entropy": 0.4672206463292241,
      "epoch": 2.1231045249883556,
      "grad_norm": 0.15426436066627502,
      "learning_rate": 2.923298678197933e-05,
      "loss": 0.4619,
      "mean_token_accuracy": 0.8513228595256805,
      "num_tokens": 736451708.0,
      "step": 23076
    },
    {
      "entropy": 0.5215802881866693,
      "epoch": 2.1231965313597967,
      "grad_norm": 0.16121964156627655,
      "learning_rate": 2.9229919955837702e-05,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8378574103116989,
      "num_tokens": 736483672.0,
      "step": 23077
    },
    {
      "entropy": 0.5579599402844906,
      "epoch": 2.123288537731238,
      "grad_norm": 0.16748046875,
      "learning_rate": 2.922685312969608e-05,
      "loss": 0.545,
      "mean_token_accuracy": 0.8290397599339485,
      "num_tokens": 736515622.0,
      "step": 23078
    },
    {
      "entropy": 0.5736380573362112,
      "epoch": 2.123380544102679,
      "grad_norm": 0.16761577129364014,
      "learning_rate": 2.922378630355445e-05,
      "loss": 0.5599,
      "mean_token_accuracy": 0.824592262506485,
      "num_tokens": 736547935.0,
      "step": 23079
    },
    {
      "entropy": 0.4704212467186153,
      "epoch": 2.1234725504741205,
      "grad_norm": 0.1582295000553131,
      "learning_rate": 2.922071947741283e-05,
      "loss": 0.4593,
      "mean_token_accuracy": 0.8537211641669273,
      "num_tokens": 736579703.0,
      "step": 23080
    },
    {
      "entropy": 0.5476798079907894,
      "epoch": 2.1235645568455617,
      "grad_norm": 0.1598007082939148,
      "learning_rate": 2.92176526512712e-05,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8312407471239567,
      "num_tokens": 736611782.0,
      "step": 23081
    },
    {
      "entropy": 0.5067064072936773,
      "epoch": 2.123656563217003,
      "grad_norm": 0.16050124168395996,
      "learning_rate": 2.9214585825129575e-05,
      "loss": 0.4941,
      "mean_token_accuracy": 0.8425111696124077,
      "num_tokens": 736643365.0,
      "step": 23082
    },
    {
      "entropy": 0.46196133084595203,
      "epoch": 2.123748569588444,
      "grad_norm": 0.15402275323867798,
      "learning_rate": 2.9211518998987946e-05,
      "loss": 0.4454,
      "mean_token_accuracy": 0.8581111654639244,
      "num_tokens": 736675175.0,
      "step": 23083
    },
    {
      "entropy": 0.5093721067532897,
      "epoch": 2.123840575959885,
      "grad_norm": 0.16433660686016083,
      "learning_rate": 2.9208452172846325e-05,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8461402207612991,
      "num_tokens": 736706235.0,
      "step": 23084
    },
    {
      "entropy": 0.41518250899389386,
      "epoch": 2.1239325823313266,
      "grad_norm": 0.14924108982086182,
      "learning_rate": 2.9205385346704696e-05,
      "loss": 0.4073,
      "mean_token_accuracy": 0.870349008589983,
      "num_tokens": 736738532.0,
      "step": 23085
    },
    {
      "entropy": 0.5824185134842992,
      "epoch": 2.124024588702768,
      "grad_norm": 0.16828428208827972,
      "learning_rate": 2.920231852056307e-05,
      "loss": 0.5681,
      "mean_token_accuracy": 0.8200098797678947,
      "num_tokens": 736770545.0,
      "step": 23086
    },
    {
      "entropy": 0.5486732888966799,
      "epoch": 2.124116595074209,
      "grad_norm": 0.16522681713104248,
      "learning_rate": 2.919925169442144e-05,
      "loss": 0.5268,
      "mean_token_accuracy": 0.8335902243852615,
      "num_tokens": 736802103.0,
      "step": 23087
    },
    {
      "entropy": 0.521499210037291,
      "epoch": 2.12420860144565,
      "grad_norm": 0.16007177531719208,
      "learning_rate": 2.919618486827982e-05,
      "loss": 0.5085,
      "mean_token_accuracy": 0.8401090763509274,
      "num_tokens": 736833839.0,
      "step": 23088
    },
    {
      "entropy": 0.5070373127236962,
      "epoch": 2.124300607817091,
      "grad_norm": 0.1634318232536316,
      "learning_rate": 2.919311804213819e-05,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8419205099344254,
      "num_tokens": 736865774.0,
      "step": 23089
    },
    {
      "entropy": 0.5276211691088974,
      "epoch": 2.1243926141885328,
      "grad_norm": 0.1645893156528473,
      "learning_rate": 2.9190051215996566e-05,
      "loss": 0.5146,
      "mean_token_accuracy": 0.8393618501722813,
      "num_tokens": 736897557.0,
      "step": 23090
    },
    {
      "entropy": 0.47562431637197733,
      "epoch": 2.124484620559974,
      "grad_norm": 0.15063761174678802,
      "learning_rate": 2.9186984389854937e-05,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8519224300980568,
      "num_tokens": 736929767.0,
      "step": 23091
    },
    {
      "entropy": 0.4971851734444499,
      "epoch": 2.124576626931415,
      "grad_norm": 0.15835461020469666,
      "learning_rate": 2.9183917563713315e-05,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8472282961010933,
      "num_tokens": 736961752.0,
      "step": 23092
    },
    {
      "entropy": 0.5617610393092036,
      "epoch": 2.124668633302856,
      "grad_norm": 0.16655360162258148,
      "learning_rate": 2.918085073757169e-05,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8277998752892017,
      "num_tokens": 736993419.0,
      "step": 23093
    },
    {
      "entropy": 0.5117303971201181,
      "epoch": 2.1247606396742973,
      "grad_norm": 0.17002961039543152,
      "learning_rate": 2.917778391143006e-05,
      "loss": 0.4979,
      "mean_token_accuracy": 0.8401886001229286,
      "num_tokens": 737024985.0,
      "step": 23094
    },
    {
      "entropy": 0.541997566819191,
      "epoch": 2.124852646045739,
      "grad_norm": 0.1622479408979416,
      "learning_rate": 2.917471708528844e-05,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8285287581384182,
      "num_tokens": 737056807.0,
      "step": 23095
    },
    {
      "entropy": 0.5791343878954649,
      "epoch": 2.12494465241718,
      "grad_norm": 0.1673947423696518,
      "learning_rate": 2.917165025914681e-05,
      "loss": 0.5605,
      "mean_token_accuracy": 0.8204672522842884,
      "num_tokens": 737088422.0,
      "step": 23096
    },
    {
      "entropy": 0.4641938703134656,
      "epoch": 2.125036658788621,
      "grad_norm": 0.1531856805086136,
      "learning_rate": 2.9168583433005188e-05,
      "loss": 0.4452,
      "mean_token_accuracy": 0.8588254787027836,
      "num_tokens": 737120773.0,
      "step": 23097
    },
    {
      "entropy": 0.5720295757055283,
      "epoch": 2.1251286651600623,
      "grad_norm": 0.16267378628253937,
      "learning_rate": 2.9165516606863556e-05,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8233392983675003,
      "num_tokens": 737152702.0,
      "step": 23098
    },
    {
      "entropy": 0.43390274327248335,
      "epoch": 2.1252206715315034,
      "grad_norm": 0.15816546976566315,
      "learning_rate": 2.9162449780721934e-05,
      "loss": 0.417,
      "mean_token_accuracy": 0.8669266477227211,
      "num_tokens": 737184023.0,
      "step": 23099
    },
    {
      "entropy": 0.5644577518105507,
      "epoch": 2.125312677902945,
      "grad_norm": 0.16323687136173248,
      "learning_rate": 2.9159382954580305e-05,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8220818527042866,
      "num_tokens": 737216016.0,
      "step": 23100
    },
    {
      "entropy": 0.4948131060227752,
      "epoch": 2.125404684274386,
      "grad_norm": 0.15574190020561218,
      "learning_rate": 2.9156316128438683e-05,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8479960523545742,
      "num_tokens": 737248305.0,
      "step": 23101
    },
    {
      "entropy": 0.5450270287692547,
      "epoch": 2.1254966906458272,
      "grad_norm": 0.1615413874387741,
      "learning_rate": 2.9153249302297055e-05,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8283499218523502,
      "num_tokens": 737280188.0,
      "step": 23102
    },
    {
      "entropy": 0.5351279526948929,
      "epoch": 2.1255886970172684,
      "grad_norm": 0.16119438409805298,
      "learning_rate": 2.915018247615543e-05,
      "loss": 0.5235,
      "mean_token_accuracy": 0.8337506279349327,
      "num_tokens": 737312459.0,
      "step": 23103
    },
    {
      "entropy": 0.4007454114034772,
      "epoch": 2.1256807033887095,
      "grad_norm": 0.14752259850502014,
      "learning_rate": 2.91471156500138e-05,
      "loss": 0.3859,
      "mean_token_accuracy": 0.8753414265811443,
      "num_tokens": 737343898.0,
      "step": 23104
    },
    {
      "entropy": 0.4816243126988411,
      "epoch": 2.125772709760151,
      "grad_norm": 0.15495988726615906,
      "learning_rate": 2.914404882387218e-05,
      "loss": 0.4748,
      "mean_token_accuracy": 0.8492911420762539,
      "num_tokens": 737376224.0,
      "step": 23105
    },
    {
      "entropy": 0.5099633447825909,
      "epoch": 2.1258647161315922,
      "grad_norm": 0.15908849239349365,
      "learning_rate": 2.914098199773055e-05,
      "loss": 0.4962,
      "mean_token_accuracy": 0.8420438282191753,
      "num_tokens": 737408649.0,
      "step": 23106
    },
    {
      "entropy": 0.5679714512079954,
      "epoch": 2.1259567225030334,
      "grad_norm": 0.16722573339939117,
      "learning_rate": 2.9137915171588924e-05,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8278626017272472,
      "num_tokens": 737440189.0,
      "step": 23107
    },
    {
      "entropy": 0.417009299620986,
      "epoch": 2.1260487288744745,
      "grad_norm": 0.14563804864883423,
      "learning_rate": 2.9134848345447296e-05,
      "loss": 0.4018,
      "mean_token_accuracy": 0.870477557182312,
      "num_tokens": 737472108.0,
      "step": 23108
    },
    {
      "entropy": 0.4859618442133069,
      "epoch": 2.1261407352459156,
      "grad_norm": 0.15686054527759552,
      "learning_rate": 2.9131781519305674e-05,
      "loss": 0.4842,
      "mean_token_accuracy": 0.846362516283989,
      "num_tokens": 737504395.0,
      "step": 23109
    },
    {
      "entropy": 0.4700957993045449,
      "epoch": 2.126232741617357,
      "grad_norm": 0.15648844838142395,
      "learning_rate": 2.9128714693164045e-05,
      "loss": 0.4614,
      "mean_token_accuracy": 0.855224285274744,
      "num_tokens": 737535955.0,
      "step": 23110
    },
    {
      "entropy": 0.5115641709417105,
      "epoch": 2.1263247479887983,
      "grad_norm": 0.1621590405702591,
      "learning_rate": 2.912564786702242e-05,
      "loss": 0.4898,
      "mean_token_accuracy": 0.8400807678699493,
      "num_tokens": 737567233.0,
      "step": 23111
    },
    {
      "entropy": 0.5852707838639617,
      "epoch": 2.1264167543602395,
      "grad_norm": 0.17055818438529968,
      "learning_rate": 2.912258104088079e-05,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8208506628870964,
      "num_tokens": 737599075.0,
      "step": 23112
    },
    {
      "entropy": 0.6123972125351429,
      "epoch": 2.1265087607316806,
      "grad_norm": 0.18108783662319183,
      "learning_rate": 2.911951421473917e-05,
      "loss": 0.6094,
      "mean_token_accuracy": 0.8078703358769417,
      "num_tokens": 737629993.0,
      "step": 23113
    },
    {
      "entropy": 0.5548908356577158,
      "epoch": 2.1266007671031217,
      "grad_norm": 0.162420853972435,
      "learning_rate": 2.911644738859754e-05,
      "loss": 0.5413,
      "mean_token_accuracy": 0.8273819200694561,
      "num_tokens": 737662322.0,
      "step": 23114
    },
    {
      "entropy": 0.461468830704689,
      "epoch": 2.1266927734745633,
      "grad_norm": 0.15173518657684326,
      "learning_rate": 2.9113380562455915e-05,
      "loss": 0.4456,
      "mean_token_accuracy": 0.8590953387320042,
      "num_tokens": 737694464.0,
      "step": 23115
    },
    {
      "entropy": 0.4670701767317951,
      "epoch": 2.1267847798460044,
      "grad_norm": 0.15495409071445465,
      "learning_rate": 2.9110313736314286e-05,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8538700826466084,
      "num_tokens": 737726521.0,
      "step": 23116
    },
    {
      "entropy": 0.5697073647752404,
      "epoch": 2.1268767862174456,
      "grad_norm": 0.16678769886493683,
      "learning_rate": 2.9107246910172664e-05,
      "loss": 0.5624,
      "mean_token_accuracy": 0.8234019093215466,
      "num_tokens": 737759247.0,
      "step": 23117
    },
    {
      "entropy": 0.5276269251480699,
      "epoch": 2.1269687925888867,
      "grad_norm": 0.16472765803337097,
      "learning_rate": 2.9104180084031035e-05,
      "loss": 0.508,
      "mean_token_accuracy": 0.8373891450464725,
      "num_tokens": 737791222.0,
      "step": 23118
    },
    {
      "entropy": 0.4632526496425271,
      "epoch": 2.127060798960328,
      "grad_norm": 0.15643922984600067,
      "learning_rate": 2.9101113257889413e-05,
      "loss": 0.457,
      "mean_token_accuracy": 0.8517072014510632,
      "num_tokens": 737823320.0,
      "step": 23119
    },
    {
      "entropy": 0.4257051581516862,
      "epoch": 2.1271528053317694,
      "grad_norm": 0.15044356882572174,
      "learning_rate": 2.909804643174778e-05,
      "loss": 0.4154,
      "mean_token_accuracy": 0.8681100979447365,
      "num_tokens": 737855257.0,
      "step": 23120
    },
    {
      "entropy": 0.5360816195607185,
      "epoch": 2.1272448117032106,
      "grad_norm": 0.16590771079063416,
      "learning_rate": 2.909497960560616e-05,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8376205563545227,
      "num_tokens": 737886808.0,
      "step": 23121
    },
    {
      "entropy": 0.6144385663792491,
      "epoch": 2.1273368180746517,
      "grad_norm": 0.17183901369571686,
      "learning_rate": 2.9091912779464537e-05,
      "loss": 0.5954,
      "mean_token_accuracy": 0.8122772052884102,
      "num_tokens": 737918663.0,
      "step": 23122
    },
    {
      "entropy": 0.5673610400408506,
      "epoch": 2.127428824446093,
      "grad_norm": 0.17234794795513153,
      "learning_rate": 2.908884595332291e-05,
      "loss": 0.5586,
      "mean_token_accuracy": 0.8264634013175964,
      "num_tokens": 737949393.0,
      "step": 23123
    },
    {
      "entropy": 0.6006079548969865,
      "epoch": 2.127520830817534,
      "grad_norm": 0.17384953796863556,
      "learning_rate": 2.9085779127181283e-05,
      "loss": 0.5848,
      "mean_token_accuracy": 0.8146601170301437,
      "num_tokens": 737981814.0,
      "step": 23124
    },
    {
      "entropy": 0.5457780389115214,
      "epoch": 2.1276128371889755,
      "grad_norm": 0.16443949937820435,
      "learning_rate": 2.9082712301039655e-05,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8325165025889874,
      "num_tokens": 738013545.0,
      "step": 23125
    },
    {
      "entropy": 0.5883706407621503,
      "epoch": 2.1277048435604167,
      "grad_norm": 0.17406214773654938,
      "learning_rate": 2.9079645474898033e-05,
      "loss": 0.5769,
      "mean_token_accuracy": 0.8180698491632938,
      "num_tokens": 738046125.0,
      "step": 23126
    },
    {
      "entropy": 0.5066366693936288,
      "epoch": 2.127796849931858,
      "grad_norm": 0.15535862743854523,
      "learning_rate": 2.9076578648756404e-05,
      "loss": 0.4942,
      "mean_token_accuracy": 0.84272351115942,
      "num_tokens": 738078555.0,
      "step": 23127
    },
    {
      "entropy": 0.46936449874192476,
      "epoch": 2.127888856303299,
      "grad_norm": 0.1558331698179245,
      "learning_rate": 2.907351182261478e-05,
      "loss": 0.4591,
      "mean_token_accuracy": 0.8534615524113178,
      "num_tokens": 738110171.0,
      "step": 23128
    },
    {
      "entropy": 0.5500239348039031,
      "epoch": 2.12798086267474,
      "grad_norm": 0.1640016734600067,
      "learning_rate": 2.907044499647315e-05,
      "loss": 0.519,
      "mean_token_accuracy": 0.8338794559240341,
      "num_tokens": 738141023.0,
      "step": 23129
    },
    {
      "entropy": 0.4671214260160923,
      "epoch": 2.1280728690461816,
      "grad_norm": 0.15607090294361115,
      "learning_rate": 2.9067378170331528e-05,
      "loss": 0.4576,
      "mean_token_accuracy": 0.856320284307003,
      "num_tokens": 738173317.0,
      "step": 23130
    },
    {
      "entropy": 0.5357344025978819,
      "epoch": 2.128164875417623,
      "grad_norm": 0.15860378742218018,
      "learning_rate": 2.90643113441899e-05,
      "loss": 0.5231,
      "mean_token_accuracy": 0.8363448530435562,
      "num_tokens": 738205360.0,
      "step": 23131
    },
    {
      "entropy": 0.4696626076474786,
      "epoch": 2.128256881789064,
      "grad_norm": 0.16033115983009338,
      "learning_rate": 2.9061244518048274e-05,
      "loss": 0.4556,
      "mean_token_accuracy": 0.8551836088299751,
      "num_tokens": 738236686.0,
      "step": 23132
    },
    {
      "entropy": 0.4228491666726768,
      "epoch": 2.128348888160505,
      "grad_norm": 0.1486940085887909,
      "learning_rate": 2.9058177691906645e-05,
      "loss": 0.4023,
      "mean_token_accuracy": 0.8705040290951729,
      "num_tokens": 738267771.0,
      "step": 23133
    },
    {
      "entropy": 0.49164372542873025,
      "epoch": 2.128440894531946,
      "grad_norm": 0.16075021028518677,
      "learning_rate": 2.9055110865765023e-05,
      "loss": 0.484,
      "mean_token_accuracy": 0.8475723043084145,
      "num_tokens": 738299626.0,
      "step": 23134
    },
    {
      "entropy": 0.4114604906644672,
      "epoch": 2.1285329009033878,
      "grad_norm": 0.15215224027633667,
      "learning_rate": 2.9052044039623394e-05,
      "loss": 0.3885,
      "mean_token_accuracy": 0.8749358467757702,
      "num_tokens": 738331043.0,
      "step": 23135
    },
    {
      "entropy": 0.4260727509390563,
      "epoch": 2.128624907274829,
      "grad_norm": 0.1460256725549698,
      "learning_rate": 2.904897721348177e-05,
      "loss": 0.4107,
      "mean_token_accuracy": 0.8678578212857246,
      "num_tokens": 738363443.0,
      "step": 23136
    },
    {
      "entropy": 0.5576546620577574,
      "epoch": 2.12871691364627,
      "grad_norm": 0.1642039269208908,
      "learning_rate": 2.904591038734014e-05,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8275554440915585,
      "num_tokens": 738395269.0,
      "step": 23137
    },
    {
      "entropy": 0.4386437723878771,
      "epoch": 2.128808920017711,
      "grad_norm": 0.14692629873752594,
      "learning_rate": 2.9042843561198518e-05,
      "loss": 0.4274,
      "mean_token_accuracy": 0.8642529807984829,
      "num_tokens": 738427902.0,
      "step": 23138
    },
    {
      "entropy": 0.5588081683963537,
      "epoch": 2.1289009263891523,
      "grad_norm": 0.16001950204372406,
      "learning_rate": 2.903977673505689e-05,
      "loss": 0.5404,
      "mean_token_accuracy": 0.8292161673307419,
      "num_tokens": 738459708.0,
      "step": 23139
    },
    {
      "entropy": 0.5167047968134284,
      "epoch": 2.128992932760594,
      "grad_norm": 0.15764367580413818,
      "learning_rate": 2.9036709908915267e-05,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8416592292487621,
      "num_tokens": 738491943.0,
      "step": 23140
    },
    {
      "entropy": 0.5800679251551628,
      "epoch": 2.129084939132035,
      "grad_norm": 0.16254058480262756,
      "learning_rate": 2.903364308277364e-05,
      "loss": 0.5533,
      "mean_token_accuracy": 0.8254889473319054,
      "num_tokens": 738523991.0,
      "step": 23141
    },
    {
      "entropy": 0.47325395606458187,
      "epoch": 2.129176945503476,
      "grad_norm": 0.1512448489665985,
      "learning_rate": 2.9030576256632013e-05,
      "loss": 0.4575,
      "mean_token_accuracy": 0.8539554551243782,
      "num_tokens": 738556095.0,
      "step": 23142
    },
    {
      "entropy": 0.4703941931948066,
      "epoch": 2.1292689518749173,
      "grad_norm": 0.15089426934719086,
      "learning_rate": 2.9027509430490385e-05,
      "loss": 0.4475,
      "mean_token_accuracy": 0.8583319187164307,
      "num_tokens": 738587757.0,
      "step": 23143
    },
    {
      "entropy": 0.49654505029320717,
      "epoch": 2.1293609582463584,
      "grad_norm": 0.1620703786611557,
      "learning_rate": 2.9024442604348763e-05,
      "loss": 0.482,
      "mean_token_accuracy": 0.8439883477985859,
      "num_tokens": 738619814.0,
      "step": 23144
    },
    {
      "entropy": 0.5572607181966305,
      "epoch": 2.1294529646178,
      "grad_norm": 0.16903874278068542,
      "learning_rate": 2.9021375778207134e-05,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8248756900429726,
      "num_tokens": 738651569.0,
      "step": 23145
    },
    {
      "entropy": 0.4368919786065817,
      "epoch": 2.129544970989241,
      "grad_norm": 0.14948788285255432,
      "learning_rate": 2.901830895206551e-05,
      "loss": 0.4258,
      "mean_token_accuracy": 0.8613190464675426,
      "num_tokens": 738684065.0,
      "step": 23146
    },
    {
      "entropy": 0.5196599625051022,
      "epoch": 2.1296369773606822,
      "grad_norm": 0.16523049771785736,
      "learning_rate": 2.901524212592388e-05,
      "loss": 0.5135,
      "mean_token_accuracy": 0.8363409861922264,
      "num_tokens": 738715141.0,
      "step": 23147
    },
    {
      "entropy": 0.5173835977911949,
      "epoch": 2.1297289837321234,
      "grad_norm": 0.16550639271736145,
      "learning_rate": 2.9012175299782258e-05,
      "loss": 0.5048,
      "mean_token_accuracy": 0.8344487175345421,
      "num_tokens": 738746587.0,
      "step": 23148
    },
    {
      "entropy": 0.5505094500258565,
      "epoch": 2.1298209901035645,
      "grad_norm": 0.16703864932060242,
      "learning_rate": 2.900910847364063e-05,
      "loss": 0.544,
      "mean_token_accuracy": 0.8263427726924419,
      "num_tokens": 738777375.0,
      "step": 23149
    },
    {
      "entropy": 0.5032160617411137,
      "epoch": 2.129912996475006,
      "grad_norm": 0.16175596415996552,
      "learning_rate": 2.9006041647499004e-05,
      "loss": 0.4984,
      "mean_token_accuracy": 0.843195766210556,
      "num_tokens": 738809221.0,
      "step": 23150
    },
    {
      "entropy": 0.6610609386116266,
      "epoch": 2.1300050028464472,
      "grad_norm": 0.17484533786773682,
      "learning_rate": 2.9002974821357375e-05,
      "loss": 0.6403,
      "mean_token_accuracy": 0.7942878939211369,
      "num_tokens": 738840841.0,
      "step": 23151
    },
    {
      "entropy": 0.5434441491961479,
      "epoch": 2.1300970092178884,
      "grad_norm": 0.16303785145282745,
      "learning_rate": 2.8999907995215753e-05,
      "loss": 0.5328,
      "mean_token_accuracy": 0.8287223353981972,
      "num_tokens": 738872846.0,
      "step": 23152
    },
    {
      "entropy": 0.4490280244499445,
      "epoch": 2.1301890155893295,
      "grad_norm": 0.15160413086414337,
      "learning_rate": 2.8996841169074128e-05,
      "loss": 0.4351,
      "mean_token_accuracy": 0.860288493335247,
      "num_tokens": 738905246.0,
      "step": 23153
    },
    {
      "entropy": 0.5713120442815125,
      "epoch": 2.1302810219607706,
      "grad_norm": 0.16368620097637177,
      "learning_rate": 2.89937743429325e-05,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8260593824088573,
      "num_tokens": 738937074.0,
      "step": 23154
    },
    {
      "entropy": 0.5363787207752466,
      "epoch": 2.130373028332212,
      "grad_norm": 0.1629754900932312,
      "learning_rate": 2.8990707516790877e-05,
      "loss": 0.5278,
      "mean_token_accuracy": 0.8309270963072777,
      "num_tokens": 738969381.0,
      "step": 23155
    },
    {
      "entropy": 0.5286064576357603,
      "epoch": 2.1304650347036533,
      "grad_norm": 0.16875731945037842,
      "learning_rate": 2.8987640690649248e-05,
      "loss": 0.515,
      "mean_token_accuracy": 0.8344334624707699,
      "num_tokens": 739001282.0,
      "step": 23156
    },
    {
      "entropy": 0.5137506145983934,
      "epoch": 2.1305570410750945,
      "grad_norm": 0.16652503609657288,
      "learning_rate": 2.8984573864507626e-05,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8417786434292793,
      "num_tokens": 739033079.0,
      "step": 23157
    },
    {
      "entropy": 0.5711315595544875,
      "epoch": 2.1306490474465356,
      "grad_norm": 0.16490687429904938,
      "learning_rate": 2.8981507038365994e-05,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8295965641736984,
      "num_tokens": 739064272.0,
      "step": 23158
    },
    {
      "entropy": 0.43296930007636547,
      "epoch": 2.1307410538179767,
      "grad_norm": 0.15052349865436554,
      "learning_rate": 2.8978440212224372e-05,
      "loss": 0.4248,
      "mean_token_accuracy": 0.8611982651054859,
      "num_tokens": 739095817.0,
      "step": 23159
    },
    {
      "entropy": 0.4022513744421303,
      "epoch": 2.1308330601894183,
      "grad_norm": 0.14580044150352478,
      "learning_rate": 2.8975373386082744e-05,
      "loss": 0.3982,
      "mean_token_accuracy": 0.8738651685416698,
      "num_tokens": 739128438.0,
      "step": 23160
    },
    {
      "entropy": 0.5895615480840206,
      "epoch": 2.1309250665608594,
      "grad_norm": 0.16335397958755493,
      "learning_rate": 2.897230655994112e-05,
      "loss": 0.5803,
      "mean_token_accuracy": 0.8163599036633968,
      "num_tokens": 739161071.0,
      "step": 23161
    },
    {
      "entropy": 0.5134705379605293,
      "epoch": 2.1310170729323006,
      "grad_norm": 0.1526172012090683,
      "learning_rate": 2.8969239733799493e-05,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8435915261507034,
      "num_tokens": 739192983.0,
      "step": 23162
    },
    {
      "entropy": 0.4947184547781944,
      "epoch": 2.1311090793037417,
      "grad_norm": 0.15321196615695953,
      "learning_rate": 2.8966172907657867e-05,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8482840731739998,
      "num_tokens": 739225751.0,
      "step": 23163
    },
    {
      "entropy": 0.4423408233560622,
      "epoch": 2.131201085675183,
      "grad_norm": 0.14929381012916565,
      "learning_rate": 2.896310608151624e-05,
      "loss": 0.4296,
      "mean_token_accuracy": 0.8610876463353634,
      "num_tokens": 739257729.0,
      "step": 23164
    },
    {
      "entropy": 0.5381987392902374,
      "epoch": 2.1312930920466244,
      "grad_norm": 0.16023926436901093,
      "learning_rate": 2.8960039255374617e-05,
      "loss": 0.5189,
      "mean_token_accuracy": 0.831583708524704,
      "num_tokens": 739289393.0,
      "step": 23165
    },
    {
      "entropy": 0.5283613502979279,
      "epoch": 2.1313850984180656,
      "grad_norm": 0.16430504620075226,
      "learning_rate": 2.8956972429232988e-05,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8365363366901875,
      "num_tokens": 739321529.0,
      "step": 23166
    },
    {
      "entropy": 0.5268253767862916,
      "epoch": 2.1314771047895067,
      "grad_norm": 0.15846969187259674,
      "learning_rate": 2.8953905603091363e-05,
      "loss": 0.5123,
      "mean_token_accuracy": 0.8371001109480858,
      "num_tokens": 739353892.0,
      "step": 23167
    },
    {
      "entropy": 0.526473255828023,
      "epoch": 2.131569111160948,
      "grad_norm": 0.1613895148038864,
      "learning_rate": 2.8950838776949734e-05,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8390063717961311,
      "num_tokens": 739386087.0,
      "step": 23168
    },
    {
      "entropy": 0.5362140005454421,
      "epoch": 2.131661117532389,
      "grad_norm": 0.1617787480354309,
      "learning_rate": 2.8947771950808112e-05,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8336824551224709,
      "num_tokens": 739417833.0,
      "step": 23169
    },
    {
      "entropy": 0.5249716220423579,
      "epoch": 2.1317531239038305,
      "grad_norm": 0.15722611546516418,
      "learning_rate": 2.8944705124666483e-05,
      "loss": 0.5107,
      "mean_token_accuracy": 0.8420993573963642,
      "num_tokens": 739450036.0,
      "step": 23170
    },
    {
      "entropy": 0.6250056941062212,
      "epoch": 2.1318451302752717,
      "grad_norm": 0.17544522881507874,
      "learning_rate": 2.8941638298524858e-05,
      "loss": 0.6104,
      "mean_token_accuracy": 0.8070855960249901,
      "num_tokens": 739481998.0,
      "step": 23171
    },
    {
      "entropy": 0.5486632075626403,
      "epoch": 2.131937136646713,
      "grad_norm": 0.16782556474208832,
      "learning_rate": 2.893857147238323e-05,
      "loss": 0.5389,
      "mean_token_accuracy": 0.8296020179986954,
      "num_tokens": 739513342.0,
      "step": 23172
    },
    {
      "entropy": 0.5731514291837811,
      "epoch": 2.132029143018154,
      "grad_norm": 0.16825813055038452,
      "learning_rate": 2.8935504646241607e-05,
      "loss": 0.5495,
      "mean_token_accuracy": 0.8253211788833141,
      "num_tokens": 739544968.0,
      "step": 23173
    },
    {
      "entropy": 0.5253212852403522,
      "epoch": 2.132121149389595,
      "grad_norm": 0.1652776449918747,
      "learning_rate": 2.893243782009998e-05,
      "loss": 0.509,
      "mean_token_accuracy": 0.8395487442612648,
      "num_tokens": 739577730.0,
      "step": 23174
    },
    {
      "entropy": 0.44631416350603104,
      "epoch": 2.1322131557610366,
      "grad_norm": 0.1519445925951004,
      "learning_rate": 2.8929370993958353e-05,
      "loss": 0.4261,
      "mean_token_accuracy": 0.863766610622406,
      "num_tokens": 739609480.0,
      "step": 23175
    },
    {
      "entropy": 0.6046900518704206,
      "epoch": 2.132305162132478,
      "grad_norm": 0.17063267529010773,
      "learning_rate": 2.8926304167816724e-05,
      "loss": 0.5926,
      "mean_token_accuracy": 0.8144297450780869,
      "num_tokens": 739640759.0,
      "step": 23176
    },
    {
      "entropy": 0.4842031029984355,
      "epoch": 2.132397168503919,
      "grad_norm": 0.15509559214115143,
      "learning_rate": 2.8923237341675102e-05,
      "loss": 0.468,
      "mean_token_accuracy": 0.8465232625603676,
      "num_tokens": 739672664.0,
      "step": 23177
    },
    {
      "entropy": 0.6052957647480071,
      "epoch": 2.13248917487536,
      "grad_norm": 0.16966640949249268,
      "learning_rate": 2.8920170515533474e-05,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8154178149998188,
      "num_tokens": 739704173.0,
      "step": 23178
    },
    {
      "entropy": 0.4889453402720392,
      "epoch": 2.132581181246801,
      "grad_norm": 0.15350544452667236,
      "learning_rate": 2.891710368939185e-05,
      "loss": 0.4584,
      "mean_token_accuracy": 0.8544853180646896,
      "num_tokens": 739736108.0,
      "step": 23179
    },
    {
      "entropy": 0.46519127674400806,
      "epoch": 2.1326731876182428,
      "grad_norm": 0.14687462151050568,
      "learning_rate": 2.891403686325022e-05,
      "loss": 0.447,
      "mean_token_accuracy": 0.8555873408913612,
      "num_tokens": 739767609.0,
      "step": 23180
    },
    {
      "entropy": 0.46057953406125307,
      "epoch": 2.132765193989684,
      "grad_norm": 0.1510486602783203,
      "learning_rate": 2.8910970037108598e-05,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8559032008051872,
      "num_tokens": 739799977.0,
      "step": 23181
    },
    {
      "entropy": 0.4472826663404703,
      "epoch": 2.132857200361125,
      "grad_norm": 0.15119723975658417,
      "learning_rate": 2.890790321096697e-05,
      "loss": 0.4325,
      "mean_token_accuracy": 0.8628150634467602,
      "num_tokens": 739831421.0,
      "step": 23182
    },
    {
      "entropy": 0.4327140599489212,
      "epoch": 2.132949206732566,
      "grad_norm": 0.15376123785972595,
      "learning_rate": 2.8904836384825347e-05,
      "loss": 0.4281,
      "mean_token_accuracy": 0.8616142980754375,
      "num_tokens": 739863756.0,
      "step": 23183
    },
    {
      "entropy": 0.589990733191371,
      "epoch": 2.1330412131040073,
      "grad_norm": 0.16550227999687195,
      "learning_rate": 2.890176955868372e-05,
      "loss": 0.5777,
      "mean_token_accuracy": 0.8198592588305473,
      "num_tokens": 739896413.0,
      "step": 23184
    },
    {
      "entropy": 0.4660560507327318,
      "epoch": 2.133133219475449,
      "grad_norm": 0.15531937777996063,
      "learning_rate": 2.8898702732542093e-05,
      "loss": 0.4597,
      "mean_token_accuracy": 0.853807058185339,
      "num_tokens": 739928614.0,
      "step": 23185
    },
    {
      "entropy": 0.567310840357095,
      "epoch": 2.13322522584689,
      "grad_norm": 0.1682511568069458,
      "learning_rate": 2.889563590640047e-05,
      "loss": 0.5601,
      "mean_token_accuracy": 0.8246331475675106,
      "num_tokens": 739960761.0,
      "step": 23186
    },
    {
      "entropy": 0.583368961699307,
      "epoch": 2.133317232218331,
      "grad_norm": 0.1722795069217682,
      "learning_rate": 2.8892569080258842e-05,
      "loss": 0.5843,
      "mean_token_accuracy": 0.8170128837227821,
      "num_tokens": 739992709.0,
      "step": 23187
    },
    {
      "entropy": 0.6225786479189992,
      "epoch": 2.1334092385897723,
      "grad_norm": 0.17701001465320587,
      "learning_rate": 2.8889502254117217e-05,
      "loss": 0.6143,
      "mean_token_accuracy": 0.806568305939436,
      "num_tokens": 740024113.0,
      "step": 23188
    },
    {
      "entropy": 0.4883500263094902,
      "epoch": 2.1335012449612134,
      "grad_norm": 0.156542107462883,
      "learning_rate": 2.8886435427975588e-05,
      "loss": 0.4819,
      "mean_token_accuracy": 0.8496495336294174,
      "num_tokens": 740055920.0,
      "step": 23189
    },
    {
      "entropy": 0.4832587242126465,
      "epoch": 2.133593251332655,
      "grad_norm": 0.158723384141922,
      "learning_rate": 2.8883368601833966e-05,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8502615317702293,
      "num_tokens": 740087752.0,
      "step": 23190
    },
    {
      "entropy": 0.549397811293602,
      "epoch": 2.133685257704096,
      "grad_norm": 0.16384625434875488,
      "learning_rate": 2.8880301775692337e-05,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8315052948892117,
      "num_tokens": 740119922.0,
      "step": 23191
    },
    {
      "entropy": 0.5496840837877244,
      "epoch": 2.1337772640755372,
      "grad_norm": 0.160543754696846,
      "learning_rate": 2.8877234949550712e-05,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8341890536248684,
      "num_tokens": 740151892.0,
      "step": 23192
    },
    {
      "entropy": 0.48938314989209175,
      "epoch": 2.1338692704469784,
      "grad_norm": 0.15324541926383972,
      "learning_rate": 2.8874168123409083e-05,
      "loss": 0.4769,
      "mean_token_accuracy": 0.8496574088931084,
      "num_tokens": 740184634.0,
      "step": 23193
    },
    {
      "entropy": 0.5195277072489262,
      "epoch": 2.1339612768184195,
      "grad_norm": 0.16720791161060333,
      "learning_rate": 2.887110129726746e-05,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8388712517917156,
      "num_tokens": 740216133.0,
      "step": 23194
    },
    {
      "entropy": 0.500020429957658,
      "epoch": 2.134053283189861,
      "grad_norm": 0.15413036942481995,
      "learning_rate": 2.8868034471125832e-05,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8454076573252678,
      "num_tokens": 740248755.0,
      "step": 23195
    },
    {
      "entropy": 0.4318434458691627,
      "epoch": 2.1341452895613022,
      "grad_norm": 0.14866523444652557,
      "learning_rate": 2.886496764498421e-05,
      "loss": 0.4212,
      "mean_token_accuracy": 0.865799468010664,
      "num_tokens": 740280589.0,
      "step": 23196
    },
    {
      "entropy": 0.5985396867617965,
      "epoch": 2.1342372959327434,
      "grad_norm": 0.16553889214992523,
      "learning_rate": 2.886190081884258e-05,
      "loss": 0.5654,
      "mean_token_accuracy": 0.8204158619046211,
      "num_tokens": 740312716.0,
      "step": 23197
    },
    {
      "entropy": 0.4914917969144881,
      "epoch": 2.1343293023041845,
      "grad_norm": 0.15733495354652405,
      "learning_rate": 2.8858833992700956e-05,
      "loss": 0.4764,
      "mean_token_accuracy": 0.8493375889956951,
      "num_tokens": 740344554.0,
      "step": 23198
    },
    {
      "entropy": 0.5315243313089013,
      "epoch": 2.1344213086756256,
      "grad_norm": 0.156772181391716,
      "learning_rate": 2.8855767166559328e-05,
      "loss": 0.5184,
      "mean_token_accuracy": 0.8355494663119316,
      "num_tokens": 740376152.0,
      "step": 23199
    },
    {
      "entropy": 0.5734927207231522,
      "epoch": 2.134513315047067,
      "grad_norm": 0.1613045483827591,
      "learning_rate": 2.8852700340417706e-05,
      "loss": 0.5546,
      "mean_token_accuracy": 0.8231820166110992,
      "num_tokens": 740408194.0,
      "step": 23200
    },
    {
      "entropy": 0.5772030688822269,
      "epoch": 2.1346053214185083,
      "grad_norm": 0.1643086075782776,
      "learning_rate": 2.8849633514276077e-05,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8243639729917049,
      "num_tokens": 740439537.0,
      "step": 23201
    },
    {
      "entropy": 0.5171600617468357,
      "epoch": 2.1346973277899495,
      "grad_norm": 0.16526181995868683,
      "learning_rate": 2.884656668813445e-05,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8382229544222355,
      "num_tokens": 740471206.0,
      "step": 23202
    },
    {
      "entropy": 0.4801423577591777,
      "epoch": 2.1347893341613906,
      "grad_norm": 0.15764547884464264,
      "learning_rate": 2.8843499861992823e-05,
      "loss": 0.4668,
      "mean_token_accuracy": 0.8504354357719421,
      "num_tokens": 740502653.0,
      "step": 23203
    },
    {
      "entropy": 0.5387161895632744,
      "epoch": 2.1348813405328317,
      "grad_norm": 0.16505186259746552,
      "learning_rate": 2.88404330358512e-05,
      "loss": 0.5263,
      "mean_token_accuracy": 0.8340415246784687,
      "num_tokens": 740534556.0,
      "step": 23204
    },
    {
      "entropy": 0.5531846140511334,
      "epoch": 2.1349733469042733,
      "grad_norm": 0.16499854624271393,
      "learning_rate": 2.8837366209709572e-05,
      "loss": 0.5427,
      "mean_token_accuracy": 0.8282453045248985,
      "num_tokens": 740566467.0,
      "step": 23205
    },
    {
      "entropy": 0.5391591317020357,
      "epoch": 2.1350653532757145,
      "grad_norm": 0.16306526958942413,
      "learning_rate": 2.8834299383567947e-05,
      "loss": 0.5267,
      "mean_token_accuracy": 0.835502877831459,
      "num_tokens": 740598101.0,
      "step": 23206
    },
    {
      "entropy": 0.4807076705619693,
      "epoch": 2.1351573596471556,
      "grad_norm": 0.1563183218240738,
      "learning_rate": 2.8831232557426318e-05,
      "loss": 0.461,
      "mean_token_accuracy": 0.8508006483316422,
      "num_tokens": 740630182.0,
      "step": 23207
    },
    {
      "entropy": 0.46794455824419856,
      "epoch": 2.1352493660185967,
      "grad_norm": 0.15438178181648254,
      "learning_rate": 2.8828165731284696e-05,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8543593697249889,
      "num_tokens": 740662581.0,
      "step": 23208
    },
    {
      "entropy": 0.4927364466711879,
      "epoch": 2.135341372390038,
      "grad_norm": 0.15540839731693268,
      "learning_rate": 2.8825098905143067e-05,
      "loss": 0.48,
      "mean_token_accuracy": 0.8463354893028736,
      "num_tokens": 740694882.0,
      "step": 23209
    },
    {
      "entropy": 0.5660964213311672,
      "epoch": 2.1354333787614794,
      "grad_norm": 0.16723893582820892,
      "learning_rate": 2.8822032079001442e-05,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8292684033513069,
      "num_tokens": 740726165.0,
      "step": 23210
    },
    {
      "entropy": 0.4682519156485796,
      "epoch": 2.1355253851329206,
      "grad_norm": 0.1580444872379303,
      "learning_rate": 2.8818965252859813e-05,
      "loss": 0.456,
      "mean_token_accuracy": 0.8585297651588917,
      "num_tokens": 740757369.0,
      "step": 23211
    },
    {
      "entropy": 0.5595266241580248,
      "epoch": 2.1356173915043617,
      "grad_norm": 0.16422030329704285,
      "learning_rate": 2.881589842671819e-05,
      "loss": 0.5419,
      "mean_token_accuracy": 0.8260028958320618,
      "num_tokens": 740788966.0,
      "step": 23212
    },
    {
      "entropy": 0.5338604832068086,
      "epoch": 2.135709397875803,
      "grad_norm": 0.15707772970199585,
      "learning_rate": 2.8812831600576563e-05,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8373957984149456,
      "num_tokens": 740821646.0,
      "step": 23213
    },
    {
      "entropy": 0.523818401619792,
      "epoch": 2.135801404247244,
      "grad_norm": 0.16297519207000732,
      "learning_rate": 2.8809764774434937e-05,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8394963629543781,
      "num_tokens": 740853177.0,
      "step": 23214
    },
    {
      "entropy": 0.40377596858888865,
      "epoch": 2.1358934106186855,
      "grad_norm": 0.14813987910747528,
      "learning_rate": 2.8806697948293315e-05,
      "loss": 0.4005,
      "mean_token_accuracy": 0.8711029849946499,
      "num_tokens": 740885291.0,
      "step": 23215
    },
    {
      "entropy": 0.5641043856739998,
      "epoch": 2.1359854169901267,
      "grad_norm": 0.17290358245372772,
      "learning_rate": 2.8803631122151687e-05,
      "loss": 0.5478,
      "mean_token_accuracy": 0.827203132212162,
      "num_tokens": 740916586.0,
      "step": 23216
    },
    {
      "entropy": 0.5986371105536819,
      "epoch": 2.136077423361568,
      "grad_norm": 0.17283202707767487,
      "learning_rate": 2.8800564296010065e-05,
      "loss": 0.5808,
      "mean_token_accuracy": 0.8170713223516941,
      "num_tokens": 740948114.0,
      "step": 23217
    },
    {
      "entropy": 0.520323564298451,
      "epoch": 2.136169429733009,
      "grad_norm": 0.1704334318637848,
      "learning_rate": 2.8797497469868436e-05,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8432340174913406,
      "num_tokens": 740977528.0,
      "step": 23218
    },
    {
      "entropy": 0.5112226372584701,
      "epoch": 2.13626143610445,
      "grad_norm": 0.15847019851207733,
      "learning_rate": 2.879443064372681e-05,
      "loss": 0.5028,
      "mean_token_accuracy": 0.842116504907608,
      "num_tokens": 741010296.0,
      "step": 23219
    },
    {
      "entropy": 0.48033329914323986,
      "epoch": 2.1363534424758917,
      "grad_norm": 0.15931715071201324,
      "learning_rate": 2.8791363817585182e-05,
      "loss": 0.4602,
      "mean_token_accuracy": 0.851326622068882,
      "num_tokens": 741041171.0,
      "step": 23220
    },
    {
      "entropy": 0.4566344153136015,
      "epoch": 2.136445448847333,
      "grad_norm": 0.1551692932844162,
      "learning_rate": 2.878829699144356e-05,
      "loss": 0.4364,
      "mean_token_accuracy": 0.8578681126236916,
      "num_tokens": 741072945.0,
      "step": 23221
    },
    {
      "entropy": 0.467099504545331,
      "epoch": 2.136537455218774,
      "grad_norm": 0.15437017381191254,
      "learning_rate": 2.878523016530193e-05,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8547022268176079,
      "num_tokens": 741104359.0,
      "step": 23222
    },
    {
      "entropy": 0.5482999477535486,
      "epoch": 2.136629461590215,
      "grad_norm": 0.16454140841960907,
      "learning_rate": 2.8782163339160306e-05,
      "loss": 0.5276,
      "mean_token_accuracy": 0.8288729265332222,
      "num_tokens": 741136061.0,
      "step": 23223
    },
    {
      "entropy": 0.537886512465775,
      "epoch": 2.136721467961656,
      "grad_norm": 0.16004058718681335,
      "learning_rate": 2.8779096513018677e-05,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8349657729268074,
      "num_tokens": 741168559.0,
      "step": 23224
    },
    {
      "entropy": 0.47341970819979906,
      "epoch": 2.1368134743330978,
      "grad_norm": 0.14901894330978394,
      "learning_rate": 2.8776029686877055e-05,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8543697521090508,
      "num_tokens": 741200754.0,
      "step": 23225
    },
    {
      "entropy": 0.4845702173188329,
      "epoch": 2.136905480704539,
      "grad_norm": 0.1526861935853958,
      "learning_rate": 2.8772962860735426e-05,
      "loss": 0.4805,
      "mean_token_accuracy": 0.8477672375738621,
      "num_tokens": 741232813.0,
      "step": 23226
    },
    {
      "entropy": 0.6012881500646472,
      "epoch": 2.13699748707598,
      "grad_norm": 0.17308960855007172,
      "learning_rate": 2.87698960345938e-05,
      "loss": 0.5931,
      "mean_token_accuracy": 0.8141300305724144,
      "num_tokens": 741264227.0,
      "step": 23227
    },
    {
      "entropy": 0.5065867304801941,
      "epoch": 2.137089493447421,
      "grad_norm": 0.15626350045204163,
      "learning_rate": 2.8766829208452172e-05,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8428983762860298,
      "num_tokens": 741296292.0,
      "step": 23228
    },
    {
      "entropy": 0.5134672466665506,
      "epoch": 2.1371814998188623,
      "grad_norm": 0.1595616638660431,
      "learning_rate": 2.876376238231055e-05,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8378497809171677,
      "num_tokens": 741327929.0,
      "step": 23229
    },
    {
      "entropy": 0.5295632937923074,
      "epoch": 2.137273506190304,
      "grad_norm": 0.16215066611766815,
      "learning_rate": 2.876069555616892e-05,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8376821391284466,
      "num_tokens": 741359512.0,
      "step": 23230
    },
    {
      "entropy": 0.5927592236548662,
      "epoch": 2.137365512561745,
      "grad_norm": 0.16807647049427032,
      "learning_rate": 2.8757628730027296e-05,
      "loss": 0.5808,
      "mean_token_accuracy": 0.818610992282629,
      "num_tokens": 741391621.0,
      "step": 23231
    },
    {
      "entropy": 0.46619167644530535,
      "epoch": 2.137457518933186,
      "grad_norm": 0.15860453248023987,
      "learning_rate": 2.8754561903885667e-05,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8558145277202129,
      "num_tokens": 741422993.0,
      "step": 23232
    },
    {
      "entropy": 0.43463163427077234,
      "epoch": 2.1375495253046273,
      "grad_norm": 0.14907939732074738,
      "learning_rate": 2.8751495077744045e-05,
      "loss": 0.4196,
      "mean_token_accuracy": 0.8638796918094158,
      "num_tokens": 741454382.0,
      "step": 23233
    },
    {
      "entropy": 0.46960093826055527,
      "epoch": 2.1376415316760684,
      "grad_norm": 0.16005456447601318,
      "learning_rate": 2.8748428251602417e-05,
      "loss": 0.459,
      "mean_token_accuracy": 0.8527140468358994,
      "num_tokens": 741486028.0,
      "step": 23234
    },
    {
      "entropy": 0.4237500783056021,
      "epoch": 2.13773353804751,
      "grad_norm": 0.15367291867733002,
      "learning_rate": 2.874536142546079e-05,
      "loss": 0.4135,
      "mean_token_accuracy": 0.8664670847356319,
      "num_tokens": 741517808.0,
      "step": 23235
    },
    {
      "entropy": 0.5791432689875364,
      "epoch": 2.137825544418951,
      "grad_norm": 0.16820672154426575,
      "learning_rate": 2.8742294599319163e-05,
      "loss": 0.5538,
      "mean_token_accuracy": 0.8205828070640564,
      "num_tokens": 741549546.0,
      "step": 23236
    },
    {
      "entropy": 0.5151622872799635,
      "epoch": 2.1379175507903923,
      "grad_norm": 0.16330954432487488,
      "learning_rate": 2.873922777317754e-05,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8417632915079594,
      "num_tokens": 741580343.0,
      "step": 23237
    },
    {
      "entropy": 0.5107581140473485,
      "epoch": 2.1380095571618334,
      "grad_norm": 0.16906726360321045,
      "learning_rate": 2.8736160947035912e-05,
      "loss": 0.5008,
      "mean_token_accuracy": 0.8387046530842781,
      "num_tokens": 741612430.0,
      "step": 23238
    },
    {
      "entropy": 0.5947691351175308,
      "epoch": 2.1381015635332745,
      "grad_norm": 0.17352358996868134,
      "learning_rate": 2.873309412089429e-05,
      "loss": 0.5857,
      "mean_token_accuracy": 0.8167214877903461,
      "num_tokens": 741644289.0,
      "step": 23239
    },
    {
      "entropy": 0.5836571513209492,
      "epoch": 2.138193569904716,
      "grad_norm": 0.1644209623336792,
      "learning_rate": 2.873002729475266e-05,
      "loss": 0.5747,
      "mean_token_accuracy": 0.8230337090790272,
      "num_tokens": 741677057.0,
      "step": 23240
    },
    {
      "entropy": 0.5593928750604391,
      "epoch": 2.1382855762761572,
      "grad_norm": 0.16055960953235626,
      "learning_rate": 2.8726960468611036e-05,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8260581195354462,
      "num_tokens": 741709434.0,
      "step": 23241
    },
    {
      "entropy": 0.45034509990364313,
      "epoch": 2.1383775826475984,
      "grad_norm": 0.15871365368366241,
      "learning_rate": 2.8723893642469407e-05,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8594659678637981,
      "num_tokens": 741741729.0,
      "step": 23242
    },
    {
      "entropy": 0.514242599485442,
      "epoch": 2.1384695890190395,
      "grad_norm": 0.1591237485408783,
      "learning_rate": 2.8720826816327785e-05,
      "loss": 0.4973,
      "mean_token_accuracy": 0.8409486897289753,
      "num_tokens": 741773215.0,
      "step": 23243
    },
    {
      "entropy": 0.46188972145318985,
      "epoch": 2.1385615953904806,
      "grad_norm": 0.16266068816184998,
      "learning_rate": 2.871775999018616e-05,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8578067347407341,
      "num_tokens": 741804450.0,
      "step": 23244
    },
    {
      "entropy": 0.47923386096954346,
      "epoch": 2.138653601761922,
      "grad_norm": 0.1572401076555252,
      "learning_rate": 2.871469316404453e-05,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8565536886453629,
      "num_tokens": 741836178.0,
      "step": 23245
    },
    {
      "entropy": 0.44758116267621517,
      "epoch": 2.1387456081333633,
      "grad_norm": 0.1530957669019699,
      "learning_rate": 2.871162633790291e-05,
      "loss": 0.4317,
      "mean_token_accuracy": 0.8603346273303032,
      "num_tokens": 741868657.0,
      "step": 23246
    },
    {
      "entropy": 0.4986031537409872,
      "epoch": 2.1388376145048045,
      "grad_norm": 0.15781709551811218,
      "learning_rate": 2.870855951176128e-05,
      "loss": 0.4819,
      "mean_token_accuracy": 0.8448217958211899,
      "num_tokens": 741900411.0,
      "step": 23247
    },
    {
      "entropy": 0.5853879377245903,
      "epoch": 2.1389296208762456,
      "grad_norm": 0.16947788000106812,
      "learning_rate": 2.8705492685619655e-05,
      "loss": 0.5716,
      "mean_token_accuracy": 0.8196653686463833,
      "num_tokens": 741932553.0,
      "step": 23248
    },
    {
      "entropy": 0.56903469376266,
      "epoch": 2.1390216272476867,
      "grad_norm": 0.1657293140888214,
      "learning_rate": 2.8702425859478026e-05,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8267125152051449,
      "num_tokens": 741964372.0,
      "step": 23249
    },
    {
      "entropy": 0.5952368378639221,
      "epoch": 2.1391136336191283,
      "grad_norm": 0.16940826177597046,
      "learning_rate": 2.8699359033336404e-05,
      "loss": 0.5819,
      "mean_token_accuracy": 0.8162714429199696,
      "num_tokens": 741995988.0,
      "step": 23250
    },
    {
      "entropy": 0.5920016448944807,
      "epoch": 2.1392056399905695,
      "grad_norm": 0.1701945811510086,
      "learning_rate": 2.8696292207194775e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.8183244615793228,
      "num_tokens": 742028006.0,
      "step": 23251
    },
    {
      "entropy": 0.46074866224080324,
      "epoch": 2.1392976463620106,
      "grad_norm": 0.15395094454288483,
      "learning_rate": 2.869322538105315e-05,
      "loss": 0.4588,
      "mean_token_accuracy": 0.8554525598883629,
      "num_tokens": 742060065.0,
      "step": 23252
    },
    {
      "entropy": 0.5675759334117174,
      "epoch": 2.1393896527334517,
      "grad_norm": 0.16949968039989471,
      "learning_rate": 2.869015855491152e-05,
      "loss": 0.5427,
      "mean_token_accuracy": 0.8284929096698761,
      "num_tokens": 742091074.0,
      "step": 23253
    },
    {
      "entropy": 0.5820810068398714,
      "epoch": 2.139481659104893,
      "grad_norm": 0.16943947970867157,
      "learning_rate": 2.86870917287699e-05,
      "loss": 0.5747,
      "mean_token_accuracy": 0.8166837766766548,
      "num_tokens": 742122534.0,
      "step": 23254
    },
    {
      "entropy": 0.5827467078343034,
      "epoch": 2.1395736654763344,
      "grad_norm": 0.16802114248275757,
      "learning_rate": 2.868402490262827e-05,
      "loss": 0.5768,
      "mean_token_accuracy": 0.8179222382605076,
      "num_tokens": 742155230.0,
      "step": 23255
    },
    {
      "entropy": 0.5645031007006764,
      "epoch": 2.1396656718477756,
      "grad_norm": 0.1629181206226349,
      "learning_rate": 2.868095807648665e-05,
      "loss": 0.5545,
      "mean_token_accuracy": 0.8277500905096531,
      "num_tokens": 742186974.0,
      "step": 23256
    },
    {
      "entropy": 0.46293772105127573,
      "epoch": 2.1397576782192167,
      "grad_norm": 0.15743663907051086,
      "learning_rate": 2.8677891250345017e-05,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8591130338609219,
      "num_tokens": 742219416.0,
      "step": 23257
    },
    {
      "entropy": 0.5291464782785624,
      "epoch": 2.139849684590658,
      "grad_norm": 0.16238591074943542,
      "learning_rate": 2.8674824424203395e-05,
      "loss": 0.5179,
      "mean_token_accuracy": 0.8346298523247242,
      "num_tokens": 742251479.0,
      "step": 23258
    },
    {
      "entropy": 0.5759221320040524,
      "epoch": 2.139941690962099,
      "grad_norm": 0.17114022374153137,
      "learning_rate": 2.8671757598061766e-05,
      "loss": 0.5636,
      "mean_token_accuracy": 0.8227950260043144,
      "num_tokens": 742283222.0,
      "step": 23259
    },
    {
      "entropy": 0.5366431940346956,
      "epoch": 2.1400336973335405,
      "grad_norm": 0.15765857696533203,
      "learning_rate": 2.8668690771920144e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8353404812514782,
      "num_tokens": 742315027.0,
      "step": 23260
    },
    {
      "entropy": 0.5453957365825772,
      "epoch": 2.1401257037049817,
      "grad_norm": 0.1613359898328781,
      "learning_rate": 2.8665623945778515e-05,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8319189921021461,
      "num_tokens": 742347586.0,
      "step": 23261
    },
    {
      "entropy": 0.5087412279099226,
      "epoch": 2.140217710076423,
      "grad_norm": 0.1620623916387558,
      "learning_rate": 2.866255711963689e-05,
      "loss": 0.5103,
      "mean_token_accuracy": 0.8378871530294418,
      "num_tokens": 742379861.0,
      "step": 23262
    },
    {
      "entropy": 0.5345679391175508,
      "epoch": 2.140309716447864,
      "grad_norm": 0.15883922576904297,
      "learning_rate": 2.865949029349526e-05,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8368633575737476,
      "num_tokens": 742411971.0,
      "step": 23263
    },
    {
      "entropy": 0.5494063647929579,
      "epoch": 2.140401722819305,
      "grad_norm": 0.16748729348182678,
      "learning_rate": 2.865642346735364e-05,
      "loss": 0.5437,
      "mean_token_accuracy": 0.8334147743880749,
      "num_tokens": 742443875.0,
      "step": 23264
    },
    {
      "entropy": 0.4720749566331506,
      "epoch": 2.1404937291907467,
      "grad_norm": 0.15973810851573944,
      "learning_rate": 2.865335664121201e-05,
      "loss": 0.4609,
      "mean_token_accuracy": 0.8546931110322475,
      "num_tokens": 742476247.0,
      "step": 23265
    },
    {
      "entropy": 0.5167452050372958,
      "epoch": 2.140585735562188,
      "grad_norm": 0.16120529174804688,
      "learning_rate": 2.8650289815070385e-05,
      "loss": 0.5135,
      "mean_token_accuracy": 0.8377984315156937,
      "num_tokens": 742508094.0,
      "step": 23266
    },
    {
      "entropy": 0.5320530193857849,
      "epoch": 2.140677741933629,
      "grad_norm": 0.1622486263513565,
      "learning_rate": 2.8647222988928756e-05,
      "loss": 0.516,
      "mean_token_accuracy": 0.8392092362046242,
      "num_tokens": 742539983.0,
      "step": 23267
    },
    {
      "entropy": 0.5141870500519872,
      "epoch": 2.14076974830507,
      "grad_norm": 0.1596747487783432,
      "learning_rate": 2.8644156162787134e-05,
      "loss": 0.4944,
      "mean_token_accuracy": 0.8413038067519665,
      "num_tokens": 742571656.0,
      "step": 23268
    },
    {
      "entropy": 0.5920631410554051,
      "epoch": 2.140861754676511,
      "grad_norm": 0.1637779027223587,
      "learning_rate": 2.8641089336645506e-05,
      "loss": 0.5873,
      "mean_token_accuracy": 0.8151412717998028,
      "num_tokens": 742603991.0,
      "step": 23269
    },
    {
      "entropy": 0.48255777545273304,
      "epoch": 2.1409537610479528,
      "grad_norm": 0.16028910875320435,
      "learning_rate": 2.863802251050388e-05,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8514119386672974,
      "num_tokens": 742634943.0,
      "step": 23270
    },
    {
      "entropy": 0.5863784588873386,
      "epoch": 2.141045767419394,
      "grad_norm": 0.17009790241718292,
      "learning_rate": 2.863495568436225e-05,
      "loss": 0.577,
      "mean_token_accuracy": 0.8167099244892597,
      "num_tokens": 742666013.0,
      "step": 23271
    },
    {
      "entropy": 0.5450926050543785,
      "epoch": 2.141137773790835,
      "grad_norm": 0.16416265070438385,
      "learning_rate": 2.863188885822063e-05,
      "loss": 0.5328,
      "mean_token_accuracy": 0.8315127231180668,
      "num_tokens": 742697961.0,
      "step": 23272
    },
    {
      "entropy": 0.553635070566088,
      "epoch": 2.141229780162276,
      "grad_norm": 0.15712030231952667,
      "learning_rate": 2.8628822032079e-05,
      "loss": 0.5399,
      "mean_token_accuracy": 0.8300972580909729,
      "num_tokens": 742730161.0,
      "step": 23273
    },
    {
      "entropy": 0.5575167797505856,
      "epoch": 2.1413217865337173,
      "grad_norm": 0.16424605250358582,
      "learning_rate": 2.8625755205937375e-05,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8228336907923222,
      "num_tokens": 742762412.0,
      "step": 23274
    },
    {
      "entropy": 0.5153187471441925,
      "epoch": 2.141413792905159,
      "grad_norm": 0.15988574922084808,
      "learning_rate": 2.8622688379795753e-05,
      "loss": 0.5057,
      "mean_token_accuracy": 0.8386416099965572,
      "num_tokens": 742794281.0,
      "step": 23275
    },
    {
      "entropy": 0.6047354023903608,
      "epoch": 2.1415057992766,
      "grad_norm": 0.17021475732326508,
      "learning_rate": 2.8619621553654125e-05,
      "loss": 0.5915,
      "mean_token_accuracy": 0.8187010213732719,
      "num_tokens": 742826787.0,
      "step": 23276
    },
    {
      "entropy": 0.5003835996612906,
      "epoch": 2.141597805648041,
      "grad_norm": 0.15560674667358398,
      "learning_rate": 2.8616554727512503e-05,
      "loss": 0.4819,
      "mean_token_accuracy": 0.8444064408540726,
      "num_tokens": 742859555.0,
      "step": 23277
    },
    {
      "entropy": 0.5587190575897694,
      "epoch": 2.1416898120194823,
      "grad_norm": 0.16678424179553986,
      "learning_rate": 2.8613487901370874e-05,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8287745788693428,
      "num_tokens": 742890902.0,
      "step": 23278
    },
    {
      "entropy": 0.5734157180413604,
      "epoch": 2.1417818183909234,
      "grad_norm": 0.16860991716384888,
      "learning_rate": 2.861042107522925e-05,
      "loss": 0.5677,
      "mean_token_accuracy": 0.8229945860803127,
      "num_tokens": 742922840.0,
      "step": 23279
    },
    {
      "entropy": 0.5451793801039457,
      "epoch": 2.141873824762365,
      "grad_norm": 0.17020119726657867,
      "learning_rate": 2.860735424908762e-05,
      "loss": 0.5198,
      "mean_token_accuracy": 0.8377978801727295,
      "num_tokens": 742954271.0,
      "step": 23280
    },
    {
      "entropy": 0.5003829095512629,
      "epoch": 2.141965831133806,
      "grad_norm": 0.15929915010929108,
      "learning_rate": 2.8604287422945998e-05,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8440065495669842,
      "num_tokens": 742987038.0,
      "step": 23281
    },
    {
      "entropy": 0.5803697891533375,
      "epoch": 2.1420578375052473,
      "grad_norm": 0.16934019327163696,
      "learning_rate": 2.860122059680437e-05,
      "loss": 0.5659,
      "mean_token_accuracy": 0.8227817378938198,
      "num_tokens": 743019273.0,
      "step": 23282
    },
    {
      "entropy": 0.5406418819911778,
      "epoch": 2.1421498438766884,
      "grad_norm": 0.16356626152992249,
      "learning_rate": 2.8598153770662744e-05,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8316172994673252,
      "num_tokens": 743051352.0,
      "step": 23283
    },
    {
      "entropy": 0.5123699251562357,
      "epoch": 2.1422418502481295,
      "grad_norm": 0.1644597351551056,
      "learning_rate": 2.8595086944521115e-05,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8385825119912624,
      "num_tokens": 743083016.0,
      "step": 23284
    },
    {
      "entropy": 0.5811642289627343,
      "epoch": 2.142333856619571,
      "grad_norm": 0.1752173751592636,
      "learning_rate": 2.8592020118379493e-05,
      "loss": 0.5737,
      "mean_token_accuracy": 0.8191650211811066,
      "num_tokens": 743114195.0,
      "step": 23285
    },
    {
      "entropy": 0.5778890671208501,
      "epoch": 2.1424258629910122,
      "grad_norm": 0.17176423966884613,
      "learning_rate": 2.8588953292237864e-05,
      "loss": 0.5694,
      "mean_token_accuracy": 0.8216324001550674,
      "num_tokens": 743145911.0,
      "step": 23286
    },
    {
      "entropy": 0.49432759499177337,
      "epoch": 2.1425178693624534,
      "grad_norm": 0.155733123421669,
      "learning_rate": 2.858588646609624e-05,
      "loss": 0.4821,
      "mean_token_accuracy": 0.8495807647705078,
      "num_tokens": 743177750.0,
      "step": 23287
    },
    {
      "entropy": 0.4451605365611613,
      "epoch": 2.1426098757338945,
      "grad_norm": 0.14878761768341064,
      "learning_rate": 2.858281963995461e-05,
      "loss": 0.4272,
      "mean_token_accuracy": 0.8635843209922314,
      "num_tokens": 743210259.0,
      "step": 23288
    },
    {
      "entropy": 0.42310275603085756,
      "epoch": 2.1427018821053356,
      "grad_norm": 0.14516878128051758,
      "learning_rate": 2.857975281381299e-05,
      "loss": 0.4088,
      "mean_token_accuracy": 0.867936298251152,
      "num_tokens": 743242973.0,
      "step": 23289
    },
    {
      "entropy": 0.5305727617815137,
      "epoch": 2.142793888476777,
      "grad_norm": 0.16265623271465302,
      "learning_rate": 2.857668598767136e-05,
      "loss": 0.5177,
      "mean_token_accuracy": 0.8364607989788055,
      "num_tokens": 743275571.0,
      "step": 23290
    },
    {
      "entropy": 0.5431813895702362,
      "epoch": 2.1428858948482183,
      "grad_norm": 0.16909779608249664,
      "learning_rate": 2.8573619161529734e-05,
      "loss": 0.5358,
      "mean_token_accuracy": 0.8294406868517399,
      "num_tokens": 743307035.0,
      "step": 23291
    },
    {
      "entropy": 0.5125799672678113,
      "epoch": 2.1429779012196595,
      "grad_norm": 0.16415730118751526,
      "learning_rate": 2.8570552335388106e-05,
      "loss": 0.5116,
      "mean_token_accuracy": 0.8379032388329506,
      "num_tokens": 743338838.0,
      "step": 23292
    },
    {
      "entropy": 0.5115627031773329,
      "epoch": 2.1430699075911006,
      "grad_norm": 0.1635432094335556,
      "learning_rate": 2.8567485509246484e-05,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8407481499016285,
      "num_tokens": 743369856.0,
      "step": 23293
    },
    {
      "entropy": 0.47495747928041965,
      "epoch": 2.1431619139625417,
      "grad_norm": 0.15119172632694244,
      "learning_rate": 2.8564418683104855e-05,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8518792241811752,
      "num_tokens": 743402609.0,
      "step": 23294
    },
    {
      "entropy": 0.5412805154919624,
      "epoch": 2.1432539203339833,
      "grad_norm": 0.161237895488739,
      "learning_rate": 2.856135185696323e-05,
      "loss": 0.5253,
      "mean_token_accuracy": 0.8368009515106678,
      "num_tokens": 743434918.0,
      "step": 23295
    },
    {
      "entropy": 0.5597788896411657,
      "epoch": 2.1433459267054245,
      "grad_norm": 0.16045494377613068,
      "learning_rate": 2.85582850308216e-05,
      "loss": 0.5382,
      "mean_token_accuracy": 0.8296356908977032,
      "num_tokens": 743466806.0,
      "step": 23296
    },
    {
      "entropy": 0.5967075601220131,
      "epoch": 2.1434379330768656,
      "grad_norm": 0.16621540486812592,
      "learning_rate": 2.855521820467998e-05,
      "loss": 0.5747,
      "mean_token_accuracy": 0.8176578804850578,
      "num_tokens": 743498627.0,
      "step": 23297
    },
    {
      "entropy": 0.5240523186512291,
      "epoch": 2.1435299394483067,
      "grad_norm": 0.15721893310546875,
      "learning_rate": 2.855215137853835e-05,
      "loss": 0.513,
      "mean_token_accuracy": 0.8413031622767448,
      "num_tokens": 743531008.0,
      "step": 23298
    },
    {
      "entropy": 0.48502062633633614,
      "epoch": 2.143621945819748,
      "grad_norm": 0.1515403687953949,
      "learning_rate": 2.8549084552396728e-05,
      "loss": 0.4626,
      "mean_token_accuracy": 0.8538060486316681,
      "num_tokens": 743562940.0,
      "step": 23299
    },
    {
      "entropy": 0.4267255365848541,
      "epoch": 2.1437139521911894,
      "grad_norm": 0.15103428065776825,
      "learning_rate": 2.85460177262551e-05,
      "loss": 0.4094,
      "mean_token_accuracy": 0.8711774088442326,
      "num_tokens": 743595403.0,
      "step": 23300
    },
    {
      "entropy": 0.481189395301044,
      "epoch": 2.1438059585626306,
      "grad_norm": 0.1546238511800766,
      "learning_rate": 2.8542950900113474e-05,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8516255356371403,
      "num_tokens": 743627314.0,
      "step": 23301
    },
    {
      "entropy": 0.4933141078799963,
      "epoch": 2.1438979649340717,
      "grad_norm": 0.1612999141216278,
      "learning_rate": 2.8539884073971845e-05,
      "loss": 0.4681,
      "mean_token_accuracy": 0.846443060785532,
      "num_tokens": 743659191.0,
      "step": 23302
    },
    {
      "entropy": 0.5496018771082163,
      "epoch": 2.143989971305513,
      "grad_norm": 0.16079863905906677,
      "learning_rate": 2.8536817247830223e-05,
      "loss": 0.5387,
      "mean_token_accuracy": 0.8295943774282932,
      "num_tokens": 743691159.0,
      "step": 23303
    },
    {
      "entropy": 0.4746436821296811,
      "epoch": 2.144081977676954,
      "grad_norm": 0.1548490971326828,
      "learning_rate": 2.8533750421688595e-05,
      "loss": 0.4523,
      "mean_token_accuracy": 0.856615673750639,
      "num_tokens": 743723294.0,
      "step": 23304
    },
    {
      "entropy": 0.4776165699586272,
      "epoch": 2.1441739840483955,
      "grad_norm": 0.16239579021930695,
      "learning_rate": 2.853068359554697e-05,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8469549641013145,
      "num_tokens": 743754754.0,
      "step": 23305
    },
    {
      "entropy": 0.5417794343084097,
      "epoch": 2.1442659904198367,
      "grad_norm": 0.16278070211410522,
      "learning_rate": 2.8527616769405347e-05,
      "loss": 0.5172,
      "mean_token_accuracy": 0.8346157670021057,
      "num_tokens": 743785794.0,
      "step": 23306
    },
    {
      "entropy": 0.5413365326821804,
      "epoch": 2.144357996791278,
      "grad_norm": 0.16313815116882324,
      "learning_rate": 2.852454994326372e-05,
      "loss": 0.5377,
      "mean_token_accuracy": 0.8287230283021927,
      "num_tokens": 743818143.0,
      "step": 23307
    },
    {
      "entropy": 0.5614390913397074,
      "epoch": 2.144450003162719,
      "grad_norm": 0.165363609790802,
      "learning_rate": 2.8521483117122093e-05,
      "loss": 0.5447,
      "mean_token_accuracy": 0.8283136151731014,
      "num_tokens": 743849461.0,
      "step": 23308
    },
    {
      "entropy": 0.6557016633450985,
      "epoch": 2.14454200953416,
      "grad_norm": 0.17811398208141327,
      "learning_rate": 2.8518416290980464e-05,
      "loss": 0.6461,
      "mean_token_accuracy": 0.7985451556742191,
      "num_tokens": 743880535.0,
      "step": 23309
    },
    {
      "entropy": 0.41366380546242,
      "epoch": 2.1446340159056017,
      "grad_norm": 0.15328092873096466,
      "learning_rate": 2.8515349464838842e-05,
      "loss": 0.3998,
      "mean_token_accuracy": 0.8704618625342846,
      "num_tokens": 743912914.0,
      "step": 23310
    },
    {
      "entropy": 0.4756177607923746,
      "epoch": 2.144726022277043,
      "grad_norm": 0.15497943758964539,
      "learning_rate": 2.8512282638697214e-05,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8545997031033039,
      "num_tokens": 743945150.0,
      "step": 23311
    },
    {
      "entropy": 0.5270340666174889,
      "epoch": 2.144818028648484,
      "grad_norm": 0.1667640656232834,
      "learning_rate": 2.8509215812555588e-05,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8396582342684269,
      "num_tokens": 743976727.0,
      "step": 23312
    },
    {
      "entropy": 0.5476831952109933,
      "epoch": 2.144910035019925,
      "grad_norm": 0.1658991426229477,
      "learning_rate": 2.850614898641396e-05,
      "loss": 0.5457,
      "mean_token_accuracy": 0.827313095331192,
      "num_tokens": 744009073.0,
      "step": 23313
    },
    {
      "entropy": 0.4657703754492104,
      "epoch": 2.145002041391366,
      "grad_norm": 0.16185875236988068,
      "learning_rate": 2.8503082160272338e-05,
      "loss": 0.4553,
      "mean_token_accuracy": 0.8565440736711025,
      "num_tokens": 744041237.0,
      "step": 23314
    },
    {
      "entropy": 0.44487542659044266,
      "epoch": 2.1450940477628078,
      "grad_norm": 0.1529327929019928,
      "learning_rate": 2.850001533413071e-05,
      "loss": 0.4315,
      "mean_token_accuracy": 0.8642019927501678,
      "num_tokens": 744073551.0,
      "step": 23315
    },
    {
      "entropy": 0.555076646618545,
      "epoch": 2.145186054134249,
      "grad_norm": 0.16754597425460815,
      "learning_rate": 2.8496948507989087e-05,
      "loss": 0.5472,
      "mean_token_accuracy": 0.8266343958675861,
      "num_tokens": 744105437.0,
      "step": 23316
    },
    {
      "entropy": 0.5413127709180117,
      "epoch": 2.14527806050569,
      "grad_norm": 0.16051168739795685,
      "learning_rate": 2.8493881681847455e-05,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8322501815855503,
      "num_tokens": 744137954.0,
      "step": 23317
    },
    {
      "entropy": 0.5161866266280413,
      "epoch": 2.145370066877131,
      "grad_norm": 0.16122625768184662,
      "learning_rate": 2.8490814855705833e-05,
      "loss": 0.5098,
      "mean_token_accuracy": 0.8414149880409241,
      "num_tokens": 744170254.0,
      "step": 23318
    },
    {
      "entropy": 0.5413804519921541,
      "epoch": 2.1454620732485723,
      "grad_norm": 0.16167312860488892,
      "learning_rate": 2.8487748029564204e-05,
      "loss": 0.5207,
      "mean_token_accuracy": 0.8342248648405075,
      "num_tokens": 744202499.0,
      "step": 23319
    },
    {
      "entropy": 0.48829830763861537,
      "epoch": 2.145554079620014,
      "grad_norm": 0.15267987549304962,
      "learning_rate": 2.8484681203422582e-05,
      "loss": 0.469,
      "mean_token_accuracy": 0.85326087474823,
      "num_tokens": 744235267.0,
      "step": 23320
    },
    {
      "entropy": 0.49090142315253615,
      "epoch": 2.145646085991455,
      "grad_norm": 0.16139782965183258,
      "learning_rate": 2.8481614377280953e-05,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8506980612874031,
      "num_tokens": 744266908.0,
      "step": 23321
    },
    {
      "entropy": 0.48516367631964386,
      "epoch": 2.145738092362896,
      "grad_norm": 0.15566948056221008,
      "learning_rate": 2.8478547551139328e-05,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8502838611602783,
      "num_tokens": 744298968.0,
      "step": 23322
    },
    {
      "entropy": 0.48936553206294775,
      "epoch": 2.1458300987343373,
      "grad_norm": 0.1515355259180069,
      "learning_rate": 2.84754807249977e-05,
      "loss": 0.4693,
      "mean_token_accuracy": 0.8503103703260422,
      "num_tokens": 744331367.0,
      "step": 23323
    },
    {
      "entropy": 0.5625440869480371,
      "epoch": 2.1459221051057784,
      "grad_norm": 0.16347189247608185,
      "learning_rate": 2.8472413898856077e-05,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8269117586314678,
      "num_tokens": 744362813.0,
      "step": 23324
    },
    {
      "entropy": 0.4860346589703113,
      "epoch": 2.14601411147722,
      "grad_norm": 0.16138501465320587,
      "learning_rate": 2.846934707271445e-05,
      "loss": 0.4656,
      "mean_token_accuracy": 0.8492652773857117,
      "num_tokens": 744394277.0,
      "step": 23325
    },
    {
      "entropy": 0.5563548374921083,
      "epoch": 2.146106117848661,
      "grad_norm": 0.1650405079126358,
      "learning_rate": 2.8466280246572823e-05,
      "loss": 0.5424,
      "mean_token_accuracy": 0.8288572207093239,
      "num_tokens": 744426195.0,
      "step": 23326
    },
    {
      "entropy": 0.5205330867320299,
      "epoch": 2.1461981242201023,
      "grad_norm": 0.1643301099538803,
      "learning_rate": 2.8463213420431194e-05,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8355981819331646,
      "num_tokens": 744457795.0,
      "step": 23327
    },
    {
      "entropy": 0.470852380618453,
      "epoch": 2.1462901305915434,
      "grad_norm": 0.15390226244926453,
      "learning_rate": 2.8460146594289572e-05,
      "loss": 0.4559,
      "mean_token_accuracy": 0.856119092553854,
      "num_tokens": 744489774.0,
      "step": 23328
    },
    {
      "entropy": 0.5319911446422338,
      "epoch": 2.1463821369629845,
      "grad_norm": 0.16653858125209808,
      "learning_rate": 2.8457079768147944e-05,
      "loss": 0.524,
      "mean_token_accuracy": 0.8320982754230499,
      "num_tokens": 744522012.0,
      "step": 23329
    },
    {
      "entropy": 0.5388133293017745,
      "epoch": 2.146474143334426,
      "grad_norm": 0.15934912860393524,
      "learning_rate": 2.845401294200632e-05,
      "loss": 0.5218,
      "mean_token_accuracy": 0.8351806811988354,
      "num_tokens": 744554426.0,
      "step": 23330
    },
    {
      "entropy": 0.49252062290906906,
      "epoch": 2.1465661497058672,
      "grad_norm": 0.15838932991027832,
      "learning_rate": 2.845094611586469e-05,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8468997441232204,
      "num_tokens": 744586623.0,
      "step": 23331
    },
    {
      "entropy": 0.48293980304151773,
      "epoch": 2.1466581560773084,
      "grad_norm": 0.15305301547050476,
      "learning_rate": 2.8447879289723068e-05,
      "loss": 0.4544,
      "mean_token_accuracy": 0.8516116477549076,
      "num_tokens": 744617869.0,
      "step": 23332
    },
    {
      "entropy": 0.565430699614808,
      "epoch": 2.1467501624487495,
      "grad_norm": 0.16371674835681915,
      "learning_rate": 2.844481246358144e-05,
      "loss": 0.5517,
      "mean_token_accuracy": 0.8272948376834393,
      "num_tokens": 744650510.0,
      "step": 23333
    },
    {
      "entropy": 0.6639653164893389,
      "epoch": 2.1468421688201906,
      "grad_norm": 0.17964686453342438,
      "learning_rate": 2.8441745637439814e-05,
      "loss": 0.6582,
      "mean_token_accuracy": 0.7948466464877129,
      "num_tokens": 744682427.0,
      "step": 23334
    },
    {
      "entropy": 0.4515158561989665,
      "epoch": 2.146934175191632,
      "grad_norm": 0.1523595154285431,
      "learning_rate": 2.8438678811298185e-05,
      "loss": 0.4389,
      "mean_token_accuracy": 0.8606476560235023,
      "num_tokens": 744714579.0,
      "step": 23335
    },
    {
      "entropy": 0.5611701216548681,
      "epoch": 2.1470261815630733,
      "grad_norm": 0.163155198097229,
      "learning_rate": 2.8435611985156563e-05,
      "loss": 0.539,
      "mean_token_accuracy": 0.8294143937528133,
      "num_tokens": 744746539.0,
      "step": 23336
    },
    {
      "entropy": 0.6765648629516363,
      "epoch": 2.1471181879345145,
      "grad_norm": 0.1771920621395111,
      "learning_rate": 2.843254515901494e-05,
      "loss": 0.6619,
      "mean_token_accuracy": 0.7960396073758602,
      "num_tokens": 744778864.0,
      "step": 23337
    },
    {
      "entropy": 0.5367965158075094,
      "epoch": 2.1472101943059556,
      "grad_norm": 0.16750745475292206,
      "learning_rate": 2.8429478332873312e-05,
      "loss": 0.5192,
      "mean_token_accuracy": 0.8364421129226685,
      "num_tokens": 744810082.0,
      "step": 23338
    },
    {
      "entropy": 0.42194920568726957,
      "epoch": 2.1473022006773967,
      "grad_norm": 0.14876675605773926,
      "learning_rate": 2.8426411506731687e-05,
      "loss": 0.413,
      "mean_token_accuracy": 0.8658715412020683,
      "num_tokens": 744842228.0,
      "step": 23339
    },
    {
      "entropy": 0.5139192622154951,
      "epoch": 2.1473942070488383,
      "grad_norm": 0.16177648305892944,
      "learning_rate": 2.8423344680590058e-05,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8409750461578369,
      "num_tokens": 744874395.0,
      "step": 23340
    },
    {
      "entropy": 0.44223692500963807,
      "epoch": 2.1474862134202795,
      "grad_norm": 0.15472617745399475,
      "learning_rate": 2.8420277854448436e-05,
      "loss": 0.4257,
      "mean_token_accuracy": 0.8649903051555157,
      "num_tokens": 744906720.0,
      "step": 23341
    },
    {
      "entropy": 0.5834575518965721,
      "epoch": 2.1475782197917206,
      "grad_norm": 0.16949881613254547,
      "learning_rate": 2.8417211028306807e-05,
      "loss": 0.5772,
      "mean_token_accuracy": 0.8187176510691643,
      "num_tokens": 744937726.0,
      "step": 23342
    },
    {
      "entropy": 0.533001497387886,
      "epoch": 2.1476702261631617,
      "grad_norm": 0.16105934977531433,
      "learning_rate": 2.8414144202165182e-05,
      "loss": 0.5222,
      "mean_token_accuracy": 0.8334355652332306,
      "num_tokens": 744969264.0,
      "step": 23343
    },
    {
      "entropy": 0.4946293383836746,
      "epoch": 2.147762232534603,
      "grad_norm": 0.1597868800163269,
      "learning_rate": 2.8411077376023553e-05,
      "loss": 0.482,
      "mean_token_accuracy": 0.8458305858075619,
      "num_tokens": 745000910.0,
      "step": 23344
    },
    {
      "entropy": 0.44934645062312484,
      "epoch": 2.1478542389060444,
      "grad_norm": 0.16374418139457703,
      "learning_rate": 2.840801054988193e-05,
      "loss": 0.4288,
      "mean_token_accuracy": 0.8645389266312122,
      "num_tokens": 745032589.0,
      "step": 23345
    },
    {
      "entropy": 0.585043067112565,
      "epoch": 2.1479462452774856,
      "grad_norm": 0.17250822484493256,
      "learning_rate": 2.8404943723740303e-05,
      "loss": 0.5714,
      "mean_token_accuracy": 0.8213891685009003,
      "num_tokens": 745064186.0,
      "step": 23346
    },
    {
      "entropy": 0.5722779808565974,
      "epoch": 2.1480382516489267,
      "grad_norm": 0.16417747735977173,
      "learning_rate": 2.8401876897598677e-05,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8239378221333027,
      "num_tokens": 745095827.0,
      "step": 23347
    },
    {
      "entropy": 0.5602258406579494,
      "epoch": 2.148130258020368,
      "grad_norm": 0.16532060503959656,
      "learning_rate": 2.839881007145705e-05,
      "loss": 0.5483,
      "mean_token_accuracy": 0.8299515806138515,
      "num_tokens": 745128125.0,
      "step": 23348
    },
    {
      "entropy": 0.4140780041925609,
      "epoch": 2.148222264391809,
      "grad_norm": 0.14169922471046448,
      "learning_rate": 2.8395743245315427e-05,
      "loss": 0.3978,
      "mean_token_accuracy": 0.8705954365432262,
      "num_tokens": 745160643.0,
      "step": 23349
    },
    {
      "entropy": 0.629798648878932,
      "epoch": 2.1483142707632505,
      "grad_norm": 0.16992737352848053,
      "learning_rate": 2.8392676419173798e-05,
      "loss": 0.6121,
      "mean_token_accuracy": 0.8111282028257847,
      "num_tokens": 745192887.0,
      "step": 23350
    },
    {
      "entropy": 0.591770994476974,
      "epoch": 2.1484062771346917,
      "grad_norm": 0.17155137658119202,
      "learning_rate": 2.8389609593032172e-05,
      "loss": 0.5751,
      "mean_token_accuracy": 0.8200083673000336,
      "num_tokens": 745224654.0,
      "step": 23351
    },
    {
      "entropy": 0.5325293764472008,
      "epoch": 2.148498283506133,
      "grad_norm": 0.16744883358478546,
      "learning_rate": 2.8386542766890544e-05,
      "loss": 0.5234,
      "mean_token_accuracy": 0.8333029039204121,
      "num_tokens": 745256074.0,
      "step": 23352
    },
    {
      "entropy": 0.4655007617548108,
      "epoch": 2.148590289877574,
      "grad_norm": 0.15517228841781616,
      "learning_rate": 2.8383475940748922e-05,
      "loss": 0.4501,
      "mean_token_accuracy": 0.8565389551222324,
      "num_tokens": 745288556.0,
      "step": 23353
    },
    {
      "entropy": 0.539492130279541,
      "epoch": 2.148682296249015,
      "grad_norm": 0.16986508667469025,
      "learning_rate": 2.8380409114607293e-05,
      "loss": 0.522,
      "mean_token_accuracy": 0.8347857035696507,
      "num_tokens": 745320104.0,
      "step": 23354
    },
    {
      "entropy": 0.4927043477073312,
      "epoch": 2.1487743026204567,
      "grad_norm": 0.15661977231502533,
      "learning_rate": 2.8377342288465668e-05,
      "loss": 0.4695,
      "mean_token_accuracy": 0.8503851406276226,
      "num_tokens": 745352257.0,
      "step": 23355
    },
    {
      "entropy": 0.5653113499283791,
      "epoch": 2.148866308991898,
      "grad_norm": 0.1693437397480011,
      "learning_rate": 2.837427546232404e-05,
      "loss": 0.5528,
      "mean_token_accuracy": 0.8279324062168598,
      "num_tokens": 745384651.0,
      "step": 23356
    },
    {
      "entropy": 0.541675501037389,
      "epoch": 2.148958315363339,
      "grad_norm": 0.16446590423583984,
      "learning_rate": 2.8371208636182417e-05,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8350731208920479,
      "num_tokens": 745416582.0,
      "step": 23357
    },
    {
      "entropy": 0.48764346167445183,
      "epoch": 2.14905032173478,
      "grad_norm": 0.15662892162799835,
      "learning_rate": 2.8368141810040788e-05,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8501162193715572,
      "num_tokens": 745448971.0,
      "step": 23358
    },
    {
      "entropy": 0.5503964452072978,
      "epoch": 2.149142328106221,
      "grad_norm": 0.1622060090303421,
      "learning_rate": 2.8365074983899166e-05,
      "loss": 0.5401,
      "mean_token_accuracy": 0.8284682966768742,
      "num_tokens": 745480830.0,
      "step": 23359
    },
    {
      "entropy": 0.49450945761054754,
      "epoch": 2.1492343344776628,
      "grad_norm": 0.1605997383594513,
      "learning_rate": 2.8362008157757537e-05,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8460445329546928,
      "num_tokens": 745512930.0,
      "step": 23360
    },
    {
      "entropy": 0.4760913043282926,
      "epoch": 2.149326340849104,
      "grad_norm": 0.15779022872447968,
      "learning_rate": 2.8358941331615912e-05,
      "loss": 0.4587,
      "mean_token_accuracy": 0.8521895855665207,
      "num_tokens": 745544582.0,
      "step": 23361
    },
    {
      "entropy": 0.5262236585840583,
      "epoch": 2.149418347220545,
      "grad_norm": 0.16202285885810852,
      "learning_rate": 2.8355874505474283e-05,
      "loss": 0.5011,
      "mean_token_accuracy": 0.8396730162203312,
      "num_tokens": 745576537.0,
      "step": 23362
    },
    {
      "entropy": 0.5123435440473258,
      "epoch": 2.149510353591986,
      "grad_norm": 0.16046592593193054,
      "learning_rate": 2.835280767933266e-05,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8393457718193531,
      "num_tokens": 745608964.0,
      "step": 23363
    },
    {
      "entropy": 0.5030562044121325,
      "epoch": 2.1496023599634273,
      "grad_norm": 0.1650247424840927,
      "learning_rate": 2.8349740853191033e-05,
      "loss": 0.4872,
      "mean_token_accuracy": 0.8430608548223972,
      "num_tokens": 745640681.0,
      "step": 23364
    },
    {
      "entropy": 0.4949344899505377,
      "epoch": 2.149694366334869,
      "grad_norm": 0.16207341849803925,
      "learning_rate": 2.8346674027049407e-05,
      "loss": 0.4776,
      "mean_token_accuracy": 0.8486727885901928,
      "num_tokens": 745673078.0,
      "step": 23365
    },
    {
      "entropy": 0.5885431747883558,
      "epoch": 2.14978637270631,
      "grad_norm": 0.17186731100082397,
      "learning_rate": 2.8343607200907785e-05,
      "loss": 0.5795,
      "mean_token_accuracy": 0.8187835216522217,
      "num_tokens": 745704812.0,
      "step": 23366
    },
    {
      "entropy": 0.47343552112579346,
      "epoch": 2.149878379077751,
      "grad_norm": 0.15301495790481567,
      "learning_rate": 2.8340540374766157e-05,
      "loss": 0.4637,
      "mean_token_accuracy": 0.8543756976723671,
      "num_tokens": 745736990.0,
      "step": 23367
    },
    {
      "entropy": 0.5795233882963657,
      "epoch": 2.1499703854491923,
      "grad_norm": 0.16253778338432312,
      "learning_rate": 2.833747354862453e-05,
      "loss": 0.563,
      "mean_token_accuracy": 0.8212206587195396,
      "num_tokens": 745768794.0,
      "step": 23368
    },
    {
      "entropy": 0.5199990896508098,
      "epoch": 2.1500623918206334,
      "grad_norm": 0.16337443888187408,
      "learning_rate": 2.8334406722482903e-05,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8375399932265282,
      "num_tokens": 745800852.0,
      "step": 23369
    },
    {
      "entropy": 0.45908898767083883,
      "epoch": 2.150154398192075,
      "grad_norm": 0.1559341698884964,
      "learning_rate": 2.833133989634128e-05,
      "loss": 0.4355,
      "mean_token_accuracy": 0.8634586594998837,
      "num_tokens": 745832510.0,
      "step": 23370
    },
    {
      "entropy": 0.6003435747697949,
      "epoch": 2.150246404563516,
      "grad_norm": 0.16584143042564392,
      "learning_rate": 2.8328273070199652e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8172398097813129,
      "num_tokens": 745865175.0,
      "step": 23371
    },
    {
      "entropy": 0.5217044604942203,
      "epoch": 2.1503384109349573,
      "grad_norm": 0.16251741349697113,
      "learning_rate": 2.8325206244058026e-05,
      "loss": 0.51,
      "mean_token_accuracy": 0.8384233675897121,
      "num_tokens": 745896275.0,
      "step": 23372
    },
    {
      "entropy": 0.5221348418854177,
      "epoch": 2.1504304173063984,
      "grad_norm": 0.15904167294502258,
      "learning_rate": 2.8322139417916398e-05,
      "loss": 0.504,
      "mean_token_accuracy": 0.8380654081702232,
      "num_tokens": 745928251.0,
      "step": 23373
    },
    {
      "entropy": 0.5202692188322544,
      "epoch": 2.1505224236778395,
      "grad_norm": 0.17003636062145233,
      "learning_rate": 2.8319072591774776e-05,
      "loss": 0.517,
      "mean_token_accuracy": 0.835818562656641,
      "num_tokens": 745959973.0,
      "step": 23374
    },
    {
      "entropy": 0.5384088875725865,
      "epoch": 2.150614430049281,
      "grad_norm": 0.16147051751613617,
      "learning_rate": 2.8316005765633147e-05,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8330411538481712,
      "num_tokens": 745992331.0,
      "step": 23375
    },
    {
      "entropy": 0.5284120896831155,
      "epoch": 2.1507064364207222,
      "grad_norm": 0.15692491829395294,
      "learning_rate": 2.8312938939491525e-05,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8432144299149513,
      "num_tokens": 746024574.0,
      "step": 23376
    },
    {
      "entropy": 0.4478382512461394,
      "epoch": 2.1507984427921634,
      "grad_norm": 0.15761972963809967,
      "learning_rate": 2.8309872113349893e-05,
      "loss": 0.4317,
      "mean_token_accuracy": 0.8627981841564178,
      "num_tokens": 746056330.0,
      "step": 23377
    },
    {
      "entropy": 0.5530404271557927,
      "epoch": 2.1508904491636045,
      "grad_norm": 0.16671758890151978,
      "learning_rate": 2.830680528720827e-05,
      "loss": 0.5475,
      "mean_token_accuracy": 0.8262438289821148,
      "num_tokens": 746088099.0,
      "step": 23378
    },
    {
      "entropy": 0.5254337675869465,
      "epoch": 2.1509824555350456,
      "grad_norm": 0.15958885848522186,
      "learning_rate": 2.8303738461066642e-05,
      "loss": 0.5101,
      "mean_token_accuracy": 0.8362896032631397,
      "num_tokens": 746120307.0,
      "step": 23379
    },
    {
      "entropy": 0.36035145772621036,
      "epoch": 2.151074461906487,
      "grad_norm": 0.1433117389678955,
      "learning_rate": 2.830067163492502e-05,
      "loss": 0.3417,
      "mean_token_accuracy": 0.8866857439279556,
      "num_tokens": 746152076.0,
      "step": 23380
    },
    {
      "entropy": 0.47302879113703966,
      "epoch": 2.1511664682779283,
      "grad_norm": 0.15205450356006622,
      "learning_rate": 2.829760480878339e-05,
      "loss": 0.4616,
      "mean_token_accuracy": 0.8584208562970161,
      "num_tokens": 746184844.0,
      "step": 23381
    },
    {
      "entropy": 0.5761929536238313,
      "epoch": 2.1512584746493695,
      "grad_norm": 0.16454672813415527,
      "learning_rate": 2.8294537982641766e-05,
      "loss": 0.5632,
      "mean_token_accuracy": 0.8242106288671494,
      "num_tokens": 746216454.0,
      "step": 23382
    },
    {
      "entropy": 0.5131681263446808,
      "epoch": 2.1513504810208106,
      "grad_norm": 0.15646101534366608,
      "learning_rate": 2.8291471156500137e-05,
      "loss": 0.5085,
      "mean_token_accuracy": 0.8388195410370827,
      "num_tokens": 746248559.0,
      "step": 23383
    },
    {
      "entropy": 0.4801428415812552,
      "epoch": 2.1514424873922517,
      "grad_norm": 0.15692241489887238,
      "learning_rate": 2.8288404330358515e-05,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8507545180618763,
      "num_tokens": 746279890.0,
      "step": 23384
    },
    {
      "entropy": 0.4985344037413597,
      "epoch": 2.1515344937636933,
      "grad_norm": 0.1562694013118744,
      "learning_rate": 2.8285337504216887e-05,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8470748364925385,
      "num_tokens": 746311637.0,
      "step": 23385
    },
    {
      "entropy": 0.4662233926355839,
      "epoch": 2.1516265001351345,
      "grad_norm": 0.15662403404712677,
      "learning_rate": 2.828227067807526e-05,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8523117750883102,
      "num_tokens": 746343236.0,
      "step": 23386
    },
    {
      "entropy": 0.44371055718511343,
      "epoch": 2.1517185065065756,
      "grad_norm": 0.16483132541179657,
      "learning_rate": 2.8279203851933633e-05,
      "loss": 0.4328,
      "mean_token_accuracy": 0.864654365926981,
      "num_tokens": 746374564.0,
      "step": 23387
    },
    {
      "entropy": 0.5264291316270828,
      "epoch": 2.1518105128780167,
      "grad_norm": 0.16766253113746643,
      "learning_rate": 2.827613702579201e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8359456919133663,
      "num_tokens": 746406120.0,
      "step": 23388
    },
    {
      "entropy": 0.5861297058872879,
      "epoch": 2.151902519249458,
      "grad_norm": 0.17068937420845032,
      "learning_rate": 2.8273070199650382e-05,
      "loss": 0.5732,
      "mean_token_accuracy": 0.823046687990427,
      "num_tokens": 746437748.0,
      "step": 23389
    },
    {
      "entropy": 0.5197410015389323,
      "epoch": 2.1519945256208994,
      "grad_norm": 0.16209368407726288,
      "learning_rate": 2.8270003373508757e-05,
      "loss": 0.5142,
      "mean_token_accuracy": 0.8368656858801842,
      "num_tokens": 746470430.0,
      "step": 23390
    },
    {
      "entropy": 0.5434769308194518,
      "epoch": 2.1520865319923406,
      "grad_norm": 0.15948033332824707,
      "learning_rate": 2.8266936547367128e-05,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8324911519885063,
      "num_tokens": 746502630.0,
      "step": 23391
    },
    {
      "entropy": 0.5930714784190059,
      "epoch": 2.1521785383637817,
      "grad_norm": 0.16588649153709412,
      "learning_rate": 2.8263869721225506e-05,
      "loss": 0.5747,
      "mean_token_accuracy": 0.8211705088615417,
      "num_tokens": 746534604.0,
      "step": 23392
    },
    {
      "entropy": 0.5728655178099871,
      "epoch": 2.152270544735223,
      "grad_norm": 0.16244839131832123,
      "learning_rate": 2.8260802895083877e-05,
      "loss": 0.5517,
      "mean_token_accuracy": 0.8286859318614006,
      "num_tokens": 746567246.0,
      "step": 23393
    },
    {
      "entropy": 0.48083787597715855,
      "epoch": 2.152362551106664,
      "grad_norm": 0.16034649312496185,
      "learning_rate": 2.8257736068942252e-05,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8487491048872471,
      "num_tokens": 746599161.0,
      "step": 23394
    },
    {
      "entropy": 0.5292590521275997,
      "epoch": 2.1524545574781055,
      "grad_norm": 0.16470156610012054,
      "learning_rate": 2.8254669242800623e-05,
      "loss": 0.51,
      "mean_token_accuracy": 0.834989856928587,
      "num_tokens": 746630523.0,
      "step": 23395
    },
    {
      "entropy": 0.5334465857595205,
      "epoch": 2.1525465638495467,
      "grad_norm": 0.15965594351291656,
      "learning_rate": 2.8251602416659e-05,
      "loss": 0.5225,
      "mean_token_accuracy": 0.8337605893611908,
      "num_tokens": 746662706.0,
      "step": 23396
    },
    {
      "entropy": 0.4958984572440386,
      "epoch": 2.152638570220988,
      "grad_norm": 0.15633311867713928,
      "learning_rate": 2.824853559051738e-05,
      "loss": 0.4851,
      "mean_token_accuracy": 0.8484735079109669,
      "num_tokens": 746695044.0,
      "step": 23397
    },
    {
      "entropy": 0.505479458719492,
      "epoch": 2.152730576592429,
      "grad_norm": 0.1594187170267105,
      "learning_rate": 2.824546876437575e-05,
      "loss": 0.4929,
      "mean_token_accuracy": 0.8432479985058308,
      "num_tokens": 746726790.0,
      "step": 23398
    },
    {
      "entropy": 0.5152829196304083,
      "epoch": 2.15282258296387,
      "grad_norm": 0.1566484570503235,
      "learning_rate": 2.8242401938234125e-05,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8432742208242416,
      "num_tokens": 746759120.0,
      "step": 23399
    },
    {
      "entropy": 0.552235959097743,
      "epoch": 2.1529145893353117,
      "grad_norm": 0.16777543723583221,
      "learning_rate": 2.8239335112092496e-05,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8303936272859573,
      "num_tokens": 746790791.0,
      "step": 23400
    },
    {
      "entropy": 0.5020347125828266,
      "epoch": 2.153006595706753,
      "grad_norm": 0.15665961802005768,
      "learning_rate": 2.8236268285950874e-05,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8422546088695526,
      "num_tokens": 746822807.0,
      "step": 23401
    },
    {
      "entropy": 0.5196290402673185,
      "epoch": 2.153098602078194,
      "grad_norm": 0.15953640639781952,
      "learning_rate": 2.8233201459809246e-05,
      "loss": 0.4944,
      "mean_token_accuracy": 0.8420410081744194,
      "num_tokens": 746854876.0,
      "step": 23402
    },
    {
      "entropy": 0.6193419918417931,
      "epoch": 2.153190608449635,
      "grad_norm": 0.16560125350952148,
      "learning_rate": 2.823013463366762e-05,
      "loss": 0.6079,
      "mean_token_accuracy": 0.8075143843889236,
      "num_tokens": 746886949.0,
      "step": 23403
    },
    {
      "entropy": 0.4978593732230365,
      "epoch": 2.153282614821076,
      "grad_norm": 0.15642084181308746,
      "learning_rate": 2.822706780752599e-05,
      "loss": 0.4803,
      "mean_token_accuracy": 0.8472082652151585,
      "num_tokens": 746919305.0,
      "step": 23404
    },
    {
      "entropy": 0.49680525064468384,
      "epoch": 2.1533746211925178,
      "grad_norm": 0.15783455967903137,
      "learning_rate": 2.822400098138437e-05,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8444799557328224,
      "num_tokens": 746951286.0,
      "step": 23405
    },
    {
      "entropy": 0.4997322456911206,
      "epoch": 2.153466627563959,
      "grad_norm": 0.1603190004825592,
      "learning_rate": 2.822093415524274e-05,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8441021852195263,
      "num_tokens": 746983252.0,
      "step": 23406
    },
    {
      "entropy": 0.5197136574424803,
      "epoch": 2.1535586339354,
      "grad_norm": 0.16315963864326477,
      "learning_rate": 2.8217867329101115e-05,
      "loss": 0.5042,
      "mean_token_accuracy": 0.8415290154516697,
      "num_tokens": 747014234.0,
      "step": 23407
    },
    {
      "entropy": 0.46837185230106115,
      "epoch": 2.153650640306841,
      "grad_norm": 0.1546141356229782,
      "learning_rate": 2.8214800502959487e-05,
      "loss": 0.4589,
      "mean_token_accuracy": 0.8523767478764057,
      "num_tokens": 747045922.0,
      "step": 23408
    },
    {
      "entropy": 0.519160027615726,
      "epoch": 2.1537426466782823,
      "grad_norm": 0.1622045338153839,
      "learning_rate": 2.8211733676817865e-05,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8415082097053528,
      "num_tokens": 747077567.0,
      "step": 23409
    },
    {
      "entropy": 0.48517479561269283,
      "epoch": 2.153834653049724,
      "grad_norm": 0.1616578996181488,
      "learning_rate": 2.8208666850676236e-05,
      "loss": 0.4736,
      "mean_token_accuracy": 0.8503417260944843,
      "num_tokens": 747109303.0,
      "step": 23410
    },
    {
      "entropy": 0.5878372620791197,
      "epoch": 2.153926659421165,
      "grad_norm": 0.16736091673374176,
      "learning_rate": 2.820560002453461e-05,
      "loss": 0.5705,
      "mean_token_accuracy": 0.8206657059490681,
      "num_tokens": 747141957.0,
      "step": 23411
    },
    {
      "entropy": 0.4420473761856556,
      "epoch": 2.154018665792606,
      "grad_norm": 0.15598176419734955,
      "learning_rate": 2.8202533198392982e-05,
      "loss": 0.4298,
      "mean_token_accuracy": 0.8655262105166912,
      "num_tokens": 747174241.0,
      "step": 23412
    },
    {
      "entropy": 0.5103033985942602,
      "epoch": 2.1541106721640473,
      "grad_norm": 0.16166849434375763,
      "learning_rate": 2.819946637225136e-05,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8432502821087837,
      "num_tokens": 747205809.0,
      "step": 23413
    },
    {
      "entropy": 0.617220526561141,
      "epoch": 2.1542026785354884,
      "grad_norm": 0.17003794014453888,
      "learning_rate": 2.819639954610973e-05,
      "loss": 0.6112,
      "mean_token_accuracy": 0.8098879344761372,
      "num_tokens": 747238185.0,
      "step": 23414
    },
    {
      "entropy": 0.5863440800458193,
      "epoch": 2.15429468490693,
      "grad_norm": 0.16486166417598724,
      "learning_rate": 2.8193332719968106e-05,
      "loss": 0.5776,
      "mean_token_accuracy": 0.8162607699632645,
      "num_tokens": 747270606.0,
      "step": 23415
    },
    {
      "entropy": 0.49287179857492447,
      "epoch": 2.154386691278371,
      "grad_norm": 0.15001410245895386,
      "learning_rate": 2.8190265893826477e-05,
      "loss": 0.4841,
      "mean_token_accuracy": 0.8452797271311283,
      "num_tokens": 747302679.0,
      "step": 23416
    },
    {
      "entropy": 0.47327770944684744,
      "epoch": 2.1544786976498123,
      "grad_norm": 0.15109902620315552,
      "learning_rate": 2.8187199067684855e-05,
      "loss": 0.4552,
      "mean_token_accuracy": 0.8565662875771523,
      "num_tokens": 747334197.0,
      "step": 23417
    },
    {
      "entropy": 0.5261021219193935,
      "epoch": 2.1545707040212534,
      "grad_norm": 0.164966881275177,
      "learning_rate": 2.8184132241543226e-05,
      "loss": 0.4972,
      "mean_token_accuracy": 0.8425910845398903,
      "num_tokens": 747365422.0,
      "step": 23418
    },
    {
      "entropy": 0.5251949857920408,
      "epoch": 2.1546627103926945,
      "grad_norm": 0.16137389838695526,
      "learning_rate": 2.8181065415401604e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8390032574534416,
      "num_tokens": 747397303.0,
      "step": 23419
    },
    {
      "entropy": 0.537420991808176,
      "epoch": 2.154754716764136,
      "grad_norm": 0.16950593888759613,
      "learning_rate": 2.8177998589259976e-05,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8338142484426498,
      "num_tokens": 747429326.0,
      "step": 23420
    },
    {
      "entropy": 0.5573554651346058,
      "epoch": 2.1548467231355772,
      "grad_norm": 0.1644422560930252,
      "learning_rate": 2.817493176311835e-05,
      "loss": 0.5399,
      "mean_token_accuracy": 0.8262745216488838,
      "num_tokens": 747460821.0,
      "step": 23421
    },
    {
      "entropy": 0.4108503647148609,
      "epoch": 2.1549387295070184,
      "grad_norm": 0.1506013125181198,
      "learning_rate": 2.817186493697672e-05,
      "loss": 0.4016,
      "mean_token_accuracy": 0.8714293912053108,
      "num_tokens": 747491949.0,
      "step": 23422
    },
    {
      "entropy": 0.5589940054342151,
      "epoch": 2.1550307358784595,
      "grad_norm": 0.16107536852359772,
      "learning_rate": 2.81687981108351e-05,
      "loss": 0.5365,
      "mean_token_accuracy": 0.8329453021287918,
      "num_tokens": 747524357.0,
      "step": 23423
    },
    {
      "entropy": 0.4415067955851555,
      "epoch": 2.1551227422499006,
      "grad_norm": 0.14593109488487244,
      "learning_rate": 2.816573128469347e-05,
      "loss": 0.4279,
      "mean_token_accuracy": 0.8630519099533558,
      "num_tokens": 747556591.0,
      "step": 23424
    },
    {
      "entropy": 0.4936859328299761,
      "epoch": 2.155214748621342,
      "grad_norm": 0.15459118783473969,
      "learning_rate": 2.8162664458551846e-05,
      "loss": 0.4831,
      "mean_token_accuracy": 0.848081286996603,
      "num_tokens": 747588561.0,
      "step": 23425
    },
    {
      "entropy": 0.5956119047477841,
      "epoch": 2.1553067549927833,
      "grad_norm": 0.16258633136749268,
      "learning_rate": 2.8159597632410217e-05,
      "loss": 0.583,
      "mean_token_accuracy": 0.8161331824958324,
      "num_tokens": 747620528.0,
      "step": 23426
    },
    {
      "entropy": 0.460949907079339,
      "epoch": 2.1553987613642245,
      "grad_norm": 0.14850854873657227,
      "learning_rate": 2.8156530806268595e-05,
      "loss": 0.4378,
      "mean_token_accuracy": 0.860158309340477,
      "num_tokens": 747652612.0,
      "step": 23427
    },
    {
      "entropy": 0.5643499307334423,
      "epoch": 2.1554907677356656,
      "grad_norm": 0.1641542613506317,
      "learning_rate": 2.815346398012697e-05,
      "loss": 0.5502,
      "mean_token_accuracy": 0.8288532122969627,
      "num_tokens": 747685052.0,
      "step": 23428
    },
    {
      "entropy": 0.4754661079496145,
      "epoch": 2.1555827741071067,
      "grad_norm": 0.15305957198143005,
      "learning_rate": 2.815039715398534e-05,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8522467650473118,
      "num_tokens": 747716189.0,
      "step": 23429
    },
    {
      "entropy": 0.578808382153511,
      "epoch": 2.1556747804785483,
      "grad_norm": 0.17098091542720795,
      "learning_rate": 2.814733032784372e-05,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8214322961866856,
      "num_tokens": 747747380.0,
      "step": 23430
    },
    {
      "entropy": 0.5209644697606564,
      "epoch": 2.1557667868499895,
      "grad_norm": 0.15723134577274323,
      "learning_rate": 2.814426350170209e-05,
      "loss": 0.5109,
      "mean_token_accuracy": 0.8348575793206692,
      "num_tokens": 747780038.0,
      "step": 23431
    },
    {
      "entropy": 0.47949178144335747,
      "epoch": 2.1558587932214306,
      "grad_norm": 0.15120293200016022,
      "learning_rate": 2.8141196675560465e-05,
      "loss": 0.4647,
      "mean_token_accuracy": 0.8540831282734871,
      "num_tokens": 747811512.0,
      "step": 23432
    },
    {
      "entropy": 0.4721435662358999,
      "epoch": 2.1559507995928717,
      "grad_norm": 0.15902309119701385,
      "learning_rate": 2.8138129849418836e-05,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8512546867132187,
      "num_tokens": 747842870.0,
      "step": 23433
    },
    {
      "entropy": 0.4948863787576556,
      "epoch": 2.156042805964313,
      "grad_norm": 0.16482451558113098,
      "learning_rate": 2.8135063023277214e-05,
      "loss": 0.48,
      "mean_token_accuracy": 0.8469477593898773,
      "num_tokens": 747874180.0,
      "step": 23434
    },
    {
      "entropy": 0.5718777887523174,
      "epoch": 2.1561348123357544,
      "grad_norm": 0.16566097736358643,
      "learning_rate": 2.8131996197135585e-05,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8276202753186226,
      "num_tokens": 747906032.0,
      "step": 23435
    },
    {
      "entropy": 0.5160844754427671,
      "epoch": 2.1562268187071956,
      "grad_norm": 0.16052626073360443,
      "learning_rate": 2.8128929370993963e-05,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8410536162555218,
      "num_tokens": 747937223.0,
      "step": 23436
    },
    {
      "entropy": 0.5655996473506093,
      "epoch": 2.1563188250786367,
      "grad_norm": 0.16484345495700836,
      "learning_rate": 2.812586254485233e-05,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8235893324017525,
      "num_tokens": 747969073.0,
      "step": 23437
    },
    {
      "entropy": 0.5721032153815031,
      "epoch": 2.156410831450078,
      "grad_norm": 0.16624456644058228,
      "learning_rate": 2.812279571871071e-05,
      "loss": 0.5575,
      "mean_token_accuracy": 0.8237679116427898,
      "num_tokens": 748001452.0,
      "step": 23438
    },
    {
      "entropy": 0.4740562466904521,
      "epoch": 2.156502837821519,
      "grad_norm": 0.16284073889255524,
      "learning_rate": 2.811972889256908e-05,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8511470779776573,
      "num_tokens": 748033367.0,
      "step": 23439
    },
    {
      "entropy": 0.4542404944077134,
      "epoch": 2.1565948441929605,
      "grad_norm": 0.15752914547920227,
      "learning_rate": 2.811666206642746e-05,
      "loss": 0.4417,
      "mean_token_accuracy": 0.8572495952248573,
      "num_tokens": 748064671.0,
      "step": 23440
    },
    {
      "entropy": 0.5588226020336151,
      "epoch": 2.1566868505644017,
      "grad_norm": 0.16518303751945496,
      "learning_rate": 2.811359524028583e-05,
      "loss": 0.5497,
      "mean_token_accuracy": 0.8267586678266525,
      "num_tokens": 748097439.0,
      "step": 23441
    },
    {
      "entropy": 0.5568671766668558,
      "epoch": 2.156778856935843,
      "grad_norm": 0.1665256917476654,
      "learning_rate": 2.8110528414144204e-05,
      "loss": 0.5485,
      "mean_token_accuracy": 0.8267584703862667,
      "num_tokens": 748129162.0,
      "step": 23442
    },
    {
      "entropy": 0.5326804388314486,
      "epoch": 2.156870863307284,
      "grad_norm": 0.16420000791549683,
      "learning_rate": 2.8107461588002576e-05,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8369474746286869,
      "num_tokens": 748160974.0,
      "step": 23443
    },
    {
      "entropy": 0.5981994923204184,
      "epoch": 2.156962869678725,
      "grad_norm": 0.17135190963745117,
      "learning_rate": 2.8104394761860954e-05,
      "loss": 0.5833,
      "mean_token_accuracy": 0.8193386904895306,
      "num_tokens": 748193122.0,
      "step": 23444
    },
    {
      "entropy": 0.5273944046348333,
      "epoch": 2.1570548760501667,
      "grad_norm": 0.1597188264131546,
      "learning_rate": 2.8101327935719325e-05,
      "loss": 0.5106,
      "mean_token_accuracy": 0.8372706323862076,
      "num_tokens": 748225759.0,
      "step": 23445
    },
    {
      "entropy": 0.4858878212980926,
      "epoch": 2.157146882421608,
      "grad_norm": 0.15583285689353943,
      "learning_rate": 2.80982611095777e-05,
      "loss": 0.4778,
      "mean_token_accuracy": 0.8469762951135635,
      "num_tokens": 748257465.0,
      "step": 23446
    },
    {
      "entropy": 0.5329191666096449,
      "epoch": 2.157238888793049,
      "grad_norm": 0.17072145640850067,
      "learning_rate": 2.809519428343607e-05,
      "loss": 0.5228,
      "mean_token_accuracy": 0.8354067727923393,
      "num_tokens": 748289480.0,
      "step": 23447
    },
    {
      "entropy": 0.4226704239845276,
      "epoch": 2.15733089516449,
      "grad_norm": 0.15220436453819275,
      "learning_rate": 2.809212745729445e-05,
      "loss": 0.4038,
      "mean_token_accuracy": 0.86734788864851,
      "num_tokens": 748320618.0,
      "step": 23448
    },
    {
      "entropy": 0.5085692908614874,
      "epoch": 2.157422901535931,
      "grad_norm": 0.15942294895648956,
      "learning_rate": 2.808906063115282e-05,
      "loss": 0.4888,
      "mean_token_accuracy": 0.8449982218444347,
      "num_tokens": 748352405.0,
      "step": 23449
    },
    {
      "entropy": 0.47348846308887005,
      "epoch": 2.1575149079073728,
      "grad_norm": 0.15385514497756958,
      "learning_rate": 2.8085993805011195e-05,
      "loss": 0.4602,
      "mean_token_accuracy": 0.85589150339365,
      "num_tokens": 748384433.0,
      "step": 23450
    },
    {
      "entropy": 0.562282214872539,
      "epoch": 2.157606914278814,
      "grad_norm": 0.17253363132476807,
      "learning_rate": 2.8082926978869566e-05,
      "loss": 0.5468,
      "mean_token_accuracy": 0.8278262689709663,
      "num_tokens": 748416516.0,
      "step": 23451
    },
    {
      "entropy": 0.4324852302670479,
      "epoch": 2.157698920650255,
      "grad_norm": 0.1520569920539856,
      "learning_rate": 2.8079860152727944e-05,
      "loss": 0.4126,
      "mean_token_accuracy": 0.8647802583873272,
      "num_tokens": 748449055.0,
      "step": 23452
    },
    {
      "entropy": 0.5905902050435543,
      "epoch": 2.157790927021696,
      "grad_norm": 0.1687912940979004,
      "learning_rate": 2.8076793326586315e-05,
      "loss": 0.5809,
      "mean_token_accuracy": 0.817687351256609,
      "num_tokens": 748481004.0,
      "step": 23453
    },
    {
      "entropy": 0.5167394056916237,
      "epoch": 2.1578829333931373,
      "grad_norm": 0.16065730154514313,
      "learning_rate": 2.807372650044469e-05,
      "loss": 0.4938,
      "mean_token_accuracy": 0.8426745198667049,
      "num_tokens": 748512828.0,
      "step": 23454
    },
    {
      "entropy": 0.4802569013554603,
      "epoch": 2.157974939764579,
      "grad_norm": 0.15349288284778595,
      "learning_rate": 2.807065967430306e-05,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8566337376832962,
      "num_tokens": 748545131.0,
      "step": 23455
    },
    {
      "entropy": 0.5476955231279135,
      "epoch": 2.15806694613602,
      "grad_norm": 0.15731972455978394,
      "learning_rate": 2.806759284816144e-05,
      "loss": 0.5316,
      "mean_token_accuracy": 0.8313037008047104,
      "num_tokens": 748577624.0,
      "step": 23456
    },
    {
      "entropy": 0.47734370082616806,
      "epoch": 2.158158952507461,
      "grad_norm": 0.1627618968486786,
      "learning_rate": 2.806452602201981e-05,
      "loss": 0.4764,
      "mean_token_accuracy": 0.847549568861723,
      "num_tokens": 748609043.0,
      "step": 23457
    },
    {
      "entropy": 0.5151209244504571,
      "epoch": 2.1582509588789023,
      "grad_norm": 0.16592298448085785,
      "learning_rate": 2.806145919587819e-05,
      "loss": 0.5055,
      "mean_token_accuracy": 0.836217287927866,
      "num_tokens": 748640449.0,
      "step": 23458
    },
    {
      "entropy": 0.5081052323803306,
      "epoch": 2.1583429652503434,
      "grad_norm": 0.1608268767595291,
      "learning_rate": 2.8058392369736563e-05,
      "loss": 0.49,
      "mean_token_accuracy": 0.8464403860270977,
      "num_tokens": 748673031.0,
      "step": 23459
    },
    {
      "entropy": 0.562757782638073,
      "epoch": 2.158434971621785,
      "grad_norm": 0.1679995357990265,
      "learning_rate": 2.8055325543594934e-05,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8259842321276665,
      "num_tokens": 748704452.0,
      "step": 23460
    },
    {
      "entropy": 0.4598269126145169,
      "epoch": 2.158526977993226,
      "grad_norm": 0.16098713874816895,
      "learning_rate": 2.8052258717453313e-05,
      "loss": 0.4539,
      "mean_token_accuracy": 0.8555351570248604,
      "num_tokens": 748735488.0,
      "step": 23461
    },
    {
      "entropy": 0.5119688119739294,
      "epoch": 2.1586189843646673,
      "grad_norm": 0.16090920567512512,
      "learning_rate": 2.8049191891311684e-05,
      "loss": 0.4959,
      "mean_token_accuracy": 0.8424451053142548,
      "num_tokens": 748767950.0,
      "step": 23462
    },
    {
      "entropy": 0.6432587709277868,
      "epoch": 2.1587109907361084,
      "grad_norm": 0.17698849737644196,
      "learning_rate": 2.804612506517006e-05,
      "loss": 0.6273,
      "mean_token_accuracy": 0.8045528754591942,
      "num_tokens": 748798636.0,
      "step": 23463
    },
    {
      "entropy": 0.519122782163322,
      "epoch": 2.1588029971075495,
      "grad_norm": 0.16432133316993713,
      "learning_rate": 2.804305823902843e-05,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8392215892672539,
      "num_tokens": 748831210.0,
      "step": 23464
    },
    {
      "entropy": 0.5664203120395541,
      "epoch": 2.158895003478991,
      "grad_norm": 0.16218310594558716,
      "learning_rate": 2.8039991412886808e-05,
      "loss": 0.5449,
      "mean_token_accuracy": 0.8293555527925491,
      "num_tokens": 748863784.0,
      "step": 23465
    },
    {
      "entropy": 0.5998277012258768,
      "epoch": 2.1589870098504322,
      "grad_norm": 0.16151203215122223,
      "learning_rate": 2.803692458674518e-05,
      "loss": 0.5705,
      "mean_token_accuracy": 0.8188213743269444,
      "num_tokens": 748896250.0,
      "step": 23466
    },
    {
      "entropy": 0.5173106230795383,
      "epoch": 2.1590790162218734,
      "grad_norm": 0.1617462933063507,
      "learning_rate": 2.8033857760603554e-05,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8401215709745884,
      "num_tokens": 748928701.0,
      "step": 23467
    },
    {
      "entropy": 0.6046184748411179,
      "epoch": 2.1591710225933145,
      "grad_norm": 0.17451760172843933,
      "learning_rate": 2.8030790934461925e-05,
      "loss": 0.589,
      "mean_token_accuracy": 0.8127662055194378,
      "num_tokens": 748960487.0,
      "step": 23468
    },
    {
      "entropy": 0.4872327616903931,
      "epoch": 2.1592630289647556,
      "grad_norm": 0.16226331889629364,
      "learning_rate": 2.8027724108320303e-05,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8484062664210796,
      "num_tokens": 748992131.0,
      "step": 23469
    },
    {
      "entropy": 0.5440181232988834,
      "epoch": 2.159355035336197,
      "grad_norm": 0.16622598469257355,
      "learning_rate": 2.8024657282178674e-05,
      "loss": 0.536,
      "mean_token_accuracy": 0.8301937319338322,
      "num_tokens": 749024306.0,
      "step": 23470
    },
    {
      "entropy": 0.5970989242196083,
      "epoch": 2.1594470417076383,
      "grad_norm": 0.17150059342384338,
      "learning_rate": 2.802159045603705e-05,
      "loss": 0.594,
      "mean_token_accuracy": 0.8124582879245281,
      "num_tokens": 749056323.0,
      "step": 23471
    },
    {
      "entropy": 0.4987118956632912,
      "epoch": 2.1595390480790795,
      "grad_norm": 0.15488897264003754,
      "learning_rate": 2.801852362989542e-05,
      "loss": 0.4841,
      "mean_token_accuracy": 0.8474741876125336,
      "num_tokens": 749088435.0,
      "step": 23472
    },
    {
      "entropy": 0.453484452329576,
      "epoch": 2.1596310544505206,
      "grad_norm": 0.15750132501125336,
      "learning_rate": 2.8015456803753798e-05,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8586261905729771,
      "num_tokens": 749119965.0,
      "step": 23473
    },
    {
      "entropy": 0.5092305559664965,
      "epoch": 2.1597230608219617,
      "grad_norm": 0.15229874849319458,
      "learning_rate": 2.801238997761217e-05,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8476951010525227,
      "num_tokens": 749152413.0,
      "step": 23474
    },
    {
      "entropy": 0.5601340373978019,
      "epoch": 2.1598150671934033,
      "grad_norm": 0.1686607003211975,
      "learning_rate": 2.8009323151470544e-05,
      "loss": 0.5381,
      "mean_token_accuracy": 0.831169605255127,
      "num_tokens": 749183443.0,
      "step": 23475
    },
    {
      "entropy": 0.4948364454321563,
      "epoch": 2.1599070735648445,
      "grad_norm": 0.15822890400886536,
      "learning_rate": 2.8006256325328915e-05,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8477302342653275,
      "num_tokens": 749215591.0,
      "step": 23476
    },
    {
      "entropy": 0.5590088404715061,
      "epoch": 2.1599990799362856,
      "grad_norm": 0.1743963062763214,
      "learning_rate": 2.8003189499187293e-05,
      "loss": 0.5437,
      "mean_token_accuracy": 0.826639924198389,
      "num_tokens": 749246859.0,
      "step": 23477
    },
    {
      "entropy": 0.6311176456511021,
      "epoch": 2.1600910863077267,
      "grad_norm": 0.1759830117225647,
      "learning_rate": 2.8000122673045665e-05,
      "loss": 0.6326,
      "mean_token_accuracy": 0.8049275614321232,
      "num_tokens": 749278621.0,
      "step": 23478
    },
    {
      "entropy": 0.5634284242987633,
      "epoch": 2.160183092679168,
      "grad_norm": 0.16827218234539032,
      "learning_rate": 2.7997055846904043e-05,
      "loss": 0.5558,
      "mean_token_accuracy": 0.8241136483848095,
      "num_tokens": 749311078.0,
      "step": 23479
    },
    {
      "entropy": 0.5239962283521891,
      "epoch": 2.1602750990506094,
      "grad_norm": 0.16442693769931793,
      "learning_rate": 2.7993989020762414e-05,
      "loss": 0.5323,
      "mean_token_accuracy": 0.8318459205329418,
      "num_tokens": 749343647.0,
      "step": 23480
    },
    {
      "entropy": 0.5988471303135157,
      "epoch": 2.1603671054220506,
      "grad_norm": 0.17763552069664001,
      "learning_rate": 2.799092219462079e-05,
      "loss": 0.5829,
      "mean_token_accuracy": 0.8127665221691132,
      "num_tokens": 749375054.0,
      "step": 23481
    },
    {
      "entropy": 0.4285264974460006,
      "epoch": 2.1604591117934917,
      "grad_norm": 0.1478218138217926,
      "learning_rate": 2.798785536847916e-05,
      "loss": 0.4149,
      "mean_token_accuracy": 0.8682870417833328,
      "num_tokens": 749407289.0,
      "step": 23482
    },
    {
      "entropy": 0.5081928675062954,
      "epoch": 2.160551118164933,
      "grad_norm": 0.1588996797800064,
      "learning_rate": 2.7984788542337538e-05,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8456437289714813,
      "num_tokens": 749438353.0,
      "step": 23483
    },
    {
      "entropy": 0.6321477498859167,
      "epoch": 2.160643124536374,
      "grad_norm": 0.17161470651626587,
      "learning_rate": 2.798172171619591e-05,
      "loss": 0.622,
      "mean_token_accuracy": 0.807765282690525,
      "num_tokens": 749469788.0,
      "step": 23484
    },
    {
      "entropy": 0.5788521599024534,
      "epoch": 2.1607351309078155,
      "grad_norm": 0.16820040345191956,
      "learning_rate": 2.7978654890054284e-05,
      "loss": 0.5688,
      "mean_token_accuracy": 0.8194657862186432,
      "num_tokens": 749502149.0,
      "step": 23485
    },
    {
      "entropy": 0.5486601823940873,
      "epoch": 2.1608271372792567,
      "grad_norm": 0.16890418529510498,
      "learning_rate": 2.7975588063912655e-05,
      "loss": 0.5332,
      "mean_token_accuracy": 0.8294647075235844,
      "num_tokens": 749534407.0,
      "step": 23486
    },
    {
      "entropy": 0.6296637393534184,
      "epoch": 2.160919143650698,
      "grad_norm": 0.17152133584022522,
      "learning_rate": 2.7972521237771033e-05,
      "loss": 0.6193,
      "mean_token_accuracy": 0.806844811886549,
      "num_tokens": 749566662.0,
      "step": 23487
    },
    {
      "entropy": 0.48248606733977795,
      "epoch": 2.161011150022139,
      "grad_norm": 0.15426206588745117,
      "learning_rate": 2.7969454411629408e-05,
      "loss": 0.4623,
      "mean_token_accuracy": 0.8508814126253128,
      "num_tokens": 749598449.0,
      "step": 23488
    },
    {
      "entropy": 0.5648638447746634,
      "epoch": 2.16110315639358,
      "grad_norm": 0.16529496014118195,
      "learning_rate": 2.796638758548778e-05,
      "loss": 0.5443,
      "mean_token_accuracy": 0.8312814421951771,
      "num_tokens": 749630372.0,
      "step": 23489
    },
    {
      "entropy": 0.5508358087390661,
      "epoch": 2.1611951627650217,
      "grad_norm": 0.1699836403131485,
      "learning_rate": 2.7963320759346157e-05,
      "loss": 0.5247,
      "mean_token_accuracy": 0.8347487263381481,
      "num_tokens": 749662205.0,
      "step": 23490
    },
    {
      "entropy": 0.5700990380719304,
      "epoch": 2.161287169136463,
      "grad_norm": 0.16638843715190887,
      "learning_rate": 2.7960253933204528e-05,
      "loss": 0.5588,
      "mean_token_accuracy": 0.8277051784098148,
      "num_tokens": 749694973.0,
      "step": 23491
    },
    {
      "entropy": 0.4902037354186177,
      "epoch": 2.161379175507904,
      "grad_norm": 0.16094845533370972,
      "learning_rate": 2.7957187107062903e-05,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8480814062058926,
      "num_tokens": 749727111.0,
      "step": 23492
    },
    {
      "entropy": 0.5341913036536425,
      "epoch": 2.161471181879345,
      "grad_norm": 0.1659790724515915,
      "learning_rate": 2.7954120280921274e-05,
      "loss": 0.534,
      "mean_token_accuracy": 0.8338386751711369,
      "num_tokens": 749759372.0,
      "step": 23493
    },
    {
      "entropy": 0.4721449613571167,
      "epoch": 2.161563188250786,
      "grad_norm": 0.1541302353143692,
      "learning_rate": 2.7951053454779652e-05,
      "loss": 0.4589,
      "mean_token_accuracy": 0.8515350110828876,
      "num_tokens": 749791589.0,
      "step": 23494
    },
    {
      "entropy": 0.5275736507028341,
      "epoch": 2.1616551946222278,
      "grad_norm": 0.16493746638298035,
      "learning_rate": 2.7947986628638023e-05,
      "loss": 0.5179,
      "mean_token_accuracy": 0.8370985947549343,
      "num_tokens": 749823964.0,
      "step": 23495
    },
    {
      "entropy": 0.6118538305163383,
      "epoch": 2.161747200993669,
      "grad_norm": 0.16800689697265625,
      "learning_rate": 2.79449198024964e-05,
      "loss": 0.6037,
      "mean_token_accuracy": 0.8117980994284153,
      "num_tokens": 749856060.0,
      "step": 23496
    },
    {
      "entropy": 0.5374245215207338,
      "epoch": 2.16183920736511,
      "grad_norm": 0.1672358363866806,
      "learning_rate": 2.794185297635477e-05,
      "loss": 0.5393,
      "mean_token_accuracy": 0.8291180022060871,
      "num_tokens": 749888198.0,
      "step": 23497
    },
    {
      "entropy": 0.47591205313801765,
      "epoch": 2.161931213736551,
      "grad_norm": 0.15862680971622467,
      "learning_rate": 2.7938786150213147e-05,
      "loss": 0.4683,
      "mean_token_accuracy": 0.8524094000458717,
      "num_tokens": 749920007.0,
      "step": 23498
    },
    {
      "entropy": 0.6243852255865932,
      "epoch": 2.1620232201079923,
      "grad_norm": 0.1703382432460785,
      "learning_rate": 2.793571932407152e-05,
      "loss": 0.6007,
      "mean_token_accuracy": 0.8133065924048424,
      "num_tokens": 749952142.0,
      "step": 23499
    },
    {
      "entropy": 0.561554005369544,
      "epoch": 2.162115226479434,
      "grad_norm": 0.16688407957553864,
      "learning_rate": 2.7932652497929897e-05,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8334239087998867,
      "num_tokens": 749984341.0,
      "step": 23500
    },
    {
      "entropy": 0.5580440768972039,
      "epoch": 2.162207232850875,
      "grad_norm": 0.16689492762088776,
      "learning_rate": 2.7929585671788268e-05,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8294375538825989,
      "num_tokens": 750016508.0,
      "step": 23501
    },
    {
      "entropy": 0.423405428417027,
      "epoch": 2.162299239222316,
      "grad_norm": 0.15076297521591187,
      "learning_rate": 2.7926518845646643e-05,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8710846193134785,
      "num_tokens": 750047788.0,
      "step": 23502
    },
    {
      "entropy": 0.5779765686020255,
      "epoch": 2.1623912455937573,
      "grad_norm": 0.16741055250167847,
      "learning_rate": 2.7923452019505014e-05,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8234153389930725,
      "num_tokens": 750079465.0,
      "step": 23503
    },
    {
      "entropy": 0.4748081723228097,
      "epoch": 2.1624832519651984,
      "grad_norm": 0.15834365785121918,
      "learning_rate": 2.7920385193363392e-05,
      "loss": 0.4612,
      "mean_token_accuracy": 0.8529379554092884,
      "num_tokens": 750111713.0,
      "step": 23504
    },
    {
      "entropy": 0.584604968316853,
      "epoch": 2.16257525833664,
      "grad_norm": 0.1672312319278717,
      "learning_rate": 2.7917318367221763e-05,
      "loss": 0.5696,
      "mean_token_accuracy": 0.8212711922824383,
      "num_tokens": 750143656.0,
      "step": 23505
    },
    {
      "entropy": 0.559148277156055,
      "epoch": 2.162667264708081,
      "grad_norm": 0.16436123847961426,
      "learning_rate": 2.7914251541080138e-05,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8298052214086056,
      "num_tokens": 750175549.0,
      "step": 23506
    },
    {
      "entropy": 0.440451149828732,
      "epoch": 2.1627592710795223,
      "grad_norm": 0.15696783363819122,
      "learning_rate": 2.791118471493851e-05,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8600175194442272,
      "num_tokens": 750208103.0,
      "step": 23507
    },
    {
      "entropy": 0.48516111820936203,
      "epoch": 2.1628512774509634,
      "grad_norm": 0.15906567871570587,
      "learning_rate": 2.7908117888796887e-05,
      "loss": 0.4829,
      "mean_token_accuracy": 0.8442528992891312,
      "num_tokens": 750240182.0,
      "step": 23508
    },
    {
      "entropy": 0.5304940305650234,
      "epoch": 2.1629432838224045,
      "grad_norm": 0.15438278019428253,
      "learning_rate": 2.790505106265526e-05,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8388989605009556,
      "num_tokens": 750272585.0,
      "step": 23509
    },
    {
      "entropy": 0.49388130754232407,
      "epoch": 2.163035290193846,
      "grad_norm": 0.16398444771766663,
      "learning_rate": 2.7901984236513633e-05,
      "loss": 0.4783,
      "mean_token_accuracy": 0.8467427082359791,
      "num_tokens": 750304159.0,
      "step": 23510
    },
    {
      "entropy": 0.5233590230345726,
      "epoch": 2.1631272965652872,
      "grad_norm": 0.1660962998867035,
      "learning_rate": 2.7898917410372004e-05,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8424194753170013,
      "num_tokens": 750336117.0,
      "step": 23511
    },
    {
      "entropy": 0.5340472301468253,
      "epoch": 2.1632193029367284,
      "grad_norm": 0.17043374478816986,
      "learning_rate": 2.7895850584230382e-05,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8305258080363274,
      "num_tokens": 750367747.0,
      "step": 23512
    },
    {
      "entropy": 0.5091903749853373,
      "epoch": 2.1633113093081695,
      "grad_norm": 0.1618773639202118,
      "learning_rate": 2.7892783758088754e-05,
      "loss": 0.4954,
      "mean_token_accuracy": 0.8426018208265305,
      "num_tokens": 750399673.0,
      "step": 23513
    },
    {
      "entropy": 0.4905216272454709,
      "epoch": 2.1634033156796106,
      "grad_norm": 0.16431476175785065,
      "learning_rate": 2.7889716931947128e-05,
      "loss": 0.485,
      "mean_token_accuracy": 0.8467851877212524,
      "num_tokens": 750430933.0,
      "step": 23514
    },
    {
      "entropy": 0.4730955818668008,
      "epoch": 2.163495322051052,
      "grad_norm": 0.15175168216228485,
      "learning_rate": 2.78866501058055e-05,
      "loss": 0.4612,
      "mean_token_accuracy": 0.8544984199106693,
      "num_tokens": 750462967.0,
      "step": 23515
    },
    {
      "entropy": 0.5519738225266337,
      "epoch": 2.1635873284224933,
      "grad_norm": 0.16438691318035126,
      "learning_rate": 2.7883583279663877e-05,
      "loss": 0.5413,
      "mean_token_accuracy": 0.8280327133834362,
      "num_tokens": 750493983.0,
      "step": 23516
    },
    {
      "entropy": 0.4943935563787818,
      "epoch": 2.1636793347939345,
      "grad_norm": 0.1578139066696167,
      "learning_rate": 2.788051645352225e-05,
      "loss": 0.4788,
      "mean_token_accuracy": 0.8513849750161171,
      "num_tokens": 750525891.0,
      "step": 23517
    },
    {
      "entropy": 0.6033784467726946,
      "epoch": 2.1637713411653756,
      "grad_norm": 0.17022839188575745,
      "learning_rate": 2.7877449627380627e-05,
      "loss": 0.5975,
      "mean_token_accuracy": 0.8097989186644554,
      "num_tokens": 750558047.0,
      "step": 23518
    },
    {
      "entropy": 0.4552340689115226,
      "epoch": 2.1638633475368168,
      "grad_norm": 0.15416878461837769,
      "learning_rate": 2.7874382801239e-05,
      "loss": 0.4415,
      "mean_token_accuracy": 0.8608532883226871,
      "num_tokens": 750589825.0,
      "step": 23519
    },
    {
      "entropy": 0.5972816254943609,
      "epoch": 2.1639553539082583,
      "grad_norm": 0.16898441314697266,
      "learning_rate": 2.7871315975097373e-05,
      "loss": 0.586,
      "mean_token_accuracy": 0.8127461522817612,
      "num_tokens": 750621378.0,
      "step": 23520
    },
    {
      "entropy": 0.504891854012385,
      "epoch": 2.1640473602796995,
      "grad_norm": 0.15479113161563873,
      "learning_rate": 2.786824914895575e-05,
      "loss": 0.487,
      "mean_token_accuracy": 0.8456603176891804,
      "num_tokens": 750653593.0,
      "step": 23521
    },
    {
      "entropy": 0.5934189278632402,
      "epoch": 2.1641393666511406,
      "grad_norm": 0.1735551804304123,
      "learning_rate": 2.7865182322814122e-05,
      "loss": 0.5716,
      "mean_token_accuracy": 0.8214004747569561,
      "num_tokens": 750684468.0,
      "step": 23522
    },
    {
      "entropy": 0.46584841050207615,
      "epoch": 2.1642313730225817,
      "grad_norm": 0.15624888241291046,
      "learning_rate": 2.7862115496672497e-05,
      "loss": 0.4351,
      "mean_token_accuracy": 0.859720092266798,
      "num_tokens": 750714990.0,
      "step": 23523
    },
    {
      "entropy": 0.5198540382552892,
      "epoch": 2.164323379394023,
      "grad_norm": 0.16089153289794922,
      "learning_rate": 2.7859048670530868e-05,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8404314927756786,
      "num_tokens": 750747364.0,
      "step": 23524
    },
    {
      "entropy": 0.5635293060913682,
      "epoch": 2.1644153857654644,
      "grad_norm": 0.16398704051971436,
      "learning_rate": 2.7855981844389246e-05,
      "loss": 0.5446,
      "mean_token_accuracy": 0.8294240832328796,
      "num_tokens": 750779474.0,
      "step": 23525
    },
    {
      "entropy": 0.6432756325230002,
      "epoch": 2.1645073921369056,
      "grad_norm": 0.1761314868927002,
      "learning_rate": 2.7852915018247617e-05,
      "loss": 0.6292,
      "mean_token_accuracy": 0.8012751489877701,
      "num_tokens": 750811218.0,
      "step": 23526
    },
    {
      "entropy": 0.5351234171539545,
      "epoch": 2.1645993985083467,
      "grad_norm": 0.16501948237419128,
      "learning_rate": 2.7849848192105992e-05,
      "loss": 0.5267,
      "mean_token_accuracy": 0.8327172212302685,
      "num_tokens": 750842840.0,
      "step": 23527
    },
    {
      "entropy": 0.5127795487642288,
      "epoch": 2.164691404879788,
      "grad_norm": 0.1625514179468155,
      "learning_rate": 2.7846781365964363e-05,
      "loss": 0.4927,
      "mean_token_accuracy": 0.8396698571741581,
      "num_tokens": 750875132.0,
      "step": 23528
    },
    {
      "entropy": 0.47016346361488104,
      "epoch": 2.164783411251229,
      "grad_norm": 0.1551342010498047,
      "learning_rate": 2.784371453982274e-05,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8568755015730858,
      "num_tokens": 750907752.0,
      "step": 23529
    },
    {
      "entropy": 0.4951634076423943,
      "epoch": 2.1648754176226706,
      "grad_norm": 0.15525226294994354,
      "learning_rate": 2.7840647713681112e-05,
      "loss": 0.4876,
      "mean_token_accuracy": 0.8452307283878326,
      "num_tokens": 750940093.0,
      "step": 23530
    },
    {
      "entropy": 0.46702174935489893,
      "epoch": 2.1649674239941117,
      "grad_norm": 0.15247321128845215,
      "learning_rate": 2.7837580887539487e-05,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8588279262185097,
      "num_tokens": 750972538.0,
      "step": 23531
    },
    {
      "entropy": 0.4906459618359804,
      "epoch": 2.165059430365553,
      "grad_norm": 0.16153042018413544,
      "learning_rate": 2.7834514061397858e-05,
      "loss": 0.4773,
      "mean_token_accuracy": 0.8458117842674255,
      "num_tokens": 751004678.0,
      "step": 23532
    },
    {
      "entropy": 0.5695738568902016,
      "epoch": 2.165151436736994,
      "grad_norm": 0.16855330765247345,
      "learning_rate": 2.7831447235256236e-05,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8246600739657879,
      "num_tokens": 751037110.0,
      "step": 23533
    },
    {
      "entropy": 0.541579857468605,
      "epoch": 2.165243443108435,
      "grad_norm": 0.16661040484905243,
      "learning_rate": 2.7828380409114608e-05,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8339703492820263,
      "num_tokens": 751068432.0,
      "step": 23534
    },
    {
      "entropy": 0.43913757940754294,
      "epoch": 2.1653354494798767,
      "grad_norm": 0.1519843339920044,
      "learning_rate": 2.7825313582972982e-05,
      "loss": 0.4272,
      "mean_token_accuracy": 0.8622751012444496,
      "num_tokens": 751100472.0,
      "step": 23535
    },
    {
      "entropy": 0.4596766531467438,
      "epoch": 2.165427455851318,
      "grad_norm": 0.15301427245140076,
      "learning_rate": 2.7822246756831353e-05,
      "loss": 0.4466,
      "mean_token_accuracy": 0.8556118719279766,
      "num_tokens": 751133240.0,
      "step": 23536
    },
    {
      "entropy": 0.4733863342553377,
      "epoch": 2.165519462222759,
      "grad_norm": 0.15736396610736847,
      "learning_rate": 2.781917993068973e-05,
      "loss": 0.4541,
      "mean_token_accuracy": 0.8556984439492226,
      "num_tokens": 751165449.0,
      "step": 23537
    },
    {
      "entropy": 0.4458550000563264,
      "epoch": 2.1656114685942,
      "grad_norm": 0.14816726744174957,
      "learning_rate": 2.7816113104548103e-05,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8589471243321896,
      "num_tokens": 751197365.0,
      "step": 23538
    },
    {
      "entropy": 0.4821312176063657,
      "epoch": 2.165703474965641,
      "grad_norm": 0.16116653382778168,
      "learning_rate": 2.781304627840648e-05,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8546739891171455,
      "num_tokens": 751229033.0,
      "step": 23539
    },
    {
      "entropy": 0.5917608318850398,
      "epoch": 2.1657954813370828,
      "grad_norm": 0.17378973960876465,
      "learning_rate": 2.7809979452264852e-05,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8179279081523418,
      "num_tokens": 751261360.0,
      "step": 23540
    },
    {
      "entropy": 0.5958796460181475,
      "epoch": 2.165887487708524,
      "grad_norm": 0.17511942982673645,
      "learning_rate": 2.7806912626123227e-05,
      "loss": 0.5888,
      "mean_token_accuracy": 0.8167043998837471,
      "num_tokens": 751293438.0,
      "step": 23541
    },
    {
      "entropy": 0.6127273426391184,
      "epoch": 2.165979494079965,
      "grad_norm": 0.1719607263803482,
      "learning_rate": 2.7803845799981598e-05,
      "loss": 0.6114,
      "mean_token_accuracy": 0.8111972734332085,
      "num_tokens": 751325499.0,
      "step": 23542
    },
    {
      "entropy": 0.46341103687882423,
      "epoch": 2.166071500451406,
      "grad_norm": 0.15062321722507477,
      "learning_rate": 2.7800778973839976e-05,
      "loss": 0.4501,
      "mean_token_accuracy": 0.8555679433047771,
      "num_tokens": 751357798.0,
      "step": 23543
    },
    {
      "entropy": 0.48495849408209324,
      "epoch": 2.1661635068228473,
      "grad_norm": 0.15872354805469513,
      "learning_rate": 2.7797712147698347e-05,
      "loss": 0.4677,
      "mean_token_accuracy": 0.8509110920131207,
      "num_tokens": 751390120.0,
      "step": 23544
    },
    {
      "entropy": 0.5077257994562387,
      "epoch": 2.166255513194289,
      "grad_norm": 0.16080856323242188,
      "learning_rate": 2.7794645321556722e-05,
      "loss": 0.4979,
      "mean_token_accuracy": 0.8398085832595825,
      "num_tokens": 751422411.0,
      "step": 23545
    },
    {
      "entropy": 0.49285715026780963,
      "epoch": 2.16634751956573,
      "grad_norm": 0.1565607786178589,
      "learning_rate": 2.7791578495415093e-05,
      "loss": 0.4688,
      "mean_token_accuracy": 0.8468815833330154,
      "num_tokens": 751454005.0,
      "step": 23546
    },
    {
      "entropy": 0.6039428389631212,
      "epoch": 2.166439525937171,
      "grad_norm": 0.16942164301872253,
      "learning_rate": 2.778851166927347e-05,
      "loss": 0.5887,
      "mean_token_accuracy": 0.8147055953741074,
      "num_tokens": 751485856.0,
      "step": 23547
    },
    {
      "entropy": 0.5052094762213528,
      "epoch": 2.1665315323086123,
      "grad_norm": 0.1626354306936264,
      "learning_rate": 2.7785444843131842e-05,
      "loss": 0.4928,
      "mean_token_accuracy": 0.8431329764425755,
      "num_tokens": 751517680.0,
      "step": 23548
    },
    {
      "entropy": 0.5134120220318437,
      "epoch": 2.1666235386800534,
      "grad_norm": 0.15904459357261658,
      "learning_rate": 2.7782378016990217e-05,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8440141789615154,
      "num_tokens": 751550411.0,
      "step": 23549
    },
    {
      "entropy": 0.4526109730359167,
      "epoch": 2.166715545051495,
      "grad_norm": 0.15456737577915192,
      "learning_rate": 2.7779311190848595e-05,
      "loss": 0.4424,
      "mean_token_accuracy": 0.8598557412624359,
      "num_tokens": 751582704.0,
      "step": 23550
    },
    {
      "entropy": 0.475658874027431,
      "epoch": 2.166807551422936,
      "grad_norm": 0.1558513343334198,
      "learning_rate": 2.7776244364706966e-05,
      "loss": 0.4631,
      "mean_token_accuracy": 0.8540302254259586,
      "num_tokens": 751614448.0,
      "step": 23551
    },
    {
      "entropy": 0.5935638975352049,
      "epoch": 2.1668995577943773,
      "grad_norm": 0.16702395677566528,
      "learning_rate": 2.777317753856534e-05,
      "loss": 0.5914,
      "mean_token_accuracy": 0.8199920803308487,
      "num_tokens": 751646030.0,
      "step": 23552
    },
    {
      "entropy": 0.5083884038031101,
      "epoch": 2.1669915641658184,
      "grad_norm": 0.1581403613090515,
      "learning_rate": 2.7770110712423712e-05,
      "loss": 0.4972,
      "mean_token_accuracy": 0.8434110060334206,
      "num_tokens": 751678745.0,
      "step": 23553
    },
    {
      "entropy": 0.4632435920648277,
      "epoch": 2.1670835705372595,
      "grad_norm": 0.15580323338508606,
      "learning_rate": 2.776704388628209e-05,
      "loss": 0.4607,
      "mean_token_accuracy": 0.8545924536883831,
      "num_tokens": 751710645.0,
      "step": 23554
    },
    {
      "entropy": 0.5339697455056012,
      "epoch": 2.167175576908701,
      "grad_norm": 0.16268415749073029,
      "learning_rate": 2.776397706014046e-05,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8344279415905476,
      "num_tokens": 751742392.0,
      "step": 23555
    },
    {
      "entropy": 0.5241322163492441,
      "epoch": 2.1672675832801422,
      "grad_norm": 0.16368548572063446,
      "learning_rate": 2.776091023399884e-05,
      "loss": 0.5098,
      "mean_token_accuracy": 0.8363889493048191,
      "num_tokens": 751774436.0,
      "step": 23556
    },
    {
      "entropy": 0.49880196806043386,
      "epoch": 2.1673595896515834,
      "grad_norm": 0.1562802791595459,
      "learning_rate": 2.7757843407857208e-05,
      "loss": 0.4809,
      "mean_token_accuracy": 0.845437865704298,
      "num_tokens": 751806592.0,
      "step": 23557
    },
    {
      "entropy": 0.4888417860493064,
      "epoch": 2.1674515960230245,
      "grad_norm": 0.15481430292129517,
      "learning_rate": 2.7754776581715586e-05,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8500776626169682,
      "num_tokens": 751838514.0,
      "step": 23558
    },
    {
      "entropy": 0.5036808652803302,
      "epoch": 2.1675436023944656,
      "grad_norm": 0.1571437567472458,
      "learning_rate": 2.7751709755573957e-05,
      "loss": 0.4956,
      "mean_token_accuracy": 0.844167910516262,
      "num_tokens": 751870637.0,
      "step": 23559
    },
    {
      "entropy": 0.5496395418886095,
      "epoch": 2.167635608765907,
      "grad_norm": 0.1584513932466507,
      "learning_rate": 2.7748642929432335e-05,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8335269689559937,
      "num_tokens": 751902835.0,
      "step": 23560
    },
    {
      "entropy": 0.5061427187174559,
      "epoch": 2.1677276151373484,
      "grad_norm": 0.15576350688934326,
      "learning_rate": 2.7745576103290706e-05,
      "loss": 0.4885,
      "mean_token_accuracy": 0.8458912633359432,
      "num_tokens": 751934576.0,
      "step": 23561
    },
    {
      "entropy": 0.5775036420673132,
      "epoch": 2.1678196215087895,
      "grad_norm": 0.16536788642406464,
      "learning_rate": 2.774250927714908e-05,
      "loss": 0.5562,
      "mean_token_accuracy": 0.8252453505992889,
      "num_tokens": 751966212.0,
      "step": 23562
    },
    {
      "entropy": 0.5228283302858472,
      "epoch": 2.1679116278802306,
      "grad_norm": 0.16041462123394012,
      "learning_rate": 2.7739442451007452e-05,
      "loss": 0.5035,
      "mean_token_accuracy": 0.8404016084969044,
      "num_tokens": 751997999.0,
      "step": 23563
    },
    {
      "entropy": 0.5577994002960622,
      "epoch": 2.1680036342516718,
      "grad_norm": 0.1628945916891098,
      "learning_rate": 2.773637562486583e-05,
      "loss": 0.541,
      "mean_token_accuracy": 0.8307371288537979,
      "num_tokens": 752029601.0,
      "step": 23564
    },
    {
      "entropy": 0.481044327840209,
      "epoch": 2.1680956406231133,
      "grad_norm": 0.15563496947288513,
      "learning_rate": 2.77333087987242e-05,
      "loss": 0.4686,
      "mean_token_accuracy": 0.8507943190634251,
      "num_tokens": 752060729.0,
      "step": 23565
    },
    {
      "entropy": 0.5850902972742915,
      "epoch": 2.1681876469945545,
      "grad_norm": 0.16668768227100372,
      "learning_rate": 2.7730241972582576e-05,
      "loss": 0.5748,
      "mean_token_accuracy": 0.8202607408165932,
      "num_tokens": 752093307.0,
      "step": 23566
    },
    {
      "entropy": 0.5076544571202248,
      "epoch": 2.1682796533659956,
      "grad_norm": 0.15964992344379425,
      "learning_rate": 2.7727175146440947e-05,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8463269434869289,
      "num_tokens": 752125263.0,
      "step": 23567
    },
    {
      "entropy": 0.48034973815083504,
      "epoch": 2.1683716597374367,
      "grad_norm": 0.15666797757148743,
      "learning_rate": 2.7724108320299325e-05,
      "loss": 0.462,
      "mean_token_accuracy": 0.8525336794555187,
      "num_tokens": 752157329.0,
      "step": 23568
    },
    {
      "entropy": 0.550719091668725,
      "epoch": 2.168463666108878,
      "grad_norm": 0.16247797012329102,
      "learning_rate": 2.7721041494157697e-05,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8293732292950153,
      "num_tokens": 752189271.0,
      "step": 23569
    },
    {
      "entropy": 0.5658561848104,
      "epoch": 2.1685556724803194,
      "grad_norm": 0.1633853316307068,
      "learning_rate": 2.771797466801607e-05,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8244852349162102,
      "num_tokens": 752221507.0,
      "step": 23570
    },
    {
      "entropy": 0.5374307427555323,
      "epoch": 2.1686476788517606,
      "grad_norm": 0.16471897065639496,
      "learning_rate": 2.7714907841874442e-05,
      "loss": 0.5222,
      "mean_token_accuracy": 0.837516974657774,
      "num_tokens": 752252488.0,
      "step": 23571
    },
    {
      "entropy": 0.4606685386970639,
      "epoch": 2.1687396852232017,
      "grad_norm": 0.15174779295921326,
      "learning_rate": 2.771184101573282e-05,
      "loss": 0.4408,
      "mean_token_accuracy": 0.8556373901665211,
      "num_tokens": 752283050.0,
      "step": 23572
    },
    {
      "entropy": 0.5427520824596286,
      "epoch": 2.168831691594643,
      "grad_norm": 0.16258922219276428,
      "learning_rate": 2.7708774189591192e-05,
      "loss": 0.5334,
      "mean_token_accuracy": 0.8328836634755135,
      "num_tokens": 752314300.0,
      "step": 23573
    },
    {
      "entropy": 0.5243508508428931,
      "epoch": 2.168923697966084,
      "grad_norm": 0.16759620606899261,
      "learning_rate": 2.7705707363449566e-05,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8384123258292675,
      "num_tokens": 752345544.0,
      "step": 23574
    },
    {
      "entropy": 0.5222736336290836,
      "epoch": 2.1690157043375256,
      "grad_norm": 0.16128437221050262,
      "learning_rate": 2.7702640537307938e-05,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8388083539903164,
      "num_tokens": 752377376.0,
      "step": 23575
    },
    {
      "entropy": 0.6256542708724737,
      "epoch": 2.1691077107089667,
      "grad_norm": 0.17684078216552734,
      "learning_rate": 2.7699573711166316e-05,
      "loss": 0.6065,
      "mean_token_accuracy": 0.8092172406613827,
      "num_tokens": 752408564.0,
      "step": 23576
    },
    {
      "entropy": 0.5687079578638077,
      "epoch": 2.169199717080408,
      "grad_norm": 0.16914983093738556,
      "learning_rate": 2.7696506885024687e-05,
      "loss": 0.5395,
      "mean_token_accuracy": 0.82804299518466,
      "num_tokens": 752440458.0,
      "step": 23577
    },
    {
      "entropy": 0.5340505857020617,
      "epoch": 2.169291723451849,
      "grad_norm": 0.1701941192150116,
      "learning_rate": 2.7693440058883065e-05,
      "loss": 0.53,
      "mean_token_accuracy": 0.8319999910891056,
      "num_tokens": 752471966.0,
      "step": 23578
    },
    {
      "entropy": 0.43196226470172405,
      "epoch": 2.16938372982329,
      "grad_norm": 0.14686809480190277,
      "learning_rate": 2.7690373232741433e-05,
      "loss": 0.4185,
      "mean_token_accuracy": 0.8663381971418858,
      "num_tokens": 752504440.0,
      "step": 23579
    },
    {
      "entropy": 0.4808097742497921,
      "epoch": 2.1694757361947317,
      "grad_norm": 0.1560438573360443,
      "learning_rate": 2.768730640659981e-05,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8530092984437943,
      "num_tokens": 752536973.0,
      "step": 23580
    },
    {
      "entropy": 0.504456015303731,
      "epoch": 2.169567742566173,
      "grad_norm": 0.16020697355270386,
      "learning_rate": 2.768423958045819e-05,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8414736948907375,
      "num_tokens": 752568644.0,
      "step": 23581
    },
    {
      "entropy": 0.38194220303557813,
      "epoch": 2.169659748937614,
      "grad_norm": 0.15416130423545837,
      "learning_rate": 2.768117275431656e-05,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8793069086968899,
      "num_tokens": 752600804.0,
      "step": 23582
    },
    {
      "entropy": 0.4809514218941331,
      "epoch": 2.169751755309055,
      "grad_norm": 0.16486616432666779,
      "learning_rate": 2.7678105928174935e-05,
      "loss": 0.4815,
      "mean_token_accuracy": 0.8490571714937687,
      "num_tokens": 752632988.0,
      "step": 23583
    },
    {
      "entropy": 0.5414765095338225,
      "epoch": 2.169843761680496,
      "grad_norm": 0.16350267827510834,
      "learning_rate": 2.7675039102033306e-05,
      "loss": 0.5276,
      "mean_token_accuracy": 0.8352518752217293,
      "num_tokens": 752664539.0,
      "step": 23584
    },
    {
      "entropy": 0.4331586379557848,
      "epoch": 2.1699357680519378,
      "grad_norm": 0.14891517162322998,
      "learning_rate": 2.7671972275891684e-05,
      "loss": 0.4199,
      "mean_token_accuracy": 0.8721631988883018,
      "num_tokens": 752696109.0,
      "step": 23585
    },
    {
      "entropy": 0.625833653844893,
      "epoch": 2.170027774423379,
      "grad_norm": 0.17495587468147278,
      "learning_rate": 2.7668905449750055e-05,
      "loss": 0.6133,
      "mean_token_accuracy": 0.8084212616086006,
      "num_tokens": 752728640.0,
      "step": 23586
    },
    {
      "entropy": 0.5325528653338552,
      "epoch": 2.17011978079482,
      "grad_norm": 0.1667054444551468,
      "learning_rate": 2.766583862360843e-05,
      "loss": 0.5227,
      "mean_token_accuracy": 0.830203715711832,
      "num_tokens": 752760000.0,
      "step": 23587
    },
    {
      "entropy": 0.5207351502031088,
      "epoch": 2.170211787166261,
      "grad_norm": 0.16123539209365845,
      "learning_rate": 2.76627717974668e-05,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8393412940204144,
      "num_tokens": 752791719.0,
      "step": 23588
    },
    {
      "entropy": 0.49465963896363974,
      "epoch": 2.1703037935377023,
      "grad_norm": 0.15821321308612823,
      "learning_rate": 2.765970497132518e-05,
      "loss": 0.4822,
      "mean_token_accuracy": 0.8457254357635975,
      "num_tokens": 752824214.0,
      "step": 23589
    },
    {
      "entropy": 0.5103702694177628,
      "epoch": 2.170395799909144,
      "grad_norm": 0.15717671811580658,
      "learning_rate": 2.765663814518355e-05,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8430751450359821,
      "num_tokens": 752855958.0,
      "step": 23590
    },
    {
      "entropy": 0.5209427010267973,
      "epoch": 2.170487806280585,
      "grad_norm": 0.15957434475421906,
      "learning_rate": 2.7653571319041925e-05,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8397612757980824,
      "num_tokens": 752888030.0,
      "step": 23591
    },
    {
      "entropy": 0.5583063922822475,
      "epoch": 2.170579812652026,
      "grad_norm": 0.16046522557735443,
      "learning_rate": 2.7650504492900296e-05,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8283785097301006,
      "num_tokens": 752919920.0,
      "step": 23592
    },
    {
      "entropy": 0.5243595875799656,
      "epoch": 2.1706718190234673,
      "grad_norm": 0.1608017086982727,
      "learning_rate": 2.7647437666758675e-05,
      "loss": 0.5054,
      "mean_token_accuracy": 0.8410822749137878,
      "num_tokens": 752952027.0,
      "step": 23593
    },
    {
      "entropy": 0.4164941553026438,
      "epoch": 2.1707638253949084,
      "grad_norm": 0.14942507445812225,
      "learning_rate": 2.7644370840617046e-05,
      "loss": 0.3958,
      "mean_token_accuracy": 0.8742103092372417,
      "num_tokens": 752984152.0,
      "step": 23594
    },
    {
      "entropy": 0.5318759884685278,
      "epoch": 2.17085583176635,
      "grad_norm": 0.16282311081886292,
      "learning_rate": 2.764130401447542e-05,
      "loss": 0.5234,
      "mean_token_accuracy": 0.8382059894502163,
      "num_tokens": 753016228.0,
      "step": 23595
    },
    {
      "entropy": 0.4418167541734874,
      "epoch": 2.170947838137791,
      "grad_norm": 0.14507968723773956,
      "learning_rate": 2.763823718833379e-05,
      "loss": 0.427,
      "mean_token_accuracy": 0.8653417713940144,
      "num_tokens": 753048883.0,
      "step": 23596
    },
    {
      "entropy": 0.5758184501901269,
      "epoch": 2.1710398445092323,
      "grad_norm": 0.16948561370372772,
      "learning_rate": 2.763517036219217e-05,
      "loss": 0.5719,
      "mean_token_accuracy": 0.8248256146907806,
      "num_tokens": 753081040.0,
      "step": 23597
    },
    {
      "entropy": 0.54065325204283,
      "epoch": 2.1711318508806734,
      "grad_norm": 0.15990611910820007,
      "learning_rate": 2.763210353605054e-05,
      "loss": 0.5295,
      "mean_token_accuracy": 0.83566889539361,
      "num_tokens": 753113626.0,
      "step": 23598
    },
    {
      "entropy": 0.4212795502971858,
      "epoch": 2.1712238572521145,
      "grad_norm": 0.14903590083122253,
      "learning_rate": 2.762903670990892e-05,
      "loss": 0.4025,
      "mean_token_accuracy": 0.8732945807278156,
      "num_tokens": 753145799.0,
      "step": 23599
    },
    {
      "entropy": 0.5420069750398397,
      "epoch": 2.171315863623556,
      "grad_norm": 0.16711321473121643,
      "learning_rate": 2.762596988376729e-05,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8371834643185139,
      "num_tokens": 753177337.0,
      "step": 23600
    },
    {
      "entropy": 0.5062887873500586,
      "epoch": 2.1714078699949972,
      "grad_norm": 0.15727591514587402,
      "learning_rate": 2.7622903057625665e-05,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8399352915585041,
      "num_tokens": 753209833.0,
      "step": 23601
    },
    {
      "entropy": 0.5152505617588758,
      "epoch": 2.1714998763664384,
      "grad_norm": 0.16374830901622772,
      "learning_rate": 2.7619836231484036e-05,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8415212370455265,
      "num_tokens": 753241338.0,
      "step": 23602
    },
    {
      "entropy": 0.5765002528205514,
      "epoch": 2.1715918827378795,
      "grad_norm": 0.16678258776664734,
      "learning_rate": 2.7616769405342414e-05,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8204906806349754,
      "num_tokens": 753272568.0,
      "step": 23603
    },
    {
      "entropy": 0.6231195833534002,
      "epoch": 2.1716838891093206,
      "grad_norm": 0.17088571190834045,
      "learning_rate": 2.7613702579200785e-05,
      "loss": 0.6147,
      "mean_token_accuracy": 0.8051037043333054,
      "num_tokens": 753304859.0,
      "step": 23604
    },
    {
      "entropy": 0.5844001602381468,
      "epoch": 2.171775895480762,
      "grad_norm": 0.16120746731758118,
      "learning_rate": 2.761063575305916e-05,
      "loss": 0.5686,
      "mean_token_accuracy": 0.8219721876084805,
      "num_tokens": 753337573.0,
      "step": 23605
    },
    {
      "entropy": 0.46525558922439814,
      "epoch": 2.1718679018522034,
      "grad_norm": 0.15840494632720947,
      "learning_rate": 2.760756892691753e-05,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8528558574616909,
      "num_tokens": 753369507.0,
      "step": 23606
    },
    {
      "entropy": 0.4502120604738593,
      "epoch": 2.1719599082236445,
      "grad_norm": 0.15458934009075165,
      "learning_rate": 2.760450210077591e-05,
      "loss": 0.4383,
      "mean_token_accuracy": 0.8612516112625599,
      "num_tokens": 753400947.0,
      "step": 23607
    },
    {
      "entropy": 0.5114734824746847,
      "epoch": 2.1720519145950856,
      "grad_norm": 0.1604742407798767,
      "learning_rate": 2.760143527463428e-05,
      "loss": 0.4914,
      "mean_token_accuracy": 0.8473717607557774,
      "num_tokens": 753433343.0,
      "step": 23608
    },
    {
      "entropy": 0.6462052799761295,
      "epoch": 2.1721439209665268,
      "grad_norm": 0.17214074730873108,
      "learning_rate": 2.7598368448492655e-05,
      "loss": 0.6305,
      "mean_token_accuracy": 0.8044315502047539,
      "num_tokens": 753465730.0,
      "step": 23609
    },
    {
      "entropy": 0.5325323008000851,
      "epoch": 2.1722359273379683,
      "grad_norm": 0.15632912516593933,
      "learning_rate": 2.7595301622351033e-05,
      "loss": 0.525,
      "mean_token_accuracy": 0.8335222825407982,
      "num_tokens": 753498192.0,
      "step": 23610
    },
    {
      "entropy": 0.5566730219870806,
      "epoch": 2.1723279337094095,
      "grad_norm": 0.1657901108264923,
      "learning_rate": 2.7592234796209405e-05,
      "loss": 0.5356,
      "mean_token_accuracy": 0.8324644081294537,
      "num_tokens": 753530778.0,
      "step": 23611
    },
    {
      "entropy": 0.556927272118628,
      "epoch": 2.1724199400808506,
      "grad_norm": 0.16577273607254028,
      "learning_rate": 2.758916797006778e-05,
      "loss": 0.5417,
      "mean_token_accuracy": 0.829172819852829,
      "num_tokens": 753562137.0,
      "step": 23612
    },
    {
      "entropy": 0.5212503843940794,
      "epoch": 2.1725119464522917,
      "grad_norm": 0.1642051488161087,
      "learning_rate": 2.758610114392615e-05,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8383208140730858,
      "num_tokens": 753594548.0,
      "step": 23613
    },
    {
      "entropy": 0.45521235186606646,
      "epoch": 2.172603952823733,
      "grad_norm": 0.15438783168792725,
      "learning_rate": 2.758303431778453e-05,
      "loss": 0.4426,
      "mean_token_accuracy": 0.8616663925349712,
      "num_tokens": 753626283.0,
      "step": 23614
    },
    {
      "entropy": 0.5498768445104361,
      "epoch": 2.1726959591951744,
      "grad_norm": 0.17022716999053955,
      "learning_rate": 2.75799674916429e-05,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8289072960615158,
      "num_tokens": 753657993.0,
      "step": 23615
    },
    {
      "entropy": 0.49401689670048654,
      "epoch": 2.1727879655666156,
      "grad_norm": 0.1500042974948883,
      "learning_rate": 2.7576900665501278e-05,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8507376946508884,
      "num_tokens": 753689841.0,
      "step": 23616
    },
    {
      "entropy": 0.5466302819550037,
      "epoch": 2.1728799719380567,
      "grad_norm": 0.16951067745685577,
      "learning_rate": 2.7573833839359646e-05,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8320077024400234,
      "num_tokens": 753721340.0,
      "step": 23617
    },
    {
      "entropy": 0.4181893840432167,
      "epoch": 2.172971978309498,
      "grad_norm": 0.15029001235961914,
      "learning_rate": 2.7570767013218024e-05,
      "loss": 0.3979,
      "mean_token_accuracy": 0.8740543127059937,
      "num_tokens": 753753705.0,
      "step": 23618
    },
    {
      "entropy": 0.5187822002917528,
      "epoch": 2.173063984680939,
      "grad_norm": 0.160674586892128,
      "learning_rate": 2.7567700187076395e-05,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8373040519654751,
      "num_tokens": 753784960.0,
      "step": 23619
    },
    {
      "entropy": 0.5751377055421472,
      "epoch": 2.1731559910523806,
      "grad_norm": 0.16703279316425323,
      "learning_rate": 2.7564633360934773e-05,
      "loss": 0.558,
      "mean_token_accuracy": 0.8223982527852058,
      "num_tokens": 753816394.0,
      "step": 23620
    },
    {
      "entropy": 0.42198326298967004,
      "epoch": 2.1732479974238217,
      "grad_norm": 0.15357249975204468,
      "learning_rate": 2.7561566534793144e-05,
      "loss": 0.4074,
      "mean_token_accuracy": 0.871946357190609,
      "num_tokens": 753848233.0,
      "step": 23621
    },
    {
      "entropy": 0.5169387515634298,
      "epoch": 2.173340003795263,
      "grad_norm": 0.16178731620311737,
      "learning_rate": 2.755849970865152e-05,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8397653102874756,
      "num_tokens": 753880055.0,
      "step": 23622
    },
    {
      "entropy": 0.5355425169691443,
      "epoch": 2.173432010166704,
      "grad_norm": 0.16514749825000763,
      "learning_rate": 2.755543288250989e-05,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8410477675497532,
      "num_tokens": 753911091.0,
      "step": 23623
    },
    {
      "entropy": 0.4867049914319068,
      "epoch": 2.173524016538145,
      "grad_norm": 0.15591517090797424,
      "learning_rate": 2.7552366056368268e-05,
      "loss": 0.4611,
      "mean_token_accuracy": 0.854348462074995,
      "num_tokens": 753943590.0,
      "step": 23624
    },
    {
      "entropy": 0.5500449566170573,
      "epoch": 2.1736160229095867,
      "grad_norm": 0.16047589480876923,
      "learning_rate": 2.754929923022664e-05,
      "loss": 0.5397,
      "mean_token_accuracy": 0.8296527154743671,
      "num_tokens": 753976293.0,
      "step": 23625
    },
    {
      "entropy": 0.45046062930487096,
      "epoch": 2.173708029281028,
      "grad_norm": 0.15236864984035492,
      "learning_rate": 2.7546232404085014e-05,
      "loss": 0.4392,
      "mean_token_accuracy": 0.8593600392341614,
      "num_tokens": 754008600.0,
      "step": 23626
    },
    {
      "entropy": 0.6412110682576895,
      "epoch": 2.173800035652469,
      "grad_norm": 0.18393243849277496,
      "learning_rate": 2.7543165577943385e-05,
      "loss": 0.6338,
      "mean_token_accuracy": 0.8057717494666576,
      "num_tokens": 754040329.0,
      "step": 23627
    },
    {
      "entropy": 0.5625120829790831,
      "epoch": 2.17389204202391,
      "grad_norm": 0.16405938565731049,
      "learning_rate": 2.7540098751801763e-05,
      "loss": 0.5686,
      "mean_token_accuracy": 0.8217721879482269,
      "num_tokens": 754072767.0,
      "step": 23628
    },
    {
      "entropy": 0.41654953407123685,
      "epoch": 2.173984048395351,
      "grad_norm": 0.15349501371383667,
      "learning_rate": 2.7537031925660135e-05,
      "loss": 0.3973,
      "mean_token_accuracy": 0.8696178607642651,
      "num_tokens": 754103396.0,
      "step": 23629
    },
    {
      "entropy": 0.48160365456715226,
      "epoch": 2.1740760547667928,
      "grad_norm": 0.15434473752975464,
      "learning_rate": 2.753396509951851e-05,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8531176671385765,
      "num_tokens": 754135631.0,
      "step": 23630
    },
    {
      "entropy": 0.4472106748726219,
      "epoch": 2.174168061138234,
      "grad_norm": 0.15512214601039886,
      "learning_rate": 2.753089827337688e-05,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8567681573331356,
      "num_tokens": 754167137.0,
      "step": 23631
    },
    {
      "entropy": 0.6431810557842255,
      "epoch": 2.174260067509675,
      "grad_norm": 0.1754617989063263,
      "learning_rate": 2.752783144723526e-05,
      "loss": 0.6276,
      "mean_token_accuracy": 0.801115732640028,
      "num_tokens": 754198836.0,
      "step": 23632
    },
    {
      "entropy": 0.4766820715740323,
      "epoch": 2.174352073881116,
      "grad_norm": 0.1534925103187561,
      "learning_rate": 2.752476462109363e-05,
      "loss": 0.4724,
      "mean_token_accuracy": 0.8492802046239376,
      "num_tokens": 754231153.0,
      "step": 23633
    },
    {
      "entropy": 0.46630226622801274,
      "epoch": 2.1744440802525573,
      "grad_norm": 0.15151071548461914,
      "learning_rate": 2.7521697794952005e-05,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8550120666623116,
      "num_tokens": 754263041.0,
      "step": 23634
    },
    {
      "entropy": 0.5031613353639841,
      "epoch": 2.174536086623999,
      "grad_norm": 0.16205881536006927,
      "learning_rate": 2.7518630968810376e-05,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8445482887327671,
      "num_tokens": 754295579.0,
      "step": 23635
    },
    {
      "entropy": 0.5471494179219007,
      "epoch": 2.17462809299544,
      "grad_norm": 0.16627241671085358,
      "learning_rate": 2.7515564142668754e-05,
      "loss": 0.5372,
      "mean_token_accuracy": 0.8275696784257889,
      "num_tokens": 754327510.0,
      "step": 23636
    },
    {
      "entropy": 0.5384451001882553,
      "epoch": 2.174720099366881,
      "grad_norm": 0.16546031832695007,
      "learning_rate": 2.7512497316527125e-05,
      "loss": 0.5242,
      "mean_token_accuracy": 0.8333648703992367,
      "num_tokens": 754359377.0,
      "step": 23637
    },
    {
      "entropy": 0.5277138622477651,
      "epoch": 2.1748121057383223,
      "grad_norm": 0.16126878559589386,
      "learning_rate": 2.7509430490385503e-05,
      "loss": 0.505,
      "mean_token_accuracy": 0.8414492085576057,
      "num_tokens": 754391008.0,
      "step": 23638
    },
    {
      "entropy": 0.5147755462676287,
      "epoch": 2.1749041121097634,
      "grad_norm": 0.15583324432373047,
      "learning_rate": 2.750636366424387e-05,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8412180580198765,
      "num_tokens": 754423164.0,
      "step": 23639
    },
    {
      "entropy": 0.42716762237250805,
      "epoch": 2.174996118481205,
      "grad_norm": 0.14755681157112122,
      "learning_rate": 2.750329683810225e-05,
      "loss": 0.4091,
      "mean_token_accuracy": 0.8695295788347721,
      "num_tokens": 754455901.0,
      "step": 23640
    },
    {
      "entropy": 0.6091063041239977,
      "epoch": 2.175088124852646,
      "grad_norm": 0.1702122837305069,
      "learning_rate": 2.7500230011960627e-05,
      "loss": 0.5907,
      "mean_token_accuracy": 0.8161290511488914,
      "num_tokens": 754487847.0,
      "step": 23641
    },
    {
      "entropy": 0.49998127296566963,
      "epoch": 2.1751801312240873,
      "grad_norm": 0.1587987244129181,
      "learning_rate": 2.7497163185819e-05,
      "loss": 0.4769,
      "mean_token_accuracy": 0.8510236293077469,
      "num_tokens": 754519798.0,
      "step": 23642
    },
    {
      "entropy": 0.5350671885535121,
      "epoch": 2.1752721375955284,
      "grad_norm": 0.16178199648857117,
      "learning_rate": 2.7494096359677373e-05,
      "loss": 0.5203,
      "mean_token_accuracy": 0.8338644504547119,
      "num_tokens": 754551787.0,
      "step": 23643
    },
    {
      "entropy": 0.5535961659625173,
      "epoch": 2.1753641439669695,
      "grad_norm": 0.16331814229488373,
      "learning_rate": 2.7491029533535744e-05,
      "loss": 0.5408,
      "mean_token_accuracy": 0.8308929055929184,
      "num_tokens": 754583806.0,
      "step": 23644
    },
    {
      "entropy": 0.4479782283306122,
      "epoch": 2.175456150338411,
      "grad_norm": 0.15667112171649933,
      "learning_rate": 2.7487962707394122e-05,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8602857030928135,
      "num_tokens": 754615957.0,
      "step": 23645
    },
    {
      "entropy": 0.40979383792728186,
      "epoch": 2.1755481567098522,
      "grad_norm": 0.14413221180438995,
      "learning_rate": 2.7484895881252494e-05,
      "loss": 0.398,
      "mean_token_accuracy": 0.8704741559922695,
      "num_tokens": 754648170.0,
      "step": 23646
    },
    {
      "entropy": 0.4250123738311231,
      "epoch": 2.1756401630812934,
      "grad_norm": 0.14771312475204468,
      "learning_rate": 2.7481829055110868e-05,
      "loss": 0.4028,
      "mean_token_accuracy": 0.869095928966999,
      "num_tokens": 754679515.0,
      "step": 23647
    },
    {
      "entropy": 0.5708408104255795,
      "epoch": 2.1757321694527345,
      "grad_norm": 0.17099963128566742,
      "learning_rate": 2.747876222896924e-05,
      "loss": 0.5639,
      "mean_token_accuracy": 0.8234424032270908,
      "num_tokens": 754711256.0,
      "step": 23648
    },
    {
      "entropy": 0.5312166027724743,
      "epoch": 2.1758241758241756,
      "grad_norm": 0.1649894118309021,
      "learning_rate": 2.7475695402827618e-05,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8355269134044647,
      "num_tokens": 754743500.0,
      "step": 23649
    },
    {
      "entropy": 0.492201242595911,
      "epoch": 2.175916182195617,
      "grad_norm": 0.1620604395866394,
      "learning_rate": 2.747262857668599e-05,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8428818359971046,
      "num_tokens": 754775684.0,
      "step": 23650
    },
    {
      "entropy": 0.5420641265809536,
      "epoch": 2.1760081885670584,
      "grad_norm": 0.17783626914024353,
      "learning_rate": 2.7469561750544363e-05,
      "loss": 0.5363,
      "mean_token_accuracy": 0.8343970887362957,
      "num_tokens": 754806734.0,
      "step": 23651
    },
    {
      "entropy": 0.4502787608653307,
      "epoch": 2.1761001949384995,
      "grad_norm": 0.1506386548280716,
      "learning_rate": 2.7466494924402735e-05,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8600744046270847,
      "num_tokens": 754839036.0,
      "step": 23652
    },
    {
      "entropy": 0.39660344552248716,
      "epoch": 2.1761922013099406,
      "grad_norm": 0.14800776541233063,
      "learning_rate": 2.7463428098261113e-05,
      "loss": 0.3891,
      "mean_token_accuracy": 0.8775628954172134,
      "num_tokens": 754870707.0,
      "step": 23653
    },
    {
      "entropy": 0.4999903244897723,
      "epoch": 2.1762842076813818,
      "grad_norm": 0.15992340445518494,
      "learning_rate": 2.7460361272119484e-05,
      "loss": 0.493,
      "mean_token_accuracy": 0.8423458561301231,
      "num_tokens": 754902108.0,
      "step": 23654
    },
    {
      "entropy": 0.4464752534404397,
      "epoch": 2.1763762140528233,
      "grad_norm": 0.15201565623283386,
      "learning_rate": 2.745729444597786e-05,
      "loss": 0.4417,
      "mean_token_accuracy": 0.8613238707184792,
      "num_tokens": 754933590.0,
      "step": 23655
    },
    {
      "entropy": 0.4675380880944431,
      "epoch": 2.1764682204242645,
      "grad_norm": 0.1610657423734665,
      "learning_rate": 2.745422761983623e-05,
      "loss": 0.4552,
      "mean_token_accuracy": 0.854716882109642,
      "num_tokens": 754964887.0,
      "step": 23656
    },
    {
      "entropy": 0.5554149858653545,
      "epoch": 2.1765602267957056,
      "grad_norm": 0.16063281893730164,
      "learning_rate": 2.7451160793694608e-05,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8300149701535702,
      "num_tokens": 754996969.0,
      "step": 23657
    },
    {
      "entropy": 0.6018376462161541,
      "epoch": 2.1766522331671467,
      "grad_norm": 0.17081329226493835,
      "learning_rate": 2.744809396755298e-05,
      "loss": 0.5867,
      "mean_token_accuracy": 0.8161436431109905,
      "num_tokens": 755029033.0,
      "step": 23658
    },
    {
      "entropy": 0.5190281253308058,
      "epoch": 2.176744239538588,
      "grad_norm": 0.15892699360847473,
      "learning_rate": 2.7445027141411357e-05,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8373443819582462,
      "num_tokens": 755060572.0,
      "step": 23659
    },
    {
      "entropy": 0.5660033859312534,
      "epoch": 2.1768362459100294,
      "grad_norm": 0.1609937697649002,
      "learning_rate": 2.744196031526973e-05,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8275868482887745,
      "num_tokens": 755092511.0,
      "step": 23660
    },
    {
      "entropy": 0.46296249493025243,
      "epoch": 2.1769282522814706,
      "grad_norm": 0.1536673903465271,
      "learning_rate": 2.7438893489128103e-05,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8565155789256096,
      "num_tokens": 755124450.0,
      "step": 23661
    },
    {
      "entropy": 0.4681877428665757,
      "epoch": 2.1770202586529117,
      "grad_norm": 0.15806370973587036,
      "learning_rate": 2.7435826662986474e-05,
      "loss": 0.4554,
      "mean_token_accuracy": 0.858518872410059,
      "num_tokens": 755156687.0,
      "step": 23662
    },
    {
      "entropy": 0.5482941279187799,
      "epoch": 2.177112265024353,
      "grad_norm": 0.15905188024044037,
      "learning_rate": 2.7432759836844852e-05,
      "loss": 0.5387,
      "mean_token_accuracy": 0.8295686431229115,
      "num_tokens": 755189318.0,
      "step": 23663
    },
    {
      "entropy": 0.5985401123762131,
      "epoch": 2.177204271395794,
      "grad_norm": 0.1756173074245453,
      "learning_rate": 2.7429693010703224e-05,
      "loss": 0.5859,
      "mean_token_accuracy": 0.8162735588848591,
      "num_tokens": 755220123.0,
      "step": 23664
    },
    {
      "entropy": 0.5263242721557617,
      "epoch": 2.1772962777672356,
      "grad_norm": 0.1681988537311554,
      "learning_rate": 2.74266261845616e-05,
      "loss": 0.5218,
      "mean_token_accuracy": 0.8347374685108662,
      "num_tokens": 755251736.0,
      "step": 23665
    },
    {
      "entropy": 0.5287044737488031,
      "epoch": 2.1773882841386767,
      "grad_norm": 0.1683071106672287,
      "learning_rate": 2.742355935841997e-05,
      "loss": 0.5307,
      "mean_token_accuracy": 0.831595990806818,
      "num_tokens": 755283062.0,
      "step": 23666
    },
    {
      "entropy": 0.43058846495114267,
      "epoch": 2.177480290510118,
      "grad_norm": 0.153971329331398,
      "learning_rate": 2.7420492532278348e-05,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8625807575881481,
      "num_tokens": 755314946.0,
      "step": 23667
    },
    {
      "entropy": 0.5244598165154457,
      "epoch": 2.177572296881559,
      "grad_norm": 0.16086459159851074,
      "learning_rate": 2.741742570613672e-05,
      "loss": 0.5137,
      "mean_token_accuracy": 0.8353030122816563,
      "num_tokens": 755347498.0,
      "step": 23668
    },
    {
      "entropy": 0.5394695568829775,
      "epoch": 2.177664303253,
      "grad_norm": 0.15550191700458527,
      "learning_rate": 2.7414358879995094e-05,
      "loss": 0.4938,
      "mean_token_accuracy": 0.8394173234701157,
      "num_tokens": 755379299.0,
      "step": 23669
    },
    {
      "entropy": 0.5387113862670958,
      "epoch": 2.1777563096244417,
      "grad_norm": 0.16594301164150238,
      "learning_rate": 2.7411292053853465e-05,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8275759406387806,
      "num_tokens": 755411749.0,
      "step": 23670
    },
    {
      "entropy": 0.5279688057489693,
      "epoch": 2.177848315995883,
      "grad_norm": 0.16463351249694824,
      "learning_rate": 2.7408225227711843e-05,
      "loss": 0.5096,
      "mean_token_accuracy": 0.8369005806744099,
      "num_tokens": 755443183.0,
      "step": 23671
    },
    {
      "entropy": 0.5555501184426248,
      "epoch": 2.177940322367324,
      "grad_norm": 0.16248950362205505,
      "learning_rate": 2.7405158401570217e-05,
      "loss": 0.5478,
      "mean_token_accuracy": 0.8281326293945312,
      "num_tokens": 755475951.0,
      "step": 23672
    },
    {
      "entropy": 0.5524555463343859,
      "epoch": 2.178032328738765,
      "grad_norm": 0.16789914667606354,
      "learning_rate": 2.740209157542859e-05,
      "loss": 0.5265,
      "mean_token_accuracy": 0.8321347944438457,
      "num_tokens": 755507373.0,
      "step": 23673
    },
    {
      "entropy": 0.4904445484280586,
      "epoch": 2.178124335110206,
      "grad_norm": 0.1594189554452896,
      "learning_rate": 2.7399024749286967e-05,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8505451157689095,
      "num_tokens": 755539375.0,
      "step": 23674
    },
    {
      "entropy": 0.5395829044282436,
      "epoch": 2.1782163414816478,
      "grad_norm": 0.1646948903799057,
      "learning_rate": 2.7395957923145338e-05,
      "loss": 0.5317,
      "mean_token_accuracy": 0.8333349302411079,
      "num_tokens": 755571809.0,
      "step": 23675
    },
    {
      "entropy": 0.3999572671018541,
      "epoch": 2.178308347853089,
      "grad_norm": 0.1482541412115097,
      "learning_rate": 2.7392891097003716e-05,
      "loss": 0.3897,
      "mean_token_accuracy": 0.8761523254215717,
      "num_tokens": 755604339.0,
      "step": 23676
    },
    {
      "entropy": 0.541537316981703,
      "epoch": 2.17840035422453,
      "grad_norm": 0.162479966878891,
      "learning_rate": 2.7389824270862084e-05,
      "loss": 0.5322,
      "mean_token_accuracy": 0.8295347429811954,
      "num_tokens": 755635335.0,
      "step": 23677
    },
    {
      "entropy": 0.5000668356660753,
      "epoch": 2.178492360595971,
      "grad_norm": 0.15999482572078705,
      "learning_rate": 2.7386757444720462e-05,
      "loss": 0.4908,
      "mean_token_accuracy": 0.8417363539338112,
      "num_tokens": 755666615.0,
      "step": 23678
    },
    {
      "entropy": 0.5144117884337902,
      "epoch": 2.1785843669674123,
      "grad_norm": 0.16529685258865356,
      "learning_rate": 2.7383690618578833e-05,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8397382944822311,
      "num_tokens": 755698532.0,
      "step": 23679
    },
    {
      "entropy": 0.5023857653141022,
      "epoch": 2.178676373338854,
      "grad_norm": 0.1580500304698944,
      "learning_rate": 2.738062379243721e-05,
      "loss": 0.4794,
      "mean_token_accuracy": 0.8445457816123962,
      "num_tokens": 755730601.0,
      "step": 23680
    },
    {
      "entropy": 0.4384928997606039,
      "epoch": 2.178768379710295,
      "grad_norm": 0.1494482159614563,
      "learning_rate": 2.7377556966295583e-05,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8640131056308746,
      "num_tokens": 755763073.0,
      "step": 23681
    },
    {
      "entropy": 0.5319747272878885,
      "epoch": 2.178860386081736,
      "grad_norm": 0.1652834564447403,
      "learning_rate": 2.7374490140153957e-05,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8417803719639778,
      "num_tokens": 755795322.0,
      "step": 23682
    },
    {
      "entropy": 0.5728955660015345,
      "epoch": 2.1789523924531773,
      "grad_norm": 0.16661016643047333,
      "learning_rate": 2.737142331401233e-05,
      "loss": 0.5536,
      "mean_token_accuracy": 0.8268703892827034,
      "num_tokens": 755826730.0,
      "step": 23683
    },
    {
      "entropy": 0.6129090080503374,
      "epoch": 2.1790443988246184,
      "grad_norm": 0.17296428978443146,
      "learning_rate": 2.7368356487870706e-05,
      "loss": 0.593,
      "mean_token_accuracy": 0.8122782409191132,
      "num_tokens": 755858752.0,
      "step": 23684
    },
    {
      "entropy": 0.5951065733097494,
      "epoch": 2.17913640519606,
      "grad_norm": 0.17193515598773956,
      "learning_rate": 2.7365289661729078e-05,
      "loss": 0.5713,
      "mean_token_accuracy": 0.8203265033662319,
      "num_tokens": 755890527.0,
      "step": 23685
    },
    {
      "entropy": 0.5120431836694479,
      "epoch": 2.179228411567501,
      "grad_norm": 0.15959133207798004,
      "learning_rate": 2.7362222835587452e-05,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8372985124588013,
      "num_tokens": 755922704.0,
      "step": 23686
    },
    {
      "entropy": 0.6011253856122494,
      "epoch": 2.1793204179389423,
      "grad_norm": 0.17335490882396698,
      "learning_rate": 2.7359156009445824e-05,
      "loss": 0.5851,
      "mean_token_accuracy": 0.8156358599662781,
      "num_tokens": 755953794.0,
      "step": 23687
    },
    {
      "entropy": 0.41471561044454575,
      "epoch": 2.1794124243103834,
      "grad_norm": 0.14206677675247192,
      "learning_rate": 2.73560891833042e-05,
      "loss": 0.4051,
      "mean_token_accuracy": 0.871703814715147,
      "num_tokens": 755985794.0,
      "step": 23688
    },
    {
      "entropy": 0.5063211154192686,
      "epoch": 2.1795044306818245,
      "grad_norm": 0.15711276233196259,
      "learning_rate": 2.7353022357162573e-05,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8476003371179104,
      "num_tokens": 756018111.0,
      "step": 23689
    },
    {
      "entropy": 0.4189927885308862,
      "epoch": 2.179596437053266,
      "grad_norm": 0.1475890874862671,
      "learning_rate": 2.7349955531020948e-05,
      "loss": 0.4022,
      "mean_token_accuracy": 0.8710589148104191,
      "num_tokens": 756050680.0,
      "step": 23690
    },
    {
      "entropy": 0.5966987144201994,
      "epoch": 2.1796884434247072,
      "grad_norm": 0.16854235529899597,
      "learning_rate": 2.734688870487932e-05,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8203394822776318,
      "num_tokens": 756082040.0,
      "step": 23691
    },
    {
      "entropy": 0.5516112968325615,
      "epoch": 2.1797804497961484,
      "grad_norm": 0.1707259863615036,
      "learning_rate": 2.7343821878737697e-05,
      "loss": 0.5452,
      "mean_token_accuracy": 0.8279221653938293,
      "num_tokens": 756113566.0,
      "step": 23692
    },
    {
      "entropy": 0.5881625171750784,
      "epoch": 2.1798724561675895,
      "grad_norm": 0.17093081772327423,
      "learning_rate": 2.7340755052596068e-05,
      "loss": 0.5865,
      "mean_token_accuracy": 0.8143822178244591,
      "num_tokens": 756144398.0,
      "step": 23693
    },
    {
      "entropy": 0.5076936467085034,
      "epoch": 2.1799644625390306,
      "grad_norm": 0.16094791889190674,
      "learning_rate": 2.7337688226454443e-05,
      "loss": 0.4949,
      "mean_token_accuracy": 0.8444409146904945,
      "num_tokens": 756175898.0,
      "step": 23694
    },
    {
      "entropy": 0.4480241434648633,
      "epoch": 2.1800564689104722,
      "grad_norm": 0.1541784703731537,
      "learning_rate": 2.7334621400312814e-05,
      "loss": 0.434,
      "mean_token_accuracy": 0.8615749999880791,
      "num_tokens": 756207136.0,
      "step": 23695
    },
    {
      "entropy": 0.603094672318548,
      "epoch": 2.1801484752819134,
      "grad_norm": 0.1727157086133957,
      "learning_rate": 2.7331554574171192e-05,
      "loss": 0.6062,
      "mean_token_accuracy": 0.8084302917122841,
      "num_tokens": 756238882.0,
      "step": 23696
    },
    {
      "entropy": 0.6289227586239576,
      "epoch": 2.1802404816533545,
      "grad_norm": 0.17701318860054016,
      "learning_rate": 2.7328487748029563e-05,
      "loss": 0.6321,
      "mean_token_accuracy": 0.804922204464674,
      "num_tokens": 756270703.0,
      "step": 23697
    },
    {
      "entropy": 0.4672899632714689,
      "epoch": 2.1803324880247956,
      "grad_norm": 0.15596912801265717,
      "learning_rate": 2.732542092188794e-05,
      "loss": 0.4587,
      "mean_token_accuracy": 0.855700183659792,
      "num_tokens": 756303047.0,
      "step": 23698
    },
    {
      "entropy": 0.576046384871006,
      "epoch": 2.1804244943962368,
      "grad_norm": 0.16864895820617676,
      "learning_rate": 2.732235409574631e-05,
      "loss": 0.5606,
      "mean_token_accuracy": 0.8231163844466209,
      "num_tokens": 756335611.0,
      "step": 23699
    },
    {
      "entropy": 0.4583329767920077,
      "epoch": 2.1805165007676783,
      "grad_norm": 0.15175704658031464,
      "learning_rate": 2.7319287269604687e-05,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8590573817491531,
      "num_tokens": 756368205.0,
      "step": 23700
    },
    {
      "entropy": 0.542741796001792,
      "epoch": 2.1806085071391195,
      "grad_norm": 0.15886889398097992,
      "learning_rate": 2.731622044346306e-05,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8320648707449436,
      "num_tokens": 756400549.0,
      "step": 23701
    },
    {
      "entropy": 0.4863789523951709,
      "epoch": 2.1807005135105606,
      "grad_norm": 0.15532748401165009,
      "learning_rate": 2.7313153617321437e-05,
      "loss": 0.4753,
      "mean_token_accuracy": 0.8474473431706429,
      "num_tokens": 756432478.0,
      "step": 23702
    },
    {
      "entropy": 0.5284036789089441,
      "epoch": 2.1807925198820017,
      "grad_norm": 0.16049398481845856,
      "learning_rate": 2.731008679117981e-05,
      "loss": 0.512,
      "mean_token_accuracy": 0.8357553370296955,
      "num_tokens": 756464547.0,
      "step": 23703
    },
    {
      "entropy": 0.47482461854815483,
      "epoch": 2.180884526253443,
      "grad_norm": 0.15597736835479736,
      "learning_rate": 2.7307019965038182e-05,
      "loss": 0.4704,
      "mean_token_accuracy": 0.8523507043719292,
      "num_tokens": 756496949.0,
      "step": 23704
    },
    {
      "entropy": 0.5005776295438409,
      "epoch": 2.1809765326248844,
      "grad_norm": 0.15983779728412628,
      "learning_rate": 2.730395313889656e-05,
      "loss": 0.4808,
      "mean_token_accuracy": 0.8476321287453175,
      "num_tokens": 756529054.0,
      "step": 23705
    },
    {
      "entropy": 0.6028060503304005,
      "epoch": 2.1810685389963256,
      "grad_norm": 0.16732577979564667,
      "learning_rate": 2.7300886312754932e-05,
      "loss": 0.591,
      "mean_token_accuracy": 0.8150758519768715,
      "num_tokens": 756561285.0,
      "step": 23706
    },
    {
      "entropy": 0.49774435721337795,
      "epoch": 2.1811605453677667,
      "grad_norm": 0.1566672921180725,
      "learning_rate": 2.7297819486613306e-05,
      "loss": 0.4747,
      "mean_token_accuracy": 0.8452741652727127,
      "num_tokens": 756593379.0,
      "step": 23707
    },
    {
      "entropy": 0.5555048389360309,
      "epoch": 2.181252551739208,
      "grad_norm": 0.16318273544311523,
      "learning_rate": 2.7294752660471678e-05,
      "loss": 0.5453,
      "mean_token_accuracy": 0.8321098946034908,
      "num_tokens": 756625771.0,
      "step": 23708
    },
    {
      "entropy": 0.5829039253294468,
      "epoch": 2.181344558110649,
      "grad_norm": 0.16742157936096191,
      "learning_rate": 2.7291685834330056e-05,
      "loss": 0.5724,
      "mean_token_accuracy": 0.8229180090129375,
      "num_tokens": 756658184.0,
      "step": 23709
    },
    {
      "entropy": 0.5194876492023468,
      "epoch": 2.1814365644820906,
      "grad_norm": 0.1624974012374878,
      "learning_rate": 2.7288619008188427e-05,
      "loss": 0.5135,
      "mean_token_accuracy": 0.8343985415995121,
      "num_tokens": 756690497.0,
      "step": 23710
    },
    {
      "entropy": 0.3808723045513034,
      "epoch": 2.1815285708535317,
      "grad_norm": 0.14883989095687866,
      "learning_rate": 2.72855521820468e-05,
      "loss": 0.3716,
      "mean_token_accuracy": 0.8801658749580383,
      "num_tokens": 756722072.0,
      "step": 23711
    },
    {
      "entropy": 0.4911482797469944,
      "epoch": 2.181620577224973,
      "grad_norm": 0.16020487248897552,
      "learning_rate": 2.7282485355905173e-05,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8491251021623611,
      "num_tokens": 756753905.0,
      "step": 23712
    },
    {
      "entropy": 0.5430525187402964,
      "epoch": 2.181712583596414,
      "grad_norm": 0.16101200878620148,
      "learning_rate": 2.727941852976355e-05,
      "loss": 0.5317,
      "mean_token_accuracy": 0.8294871039688587,
      "num_tokens": 756786002.0,
      "step": 23713
    },
    {
      "entropy": 0.5730591993778944,
      "epoch": 2.181804589967855,
      "grad_norm": 0.17002741992473602,
      "learning_rate": 2.7276351703621922e-05,
      "loss": 0.5701,
      "mean_token_accuracy": 0.8209846504032612,
      "num_tokens": 756817844.0,
      "step": 23714
    },
    {
      "entropy": 0.597935794852674,
      "epoch": 2.1818965963392967,
      "grad_norm": 0.17075107991695404,
      "learning_rate": 2.7273284877480297e-05,
      "loss": 0.5892,
      "mean_token_accuracy": 0.8148635923862457,
      "num_tokens": 756849454.0,
      "step": 23715
    },
    {
      "entropy": 0.5994333205744624,
      "epoch": 2.181988602710738,
      "grad_norm": 0.17020195722579956,
      "learning_rate": 2.7270218051338668e-05,
      "loss": 0.5812,
      "mean_token_accuracy": 0.8154087066650391,
      "num_tokens": 756881945.0,
      "step": 23716
    },
    {
      "entropy": 0.5848719058558345,
      "epoch": 2.182080609082179,
      "grad_norm": 0.1735139638185501,
      "learning_rate": 2.7267151225197046e-05,
      "loss": 0.5746,
      "mean_token_accuracy": 0.8220936618745327,
      "num_tokens": 756913772.0,
      "step": 23717
    },
    {
      "entropy": 0.5142320338636637,
      "epoch": 2.18217261545362,
      "grad_norm": 0.15754184126853943,
      "learning_rate": 2.7264084399055417e-05,
      "loss": 0.4841,
      "mean_token_accuracy": 0.8472132347524166,
      "num_tokens": 756945229.0,
      "step": 23718
    },
    {
      "entropy": 0.5277885822579265,
      "epoch": 2.182264621825061,
      "grad_norm": 0.1618301421403885,
      "learning_rate": 2.7261017572913795e-05,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8399118334054947,
      "num_tokens": 756977474.0,
      "step": 23719
    },
    {
      "entropy": 0.5600571124814451,
      "epoch": 2.182356628196503,
      "grad_norm": 0.1602032333612442,
      "learning_rate": 2.7257950746772167e-05,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8326543234288692,
      "num_tokens": 757009495.0,
      "step": 23720
    },
    {
      "entropy": 0.4860941111110151,
      "epoch": 2.182448634567944,
      "grad_norm": 0.15083695948123932,
      "learning_rate": 2.725488392063054e-05,
      "loss": 0.468,
      "mean_token_accuracy": 0.8519751951098442,
      "num_tokens": 757041432.0,
      "step": 23721
    },
    {
      "entropy": 0.5186964082531631,
      "epoch": 2.182540640939385,
      "grad_norm": 0.16653591394424438,
      "learning_rate": 2.7251817094488913e-05,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8395251892507076,
      "num_tokens": 757072876.0,
      "step": 23722
    },
    {
      "entropy": 0.48088933038525283,
      "epoch": 2.182632647310826,
      "grad_norm": 0.1582144796848297,
      "learning_rate": 2.724875026834729e-05,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8496546484529972,
      "num_tokens": 757105449.0,
      "step": 23723
    },
    {
      "entropy": 0.5091195870190859,
      "epoch": 2.1827246536822673,
      "grad_norm": 0.16662262380123138,
      "learning_rate": 2.7245683442205662e-05,
      "loss": 0.4943,
      "mean_token_accuracy": 0.8417821303009987,
      "num_tokens": 757138053.0,
      "step": 23724
    },
    {
      "entropy": 0.4593860500026494,
      "epoch": 2.182816660053709,
      "grad_norm": 0.15090414881706238,
      "learning_rate": 2.7242616616064037e-05,
      "loss": 0.4352,
      "mean_token_accuracy": 0.8601684309542179,
      "num_tokens": 757170267.0,
      "step": 23725
    },
    {
      "entropy": 0.6088556349277496,
      "epoch": 2.18290866642515,
      "grad_norm": 0.17299504578113556,
      "learning_rate": 2.7239549789922408e-05,
      "loss": 0.5887,
      "mean_token_accuracy": 0.8171259984374046,
      "num_tokens": 757202382.0,
      "step": 23726
    },
    {
      "entropy": 0.451709499116987,
      "epoch": 2.183000672796591,
      "grad_norm": 0.14545489847660065,
      "learning_rate": 2.7236482963780786e-05,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8637298941612244,
      "num_tokens": 757235003.0,
      "step": 23727
    },
    {
      "entropy": 0.4784110737964511,
      "epoch": 2.1830926791680323,
      "grad_norm": 0.1559625118970871,
      "learning_rate": 2.7233416137639157e-05,
      "loss": 0.4622,
      "mean_token_accuracy": 0.8507075533270836,
      "num_tokens": 757266999.0,
      "step": 23728
    },
    {
      "entropy": 0.5713079175911844,
      "epoch": 2.1831846855394734,
      "grad_norm": 0.16390889883041382,
      "learning_rate": 2.7230349311497532e-05,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8254763036966324,
      "num_tokens": 757299767.0,
      "step": 23729
    },
    {
      "entropy": 0.48929715156555176,
      "epoch": 2.183276691910915,
      "grad_norm": 0.15925459563732147,
      "learning_rate": 2.7227282485355903e-05,
      "loss": 0.4803,
      "mean_token_accuracy": 0.8481694236397743,
      "num_tokens": 757331860.0,
      "step": 23730
    },
    {
      "entropy": 0.5413746880367398,
      "epoch": 2.183368698282356,
      "grad_norm": 0.16103363037109375,
      "learning_rate": 2.722421565921428e-05,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8330196179449558,
      "num_tokens": 757364040.0,
      "step": 23731
    },
    {
      "entropy": 0.5330888647586107,
      "epoch": 2.1834607046537973,
      "grad_norm": 0.16140414774417877,
      "learning_rate": 2.7221148833072656e-05,
      "loss": 0.5242,
      "mean_token_accuracy": 0.8339665681123734,
      "num_tokens": 757395310.0,
      "step": 23732
    },
    {
      "entropy": 0.5151425404474139,
      "epoch": 2.1835527110252384,
      "grad_norm": 0.16652099788188934,
      "learning_rate": 2.7218082006931027e-05,
      "loss": 0.5004,
      "mean_token_accuracy": 0.8402676694095135,
      "num_tokens": 757427219.0,
      "step": 23733
    },
    {
      "entropy": 0.4931772309355438,
      "epoch": 2.1836447173966795,
      "grad_norm": 0.16293372213840485,
      "learning_rate": 2.7215015180789405e-05,
      "loss": 0.4811,
      "mean_token_accuracy": 0.8474881686270237,
      "num_tokens": 757459307.0,
      "step": 23734
    },
    {
      "entropy": 0.50099135376513,
      "epoch": 2.183736723768121,
      "grad_norm": 0.16221080720424652,
      "learning_rate": 2.7211948354647776e-05,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8416542746126652,
      "num_tokens": 757491359.0,
      "step": 23735
    },
    {
      "entropy": 0.4609315637499094,
      "epoch": 2.1838287301395622,
      "grad_norm": 0.1585645228624344,
      "learning_rate": 2.7208881528506154e-05,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8586165979504585,
      "num_tokens": 757523380.0,
      "step": 23736
    },
    {
      "entropy": 0.5385189331136644,
      "epoch": 2.1839207365110034,
      "grad_norm": 0.16594937443733215,
      "learning_rate": 2.7205814702364522e-05,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8337184153497219,
      "num_tokens": 757555613.0,
      "step": 23737
    },
    {
      "entropy": 0.5175488889217377,
      "epoch": 2.1840127428824445,
      "grad_norm": 0.1622328907251358,
      "learning_rate": 2.72027478762229e-05,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8419675454497337,
      "num_tokens": 757588009.0,
      "step": 23738
    },
    {
      "entropy": 0.5283892694860697,
      "epoch": 2.1841047492538856,
      "grad_norm": 0.1650855839252472,
      "learning_rate": 2.719968105008127e-05,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8396175615489483,
      "num_tokens": 757619933.0,
      "step": 23739
    },
    {
      "entropy": 0.534344166982919,
      "epoch": 2.1841967556253272,
      "grad_norm": 0.16366459429264069,
      "learning_rate": 2.719661422393965e-05,
      "loss": 0.5206,
      "mean_token_accuracy": 0.8361436538398266,
      "num_tokens": 757652021.0,
      "step": 23740
    },
    {
      "entropy": 0.5384791204705834,
      "epoch": 2.1842887619967684,
      "grad_norm": 0.15946906805038452,
      "learning_rate": 2.719354739779802e-05,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8316628709435463,
      "num_tokens": 757684265.0,
      "step": 23741
    },
    {
      "entropy": 0.47322862688452005,
      "epoch": 2.1843807683682095,
      "grad_norm": 0.1595618575811386,
      "learning_rate": 2.7190480571656395e-05,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8520045839250088,
      "num_tokens": 757715847.0,
      "step": 23742
    },
    {
      "entropy": 0.5500846542418003,
      "epoch": 2.1844727747396506,
      "grad_norm": 0.16215692460536957,
      "learning_rate": 2.7187413745514767e-05,
      "loss": 0.5347,
      "mean_token_accuracy": 0.8325664810836315,
      "num_tokens": 757748119.0,
      "step": 23743
    },
    {
      "entropy": 0.4934967001900077,
      "epoch": 2.1845647811110918,
      "grad_norm": 0.15479157865047455,
      "learning_rate": 2.7184346919373145e-05,
      "loss": 0.4831,
      "mean_token_accuracy": 0.8484692610800266,
      "num_tokens": 757780199.0,
      "step": 23744
    },
    {
      "entropy": 0.5398937575519085,
      "epoch": 2.1846567874825333,
      "grad_norm": 0.16216106712818146,
      "learning_rate": 2.7181280093231516e-05,
      "loss": 0.5399,
      "mean_token_accuracy": 0.8307509906589985,
      "num_tokens": 757812710.0,
      "step": 23745
    },
    {
      "entropy": 0.6010965956375003,
      "epoch": 2.1847487938539745,
      "grad_norm": 0.16827607154846191,
      "learning_rate": 2.717821326708989e-05,
      "loss": 0.5806,
      "mean_token_accuracy": 0.8153496608138084,
      "num_tokens": 757844596.0,
      "step": 23746
    },
    {
      "entropy": 0.5676482012495399,
      "epoch": 2.1848408002254156,
      "grad_norm": 0.16435697674751282,
      "learning_rate": 2.7175146440948262e-05,
      "loss": 0.5543,
      "mean_token_accuracy": 0.8243623860180378,
      "num_tokens": 757876770.0,
      "step": 23747
    },
    {
      "entropy": 0.5399671280756593,
      "epoch": 2.1849328065968567,
      "grad_norm": 0.16695153713226318,
      "learning_rate": 2.717207961480664e-05,
      "loss": 0.5221,
      "mean_token_accuracy": 0.8346253708004951,
      "num_tokens": 757909365.0,
      "step": 23748
    },
    {
      "entropy": 0.5914628067985177,
      "epoch": 2.185024812968298,
      "grad_norm": 0.1654667854309082,
      "learning_rate": 2.716901278866501e-05,
      "loss": 0.5805,
      "mean_token_accuracy": 0.817086536437273,
      "num_tokens": 757941211.0,
      "step": 23749
    },
    {
      "entropy": 0.5942269591614604,
      "epoch": 2.1851168193397394,
      "grad_norm": 0.16724610328674316,
      "learning_rate": 2.7165945962523386e-05,
      "loss": 0.575,
      "mean_token_accuracy": 0.8166666254401207,
      "num_tokens": 757973503.0,
      "step": 23750
    },
    {
      "entropy": 0.5887294104322791,
      "epoch": 2.1852088257111806,
      "grad_norm": 0.16293290257453918,
      "learning_rate": 2.7162879136381757e-05,
      "loss": 0.567,
      "mean_token_accuracy": 0.8207765519618988,
      "num_tokens": 758005800.0,
      "step": 23751
    },
    {
      "entropy": 0.5187226086854935,
      "epoch": 2.1853008320826217,
      "grad_norm": 0.16725029051303864,
      "learning_rate": 2.7159812310240135e-05,
      "loss": 0.5119,
      "mean_token_accuracy": 0.835159782320261,
      "num_tokens": 758037601.0,
      "step": 23752
    },
    {
      "entropy": 0.5328169642016292,
      "epoch": 2.185392838454063,
      "grad_norm": 0.1685868501663208,
      "learning_rate": 2.7156745484098506e-05,
      "loss": 0.5196,
      "mean_token_accuracy": 0.8363794833421707,
      "num_tokens": 758069633.0,
      "step": 23753
    },
    {
      "entropy": 0.5469092046841979,
      "epoch": 2.185484844825504,
      "grad_norm": 0.16068069636821747,
      "learning_rate": 2.715367865795688e-05,
      "loss": 0.5386,
      "mean_token_accuracy": 0.8301772326231003,
      "num_tokens": 758101752.0,
      "step": 23754
    },
    {
      "entropy": 0.6085434313863516,
      "epoch": 2.1855768511969456,
      "grad_norm": 0.1719134896993637,
      "learning_rate": 2.7150611831815252e-05,
      "loss": 0.5955,
      "mean_token_accuracy": 0.8108835183084011,
      "num_tokens": 758133038.0,
      "step": 23755
    },
    {
      "entropy": 0.4544774852693081,
      "epoch": 2.1856688575683867,
      "grad_norm": 0.16152223944664001,
      "learning_rate": 2.714754500567363e-05,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8558289557695389,
      "num_tokens": 758164242.0,
      "step": 23756
    },
    {
      "entropy": 0.49539794120937586,
      "epoch": 2.185760863939828,
      "grad_norm": 0.16023893654346466,
      "learning_rate": 2.7144478179532e-05,
      "loss": 0.4846,
      "mean_token_accuracy": 0.8434996418654919,
      "num_tokens": 758196243.0,
      "step": 23757
    },
    {
      "entropy": 0.5244387537240982,
      "epoch": 2.185852870311269,
      "grad_norm": 0.16059288382530212,
      "learning_rate": 2.714141135339038e-05,
      "loss": 0.5044,
      "mean_token_accuracy": 0.8378670141100883,
      "num_tokens": 758227858.0,
      "step": 23758
    },
    {
      "entropy": 0.532851202879101,
      "epoch": 2.18594487668271,
      "grad_norm": 0.1603340208530426,
      "learning_rate": 2.7138344527248747e-05,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8359714075922966,
      "num_tokens": 758259472.0,
      "step": 23759
    },
    {
      "entropy": 0.5132010020315647,
      "epoch": 2.1860368830541517,
      "grad_norm": 0.1647665798664093,
      "learning_rate": 2.7135277701107125e-05,
      "loss": 0.5111,
      "mean_token_accuracy": 0.8382012359797955,
      "num_tokens": 758291817.0,
      "step": 23760
    },
    {
      "entropy": 0.5900923488661647,
      "epoch": 2.186128889425593,
      "grad_norm": 0.16724541783332825,
      "learning_rate": 2.7132210874965497e-05,
      "loss": 0.5761,
      "mean_token_accuracy": 0.8214148841798306,
      "num_tokens": 758323489.0,
      "step": 23761
    },
    {
      "entropy": 0.5777246858924627,
      "epoch": 2.186220895797034,
      "grad_norm": 0.17169678211212158,
      "learning_rate": 2.7129144048823875e-05,
      "loss": 0.5634,
      "mean_token_accuracy": 0.8190270774066448,
      "num_tokens": 758355213.0,
      "step": 23762
    },
    {
      "entropy": 0.3969243075698614,
      "epoch": 2.186312902168475,
      "grad_norm": 0.14326122403144836,
      "learning_rate": 2.712607722268225e-05,
      "loss": 0.3869,
      "mean_token_accuracy": 0.8723056577146053,
      "num_tokens": 758387665.0,
      "step": 23763
    },
    {
      "entropy": 0.5617887144908309,
      "epoch": 2.186404908539916,
      "grad_norm": 0.16499751806259155,
      "learning_rate": 2.712301039654062e-05,
      "loss": 0.5574,
      "mean_token_accuracy": 0.8224309608340263,
      "num_tokens": 758420010.0,
      "step": 23764
    },
    {
      "entropy": 0.47784499218687415,
      "epoch": 2.186496914911358,
      "grad_norm": 0.15363676846027374,
      "learning_rate": 2.7119943570399e-05,
      "loss": 0.4607,
      "mean_token_accuracy": 0.8526087142527103,
      "num_tokens": 758452506.0,
      "step": 23765
    },
    {
      "entropy": 0.5385084487497807,
      "epoch": 2.186588921282799,
      "grad_norm": 0.16274559497833252,
      "learning_rate": 2.711687674425737e-05,
      "loss": 0.5217,
      "mean_token_accuracy": 0.8334378972649574,
      "num_tokens": 758484578.0,
      "step": 23766
    },
    {
      "entropy": 0.4771590167656541,
      "epoch": 2.18668092765424,
      "grad_norm": 0.15859480202198029,
      "learning_rate": 2.7113809918115745e-05,
      "loss": 0.4567,
      "mean_token_accuracy": 0.852822732180357,
      "num_tokens": 758516251.0,
      "step": 23767
    },
    {
      "entropy": 0.3456671244930476,
      "epoch": 2.186772934025681,
      "grad_norm": 0.13526275753974915,
      "learning_rate": 2.7110743091974116e-05,
      "loss": 0.3312,
      "mean_token_accuracy": 0.8930121175944805,
      "num_tokens": 758548635.0,
      "step": 23768
    },
    {
      "entropy": 0.5031196111813188,
      "epoch": 2.1868649403971223,
      "grad_norm": 0.1598338931798935,
      "learning_rate": 2.7107676265832494e-05,
      "loss": 0.4841,
      "mean_token_accuracy": 0.8433161601424217,
      "num_tokens": 758580003.0,
      "step": 23769
    },
    {
      "entropy": 0.5670165481278673,
      "epoch": 2.186956946768564,
      "grad_norm": 0.1642935574054718,
      "learning_rate": 2.7104609439690865e-05,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8267302848398685,
      "num_tokens": 758612016.0,
      "step": 23770
    },
    {
      "entropy": 0.5395582262426615,
      "epoch": 2.187048953140005,
      "grad_norm": 0.16523779928684235,
      "learning_rate": 2.710154261354924e-05,
      "loss": 0.528,
      "mean_token_accuracy": 0.8327017650008202,
      "num_tokens": 758643334.0,
      "step": 23771
    },
    {
      "entropy": 0.48379847314208746,
      "epoch": 2.187140959511446,
      "grad_norm": 0.1562117338180542,
      "learning_rate": 2.709847578740761e-05,
      "loss": 0.4663,
      "mean_token_accuracy": 0.8480079956352711,
      "num_tokens": 758675132.0,
      "step": 23772
    },
    {
      "entropy": 0.4955163775011897,
      "epoch": 2.1872329658828873,
      "grad_norm": 0.16332295536994934,
      "learning_rate": 2.709540896126599e-05,
      "loss": 0.4895,
      "mean_token_accuracy": 0.844387698918581,
      "num_tokens": 758707720.0,
      "step": 23773
    },
    {
      "entropy": 0.44267973862588406,
      "epoch": 2.1873249722543284,
      "grad_norm": 0.15093153715133667,
      "learning_rate": 2.709234213512436e-05,
      "loss": 0.4354,
      "mean_token_accuracy": 0.8608657233417034,
      "num_tokens": 758739803.0,
      "step": 23774
    },
    {
      "entropy": 0.5947998724877834,
      "epoch": 2.18741697862577,
      "grad_norm": 0.16730555891990662,
      "learning_rate": 2.7089275308982735e-05,
      "loss": 0.5813,
      "mean_token_accuracy": 0.8169506378471851,
      "num_tokens": 758771978.0,
      "step": 23775
    },
    {
      "entropy": 0.5383327398449183,
      "epoch": 2.187508984997211,
      "grad_norm": 0.16515058279037476,
      "learning_rate": 2.7086208482841106e-05,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8343851380050182,
      "num_tokens": 758804438.0,
      "step": 23776
    },
    {
      "entropy": 0.521900984691456,
      "epoch": 2.1876009913686523,
      "grad_norm": 0.16535314917564392,
      "learning_rate": 2.7083141656699484e-05,
      "loss": 0.5186,
      "mean_token_accuracy": 0.8376488015055656,
      "num_tokens": 758836700.0,
      "step": 23777
    },
    {
      "entropy": 0.4144380586221814,
      "epoch": 2.1876929977400934,
      "grad_norm": 0.14870786666870117,
      "learning_rate": 2.7080074830557856e-05,
      "loss": 0.408,
      "mean_token_accuracy": 0.8694970794022083,
      "num_tokens": 758869194.0,
      "step": 23778
    },
    {
      "entropy": 0.54081907030195,
      "epoch": 2.1877850041115345,
      "grad_norm": 0.1664194017648697,
      "learning_rate": 2.7077008004416234e-05,
      "loss": 0.5267,
      "mean_token_accuracy": 0.830532617866993,
      "num_tokens": 758901421.0,
      "step": 23779
    },
    {
      "entropy": 0.5786362625658512,
      "epoch": 2.187877010482976,
      "grad_norm": 0.16970284283161163,
      "learning_rate": 2.7073941178274605e-05,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8199919648468494,
      "num_tokens": 758933149.0,
      "step": 23780
    },
    {
      "entropy": 0.47469058353453875,
      "epoch": 2.1879690168544172,
      "grad_norm": 0.15678396821022034,
      "learning_rate": 2.707087435213298e-05,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8518393710255623,
      "num_tokens": 758964818.0,
      "step": 23781
    },
    {
      "entropy": 0.5425951573997736,
      "epoch": 2.1880610232258584,
      "grad_norm": 0.16428102552890778,
      "learning_rate": 2.706780752599135e-05,
      "loss": 0.5293,
      "mean_token_accuracy": 0.832843229174614,
      "num_tokens": 758997031.0,
      "step": 23782
    },
    {
      "entropy": 0.6135748289525509,
      "epoch": 2.1881530295972995,
      "grad_norm": 0.1689968705177307,
      "learning_rate": 2.706474069984973e-05,
      "loss": 0.6098,
      "mean_token_accuracy": 0.8091376163065434,
      "num_tokens": 759029595.0,
      "step": 23783
    },
    {
      "entropy": 0.6933795106597245,
      "epoch": 2.1882450359687406,
      "grad_norm": 0.17834557592868805,
      "learning_rate": 2.70616738737081e-05,
      "loss": 0.6743,
      "mean_token_accuracy": 0.7953371703624725,
      "num_tokens": 759061272.0,
      "step": 23784
    },
    {
      "entropy": 0.5295474855229259,
      "epoch": 2.1883370423401822,
      "grad_norm": 0.15764114260673523,
      "learning_rate": 2.7058607047566475e-05,
      "loss": 0.5119,
      "mean_token_accuracy": 0.8397271819412708,
      "num_tokens": 759093598.0,
      "step": 23785
    },
    {
      "entropy": 0.5615310817956924,
      "epoch": 2.1884290487116234,
      "grad_norm": 0.16892854869365692,
      "learning_rate": 2.7055540221424846e-05,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8211745135486126,
      "num_tokens": 759124690.0,
      "step": 23786
    },
    {
      "entropy": 0.4329511453397572,
      "epoch": 2.1885210550830645,
      "grad_norm": 0.1479867696762085,
      "learning_rate": 2.7052473395283224e-05,
      "loss": 0.4182,
      "mean_token_accuracy": 0.8693915456533432,
      "num_tokens": 759157307.0,
      "step": 23787
    },
    {
      "entropy": 0.5069901756942272,
      "epoch": 2.1886130614545056,
      "grad_norm": 0.1601894646883011,
      "learning_rate": 2.7049406569141595e-05,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8421868085861206,
      "num_tokens": 759189189.0,
      "step": 23788
    },
    {
      "entropy": 0.5335832312703133,
      "epoch": 2.1887050678259468,
      "grad_norm": 0.15985576808452606,
      "learning_rate": 2.704633974299997e-05,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8366073369979858,
      "num_tokens": 759221404.0,
      "step": 23789
    },
    {
      "entropy": 0.4734414694830775,
      "epoch": 2.1887970741973883,
      "grad_norm": 0.1573713719844818,
      "learning_rate": 2.704327291685834e-05,
      "loss": 0.4559,
      "mean_token_accuracy": 0.8545776158571243,
      "num_tokens": 759253024.0,
      "step": 23790
    },
    {
      "entropy": 0.4018732188269496,
      "epoch": 2.1888890805688295,
      "grad_norm": 0.14465515315532684,
      "learning_rate": 2.704020609071672e-05,
      "loss": 0.3894,
      "mean_token_accuracy": 0.8716009892523289,
      "num_tokens": 759285057.0,
      "step": 23791
    },
    {
      "entropy": 0.5684047918766737,
      "epoch": 2.1889810869402706,
      "grad_norm": 0.1680024415254593,
      "learning_rate": 2.703713926457509e-05,
      "loss": 0.5721,
      "mean_token_accuracy": 0.8208846896886826,
      "num_tokens": 759317673.0,
      "step": 23792
    },
    {
      "entropy": 0.5751814404502511,
      "epoch": 2.1890730933117117,
      "grad_norm": 0.16985486447811127,
      "learning_rate": 2.7034072438433465e-05,
      "loss": 0.5556,
      "mean_token_accuracy": 0.8269699364900589,
      "num_tokens": 759349180.0,
      "step": 23793
    },
    {
      "entropy": 0.5538265164941549,
      "epoch": 2.189165099683153,
      "grad_norm": 0.16595907509326935,
      "learning_rate": 2.7031005612291843e-05,
      "loss": 0.5449,
      "mean_token_accuracy": 0.8284485265612602,
      "num_tokens": 759381674.0,
      "step": 23794
    },
    {
      "entropy": 0.548831813968718,
      "epoch": 2.1892571060545944,
      "grad_norm": 0.1608937680721283,
      "learning_rate": 2.7027938786150214e-05,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8305912688374519,
      "num_tokens": 759414167.0,
      "step": 23795
    },
    {
      "entropy": 0.49691994953900576,
      "epoch": 2.1893491124260356,
      "grad_norm": 0.15400315821170807,
      "learning_rate": 2.7024871960008592e-05,
      "loss": 0.4776,
      "mean_token_accuracy": 0.849508959800005,
      "num_tokens": 759446743.0,
      "step": 23796
    },
    {
      "entropy": 0.5878166318871081,
      "epoch": 2.1894411187974767,
      "grad_norm": 0.1675867885351181,
      "learning_rate": 2.702180513386696e-05,
      "loss": 0.5723,
      "mean_token_accuracy": 0.8163991831243038,
      "num_tokens": 759478781.0,
      "step": 23797
    },
    {
      "entropy": 0.5083350632339716,
      "epoch": 2.189533125168918,
      "grad_norm": 0.1582464575767517,
      "learning_rate": 2.701873830772534e-05,
      "loss": 0.4994,
      "mean_token_accuracy": 0.8435896672308445,
      "num_tokens": 759511042.0,
      "step": 23798
    },
    {
      "entropy": 0.5177946551702917,
      "epoch": 2.189625131540359,
      "grad_norm": 0.1595437079668045,
      "learning_rate": 2.701567148158371e-05,
      "loss": 0.4971,
      "mean_token_accuracy": 0.8436611406505108,
      "num_tokens": 759542916.0,
      "step": 23799
    },
    {
      "entropy": 0.5833287257701159,
      "epoch": 2.1897171379118006,
      "grad_norm": 0.1709306389093399,
      "learning_rate": 2.7012604655442088e-05,
      "loss": 0.5746,
      "mean_token_accuracy": 0.8164657391607761,
      "num_tokens": 759574831.0,
      "step": 23800
    },
    {
      "entropy": 0.5158677615690976,
      "epoch": 2.1898091442832417,
      "grad_norm": 0.15563532710075378,
      "learning_rate": 2.700953782930046e-05,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8457126133143902,
      "num_tokens": 759607099.0,
      "step": 23801
    },
    {
      "entropy": 0.49806624511256814,
      "epoch": 2.189901150654683,
      "grad_norm": 0.15422919392585754,
      "learning_rate": 2.7006471003158834e-05,
      "loss": 0.4949,
      "mean_token_accuracy": 0.8466300517320633,
      "num_tokens": 759639106.0,
      "step": 23802
    },
    {
      "entropy": 0.6026915144175291,
      "epoch": 2.189993157026124,
      "grad_norm": 0.16635890305042267,
      "learning_rate": 2.7003404177017205e-05,
      "loss": 0.5869,
      "mean_token_accuracy": 0.8154717646539211,
      "num_tokens": 759670644.0,
      "step": 23803
    },
    {
      "entropy": 0.4735979549586773,
      "epoch": 2.190085163397565,
      "grad_norm": 0.15800276398658752,
      "learning_rate": 2.7000337350875583e-05,
      "loss": 0.4635,
      "mean_token_accuracy": 0.84751296043396,
      "num_tokens": 759702563.0,
      "step": 23804
    },
    {
      "entropy": 0.4413170637562871,
      "epoch": 2.1901771697690067,
      "grad_norm": 0.15064391493797302,
      "learning_rate": 2.6997270524733954e-05,
      "loss": 0.422,
      "mean_token_accuracy": 0.869357455521822,
      "num_tokens": 759734262.0,
      "step": 23805
    },
    {
      "entropy": 0.5727147958241403,
      "epoch": 2.190269176140448,
      "grad_norm": 0.162059485912323,
      "learning_rate": 2.699420369859233e-05,
      "loss": 0.5467,
      "mean_token_accuracy": 0.8244815208017826,
      "num_tokens": 759766355.0,
      "step": 23806
    },
    {
      "entropy": 0.5347528662532568,
      "epoch": 2.190361182511889,
      "grad_norm": 0.16541843116283417,
      "learning_rate": 2.69911368724507e-05,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8341879621148109,
      "num_tokens": 759798104.0,
      "step": 23807
    },
    {
      "entropy": 0.4856530660763383,
      "epoch": 2.19045318888333,
      "grad_norm": 0.1525842845439911,
      "learning_rate": 2.6988070046309078e-05,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8464928492903709,
      "num_tokens": 759829805.0,
      "step": 23808
    },
    {
      "entropy": 0.47147962381131947,
      "epoch": 2.190545195254771,
      "grad_norm": 0.1519036740064621,
      "learning_rate": 2.698500322016745e-05,
      "loss": 0.4529,
      "mean_token_accuracy": 0.8569317534565926,
      "num_tokens": 759861117.0,
      "step": 23809
    },
    {
      "entropy": 0.47499594278633595,
      "epoch": 2.190637201626213,
      "grad_norm": 0.1569909006357193,
      "learning_rate": 2.6981936394025824e-05,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8536520153284073,
      "num_tokens": 759892505.0,
      "step": 23810
    },
    {
      "entropy": 0.5231226468458772,
      "epoch": 2.190729207997654,
      "grad_norm": 0.16276466846466064,
      "learning_rate": 2.6978869567884195e-05,
      "loss": 0.5038,
      "mean_token_accuracy": 0.8389384411275387,
      "num_tokens": 759925102.0,
      "step": 23811
    },
    {
      "entropy": 0.5669290823861957,
      "epoch": 2.190821214369095,
      "grad_norm": 0.1613745093345642,
      "learning_rate": 2.6975802741742573e-05,
      "loss": 0.5489,
      "mean_token_accuracy": 0.8284406252205372,
      "num_tokens": 759957385.0,
      "step": 23812
    },
    {
      "entropy": 0.4150994410738349,
      "epoch": 2.190913220740536,
      "grad_norm": 0.14834314584732056,
      "learning_rate": 2.6972735915600945e-05,
      "loss": 0.3969,
      "mean_token_accuracy": 0.8719798140227795,
      "num_tokens": 759989203.0,
      "step": 23813
    },
    {
      "entropy": 0.6023230478167534,
      "epoch": 2.1910052271119773,
      "grad_norm": 0.17113271355628967,
      "learning_rate": 2.696966908945932e-05,
      "loss": 0.5769,
      "mean_token_accuracy": 0.8174680881202221,
      "num_tokens": 760020481.0,
      "step": 23814
    },
    {
      "entropy": 0.5455618768464774,
      "epoch": 2.191097233483419,
      "grad_norm": 0.1623707115650177,
      "learning_rate": 2.696660226331769e-05,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8298507630825043,
      "num_tokens": 760052579.0,
      "step": 23815
    },
    {
      "entropy": 0.583277290686965,
      "epoch": 2.19118923985486,
      "grad_norm": 0.17166562378406525,
      "learning_rate": 2.696353543717607e-05,
      "loss": 0.5757,
      "mean_token_accuracy": 0.8221931532025337,
      "num_tokens": 760083225.0,
      "step": 23816
    },
    {
      "entropy": 0.4941352792084217,
      "epoch": 2.191281246226301,
      "grad_norm": 0.1619686484336853,
      "learning_rate": 2.696046861103444e-05,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8472783342003822,
      "num_tokens": 760114807.0,
      "step": 23817
    },
    {
      "entropy": 0.5564689505845308,
      "epoch": 2.1913732525977423,
      "grad_norm": 0.15965600311756134,
      "learning_rate": 2.6957401784892818e-05,
      "loss": 0.5433,
      "mean_token_accuracy": 0.8310802653431892,
      "num_tokens": 760147279.0,
      "step": 23818
    },
    {
      "entropy": 0.5595513544976711,
      "epoch": 2.1914652589691834,
      "grad_norm": 0.16577883064746857,
      "learning_rate": 2.6954334958751186e-05,
      "loss": 0.5535,
      "mean_token_accuracy": 0.8253929205238819,
      "num_tokens": 760179800.0,
      "step": 23819
    },
    {
      "entropy": 0.4342053709551692,
      "epoch": 2.191557265340625,
      "grad_norm": 0.14770345389842987,
      "learning_rate": 2.6951268132609564e-05,
      "loss": 0.4254,
      "mean_token_accuracy": 0.8642414696514606,
      "num_tokens": 760211846.0,
      "step": 23820
    },
    {
      "entropy": 0.5916228238493204,
      "epoch": 2.191649271712066,
      "grad_norm": 0.1665831357240677,
      "learning_rate": 2.6948201306467935e-05,
      "loss": 0.5782,
      "mean_token_accuracy": 0.8188469409942627,
      "num_tokens": 760244339.0,
      "step": 23821
    },
    {
      "entropy": 0.4869741462171078,
      "epoch": 2.1917412780835073,
      "grad_norm": 0.16294357180595398,
      "learning_rate": 2.6945134480326313e-05,
      "loss": 0.4794,
      "mean_token_accuracy": 0.8486057221889496,
      "num_tokens": 760275689.0,
      "step": 23822
    },
    {
      "entropy": 0.4928282715845853,
      "epoch": 2.1918332844549484,
      "grad_norm": 0.15540896356105804,
      "learning_rate": 2.6942067654184684e-05,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8473732098937035,
      "num_tokens": 760307334.0,
      "step": 23823
    },
    {
      "entropy": 0.578589865937829,
      "epoch": 2.1919252908263895,
      "grad_norm": 0.1776258796453476,
      "learning_rate": 2.693900082804306e-05,
      "loss": 0.5719,
      "mean_token_accuracy": 0.8197028040885925,
      "num_tokens": 760339159.0,
      "step": 23824
    },
    {
      "entropy": 0.574146943166852,
      "epoch": 2.192017297197831,
      "grad_norm": 0.16611576080322266,
      "learning_rate": 2.6935934001901437e-05,
      "loss": 0.5544,
      "mean_token_accuracy": 0.8246815949678421,
      "num_tokens": 760371239.0,
      "step": 23825
    },
    {
      "entropy": 0.47126574511639774,
      "epoch": 2.1921093035692722,
      "grad_norm": 0.15437957644462585,
      "learning_rate": 2.6932867175759808e-05,
      "loss": 0.4603,
      "mean_token_accuracy": 0.8581824861466885,
      "num_tokens": 760403471.0,
      "step": 23826
    },
    {
      "entropy": 0.4510419415310025,
      "epoch": 2.1922013099407134,
      "grad_norm": 0.1536766141653061,
      "learning_rate": 2.6929800349618183e-05,
      "loss": 0.4299,
      "mean_token_accuracy": 0.857527133077383,
      "num_tokens": 760434560.0,
      "step": 23827
    },
    {
      "entropy": 0.48959570564329624,
      "epoch": 2.1922933163121545,
      "grad_norm": 0.1507236212491989,
      "learning_rate": 2.6926733523476554e-05,
      "loss": 0.4761,
      "mean_token_accuracy": 0.8515100888907909,
      "num_tokens": 760466925.0,
      "step": 23828
    },
    {
      "entropy": 0.5614915583282709,
      "epoch": 2.1923853226835956,
      "grad_norm": 0.15993788838386536,
      "learning_rate": 2.6923666697334932e-05,
      "loss": 0.5441,
      "mean_token_accuracy": 0.825629536062479,
      "num_tokens": 760499063.0,
      "step": 23829
    },
    {
      "entropy": 0.49306635558605194,
      "epoch": 2.1924773290550372,
      "grad_norm": 0.16267350316047668,
      "learning_rate": 2.6920599871193303e-05,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8481754809617996,
      "num_tokens": 760530178.0,
      "step": 23830
    },
    {
      "entropy": 0.38733710860833526,
      "epoch": 2.1925693354264784,
      "grad_norm": 0.1470993459224701,
      "learning_rate": 2.6917533045051678e-05,
      "loss": 0.3819,
      "mean_token_accuracy": 0.8763630203902721,
      "num_tokens": 760561828.0,
      "step": 23831
    },
    {
      "entropy": 0.39886833261698484,
      "epoch": 2.1926613417979195,
      "grad_norm": 0.14781959354877472,
      "learning_rate": 2.691446621891005e-05,
      "loss": 0.381,
      "mean_token_accuracy": 0.8784703984856606,
      "num_tokens": 760593714.0,
      "step": 23832
    },
    {
      "entropy": 0.6244600862264633,
      "epoch": 2.1927533481693606,
      "grad_norm": 0.1751101315021515,
      "learning_rate": 2.6911399392768427e-05,
      "loss": 0.6063,
      "mean_token_accuracy": 0.8056793175637722,
      "num_tokens": 760624798.0,
      "step": 23833
    },
    {
      "entropy": 0.513577233068645,
      "epoch": 2.1928453545408018,
      "grad_norm": 0.16182841360569,
      "learning_rate": 2.69083325666268e-05,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8442580699920654,
      "num_tokens": 760656653.0,
      "step": 23834
    },
    {
      "entropy": 0.5141155561432242,
      "epoch": 2.1929373609122433,
      "grad_norm": 0.16006620228290558,
      "learning_rate": 2.6905265740485173e-05,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8410985544323921,
      "num_tokens": 760688795.0,
      "step": 23835
    },
    {
      "entropy": 0.4360063043422997,
      "epoch": 2.1930293672836845,
      "grad_norm": 0.15128551423549652,
      "learning_rate": 2.6902198914343544e-05,
      "loss": 0.421,
      "mean_token_accuracy": 0.8636649250984192,
      "num_tokens": 760720988.0,
      "step": 23836
    },
    {
      "entropy": 0.5020132027566433,
      "epoch": 2.1931213736551256,
      "grad_norm": 0.15785102546215057,
      "learning_rate": 2.6899132088201923e-05,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8440614975988865,
      "num_tokens": 760753306.0,
      "step": 23837
    },
    {
      "entropy": 0.443241496803239,
      "epoch": 2.1932133800265667,
      "grad_norm": 0.15858392417430878,
      "learning_rate": 2.6896065262060294e-05,
      "loss": 0.4372,
      "mean_token_accuracy": 0.8610507734119892,
      "num_tokens": 760784329.0,
      "step": 23838
    },
    {
      "entropy": 0.48549686186015606,
      "epoch": 2.193305386398008,
      "grad_norm": 0.1593698114156723,
      "learning_rate": 2.6892998435918672e-05,
      "loss": 0.4736,
      "mean_token_accuracy": 0.8474798835813999,
      "num_tokens": 760815575.0,
      "step": 23839
    },
    {
      "entropy": 0.4895594324916601,
      "epoch": 2.1933973927694494,
      "grad_norm": 0.15520605444908142,
      "learning_rate": 2.6889931609777043e-05,
      "loss": 0.4658,
      "mean_token_accuracy": 0.849860668182373,
      "num_tokens": 760847412.0,
      "step": 23840
    },
    {
      "entropy": 0.5464309435337782,
      "epoch": 2.1934893991408906,
      "grad_norm": 0.1702042520046234,
      "learning_rate": 2.6886864783635418e-05,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8296872079372406,
      "num_tokens": 760879245.0,
      "step": 23841
    },
    {
      "entropy": 0.5097060045227408,
      "epoch": 2.1935814055123317,
      "grad_norm": 0.16050253808498383,
      "learning_rate": 2.688379795749379e-05,
      "loss": 0.4888,
      "mean_token_accuracy": 0.8447592072188854,
      "num_tokens": 760910473.0,
      "step": 23842
    },
    {
      "entropy": 0.5359865557402372,
      "epoch": 2.193673411883773,
      "grad_norm": 0.15880931913852692,
      "learning_rate": 2.6880731131352167e-05,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8334648087620735,
      "num_tokens": 760943061.0,
      "step": 23843
    },
    {
      "entropy": 0.46295816358178854,
      "epoch": 2.193765418255214,
      "grad_norm": 0.1523503214120865,
      "learning_rate": 2.6877664305210538e-05,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8562530837953091,
      "num_tokens": 760975686.0,
      "step": 23844
    },
    {
      "entropy": 0.41816407069563866,
      "epoch": 2.1938574246266556,
      "grad_norm": 0.14919306337833405,
      "learning_rate": 2.6874597479068913e-05,
      "loss": 0.4026,
      "mean_token_accuracy": 0.8706895224750042,
      "num_tokens": 761007805.0,
      "step": 23845
    },
    {
      "entropy": 0.414823554456234,
      "epoch": 2.1939494309980967,
      "grad_norm": 0.1452331393957138,
      "learning_rate": 2.6871530652927284e-05,
      "loss": 0.4107,
      "mean_token_accuracy": 0.8696855194866657,
      "num_tokens": 761040542.0,
      "step": 23846
    },
    {
      "entropy": 0.5308820819482207,
      "epoch": 2.194041437369538,
      "grad_norm": 0.15893562138080597,
      "learning_rate": 2.6868463826785662e-05,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8342249244451523,
      "num_tokens": 761072547.0,
      "step": 23847
    },
    {
      "entropy": 0.523789837025106,
      "epoch": 2.194133443740979,
      "grad_norm": 0.16097860038280487,
      "learning_rate": 2.6865397000644033e-05,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8347685821354389,
      "num_tokens": 761104749.0,
      "step": 23848
    },
    {
      "entropy": 0.5926093519665301,
      "epoch": 2.19422545011242,
      "grad_norm": 0.1728430688381195,
      "learning_rate": 2.6862330174502408e-05,
      "loss": 0.5952,
      "mean_token_accuracy": 0.8169688582420349,
      "num_tokens": 761137200.0,
      "step": 23849
    },
    {
      "entropy": 0.5218187049031258,
      "epoch": 2.1943174564838617,
      "grad_norm": 0.16457179188728333,
      "learning_rate": 2.685926334836078e-05,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8346878923475742,
      "num_tokens": 761169410.0,
      "step": 23850
    },
    {
      "entropy": 0.48994527477771044,
      "epoch": 2.194409462855303,
      "grad_norm": 0.15739598870277405,
      "learning_rate": 2.6856196522219157e-05,
      "loss": 0.4828,
      "mean_token_accuracy": 0.8452988639473915,
      "num_tokens": 761201983.0,
      "step": 23851
    },
    {
      "entropy": 0.47110319836065173,
      "epoch": 2.194501469226744,
      "grad_norm": 0.15655402839183807,
      "learning_rate": 2.685312969607753e-05,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8560862988233566,
      "num_tokens": 761233198.0,
      "step": 23852
    },
    {
      "entropy": 0.4632733450271189,
      "epoch": 2.194593475598185,
      "grad_norm": 0.15316885709762573,
      "learning_rate": 2.6850062869935903e-05,
      "loss": 0.4499,
      "mean_token_accuracy": 0.853406984359026,
      "num_tokens": 761265235.0,
      "step": 23853
    },
    {
      "entropy": 0.4958559339866042,
      "epoch": 2.194685481969626,
      "grad_norm": 0.1540669947862625,
      "learning_rate": 2.684699604379428e-05,
      "loss": 0.477,
      "mean_token_accuracy": 0.8519904837012291,
      "num_tokens": 761297046.0,
      "step": 23854
    },
    {
      "entropy": 0.6204333212226629,
      "epoch": 2.194777488341068,
      "grad_norm": 0.174019455909729,
      "learning_rate": 2.6843929217652653e-05,
      "loss": 0.5851,
      "mean_token_accuracy": 0.8145680725574493,
      "num_tokens": 761328455.0,
      "step": 23855
    },
    {
      "entropy": 0.4414678995963186,
      "epoch": 2.194869494712509,
      "grad_norm": 0.14863085746765137,
      "learning_rate": 2.684086239151103e-05,
      "loss": 0.425,
      "mean_token_accuracy": 0.8631187528371811,
      "num_tokens": 761360838.0,
      "step": 23856
    },
    {
      "entropy": 0.5118047120049596,
      "epoch": 2.19496150108395,
      "grad_norm": 0.1559344381093979,
      "learning_rate": 2.68377955653694e-05,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8425546735525131,
      "num_tokens": 761392984.0,
      "step": 23857
    },
    {
      "entropy": 0.4862797064706683,
      "epoch": 2.195053507455391,
      "grad_norm": 0.15256869792938232,
      "learning_rate": 2.6834728739227777e-05,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8500564247369766,
      "num_tokens": 761425432.0,
      "step": 23858
    },
    {
      "entropy": 0.5376795576885343,
      "epoch": 2.1951455138268323,
      "grad_norm": 0.1636016070842743,
      "learning_rate": 2.6831661913086148e-05,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8339269272983074,
      "num_tokens": 761457413.0,
      "step": 23859
    },
    {
      "entropy": 0.4756336407735944,
      "epoch": 2.195237520198274,
      "grad_norm": 0.15115591883659363,
      "learning_rate": 2.6828595086944526e-05,
      "loss": 0.4584,
      "mean_token_accuracy": 0.8497894406318665,
      "num_tokens": 761488690.0,
      "step": 23860
    },
    {
      "entropy": 0.6548992209136486,
      "epoch": 2.195329526569715,
      "grad_norm": 0.17501308023929596,
      "learning_rate": 2.6825528260802897e-05,
      "loss": 0.642,
      "mean_token_accuracy": 0.7963238134980202,
      "num_tokens": 761521040.0,
      "step": 23861
    },
    {
      "entropy": 0.6228053513914347,
      "epoch": 2.195421532941156,
      "grad_norm": 0.17163200676441193,
      "learning_rate": 2.6822461434661272e-05,
      "loss": 0.6147,
      "mean_token_accuracy": 0.8079055696725845,
      "num_tokens": 761553507.0,
      "step": 23862
    },
    {
      "entropy": 0.49400899559259415,
      "epoch": 2.1955135393125973,
      "grad_norm": 0.16400010883808136,
      "learning_rate": 2.6819394608519643e-05,
      "loss": 0.4928,
      "mean_token_accuracy": 0.8448725193738937,
      "num_tokens": 761585316.0,
      "step": 23863
    },
    {
      "entropy": 0.5198334637098014,
      "epoch": 2.1956055456840384,
      "grad_norm": 0.16318215429782867,
      "learning_rate": 2.681632778237802e-05,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8394715636968613,
      "num_tokens": 761617372.0,
      "step": 23864
    },
    {
      "entropy": 0.5139255747199059,
      "epoch": 2.19569755205548,
      "grad_norm": 0.15948408842086792,
      "learning_rate": 2.6813260956236392e-05,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8423855938017368,
      "num_tokens": 761649122.0,
      "step": 23865
    },
    {
      "entropy": 0.5333412932232022,
      "epoch": 2.195789558426921,
      "grad_norm": 0.16458731889724731,
      "learning_rate": 2.6810194130094767e-05,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8305684365332127,
      "num_tokens": 761680909.0,
      "step": 23866
    },
    {
      "entropy": 0.4973894264549017,
      "epoch": 2.1958815647983623,
      "grad_norm": 0.15860086679458618,
      "learning_rate": 2.6807127303953138e-05,
      "loss": 0.485,
      "mean_token_accuracy": 0.844017818570137,
      "num_tokens": 761713121.0,
      "step": 23867
    },
    {
      "entropy": 0.5426563906949013,
      "epoch": 2.1959735711698034,
      "grad_norm": 0.1647251844406128,
      "learning_rate": 2.6804060477811516e-05,
      "loss": 0.5268,
      "mean_token_accuracy": 0.8356899917125702,
      "num_tokens": 761744504.0,
      "step": 23868
    },
    {
      "entropy": 0.5566534176468849,
      "epoch": 2.1960655775412445,
      "grad_norm": 0.16485296189785004,
      "learning_rate": 2.6800993651669888e-05,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8278569839894772,
      "num_tokens": 761776790.0,
      "step": 23869
    },
    {
      "entropy": 0.5174671830609441,
      "epoch": 2.196157583912686,
      "grad_norm": 0.1632506102323532,
      "learning_rate": 2.6797926825528262e-05,
      "loss": 0.5057,
      "mean_token_accuracy": 0.8394206464290619,
      "num_tokens": 761808919.0,
      "step": 23870
    },
    {
      "entropy": 0.5077793067321181,
      "epoch": 2.1962495902841273,
      "grad_norm": 0.16226962208747864,
      "learning_rate": 2.6794859999386633e-05,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8453873917460442,
      "num_tokens": 761840675.0,
      "step": 23871
    },
    {
      "entropy": 0.5280186019372195,
      "epoch": 2.1963415966555684,
      "grad_norm": 0.16102856397628784,
      "learning_rate": 2.679179317324501e-05,
      "loss": 0.5082,
      "mean_token_accuracy": 0.839551292359829,
      "num_tokens": 761872435.0,
      "step": 23872
    },
    {
      "entropy": 0.6251271329820156,
      "epoch": 2.1964336030270095,
      "grad_norm": 0.17130516469478607,
      "learning_rate": 2.6788726347103383e-05,
      "loss": 0.6098,
      "mean_token_accuracy": 0.8086183145642281,
      "num_tokens": 761904998.0,
      "step": 23873
    },
    {
      "entropy": 0.524519762955606,
      "epoch": 2.1965256093984507,
      "grad_norm": 0.16186025738716125,
      "learning_rate": 2.6785659520961757e-05,
      "loss": 0.5061,
      "mean_token_accuracy": 0.8335610739886761,
      "num_tokens": 761937061.0,
      "step": 23874
    },
    {
      "entropy": 0.5726151447743177,
      "epoch": 2.1966176157698922,
      "grad_norm": 0.16412532329559326,
      "learning_rate": 2.678259269482013e-05,
      "loss": 0.5553,
      "mean_token_accuracy": 0.8271565958857536,
      "num_tokens": 761969389.0,
      "step": 23875
    },
    {
      "entropy": 0.5088929338380694,
      "epoch": 2.1967096221413334,
      "grad_norm": 0.16132007539272308,
      "learning_rate": 2.6779525868678507e-05,
      "loss": 0.4909,
      "mean_token_accuracy": 0.8433825634419918,
      "num_tokens": 762001883.0,
      "step": 23876
    },
    {
      "entropy": 0.57261170912534,
      "epoch": 2.1968016285127745,
      "grad_norm": 0.1681133657693863,
      "learning_rate": 2.6776459042536878e-05,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8221142627298832,
      "num_tokens": 762033233.0,
      "step": 23877
    },
    {
      "entropy": 0.5485435910522938,
      "epoch": 2.1968936348842156,
      "grad_norm": 0.163797065615654,
      "learning_rate": 2.6773392216395256e-05,
      "loss": 0.5269,
      "mean_token_accuracy": 0.833943534642458,
      "num_tokens": 762065339.0,
      "step": 23878
    },
    {
      "entropy": 0.5073887938633561,
      "epoch": 2.1969856412556568,
      "grad_norm": 0.1620606631040573,
      "learning_rate": 2.6770325390253624e-05,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8434332832694054,
      "num_tokens": 762097512.0,
      "step": 23879
    },
    {
      "entropy": 0.4191779689863324,
      "epoch": 2.1970776476270983,
      "grad_norm": 0.15407699346542358,
      "learning_rate": 2.6767258564112002e-05,
      "loss": 0.4081,
      "mean_token_accuracy": 0.8710489645600319,
      "num_tokens": 762128628.0,
      "step": 23880
    },
    {
      "entropy": 0.4899896909482777,
      "epoch": 2.1971696539985395,
      "grad_norm": 0.1604280024766922,
      "learning_rate": 2.6764191737970373e-05,
      "loss": 0.4827,
      "mean_token_accuracy": 0.8500471115112305,
      "num_tokens": 762160814.0,
      "step": 23881
    },
    {
      "entropy": 0.5154887409880757,
      "epoch": 2.1972616603699806,
      "grad_norm": 0.16071002185344696,
      "learning_rate": 2.676112491182875e-05,
      "loss": 0.506,
      "mean_token_accuracy": 0.8373332768678665,
      "num_tokens": 762193426.0,
      "step": 23882
    },
    {
      "entropy": 0.5200560146477073,
      "epoch": 2.1973536667414217,
      "grad_norm": 0.16442076861858368,
      "learning_rate": 2.6758058085687122e-05,
      "loss": 0.5012,
      "mean_token_accuracy": 0.8424103930592537,
      "num_tokens": 762224528.0,
      "step": 23883
    },
    {
      "entropy": 0.5619708215817809,
      "epoch": 2.197445673112863,
      "grad_norm": 0.16523340344429016,
      "learning_rate": 2.6754991259545497e-05,
      "loss": 0.5508,
      "mean_token_accuracy": 0.8242597579956055,
      "num_tokens": 762255990.0,
      "step": 23884
    },
    {
      "entropy": 0.5185626614838839,
      "epoch": 2.1975376794843045,
      "grad_norm": 0.16037997603416443,
      "learning_rate": 2.6751924433403875e-05,
      "loss": 0.5048,
      "mean_token_accuracy": 0.8366673141717911,
      "num_tokens": 762288403.0,
      "step": 23885
    },
    {
      "entropy": 0.5906682247295976,
      "epoch": 2.1976296858557456,
      "grad_norm": 0.16833089292049408,
      "learning_rate": 2.6748857607262246e-05,
      "loss": 0.5799,
      "mean_token_accuracy": 0.8184047043323517,
      "num_tokens": 762320539.0,
      "step": 23886
    },
    {
      "entropy": 0.4870938416570425,
      "epoch": 2.1977216922271867,
      "grad_norm": 0.1546534299850464,
      "learning_rate": 2.674579078112062e-05,
      "loss": 0.4695,
      "mean_token_accuracy": 0.8503160141408443,
      "num_tokens": 762352623.0,
      "step": 23887
    },
    {
      "entropy": 0.5661062318831682,
      "epoch": 2.197813698598628,
      "grad_norm": 0.1625399887561798,
      "learning_rate": 2.6742723954978992e-05,
      "loss": 0.5402,
      "mean_token_accuracy": 0.8295792788267136,
      "num_tokens": 762384459.0,
      "step": 23888
    },
    {
      "entropy": 0.5174448592588305,
      "epoch": 2.197905704970069,
      "grad_norm": 0.15953534841537476,
      "learning_rate": 2.673965712883737e-05,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8388465531170368,
      "num_tokens": 762417151.0,
      "step": 23889
    },
    {
      "entropy": 0.5060747554525733,
      "epoch": 2.1979977113415106,
      "grad_norm": 0.1666906476020813,
      "learning_rate": 2.673659030269574e-05,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8426966853439808,
      "num_tokens": 762448470.0,
      "step": 23890
    },
    {
      "entropy": 0.5088634905405343,
      "epoch": 2.1980897177129517,
      "grad_norm": 0.16008727252483368,
      "learning_rate": 2.6733523476554116e-05,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8445515111088753,
      "num_tokens": 762480682.0,
      "step": 23891
    },
    {
      "entropy": 0.5725460685789585,
      "epoch": 2.198181724084393,
      "grad_norm": 0.16455329954624176,
      "learning_rate": 2.6730456650412487e-05,
      "loss": 0.5625,
      "mean_token_accuracy": 0.8204195499420166,
      "num_tokens": 762512647.0,
      "step": 23892
    },
    {
      "entropy": 0.5988190863281488,
      "epoch": 2.198273730455834,
      "grad_norm": 0.16629591584205627,
      "learning_rate": 2.6727389824270866e-05,
      "loss": 0.5801,
      "mean_token_accuracy": 0.8144536800682545,
      "num_tokens": 762544845.0,
      "step": 23893
    },
    {
      "entropy": 0.514699381776154,
      "epoch": 2.198365736827275,
      "grad_norm": 0.15556223690509796,
      "learning_rate": 2.6724322998129237e-05,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8434582874178886,
      "num_tokens": 762576830.0,
      "step": 23894
    },
    {
      "entropy": 0.5009180130437016,
      "epoch": 2.1984577431987167,
      "grad_norm": 0.15802831947803497,
      "learning_rate": 2.672125617198761e-05,
      "loss": 0.4898,
      "mean_token_accuracy": 0.8428050838410854,
      "num_tokens": 762609185.0,
      "step": 23895
    },
    {
      "entropy": 0.511470252647996,
      "epoch": 2.198549749570158,
      "grad_norm": 0.16059549152851105,
      "learning_rate": 2.6718189345845983e-05,
      "loss": 0.501,
      "mean_token_accuracy": 0.8392354175448418,
      "num_tokens": 762640998.0,
      "step": 23896
    },
    {
      "entropy": 0.5625744517892599,
      "epoch": 2.198641755941599,
      "grad_norm": 0.1707785278558731,
      "learning_rate": 2.671512251970436e-05,
      "loss": 0.5423,
      "mean_token_accuracy": 0.8288163915276527,
      "num_tokens": 762671638.0,
      "step": 23897
    },
    {
      "entropy": 0.5118923354893923,
      "epoch": 2.19873376231304,
      "grad_norm": 0.16253754496574402,
      "learning_rate": 2.6712055693562732e-05,
      "loss": 0.5083,
      "mean_token_accuracy": 0.8376931697130203,
      "num_tokens": 762703557.0,
      "step": 23898
    },
    {
      "entropy": 0.4462644997984171,
      "epoch": 2.198825768684481,
      "grad_norm": 0.1630004644393921,
      "learning_rate": 2.670898886742111e-05,
      "loss": 0.4247,
      "mean_token_accuracy": 0.8616640754044056,
      "num_tokens": 762734737.0,
      "step": 23899
    },
    {
      "entropy": 0.5374566707760096,
      "epoch": 2.198917775055923,
      "grad_norm": 0.15807585418224335,
      "learning_rate": 2.670592204127948e-05,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8334488272666931,
      "num_tokens": 762766648.0,
      "step": 23900
    },
    {
      "entropy": 0.450707426876761,
      "epoch": 2.199009781427364,
      "grad_norm": 0.15249235928058624,
      "learning_rate": 2.6702855215137856e-05,
      "loss": 0.434,
      "mean_token_accuracy": 0.8595440723001957,
      "num_tokens": 762797857.0,
      "step": 23901
    },
    {
      "entropy": 0.5362517666071653,
      "epoch": 2.199101787798805,
      "grad_norm": 0.1631220430135727,
      "learning_rate": 2.6699788388996227e-05,
      "loss": 0.533,
      "mean_token_accuracy": 0.8340910114347935,
      "num_tokens": 762830289.0,
      "step": 23902
    },
    {
      "entropy": 0.42659678193740547,
      "epoch": 2.199193794170246,
      "grad_norm": 0.14866095781326294,
      "learning_rate": 2.6696721562854605e-05,
      "loss": 0.4156,
      "mean_token_accuracy": 0.869159284979105,
      "num_tokens": 762862360.0,
      "step": 23903
    },
    {
      "entropy": 0.5557638900354505,
      "epoch": 2.1992858005416873,
      "grad_norm": 0.17388896644115448,
      "learning_rate": 2.6693654736712976e-05,
      "loss": 0.5469,
      "mean_token_accuracy": 0.8271937742829323,
      "num_tokens": 762893477.0,
      "step": 23904
    },
    {
      "entropy": 0.556524976156652,
      "epoch": 2.199377806913129,
      "grad_norm": 0.1590351015329361,
      "learning_rate": 2.669058791057135e-05,
      "loss": 0.5375,
      "mean_token_accuracy": 0.8284084312617779,
      "num_tokens": 762925825.0,
      "step": 23905
    },
    {
      "entropy": 0.49678466469049454,
      "epoch": 2.19946981328457,
      "grad_norm": 0.1586136668920517,
      "learning_rate": 2.6687521084429722e-05,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8442197218537331,
      "num_tokens": 762957902.0,
      "step": 23906
    },
    {
      "entropy": 0.5818054601550102,
      "epoch": 2.199561819656011,
      "grad_norm": 0.17020206153392792,
      "learning_rate": 2.66844542582881e-05,
      "loss": 0.5619,
      "mean_token_accuracy": 0.8249221406877041,
      "num_tokens": 762990175.0,
      "step": 23907
    },
    {
      "entropy": 0.55581618193537,
      "epoch": 2.1996538260274523,
      "grad_norm": 0.16133366525173187,
      "learning_rate": 2.668138743214647e-05,
      "loss": 0.5401,
      "mean_token_accuracy": 0.8312729336321354,
      "num_tokens": 763022537.0,
      "step": 23908
    },
    {
      "entropy": 0.5206970721483231,
      "epoch": 2.1997458323988934,
      "grad_norm": 0.15973475575447083,
      "learning_rate": 2.6678320606004846e-05,
      "loss": 0.5091,
      "mean_token_accuracy": 0.84162962064147,
      "num_tokens": 763054813.0,
      "step": 23909
    },
    {
      "entropy": 0.7162657994776964,
      "epoch": 2.199837838770335,
      "grad_norm": 0.1838168203830719,
      "learning_rate": 2.6675253779863218e-05,
      "loss": 0.692,
      "mean_token_accuracy": 0.7832848951220512,
      "num_tokens": 763086551.0,
      "step": 23910
    },
    {
      "entropy": 0.527633662102744,
      "epoch": 2.199929845141776,
      "grad_norm": 0.16136038303375244,
      "learning_rate": 2.6672186953721596e-05,
      "loss": 0.509,
      "mean_token_accuracy": 0.8349039070308208,
      "num_tokens": 763118482.0,
      "step": 23911
    },
    {
      "entropy": 0.444749849033542,
      "epoch": 2.2000218515132173,
      "grad_norm": 0.15326718986034393,
      "learning_rate": 2.6669120127579967e-05,
      "loss": 0.4206,
      "mean_token_accuracy": 0.8620240762829781,
      "num_tokens": 763150831.0,
      "step": 23912
    },
    {
      "entropy": 0.5400611627846956,
      "epoch": 2.2001138578846584,
      "grad_norm": 0.1656465083360672,
      "learning_rate": 2.666605330143834e-05,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8311649262905121,
      "num_tokens": 763182613.0,
      "step": 23913
    },
    {
      "entropy": 0.4767199109774083,
      "epoch": 2.2002058642560995,
      "grad_norm": 0.15583904087543488,
      "learning_rate": 2.6662986475296713e-05,
      "loss": 0.4649,
      "mean_token_accuracy": 0.8539392836391926,
      "num_tokens": 763214147.0,
      "step": 23914
    },
    {
      "entropy": 0.4891490451991558,
      "epoch": 2.200297870627541,
      "grad_norm": 0.15329042077064514,
      "learning_rate": 2.665991964915509e-05,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8502002023160458,
      "num_tokens": 763246682.0,
      "step": 23915
    },
    {
      "entropy": 0.518265746999532,
      "epoch": 2.2003898769989823,
      "grad_norm": 0.16100290417671204,
      "learning_rate": 2.665685282301347e-05,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8397911414504051,
      "num_tokens": 763279110.0,
      "step": 23916
    },
    {
      "entropy": 0.4912858745083213,
      "epoch": 2.2004818833704234,
      "grad_norm": 0.15494999289512634,
      "learning_rate": 2.665378599687184e-05,
      "loss": 0.4836,
      "mean_token_accuracy": 0.8472531400620937,
      "num_tokens": 763310471.0,
      "step": 23917
    },
    {
      "entropy": 0.5667171312961727,
      "epoch": 2.2005738897418645,
      "grad_norm": 0.16204412281513214,
      "learning_rate": 2.6650719170730215e-05,
      "loss": 0.556,
      "mean_token_accuracy": 0.8282780051231384,
      "num_tokens": 763342872.0,
      "step": 23918
    },
    {
      "entropy": 0.47775177052244544,
      "epoch": 2.2006658961133057,
      "grad_norm": 0.16114436089992523,
      "learning_rate": 2.6647652344588586e-05,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8494814150035381,
      "num_tokens": 763374470.0,
      "step": 23919
    },
    {
      "entropy": 0.4727697968482971,
      "epoch": 2.2007579024847472,
      "grad_norm": 0.16232417523860931,
      "learning_rate": 2.6644585518446964e-05,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8532162383198738,
      "num_tokens": 763406117.0,
      "step": 23920
    },
    {
      "entropy": 0.5647092573344707,
      "epoch": 2.2008499088561884,
      "grad_norm": 0.16617697477340698,
      "learning_rate": 2.6641518692305335e-05,
      "loss": 0.5509,
      "mean_token_accuracy": 0.822596188634634,
      "num_tokens": 763437865.0,
      "step": 23921
    },
    {
      "entropy": 0.5425895480439067,
      "epoch": 2.2009419152276295,
      "grad_norm": 0.1618974506855011,
      "learning_rate": 2.663845186616371e-05,
      "loss": 0.5271,
      "mean_token_accuracy": 0.8319633714854717,
      "num_tokens": 763469254.0,
      "step": 23922
    },
    {
      "entropy": 0.618665736168623,
      "epoch": 2.2010339215990706,
      "grad_norm": 0.1710975617170334,
      "learning_rate": 2.663538504002208e-05,
      "loss": 0.6129,
      "mean_token_accuracy": 0.8110185377299786,
      "num_tokens": 763501579.0,
      "step": 23923
    },
    {
      "entropy": 0.5103622004389763,
      "epoch": 2.2011259279705118,
      "grad_norm": 0.16472594439983368,
      "learning_rate": 2.663231821388046e-05,
      "loss": 0.5035,
      "mean_token_accuracy": 0.8406838811933994,
      "num_tokens": 763532804.0,
      "step": 23924
    },
    {
      "entropy": 0.5852285670116544,
      "epoch": 2.2012179343419533,
      "grad_norm": 0.16992193460464478,
      "learning_rate": 2.662925138773883e-05,
      "loss": 0.5681,
      "mean_token_accuracy": 0.8219580352306366,
      "num_tokens": 763564401.0,
      "step": 23925
    },
    {
      "entropy": 0.5272074174135923,
      "epoch": 2.2013099407133945,
      "grad_norm": 0.16686852276325226,
      "learning_rate": 2.6626184561597205e-05,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8348043672740459,
      "num_tokens": 763596236.0,
      "step": 23926
    },
    {
      "entropy": 0.5814924854785204,
      "epoch": 2.2014019470848356,
      "grad_norm": 0.1669013351202011,
      "learning_rate": 2.6623117735455576e-05,
      "loss": 0.5655,
      "mean_token_accuracy": 0.821118887513876,
      "num_tokens": 763628069.0,
      "step": 23927
    },
    {
      "entropy": 0.562896073795855,
      "epoch": 2.2014939534562767,
      "grad_norm": 0.16245561838150024,
      "learning_rate": 2.6620050909313954e-05,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8258851021528244,
      "num_tokens": 763659730.0,
      "step": 23928
    },
    {
      "entropy": 0.49891030602157116,
      "epoch": 2.201585959827718,
      "grad_norm": 0.16121496260166168,
      "learning_rate": 2.6616984083172326e-05,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8390556201338768,
      "num_tokens": 763691557.0,
      "step": 23929
    },
    {
      "entropy": 0.4872906804084778,
      "epoch": 2.2016779661991595,
      "grad_norm": 0.15652237832546234,
      "learning_rate": 2.66139172570307e-05,
      "loss": 0.477,
      "mean_token_accuracy": 0.8477944284677505,
      "num_tokens": 763723506.0,
      "step": 23930
    },
    {
      "entropy": 0.5306161171756685,
      "epoch": 2.2017699725706006,
      "grad_norm": 0.16478799283504486,
      "learning_rate": 2.661085043088907e-05,
      "loss": 0.516,
      "mean_token_accuracy": 0.8370342776179314,
      "num_tokens": 763755497.0,
      "step": 23931
    },
    {
      "entropy": 0.43972060456871986,
      "epoch": 2.2018619789420417,
      "grad_norm": 0.15708491206169128,
      "learning_rate": 2.660778360474745e-05,
      "loss": 0.4183,
      "mean_token_accuracy": 0.8633726686239243,
      "num_tokens": 763786954.0,
      "step": 23932
    },
    {
      "entropy": 0.572416833601892,
      "epoch": 2.201953985313483,
      "grad_norm": 0.1649593561887741,
      "learning_rate": 2.660471677860582e-05,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8268798552453518,
      "num_tokens": 763819145.0,
      "step": 23933
    },
    {
      "entropy": 0.6032143402844667,
      "epoch": 2.202045991684924,
      "grad_norm": 0.1692012995481491,
      "learning_rate": 2.6601649952464196e-05,
      "loss": 0.5912,
      "mean_token_accuracy": 0.8137346468865871,
      "num_tokens": 763851004.0,
      "step": 23934
    },
    {
      "entropy": 0.6101683676242828,
      "epoch": 2.2021379980563656,
      "grad_norm": 0.1689339280128479,
      "learning_rate": 2.6598583126322567e-05,
      "loss": 0.5981,
      "mean_token_accuracy": 0.8161579482257366,
      "num_tokens": 763883536.0,
      "step": 23935
    },
    {
      "entropy": 0.5444315392524004,
      "epoch": 2.2022300044278067,
      "grad_norm": 0.16083446145057678,
      "learning_rate": 2.6595516300180945e-05,
      "loss": 0.5332,
      "mean_token_accuracy": 0.834735207259655,
      "num_tokens": 763915418.0,
      "step": 23936
    },
    {
      "entropy": 0.6021169815212488,
      "epoch": 2.202322010799248,
      "grad_norm": 0.16972903907299042,
      "learning_rate": 2.6592449474039316e-05,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8146600984036922,
      "num_tokens": 763947388.0,
      "step": 23937
    },
    {
      "entropy": 0.5807130998000503,
      "epoch": 2.202414017170689,
      "grad_norm": 0.17059025168418884,
      "learning_rate": 2.6589382647897694e-05,
      "loss": 0.5716,
      "mean_token_accuracy": 0.8196711838245392,
      "num_tokens": 763978728.0,
      "step": 23938
    },
    {
      "entropy": 0.5204531662166119,
      "epoch": 2.20250602354213,
      "grad_norm": 0.15295737981796265,
      "learning_rate": 2.6586315821756065e-05,
      "loss": 0.5008,
      "mean_token_accuracy": 0.8423162512481213,
      "num_tokens": 764010956.0,
      "step": 23939
    },
    {
      "entropy": 0.4803571356460452,
      "epoch": 2.2025980299135717,
      "grad_norm": 0.1576010286808014,
      "learning_rate": 2.658324899561444e-05,
      "loss": 0.459,
      "mean_token_accuracy": 0.8552361726760864,
      "num_tokens": 764042795.0,
      "step": 23940
    },
    {
      "entropy": 0.5886522065848112,
      "epoch": 2.202690036285013,
      "grad_norm": 0.1711193472146988,
      "learning_rate": 2.658018216947281e-05,
      "loss": 0.5765,
      "mean_token_accuracy": 0.8222906067967415,
      "num_tokens": 764074277.0,
      "step": 23941
    },
    {
      "entropy": 0.5970562603324652,
      "epoch": 2.202782042656454,
      "grad_norm": 0.17231369018554688,
      "learning_rate": 2.657711534333119e-05,
      "loss": 0.5912,
      "mean_token_accuracy": 0.8128959015011787,
      "num_tokens": 764104745.0,
      "step": 23942
    },
    {
      "entropy": 0.5261580310761929,
      "epoch": 2.202874049027895,
      "grad_norm": 0.16394658386707306,
      "learning_rate": 2.657404851718956e-05,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8388221971690655,
      "num_tokens": 764136715.0,
      "step": 23943
    },
    {
      "entropy": 0.5508918222039938,
      "epoch": 2.202966055399336,
      "grad_norm": 0.16913574934005737,
      "learning_rate": 2.6570981691047935e-05,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8286602273583412,
      "num_tokens": 764169090.0,
      "step": 23944
    },
    {
      "entropy": 0.5544403516687453,
      "epoch": 2.203058061770778,
      "grad_norm": 0.16548798978328705,
      "learning_rate": 2.6567914864906307e-05,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8301277682185173,
      "num_tokens": 764200946.0,
      "step": 23945
    },
    {
      "entropy": 0.5053687272593379,
      "epoch": 2.203150068142219,
      "grad_norm": 0.16305659711360931,
      "learning_rate": 2.6564848038764685e-05,
      "loss": 0.4884,
      "mean_token_accuracy": 0.8418848253786564,
      "num_tokens": 764232615.0,
      "step": 23946
    },
    {
      "entropy": 0.5848883632570505,
      "epoch": 2.20324207451366,
      "grad_norm": 0.16821731626987457,
      "learning_rate": 2.656178121262306e-05,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8217972815036774,
      "num_tokens": 764263510.0,
      "step": 23947
    },
    {
      "entropy": 0.5273492205888033,
      "epoch": 2.203334080885101,
      "grad_norm": 0.15823791921138763,
      "learning_rate": 2.655871438648143e-05,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8405602499842644,
      "num_tokens": 764295590.0,
      "step": 23948
    },
    {
      "entropy": 0.5648670317605138,
      "epoch": 2.2034260872565423,
      "grad_norm": 0.16587208211421967,
      "learning_rate": 2.655564756033981e-05,
      "loss": 0.559,
      "mean_token_accuracy": 0.8235930129885674,
      "num_tokens": 764327416.0,
      "step": 23949
    },
    {
      "entropy": 0.5772586259990931,
      "epoch": 2.203518093627984,
      "grad_norm": 0.1652827113866806,
      "learning_rate": 2.655258073419818e-05,
      "loss": 0.5736,
      "mean_token_accuracy": 0.8208182938396931,
      "num_tokens": 764359525.0,
      "step": 23950
    },
    {
      "entropy": 0.5854279845952988,
      "epoch": 2.203610099999425,
      "grad_norm": 0.1688268482685089,
      "learning_rate": 2.6549513908056554e-05,
      "loss": 0.5654,
      "mean_token_accuracy": 0.8184000179171562,
      "num_tokens": 764390729.0,
      "step": 23951
    },
    {
      "entropy": 0.562949325889349,
      "epoch": 2.203702106370866,
      "grad_norm": 0.17031051218509674,
      "learning_rate": 2.6546447081914926e-05,
      "loss": 0.5453,
      "mean_token_accuracy": 0.829937506467104,
      "num_tokens": 764422121.0,
      "step": 23952
    },
    {
      "entropy": 0.45465375669300556,
      "epoch": 2.2037941127423073,
      "grad_norm": 0.1521497368812561,
      "learning_rate": 2.6543380255773304e-05,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8595958612859249,
      "num_tokens": 764452898.0,
      "step": 23953
    },
    {
      "entropy": 0.5852860864251852,
      "epoch": 2.2038861191137484,
      "grad_norm": 0.16529709100723267,
      "learning_rate": 2.6540313429631675e-05,
      "loss": 0.565,
      "mean_token_accuracy": 0.8235778249800205,
      "num_tokens": 764485139.0,
      "step": 23954
    },
    {
      "entropy": 0.4771058149635792,
      "epoch": 2.20397812548519,
      "grad_norm": 0.153773233294487,
      "learning_rate": 2.6537246603490053e-05,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8516406305134296,
      "num_tokens": 764517148.0,
      "step": 23955
    },
    {
      "entropy": 0.5097680049948394,
      "epoch": 2.204070131856631,
      "grad_norm": 0.15332327783107758,
      "learning_rate": 2.653417977734842e-05,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8446888662874699,
      "num_tokens": 764549310.0,
      "step": 23956
    },
    {
      "entropy": 0.42364570731297135,
      "epoch": 2.2041621382280723,
      "grad_norm": 0.149727463722229,
      "learning_rate": 2.65311129512068e-05,
      "loss": 0.4104,
      "mean_token_accuracy": 0.8679227381944656,
      "num_tokens": 764581425.0,
      "step": 23957
    },
    {
      "entropy": 0.4483633217168972,
      "epoch": 2.2042541445995134,
      "grad_norm": 0.1544465720653534,
      "learning_rate": 2.652804612506517e-05,
      "loss": 0.4411,
      "mean_token_accuracy": 0.864039234817028,
      "num_tokens": 764613894.0,
      "step": 23958
    },
    {
      "entropy": 0.4533162140287459,
      "epoch": 2.2043461509709545,
      "grad_norm": 0.1518765687942505,
      "learning_rate": 2.6524979298923548e-05,
      "loss": 0.4402,
      "mean_token_accuracy": 0.8602524399757385,
      "num_tokens": 764645786.0,
      "step": 23959
    },
    {
      "entropy": 0.5654987208545208,
      "epoch": 2.204438157342396,
      "grad_norm": 0.17083704471588135,
      "learning_rate": 2.652191247278192e-05,
      "loss": 0.5589,
      "mean_token_accuracy": 0.8216442130506039,
      "num_tokens": 764677723.0,
      "step": 23960
    },
    {
      "entropy": 0.5567329805344343,
      "epoch": 2.2045301637138373,
      "grad_norm": 0.16560840606689453,
      "learning_rate": 2.6518845646640294e-05,
      "loss": 0.5393,
      "mean_token_accuracy": 0.8321106657385826,
      "num_tokens": 764709579.0,
      "step": 23961
    },
    {
      "entropy": 0.45920155360363424,
      "epoch": 2.2046221700852784,
      "grad_norm": 0.1559060961008072,
      "learning_rate": 2.6515778820498665e-05,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8571662679314613,
      "num_tokens": 764740770.0,
      "step": 23962
    },
    {
      "entropy": 0.43208590941503644,
      "epoch": 2.2047141764567195,
      "grad_norm": 0.15608006715774536,
      "learning_rate": 2.6512711994357043e-05,
      "loss": 0.4264,
      "mean_token_accuracy": 0.8672691136598587,
      "num_tokens": 764772688.0,
      "step": 23963
    },
    {
      "entropy": 0.5136819817125797,
      "epoch": 2.2048061828281607,
      "grad_norm": 0.15905331075191498,
      "learning_rate": 2.6509645168215415e-05,
      "loss": 0.5061,
      "mean_token_accuracy": 0.8392887338995934,
      "num_tokens": 764804889.0,
      "step": 23964
    },
    {
      "entropy": 0.6028735786676407,
      "epoch": 2.2048981891996022,
      "grad_norm": 0.16995371878147125,
      "learning_rate": 2.650657834207379e-05,
      "loss": 0.6035,
      "mean_token_accuracy": 0.8097134083509445,
      "num_tokens": 764836963.0,
      "step": 23965
    },
    {
      "entropy": 0.48672191333025694,
      "epoch": 2.2049901955710434,
      "grad_norm": 0.15958307683467865,
      "learning_rate": 2.650351151593216e-05,
      "loss": 0.4749,
      "mean_token_accuracy": 0.8476593904197216,
      "num_tokens": 764868612.0,
      "step": 23966
    },
    {
      "entropy": 0.4799659373238683,
      "epoch": 2.2050822019424845,
      "grad_norm": 0.16033105552196503,
      "learning_rate": 2.650044468979054e-05,
      "loss": 0.4732,
      "mean_token_accuracy": 0.8535904511809349,
      "num_tokens": 764899732.0,
      "step": 23967
    },
    {
      "entropy": 0.5242319600656629,
      "epoch": 2.2051742083139256,
      "grad_norm": 0.16019481420516968,
      "learning_rate": 2.649737786364891e-05,
      "loss": 0.5108,
      "mean_token_accuracy": 0.836856085807085,
      "num_tokens": 764931421.0,
      "step": 23968
    },
    {
      "entropy": 0.5513105038553476,
      "epoch": 2.2052662146853668,
      "grad_norm": 0.1662873923778534,
      "learning_rate": 2.6494311037507285e-05,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8286644034087658,
      "num_tokens": 764963168.0,
      "step": 23969
    },
    {
      "entropy": 0.5335994530469179,
      "epoch": 2.2053582210568083,
      "grad_norm": 0.1656167358160019,
      "learning_rate": 2.6491244211365656e-05,
      "loss": 0.5118,
      "mean_token_accuracy": 0.8346931077539921,
      "num_tokens": 764994064.0,
      "step": 23970
    },
    {
      "entropy": 0.465512472204864,
      "epoch": 2.2054502274282495,
      "grad_norm": 0.15478602051734924,
      "learning_rate": 2.6488177385224034e-05,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8558158539235592,
      "num_tokens": 765026187.0,
      "step": 23971
    },
    {
      "entropy": 0.5632661283016205,
      "epoch": 2.2055422337996906,
      "grad_norm": 0.16487561166286469,
      "learning_rate": 2.6485110559082405e-05,
      "loss": 0.5561,
      "mean_token_accuracy": 0.8245341069996357,
      "num_tokens": 765058211.0,
      "step": 23972
    },
    {
      "entropy": 0.5019720112904906,
      "epoch": 2.2056342401711317,
      "grad_norm": 0.15683336555957794,
      "learning_rate": 2.648204373294078e-05,
      "loss": 0.4827,
      "mean_token_accuracy": 0.8448794335126877,
      "num_tokens": 765089735.0,
      "step": 23973
    },
    {
      "entropy": 0.5615366455167532,
      "epoch": 2.205726246542573,
      "grad_norm": 0.1638963222503662,
      "learning_rate": 2.647897690679915e-05,
      "loss": 0.5393,
      "mean_token_accuracy": 0.8295859284698963,
      "num_tokens": 765121926.0,
      "step": 23974
    },
    {
      "entropy": 0.5336444722488523,
      "epoch": 2.2058182529140145,
      "grad_norm": 0.164345383644104,
      "learning_rate": 2.647591008065753e-05,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8360786326229572,
      "num_tokens": 765154021.0,
      "step": 23975
    },
    {
      "entropy": 0.5615490637719631,
      "epoch": 2.2059102592854556,
      "grad_norm": 0.1597108393907547,
      "learning_rate": 2.6472843254515907e-05,
      "loss": 0.548,
      "mean_token_accuracy": 0.8275507651269436,
      "num_tokens": 765185989.0,
      "step": 23976
    },
    {
      "entropy": 0.5768921477720141,
      "epoch": 2.2060022656568967,
      "grad_norm": 0.16978813707828522,
      "learning_rate": 2.6469776428374278e-05,
      "loss": 0.5609,
      "mean_token_accuracy": 0.8230589404702187,
      "num_tokens": 765217337.0,
      "step": 23977
    },
    {
      "entropy": 0.4558490440249443,
      "epoch": 2.206094272028338,
      "grad_norm": 0.1494910567998886,
      "learning_rate": 2.6466709602232653e-05,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8581074178218842,
      "num_tokens": 765249599.0,
      "step": 23978
    },
    {
      "entropy": 0.515255355508998,
      "epoch": 2.206186278399779,
      "grad_norm": 0.1569434404373169,
      "learning_rate": 2.6463642776091024e-05,
      "loss": 0.5083,
      "mean_token_accuracy": 0.8389475382864475,
      "num_tokens": 765281775.0,
      "step": 23979
    },
    {
      "entropy": 0.54369224794209,
      "epoch": 2.2062782847712206,
      "grad_norm": 0.16864365339279175,
      "learning_rate": 2.6460575949949402e-05,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8330983221530914,
      "num_tokens": 765313509.0,
      "step": 23980
    },
    {
      "entropy": 0.5298580916132778,
      "epoch": 2.2063702911426617,
      "grad_norm": 0.16364848613739014,
      "learning_rate": 2.6457509123807774e-05,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8369169868528843,
      "num_tokens": 765345256.0,
      "step": 23981
    },
    {
      "entropy": 0.5910944864153862,
      "epoch": 2.206462297514103,
      "grad_norm": 0.1686873435974121,
      "learning_rate": 2.6454442297666148e-05,
      "loss": 0.5918,
      "mean_token_accuracy": 0.8125168345868587,
      "num_tokens": 765377634.0,
      "step": 23982
    },
    {
      "entropy": 0.4699786650016904,
      "epoch": 2.206554303885544,
      "grad_norm": 0.16271041333675385,
      "learning_rate": 2.645137547152452e-05,
      "loss": 0.4604,
      "mean_token_accuracy": 0.8542009703814983,
      "num_tokens": 765409283.0,
      "step": 23983
    },
    {
      "entropy": 0.4875414287671447,
      "epoch": 2.206646310256985,
      "grad_norm": 0.15592241287231445,
      "learning_rate": 2.6448308645382897e-05,
      "loss": 0.4769,
      "mean_token_accuracy": 0.8494719602167606,
      "num_tokens": 765441131.0,
      "step": 23984
    },
    {
      "entropy": 0.5107776448130608,
      "epoch": 2.2067383166284267,
      "grad_norm": 0.15358436107635498,
      "learning_rate": 2.644524181924127e-05,
      "loss": 0.4944,
      "mean_token_accuracy": 0.8430284969508648,
      "num_tokens": 765473609.0,
      "step": 23985
    },
    {
      "entropy": 0.5104899303987622,
      "epoch": 2.206830322999868,
      "grad_norm": 0.15376712381839752,
      "learning_rate": 2.6442174993099643e-05,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8415204994380474,
      "num_tokens": 765505836.0,
      "step": 23986
    },
    {
      "entropy": 0.5676135281100869,
      "epoch": 2.206922329371309,
      "grad_norm": 0.17077681422233582,
      "learning_rate": 2.6439108166958015e-05,
      "loss": 0.5576,
      "mean_token_accuracy": 0.8245835117995739,
      "num_tokens": 765537673.0,
      "step": 23987
    },
    {
      "entropy": 0.5258929450064898,
      "epoch": 2.20701433574275,
      "grad_norm": 0.16340123116970062,
      "learning_rate": 2.6436041340816393e-05,
      "loss": 0.519,
      "mean_token_accuracy": 0.8344720900058746,
      "num_tokens": 765569307.0,
      "step": 23988
    },
    {
      "entropy": 0.5771164931356907,
      "epoch": 2.207106342114191,
      "grad_norm": 0.16782891750335693,
      "learning_rate": 2.6432974514674764e-05,
      "loss": 0.5663,
      "mean_token_accuracy": 0.8231719397008419,
      "num_tokens": 765601591.0,
      "step": 23989
    },
    {
      "entropy": 0.5906593706458807,
      "epoch": 2.207198348485633,
      "grad_norm": 0.16488218307495117,
      "learning_rate": 2.642990768853314e-05,
      "loss": 0.5679,
      "mean_token_accuracy": 0.8244653642177582,
      "num_tokens": 765633729.0,
      "step": 23990
    },
    {
      "entropy": 0.5455995248630643,
      "epoch": 2.207290354857074,
      "grad_norm": 0.16211414337158203,
      "learning_rate": 2.642684086239151e-05,
      "loss": 0.5357,
      "mean_token_accuracy": 0.8312725648283958,
      "num_tokens": 765666247.0,
      "step": 23991
    },
    {
      "entropy": 0.5599164823070168,
      "epoch": 2.207382361228515,
      "grad_norm": 0.16332457959651947,
      "learning_rate": 2.6423774036249888e-05,
      "loss": 0.5528,
      "mean_token_accuracy": 0.8257576636970043,
      "num_tokens": 765697775.0,
      "step": 23992
    },
    {
      "entropy": 0.6077757123857737,
      "epoch": 2.207474367599956,
      "grad_norm": 0.1721692532300949,
      "learning_rate": 2.642070721010826e-05,
      "loss": 0.5978,
      "mean_token_accuracy": 0.813025064766407,
      "num_tokens": 765729140.0,
      "step": 23993
    },
    {
      "entropy": 0.49618340376764536,
      "epoch": 2.2075663739713973,
      "grad_norm": 0.15353308618068695,
      "learning_rate": 2.6417640383966634e-05,
      "loss": 0.4863,
      "mean_token_accuracy": 0.8452802933752537,
      "num_tokens": 765761760.0,
      "step": 23994
    },
    {
      "entropy": 0.46226145047694445,
      "epoch": 2.207658380342839,
      "grad_norm": 0.15009719133377075,
      "learning_rate": 2.6414573557825005e-05,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8580060303211212,
      "num_tokens": 765794433.0,
      "step": 23995
    },
    {
      "entropy": 0.49155959114432335,
      "epoch": 2.20775038671428,
      "grad_norm": 0.1590820550918579,
      "learning_rate": 2.6411506731683383e-05,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8498465940356255,
      "num_tokens": 765826707.0,
      "step": 23996
    },
    {
      "entropy": 0.5314888518769294,
      "epoch": 2.207842393085721,
      "grad_norm": 0.16655096411705017,
      "learning_rate": 2.6408439905541754e-05,
      "loss": 0.5217,
      "mean_token_accuracy": 0.8346432559192181,
      "num_tokens": 765858900.0,
      "step": 23997
    },
    {
      "entropy": 0.4881458142772317,
      "epoch": 2.2079343994571623,
      "grad_norm": 0.16165980696678162,
      "learning_rate": 2.6405373079400132e-05,
      "loss": 0.4804,
      "mean_token_accuracy": 0.8452674597501755,
      "num_tokens": 765890621.0,
      "step": 23998
    },
    {
      "entropy": 0.600182079244405,
      "epoch": 2.2080264058286034,
      "grad_norm": 0.1756221354007721,
      "learning_rate": 2.6402306253258504e-05,
      "loss": 0.5805,
      "mean_token_accuracy": 0.8165631592273712,
      "num_tokens": 765922578.0,
      "step": 23999
    },
    {
      "entropy": 0.5317483758553863,
      "epoch": 2.208118412200045,
      "grad_norm": 0.16364751756191254,
      "learning_rate": 2.6399239427116878e-05,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8356156684458256,
      "num_tokens": 765954167.0,
      "step": 24000
    },
    {
      "entropy": 0.36891785194166005,
      "epoch": 2.208210418571486,
      "grad_norm": 0.1393590122461319,
      "learning_rate": 2.639617260097525e-05,
      "loss": 0.3523,
      "mean_token_accuracy": 0.8860837742686272,
      "num_tokens": 765986277.0,
      "step": 24001
    },
    {
      "entropy": 0.5384258171543479,
      "epoch": 2.2083024249429273,
      "grad_norm": 0.1675165295600891,
      "learning_rate": 2.6393105774833628e-05,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8309107348322868,
      "num_tokens": 766017434.0,
      "step": 24002
    },
    {
      "entropy": 0.4380256317090243,
      "epoch": 2.2083944313143684,
      "grad_norm": 0.14572268724441528,
      "learning_rate": 2.6390038948692e-05,
      "loss": 0.4106,
      "mean_token_accuracy": 0.8700767420232296,
      "num_tokens": 766049357.0,
      "step": 24003
    },
    {
      "entropy": 0.49978202814236283,
      "epoch": 2.2084864376858095,
      "grad_norm": 0.15562298893928528,
      "learning_rate": 2.6386972122550373e-05,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8456371836364269,
      "num_tokens": 766081313.0,
      "step": 24004
    },
    {
      "entropy": 0.600912207737565,
      "epoch": 2.208578444057251,
      "grad_norm": 0.16925887763500214,
      "learning_rate": 2.6383905296408745e-05,
      "loss": 0.5977,
      "mean_token_accuracy": 0.8079326637089252,
      "num_tokens": 766113139.0,
      "step": 24005
    },
    {
      "entropy": 0.5665336735546589,
      "epoch": 2.2086704504286923,
      "grad_norm": 0.16540034115314484,
      "learning_rate": 2.6380838470267123e-05,
      "loss": 0.545,
      "mean_token_accuracy": 0.8320765309035778,
      "num_tokens": 766144724.0,
      "step": 24006
    },
    {
      "entropy": 0.5072391647845507,
      "epoch": 2.2087624568001334,
      "grad_norm": 0.15450666844844818,
      "learning_rate": 2.6377771644125497e-05,
      "loss": 0.4887,
      "mean_token_accuracy": 0.8428581431508064,
      "num_tokens": 766176651.0,
      "step": 24007
    },
    {
      "entropy": 0.5103387013077736,
      "epoch": 2.2088544631715745,
      "grad_norm": 0.15476581454277039,
      "learning_rate": 2.637470481798387e-05,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8417886048555374,
      "num_tokens": 766208804.0,
      "step": 24008
    },
    {
      "entropy": 0.5539711527526379,
      "epoch": 2.2089464695430157,
      "grad_norm": 0.1705973744392395,
      "learning_rate": 2.6371637991842247e-05,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8228261321783066,
      "num_tokens": 766240411.0,
      "step": 24009
    },
    {
      "entropy": 0.5857567719649523,
      "epoch": 2.2090384759144572,
      "grad_norm": 0.1720418483018875,
      "learning_rate": 2.6368571165700618e-05,
      "loss": 0.5782,
      "mean_token_accuracy": 0.8193559423089027,
      "num_tokens": 766272550.0,
      "step": 24010
    },
    {
      "entropy": 0.5432122210040689,
      "epoch": 2.2091304822858984,
      "grad_norm": 0.1626383364200592,
      "learning_rate": 2.6365504339558993e-05,
      "loss": 0.5176,
      "mean_token_accuracy": 0.8330299369990826,
      "num_tokens": 766304062.0,
      "step": 24011
    },
    {
      "entropy": 0.5488332007080317,
      "epoch": 2.2092224886573395,
      "grad_norm": 0.16398142278194427,
      "learning_rate": 2.6362437513417364e-05,
      "loss": 0.5381,
      "mean_token_accuracy": 0.8294438868761063,
      "num_tokens": 766336431.0,
      "step": 24012
    },
    {
      "entropy": 0.5567692518234253,
      "epoch": 2.2093144950287806,
      "grad_norm": 0.16571888327598572,
      "learning_rate": 2.6359370687275742e-05,
      "loss": 0.5443,
      "mean_token_accuracy": 0.8304989486932755,
      "num_tokens": 766368621.0,
      "step": 24013
    },
    {
      "entropy": 0.4961389824748039,
      "epoch": 2.2094065014002218,
      "grad_norm": 0.1612730771303177,
      "learning_rate": 2.6356303861134113e-05,
      "loss": 0.4831,
      "mean_token_accuracy": 0.8472606390714645,
      "num_tokens": 766399962.0,
      "step": 24014
    },
    {
      "entropy": 0.5078646456822753,
      "epoch": 2.2094985077716633,
      "grad_norm": 0.15545666217803955,
      "learning_rate": 2.635323703499249e-05,
      "loss": 0.5042,
      "mean_token_accuracy": 0.8417026661336422,
      "num_tokens": 766431678.0,
      "step": 24015
    },
    {
      "entropy": 0.47606306429952383,
      "epoch": 2.2095905141431045,
      "grad_norm": 0.1533505916595459,
      "learning_rate": 2.635017020885086e-05,
      "loss": 0.4604,
      "mean_token_accuracy": 0.851064894348383,
      "num_tokens": 766464052.0,
      "step": 24016
    },
    {
      "entropy": 0.5527056446298957,
      "epoch": 2.2096825205145456,
      "grad_norm": 0.16247284412384033,
      "learning_rate": 2.6347103382709237e-05,
      "loss": 0.5333,
      "mean_token_accuracy": 0.8341569490730762,
      "num_tokens": 766495427.0,
      "step": 24017
    },
    {
      "entropy": 0.5685329944826663,
      "epoch": 2.2097745268859867,
      "grad_norm": 0.16034391522407532,
      "learning_rate": 2.634403655656761e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8258161768317223,
      "num_tokens": 766527551.0,
      "step": 24018
    },
    {
      "entropy": 0.5150396646931767,
      "epoch": 2.209866533257428,
      "grad_norm": 0.15768472850322723,
      "learning_rate": 2.6340969730425986e-05,
      "loss": 0.4967,
      "mean_token_accuracy": 0.8450462967157364,
      "num_tokens": 766559987.0,
      "step": 24019
    },
    {
      "entropy": 0.4948729551397264,
      "epoch": 2.2099585396288695,
      "grad_norm": 0.16389727592468262,
      "learning_rate": 2.6337902904284358e-05,
      "loss": 0.4876,
      "mean_token_accuracy": 0.8448190279304981,
      "num_tokens": 766592299.0,
      "step": 24020
    },
    {
      "entropy": 0.5510890800505877,
      "epoch": 2.2100505460003106,
      "grad_norm": 0.16484488546848297,
      "learning_rate": 2.6334836078142732e-05,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8282687962055206,
      "num_tokens": 766624495.0,
      "step": 24021
    },
    {
      "entropy": 0.5893961284309626,
      "epoch": 2.2101425523717517,
      "grad_norm": 0.16874688863754272,
      "learning_rate": 2.6331769252001104e-05,
      "loss": 0.5773,
      "mean_token_accuracy": 0.8204452618956566,
      "num_tokens": 766656890.0,
      "step": 24022
    },
    {
      "entropy": 0.484337393194437,
      "epoch": 2.210234558743193,
      "grad_norm": 0.16068217158317566,
      "learning_rate": 2.632870242585948e-05,
      "loss": 0.4721,
      "mean_token_accuracy": 0.8488460630178452,
      "num_tokens": 766688480.0,
      "step": 24023
    },
    {
      "entropy": 0.4782763598486781,
      "epoch": 2.210326565114634,
      "grad_norm": 0.16014021635055542,
      "learning_rate": 2.6325635599717853e-05,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8506838269531727,
      "num_tokens": 766720106.0,
      "step": 24024
    },
    {
      "entropy": 0.5352312009781599,
      "epoch": 2.2104185714860756,
      "grad_norm": 0.16480456292629242,
      "learning_rate": 2.6322568773576228e-05,
      "loss": 0.5319,
      "mean_token_accuracy": 0.8301126286387444,
      "num_tokens": 766752370.0,
      "step": 24025
    },
    {
      "entropy": 0.557702730409801,
      "epoch": 2.2105105778575167,
      "grad_norm": 0.16410891711711884,
      "learning_rate": 2.63195019474346e-05,
      "loss": 0.5484,
      "mean_token_accuracy": 0.8277337588369846,
      "num_tokens": 766784815.0,
      "step": 24026
    },
    {
      "entropy": 0.5436742510646582,
      "epoch": 2.210602584228958,
      "grad_norm": 0.16632165014743805,
      "learning_rate": 2.6316435121292977e-05,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8354789018630981,
      "num_tokens": 766816474.0,
      "step": 24027
    },
    {
      "entropy": 0.5904379438143224,
      "epoch": 2.210694590600399,
      "grad_norm": 0.17232416570186615,
      "learning_rate": 2.6313368295151348e-05,
      "loss": 0.5722,
      "mean_token_accuracy": 0.818608608096838,
      "num_tokens": 766848243.0,
      "step": 24028
    },
    {
      "entropy": 0.48698143963702023,
      "epoch": 2.21078659697184,
      "grad_norm": 0.15377192199230194,
      "learning_rate": 2.6310301469009723e-05,
      "loss": 0.4799,
      "mean_token_accuracy": 0.8496409021317959,
      "num_tokens": 766880105.0,
      "step": 24029
    },
    {
      "entropy": 0.5313947191461921,
      "epoch": 2.2108786033432817,
      "grad_norm": 0.15833313763141632,
      "learning_rate": 2.6307234642868094e-05,
      "loss": 0.5198,
      "mean_token_accuracy": 0.831680104136467,
      "num_tokens": 766911256.0,
      "step": 24030
    },
    {
      "entropy": 0.4975805194117129,
      "epoch": 2.210970609714723,
      "grad_norm": 0.1586233377456665,
      "learning_rate": 2.6304167816726472e-05,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8489290066063404,
      "num_tokens": 766942891.0,
      "step": 24031
    },
    {
      "entropy": 0.502127829939127,
      "epoch": 2.211062616086164,
      "grad_norm": 0.16252484917640686,
      "learning_rate": 2.6301100990584843e-05,
      "loss": 0.4963,
      "mean_token_accuracy": 0.8414627648890018,
      "num_tokens": 766975176.0,
      "step": 24032
    },
    {
      "entropy": 0.5865205486770719,
      "epoch": 2.211154622457605,
      "grad_norm": 0.1646774262189865,
      "learning_rate": 2.6298034164443218e-05,
      "loss": 0.5724,
      "mean_token_accuracy": 0.8170786909759045,
      "num_tokens": 767007143.0,
      "step": 24033
    },
    {
      "entropy": 0.5321864411234856,
      "epoch": 2.211246628829046,
      "grad_norm": 0.16295655071735382,
      "learning_rate": 2.629496733830159e-05,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8364315889775753,
      "num_tokens": 767039217.0,
      "step": 24034
    },
    {
      "entropy": 0.4676116667687893,
      "epoch": 2.211338635200488,
      "grad_norm": 0.14963535964488983,
      "learning_rate": 2.6291900512159967e-05,
      "loss": 0.4507,
      "mean_token_accuracy": 0.8548535145819187,
      "num_tokens": 767071185.0,
      "step": 24035
    },
    {
      "entropy": 0.5452713733538985,
      "epoch": 2.211430641571929,
      "grad_norm": 0.1617252081632614,
      "learning_rate": 2.628883368601834e-05,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8300256468355656,
      "num_tokens": 767103953.0,
      "step": 24036
    },
    {
      "entropy": 0.4303496638312936,
      "epoch": 2.21152264794337,
      "grad_norm": 0.1488451212644577,
      "learning_rate": 2.6285766859876717e-05,
      "loss": 0.4183,
      "mean_token_accuracy": 0.8622878417372704,
      "num_tokens": 767136508.0,
      "step": 24037
    },
    {
      "entropy": 0.4428932892624289,
      "epoch": 2.211614654314811,
      "grad_norm": 0.14992669224739075,
      "learning_rate": 2.628270003373509e-05,
      "loss": 0.4296,
      "mean_token_accuracy": 0.8629077337682247,
      "num_tokens": 767169014.0,
      "step": 24038
    },
    {
      "entropy": 0.5949867460876703,
      "epoch": 2.2117066606862523,
      "grad_norm": 0.1625298410654068,
      "learning_rate": 2.6279633207593462e-05,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8201784417033195,
      "num_tokens": 767201480.0,
      "step": 24039
    },
    {
      "entropy": 0.5247029438614845,
      "epoch": 2.211798667057694,
      "grad_norm": 0.16338884830474854,
      "learning_rate": 2.627656638145184e-05,
      "loss": 0.5154,
      "mean_token_accuracy": 0.8378831371665001,
      "num_tokens": 767232926.0,
      "step": 24040
    },
    {
      "entropy": 0.4965231576934457,
      "epoch": 2.211890673429135,
      "grad_norm": 0.16057677567005157,
      "learning_rate": 2.6273499555310212e-05,
      "loss": 0.4822,
      "mean_token_accuracy": 0.8481227681040764,
      "num_tokens": 767264710.0,
      "step": 24041
    },
    {
      "entropy": 0.46862412616610527,
      "epoch": 2.211982679800576,
      "grad_norm": 0.1505243182182312,
      "learning_rate": 2.6270432729168586e-05,
      "loss": 0.453,
      "mean_token_accuracy": 0.855161726474762,
      "num_tokens": 767296762.0,
      "step": 24042
    },
    {
      "entropy": 0.5755218183621764,
      "epoch": 2.2120746861720173,
      "grad_norm": 0.176219180226326,
      "learning_rate": 2.6267365903026958e-05,
      "loss": 0.5686,
      "mean_token_accuracy": 0.8214490860700607,
      "num_tokens": 767328633.0,
      "step": 24043
    },
    {
      "entropy": 0.4739637959282845,
      "epoch": 2.2121666925434584,
      "grad_norm": 0.15079019963741302,
      "learning_rate": 2.6264299076885336e-05,
      "loss": 0.4592,
      "mean_token_accuracy": 0.8535439856350422,
      "num_tokens": 767360824.0,
      "step": 24044
    },
    {
      "entropy": 0.46660470543429255,
      "epoch": 2.2122586989149,
      "grad_norm": 0.15454262495040894,
      "learning_rate": 2.6261232250743707e-05,
      "loss": 0.446,
      "mean_token_accuracy": 0.8570865243673325,
      "num_tokens": 767392545.0,
      "step": 24045
    },
    {
      "entropy": 0.4369996802415699,
      "epoch": 2.212350705286341,
      "grad_norm": 0.150192990899086,
      "learning_rate": 2.625816542460208e-05,
      "loss": 0.4195,
      "mean_token_accuracy": 0.8631003946065903,
      "num_tokens": 767424444.0,
      "step": 24046
    },
    {
      "entropy": 0.5171256372705102,
      "epoch": 2.2124427116577823,
      "grad_norm": 0.16650909185409546,
      "learning_rate": 2.6255098598460453e-05,
      "loss": 0.4988,
      "mean_token_accuracy": 0.8423870541155338,
      "num_tokens": 767456255.0,
      "step": 24047
    },
    {
      "entropy": 0.5108200404793024,
      "epoch": 2.2125347180292234,
      "grad_norm": 0.16330547630786896,
      "learning_rate": 2.625203177231883e-05,
      "loss": 0.5037,
      "mean_token_accuracy": 0.840021125972271,
      "num_tokens": 767488749.0,
      "step": 24048
    },
    {
      "entropy": 0.5816919449716806,
      "epoch": 2.2126267244006645,
      "grad_norm": 0.16694240272045135,
      "learning_rate": 2.6248964946177202e-05,
      "loss": 0.5691,
      "mean_token_accuracy": 0.8177032358944416,
      "num_tokens": 767520996.0,
      "step": 24049
    },
    {
      "entropy": 0.4476510928943753,
      "epoch": 2.212718730772106,
      "grad_norm": 0.1592770367860794,
      "learning_rate": 2.6245898120035577e-05,
      "loss": 0.4355,
      "mean_token_accuracy": 0.8627542294561863,
      "num_tokens": 767552897.0,
      "step": 24050
    },
    {
      "entropy": 0.5284497775137424,
      "epoch": 2.2128107371435473,
      "grad_norm": 0.16278137266635895,
      "learning_rate": 2.6242831293893948e-05,
      "loss": 0.5143,
      "mean_token_accuracy": 0.8358217440545559,
      "num_tokens": 767584523.0,
      "step": 24051
    },
    {
      "entropy": 0.5546589283039793,
      "epoch": 2.2129027435149884,
      "grad_norm": 0.1642400473356247,
      "learning_rate": 2.6239764467752326e-05,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8307456597685814,
      "num_tokens": 767616189.0,
      "step": 24052
    },
    {
      "entropy": 0.5767873842269182,
      "epoch": 2.2129947498864295,
      "grad_norm": 0.16769619286060333,
      "learning_rate": 2.6236697641610697e-05,
      "loss": 0.5624,
      "mean_token_accuracy": 0.8210429884493351,
      "num_tokens": 767647104.0,
      "step": 24053
    },
    {
      "entropy": 0.5031143622472882,
      "epoch": 2.2130867562578707,
      "grad_norm": 0.16199207305908203,
      "learning_rate": 2.6233630815469072e-05,
      "loss": 0.4951,
      "mean_token_accuracy": 0.8450396955013275,
      "num_tokens": 767679459.0,
      "step": 24054
    },
    {
      "entropy": 0.5046978080645204,
      "epoch": 2.2131787626293122,
      "grad_norm": 0.16178983449935913,
      "learning_rate": 2.6230563989327443e-05,
      "loss": 0.4961,
      "mean_token_accuracy": 0.8429437056183815,
      "num_tokens": 767711290.0,
      "step": 24055
    },
    {
      "entropy": 0.4539424069225788,
      "epoch": 2.2132707690007534,
      "grad_norm": 0.15292729437351227,
      "learning_rate": 2.622749716318582e-05,
      "loss": 0.441,
      "mean_token_accuracy": 0.8562205769121647,
      "num_tokens": 767742874.0,
      "step": 24056
    },
    {
      "entropy": 0.68434764072299,
      "epoch": 2.2133627753721945,
      "grad_norm": 0.17810790240764618,
      "learning_rate": 2.6224430337044193e-05,
      "loss": 0.6675,
      "mean_token_accuracy": 0.7910949699580669,
      "num_tokens": 767775065.0,
      "step": 24057
    },
    {
      "entropy": 0.508543549105525,
      "epoch": 2.2134547817436356,
      "grad_norm": 0.15812020003795624,
      "learning_rate": 2.622136351090257e-05,
      "loss": 0.4943,
      "mean_token_accuracy": 0.8404782153666019,
      "num_tokens": 767807554.0,
      "step": 24058
    },
    {
      "entropy": 0.6023032935336232,
      "epoch": 2.2135467881150768,
      "grad_norm": 0.17847855389118195,
      "learning_rate": 2.6218296684760942e-05,
      "loss": 0.5976,
      "mean_token_accuracy": 0.8138238973915577,
      "num_tokens": 767839398.0,
      "step": 24059
    },
    {
      "entropy": 0.37412178702652454,
      "epoch": 2.2136387944865183,
      "grad_norm": 0.14635999500751495,
      "learning_rate": 2.6215229858619316e-05,
      "loss": 0.3653,
      "mean_token_accuracy": 0.8801523894071579,
      "num_tokens": 767871966.0,
      "step": 24060
    },
    {
      "entropy": 0.48853588476777077,
      "epoch": 2.2137308008579595,
      "grad_norm": 0.15578199923038483,
      "learning_rate": 2.6212163032477688e-05,
      "loss": 0.4747,
      "mean_token_accuracy": 0.847950167953968,
      "num_tokens": 767904143.0,
      "step": 24061
    },
    {
      "entropy": 0.5254276981577277,
      "epoch": 2.2138228072294006,
      "grad_norm": 0.1589752584695816,
      "learning_rate": 2.6209096206336066e-05,
      "loss": 0.499,
      "mean_token_accuracy": 0.8409887626767159,
      "num_tokens": 767935875.0,
      "step": 24062
    },
    {
      "entropy": 0.49342751130461693,
      "epoch": 2.2139148136008417,
      "grad_norm": 0.15691231191158295,
      "learning_rate": 2.6206029380194437e-05,
      "loss": 0.484,
      "mean_token_accuracy": 0.8459076508879662,
      "num_tokens": 767968163.0,
      "step": 24063
    },
    {
      "entropy": 0.5669362563639879,
      "epoch": 2.214006819972283,
      "grad_norm": 0.1680675595998764,
      "learning_rate": 2.620296255405281e-05,
      "loss": 0.5655,
      "mean_token_accuracy": 0.8223720639944077,
      "num_tokens": 768000348.0,
      "step": 24064
    },
    {
      "entropy": 0.5546128805726767,
      "epoch": 2.2140988263437245,
      "grad_norm": 0.1689557135105133,
      "learning_rate": 2.6199895727911183e-05,
      "loss": 0.5488,
      "mean_token_accuracy": 0.8270429335534573,
      "num_tokens": 768031804.0,
      "step": 24065
    },
    {
      "entropy": 0.4492250815965235,
      "epoch": 2.2141908327151656,
      "grad_norm": 0.15291757881641388,
      "learning_rate": 2.619682890176956e-05,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8605100363492966,
      "num_tokens": 768064135.0,
      "step": 24066
    },
    {
      "entropy": 0.5164509117603302,
      "epoch": 2.2142828390866067,
      "grad_norm": 0.156798854470253,
      "learning_rate": 2.6193762075627932e-05,
      "loss": 0.5046,
      "mean_token_accuracy": 0.840927466750145,
      "num_tokens": 768096688.0,
      "step": 24067
    },
    {
      "entropy": 0.5459316847845912,
      "epoch": 2.214374845458048,
      "grad_norm": 0.1636161506175995,
      "learning_rate": 2.6190695249486307e-05,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8317046947777271,
      "num_tokens": 768128485.0,
      "step": 24068
    },
    {
      "entropy": 0.48328449577093124,
      "epoch": 2.214466851829489,
      "grad_norm": 0.1528412252664566,
      "learning_rate": 2.6187628423344685e-05,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8505640998482704,
      "num_tokens": 768161231.0,
      "step": 24069
    },
    {
      "entropy": 0.49227343313395977,
      "epoch": 2.2145588582009306,
      "grad_norm": 0.16455641388893127,
      "learning_rate": 2.6184561597203056e-05,
      "loss": 0.4782,
      "mean_token_accuracy": 0.8471330627799034,
      "num_tokens": 768192233.0,
      "step": 24070
    },
    {
      "entropy": 0.5533964144997299,
      "epoch": 2.2146508645723717,
      "grad_norm": 0.16813957691192627,
      "learning_rate": 2.618149477106143e-05,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8285398185253143,
      "num_tokens": 768224155.0,
      "step": 24071
    },
    {
      "entropy": 0.5430130725726485,
      "epoch": 2.214742870943813,
      "grad_norm": 0.1604178249835968,
      "learning_rate": 2.6178427944919802e-05,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8287432752549648,
      "num_tokens": 768256923.0,
      "step": 24072
    },
    {
      "entropy": 0.39302029460668564,
      "epoch": 2.214834877315254,
      "grad_norm": 0.14211469888687134,
      "learning_rate": 2.617536111877818e-05,
      "loss": 0.3795,
      "mean_token_accuracy": 0.8776410855352879,
      "num_tokens": 768289525.0,
      "step": 24073
    },
    {
      "entropy": 0.4869801802560687,
      "epoch": 2.214926883686695,
      "grad_norm": 0.15735618770122528,
      "learning_rate": 2.617229429263655e-05,
      "loss": 0.4771,
      "mean_token_accuracy": 0.8490805923938751,
      "num_tokens": 768321857.0,
      "step": 24074
    },
    {
      "entropy": 0.472880607820116,
      "epoch": 2.2150188900581367,
      "grad_norm": 0.15510278940200806,
      "learning_rate": 2.616922746649493e-05,
      "loss": 0.4557,
      "mean_token_accuracy": 0.852123461663723,
      "num_tokens": 768354254.0,
      "step": 24075
    },
    {
      "entropy": 0.5909294448792934,
      "epoch": 2.215110896429578,
      "grad_norm": 0.1725175976753235,
      "learning_rate": 2.6166160640353297e-05,
      "loss": 0.5831,
      "mean_token_accuracy": 0.8170455507934093,
      "num_tokens": 768386109.0,
      "step": 24076
    },
    {
      "entropy": 0.49207106325775385,
      "epoch": 2.215202902801019,
      "grad_norm": 0.1585988700389862,
      "learning_rate": 2.6163093814211675e-05,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8449298776686192,
      "num_tokens": 768418316.0,
      "step": 24077
    },
    {
      "entropy": 0.538598945364356,
      "epoch": 2.21529490917246,
      "grad_norm": 0.1622481346130371,
      "learning_rate": 2.6160026988070047e-05,
      "loss": 0.5262,
      "mean_token_accuracy": 0.8318434692919254,
      "num_tokens": 768450123.0,
      "step": 24078
    },
    {
      "entropy": 0.5110766487196088,
      "epoch": 2.215386915543901,
      "grad_norm": 0.16051459312438965,
      "learning_rate": 2.6156960161928425e-05,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8433534689247608,
      "num_tokens": 768482078.0,
      "step": 24079
    },
    {
      "entropy": 0.4692382859066129,
      "epoch": 2.215478921915343,
      "grad_norm": 0.16236044466495514,
      "learning_rate": 2.6153893335786796e-05,
      "loss": 0.4642,
      "mean_token_accuracy": 0.8540223874151707,
      "num_tokens": 768514451.0,
      "step": 24080
    },
    {
      "entropy": 0.46861628256738186,
      "epoch": 2.215570928286784,
      "grad_norm": 0.1509227156639099,
      "learning_rate": 2.615082650964517e-05,
      "loss": 0.4472,
      "mean_token_accuracy": 0.856971126049757,
      "num_tokens": 768546674.0,
      "step": 24081
    },
    {
      "entropy": 0.570642601698637,
      "epoch": 2.215662934658225,
      "grad_norm": 0.17187529802322388,
      "learning_rate": 2.6147759683503542e-05,
      "loss": 0.5633,
      "mean_token_accuracy": 0.8209160752594471,
      "num_tokens": 768578562.0,
      "step": 24082
    },
    {
      "entropy": 0.4045756063424051,
      "epoch": 2.215754941029666,
      "grad_norm": 0.15134476125240326,
      "learning_rate": 2.614469285736192e-05,
      "loss": 0.3936,
      "mean_token_accuracy": 0.8784934803843498,
      "num_tokens": 768610860.0,
      "step": 24083
    },
    {
      "entropy": 0.5526494202204049,
      "epoch": 2.2158469474011073,
      "grad_norm": 0.16610614955425262,
      "learning_rate": 2.614162603122029e-05,
      "loss": 0.531,
      "mean_token_accuracy": 0.8304121755063534,
      "num_tokens": 768642349.0,
      "step": 24084
    },
    {
      "entropy": 0.5257638655602932,
      "epoch": 2.215938953772549,
      "grad_norm": 0.16973930597305298,
      "learning_rate": 2.6138559205078666e-05,
      "loss": 0.5208,
      "mean_token_accuracy": 0.8368773572146893,
      "num_tokens": 768673761.0,
      "step": 24085
    },
    {
      "entropy": 0.4299654527567327,
      "epoch": 2.21603096014399,
      "grad_norm": 0.14887455105781555,
      "learning_rate": 2.6135492378937037e-05,
      "loss": 0.424,
      "mean_token_accuracy": 0.8626262359321117,
      "num_tokens": 768706360.0,
      "step": 24086
    },
    {
      "entropy": 0.62607217207551,
      "epoch": 2.216122966515431,
      "grad_norm": 0.17458787560462952,
      "learning_rate": 2.6132425552795415e-05,
      "loss": 0.6127,
      "mean_token_accuracy": 0.8061243630945683,
      "num_tokens": 768738594.0,
      "step": 24087
    },
    {
      "entropy": 0.5236664013937116,
      "epoch": 2.2162149728868723,
      "grad_norm": 0.16066375374794006,
      "learning_rate": 2.6129358726653786e-05,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8405944146215916,
      "num_tokens": 768769722.0,
      "step": 24088
    },
    {
      "entropy": 0.474781284108758,
      "epoch": 2.2163069792583134,
      "grad_norm": 0.15469864010810852,
      "learning_rate": 2.612629190051216e-05,
      "loss": 0.4533,
      "mean_token_accuracy": 0.8567617274820805,
      "num_tokens": 768801838.0,
      "step": 24089
    },
    {
      "entropy": 0.43614475056529045,
      "epoch": 2.216398985629755,
      "grad_norm": 0.15131434798240662,
      "learning_rate": 2.6123225074370532e-05,
      "loss": 0.4326,
      "mean_token_accuracy": 0.8635197840631008,
      "num_tokens": 768834606.0,
      "step": 24090
    },
    {
      "entropy": 0.4126242958009243,
      "epoch": 2.216490992001196,
      "grad_norm": 0.14551293849945068,
      "learning_rate": 2.612015824822891e-05,
      "loss": 0.3908,
      "mean_token_accuracy": 0.8736629784107208,
      "num_tokens": 768866918.0,
      "step": 24091
    },
    {
      "entropy": 0.5124367689713836,
      "epoch": 2.2165829983726373,
      "grad_norm": 0.1632121205329895,
      "learning_rate": 2.611709142208728e-05,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8406545296311378,
      "num_tokens": 768897802.0,
      "step": 24092
    },
    {
      "entropy": 0.6120464410632849,
      "epoch": 2.2166750047440784,
      "grad_norm": 0.16825607419013977,
      "learning_rate": 2.6114024595945656e-05,
      "loss": 0.5973,
      "mean_token_accuracy": 0.8165067844092846,
      "num_tokens": 768930127.0,
      "step": 24093
    },
    {
      "entropy": 0.4345440175384283,
      "epoch": 2.2167670111155195,
      "grad_norm": 0.14337751269340515,
      "learning_rate": 2.6110957769804027e-05,
      "loss": 0.4276,
      "mean_token_accuracy": 0.8639183789491653,
      "num_tokens": 768962627.0,
      "step": 24094
    },
    {
      "entropy": 0.5614370983093977,
      "epoch": 2.216859017486961,
      "grad_norm": 0.16849683225154877,
      "learning_rate": 2.6107890943662405e-05,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8263925164937973,
      "num_tokens": 768994305.0,
      "step": 24095
    },
    {
      "entropy": 0.4391382117755711,
      "epoch": 2.2169510238584023,
      "grad_norm": 0.15144671499729156,
      "learning_rate": 2.6104824117520777e-05,
      "loss": 0.4288,
      "mean_token_accuracy": 0.8607584461569786,
      "num_tokens": 769026280.0,
      "step": 24096
    },
    {
      "entropy": 0.5238379323855042,
      "epoch": 2.2170430302298434,
      "grad_norm": 0.16336342692375183,
      "learning_rate": 2.6101757291379155e-05,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8407543823122978,
      "num_tokens": 769058355.0,
      "step": 24097
    },
    {
      "entropy": 0.6174629312008619,
      "epoch": 2.2171350366012845,
      "grad_norm": 0.16578218340873718,
      "learning_rate": 2.609869046523753e-05,
      "loss": 0.6036,
      "mean_token_accuracy": 0.8098328895866871,
      "num_tokens": 769090802.0,
      "step": 24098
    },
    {
      "entropy": 0.5270666875876486,
      "epoch": 2.2172270429727257,
      "grad_norm": 0.15411385893821716,
      "learning_rate": 2.60956236390959e-05,
      "loss": 0.515,
      "mean_token_accuracy": 0.8352870345115662,
      "num_tokens": 769123504.0,
      "step": 24099
    },
    {
      "entropy": 0.456963793374598,
      "epoch": 2.2173190493441672,
      "grad_norm": 0.1619601845741272,
      "learning_rate": 2.609255681295428e-05,
      "loss": 0.4542,
      "mean_token_accuracy": 0.855462945997715,
      "num_tokens": 769155306.0,
      "step": 24100
    },
    {
      "entropy": 0.5613561877980828,
      "epoch": 2.2174110557156084,
      "grad_norm": 0.1668044924736023,
      "learning_rate": 2.608948998681265e-05,
      "loss": 0.5544,
      "mean_token_accuracy": 0.8298528119921684,
      "num_tokens": 769186523.0,
      "step": 24101
    },
    {
      "entropy": 0.5321781723760068,
      "epoch": 2.2175030620870495,
      "grad_norm": 0.16078700125217438,
      "learning_rate": 2.6086423160671025e-05,
      "loss": 0.5061,
      "mean_token_accuracy": 0.8379521742463112,
      "num_tokens": 769218999.0,
      "step": 24102
    },
    {
      "entropy": 0.4454213213175535,
      "epoch": 2.2175950684584906,
      "grad_norm": 0.15027998387813568,
      "learning_rate": 2.6083356334529396e-05,
      "loss": 0.4218,
      "mean_token_accuracy": 0.8634618893265724,
      "num_tokens": 769251191.0,
      "step": 24103
    },
    {
      "entropy": 0.6184800211340189,
      "epoch": 2.2176870748299318,
      "grad_norm": 0.17652636766433716,
      "learning_rate": 2.6080289508387774e-05,
      "loss": 0.6084,
      "mean_token_accuracy": 0.8132140710949898,
      "num_tokens": 769282896.0,
      "step": 24104
    },
    {
      "entropy": 0.49645992647856474,
      "epoch": 2.2177790812013733,
      "grad_norm": 0.15132947266101837,
      "learning_rate": 2.6077222682246145e-05,
      "loss": 0.4794,
      "mean_token_accuracy": 0.8491695150732994,
      "num_tokens": 769315664.0,
      "step": 24105
    },
    {
      "entropy": 0.5156567431986332,
      "epoch": 2.2178710875728145,
      "grad_norm": 0.16544559597969055,
      "learning_rate": 2.607415585610452e-05,
      "loss": 0.503,
      "mean_token_accuracy": 0.8398311957716942,
      "num_tokens": 769347169.0,
      "step": 24106
    },
    {
      "entropy": 0.5340972850099206,
      "epoch": 2.2179630939442556,
      "grad_norm": 0.16743306815624237,
      "learning_rate": 2.607108902996289e-05,
      "loss": 0.5218,
      "mean_token_accuracy": 0.8360584154725075,
      "num_tokens": 769378699.0,
      "step": 24107
    },
    {
      "entropy": 0.5076277223415673,
      "epoch": 2.2180551003156967,
      "grad_norm": 0.16214731335639954,
      "learning_rate": 2.606802220382127e-05,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8430681750178337,
      "num_tokens": 769410765.0,
      "step": 24108
    },
    {
      "entropy": 0.5770067945122719,
      "epoch": 2.218147106687138,
      "grad_norm": 0.17169605195522308,
      "learning_rate": 2.606495537767964e-05,
      "loss": 0.5696,
      "mean_token_accuracy": 0.8211358971893787,
      "num_tokens": 769442786.0,
      "step": 24109
    },
    {
      "entropy": 0.5594105487689376,
      "epoch": 2.2182391130585795,
      "grad_norm": 0.1641508936882019,
      "learning_rate": 2.6061888551538015e-05,
      "loss": 0.5479,
      "mean_token_accuracy": 0.8271772116422653,
      "num_tokens": 769475135.0,
      "step": 24110
    },
    {
      "entropy": 0.48537838738411665,
      "epoch": 2.2183311194300206,
      "grad_norm": 0.15699556469917297,
      "learning_rate": 2.6058821725396386e-05,
      "loss": 0.4863,
      "mean_token_accuracy": 0.8465825840830803,
      "num_tokens": 769507358.0,
      "step": 24111
    },
    {
      "entropy": 0.5231360197067261,
      "epoch": 2.2184231258014617,
      "grad_norm": 0.16288383305072784,
      "learning_rate": 2.6055754899254764e-05,
      "loss": 0.507,
      "mean_token_accuracy": 0.8396123461425304,
      "num_tokens": 769538760.0,
      "step": 24112
    },
    {
      "entropy": 0.5669701192528009,
      "epoch": 2.218515132172903,
      "grad_norm": 0.16383010149002075,
      "learning_rate": 2.6052688073113136e-05,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8269264586269855,
      "num_tokens": 769570824.0,
      "step": 24113
    },
    {
      "entropy": 0.5126207331195474,
      "epoch": 2.218607138544344,
      "grad_norm": 0.15697240829467773,
      "learning_rate": 2.604962124697151e-05,
      "loss": 0.4949,
      "mean_token_accuracy": 0.8429424725472927,
      "num_tokens": 769603261.0,
      "step": 24114
    },
    {
      "entropy": 0.5271755270659924,
      "epoch": 2.2186991449157856,
      "grad_norm": 0.1598556935787201,
      "learning_rate": 2.604655442082988e-05,
      "loss": 0.5121,
      "mean_token_accuracy": 0.8411515317857265,
      "num_tokens": 769635486.0,
      "step": 24115
    },
    {
      "entropy": 0.5224324315786362,
      "epoch": 2.2187911512872267,
      "grad_norm": 0.15820761024951935,
      "learning_rate": 2.604348759468826e-05,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8372800424695015,
      "num_tokens": 769667866.0,
      "step": 24116
    },
    {
      "entropy": 0.5391001706011593,
      "epoch": 2.218883157658668,
      "grad_norm": 0.16259688138961792,
      "learning_rate": 2.604042076854663e-05,
      "loss": 0.5306,
      "mean_token_accuracy": 0.8336918838322163,
      "num_tokens": 769700314.0,
      "step": 24117
    },
    {
      "entropy": 0.5647189691662788,
      "epoch": 2.218975164030109,
      "grad_norm": 0.16588406264781952,
      "learning_rate": 2.603735394240501e-05,
      "loss": 0.5623,
      "mean_token_accuracy": 0.821798350661993,
      "num_tokens": 769732904.0,
      "step": 24118
    },
    {
      "entropy": 0.5979863693937659,
      "epoch": 2.21906717040155,
      "grad_norm": 0.17053662240505219,
      "learning_rate": 2.603428711626338e-05,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8198774382472038,
      "num_tokens": 769765037.0,
      "step": 24119
    },
    {
      "entropy": 0.5964881088584661,
      "epoch": 2.2191591767729917,
      "grad_norm": 0.17379604279994965,
      "learning_rate": 2.6031220290121755e-05,
      "loss": 0.5965,
      "mean_token_accuracy": 0.8134577982127666,
      "num_tokens": 769796942.0,
      "step": 24120
    },
    {
      "entropy": 0.5283995289355516,
      "epoch": 2.219251183144433,
      "grad_norm": 0.15946917235851288,
      "learning_rate": 2.6028153463980126e-05,
      "loss": 0.518,
      "mean_token_accuracy": 0.8352918885648251,
      "num_tokens": 769829231.0,
      "step": 24121
    },
    {
      "entropy": 0.4867911245673895,
      "epoch": 2.219343189515874,
      "grad_norm": 0.15785814821720123,
      "learning_rate": 2.6025086637838504e-05,
      "loss": 0.4753,
      "mean_token_accuracy": 0.848216887563467,
      "num_tokens": 769861997.0,
      "step": 24122
    },
    {
      "entropy": 0.5264019891619682,
      "epoch": 2.219435195887315,
      "grad_norm": 0.16321037709712982,
      "learning_rate": 2.6022019811696875e-05,
      "loss": 0.5141,
      "mean_token_accuracy": 0.8335185348987579,
      "num_tokens": 769892968.0,
      "step": 24123
    },
    {
      "entropy": 0.5249070012941957,
      "epoch": 2.219527202258756,
      "grad_norm": 0.1632147580385208,
      "learning_rate": 2.601895298555525e-05,
      "loss": 0.5186,
      "mean_token_accuracy": 0.8322250097990036,
      "num_tokens": 769925209.0,
      "step": 24124
    },
    {
      "entropy": 0.5725180003792048,
      "epoch": 2.219619208630198,
      "grad_norm": 0.17284472286701202,
      "learning_rate": 2.601588615941362e-05,
      "loss": 0.565,
      "mean_token_accuracy": 0.8242493011057377,
      "num_tokens": 769956599.0,
      "step": 24125
    },
    {
      "entropy": 0.4814272029325366,
      "epoch": 2.219711215001639,
      "grad_norm": 0.15213355422019958,
      "learning_rate": 2.6012819333272e-05,
      "loss": 0.4763,
      "mean_token_accuracy": 0.8485010787844658,
      "num_tokens": 769989329.0,
      "step": 24126
    },
    {
      "entropy": 0.5415396727621555,
      "epoch": 2.21980322137308,
      "grad_norm": 0.1628645360469818,
      "learning_rate": 2.600975250713037e-05,
      "loss": 0.54,
      "mean_token_accuracy": 0.8294633440673351,
      "num_tokens": 770020888.0,
      "step": 24127
    },
    {
      "entropy": 0.532535427249968,
      "epoch": 2.219895227744521,
      "grad_norm": 0.15922997891902924,
      "learning_rate": 2.6006685680988745e-05,
      "loss": 0.5081,
      "mean_token_accuracy": 0.8364602588117123,
      "num_tokens": 770052936.0,
      "step": 24128
    },
    {
      "entropy": 0.44204655243083835,
      "epoch": 2.2199872341159623,
      "grad_norm": 0.14731086790561676,
      "learning_rate": 2.6003618854847123e-05,
      "loss": 0.4252,
      "mean_token_accuracy": 0.8673265390098095,
      "num_tokens": 770084897.0,
      "step": 24129
    },
    {
      "entropy": 0.5431371685117483,
      "epoch": 2.220079240487404,
      "grad_norm": 0.16157712042331696,
      "learning_rate": 2.6000552028705494e-05,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8354815244674683,
      "num_tokens": 770116542.0,
      "step": 24130
    },
    {
      "entropy": 0.44536241330206394,
      "epoch": 2.220171246858845,
      "grad_norm": 0.1555860936641693,
      "learning_rate": 2.599748520256387e-05,
      "loss": 0.4382,
      "mean_token_accuracy": 0.8609196953475475,
      "num_tokens": 770148809.0,
      "step": 24131
    },
    {
      "entropy": 0.5366971967741847,
      "epoch": 2.220263253230286,
      "grad_norm": 0.16002878546714783,
      "learning_rate": 2.599441837642224e-05,
      "loss": 0.5204,
      "mean_token_accuracy": 0.838856179267168,
      "num_tokens": 770180555.0,
      "step": 24132
    },
    {
      "entropy": 0.5412471182644367,
      "epoch": 2.2203552596017273,
      "grad_norm": 0.159589022397995,
      "learning_rate": 2.5991351550280618e-05,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8303914591670036,
      "num_tokens": 770212191.0,
      "step": 24133
    },
    {
      "entropy": 0.5021136768627912,
      "epoch": 2.2204472659731684,
      "grad_norm": 0.159267857670784,
      "learning_rate": 2.598828472413899e-05,
      "loss": 0.4839,
      "mean_token_accuracy": 0.8472291715443134,
      "num_tokens": 770243764.0,
      "step": 24134
    },
    {
      "entropy": 0.5125138089060783,
      "epoch": 2.22053927234461,
      "grad_norm": 0.15979747474193573,
      "learning_rate": 2.5985217897997368e-05,
      "loss": 0.5035,
      "mean_token_accuracy": 0.8386246934533119,
      "num_tokens": 770275971.0,
      "step": 24135
    },
    {
      "entropy": 0.5389091959223151,
      "epoch": 2.220631278716051,
      "grad_norm": 0.1606779843568802,
      "learning_rate": 2.5982151071855735e-05,
      "loss": 0.5206,
      "mean_token_accuracy": 0.8349708281457424,
      "num_tokens": 770307557.0,
      "step": 24136
    },
    {
      "entropy": 0.5271124187856913,
      "epoch": 2.2207232850874923,
      "grad_norm": 0.16577214002609253,
      "learning_rate": 2.5979084245714113e-05,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8420873731374741,
      "num_tokens": 770339590.0,
      "step": 24137
    },
    {
      "entropy": 0.6008900450542569,
      "epoch": 2.2208152914589334,
      "grad_norm": 0.16818198561668396,
      "learning_rate": 2.5976017419572485e-05,
      "loss": 0.5869,
      "mean_token_accuracy": 0.8177864886820316,
      "num_tokens": 770371766.0,
      "step": 24138
    },
    {
      "entropy": 0.569618072360754,
      "epoch": 2.2209072978303745,
      "grad_norm": 0.16692565381526947,
      "learning_rate": 2.5972950593430863e-05,
      "loss": 0.5327,
      "mean_token_accuracy": 0.8318041600286961,
      "num_tokens": 770403323.0,
      "step": 24139
    },
    {
      "entropy": 0.5514718545600772,
      "epoch": 2.220999304201816,
      "grad_norm": 0.1654246300458908,
      "learning_rate": 2.5969883767289234e-05,
      "loss": 0.534,
      "mean_token_accuracy": 0.8292519636452198,
      "num_tokens": 770435214.0,
      "step": 24140
    },
    {
      "entropy": 0.32402163185179234,
      "epoch": 2.2210913105732573,
      "grad_norm": 0.1364925652742386,
      "learning_rate": 2.596681694114761e-05,
      "loss": 0.3079,
      "mean_token_accuracy": 0.8990892134606838,
      "num_tokens": 770467472.0,
      "step": 24141
    },
    {
      "entropy": 0.510351060424,
      "epoch": 2.2211833169446984,
      "grad_norm": 0.1595855951309204,
      "learning_rate": 2.596375011500598e-05,
      "loss": 0.4975,
      "mean_token_accuracy": 0.839462012052536,
      "num_tokens": 770498740.0,
      "step": 24142
    },
    {
      "entropy": 0.5306610092520714,
      "epoch": 2.2212753233161395,
      "grad_norm": 0.1641390025615692,
      "learning_rate": 2.5960683288864358e-05,
      "loss": 0.5093,
      "mean_token_accuracy": 0.83751380443573,
      "num_tokens": 770530698.0,
      "step": 24143
    },
    {
      "entropy": 0.5125164221972227,
      "epoch": 2.2213673296875807,
      "grad_norm": 0.1620316356420517,
      "learning_rate": 2.595761646272273e-05,
      "loss": 0.5082,
      "mean_token_accuracy": 0.8410204201936722,
      "num_tokens": 770563121.0,
      "step": 24144
    },
    {
      "entropy": 0.5160267492756248,
      "epoch": 2.2214593360590222,
      "grad_norm": 0.16471782326698303,
      "learning_rate": 2.5954549636581104e-05,
      "loss": 0.511,
      "mean_token_accuracy": 0.8398899771273136,
      "num_tokens": 770595770.0,
      "step": 24145
    },
    {
      "entropy": 0.4066947717219591,
      "epoch": 2.2215513424304634,
      "grad_norm": 0.15167386829853058,
      "learning_rate": 2.5951482810439475e-05,
      "loss": 0.3948,
      "mean_token_accuracy": 0.8729531653225422,
      "num_tokens": 770628501.0,
      "step": 24146
    },
    {
      "entropy": 0.6126541383564472,
      "epoch": 2.2216433488019045,
      "grad_norm": 0.17583829164505005,
      "learning_rate": 2.5948415984297853e-05,
      "loss": 0.6089,
      "mean_token_accuracy": 0.8098628968000412,
      "num_tokens": 770660013.0,
      "step": 24147
    },
    {
      "entropy": 0.38409616658464074,
      "epoch": 2.2217353551733456,
      "grad_norm": 0.14576859772205353,
      "learning_rate": 2.5945349158156224e-05,
      "loss": 0.3721,
      "mean_token_accuracy": 0.8842816948890686,
      "num_tokens": 770691985.0,
      "step": 24148
    },
    {
      "entropy": 0.6210328340530396,
      "epoch": 2.2218273615447868,
      "grad_norm": 0.16729964315891266,
      "learning_rate": 2.59422823320146e-05,
      "loss": 0.6243,
      "mean_token_accuracy": 0.8056728318333626,
      "num_tokens": 770724297.0,
      "step": 24149
    },
    {
      "entropy": 0.5068515921011567,
      "epoch": 2.2219193679162283,
      "grad_norm": 0.15809476375579834,
      "learning_rate": 2.593921550587297e-05,
      "loss": 0.5027,
      "mean_token_accuracy": 0.8389965109527111,
      "num_tokens": 770756625.0,
      "step": 24150
    },
    {
      "entropy": 0.47415505442768335,
      "epoch": 2.2220113742876695,
      "grad_norm": 0.1567959189414978,
      "learning_rate": 2.593614867973135e-05,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8508407473564148,
      "num_tokens": 770788184.0,
      "step": 24151
    },
    {
      "entropy": 0.5476826131343842,
      "epoch": 2.2221033806591106,
      "grad_norm": 0.16761381924152374,
      "learning_rate": 2.593308185358972e-05,
      "loss": 0.5377,
      "mean_token_accuracy": 0.8312164135277271,
      "num_tokens": 770820952.0,
      "step": 24152
    },
    {
      "entropy": 0.44882893189787865,
      "epoch": 2.2221953870305517,
      "grad_norm": 0.15038596093654633,
      "learning_rate": 2.5930015027448094e-05,
      "loss": 0.4308,
      "mean_token_accuracy": 0.8635087497532368,
      "num_tokens": 770853294.0,
      "step": 24153
    },
    {
      "entropy": 0.6774400509893894,
      "epoch": 2.222287393401993,
      "grad_norm": 0.17577053606510162,
      "learning_rate": 2.5926948201306466e-05,
      "loss": 0.6652,
      "mean_token_accuracy": 0.7928628474473953,
      "num_tokens": 770886050.0,
      "step": 24154
    },
    {
      "entropy": 0.4569920711219311,
      "epoch": 2.2223793997734345,
      "grad_norm": 0.15774744749069214,
      "learning_rate": 2.5923881375164844e-05,
      "loss": 0.4434,
      "mean_token_accuracy": 0.8573924526572227,
      "num_tokens": 770917970.0,
      "step": 24155
    },
    {
      "entropy": 0.449383984785527,
      "epoch": 2.2224714061448756,
      "grad_norm": 0.1473633050918579,
      "learning_rate": 2.5920814549023215e-05,
      "loss": 0.4273,
      "mean_token_accuracy": 0.8640789873898029,
      "num_tokens": 770950025.0,
      "step": 24156
    },
    {
      "entropy": 0.5928336856886744,
      "epoch": 2.2225634125163167,
      "grad_norm": 0.17341946065425873,
      "learning_rate": 2.5917747722881593e-05,
      "loss": 0.5851,
      "mean_token_accuracy": 0.8160858973860741,
      "num_tokens": 770981603.0,
      "step": 24157
    },
    {
      "entropy": 0.5164566747844219,
      "epoch": 2.222655418887758,
      "grad_norm": 0.16849513351917267,
      "learning_rate": 2.591468089673996e-05,
      "loss": 0.5097,
      "mean_token_accuracy": 0.8378558494150639,
      "num_tokens": 771013488.0,
      "step": 24158
    },
    {
      "entropy": 0.4392341426573694,
      "epoch": 2.222747425259199,
      "grad_norm": 0.1564790904521942,
      "learning_rate": 2.591161407059834e-05,
      "loss": 0.4353,
      "mean_token_accuracy": 0.864222526550293,
      "num_tokens": 771045418.0,
      "step": 24159
    },
    {
      "entropy": 0.5884446613490582,
      "epoch": 2.2228394316306406,
      "grad_norm": 0.16818274557590485,
      "learning_rate": 2.5908547244456717e-05,
      "loss": 0.5658,
      "mean_token_accuracy": 0.820187371224165,
      "num_tokens": 771077256.0,
      "step": 24160
    },
    {
      "entropy": 0.5775995254516602,
      "epoch": 2.2229314380020817,
      "grad_norm": 0.16729046404361725,
      "learning_rate": 2.5905480418315088e-05,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8220489323139191,
      "num_tokens": 771109067.0,
      "step": 24161
    },
    {
      "entropy": 0.4814216438680887,
      "epoch": 2.223023444373523,
      "grad_norm": 0.1523977369070053,
      "learning_rate": 2.5902413592173463e-05,
      "loss": 0.4675,
      "mean_token_accuracy": 0.8497374020516872,
      "num_tokens": 771141202.0,
      "step": 24162
    },
    {
      "entropy": 0.4626414147205651,
      "epoch": 2.223115450744964,
      "grad_norm": 0.14276228845119476,
      "learning_rate": 2.5899346766031834e-05,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8595510981976986,
      "num_tokens": 771173716.0,
      "step": 24163
    },
    {
      "entropy": 0.5447672251611948,
      "epoch": 2.223207457116405,
      "grad_norm": 0.1647256761789322,
      "learning_rate": 2.5896279939890212e-05,
      "loss": 0.5327,
      "mean_token_accuracy": 0.833054181188345,
      "num_tokens": 771205458.0,
      "step": 24164
    },
    {
      "entropy": 0.5558562851510942,
      "epoch": 2.2232994634878467,
      "grad_norm": 0.16975969076156616,
      "learning_rate": 2.5893213113748583e-05,
      "loss": 0.5479,
      "mean_token_accuracy": 0.8280544020235538,
      "num_tokens": 771237274.0,
      "step": 24165
    },
    {
      "entropy": 0.49836369417607784,
      "epoch": 2.223391469859288,
      "grad_norm": 0.15425027906894684,
      "learning_rate": 2.5890146287606958e-05,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8465288132429123,
      "num_tokens": 771269119.0,
      "step": 24166
    },
    {
      "entropy": 0.5306888502091169,
      "epoch": 2.223483476230729,
      "grad_norm": 0.15923303365707397,
      "learning_rate": 2.588707946146533e-05,
      "loss": 0.5231,
      "mean_token_accuracy": 0.8345753848552704,
      "num_tokens": 771300540.0,
      "step": 24167
    },
    {
      "entropy": 0.4670058134943247,
      "epoch": 2.22357548260217,
      "grad_norm": 0.14808879792690277,
      "learning_rate": 2.5884012635323707e-05,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8560908921062946,
      "num_tokens": 771333174.0,
      "step": 24168
    },
    {
      "entropy": 0.5321493453811854,
      "epoch": 2.223667488973611,
      "grad_norm": 0.15962429344654083,
      "learning_rate": 2.588094580918208e-05,
      "loss": 0.5126,
      "mean_token_accuracy": 0.8396625630557537,
      "num_tokens": 771365284.0,
      "step": 24169
    },
    {
      "entropy": 0.574725678190589,
      "epoch": 2.223759495345053,
      "grad_norm": 0.16324111819267273,
      "learning_rate": 2.5877878983040453e-05,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8282006271183491,
      "num_tokens": 771397380.0,
      "step": 24170
    },
    {
      "entropy": 0.5360817294567823,
      "epoch": 2.223851501716494,
      "grad_norm": 0.17075231671333313,
      "learning_rate": 2.5874812156898824e-05,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8308601267635822,
      "num_tokens": 771429107.0,
      "step": 24171
    },
    {
      "entropy": 0.5332609750330448,
      "epoch": 2.223943508087935,
      "grad_norm": 0.16448351740837097,
      "learning_rate": 2.5871745330757202e-05,
      "loss": 0.5291,
      "mean_token_accuracy": 0.8334911167621613,
      "num_tokens": 771460356.0,
      "step": 24172
    },
    {
      "entropy": 0.5332767898216844,
      "epoch": 2.224035514459376,
      "grad_norm": 0.16292357444763184,
      "learning_rate": 2.5868678504615574e-05,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8363931998610497,
      "num_tokens": 771492597.0,
      "step": 24173
    },
    {
      "entropy": 0.5575981047004461,
      "epoch": 2.2241275208308173,
      "grad_norm": 0.1589953452348709,
      "learning_rate": 2.586561167847395e-05,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8288236670196056,
      "num_tokens": 771525233.0,
      "step": 24174
    },
    {
      "entropy": 0.5755680873990059,
      "epoch": 2.224219527202259,
      "grad_norm": 0.16782258450984955,
      "learning_rate": 2.586254485233232e-05,
      "loss": 0.5746,
      "mean_token_accuracy": 0.815997838973999,
      "num_tokens": 771556781.0,
      "step": 24175
    },
    {
      "entropy": 0.5576616041362286,
      "epoch": 2.2243115335737,
      "grad_norm": 0.16060659289360046,
      "learning_rate": 2.5859478026190698e-05,
      "loss": 0.5443,
      "mean_token_accuracy": 0.8284975551068783,
      "num_tokens": 771589152.0,
      "step": 24176
    },
    {
      "entropy": 0.47238989919424057,
      "epoch": 2.224403539945141,
      "grad_norm": 0.16017372906208038,
      "learning_rate": 2.585641120004907e-05,
      "loss": 0.4621,
      "mean_token_accuracy": 0.8538109250366688,
      "num_tokens": 771620987.0,
      "step": 24177
    },
    {
      "entropy": 0.5230642086826265,
      "epoch": 2.2244955463165823,
      "grad_norm": 0.16441266238689423,
      "learning_rate": 2.5853344373907447e-05,
      "loss": 0.4936,
      "mean_token_accuracy": 0.8403279222548008,
      "num_tokens": 771652180.0,
      "step": 24178
    },
    {
      "entropy": 0.5188231791835278,
      "epoch": 2.2245875526880234,
      "grad_norm": 0.15875445306301117,
      "learning_rate": 2.5850277547765818e-05,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8429018259048462,
      "num_tokens": 771683944.0,
      "step": 24179
    },
    {
      "entropy": 0.6160870548337698,
      "epoch": 2.224679559059465,
      "grad_norm": 0.17091815173625946,
      "learning_rate": 2.5847210721624193e-05,
      "loss": 0.6047,
      "mean_token_accuracy": 0.8082520514726639,
      "num_tokens": 771716293.0,
      "step": 24180
    },
    {
      "entropy": 0.5460703447461128,
      "epoch": 2.224771565430906,
      "grad_norm": 0.16684061288833618,
      "learning_rate": 2.5844143895482564e-05,
      "loss": 0.5293,
      "mean_token_accuracy": 0.8328435011208057,
      "num_tokens": 771748403.0,
      "step": 24181
    },
    {
      "entropy": 0.537287502666004,
      "epoch": 2.2248635718023473,
      "grad_norm": 0.16140489280223846,
      "learning_rate": 2.5841077069340942e-05,
      "loss": 0.521,
      "mean_token_accuracy": 0.8359015174210072,
      "num_tokens": 771780516.0,
      "step": 24182
    },
    {
      "entropy": 0.47824247647076845,
      "epoch": 2.2249555781737884,
      "grad_norm": 0.15476438403129578,
      "learning_rate": 2.5838010243199313e-05,
      "loss": 0.4549,
      "mean_token_accuracy": 0.8532870672643185,
      "num_tokens": 771812650.0,
      "step": 24183
    },
    {
      "entropy": 0.5063483193516731,
      "epoch": 2.2250475845452296,
      "grad_norm": 0.1568438857793808,
      "learning_rate": 2.5834943417057688e-05,
      "loss": 0.4987,
      "mean_token_accuracy": 0.8476503193378448,
      "num_tokens": 771844688.0,
      "step": 24184
    },
    {
      "entropy": 0.4278112929314375,
      "epoch": 2.225139590916671,
      "grad_norm": 0.1478024572134018,
      "learning_rate": 2.583187659091606e-05,
      "loss": 0.4117,
      "mean_token_accuracy": 0.8699154667556286,
      "num_tokens": 771876382.0,
      "step": 24185
    },
    {
      "entropy": 0.523198489099741,
      "epoch": 2.2252315972881123,
      "grad_norm": 0.16109131276607513,
      "learning_rate": 2.5828809764774437e-05,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8356128223240376,
      "num_tokens": 771908725.0,
      "step": 24186
    },
    {
      "entropy": 0.6150384172797203,
      "epoch": 2.2253236036595534,
      "grad_norm": 0.17417985200881958,
      "learning_rate": 2.582574293863281e-05,
      "loss": 0.6096,
      "mean_token_accuracy": 0.807486642152071,
      "num_tokens": 771941261.0,
      "step": 24187
    },
    {
      "entropy": 0.5670816535130143,
      "epoch": 2.2254156100309945,
      "grad_norm": 0.1661311835050583,
      "learning_rate": 2.5822676112491183e-05,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8303897231817245,
      "num_tokens": 771973065.0,
      "step": 24188
    },
    {
      "entropy": 0.5940615171566606,
      "epoch": 2.2255076164024357,
      "grad_norm": 0.17367397248744965,
      "learning_rate": 2.5819609286349555e-05,
      "loss": 0.5762,
      "mean_token_accuracy": 0.8178001493215561,
      "num_tokens": 772004090.0,
      "step": 24189
    },
    {
      "entropy": 0.4830694762058556,
      "epoch": 2.2255996227738772,
      "grad_norm": 0.15897896885871887,
      "learning_rate": 2.5816542460207933e-05,
      "loss": 0.4695,
      "mean_token_accuracy": 0.8504563868045807,
      "num_tokens": 772035645.0,
      "step": 24190
    },
    {
      "entropy": 0.46350731048732996,
      "epoch": 2.2256916291453184,
      "grad_norm": 0.15585826337337494,
      "learning_rate": 2.5813475634066307e-05,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8525413908064365,
      "num_tokens": 772067535.0,
      "step": 24191
    },
    {
      "entropy": 0.5706895980983973,
      "epoch": 2.2257836355167595,
      "grad_norm": 0.16769146919250488,
      "learning_rate": 2.581040880792468e-05,
      "loss": 0.5615,
      "mean_token_accuracy": 0.820682629942894,
      "num_tokens": 772100068.0,
      "step": 24192
    },
    {
      "entropy": 0.5525089921429753,
      "epoch": 2.2258756418882006,
      "grad_norm": 0.16313698887825012,
      "learning_rate": 2.5807341981783056e-05,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8290542811155319,
      "num_tokens": 772132548.0,
      "step": 24193
    },
    {
      "entropy": 0.5156596023589373,
      "epoch": 2.2259676482596418,
      "grad_norm": 0.15946556627750397,
      "learning_rate": 2.5804275155641428e-05,
      "loss": 0.4999,
      "mean_token_accuracy": 0.8418563529849052,
      "num_tokens": 772163755.0,
      "step": 24194
    },
    {
      "entropy": 0.5200337516143918,
      "epoch": 2.2260596546310834,
      "grad_norm": 0.1578545719385147,
      "learning_rate": 2.5801208329499806e-05,
      "loss": 0.4967,
      "mean_token_accuracy": 0.8372015208005905,
      "num_tokens": 772195866.0,
      "step": 24195
    },
    {
      "entropy": 0.5080026369541883,
      "epoch": 2.2261516610025245,
      "grad_norm": 0.15749460458755493,
      "learning_rate": 2.5798141503358174e-05,
      "loss": 0.4949,
      "mean_token_accuracy": 0.8402348533272743,
      "num_tokens": 772227352.0,
      "step": 24196
    },
    {
      "entropy": 0.4938706997781992,
      "epoch": 2.2262436673739656,
      "grad_norm": 0.15688635408878326,
      "learning_rate": 2.579507467721655e-05,
      "loss": 0.4836,
      "mean_token_accuracy": 0.8456770330667496,
      "num_tokens": 772259708.0,
      "step": 24197
    },
    {
      "entropy": 0.4143405146896839,
      "epoch": 2.2263356737454068,
      "grad_norm": 0.15118391811847687,
      "learning_rate": 2.5792007851074923e-05,
      "loss": 0.3988,
      "mean_token_accuracy": 0.8716897815465927,
      "num_tokens": 772291694.0,
      "step": 24198
    },
    {
      "entropy": 0.553772839717567,
      "epoch": 2.226427680116848,
      "grad_norm": 0.16173598170280457,
      "learning_rate": 2.57889410249333e-05,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8283601924777031,
      "num_tokens": 772324425.0,
      "step": 24199
    },
    {
      "entropy": 0.6061022141948342,
      "epoch": 2.2265196864882895,
      "grad_norm": 0.1726573258638382,
      "learning_rate": 2.5785874198791672e-05,
      "loss": 0.5922,
      "mean_token_accuracy": 0.8177211470901966,
      "num_tokens": 772356516.0,
      "step": 24200
    },
    {
      "entropy": 0.4310205499641597,
      "epoch": 2.2266116928597306,
      "grad_norm": 0.14938883483409882,
      "learning_rate": 2.5782807372650047e-05,
      "loss": 0.4113,
      "mean_token_accuracy": 0.8655320778489113,
      "num_tokens": 772388422.0,
      "step": 24201
    },
    {
      "entropy": 0.49291414953768253,
      "epoch": 2.2267036992311717,
      "grad_norm": 0.15434715151786804,
      "learning_rate": 2.5779740546508418e-05,
      "loss": 0.4824,
      "mean_token_accuracy": 0.8449989855289459,
      "num_tokens": 772420416.0,
      "step": 24202
    },
    {
      "entropy": 0.4596136147156358,
      "epoch": 2.226795705602613,
      "grad_norm": 0.15812233090400696,
      "learning_rate": 2.5776673720366796e-05,
      "loss": 0.4602,
      "mean_token_accuracy": 0.8519364185631275,
      "num_tokens": 772452419.0,
      "step": 24203
    },
    {
      "entropy": 0.5959602296352386,
      "epoch": 2.226887711974054,
      "grad_norm": 0.17500127851963043,
      "learning_rate": 2.5773606894225167e-05,
      "loss": 0.5848,
      "mean_token_accuracy": 0.8153360225260258,
      "num_tokens": 772484418.0,
      "step": 24204
    },
    {
      "entropy": 0.5585185028612614,
      "epoch": 2.2269797183454956,
      "grad_norm": 0.15680742263793945,
      "learning_rate": 2.5770540068083542e-05,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8327614329755306,
      "num_tokens": 772516521.0,
      "step": 24205
    },
    {
      "entropy": 0.5843577766790986,
      "epoch": 2.2270717247169367,
      "grad_norm": 0.16620823740959167,
      "learning_rate": 2.5767473241941913e-05,
      "loss": 0.5652,
      "mean_token_accuracy": 0.8184045515954494,
      "num_tokens": 772548359.0,
      "step": 24206
    },
    {
      "entropy": 0.5589355072006583,
      "epoch": 2.227163731088378,
      "grad_norm": 0.1671816110610962,
      "learning_rate": 2.576440641580029e-05,
      "loss": 0.5541,
      "mean_token_accuracy": 0.8254936523735523,
      "num_tokens": 772579962.0,
      "step": 24207
    },
    {
      "entropy": 0.5127271264791489,
      "epoch": 2.227255737459819,
      "grad_norm": 0.16154004633426666,
      "learning_rate": 2.5761339589658663e-05,
      "loss": 0.505,
      "mean_token_accuracy": 0.8385627493262291,
      "num_tokens": 772611377.0,
      "step": 24208
    },
    {
      "entropy": 0.5489284992218018,
      "epoch": 2.22734774383126,
      "grad_norm": 0.17002421617507935,
      "learning_rate": 2.5758272763517037e-05,
      "loss": 0.5324,
      "mean_token_accuracy": 0.8332307003438473,
      "num_tokens": 772643230.0,
      "step": 24209
    },
    {
      "entropy": 0.564229927957058,
      "epoch": 2.2274397502027017,
      "grad_norm": 0.16743972897529602,
      "learning_rate": 2.575520593737541e-05,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8281581848859787,
      "num_tokens": 772675419.0,
      "step": 24210
    },
    {
      "entropy": 0.5161619298160076,
      "epoch": 2.227531756574143,
      "grad_norm": 0.16041989624500275,
      "learning_rate": 2.5752139111233787e-05,
      "loss": 0.4988,
      "mean_token_accuracy": 0.8428904488682747,
      "num_tokens": 772707458.0,
      "step": 24211
    },
    {
      "entropy": 0.5086976392194629,
      "epoch": 2.227623762945584,
      "grad_norm": 0.15733115375041962,
      "learning_rate": 2.5749072285092158e-05,
      "loss": 0.4803,
      "mean_token_accuracy": 0.8486697934567928,
      "num_tokens": 772739430.0,
      "step": 24212
    },
    {
      "entropy": 0.48994227312505245,
      "epoch": 2.227715769317025,
      "grad_norm": 0.15414069592952728,
      "learning_rate": 2.5746005458950533e-05,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8514166921377182,
      "num_tokens": 772771896.0,
      "step": 24213
    },
    {
      "entropy": 0.5538936164230108,
      "epoch": 2.227807775688466,
      "grad_norm": 0.16866429150104523,
      "learning_rate": 2.5742938632808904e-05,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8272120133042336,
      "num_tokens": 772803607.0,
      "step": 24214
    },
    {
      "entropy": 0.5754121374338865,
      "epoch": 2.227899782059908,
      "grad_norm": 0.16851167380809784,
      "learning_rate": 2.5739871806667282e-05,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8201387748122215,
      "num_tokens": 772834918.0,
      "step": 24215
    },
    {
      "entropy": 0.5977393332868814,
      "epoch": 2.227991788431349,
      "grad_norm": 0.16975998878479004,
      "learning_rate": 2.5736804980525653e-05,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8151269294321537,
      "num_tokens": 772866826.0,
      "step": 24216
    },
    {
      "entropy": 0.5858702547848225,
      "epoch": 2.22808379480279,
      "grad_norm": 0.16681984066963196,
      "learning_rate": 2.573373815438403e-05,
      "loss": 0.5785,
      "mean_token_accuracy": 0.817340075969696,
      "num_tokens": 772899349.0,
      "step": 24217
    },
    {
      "entropy": 0.5012794428039342,
      "epoch": 2.228175801174231,
      "grad_norm": 0.15877966582775116,
      "learning_rate": 2.57306713282424e-05,
      "loss": 0.4771,
      "mean_token_accuracy": 0.8487435318529606,
      "num_tokens": 772931146.0,
      "step": 24218
    },
    {
      "entropy": 0.5101269790902734,
      "epoch": 2.2282678075456723,
      "grad_norm": 0.16086624562740326,
      "learning_rate": 2.5727604502100777e-05,
      "loss": 0.5042,
      "mean_token_accuracy": 0.840113777667284,
      "num_tokens": 772963065.0,
      "step": 24219
    },
    {
      "entropy": 0.5012874277308583,
      "epoch": 2.228359813917114,
      "grad_norm": 0.15265828371047974,
      "learning_rate": 2.5724537675959155e-05,
      "loss": 0.4961,
      "mean_token_accuracy": 0.8404385633766651,
      "num_tokens": 772995654.0,
      "step": 24220
    },
    {
      "entropy": 0.5050824664067477,
      "epoch": 2.228451820288555,
      "grad_norm": 0.16120262444019318,
      "learning_rate": 2.5721470849817526e-05,
      "loss": 0.4904,
      "mean_token_accuracy": 0.8437262959778309,
      "num_tokens": 773027145.0,
      "step": 24221
    },
    {
      "entropy": 0.5771566163748503,
      "epoch": 2.228543826659996,
      "grad_norm": 0.16911399364471436,
      "learning_rate": 2.57184040236759e-05,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8201207742094994,
      "num_tokens": 773059392.0,
      "step": 24222
    },
    {
      "entropy": 0.5430606752634048,
      "epoch": 2.2286358330314373,
      "grad_norm": 0.16423708200454712,
      "learning_rate": 2.5715337197534272e-05,
      "loss": 0.529,
      "mean_token_accuracy": 0.8337993435561657,
      "num_tokens": 773091392.0,
      "step": 24223
    },
    {
      "entropy": 0.4864490544423461,
      "epoch": 2.228727839402879,
      "grad_norm": 0.15861792862415314,
      "learning_rate": 2.571227037139265e-05,
      "loss": 0.4753,
      "mean_token_accuracy": 0.8489278629422188,
      "num_tokens": 773123459.0,
      "step": 24224
    },
    {
      "entropy": 0.6029484262689948,
      "epoch": 2.22881984577432,
      "grad_norm": 0.17532320320606232,
      "learning_rate": 2.570920354525102e-05,
      "loss": 0.6017,
      "mean_token_accuracy": 0.8108507916331291,
      "num_tokens": 773155223.0,
      "step": 24225
    },
    {
      "entropy": 0.5491222362034023,
      "epoch": 2.228911852145761,
      "grad_norm": 0.16129957139492035,
      "learning_rate": 2.5706136719109396e-05,
      "loss": 0.5349,
      "mean_token_accuracy": 0.8339671231806278,
      "num_tokens": 773187469.0,
      "step": 24226
    },
    {
      "entropy": 0.585704330354929,
      "epoch": 2.2290038585172023,
      "grad_norm": 0.16762952506542206,
      "learning_rate": 2.5703069892967767e-05,
      "loss": 0.5839,
      "mean_token_accuracy": 0.8176422454416752,
      "num_tokens": 773219246.0,
      "step": 24227
    },
    {
      "entropy": 0.5796340871602297,
      "epoch": 2.2290958648886434,
      "grad_norm": 0.1688656210899353,
      "learning_rate": 2.5700003066826145e-05,
      "loss": 0.5795,
      "mean_token_accuracy": 0.8189657106995583,
      "num_tokens": 773250555.0,
      "step": 24228
    },
    {
      "entropy": 0.5116355104837567,
      "epoch": 2.229187871260085,
      "grad_norm": 0.15714778006076813,
      "learning_rate": 2.5696936240684517e-05,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8405769281089306,
      "num_tokens": 773282004.0,
      "step": 24229
    },
    {
      "entropy": 0.5378795489668846,
      "epoch": 2.229279877631526,
      "grad_norm": 0.16276484727859497,
      "learning_rate": 2.569386941454289e-05,
      "loss": 0.5243,
      "mean_token_accuracy": 0.8358909711241722,
      "num_tokens": 773314412.0,
      "step": 24230
    },
    {
      "entropy": 0.4740041419863701,
      "epoch": 2.2293718840029673,
      "grad_norm": 0.15510046482086182,
      "learning_rate": 2.5690802588401263e-05,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8510599806904793,
      "num_tokens": 773346967.0,
      "step": 24231
    },
    {
      "entropy": 0.4325535265961662,
      "epoch": 2.2294638903744084,
      "grad_norm": 0.15080861747264862,
      "learning_rate": 2.568773576225964e-05,
      "loss": 0.425,
      "mean_token_accuracy": 0.8665114678442478,
      "num_tokens": 773379396.0,
      "step": 24232
    },
    {
      "entropy": 0.5336015238426626,
      "epoch": 2.2295558967458495,
      "grad_norm": 0.1621493697166443,
      "learning_rate": 2.5684668936118012e-05,
      "loss": 0.5228,
      "mean_token_accuracy": 0.8368220180273056,
      "num_tokens": 773411274.0,
      "step": 24233
    },
    {
      "entropy": 0.4344001426361501,
      "epoch": 2.229647903117291,
      "grad_norm": 0.1500534564256668,
      "learning_rate": 2.5681602109976387e-05,
      "loss": 0.4085,
      "mean_token_accuracy": 0.8652683570981026,
      "num_tokens": 773443163.0,
      "step": 24234
    },
    {
      "entropy": 0.493643709924072,
      "epoch": 2.2297399094887322,
      "grad_norm": 0.1514306515455246,
      "learning_rate": 2.5678535283834758e-05,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8520896062254906,
      "num_tokens": 773475919.0,
      "step": 24235
    },
    {
      "entropy": 0.5689303362742066,
      "epoch": 2.2298319158601734,
      "grad_norm": 0.17050206661224365,
      "learning_rate": 2.5675468457693136e-05,
      "loss": 0.554,
      "mean_token_accuracy": 0.8232903219759464,
      "num_tokens": 773507598.0,
      "step": 24236
    },
    {
      "entropy": 0.465215896256268,
      "epoch": 2.2299239222316145,
      "grad_norm": 0.15147949755191803,
      "learning_rate": 2.5672401631551507e-05,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8571645617485046,
      "num_tokens": 773540365.0,
      "step": 24237
    },
    {
      "entropy": 0.5292282016016543,
      "epoch": 2.2300159286030556,
      "grad_norm": 0.15871350467205048,
      "learning_rate": 2.5669334805409885e-05,
      "loss": 0.5081,
      "mean_token_accuracy": 0.8370634689927101,
      "num_tokens": 773572628.0,
      "step": 24238
    },
    {
      "entropy": 0.5127226365730166,
      "epoch": 2.230107934974497,
      "grad_norm": 0.15640512108802795,
      "learning_rate": 2.5666267979268256e-05,
      "loss": 0.5052,
      "mean_token_accuracy": 0.8419892676174641,
      "num_tokens": 773604617.0,
      "step": 24239
    },
    {
      "entropy": 0.5863154325634241,
      "epoch": 2.2301999413459384,
      "grad_norm": 0.1670151948928833,
      "learning_rate": 2.566320115312663e-05,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8201940767467022,
      "num_tokens": 773636143.0,
      "step": 24240
    },
    {
      "entropy": 0.5516415745951235,
      "epoch": 2.2302919477173795,
      "grad_norm": 0.1717502325773239,
      "learning_rate": 2.5660134326985002e-05,
      "loss": 0.5446,
      "mean_token_accuracy": 0.8303071334958076,
      "num_tokens": 773666826.0,
      "step": 24241
    },
    {
      "entropy": 0.5943624414503574,
      "epoch": 2.2303839540888206,
      "grad_norm": 0.17130158841609955,
      "learning_rate": 2.565706750084338e-05,
      "loss": 0.5896,
      "mean_token_accuracy": 0.8159093298017979,
      "num_tokens": 773698723.0,
      "step": 24242
    },
    {
      "entropy": 0.528668228071183,
      "epoch": 2.2304759604602618,
      "grad_norm": 0.1600949466228485,
      "learning_rate": 2.565400067470175e-05,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8370017819106579,
      "num_tokens": 773729921.0,
      "step": 24243
    },
    {
      "entropy": 0.5921947965398431,
      "epoch": 2.2305679668317033,
      "grad_norm": 0.16292038559913635,
      "learning_rate": 2.5650933848560126e-05,
      "loss": 0.5792,
      "mean_token_accuracy": 0.8173513859510422,
      "num_tokens": 773762118.0,
      "step": 24244
    },
    {
      "entropy": 0.5449336771853268,
      "epoch": 2.2306599732031445,
      "grad_norm": 0.1611427366733551,
      "learning_rate": 2.5647867022418498e-05,
      "loss": 0.5265,
      "mean_token_accuracy": 0.8338914811611176,
      "num_tokens": 773793992.0,
      "step": 24245
    },
    {
      "entropy": 0.5195654160343111,
      "epoch": 2.2307519795745856,
      "grad_norm": 0.1571577936410904,
      "learning_rate": 2.5644800196276876e-05,
      "loss": 0.5008,
      "mean_token_accuracy": 0.8417230173945427,
      "num_tokens": 773826532.0,
      "step": 24246
    },
    {
      "entropy": 0.39614779665134847,
      "epoch": 2.2308439859460267,
      "grad_norm": 0.14596448838710785,
      "learning_rate": 2.5641733370135247e-05,
      "loss": 0.3901,
      "mean_token_accuracy": 0.8746721558272839,
      "num_tokens": 773858251.0,
      "step": 24247
    },
    {
      "entropy": 0.43355192616581917,
      "epoch": 2.230935992317468,
      "grad_norm": 0.15388889610767365,
      "learning_rate": 2.563866654399362e-05,
      "loss": 0.4209,
      "mean_token_accuracy": 0.8657216727733612,
      "num_tokens": 773889227.0,
      "step": 24248
    },
    {
      "entropy": 0.5757452938705683,
      "epoch": 2.2310279986889094,
      "grad_norm": 0.1662476360797882,
      "learning_rate": 2.5635599717851993e-05,
      "loss": 0.5617,
      "mean_token_accuracy": 0.8224816918373108,
      "num_tokens": 773921023.0,
      "step": 24249
    },
    {
      "entropy": 0.5306198676116765,
      "epoch": 2.2311200050603506,
      "grad_norm": 0.16432391107082367,
      "learning_rate": 2.563253289171037e-05,
      "loss": 0.5248,
      "mean_token_accuracy": 0.8362682573497295,
      "num_tokens": 773952802.0,
      "step": 24250
    },
    {
      "entropy": 0.5843898681923747,
      "epoch": 2.2312120114317917,
      "grad_norm": 0.16694867610931396,
      "learning_rate": 2.5629466065568745e-05,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8194572068750858,
      "num_tokens": 773984613.0,
      "step": 24251
    },
    {
      "entropy": 0.4977124221622944,
      "epoch": 2.231304017803233,
      "grad_norm": 0.15905877947807312,
      "learning_rate": 2.5626399239427117e-05,
      "loss": 0.4859,
      "mean_token_accuracy": 0.8435473926365376,
      "num_tokens": 774017167.0,
      "step": 24252
    },
    {
      "entropy": 0.513339180033654,
      "epoch": 2.231396024174674,
      "grad_norm": 0.16109547019004822,
      "learning_rate": 2.5623332413285495e-05,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8448058068752289,
      "num_tokens": 774048710.0,
      "step": 24253
    },
    {
      "entropy": 0.560325033031404,
      "epoch": 2.2314880305461156,
      "grad_norm": 0.16057373583316803,
      "learning_rate": 2.5620265587143866e-05,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8281800672411919,
      "num_tokens": 774081128.0,
      "step": 24254
    },
    {
      "entropy": 0.5906694158911705,
      "epoch": 2.2315800369175567,
      "grad_norm": 0.16647708415985107,
      "learning_rate": 2.5617198761002244e-05,
      "loss": 0.5815,
      "mean_token_accuracy": 0.8204065449535847,
      "num_tokens": 774112784.0,
      "step": 24255
    },
    {
      "entropy": 0.6054695583879948,
      "epoch": 2.231672043288998,
      "grad_norm": 0.1675838828086853,
      "learning_rate": 2.5614131934860612e-05,
      "loss": 0.5936,
      "mean_token_accuracy": 0.8121442943811417,
      "num_tokens": 774145091.0,
      "step": 24256
    },
    {
      "entropy": 0.49472575075924397,
      "epoch": 2.231764049660439,
      "grad_norm": 0.15979841351509094,
      "learning_rate": 2.561106510871899e-05,
      "loss": 0.4836,
      "mean_token_accuracy": 0.8464874811470509,
      "num_tokens": 774177348.0,
      "step": 24257
    },
    {
      "entropy": 0.5805886466987431,
      "epoch": 2.23185605603188,
      "grad_norm": 0.16818678379058838,
      "learning_rate": 2.560799828257736e-05,
      "loss": 0.5716,
      "mean_token_accuracy": 0.8193923197686672,
      "num_tokens": 774209134.0,
      "step": 24258
    },
    {
      "entropy": 0.57272869348526,
      "epoch": 2.2319480624033217,
      "grad_norm": 0.16775625944137573,
      "learning_rate": 2.560493145643574e-05,
      "loss": 0.563,
      "mean_token_accuracy": 0.8246313594281673,
      "num_tokens": 774240662.0,
      "step": 24259
    },
    {
      "entropy": 0.5190178984776139,
      "epoch": 2.232040068774763,
      "grad_norm": 0.1613803505897522,
      "learning_rate": 2.560186463029411e-05,
      "loss": 0.5094,
      "mean_token_accuracy": 0.8338312283158302,
      "num_tokens": 774272755.0,
      "step": 24260
    },
    {
      "entropy": 0.5484495032578707,
      "epoch": 2.232132075146204,
      "grad_norm": 0.15859757363796234,
      "learning_rate": 2.5598797804152485e-05,
      "loss": 0.536,
      "mean_token_accuracy": 0.8311124481260777,
      "num_tokens": 774304785.0,
      "step": 24261
    },
    {
      "entropy": 0.5121656004339457,
      "epoch": 2.232224081517645,
      "grad_norm": 0.15812642872333527,
      "learning_rate": 2.5595730978010856e-05,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8381439261138439,
      "num_tokens": 774337207.0,
      "step": 24262
    },
    {
      "entropy": 0.6150475703179836,
      "epoch": 2.232316087889086,
      "grad_norm": 0.1715523600578308,
      "learning_rate": 2.5592664151869234e-05,
      "loss": 0.6123,
      "mean_token_accuracy": 0.8124958612024784,
      "num_tokens": 774369213.0,
      "step": 24263
    },
    {
      "entropy": 0.582396300509572,
      "epoch": 2.2324080942605278,
      "grad_norm": 0.17195047438144684,
      "learning_rate": 2.5589597325727606e-05,
      "loss": 0.5681,
      "mean_token_accuracy": 0.8239339292049408,
      "num_tokens": 774400969.0,
      "step": 24264
    },
    {
      "entropy": 0.5621020724065602,
      "epoch": 2.232500100631969,
      "grad_norm": 0.17193754017353058,
      "learning_rate": 2.558653049958598e-05,
      "loss": 0.5506,
      "mean_token_accuracy": 0.8252263516187668,
      "num_tokens": 774433312.0,
      "step": 24265
    },
    {
      "entropy": 0.4957766458392143,
      "epoch": 2.23259210700341,
      "grad_norm": 0.15749599039554596,
      "learning_rate": 2.558346367344435e-05,
      "loss": 0.4824,
      "mean_token_accuracy": 0.8445932827889919,
      "num_tokens": 774465675.0,
      "step": 24266
    },
    {
      "entropy": 0.46912147011607885,
      "epoch": 2.232684113374851,
      "grad_norm": 0.1507721245288849,
      "learning_rate": 2.558039684730273e-05,
      "loss": 0.4584,
      "mean_token_accuracy": 0.8548216074705124,
      "num_tokens": 774498229.0,
      "step": 24267
    },
    {
      "entropy": 0.4421043526381254,
      "epoch": 2.2327761197462923,
      "grad_norm": 0.14998717606067657,
      "learning_rate": 2.55773300211611e-05,
      "loss": 0.436,
      "mean_token_accuracy": 0.8635517805814743,
      "num_tokens": 774530361.0,
      "step": 24268
    },
    {
      "entropy": 0.5330431517213583,
      "epoch": 2.232868126117734,
      "grad_norm": 0.16549460589885712,
      "learning_rate": 2.5574263195019475e-05,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8307439424097538,
      "num_tokens": 774562941.0,
      "step": 24269
    },
    {
      "entropy": 0.47594837099313736,
      "epoch": 2.232960132489175,
      "grad_norm": 0.15344706177711487,
      "learning_rate": 2.5571196368877847e-05,
      "loss": 0.456,
      "mean_token_accuracy": 0.8536700420081615,
      "num_tokens": 774595287.0,
      "step": 24270
    },
    {
      "entropy": 0.5945634935051203,
      "epoch": 2.233052138860616,
      "grad_norm": 0.16901706159114838,
      "learning_rate": 2.5568129542736225e-05,
      "loss": 0.5808,
      "mean_token_accuracy": 0.8176503479480743,
      "num_tokens": 774627803.0,
      "step": 24271
    },
    {
      "entropy": 0.5704577057622373,
      "epoch": 2.2331441452320573,
      "grad_norm": 0.16741943359375,
      "learning_rate": 2.5565062716594596e-05,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8215742483735085,
      "num_tokens": 774659472.0,
      "step": 24272
    },
    {
      "entropy": 0.4920711163431406,
      "epoch": 2.2332361516034984,
      "grad_norm": 0.1527986079454422,
      "learning_rate": 2.556199589045297e-05,
      "loss": 0.4732,
      "mean_token_accuracy": 0.8486400172114372,
      "num_tokens": 774692212.0,
      "step": 24273
    },
    {
      "entropy": 0.5392793854698539,
      "epoch": 2.23332815797494,
      "grad_norm": 0.16326214373111725,
      "learning_rate": 2.5558929064311342e-05,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8352835848927498,
      "num_tokens": 774723334.0,
      "step": 24274
    },
    {
      "entropy": 0.5053826533257961,
      "epoch": 2.233420164346381,
      "grad_norm": 0.1564423143863678,
      "learning_rate": 2.555586223816972e-05,
      "loss": 0.4837,
      "mean_token_accuracy": 0.8443369828164577,
      "num_tokens": 774755022.0,
      "step": 24275
    },
    {
      "entropy": 0.48460587300360203,
      "epoch": 2.2335121707178223,
      "grad_norm": 0.1601904183626175,
      "learning_rate": 2.555279541202809e-05,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8477864861488342,
      "num_tokens": 774786800.0,
      "step": 24276
    },
    {
      "entropy": 0.5612047985196114,
      "epoch": 2.2336041770892634,
      "grad_norm": 0.16863031685352325,
      "learning_rate": 2.554972858588647e-05,
      "loss": 0.5552,
      "mean_token_accuracy": 0.8273339867591858,
      "num_tokens": 774817936.0,
      "step": 24277
    },
    {
      "entropy": 0.46157117327675223,
      "epoch": 2.2336961834607045,
      "grad_norm": 0.1546473503112793,
      "learning_rate": 2.5546661759744837e-05,
      "loss": 0.4509,
      "mean_token_accuracy": 0.855348814278841,
      "num_tokens": 774849605.0,
      "step": 24278
    },
    {
      "entropy": 0.5831308588385582,
      "epoch": 2.233788189832146,
      "grad_norm": 0.16603517532348633,
      "learning_rate": 2.5543594933603215e-05,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8183613494038582,
      "num_tokens": 774881848.0,
      "step": 24279
    },
    {
      "entropy": 0.5458133108913898,
      "epoch": 2.2338801962035872,
      "grad_norm": 0.16122125089168549,
      "learning_rate": 2.5540528107461586e-05,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8289372846484184,
      "num_tokens": 774913545.0,
      "step": 24280
    },
    {
      "entropy": 0.5482013402506709,
      "epoch": 2.2339722025750284,
      "grad_norm": 0.16053934395313263,
      "learning_rate": 2.5537461281319964e-05,
      "loss": 0.5347,
      "mean_token_accuracy": 0.8302612341940403,
      "num_tokens": 774945524.0,
      "step": 24281
    },
    {
      "entropy": 0.5434949528425932,
      "epoch": 2.2340642089464695,
      "grad_norm": 0.16506780683994293,
      "learning_rate": 2.553439445517834e-05,
      "loss": 0.5249,
      "mean_token_accuracy": 0.8328287079930305,
      "num_tokens": 774976715.0,
      "step": 24282
    },
    {
      "entropy": 0.5656422041356564,
      "epoch": 2.2341562153179106,
      "grad_norm": 0.16663098335266113,
      "learning_rate": 2.553132762903671e-05,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8239072225987911,
      "num_tokens": 775008593.0,
      "step": 24283
    },
    {
      "entropy": 0.6029374152421951,
      "epoch": 2.234248221689352,
      "grad_norm": 0.17342090606689453,
      "learning_rate": 2.552826080289509e-05,
      "loss": 0.6053,
      "mean_token_accuracy": 0.8110905401408672,
      "num_tokens": 775040898.0,
      "step": 24284
    },
    {
      "entropy": 0.4864859892986715,
      "epoch": 2.2343402280607934,
      "grad_norm": 0.15485118329524994,
      "learning_rate": 2.552519397675346e-05,
      "loss": 0.474,
      "mean_token_accuracy": 0.8495302982628345,
      "num_tokens": 775073073.0,
      "step": 24285
    },
    {
      "entropy": 0.5851441640406847,
      "epoch": 2.2344322344322345,
      "grad_norm": 0.17190493643283844,
      "learning_rate": 2.5522127150611834e-05,
      "loss": 0.5672,
      "mean_token_accuracy": 0.8210060261189938,
      "num_tokens": 775104780.0,
      "step": 24286
    },
    {
      "entropy": 0.5287623098120093,
      "epoch": 2.2345242408036756,
      "grad_norm": 0.16459810733795166,
      "learning_rate": 2.5519060324470206e-05,
      "loss": 0.521,
      "mean_token_accuracy": 0.8359221778810024,
      "num_tokens": 775137028.0,
      "step": 24287
    },
    {
      "entropy": 0.5121986698359251,
      "epoch": 2.2346162471751168,
      "grad_norm": 0.1625107079744339,
      "learning_rate": 2.5515993498328584e-05,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8438336588442326,
      "num_tokens": 775169518.0,
      "step": 24288
    },
    {
      "entropy": 0.4842740138992667,
      "epoch": 2.2347082535465583,
      "grad_norm": 0.15271642804145813,
      "learning_rate": 2.5512926672186955e-05,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8518194742500782,
      "num_tokens": 775201041.0,
      "step": 24289
    },
    {
      "entropy": 0.48872956773266196,
      "epoch": 2.2348002599179995,
      "grad_norm": 0.15585537254810333,
      "learning_rate": 2.550985984604533e-05,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8481236025691032,
      "num_tokens": 775233344.0,
      "step": 24290
    },
    {
      "entropy": 0.5497427005320787,
      "epoch": 2.2348922662894406,
      "grad_norm": 0.16481535136699677,
      "learning_rate": 2.55067930199037e-05,
      "loss": 0.5336,
      "mean_token_accuracy": 0.8313035257160664,
      "num_tokens": 775265026.0,
      "step": 24291
    },
    {
      "entropy": 0.5004975832998753,
      "epoch": 2.2349842726608817,
      "grad_norm": 0.15798068046569824,
      "learning_rate": 2.550372619376208e-05,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8422435596585274,
      "num_tokens": 775296832.0,
      "step": 24292
    },
    {
      "entropy": 0.6179549284279346,
      "epoch": 2.235076279032323,
      "grad_norm": 0.16982732713222504,
      "learning_rate": 2.550065936762045e-05,
      "loss": 0.5949,
      "mean_token_accuracy": 0.8118578419089317,
      "num_tokens": 775328874.0,
      "step": 24293
    },
    {
      "entropy": 0.5319094015285373,
      "epoch": 2.2351682854037644,
      "grad_norm": 0.15892596542835236,
      "learning_rate": 2.5497592541478825e-05,
      "loss": 0.5086,
      "mean_token_accuracy": 0.835862472653389,
      "num_tokens": 775360463.0,
      "step": 24294
    },
    {
      "entropy": 0.5362688163295388,
      "epoch": 2.2352602917752056,
      "grad_norm": 0.16037821769714355,
      "learning_rate": 2.5494525715337196e-05,
      "loss": 0.5241,
      "mean_token_accuracy": 0.833959873765707,
      "num_tokens": 775392579.0,
      "step": 24295
    },
    {
      "entropy": 0.5722339199855924,
      "epoch": 2.2353522981466467,
      "grad_norm": 0.16844357550144196,
      "learning_rate": 2.5491458889195574e-05,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8246685266494751,
      "num_tokens": 775424547.0,
      "step": 24296
    },
    {
      "entropy": 0.5404821471311152,
      "epoch": 2.235444304518088,
      "grad_norm": 0.16975654661655426,
      "learning_rate": 2.5488392063053945e-05,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8288358375430107,
      "num_tokens": 775455877.0,
      "step": 24297
    },
    {
      "entropy": 0.6506285755895078,
      "epoch": 2.235536310889529,
      "grad_norm": 0.1706564724445343,
      "learning_rate": 2.5485325236912323e-05,
      "loss": 0.6363,
      "mean_token_accuracy": 0.8011028692126274,
      "num_tokens": 775487875.0,
      "step": 24298
    },
    {
      "entropy": 0.4808059260249138,
      "epoch": 2.2356283172609706,
      "grad_norm": 0.15541671216487885,
      "learning_rate": 2.5482258410770695e-05,
      "loss": 0.4611,
      "mean_token_accuracy": 0.8506333529949188,
      "num_tokens": 775519790.0,
      "step": 24299
    },
    {
      "entropy": 0.50106131285429,
      "epoch": 2.2357203236324117,
      "grad_norm": 0.16024284064769745,
      "learning_rate": 2.547919158462907e-05,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8424068167805672,
      "num_tokens": 775551762.0,
      "step": 24300
    },
    {
      "entropy": 0.495174958370626,
      "epoch": 2.235812330003853,
      "grad_norm": 0.16298136115074158,
      "learning_rate": 2.547612475848744e-05,
      "loss": 0.4827,
      "mean_token_accuracy": 0.8485866822302341,
      "num_tokens": 775582818.0,
      "step": 24301
    },
    {
      "entropy": 0.5220433715730906,
      "epoch": 2.235904336375294,
      "grad_norm": 0.17122897505760193,
      "learning_rate": 2.547305793234582e-05,
      "loss": 0.5144,
      "mean_token_accuracy": 0.8308790065348148,
      "num_tokens": 775615076.0,
      "step": 24302
    },
    {
      "entropy": 0.4616565643809736,
      "epoch": 2.235996342746735,
      "grad_norm": 0.15702515840530396,
      "learning_rate": 2.546999110620419e-05,
      "loss": 0.4398,
      "mean_token_accuracy": 0.8569435998797417,
      "num_tokens": 775646447.0,
      "step": 24303
    },
    {
      "entropy": 0.5251553254202008,
      "epoch": 2.2360883491181767,
      "grad_norm": 0.16119636595249176,
      "learning_rate": 2.5466924280062564e-05,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8412710428237915,
      "num_tokens": 775678218.0,
      "step": 24304
    },
    {
      "entropy": 0.5119995903223753,
      "epoch": 2.236180355489618,
      "grad_norm": 0.15722234547138214,
      "learning_rate": 2.5463857453920936e-05,
      "loss": 0.5006,
      "mean_token_accuracy": 0.8385735973715782,
      "num_tokens": 775710356.0,
      "step": 24305
    },
    {
      "entropy": 0.5194174293428659,
      "epoch": 2.236272361861059,
      "grad_norm": 0.1602988988161087,
      "learning_rate": 2.5460790627779314e-05,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8397184312343597,
      "num_tokens": 775741908.0,
      "step": 24306
    },
    {
      "entropy": 0.6021772241219878,
      "epoch": 2.2363643682325,
      "grad_norm": 0.17033201456069946,
      "learning_rate": 2.5457723801637685e-05,
      "loss": 0.5825,
      "mean_token_accuracy": 0.819297481328249,
      "num_tokens": 775773881.0,
      "step": 24307
    },
    {
      "entropy": 0.5035825171507895,
      "epoch": 2.236456374603941,
      "grad_norm": 0.15545876324176788,
      "learning_rate": 2.545465697549606e-05,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8446627594530582,
      "num_tokens": 775805793.0,
      "step": 24308
    },
    {
      "entropy": 0.5143731497228146,
      "epoch": 2.2365483809753828,
      "grad_norm": 0.16165082156658173,
      "learning_rate": 2.545159014935443e-05,
      "loss": 0.509,
      "mean_token_accuracy": 0.8400025926530361,
      "num_tokens": 775837419.0,
      "step": 24309
    },
    {
      "entropy": 0.6064698565751314,
      "epoch": 2.236640387346824,
      "grad_norm": 0.16827626526355743,
      "learning_rate": 2.544852332321281e-05,
      "loss": 0.5803,
      "mean_token_accuracy": 0.8159987926483154,
      "num_tokens": 775869253.0,
      "step": 24310
    },
    {
      "entropy": 0.5096973879262805,
      "epoch": 2.236732393718265,
      "grad_norm": 0.15669800341129303,
      "learning_rate": 2.544545649707118e-05,
      "loss": 0.4851,
      "mean_token_accuracy": 0.8466968983411789,
      "num_tokens": 775900950.0,
      "step": 24311
    },
    {
      "entropy": 0.36031862278468907,
      "epoch": 2.236824400089706,
      "grad_norm": 0.1369396299123764,
      "learning_rate": 2.5442389670929555e-05,
      "loss": 0.3466,
      "mean_token_accuracy": 0.8887723349034786,
      "num_tokens": 775933392.0,
      "step": 24312
    },
    {
      "entropy": 0.5206055929884315,
      "epoch": 2.2369164064611473,
      "grad_norm": 0.1572647988796234,
      "learning_rate": 2.5439322844787933e-05,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8405206575989723,
      "num_tokens": 775965465.0,
      "step": 24313
    },
    {
      "entropy": 0.49490527040325105,
      "epoch": 2.237008412832589,
      "grad_norm": 0.1548217535018921,
      "learning_rate": 2.5436256018646304e-05,
      "loss": 0.4793,
      "mean_token_accuracy": 0.8490888811647892,
      "num_tokens": 775997520.0,
      "step": 24314
    },
    {
      "entropy": 0.5647716643288732,
      "epoch": 2.23710041920403,
      "grad_norm": 0.16890667378902435,
      "learning_rate": 2.5433189192504682e-05,
      "loss": 0.5592,
      "mean_token_accuracy": 0.8211259469389915,
      "num_tokens": 776030114.0,
      "step": 24315
    },
    {
      "entropy": 0.5357442833483219,
      "epoch": 2.237192425575471,
      "grad_norm": 0.1581505686044693,
      "learning_rate": 2.543012236636305e-05,
      "loss": 0.5181,
      "mean_token_accuracy": 0.8353151753544807,
      "num_tokens": 776062342.0,
      "step": 24316
    },
    {
      "entropy": 0.5412068255245686,
      "epoch": 2.2372844319469123,
      "grad_norm": 0.16845126450061798,
      "learning_rate": 2.5427055540221428e-05,
      "loss": 0.5223,
      "mean_token_accuracy": 0.8339249566197395,
      "num_tokens": 776094136.0,
      "step": 24317
    },
    {
      "entropy": 0.5481019765138626,
      "epoch": 2.2373764383183534,
      "grad_norm": 0.16919121146202087,
      "learning_rate": 2.54239887140798e-05,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8315428495407104,
      "num_tokens": 776125166.0,
      "step": 24318
    },
    {
      "entropy": 0.4751322837546468,
      "epoch": 2.237468444689795,
      "grad_norm": 0.15921860933303833,
      "learning_rate": 2.5420921887938177e-05,
      "loss": 0.458,
      "mean_token_accuracy": 0.8537267185747623,
      "num_tokens": 776156927.0,
      "step": 24319
    },
    {
      "entropy": 0.506112705450505,
      "epoch": 2.237560451061236,
      "grad_norm": 0.15846627950668335,
      "learning_rate": 2.541785506179655e-05,
      "loss": 0.4876,
      "mean_token_accuracy": 0.8442728407680988,
      "num_tokens": 776188487.0,
      "step": 24320
    },
    {
      "entropy": 0.5820613242685795,
      "epoch": 2.2376524574326773,
      "grad_norm": 0.17129819095134735,
      "learning_rate": 2.5414788235654923e-05,
      "loss": 0.5792,
      "mean_token_accuracy": 0.8159822188317776,
      "num_tokens": 776220894.0,
      "step": 24321
    },
    {
      "entropy": 0.5045490614138544,
      "epoch": 2.2377444638041184,
      "grad_norm": 0.16366977989673615,
      "learning_rate": 2.5411721409513295e-05,
      "loss": 0.4893,
      "mean_token_accuracy": 0.8440272435545921,
      "num_tokens": 776252976.0,
      "step": 24322
    },
    {
      "entropy": 0.5627283174544573,
      "epoch": 2.2378364701755595,
      "grad_norm": 0.16790516674518585,
      "learning_rate": 2.5408654583371673e-05,
      "loss": 0.5539,
      "mean_token_accuracy": 0.823657676577568,
      "num_tokens": 776284690.0,
      "step": 24323
    },
    {
      "entropy": 0.5292501971125603,
      "epoch": 2.237928476547001,
      "grad_norm": 0.16272322833538055,
      "learning_rate": 2.5405587757230044e-05,
      "loss": 0.509,
      "mean_token_accuracy": 0.8403150960803032,
      "num_tokens": 776317458.0,
      "step": 24324
    },
    {
      "entropy": 0.5217741560190916,
      "epoch": 2.2380204829184422,
      "grad_norm": 0.1655840426683426,
      "learning_rate": 2.540252093108842e-05,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8416092246770859,
      "num_tokens": 776349056.0,
      "step": 24325
    },
    {
      "entropy": 0.5828493973240256,
      "epoch": 2.2381124892898834,
      "grad_norm": 0.17143860459327698,
      "learning_rate": 2.539945410494679e-05,
      "loss": 0.5751,
      "mean_token_accuracy": 0.817868996411562,
      "num_tokens": 776381336.0,
      "step": 24326
    },
    {
      "entropy": 0.4871199680492282,
      "epoch": 2.2382044956613245,
      "grad_norm": 0.16225358843803406,
      "learning_rate": 2.5396387278805168e-05,
      "loss": 0.4913,
      "mean_token_accuracy": 0.843885026872158,
      "num_tokens": 776413088.0,
      "step": 24327
    },
    {
      "entropy": 0.5522152204066515,
      "epoch": 2.2382965020327656,
      "grad_norm": 0.1692233383655548,
      "learning_rate": 2.539332045266354e-05,
      "loss": 0.5388,
      "mean_token_accuracy": 0.8300964161753654,
      "num_tokens": 776444675.0,
      "step": 24328
    },
    {
      "entropy": 0.5452895136550069,
      "epoch": 2.2383885084042072,
      "grad_norm": 0.16591620445251465,
      "learning_rate": 2.5390253626521914e-05,
      "loss": 0.5407,
      "mean_token_accuracy": 0.8283675536513329,
      "num_tokens": 776476441.0,
      "step": 24329
    },
    {
      "entropy": 0.5817286632955074,
      "epoch": 2.2384805147756484,
      "grad_norm": 0.17347632348537445,
      "learning_rate": 2.5387186800380285e-05,
      "loss": 0.5759,
      "mean_token_accuracy": 0.8183329701423645,
      "num_tokens": 776508159.0,
      "step": 24330
    },
    {
      "entropy": 0.5358168045058846,
      "epoch": 2.2385725211470895,
      "grad_norm": 0.16013237833976746,
      "learning_rate": 2.5384119974238663e-05,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8308964818716049,
      "num_tokens": 776540591.0,
      "step": 24331
    },
    {
      "entropy": 0.5316958306357265,
      "epoch": 2.2386645275185306,
      "grad_norm": 0.16230374574661255,
      "learning_rate": 2.5381053148097034e-05,
      "loss": 0.5184,
      "mean_token_accuracy": 0.8297245986759663,
      "num_tokens": 776571687.0,
      "step": 24332
    },
    {
      "entropy": 0.5510586407035589,
      "epoch": 2.2387565338899718,
      "grad_norm": 0.17501583695411682,
      "learning_rate": 2.537798632195541e-05,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8324394300580025,
      "num_tokens": 776603634.0,
      "step": 24333
    },
    {
      "entropy": 0.5334829948842525,
      "epoch": 2.2388485402614133,
      "grad_norm": 0.16416391730308533,
      "learning_rate": 2.537491949581378e-05,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8328782953321934,
      "num_tokens": 776635711.0,
      "step": 24334
    },
    {
      "entropy": 0.484164847061038,
      "epoch": 2.2389405466328545,
      "grad_norm": 0.1538373827934265,
      "learning_rate": 2.5371852669672158e-05,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8493463285267353,
      "num_tokens": 776668313.0,
      "step": 24335
    },
    {
      "entropy": 0.47262758878059685,
      "epoch": 2.2390325530042956,
      "grad_norm": 0.15219049155712128,
      "learning_rate": 2.536878584353053e-05,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8546391576528549,
      "num_tokens": 776700190.0,
      "step": 24336
    },
    {
      "entropy": 0.5487995930016041,
      "epoch": 2.2391245593757367,
      "grad_norm": 0.15739716589450836,
      "learning_rate": 2.5365719017388907e-05,
      "loss": 0.5285,
      "mean_token_accuracy": 0.8381688259541988,
      "num_tokens": 776732687.0,
      "step": 24337
    },
    {
      "entropy": 0.5696483151987195,
      "epoch": 2.239216565747178,
      "grad_norm": 0.17243815958499908,
      "learning_rate": 2.5362652191247275e-05,
      "loss": 0.5667,
      "mean_token_accuracy": 0.8219850175082684,
      "num_tokens": 776764468.0,
      "step": 24338
    },
    {
      "entropy": 0.5513638705015182,
      "epoch": 2.2393085721186194,
      "grad_norm": 0.1593128889799118,
      "learning_rate": 2.5359585365105653e-05,
      "loss": 0.5346,
      "mean_token_accuracy": 0.8316337466239929,
      "num_tokens": 776797010.0,
      "step": 24339
    },
    {
      "entropy": 0.4262244099518284,
      "epoch": 2.2394005784900606,
      "grad_norm": 0.14796385169029236,
      "learning_rate": 2.5356518538964025e-05,
      "loss": 0.4091,
      "mean_token_accuracy": 0.8688991665840149,
      "num_tokens": 776829158.0,
      "step": 24340
    },
    {
      "entropy": 0.48363923362921923,
      "epoch": 2.2394925848615017,
      "grad_norm": 0.1592073291540146,
      "learning_rate": 2.5353451712822403e-05,
      "loss": 0.4715,
      "mean_token_accuracy": 0.8520999848842621,
      "num_tokens": 776861767.0,
      "step": 24341
    },
    {
      "entropy": 0.6677848733961582,
      "epoch": 2.239584591232943,
      "grad_norm": 0.18035776913166046,
      "learning_rate": 2.5350384886680777e-05,
      "loss": 0.6548,
      "mean_token_accuracy": 0.790160272270441,
      "num_tokens": 776893197.0,
      "step": 24342
    },
    {
      "entropy": 0.5356016643345356,
      "epoch": 2.239676597604384,
      "grad_norm": 0.16154247522354126,
      "learning_rate": 2.534731806053915e-05,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8373936228454113,
      "num_tokens": 776924823.0,
      "step": 24343
    },
    {
      "entropy": 0.5159068126231432,
      "epoch": 2.2397686039758256,
      "grad_norm": 0.16089223325252533,
      "learning_rate": 2.5344251234397527e-05,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8412481993436813,
      "num_tokens": 776957192.0,
      "step": 24344
    },
    {
      "entropy": 0.5054287742823362,
      "epoch": 2.2398606103472667,
      "grad_norm": 0.1536366194486618,
      "learning_rate": 2.5341184408255898e-05,
      "loss": 0.4912,
      "mean_token_accuracy": 0.8435785584151745,
      "num_tokens": 776989706.0,
      "step": 24345
    },
    {
      "entropy": 0.5940820770338178,
      "epoch": 2.239952616718708,
      "grad_norm": 0.17417249083518982,
      "learning_rate": 2.5338117582114273e-05,
      "loss": 0.5912,
      "mean_token_accuracy": 0.8197538778185844,
      "num_tokens": 777021419.0,
      "step": 24346
    },
    {
      "entropy": 0.49537994572892785,
      "epoch": 2.240044623090149,
      "grad_norm": 0.16150154173374176,
      "learning_rate": 2.5335050755972644e-05,
      "loss": 0.4797,
      "mean_token_accuracy": 0.8461641930043697,
      "num_tokens": 777052494.0,
      "step": 24347
    },
    {
      "entropy": 0.5339823719114065,
      "epoch": 2.24013662946159,
      "grad_norm": 0.16548269987106323,
      "learning_rate": 2.5331983929831022e-05,
      "loss": 0.5215,
      "mean_token_accuracy": 0.8291599564254284,
      "num_tokens": 777084375.0,
      "step": 24348
    },
    {
      "entropy": 0.41368464194238186,
      "epoch": 2.2402286358330317,
      "grad_norm": 0.14939239621162415,
      "learning_rate": 2.5328917103689393e-05,
      "loss": 0.3997,
      "mean_token_accuracy": 0.8708359077572823,
      "num_tokens": 777116371.0,
      "step": 24349
    },
    {
      "entropy": 0.4702588440850377,
      "epoch": 2.240320642204473,
      "grad_norm": 0.15341639518737793,
      "learning_rate": 2.5325850277547768e-05,
      "loss": 0.4592,
      "mean_token_accuracy": 0.8552411906421185,
      "num_tokens": 777148203.0,
      "step": 24350
    },
    {
      "entropy": 0.4613325875252485,
      "epoch": 2.240412648575914,
      "grad_norm": 0.16294576227664948,
      "learning_rate": 2.532278345140614e-05,
      "loss": 0.4464,
      "mean_token_accuracy": 0.8577227406203747,
      "num_tokens": 777180197.0,
      "step": 24351
    },
    {
      "entropy": 0.5858495971187949,
      "epoch": 2.240504654947355,
      "grad_norm": 0.16987313330173492,
      "learning_rate": 2.5319716625264517e-05,
      "loss": 0.5803,
      "mean_token_accuracy": 0.8179271593689919,
      "num_tokens": 777212158.0,
      "step": 24352
    },
    {
      "entropy": 0.4230671380646527,
      "epoch": 2.240596661318796,
      "grad_norm": 0.1467333734035492,
      "learning_rate": 2.5316649799122888e-05,
      "loss": 0.4099,
      "mean_token_accuracy": 0.8649024739861488,
      "num_tokens": 777244697.0,
      "step": 24353
    },
    {
      "entropy": 0.4792892299592495,
      "epoch": 2.240688667690238,
      "grad_norm": 0.15717226266860962,
      "learning_rate": 2.5313582972981263e-05,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8521118760108948,
      "num_tokens": 777276869.0,
      "step": 24354
    },
    {
      "entropy": 0.5828973157331347,
      "epoch": 2.240780674061679,
      "grad_norm": 0.16724304854869843,
      "learning_rate": 2.5310516146839634e-05,
      "loss": 0.5668,
      "mean_token_accuracy": 0.8199716433882713,
      "num_tokens": 777308440.0,
      "step": 24355
    },
    {
      "entropy": 0.43065671622753143,
      "epoch": 2.24087268043312,
      "grad_norm": 0.15203317999839783,
      "learning_rate": 2.5307449320698012e-05,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8657919950783253,
      "num_tokens": 777340816.0,
      "step": 24356
    },
    {
      "entropy": 0.5334430038928986,
      "epoch": 2.240964686804561,
      "grad_norm": 0.16005828976631165,
      "learning_rate": 2.5304382494556383e-05,
      "loss": 0.5292,
      "mean_token_accuracy": 0.8323092460632324,
      "num_tokens": 777373370.0,
      "step": 24357
    },
    {
      "entropy": 0.5612389110028744,
      "epoch": 2.2410566931760023,
      "grad_norm": 0.16833429038524628,
      "learning_rate": 2.530131566841476e-05,
      "loss": 0.542,
      "mean_token_accuracy": 0.8265917338430882,
      "num_tokens": 777404648.0,
      "step": 24358
    },
    {
      "entropy": 0.46492713689804077,
      "epoch": 2.241148699547444,
      "grad_norm": 0.15303367376327515,
      "learning_rate": 2.5298248842273133e-05,
      "loss": 0.4521,
      "mean_token_accuracy": 0.8549667969346046,
      "num_tokens": 777436968.0,
      "step": 24359
    },
    {
      "entropy": 0.4836890965234488,
      "epoch": 2.241240705918885,
      "grad_norm": 0.1573108732700348,
      "learning_rate": 2.5295182016131507e-05,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8484896086156368,
      "num_tokens": 777469261.0,
      "step": 24360
    },
    {
      "entropy": 0.5168397426605225,
      "epoch": 2.241332712290326,
      "grad_norm": 0.1639476716518402,
      "learning_rate": 2.529211518998988e-05,
      "loss": 0.4992,
      "mean_token_accuracy": 0.84162612631917,
      "num_tokens": 777500524.0,
      "step": 24361
    },
    {
      "entropy": 0.5430391766130924,
      "epoch": 2.2414247186617673,
      "grad_norm": 0.1646198034286499,
      "learning_rate": 2.5289048363848257e-05,
      "loss": 0.5276,
      "mean_token_accuracy": 0.8345617540180683,
      "num_tokens": 777532451.0,
      "step": 24362
    },
    {
      "entropy": 0.476382315158844,
      "epoch": 2.2415167250332084,
      "grad_norm": 0.16003717482089996,
      "learning_rate": 2.5285981537706628e-05,
      "loss": 0.4609,
      "mean_token_accuracy": 0.8531097061932087,
      "num_tokens": 777564273.0,
      "step": 24363
    },
    {
      "entropy": 0.6268751453608274,
      "epoch": 2.24160873140465,
      "grad_norm": 0.17868517339229584,
      "learning_rate": 2.5282914711565003e-05,
      "loss": 0.6191,
      "mean_token_accuracy": 0.8056418374180794,
      "num_tokens": 777596072.0,
      "step": 24364
    },
    {
      "entropy": 0.5439441259950399,
      "epoch": 2.241700737776091,
      "grad_norm": 0.16243098676204681,
      "learning_rate": 2.5279847885423374e-05,
      "loss": 0.5413,
      "mean_token_accuracy": 0.8330697789788246,
      "num_tokens": 777628480.0,
      "step": 24365
    },
    {
      "entropy": 0.49746915604919195,
      "epoch": 2.2417927441475323,
      "grad_norm": 0.15644817054271698,
      "learning_rate": 2.5276781059281752e-05,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8452222049236298,
      "num_tokens": 777660763.0,
      "step": 24366
    },
    {
      "entropy": 0.5083156190812588,
      "epoch": 2.2418847505189734,
      "grad_norm": 0.16087040305137634,
      "learning_rate": 2.5273714233140123e-05,
      "loss": 0.4847,
      "mean_token_accuracy": 0.8468571379780769,
      "num_tokens": 777691828.0,
      "step": 24367
    },
    {
      "entropy": 0.5523584932088852,
      "epoch": 2.2419767568904145,
      "grad_norm": 0.1656283140182495,
      "learning_rate": 2.5270647406998498e-05,
      "loss": 0.54,
      "mean_token_accuracy": 0.8332433439791203,
      "num_tokens": 777723193.0,
      "step": 24368
    },
    {
      "entropy": 0.5517128221690655,
      "epoch": 2.242068763261856,
      "grad_norm": 0.16117045283317566,
      "learning_rate": 2.526758058085687e-05,
      "loss": 0.5395,
      "mean_token_accuracy": 0.8293881081044674,
      "num_tokens": 777755159.0,
      "step": 24369
    },
    {
      "entropy": 0.43457604572176933,
      "epoch": 2.2421607696332972,
      "grad_norm": 0.14923226833343506,
      "learning_rate": 2.5264513754715247e-05,
      "loss": 0.4246,
      "mean_token_accuracy": 0.8633482195436954,
      "num_tokens": 777787014.0,
      "step": 24370
    },
    {
      "entropy": 0.5166741041466594,
      "epoch": 2.2422527760047384,
      "grad_norm": 0.15770655870437622,
      "learning_rate": 2.526144692857362e-05,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8428030759096146,
      "num_tokens": 777818252.0,
      "step": 24371
    },
    {
      "entropy": 0.5246403128840029,
      "epoch": 2.2423447823761795,
      "grad_norm": 0.1602787971496582,
      "learning_rate": 2.5258380102431993e-05,
      "loss": 0.4994,
      "mean_token_accuracy": 0.8391195870935917,
      "num_tokens": 777850635.0,
      "step": 24372
    },
    {
      "entropy": 0.49173396825790405,
      "epoch": 2.2424367887476206,
      "grad_norm": 0.1604403555393219,
      "learning_rate": 2.525531327629037e-05,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8434641771018505,
      "num_tokens": 777881958.0,
      "step": 24373
    },
    {
      "entropy": 0.5485919881612062,
      "epoch": 2.2425287951190622,
      "grad_norm": 0.16202595829963684,
      "learning_rate": 2.5252246450148742e-05,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8337739296257496,
      "num_tokens": 777913737.0,
      "step": 24374
    },
    {
      "entropy": 0.49599481746554375,
      "epoch": 2.2426208014905034,
      "grad_norm": 0.15658308565616608,
      "learning_rate": 2.524917962400712e-05,
      "loss": 0.4786,
      "mean_token_accuracy": 0.8478347584605217,
      "num_tokens": 777945670.0,
      "step": 24375
    },
    {
      "entropy": 0.5924332290887833,
      "epoch": 2.2427128078619445,
      "grad_norm": 0.17418518662452698,
      "learning_rate": 2.5246112797865488e-05,
      "loss": 0.5813,
      "mean_token_accuracy": 0.8177294619381428,
      "num_tokens": 777977196.0,
      "step": 24376
    },
    {
      "entropy": 0.4989653145894408,
      "epoch": 2.2428048142333856,
      "grad_norm": 0.16107451915740967,
      "learning_rate": 2.5243045971723866e-05,
      "loss": 0.4928,
      "mean_token_accuracy": 0.8466594256460667,
      "num_tokens": 778009415.0,
      "step": 24377
    },
    {
      "entropy": 0.527044277638197,
      "epoch": 2.2428968206048268,
      "grad_norm": 0.16310612857341766,
      "learning_rate": 2.5239979145582238e-05,
      "loss": 0.5184,
      "mean_token_accuracy": 0.8345987237989902,
      "num_tokens": 778040767.0,
      "step": 24378
    },
    {
      "entropy": 0.46537351119332016,
      "epoch": 2.2429888269762683,
      "grad_norm": 0.15448464453220367,
      "learning_rate": 2.5236912319440616e-05,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8554851114749908,
      "num_tokens": 778072063.0,
      "step": 24379
    },
    {
      "entropy": 0.5333629101514816,
      "epoch": 2.2430808333477095,
      "grad_norm": 0.165225550532341,
      "learning_rate": 2.5233845493298987e-05,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8333745561540127,
      "num_tokens": 778103892.0,
      "step": 24380
    },
    {
      "entropy": 0.5864412002265453,
      "epoch": 2.2431728397191506,
      "grad_norm": 0.16616566479206085,
      "learning_rate": 2.523077866715736e-05,
      "loss": 0.5766,
      "mean_token_accuracy": 0.8167502582073212,
      "num_tokens": 778136286.0,
      "step": 24381
    },
    {
      "entropy": 0.47023992077447474,
      "epoch": 2.2432648460905917,
      "grad_norm": 0.15855543315410614,
      "learning_rate": 2.5227711841015733e-05,
      "loss": 0.4586,
      "mean_token_accuracy": 0.8508500009775162,
      "num_tokens": 778168105.0,
      "step": 24382
    },
    {
      "entropy": 0.5161020485684276,
      "epoch": 2.243356852462033,
      "grad_norm": 0.15277431905269623,
      "learning_rate": 2.522464501487411e-05,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8432872779667377,
      "num_tokens": 778200245.0,
      "step": 24383
    },
    {
      "entropy": 0.5153476647101343,
      "epoch": 2.2434488588334744,
      "grad_norm": 0.1645563542842865,
      "learning_rate": 2.5221578188732482e-05,
      "loss": 0.5044,
      "mean_token_accuracy": 0.840835515409708,
      "num_tokens": 778232158.0,
      "step": 24384
    },
    {
      "entropy": 0.5711097605526447,
      "epoch": 2.2435408652049156,
      "grad_norm": 0.1661444753408432,
      "learning_rate": 2.5218511362590857e-05,
      "loss": 0.5519,
      "mean_token_accuracy": 0.8210692182183266,
      "num_tokens": 778263929.0,
      "step": 24385
    },
    {
      "entropy": 0.5389080327004194,
      "epoch": 2.2436328715763567,
      "grad_norm": 0.16397835314273834,
      "learning_rate": 2.5215444536449228e-05,
      "loss": 0.5255,
      "mean_token_accuracy": 0.8336172662675381,
      "num_tokens": 778294888.0,
      "step": 24386
    },
    {
      "entropy": 0.5723375994712114,
      "epoch": 2.243724877947798,
      "grad_norm": 0.1744144856929779,
      "learning_rate": 2.5212377710307606e-05,
      "loss": 0.5548,
      "mean_token_accuracy": 0.8236683160066605,
      "num_tokens": 778326797.0,
      "step": 24387
    },
    {
      "entropy": 0.5896652229130268,
      "epoch": 2.243816884319239,
      "grad_norm": 0.1693420559167862,
      "learning_rate": 2.5209310884165977e-05,
      "loss": 0.58,
      "mean_token_accuracy": 0.8153375424444675,
      "num_tokens": 778359139.0,
      "step": 24388
    },
    {
      "entropy": 0.5235397852957249,
      "epoch": 2.2439088906906806,
      "grad_norm": 0.16081227362155914,
      "learning_rate": 2.5206244058024352e-05,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8391230143606663,
      "num_tokens": 778391249.0,
      "step": 24389
    },
    {
      "entropy": 0.5020115179941058,
      "epoch": 2.2440008970621217,
      "grad_norm": 0.15916767716407776,
      "learning_rate": 2.5203177231882723e-05,
      "loss": 0.4789,
      "mean_token_accuracy": 0.8484097085893154,
      "num_tokens": 778422992.0,
      "step": 24390
    },
    {
      "entropy": 0.5840877443552017,
      "epoch": 2.244092903433563,
      "grad_norm": 0.16935677826404572,
      "learning_rate": 2.52001104057411e-05,
      "loss": 0.5759,
      "mean_token_accuracy": 0.8173581138253212,
      "num_tokens": 778455525.0,
      "step": 24391
    },
    {
      "entropy": 0.5147940805181861,
      "epoch": 2.244184909805004,
      "grad_norm": 0.1608891636133194,
      "learning_rate": 2.5197043579599472e-05,
      "loss": 0.5063,
      "mean_token_accuracy": 0.83915850892663,
      "num_tokens": 778487630.0,
      "step": 24392
    },
    {
      "entropy": 0.5169781651347876,
      "epoch": 2.244276916176445,
      "grad_norm": 0.16299013793468475,
      "learning_rate": 2.5193976753457847e-05,
      "loss": 0.497,
      "mean_token_accuracy": 0.836613267660141,
      "num_tokens": 778519234.0,
      "step": 24393
    },
    {
      "entropy": 0.5107710575684905,
      "epoch": 2.2443689225478867,
      "grad_norm": 0.16086171567440033,
      "learning_rate": 2.519090992731622e-05,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8425137288868427,
      "num_tokens": 778550769.0,
      "step": 24394
    },
    {
      "entropy": 0.5328623354434967,
      "epoch": 2.244460928919328,
      "grad_norm": 0.16885226964950562,
      "learning_rate": 2.5187843101174596e-05,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8348193950951099,
      "num_tokens": 778582377.0,
      "step": 24395
    },
    {
      "entropy": 0.4743041954934597,
      "epoch": 2.244552935290769,
      "grad_norm": 0.15618373453617096,
      "learning_rate": 2.5184776275032968e-05,
      "loss": 0.4648,
      "mean_token_accuracy": 0.8546303920447826,
      "num_tokens": 778614125.0,
      "step": 24396
    },
    {
      "entropy": 0.5729340286925435,
      "epoch": 2.24464494166221,
      "grad_norm": 0.1672794669866562,
      "learning_rate": 2.5181709448891346e-05,
      "loss": 0.5597,
      "mean_token_accuracy": 0.820523664355278,
      "num_tokens": 778646511.0,
      "step": 24397
    },
    {
      "entropy": 0.43367901910096407,
      "epoch": 2.244736948033651,
      "grad_norm": 0.15449738502502441,
      "learning_rate": 2.5178642622749714e-05,
      "loss": 0.421,
      "mean_token_accuracy": 0.8664235845208168,
      "num_tokens": 778677162.0,
      "step": 24398
    },
    {
      "entropy": 0.48690815502777696,
      "epoch": 2.244828954405093,
      "grad_norm": 0.15274816751480103,
      "learning_rate": 2.517557579660809e-05,
      "loss": 0.4641,
      "mean_token_accuracy": 0.8516839258372784,
      "num_tokens": 778709755.0,
      "step": 24399
    },
    {
      "entropy": 0.541214563883841,
      "epoch": 2.244920960776534,
      "grad_norm": 0.16176864504814148,
      "learning_rate": 2.5172508970466463e-05,
      "loss": 0.5395,
      "mean_token_accuracy": 0.8299031779170036,
      "num_tokens": 778741990.0,
      "step": 24400
    },
    {
      "entropy": 0.534924385137856,
      "epoch": 2.245012967147975,
      "grad_norm": 0.1629367619752884,
      "learning_rate": 2.516944214432484e-05,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8327656723558903,
      "num_tokens": 778774305.0,
      "step": 24401
    },
    {
      "entropy": 0.4424242451786995,
      "epoch": 2.245104973519416,
      "grad_norm": 0.15352459251880646,
      "learning_rate": 2.5166375318183212e-05,
      "loss": 0.4287,
      "mean_token_accuracy": 0.8639959990978241,
      "num_tokens": 778806795.0,
      "step": 24402
    },
    {
      "entropy": 0.5612744097597897,
      "epoch": 2.2451969798908573,
      "grad_norm": 0.1678120493888855,
      "learning_rate": 2.5163308492041587e-05,
      "loss": 0.5422,
      "mean_token_accuracy": 0.828279260545969,
      "num_tokens": 778838043.0,
      "step": 24403
    },
    {
      "entropy": 0.624454703181982,
      "epoch": 2.245288986262299,
      "grad_norm": 0.1742134541273117,
      "learning_rate": 2.5160241665899965e-05,
      "loss": 0.6149,
      "mean_token_accuracy": 0.8054264448583126,
      "num_tokens": 778869888.0,
      "step": 24404
    },
    {
      "entropy": 0.4553936976008117,
      "epoch": 2.24538099263374,
      "grad_norm": 0.14954820275306702,
      "learning_rate": 2.5157174839758336e-05,
      "loss": 0.4424,
      "mean_token_accuracy": 0.8556513711810112,
      "num_tokens": 778902048.0,
      "step": 24405
    },
    {
      "entropy": 0.5423066350631416,
      "epoch": 2.245472999005181,
      "grad_norm": 0.16385459899902344,
      "learning_rate": 2.515410801361671e-05,
      "loss": 0.5252,
      "mean_token_accuracy": 0.8335962072014809,
      "num_tokens": 778933764.0,
      "step": 24406
    },
    {
      "entropy": 0.5633779326453805,
      "epoch": 2.2455650053766223,
      "grad_norm": 0.16928386688232422,
      "learning_rate": 2.5151041187475082e-05,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8284667693078518,
      "num_tokens": 778965664.0,
      "step": 24407
    },
    {
      "entropy": 0.5440976684913039,
      "epoch": 2.2456570117480634,
      "grad_norm": 0.16413433849811554,
      "learning_rate": 2.514797436133346e-05,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8332220017910004,
      "num_tokens": 778997309.0,
      "step": 24408
    },
    {
      "entropy": 0.42302100267261267,
      "epoch": 2.245749018119505,
      "grad_norm": 0.14876329898834229,
      "learning_rate": 2.514490753519183e-05,
      "loss": 0.4087,
      "mean_token_accuracy": 0.8708441369235516,
      "num_tokens": 779029692.0,
      "step": 24409
    },
    {
      "entropy": 0.4372321502305567,
      "epoch": 2.245841024490946,
      "grad_norm": 0.15705755352973938,
      "learning_rate": 2.5141840709050206e-05,
      "loss": 0.4257,
      "mean_token_accuracy": 0.8605253174901009,
      "num_tokens": 779061958.0,
      "step": 24410
    },
    {
      "entropy": 0.5268369314726442,
      "epoch": 2.2459330308623873,
      "grad_norm": 0.1642853170633316,
      "learning_rate": 2.5138773882908577e-05,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8378218971192837,
      "num_tokens": 779093949.0,
      "step": 24411
    },
    {
      "entropy": 0.5627773851156235,
      "epoch": 2.2460250372338284,
      "grad_norm": 0.16713640093803406,
      "learning_rate": 2.5135707056766955e-05,
      "loss": 0.539,
      "mean_token_accuracy": 0.8283892013132572,
      "num_tokens": 779125237.0,
      "step": 24412
    },
    {
      "entropy": 0.5697733114939183,
      "epoch": 2.2461170436052695,
      "grad_norm": 0.16689471900463104,
      "learning_rate": 2.5132640230625326e-05,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8266318514943123,
      "num_tokens": 779157015.0,
      "step": 24413
    },
    {
      "entropy": 0.5174546595662832,
      "epoch": 2.246209049976711,
      "grad_norm": 0.1619732826948166,
      "learning_rate": 2.51295734044837e-05,
      "loss": 0.4936,
      "mean_token_accuracy": 0.8468924798071384,
      "num_tokens": 779189493.0,
      "step": 24414
    },
    {
      "entropy": 0.5430443864315748,
      "epoch": 2.2463010563481522,
      "grad_norm": 0.1648133248090744,
      "learning_rate": 2.5126506578342072e-05,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8351161368191242,
      "num_tokens": 779220934.0,
      "step": 24415
    },
    {
      "entropy": 0.4496206697076559,
      "epoch": 2.2463930627195934,
      "grad_norm": 0.15882469713687897,
      "learning_rate": 2.512343975220045e-05,
      "loss": 0.439,
      "mean_token_accuracy": 0.8610536381602287,
      "num_tokens": 779252469.0,
      "step": 24416
    },
    {
      "entropy": 0.525094854645431,
      "epoch": 2.2464850690910345,
      "grad_norm": 0.16478121280670166,
      "learning_rate": 2.512037292605882e-05,
      "loss": 0.5123,
      "mean_token_accuracy": 0.8357397168874741,
      "num_tokens": 779283898.0,
      "step": 24417
    },
    {
      "entropy": 0.5107376324012876,
      "epoch": 2.2465770754624756,
      "grad_norm": 0.1605338603258133,
      "learning_rate": 2.51173060999172e-05,
      "loss": 0.5048,
      "mean_token_accuracy": 0.8405421487987041,
      "num_tokens": 779316076.0,
      "step": 24418
    },
    {
      "entropy": 0.5582011677324772,
      "epoch": 2.2466690818339172,
      "grad_norm": 0.1673109233379364,
      "learning_rate": 2.511423927377557e-05,
      "loss": 0.5576,
      "mean_token_accuracy": 0.8199644796550274,
      "num_tokens": 779347197.0,
      "step": 24419
    },
    {
      "entropy": 0.4149530476424843,
      "epoch": 2.2467610882053584,
      "grad_norm": 0.15451844036579132,
      "learning_rate": 2.5111172447633946e-05,
      "loss": 0.401,
      "mean_token_accuracy": 0.8731045313179493,
      "num_tokens": 779378756.0,
      "step": 24420
    },
    {
      "entropy": 0.6113915294408798,
      "epoch": 2.2468530945767995,
      "grad_norm": 0.1740831881761551,
      "learning_rate": 2.5108105621492317e-05,
      "loss": 0.5984,
      "mean_token_accuracy": 0.8123679831624031,
      "num_tokens": 779410420.0,
      "step": 24421
    },
    {
      "entropy": 0.40971843036822975,
      "epoch": 2.2469451009482406,
      "grad_norm": 0.14400969445705414,
      "learning_rate": 2.5105038795350695e-05,
      "loss": 0.3882,
      "mean_token_accuracy": 0.8764366805553436,
      "num_tokens": 779442639.0,
      "step": 24422
    },
    {
      "entropy": 0.6040834290906787,
      "epoch": 2.2470371073196818,
      "grad_norm": 0.16814790666103363,
      "learning_rate": 2.5101971969209066e-05,
      "loss": 0.5946,
      "mean_token_accuracy": 0.815187469124794,
      "num_tokens": 779474526.0,
      "step": 24423
    },
    {
      "entropy": 0.48675424978137016,
      "epoch": 2.2471291136911233,
      "grad_norm": 0.1575809270143509,
      "learning_rate": 2.509890514306744e-05,
      "loss": 0.482,
      "mean_token_accuracy": 0.8450056612491608,
      "num_tokens": 779506149.0,
      "step": 24424
    },
    {
      "entropy": 0.5450378656387329,
      "epoch": 2.2472211200625645,
      "grad_norm": 0.16566257178783417,
      "learning_rate": 2.5095838316925812e-05,
      "loss": 0.5299,
      "mean_token_accuracy": 0.8349120728671551,
      "num_tokens": 779537564.0,
      "step": 24425
    },
    {
      "entropy": 0.4686151137575507,
      "epoch": 2.2473131264340056,
      "grad_norm": 0.1534123569726944,
      "learning_rate": 2.509277149078419e-05,
      "loss": 0.4454,
      "mean_token_accuracy": 0.8554322235286236,
      "num_tokens": 779569033.0,
      "step": 24426
    },
    {
      "entropy": 0.6027639470994473,
      "epoch": 2.2474051328054467,
      "grad_norm": 0.17055819928646088,
      "learning_rate": 2.508970466464256e-05,
      "loss": 0.5915,
      "mean_token_accuracy": 0.8152408674359322,
      "num_tokens": 779600843.0,
      "step": 24427
    },
    {
      "entropy": 0.4636322585865855,
      "epoch": 2.247497139176888,
      "grad_norm": 0.15232105553150177,
      "learning_rate": 2.5086637838500936e-05,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8564231395721436,
      "num_tokens": 779633129.0,
      "step": 24428
    },
    {
      "entropy": 0.48747340543195605,
      "epoch": 2.2475891455483294,
      "grad_norm": 0.15097759664058685,
      "learning_rate": 2.5083571012359307e-05,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8508109115064144,
      "num_tokens": 779665759.0,
      "step": 24429
    },
    {
      "entropy": 0.5263208476826549,
      "epoch": 2.2476811519197706,
      "grad_norm": 0.15841029584407806,
      "learning_rate": 2.5080504186217685e-05,
      "loss": 0.516,
      "mean_token_accuracy": 0.8361768573522568,
      "num_tokens": 779698372.0,
      "step": 24430
    },
    {
      "entropy": 0.4946761140599847,
      "epoch": 2.2477731582912117,
      "grad_norm": 0.16540172696113586,
      "learning_rate": 2.5077437360076057e-05,
      "loss": 0.482,
      "mean_token_accuracy": 0.8508294709026814,
      "num_tokens": 779729357.0,
      "step": 24431
    },
    {
      "entropy": 0.4124473789706826,
      "epoch": 2.247865164662653,
      "grad_norm": 0.14873120188713074,
      "learning_rate": 2.507437053393443e-05,
      "loss": 0.4015,
      "mean_token_accuracy": 0.8680865317583084,
      "num_tokens": 779761726.0,
      "step": 24432
    },
    {
      "entropy": 0.5100370291620493,
      "epoch": 2.247957171034094,
      "grad_norm": 0.16215935349464417,
      "learning_rate": 2.5071303707792803e-05,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8409861102700233,
      "num_tokens": 779792995.0,
      "step": 24433
    },
    {
      "entropy": 0.4744536383077502,
      "epoch": 2.2480491774055356,
      "grad_norm": 0.15314799547195435,
      "learning_rate": 2.506823688165118e-05,
      "loss": 0.454,
      "mean_token_accuracy": 0.8538881242275238,
      "num_tokens": 779824198.0,
      "step": 24434
    },
    {
      "entropy": 0.5716152116656303,
      "epoch": 2.2481411837769767,
      "grad_norm": 0.16827519237995148,
      "learning_rate": 2.506517005550956e-05,
      "loss": 0.5665,
      "mean_token_accuracy": 0.820639468729496,
      "num_tokens": 779856746.0,
      "step": 24435
    },
    {
      "entropy": 0.5480635203421116,
      "epoch": 2.248233190148418,
      "grad_norm": 0.16142991185188293,
      "learning_rate": 2.5062103229367926e-05,
      "loss": 0.5262,
      "mean_token_accuracy": 0.834025152027607,
      "num_tokens": 779888913.0,
      "step": 24436
    },
    {
      "entropy": 0.44954340253025293,
      "epoch": 2.248325196519859,
      "grad_norm": 0.15123535692691803,
      "learning_rate": 2.5059036403226304e-05,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8589532449841499,
      "num_tokens": 779921335.0,
      "step": 24437
    },
    {
      "entropy": 0.4658341915346682,
      "epoch": 2.2484172028913,
      "grad_norm": 0.15378868579864502,
      "learning_rate": 2.5055969577084676e-05,
      "loss": 0.4444,
      "mean_token_accuracy": 0.8579787760972977,
      "num_tokens": 779953459.0,
      "step": 24438
    },
    {
      "entropy": 0.4356749430298805,
      "epoch": 2.2485092092627417,
      "grad_norm": 0.14505207538604736,
      "learning_rate": 2.5052902750943054e-05,
      "loss": 0.4054,
      "mean_token_accuracy": 0.8651524260640144,
      "num_tokens": 779985021.0,
      "step": 24439
    },
    {
      "entropy": 0.5238208631053567,
      "epoch": 2.248601215634183,
      "grad_norm": 0.16008180379867554,
      "learning_rate": 2.5049835924801425e-05,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8424661196768284,
      "num_tokens": 780017200.0,
      "step": 24440
    },
    {
      "entropy": 0.47878003399819136,
      "epoch": 2.248693222005624,
      "grad_norm": 0.15838241577148438,
      "learning_rate": 2.50467690986598e-05,
      "loss": 0.4705,
      "mean_token_accuracy": 0.8517659865319729,
      "num_tokens": 780049261.0,
      "step": 24441
    },
    {
      "entropy": 0.4969166601076722,
      "epoch": 2.248785228377065,
      "grad_norm": 0.158899188041687,
      "learning_rate": 2.504370227251817e-05,
      "loss": 0.4857,
      "mean_token_accuracy": 0.847134705632925,
      "num_tokens": 780081051.0,
      "step": 24442
    },
    {
      "entropy": 0.5447862390428782,
      "epoch": 2.248877234748506,
      "grad_norm": 0.1681572049856186,
      "learning_rate": 2.504063544637655e-05,
      "loss": 0.5358,
      "mean_token_accuracy": 0.82859867811203,
      "num_tokens": 780112425.0,
      "step": 24443
    },
    {
      "entropy": 0.43117644102312624,
      "epoch": 2.248969241119948,
      "grad_norm": 0.14419656991958618,
      "learning_rate": 2.503756862023492e-05,
      "loss": 0.4216,
      "mean_token_accuracy": 0.862595684826374,
      "num_tokens": 780144590.0,
      "step": 24444
    },
    {
      "entropy": 0.6444276459515095,
      "epoch": 2.249061247491389,
      "grad_norm": 0.1797318458557129,
      "learning_rate": 2.5034501794093295e-05,
      "loss": 0.639,
      "mean_token_accuracy": 0.7975261099636555,
      "num_tokens": 780176487.0,
      "step": 24445
    },
    {
      "entropy": 0.47039001155644655,
      "epoch": 2.24915325386283,
      "grad_norm": 0.15178389847278595,
      "learning_rate": 2.5031434967951666e-05,
      "loss": 0.4669,
      "mean_token_accuracy": 0.8556609228253365,
      "num_tokens": 780208494.0,
      "step": 24446
    },
    {
      "entropy": 0.504041506908834,
      "epoch": 2.249245260234271,
      "grad_norm": 0.1651492416858673,
      "learning_rate": 2.5028368141810044e-05,
      "loss": 0.501,
      "mean_token_accuracy": 0.8373551145195961,
      "num_tokens": 780239818.0,
      "step": 24447
    },
    {
      "entropy": 0.5108443954959512,
      "epoch": 2.2493372666057123,
      "grad_norm": 0.16348709166049957,
      "learning_rate": 2.5025301315668415e-05,
      "loss": 0.4974,
      "mean_token_accuracy": 0.8399273790419102,
      "num_tokens": 780271317.0,
      "step": 24448
    },
    {
      "entropy": 0.5257391878403723,
      "epoch": 2.249429272977154,
      "grad_norm": 0.16166585683822632,
      "learning_rate": 2.502223448952679e-05,
      "loss": 0.5168,
      "mean_token_accuracy": 0.8392247445881367,
      "num_tokens": 780302684.0,
      "step": 24449
    },
    {
      "entropy": 0.5033611818216741,
      "epoch": 2.249521279348595,
      "grad_norm": 0.1613195836544037,
      "learning_rate": 2.501916766338516e-05,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8473200723528862,
      "num_tokens": 780334931.0,
      "step": 24450
    },
    {
      "entropy": 0.40127606107853353,
      "epoch": 2.249613285720036,
      "grad_norm": 0.15066500008106232,
      "learning_rate": 2.501610083724354e-05,
      "loss": 0.3885,
      "mean_token_accuracy": 0.8750907555222511,
      "num_tokens": 780367212.0,
      "step": 24451
    },
    {
      "entropy": 0.48405956383794546,
      "epoch": 2.2497052920914773,
      "grad_norm": 0.1550741195678711,
      "learning_rate": 2.501303401110191e-05,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8488504700362682,
      "num_tokens": 780399513.0,
      "step": 24452
    },
    {
      "entropy": 0.5044801523908973,
      "epoch": 2.2497972984629184,
      "grad_norm": 0.15507839620113373,
      "learning_rate": 2.5009967184960285e-05,
      "loss": 0.488,
      "mean_token_accuracy": 0.842807050794363,
      "num_tokens": 780431327.0,
      "step": 24453
    },
    {
      "entropy": 0.5630951095372438,
      "epoch": 2.24988930483436,
      "grad_norm": 0.16077059507369995,
      "learning_rate": 2.5006900358818657e-05,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8298156447708607,
      "num_tokens": 780463246.0,
      "step": 24454
    },
    {
      "entropy": 0.4360450751846656,
      "epoch": 2.249981311205801,
      "grad_norm": 0.1492619663476944,
      "learning_rate": 2.5003833532677035e-05,
      "loss": 0.423,
      "mean_token_accuracy": 0.8646026328206062,
      "num_tokens": 780495539.0,
      "step": 24455
    },
    {
      "entropy": 0.545049782609567,
      "epoch": 2.2500733175772423,
      "grad_norm": 0.16018982231616974,
      "learning_rate": 2.5000766706535406e-05,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8372747153043747,
      "num_tokens": 780527880.0,
      "step": 24456
    },
    {
      "entropy": 0.6002118424512446,
      "epoch": 2.2501653239486834,
      "grad_norm": 0.1648840755224228,
      "learning_rate": 2.4997699880393784e-05,
      "loss": 0.5872,
      "mean_token_accuracy": 0.8171718716621399,
      "num_tokens": 780560581.0,
      "step": 24457
    },
    {
      "entropy": 0.45802199421450496,
      "epoch": 2.2502573303201245,
      "grad_norm": 0.14949658513069153,
      "learning_rate": 2.4994633054252155e-05,
      "loss": 0.4363,
      "mean_token_accuracy": 0.858353715389967,
      "num_tokens": 780592693.0,
      "step": 24458
    },
    {
      "entropy": 0.38501535763498396,
      "epoch": 2.250349336691566,
      "grad_norm": 0.14349797368049622,
      "learning_rate": 2.499156622811053e-05,
      "loss": 0.3733,
      "mean_token_accuracy": 0.8800480663776398,
      "num_tokens": 780624905.0,
      "step": 24459
    },
    {
      "entropy": 0.5534564238041639,
      "epoch": 2.2504413430630072,
      "grad_norm": 0.16110654175281525,
      "learning_rate": 2.4988499401968904e-05,
      "loss": 0.5576,
      "mean_token_accuracy": 0.8260485045611858,
      "num_tokens": 780657508.0,
      "step": 24460
    },
    {
      "entropy": 0.5246065864339471,
      "epoch": 2.2505333494344484,
      "grad_norm": 0.1625121533870697,
      "learning_rate": 2.498543257582728e-05,
      "loss": 0.5082,
      "mean_token_accuracy": 0.8404434099793434,
      "num_tokens": 780689331.0,
      "step": 24461
    },
    {
      "entropy": 0.5219376608729362,
      "epoch": 2.2506253558058895,
      "grad_norm": 0.16134411096572876,
      "learning_rate": 2.498236574968565e-05,
      "loss": 0.5073,
      "mean_token_accuracy": 0.8390005268156528,
      "num_tokens": 780720606.0,
      "step": 24462
    },
    {
      "entropy": 0.42246348038315773,
      "epoch": 2.2507173621773306,
      "grad_norm": 0.1560068130493164,
      "learning_rate": 2.4979298923544025e-05,
      "loss": 0.4057,
      "mean_token_accuracy": 0.8687624000012875,
      "num_tokens": 780752874.0,
      "step": 24463
    },
    {
      "entropy": 0.6125754397362471,
      "epoch": 2.2508093685487722,
      "grad_norm": 0.17022131383419037,
      "learning_rate": 2.49762320974024e-05,
      "loss": 0.5954,
      "mean_token_accuracy": 0.8155272156000137,
      "num_tokens": 780784395.0,
      "step": 24464
    },
    {
      "entropy": 0.5393954768078402,
      "epoch": 2.2509013749202134,
      "grad_norm": 0.15903140604496002,
      "learning_rate": 2.4973165271260774e-05,
      "loss": 0.517,
      "mean_token_accuracy": 0.8344217278063297,
      "num_tokens": 780816642.0,
      "step": 24465
    },
    {
      "entropy": 0.5208977954462171,
      "epoch": 2.2509933812916545,
      "grad_norm": 0.16028986871242523,
      "learning_rate": 2.4970098445119146e-05,
      "loss": 0.5075,
      "mean_token_accuracy": 0.8360008150339127,
      "num_tokens": 780848738.0,
      "step": 24466
    },
    {
      "entropy": 0.5296133942902088,
      "epoch": 2.2510853876630956,
      "grad_norm": 0.16917018592357635,
      "learning_rate": 2.496703161897752e-05,
      "loss": 0.5228,
      "mean_token_accuracy": 0.842368271201849,
      "num_tokens": 780880888.0,
      "step": 24467
    },
    {
      "entropy": 0.5753062106668949,
      "epoch": 2.2511773940345368,
      "grad_norm": 0.17010529339313507,
      "learning_rate": 2.4963964792835895e-05,
      "loss": 0.5652,
      "mean_token_accuracy": 0.820952046662569,
      "num_tokens": 780911617.0,
      "step": 24468
    },
    {
      "entropy": 0.41842825151979923,
      "epoch": 2.2512694004059783,
      "grad_norm": 0.15192735195159912,
      "learning_rate": 2.496089796669427e-05,
      "loss": 0.398,
      "mean_token_accuracy": 0.8739656396210194,
      "num_tokens": 780943322.0,
      "step": 24469
    },
    {
      "entropy": 0.4780424479395151,
      "epoch": 2.2513614067774195,
      "grad_norm": 0.1526806801557541,
      "learning_rate": 2.4957831140552644e-05,
      "loss": 0.4698,
      "mean_token_accuracy": 0.85159981995821,
      "num_tokens": 780975474.0,
      "step": 24470
    },
    {
      "entropy": 0.5382215604186058,
      "epoch": 2.2514534131488606,
      "grad_norm": 0.16390009224414825,
      "learning_rate": 2.4954764314411015e-05,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8378297947347164,
      "num_tokens": 781007568.0,
      "step": 24471
    },
    {
      "entropy": 0.5613689785823226,
      "epoch": 2.2515454195203017,
      "grad_norm": 0.1624917834997177,
      "learning_rate": 2.495169748826939e-05,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8307937569916248,
      "num_tokens": 781039532.0,
      "step": 24472
    },
    {
      "entropy": 0.5461323410272598,
      "epoch": 2.251637425891743,
      "grad_norm": 0.16408802568912506,
      "learning_rate": 2.4948630662127765e-05,
      "loss": 0.5276,
      "mean_token_accuracy": 0.8349205777049065,
      "num_tokens": 781071415.0,
      "step": 24473
    },
    {
      "entropy": 0.44381635589525104,
      "epoch": 2.2517294322631844,
      "grad_norm": 0.15590700507164001,
      "learning_rate": 2.494556383598614e-05,
      "loss": 0.4233,
      "mean_token_accuracy": 0.8623992577195168,
      "num_tokens": 781102286.0,
      "step": 24474
    },
    {
      "entropy": 0.5766895832493901,
      "epoch": 2.2518214386346256,
      "grad_norm": 0.1660001426935196,
      "learning_rate": 2.494249700984451e-05,
      "loss": 0.5736,
      "mean_token_accuracy": 0.8201344534754753,
      "num_tokens": 781134467.0,
      "step": 24475
    },
    {
      "entropy": 0.5389787200838327,
      "epoch": 2.2519134450060667,
      "grad_norm": 0.1606208086013794,
      "learning_rate": 2.4939430183702885e-05,
      "loss": 0.5277,
      "mean_token_accuracy": 0.8312935754656792,
      "num_tokens": 781166486.0,
      "step": 24476
    },
    {
      "entropy": 0.5789657924324274,
      "epoch": 2.252005451377508,
      "grad_norm": 0.17104192078113556,
      "learning_rate": 2.493636335756126e-05,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8226261362433434,
      "num_tokens": 781197764.0,
      "step": 24477
    },
    {
      "entropy": 0.5201153568923473,
      "epoch": 2.252097457748949,
      "grad_norm": 0.16283902525901794,
      "learning_rate": 2.4933296531419635e-05,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8411730080842972,
      "num_tokens": 781229997.0,
      "step": 24478
    },
    {
      "entropy": 0.5015730762388557,
      "epoch": 2.2521894641203906,
      "grad_norm": 0.16043440997600555,
      "learning_rate": 2.493022970527801e-05,
      "loss": 0.4904,
      "mean_token_accuracy": 0.845820352435112,
      "num_tokens": 781261867.0,
      "step": 24479
    },
    {
      "entropy": 0.5474381912499666,
      "epoch": 2.2522814704918317,
      "grad_norm": 0.16937081515789032,
      "learning_rate": 2.4927162879136384e-05,
      "loss": 0.5356,
      "mean_token_accuracy": 0.8312846198678017,
      "num_tokens": 781292713.0,
      "step": 24480
    },
    {
      "entropy": 0.5576001303270459,
      "epoch": 2.252373476863273,
      "grad_norm": 0.16416025161743164,
      "learning_rate": 2.492409605299476e-05,
      "loss": 0.5452,
      "mean_token_accuracy": 0.8265961073338985,
      "num_tokens": 781324828.0,
      "step": 24481
    },
    {
      "entropy": 0.6574952807277441,
      "epoch": 2.252465483234714,
      "grad_norm": 0.17643921077251434,
      "learning_rate": 2.4921029226853133e-05,
      "loss": 0.6465,
      "mean_token_accuracy": 0.7983353361487389,
      "num_tokens": 781357024.0,
      "step": 24482
    },
    {
      "entropy": 0.5279461070895195,
      "epoch": 2.252557489606155,
      "grad_norm": 0.15999841690063477,
      "learning_rate": 2.4917962400711504e-05,
      "loss": 0.5108,
      "mean_token_accuracy": 0.8401323445141315,
      "num_tokens": 781388436.0,
      "step": 24483
    },
    {
      "entropy": 0.471074684523046,
      "epoch": 2.2526494959775967,
      "grad_norm": 0.15115118026733398,
      "learning_rate": 2.491489557456988e-05,
      "loss": 0.4567,
      "mean_token_accuracy": 0.8565506935119629,
      "num_tokens": 781420330.0,
      "step": 24484
    },
    {
      "entropy": 0.4694964112713933,
      "epoch": 2.252741502349038,
      "grad_norm": 0.15591749548912048,
      "learning_rate": 2.4911828748428254e-05,
      "loss": 0.4634,
      "mean_token_accuracy": 0.8541263528168201,
      "num_tokens": 781452751.0,
      "step": 24485
    },
    {
      "entropy": 0.5175333917140961,
      "epoch": 2.252833508720479,
      "grad_norm": 0.16194859147071838,
      "learning_rate": 2.490876192228663e-05,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8403844945132732,
      "num_tokens": 781484653.0,
      "step": 24486
    },
    {
      "entropy": 0.46133989025838673,
      "epoch": 2.25292551509192,
      "grad_norm": 0.1502697616815567,
      "learning_rate": 2.4905695096145003e-05,
      "loss": 0.4299,
      "mean_token_accuracy": 0.8624884113669395,
      "num_tokens": 781516195.0,
      "step": 24487
    },
    {
      "entropy": 0.6106804134324193,
      "epoch": 2.253017521463361,
      "grad_norm": 0.17013897001743317,
      "learning_rate": 2.4902628270003374e-05,
      "loss": 0.5927,
      "mean_token_accuracy": 0.8112868815660477,
      "num_tokens": 781548124.0,
      "step": 24488
    },
    {
      "entropy": 0.46159154525958,
      "epoch": 2.253109527834803,
      "grad_norm": 0.16044831275939941,
      "learning_rate": 2.489956144386175e-05,
      "loss": 0.4457,
      "mean_token_accuracy": 0.8588029332458973,
      "num_tokens": 781580351.0,
      "step": 24489
    },
    {
      "entropy": 0.520889014005661,
      "epoch": 2.253201534206244,
      "grad_norm": 0.1641312688589096,
      "learning_rate": 2.4896494617720124e-05,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8346798755228519,
      "num_tokens": 781612830.0,
      "step": 24490
    },
    {
      "entropy": 0.6473383642733097,
      "epoch": 2.253293540577685,
      "grad_norm": 0.17889389395713806,
      "learning_rate": 2.4893427791578498e-05,
      "loss": 0.6328,
      "mean_token_accuracy": 0.8036094158887863,
      "num_tokens": 781644872.0,
      "step": 24491
    },
    {
      "entropy": 0.549550611525774,
      "epoch": 2.253385546949126,
      "grad_norm": 0.1616404950618744,
      "learning_rate": 2.489036096543687e-05,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8318069986999035,
      "num_tokens": 781676547.0,
      "step": 24492
    },
    {
      "entropy": 0.45412088790908456,
      "epoch": 2.2534775533205673,
      "grad_norm": 0.1519334614276886,
      "learning_rate": 2.4887294139295244e-05,
      "loss": 0.4412,
      "mean_token_accuracy": 0.8587366156280041,
      "num_tokens": 781708628.0,
      "step": 24493
    },
    {
      "entropy": 0.4901073444634676,
      "epoch": 2.253569559692009,
      "grad_norm": 0.16024750471115112,
      "learning_rate": 2.488422731315362e-05,
      "loss": 0.4748,
      "mean_token_accuracy": 0.8471357449889183,
      "num_tokens": 781740297.0,
      "step": 24494
    },
    {
      "entropy": 0.5531571842730045,
      "epoch": 2.25366156606345,
      "grad_norm": 0.1645442545413971,
      "learning_rate": 2.4881160487011993e-05,
      "loss": 0.5372,
      "mean_token_accuracy": 0.8301925919950008,
      "num_tokens": 781771753.0,
      "step": 24495
    },
    {
      "entropy": 0.5918507222086191,
      "epoch": 2.253753572434891,
      "grad_norm": 0.17048805952072144,
      "learning_rate": 2.4878093660870365e-05,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8213001601397991,
      "num_tokens": 781802742.0,
      "step": 24496
    },
    {
      "entropy": 0.501718427054584,
      "epoch": 2.2538455788063323,
      "grad_norm": 0.15854963660240173,
      "learning_rate": 2.487502683472874e-05,
      "loss": 0.4892,
      "mean_token_accuracy": 0.8466691970825195,
      "num_tokens": 781834588.0,
      "step": 24497
    },
    {
      "entropy": 0.4986535422503948,
      "epoch": 2.2539375851777734,
      "grad_norm": 0.1664598435163498,
      "learning_rate": 2.4871960008587114e-05,
      "loss": 0.4883,
      "mean_token_accuracy": 0.8417487591505051,
      "num_tokens": 781866257.0,
      "step": 24498
    },
    {
      "entropy": 0.584959477186203,
      "epoch": 2.254029591549215,
      "grad_norm": 0.16752810776233673,
      "learning_rate": 2.486889318244549e-05,
      "loss": 0.5696,
      "mean_token_accuracy": 0.8182439506053925,
      "num_tokens": 781897811.0,
      "step": 24499
    },
    {
      "entropy": 0.475332111120224,
      "epoch": 2.254121597920656,
      "grad_norm": 0.15829338133335114,
      "learning_rate": 2.4865826356303863e-05,
      "loss": 0.4549,
      "mean_token_accuracy": 0.854022353887558,
      "num_tokens": 781929011.0,
      "step": 24500
    },
    {
      "entropy": 0.48116996232420206,
      "epoch": 2.2542136042920973,
      "grad_norm": 0.16026760637760162,
      "learning_rate": 2.4862759530162234e-05,
      "loss": 0.469,
      "mean_token_accuracy": 0.8504132404923439,
      "num_tokens": 781960808.0,
      "step": 24501
    },
    {
      "entropy": 0.5311038568615913,
      "epoch": 2.2543056106635384,
      "grad_norm": 0.1617385596036911,
      "learning_rate": 2.485969270402061e-05,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8295583054423332,
      "num_tokens": 781993376.0,
      "step": 24502
    },
    {
      "entropy": 0.5148683488368988,
      "epoch": 2.2543976170349795,
      "grad_norm": 0.15333020687103271,
      "learning_rate": 2.4856625877878984e-05,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8399798423051834,
      "num_tokens": 782025895.0,
      "step": 24503
    },
    {
      "entropy": 0.47817657608538866,
      "epoch": 2.254489623406421,
      "grad_norm": 0.15285184979438782,
      "learning_rate": 2.485355905173736e-05,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8536534532904625,
      "num_tokens": 782058360.0,
      "step": 24504
    },
    {
      "entropy": 0.5585655486211181,
      "epoch": 2.2545816297778623,
      "grad_norm": 0.170372873544693,
      "learning_rate": 2.485049222559573e-05,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8293003253638744,
      "num_tokens": 782089844.0,
      "step": 24505
    },
    {
      "entropy": 0.6018581814132631,
      "epoch": 2.2546736361493034,
      "grad_norm": 0.17004404962062836,
      "learning_rate": 2.4847425399454104e-05,
      "loss": 0.5874,
      "mean_token_accuracy": 0.8147604614496231,
      "num_tokens": 782122152.0,
      "step": 24506
    },
    {
      "entropy": 0.6774228140711784,
      "epoch": 2.2547656425207445,
      "grad_norm": 0.1819961667060852,
      "learning_rate": 2.484435857331248e-05,
      "loss": 0.6551,
      "mean_token_accuracy": 0.7967884168028831,
      "num_tokens": 782153598.0,
      "step": 24507
    },
    {
      "entropy": 0.4220258230343461,
      "epoch": 2.2548576488921857,
      "grad_norm": 0.15510942041873932,
      "learning_rate": 2.4841291747170854e-05,
      "loss": 0.4141,
      "mean_token_accuracy": 0.8674763739109039,
      "num_tokens": 782185005.0,
      "step": 24508
    },
    {
      "entropy": 0.4285238836891949,
      "epoch": 2.2549496552636272,
      "grad_norm": 0.155912846326828,
      "learning_rate": 2.4838224921029228e-05,
      "loss": 0.4116,
      "mean_token_accuracy": 0.8656356893479824,
      "num_tokens": 782216040.0,
      "step": 24509
    },
    {
      "entropy": 0.5349156651645899,
      "epoch": 2.2550416616350684,
      "grad_norm": 0.16345177590847015,
      "learning_rate": 2.4835158094887603e-05,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8312010541558266,
      "num_tokens": 782247007.0,
      "step": 24510
    },
    {
      "entropy": 0.5115064606070518,
      "epoch": 2.2551336680065095,
      "grad_norm": 0.15934626758098602,
      "learning_rate": 2.4832091268745978e-05,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8407339528203011,
      "num_tokens": 782279710.0,
      "step": 24511
    },
    {
      "entropy": 0.4831138700246811,
      "epoch": 2.2552256743779506,
      "grad_norm": 0.16171862185001373,
      "learning_rate": 2.4829024442604352e-05,
      "loss": 0.476,
      "mean_token_accuracy": 0.8500587604939938,
      "num_tokens": 782311075.0,
      "step": 24512
    },
    {
      "entropy": 0.49334053695201874,
      "epoch": 2.2553176807493918,
      "grad_norm": 0.16397272050380707,
      "learning_rate": 2.4825957616462723e-05,
      "loss": 0.4763,
      "mean_token_accuracy": 0.8462398946285248,
      "num_tokens": 782342697.0,
      "step": 24513
    },
    {
      "entropy": 0.5089993346482515,
      "epoch": 2.2554096871208333,
      "grad_norm": 0.15748484432697296,
      "learning_rate": 2.4822890790321098e-05,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8436431288719177,
      "num_tokens": 782375465.0,
      "step": 24514
    },
    {
      "entropy": 0.6928789634257555,
      "epoch": 2.2555016934922745,
      "grad_norm": 0.17472220957279205,
      "learning_rate": 2.4819823964179473e-05,
      "loss": 0.6791,
      "mean_token_accuracy": 0.7882322818040848,
      "num_tokens": 782407447.0,
      "step": 24515
    },
    {
      "entropy": 0.5123721659183502,
      "epoch": 2.2555936998637156,
      "grad_norm": 0.1610775589942932,
      "learning_rate": 2.4816757138037847e-05,
      "loss": 0.4915,
      "mean_token_accuracy": 0.8403894454240799,
      "num_tokens": 782439496.0,
      "step": 24516
    },
    {
      "entropy": 0.5833586417138577,
      "epoch": 2.2556857062351567,
      "grad_norm": 0.1626797765493393,
      "learning_rate": 2.4813690311896222e-05,
      "loss": 0.5705,
      "mean_token_accuracy": 0.8220996707677841,
      "num_tokens": 782471834.0,
      "step": 24517
    },
    {
      "entropy": 0.43609125819057226,
      "epoch": 2.255777712606598,
      "grad_norm": 0.15197442471981049,
      "learning_rate": 2.4810623485754593e-05,
      "loss": 0.4108,
      "mean_token_accuracy": 0.8651783876121044,
      "num_tokens": 782503200.0,
      "step": 24518
    },
    {
      "entropy": 0.5956856124103069,
      "epoch": 2.2558697189780395,
      "grad_norm": 0.16496317088603973,
      "learning_rate": 2.4807556659612968e-05,
      "loss": 0.5618,
      "mean_token_accuracy": 0.8219750337302685,
      "num_tokens": 782534745.0,
      "step": 24519
    },
    {
      "entropy": 0.5714917960576713,
      "epoch": 2.2559617253494806,
      "grad_norm": 0.16621796786785126,
      "learning_rate": 2.4804489833471343e-05,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8245766907930374,
      "num_tokens": 782567335.0,
      "step": 24520
    },
    {
      "entropy": 0.540440620854497,
      "epoch": 2.2560537317209217,
      "grad_norm": 0.17023427784442902,
      "learning_rate": 2.4801423007329717e-05,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8282462023198605,
      "num_tokens": 782598384.0,
      "step": 24521
    },
    {
      "entropy": 0.5564644341357052,
      "epoch": 2.256145738092363,
      "grad_norm": 0.16531458497047424,
      "learning_rate": 2.479835618118809e-05,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8233073800802231,
      "num_tokens": 782630802.0,
      "step": 24522
    },
    {
      "entropy": 0.48090568371117115,
      "epoch": 2.256237744463804,
      "grad_norm": 0.1566942036151886,
      "learning_rate": 2.4795289355046463e-05,
      "loss": 0.4593,
      "mean_token_accuracy": 0.8497747778892517,
      "num_tokens": 782662065.0,
      "step": 24523
    },
    {
      "entropy": 0.51016257237643,
      "epoch": 2.2563297508352456,
      "grad_norm": 0.1648634374141693,
      "learning_rate": 2.4792222528904838e-05,
      "loss": 0.4954,
      "mean_token_accuracy": 0.8422924727201462,
      "num_tokens": 782694070.0,
      "step": 24524
    },
    {
      "entropy": 0.509133679792285,
      "epoch": 2.2564217572066867,
      "grad_norm": 0.17088322341442108,
      "learning_rate": 2.4789155702763212e-05,
      "loss": 0.5055,
      "mean_token_accuracy": 0.8401783108711243,
      "num_tokens": 782725425.0,
      "step": 24525
    },
    {
      "entropy": 0.5174186616204679,
      "epoch": 2.256513763578128,
      "grad_norm": 0.1636975109577179,
      "learning_rate": 2.4786088876621584e-05,
      "loss": 0.4949,
      "mean_token_accuracy": 0.8454762101173401,
      "num_tokens": 782757687.0,
      "step": 24526
    },
    {
      "entropy": 0.5669670556671917,
      "epoch": 2.256605769949569,
      "grad_norm": 0.16512200236320496,
      "learning_rate": 2.478302205047996e-05,
      "loss": 0.5519,
      "mean_token_accuracy": 0.8250644579529762,
      "num_tokens": 782790071.0,
      "step": 24527
    },
    {
      "entropy": 0.5113000916317105,
      "epoch": 2.25669777632101,
      "grad_norm": 0.1612825095653534,
      "learning_rate": 2.4779955224338333e-05,
      "loss": 0.5029,
      "mean_token_accuracy": 0.8374633640050888,
      "num_tokens": 782822241.0,
      "step": 24528
    },
    {
      "entropy": 0.46844421699643135,
      "epoch": 2.2567897826924517,
      "grad_norm": 0.1479194313287735,
      "learning_rate": 2.4776888398196708e-05,
      "loss": 0.4562,
      "mean_token_accuracy": 0.8551558628678322,
      "num_tokens": 782854623.0,
      "step": 24529
    },
    {
      "entropy": 0.6047173850238323,
      "epoch": 2.256881789063893,
      "grad_norm": 0.17354673147201538,
      "learning_rate": 2.4773821572055082e-05,
      "loss": 0.6004,
      "mean_token_accuracy": 0.8110603019595146,
      "num_tokens": 782886561.0,
      "step": 24530
    },
    {
      "entropy": 0.4990354971960187,
      "epoch": 2.256973795435334,
      "grad_norm": 0.15737511217594147,
      "learning_rate": 2.4770754745913454e-05,
      "loss": 0.4819,
      "mean_token_accuracy": 0.8438973352313042,
      "num_tokens": 782918129.0,
      "step": 24531
    },
    {
      "entropy": 0.5194796742871404,
      "epoch": 2.257065801806775,
      "grad_norm": 0.15906347334384918,
      "learning_rate": 2.4767687919771828e-05,
      "loss": 0.5052,
      "mean_token_accuracy": 0.8376389518380165,
      "num_tokens": 782950225.0,
      "step": 24532
    },
    {
      "entropy": 0.6211152318865061,
      "epoch": 2.257157808178216,
      "grad_norm": 0.17936137318611145,
      "learning_rate": 2.4764621093630203e-05,
      "loss": 0.6169,
      "mean_token_accuracy": 0.8096489198505878,
      "num_tokens": 782980816.0,
      "step": 24533
    },
    {
      "entropy": 0.5435153124853969,
      "epoch": 2.257249814549658,
      "grad_norm": 0.17536482214927673,
      "learning_rate": 2.4761554267488578e-05,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8356560617685318,
      "num_tokens": 783012137.0,
      "step": 24534
    },
    {
      "entropy": 0.3533323109149933,
      "epoch": 2.257341820921099,
      "grad_norm": 0.1431332528591156,
      "learning_rate": 2.475848744134695e-05,
      "loss": 0.3393,
      "mean_token_accuracy": 0.891213234513998,
      "num_tokens": 783044064.0,
      "step": 24535
    },
    {
      "entropy": 0.5871488004922867,
      "epoch": 2.25743382729254,
      "grad_norm": 0.1686410903930664,
      "learning_rate": 2.4755420615205323e-05,
      "loss": 0.5875,
      "mean_token_accuracy": 0.816766157746315,
      "num_tokens": 783076146.0,
      "step": 24536
    },
    {
      "entropy": 0.5035870256833732,
      "epoch": 2.257525833663981,
      "grad_norm": 0.16289837658405304,
      "learning_rate": 2.4752353789063698e-05,
      "loss": 0.4946,
      "mean_token_accuracy": 0.8414501920342445,
      "num_tokens": 783108254.0,
      "step": 24537
    },
    {
      "entropy": 0.5140165844932199,
      "epoch": 2.2576178400354223,
      "grad_norm": 0.15949609875679016,
      "learning_rate": 2.4749286962922073e-05,
      "loss": 0.4915,
      "mean_token_accuracy": 0.8420959450304508,
      "num_tokens": 783140112.0,
      "step": 24538
    },
    {
      "entropy": 0.5072447564452887,
      "epoch": 2.257709846406864,
      "grad_norm": 0.155429407954216,
      "learning_rate": 2.4746220136780447e-05,
      "loss": 0.4941,
      "mean_token_accuracy": 0.8427579551935196,
      "num_tokens": 783172161.0,
      "step": 24539
    },
    {
      "entropy": 0.4986793613061309,
      "epoch": 2.257801852778305,
      "grad_norm": 0.15994855761528015,
      "learning_rate": 2.474315331063882e-05,
      "loss": 0.484,
      "mean_token_accuracy": 0.8471425473690033,
      "num_tokens": 783204560.0,
      "step": 24540
    },
    {
      "entropy": 0.5655517764389515,
      "epoch": 2.257893859149746,
      "grad_norm": 0.16681864857673645,
      "learning_rate": 2.4740086484497197e-05,
      "loss": 0.564,
      "mean_token_accuracy": 0.8241430595517159,
      "num_tokens": 783236323.0,
      "step": 24541
    },
    {
      "entropy": 0.549525011330843,
      "epoch": 2.2579858655211873,
      "grad_norm": 0.16640271246433258,
      "learning_rate": 2.473701965835557e-05,
      "loss": 0.5408,
      "mean_token_accuracy": 0.8285206146538258,
      "num_tokens": 783268474.0,
      "step": 24542
    },
    {
      "entropy": 0.4294705423526466,
      "epoch": 2.2580778718926284,
      "grad_norm": 0.145483136177063,
      "learning_rate": 2.4733952832213943e-05,
      "loss": 0.4172,
      "mean_token_accuracy": 0.8666560985147953,
      "num_tokens": 783301010.0,
      "step": 24543
    },
    {
      "entropy": 0.5375892482697964,
      "epoch": 2.25816987826407,
      "grad_norm": 0.16489608585834503,
      "learning_rate": 2.4730886006072317e-05,
      "loss": 0.5124,
      "mean_token_accuracy": 0.8383024483919144,
      "num_tokens": 783332217.0,
      "step": 24544
    },
    {
      "entropy": 0.517590269446373,
      "epoch": 2.258261884635511,
      "grad_norm": 0.16224044561386108,
      "learning_rate": 2.4727819179930692e-05,
      "loss": 0.4948,
      "mean_token_accuracy": 0.8458204306662083,
      "num_tokens": 783364052.0,
      "step": 24545
    },
    {
      "entropy": 0.5708387503400445,
      "epoch": 2.2583538910069523,
      "grad_norm": 0.1646776795387268,
      "learning_rate": 2.4724752353789067e-05,
      "loss": 0.5673,
      "mean_token_accuracy": 0.8216243125498295,
      "num_tokens": 783396014.0,
      "step": 24546
    },
    {
      "entropy": 0.5862235352396965,
      "epoch": 2.2584458973783934,
      "grad_norm": 0.16728754341602325,
      "learning_rate": 2.472168552764744e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8219691850244999,
      "num_tokens": 783428174.0,
      "step": 24547
    },
    {
      "entropy": 0.47828277200460434,
      "epoch": 2.2585379037498345,
      "grad_norm": 0.15088939666748047,
      "learning_rate": 2.4718618701505812e-05,
      "loss": 0.4569,
      "mean_token_accuracy": 0.851401686668396,
      "num_tokens": 783460694.0,
      "step": 24548
    },
    {
      "entropy": 0.40027841553092003,
      "epoch": 2.258629910121276,
      "grad_norm": 0.1434927135705948,
      "learning_rate": 2.4715551875364187e-05,
      "loss": 0.3884,
      "mean_token_accuracy": 0.8751315958797932,
      "num_tokens": 783493200.0,
      "step": 24549
    },
    {
      "entropy": 0.6109405881725252,
      "epoch": 2.2587219164927173,
      "grad_norm": 0.17207078635692596,
      "learning_rate": 2.4712485049222562e-05,
      "loss": 0.5991,
      "mean_token_accuracy": 0.8106729052960873,
      "num_tokens": 783525220.0,
      "step": 24550
    },
    {
      "entropy": 0.5302469884045422,
      "epoch": 2.2588139228641584,
      "grad_norm": 0.1606551706790924,
      "learning_rate": 2.4709418223080936e-05,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8389846384525299,
      "num_tokens": 783557580.0,
      "step": 24551
    },
    {
      "entropy": 0.4937970908358693,
      "epoch": 2.2589059292355995,
      "grad_norm": 0.1588021069765091,
      "learning_rate": 2.4706351396939308e-05,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8491972386837006,
      "num_tokens": 783589421.0,
      "step": 24552
    },
    {
      "entropy": 0.43799124751240015,
      "epoch": 2.2589979356070407,
      "grad_norm": 0.14956773817539215,
      "learning_rate": 2.4703284570797682e-05,
      "loss": 0.4218,
      "mean_token_accuracy": 0.8641326576471329,
      "num_tokens": 783620996.0,
      "step": 24553
    },
    {
      "entropy": 0.47200826997868717,
      "epoch": 2.2590899419784822,
      "grad_norm": 0.1518024206161499,
      "learning_rate": 2.4700217744656057e-05,
      "loss": 0.4456,
      "mean_token_accuracy": 0.8550765588879585,
      "num_tokens": 783652552.0,
      "step": 24554
    },
    {
      "entropy": 0.535239209420979,
      "epoch": 2.2591819483499234,
      "grad_norm": 0.16255293786525726,
      "learning_rate": 2.469715091851443e-05,
      "loss": 0.526,
      "mean_token_accuracy": 0.8352057933807373,
      "num_tokens": 783684682.0,
      "step": 24555
    },
    {
      "entropy": 0.6082567740231752,
      "epoch": 2.2592739547213645,
      "grad_norm": 0.17004114389419556,
      "learning_rate": 2.4694084092372803e-05,
      "loss": 0.5961,
      "mean_token_accuracy": 0.8115198276937008,
      "num_tokens": 783716548.0,
      "step": 24556
    },
    {
      "entropy": 0.46990737272426486,
      "epoch": 2.2593659610928056,
      "grad_norm": 0.1491926908493042,
      "learning_rate": 2.4691017266231177e-05,
      "loss": 0.4518,
      "mean_token_accuracy": 0.854059312492609,
      "num_tokens": 783748524.0,
      "step": 24557
    },
    {
      "entropy": 0.5267287963069975,
      "epoch": 2.2594579674642468,
      "grad_norm": 0.16830958425998688,
      "learning_rate": 2.4687950440089552e-05,
      "loss": 0.5117,
      "mean_token_accuracy": 0.8384485021233559,
      "num_tokens": 783778545.0,
      "step": 24558
    },
    {
      "entropy": 0.4913132069632411,
      "epoch": 2.2595499738356883,
      "grad_norm": 0.16261932253837585,
      "learning_rate": 2.4684883613947927e-05,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8495109640061855,
      "num_tokens": 783810581.0,
      "step": 24559
    },
    {
      "entropy": 0.6566716469824314,
      "epoch": 2.2596419802071295,
      "grad_norm": 0.17744749784469604,
      "learning_rate": 2.46818167878063e-05,
      "loss": 0.6459,
      "mean_token_accuracy": 0.7960399203002453,
      "num_tokens": 783842929.0,
      "step": 24560
    },
    {
      "entropy": 0.6291407654061913,
      "epoch": 2.2597339865785706,
      "grad_norm": 0.17627926170825958,
      "learning_rate": 2.4678749961664673e-05,
      "loss": 0.6203,
      "mean_token_accuracy": 0.8053965270519257,
      "num_tokens": 783874552.0,
      "step": 24561
    },
    {
      "entropy": 0.5177642283961177,
      "epoch": 2.2598259929500117,
      "grad_norm": 0.1649087816476822,
      "learning_rate": 2.4675683135523047e-05,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8348177187144756,
      "num_tokens": 783907053.0,
      "step": 24562
    },
    {
      "entropy": 0.5501558519899845,
      "epoch": 2.259917999321453,
      "grad_norm": 0.16829080879688263,
      "learning_rate": 2.4672616309381422e-05,
      "loss": 0.5353,
      "mean_token_accuracy": 0.8323058933019638,
      "num_tokens": 783939068.0,
      "step": 24563
    },
    {
      "entropy": 0.6146047962829471,
      "epoch": 2.2600100056928945,
      "grad_norm": 0.17838849127292633,
      "learning_rate": 2.4669549483239797e-05,
      "loss": 0.6069,
      "mean_token_accuracy": 0.810126181691885,
      "num_tokens": 783970776.0,
      "step": 24564
    },
    {
      "entropy": 0.4412050899118185,
      "epoch": 2.2601020120643356,
      "grad_norm": 0.1550249308347702,
      "learning_rate": 2.4666482657098168e-05,
      "loss": 0.4225,
      "mean_token_accuracy": 0.8615644536912441,
      "num_tokens": 784002607.0,
      "step": 24565
    },
    {
      "entropy": 0.5150998014723882,
      "epoch": 2.2601940184357767,
      "grad_norm": 0.161502406001091,
      "learning_rate": 2.4663415830956543e-05,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8421926386654377,
      "num_tokens": 784034268.0,
      "step": 24566
    },
    {
      "entropy": 0.4645986035466194,
      "epoch": 2.260286024807218,
      "grad_norm": 0.15727870166301727,
      "learning_rate": 2.4660349004814917e-05,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8566266112029552,
      "num_tokens": 784066159.0,
      "step": 24567
    },
    {
      "entropy": 0.4918628418818116,
      "epoch": 2.260378031178659,
      "grad_norm": 0.15515749156475067,
      "learning_rate": 2.4657282178673292e-05,
      "loss": 0.467,
      "mean_token_accuracy": 0.8513886220753193,
      "num_tokens": 784098088.0,
      "step": 24568
    },
    {
      "entropy": 0.5078882230445743,
      "epoch": 2.2604700375501006,
      "grad_norm": 0.15695063769817352,
      "learning_rate": 2.4654215352531666e-05,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8414461426436901,
      "num_tokens": 784130273.0,
      "step": 24569
    },
    {
      "entropy": 0.5596571043133736,
      "epoch": 2.2605620439215417,
      "grad_norm": 0.16891182959079742,
      "learning_rate": 2.4651148526390038e-05,
      "loss": 0.5555,
      "mean_token_accuracy": 0.8246101923286915,
      "num_tokens": 784161672.0,
      "step": 24570
    },
    {
      "entropy": 0.4478213884867728,
      "epoch": 2.260654050292983,
      "grad_norm": 0.15142232179641724,
      "learning_rate": 2.4648081700248416e-05,
      "loss": 0.436,
      "mean_token_accuracy": 0.8658694066107273,
      "num_tokens": 784193681.0,
      "step": 24571
    },
    {
      "entropy": 0.4635441112332046,
      "epoch": 2.260746056664424,
      "grad_norm": 0.1574452966451645,
      "learning_rate": 2.464501487410679e-05,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8526921980082989,
      "num_tokens": 784225383.0,
      "step": 24572
    },
    {
      "entropy": 0.5709956455975771,
      "epoch": 2.260838063035865,
      "grad_norm": 0.168389230966568,
      "learning_rate": 2.464194804796516e-05,
      "loss": 0.5614,
      "mean_token_accuracy": 0.8231245279312134,
      "num_tokens": 784257104.0,
      "step": 24573
    },
    {
      "entropy": 0.3983075665310025,
      "epoch": 2.2609300694073067,
      "grad_norm": 0.14919939637184143,
      "learning_rate": 2.4638881221823536e-05,
      "loss": 0.3853,
      "mean_token_accuracy": 0.8719161339104176,
      "num_tokens": 784288898.0,
      "step": 24574
    },
    {
      "entropy": 0.6055213734507561,
      "epoch": 2.261022075778748,
      "grad_norm": 0.16822756826877594,
      "learning_rate": 2.463581439568191e-05,
      "loss": 0.5921,
      "mean_token_accuracy": 0.8165372014045715,
      "num_tokens": 784320886.0,
      "step": 24575
    },
    {
      "entropy": 0.5183099191635847,
      "epoch": 2.261114082150189,
      "grad_norm": 0.16231697797775269,
      "learning_rate": 2.4632747569540286e-05,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8373265005648136,
      "num_tokens": 784352704.0,
      "step": 24576
    },
    {
      "entropy": 0.39107468351721764,
      "epoch": 2.26120608852163,
      "grad_norm": 0.14610712230205536,
      "learning_rate": 2.462968074339866e-05,
      "loss": 0.3725,
      "mean_token_accuracy": 0.8809253200888634,
      "num_tokens": 784384399.0,
      "step": 24577
    },
    {
      "entropy": 0.48910402320325375,
      "epoch": 2.261298094893071,
      "grad_norm": 0.1570131927728653,
      "learning_rate": 2.462661391725703e-05,
      "loss": 0.4705,
      "mean_token_accuracy": 0.8513090088963509,
      "num_tokens": 784416473.0,
      "step": 24578
    },
    {
      "entropy": 0.5405677361413836,
      "epoch": 2.261390101264513,
      "grad_norm": 0.1630258560180664,
      "learning_rate": 2.4623547091115406e-05,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8282141461968422,
      "num_tokens": 784448268.0,
      "step": 24579
    },
    {
      "entropy": 0.5196483638137579,
      "epoch": 2.261482107635954,
      "grad_norm": 0.16017219424247742,
      "learning_rate": 2.462048026497378e-05,
      "loss": 0.505,
      "mean_token_accuracy": 0.8393938057124615,
      "num_tokens": 784480767.0,
      "step": 24580
    },
    {
      "entropy": 0.5094812074676156,
      "epoch": 2.261574114007395,
      "grad_norm": 0.15653620660305023,
      "learning_rate": 2.4617413438832155e-05,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8429575748741627,
      "num_tokens": 784512762.0,
      "step": 24581
    },
    {
      "entropy": 0.5012815929949284,
      "epoch": 2.261666120378836,
      "grad_norm": 0.15092526376247406,
      "learning_rate": 2.4614346612690527e-05,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8474145606160164,
      "num_tokens": 784545051.0,
      "step": 24582
    },
    {
      "entropy": 0.43979067355394363,
      "epoch": 2.2617581267502773,
      "grad_norm": 0.15273168683052063,
      "learning_rate": 2.46112797865489e-05,
      "loss": 0.4277,
      "mean_token_accuracy": 0.8655994795262814,
      "num_tokens": 784576940.0,
      "step": 24583
    },
    {
      "entropy": 0.5013016103766859,
      "epoch": 2.261850133121719,
      "grad_norm": 0.15497753024101257,
      "learning_rate": 2.4608212960407276e-05,
      "loss": 0.495,
      "mean_token_accuracy": 0.8441880196332932,
      "num_tokens": 784609041.0,
      "step": 24584
    },
    {
      "entropy": 0.5041076671332121,
      "epoch": 2.26194213949316,
      "grad_norm": 0.16097766160964966,
      "learning_rate": 2.460514613426565e-05,
      "loss": 0.487,
      "mean_token_accuracy": 0.8423672839999199,
      "num_tokens": 784641016.0,
      "step": 24585
    },
    {
      "entropy": 0.587452826090157,
      "epoch": 2.262034145864601,
      "grad_norm": 0.1714186668395996,
      "learning_rate": 2.4602079308124022e-05,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8200637772679329,
      "num_tokens": 784672738.0,
      "step": 24586
    },
    {
      "entropy": 0.5067586393561214,
      "epoch": 2.2621261522360423,
      "grad_norm": 0.15675923228263855,
      "learning_rate": 2.4599012481982397e-05,
      "loss": 0.4829,
      "mean_token_accuracy": 0.8411823511123657,
      "num_tokens": 784704207.0,
      "step": 24587
    },
    {
      "entropy": 0.39393299631774426,
      "epoch": 2.2622181586074834,
      "grad_norm": 0.1383354514837265,
      "learning_rate": 2.459594565584077e-05,
      "loss": 0.3717,
      "mean_token_accuracy": 0.879996944218874,
      "num_tokens": 784736297.0,
      "step": 24588
    },
    {
      "entropy": 0.506431931629777,
      "epoch": 2.262310164978925,
      "grad_norm": 0.16159066557884216,
      "learning_rate": 2.4592878829699146e-05,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8430681303143501,
      "num_tokens": 784768148.0,
      "step": 24589
    },
    {
      "entropy": 0.507101443130523,
      "epoch": 2.262402171350366,
      "grad_norm": 0.16303688287734985,
      "learning_rate": 2.458981200355752e-05,
      "loss": 0.5026,
      "mean_token_accuracy": 0.8402571529150009,
      "num_tokens": 784800454.0,
      "step": 24590
    },
    {
      "entropy": 0.5112983472645283,
      "epoch": 2.2624941777218073,
      "grad_norm": 0.15917035937309265,
      "learning_rate": 2.4586745177415892e-05,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8410499393939972,
      "num_tokens": 784832413.0,
      "step": 24591
    },
    {
      "entropy": 0.47784466156736016,
      "epoch": 2.2625861840932484,
      "grad_norm": 0.15865284204483032,
      "learning_rate": 2.4583678351274266e-05,
      "loss": 0.4647,
      "mean_token_accuracy": 0.8511634208261967,
      "num_tokens": 784864412.0,
      "step": 24592
    },
    {
      "entropy": 0.5040627382695675,
      "epoch": 2.2626781904646895,
      "grad_norm": 0.163718581199646,
      "learning_rate": 2.458061152513264e-05,
      "loss": 0.4961,
      "mean_token_accuracy": 0.8422131165862083,
      "num_tokens": 784895866.0,
      "step": 24593
    },
    {
      "entropy": 0.4755829409696162,
      "epoch": 2.262770196836131,
      "grad_norm": 0.15398292243480682,
      "learning_rate": 2.4577544698991016e-05,
      "loss": 0.4705,
      "mean_token_accuracy": 0.8518611155450344,
      "num_tokens": 784927778.0,
      "step": 24594
    },
    {
      "entropy": 0.48168890713714063,
      "epoch": 2.2628622032075723,
      "grad_norm": 0.15373359620571136,
      "learning_rate": 2.4574477872849387e-05,
      "loss": 0.4688,
      "mean_token_accuracy": 0.8520983085036278,
      "num_tokens": 784960176.0,
      "step": 24595
    },
    {
      "entropy": 0.47748640179634094,
      "epoch": 2.2629542095790134,
      "grad_norm": 0.15286210179328918,
      "learning_rate": 2.457141104670776e-05,
      "loss": 0.4676,
      "mean_token_accuracy": 0.8503114357590675,
      "num_tokens": 784991990.0,
      "step": 24596
    },
    {
      "entropy": 0.5999750681221485,
      "epoch": 2.2630462159504545,
      "grad_norm": 0.17166060209274292,
      "learning_rate": 2.4568344220566136e-05,
      "loss": 0.603,
      "mean_token_accuracy": 0.8160476870834827,
      "num_tokens": 785024717.0,
      "step": 24597
    },
    {
      "entropy": 0.5200183093547821,
      "epoch": 2.2631382223218957,
      "grad_norm": 0.15564264357089996,
      "learning_rate": 2.456527739442451e-05,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8392484188079834,
      "num_tokens": 785057175.0,
      "step": 24598
    },
    {
      "entropy": 0.5562101055402309,
      "epoch": 2.2632302286933372,
      "grad_norm": 0.1685962826013565,
      "learning_rate": 2.4562210568282886e-05,
      "loss": 0.5463,
      "mean_token_accuracy": 0.8301191404461861,
      "num_tokens": 785089113.0,
      "step": 24599
    },
    {
      "entropy": 0.5113824470899999,
      "epoch": 2.2633222350647784,
      "grad_norm": 0.16472026705741882,
      "learning_rate": 2.4559143742141257e-05,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8428995609283447,
      "num_tokens": 785121261.0,
      "step": 24600
    },
    {
      "entropy": 0.4193472333718091,
      "epoch": 2.2634142414362195,
      "grad_norm": 0.14568482339382172,
      "learning_rate": 2.455607691599963e-05,
      "loss": 0.4071,
      "mean_token_accuracy": 0.8696447610855103,
      "num_tokens": 785153231.0,
      "step": 24601
    },
    {
      "entropy": 0.42052369471639395,
      "epoch": 2.2635062478076606,
      "grad_norm": 0.14793668687343597,
      "learning_rate": 2.455301008985801e-05,
      "loss": 0.4145,
      "mean_token_accuracy": 0.8693816587328911,
      "num_tokens": 785185740.0,
      "step": 24602
    },
    {
      "entropy": 0.49737492110580206,
      "epoch": 2.2635982541791018,
      "grad_norm": 0.1608361303806305,
      "learning_rate": 2.454994326371638e-05,
      "loss": 0.4776,
      "mean_token_accuracy": 0.848260436207056,
      "num_tokens": 785217341.0,
      "step": 24603
    },
    {
      "entropy": 0.569516334682703,
      "epoch": 2.2636902605505433,
      "grad_norm": 0.16350765526294708,
      "learning_rate": 2.4546876437574755e-05,
      "loss": 0.5569,
      "mean_token_accuracy": 0.8252685368061066,
      "num_tokens": 785249580.0,
      "step": 24604
    },
    {
      "entropy": 0.47508107451722026,
      "epoch": 2.2637822669219845,
      "grad_norm": 0.15285125374794006,
      "learning_rate": 2.454380961143313e-05,
      "loss": 0.4603,
      "mean_token_accuracy": 0.8545135371387005,
      "num_tokens": 785282237.0,
      "step": 24605
    },
    {
      "entropy": 0.46010617911815643,
      "epoch": 2.2638742732934256,
      "grad_norm": 0.15449948608875275,
      "learning_rate": 2.4540742785291505e-05,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8561110757291317,
      "num_tokens": 785314079.0,
      "step": 24606
    },
    {
      "entropy": 0.4993338081985712,
      "epoch": 2.2639662796648667,
      "grad_norm": 0.160341277718544,
      "learning_rate": 2.453767595914988e-05,
      "loss": 0.4821,
      "mean_token_accuracy": 0.8466052822768688,
      "num_tokens": 785345514.0,
      "step": 24607
    },
    {
      "entropy": 0.6441259980201721,
      "epoch": 2.264058286036308,
      "grad_norm": 0.17642800509929657,
      "learning_rate": 2.453460913300825e-05,
      "loss": 0.6249,
      "mean_token_accuracy": 0.8027389906346798,
      "num_tokens": 785377197.0,
      "step": 24608
    },
    {
      "entropy": 0.572479821741581,
      "epoch": 2.2641502924077495,
      "grad_norm": 0.1659451127052307,
      "learning_rate": 2.4531542306866625e-05,
      "loss": 0.5514,
      "mean_token_accuracy": 0.8246876671910286,
      "num_tokens": 785408844.0,
      "step": 24609
    },
    {
      "entropy": 0.4621280967257917,
      "epoch": 2.2642422987791906,
      "grad_norm": 0.1535191386938095,
      "learning_rate": 2.4528475480725e-05,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8580085337162018,
      "num_tokens": 785441265.0,
      "step": 24610
    },
    {
      "entropy": 0.5411813259124756,
      "epoch": 2.2643343051506317,
      "grad_norm": 0.1647164225578308,
      "learning_rate": 2.4525408654583375e-05,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8342201821506023,
      "num_tokens": 785473308.0,
      "step": 24611
    },
    {
      "entropy": 0.6243365965783596,
      "epoch": 2.264426311522073,
      "grad_norm": 0.1719638556241989,
      "learning_rate": 2.4522341828441746e-05,
      "loss": 0.6081,
      "mean_token_accuracy": 0.810581523925066,
      "num_tokens": 785506046.0,
      "step": 24612
    },
    {
      "entropy": 0.4782022722065449,
      "epoch": 2.264518317893514,
      "grad_norm": 0.15711839497089386,
      "learning_rate": 2.451927500230012e-05,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8498905338346958,
      "num_tokens": 785537244.0,
      "step": 24613
    },
    {
      "entropy": 0.49029276706278324,
      "epoch": 2.2646103242649556,
      "grad_norm": 0.15835067629814148,
      "learning_rate": 2.4516208176158495e-05,
      "loss": 0.486,
      "mean_token_accuracy": 0.8457020558416843,
      "num_tokens": 785569313.0,
      "step": 24614
    },
    {
      "entropy": 0.48105119867250323,
      "epoch": 2.2647023306363967,
      "grad_norm": 0.152831569314003,
      "learning_rate": 2.451314135001687e-05,
      "loss": 0.4694,
      "mean_token_accuracy": 0.8519275635480881,
      "num_tokens": 785601132.0,
      "step": 24615
    },
    {
      "entropy": 0.5301421862095594,
      "epoch": 2.264794337007838,
      "grad_norm": 0.16604310274124146,
      "learning_rate": 2.451007452387524e-05,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8359375968575478,
      "num_tokens": 785632171.0,
      "step": 24616
    },
    {
      "entropy": 0.5793731543235481,
      "epoch": 2.264886343379279,
      "grad_norm": 0.16653139889240265,
      "learning_rate": 2.4507007697733616e-05,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8186024539172649,
      "num_tokens": 785664100.0,
      "step": 24617
    },
    {
      "entropy": 0.527875606669113,
      "epoch": 2.26497834975072,
      "grad_norm": 0.16406284272670746,
      "learning_rate": 2.450394087159199e-05,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8363153375685215,
      "num_tokens": 785696868.0,
      "step": 24618
    },
    {
      "entropy": 0.5872726868838072,
      "epoch": 2.2650703561221617,
      "grad_norm": 0.170766681432724,
      "learning_rate": 2.4500874045450365e-05,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8217771239578724,
      "num_tokens": 785728471.0,
      "step": 24619
    },
    {
      "entropy": 0.5070331515744328,
      "epoch": 2.265162362493603,
      "grad_norm": 0.15831893682479858,
      "learning_rate": 2.449780721930874e-05,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8442956060171127,
      "num_tokens": 785760369.0,
      "step": 24620
    },
    {
      "entropy": 0.601302036549896,
      "epoch": 2.265254368865044,
      "grad_norm": 0.17003385722637177,
      "learning_rate": 2.449474039316711e-05,
      "loss": 0.5839,
      "mean_token_accuracy": 0.8161605671048164,
      "num_tokens": 785792499.0,
      "step": 24621
    },
    {
      "entropy": 0.4814975820481777,
      "epoch": 2.265346375236485,
      "grad_norm": 0.15550051629543304,
      "learning_rate": 2.4491673567025486e-05,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8490251563489437,
      "num_tokens": 785824648.0,
      "step": 24622
    },
    {
      "entropy": 0.5159069839864969,
      "epoch": 2.265438381607926,
      "grad_norm": 0.15916121006011963,
      "learning_rate": 2.448860674088386e-05,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8353670686483383,
      "num_tokens": 785856228.0,
      "step": 24623
    },
    {
      "entropy": 0.5356450527906418,
      "epoch": 2.265530387979368,
      "grad_norm": 0.1728867143392563,
      "learning_rate": 2.4485539914742235e-05,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8308768682181835,
      "num_tokens": 785888252.0,
      "step": 24624
    },
    {
      "entropy": 0.47898038383573294,
      "epoch": 2.265622394350809,
      "grad_norm": 0.15496112406253815,
      "learning_rate": 2.4482473088600606e-05,
      "loss": 0.4717,
      "mean_token_accuracy": 0.8484927862882614,
      "num_tokens": 785920627.0,
      "step": 24625
    },
    {
      "entropy": 0.477760492824018,
      "epoch": 2.26571440072225,
      "grad_norm": 0.1672181487083435,
      "learning_rate": 2.447940626245898e-05,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8529118597507477,
      "num_tokens": 785951744.0,
      "step": 24626
    },
    {
      "entropy": 0.4586123190820217,
      "epoch": 2.265806407093691,
      "grad_norm": 0.15514370799064636,
      "learning_rate": 2.4476339436317355e-05,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8579793274402618,
      "num_tokens": 785983615.0,
      "step": 24627
    },
    {
      "entropy": 0.4541139341890812,
      "epoch": 2.2658984134651323,
      "grad_norm": 0.15295706689357758,
      "learning_rate": 2.447327261017573e-05,
      "loss": 0.4473,
      "mean_token_accuracy": 0.8588734157383442,
      "num_tokens": 786015235.0,
      "step": 24628
    },
    {
      "entropy": 0.498727485537529,
      "epoch": 2.265990419836574,
      "grad_norm": 0.1632532924413681,
      "learning_rate": 2.4470205784034105e-05,
      "loss": 0.4842,
      "mean_token_accuracy": 0.8442413993179798,
      "num_tokens": 786047217.0,
      "step": 24629
    },
    {
      "entropy": 0.4707432081922889,
      "epoch": 2.266082426208015,
      "grad_norm": 0.1528080254793167,
      "learning_rate": 2.4467138957892476e-05,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8510444164276123,
      "num_tokens": 786079367.0,
      "step": 24630
    },
    {
      "entropy": 0.4807713436894119,
      "epoch": 2.266174432579456,
      "grad_norm": 0.1603313386440277,
      "learning_rate": 2.446407213175085e-05,
      "loss": 0.4695,
      "mean_token_accuracy": 0.8529511503875256,
      "num_tokens": 786110267.0,
      "step": 24631
    },
    {
      "entropy": 0.4961147643625736,
      "epoch": 2.2662664389508973,
      "grad_norm": 0.15681007504463196,
      "learning_rate": 2.446100530560923e-05,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8486829213798046,
      "num_tokens": 786142168.0,
      "step": 24632
    },
    {
      "entropy": 0.4858971694484353,
      "epoch": 2.2663584453223384,
      "grad_norm": 0.15788188576698303,
      "learning_rate": 2.44579384794676e-05,
      "loss": 0.4605,
      "mean_token_accuracy": 0.8538813591003418,
      "num_tokens": 786173765.0,
      "step": 24633
    },
    {
      "entropy": 0.509690348058939,
      "epoch": 2.26645045169378,
      "grad_norm": 0.16234709322452545,
      "learning_rate": 2.4454871653325975e-05,
      "loss": 0.4932,
      "mean_token_accuracy": 0.8487284369766712,
      "num_tokens": 786205665.0,
      "step": 24634
    },
    {
      "entropy": 0.5065703121945262,
      "epoch": 2.266542458065221,
      "grad_norm": 0.1578988879919052,
      "learning_rate": 2.445180482718435e-05,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8431923016905785,
      "num_tokens": 786237860.0,
      "step": 24635
    },
    {
      "entropy": 0.4165531061589718,
      "epoch": 2.2666344644366623,
      "grad_norm": 0.150763601064682,
      "learning_rate": 2.4448738001042724e-05,
      "loss": 0.4076,
      "mean_token_accuracy": 0.8697074055671692,
      "num_tokens": 786269903.0,
      "step": 24636
    },
    {
      "entropy": 0.43490260327234864,
      "epoch": 2.2667264708081034,
      "grad_norm": 0.15556414425373077,
      "learning_rate": 2.44456711749011e-05,
      "loss": 0.4121,
      "mean_token_accuracy": 0.8659689426422119,
      "num_tokens": 786301589.0,
      "step": 24637
    },
    {
      "entropy": 0.53483344335109,
      "epoch": 2.2668184771795445,
      "grad_norm": 0.16524842381477356,
      "learning_rate": 2.444260434875947e-05,
      "loss": 0.5253,
      "mean_token_accuracy": 0.8310314528644085,
      "num_tokens": 786332925.0,
      "step": 24638
    },
    {
      "entropy": 0.47245862893760204,
      "epoch": 2.266910483550986,
      "grad_norm": 0.15218400955200195,
      "learning_rate": 2.4439537522617844e-05,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8557784222066402,
      "num_tokens": 786365363.0,
      "step": 24639
    },
    {
      "entropy": 0.6708200061693788,
      "epoch": 2.2670024899224273,
      "grad_norm": 0.18086713552474976,
      "learning_rate": 2.443647069647622e-05,
      "loss": 0.6603,
      "mean_token_accuracy": 0.7924203276634216,
      "num_tokens": 786397068.0,
      "step": 24640
    },
    {
      "entropy": 0.45811540307477117,
      "epoch": 2.2670944962938684,
      "grad_norm": 0.1534530073404312,
      "learning_rate": 2.4433403870334594e-05,
      "loss": 0.4418,
      "mean_token_accuracy": 0.8583534136414528,
      "num_tokens": 786429193.0,
      "step": 24641
    },
    {
      "entropy": 0.5382307795807719,
      "epoch": 2.2671865026653095,
      "grad_norm": 0.16451695561408997,
      "learning_rate": 2.4430337044192965e-05,
      "loss": 0.5278,
      "mean_token_accuracy": 0.8338921740651131,
      "num_tokens": 786461105.0,
      "step": 24642
    },
    {
      "entropy": 0.5458865556865931,
      "epoch": 2.2672785090367507,
      "grad_norm": 0.16304361820220947,
      "learning_rate": 2.442727021805134e-05,
      "loss": 0.5298,
      "mean_token_accuracy": 0.8337337970733643,
      "num_tokens": 786492930.0,
      "step": 24643
    },
    {
      "entropy": 0.5094610939268023,
      "epoch": 2.2673705154081922,
      "grad_norm": 0.15820640325546265,
      "learning_rate": 2.4424203391909714e-05,
      "loss": 0.499,
      "mean_token_accuracy": 0.8386369161307812,
      "num_tokens": 786525237.0,
      "step": 24644
    },
    {
      "entropy": 0.5212052734568715,
      "epoch": 2.2674625217796334,
      "grad_norm": 0.1618703007698059,
      "learning_rate": 2.442113656576809e-05,
      "loss": 0.5,
      "mean_token_accuracy": 0.8381835073232651,
      "num_tokens": 786556595.0,
      "step": 24645
    },
    {
      "entropy": 0.39505948266014457,
      "epoch": 2.2675545281510745,
      "grad_norm": 0.14351169764995575,
      "learning_rate": 2.4418069739626464e-05,
      "loss": 0.3747,
      "mean_token_accuracy": 0.8795206546783447,
      "num_tokens": 786588093.0,
      "step": 24646
    },
    {
      "entropy": 0.4508899771608412,
      "epoch": 2.2676465345225156,
      "grad_norm": 0.15223978459835052,
      "learning_rate": 2.4415002913484835e-05,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8614583611488342,
      "num_tokens": 786619675.0,
      "step": 24647
    },
    {
      "entropy": 0.6121963113546371,
      "epoch": 2.2677385408939568,
      "grad_norm": 0.17256255447864532,
      "learning_rate": 2.441193608734321e-05,
      "loss": 0.6012,
      "mean_token_accuracy": 0.8087837286293507,
      "num_tokens": 786651855.0,
      "step": 24648
    },
    {
      "entropy": 0.5380998570472002,
      "epoch": 2.2678305472653983,
      "grad_norm": 0.1633450835943222,
      "learning_rate": 2.4408869261201584e-05,
      "loss": 0.5114,
      "mean_token_accuracy": 0.8343476988375187,
      "num_tokens": 786683675.0,
      "step": 24649
    },
    {
      "entropy": 0.6481369528919458,
      "epoch": 2.2679225536368395,
      "grad_norm": 0.17098495364189148,
      "learning_rate": 2.440580243505996e-05,
      "loss": 0.6479,
      "mean_token_accuracy": 0.7989391125738621,
      "num_tokens": 786714849.0,
      "step": 24650
    },
    {
      "entropy": 0.5155633063986897,
      "epoch": 2.2680145600082806,
      "grad_norm": 0.16313019394874573,
      "learning_rate": 2.440273560891833e-05,
      "loss": 0.5109,
      "mean_token_accuracy": 0.8378984332084656,
      "num_tokens": 786746570.0,
      "step": 24651
    },
    {
      "entropy": 0.5715627064928412,
      "epoch": 2.2681065663797217,
      "grad_norm": 0.16955482959747314,
      "learning_rate": 2.4399668782776705e-05,
      "loss": 0.5628,
      "mean_token_accuracy": 0.8239427506923676,
      "num_tokens": 786778427.0,
      "step": 24652
    },
    {
      "entropy": 0.519910080358386,
      "epoch": 2.268198572751163,
      "grad_norm": 0.15750010311603546,
      "learning_rate": 2.439660195663508e-05,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8375006765127182,
      "num_tokens": 786810695.0,
      "step": 24653
    },
    {
      "entropy": 0.6179114878177643,
      "epoch": 2.2682905791226045,
      "grad_norm": 0.1727377027273178,
      "learning_rate": 2.4393535130493454e-05,
      "loss": 0.61,
      "mean_token_accuracy": 0.8054852075874805,
      "num_tokens": 786842109.0,
      "step": 24654
    },
    {
      "entropy": 0.4401693446561694,
      "epoch": 2.2683825854940456,
      "grad_norm": 0.14991846680641174,
      "learning_rate": 2.4390468304351825e-05,
      "loss": 0.4285,
      "mean_token_accuracy": 0.8627111092209816,
      "num_tokens": 786874554.0,
      "step": 24655
    },
    {
      "entropy": 0.5511954082176089,
      "epoch": 2.2684745918654867,
      "grad_norm": 0.1668449491262436,
      "learning_rate": 2.43874014782102e-05,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8297596238553524,
      "num_tokens": 786906956.0,
      "step": 24656
    },
    {
      "entropy": 0.5152017311193049,
      "epoch": 2.268566598236928,
      "grad_norm": 0.16057193279266357,
      "learning_rate": 2.4384334652068574e-05,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8409768156707287,
      "num_tokens": 786938796.0,
      "step": 24657
    },
    {
      "entropy": 0.3767043468542397,
      "epoch": 2.268658604608369,
      "grad_norm": 0.14043666422367096,
      "learning_rate": 2.438126782592695e-05,
      "loss": 0.3596,
      "mean_token_accuracy": 0.8845101669430733,
      "num_tokens": 786971359.0,
      "step": 24658
    },
    {
      "entropy": 0.5439139660447836,
      "epoch": 2.2687506109798106,
      "grad_norm": 0.16068032383918762,
      "learning_rate": 2.4378200999785324e-05,
      "loss": 0.5302,
      "mean_token_accuracy": 0.8337659575045109,
      "num_tokens": 787004117.0,
      "step": 24659
    },
    {
      "entropy": 0.5393258780241013,
      "epoch": 2.2688426173512517,
      "grad_norm": 0.16388250887393951,
      "learning_rate": 2.4375134173643695e-05,
      "loss": 0.528,
      "mean_token_accuracy": 0.8337978646159172,
      "num_tokens": 787036663.0,
      "step": 24660
    },
    {
      "entropy": 0.5403086841106415,
      "epoch": 2.268934623722693,
      "grad_norm": 0.16199447214603424,
      "learning_rate": 2.437206734750207e-05,
      "loss": 0.5216,
      "mean_token_accuracy": 0.83399348706007,
      "num_tokens": 787068492.0,
      "step": 24661
    },
    {
      "entropy": 0.5095596835017204,
      "epoch": 2.269026630094134,
      "grad_norm": 0.15844812989234924,
      "learning_rate": 2.4369000521360444e-05,
      "loss": 0.4915,
      "mean_token_accuracy": 0.8441050536930561,
      "num_tokens": 787100084.0,
      "step": 24662
    },
    {
      "entropy": 0.5616978406906128,
      "epoch": 2.269118636465575,
      "grad_norm": 0.16142219305038452,
      "learning_rate": 2.436593369521882e-05,
      "loss": 0.5353,
      "mean_token_accuracy": 0.8265562169253826,
      "num_tokens": 787131548.0,
      "step": 24663
    },
    {
      "entropy": 0.5133373197168112,
      "epoch": 2.2692106428370167,
      "grad_norm": 0.16376936435699463,
      "learning_rate": 2.4362866869077194e-05,
      "loss": 0.4981,
      "mean_token_accuracy": 0.8440220318734646,
      "num_tokens": 787164037.0,
      "step": 24664
    },
    {
      "entropy": 0.5245968773961067,
      "epoch": 2.269302649208458,
      "grad_norm": 0.15457439422607422,
      "learning_rate": 2.4359800042935568e-05,
      "loss": 0.5114,
      "mean_token_accuracy": 0.8374164178967476,
      "num_tokens": 787195675.0,
      "step": 24665
    },
    {
      "entropy": 0.5232439725659788,
      "epoch": 2.269394655579899,
      "grad_norm": 0.16735953092575073,
      "learning_rate": 2.4356733216793943e-05,
      "loss": 0.498,
      "mean_token_accuracy": 0.8407024815678596,
      "num_tokens": 787227317.0,
      "step": 24666
    },
    {
      "entropy": 0.4913551793433726,
      "epoch": 2.26948666195134,
      "grad_norm": 0.15691642463207245,
      "learning_rate": 2.4353666390652318e-05,
      "loss": 0.4747,
      "mean_token_accuracy": 0.8490029610693455,
      "num_tokens": 787258074.0,
      "step": 24667
    },
    {
      "entropy": 0.505562761798501,
      "epoch": 2.269578668322781,
      "grad_norm": 0.16079182922840118,
      "learning_rate": 2.435059956451069e-05,
      "loss": 0.4952,
      "mean_token_accuracy": 0.8409130498766899,
      "num_tokens": 787290285.0,
      "step": 24668
    },
    {
      "entropy": 0.4326065890491009,
      "epoch": 2.269670674694223,
      "grad_norm": 0.15048451721668243,
      "learning_rate": 2.4347532738369063e-05,
      "loss": 0.4246,
      "mean_token_accuracy": 0.8662302866578102,
      "num_tokens": 787322433.0,
      "step": 24669
    },
    {
      "entropy": 0.4891739869490266,
      "epoch": 2.269762681065664,
      "grad_norm": 0.1619122177362442,
      "learning_rate": 2.4344465912227438e-05,
      "loss": 0.4741,
      "mean_token_accuracy": 0.8503320701420307,
      "num_tokens": 787353717.0,
      "step": 24670
    },
    {
      "entropy": 0.5692709526047111,
      "epoch": 2.269854687437105,
      "grad_norm": 0.16790790855884552,
      "learning_rate": 2.4341399086085813e-05,
      "loss": 0.5711,
      "mean_token_accuracy": 0.8165594302117825,
      "num_tokens": 787385975.0,
      "step": 24671
    },
    {
      "entropy": 0.5025130156427622,
      "epoch": 2.269946693808546,
      "grad_norm": 0.16399770975112915,
      "learning_rate": 2.4338332259944184e-05,
      "loss": 0.4956,
      "mean_token_accuracy": 0.843120351433754,
      "num_tokens": 787418329.0,
      "step": 24672
    },
    {
      "entropy": 0.5521039087325335,
      "epoch": 2.2700387001799873,
      "grad_norm": 0.16030244529247284,
      "learning_rate": 2.433526543380256e-05,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8329082429409027,
      "num_tokens": 787450480.0,
      "step": 24673
    },
    {
      "entropy": 0.5099990982562304,
      "epoch": 2.270130706551429,
      "grad_norm": 0.16597434878349304,
      "learning_rate": 2.4332198607660933e-05,
      "loss": 0.5012,
      "mean_token_accuracy": 0.8384226635098457,
      "num_tokens": 787482496.0,
      "step": 24674
    },
    {
      "entropy": 0.5740559697151184,
      "epoch": 2.27022271292287,
      "grad_norm": 0.16908711194992065,
      "learning_rate": 2.4329131781519308e-05,
      "loss": 0.5765,
      "mean_token_accuracy": 0.82041335105896,
      "num_tokens": 787514882.0,
      "step": 24675
    },
    {
      "entropy": 0.505657265894115,
      "epoch": 2.270314719294311,
      "grad_norm": 0.16363483667373657,
      "learning_rate": 2.4326064955377683e-05,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8481330461800098,
      "num_tokens": 787546477.0,
      "step": 24676
    },
    {
      "entropy": 0.5014156298711896,
      "epoch": 2.2704067256657523,
      "grad_norm": 0.15535040199756622,
      "learning_rate": 2.4322998129236054e-05,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8468810841441154,
      "num_tokens": 787578423.0,
      "step": 24677
    },
    {
      "entropy": 0.4327797684818506,
      "epoch": 2.2704987320371934,
      "grad_norm": 0.15430137515068054,
      "learning_rate": 2.431993130309443e-05,
      "loss": 0.4227,
      "mean_token_accuracy": 0.8645026534795761,
      "num_tokens": 787611108.0,
      "step": 24678
    },
    {
      "entropy": 0.5642021158710122,
      "epoch": 2.270590738408635,
      "grad_norm": 0.16541652381420135,
      "learning_rate": 2.4316864476952803e-05,
      "loss": 0.5561,
      "mean_token_accuracy": 0.8216247446835041,
      "num_tokens": 787643485.0,
      "step": 24679
    },
    {
      "entropy": 0.494864150416106,
      "epoch": 2.270682744780076,
      "grad_norm": 0.15950293838977814,
      "learning_rate": 2.4313797650811178e-05,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8448188230395317,
      "num_tokens": 787675336.0,
      "step": 24680
    },
    {
      "entropy": 0.5247862562537193,
      "epoch": 2.2707747511515173,
      "grad_norm": 0.15846924483776093,
      "learning_rate": 2.431073082466955e-05,
      "loss": 0.5011,
      "mean_token_accuracy": 0.8400214724242687,
      "num_tokens": 787706850.0,
      "step": 24681
    },
    {
      "entropy": 0.6482155006378889,
      "epoch": 2.2708667575229584,
      "grad_norm": 0.18320755660533905,
      "learning_rate": 2.4307663998527924e-05,
      "loss": 0.6268,
      "mean_token_accuracy": 0.7994188405573368,
      "num_tokens": 787738075.0,
      "step": 24682
    },
    {
      "entropy": 0.6313112694770098,
      "epoch": 2.2709587638943995,
      "grad_norm": 0.1751837283372879,
      "learning_rate": 2.43045971723863e-05,
      "loss": 0.6158,
      "mean_token_accuracy": 0.8071538843214512,
      "num_tokens": 787769894.0,
      "step": 24683
    },
    {
      "entropy": 0.5671813599765301,
      "epoch": 2.271050770265841,
      "grad_norm": 0.16784368455410004,
      "learning_rate": 2.4301530346244673e-05,
      "loss": 0.5603,
      "mean_token_accuracy": 0.825628224760294,
      "num_tokens": 787801541.0,
      "step": 24684
    },
    {
      "entropy": 0.51019108411856,
      "epoch": 2.2711427766372823,
      "grad_norm": 0.15379920601844788,
      "learning_rate": 2.4298463520103044e-05,
      "loss": 0.489,
      "mean_token_accuracy": 0.848400391638279,
      "num_tokens": 787834280.0,
      "step": 24685
    },
    {
      "entropy": 0.5104866530746222,
      "epoch": 2.2712347830087234,
      "grad_norm": 0.15704908967018127,
      "learning_rate": 2.429539669396142e-05,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8426595889031887,
      "num_tokens": 787866701.0,
      "step": 24686
    },
    {
      "entropy": 0.45555978431366384,
      "epoch": 2.2713267893801645,
      "grad_norm": 0.1517704725265503,
      "learning_rate": 2.4292329867819794e-05,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8568983636796474,
      "num_tokens": 787898892.0,
      "step": 24687
    },
    {
      "entropy": 0.4647428523749113,
      "epoch": 2.2714187957516057,
      "grad_norm": 0.15136171877384186,
      "learning_rate": 2.4289263041678168e-05,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8574569709599018,
      "num_tokens": 787931233.0,
      "step": 24688
    },
    {
      "entropy": 0.4809472148772329,
      "epoch": 2.2715108021230472,
      "grad_norm": 0.15752379596233368,
      "learning_rate": 2.4286196215536543e-05,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8510449267923832,
      "num_tokens": 787962810.0,
      "step": 24689
    },
    {
      "entropy": 0.45497375912964344,
      "epoch": 2.2716028084944884,
      "grad_norm": 0.15759451687335968,
      "learning_rate": 2.4283129389394914e-05,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8584309667348862,
      "num_tokens": 787994341.0,
      "step": 24690
    },
    {
      "entropy": 0.4803036190569401,
      "epoch": 2.2716948148659295,
      "grad_norm": 0.155232772231102,
      "learning_rate": 2.428006256325329e-05,
      "loss": 0.4741,
      "mean_token_accuracy": 0.8480013087391853,
      "num_tokens": 788026029.0,
      "step": 24691
    },
    {
      "entropy": 0.5106448289006948,
      "epoch": 2.2717868212373706,
      "grad_norm": 0.15337815880775452,
      "learning_rate": 2.4276995737111663e-05,
      "loss": 0.487,
      "mean_token_accuracy": 0.8449308685958385,
      "num_tokens": 788058247.0,
      "step": 24692
    },
    {
      "entropy": 0.5276581738144159,
      "epoch": 2.2718788276088118,
      "grad_norm": 0.1616590917110443,
      "learning_rate": 2.4273928910970038e-05,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8378995172679424,
      "num_tokens": 788090237.0,
      "step": 24693
    },
    {
      "entropy": 0.5112798875197768,
      "epoch": 2.2719708339802533,
      "grad_norm": 0.16211561858654022,
      "learning_rate": 2.4270862084828413e-05,
      "loss": 0.5067,
      "mean_token_accuracy": 0.8412985280156136,
      "num_tokens": 788121974.0,
      "step": 24694
    },
    {
      "entropy": 0.5725947394967079,
      "epoch": 2.2720628403516945,
      "grad_norm": 0.16960158944129944,
      "learning_rate": 2.4267795258686787e-05,
      "loss": 0.5487,
      "mean_token_accuracy": 0.827119879424572,
      "num_tokens": 788153357.0,
      "step": 24695
    },
    {
      "entropy": 0.5762394834309816,
      "epoch": 2.2721548467231356,
      "grad_norm": 0.16974899172782898,
      "learning_rate": 2.4264728432545162e-05,
      "loss": 0.5762,
      "mean_token_accuracy": 0.8192118480801582,
      "num_tokens": 788184849.0,
      "step": 24696
    },
    {
      "entropy": 0.4883555117994547,
      "epoch": 2.2722468530945767,
      "grad_norm": 0.1641780585050583,
      "learning_rate": 2.4261661606403537e-05,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8501360937952995,
      "num_tokens": 788216221.0,
      "step": 24697
    },
    {
      "entropy": 0.5125426761806011,
      "epoch": 2.272338859466018,
      "grad_norm": 0.16012997925281525,
      "learning_rate": 2.4258594780261908e-05,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8447576127946377,
      "num_tokens": 788247878.0,
      "step": 24698
    },
    {
      "entropy": 0.5247767297551036,
      "epoch": 2.2724308658374595,
      "grad_norm": 0.16091695427894592,
      "learning_rate": 2.4255527954120283e-05,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8375959806144238,
      "num_tokens": 788279658.0,
      "step": 24699
    },
    {
      "entropy": 0.4400638049701229,
      "epoch": 2.2725228722089006,
      "grad_norm": 0.15266135334968567,
      "learning_rate": 2.4252461127978657e-05,
      "loss": 0.4313,
      "mean_token_accuracy": 0.8658205829560757,
      "num_tokens": 788311716.0,
      "step": 24700
    },
    {
      "entropy": 0.6096321661025286,
      "epoch": 2.2726148785803417,
      "grad_norm": 0.16778643429279327,
      "learning_rate": 2.4249394301837032e-05,
      "loss": 0.6037,
      "mean_token_accuracy": 0.8109500482678413,
      "num_tokens": 788344256.0,
      "step": 24701
    },
    {
      "entropy": 0.5603882223367691,
      "epoch": 2.272706884951783,
      "grad_norm": 0.1633259505033493,
      "learning_rate": 2.4246327475695403e-05,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8333852253854275,
      "num_tokens": 788376567.0,
      "step": 24702
    },
    {
      "entropy": 0.6444603614509106,
      "epoch": 2.272798891323224,
      "grad_norm": 0.17787116765975952,
      "learning_rate": 2.4243260649553778e-05,
      "loss": 0.6214,
      "mean_token_accuracy": 0.8037006296217442,
      "num_tokens": 788408283.0,
      "step": 24703
    },
    {
      "entropy": 0.4969226629473269,
      "epoch": 2.2728908976946656,
      "grad_norm": 0.1567903757095337,
      "learning_rate": 2.4240193823412152e-05,
      "loss": 0.4748,
      "mean_token_accuracy": 0.8492328077554703,
      "num_tokens": 788440743.0,
      "step": 24704
    },
    {
      "entropy": 0.6229052357375622,
      "epoch": 2.2729829040661067,
      "grad_norm": 0.17287395894527435,
      "learning_rate": 2.4237126997270527e-05,
      "loss": 0.6072,
      "mean_token_accuracy": 0.8125834837555885,
      "num_tokens": 788472363.0,
      "step": 24705
    },
    {
      "entropy": 0.44665378192439675,
      "epoch": 2.273074910437548,
      "grad_norm": 0.15772517025470734,
      "learning_rate": 2.4234060171128902e-05,
      "loss": 0.4299,
      "mean_token_accuracy": 0.8594197817146778,
      "num_tokens": 788504153.0,
      "step": 24706
    },
    {
      "entropy": 0.6565308980643749,
      "epoch": 2.273166916808989,
      "grad_norm": 0.17874042689800262,
      "learning_rate": 2.4230993344987273e-05,
      "loss": 0.6389,
      "mean_token_accuracy": 0.8015518113970757,
      "num_tokens": 788536030.0,
      "step": 24707
    },
    {
      "entropy": 0.5427574710920453,
      "epoch": 2.27325892318043,
      "grad_norm": 0.16160371899604797,
      "learning_rate": 2.4227926518845648e-05,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8319182842969894,
      "num_tokens": 788568049.0,
      "step": 24708
    },
    {
      "entropy": 0.5847585126757622,
      "epoch": 2.2733509295518717,
      "grad_norm": 0.16986532509326935,
      "learning_rate": 2.4224859692704022e-05,
      "loss": 0.5738,
      "mean_token_accuracy": 0.8207962922751904,
      "num_tokens": 788599626.0,
      "step": 24709
    },
    {
      "entropy": 0.4527110527269542,
      "epoch": 2.273442935923313,
      "grad_norm": 0.15161454677581787,
      "learning_rate": 2.4221792866562397e-05,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8550333082675934,
      "num_tokens": 788631412.0,
      "step": 24710
    },
    {
      "entropy": 0.5533726699650288,
      "epoch": 2.273534942294754,
      "grad_norm": 0.16623330116271973,
      "learning_rate": 2.4218726040420768e-05,
      "loss": 0.5381,
      "mean_token_accuracy": 0.8308702409267426,
      "num_tokens": 788663245.0,
      "step": 24711
    },
    {
      "entropy": 0.4832653347402811,
      "epoch": 2.273626948666195,
      "grad_norm": 0.1588170826435089,
      "learning_rate": 2.4215659214279143e-05,
      "loss": 0.4805,
      "mean_token_accuracy": 0.8479926697909832,
      "num_tokens": 788695335.0,
      "step": 24712
    },
    {
      "entropy": 0.5890488885343075,
      "epoch": 2.273718955037636,
      "grad_norm": 0.16972137987613678,
      "learning_rate": 2.4212592388137517e-05,
      "loss": 0.5776,
      "mean_token_accuracy": 0.8181654699146748,
      "num_tokens": 788727212.0,
      "step": 24713
    },
    {
      "entropy": 0.6064128987491131,
      "epoch": 2.273810961409078,
      "grad_norm": 0.17372669279575348,
      "learning_rate": 2.4209525561995892e-05,
      "loss": 0.5893,
      "mean_token_accuracy": 0.8121421784162521,
      "num_tokens": 788758665.0,
      "step": 24714
    },
    {
      "entropy": 0.5868850825354457,
      "epoch": 2.273902967780519,
      "grad_norm": 0.1651320606470108,
      "learning_rate": 2.4206458735854263e-05,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8169956989586353,
      "num_tokens": 788790789.0,
      "step": 24715
    },
    {
      "entropy": 0.5372875593602657,
      "epoch": 2.27399497415196,
      "grad_norm": 0.1614305078983307,
      "learning_rate": 2.4203391909712638e-05,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8357325196266174,
      "num_tokens": 788822804.0,
      "step": 24716
    },
    {
      "entropy": 0.5737000126391649,
      "epoch": 2.274086980523401,
      "grad_norm": 0.1702767014503479,
      "learning_rate": 2.4200325083571013e-05,
      "loss": 0.5652,
      "mean_token_accuracy": 0.8221546299755573,
      "num_tokens": 788855030.0,
      "step": 24717
    },
    {
      "entropy": 0.5665901154279709,
      "epoch": 2.2741789868948423,
      "grad_norm": 0.1649961918592453,
      "learning_rate": 2.4197258257429387e-05,
      "loss": 0.5596,
      "mean_token_accuracy": 0.8204909041523933,
      "num_tokens": 788887595.0,
      "step": 24718
    },
    {
      "entropy": 0.5333996484987438,
      "epoch": 2.274270993266284,
      "grad_norm": 0.16159088909626007,
      "learning_rate": 2.4194191431287762e-05,
      "loss": 0.528,
      "mean_token_accuracy": 0.8345103226602077,
      "num_tokens": 788920044.0,
      "step": 24719
    },
    {
      "entropy": 0.45500810677185655,
      "epoch": 2.274362999637725,
      "grad_norm": 0.1548926830291748,
      "learning_rate": 2.4191124605146133e-05,
      "loss": 0.4424,
      "mean_token_accuracy": 0.8585934042930603,
      "num_tokens": 788951819.0,
      "step": 24720
    },
    {
      "entropy": 0.5390315186232328,
      "epoch": 2.274455006009166,
      "grad_norm": 0.17142854630947113,
      "learning_rate": 2.4188057779004508e-05,
      "loss": 0.5255,
      "mean_token_accuracy": 0.8310644924640656,
      "num_tokens": 788983543.0,
      "step": 24721
    },
    {
      "entropy": 0.5121290748938918,
      "epoch": 2.2745470123806073,
      "grad_norm": 0.1675376147031784,
      "learning_rate": 2.4184990952862883e-05,
      "loss": 0.5021,
      "mean_token_accuracy": 0.8380309157073498,
      "num_tokens": 789015061.0,
      "step": 24722
    },
    {
      "entropy": 0.5287897065281868,
      "epoch": 2.2746390187520484,
      "grad_norm": 0.16386616230010986,
      "learning_rate": 2.4181924126721257e-05,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8397675566375256,
      "num_tokens": 789046728.0,
      "step": 24723
    },
    {
      "entropy": 0.487330736592412,
      "epoch": 2.27473102512349,
      "grad_norm": 0.1576896458864212,
      "learning_rate": 2.4178857300579632e-05,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8465719856321812,
      "num_tokens": 789079056.0,
      "step": 24724
    },
    {
      "entropy": 0.4926240164786577,
      "epoch": 2.274823031494931,
      "grad_norm": 0.1581512987613678,
      "learning_rate": 2.4175790474438006e-05,
      "loss": 0.4761,
      "mean_token_accuracy": 0.850345067679882,
      "num_tokens": 789110515.0,
      "step": 24725
    },
    {
      "entropy": 0.5417800322175026,
      "epoch": 2.2749150378663723,
      "grad_norm": 0.16554337739944458,
      "learning_rate": 2.417272364829638e-05,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8316761329770088,
      "num_tokens": 789142966.0,
      "step": 24726
    },
    {
      "entropy": 0.5513838902115822,
      "epoch": 2.2750070442378134,
      "grad_norm": 0.1643492579460144,
      "learning_rate": 2.4169656822154756e-05,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8293828703463078,
      "num_tokens": 789174275.0,
      "step": 24727
    },
    {
      "entropy": 0.5727893030270934,
      "epoch": 2.2750990506092545,
      "grad_norm": 0.16941767930984497,
      "learning_rate": 2.4166589996013127e-05,
      "loss": 0.5689,
      "mean_token_accuracy": 0.8255870975553989,
      "num_tokens": 789206152.0,
      "step": 24728
    },
    {
      "entropy": 0.5110635901801288,
      "epoch": 2.275191056980696,
      "grad_norm": 0.15954247117042542,
      "learning_rate": 2.41635231698715e-05,
      "loss": 0.5005,
      "mean_token_accuracy": 0.8426895253360271,
      "num_tokens": 789237835.0,
      "step": 24729
    },
    {
      "entropy": 0.5338449489790946,
      "epoch": 2.2752830633521373,
      "grad_norm": 0.15949875116348267,
      "learning_rate": 2.4160456343729876e-05,
      "loss": 0.5143,
      "mean_token_accuracy": 0.8359921500086784,
      "num_tokens": 789269851.0,
      "step": 24730
    },
    {
      "entropy": 0.5003840615972877,
      "epoch": 2.2753750697235784,
      "grad_norm": 0.16494962573051453,
      "learning_rate": 2.415738951758825e-05,
      "loss": 0.4823,
      "mean_token_accuracy": 0.8481031954288483,
      "num_tokens": 789301173.0,
      "step": 24731
    },
    {
      "entropy": 0.5488938316702843,
      "epoch": 2.2754670760950195,
      "grad_norm": 0.1665111780166626,
      "learning_rate": 2.4154322691446622e-05,
      "loss": 0.5326,
      "mean_token_accuracy": 0.830007329583168,
      "num_tokens": 789333158.0,
      "step": 24732
    },
    {
      "entropy": 0.4559368006885052,
      "epoch": 2.2755590824664607,
      "grad_norm": 0.15650929510593414,
      "learning_rate": 2.4151255865304997e-05,
      "loss": 0.439,
      "mean_token_accuracy": 0.8592418991029263,
      "num_tokens": 789364568.0,
      "step": 24733
    },
    {
      "entropy": 0.5334206586703658,
      "epoch": 2.2756510888379022,
      "grad_norm": 0.16007061302661896,
      "learning_rate": 2.414818903916337e-05,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8377645574510098,
      "num_tokens": 789396758.0,
      "step": 24734
    },
    {
      "entropy": 0.45856356993317604,
      "epoch": 2.2757430952093434,
      "grad_norm": 0.15311352908611298,
      "learning_rate": 2.4145122213021746e-05,
      "loss": 0.4424,
      "mean_token_accuracy": 0.8617241084575653,
      "num_tokens": 789429052.0,
      "step": 24735
    },
    {
      "entropy": 0.5068367337808013,
      "epoch": 2.2758351015807845,
      "grad_norm": 0.16762423515319824,
      "learning_rate": 2.414205538688012e-05,
      "loss": 0.4951,
      "mean_token_accuracy": 0.8445855565369129,
      "num_tokens": 789459970.0,
      "step": 24736
    },
    {
      "entropy": 0.49439932126551867,
      "epoch": 2.2759271079522256,
      "grad_norm": 0.1557445526123047,
      "learning_rate": 2.4138988560738492e-05,
      "loss": 0.4835,
      "mean_token_accuracy": 0.8476398885250092,
      "num_tokens": 789492077.0,
      "step": 24737
    },
    {
      "entropy": 0.5180305419489741,
      "epoch": 2.2760191143236668,
      "grad_norm": 0.16046170890331268,
      "learning_rate": 2.4135921734596867e-05,
      "loss": 0.507,
      "mean_token_accuracy": 0.8393427021801472,
      "num_tokens": 789524329.0,
      "step": 24738
    },
    {
      "entropy": 0.5797810519579798,
      "epoch": 2.2761111206951083,
      "grad_norm": 0.16414493322372437,
      "learning_rate": 2.413285490845524e-05,
      "loss": 0.5605,
      "mean_token_accuracy": 0.8260746225714684,
      "num_tokens": 789556847.0,
      "step": 24739
    },
    {
      "entropy": 0.503440666012466,
      "epoch": 2.2762031270665495,
      "grad_norm": 0.16121579706668854,
      "learning_rate": 2.4129788082313616e-05,
      "loss": 0.4877,
      "mean_token_accuracy": 0.8484753705561161,
      "num_tokens": 789587864.0,
      "step": 24740
    },
    {
      "entropy": 0.5188147900626063,
      "epoch": 2.2762951334379906,
      "grad_norm": 0.16296526789665222,
      "learning_rate": 2.4126721256171987e-05,
      "loss": 0.5116,
      "mean_token_accuracy": 0.8380853310227394,
      "num_tokens": 789619956.0,
      "step": 24741
    },
    {
      "entropy": 0.5466853519901633,
      "epoch": 2.2763871398094317,
      "grad_norm": 0.16065578162670135,
      "learning_rate": 2.4123654430030362e-05,
      "loss": 0.5336,
      "mean_token_accuracy": 0.8291654586791992,
      "num_tokens": 789652388.0,
      "step": 24742
    },
    {
      "entropy": 0.6092647815821692,
      "epoch": 2.276479146180873,
      "grad_norm": 0.17093822360038757,
      "learning_rate": 2.4120587603888737e-05,
      "loss": 0.5961,
      "mean_token_accuracy": 0.8152580447494984,
      "num_tokens": 789684022.0,
      "step": 24743
    },
    {
      "entropy": 0.5746552180498838,
      "epoch": 2.2765711525523145,
      "grad_norm": 0.1667526811361313,
      "learning_rate": 2.411752077774711e-05,
      "loss": 0.5546,
      "mean_token_accuracy": 0.8260720558464527,
      "num_tokens": 789716098.0,
      "step": 24744
    },
    {
      "entropy": 0.46033391007222235,
      "epoch": 2.2766631589237556,
      "grad_norm": 0.1552361100912094,
      "learning_rate": 2.4114453951605482e-05,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8559731058776379,
      "num_tokens": 789748575.0,
      "step": 24745
    },
    {
      "entropy": 0.6050232211127877,
      "epoch": 2.2767551652951967,
      "grad_norm": 0.16557703912258148,
      "learning_rate": 2.4111387125463857e-05,
      "loss": 0.586,
      "mean_token_accuracy": 0.8158828467130661,
      "num_tokens": 789780691.0,
      "step": 24746
    },
    {
      "entropy": 0.5554639925248921,
      "epoch": 2.276847171666638,
      "grad_norm": 0.17069777846336365,
      "learning_rate": 2.4108320299322232e-05,
      "loss": 0.5452,
      "mean_token_accuracy": 0.8260688409209251,
      "num_tokens": 789811614.0,
      "step": 24747
    },
    {
      "entropy": 0.45241294940933585,
      "epoch": 2.276939178038079,
      "grad_norm": 0.15671518445014954,
      "learning_rate": 2.4105253473180606e-05,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8576531521975994,
      "num_tokens": 789843241.0,
      "step": 24748
    },
    {
      "entropy": 0.6374447923153639,
      "epoch": 2.2770311844095206,
      "grad_norm": 0.17125245928764343,
      "learning_rate": 2.410218664703898e-05,
      "loss": 0.624,
      "mean_token_accuracy": 0.8058015406131744,
      "num_tokens": 789875248.0,
      "step": 24749
    },
    {
      "entropy": 0.4396197374444455,
      "epoch": 2.2771231907809617,
      "grad_norm": 0.15776802599430084,
      "learning_rate": 2.4099119820897352e-05,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8612487055361271,
      "num_tokens": 789906645.0,
      "step": 24750
    },
    {
      "entropy": 0.5224968413822353,
      "epoch": 2.277215197152403,
      "grad_norm": 0.1611693948507309,
      "learning_rate": 2.4096052994755727e-05,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8376149721443653,
      "num_tokens": 789938939.0,
      "step": 24751
    },
    {
      "entropy": 0.5329053746536374,
      "epoch": 2.277307203523844,
      "grad_norm": 0.16787222027778625,
      "learning_rate": 2.40929861686141e-05,
      "loss": 0.5299,
      "mean_token_accuracy": 0.8352250680327415,
      "num_tokens": 789970384.0,
      "step": 24752
    },
    {
      "entropy": 0.49156288243830204,
      "epoch": 2.277399209895285,
      "grad_norm": 0.15600453317165375,
      "learning_rate": 2.4089919342472476e-05,
      "loss": 0.4836,
      "mean_token_accuracy": 0.8450821042060852,
      "num_tokens": 790002965.0,
      "step": 24753
    },
    {
      "entropy": 0.5280759800225496,
      "epoch": 2.2774912162667267,
      "grad_norm": 0.16500382125377655,
      "learning_rate": 2.408685251633085e-05,
      "loss": 0.5241,
      "mean_token_accuracy": 0.835914671421051,
      "num_tokens": 790035205.0,
      "step": 24754
    },
    {
      "entropy": 0.5171877667307854,
      "epoch": 2.277583222638168,
      "grad_norm": 0.15985018014907837,
      "learning_rate": 2.4083785690189226e-05,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8441400676965714,
      "num_tokens": 790067535.0,
      "step": 24755
    },
    {
      "entropy": 0.5208772104233503,
      "epoch": 2.277675229009609,
      "grad_norm": 0.15967293083667755,
      "learning_rate": 2.40807188640476e-05,
      "loss": 0.4992,
      "mean_token_accuracy": 0.8392755836248398,
      "num_tokens": 790099764.0,
      "step": 24756
    },
    {
      "entropy": 0.5466566616669297,
      "epoch": 2.27776723538105,
      "grad_norm": 0.16428852081298828,
      "learning_rate": 2.4077652037905975e-05,
      "loss": 0.5377,
      "mean_token_accuracy": 0.8301212526857853,
      "num_tokens": 790131597.0,
      "step": 24757
    },
    {
      "entropy": 0.4885553941130638,
      "epoch": 2.277859241752491,
      "grad_norm": 0.15445972979068756,
      "learning_rate": 2.4074585211764346e-05,
      "loss": 0.474,
      "mean_token_accuracy": 0.8486098796129227,
      "num_tokens": 790163316.0,
      "step": 24758
    },
    {
      "entropy": 0.5435596145689487,
      "epoch": 2.277951248123933,
      "grad_norm": 0.16426284611225128,
      "learning_rate": 2.407151838562272e-05,
      "loss": 0.5336,
      "mean_token_accuracy": 0.8303561769425869,
      "num_tokens": 790195842.0,
      "step": 24759
    },
    {
      "entropy": 0.4474881496280432,
      "epoch": 2.278043254495374,
      "grad_norm": 0.15952454507350922,
      "learning_rate": 2.4068451559481095e-05,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8634869270026684,
      "num_tokens": 790227849.0,
      "step": 24760
    },
    {
      "entropy": 0.4621778232976794,
      "epoch": 2.278135260866815,
      "grad_norm": 0.15532535314559937,
      "learning_rate": 2.406538473333947e-05,
      "loss": 0.4431,
      "mean_token_accuracy": 0.8566263169050217,
      "num_tokens": 790259693.0,
      "step": 24761
    },
    {
      "entropy": 0.5648450311273336,
      "epoch": 2.278227267238256,
      "grad_norm": 0.16556328535079956,
      "learning_rate": 2.406231790719784e-05,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8273113965988159,
      "num_tokens": 790291645.0,
      "step": 24762
    },
    {
      "entropy": 0.4686594791710377,
      "epoch": 2.2783192736096973,
      "grad_norm": 0.1569826602935791,
      "learning_rate": 2.4059251081056216e-05,
      "loss": 0.4559,
      "mean_token_accuracy": 0.8555727526545525,
      "num_tokens": 790323926.0,
      "step": 24763
    },
    {
      "entropy": 0.4988693450577557,
      "epoch": 2.278411279981139,
      "grad_norm": 0.15972113609313965,
      "learning_rate": 2.405618425491459e-05,
      "loss": 0.4846,
      "mean_token_accuracy": 0.8492051288485527,
      "num_tokens": 790355736.0,
      "step": 24764
    },
    {
      "entropy": 0.49670882103964686,
      "epoch": 2.27850328635258,
      "grad_norm": 0.1559053212404251,
      "learning_rate": 2.4053117428772965e-05,
      "loss": 0.4813,
      "mean_token_accuracy": 0.8465516157448292,
      "num_tokens": 790387559.0,
      "step": 24765
    },
    {
      "entropy": 0.6653489135205746,
      "epoch": 2.278595292724021,
      "grad_norm": 0.17356657981872559,
      "learning_rate": 2.405005060263134e-05,
      "loss": 0.6491,
      "mean_token_accuracy": 0.7951400429010391,
      "num_tokens": 790419610.0,
      "step": 24766
    },
    {
      "entropy": 0.499685381539166,
      "epoch": 2.2786872990954623,
      "grad_norm": 0.1582302302122116,
      "learning_rate": 2.404698377648971e-05,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8445202670991421,
      "num_tokens": 790450953.0,
      "step": 24767
    },
    {
      "entropy": 0.5271407812833786,
      "epoch": 2.2787793054669034,
      "grad_norm": 0.15532459318637848,
      "learning_rate": 2.4043916950348086e-05,
      "loss": 0.5072,
      "mean_token_accuracy": 0.840470477938652,
      "num_tokens": 790483279.0,
      "step": 24768
    },
    {
      "entropy": 0.6115701561793685,
      "epoch": 2.278871311838345,
      "grad_norm": 0.17567729949951172,
      "learning_rate": 2.404085012420646e-05,
      "loss": 0.6003,
      "mean_token_accuracy": 0.8132058680057526,
      "num_tokens": 790515212.0,
      "step": 24769
    },
    {
      "entropy": 0.5238954108208418,
      "epoch": 2.278963318209786,
      "grad_norm": 0.15969489514827728,
      "learning_rate": 2.4037783298064835e-05,
      "loss": 0.5243,
      "mean_token_accuracy": 0.8358677104115486,
      "num_tokens": 790547732.0,
      "step": 24770
    },
    {
      "entropy": 0.486124612390995,
      "epoch": 2.2790553245812273,
      "grad_norm": 0.15560205280780792,
      "learning_rate": 2.4034716471923206e-05,
      "loss": 0.4804,
      "mean_token_accuracy": 0.8488080352544785,
      "num_tokens": 790579984.0,
      "step": 24771
    },
    {
      "entropy": 0.5430626263841987,
      "epoch": 2.2791473309526684,
      "grad_norm": 0.1680562049150467,
      "learning_rate": 2.403164964578158e-05,
      "loss": 0.52,
      "mean_token_accuracy": 0.8333760276436806,
      "num_tokens": 790611844.0,
      "step": 24772
    },
    {
      "entropy": 0.45480002649128437,
      "epoch": 2.2792393373241095,
      "grad_norm": 0.14967656135559082,
      "learning_rate": 2.4028582819639956e-05,
      "loss": 0.4454,
      "mean_token_accuracy": 0.8596851043403149,
      "num_tokens": 790643691.0,
      "step": 24773
    },
    {
      "entropy": 0.5036928560584784,
      "epoch": 2.279331343695551,
      "grad_norm": 0.1573992371559143,
      "learning_rate": 2.402551599349833e-05,
      "loss": 0.4938,
      "mean_token_accuracy": 0.8434091471135616,
      "num_tokens": 790675437.0,
      "step": 24774
    },
    {
      "entropy": 0.5454310532659292,
      "epoch": 2.2794233500669923,
      "grad_norm": 0.1634073406457901,
      "learning_rate": 2.40224491673567e-05,
      "loss": 0.5346,
      "mean_token_accuracy": 0.8286334723234177,
      "num_tokens": 790707272.0,
      "step": 24775
    },
    {
      "entropy": 0.5640181768685579,
      "epoch": 2.2795153564384334,
      "grad_norm": 0.16127923130989075,
      "learning_rate": 2.4019382341215076e-05,
      "loss": 0.5543,
      "mean_token_accuracy": 0.8260014690458775,
      "num_tokens": 790739920.0,
      "step": 24776
    },
    {
      "entropy": 0.49141199234873056,
      "epoch": 2.2796073628098745,
      "grad_norm": 0.1588038057088852,
      "learning_rate": 2.401631551507345e-05,
      "loss": 0.4785,
      "mean_token_accuracy": 0.846334557980299,
      "num_tokens": 790771445.0,
      "step": 24777
    },
    {
      "entropy": 0.540071171708405,
      "epoch": 2.2796993691813157,
      "grad_norm": 0.1616222858428955,
      "learning_rate": 2.4013248688931826e-05,
      "loss": 0.5287,
      "mean_token_accuracy": 0.837671410292387,
      "num_tokens": 790803267.0,
      "step": 24778
    },
    {
      "entropy": 0.5126354503445327,
      "epoch": 2.2797913755527572,
      "grad_norm": 0.16327165067195892,
      "learning_rate": 2.40101818627902e-05,
      "loss": 0.5038,
      "mean_token_accuracy": 0.8397195860743523,
      "num_tokens": 790834864.0,
      "step": 24779
    },
    {
      "entropy": 0.4558897474780679,
      "epoch": 2.2798833819241984,
      "grad_norm": 0.15951228141784668,
      "learning_rate": 2.400711503664857e-05,
      "loss": 0.446,
      "mean_token_accuracy": 0.8582405410706997,
      "num_tokens": 790867111.0,
      "step": 24780
    },
    {
      "entropy": 0.5208575315773487,
      "epoch": 2.2799753882956395,
      "grad_norm": 0.15898394584655762,
      "learning_rate": 2.4004048210506946e-05,
      "loss": 0.5083,
      "mean_token_accuracy": 0.8396087884902954,
      "num_tokens": 790899216.0,
      "step": 24781
    },
    {
      "entropy": 0.48636922985315323,
      "epoch": 2.2800673946670806,
      "grad_norm": 0.15734192728996277,
      "learning_rate": 2.400098138436532e-05,
      "loss": 0.467,
      "mean_token_accuracy": 0.8505772165954113,
      "num_tokens": 790930047.0,
      "step": 24782
    },
    {
      "entropy": 0.46577146649360657,
      "epoch": 2.2801594010385218,
      "grad_norm": 0.15601052343845367,
      "learning_rate": 2.3997914558223695e-05,
      "loss": 0.4437,
      "mean_token_accuracy": 0.8564532808959484,
      "num_tokens": 790961935.0,
      "step": 24783
    },
    {
      "entropy": 0.37171759037300944,
      "epoch": 2.2802514074099633,
      "grad_norm": 0.13959459960460663,
      "learning_rate": 2.3994847732082067e-05,
      "loss": 0.3566,
      "mean_token_accuracy": 0.8843108080327511,
      "num_tokens": 790993647.0,
      "step": 24784
    },
    {
      "entropy": 0.4899860490113497,
      "epoch": 2.2803434137814045,
      "grad_norm": 0.16106028854846954,
      "learning_rate": 2.3991780905940445e-05,
      "loss": 0.4797,
      "mean_token_accuracy": 0.8459254540503025,
      "num_tokens": 791025423.0,
      "step": 24785
    },
    {
      "entropy": 0.5589744085446,
      "epoch": 2.2804354201528456,
      "grad_norm": 0.1688041090965271,
      "learning_rate": 2.398871407979882e-05,
      "loss": 0.548,
      "mean_token_accuracy": 0.8261574544012547,
      "num_tokens": 791056446.0,
      "step": 24786
    },
    {
      "entropy": 0.5728647038340569,
      "epoch": 2.2805274265242867,
      "grad_norm": 0.16749505698680878,
      "learning_rate": 2.3985647253657194e-05,
      "loss": 0.5469,
      "mean_token_accuracy": 0.8277237340807915,
      "num_tokens": 791088576.0,
      "step": 24787
    },
    {
      "entropy": 0.5854055266827345,
      "epoch": 2.280619432895728,
      "grad_norm": 0.17314982414245605,
      "learning_rate": 2.3982580427515565e-05,
      "loss": 0.5755,
      "mean_token_accuracy": 0.8199371434748173,
      "num_tokens": 791120047.0,
      "step": 24788
    },
    {
      "entropy": 0.5484726140275598,
      "epoch": 2.2807114392671695,
      "grad_norm": 0.16455323994159698,
      "learning_rate": 2.397951360137394e-05,
      "loss": 0.5382,
      "mean_token_accuracy": 0.8308266848325729,
      "num_tokens": 791151700.0,
      "step": 24789
    },
    {
      "entropy": 0.5946166310459375,
      "epoch": 2.2808034456386106,
      "grad_norm": 0.16933166980743408,
      "learning_rate": 2.3976446775232315e-05,
      "loss": 0.5806,
      "mean_token_accuracy": 0.8195050582289696,
      "num_tokens": 791183795.0,
      "step": 24790
    },
    {
      "entropy": 0.46106779482215643,
      "epoch": 2.2808954520100517,
      "grad_norm": 0.154572993516922,
      "learning_rate": 2.397337994909069e-05,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8560198359191418,
      "num_tokens": 791215946.0,
      "step": 24791
    },
    {
      "entropy": 0.5518293082714081,
      "epoch": 2.280987458381493,
      "grad_norm": 0.1669662743806839,
      "learning_rate": 2.397031312294906e-05,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8271403722465038,
      "num_tokens": 791247698.0,
      "step": 24792
    },
    {
      "entropy": 0.4622586041223258,
      "epoch": 2.281079464752934,
      "grad_norm": 0.1598619669675827,
      "learning_rate": 2.3967246296807435e-05,
      "loss": 0.4473,
      "mean_token_accuracy": 0.8568647652864456,
      "num_tokens": 791279556.0,
      "step": 24793
    },
    {
      "entropy": 0.5448378091678023,
      "epoch": 2.2811714711243756,
      "grad_norm": 0.1617441475391388,
      "learning_rate": 2.396417947066581e-05,
      "loss": 0.5341,
      "mean_token_accuracy": 0.8339422158896923,
      "num_tokens": 791311405.0,
      "step": 24794
    },
    {
      "entropy": 0.4678119895979762,
      "epoch": 2.2812634774958167,
      "grad_norm": 0.15835782885551453,
      "learning_rate": 2.3961112644524184e-05,
      "loss": 0.4644,
      "mean_token_accuracy": 0.8502156436443329,
      "num_tokens": 791343025.0,
      "step": 24795
    },
    {
      "entropy": 0.5538558587431908,
      "epoch": 2.281355483867258,
      "grad_norm": 0.17029927670955658,
      "learning_rate": 2.395804581838256e-05,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8252853453159332,
      "num_tokens": 791373791.0,
      "step": 24796
    },
    {
      "entropy": 0.510709134163335,
      "epoch": 2.281447490238699,
      "grad_norm": 0.15592913329601288,
      "learning_rate": 2.395497899224093e-05,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8452670387923717,
      "num_tokens": 791405693.0,
      "step": 24797
    },
    {
      "entropy": 0.4610164910554886,
      "epoch": 2.28153949661014,
      "grad_norm": 0.15347999334335327,
      "learning_rate": 2.3951912166099305e-05,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8577833510935307,
      "num_tokens": 791437528.0,
      "step": 24798
    },
    {
      "entropy": 0.46515609370544553,
      "epoch": 2.2816315029815817,
      "grad_norm": 0.15384450554847717,
      "learning_rate": 2.394884533995768e-05,
      "loss": 0.4547,
      "mean_token_accuracy": 0.8565889075398445,
      "num_tokens": 791470296.0,
      "step": 24799
    },
    {
      "entropy": 0.6078530941158533,
      "epoch": 2.281723509353023,
      "grad_norm": 0.1732209026813507,
      "learning_rate": 2.3945778513816054e-05,
      "loss": 0.6064,
      "mean_token_accuracy": 0.8115341551601887,
      "num_tokens": 791502254.0,
      "step": 24800
    },
    {
      "entropy": 0.5344701362773776,
      "epoch": 2.281815515724464,
      "grad_norm": 0.1649043709039688,
      "learning_rate": 2.3942711687674425e-05,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8312117159366608,
      "num_tokens": 791533985.0,
      "step": 24801
    },
    {
      "entropy": 0.6104875914752483,
      "epoch": 2.281907522095905,
      "grad_norm": 0.17123457789421082,
      "learning_rate": 2.39396448615328e-05,
      "loss": 0.5931,
      "mean_token_accuracy": 0.816267341375351,
      "num_tokens": 791564956.0,
      "step": 24802
    },
    {
      "entropy": 0.4413899378851056,
      "epoch": 2.281999528467346,
      "grad_norm": 0.15201131999492645,
      "learning_rate": 2.3936578035391175e-05,
      "loss": 0.4288,
      "mean_token_accuracy": 0.8614715933799744,
      "num_tokens": 791596963.0,
      "step": 24803
    },
    {
      "entropy": 0.5045588286593556,
      "epoch": 2.282091534838788,
      "grad_norm": 0.15923380851745605,
      "learning_rate": 2.393351120924955e-05,
      "loss": 0.4875,
      "mean_token_accuracy": 0.8449950590729713,
      "num_tokens": 791629059.0,
      "step": 24804
    },
    {
      "entropy": 0.47442026529461145,
      "epoch": 2.282183541210229,
      "grad_norm": 0.15318168699741364,
      "learning_rate": 2.393044438310792e-05,
      "loss": 0.4562,
      "mean_token_accuracy": 0.8504163809120655,
      "num_tokens": 791660510.0,
      "step": 24805
    },
    {
      "entropy": 0.4882506625726819,
      "epoch": 2.28227554758167,
      "grad_norm": 0.1615293025970459,
      "learning_rate": 2.3927377556966295e-05,
      "loss": 0.4758,
      "mean_token_accuracy": 0.8465575687587261,
      "num_tokens": 791692490.0,
      "step": 24806
    },
    {
      "entropy": 0.551159480586648,
      "epoch": 2.282367553953111,
      "grad_norm": 0.16946566104888916,
      "learning_rate": 2.392431073082467e-05,
      "loss": 0.5323,
      "mean_token_accuracy": 0.8313930593430996,
      "num_tokens": 791724267.0,
      "step": 24807
    },
    {
      "entropy": 0.5299145970493555,
      "epoch": 2.2824595603245523,
      "grad_norm": 0.16257037222385406,
      "learning_rate": 2.3921243904683045e-05,
      "loss": 0.5137,
      "mean_token_accuracy": 0.8339723274111748,
      "num_tokens": 791755991.0,
      "step": 24808
    },
    {
      "entropy": 0.4500005654990673,
      "epoch": 2.282551566695994,
      "grad_norm": 0.15196345746517181,
      "learning_rate": 2.391817707854142e-05,
      "loss": 0.4493,
      "mean_token_accuracy": 0.858436830341816,
      "num_tokens": 791788286.0,
      "step": 24809
    },
    {
      "entropy": 0.5027348008006811,
      "epoch": 2.282643573067435,
      "grad_norm": 0.1600855439901352,
      "learning_rate": 2.391511025239979e-05,
      "loss": 0.4921,
      "mean_token_accuracy": 0.8421994857490063,
      "num_tokens": 791819985.0,
      "step": 24810
    },
    {
      "entropy": 0.5297353584319353,
      "epoch": 2.282735579438876,
      "grad_norm": 0.16726653277873993,
      "learning_rate": 2.3912043426258165e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8369172066450119,
      "num_tokens": 791851793.0,
      "step": 24811
    },
    {
      "entropy": 0.42849821876734495,
      "epoch": 2.2828275858103173,
      "grad_norm": 0.151315838098526,
      "learning_rate": 2.390897660011654e-05,
      "loss": 0.421,
      "mean_token_accuracy": 0.8652885779738426,
      "num_tokens": 791883999.0,
      "step": 24812
    },
    {
      "entropy": 0.5561282448470592,
      "epoch": 2.2829195921817584,
      "grad_norm": 0.16204121708869934,
      "learning_rate": 2.3905909773974914e-05,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8308124803006649,
      "num_tokens": 791915578.0,
      "step": 24813
    },
    {
      "entropy": 0.4788082279264927,
      "epoch": 2.2830115985532,
      "grad_norm": 0.1507711261510849,
      "learning_rate": 2.3902842947833286e-05,
      "loss": 0.4598,
      "mean_token_accuracy": 0.8515395745635033,
      "num_tokens": 791947904.0,
      "step": 24814
    },
    {
      "entropy": 0.4490137444809079,
      "epoch": 2.283103604924641,
      "grad_norm": 0.15423455834388733,
      "learning_rate": 2.3899776121691664e-05,
      "loss": 0.4315,
      "mean_token_accuracy": 0.858571495860815,
      "num_tokens": 791979334.0,
      "step": 24815
    },
    {
      "entropy": 0.43072577752172947,
      "epoch": 2.2831956112960823,
      "grad_norm": 0.14879333972930908,
      "learning_rate": 2.389670929555004e-05,
      "loss": 0.4172,
      "mean_token_accuracy": 0.8696215786039829,
      "num_tokens": 792011399.0,
      "step": 24816
    },
    {
      "entropy": 0.4295141980983317,
      "epoch": 2.2832876176675234,
      "grad_norm": 0.1480940878391266,
      "learning_rate": 2.3893642469408413e-05,
      "loss": 0.4162,
      "mean_token_accuracy": 0.8686994016170502,
      "num_tokens": 792043187.0,
      "step": 24817
    },
    {
      "entropy": 0.4499346436932683,
      "epoch": 2.2833796240389646,
      "grad_norm": 0.14978231489658356,
      "learning_rate": 2.3890575643266784e-05,
      "loss": 0.429,
      "mean_token_accuracy": 0.8648983612656593,
      "num_tokens": 792075536.0,
      "step": 24818
    },
    {
      "entropy": 0.5355925555340946,
      "epoch": 2.283471630410406,
      "grad_norm": 0.15664741396903992,
      "learning_rate": 2.388750881712516e-05,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8355017006397247,
      "num_tokens": 792107341.0,
      "step": 24819
    },
    {
      "entropy": 0.5331519423052669,
      "epoch": 2.2835636367818473,
      "grad_norm": 0.1627216339111328,
      "learning_rate": 2.3884441990983534e-05,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8328422345221043,
      "num_tokens": 792139357.0,
      "step": 24820
    },
    {
      "entropy": 0.5187115352600813,
      "epoch": 2.2836556431532884,
      "grad_norm": 0.16298623383045197,
      "learning_rate": 2.3881375164841908e-05,
      "loss": 0.5098,
      "mean_token_accuracy": 0.8378810435533524,
      "num_tokens": 792170227.0,
      "step": 24821
    },
    {
      "entropy": 0.4249349208548665,
      "epoch": 2.2837476495247295,
      "grad_norm": 0.15532466769218445,
      "learning_rate": 2.387830833870028e-05,
      "loss": 0.4046,
      "mean_token_accuracy": 0.8710414357483387,
      "num_tokens": 792202293.0,
      "step": 24822
    },
    {
      "entropy": 0.5577663253061473,
      "epoch": 2.2838396558961707,
      "grad_norm": 0.17101067304611206,
      "learning_rate": 2.3875241512558654e-05,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8272484838962555,
      "num_tokens": 792234446.0,
      "step": 24823
    },
    {
      "entropy": 0.5460330191999674,
      "epoch": 2.2839316622676122,
      "grad_norm": 0.17109671235084534,
      "learning_rate": 2.387217468641703e-05,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8329455070197582,
      "num_tokens": 792266331.0,
      "step": 24824
    },
    {
      "entropy": 0.47528935014270246,
      "epoch": 2.2840236686390534,
      "grad_norm": 0.1649189293384552,
      "learning_rate": 2.3869107860275403e-05,
      "loss": 0.4728,
      "mean_token_accuracy": 0.855821929872036,
      "num_tokens": 792297812.0,
      "step": 24825
    },
    {
      "entropy": 0.5040144175291061,
      "epoch": 2.2841156750104945,
      "grad_norm": 0.15947218239307404,
      "learning_rate": 2.3866041034133778e-05,
      "loss": 0.492,
      "mean_token_accuracy": 0.8391249477863312,
      "num_tokens": 792330003.0,
      "step": 24826
    },
    {
      "entropy": 0.5749142458662391,
      "epoch": 2.2842076813819356,
      "grad_norm": 0.16964176297187805,
      "learning_rate": 2.386297420799215e-05,
      "loss": 0.5673,
      "mean_token_accuracy": 0.8201099820435047,
      "num_tokens": 792362261.0,
      "step": 24827
    },
    {
      "entropy": 0.5782619612291455,
      "epoch": 2.2842996877533768,
      "grad_norm": 0.1663077175617218,
      "learning_rate": 2.3859907381850524e-05,
      "loss": 0.5742,
      "mean_token_accuracy": 0.818776935338974,
      "num_tokens": 792394430.0,
      "step": 24828
    },
    {
      "entropy": 0.4593889620155096,
      "epoch": 2.2843916941248184,
      "grad_norm": 0.1477755904197693,
      "learning_rate": 2.38568405557089e-05,
      "loss": 0.4413,
      "mean_token_accuracy": 0.8597980812191963,
      "num_tokens": 792426810.0,
      "step": 24829
    },
    {
      "entropy": 0.4442543718032539,
      "epoch": 2.2844837004962595,
      "grad_norm": 0.14572575688362122,
      "learning_rate": 2.3853773729567273e-05,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8618288040161133,
      "num_tokens": 792459497.0,
      "step": 24830
    },
    {
      "entropy": 0.5595119148492813,
      "epoch": 2.2845757068677006,
      "grad_norm": 0.16482427716255188,
      "learning_rate": 2.3850706903425645e-05,
      "loss": 0.5294,
      "mean_token_accuracy": 0.8319618664681911,
      "num_tokens": 792491225.0,
      "step": 24831
    },
    {
      "entropy": 0.5267618531361222,
      "epoch": 2.2846677132391418,
      "grad_norm": 0.1620045006275177,
      "learning_rate": 2.384764007728402e-05,
      "loss": 0.513,
      "mean_token_accuracy": 0.8392653577029705,
      "num_tokens": 792523468.0,
      "step": 24832
    },
    {
      "entropy": 0.5546950157731771,
      "epoch": 2.284759719610583,
      "grad_norm": 0.1670035868883133,
      "learning_rate": 2.3844573251142394e-05,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8272733986377716,
      "num_tokens": 792555101.0,
      "step": 24833
    },
    {
      "entropy": 0.49882730562239885,
      "epoch": 2.2848517259820245,
      "grad_norm": 0.16393402218818665,
      "learning_rate": 2.384150642500077e-05,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8480370156466961,
      "num_tokens": 792586935.0,
      "step": 24834
    },
    {
      "entropy": 0.5304940789937973,
      "epoch": 2.2849437323534656,
      "grad_norm": 0.16163566708564758,
      "learning_rate": 2.383843959885914e-05,
      "loss": 0.518,
      "mean_token_accuracy": 0.8351639471948147,
      "num_tokens": 792618867.0,
      "step": 24835
    },
    {
      "entropy": 0.5405516214668751,
      "epoch": 2.2850357387249067,
      "grad_norm": 0.1640494465827942,
      "learning_rate": 2.3835372772717514e-05,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8326949253678322,
      "num_tokens": 792651123.0,
      "step": 24836
    },
    {
      "entropy": 0.5379005651921034,
      "epoch": 2.285127745096348,
      "grad_norm": 0.16129571199417114,
      "learning_rate": 2.383230594657589e-05,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8378556668758392,
      "num_tokens": 792683058.0,
      "step": 24837
    },
    {
      "entropy": 0.5988846025429666,
      "epoch": 2.285219751467789,
      "grad_norm": 0.1678108274936676,
      "learning_rate": 2.3829239120434264e-05,
      "loss": 0.5951,
      "mean_token_accuracy": 0.8139962330460548,
      "num_tokens": 792714637.0,
      "step": 24838
    },
    {
      "entropy": 0.567249515093863,
      "epoch": 2.2853117578392306,
      "grad_norm": 0.16749542951583862,
      "learning_rate": 2.382617229429264e-05,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8248436078429222,
      "num_tokens": 792746782.0,
      "step": 24839
    },
    {
      "entropy": 0.5368151385337114,
      "epoch": 2.2854037642106717,
      "grad_norm": 0.16476984322071075,
      "learning_rate": 2.382310546815101e-05,
      "loss": 0.5176,
      "mean_token_accuracy": 0.8377166353166103,
      "num_tokens": 792777314.0,
      "step": 24840
    },
    {
      "entropy": 0.5799811342731118,
      "epoch": 2.285495770582113,
      "grad_norm": 0.1655520349740982,
      "learning_rate": 2.3820038642009384e-05,
      "loss": 0.5749,
      "mean_token_accuracy": 0.8206891976296902,
      "num_tokens": 792808672.0,
      "step": 24841
    },
    {
      "entropy": 0.4866250157356262,
      "epoch": 2.285587776953554,
      "grad_norm": 0.1513582170009613,
      "learning_rate": 2.381697181586776e-05,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8500857129693031,
      "num_tokens": 792841388.0,
      "step": 24842
    },
    {
      "entropy": 0.4979741391725838,
      "epoch": 2.285679783324995,
      "grad_norm": 0.15846651792526245,
      "learning_rate": 2.3813904989726134e-05,
      "loss": 0.4841,
      "mean_token_accuracy": 0.8451021537184715,
      "num_tokens": 792872818.0,
      "step": 24843
    },
    {
      "entropy": 0.49096371699124575,
      "epoch": 2.2857717896964367,
      "grad_norm": 0.16009491682052612,
      "learning_rate": 2.3810838163584505e-05,
      "loss": 0.4749,
      "mean_token_accuracy": 0.8484003469347954,
      "num_tokens": 792905337.0,
      "step": 24844
    },
    {
      "entropy": 0.45691578835248947,
      "epoch": 2.285863796067878,
      "grad_norm": 0.14683660864830017,
      "learning_rate": 2.380777133744288e-05,
      "loss": 0.4482,
      "mean_token_accuracy": 0.8558922633528709,
      "num_tokens": 792937819.0,
      "step": 24845
    },
    {
      "entropy": 0.4566968632861972,
      "epoch": 2.285955802439319,
      "grad_norm": 0.16040609776973724,
      "learning_rate": 2.3804704511301258e-05,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8533603362739086,
      "num_tokens": 792969808.0,
      "step": 24846
    },
    {
      "entropy": 0.5422802092507482,
      "epoch": 2.28604780881076,
      "grad_norm": 0.1609673649072647,
      "learning_rate": 2.3801637685159632e-05,
      "loss": 0.5216,
      "mean_token_accuracy": 0.8359594866633415,
      "num_tokens": 793001119.0,
      "step": 24847
    },
    {
      "entropy": 0.436546198092401,
      "epoch": 2.286139815182201,
      "grad_norm": 0.15388156473636627,
      "learning_rate": 2.3798570859018003e-05,
      "loss": 0.4284,
      "mean_token_accuracy": 0.8619458042085171,
      "num_tokens": 793033222.0,
      "step": 24848
    },
    {
      "entropy": 0.5969141526147723,
      "epoch": 2.286231821553643,
      "grad_norm": 0.16694167256355286,
      "learning_rate": 2.3795504032876378e-05,
      "loss": 0.5734,
      "mean_token_accuracy": 0.8169064782559872,
      "num_tokens": 793064889.0,
      "step": 24849
    },
    {
      "entropy": 0.4826177656650543,
      "epoch": 2.286323827925084,
      "grad_norm": 0.15813980996608734,
      "learning_rate": 2.3792437206734753e-05,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8459763564169407,
      "num_tokens": 793095831.0,
      "step": 24850
    },
    {
      "entropy": 0.6163155829999596,
      "epoch": 2.286415834296525,
      "grad_norm": 0.17658652365207672,
      "learning_rate": 2.3789370380593127e-05,
      "loss": 0.6057,
      "mean_token_accuracy": 0.8145327270030975,
      "num_tokens": 793127541.0,
      "step": 24851
    },
    {
      "entropy": 0.48891609720885754,
      "epoch": 2.286507840667966,
      "grad_norm": 0.15677377581596375,
      "learning_rate": 2.37863035544515e-05,
      "loss": 0.4769,
      "mean_token_accuracy": 0.8470417633652687,
      "num_tokens": 793160148.0,
      "step": 24852
    },
    {
      "entropy": 0.5009251795709133,
      "epoch": 2.2865998470394073,
      "grad_norm": 0.15830300748348236,
      "learning_rate": 2.3783236728309873e-05,
      "loss": 0.4954,
      "mean_token_accuracy": 0.8442185930907726,
      "num_tokens": 793192624.0,
      "step": 24853
    },
    {
      "entropy": 0.4645731821656227,
      "epoch": 2.286691853410849,
      "grad_norm": 0.15075166523456573,
      "learning_rate": 2.3780169902168248e-05,
      "loss": 0.4506,
      "mean_token_accuracy": 0.8552990257740021,
      "num_tokens": 793224860.0,
      "step": 24854
    },
    {
      "entropy": 0.562318854033947,
      "epoch": 2.28678385978229,
      "grad_norm": 0.16468562185764313,
      "learning_rate": 2.3777103076026623e-05,
      "loss": 0.5578,
      "mean_token_accuracy": 0.8265339769423008,
      "num_tokens": 793257325.0,
      "step": 24855
    },
    {
      "entropy": 0.5233801235444844,
      "epoch": 2.286875866153731,
      "grad_norm": 0.16498275101184845,
      "learning_rate": 2.3774036249884997e-05,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8422559536993504,
      "num_tokens": 793288865.0,
      "step": 24856
    },
    {
      "entropy": 0.39774706959724426,
      "epoch": 2.2869678725251723,
      "grad_norm": 0.14888112246990204,
      "learning_rate": 2.377096942374337e-05,
      "loss": 0.3871,
      "mean_token_accuracy": 0.8776350058615208,
      "num_tokens": 793321448.0,
      "step": 24857
    },
    {
      "entropy": 0.48350213887169957,
      "epoch": 2.2870598788966134,
      "grad_norm": 0.1567506343126297,
      "learning_rate": 2.3767902597601743e-05,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8544227033853531,
      "num_tokens": 793353179.0,
      "step": 24858
    },
    {
      "entropy": 0.5305229439400136,
      "epoch": 2.287151885268055,
      "grad_norm": 0.15970700979232788,
      "learning_rate": 2.3764835771460118e-05,
      "loss": 0.5198,
      "mean_token_accuracy": 0.8407006859779358,
      "num_tokens": 793385731.0,
      "step": 24859
    },
    {
      "entropy": 0.5660337107256055,
      "epoch": 2.287243891639496,
      "grad_norm": 0.1712346225976944,
      "learning_rate": 2.3761768945318492e-05,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8273665979504585,
      "num_tokens": 793417088.0,
      "step": 24860
    },
    {
      "entropy": 0.55980743188411,
      "epoch": 2.2873358980109373,
      "grad_norm": 0.16241925954818726,
      "learning_rate": 2.3758702119176864e-05,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8272862732410431,
      "num_tokens": 793448938.0,
      "step": 24861
    },
    {
      "entropy": 0.5008371602743864,
      "epoch": 2.2874279043823784,
      "grad_norm": 0.1607091873884201,
      "learning_rate": 2.375563529303524e-05,
      "loss": 0.488,
      "mean_token_accuracy": 0.8471826799213886,
      "num_tokens": 793480849.0,
      "step": 24862
    },
    {
      "entropy": 0.5121421338990331,
      "epoch": 2.2875199107538196,
      "grad_norm": 0.1599029153585434,
      "learning_rate": 2.3752568466893613e-05,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8410870581865311,
      "num_tokens": 793513131.0,
      "step": 24863
    },
    {
      "entropy": 0.5173291489481926,
      "epoch": 2.287611917125261,
      "grad_norm": 0.16081157326698303,
      "learning_rate": 2.3749501640751988e-05,
      "loss": 0.5048,
      "mean_token_accuracy": 0.838460274040699,
      "num_tokens": 793544982.0,
      "step": 24864
    },
    {
      "entropy": 0.4626592234708369,
      "epoch": 2.2877039234967023,
      "grad_norm": 0.15120665729045868,
      "learning_rate": 2.374643481461036e-05,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8600866496562958,
      "num_tokens": 793576758.0,
      "step": 24865
    },
    {
      "entropy": 0.49226532503962517,
      "epoch": 2.2877959298681434,
      "grad_norm": 0.1529003083705902,
      "learning_rate": 2.3743367988468734e-05,
      "loss": 0.4845,
      "mean_token_accuracy": 0.8442454822361469,
      "num_tokens": 793608725.0,
      "step": 24866
    },
    {
      "entropy": 0.49319954123348,
      "epoch": 2.2878879362395845,
      "grad_norm": 0.15869943797588348,
      "learning_rate": 2.3740301162327108e-05,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8455931283533573,
      "num_tokens": 793640839.0,
      "step": 24867
    },
    {
      "entropy": 0.5275928564369678,
      "epoch": 2.2879799426110257,
      "grad_norm": 0.16878071427345276,
      "learning_rate": 2.3737234336185483e-05,
      "loss": 0.5225,
      "mean_token_accuracy": 0.8354265205562115,
      "num_tokens": 793672334.0,
      "step": 24868
    },
    {
      "entropy": 0.5378963500261307,
      "epoch": 2.2880719489824672,
      "grad_norm": 0.15558236837387085,
      "learning_rate": 2.3734167510043857e-05,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8359597064554691,
      "num_tokens": 793704429.0,
      "step": 24869
    },
    {
      "entropy": 0.5604167282581329,
      "epoch": 2.2881639553539084,
      "grad_norm": 0.1706591546535492,
      "learning_rate": 2.373110068390223e-05,
      "loss": 0.5622,
      "mean_token_accuracy": 0.8249838165938854,
      "num_tokens": 793736462.0,
      "step": 24870
    },
    {
      "entropy": 0.6215525325387716,
      "epoch": 2.2882559617253495,
      "grad_norm": 0.1735486388206482,
      "learning_rate": 2.3728033857760603e-05,
      "loss": 0.6094,
      "mean_token_accuracy": 0.8098603151738644,
      "num_tokens": 793767537.0,
      "step": 24871
    },
    {
      "entropy": 0.44446720462292433,
      "epoch": 2.2883479680967906,
      "grad_norm": 0.1616355925798416,
      "learning_rate": 2.3724967031618978e-05,
      "loss": 0.4361,
      "mean_token_accuracy": 0.860302560031414,
      "num_tokens": 793799400.0,
      "step": 24872
    },
    {
      "entropy": 0.5838195476680994,
      "epoch": 2.2884399744682318,
      "grad_norm": 0.1735939234495163,
      "learning_rate": 2.3721900205477353e-05,
      "loss": 0.5714,
      "mean_token_accuracy": 0.8199001736938953,
      "num_tokens": 793830307.0,
      "step": 24873
    },
    {
      "entropy": 0.5196483000181615,
      "epoch": 2.2885319808396734,
      "grad_norm": 0.16798022389411926,
      "learning_rate": 2.3718833379335724e-05,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8390682078897953,
      "num_tokens": 793861824.0,
      "step": 24874
    },
    {
      "entropy": 0.49593030381947756,
      "epoch": 2.2886239872111145,
      "grad_norm": 0.1553063839673996,
      "learning_rate": 2.37157665531941e-05,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8430583588778973,
      "num_tokens": 793894019.0,
      "step": 24875
    },
    {
      "entropy": 0.4773111455142498,
      "epoch": 2.2887159935825556,
      "grad_norm": 0.1506500542163849,
      "learning_rate": 2.3712699727052477e-05,
      "loss": 0.4547,
      "mean_token_accuracy": 0.8549998998641968,
      "num_tokens": 793925963.0,
      "step": 24876
    },
    {
      "entropy": 0.4746790425851941,
      "epoch": 2.2888079999539968,
      "grad_norm": 0.15540798008441925,
      "learning_rate": 2.370963290091085e-05,
      "loss": 0.4605,
      "mean_token_accuracy": 0.8527517132461071,
      "num_tokens": 793957729.0,
      "step": 24877
    },
    {
      "entropy": 0.4879652170930058,
      "epoch": 2.288900006325438,
      "grad_norm": 0.15284959971904755,
      "learning_rate": 2.3706566074769223e-05,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8516496866941452,
      "num_tokens": 793989891.0,
      "step": 24878
    },
    {
      "entropy": 0.548241501674056,
      "epoch": 2.2889920126968795,
      "grad_norm": 0.16330325603485107,
      "learning_rate": 2.3703499248627597e-05,
      "loss": 0.5337,
      "mean_token_accuracy": 0.8324509970843792,
      "num_tokens": 794021365.0,
      "step": 24879
    },
    {
      "entropy": 0.6628760285675526,
      "epoch": 2.2890840190683206,
      "grad_norm": 0.18108564615249634,
      "learning_rate": 2.3700432422485972e-05,
      "loss": 0.6495,
      "mean_token_accuracy": 0.7971770539879799,
      "num_tokens": 794052910.0,
      "step": 24880
    },
    {
      "entropy": 0.5601666020229459,
      "epoch": 2.2891760254397617,
      "grad_norm": 0.1684601902961731,
      "learning_rate": 2.3697365596344346e-05,
      "loss": 0.5636,
      "mean_token_accuracy": 0.8236116506159306,
      "num_tokens": 794084922.0,
      "step": 24881
    },
    {
      "entropy": 0.49340156139805913,
      "epoch": 2.289268031811203,
      "grad_norm": 0.15913309156894684,
      "learning_rate": 2.3694298770202718e-05,
      "loss": 0.4821,
      "mean_token_accuracy": 0.8451677896082401,
      "num_tokens": 794117051.0,
      "step": 24882
    },
    {
      "entropy": 0.5375500526279211,
      "epoch": 2.289360038182644,
      "grad_norm": 0.1632024496793747,
      "learning_rate": 2.3691231944061092e-05,
      "loss": 0.5058,
      "mean_token_accuracy": 0.835398904979229,
      "num_tokens": 794148299.0,
      "step": 24883
    },
    {
      "entropy": 0.5176742672920227,
      "epoch": 2.2894520445540856,
      "grad_norm": 0.1611640900373459,
      "learning_rate": 2.3688165117919467e-05,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8412742167711258,
      "num_tokens": 794179535.0,
      "step": 24884
    },
    {
      "entropy": 0.5000151889398694,
      "epoch": 2.2895440509255267,
      "grad_norm": 0.1591489464044571,
      "learning_rate": 2.368509829177784e-05,
      "loss": 0.4827,
      "mean_token_accuracy": 0.8456514216959476,
      "num_tokens": 794211121.0,
      "step": 24885
    },
    {
      "entropy": 0.5212646559812129,
      "epoch": 2.289636057296968,
      "grad_norm": 0.1624983847141266,
      "learning_rate": 2.3682031465636216e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8402030803263187,
      "num_tokens": 794243303.0,
      "step": 24886
    },
    {
      "entropy": 0.48617367167025805,
      "epoch": 2.289728063668409,
      "grad_norm": 0.15178504586219788,
      "learning_rate": 2.3678964639494588e-05,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8515279218554497,
      "num_tokens": 794275409.0,
      "step": 24887
    },
    {
      "entropy": 0.5560215450823307,
      "epoch": 2.28982007003985,
      "grad_norm": 0.1634722203016281,
      "learning_rate": 2.3675897813352962e-05,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8272469937801361,
      "num_tokens": 794307613.0,
      "step": 24888
    },
    {
      "entropy": 0.5791906900703907,
      "epoch": 2.2899120764112917,
      "grad_norm": 0.1649608165025711,
      "learning_rate": 2.3672830987211337e-05,
      "loss": 0.561,
      "mean_token_accuracy": 0.822334673255682,
      "num_tokens": 794339232.0,
      "step": 24889
    },
    {
      "entropy": 0.5460730958729982,
      "epoch": 2.290004082782733,
      "grad_norm": 0.1637134999036789,
      "learning_rate": 2.366976416106971e-05,
      "loss": 0.5402,
      "mean_token_accuracy": 0.8284619115293026,
      "num_tokens": 794371334.0,
      "step": 24890
    },
    {
      "entropy": 0.3595392741262913,
      "epoch": 2.290096089154174,
      "grad_norm": 0.13924045860767365,
      "learning_rate": 2.3666697334928083e-05,
      "loss": 0.3455,
      "mean_token_accuracy": 0.890695832669735,
      "num_tokens": 794403458.0,
      "step": 24891
    },
    {
      "entropy": 0.5482819201424718,
      "epoch": 2.290188095525615,
      "grad_norm": 0.17220479249954224,
      "learning_rate": 2.3663630508786457e-05,
      "loss": 0.5518,
      "mean_token_accuracy": 0.8342283852398396,
      "num_tokens": 794435305.0,
      "step": 24892
    },
    {
      "entropy": 0.46083150804042816,
      "epoch": 2.290280101897056,
      "grad_norm": 0.18036280572414398,
      "learning_rate": 2.3660563682644832e-05,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8544235862791538,
      "num_tokens": 794466723.0,
      "step": 24893
    },
    {
      "entropy": 0.49965286161750555,
      "epoch": 2.290372108268498,
      "grad_norm": 0.15604689717292786,
      "learning_rate": 2.3657496856503207e-05,
      "loss": 0.4836,
      "mean_token_accuracy": 0.8497989289462566,
      "num_tokens": 794498949.0,
      "step": 24894
    },
    {
      "entropy": 0.5785507261753082,
      "epoch": 2.290464114639939,
      "grad_norm": 0.17067939043045044,
      "learning_rate": 2.3654430030361578e-05,
      "loss": 0.5777,
      "mean_token_accuracy": 0.8159575909376144,
      "num_tokens": 794530481.0,
      "step": 24895
    },
    {
      "entropy": 0.5533730015158653,
      "epoch": 2.29055612101138,
      "grad_norm": 0.163491889834404,
      "learning_rate": 2.3651363204219953e-05,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8289369903504848,
      "num_tokens": 794562266.0,
      "step": 24896
    },
    {
      "entropy": 0.5203097239136696,
      "epoch": 2.290648127382821,
      "grad_norm": 0.16478732228279114,
      "learning_rate": 2.3648296378078327e-05,
      "loss": 0.5028,
      "mean_token_accuracy": 0.839831605553627,
      "num_tokens": 794594297.0,
      "step": 24897
    },
    {
      "entropy": 0.5097546866163611,
      "epoch": 2.2907401337542623,
      "grad_norm": 0.16548706591129303,
      "learning_rate": 2.3645229551936702e-05,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8471565507352352,
      "num_tokens": 794626112.0,
      "step": 24898
    },
    {
      "entropy": 0.6701088100671768,
      "epoch": 2.290832140125704,
      "grad_norm": 0.1767224818468094,
      "learning_rate": 2.3642162725795077e-05,
      "loss": 0.6604,
      "mean_token_accuracy": 0.7927133813500404,
      "num_tokens": 794658071.0,
      "step": 24899
    },
    {
      "entropy": 0.49892285466194153,
      "epoch": 2.290924146497145,
      "grad_norm": 0.15428024530410767,
      "learning_rate": 2.3639095899653448e-05,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8453267626464367,
      "num_tokens": 794690297.0,
      "step": 24900
    },
    {
      "entropy": 0.4492517402395606,
      "epoch": 2.291016152868586,
      "grad_norm": 0.14764510095119476,
      "learning_rate": 2.3636029073511822e-05,
      "loss": 0.4272,
      "mean_token_accuracy": 0.8607335835695267,
      "num_tokens": 794721980.0,
      "step": 24901
    },
    {
      "entropy": 0.5611344091594219,
      "epoch": 2.2911081592400273,
      "grad_norm": 0.16714975237846375,
      "learning_rate": 2.3632962247370197e-05,
      "loss": 0.5514,
      "mean_token_accuracy": 0.8277269601821899,
      "num_tokens": 794753810.0,
      "step": 24902
    },
    {
      "entropy": 0.5589141305536032,
      "epoch": 2.2912001656114684,
      "grad_norm": 0.16229067742824554,
      "learning_rate": 2.3629895421228572e-05,
      "loss": 0.5452,
      "mean_token_accuracy": 0.8296162858605385,
      "num_tokens": 794785872.0,
      "step": 24903
    },
    {
      "entropy": 0.47427027812227607,
      "epoch": 2.29129217198291,
      "grad_norm": 0.15505018830299377,
      "learning_rate": 2.3626828595086943e-05,
      "loss": 0.4635,
      "mean_token_accuracy": 0.8541552573442459,
      "num_tokens": 794817973.0,
      "step": 24904
    },
    {
      "entropy": 0.5825018975883722,
      "epoch": 2.291384178354351,
      "grad_norm": 0.1621822863817215,
      "learning_rate": 2.3623761768945318e-05,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8214263580739498,
      "num_tokens": 794850118.0,
      "step": 24905
    },
    {
      "entropy": 0.4485404046718031,
      "epoch": 2.2914761847257923,
      "grad_norm": 0.14725925028324127,
      "learning_rate": 2.3620694942803692e-05,
      "loss": 0.4319,
      "mean_token_accuracy": 0.8630339130759239,
      "num_tokens": 794882369.0,
      "step": 24906
    },
    {
      "entropy": 0.5990163937676698,
      "epoch": 2.2915681910972334,
      "grad_norm": 0.16861175000667572,
      "learning_rate": 2.361762811666207e-05,
      "loss": 0.5749,
      "mean_token_accuracy": 0.8198669441044331,
      "num_tokens": 794914211.0,
      "step": 24907
    },
    {
      "entropy": 0.47966704796999693,
      "epoch": 2.2916601974686746,
      "grad_norm": 0.1593606173992157,
      "learning_rate": 2.361456129052044e-05,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8502033576369286,
      "num_tokens": 794945757.0,
      "step": 24908
    },
    {
      "entropy": 0.5142995631322265,
      "epoch": 2.291752203840116,
      "grad_norm": 0.1601037085056305,
      "learning_rate": 2.3611494464378816e-05,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8429182805120945,
      "num_tokens": 794977975.0,
      "step": 24909
    },
    {
      "entropy": 0.5072799725458026,
      "epoch": 2.2918442102115573,
      "grad_norm": 0.15894857048988342,
      "learning_rate": 2.360842763823719e-05,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8451592922210693,
      "num_tokens": 795009550.0,
      "step": 24910
    },
    {
      "entropy": 0.455021345987916,
      "epoch": 2.2919362165829984,
      "grad_norm": 0.154002845287323,
      "learning_rate": 2.3605360812095566e-05,
      "loss": 0.4413,
      "mean_token_accuracy": 0.8605850525200367,
      "num_tokens": 795041747.0,
      "step": 24911
    },
    {
      "entropy": 0.5322994254529476,
      "epoch": 2.2920282229544395,
      "grad_norm": 0.1622229963541031,
      "learning_rate": 2.3602293985953937e-05,
      "loss": 0.5184,
      "mean_token_accuracy": 0.8332306668162346,
      "num_tokens": 795073529.0,
      "step": 24912
    },
    {
      "entropy": 0.5200427970848978,
      "epoch": 2.2921202293258807,
      "grad_norm": 0.1637488156557083,
      "learning_rate": 2.359922715981231e-05,
      "loss": 0.5075,
      "mean_token_accuracy": 0.836538553237915,
      "num_tokens": 795105089.0,
      "step": 24913
    },
    {
      "entropy": 0.4278418552130461,
      "epoch": 2.2922122356973222,
      "grad_norm": 0.1509123593568802,
      "learning_rate": 2.3596160333670686e-05,
      "loss": 0.413,
      "mean_token_accuracy": 0.8666439987719059,
      "num_tokens": 795137579.0,
      "step": 24914
    },
    {
      "entropy": 0.4667191943153739,
      "epoch": 2.2923042420687634,
      "grad_norm": 0.15407437086105347,
      "learning_rate": 2.359309350752906e-05,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8538168370723724,
      "num_tokens": 795169549.0,
      "step": 24915
    },
    {
      "entropy": 0.5721341855823994,
      "epoch": 2.2923962484402045,
      "grad_norm": 0.16815850138664246,
      "learning_rate": 2.3590026681387435e-05,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8263775780797005,
      "num_tokens": 795200966.0,
      "step": 24916
    },
    {
      "entropy": 0.5707619192544371,
      "epoch": 2.2924882548116456,
      "grad_norm": 0.17238500714302063,
      "learning_rate": 2.3586959855245807e-05,
      "loss": 0.5556,
      "mean_token_accuracy": 0.8277201130986214,
      "num_tokens": 795232603.0,
      "step": 24917
    },
    {
      "entropy": 0.5961763095110655,
      "epoch": 2.2925802611830868,
      "grad_norm": 0.17218802869319916,
      "learning_rate": 2.358389302910418e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8143988735973835,
      "num_tokens": 795264555.0,
      "step": 24918
    },
    {
      "entropy": 0.5173727568471804,
      "epoch": 2.2926722675545284,
      "grad_norm": 0.16452138125896454,
      "learning_rate": 2.3580826202962556e-05,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8388156853616238,
      "num_tokens": 795296163.0,
      "step": 24919
    },
    {
      "entropy": 0.4490581825375557,
      "epoch": 2.2927642739259695,
      "grad_norm": 0.1591854840517044,
      "learning_rate": 2.357775937682093e-05,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8585426360368729,
      "num_tokens": 795328191.0,
      "step": 24920
    },
    {
      "entropy": 0.5248182341456413,
      "epoch": 2.2928562802974106,
      "grad_norm": 0.15619227290153503,
      "learning_rate": 2.3574692550679302e-05,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8397159390151501,
      "num_tokens": 795360887.0,
      "step": 24921
    },
    {
      "entropy": 0.48601557791698724,
      "epoch": 2.2929482866688518,
      "grad_norm": 0.16007116436958313,
      "learning_rate": 2.3571625724537677e-05,
      "loss": 0.4707,
      "mean_token_accuracy": 0.851886011660099,
      "num_tokens": 795391993.0,
      "step": 24922
    },
    {
      "entropy": 0.6015950590372086,
      "epoch": 2.293040293040293,
      "grad_norm": 0.16633644700050354,
      "learning_rate": 2.356855889839605e-05,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8157526887953281,
      "num_tokens": 795424060.0,
      "step": 24923
    },
    {
      "entropy": 0.5174860870465636,
      "epoch": 2.2931322994117345,
      "grad_norm": 0.15745118260383606,
      "learning_rate": 2.3565492072254426e-05,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8365189023315907,
      "num_tokens": 795456585.0,
      "step": 24924
    },
    {
      "entropy": 0.48351908661425114,
      "epoch": 2.2932243057831756,
      "grad_norm": 0.15613022446632385,
      "learning_rate": 2.3562425246112797e-05,
      "loss": 0.4701,
      "mean_token_accuracy": 0.8487137109041214,
      "num_tokens": 795488718.0,
      "step": 24925
    },
    {
      "entropy": 0.5244168797507882,
      "epoch": 2.2933163121546167,
      "grad_norm": 0.16309314966201782,
      "learning_rate": 2.3559358419971172e-05,
      "loss": 0.5204,
      "mean_token_accuracy": 0.8353947810828686,
      "num_tokens": 795521092.0,
      "step": 24926
    },
    {
      "entropy": 0.5111340545117855,
      "epoch": 2.293408318526058,
      "grad_norm": 0.1566571742296219,
      "learning_rate": 2.3556291593829546e-05,
      "loss": 0.4925,
      "mean_token_accuracy": 0.843125719577074,
      "num_tokens": 795553225.0,
      "step": 24927
    },
    {
      "entropy": 0.49534939881414175,
      "epoch": 2.293500324897499,
      "grad_norm": 0.1584838181734085,
      "learning_rate": 2.355322476768792e-05,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8441843204200268,
      "num_tokens": 795584905.0,
      "step": 24928
    },
    {
      "entropy": 0.4930248875170946,
      "epoch": 2.2935923312689406,
      "grad_norm": 0.1552295684814453,
      "learning_rate": 2.3550157941546296e-05,
      "loss": 0.485,
      "mean_token_accuracy": 0.8470005393028259,
      "num_tokens": 795616891.0,
      "step": 24929
    },
    {
      "entropy": 0.5670562302693725,
      "epoch": 2.2936843376403817,
      "grad_norm": 0.1619044989347458,
      "learning_rate": 2.3547091115404667e-05,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8271258771419525,
      "num_tokens": 795648767.0,
      "step": 24930
    },
    {
      "entropy": 0.48993699438869953,
      "epoch": 2.293776344011823,
      "grad_norm": 0.15871202945709229,
      "learning_rate": 2.354402428926304e-05,
      "loss": 0.4892,
      "mean_token_accuracy": 0.8455793559551239,
      "num_tokens": 795681010.0,
      "step": 24931
    },
    {
      "entropy": 0.6114970068447292,
      "epoch": 2.293868350383264,
      "grad_norm": 0.1773960441350937,
      "learning_rate": 2.3540957463121416e-05,
      "loss": 0.6077,
      "mean_token_accuracy": 0.8090360276401043,
      "num_tokens": 795712992.0,
      "step": 24932
    },
    {
      "entropy": 0.586564190685749,
      "epoch": 2.293960356754705,
      "grad_norm": 0.17255344986915588,
      "learning_rate": 2.353789063697979e-05,
      "loss": 0.5878,
      "mean_token_accuracy": 0.8187693580985069,
      "num_tokens": 795745009.0,
      "step": 24933
    },
    {
      "entropy": 0.45229259273037314,
      "epoch": 2.2940523631261467,
      "grad_norm": 0.15265166759490967,
      "learning_rate": 2.3534823810838162e-05,
      "loss": 0.4392,
      "mean_token_accuracy": 0.8599123284220695,
      "num_tokens": 795777450.0,
      "step": 24934
    },
    {
      "entropy": 0.5342502975836396,
      "epoch": 2.294144369497588,
      "grad_norm": 0.16088582575321198,
      "learning_rate": 2.3531756984696537e-05,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8305434547364712,
      "num_tokens": 795809088.0,
      "step": 24935
    },
    {
      "entropy": 0.3938576662912965,
      "epoch": 2.294236375869029,
      "grad_norm": 0.1445605605840683,
      "learning_rate": 2.352869015855491e-05,
      "loss": 0.3745,
      "mean_token_accuracy": 0.8754732981324196,
      "num_tokens": 795840601.0,
      "step": 24936
    },
    {
      "entropy": 0.6105713210999966,
      "epoch": 2.29432838224047,
      "grad_norm": 0.17179830372333527,
      "learning_rate": 2.352562333241329e-05,
      "loss": 0.5941,
      "mean_token_accuracy": 0.8149853721261024,
      "num_tokens": 795872798.0,
      "step": 24937
    },
    {
      "entropy": 0.44254941027611494,
      "epoch": 2.294420388611911,
      "grad_norm": 0.1499170958995819,
      "learning_rate": 2.352255650627166e-05,
      "loss": 0.4242,
      "mean_token_accuracy": 0.8636000342667103,
      "num_tokens": 795904651.0,
      "step": 24938
    },
    {
      "entropy": 0.5381370261311531,
      "epoch": 2.294512394983353,
      "grad_norm": 0.16814732551574707,
      "learning_rate": 2.3519489680130035e-05,
      "loss": 0.5322,
      "mean_token_accuracy": 0.8310924954712391,
      "num_tokens": 795935845.0,
      "step": 24939
    },
    {
      "entropy": 0.4841436455026269,
      "epoch": 2.294604401354794,
      "grad_norm": 0.15658968687057495,
      "learning_rate": 2.351642285398841e-05,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8511376678943634,
      "num_tokens": 795967812.0,
      "step": 24940
    },
    {
      "entropy": 0.46690370817668736,
      "epoch": 2.294696407726235,
      "grad_norm": 0.1546401083469391,
      "learning_rate": 2.3513356027846785e-05,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8549908436834812,
      "num_tokens": 795999747.0,
      "step": 24941
    },
    {
      "entropy": 0.5135745285078883,
      "epoch": 2.294788414097676,
      "grad_norm": 0.15344947576522827,
      "learning_rate": 2.3510289201705156e-05,
      "loss": 0.4936,
      "mean_token_accuracy": 0.8460807986557484,
      "num_tokens": 796031859.0,
      "step": 24942
    },
    {
      "entropy": 0.5513822785578668,
      "epoch": 2.2948804204691173,
      "grad_norm": 0.16375668346881866,
      "learning_rate": 2.350722237556353e-05,
      "loss": 0.533,
      "mean_token_accuracy": 0.8333584107458591,
      "num_tokens": 796063750.0,
      "step": 24943
    },
    {
      "entropy": 0.5712618175894022,
      "epoch": 2.294972426840559,
      "grad_norm": 0.162966787815094,
      "learning_rate": 2.3504155549421905e-05,
      "loss": 0.5533,
      "mean_token_accuracy": 0.8253410346806049,
      "num_tokens": 796095722.0,
      "step": 24944
    },
    {
      "entropy": 0.5421283128671348,
      "epoch": 2.295064433212,
      "grad_norm": 0.16090109944343567,
      "learning_rate": 2.350108872328028e-05,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8325337544083595,
      "num_tokens": 796128035.0,
      "step": 24945
    },
    {
      "entropy": 0.5856720940209925,
      "epoch": 2.295156439583441,
      "grad_norm": 0.1675437092781067,
      "learning_rate": 2.3498021897138655e-05,
      "loss": 0.57,
      "mean_token_accuracy": 0.8219611272215843,
      "num_tokens": 796160445.0,
      "step": 24946
    },
    {
      "entropy": 0.5179970301687717,
      "epoch": 2.2952484459548823,
      "grad_norm": 0.16377781331539154,
      "learning_rate": 2.3494955070997026e-05,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8379718959331512,
      "num_tokens": 796192032.0,
      "step": 24947
    },
    {
      "entropy": 0.5205935016274452,
      "epoch": 2.2953404523263234,
      "grad_norm": 0.1659930944442749,
      "learning_rate": 2.34918882448554e-05,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8413961715996265,
      "num_tokens": 796223252.0,
      "step": 24948
    },
    {
      "entropy": 0.5642033815383911,
      "epoch": 2.295432458697765,
      "grad_norm": 0.16385924816131592,
      "learning_rate": 2.3488821418713775e-05,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8270787969231606,
      "num_tokens": 796254330.0,
      "step": 24949
    },
    {
      "entropy": 0.548665358684957,
      "epoch": 2.295524465069206,
      "grad_norm": 0.16018140316009521,
      "learning_rate": 2.348575459257215e-05,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8301636427640915,
      "num_tokens": 796286009.0,
      "step": 24950
    },
    {
      "entropy": 0.6212998833507299,
      "epoch": 2.2956164714406473,
      "grad_norm": 0.16484177112579346,
      "learning_rate": 2.348268776643052e-05,
      "loss": 0.6066,
      "mean_token_accuracy": 0.8102474957704544,
      "num_tokens": 796318482.0,
      "step": 24951
    },
    {
      "entropy": 0.4676187774166465,
      "epoch": 2.2957084778120884,
      "grad_norm": 0.151743084192276,
      "learning_rate": 2.3479620940288896e-05,
      "loss": 0.4608,
      "mean_token_accuracy": 0.85307552292943,
      "num_tokens": 796350242.0,
      "step": 24952
    },
    {
      "entropy": 0.5809926623478532,
      "epoch": 2.2958004841835296,
      "grad_norm": 0.16832545399665833,
      "learning_rate": 2.347655411414727e-05,
      "loss": 0.5697,
      "mean_token_accuracy": 0.8227425813674927,
      "num_tokens": 796382681.0,
      "step": 24953
    },
    {
      "entropy": 0.43061370751820505,
      "epoch": 2.295892490554971,
      "grad_norm": 0.1485719382762909,
      "learning_rate": 2.3473487288005645e-05,
      "loss": 0.4075,
      "mean_token_accuracy": 0.8691484481096268,
      "num_tokens": 796414356.0,
      "step": 24954
    },
    {
      "entropy": 0.45010579109657556,
      "epoch": 2.2959844969264123,
      "grad_norm": 0.15489846467971802,
      "learning_rate": 2.3470420461864016e-05,
      "loss": 0.4494,
      "mean_token_accuracy": 0.8588790334761143,
      "num_tokens": 796446110.0,
      "step": 24955
    },
    {
      "entropy": 0.5378004387021065,
      "epoch": 2.2960765032978534,
      "grad_norm": 0.16193413734436035,
      "learning_rate": 2.346735363572239e-05,
      "loss": 0.5255,
      "mean_token_accuracy": 0.8337532803416252,
      "num_tokens": 796477989.0,
      "step": 24956
    },
    {
      "entropy": 0.5843570046126842,
      "epoch": 2.2961685096692945,
      "grad_norm": 0.1709117442369461,
      "learning_rate": 2.3464286809580765e-05,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8208719603717327,
      "num_tokens": 796510001.0,
      "step": 24957
    },
    {
      "entropy": 0.5443071480840445,
      "epoch": 2.2962605160407357,
      "grad_norm": 0.164486825466156,
      "learning_rate": 2.346121998343914e-05,
      "loss": 0.5417,
      "mean_token_accuracy": 0.8305964656174183,
      "num_tokens": 796542194.0,
      "step": 24958
    },
    {
      "entropy": 0.4964566472917795,
      "epoch": 2.2963525224121772,
      "grad_norm": 0.15994606912136078,
      "learning_rate": 2.3458153157297515e-05,
      "loss": 0.473,
      "mean_token_accuracy": 0.8508971706032753,
      "num_tokens": 796573547.0,
      "step": 24959
    },
    {
      "entropy": 0.5075732180848718,
      "epoch": 2.2964445287836184,
      "grad_norm": 0.15976104140281677,
      "learning_rate": 2.3455086331155886e-05,
      "loss": 0.4971,
      "mean_token_accuracy": 0.8409221693873405,
      "num_tokens": 796604694.0,
      "step": 24960
    },
    {
      "entropy": 0.5177632849663496,
      "epoch": 2.2965365351550595,
      "grad_norm": 0.16863712668418884,
      "learning_rate": 2.345201950501426e-05,
      "loss": 0.5142,
      "mean_token_accuracy": 0.8387355506420135,
      "num_tokens": 796637005.0,
      "step": 24961
    },
    {
      "entropy": 0.5320279663428664,
      "epoch": 2.2966285415265006,
      "grad_norm": 0.16566243767738342,
      "learning_rate": 2.3448952678872635e-05,
      "loss": 0.5214,
      "mean_token_accuracy": 0.8319780826568604,
      "num_tokens": 796668393.0,
      "step": 24962
    },
    {
      "entropy": 0.5555369248613715,
      "epoch": 2.2967205478979418,
      "grad_norm": 0.16381089389324188,
      "learning_rate": 2.344588585273101e-05,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8308431208133698,
      "num_tokens": 796700170.0,
      "step": 24963
    },
    {
      "entropy": 0.4244365638587624,
      "epoch": 2.2968125542693834,
      "grad_norm": 0.14799407124519348,
      "learning_rate": 2.344281902658938e-05,
      "loss": 0.406,
      "mean_token_accuracy": 0.8723683804273605,
      "num_tokens": 796732547.0,
      "step": 24964
    },
    {
      "entropy": 0.5403442396782339,
      "epoch": 2.2969045606408245,
      "grad_norm": 0.15996474027633667,
      "learning_rate": 2.3439752200447756e-05,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8325475715100765,
      "num_tokens": 796764959.0,
      "step": 24965
    },
    {
      "entropy": 0.6044024098664522,
      "epoch": 2.2969965670122656,
      "grad_norm": 0.17677073180675507,
      "learning_rate": 2.343668537430613e-05,
      "loss": 0.5879,
      "mean_token_accuracy": 0.8146393075585365,
      "num_tokens": 796796208.0,
      "step": 24966
    },
    {
      "entropy": 0.5064306296408176,
      "epoch": 2.2970885733837068,
      "grad_norm": 0.16365602612495422,
      "learning_rate": 2.3433618548164505e-05,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8391654267907143,
      "num_tokens": 796827898.0,
      "step": 24967
    },
    {
      "entropy": 0.4182888828217983,
      "epoch": 2.297180579755148,
      "grad_norm": 0.15554232895374298,
      "learning_rate": 2.343055172202288e-05,
      "loss": 0.3996,
      "mean_token_accuracy": 0.8720531463623047,
      "num_tokens": 796859430.0,
      "step": 24968
    },
    {
      "entropy": 0.5578549783676863,
      "epoch": 2.2972725861265895,
      "grad_norm": 0.1632823348045349,
      "learning_rate": 2.3427484895881254e-05,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8269779570400715,
      "num_tokens": 796890613.0,
      "step": 24969
    },
    {
      "entropy": 0.5353901423513889,
      "epoch": 2.2973645924980306,
      "grad_norm": 0.16182173788547516,
      "learning_rate": 2.342441806973963e-05,
      "loss": 0.5101,
      "mean_token_accuracy": 0.8350053392350674,
      "num_tokens": 796922037.0,
      "step": 24970
    },
    {
      "entropy": 0.5821937825530767,
      "epoch": 2.2974565988694717,
      "grad_norm": 0.16786684095859528,
      "learning_rate": 2.3421351243598004e-05,
      "loss": 0.5757,
      "mean_token_accuracy": 0.8177402876317501,
      "num_tokens": 796953949.0,
      "step": 24971
    },
    {
      "entropy": 0.5102857337333262,
      "epoch": 2.297548605240913,
      "grad_norm": 0.1591046154499054,
      "learning_rate": 2.3418284417456375e-05,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8408326357603073,
      "num_tokens": 796986484.0,
      "step": 24972
    },
    {
      "entropy": 0.5100459307432175,
      "epoch": 2.297640611612354,
      "grad_norm": 0.16351282596588135,
      "learning_rate": 2.341521759131475e-05,
      "loss": 0.5011,
      "mean_token_accuracy": 0.839546836912632,
      "num_tokens": 797018527.0,
      "step": 24973
    },
    {
      "entropy": 0.42257893295027316,
      "epoch": 2.2977326179837956,
      "grad_norm": 0.14713886380195618,
      "learning_rate": 2.3412150765173124e-05,
      "loss": 0.4082,
      "mean_token_accuracy": 0.8704797327518463,
      "num_tokens": 797051031.0,
      "step": 24974
    },
    {
      "entropy": 0.5161332236602902,
      "epoch": 2.2978246243552367,
      "grad_norm": 0.15928435325622559,
      "learning_rate": 2.34090839390315e-05,
      "loss": 0.5111,
      "mean_token_accuracy": 0.8382378555834293,
      "num_tokens": 797083099.0,
      "step": 24975
    },
    {
      "entropy": 0.5456962063908577,
      "epoch": 2.297916630726678,
      "grad_norm": 0.1701926290988922,
      "learning_rate": 2.3406017112889874e-05,
      "loss": 0.5356,
      "mean_token_accuracy": 0.832642700523138,
      "num_tokens": 797114713.0,
      "step": 24976
    },
    {
      "entropy": 0.5356310708448291,
      "epoch": 2.298008637098119,
      "grad_norm": 0.15978501737117767,
      "learning_rate": 2.3402950286748245e-05,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8345286212861538,
      "num_tokens": 797147113.0,
      "step": 24977
    },
    {
      "entropy": 0.5568617582321167,
      "epoch": 2.29810064346956,
      "grad_norm": 0.16730789840221405,
      "learning_rate": 2.339988346060662e-05,
      "loss": 0.5468,
      "mean_token_accuracy": 0.8283652849495411,
      "num_tokens": 797179635.0,
      "step": 24978
    },
    {
      "entropy": 0.4626257074996829,
      "epoch": 2.2981926498410017,
      "grad_norm": 0.15678471326828003,
      "learning_rate": 2.3396816634464994e-05,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8594992831349373,
      "num_tokens": 797211606.0,
      "step": 24979
    },
    {
      "entropy": 0.5477163000032306,
      "epoch": 2.298284656212443,
      "grad_norm": 0.16309750080108643,
      "learning_rate": 2.339374980832337e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8308718129992485,
      "num_tokens": 797243889.0,
      "step": 24980
    },
    {
      "entropy": 0.4642360284924507,
      "epoch": 2.298376662583884,
      "grad_norm": 0.15226751565933228,
      "learning_rate": 2.339068298218174e-05,
      "loss": 0.459,
      "mean_token_accuracy": 0.8540207669138908,
      "num_tokens": 797276149.0,
      "step": 24981
    },
    {
      "entropy": 0.5039901188574731,
      "epoch": 2.298468668955325,
      "grad_norm": 0.16104307770729065,
      "learning_rate": 2.3387616156040115e-05,
      "loss": 0.48,
      "mean_token_accuracy": 0.8484782315790653,
      "num_tokens": 797307789.0,
      "step": 24982
    },
    {
      "entropy": 0.49371496494859457,
      "epoch": 2.2985606753267662,
      "grad_norm": 0.16020765900611877,
      "learning_rate": 2.338454932989849e-05,
      "loss": 0.4805,
      "mean_token_accuracy": 0.8454725258052349,
      "num_tokens": 797339825.0,
      "step": 24983
    },
    {
      "entropy": 0.64998834207654,
      "epoch": 2.298652681698208,
      "grad_norm": 0.17353829741477966,
      "learning_rate": 2.3381482503756864e-05,
      "loss": 0.6399,
      "mean_token_accuracy": 0.799845028668642,
      "num_tokens": 797372401.0,
      "step": 24984
    },
    {
      "entropy": 0.5478743710555136,
      "epoch": 2.298744688069649,
      "grad_norm": 0.1593879759311676,
      "learning_rate": 2.3378415677615235e-05,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8294363804161549,
      "num_tokens": 797405154.0,
      "step": 24985
    },
    {
      "entropy": 0.48827136820182204,
      "epoch": 2.29883669444109,
      "grad_norm": 0.15786112844944,
      "learning_rate": 2.337534885147361e-05,
      "loss": 0.469,
      "mean_token_accuracy": 0.8495898805558681,
      "num_tokens": 797437198.0,
      "step": 24986
    },
    {
      "entropy": 0.526643147226423,
      "epoch": 2.298928700812531,
      "grad_norm": 0.15669922530651093,
      "learning_rate": 2.3372282025331985e-05,
      "loss": 0.5006,
      "mean_token_accuracy": 0.8411644622683525,
      "num_tokens": 797469172.0,
      "step": 24987
    },
    {
      "entropy": 0.5383353736251593,
      "epoch": 2.2990207071839723,
      "grad_norm": 0.1556607037782669,
      "learning_rate": 2.336921519919036e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.835374653339386,
      "num_tokens": 797501464.0,
      "step": 24988
    },
    {
      "entropy": 0.4339510863646865,
      "epoch": 2.299112713555414,
      "grad_norm": 0.14784646034240723,
      "learning_rate": 2.3366148373048734e-05,
      "loss": 0.4092,
      "mean_token_accuracy": 0.8681871369481087,
      "num_tokens": 797533839.0,
      "step": 24989
    },
    {
      "entropy": 0.6817240286618471,
      "epoch": 2.299204719926855,
      "grad_norm": 0.17683304846286774,
      "learning_rate": 2.3363081546907105e-05,
      "loss": 0.6697,
      "mean_token_accuracy": 0.7923115715384483,
      "num_tokens": 797565619.0,
      "step": 24990
    },
    {
      "entropy": 0.4101019313093275,
      "epoch": 2.299296726298296,
      "grad_norm": 0.14064453542232513,
      "learning_rate": 2.336001472076548e-05,
      "loss": 0.3996,
      "mean_token_accuracy": 0.871000237762928,
      "num_tokens": 797598387.0,
      "step": 24991
    },
    {
      "entropy": 0.46188132790848613,
      "epoch": 2.2993887326697373,
      "grad_norm": 0.15440036356449127,
      "learning_rate": 2.3356947894623854e-05,
      "loss": 0.4476,
      "mean_token_accuracy": 0.8589780777692795,
      "num_tokens": 797630488.0,
      "step": 24992
    },
    {
      "entropy": 0.4993925429880619,
      "epoch": 2.2994807390411784,
      "grad_norm": 0.16107048094272614,
      "learning_rate": 2.335388106848223e-05,
      "loss": 0.4883,
      "mean_token_accuracy": 0.8412846252322197,
      "num_tokens": 797661848.0,
      "step": 24993
    },
    {
      "entropy": 0.4820429487153888,
      "epoch": 2.29957274541262,
      "grad_norm": 0.15712673962116241,
      "learning_rate": 2.33508142423406e-05,
      "loss": 0.4747,
      "mean_token_accuracy": 0.8500072360038757,
      "num_tokens": 797693687.0,
      "step": 24994
    },
    {
      "entropy": 0.5412921579554677,
      "epoch": 2.299664751784061,
      "grad_norm": 0.16397319734096527,
      "learning_rate": 2.3347747416198975e-05,
      "loss": 0.5158,
      "mean_token_accuracy": 0.8360274732112885,
      "num_tokens": 797724889.0,
      "step": 24995
    },
    {
      "entropy": 0.5370602821931243,
      "epoch": 2.2997567581555023,
      "grad_norm": 0.16405412554740906,
      "learning_rate": 2.334468059005735e-05,
      "loss": 0.5225,
      "mean_token_accuracy": 0.8342984840273857,
      "num_tokens": 797756949.0,
      "step": 24996
    },
    {
      "entropy": 0.48127812519669533,
      "epoch": 2.2998487645269434,
      "grad_norm": 0.1565246731042862,
      "learning_rate": 2.3341613763915724e-05,
      "loss": 0.4744,
      "mean_token_accuracy": 0.84727793186903,
      "num_tokens": 797788854.0,
      "step": 24997
    },
    {
      "entropy": 0.37910995120182633,
      "epoch": 2.2999407708983846,
      "grad_norm": 0.1454056054353714,
      "learning_rate": 2.33385469377741e-05,
      "loss": 0.3646,
      "mean_token_accuracy": 0.8825919143855572,
      "num_tokens": 797820997.0,
      "step": 24998
    },
    {
      "entropy": 0.5743078459054232,
      "epoch": 2.300032777269826,
      "grad_norm": 0.16802534461021423,
      "learning_rate": 2.3335480111632474e-05,
      "loss": 0.5713,
      "mean_token_accuracy": 0.818975817412138,
      "num_tokens": 797852997.0,
      "step": 24999
    },
    {
      "entropy": 0.4522051736712456,
      "epoch": 2.3001247836412673,
      "grad_norm": 0.16167418658733368,
      "learning_rate": 2.3332413285490848e-05,
      "loss": 0.4472,
      "mean_token_accuracy": 0.8544017411768436,
      "num_tokens": 797883455.0,
      "step": 25000
    },
    {
      "entropy": 0.573704537935555,
      "epoch": 2.3002167900127084,
      "grad_norm": 0.16598133742809296,
      "learning_rate": 2.3329346459349223e-05,
      "loss": 0.5644,
      "mean_token_accuracy": 0.8259575963020325,
      "num_tokens": 797916019.0,
      "step": 25001
    },
    {
      "entropy": 0.542986296582967,
      "epoch": 2.3003087963841495,
      "grad_norm": 0.16387507319450378,
      "learning_rate": 2.3326279633207594e-05,
      "loss": 0.5336,
      "mean_token_accuracy": 0.8310545831918716,
      "num_tokens": 797947764.0,
      "step": 25002
    },
    {
      "entropy": 0.556487925350666,
      "epoch": 2.3004008027555907,
      "grad_norm": 0.1666967272758484,
      "learning_rate": 2.332321280706597e-05,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8257996551692486,
      "num_tokens": 797979338.0,
      "step": 25003
    },
    {
      "entropy": 0.5515676881186664,
      "epoch": 2.3004928091270322,
      "grad_norm": 0.16238993406295776,
      "learning_rate": 2.3320145980924343e-05,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8281550630927086,
      "num_tokens": 798011626.0,
      "step": 25004
    },
    {
      "entropy": 0.510784930549562,
      "epoch": 2.3005848154984734,
      "grad_norm": 0.16472209990024567,
      "learning_rate": 2.3317079154782718e-05,
      "loss": 0.4881,
      "mean_token_accuracy": 0.845031201839447,
      "num_tokens": 798043174.0,
      "step": 25005
    },
    {
      "entropy": 0.4637219519354403,
      "epoch": 2.3006768218699145,
      "grad_norm": 0.1510649174451828,
      "learning_rate": 2.3314012328641093e-05,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8568424805998802,
      "num_tokens": 798075139.0,
      "step": 25006
    },
    {
      "entropy": 0.5221775269601494,
      "epoch": 2.3007688282413556,
      "grad_norm": 0.1573042869567871,
      "learning_rate": 2.3310945502499464e-05,
      "loss": 0.4982,
      "mean_token_accuracy": 0.8408615477383137,
      "num_tokens": 798107194.0,
      "step": 25007
    },
    {
      "entropy": 0.49664491415023804,
      "epoch": 2.300860834612797,
      "grad_norm": 0.15647441148757935,
      "learning_rate": 2.330787867635784e-05,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8485538363456726,
      "num_tokens": 798139077.0,
      "step": 25008
    },
    {
      "entropy": 0.5277047120034695,
      "epoch": 2.3009528409842384,
      "grad_norm": 0.1637730747461319,
      "learning_rate": 2.3304811850216213e-05,
      "loss": 0.5145,
      "mean_token_accuracy": 0.8366292156279087,
      "num_tokens": 798170849.0,
      "step": 25009
    },
    {
      "entropy": 0.5406338348984718,
      "epoch": 2.3010448473556795,
      "grad_norm": 0.1575423926115036,
      "learning_rate": 2.3301745024074588e-05,
      "loss": 0.5267,
      "mean_token_accuracy": 0.835480272769928,
      "num_tokens": 798203307.0,
      "step": 25010
    },
    {
      "entropy": 0.5491605685092509,
      "epoch": 2.3011368537271206,
      "grad_norm": 0.1648874282836914,
      "learning_rate": 2.329867819793296e-05,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8299431502819061,
      "num_tokens": 798235226.0,
      "step": 25011
    },
    {
      "entropy": 0.3882574064191431,
      "epoch": 2.3012288600985618,
      "grad_norm": 0.13931074738502502,
      "learning_rate": 2.3295611371791334e-05,
      "loss": 0.3831,
      "mean_token_accuracy": 0.8780979625880718,
      "num_tokens": 798267293.0,
      "step": 25012
    },
    {
      "entropy": 0.6018684739246964,
      "epoch": 2.301320866470003,
      "grad_norm": 0.170615553855896,
      "learning_rate": 2.329254454564971e-05,
      "loss": 0.5894,
      "mean_token_accuracy": 0.8134896121919155,
      "num_tokens": 798299409.0,
      "step": 25013
    },
    {
      "entropy": 0.5268351300619543,
      "epoch": 2.3014128728414445,
      "grad_norm": 0.16312454640865326,
      "learning_rate": 2.3289477719508083e-05,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8358889706432819,
      "num_tokens": 798330317.0,
      "step": 25014
    },
    {
      "entropy": 0.4698348673991859,
      "epoch": 2.3015048792128856,
      "grad_norm": 0.1551775187253952,
      "learning_rate": 2.3286410893366454e-05,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8586615100502968,
      "num_tokens": 798362371.0,
      "step": 25015
    },
    {
      "entropy": 0.47187452763319016,
      "epoch": 2.3015968855843267,
      "grad_norm": 0.15277709066867828,
      "learning_rate": 2.328334406722483e-05,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8553233034908772,
      "num_tokens": 798394464.0,
      "step": 25016
    },
    {
      "entropy": 0.542584921233356,
      "epoch": 2.301688891955768,
      "grad_norm": 0.15943118929862976,
      "learning_rate": 2.3280277241083204e-05,
      "loss": 0.535,
      "mean_token_accuracy": 0.8331965543329716,
      "num_tokens": 798427171.0,
      "step": 25017
    },
    {
      "entropy": 0.5363751072436571,
      "epoch": 2.301780898327209,
      "grad_norm": 0.1622680276632309,
      "learning_rate": 2.3277210414941578e-05,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8321875743567944,
      "num_tokens": 798459592.0,
      "step": 25018
    },
    {
      "entropy": 0.5356817087158561,
      "epoch": 2.3018729046986506,
      "grad_norm": 0.16223862767219543,
      "learning_rate": 2.3274143588799953e-05,
      "loss": 0.5249,
      "mean_token_accuracy": 0.8331377319991589,
      "num_tokens": 798491487.0,
      "step": 25019
    },
    {
      "entropy": 0.4157937918789685,
      "epoch": 2.3019649110700917,
      "grad_norm": 0.14884300529956818,
      "learning_rate": 2.3271076762658324e-05,
      "loss": 0.4043,
      "mean_token_accuracy": 0.8698815256357193,
      "num_tokens": 798523815.0,
      "step": 25020
    },
    {
      "entropy": 0.5230802316218615,
      "epoch": 2.302056917441533,
      "grad_norm": 0.1626794934272766,
      "learning_rate": 2.32680099365167e-05,
      "loss": 0.5255,
      "mean_token_accuracy": 0.8320540115237236,
      "num_tokens": 798556279.0,
      "step": 25021
    },
    {
      "entropy": 0.5540210418403149,
      "epoch": 2.302148923812974,
      "grad_norm": 0.16076040267944336,
      "learning_rate": 2.3264943110375074e-05,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8270393945276737,
      "num_tokens": 798588948.0,
      "step": 25022
    },
    {
      "entropy": 0.5801950562745333,
      "epoch": 2.302240930184415,
      "grad_norm": 0.17590190470218658,
      "learning_rate": 2.3261876284233448e-05,
      "loss": 0.5859,
      "mean_token_accuracy": 0.8155939243733883,
      "num_tokens": 798620152.0,
      "step": 25023
    },
    {
      "entropy": 0.4727565599605441,
      "epoch": 2.3023329365558567,
      "grad_norm": 0.1548701822757721,
      "learning_rate": 2.325880945809182e-05,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8519293144345284,
      "num_tokens": 798652754.0,
      "step": 25024
    },
    {
      "entropy": 0.559745317324996,
      "epoch": 2.302424942927298,
      "grad_norm": 0.16608642041683197,
      "learning_rate": 2.3255742631950194e-05,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8266131542623043,
      "num_tokens": 798684257.0,
      "step": 25025
    },
    {
      "entropy": 0.47721599601209164,
      "epoch": 2.302516949298739,
      "grad_norm": 0.16395002603530884,
      "learning_rate": 2.325267580580857e-05,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8517101593315601,
      "num_tokens": 798715523.0,
      "step": 25026
    },
    {
      "entropy": 0.5234392583370209,
      "epoch": 2.30260895567018,
      "grad_norm": 0.16691024601459503,
      "learning_rate": 2.3249608979666943e-05,
      "loss": 0.5067,
      "mean_token_accuracy": 0.8377096019685268,
      "num_tokens": 798746647.0,
      "step": 25027
    },
    {
      "entropy": 0.5228827223181725,
      "epoch": 2.3027009620416212,
      "grad_norm": 0.15966582298278809,
      "learning_rate": 2.3246542153525318e-05,
      "loss": 0.5164,
      "mean_token_accuracy": 0.8381769210100174,
      "num_tokens": 798778919.0,
      "step": 25028
    },
    {
      "entropy": 0.4710886203683913,
      "epoch": 2.302792968413063,
      "grad_norm": 0.15883924067020416,
      "learning_rate": 2.3243475327383693e-05,
      "loss": 0.4587,
      "mean_token_accuracy": 0.8528469391167164,
      "num_tokens": 798810842.0,
      "step": 25029
    },
    {
      "entropy": 0.600239546969533,
      "epoch": 2.302884974784504,
      "grad_norm": 0.17189782857894897,
      "learning_rate": 2.3240408501242067e-05,
      "loss": 0.6006,
      "mean_token_accuracy": 0.8126062527298927,
      "num_tokens": 798842369.0,
      "step": 25030
    },
    {
      "entropy": 0.45430911239236593,
      "epoch": 2.302976981155945,
      "grad_norm": 0.15923839807510376,
      "learning_rate": 2.3237341675100442e-05,
      "loss": 0.4292,
      "mean_token_accuracy": 0.8643707111477852,
      "num_tokens": 798873758.0,
      "step": 25031
    },
    {
      "entropy": 0.5584908965975046,
      "epoch": 2.303068987527386,
      "grad_norm": 0.16649048030376434,
      "learning_rate": 2.3234274848958813e-05,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8284660913050175,
      "num_tokens": 798906049.0,
      "step": 25032
    },
    {
      "entropy": 0.5353819448500872,
      "epoch": 2.3031609938988273,
      "grad_norm": 0.16170541942119598,
      "learning_rate": 2.3231208022817188e-05,
      "loss": 0.522,
      "mean_token_accuracy": 0.8380618430674076,
      "num_tokens": 798938043.0,
      "step": 25033
    },
    {
      "entropy": 0.5141765307635069,
      "epoch": 2.303253000270269,
      "grad_norm": 0.16010402143001556,
      "learning_rate": 2.3228141196675563e-05,
      "loss": 0.4935,
      "mean_token_accuracy": 0.8468645997345448,
      "num_tokens": 798969508.0,
      "step": 25034
    },
    {
      "entropy": 0.5798582714051008,
      "epoch": 2.30334500664171,
      "grad_norm": 0.16393455862998962,
      "learning_rate": 2.3225074370533937e-05,
      "loss": 0.5652,
      "mean_token_accuracy": 0.8227803371846676,
      "num_tokens": 799001783.0,
      "step": 25035
    },
    {
      "entropy": 0.6525840489193797,
      "epoch": 2.303437013013151,
      "grad_norm": 0.1774030476808548,
      "learning_rate": 2.3222007544392312e-05,
      "loss": 0.6416,
      "mean_token_accuracy": 0.7973799519240856,
      "num_tokens": 799033408.0,
      "step": 25036
    },
    {
      "entropy": 0.5309966504573822,
      "epoch": 2.3035290193845923,
      "grad_norm": 0.1660016030073166,
      "learning_rate": 2.3218940718250683e-05,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8362983539700508,
      "num_tokens": 799064881.0,
      "step": 25037
    },
    {
      "entropy": 0.4831692557781935,
      "epoch": 2.3036210257560334,
      "grad_norm": 0.1569977104663849,
      "learning_rate": 2.3215873892109058e-05,
      "loss": 0.4674,
      "mean_token_accuracy": 0.8479186743497849,
      "num_tokens": 799096483.0,
      "step": 25038
    },
    {
      "entropy": 0.543929498642683,
      "epoch": 2.303713032127475,
      "grad_norm": 0.16292299330234528,
      "learning_rate": 2.3212807065967432e-05,
      "loss": 0.5325,
      "mean_token_accuracy": 0.8310500755906105,
      "num_tokens": 799128419.0,
      "step": 25039
    },
    {
      "entropy": 0.45007783733308315,
      "epoch": 2.303805038498916,
      "grad_norm": 0.14647245407104492,
      "learning_rate": 2.3209740239825807e-05,
      "loss": 0.4286,
      "mean_token_accuracy": 0.8597122579813004,
      "num_tokens": 799160147.0,
      "step": 25040
    },
    {
      "entropy": 0.387152933049947,
      "epoch": 2.3038970448703573,
      "grad_norm": 0.14576949179172516,
      "learning_rate": 2.3206673413684178e-05,
      "loss": 0.3745,
      "mean_token_accuracy": 0.877670418471098,
      "num_tokens": 799191831.0,
      "step": 25041
    },
    {
      "entropy": 0.4603413389995694,
      "epoch": 2.3039890512417984,
      "grad_norm": 0.15220105648040771,
      "learning_rate": 2.3203606587542553e-05,
      "loss": 0.4576,
      "mean_token_accuracy": 0.857893168926239,
      "num_tokens": 799224175.0,
      "step": 25042
    },
    {
      "entropy": 0.5522324643097818,
      "epoch": 2.3040810576132396,
      "grad_norm": 0.1619063764810562,
      "learning_rate": 2.3200539761400928e-05,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8350371718406677,
      "num_tokens": 799255964.0,
      "step": 25043
    },
    {
      "entropy": 0.5455350838601589,
      "epoch": 2.304173063984681,
      "grad_norm": 0.1603131890296936,
      "learning_rate": 2.3197472935259302e-05,
      "loss": 0.5345,
      "mean_token_accuracy": 0.8343108855187893,
      "num_tokens": 799288578.0,
      "step": 25044
    },
    {
      "entropy": 0.37316273991018534,
      "epoch": 2.3042650703561223,
      "grad_norm": 0.1393585205078125,
      "learning_rate": 2.3194406109117673e-05,
      "loss": 0.3603,
      "mean_token_accuracy": 0.883416473865509,
      "num_tokens": 799320758.0,
      "step": 25045
    },
    {
      "entropy": 0.5099205302540213,
      "epoch": 2.3043570767275634,
      "grad_norm": 0.15679709613323212,
      "learning_rate": 2.3191339282976048e-05,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8415416143834591,
      "num_tokens": 799352318.0,
      "step": 25046
    },
    {
      "entropy": 0.5192297101020813,
      "epoch": 2.3044490830990045,
      "grad_norm": 0.1596827656030655,
      "learning_rate": 2.3188272456834423e-05,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8366102576255798,
      "num_tokens": 799384920.0,
      "step": 25047
    },
    {
      "entropy": 0.5520058549009264,
      "epoch": 2.3045410894704457,
      "grad_norm": 0.15722572803497314,
      "learning_rate": 2.3185205630692797e-05,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8331661336123943,
      "num_tokens": 799417505.0,
      "step": 25048
    },
    {
      "entropy": 0.5144195463508368,
      "epoch": 2.3046330958418872,
      "grad_norm": 0.15668156743049622,
      "learning_rate": 2.3182138804551172e-05,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8419928625226021,
      "num_tokens": 799449747.0,
      "step": 25049
    },
    {
      "entropy": 0.5897178892046213,
      "epoch": 2.3047251022133284,
      "grad_norm": 0.1691664755344391,
      "learning_rate": 2.3179071978409543e-05,
      "loss": 0.5822,
      "mean_token_accuracy": 0.8187034912407398,
      "num_tokens": 799481460.0,
      "step": 25050
    },
    {
      "entropy": 0.5493861918803304,
      "epoch": 2.3048171085847695,
      "grad_norm": 0.1646900773048401,
      "learning_rate": 2.3176005152267918e-05,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8321338444948196,
      "num_tokens": 799513740.0,
      "step": 25051
    },
    {
      "entropy": 0.5462896656244993,
      "epoch": 2.3049091149562106,
      "grad_norm": 0.1628929078578949,
      "learning_rate": 2.3172938326126293e-05,
      "loss": 0.5379,
      "mean_token_accuracy": 0.827936127781868,
      "num_tokens": 799546246.0,
      "step": 25052
    },
    {
      "entropy": 0.5287166219204664,
      "epoch": 2.305001121327652,
      "grad_norm": 0.16077464818954468,
      "learning_rate": 2.3169871499984667e-05,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8366714529693127,
      "num_tokens": 799577385.0,
      "step": 25053
    },
    {
      "entropy": 0.5033347178250551,
      "epoch": 2.3050931276990934,
      "grad_norm": 0.15845108032226562,
      "learning_rate": 2.316680467384304e-05,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8442805707454681,
      "num_tokens": 799609878.0,
      "step": 25054
    },
    {
      "entropy": 0.5419230042025447,
      "epoch": 2.3051851340705345,
      "grad_norm": 0.16616208851337433,
      "learning_rate": 2.3163737847701413e-05,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8347177654504776,
      "num_tokens": 799641214.0,
      "step": 25055
    },
    {
      "entropy": 0.5240773819386959,
      "epoch": 2.3052771404419756,
      "grad_norm": 0.16346341371536255,
      "learning_rate": 2.3160671021559788e-05,
      "loss": 0.5037,
      "mean_token_accuracy": 0.837718065828085,
      "num_tokens": 799672945.0,
      "step": 25056
    },
    {
      "entropy": 0.5986852869391441,
      "epoch": 2.3053691468134168,
      "grad_norm": 0.181003138422966,
      "learning_rate": 2.3157604195418162e-05,
      "loss": 0.589,
      "mean_token_accuracy": 0.8176357969641685,
      "num_tokens": 799703430.0,
      "step": 25057
    },
    {
      "entropy": 0.5232339166104794,
      "epoch": 2.305461153184858,
      "grad_norm": 0.16287831962108612,
      "learning_rate": 2.3154537369276537e-05,
      "loss": 0.5057,
      "mean_token_accuracy": 0.838965866714716,
      "num_tokens": 799734722.0,
      "step": 25058
    },
    {
      "entropy": 0.4712341530248523,
      "epoch": 2.3055531595562995,
      "grad_norm": 0.161799818277359,
      "learning_rate": 2.3151470543134912e-05,
      "loss": 0.4533,
      "mean_token_accuracy": 0.8530858419835567,
      "num_tokens": 799765759.0,
      "step": 25059
    },
    {
      "entropy": 0.4813461899757385,
      "epoch": 2.3056451659277406,
      "grad_norm": 0.15115392208099365,
      "learning_rate": 2.3148403716993286e-05,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8520457074046135,
      "num_tokens": 799798162.0,
      "step": 25060
    },
    {
      "entropy": 0.5240558022633195,
      "epoch": 2.3057371722991817,
      "grad_norm": 0.15972061455249786,
      "learning_rate": 2.314533689085166e-05,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8384142406284809,
      "num_tokens": 799829834.0,
      "step": 25061
    },
    {
      "entropy": 0.5789448907598853,
      "epoch": 2.305829178670623,
      "grad_norm": 0.16523559391498566,
      "learning_rate": 2.3142270064710032e-05,
      "loss": 0.5721,
      "mean_token_accuracy": 0.8228581845760345,
      "num_tokens": 799861815.0,
      "step": 25062
    },
    {
      "entropy": 0.5336922565475106,
      "epoch": 2.305921185042064,
      "grad_norm": 0.16176500916481018,
      "learning_rate": 2.3139203238568407e-05,
      "loss": 0.5191,
      "mean_token_accuracy": 0.8370240330696106,
      "num_tokens": 799893604.0,
      "step": 25063
    },
    {
      "entropy": 0.5950111523270607,
      "epoch": 2.3060131914135056,
      "grad_norm": 0.16951261460781097,
      "learning_rate": 2.313613641242678e-05,
      "loss": 0.5812,
      "mean_token_accuracy": 0.8161422237753868,
      "num_tokens": 799925663.0,
      "step": 25064
    },
    {
      "entropy": 0.4868598263710737,
      "epoch": 2.3061051977849467,
      "grad_norm": 0.15983861684799194,
      "learning_rate": 2.3133069586285156e-05,
      "loss": 0.474,
      "mean_token_accuracy": 0.8496010638773441,
      "num_tokens": 799956959.0,
      "step": 25065
    },
    {
      "entropy": 0.5109472097828984,
      "epoch": 2.306197204156388,
      "grad_norm": 0.16228094696998596,
      "learning_rate": 2.313000276014353e-05,
      "loss": 0.5098,
      "mean_token_accuracy": 0.8389973305165768,
      "num_tokens": 799988907.0,
      "step": 25066
    },
    {
      "entropy": 0.579601775854826,
      "epoch": 2.306289210527829,
      "grad_norm": 0.17023198306560516,
      "learning_rate": 2.3126935934001902e-05,
      "loss": 0.5702,
      "mean_token_accuracy": 0.8230413384735584,
      "num_tokens": 800020298.0,
      "step": 25067
    },
    {
      "entropy": 0.4476083414629102,
      "epoch": 2.30638121689927,
      "grad_norm": 0.1545162945985794,
      "learning_rate": 2.3123869107860277e-05,
      "loss": 0.4352,
      "mean_token_accuracy": 0.8632548376917839,
      "num_tokens": 800052260.0,
      "step": 25068
    },
    {
      "entropy": 0.44923875480890274,
      "epoch": 2.3064732232707117,
      "grad_norm": 0.1523221880197525,
      "learning_rate": 2.312080228171865e-05,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8623043410480022,
      "num_tokens": 800084019.0,
      "step": 25069
    },
    {
      "entropy": 0.5078659197315574,
      "epoch": 2.306565229642153,
      "grad_norm": 0.16141262650489807,
      "learning_rate": 2.3117735455577026e-05,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8392600379884243,
      "num_tokens": 800115153.0,
      "step": 25070
    },
    {
      "entropy": 0.573445507325232,
      "epoch": 2.306657236013594,
      "grad_norm": 0.16501395404338837,
      "learning_rate": 2.3114668629435397e-05,
      "loss": 0.5594,
      "mean_token_accuracy": 0.8241289518773556,
      "num_tokens": 800147833.0,
      "step": 25071
    },
    {
      "entropy": 0.4892126675695181,
      "epoch": 2.306749242385035,
      "grad_norm": 0.15432143211364746,
      "learning_rate": 2.3111601803293772e-05,
      "loss": 0.4603,
      "mean_token_accuracy": 0.8525551110506058,
      "num_tokens": 800179859.0,
      "step": 25072
    },
    {
      "entropy": 0.5024746144190431,
      "epoch": 2.3068412487564762,
      "grad_norm": 0.1525515913963318,
      "learning_rate": 2.3108534977152147e-05,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8433887846767902,
      "num_tokens": 800212259.0,
      "step": 25073
    },
    {
      "entropy": 0.5817473966162652,
      "epoch": 2.306933255127918,
      "grad_norm": 0.1677529215812683,
      "learning_rate": 2.310546815101052e-05,
      "loss": 0.5732,
      "mean_token_accuracy": 0.8226475566625595,
      "num_tokens": 800244247.0,
      "step": 25074
    },
    {
      "entropy": 0.5047985324636102,
      "epoch": 2.307025261499359,
      "grad_norm": 0.16188642382621765,
      "learning_rate": 2.3102401324868893e-05,
      "loss": 0.4948,
      "mean_token_accuracy": 0.8449255600571632,
      "num_tokens": 800276872.0,
      "step": 25075
    },
    {
      "entropy": 0.4642277415841818,
      "epoch": 2.3071172678708,
      "grad_norm": 0.15810075402259827,
      "learning_rate": 2.3099334498727267e-05,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8553129583597183,
      "num_tokens": 800309001.0,
      "step": 25076
    },
    {
      "entropy": 0.5177392568439245,
      "epoch": 2.307209274242241,
      "grad_norm": 0.1612432599067688,
      "learning_rate": 2.3096267672585642e-05,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8415494300425053,
      "num_tokens": 800340835.0,
      "step": 25077
    },
    {
      "entropy": 0.587585206143558,
      "epoch": 2.3073012806136823,
      "grad_norm": 0.16672943532466888,
      "learning_rate": 2.3093200846444017e-05,
      "loss": 0.5692,
      "mean_token_accuracy": 0.8212103135883808,
      "num_tokens": 800372630.0,
      "step": 25078
    },
    {
      "entropy": 0.5232138806022704,
      "epoch": 2.307393286985124,
      "grad_norm": 0.16347703337669373,
      "learning_rate": 2.309013402030239e-05,
      "loss": 0.4969,
      "mean_token_accuracy": 0.8420384041965008,
      "num_tokens": 800403015.0,
      "step": 25079
    },
    {
      "entropy": 0.5059494571760297,
      "epoch": 2.307485293356565,
      "grad_norm": 0.15956419706344604,
      "learning_rate": 2.3087067194160762e-05,
      "loss": 0.4844,
      "mean_token_accuracy": 0.8477950878441334,
      "num_tokens": 800435129.0,
      "step": 25080
    },
    {
      "entropy": 0.508128747344017,
      "epoch": 2.307577299728006,
      "grad_norm": 0.15494941174983978,
      "learning_rate": 2.3084000368019137e-05,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8470318280160427,
      "num_tokens": 800467327.0,
      "step": 25081
    },
    {
      "entropy": 0.5307963148225099,
      "epoch": 2.3076693060994473,
      "grad_norm": 0.1641799807548523,
      "learning_rate": 2.3080933541877512e-05,
      "loss": 0.5222,
      "mean_token_accuracy": 0.8385047502815723,
      "num_tokens": 800500029.0,
      "step": 25082
    },
    {
      "entropy": 0.5780248083174229,
      "epoch": 2.3077613124708884,
      "grad_norm": 0.17292927205562592,
      "learning_rate": 2.3077866715735886e-05,
      "loss": 0.5761,
      "mean_token_accuracy": 0.818902276456356,
      "num_tokens": 800531696.0,
      "step": 25083
    },
    {
      "entropy": 0.5908965896815062,
      "epoch": 2.30785331884233,
      "grad_norm": 0.16733981668949127,
      "learning_rate": 2.3074799889594258e-05,
      "loss": 0.5864,
      "mean_token_accuracy": 0.819146353751421,
      "num_tokens": 800564307.0,
      "step": 25084
    },
    {
      "entropy": 0.5115935318171978,
      "epoch": 2.307945325213771,
      "grad_norm": 0.1645452231168747,
      "learning_rate": 2.3071733063452632e-05,
      "loss": 0.5111,
      "mean_token_accuracy": 0.8378673121333122,
      "num_tokens": 800596455.0,
      "step": 25085
    },
    {
      "entropy": 0.5613667014986277,
      "epoch": 2.3080373315852123,
      "grad_norm": 0.16661587357521057,
      "learning_rate": 2.3068666237311007e-05,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8261423707008362,
      "num_tokens": 800627911.0,
      "step": 25086
    },
    {
      "entropy": 0.4991748573957011,
      "epoch": 2.3081293379566534,
      "grad_norm": 0.15190881490707397,
      "learning_rate": 2.306559941116938e-05,
      "loss": 0.4841,
      "mean_token_accuracy": 0.8465716391801834,
      "num_tokens": 800660619.0,
      "step": 25087
    },
    {
      "entropy": 0.5211432576179504,
      "epoch": 2.3082213443280946,
      "grad_norm": 0.15366488695144653,
      "learning_rate": 2.3062532585027756e-05,
      "loss": 0.4984,
      "mean_token_accuracy": 0.842158630490303,
      "num_tokens": 800692873.0,
      "step": 25088
    },
    {
      "entropy": 0.5556977866217494,
      "epoch": 2.308313350699536,
      "grad_norm": 0.16464035212993622,
      "learning_rate": 2.3059465758886127e-05,
      "loss": 0.546,
      "mean_token_accuracy": 0.8315174728631973,
      "num_tokens": 800725033.0,
      "step": 25089
    },
    {
      "entropy": 0.5368934394791722,
      "epoch": 2.3084053570709773,
      "grad_norm": 0.17196635901927948,
      "learning_rate": 2.3056398932744505e-05,
      "loss": 0.5108,
      "mean_token_accuracy": 0.8340272195637226,
      "num_tokens": 800755925.0,
      "step": 25090
    },
    {
      "entropy": 0.487304525449872,
      "epoch": 2.3084973634424184,
      "grad_norm": 0.15760914981365204,
      "learning_rate": 2.305333210660288e-05,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8506973460316658,
      "num_tokens": 800788062.0,
      "step": 25091
    },
    {
      "entropy": 0.4125743641052395,
      "epoch": 2.3085893698138595,
      "grad_norm": 0.15316878259181976,
      "learning_rate": 2.305026528046125e-05,
      "loss": 0.4069,
      "mean_token_accuracy": 0.8708319216966629,
      "num_tokens": 800820008.0,
      "step": 25092
    },
    {
      "entropy": 0.5863271867856383,
      "epoch": 2.3086813761853007,
      "grad_norm": 0.17088709771633148,
      "learning_rate": 2.3047198454319626e-05,
      "loss": 0.58,
      "mean_token_accuracy": 0.8173819109797478,
      "num_tokens": 800851608.0,
      "step": 25093
    },
    {
      "entropy": 0.5544433481991291,
      "epoch": 2.3087733825567422,
      "grad_norm": 0.1634879857301712,
      "learning_rate": 2.3044131628178e-05,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8315256498754025,
      "num_tokens": 800883184.0,
      "step": 25094
    },
    {
      "entropy": 0.5332716917619109,
      "epoch": 2.3088653889281834,
      "grad_norm": 0.16836494207382202,
      "learning_rate": 2.3041064802036375e-05,
      "loss": 0.518,
      "mean_token_accuracy": 0.8355391174554825,
      "num_tokens": 800915060.0,
      "step": 25095
    },
    {
      "entropy": 0.42099947039969265,
      "epoch": 2.3089573952996245,
      "grad_norm": 0.15256524085998535,
      "learning_rate": 2.303799797589475e-05,
      "loss": 0.4178,
      "mean_token_accuracy": 0.8670454695820808,
      "num_tokens": 800947423.0,
      "step": 25096
    },
    {
      "entropy": 0.560181487351656,
      "epoch": 2.3090494016710656,
      "grad_norm": 0.16556091606616974,
      "learning_rate": 2.303493114975312e-05,
      "loss": 0.5477,
      "mean_token_accuracy": 0.82674690335989,
      "num_tokens": 800978251.0,
      "step": 25097
    },
    {
      "entropy": 0.592619638890028,
      "epoch": 2.309141408042507,
      "grad_norm": 0.16346384584903717,
      "learning_rate": 2.3031864323611496e-05,
      "loss": 0.5796,
      "mean_token_accuracy": 0.8168842270970345,
      "num_tokens": 801010785.0,
      "step": 25098
    },
    {
      "entropy": 0.5172713659703732,
      "epoch": 2.3092334144139484,
      "grad_norm": 0.16130173206329346,
      "learning_rate": 2.302879749746987e-05,
      "loss": 0.4954,
      "mean_token_accuracy": 0.8435904085636139,
      "num_tokens": 801042731.0,
      "step": 25099
    },
    {
      "entropy": 0.5633503054268658,
      "epoch": 2.3093254207853895,
      "grad_norm": 0.16621722280979156,
      "learning_rate": 2.3025730671328245e-05,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8241169229149818,
      "num_tokens": 801075496.0,
      "step": 25100
    },
    {
      "entropy": 0.38763618003576994,
      "epoch": 2.3094174271568306,
      "grad_norm": 0.1506165713071823,
      "learning_rate": 2.3022663845186616e-05,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8817378431558609,
      "num_tokens": 801106738.0,
      "step": 25101
    },
    {
      "entropy": 0.4778105393052101,
      "epoch": 2.3095094335282718,
      "grad_norm": 0.15745659172534943,
      "learning_rate": 2.301959701904499e-05,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8532241769134998,
      "num_tokens": 801138297.0,
      "step": 25102
    },
    {
      "entropy": 0.5135765101294965,
      "epoch": 2.309601439899713,
      "grad_norm": 0.15642371773719788,
      "learning_rate": 2.3016530192903366e-05,
      "loss": 0.5046,
      "mean_token_accuracy": 0.8423734568059444,
      "num_tokens": 801170430.0,
      "step": 25103
    },
    {
      "entropy": 0.5253404015675187,
      "epoch": 2.3096934462711545,
      "grad_norm": 0.1604124754667282,
      "learning_rate": 2.301346336676174e-05,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8347527533769608,
      "num_tokens": 801202246.0,
      "step": 25104
    },
    {
      "entropy": 0.5203108263667673,
      "epoch": 2.3097854526425956,
      "grad_norm": 0.165006622672081,
      "learning_rate": 2.301039654062011e-05,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8386775590479374,
      "num_tokens": 801234595.0,
      "step": 25105
    },
    {
      "entropy": 0.5568141918629408,
      "epoch": 2.3098774590140367,
      "grad_norm": 0.1689203828573227,
      "learning_rate": 2.3007329714478486e-05,
      "loss": 0.5585,
      "mean_token_accuracy": 0.8231970071792603,
      "num_tokens": 801267070.0,
      "step": 25106
    },
    {
      "entropy": 0.5561237297952175,
      "epoch": 2.309969465385478,
      "grad_norm": 0.16523069143295288,
      "learning_rate": 2.300426288833686e-05,
      "loss": 0.5443,
      "mean_token_accuracy": 0.8281629048287868,
      "num_tokens": 801298704.0,
      "step": 25107
    },
    {
      "entropy": 0.5211903001181781,
      "epoch": 2.310061471756919,
      "grad_norm": 0.16045738756656647,
      "learning_rate": 2.3001196062195236e-05,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8400757722556591,
      "num_tokens": 801330956.0,
      "step": 25108
    },
    {
      "entropy": 0.5183837926015258,
      "epoch": 2.3101534781283606,
      "grad_norm": 0.16051141917705536,
      "learning_rate": 2.299812923605361e-05,
      "loss": 0.507,
      "mean_token_accuracy": 0.8391328193247318,
      "num_tokens": 801362088.0,
      "step": 25109
    },
    {
      "entropy": 0.6264244932681322,
      "epoch": 2.3102454844998017,
      "grad_norm": 0.17164823412895203,
      "learning_rate": 2.299506240991198e-05,
      "loss": 0.6106,
      "mean_token_accuracy": 0.8078250661492348,
      "num_tokens": 801394228.0,
      "step": 25110
    },
    {
      "entropy": 0.47894513607025146,
      "epoch": 2.310337490871243,
      "grad_norm": 0.15480270981788635,
      "learning_rate": 2.2991995583770356e-05,
      "loss": 0.477,
      "mean_token_accuracy": 0.8480948358774185,
      "num_tokens": 801426931.0,
      "step": 25111
    },
    {
      "entropy": 0.5370767810381949,
      "epoch": 2.310429497242684,
      "grad_norm": 0.16356615722179413,
      "learning_rate": 2.298892875762873e-05,
      "loss": 0.5267,
      "mean_token_accuracy": 0.836113266646862,
      "num_tokens": 801458767.0,
      "step": 25112
    },
    {
      "entropy": 0.521291671320796,
      "epoch": 2.310521503614125,
      "grad_norm": 0.16527247428894043,
      "learning_rate": 2.2985861931487105e-05,
      "loss": 0.5113,
      "mean_token_accuracy": 0.8367268815636635,
      "num_tokens": 801489833.0,
      "step": 25113
    },
    {
      "entropy": 0.5205995147116482,
      "epoch": 2.3106135099855667,
      "grad_norm": 0.16198702156543732,
      "learning_rate": 2.2982795105345477e-05,
      "loss": 0.5053,
      "mean_token_accuracy": 0.839653629809618,
      "num_tokens": 801521860.0,
      "step": 25114
    },
    {
      "entropy": 0.603307643905282,
      "epoch": 2.310705516357008,
      "grad_norm": 0.16802026331424713,
      "learning_rate": 2.297972827920385e-05,
      "loss": 0.5796,
      "mean_token_accuracy": 0.8132590837776661,
      "num_tokens": 801553436.0,
      "step": 25115
    },
    {
      "entropy": 0.5241159675642848,
      "epoch": 2.310797522728449,
      "grad_norm": 0.17274372279644012,
      "learning_rate": 2.2976661453062226e-05,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8352809324860573,
      "num_tokens": 801584959.0,
      "step": 25116
    },
    {
      "entropy": 0.4449713643407449,
      "epoch": 2.31088952909989,
      "grad_norm": 0.15693499147891998,
      "learning_rate": 2.29735946269206e-05,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8596192747354507,
      "num_tokens": 801616342.0,
      "step": 25117
    },
    {
      "entropy": 0.5675513260066509,
      "epoch": 2.3109815354713312,
      "grad_norm": 0.16450296342372894,
      "learning_rate": 2.2970527800778975e-05,
      "loss": 0.5535,
      "mean_token_accuracy": 0.8245981074869633,
      "num_tokens": 801648615.0,
      "step": 25118
    },
    {
      "entropy": 0.5148673821240664,
      "epoch": 2.311073541842773,
      "grad_norm": 0.16148169338703156,
      "learning_rate": 2.2967460974637347e-05,
      "loss": 0.5075,
      "mean_token_accuracy": 0.8366937376558781,
      "num_tokens": 801680512.0,
      "step": 25119
    },
    {
      "entropy": 0.48999247048050165,
      "epoch": 2.311165548214214,
      "grad_norm": 0.16096745431423187,
      "learning_rate": 2.2964394148495725e-05,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8475928716361523,
      "num_tokens": 801712677.0,
      "step": 25120
    },
    {
      "entropy": 0.5093931341543794,
      "epoch": 2.311257554585655,
      "grad_norm": 0.15802393853664398,
      "learning_rate": 2.29613273223541e-05,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8453201241791248,
      "num_tokens": 801745279.0,
      "step": 25121
    },
    {
      "entropy": 0.5526436194777489,
      "epoch": 2.311349560957096,
      "grad_norm": 0.16581712663173676,
      "learning_rate": 2.295826049621247e-05,
      "loss": 0.5388,
      "mean_token_accuracy": 0.8307658173143864,
      "num_tokens": 801776957.0,
      "step": 25122
    },
    {
      "entropy": 0.5462177693843842,
      "epoch": 2.3114415673285373,
      "grad_norm": 0.16720299422740936,
      "learning_rate": 2.2955193670070845e-05,
      "loss": 0.533,
      "mean_token_accuracy": 0.8345778696238995,
      "num_tokens": 801809073.0,
      "step": 25123
    },
    {
      "entropy": 0.5126902544870973,
      "epoch": 2.311533573699979,
      "grad_norm": 0.15832528471946716,
      "learning_rate": 2.295212684392922e-05,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8422737047076225,
      "num_tokens": 801840428.0,
      "step": 25124
    },
    {
      "entropy": 0.5468550957739353,
      "epoch": 2.31162558007142,
      "grad_norm": 0.16504491865634918,
      "learning_rate": 2.2949060017787594e-05,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8308915793895721,
      "num_tokens": 801872018.0,
      "step": 25125
    },
    {
      "entropy": 0.48075492307543755,
      "epoch": 2.311717586442861,
      "grad_norm": 0.1567101925611496,
      "learning_rate": 2.294599319164597e-05,
      "loss": 0.4779,
      "mean_token_accuracy": 0.847786296159029,
      "num_tokens": 801903866.0,
      "step": 25126
    },
    {
      "entropy": 0.4677213979884982,
      "epoch": 2.3118095928143023,
      "grad_norm": 0.15359170734882355,
      "learning_rate": 2.294292636550434e-05,
      "loss": 0.4563,
      "mean_token_accuracy": 0.85204141959548,
      "num_tokens": 801935714.0,
      "step": 25127
    },
    {
      "entropy": 0.5399137884378433,
      "epoch": 2.3119015991857434,
      "grad_norm": 0.16316568851470947,
      "learning_rate": 2.2939859539362715e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.835649199783802,
      "num_tokens": 801966680.0,
      "step": 25128
    },
    {
      "entropy": 0.5589268300682306,
      "epoch": 2.311993605557185,
      "grad_norm": 0.16556961834430695,
      "learning_rate": 2.293679271322109e-05,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8262122198939323,
      "num_tokens": 801998420.0,
      "step": 25129
    },
    {
      "entropy": 0.5071427263319492,
      "epoch": 2.312085611928626,
      "grad_norm": 0.15784604847431183,
      "learning_rate": 2.2933725887079464e-05,
      "loss": 0.4929,
      "mean_token_accuracy": 0.8423387296497822,
      "num_tokens": 802030869.0,
      "step": 25130
    },
    {
      "entropy": 0.47032454796135426,
      "epoch": 2.3121776183000673,
      "grad_norm": 0.15408097207546234,
      "learning_rate": 2.2930659060937836e-05,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8562861308455467,
      "num_tokens": 802063347.0,
      "step": 25131
    },
    {
      "entropy": 0.5528808543458581,
      "epoch": 2.3122696246715084,
      "grad_norm": 0.16859236359596252,
      "learning_rate": 2.292759223479621e-05,
      "loss": 0.5568,
      "mean_token_accuracy": 0.8311533033847809,
      "num_tokens": 802094586.0,
      "step": 25132
    },
    {
      "entropy": 0.5495193712413311,
      "epoch": 2.3123616310429496,
      "grad_norm": 0.16659700870513916,
      "learning_rate": 2.2924525408654585e-05,
      "loss": 0.5445,
      "mean_token_accuracy": 0.8276583068072796,
      "num_tokens": 802127157.0,
      "step": 25133
    },
    {
      "entropy": 0.5940435333177447,
      "epoch": 2.312453637414391,
      "grad_norm": 0.1683439314365387,
      "learning_rate": 2.292145858251296e-05,
      "loss": 0.5763,
      "mean_token_accuracy": 0.81727085262537,
      "num_tokens": 802159008.0,
      "step": 25134
    },
    {
      "entropy": 0.550460031372495,
      "epoch": 2.3125456437858323,
      "grad_norm": 0.155048668384552,
      "learning_rate": 2.291839175637133e-05,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8316154293715954,
      "num_tokens": 802191446.0,
      "step": 25135
    },
    {
      "entropy": 0.5777712520211935,
      "epoch": 2.3126376501572734,
      "grad_norm": 0.16863323748111725,
      "learning_rate": 2.2915324930229705e-05,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8207007758319378,
      "num_tokens": 802223340.0,
      "step": 25136
    },
    {
      "entropy": 0.5270719965919852,
      "epoch": 2.3127296565287145,
      "grad_norm": 0.15947872400283813,
      "learning_rate": 2.291225810408808e-05,
      "loss": 0.5108,
      "mean_token_accuracy": 0.8391508534550667,
      "num_tokens": 802255406.0,
      "step": 25137
    },
    {
      "entropy": 0.5631100181490183,
      "epoch": 2.3128216629001557,
      "grad_norm": 0.16693232953548431,
      "learning_rate": 2.2909191277946455e-05,
      "loss": 0.5525,
      "mean_token_accuracy": 0.8253905475139618,
      "num_tokens": 802287187.0,
      "step": 25138
    },
    {
      "entropy": 0.4276406718418002,
      "epoch": 2.3129136692715973,
      "grad_norm": 0.14336615800857544,
      "learning_rate": 2.290612445180483e-05,
      "loss": 0.4112,
      "mean_token_accuracy": 0.8653998672962189,
      "num_tokens": 802319313.0,
      "step": 25139
    },
    {
      "entropy": 0.5831938814371824,
      "epoch": 2.3130056756430384,
      "grad_norm": 0.16829758882522583,
      "learning_rate": 2.29030576256632e-05,
      "loss": 0.5635,
      "mean_token_accuracy": 0.8225157931447029,
      "num_tokens": 802349933.0,
      "step": 25140
    },
    {
      "entropy": 0.3779500452801585,
      "epoch": 2.3130976820144795,
      "grad_norm": 0.13876792788505554,
      "learning_rate": 2.2899990799521575e-05,
      "loss": 0.361,
      "mean_token_accuracy": 0.8832645788788795,
      "num_tokens": 802381776.0,
      "step": 25141
    },
    {
      "entropy": 0.5711993556469679,
      "epoch": 2.3131896883859207,
      "grad_norm": 0.1650266945362091,
      "learning_rate": 2.289692397337995e-05,
      "loss": 0.5652,
      "mean_token_accuracy": 0.8223155215382576,
      "num_tokens": 802413891.0,
      "step": 25142
    },
    {
      "entropy": 0.5325160175561905,
      "epoch": 2.313281694757362,
      "grad_norm": 0.16441422700881958,
      "learning_rate": 2.2893857147238325e-05,
      "loss": 0.5239,
      "mean_token_accuracy": 0.839646726846695,
      "num_tokens": 802445629.0,
      "step": 25143
    },
    {
      "entropy": 0.4805815778672695,
      "epoch": 2.3133737011288034,
      "grad_norm": 0.15367460250854492,
      "learning_rate": 2.2890790321096696e-05,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8484525382518768,
      "num_tokens": 802478052.0,
      "step": 25144
    },
    {
      "entropy": 0.5247154654935002,
      "epoch": 2.3134657075002445,
      "grad_norm": 0.15713748335838318,
      "learning_rate": 2.288772349495507e-05,
      "loss": 0.508,
      "mean_token_accuracy": 0.8384093306958675,
      "num_tokens": 802510271.0,
      "step": 25145
    },
    {
      "entropy": 0.5391512140631676,
      "epoch": 2.3135577138716856,
      "grad_norm": 0.17083293199539185,
      "learning_rate": 2.2884656668813445e-05,
      "loss": 0.5387,
      "mean_token_accuracy": 0.8292112201452255,
      "num_tokens": 802541798.0,
      "step": 25146
    },
    {
      "entropy": 0.5033819107338786,
      "epoch": 2.3136497202431268,
      "grad_norm": 0.1632751226425171,
      "learning_rate": 2.288158984267182e-05,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8437625132501125,
      "num_tokens": 802573206.0,
      "step": 25147
    },
    {
      "entropy": 0.47403513081371784,
      "epoch": 2.313741726614568,
      "grad_norm": 0.15363970398902893,
      "learning_rate": 2.2878523016530194e-05,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8484965600073338,
      "num_tokens": 802605438.0,
      "step": 25148
    },
    {
      "entropy": 0.5413219192996621,
      "epoch": 2.3138337329860095,
      "grad_norm": 0.17305922508239746,
      "learning_rate": 2.2875456190388566e-05,
      "loss": 0.5371,
      "mean_token_accuracy": 0.8319943808019161,
      "num_tokens": 802636881.0,
      "step": 25149
    },
    {
      "entropy": 0.44457135756965727,
      "epoch": 2.3139257393574506,
      "grad_norm": 0.14872771501541138,
      "learning_rate": 2.287238936424694e-05,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8625831939280033,
      "num_tokens": 802668905.0,
      "step": 25150
    },
    {
      "entropy": 0.5673285154625773,
      "epoch": 2.3140177457288917,
      "grad_norm": 0.1656716912984848,
      "learning_rate": 2.286932253810532e-05,
      "loss": 0.5505,
      "mean_token_accuracy": 0.8265413828194141,
      "num_tokens": 802701415.0,
      "step": 25151
    },
    {
      "entropy": 0.5186083680018783,
      "epoch": 2.314109752100333,
      "grad_norm": 0.16621993482112885,
      "learning_rate": 2.286625571196369e-05,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8398144245147705,
      "num_tokens": 802732939.0,
      "step": 25152
    },
    {
      "entropy": 0.6011839937418699,
      "epoch": 2.314201758471774,
      "grad_norm": 0.17094004154205322,
      "learning_rate": 2.2863188885822064e-05,
      "loss": 0.5848,
      "mean_token_accuracy": 0.8159437291324139,
      "num_tokens": 802765217.0,
      "step": 25153
    },
    {
      "entropy": 0.4336349132936448,
      "epoch": 2.3142937648432156,
      "grad_norm": 0.15089410543441772,
      "learning_rate": 2.286012205968044e-05,
      "loss": 0.4191,
      "mean_token_accuracy": 0.8659469448029995,
      "num_tokens": 802797179.0,
      "step": 25154
    },
    {
      "entropy": 0.4773937058635056,
      "epoch": 2.3143857712146567,
      "grad_norm": 0.1634649783372879,
      "learning_rate": 2.2857055233538814e-05,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8531189672648907,
      "num_tokens": 802829365.0,
      "step": 25155
    },
    {
      "entropy": 0.4303214121609926,
      "epoch": 2.314477777586098,
      "grad_norm": 0.14404502511024475,
      "learning_rate": 2.2853988407397188e-05,
      "loss": 0.4182,
      "mean_token_accuracy": 0.8676212020218372,
      "num_tokens": 802861608.0,
      "step": 25156
    },
    {
      "entropy": 0.4928422295488417,
      "epoch": 2.314569783957539,
      "grad_norm": 0.1635335385799408,
      "learning_rate": 2.285092158125556e-05,
      "loss": 0.4898,
      "mean_token_accuracy": 0.8466669730842113,
      "num_tokens": 802893066.0,
      "step": 25157
    },
    {
      "entropy": 0.4756348291411996,
      "epoch": 2.31466179032898,
      "grad_norm": 0.15469782054424286,
      "learning_rate": 2.2847854755113934e-05,
      "loss": 0.4665,
      "mean_token_accuracy": 0.853121105581522,
      "num_tokens": 802925287.0,
      "step": 25158
    },
    {
      "entropy": 0.48440923262387514,
      "epoch": 2.3147537967004217,
      "grad_norm": 0.15464279055595398,
      "learning_rate": 2.284478792897231e-05,
      "loss": 0.4668,
      "mean_token_accuracy": 0.8514199703931808,
      "num_tokens": 802957099.0,
      "step": 25159
    },
    {
      "entropy": 0.3486238259356469,
      "epoch": 2.314845803071863,
      "grad_norm": 0.1383250206708908,
      "learning_rate": 2.2841721102830683e-05,
      "loss": 0.335,
      "mean_token_accuracy": 0.8914608024060726,
      "num_tokens": 802989323.0,
      "step": 25160
    },
    {
      "entropy": 0.5298748761415482,
      "epoch": 2.314937809443304,
      "grad_norm": 0.17019428312778473,
      "learning_rate": 2.2838654276689055e-05,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8377180993556976,
      "num_tokens": 803020205.0,
      "step": 25161
    },
    {
      "entropy": 0.5788375707343221,
      "epoch": 2.315029815814745,
      "grad_norm": 0.16985264420509338,
      "learning_rate": 2.283558745054743e-05,
      "loss": 0.5614,
      "mean_token_accuracy": 0.8222916126251221,
      "num_tokens": 803051067.0,
      "step": 25162
    },
    {
      "entropy": 0.5745835993438959,
      "epoch": 2.3151218221861862,
      "grad_norm": 0.16652950644493103,
      "learning_rate": 2.2832520624405804e-05,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8271581195294857,
      "num_tokens": 803082708.0,
      "step": 25163
    },
    {
      "entropy": 0.5139192901551723,
      "epoch": 2.315213828557628,
      "grad_norm": 0.16506099700927734,
      "learning_rate": 2.282945379826418e-05,
      "loss": 0.5094,
      "mean_token_accuracy": 0.836444303393364,
      "num_tokens": 803114570.0,
      "step": 25164
    },
    {
      "entropy": 0.5005513913929462,
      "epoch": 2.315305834929069,
      "grad_norm": 0.15524831414222717,
      "learning_rate": 2.282638697212255e-05,
      "loss": 0.4969,
      "mean_token_accuracy": 0.8409740440547466,
      "num_tokens": 803146766.0,
      "step": 25165
    },
    {
      "entropy": 0.49887625221163034,
      "epoch": 2.31539784130051,
      "grad_norm": 0.15886279940605164,
      "learning_rate": 2.2823320145980925e-05,
      "loss": 0.4782,
      "mean_token_accuracy": 0.8482112735509872,
      "num_tokens": 803178880.0,
      "step": 25166
    },
    {
      "entropy": 0.521731112152338,
      "epoch": 2.315489847671951,
      "grad_norm": 0.1641351282596588,
      "learning_rate": 2.28202533198393e-05,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8404214754700661,
      "num_tokens": 803210312.0,
      "step": 25167
    },
    {
      "entropy": 0.4901375784538686,
      "epoch": 2.3155818540433923,
      "grad_norm": 0.16230282187461853,
      "learning_rate": 2.2817186493697674e-05,
      "loss": 0.4781,
      "mean_token_accuracy": 0.8490314856171608,
      "num_tokens": 803242447.0,
      "step": 25168
    },
    {
      "entropy": 0.5237335981801152,
      "epoch": 2.315673860414834,
      "grad_norm": 0.16495417058467865,
      "learning_rate": 2.281411966755605e-05,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8387499488890171,
      "num_tokens": 803273794.0,
      "step": 25169
    },
    {
      "entropy": 0.5307688377797604,
      "epoch": 2.315765866786275,
      "grad_norm": 0.17171306908130646,
      "learning_rate": 2.281105284141442e-05,
      "loss": 0.52,
      "mean_token_accuracy": 0.8317863047122955,
      "num_tokens": 803304733.0,
      "step": 25170
    },
    {
      "entropy": 0.5729847783222795,
      "epoch": 2.315857873157716,
      "grad_norm": 0.16252827644348145,
      "learning_rate": 2.2807986015272794e-05,
      "loss": 0.562,
      "mean_token_accuracy": 0.8230206407606602,
      "num_tokens": 803336977.0,
      "step": 25171
    },
    {
      "entropy": 0.521293324418366,
      "epoch": 2.3159498795291573,
      "grad_norm": 0.15487059950828552,
      "learning_rate": 2.280491918913117e-05,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8405033349990845,
      "num_tokens": 803368507.0,
      "step": 25172
    },
    {
      "entropy": 0.482604016084224,
      "epoch": 2.3160418859005985,
      "grad_norm": 0.15437108278274536,
      "learning_rate": 2.2801852362989544e-05,
      "loss": 0.4771,
      "mean_token_accuracy": 0.8489557802677155,
      "num_tokens": 803400582.0,
      "step": 25173
    },
    {
      "entropy": 0.4989618135150522,
      "epoch": 2.31613389227204,
      "grad_norm": 0.15195603668689728,
      "learning_rate": 2.2798785536847915e-05,
      "loss": 0.4905,
      "mean_token_accuracy": 0.8433856666088104,
      "num_tokens": 803433335.0,
      "step": 25174
    },
    {
      "entropy": 0.5777119947597384,
      "epoch": 2.316225898643481,
      "grad_norm": 0.1696212887763977,
      "learning_rate": 2.279571871070629e-05,
      "loss": 0.5701,
      "mean_token_accuracy": 0.8216627277433872,
      "num_tokens": 803464546.0,
      "step": 25175
    },
    {
      "entropy": 0.45725591108202934,
      "epoch": 2.3163179050149223,
      "grad_norm": 0.15522736310958862,
      "learning_rate": 2.2792651884564664e-05,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8574786894023418,
      "num_tokens": 803496500.0,
      "step": 25176
    },
    {
      "entropy": 0.5501332115381956,
      "epoch": 2.3164099113863634,
      "grad_norm": 0.16792437434196472,
      "learning_rate": 2.278958505842304e-05,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8301997557282448,
      "num_tokens": 803526963.0,
      "step": 25177
    },
    {
      "entropy": 0.5353647300507873,
      "epoch": 2.3165019177578046,
      "grad_norm": 0.16323302686214447,
      "learning_rate": 2.2786518232281413e-05,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8381466493010521,
      "num_tokens": 803558953.0,
      "step": 25178
    },
    {
      "entropy": 0.5830913949757814,
      "epoch": 2.316593924129246,
      "grad_norm": 0.1709592640399933,
      "learning_rate": 2.2783451406139785e-05,
      "loss": 0.5775,
      "mean_token_accuracy": 0.8214461989700794,
      "num_tokens": 803590811.0,
      "step": 25179
    },
    {
      "entropy": 0.5657001622021198,
      "epoch": 2.3166859305006873,
      "grad_norm": 0.17079155147075653,
      "learning_rate": 2.278038457999816e-05,
      "loss": 0.5608,
      "mean_token_accuracy": 0.8216197565197945,
      "num_tokens": 803622542.0,
      "step": 25180
    },
    {
      "entropy": 0.4258003393188119,
      "epoch": 2.3167779368721284,
      "grad_norm": 0.14782236516475677,
      "learning_rate": 2.2777317753856537e-05,
      "loss": 0.4164,
      "mean_token_accuracy": 0.8702001012861729,
      "num_tokens": 803654499.0,
      "step": 25181
    },
    {
      "entropy": 0.5043971594423056,
      "epoch": 2.3168699432435695,
      "grad_norm": 0.1560724377632141,
      "learning_rate": 2.277425092771491e-05,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8443129844963551,
      "num_tokens": 803685567.0,
      "step": 25182
    },
    {
      "entropy": 0.47293971106410027,
      "epoch": 2.3169619496150107,
      "grad_norm": 0.1504366248846054,
      "learning_rate": 2.2771184101573283e-05,
      "loss": 0.4618,
      "mean_token_accuracy": 0.8570609018206596,
      "num_tokens": 803718032.0,
      "step": 25183
    },
    {
      "entropy": 0.49766878224909306,
      "epoch": 2.3170539559864523,
      "grad_norm": 0.16081944108009338,
      "learning_rate": 2.2768117275431658e-05,
      "loss": 0.479,
      "mean_token_accuracy": 0.8464933075010777,
      "num_tokens": 803749261.0,
      "step": 25184
    },
    {
      "entropy": 0.5400229971855879,
      "epoch": 2.3171459623578934,
      "grad_norm": 0.16796182096004486,
      "learning_rate": 2.2765050449290033e-05,
      "loss": 0.5241,
      "mean_token_accuracy": 0.8304631374776363,
      "num_tokens": 803780801.0,
      "step": 25185
    },
    {
      "entropy": 0.4860871620476246,
      "epoch": 2.3172379687293345,
      "grad_norm": 0.16589924693107605,
      "learning_rate": 2.2761983623148407e-05,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8508602678775787,
      "num_tokens": 803812379.0,
      "step": 25186
    },
    {
      "entropy": 0.46558887138962746,
      "epoch": 2.3173299751007757,
      "grad_norm": 0.15394796431064606,
      "learning_rate": 2.275891679700678e-05,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8549349345266819,
      "num_tokens": 803844505.0,
      "step": 25187
    },
    {
      "entropy": 0.5054906373843551,
      "epoch": 2.317421981472217,
      "grad_norm": 0.15732145309448242,
      "learning_rate": 2.2755849970865153e-05,
      "loss": 0.4883,
      "mean_token_accuracy": 0.8428374566137791,
      "num_tokens": 803876752.0,
      "step": 25188
    },
    {
      "entropy": 0.5174683043733239,
      "epoch": 2.3175139878436584,
      "grad_norm": 0.16533352434635162,
      "learning_rate": 2.2752783144723528e-05,
      "loss": 0.5048,
      "mean_token_accuracy": 0.841178122907877,
      "num_tokens": 803908067.0,
      "step": 25189
    },
    {
      "entropy": 0.541555866599083,
      "epoch": 2.3176059942150995,
      "grad_norm": 0.16224367916584015,
      "learning_rate": 2.2749716318581902e-05,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8345691710710526,
      "num_tokens": 803940302.0,
      "step": 25190
    },
    {
      "entropy": 0.48255342547781765,
      "epoch": 2.3176980005865406,
      "grad_norm": 0.1609201282262802,
      "learning_rate": 2.2746649492440274e-05,
      "loss": 0.4737,
      "mean_token_accuracy": 0.8529666215181351,
      "num_tokens": 803971945.0,
      "step": 25191
    },
    {
      "entropy": 0.503317579627037,
      "epoch": 2.3177900069579818,
      "grad_norm": 0.160977303981781,
      "learning_rate": 2.274358266629865e-05,
      "loss": 0.4948,
      "mean_token_accuracy": 0.8402017876505852,
      "num_tokens": 804003426.0,
      "step": 25192
    },
    {
      "entropy": 0.5029942309483886,
      "epoch": 2.317882013329423,
      "grad_norm": 0.16202040016651154,
      "learning_rate": 2.2740515840157023e-05,
      "loss": 0.5008,
      "mean_token_accuracy": 0.8405500985682011,
      "num_tokens": 804035715.0,
      "step": 25193
    },
    {
      "entropy": 0.555846506729722,
      "epoch": 2.3179740197008645,
      "grad_norm": 0.16355912387371063,
      "learning_rate": 2.2737449014015398e-05,
      "loss": 0.5381,
      "mean_token_accuracy": 0.8291481770575047,
      "num_tokens": 804067936.0,
      "step": 25194
    },
    {
      "entropy": 0.567995335906744,
      "epoch": 2.3180660260723056,
      "grad_norm": 0.17002061009407043,
      "learning_rate": 2.273438218787377e-05,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8217899091541767,
      "num_tokens": 804099982.0,
      "step": 25195
    },
    {
      "entropy": 0.5723052378743887,
      "epoch": 2.3181580324437467,
      "grad_norm": 0.17129474878311157,
      "learning_rate": 2.2731315361732144e-05,
      "loss": 0.5629,
      "mean_token_accuracy": 0.8216893412172794,
      "num_tokens": 804131660.0,
      "step": 25196
    },
    {
      "entropy": 0.5043212556047365,
      "epoch": 2.318250038815188,
      "grad_norm": 0.16054761409759521,
      "learning_rate": 2.2728248535590518e-05,
      "loss": 0.4969,
      "mean_token_accuracy": 0.84087048843503,
      "num_tokens": 804163288.0,
      "step": 25197
    },
    {
      "entropy": 0.566611603833735,
      "epoch": 2.318342045186629,
      "grad_norm": 0.16592496633529663,
      "learning_rate": 2.2725181709448893e-05,
      "loss": 0.5455,
      "mean_token_accuracy": 0.8283709958195686,
      "num_tokens": 804195427.0,
      "step": 25198
    },
    {
      "entropy": 0.5048708296380937,
      "epoch": 2.3184340515580706,
      "grad_norm": 0.16156592965126038,
      "learning_rate": 2.2722114883307268e-05,
      "loss": 0.4923,
      "mean_token_accuracy": 0.8451996073126793,
      "num_tokens": 804226971.0,
      "step": 25199
    },
    {
      "entropy": 0.5387132558971643,
      "epoch": 2.3185260579295117,
      "grad_norm": 0.16572754085063934,
      "learning_rate": 2.271904805716564e-05,
      "loss": 0.521,
      "mean_token_accuracy": 0.8344298861920834,
      "num_tokens": 804258376.0,
      "step": 25200
    },
    {
      "entropy": 0.5559044014662504,
      "epoch": 2.318618064300953,
      "grad_norm": 0.16048184037208557,
      "learning_rate": 2.2715981231024013e-05,
      "loss": 0.5399,
      "mean_token_accuracy": 0.8287432864308357,
      "num_tokens": 804291144.0,
      "step": 25201
    },
    {
      "entropy": 0.5248086391948164,
      "epoch": 2.318710070672394,
      "grad_norm": 0.1595417559146881,
      "learning_rate": 2.2712914404882388e-05,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8378454335033894,
      "num_tokens": 804323607.0,
      "step": 25202
    },
    {
      "entropy": 0.4534313874319196,
      "epoch": 2.318802077043835,
      "grad_norm": 0.15389803051948547,
      "learning_rate": 2.2709847578740763e-05,
      "loss": 0.4479,
      "mean_token_accuracy": 0.8574574515223503,
      "num_tokens": 804356009.0,
      "step": 25203
    },
    {
      "entropy": 0.6064914558082819,
      "epoch": 2.3188940834152767,
      "grad_norm": 0.17053990066051483,
      "learning_rate": 2.2706780752599134e-05,
      "loss": 0.5913,
      "mean_token_accuracy": 0.8114228844642639,
      "num_tokens": 804387665.0,
      "step": 25204
    },
    {
      "entropy": 0.5652018263936043,
      "epoch": 2.318986089786718,
      "grad_norm": 0.16565604507923126,
      "learning_rate": 2.270371392645751e-05,
      "loss": 0.5608,
      "mean_token_accuracy": 0.8259492926299572,
      "num_tokens": 804419995.0,
      "step": 25205
    },
    {
      "entropy": 0.47455823980271816,
      "epoch": 2.319078096158159,
      "grad_norm": 0.15407375991344452,
      "learning_rate": 2.2700647100315883e-05,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8539148680865765,
      "num_tokens": 804451619.0,
      "step": 25206
    },
    {
      "entropy": 0.4750592513009906,
      "epoch": 2.3191701025296,
      "grad_norm": 0.15382924675941467,
      "learning_rate": 2.2697580274174258e-05,
      "loss": 0.4693,
      "mean_token_accuracy": 0.8519616574048996,
      "num_tokens": 804483518.0,
      "step": 25207
    },
    {
      "entropy": 0.5098492819815874,
      "epoch": 2.3192621089010412,
      "grad_norm": 0.16363582015037537,
      "learning_rate": 2.2694513448032633e-05,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8387256860733032,
      "num_tokens": 804515426.0,
      "step": 25208
    },
    {
      "entropy": 0.5752709470689297,
      "epoch": 2.319354115272483,
      "grad_norm": 0.17038366198539734,
      "learning_rate": 2.2691446621891004e-05,
      "loss": 0.5647,
      "mean_token_accuracy": 0.8205769583582878,
      "num_tokens": 804547617.0,
      "step": 25209
    },
    {
      "entropy": 0.5358101788442582,
      "epoch": 2.319446121643924,
      "grad_norm": 0.1612497866153717,
      "learning_rate": 2.268837979574938e-05,
      "loss": 0.5244,
      "mean_token_accuracy": 0.8357739374041557,
      "num_tokens": 804580225.0,
      "step": 25210
    },
    {
      "entropy": 0.4933468885719776,
      "epoch": 2.319538128015365,
      "grad_norm": 0.1575235277414322,
      "learning_rate": 2.2685312969607753e-05,
      "loss": 0.4739,
      "mean_token_accuracy": 0.8487297296524048,
      "num_tokens": 804612556.0,
      "step": 25211
    },
    {
      "entropy": 0.4131561703979969,
      "epoch": 2.319630134386806,
      "grad_norm": 0.1419726312160492,
      "learning_rate": 2.2682246143466128e-05,
      "loss": 0.4021,
      "mean_token_accuracy": 0.8745204694569111,
      "num_tokens": 804645232.0,
      "step": 25212
    },
    {
      "entropy": 0.5612348727881908,
      "epoch": 2.3197221407582473,
      "grad_norm": 0.16777609288692474,
      "learning_rate": 2.2679179317324502e-05,
      "loss": 0.5525,
      "mean_token_accuracy": 0.8262250982224941,
      "num_tokens": 804677516.0,
      "step": 25213
    },
    {
      "entropy": 0.464638854842633,
      "epoch": 2.319814147129689,
      "grad_norm": 0.1539989858865738,
      "learning_rate": 2.2676112491182877e-05,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8553230725228786,
      "num_tokens": 804709772.0,
      "step": 25214
    },
    {
      "entropy": 0.55633051879704,
      "epoch": 2.31990615350113,
      "grad_norm": 0.16518056392669678,
      "learning_rate": 2.2673045665041252e-05,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8306923322379589,
      "num_tokens": 804742159.0,
      "step": 25215
    },
    {
      "entropy": 0.6377935139462352,
      "epoch": 2.319998159872571,
      "grad_norm": 0.17439530789852142,
      "learning_rate": 2.2669978838899626e-05,
      "loss": 0.6311,
      "mean_token_accuracy": 0.8020104803144932,
      "num_tokens": 804774737.0,
      "step": 25216
    },
    {
      "entropy": 0.5594222955405712,
      "epoch": 2.3200901662440123,
      "grad_norm": 0.16630472242832184,
      "learning_rate": 2.2666912012757998e-05,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8242732994258404,
      "num_tokens": 804807182.0,
      "step": 25217
    },
    {
      "entropy": 0.5390727538615465,
      "epoch": 2.3201821726154535,
      "grad_norm": 0.1583838313817978,
      "learning_rate": 2.2663845186616372e-05,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8355869799852371,
      "num_tokens": 804839347.0,
      "step": 25218
    },
    {
      "entropy": 0.5026420904323459,
      "epoch": 2.320274178986895,
      "grad_norm": 0.1621468961238861,
      "learning_rate": 2.2660778360474747e-05,
      "loss": 0.4871,
      "mean_token_accuracy": 0.843279805034399,
      "num_tokens": 804871331.0,
      "step": 25219
    },
    {
      "entropy": 0.5730227753520012,
      "epoch": 2.320366185358336,
      "grad_norm": 0.1693664938211441,
      "learning_rate": 2.265771153433312e-05,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8244024887681007,
      "num_tokens": 804903709.0,
      "step": 25220
    },
    {
      "entropy": 0.5221093064174056,
      "epoch": 2.3204581917297773,
      "grad_norm": 0.16093853116035461,
      "learning_rate": 2.2654644708191493e-05,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8386682122945786,
      "num_tokens": 804935082.0,
      "step": 25221
    },
    {
      "entropy": 0.46300294902175665,
      "epoch": 2.3205501981012184,
      "grad_norm": 0.1518557220697403,
      "learning_rate": 2.2651577882049867e-05,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8565740138292313,
      "num_tokens": 804967206.0,
      "step": 25222
    },
    {
      "entropy": 0.4746914766728878,
      "epoch": 2.3206422044726596,
      "grad_norm": 0.15734148025512695,
      "learning_rate": 2.2648511055908242e-05,
      "loss": 0.4626,
      "mean_token_accuracy": 0.8528572730720043,
      "num_tokens": 804999239.0,
      "step": 25223
    },
    {
      "entropy": 0.4472365779802203,
      "epoch": 2.320734210844101,
      "grad_norm": 0.14559955894947052,
      "learning_rate": 2.2645444229766617e-05,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8642861023545265,
      "num_tokens": 805031346.0,
      "step": 25224
    },
    {
      "entropy": 0.5803207764402032,
      "epoch": 2.3208262172155423,
      "grad_norm": 0.17105905711650848,
      "learning_rate": 2.2642377403624988e-05,
      "loss": 0.5736,
      "mean_token_accuracy": 0.8174317255616188,
      "num_tokens": 805063179.0,
      "step": 25225
    },
    {
      "entropy": 0.5662807431071997,
      "epoch": 2.3209182235869834,
      "grad_norm": 0.16659127175807953,
      "learning_rate": 2.2639310577483363e-05,
      "loss": 0.5446,
      "mean_token_accuracy": 0.8260573744773865,
      "num_tokens": 805094570.0,
      "step": 25226
    },
    {
      "entropy": 0.5299295638687909,
      "epoch": 2.3210102299584245,
      "grad_norm": 0.15694959461688995,
      "learning_rate": 2.2636243751341737e-05,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8376584127545357,
      "num_tokens": 805126859.0,
      "step": 25227
    },
    {
      "entropy": 0.6318655046634376,
      "epoch": 2.3211022363298657,
      "grad_norm": 0.1709318906068802,
      "learning_rate": 2.2633176925200112e-05,
      "loss": 0.6103,
      "mean_token_accuracy": 0.8089192993938923,
      "num_tokens": 805158935.0,
      "step": 25228
    },
    {
      "entropy": 0.5166059620678425,
      "epoch": 2.3211942427013073,
      "grad_norm": 0.16296926140785217,
      "learning_rate": 2.2630110099058487e-05,
      "loss": 0.4988,
      "mean_token_accuracy": 0.8398605845868587,
      "num_tokens": 805191586.0,
      "step": 25229
    },
    {
      "entropy": 0.5596910640597343,
      "epoch": 2.3212862490727484,
      "grad_norm": 0.16387653350830078,
      "learning_rate": 2.2627043272916858e-05,
      "loss": 0.5374,
      "mean_token_accuracy": 0.831169381737709,
      "num_tokens": 805223804.0,
      "step": 25230
    },
    {
      "entropy": 0.42087836319115013,
      "epoch": 2.3213782554441895,
      "grad_norm": 0.14565923810005188,
      "learning_rate": 2.2623976446775233e-05,
      "loss": 0.4008,
      "mean_token_accuracy": 0.8735095374286175,
      "num_tokens": 805255204.0,
      "step": 25231
    },
    {
      "entropy": 0.5071633188053966,
      "epoch": 2.3214702618156307,
      "grad_norm": 0.16076688468456268,
      "learning_rate": 2.2620909620633607e-05,
      "loss": 0.4906,
      "mean_token_accuracy": 0.8414338231086731,
      "num_tokens": 805287167.0,
      "step": 25232
    },
    {
      "entropy": 0.522416764870286,
      "epoch": 2.321562268187072,
      "grad_norm": 0.16898776590824127,
      "learning_rate": 2.2617842794491982e-05,
      "loss": 0.5189,
      "mean_token_accuracy": 0.8345462158322334,
      "num_tokens": 805318488.0,
      "step": 25233
    },
    {
      "entropy": 0.464611092582345,
      "epoch": 2.3216542745585134,
      "grad_norm": 0.15734177827835083,
      "learning_rate": 2.2614775968350353e-05,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8551323600113392,
      "num_tokens": 805350167.0,
      "step": 25234
    },
    {
      "entropy": 0.5167939066886902,
      "epoch": 2.3217462809299545,
      "grad_norm": 0.15985943377017975,
      "learning_rate": 2.2611709142208728e-05,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8363000825047493,
      "num_tokens": 805382094.0,
      "step": 25235
    },
    {
      "entropy": 0.46233461587689817,
      "epoch": 2.3218382873013956,
      "grad_norm": 0.15320970118045807,
      "learning_rate": 2.2608642316067102e-05,
      "loss": 0.4452,
      "mean_token_accuracy": 0.8590850755572319,
      "num_tokens": 805413661.0,
      "step": 25236
    },
    {
      "entropy": 0.46655827295035124,
      "epoch": 2.3219302936728368,
      "grad_norm": 0.15763907134532928,
      "learning_rate": 2.2605575489925477e-05,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8527010269463062,
      "num_tokens": 805445766.0,
      "step": 25237
    },
    {
      "entropy": 0.4692130731418729,
      "epoch": 2.322022300044278,
      "grad_norm": 0.15447746217250824,
      "learning_rate": 2.260250866378385e-05,
      "loss": 0.4583,
      "mean_token_accuracy": 0.85067493095994,
      "num_tokens": 805477474.0,
      "step": 25238
    },
    {
      "entropy": 0.5769457519054413,
      "epoch": 2.3221143064157195,
      "grad_norm": 0.1706840842962265,
      "learning_rate": 2.2599441837642223e-05,
      "loss": 0.568,
      "mean_token_accuracy": 0.8169513642787933,
      "num_tokens": 805508783.0,
      "step": 25239
    },
    {
      "entropy": 0.4845670238137245,
      "epoch": 2.3222063127871606,
      "grad_norm": 0.15856824815273285,
      "learning_rate": 2.2596375011500598e-05,
      "loss": 0.4738,
      "mean_token_accuracy": 0.8478328511118889,
      "num_tokens": 805541023.0,
      "step": 25240
    },
    {
      "entropy": 0.5740445703268051,
      "epoch": 2.3222983191586017,
      "grad_norm": 0.16654926538467407,
      "learning_rate": 2.2593308185358972e-05,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8246989548206329,
      "num_tokens": 805572815.0,
      "step": 25241
    },
    {
      "entropy": 0.45032198866829276,
      "epoch": 2.322390325530043,
      "grad_norm": 0.15144847333431244,
      "learning_rate": 2.2590241359217347e-05,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8597037196159363,
      "num_tokens": 805605149.0,
      "step": 25242
    },
    {
      "entropy": 0.53102639131248,
      "epoch": 2.322482331901484,
      "grad_norm": 0.16614970564842224,
      "learning_rate": 2.258717453307572e-05,
      "loss": 0.528,
      "mean_token_accuracy": 0.8312471136450768,
      "num_tokens": 805636056.0,
      "step": 25243
    },
    {
      "entropy": 0.5520358122885227,
      "epoch": 2.3225743382729256,
      "grad_norm": 0.1632205694913864,
      "learning_rate": 2.2584107706934096e-05,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8308334536850452,
      "num_tokens": 805667739.0,
      "step": 25244
    },
    {
      "entropy": 0.4419072913005948,
      "epoch": 2.3226663446443667,
      "grad_norm": 0.15129660069942474,
      "learning_rate": 2.258104088079247e-05,
      "loss": 0.4364,
      "mean_token_accuracy": 0.8613159954547882,
      "num_tokens": 805699346.0,
      "step": 25245
    },
    {
      "entropy": 0.49611949105747044,
      "epoch": 2.322758351015808,
      "grad_norm": 0.1576724648475647,
      "learning_rate": 2.2577974054650845e-05,
      "loss": 0.4932,
      "mean_token_accuracy": 0.8436410874128342,
      "num_tokens": 805731996.0,
      "step": 25246
    },
    {
      "entropy": 0.5840349914506078,
      "epoch": 2.322850357387249,
      "grad_norm": 0.17149372398853302,
      "learning_rate": 2.2574907228509217e-05,
      "loss": 0.5726,
      "mean_token_accuracy": 0.8171418569982052,
      "num_tokens": 805763362.0,
      "step": 25247
    },
    {
      "entropy": 0.5203036088496447,
      "epoch": 2.32294236375869,
      "grad_norm": 0.1557326763868332,
      "learning_rate": 2.257184040236759e-05,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8417237140238285,
      "num_tokens": 805795778.0,
      "step": 25248
    },
    {
      "entropy": 0.36814146023243666,
      "epoch": 2.3230343701301317,
      "grad_norm": 0.13802173733711243,
      "learning_rate": 2.2568773576225966e-05,
      "loss": 0.3492,
      "mean_token_accuracy": 0.8866328820586205,
      "num_tokens": 805828546.0,
      "step": 25249
    },
    {
      "entropy": 0.48979244008660316,
      "epoch": 2.323126376501573,
      "grad_norm": 0.15016281604766846,
      "learning_rate": 2.256570675008434e-05,
      "loss": 0.4841,
      "mean_token_accuracy": 0.844223253428936,
      "num_tokens": 805861314.0,
      "step": 25250
    },
    {
      "entropy": 0.435586123727262,
      "epoch": 2.323218382873014,
      "grad_norm": 0.15053527057170868,
      "learning_rate": 2.2562639923942712e-05,
      "loss": 0.426,
      "mean_token_accuracy": 0.865400493144989,
      "num_tokens": 805893389.0,
      "step": 25251
    },
    {
      "entropy": 0.49870393809396774,
      "epoch": 2.323310389244455,
      "grad_norm": 0.15321825444698334,
      "learning_rate": 2.2559573097801087e-05,
      "loss": 0.481,
      "mean_token_accuracy": 0.8471186980605125,
      "num_tokens": 805925752.0,
      "step": 25252
    },
    {
      "entropy": 0.5479178335517645,
      "epoch": 2.3234023956158962,
      "grad_norm": 0.1635541468858719,
      "learning_rate": 2.255650627165946e-05,
      "loss": 0.5372,
      "mean_token_accuracy": 0.8283802904188633,
      "num_tokens": 805958041.0,
      "step": 25253
    },
    {
      "entropy": 0.5665535856969655,
      "epoch": 2.323494401987338,
      "grad_norm": 0.1662689447402954,
      "learning_rate": 2.2553439445517836e-05,
      "loss": 0.5365,
      "mean_token_accuracy": 0.8308361023664474,
      "num_tokens": 805989540.0,
      "step": 25254
    },
    {
      "entropy": 0.6274078041315079,
      "epoch": 2.323586408358779,
      "grad_norm": 0.17443795502185822,
      "learning_rate": 2.2550372619376207e-05,
      "loss": 0.6215,
      "mean_token_accuracy": 0.805731613188982,
      "num_tokens": 806021566.0,
      "step": 25255
    },
    {
      "entropy": 0.46160740591585636,
      "epoch": 2.32367841473022,
      "grad_norm": 0.15064309537410736,
      "learning_rate": 2.2547305793234582e-05,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8563450872898102,
      "num_tokens": 806053227.0,
      "step": 25256
    },
    {
      "entropy": 0.6326303295791149,
      "epoch": 2.323770421101661,
      "grad_norm": 0.1777363270521164,
      "learning_rate": 2.2544238967092956e-05,
      "loss": 0.6302,
      "mean_token_accuracy": 0.7998597174882889,
      "num_tokens": 806084596.0,
      "step": 25257
    },
    {
      "entropy": 0.5275838477537036,
      "epoch": 2.3238624274731023,
      "grad_norm": 0.15417209267616272,
      "learning_rate": 2.254117214095133e-05,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8420024998486042,
      "num_tokens": 806116976.0,
      "step": 25258
    },
    {
      "entropy": 0.5397563241422176,
      "epoch": 2.323954433844544,
      "grad_norm": 0.15948766469955444,
      "learning_rate": 2.2538105314809706e-05,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8353096209466457,
      "num_tokens": 806148722.0,
      "step": 25259
    },
    {
      "entropy": 0.5337205659598112,
      "epoch": 2.324046440215985,
      "grad_norm": 0.1616434007883072,
      "learning_rate": 2.2535038488668077e-05,
      "loss": 0.5261,
      "mean_token_accuracy": 0.8327205292880535,
      "num_tokens": 806181256.0,
      "step": 25260
    },
    {
      "entropy": 0.5627286774106324,
      "epoch": 2.324138446587426,
      "grad_norm": 0.16881346702575684,
      "learning_rate": 2.253197166252645e-05,
      "loss": 0.5463,
      "mean_token_accuracy": 0.8255730718374252,
      "num_tokens": 806211952.0,
      "step": 25261
    },
    {
      "entropy": 0.5993702607229352,
      "epoch": 2.3242304529588673,
      "grad_norm": 0.1677011251449585,
      "learning_rate": 2.2528904836384826e-05,
      "loss": 0.5724,
      "mean_token_accuracy": 0.8206595927476883,
      "num_tokens": 806243457.0,
      "step": 25262
    },
    {
      "entropy": 0.5014312716666609,
      "epoch": 2.3243224593303085,
      "grad_norm": 0.16331541538238525,
      "learning_rate": 2.25258380102432e-05,
      "loss": 0.4943,
      "mean_token_accuracy": 0.8437696397304535,
      "num_tokens": 806275829.0,
      "step": 25263
    },
    {
      "entropy": 0.5875299340113997,
      "epoch": 2.32441446570175,
      "grad_norm": 0.17071625590324402,
      "learning_rate": 2.2522771184101572e-05,
      "loss": 0.5769,
      "mean_token_accuracy": 0.8170618042349815,
      "num_tokens": 806307893.0,
      "step": 25264
    },
    {
      "entropy": 0.5991534586064517,
      "epoch": 2.324506472073191,
      "grad_norm": 0.17375700175762177,
      "learning_rate": 2.2519704357959947e-05,
      "loss": 0.5892,
      "mean_token_accuracy": 0.8186375424265862,
      "num_tokens": 806339788.0,
      "step": 25265
    },
    {
      "entropy": 0.5850745467469096,
      "epoch": 2.3245984784446323,
      "grad_norm": 0.1685900241136551,
      "learning_rate": 2.251663753181832e-05,
      "loss": 0.5908,
      "mean_token_accuracy": 0.8155606500804424,
      "num_tokens": 806371516.0,
      "step": 25266
    },
    {
      "entropy": 0.5905821146443486,
      "epoch": 2.3246904848160734,
      "grad_norm": 0.170180544257164,
      "learning_rate": 2.2513570705676696e-05,
      "loss": 0.5837,
      "mean_token_accuracy": 0.8147497475147247,
      "num_tokens": 806403263.0,
      "step": 25267
    },
    {
      "entropy": 0.5744648696854711,
      "epoch": 2.3247824911875146,
      "grad_norm": 0.1662142276763916,
      "learning_rate": 2.251050387953507e-05,
      "loss": 0.57,
      "mean_token_accuracy": 0.8175493255257607,
      "num_tokens": 806434870.0,
      "step": 25268
    },
    {
      "entropy": 0.46060498245060444,
      "epoch": 2.324874497558956,
      "grad_norm": 0.1557784229516983,
      "learning_rate": 2.2507437053393442e-05,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8559170812368393,
      "num_tokens": 806466763.0,
      "step": 25269
    },
    {
      "entropy": 0.5618395917117596,
      "epoch": 2.3249665039303973,
      "grad_norm": 0.1663181483745575,
      "learning_rate": 2.2504370227251817e-05,
      "loss": 0.5354,
      "mean_token_accuracy": 0.827781118452549,
      "num_tokens": 806498566.0,
      "step": 25270
    },
    {
      "entropy": 0.517556032165885,
      "epoch": 2.3250585103018384,
      "grad_norm": 0.15603850781917572,
      "learning_rate": 2.250130340111019e-05,
      "loss": 0.505,
      "mean_token_accuracy": 0.8395987264811993,
      "num_tokens": 806530673.0,
      "step": 25271
    },
    {
      "entropy": 0.5302988439798355,
      "epoch": 2.3251505166732795,
      "grad_norm": 0.16300927102565765,
      "learning_rate": 2.2498236574968566e-05,
      "loss": 0.5194,
      "mean_token_accuracy": 0.8366815745830536,
      "num_tokens": 806562806.0,
      "step": 25272
    },
    {
      "entropy": 0.46990113286301494,
      "epoch": 2.3252425230447207,
      "grad_norm": 0.1551915407180786,
      "learning_rate": 2.249516974882694e-05,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8520937263965607,
      "num_tokens": 806594685.0,
      "step": 25273
    },
    {
      "entropy": 0.6602779142558575,
      "epoch": 2.3253345294161623,
      "grad_norm": 0.17204445600509644,
      "learning_rate": 2.2492102922685315e-05,
      "loss": 0.6493,
      "mean_token_accuracy": 0.7993394732475281,
      "num_tokens": 806627367.0,
      "step": 25274
    },
    {
      "entropy": 0.49688850925303996,
      "epoch": 2.3254265357876034,
      "grad_norm": 0.16139920055866241,
      "learning_rate": 2.248903609654369e-05,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8481481336057186,
      "num_tokens": 806659751.0,
      "step": 25275
    },
    {
      "entropy": 0.585121663287282,
      "epoch": 2.3255185421590445,
      "grad_norm": 0.1731739044189453,
      "learning_rate": 2.2485969270402065e-05,
      "loss": 0.5755,
      "mean_token_accuracy": 0.8209249973297119,
      "num_tokens": 806691958.0,
      "step": 25276
    },
    {
      "entropy": 0.5589048620313406,
      "epoch": 2.3256105485304857,
      "grad_norm": 0.1637868732213974,
      "learning_rate": 2.2482902444260436e-05,
      "loss": 0.5399,
      "mean_token_accuracy": 0.8304075486958027,
      "num_tokens": 806723980.0,
      "step": 25277
    },
    {
      "entropy": 0.4906876692548394,
      "epoch": 2.325702554901927,
      "grad_norm": 0.15873867273330688,
      "learning_rate": 2.247983561811881e-05,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8460990339517593,
      "num_tokens": 806756387.0,
      "step": 25278
    },
    {
      "entropy": 0.49782656878232956,
      "epoch": 2.3257945612733684,
      "grad_norm": 0.15222205221652985,
      "learning_rate": 2.2476768791977185e-05,
      "loss": 0.483,
      "mean_token_accuracy": 0.847039233893156,
      "num_tokens": 806787822.0,
      "step": 25279
    },
    {
      "entropy": 0.5107954195700586,
      "epoch": 2.3258865676448095,
      "grad_norm": 0.15749944746494293,
      "learning_rate": 2.247370196583556e-05,
      "loss": 0.482,
      "mean_token_accuracy": 0.8484946340322495,
      "num_tokens": 806819775.0,
      "step": 25280
    },
    {
      "entropy": 0.4805318620055914,
      "epoch": 2.3259785740162506,
      "grad_norm": 0.15720731019973755,
      "learning_rate": 2.247063513969393e-05,
      "loss": 0.457,
      "mean_token_accuracy": 0.8536339625716209,
      "num_tokens": 806851719.0,
      "step": 25281
    },
    {
      "entropy": 0.4913817858323455,
      "epoch": 2.3260705803876918,
      "grad_norm": 0.15927647054195404,
      "learning_rate": 2.2467568313552306e-05,
      "loss": 0.4803,
      "mean_token_accuracy": 0.8485964499413967,
      "num_tokens": 806884158.0,
      "step": 25282
    },
    {
      "entropy": 0.5139999995008111,
      "epoch": 2.326162586759133,
      "grad_norm": 0.1595560610294342,
      "learning_rate": 2.246450148741068e-05,
      "loss": 0.4842,
      "mean_token_accuracy": 0.8421072699129581,
      "num_tokens": 806915829.0,
      "step": 25283
    },
    {
      "entropy": 0.43805744498968124,
      "epoch": 2.3262545931305745,
      "grad_norm": 0.1503380537033081,
      "learning_rate": 2.2461434661269055e-05,
      "loss": 0.4241,
      "mean_token_accuracy": 0.8636017814278603,
      "num_tokens": 806947835.0,
      "step": 25284
    },
    {
      "entropy": 0.4500849745236337,
      "epoch": 2.3263465995020156,
      "grad_norm": 0.1552901715040207,
      "learning_rate": 2.2458367835127426e-05,
      "loss": 0.4426,
      "mean_token_accuracy": 0.8570306897163391,
      "num_tokens": 806979513.0,
      "step": 25285
    },
    {
      "entropy": 0.5296015292406082,
      "epoch": 2.3264386058734567,
      "grad_norm": 0.1633971929550171,
      "learning_rate": 2.24553010089858e-05,
      "loss": 0.5243,
      "mean_token_accuracy": 0.8338333889842033,
      "num_tokens": 807011910.0,
      "step": 25286
    },
    {
      "entropy": 0.4961011167615652,
      "epoch": 2.326530612244898,
      "grad_norm": 0.1619192361831665,
      "learning_rate": 2.2452234182844176e-05,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8443731442093849,
      "num_tokens": 807043734.0,
      "step": 25287
    },
    {
      "entropy": 0.47992356820032,
      "epoch": 2.326622618616339,
      "grad_norm": 0.15655048191547394,
      "learning_rate": 2.244916735670255e-05,
      "loss": 0.4556,
      "mean_token_accuracy": 0.8539511151611805,
      "num_tokens": 807074867.0,
      "step": 25288
    },
    {
      "entropy": 0.4702442013658583,
      "epoch": 2.3267146249877806,
      "grad_norm": 0.1598350554704666,
      "learning_rate": 2.2446100530560925e-05,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8560150377452374,
      "num_tokens": 807106802.0,
      "step": 25289
    },
    {
      "entropy": 0.549662621691823,
      "epoch": 2.3268066313592217,
      "grad_norm": 0.17233514785766602,
      "learning_rate": 2.2443033704419296e-05,
      "loss": 0.5534,
      "mean_token_accuracy": 0.8245845809578896,
      "num_tokens": 807137421.0,
      "step": 25290
    },
    {
      "entropy": 0.5134260524064302,
      "epoch": 2.326898637730663,
      "grad_norm": 0.15803903341293335,
      "learning_rate": 2.243996687827767e-05,
      "loss": 0.4993,
      "mean_token_accuracy": 0.840189665555954,
      "num_tokens": 807169435.0,
      "step": 25291
    },
    {
      "entropy": 0.5787823712453246,
      "epoch": 2.326990644102104,
      "grad_norm": 0.16712325811386108,
      "learning_rate": 2.2436900052136045e-05,
      "loss": 0.5741,
      "mean_token_accuracy": 0.8169124685227871,
      "num_tokens": 807201830.0,
      "step": 25292
    },
    {
      "entropy": 0.4480204628780484,
      "epoch": 2.327082650473545,
      "grad_norm": 0.14973084628582,
      "learning_rate": 2.243383322599442e-05,
      "loss": 0.4384,
      "mean_token_accuracy": 0.8609937280416489,
      "num_tokens": 807234014.0,
      "step": 25293
    },
    {
      "entropy": 0.5487115727737546,
      "epoch": 2.3271746568449867,
      "grad_norm": 0.16410048305988312,
      "learning_rate": 2.243076639985279e-05,
      "loss": 0.5214,
      "mean_token_accuracy": 0.8376666754484177,
      "num_tokens": 807266226.0,
      "step": 25294
    },
    {
      "entropy": 0.4467361057177186,
      "epoch": 2.327266663216428,
      "grad_norm": 0.15358476340770721,
      "learning_rate": 2.2427699573711166e-05,
      "loss": 0.4315,
      "mean_token_accuracy": 0.8608063459396362,
      "num_tokens": 807298622.0,
      "step": 25295
    },
    {
      "entropy": 0.42931118328124285,
      "epoch": 2.327358669587869,
      "grad_norm": 0.14715297520160675,
      "learning_rate": 2.242463274756954e-05,
      "loss": 0.4118,
      "mean_token_accuracy": 0.8691650964319706,
      "num_tokens": 807330827.0,
      "step": 25296
    },
    {
      "entropy": 0.5383354439400136,
      "epoch": 2.32745067595931,
      "grad_norm": 0.16130311787128448,
      "learning_rate": 2.2421565921427915e-05,
      "loss": 0.5215,
      "mean_token_accuracy": 0.8363952040672302,
      "num_tokens": 807362676.0,
      "step": 25297
    },
    {
      "entropy": 0.5635715089738369,
      "epoch": 2.3275426823307512,
      "grad_norm": 0.16118952631950378,
      "learning_rate": 2.241849909528629e-05,
      "loss": 0.5427,
      "mean_token_accuracy": 0.8258894607424736,
      "num_tokens": 807394581.0,
      "step": 25298
    },
    {
      "entropy": 0.4670289261266589,
      "epoch": 2.327634688702193,
      "grad_norm": 0.15736955404281616,
      "learning_rate": 2.241543226914466e-05,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8580460734665394,
      "num_tokens": 807426988.0,
      "step": 25299
    },
    {
      "entropy": 0.5363371041603386,
      "epoch": 2.327726695073634,
      "grad_norm": 0.16444867849349976,
      "learning_rate": 2.2412365443003036e-05,
      "loss": 0.5234,
      "mean_token_accuracy": 0.8357298225164413,
      "num_tokens": 807458734.0,
      "step": 25300
    },
    {
      "entropy": 0.5020619736751541,
      "epoch": 2.327818701445075,
      "grad_norm": 0.15556035935878754,
      "learning_rate": 2.240929861686141e-05,
      "loss": 0.4932,
      "mean_token_accuracy": 0.844909630715847,
      "num_tokens": 807490823.0,
      "step": 25301
    },
    {
      "entropy": 0.5483325943350792,
      "epoch": 2.327910707816516,
      "grad_norm": 0.16969871520996094,
      "learning_rate": 2.2406231790719785e-05,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8270541094243526,
      "num_tokens": 807521997.0,
      "step": 25302
    },
    {
      "entropy": 0.5333770802244544,
      "epoch": 2.3280027141879573,
      "grad_norm": 0.15837524831295013,
      "learning_rate": 2.240316496457816e-05,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8366302251815796,
      "num_tokens": 807554234.0,
      "step": 25303
    },
    {
      "entropy": 0.45480289217084646,
      "epoch": 2.328094720559399,
      "grad_norm": 0.15819667279720306,
      "learning_rate": 2.2400098138436534e-05,
      "loss": 0.45,
      "mean_token_accuracy": 0.8545272536575794,
      "num_tokens": 807585234.0,
      "step": 25304
    },
    {
      "entropy": 0.5847405027598143,
      "epoch": 2.32818672693084,
      "grad_norm": 0.16886873543262482,
      "learning_rate": 2.239703131229491e-05,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8190629854798317,
      "num_tokens": 807616071.0,
      "step": 25305
    },
    {
      "entropy": 0.5408702362328768,
      "epoch": 2.328278733302281,
      "grad_norm": 0.16609954833984375,
      "learning_rate": 2.2393964486153284e-05,
      "loss": 0.525,
      "mean_token_accuracy": 0.8323307037353516,
      "num_tokens": 807647932.0,
      "step": 25306
    },
    {
      "entropy": 0.49477491015568376,
      "epoch": 2.3283707396737223,
      "grad_norm": 0.16345322132110596,
      "learning_rate": 2.2390897660011655e-05,
      "loss": 0.4851,
      "mean_token_accuracy": 0.8461481966078281,
      "num_tokens": 807680209.0,
      "step": 25307
    },
    {
      "entropy": 0.5070097837597132,
      "epoch": 2.3284627460451635,
      "grad_norm": 0.15643036365509033,
      "learning_rate": 2.238783083387003e-05,
      "loss": 0.5011,
      "mean_token_accuracy": 0.8401672020554543,
      "num_tokens": 807712159.0,
      "step": 25308
    },
    {
      "entropy": 0.5581168457865715,
      "epoch": 2.328554752416605,
      "grad_norm": 0.1660395711660385,
      "learning_rate": 2.2384764007728404e-05,
      "loss": 0.5628,
      "mean_token_accuracy": 0.8240561932325363,
      "num_tokens": 807744676.0,
      "step": 25309
    },
    {
      "entropy": 0.4071207600645721,
      "epoch": 2.328646758788046,
      "grad_norm": 0.14500415325164795,
      "learning_rate": 2.238169718158678e-05,
      "loss": 0.3892,
      "mean_token_accuracy": 0.8767443001270294,
      "num_tokens": 807777304.0,
      "step": 25310
    },
    {
      "entropy": 0.4869092870503664,
      "epoch": 2.3287387651594873,
      "grad_norm": 0.15456517040729523,
      "learning_rate": 2.237863035544515e-05,
      "loss": 0.4759,
      "mean_token_accuracy": 0.848826453089714,
      "num_tokens": 807809651.0,
      "step": 25311
    },
    {
      "entropy": 0.5094773955643177,
      "epoch": 2.3288307715309284,
      "grad_norm": 0.16163499653339386,
      "learning_rate": 2.2375563529303525e-05,
      "loss": 0.4939,
      "mean_token_accuracy": 0.8399196974933147,
      "num_tokens": 807841247.0,
      "step": 25312
    },
    {
      "entropy": 0.575840862467885,
      "epoch": 2.3289227779023696,
      "grad_norm": 0.17067740857601166,
      "learning_rate": 2.23724967031619e-05,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8171841613948345,
      "num_tokens": 807872791.0,
      "step": 25313
    },
    {
      "entropy": 0.5593357558827847,
      "epoch": 2.329014784273811,
      "grad_norm": 0.1648123413324356,
      "learning_rate": 2.2369429877020274e-05,
      "loss": 0.5444,
      "mean_token_accuracy": 0.8280944675207138,
      "num_tokens": 807905035.0,
      "step": 25314
    },
    {
      "entropy": 0.4431611364707351,
      "epoch": 2.3291067906452523,
      "grad_norm": 0.15308372676372528,
      "learning_rate": 2.2366363050878645e-05,
      "loss": 0.4415,
      "mean_token_accuracy": 0.8569737486541271,
      "num_tokens": 807937155.0,
      "step": 25315
    },
    {
      "entropy": 0.5644648743327707,
      "epoch": 2.3291987970166934,
      "grad_norm": 0.16794084012508392,
      "learning_rate": 2.236329622473702e-05,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8266159072518349,
      "num_tokens": 807969498.0,
      "step": 25316
    },
    {
      "entropy": 0.5713472636416554,
      "epoch": 2.3292908033881345,
      "grad_norm": 0.1684800237417221,
      "learning_rate": 2.2360229398595395e-05,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8260368779301643,
      "num_tokens": 808000570.0,
      "step": 25317
    },
    {
      "entropy": 0.5063774100271985,
      "epoch": 2.3293828097595757,
      "grad_norm": 0.15399779379367828,
      "learning_rate": 2.235716257245377e-05,
      "loss": 0.491,
      "mean_token_accuracy": 0.8444665297865868,
      "num_tokens": 808032815.0,
      "step": 25318
    },
    {
      "entropy": 0.5249217413365841,
      "epoch": 2.3294748161310173,
      "grad_norm": 0.16622602939605713,
      "learning_rate": 2.2354095746312144e-05,
      "loss": 0.5075,
      "mean_token_accuracy": 0.839534804224968,
      "num_tokens": 808064367.0,
      "step": 25319
    },
    {
      "entropy": 0.416641958989203,
      "epoch": 2.3295668225024584,
      "grad_norm": 0.14809897541999817,
      "learning_rate": 2.2351028920170515e-05,
      "loss": 0.4077,
      "mean_token_accuracy": 0.8688831925392151,
      "num_tokens": 808096915.0,
      "step": 25320
    },
    {
      "entropy": 0.5357289109379053,
      "epoch": 2.3296588288738995,
      "grad_norm": 0.16331736743450165,
      "learning_rate": 2.234796209402889e-05,
      "loss": 0.526,
      "mean_token_accuracy": 0.8367569670081139,
      "num_tokens": 808128833.0,
      "step": 25321
    },
    {
      "entropy": 0.4891517765354365,
      "epoch": 2.3297508352453407,
      "grad_norm": 0.15299789607524872,
      "learning_rate": 2.2344895267887264e-05,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8517029881477356,
      "num_tokens": 808161076.0,
      "step": 25322
    },
    {
      "entropy": 0.5391707429662347,
      "epoch": 2.329842841616782,
      "grad_norm": 0.16708478331565857,
      "learning_rate": 2.234182844174564e-05,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8332529813051224,
      "num_tokens": 808192582.0,
      "step": 25323
    },
    {
      "entropy": 0.5286461990326643,
      "epoch": 2.3299348479882234,
      "grad_norm": 0.1602764129638672,
      "learning_rate": 2.233876161560401e-05,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8338894844055176,
      "num_tokens": 808224546.0,
      "step": 25324
    },
    {
      "entropy": 0.4832066148519516,
      "epoch": 2.3300268543596645,
      "grad_norm": 0.16078431904315948,
      "learning_rate": 2.2335694789462385e-05,
      "loss": 0.4615,
      "mean_token_accuracy": 0.8501732908189297,
      "num_tokens": 808256797.0,
      "step": 25325
    },
    {
      "entropy": 0.5751765272580087,
      "epoch": 2.3301188607311056,
      "grad_norm": 0.16568978130817413,
      "learning_rate": 2.233262796332076e-05,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8269211649894714,
      "num_tokens": 808288730.0,
      "step": 25326
    },
    {
      "entropy": 0.4904527524486184,
      "epoch": 2.3302108671025468,
      "grad_norm": 0.15781332552433014,
      "learning_rate": 2.2329561137179134e-05,
      "loss": 0.4753,
      "mean_token_accuracy": 0.8472790978848934,
      "num_tokens": 808320857.0,
      "step": 25327
    },
    {
      "entropy": 0.42806294839829206,
      "epoch": 2.330302873473988,
      "grad_norm": 0.14415407180786133,
      "learning_rate": 2.232649431103751e-05,
      "loss": 0.4163,
      "mean_token_accuracy": 0.8684568926692009,
      "num_tokens": 808352604.0,
      "step": 25328
    },
    {
      "entropy": 0.5315435491502285,
      "epoch": 2.3303948798454295,
      "grad_norm": 0.16653040051460266,
      "learning_rate": 2.232342748489588e-05,
      "loss": 0.522,
      "mean_token_accuracy": 0.8374136686325073,
      "num_tokens": 808383966.0,
      "step": 25329
    },
    {
      "entropy": 0.5064332010224462,
      "epoch": 2.3304868862168706,
      "grad_norm": 0.16007103025913239,
      "learning_rate": 2.2320360658754255e-05,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8436578214168549,
      "num_tokens": 808415095.0,
      "step": 25330
    },
    {
      "entropy": 0.4797869324684143,
      "epoch": 2.3305788925883117,
      "grad_norm": 0.15531232953071594,
      "learning_rate": 2.231729383261263e-05,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8483793921768665,
      "num_tokens": 808446886.0,
      "step": 25331
    },
    {
      "entropy": 0.5590710081160069,
      "epoch": 2.330670898959753,
      "grad_norm": 0.16561336815357208,
      "learning_rate": 2.2314227006471004e-05,
      "loss": 0.5493,
      "mean_token_accuracy": 0.8264541961252689,
      "num_tokens": 808479486.0,
      "step": 25332
    },
    {
      "entropy": 0.5168212484568357,
      "epoch": 2.330762905331194,
      "grad_norm": 0.16343890130519867,
      "learning_rate": 2.2311160180329375e-05,
      "loss": 0.5063,
      "mean_token_accuracy": 0.8373972326517105,
      "num_tokens": 808510748.0,
      "step": 25333
    },
    {
      "entropy": 0.423238436691463,
      "epoch": 2.3308549117026356,
      "grad_norm": 0.15363101661205292,
      "learning_rate": 2.2308093354187753e-05,
      "loss": 0.4098,
      "mean_token_accuracy": 0.8679915741086006,
      "num_tokens": 808542447.0,
      "step": 25334
    },
    {
      "entropy": 0.5928523931652308,
      "epoch": 2.3309469180740767,
      "grad_norm": 0.17440451681613922,
      "learning_rate": 2.2305026528046128e-05,
      "loss": 0.5816,
      "mean_token_accuracy": 0.8182483315467834,
      "num_tokens": 808573959.0,
      "step": 25335
    },
    {
      "entropy": 0.5032478733919561,
      "epoch": 2.331038924445518,
      "grad_norm": 0.15593670308589935,
      "learning_rate": 2.2301959701904503e-05,
      "loss": 0.4896,
      "mean_token_accuracy": 0.8469676226377487,
      "num_tokens": 808606494.0,
      "step": 25336
    },
    {
      "entropy": 0.5383511018007994,
      "epoch": 2.331130930816959,
      "grad_norm": 0.16243354976177216,
      "learning_rate": 2.2298892875762874e-05,
      "loss": 0.5218,
      "mean_token_accuracy": 0.8356500826776028,
      "num_tokens": 808638252.0,
      "step": 25337
    },
    {
      "entropy": 0.5903336480259895,
      "epoch": 2.3312229371884,
      "grad_norm": 0.16714531183242798,
      "learning_rate": 2.229582604962125e-05,
      "loss": 0.58,
      "mean_token_accuracy": 0.816331047564745,
      "num_tokens": 808670261.0,
      "step": 25338
    },
    {
      "entropy": 0.5442987084388733,
      "epoch": 2.3313149435598417,
      "grad_norm": 0.1654207557439804,
      "learning_rate": 2.2292759223479623e-05,
      "loss": 0.529,
      "mean_token_accuracy": 0.8294319473206997,
      "num_tokens": 808701414.0,
      "step": 25339
    },
    {
      "entropy": 0.45419708639383316,
      "epoch": 2.331406949931283,
      "grad_norm": 0.15650704503059387,
      "learning_rate": 2.2289692397337998e-05,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8606525361537933,
      "num_tokens": 808733469.0,
      "step": 25340
    },
    {
      "entropy": 0.5726649798452854,
      "epoch": 2.331498956302724,
      "grad_norm": 0.16959552466869354,
      "learning_rate": 2.228662557119637e-05,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8255394995212555,
      "num_tokens": 808765394.0,
      "step": 25341
    },
    {
      "entropy": 0.6474495809525251,
      "epoch": 2.331590962674165,
      "grad_norm": 0.1769135296344757,
      "learning_rate": 2.2283558745054744e-05,
      "loss": 0.6354,
      "mean_token_accuracy": 0.8012820519506931,
      "num_tokens": 808796782.0,
      "step": 25342
    },
    {
      "entropy": 0.5383284753188491,
      "epoch": 2.3316829690456062,
      "grad_norm": 0.1604040414094925,
      "learning_rate": 2.228049191891312e-05,
      "loss": 0.5293,
      "mean_token_accuracy": 0.835674699395895,
      "num_tokens": 808829076.0,
      "step": 25343
    },
    {
      "entropy": 0.5642052702605724,
      "epoch": 2.331774975417048,
      "grad_norm": 0.1644805669784546,
      "learning_rate": 2.2277425092771493e-05,
      "loss": 0.545,
      "mean_token_accuracy": 0.8250079825520515,
      "num_tokens": 808861455.0,
      "step": 25344
    },
    {
      "entropy": 0.5498912213370204,
      "epoch": 2.331866981788489,
      "grad_norm": 0.1664866954088211,
      "learning_rate": 2.2274358266629864e-05,
      "loss": 0.5394,
      "mean_token_accuracy": 0.829914141446352,
      "num_tokens": 808893518.0,
      "step": 25345
    },
    {
      "entropy": 0.514979163184762,
      "epoch": 2.33195898815993,
      "grad_norm": 0.16262291371822357,
      "learning_rate": 2.227129144048824e-05,
      "loss": 0.5041,
      "mean_token_accuracy": 0.8392904661595821,
      "num_tokens": 808925504.0,
      "step": 25346
    },
    {
      "entropy": 0.44679208798334,
      "epoch": 2.332050994531371,
      "grad_norm": 0.15193161368370056,
      "learning_rate": 2.2268224614346614e-05,
      "loss": 0.4291,
      "mean_token_accuracy": 0.8596961908042431,
      "num_tokens": 808956983.0,
      "step": 25347
    },
    {
      "entropy": 0.5265334276482463,
      "epoch": 2.3321430009028123,
      "grad_norm": 0.1626061499118805,
      "learning_rate": 2.226515778820499e-05,
      "loss": 0.5206,
      "mean_token_accuracy": 0.8345176242291927,
      "num_tokens": 808988507.0,
      "step": 25348
    },
    {
      "entropy": 0.6713176537305117,
      "epoch": 2.332235007274254,
      "grad_norm": 0.17828331887722015,
      "learning_rate": 2.2262090962063363e-05,
      "loss": 0.6505,
      "mean_token_accuracy": 0.7966595105826855,
      "num_tokens": 809020676.0,
      "step": 25349
    },
    {
      "entropy": 0.4456854937598109,
      "epoch": 2.332327013645695,
      "grad_norm": 0.15659014880657196,
      "learning_rate": 2.2259024135921734e-05,
      "loss": 0.4374,
      "mean_token_accuracy": 0.8563956245779991,
      "num_tokens": 809053166.0,
      "step": 25350
    },
    {
      "entropy": 0.5930092064663768,
      "epoch": 2.332419020017136,
      "grad_norm": 0.1771264374256134,
      "learning_rate": 2.225595730978011e-05,
      "loss": 0.5741,
      "mean_token_accuracy": 0.8199868462979794,
      "num_tokens": 809084763.0,
      "step": 25351
    },
    {
      "entropy": 0.517909569432959,
      "epoch": 2.3325110263885773,
      "grad_norm": 0.15808482468128204,
      "learning_rate": 2.2252890483638484e-05,
      "loss": 0.5021,
      "mean_token_accuracy": 0.8414645977318287,
      "num_tokens": 809116452.0,
      "step": 25352
    },
    {
      "entropy": 0.5196388768963516,
      "epoch": 2.3326030327600185,
      "grad_norm": 0.16216881573200226,
      "learning_rate": 2.2249823657496858e-05,
      "loss": 0.5006,
      "mean_token_accuracy": 0.8424673825502396,
      "num_tokens": 809147859.0,
      "step": 25353
    },
    {
      "entropy": 0.5933863110840321,
      "epoch": 2.33269503913146,
      "grad_norm": 0.16252820193767548,
      "learning_rate": 2.224675683135523e-05,
      "loss": 0.5791,
      "mean_token_accuracy": 0.8159246481955051,
      "num_tokens": 809179983.0,
      "step": 25354
    },
    {
      "entropy": 0.5924408752471209,
      "epoch": 2.332787045502901,
      "grad_norm": 0.17183955013751984,
      "learning_rate": 2.2243690005213604e-05,
      "loss": 0.5885,
      "mean_token_accuracy": 0.8159428536891937,
      "num_tokens": 809211887.0,
      "step": 25355
    },
    {
      "entropy": 0.4902804708108306,
      "epoch": 2.3328790518743423,
      "grad_norm": 0.1584063172340393,
      "learning_rate": 2.224062317907198e-05,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8458381555974483,
      "num_tokens": 809243399.0,
      "step": 25356
    },
    {
      "entropy": 0.5489346319809556,
      "epoch": 2.3329710582457834,
      "grad_norm": 0.16161008179187775,
      "learning_rate": 2.2237556352930353e-05,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8309973515570164,
      "num_tokens": 809275028.0,
      "step": 25357
    },
    {
      "entropy": 0.4841193223837763,
      "epoch": 2.3330630646172246,
      "grad_norm": 0.15996302664279938,
      "learning_rate": 2.2234489526788728e-05,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8467593230307102,
      "num_tokens": 809306659.0,
      "step": 25358
    },
    {
      "entropy": 0.509845266584307,
      "epoch": 2.333155070988666,
      "grad_norm": 0.15811485052108765,
      "learning_rate": 2.22314227006471e-05,
      "loss": 0.5047,
      "mean_token_accuracy": 0.8421486653387547,
      "num_tokens": 809338421.0,
      "step": 25359
    },
    {
      "entropy": 0.5511269182898104,
      "epoch": 2.3332470773601073,
      "grad_norm": 0.16930364072322845,
      "learning_rate": 2.2228355874505474e-05,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8281344510614872,
      "num_tokens": 809370595.0,
      "step": 25360
    },
    {
      "entropy": 0.5358506661141291,
      "epoch": 2.3333390837315484,
      "grad_norm": 0.16399674117565155,
      "learning_rate": 2.222528904836385e-05,
      "loss": 0.519,
      "mean_token_accuracy": 0.8380574956536293,
      "num_tokens": 809402535.0,
      "step": 25361
    },
    {
      "entropy": 0.5191839272156358,
      "epoch": 2.3334310901029895,
      "grad_norm": 0.16718804836273193,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.5007,
      "mean_token_accuracy": 0.838709183037281,
      "num_tokens": 809434340.0,
      "step": 25362
    },
    {
      "entropy": 0.501044737175107,
      "epoch": 2.3335230964744307,
      "grad_norm": 0.15425239503383636,
      "learning_rate": 2.2219155396080595e-05,
      "loss": 0.485,
      "mean_token_accuracy": 0.8459467440843582,
      "num_tokens": 809466781.0,
      "step": 25363
    },
    {
      "entropy": 0.5082686692476273,
      "epoch": 2.3336151028458723,
      "grad_norm": 0.15831731259822845,
      "learning_rate": 2.2216088569938973e-05,
      "loss": 0.4978,
      "mean_token_accuracy": 0.8415742330253124,
      "num_tokens": 809499090.0,
      "step": 25364
    },
    {
      "entropy": 0.5153149180114269,
      "epoch": 2.3337071092173134,
      "grad_norm": 0.16209851205348969,
      "learning_rate": 2.2213021743797347e-05,
      "loss": 0.5116,
      "mean_token_accuracy": 0.8381311967968941,
      "num_tokens": 809530582.0,
      "step": 25365
    },
    {
      "entropy": 0.43162071448750794,
      "epoch": 2.3337991155887545,
      "grad_norm": 0.14309921860694885,
      "learning_rate": 2.2209954917655722e-05,
      "loss": 0.4186,
      "mean_token_accuracy": 0.8683476746082306,
      "num_tokens": 809563277.0,
      "step": 25366
    },
    {
      "entropy": 0.5854429062455893,
      "epoch": 2.3338911219601957,
      "grad_norm": 0.16789411008358002,
      "learning_rate": 2.2206888091514093e-05,
      "loss": 0.574,
      "mean_token_accuracy": 0.8204487711191177,
      "num_tokens": 809594721.0,
      "step": 25367
    },
    {
      "entropy": 0.5583984591066837,
      "epoch": 2.333983128331637,
      "grad_norm": 0.17682942748069763,
      "learning_rate": 2.2203821265372468e-05,
      "loss": 0.5509,
      "mean_token_accuracy": 0.827939223498106,
      "num_tokens": 809625191.0,
      "step": 25368
    },
    {
      "entropy": 0.4724930548109114,
      "epoch": 2.3340751347030784,
      "grad_norm": 0.1547667235136032,
      "learning_rate": 2.2200754439230842e-05,
      "loss": 0.4613,
      "mean_token_accuracy": 0.85739840939641,
      "num_tokens": 809656911.0,
      "step": 25369
    },
    {
      "entropy": 0.49017384089529514,
      "epoch": 2.3341671410745195,
      "grad_norm": 0.15827412903308868,
      "learning_rate": 2.2197687613089217e-05,
      "loss": 0.4735,
      "mean_token_accuracy": 0.8467213846743107,
      "num_tokens": 809688767.0,
      "step": 25370
    },
    {
      "entropy": 0.5495691411197186,
      "epoch": 2.3342591474459606,
      "grad_norm": 0.1615716964006424,
      "learning_rate": 2.219462078694759e-05,
      "loss": 0.5293,
      "mean_token_accuracy": 0.8338403254747391,
      "num_tokens": 809720731.0,
      "step": 25371
    },
    {
      "entropy": 0.47419248055666685,
      "epoch": 2.3343511538174018,
      "grad_norm": 0.15711133182048798,
      "learning_rate": 2.2191553960805963e-05,
      "loss": 0.4609,
      "mean_token_accuracy": 0.8539056032896042,
      "num_tokens": 809752150.0,
      "step": 25372
    },
    {
      "entropy": 0.5296588987112045,
      "epoch": 2.334443160188843,
      "grad_norm": 0.1701837033033371,
      "learning_rate": 2.2188487134664338e-05,
      "loss": 0.5144,
      "mean_token_accuracy": 0.839866403490305,
      "num_tokens": 809783345.0,
      "step": 25373
    },
    {
      "entropy": 0.5614342270419002,
      "epoch": 2.3345351665602845,
      "grad_norm": 0.16531306505203247,
      "learning_rate": 2.2185420308522712e-05,
      "loss": 0.5505,
      "mean_token_accuracy": 0.8242644406855106,
      "num_tokens": 809815202.0,
      "step": 25374
    },
    {
      "entropy": 0.48042873200029135,
      "epoch": 2.3346271729317256,
      "grad_norm": 0.1559966802597046,
      "learning_rate": 2.2182353482381087e-05,
      "loss": 0.4681,
      "mean_token_accuracy": 0.8495835177600384,
      "num_tokens": 809846966.0,
      "step": 25375
    },
    {
      "entropy": 0.5087133548222482,
      "epoch": 2.3347191793031667,
      "grad_norm": 0.15831904113292694,
      "learning_rate": 2.2179286656239458e-05,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8448728993535042,
      "num_tokens": 809879123.0,
      "step": 25376
    },
    {
      "entropy": 0.4935970762744546,
      "epoch": 2.334811185674608,
      "grad_norm": 0.15882745385169983,
      "learning_rate": 2.2176219830097833e-05,
      "loss": 0.486,
      "mean_token_accuracy": 0.8468896709382534,
      "num_tokens": 809910907.0,
      "step": 25377
    },
    {
      "entropy": 0.5879342462867498,
      "epoch": 2.334903192046049,
      "grad_norm": 0.1708189696073532,
      "learning_rate": 2.2173153003956207e-05,
      "loss": 0.5734,
      "mean_token_accuracy": 0.8189339265227318,
      "num_tokens": 809942580.0,
      "step": 25378
    },
    {
      "entropy": 0.661884868517518,
      "epoch": 2.3349951984174906,
      "grad_norm": 0.17488619685173035,
      "learning_rate": 2.2170086177814582e-05,
      "loss": 0.6476,
      "mean_token_accuracy": 0.8001497387886047,
      "num_tokens": 809974892.0,
      "step": 25379
    },
    {
      "entropy": 0.5646666269749403,
      "epoch": 2.3350872047889317,
      "grad_norm": 0.16706274449825287,
      "learning_rate": 2.2167019351672953e-05,
      "loss": 0.5471,
      "mean_token_accuracy": 0.8231784962117672,
      "num_tokens": 810006408.0,
      "step": 25380
    },
    {
      "entropy": 0.42927661957219243,
      "epoch": 2.335179211160373,
      "grad_norm": 0.15657413005828857,
      "learning_rate": 2.2163952525531328e-05,
      "loss": 0.4141,
      "mean_token_accuracy": 0.8687731809914112,
      "num_tokens": 810038748.0,
      "step": 25381
    },
    {
      "entropy": 0.3863818140234798,
      "epoch": 2.335271217531814,
      "grad_norm": 0.146961510181427,
      "learning_rate": 2.2160885699389703e-05,
      "loss": 0.3837,
      "mean_token_accuracy": 0.8756027743220329,
      "num_tokens": 810070160.0,
      "step": 25382
    },
    {
      "entropy": 0.5541303360369056,
      "epoch": 2.335363223903255,
      "grad_norm": 0.16087929904460907,
      "learning_rate": 2.2157818873248077e-05,
      "loss": 0.5409,
      "mean_token_accuracy": 0.831233125180006,
      "num_tokens": 810102467.0,
      "step": 25383
    },
    {
      "entropy": 0.5994098596274853,
      "epoch": 2.3354552302746967,
      "grad_norm": 0.17329667508602142,
      "learning_rate": 2.215475204710645e-05,
      "loss": 0.5964,
      "mean_token_accuracy": 0.8143153078854084,
      "num_tokens": 810134130.0,
      "step": 25384
    },
    {
      "entropy": 0.5080065354704857,
      "epoch": 2.335547236646138,
      "grad_norm": 0.16281749308109283,
      "learning_rate": 2.2151685220964823e-05,
      "loss": 0.4859,
      "mean_token_accuracy": 0.8428653217852116,
      "num_tokens": 810165292.0,
      "step": 25385
    },
    {
      "entropy": 0.5478651546873152,
      "epoch": 2.335639243017579,
      "grad_norm": 0.1620418280363083,
      "learning_rate": 2.2148618394823198e-05,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8291626311838627,
      "num_tokens": 810197670.0,
      "step": 25386
    },
    {
      "entropy": 0.5432215761393309,
      "epoch": 2.33573124938902,
      "grad_norm": 0.16133522987365723,
      "learning_rate": 2.2145551568681573e-05,
      "loss": 0.5418,
      "mean_token_accuracy": 0.8310026824474335,
      "num_tokens": 810230438.0,
      "step": 25387
    },
    {
      "entropy": 0.5593562228605151,
      "epoch": 2.3358232557604612,
      "grad_norm": 0.16469722986221313,
      "learning_rate": 2.2142484742539947e-05,
      "loss": 0.5506,
      "mean_token_accuracy": 0.8262763693928719,
      "num_tokens": 810261890.0,
      "step": 25388
    },
    {
      "entropy": 0.5262035776395351,
      "epoch": 2.335915262131903,
      "grad_norm": 0.162392720580101,
      "learning_rate": 2.213941791639832e-05,
      "loss": 0.5113,
      "mean_token_accuracy": 0.8388334214687347,
      "num_tokens": 810293768.0,
      "step": 25389
    },
    {
      "entropy": 0.5778560787439346,
      "epoch": 2.336007268503344,
      "grad_norm": 0.164114847779274,
      "learning_rate": 2.2136351090256693e-05,
      "loss": 0.5609,
      "mean_token_accuracy": 0.8216534145176411,
      "num_tokens": 810325200.0,
      "step": 25390
    },
    {
      "entropy": 0.6122992150485516,
      "epoch": 2.336099274874785,
      "grad_norm": 0.1751924455165863,
      "learning_rate": 2.2133284264115068e-05,
      "loss": 0.6049,
      "mean_token_accuracy": 0.807255070656538,
      "num_tokens": 810356275.0,
      "step": 25391
    },
    {
      "entropy": 0.559012109413743,
      "epoch": 2.336191281246226,
      "grad_norm": 0.16548509895801544,
      "learning_rate": 2.2130217437973442e-05,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8277681022882462,
      "num_tokens": 810386808.0,
      "step": 25392
    },
    {
      "entropy": 0.5149222216568887,
      "epoch": 2.3362832876176673,
      "grad_norm": 0.1554878056049347,
      "learning_rate": 2.2127150611831814e-05,
      "loss": 0.5029,
      "mean_token_accuracy": 0.8422015905380249,
      "num_tokens": 810419165.0,
      "step": 25393
    },
    {
      "entropy": 0.530878460034728,
      "epoch": 2.336375293989109,
      "grad_norm": 0.1633700132369995,
      "learning_rate": 2.2124083785690188e-05,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8351241610944271,
      "num_tokens": 810451585.0,
      "step": 25394
    },
    {
      "entropy": 0.661050459370017,
      "epoch": 2.33646730036055,
      "grad_norm": 0.17766764760017395,
      "learning_rate": 2.2121016959548566e-05,
      "loss": 0.6541,
      "mean_token_accuracy": 0.7957199364900589,
      "num_tokens": 810483480.0,
      "step": 25395
    },
    {
      "entropy": 0.49432891327887774,
      "epoch": 2.336559306731991,
      "grad_norm": 0.16211602091789246,
      "learning_rate": 2.211795013340694e-05,
      "loss": 0.4802,
      "mean_token_accuracy": 0.8497466035187244,
      "num_tokens": 810515221.0,
      "step": 25396
    },
    {
      "entropy": 0.6577170696109533,
      "epoch": 2.3366513131034323,
      "grad_norm": 0.17422159016132355,
      "learning_rate": 2.2114883307265312e-05,
      "loss": 0.6354,
      "mean_token_accuracy": 0.80268394947052,
      "num_tokens": 810546936.0,
      "step": 25397
    },
    {
      "entropy": 0.47450566850602627,
      "epoch": 2.3367433194748735,
      "grad_norm": 0.15509864687919617,
      "learning_rate": 2.2111816481123687e-05,
      "loss": 0.4628,
      "mean_token_accuracy": 0.8525736704468727,
      "num_tokens": 810579039.0,
      "step": 25398
    },
    {
      "entropy": 0.5263293832540512,
      "epoch": 2.336835325846315,
      "grad_norm": 0.1551290899515152,
      "learning_rate": 2.210874965498206e-05,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8410601131618023,
      "num_tokens": 810610896.0,
      "step": 25399
    },
    {
      "entropy": 0.5451594609767199,
      "epoch": 2.336927332217756,
      "grad_norm": 0.16618581116199493,
      "learning_rate": 2.2105682828840436e-05,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8338431306183338,
      "num_tokens": 810643133.0,
      "step": 25400
    },
    {
      "entropy": 0.6198538462631404,
      "epoch": 2.3370193385891973,
      "grad_norm": 0.1716468185186386,
      "learning_rate": 2.2102616002698807e-05,
      "loss": 0.6001,
      "mean_token_accuracy": 0.8098901100456715,
      "num_tokens": 810674454.0,
      "step": 25401
    },
    {
      "entropy": 0.4519251622259617,
      "epoch": 2.3371113449606384,
      "grad_norm": 0.15167509019374847,
      "learning_rate": 2.2099549176557182e-05,
      "loss": 0.4412,
      "mean_token_accuracy": 0.8592911139130592,
      "num_tokens": 810706545.0,
      "step": 25402
    },
    {
      "entropy": 0.5339863658882678,
      "epoch": 2.3372033513320796,
      "grad_norm": 0.1649412363767624,
      "learning_rate": 2.2096482350415557e-05,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8351983316242695,
      "num_tokens": 810738677.0,
      "step": 25403
    },
    {
      "entropy": 0.4812779682688415,
      "epoch": 2.337295357703521,
      "grad_norm": 0.15745431184768677,
      "learning_rate": 2.209341552427393e-05,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8551194258034229,
      "num_tokens": 810770520.0,
      "step": 25404
    },
    {
      "entropy": 0.5391188897192478,
      "epoch": 2.3373873640749623,
      "grad_norm": 0.16303743422031403,
      "learning_rate": 2.2090348698132306e-05,
      "loss": 0.53,
      "mean_token_accuracy": 0.8308138214051723,
      "num_tokens": 810802343.0,
      "step": 25405
    },
    {
      "entropy": 0.5197692699730396,
      "epoch": 2.3374793704464034,
      "grad_norm": 0.15881691873073578,
      "learning_rate": 2.2087281871990677e-05,
      "loss": 0.5093,
      "mean_token_accuracy": 0.8400153405964375,
      "num_tokens": 810834668.0,
      "step": 25406
    },
    {
      "entropy": 0.42456015234347433,
      "epoch": 2.3375713768178445,
      "grad_norm": 0.15566720068454742,
      "learning_rate": 2.2084215045849052e-05,
      "loss": 0.4122,
      "mean_token_accuracy": 0.868404284119606,
      "num_tokens": 810866906.0,
      "step": 25407
    },
    {
      "entropy": 0.5600406974554062,
      "epoch": 2.3376633831892857,
      "grad_norm": 0.16733990609645844,
      "learning_rate": 2.2081148219707427e-05,
      "loss": 0.5514,
      "mean_token_accuracy": 0.8253092244267464,
      "num_tokens": 810898928.0,
      "step": 25408
    },
    {
      "entropy": 0.5697155520319939,
      "epoch": 2.3377553895607273,
      "grad_norm": 0.16769443452358246,
      "learning_rate": 2.20780813935658e-05,
      "loss": 0.56,
      "mean_token_accuracy": 0.8209701664745808,
      "num_tokens": 810930936.0,
      "step": 25409
    },
    {
      "entropy": 0.5508955083787441,
      "epoch": 2.3378473959321684,
      "grad_norm": 0.1737731546163559,
      "learning_rate": 2.2075014567424172e-05,
      "loss": 0.5336,
      "mean_token_accuracy": 0.828195258975029,
      "num_tokens": 810961992.0,
      "step": 25410
    },
    {
      "entropy": 0.3904349207878113,
      "epoch": 2.3379394023036095,
      "grad_norm": 0.14454500377178192,
      "learning_rate": 2.2071947741282547e-05,
      "loss": 0.3828,
      "mean_token_accuracy": 0.8799998946487904,
      "num_tokens": 810994372.0,
      "step": 25411
    },
    {
      "entropy": 0.49359708873089403,
      "epoch": 2.3380314086750507,
      "grad_norm": 0.15281426906585693,
      "learning_rate": 2.2068880915140922e-05,
      "loss": 0.4729,
      "mean_token_accuracy": 0.8473955243825912,
      "num_tokens": 811026534.0,
      "step": 25412
    },
    {
      "entropy": 0.5524979662150145,
      "epoch": 2.338123415046492,
      "grad_norm": 0.16228929162025452,
      "learning_rate": 2.2065814088999296e-05,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8275744095444679,
      "num_tokens": 811058779.0,
      "step": 25413
    },
    {
      "entropy": 0.44330946961417794,
      "epoch": 2.3382154214179334,
      "grad_norm": 0.15368399024009705,
      "learning_rate": 2.2062747262857668e-05,
      "loss": 0.4269,
      "mean_token_accuracy": 0.8616804853081703,
      "num_tokens": 811090210.0,
      "step": 25414
    },
    {
      "entropy": 0.5080028939992189,
      "epoch": 2.3383074277893745,
      "grad_norm": 0.1572154015302658,
      "learning_rate": 2.2059680436716042e-05,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8395472019910812,
      "num_tokens": 811122156.0,
      "step": 25415
    },
    {
      "entropy": 0.541153128258884,
      "epoch": 2.3383994341608156,
      "grad_norm": 0.1660388708114624,
      "learning_rate": 2.2056613610574417e-05,
      "loss": 0.5317,
      "mean_token_accuracy": 0.8290988095104694,
      "num_tokens": 811153268.0,
      "step": 25416
    },
    {
      "entropy": 0.5438232850283384,
      "epoch": 2.3384914405322568,
      "grad_norm": 0.1609288454055786,
      "learning_rate": 2.205354678443279e-05,
      "loss": 0.5313,
      "mean_token_accuracy": 0.831424206495285,
      "num_tokens": 811185342.0,
      "step": 25417
    },
    {
      "entropy": 0.4951083578635007,
      "epoch": 2.338583446903698,
      "grad_norm": 0.1562614142894745,
      "learning_rate": 2.2050479958291166e-05,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8472444824874401,
      "num_tokens": 811216640.0,
      "step": 25418
    },
    {
      "entropy": 0.540538027882576,
      "epoch": 2.3386754532751395,
      "grad_norm": 0.16533933579921722,
      "learning_rate": 2.2047413132149538e-05,
      "loss": 0.5241,
      "mean_token_accuracy": 0.831847008317709,
      "num_tokens": 811248230.0,
      "step": 25419
    },
    {
      "entropy": 0.46537014364730567,
      "epoch": 2.3387674596465806,
      "grad_norm": 0.15307505428791046,
      "learning_rate": 2.2044346306007912e-05,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8551991395652294,
      "num_tokens": 811280278.0,
      "step": 25420
    },
    {
      "entropy": 0.5274637173861265,
      "epoch": 2.3388594660180217,
      "grad_norm": 0.16397203505039215,
      "learning_rate": 2.2041279479866287e-05,
      "loss": 0.5142,
      "mean_token_accuracy": 0.8382059410214424,
      "num_tokens": 811312293.0,
      "step": 25421
    },
    {
      "entropy": 0.43726425245404243,
      "epoch": 2.338951472389463,
      "grad_norm": 0.14984266459941864,
      "learning_rate": 2.203821265372466e-05,
      "loss": 0.4267,
      "mean_token_accuracy": 0.8659623898565769,
      "num_tokens": 811345061.0,
      "step": 25422
    },
    {
      "entropy": 0.5168485045433044,
      "epoch": 2.339043478760904,
      "grad_norm": 0.1667502373456955,
      "learning_rate": 2.2035145827583033e-05,
      "loss": 0.5055,
      "mean_token_accuracy": 0.8407119363546371,
      "num_tokens": 811375793.0,
      "step": 25423
    },
    {
      "entropy": 0.5930346995592117,
      "epoch": 2.3391354851323456,
      "grad_norm": 0.1653212606906891,
      "learning_rate": 2.2032079001441407e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8171431235969067,
      "num_tokens": 811407945.0,
      "step": 25424
    },
    {
      "entropy": 0.49350348860025406,
      "epoch": 2.3392274915037867,
      "grad_norm": 0.1610638052225113,
      "learning_rate": 2.2029012175299785e-05,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8530236631631851,
      "num_tokens": 811439992.0,
      "step": 25425
    },
    {
      "entropy": 0.5119087938219309,
      "epoch": 2.339319497875228,
      "grad_norm": 0.15948383510112762,
      "learning_rate": 2.202594534915816e-05,
      "loss": 0.509,
      "mean_token_accuracy": 0.8398580774664879,
      "num_tokens": 811471891.0,
      "step": 25426
    },
    {
      "entropy": 0.499005775898695,
      "epoch": 2.339411504246669,
      "grad_norm": 0.15573816001415253,
      "learning_rate": 2.202287852301653e-05,
      "loss": 0.4968,
      "mean_token_accuracy": 0.8448676876723766,
      "num_tokens": 811503862.0,
      "step": 25427
    },
    {
      "entropy": 0.49761854857206345,
      "epoch": 2.33950351061811,
      "grad_norm": 0.1672133058309555,
      "learning_rate": 2.2019811696874906e-05,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8427280634641647,
      "num_tokens": 811536251.0,
      "step": 25428
    },
    {
      "entropy": 0.5498074200004339,
      "epoch": 2.3395955169895517,
      "grad_norm": 0.1738906353712082,
      "learning_rate": 2.201674487073328e-05,
      "loss": 0.5487,
      "mean_token_accuracy": 0.8275480717420578,
      "num_tokens": 811568956.0,
      "step": 25429
    },
    {
      "entropy": 0.41655429173260927,
      "epoch": 2.339687523360993,
      "grad_norm": 0.14953108131885529,
      "learning_rate": 2.2013678044591655e-05,
      "loss": 0.4036,
      "mean_token_accuracy": 0.8714198470115662,
      "num_tokens": 811600875.0,
      "step": 25430
    },
    {
      "entropy": 0.5338136777281761,
      "epoch": 2.339779529732434,
      "grad_norm": 0.16223539412021637,
      "learning_rate": 2.2010611218450027e-05,
      "loss": 0.5142,
      "mean_token_accuracy": 0.8340826891362667,
      "num_tokens": 811632623.0,
      "step": 25431
    },
    {
      "entropy": 0.5747783575206995,
      "epoch": 2.339871536103875,
      "grad_norm": 0.17155613005161285,
      "learning_rate": 2.20075443923084e-05,
      "loss": 0.5653,
      "mean_token_accuracy": 0.8242207430303097,
      "num_tokens": 811663956.0,
      "step": 25432
    },
    {
      "entropy": 0.5507668852806091,
      "epoch": 2.3399635424753162,
      "grad_norm": 0.16591843962669373,
      "learning_rate": 2.2004477566166776e-05,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8347701393067837,
      "num_tokens": 811695292.0,
      "step": 25433
    },
    {
      "entropy": 0.4765307195484638,
      "epoch": 2.340055548846758,
      "grad_norm": 0.15455569326877594,
      "learning_rate": 2.200141074002515e-05,
      "loss": 0.4556,
      "mean_token_accuracy": 0.8539902605116367,
      "num_tokens": 811727082.0,
      "step": 25434
    },
    {
      "entropy": 0.6213620640337467,
      "epoch": 2.340147555218199,
      "grad_norm": 0.17301611602306366,
      "learning_rate": 2.1998343913883525e-05,
      "loss": 0.5974,
      "mean_token_accuracy": 0.8117490820586681,
      "num_tokens": 811758286.0,
      "step": 25435
    },
    {
      "entropy": 0.5135159976780415,
      "epoch": 2.34023956158964,
      "grad_norm": 0.16138634085655212,
      "learning_rate": 2.1995277087741896e-05,
      "loss": 0.5039,
      "mean_token_accuracy": 0.8432753644883633,
      "num_tokens": 811790530.0,
      "step": 25436
    },
    {
      "entropy": 0.4588173530064523,
      "epoch": 2.340331567961081,
      "grad_norm": 0.15356647968292236,
      "learning_rate": 2.199221026160027e-05,
      "loss": 0.4493,
      "mean_token_accuracy": 0.858758557587862,
      "num_tokens": 811822867.0,
      "step": 25437
    },
    {
      "entropy": 0.535624336451292,
      "epoch": 2.3404235743325223,
      "grad_norm": 0.15989632904529572,
      "learning_rate": 2.1989143435458646e-05,
      "loss": 0.5284,
      "mean_token_accuracy": 0.8327978514134884,
      "num_tokens": 811854894.0,
      "step": 25438
    },
    {
      "entropy": 0.5066377860493958,
      "epoch": 2.340515580703964,
      "grad_norm": 0.15985655784606934,
      "learning_rate": 2.198607660931702e-05,
      "loss": 0.4905,
      "mean_token_accuracy": 0.8415816687047482,
      "num_tokens": 811886987.0,
      "step": 25439
    },
    {
      "entropy": 0.5810728194192052,
      "epoch": 2.340607587075405,
      "grad_norm": 0.17275995016098022,
      "learning_rate": 2.198300978317539e-05,
      "loss": 0.5656,
      "mean_token_accuracy": 0.823616623878479,
      "num_tokens": 811919460.0,
      "step": 25440
    },
    {
      "entropy": 0.5020916284993291,
      "epoch": 2.340699593446846,
      "grad_norm": 0.1567457914352417,
      "learning_rate": 2.1979942957033766e-05,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8424880169332027,
      "num_tokens": 811951909.0,
      "step": 25441
    },
    {
      "entropy": 0.5182893760502338,
      "epoch": 2.3407915998182873,
      "grad_norm": 0.16413837671279907,
      "learning_rate": 2.197687613089214e-05,
      "loss": 0.5072,
      "mean_token_accuracy": 0.8377948887646198,
      "num_tokens": 811983519.0,
      "step": 25442
    },
    {
      "entropy": 0.5976263815537095,
      "epoch": 2.3408836061897285,
      "grad_norm": 0.17083729803562164,
      "learning_rate": 2.1973809304750516e-05,
      "loss": 0.5924,
      "mean_token_accuracy": 0.8131902106106281,
      "num_tokens": 812015837.0,
      "step": 25443
    },
    {
      "entropy": 0.5735164266079664,
      "epoch": 2.34097561256117,
      "grad_norm": 0.16816578805446625,
      "learning_rate": 2.1970742478608887e-05,
      "loss": 0.5715,
      "mean_token_accuracy": 0.8206540308892727,
      "num_tokens": 812047850.0,
      "step": 25444
    },
    {
      "entropy": 0.5177031904459,
      "epoch": 2.341067618932611,
      "grad_norm": 0.15826773643493652,
      "learning_rate": 2.196767565246726e-05,
      "loss": 0.5039,
      "mean_token_accuracy": 0.8379562273621559,
      "num_tokens": 812080185.0,
      "step": 25445
    },
    {
      "entropy": 0.6199878486804664,
      "epoch": 2.3411596253040523,
      "grad_norm": 0.1714019626379013,
      "learning_rate": 2.1964608826325636e-05,
      "loss": 0.6113,
      "mean_token_accuracy": 0.8069978058338165,
      "num_tokens": 812112131.0,
      "step": 25446
    },
    {
      "entropy": 0.5224768058396876,
      "epoch": 2.3412516316754934,
      "grad_norm": 0.16532304883003235,
      "learning_rate": 2.196154200018401e-05,
      "loss": 0.5123,
      "mean_token_accuracy": 0.8372174762189388,
      "num_tokens": 812142294.0,
      "step": 25447
    },
    {
      "entropy": 0.5454733148217201,
      "epoch": 2.3413436380469346,
      "grad_norm": 0.1639668196439743,
      "learning_rate": 2.1958475174042385e-05,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8288232609629631,
      "num_tokens": 812174380.0,
      "step": 25448
    },
    {
      "entropy": 0.5418479833751917,
      "epoch": 2.341435644418376,
      "grad_norm": 0.1584101915359497,
      "learning_rate": 2.1955408347900757e-05,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8346309699118137,
      "num_tokens": 812206361.0,
      "step": 25449
    },
    {
      "entropy": 0.5203007548116148,
      "epoch": 2.3415276507898173,
      "grad_norm": 0.16085536777973175,
      "learning_rate": 2.195234152175913e-05,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8319733627140522,
      "num_tokens": 812238566.0,
      "step": 25450
    },
    {
      "entropy": 0.4998633998911828,
      "epoch": 2.3416196571612584,
      "grad_norm": 0.1556505560874939,
      "learning_rate": 2.1949274695617506e-05,
      "loss": 0.481,
      "mean_token_accuracy": 0.8480337485671043,
      "num_tokens": 812270495.0,
      "step": 25451
    },
    {
      "entropy": 0.4017998707713559,
      "epoch": 2.3417116635326995,
      "grad_norm": 0.1394970715045929,
      "learning_rate": 2.194620786947588e-05,
      "loss": 0.3807,
      "mean_token_accuracy": 0.8769436553120613,
      "num_tokens": 812302297.0,
      "step": 25452
    },
    {
      "entropy": 0.5944263525307178,
      "epoch": 2.3418036699041407,
      "grad_norm": 0.1681579202413559,
      "learning_rate": 2.1943141043334252e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.8182437345385551,
      "num_tokens": 812334181.0,
      "step": 25453
    },
    {
      "entropy": 0.42600085120648146,
      "epoch": 2.3418956762755823,
      "grad_norm": 0.14524385333061218,
      "learning_rate": 2.1940074217192626e-05,
      "loss": 0.4072,
      "mean_token_accuracy": 0.8663834519684315,
      "num_tokens": 812366115.0,
      "step": 25454
    },
    {
      "entropy": 0.5549150509759784,
      "epoch": 2.3419876826470234,
      "grad_norm": 0.16258782148361206,
      "learning_rate": 2.1937007391051e-05,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8301172442734241,
      "num_tokens": 812398883.0,
      "step": 25455
    },
    {
      "entropy": 0.42173649184405804,
      "epoch": 2.3420796890184645,
      "grad_norm": 0.144321471452713,
      "learning_rate": 2.193394056490938e-05,
      "loss": 0.4096,
      "mean_token_accuracy": 0.8691992536187172,
      "num_tokens": 812430459.0,
      "step": 25456
    },
    {
      "entropy": 0.5079072220250964,
      "epoch": 2.3421716953899057,
      "grad_norm": 0.15350230038166046,
      "learning_rate": 2.193087373876775e-05,
      "loss": 0.4948,
      "mean_token_accuracy": 0.8424523696303368,
      "num_tokens": 812463227.0,
      "step": 25457
    },
    {
      "entropy": 0.5823821667581797,
      "epoch": 2.342263701761347,
      "grad_norm": 0.17033424973487854,
      "learning_rate": 2.1927806912626125e-05,
      "loss": 0.5782,
      "mean_token_accuracy": 0.818702831864357,
      "num_tokens": 812495181.0,
      "step": 25458
    },
    {
      "entropy": 0.5308910086750984,
      "epoch": 2.3423557081327884,
      "grad_norm": 0.1621832549571991,
      "learning_rate": 2.19247400864845e-05,
      "loss": 0.5271,
      "mean_token_accuracy": 0.8366891220211983,
      "num_tokens": 812527038.0,
      "step": 25459
    },
    {
      "entropy": 0.548253545537591,
      "epoch": 2.3424477145042295,
      "grad_norm": 0.1677302122116089,
      "learning_rate": 2.1921673260342874e-05,
      "loss": 0.5371,
      "mean_token_accuracy": 0.8302614837884903,
      "num_tokens": 812559574.0,
      "step": 25460
    },
    {
      "entropy": 0.5586397424340248,
      "epoch": 2.3425397208756706,
      "grad_norm": 0.17447374761104584,
      "learning_rate": 2.1918606434201246e-05,
      "loss": 0.5419,
      "mean_token_accuracy": 0.8244168795645237,
      "num_tokens": 812589056.0,
      "step": 25461
    },
    {
      "entropy": 0.4807878490537405,
      "epoch": 2.3426317272471118,
      "grad_norm": 0.16216105222702026,
      "learning_rate": 2.191553960805962e-05,
      "loss": 0.4833,
      "mean_token_accuracy": 0.8459464237093925,
      "num_tokens": 812621391.0,
      "step": 25462
    },
    {
      "entropy": 0.5951250642538071,
      "epoch": 2.342723733618553,
      "grad_norm": 0.16795437037944794,
      "learning_rate": 2.1912472781917995e-05,
      "loss": 0.592,
      "mean_token_accuracy": 0.81300925090909,
      "num_tokens": 812653615.0,
      "step": 25463
    },
    {
      "entropy": 0.540901679545641,
      "epoch": 2.3428157399899945,
      "grad_norm": 0.15949152410030365,
      "learning_rate": 2.190940595577637e-05,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8353219628334045,
      "num_tokens": 812685842.0,
      "step": 25464
    },
    {
      "entropy": 0.6479253461584449,
      "epoch": 2.3429077463614356,
      "grad_norm": 0.18218378722667694,
      "learning_rate": 2.1906339129634744e-05,
      "loss": 0.6355,
      "mean_token_accuracy": 0.8030321076512337,
      "num_tokens": 812717335.0,
      "step": 25465
    },
    {
      "entropy": 0.6048780828714371,
      "epoch": 2.3429997527328768,
      "grad_norm": 0.16796055436134338,
      "learning_rate": 2.1903272303493115e-05,
      "loss": 0.5977,
      "mean_token_accuracy": 0.8125480636954308,
      "num_tokens": 812749515.0,
      "step": 25466
    },
    {
      "entropy": 0.5847743395715952,
      "epoch": 2.343091759104318,
      "grad_norm": 0.16210268437862396,
      "learning_rate": 2.190020547735149e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8220599256455898,
      "num_tokens": 812781825.0,
      "step": 25467
    },
    {
      "entropy": 0.5010995636694133,
      "epoch": 2.343183765475759,
      "grad_norm": 0.15597756206989288,
      "learning_rate": 2.1897138651209865e-05,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8420792371034622,
      "num_tokens": 812813740.0,
      "step": 25468
    },
    {
      "entropy": 0.46909144427627325,
      "epoch": 2.3432757718472006,
      "grad_norm": 0.15258149802684784,
      "learning_rate": 2.189407182506824e-05,
      "loss": 0.4456,
      "mean_token_accuracy": 0.856033630669117,
      "num_tokens": 812845589.0,
      "step": 25469
    },
    {
      "entropy": 0.5849519092589617,
      "epoch": 2.3433677782186417,
      "grad_norm": 0.16907817125320435,
      "learning_rate": 2.189100499892661e-05,
      "loss": 0.5788,
      "mean_token_accuracy": 0.819702785462141,
      "num_tokens": 812878002.0,
      "step": 25470
    },
    {
      "entropy": 0.4791831113398075,
      "epoch": 2.343459784590083,
      "grad_norm": 0.15499307215213776,
      "learning_rate": 2.1887938172784985e-05,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8488266356289387,
      "num_tokens": 812909744.0,
      "step": 25471
    },
    {
      "entropy": 0.5931985691422597,
      "epoch": 2.343551790961524,
      "grad_norm": 0.1661878526210785,
      "learning_rate": 2.188487134664336e-05,
      "loss": 0.5884,
      "mean_token_accuracy": 0.8178095668554306,
      "num_tokens": 812942272.0,
      "step": 25472
    },
    {
      "entropy": 0.4364624600857496,
      "epoch": 2.343643797332965,
      "grad_norm": 0.1471676379442215,
      "learning_rate": 2.1881804520501735e-05,
      "loss": 0.4212,
      "mean_token_accuracy": 0.8646317385137081,
      "num_tokens": 812974429.0,
      "step": 25473
    },
    {
      "entropy": 0.5968209709972143,
      "epoch": 2.3437358037044067,
      "grad_norm": 0.16855081915855408,
      "learning_rate": 2.1878737694360106e-05,
      "loss": 0.5866,
      "mean_token_accuracy": 0.814894150942564,
      "num_tokens": 813006050.0,
      "step": 25474
    },
    {
      "entropy": 0.46720799524337053,
      "epoch": 2.343827810075848,
      "grad_norm": 0.15784315764904022,
      "learning_rate": 2.187567086821848e-05,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8521753400564194,
      "num_tokens": 813038005.0,
      "step": 25475
    },
    {
      "entropy": 0.5766229946166277,
      "epoch": 2.343919816447289,
      "grad_norm": 0.16480837762355804,
      "learning_rate": 2.1872604042076855e-05,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8236271068453789,
      "num_tokens": 813069879.0,
      "step": 25476
    },
    {
      "entropy": 0.5192772122099996,
      "epoch": 2.34401182281873,
      "grad_norm": 0.16236136853694916,
      "learning_rate": 2.186953721593523e-05,
      "loss": 0.5018,
      "mean_token_accuracy": 0.8390846401453018,
      "num_tokens": 813101466.0,
      "step": 25477
    },
    {
      "entropy": 0.4606723878532648,
      "epoch": 2.3441038291901712,
      "grad_norm": 0.1524319350719452,
      "learning_rate": 2.1866470389793604e-05,
      "loss": 0.4503,
      "mean_token_accuracy": 0.8538313545286655,
      "num_tokens": 813133343.0,
      "step": 25478
    },
    {
      "entropy": 0.4967601280659437,
      "epoch": 2.344195835561613,
      "grad_norm": 0.16239438951015472,
      "learning_rate": 2.1863403563651976e-05,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8446178585290909,
      "num_tokens": 813165547.0,
      "step": 25479
    },
    {
      "entropy": 0.41867832792922854,
      "epoch": 2.344287841933054,
      "grad_norm": 0.14488038420677185,
      "learning_rate": 2.186033673751035e-05,
      "loss": 0.4057,
      "mean_token_accuracy": 0.8691241145133972,
      "num_tokens": 813197629.0,
      "step": 25480
    },
    {
      "entropy": 0.5686537232249975,
      "epoch": 2.344379848304495,
      "grad_norm": 0.1667676866054535,
      "learning_rate": 2.1857269911368725e-05,
      "loss": 0.5615,
      "mean_token_accuracy": 0.826090432703495,
      "num_tokens": 813229091.0,
      "step": 25481
    },
    {
      "entropy": 0.5186142884194851,
      "epoch": 2.344471854675936,
      "grad_norm": 0.16285906732082367,
      "learning_rate": 2.18542030852271e-05,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8337593525648117,
      "num_tokens": 813260911.0,
      "step": 25482
    },
    {
      "entropy": 0.4975816700607538,
      "epoch": 2.3445638610473774,
      "grad_norm": 0.16110311448574066,
      "learning_rate": 2.185113625908547e-05,
      "loss": 0.4782,
      "mean_token_accuracy": 0.8495219573378563,
      "num_tokens": 813291858.0,
      "step": 25483
    },
    {
      "entropy": 0.5528094246983528,
      "epoch": 2.344655867418819,
      "grad_norm": 0.16456927359104156,
      "learning_rate": 2.1848069432943846e-05,
      "loss": 0.547,
      "mean_token_accuracy": 0.8264799863100052,
      "num_tokens": 813324056.0,
      "step": 25484
    },
    {
      "entropy": 0.4648997550830245,
      "epoch": 2.34474787379026,
      "grad_norm": 0.15551625192165375,
      "learning_rate": 2.184500260680222e-05,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8555510379374027,
      "num_tokens": 813356099.0,
      "step": 25485
    },
    {
      "entropy": 0.5076014376245439,
      "epoch": 2.344839880161701,
      "grad_norm": 0.16184164583683014,
      "learning_rate": 2.1841935780660598e-05,
      "loss": 0.4952,
      "mean_token_accuracy": 0.8434931002557278,
      "num_tokens": 813386974.0,
      "step": 25486
    },
    {
      "entropy": 0.5601502731442451,
      "epoch": 2.3449318865331423,
      "grad_norm": 0.16461370885372162,
      "learning_rate": 2.183886895451897e-05,
      "loss": 0.5417,
      "mean_token_accuracy": 0.8263065554201603,
      "num_tokens": 813418985.0,
      "step": 25487
    },
    {
      "entropy": 0.44750552135519683,
      "epoch": 2.3450238929045835,
      "grad_norm": 0.15027721226215363,
      "learning_rate": 2.1835802128377344e-05,
      "loss": 0.4305,
      "mean_token_accuracy": 0.8621895685791969,
      "num_tokens": 813451709.0,
      "step": 25488
    },
    {
      "entropy": 0.5464363377541304,
      "epoch": 2.345115899276025,
      "grad_norm": 0.1684400588274002,
      "learning_rate": 2.183273530223572e-05,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8321278877556324,
      "num_tokens": 813483577.0,
      "step": 25489
    },
    {
      "entropy": 0.429918747395277,
      "epoch": 2.345207905647466,
      "grad_norm": 0.14907759428024292,
      "learning_rate": 2.1829668476094093e-05,
      "loss": 0.4145,
      "mean_token_accuracy": 0.866951160132885,
      "num_tokens": 813515318.0,
      "step": 25490
    },
    {
      "entropy": 0.5515653565526009,
      "epoch": 2.3452999120189073,
      "grad_norm": 0.16412052512168884,
      "learning_rate": 2.1826601649952465e-05,
      "loss": 0.5381,
      "mean_token_accuracy": 0.8276832774281502,
      "num_tokens": 813547927.0,
      "step": 25491
    },
    {
      "entropy": 0.49337410973384976,
      "epoch": 2.3453919183903484,
      "grad_norm": 0.1568285971879959,
      "learning_rate": 2.182353482381084e-05,
      "loss": 0.4781,
      "mean_token_accuracy": 0.8458174802362919,
      "num_tokens": 813579667.0,
      "step": 25492
    },
    {
      "entropy": 0.5295086628757417,
      "epoch": 2.3454839247617896,
      "grad_norm": 0.1575102061033249,
      "learning_rate": 2.1820467997669214e-05,
      "loss": 0.5229,
      "mean_token_accuracy": 0.837444469332695,
      "num_tokens": 813611643.0,
      "step": 25493
    },
    {
      "entropy": 0.538065099157393,
      "epoch": 2.345575931133231,
      "grad_norm": 0.1723933219909668,
      "learning_rate": 2.181740117152759e-05,
      "loss": 0.5356,
      "mean_token_accuracy": 0.8311313688755035,
      "num_tokens": 813642795.0,
      "step": 25494
    },
    {
      "entropy": 0.5819179248064756,
      "epoch": 2.3456679375046723,
      "grad_norm": 0.17238649725914001,
      "learning_rate": 2.1814334345385963e-05,
      "loss": 0.581,
      "mean_token_accuracy": 0.8195812441408634,
      "num_tokens": 813674826.0,
      "step": 25495
    },
    {
      "entropy": 0.504918853752315,
      "epoch": 2.3457599438761134,
      "grad_norm": 0.16363511979579926,
      "learning_rate": 2.1811267519244335e-05,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8464325591921806,
      "num_tokens": 813706278.0,
      "step": 25496
    },
    {
      "entropy": 0.44699965324252844,
      "epoch": 2.3458519502475546,
      "grad_norm": 0.1540464162826538,
      "learning_rate": 2.180820069310271e-05,
      "loss": 0.4316,
      "mean_token_accuracy": 0.8659435659646988,
      "num_tokens": 813738413.0,
      "step": 25497
    },
    {
      "entropy": 0.4806839635130018,
      "epoch": 2.3459439566189957,
      "grad_norm": 0.1576417088508606,
      "learning_rate": 2.1805133866961084e-05,
      "loss": 0.466,
      "mean_token_accuracy": 0.8492762558162212,
      "num_tokens": 813770122.0,
      "step": 25498
    },
    {
      "entropy": 0.5260245096869767,
      "epoch": 2.3460359629904373,
      "grad_norm": 0.15833289921283722,
      "learning_rate": 2.180206704081946e-05,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8439675159752369,
      "num_tokens": 813801452.0,
      "step": 25499
    },
    {
      "entropy": 0.45166791416704655,
      "epoch": 2.3461279693618784,
      "grad_norm": 0.157806396484375,
      "learning_rate": 2.179900021467783e-05,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8558870479464531,
      "num_tokens": 813832875.0,
      "step": 25500
    },
    {
      "entropy": 0.5454323943704367,
      "epoch": 2.3462199757333195,
      "grad_norm": 0.16502529382705688,
      "learning_rate": 2.1795933388536204e-05,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8300463147461414,
      "num_tokens": 813864819.0,
      "step": 25501
    },
    {
      "entropy": 0.5114821959286928,
      "epoch": 2.3463119821047607,
      "grad_norm": 0.15662096440792084,
      "learning_rate": 2.179286656239458e-05,
      "loss": 0.4963,
      "mean_token_accuracy": 0.8400674127042294,
      "num_tokens": 813896102.0,
      "step": 25502
    },
    {
      "entropy": 0.5863871518522501,
      "epoch": 2.346403988476202,
      "grad_norm": 0.1707482635974884,
      "learning_rate": 2.1789799736252954e-05,
      "loss": 0.5772,
      "mean_token_accuracy": 0.8152374438941479,
      "num_tokens": 813927884.0,
      "step": 25503
    },
    {
      "entropy": 0.5480751488357782,
      "epoch": 2.3464959948476434,
      "grad_norm": 0.15938855707645416,
      "learning_rate": 2.1786732910111325e-05,
      "loss": 0.533,
      "mean_token_accuracy": 0.8294633589684963,
      "num_tokens": 813959993.0,
      "step": 25504
    },
    {
      "entropy": 0.5175322098657489,
      "epoch": 2.3465880012190845,
      "grad_norm": 0.16832582652568817,
      "learning_rate": 2.17836660839697e-05,
      "loss": 0.5107,
      "mean_token_accuracy": 0.8397419862449169,
      "num_tokens": 813991485.0,
      "step": 25505
    },
    {
      "entropy": 0.49229856487363577,
      "epoch": 2.3466800075905256,
      "grad_norm": 0.16168928146362305,
      "learning_rate": 2.1780599257828074e-05,
      "loss": 0.4863,
      "mean_token_accuracy": 0.8445522449910641,
      "num_tokens": 814024079.0,
      "step": 25506
    },
    {
      "entropy": 0.4584890897385776,
      "epoch": 2.3467720139619668,
      "grad_norm": 0.15089643001556396,
      "learning_rate": 2.177753243168645e-05,
      "loss": 0.4346,
      "mean_token_accuracy": 0.8629095293581486,
      "num_tokens": 814056386.0,
      "step": 25507
    },
    {
      "entropy": 0.5919065279886127,
      "epoch": 2.346864020333408,
      "grad_norm": 0.16207517683506012,
      "learning_rate": 2.1774465605544824e-05,
      "loss": 0.5744,
      "mean_token_accuracy": 0.8202410787343979,
      "num_tokens": 814088800.0,
      "step": 25508
    },
    {
      "entropy": 0.5766503922641277,
      "epoch": 2.3469560267048495,
      "grad_norm": 0.166274756193161,
      "learning_rate": 2.1771398779403195e-05,
      "loss": 0.5572,
      "mean_token_accuracy": 0.8196210414171219,
      "num_tokens": 814120521.0,
      "step": 25509
    },
    {
      "entropy": 0.4965647291392088,
      "epoch": 2.3470480330762906,
      "grad_norm": 0.155220627784729,
      "learning_rate": 2.176833195326157e-05,
      "loss": 0.4929,
      "mean_token_accuracy": 0.8464587591588497,
      "num_tokens": 814152641.0,
      "step": 25510
    },
    {
      "entropy": 0.5957510583102703,
      "epoch": 2.3471400394477318,
      "grad_norm": 0.16803795099258423,
      "learning_rate": 2.1765265127119944e-05,
      "loss": 0.5842,
      "mean_token_accuracy": 0.8155211731791496,
      "num_tokens": 814184243.0,
      "step": 25511
    },
    {
      "entropy": 0.4789503887295723,
      "epoch": 2.347232045819173,
      "grad_norm": 0.15982306003570557,
      "learning_rate": 2.176219830097832e-05,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8534240648150444,
      "num_tokens": 814214871.0,
      "step": 25512
    },
    {
      "entropy": 0.5084677087143064,
      "epoch": 2.347324052190614,
      "grad_norm": 0.1533605456352234,
      "learning_rate": 2.175913147483669e-05,
      "loss": 0.4967,
      "mean_token_accuracy": 0.8430599272251129,
      "num_tokens": 814247481.0,
      "step": 25513
    },
    {
      "entropy": 0.592767708003521,
      "epoch": 2.3474160585620556,
      "grad_norm": 0.17453302443027496,
      "learning_rate": 2.1756064648695065e-05,
      "loss": 0.5785,
      "mean_token_accuracy": 0.8209718130528927,
      "num_tokens": 814278566.0,
      "step": 25514
    },
    {
      "entropy": 0.4881215086206794,
      "epoch": 2.3475080649334967,
      "grad_norm": 0.16354061663150787,
      "learning_rate": 2.175299782255344e-05,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8462105765938759,
      "num_tokens": 814309567.0,
      "step": 25515
    },
    {
      "entropy": 0.5215962994843721,
      "epoch": 2.347600071304938,
      "grad_norm": 0.16851472854614258,
      "learning_rate": 2.1749930996411814e-05,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8378285691142082,
      "num_tokens": 814341699.0,
      "step": 25516
    },
    {
      "entropy": 0.48173865489661694,
      "epoch": 2.347692077676379,
      "grad_norm": 0.15348295867443085,
      "learning_rate": 2.174686417027019e-05,
      "loss": 0.4665,
      "mean_token_accuracy": 0.8547941297292709,
      "num_tokens": 814373809.0,
      "step": 25517
    },
    {
      "entropy": 0.4607549086213112,
      "epoch": 2.34778408404782,
      "grad_norm": 0.15383747220039368,
      "learning_rate": 2.1743797344128563e-05,
      "loss": 0.4478,
      "mean_token_accuracy": 0.8569968976080418,
      "num_tokens": 814405576.0,
      "step": 25518
    },
    {
      "entropy": 0.4503728412091732,
      "epoch": 2.3478760904192617,
      "grad_norm": 0.1508798748254776,
      "learning_rate": 2.1740730517986938e-05,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8628590814769268,
      "num_tokens": 814437756.0,
      "step": 25519
    },
    {
      "entropy": 0.511397959664464,
      "epoch": 2.347968096790703,
      "grad_norm": 0.16012464463710785,
      "learning_rate": 2.1737663691845313e-05,
      "loss": 0.4877,
      "mean_token_accuracy": 0.8447330221533775,
      "num_tokens": 814468867.0,
      "step": 25520
    },
    {
      "entropy": 0.5165605563670397,
      "epoch": 2.348060103162144,
      "grad_norm": 0.1656806617975235,
      "learning_rate": 2.1734596865703684e-05,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8389580100774765,
      "num_tokens": 814501288.0,
      "step": 25521
    },
    {
      "entropy": 0.5659691393375397,
      "epoch": 2.348152109533585,
      "grad_norm": 0.16268502175807953,
      "learning_rate": 2.173153003956206e-05,
      "loss": 0.552,
      "mean_token_accuracy": 0.8266013823449612,
      "num_tokens": 814533416.0,
      "step": 25522
    },
    {
      "entropy": 0.47721230145543814,
      "epoch": 2.3482441159050262,
      "grad_norm": 0.15361928939819336,
      "learning_rate": 2.1728463213420433e-05,
      "loss": 0.4749,
      "mean_token_accuracy": 0.8488469272851944,
      "num_tokens": 814565689.0,
      "step": 25523
    },
    {
      "entropy": 0.42556167021393776,
      "epoch": 2.348336122276468,
      "grad_norm": 0.14613181352615356,
      "learning_rate": 2.1725396387278808e-05,
      "loss": 0.4092,
      "mean_token_accuracy": 0.8679407723248005,
      "num_tokens": 814598440.0,
      "step": 25524
    },
    {
      "entropy": 0.47475553129334,
      "epoch": 2.348428128647909,
      "grad_norm": 0.16238601505756378,
      "learning_rate": 2.1722329561137182e-05,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8514968678355217,
      "num_tokens": 814629646.0,
      "step": 25525
    },
    {
      "entropy": 0.5209665801376104,
      "epoch": 2.34852013501935,
      "grad_norm": 0.16322912275791168,
      "learning_rate": 2.1719262734995554e-05,
      "loss": 0.5105,
      "mean_token_accuracy": 0.8381665386259556,
      "num_tokens": 814661503.0,
      "step": 25526
    },
    {
      "entropy": 0.4956805529072881,
      "epoch": 2.348612141390791,
      "grad_norm": 0.16124330461025238,
      "learning_rate": 2.171619590885393e-05,
      "loss": 0.4952,
      "mean_token_accuracy": 0.8402320146560669,
      "num_tokens": 814693059.0,
      "step": 25527
    },
    {
      "entropy": 0.47757432656362653,
      "epoch": 2.3487041477622324,
      "grad_norm": 0.15168911218643188,
      "learning_rate": 2.1713129082712303e-05,
      "loss": 0.4674,
      "mean_token_accuracy": 0.8503033556044102,
      "num_tokens": 814725325.0,
      "step": 25528
    },
    {
      "entropy": 0.491467522457242,
      "epoch": 2.348796154133674,
      "grad_norm": 0.15819840133190155,
      "learning_rate": 2.1710062256570678e-05,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8510497473180294,
      "num_tokens": 814757647.0,
      "step": 25529
    },
    {
      "entropy": 0.5125586902722716,
      "epoch": 2.348888160505115,
      "grad_norm": 0.15860676765441895,
      "learning_rate": 2.170699543042905e-05,
      "loss": 0.5016,
      "mean_token_accuracy": 0.8412727862596512,
      "num_tokens": 814789333.0,
      "step": 25530
    },
    {
      "entropy": 0.4345680670812726,
      "epoch": 2.348980166876556,
      "grad_norm": 0.15004311501979828,
      "learning_rate": 2.1703928604287424e-05,
      "loss": 0.4251,
      "mean_token_accuracy": 0.8621978163719177,
      "num_tokens": 814821717.0,
      "step": 25531
    },
    {
      "entropy": 0.4730522462632507,
      "epoch": 2.3490721732479973,
      "grad_norm": 0.15879642963409424,
      "learning_rate": 2.1700861778145798e-05,
      "loss": 0.4703,
      "mean_token_accuracy": 0.851311594247818,
      "num_tokens": 814853722.0,
      "step": 25532
    },
    {
      "entropy": 0.48284289706498384,
      "epoch": 2.3491641796194385,
      "grad_norm": 0.15561191737651825,
      "learning_rate": 2.1697794952004173e-05,
      "loss": 0.4706,
      "mean_token_accuracy": 0.8506996184587479,
      "num_tokens": 814885627.0,
      "step": 25533
    },
    {
      "entropy": 0.49128034338355064,
      "epoch": 2.34925618599088,
      "grad_norm": 0.1576792448759079,
      "learning_rate": 2.1694728125862544e-05,
      "loss": 0.4823,
      "mean_token_accuracy": 0.8460208922624588,
      "num_tokens": 814918173.0,
      "step": 25534
    },
    {
      "entropy": 0.5760758807882667,
      "epoch": 2.349348192362321,
      "grad_norm": 0.1724141240119934,
      "learning_rate": 2.169166129972092e-05,
      "loss": 0.5636,
      "mean_token_accuracy": 0.8246931992471218,
      "num_tokens": 814950030.0,
      "step": 25535
    },
    {
      "entropy": 0.47429997799918056,
      "epoch": 2.3494401987337623,
      "grad_norm": 0.15622468292713165,
      "learning_rate": 2.1688594473579293e-05,
      "loss": 0.4676,
      "mean_token_accuracy": 0.8498798124492168,
      "num_tokens": 814981339.0,
      "step": 25536
    },
    {
      "entropy": 0.5182853741571307,
      "epoch": 2.3495322051052034,
      "grad_norm": 0.16080935299396515,
      "learning_rate": 2.1685527647437668e-05,
      "loss": 0.4945,
      "mean_token_accuracy": 0.8387531638145447,
      "num_tokens": 815013185.0,
      "step": 25537
    },
    {
      "entropy": 0.482073143357411,
      "epoch": 2.3496242114766446,
      "grad_norm": 0.15685586631298065,
      "learning_rate": 2.1682460821296043e-05,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8518820479512215,
      "num_tokens": 815045159.0,
      "step": 25538
    },
    {
      "entropy": 0.49934902880340815,
      "epoch": 2.349716217848086,
      "grad_norm": 0.1566418558359146,
      "learning_rate": 2.1679393995154414e-05,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8424873054027557,
      "num_tokens": 815077428.0,
      "step": 25539
    },
    {
      "entropy": 0.5621266420930624,
      "epoch": 2.3498082242195273,
      "grad_norm": 0.16520380973815918,
      "learning_rate": 2.167632716901279e-05,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8268767930567265,
      "num_tokens": 815109239.0,
      "step": 25540
    },
    {
      "entropy": 0.4683609385974705,
      "epoch": 2.3499002305909684,
      "grad_norm": 0.15029147267341614,
      "learning_rate": 2.1673260342871163e-05,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8527080044150352,
      "num_tokens": 815141673.0,
      "step": 25541
    },
    {
      "entropy": 0.5546943414956331,
      "epoch": 2.3499922369624096,
      "grad_norm": 0.16309970617294312,
      "learning_rate": 2.1670193516729538e-05,
      "loss": 0.5348,
      "mean_token_accuracy": 0.8336633369326591,
      "num_tokens": 815173647.0,
      "step": 25542
    },
    {
      "entropy": 0.5259770657867193,
      "epoch": 2.3500842433338507,
      "grad_norm": 0.16887277364730835,
      "learning_rate": 2.166712669058791e-05,
      "loss": 0.5171,
      "mean_token_accuracy": 0.8369038291275501,
      "num_tokens": 815205197.0,
      "step": 25543
    },
    {
      "entropy": 0.4926290474832058,
      "epoch": 2.3501762497052923,
      "grad_norm": 0.15843896567821503,
      "learning_rate": 2.1664059864446284e-05,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8489904142916203,
      "num_tokens": 815237319.0,
      "step": 25544
    },
    {
      "entropy": 0.48088109493255615,
      "epoch": 2.3502682560767334,
      "grad_norm": 0.15523850917816162,
      "learning_rate": 2.166099303830466e-05,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8518723025918007,
      "num_tokens": 815268535.0,
      "step": 25545
    },
    {
      "entropy": 0.45389842335134745,
      "epoch": 2.3503602624481745,
      "grad_norm": 0.15607406198978424,
      "learning_rate": 2.1657926212163033e-05,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8589620925486088,
      "num_tokens": 815301036.0,
      "step": 25546
    },
    {
      "entropy": 0.51224719802849,
      "epoch": 2.3504522688196157,
      "grad_norm": 0.1598283052444458,
      "learning_rate": 2.1654859386021408e-05,
      "loss": 0.5027,
      "mean_token_accuracy": 0.8421202674508095,
      "num_tokens": 815332764.0,
      "step": 25547
    },
    {
      "entropy": 0.4872558615170419,
      "epoch": 2.350544275191057,
      "grad_norm": 0.1564982831478119,
      "learning_rate": 2.1651792559879782e-05,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8510568663477898,
      "num_tokens": 815364815.0,
      "step": 25548
    },
    {
      "entropy": 0.5222828332334757,
      "epoch": 2.3506362815624984,
      "grad_norm": 0.16078375279903412,
      "learning_rate": 2.1648725733738157e-05,
      "loss": 0.5213,
      "mean_token_accuracy": 0.8376134000718594,
      "num_tokens": 815396931.0,
      "step": 25549
    },
    {
      "entropy": 0.5678416937589645,
      "epoch": 2.3507282879339395,
      "grad_norm": 0.16722029447555542,
      "learning_rate": 2.164565890759653e-05,
      "loss": 0.5619,
      "mean_token_accuracy": 0.8236617892980576,
      "num_tokens": 815428765.0,
      "step": 25550
    },
    {
      "entropy": 0.5564559602644295,
      "epoch": 2.3508202943053806,
      "grad_norm": 0.16449666023254395,
      "learning_rate": 2.1642592081454903e-05,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8328217566013336,
      "num_tokens": 815461189.0,
      "step": 25551
    },
    {
      "entropy": 0.6568860225379467,
      "epoch": 2.3509123006768218,
      "grad_norm": 0.18105565011501312,
      "learning_rate": 2.1639525255313278e-05,
      "loss": 0.6464,
      "mean_token_accuracy": 0.7991916909813881,
      "num_tokens": 815492685.0,
      "step": 25552
    },
    {
      "entropy": 0.5091654974967241,
      "epoch": 2.351004307048263,
      "grad_norm": 0.15962937474250793,
      "learning_rate": 2.1636458429171652e-05,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8437994681298733,
      "num_tokens": 815525028.0,
      "step": 25553
    },
    {
      "entropy": 0.5127290720120072,
      "epoch": 2.3510963134197045,
      "grad_norm": 0.15479204058647156,
      "learning_rate": 2.1633391603030027e-05,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8405519053339958,
      "num_tokens": 815557151.0,
      "step": 25554
    },
    {
      "entropy": 0.5216312250122428,
      "epoch": 2.3511883197911456,
      "grad_norm": 0.16369125247001648,
      "learning_rate": 2.16303247768884e-05,
      "loss": 0.512,
      "mean_token_accuracy": 0.8362464383244514,
      "num_tokens": 815587898.0,
      "step": 25555
    },
    {
      "entropy": 0.4512544125318527,
      "epoch": 2.3512803261625868,
      "grad_norm": 0.14999957382678986,
      "learning_rate": 2.1627257950746773e-05,
      "loss": 0.439,
      "mean_token_accuracy": 0.8572876080870628,
      "num_tokens": 815620365.0,
      "step": 25556
    },
    {
      "entropy": 0.5173904364928603,
      "epoch": 2.351372332534028,
      "grad_norm": 0.1531398445367813,
      "learning_rate": 2.1624191124605147e-05,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8372313007712364,
      "num_tokens": 815653133.0,
      "step": 25557
    },
    {
      "entropy": 0.5663134381175041,
      "epoch": 2.351464338905469,
      "grad_norm": 0.17110680043697357,
      "learning_rate": 2.1621124298463522e-05,
      "loss": 0.5577,
      "mean_token_accuracy": 0.8219251595437527,
      "num_tokens": 815683968.0,
      "step": 25558
    },
    {
      "entropy": 0.5994828175753355,
      "epoch": 2.3515563452769106,
      "grad_norm": 0.1700032353401184,
      "learning_rate": 2.1618057472321897e-05,
      "loss": 0.5806,
      "mean_token_accuracy": 0.8138868138194084,
      "num_tokens": 815715592.0,
      "step": 25559
    },
    {
      "entropy": 0.6028104955330491,
      "epoch": 2.3516483516483517,
      "grad_norm": 0.17332342267036438,
      "learning_rate": 2.1614990646180268e-05,
      "loss": 0.5983,
      "mean_token_accuracy": 0.8103469535708427,
      "num_tokens": 815747399.0,
      "step": 25560
    },
    {
      "entropy": 0.46437558624893427,
      "epoch": 2.351740358019793,
      "grad_norm": 0.15734434127807617,
      "learning_rate": 2.1611923820038643e-05,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8530928008258343,
      "num_tokens": 815779496.0,
      "step": 25561
    },
    {
      "entropy": 0.6076374407857656,
      "epoch": 2.351832364391234,
      "grad_norm": 0.17800669372081757,
      "learning_rate": 2.1608856993897017e-05,
      "loss": 0.5965,
      "mean_token_accuracy": 0.8121157363057137,
      "num_tokens": 815811242.0,
      "step": 25562
    },
    {
      "entropy": 0.6045909151434898,
      "epoch": 2.351924370762675,
      "grad_norm": 0.17616799473762512,
      "learning_rate": 2.1605790167755392e-05,
      "loss": 0.5988,
      "mean_token_accuracy": 0.8118979968130589,
      "num_tokens": 815842706.0,
      "step": 25563
    },
    {
      "entropy": 0.4663533801212907,
      "epoch": 2.3520163771341167,
      "grad_norm": 0.15826363861560822,
      "learning_rate": 2.1602723341613763e-05,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8530286252498627,
      "num_tokens": 815874274.0,
      "step": 25564
    },
    {
      "entropy": 0.5749657060950994,
      "epoch": 2.352108383505558,
      "grad_norm": 0.16476579010486603,
      "learning_rate": 2.1599656515472138e-05,
      "loss": 0.5662,
      "mean_token_accuracy": 0.8197138495743275,
      "num_tokens": 815906434.0,
      "step": 25565
    },
    {
      "entropy": 0.5309097203426063,
      "epoch": 2.352200389876999,
      "grad_norm": 0.16898392140865326,
      "learning_rate": 2.1596589689330512e-05,
      "loss": 0.5233,
      "mean_token_accuracy": 0.837357122451067,
      "num_tokens": 815938046.0,
      "step": 25566
    },
    {
      "entropy": 0.43522981321439147,
      "epoch": 2.35229239624844,
      "grad_norm": 0.1541394144296646,
      "learning_rate": 2.1593522863188887e-05,
      "loss": 0.4198,
      "mean_token_accuracy": 0.8650204986333847,
      "num_tokens": 815970110.0,
      "step": 25567
    },
    {
      "entropy": 0.4052933019120246,
      "epoch": 2.3523844026198812,
      "grad_norm": 0.14180152118206024,
      "learning_rate": 2.1590456037047262e-05,
      "loss": 0.3875,
      "mean_token_accuracy": 0.8753866739571095,
      "num_tokens": 816002544.0,
      "step": 25568
    },
    {
      "entropy": 0.5516154859215021,
      "epoch": 2.352476408991323,
      "grad_norm": 0.16606026887893677,
      "learning_rate": 2.1587389210905633e-05,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8358344919979572,
      "num_tokens": 816034626.0,
      "step": 25569
    },
    {
      "entropy": 0.6056702909991145,
      "epoch": 2.352568415362764,
      "grad_norm": 0.1668311506509781,
      "learning_rate": 2.1584322384764008e-05,
      "loss": 0.5869,
      "mean_token_accuracy": 0.818265225738287,
      "num_tokens": 816066660.0,
      "step": 25570
    },
    {
      "entropy": 0.4883196750888601,
      "epoch": 2.352660421734205,
      "grad_norm": 0.15798375010490417,
      "learning_rate": 2.1581255558622382e-05,
      "loss": 0.472,
      "mean_token_accuracy": 0.8503862135112286,
      "num_tokens": 816098492.0,
      "step": 25571
    },
    {
      "entropy": 0.5110103543847799,
      "epoch": 2.352752428105646,
      "grad_norm": 0.15466390550136566,
      "learning_rate": 2.1578188732480757e-05,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8422628752887249,
      "num_tokens": 816131200.0,
      "step": 25572
    },
    {
      "entropy": 0.4767919606529176,
      "epoch": 2.3528444344770874,
      "grad_norm": 0.15888220071792603,
      "learning_rate": 2.1575121906339128e-05,
      "loss": 0.462,
      "mean_token_accuracy": 0.8547418713569641,
      "num_tokens": 816163572.0,
      "step": 25573
    },
    {
      "entropy": 0.5476997280493379,
      "epoch": 2.352936440848529,
      "grad_norm": 0.16318702697753906,
      "learning_rate": 2.1572055080197503e-05,
      "loss": 0.5332,
      "mean_token_accuracy": 0.8341124504804611,
      "num_tokens": 816195469.0,
      "step": 25574
    },
    {
      "entropy": 0.5669013652950525,
      "epoch": 2.35302844721997,
      "grad_norm": 0.16905243694782257,
      "learning_rate": 2.1568988254055878e-05,
      "loss": 0.5556,
      "mean_token_accuracy": 0.824676651507616,
      "num_tokens": 816226974.0,
      "step": 25575
    },
    {
      "entropy": 0.49102861247956753,
      "epoch": 2.353120453591411,
      "grad_norm": 0.15948379039764404,
      "learning_rate": 2.1565921427914252e-05,
      "loss": 0.4764,
      "mean_token_accuracy": 0.8489497527480125,
      "num_tokens": 816257993.0,
      "step": 25576
    },
    {
      "entropy": 0.5559776686131954,
      "epoch": 2.3532124599628523,
      "grad_norm": 0.1703823357820511,
      "learning_rate": 2.1562854601772627e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8217390589416027,
      "num_tokens": 816290541.0,
      "step": 25577
    },
    {
      "entropy": 0.4941693707369268,
      "epoch": 2.3533044663342935,
      "grad_norm": 0.15395547449588776,
      "learning_rate": 2.1559787775631e-05,
      "loss": 0.475,
      "mean_token_accuracy": 0.8487050831317902,
      "num_tokens": 816322879.0,
      "step": 25578
    },
    {
      "entropy": 0.5545475562103093,
      "epoch": 2.353396472705735,
      "grad_norm": 0.16893307864665985,
      "learning_rate": 2.1556720949489376e-05,
      "loss": 0.5444,
      "mean_token_accuracy": 0.8308612927794456,
      "num_tokens": 816354823.0,
      "step": 25579
    },
    {
      "entropy": 0.47842646134085953,
      "epoch": 2.353488479077176,
      "grad_norm": 0.1566786766052246,
      "learning_rate": 2.155365412334775e-05,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8542296215891838,
      "num_tokens": 816386792.0,
      "step": 25580
    },
    {
      "entropy": 0.5240861345082521,
      "epoch": 2.3535804854486173,
      "grad_norm": 0.1621512770652771,
      "learning_rate": 2.1550587297206122e-05,
      "loss": 0.5093,
      "mean_token_accuracy": 0.841554056853056,
      "num_tokens": 816417915.0,
      "step": 25581
    },
    {
      "entropy": 0.4336992676835507,
      "epoch": 2.3536724918200584,
      "grad_norm": 0.14858101308345795,
      "learning_rate": 2.1547520471064497e-05,
      "loss": 0.4281,
      "mean_token_accuracy": 0.8604051545262337,
      "num_tokens": 816449582.0,
      "step": 25582
    },
    {
      "entropy": 0.5198017777875066,
      "epoch": 2.3537644981914996,
      "grad_norm": 0.16624000668525696,
      "learning_rate": 2.154445364492287e-05,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8378312066197395,
      "num_tokens": 816481962.0,
      "step": 25583
    },
    {
      "entropy": 0.44757992774248123,
      "epoch": 2.353856504562941,
      "grad_norm": 0.14780934154987335,
      "learning_rate": 2.1541386818781246e-05,
      "loss": 0.4271,
      "mean_token_accuracy": 0.862838875502348,
      "num_tokens": 816513524.0,
      "step": 25584
    },
    {
      "entropy": 0.5664929980412126,
      "epoch": 2.3539485109343823,
      "grad_norm": 0.1639649122953415,
      "learning_rate": 2.153831999263962e-05,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8272959515452385,
      "num_tokens": 816545916.0,
      "step": 25585
    },
    {
      "entropy": 0.582987162284553,
      "epoch": 2.3540405173058234,
      "grad_norm": 0.16679075360298157,
      "learning_rate": 2.1535253166497992e-05,
      "loss": 0.5684,
      "mean_token_accuracy": 0.8194297179579735,
      "num_tokens": 816578333.0,
      "step": 25586
    },
    {
      "entropy": 0.5221044742502272,
      "epoch": 2.3541325236772646,
      "grad_norm": 0.15885941684246063,
      "learning_rate": 2.1532186340356367e-05,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8361896984279156,
      "num_tokens": 816610057.0,
      "step": 25587
    },
    {
      "entropy": 0.44473579013720155,
      "epoch": 2.3542245300487057,
      "grad_norm": 0.1488819569349289,
      "learning_rate": 2.152911951421474e-05,
      "loss": 0.4275,
      "mean_token_accuracy": 0.8619626276195049,
      "num_tokens": 816642825.0,
      "step": 25588
    },
    {
      "entropy": 0.5917569305747747,
      "epoch": 2.3543165364201473,
      "grad_norm": 0.16910360753536224,
      "learning_rate": 2.1526052688073116e-05,
      "loss": 0.5703,
      "mean_token_accuracy": 0.8195075578987598,
      "num_tokens": 816673935.0,
      "step": 25589
    },
    {
      "entropy": 0.4920499362051487,
      "epoch": 2.3544085427915884,
      "grad_norm": 0.1546858698129654,
      "learning_rate": 2.1522985861931487e-05,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8429812155663967,
      "num_tokens": 816705699.0,
      "step": 25590
    },
    {
      "entropy": 0.5875488426536322,
      "epoch": 2.3545005491630295,
      "grad_norm": 0.1708909571170807,
      "learning_rate": 2.1519919035789862e-05,
      "loss": 0.5777,
      "mean_token_accuracy": 0.8203585632145405,
      "num_tokens": 816737839.0,
      "step": 25591
    },
    {
      "entropy": 0.4680969570763409,
      "epoch": 2.3545925555344707,
      "grad_norm": 0.15661120414733887,
      "learning_rate": 2.1516852209648236e-05,
      "loss": 0.4583,
      "mean_token_accuracy": 0.852925144135952,
      "num_tokens": 816769647.0,
      "step": 25592
    },
    {
      "entropy": 0.45414275932125747,
      "epoch": 2.354684561905912,
      "grad_norm": 0.14919233322143555,
      "learning_rate": 2.151378538350661e-05,
      "loss": 0.4491,
      "mean_token_accuracy": 0.8569858334958553,
      "num_tokens": 816802415.0,
      "step": 25593
    },
    {
      "entropy": 0.4731817925348878,
      "epoch": 2.3547765682773534,
      "grad_norm": 0.15471072494983673,
      "learning_rate": 2.1510718557364982e-05,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8501587361097336,
      "num_tokens": 816834802.0,
      "step": 25594
    },
    {
      "entropy": 0.4534338731318712,
      "epoch": 2.3548685746487945,
      "grad_norm": 0.15989159047603607,
      "learning_rate": 2.1507651731223357e-05,
      "loss": 0.4417,
      "mean_token_accuracy": 0.857146292924881,
      "num_tokens": 816866650.0,
      "step": 25595
    },
    {
      "entropy": 0.5043223090469837,
      "epoch": 2.3549605810202356,
      "grad_norm": 0.15697763860225677,
      "learning_rate": 2.150458490508173e-05,
      "loss": 0.4982,
      "mean_token_accuracy": 0.8417927771806717,
      "num_tokens": 816898848.0,
      "step": 25596
    },
    {
      "entropy": 0.5263870419003069,
      "epoch": 2.3550525873916768,
      "grad_norm": 0.17125031352043152,
      "learning_rate": 2.1501518078940106e-05,
      "loss": 0.5047,
      "mean_token_accuracy": 0.840131115168333,
      "num_tokens": 816929934.0,
      "step": 25597
    },
    {
      "entropy": 0.49979779589921236,
      "epoch": 2.355144593763118,
      "grad_norm": 0.1604481339454651,
      "learning_rate": 2.149845125279848e-05,
      "loss": 0.477,
      "mean_token_accuracy": 0.8461016342043877,
      "num_tokens": 816960787.0,
      "step": 25598
    },
    {
      "entropy": 0.5557303186506033,
      "epoch": 2.3552366001345595,
      "grad_norm": 0.16758623719215393,
      "learning_rate": 2.1495384426656852e-05,
      "loss": 0.5433,
      "mean_token_accuracy": 0.8315826915204525,
      "num_tokens": 816992090.0,
      "step": 25599
    },
    {
      "entropy": 0.5300596617162228,
      "epoch": 2.3553286065060006,
      "grad_norm": 0.1611204892396927,
      "learning_rate": 2.1492317600515227e-05,
      "loss": 0.5169,
      "mean_token_accuracy": 0.834592416882515,
      "num_tokens": 817024702.0,
      "step": 25600
    },
    {
      "entropy": 0.5046031130477786,
      "epoch": 2.3554206128774418,
      "grad_norm": 0.1574343591928482,
      "learning_rate": 2.14892507743736e-05,
      "loss": 0.4979,
      "mean_token_accuracy": 0.8427225165069103,
      "num_tokens": 817056690.0,
      "step": 25601
    },
    {
      "entropy": 0.5126386061310768,
      "epoch": 2.355512619248883,
      "grad_norm": 0.162128746509552,
      "learning_rate": 2.1486183948231976e-05,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8402947336435318,
      "num_tokens": 817087777.0,
      "step": 25602
    },
    {
      "entropy": 0.5058175707235932,
      "epoch": 2.355604625620324,
      "grad_norm": 0.15806211531162262,
      "learning_rate": 2.1483117122090347e-05,
      "loss": 0.4846,
      "mean_token_accuracy": 0.8425051495432854,
      "num_tokens": 817119012.0,
      "step": 25603
    },
    {
      "entropy": 0.5649839667603374,
      "epoch": 2.3556966319917656,
      "grad_norm": 0.16771627962589264,
      "learning_rate": 2.1480050295948722e-05,
      "loss": 0.5438,
      "mean_token_accuracy": 0.8253143057227135,
      "num_tokens": 817150397.0,
      "step": 25604
    },
    {
      "entropy": 0.5660046618431807,
      "epoch": 2.3557886383632067,
      "grad_norm": 0.17099477350711823,
      "learning_rate": 2.1476983469807097e-05,
      "loss": 0.544,
      "mean_token_accuracy": 0.8274915181100368,
      "num_tokens": 817182293.0,
      "step": 25605
    },
    {
      "entropy": 0.5316700395196676,
      "epoch": 2.355880644734648,
      "grad_norm": 0.16503159701824188,
      "learning_rate": 2.147391664366547e-05,
      "loss": 0.5231,
      "mean_token_accuracy": 0.8363272920250893,
      "num_tokens": 817214170.0,
      "step": 25606
    },
    {
      "entropy": 0.5452999756671488,
      "epoch": 2.355972651106089,
      "grad_norm": 0.1652543842792511,
      "learning_rate": 2.1470849817523846e-05,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8308203332126141,
      "num_tokens": 817246729.0,
      "step": 25607
    },
    {
      "entropy": 0.5357951140031219,
      "epoch": 2.35606465747753,
      "grad_norm": 0.16268602013587952,
      "learning_rate": 2.146778299138222e-05,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8359585143625736,
      "num_tokens": 817278576.0,
      "step": 25608
    },
    {
      "entropy": 0.5215537939220667,
      "epoch": 2.3561566638489717,
      "grad_norm": 0.16082532703876495,
      "learning_rate": 2.1464716165240595e-05,
      "loss": 0.5085,
      "mean_token_accuracy": 0.8383247777819633,
      "num_tokens": 817310690.0,
      "step": 25609
    },
    {
      "entropy": 0.5294525735080242,
      "epoch": 2.356248670220413,
      "grad_norm": 0.1600373238325119,
      "learning_rate": 2.146164933909897e-05,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8371956571936607,
      "num_tokens": 817343236.0,
      "step": 25610
    },
    {
      "entropy": 0.5656447308138013,
      "epoch": 2.356340676591854,
      "grad_norm": 0.16117241978645325,
      "learning_rate": 2.145858251295734e-05,
      "loss": 0.5475,
      "mean_token_accuracy": 0.8306212536990643,
      "num_tokens": 817375568.0,
      "step": 25611
    },
    {
      "entropy": 0.48194982297718525,
      "epoch": 2.356432682963295,
      "grad_norm": 0.15366755425930023,
      "learning_rate": 2.1455515686815716e-05,
      "loss": 0.4721,
      "mean_token_accuracy": 0.8473949432373047,
      "num_tokens": 817407555.0,
      "step": 25612
    },
    {
      "entropy": 0.5648027565330267,
      "epoch": 2.3565246893347362,
      "grad_norm": 0.17056074738502502,
      "learning_rate": 2.145244886067409e-05,
      "loss": 0.5592,
      "mean_token_accuracy": 0.8233402110636234,
      "num_tokens": 817439453.0,
      "step": 25613
    },
    {
      "entropy": 0.4962919168174267,
      "epoch": 2.356616695706178,
      "grad_norm": 0.1594206690788269,
      "learning_rate": 2.1449382034532465e-05,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8511606268584728,
      "num_tokens": 817471678.0,
      "step": 25614
    },
    {
      "entropy": 0.5171406408771873,
      "epoch": 2.356708702077619,
      "grad_norm": 0.16713212430477142,
      "learning_rate": 2.144631520839084e-05,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8375032395124435,
      "num_tokens": 817502897.0,
      "step": 25615
    },
    {
      "entropy": 0.54277806263417,
      "epoch": 2.35680070844906,
      "grad_norm": 0.16235610842704773,
      "learning_rate": 2.144324838224921e-05,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8296985700726509,
      "num_tokens": 817534324.0,
      "step": 25616
    },
    {
      "entropy": 0.5825234055519104,
      "epoch": 2.3568927148205012,
      "grad_norm": 0.17670093476772308,
      "learning_rate": 2.1440181556107586e-05,
      "loss": 0.5694,
      "mean_token_accuracy": 0.8185390345752239,
      "num_tokens": 817565683.0,
      "step": 25617
    },
    {
      "entropy": 0.5691337902098894,
      "epoch": 2.3569847211919424,
      "grad_norm": 0.16679124534130096,
      "learning_rate": 2.143711472996596e-05,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8257592432200909,
      "num_tokens": 817598091.0,
      "step": 25618
    },
    {
      "entropy": 0.4139093500562012,
      "epoch": 2.357076727563384,
      "grad_norm": 0.1456737220287323,
      "learning_rate": 2.1434047903824335e-05,
      "loss": 0.4076,
      "mean_token_accuracy": 0.8706641644239426,
      "num_tokens": 817630016.0,
      "step": 25619
    },
    {
      "entropy": 0.5804222822189331,
      "epoch": 2.357168733934825,
      "grad_norm": 0.16375882923603058,
      "learning_rate": 2.1430981077682706e-05,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8216579481959343,
      "num_tokens": 817661997.0,
      "step": 25620
    },
    {
      "entropy": 0.5324945393949747,
      "epoch": 2.357260740306266,
      "grad_norm": 0.159685879945755,
      "learning_rate": 2.142791425154108e-05,
      "loss": 0.5215,
      "mean_token_accuracy": 0.8338510654866695,
      "num_tokens": 817694164.0,
      "step": 25621
    },
    {
      "entropy": 0.5544083118438721,
      "epoch": 2.3573527466777073,
      "grad_norm": 0.15835680067539215,
      "learning_rate": 2.1424847425399455e-05,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8290276080369949,
      "num_tokens": 817726635.0,
      "step": 25622
    },
    {
      "entropy": 0.5337888263165951,
      "epoch": 2.3574447530491485,
      "grad_norm": 0.1615951508283615,
      "learning_rate": 2.142178059925783e-05,
      "loss": 0.5203,
      "mean_token_accuracy": 0.8379697278141975,
      "num_tokens": 817759077.0,
      "step": 25623
    },
    {
      "entropy": 0.5940871611237526,
      "epoch": 2.35753675942059,
      "grad_norm": 0.1735890507698059,
      "learning_rate": 2.14187137731162e-05,
      "loss": 0.581,
      "mean_token_accuracy": 0.8143941722810268,
      "num_tokens": 817789969.0,
      "step": 25624
    },
    {
      "entropy": 0.4898594841361046,
      "epoch": 2.357628765792031,
      "grad_norm": 0.15761186182498932,
      "learning_rate": 2.1415646946974576e-05,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8472796343266964,
      "num_tokens": 817821580.0,
      "step": 25625
    },
    {
      "entropy": 0.5220054658129811,
      "epoch": 2.3577207721634723,
      "grad_norm": 0.16297467052936554,
      "learning_rate": 2.141258012083295e-05,
      "loss": 0.516,
      "mean_token_accuracy": 0.8359979130327702,
      "num_tokens": 817853039.0,
      "step": 25626
    },
    {
      "entropy": 0.5490161459892988,
      "epoch": 2.3578127785349134,
      "grad_norm": 0.163652703166008,
      "learning_rate": 2.1409513294691325e-05,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8304113484919071,
      "num_tokens": 817884450.0,
      "step": 25627
    },
    {
      "entropy": 0.6573353987187147,
      "epoch": 2.3579047849063546,
      "grad_norm": 0.172237828373909,
      "learning_rate": 2.14064464685497e-05,
      "loss": 0.6494,
      "mean_token_accuracy": 0.7953173629939556,
      "num_tokens": 817916576.0,
      "step": 25628
    },
    {
      "entropy": 0.5610318090766668,
      "epoch": 2.357996791277796,
      "grad_norm": 0.16545654833316803,
      "learning_rate": 2.140337964240807e-05,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8292854055762291,
      "num_tokens": 817948513.0,
      "step": 25629
    },
    {
      "entropy": 0.5298922244692221,
      "epoch": 2.3580887976492373,
      "grad_norm": 0.16303971409797668,
      "learning_rate": 2.1400312816266446e-05,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8384078964591026,
      "num_tokens": 817980596.0,
      "step": 25630
    },
    {
      "entropy": 0.460639710072428,
      "epoch": 2.3581808040206784,
      "grad_norm": 0.15715622901916504,
      "learning_rate": 2.139724599012482e-05,
      "loss": 0.446,
      "mean_token_accuracy": 0.8567503541707993,
      "num_tokens": 818012772.0,
      "step": 25631
    },
    {
      "entropy": 0.401648523285985,
      "epoch": 2.3582728103921196,
      "grad_norm": 0.142316073179245,
      "learning_rate": 2.1394179163983195e-05,
      "loss": 0.3913,
      "mean_token_accuracy": 0.8743421919643879,
      "num_tokens": 818044612.0,
      "step": 25632
    },
    {
      "entropy": 0.5347359925508499,
      "epoch": 2.3583648167635607,
      "grad_norm": 0.15815231204032898,
      "learning_rate": 2.1391112337841566e-05,
      "loss": 0.5181,
      "mean_token_accuracy": 0.835111066699028,
      "num_tokens": 818076602.0,
      "step": 25633
    },
    {
      "entropy": 0.5849387049674988,
      "epoch": 2.3584568231350023,
      "grad_norm": 0.16727787256240845,
      "learning_rate": 2.138804551169994e-05,
      "loss": 0.5641,
      "mean_token_accuracy": 0.8232447691261768,
      "num_tokens": 818109265.0,
      "step": 25634
    },
    {
      "entropy": 0.5144660184159875,
      "epoch": 2.3585488295064434,
      "grad_norm": 0.1582108736038208,
      "learning_rate": 2.1384978685558316e-05,
      "loss": 0.4994,
      "mean_token_accuracy": 0.8391495235264301,
      "num_tokens": 818141230.0,
      "step": 25635
    },
    {
      "entropy": 0.5226442022249103,
      "epoch": 2.3586408358778845,
      "grad_norm": 0.16378837823867798,
      "learning_rate": 2.138191185941669e-05,
      "loss": 0.5119,
      "mean_token_accuracy": 0.8377886787056923,
      "num_tokens": 818173588.0,
      "step": 25636
    },
    {
      "entropy": 0.5374631192535162,
      "epoch": 2.3587328422493257,
      "grad_norm": 0.1631178855895996,
      "learning_rate": 2.1378845033275065e-05,
      "loss": 0.5347,
      "mean_token_accuracy": 0.8322190791368484,
      "num_tokens": 818205575.0,
      "step": 25637
    },
    {
      "entropy": 0.5249018846079707,
      "epoch": 2.358824848620767,
      "grad_norm": 0.16055016219615936,
      "learning_rate": 2.1375778207133436e-05,
      "loss": 0.5255,
      "mean_token_accuracy": 0.8373996093869209,
      "num_tokens": 818238190.0,
      "step": 25638
    },
    {
      "entropy": 0.5441759722307324,
      "epoch": 2.3589168549922084,
      "grad_norm": 0.17094209790229797,
      "learning_rate": 2.1372711380991814e-05,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8314224034547806,
      "num_tokens": 818269098.0,
      "step": 25639
    },
    {
      "entropy": 0.4835845325142145,
      "epoch": 2.3590088613636495,
      "grad_norm": 0.15487952530384064,
      "learning_rate": 2.136964455485019e-05,
      "loss": 0.4822,
      "mean_token_accuracy": 0.844808503985405,
      "num_tokens": 818301795.0,
      "step": 25640
    },
    {
      "entropy": 0.5010161623358727,
      "epoch": 2.3591008677350906,
      "grad_norm": 0.16197998821735382,
      "learning_rate": 2.136657772870856e-05,
      "loss": 0.4946,
      "mean_token_accuracy": 0.8422890529036522,
      "num_tokens": 818333782.0,
      "step": 25641
    },
    {
      "entropy": 0.5555134043097496,
      "epoch": 2.359192874106532,
      "grad_norm": 0.1638532429933548,
      "learning_rate": 2.1363510902566935e-05,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8296592645347118,
      "num_tokens": 818366550.0,
      "step": 25642
    },
    {
      "entropy": 0.6334232771769166,
      "epoch": 2.359284880477973,
      "grad_norm": 0.16914691030979156,
      "learning_rate": 2.136044407642531e-05,
      "loss": 0.611,
      "mean_token_accuracy": 0.8073679208755493,
      "num_tokens": 818398237.0,
      "step": 25643
    },
    {
      "entropy": 0.5169970709830523,
      "epoch": 2.3593768868494145,
      "grad_norm": 0.16125227510929108,
      "learning_rate": 2.1357377250283684e-05,
      "loss": 0.5047,
      "mean_token_accuracy": 0.8393509462475777,
      "num_tokens": 818430331.0,
      "step": 25644
    },
    {
      "entropy": 0.44453758536837995,
      "epoch": 2.3594688932208556,
      "grad_norm": 0.15885335206985474,
      "learning_rate": 2.135431042414206e-05,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8641091249883175,
      "num_tokens": 818462205.0,
      "step": 25645
    },
    {
      "entropy": 0.5509883603081107,
      "epoch": 2.3595608995922968,
      "grad_norm": 0.1672956496477127,
      "learning_rate": 2.135124359800043e-05,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8308926783502102,
      "num_tokens": 818493631.0,
      "step": 25646
    },
    {
      "entropy": 0.5632276467513293,
      "epoch": 2.359652905963738,
      "grad_norm": 0.1640898883342743,
      "learning_rate": 2.1348176771858805e-05,
      "loss": 0.5545,
      "mean_token_accuracy": 0.8273324333131313,
      "num_tokens": 818526280.0,
      "step": 25647
    },
    {
      "entropy": 0.5299253519624472,
      "epoch": 2.359744912335179,
      "grad_norm": 0.16112419962882996,
      "learning_rate": 2.134510994571718e-05,
      "loss": 0.5144,
      "mean_token_accuracy": 0.836261685937643,
      "num_tokens": 818558421.0,
      "step": 25648
    },
    {
      "entropy": 0.46059264382347465,
      "epoch": 2.3598369187066206,
      "grad_norm": 0.15030542016029358,
      "learning_rate": 2.1342043119575554e-05,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8574669621884823,
      "num_tokens": 818590597.0,
      "step": 25649
    },
    {
      "entropy": 0.4746459466405213,
      "epoch": 2.3599289250780617,
      "grad_norm": 0.15569260716438293,
      "learning_rate": 2.1338976293433925e-05,
      "loss": 0.4564,
      "mean_token_accuracy": 0.8553053475916386,
      "num_tokens": 818622011.0,
      "step": 25650
    },
    {
      "entropy": 0.6289124339818954,
      "epoch": 2.360020931449503,
      "grad_norm": 0.17097528278827667,
      "learning_rate": 2.13359094672923e-05,
      "loss": 0.6182,
      "mean_token_accuracy": 0.8076122365891933,
      "num_tokens": 818653828.0,
      "step": 25651
    },
    {
      "entropy": 0.5224725855514407,
      "epoch": 2.360112937820944,
      "grad_norm": 0.16148005425930023,
      "learning_rate": 2.1332842641150675e-05,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8363118916749954,
      "num_tokens": 818685098.0,
      "step": 25652
    },
    {
      "entropy": 0.533381421584636,
      "epoch": 2.360204944192385,
      "grad_norm": 0.1621686965227127,
      "learning_rate": 2.132977581500905e-05,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8366015180945396,
      "num_tokens": 818716788.0,
      "step": 25653
    },
    {
      "entropy": 0.5374051239341497,
      "epoch": 2.3602969505638267,
      "grad_norm": 0.16353850066661835,
      "learning_rate": 2.132670898886742e-05,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8333742171525955,
      "num_tokens": 818748376.0,
      "step": 25654
    },
    {
      "entropy": 0.46046037413179874,
      "epoch": 2.360388956935268,
      "grad_norm": 0.1600683331489563,
      "learning_rate": 2.1323642162725795e-05,
      "loss": 0.4629,
      "mean_token_accuracy": 0.853199802339077,
      "num_tokens": 818781144.0,
      "step": 25655
    },
    {
      "entropy": 0.5612894715741277,
      "epoch": 2.360480963306709,
      "grad_norm": 0.16403280198574066,
      "learning_rate": 2.132057533658417e-05,
      "loss": 0.5589,
      "mean_token_accuracy": 0.8253719843924046,
      "num_tokens": 818813276.0,
      "step": 25656
    },
    {
      "entropy": 0.46353160217404366,
      "epoch": 2.36057296967815,
      "grad_norm": 0.16003288328647614,
      "learning_rate": 2.1317508510442544e-05,
      "loss": 0.4592,
      "mean_token_accuracy": 0.8582835346460342,
      "num_tokens": 818844326.0,
      "step": 25657
    },
    {
      "entropy": 0.42991546797566116,
      "epoch": 2.3606649760495912,
      "grad_norm": 0.14212660491466522,
      "learning_rate": 2.131444168430092e-05,
      "loss": 0.4166,
      "mean_token_accuracy": 0.8707194477319717,
      "num_tokens": 818877008.0,
      "step": 25658
    },
    {
      "entropy": 0.5596249103546143,
      "epoch": 2.360756982421033,
      "grad_norm": 0.16710585355758667,
      "learning_rate": 2.131137485815929e-05,
      "loss": 0.5388,
      "mean_token_accuracy": 0.8286410495638847,
      "num_tokens": 818908099.0,
      "step": 25659
    },
    {
      "entropy": 0.44677979312837124,
      "epoch": 2.360848988792474,
      "grad_norm": 0.1460222452878952,
      "learning_rate": 2.1308308032017665e-05,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8643835596740246,
      "num_tokens": 818940044.0,
      "step": 25660
    },
    {
      "entropy": 0.4947759071364999,
      "epoch": 2.360940995163915,
      "grad_norm": 0.15829357504844666,
      "learning_rate": 2.130524120587604e-05,
      "loss": 0.475,
      "mean_token_accuracy": 0.8470585867762566,
      "num_tokens": 818970547.0,
      "step": 25661
    },
    {
      "entropy": 0.5952994700055569,
      "epoch": 2.3610330015353562,
      "grad_norm": 0.1643376499414444,
      "learning_rate": 2.1302174379734414e-05,
      "loss": 0.583,
      "mean_token_accuracy": 0.8134019114077091,
      "num_tokens": 819002541.0,
      "step": 25662
    },
    {
      "entropy": 0.5006870506331325,
      "epoch": 2.3611250079067974,
      "grad_norm": 0.16173996031284332,
      "learning_rate": 2.1299107553592786e-05,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8452001288533211,
      "num_tokens": 819034298.0,
      "step": 25663
    },
    {
      "entropy": 0.5105008780956268,
      "epoch": 2.361217014278239,
      "grad_norm": 0.16085775196552277,
      "learning_rate": 2.129604072745116e-05,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8398134522140026,
      "num_tokens": 819066283.0,
      "step": 25664
    },
    {
      "entropy": 0.4614974050782621,
      "epoch": 2.36130902064968,
      "grad_norm": 0.14666885137557983,
      "learning_rate": 2.1292973901309535e-05,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8545823283493519,
      "num_tokens": 819098889.0,
      "step": 25665
    },
    {
      "entropy": 0.42232658877037466,
      "epoch": 2.361401027021121,
      "grad_norm": 0.14723412692546844,
      "learning_rate": 2.128990707516791e-05,
      "loss": 0.4066,
      "mean_token_accuracy": 0.8671836778521538,
      "num_tokens": 819131657.0,
      "step": 25666
    },
    {
      "entropy": 0.4784840438514948,
      "epoch": 2.3614930333925623,
      "grad_norm": 0.1547611504793167,
      "learning_rate": 2.1286840249026284e-05,
      "loss": 0.4493,
      "mean_token_accuracy": 0.8534877337515354,
      "num_tokens": 819163054.0,
      "step": 25667
    },
    {
      "entropy": 0.49005863768979907,
      "epoch": 2.3615850397640035,
      "grad_norm": 0.1549265831708908,
      "learning_rate": 2.1283773422884655e-05,
      "loss": 0.466,
      "mean_token_accuracy": 0.8526225723326206,
      "num_tokens": 819195313.0,
      "step": 25668
    },
    {
      "entropy": 0.5051598283462226,
      "epoch": 2.361677046135445,
      "grad_norm": 0.16190890967845917,
      "learning_rate": 2.1280706596743033e-05,
      "loss": 0.488,
      "mean_token_accuracy": 0.8460114970803261,
      "num_tokens": 819227554.0,
      "step": 25669
    },
    {
      "entropy": 0.638963496312499,
      "epoch": 2.361769052506886,
      "grad_norm": 0.17847022414207458,
      "learning_rate": 2.1277639770601408e-05,
      "loss": 0.6368,
      "mean_token_accuracy": 0.7999412454664707,
      "num_tokens": 819259394.0,
      "step": 25670
    },
    {
      "entropy": 0.5083031887188554,
      "epoch": 2.3618610588783273,
      "grad_norm": 0.15887418389320374,
      "learning_rate": 2.127457294445978e-05,
      "loss": 0.4944,
      "mean_token_accuracy": 0.841127060353756,
      "num_tokens": 819290809.0,
      "step": 25671
    },
    {
      "entropy": 0.5590608641505241,
      "epoch": 2.3619530652497684,
      "grad_norm": 0.1602736860513687,
      "learning_rate": 2.1271506118318154e-05,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8290139250457287,
      "num_tokens": 819323285.0,
      "step": 25672
    },
    {
      "entropy": 0.5235464442521334,
      "epoch": 2.3620450716212096,
      "grad_norm": 0.15893037617206573,
      "learning_rate": 2.126843929217653e-05,
      "loss": 0.5072,
      "mean_token_accuracy": 0.8405635990202427,
      "num_tokens": 819355759.0,
      "step": 25673
    },
    {
      "entropy": 0.6533501651138067,
      "epoch": 2.362137077992651,
      "grad_norm": 0.17175954580307007,
      "learning_rate": 2.1265372466034903e-05,
      "loss": 0.6301,
      "mean_token_accuracy": 0.8030939064919949,
      "num_tokens": 819387577.0,
      "step": 25674
    },
    {
      "entropy": 0.5248013073578477,
      "epoch": 2.3622290843640923,
      "grad_norm": 0.16137653589248657,
      "learning_rate": 2.1262305639893278e-05,
      "loss": 0.5097,
      "mean_token_accuracy": 0.8388189747929573,
      "num_tokens": 819419267.0,
      "step": 25675
    },
    {
      "entropy": 0.49896268686279655,
      "epoch": 2.3623210907355334,
      "grad_norm": 0.1542089730501175,
      "learning_rate": 2.125923881375165e-05,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8495366387069225,
      "num_tokens": 819451457.0,
      "step": 25676
    },
    {
      "entropy": 0.624297708272934,
      "epoch": 2.3624130971069746,
      "grad_norm": 0.16729743778705597,
      "learning_rate": 2.1256171987610024e-05,
      "loss": 0.619,
      "mean_token_accuracy": 0.8054996095597744,
      "num_tokens": 819483776.0,
      "step": 25677
    },
    {
      "entropy": 0.5003483658656478,
      "epoch": 2.3625051034784157,
      "grad_norm": 0.15463362634181976,
      "learning_rate": 2.12531051614684e-05,
      "loss": 0.4951,
      "mean_token_accuracy": 0.8446212187409401,
      "num_tokens": 819515914.0,
      "step": 25678
    },
    {
      "entropy": 0.5193071626126766,
      "epoch": 2.3625971098498573,
      "grad_norm": 0.1633939892053604,
      "learning_rate": 2.1250038335326773e-05,
      "loss": 0.5197,
      "mean_token_accuracy": 0.8377560786902905,
      "num_tokens": 819547817.0,
      "step": 25679
    },
    {
      "entropy": 0.5015627518296242,
      "epoch": 2.3626891162212984,
      "grad_norm": 0.1577436327934265,
      "learning_rate": 2.1246971509185144e-05,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8430968709290028,
      "num_tokens": 819580289.0,
      "step": 25680
    },
    {
      "entropy": 0.46181901544332504,
      "epoch": 2.3627811225927395,
      "grad_norm": 0.15437297523021698,
      "learning_rate": 2.124390468304352e-05,
      "loss": 0.4528,
      "mean_token_accuracy": 0.855523195117712,
      "num_tokens": 819612715.0,
      "step": 25681
    },
    {
      "entropy": 0.57676991680637,
      "epoch": 2.3628731289641807,
      "grad_norm": 0.17351803183555603,
      "learning_rate": 2.1240837856901894e-05,
      "loss": 0.5764,
      "mean_token_accuracy": 0.8182020671665668,
      "num_tokens": 819644264.0,
      "step": 25682
    },
    {
      "entropy": 0.621406284160912,
      "epoch": 2.362965135335622,
      "grad_norm": 0.18057160079479218,
      "learning_rate": 2.123777103076027e-05,
      "loss": 0.6194,
      "mean_token_accuracy": 0.8055323250591755,
      "num_tokens": 819675640.0,
      "step": 25683
    },
    {
      "entropy": 0.4766353666782379,
      "epoch": 2.3630571417070634,
      "grad_norm": 0.15681780874729156,
      "learning_rate": 2.123470420461864e-05,
      "loss": 0.465,
      "mean_token_accuracy": 0.8488666154444218,
      "num_tokens": 819707710.0,
      "step": 25684
    },
    {
      "entropy": 0.5231007458642125,
      "epoch": 2.3631491480785045,
      "grad_norm": 0.16033191978931427,
      "learning_rate": 2.1231637378477014e-05,
      "loss": 0.5042,
      "mean_token_accuracy": 0.8387314714491367,
      "num_tokens": 819739524.0,
      "step": 25685
    },
    {
      "entropy": 0.5677141463384032,
      "epoch": 2.3632411544499456,
      "grad_norm": 0.16525471210479736,
      "learning_rate": 2.122857055233539e-05,
      "loss": 0.5568,
      "mean_token_accuracy": 0.8246969468891621,
      "num_tokens": 819771317.0,
      "step": 25686
    },
    {
      "entropy": 0.48540474474430084,
      "epoch": 2.363333160821387,
      "grad_norm": 0.1610558032989502,
      "learning_rate": 2.1225503726193764e-05,
      "loss": 0.4695,
      "mean_token_accuracy": 0.849683340638876,
      "num_tokens": 819802320.0,
      "step": 25687
    },
    {
      "entropy": 0.46347806276753545,
      "epoch": 2.363425167192828,
      "grad_norm": 0.1558840125799179,
      "learning_rate": 2.1222436900052138e-05,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8561744876205921,
      "num_tokens": 819834090.0,
      "step": 25688
    },
    {
      "entropy": 0.4966772710904479,
      "epoch": 2.3635171735642695,
      "grad_norm": 0.15858572721481323,
      "learning_rate": 2.121937007391051e-05,
      "loss": 0.4776,
      "mean_token_accuracy": 0.846134789288044,
      "num_tokens": 819865492.0,
      "step": 25689
    },
    {
      "entropy": 0.518487511202693,
      "epoch": 2.3636091799357106,
      "grad_norm": 0.16446553170681,
      "learning_rate": 2.1216303247768884e-05,
      "loss": 0.5117,
      "mean_token_accuracy": 0.8384512104094028,
      "num_tokens": 819897260.0,
      "step": 25690
    },
    {
      "entropy": 0.5513265877962112,
      "epoch": 2.3637011863071518,
      "grad_norm": 0.1671285629272461,
      "learning_rate": 2.121323642162726e-05,
      "loss": 0.5322,
      "mean_token_accuracy": 0.8311706222593784,
      "num_tokens": 819929235.0,
      "step": 25691
    },
    {
      "entropy": 0.4713011793792248,
      "epoch": 2.363793192678593,
      "grad_norm": 0.16054388880729675,
      "learning_rate": 2.1210169595485633e-05,
      "loss": 0.46,
      "mean_token_accuracy": 0.853533186018467,
      "num_tokens": 819961103.0,
      "step": 25692
    },
    {
      "entropy": 0.5084033980965614,
      "epoch": 2.363885199050034,
      "grad_norm": 0.15687546133995056,
      "learning_rate": 2.1207102769344005e-05,
      "loss": 0.496,
      "mean_token_accuracy": 0.8401130512356758,
      "num_tokens": 819993610.0,
      "step": 25693
    },
    {
      "entropy": 0.45735959336161613,
      "epoch": 2.3639772054214756,
      "grad_norm": 0.15858663618564606,
      "learning_rate": 2.120403594320238e-05,
      "loss": 0.442,
      "mean_token_accuracy": 0.855887345969677,
      "num_tokens": 820025509.0,
      "step": 25694
    },
    {
      "entropy": 0.519312085583806,
      "epoch": 2.3640692117929167,
      "grad_norm": 0.16093553602695465,
      "learning_rate": 2.1200969117060754e-05,
      "loss": 0.5113,
      "mean_token_accuracy": 0.8392850756645203,
      "num_tokens": 820057552.0,
      "step": 25695
    },
    {
      "entropy": 0.5251762792468071,
      "epoch": 2.364161218164358,
      "grad_norm": 0.1660960614681244,
      "learning_rate": 2.119790229091913e-05,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8390705399215221,
      "num_tokens": 820089332.0,
      "step": 25696
    },
    {
      "entropy": 0.4591878242790699,
      "epoch": 2.364253224535799,
      "grad_norm": 0.15737462043762207,
      "learning_rate": 2.1194835464777503e-05,
      "loss": 0.439,
      "mean_token_accuracy": 0.8595554605126381,
      "num_tokens": 820120902.0,
      "step": 25697
    },
    {
      "entropy": 0.5101629449054599,
      "epoch": 2.36434523090724,
      "grad_norm": 0.15462233126163483,
      "learning_rate": 2.1191768638635874e-05,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8450489826500416,
      "num_tokens": 820153603.0,
      "step": 25698
    },
    {
      "entropy": 0.3972784634679556,
      "epoch": 2.3644372372786817,
      "grad_norm": 0.1425456404685974,
      "learning_rate": 2.118870181249425e-05,
      "loss": 0.3785,
      "mean_token_accuracy": 0.8772687613964081,
      "num_tokens": 820186144.0,
      "step": 25699
    },
    {
      "entropy": 0.3983548451215029,
      "epoch": 2.364529243650123,
      "grad_norm": 0.14440011978149414,
      "learning_rate": 2.1185634986352627e-05,
      "loss": 0.3839,
      "mean_token_accuracy": 0.8776392042636871,
      "num_tokens": 820218145.0,
      "step": 25700
    },
    {
      "entropy": 0.3993650730699301,
      "epoch": 2.364621250021564,
      "grad_norm": 0.14652124047279358,
      "learning_rate": 2.1182568160211e-05,
      "loss": 0.3904,
      "mean_token_accuracy": 0.8724453300237656,
      "num_tokens": 820249809.0,
      "step": 25701
    },
    {
      "entropy": 0.6127847479656339,
      "epoch": 2.364713256393005,
      "grad_norm": 0.1674865186214447,
      "learning_rate": 2.1179501334069373e-05,
      "loss": 0.5941,
      "mean_token_accuracy": 0.8152335546910763,
      "num_tokens": 820281994.0,
      "step": 25702
    },
    {
      "entropy": 0.5431090565398335,
      "epoch": 2.3648052627644462,
      "grad_norm": 0.16304847598075867,
      "learning_rate": 2.1176434507927748e-05,
      "loss": 0.516,
      "mean_token_accuracy": 0.8318220749497414,
      "num_tokens": 820313870.0,
      "step": 25703
    },
    {
      "entropy": 0.48913456685841084,
      "epoch": 2.364897269135888,
      "grad_norm": 0.15445482730865479,
      "learning_rate": 2.1173367681786122e-05,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8512603640556335,
      "num_tokens": 820345824.0,
      "step": 25704
    },
    {
      "entropy": 0.5507302349433303,
      "epoch": 2.364989275507329,
      "grad_norm": 0.15819138288497925,
      "learning_rate": 2.1170300855644497e-05,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8305819109082222,
      "num_tokens": 820378538.0,
      "step": 25705
    },
    {
      "entropy": 0.5327284382656217,
      "epoch": 2.36508128187877,
      "grad_norm": 0.1635805368423462,
      "learning_rate": 2.1167234029502868e-05,
      "loss": 0.5195,
      "mean_token_accuracy": 0.834996659308672,
      "num_tokens": 820410404.0,
      "step": 25706
    },
    {
      "entropy": 0.45610903261695057,
      "epoch": 2.3651732882502112,
      "grad_norm": 0.15121060609817505,
      "learning_rate": 2.1164167203361243e-05,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8549386449158192,
      "num_tokens": 820442157.0,
      "step": 25707
    },
    {
      "entropy": 0.5540971765294671,
      "epoch": 2.3652652946216524,
      "grad_norm": 0.17320837080478668,
      "learning_rate": 2.1161100377219618e-05,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8223690651357174,
      "num_tokens": 820473672.0,
      "step": 25708
    },
    {
      "entropy": 0.5366207342594862,
      "epoch": 2.365357300993094,
      "grad_norm": 0.164528027176857,
      "learning_rate": 2.1158033551077992e-05,
      "loss": 0.5137,
      "mean_token_accuracy": 0.8383371941745281,
      "num_tokens": 820505173.0,
      "step": 25709
    },
    {
      "entropy": 0.5115102855488658,
      "epoch": 2.365449307364535,
      "grad_norm": 0.1559453010559082,
      "learning_rate": 2.1154966724936363e-05,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8370702043175697,
      "num_tokens": 820536849.0,
      "step": 25710
    },
    {
      "entropy": 0.5211611315608025,
      "epoch": 2.365541313735976,
      "grad_norm": 0.16512273252010345,
      "learning_rate": 2.1151899898794738e-05,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8386046960949898,
      "num_tokens": 820568476.0,
      "step": 25711
    },
    {
      "entropy": 0.46911942306905985,
      "epoch": 2.3656333201074173,
      "grad_norm": 0.15767036378383636,
      "learning_rate": 2.1148833072653113e-05,
      "loss": 0.4473,
      "mean_token_accuracy": 0.8553788475692272,
      "num_tokens": 820600692.0,
      "step": 25712
    },
    {
      "entropy": 0.5009477629791945,
      "epoch": 2.3657253264788585,
      "grad_norm": 0.15735265612602234,
      "learning_rate": 2.1145766246511487e-05,
      "loss": 0.4914,
      "mean_token_accuracy": 0.8470648750662804,
      "num_tokens": 820633254.0,
      "step": 25713
    },
    {
      "entropy": 0.5646863337606192,
      "epoch": 2.3658173328503,
      "grad_norm": 0.16394087672233582,
      "learning_rate": 2.114269942036986e-05,
      "loss": 0.545,
      "mean_token_accuracy": 0.8259791918098927,
      "num_tokens": 820665668.0,
      "step": 25714
    },
    {
      "entropy": 0.4941774997860193,
      "epoch": 2.365909339221741,
      "grad_norm": 0.16051506996154785,
      "learning_rate": 2.1139632594228233e-05,
      "loss": 0.4819,
      "mean_token_accuracy": 0.8436015024781227,
      "num_tokens": 820697902.0,
      "step": 25715
    },
    {
      "entropy": 0.5625868830829859,
      "epoch": 2.3660013455931823,
      "grad_norm": 0.1665610522031784,
      "learning_rate": 2.1136565768086608e-05,
      "loss": 0.5583,
      "mean_token_accuracy": 0.8265301026403904,
      "num_tokens": 820729987.0,
      "step": 25716
    },
    {
      "entropy": 0.5923120174556971,
      "epoch": 2.3660933519646234,
      "grad_norm": 0.17234516143798828,
      "learning_rate": 2.1133498941944983e-05,
      "loss": 0.5906,
      "mean_token_accuracy": 0.8129290044307709,
      "num_tokens": 820761655.0,
      "step": 25717
    },
    {
      "entropy": 0.4919923224952072,
      "epoch": 2.3661853583360646,
      "grad_norm": 0.15724758803844452,
      "learning_rate": 2.1130432115803357e-05,
      "loss": 0.4829,
      "mean_token_accuracy": 0.8461174182593822,
      "num_tokens": 820794258.0,
      "step": 25718
    },
    {
      "entropy": 0.5572485327720642,
      "epoch": 2.366277364707506,
      "grad_norm": 0.1654139906167984,
      "learning_rate": 2.112736528966173e-05,
      "loss": 0.5552,
      "mean_token_accuracy": 0.8240667097270489,
      "num_tokens": 820826792.0,
      "step": 25719
    },
    {
      "entropy": 0.5475485995411873,
      "epoch": 2.3663693710789473,
      "grad_norm": 0.15797550976276398,
      "learning_rate": 2.1124298463520103e-05,
      "loss": 0.5386,
      "mean_token_accuracy": 0.8271421790122986,
      "num_tokens": 820859187.0,
      "step": 25720
    },
    {
      "entropy": 0.5447046568151563,
      "epoch": 2.3664613774503884,
      "grad_norm": 0.16315098106861115,
      "learning_rate": 2.1121231637378478e-05,
      "loss": 0.5327,
      "mean_token_accuracy": 0.8312439173460007,
      "num_tokens": 820891627.0,
      "step": 25721
    },
    {
      "entropy": 0.5018086759373546,
      "epoch": 2.3665533838218296,
      "grad_norm": 0.1548866182565689,
      "learning_rate": 2.1118164811236852e-05,
      "loss": 0.4881,
      "mean_token_accuracy": 0.8459268286824226,
      "num_tokens": 820923931.0,
      "step": 25722
    },
    {
      "entropy": 0.46130844112485647,
      "epoch": 2.3666453901932707,
      "grad_norm": 0.15595075488090515,
      "learning_rate": 2.1115097985095224e-05,
      "loss": 0.45,
      "mean_token_accuracy": 0.8600748255848885,
      "num_tokens": 820955883.0,
      "step": 25723
    },
    {
      "entropy": 0.5311868879944086,
      "epoch": 2.3667373965647123,
      "grad_norm": 0.16379398107528687,
      "learning_rate": 2.11120311589536e-05,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8339813239872456,
      "num_tokens": 820988180.0,
      "step": 25724
    },
    {
      "entropy": 0.7057678773999214,
      "epoch": 2.3668294029361534,
      "grad_norm": 0.17664511501789093,
      "learning_rate": 2.1108964332811973e-05,
      "loss": 0.6934,
      "mean_token_accuracy": 0.7831022776663303,
      "num_tokens": 821020137.0,
      "step": 25725
    },
    {
      "entropy": 0.49022599845193326,
      "epoch": 2.3669214093075945,
      "grad_norm": 0.15472368896007538,
      "learning_rate": 2.1105897506670348e-05,
      "loss": 0.4786,
      "mean_token_accuracy": 0.8507877290248871,
      "num_tokens": 821052905.0,
      "step": 25726
    },
    {
      "entropy": 0.632229994982481,
      "epoch": 2.3670134156790357,
      "grad_norm": 0.1712760180234909,
      "learning_rate": 2.1102830680528722e-05,
      "loss": 0.614,
      "mean_token_accuracy": 0.8100426942110062,
      "num_tokens": 821084823.0,
      "step": 25727
    },
    {
      "entropy": 0.43960240576416254,
      "epoch": 2.367105422050477,
      "grad_norm": 0.1508485972881317,
      "learning_rate": 2.1099763854387094e-05,
      "loss": 0.4228,
      "mean_token_accuracy": 0.8635807037353516,
      "num_tokens": 821117295.0,
      "step": 25728
    },
    {
      "entropy": 0.5663454092573375,
      "epoch": 2.3671974284219184,
      "grad_norm": 0.16131986677646637,
      "learning_rate": 2.1096697028245468e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8291001580655575,
      "num_tokens": 821150035.0,
      "step": 25729
    },
    {
      "entropy": 0.4654030150268227,
      "epoch": 2.3672894347933595,
      "grad_norm": 0.15320336818695068,
      "learning_rate": 2.1093630202103846e-05,
      "loss": 0.4608,
      "mean_token_accuracy": 0.85289566218853,
      "num_tokens": 821182386.0,
      "step": 25730
    },
    {
      "entropy": 0.6025048941373825,
      "epoch": 2.3673814411648006,
      "grad_norm": 0.16826917231082916,
      "learning_rate": 2.1090563375962218e-05,
      "loss": 0.581,
      "mean_token_accuracy": 0.8149486295878887,
      "num_tokens": 821214062.0,
      "step": 25731
    },
    {
      "entropy": 0.44275127351284027,
      "epoch": 2.367473447536242,
      "grad_norm": 0.14690594375133514,
      "learning_rate": 2.1087496549820592e-05,
      "loss": 0.427,
      "mean_token_accuracy": 0.8633382879197598,
      "num_tokens": 821246354.0,
      "step": 25732
    },
    {
      "entropy": 0.4502996081719175,
      "epoch": 2.367565453907683,
      "grad_norm": 0.1591474860906601,
      "learning_rate": 2.1084429723678967e-05,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8599514588713646,
      "num_tokens": 821278732.0,
      "step": 25733
    },
    {
      "entropy": 0.5489085335284472,
      "epoch": 2.3676574602791245,
      "grad_norm": 0.16640715301036835,
      "learning_rate": 2.108136289753734e-05,
      "loss": 0.5551,
      "mean_token_accuracy": 0.8282125815749168,
      "num_tokens": 821310264.0,
      "step": 25734
    },
    {
      "entropy": 0.4252933629322797,
      "epoch": 2.3677494666505656,
      "grad_norm": 0.1503095179796219,
      "learning_rate": 2.1078296071395716e-05,
      "loss": 0.4061,
      "mean_token_accuracy": 0.871965903788805,
      "num_tokens": 821342381.0,
      "step": 25735
    },
    {
      "entropy": 0.4762898162007332,
      "epoch": 2.3678414730220068,
      "grad_norm": 0.15629935264587402,
      "learning_rate": 2.1075229245254087e-05,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8514681309461594,
      "num_tokens": 821373393.0,
      "step": 25736
    },
    {
      "entropy": 0.615411045961082,
      "epoch": 2.367933479393448,
      "grad_norm": 0.17545926570892334,
      "learning_rate": 2.1072162419112462e-05,
      "loss": 0.6062,
      "mean_token_accuracy": 0.8088439367711544,
      "num_tokens": 821404891.0,
      "step": 25737
    },
    {
      "entropy": 0.5738867959007621,
      "epoch": 2.368025485764889,
      "grad_norm": 0.1639767587184906,
      "learning_rate": 2.1069095592970837e-05,
      "loss": 0.5696,
      "mean_token_accuracy": 0.8225175738334656,
      "num_tokens": 821437561.0,
      "step": 25738
    },
    {
      "entropy": 0.5439132479950786,
      "epoch": 2.3681174921363306,
      "grad_norm": 0.16632649302482605,
      "learning_rate": 2.106602876682921e-05,
      "loss": 0.5418,
      "mean_token_accuracy": 0.8274118453264236,
      "num_tokens": 821469566.0,
      "step": 25739
    },
    {
      "entropy": 0.59254165366292,
      "epoch": 2.3682094985077717,
      "grad_norm": 0.17250853776931763,
      "learning_rate": 2.1062961940687583e-05,
      "loss": 0.5854,
      "mean_token_accuracy": 0.8174435868859291,
      "num_tokens": 821500709.0,
      "step": 25740
    },
    {
      "entropy": 0.6052577868103981,
      "epoch": 2.368301504879213,
      "grad_norm": 0.16715696454048157,
      "learning_rate": 2.1059895114545957e-05,
      "loss": 0.5936,
      "mean_token_accuracy": 0.8149029575288296,
      "num_tokens": 821532354.0,
      "step": 25741
    },
    {
      "entropy": 0.4406666802242398,
      "epoch": 2.368393511250654,
      "grad_norm": 0.1513926088809967,
      "learning_rate": 2.1056828288404332e-05,
      "loss": 0.4289,
      "mean_token_accuracy": 0.863419346511364,
      "num_tokens": 821564470.0,
      "step": 25742
    },
    {
      "entropy": 0.5535288397222757,
      "epoch": 2.368485517622095,
      "grad_norm": 0.16873767971992493,
      "learning_rate": 2.1053761462262707e-05,
      "loss": 0.5486,
      "mean_token_accuracy": 0.8300113826990128,
      "num_tokens": 821595767.0,
      "step": 25743
    },
    {
      "entropy": 0.4975919555872679,
      "epoch": 2.3685775239935367,
      "grad_norm": 0.15390226244926453,
      "learning_rate": 2.1050694636121078e-05,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8465742543339729,
      "num_tokens": 821628535.0,
      "step": 25744
    },
    {
      "entropy": 0.42307151947170496,
      "epoch": 2.368669530364978,
      "grad_norm": 0.14770573377609253,
      "learning_rate": 2.1047627809979452e-05,
      "loss": 0.4066,
      "mean_token_accuracy": 0.8709238357841969,
      "num_tokens": 821660577.0,
      "step": 25745
    },
    {
      "entropy": 0.5529237352311611,
      "epoch": 2.368761536736419,
      "grad_norm": 0.16631963849067688,
      "learning_rate": 2.1044560983837827e-05,
      "loss": 0.5381,
      "mean_token_accuracy": 0.8297855518758297,
      "num_tokens": 821692017.0,
      "step": 25746
    },
    {
      "entropy": 0.4851539172232151,
      "epoch": 2.36885354310786,
      "grad_norm": 0.15580929815769196,
      "learning_rate": 2.1041494157696202e-05,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8546341173350811,
      "num_tokens": 821724265.0,
      "step": 25747
    },
    {
      "entropy": 0.49888229090720415,
      "epoch": 2.3689455494793012,
      "grad_norm": 0.1576162576675415,
      "learning_rate": 2.1038427331554576e-05,
      "loss": 0.4761,
      "mean_token_accuracy": 0.8494114577770233,
      "num_tokens": 821756021.0,
      "step": 25748
    },
    {
      "entropy": 0.4295121030882001,
      "epoch": 2.369037555850743,
      "grad_norm": 0.1448814868927002,
      "learning_rate": 2.1035360505412948e-05,
      "loss": 0.4222,
      "mean_token_accuracy": 0.8654438890516758,
      "num_tokens": 821788333.0,
      "step": 25749
    },
    {
      "entropy": 0.5018582521006465,
      "epoch": 2.369129562222184,
      "grad_norm": 0.15734438598155975,
      "learning_rate": 2.1032293679271322e-05,
      "loss": 0.4929,
      "mean_token_accuracy": 0.8447303138673306,
      "num_tokens": 821820078.0,
      "step": 25750
    },
    {
      "entropy": 0.6123523069545627,
      "epoch": 2.369221568593625,
      "grad_norm": 0.17347833514213562,
      "learning_rate": 2.1029226853129697e-05,
      "loss": 0.6028,
      "mean_token_accuracy": 0.808652650564909,
      "num_tokens": 821851927.0,
      "step": 25751
    },
    {
      "entropy": 0.5102202082052827,
      "epoch": 2.3693135749650662,
      "grad_norm": 0.15544728934764862,
      "learning_rate": 2.102616002698807e-05,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8464146107435226,
      "num_tokens": 821884349.0,
      "step": 25752
    },
    {
      "entropy": 0.6474671652540565,
      "epoch": 2.3694055813365074,
      "grad_norm": 0.17152109742164612,
      "learning_rate": 2.1023093200846443e-05,
      "loss": 0.6275,
      "mean_token_accuracy": 0.807854488492012,
      "num_tokens": 821916161.0,
      "step": 25753
    },
    {
      "entropy": 0.5141332177445292,
      "epoch": 2.369497587707949,
      "grad_norm": 0.16125135123729706,
      "learning_rate": 2.1020026374704817e-05,
      "loss": 0.5013,
      "mean_token_accuracy": 0.8418369889259338,
      "num_tokens": 821948348.0,
      "step": 25754
    },
    {
      "entropy": 0.6355437748134136,
      "epoch": 2.36958959407939,
      "grad_norm": 0.1789812594652176,
      "learning_rate": 2.1016959548563192e-05,
      "loss": 0.6219,
      "mean_token_accuracy": 0.8049910068511963,
      "num_tokens": 821979734.0,
      "step": 25755
    },
    {
      "entropy": 0.5301446299999952,
      "epoch": 2.369681600450831,
      "grad_norm": 0.1577194184064865,
      "learning_rate": 2.1013892722421567e-05,
      "loss": 0.5127,
      "mean_token_accuracy": 0.8351097963750362,
      "num_tokens": 822011694.0,
      "step": 25756
    },
    {
      "entropy": 0.5065490459091961,
      "epoch": 2.3697736068222723,
      "grad_norm": 0.15539376437664032,
      "learning_rate": 2.101082589627994e-05,
      "loss": 0.4965,
      "mean_token_accuracy": 0.842429019510746,
      "num_tokens": 822044255.0,
      "step": 25757
    },
    {
      "entropy": 0.52814756706357,
      "epoch": 2.3698656131937135,
      "grad_norm": 0.16112078726291656,
      "learning_rate": 2.1007759070138313e-05,
      "loss": 0.514,
      "mean_token_accuracy": 0.8381861560046673,
      "num_tokens": 822076758.0,
      "step": 25758
    },
    {
      "entropy": 0.49259378761053085,
      "epoch": 2.369957619565155,
      "grad_norm": 0.15316623449325562,
      "learning_rate": 2.1004692243996687e-05,
      "loss": 0.4805,
      "mean_token_accuracy": 0.8473135493695736,
      "num_tokens": 822108404.0,
      "step": 25759
    },
    {
      "entropy": 0.6340024191886187,
      "epoch": 2.370049625936596,
      "grad_norm": 0.17773103713989258,
      "learning_rate": 2.1001625417855062e-05,
      "loss": 0.6207,
      "mean_token_accuracy": 0.8034694641828537,
      "num_tokens": 822139920.0,
      "step": 25760
    },
    {
      "entropy": 0.5585168699035421,
      "epoch": 2.3701416323080373,
      "grad_norm": 0.1671450287103653,
      "learning_rate": 2.0998558591713437e-05,
      "loss": 0.5432,
      "mean_token_accuracy": 0.8271000497043133,
      "num_tokens": 822171239.0,
      "step": 25761
    },
    {
      "entropy": 0.4308317070826888,
      "epoch": 2.3702336386794784,
      "grad_norm": 0.14764322340488434,
      "learning_rate": 2.099549176557181e-05,
      "loss": 0.4178,
      "mean_token_accuracy": 0.866690456867218,
      "num_tokens": 822203868.0,
      "step": 25762
    },
    {
      "entropy": 0.5015133824199438,
      "epoch": 2.3703256450509196,
      "grad_norm": 0.16467086970806122,
      "learning_rate": 2.0992424939430186e-05,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8456373326480389,
      "num_tokens": 822235435.0,
      "step": 25763
    },
    {
      "entropy": 0.46878791879862547,
      "epoch": 2.370417651422361,
      "grad_norm": 0.15808166563510895,
      "learning_rate": 2.098935811328856e-05,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8524835035204887,
      "num_tokens": 822267370.0,
      "step": 25764
    },
    {
      "entropy": 0.4563261987641454,
      "epoch": 2.3705096577938023,
      "grad_norm": 0.15550126135349274,
      "learning_rate": 2.0986291287146935e-05,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8568753451108932,
      "num_tokens": 822298522.0,
      "step": 25765
    },
    {
      "entropy": 0.4650128213688731,
      "epoch": 2.3706016641652434,
      "grad_norm": 0.1558605283498764,
      "learning_rate": 2.0983224461005306e-05,
      "loss": 0.4448,
      "mean_token_accuracy": 0.8556360937654972,
      "num_tokens": 822330526.0,
      "step": 25766
    },
    {
      "entropy": 0.48758196271955967,
      "epoch": 2.3706936705366846,
      "grad_norm": 0.1632997691631317,
      "learning_rate": 2.098015763486368e-05,
      "loss": 0.472,
      "mean_token_accuracy": 0.8490259796380997,
      "num_tokens": 822363014.0,
      "step": 25767
    },
    {
      "entropy": 0.49846268631517887,
      "epoch": 2.3707856769081257,
      "grad_norm": 0.15963609516620636,
      "learning_rate": 2.0977090808722056e-05,
      "loss": 0.4828,
      "mean_token_accuracy": 0.8436748273670673,
      "num_tokens": 822395047.0,
      "step": 25768
    },
    {
      "entropy": 0.524173797108233,
      "epoch": 2.3708776832795673,
      "grad_norm": 0.16565746068954468,
      "learning_rate": 2.097402398258043e-05,
      "loss": 0.5066,
      "mean_token_accuracy": 0.8362104259431362,
      "num_tokens": 822426556.0,
      "step": 25769
    },
    {
      "entropy": 0.4732971340417862,
      "epoch": 2.3709696896510084,
      "grad_norm": 0.15661479532718658,
      "learning_rate": 2.09709571564388e-05,
      "loss": 0.4645,
      "mean_token_accuracy": 0.8528907150030136,
      "num_tokens": 822458805.0,
      "step": 25770
    },
    {
      "entropy": 0.5379225313663483,
      "epoch": 2.3710616960224495,
      "grad_norm": 0.1667531579732895,
      "learning_rate": 2.0967890330297176e-05,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8310831859707832,
      "num_tokens": 822490365.0,
      "step": 25771
    },
    {
      "entropy": 0.5540529433637857,
      "epoch": 2.3711537023938907,
      "grad_norm": 0.16614598035812378,
      "learning_rate": 2.096482350415555e-05,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8259942829608917,
      "num_tokens": 822522504.0,
      "step": 25772
    },
    {
      "entropy": 0.5293927513994277,
      "epoch": 2.371245708765332,
      "grad_norm": 0.16049982607364655,
      "learning_rate": 2.0961756678013926e-05,
      "loss": 0.5247,
      "mean_token_accuracy": 0.8361855000257492,
      "num_tokens": 822555183.0,
      "step": 25773
    },
    {
      "entropy": 0.36347160884179175,
      "epoch": 2.3713377151367734,
      "grad_norm": 0.13626256585121155,
      "learning_rate": 2.0958689851872297e-05,
      "loss": 0.3496,
      "mean_token_accuracy": 0.8878738582134247,
      "num_tokens": 822587691.0,
      "step": 25774
    },
    {
      "entropy": 0.4831793177872896,
      "epoch": 2.3714297215082145,
      "grad_norm": 0.15644410252571106,
      "learning_rate": 2.095562302573067e-05,
      "loss": 0.4716,
      "mean_token_accuracy": 0.8474988453090191,
      "num_tokens": 822619775.0,
      "step": 25775
    },
    {
      "entropy": 0.464958727825433,
      "epoch": 2.3715217278796556,
      "grad_norm": 0.15172705054283142,
      "learning_rate": 2.0952556199589046e-05,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8537873961031437,
      "num_tokens": 822652295.0,
      "step": 25776
    },
    {
      "entropy": 0.5472142514772713,
      "epoch": 2.371613734251097,
      "grad_norm": 0.16517634689807892,
      "learning_rate": 2.094948937344742e-05,
      "loss": 0.5356,
      "mean_token_accuracy": 0.8326831869781017,
      "num_tokens": 822684382.0,
      "step": 25777
    },
    {
      "entropy": 0.5856922988314182,
      "epoch": 2.371705740622538,
      "grad_norm": 0.16783694922924042,
      "learning_rate": 2.0946422547305795e-05,
      "loss": 0.5658,
      "mean_token_accuracy": 0.8187438398599625,
      "num_tokens": 822716205.0,
      "step": 25778
    },
    {
      "entropy": 0.5471664359793067,
      "epoch": 2.3717977469939795,
      "grad_norm": 0.17399492859840393,
      "learning_rate": 2.0943355721164167e-05,
      "loss": 0.5437,
      "mean_token_accuracy": 0.8277795426547527,
      "num_tokens": 822747792.0,
      "step": 25779
    },
    {
      "entropy": 0.5556042492389679,
      "epoch": 2.3718897533654206,
      "grad_norm": 0.16423919796943665,
      "learning_rate": 2.094028889502254e-05,
      "loss": 0.5479,
      "mean_token_accuracy": 0.8266471214592457,
      "num_tokens": 822779970.0,
      "step": 25780
    },
    {
      "entropy": 0.49559195432811975,
      "epoch": 2.3719817597368618,
      "grad_norm": 0.16297782957553864,
      "learning_rate": 2.0937222068880916e-05,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8459451422095299,
      "num_tokens": 822811538.0,
      "step": 25781
    },
    {
      "entropy": 0.46217478066682816,
      "epoch": 2.372073766108303,
      "grad_norm": 0.151994988322258,
      "learning_rate": 2.093415524273929e-05,
      "loss": 0.4527,
      "mean_token_accuracy": 0.852303072810173,
      "num_tokens": 822843470.0,
      "step": 25782
    },
    {
      "entropy": 0.5444091716781259,
      "epoch": 2.372165772479744,
      "grad_norm": 0.16675707697868347,
      "learning_rate": 2.0931088416597662e-05,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8296445608139038,
      "num_tokens": 822875748.0,
      "step": 25783
    },
    {
      "entropy": 0.5583274615928531,
      "epoch": 2.3722577788511856,
      "grad_norm": 0.16313278675079346,
      "learning_rate": 2.0928021590456037e-05,
      "loss": 0.539,
      "mean_token_accuracy": 0.8289195895195007,
      "num_tokens": 822907782.0,
      "step": 25784
    },
    {
      "entropy": 0.514779285993427,
      "epoch": 2.3723497852226267,
      "grad_norm": 0.1597016602754593,
      "learning_rate": 2.092495476431441e-05,
      "loss": 0.5004,
      "mean_token_accuracy": 0.8421835787594318,
      "num_tokens": 822939760.0,
      "step": 25785
    },
    {
      "entropy": 0.6073905350640416,
      "epoch": 2.372441791594068,
      "grad_norm": 0.16992001235485077,
      "learning_rate": 2.0921887938172786e-05,
      "loss": 0.5959,
      "mean_token_accuracy": 0.8083424009382725,
      "num_tokens": 822971495.0,
      "step": 25786
    },
    {
      "entropy": 0.452839570119977,
      "epoch": 2.372533797965509,
      "grad_norm": 0.15459652245044708,
      "learning_rate": 2.091882111203116e-05,
      "loss": 0.4408,
      "mean_token_accuracy": 0.8574892058968544,
      "num_tokens": 823003774.0,
      "step": 25787
    },
    {
      "entropy": 0.4943582871928811,
      "epoch": 2.37262580433695,
      "grad_norm": 0.15920698642730713,
      "learning_rate": 2.0915754285889532e-05,
      "loss": 0.478,
      "mean_token_accuracy": 0.8465146608650684,
      "num_tokens": 823035035.0,
      "step": 25788
    },
    {
      "entropy": 0.4447372378781438,
      "epoch": 2.3727178107083917,
      "grad_norm": 0.1505594551563263,
      "learning_rate": 2.0912687459747906e-05,
      "loss": 0.4359,
      "mean_token_accuracy": 0.8595007695257664,
      "num_tokens": 823066899.0,
      "step": 25789
    },
    {
      "entropy": 0.38903853157535195,
      "epoch": 2.372809817079833,
      "grad_norm": 0.14110662043094635,
      "learning_rate": 2.090962063360628e-05,
      "loss": 0.3694,
      "mean_token_accuracy": 0.8823918662965298,
      "num_tokens": 823099354.0,
      "step": 25790
    },
    {
      "entropy": 0.42990896943956614,
      "epoch": 2.372901823451274,
      "grad_norm": 0.1478475034236908,
      "learning_rate": 2.0906553807464656e-05,
      "loss": 0.417,
      "mean_token_accuracy": 0.8650495409965515,
      "num_tokens": 823131076.0,
      "step": 25791
    },
    {
      "entropy": 0.4521518498659134,
      "epoch": 2.372993829822715,
      "grad_norm": 0.15019738674163818,
      "learning_rate": 2.090348698132303e-05,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8603444024920464,
      "num_tokens": 823163844.0,
      "step": 25792
    },
    {
      "entropy": 0.5570888975635171,
      "epoch": 2.3730858361941563,
      "grad_norm": 0.16687080264091492,
      "learning_rate": 2.0900420155181405e-05,
      "loss": 0.552,
      "mean_token_accuracy": 0.8276279680430889,
      "num_tokens": 823196054.0,
      "step": 25793
    },
    {
      "entropy": 0.5491012707352638,
      "epoch": 2.373177842565598,
      "grad_norm": 0.16072969138622284,
      "learning_rate": 2.089735332903978e-05,
      "loss": 0.5324,
      "mean_token_accuracy": 0.8311360627412796,
      "num_tokens": 823227885.0,
      "step": 25794
    },
    {
      "entropy": 0.5041112243197858,
      "epoch": 2.373269848937039,
      "grad_norm": 0.16113553941249847,
      "learning_rate": 2.0894286502898154e-05,
      "loss": 0.4842,
      "mean_token_accuracy": 0.8435670956969261,
      "num_tokens": 823259473.0,
      "step": 25795
    },
    {
      "entropy": 0.4806255176663399,
      "epoch": 2.37336185530848,
      "grad_norm": 0.1556052714586258,
      "learning_rate": 2.0891219676756526e-05,
      "loss": 0.4848,
      "mean_token_accuracy": 0.851337306201458,
      "num_tokens": 823290788.0,
      "step": 25796
    },
    {
      "entropy": 0.4963030191138387,
      "epoch": 2.3734538616799212,
      "grad_norm": 0.15276473760604858,
      "learning_rate": 2.08881528506149e-05,
      "loss": 0.4892,
      "mean_token_accuracy": 0.8470772244036198,
      "num_tokens": 823323280.0,
      "step": 25797
    },
    {
      "entropy": 0.6217903196811676,
      "epoch": 2.3735458680513624,
      "grad_norm": 0.16540300846099854,
      "learning_rate": 2.0885086024473275e-05,
      "loss": 0.6112,
      "mean_token_accuracy": 0.8108683452010155,
      "num_tokens": 823355408.0,
      "step": 25798
    },
    {
      "entropy": 0.513604924082756,
      "epoch": 2.373637874422804,
      "grad_norm": 0.15936753153800964,
      "learning_rate": 2.088201919833165e-05,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8409341163933277,
      "num_tokens": 823387868.0,
      "step": 25799
    },
    {
      "entropy": 0.6281321458518505,
      "epoch": 2.373729880794245,
      "grad_norm": 0.1737850308418274,
      "learning_rate": 2.087895237219002e-05,
      "loss": 0.6137,
      "mean_token_accuracy": 0.8116971589624882,
      "num_tokens": 823420219.0,
      "step": 25800
    },
    {
      "entropy": 0.5248638954944909,
      "epoch": 2.373821887165686,
      "grad_norm": 0.16005992889404297,
      "learning_rate": 2.0875885546048395e-05,
      "loss": 0.5097,
      "mean_token_accuracy": 0.8409692868590355,
      "num_tokens": 823452106.0,
      "step": 25801
    },
    {
      "entropy": 0.6044590920209885,
      "epoch": 2.3739138935371273,
      "grad_norm": 0.17443960905075073,
      "learning_rate": 2.087281871990677e-05,
      "loss": 0.5851,
      "mean_token_accuracy": 0.8149963766336441,
      "num_tokens": 823484037.0,
      "step": 25802
    },
    {
      "entropy": 0.5352978371083736,
      "epoch": 2.3740058999085685,
      "grad_norm": 0.16177284717559814,
      "learning_rate": 2.0869751893765145e-05,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8346748873591423,
      "num_tokens": 823515044.0,
      "step": 25803
    },
    {
      "entropy": 0.5031782519072294,
      "epoch": 2.37409790628001,
      "grad_norm": 0.1568748950958252,
      "learning_rate": 2.0866685067623516e-05,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8418217822909355,
      "num_tokens": 823547382.0,
      "step": 25804
    },
    {
      "entropy": 0.5692551359534264,
      "epoch": 2.374189912651451,
      "grad_norm": 0.1723451316356659,
      "learning_rate": 2.086361824148189e-05,
      "loss": 0.5543,
      "mean_token_accuracy": 0.8291160687804222,
      "num_tokens": 823579964.0,
      "step": 25805
    },
    {
      "entropy": 0.45489203627221286,
      "epoch": 2.3742819190228923,
      "grad_norm": 0.15250860154628754,
      "learning_rate": 2.0860551415340265e-05,
      "loss": 0.4445,
      "mean_token_accuracy": 0.8625864833593369,
      "num_tokens": 823611059.0,
      "step": 25806
    },
    {
      "entropy": 0.561831939034164,
      "epoch": 2.3743739253943335,
      "grad_norm": 0.1703481823205948,
      "learning_rate": 2.085748458919864e-05,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8242750279605389,
      "num_tokens": 823642416.0,
      "step": 25807
    },
    {
      "entropy": 0.5133491074666381,
      "epoch": 2.3744659317657746,
      "grad_norm": 0.1615981161594391,
      "learning_rate": 2.0854417763057015e-05,
      "loss": 0.5085,
      "mean_token_accuracy": 0.8375977016985416,
      "num_tokens": 823675184.0,
      "step": 25808
    },
    {
      "entropy": 0.5209038117900491,
      "epoch": 2.374557938137216,
      "grad_norm": 0.1559821367263794,
      "learning_rate": 2.0851350936915386e-05,
      "loss": 0.5113,
      "mean_token_accuracy": 0.8386948518455029,
      "num_tokens": 823706802.0,
      "step": 25809
    },
    {
      "entropy": 0.6093788053840399,
      "epoch": 2.3746499445086573,
      "grad_norm": 0.1648399531841278,
      "learning_rate": 2.084828411077376e-05,
      "loss": 0.5897,
      "mean_token_accuracy": 0.815287534147501,
      "num_tokens": 823738715.0,
      "step": 25810
    },
    {
      "entropy": 0.5051299049519002,
      "epoch": 2.3747419508800984,
      "grad_norm": 0.16083866357803345,
      "learning_rate": 2.0845217284632135e-05,
      "loss": 0.4891,
      "mean_token_accuracy": 0.846304003149271,
      "num_tokens": 823770260.0,
      "step": 25811
    },
    {
      "entropy": 0.5385047430172563,
      "epoch": 2.3748339572515396,
      "grad_norm": 0.16397373378276825,
      "learning_rate": 2.084215045849051e-05,
      "loss": 0.531,
      "mean_token_accuracy": 0.8300644569098949,
      "num_tokens": 823802593.0,
      "step": 25812
    },
    {
      "entropy": 0.5011907620355487,
      "epoch": 2.3749259636229807,
      "grad_norm": 0.1585693210363388,
      "learning_rate": 2.083908363234888e-05,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8440257906913757,
      "num_tokens": 823834390.0,
      "step": 25813
    },
    {
      "entropy": 0.48879929911345243,
      "epoch": 2.3750179699944223,
      "grad_norm": 0.1602981686592102,
      "learning_rate": 2.0836016806207256e-05,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8475508689880371,
      "num_tokens": 823866160.0,
      "step": 25814
    },
    {
      "entropy": 0.4711970016360283,
      "epoch": 2.3751099763658634,
      "grad_norm": 0.1609192043542862,
      "learning_rate": 2.083294998006563e-05,
      "loss": 0.4712,
      "mean_token_accuracy": 0.8505725264549255,
      "num_tokens": 823898338.0,
      "step": 25815
    },
    {
      "entropy": 0.5266372212208807,
      "epoch": 2.3752019827373045,
      "grad_norm": 0.15724918246269226,
      "learning_rate": 2.0829883153924005e-05,
      "loss": 0.52,
      "mean_token_accuracy": 0.833451509475708,
      "num_tokens": 823930660.0,
      "step": 25816
    },
    {
      "entropy": 0.5191068975254893,
      "epoch": 2.3752939891087457,
      "grad_norm": 0.1605096012353897,
      "learning_rate": 2.082681632778238e-05,
      "loss": 0.5019,
      "mean_token_accuracy": 0.841172605752945,
      "num_tokens": 823962558.0,
      "step": 25817
    },
    {
      "entropy": 0.5154095208272338,
      "epoch": 2.375385995480187,
      "grad_norm": 0.16898825764656067,
      "learning_rate": 2.082374950164075e-05,
      "loss": 0.5175,
      "mean_token_accuracy": 0.834997795522213,
      "num_tokens": 823993681.0,
      "step": 25818
    },
    {
      "entropy": 0.510389216709882,
      "epoch": 2.3754780018516284,
      "grad_norm": 0.1633961796760559,
      "learning_rate": 2.0820682675499126e-05,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8385244682431221,
      "num_tokens": 824026072.0,
      "step": 25819
    },
    {
      "entropy": 0.546795504167676,
      "epoch": 2.3755700082230695,
      "grad_norm": 0.1623736321926117,
      "learning_rate": 2.08176158493575e-05,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8368657864630222,
      "num_tokens": 824057878.0,
      "step": 25820
    },
    {
      "entropy": 0.5266191866248846,
      "epoch": 2.3756620145945107,
      "grad_norm": 0.16650384664535522,
      "learning_rate": 2.0814549023215875e-05,
      "loss": 0.5191,
      "mean_token_accuracy": 0.8360488601028919,
      "num_tokens": 824089873.0,
      "step": 25821
    },
    {
      "entropy": 0.4439473347738385,
      "epoch": 2.375754020965952,
      "grad_norm": 0.1518225371837616,
      "learning_rate": 2.081148219707425e-05,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8638092130422592,
      "num_tokens": 824121989.0,
      "step": 25822
    },
    {
      "entropy": 0.4985537799075246,
      "epoch": 2.375846027337393,
      "grad_norm": 0.15935808420181274,
      "learning_rate": 2.0808415370932624e-05,
      "loss": 0.4856,
      "mean_token_accuracy": 0.8474927172064781,
      "num_tokens": 824153603.0,
      "step": 25823
    },
    {
      "entropy": 0.49546638736501336,
      "epoch": 2.3759380337088345,
      "grad_norm": 0.16280370950698853,
      "learning_rate": 2.0805348544791e-05,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8460226431488991,
      "num_tokens": 824185464.0,
      "step": 25824
    },
    {
      "entropy": 0.5156042613089085,
      "epoch": 2.3760300400802756,
      "grad_norm": 0.16080118715763092,
      "learning_rate": 2.0802281718649373e-05,
      "loss": 0.5001,
      "mean_token_accuracy": 0.8387511745095253,
      "num_tokens": 824217821.0,
      "step": 25825
    },
    {
      "entropy": 0.5073393601924181,
      "epoch": 2.3761220464517168,
      "grad_norm": 0.15903666615486145,
      "learning_rate": 2.0799214892507745e-05,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8442962542176247,
      "num_tokens": 824248859.0,
      "step": 25826
    },
    {
      "entropy": 0.4591219602152705,
      "epoch": 2.376214052823158,
      "grad_norm": 0.1497235745191574,
      "learning_rate": 2.079614806636612e-05,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8598162904381752,
      "num_tokens": 824280756.0,
      "step": 25827
    },
    {
      "entropy": 0.5608613919466734,
      "epoch": 2.376306059194599,
      "grad_norm": 0.1694476157426834,
      "learning_rate": 2.0793081240224494e-05,
      "loss": 0.5508,
      "mean_token_accuracy": 0.8247724510729313,
      "num_tokens": 824311874.0,
      "step": 25828
    },
    {
      "entropy": 0.5635760352015495,
      "epoch": 2.3763980655660406,
      "grad_norm": 0.16263307631015778,
      "learning_rate": 2.079001441408287e-05,
      "loss": 0.552,
      "mean_token_accuracy": 0.8235451467335224,
      "num_tokens": 824344189.0,
      "step": 25829
    },
    {
      "entropy": 0.5442386139184237,
      "epoch": 2.3764900719374817,
      "grad_norm": 0.16632680594921112,
      "learning_rate": 2.078694758794124e-05,
      "loss": 0.5327,
      "mean_token_accuracy": 0.8324128873646259,
      "num_tokens": 824375602.0,
      "step": 25830
    },
    {
      "entropy": 0.6126197027042508,
      "epoch": 2.376582078308923,
      "grad_norm": 0.17415748536586761,
      "learning_rate": 2.0783880761799615e-05,
      "loss": 0.5882,
      "mean_token_accuracy": 0.817116491496563,
      "num_tokens": 824406831.0,
      "step": 25831
    },
    {
      "entropy": 0.5734748244285583,
      "epoch": 2.376674084680364,
      "grad_norm": 0.16452011466026306,
      "learning_rate": 2.078081393565799e-05,
      "loss": 0.5594,
      "mean_token_accuracy": 0.8210068233311176,
      "num_tokens": 824438840.0,
      "step": 25832
    },
    {
      "entropy": 0.48877802793867886,
      "epoch": 2.376766091051805,
      "grad_norm": 0.15810152888298035,
      "learning_rate": 2.0777747109516364e-05,
      "loss": 0.4818,
      "mean_token_accuracy": 0.8498333245515823,
      "num_tokens": 824470030.0,
      "step": 25833
    },
    {
      "entropy": 0.5695974947884679,
      "epoch": 2.3768580974232467,
      "grad_norm": 0.16661344468593597,
      "learning_rate": 2.0774680283374735e-05,
      "loss": 0.5554,
      "mean_token_accuracy": 0.8224705681204796,
      "num_tokens": 824501905.0,
      "step": 25834
    },
    {
      "entropy": 0.49095697957091033,
      "epoch": 2.376950103794688,
      "grad_norm": 0.15634861588478088,
      "learning_rate": 2.077161345723311e-05,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8466938734054565,
      "num_tokens": 824534050.0,
      "step": 25835
    },
    {
      "entropy": 0.5745888072997332,
      "epoch": 2.377042110166129,
      "grad_norm": 0.16464343667030334,
      "learning_rate": 2.0768546631091484e-05,
      "loss": 0.5698,
      "mean_token_accuracy": 0.8163165599107742,
      "num_tokens": 824566818.0,
      "step": 25836
    },
    {
      "entropy": 0.47202992625534534,
      "epoch": 2.37713411653757,
      "grad_norm": 0.15492796897888184,
      "learning_rate": 2.076547980494986e-05,
      "loss": 0.4615,
      "mean_token_accuracy": 0.853841420263052,
      "num_tokens": 824599409.0,
      "step": 25837
    },
    {
      "entropy": 0.4600107320584357,
      "epoch": 2.3772261229090113,
      "grad_norm": 0.1580265760421753,
      "learning_rate": 2.0762412978808234e-05,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8573556132614613,
      "num_tokens": 824631685.0,
      "step": 25838
    },
    {
      "entropy": 0.5093472097069025,
      "epoch": 2.377318129280453,
      "grad_norm": 0.16146047413349152,
      "learning_rate": 2.0759346152666605e-05,
      "loss": 0.4946,
      "mean_token_accuracy": 0.8458077311515808,
      "num_tokens": 824663610.0,
      "step": 25839
    },
    {
      "entropy": 0.46023987978696823,
      "epoch": 2.377410135651894,
      "grad_norm": 0.151123508810997,
      "learning_rate": 2.075627932652498e-05,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8564922623336315,
      "num_tokens": 824695959.0,
      "step": 25840
    },
    {
      "entropy": 0.5622863054741174,
      "epoch": 2.377502142023335,
      "grad_norm": 0.16520239412784576,
      "learning_rate": 2.0753212500383354e-05,
      "loss": 0.546,
      "mean_token_accuracy": 0.8281807377934456,
      "num_tokens": 824727901.0,
      "step": 25841
    },
    {
      "entropy": 0.5312357593793422,
      "epoch": 2.3775941483947762,
      "grad_norm": 0.16383448243141174,
      "learning_rate": 2.075014567424173e-05,
      "loss": 0.5139,
      "mean_token_accuracy": 0.8413306064903736,
      "num_tokens": 824759693.0,
      "step": 25842
    },
    {
      "entropy": 0.5781342200934887,
      "epoch": 2.3776861547662174,
      "grad_norm": 0.16884425282478333,
      "learning_rate": 2.07470788481001e-05,
      "loss": 0.5713,
      "mean_token_accuracy": 0.8172883540391922,
      "num_tokens": 824790956.0,
      "step": 25843
    },
    {
      "entropy": 0.5668205972760916,
      "epoch": 2.377778161137659,
      "grad_norm": 0.16102944314479828,
      "learning_rate": 2.0744012021958475e-05,
      "loss": 0.551,
      "mean_token_accuracy": 0.8260858729481697,
      "num_tokens": 824823544.0,
      "step": 25844
    },
    {
      "entropy": 0.5133787859231234,
      "epoch": 2.3778701675091,
      "grad_norm": 0.16599534451961517,
      "learning_rate": 2.074094519581685e-05,
      "loss": 0.503,
      "mean_token_accuracy": 0.8387409225106239,
      "num_tokens": 824855545.0,
      "step": 25845
    },
    {
      "entropy": 0.4233817118220031,
      "epoch": 2.377962173880541,
      "grad_norm": 0.1526292860507965,
      "learning_rate": 2.0737878369675224e-05,
      "loss": 0.4199,
      "mean_token_accuracy": 0.8617386743426323,
      "num_tokens": 824887356.0,
      "step": 25846
    },
    {
      "entropy": 0.5533284395933151,
      "epoch": 2.3780541802519823,
      "grad_norm": 0.15558716654777527,
      "learning_rate": 2.07348115435336e-05,
      "loss": 0.5233,
      "mean_token_accuracy": 0.833095196634531,
      "num_tokens": 824918843.0,
      "step": 25847
    },
    {
      "entropy": 0.524910387583077,
      "epoch": 2.3781461866234235,
      "grad_norm": 0.15965649485588074,
      "learning_rate": 2.073174471739197e-05,
      "loss": 0.5181,
      "mean_token_accuracy": 0.8353317901492119,
      "num_tokens": 824950268.0,
      "step": 25848
    },
    {
      "entropy": 0.5302088316529989,
      "epoch": 2.378238192994865,
      "grad_norm": 0.1693081259727478,
      "learning_rate": 2.0728677891250345e-05,
      "loss": 0.5162,
      "mean_token_accuracy": 0.832400631159544,
      "num_tokens": 824981040.0,
      "step": 25849
    },
    {
      "entropy": 0.4801434511318803,
      "epoch": 2.378330199366306,
      "grad_norm": 0.15709340572357178,
      "learning_rate": 2.072561106510872e-05,
      "loss": 0.479,
      "mean_token_accuracy": 0.8490546867251396,
      "num_tokens": 825013375.0,
      "step": 25850
    },
    {
      "entropy": 0.47138975467532873,
      "epoch": 2.3784222057377473,
      "grad_norm": 0.15363475680351257,
      "learning_rate": 2.0722544238967094e-05,
      "loss": 0.4551,
      "mean_token_accuracy": 0.8543913178145885,
      "num_tokens": 825045568.0,
      "step": 25851
    },
    {
      "entropy": 0.4998994735069573,
      "epoch": 2.3785142121091885,
      "grad_norm": 0.15593880414962769,
      "learning_rate": 2.071947741282547e-05,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8487939387559891,
      "num_tokens": 825077607.0,
      "step": 25852
    },
    {
      "entropy": 0.3991223117336631,
      "epoch": 2.3786062184806296,
      "grad_norm": 0.14233161509037018,
      "learning_rate": 2.0716410586683843e-05,
      "loss": 0.3864,
      "mean_token_accuracy": 0.8750565350055695,
      "num_tokens": 825109851.0,
      "step": 25853
    },
    {
      "entropy": 0.5227622790262103,
      "epoch": 2.378698224852071,
      "grad_norm": 0.15641175210475922,
      "learning_rate": 2.0713343760542218e-05,
      "loss": 0.5148,
      "mean_token_accuracy": 0.8384633287787437,
      "num_tokens": 825141927.0,
      "step": 25854
    },
    {
      "entropy": 0.51039969176054,
      "epoch": 2.3787902312235123,
      "grad_norm": 0.1605255901813507,
      "learning_rate": 2.0710276934400593e-05,
      "loss": 0.513,
      "mean_token_accuracy": 0.8412904590368271,
      "num_tokens": 825173653.0,
      "step": 25855
    },
    {
      "entropy": 0.525703689083457,
      "epoch": 2.3788822375949534,
      "grad_norm": 0.1616133451461792,
      "learning_rate": 2.0707210108258964e-05,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8352042362093925,
      "num_tokens": 825205946.0,
      "step": 25856
    },
    {
      "entropy": 0.5758435390889645,
      "epoch": 2.3789742439663946,
      "grad_norm": 0.16650500893592834,
      "learning_rate": 2.070414328211734e-05,
      "loss": 0.5657,
      "mean_token_accuracy": 0.8225490376353264,
      "num_tokens": 825238074.0,
      "step": 25857
    },
    {
      "entropy": 0.5174146029166877,
      "epoch": 2.3790662503378357,
      "grad_norm": 0.16120734810829163,
      "learning_rate": 2.0701076455975713e-05,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8386805467307568,
      "num_tokens": 825269664.0,
      "step": 25858
    },
    {
      "entropy": 0.39651605440303683,
      "epoch": 2.3791582567092773,
      "grad_norm": 0.14601902663707733,
      "learning_rate": 2.0698009629834088e-05,
      "loss": 0.3916,
      "mean_token_accuracy": 0.8745321370661259,
      "num_tokens": 825302008.0,
      "step": 25859
    },
    {
      "entropy": 0.4729943023994565,
      "epoch": 2.3792502630807184,
      "grad_norm": 0.15114152431488037,
      "learning_rate": 2.069494280369246e-05,
      "loss": 0.4501,
      "mean_token_accuracy": 0.8587107919156551,
      "num_tokens": 825334558.0,
      "step": 25860
    },
    {
      "entropy": 0.6731477044522762,
      "epoch": 2.3793422694521595,
      "grad_norm": 0.17072172462940216,
      "learning_rate": 2.0691875977550834e-05,
      "loss": 0.6503,
      "mean_token_accuracy": 0.7962285540997982,
      "num_tokens": 825367014.0,
      "step": 25861
    },
    {
      "entropy": 0.5252726580947638,
      "epoch": 2.3794342758236007,
      "grad_norm": 0.16014456748962402,
      "learning_rate": 2.0688809151409208e-05,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8373378366231918,
      "num_tokens": 825399731.0,
      "step": 25862
    },
    {
      "entropy": 0.5416075633838773,
      "epoch": 2.379526282195042,
      "grad_norm": 0.16116903722286224,
      "learning_rate": 2.0685742325267583e-05,
      "loss": 0.5194,
      "mean_token_accuracy": 0.8347915820777416,
      "num_tokens": 825431137.0,
      "step": 25863
    },
    {
      "entropy": 0.5712365116924047,
      "epoch": 2.3796182885664834,
      "grad_norm": 0.16396665573120117,
      "learning_rate": 2.0682675499125954e-05,
      "loss": 0.5497,
      "mean_token_accuracy": 0.8279120326042175,
      "num_tokens": 825463205.0,
      "step": 25864
    },
    {
      "entropy": 0.47906726598739624,
      "epoch": 2.3797102949379245,
      "grad_norm": 0.15877453982830048,
      "learning_rate": 2.067960867298433e-05,
      "loss": 0.4719,
      "mean_token_accuracy": 0.8490740917623043,
      "num_tokens": 825494932.0,
      "step": 25865
    },
    {
      "entropy": 0.5789443185785785,
      "epoch": 2.3798023013093657,
      "grad_norm": 0.16969341039657593,
      "learning_rate": 2.0676541846842703e-05,
      "loss": 0.5585,
      "mean_token_accuracy": 0.8215925544500351,
      "num_tokens": 825526185.0,
      "step": 25866
    },
    {
      "entropy": 0.5045758378691971,
      "epoch": 2.379894307680807,
      "grad_norm": 0.15824896097183228,
      "learning_rate": 2.0673475020701078e-05,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8447554782032967,
      "num_tokens": 825557725.0,
      "step": 25867
    },
    {
      "entropy": 0.46349386032670736,
      "epoch": 2.379986314052248,
      "grad_norm": 0.15214894711971283,
      "learning_rate": 2.0670408194559453e-05,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8557786010205746,
      "num_tokens": 825589986.0,
      "step": 25868
    },
    {
      "entropy": 0.507754372432828,
      "epoch": 2.3800783204236895,
      "grad_norm": 0.15921451151371002,
      "learning_rate": 2.0667341368417824e-05,
      "loss": 0.4892,
      "mean_token_accuracy": 0.8436979092657566,
      "num_tokens": 825621781.0,
      "step": 25869
    },
    {
      "entropy": 0.4974041630048305,
      "epoch": 2.3801703267951306,
      "grad_norm": 0.1597600281238556,
      "learning_rate": 2.06642745422762e-05,
      "loss": 0.4781,
      "mean_token_accuracy": 0.8486153483390808,
      "num_tokens": 825653160.0,
      "step": 25870
    },
    {
      "entropy": 0.5472579477354884,
      "epoch": 2.3802623331665718,
      "grad_norm": 0.16791489720344543,
      "learning_rate": 2.0661207716134573e-05,
      "loss": 0.5421,
      "mean_token_accuracy": 0.8297888077795506,
      "num_tokens": 825685277.0,
      "step": 25871
    },
    {
      "entropy": 0.5718789957463741,
      "epoch": 2.380354339538013,
      "grad_norm": 0.16938020288944244,
      "learning_rate": 2.0658140889992948e-05,
      "loss": 0.562,
      "mean_token_accuracy": 0.8221174515783787,
      "num_tokens": 825717041.0,
      "step": 25872
    },
    {
      "entropy": 0.555017426609993,
      "epoch": 2.380446345909454,
      "grad_norm": 0.16075724363327026,
      "learning_rate": 2.065507406385132e-05,
      "loss": 0.538,
      "mean_token_accuracy": 0.8289935365319252,
      "num_tokens": 825748553.0,
      "step": 25873
    },
    {
      "entropy": 0.49509680178016424,
      "epoch": 2.3805383522808956,
      "grad_norm": 0.15924659371376038,
      "learning_rate": 2.0652007237709694e-05,
      "loss": 0.4816,
      "mean_token_accuracy": 0.845067135989666,
      "num_tokens": 825780483.0,
      "step": 25874
    },
    {
      "entropy": 0.47908139368519187,
      "epoch": 2.3806303586523367,
      "grad_norm": 0.1543668806552887,
      "learning_rate": 2.064894041156807e-05,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8498321585357189,
      "num_tokens": 825813033.0,
      "step": 25875
    },
    {
      "entropy": 0.5066130328923464,
      "epoch": 2.380722365023778,
      "grad_norm": 0.16175256669521332,
      "learning_rate": 2.0645873585426443e-05,
      "loss": 0.4971,
      "mean_token_accuracy": 0.8430140689015388,
      "num_tokens": 825845125.0,
      "step": 25876
    },
    {
      "entropy": 0.568053275346756,
      "epoch": 2.380814371395219,
      "grad_norm": 0.16927622258663177,
      "learning_rate": 2.0642806759284818e-05,
      "loss": 0.5636,
      "mean_token_accuracy": 0.8226006254553795,
      "num_tokens": 825876435.0,
      "step": 25877
    },
    {
      "entropy": 0.5931488643400371,
      "epoch": 2.38090637776666,
      "grad_norm": 0.16693143546581268,
      "learning_rate": 2.063973993314319e-05,
      "loss": 0.5869,
      "mean_token_accuracy": 0.8163832388818264,
      "num_tokens": 825908268.0,
      "step": 25878
    },
    {
      "entropy": 0.5475636799819767,
      "epoch": 2.3809983841381017,
      "grad_norm": 0.16924087703227997,
      "learning_rate": 2.0636673107001564e-05,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8292589224874973,
      "num_tokens": 825939524.0,
      "step": 25879
    },
    {
      "entropy": 0.5204731160774827,
      "epoch": 2.381090390509543,
      "grad_norm": 0.16391995549201965,
      "learning_rate": 2.063360628085994e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8394138142466545,
      "num_tokens": 825970816.0,
      "step": 25880
    },
    {
      "entropy": 0.5219220295548439,
      "epoch": 2.381182396880984,
      "grad_norm": 0.1562340408563614,
      "learning_rate": 2.0630539454718313e-05,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8403976410627365,
      "num_tokens": 826003036.0,
      "step": 25881
    },
    {
      "entropy": 0.5775115136057138,
      "epoch": 2.381274403252425,
      "grad_norm": 0.1696782410144806,
      "learning_rate": 2.0627472628576684e-05,
      "loss": 0.5705,
      "mean_token_accuracy": 0.8203858509659767,
      "num_tokens": 826035183.0,
      "step": 25882
    },
    {
      "entropy": 0.5678508169949055,
      "epoch": 2.3813664096238663,
      "grad_norm": 0.16177619993686676,
      "learning_rate": 2.0624405802435062e-05,
      "loss": 0.5533,
      "mean_token_accuracy": 0.8243931084871292,
      "num_tokens": 826067279.0,
      "step": 25883
    },
    {
      "entropy": 0.5427110437303782,
      "epoch": 2.381458415995308,
      "grad_norm": 0.16702044010162354,
      "learning_rate": 2.0621338976293437e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8319164216518402,
      "num_tokens": 826098474.0,
      "step": 25884
    },
    {
      "entropy": 0.42329040821641684,
      "epoch": 2.381550422366749,
      "grad_norm": 0.14473743736743927,
      "learning_rate": 2.061827215015181e-05,
      "loss": 0.4088,
      "mean_token_accuracy": 0.8679927960038185,
      "num_tokens": 826130749.0,
      "step": 25885
    },
    {
      "entropy": 0.45802905410528183,
      "epoch": 2.38164242873819,
      "grad_norm": 0.15500357747077942,
      "learning_rate": 2.0615205324010183e-05,
      "loss": 0.4481,
      "mean_token_accuracy": 0.8589239455759525,
      "num_tokens": 826162938.0,
      "step": 25886
    },
    {
      "entropy": 0.49927312368527055,
      "epoch": 2.3817344351096312,
      "grad_norm": 0.15867415070533752,
      "learning_rate": 2.0612138497868558e-05,
      "loss": 0.4845,
      "mean_token_accuracy": 0.8445011004805565,
      "num_tokens": 826195404.0,
      "step": 25887
    },
    {
      "entropy": 0.5473742266185582,
      "epoch": 2.3818264414810724,
      "grad_norm": 0.16004277765750885,
      "learning_rate": 2.0609071671726932e-05,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8310508988797665,
      "num_tokens": 826227550.0,
      "step": 25888
    },
    {
      "entropy": 0.5131469331681728,
      "epoch": 2.381918447852514,
      "grad_norm": 0.15568168461322784,
      "learning_rate": 2.0606004845585307e-05,
      "loss": 0.4999,
      "mean_token_accuracy": 0.8396494537591934,
      "num_tokens": 826259927.0,
      "step": 25889
    },
    {
      "entropy": 0.5377856781706214,
      "epoch": 2.382010454223955,
      "grad_norm": 0.16173651814460754,
      "learning_rate": 2.0602938019443678e-05,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8364325650036335,
      "num_tokens": 826291978.0,
      "step": 25890
    },
    {
      "entropy": 0.46619252022355795,
      "epoch": 2.382102460595396,
      "grad_norm": 0.15599152445793152,
      "learning_rate": 2.0599871193302053e-05,
      "loss": 0.4592,
      "mean_token_accuracy": 0.8530080057680607,
      "num_tokens": 826324229.0,
      "step": 25891
    },
    {
      "entropy": 0.6853569112718105,
      "epoch": 2.3821944669668373,
      "grad_norm": 0.18118664622306824,
      "learning_rate": 2.0596804367160427e-05,
      "loss": 0.6756,
      "mean_token_accuracy": 0.7858183681964874,
      "num_tokens": 826355807.0,
      "step": 25892
    },
    {
      "entropy": 0.5088647180236876,
      "epoch": 2.3822864733382785,
      "grad_norm": 0.16011229157447815,
      "learning_rate": 2.0593737541018802e-05,
      "loss": 0.4943,
      "mean_token_accuracy": 0.8424270749092102,
      "num_tokens": 826386982.0,
      "step": 25893
    },
    {
      "entropy": 0.5582432728260756,
      "epoch": 2.38237847970972,
      "grad_norm": 0.1750139594078064,
      "learning_rate": 2.0590670714877173e-05,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8225109502673149,
      "num_tokens": 826418875.0,
      "step": 25894
    },
    {
      "entropy": 0.5631539290770888,
      "epoch": 2.382470486081161,
      "grad_norm": 0.16580963134765625,
      "learning_rate": 2.0587603888735548e-05,
      "loss": 0.5567,
      "mean_token_accuracy": 0.8264936096966267,
      "num_tokens": 826450746.0,
      "step": 25895
    },
    {
      "entropy": 0.4684787499718368,
      "epoch": 2.3825624924526023,
      "grad_norm": 0.17025139927864075,
      "learning_rate": 2.0584537062593923e-05,
      "loss": 0.4599,
      "mean_token_accuracy": 0.849850632250309,
      "num_tokens": 826481945.0,
      "step": 25896
    },
    {
      "entropy": 0.5394230782985687,
      "epoch": 2.3826544988240435,
      "grad_norm": 0.1638023406267166,
      "learning_rate": 2.0581470236452297e-05,
      "loss": 0.5352,
      "mean_token_accuracy": 0.8331033326685429,
      "num_tokens": 826514396.0,
      "step": 25897
    },
    {
      "entropy": 0.5824037007987499,
      "epoch": 2.3827465051954846,
      "grad_norm": 0.17241357266902924,
      "learning_rate": 2.0578403410310672e-05,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8194015771150589,
      "num_tokens": 826545808.0,
      "step": 25898
    },
    {
      "entropy": 0.5612873532809317,
      "epoch": 2.382838511566926,
      "grad_norm": 0.17056867480278015,
      "learning_rate": 2.0575336584169043e-05,
      "loss": 0.5462,
      "mean_token_accuracy": 0.826900340616703,
      "num_tokens": 826576548.0,
      "step": 25899
    },
    {
      "entropy": 0.5581387337297201,
      "epoch": 2.3829305179383673,
      "grad_norm": 0.16734643280506134,
      "learning_rate": 2.0572269758027418e-05,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8249708041548729,
      "num_tokens": 826608930.0,
      "step": 25900
    },
    {
      "entropy": 0.5873732529580593,
      "epoch": 2.3830225243098084,
      "grad_norm": 0.1713734120130539,
      "learning_rate": 2.0569202931885792e-05,
      "loss": 0.5686,
      "mean_token_accuracy": 0.8185121640563011,
      "num_tokens": 826640007.0,
      "step": 25901
    },
    {
      "entropy": 0.5119242733344436,
      "epoch": 2.3831145306812496,
      "grad_norm": 0.1574266254901886,
      "learning_rate": 2.0566136105744167e-05,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8412097133696079,
      "num_tokens": 826671941.0,
      "step": 25902
    },
    {
      "entropy": 0.5301620168611407,
      "epoch": 2.3832065370526907,
      "grad_norm": 0.1615232229232788,
      "learning_rate": 2.056306927960254e-05,
      "loss": 0.5171,
      "mean_token_accuracy": 0.8362441919744015,
      "num_tokens": 826703868.0,
      "step": 25903
    },
    {
      "entropy": 0.5355892949737608,
      "epoch": 2.3832985434241323,
      "grad_norm": 0.1556517779827118,
      "learning_rate": 2.0560002453460913e-05,
      "loss": 0.5081,
      "mean_token_accuracy": 0.8397416025400162,
      "num_tokens": 826735961.0,
      "step": 25904
    },
    {
      "entropy": 0.5464420272037387,
      "epoch": 2.3833905497955734,
      "grad_norm": 0.16142752766609192,
      "learning_rate": 2.0556935627319288e-05,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8350022546947002,
      "num_tokens": 826767957.0,
      "step": 25905
    },
    {
      "entropy": 0.5104989111423492,
      "epoch": 2.3834825561670145,
      "grad_norm": 0.16456343233585358,
      "learning_rate": 2.0553868801177662e-05,
      "loss": 0.4943,
      "mean_token_accuracy": 0.8419021442532539,
      "num_tokens": 826800287.0,
      "step": 25906
    },
    {
      "entropy": 0.5672890618443489,
      "epoch": 2.3835745625384557,
      "grad_norm": 0.16622059047222137,
      "learning_rate": 2.0550801975036037e-05,
      "loss": 0.5544,
      "mean_token_accuracy": 0.82893967628479,
      "num_tokens": 826832426.0,
      "step": 25907
    },
    {
      "entropy": 0.5227971961721778,
      "epoch": 2.383666568909897,
      "grad_norm": 0.161079540848732,
      "learning_rate": 2.0547735148894408e-05,
      "loss": 0.5146,
      "mean_token_accuracy": 0.8371739946305752,
      "num_tokens": 826864581.0,
      "step": 25908
    },
    {
      "entropy": 0.496181447757408,
      "epoch": 2.3837585752813384,
      "grad_norm": 0.15435706079006195,
      "learning_rate": 2.0544668322752783e-05,
      "loss": 0.4782,
      "mean_token_accuracy": 0.849008996039629,
      "num_tokens": 826897207.0,
      "step": 25909
    },
    {
      "entropy": 0.4524620436131954,
      "epoch": 2.3838505816527795,
      "grad_norm": 0.14914052188396454,
      "learning_rate": 2.0541601496611157e-05,
      "loss": 0.4333,
      "mean_token_accuracy": 0.861801877617836,
      "num_tokens": 826929417.0,
      "step": 25910
    },
    {
      "entropy": 0.5379295321181417,
      "epoch": 2.3839425880242207,
      "grad_norm": 0.16358782351016998,
      "learning_rate": 2.0538534670469532e-05,
      "loss": 0.5327,
      "mean_token_accuracy": 0.831787459552288,
      "num_tokens": 826961608.0,
      "step": 25911
    },
    {
      "entropy": 0.5400712471455336,
      "epoch": 2.384034594395662,
      "grad_norm": 0.16207663714885712,
      "learning_rate": 2.0535467844327903e-05,
      "loss": 0.5319,
      "mean_token_accuracy": 0.8283062428236008,
      "num_tokens": 826994017.0,
      "step": 25912
    },
    {
      "entropy": 0.5749440554063767,
      "epoch": 2.384126600767103,
      "grad_norm": 0.16806796193122864,
      "learning_rate": 2.053240101818628e-05,
      "loss": 0.5634,
      "mean_token_accuracy": 0.8208496533334255,
      "num_tokens": 827025707.0,
      "step": 25913
    },
    {
      "entropy": 0.5727726984769106,
      "epoch": 2.3842186071385445,
      "grad_norm": 0.17198924720287323,
      "learning_rate": 2.0529334192044656e-05,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8233381845057011,
      "num_tokens": 827057205.0,
      "step": 25914
    },
    {
      "entropy": 0.4825392896309495,
      "epoch": 2.3843106135099856,
      "grad_norm": 0.15945225954055786,
      "learning_rate": 2.052626736590303e-05,
      "loss": 0.4688,
      "mean_token_accuracy": 0.8477211706340313,
      "num_tokens": 827089182.0,
      "step": 25915
    },
    {
      "entropy": 0.4957933556288481,
      "epoch": 2.3844026198814268,
      "grad_norm": 0.16051021218299866,
      "learning_rate": 2.0523200539761402e-05,
      "loss": 0.4895,
      "mean_token_accuracy": 0.8457416743040085,
      "num_tokens": 827120712.0,
      "step": 25916
    },
    {
      "entropy": 0.47375369165092707,
      "epoch": 2.384494626252868,
      "grad_norm": 0.1528744101524353,
      "learning_rate": 2.0520133713619777e-05,
      "loss": 0.4622,
      "mean_token_accuracy": 0.8504408448934555,
      "num_tokens": 827153208.0,
      "step": 25917
    },
    {
      "entropy": 0.4149596532806754,
      "epoch": 2.384586632624309,
      "grad_norm": 0.14634756743907928,
      "learning_rate": 2.051706688747815e-05,
      "loss": 0.3959,
      "mean_token_accuracy": 0.8714150972664356,
      "num_tokens": 827185571.0,
      "step": 25918
    },
    {
      "entropy": 0.6283595031127334,
      "epoch": 2.3846786389957506,
      "grad_norm": 0.17380118370056152,
      "learning_rate": 2.0514000061336526e-05,
      "loss": 0.6283,
      "mean_token_accuracy": 0.8046891242265701,
      "num_tokens": 827217375.0,
      "step": 25919
    },
    {
      "entropy": 0.6556400395929813,
      "epoch": 2.3847706453671917,
      "grad_norm": 0.1769297868013382,
      "learning_rate": 2.0510933235194897e-05,
      "loss": 0.6453,
      "mean_token_accuracy": 0.7988788336515427,
      "num_tokens": 827249582.0,
      "step": 25920
    },
    {
      "entropy": 0.5786352567374706,
      "epoch": 2.384862651738633,
      "grad_norm": 0.16963008046150208,
      "learning_rate": 2.0507866409053272e-05,
      "loss": 0.5761,
      "mean_token_accuracy": 0.8210183680057526,
      "num_tokens": 827281614.0,
      "step": 25921
    },
    {
      "entropy": 0.5124780032783747,
      "epoch": 2.384954658110074,
      "grad_norm": 0.1610480397939682,
      "learning_rate": 2.0504799582911646e-05,
      "loss": 0.4999,
      "mean_token_accuracy": 0.8410065174102783,
      "num_tokens": 827312826.0,
      "step": 25922
    },
    {
      "entropy": 0.5059288162738085,
      "epoch": 2.385046664481515,
      "grad_norm": 0.16461288928985596,
      "learning_rate": 2.050173275677002e-05,
      "loss": 0.4866,
      "mean_token_accuracy": 0.8455861508846283,
      "num_tokens": 827344626.0,
      "step": 25923
    },
    {
      "entropy": 0.5014519435353577,
      "epoch": 2.3851386708529567,
      "grad_norm": 0.1577492654323578,
      "learning_rate": 2.0498665930628392e-05,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8489733561873436,
      "num_tokens": 827376363.0,
      "step": 25924
    },
    {
      "entropy": 0.5778761389665306,
      "epoch": 2.385230677224398,
      "grad_norm": 0.16530567407608032,
      "learning_rate": 2.0495599104486767e-05,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8256405740976334,
      "num_tokens": 827407888.0,
      "step": 25925
    },
    {
      "entropy": 0.47645785426720977,
      "epoch": 2.385322683595839,
      "grad_norm": 0.1539819985628128,
      "learning_rate": 2.049253227834514e-05,
      "loss": 0.4559,
      "mean_token_accuracy": 0.8531432747840881,
      "num_tokens": 827440022.0,
      "step": 25926
    },
    {
      "entropy": 0.4857247518375516,
      "epoch": 2.38541468996728,
      "grad_norm": 0.15596149861812592,
      "learning_rate": 2.0489465452203516e-05,
      "loss": 0.4845,
      "mean_token_accuracy": 0.8452218733727932,
      "num_tokens": 827472327.0,
      "step": 25927
    },
    {
      "entropy": 0.5775676919147372,
      "epoch": 2.3855066963387213,
      "grad_norm": 0.1651666909456253,
      "learning_rate": 2.048639862606189e-05,
      "loss": 0.5656,
      "mean_token_accuracy": 0.8215981163084507,
      "num_tokens": 827504235.0,
      "step": 25928
    },
    {
      "entropy": 0.46402856428176165,
      "epoch": 2.385598702710163,
      "grad_norm": 0.1551714986562729,
      "learning_rate": 2.0483331799920262e-05,
      "loss": 0.4523,
      "mean_token_accuracy": 0.8568431250751019,
      "num_tokens": 827536254.0,
      "step": 25929
    },
    {
      "entropy": 0.5684116929769516,
      "epoch": 2.385690709081604,
      "grad_norm": 0.1703130304813385,
      "learning_rate": 2.0480264973778637e-05,
      "loss": 0.5445,
      "mean_token_accuracy": 0.8243907988071442,
      "num_tokens": 827567736.0,
      "step": 25930
    },
    {
      "entropy": 0.5204666946083307,
      "epoch": 2.385782715453045,
      "grad_norm": 0.1596682369709015,
      "learning_rate": 2.047719814763701e-05,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8443445786833763,
      "num_tokens": 827599920.0,
      "step": 25931
    },
    {
      "entropy": 0.5063245352357626,
      "epoch": 2.3858747218244862,
      "grad_norm": 0.15960073471069336,
      "learning_rate": 2.0474131321495386e-05,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8431854695081711,
      "num_tokens": 827631902.0,
      "step": 25932
    },
    {
      "entropy": 0.5391277363523841,
      "epoch": 2.3859667281959274,
      "grad_norm": 0.16594447195529938,
      "learning_rate": 2.0471064495353757e-05,
      "loss": 0.5273,
      "mean_token_accuracy": 0.835563912987709,
      "num_tokens": 827663497.0,
      "step": 25933
    },
    {
      "entropy": 0.5754137029871345,
      "epoch": 2.386058734567369,
      "grad_norm": 0.17548662424087524,
      "learning_rate": 2.0467997669212132e-05,
      "loss": 0.5673,
      "mean_token_accuracy": 0.8226985409855843,
      "num_tokens": 827694299.0,
      "step": 25934
    },
    {
      "entropy": 0.5554325599223375,
      "epoch": 2.38615074093881,
      "grad_norm": 0.1627846211194992,
      "learning_rate": 2.0464930843070507e-05,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8340284712612629,
      "num_tokens": 827726255.0,
      "step": 25935
    },
    {
      "entropy": 0.5705868266522884,
      "epoch": 2.386242747310251,
      "grad_norm": 0.17726194858551025,
      "learning_rate": 2.046186401692888e-05,
      "loss": 0.5551,
      "mean_token_accuracy": 0.8250541873276234,
      "num_tokens": 827758061.0,
      "step": 25936
    },
    {
      "entropy": 0.5847843773663044,
      "epoch": 2.3863347536816923,
      "grad_norm": 0.16787011921405792,
      "learning_rate": 2.0458797190787256e-05,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8227411732077599,
      "num_tokens": 827789915.0,
      "step": 25937
    },
    {
      "entropy": 0.5205550193786621,
      "epoch": 2.3864267600531335,
      "grad_norm": 0.1665075570344925,
      "learning_rate": 2.0455730364645627e-05,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8374015651643276,
      "num_tokens": 827821696.0,
      "step": 25938
    },
    {
      "entropy": 0.5709926299750805,
      "epoch": 2.386518766424575,
      "grad_norm": 0.16650979220867157,
      "learning_rate": 2.0452663538504002e-05,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8233049996197224,
      "num_tokens": 827853647.0,
      "step": 25939
    },
    {
      "entropy": 0.5375868659466505,
      "epoch": 2.386610772796016,
      "grad_norm": 0.16289108991622925,
      "learning_rate": 2.0449596712362377e-05,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8340267464518547,
      "num_tokens": 827885804.0,
      "step": 25940
    },
    {
      "entropy": 0.521522032096982,
      "epoch": 2.3867027791674573,
      "grad_norm": 0.15588133037090302,
      "learning_rate": 2.044652988622075e-05,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8394428193569183,
      "num_tokens": 827917655.0,
      "step": 25941
    },
    {
      "entropy": 0.5363154988735914,
      "epoch": 2.3867947855388985,
      "grad_norm": 0.16259761154651642,
      "learning_rate": 2.0443463060079122e-05,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8343390971422195,
      "num_tokens": 827949767.0,
      "step": 25942
    },
    {
      "entropy": 0.6084919176064432,
      "epoch": 2.3868867919103396,
      "grad_norm": 0.17493607103824615,
      "learning_rate": 2.0440396233937497e-05,
      "loss": 0.6015,
      "mean_token_accuracy": 0.8135214485228062,
      "num_tokens": 827981504.0,
      "step": 25943
    },
    {
      "entropy": 0.4520797664299607,
      "epoch": 2.386978798281781,
      "grad_norm": 0.1522778421640396,
      "learning_rate": 2.0437329407795875e-05,
      "loss": 0.4471,
      "mean_token_accuracy": 0.8573425561189651,
      "num_tokens": 828013130.0,
      "step": 25944
    },
    {
      "entropy": 0.49557905830442905,
      "epoch": 2.3870708046532223,
      "grad_norm": 0.15854886174201965,
      "learning_rate": 2.043426258165425e-05,
      "loss": 0.4844,
      "mean_token_accuracy": 0.8479817025363445,
      "num_tokens": 828044856.0,
      "step": 25945
    },
    {
      "entropy": 0.5634048748761415,
      "epoch": 2.3871628110246634,
      "grad_norm": 0.16232487559318542,
      "learning_rate": 2.043119575551262e-05,
      "loss": 0.5426,
      "mean_token_accuracy": 0.829259816557169,
      "num_tokens": 828077367.0,
      "step": 25946
    },
    {
      "entropy": 0.5560693722218275,
      "epoch": 2.3872548173961046,
      "grad_norm": 0.17319342494010925,
      "learning_rate": 2.0428128929370996e-05,
      "loss": 0.5599,
      "mean_token_accuracy": 0.8217028677463531,
      "num_tokens": 828108330.0,
      "step": 25947
    },
    {
      "entropy": 0.5228100321255624,
      "epoch": 2.3873468237675457,
      "grad_norm": 0.1639045923948288,
      "learning_rate": 2.042506210322937e-05,
      "loss": 0.5158,
      "mean_token_accuracy": 0.8382198289036751,
      "num_tokens": 828140179.0,
      "step": 25948
    },
    {
      "entropy": 0.5802959706634283,
      "epoch": 2.3874388301389873,
      "grad_norm": 0.16733771562576294,
      "learning_rate": 2.0421995277087745e-05,
      "loss": 0.5596,
      "mean_token_accuracy": 0.8207046687602997,
      "num_tokens": 828172140.0,
      "step": 25949
    },
    {
      "entropy": 0.4287662226706743,
      "epoch": 2.3875308365104284,
      "grad_norm": 0.14804145693778992,
      "learning_rate": 2.0418928450946116e-05,
      "loss": 0.4104,
      "mean_token_accuracy": 0.8659519590437412,
      "num_tokens": 828204454.0,
      "step": 25950
    },
    {
      "entropy": 0.5848167268559337,
      "epoch": 2.3876228428818695,
      "grad_norm": 0.17001955211162567,
      "learning_rate": 2.041586162480449e-05,
      "loss": 0.5658,
      "mean_token_accuracy": 0.8193349055945873,
      "num_tokens": 828235765.0,
      "step": 25951
    },
    {
      "entropy": 0.5612141815945506,
      "epoch": 2.3877148492533107,
      "grad_norm": 0.1649446040391922,
      "learning_rate": 2.0412794798662866e-05,
      "loss": 0.5517,
      "mean_token_accuracy": 0.8267367742955685,
      "num_tokens": 828267836.0,
      "step": 25952
    },
    {
      "entropy": 0.4998874068260193,
      "epoch": 2.387806855624752,
      "grad_norm": 0.15527430176734924,
      "learning_rate": 2.040972797252124e-05,
      "loss": 0.483,
      "mean_token_accuracy": 0.8470882773399353,
      "num_tokens": 828299801.0,
      "step": 25953
    },
    {
      "entropy": 0.5662987995892763,
      "epoch": 2.3878988619961934,
      "grad_norm": 0.16614367067813873,
      "learning_rate": 2.040666114637961e-05,
      "loss": 0.559,
      "mean_token_accuracy": 0.8231033943593502,
      "num_tokens": 828332014.0,
      "step": 25954
    },
    {
      "entropy": 0.5497546368278563,
      "epoch": 2.3879908683676345,
      "grad_norm": 0.16762377321720123,
      "learning_rate": 2.0403594320237986e-05,
      "loss": 0.5262,
      "mean_token_accuracy": 0.8340549916028976,
      "num_tokens": 828363888.0,
      "step": 25955
    },
    {
      "entropy": 0.5041825827211142,
      "epoch": 2.3880828747390757,
      "grad_norm": 0.16508695483207703,
      "learning_rate": 2.040052749409636e-05,
      "loss": 0.4959,
      "mean_token_accuracy": 0.8427244946360588,
      "num_tokens": 828395550.0,
      "step": 25956
    },
    {
      "entropy": 0.459914137609303,
      "epoch": 2.388174881110517,
      "grad_norm": 0.14749710261821747,
      "learning_rate": 2.0397460667954735e-05,
      "loss": 0.441,
      "mean_token_accuracy": 0.8578957505524158,
      "num_tokens": 828427565.0,
      "step": 25957
    },
    {
      "entropy": 0.4936188291758299,
      "epoch": 2.388266887481958,
      "grad_norm": 0.15920810401439667,
      "learning_rate": 2.039439384181311e-05,
      "loss": 0.4899,
      "mean_token_accuracy": 0.8462695069611073,
      "num_tokens": 828459584.0,
      "step": 25958
    },
    {
      "entropy": 0.5253527651075274,
      "epoch": 2.3883588938533995,
      "grad_norm": 0.16093593835830688,
      "learning_rate": 2.039132701567148e-05,
      "loss": 0.5206,
      "mean_token_accuracy": 0.8351276442408562,
      "num_tokens": 828492231.0,
      "step": 25959
    },
    {
      "entropy": 0.47681043297052383,
      "epoch": 2.3884509002248406,
      "grad_norm": 0.15361493825912476,
      "learning_rate": 2.0388260189529856e-05,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8511020354926586,
      "num_tokens": 828524417.0,
      "step": 25960
    },
    {
      "entropy": 0.43319614813663065,
      "epoch": 2.3885429065962818,
      "grad_norm": 0.1468287855386734,
      "learning_rate": 2.038519336338823e-05,
      "loss": 0.4155,
      "mean_token_accuracy": 0.86770274117589,
      "num_tokens": 828557185.0,
      "step": 25961
    },
    {
      "entropy": 0.5625443886965513,
      "epoch": 2.388634912967723,
      "grad_norm": 0.1716351956129074,
      "learning_rate": 2.0382126537246605e-05,
      "loss": 0.5518,
      "mean_token_accuracy": 0.8264727778732777,
      "num_tokens": 828588789.0,
      "step": 25962
    },
    {
      "entropy": 0.5238080332055688,
      "epoch": 2.388726919339164,
      "grad_norm": 0.16157086193561554,
      "learning_rate": 2.0379059711104977e-05,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8389193415641785,
      "num_tokens": 828620012.0,
      "step": 25963
    },
    {
      "entropy": 0.6511806193739176,
      "epoch": 2.3888189257106056,
      "grad_norm": 0.18082129955291748,
      "learning_rate": 2.037599288496335e-05,
      "loss": 0.6414,
      "mean_token_accuracy": 0.7999005056917667,
      "num_tokens": 828651808.0,
      "step": 25964
    },
    {
      "entropy": 0.4939250126481056,
      "epoch": 2.3889109320820467,
      "grad_norm": 0.15600308775901794,
      "learning_rate": 2.0372926058821726e-05,
      "loss": 0.478,
      "mean_token_accuracy": 0.8490507453680038,
      "num_tokens": 828683906.0,
      "step": 25965
    },
    {
      "entropy": 0.4829183118417859,
      "epoch": 2.389002938453488,
      "grad_norm": 0.16200979053974152,
      "learning_rate": 2.03698592326801e-05,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8496218919754028,
      "num_tokens": 828715561.0,
      "step": 25966
    },
    {
      "entropy": 0.46850717812776566,
      "epoch": 2.389094944824929,
      "grad_norm": 0.16013586521148682,
      "learning_rate": 2.0366792406538475e-05,
      "loss": 0.4624,
      "mean_token_accuracy": 0.853613842278719,
      "num_tokens": 828747195.0,
      "step": 25967
    },
    {
      "entropy": 0.4238426419906318,
      "epoch": 2.38918695119637,
      "grad_norm": 0.14450369775295258,
      "learning_rate": 2.0363725580396846e-05,
      "loss": 0.4061,
      "mean_token_accuracy": 0.8696884214878082,
      "num_tokens": 828779475.0,
      "step": 25968
    },
    {
      "entropy": 0.5405379133298993,
      "epoch": 2.3892789575678117,
      "grad_norm": 0.16569910943508148,
      "learning_rate": 2.036065875425522e-05,
      "loss": 0.5217,
      "mean_token_accuracy": 0.8358982391655445,
      "num_tokens": 828811866.0,
      "step": 25969
    },
    {
      "entropy": 0.4457728024572134,
      "epoch": 2.389370963939253,
      "grad_norm": 0.14337733387947083,
      "learning_rate": 2.0357591928113596e-05,
      "loss": 0.4309,
      "mean_token_accuracy": 0.8648624047636986,
      "num_tokens": 828844301.0,
      "step": 25970
    },
    {
      "entropy": 0.5555729288607836,
      "epoch": 2.389462970310694,
      "grad_norm": 0.16829638183116913,
      "learning_rate": 2.035452510197197e-05,
      "loss": 0.5512,
      "mean_token_accuracy": 0.8251438140869141,
      "num_tokens": 828876505.0,
      "step": 25971
    },
    {
      "entropy": 0.4454543311148882,
      "epoch": 2.389554976682135,
      "grad_norm": 0.15433885157108307,
      "learning_rate": 2.035145827583034e-05,
      "loss": 0.4305,
      "mean_token_accuracy": 0.8636614345014095,
      "num_tokens": 828908550.0,
      "step": 25972
    },
    {
      "entropy": 0.5468928655609488,
      "epoch": 2.3896469830535763,
      "grad_norm": 0.1668354570865631,
      "learning_rate": 2.0348391449688716e-05,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8249831311404705,
      "num_tokens": 828940908.0,
      "step": 25973
    },
    {
      "entropy": 0.4693896872922778,
      "epoch": 2.389738989425018,
      "grad_norm": 0.157209575176239,
      "learning_rate": 2.0345324623547094e-05,
      "loss": 0.4595,
      "mean_token_accuracy": 0.855934701859951,
      "num_tokens": 828972803.0,
      "step": 25974
    },
    {
      "entropy": 0.5413063913583755,
      "epoch": 2.389830995796459,
      "grad_norm": 0.16679294407367706,
      "learning_rate": 2.034225779740547e-05,
      "loss": 0.5307,
      "mean_token_accuracy": 0.8343832418322563,
      "num_tokens": 829004880.0,
      "step": 25975
    },
    {
      "entropy": 0.5235282918438315,
      "epoch": 2.3899230021679,
      "grad_norm": 0.15967299044132233,
      "learning_rate": 2.033919097126384e-05,
      "loss": 0.5094,
      "mean_token_accuracy": 0.8419100046157837,
      "num_tokens": 829036908.0,
      "step": 25976
    },
    {
      "entropy": 0.5553384206723422,
      "epoch": 2.3900150085393412,
      "grad_norm": 0.16670413315296173,
      "learning_rate": 2.0336124145122215e-05,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8246643766760826,
      "num_tokens": 829068704.0,
      "step": 25977
    },
    {
      "entropy": 0.5355185903608799,
      "epoch": 2.3901070149107824,
      "grad_norm": 0.16209276020526886,
      "learning_rate": 2.033305731898059e-05,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8301995657384396,
      "num_tokens": 829100997.0,
      "step": 25978
    },
    {
      "entropy": 0.45862957648932934,
      "epoch": 2.390199021282224,
      "grad_norm": 0.14864620566368103,
      "learning_rate": 2.0329990492838964e-05,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8586868606507778,
      "num_tokens": 829133289.0,
      "step": 25979
    },
    {
      "entropy": 0.5359608186408877,
      "epoch": 2.390291027653665,
      "grad_norm": 0.16002649068832397,
      "learning_rate": 2.0326923666697335e-05,
      "loss": 0.5248,
      "mean_token_accuracy": 0.8338039517402649,
      "num_tokens": 829165652.0,
      "step": 25980
    },
    {
      "entropy": 0.48572508222423494,
      "epoch": 2.390383034025106,
      "grad_norm": 0.15606780350208282,
      "learning_rate": 2.032385684055571e-05,
      "loss": 0.4803,
      "mean_token_accuracy": 0.8507938347756863,
      "num_tokens": 829197164.0,
      "step": 25981
    },
    {
      "entropy": 0.514821095392108,
      "epoch": 2.3904750403965473,
      "grad_norm": 0.16127437353134155,
      "learning_rate": 2.0320790014414085e-05,
      "loss": 0.505,
      "mean_token_accuracy": 0.8415593914687634,
      "num_tokens": 829229066.0,
      "step": 25982
    },
    {
      "entropy": 0.389470461755991,
      "epoch": 2.3905670467679885,
      "grad_norm": 0.13980945944786072,
      "learning_rate": 2.031772318827246e-05,
      "loss": 0.3688,
      "mean_token_accuracy": 0.8796819224953651,
      "num_tokens": 829261197.0,
      "step": 25983
    },
    {
      "entropy": 0.5615594908595085,
      "epoch": 2.39065905313943,
      "grad_norm": 0.16774606704711914,
      "learning_rate": 2.031465636213083e-05,
      "loss": 0.5513,
      "mean_token_accuracy": 0.8229242525994778,
      "num_tokens": 829292680.0,
      "step": 25984
    },
    {
      "entropy": 0.4209254572633654,
      "epoch": 2.390751059510871,
      "grad_norm": 0.1478249877691269,
      "learning_rate": 2.0311589535989205e-05,
      "loss": 0.4089,
      "mean_token_accuracy": 0.8707721531391144,
      "num_tokens": 829325373.0,
      "step": 25985
    },
    {
      "entropy": 0.5015450706705451,
      "epoch": 2.3908430658823123,
      "grad_norm": 0.15949799120426178,
      "learning_rate": 2.030852270984758e-05,
      "loss": 0.484,
      "mean_token_accuracy": 0.8452342599630356,
      "num_tokens": 829356944.0,
      "step": 25986
    },
    {
      "entropy": 0.5119337262585759,
      "epoch": 2.3909350722537535,
      "grad_norm": 0.15163353085517883,
      "learning_rate": 2.0305455883705955e-05,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8452849313616753,
      "num_tokens": 829389514.0,
      "step": 25987
    },
    {
      "entropy": 0.5045875515788794,
      "epoch": 2.3910270786251946,
      "grad_norm": 0.15627942979335785,
      "learning_rate": 2.030238905756433e-05,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8434302359819412,
      "num_tokens": 829421215.0,
      "step": 25988
    },
    {
      "entropy": 0.5984203098341823,
      "epoch": 2.391119084996636,
      "grad_norm": 0.16595697402954102,
      "learning_rate": 2.02993222314227e-05,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8175735026597977,
      "num_tokens": 829453733.0,
      "step": 25989
    },
    {
      "entropy": 0.3967403012793511,
      "epoch": 2.3912110913680773,
      "grad_norm": 0.15017975866794586,
      "learning_rate": 2.0296255405281075e-05,
      "loss": 0.3891,
      "mean_token_accuracy": 0.8715414255857468,
      "num_tokens": 829486138.0,
      "step": 25990
    },
    {
      "entropy": 0.5349061954766512,
      "epoch": 2.3913030977395184,
      "grad_norm": 0.16029107570648193,
      "learning_rate": 2.029318857913945e-05,
      "loss": 0.5126,
      "mean_token_accuracy": 0.8351424336433411,
      "num_tokens": 829518101.0,
      "step": 25991
    },
    {
      "entropy": 0.5353635335341096,
      "epoch": 2.3913951041109596,
      "grad_norm": 0.1637115627527237,
      "learning_rate": 2.0290121752997824e-05,
      "loss": 0.5126,
      "mean_token_accuracy": 0.8371891602873802,
      "num_tokens": 829549765.0,
      "step": 25992
    },
    {
      "entropy": 0.5096214506775141,
      "epoch": 2.3914871104824007,
      "grad_norm": 0.1658657342195511,
      "learning_rate": 2.0287054926856196e-05,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8417334407567978,
      "num_tokens": 829580996.0,
      "step": 25993
    },
    {
      "entropy": 0.5203851251862943,
      "epoch": 2.3915791168538423,
      "grad_norm": 0.16244234144687653,
      "learning_rate": 2.028398810071457e-05,
      "loss": 0.5017,
      "mean_token_accuracy": 0.842563945800066,
      "num_tokens": 829612554.0,
      "step": 25994
    },
    {
      "entropy": 0.5352746592834592,
      "epoch": 2.3916711232252834,
      "grad_norm": 0.16451333463191986,
      "learning_rate": 2.0280921274572945e-05,
      "loss": 0.5278,
      "mean_token_accuracy": 0.8288369253277779,
      "num_tokens": 829644192.0,
      "step": 25995
    },
    {
      "entropy": 0.6423499248921871,
      "epoch": 2.3917631295967245,
      "grad_norm": 0.17205002903938293,
      "learning_rate": 2.027785444843132e-05,
      "loss": 0.6297,
      "mean_token_accuracy": 0.8003743104636669,
      "num_tokens": 829676518.0,
      "step": 25996
    },
    {
      "entropy": 0.5219212556257844,
      "epoch": 2.3918551359681657,
      "grad_norm": 0.1623709499835968,
      "learning_rate": 2.0274787622289694e-05,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8404171988368034,
      "num_tokens": 829708722.0,
      "step": 25997
    },
    {
      "entropy": 0.44460027548484504,
      "epoch": 2.391947142339607,
      "grad_norm": 0.1540852189064026,
      "learning_rate": 2.0271720796148065e-05,
      "loss": 0.4419,
      "mean_token_accuracy": 0.8592716827988625,
      "num_tokens": 829741443.0,
      "step": 25998
    },
    {
      "entropy": 0.5686722584068775,
      "epoch": 2.3920391487110484,
      "grad_norm": 0.17085030674934387,
      "learning_rate": 2.026865397000644e-05,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8237248249351978,
      "num_tokens": 829772217.0,
      "step": 25999
    },
    {
      "entropy": 0.6125963572412729,
      "epoch": 2.3921311550824895,
      "grad_norm": 0.17265217006206512,
      "learning_rate": 2.0265587143864815e-05,
      "loss": 0.6069,
      "mean_token_accuracy": 0.8084175996482372,
      "num_tokens": 829804070.0,
      "step": 26000
    },
    {
      "entropy": 0.631151108071208,
      "epoch": 2.3922231614539307,
      "grad_norm": 0.1747327595949173,
      "learning_rate": 2.026252031772319e-05,
      "loss": 0.6192,
      "mean_token_accuracy": 0.8055436126887798,
      "num_tokens": 829835861.0,
      "step": 26001
    },
    {
      "entropy": 0.4415526366792619,
      "epoch": 2.392315167825372,
      "grad_norm": 0.1539824903011322,
      "learning_rate": 2.025945349158156e-05,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8608113937079906,
      "num_tokens": 829868190.0,
      "step": 26002
    },
    {
      "entropy": 0.6312031988054514,
      "epoch": 2.392407174196813,
      "grad_norm": 0.16892504692077637,
      "learning_rate": 2.0256386665439935e-05,
      "loss": 0.6188,
      "mean_token_accuracy": 0.8072929680347443,
      "num_tokens": 829900349.0,
      "step": 26003
    },
    {
      "entropy": 0.41552689159289,
      "epoch": 2.3924991805682545,
      "grad_norm": 0.15337488055229187,
      "learning_rate": 2.025331983929831e-05,
      "loss": 0.405,
      "mean_token_accuracy": 0.8721739165484905,
      "num_tokens": 829932299.0,
      "step": 26004
    },
    {
      "entropy": 0.45775606483221054,
      "epoch": 2.3925911869396956,
      "grad_norm": 0.15683186054229736,
      "learning_rate": 2.0250253013156688e-05,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8584898598492146,
      "num_tokens": 829964399.0,
      "step": 26005
    },
    {
      "entropy": 0.48467146046459675,
      "epoch": 2.3926831933111368,
      "grad_norm": 0.1550440788269043,
      "learning_rate": 2.024718618701506e-05,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8477933444082737,
      "num_tokens": 829996585.0,
      "step": 26006
    },
    {
      "entropy": 0.4871612377464771,
      "epoch": 2.392775199682578,
      "grad_norm": 0.15604333579540253,
      "learning_rate": 2.0244119360873434e-05,
      "loss": 0.4785,
      "mean_token_accuracy": 0.8451092578470707,
      "num_tokens": 830029062.0,
      "step": 26007
    },
    {
      "entropy": 0.5411425498314202,
      "epoch": 2.392867206054019,
      "grad_norm": 0.1585625559091568,
      "learning_rate": 2.024105253473181e-05,
      "loss": 0.5168,
      "mean_token_accuracy": 0.8345232643187046,
      "num_tokens": 830061010.0,
      "step": 26008
    },
    {
      "entropy": 0.6369503866881132,
      "epoch": 2.3929592124254606,
      "grad_norm": 0.17015612125396729,
      "learning_rate": 2.0237985708590183e-05,
      "loss": 0.62,
      "mean_token_accuracy": 0.8057003878057003,
      "num_tokens": 830093234.0,
      "step": 26009
    },
    {
      "entropy": 0.564261592226103,
      "epoch": 2.3930512187969017,
      "grad_norm": 0.16829314827919006,
      "learning_rate": 2.0234918882448554e-05,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8266493417322636,
      "num_tokens": 830125416.0,
      "step": 26010
    },
    {
      "entropy": 0.4938578298315406,
      "epoch": 2.393143225168343,
      "grad_norm": 0.15025709569454193,
      "learning_rate": 2.023185205630693e-05,
      "loss": 0.4816,
      "mean_token_accuracy": 0.844615675508976,
      "num_tokens": 830158029.0,
      "step": 26011
    },
    {
      "entropy": 0.6126210764050484,
      "epoch": 2.393235231539784,
      "grad_norm": 0.16685135662555695,
      "learning_rate": 2.0228785230165304e-05,
      "loss": 0.594,
      "mean_token_accuracy": 0.8149754293262959,
      "num_tokens": 830190419.0,
      "step": 26012
    },
    {
      "entropy": 0.5106895379722118,
      "epoch": 2.393327237911225,
      "grad_norm": 0.1553761065006256,
      "learning_rate": 2.022571840402368e-05,
      "loss": 0.4982,
      "mean_token_accuracy": 0.8458468578755856,
      "num_tokens": 830222780.0,
      "step": 26013
    },
    {
      "entropy": 0.3954031402245164,
      "epoch": 2.3934192442826667,
      "grad_norm": 0.14773251116275787,
      "learning_rate": 2.022265157788205e-05,
      "loss": 0.3853,
      "mean_token_accuracy": 0.8756941705942154,
      "num_tokens": 830254430.0,
      "step": 26014
    },
    {
      "entropy": 0.5372840855270624,
      "epoch": 2.393511250654108,
      "grad_norm": 0.16037417948246002,
      "learning_rate": 2.0219584751740424e-05,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8333330415189266,
      "num_tokens": 830286069.0,
      "step": 26015
    },
    {
      "entropy": 0.5037840246222913,
      "epoch": 2.393603257025549,
      "grad_norm": 0.1600358784198761,
      "learning_rate": 2.02165179255988e-05,
      "loss": 0.4915,
      "mean_token_accuracy": 0.8460173606872559,
      "num_tokens": 830317786.0,
      "step": 26016
    },
    {
      "entropy": 0.6054597403854132,
      "epoch": 2.39369526339699,
      "grad_norm": 0.1694025993347168,
      "learning_rate": 2.0213451099457174e-05,
      "loss": 0.59,
      "mean_token_accuracy": 0.8158861622214317,
      "num_tokens": 830350076.0,
      "step": 26017
    },
    {
      "entropy": 0.5644578342325985,
      "epoch": 2.3937872697684313,
      "grad_norm": 0.16162483394145966,
      "learning_rate": 2.0210384273315548e-05,
      "loss": 0.5496,
      "mean_token_accuracy": 0.8278141617774963,
      "num_tokens": 830382721.0,
      "step": 26018
    },
    {
      "entropy": 0.5219134036451578,
      "epoch": 2.393879276139873,
      "grad_norm": 0.16227301955223083,
      "learning_rate": 2.020731744717392e-05,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8411712422966957,
      "num_tokens": 830414894.0,
      "step": 26019
    },
    {
      "entropy": 0.4884016253054142,
      "epoch": 2.393971282511314,
      "grad_norm": 0.1604190319776535,
      "learning_rate": 2.0204250621032294e-05,
      "loss": 0.48,
      "mean_token_accuracy": 0.8480271100997925,
      "num_tokens": 830447266.0,
      "step": 26020
    },
    {
      "entropy": 0.49672401137650013,
      "epoch": 2.394063288882755,
      "grad_norm": 0.15787571668624878,
      "learning_rate": 2.020118379489067e-05,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8441519923508167,
      "num_tokens": 830479373.0,
      "step": 26021
    },
    {
      "entropy": 0.4772544445004314,
      "epoch": 2.3941552952541962,
      "grad_norm": 0.15719541907310486,
      "learning_rate": 2.0198116968749043e-05,
      "loss": 0.4681,
      "mean_token_accuracy": 0.849816057831049,
      "num_tokens": 830510975.0,
      "step": 26022
    },
    {
      "entropy": 0.42917713103815913,
      "epoch": 2.3942473016256374,
      "grad_norm": 0.14540989696979523,
      "learning_rate": 2.0195050142607415e-05,
      "loss": 0.4236,
      "mean_token_accuracy": 0.8670025281608105,
      "num_tokens": 830543481.0,
      "step": 26023
    },
    {
      "entropy": 0.476001750677824,
      "epoch": 2.394339307997079,
      "grad_norm": 0.16162504255771637,
      "learning_rate": 2.019198331646579e-05,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8480260446667671,
      "num_tokens": 830575515.0,
      "step": 26024
    },
    {
      "entropy": 0.5613339236006141,
      "epoch": 2.39443131436852,
      "grad_norm": 0.1646254062652588,
      "learning_rate": 2.0188916490324164e-05,
      "loss": 0.5401,
      "mean_token_accuracy": 0.8268464803695679,
      "num_tokens": 830606950.0,
      "step": 26025
    },
    {
      "entropy": 0.5033455537632108,
      "epoch": 2.394523320739961,
      "grad_norm": 0.16159473359584808,
      "learning_rate": 2.018584966418254e-05,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8418337926268578,
      "num_tokens": 830638379.0,
      "step": 26026
    },
    {
      "entropy": 0.5357796642929316,
      "epoch": 2.3946153271114023,
      "grad_norm": 0.1678950935602188,
      "learning_rate": 2.0182782838040913e-05,
      "loss": 0.519,
      "mean_token_accuracy": 0.8345823660492897,
      "num_tokens": 830670287.0,
      "step": 26027
    },
    {
      "entropy": 0.5352880721911788,
      "epoch": 2.3947073334828435,
      "grad_norm": 0.16914556920528412,
      "learning_rate": 2.0179716011899285e-05,
      "loss": 0.5262,
      "mean_token_accuracy": 0.8347442746162415,
      "num_tokens": 830701518.0,
      "step": 26028
    },
    {
      "entropy": 0.5537536963820457,
      "epoch": 2.394799339854285,
      "grad_norm": 0.16302576661109924,
      "learning_rate": 2.017664918575766e-05,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8331256695091724,
      "num_tokens": 830733575.0,
      "step": 26029
    },
    {
      "entropy": 0.42856897972524166,
      "epoch": 2.394891346225726,
      "grad_norm": 0.15741080045700073,
      "learning_rate": 2.0173582359616034e-05,
      "loss": 0.4193,
      "mean_token_accuracy": 0.8647625222802162,
      "num_tokens": 830763719.0,
      "step": 26030
    },
    {
      "entropy": 0.6199511354789138,
      "epoch": 2.3949833525971673,
      "grad_norm": 0.1676214188337326,
      "learning_rate": 2.017051553347441e-05,
      "loss": 0.5902,
      "mean_token_accuracy": 0.8135969750583172,
      "num_tokens": 830795100.0,
      "step": 26031
    },
    {
      "entropy": 0.5524960933253169,
      "epoch": 2.3950753589686085,
      "grad_norm": 0.1656792312860489,
      "learning_rate": 2.016744870733278e-05,
      "loss": 0.5404,
      "mean_token_accuracy": 0.8289933279156685,
      "num_tokens": 830826196.0,
      "step": 26032
    },
    {
      "entropy": 0.524589660577476,
      "epoch": 2.3951673653400496,
      "grad_norm": 0.16759037971496582,
      "learning_rate": 2.0164381881191154e-05,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8368616737425327,
      "num_tokens": 830857847.0,
      "step": 26033
    },
    {
      "entropy": 0.49963274598121643,
      "epoch": 2.395259371711491,
      "grad_norm": 0.1541866809129715,
      "learning_rate": 2.016131505504953e-05,
      "loss": 0.491,
      "mean_token_accuracy": 0.8448562957346439,
      "num_tokens": 830889521.0,
      "step": 26034
    },
    {
      "entropy": 0.5411881469190121,
      "epoch": 2.3953513780829323,
      "grad_norm": 0.16285382211208344,
      "learning_rate": 2.0158248228907907e-05,
      "loss": 0.5345,
      "mean_token_accuracy": 0.8328080326318741,
      "num_tokens": 830921455.0,
      "step": 26035
    },
    {
      "entropy": 0.47807369637303054,
      "epoch": 2.3954433844543734,
      "grad_norm": 0.15694136917591095,
      "learning_rate": 2.015518140276628e-05,
      "loss": 0.4815,
      "mean_token_accuracy": 0.8492404222488403,
      "num_tokens": 830953340.0,
      "step": 26036
    },
    {
      "entropy": 0.4745467712637037,
      "epoch": 2.3955353908258146,
      "grad_norm": 0.15193109214305878,
      "learning_rate": 2.0152114576624653e-05,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8540313877165318,
      "num_tokens": 830985839.0,
      "step": 26037
    },
    {
      "entropy": 0.47539366595447063,
      "epoch": 2.3956273971972557,
      "grad_norm": 0.1577988862991333,
      "learning_rate": 2.0149047750483028e-05,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8513781428337097,
      "num_tokens": 831017755.0,
      "step": 26038
    },
    {
      "entropy": 0.638854943215847,
      "epoch": 2.3957194035686973,
      "grad_norm": 0.17587482929229736,
      "learning_rate": 2.0145980924341402e-05,
      "loss": 0.6281,
      "mean_token_accuracy": 0.8009896129369736,
      "num_tokens": 831049808.0,
      "step": 26039
    },
    {
      "entropy": 0.5787035822868347,
      "epoch": 2.3958114099401384,
      "grad_norm": 0.16511185467243195,
      "learning_rate": 2.0142914098199774e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8236297704279423,
      "num_tokens": 831082052.0,
      "step": 26040
    },
    {
      "entropy": 0.4850322543643415,
      "epoch": 2.3959034163115795,
      "grad_norm": 0.15112094581127167,
      "learning_rate": 2.0139847272058148e-05,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8500274196267128,
      "num_tokens": 831113662.0,
      "step": 26041
    },
    {
      "entropy": 0.5151265431195498,
      "epoch": 2.3959954226830207,
      "grad_norm": 0.15440376102924347,
      "learning_rate": 2.0136780445916523e-05,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8400571011006832,
      "num_tokens": 831146294.0,
      "step": 26042
    },
    {
      "entropy": 0.5492872213944793,
      "epoch": 2.396087429054462,
      "grad_norm": 0.15888239443302155,
      "learning_rate": 2.0133713619774897e-05,
      "loss": 0.5291,
      "mean_token_accuracy": 0.8325620889663696,
      "num_tokens": 831178096.0,
      "step": 26043
    },
    {
      "entropy": 0.4910870515741408,
      "epoch": 2.3961794354259034,
      "grad_norm": 0.1558455228805542,
      "learning_rate": 2.013064679363327e-05,
      "loss": 0.4785,
      "mean_token_accuracy": 0.8485259339213371,
      "num_tokens": 831210016.0,
      "step": 26044
    },
    {
      "entropy": 0.5429586675018072,
      "epoch": 2.3962714417973445,
      "grad_norm": 0.16100266575813293,
      "learning_rate": 2.0127579967491643e-05,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8338323533535004,
      "num_tokens": 831241680.0,
      "step": 26045
    },
    {
      "entropy": 0.4339414285495877,
      "epoch": 2.3963634481687857,
      "grad_norm": 0.1477964073419571,
      "learning_rate": 2.0124513141350018e-05,
      "loss": 0.41,
      "mean_token_accuracy": 0.8669713661074638,
      "num_tokens": 831273785.0,
      "step": 26046
    },
    {
      "entropy": 0.5567432390525937,
      "epoch": 2.396455454540227,
      "grad_norm": 0.16522106528282166,
      "learning_rate": 2.0121446315208393e-05,
      "loss": 0.5415,
      "mean_token_accuracy": 0.826456680893898,
      "num_tokens": 831305413.0,
      "step": 26047
    },
    {
      "entropy": 0.5010709688067436,
      "epoch": 2.396547460911668,
      "grad_norm": 0.15987305343151093,
      "learning_rate": 2.0118379489066767e-05,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8411732241511345,
      "num_tokens": 831338047.0,
      "step": 26048
    },
    {
      "entropy": 0.4972944725304842,
      "epoch": 2.3966394672831095,
      "grad_norm": 0.1550826132297516,
      "learning_rate": 2.011531266292514e-05,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8495016731321812,
      "num_tokens": 831370181.0,
      "step": 26049
    },
    {
      "entropy": 0.6000113915652037,
      "epoch": 2.3967314736545506,
      "grad_norm": 0.17149531841278076,
      "learning_rate": 2.0112245836783513e-05,
      "loss": 0.6045,
      "mean_token_accuracy": 0.8102115504443645,
      "num_tokens": 831402457.0,
      "step": 26050
    },
    {
      "entropy": 0.582507386803627,
      "epoch": 2.3968234800259918,
      "grad_norm": 0.16604025661945343,
      "learning_rate": 2.0109179010641888e-05,
      "loss": 0.576,
      "mean_token_accuracy": 0.8197656609117985,
      "num_tokens": 831433862.0,
      "step": 26051
    },
    {
      "entropy": 0.5095523027703166,
      "epoch": 2.396915486397433,
      "grad_norm": 0.1601039469242096,
      "learning_rate": 2.0106112184500263e-05,
      "loss": 0.4978,
      "mean_token_accuracy": 0.8420360162854195,
      "num_tokens": 831466297.0,
      "step": 26052
    },
    {
      "entropy": 0.46695517376065254,
      "epoch": 2.397007492768874,
      "grad_norm": 0.15426020324230194,
      "learning_rate": 2.0103045358358634e-05,
      "loss": 0.461,
      "mean_token_accuracy": 0.852199163287878,
      "num_tokens": 831498717.0,
      "step": 26053
    },
    {
      "entropy": 0.5336021007969975,
      "epoch": 2.3970994991403156,
      "grad_norm": 0.15991365909576416,
      "learning_rate": 2.009997853221701e-05,
      "loss": 0.5232,
      "mean_token_accuracy": 0.8353327624499798,
      "num_tokens": 831530335.0,
      "step": 26054
    },
    {
      "entropy": 0.435227551497519,
      "epoch": 2.3971915055117567,
      "grad_norm": 0.14951059222221375,
      "learning_rate": 2.0096911706075383e-05,
      "loss": 0.421,
      "mean_token_accuracy": 0.866148978471756,
      "num_tokens": 831562407.0,
      "step": 26055
    },
    {
      "entropy": 0.5544738452881575,
      "epoch": 2.397283511883198,
      "grad_norm": 0.16901527345180511,
      "learning_rate": 2.0093844879933758e-05,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8287201970815659,
      "num_tokens": 831593548.0,
      "step": 26056
    },
    {
      "entropy": 0.6090597701258957,
      "epoch": 2.397375518254639,
      "grad_norm": 0.16974718868732452,
      "learning_rate": 2.0090778053792132e-05,
      "loss": 0.5971,
      "mean_token_accuracy": 0.814228992909193,
      "num_tokens": 831625485.0,
      "step": 26057
    },
    {
      "entropy": 0.602286834269762,
      "epoch": 2.39746752462608,
      "grad_norm": 0.17154717445373535,
      "learning_rate": 2.0087711227650504e-05,
      "loss": 0.5961,
      "mean_token_accuracy": 0.8139681443572044,
      "num_tokens": 831657851.0,
      "step": 26058
    },
    {
      "entropy": 0.5114651108160615,
      "epoch": 2.3975595309975217,
      "grad_norm": 0.15650378167629242,
      "learning_rate": 2.0084644401508878e-05,
      "loss": 0.475,
      "mean_token_accuracy": 0.8454479612410069,
      "num_tokens": 831688811.0,
      "step": 26059
    },
    {
      "entropy": 0.44123004481662065,
      "epoch": 2.397651537368963,
      "grad_norm": 0.14725413918495178,
      "learning_rate": 2.0081577575367253e-05,
      "loss": 0.4178,
      "mean_token_accuracy": 0.8650633059442043,
      "num_tokens": 831720559.0,
      "step": 26060
    },
    {
      "entropy": 0.48839156702160835,
      "epoch": 2.397743543740404,
      "grad_norm": 0.15455183386802673,
      "learning_rate": 2.0078510749225628e-05,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8496154807507992,
      "num_tokens": 831752794.0,
      "step": 26061
    },
    {
      "entropy": 0.5635333973914385,
      "epoch": 2.397835550111845,
      "grad_norm": 0.16796444356441498,
      "learning_rate": 2.0075443923084e-05,
      "loss": 0.5503,
      "mean_token_accuracy": 0.826041653752327,
      "num_tokens": 831784155.0,
      "step": 26062
    },
    {
      "entropy": 0.5845639547333121,
      "epoch": 2.3979275564832863,
      "grad_norm": 0.16981001198291779,
      "learning_rate": 2.0072377096942374e-05,
      "loss": 0.5798,
      "mean_token_accuracy": 0.8187268003821373,
      "num_tokens": 831815231.0,
      "step": 26063
    },
    {
      "entropy": 0.4572769245132804,
      "epoch": 2.398019562854728,
      "grad_norm": 0.1607714295387268,
      "learning_rate": 2.0069310270800748e-05,
      "loss": 0.457,
      "mean_token_accuracy": 0.8573579043149948,
      "num_tokens": 831847024.0,
      "step": 26064
    },
    {
      "entropy": 0.515961455181241,
      "epoch": 2.398111569226169,
      "grad_norm": 0.1711684912443161,
      "learning_rate": 2.0066243444659123e-05,
      "loss": 0.5008,
      "mean_token_accuracy": 0.8393827266991138,
      "num_tokens": 831877338.0,
      "step": 26065
    },
    {
      "entropy": 0.38748012529686093,
      "epoch": 2.39820357559761,
      "grad_norm": 0.14191259443759918,
      "learning_rate": 2.0063176618517497e-05,
      "loss": 0.3679,
      "mean_token_accuracy": 0.8788479715585709,
      "num_tokens": 831909669.0,
      "step": 26066
    },
    {
      "entropy": 0.4345169821754098,
      "epoch": 2.3982955819690512,
      "grad_norm": 0.15276393294334412,
      "learning_rate": 2.0060109792375872e-05,
      "loss": 0.4187,
      "mean_token_accuracy": 0.8650760762393475,
      "num_tokens": 831941807.0,
      "step": 26067
    },
    {
      "entropy": 0.4727937269490212,
      "epoch": 2.3983875883404924,
      "grad_norm": 0.1570163518190384,
      "learning_rate": 2.0057042966234247e-05,
      "loss": 0.468,
      "mean_token_accuracy": 0.8527010269463062,
      "num_tokens": 831973914.0,
      "step": 26068
    },
    {
      "entropy": 0.5157396048307419,
      "epoch": 2.398479594711934,
      "grad_norm": 0.16414761543273926,
      "learning_rate": 2.005397614009262e-05,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8370530717074871,
      "num_tokens": 832004825.0,
      "step": 26069
    },
    {
      "entropy": 0.5410838504321873,
      "epoch": 2.398571601083375,
      "grad_norm": 0.15947595238685608,
      "learning_rate": 2.0050909313950993e-05,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8358268141746521,
      "num_tokens": 832037593.0,
      "step": 26070
    },
    {
      "entropy": 0.6262516146525741,
      "epoch": 2.398663607454816,
      "grad_norm": 0.17468278110027313,
      "learning_rate": 2.0047842487809367e-05,
      "loss": 0.6236,
      "mean_token_accuracy": 0.8077353835105896,
      "num_tokens": 832070017.0,
      "step": 26071
    },
    {
      "entropy": 0.594214990735054,
      "epoch": 2.3987556138262573,
      "grad_norm": 0.16980233788490295,
      "learning_rate": 2.0044775661667742e-05,
      "loss": 0.5849,
      "mean_token_accuracy": 0.8172344602644444,
      "num_tokens": 832101996.0,
      "step": 26072
    },
    {
      "entropy": 0.5990521935746074,
      "epoch": 2.3988476201976985,
      "grad_norm": 0.17134879529476166,
      "learning_rate": 2.0041708835526117e-05,
      "loss": 0.5862,
      "mean_token_accuracy": 0.8159698359668255,
      "num_tokens": 832134096.0,
      "step": 26073
    },
    {
      "entropy": 0.5055799977853894,
      "epoch": 2.39893962656914,
      "grad_norm": 0.156584233045578,
      "learning_rate": 2.0038642009384488e-05,
      "loss": 0.4916,
      "mean_token_accuracy": 0.847769882529974,
      "num_tokens": 832166675.0,
      "step": 26074
    },
    {
      "entropy": 0.47438438422977924,
      "epoch": 2.399031632940581,
      "grad_norm": 0.15551209449768066,
      "learning_rate": 2.0035575183242863e-05,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8521674014627934,
      "num_tokens": 832198554.0,
      "step": 26075
    },
    {
      "entropy": 0.4772354420274496,
      "epoch": 2.3991236393120223,
      "grad_norm": 0.15391185879707336,
      "learning_rate": 2.0032508357101237e-05,
      "loss": 0.4668,
      "mean_token_accuracy": 0.8517738804221153,
      "num_tokens": 832230242.0,
      "step": 26076
    },
    {
      "entropy": 0.5492698438465595,
      "epoch": 2.3992156456834635,
      "grad_norm": 0.16723167896270752,
      "learning_rate": 2.0029441530959612e-05,
      "loss": 0.5389,
      "mean_token_accuracy": 0.8271316811442375,
      "num_tokens": 832262550.0,
      "step": 26077
    },
    {
      "entropy": 0.4909756500273943,
      "epoch": 2.3993076520549046,
      "grad_norm": 0.15995779633522034,
      "learning_rate": 2.0026374704817986e-05,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8455169722437859,
      "num_tokens": 832294646.0,
      "step": 26078
    },
    {
      "entropy": 0.4712241441011429,
      "epoch": 2.399399658426346,
      "grad_norm": 0.15777873992919922,
      "learning_rate": 2.0023307878676358e-05,
      "loss": 0.461,
      "mean_token_accuracy": 0.8531881980597973,
      "num_tokens": 832325830.0,
      "step": 26079
    },
    {
      "entropy": 0.5320056870114058,
      "epoch": 2.3994916647977873,
      "grad_norm": 0.1623464673757553,
      "learning_rate": 2.0020241052534732e-05,
      "loss": 0.511,
      "mean_token_accuracy": 0.8359527960419655,
      "num_tokens": 832358075.0,
      "step": 26080
    },
    {
      "entropy": 0.5347991976886988,
      "epoch": 2.3995836711692284,
      "grad_norm": 0.16618967056274414,
      "learning_rate": 2.0017174226393107e-05,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8344014324247837,
      "num_tokens": 832389188.0,
      "step": 26081
    },
    {
      "entropy": 0.5000804769806564,
      "epoch": 2.3996756775406696,
      "grad_norm": 0.16082650423049927,
      "learning_rate": 2.001410740025148e-05,
      "loss": 0.4938,
      "mean_token_accuracy": 0.8428866043686867,
      "num_tokens": 832421126.0,
      "step": 26082
    },
    {
      "entropy": 0.5239094155840576,
      "epoch": 2.3997676839121107,
      "grad_norm": 0.16319337487220764,
      "learning_rate": 2.0011040574109853e-05,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8374990187585354,
      "num_tokens": 832452785.0,
      "step": 26083
    },
    {
      "entropy": 0.4970861207693815,
      "epoch": 2.3998596902835523,
      "grad_norm": 0.15907536447048187,
      "learning_rate": 2.0007973747968228e-05,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8465884141623974,
      "num_tokens": 832484017.0,
      "step": 26084
    },
    {
      "entropy": 0.5909769013524055,
      "epoch": 2.3999516966549934,
      "grad_norm": 0.1674467772245407,
      "learning_rate": 2.0004906921826602e-05,
      "loss": 0.5699,
      "mean_token_accuracy": 0.8224213160574436,
      "num_tokens": 832516097.0,
      "step": 26085
    },
    {
      "entropy": 0.5012580053880811,
      "epoch": 2.4000437030264345,
      "grad_norm": 0.16471418738365173,
      "learning_rate": 2.0001840095684977e-05,
      "loss": 0.4971,
      "mean_token_accuracy": 0.8449340686202049,
      "num_tokens": 832547951.0,
      "step": 26086
    },
    {
      "entropy": 0.44026640336960554,
      "epoch": 2.4001357093978757,
      "grad_norm": 0.1512013077735901,
      "learning_rate": 1.999877326954335e-05,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8637932389974594,
      "num_tokens": 832579869.0,
      "step": 26087
    },
    {
      "entropy": 0.6022867802530527,
      "epoch": 2.400227715769317,
      "grad_norm": 0.17311398684978485,
      "learning_rate": 1.9995706443401723e-05,
      "loss": 0.6024,
      "mean_token_accuracy": 0.8110047616064548,
      "num_tokens": 832611696.0,
      "step": 26088
    },
    {
      "entropy": 0.5694153411313891,
      "epoch": 2.4003197221407584,
      "grad_norm": 0.16551218926906586,
      "learning_rate": 1.9992639617260097e-05,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8224677927792072,
      "num_tokens": 832643784.0,
      "step": 26089
    },
    {
      "entropy": 0.5503042470663786,
      "epoch": 2.4004117285121995,
      "grad_norm": 0.16678497195243835,
      "learning_rate": 1.9989572791118472e-05,
      "loss": 0.5448,
      "mean_token_accuracy": 0.829307921230793,
      "num_tokens": 832676107.0,
      "step": 26090
    },
    {
      "entropy": 0.5087526831775904,
      "epoch": 2.4005037348836407,
      "grad_norm": 0.15966349840164185,
      "learning_rate": 1.9986505964976847e-05,
      "loss": 0.4928,
      "mean_token_accuracy": 0.8428131490945816,
      "num_tokens": 832708353.0,
      "step": 26091
    },
    {
      "entropy": 0.5795220043510199,
      "epoch": 2.400595741255082,
      "grad_norm": 0.1752215176820755,
      "learning_rate": 1.9983439138835218e-05,
      "loss": 0.581,
      "mean_token_accuracy": 0.8173481896519661,
      "num_tokens": 832740599.0,
      "step": 26092
    },
    {
      "entropy": 0.5285606784746051,
      "epoch": 2.400687747626523,
      "grad_norm": 0.15765516459941864,
      "learning_rate": 1.9980372312693593e-05,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8392071351408958,
      "num_tokens": 832772480.0,
      "step": 26093
    },
    {
      "entropy": 0.4194494115654379,
      "epoch": 2.4007797539979645,
      "grad_norm": 0.14045113325119019,
      "learning_rate": 1.9977305486551967e-05,
      "loss": 0.4012,
      "mean_token_accuracy": 0.8710881657898426,
      "num_tokens": 832804580.0,
      "step": 26094
    },
    {
      "entropy": 0.5413951482623816,
      "epoch": 2.4008717603694056,
      "grad_norm": 0.16552762687206268,
      "learning_rate": 1.9974238660410342e-05,
      "loss": 0.5294,
      "mean_token_accuracy": 0.8332085013389587,
      "num_tokens": 832835749.0,
      "step": 26095
    },
    {
      "entropy": 0.49575994070619345,
      "epoch": 2.4009637667408468,
      "grad_norm": 0.1549571305513382,
      "learning_rate": 1.9971171834268717e-05,
      "loss": 0.4782,
      "mean_token_accuracy": 0.84889742359519,
      "num_tokens": 832868062.0,
      "step": 26096
    },
    {
      "entropy": 0.4575978491920978,
      "epoch": 2.401055773112288,
      "grad_norm": 0.15514639019966125,
      "learning_rate": 1.996810500812709e-05,
      "loss": 0.4555,
      "mean_token_accuracy": 0.855836670845747,
      "num_tokens": 832900104.0,
      "step": 26097
    },
    {
      "entropy": 0.5550529574975371,
      "epoch": 2.401147779483729,
      "grad_norm": 0.16236792504787445,
      "learning_rate": 1.9965038181985466e-05,
      "loss": 0.5425,
      "mean_token_accuracy": 0.8275897316634655,
      "num_tokens": 832932662.0,
      "step": 26098
    },
    {
      "entropy": 0.5399184301495552,
      "epoch": 2.4012397858551706,
      "grad_norm": 0.16672390699386597,
      "learning_rate": 1.996197135584384e-05,
      "loss": 0.5336,
      "mean_token_accuracy": 0.8341385237872601,
      "num_tokens": 832965177.0,
      "step": 26099
    },
    {
      "entropy": 0.4806796833872795,
      "epoch": 2.4013317922266117,
      "grad_norm": 0.16162320971488953,
      "learning_rate": 1.9958904529702212e-05,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8509945310652256,
      "num_tokens": 832996419.0,
      "step": 26100
    },
    {
      "entropy": 0.47941113635897636,
      "epoch": 2.401423798598053,
      "grad_norm": 0.1553298383951187,
      "learning_rate": 1.9955837703560586e-05,
      "loss": 0.4668,
      "mean_token_accuracy": 0.850041076540947,
      "num_tokens": 833028455.0,
      "step": 26101
    },
    {
      "entropy": 0.5035446118563414,
      "epoch": 2.401515804969494,
      "grad_norm": 0.15632563829421997,
      "learning_rate": 1.995277087741896e-05,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8473674319684505,
      "num_tokens": 833059959.0,
      "step": 26102
    },
    {
      "entropy": 0.5461089480668306,
      "epoch": 2.401607811340935,
      "grad_norm": 0.16495907306671143,
      "learning_rate": 1.9949704051277336e-05,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8287323564291,
      "num_tokens": 833091401.0,
      "step": 26103
    },
    {
      "entropy": 0.5941068064421415,
      "epoch": 2.4016998177123767,
      "grad_norm": 0.1709076315164566,
      "learning_rate": 1.994663722513571e-05,
      "loss": 0.589,
      "mean_token_accuracy": 0.8160570599138737,
      "num_tokens": 833123717.0,
      "step": 26104
    },
    {
      "entropy": 0.6024346249178052,
      "epoch": 2.401791824083818,
      "grad_norm": 0.169118270277977,
      "learning_rate": 1.994357039899408e-05,
      "loss": 0.5913,
      "mean_token_accuracy": 0.8134704791009426,
      "num_tokens": 833155720.0,
      "step": 26105
    },
    {
      "entropy": 0.4916317854076624,
      "epoch": 2.401883830455259,
      "grad_norm": 0.15778358280658722,
      "learning_rate": 1.9940503572852456e-05,
      "loss": 0.4856,
      "mean_token_accuracy": 0.8499465435743332,
      "num_tokens": 833187435.0,
      "step": 26106
    },
    {
      "entropy": 0.5366482578683645,
      "epoch": 2.4019758368267,
      "grad_norm": 0.15745733678340912,
      "learning_rate": 1.993743674671083e-05,
      "loss": 0.5198,
      "mean_token_accuracy": 0.8396179154515266,
      "num_tokens": 833219347.0,
      "step": 26107
    },
    {
      "entropy": 0.4741659788414836,
      "epoch": 2.4020678431981413,
      "grad_norm": 0.15432728826999664,
      "learning_rate": 1.9934369920569206e-05,
      "loss": 0.4717,
      "mean_token_accuracy": 0.850059274584055,
      "num_tokens": 833251285.0,
      "step": 26108
    },
    {
      "entropy": 0.5376030448824167,
      "epoch": 2.402159849569583,
      "grad_norm": 0.16487079858779907,
      "learning_rate": 1.9931303094427577e-05,
      "loss": 0.5207,
      "mean_token_accuracy": 0.8361648432910442,
      "num_tokens": 833282862.0,
      "step": 26109
    },
    {
      "entropy": 0.5739565994590521,
      "epoch": 2.402251855941024,
      "grad_norm": 0.16851800680160522,
      "learning_rate": 1.992823626828595e-05,
      "loss": 0.5742,
      "mean_token_accuracy": 0.8213236592710018,
      "num_tokens": 833315317.0,
      "step": 26110
    },
    {
      "entropy": 0.5439188005402684,
      "epoch": 2.402343862312465,
      "grad_norm": 0.16193464398384094,
      "learning_rate": 1.9925169442144326e-05,
      "loss": 0.5317,
      "mean_token_accuracy": 0.8327164314687252,
      "num_tokens": 833347002.0,
      "step": 26111
    },
    {
      "entropy": 0.4639369649812579,
      "epoch": 2.4024358686839062,
      "grad_norm": 0.15510104596614838,
      "learning_rate": 1.99221026160027e-05,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8557242080569267,
      "num_tokens": 833378581.0,
      "step": 26112
    },
    {
      "entropy": 0.4847737681120634,
      "epoch": 2.4025278750553474,
      "grad_norm": 0.1554267555475235,
      "learning_rate": 1.9919035789861072e-05,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8461725004017353,
      "num_tokens": 833410353.0,
      "step": 26113
    },
    {
      "entropy": 0.5539296194911003,
      "epoch": 2.402619881426789,
      "grad_norm": 0.16433024406433105,
      "learning_rate": 1.9915968963719447e-05,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8298123702406883,
      "num_tokens": 833442056.0,
      "step": 26114
    },
    {
      "entropy": 0.36323997378349304,
      "epoch": 2.40271188779823,
      "grad_norm": 0.13966476917266846,
      "learning_rate": 1.991290213757782e-05,
      "loss": 0.3527,
      "mean_token_accuracy": 0.8868836499750614,
      "num_tokens": 833474264.0,
      "step": 26115
    },
    {
      "entropy": 0.5284060705453157,
      "epoch": 2.402803894169671,
      "grad_norm": 0.16864274442195892,
      "learning_rate": 1.9909835311436196e-05,
      "loss": 0.5311,
      "mean_token_accuracy": 0.830018911510706,
      "num_tokens": 833505948.0,
      "step": 26116
    },
    {
      "entropy": 0.6270012836903334,
      "epoch": 2.4028959005411124,
      "grad_norm": 0.17528985440731049,
      "learning_rate": 1.990676848529457e-05,
      "loss": 0.6145,
      "mean_token_accuracy": 0.8056369461119175,
      "num_tokens": 833537091.0,
      "step": 26117
    },
    {
      "entropy": 0.5678366255015135,
      "epoch": 2.4029879069125535,
      "grad_norm": 0.16650883853435516,
      "learning_rate": 1.9903701659152942e-05,
      "loss": 0.5627,
      "mean_token_accuracy": 0.8218117095530033,
      "num_tokens": 833569493.0,
      "step": 26118
    },
    {
      "entropy": 0.626291430555284,
      "epoch": 2.403079913283995,
      "grad_norm": 0.1736854761838913,
      "learning_rate": 1.9900634833011317e-05,
      "loss": 0.6171,
      "mean_token_accuracy": 0.807959470897913,
      "num_tokens": 833601546.0,
      "step": 26119
    },
    {
      "entropy": 0.5429552933201194,
      "epoch": 2.403171919655436,
      "grad_norm": 0.15873397886753082,
      "learning_rate": 1.989756800686969e-05,
      "loss": 0.522,
      "mean_token_accuracy": 0.8342181071639061,
      "num_tokens": 833633807.0,
      "step": 26120
    },
    {
      "entropy": 0.4651013354305178,
      "epoch": 2.4032639260268773,
      "grad_norm": 0.1522996723651886,
      "learning_rate": 1.9894501180728066e-05,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8560621291399002,
      "num_tokens": 833665199.0,
      "step": 26121
    },
    {
      "entropy": 0.4650922240689397,
      "epoch": 2.4033559323983185,
      "grad_norm": 0.1523481011390686,
      "learning_rate": 1.9891434354586437e-05,
      "loss": 0.4585,
      "mean_token_accuracy": 0.8545737601816654,
      "num_tokens": 833697967.0,
      "step": 26122
    },
    {
      "entropy": 0.4189236257225275,
      "epoch": 2.4034479387697596,
      "grad_norm": 0.14493297040462494,
      "learning_rate": 1.9888367528444812e-05,
      "loss": 0.4104,
      "mean_token_accuracy": 0.8700653202831745,
      "num_tokens": 833730221.0,
      "step": 26123
    },
    {
      "entropy": 0.496920840931125,
      "epoch": 2.403539945141201,
      "grad_norm": 0.1605994999408722,
      "learning_rate": 1.9885300702303186e-05,
      "loss": 0.4929,
      "mean_token_accuracy": 0.8424192108213902,
      "num_tokens": 833762238.0,
      "step": 26124
    },
    {
      "entropy": 0.44246829207986593,
      "epoch": 2.4036319515126423,
      "grad_norm": 0.15260624885559082,
      "learning_rate": 1.988223387616156e-05,
      "loss": 0.4352,
      "mean_token_accuracy": 0.8599189482629299,
      "num_tokens": 833793753.0,
      "step": 26125
    },
    {
      "entropy": 0.5027832518098876,
      "epoch": 2.4037239578840834,
      "grad_norm": 0.15351541340351105,
      "learning_rate": 1.9879167050019936e-05,
      "loss": 0.4803,
      "mean_token_accuracy": 0.8464512266218662,
      "num_tokens": 833825351.0,
      "step": 26126
    },
    {
      "entropy": 0.5465493770316243,
      "epoch": 2.4038159642555246,
      "grad_norm": 0.16686376929283142,
      "learning_rate": 1.987610022387831e-05,
      "loss": 0.5439,
      "mean_token_accuracy": 0.825518611818552,
      "num_tokens": 833857545.0,
      "step": 26127
    },
    {
      "entropy": 0.4970105937682092,
      "epoch": 2.4039079706269657,
      "grad_norm": 0.15429702401161194,
      "learning_rate": 1.9873033397736685e-05,
      "loss": 0.4811,
      "mean_token_accuracy": 0.8487188555300236,
      "num_tokens": 833889873.0,
      "step": 26128
    },
    {
      "entropy": 0.5495276441797614,
      "epoch": 2.4039999769984073,
      "grad_norm": 0.16619472205638885,
      "learning_rate": 1.986996657159506e-05,
      "loss": 0.534,
      "mean_token_accuracy": 0.834750808775425,
      "num_tokens": 833921788.0,
      "step": 26129
    },
    {
      "entropy": 0.4298303509131074,
      "epoch": 2.4040919833698484,
      "grad_norm": 0.1441943198442459,
      "learning_rate": 1.986689974545343e-05,
      "loss": 0.4123,
      "mean_token_accuracy": 0.8696460500359535,
      "num_tokens": 833954442.0,
      "step": 26130
    },
    {
      "entropy": 0.5636512143537402,
      "epoch": 2.4041839897412896,
      "grad_norm": 0.16196200251579285,
      "learning_rate": 1.9863832919311805e-05,
      "loss": 0.552,
      "mean_token_accuracy": 0.8253497071564198,
      "num_tokens": 833986423.0,
      "step": 26131
    },
    {
      "entropy": 0.5076626902446151,
      "epoch": 2.4042759961127307,
      "grad_norm": 0.1628219485282898,
      "learning_rate": 1.986076609317018e-05,
      "loss": 0.4944,
      "mean_token_accuracy": 0.8428264148533344,
      "num_tokens": 834018115.0,
      "step": 26132
    },
    {
      "entropy": 0.5878603615565225,
      "epoch": 2.404368002484172,
      "grad_norm": 0.1691756248474121,
      "learning_rate": 1.9857699267028555e-05,
      "loss": 0.573,
      "mean_token_accuracy": 0.8193567991256714,
      "num_tokens": 834049315.0,
      "step": 26133
    },
    {
      "entropy": 0.5436814215499908,
      "epoch": 2.4044600088556134,
      "grad_norm": 0.1635713279247284,
      "learning_rate": 1.985463244088693e-05,
      "loss": 0.5203,
      "mean_token_accuracy": 0.8322717286646366,
      "num_tokens": 834081443.0,
      "step": 26134
    },
    {
      "entropy": 0.47472541593015194,
      "epoch": 2.4045520152270545,
      "grad_norm": 0.1513386368751526,
      "learning_rate": 1.98515656147453e-05,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8545581512153149,
      "num_tokens": 834113637.0,
      "step": 26135
    },
    {
      "entropy": 0.48759076464921236,
      "epoch": 2.4046440215984957,
      "grad_norm": 0.1574689894914627,
      "learning_rate": 1.9848498788603675e-05,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8483472131192684,
      "num_tokens": 834144860.0,
      "step": 26136
    },
    {
      "entropy": 0.4218065803870559,
      "epoch": 2.404736027969937,
      "grad_norm": 0.1552276313304901,
      "learning_rate": 1.984543196246205e-05,
      "loss": 0.4167,
      "mean_token_accuracy": 0.867742508649826,
      "num_tokens": 834175925.0,
      "step": 26137
    },
    {
      "entropy": 0.436872529797256,
      "epoch": 2.404828034341378,
      "grad_norm": 0.1534586399793625,
      "learning_rate": 1.9842365136320425e-05,
      "loss": 0.4221,
      "mean_token_accuracy": 0.8676673844456673,
      "num_tokens": 834207294.0,
      "step": 26138
    },
    {
      "entropy": 0.47464714432135224,
      "epoch": 2.4049200407128195,
      "grad_norm": 0.155690997838974,
      "learning_rate": 1.9839298310178796e-05,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8528567291796207,
      "num_tokens": 834239017.0,
      "step": 26139
    },
    {
      "entropy": 0.5511893527582288,
      "epoch": 2.4050120470842606,
      "grad_norm": 0.17076759040355682,
      "learning_rate": 1.983623148403717e-05,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8312119916081429,
      "num_tokens": 834270901.0,
      "step": 26140
    },
    {
      "entropy": 0.43868565070442855,
      "epoch": 2.4051040534557018,
      "grad_norm": 0.15244068205356598,
      "learning_rate": 1.9833164657895545e-05,
      "loss": 0.4181,
      "mean_token_accuracy": 0.8628141582012177,
      "num_tokens": 834301310.0,
      "step": 26141
    },
    {
      "entropy": 0.6227294616401196,
      "epoch": 2.405196059827143,
      "grad_norm": 0.17622262239456177,
      "learning_rate": 1.983009783175392e-05,
      "loss": 0.6135,
      "mean_token_accuracy": 0.808188185095787,
      "num_tokens": 834332780.0,
      "step": 26142
    },
    {
      "entropy": 0.5099688521586359,
      "epoch": 2.405288066198584,
      "grad_norm": 0.15865379571914673,
      "learning_rate": 1.982703100561229e-05,
      "loss": 0.4842,
      "mean_token_accuracy": 0.8486400693655014,
      "num_tokens": 834364374.0,
      "step": 26143
    },
    {
      "entropy": 0.5395197607576847,
      "epoch": 2.4053800725700256,
      "grad_norm": 0.16220954060554504,
      "learning_rate": 1.9823964179470666e-05,
      "loss": 0.5277,
      "mean_token_accuracy": 0.8307577706873417,
      "num_tokens": 834396286.0,
      "step": 26144
    },
    {
      "entropy": 0.6877842843532562,
      "epoch": 2.4054720789414668,
      "grad_norm": 0.17906774580478668,
      "learning_rate": 1.982089735332904e-05,
      "loss": 0.6704,
      "mean_token_accuracy": 0.7893979251384735,
      "num_tokens": 834427694.0,
      "step": 26145
    },
    {
      "entropy": 0.5318983281031251,
      "epoch": 2.405564085312908,
      "grad_norm": 0.15839402377605438,
      "learning_rate": 1.9817830527187415e-05,
      "loss": 0.5073,
      "mean_token_accuracy": 0.8388624750077724,
      "num_tokens": 834459434.0,
      "step": 26146
    },
    {
      "entropy": 0.5946611035615206,
      "epoch": 2.405656091684349,
      "grad_norm": 0.17635153234004974,
      "learning_rate": 1.981476370104579e-05,
      "loss": 0.5837,
      "mean_token_accuracy": 0.8179574571549892,
      "num_tokens": 834490452.0,
      "step": 26147
    },
    {
      "entropy": 0.5721626002341509,
      "epoch": 2.40574809805579,
      "grad_norm": 0.16793763637542725,
      "learning_rate": 1.981169687490416e-05,
      "loss": 0.559,
      "mean_token_accuracy": 0.8260501138865948,
      "num_tokens": 834522469.0,
      "step": 26148
    },
    {
      "entropy": 0.5004021525382996,
      "epoch": 2.4058401044272317,
      "grad_norm": 0.15997838973999023,
      "learning_rate": 1.9808630048762536e-05,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8457598946988583,
      "num_tokens": 834554204.0,
      "step": 26149
    },
    {
      "entropy": 0.5237903948873281,
      "epoch": 2.405932110798673,
      "grad_norm": 0.16037040948867798,
      "learning_rate": 1.980556322262091e-05,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8432026617228985,
      "num_tokens": 834585533.0,
      "step": 26150
    },
    {
      "entropy": 0.6057842951267958,
      "epoch": 2.406024117170114,
      "grad_norm": 0.16564974188804626,
      "learning_rate": 1.9802496396479285e-05,
      "loss": 0.6012,
      "mean_token_accuracy": 0.8129435889422894,
      "num_tokens": 834617718.0,
      "step": 26151
    },
    {
      "entropy": 0.4399915081448853,
      "epoch": 2.406116123541555,
      "grad_norm": 0.15008413791656494,
      "learning_rate": 1.9799429570337656e-05,
      "loss": 0.4305,
      "mean_token_accuracy": 0.8626795075833797,
      "num_tokens": 834650101.0,
      "step": 26152
    },
    {
      "entropy": 0.5862985905259848,
      "epoch": 2.4062081299129963,
      "grad_norm": 0.17090533673763275,
      "learning_rate": 1.979636274419603e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8190912716090679,
      "num_tokens": 834681654.0,
      "step": 26153
    },
    {
      "entropy": 0.4691485594958067,
      "epoch": 2.406300136284438,
      "grad_norm": 0.15388737618923187,
      "learning_rate": 1.9793295918054405e-05,
      "loss": 0.4607,
      "mean_token_accuracy": 0.8517224825918674,
      "num_tokens": 834713891.0,
      "step": 26154
    },
    {
      "entropy": 0.49331429693847895,
      "epoch": 2.406392142655879,
      "grad_norm": 0.16268716752529144,
      "learning_rate": 1.979022909191278e-05,
      "loss": 0.4906,
      "mean_token_accuracy": 0.844536304473877,
      "num_tokens": 834745788.0,
      "step": 26155
    },
    {
      "entropy": 0.5126431211829185,
      "epoch": 2.40648414902732,
      "grad_norm": 0.1587390899658203,
      "learning_rate": 1.9787162265771155e-05,
      "loss": 0.5008,
      "mean_token_accuracy": 0.8396257646381855,
      "num_tokens": 834778287.0,
      "step": 26156
    },
    {
      "entropy": 0.4907363534439355,
      "epoch": 2.4065761553987612,
      "grad_norm": 0.1556762009859085,
      "learning_rate": 1.978409543962953e-05,
      "loss": 0.4792,
      "mean_token_accuracy": 0.8444112874567509,
      "num_tokens": 834810716.0,
      "step": 26157
    },
    {
      "entropy": 0.5506733329966664,
      "epoch": 2.4066681617702024,
      "grad_norm": 0.16224651038646698,
      "learning_rate": 1.9781028613487904e-05,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8298027329146862,
      "num_tokens": 834842798.0,
      "step": 26158
    },
    {
      "entropy": 0.5488789435476065,
      "epoch": 2.406760168141644,
      "grad_norm": 0.16026923060417175,
      "learning_rate": 1.977796178734628e-05,
      "loss": 0.5447,
      "mean_token_accuracy": 0.8285291530191898,
      "num_tokens": 834875374.0,
      "step": 26159
    },
    {
      "entropy": 0.5143155134283006,
      "epoch": 2.406852174513085,
      "grad_norm": 0.15929286181926727,
      "learning_rate": 1.977489496120465e-05,
      "loss": 0.5123,
      "mean_token_accuracy": 0.8370455056428909,
      "num_tokens": 834907977.0,
      "step": 26160
    },
    {
      "entropy": 0.6576433531008661,
      "epoch": 2.406944180884526,
      "grad_norm": 0.17253950238227844,
      "learning_rate": 1.9771828135063025e-05,
      "loss": 0.641,
      "mean_token_accuracy": 0.7985675372183323,
      "num_tokens": 834939677.0,
      "step": 26161
    },
    {
      "entropy": 0.5176816359162331,
      "epoch": 2.4070361872559674,
      "grad_norm": 0.15900221467018127,
      "learning_rate": 1.97687613089214e-05,
      "loss": 0.4913,
      "mean_token_accuracy": 0.8420113995671272,
      "num_tokens": 834970339.0,
      "step": 26162
    },
    {
      "entropy": 0.5000947397202253,
      "epoch": 2.4071281936274085,
      "grad_norm": 0.15268085896968842,
      "learning_rate": 1.9765694482779774e-05,
      "loss": 0.4955,
      "mean_token_accuracy": 0.841589018702507,
      "num_tokens": 835002899.0,
      "step": 26163
    },
    {
      "entropy": 0.5173370828852057,
      "epoch": 2.40722019999885,
      "grad_norm": 0.1608669012784958,
      "learning_rate": 1.976262765663815e-05,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8419556468725204,
      "num_tokens": 835035366.0,
      "step": 26164
    },
    {
      "entropy": 0.5583915486931801,
      "epoch": 2.407312206370291,
      "grad_norm": 0.16267837584018707,
      "learning_rate": 1.975956083049652e-05,
      "loss": 0.549,
      "mean_token_accuracy": 0.823723278939724,
      "num_tokens": 835067219.0,
      "step": 26165
    },
    {
      "entropy": 0.5034133521839976,
      "epoch": 2.4074042127417323,
      "grad_norm": 0.1596042960882187,
      "learning_rate": 1.9756494004354894e-05,
      "loss": 0.4938,
      "mean_token_accuracy": 0.8444642052054405,
      "num_tokens": 835099154.0,
      "step": 26166
    },
    {
      "entropy": 0.5700008072890341,
      "epoch": 2.4074962191131735,
      "grad_norm": 0.16875316202640533,
      "learning_rate": 1.975342717821327e-05,
      "loss": 0.5604,
      "mean_token_accuracy": 0.822108656167984,
      "num_tokens": 835130114.0,
      "step": 26167
    },
    {
      "entropy": 0.5504864891991019,
      "epoch": 2.4075882254846146,
      "grad_norm": 0.16675639152526855,
      "learning_rate": 1.9750360352071644e-05,
      "loss": 0.5285,
      "mean_token_accuracy": 0.831296369433403,
      "num_tokens": 835161729.0,
      "step": 26168
    },
    {
      "entropy": 0.5486484654247761,
      "epoch": 2.407680231856056,
      "grad_norm": 0.1598239690065384,
      "learning_rate": 1.9747293525930015e-05,
      "loss": 0.5402,
      "mean_token_accuracy": 0.82812275364995,
      "num_tokens": 835193620.0,
      "step": 26169
    },
    {
      "entropy": 0.4946651067584753,
      "epoch": 2.4077722382274973,
      "grad_norm": 0.16039347648620605,
      "learning_rate": 1.974422669978839e-05,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8481707572937012,
      "num_tokens": 835225332.0,
      "step": 26170
    },
    {
      "entropy": 0.4504698282107711,
      "epoch": 2.4078642445989384,
      "grad_norm": 0.14583466947078705,
      "learning_rate": 1.9741159873646764e-05,
      "loss": 0.4456,
      "mean_token_accuracy": 0.8579606600105762,
      "num_tokens": 835257847.0,
      "step": 26171
    },
    {
      "entropy": 0.6095360778272152,
      "epoch": 2.4079562509703796,
      "grad_norm": 0.1684550791978836,
      "learning_rate": 1.973809304750514e-05,
      "loss": 0.6032,
      "mean_token_accuracy": 0.8112558573484421,
      "num_tokens": 835290024.0,
      "step": 26172
    },
    {
      "entropy": 0.49461911898106337,
      "epoch": 2.4080482573418207,
      "grad_norm": 0.16005763411521912,
      "learning_rate": 1.973502622136351e-05,
      "loss": 0.476,
      "mean_token_accuracy": 0.8483422957360744,
      "num_tokens": 835321847.0,
      "step": 26173
    },
    {
      "entropy": 0.47998157516121864,
      "epoch": 2.4081402637132623,
      "grad_norm": 0.15577924251556396,
      "learning_rate": 1.9731959395221885e-05,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8542527668178082,
      "num_tokens": 835354220.0,
      "step": 26174
    },
    {
      "entropy": 0.5700126104056835,
      "epoch": 2.4082322700847034,
      "grad_norm": 0.16442428529262543,
      "learning_rate": 1.972889256908026e-05,
      "loss": 0.5603,
      "mean_token_accuracy": 0.8235969990491867,
      "num_tokens": 835386126.0,
      "step": 26175
    },
    {
      "entropy": 0.5441691167652607,
      "epoch": 2.4083242764561446,
      "grad_norm": 0.1586681753396988,
      "learning_rate": 1.9725825742938634e-05,
      "loss": 0.5267,
      "mean_token_accuracy": 0.83403554931283,
      "num_tokens": 835417793.0,
      "step": 26176
    },
    {
      "entropy": 0.5227386876940727,
      "epoch": 2.4084162828275857,
      "grad_norm": 0.16068100929260254,
      "learning_rate": 1.972275891679701e-05,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8380877450108528,
      "num_tokens": 835449930.0,
      "step": 26177
    },
    {
      "entropy": 0.5156838037073612,
      "epoch": 2.408508289199027,
      "grad_norm": 0.15904264152050018,
      "learning_rate": 1.971969209065538e-05,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8407746180891991,
      "num_tokens": 835481911.0,
      "step": 26178
    },
    {
      "entropy": 0.5745915872976184,
      "epoch": 2.4086002955704684,
      "grad_norm": 0.17171798646450043,
      "learning_rate": 1.9716625264513755e-05,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8248511105775833,
      "num_tokens": 835513630.0,
      "step": 26179
    },
    {
      "entropy": 0.47537724394351244,
      "epoch": 2.4086923019419095,
      "grad_norm": 0.15335410833358765,
      "learning_rate": 1.971355843837213e-05,
      "loss": 0.46,
      "mean_token_accuracy": 0.8520065881311893,
      "num_tokens": 835546084.0,
      "step": 26180
    },
    {
      "entropy": 0.5490869274362922,
      "epoch": 2.4087843083133507,
      "grad_norm": 0.16218088567256927,
      "learning_rate": 1.9710491612230504e-05,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8301462382078171,
      "num_tokens": 835578442.0,
      "step": 26181
    },
    {
      "entropy": 0.532496533356607,
      "epoch": 2.408876314684792,
      "grad_norm": 0.1585909128189087,
      "learning_rate": 1.9707424786088875e-05,
      "loss": 0.5119,
      "mean_token_accuracy": 0.8368688896298409,
      "num_tokens": 835610684.0,
      "step": 26182
    },
    {
      "entropy": 0.5496211852878332,
      "epoch": 2.408968321056233,
      "grad_norm": 0.16562092304229736,
      "learning_rate": 1.970435795994725e-05,
      "loss": 0.5357,
      "mean_token_accuracy": 0.8300470225512981,
      "num_tokens": 835643111.0,
      "step": 26183
    },
    {
      "entropy": 0.5724078798666596,
      "epoch": 2.4090603274276745,
      "grad_norm": 0.16878551244735718,
      "learning_rate": 1.9701291133805625e-05,
      "loss": 0.5631,
      "mean_token_accuracy": 0.8233716115355492,
      "num_tokens": 835675508.0,
      "step": 26184
    },
    {
      "entropy": 0.5177913485094905,
      "epoch": 2.4091523337991156,
      "grad_norm": 0.1622123122215271,
      "learning_rate": 1.9698224307664e-05,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8366907685995102,
      "num_tokens": 835708067.0,
      "step": 26185
    },
    {
      "entropy": 0.5193694923073053,
      "epoch": 2.4092443401705568,
      "grad_norm": 0.16082893311977386,
      "learning_rate": 1.9695157481522374e-05,
      "loss": 0.5057,
      "mean_token_accuracy": 0.8390460312366486,
      "num_tokens": 835740410.0,
      "step": 26186
    },
    {
      "entropy": 0.5525753665715456,
      "epoch": 2.409336346541998,
      "grad_norm": 0.16915076971054077,
      "learning_rate": 1.9692090655380745e-05,
      "loss": 0.5427,
      "mean_token_accuracy": 0.8278582692146301,
      "num_tokens": 835772428.0,
      "step": 26187
    },
    {
      "entropy": 0.45162942819297314,
      "epoch": 2.409428352913439,
      "grad_norm": 0.1585538387298584,
      "learning_rate": 1.9689023829239123e-05,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8564438000321388,
      "num_tokens": 835804399.0,
      "step": 26188
    },
    {
      "entropy": 0.501739751547575,
      "epoch": 2.4095203592848806,
      "grad_norm": 0.1637091040611267,
      "learning_rate": 1.9685957003097498e-05,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8451672196388245,
      "num_tokens": 835835111.0,
      "step": 26189
    },
    {
      "entropy": 0.44216250255703926,
      "epoch": 2.4096123656563218,
      "grad_norm": 0.15736733376979828,
      "learning_rate": 1.968289017695587e-05,
      "loss": 0.433,
      "mean_token_accuracy": 0.8595384918153286,
      "num_tokens": 835866916.0,
      "step": 26190
    },
    {
      "entropy": 0.551089646993205,
      "epoch": 2.409704372027763,
      "grad_norm": 0.16787447035312653,
      "learning_rate": 1.9679823350814244e-05,
      "loss": 0.5478,
      "mean_token_accuracy": 0.8304443284869194,
      "num_tokens": 835898661.0,
      "step": 26191
    },
    {
      "entropy": 0.5683368127793074,
      "epoch": 2.409796378399204,
      "grad_norm": 0.1675398200750351,
      "learning_rate": 1.967675652467262e-05,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8201028406620026,
      "num_tokens": 835930228.0,
      "step": 26192
    },
    {
      "entropy": 0.566498314961791,
      "epoch": 2.409888384770645,
      "grad_norm": 0.17167754471302032,
      "learning_rate": 1.9673689698530993e-05,
      "loss": 0.57,
      "mean_token_accuracy": 0.8234076611697674,
      "num_tokens": 835962006.0,
      "step": 26193
    },
    {
      "entropy": 0.5258171856403351,
      "epoch": 2.4099803911420867,
      "grad_norm": 0.163818821310997,
      "learning_rate": 1.9670622872389368e-05,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8332011476159096,
      "num_tokens": 835993018.0,
      "step": 26194
    },
    {
      "entropy": 0.4401315273717046,
      "epoch": 2.410072397513528,
      "grad_norm": 0.15250156819820404,
      "learning_rate": 1.966755604624774e-05,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8632313907146454,
      "num_tokens": 836025135.0,
      "step": 26195
    },
    {
      "entropy": 0.666875347495079,
      "epoch": 2.410164403884969,
      "grad_norm": 0.17506107687950134,
      "learning_rate": 1.9664489220106114e-05,
      "loss": 0.6572,
      "mean_token_accuracy": 0.7939229756593704,
      "num_tokens": 836057052.0,
      "step": 26196
    },
    {
      "entropy": 0.5516678839921951,
      "epoch": 2.41025641025641,
      "grad_norm": 0.16834084689617157,
      "learning_rate": 1.9661422393964488e-05,
      "loss": 0.5439,
      "mean_token_accuracy": 0.830030370503664,
      "num_tokens": 836088471.0,
      "step": 26197
    },
    {
      "entropy": 0.622870909050107,
      "epoch": 2.4103484166278513,
      "grad_norm": 0.1728067249059677,
      "learning_rate": 1.9658355567822863e-05,
      "loss": 0.6164,
      "mean_token_accuracy": 0.8089278824627399,
      "num_tokens": 836120396.0,
      "step": 26198
    },
    {
      "entropy": 0.5281568029895425,
      "epoch": 2.410440422999293,
      "grad_norm": 0.16271060705184937,
      "learning_rate": 1.9655288741681234e-05,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8392013944685459,
      "num_tokens": 836152039.0,
      "step": 26199
    },
    {
      "entropy": 0.4584257567767054,
      "epoch": 2.410532429370734,
      "grad_norm": 0.15068168938159943,
      "learning_rate": 1.965222191553961e-05,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8554592095315456,
      "num_tokens": 836184807.0,
      "step": 26200
    },
    {
      "entropy": 0.5838647969067097,
      "epoch": 2.410624435742175,
      "grad_norm": 0.16621986031532288,
      "learning_rate": 1.9649155089397983e-05,
      "loss": 0.552,
      "mean_token_accuracy": 0.8249569609761238,
      "num_tokens": 836216360.0,
      "step": 26201
    },
    {
      "entropy": 0.47963701747357845,
      "epoch": 2.4107164421136162,
      "grad_norm": 0.15339843928813934,
      "learning_rate": 1.9646088263256358e-05,
      "loss": 0.4753,
      "mean_token_accuracy": 0.8487159945070744,
      "num_tokens": 836248730.0,
      "step": 26202
    },
    {
      "entropy": 0.5093176877126098,
      "epoch": 2.4108084484850574,
      "grad_norm": 0.15838785469532013,
      "learning_rate": 1.964302143711473e-05,
      "loss": 0.4908,
      "mean_token_accuracy": 0.8415728993713856,
      "num_tokens": 836279994.0,
      "step": 26203
    },
    {
      "entropy": 0.6068449057638645,
      "epoch": 2.410900454856499,
      "grad_norm": 0.17246170341968536,
      "learning_rate": 1.9639954610973104e-05,
      "loss": 0.5964,
      "mean_token_accuracy": 0.8141794428229332,
      "num_tokens": 836311335.0,
      "step": 26204
    },
    {
      "entropy": 0.5046912579564378,
      "epoch": 2.41099246122794,
      "grad_norm": 0.15430116653442383,
      "learning_rate": 1.963688778483148e-05,
      "loss": 0.4892,
      "mean_token_accuracy": 0.845535159111023,
      "num_tokens": 836343201.0,
      "step": 26205
    },
    {
      "entropy": 0.41667114570736885,
      "epoch": 2.411084467599381,
      "grad_norm": 0.14496630430221558,
      "learning_rate": 1.9633820958689853e-05,
      "loss": 0.4049,
      "mean_token_accuracy": 0.8703527823090553,
      "num_tokens": 836375533.0,
      "step": 26206
    },
    {
      "entropy": 0.5194742735475302,
      "epoch": 2.4111764739708224,
      "grad_norm": 0.15884047746658325,
      "learning_rate": 1.9630754132548228e-05,
      "loss": 0.501,
      "mean_token_accuracy": 0.8421749472618103,
      "num_tokens": 836406969.0,
      "step": 26207
    },
    {
      "entropy": 0.5766948256641626,
      "epoch": 2.4112684803422635,
      "grad_norm": 0.16393762826919556,
      "learning_rate": 1.96276873064066e-05,
      "loss": 0.5562,
      "mean_token_accuracy": 0.8228850141167641,
      "num_tokens": 836439259.0,
      "step": 26208
    },
    {
      "entropy": 0.5128860948607326,
      "epoch": 2.411360486713705,
      "grad_norm": 0.15301668643951416,
      "learning_rate": 1.9624620480264974e-05,
      "loss": 0.4882,
      "mean_token_accuracy": 0.8453052118420601,
      "num_tokens": 836471744.0,
      "step": 26209
    },
    {
      "entropy": 0.5737331677228212,
      "epoch": 2.411452493085146,
      "grad_norm": 0.1653514951467514,
      "learning_rate": 1.962155365412335e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8214019648730755,
      "num_tokens": 836503354.0,
      "step": 26210
    },
    {
      "entropy": 0.4689354752190411,
      "epoch": 2.4115444994565873,
      "grad_norm": 0.14921784400939941,
      "learning_rate": 1.9618486827981723e-05,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8552605472505093,
      "num_tokens": 836535044.0,
      "step": 26211
    },
    {
      "entropy": 0.4962885696440935,
      "epoch": 2.4116365058280285,
      "grad_norm": 0.15298448503017426,
      "learning_rate": 1.9615420001840094e-05,
      "loss": 0.4788,
      "mean_token_accuracy": 0.8470564819872379,
      "num_tokens": 836567112.0,
      "step": 26212
    },
    {
      "entropy": 0.5189344292739406,
      "epoch": 2.4117285121994696,
      "grad_norm": 0.15814140439033508,
      "learning_rate": 1.961235317569847e-05,
      "loss": 0.5116,
      "mean_token_accuracy": 0.8401139415800571,
      "num_tokens": 836599576.0,
      "step": 26213
    },
    {
      "entropy": 0.5883199721574783,
      "epoch": 2.411820518570911,
      "grad_norm": 0.1666446030139923,
      "learning_rate": 1.9609286349556844e-05,
      "loss": 0.5734,
      "mean_token_accuracy": 0.8196572922170162,
      "num_tokens": 836632123.0,
      "step": 26214
    },
    {
      "entropy": 0.4568587576504797,
      "epoch": 2.4119125249423523,
      "grad_norm": 0.1491430252790451,
      "learning_rate": 1.9606219523415218e-05,
      "loss": 0.4556,
      "mean_token_accuracy": 0.856887012720108,
      "num_tokens": 836664316.0,
      "step": 26215
    },
    {
      "entropy": 0.5194780062884092,
      "epoch": 2.4120045313137934,
      "grad_norm": 0.16833798587322235,
      "learning_rate": 1.9603152697273593e-05,
      "loss": 0.5142,
      "mean_token_accuracy": 0.8405493274331093,
      "num_tokens": 836695865.0,
      "step": 26216
    },
    {
      "entropy": 0.51119440398179,
      "epoch": 2.4120965376852346,
      "grad_norm": 0.1624678373336792,
      "learning_rate": 1.9600085871131964e-05,
      "loss": 0.5092,
      "mean_token_accuracy": 0.834086462855339,
      "num_tokens": 836728633.0,
      "step": 26217
    },
    {
      "entropy": 0.5044340533204377,
      "epoch": 2.4121885440566757,
      "grad_norm": 0.15620867908000946,
      "learning_rate": 1.9597019044990342e-05,
      "loss": 0.4969,
      "mean_token_accuracy": 0.84138373285532,
      "num_tokens": 836761401.0,
      "step": 26218
    },
    {
      "entropy": 0.48979128058999777,
      "epoch": 2.4122805504281173,
      "grad_norm": 0.1661367267370224,
      "learning_rate": 1.9593952218848717e-05,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8487119600176811,
      "num_tokens": 836793080.0,
      "step": 26219
    },
    {
      "entropy": 0.47969215526245534,
      "epoch": 2.4123725567995584,
      "grad_norm": 0.15215742588043213,
      "learning_rate": 1.9590885392707088e-05,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8511816598474979,
      "num_tokens": 836825266.0,
      "step": 26220
    },
    {
      "entropy": 0.5266117288265377,
      "epoch": 2.4124645631709996,
      "grad_norm": 0.15478961169719696,
      "learning_rate": 1.9587818566565463e-05,
      "loss": 0.5142,
      "mean_token_accuracy": 0.840375192463398,
      "num_tokens": 836857596.0,
      "step": 26221
    },
    {
      "entropy": 0.4685056898742914,
      "epoch": 2.4125565695424407,
      "grad_norm": 0.15298953652381897,
      "learning_rate": 1.9584751740423837e-05,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8557789288461208,
      "num_tokens": 836888944.0,
      "step": 26222
    },
    {
      "entropy": 0.49530377658084035,
      "epoch": 2.412648575913882,
      "grad_norm": 0.15873616933822632,
      "learning_rate": 1.9581684914282212e-05,
      "loss": 0.4823,
      "mean_token_accuracy": 0.846111610531807,
      "num_tokens": 836921074.0,
      "step": 26223
    },
    {
      "entropy": 0.48755086911842227,
      "epoch": 2.4127405822853234,
      "grad_norm": 0.16097377240657806,
      "learning_rate": 1.9578618088140587e-05,
      "loss": 0.4846,
      "mean_token_accuracy": 0.8461832217872143,
      "num_tokens": 836953374.0,
      "step": 26224
    },
    {
      "entropy": 0.374448636546731,
      "epoch": 2.4128325886567645,
      "grad_norm": 0.14283087849617004,
      "learning_rate": 1.9575551261998958e-05,
      "loss": 0.3548,
      "mean_token_accuracy": 0.8830354921519756,
      "num_tokens": 836984767.0,
      "step": 26225
    },
    {
      "entropy": 0.5134054301306605,
      "epoch": 2.4129245950282057,
      "grad_norm": 0.15771128237247467,
      "learning_rate": 1.9572484435857333e-05,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8412019498646259,
      "num_tokens": 837016962.0,
      "step": 26226
    },
    {
      "entropy": 0.5094224133063108,
      "epoch": 2.413016601399647,
      "grad_norm": 0.16338026523590088,
      "learning_rate": 1.9569417609715707e-05,
      "loss": 0.5121,
      "mean_token_accuracy": 0.8366847932338715,
      "num_tokens": 837048740.0,
      "step": 26227
    },
    {
      "entropy": 0.4351937265601009,
      "epoch": 2.413108607771088,
      "grad_norm": 0.1448437124490738,
      "learning_rate": 1.9566350783574082e-05,
      "loss": 0.4137,
      "mean_token_accuracy": 0.866161897778511,
      "num_tokens": 837080753.0,
      "step": 26228
    },
    {
      "entropy": 0.4966364190913737,
      "epoch": 2.4132006141425295,
      "grad_norm": 0.15721730887889862,
      "learning_rate": 1.9563283957432453e-05,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8491725437343121,
      "num_tokens": 837112097.0,
      "step": 26229
    },
    {
      "entropy": 0.5031116493046284,
      "epoch": 2.4132926205139706,
      "grad_norm": 0.15266487002372742,
      "learning_rate": 1.9560217131290828e-05,
      "loss": 0.4954,
      "mean_token_accuracy": 0.8447023220360279,
      "num_tokens": 837144035.0,
      "step": 26230
    },
    {
      "entropy": 0.5095049720257521,
      "epoch": 2.4133846268854118,
      "grad_norm": 0.15993425250053406,
      "learning_rate": 1.9557150305149202e-05,
      "loss": 0.4909,
      "mean_token_accuracy": 0.8416613079607487,
      "num_tokens": 837175340.0,
      "step": 26231
    },
    {
      "entropy": 0.4963711479213089,
      "epoch": 2.413476633256853,
      "grad_norm": 0.15718436241149902,
      "learning_rate": 1.9554083479007577e-05,
      "loss": 0.469,
      "mean_token_accuracy": 0.8469946458935738,
      "num_tokens": 837206839.0,
      "step": 26232
    },
    {
      "entropy": 0.4963783025741577,
      "epoch": 2.413568639628294,
      "grad_norm": 0.1561092734336853,
      "learning_rate": 1.955101665286595e-05,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8472450375556946,
      "num_tokens": 837239082.0,
      "step": 26233
    },
    {
      "entropy": 0.5451268125325441,
      "epoch": 2.4136606459997356,
      "grad_norm": 0.16140303015708923,
      "learning_rate": 1.9547949826724323e-05,
      "loss": 0.5248,
      "mean_token_accuracy": 0.833341583609581,
      "num_tokens": 837270538.0,
      "step": 26234
    },
    {
      "entropy": 0.4405608088709414,
      "epoch": 2.4137526523711768,
      "grad_norm": 0.14937783777713776,
      "learning_rate": 1.9544883000582698e-05,
      "loss": 0.415,
      "mean_token_accuracy": 0.8660451211035252,
      "num_tokens": 837301891.0,
      "step": 26235
    },
    {
      "entropy": 0.4849679758772254,
      "epoch": 2.413844658742618,
      "grad_norm": 0.1526428461074829,
      "learning_rate": 1.9541816174441072e-05,
      "loss": 0.4679,
      "mean_token_accuracy": 0.8496525138616562,
      "num_tokens": 837334010.0,
      "step": 26236
    },
    {
      "entropy": 0.5278410203754902,
      "epoch": 2.413936665114059,
      "grad_norm": 0.15299832820892334,
      "learning_rate": 1.9538749348299447e-05,
      "loss": 0.5127,
      "mean_token_accuracy": 0.8361075595021248,
      "num_tokens": 837366491.0,
      "step": 26237
    },
    {
      "entropy": 0.5269639454782009,
      "epoch": 2.4140286714855,
      "grad_norm": 0.16160568594932556,
      "learning_rate": 1.9535682522157818e-05,
      "loss": 0.5207,
      "mean_token_accuracy": 0.8348600007593632,
      "num_tokens": 837398490.0,
      "step": 26238
    },
    {
      "entropy": 0.4363840511068702,
      "epoch": 2.4141206778569417,
      "grad_norm": 0.14537681639194489,
      "learning_rate": 1.9532615696016193e-05,
      "loss": 0.4225,
      "mean_token_accuracy": 0.8637517169117928,
      "num_tokens": 837430143.0,
      "step": 26239
    },
    {
      "entropy": 0.5347416959702969,
      "epoch": 2.414212684228383,
      "grad_norm": 0.1654449850320816,
      "learning_rate": 1.9529548869874568e-05,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8276915214955807,
      "num_tokens": 837462104.0,
      "step": 26240
    },
    {
      "entropy": 0.4142713313922286,
      "epoch": 2.414304690599824,
      "grad_norm": 0.14597363770008087,
      "learning_rate": 1.9526482043732942e-05,
      "loss": 0.4136,
      "mean_token_accuracy": 0.8677781820297241,
      "num_tokens": 837494613.0,
      "step": 26241
    },
    {
      "entropy": 0.4071981282904744,
      "epoch": 2.414396696971265,
      "grad_norm": 0.14257872104644775,
      "learning_rate": 1.9523415217591313e-05,
      "loss": 0.3829,
      "mean_token_accuracy": 0.8753468729555607,
      "num_tokens": 837526558.0,
      "step": 26242
    },
    {
      "entropy": 0.5490435119718313,
      "epoch": 2.4144887033427063,
      "grad_norm": 0.16245099902153015,
      "learning_rate": 1.9520348391449688e-05,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8306854590773582,
      "num_tokens": 837558264.0,
      "step": 26243
    },
    {
      "entropy": 0.4843086190521717,
      "epoch": 2.414580709714148,
      "grad_norm": 0.15672263503074646,
      "learning_rate": 1.9517281565308063e-05,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8486210815608501,
      "num_tokens": 837590336.0,
      "step": 26244
    },
    {
      "entropy": 0.4871514793485403,
      "epoch": 2.414672716085589,
      "grad_norm": 0.15796081721782684,
      "learning_rate": 1.9514214739166437e-05,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8500645458698273,
      "num_tokens": 837622658.0,
      "step": 26245
    },
    {
      "entropy": 0.5504066585563123,
      "epoch": 2.41476472245703,
      "grad_norm": 0.1652960628271103,
      "learning_rate": 1.9511147913024812e-05,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8318702280521393,
      "num_tokens": 837654661.0,
      "step": 26246
    },
    {
      "entropy": 0.5364550221711397,
      "epoch": 2.4148567288284712,
      "grad_norm": 0.1655900478363037,
      "learning_rate": 1.9508081086883183e-05,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8359216004610062,
      "num_tokens": 837686727.0,
      "step": 26247
    },
    {
      "entropy": 0.46874054009094834,
      "epoch": 2.4149487351999124,
      "grad_norm": 0.15939263999462128,
      "learning_rate": 1.9505014260741558e-05,
      "loss": 0.4523,
      "mean_token_accuracy": 0.8539531454443932,
      "num_tokens": 837718393.0,
      "step": 26248
    },
    {
      "entropy": 0.60346167255193,
      "epoch": 2.415040741571354,
      "grad_norm": 0.16616109013557434,
      "learning_rate": 1.9501947434599936e-05,
      "loss": 0.5865,
      "mean_token_accuracy": 0.8175468854606152,
      "num_tokens": 837750519.0,
      "step": 26249
    },
    {
      "entropy": 0.5907868351787329,
      "epoch": 2.415132747942795,
      "grad_norm": 0.1695909947156906,
      "learning_rate": 1.9498880608458307e-05,
      "loss": 0.5772,
      "mean_token_accuracy": 0.8177648894488811,
      "num_tokens": 837782909.0,
      "step": 26250
    },
    {
      "entropy": 0.5797215385828167,
      "epoch": 2.415224754314236,
      "grad_norm": 0.17291800677776337,
      "learning_rate": 1.9495813782316682e-05,
      "loss": 0.5764,
      "mean_token_accuracy": 0.8179140500724316,
      "num_tokens": 837814751.0,
      "step": 26251
    },
    {
      "entropy": 0.5038789468817413,
      "epoch": 2.4153167606856774,
      "grad_norm": 0.15576066076755524,
      "learning_rate": 1.9492746956175057e-05,
      "loss": 0.4962,
      "mean_token_accuracy": 0.8391447253525257,
      "num_tokens": 837847189.0,
      "step": 26252
    },
    {
      "entropy": 0.5486738737672567,
      "epoch": 2.4154087670571185,
      "grad_norm": 0.16350534558296204,
      "learning_rate": 1.948968013003343e-05,
      "loss": 0.5286,
      "mean_token_accuracy": 0.831985104829073,
      "num_tokens": 837879482.0,
      "step": 26253
    },
    {
      "entropy": 0.5130103859119117,
      "epoch": 2.41550077342856,
      "grad_norm": 0.16125698387622833,
      "learning_rate": 1.9486613303891806e-05,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8415835909545422,
      "num_tokens": 837911869.0,
      "step": 26254
    },
    {
      "entropy": 0.5280800107866526,
      "epoch": 2.415592779800001,
      "grad_norm": 0.16021665930747986,
      "learning_rate": 1.9483546477750177e-05,
      "loss": 0.5149,
      "mean_token_accuracy": 0.834221500903368,
      "num_tokens": 837944212.0,
      "step": 26255
    },
    {
      "entropy": 0.49560053646564484,
      "epoch": 2.4156847861714423,
      "grad_norm": 0.16074839234352112,
      "learning_rate": 1.9480479651608552e-05,
      "loss": 0.4839,
      "mean_token_accuracy": 0.8436659686267376,
      "num_tokens": 837976623.0,
      "step": 26256
    },
    {
      "entropy": 0.5105428099632263,
      "epoch": 2.4157767925428835,
      "grad_norm": 0.1597680300474167,
      "learning_rate": 1.9477412825466926e-05,
      "loss": 0.5012,
      "mean_token_accuracy": 0.842586163431406,
      "num_tokens": 838009155.0,
      "step": 26257
    },
    {
      "entropy": 0.5455433391034603,
      "epoch": 2.4158687989143246,
      "grad_norm": 0.1604437679052353,
      "learning_rate": 1.94743459993253e-05,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8323696702718735,
      "num_tokens": 838041322.0,
      "step": 26258
    },
    {
      "entropy": 0.5649656597524881,
      "epoch": 2.415960805285766,
      "grad_norm": 0.17108139395713806,
      "learning_rate": 1.9471279173183672e-05,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8233682923018932,
      "num_tokens": 838072371.0,
      "step": 26259
    },
    {
      "entropy": 0.5935495449230075,
      "epoch": 2.4160528116572073,
      "grad_norm": 0.1693616658449173,
      "learning_rate": 1.9468212347042047e-05,
      "loss": 0.5825,
      "mean_token_accuracy": 0.8144504241645336,
      "num_tokens": 838104874.0,
      "step": 26260
    },
    {
      "entropy": 0.5206452924758196,
      "epoch": 2.4161448180286484,
      "grad_norm": 0.16266000270843506,
      "learning_rate": 1.946514552090042e-05,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8383431136608124,
      "num_tokens": 838137125.0,
      "step": 26261
    },
    {
      "entropy": 0.5258727641776204,
      "epoch": 2.4162368244000896,
      "grad_norm": 0.16262578964233398,
      "learning_rate": 1.9462078694758796e-05,
      "loss": 0.5135,
      "mean_token_accuracy": 0.8365604355931282,
      "num_tokens": 838169201.0,
      "step": 26262
    },
    {
      "entropy": 0.3992266012355685,
      "epoch": 2.4163288307715307,
      "grad_norm": 0.14321324229240417,
      "learning_rate": 1.9459011868617167e-05,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8778442852199078,
      "num_tokens": 838201056.0,
      "step": 26263
    },
    {
      "entropy": 0.4539053849875927,
      "epoch": 2.4164208371429723,
      "grad_norm": 0.1485748440027237,
      "learning_rate": 1.9455945042475542e-05,
      "loss": 0.4384,
      "mean_token_accuracy": 0.8596286587417126,
      "num_tokens": 838233084.0,
      "step": 26264
    },
    {
      "entropy": 0.48178022948559374,
      "epoch": 2.4165128435144134,
      "grad_norm": 0.15234102308750153,
      "learning_rate": 1.9452878216333917e-05,
      "loss": 0.4681,
      "mean_token_accuracy": 0.8518524393439293,
      "num_tokens": 838265085.0,
      "step": 26265
    },
    {
      "entropy": 0.43587500136345625,
      "epoch": 2.4166048498858546,
      "grad_norm": 0.15751948952674866,
      "learning_rate": 1.944981139019229e-05,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8612751476466656,
      "num_tokens": 838297285.0,
      "step": 26266
    },
    {
      "entropy": 0.4866376486606896,
      "epoch": 2.4166968562572957,
      "grad_norm": 0.1580386608839035,
      "learning_rate": 1.9446744564050666e-05,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8514489755034447,
      "num_tokens": 838329457.0,
      "step": 26267
    },
    {
      "entropy": 0.5845548287034035,
      "epoch": 2.416788862628737,
      "grad_norm": 0.16486333310604095,
      "learning_rate": 1.9443677737909037e-05,
      "loss": 0.5712,
      "mean_token_accuracy": 0.8231195397675037,
      "num_tokens": 838361985.0,
      "step": 26268
    },
    {
      "entropy": 0.6318444134667516,
      "epoch": 2.4168808690001784,
      "grad_norm": 0.1778215765953064,
      "learning_rate": 1.9440610911767412e-05,
      "loss": 0.6133,
      "mean_token_accuracy": 0.8071498870849609,
      "num_tokens": 838392833.0,
      "step": 26269
    },
    {
      "entropy": 0.576216472312808,
      "epoch": 2.4169728753716195,
      "grad_norm": 0.16798865795135498,
      "learning_rate": 1.9437544085625787e-05,
      "loss": 0.573,
      "mean_token_accuracy": 0.8225463666021824,
      "num_tokens": 838423548.0,
      "step": 26270
    },
    {
      "entropy": 0.5432569282129407,
      "epoch": 2.4170648817430607,
      "grad_norm": 0.16493427753448486,
      "learning_rate": 1.943447725948416e-05,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8316544778645039,
      "num_tokens": 838456028.0,
      "step": 26271
    },
    {
      "entropy": 0.5122611143160611,
      "epoch": 2.417156888114502,
      "grad_norm": 0.16457322239875793,
      "learning_rate": 1.9431410433342533e-05,
      "loss": 0.5117,
      "mean_token_accuracy": 0.839708000421524,
      "num_tokens": 838486545.0,
      "step": 26272
    },
    {
      "entropy": 0.5224698625970632,
      "epoch": 2.417248894485943,
      "grad_norm": 0.16017672419548035,
      "learning_rate": 1.9428343607200907e-05,
      "loss": 0.521,
      "mean_token_accuracy": 0.8359888158738613,
      "num_tokens": 838519187.0,
      "step": 26273
    },
    {
      "entropy": 0.4691917123273015,
      "epoch": 2.4173409008573845,
      "grad_norm": 0.15059880912303925,
      "learning_rate": 1.9425276781059282e-05,
      "loss": 0.46,
      "mean_token_accuracy": 0.8541224114596844,
      "num_tokens": 838551275.0,
      "step": 26274
    },
    {
      "entropy": 0.40597254363819957,
      "epoch": 2.4174329072288256,
      "grad_norm": 0.14573422074317932,
      "learning_rate": 1.9422209954917656e-05,
      "loss": 0.3962,
      "mean_token_accuracy": 0.8742405325174332,
      "num_tokens": 838584024.0,
      "step": 26275
    },
    {
      "entropy": 0.4990108795464039,
      "epoch": 2.417524913600267,
      "grad_norm": 0.1601148098707199,
      "learning_rate": 1.941914312877603e-05,
      "loss": 0.4776,
      "mean_token_accuracy": 0.8464471437036991,
      "num_tokens": 838614863.0,
      "step": 26276
    },
    {
      "entropy": 0.5564038935117424,
      "epoch": 2.417616919971708,
      "grad_norm": 0.1664597988128662,
      "learning_rate": 1.9416076302634402e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8248379528522491,
      "num_tokens": 838647146.0,
      "step": 26277
    },
    {
      "entropy": 0.5277552148327231,
      "epoch": 2.417708926343149,
      "grad_norm": 0.162570521235466,
      "learning_rate": 1.9413009476492777e-05,
      "loss": 0.518,
      "mean_token_accuracy": 0.8336757607758045,
      "num_tokens": 838678830.0,
      "step": 26278
    },
    {
      "entropy": 0.46087318286299706,
      "epoch": 2.4178009327145906,
      "grad_norm": 0.15471142530441284,
      "learning_rate": 1.9409942650351155e-05,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8599086925387383,
      "num_tokens": 838711211.0,
      "step": 26279
    },
    {
      "entropy": 0.5195219875313342,
      "epoch": 2.4178929390860318,
      "grad_norm": 0.16375814378261566,
      "learning_rate": 1.9406875824209526e-05,
      "loss": 0.5115,
      "mean_token_accuracy": 0.839448805898428,
      "num_tokens": 838742215.0,
      "step": 26280
    },
    {
      "entropy": 0.4953533988445997,
      "epoch": 2.417984945457473,
      "grad_norm": 0.1612648069858551,
      "learning_rate": 1.94038089980679e-05,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8422178849577904,
      "num_tokens": 838773952.0,
      "step": 26281
    },
    {
      "entropy": 0.48253753781318665,
      "epoch": 2.418076951828914,
      "grad_norm": 0.15085934102535248,
      "learning_rate": 1.9400742171926276e-05,
      "loss": 0.4694,
      "mean_token_accuracy": 0.8506710156798363,
      "num_tokens": 838806371.0,
      "step": 26282
    },
    {
      "entropy": 0.586578294634819,
      "epoch": 2.418168958200355,
      "grad_norm": 0.16372935473918915,
      "learning_rate": 1.939767534578465e-05,
      "loss": 0.579,
      "mean_token_accuracy": 0.8224802725017071,
      "num_tokens": 838838784.0,
      "step": 26283
    },
    {
      "entropy": 0.5037675146013498,
      "epoch": 2.4182609645717967,
      "grad_norm": 0.1605452597141266,
      "learning_rate": 1.9394608519643025e-05,
      "loss": 0.4895,
      "mean_token_accuracy": 0.8462951444089413,
      "num_tokens": 838870483.0,
      "step": 26284
    },
    {
      "entropy": 0.6188630312681198,
      "epoch": 2.418352970943238,
      "grad_norm": 0.17192287743091583,
      "learning_rate": 1.9391541693501396e-05,
      "loss": 0.6148,
      "mean_token_accuracy": 0.8068852312862873,
      "num_tokens": 838902262.0,
      "step": 26285
    },
    {
      "entropy": 0.5354252075776458,
      "epoch": 2.418444977314679,
      "grad_norm": 0.16095364093780518,
      "learning_rate": 1.938847486735977e-05,
      "loss": 0.5294,
      "mean_token_accuracy": 0.8334634825587273,
      "num_tokens": 838934463.0,
      "step": 26286
    },
    {
      "entropy": 0.4959522243589163,
      "epoch": 2.41853698368612,
      "grad_norm": 0.15910163521766663,
      "learning_rate": 1.9385408041218145e-05,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8425245471298695,
      "num_tokens": 838966872.0,
      "step": 26287
    },
    {
      "entropy": 0.4486150350421667,
      "epoch": 2.4186289900575613,
      "grad_norm": 0.1495533138513565,
      "learning_rate": 1.938234121507652e-05,
      "loss": 0.4306,
      "mean_token_accuracy": 0.8633880726993084,
      "num_tokens": 838998199.0,
      "step": 26288
    },
    {
      "entropy": 0.5216753333806992,
      "epoch": 2.418720996429003,
      "grad_norm": 0.16135196387767792,
      "learning_rate": 1.937927438893489e-05,
      "loss": 0.4961,
      "mean_token_accuracy": 0.8464280739426613,
      "num_tokens": 839030086.0,
      "step": 26289
    },
    {
      "entropy": 0.6740218196064234,
      "epoch": 2.418813002800444,
      "grad_norm": 0.1789267212152481,
      "learning_rate": 1.9376207562793266e-05,
      "loss": 0.6744,
      "mean_token_accuracy": 0.7941544204950333,
      "num_tokens": 839062110.0,
      "step": 26290
    },
    {
      "entropy": 0.6040378306061029,
      "epoch": 2.418905009171885,
      "grad_norm": 10.45012378692627,
      "learning_rate": 1.937314073665164e-05,
      "loss": 1.1112,
      "mean_token_accuracy": 0.7924032777082175,
      "num_tokens": 839094132.0,
      "step": 26291
    },
    {
      "entropy": 0.5089667178690434,
      "epoch": 2.4189970155433262,
      "grad_norm": 0.16180169582366943,
      "learning_rate": 1.9370073910510015e-05,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8395521901547909,
      "num_tokens": 839125837.0,
      "step": 26292
    },
    {
      "entropy": 0.5547654805704951,
      "epoch": 2.4190890219147674,
      "grad_norm": 0.16854777932167053,
      "learning_rate": 1.9367007084368387e-05,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8266929276287556,
      "num_tokens": 839156976.0,
      "step": 26293
    },
    {
      "entropy": 0.5767091549932957,
      "epoch": 2.419181028286209,
      "grad_norm": 0.17813105881214142,
      "learning_rate": 1.936394025822676e-05,
      "loss": 0.5754,
      "mean_token_accuracy": 0.8186828680336475,
      "num_tokens": 839187918.0,
      "step": 26294
    },
    {
      "entropy": 0.4951028909999877,
      "epoch": 2.41927303465765,
      "grad_norm": 0.16786326467990875,
      "learning_rate": 1.9360873432085136e-05,
      "loss": 0.4788,
      "mean_token_accuracy": 0.8490370661020279,
      "num_tokens": 839219937.0,
      "step": 26295
    },
    {
      "entropy": 0.5835617445409298,
      "epoch": 2.4193650410290912,
      "grad_norm": 0.19421608746051788,
      "learning_rate": 1.935780660594351e-05,
      "loss": 0.5577,
      "mean_token_accuracy": 0.8246928118169308,
      "num_tokens": 839251023.0,
      "step": 26296
    },
    {
      "entropy": 0.5719884280115366,
      "epoch": 2.4194570474005324,
      "grad_norm": 0.19049908220767975,
      "learning_rate": 1.9354739779801885e-05,
      "loss": 0.5552,
      "mean_token_accuracy": 0.8271919265389442,
      "num_tokens": 839283016.0,
      "step": 26297
    },
    {
      "entropy": 0.5573744382709265,
      "epoch": 2.4195490537719735,
      "grad_norm": 0.18793749809265137,
      "learning_rate": 1.9351672953660256e-05,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8311035893857479,
      "num_tokens": 839314958.0,
      "step": 26298
    },
    {
      "entropy": 0.46159248938784003,
      "epoch": 2.419641060143415,
      "grad_norm": 0.17295727133750916,
      "learning_rate": 1.934860612751863e-05,
      "loss": 0.4523,
      "mean_token_accuracy": 0.8545262552797794,
      "num_tokens": 839347188.0,
      "step": 26299
    },
    {
      "entropy": 0.49550957046449184,
      "epoch": 2.419733066514856,
      "grad_norm": 0.16879190504550934,
      "learning_rate": 1.9345539301377006e-05,
      "loss": 0.4741,
      "mean_token_accuracy": 0.8458736091852188,
      "num_tokens": 839379386.0,
      "step": 26300
    },
    {
      "entropy": 0.5345104052685201,
      "epoch": 2.4198250728862973,
      "grad_norm": 0.16282372176647186,
      "learning_rate": 1.934247247523538e-05,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8390679359436035,
      "num_tokens": 839411314.0,
      "step": 26301
    },
    {
      "entropy": 0.389999620616436,
      "epoch": 2.4199170792577385,
      "grad_norm": 0.15068212151527405,
      "learning_rate": 1.933940564909375e-05,
      "loss": 0.3794,
      "mean_token_accuracy": 0.8826135359704494,
      "num_tokens": 839444075.0,
      "step": 26302
    },
    {
      "entropy": 0.657831272110343,
      "epoch": 2.4200090856291796,
      "grad_norm": 0.17597220838069916,
      "learning_rate": 1.9336338822952126e-05,
      "loss": 0.6447,
      "mean_token_accuracy": 0.7984032034873962,
      "num_tokens": 839475342.0,
      "step": 26303
    },
    {
      "entropy": 0.4946559313684702,
      "epoch": 2.420101092000621,
      "grad_norm": 0.15183591842651367,
      "learning_rate": 1.93332719968105e-05,
      "loss": 0.4797,
      "mean_token_accuracy": 0.8491266593337059,
      "num_tokens": 839508071.0,
      "step": 26304
    },
    {
      "entropy": 0.48641214426606894,
      "epoch": 2.4201930983720623,
      "grad_norm": 0.15804404020309448,
      "learning_rate": 1.9330205170668876e-05,
      "loss": 0.4648,
      "mean_token_accuracy": 0.8554648794233799,
      "num_tokens": 839539078.0,
      "step": 26305
    },
    {
      "entropy": 0.5438044406473637,
      "epoch": 2.4202851047435034,
      "grad_norm": 0.17167679965496063,
      "learning_rate": 1.932713834452725e-05,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8335687108337879,
      "num_tokens": 839569912.0,
      "step": 26306
    },
    {
      "entropy": 0.5968007082119584,
      "epoch": 2.4203771111149446,
      "grad_norm": 0.16878047585487366,
      "learning_rate": 1.932407151838562e-05,
      "loss": 0.5913,
      "mean_token_accuracy": 0.813680037856102,
      "num_tokens": 839600794.0,
      "step": 26307
    },
    {
      "entropy": 0.4875752925872803,
      "epoch": 2.4204691174863857,
      "grad_norm": 0.16011981666088104,
      "learning_rate": 1.9321004692243996e-05,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8438336290419102,
      "num_tokens": 839632160.0,
      "step": 26308
    },
    {
      "entropy": 0.5507159680128098,
      "epoch": 2.4205611238578273,
      "grad_norm": 0.16587483882904053,
      "learning_rate": 1.931793786610237e-05,
      "loss": 0.5447,
      "mean_token_accuracy": 0.827044915407896,
      "num_tokens": 839664427.0,
      "step": 26309
    },
    {
      "entropy": 0.43580812541767955,
      "epoch": 2.4206531302292684,
      "grad_norm": 0.1451251357793808,
      "learning_rate": 1.9314871039960745e-05,
      "loss": 0.4295,
      "mean_token_accuracy": 0.8655190914869308,
      "num_tokens": 839696999.0,
      "step": 26310
    },
    {
      "entropy": 0.43587171379476786,
      "epoch": 2.4207451366007096,
      "grad_norm": 0.1448565572500229,
      "learning_rate": 1.931180421381912e-05,
      "loss": 0.4202,
      "mean_token_accuracy": 0.8658527433872223,
      "num_tokens": 839728788.0,
      "step": 26311
    },
    {
      "entropy": 0.5595165975391865,
      "epoch": 2.4208371429721507,
      "grad_norm": 0.16860570013523102,
      "learning_rate": 1.9308737387677495e-05,
      "loss": 0.5465,
      "mean_token_accuracy": 0.823134396225214,
      "num_tokens": 839759095.0,
      "step": 26312
    },
    {
      "entropy": 0.4763611676171422,
      "epoch": 2.420929149343592,
      "grad_norm": 0.15162308514118195,
      "learning_rate": 1.930567056153587e-05,
      "loss": 0.4693,
      "mean_token_accuracy": 0.8516562767326832,
      "num_tokens": 839791664.0,
      "step": 26313
    },
    {
      "entropy": 0.5556436481419951,
      "epoch": 2.4210211557150334,
      "grad_norm": 0.16727206110954285,
      "learning_rate": 1.9302603735394244e-05,
      "loss": 0.5455,
      "mean_token_accuracy": 0.829166367650032,
      "num_tokens": 839823729.0,
      "step": 26314
    },
    {
      "entropy": 0.5841872552409768,
      "epoch": 2.4211131620864745,
      "grad_norm": 0.16811227798461914,
      "learning_rate": 1.9299536909252615e-05,
      "loss": 0.5696,
      "mean_token_accuracy": 0.8195250518620014,
      "num_tokens": 839855675.0,
      "step": 26315
    },
    {
      "entropy": 0.5547017231583595,
      "epoch": 2.4212051684579157,
      "grad_norm": 0.16645769774913788,
      "learning_rate": 1.929647008311099e-05,
      "loss": 0.5501,
      "mean_token_accuracy": 0.824868880212307,
      "num_tokens": 839887335.0,
      "step": 26316
    },
    {
      "entropy": 0.4835600731894374,
      "epoch": 2.421297174829357,
      "grad_norm": 0.1551358699798584,
      "learning_rate": 1.9293403256969365e-05,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8484980836510658,
      "num_tokens": 839919588.0,
      "step": 26317
    },
    {
      "entropy": 0.5331715727224946,
      "epoch": 2.421389181200798,
      "grad_norm": 0.16595174372196198,
      "learning_rate": 1.929033643082774e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8327179662883282,
      "num_tokens": 839951310.0,
      "step": 26318
    },
    {
      "entropy": 0.5180539861321449,
      "epoch": 2.4214811875722395,
      "grad_norm": 0.1615906059741974,
      "learning_rate": 1.928726960468611e-05,
      "loss": 0.5061,
      "mean_token_accuracy": 0.8400950580835342,
      "num_tokens": 839983330.0,
      "step": 26319
    },
    {
      "entropy": 0.5798068698495626,
      "epoch": 2.4215731939436806,
      "grad_norm": 0.16968697309494019,
      "learning_rate": 1.9284202778544485e-05,
      "loss": 0.5672,
      "mean_token_accuracy": 0.8202954865992069,
      "num_tokens": 840015404.0,
      "step": 26320
    },
    {
      "entropy": 0.5356476390734315,
      "epoch": 2.421665200315122,
      "grad_norm": 0.16196531057357788,
      "learning_rate": 1.928113595240286e-05,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8376455083489418,
      "num_tokens": 840046976.0,
      "step": 26321
    },
    {
      "entropy": 0.4743208307772875,
      "epoch": 2.421757206686563,
      "grad_norm": 0.15581205487251282,
      "learning_rate": 1.9278069126261234e-05,
      "loss": 0.462,
      "mean_token_accuracy": 0.8588723428547382,
      "num_tokens": 840078882.0,
      "step": 26322
    },
    {
      "entropy": 0.4132991652004421,
      "epoch": 2.421849213058004,
      "grad_norm": 0.14434105157852173,
      "learning_rate": 1.9275002300119606e-05,
      "loss": 0.4097,
      "mean_token_accuracy": 0.8691697120666504,
      "num_tokens": 840111318.0,
      "step": 26323
    },
    {
      "entropy": 0.5746529959142208,
      "epoch": 2.4219412194294456,
      "grad_norm": 0.16632752120494843,
      "learning_rate": 1.927193547397798e-05,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8223824575543404,
      "num_tokens": 840143659.0,
      "step": 26324
    },
    {
      "entropy": 0.5318578006699681,
      "epoch": 2.4220332258008868,
      "grad_norm": 0.15949025750160217,
      "learning_rate": 1.9268868647836355e-05,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8406442366540432,
      "num_tokens": 840175192.0,
      "step": 26325
    },
    {
      "entropy": 0.56197215244174,
      "epoch": 2.422125232172328,
      "grad_norm": 0.1632312685251236,
      "learning_rate": 1.926580182169473e-05,
      "loss": 0.5483,
      "mean_token_accuracy": 0.8245162777602673,
      "num_tokens": 840206686.0,
      "step": 26326
    },
    {
      "entropy": 0.4964063595980406,
      "epoch": 2.422217238543769,
      "grad_norm": 0.15511052310466766,
      "learning_rate": 1.9262734995553104e-05,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8461114354431629,
      "num_tokens": 840238557.0,
      "step": 26327
    },
    {
      "entropy": 0.5429633241146803,
      "epoch": 2.42230924491521,
      "grad_norm": 0.16501085460186005,
      "learning_rate": 1.9259668169411476e-05,
      "loss": 0.5262,
      "mean_token_accuracy": 0.8321288786828518,
      "num_tokens": 840270545.0,
      "step": 26328
    },
    {
      "entropy": 0.6148453839123249,
      "epoch": 2.4224012512866517,
      "grad_norm": 0.17230193316936493,
      "learning_rate": 1.925660134326985e-05,
      "loss": 0.61,
      "mean_token_accuracy": 0.8078610450029373,
      "num_tokens": 840302013.0,
      "step": 26329
    },
    {
      "entropy": 0.4358662061858922,
      "epoch": 2.422493257658093,
      "grad_norm": 0.14612968266010284,
      "learning_rate": 1.9253534517128225e-05,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8629809059202671,
      "num_tokens": 840334473.0,
      "step": 26330
    },
    {
      "entropy": 0.5274232737720013,
      "epoch": 2.422585264029534,
      "grad_norm": 0.15932336449623108,
      "learning_rate": 1.92504676909866e-05,
      "loss": 0.5119,
      "mean_token_accuracy": 0.8373482748866081,
      "num_tokens": 840366697.0,
      "step": 26331
    },
    {
      "entropy": 0.4525425508618355,
      "epoch": 2.422677270400975,
      "grad_norm": 0.1557856798171997,
      "learning_rate": 1.924740086484497e-05,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8532012067735195,
      "num_tokens": 840397935.0,
      "step": 26332
    },
    {
      "entropy": 0.4880889384075999,
      "epoch": 2.4227692767724163,
      "grad_norm": 0.157795712351799,
      "learning_rate": 1.9244334038703345e-05,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8479533717036247,
      "num_tokens": 840429663.0,
      "step": 26333
    },
    {
      "entropy": 0.5967187080532312,
      "epoch": 2.422861283143858,
      "grad_norm": 0.17415113747119904,
      "learning_rate": 1.924126721256172e-05,
      "loss": 0.5823,
      "mean_token_accuracy": 0.8167623430490494,
      "num_tokens": 840461460.0,
      "step": 26334
    },
    {
      "entropy": 0.5609104856848717,
      "epoch": 2.422953289515299,
      "grad_norm": 0.16638244688510895,
      "learning_rate": 1.9238200386420095e-05,
      "loss": 0.5503,
      "mean_token_accuracy": 0.8219676241278648,
      "num_tokens": 840493465.0,
      "step": 26335
    },
    {
      "entropy": 0.5286913220770657,
      "epoch": 2.42304529588674,
      "grad_norm": 0.16410218179225922,
      "learning_rate": 1.923513356027847e-05,
      "loss": 0.5172,
      "mean_token_accuracy": 0.8338008970022202,
      "num_tokens": 840524789.0,
      "step": 26336
    },
    {
      "entropy": 0.5106750782579184,
      "epoch": 2.4231373022581812,
      "grad_norm": 0.16160902380943298,
      "learning_rate": 1.923206673413684e-05,
      "loss": 0.5097,
      "mean_token_accuracy": 0.8403234221041203,
      "num_tokens": 840556961.0,
      "step": 26337
    },
    {
      "entropy": 0.5146306026726961,
      "epoch": 2.4232293086296224,
      "grad_norm": 0.15796782076358795,
      "learning_rate": 1.9228999907995215e-05,
      "loss": 0.5073,
      "mean_token_accuracy": 0.8385555222630501,
      "num_tokens": 840589378.0,
      "step": 26338
    },
    {
      "entropy": 0.4060762315057218,
      "epoch": 2.423321315001064,
      "grad_norm": 0.14326047897338867,
      "learning_rate": 1.922593308185359e-05,
      "loss": 0.3844,
      "mean_token_accuracy": 0.8773957192897797,
      "num_tokens": 840621300.0,
      "step": 26339
    },
    {
      "entropy": 0.5279860757291317,
      "epoch": 2.423413321372505,
      "grad_norm": 0.16410650312900543,
      "learning_rate": 1.9222866255711965e-05,
      "loss": 0.5177,
      "mean_token_accuracy": 0.8360258676111698,
      "num_tokens": 840653274.0,
      "step": 26340
    },
    {
      "entropy": 0.40760072926059365,
      "epoch": 2.4235053277439462,
      "grad_norm": 0.14673586189746857,
      "learning_rate": 1.921979942957034e-05,
      "loss": 0.3914,
      "mean_token_accuracy": 0.8723608441650867,
      "num_tokens": 840685424.0,
      "step": 26341
    },
    {
      "entropy": 0.421301880851388,
      "epoch": 2.4235973341153874,
      "grad_norm": 0.14882725477218628,
      "learning_rate": 1.9216732603428714e-05,
      "loss": 0.4168,
      "mean_token_accuracy": 0.8660787008702755,
      "num_tokens": 840717849.0,
      "step": 26342
    },
    {
      "entropy": 0.5670976042747498,
      "epoch": 2.4236893404868285,
      "grad_norm": 0.16972021758556366,
      "learning_rate": 1.921366577728709e-05,
      "loss": 0.5579,
      "mean_token_accuracy": 0.8225084878504276,
      "num_tokens": 840749757.0,
      "step": 26343
    },
    {
      "entropy": 0.4893368138000369,
      "epoch": 2.42378134685827,
      "grad_norm": 0.1591264307498932,
      "learning_rate": 1.9210598951145463e-05,
      "loss": 0.4803,
      "mean_token_accuracy": 0.8493402190506458,
      "num_tokens": 840781934.0,
      "step": 26344
    },
    {
      "entropy": 0.575337715446949,
      "epoch": 2.423873353229711,
      "grad_norm": 0.16453485190868378,
      "learning_rate": 1.9207532125003834e-05,
      "loss": 0.5691,
      "mean_token_accuracy": 0.823400691151619,
      "num_tokens": 840814017.0,
      "step": 26345
    },
    {
      "entropy": 0.5361615028232336,
      "epoch": 2.4239653596011523,
      "grad_norm": 0.16516035795211792,
      "learning_rate": 1.920446529886221e-05,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8344879001379013,
      "num_tokens": 840846076.0,
      "step": 26346
    },
    {
      "entropy": 0.5026085898280144,
      "epoch": 2.4240573659725935,
      "grad_norm": 0.15369142591953278,
      "learning_rate": 1.9201398472720584e-05,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8439575806260109,
      "num_tokens": 840878714.0,
      "step": 26347
    },
    {
      "entropy": 0.473145738709718,
      "epoch": 2.4241493723440346,
      "grad_norm": 0.15297989547252655,
      "learning_rate": 1.919833164657896e-05,
      "loss": 0.4547,
      "mean_token_accuracy": 0.8570959568023682,
      "num_tokens": 840910555.0,
      "step": 26348
    },
    {
      "entropy": 0.6095546903088689,
      "epoch": 2.424241378715476,
      "grad_norm": 0.1695767343044281,
      "learning_rate": 1.919526482043733e-05,
      "loss": 0.6042,
      "mean_token_accuracy": 0.8085300885140896,
      "num_tokens": 840942336.0,
      "step": 26349
    },
    {
      "entropy": 0.5817014295607805,
      "epoch": 2.4243333850869173,
      "grad_norm": 0.17307670414447784,
      "learning_rate": 1.9192197994295704e-05,
      "loss": 0.5729,
      "mean_token_accuracy": 0.8197600431740284,
      "num_tokens": 840972476.0,
      "step": 26350
    },
    {
      "entropy": 0.5080250641331077,
      "epoch": 2.4244253914583584,
      "grad_norm": 0.15726302564144135,
      "learning_rate": 1.918913116815408e-05,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8448506034910679,
      "num_tokens": 841004988.0,
      "step": 26351
    },
    {
      "entropy": 0.4861973123624921,
      "epoch": 2.4245173978297996,
      "grad_norm": 0.15373507142066956,
      "learning_rate": 1.9186064342012454e-05,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8505481071770191,
      "num_tokens": 841036401.0,
      "step": 26352
    },
    {
      "entropy": 0.5454591382294893,
      "epoch": 2.4246094042012407,
      "grad_norm": 0.1616116613149643,
      "learning_rate": 1.9182997515870825e-05,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8312272652983665,
      "num_tokens": 841068069.0,
      "step": 26353
    },
    {
      "entropy": 0.5528915505856276,
      "epoch": 2.4247014105726823,
      "grad_norm": 0.16695833206176758,
      "learning_rate": 1.91799306897292e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8290524370968342,
      "num_tokens": 841100571.0,
      "step": 26354
    },
    {
      "entropy": 0.48408691957592964,
      "epoch": 2.4247934169441234,
      "grad_norm": 0.15844298899173737,
      "learning_rate": 1.9176863863587574e-05,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8455912508070469,
      "num_tokens": 841131254.0,
      "step": 26355
    },
    {
      "entropy": 0.4846726432442665,
      "epoch": 2.4248854233155646,
      "grad_norm": 0.15361535549163818,
      "learning_rate": 1.917379703744595e-05,
      "loss": 0.4803,
      "mean_token_accuracy": 0.8491148576140404,
      "num_tokens": 841163547.0,
      "step": 26356
    },
    {
      "entropy": 0.5472778268158436,
      "epoch": 2.4249774296870057,
      "grad_norm": 0.16259868443012238,
      "learning_rate": 1.9170730211304323e-05,
      "loss": 0.5332,
      "mean_token_accuracy": 0.8293956778943539,
      "num_tokens": 841195453.0,
      "step": 26357
    },
    {
      "entropy": 0.5021707371342927,
      "epoch": 2.425069436058447,
      "grad_norm": 0.1496133804321289,
      "learning_rate": 1.9167663385162695e-05,
      "loss": 0.4871,
      "mean_token_accuracy": 0.8482457548379898,
      "num_tokens": 841227768.0,
      "step": 26358
    },
    {
      "entropy": 0.4832816356793046,
      "epoch": 2.4251614424298884,
      "grad_norm": 0.15204863250255585,
      "learning_rate": 1.916459655902107e-05,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8531017974019051,
      "num_tokens": 841260333.0,
      "step": 26359
    },
    {
      "entropy": 0.4839537162333727,
      "epoch": 2.4252534488013295,
      "grad_norm": 0.146095871925354,
      "learning_rate": 1.9161529732879444e-05,
      "loss": 0.4643,
      "mean_token_accuracy": 0.8549401499330997,
      "num_tokens": 841293101.0,
      "step": 26360
    },
    {
      "entropy": 0.6112568713724613,
      "epoch": 2.4253454551727707,
      "grad_norm": 0.17255249619483948,
      "learning_rate": 1.915846290673782e-05,
      "loss": 0.603,
      "mean_token_accuracy": 0.8122915364801884,
      "num_tokens": 841325573.0,
      "step": 26361
    },
    {
      "entropy": 0.5215438511222601,
      "epoch": 2.425437461544212,
      "grad_norm": 0.1624741554260254,
      "learning_rate": 1.915539608059619e-05,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8418176285922527,
      "num_tokens": 841357317.0,
      "step": 26362
    },
    {
      "entropy": 0.41702471766620874,
      "epoch": 2.425529467915653,
      "grad_norm": 0.14958316087722778,
      "learning_rate": 1.9152329254454564e-05,
      "loss": 0.4041,
      "mean_token_accuracy": 0.8690393008291721,
      "num_tokens": 841389430.0,
      "step": 26363
    },
    {
      "entropy": 0.4647013349458575,
      "epoch": 2.4256214742870945,
      "grad_norm": 0.1514224112033844,
      "learning_rate": 1.914926242831294e-05,
      "loss": 0.44,
      "mean_token_accuracy": 0.8574097454547882,
      "num_tokens": 841421116.0,
      "step": 26364
    },
    {
      "entropy": 0.4363692350452766,
      "epoch": 2.4257134806585356,
      "grad_norm": 0.15051701664924622,
      "learning_rate": 1.9146195602171314e-05,
      "loss": 0.4236,
      "mean_token_accuracy": 0.8645518682897091,
      "num_tokens": 841452917.0,
      "step": 26365
    },
    {
      "entropy": 0.5218030971009284,
      "epoch": 2.425805487029977,
      "grad_norm": 0.1602536141872406,
      "learning_rate": 1.914312877602969e-05,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8407766073942184,
      "num_tokens": 841485071.0,
      "step": 26366
    },
    {
      "entropy": 0.524603858590126,
      "epoch": 2.425897493401418,
      "grad_norm": 0.16175797581672668,
      "learning_rate": 1.914006194988806e-05,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8360143005847931,
      "num_tokens": 841516269.0,
      "step": 26367
    },
    {
      "entropy": 0.6892434675246477,
      "epoch": 2.425989499772859,
      "grad_norm": 0.17600466310977936,
      "learning_rate": 1.9136995123746434e-05,
      "loss": 0.6733,
      "mean_token_accuracy": 0.789639126509428,
      "num_tokens": 841547730.0,
      "step": 26368
    },
    {
      "entropy": 0.5450124656781554,
      "epoch": 2.4260815061443006,
      "grad_norm": 0.16644905507564545,
      "learning_rate": 1.913392829760481e-05,
      "loss": 0.5438,
      "mean_token_accuracy": 0.8332631997764111,
      "num_tokens": 841578909.0,
      "step": 26369
    },
    {
      "entropy": 0.5328117804601789,
      "epoch": 2.4261735125157418,
      "grad_norm": 0.15999124944210052,
      "learning_rate": 1.9130861471463184e-05,
      "loss": 0.5102,
      "mean_token_accuracy": 0.836312785744667,
      "num_tokens": 841610536.0,
      "step": 26370
    },
    {
      "entropy": 0.561974854208529,
      "epoch": 2.426265518887183,
      "grad_norm": 0.1708362102508545,
      "learning_rate": 1.9127794645321558e-05,
      "loss": 0.5495,
      "mean_token_accuracy": 0.8253428116440773,
      "num_tokens": 841641767.0,
      "step": 26371
    },
    {
      "entropy": 0.5242815129458904,
      "epoch": 2.426357525258624,
      "grad_norm": 0.16083751618862152,
      "learning_rate": 1.9124727819179933e-05,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8369554951786995,
      "num_tokens": 841674198.0,
      "step": 26372
    },
    {
      "entropy": 0.47312218917068094,
      "epoch": 2.426449531630065,
      "grad_norm": 0.1545485258102417,
      "learning_rate": 1.9121660993038308e-05,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8522380851209164,
      "num_tokens": 841705908.0,
      "step": 26373
    },
    {
      "entropy": 0.5107283480465412,
      "epoch": 2.4265415380015067,
      "grad_norm": 0.15571485459804535,
      "learning_rate": 1.9118594166896682e-05,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8428943939507008,
      "num_tokens": 841737928.0,
      "step": 26374
    },
    {
      "entropy": 0.5068120462819934,
      "epoch": 2.426633544372948,
      "grad_norm": 0.16164450347423553,
      "learning_rate": 1.9115527340755053e-05,
      "loss": 0.5094,
      "mean_token_accuracy": 0.8390782438218594,
      "num_tokens": 841769925.0,
      "step": 26375
    },
    {
      "entropy": 0.556636082008481,
      "epoch": 2.426725550744389,
      "grad_norm": 0.16601517796516418,
      "learning_rate": 1.9112460514613428e-05,
      "loss": 0.5446,
      "mean_token_accuracy": 0.8268909119069576,
      "num_tokens": 841802116.0,
      "step": 26376
    },
    {
      "entropy": 0.5831893775612116,
      "epoch": 2.42681755711583,
      "grad_norm": 0.16955973207950592,
      "learning_rate": 1.9109393688471803e-05,
      "loss": 0.5759,
      "mean_token_accuracy": 0.821335643529892,
      "num_tokens": 841833925.0,
      "step": 26377
    },
    {
      "entropy": 0.6095866337418556,
      "epoch": 2.4269095634872713,
      "grad_norm": 0.17298276722431183,
      "learning_rate": 1.9106326862330177e-05,
      "loss": 0.5893,
      "mean_token_accuracy": 0.8152382262051105,
      "num_tokens": 841864904.0,
      "step": 26378
    },
    {
      "entropy": 0.43289319332689047,
      "epoch": 2.427001569858713,
      "grad_norm": 0.14972637593746185,
      "learning_rate": 1.910326003618855e-05,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8667226918041706,
      "num_tokens": 841896659.0,
      "step": 26379
    },
    {
      "entropy": 0.6117396298795938,
      "epoch": 2.427093576230154,
      "grad_norm": 0.17140159010887146,
      "learning_rate": 1.9100193210046923e-05,
      "loss": 0.6036,
      "mean_token_accuracy": 0.8164207227528095,
      "num_tokens": 841928970.0,
      "step": 26380
    },
    {
      "entropy": 0.5028997287154198,
      "epoch": 2.427185582601595,
      "grad_norm": 0.15785446763038635,
      "learning_rate": 1.9097126383905298e-05,
      "loss": 0.4928,
      "mean_token_accuracy": 0.8475314378738403,
      "num_tokens": 841961067.0,
      "step": 26381
    },
    {
      "entropy": 0.4985145041719079,
      "epoch": 2.4272775889730362,
      "grad_norm": 0.15591523051261902,
      "learning_rate": 1.9094059557763673e-05,
      "loss": 0.4778,
      "mean_token_accuracy": 0.8486734516918659,
      "num_tokens": 841993508.0,
      "step": 26382
    },
    {
      "entropy": 0.42063776310533285,
      "epoch": 2.4273695953444774,
      "grad_norm": 0.1461179256439209,
      "learning_rate": 1.9090992731622044e-05,
      "loss": 0.4028,
      "mean_token_accuracy": 0.8686473071575165,
      "num_tokens": 842025756.0,
      "step": 26383
    },
    {
      "entropy": 0.5267479866743088,
      "epoch": 2.427461601715919,
      "grad_norm": 0.15950500965118408,
      "learning_rate": 1.908792590548042e-05,
      "loss": 0.5164,
      "mean_token_accuracy": 0.8344012312591076,
      "num_tokens": 842057188.0,
      "step": 26384
    },
    {
      "entropy": 0.5571404322981834,
      "epoch": 2.42755360808736,
      "grad_norm": 0.16616354882717133,
      "learning_rate": 1.9084859079338793e-05,
      "loss": 0.5437,
      "mean_token_accuracy": 0.826896671205759,
      "num_tokens": 842088934.0,
      "step": 26385
    },
    {
      "entropy": 0.5691304467618465,
      "epoch": 2.4276456144588012,
      "grad_norm": 0.16632886230945587,
      "learning_rate": 1.9081792253197168e-05,
      "loss": 0.5508,
      "mean_token_accuracy": 0.8259717747569084,
      "num_tokens": 842120243.0,
      "step": 26386
    },
    {
      "entropy": 0.508428543806076,
      "epoch": 2.4277376208302424,
      "grad_norm": 0.16031059622764587,
      "learning_rate": 1.9078725427055542e-05,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8438622392714024,
      "num_tokens": 842151851.0,
      "step": 26387
    },
    {
      "entropy": 0.5717435176484287,
      "epoch": 2.4278296272016835,
      "grad_norm": 0.16833116114139557,
      "learning_rate": 1.9075658600913914e-05,
      "loss": 0.551,
      "mean_token_accuracy": 0.8261377476155758,
      "num_tokens": 842183111.0,
      "step": 26388
    },
    {
      "entropy": 0.5365288658067584,
      "epoch": 2.427921633573125,
      "grad_norm": 0.16139355301856995,
      "learning_rate": 1.907259177477229e-05,
      "loss": 0.5268,
      "mean_token_accuracy": 0.833586785942316,
      "num_tokens": 842215528.0,
      "step": 26389
    },
    {
      "entropy": 0.6250439547002316,
      "epoch": 2.428013639944566,
      "grad_norm": 0.16856400668621063,
      "learning_rate": 1.9069524948630663e-05,
      "loss": 0.6045,
      "mean_token_accuracy": 0.8095130696892738,
      "num_tokens": 842247638.0,
      "step": 26390
    },
    {
      "entropy": 0.4381772964261472,
      "epoch": 2.4281056463160073,
      "grad_norm": 0.14924092590808868,
      "learning_rate": 1.9066458122489038e-05,
      "loss": 0.4273,
      "mean_token_accuracy": 0.8644360639154911,
      "num_tokens": 842279644.0,
      "step": 26391
    },
    {
      "entropy": 0.49508773209527135,
      "epoch": 2.4281976526874485,
      "grad_norm": 0.15561601519584656,
      "learning_rate": 1.906339129634741e-05,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8456987403333187,
      "num_tokens": 842311808.0,
      "step": 26392
    },
    {
      "entropy": 0.4532238414976746,
      "epoch": 2.4282896590588896,
      "grad_norm": 0.1533341258764267,
      "learning_rate": 1.9060324470205784e-05,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8597273267805576,
      "num_tokens": 842343885.0,
      "step": 26393
    },
    {
      "entropy": 0.5414810217916965,
      "epoch": 2.428381665430331,
      "grad_norm": 0.15368421375751495,
      "learning_rate": 1.9057257644064158e-05,
      "loss": 0.5216,
      "mean_token_accuracy": 0.8376041576266289,
      "num_tokens": 842376181.0,
      "step": 26394
    },
    {
      "entropy": 0.506747517734766,
      "epoch": 2.4284736718017723,
      "grad_norm": 0.15920305252075195,
      "learning_rate": 1.9054190817922533e-05,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8436495773494244,
      "num_tokens": 842408705.0,
      "step": 26395
    },
    {
      "entropy": 0.47378688491880894,
      "epoch": 2.4285656781732134,
      "grad_norm": 0.1599651426076889,
      "learning_rate": 1.9051123991780908e-05,
      "loss": 0.4643,
      "mean_token_accuracy": 0.8542438261210918,
      "num_tokens": 842440158.0,
      "step": 26396
    },
    {
      "entropy": 0.49952477449551225,
      "epoch": 2.4286576845446546,
      "grad_norm": 0.15906760096549988,
      "learning_rate": 1.904805716563928e-05,
      "loss": 0.4856,
      "mean_token_accuracy": 0.846155509352684,
      "num_tokens": 842471549.0,
      "step": 26397
    },
    {
      "entropy": 0.4464201182126999,
      "epoch": 2.4287496909160957,
      "grad_norm": 0.15717452764511108,
      "learning_rate": 1.9044990339497653e-05,
      "loss": 0.446,
      "mean_token_accuracy": 0.856938436627388,
      "num_tokens": 842503292.0,
      "step": 26398
    },
    {
      "entropy": 0.5538794696331024,
      "epoch": 2.4288416972875373,
      "grad_norm": 0.1660597175359726,
      "learning_rate": 1.9041923513356028e-05,
      "loss": 0.5466,
      "mean_token_accuracy": 0.829655222594738,
      "num_tokens": 842535945.0,
      "step": 26399
    },
    {
      "entropy": 0.5141926323994994,
      "epoch": 2.4289337036589784,
      "grad_norm": 0.15657979249954224,
      "learning_rate": 1.9038856687214403e-05,
      "loss": 0.5046,
      "mean_token_accuracy": 0.84077313169837,
      "num_tokens": 842568696.0,
      "step": 26400
    },
    {
      "entropy": 0.5303751819301397,
      "epoch": 2.4290257100304196,
      "grad_norm": 0.1648673266172409,
      "learning_rate": 1.9035789861072777e-05,
      "loss": 0.5204,
      "mean_token_accuracy": 0.8345928974449635,
      "num_tokens": 842600231.0,
      "step": 26401
    },
    {
      "entropy": 0.5350354071706533,
      "epoch": 2.4291177164018607,
      "grad_norm": 0.16497959196567535,
      "learning_rate": 1.9032723034931152e-05,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8318389356136322,
      "num_tokens": 842632328.0,
      "step": 26402
    },
    {
      "entropy": 0.37780476873740554,
      "epoch": 2.429209722773302,
      "grad_norm": 0.14543357491493225,
      "learning_rate": 1.9029656208789527e-05,
      "loss": 0.3698,
      "mean_token_accuracy": 0.8807399049401283,
      "num_tokens": 842664647.0,
      "step": 26403
    },
    {
      "entropy": 0.5304428953677416,
      "epoch": 2.4293017291447434,
      "grad_norm": 0.15875627100467682,
      "learning_rate": 1.90265893826479e-05,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8308022655546665,
      "num_tokens": 842696642.0,
      "step": 26404
    },
    {
      "entropy": 0.5234688948839903,
      "epoch": 2.4293937355161845,
      "grad_norm": 0.16220252215862274,
      "learning_rate": 1.9023522556506273e-05,
      "loss": 0.509,
      "mean_token_accuracy": 0.8381756283342838,
      "num_tokens": 842728904.0,
      "step": 26405
    },
    {
      "entropy": 0.549917901866138,
      "epoch": 2.4294857418876257,
      "grad_norm": 0.16036202013492584,
      "learning_rate": 1.9020455730364647e-05,
      "loss": 0.54,
      "mean_token_accuracy": 0.8286069892346859,
      "num_tokens": 842761447.0,
      "step": 26406
    },
    {
      "entropy": 0.596154348924756,
      "epoch": 2.429577748259067,
      "grad_norm": 0.1686651110649109,
      "learning_rate": 1.9017388904223022e-05,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8129147887229919,
      "num_tokens": 842794114.0,
      "step": 26407
    },
    {
      "entropy": 0.5150248715654016,
      "epoch": 2.429669754630508,
      "grad_norm": 0.16301317512989044,
      "learning_rate": 1.9014322078081397e-05,
      "loss": 0.5073,
      "mean_token_accuracy": 0.8377518616616726,
      "num_tokens": 842825700.0,
      "step": 26408
    },
    {
      "entropy": 0.5705765541642904,
      "epoch": 2.4297617610019495,
      "grad_norm": 0.1686079055070877,
      "learning_rate": 1.9011255251939768e-05,
      "loss": 0.5602,
      "mean_token_accuracy": 0.8229309618473053,
      "num_tokens": 842857627.0,
      "step": 26409
    },
    {
      "entropy": 0.5022929066326469,
      "epoch": 2.4298537673733906,
      "grad_norm": 0.15603461861610413,
      "learning_rate": 1.9008188425798142e-05,
      "loss": 0.4898,
      "mean_token_accuracy": 0.8436843231320381,
      "num_tokens": 842889402.0,
      "step": 26410
    },
    {
      "entropy": 0.573685297742486,
      "epoch": 2.429945773744832,
      "grad_norm": 0.16710203886032104,
      "learning_rate": 1.9005121599656517e-05,
      "loss": 0.5694,
      "mean_token_accuracy": 0.819542795419693,
      "num_tokens": 842921560.0,
      "step": 26411
    },
    {
      "entropy": 0.5560936955735087,
      "epoch": 2.430037780116273,
      "grad_norm": 0.16854842007160187,
      "learning_rate": 1.9002054773514892e-05,
      "loss": 0.5463,
      "mean_token_accuracy": 0.8338998705148697,
      "num_tokens": 842953572.0,
      "step": 26412
    },
    {
      "entropy": 0.554400846362114,
      "epoch": 2.430129786487714,
      "grad_norm": 0.16289512813091278,
      "learning_rate": 1.8998987947373263e-05,
      "loss": 0.5372,
      "mean_token_accuracy": 0.8312795460224152,
      "num_tokens": 842985480.0,
      "step": 26413
    },
    {
      "entropy": 0.5135478805750608,
      "epoch": 2.4302217928591556,
      "grad_norm": 0.15819278359413147,
      "learning_rate": 1.8995921121231638e-05,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8432147204875946,
      "num_tokens": 843016900.0,
      "step": 26414
    },
    {
      "entropy": 0.4479727717116475,
      "epoch": 2.4303137992305968,
      "grad_norm": 0.15225189924240112,
      "learning_rate": 1.8992854295090012e-05,
      "loss": 0.43,
      "mean_token_accuracy": 0.8657206669449806,
      "num_tokens": 843049277.0,
      "step": 26415
    },
    {
      "entropy": 0.5782917588949203,
      "epoch": 2.430405805602038,
      "grad_norm": 0.17283381521701813,
      "learning_rate": 1.8989787468948387e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8221362084150314,
      "num_tokens": 843080418.0,
      "step": 26416
    },
    {
      "entropy": 0.5114531894214451,
      "epoch": 2.430497811973479,
      "grad_norm": 0.15788137912750244,
      "learning_rate": 1.898672064280676e-05,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8424296900629997,
      "num_tokens": 843112912.0,
      "step": 26417
    },
    {
      "entropy": 0.3926129201427102,
      "epoch": 2.43058981834492,
      "grad_norm": 0.14834360778331757,
      "learning_rate": 1.8983653816665133e-05,
      "loss": 0.3794,
      "mean_token_accuracy": 0.878724966198206,
      "num_tokens": 843145680.0,
      "step": 26418
    },
    {
      "entropy": 0.5368717974051833,
      "epoch": 2.4306818247163617,
      "grad_norm": 0.16566133499145508,
      "learning_rate": 1.8980586990523507e-05,
      "loss": 0.5285,
      "mean_token_accuracy": 0.8354651071131229,
      "num_tokens": 843177865.0,
      "step": 26419
    },
    {
      "entropy": 0.5447157649323344,
      "epoch": 2.430773831087803,
      "grad_norm": 0.1624753624200821,
      "learning_rate": 1.8977520164381882e-05,
      "loss": 0.524,
      "mean_token_accuracy": 0.8318611867725849,
      "num_tokens": 843209261.0,
      "step": 26420
    },
    {
      "entropy": 0.567595798522234,
      "epoch": 2.430865837459244,
      "grad_norm": 0.16578657925128937,
      "learning_rate": 1.8974453338240257e-05,
      "loss": 0.5605,
      "mean_token_accuracy": 0.8258330188691616,
      "num_tokens": 843241422.0,
      "step": 26421
    },
    {
      "entropy": 0.5479961773380637,
      "epoch": 2.430957843830685,
      "grad_norm": 0.16514690220355988,
      "learning_rate": 1.8971386512098628e-05,
      "loss": 0.537,
      "mean_token_accuracy": 0.8337521106004715,
      "num_tokens": 843273714.0,
      "step": 26422
    },
    {
      "entropy": 0.5945105161517859,
      "epoch": 2.4310498502021263,
      "grad_norm": 0.17239640653133392,
      "learning_rate": 1.8968319685957003e-05,
      "loss": 0.5919,
      "mean_token_accuracy": 0.818091943860054,
      "num_tokens": 843304820.0,
      "step": 26423
    },
    {
      "entropy": 0.5287240501493216,
      "epoch": 2.431141856573568,
      "grad_norm": 0.1622384637594223,
      "learning_rate": 1.8965252859815377e-05,
      "loss": 0.519,
      "mean_token_accuracy": 0.8352304548025131,
      "num_tokens": 843336577.0,
      "step": 26424
    },
    {
      "entropy": 0.5531487744301558,
      "epoch": 2.431233862945009,
      "grad_norm": 0.16076220571994781,
      "learning_rate": 1.8962186033673752e-05,
      "loss": 0.5401,
      "mean_token_accuracy": 0.8312461711466312,
      "num_tokens": 843368569.0,
      "step": 26425
    },
    {
      "entropy": 0.545215705409646,
      "epoch": 2.43132586931645,
      "grad_norm": 0.16423870623111725,
      "learning_rate": 1.8959119207532127e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8273362629115582,
      "num_tokens": 843400492.0,
      "step": 26426
    },
    {
      "entropy": 0.4097695597447455,
      "epoch": 2.4314178756878913,
      "grad_norm": 0.14655855298042297,
      "learning_rate": 1.8956052381390498e-05,
      "loss": 0.3943,
      "mean_token_accuracy": 0.8706651665270329,
      "num_tokens": 843432500.0,
      "step": 26427
    },
    {
      "entropy": 0.5229513943195343,
      "epoch": 2.4315098820593324,
      "grad_norm": 0.15998803079128265,
      "learning_rate": 1.8952985555248873e-05,
      "loss": 0.522,
      "mean_token_accuracy": 0.8374555706977844,
      "num_tokens": 843464937.0,
      "step": 26428
    },
    {
      "entropy": 0.46127200801856816,
      "epoch": 2.431601888430774,
      "grad_norm": 0.15474075078964233,
      "learning_rate": 1.8949918729107247e-05,
      "loss": 0.447,
      "mean_token_accuracy": 0.857968408614397,
      "num_tokens": 843496608.0,
      "step": 26429
    },
    {
      "entropy": 0.4721459543798119,
      "epoch": 2.431693894802215,
      "grad_norm": 0.1557067185640335,
      "learning_rate": 1.8946851902965622e-05,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8511115200817585,
      "num_tokens": 843528867.0,
      "step": 26430
    },
    {
      "entropy": 0.5286300405859947,
      "epoch": 2.4317859011736562,
      "grad_norm": 0.15563124418258667,
      "learning_rate": 1.8943785076823993e-05,
      "loss": 0.5145,
      "mean_token_accuracy": 0.8385481052100658,
      "num_tokens": 843560786.0,
      "step": 26431
    },
    {
      "entropy": 0.5302146538160741,
      "epoch": 2.4318779075450974,
      "grad_norm": 0.16369450092315674,
      "learning_rate": 1.894071825068237e-05,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8400898650288582,
      "num_tokens": 843592044.0,
      "step": 26432
    },
    {
      "entropy": 0.4862352660857141,
      "epoch": 2.4319699139165385,
      "grad_norm": 0.15705712139606476,
      "learning_rate": 1.8937651424540746e-05,
      "loss": 0.4741,
      "mean_token_accuracy": 0.8494297973811626,
      "num_tokens": 843623524.0,
      "step": 26433
    },
    {
      "entropy": 0.5613699499517679,
      "epoch": 2.43206192028798,
      "grad_norm": 0.1665760725736618,
      "learning_rate": 1.893458459839912e-05,
      "loss": 0.5433,
      "mean_token_accuracy": 0.8258237689733505,
      "num_tokens": 843655545.0,
      "step": 26434
    },
    {
      "entropy": 0.42755202297121286,
      "epoch": 2.432153926659421,
      "grad_norm": 0.1533334106206894,
      "learning_rate": 1.893151777225749e-05,
      "loss": 0.4226,
      "mean_token_accuracy": 0.8642271682620049,
      "num_tokens": 843687486.0,
      "step": 26435
    },
    {
      "entropy": 0.5508307926356792,
      "epoch": 2.4322459330308623,
      "grad_norm": 0.16363604366779327,
      "learning_rate": 1.8928450946115866e-05,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8295532837510109,
      "num_tokens": 843719251.0,
      "step": 26436
    },
    {
      "entropy": 0.5750130980741233,
      "epoch": 2.4323379394023035,
      "grad_norm": 0.16031308472156525,
      "learning_rate": 1.892538411997424e-05,
      "loss": 0.5592,
      "mean_token_accuracy": 0.8231790438294411,
      "num_tokens": 843751394.0,
      "step": 26437
    },
    {
      "entropy": 0.4593544630333781,
      "epoch": 2.4324299457737446,
      "grad_norm": 0.15498818457126617,
      "learning_rate": 1.8922317293832616e-05,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8535800576210022,
      "num_tokens": 843783580.0,
      "step": 26438
    },
    {
      "entropy": 0.5168400052934885,
      "epoch": 2.432521952145186,
      "grad_norm": 0.16212111711502075,
      "learning_rate": 1.8919250467690987e-05,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8405255749821663,
      "num_tokens": 843815015.0,
      "step": 26439
    },
    {
      "entropy": 0.5107456017285585,
      "epoch": 2.4326139585166273,
      "grad_norm": 0.16426922380924225,
      "learning_rate": 1.891618364154936e-05,
      "loss": 0.4968,
      "mean_token_accuracy": 0.8407432772219181,
      "num_tokens": 843845044.0,
      "step": 26440
    },
    {
      "entropy": 0.4860374005511403,
      "epoch": 2.4327059648880685,
      "grad_norm": 0.15542003512382507,
      "learning_rate": 1.8913116815407736e-05,
      "loss": 0.4699,
      "mean_token_accuracy": 0.8473839834332466,
      "num_tokens": 843877127.0,
      "step": 26441
    },
    {
      "entropy": 0.48779444582760334,
      "epoch": 2.4327979712595096,
      "grad_norm": 0.16310851275920868,
      "learning_rate": 1.891004998926611e-05,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8500749878585339,
      "num_tokens": 843908707.0,
      "step": 26442
    },
    {
      "entropy": 0.4257013383321464,
      "epoch": 2.4328899776309507,
      "grad_norm": 0.1475527137517929,
      "learning_rate": 1.8906983163124482e-05,
      "loss": 0.4153,
      "mean_token_accuracy": 0.8676545917987823,
      "num_tokens": 843940406.0,
      "step": 26443
    },
    {
      "entropy": 0.6361409611999989,
      "epoch": 2.4329819840023923,
      "grad_norm": 0.1715574562549591,
      "learning_rate": 1.8903916336982857e-05,
      "loss": 0.6249,
      "mean_token_accuracy": 0.8055192977190018,
      "num_tokens": 843972539.0,
      "step": 26444
    },
    {
      "entropy": 0.5862561732064933,
      "epoch": 2.4330739903738334,
      "grad_norm": 0.1692807376384735,
      "learning_rate": 1.890084951084123e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8204825483262539,
      "num_tokens": 844004705.0,
      "step": 26445
    },
    {
      "entropy": 0.5151276690885425,
      "epoch": 2.4331659967452746,
      "grad_norm": 0.15311890840530396,
      "learning_rate": 1.8897782684699606e-05,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8371082954108715,
      "num_tokens": 844036916.0,
      "step": 26446
    },
    {
      "entropy": 0.46743960375897586,
      "epoch": 2.4332580031167157,
      "grad_norm": 0.15214215219020844,
      "learning_rate": 1.889471585855798e-05,
      "loss": 0.4544,
      "mean_token_accuracy": 0.85458604991436,
      "num_tokens": 844069523.0,
      "step": 26447
    },
    {
      "entropy": 0.6020403634756804,
      "epoch": 2.433350009488157,
      "grad_norm": 0.1714521050453186,
      "learning_rate": 1.8891649032416352e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8170080929994583,
      "num_tokens": 844100879.0,
      "step": 26448
    },
    {
      "entropy": 0.5461707198992372,
      "epoch": 2.4334420158595984,
      "grad_norm": 0.16169370710849762,
      "learning_rate": 1.8888582206274727e-05,
      "loss": 0.5293,
      "mean_token_accuracy": 0.8302256613969803,
      "num_tokens": 844132471.0,
      "step": 26449
    },
    {
      "entropy": 0.4733623117208481,
      "epoch": 2.4335340222310395,
      "grad_norm": 0.15563073754310608,
      "learning_rate": 1.88855153801331e-05,
      "loss": 0.4628,
      "mean_token_accuracy": 0.8500926643610001,
      "num_tokens": 844164596.0,
      "step": 26450
    },
    {
      "entropy": 0.6047813650220633,
      "epoch": 2.4336260286024807,
      "grad_norm": 0.1689334660768509,
      "learning_rate": 1.8882448553991476e-05,
      "loss": 0.5995,
      "mean_token_accuracy": 0.8123579025268555,
      "num_tokens": 844196471.0,
      "step": 26451
    },
    {
      "entropy": 0.5000590747222304,
      "epoch": 2.433718034973922,
      "grad_norm": 0.15501920878887177,
      "learning_rate": 1.8879381727849847e-05,
      "loss": 0.485,
      "mean_token_accuracy": 0.8449008204042912,
      "num_tokens": 844228403.0,
      "step": 26452
    },
    {
      "entropy": 0.4718670751899481,
      "epoch": 2.433810041345363,
      "grad_norm": 0.1533588171005249,
      "learning_rate": 1.8876314901708222e-05,
      "loss": 0.4634,
      "mean_token_accuracy": 0.8522845543920994,
      "num_tokens": 844260509.0,
      "step": 26453
    },
    {
      "entropy": 0.4753407584503293,
      "epoch": 2.4339020477168045,
      "grad_norm": 0.16576911509037018,
      "learning_rate": 1.8873248075566596e-05,
      "loss": 0.4732,
      "mean_token_accuracy": 0.852683812379837,
      "num_tokens": 844291882.0,
      "step": 26454
    },
    {
      "entropy": 0.5143645610660315,
      "epoch": 2.4339940540882457,
      "grad_norm": 0.1663440316915512,
      "learning_rate": 1.887018124942497e-05,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8357102982699871,
      "num_tokens": 844323145.0,
      "step": 26455
    },
    {
      "entropy": 0.5303510185331106,
      "epoch": 2.434086060459687,
      "grad_norm": 0.1562710702419281,
      "learning_rate": 1.8867114423283346e-05,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8377726599574089,
      "num_tokens": 844355860.0,
      "step": 26456
    },
    {
      "entropy": 0.4617764725117013,
      "epoch": 2.434178066831128,
      "grad_norm": 0.14857591688632965,
      "learning_rate": 1.8864047597141717e-05,
      "loss": 0.4384,
      "mean_token_accuracy": 0.859980009496212,
      "num_tokens": 844387609.0,
      "step": 26457
    },
    {
      "entropy": 0.638699708506465,
      "epoch": 2.434270073202569,
      "grad_norm": 0.1664532870054245,
      "learning_rate": 1.886098077100009e-05,
      "loss": 0.6198,
      "mean_token_accuracy": 0.8027259930968285,
      "num_tokens": 844419949.0,
      "step": 26458
    },
    {
      "entropy": 0.6290216073393822,
      "epoch": 2.4343620795740106,
      "grad_norm": 0.17070871591567993,
      "learning_rate": 1.8857913944858466e-05,
      "loss": 0.6161,
      "mean_token_accuracy": 0.8056963346898556,
      "num_tokens": 844451625.0,
      "step": 26459
    },
    {
      "entropy": 0.5285453386604786,
      "epoch": 2.4344540859454518,
      "grad_norm": 0.1584034413099289,
      "learning_rate": 1.885484711871684e-05,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8389306478202343,
      "num_tokens": 844484142.0,
      "step": 26460
    },
    {
      "entropy": 0.5351865887641907,
      "epoch": 2.434546092316893,
      "grad_norm": 0.15800069272518158,
      "learning_rate": 1.8851780292575212e-05,
      "loss": 0.5145,
      "mean_token_accuracy": 0.838535126298666,
      "num_tokens": 844515944.0,
      "step": 26461
    },
    {
      "entropy": 0.4629022506996989,
      "epoch": 2.434638098688334,
      "grad_norm": 0.1551232784986496,
      "learning_rate": 1.884871346643359e-05,
      "loss": 0.4432,
      "mean_token_accuracy": 0.8574996516108513,
      "num_tokens": 844548348.0,
      "step": 26462
    },
    {
      "entropy": 0.4913308206014335,
      "epoch": 2.434730105059775,
      "grad_norm": 0.15265750885009766,
      "learning_rate": 1.8845646640291965e-05,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8479638397693634,
      "num_tokens": 844580275.0,
      "step": 26463
    },
    {
      "entropy": 0.5030977018177509,
      "epoch": 2.4348221114312167,
      "grad_norm": 0.15846049785614014,
      "learning_rate": 1.884257981415034e-05,
      "loss": 0.5006,
      "mean_token_accuracy": 0.842519111931324,
      "num_tokens": 844612561.0,
      "step": 26464
    },
    {
      "entropy": 0.4636552897281945,
      "epoch": 2.434914117802658,
      "grad_norm": 0.155861034989357,
      "learning_rate": 1.883951298800871e-05,
      "loss": 0.454,
      "mean_token_accuracy": 0.8542433604598045,
      "num_tokens": 844644480.0,
      "step": 26465
    },
    {
      "entropy": 0.47915175650268793,
      "epoch": 2.435006124174099,
      "grad_norm": 0.15424738824367523,
      "learning_rate": 1.8836446161867085e-05,
      "loss": 0.4689,
      "mean_token_accuracy": 0.850907240062952,
      "num_tokens": 844676434.0,
      "step": 26466
    },
    {
      "entropy": 0.5463482774794102,
      "epoch": 2.43509813054554,
      "grad_norm": 0.16593477129936218,
      "learning_rate": 1.883337933572546e-05,
      "loss": 0.5311,
      "mean_token_accuracy": 0.8315225541591644,
      "num_tokens": 844708413.0,
      "step": 26467
    },
    {
      "entropy": 0.4098454536870122,
      "epoch": 2.4351901369169813,
      "grad_norm": 0.1455949991941452,
      "learning_rate": 1.8830312509583835e-05,
      "loss": 0.4067,
      "mean_token_accuracy": 0.8714805319905281,
      "num_tokens": 844740773.0,
      "step": 26468
    },
    {
      "entropy": 0.5783689767122269,
      "epoch": 2.435282143288423,
      "grad_norm": 0.17196060717105865,
      "learning_rate": 1.8827245683442206e-05,
      "loss": 0.575,
      "mean_token_accuracy": 0.8173173815011978,
      "num_tokens": 844772315.0,
      "step": 26469
    },
    {
      "entropy": 0.6020664116367698,
      "epoch": 2.435374149659864,
      "grad_norm": 0.1746930032968521,
      "learning_rate": 1.882417885730058e-05,
      "loss": 0.5813,
      "mean_token_accuracy": 0.8197579421103001,
      "num_tokens": 844804389.0,
      "step": 26470
    },
    {
      "entropy": 0.5247920528054237,
      "epoch": 2.435466156031305,
      "grad_norm": 0.16272498667240143,
      "learning_rate": 1.8821112031158955e-05,
      "loss": 0.5131,
      "mean_token_accuracy": 0.8363627940416336,
      "num_tokens": 844836012.0,
      "step": 26471
    },
    {
      "entropy": 0.4948287596926093,
      "epoch": 2.4355581624027463,
      "grad_norm": 0.15564745664596558,
      "learning_rate": 1.881804520501733e-05,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8448917642235756,
      "num_tokens": 844868610.0,
      "step": 26472
    },
    {
      "entropy": 0.5656816326081753,
      "epoch": 2.4356501687741874,
      "grad_norm": 0.16754937171936035,
      "learning_rate": 1.88149783788757e-05,
      "loss": 0.5508,
      "mean_token_accuracy": 0.8234576359391212,
      "num_tokens": 844900323.0,
      "step": 26473
    },
    {
      "entropy": 0.49773271195590496,
      "epoch": 2.435742175145629,
      "grad_norm": 0.16508732736110687,
      "learning_rate": 1.8811911552734076e-05,
      "loss": 0.4802,
      "mean_token_accuracy": 0.8476957455277443,
      "num_tokens": 844932137.0,
      "step": 26474
    },
    {
      "entropy": 0.5422661025077105,
      "epoch": 2.43583418151707,
      "grad_norm": 0.1658744364976883,
      "learning_rate": 1.880884472659245e-05,
      "loss": 0.5326,
      "mean_token_accuracy": 0.8339910767972469,
      "num_tokens": 844964196.0,
      "step": 26475
    },
    {
      "entropy": 0.5437895683571696,
      "epoch": 2.4359261878885112,
      "grad_norm": 0.1624656319618225,
      "learning_rate": 1.8805777900450825e-05,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8326182700693607,
      "num_tokens": 844995715.0,
      "step": 26476
    },
    {
      "entropy": 0.5690386928617954,
      "epoch": 2.4360181942599524,
      "grad_norm": 0.17032332718372345,
      "learning_rate": 1.88027110743092e-05,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8262107968330383,
      "num_tokens": 845026792.0,
      "step": 26477
    },
    {
      "entropy": 0.5519465850666165,
      "epoch": 2.4361102006313935,
      "grad_norm": 0.16518859565258026,
      "learning_rate": 1.879964424816757e-05,
      "loss": 0.5455,
      "mean_token_accuracy": 0.8290327861905098,
      "num_tokens": 845059210.0,
      "step": 26478
    },
    {
      "entropy": 0.47338774241507053,
      "epoch": 2.436202207002835,
      "grad_norm": 0.15412692725658417,
      "learning_rate": 1.8796577422025946e-05,
      "loss": 0.462,
      "mean_token_accuracy": 0.8546199277043343,
      "num_tokens": 845091817.0,
      "step": 26479
    },
    {
      "entropy": 0.524635139037855,
      "epoch": 2.436294213374276,
      "grad_norm": 0.16146455705165863,
      "learning_rate": 1.879351059588432e-05,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8363635316491127,
      "num_tokens": 845124190.0,
      "step": 26480
    },
    {
      "entropy": 0.5402294341474771,
      "epoch": 2.4363862197457173,
      "grad_norm": 0.16220276057720184,
      "learning_rate": 1.8790443769742695e-05,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8318869210779667,
      "num_tokens": 845156057.0,
      "step": 26481
    },
    {
      "entropy": 0.5197883658111095,
      "epoch": 2.4364782261171585,
      "grad_norm": 0.1596360206604004,
      "learning_rate": 1.8787376943601066e-05,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8405693657696247,
      "num_tokens": 845187318.0,
      "step": 26482
    },
    {
      "entropy": 0.5468781930394471,
      "epoch": 2.4365702324885996,
      "grad_norm": 0.1661984920501709,
      "learning_rate": 1.878431011745944e-05,
      "loss": 0.5421,
      "mean_token_accuracy": 0.8321387581527233,
      "num_tokens": 845218358.0,
      "step": 26483
    },
    {
      "entropy": 0.5472626704722643,
      "epoch": 2.436662238860041,
      "grad_norm": 0.16631589829921722,
      "learning_rate": 1.8781243291317816e-05,
      "loss": 0.5302,
      "mean_token_accuracy": 0.8267102800309658,
      "num_tokens": 845249973.0,
      "step": 26484
    },
    {
      "entropy": 0.5261906404048204,
      "epoch": 2.4367542452314823,
      "grad_norm": 0.16010895371437073,
      "learning_rate": 1.877817646517619e-05,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8382529765367508,
      "num_tokens": 845281753.0,
      "step": 26485
    },
    {
      "entropy": 0.48264792934060097,
      "epoch": 2.4368462516029235,
      "grad_norm": 0.1557435393333435,
      "learning_rate": 1.8775109639034565e-05,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8497750423848629,
      "num_tokens": 845313378.0,
      "step": 26486
    },
    {
      "entropy": 0.488568427041173,
      "epoch": 2.4369382579743646,
      "grad_norm": 0.1631707400083542,
      "learning_rate": 1.8772042812892936e-05,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8506179004907608,
      "num_tokens": 845344827.0,
      "step": 26487
    },
    {
      "entropy": 0.5453840913251042,
      "epoch": 2.4370302643458057,
      "grad_norm": 0.1584467589855194,
      "learning_rate": 1.876897598675131e-05,
      "loss": 0.5399,
      "mean_token_accuracy": 0.8334460519254208,
      "num_tokens": 845377360.0,
      "step": 26488
    },
    {
      "entropy": 0.4876393638551235,
      "epoch": 2.4371222707172473,
      "grad_norm": 0.16365781426429749,
      "learning_rate": 1.8765909160609685e-05,
      "loss": 0.4829,
      "mean_token_accuracy": 0.8473892137408257,
      "num_tokens": 845409360.0,
      "step": 26489
    },
    {
      "entropy": 0.5414368342608213,
      "epoch": 2.4372142770886884,
      "grad_norm": 0.16447220742702484,
      "learning_rate": 1.876284233446806e-05,
      "loss": 0.5337,
      "mean_token_accuracy": 0.8305318132042885,
      "num_tokens": 845441147.0,
      "step": 26490
    },
    {
      "entropy": 0.5782316578552127,
      "epoch": 2.4373062834601296,
      "grad_norm": 0.16601069271564484,
      "learning_rate": 1.875977550832643e-05,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8213752619922161,
      "num_tokens": 845473026.0,
      "step": 26491
    },
    {
      "entropy": 0.49231055565178394,
      "epoch": 2.4373982898315707,
      "grad_norm": 0.15700644254684448,
      "learning_rate": 1.8756708682184806e-05,
      "loss": 0.4785,
      "mean_token_accuracy": 0.849011555314064,
      "num_tokens": 845504530.0,
      "step": 26492
    },
    {
      "entropy": 0.4998061787337065,
      "epoch": 2.437490296203012,
      "grad_norm": 0.15807406604290009,
      "learning_rate": 1.8753641856043184e-05,
      "loss": 0.4989,
      "mean_token_accuracy": 0.8426844142377377,
      "num_tokens": 845537041.0,
      "step": 26493
    },
    {
      "entropy": 0.6192526295781136,
      "epoch": 2.4375823025744534,
      "grad_norm": 0.18288707733154297,
      "learning_rate": 1.875057502990156e-05,
      "loss": 0.6122,
      "mean_token_accuracy": 0.8030492588877678,
      "num_tokens": 845567567.0,
      "step": 26494
    },
    {
      "entropy": 0.47695410531014204,
      "epoch": 2.4376743089458945,
      "grad_norm": 0.15891367197036743,
      "learning_rate": 1.874750820375993e-05,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8482069484889507,
      "num_tokens": 845598994.0,
      "step": 26495
    },
    {
      "entropy": 0.5384460296481848,
      "epoch": 2.4377663153173357,
      "grad_norm": 0.16200774908065796,
      "learning_rate": 1.8744441377618305e-05,
      "loss": 0.5179,
      "mean_token_accuracy": 0.8359892033040524,
      "num_tokens": 845631091.0,
      "step": 26496
    },
    {
      "entropy": 0.49195805145427585,
      "epoch": 2.437858321688777,
      "grad_norm": 0.15244647860527039,
      "learning_rate": 1.874137455147668e-05,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8495751358568668,
      "num_tokens": 845662752.0,
      "step": 26497
    },
    {
      "entropy": 0.5205183275975287,
      "epoch": 2.437950328060218,
      "grad_norm": 0.1555410921573639,
      "learning_rate": 1.8738307725335054e-05,
      "loss": 0.5085,
      "mean_token_accuracy": 0.8398414701223373,
      "num_tokens": 845694743.0,
      "step": 26498
    },
    {
      "entropy": 0.511821857187897,
      "epoch": 2.4380423344316595,
      "grad_norm": 0.15946052968502045,
      "learning_rate": 1.8735240899193425e-05,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8396888561546803,
      "num_tokens": 845726338.0,
      "step": 26499
    },
    {
      "entropy": 0.41497904784046113,
      "epoch": 2.4381343408031007,
      "grad_norm": 0.14245451986789703,
      "learning_rate": 1.87321740730518e-05,
      "loss": 0.3902,
      "mean_token_accuracy": 0.8756228312849998,
      "num_tokens": 845758377.0,
      "step": 26500
    },
    {
      "entropy": 0.5169788710772991,
      "epoch": 2.438226347174542,
      "grad_norm": 0.16340816020965576,
      "learning_rate": 1.8729107246910174e-05,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8389957398176193,
      "num_tokens": 845790255.0,
      "step": 26501
    },
    {
      "entropy": 0.5493694730103016,
      "epoch": 2.438318353545983,
      "grad_norm": 0.16670432686805725,
      "learning_rate": 1.872604042076855e-05,
      "loss": 0.5238,
      "mean_token_accuracy": 0.832218274474144,
      "num_tokens": 845822147.0,
      "step": 26502
    },
    {
      "entropy": 0.4774694324005395,
      "epoch": 2.438410359917424,
      "grad_norm": 0.1491498202085495,
      "learning_rate": 1.872297359462692e-05,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8514072336256504,
      "num_tokens": 845854307.0,
      "step": 26503
    },
    {
      "entropy": 0.5310746151953936,
      "epoch": 2.4385023662888656,
      "grad_norm": 0.16522721946239471,
      "learning_rate": 1.8719906768485295e-05,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8338141441345215,
      "num_tokens": 845885552.0,
      "step": 26504
    },
    {
      "entropy": 0.5195554653182626,
      "epoch": 2.4385943726603068,
      "grad_norm": 0.1566237211227417,
      "learning_rate": 1.871683994234367e-05,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8346198089420795,
      "num_tokens": 845917831.0,
      "step": 26505
    },
    {
      "entropy": 0.5313415955752134,
      "epoch": 2.438686379031748,
      "grad_norm": 0.1642470359802246,
      "learning_rate": 1.8713773116202044e-05,
      "loss": 0.5202,
      "mean_token_accuracy": 0.8337301909923553,
      "num_tokens": 845949089.0,
      "step": 26506
    },
    {
      "entropy": 0.49356034910306334,
      "epoch": 2.438778385403189,
      "grad_norm": 0.16136328876018524,
      "learning_rate": 1.871070629006042e-05,
      "loss": 0.4874,
      "mean_token_accuracy": 0.8455701768398285,
      "num_tokens": 845981400.0,
      "step": 26507
    },
    {
      "entropy": 0.5154252250213176,
      "epoch": 2.43887039177463,
      "grad_norm": 0.15491247177124023,
      "learning_rate": 1.870763946391879e-05,
      "loss": 0.4954,
      "mean_token_accuracy": 0.8396467231214046,
      "num_tokens": 846012928.0,
      "step": 26508
    },
    {
      "entropy": 0.5044732703827322,
      "epoch": 2.4389623981460717,
      "grad_norm": 0.1599520444869995,
      "learning_rate": 1.8704572637777165e-05,
      "loss": 0.4928,
      "mean_token_accuracy": 0.8442720361053944,
      "num_tokens": 846044880.0,
      "step": 26509
    },
    {
      "entropy": 0.4822880318388343,
      "epoch": 2.439054404517513,
      "grad_norm": 0.15488900244235992,
      "learning_rate": 1.870150581163554e-05,
      "loss": 0.4635,
      "mean_token_accuracy": 0.8504889123141766,
      "num_tokens": 846076670.0,
      "step": 26510
    },
    {
      "entropy": 0.540128780528903,
      "epoch": 2.439146410888954,
      "grad_norm": 0.16053970158100128,
      "learning_rate": 1.8698438985493914e-05,
      "loss": 0.545,
      "mean_token_accuracy": 0.8284750655293465,
      "num_tokens": 846108836.0,
      "step": 26511
    },
    {
      "entropy": 0.5499727576971054,
      "epoch": 2.439238417260395,
      "grad_norm": 0.16191799938678741,
      "learning_rate": 1.8695372159352285e-05,
      "loss": 0.5414,
      "mean_token_accuracy": 0.8284496963024139,
      "num_tokens": 846141349.0,
      "step": 26512
    },
    {
      "entropy": 0.6557327751070261,
      "epoch": 2.4393304236318363,
      "grad_norm": 0.17283737659454346,
      "learning_rate": 1.869230533321066e-05,
      "loss": 0.6344,
      "mean_token_accuracy": 0.8007085993885994,
      "num_tokens": 846173331.0,
      "step": 26513
    },
    {
      "entropy": 0.5511623192578554,
      "epoch": 2.439422430003278,
      "grad_norm": 0.1614423543214798,
      "learning_rate": 1.8689238507069035e-05,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8338131196796894,
      "num_tokens": 846205753.0,
      "step": 26514
    },
    {
      "entropy": 0.5472668316215277,
      "epoch": 2.439514436374719,
      "grad_norm": 0.16540482640266418,
      "learning_rate": 1.868617168092741e-05,
      "loss": 0.53,
      "mean_token_accuracy": 0.82961181178689,
      "num_tokens": 846238089.0,
      "step": 26515
    },
    {
      "entropy": 0.5349667863920331,
      "epoch": 2.43960644274616,
      "grad_norm": 0.16307352483272552,
      "learning_rate": 1.8683104854785784e-05,
      "loss": 0.5204,
      "mean_token_accuracy": 0.8363047614693642,
      "num_tokens": 846270644.0,
      "step": 26516
    },
    {
      "entropy": 0.5070954160764813,
      "epoch": 2.4396984491176013,
      "grad_norm": 0.15420368313789368,
      "learning_rate": 1.8680038028644155e-05,
      "loss": 0.5001,
      "mean_token_accuracy": 0.8454757779836655,
      "num_tokens": 846302804.0,
      "step": 26517
    },
    {
      "entropy": 0.4562827639747411,
      "epoch": 2.4397904554890424,
      "grad_norm": 0.1511267125606537,
      "learning_rate": 1.867697120250253e-05,
      "loss": 0.4467,
      "mean_token_accuracy": 0.8579824827611446,
      "num_tokens": 846334333.0,
      "step": 26518
    },
    {
      "entropy": 0.44073343463242054,
      "epoch": 2.439882461860484,
      "grad_norm": 0.1525479257106781,
      "learning_rate": 1.8673904376360904e-05,
      "loss": 0.4385,
      "mean_token_accuracy": 0.8606233932077885,
      "num_tokens": 846366695.0,
      "step": 26519
    },
    {
      "entropy": 0.39282661490142345,
      "epoch": 2.439974468231925,
      "grad_norm": 0.14507198333740234,
      "learning_rate": 1.867083755021928e-05,
      "loss": 0.3863,
      "mean_token_accuracy": 0.8741831444203854,
      "num_tokens": 846398550.0,
      "step": 26520
    },
    {
      "entropy": 0.38844721973873675,
      "epoch": 2.4400664746033662,
      "grad_norm": 0.1407991498708725,
      "learning_rate": 1.866777072407765e-05,
      "loss": 0.3758,
      "mean_token_accuracy": 0.8799747042357922,
      "num_tokens": 846430552.0,
      "step": 26521
    },
    {
      "entropy": 0.5819631181657314,
      "epoch": 2.4401584809748074,
      "grad_norm": 0.16496753692626953,
      "learning_rate": 1.8664703897936025e-05,
      "loss": 0.5632,
      "mean_token_accuracy": 0.8219483718276024,
      "num_tokens": 846461523.0,
      "step": 26522
    },
    {
      "entropy": 0.43176891142502427,
      "epoch": 2.4402504873462485,
      "grad_norm": 0.14567537605762482,
      "learning_rate": 1.8661637071794403e-05,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8682219386100769,
      "num_tokens": 846493722.0,
      "step": 26523
    },
    {
      "entropy": 0.4924029354006052,
      "epoch": 2.44034249371769,
      "grad_norm": 0.15720777213573456,
      "learning_rate": 1.8658570245652778e-05,
      "loss": 0.468,
      "mean_token_accuracy": 0.8488271608948708,
      "num_tokens": 846524664.0,
      "step": 26524
    },
    {
      "entropy": 0.4213243881240487,
      "epoch": 2.440434500089131,
      "grad_norm": 0.1468995213508606,
      "learning_rate": 1.865550341951115e-05,
      "loss": 0.4135,
      "mean_token_accuracy": 0.8669003024697304,
      "num_tokens": 846557017.0,
      "step": 26525
    },
    {
      "entropy": 0.5388267934322357,
      "epoch": 2.4405265064605723,
      "grad_norm": 0.17223197221755981,
      "learning_rate": 1.8652436593369524e-05,
      "loss": 0.5381,
      "mean_token_accuracy": 0.8316255211830139,
      "num_tokens": 846587986.0,
      "step": 26526
    },
    {
      "entropy": 0.4349020039662719,
      "epoch": 2.4406185128320135,
      "grad_norm": 0.14473088085651398,
      "learning_rate": 1.8649369767227898e-05,
      "loss": 0.4115,
      "mean_token_accuracy": 0.8665308393537998,
      "num_tokens": 846619896.0,
      "step": 26527
    },
    {
      "entropy": 0.45154910162091255,
      "epoch": 2.4407105192034546,
      "grad_norm": 0.1572769433259964,
      "learning_rate": 1.8646302941086273e-05,
      "loss": 0.4411,
      "mean_token_accuracy": 0.8557233437895775,
      "num_tokens": 846650581.0,
      "step": 26528
    },
    {
      "entropy": 0.4533130875788629,
      "epoch": 2.440802525574896,
      "grad_norm": 0.15467266738414764,
      "learning_rate": 1.8643236114944644e-05,
      "loss": 0.435,
      "mean_token_accuracy": 0.8618859462440014,
      "num_tokens": 846682294.0,
      "step": 26529
    },
    {
      "entropy": 0.4429628795478493,
      "epoch": 2.4408945319463373,
      "grad_norm": 0.15145574510097504,
      "learning_rate": 1.864016928880302e-05,
      "loss": 0.4337,
      "mean_token_accuracy": 0.8602314181625843,
      "num_tokens": 846714089.0,
      "step": 26530
    },
    {
      "entropy": 0.6347021656110883,
      "epoch": 2.4409865383177785,
      "grad_norm": 0.16981753706932068,
      "learning_rate": 1.8637102462661393e-05,
      "loss": 0.6202,
      "mean_token_accuracy": 0.8065297938883305,
      "num_tokens": 846746030.0,
      "step": 26531
    },
    {
      "entropy": 0.42200299235992134,
      "epoch": 2.4410785446892196,
      "grad_norm": 0.14808374643325806,
      "learning_rate": 1.8634035636519768e-05,
      "loss": 0.4156,
      "mean_token_accuracy": 0.8659582883119583,
      "num_tokens": 846778451.0,
      "step": 26532
    },
    {
      "entropy": 0.5413486175239086,
      "epoch": 2.4411705510606607,
      "grad_norm": 0.16413143277168274,
      "learning_rate": 1.863096881037814e-05,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8328112326562405,
      "num_tokens": 846810739.0,
      "step": 26533
    },
    {
      "entropy": 0.5335094649344683,
      "epoch": 2.4412625574321023,
      "grad_norm": 0.1587821990251541,
      "learning_rate": 1.8627901984236514e-05,
      "loss": 0.5131,
      "mean_token_accuracy": 0.8365808017551899,
      "num_tokens": 846843413.0,
      "step": 26534
    },
    {
      "entropy": 0.5175756961107254,
      "epoch": 2.4413545638035434,
      "grad_norm": 0.1583515703678131,
      "learning_rate": 1.862483515809489e-05,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8406030647456646,
      "num_tokens": 846874984.0,
      "step": 26535
    },
    {
      "entropy": 0.5304574258625507,
      "epoch": 2.4414465701749846,
      "grad_norm": 0.16338886320590973,
      "learning_rate": 1.8621768331953263e-05,
      "loss": 0.5144,
      "mean_token_accuracy": 0.8395533300936222,
      "num_tokens": 846906689.0,
      "step": 26536
    },
    {
      "entropy": 0.5373472105711699,
      "epoch": 2.4415385765464257,
      "grad_norm": 0.16789962351322174,
      "learning_rate": 1.8618701505811638e-05,
      "loss": 0.5328,
      "mean_token_accuracy": 0.8308085612952709,
      "num_tokens": 846938289.0,
      "step": 26537
    },
    {
      "entropy": 0.6227989178150892,
      "epoch": 2.441630582917867,
      "grad_norm": 0.1679866760969162,
      "learning_rate": 1.861563467967001e-05,
      "loss": 0.6055,
      "mean_token_accuracy": 0.8090629801154137,
      "num_tokens": 846970349.0,
      "step": 26538
    },
    {
      "entropy": 0.5029485132545233,
      "epoch": 2.4417225892893084,
      "grad_norm": 0.16299989819526672,
      "learning_rate": 1.8612567853528384e-05,
      "loss": 0.482,
      "mean_token_accuracy": 0.8489068225026131,
      "num_tokens": 847001935.0,
      "step": 26539
    },
    {
      "entropy": 0.5236786124296486,
      "epoch": 2.4418145956607495,
      "grad_norm": 0.15963660180568695,
      "learning_rate": 1.860950102738676e-05,
      "loss": 0.5097,
      "mean_token_accuracy": 0.8383860103785992,
      "num_tokens": 847034275.0,
      "step": 26540
    },
    {
      "entropy": 0.5901141837239265,
      "epoch": 2.4419066020321907,
      "grad_norm": 0.16785979270935059,
      "learning_rate": 1.8606434201245133e-05,
      "loss": 0.571,
      "mean_token_accuracy": 0.8217617459595203,
      "num_tokens": 847065849.0,
      "step": 26541
    },
    {
      "entropy": 0.4914428163319826,
      "epoch": 2.441998608403632,
      "grad_norm": 0.1599930077791214,
      "learning_rate": 1.8603367375103504e-05,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8465044014155865,
      "num_tokens": 847098077.0,
      "step": 26542
    },
    {
      "entropy": 0.4647759674116969,
      "epoch": 2.442090614775073,
      "grad_norm": 0.1463533341884613,
      "learning_rate": 1.860030054896188e-05,
      "loss": 0.4587,
      "mean_token_accuracy": 0.8540050536394119,
      "num_tokens": 847130002.0,
      "step": 26543
    },
    {
      "entropy": 0.5541572635993361,
      "epoch": 2.4421826211465145,
      "grad_norm": 0.1613193154335022,
      "learning_rate": 1.8597233722820254e-05,
      "loss": 0.5469,
      "mean_token_accuracy": 0.8318381644785404,
      "num_tokens": 847162461.0,
      "step": 26544
    },
    {
      "entropy": 0.49421111587435007,
      "epoch": 2.4422746275179557,
      "grad_norm": 0.15455909073352814,
      "learning_rate": 1.859416689667863e-05,
      "loss": 0.4826,
      "mean_token_accuracy": 0.8452801629900932,
      "num_tokens": 847194668.0,
      "step": 26545
    },
    {
      "entropy": 0.5922011351212859,
      "epoch": 2.442366633889397,
      "grad_norm": 0.17172591388225555,
      "learning_rate": 1.8591100070537003e-05,
      "loss": 0.5867,
      "mean_token_accuracy": 0.8158132284879684,
      "num_tokens": 847225709.0,
      "step": 26546
    },
    {
      "entropy": 0.5173297170549631,
      "epoch": 2.442458640260838,
      "grad_norm": 0.15712973475456238,
      "learning_rate": 1.8588033244395374e-05,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8414908200502396,
      "num_tokens": 847258037.0,
      "step": 26547
    },
    {
      "entropy": 0.5343514382839203,
      "epoch": 2.442550646632279,
      "grad_norm": 0.15706805884838104,
      "learning_rate": 1.858496641825375e-05,
      "loss": 0.5189,
      "mean_token_accuracy": 0.8373443819582462,
      "num_tokens": 847290452.0,
      "step": 26548
    },
    {
      "entropy": 0.5377911385148764,
      "epoch": 2.4426426530037206,
      "grad_norm": 0.16441108286380768,
      "learning_rate": 1.8581899592112124e-05,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8327508606016636,
      "num_tokens": 847321619.0,
      "step": 26549
    },
    {
      "entropy": 0.41248698672279716,
      "epoch": 2.4427346593751618,
      "grad_norm": 0.1459769308567047,
      "learning_rate": 1.8578832765970498e-05,
      "loss": 0.4012,
      "mean_token_accuracy": 0.8700261153280735,
      "num_tokens": 847353788.0,
      "step": 26550
    },
    {
      "entropy": 0.5519882179796696,
      "epoch": 2.442826665746603,
      "grad_norm": 0.16373786330223083,
      "learning_rate": 1.857576593982887e-05,
      "loss": 0.5371,
      "mean_token_accuracy": 0.8310258500277996,
      "num_tokens": 847385859.0,
      "step": 26551
    },
    {
      "entropy": 0.5501159597188234,
      "epoch": 2.442918672118044,
      "grad_norm": 0.1640264093875885,
      "learning_rate": 1.8572699113687244e-05,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8310233131051064,
      "num_tokens": 847418088.0,
      "step": 26552
    },
    {
      "entropy": 0.5150203483644873,
      "epoch": 2.443010678489485,
      "grad_norm": 0.15921460092067719,
      "learning_rate": 1.856963228754562e-05,
      "loss": 0.5156,
      "mean_token_accuracy": 0.8378450945019722,
      "num_tokens": 847450147.0,
      "step": 26553
    },
    {
      "entropy": 0.5041139628738165,
      "epoch": 2.4431026848609267,
      "grad_norm": 0.15283288061618805,
      "learning_rate": 1.8566565461403997e-05,
      "loss": 0.4829,
      "mean_token_accuracy": 0.8467734977602959,
      "num_tokens": 847482018.0,
      "step": 26554
    },
    {
      "entropy": 0.4531585378572345,
      "epoch": 2.443194691232368,
      "grad_norm": 0.1494520604610443,
      "learning_rate": 1.8563498635262368e-05,
      "loss": 0.4415,
      "mean_token_accuracy": 0.8578243441879749,
      "num_tokens": 847514370.0,
      "step": 26555
    },
    {
      "entropy": 0.5547129884362221,
      "epoch": 2.443286697603809,
      "grad_norm": 0.1631826013326645,
      "learning_rate": 1.8560431809120743e-05,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8308441787958145,
      "num_tokens": 847545947.0,
      "step": 26556
    },
    {
      "entropy": 0.7005666960030794,
      "epoch": 2.44337870397525,
      "grad_norm": 0.17866073548793793,
      "learning_rate": 1.8557364982979117e-05,
      "loss": 0.6849,
      "mean_token_accuracy": 0.7881284281611443,
      "num_tokens": 847578143.0,
      "step": 26557
    },
    {
      "entropy": 0.48899462539702654,
      "epoch": 2.4434707103466913,
      "grad_norm": 0.16010811924934387,
      "learning_rate": 1.8554298156837492e-05,
      "loss": 0.4761,
      "mean_token_accuracy": 0.8502323105931282,
      "num_tokens": 847610237.0,
      "step": 26558
    },
    {
      "entropy": 0.48509029392153025,
      "epoch": 2.443562716718133,
      "grad_norm": 0.15294532477855682,
      "learning_rate": 1.8551231330695863e-05,
      "loss": 0.4645,
      "mean_token_accuracy": 0.8499435894191265,
      "num_tokens": 847641773.0,
      "step": 26559
    },
    {
      "entropy": 0.5185509445145726,
      "epoch": 2.443654723089574,
      "grad_norm": 0.16138572990894318,
      "learning_rate": 1.8548164504554238e-05,
      "loss": 0.5131,
      "mean_token_accuracy": 0.8374294117093086,
      "num_tokens": 847673293.0,
      "step": 26560
    },
    {
      "entropy": 0.4501164183020592,
      "epoch": 2.443746729461015,
      "grad_norm": 0.1519862711429596,
      "learning_rate": 1.8545097678412613e-05,
      "loss": 0.4436,
      "mean_token_accuracy": 0.8592123799026012,
      "num_tokens": 847705511.0,
      "step": 26561
    },
    {
      "entropy": 0.5280965778511018,
      "epoch": 2.4438387358324563,
      "grad_norm": 0.1617654412984848,
      "learning_rate": 1.8542030852270987e-05,
      "loss": 0.5103,
      "mean_token_accuracy": 0.8345306478440762,
      "num_tokens": 847737582.0,
      "step": 26562
    },
    {
      "entropy": 0.47611571941524744,
      "epoch": 2.4439307422038974,
      "grad_norm": 0.15460152924060822,
      "learning_rate": 1.853896402612936e-05,
      "loss": 0.47,
      "mean_token_accuracy": 0.8539528883993626,
      "num_tokens": 847769809.0,
      "step": 26563
    },
    {
      "entropy": 0.5439370591193438,
      "epoch": 2.444022748575339,
      "grad_norm": 0.1604483425617218,
      "learning_rate": 1.8535897199987733e-05,
      "loss": 0.5365,
      "mean_token_accuracy": 0.830622635781765,
      "num_tokens": 847801746.0,
      "step": 26564
    },
    {
      "entropy": 0.5139882926596329,
      "epoch": 2.44411475494678,
      "grad_norm": 0.16074596345424652,
      "learning_rate": 1.8532830373846108e-05,
      "loss": 0.5003,
      "mean_token_accuracy": 0.840091809630394,
      "num_tokens": 847834003.0,
      "step": 26565
    },
    {
      "entropy": 0.4616657104343176,
      "epoch": 2.4442067613182212,
      "grad_norm": 0.15460404753684998,
      "learning_rate": 1.8529763547704482e-05,
      "loss": 0.4624,
      "mean_token_accuracy": 0.8538376465439796,
      "num_tokens": 847865804.0,
      "step": 26566
    },
    {
      "entropy": 0.5517394915223122,
      "epoch": 2.4442987676896624,
      "grad_norm": 0.16331028938293457,
      "learning_rate": 1.8526696721562857e-05,
      "loss": 0.5424,
      "mean_token_accuracy": 0.827212031930685,
      "num_tokens": 847898070.0,
      "step": 26567
    },
    {
      "entropy": 0.5868012211285532,
      "epoch": 2.4443907740611035,
      "grad_norm": 0.16604594886302948,
      "learning_rate": 1.852362989542123e-05,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8245608061552048,
      "num_tokens": 847929459.0,
      "step": 26568
    },
    {
      "entropy": 0.4744437262415886,
      "epoch": 2.444482780432545,
      "grad_norm": 0.15514987707138062,
      "learning_rate": 1.8520563069279603e-05,
      "loss": 0.4674,
      "mean_token_accuracy": 0.8498916663229465,
      "num_tokens": 847961536.0,
      "step": 26569
    },
    {
      "entropy": 0.5998685900121927,
      "epoch": 2.444574786803986,
      "grad_norm": 0.1696644127368927,
      "learning_rate": 1.8517496243137978e-05,
      "loss": 0.5869,
      "mean_token_accuracy": 0.811835415661335,
      "num_tokens": 847993279.0,
      "step": 26570
    },
    {
      "entropy": 0.5342165417969227,
      "epoch": 2.4446667931754273,
      "grad_norm": 0.16029515862464905,
      "learning_rate": 1.8514429416996352e-05,
      "loss": 0.5322,
      "mean_token_accuracy": 0.8338080123066902,
      "num_tokens": 848025715.0,
      "step": 26571
    },
    {
      "entropy": 0.5813234527595341,
      "epoch": 2.4447587995468685,
      "grad_norm": 0.1703179031610489,
      "learning_rate": 1.8511362590854724e-05,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8238332010805607,
      "num_tokens": 848058408.0,
      "step": 26572
    },
    {
      "entropy": 0.4018043056130409,
      "epoch": 2.4448508059183096,
      "grad_norm": 0.1455545425415039,
      "learning_rate": 1.8508295764713098e-05,
      "loss": 0.3879,
      "mean_token_accuracy": 0.87525350227952,
      "num_tokens": 848090578.0,
      "step": 26573
    },
    {
      "entropy": 0.5184605638496578,
      "epoch": 2.444942812289751,
      "grad_norm": 0.159664586186409,
      "learning_rate": 1.8505228938571473e-05,
      "loss": 0.5075,
      "mean_token_accuracy": 0.8413572981953621,
      "num_tokens": 848122145.0,
      "step": 26574
    },
    {
      "entropy": 0.5383615307509899,
      "epoch": 2.4450348186611923,
      "grad_norm": 0.15957333147525787,
      "learning_rate": 1.8502162112429847e-05,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8380520045757294,
      "num_tokens": 848153879.0,
      "step": 26575
    },
    {
      "entropy": 0.4609917178750038,
      "epoch": 2.4451268250326335,
      "grad_norm": 0.1536044478416443,
      "learning_rate": 1.8499095286288222e-05,
      "loss": 0.4468,
      "mean_token_accuracy": 0.8539705313742161,
      "num_tokens": 848185763.0,
      "step": 26576
    },
    {
      "entropy": 0.566728160250932,
      "epoch": 2.4452188314040746,
      "grad_norm": 0.16566285490989685,
      "learning_rate": 1.8496028460146593e-05,
      "loss": 0.5555,
      "mean_token_accuracy": 0.8239620104432106,
      "num_tokens": 848218281.0,
      "step": 26577
    },
    {
      "entropy": 0.5345778530463576,
      "epoch": 2.4453108377755157,
      "grad_norm": 0.1628141701221466,
      "learning_rate": 1.8492961634004968e-05,
      "loss": 0.5279,
      "mean_token_accuracy": 0.8314963020384312,
      "num_tokens": 848250468.0,
      "step": 26578
    },
    {
      "entropy": 0.531460126163438,
      "epoch": 2.4454028441469573,
      "grad_norm": 0.15633848309516907,
      "learning_rate": 1.8489894807863343e-05,
      "loss": 0.5126,
      "mean_token_accuracy": 0.8389285914599895,
      "num_tokens": 848282303.0,
      "step": 26579
    },
    {
      "entropy": 0.574982182122767,
      "epoch": 2.4454948505183984,
      "grad_norm": 0.16380955278873444,
      "learning_rate": 1.8486827981721717e-05,
      "loss": 0.5491,
      "mean_token_accuracy": 0.8260463699698448,
      "num_tokens": 848314472.0,
      "step": 26580
    },
    {
      "entropy": 0.5172215141355991,
      "epoch": 2.4455868568898396,
      "grad_norm": 0.15760117769241333,
      "learning_rate": 1.848376115558009e-05,
      "loss": 0.51,
      "mean_token_accuracy": 0.8389918133616447,
      "num_tokens": 848347013.0,
      "step": 26581
    },
    {
      "entropy": 0.5220510261133313,
      "epoch": 2.4456788632612807,
      "grad_norm": 0.16193220019340515,
      "learning_rate": 1.8480694329438463e-05,
      "loss": 0.5149,
      "mean_token_accuracy": 0.8384122885763645,
      "num_tokens": 848379167.0,
      "step": 26582
    },
    {
      "entropy": 0.537972828373313,
      "epoch": 2.445770869632722,
      "grad_norm": 0.15981072187423706,
      "learning_rate": 1.8477627503296838e-05,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8354108817875385,
      "num_tokens": 848411317.0,
      "step": 26583
    },
    {
      "entropy": 0.49579527974128723,
      "epoch": 2.4458628760041634,
      "grad_norm": 0.1548408418893814,
      "learning_rate": 1.8474560677155216e-05,
      "loss": 0.4788,
      "mean_token_accuracy": 0.846980944275856,
      "num_tokens": 848443226.0,
      "step": 26584
    },
    {
      "entropy": 0.5450845267623663,
      "epoch": 2.4459548823756045,
      "grad_norm": 0.16129811108112335,
      "learning_rate": 1.8471493851013587e-05,
      "loss": 0.5177,
      "mean_token_accuracy": 0.8383013233542442,
      "num_tokens": 848474958.0,
      "step": 26585
    },
    {
      "entropy": 0.4706589002162218,
      "epoch": 2.4460468887470457,
      "grad_norm": 0.1517769992351532,
      "learning_rate": 1.8468427024871962e-05,
      "loss": 0.4568,
      "mean_token_accuracy": 0.8533821105957031,
      "num_tokens": 848507445.0,
      "step": 26586
    },
    {
      "entropy": 0.5429400773718953,
      "epoch": 2.446138895118487,
      "grad_norm": 0.16439801454544067,
      "learning_rate": 1.8465360198730336e-05,
      "loss": 0.5358,
      "mean_token_accuracy": 0.8307908624410629,
      "num_tokens": 848539321.0,
      "step": 26587
    },
    {
      "entropy": 0.5241084042936563,
      "epoch": 2.446230901489928,
      "grad_norm": 0.15635375678539276,
      "learning_rate": 1.846229337258871e-05,
      "loss": 0.5142,
      "mean_token_accuracy": 0.8372743837535381,
      "num_tokens": 848571648.0,
      "step": 26588
    },
    {
      "entropy": 0.5669491980224848,
      "epoch": 2.4463229078613695,
      "grad_norm": 0.16445651650428772,
      "learning_rate": 1.8459226546447082e-05,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8216875307261944,
      "num_tokens": 848603373.0,
      "step": 26589
    },
    {
      "entropy": 0.4266950888559222,
      "epoch": 2.4464149142328107,
      "grad_norm": 0.1524423211812973,
      "learning_rate": 1.8456159720305457e-05,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8675834126770496,
      "num_tokens": 848635630.0,
      "step": 26590
    },
    {
      "entropy": 0.5247887363657355,
      "epoch": 2.446506920604252,
      "grad_norm": 0.16030281782150269,
      "learning_rate": 1.845309289416383e-05,
      "loss": 0.5073,
      "mean_token_accuracy": 0.8391844630241394,
      "num_tokens": 848667669.0,
      "step": 26591
    },
    {
      "entropy": 0.5212376099079847,
      "epoch": 2.446598926975693,
      "grad_norm": 0.15966767072677612,
      "learning_rate": 1.8450026068022206e-05,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8404590673744678,
      "num_tokens": 848699438.0,
      "step": 26592
    },
    {
      "entropy": 0.47433702554553747,
      "epoch": 2.446690933347134,
      "grad_norm": 0.15560553967952728,
      "learning_rate": 1.8446959241880578e-05,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8492574170231819,
      "num_tokens": 848731479.0,
      "step": 26593
    },
    {
      "entropy": 0.4825132703408599,
      "epoch": 2.4467829397185756,
      "grad_norm": 0.15179699659347534,
      "learning_rate": 1.8443892415738952e-05,
      "loss": 0.475,
      "mean_token_accuracy": 0.8496580347418785,
      "num_tokens": 848764247.0,
      "step": 26594
    },
    {
      "entropy": 0.4690811721375212,
      "epoch": 2.4468749460900168,
      "grad_norm": 0.15746818482875824,
      "learning_rate": 1.8440825589597327e-05,
      "loss": 0.4606,
      "mean_token_accuracy": 0.8546302728354931,
      "num_tokens": 848796583.0,
      "step": 26595
    },
    {
      "entropy": 0.5154937696643174,
      "epoch": 2.446966952461458,
      "grad_norm": 0.15818707644939423,
      "learning_rate": 1.84377587634557e-05,
      "loss": 0.5069,
      "mean_token_accuracy": 0.8404110036790371,
      "num_tokens": 848828578.0,
      "step": 26596
    },
    {
      "entropy": 0.5618832837790251,
      "epoch": 2.447058958832899,
      "grad_norm": 0.16608327627182007,
      "learning_rate": 1.8434691937314076e-05,
      "loss": 0.561,
      "mean_token_accuracy": 0.8210473693907261,
      "num_tokens": 848860654.0,
      "step": 26597
    },
    {
      "entropy": 0.3980123372748494,
      "epoch": 2.44715096520434,
      "grad_norm": 0.1460348218679428,
      "learning_rate": 1.8431625111172447e-05,
      "loss": 0.3929,
      "mean_token_accuracy": 0.8726183101534843,
      "num_tokens": 848892895.0,
      "step": 26598
    },
    {
      "entropy": 0.5493304282426834,
      "epoch": 2.4472429715757817,
      "grad_norm": 0.16480591893196106,
      "learning_rate": 1.8428558285030822e-05,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8281089402735233,
      "num_tokens": 848925253.0,
      "step": 26599
    },
    {
      "entropy": 0.5258596912026405,
      "epoch": 2.447334977947223,
      "grad_norm": 0.15846146643161774,
      "learning_rate": 1.8425491458889197e-05,
      "loss": 0.5113,
      "mean_token_accuracy": 0.8363987915217876,
      "num_tokens": 848957005.0,
      "step": 26600
    },
    {
      "entropy": 0.48861092515289783,
      "epoch": 2.447426984318664,
      "grad_norm": 0.16220547258853912,
      "learning_rate": 1.842242463274757e-05,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8477108031511307,
      "num_tokens": 848987939.0,
      "step": 26601
    },
    {
      "entropy": 0.4353056442923844,
      "epoch": 2.447518990690105,
      "grad_norm": 0.1486930549144745,
      "learning_rate": 1.8419357806605943e-05,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8606814332306385,
      "num_tokens": 849019844.0,
      "step": 26602
    },
    {
      "entropy": 0.5149758821353316,
      "epoch": 2.4476109970615463,
      "grad_norm": 0.1578935831785202,
      "learning_rate": 1.8416290980464317e-05,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8389247097074986,
      "num_tokens": 849052362.0,
      "step": 26603
    },
    {
      "entropy": 0.6458366475999355,
      "epoch": 2.447703003432988,
      "grad_norm": 0.17619992792606354,
      "learning_rate": 1.8413224154322692e-05,
      "loss": 0.6363,
      "mean_token_accuracy": 0.8030641637742519,
      "num_tokens": 849084659.0,
      "step": 26604
    },
    {
      "entropy": 0.5401275549083948,
      "epoch": 2.447795009804429,
      "grad_norm": 0.16348546743392944,
      "learning_rate": 1.8410157328181067e-05,
      "loss": 0.5292,
      "mean_token_accuracy": 0.832720447331667,
      "num_tokens": 849116504.0,
      "step": 26605
    },
    {
      "entropy": 0.3973962508607656,
      "epoch": 2.44788701617587,
      "grad_norm": 0.1375880241394043,
      "learning_rate": 1.840709050203944e-05,
      "loss": 0.3929,
      "mean_token_accuracy": 0.8759637959301472,
      "num_tokens": 849148493.0,
      "step": 26606
    },
    {
      "entropy": 0.6075845239683986,
      "epoch": 2.4479790225473113,
      "grad_norm": 0.17244426906108856,
      "learning_rate": 1.8404023675897812e-05,
      "loss": 0.5849,
      "mean_token_accuracy": 0.8166434280574322,
      "num_tokens": 849180306.0,
      "step": 26607
    },
    {
      "entropy": 0.4023278448730707,
      "epoch": 2.4480710289187524,
      "grad_norm": 0.1409095972776413,
      "learning_rate": 1.8400956849756187e-05,
      "loss": 0.3783,
      "mean_token_accuracy": 0.8758637607097626,
      "num_tokens": 849212342.0,
      "step": 26608
    },
    {
      "entropy": 0.527533614076674,
      "epoch": 2.448163035290194,
      "grad_norm": 0.15658335387706757,
      "learning_rate": 1.8397890023614562e-05,
      "loss": 0.5096,
      "mean_token_accuracy": 0.8376094214618206,
      "num_tokens": 849244506.0,
      "step": 26609
    },
    {
      "entropy": 0.48238059505820274,
      "epoch": 2.448255041661635,
      "grad_norm": 0.1516648828983307,
      "learning_rate": 1.8394823197472936e-05,
      "loss": 0.4629,
      "mean_token_accuracy": 0.8520570434629917,
      "num_tokens": 849277147.0,
      "step": 26610
    },
    {
      "entropy": 0.6266539767384529,
      "epoch": 2.4483470480330762,
      "grad_norm": 0.17544278502464294,
      "learning_rate": 1.8391756371331308e-05,
      "loss": 0.6206,
      "mean_token_accuracy": 0.8074145205318928,
      "num_tokens": 849309015.0,
      "step": 26611
    },
    {
      "entropy": 0.5171671421267092,
      "epoch": 2.4484390544045174,
      "grad_norm": 0.16321054100990295,
      "learning_rate": 1.8388689545189682e-05,
      "loss": 0.5047,
      "mean_token_accuracy": 0.8424532674252987,
      "num_tokens": 849340987.0,
      "step": 26612
    },
    {
      "entropy": 0.4390060221776366,
      "epoch": 2.4485310607759585,
      "grad_norm": 0.14845627546310425,
      "learning_rate": 1.8385622719048057e-05,
      "loss": 0.4191,
      "mean_token_accuracy": 0.8667279481887817,
      "num_tokens": 849372995.0,
      "step": 26613
    },
    {
      "entropy": 0.5557491881772876,
      "epoch": 2.4486230671474,
      "grad_norm": 0.16660597920417786,
      "learning_rate": 1.838255589290643e-05,
      "loss": 0.5504,
      "mean_token_accuracy": 0.8232576064765453,
      "num_tokens": 849404889.0,
      "step": 26614
    },
    {
      "entropy": 0.4619093043729663,
      "epoch": 2.448715073518841,
      "grad_norm": 0.1471141129732132,
      "learning_rate": 1.8379489066764806e-05,
      "loss": 0.4436,
      "mean_token_accuracy": 0.8574857823550701,
      "num_tokens": 849436710.0,
      "step": 26615
    },
    {
      "entropy": 0.48294012201949954,
      "epoch": 2.4488070798902823,
      "grad_norm": 0.15415653586387634,
      "learning_rate": 1.837642224062318e-05,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8471771851181984,
      "num_tokens": 849469075.0,
      "step": 26616
    },
    {
      "entropy": 0.4923661109060049,
      "epoch": 2.4488990862617235,
      "grad_norm": 0.15711908042430878,
      "learning_rate": 1.8373355414481556e-05,
      "loss": 0.4846,
      "mean_token_accuracy": 0.8443669676780701,
      "num_tokens": 849501422.0,
      "step": 26617
    },
    {
      "entropy": 0.5673062652349472,
      "epoch": 2.4489910926331646,
      "grad_norm": 0.16419053077697754,
      "learning_rate": 1.837028858833993e-05,
      "loss": 0.5515,
      "mean_token_accuracy": 0.8237439692020416,
      "num_tokens": 849533905.0,
      "step": 26618
    },
    {
      "entropy": 0.49948814837262034,
      "epoch": 2.449083099004606,
      "grad_norm": 0.1522749662399292,
      "learning_rate": 1.83672217621983e-05,
      "loss": 0.4846,
      "mean_token_accuracy": 0.847310334444046,
      "num_tokens": 849566148.0,
      "step": 26619
    },
    {
      "entropy": 0.4398367127869278,
      "epoch": 2.4491751053760473,
      "grad_norm": 0.14836078882217407,
      "learning_rate": 1.8364154936056676e-05,
      "loss": 0.4197,
      "mean_token_accuracy": 0.8672992326319218,
      "num_tokens": 849598738.0,
      "step": 26620
    },
    {
      "entropy": 0.4484964357689023,
      "epoch": 2.4492671117474885,
      "grad_norm": 0.14300374686717987,
      "learning_rate": 1.836108810991505e-05,
      "loss": 0.437,
      "mean_token_accuracy": 0.8605951555073261,
      "num_tokens": 849631120.0,
      "step": 26621
    },
    {
      "entropy": 0.5023294860729948,
      "epoch": 2.4493591181189296,
      "grad_norm": 0.156409353017807,
      "learning_rate": 1.8358021283773425e-05,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8480755425989628,
      "num_tokens": 849662446.0,
      "step": 26622
    },
    {
      "entropy": 0.4597419190686196,
      "epoch": 2.4494511244903707,
      "grad_norm": 0.15369774401187897,
      "learning_rate": 1.8354954457631797e-05,
      "loss": 0.4434,
      "mean_token_accuracy": 0.857905350625515,
      "num_tokens": 849694609.0,
      "step": 26623
    },
    {
      "entropy": 0.5570343043655157,
      "epoch": 2.4495431308618123,
      "grad_norm": 0.16687346994876862,
      "learning_rate": 1.835188763149017e-05,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8290382809937,
      "num_tokens": 849726820.0,
      "step": 26624
    },
    {
      "entropy": 0.5187474526464939,
      "epoch": 2.4496351372332534,
      "grad_norm": 0.15541744232177734,
      "learning_rate": 1.8348820805348546e-05,
      "loss": 0.5146,
      "mean_token_accuracy": 0.8396404311060905,
      "num_tokens": 849759437.0,
      "step": 26625
    },
    {
      "entropy": 0.4302801387384534,
      "epoch": 2.4497271436046946,
      "grad_norm": 0.14984720945358276,
      "learning_rate": 1.834575397920692e-05,
      "loss": 0.4239,
      "mean_token_accuracy": 0.8648076318204403,
      "num_tokens": 849791338.0,
      "step": 26626
    },
    {
      "entropy": 0.5130422245711088,
      "epoch": 2.4498191499761357,
      "grad_norm": 0.15678462386131287,
      "learning_rate": 1.8342687153065295e-05,
      "loss": 0.5048,
      "mean_token_accuracy": 0.8371996134519577,
      "num_tokens": 849823386.0,
      "step": 26627
    },
    {
      "entropy": 0.5348533252254128,
      "epoch": 2.449911156347577,
      "grad_norm": 0.1592174768447876,
      "learning_rate": 1.8339620326923667e-05,
      "loss": 0.5234,
      "mean_token_accuracy": 0.8365197740495205,
      "num_tokens": 849855789.0,
      "step": 26628
    },
    {
      "entropy": 0.5241115968674421,
      "epoch": 2.4500031627190184,
      "grad_norm": 0.16492266952991486,
      "learning_rate": 1.833655350078204e-05,
      "loss": 0.5149,
      "mean_token_accuracy": 0.8350529298186302,
      "num_tokens": 849887425.0,
      "step": 26629
    },
    {
      "entropy": 0.5576102510094643,
      "epoch": 2.4500951690904595,
      "grad_norm": 0.1678747832775116,
      "learning_rate": 1.8333486674640416e-05,
      "loss": 0.5664,
      "mean_token_accuracy": 0.8267012350261211,
      "num_tokens": 849919740.0,
      "step": 26630
    },
    {
      "entropy": 0.53632936719805,
      "epoch": 2.4501871754619007,
      "grad_norm": 0.15803293883800507,
      "learning_rate": 1.833041984849879e-05,
      "loss": 0.5233,
      "mean_token_accuracy": 0.830446470528841,
      "num_tokens": 849951513.0,
      "step": 26631
    },
    {
      "entropy": 0.5141473961994052,
      "epoch": 2.450279181833342,
      "grad_norm": 0.1638219803571701,
      "learning_rate": 1.8327353022357162e-05,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8406458832323551,
      "num_tokens": 849983202.0,
      "step": 26632
    },
    {
      "entropy": 0.5346006322652102,
      "epoch": 2.450371188204783,
      "grad_norm": 0.16242893040180206,
      "learning_rate": 1.8324286196215536e-05,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8362992480397224,
      "num_tokens": 850015393.0,
      "step": 26633
    },
    {
      "entropy": 0.5016578882932663,
      "epoch": 2.4504631945762245,
      "grad_norm": 0.15883463621139526,
      "learning_rate": 1.832121937007391e-05,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8421708419919014,
      "num_tokens": 850047171.0,
      "step": 26634
    },
    {
      "entropy": 0.5165531504899263,
      "epoch": 2.4505552009476657,
      "grad_norm": 0.15952220559120178,
      "learning_rate": 1.8318152543932286e-05,
      "loss": 0.5018,
      "mean_token_accuracy": 0.8418941386044025,
      "num_tokens": 850079425.0,
      "step": 26635
    },
    {
      "entropy": 0.5966058429330587,
      "epoch": 2.450647207319107,
      "grad_norm": 0.16664338111877441,
      "learning_rate": 1.831508571779066e-05,
      "loss": 0.5781,
      "mean_token_accuracy": 0.8228876739740372,
      "num_tokens": 850111550.0,
      "step": 26636
    },
    {
      "entropy": 0.5782178714871407,
      "epoch": 2.450739213690548,
      "grad_norm": 0.17323993146419525,
      "learning_rate": 1.831201889164903e-05,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8209275715053082,
      "num_tokens": 850143245.0,
      "step": 26637
    },
    {
      "entropy": 0.5477541852742434,
      "epoch": 2.450831220061989,
      "grad_norm": 0.16175755858421326,
      "learning_rate": 1.8308952065507406e-05,
      "loss": 0.5322,
      "mean_token_accuracy": 0.8288710229098797,
      "num_tokens": 850175360.0,
      "step": 26638
    },
    {
      "entropy": 0.5801816135644913,
      "epoch": 2.4509232264334306,
      "grad_norm": 0.17866311967372894,
      "learning_rate": 1.830588523936578e-05,
      "loss": 0.5894,
      "mean_token_accuracy": 0.813357125967741,
      "num_tokens": 850205406.0,
      "step": 26639
    },
    {
      "entropy": 0.5890486594289541,
      "epoch": 2.4510152328048718,
      "grad_norm": 0.1675492376089096,
      "learning_rate": 1.8302818413224156e-05,
      "loss": 0.5646,
      "mean_token_accuracy": 0.8204379752278328,
      "num_tokens": 850237446.0,
      "step": 26640
    },
    {
      "entropy": 0.5006236946210265,
      "epoch": 2.451107239176313,
      "grad_norm": 0.15162739157676697,
      "learning_rate": 1.8299751587082527e-05,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8433481156826019,
      "num_tokens": 850269229.0,
      "step": 26641
    },
    {
      "entropy": 0.47729366400744766,
      "epoch": 2.451199245547754,
      "grad_norm": 0.15656541287899017,
      "learning_rate": 1.82966847609409e-05,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8501835279166698,
      "num_tokens": 850301603.0,
      "step": 26642
    },
    {
      "entropy": 0.6119969002902508,
      "epoch": 2.451291251919195,
      "grad_norm": 0.17023608088493347,
      "learning_rate": 1.8293617934799276e-05,
      "loss": 0.5981,
      "mean_token_accuracy": 0.8128780797123909,
      "num_tokens": 850333806.0,
      "step": 26643
    },
    {
      "entropy": 0.48245594184845686,
      "epoch": 2.4513832582906367,
      "grad_norm": 0.15760329365730286,
      "learning_rate": 1.829055110865765e-05,
      "loss": 0.46,
      "mean_token_accuracy": 0.8535502105951309,
      "num_tokens": 850365668.0,
      "step": 26644
    },
    {
      "entropy": 0.6097371838986874,
      "epoch": 2.451475264662078,
      "grad_norm": 0.1691809743642807,
      "learning_rate": 1.8287484282516025e-05,
      "loss": 0.5966,
      "mean_token_accuracy": 0.8132809400558472,
      "num_tokens": 850397206.0,
      "step": 26645
    },
    {
      "entropy": 0.6163602918386459,
      "epoch": 2.451567271033519,
      "grad_norm": 0.17404916882514954,
      "learning_rate": 1.82844174563744e-05,
      "loss": 0.6077,
      "mean_token_accuracy": 0.8106948919594288,
      "num_tokens": 850428919.0,
      "step": 26646
    },
    {
      "entropy": 0.4850107138045132,
      "epoch": 2.45165927740496,
      "grad_norm": 0.15107597410678864,
      "learning_rate": 1.8281350630232775e-05,
      "loss": 0.4679,
      "mean_token_accuracy": 0.8530442416667938,
      "num_tokens": 850461635.0,
      "step": 26647
    },
    {
      "entropy": 0.4895791634917259,
      "epoch": 2.4517512837764013,
      "grad_norm": 0.1525314599275589,
      "learning_rate": 1.827828380409115e-05,
      "loss": 0.477,
      "mean_token_accuracy": 0.8437366299331188,
      "num_tokens": 850493773.0,
      "step": 26648
    },
    {
      "entropy": 0.4906221032142639,
      "epoch": 2.451843290147843,
      "grad_norm": 0.15880268812179565,
      "learning_rate": 1.827521697794952e-05,
      "loss": 0.4763,
      "mean_token_accuracy": 0.8499248251318932,
      "num_tokens": 850525648.0,
      "step": 26649
    },
    {
      "entropy": 0.5160438632592559,
      "epoch": 2.451935296519284,
      "grad_norm": 0.1600552797317505,
      "learning_rate": 1.8272150151807895e-05,
      "loss": 0.514,
      "mean_token_accuracy": 0.836430698633194,
      "num_tokens": 850557897.0,
      "step": 26650
    },
    {
      "entropy": 0.48010733001865447,
      "epoch": 2.452027302890725,
      "grad_norm": 0.15512505173683167,
      "learning_rate": 1.826908332566627e-05,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8499287962913513,
      "num_tokens": 850589095.0,
      "step": 26651
    },
    {
      "entropy": 0.5288979751057923,
      "epoch": 2.4521193092621663,
      "grad_norm": 0.15992572903633118,
      "learning_rate": 1.8266016499524645e-05,
      "loss": 0.5218,
      "mean_token_accuracy": 0.8359533995389938,
      "num_tokens": 850621150.0,
      "step": 26652
    },
    {
      "entropy": 0.5486281882040203,
      "epoch": 2.4522113156336074,
      "grad_norm": 0.16535386443138123,
      "learning_rate": 1.8262949673383016e-05,
      "loss": 0.5328,
      "mean_token_accuracy": 0.8297893181443214,
      "num_tokens": 850653169.0,
      "step": 26653
    },
    {
      "entropy": 0.6205538287758827,
      "epoch": 2.452303322005049,
      "grad_norm": 0.1680159568786621,
      "learning_rate": 1.825988284724139e-05,
      "loss": 0.5963,
      "mean_token_accuracy": 0.8105881661176682,
      "num_tokens": 850684777.0,
      "step": 26654
    },
    {
      "entropy": 0.532672893255949,
      "epoch": 2.45239532837649,
      "grad_norm": 0.16393299400806427,
      "learning_rate": 1.8256816021099765e-05,
      "loss": 0.5293,
      "mean_token_accuracy": 0.8308009244501591,
      "num_tokens": 850716639.0,
      "step": 26655
    },
    {
      "entropy": 0.4736121064051986,
      "epoch": 2.4524873347479312,
      "grad_norm": 0.16489970684051514,
      "learning_rate": 1.825374919495814e-05,
      "loss": 0.4701,
      "mean_token_accuracy": 0.8545492105185986,
      "num_tokens": 850748088.0,
      "step": 26656
    },
    {
      "entropy": 0.49915322568267584,
      "epoch": 2.4525793411193724,
      "grad_norm": 0.15290111303329468,
      "learning_rate": 1.8250682368816514e-05,
      "loss": 0.4805,
      "mean_token_accuracy": 0.8480291180312634,
      "num_tokens": 850780280.0,
      "step": 26657
    },
    {
      "entropy": 0.4356419234536588,
      "epoch": 2.4526713474908135,
      "grad_norm": 0.1475526988506317,
      "learning_rate": 1.8247615542674886e-05,
      "loss": 0.4184,
      "mean_token_accuracy": 0.8662915863096714,
      "num_tokens": 850811826.0,
      "step": 26658
    },
    {
      "entropy": 0.5162115348502994,
      "epoch": 2.452763353862255,
      "grad_norm": 0.16063323616981506,
      "learning_rate": 1.824454871653326e-05,
      "loss": 0.5107,
      "mean_token_accuracy": 0.8386348225176334,
      "num_tokens": 850843850.0,
      "step": 26659
    },
    {
      "entropy": 0.5408323155716062,
      "epoch": 2.452855360233696,
      "grad_norm": 0.16425886750221252,
      "learning_rate": 1.8241481890391635e-05,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8336763866245747,
      "num_tokens": 850876177.0,
      "step": 26660
    },
    {
      "entropy": 0.47622725926339626,
      "epoch": 2.4529473666051373,
      "grad_norm": 0.15521526336669922,
      "learning_rate": 1.823841506425001e-05,
      "loss": 0.4576,
      "mean_token_accuracy": 0.851473655551672,
      "num_tokens": 850907563.0,
      "step": 26661
    },
    {
      "entropy": 0.45898999366909266,
      "epoch": 2.4530393729765785,
      "grad_norm": 0.15123850107192993,
      "learning_rate": 1.823534823810838e-05,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8556359671056271,
      "num_tokens": 850939366.0,
      "step": 26662
    },
    {
      "entropy": 0.46173661202192307,
      "epoch": 2.4531313793480196,
      "grad_norm": 0.1504283994436264,
      "learning_rate": 1.8232281411966755e-05,
      "loss": 0.441,
      "mean_token_accuracy": 0.8549915365874767,
      "num_tokens": 850970424.0,
      "step": 26663
    },
    {
      "entropy": 0.5588316535577178,
      "epoch": 2.453223385719461,
      "grad_norm": 0.16327565908432007,
      "learning_rate": 1.822921458582513e-05,
      "loss": 0.5558,
      "mean_token_accuracy": 0.8255641125142574,
      "num_tokens": 851002537.0,
      "step": 26664
    },
    {
      "entropy": 0.4861026583239436,
      "epoch": 2.4533153920909023,
      "grad_norm": 0.15591205656528473,
      "learning_rate": 1.8226147759683505e-05,
      "loss": 0.467,
      "mean_token_accuracy": 0.8539020605385303,
      "num_tokens": 851034506.0,
      "step": 26665
    },
    {
      "entropy": 0.4768067253753543,
      "epoch": 2.4534073984623435,
      "grad_norm": 0.1638830453157425,
      "learning_rate": 1.822308093354188e-05,
      "loss": 0.4626,
      "mean_token_accuracy": 0.8551753610372543,
      "num_tokens": 851065962.0,
      "step": 26666
    },
    {
      "entropy": 0.5423914436250925,
      "epoch": 2.4534994048337846,
      "grad_norm": 0.16700656712055206,
      "learning_rate": 1.822001410740025e-05,
      "loss": 0.531,
      "mean_token_accuracy": 0.8303971067070961,
      "num_tokens": 851098185.0,
      "step": 26667
    },
    {
      "entropy": 0.4485407853499055,
      "epoch": 2.4535914112052257,
      "grad_norm": 0.14700640738010406,
      "learning_rate": 1.8216947281258625e-05,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8599894754588604,
      "num_tokens": 851130517.0,
      "step": 26668
    },
    {
      "entropy": 0.5294692534953356,
      "epoch": 2.4536834175766673,
      "grad_norm": 0.16288946568965912,
      "learning_rate": 1.8213880455117e-05,
      "loss": 0.5149,
      "mean_token_accuracy": 0.8357882760465145,
      "num_tokens": 851163146.0,
      "step": 26669
    },
    {
      "entropy": 0.5092012835666537,
      "epoch": 2.4537754239481084,
      "grad_norm": 0.15332794189453125,
      "learning_rate": 1.8210813628975375e-05,
      "loss": 0.5114,
      "mean_token_accuracy": 0.8427055925130844,
      "num_tokens": 851195022.0,
      "step": 26670
    },
    {
      "entropy": 0.527759918011725,
      "epoch": 2.4538674303195496,
      "grad_norm": 0.1608283370733261,
      "learning_rate": 1.8207746802833746e-05,
      "loss": 0.5184,
      "mean_token_accuracy": 0.8360292315483093,
      "num_tokens": 851227167.0,
      "step": 26671
    },
    {
      "entropy": 0.4430248262360692,
      "epoch": 2.4539594366909907,
      "grad_norm": 0.15134674310684204,
      "learning_rate": 1.820467997669212e-05,
      "loss": 0.4326,
      "mean_token_accuracy": 0.8624053336679935,
      "num_tokens": 851258495.0,
      "step": 26672
    },
    {
      "entropy": 0.6080379590857774,
      "epoch": 2.454051443062432,
      "grad_norm": 0.16836650669574738,
      "learning_rate": 1.8201613150550495e-05,
      "loss": 0.5915,
      "mean_token_accuracy": 0.812820740044117,
      "num_tokens": 851290244.0,
      "step": 26673
    },
    {
      "entropy": 0.43579821940511465,
      "epoch": 2.4541434494338734,
      "grad_norm": 0.14855755865573883,
      "learning_rate": 1.819854632440887e-05,
      "loss": 0.4166,
      "mean_token_accuracy": 0.8644827157258987,
      "num_tokens": 851322610.0,
      "step": 26674
    },
    {
      "entropy": 0.5651506949216127,
      "epoch": 2.4542354558053145,
      "grad_norm": 0.16688032448291779,
      "learning_rate": 1.819547949826724e-05,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8252209238708019,
      "num_tokens": 851354353.0,
      "step": 26675
    },
    {
      "entropy": 0.6190228052437305,
      "epoch": 2.4543274621767557,
      "grad_norm": 0.175596222281456,
      "learning_rate": 1.819241267212562e-05,
      "loss": 0.6157,
      "mean_token_accuracy": 0.807651162147522,
      "num_tokens": 851385803.0,
      "step": 26676
    },
    {
      "entropy": 0.5624864392448217,
      "epoch": 2.454419468548197,
      "grad_norm": 0.16449807584285736,
      "learning_rate": 1.8189345845983994e-05,
      "loss": 0.5525,
      "mean_token_accuracy": 0.82438775151968,
      "num_tokens": 851417890.0,
      "step": 26677
    },
    {
      "entropy": 0.5055980309844017,
      "epoch": 2.454511474919638,
      "grad_norm": 0.15972024202346802,
      "learning_rate": 1.818627901984237e-05,
      "loss": 0.4909,
      "mean_token_accuracy": 0.8445568978786469,
      "num_tokens": 851449919.0,
      "step": 26678
    },
    {
      "entropy": 0.446243685670197,
      "epoch": 2.4546034812910795,
      "grad_norm": 0.14770659804344177,
      "learning_rate": 1.818321219370074e-05,
      "loss": 0.4221,
      "mean_token_accuracy": 0.864593718200922,
      "num_tokens": 851482420.0,
      "step": 26679
    },
    {
      "entropy": 0.49311858462169766,
      "epoch": 2.4546954876625207,
      "grad_norm": 0.15495805442333221,
      "learning_rate": 1.8180145367559114e-05,
      "loss": 0.4789,
      "mean_token_accuracy": 0.8479847311973572,
      "num_tokens": 851514698.0,
      "step": 26680
    },
    {
      "entropy": 0.4553789743222296,
      "epoch": 2.454787494033962,
      "grad_norm": 0.15230512619018555,
      "learning_rate": 1.817707854141749e-05,
      "loss": 0.4364,
      "mean_token_accuracy": 0.8601022846996784,
      "num_tokens": 851545127.0,
      "step": 26681
    },
    {
      "entropy": 0.5119462627917528,
      "epoch": 2.454879500405403,
      "grad_norm": 0.16018636524677277,
      "learning_rate": 1.8174011715275864e-05,
      "loss": 0.4921,
      "mean_token_accuracy": 0.8426245860755444,
      "num_tokens": 851577038.0,
      "step": 26682
    },
    {
      "entropy": 0.41922272252850235,
      "epoch": 2.454971506776844,
      "grad_norm": 0.14464543759822845,
      "learning_rate": 1.8170944889134235e-05,
      "loss": 0.3991,
      "mean_token_accuracy": 0.8717711456120014,
      "num_tokens": 851608731.0,
      "step": 26683
    },
    {
      "entropy": 0.36261895718052983,
      "epoch": 2.4550635131482856,
      "grad_norm": 0.13733772933483124,
      "learning_rate": 1.816787806299261e-05,
      "loss": 0.3518,
      "mean_token_accuracy": 0.8826399929821491,
      "num_tokens": 851641421.0,
      "step": 26684
    },
    {
      "entropy": 0.4856330493930727,
      "epoch": 2.4551555195197268,
      "grad_norm": 0.15526016056537628,
      "learning_rate": 1.8164811236850984e-05,
      "loss": 0.484,
      "mean_token_accuracy": 0.8458799421787262,
      "num_tokens": 851673415.0,
      "step": 26685
    },
    {
      "entropy": 0.42934050410985947,
      "epoch": 2.455247525891168,
      "grad_norm": 0.15277506411075592,
      "learning_rate": 1.816174441070936e-05,
      "loss": 0.4172,
      "mean_token_accuracy": 0.8641983009874821,
      "num_tokens": 851705659.0,
      "step": 26686
    },
    {
      "entropy": 0.5026910174638033,
      "epoch": 2.455339532262609,
      "grad_norm": 0.157871276140213,
      "learning_rate": 1.8158677584567733e-05,
      "loss": 0.4749,
      "mean_token_accuracy": 0.8446915969252586,
      "num_tokens": 851737080.0,
      "step": 26687
    },
    {
      "entropy": 0.45160098746418953,
      "epoch": 2.45543153863405,
      "grad_norm": 0.15597669780254364,
      "learning_rate": 1.8155610758426105e-05,
      "loss": 0.4406,
      "mean_token_accuracy": 0.859335046261549,
      "num_tokens": 851768339.0,
      "step": 26688
    },
    {
      "entropy": 0.6569693312048912,
      "epoch": 2.4555235450054917,
      "grad_norm": 0.1791491061449051,
      "learning_rate": 1.815254393228448e-05,
      "loss": 0.6483,
      "mean_token_accuracy": 0.7957662679255009,
      "num_tokens": 851800054.0,
      "step": 26689
    },
    {
      "entropy": 0.4677579812705517,
      "epoch": 2.455615551376933,
      "grad_norm": 0.15788649022579193,
      "learning_rate": 1.8149477106142854e-05,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8536218293011189,
      "num_tokens": 851831702.0,
      "step": 26690
    },
    {
      "entropy": 0.49566161073744297,
      "epoch": 2.455707557748374,
      "grad_norm": 0.15777134895324707,
      "learning_rate": 1.814641028000123e-05,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8499150760471821,
      "num_tokens": 851863531.0,
      "step": 26691
    },
    {
      "entropy": 0.5081593114882708,
      "epoch": 2.455799564119815,
      "grad_norm": 0.15334776043891907,
      "learning_rate": 1.81433434538596e-05,
      "loss": 0.4941,
      "mean_token_accuracy": 0.8430546671152115,
      "num_tokens": 851895615.0,
      "step": 26692
    },
    {
      "entropy": 0.4558687340468168,
      "epoch": 2.4558915704912563,
      "grad_norm": 0.15579569339752197,
      "learning_rate": 1.8140276627717975e-05,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8546346798539162,
      "num_tokens": 851927685.0,
      "step": 26693
    },
    {
      "entropy": 0.5025640479288995,
      "epoch": 2.455983576862698,
      "grad_norm": 0.15983597934246063,
      "learning_rate": 1.813720980157635e-05,
      "loss": 0.493,
      "mean_token_accuracy": 0.8434934616088867,
      "num_tokens": 851959860.0,
      "step": 26694
    },
    {
      "entropy": 0.5305973794311285,
      "epoch": 2.456075583234139,
      "grad_norm": 0.16246291995048523,
      "learning_rate": 1.8134142975434724e-05,
      "loss": 0.5198,
      "mean_token_accuracy": 0.8328907266259193,
      "num_tokens": 851991377.0,
      "step": 26695
    },
    {
      "entropy": 0.5934715969488025,
      "epoch": 2.45616758960558,
      "grad_norm": 0.16497336328029633,
      "learning_rate": 1.81310761492931e-05,
      "loss": 0.579,
      "mean_token_accuracy": 0.815651960670948,
      "num_tokens": 852023213.0,
      "step": 26696
    },
    {
      "entropy": 0.4537722086533904,
      "epoch": 2.4562595959770213,
      "grad_norm": 0.15099771320819855,
      "learning_rate": 1.812800932315147e-05,
      "loss": 0.4553,
      "mean_token_accuracy": 0.8551701568067074,
      "num_tokens": 852055870.0,
      "step": 26697
    },
    {
      "entropy": 0.5311544728465378,
      "epoch": 2.4563516023484624,
      "grad_norm": 0.1595541536808014,
      "learning_rate": 1.8124942497009844e-05,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8350011892616749,
      "num_tokens": 852087789.0,
      "step": 26698
    },
    {
      "entropy": 0.5630695624276996,
      "epoch": 2.456443608719904,
      "grad_norm": 0.16634723544120789,
      "learning_rate": 1.812187567086822e-05,
      "loss": 0.5408,
      "mean_token_accuracy": 0.8290241360664368,
      "num_tokens": 852119483.0,
      "step": 26699
    },
    {
      "entropy": 0.48798003140836954,
      "epoch": 2.456535615091345,
      "grad_norm": 0.16048696637153625,
      "learning_rate": 1.8118808844726594e-05,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8491269499063492,
      "num_tokens": 852151676.0,
      "step": 26700
    },
    {
      "entropy": 0.5845205751247704,
      "epoch": 2.4566276214627862,
      "grad_norm": 0.17067702114582062,
      "learning_rate": 1.8115742018584965e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8196885138750076,
      "num_tokens": 852182534.0,
      "step": 26701
    },
    {
      "entropy": 0.5995588628575206,
      "epoch": 2.4567196278342274,
      "grad_norm": 0.1783478856086731,
      "learning_rate": 1.811267519244334e-05,
      "loss": 0.6009,
      "mean_token_accuracy": 0.8134289309382439,
      "num_tokens": 852213820.0,
      "step": 26702
    },
    {
      "entropy": 0.5246204454451799,
      "epoch": 2.4568116342056685,
      "grad_norm": 0.16415666043758392,
      "learning_rate": 1.8109608366301714e-05,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8343973383307457,
      "num_tokens": 852245964.0,
      "step": 26703
    },
    {
      "entropy": 0.6430352907627821,
      "epoch": 2.45690364057711,
      "grad_norm": 0.17001774907112122,
      "learning_rate": 1.810654154016009e-05,
      "loss": 0.6275,
      "mean_token_accuracy": 0.8045362830162048,
      "num_tokens": 852278298.0,
      "step": 26704
    },
    {
      "entropy": 0.5694993976503611,
      "epoch": 2.456995646948551,
      "grad_norm": 0.17255952954292297,
      "learning_rate": 1.810347471401846e-05,
      "loss": 0.5655,
      "mean_token_accuracy": 0.8233221918344498,
      "num_tokens": 852309822.0,
      "step": 26705
    },
    {
      "entropy": 0.43820713460445404,
      "epoch": 2.4570876533199923,
      "grad_norm": 0.14748454093933105,
      "learning_rate": 1.8100407887876838e-05,
      "loss": 0.429,
      "mean_token_accuracy": 0.8663835115730762,
      "num_tokens": 852342217.0,
      "step": 26706
    },
    {
      "entropy": 0.6123045235872269,
      "epoch": 2.4571796596914335,
      "grad_norm": 0.1682884693145752,
      "learning_rate": 1.8097341061735213e-05,
      "loss": 0.5992,
      "mean_token_accuracy": 0.8143976368010044,
      "num_tokens": 852373154.0,
      "step": 26707
    },
    {
      "entropy": 0.6145134642720222,
      "epoch": 2.4572716660628746,
      "grad_norm": 0.1663305163383484,
      "learning_rate": 1.8094274235593588e-05,
      "loss": 0.6022,
      "mean_token_accuracy": 0.8102940917015076,
      "num_tokens": 852405405.0,
      "step": 26708
    },
    {
      "entropy": 0.6022633938118815,
      "epoch": 2.457363672434316,
      "grad_norm": 0.17172478139400482,
      "learning_rate": 1.809120740945196e-05,
      "loss": 0.5897,
      "mean_token_accuracy": 0.8107917569577694,
      "num_tokens": 852437413.0,
      "step": 26709
    },
    {
      "entropy": 0.5525014596059918,
      "epoch": 2.4574556788057573,
      "grad_norm": 0.16728727519512177,
      "learning_rate": 1.8088140583310333e-05,
      "loss": 0.5355,
      "mean_token_accuracy": 0.8310388773679733,
      "num_tokens": 852469153.0,
      "step": 26710
    },
    {
      "entropy": 0.5729895541444421,
      "epoch": 2.4575476851771985,
      "grad_norm": 0.1634235382080078,
      "learning_rate": 1.8085073757168708e-05,
      "loss": 0.5656,
      "mean_token_accuracy": 0.8212672285735607,
      "num_tokens": 852501033.0,
      "step": 26711
    },
    {
      "entropy": 0.47050097468309104,
      "epoch": 2.4576396915486396,
      "grad_norm": 0.15247073769569397,
      "learning_rate": 1.8082006931027083e-05,
      "loss": 0.46,
      "mean_token_accuracy": 0.853355024009943,
      "num_tokens": 852533417.0,
      "step": 26712
    },
    {
      "entropy": 0.5216265879571438,
      "epoch": 2.4577316979200807,
      "grad_norm": 0.16547437012195587,
      "learning_rate": 1.8078940104885454e-05,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8352895565330982,
      "num_tokens": 852565966.0,
      "step": 26713
    },
    {
      "entropy": 0.4880653088912368,
      "epoch": 2.4578237042915223,
      "grad_norm": 0.1566626876592636,
      "learning_rate": 1.807587327874383e-05,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8434318602085114,
      "num_tokens": 852597390.0,
      "step": 26714
    },
    {
      "entropy": 0.5128579456359148,
      "epoch": 2.4579157106629634,
      "grad_norm": 0.16322767734527588,
      "learning_rate": 1.8072806452602203e-05,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8462017104029655,
      "num_tokens": 852628156.0,
      "step": 26715
    },
    {
      "entropy": 0.42936704820021987,
      "epoch": 2.4580077170344046,
      "grad_norm": 0.14796796441078186,
      "learning_rate": 1.8069739626460578e-05,
      "loss": 0.4164,
      "mean_token_accuracy": 0.8631087690591812,
      "num_tokens": 852660092.0,
      "step": 26716
    },
    {
      "entropy": 0.5159088959917426,
      "epoch": 2.4580997234058457,
      "grad_norm": 0.16678783297538757,
      "learning_rate": 1.8066672800318953e-05,
      "loss": 0.5125,
      "mean_token_accuracy": 0.8378639109432697,
      "num_tokens": 852692226.0,
      "step": 26717
    },
    {
      "entropy": 0.5114379739388824,
      "epoch": 2.458191729777287,
      "grad_norm": 0.16231435537338257,
      "learning_rate": 1.8063605974177324e-05,
      "loss": 0.5146,
      "mean_token_accuracy": 0.8386014848947525,
      "num_tokens": 852723864.0,
      "step": 26718
    },
    {
      "entropy": 0.4634872036986053,
      "epoch": 2.4582837361487284,
      "grad_norm": 0.14810463786125183,
      "learning_rate": 1.80605391480357e-05,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8565805926918983,
      "num_tokens": 852756152.0,
      "step": 26719
    },
    {
      "entropy": 0.4298457899130881,
      "epoch": 2.4583757425201695,
      "grad_norm": 0.14503352344036102,
      "learning_rate": 1.8057472321894073e-05,
      "loss": 0.4184,
      "mean_token_accuracy": 0.8666068241000175,
      "num_tokens": 852788142.0,
      "step": 26720
    },
    {
      "entropy": 0.4879705375060439,
      "epoch": 2.4584677488916107,
      "grad_norm": 0.15059098601341248,
      "learning_rate": 1.8054405495752448e-05,
      "loss": 0.476,
      "mean_token_accuracy": 0.8498167991638184,
      "num_tokens": 852820210.0,
      "step": 26721
    },
    {
      "entropy": 0.5096382941119373,
      "epoch": 2.458559755263052,
      "grad_norm": 0.15862701833248138,
      "learning_rate": 1.805133866961082e-05,
      "loss": 0.5042,
      "mean_token_accuracy": 0.8418942913413048,
      "num_tokens": 852852695.0,
      "step": 26722
    },
    {
      "entropy": 0.4454883683938533,
      "epoch": 2.458651761634493,
      "grad_norm": 0.15139663219451904,
      "learning_rate": 1.8048271843469194e-05,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8678354769945145,
      "num_tokens": 852883397.0,
      "step": 26723
    },
    {
      "entropy": 0.44497907534241676,
      "epoch": 2.4587437680059345,
      "grad_norm": 0.147821843624115,
      "learning_rate": 1.804520501732757e-05,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8664252609014511,
      "num_tokens": 852915929.0,
      "step": 26724
    },
    {
      "entropy": 0.5095361424610019,
      "epoch": 2.4588357743773757,
      "grad_norm": 0.15944302082061768,
      "learning_rate": 1.8042138191185943e-05,
      "loss": 0.4987,
      "mean_token_accuracy": 0.840814184397459,
      "num_tokens": 852947990.0,
      "step": 26725
    },
    {
      "entropy": 0.6025244239717722,
      "epoch": 2.458927780748817,
      "grad_norm": 0.16754144430160522,
      "learning_rate": 1.8039071365044318e-05,
      "loss": 0.5861,
      "mean_token_accuracy": 0.8159728683531284,
      "num_tokens": 852980213.0,
      "step": 26726
    },
    {
      "entropy": 0.5284077962860465,
      "epoch": 2.459019787120258,
      "grad_norm": 0.1559409350156784,
      "learning_rate": 1.803600453890269e-05,
      "loss": 0.5088,
      "mean_token_accuracy": 0.8398403599858284,
      "num_tokens": 853012446.0,
      "step": 26727
    },
    {
      "entropy": 0.5296865440905094,
      "epoch": 2.459111793491699,
      "grad_norm": 0.16249507665634155,
      "learning_rate": 1.8032937712761064e-05,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8315144069492817,
      "num_tokens": 853044269.0,
      "step": 26728
    },
    {
      "entropy": 0.48009470745455474,
      "epoch": 2.4592037998631406,
      "grad_norm": 0.1498861014842987,
      "learning_rate": 1.8029870886619438e-05,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8549606688320637,
      "num_tokens": 853075713.0,
      "step": 26729
    },
    {
      "entropy": 0.5870454981923103,
      "epoch": 2.4592958062345818,
      "grad_norm": 0.1674538105726242,
      "learning_rate": 1.8026804060477813e-05,
      "loss": 0.5722,
      "mean_token_accuracy": 0.8201033920049667,
      "num_tokens": 853107871.0,
      "step": 26730
    },
    {
      "entropy": 0.5668134419247508,
      "epoch": 2.459387812606023,
      "grad_norm": 0.16688548028469086,
      "learning_rate": 1.8023737234336184e-05,
      "loss": 0.5608,
      "mean_token_accuracy": 0.8239430338144302,
      "num_tokens": 853139988.0,
      "step": 26731
    },
    {
      "entropy": 0.5039502409053966,
      "epoch": 2.459479818977464,
      "grad_norm": 0.15526799857616425,
      "learning_rate": 1.802067040819456e-05,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8447184711694717,
      "num_tokens": 853171087.0,
      "step": 26732
    },
    {
      "entropy": 0.4422299196012318,
      "epoch": 2.459571825348905,
      "grad_norm": 0.14757326245307922,
      "learning_rate": 1.8017603582052933e-05,
      "loss": 0.4248,
      "mean_token_accuracy": 0.8628758192062378,
      "num_tokens": 853203342.0,
      "step": 26733
    },
    {
      "entropy": 0.4698693500831723,
      "epoch": 2.4596638317203467,
      "grad_norm": 0.15778930485248566,
      "learning_rate": 1.8014536755911308e-05,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8521520085632801,
      "num_tokens": 853235338.0,
      "step": 26734
    },
    {
      "entropy": 0.5559359593316913,
      "epoch": 2.459755838091788,
      "grad_norm": 0.16219988465309143,
      "learning_rate": 1.801146992976968e-05,
      "loss": 0.5292,
      "mean_token_accuracy": 0.8319410532712936,
      "num_tokens": 853266491.0,
      "step": 26735
    },
    {
      "entropy": 0.5654996447265148,
      "epoch": 2.459847844463229,
      "grad_norm": 0.17074260115623474,
      "learning_rate": 1.8008403103628054e-05,
      "loss": 0.5538,
      "mean_token_accuracy": 0.8229511417448521,
      "num_tokens": 853297917.0,
      "step": 26736
    },
    {
      "entropy": 0.5397288338281214,
      "epoch": 2.45993985083467,
      "grad_norm": 0.15816287696361542,
      "learning_rate": 1.8005336277486432e-05,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8303338401019573,
      "num_tokens": 853329914.0,
      "step": 26737
    },
    {
      "entropy": 0.5077033713459969,
      "epoch": 2.4600318572061113,
      "grad_norm": 0.16032427549362183,
      "learning_rate": 1.8002269451344807e-05,
      "loss": 0.4831,
      "mean_token_accuracy": 0.8457691632211208,
      "num_tokens": 853361944.0,
      "step": 26738
    },
    {
      "entropy": 0.5758383786305785,
      "epoch": 2.460123863577553,
      "grad_norm": 0.16921909153461456,
      "learning_rate": 1.7999202625203178e-05,
      "loss": 0.5614,
      "mean_token_accuracy": 0.8210838846862316,
      "num_tokens": 853393364.0,
      "step": 26739
    },
    {
      "entropy": 0.49434982892125845,
      "epoch": 2.460215869948994,
      "grad_norm": 0.15700605511665344,
      "learning_rate": 1.7996135799061553e-05,
      "loss": 0.4838,
      "mean_token_accuracy": 0.849987305700779,
      "num_tokens": 853425917.0,
      "step": 26740
    },
    {
      "entropy": 0.5575208757072687,
      "epoch": 2.460307876320435,
      "grad_norm": 0.16908203065395355,
      "learning_rate": 1.7993068972919927e-05,
      "loss": 0.5525,
      "mean_token_accuracy": 0.8248108439147472,
      "num_tokens": 853458070.0,
      "step": 26741
    },
    {
      "entropy": 0.42944913264364004,
      "epoch": 2.4603998826918763,
      "grad_norm": 0.15262503921985626,
      "learning_rate": 1.7990002146778302e-05,
      "loss": 0.4256,
      "mean_token_accuracy": 0.865144319832325,
      "num_tokens": 853489718.0,
      "step": 26742
    },
    {
      "entropy": 0.5572548545897007,
      "epoch": 2.4604918890633174,
      "grad_norm": 0.17130818963050842,
      "learning_rate": 1.7986935320636673e-05,
      "loss": 0.5491,
      "mean_token_accuracy": 0.8275665417313576,
      "num_tokens": 853521123.0,
      "step": 26743
    },
    {
      "entropy": 0.541086355689913,
      "epoch": 2.460583895434759,
      "grad_norm": 0.16585274040699005,
      "learning_rate": 1.7983868494495048e-05,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8331703953444958,
      "num_tokens": 853552789.0,
      "step": 26744
    },
    {
      "entropy": 0.4416524115949869,
      "epoch": 2.4606759018062,
      "grad_norm": 0.15015777945518494,
      "learning_rate": 1.7980801668353422e-05,
      "loss": 0.4382,
      "mean_token_accuracy": 0.8603246323764324,
      "num_tokens": 853585492.0,
      "step": 26745
    },
    {
      "entropy": 0.47589577082544565,
      "epoch": 2.4607679081776412,
      "grad_norm": 0.15468673408031464,
      "learning_rate": 1.7977734842211797e-05,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8507549874484539,
      "num_tokens": 853618003.0,
      "step": 26746
    },
    {
      "entropy": 0.5205119354650378,
      "epoch": 2.4608599145490824,
      "grad_norm": 0.1628328114748001,
      "learning_rate": 1.797466801607017e-05,
      "loss": 0.51,
      "mean_token_accuracy": 0.8406395614147186,
      "num_tokens": 853649672.0,
      "step": 26747
    },
    {
      "entropy": 0.5196724133566022,
      "epoch": 2.4609519209205235,
      "grad_norm": 0.15883246064186096,
      "learning_rate": 1.7971601189928543e-05,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8413500972092152,
      "num_tokens": 853681980.0,
      "step": 26748
    },
    {
      "entropy": 0.4974817130714655,
      "epoch": 2.461043927291965,
      "grad_norm": 0.16113708913326263,
      "learning_rate": 1.7968534363786918e-05,
      "loss": 0.4847,
      "mean_token_accuracy": 0.8438088856637478,
      "num_tokens": 853713022.0,
      "step": 26749
    },
    {
      "entropy": 0.4701250554062426,
      "epoch": 2.461135933663406,
      "grad_norm": 0.1585799604654312,
      "learning_rate": 1.7965467537645292e-05,
      "loss": 0.4521,
      "mean_token_accuracy": 0.8541697971522808,
      "num_tokens": 853744197.0,
      "step": 26750
    },
    {
      "entropy": 0.40619861287996173,
      "epoch": 2.4612279400348474,
      "grad_norm": 0.14809933304786682,
      "learning_rate": 1.7962400711503667e-05,
      "loss": 0.3988,
      "mean_token_accuracy": 0.8691612631082535,
      "num_tokens": 853776488.0,
      "step": 26751
    },
    {
      "entropy": 0.4671705085784197,
      "epoch": 2.4613199464062885,
      "grad_norm": 0.14678601920604706,
      "learning_rate": 1.7959333885362038e-05,
      "loss": 0.4497,
      "mean_token_accuracy": 0.8544729575514793,
      "num_tokens": 853808681.0,
      "step": 26752
    },
    {
      "entropy": 0.5273145223036408,
      "epoch": 2.4614119527777296,
      "grad_norm": 0.15582334995269775,
      "learning_rate": 1.7956267059220413e-05,
      "loss": 0.5125,
      "mean_token_accuracy": 0.8389767222106457,
      "num_tokens": 853840684.0,
      "step": 26753
    },
    {
      "entropy": 0.44650526950135827,
      "epoch": 2.461503959149171,
      "grad_norm": 0.15448634326457977,
      "learning_rate": 1.7953200233078787e-05,
      "loss": 0.4347,
      "mean_token_accuracy": 0.8547153547406197,
      "num_tokens": 853872477.0,
      "step": 26754
    },
    {
      "entropy": 0.49122015945613384,
      "epoch": 2.4615959655206123,
      "grad_norm": 0.15399539470672607,
      "learning_rate": 1.7950133406937162e-05,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8498180396854877,
      "num_tokens": 853904114.0,
      "step": 26755
    },
    {
      "entropy": 0.4990622114855796,
      "epoch": 2.4616879718920535,
      "grad_norm": 0.15320375561714172,
      "learning_rate": 1.7947066580795537e-05,
      "loss": 0.478,
      "mean_token_accuracy": 0.8468683958053589,
      "num_tokens": 853935746.0,
      "step": 26756
    },
    {
      "entropy": 0.5828517116606236,
      "epoch": 2.4617799782634946,
      "grad_norm": 0.17054755985736847,
      "learning_rate": 1.7943999754653908e-05,
      "loss": 0.5847,
      "mean_token_accuracy": 0.8172935917973518,
      "num_tokens": 853968514.0,
      "step": 26757
    },
    {
      "entropy": 0.5114091522991657,
      "epoch": 2.4618719846349357,
      "grad_norm": 0.1600644290447235,
      "learning_rate": 1.7940932928512283e-05,
      "loss": 0.4935,
      "mean_token_accuracy": 0.842549379914999,
      "num_tokens": 854000715.0,
      "step": 26758
    },
    {
      "entropy": 0.47866894002072513,
      "epoch": 2.4619639910063773,
      "grad_norm": 0.15746648609638214,
      "learning_rate": 1.7937866102370657e-05,
      "loss": 0.4648,
      "mean_token_accuracy": 0.8499566800892353,
      "num_tokens": 854032676.0,
      "step": 26759
    },
    {
      "entropy": 0.5750427972525358,
      "epoch": 2.4620559973778184,
      "grad_norm": 0.16626141965389252,
      "learning_rate": 1.7934799276229032e-05,
      "loss": 0.5664,
      "mean_token_accuracy": 0.8254901729524136,
      "num_tokens": 854065120.0,
      "step": 26760
    },
    {
      "entropy": 0.5391187081113458,
      "epoch": 2.4621480037492596,
      "grad_norm": 0.16305436193943024,
      "learning_rate": 1.7931732450087403e-05,
      "loss": 0.5333,
      "mean_token_accuracy": 0.8318696655333042,
      "num_tokens": 854096713.0,
      "step": 26761
    },
    {
      "entropy": 0.5070473055820912,
      "epoch": 2.4622400101207007,
      "grad_norm": 0.15650096535682678,
      "learning_rate": 1.7928665623945778e-05,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8449875079095364,
      "num_tokens": 854128276.0,
      "step": 26762
    },
    {
      "entropy": 0.6317997016012669,
      "epoch": 2.462332016492142,
      "grad_norm": 0.1696479320526123,
      "learning_rate": 1.7925598797804152e-05,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8063080683350563,
      "num_tokens": 854160606.0,
      "step": 26763
    },
    {
      "entropy": 0.5073451511561871,
      "epoch": 2.4624240228635834,
      "grad_norm": 0.16625508666038513,
      "learning_rate": 1.7922531971662527e-05,
      "loss": 0.4999,
      "mean_token_accuracy": 0.8463749028742313,
      "num_tokens": 854192865.0,
      "step": 26764
    },
    {
      "entropy": 0.5083900894969702,
      "epoch": 2.4625160292350246,
      "grad_norm": 0.16229355335235596,
      "learning_rate": 1.79194651455209e-05,
      "loss": 0.4879,
      "mean_token_accuracy": 0.8426254913210869,
      "num_tokens": 854224770.0,
      "step": 26765
    },
    {
      "entropy": 0.47985936515033245,
      "epoch": 2.4626080356064657,
      "grad_norm": 0.15406718850135803,
      "learning_rate": 1.7916398319379273e-05,
      "loss": 0.4668,
      "mean_token_accuracy": 0.8493351712822914,
      "num_tokens": 854257004.0,
      "step": 26766
    },
    {
      "entropy": 0.510482863523066,
      "epoch": 2.462700041977907,
      "grad_norm": 0.15946565568447113,
      "learning_rate": 1.791333149323765e-05,
      "loss": 0.4839,
      "mean_token_accuracy": 0.8439178541302681,
      "num_tokens": 854287521.0,
      "step": 26767
    },
    {
      "entropy": 0.4842119636014104,
      "epoch": 2.462792048349348,
      "grad_norm": 0.15466271340847015,
      "learning_rate": 1.7910264667096026e-05,
      "loss": 0.4769,
      "mean_token_accuracy": 0.8495886214077473,
      "num_tokens": 854319648.0,
      "step": 26768
    },
    {
      "entropy": 0.49842571280896664,
      "epoch": 2.4628840547207895,
      "grad_norm": 0.15572020411491394,
      "learning_rate": 1.7907197840954397e-05,
      "loss": 0.4856,
      "mean_token_accuracy": 0.846385546028614,
      "num_tokens": 854351747.0,
      "step": 26769
    },
    {
      "entropy": 0.47517778910696507,
      "epoch": 2.4629760610922307,
      "grad_norm": 0.15338382124900818,
      "learning_rate": 1.790413101481277e-05,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8530594184994698,
      "num_tokens": 854384502.0,
      "step": 26770
    },
    {
      "entropy": 0.6280003990978003,
      "epoch": 2.463068067463672,
      "grad_norm": 0.16850262880325317,
      "learning_rate": 1.7901064188671146e-05,
      "loss": 0.611,
      "mean_token_accuracy": 0.8076197728514671,
      "num_tokens": 854416710.0,
      "step": 26771
    },
    {
      "entropy": 0.5230775522068143,
      "epoch": 2.463160073835113,
      "grad_norm": 0.16641362011432648,
      "learning_rate": 1.789799736252952e-05,
      "loss": 0.5132,
      "mean_token_accuracy": 0.83647620677948,
      "num_tokens": 854448067.0,
      "step": 26772
    },
    {
      "entropy": 0.5027418369427323,
      "epoch": 2.463252080206554,
      "grad_norm": 0.1619746834039688,
      "learning_rate": 1.7894930536387892e-05,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8413928486406803,
      "num_tokens": 854480055.0,
      "step": 26773
    },
    {
      "entropy": 0.5062570357695222,
      "epoch": 2.4633440865779956,
      "grad_norm": 0.15692052245140076,
      "learning_rate": 1.7891863710246267e-05,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8425074107944965,
      "num_tokens": 854512474.0,
      "step": 26774
    },
    {
      "entropy": 0.5411133859306574,
      "epoch": 2.4634360929494368,
      "grad_norm": 0.16477133333683014,
      "learning_rate": 1.788879688410464e-05,
      "loss": 0.537,
      "mean_token_accuracy": 0.8276866786181927,
      "num_tokens": 854544264.0,
      "step": 26775
    },
    {
      "entropy": 0.6088929194957018,
      "epoch": 2.463528099320878,
      "grad_norm": 0.17120255529880524,
      "learning_rate": 1.7885730057963016e-05,
      "loss": 0.593,
      "mean_token_accuracy": 0.8119924701750278,
      "num_tokens": 854576180.0,
      "step": 26776
    },
    {
      "entropy": 0.5429066261276603,
      "epoch": 2.463620105692319,
      "grad_norm": 0.16181637346744537,
      "learning_rate": 1.788266323182139e-05,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8287947177886963,
      "num_tokens": 854608475.0,
      "step": 26777
    },
    {
      "entropy": 0.5277280155569315,
      "epoch": 2.46371211206376,
      "grad_norm": 0.16398608684539795,
      "learning_rate": 1.7879596405679762e-05,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8373169302940369,
      "num_tokens": 854640516.0,
      "step": 26778
    },
    {
      "entropy": 0.544672878459096,
      "epoch": 2.4638041184352018,
      "grad_norm": 0.1708213984966278,
      "learning_rate": 1.7876529579538137e-05,
      "loss": 0.5427,
      "mean_token_accuracy": 0.8303838558495045,
      "num_tokens": 854671912.0,
      "step": 26779
    },
    {
      "entropy": 0.4967766311019659,
      "epoch": 2.463896124806643,
      "grad_norm": 0.1548825353384018,
      "learning_rate": 1.787346275339651e-05,
      "loss": 0.4913,
      "mean_token_accuracy": 0.841691829264164,
      "num_tokens": 854703895.0,
      "step": 26780
    },
    {
      "entropy": 0.5725725200027227,
      "epoch": 2.463988131178084,
      "grad_norm": 0.1697545200586319,
      "learning_rate": 1.7870395927254886e-05,
      "loss": 0.5596,
      "mean_token_accuracy": 0.8214158862829208,
      "num_tokens": 854735731.0,
      "step": 26781
    },
    {
      "entropy": 0.5847976990044117,
      "epoch": 2.464080137549525,
      "grad_norm": 0.16844215989112854,
      "learning_rate": 1.7867329101113257e-05,
      "loss": 0.5767,
      "mean_token_accuracy": 0.8168876431882381,
      "num_tokens": 854767079.0,
      "step": 26782
    },
    {
      "entropy": 0.5666276002302766,
      "epoch": 2.4641721439209663,
      "grad_norm": 0.17095205187797546,
      "learning_rate": 1.7864262274971632e-05,
      "loss": 0.5626,
      "mean_token_accuracy": 0.8197958990931511,
      "num_tokens": 854799341.0,
      "step": 26783
    },
    {
      "entropy": 0.43518006429076195,
      "epoch": 2.464264150292408,
      "grad_norm": 0.14805176854133606,
      "learning_rate": 1.7861195448830007e-05,
      "loss": 0.4246,
      "mean_token_accuracy": 0.8650886788964272,
      "num_tokens": 854831402.0,
      "step": 26784
    },
    {
      "entropy": 0.4661131100729108,
      "epoch": 2.464356156663849,
      "grad_norm": 0.15333832800388336,
      "learning_rate": 1.785812862268838e-05,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8535229153931141,
      "num_tokens": 854863738.0,
      "step": 26785
    },
    {
      "entropy": 0.5533495144918561,
      "epoch": 2.46444816303529,
      "grad_norm": 0.16644369065761566,
      "learning_rate": 1.7855061796546756e-05,
      "loss": 0.5301,
      "mean_token_accuracy": 0.8330163285136223,
      "num_tokens": 854895318.0,
      "step": 26786
    },
    {
      "entropy": 0.5866133654490113,
      "epoch": 2.4645401694067313,
      "grad_norm": 0.1635887622833252,
      "learning_rate": 1.7851994970405127e-05,
      "loss": 0.5705,
      "mean_token_accuracy": 0.8220289498567581,
      "num_tokens": 854927917.0,
      "step": 26787
    },
    {
      "entropy": 0.547867439687252,
      "epoch": 2.4646321757781724,
      "grad_norm": 0.1587936133146286,
      "learning_rate": 1.7848928144263502e-05,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8324003517627716,
      "num_tokens": 854960395.0,
      "step": 26788
    },
    {
      "entropy": 0.5298930983990431,
      "epoch": 2.464724182149614,
      "grad_norm": 0.16291746497154236,
      "learning_rate": 1.7845861318121876e-05,
      "loss": 0.5208,
      "mean_token_accuracy": 0.8359316810965538,
      "num_tokens": 854991796.0,
      "step": 26789
    },
    {
      "entropy": 0.5475696511566639,
      "epoch": 2.464816188521055,
      "grad_norm": 0.1635655164718628,
      "learning_rate": 1.784279449198025e-05,
      "loss": 0.5418,
      "mean_token_accuracy": 0.8322382867336273,
      "num_tokens": 855024331.0,
      "step": 26790
    },
    {
      "entropy": 0.4737059115432203,
      "epoch": 2.4649081948924962,
      "grad_norm": 0.1570877730846405,
      "learning_rate": 1.7839727665838622e-05,
      "loss": 0.4568,
      "mean_token_accuracy": 0.8535628654062748,
      "num_tokens": 855055816.0,
      "step": 26791
    },
    {
      "entropy": 0.5730172470211983,
      "epoch": 2.4650002012639374,
      "grad_norm": 0.16424442827701569,
      "learning_rate": 1.7836660839696997e-05,
      "loss": 0.5617,
      "mean_token_accuracy": 0.8215620703995228,
      "num_tokens": 855088399.0,
      "step": 26792
    },
    {
      "entropy": 0.4840606590732932,
      "epoch": 2.4650922076353785,
      "grad_norm": 0.15537512302398682,
      "learning_rate": 1.783359401355537e-05,
      "loss": 0.479,
      "mean_token_accuracy": 0.8468677625060081,
      "num_tokens": 855121003.0,
      "step": 26793
    },
    {
      "entropy": 0.54824208188802,
      "epoch": 2.46518421400682,
      "grad_norm": 0.16380780935287476,
      "learning_rate": 1.7830527187413746e-05,
      "loss": 0.5469,
      "mean_token_accuracy": 0.830609604716301,
      "num_tokens": 855153585.0,
      "step": 26794
    },
    {
      "entropy": 0.45554118510335684,
      "epoch": 2.465276220378261,
      "grad_norm": 0.15032505989074707,
      "learning_rate": 1.7827460361272117e-05,
      "loss": 0.4438,
      "mean_token_accuracy": 0.8568514063954353,
      "num_tokens": 855186044.0,
      "step": 26795
    },
    {
      "entropy": 0.5311324959620833,
      "epoch": 2.4653682267497024,
      "grad_norm": 0.1627340465784073,
      "learning_rate": 1.7824393535130492e-05,
      "loss": 0.5214,
      "mean_token_accuracy": 0.8329229801893234,
      "num_tokens": 855217767.0,
      "step": 26796
    },
    {
      "entropy": 0.6566183101385832,
      "epoch": 2.4654602331211435,
      "grad_norm": 0.17668990790843964,
      "learning_rate": 1.7821326708988867e-05,
      "loss": 0.6514,
      "mean_token_accuracy": 0.7951206639409065,
      "num_tokens": 855249523.0,
      "step": 26797
    },
    {
      "entropy": 0.504728083498776,
      "epoch": 2.4655522394925846,
      "grad_norm": 0.16358482837677002,
      "learning_rate": 1.7818259882847245e-05,
      "loss": 0.4881,
      "mean_token_accuracy": 0.8418394960463047,
      "num_tokens": 855281505.0,
      "step": 26798
    },
    {
      "entropy": 0.5268937461078167,
      "epoch": 2.465644245864026,
      "grad_norm": 0.16519950330257416,
      "learning_rate": 1.7815193056705616e-05,
      "loss": 0.515,
      "mean_token_accuracy": 0.8362377323210239,
      "num_tokens": 855313326.0,
      "step": 26799
    },
    {
      "entropy": 0.643142681568861,
      "epoch": 2.4657362522354673,
      "grad_norm": 0.1701156497001648,
      "learning_rate": 1.781212623056399e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.8033558875322342,
      "num_tokens": 855345168.0,
      "step": 26800
    },
    {
      "entropy": 0.5145516963675618,
      "epoch": 2.4658282586069085,
      "grad_norm": 0.1631576120853424,
      "learning_rate": 1.7809059404422365e-05,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8403853438794613,
      "num_tokens": 855376327.0,
      "step": 26801
    },
    {
      "entropy": 0.46397147607058287,
      "epoch": 2.4659202649783496,
      "grad_norm": 0.15421009063720703,
      "learning_rate": 1.780599257828074e-05,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8528709448873997,
      "num_tokens": 855408130.0,
      "step": 26802
    },
    {
      "entropy": 0.5533314580097795,
      "epoch": 2.4660122713497907,
      "grad_norm": 0.16541776061058044,
      "learning_rate": 1.780292575213911e-05,
      "loss": 0.5449,
      "mean_token_accuracy": 0.8289039097726345,
      "num_tokens": 855440129.0,
      "step": 26803
    },
    {
      "entropy": 0.5536783412098885,
      "epoch": 2.4661042777212323,
      "grad_norm": 0.17046047747135162,
      "learning_rate": 1.7799858925997486e-05,
      "loss": 0.541,
      "mean_token_accuracy": 0.8250796012580395,
      "num_tokens": 855471553.0,
      "step": 26804
    },
    {
      "entropy": 0.615309719927609,
      "epoch": 2.4661962840926734,
      "grad_norm": 0.17470000684261322,
      "learning_rate": 1.779679209985586e-05,
      "loss": 0.6123,
      "mean_token_accuracy": 0.8098565638065338,
      "num_tokens": 855502540.0,
      "step": 26805
    },
    {
      "entropy": 0.4959853107575327,
      "epoch": 2.4662882904641146,
      "grad_norm": 0.14972001314163208,
      "learning_rate": 1.7793725273714235e-05,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8507314547896385,
      "num_tokens": 855535089.0,
      "step": 26806
    },
    {
      "entropy": 0.5359093081206083,
      "epoch": 2.4663802968355557,
      "grad_norm": 0.1625102311372757,
      "learning_rate": 1.779065844757261e-05,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8353929854929447,
      "num_tokens": 855566933.0,
      "step": 26807
    },
    {
      "entropy": 0.5724209994077682,
      "epoch": 2.466472303206997,
      "grad_norm": 0.16647639870643616,
      "learning_rate": 1.778759162143098e-05,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8217633590102196,
      "num_tokens": 855598378.0,
      "step": 26808
    },
    {
      "entropy": 0.5279050897806883,
      "epoch": 2.4665643095784384,
      "grad_norm": 0.16215939819812775,
      "learning_rate": 1.7784524795289356e-05,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8395560085773468,
      "num_tokens": 855630001.0,
      "step": 26809
    },
    {
      "entropy": 0.4911066599888727,
      "epoch": 2.4666563159498796,
      "grad_norm": 0.15347404778003693,
      "learning_rate": 1.778145796914773e-05,
      "loss": 0.4725,
      "mean_token_accuracy": 0.8510889373719692,
      "num_tokens": 855661800.0,
      "step": 26810
    },
    {
      "entropy": 0.5129545191302896,
      "epoch": 2.4667483223213207,
      "grad_norm": 0.15787404775619507,
      "learning_rate": 1.7778391143006105e-05,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8364833071827888,
      "num_tokens": 855693752.0,
      "step": 26811
    },
    {
      "entropy": 0.585635619238019,
      "epoch": 2.466840328692762,
      "grad_norm": 0.16540351510047913,
      "learning_rate": 1.7775324316864476e-05,
      "loss": 0.5772,
      "mean_token_accuracy": 0.8187939487397671,
      "num_tokens": 855726146.0,
      "step": 26812
    },
    {
      "entropy": 0.4776221443898976,
      "epoch": 2.466932335064203,
      "grad_norm": 0.15724334120750427,
      "learning_rate": 1.777225749072285e-05,
      "loss": 0.4792,
      "mean_token_accuracy": 0.8500911593437195,
      "num_tokens": 855758071.0,
      "step": 26813
    },
    {
      "entropy": 0.49374441197142005,
      "epoch": 2.4670243414356445,
      "grad_norm": 0.15939593315124512,
      "learning_rate": 1.7769190664581226e-05,
      "loss": 0.4736,
      "mean_token_accuracy": 0.8482744619250298,
      "num_tokens": 855789330.0,
      "step": 26814
    },
    {
      "entropy": 0.5852363333106041,
      "epoch": 2.4671163478070857,
      "grad_norm": 0.17163397371768951,
      "learning_rate": 1.77661238384396e-05,
      "loss": 0.5796,
      "mean_token_accuracy": 0.8146022483706474,
      "num_tokens": 855820889.0,
      "step": 26815
    },
    {
      "entropy": 0.38054176000878215,
      "epoch": 2.467208354178527,
      "grad_norm": 0.14373750984668732,
      "learning_rate": 1.7763057012297975e-05,
      "loss": 0.3725,
      "mean_token_accuracy": 0.8822458535432816,
      "num_tokens": 855852980.0,
      "step": 26816
    },
    {
      "entropy": 0.4795304290018976,
      "epoch": 2.467300360549968,
      "grad_norm": 0.15225303173065186,
      "learning_rate": 1.7759990186156346e-05,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8526782505214214,
      "num_tokens": 855885299.0,
      "step": 26817
    },
    {
      "entropy": 0.5317700689192861,
      "epoch": 2.467392366921409,
      "grad_norm": 0.15999403595924377,
      "learning_rate": 1.775692336001472e-05,
      "loss": 0.5195,
      "mean_token_accuracy": 0.8345112912356853,
      "num_tokens": 855916728.0,
      "step": 26818
    },
    {
      "entropy": 0.5544528216123581,
      "epoch": 2.4674843732928506,
      "grad_norm": 0.16487745940685272,
      "learning_rate": 1.7753856533873095e-05,
      "loss": 0.5488,
      "mean_token_accuracy": 0.8272390030324459,
      "num_tokens": 855947948.0,
      "step": 26819
    },
    {
      "entropy": 0.5277422210201621,
      "epoch": 2.4675763796642918,
      "grad_norm": 0.16429874300956726,
      "learning_rate": 1.775078970773147e-05,
      "loss": 0.5186,
      "mean_token_accuracy": 0.836632139980793,
      "num_tokens": 855979719.0,
      "step": 26820
    },
    {
      "entropy": 0.48772989632561803,
      "epoch": 2.467668386035733,
      "grad_norm": 0.15765351057052612,
      "learning_rate": 1.774772288158984e-05,
      "loss": 0.4783,
      "mean_token_accuracy": 0.8474642634391785,
      "num_tokens": 856011819.0,
      "step": 26821
    },
    {
      "entropy": 0.5292754014953971,
      "epoch": 2.467760392407174,
      "grad_norm": 0.1641819328069687,
      "learning_rate": 1.7744656055448216e-05,
      "loss": 0.5148,
      "mean_token_accuracy": 0.8372265249490738,
      "num_tokens": 856043942.0,
      "step": 26822
    },
    {
      "entropy": 0.47202615067362785,
      "epoch": 2.467852398778615,
      "grad_norm": 0.15231028199195862,
      "learning_rate": 1.774158922930659e-05,
      "loss": 0.4548,
      "mean_token_accuracy": 0.85714515671134,
      "num_tokens": 856075925.0,
      "step": 26823
    },
    {
      "entropy": 0.5043285945430398,
      "epoch": 2.4679444051500568,
      "grad_norm": 0.16083568334579468,
      "learning_rate": 1.7738522403164965e-05,
      "loss": 0.4959,
      "mean_token_accuracy": 0.8408990725874901,
      "num_tokens": 856107212.0,
      "step": 26824
    },
    {
      "entropy": 0.5750737767666578,
      "epoch": 2.468036411521498,
      "grad_norm": 0.16571129858493805,
      "learning_rate": 1.773545557702334e-05,
      "loss": 0.565,
      "mean_token_accuracy": 0.8227071389555931,
      "num_tokens": 856139087.0,
      "step": 26825
    },
    {
      "entropy": 0.5122892577201128,
      "epoch": 2.468128417892939,
      "grad_norm": 0.15795673429965973,
      "learning_rate": 1.773238875088171e-05,
      "loss": 0.5008,
      "mean_token_accuracy": 0.8403066918253899,
      "num_tokens": 856171409.0,
      "step": 26826
    },
    {
      "entropy": 0.5542438719421625,
      "epoch": 2.46822042426438,
      "grad_norm": 0.16375865042209625,
      "learning_rate": 1.7729321924740086e-05,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8271817266941071,
      "num_tokens": 856203950.0,
      "step": 26827
    },
    {
      "entropy": 0.5793125480413437,
      "epoch": 2.4683124306358213,
      "grad_norm": 0.16501984000205994,
      "learning_rate": 1.7726255098598464e-05,
      "loss": 0.5616,
      "mean_token_accuracy": 0.8237934224307537,
      "num_tokens": 856235978.0,
      "step": 26828
    },
    {
      "entropy": 0.5205903127789497,
      "epoch": 2.468404437007263,
      "grad_norm": 0.1639954149723053,
      "learning_rate": 1.7723188272456835e-05,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8447838164865971,
      "num_tokens": 856267504.0,
      "step": 26829
    },
    {
      "entropy": 0.504973940551281,
      "epoch": 2.468496443378704,
      "grad_norm": 0.16268205642700195,
      "learning_rate": 1.772012144631521e-05,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8424904569983482,
      "num_tokens": 856299819.0,
      "step": 26830
    },
    {
      "entropy": 0.5243184873834252,
      "epoch": 2.468588449750145,
      "grad_norm": 0.1647782325744629,
      "learning_rate": 1.7717054620173584e-05,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8386669717729092,
      "num_tokens": 856331984.0,
      "step": 26831
    },
    {
      "entropy": 0.5613729069009423,
      "epoch": 2.4686804561215863,
      "grad_norm": 0.1722932904958725,
      "learning_rate": 1.771398779403196e-05,
      "loss": 0.5447,
      "mean_token_accuracy": 0.827180240303278,
      "num_tokens": 856363746.0,
      "step": 26832
    },
    {
      "entropy": 0.4844993338920176,
      "epoch": 2.4687724624930274,
      "grad_norm": 0.15362945199012756,
      "learning_rate": 1.7710920967890334e-05,
      "loss": 0.4676,
      "mean_token_accuracy": 0.8497374504804611,
      "num_tokens": 856395763.0,
      "step": 26833
    },
    {
      "entropy": 0.5378889236599207,
      "epoch": 2.468864468864469,
      "grad_norm": 0.1664133220911026,
      "learning_rate": 1.7707854141748705e-05,
      "loss": 0.5235,
      "mean_token_accuracy": 0.8322470225393772,
      "num_tokens": 856427860.0,
      "step": 26834
    },
    {
      "entropy": 0.5713976714760065,
      "epoch": 2.46895647523591,
      "grad_norm": 0.16656608879566193,
      "learning_rate": 1.770478731560708e-05,
      "loss": 0.559,
      "mean_token_accuracy": 0.8263789862394333,
      "num_tokens": 856460241.0,
      "step": 26835
    },
    {
      "entropy": 0.40388107392936945,
      "epoch": 2.4690484816073512,
      "grad_norm": 0.13700446486473083,
      "learning_rate": 1.7701720489465454e-05,
      "loss": 0.3864,
      "mean_token_accuracy": 0.8778172470629215,
      "num_tokens": 856492085.0,
      "step": 26836
    },
    {
      "entropy": 0.4982284919824451,
      "epoch": 2.4691404879787924,
      "grad_norm": 0.16251473128795624,
      "learning_rate": 1.769865366332383e-05,
      "loss": 0.4909,
      "mean_token_accuracy": 0.8412791155278683,
      "num_tokens": 856522849.0,
      "step": 26837
    },
    {
      "entropy": 0.378075132612139,
      "epoch": 2.4692324943502335,
      "grad_norm": 0.1376480758190155,
      "learning_rate": 1.76955868371822e-05,
      "loss": 0.3573,
      "mean_token_accuracy": 0.882509671151638,
      "num_tokens": 856554764.0,
      "step": 26838
    },
    {
      "entropy": 0.5312943514436483,
      "epoch": 2.469324500721675,
      "grad_norm": 0.16029033064842224,
      "learning_rate": 1.7692520011040575e-05,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8349984176456928,
      "num_tokens": 856586911.0,
      "step": 26839
    },
    {
      "entropy": 0.4998412299901247,
      "epoch": 2.469416507093116,
      "grad_norm": 0.1555105596780777,
      "learning_rate": 1.768945318489895e-05,
      "loss": 0.4885,
      "mean_token_accuracy": 0.8433176688849926,
      "num_tokens": 856618569.0,
      "step": 26840
    },
    {
      "entropy": 0.4931703981710598,
      "epoch": 2.4695085134645574,
      "grad_norm": 0.1579950600862503,
      "learning_rate": 1.7686386358757324e-05,
      "loss": 0.4822,
      "mean_token_accuracy": 0.844871811568737,
      "num_tokens": 856650605.0,
      "step": 26841
    },
    {
      "entropy": 0.5284779686480761,
      "epoch": 2.4696005198359985,
      "grad_norm": 0.1639154553413391,
      "learning_rate": 1.7683319532615695e-05,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8344172164797783,
      "num_tokens": 856681845.0,
      "step": 26842
    },
    {
      "entropy": 0.36085614119656384,
      "epoch": 2.4696925262074396,
      "grad_norm": 0.13846448063850403,
      "learning_rate": 1.768025270647407e-05,
      "loss": 0.347,
      "mean_token_accuracy": 0.8877190798521042,
      "num_tokens": 856714239.0,
      "step": 26843
    },
    {
      "entropy": 0.49424982629716396,
      "epoch": 2.469784532578881,
      "grad_norm": 0.16110427677631378,
      "learning_rate": 1.7677185880332445e-05,
      "loss": 0.491,
      "mean_token_accuracy": 0.8399295583367348,
      "num_tokens": 856745898.0,
      "step": 26844
    },
    {
      "entropy": 0.49334901105612516,
      "epoch": 2.4698765389503223,
      "grad_norm": 0.15201137959957123,
      "learning_rate": 1.767411905419082e-05,
      "loss": 0.4805,
      "mean_token_accuracy": 0.8473118953406811,
      "num_tokens": 856778352.0,
      "step": 26845
    },
    {
      "entropy": 0.5110820913687348,
      "epoch": 2.4699685453217635,
      "grad_norm": 0.15845634043216705,
      "learning_rate": 1.7671052228049194e-05,
      "loss": 0.5061,
      "mean_token_accuracy": 0.8370517082512379,
      "num_tokens": 856810445.0,
      "step": 26846
    },
    {
      "entropy": 0.433079665992409,
      "epoch": 2.4700605516932046,
      "grad_norm": 0.1469019055366516,
      "learning_rate": 1.7667985401907565e-05,
      "loss": 0.4179,
      "mean_token_accuracy": 0.8657020814716816,
      "num_tokens": 856842177.0,
      "step": 26847
    },
    {
      "entropy": 0.43781242007389665,
      "epoch": 2.4701525580646457,
      "grad_norm": 0.14736665785312653,
      "learning_rate": 1.766491857576594e-05,
      "loss": 0.4245,
      "mean_token_accuracy": 0.8655923902988434,
      "num_tokens": 856874778.0,
      "step": 26848
    },
    {
      "entropy": 0.5167842591181397,
      "epoch": 2.4702445644360873,
      "grad_norm": 0.15385569632053375,
      "learning_rate": 1.7661851749624315e-05,
      "loss": 0.5063,
      "mean_token_accuracy": 0.8389474712312222,
      "num_tokens": 856907065.0,
      "step": 26849
    },
    {
      "entropy": 0.5185169437900186,
      "epoch": 2.4703365708075284,
      "grad_norm": 0.15908579528331757,
      "learning_rate": 1.765878492348269e-05,
      "loss": 0.5119,
      "mean_token_accuracy": 0.8371010199189186,
      "num_tokens": 856939380.0,
      "step": 26850
    },
    {
      "entropy": 0.5756661877967417,
      "epoch": 2.4704285771789696,
      "grad_norm": 0.16491323709487915,
      "learning_rate": 1.765571809734106e-05,
      "loss": 0.56,
      "mean_token_accuracy": 0.8220053240656853,
      "num_tokens": 856971196.0,
      "step": 26851
    },
    {
      "entropy": 0.5825381129980087,
      "epoch": 2.4705205835504107,
      "grad_norm": 0.16729949414730072,
      "learning_rate": 1.7652651271199435e-05,
      "loss": 0.5639,
      "mean_token_accuracy": 0.8206972442567348,
      "num_tokens": 857002171.0,
      "step": 26852
    },
    {
      "entropy": 0.5776198515668511,
      "epoch": 2.470612589921852,
      "grad_norm": 0.1713983118534088,
      "learning_rate": 1.764958444505781e-05,
      "loss": 0.5711,
      "mean_token_accuracy": 0.8190323896706104,
      "num_tokens": 857034382.0,
      "step": 26853
    },
    {
      "entropy": 0.5582004999741912,
      "epoch": 2.4707045962932934,
      "grad_norm": 0.16391275823116302,
      "learning_rate": 1.7646517618916184e-05,
      "loss": 0.5576,
      "mean_token_accuracy": 0.8244739510118961,
      "num_tokens": 857066822.0,
      "step": 26854
    },
    {
      "entropy": 0.54944814927876,
      "epoch": 2.4707966026647346,
      "grad_norm": 0.1637251228094101,
      "learning_rate": 1.764345079277456e-05,
      "loss": 0.5346,
      "mean_token_accuracy": 0.8291910588741302,
      "num_tokens": 857099161.0,
      "step": 26855
    },
    {
      "entropy": 0.5392610523849726,
      "epoch": 2.4708886090361757,
      "grad_norm": 0.1610684096813202,
      "learning_rate": 1.764038396663293e-05,
      "loss": 0.5315,
      "mean_token_accuracy": 0.8312616050243378,
      "num_tokens": 857130839.0,
      "step": 26856
    },
    {
      "entropy": 0.40309116896241903,
      "epoch": 2.470980615407617,
      "grad_norm": 0.14797434210777283,
      "learning_rate": 1.7637317140491305e-05,
      "loss": 0.3892,
      "mean_token_accuracy": 0.8739227056503296,
      "num_tokens": 857162805.0,
      "step": 26857
    },
    {
      "entropy": 0.617240134626627,
      "epoch": 2.471072621779058,
      "grad_norm": 0.16941772401332855,
      "learning_rate": 1.763425031434968e-05,
      "loss": 0.6089,
      "mean_token_accuracy": 0.8095221631228924,
      "num_tokens": 857195256.0,
      "step": 26858
    },
    {
      "entropy": 0.5587872546166182,
      "epoch": 2.4711646281504995,
      "grad_norm": 0.1703302413225174,
      "learning_rate": 1.7631183488208054e-05,
      "loss": 0.548,
      "mean_token_accuracy": 0.8235754035413265,
      "num_tokens": 857227423.0,
      "step": 26859
    },
    {
      "entropy": 0.6001679757609963,
      "epoch": 2.4712566345219407,
      "grad_norm": 0.16930729150772095,
      "learning_rate": 1.762811666206643e-05,
      "loss": 0.5838,
      "mean_token_accuracy": 0.8155282139778137,
      "num_tokens": 857259834.0,
      "step": 26860
    },
    {
      "entropy": 0.464911344810389,
      "epoch": 2.471348640893382,
      "grad_norm": 0.1493925005197525,
      "learning_rate": 1.7625049835924804e-05,
      "loss": 0.451,
      "mean_token_accuracy": 0.8586611747741699,
      "num_tokens": 857292228.0,
      "step": 26861
    },
    {
      "entropy": 0.4495681303087622,
      "epoch": 2.471440647264823,
      "grad_norm": 0.15550467371940613,
      "learning_rate": 1.7621983009783178e-05,
      "loss": 0.4407,
      "mean_token_accuracy": 0.8592879213392735,
      "num_tokens": 857323735.0,
      "step": 26862
    },
    {
      "entropy": 0.5394686833024025,
      "epoch": 2.471532653636264,
      "grad_norm": 0.16567540168762207,
      "learning_rate": 1.7618916183641553e-05,
      "loss": 0.5432,
      "mean_token_accuracy": 0.8341378830373287,
      "num_tokens": 857355799.0,
      "step": 26863
    },
    {
      "entropy": 0.5799036622047424,
      "epoch": 2.4716246600077056,
      "grad_norm": 0.1687679886817932,
      "learning_rate": 1.7615849357499924e-05,
      "loss": 0.5668,
      "mean_token_accuracy": 0.8211125954985619,
      "num_tokens": 857387844.0,
      "step": 26864
    },
    {
      "entropy": 0.5607181619852781,
      "epoch": 2.4717166663791468,
      "grad_norm": 0.16650834679603577,
      "learning_rate": 1.76127825313583e-05,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8290707804262638,
      "num_tokens": 857419317.0,
      "step": 26865
    },
    {
      "entropy": 0.5486329551786184,
      "epoch": 2.471808672750588,
      "grad_norm": 0.16197045147418976,
      "learning_rate": 1.7609715705216673e-05,
      "loss": 0.5276,
      "mean_token_accuracy": 0.8324611335992813,
      "num_tokens": 857451498.0,
      "step": 26866
    },
    {
      "entropy": 0.6059907842427492,
      "epoch": 2.471900679122029,
      "grad_norm": 0.17655432224273682,
      "learning_rate": 1.7606648879075048e-05,
      "loss": 0.5976,
      "mean_token_accuracy": 0.813653226941824,
      "num_tokens": 857482792.0,
      "step": 26867
    },
    {
      "entropy": 0.5520404493436217,
      "epoch": 2.47199268549347,
      "grad_norm": 0.16970349848270416,
      "learning_rate": 1.760358205293342e-05,
      "loss": 0.5424,
      "mean_token_accuracy": 0.8284584507346153,
      "num_tokens": 857514076.0,
      "step": 26868
    },
    {
      "entropy": 0.5436621569097042,
      "epoch": 2.4720846918649118,
      "grad_norm": 0.1599084585905075,
      "learning_rate": 1.7600515226791794e-05,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8282278552651405,
      "num_tokens": 857546208.0,
      "step": 26869
    },
    {
      "entropy": 0.5761981420218945,
      "epoch": 2.472176698236353,
      "grad_norm": 0.16658565402030945,
      "learning_rate": 1.759744840065017e-05,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8237046599388123,
      "num_tokens": 857578299.0,
      "step": 26870
    },
    {
      "entropy": 0.5674995081499219,
      "epoch": 2.472268704607794,
      "grad_norm": 0.16436941921710968,
      "learning_rate": 1.7594381574508543e-05,
      "loss": 0.5586,
      "mean_token_accuracy": 0.8225355967879295,
      "num_tokens": 857610687.0,
      "step": 26871
    },
    {
      "entropy": 0.5219568219035864,
      "epoch": 2.472360710979235,
      "grad_norm": 0.1590382605791092,
      "learning_rate": 1.7591314748366915e-05,
      "loss": 0.517,
      "mean_token_accuracy": 0.8366209305822849,
      "num_tokens": 857642927.0,
      "step": 26872
    },
    {
      "entropy": 0.5504006333649158,
      "epoch": 2.4724527173506763,
      "grad_norm": 0.1660768687725067,
      "learning_rate": 1.758824792222529e-05,
      "loss": 0.5416,
      "mean_token_accuracy": 0.8294902220368385,
      "num_tokens": 857675164.0,
      "step": 26873
    },
    {
      "entropy": 0.5456795445643365,
      "epoch": 2.472544723722118,
      "grad_norm": 0.16013352572917938,
      "learning_rate": 1.7585181096083664e-05,
      "loss": 0.5331,
      "mean_token_accuracy": 0.8357034511864185,
      "num_tokens": 857707708.0,
      "step": 26874
    },
    {
      "entropy": 0.4097641110420227,
      "epoch": 2.472636730093559,
      "grad_norm": 0.14573509991168976,
      "learning_rate": 1.758211426994204e-05,
      "loss": 0.404,
      "mean_token_accuracy": 0.8708170503377914,
      "num_tokens": 857740476.0,
      "step": 26875
    },
    {
      "entropy": 0.5197221552953124,
      "epoch": 2.472728736465,
      "grad_norm": 0.1615038514137268,
      "learning_rate": 1.7579047443800413e-05,
      "loss": 0.5053,
      "mean_token_accuracy": 0.8427731096744537,
      "num_tokens": 857771729.0,
      "step": 26876
    },
    {
      "entropy": 0.5779394768178463,
      "epoch": 2.4728207428364413,
      "grad_norm": 0.16897493600845337,
      "learning_rate": 1.7575980617658784e-05,
      "loss": 0.5744,
      "mean_token_accuracy": 0.8222194314002991,
      "num_tokens": 857803111.0,
      "step": 26877
    },
    {
      "entropy": 0.5326984254643321,
      "epoch": 2.4729127492078824,
      "grad_norm": 0.16437916457653046,
      "learning_rate": 1.757291379151716e-05,
      "loss": 0.5197,
      "mean_token_accuracy": 0.8346062377095222,
      "num_tokens": 857834323.0,
      "step": 26878
    },
    {
      "entropy": 0.48159098718315363,
      "epoch": 2.473004755579324,
      "grad_norm": 0.15267778933048248,
      "learning_rate": 1.7569846965375534e-05,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8545452244579792,
      "num_tokens": 857866127.0,
      "step": 26879
    },
    {
      "entropy": 0.5861514067510143,
      "epoch": 2.473096761950765,
      "grad_norm": 0.17217664420604706,
      "learning_rate": 1.7566780139233908e-05,
      "loss": 0.5839,
      "mean_token_accuracy": 0.8178610503673553,
      "num_tokens": 857897513.0,
      "step": 26880
    },
    {
      "entropy": 0.568843842484057,
      "epoch": 2.4731887683222062,
      "grad_norm": 0.1694268435239792,
      "learning_rate": 1.756371331309228e-05,
      "loss": 0.5599,
      "mean_token_accuracy": 0.824450146406889,
      "num_tokens": 857928809.0,
      "step": 26881
    },
    {
      "entropy": 0.5587967485189438,
      "epoch": 2.4732807746936474,
      "grad_norm": 0.16616378724575043,
      "learning_rate": 1.7560646486950654e-05,
      "loss": 0.547,
      "mean_token_accuracy": 0.8285246826708317,
      "num_tokens": 857960764.0,
      "step": 26882
    },
    {
      "entropy": 0.48086500354111195,
      "epoch": 2.4733727810650885,
      "grad_norm": 0.15146487951278687,
      "learning_rate": 1.755757966080903e-05,
      "loss": 0.47,
      "mean_token_accuracy": 0.8501364029943943,
      "num_tokens": 857993171.0,
      "step": 26883
    },
    {
      "entropy": 0.47325235418975353,
      "epoch": 2.47346478743653,
      "grad_norm": 0.15795767307281494,
      "learning_rate": 1.7554512834667404e-05,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8550068326294422,
      "num_tokens": 858025136.0,
      "step": 26884
    },
    {
      "entropy": 0.3901876574382186,
      "epoch": 2.473556793807971,
      "grad_norm": 0.14583104848861694,
      "learning_rate": 1.7551446008525778e-05,
      "loss": 0.3855,
      "mean_token_accuracy": 0.8733292035758495,
      "num_tokens": 858057357.0,
      "step": 26885
    },
    {
      "entropy": 0.5254912730306387,
      "epoch": 2.4736488001794124,
      "grad_norm": 0.16282010078430176,
      "learning_rate": 1.754837918238415e-05,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8372178412973881,
      "num_tokens": 858089257.0,
      "step": 26886
    },
    {
      "entropy": 0.5618502739816904,
      "epoch": 2.4737408065508535,
      "grad_norm": 0.16059257090091705,
      "learning_rate": 1.7545312356242524e-05,
      "loss": 0.5393,
      "mean_token_accuracy": 0.8301367498934269,
      "num_tokens": 858121298.0,
      "step": 26887
    },
    {
      "entropy": 0.560957585927099,
      "epoch": 2.4738328129222946,
      "grad_norm": 0.17034606635570526,
      "learning_rate": 1.75422455301009e-05,
      "loss": 0.5533,
      "mean_token_accuracy": 0.8218100629746914,
      "num_tokens": 858153348.0,
      "step": 26888
    },
    {
      "entropy": 0.49748319014906883,
      "epoch": 2.473924819293736,
      "grad_norm": 0.1501895934343338,
      "learning_rate": 1.7539178703959273e-05,
      "loss": 0.4878,
      "mean_token_accuracy": 0.846405602991581,
      "num_tokens": 858185443.0,
      "step": 26889
    },
    {
      "entropy": 0.528194825630635,
      "epoch": 2.4740168256651773,
      "grad_norm": 0.16769009828567505,
      "learning_rate": 1.7536111877817648e-05,
      "loss": 0.5216,
      "mean_token_accuracy": 0.8341872468590736,
      "num_tokens": 858217706.0,
      "step": 26890
    },
    {
      "entropy": 0.4874874157831073,
      "epoch": 2.4741088320366185,
      "grad_norm": 0.15502001345157623,
      "learning_rate": 1.7533045051676023e-05,
      "loss": 0.468,
      "mean_token_accuracy": 0.8524917960166931,
      "num_tokens": 858249588.0,
      "step": 26891
    },
    {
      "entropy": 0.42766309226863086,
      "epoch": 2.4742008384080596,
      "grad_norm": 0.15220196545124054,
      "learning_rate": 1.7529978225534397e-05,
      "loss": 0.4076,
      "mean_token_accuracy": 0.8681859821081161,
      "num_tokens": 858280935.0,
      "step": 26892
    },
    {
      "entropy": 0.5191182196140289,
      "epoch": 2.4742928447795007,
      "grad_norm": 0.15727479755878448,
      "learning_rate": 1.7526911399392772e-05,
      "loss": 0.5006,
      "mean_token_accuracy": 0.8431180715560913,
      "num_tokens": 858313154.0,
      "step": 26893
    },
    {
      "entropy": 0.41885285964235663,
      "epoch": 2.4743848511509423,
      "grad_norm": 0.14525897800922394,
      "learning_rate": 1.7523844573251143e-05,
      "loss": 0.4025,
      "mean_token_accuracy": 0.8709408454596996,
      "num_tokens": 858345555.0,
      "step": 26894
    },
    {
      "entropy": 0.5175558412447572,
      "epoch": 2.4744768575223834,
      "grad_norm": 0.15602128207683563,
      "learning_rate": 1.7520777747109518e-05,
      "loss": 0.4978,
      "mean_token_accuracy": 0.8396695517003536,
      "num_tokens": 858377592.0,
      "step": 26895
    },
    {
      "entropy": 0.5226951043587178,
      "epoch": 2.4745688638938246,
      "grad_norm": 0.16184349358081818,
      "learning_rate": 1.7517710920967893e-05,
      "loss": 0.5097,
      "mean_token_accuracy": 0.8397239856421947,
      "num_tokens": 858408677.0,
      "step": 26896
    },
    {
      "entropy": 0.5727191986516118,
      "epoch": 2.4746608702652657,
      "grad_norm": 0.16859398782253265,
      "learning_rate": 1.7514644094826267e-05,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8238265663385391,
      "num_tokens": 858440325.0,
      "step": 26897
    },
    {
      "entropy": 0.5509033761918545,
      "epoch": 2.474752876636707,
      "grad_norm": 0.16900534927845,
      "learning_rate": 1.751157726868464e-05,
      "loss": 0.531,
      "mean_token_accuracy": 0.8347379229962826,
      "num_tokens": 858471177.0,
      "step": 26898
    },
    {
      "entropy": 0.6518062688410282,
      "epoch": 2.4748448830081484,
      "grad_norm": 0.17649728059768677,
      "learning_rate": 1.7508510442543013e-05,
      "loss": 0.6468,
      "mean_token_accuracy": 0.7947249934077263,
      "num_tokens": 858502520.0,
      "step": 26899
    },
    {
      "entropy": 0.42798755515832454,
      "epoch": 2.4749368893795896,
      "grad_norm": 0.15218780934810638,
      "learning_rate": 1.7505443616401388e-05,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8613183610141277,
      "num_tokens": 858534307.0,
      "step": 26900
    },
    {
      "entropy": 0.5374296670779586,
      "epoch": 2.4750288957510307,
      "grad_norm": 0.17063893377780914,
      "learning_rate": 1.7502376790259762e-05,
      "loss": 0.5291,
      "mean_token_accuracy": 0.8339495174586773,
      "num_tokens": 858565758.0,
      "step": 26901
    },
    {
      "entropy": 0.49855064041912556,
      "epoch": 2.475120902122472,
      "grad_norm": 0.15981481969356537,
      "learning_rate": 1.7499309964118134e-05,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8440459854900837,
      "num_tokens": 858598037.0,
      "step": 26902
    },
    {
      "entropy": 0.4491792665794492,
      "epoch": 2.475212908493913,
      "grad_norm": 0.14915569126605988,
      "learning_rate": 1.7496243137976508e-05,
      "loss": 0.4314,
      "mean_token_accuracy": 0.8598180413246155,
      "num_tokens": 858630023.0,
      "step": 26903
    },
    {
      "entropy": 0.5649998364970088,
      "epoch": 2.4753049148653545,
      "grad_norm": 0.16467978060245514,
      "learning_rate": 1.7493176311834883e-05,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8245684988796711,
      "num_tokens": 858661845.0,
      "step": 26904
    },
    {
      "entropy": 0.4296649405732751,
      "epoch": 2.4753969212367957,
      "grad_norm": 0.1443108469247818,
      "learning_rate": 1.7490109485693258e-05,
      "loss": 0.411,
      "mean_token_accuracy": 0.8693005330860615,
      "num_tokens": 858693665.0,
      "step": 26905
    },
    {
      "entropy": 0.47125783283263445,
      "epoch": 2.475488927608237,
      "grad_norm": 0.14712795615196228,
      "learning_rate": 1.7487042659551632e-05,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8540919423103333,
      "num_tokens": 858725869.0,
      "step": 26906
    },
    {
      "entropy": 0.5417893612757325,
      "epoch": 2.475580933979678,
      "grad_norm": 0.16314905881881714,
      "learning_rate": 1.7483975833410003e-05,
      "loss": 0.5375,
      "mean_token_accuracy": 0.8304567635059357,
      "num_tokens": 858756820.0,
      "step": 26907
    },
    {
      "entropy": 0.5669939089566469,
      "epoch": 2.475672940351119,
      "grad_norm": 0.16951246559619904,
      "learning_rate": 1.7480909007268378e-05,
      "loss": 0.5577,
      "mean_token_accuracy": 0.8226274251937866,
      "num_tokens": 858787933.0,
      "step": 26908
    },
    {
      "entropy": 0.5582329891622066,
      "epoch": 2.4757649467225606,
      "grad_norm": 0.16367681324481964,
      "learning_rate": 1.7477842181126753e-05,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8238886930048466,
      "num_tokens": 858820015.0,
      "step": 26909
    },
    {
      "entropy": 0.5295727988705039,
      "epoch": 2.4758569530940018,
      "grad_norm": 0.16150425374507904,
      "learning_rate": 1.7474775354985127e-05,
      "loss": 0.5109,
      "mean_token_accuracy": 0.8359940648078918,
      "num_tokens": 858851521.0,
      "step": 26910
    },
    {
      "entropy": 0.5421970430761576,
      "epoch": 2.475948959465443,
      "grad_norm": 0.16410060226917267,
      "learning_rate": 1.74717085288435e-05,
      "loss": 0.5231,
      "mean_token_accuracy": 0.8334031291306019,
      "num_tokens": 858883271.0,
      "step": 26911
    },
    {
      "entropy": 0.48472131229937077,
      "epoch": 2.476040965836884,
      "grad_norm": 0.15406233072280884,
      "learning_rate": 1.7468641702701873e-05,
      "loss": 0.4636,
      "mean_token_accuracy": 0.852374579757452,
      "num_tokens": 858915524.0,
      "step": 26912
    },
    {
      "entropy": 0.53501280490309,
      "epoch": 2.476132972208325,
      "grad_norm": 0.16956265270709991,
      "learning_rate": 1.7465574876560248e-05,
      "loss": 0.5341,
      "mean_token_accuracy": 0.8336951434612274,
      "num_tokens": 858947014.0,
      "step": 26913
    },
    {
      "entropy": 0.4566282966407016,
      "epoch": 2.4762249785797668,
      "grad_norm": 0.15425030887126923,
      "learning_rate": 1.7462508050418623e-05,
      "loss": 0.4634,
      "mean_token_accuracy": 0.8529374450445175,
      "num_tokens": 858979143.0,
      "step": 26914
    },
    {
      "entropy": 0.6143694585189223,
      "epoch": 2.476316984951208,
      "grad_norm": 0.17202623188495636,
      "learning_rate": 1.7459441224276997e-05,
      "loss": 0.6143,
      "mean_token_accuracy": 0.8074575997889042,
      "num_tokens": 859011696.0,
      "step": 26915
    },
    {
      "entropy": 0.4735669884830713,
      "epoch": 2.476408991322649,
      "grad_norm": 0.15966814756393433,
      "learning_rate": 1.745637439813537e-05,
      "loss": 0.4622,
      "mean_token_accuracy": 0.8532616645097733,
      "num_tokens": 859043771.0,
      "step": 26916
    },
    {
      "entropy": 0.5719269178807735,
      "epoch": 2.47650099769409,
      "grad_norm": 0.1689932644367218,
      "learning_rate": 1.7453307571993743e-05,
      "loss": 0.5599,
      "mean_token_accuracy": 0.8231121078133583,
      "num_tokens": 859075822.0,
      "step": 26917
    },
    {
      "entropy": 0.52149080298841,
      "epoch": 2.4765930040655313,
      "grad_norm": 0.16087189316749573,
      "learning_rate": 1.7450240745852118e-05,
      "loss": 0.5137,
      "mean_token_accuracy": 0.8385207019746304,
      "num_tokens": 859108111.0,
      "step": 26918
    },
    {
      "entropy": 0.49253156408667564,
      "epoch": 2.476685010436973,
      "grad_norm": 0.155195951461792,
      "learning_rate": 1.7447173919710492e-05,
      "loss": 0.4712,
      "mean_token_accuracy": 0.8494621440768242,
      "num_tokens": 859139236.0,
      "step": 26919
    },
    {
      "entropy": 0.4212112558307126,
      "epoch": 2.476777016808414,
      "grad_norm": 0.14378121495246887,
      "learning_rate": 1.7444107093568867e-05,
      "loss": 0.4071,
      "mean_token_accuracy": 0.8686141334474087,
      "num_tokens": 859171357.0,
      "step": 26920
    },
    {
      "entropy": 0.48249124246649444,
      "epoch": 2.476869023179855,
      "grad_norm": 0.15084858238697052,
      "learning_rate": 1.7441040267427242e-05,
      "loss": 0.4679,
      "mean_token_accuracy": 0.8500818610191345,
      "num_tokens": 859203059.0,
      "step": 26921
    },
    {
      "entropy": 0.548901482950896,
      "epoch": 2.4769610295512963,
      "grad_norm": 0.1609858125448227,
      "learning_rate": 1.7437973441285616e-05,
      "loss": 0.5275,
      "mean_token_accuracy": 0.8306460902094841,
      "num_tokens": 859235253.0,
      "step": 26922
    },
    {
      "entropy": 0.42490996117703617,
      "epoch": 2.4770530359227374,
      "grad_norm": 0.14630474150180817,
      "learning_rate": 1.743490661514399e-05,
      "loss": 0.4034,
      "mean_token_accuracy": 0.8716843388974667,
      "num_tokens": 859267316.0,
      "step": 26923
    },
    {
      "entropy": 0.5814698841422796,
      "epoch": 2.477145042294179,
      "grad_norm": 0.16827258467674255,
      "learning_rate": 1.7431839789002362e-05,
      "loss": 0.5666,
      "mean_token_accuracy": 0.8206615969538689,
      "num_tokens": 859299018.0,
      "step": 26924
    },
    {
      "entropy": 0.571176050696522,
      "epoch": 2.47723704866562,
      "grad_norm": 0.16400475800037384,
      "learning_rate": 1.7428772962860737e-05,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8226020038127899,
      "num_tokens": 859331128.0,
      "step": 26925
    },
    {
      "entropy": 0.47766655031591654,
      "epoch": 2.4773290550370612,
      "grad_norm": 0.15814684331417084,
      "learning_rate": 1.742570613671911e-05,
      "loss": 0.4619,
      "mean_token_accuracy": 0.8506523184478283,
      "num_tokens": 859363215.0,
      "step": 26926
    },
    {
      "entropy": 0.5787805637810379,
      "epoch": 2.4774210614085024,
      "grad_norm": 0.17185969650745392,
      "learning_rate": 1.7422639310577486e-05,
      "loss": 0.5562,
      "mean_token_accuracy": 0.8235506750643253,
      "num_tokens": 859394007.0,
      "step": 26927
    },
    {
      "entropy": 0.5253278631716967,
      "epoch": 2.4775130677799435,
      "grad_norm": 0.1607929766178131,
      "learning_rate": 1.7419572484435858e-05,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8375841751694679,
      "num_tokens": 859425503.0,
      "step": 26928
    },
    {
      "entropy": 0.5332245398312807,
      "epoch": 2.477605074151385,
      "grad_norm": 0.16239577531814575,
      "learning_rate": 1.7416505658294232e-05,
      "loss": 0.5256,
      "mean_token_accuracy": 0.8326383084058762,
      "num_tokens": 859458206.0,
      "step": 26929
    },
    {
      "entropy": 0.5594249069690704,
      "epoch": 2.4776970805228262,
      "grad_norm": 0.16563861072063446,
      "learning_rate": 1.7413438832152607e-05,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8272388353943825,
      "num_tokens": 859490319.0,
      "step": 26930
    },
    {
      "entropy": 0.39747579500544816,
      "epoch": 2.4777890868942674,
      "grad_norm": 0.14387792348861694,
      "learning_rate": 1.741037200601098e-05,
      "loss": 0.3969,
      "mean_token_accuracy": 0.8750868737697601,
      "num_tokens": 859522055.0,
      "step": 26931
    },
    {
      "entropy": 0.4974163235165179,
      "epoch": 2.4778810932657085,
      "grad_norm": 0.1592860370874405,
      "learning_rate": 1.7407305179869353e-05,
      "loss": 0.4835,
      "mean_token_accuracy": 0.8474000208079815,
      "num_tokens": 859553645.0,
      "step": 26932
    },
    {
      "entropy": 0.5033198448363692,
      "epoch": 2.4779730996371496,
      "grad_norm": 0.1589294970035553,
      "learning_rate": 1.7404238353727727e-05,
      "loss": 0.4904,
      "mean_token_accuracy": 0.8427360877394676,
      "num_tokens": 859585480.0,
      "step": 26933
    },
    {
      "entropy": 0.4552988560171798,
      "epoch": 2.478065106008591,
      "grad_norm": 0.14758813381195068,
      "learning_rate": 1.7401171527586102e-05,
      "loss": 0.4443,
      "mean_token_accuracy": 0.8580373674631119,
      "num_tokens": 859617485.0,
      "step": 26934
    },
    {
      "entropy": 0.439623243175447,
      "epoch": 2.4781571123800323,
      "grad_norm": 0.1517437845468521,
      "learning_rate": 1.7398104701444477e-05,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8617992326617241,
      "num_tokens": 859649824.0,
      "step": 26935
    },
    {
      "entropy": 0.46485308883711696,
      "epoch": 2.4782491187514735,
      "grad_norm": 0.15629561245441437,
      "learning_rate": 1.739503787530285e-05,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8516733795404434,
      "num_tokens": 859681484.0,
      "step": 26936
    },
    {
      "entropy": 0.4735565008595586,
      "epoch": 2.4783411251229146,
      "grad_norm": 0.1575748175382614,
      "learning_rate": 1.7391971049161223e-05,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8465184308588505,
      "num_tokens": 859713834.0,
      "step": 26937
    },
    {
      "entropy": 0.49542104080319405,
      "epoch": 2.4784331314943557,
      "grad_norm": 0.16034021973609924,
      "learning_rate": 1.7388904223019597e-05,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8460045456886292,
      "num_tokens": 859745614.0,
      "step": 26938
    },
    {
      "entropy": 0.5810815510340035,
      "epoch": 2.4785251378657973,
      "grad_norm": 0.16318392753601074,
      "learning_rate": 1.7385837396877972e-05,
      "loss": 0.5689,
      "mean_token_accuracy": 0.820832148194313,
      "num_tokens": 859777841.0,
      "step": 26939
    },
    {
      "entropy": 0.5193031877279282,
      "epoch": 2.4786171442372384,
      "grad_norm": 0.15935878455638885,
      "learning_rate": 1.7382770570736347e-05,
      "loss": 0.5117,
      "mean_token_accuracy": 0.8386777155101299,
      "num_tokens": 859809936.0,
      "step": 26940
    },
    {
      "entropy": 0.5050076702609658,
      "epoch": 2.4787091506086796,
      "grad_norm": 0.15987487137317657,
      "learning_rate": 1.7379703744594718e-05,
      "loss": 0.4961,
      "mean_token_accuracy": 0.8357764072716236,
      "num_tokens": 859841338.0,
      "step": 26941
    },
    {
      "entropy": 0.5507608335465193,
      "epoch": 2.4788011569801207,
      "grad_norm": 0.16349279880523682,
      "learning_rate": 1.7376636918453092e-05,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8309077471494675,
      "num_tokens": 859873146.0,
      "step": 26942
    },
    {
      "entropy": 0.47251275926828384,
      "epoch": 2.4788931633515623,
      "grad_norm": 0.14741279184818268,
      "learning_rate": 1.7373570092311467e-05,
      "loss": 0.4635,
      "mean_token_accuracy": 0.8521113768219948,
      "num_tokens": 859905229.0,
      "step": 26943
    },
    {
      "entropy": 0.5479840952903032,
      "epoch": 2.4789851697230034,
      "grad_norm": 0.16725380718708038,
      "learning_rate": 1.7370503266169842e-05,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8271041065454483,
      "num_tokens": 859937286.0,
      "step": 26944
    },
    {
      "entropy": 0.5543686989694834,
      "epoch": 2.4790771760944446,
      "grad_norm": 0.1641460508108139,
      "learning_rate": 1.7367436440028216e-05,
      "loss": 0.5396,
      "mean_token_accuracy": 0.8302730359137058,
      "num_tokens": 859968875.0,
      "step": 26945
    },
    {
      "entropy": 0.5524492124095559,
      "epoch": 2.4791691824658857,
      "grad_norm": 0.16643942892551422,
      "learning_rate": 1.7364369613886588e-05,
      "loss": 0.5389,
      "mean_token_accuracy": 0.8275033533573151,
      "num_tokens": 860001089.0,
      "step": 26946
    },
    {
      "entropy": 0.5381143372505903,
      "epoch": 2.479261188837327,
      "grad_norm": 0.1652940958738327,
      "learning_rate": 1.7361302787744962e-05,
      "loss": 0.5353,
      "mean_token_accuracy": 0.8339348249137402,
      "num_tokens": 860032773.0,
      "step": 26947
    },
    {
      "entropy": 0.4232986280694604,
      "epoch": 2.4793531952087684,
      "grad_norm": 0.14630965888500214,
      "learning_rate": 1.7358235961603337e-05,
      "loss": 0.4047,
      "mean_token_accuracy": 0.8701019696891308,
      "num_tokens": 860065445.0,
      "step": 26948
    },
    {
      "entropy": 0.4639157960191369,
      "epoch": 2.4794452015802095,
      "grad_norm": 0.1486344337463379,
      "learning_rate": 1.735516913546171e-05,
      "loss": 0.4515,
      "mean_token_accuracy": 0.8548243977129459,
      "num_tokens": 860098101.0,
      "step": 26949
    },
    {
      "entropy": 0.4810981876216829,
      "epoch": 2.4795372079516507,
      "grad_norm": 0.15256111323833466,
      "learning_rate": 1.7352102309320086e-05,
      "loss": 0.4641,
      "mean_token_accuracy": 0.855274610221386,
      "num_tokens": 860129781.0,
      "step": 26950
    },
    {
      "entropy": 0.4935913719236851,
      "epoch": 2.479629214323092,
      "grad_norm": 0.15695206820964813,
      "learning_rate": 1.734903548317846e-05,
      "loss": 0.485,
      "mean_token_accuracy": 0.8445572033524513,
      "num_tokens": 860162342.0,
      "step": 26951
    },
    {
      "entropy": 0.5717246271669865,
      "epoch": 2.479721220694533,
      "grad_norm": 0.1603814661502838,
      "learning_rate": 1.7345968657036835e-05,
      "loss": 0.553,
      "mean_token_accuracy": 0.8248050399124622,
      "num_tokens": 860194605.0,
      "step": 26952
    },
    {
      "entropy": 0.4775671996176243,
      "epoch": 2.4798132270659745,
      "grad_norm": 0.15853220224380493,
      "learning_rate": 1.734290183089521e-05,
      "loss": 0.4647,
      "mean_token_accuracy": 0.8476717062294483,
      "num_tokens": 860226658.0,
      "step": 26953
    },
    {
      "entropy": 0.4029160104691982,
      "epoch": 2.4799052334374156,
      "grad_norm": 0.14485609531402588,
      "learning_rate": 1.733983500475358e-05,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8702585510909557,
      "num_tokens": 860258524.0,
      "step": 26954
    },
    {
      "entropy": 0.6188083402812481,
      "epoch": 2.479997239808857,
      "grad_norm": 0.17128701508045197,
      "learning_rate": 1.7336768178611956e-05,
      "loss": 0.6055,
      "mean_token_accuracy": 0.8094347007572651,
      "num_tokens": 860290402.0,
      "step": 26955
    },
    {
      "entropy": 0.535370359197259,
      "epoch": 2.480089246180298,
      "grad_norm": 0.16176313161849976,
      "learning_rate": 1.733370135247033e-05,
      "loss": 0.5283,
      "mean_token_accuracy": 0.832008358091116,
      "num_tokens": 860321844.0,
      "step": 26956
    },
    {
      "entropy": 0.5229994086548686,
      "epoch": 2.480181252551739,
      "grad_norm": 0.1544422209262848,
      "learning_rate": 1.7330634526328705e-05,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8401156365871429,
      "num_tokens": 860353830.0,
      "step": 26957
    },
    {
      "entropy": 0.5449258533772081,
      "epoch": 2.4802732589231806,
      "grad_norm": 0.15684296190738678,
      "learning_rate": 1.7327567700187077e-05,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8333967663347721,
      "num_tokens": 860386085.0,
      "step": 26958
    },
    {
      "entropy": 0.5302501674741507,
      "epoch": 2.4803652652946218,
      "grad_norm": 0.16288265585899353,
      "learning_rate": 1.732450087404545e-05,
      "loss": 0.518,
      "mean_token_accuracy": 0.836209163069725,
      "num_tokens": 860418504.0,
      "step": 26959
    },
    {
      "entropy": 0.41889889538288116,
      "epoch": 2.480457271666063,
      "grad_norm": 0.1515583097934723,
      "learning_rate": 1.7321434047903826e-05,
      "loss": 0.3994,
      "mean_token_accuracy": 0.8674222193658352,
      "num_tokens": 860449732.0,
      "step": 26960
    },
    {
      "entropy": 0.5706712789833546,
      "epoch": 2.480549278037504,
      "grad_norm": 0.1701822280883789,
      "learning_rate": 1.73183672217622e-05,
      "loss": 0.5493,
      "mean_token_accuracy": 0.8266528137028217,
      "num_tokens": 860481249.0,
      "step": 26961
    },
    {
      "entropy": 0.4655493386089802,
      "epoch": 2.480641284408945,
      "grad_norm": 0.15649761259555817,
      "learning_rate": 1.7315300395620572e-05,
      "loss": 0.456,
      "mean_token_accuracy": 0.8525662049651146,
      "num_tokens": 860512939.0,
      "step": 26962
    },
    {
      "entropy": 0.5455471444875002,
      "epoch": 2.4807332907803867,
      "grad_norm": 0.16575659811496735,
      "learning_rate": 1.7312233569478946e-05,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8301913514733315,
      "num_tokens": 860544330.0,
      "step": 26963
    },
    {
      "entropy": 0.6250993013381958,
      "epoch": 2.480825297151828,
      "grad_norm": 0.17238470911979675,
      "learning_rate": 1.730916674333732e-05,
      "loss": 0.6224,
      "mean_token_accuracy": 0.8080897144973278,
      "num_tokens": 860575786.0,
      "step": 26964
    },
    {
      "entropy": 0.5031236596405506,
      "epoch": 2.480917303523269,
      "grad_norm": 0.15727289021015167,
      "learning_rate": 1.7306099917195696e-05,
      "loss": 0.4881,
      "mean_token_accuracy": 0.8425105698406696,
      "num_tokens": 860608374.0,
      "step": 26965
    },
    {
      "entropy": 0.5339007419534028,
      "epoch": 2.48100930989471,
      "grad_norm": 0.16280002892017365,
      "learning_rate": 1.730303309105407e-05,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8325917460024357,
      "num_tokens": 860640028.0,
      "step": 26966
    },
    {
      "entropy": 0.6077867140993476,
      "epoch": 2.4811013162661513,
      "grad_norm": 0.17506177723407745,
      "learning_rate": 1.729996626491244e-05,
      "loss": 0.5939,
      "mean_token_accuracy": 0.8089856691658497,
      "num_tokens": 860671161.0,
      "step": 26967
    },
    {
      "entropy": 0.49621977284550667,
      "epoch": 2.481193322637593,
      "grad_norm": 0.15869340300559998,
      "learning_rate": 1.7296899438770816e-05,
      "loss": 0.4776,
      "mean_token_accuracy": 0.8466026745736599,
      "num_tokens": 860702880.0,
      "step": 26968
    },
    {
      "entropy": 0.49089630506932735,
      "epoch": 2.481285329009034,
      "grad_norm": 0.1546841263771057,
      "learning_rate": 1.729383261262919e-05,
      "loss": 0.4789,
      "mean_token_accuracy": 0.8503386564552784,
      "num_tokens": 860734180.0,
      "step": 26969
    },
    {
      "entropy": 0.5542250201106071,
      "epoch": 2.481377335380475,
      "grad_norm": 0.16443796455860138,
      "learning_rate": 1.7290765786487566e-05,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8242576941847801,
      "num_tokens": 860766603.0,
      "step": 26970
    },
    {
      "entropy": 0.5040615806356072,
      "epoch": 2.4814693417519162,
      "grad_norm": 0.15747833251953125,
      "learning_rate": 1.7287698960345937e-05,
      "loss": 0.4977,
      "mean_token_accuracy": 0.8409611880779266,
      "num_tokens": 860798824.0,
      "step": 26971
    },
    {
      "entropy": 0.5029306774958968,
      "epoch": 2.4815613481233574,
      "grad_norm": 0.15502245724201202,
      "learning_rate": 1.728463213420431e-05,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8444384671747684,
      "num_tokens": 860830415.0,
      "step": 26972
    },
    {
      "entropy": 0.5095021799206734,
      "epoch": 2.481653354494799,
      "grad_norm": 0.15646550059318542,
      "learning_rate": 1.7281565308062686e-05,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8436222672462463,
      "num_tokens": 860861818.0,
      "step": 26973
    },
    {
      "entropy": 0.5105061733629555,
      "epoch": 2.48174536086624,
      "grad_norm": 0.15952664613723755,
      "learning_rate": 1.727849848192106e-05,
      "loss": 0.4963,
      "mean_token_accuracy": 0.8429428897798061,
      "num_tokens": 860892865.0,
      "step": 26974
    },
    {
      "entropy": 0.5476577170193195,
      "epoch": 2.4818373672376812,
      "grad_norm": 0.16642603278160095,
      "learning_rate": 1.7275431655779435e-05,
      "loss": 0.5315,
      "mean_token_accuracy": 0.8315054439008236,
      "num_tokens": 860924412.0,
      "step": 26975
    },
    {
      "entropy": 0.4414223888888955,
      "epoch": 2.4819293736091224,
      "grad_norm": 0.1515299528837204,
      "learning_rate": 1.7272364829637807e-05,
      "loss": 0.4295,
      "mean_token_accuracy": 0.8606608994305134,
      "num_tokens": 860956441.0,
      "step": 26976
    },
    {
      "entropy": 0.5463256519287825,
      "epoch": 2.4820213799805635,
      "grad_norm": 0.1627928912639618,
      "learning_rate": 1.726929800349618e-05,
      "loss": 0.5344,
      "mean_token_accuracy": 0.8313192203640938,
      "num_tokens": 860988565.0,
      "step": 26977
    },
    {
      "entropy": 0.4818577696569264,
      "epoch": 2.482113386352005,
      "grad_norm": 0.15512391924858093,
      "learning_rate": 1.7266231177354556e-05,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8526559323072433,
      "num_tokens": 861019628.0,
      "step": 26978
    },
    {
      "entropy": 0.5658346163108945,
      "epoch": 2.482205392723446,
      "grad_norm": 0.16812972724437714,
      "learning_rate": 1.726316435121293e-05,
      "loss": 0.5483,
      "mean_token_accuracy": 0.8255239315330982,
      "num_tokens": 861051988.0,
      "step": 26979
    },
    {
      "entropy": 0.5890794731676579,
      "epoch": 2.4822973990948873,
      "grad_norm": 0.16808244585990906,
      "learning_rate": 1.7260097525071302e-05,
      "loss": 0.5767,
      "mean_token_accuracy": 0.8197214193642139,
      "num_tokens": 861083899.0,
      "step": 26980
    },
    {
      "entropy": 0.5229503409937024,
      "epoch": 2.4823894054663285,
      "grad_norm": 0.15749293565750122,
      "learning_rate": 1.725703069892968e-05,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8353167064487934,
      "num_tokens": 861116424.0,
      "step": 26981
    },
    {
      "entropy": 0.43585914466530085,
      "epoch": 2.4824814118377696,
      "grad_norm": 0.14775380492210388,
      "learning_rate": 1.7253963872788055e-05,
      "loss": 0.4278,
      "mean_token_accuracy": 0.8621797040104866,
      "num_tokens": 861147443.0,
      "step": 26982
    },
    {
      "entropy": 0.5564222801476717,
      "epoch": 2.482573418209211,
      "grad_norm": 0.1639942079782486,
      "learning_rate": 1.725089704664643e-05,
      "loss": 0.5423,
      "mean_token_accuracy": 0.8302844502031803,
      "num_tokens": 861179324.0,
      "step": 26983
    },
    {
      "entropy": 0.5518215019255877,
      "epoch": 2.4826654245806523,
      "grad_norm": 0.16775870323181152,
      "learning_rate": 1.72478302205048e-05,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8272621408104897,
      "num_tokens": 861210668.0,
      "step": 26984
    },
    {
      "entropy": 0.5067162488121539,
      "epoch": 2.4827574309520934,
      "grad_norm": 0.1546330600976944,
      "learning_rate": 1.7244763394363175e-05,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8432163074612617,
      "num_tokens": 861242830.0,
      "step": 26985
    },
    {
      "entropy": 0.4554701577872038,
      "epoch": 2.4828494373235346,
      "grad_norm": 0.15165317058563232,
      "learning_rate": 1.724169656822155e-05,
      "loss": 0.4425,
      "mean_token_accuracy": 0.8576950579881668,
      "num_tokens": 861274505.0,
      "step": 26986
    },
    {
      "entropy": 0.5715316031128168,
      "epoch": 2.4829414436949757,
      "grad_norm": 0.1682204306125641,
      "learning_rate": 1.7238629742079924e-05,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8198893554508686,
      "num_tokens": 861306781.0,
      "step": 26987
    },
    {
      "entropy": 0.4721391573548317,
      "epoch": 2.4830334500664173,
      "grad_norm": 0.15285834670066833,
      "learning_rate": 1.7235562915938296e-05,
      "loss": 0.4541,
      "mean_token_accuracy": 0.8527084924280643,
      "num_tokens": 861337879.0,
      "step": 26988
    },
    {
      "entropy": 0.4749219361692667,
      "epoch": 2.4831254564378584,
      "grad_norm": 0.15464140474796295,
      "learning_rate": 1.723249608979667e-05,
      "loss": 0.4621,
      "mean_token_accuracy": 0.8535813391208649,
      "num_tokens": 861369966.0,
      "step": 26989
    },
    {
      "entropy": 0.5904120150953531,
      "epoch": 2.4832174628092996,
      "grad_norm": 0.16876843571662903,
      "learning_rate": 1.7229429263655045e-05,
      "loss": 0.5871,
      "mean_token_accuracy": 0.8175771571695805,
      "num_tokens": 861401569.0,
      "step": 26990
    },
    {
      "entropy": 0.4414580473676324,
      "epoch": 2.4833094691807407,
      "grad_norm": 0.15360529720783234,
      "learning_rate": 1.722636243751342e-05,
      "loss": 0.4342,
      "mean_token_accuracy": 0.8583322465419769,
      "num_tokens": 861433142.0,
      "step": 26991
    },
    {
      "entropy": 0.4842554032802582,
      "epoch": 2.483401475552182,
      "grad_norm": 0.1523999571800232,
      "learning_rate": 1.722329561137179e-05,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8498064763844013,
      "num_tokens": 861465279.0,
      "step": 26992
    },
    {
      "entropy": 0.5499645015224814,
      "epoch": 2.4834934819236234,
      "grad_norm": 0.1609460860490799,
      "learning_rate": 1.7220228785230166e-05,
      "loss": 0.5203,
      "mean_token_accuracy": 0.8362118154764175,
      "num_tokens": 861496216.0,
      "step": 26993
    },
    {
      "entropy": 0.4922052752226591,
      "epoch": 2.4835854882950645,
      "grad_norm": 0.16383549571037292,
      "learning_rate": 1.721716195908854e-05,
      "loss": 0.4831,
      "mean_token_accuracy": 0.8449182920157909,
      "num_tokens": 861528116.0,
      "step": 26994
    },
    {
      "entropy": 0.5937807131558657,
      "epoch": 2.4836774946665057,
      "grad_norm": 0.16900883615016937,
      "learning_rate": 1.7214095132946915e-05,
      "loss": 0.5809,
      "mean_token_accuracy": 0.817290086299181,
      "num_tokens": 861559475.0,
      "step": 26995
    },
    {
      "entropy": 0.524427080526948,
      "epoch": 2.483769501037947,
      "grad_norm": 0.15570838749408722,
      "learning_rate": 1.721102830680529e-05,
      "loss": 0.5102,
      "mean_token_accuracy": 0.8377375304698944,
      "num_tokens": 861591629.0,
      "step": 26996
    },
    {
      "entropy": 0.5600094515830278,
      "epoch": 2.483861507409388,
      "grad_norm": 0.169827401638031,
      "learning_rate": 1.720796148066366e-05,
      "loss": 0.5437,
      "mean_token_accuracy": 0.8311080411076546,
      "num_tokens": 861623289.0,
      "step": 26997
    },
    {
      "entropy": 0.45230552973225713,
      "epoch": 2.4839535137808295,
      "grad_norm": 0.1534481942653656,
      "learning_rate": 1.7204894654522035e-05,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8615865781903267,
      "num_tokens": 861655398.0,
      "step": 26998
    },
    {
      "entropy": 0.47870420571416616,
      "epoch": 2.4840455201522706,
      "grad_norm": 0.15304110944271088,
      "learning_rate": 1.720182782838041e-05,
      "loss": 0.4639,
      "mean_token_accuracy": 0.8512802235782146,
      "num_tokens": 861687581.0,
      "step": 26999
    },
    {
      "entropy": 0.5543196313083172,
      "epoch": 2.484137526523712,
      "grad_norm": 0.1684611737728119,
      "learning_rate": 1.7198761002238785e-05,
      "loss": 0.5386,
      "mean_token_accuracy": 0.8313134834170341,
      "num_tokens": 861719267.0,
      "step": 27000
    },
    {
      "entropy": 0.4095446215942502,
      "epoch": 2.484229532895153,
      "grad_norm": 0.14926382899284363,
      "learning_rate": 1.7195694176097156e-05,
      "loss": 0.3946,
      "mean_token_accuracy": 0.8723492175340652,
      "num_tokens": 861750845.0,
      "step": 27001
    },
    {
      "entropy": 0.534051826223731,
      "epoch": 2.484321539266594,
      "grad_norm": 0.1564432978630066,
      "learning_rate": 1.719262734995553e-05,
      "loss": 0.5197,
      "mean_token_accuracy": 0.837386392056942,
      "num_tokens": 861783108.0,
      "step": 27002
    },
    {
      "entropy": 0.5271534575149417,
      "epoch": 2.4844135456380356,
      "grad_norm": 0.1638278216123581,
      "learning_rate": 1.7189560523813905e-05,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8355094343423843,
      "num_tokens": 861814937.0,
      "step": 27003
    },
    {
      "entropy": 0.33123617712408304,
      "epoch": 2.4845055520094768,
      "grad_norm": 0.12957432866096497,
      "learning_rate": 1.718649369767228e-05,
      "loss": 0.3154,
      "mean_token_accuracy": 0.8984794802963734,
      "num_tokens": 861847705.0,
      "step": 27004
    },
    {
      "entropy": 0.5733647001907229,
      "epoch": 2.484597558380918,
      "grad_norm": 0.17083929479122162,
      "learning_rate": 1.7183426871530655e-05,
      "loss": 0.5618,
      "mean_token_accuracy": 0.8214659467339516,
      "num_tokens": 861879063.0,
      "step": 27005
    },
    {
      "entropy": 0.5521057639271021,
      "epoch": 2.484689564752359,
      "grad_norm": 0.1645837426185608,
      "learning_rate": 1.7180360045389026e-05,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8319636881351471,
      "num_tokens": 861911082.0,
      "step": 27006
    },
    {
      "entropy": 0.5452729053795338,
      "epoch": 2.4847815711238,
      "grad_norm": 0.16488100588321686,
      "learning_rate": 1.71772932192474e-05,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8280398808419704,
      "num_tokens": 861942902.0,
      "step": 27007
    },
    {
      "entropy": 0.5422455202788115,
      "epoch": 2.4848735774952417,
      "grad_norm": 0.16800974309444427,
      "learning_rate": 1.7174226393105775e-05,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8284241519868374,
      "num_tokens": 861975193.0,
      "step": 27008
    },
    {
      "entropy": 0.6086358781903982,
      "epoch": 2.484965583866683,
      "grad_norm": 0.17170852422714233,
      "learning_rate": 1.717115956696415e-05,
      "loss": 0.6033,
      "mean_token_accuracy": 0.8103258088231087,
      "num_tokens": 862006879.0,
      "step": 27009
    },
    {
      "entropy": 0.42874744441360235,
      "epoch": 2.485057590238124,
      "grad_norm": 0.14701233804225922,
      "learning_rate": 1.716809274082252e-05,
      "loss": 0.414,
      "mean_token_accuracy": 0.8628704957664013,
      "num_tokens": 862038830.0,
      "step": 27010
    },
    {
      "entropy": 0.49673283752053976,
      "epoch": 2.485149596609565,
      "grad_norm": 0.15420666337013245,
      "learning_rate": 1.71650259146809e-05,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8449597619473934,
      "num_tokens": 862071164.0,
      "step": 27011
    },
    {
      "entropy": 0.5750799868255854,
      "epoch": 2.4852416029810063,
      "grad_norm": 0.16511820256710052,
      "learning_rate": 1.7161959088539274e-05,
      "loss": 0.5649,
      "mean_token_accuracy": 0.8195070363581181,
      "num_tokens": 862102727.0,
      "step": 27012
    },
    {
      "entropy": 0.5196877233684063,
      "epoch": 2.485333609352448,
      "grad_norm": 0.15979669988155365,
      "learning_rate": 1.715889226239765e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.835473321378231,
      "num_tokens": 862135434.0,
      "step": 27013
    },
    {
      "entropy": 0.4570322511717677,
      "epoch": 2.485425615723889,
      "grad_norm": 0.15357311069965363,
      "learning_rate": 1.715582543625602e-05,
      "loss": 0.4462,
      "mean_token_accuracy": 0.8583311140537262,
      "num_tokens": 862167277.0,
      "step": 27014
    },
    {
      "entropy": 0.5371008664369583,
      "epoch": 2.48551762209533,
      "grad_norm": 0.15382297337055206,
      "learning_rate": 1.7152758610114394e-05,
      "loss": 0.5234,
      "mean_token_accuracy": 0.8360143676400185,
      "num_tokens": 862199640.0,
      "step": 27015
    },
    {
      "entropy": 0.5411534085869789,
      "epoch": 2.4856096284667712,
      "grad_norm": 0.1612299233675003,
      "learning_rate": 1.714969178397277e-05,
      "loss": 0.5307,
      "mean_token_accuracy": 0.8299450613558292,
      "num_tokens": 862231741.0,
      "step": 27016
    },
    {
      "entropy": 0.5126406261697412,
      "epoch": 2.4857016348382124,
      "grad_norm": 0.16124093532562256,
      "learning_rate": 1.7146624957831144e-05,
      "loss": 0.4963,
      "mean_token_accuracy": 0.841754361987114,
      "num_tokens": 862262719.0,
      "step": 27017
    },
    {
      "entropy": 0.6381918415427208,
      "epoch": 2.485793641209654,
      "grad_norm": 0.17615698277950287,
      "learning_rate": 1.7143558131689515e-05,
      "loss": 0.6343,
      "mean_token_accuracy": 0.8042179569602013,
      "num_tokens": 862295306.0,
      "step": 27018
    },
    {
      "entropy": 0.5011485880240798,
      "epoch": 2.485885647581095,
      "grad_norm": 0.15339751541614532,
      "learning_rate": 1.714049130554789e-05,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8465350680053234,
      "num_tokens": 862326473.0,
      "step": 27019
    },
    {
      "entropy": 0.6382611803710461,
      "epoch": 2.4859776539525362,
      "grad_norm": 0.1777593493461609,
      "learning_rate": 1.7137424479406264e-05,
      "loss": 0.6272,
      "mean_token_accuracy": 0.8024968020617962,
      "num_tokens": 862357821.0,
      "step": 27020
    },
    {
      "entropy": 0.5336139444261789,
      "epoch": 2.4860696603239774,
      "grad_norm": 0.16941116750240326,
      "learning_rate": 1.713435765326464e-05,
      "loss": 0.5211,
      "mean_token_accuracy": 0.8368829935789108,
      "num_tokens": 862388657.0,
      "step": 27021
    },
    {
      "entropy": 0.5768440105021,
      "epoch": 2.4861616666954185,
      "grad_norm": 0.16376826167106628,
      "learning_rate": 1.713129082712301e-05,
      "loss": 0.5697,
      "mean_token_accuracy": 0.8220066353678703,
      "num_tokens": 862420727.0,
      "step": 27022
    },
    {
      "entropy": 0.5410972470417619,
      "epoch": 2.48625367306686,
      "grad_norm": 0.1640252321958542,
      "learning_rate": 1.7128224000981385e-05,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8318216614425182,
      "num_tokens": 862452163.0,
      "step": 27023
    },
    {
      "entropy": 0.4896414680406451,
      "epoch": 2.486345679438301,
      "grad_norm": 0.1584080159664154,
      "learning_rate": 1.712515717483976e-05,
      "loss": 0.4882,
      "mean_token_accuracy": 0.8445382490754128,
      "num_tokens": 862483974.0,
      "step": 27024
    },
    {
      "entropy": 0.44797405879944563,
      "epoch": 2.4864376858097423,
      "grad_norm": 0.15135911107063293,
      "learning_rate": 1.7122090348698134e-05,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8593875728547573,
      "num_tokens": 862515719.0,
      "step": 27025
    },
    {
      "entropy": 0.5680500315502286,
      "epoch": 2.4865296921811835,
      "grad_norm": 0.16470970213413239,
      "learning_rate": 1.711902352255651e-05,
      "loss": 0.5612,
      "mean_token_accuracy": 0.8234197162091732,
      "num_tokens": 862547869.0,
      "step": 27026
    },
    {
      "entropy": 0.6229239087551832,
      "epoch": 2.4866216985526246,
      "grad_norm": 0.17031888663768768,
      "learning_rate": 1.711595669641488e-05,
      "loss": 0.5991,
      "mean_token_accuracy": 0.8119718693196774,
      "num_tokens": 862580602.0,
      "step": 27027
    },
    {
      "entropy": 0.4024381865747273,
      "epoch": 2.486713704924066,
      "grad_norm": 0.1460733562707901,
      "learning_rate": 1.7112889870273254e-05,
      "loss": 0.3885,
      "mean_token_accuracy": 0.8726622015237808,
      "num_tokens": 862612602.0,
      "step": 27028
    },
    {
      "entropy": 0.5889940671622753,
      "epoch": 2.4868057112955073,
      "grad_norm": 0.1713707447052002,
      "learning_rate": 1.710982304413163e-05,
      "loss": 0.5754,
      "mean_token_accuracy": 0.8189929574728012,
      "num_tokens": 862644792.0,
      "step": 27029
    },
    {
      "entropy": 0.5252693500369787,
      "epoch": 2.4868977176669484,
      "grad_norm": 0.167427197098732,
      "learning_rate": 1.7106756217990004e-05,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8389829881489277,
      "num_tokens": 862677046.0,
      "step": 27030
    },
    {
      "entropy": 0.6199880540370941,
      "epoch": 2.4869897240383896,
      "grad_norm": 0.17175409197807312,
      "learning_rate": 1.7103689391848375e-05,
      "loss": 0.612,
      "mean_token_accuracy": 0.8111811093986034,
      "num_tokens": 862708662.0,
      "step": 27031
    },
    {
      "entropy": 0.6114432495087385,
      "epoch": 2.4870817304098307,
      "grad_norm": 0.17210939526557922,
      "learning_rate": 1.710062256570675e-05,
      "loss": 0.6011,
      "mean_token_accuracy": 0.8079631105065346,
      "num_tokens": 862740680.0,
      "step": 27032
    },
    {
      "entropy": 0.5122226849198341,
      "epoch": 2.4871737367812723,
      "grad_norm": 0.1628570556640625,
      "learning_rate": 1.7097555739565124e-05,
      "loss": 0.5075,
      "mean_token_accuracy": 0.8407013081014156,
      "num_tokens": 862771924.0,
      "step": 27033
    },
    {
      "entropy": 0.5159757458604872,
      "epoch": 2.4872657431527134,
      "grad_norm": 0.15380726754665375,
      "learning_rate": 1.70944889134235e-05,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8426057025790215,
      "num_tokens": 862803713.0,
      "step": 27034
    },
    {
      "entropy": 0.49932063929736614,
      "epoch": 2.4873577495241546,
      "grad_norm": 0.15687288343906403,
      "learning_rate": 1.7091422087281874e-05,
      "loss": 0.4773,
      "mean_token_accuracy": 0.8470751270651817,
      "num_tokens": 862835448.0,
      "step": 27035
    },
    {
      "entropy": 0.5545489061623812,
      "epoch": 2.4874497558955957,
      "grad_norm": 0.1637856364250183,
      "learning_rate": 1.7088355261140245e-05,
      "loss": 0.5408,
      "mean_token_accuracy": 0.8301152288913727,
      "num_tokens": 862867358.0,
      "step": 27036
    },
    {
      "entropy": 0.5651116725057364,
      "epoch": 2.487541762267037,
      "grad_norm": 0.16468088328838348,
      "learning_rate": 1.708528843499862e-05,
      "loss": 0.5523,
      "mean_token_accuracy": 0.8264228142797947,
      "num_tokens": 862900126.0,
      "step": 27037
    },
    {
      "entropy": 0.5082933218218386,
      "epoch": 2.4876337686384784,
      "grad_norm": 0.1571376919746399,
      "learning_rate": 1.7082221608856994e-05,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8418377488851547,
      "num_tokens": 862932639.0,
      "step": 27038
    },
    {
      "entropy": 0.5254828240722418,
      "epoch": 2.4877257750099195,
      "grad_norm": 0.16113637387752533,
      "learning_rate": 1.707915478271537e-05,
      "loss": 0.5203,
      "mean_token_accuracy": 0.834993701428175,
      "num_tokens": 862964878.0,
      "step": 27039
    },
    {
      "entropy": 0.5856927931308746,
      "epoch": 2.4878177813813607,
      "grad_norm": 0.16932478547096252,
      "learning_rate": 1.707608795657374e-05,
      "loss": 0.5756,
      "mean_token_accuracy": 0.8216099590063095,
      "num_tokens": 862996215.0,
      "step": 27040
    },
    {
      "entropy": 0.48061622702516615,
      "epoch": 2.487909787752802,
      "grad_norm": 0.1501913070678711,
      "learning_rate": 1.7073021130432115e-05,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8511023409664631,
      "num_tokens": 863028198.0,
      "step": 27041
    },
    {
      "entropy": 0.498581413179636,
      "epoch": 2.488001794124243,
      "grad_norm": 0.16153207421302795,
      "learning_rate": 1.7069954304290493e-05,
      "loss": 0.4893,
      "mean_token_accuracy": 0.845245249569416,
      "num_tokens": 863060424.0,
      "step": 27042
    },
    {
      "entropy": 0.520514382282272,
      "epoch": 2.4880938004956845,
      "grad_norm": 0.156950905919075,
      "learning_rate": 1.7066887478148867e-05,
      "loss": 0.5114,
      "mean_token_accuracy": 0.8337589800357819,
      "num_tokens": 863092251.0,
      "step": 27043
    },
    {
      "entropy": 0.5721728727221489,
      "epoch": 2.4881858068671256,
      "grad_norm": 0.16869403421878815,
      "learning_rate": 1.706382065200724e-05,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8278770335018635,
      "num_tokens": 863124440.0,
      "step": 27044
    },
    {
      "entropy": 0.4560974187916145,
      "epoch": 2.488277813238567,
      "grad_norm": 0.14984993636608124,
      "learning_rate": 1.7060753825865613e-05,
      "loss": 0.4434,
      "mean_token_accuracy": 0.8623925410211086,
      "num_tokens": 863156654.0,
      "step": 27045
    },
    {
      "entropy": 0.4636584408581257,
      "epoch": 2.488369819610008,
      "grad_norm": 0.1586962640285492,
      "learning_rate": 1.7057686999723988e-05,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8524904884397984,
      "num_tokens": 863189095.0,
      "step": 27046
    },
    {
      "entropy": 0.5113141434267163,
      "epoch": 2.488461825981449,
      "grad_norm": 0.15960216522216797,
      "learning_rate": 1.7054620173582363e-05,
      "loss": 0.4912,
      "mean_token_accuracy": 0.8418457023799419,
      "num_tokens": 863220617.0,
      "step": 27047
    },
    {
      "entropy": 0.5954549638554454,
      "epoch": 2.4885538323528906,
      "grad_norm": 0.17196395993232727,
      "learning_rate": 1.7051553347440734e-05,
      "loss": 0.5757,
      "mean_token_accuracy": 0.8174654617905617,
      "num_tokens": 863252799.0,
      "step": 27048
    },
    {
      "entropy": 0.5678101485827938,
      "epoch": 2.4886458387243318,
      "grad_norm": 0.16757822036743164,
      "learning_rate": 1.704848652129911e-05,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8302439302206039,
      "num_tokens": 863285092.0,
      "step": 27049
    },
    {
      "entropy": 0.5816477024927735,
      "epoch": 2.488737845095773,
      "grad_norm": 0.17043234407901764,
      "learning_rate": 1.7045419695157483e-05,
      "loss": 0.5709,
      "mean_token_accuracy": 0.8245456181466579,
      "num_tokens": 863316179.0,
      "step": 27050
    },
    {
      "entropy": 0.49046191573143005,
      "epoch": 2.488829851467214,
      "grad_norm": 0.16502012312412262,
      "learning_rate": 1.7042352869015858e-05,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8474429734051228,
      "num_tokens": 863348527.0,
      "step": 27051
    },
    {
      "entropy": 0.5734926522709429,
      "epoch": 2.488921857838655,
      "grad_norm": 0.16239479184150696,
      "learning_rate": 1.703928604287423e-05,
      "loss": 0.5559,
      "mean_token_accuracy": 0.826150793582201,
      "num_tokens": 863380182.0,
      "step": 27052
    },
    {
      "entropy": 0.44396356493234634,
      "epoch": 2.4890138642100967,
      "grad_norm": 0.1547837108373642,
      "learning_rate": 1.7036219216732604e-05,
      "loss": 0.4373,
      "mean_token_accuracy": 0.8610099852085114,
      "num_tokens": 863412387.0,
      "step": 27053
    },
    {
      "entropy": 0.4536458239890635,
      "epoch": 2.489105870581538,
      "grad_norm": 0.15396054089069366,
      "learning_rate": 1.703315239059098e-05,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8577491380274296,
      "num_tokens": 863444384.0,
      "step": 27054
    },
    {
      "entropy": 0.5124621465802193,
      "epoch": 2.489197876952979,
      "grad_norm": 0.15948541462421417,
      "learning_rate": 1.7030085564449353e-05,
      "loss": 0.5026,
      "mean_token_accuracy": 0.8382447920739651,
      "num_tokens": 863476694.0,
      "step": 27055
    },
    {
      "entropy": 0.5835613375529647,
      "epoch": 2.48928988332442,
      "grad_norm": 0.1784055233001709,
      "learning_rate": 1.7027018738307728e-05,
      "loss": 0.5818,
      "mean_token_accuracy": 0.8176175951957703,
      "num_tokens": 863508100.0,
      "step": 27056
    },
    {
      "entropy": 0.5677222292870283,
      "epoch": 2.4893818896958613,
      "grad_norm": 0.16950908303260803,
      "learning_rate": 1.70239519121661e-05,
      "loss": 0.5562,
      "mean_token_accuracy": 0.8238963931798935,
      "num_tokens": 863539855.0,
      "step": 27057
    },
    {
      "entropy": 0.48247809056192636,
      "epoch": 2.489473896067303,
      "grad_norm": 0.15459443628787994,
      "learning_rate": 1.7020885086024474e-05,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8466948680579662,
      "num_tokens": 863571443.0,
      "step": 27058
    },
    {
      "entropy": 0.599893250502646,
      "epoch": 2.489565902438744,
      "grad_norm": 0.16596205532550812,
      "learning_rate": 1.7017818259882848e-05,
      "loss": 0.5991,
      "mean_token_accuracy": 0.8148441910743713,
      "num_tokens": 863603989.0,
      "step": 27059
    },
    {
      "entropy": 0.5778701901435852,
      "epoch": 2.489657908810185,
      "grad_norm": 0.17175792157649994,
      "learning_rate": 1.7014751433741223e-05,
      "loss": 0.5765,
      "mean_token_accuracy": 0.8151457570493221,
      "num_tokens": 863636362.0,
      "step": 27060
    },
    {
      "entropy": 0.4363351615611464,
      "epoch": 2.4897499151816262,
      "grad_norm": 0.150569349527359,
      "learning_rate": 1.7011684607599594e-05,
      "loss": 0.431,
      "mean_token_accuracy": 0.8635459281504154,
      "num_tokens": 863668785.0,
      "step": 27061
    },
    {
      "entropy": 0.5150648010894656,
      "epoch": 2.4898419215530674,
      "grad_norm": 0.16270124912261963,
      "learning_rate": 1.700861778145797e-05,
      "loss": 0.513,
      "mean_token_accuracy": 0.8354131616652012,
      "num_tokens": 863700549.0,
      "step": 27062
    },
    {
      "entropy": 0.501937665976584,
      "epoch": 2.489933927924509,
      "grad_norm": 0.1579204499721527,
      "learning_rate": 1.7005550955316343e-05,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8421862609684467,
      "num_tokens": 863732871.0,
      "step": 27063
    },
    {
      "entropy": 0.48978302569594234,
      "epoch": 2.49002593429595,
      "grad_norm": 0.15665733814239502,
      "learning_rate": 1.7002484129174718e-05,
      "loss": 0.4803,
      "mean_token_accuracy": 0.8481318913400173,
      "num_tokens": 863763871.0,
      "step": 27064
    },
    {
      "entropy": 0.5086800919380039,
      "epoch": 2.4901179406673912,
      "grad_norm": 0.16115927696228027,
      "learning_rate": 1.6999417303033093e-05,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8387540467083454,
      "num_tokens": 863795976.0,
      "step": 27065
    },
    {
      "entropy": 0.545621263794601,
      "epoch": 2.4902099470388324,
      "grad_norm": 0.16510282456874847,
      "learning_rate": 1.6996350476891464e-05,
      "loss": 0.5265,
      "mean_token_accuracy": 0.836009468883276,
      "num_tokens": 863827878.0,
      "step": 27066
    },
    {
      "entropy": 0.5374722965061665,
      "epoch": 2.4903019534102735,
      "grad_norm": 0.163176029920578,
      "learning_rate": 1.699328365074984e-05,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8340047188103199,
      "num_tokens": 863859238.0,
      "step": 27067
    },
    {
      "entropy": 0.5590057019144297,
      "epoch": 2.490393959781715,
      "grad_norm": 0.16405555605888367,
      "learning_rate": 1.6990216824608213e-05,
      "loss": 0.5389,
      "mean_token_accuracy": 0.829185176640749,
      "num_tokens": 863891187.0,
      "step": 27068
    },
    {
      "entropy": 0.5406850930303335,
      "epoch": 2.490485966153156,
      "grad_norm": 0.1527196615934372,
      "learning_rate": 1.6987149998466588e-05,
      "loss": 0.5156,
      "mean_token_accuracy": 0.8371396139264107,
      "num_tokens": 863922776.0,
      "step": 27069
    },
    {
      "entropy": 0.5408521969802678,
      "epoch": 2.4905779725245973,
      "grad_norm": 0.15824396908283234,
      "learning_rate": 1.698408317232496e-05,
      "loss": 0.5216,
      "mean_token_accuracy": 0.8369349129498005,
      "num_tokens": 863954988.0,
      "step": 27070
    },
    {
      "entropy": 0.5941262962296605,
      "epoch": 2.4906699788960385,
      "grad_norm": 0.1679355651140213,
      "learning_rate": 1.6981016346183334e-05,
      "loss": 0.576,
      "mean_token_accuracy": 0.8196395225822926,
      "num_tokens": 863987072.0,
      "step": 27071
    },
    {
      "entropy": 0.5492814308963716,
      "epoch": 2.4907619852674796,
      "grad_norm": 0.1605878472328186,
      "learning_rate": 1.6977949520041712e-05,
      "loss": 0.534,
      "mean_token_accuracy": 0.8316535241901875,
      "num_tokens": 864019391.0,
      "step": 27072
    },
    {
      "entropy": 0.45814819168299437,
      "epoch": 2.490853991638921,
      "grad_norm": 0.1484450101852417,
      "learning_rate": 1.6974882693900087e-05,
      "loss": 0.4367,
      "mean_token_accuracy": 0.8607558310031891,
      "num_tokens": 864051543.0,
      "step": 27073
    },
    {
      "entropy": 0.5582690881565213,
      "epoch": 2.4909459980103623,
      "grad_norm": 0.16145147383213043,
      "learning_rate": 1.6971815867758458e-05,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8259296007454395,
      "num_tokens": 864083398.0,
      "step": 27074
    },
    {
      "entropy": 0.5266843067947775,
      "epoch": 2.4910380043818035,
      "grad_norm": 0.15592758357524872,
      "learning_rate": 1.6968749041616832e-05,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8349461555480957,
      "num_tokens": 864115540.0,
      "step": 27075
    },
    {
      "entropy": 0.5164959083776921,
      "epoch": 2.4911300107532446,
      "grad_norm": 0.1620248258113861,
      "learning_rate": 1.6965682215475207e-05,
      "loss": 0.4963,
      "mean_token_accuracy": 0.8414074629545212,
      "num_tokens": 864146421.0,
      "step": 27076
    },
    {
      "entropy": 0.5649430891498923,
      "epoch": 2.4912220171246857,
      "grad_norm": 0.16819241642951965,
      "learning_rate": 1.6962615389333582e-05,
      "loss": 0.5561,
      "mean_token_accuracy": 0.8263146989047527,
      "num_tokens": 864177559.0,
      "step": 27077
    },
    {
      "entropy": 0.5143844978883862,
      "epoch": 2.4913140234961273,
      "grad_norm": 0.16109462082386017,
      "learning_rate": 1.6959548563191953e-05,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8401812128722668,
      "num_tokens": 864209635.0,
      "step": 27078
    },
    {
      "entropy": 0.5596768893301487,
      "epoch": 2.4914060298675684,
      "grad_norm": 0.16155818104743958,
      "learning_rate": 1.6956481737050328e-05,
      "loss": 0.5449,
      "mean_token_accuracy": 0.8273086473345757,
      "num_tokens": 864241602.0,
      "step": 27079
    },
    {
      "entropy": 0.5281992387026548,
      "epoch": 2.4914980362390096,
      "grad_norm": 0.15943922102451324,
      "learning_rate": 1.6953414910908702e-05,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8361123949289322,
      "num_tokens": 864273965.0,
      "step": 27080
    },
    {
      "entropy": 0.5370630798861384,
      "epoch": 2.4915900426104507,
      "grad_norm": 0.15945008397102356,
      "learning_rate": 1.6950348084767077e-05,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8312040083110332,
      "num_tokens": 864306244.0,
      "step": 27081
    },
    {
      "entropy": 0.49420143687166274,
      "epoch": 2.491682048981892,
      "grad_norm": 0.16097983717918396,
      "learning_rate": 1.6947281258625448e-05,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8478344157338142,
      "num_tokens": 864337481.0,
      "step": 27082
    },
    {
      "entropy": 0.5978228021413088,
      "epoch": 2.4917740553533334,
      "grad_norm": 0.17153562605381012,
      "learning_rate": 1.6944214432483823e-05,
      "loss": 0.5975,
      "mean_token_accuracy": 0.8108211532235146,
      "num_tokens": 864369150.0,
      "step": 27083
    },
    {
      "entropy": 0.544031661003828,
      "epoch": 2.4918660617247745,
      "grad_norm": 0.1634170114994049,
      "learning_rate": 1.6941147606342197e-05,
      "loss": 0.5396,
      "mean_token_accuracy": 0.8314764685928822,
      "num_tokens": 864400459.0,
      "step": 27084
    },
    {
      "entropy": 0.46968803624622524,
      "epoch": 2.4919580680962157,
      "grad_norm": 0.1533164083957672,
      "learning_rate": 1.6938080780200572e-05,
      "loss": 0.457,
      "mean_token_accuracy": 0.8537419475615025,
      "num_tokens": 864432166.0,
      "step": 27085
    },
    {
      "entropy": 0.47280273772776127,
      "epoch": 2.492050074467657,
      "grad_norm": 0.1559423804283142,
      "learning_rate": 1.6935013954058947e-05,
      "loss": 0.4596,
      "mean_token_accuracy": 0.8538725227117538,
      "num_tokens": 864464366.0,
      "step": 27086
    },
    {
      "entropy": 0.49060697108507156,
      "epoch": 2.492142080839098,
      "grad_norm": 0.153002068400383,
      "learning_rate": 1.6931947127917318e-05,
      "loss": 0.4784,
      "mean_token_accuracy": 0.8475185707211494,
      "num_tokens": 864496499.0,
      "step": 27087
    },
    {
      "entropy": 0.5394955398514867,
      "epoch": 2.4922340872105395,
      "grad_norm": 0.1620476096868515,
      "learning_rate": 1.6928880301775693e-05,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8357813097536564,
      "num_tokens": 864528258.0,
      "step": 27088
    },
    {
      "entropy": 0.4758501220494509,
      "epoch": 2.4923260935819807,
      "grad_norm": 0.15504099428653717,
      "learning_rate": 1.6925813475634067e-05,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8534991629421711,
      "num_tokens": 864559796.0,
      "step": 27089
    },
    {
      "entropy": 0.5118056247010827,
      "epoch": 2.492418099953422,
      "grad_norm": 0.16243471205234528,
      "learning_rate": 1.6922746649492442e-05,
      "loss": 0.4863,
      "mean_token_accuracy": 0.8438524901866913,
      "num_tokens": 864591737.0,
      "step": 27090
    },
    {
      "entropy": 0.5400791494175792,
      "epoch": 2.492510106324863,
      "grad_norm": 0.15601591765880585,
      "learning_rate": 1.6919679823350813e-05,
      "loss": 0.5257,
      "mean_token_accuracy": 0.833458174020052,
      "num_tokens": 864623993.0,
      "step": 27091
    },
    {
      "entropy": 0.5235188659280539,
      "epoch": 2.492602112696304,
      "grad_norm": 0.1628274917602539,
      "learning_rate": 1.6916612997209188e-05,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8380469866096973,
      "num_tokens": 864655218.0,
      "step": 27092
    },
    {
      "entropy": 0.5666271345689893,
      "epoch": 2.4926941190677456,
      "grad_norm": 0.16934143006801605,
      "learning_rate": 1.6913546171067563e-05,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8216351307928562,
      "num_tokens": 864685975.0,
      "step": 27093
    },
    {
      "entropy": 0.4400050528347492,
      "epoch": 2.4927861254391868,
      "grad_norm": 0.14829832315444946,
      "learning_rate": 1.6910479344925937e-05,
      "loss": 0.4353,
      "mean_token_accuracy": 0.8606753535568714,
      "num_tokens": 864717828.0,
      "step": 27094
    },
    {
      "entropy": 0.5179439205676317,
      "epoch": 2.492878131810628,
      "grad_norm": 0.1664547622203827,
      "learning_rate": 1.6907412518784312e-05,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8355581760406494,
      "num_tokens": 864749827.0,
      "step": 27095
    },
    {
      "entropy": 0.45895533822476864,
      "epoch": 2.492970138182069,
      "grad_norm": 0.15343932807445526,
      "learning_rate": 1.6904345692642683e-05,
      "loss": 0.4501,
      "mean_token_accuracy": 0.8538447953760624,
      "num_tokens": 864781845.0,
      "step": 27096
    },
    {
      "entropy": 0.450878219678998,
      "epoch": 2.49306214455351,
      "grad_norm": 0.15507271885871887,
      "learning_rate": 1.6901278866501058e-05,
      "loss": 0.4401,
      "mean_token_accuracy": 0.8573752529919147,
      "num_tokens": 864812436.0,
      "step": 27097
    },
    {
      "entropy": 0.6275463495403528,
      "epoch": 2.4931541509249517,
      "grad_norm": 0.17292697727680206,
      "learning_rate": 1.6898212040359432e-05,
      "loss": 0.6111,
      "mean_token_accuracy": 0.8052139356732368,
      "num_tokens": 864843178.0,
      "step": 27098
    },
    {
      "entropy": 0.5026400880888104,
      "epoch": 2.493246157296393,
      "grad_norm": 0.15847846865653992,
      "learning_rate": 1.6895145214217807e-05,
      "loss": 0.4749,
      "mean_token_accuracy": 0.8475413583219051,
      "num_tokens": 864874282.0,
      "step": 27099
    },
    {
      "entropy": 0.5207033324986696,
      "epoch": 2.493338163667834,
      "grad_norm": 0.1593378633260727,
      "learning_rate": 1.6892078388076178e-05,
      "loss": 0.5076,
      "mean_token_accuracy": 0.839575283229351,
      "num_tokens": 864906427.0,
      "step": 27100
    },
    {
      "entropy": 0.459060488268733,
      "epoch": 2.493430170039275,
      "grad_norm": 0.157873272895813,
      "learning_rate": 1.6889011561934553e-05,
      "loss": 0.4526,
      "mean_token_accuracy": 0.8548579514026642,
      "num_tokens": 864938088.0,
      "step": 27101
    },
    {
      "entropy": 0.5621326379477978,
      "epoch": 2.4935221764107163,
      "grad_norm": 0.16284795105457306,
      "learning_rate": 1.6885944735792928e-05,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8279316686093807,
      "num_tokens": 864969943.0,
      "step": 27102
    },
    {
      "entropy": 0.5266023413278162,
      "epoch": 2.493614182782158,
      "grad_norm": 0.1617995798587799,
      "learning_rate": 1.6882877909651306e-05,
      "loss": 0.514,
      "mean_token_accuracy": 0.8395575322210789,
      "num_tokens": 865001281.0,
      "step": 27103
    },
    {
      "entropy": 0.4999895775690675,
      "epoch": 2.493706189153599,
      "grad_norm": 0.1566380262374878,
      "learning_rate": 1.6879811083509677e-05,
      "loss": 0.483,
      "mean_token_accuracy": 0.8453840054571629,
      "num_tokens": 865032696.0,
      "step": 27104
    },
    {
      "entropy": 0.6032780492678285,
      "epoch": 2.49379819552504,
      "grad_norm": 0.16454647481441498,
      "learning_rate": 1.687674425736805e-05,
      "loss": 0.592,
      "mean_token_accuracy": 0.8153151348233223,
      "num_tokens": 865065262.0,
      "step": 27105
    },
    {
      "entropy": 0.5375175811350346,
      "epoch": 2.4938902018964813,
      "grad_norm": 0.16507229208946228,
      "learning_rate": 1.6873677431226426e-05,
      "loss": 0.5302,
      "mean_token_accuracy": 0.8291197158396244,
      "num_tokens": 865097499.0,
      "step": 27106
    },
    {
      "entropy": 0.517656734213233,
      "epoch": 2.4939822082679224,
      "grad_norm": 0.15659821033477783,
      "learning_rate": 1.68706106050848e-05,
      "loss": 0.5066,
      "mean_token_accuracy": 0.8390302993357182,
      "num_tokens": 865130132.0,
      "step": 27107
    },
    {
      "entropy": 0.5919643845409155,
      "epoch": 2.494074214639364,
      "grad_norm": 0.1675519496202469,
      "learning_rate": 1.6867543778943172e-05,
      "loss": 0.5785,
      "mean_token_accuracy": 0.8139462918043137,
      "num_tokens": 865161963.0,
      "step": 27108
    },
    {
      "entropy": 0.5537127107381821,
      "epoch": 2.494166221010805,
      "grad_norm": 0.16293387115001678,
      "learning_rate": 1.6864476952801547e-05,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8273540213704109,
      "num_tokens": 865193600.0,
      "step": 27109
    },
    {
      "entropy": 0.6373244696296751,
      "epoch": 2.4942582273822462,
      "grad_norm": 0.1667211353778839,
      "learning_rate": 1.686141012665992e-05,
      "loss": 0.6228,
      "mean_token_accuracy": 0.8015719205141068,
      "num_tokens": 865226129.0,
      "step": 27110
    },
    {
      "entropy": 0.5757580529898405,
      "epoch": 2.4943502337536874,
      "grad_norm": 0.16969501972198486,
      "learning_rate": 1.6858343300518296e-05,
      "loss": 0.5646,
      "mean_token_accuracy": 0.8181070871651173,
      "num_tokens": 865257689.0,
      "step": 27111
    },
    {
      "entropy": 0.5657982612028718,
      "epoch": 2.4944422401251285,
      "grad_norm": 0.16968843340873718,
      "learning_rate": 1.6855276474376667e-05,
      "loss": 0.5577,
      "mean_token_accuracy": 0.8216737136244774,
      "num_tokens": 865289221.0,
      "step": 27112
    },
    {
      "entropy": 0.5599340950138867,
      "epoch": 2.49453424649657,
      "grad_norm": 0.16316258907318115,
      "learning_rate": 1.6852209648235042e-05,
      "loss": 0.5488,
      "mean_token_accuracy": 0.8276719860732555,
      "num_tokens": 865321094.0,
      "step": 27113
    },
    {
      "entropy": 0.5150175783783197,
      "epoch": 2.494626252868011,
      "grad_norm": 0.16107231378555298,
      "learning_rate": 1.6849142822093417e-05,
      "loss": 0.5006,
      "mean_token_accuracy": 0.8398143574595451,
      "num_tokens": 865352728.0,
      "step": 27114
    },
    {
      "entropy": 0.4961637333035469,
      "epoch": 2.4947182592394523,
      "grad_norm": 0.16020193696022034,
      "learning_rate": 1.684607599595179e-05,
      "loss": 0.4817,
      "mean_token_accuracy": 0.848249863833189,
      "num_tokens": 865383902.0,
      "step": 27115
    },
    {
      "entropy": 0.46083161793649197,
      "epoch": 2.4948102656108935,
      "grad_norm": 0.15364105999469757,
      "learning_rate": 1.6843009169810166e-05,
      "loss": 0.4537,
      "mean_token_accuracy": 0.854517288506031,
      "num_tokens": 865416296.0,
      "step": 27116
    },
    {
      "entropy": 0.5239020606968552,
      "epoch": 2.4949022719823346,
      "grad_norm": 0.15786245465278625,
      "learning_rate": 1.6839942343668537e-05,
      "loss": 0.5221,
      "mean_token_accuracy": 0.8364038057625294,
      "num_tokens": 865448429.0,
      "step": 27117
    },
    {
      "entropy": 0.45509192906320095,
      "epoch": 2.494994278353776,
      "grad_norm": 0.153212770819664,
      "learning_rate": 1.6836875517526912e-05,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8627060726284981,
      "num_tokens": 865480860.0,
      "step": 27118
    },
    {
      "entropy": 0.5718605152796954,
      "epoch": 2.4950862847252173,
      "grad_norm": 0.1691250503063202,
      "learning_rate": 1.6833808691385286e-05,
      "loss": 0.5746,
      "mean_token_accuracy": 0.8207387700676918,
      "num_tokens": 865512814.0,
      "step": 27119
    },
    {
      "entropy": 0.4831044254824519,
      "epoch": 2.4951782910966585,
      "grad_norm": 0.1569659411907196,
      "learning_rate": 1.683074186524366e-05,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8485098108649254,
      "num_tokens": 865544997.0,
      "step": 27120
    },
    {
      "entropy": 0.5525877326726913,
      "epoch": 2.4952702974680996,
      "grad_norm": 0.1600838303565979,
      "learning_rate": 1.6827675039102032e-05,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8279799707233906,
      "num_tokens": 865577765.0,
      "step": 27121
    },
    {
      "entropy": 0.40581033658236265,
      "epoch": 2.4953623038395407,
      "grad_norm": 0.14633455872535706,
      "learning_rate": 1.6824608212960407e-05,
      "loss": 0.4051,
      "mean_token_accuracy": 0.8691484853625298,
      "num_tokens": 865610423.0,
      "step": 27122
    },
    {
      "entropy": 0.4635687004774809,
      "epoch": 2.4954543102109823,
      "grad_norm": 0.1435624659061432,
      "learning_rate": 1.682154138681878e-05,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8581909947097301,
      "num_tokens": 865643185.0,
      "step": 27123
    },
    {
      "entropy": 0.5430573839694262,
      "epoch": 2.4955463165824234,
      "grad_norm": 0.1549793928861618,
      "learning_rate": 1.6818474560677156e-05,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8368456140160561,
      "num_tokens": 865675346.0,
      "step": 27124
    },
    {
      "entropy": 0.5090134828351438,
      "epoch": 2.4956383229538646,
      "grad_norm": 0.1602858603000641,
      "learning_rate": 1.681540773453553e-05,
      "loss": 0.4894,
      "mean_token_accuracy": 0.847517766058445,
      "num_tokens": 865706964.0,
      "step": 27125
    },
    {
      "entropy": 0.40901371696963906,
      "epoch": 2.4957303293253057,
      "grad_norm": 0.1428505927324295,
      "learning_rate": 1.6812340908393902e-05,
      "loss": 0.4041,
      "mean_token_accuracy": 0.8705340996384621,
      "num_tokens": 865739168.0,
      "step": 27126
    },
    {
      "entropy": 0.41968059726059437,
      "epoch": 2.495822335696747,
      "grad_norm": 0.14425048232078552,
      "learning_rate": 1.6809274082252277e-05,
      "loss": 0.4057,
      "mean_token_accuracy": 0.8703207597136497,
      "num_tokens": 865771915.0,
      "step": 27127
    },
    {
      "entropy": 0.5657970085740089,
      "epoch": 2.4959143420681884,
      "grad_norm": 0.1649303287267685,
      "learning_rate": 1.680620725611065e-05,
      "loss": 0.558,
      "mean_token_accuracy": 0.8271622471511364,
      "num_tokens": 865804471.0,
      "step": 27128
    },
    {
      "entropy": 0.46038202568888664,
      "epoch": 2.4960063484396295,
      "grad_norm": 0.1488354206085205,
      "learning_rate": 1.6803140429969026e-05,
      "loss": 0.4473,
      "mean_token_accuracy": 0.8562032766640186,
      "num_tokens": 865837035.0,
      "step": 27129
    },
    {
      "entropy": 0.5566859398968518,
      "epoch": 2.4960983548110707,
      "grad_norm": 0.16660431027412415,
      "learning_rate": 1.6800073603827397e-05,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8303965032100677,
      "num_tokens": 865869413.0,
      "step": 27130
    },
    {
      "entropy": 0.5753195148427039,
      "epoch": 2.496190361182512,
      "grad_norm": 0.16764578223228455,
      "learning_rate": 1.6797006777685772e-05,
      "loss": 0.5633,
      "mean_token_accuracy": 0.8259486705064774,
      "num_tokens": 865901743.0,
      "step": 27131
    },
    {
      "entropy": 0.5165434945374727,
      "epoch": 2.496282367553953,
      "grad_norm": 0.15619045495986938,
      "learning_rate": 1.6793939951544147e-05,
      "loss": 0.5061,
      "mean_token_accuracy": 0.838468573987484,
      "num_tokens": 865933911.0,
      "step": 27132
    },
    {
      "entropy": 0.4970728103071451,
      "epoch": 2.4963743739253945,
      "grad_norm": 0.15970014035701752,
      "learning_rate": 1.6790873125402525e-05,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8463414944708347,
      "num_tokens": 865966324.0,
      "step": 27133
    },
    {
      "entropy": 0.4791879206895828,
      "epoch": 2.4964663802968357,
      "grad_norm": 0.15429987013339996,
      "learning_rate": 1.6787806299260896e-05,
      "loss": 0.4628,
      "mean_token_accuracy": 0.8509766198694706,
      "num_tokens": 865998307.0,
      "step": 27134
    },
    {
      "entropy": 0.5571706090122461,
      "epoch": 2.496558386668277,
      "grad_norm": 0.1684926301240921,
      "learning_rate": 1.678473947311927e-05,
      "loss": 0.548,
      "mean_token_accuracy": 0.8277069106698036,
      "num_tokens": 866030376.0,
      "step": 27135
    },
    {
      "entropy": 0.454028460662812,
      "epoch": 2.496650393039718,
      "grad_norm": 0.15068258345127106,
      "learning_rate": 1.6781672646977645e-05,
      "loss": 0.441,
      "mean_token_accuracy": 0.8604592010378838,
      "num_tokens": 866062133.0,
      "step": 27136
    },
    {
      "entropy": 0.4871986676007509,
      "epoch": 2.496742399411159,
      "grad_norm": 0.1606561243534088,
      "learning_rate": 1.677860582083602e-05,
      "loss": 0.4801,
      "mean_token_accuracy": 0.846719354391098,
      "num_tokens": 866094586.0,
      "step": 27137
    },
    {
      "entropy": 0.5425303189549595,
      "epoch": 2.4968344057826006,
      "grad_norm": 0.16421514749526978,
      "learning_rate": 1.677553899469439e-05,
      "loss": 0.5417,
      "mean_token_accuracy": 0.8329952172935009,
      "num_tokens": 866126628.0,
      "step": 27138
    },
    {
      "entropy": 0.5335106775164604,
      "epoch": 2.4969264121540418,
      "grad_norm": 0.16206233203411102,
      "learning_rate": 1.6772472168552766e-05,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8365765623748302,
      "num_tokens": 866158567.0,
      "step": 27139
    },
    {
      "entropy": 0.5806295676156878,
      "epoch": 2.497018418525483,
      "grad_norm": 0.16627387702465057,
      "learning_rate": 1.676940534241114e-05,
      "loss": 0.5592,
      "mean_token_accuracy": 0.8214837498962879,
      "num_tokens": 866190325.0,
      "step": 27140
    },
    {
      "entropy": 0.4637767574749887,
      "epoch": 2.497110424896924,
      "grad_norm": 0.15355604887008667,
      "learning_rate": 1.6766338516269515e-05,
      "loss": 0.4574,
      "mean_token_accuracy": 0.8528719656169415,
      "num_tokens": 866222736.0,
      "step": 27141
    },
    {
      "entropy": 0.5822288449853659,
      "epoch": 2.497202431268365,
      "grad_norm": 0.16672061383724213,
      "learning_rate": 1.6763271690127886e-05,
      "loss": 0.5678,
      "mean_token_accuracy": 0.8172484710812569,
      "num_tokens": 866254810.0,
      "step": 27142
    },
    {
      "entropy": 0.5135874757543206,
      "epoch": 2.4972944376398067,
      "grad_norm": 0.16142438352108002,
      "learning_rate": 1.676020486398626e-05,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8430912792682648,
      "num_tokens": 866287228.0,
      "step": 27143
    },
    {
      "entropy": 0.5122102051973343,
      "epoch": 2.497386444011248,
      "grad_norm": 0.1590030938386917,
      "learning_rate": 1.6757138037844636e-05,
      "loss": 0.495,
      "mean_token_accuracy": 0.8396421447396278,
      "num_tokens": 866319168.0,
      "step": 27144
    },
    {
      "entropy": 0.45932399574667215,
      "epoch": 2.497478450382689,
      "grad_norm": 0.15379150211811066,
      "learning_rate": 1.675407121170301e-05,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8565047457814217,
      "num_tokens": 866351250.0,
      "step": 27145
    },
    {
      "entropy": 0.5249312845990062,
      "epoch": 2.49757045675413,
      "grad_norm": 0.15952827036380768,
      "learning_rate": 1.6751004385561385e-05,
      "loss": 0.5192,
      "mean_token_accuracy": 0.8318683244287968,
      "num_tokens": 866383964.0,
      "step": 27146
    },
    {
      "entropy": 0.5339622460305691,
      "epoch": 2.4976624631255713,
      "grad_norm": 0.16286110877990723,
      "learning_rate": 1.6747937559419756e-05,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8335576951503754,
      "num_tokens": 866416002.0,
      "step": 27147
    },
    {
      "entropy": 0.44001844711601734,
      "epoch": 2.497754469497013,
      "grad_norm": 0.15054334700107574,
      "learning_rate": 1.674487073327813e-05,
      "loss": 0.4293,
      "mean_token_accuracy": 0.863052487373352,
      "num_tokens": 866447167.0,
      "step": 27148
    },
    {
      "entropy": 0.5769801829010248,
      "epoch": 2.497846475868454,
      "grad_norm": 0.16717112064361572,
      "learning_rate": 1.6741803907136506e-05,
      "loss": 0.5611,
      "mean_token_accuracy": 0.8204253539443016,
      "num_tokens": 866478354.0,
      "step": 27149
    },
    {
      "entropy": 0.4637340164044872,
      "epoch": 2.497938482239895,
      "grad_norm": 0.15791107714176178,
      "learning_rate": 1.673873708099488e-05,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8585390672087669,
      "num_tokens": 866509080.0,
      "step": 27150
    },
    {
      "entropy": 0.5115574449300766,
      "epoch": 2.4980304886113363,
      "grad_norm": 0.16080188751220703,
      "learning_rate": 1.673567025485325e-05,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8435153886675835,
      "num_tokens": 866540930.0,
      "step": 27151
    },
    {
      "entropy": 0.5486213993281126,
      "epoch": 2.4981224949827774,
      "grad_norm": 0.160066619515419,
      "learning_rate": 1.6732603428711626e-05,
      "loss": 0.5249,
      "mean_token_accuracy": 0.8303378745913506,
      "num_tokens": 866572240.0,
      "step": 27152
    },
    {
      "entropy": 0.45785934943705797,
      "epoch": 2.498214501354219,
      "grad_norm": 0.15622860193252563,
      "learning_rate": 1.672953660257e-05,
      "loss": 0.451,
      "mean_token_accuracy": 0.8534194603562355,
      "num_tokens": 866603876.0,
      "step": 27153
    },
    {
      "entropy": 0.5610766243189573,
      "epoch": 2.49830650772566,
      "grad_norm": 0.1655724197626114,
      "learning_rate": 1.6726469776428375e-05,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8253033719956875,
      "num_tokens": 866635998.0,
      "step": 27154
    },
    {
      "entropy": 0.47404667316004634,
      "epoch": 2.4983985140971012,
      "grad_norm": 0.1584269106388092,
      "learning_rate": 1.672340295028675e-05,
      "loss": 0.4664,
      "mean_token_accuracy": 0.8512740582227707,
      "num_tokens": 866667753.0,
      "step": 27155
    },
    {
      "entropy": 0.4983358010649681,
      "epoch": 2.4984905204685424,
      "grad_norm": 0.15714895725250244,
      "learning_rate": 1.672033612414512e-05,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8460053615272045,
      "num_tokens": 866699868.0,
      "step": 27156
    },
    {
      "entropy": 0.5599870393052697,
      "epoch": 2.4985825268399835,
      "grad_norm": 0.16484574973583221,
      "learning_rate": 1.6717269298003496e-05,
      "loss": 0.5517,
      "mean_token_accuracy": 0.8277806676924229,
      "num_tokens": 866730828.0,
      "step": 27157
    },
    {
      "entropy": 0.5762367686256766,
      "epoch": 2.498674533211425,
      "grad_norm": 0.1716500222682953,
      "learning_rate": 1.671420247186187e-05,
      "loss": 0.5628,
      "mean_token_accuracy": 0.820235975086689,
      "num_tokens": 866761826.0,
      "step": 27158
    },
    {
      "entropy": 0.6198794357478619,
      "epoch": 2.498766539582866,
      "grad_norm": 0.17072942852973938,
      "learning_rate": 1.6711135645720245e-05,
      "loss": 0.6105,
      "mean_token_accuracy": 0.806977067142725,
      "num_tokens": 866794294.0,
      "step": 27159
    },
    {
      "entropy": 0.45156453270465136,
      "epoch": 2.4988585459543073,
      "grad_norm": 0.14723259210586548,
      "learning_rate": 1.6708068819578616e-05,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8593243807554245,
      "num_tokens": 866826665.0,
      "step": 27160
    },
    {
      "entropy": 0.4082418151665479,
      "epoch": 2.4989505523257485,
      "grad_norm": 0.14660592377185822,
      "learning_rate": 1.670500199343699e-05,
      "loss": 0.4002,
      "mean_token_accuracy": 0.8737996853888035,
      "num_tokens": 866858458.0,
      "step": 27161
    },
    {
      "entropy": 0.5212910510599613,
      "epoch": 2.4990425586971896,
      "grad_norm": 0.16355876624584198,
      "learning_rate": 1.6701935167295366e-05,
      "loss": 0.512,
      "mean_token_accuracy": 0.8354893438518047,
      "num_tokens": 866891007.0,
      "step": 27162
    },
    {
      "entropy": 0.4749809140339494,
      "epoch": 2.499134565068631,
      "grad_norm": 0.15686818957328796,
      "learning_rate": 1.669886834115374e-05,
      "loss": 0.4634,
      "mean_token_accuracy": 0.855939507484436,
      "num_tokens": 866922443.0,
      "step": 27163
    },
    {
      "entropy": 0.45315312198363245,
      "epoch": 2.4992265714400723,
      "grad_norm": 0.15194858610630035,
      "learning_rate": 1.6695801515012115e-05,
      "loss": 0.4472,
      "mean_token_accuracy": 0.8597737364470959,
      "num_tokens": 866954415.0,
      "step": 27164
    },
    {
      "entropy": 0.5481878411956131,
      "epoch": 2.4993185778115135,
      "grad_norm": 0.15810824930667877,
      "learning_rate": 1.669273468887049e-05,
      "loss": 0.5469,
      "mean_token_accuracy": 0.8284385055303574,
      "num_tokens": 866986344.0,
      "step": 27165
    },
    {
      "entropy": 0.5592481608036906,
      "epoch": 2.4994105841829546,
      "grad_norm": 0.16588523983955383,
      "learning_rate": 1.6689667862728864e-05,
      "loss": 0.5508,
      "mean_token_accuracy": 0.8263690620660782,
      "num_tokens": 867018755.0,
      "step": 27166
    },
    {
      "entropy": 0.52544039580971,
      "epoch": 2.4995025905543957,
      "grad_norm": 0.1615496724843979,
      "learning_rate": 1.668660103658724e-05,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8386763408780098,
      "num_tokens": 867050984.0,
      "step": 27167
    },
    {
      "entropy": 0.4255006015300751,
      "epoch": 2.4995945969258373,
      "grad_norm": 0.1453103870153427,
      "learning_rate": 1.668353421044561e-05,
      "loss": 0.4128,
      "mean_token_accuracy": 0.8672592677175999,
      "num_tokens": 867083382.0,
      "step": 27168
    },
    {
      "entropy": 0.5395193789154291,
      "epoch": 2.4996866032972784,
      "grad_norm": 0.1601058393716812,
      "learning_rate": 1.6680467384303985e-05,
      "loss": 0.5326,
      "mean_token_accuracy": 0.8306919224560261,
      "num_tokens": 867115440.0,
      "step": 27169
    },
    {
      "entropy": 0.5995480660349131,
      "epoch": 2.4997786096687196,
      "grad_norm": 0.17216284573078156,
      "learning_rate": 1.667740055816236e-05,
      "loss": 0.5828,
      "mean_token_accuracy": 0.8177907876670361,
      "num_tokens": 867147647.0,
      "step": 27170
    },
    {
      "entropy": 0.5491877375170588,
      "epoch": 2.4998706160401607,
      "grad_norm": 0.16508464515209198,
      "learning_rate": 1.6674333732020734e-05,
      "loss": 0.5471,
      "mean_token_accuracy": 0.8302570320665836,
      "num_tokens": 867179705.0,
      "step": 27171
    },
    {
      "entropy": 0.45571235893294215,
      "epoch": 2.499962622411602,
      "grad_norm": 0.15475133061408997,
      "learning_rate": 1.6671266905879105e-05,
      "loss": 0.4452,
      "mean_token_accuracy": 0.8589723221957684,
      "num_tokens": 867211115.0,
      "step": 27172
    },
    {
      "entropy": 0.5214428342878819,
      "epoch": 2.500054628783043,
      "grad_norm": 0.16274124383926392,
      "learning_rate": 1.666820007973748e-05,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8395552523434162,
      "num_tokens": 867242566.0,
      "step": 27173
    },
    {
      "entropy": 0.4188756311777979,
      "epoch": 2.5001466351544845,
      "grad_norm": 0.1460035890340805,
      "learning_rate": 1.6665133253595855e-05,
      "loss": 0.4023,
      "mean_token_accuracy": 0.8675454743206501,
      "num_tokens": 867274512.0,
      "step": 27174
    },
    {
      "entropy": 0.5405476200394332,
      "epoch": 2.5002386415259257,
      "grad_norm": 0.1552581489086151,
      "learning_rate": 1.666206642745423e-05,
      "loss": 0.5214,
      "mean_token_accuracy": 0.8326511234045029,
      "num_tokens": 867306741.0,
      "step": 27175
    },
    {
      "entropy": 0.44116888102144003,
      "epoch": 2.500330647897367,
      "grad_norm": 0.14957819879055023,
      "learning_rate": 1.6658999601312604e-05,
      "loss": 0.4202,
      "mean_token_accuracy": 0.8664242848753929,
      "num_tokens": 867338937.0,
      "step": 27176
    },
    {
      "entropy": 0.4983610026538372,
      "epoch": 2.5004226542688084,
      "grad_norm": 0.15699787437915802,
      "learning_rate": 1.6655932775170975e-05,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8465467318892479,
      "num_tokens": 867371093.0,
      "step": 27177
    },
    {
      "entropy": 0.48897640127688646,
      "epoch": 2.500514660640249,
      "grad_norm": 0.15488770604133606,
      "learning_rate": 1.665286594902935e-05,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8497091829776764,
      "num_tokens": 867402575.0,
      "step": 27178
    },
    {
      "entropy": 0.4356646682135761,
      "epoch": 2.5006066670116907,
      "grad_norm": 0.15653081238269806,
      "learning_rate": 1.6649799122887725e-05,
      "loss": 0.4233,
      "mean_token_accuracy": 0.8641145117580891,
      "num_tokens": 867434481.0,
      "step": 27179
    },
    {
      "entropy": 0.5926628322340548,
      "epoch": 2.500698673383132,
      "grad_norm": 0.1677621454000473,
      "learning_rate": 1.66467322967461e-05,
      "loss": 0.58,
      "mean_token_accuracy": 0.8153310976922512,
      "num_tokens": 867466067.0,
      "step": 27180
    },
    {
      "entropy": 0.4615567720029503,
      "epoch": 2.500790679754573,
      "grad_norm": 0.1514979749917984,
      "learning_rate": 1.664366547060447e-05,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8549078404903412,
      "num_tokens": 867498336.0,
      "step": 27181
    },
    {
      "entropy": 0.49097431916743517,
      "epoch": 2.5008826861260145,
      "grad_norm": 0.154726043343544,
      "learning_rate": 1.6640598644462845e-05,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8457979634404182,
      "num_tokens": 867529886.0,
      "step": 27182
    },
    {
      "entropy": 0.5473486250266433,
      "epoch": 2.500974692497455,
      "grad_norm": 0.16107168793678284,
      "learning_rate": 1.663753181832122e-05,
      "loss": 0.5386,
      "mean_token_accuracy": 0.8303739130496979,
      "num_tokens": 867562243.0,
      "step": 27183
    },
    {
      "entropy": 0.46222113259136677,
      "epoch": 2.5010666988688968,
      "grad_norm": 0.1541031450033188,
      "learning_rate": 1.6634464992179594e-05,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8545008152723312,
      "num_tokens": 867594316.0,
      "step": 27184
    },
    {
      "entropy": 0.523082516156137,
      "epoch": 2.501158705240338,
      "grad_norm": 0.16368021070957184,
      "learning_rate": 1.663139816603797e-05,
      "loss": 0.5146,
      "mean_token_accuracy": 0.83724519982934,
      "num_tokens": 867625973.0,
      "step": 27185
    },
    {
      "entropy": 0.4521826785057783,
      "epoch": 2.501250711611779,
      "grad_norm": 0.14688967168331146,
      "learning_rate": 1.662833133989634e-05,
      "loss": 0.4284,
      "mean_token_accuracy": 0.8599988706409931,
      "num_tokens": 867657844.0,
      "step": 27186
    },
    {
      "entropy": 0.5034706555306911,
      "epoch": 2.5013427179832206,
      "grad_norm": 0.15955419838428497,
      "learning_rate": 1.6625264513754715e-05,
      "loss": 0.4936,
      "mean_token_accuracy": 0.8434852585196495,
      "num_tokens": 867690270.0,
      "step": 27187
    },
    {
      "entropy": 0.5499849766492844,
      "epoch": 2.5014347243546613,
      "grad_norm": 0.16850323975086212,
      "learning_rate": 1.662219768761309e-05,
      "loss": 0.5449,
      "mean_token_accuracy": 0.8327248990535736,
      "num_tokens": 867722049.0,
      "step": 27188
    },
    {
      "entropy": 0.5543526688124985,
      "epoch": 2.501526730726103,
      "grad_norm": 0.16273263096809387,
      "learning_rate": 1.6619130861471464e-05,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8270731866359711,
      "num_tokens": 867754178.0,
      "step": 27189
    },
    {
      "entropy": 0.5463441082974896,
      "epoch": 2.501618737097544,
      "grad_norm": 0.16081711649894714,
      "learning_rate": 1.6616064035329836e-05,
      "loss": 0.5324,
      "mean_token_accuracy": 0.8347094319760799,
      "num_tokens": 867786401.0,
      "step": 27190
    },
    {
      "entropy": 0.47751805651932955,
      "epoch": 2.501710743468985,
      "grad_norm": 0.15125037729740143,
      "learning_rate": 1.661299720918821e-05,
      "loss": 0.4648,
      "mean_token_accuracy": 0.8526227474212646,
      "num_tokens": 867818797.0,
      "step": 27191
    },
    {
      "entropy": 0.4284177925437689,
      "epoch": 2.5018027498404267,
      "grad_norm": 0.14597339928150177,
      "learning_rate": 1.6609930383046585e-05,
      "loss": 0.416,
      "mean_token_accuracy": 0.8652424812316895,
      "num_tokens": 867851009.0,
      "step": 27192
    },
    {
      "entropy": 0.5386719331145287,
      "epoch": 2.5018947562118674,
      "grad_norm": 0.17016077041625977,
      "learning_rate": 1.660686355690496e-05,
      "loss": 0.516,
      "mean_token_accuracy": 0.8372123129665852,
      "num_tokens": 867882793.0,
      "step": 27193
    },
    {
      "entropy": 0.5725612607784569,
      "epoch": 2.501986762583309,
      "grad_norm": 0.17113164067268372,
      "learning_rate": 1.6603796730763334e-05,
      "loss": 0.568,
      "mean_token_accuracy": 0.8227527812123299,
      "num_tokens": 867915359.0,
      "step": 27194
    },
    {
      "entropy": 0.5540238628163934,
      "epoch": 2.50207876895475,
      "grad_norm": 0.16032251715660095,
      "learning_rate": 1.660072990462171e-05,
      "loss": 0.5463,
      "mean_token_accuracy": 0.832856759428978,
      "num_tokens": 867947608.0,
      "step": 27195
    },
    {
      "entropy": 0.4960636291652918,
      "epoch": 2.5021707753261913,
      "grad_norm": 0.15690167248249054,
      "learning_rate": 1.6597663078480083e-05,
      "loss": 0.483,
      "mean_token_accuracy": 0.8452126011252403,
      "num_tokens": 867980024.0,
      "step": 27196
    },
    {
      "entropy": 0.515373133122921,
      "epoch": 2.502262781697633,
      "grad_norm": 0.1545892357826233,
      "learning_rate": 1.6594596252338458e-05,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8421556502580643,
      "num_tokens": 868012401.0,
      "step": 27197
    },
    {
      "entropy": 0.4873743765056133,
      "epoch": 2.5023547880690735,
      "grad_norm": 0.1558798998594284,
      "learning_rate": 1.659152942619683e-05,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8473915718495846,
      "num_tokens": 868044877.0,
      "step": 27198
    },
    {
      "entropy": 0.49649125477299094,
      "epoch": 2.502446794440515,
      "grad_norm": 0.1572885364294052,
      "learning_rate": 1.6588462600055204e-05,
      "loss": 0.4871,
      "mean_token_accuracy": 0.843066856265068,
      "num_tokens": 868077105.0,
      "step": 27199
    },
    {
      "entropy": 0.469110275618732,
      "epoch": 2.5025388008119562,
      "grad_norm": 0.15132303535938263,
      "learning_rate": 1.658539577391358e-05,
      "loss": 0.4587,
      "mean_token_accuracy": 0.8545210473239422,
      "num_tokens": 868109477.0,
      "step": 27200
    },
    {
      "entropy": 0.610972736030817,
      "epoch": 2.5026308071833974,
      "grad_norm": 0.17030414938926697,
      "learning_rate": 1.6582328947771953e-05,
      "loss": 0.6038,
      "mean_token_accuracy": 0.810700461268425,
      "num_tokens": 868141254.0,
      "step": 27201
    },
    {
      "entropy": 0.5912487674504519,
      "epoch": 2.502722813554839,
      "grad_norm": 0.17055192589759827,
      "learning_rate": 1.6579262121630325e-05,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8194355219602585,
      "num_tokens": 868173288.0,
      "step": 27202
    },
    {
      "entropy": 0.44947826862335205,
      "epoch": 2.5028148199262796,
      "grad_norm": 0.1527615338563919,
      "learning_rate": 1.65761952954887e-05,
      "loss": 0.4338,
      "mean_token_accuracy": 0.864036463201046,
      "num_tokens": 868205430.0,
      "step": 27203
    },
    {
      "entropy": 0.633103309199214,
      "epoch": 2.502906826297721,
      "grad_norm": 0.1740143597126007,
      "learning_rate": 1.6573128469347074e-05,
      "loss": 0.6225,
      "mean_token_accuracy": 0.8051256500184536,
      "num_tokens": 868236128.0,
      "step": 27204
    },
    {
      "entropy": 0.5247269282117486,
      "epoch": 2.5029988326691623,
      "grad_norm": 0.16000588238239288,
      "learning_rate": 1.657006164320545e-05,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8390000909566879,
      "num_tokens": 868267631.0,
      "step": 27205
    },
    {
      "entropy": 0.4466683503706008,
      "epoch": 2.5030908390406035,
      "grad_norm": 0.14977194368839264,
      "learning_rate": 1.6566994817063823e-05,
      "loss": 0.4336,
      "mean_token_accuracy": 0.8590404614806175,
      "num_tokens": 868299480.0,
      "step": 27206
    },
    {
      "entropy": 0.5427457811310887,
      "epoch": 2.503182845412045,
      "grad_norm": 0.1676873415708542,
      "learning_rate": 1.6563927990922194e-05,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8305167518556118,
      "num_tokens": 868331563.0,
      "step": 27207
    },
    {
      "entropy": 0.5999596510082483,
      "epoch": 2.5032748517834857,
      "grad_norm": 0.17381757497787476,
      "learning_rate": 1.656086116478057e-05,
      "loss": 0.5997,
      "mean_token_accuracy": 0.8114734478294849,
      "num_tokens": 868362345.0,
      "step": 27208
    },
    {
      "entropy": 0.44535367749631405,
      "epoch": 2.5033668581549273,
      "grad_norm": 0.15063001215457916,
      "learning_rate": 1.6557794338638944e-05,
      "loss": 0.431,
      "mean_token_accuracy": 0.8647404089570045,
      "num_tokens": 868393847.0,
      "step": 27209
    },
    {
      "entropy": 0.42902555502951145,
      "epoch": 2.5034588645263685,
      "grad_norm": 0.15312828123569489,
      "learning_rate": 1.655472751249732e-05,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8670597113668919,
      "num_tokens": 868426420.0,
      "step": 27210
    },
    {
      "entropy": 0.6062885411083698,
      "epoch": 2.5035508708978096,
      "grad_norm": 0.1725195199251175,
      "learning_rate": 1.655166068635569e-05,
      "loss": 0.5962,
      "mean_token_accuracy": 0.811867605894804,
      "num_tokens": 868458258.0,
      "step": 27211
    },
    {
      "entropy": 0.524228603579104,
      "epoch": 2.503642877269251,
      "grad_norm": 0.1595715880393982,
      "learning_rate": 1.6548593860214064e-05,
      "loss": 0.5258,
      "mean_token_accuracy": 0.835303008556366,
      "num_tokens": 868490347.0,
      "step": 27212
    },
    {
      "entropy": 0.5637189541012049,
      "epoch": 2.503734883640692,
      "grad_norm": 0.16344305872917175,
      "learning_rate": 1.654552703407244e-05,
      "loss": 0.5545,
      "mean_token_accuracy": 0.8230182603001595,
      "num_tokens": 868522404.0,
      "step": 27213
    },
    {
      "entropy": 0.6300210282206535,
      "epoch": 2.5038268900121334,
      "grad_norm": 0.17643220722675323,
      "learning_rate": 1.6542460207930814e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8084288015961647,
      "num_tokens": 868553683.0,
      "step": 27214
    },
    {
      "entropy": 0.44790993025526404,
      "epoch": 2.5039188963835746,
      "grad_norm": 0.14721418917179108,
      "learning_rate": 1.6539393381789188e-05,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8625769950449467,
      "num_tokens": 868585809.0,
      "step": 27215
    },
    {
      "entropy": 0.43055208306759596,
      "epoch": 2.5040109027550157,
      "grad_norm": 0.14952488243579865,
      "learning_rate": 1.653632655564756e-05,
      "loss": 0.4195,
      "mean_token_accuracy": 0.8647691980004311,
      "num_tokens": 868617795.0,
      "step": 27216
    },
    {
      "entropy": 0.5518979295156896,
      "epoch": 2.5041029091264573,
      "grad_norm": 0.16420741379261017,
      "learning_rate": 1.6533259729505934e-05,
      "loss": 0.5419,
      "mean_token_accuracy": 0.8314022682607174,
      "num_tokens": 868649991.0,
      "step": 27217
    },
    {
      "entropy": 0.48623468540608883,
      "epoch": 2.504194915497898,
      "grad_norm": 0.15891915559768677,
      "learning_rate": 1.653019290336431e-05,
      "loss": 0.4758,
      "mean_token_accuracy": 0.8462763763964176,
      "num_tokens": 868680926.0,
      "step": 27218
    },
    {
      "entropy": 0.46594120655208826,
      "epoch": 2.5042869218693395,
      "grad_norm": 0.15907537937164307,
      "learning_rate": 1.6527126077222683e-05,
      "loss": 0.4656,
      "mean_token_accuracy": 0.8507650531828403,
      "num_tokens": 868713089.0,
      "step": 27219
    },
    {
      "entropy": 0.5162785425782204,
      "epoch": 2.5043789282407807,
      "grad_norm": 0.15826886892318726,
      "learning_rate": 1.6524059251081055e-05,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8383590616285801,
      "num_tokens": 868744771.0,
      "step": 27220
    },
    {
      "entropy": 0.46331943944096565,
      "epoch": 2.504470934612222,
      "grad_norm": 0.15549786388874054,
      "learning_rate": 1.652099242493943e-05,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8538318388164043,
      "num_tokens": 868776332.0,
      "step": 27221
    },
    {
      "entropy": 0.5150128616951406,
      "epoch": 2.5045629409836634,
      "grad_norm": 0.16042156517505646,
      "learning_rate": 1.6517925598797804e-05,
      "loss": 0.5058,
      "mean_token_accuracy": 0.8401549682021141,
      "num_tokens": 868808532.0,
      "step": 27222
    },
    {
      "entropy": 0.5596203710883856,
      "epoch": 2.504654947355104,
      "grad_norm": 0.16918465495109558,
      "learning_rate": 1.651485877265618e-05,
      "loss": 0.5503,
      "mean_token_accuracy": 0.8233125805854797,
      "num_tokens": 868839503.0,
      "step": 27223
    },
    {
      "entropy": 0.4826796390116215,
      "epoch": 2.5047469537265457,
      "grad_norm": 0.1537577360868454,
      "learning_rate": 1.651179194651455e-05,
      "loss": 0.4648,
      "mean_token_accuracy": 0.8512575067579746,
      "num_tokens": 868871620.0,
      "step": 27224
    },
    {
      "entropy": 0.570859344676137,
      "epoch": 2.504838960097987,
      "grad_norm": 0.16557712852954865,
      "learning_rate": 1.6508725120372928e-05,
      "loss": 0.5662,
      "mean_token_accuracy": 0.8224595747888088,
      "num_tokens": 868903827.0,
      "step": 27225
    },
    {
      "entropy": 0.53622901532799,
      "epoch": 2.504930966469428,
      "grad_norm": 0.16242194175720215,
      "learning_rate": 1.6505658294231303e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8345791772007942,
      "num_tokens": 868936053.0,
      "step": 27226
    },
    {
      "entropy": 0.4236547858454287,
      "epoch": 2.5050229728408695,
      "grad_norm": 0.14361628890037537,
      "learning_rate": 1.6502591468089677e-05,
      "loss": 0.4055,
      "mean_token_accuracy": 0.870853029191494,
      "num_tokens": 868967476.0,
      "step": 27227
    },
    {
      "entropy": 0.5694685354828835,
      "epoch": 2.50511497921231,
      "grad_norm": 0.1625366508960724,
      "learning_rate": 1.649952464194805e-05,
      "loss": 0.5618,
      "mean_token_accuracy": 0.8238913267850876,
      "num_tokens": 868999795.0,
      "step": 27228
    },
    {
      "entropy": 0.47845237189903855,
      "epoch": 2.5052069855837518,
      "grad_norm": 0.15244247019290924,
      "learning_rate": 1.6496457815806423e-05,
      "loss": 0.4603,
      "mean_token_accuracy": 0.8522653952240944,
      "num_tokens": 869031402.0,
      "step": 27229
    },
    {
      "entropy": 0.5360496900975704,
      "epoch": 2.505298991955193,
      "grad_norm": 0.16595807671546936,
      "learning_rate": 1.6493390989664798e-05,
      "loss": 0.5311,
      "mean_token_accuracy": 0.8306511901319027,
      "num_tokens": 869062869.0,
      "step": 27230
    },
    {
      "entropy": 0.48203565925359726,
      "epoch": 2.505390998326634,
      "grad_norm": 0.1620766669511795,
      "learning_rate": 1.6490324163523172e-05,
      "loss": 0.4582,
      "mean_token_accuracy": 0.8506702184677124,
      "num_tokens": 869093867.0,
      "step": 27231
    },
    {
      "entropy": 0.5465827472507954,
      "epoch": 2.5054830046980756,
      "grad_norm": 0.16295498609542847,
      "learning_rate": 1.6487257337381544e-05,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8282242268323898,
      "num_tokens": 869126635.0,
      "step": 27232
    },
    {
      "entropy": 0.5694139888510108,
      "epoch": 2.5055750110695163,
      "grad_norm": 0.16509808599948883,
      "learning_rate": 1.648419051123992e-05,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8253688327968121,
      "num_tokens": 869158504.0,
      "step": 27233
    },
    {
      "entropy": 0.4623823086731136,
      "epoch": 2.505667017440958,
      "grad_norm": 0.15618398785591125,
      "learning_rate": 1.6481123685098293e-05,
      "loss": 0.4527,
      "mean_token_accuracy": 0.8537643998861313,
      "num_tokens": 869190464.0,
      "step": 27234
    },
    {
      "entropy": 0.6014164164662361,
      "epoch": 2.505759023812399,
      "grad_norm": 0.17194025218486786,
      "learning_rate": 1.6478056858956668e-05,
      "loss": 0.6041,
      "mean_token_accuracy": 0.8092698715627193,
      "num_tokens": 869222764.0,
      "step": 27235
    },
    {
      "entropy": 0.5530129047110677,
      "epoch": 2.50585103018384,
      "grad_norm": 0.16558867692947388,
      "learning_rate": 1.6474990032815042e-05,
      "loss": 0.5487,
      "mean_token_accuracy": 0.8314695172011852,
      "num_tokens": 869254196.0,
      "step": 27236
    },
    {
      "entropy": 0.5361601561307907,
      "epoch": 2.5059430365552817,
      "grad_norm": 0.1612609177827835,
      "learning_rate": 1.6471923206673414e-05,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8370138667523861,
      "num_tokens": 869286686.0,
      "step": 27237
    },
    {
      "entropy": 0.5449646092019975,
      "epoch": 2.5060350429267224,
      "grad_norm": 0.16229864954948425,
      "learning_rate": 1.6468856380531788e-05,
      "loss": 0.5363,
      "mean_token_accuracy": 0.8307640440762043,
      "num_tokens": 869319071.0,
      "step": 27238
    },
    {
      "entropy": 0.5302569391205907,
      "epoch": 2.506127049298164,
      "grad_norm": 0.15917547047138214,
      "learning_rate": 1.6465789554390163e-05,
      "loss": 0.5195,
      "mean_token_accuracy": 0.837843980640173,
      "num_tokens": 869350895.0,
      "step": 27239
    },
    {
      "entropy": 0.4770815437659621,
      "epoch": 2.506219055669605,
      "grad_norm": 0.15550585091114044,
      "learning_rate": 1.6462722728248537e-05,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8537984862923622,
      "num_tokens": 869383306.0,
      "step": 27240
    },
    {
      "entropy": 0.5791897168383002,
      "epoch": 2.5063110620410463,
      "grad_norm": 0.16181965172290802,
      "learning_rate": 1.645965590210691e-05,
      "loss": 0.5477,
      "mean_token_accuracy": 0.8256833031773567,
      "num_tokens": 869414946.0,
      "step": 27241
    },
    {
      "entropy": 0.563239187002182,
      "epoch": 2.506403068412488,
      "grad_norm": 0.16495740413665771,
      "learning_rate": 1.6456589075965283e-05,
      "loss": 0.5541,
      "mean_token_accuracy": 0.8281440958380699,
      "num_tokens": 869447106.0,
      "step": 27242
    },
    {
      "entropy": 0.5943867992609739,
      "epoch": 2.5064950747839285,
      "grad_norm": 0.166884183883667,
      "learning_rate": 1.6453522249823658e-05,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8186848387122154,
      "num_tokens": 869479610.0,
      "step": 27243
    },
    {
      "entropy": 0.4643085291609168,
      "epoch": 2.50658708115537,
      "grad_norm": 0.1582583487033844,
      "learning_rate": 1.6450455423682033e-05,
      "loss": 0.455,
      "mean_token_accuracy": 0.8573878891766071,
      "num_tokens": 869510676.0,
      "step": 27244
    },
    {
      "entropy": 0.5262798070907593,
      "epoch": 2.5066790875268112,
      "grad_norm": 0.16078925132751465,
      "learning_rate": 1.6447388597540407e-05,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8357656300067902,
      "num_tokens": 869542736.0,
      "step": 27245
    },
    {
      "entropy": 0.5114972814917564,
      "epoch": 2.5067710938982524,
      "grad_norm": 0.15387606620788574,
      "learning_rate": 1.644432177139878e-05,
      "loss": 0.5038,
      "mean_token_accuracy": 0.8400791697204113,
      "num_tokens": 869575174.0,
      "step": 27246
    },
    {
      "entropy": 0.51633754465729,
      "epoch": 2.506863100269694,
      "grad_norm": 0.16329076886177063,
      "learning_rate": 1.6441254945257153e-05,
      "loss": 0.4944,
      "mean_token_accuracy": 0.8409492634236813,
      "num_tokens": 869606458.0,
      "step": 27247
    },
    {
      "entropy": 0.6231646183878183,
      "epoch": 2.5069551066411346,
      "grad_norm": 0.17134925723075867,
      "learning_rate": 1.6438188119115528e-05,
      "loss": 0.6035,
      "mean_token_accuracy": 0.8139513358473778,
      "num_tokens": 869638535.0,
      "step": 27248
    },
    {
      "entropy": 0.5380160007625818,
      "epoch": 2.507047113012576,
      "grad_norm": 0.15896500647068024,
      "learning_rate": 1.6435121292973903e-05,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8378744833171368,
      "num_tokens": 869670940.0,
      "step": 27249
    },
    {
      "entropy": 0.5141032373066992,
      "epoch": 2.5071391193840173,
      "grad_norm": 0.15968142449855804,
      "learning_rate": 1.6432054466832274e-05,
      "loss": 0.4951,
      "mean_token_accuracy": 0.8421454094350338,
      "num_tokens": 869702992.0,
      "step": 27250
    },
    {
      "entropy": 0.5119600370526314,
      "epoch": 2.5072311257554585,
      "grad_norm": 0.16182978451251984,
      "learning_rate": 1.642898764069065e-05,
      "loss": 0.5018,
      "mean_token_accuracy": 0.8395995572209358,
      "num_tokens": 869734743.0,
      "step": 27251
    },
    {
      "entropy": 0.43412564881145954,
      "epoch": 2.5073231321269,
      "grad_norm": 0.15559828281402588,
      "learning_rate": 1.6425920814549023e-05,
      "loss": 0.4231,
      "mean_token_accuracy": 0.8642477840185165,
      "num_tokens": 869767100.0,
      "step": 27252
    },
    {
      "entropy": 0.5653922986239195,
      "epoch": 2.5074151384983407,
      "grad_norm": 0.16697345674037933,
      "learning_rate": 1.6422853988407398e-05,
      "loss": 0.5614,
      "mean_token_accuracy": 0.8200481869280338,
      "num_tokens": 869798457.0,
      "step": 27253
    },
    {
      "entropy": 0.5355486460030079,
      "epoch": 2.5075071448697823,
      "grad_norm": 0.1650409698486328,
      "learning_rate": 1.641978716226577e-05,
      "loss": 0.5262,
      "mean_token_accuracy": 0.8339042514562607,
      "num_tokens": 869830280.0,
      "step": 27254
    },
    {
      "entropy": 0.48371898382902145,
      "epoch": 2.5075991512412235,
      "grad_norm": 0.15929394960403442,
      "learning_rate": 1.6416720336124147e-05,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8517575711011887,
      "num_tokens": 869861152.0,
      "step": 27255
    },
    {
      "entropy": 0.43803169298917055,
      "epoch": 2.5076911576126646,
      "grad_norm": 0.1540970504283905,
      "learning_rate": 1.641365350998252e-05,
      "loss": 0.4373,
      "mean_token_accuracy": 0.8609082922339439,
      "num_tokens": 869893303.0,
      "step": 27256
    },
    {
      "entropy": 0.5019344985485077,
      "epoch": 2.507783163984106,
      "grad_norm": 0.16099002957344055,
      "learning_rate": 1.6410586683840896e-05,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8427834287285805,
      "num_tokens": 869925567.0,
      "step": 27257
    },
    {
      "entropy": 0.5677878288552165,
      "epoch": 2.507875170355547,
      "grad_norm": 0.1632666140794754,
      "learning_rate": 1.6407519857699268e-05,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8241563700139523,
      "num_tokens": 869957030.0,
      "step": 27258
    },
    {
      "entropy": 0.44450607150793076,
      "epoch": 2.5079671767269884,
      "grad_norm": 0.15050168335437775,
      "learning_rate": 1.6404453031557642e-05,
      "loss": 0.437,
      "mean_token_accuracy": 0.8614283911883831,
      "num_tokens": 869989374.0,
      "step": 27259
    },
    {
      "entropy": 0.4490528148598969,
      "epoch": 2.5080591830984296,
      "grad_norm": 0.15214599668979645,
      "learning_rate": 1.6401386205416017e-05,
      "loss": 0.435,
      "mean_token_accuracy": 0.8603278808295727,
      "num_tokens": 870021518.0,
      "step": 27260
    },
    {
      "entropy": 0.5749696269631386,
      "epoch": 2.5081511894698707,
      "grad_norm": 0.16733598709106445,
      "learning_rate": 1.639831937927439e-05,
      "loss": 0.5698,
      "mean_token_accuracy": 0.8210006766021252,
      "num_tokens": 870054181.0,
      "step": 27261
    },
    {
      "entropy": 0.4882448147982359,
      "epoch": 2.5082431958413123,
      "grad_norm": 0.1556159406900406,
      "learning_rate": 1.6395252553132763e-05,
      "loss": 0.469,
      "mean_token_accuracy": 0.8482438959181309,
      "num_tokens": 870086463.0,
      "step": 27262
    },
    {
      "entropy": 0.5256573054939508,
      "epoch": 2.508335202212753,
      "grad_norm": 0.1626521348953247,
      "learning_rate": 1.6392185726991137e-05,
      "loss": 0.5119,
      "mean_token_accuracy": 0.8378248624503613,
      "num_tokens": 870119038.0,
      "step": 27263
    },
    {
      "entropy": 0.4960642894729972,
      "epoch": 2.5084272085841945,
      "grad_norm": 0.1582111120223999,
      "learning_rate": 1.6389118900849512e-05,
      "loss": 0.4835,
      "mean_token_accuracy": 0.8484436795115471,
      "num_tokens": 870150869.0,
      "step": 27264
    },
    {
      "entropy": 0.5658022165298462,
      "epoch": 2.5085192149556357,
      "grad_norm": 0.16643761098384857,
      "learning_rate": 1.6386052074707887e-05,
      "loss": 0.5544,
      "mean_token_accuracy": 0.821894109249115,
      "num_tokens": 870182242.0,
      "step": 27265
    },
    {
      "entropy": 0.5016528405249119,
      "epoch": 2.508611221327077,
      "grad_norm": 0.1582150012254715,
      "learning_rate": 1.638298524856626e-05,
      "loss": 0.492,
      "mean_token_accuracy": 0.8421121686697006,
      "num_tokens": 870214452.0,
      "step": 27266
    },
    {
      "entropy": 0.5860534822568297,
      "epoch": 2.5087032276985184,
      "grad_norm": 0.17070937156677246,
      "learning_rate": 1.6379918422424633e-05,
      "loss": 0.5765,
      "mean_token_accuracy": 0.822305653244257,
      "num_tokens": 870245880.0,
      "step": 27267
    },
    {
      "entropy": 0.45323833264410496,
      "epoch": 2.508795234069959,
      "grad_norm": 0.14953738451004028,
      "learning_rate": 1.6376851596283007e-05,
      "loss": 0.4381,
      "mean_token_accuracy": 0.8581569269299507,
      "num_tokens": 870278542.0,
      "step": 27268
    },
    {
      "entropy": 0.4557606466114521,
      "epoch": 2.5088872404414007,
      "grad_norm": 0.15327931940555573,
      "learning_rate": 1.6373784770141382e-05,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8580162227153778,
      "num_tokens": 870310725.0,
      "step": 27269
    },
    {
      "entropy": 0.44891584501601756,
      "epoch": 2.508979246812842,
      "grad_norm": 0.14626148343086243,
      "learning_rate": 1.6370717943999757e-05,
      "loss": 0.4282,
      "mean_token_accuracy": 0.865389995276928,
      "num_tokens": 870342594.0,
      "step": 27270
    },
    {
      "entropy": 0.543222757987678,
      "epoch": 2.509071253184283,
      "grad_norm": 0.16039639711380005,
      "learning_rate": 1.6367651117858128e-05,
      "loss": 0.5262,
      "mean_token_accuracy": 0.8306580260396004,
      "num_tokens": 870374532.0,
      "step": 27271
    },
    {
      "entropy": 0.5473461616784334,
      "epoch": 2.5091632595557245,
      "grad_norm": 0.1682865023612976,
      "learning_rate": 1.6364584291716502e-05,
      "loss": 0.5366,
      "mean_token_accuracy": 0.830856192857027,
      "num_tokens": 870406532.0,
      "step": 27272
    },
    {
      "entropy": 0.5398308141157031,
      "epoch": 2.509255265927165,
      "grad_norm": 0.16247177124023438,
      "learning_rate": 1.6361517465574877e-05,
      "loss": 0.528,
      "mean_token_accuracy": 0.8351245671510696,
      "num_tokens": 870439300.0,
      "step": 27273
    },
    {
      "entropy": 0.4691578187048435,
      "epoch": 2.5093472722986068,
      "grad_norm": 0.15207727253437042,
      "learning_rate": 1.6358450639433252e-05,
      "loss": 0.4491,
      "mean_token_accuracy": 0.8548619747161865,
      "num_tokens": 870471139.0,
      "step": 27274
    },
    {
      "entropy": 0.502510279417038,
      "epoch": 2.509439278670048,
      "grad_norm": 0.16719263792037964,
      "learning_rate": 1.6355383813291626e-05,
      "loss": 0.4974,
      "mean_token_accuracy": 0.8413188494741917,
      "num_tokens": 870502380.0,
      "step": 27275
    },
    {
      "entropy": 0.4761591460555792,
      "epoch": 2.509531285041489,
      "grad_norm": 0.15221677720546722,
      "learning_rate": 1.6352316987149998e-05,
      "loss": 0.4623,
      "mean_token_accuracy": 0.8524927571415901,
      "num_tokens": 870534712.0,
      "step": 27276
    },
    {
      "entropy": 0.5524165194947273,
      "epoch": 2.5096232914129306,
      "grad_norm": 0.16413874924182892,
      "learning_rate": 1.6349250161008372e-05,
      "loss": 0.542,
      "mean_token_accuracy": 0.8293341808021069,
      "num_tokens": 870565918.0,
      "step": 27277
    },
    {
      "entropy": 0.4848499083891511,
      "epoch": 2.5097152977843713,
      "grad_norm": 0.1566833108663559,
      "learning_rate": 1.6346183334866747e-05,
      "loss": 0.4739,
      "mean_token_accuracy": 0.8458973504602909,
      "num_tokens": 870598113.0,
      "step": 27278
    },
    {
      "entropy": 0.532508990727365,
      "epoch": 2.509807304155813,
      "grad_norm": 0.16592878103256226,
      "learning_rate": 1.634311650872512e-05,
      "loss": 0.5216,
      "mean_token_accuracy": 0.8354321829974651,
      "num_tokens": 870630216.0,
      "step": 27279
    },
    {
      "entropy": 0.5699373371899128,
      "epoch": 2.509899310527254,
      "grad_norm": 0.16476163268089294,
      "learning_rate": 1.6340049682583493e-05,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8167874924838543,
      "num_tokens": 870662467.0,
      "step": 27280
    },
    {
      "entropy": 0.5184783507138491,
      "epoch": 2.509991316898695,
      "grad_norm": 0.15912941098213196,
      "learning_rate": 1.6336982856441868e-05,
      "loss": 0.514,
      "mean_token_accuracy": 0.8378495499491692,
      "num_tokens": 870694839.0,
      "step": 27281
    },
    {
      "entropy": 0.5815288871526718,
      "epoch": 2.5100833232701367,
      "grad_norm": 0.16371795535087585,
      "learning_rate": 1.6333916030300242e-05,
      "loss": 0.5671,
      "mean_token_accuracy": 0.8213624358177185,
      "num_tokens": 870727410.0,
      "step": 27282
    },
    {
      "entropy": 0.49952840618789196,
      "epoch": 2.5101753296415774,
      "grad_norm": 0.16089484095573425,
      "learning_rate": 1.6330849204158617e-05,
      "loss": 0.4988,
      "mean_token_accuracy": 0.8431074358522892,
      "num_tokens": 870759788.0,
      "step": 27283
    },
    {
      "entropy": 0.41906084935180843,
      "epoch": 2.510267336013019,
      "grad_norm": 0.15122567117214203,
      "learning_rate": 1.6327782378016988e-05,
      "loss": 0.4135,
      "mean_token_accuracy": 0.8683419860899448,
      "num_tokens": 870791293.0,
      "step": 27284
    },
    {
      "entropy": 0.5078394776210189,
      "epoch": 2.51035934238446,
      "grad_norm": 0.16225160658359528,
      "learning_rate": 1.6324715551875363e-05,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8425630442798138,
      "num_tokens": 870823866.0,
      "step": 27285
    },
    {
      "entropy": 0.548699707724154,
      "epoch": 2.5104513487559013,
      "grad_norm": 0.1592131406068802,
      "learning_rate": 1.632164872573374e-05,
      "loss": 0.5425,
      "mean_token_accuracy": 0.8277553543448448,
      "num_tokens": 870856389.0,
      "step": 27286
    },
    {
      "entropy": 0.5389552935957909,
      "epoch": 2.510543355127343,
      "grad_norm": 0.1606113612651825,
      "learning_rate": 1.6318581899592115e-05,
      "loss": 0.5306,
      "mean_token_accuracy": 0.8366191610693932,
      "num_tokens": 870888845.0,
      "step": 27287
    },
    {
      "entropy": 0.5620732298120856,
      "epoch": 2.5106353614987835,
      "grad_norm": 0.16411571204662323,
      "learning_rate": 1.6315515073450487e-05,
      "loss": 0.549,
      "mean_token_accuracy": 0.8269773460924625,
      "num_tokens": 870920577.0,
      "step": 27288
    },
    {
      "entropy": 0.6551814991980791,
      "epoch": 2.510727367870225,
      "grad_norm": 0.17015638947486877,
      "learning_rate": 1.631244824730886e-05,
      "loss": 0.6421,
      "mean_token_accuracy": 0.7971304878592491,
      "num_tokens": 870952785.0,
      "step": 27289
    },
    {
      "entropy": 0.5268676816485822,
      "epoch": 2.5108193742416662,
      "grad_norm": 0.16712723672389984,
      "learning_rate": 1.6309381421167236e-05,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8340361677110195,
      "num_tokens": 870985016.0,
      "step": 27290
    },
    {
      "entropy": 0.4494405922014266,
      "epoch": 2.5109113806131074,
      "grad_norm": 0.15136274695396423,
      "learning_rate": 1.630631459502561e-05,
      "loss": 0.4338,
      "mean_token_accuracy": 0.863076277077198,
      "num_tokens": 871016565.0,
      "step": 27291
    },
    {
      "entropy": 0.5559796467423439,
      "epoch": 2.511003386984549,
      "grad_norm": 0.16534116864204407,
      "learning_rate": 1.6303247768883982e-05,
      "loss": 0.5271,
      "mean_token_accuracy": 0.8352772295475006,
      "num_tokens": 871049333.0,
      "step": 27292
    },
    {
      "entropy": 0.46432723570615053,
      "epoch": 2.5110953933559896,
      "grad_norm": 0.15152141451835632,
      "learning_rate": 1.6300180942742357e-05,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8527020923793316,
      "num_tokens": 871081783.0,
      "step": 27293
    },
    {
      "entropy": 0.5716440845280886,
      "epoch": 2.511187399727431,
      "grad_norm": 0.1655314564704895,
      "learning_rate": 1.629711411660073e-05,
      "loss": 0.5587,
      "mean_token_accuracy": 0.8270396552979946,
      "num_tokens": 871114124.0,
      "step": 27294
    },
    {
      "entropy": 0.5733606070280075,
      "epoch": 2.5112794060988723,
      "grad_norm": 0.1718057245016098,
      "learning_rate": 1.6294047290459106e-05,
      "loss": 0.579,
      "mean_token_accuracy": 0.8167405761778355,
      "num_tokens": 871145649.0,
      "step": 27295
    },
    {
      "entropy": 0.5469632185995579,
      "epoch": 2.5113714124703135,
      "grad_norm": 0.16190919280052185,
      "learning_rate": 1.629098046431748e-05,
      "loss": 0.5395,
      "mean_token_accuracy": 0.8282879665493965,
      "num_tokens": 871178115.0,
      "step": 27296
    },
    {
      "entropy": 0.5115541024133563,
      "epoch": 2.511463418841755,
      "grad_norm": 0.1609124094247818,
      "learning_rate": 1.6287913638175852e-05,
      "loss": 0.4961,
      "mean_token_accuracy": 0.8440234288573265,
      "num_tokens": 871210668.0,
      "step": 27297
    },
    {
      "entropy": 0.5065490796696395,
      "epoch": 2.5115554252131957,
      "grad_norm": 0.16211511194705963,
      "learning_rate": 1.6284846812034226e-05,
      "loss": 0.5011,
      "mean_token_accuracy": 0.8393410183489323,
      "num_tokens": 871242507.0,
      "step": 27298
    },
    {
      "entropy": 0.5115646347403526,
      "epoch": 2.5116474315846373,
      "grad_norm": 0.16170728206634521,
      "learning_rate": 1.62817799858926e-05,
      "loss": 0.4912,
      "mean_token_accuracy": 0.8399313613772392,
      "num_tokens": 871273925.0,
      "step": 27299
    },
    {
      "entropy": 0.42442187294363976,
      "epoch": 2.5117394379560785,
      "grad_norm": 0.14864636957645416,
      "learning_rate": 1.6278713159750976e-05,
      "loss": 0.4153,
      "mean_token_accuracy": 0.8681810274720192,
      "num_tokens": 871305745.0,
      "step": 27300
    },
    {
      "entropy": 0.516277099493891,
      "epoch": 2.5118314443275196,
      "grad_norm": 0.15707072615623474,
      "learning_rate": 1.6275646333609347e-05,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8410394042730331,
      "num_tokens": 871337783.0,
      "step": 27301
    },
    {
      "entropy": 0.5621011666953564,
      "epoch": 2.511923450698961,
      "grad_norm": 0.1666930615901947,
      "learning_rate": 1.627257950746772e-05,
      "loss": 0.5534,
      "mean_token_accuracy": 0.8272883035242558,
      "num_tokens": 871369163.0,
      "step": 27302
    },
    {
      "entropy": 0.6234706649556756,
      "epoch": 2.512015457070402,
      "grad_norm": 0.1702185869216919,
      "learning_rate": 1.6269512681326096e-05,
      "loss": 0.6183,
      "mean_token_accuracy": 0.8090721257030964,
      "num_tokens": 871401070.0,
      "step": 27303
    },
    {
      "entropy": 0.4289499609731138,
      "epoch": 2.5121074634418434,
      "grad_norm": 0.14284799993038177,
      "learning_rate": 1.626644585518447e-05,
      "loss": 0.4105,
      "mean_token_accuracy": 0.8695724159479141,
      "num_tokens": 871433481.0,
      "step": 27304
    },
    {
      "entropy": 0.6088764481246471,
      "epoch": 2.5121994698132846,
      "grad_norm": 0.1720663160085678,
      "learning_rate": 1.6263379029042846e-05,
      "loss": 0.5972,
      "mean_token_accuracy": 0.8124589063227177,
      "num_tokens": 871464148.0,
      "step": 27305
    },
    {
      "entropy": 0.4875169894658029,
      "epoch": 2.5122914761847257,
      "grad_norm": 0.15151870250701904,
      "learning_rate": 1.6260312202901217e-05,
      "loss": 0.464,
      "mean_token_accuracy": 0.8519046977162361,
      "num_tokens": 871496313.0,
      "step": 27306
    },
    {
      "entropy": 0.5110763711854815,
      "epoch": 2.5123834825561673,
      "grad_norm": 0.15804500877857208,
      "learning_rate": 1.625724537675959e-05,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8406185172498226,
      "num_tokens": 871528338.0,
      "step": 27307
    },
    {
      "entropy": 0.4311844832263887,
      "epoch": 2.512475488927608,
      "grad_norm": 0.14220060408115387,
      "learning_rate": 1.6254178550617966e-05,
      "loss": 0.4148,
      "mean_token_accuracy": 0.8680256754159927,
      "num_tokens": 871560952.0,
      "step": 27308
    },
    {
      "entropy": 0.5216938182711601,
      "epoch": 2.5125674952990495,
      "grad_norm": 0.16328288614749908,
      "learning_rate": 1.625111172447634e-05,
      "loss": 0.5202,
      "mean_token_accuracy": 0.8370471745729446,
      "num_tokens": 871593110.0,
      "step": 27309
    },
    {
      "entropy": 0.4894604757428169,
      "epoch": 2.5126595016704907,
      "grad_norm": 0.15125149488449097,
      "learning_rate": 1.6248044898334712e-05,
      "loss": 0.4785,
      "mean_token_accuracy": 0.8468519374728203,
      "num_tokens": 871624890.0,
      "step": 27310
    },
    {
      "entropy": 0.42757780384272337,
      "epoch": 2.512751508041932,
      "grad_norm": 0.1472858190536499,
      "learning_rate": 1.6244978072193087e-05,
      "loss": 0.4205,
      "mean_token_accuracy": 0.8637354001402855,
      "num_tokens": 871656741.0,
      "step": 27311
    },
    {
      "entropy": 0.5550139313563704,
      "epoch": 2.5128435144133734,
      "grad_norm": 0.16487830877304077,
      "learning_rate": 1.624191124605146e-05,
      "loss": 0.5489,
      "mean_token_accuracy": 0.8287834860384464,
      "num_tokens": 871688499.0,
      "step": 27312
    },
    {
      "entropy": 0.5586722590960562,
      "epoch": 2.512935520784814,
      "grad_norm": 0.16693705320358276,
      "learning_rate": 1.6238844419909836e-05,
      "loss": 0.5528,
      "mean_token_accuracy": 0.8246204480528831,
      "num_tokens": 871721095.0,
      "step": 27313
    },
    {
      "entropy": 0.559432290494442,
      "epoch": 2.5130275271562557,
      "grad_norm": 0.1637842357158661,
      "learning_rate": 1.6235777593768207e-05,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8273477479815483,
      "num_tokens": 871752031.0,
      "step": 27314
    },
    {
      "entropy": 0.5904718916863203,
      "epoch": 2.513119533527697,
      "grad_norm": 0.17185701429843903,
      "learning_rate": 1.6232710767626582e-05,
      "loss": 0.5816,
      "mean_token_accuracy": 0.8124543018639088,
      "num_tokens": 871783598.0,
      "step": 27315
    },
    {
      "entropy": 0.5569889564067125,
      "epoch": 2.513211539899138,
      "grad_norm": 0.16683340072631836,
      "learning_rate": 1.622964394148496e-05,
      "loss": 0.5486,
      "mean_token_accuracy": 0.8208978995680809,
      "num_tokens": 871815122.0,
      "step": 27316
    },
    {
      "entropy": 0.5561720747500658,
      "epoch": 2.5133035462705795,
      "grad_norm": 0.16686441004276276,
      "learning_rate": 1.6226577115343335e-05,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8236128687858582,
      "num_tokens": 871846936.0,
      "step": 27317
    },
    {
      "entropy": 0.5621066745370626,
      "epoch": 2.51339555264202,
      "grad_norm": 0.16629917919635773,
      "learning_rate": 1.6223510289201706e-05,
      "loss": 0.5489,
      "mean_token_accuracy": 0.8290707431733608,
      "num_tokens": 871878945.0,
      "step": 27318
    },
    {
      "entropy": 0.48783658538013697,
      "epoch": 2.5134875590134618,
      "grad_norm": 0.1563485860824585,
      "learning_rate": 1.622044346306008e-05,
      "loss": 0.4753,
      "mean_token_accuracy": 0.8465032987296581,
      "num_tokens": 871911031.0,
      "step": 27319
    },
    {
      "entropy": 0.5503398990258574,
      "epoch": 2.513579565384903,
      "grad_norm": 0.16624018549919128,
      "learning_rate": 1.6217376636918455e-05,
      "loss": 0.5416,
      "mean_token_accuracy": 0.8276874385774136,
      "num_tokens": 871942680.0,
      "step": 27320
    },
    {
      "entropy": 0.5188472752925009,
      "epoch": 2.513671571756344,
      "grad_norm": 0.16028647124767303,
      "learning_rate": 1.621430981077683e-05,
      "loss": 0.5105,
      "mean_token_accuracy": 0.8392045758664608,
      "num_tokens": 871974248.0,
      "step": 27321
    },
    {
      "entropy": 0.5068162400275469,
      "epoch": 2.5137635781277856,
      "grad_norm": 0.15573029220104218,
      "learning_rate": 1.62112429846352e-05,
      "loss": 0.4851,
      "mean_token_accuracy": 0.844241552054882,
      "num_tokens": 872006518.0,
      "step": 27322
    },
    {
      "entropy": 0.5400262624025345,
      "epoch": 2.5138555844992263,
      "grad_norm": 0.16604219377040863,
      "learning_rate": 1.6208176158493576e-05,
      "loss": 0.5222,
      "mean_token_accuracy": 0.8328912146389484,
      "num_tokens": 872038189.0,
      "step": 27323
    },
    {
      "entropy": 0.6170321833342314,
      "epoch": 2.513947590870668,
      "grad_norm": 0.16937682032585144,
      "learning_rate": 1.620510933235195e-05,
      "loss": 0.6012,
      "mean_token_accuracy": 0.8109753280878067,
      "num_tokens": 872070555.0,
      "step": 27324
    },
    {
      "entropy": 0.4336993810720742,
      "epoch": 2.514039597242109,
      "grad_norm": 0.15100562572479248,
      "learning_rate": 1.6202042506210325e-05,
      "loss": 0.4183,
      "mean_token_accuracy": 0.8655639477074146,
      "num_tokens": 872102475.0,
      "step": 27325
    },
    {
      "entropy": 0.5380606083199382,
      "epoch": 2.51413160361355,
      "grad_norm": 0.16286827623844147,
      "learning_rate": 1.61989756800687e-05,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8315814733505249,
      "num_tokens": 872133696.0,
      "step": 27326
    },
    {
      "entropy": 0.5224732924252748,
      "epoch": 2.5142236099849917,
      "grad_norm": 0.16622228920459747,
      "learning_rate": 1.619590885392707e-05,
      "loss": 0.5204,
      "mean_token_accuracy": 0.8371834047138691,
      "num_tokens": 872166053.0,
      "step": 27327
    },
    {
      "entropy": 0.4530913205817342,
      "epoch": 2.5143156163564324,
      "grad_norm": 0.15323829650878906,
      "learning_rate": 1.6192842027785445e-05,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8594505824148655,
      "num_tokens": 872198362.0,
      "step": 27328
    },
    {
      "entropy": 0.4774752839002758,
      "epoch": 2.514407622727874,
      "grad_norm": 0.1532161384820938,
      "learning_rate": 1.618977520164382e-05,
      "loss": 0.4566,
      "mean_token_accuracy": 0.852659847587347,
      "num_tokens": 872230319.0,
      "step": 27329
    },
    {
      "entropy": 0.5511252209544182,
      "epoch": 2.514499629099315,
      "grad_norm": 0.1621730923652649,
      "learning_rate": 1.6186708375502195e-05,
      "loss": 0.5446,
      "mean_token_accuracy": 0.8271098956465721,
      "num_tokens": 872262824.0,
      "step": 27330
    },
    {
      "entropy": 0.6171340271830559,
      "epoch": 2.5145916354707563,
      "grad_norm": 0.17404523491859436,
      "learning_rate": 1.6183641549360566e-05,
      "loss": 0.6037,
      "mean_token_accuracy": 0.8085016310214996,
      "num_tokens": 872294169.0,
      "step": 27331
    },
    {
      "entropy": 0.5040015261620283,
      "epoch": 2.514683641842198,
      "grad_norm": 0.15880663692951202,
      "learning_rate": 1.618057472321894e-05,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8426836095750332,
      "num_tokens": 872325596.0,
      "step": 27332
    },
    {
      "entropy": 0.3912375245708972,
      "epoch": 2.5147756482136385,
      "grad_norm": 0.1427338868379593,
      "learning_rate": 1.6177507897077315e-05,
      "loss": 0.3769,
      "mean_token_accuracy": 0.8761796616017818,
      "num_tokens": 872356550.0,
      "step": 27333
    },
    {
      "entropy": 0.5545766651630402,
      "epoch": 2.51486765458508,
      "grad_norm": 0.16699329018592834,
      "learning_rate": 1.617444107093569e-05,
      "loss": 0.55,
      "mean_token_accuracy": 0.8274734616279602,
      "num_tokens": 872388368.0,
      "step": 27334
    },
    {
      "entropy": 0.5144259417429566,
      "epoch": 2.5149596609565212,
      "grad_norm": 0.15882156789302826,
      "learning_rate": 1.6171374244794065e-05,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8359752297401428,
      "num_tokens": 872420279.0,
      "step": 27335
    },
    {
      "entropy": 0.47829891461879015,
      "epoch": 2.5150516673279624,
      "grad_norm": 0.15179988741874695,
      "learning_rate": 1.6168307418652436e-05,
      "loss": 0.4581,
      "mean_token_accuracy": 0.852763470262289,
      "num_tokens": 872452394.0,
      "step": 27336
    },
    {
      "entropy": 0.48446539393626153,
      "epoch": 2.515143673699404,
      "grad_norm": 0.15410739183425903,
      "learning_rate": 1.616524059251081e-05,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8511681035161018,
      "num_tokens": 872484390.0,
      "step": 27337
    },
    {
      "entropy": 0.6038760058581829,
      "epoch": 2.5152356800708446,
      "grad_norm": 0.1744147092103958,
      "learning_rate": 1.6162173766369185e-05,
      "loss": 0.5967,
      "mean_token_accuracy": 0.8121825009584427,
      "num_tokens": 872515556.0,
      "step": 27338
    },
    {
      "entropy": 0.5212838733568788,
      "epoch": 2.515327686442286,
      "grad_norm": 0.16570964455604553,
      "learning_rate": 1.615910694022756e-05,
      "loss": 0.5196,
      "mean_token_accuracy": 0.8344423696398735,
      "num_tokens": 872548042.0,
      "step": 27339
    },
    {
      "entropy": 0.4989027122501284,
      "epoch": 2.5154196928137273,
      "grad_norm": 0.1627015322446823,
      "learning_rate": 1.615604011408593e-05,
      "loss": 0.4928,
      "mean_token_accuracy": 0.8422199860215187,
      "num_tokens": 872579230.0,
      "step": 27340
    },
    {
      "entropy": 0.5882809516042471,
      "epoch": 2.5155116991851685,
      "grad_norm": 0.1685372292995453,
      "learning_rate": 1.6152973287944306e-05,
      "loss": 0.5843,
      "mean_token_accuracy": 0.8154678791761398,
      "num_tokens": 872611209.0,
      "step": 27341
    },
    {
      "entropy": 0.4851894788444042,
      "epoch": 2.51560370555661,
      "grad_norm": 0.15407022833824158,
      "learning_rate": 1.614990646180268e-05,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8493667654693127,
      "num_tokens": 872643610.0,
      "step": 27342
    },
    {
      "entropy": 0.6322556678205729,
      "epoch": 2.5156957119280507,
      "grad_norm": 0.16994355618953705,
      "learning_rate": 1.6146839635661055e-05,
      "loss": 0.619,
      "mean_token_accuracy": 0.8054135739803314,
      "num_tokens": 872676035.0,
      "step": 27343
    },
    {
      "entropy": 0.49711337708868086,
      "epoch": 2.5157877182994923,
      "grad_norm": 0.15733668208122253,
      "learning_rate": 1.6143772809519426e-05,
      "loss": 0.4786,
      "mean_token_accuracy": 0.8445518985390663,
      "num_tokens": 872708411.0,
      "step": 27344
    },
    {
      "entropy": 0.5649159047752619,
      "epoch": 2.5158797246709335,
      "grad_norm": 0.1627017855644226,
      "learning_rate": 1.61407059833778e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8270583786070347,
      "num_tokens": 872740056.0,
      "step": 27345
    },
    {
      "entropy": 0.5081013040617108,
      "epoch": 2.5159717310423746,
      "grad_norm": 0.1561659872531891,
      "learning_rate": 1.6137639157236176e-05,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8439699299633503,
      "num_tokens": 872771617.0,
      "step": 27346
    },
    {
      "entropy": 0.44329064060002565,
      "epoch": 2.516063737413816,
      "grad_norm": 0.15429925918579102,
      "learning_rate": 1.6134572331094554e-05,
      "loss": 0.431,
      "mean_token_accuracy": 0.8579886257648468,
      "num_tokens": 872802607.0,
      "step": 27347
    },
    {
      "entropy": 0.562725548632443,
      "epoch": 2.516155743785257,
      "grad_norm": 0.16414380073547363,
      "learning_rate": 1.6131505504952925e-05,
      "loss": 0.5582,
      "mean_token_accuracy": 0.8239606730639935,
      "num_tokens": 872834837.0,
      "step": 27348
    },
    {
      "entropy": 0.5327634224668145,
      "epoch": 2.5162477501566984,
      "grad_norm": 0.16787871718406677,
      "learning_rate": 1.61284386788113e-05,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8359359391033649,
      "num_tokens": 872866374.0,
      "step": 27349
    },
    {
      "entropy": 0.4403064213693142,
      "epoch": 2.5163397565281396,
      "grad_norm": 0.14938019216060638,
      "learning_rate": 1.6125371852669674e-05,
      "loss": 0.4196,
      "mean_token_accuracy": 0.8648174740374088,
      "num_tokens": 872897742.0,
      "step": 27350
    },
    {
      "entropy": 0.48852445650845766,
      "epoch": 2.5164317628995807,
      "grad_norm": 0.15427175164222717,
      "learning_rate": 1.612230502652805e-05,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8537684790790081,
      "num_tokens": 872929458.0,
      "step": 27351
    },
    {
      "entropy": 0.4400868574157357,
      "epoch": 2.5165237692710223,
      "grad_norm": 0.15252923965454102,
      "learning_rate": 1.611923820038642e-05,
      "loss": 0.4272,
      "mean_token_accuracy": 0.8647445812821388,
      "num_tokens": 872961347.0,
      "step": 27352
    },
    {
      "entropy": 0.553059303201735,
      "epoch": 2.516615775642463,
      "grad_norm": 0.16863089799880981,
      "learning_rate": 1.6116171374244795e-05,
      "loss": 0.5479,
      "mean_token_accuracy": 0.828367080539465,
      "num_tokens": 872993153.0,
      "step": 27353
    },
    {
      "entropy": 0.5393837792798877,
      "epoch": 2.5167077820139045,
      "grad_norm": 0.1609755903482437,
      "learning_rate": 1.611310454810317e-05,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8306908048689365,
      "num_tokens": 873025215.0,
      "step": 27354
    },
    {
      "entropy": 0.4486816180869937,
      "epoch": 2.5167997883853457,
      "grad_norm": 0.1500260978937149,
      "learning_rate": 1.6110037721961544e-05,
      "loss": 0.4233,
      "mean_token_accuracy": 0.8623133860528469,
      "num_tokens": 873057442.0,
      "step": 27355
    },
    {
      "entropy": 0.49229345843195915,
      "epoch": 2.516891794756787,
      "grad_norm": 0.15831784904003143,
      "learning_rate": 1.610697089581992e-05,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8508256860077381,
      "num_tokens": 873088670.0,
      "step": 27356
    },
    {
      "entropy": 0.5817677639424801,
      "epoch": 2.5169838011282284,
      "grad_norm": 0.17289134860038757,
      "learning_rate": 1.610390406967829e-05,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8210476748645306,
      "num_tokens": 873119035.0,
      "step": 27357
    },
    {
      "entropy": 0.499105891212821,
      "epoch": 2.517075807499669,
      "grad_norm": 0.15032711625099182,
      "learning_rate": 1.6100837243536665e-05,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8501176051795483,
      "num_tokens": 873151312.0,
      "step": 27358
    },
    {
      "entropy": 0.6008208636194468,
      "epoch": 2.5171678138711107,
      "grad_norm": 0.16836901009082794,
      "learning_rate": 1.609777041739504e-05,
      "loss": 0.5917,
      "mean_token_accuracy": 0.810894925147295,
      "num_tokens": 873183279.0,
      "step": 27359
    },
    {
      "entropy": 0.48385611653793603,
      "epoch": 2.517259820242552,
      "grad_norm": 0.15739081799983978,
      "learning_rate": 1.6094703591253414e-05,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8458703719079494,
      "num_tokens": 873215067.0,
      "step": 27360
    },
    {
      "entropy": 0.47587295714765787,
      "epoch": 2.517351826613993,
      "grad_norm": 0.15448878705501556,
      "learning_rate": 1.6091636765111785e-05,
      "loss": 0.4727,
      "mean_token_accuracy": 0.847880195826292,
      "num_tokens": 873247020.0,
      "step": 27361
    },
    {
      "entropy": 0.48009174410253763,
      "epoch": 2.5174438329854345,
      "grad_norm": 0.15720973908901215,
      "learning_rate": 1.608856993897016e-05,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8523613773286343,
      "num_tokens": 873277627.0,
      "step": 27362
    },
    {
      "entropy": 0.41928697749972343,
      "epoch": 2.517535839356875,
      "grad_norm": 0.1470060497522354,
      "learning_rate": 1.6085503112828534e-05,
      "loss": 0.4103,
      "mean_token_accuracy": 0.8690248541533947,
      "num_tokens": 873310215.0,
      "step": 27363
    },
    {
      "entropy": 0.5179541986435652,
      "epoch": 2.5176278457283168,
      "grad_norm": 0.15831492841243744,
      "learning_rate": 1.608243628668691e-05,
      "loss": 0.5099,
      "mean_token_accuracy": 0.8403728306293488,
      "num_tokens": 873342471.0,
      "step": 27364
    },
    {
      "entropy": 0.5830502640455961,
      "epoch": 2.517719852099758,
      "grad_norm": 0.1656288355588913,
      "learning_rate": 1.6079369460545284e-05,
      "loss": 0.5747,
      "mean_token_accuracy": 0.818856094032526,
      "num_tokens": 873374813.0,
      "step": 27365
    },
    {
      "entropy": 0.4498228342272341,
      "epoch": 2.517811858471199,
      "grad_norm": 0.14895884692668915,
      "learning_rate": 1.6076302634403655e-05,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8636327013373375,
      "num_tokens": 873407078.0,
      "step": 27366
    },
    {
      "entropy": 0.5148942060768604,
      "epoch": 2.5179038648426406,
      "grad_norm": 0.15631800889968872,
      "learning_rate": 1.607323580826203e-05,
      "loss": 0.5013,
      "mean_token_accuracy": 0.8439406491816044,
      "num_tokens": 873439485.0,
      "step": 27367
    },
    {
      "entropy": 0.5353624848648906,
      "epoch": 2.5179958712140813,
      "grad_norm": 0.1622074842453003,
      "learning_rate": 1.6070168982120404e-05,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8355393782258034,
      "num_tokens": 873470881.0,
      "step": 27368
    },
    {
      "entropy": 0.5478932117111981,
      "epoch": 2.518087877585523,
      "grad_norm": 0.16138894855976105,
      "learning_rate": 1.606710215597878e-05,
      "loss": 0.542,
      "mean_token_accuracy": 0.828082200139761,
      "num_tokens": 873502843.0,
      "step": 27369
    },
    {
      "entropy": 0.5139928865246475,
      "epoch": 2.518179883956964,
      "grad_norm": 0.16412080824375153,
      "learning_rate": 1.606403532983715e-05,
      "loss": 0.5096,
      "mean_token_accuracy": 0.8381833843886852,
      "num_tokens": 873535087.0,
      "step": 27370
    },
    {
      "entropy": 0.5086610638536513,
      "epoch": 2.518271890328405,
      "grad_norm": 0.15716910362243652,
      "learning_rate": 1.6060968503695525e-05,
      "loss": 0.493,
      "mean_token_accuracy": 0.8428498990833759,
      "num_tokens": 873567438.0,
      "step": 27371
    },
    {
      "entropy": 0.4617308476008475,
      "epoch": 2.5183638966998467,
      "grad_norm": 0.15141859650611877,
      "learning_rate": 1.60579016775539e-05,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8557691872119904,
      "num_tokens": 873599697.0,
      "step": 27372
    },
    {
      "entropy": 0.6098173260688782,
      "epoch": 2.5184559030712874,
      "grad_norm": 0.1756729930639267,
      "learning_rate": 1.6054834851412274e-05,
      "loss": 0.5947,
      "mean_token_accuracy": 0.8125628940761089,
      "num_tokens": 873631910.0,
      "step": 27373
    },
    {
      "entropy": 0.6165027506649494,
      "epoch": 2.518547909442729,
      "grad_norm": 0.17201462388038635,
      "learning_rate": 1.6051768025270645e-05,
      "loss": 0.6039,
      "mean_token_accuracy": 0.8072566762566566,
      "num_tokens": 873663529.0,
      "step": 27374
    },
    {
      "entropy": 0.5262299086898565,
      "epoch": 2.51863991581417,
      "grad_norm": 0.16517052054405212,
      "learning_rate": 1.604870119912902e-05,
      "loss": 0.5029,
      "mean_token_accuracy": 0.8392763212323189,
      "num_tokens": 873694463.0,
      "step": 27375
    },
    {
      "entropy": 0.5126118683256209,
      "epoch": 2.5187319221856113,
      "grad_norm": 0.16343125700950623,
      "learning_rate": 1.6045634372987395e-05,
      "loss": 0.498,
      "mean_token_accuracy": 0.8387521579861641,
      "num_tokens": 873726252.0,
      "step": 27376
    },
    {
      "entropy": 0.4579858845099807,
      "epoch": 2.518823928557053,
      "grad_norm": 0.15232136845588684,
      "learning_rate": 1.6042567546845773e-05,
      "loss": 0.4419,
      "mean_token_accuracy": 0.8565392717719078,
      "num_tokens": 873758492.0,
      "step": 27377
    },
    {
      "entropy": 0.5461467448621988,
      "epoch": 2.5189159349284935,
      "grad_norm": 0.16369883716106415,
      "learning_rate": 1.6039500720704144e-05,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8312388397753239,
      "num_tokens": 873790317.0,
      "step": 27378
    },
    {
      "entropy": 0.43956895684823394,
      "epoch": 2.519007941299935,
      "grad_norm": 0.15131250023841858,
      "learning_rate": 1.603643389456252e-05,
      "loss": 0.4314,
      "mean_token_accuracy": 0.8628479726612568,
      "num_tokens": 873822002.0,
      "step": 27379
    },
    {
      "entropy": 0.38500533858314157,
      "epoch": 2.5190999476713762,
      "grad_norm": 0.13957661390304565,
      "learning_rate": 1.6033367068420893e-05,
      "loss": 0.3696,
      "mean_token_accuracy": 0.8820347562432289,
      "num_tokens": 873853742.0,
      "step": 27380
    },
    {
      "entropy": 0.40893311984837055,
      "epoch": 2.5191919540428174,
      "grad_norm": 0.14826303720474243,
      "learning_rate": 1.6030300242279268e-05,
      "loss": 0.3961,
      "mean_token_accuracy": 0.8738729543983936,
      "num_tokens": 873886263.0,
      "step": 27381
    },
    {
      "entropy": 0.5157372066751122,
      "epoch": 2.519283960414259,
      "grad_norm": 0.16359320282936096,
      "learning_rate": 1.602723341613764e-05,
      "loss": 0.512,
      "mean_token_accuracy": 0.8394507206976414,
      "num_tokens": 873917829.0,
      "step": 27382
    },
    {
      "entropy": 0.5357468519359827,
      "epoch": 2.5193759667856996,
      "grad_norm": 0.15807276964187622,
      "learning_rate": 1.6024166589996014e-05,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8325831666588783,
      "num_tokens": 873950059.0,
      "step": 27383
    },
    {
      "entropy": 0.4871191466227174,
      "epoch": 2.519467973157141,
      "grad_norm": 0.1547524482011795,
      "learning_rate": 1.602109976385439e-05,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8510080799460411,
      "num_tokens": 873981332.0,
      "step": 27384
    },
    {
      "entropy": 0.5828242935240269,
      "epoch": 2.5195599795285823,
      "grad_norm": 0.16879898309707642,
      "learning_rate": 1.6018032937712763e-05,
      "loss": 0.5721,
      "mean_token_accuracy": 0.8195721432566643,
      "num_tokens": 874012996.0,
      "step": 27385
    },
    {
      "entropy": 0.497200945392251,
      "epoch": 2.5196519859000235,
      "grad_norm": 0.1545996069908142,
      "learning_rate": 1.6014966111571138e-05,
      "loss": 0.4859,
      "mean_token_accuracy": 0.8461237587034702,
      "num_tokens": 874045192.0,
      "step": 27386
    },
    {
      "entropy": 0.5487806918099523,
      "epoch": 2.519743992271465,
      "grad_norm": 0.16419769823551178,
      "learning_rate": 1.601189928542951e-05,
      "loss": 0.5286,
      "mean_token_accuracy": 0.8321946375072002,
      "num_tokens": 874077186.0,
      "step": 27387
    },
    {
      "entropy": 0.5343973506242037,
      "epoch": 2.5198359986429057,
      "grad_norm": 0.16236022114753723,
      "learning_rate": 1.6008832459287884e-05,
      "loss": 0.5247,
      "mean_token_accuracy": 0.8338335603475571,
      "num_tokens": 874108309.0,
      "step": 27388
    },
    {
      "entropy": 0.4682199206436053,
      "epoch": 2.5199280050143473,
      "grad_norm": 0.1570669561624527,
      "learning_rate": 1.600576563314626e-05,
      "loss": 0.4582,
      "mean_token_accuracy": 0.8546743765473366,
      "num_tokens": 874140537.0,
      "step": 27389
    },
    {
      "entropy": 0.4792602742090821,
      "epoch": 2.5200200113857885,
      "grad_norm": 0.1492931842803955,
      "learning_rate": 1.6002698807004633e-05,
      "loss": 0.4699,
      "mean_token_accuracy": 0.8502393923699856,
      "num_tokens": 874172792.0,
      "step": 27390
    },
    {
      "entropy": 0.4895400167442858,
      "epoch": 2.5201120177572296,
      "grad_norm": 0.1574285924434662,
      "learning_rate": 1.5999631980863004e-05,
      "loss": 0.4765,
      "mean_token_accuracy": 0.8484074138104916,
      "num_tokens": 874205005.0,
      "step": 27391
    },
    {
      "entropy": 0.6324395909905434,
      "epoch": 2.520204024128671,
      "grad_norm": 0.1738874465227127,
      "learning_rate": 1.599656515472138e-05,
      "loss": 0.6226,
      "mean_token_accuracy": 0.8037400245666504,
      "num_tokens": 874236541.0,
      "step": 27392
    },
    {
      "entropy": 0.5708115622401237,
      "epoch": 2.520296030500112,
      "grad_norm": 0.1638810634613037,
      "learning_rate": 1.5993498328579754e-05,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8257911391556263,
      "num_tokens": 874268938.0,
      "step": 27393
    },
    {
      "entropy": 0.6009352440014482,
      "epoch": 2.5203880368715534,
      "grad_norm": 0.16584403812885284,
      "learning_rate": 1.5990431502438128e-05,
      "loss": 0.5847,
      "mean_token_accuracy": 0.8160592913627625,
      "num_tokens": 874301261.0,
      "step": 27394
    },
    {
      "entropy": 0.5669063422828913,
      "epoch": 2.5204800432429946,
      "grad_norm": 0.16362112760543823,
      "learning_rate": 1.5987364676296503e-05,
      "loss": 0.5613,
      "mean_token_accuracy": 0.820243775844574,
      "num_tokens": 874333512.0,
      "step": 27395
    },
    {
      "entropy": 0.44478536397218704,
      "epoch": 2.5205720496144357,
      "grad_norm": 0.1522369235754013,
      "learning_rate": 1.5984297850154874e-05,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8635677434504032,
      "num_tokens": 874365812.0,
      "step": 27396
    },
    {
      "entropy": 0.4694756790995598,
      "epoch": 2.5206640559858773,
      "grad_norm": 0.15832118690013885,
      "learning_rate": 1.598123102401325e-05,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8543919026851654,
      "num_tokens": 874397905.0,
      "step": 27397
    },
    {
      "entropy": 0.5247847940772772,
      "epoch": 2.520756062357318,
      "grad_norm": 0.16309210658073425,
      "learning_rate": 1.5978164197871623e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8340024538338184,
      "num_tokens": 874429853.0,
      "step": 27398
    },
    {
      "entropy": 0.48008991219103336,
      "epoch": 2.5208480687287596,
      "grad_norm": 0.15751798450946808,
      "learning_rate": 1.5975097371729998e-05,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8515659123659134,
      "num_tokens": 874461491.0,
      "step": 27399
    },
    {
      "entropy": 0.47537325602024794,
      "epoch": 2.5209400751002007,
      "grad_norm": 0.15624359250068665,
      "learning_rate": 1.597203054558837e-05,
      "loss": 0.4665,
      "mean_token_accuracy": 0.8493224009871483,
      "num_tokens": 874493734.0,
      "step": 27400
    },
    {
      "entropy": 0.5270224148407578,
      "epoch": 2.521032081471642,
      "grad_norm": 0.16454975306987762,
      "learning_rate": 1.5968963719446744e-05,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8329074494540691,
      "num_tokens": 874526036.0,
      "step": 27401
    },
    {
      "entropy": 0.5470715090632439,
      "epoch": 2.5211240878430834,
      "grad_norm": 0.16190378367900848,
      "learning_rate": 1.596589689330512e-05,
      "loss": 0.5404,
      "mean_token_accuracy": 0.8290186487138271,
      "num_tokens": 874557799.0,
      "step": 27402
    },
    {
      "entropy": 0.4984541479498148,
      "epoch": 2.521216094214524,
      "grad_norm": 0.15699659287929535,
      "learning_rate": 1.5962830067163493e-05,
      "loss": 0.4797,
      "mean_token_accuracy": 0.845769289880991,
      "num_tokens": 874589797.0,
      "step": 27403
    },
    {
      "entropy": 0.48988337349146605,
      "epoch": 2.5213081005859657,
      "grad_norm": 0.16161449253559113,
      "learning_rate": 1.5959763241021864e-05,
      "loss": 0.4862,
      "mean_token_accuracy": 0.8481257036328316,
      "num_tokens": 874621382.0,
      "step": 27404
    },
    {
      "entropy": 0.5123094888404012,
      "epoch": 2.521400106957407,
      "grad_norm": 0.1672327071428299,
      "learning_rate": 1.595669641488024e-05,
      "loss": 0.5085,
      "mean_token_accuracy": 0.8372073695063591,
      "num_tokens": 874652786.0,
      "step": 27405
    },
    {
      "entropy": 0.5565647054463625,
      "epoch": 2.521492113328848,
      "grad_norm": 0.16806094348430634,
      "learning_rate": 1.5953629588738614e-05,
      "loss": 0.5519,
      "mean_token_accuracy": 0.8226103819906712,
      "num_tokens": 874684617.0,
      "step": 27406
    },
    {
      "entropy": 0.5002845535054803,
      "epoch": 2.5215841197002895,
      "grad_norm": 0.15713122487068176,
      "learning_rate": 1.595056276259699e-05,
      "loss": 0.4893,
      "mean_token_accuracy": 0.8422419093549252,
      "num_tokens": 874716711.0,
      "step": 27407
    },
    {
      "entropy": 0.5224656462669373,
      "epoch": 2.52167612607173,
      "grad_norm": 0.16447943449020386,
      "learning_rate": 1.5947495936455363e-05,
      "loss": 0.5142,
      "mean_token_accuracy": 0.8351786844432354,
      "num_tokens": 874748813.0,
      "step": 27408
    },
    {
      "entropy": 0.587095096707344,
      "epoch": 2.5217681324431718,
      "grad_norm": 0.16795964539051056,
      "learning_rate": 1.5944429110313738e-05,
      "loss": 0.5673,
      "mean_token_accuracy": 0.8163065277040005,
      "num_tokens": 874780072.0,
      "step": 27409
    },
    {
      "entropy": 0.5127437796909362,
      "epoch": 2.521860138814613,
      "grad_norm": 0.16352304816246033,
      "learning_rate": 1.5941362284172112e-05,
      "loss": 0.5018,
      "mean_token_accuracy": 0.844820462167263,
      "num_tokens": 874811669.0,
      "step": 27410
    },
    {
      "entropy": 0.48183506075292826,
      "epoch": 2.521952145186054,
      "grad_norm": 0.15877023339271545,
      "learning_rate": 1.5938295458030487e-05,
      "loss": 0.4704,
      "mean_token_accuracy": 0.8491338044404984,
      "num_tokens": 874844050.0,
      "step": 27411
    },
    {
      "entropy": 0.4896266176365316,
      "epoch": 2.5220441515574956,
      "grad_norm": 0.15315556526184082,
      "learning_rate": 1.5935228631888858e-05,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8464682027697563,
      "num_tokens": 874876281.0,
      "step": 27412
    },
    {
      "entropy": 0.43802048824727535,
      "epoch": 2.5221361579289363,
      "grad_norm": 0.15659768879413605,
      "learning_rate": 1.5932161805747233e-05,
      "loss": 0.4297,
      "mean_token_accuracy": 0.8632519133388996,
      "num_tokens": 874907412.0,
      "step": 27413
    },
    {
      "entropy": 0.4858498042449355,
      "epoch": 2.522228164300378,
      "grad_norm": 0.15265125036239624,
      "learning_rate": 1.5929094979605608e-05,
      "loss": 0.478,
      "mean_token_accuracy": 0.8477685339748859,
      "num_tokens": 874939584.0,
      "step": 27414
    },
    {
      "entropy": 0.5436322437599301,
      "epoch": 2.522320170671819,
      "grad_norm": 0.16111387312412262,
      "learning_rate": 1.5926028153463982e-05,
      "loss": 0.5344,
      "mean_token_accuracy": 0.8334721848368645,
      "num_tokens": 874972041.0,
      "step": 27415
    },
    {
      "entropy": 0.48557755071669817,
      "epoch": 2.52241217704326,
      "grad_norm": 0.15844425559043884,
      "learning_rate": 1.5922961327322357e-05,
      "loss": 0.4722,
      "mean_token_accuracy": 0.8456205353140831,
      "num_tokens": 875003990.0,
      "step": 27416
    },
    {
      "entropy": 0.5152215557172894,
      "epoch": 2.5225041834147017,
      "grad_norm": 0.16447441279888153,
      "learning_rate": 1.5919894501180728e-05,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8388889171183109,
      "num_tokens": 875035901.0,
      "step": 27417
    },
    {
      "entropy": 0.5126299284165725,
      "epoch": 2.5225961897861424,
      "grad_norm": 0.1571342647075653,
      "learning_rate": 1.5916827675039103e-05,
      "loss": 0.4912,
      "mean_token_accuracy": 0.8452944941818714,
      "num_tokens": 875068496.0,
      "step": 27418
    },
    {
      "entropy": 0.5165267731063068,
      "epoch": 2.522688196157584,
      "grad_norm": 0.15632468461990356,
      "learning_rate": 1.5913760848897477e-05,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8446237593889236,
      "num_tokens": 875100420.0,
      "step": 27419
    },
    {
      "entropy": 0.5491517521440983,
      "epoch": 2.522780202529025,
      "grad_norm": 0.1606752723455429,
      "learning_rate": 1.5910694022755852e-05,
      "loss": 0.5326,
      "mean_token_accuracy": 0.8309700153768063,
      "num_tokens": 875132040.0,
      "step": 27420
    },
    {
      "entropy": 0.5190220223739743,
      "epoch": 2.5228722089004663,
      "grad_norm": 0.15830779075622559,
      "learning_rate": 1.5907627196614223e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8377459086477757,
      "num_tokens": 875164699.0,
      "step": 27421
    },
    {
      "entropy": 0.5430251490324736,
      "epoch": 2.522964215271908,
      "grad_norm": 0.1629553735256195,
      "learning_rate": 1.5904560370472598e-05,
      "loss": 0.5302,
      "mean_token_accuracy": 0.8348071984946728,
      "num_tokens": 875196209.0,
      "step": 27422
    },
    {
      "entropy": 0.47360848635435104,
      "epoch": 2.5230562216433485,
      "grad_norm": 0.15489597618579865,
      "learning_rate": 1.5901493544330973e-05,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8528396598994732,
      "num_tokens": 875228204.0,
      "step": 27423
    },
    {
      "entropy": 0.5196735952049494,
      "epoch": 2.52314822801479,
      "grad_norm": 0.16496476531028748,
      "learning_rate": 1.5898426718189347e-05,
      "loss": 0.5176,
      "mean_token_accuracy": 0.8400709331035614,
      "num_tokens": 875259762.0,
      "step": 27424
    },
    {
      "entropy": 0.5847544841235504,
      "epoch": 2.5232402343862312,
      "grad_norm": 0.164348766207695,
      "learning_rate": 1.5895359892047722e-05,
      "loss": 0.575,
      "mean_token_accuracy": 0.8207120969891548,
      "num_tokens": 875292000.0,
      "step": 27425
    },
    {
      "entropy": 0.5561698793899268,
      "epoch": 2.5233322407576724,
      "grad_norm": 0.17496031522750854,
      "learning_rate": 1.5892293065906093e-05,
      "loss": 0.5596,
      "mean_token_accuracy": 0.8232961259782314,
      "num_tokens": 875323933.0,
      "step": 27426
    },
    {
      "entropy": 0.5137750906869769,
      "epoch": 2.523424247129114,
      "grad_norm": 0.16097278892993927,
      "learning_rate": 1.5889226239764468e-05,
      "loss": 0.5111,
      "mean_token_accuracy": 0.8392200469970703,
      "num_tokens": 875355926.0,
      "step": 27427
    },
    {
      "entropy": 0.520591888576746,
      "epoch": 2.5235162535005546,
      "grad_norm": 0.15591204166412354,
      "learning_rate": 1.5886159413622842e-05,
      "loss": 0.5152,
      "mean_token_accuracy": 0.83790597692132,
      "num_tokens": 875388530.0,
      "step": 27428
    },
    {
      "entropy": 0.5471139415167272,
      "epoch": 2.523608259871996,
      "grad_norm": 0.16030386090278625,
      "learning_rate": 1.5883092587481217e-05,
      "loss": 0.5317,
      "mean_token_accuracy": 0.8331948034465313,
      "num_tokens": 875419995.0,
      "step": 27429
    },
    {
      "entropy": 0.5092904400080442,
      "epoch": 2.5237002662434374,
      "grad_norm": 0.16275154054164886,
      "learning_rate": 1.588002576133959e-05,
      "loss": 0.4821,
      "mean_token_accuracy": 0.8422192074358463,
      "num_tokens": 875451516.0,
      "step": 27430
    },
    {
      "entropy": 0.538079890422523,
      "epoch": 2.5237922726148785,
      "grad_norm": 0.1626659333705902,
      "learning_rate": 1.5876958935197963e-05,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8309925049543381,
      "num_tokens": 875484106.0,
      "step": 27431
    },
    {
      "entropy": 0.6103739340323955,
      "epoch": 2.52388427898632,
      "grad_norm": 0.16759122908115387,
      "learning_rate": 1.5873892109056338e-05,
      "loss": 0.5978,
      "mean_token_accuracy": 0.8119603954255581,
      "num_tokens": 875516234.0,
      "step": 27432
    },
    {
      "entropy": 0.5045324107632041,
      "epoch": 2.5239762853577608,
      "grad_norm": 0.15990100800991058,
      "learning_rate": 1.5870825282914712e-05,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8405160307884216,
      "num_tokens": 875548622.0,
      "step": 27433
    },
    {
      "entropy": 0.55286521371454,
      "epoch": 2.5240682917292023,
      "grad_norm": 0.1586671769618988,
      "learning_rate": 1.5867758456773084e-05,
      "loss": 0.545,
      "mean_token_accuracy": 0.8285384848713875,
      "num_tokens": 875580888.0,
      "step": 27434
    },
    {
      "entropy": 0.4328466239385307,
      "epoch": 2.5241602981006435,
      "grad_norm": 0.14809869229793549,
      "learning_rate": 1.5864691630631458e-05,
      "loss": 0.4167,
      "mean_token_accuracy": 0.86696857213974,
      "num_tokens": 875613251.0,
      "step": 27435
    },
    {
      "entropy": 0.47596043860539794,
      "epoch": 2.5242523044720846,
      "grad_norm": 0.14769789576530457,
      "learning_rate": 1.5861624804489833e-05,
      "loss": 0.461,
      "mean_token_accuracy": 0.8540057763457298,
      "num_tokens": 875645316.0,
      "step": 27436
    },
    {
      "entropy": 0.5226264391094446,
      "epoch": 2.524344310843526,
      "grad_norm": 0.1602407991886139,
      "learning_rate": 1.5858557978348208e-05,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8393892012536526,
      "num_tokens": 875677251.0,
      "step": 27437
    },
    {
      "entropy": 0.4491984536871314,
      "epoch": 2.524436317214967,
      "grad_norm": 0.1431845724582672,
      "learning_rate": 1.5855491152206582e-05,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8617666624486446,
      "num_tokens": 875709297.0,
      "step": 27438
    },
    {
      "entropy": 0.480281338095665,
      "epoch": 2.5245283235864084,
      "grad_norm": 0.15728360414505005,
      "learning_rate": 1.5852424326064957e-05,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8523573614656925,
      "num_tokens": 875741502.0,
      "step": 27439
    },
    {
      "entropy": 0.4958770480006933,
      "epoch": 2.5246203299578496,
      "grad_norm": 0.15608327090740204,
      "learning_rate": 1.584935749992333e-05,
      "loss": 0.4872,
      "mean_token_accuracy": 0.8426183462142944,
      "num_tokens": 875773455.0,
      "step": 27440
    },
    {
      "entropy": 0.5277892239391804,
      "epoch": 2.5247123363292907,
      "grad_norm": 0.1560017168521881,
      "learning_rate": 1.5846290673781706e-05,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8387358821928501,
      "num_tokens": 875805027.0,
      "step": 27441
    },
    {
      "entropy": 0.46134413219988346,
      "epoch": 2.5248043427007323,
      "grad_norm": 0.1477513611316681,
      "learning_rate": 1.5843223847640077e-05,
      "loss": 0.4468,
      "mean_token_accuracy": 0.8559309840202332,
      "num_tokens": 875837128.0,
      "step": 27442
    },
    {
      "entropy": 0.5144088473170996,
      "epoch": 2.524896349072173,
      "grad_norm": 0.15814988315105438,
      "learning_rate": 1.5840157021498452e-05,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8428749702870846,
      "num_tokens": 875868640.0,
      "step": 27443
    },
    {
      "entropy": 0.4711019927635789,
      "epoch": 2.5249883554436146,
      "grad_norm": 0.15537592768669128,
      "learning_rate": 1.5837090195356827e-05,
      "loss": 0.4656,
      "mean_token_accuracy": 0.8519451692700386,
      "num_tokens": 875900921.0,
      "step": 27444
    },
    {
      "entropy": 0.44539497466757894,
      "epoch": 2.5250803618150557,
      "grad_norm": 0.1476946324110031,
      "learning_rate": 1.58340233692152e-05,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8592703007161617,
      "num_tokens": 875932902.0,
      "step": 27445
    },
    {
      "entropy": 0.5789714399725199,
      "epoch": 2.525172368186497,
      "grad_norm": 0.1657513529062271,
      "learning_rate": 1.5830956543073576e-05,
      "loss": 0.5694,
      "mean_token_accuracy": 0.8237077705562115,
      "num_tokens": 875965547.0,
      "step": 27446
    },
    {
      "entropy": 0.6030786978080869,
      "epoch": 2.5252643745579384,
      "grad_norm": 0.16530683636665344,
      "learning_rate": 1.5827889716931947e-05,
      "loss": 0.5869,
      "mean_token_accuracy": 0.8166493698954582,
      "num_tokens": 875997748.0,
      "step": 27447
    },
    {
      "entropy": 0.533460165373981,
      "epoch": 2.525356380929379,
      "grad_norm": 0.16507165133953094,
      "learning_rate": 1.5824822890790322e-05,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8365520611405373,
      "num_tokens": 876029263.0,
      "step": 27448
    },
    {
      "entropy": 0.5153732653707266,
      "epoch": 2.5254483873008207,
      "grad_norm": 0.1589367538690567,
      "learning_rate": 1.5821756064648697e-05,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8401929587125778,
      "num_tokens": 876062031.0,
      "step": 27449
    },
    {
      "entropy": 0.4786268789321184,
      "epoch": 2.525540393672262,
      "grad_norm": 0.14933931827545166,
      "learning_rate": 1.581868923850707e-05,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8536046929657459,
      "num_tokens": 876094124.0,
      "step": 27450
    },
    {
      "entropy": 0.4765734316315502,
      "epoch": 2.525632400043703,
      "grad_norm": 0.15111184120178223,
      "learning_rate": 1.5815622412365442e-05,
      "loss": 0.464,
      "mean_token_accuracy": 0.8547043800354004,
      "num_tokens": 876126474.0,
      "step": 27451
    },
    {
      "entropy": 0.4707715711556375,
      "epoch": 2.5257244064151445,
      "grad_norm": 0.15071971714496613,
      "learning_rate": 1.5812555586223817e-05,
      "loss": 0.4527,
      "mean_token_accuracy": 0.8558223694562912,
      "num_tokens": 876158901.0,
      "step": 27452
    },
    {
      "entropy": 0.5306273940950632,
      "epoch": 2.525816412786585,
      "grad_norm": 0.1573346108198166,
      "learning_rate": 1.5809488760082192e-05,
      "loss": 0.5114,
      "mean_token_accuracy": 0.8396140336990356,
      "num_tokens": 876191262.0,
      "step": 27453
    },
    {
      "entropy": 0.5021675489842892,
      "epoch": 2.5259084191580268,
      "grad_norm": 0.16539280116558075,
      "learning_rate": 1.5806421933940566e-05,
      "loss": 0.5041,
      "mean_token_accuracy": 0.8412122093141079,
      "num_tokens": 876223065.0,
      "step": 27454
    },
    {
      "entropy": 0.5648883078247309,
      "epoch": 2.526000425529468,
      "grad_norm": 0.16555853188037872,
      "learning_rate": 1.580335510779894e-05,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8233447447419167,
      "num_tokens": 876254750.0,
      "step": 27455
    },
    {
      "entropy": 0.5124786077067256,
      "epoch": 2.526092431900909,
      "grad_norm": 0.16242370009422302,
      "learning_rate": 1.5800288281657312e-05,
      "loss": 0.501,
      "mean_token_accuracy": 0.8422171026468277,
      "num_tokens": 876286556.0,
      "step": 27456
    },
    {
      "entropy": 0.46018351381644607,
      "epoch": 2.5261844382723506,
      "grad_norm": 0.15337993204593658,
      "learning_rate": 1.5797221455515687e-05,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8555104061961174,
      "num_tokens": 876317892.0,
      "step": 27457
    },
    {
      "entropy": 0.5231654159724712,
      "epoch": 2.5262764446437913,
      "grad_norm": 0.15943196415901184,
      "learning_rate": 1.579415462937406e-05,
      "loss": 0.4956,
      "mean_token_accuracy": 0.838602177798748,
      "num_tokens": 876348555.0,
      "step": 27458
    },
    {
      "entropy": 0.46921875327825546,
      "epoch": 2.526368451015233,
      "grad_norm": 0.1530439555644989,
      "learning_rate": 1.5791087803232436e-05,
      "loss": 0.4598,
      "mean_token_accuracy": 0.8528440818190575,
      "num_tokens": 876380190.0,
      "step": 27459
    },
    {
      "entropy": 0.5228902316885069,
      "epoch": 2.526460457386674,
      "grad_norm": 0.16136863827705383,
      "learning_rate": 1.5788020977090807e-05,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8399910256266594,
      "num_tokens": 876411694.0,
      "step": 27460
    },
    {
      "entropy": 0.44269856717437506,
      "epoch": 2.526552463758115,
      "grad_norm": 0.1454673707485199,
      "learning_rate": 1.5784954150949182e-05,
      "loss": 0.4181,
      "mean_token_accuracy": 0.8657635524868965,
      "num_tokens": 876444034.0,
      "step": 27461
    },
    {
      "entropy": 0.4654745073057711,
      "epoch": 2.5266444701295567,
      "grad_norm": 0.15138272941112518,
      "learning_rate": 1.5781887324807557e-05,
      "loss": 0.4454,
      "mean_token_accuracy": 0.859241783618927,
      "num_tokens": 876476366.0,
      "step": 27462
    },
    {
      "entropy": 0.493172368966043,
      "epoch": 2.5267364765009974,
      "grad_norm": 0.15462853014469147,
      "learning_rate": 1.577882049866593e-05,
      "loss": 0.4785,
      "mean_token_accuracy": 0.8462599217891693,
      "num_tokens": 876508913.0,
      "step": 27463
    },
    {
      "entropy": 0.5222478155046701,
      "epoch": 2.526828482872439,
      "grad_norm": 0.1633201688528061,
      "learning_rate": 1.5775753672524303e-05,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8364397324621677,
      "num_tokens": 876540372.0,
      "step": 27464
    },
    {
      "entropy": 0.5099381704349071,
      "epoch": 2.52692048924388,
      "grad_norm": 0.16216552257537842,
      "learning_rate": 1.5772686846382677e-05,
      "loss": 0.4963,
      "mean_token_accuracy": 0.8413417264819145,
      "num_tokens": 876571680.0,
      "step": 27465
    },
    {
      "entropy": 0.625681210309267,
      "epoch": 2.5270124956153213,
      "grad_norm": 0.16886277496814728,
      "learning_rate": 1.5769620020241052e-05,
      "loss": 0.6094,
      "mean_token_accuracy": 0.8118008188903332,
      "num_tokens": 876604031.0,
      "step": 27466
    },
    {
      "entropy": 0.5147204101085663,
      "epoch": 2.527104501986763,
      "grad_norm": 0.15502889454364777,
      "learning_rate": 1.5766553194099427e-05,
      "loss": 0.506,
      "mean_token_accuracy": 0.8406480476260185,
      "num_tokens": 876635800.0,
      "step": 27467
    },
    {
      "entropy": 0.5153487864881754,
      "epoch": 2.5271965083582035,
      "grad_norm": 0.1647031158208847,
      "learning_rate": 1.57634863679578e-05,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8342233933508396,
      "num_tokens": 876666976.0,
      "step": 27468
    },
    {
      "entropy": 0.5815051114186645,
      "epoch": 2.527288514729645,
      "grad_norm": 0.1691960245370865,
      "learning_rate": 1.5760419541816176e-05,
      "loss": 0.5752,
      "mean_token_accuracy": 0.8184362053871155,
      "num_tokens": 876699167.0,
      "step": 27469
    },
    {
      "entropy": 0.5391943622380495,
      "epoch": 2.5273805211010862,
      "grad_norm": 0.16202569007873535,
      "learning_rate": 1.575735271567455e-05,
      "loss": 0.523,
      "mean_token_accuracy": 0.833130493760109,
      "num_tokens": 876730671.0,
      "step": 27470
    },
    {
      "entropy": 0.5108699491247535,
      "epoch": 2.5274725274725274,
      "grad_norm": 0.16196583211421967,
      "learning_rate": 1.5754285889532925e-05,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8409392274916172,
      "num_tokens": 876761538.0,
      "step": 27471
    },
    {
      "entropy": 0.5552132949233055,
      "epoch": 2.527564533843969,
      "grad_norm": 0.16806524991989136,
      "learning_rate": 1.5751219063391296e-05,
      "loss": 0.5389,
      "mean_token_accuracy": 0.8274839222431183,
      "num_tokens": 876792149.0,
      "step": 27472
    },
    {
      "entropy": 0.5076961545273662,
      "epoch": 2.5276565402154096,
      "grad_norm": 0.16317634284496307,
      "learning_rate": 1.574815223724967e-05,
      "loss": 0.496,
      "mean_token_accuracy": 0.8409505859017372,
      "num_tokens": 876823970.0,
      "step": 27473
    },
    {
      "entropy": 0.453724421095103,
      "epoch": 2.527748546586851,
      "grad_norm": 0.15111248195171356,
      "learning_rate": 1.5745085411108046e-05,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8595317415893078,
      "num_tokens": 876856069.0,
      "step": 27474
    },
    {
      "entropy": 0.5249647868331522,
      "epoch": 2.5278405529582924,
      "grad_norm": 0.16227839887142181,
      "learning_rate": 1.574201858496642e-05,
      "loss": 0.5135,
      "mean_token_accuracy": 0.8368845656514168,
      "num_tokens": 876888118.0,
      "step": 27475
    },
    {
      "entropy": 0.4838700145483017,
      "epoch": 2.5279325593297335,
      "grad_norm": 0.15410631895065308,
      "learning_rate": 1.5738951758824795e-05,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8504349552094936,
      "num_tokens": 876920159.0,
      "step": 27476
    },
    {
      "entropy": 0.46903597144410014,
      "epoch": 2.528024565701175,
      "grad_norm": 0.15238124132156372,
      "learning_rate": 1.5735884932683166e-05,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8526997230947018,
      "num_tokens": 876951636.0,
      "step": 27477
    },
    {
      "entropy": 0.48863217420876026,
      "epoch": 2.5281165720726158,
      "grad_norm": 0.1586627960205078,
      "learning_rate": 1.573281810654154e-05,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8480163365602493,
      "num_tokens": 876984137.0,
      "step": 27478
    },
    {
      "entropy": 0.43411816586740315,
      "epoch": 2.5282085784440573,
      "grad_norm": 0.15043844282627106,
      "learning_rate": 1.5729751280399916e-05,
      "loss": 0.4115,
      "mean_token_accuracy": 0.8660272471606731,
      "num_tokens": 877015447.0,
      "step": 27479
    },
    {
      "entropy": 0.4489554844330996,
      "epoch": 2.5283005848154985,
      "grad_norm": 0.16131889820098877,
      "learning_rate": 1.572668445425829e-05,
      "loss": 0.4336,
      "mean_token_accuracy": 0.8609674721956253,
      "num_tokens": 877046771.0,
      "step": 27480
    },
    {
      "entropy": 0.5008474923670292,
      "epoch": 2.5283925911869396,
      "grad_norm": 0.1575571447610855,
      "learning_rate": 1.572361762811666e-05,
      "loss": 0.5018,
      "mean_token_accuracy": 0.838309857994318,
      "num_tokens": 877079026.0,
      "step": 27481
    },
    {
      "entropy": 0.4680535839870572,
      "epoch": 2.528484597558381,
      "grad_norm": 0.16808687150478363,
      "learning_rate": 1.5720550801975036e-05,
      "loss": 0.46,
      "mean_token_accuracy": 0.8578804023563862,
      "num_tokens": 877111037.0,
      "step": 27482
    },
    {
      "entropy": 0.5068857748992741,
      "epoch": 2.528576603929822,
      "grad_norm": 0.16557247936725616,
      "learning_rate": 1.571748397583341e-05,
      "loss": 0.5014,
      "mean_token_accuracy": 0.842196736484766,
      "num_tokens": 877143118.0,
      "step": 27483
    },
    {
      "entropy": 0.48419783008284867,
      "epoch": 2.5286686103012634,
      "grad_norm": 0.1551390290260315,
      "learning_rate": 1.5714417149691785e-05,
      "loss": 0.4658,
      "mean_token_accuracy": 0.8538440093398094,
      "num_tokens": 877175208.0,
      "step": 27484
    },
    {
      "entropy": 0.5375168737955391,
      "epoch": 2.5287606166727046,
      "grad_norm": 0.1604606807231903,
      "learning_rate": 1.571135032355016e-05,
      "loss": 0.524,
      "mean_token_accuracy": 0.8342262208461761,
      "num_tokens": 877206741.0,
      "step": 27485
    },
    {
      "entropy": 0.4998793713748455,
      "epoch": 2.5288526230441457,
      "grad_norm": 0.16161492466926575,
      "learning_rate": 1.570828349740853e-05,
      "loss": 0.4875,
      "mean_token_accuracy": 0.8448375500738621,
      "num_tokens": 877238768.0,
      "step": 27486
    },
    {
      "entropy": 0.5300751263275743,
      "epoch": 2.5289446294155873,
      "grad_norm": 0.1596415936946869,
      "learning_rate": 1.5705216671266906e-05,
      "loss": 0.5284,
      "mean_token_accuracy": 0.8318420648574829,
      "num_tokens": 877270753.0,
      "step": 27487
    },
    {
      "entropy": 0.4440357219427824,
      "epoch": 2.529036635787028,
      "grad_norm": 0.1501729041337967,
      "learning_rate": 1.570214984512528e-05,
      "loss": 0.434,
      "mean_token_accuracy": 0.8629393614828587,
      "num_tokens": 877302879.0,
      "step": 27488
    },
    {
      "entropy": 0.5181625299155712,
      "epoch": 2.5291286421584696,
      "grad_norm": 0.16118228435516357,
      "learning_rate": 1.5699083018983655e-05,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8412109948694706,
      "num_tokens": 877334869.0,
      "step": 27489
    },
    {
      "entropy": 0.552944028750062,
      "epoch": 2.5292206485299107,
      "grad_norm": 0.161992609500885,
      "learning_rate": 1.5696016192842027e-05,
      "loss": 0.5393,
      "mean_token_accuracy": 0.8309062458574772,
      "num_tokens": 877366678.0,
      "step": 27490
    },
    {
      "entropy": 0.5928023848682642,
      "epoch": 2.529312654901352,
      "grad_norm": 0.17848412692546844,
      "learning_rate": 1.56929493667004e-05,
      "loss": 0.5804,
      "mean_token_accuracy": 0.8165985345840454,
      "num_tokens": 877397246.0,
      "step": 27491
    },
    {
      "entropy": 0.5797601705417037,
      "epoch": 2.5294046612727934,
      "grad_norm": 0.16510623693466187,
      "learning_rate": 1.5689882540558776e-05,
      "loss": 0.5638,
      "mean_token_accuracy": 0.8205288536846638,
      "num_tokens": 877429168.0,
      "step": 27492
    },
    {
      "entropy": 0.44320044107735157,
      "epoch": 2.529496667644234,
      "grad_norm": 0.15158165991306305,
      "learning_rate": 1.568681571441715e-05,
      "loss": 0.4354,
      "mean_token_accuracy": 0.8613980151712894,
      "num_tokens": 877461511.0,
      "step": 27493
    },
    {
      "entropy": 0.5069229165092111,
      "epoch": 2.5295886740156757,
      "grad_norm": 0.15264293551445007,
      "learning_rate": 1.5683748888275522e-05,
      "loss": 0.4961,
      "mean_token_accuracy": 0.8441086485981941,
      "num_tokens": 877494272.0,
      "step": 27494
    },
    {
      "entropy": 0.4640063834376633,
      "epoch": 2.529680680387117,
      "grad_norm": 0.14968790113925934,
      "learning_rate": 1.5680682062133896e-05,
      "loss": 0.4555,
      "mean_token_accuracy": 0.8573169037699699,
      "num_tokens": 877526801.0,
      "step": 27495
    },
    {
      "entropy": 0.5228749876841903,
      "epoch": 2.529772686758558,
      "grad_norm": 0.16092586517333984,
      "learning_rate": 1.567761523599227e-05,
      "loss": 0.5145,
      "mean_token_accuracy": 0.8384511545300484,
      "num_tokens": 877558630.0,
      "step": 27496
    },
    {
      "entropy": 0.5488126617856324,
      "epoch": 2.5298646931299995,
      "grad_norm": 0.16779932379722595,
      "learning_rate": 1.5674548409850646e-05,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8311196751892567,
      "num_tokens": 877589590.0,
      "step": 27497
    },
    {
      "entropy": 0.5185813987627625,
      "epoch": 2.52995669950144,
      "grad_norm": 0.1639403998851776,
      "learning_rate": 1.567148158370902e-05,
      "loss": 0.5223,
      "mean_token_accuracy": 0.8365667760372162,
      "num_tokens": 877621462.0,
      "step": 27498
    },
    {
      "entropy": 0.576180106960237,
      "epoch": 2.5300487058728818,
      "grad_norm": 0.16273555159568787,
      "learning_rate": 1.5668414757567395e-05,
      "loss": 0.5661,
      "mean_token_accuracy": 0.8178860172629356,
      "num_tokens": 877653920.0,
      "step": 27499
    },
    {
      "entropy": 0.542471868917346,
      "epoch": 2.530140712244323,
      "grad_norm": 0.16057731211185455,
      "learning_rate": 1.566534793142577e-05,
      "loss": 0.5241,
      "mean_token_accuracy": 0.8355162926018238,
      "num_tokens": 877686269.0,
      "step": 27500
    },
    {
      "entropy": 0.4237325391732156,
      "epoch": 2.530232718615764,
      "grad_norm": 0.14807851612567902,
      "learning_rate": 1.5662281105284144e-05,
      "loss": 0.4168,
      "mean_token_accuracy": 0.8690475597977638,
      "num_tokens": 877717913.0,
      "step": 27501
    },
    {
      "entropy": 0.47112299501895905,
      "epoch": 2.5303247249872056,
      "grad_norm": 0.15171673893928528,
      "learning_rate": 1.5659214279142516e-05,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8484062030911446,
      "num_tokens": 877750681.0,
      "step": 27502
    },
    {
      "entropy": 0.40473176073282957,
      "epoch": 2.5304167313586463,
      "grad_norm": 0.14488790929317474,
      "learning_rate": 1.565614745300089e-05,
      "loss": 0.3877,
      "mean_token_accuracy": 0.8756334781646729,
      "num_tokens": 877782772.0,
      "step": 27503
    },
    {
      "entropy": 0.6737252986058593,
      "epoch": 2.530508737730088,
      "grad_norm": 0.17597152292728424,
      "learning_rate": 1.5653080626859265e-05,
      "loss": 0.6573,
      "mean_token_accuracy": 0.7949530892074108,
      "num_tokens": 877815010.0,
      "step": 27504
    },
    {
      "entropy": 0.5384599673561752,
      "epoch": 2.530600744101529,
      "grad_norm": 0.15801332890987396,
      "learning_rate": 1.565001380071764e-05,
      "loss": 0.5381,
      "mean_token_accuracy": 0.8340439237654209,
      "num_tokens": 877847285.0,
      "step": 27505
    },
    {
      "entropy": 0.5098370928317308,
      "epoch": 2.53069275047297,
      "grad_norm": 0.15762268006801605,
      "learning_rate": 1.5646946974576014e-05,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8451930955052376,
      "num_tokens": 877879727.0,
      "step": 27506
    },
    {
      "entropy": 0.572915150783956,
      "epoch": 2.5307847568444117,
      "grad_norm": 0.16928155720233917,
      "learning_rate": 1.5643880148434385e-05,
      "loss": 0.5535,
      "mean_token_accuracy": 0.8245499096810818,
      "num_tokens": 877911218.0,
      "step": 27507
    },
    {
      "entropy": 0.5638852482661605,
      "epoch": 2.5308767632158524,
      "grad_norm": 0.1634128987789154,
      "learning_rate": 1.564081332229276e-05,
      "loss": 0.5506,
      "mean_token_accuracy": 0.8265179917216301,
      "num_tokens": 877943828.0,
      "step": 27508
    },
    {
      "entropy": 0.4954165145754814,
      "epoch": 2.530968769587294,
      "grad_norm": 0.15369915962219238,
      "learning_rate": 1.5637746496151135e-05,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8478402085602283,
      "num_tokens": 877975606.0,
      "step": 27509
    },
    {
      "entropy": 0.5969237210229039,
      "epoch": 2.531060775958735,
      "grad_norm": 0.16503946483135223,
      "learning_rate": 1.563467967000951e-05,
      "loss": 0.5844,
      "mean_token_accuracy": 0.8156669810414314,
      "num_tokens": 878007543.0,
      "step": 27510
    },
    {
      "entropy": 0.47843778133392334,
      "epoch": 2.5311527823301763,
      "grad_norm": 0.15193508565425873,
      "learning_rate": 1.563161284386788e-05,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8544701933860779,
      "num_tokens": 878039478.0,
      "step": 27511
    },
    {
      "entropy": 0.5255174972116947,
      "epoch": 2.531244788701618,
      "grad_norm": 0.15960155427455902,
      "learning_rate": 1.5628546017726255e-05,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8393499813973904,
      "num_tokens": 878071432.0,
      "step": 27512
    },
    {
      "entropy": 0.490228065289557,
      "epoch": 2.5313367950730585,
      "grad_norm": 0.1575915664434433,
      "learning_rate": 1.562547919158463e-05,
      "loss": 0.4714,
      "mean_token_accuracy": 0.8486229069530964,
      "num_tokens": 878102628.0,
      "step": 27513
    },
    {
      "entropy": 0.6649123206734657,
      "epoch": 2.5314288014445,
      "grad_norm": 0.174725741147995,
      "learning_rate": 1.5622412365443005e-05,
      "loss": 0.6497,
      "mean_token_accuracy": 0.7932644970715046,
      "num_tokens": 878134261.0,
      "step": 27514
    },
    {
      "entropy": 0.44295182218775153,
      "epoch": 2.5315208078159412,
      "grad_norm": 0.15164050459861755,
      "learning_rate": 1.561934553930138e-05,
      "loss": 0.436,
      "mean_token_accuracy": 0.8584431186318398,
      "num_tokens": 878166182.0,
      "step": 27515
    },
    {
      "entropy": 0.5781790313776582,
      "epoch": 2.5316128141873824,
      "grad_norm": 0.16627256572246552,
      "learning_rate": 1.561627871315975e-05,
      "loss": 0.5697,
      "mean_token_accuracy": 0.8182057961821556,
      "num_tokens": 878197981.0,
      "step": 27516
    },
    {
      "entropy": 0.5150677300989628,
      "epoch": 2.531704820558824,
      "grad_norm": 0.16079694032669067,
      "learning_rate": 1.5613211887018125e-05,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8475242964923382,
      "num_tokens": 878229392.0,
      "step": 27517
    },
    {
      "entropy": 0.4259915598668158,
      "epoch": 2.5317968269302646,
      "grad_norm": 0.1483437418937683,
      "learning_rate": 1.56101450608765e-05,
      "loss": 0.4203,
      "mean_token_accuracy": 0.8679146505892277,
      "num_tokens": 878261636.0,
      "step": 27518
    },
    {
      "entropy": 0.39985561871435493,
      "epoch": 2.531888833301706,
      "grad_norm": 0.1446811556816101,
      "learning_rate": 1.5607078234734874e-05,
      "loss": 0.3929,
      "mean_token_accuracy": 0.8754244074225426,
      "num_tokens": 878293344.0,
      "step": 27519
    },
    {
      "entropy": 0.49615045823156834,
      "epoch": 2.5319808396731474,
      "grad_norm": 0.15605440735816956,
      "learning_rate": 1.5604011408593246e-05,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8433062136173248,
      "num_tokens": 878325826.0,
      "step": 27520
    },
    {
      "entropy": 0.5003594653680921,
      "epoch": 2.5320728460445885,
      "grad_norm": 0.15684251487255096,
      "learning_rate": 1.560094458245162e-05,
      "loss": 0.4882,
      "mean_token_accuracy": 0.8449111469089985,
      "num_tokens": 878357838.0,
      "step": 27521
    },
    {
      "entropy": 0.5873441081494093,
      "epoch": 2.53216485241603,
      "grad_norm": 0.1657245010137558,
      "learning_rate": 1.5597877756309995e-05,
      "loss": 0.5634,
      "mean_token_accuracy": 0.8229261711239815,
      "num_tokens": 878389677.0,
      "step": 27522
    },
    {
      "entropy": 0.42485205084085464,
      "epoch": 2.532256858787471,
      "grad_norm": 0.15110670030117035,
      "learning_rate": 1.559481093016837e-05,
      "loss": 0.421,
      "mean_token_accuracy": 0.8659634590148926,
      "num_tokens": 878422111.0,
      "step": 27523
    },
    {
      "entropy": 0.5182520179077983,
      "epoch": 2.5323488651589123,
      "grad_norm": 0.1598394215106964,
      "learning_rate": 1.559174410402674e-05,
      "loss": 0.5069,
      "mean_token_accuracy": 0.8393749222159386,
      "num_tokens": 878453637.0,
      "step": 27524
    },
    {
      "entropy": 0.49528797157108784,
      "epoch": 2.5324408715303535,
      "grad_norm": 0.1520298570394516,
      "learning_rate": 1.5588677277885116e-05,
      "loss": 0.4826,
      "mean_token_accuracy": 0.8451217748224735,
      "num_tokens": 878485625.0,
      "step": 27525
    },
    {
      "entropy": 0.49750399589538574,
      "epoch": 2.5325328779017946,
      "grad_norm": 0.1577782779932022,
      "learning_rate": 1.558561045174349e-05,
      "loss": 0.4793,
      "mean_token_accuracy": 0.8461775258183479,
      "num_tokens": 878517458.0,
      "step": 27526
    },
    {
      "entropy": 0.4178681466728449,
      "epoch": 2.532624884273236,
      "grad_norm": 0.147279292345047,
      "learning_rate": 1.5582543625601865e-05,
      "loss": 0.4093,
      "mean_token_accuracy": 0.867997832596302,
      "num_tokens": 878549771.0,
      "step": 27527
    },
    {
      "entropy": 0.5520017528906465,
      "epoch": 2.5327168906446773,
      "grad_norm": 0.16381411254405975,
      "learning_rate": 1.557947679946024e-05,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8294366076588631,
      "num_tokens": 878581632.0,
      "step": 27528
    },
    {
      "entropy": 0.5695761777460575,
      "epoch": 2.5328088970161184,
      "grad_norm": 0.17003528773784637,
      "learning_rate": 1.557640997331861e-05,
      "loss": 0.562,
      "mean_token_accuracy": 0.8219980448484421,
      "num_tokens": 878613468.0,
      "step": 27529
    },
    {
      "entropy": 0.5030977847054601,
      "epoch": 2.5329009033875596,
      "grad_norm": 0.16191866993904114,
      "learning_rate": 1.557334314717699e-05,
      "loss": 0.4994,
      "mean_token_accuracy": 0.8451092131435871,
      "num_tokens": 878645871.0,
      "step": 27530
    },
    {
      "entropy": 0.5385425696149468,
      "epoch": 2.5329929097590007,
      "grad_norm": 0.16157513856887817,
      "learning_rate": 1.5570276321035363e-05,
      "loss": 0.5325,
      "mean_token_accuracy": 0.8316142112016678,
      "num_tokens": 878677993.0,
      "step": 27531
    },
    {
      "entropy": 0.4577763946726918,
      "epoch": 2.5330849161304423,
      "grad_norm": 0.15643025934696198,
      "learning_rate": 1.5567209494893738e-05,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8573017753660679,
      "num_tokens": 878709796.0,
      "step": 27532
    },
    {
      "entropy": 0.48095822241157293,
      "epoch": 2.5331769225018834,
      "grad_norm": 0.15646705031394958,
      "learning_rate": 1.556414266875211e-05,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8464099317789078,
      "num_tokens": 878741509.0,
      "step": 27533
    },
    {
      "entropy": 0.6192661803215742,
      "epoch": 2.5332689288733246,
      "grad_norm": 0.1668761968612671,
      "learning_rate": 1.5561075842610484e-05,
      "loss": 0.5979,
      "mean_token_accuracy": 0.8127796463668346,
      "num_tokens": 878773413.0,
      "step": 27534
    },
    {
      "entropy": 0.5826269127428532,
      "epoch": 2.5333609352447657,
      "grad_norm": 0.1697874218225479,
      "learning_rate": 1.555800901646886e-05,
      "loss": 0.5747,
      "mean_token_accuracy": 0.8180871568620205,
      "num_tokens": 878805909.0,
      "step": 27535
    },
    {
      "entropy": 0.49278664239682257,
      "epoch": 2.533452941616207,
      "grad_norm": 0.15947306156158447,
      "learning_rate": 1.5554942190327233e-05,
      "loss": 0.479,
      "mean_token_accuracy": 0.8430229350924492,
      "num_tokens": 878837740.0,
      "step": 27536
    },
    {
      "entropy": 0.46095906267873943,
      "epoch": 2.5335449479876484,
      "grad_norm": 0.1497519612312317,
      "learning_rate": 1.5551875364185605e-05,
      "loss": 0.4466,
      "mean_token_accuracy": 0.8584785647690296,
      "num_tokens": 878869100.0,
      "step": 27537
    },
    {
      "entropy": 0.5220719871576875,
      "epoch": 2.5336369543590895,
      "grad_norm": 0.15664254128932953,
      "learning_rate": 1.554880853804398e-05,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8403407670557499,
      "num_tokens": 878900300.0,
      "step": 27538
    },
    {
      "entropy": 0.5680605680681765,
      "epoch": 2.5337289607305307,
      "grad_norm": 0.16802409291267395,
      "learning_rate": 1.5545741711902354e-05,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8237234465777874,
      "num_tokens": 878932278.0,
      "step": 27539
    },
    {
      "entropy": 0.4819596279412508,
      "epoch": 2.533820967101972,
      "grad_norm": 0.15234801173210144,
      "learning_rate": 1.554267488576073e-05,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8489962443709373,
      "num_tokens": 878964777.0,
      "step": 27540
    },
    {
      "entropy": 0.47601051442325115,
      "epoch": 2.533912973473413,
      "grad_norm": 0.1551257222890854,
      "learning_rate": 1.55396080596191e-05,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8511253818869591,
      "num_tokens": 878997254.0,
      "step": 27541
    },
    {
      "entropy": 0.43455787654966116,
      "epoch": 2.5340049798448545,
      "grad_norm": 0.15138036012649536,
      "learning_rate": 1.5536541233477474e-05,
      "loss": 0.422,
      "mean_token_accuracy": 0.8654058538377285,
      "num_tokens": 879029446.0,
      "step": 27542
    },
    {
      "entropy": 0.4964918391779065,
      "epoch": 2.5340969862162956,
      "grad_norm": 0.15689831972122192,
      "learning_rate": 1.553347440733585e-05,
      "loss": 0.48,
      "mean_token_accuracy": 0.8476653210818768,
      "num_tokens": 879061652.0,
      "step": 27543
    },
    {
      "entropy": 0.44688510918058455,
      "epoch": 2.5341889925877368,
      "grad_norm": 0.1521461009979248,
      "learning_rate": 1.5530407581194224e-05,
      "loss": 0.4355,
      "mean_token_accuracy": 0.8606016784906387,
      "num_tokens": 879093484.0,
      "step": 27544
    },
    {
      "entropy": 0.6105732526630163,
      "epoch": 2.534280998959178,
      "grad_norm": 0.1743333786725998,
      "learning_rate": 1.55273407550526e-05,
      "loss": 0.5954,
      "mean_token_accuracy": 0.8116840310394764,
      "num_tokens": 879125414.0,
      "step": 27545
    },
    {
      "entropy": 0.5325533449649811,
      "epoch": 2.534373005330619,
      "grad_norm": 0.16030284762382507,
      "learning_rate": 1.552427392891097e-05,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8320506699383259,
      "num_tokens": 879157581.0,
      "step": 27546
    },
    {
      "entropy": 0.5078556090593338,
      "epoch": 2.5344650117020606,
      "grad_norm": 0.15770936012268066,
      "learning_rate": 1.5521207102769344e-05,
      "loss": 0.4962,
      "mean_token_accuracy": 0.8423919565975666,
      "num_tokens": 879189121.0,
      "step": 27547
    },
    {
      "entropy": 0.5171134134288877,
      "epoch": 2.5345570180735018,
      "grad_norm": 0.16259197890758514,
      "learning_rate": 1.551814027662772e-05,
      "loss": 0.4995,
      "mean_token_accuracy": 0.8397371433675289,
      "num_tokens": 879220875.0,
      "step": 27548
    },
    {
      "entropy": 0.5924079604446888,
      "epoch": 2.534649024444943,
      "grad_norm": 0.16614876687526703,
      "learning_rate": 1.5515073450486094e-05,
      "loss": 0.581,
      "mean_token_accuracy": 0.8182885572314262,
      "num_tokens": 879252868.0,
      "step": 27549
    },
    {
      "entropy": 0.4132724069058895,
      "epoch": 2.534741030816384,
      "grad_norm": 0.14562633633613586,
      "learning_rate": 1.5512006624344465e-05,
      "loss": 0.3927,
      "mean_token_accuracy": 0.8751840628683567,
      "num_tokens": 879285206.0,
      "step": 27550
    },
    {
      "entropy": 0.4482438648119569,
      "epoch": 2.534833037187825,
      "grad_norm": 0.1495952308177948,
      "learning_rate": 1.550893979820284e-05,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8601407743990421,
      "num_tokens": 879316985.0,
      "step": 27551
    },
    {
      "entropy": 0.4974362001521513,
      "epoch": 2.5349250435592667,
      "grad_norm": 0.16023565828800201,
      "learning_rate": 1.5505872972061214e-05,
      "loss": 0.4872,
      "mean_token_accuracy": 0.8446488752961159,
      "num_tokens": 879349128.0,
      "step": 27552
    },
    {
      "entropy": 0.6484391037374735,
      "epoch": 2.535017049930708,
      "grad_norm": 0.1748078167438507,
      "learning_rate": 1.550280614591959e-05,
      "loss": 0.6325,
      "mean_token_accuracy": 0.7991555705666542,
      "num_tokens": 879381131.0,
      "step": 27553
    },
    {
      "entropy": 0.556187467649579,
      "epoch": 2.535109056302149,
      "grad_norm": 0.1685553342103958,
      "learning_rate": 1.5499739319777963e-05,
      "loss": 0.5388,
      "mean_token_accuracy": 0.8274469263851643,
      "num_tokens": 879412710.0,
      "step": 27554
    },
    {
      "entropy": 0.4634636649861932,
      "epoch": 2.53520106267359,
      "grad_norm": 0.15074095129966736,
      "learning_rate": 1.5496672493636335e-05,
      "loss": 0.4587,
      "mean_token_accuracy": 0.8564459793269634,
      "num_tokens": 879444738.0,
      "step": 27555
    },
    {
      "entropy": 0.5726642394438386,
      "epoch": 2.5352930690450313,
      "grad_norm": 0.17194893956184387,
      "learning_rate": 1.549360566749471e-05,
      "loss": 0.5585,
      "mean_token_accuracy": 0.8237759731709957,
      "num_tokens": 879476297.0,
      "step": 27556
    },
    {
      "entropy": 0.5030915401875973,
      "epoch": 2.535385075416473,
      "grad_norm": 0.16449514031410217,
      "learning_rate": 1.5490538841353084e-05,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8420294709503651,
      "num_tokens": 879507560.0,
      "step": 27557
    },
    {
      "entropy": 0.5641990602016449,
      "epoch": 2.535477081787914,
      "grad_norm": 0.1598387062549591,
      "learning_rate": 1.548747201521146e-05,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8265436068177223,
      "num_tokens": 879539635.0,
      "step": 27558
    },
    {
      "entropy": 0.55881267786026,
      "epoch": 2.535569088159355,
      "grad_norm": 0.1656816303730011,
      "learning_rate": 1.548440518906983e-05,
      "loss": 0.5521,
      "mean_token_accuracy": 0.829938679933548,
      "num_tokens": 879571067.0,
      "step": 27559
    },
    {
      "entropy": 0.592491302639246,
      "epoch": 2.5356610945307962,
      "grad_norm": 0.17507268488407135,
      "learning_rate": 1.5481338362928208e-05,
      "loss": 0.5869,
      "mean_token_accuracy": 0.8149746581912041,
      "num_tokens": 879603106.0,
      "step": 27560
    },
    {
      "entropy": 0.5665137921459973,
      "epoch": 2.5357531009022374,
      "grad_norm": 0.16712796688079834,
      "learning_rate": 1.5478271536786583e-05,
      "loss": 0.5536,
      "mean_token_accuracy": 0.8244073204696178,
      "num_tokens": 879635058.0,
      "step": 27561
    },
    {
      "entropy": 0.5120168013963848,
      "epoch": 2.535845107273679,
      "grad_norm": 0.15688923001289368,
      "learning_rate": 1.5475204710644957e-05,
      "loss": 0.4979,
      "mean_token_accuracy": 0.8450286947190762,
      "num_tokens": 879667509.0,
      "step": 27562
    },
    {
      "entropy": 0.42200299352407455,
      "epoch": 2.53593711364512,
      "grad_norm": 0.15220847725868225,
      "learning_rate": 1.547213788450333e-05,
      "loss": 0.404,
      "mean_token_accuracy": 0.8693712130188942,
      "num_tokens": 879699080.0,
      "step": 27563
    },
    {
      "entropy": 0.4918134310282767,
      "epoch": 2.5360291200165612,
      "grad_norm": 0.15450309216976166,
      "learning_rate": 1.5469071058361703e-05,
      "loss": 0.4783,
      "mean_token_accuracy": 0.849661149084568,
      "num_tokens": 879731678.0,
      "step": 27564
    },
    {
      "entropy": 0.4454573525581509,
      "epoch": 2.5361211263880024,
      "grad_norm": 0.14972184598445892,
      "learning_rate": 1.5466004232220078e-05,
      "loss": 0.4386,
      "mean_token_accuracy": 0.8607419654726982,
      "num_tokens": 879763784.0,
      "step": 27565
    },
    {
      "entropy": 0.5695819184184074,
      "epoch": 2.5362131327594435,
      "grad_norm": 0.17401635646820068,
      "learning_rate": 1.5462937406078452e-05,
      "loss": 0.5725,
      "mean_token_accuracy": 0.8204970583319664,
      "num_tokens": 879794525.0,
      "step": 27566
    },
    {
      "entropy": 0.4679831634275615,
      "epoch": 2.536305139130885,
      "grad_norm": 0.1520656943321228,
      "learning_rate": 1.5459870579936824e-05,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8548391051590443,
      "num_tokens": 879826518.0,
      "step": 27567
    },
    {
      "entropy": 0.5511993379332125,
      "epoch": 2.536397145502326,
      "grad_norm": 0.1631087064743042,
      "learning_rate": 1.5456803753795198e-05,
      "loss": 0.544,
      "mean_token_accuracy": 0.8275901675224304,
      "num_tokens": 879858876.0,
      "step": 27568
    },
    {
      "entropy": 0.6095849238336086,
      "epoch": 2.5364891518737673,
      "grad_norm": 0.17546944320201874,
      "learning_rate": 1.5453736927653573e-05,
      "loss": 0.6059,
      "mean_token_accuracy": 0.8076796308159828,
      "num_tokens": 879891192.0,
      "step": 27569
    },
    {
      "entropy": 0.5584848560392857,
      "epoch": 2.5365811582452085,
      "grad_norm": 0.16083188354969025,
      "learning_rate": 1.5450670101511948e-05,
      "loss": 0.5426,
      "mean_token_accuracy": 0.8324565067887306,
      "num_tokens": 879922921.0,
      "step": 27570
    },
    {
      "entropy": 0.5324679296463728,
      "epoch": 2.5366731646166496,
      "grad_norm": 0.16278637945652008,
      "learning_rate": 1.544760327537032e-05,
      "loss": 0.5208,
      "mean_token_accuracy": 0.8348008207976818,
      "num_tokens": 879955471.0,
      "step": 27571
    },
    {
      "entropy": 0.5917774718254805,
      "epoch": 2.536765170988091,
      "grad_norm": 0.17060159146785736,
      "learning_rate": 1.5444536449228693e-05,
      "loss": 0.5818,
      "mean_token_accuracy": 0.8204889670014381,
      "num_tokens": 879986094.0,
      "step": 27572
    },
    {
      "entropy": 0.5651383483782411,
      "epoch": 2.5368571773595323,
      "grad_norm": 0.16964417695999146,
      "learning_rate": 1.5441469623087068e-05,
      "loss": 0.5609,
      "mean_token_accuracy": 0.8207869678735733,
      "num_tokens": 880017796.0,
      "step": 27573
    },
    {
      "entropy": 0.5171219250187278,
      "epoch": 2.5369491837309734,
      "grad_norm": 0.16469840705394745,
      "learning_rate": 1.5438402796945443e-05,
      "loss": 0.5088,
      "mean_token_accuracy": 0.8382349088788033,
      "num_tokens": 880049642.0,
      "step": 27574
    },
    {
      "entropy": 0.48037294670939445,
      "epoch": 2.5370411901024146,
      "grad_norm": 0.15556667745113373,
      "learning_rate": 1.5435335970803817e-05,
      "loss": 0.473,
      "mean_token_accuracy": 0.8509038984775543,
      "num_tokens": 880081875.0,
      "step": 27575
    },
    {
      "entropy": 0.5542248438578099,
      "epoch": 2.5371331964738557,
      "grad_norm": 0.16164276003837585,
      "learning_rate": 1.543226914466219e-05,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8298530802130699,
      "num_tokens": 880113769.0,
      "step": 27576
    },
    {
      "entropy": 0.577095303684473,
      "epoch": 2.5372252028452973,
      "grad_norm": 0.16826091706752777,
      "learning_rate": 1.5429202318520563e-05,
      "loss": 0.5779,
      "mean_token_accuracy": 0.8186462260782719,
      "num_tokens": 880145559.0,
      "step": 27577
    },
    {
      "entropy": 0.5374100394546986,
      "epoch": 2.5373172092167384,
      "grad_norm": 0.15742158889770508,
      "learning_rate": 1.5426135492378938e-05,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8357390463352203,
      "num_tokens": 880177903.0,
      "step": 27578
    },
    {
      "entropy": 0.5325026493519545,
      "epoch": 2.5374092155881796,
      "grad_norm": 0.16090092062950134,
      "learning_rate": 1.5423068666237313e-05,
      "loss": 0.4951,
      "mean_token_accuracy": 0.8386152908205986,
      "num_tokens": 880208805.0,
      "step": 27579
    },
    {
      "entropy": 0.5763975875452161,
      "epoch": 2.5375012219596207,
      "grad_norm": 0.16586683690547943,
      "learning_rate": 1.5420001840095684e-05,
      "loss": 0.5641,
      "mean_token_accuracy": 0.8241145461797714,
      "num_tokens": 880241488.0,
      "step": 27580
    },
    {
      "entropy": 0.5022317287512124,
      "epoch": 2.537593228331062,
      "grad_norm": 0.15440583229064941,
      "learning_rate": 1.541693501395406e-05,
      "loss": 0.4871,
      "mean_token_accuracy": 0.8490348570048809,
      "num_tokens": 880273497.0,
      "step": 27581
    },
    {
      "entropy": 0.5608067186549306,
      "epoch": 2.5376852347025034,
      "grad_norm": 0.16230450570583344,
      "learning_rate": 1.5413868187812433e-05,
      "loss": 0.5375,
      "mean_token_accuracy": 0.8317755833268166,
      "num_tokens": 880305892.0,
      "step": 27582
    },
    {
      "entropy": 0.5164709901437163,
      "epoch": 2.5377772410739445,
      "grad_norm": 0.16382327675819397,
      "learning_rate": 1.5410801361670808e-05,
      "loss": 0.5097,
      "mean_token_accuracy": 0.8391444645822048,
      "num_tokens": 880338198.0,
      "step": 27583
    },
    {
      "entropy": 0.540534595027566,
      "epoch": 2.5378692474453857,
      "grad_norm": 0.15782086551189423,
      "learning_rate": 1.5407734535529182e-05,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8355664946138859,
      "num_tokens": 880370706.0,
      "step": 27584
    },
    {
      "entropy": 0.5245020408183336,
      "epoch": 2.537961253816827,
      "grad_norm": 0.16070646047592163,
      "learning_rate": 1.5404667709387554e-05,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8381564021110535,
      "num_tokens": 880402728.0,
      "step": 27585
    },
    {
      "entropy": 0.4959778403863311,
      "epoch": 2.538053260188268,
      "grad_norm": 0.15590156614780426,
      "learning_rate": 1.540160088324593e-05,
      "loss": 0.4804,
      "mean_token_accuracy": 0.8509062193334103,
      "num_tokens": 880434561.0,
      "step": 27586
    },
    {
      "entropy": 0.5044985897839069,
      "epoch": 2.5381452665597095,
      "grad_norm": 0.15616223216056824,
      "learning_rate": 1.5398534057104303e-05,
      "loss": 0.4892,
      "mean_token_accuracy": 0.8459640517830849,
      "num_tokens": 880466211.0,
      "step": 27587
    },
    {
      "entropy": 0.5006192899309099,
      "epoch": 2.5382372729311506,
      "grad_norm": 0.15885084867477417,
      "learning_rate": 1.5395467230962678e-05,
      "loss": 0.4815,
      "mean_token_accuracy": 0.8494201749563217,
      "num_tokens": 880498475.0,
      "step": 27588
    },
    {
      "entropy": 0.4662532676011324,
      "epoch": 2.538329279302592,
      "grad_norm": 0.16017118096351624,
      "learning_rate": 1.539240040482105e-05,
      "loss": 0.4514,
      "mean_token_accuracy": 0.8550592958927155,
      "num_tokens": 880530014.0,
      "step": 27589
    },
    {
      "entropy": 0.42780629452317953,
      "epoch": 2.538421285674033,
      "grad_norm": 0.1468440443277359,
      "learning_rate": 1.5389333578679424e-05,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8644977025687695,
      "num_tokens": 880562512.0,
      "step": 27590
    },
    {
      "entropy": 0.6039804928004742,
      "epoch": 2.538513292045474,
      "grad_norm": 0.174861878156662,
      "learning_rate": 1.53862667525378e-05,
      "loss": 0.5931,
      "mean_token_accuracy": 0.8096238560974598,
      "num_tokens": 880593771.0,
      "step": 27591
    },
    {
      "entropy": 0.46193862380459905,
      "epoch": 2.5386052984169156,
      "grad_norm": 0.14949513971805573,
      "learning_rate": 1.5383199926396176e-05,
      "loss": 0.4478,
      "mean_token_accuracy": 0.8536109663546085,
      "num_tokens": 880625498.0,
      "step": 27592
    },
    {
      "entropy": 0.587189300917089,
      "epoch": 2.5386973047883568,
      "grad_norm": 0.16569775342941284,
      "learning_rate": 1.5380133100254548e-05,
      "loss": 0.5803,
      "mean_token_accuracy": 0.8186362870037556,
      "num_tokens": 880657794.0,
      "step": 27593
    },
    {
      "entropy": 0.5107993260025978,
      "epoch": 2.538789311159798,
      "grad_norm": 0.16070953011512756,
      "learning_rate": 1.5377066274112922e-05,
      "loss": 0.5116,
      "mean_token_accuracy": 0.8375945501029491,
      "num_tokens": 880690247.0,
      "step": 27594
    },
    {
      "entropy": 0.5958576677367091,
      "epoch": 2.538881317531239,
      "grad_norm": 0.17351379990577698,
      "learning_rate": 1.5373999447971297e-05,
      "loss": 0.5863,
      "mean_token_accuracy": 0.8158215060830116,
      "num_tokens": 880721902.0,
      "step": 27595
    },
    {
      "entropy": 0.5418737009167671,
      "epoch": 2.53897332390268,
      "grad_norm": 0.16524630784988403,
      "learning_rate": 1.537093262182967e-05,
      "loss": 0.5358,
      "mean_token_accuracy": 0.8324750624597073,
      "num_tokens": 880754032.0,
      "step": 27596
    },
    {
      "entropy": 0.519491643179208,
      "epoch": 2.5390653302741217,
      "grad_norm": 0.1593904048204422,
      "learning_rate": 1.5367865795688043e-05,
      "loss": 0.4954,
      "mean_token_accuracy": 0.840717040002346,
      "num_tokens": 880785718.0,
      "step": 27597
    },
    {
      "entropy": 0.4727132199332118,
      "epoch": 2.539157336645563,
      "grad_norm": 0.14594265818595886,
      "learning_rate": 1.5364798969546417e-05,
      "loss": 0.4594,
      "mean_token_accuracy": 0.853216215968132,
      "num_tokens": 880817395.0,
      "step": 27598
    },
    {
      "entropy": 0.5041998354718089,
      "epoch": 2.539249343017004,
      "grad_norm": 0.15838584303855896,
      "learning_rate": 1.5361732143404792e-05,
      "loss": 0.4951,
      "mean_token_accuracy": 0.8466919250786304,
      "num_tokens": 880849427.0,
      "step": 27599
    },
    {
      "entropy": 0.5518663665279746,
      "epoch": 2.539341349388445,
      "grad_norm": 0.1661151945590973,
      "learning_rate": 1.5358665317263167e-05,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8301492221653461,
      "num_tokens": 880880506.0,
      "step": 27600
    },
    {
      "entropy": 0.5378349586389959,
      "epoch": 2.5394333557598863,
      "grad_norm": 0.1616792380809784,
      "learning_rate": 1.5355598491121538e-05,
      "loss": 0.5324,
      "mean_token_accuracy": 0.8320276662707329,
      "num_tokens": 880912779.0,
      "step": 27601
    },
    {
      "entropy": 0.5390320625156164,
      "epoch": 2.539525362131328,
      "grad_norm": 0.1616346538066864,
      "learning_rate": 1.5352531664979913e-05,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8320247940719128,
      "num_tokens": 880944644.0,
      "step": 27602
    },
    {
      "entropy": 0.5486834719777107,
      "epoch": 2.539617368502769,
      "grad_norm": 0.16791708767414093,
      "learning_rate": 1.5349464838838287e-05,
      "loss": 0.5523,
      "mean_token_accuracy": 0.827607586979866,
      "num_tokens": 880977199.0,
      "step": 27603
    },
    {
      "entropy": 0.6612535305321217,
      "epoch": 2.53970937487421,
      "grad_norm": 0.17400343716144562,
      "learning_rate": 1.5346398012696662e-05,
      "loss": 0.6621,
      "mean_token_accuracy": 0.7962509542703629,
      "num_tokens": 881009567.0,
      "step": 27604
    },
    {
      "entropy": 0.4916988154873252,
      "epoch": 2.5398013812456512,
      "grad_norm": 0.16051647067070007,
      "learning_rate": 1.5343331186555037e-05,
      "loss": 0.4738,
      "mean_token_accuracy": 0.8481092974543571,
      "num_tokens": 881041274.0,
      "step": 27605
    },
    {
      "entropy": 0.5378235802054405,
      "epoch": 2.5398933876170924,
      "grad_norm": 0.16192492842674255,
      "learning_rate": 1.5340264360413408e-05,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8343631029129028,
      "num_tokens": 881074031.0,
      "step": 27606
    },
    {
      "entropy": 0.5712324315682054,
      "epoch": 2.539985393988534,
      "grad_norm": 0.16152586042881012,
      "learning_rate": 1.5337197534271782e-05,
      "loss": 0.5611,
      "mean_token_accuracy": 0.8243147544562817,
      "num_tokens": 881106629.0,
      "step": 27607
    },
    {
      "entropy": 0.4254491897299886,
      "epoch": 2.540077400359975,
      "grad_norm": 0.16030927002429962,
      "learning_rate": 1.5334130708130157e-05,
      "loss": 0.4167,
      "mean_token_accuracy": 0.8682979121804237,
      "num_tokens": 881139160.0,
      "step": 27608
    },
    {
      "entropy": 0.5812245756387711,
      "epoch": 2.5401694067314162,
      "grad_norm": 0.16320791840553284,
      "learning_rate": 1.5331063881988532e-05,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8247861489653587,
      "num_tokens": 881171000.0,
      "step": 27609
    },
    {
      "entropy": 0.40353567618876696,
      "epoch": 2.5402614131028574,
      "grad_norm": 0.13975918292999268,
      "learning_rate": 1.5327997055846903e-05,
      "loss": 0.3834,
      "mean_token_accuracy": 0.8747483417391777,
      "num_tokens": 881203117.0,
      "step": 27610
    },
    {
      "entropy": 0.484322932548821,
      "epoch": 2.5403534194742985,
      "grad_norm": 0.15319587290287018,
      "learning_rate": 1.5324930229705278e-05,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8512351773679256,
      "num_tokens": 881234879.0,
      "step": 27611
    },
    {
      "entropy": 0.48105441592633724,
      "epoch": 2.54044542584574,
      "grad_norm": 0.15590880811214447,
      "learning_rate": 1.5321863403563652e-05,
      "loss": 0.4662,
      "mean_token_accuracy": 0.8525304831564426,
      "num_tokens": 881266454.0,
      "step": 27612
    },
    {
      "entropy": 0.5996843408793211,
      "epoch": 2.540537432217181,
      "grad_norm": 0.16865409910678864,
      "learning_rate": 1.5318796577422027e-05,
      "loss": 0.5891,
      "mean_token_accuracy": 0.8144354484975338,
      "num_tokens": 881298342.0,
      "step": 27613
    },
    {
      "entropy": 0.6294915173202753,
      "epoch": 2.5406294385886223,
      "grad_norm": 0.16483493149280548,
      "learning_rate": 1.53157297512804e-05,
      "loss": 0.6197,
      "mean_token_accuracy": 0.8052841275930405,
      "num_tokens": 881330516.0,
      "step": 27614
    },
    {
      "entropy": 0.4838863480836153,
      "epoch": 2.5407214449600635,
      "grad_norm": 0.1554408073425293,
      "learning_rate": 1.5312662925138773e-05,
      "loss": 0.4716,
      "mean_token_accuracy": 0.8507682234048843,
      "num_tokens": 881362050.0,
      "step": 27615
    },
    {
      "entropy": 0.5000070165842772,
      "epoch": 2.5408134513315046,
      "grad_norm": 0.1569204330444336,
      "learning_rate": 1.5309596098997147e-05,
      "loss": 0.4803,
      "mean_token_accuracy": 0.8468875698745251,
      "num_tokens": 881393982.0,
      "step": 27616
    },
    {
      "entropy": 0.4828872215002775,
      "epoch": 2.540905457702946,
      "grad_norm": 0.16240853071212769,
      "learning_rate": 1.5306529272855522e-05,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8478628918528557,
      "num_tokens": 881425101.0,
      "step": 27617
    },
    {
      "entropy": 0.4736256320029497,
      "epoch": 2.5409974640743873,
      "grad_norm": 0.15643422305583954,
      "learning_rate": 1.5303462446713897e-05,
      "loss": 0.4647,
      "mean_token_accuracy": 0.8548098206520081,
      "num_tokens": 881456617.0,
      "step": 27618
    },
    {
      "entropy": 0.6145694516599178,
      "epoch": 2.5410894704458284,
      "grad_norm": 0.17511725425720215,
      "learning_rate": 1.5300395620572268e-05,
      "loss": 0.6041,
      "mean_token_accuracy": 0.8060619086027145,
      "num_tokens": 881488004.0,
      "step": 27619
    },
    {
      "entropy": 0.46118696592748165,
      "epoch": 2.5411814768172696,
      "grad_norm": 0.1564871221780777,
      "learning_rate": 1.5297328794430643e-05,
      "loss": 0.4514,
      "mean_token_accuracy": 0.8558289706707001,
      "num_tokens": 881519576.0,
      "step": 27620
    },
    {
      "entropy": 0.4322966740000993,
      "epoch": 2.5412734831887107,
      "grad_norm": 0.14540880918502808,
      "learning_rate": 1.529426196828902e-05,
      "loss": 0.4198,
      "mean_token_accuracy": 0.8691730499267578,
      "num_tokens": 881551809.0,
      "step": 27621
    },
    {
      "entropy": 0.4932854790240526,
      "epoch": 2.5413654895601523,
      "grad_norm": 0.15650662779808044,
      "learning_rate": 1.5291195142147395e-05,
      "loss": 0.4868,
      "mean_token_accuracy": 0.8480531759560108,
      "num_tokens": 881583303.0,
      "step": 27622
    },
    {
      "entropy": 0.4855492733186111,
      "epoch": 2.5414574959315934,
      "grad_norm": 0.15547356009483337,
      "learning_rate": 1.5288128316005767e-05,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8484455943107605,
      "num_tokens": 881615438.0,
      "step": 27623
    },
    {
      "entropy": 0.5438357023522258,
      "epoch": 2.5415495023030346,
      "grad_norm": 0.15946733951568604,
      "learning_rate": 1.528506148986414e-05,
      "loss": 0.5249,
      "mean_token_accuracy": 0.8334803469479084,
      "num_tokens": 881647057.0,
      "step": 27624
    },
    {
      "entropy": 0.5175397987477481,
      "epoch": 2.5416415086744757,
      "grad_norm": 0.15492013096809387,
      "learning_rate": 1.5281994663722516e-05,
      "loss": 0.5042,
      "mean_token_accuracy": 0.8414860516786575,
      "num_tokens": 881678880.0,
      "step": 27625
    },
    {
      "entropy": 0.5837755016982555,
      "epoch": 2.541733515045917,
      "grad_norm": 0.16832557320594788,
      "learning_rate": 1.527892783758089e-05,
      "loss": 0.5807,
      "mean_token_accuracy": 0.8181545846164227,
      "num_tokens": 881711415.0,
      "step": 27626
    },
    {
      "entropy": 0.414990259334445,
      "epoch": 2.5418255214173584,
      "grad_norm": 0.1459352970123291,
      "learning_rate": 1.5275861011439262e-05,
      "loss": 0.404,
      "mean_token_accuracy": 0.8716326989233494,
      "num_tokens": 881743135.0,
      "step": 27627
    },
    {
      "entropy": 0.37271126406267285,
      "epoch": 2.5419175277887995,
      "grad_norm": 0.14419855177402496,
      "learning_rate": 1.5272794185297636e-05,
      "loss": 0.3564,
      "mean_token_accuracy": 0.8843738548457623,
      "num_tokens": 881775072.0,
      "step": 27628
    },
    {
      "entropy": 0.5359947262331843,
      "epoch": 2.5420095341602407,
      "grad_norm": 0.15866884589195251,
      "learning_rate": 1.526972735915601e-05,
      "loss": 0.5302,
      "mean_token_accuracy": 0.8351451717317104,
      "num_tokens": 881807106.0,
      "step": 27629
    },
    {
      "entropy": 0.4790884441463277,
      "epoch": 2.542101540531682,
      "grad_norm": 0.1480739414691925,
      "learning_rate": 1.5266660533014386e-05,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8512066230177879,
      "num_tokens": 881839159.0,
      "step": 27630
    },
    {
      "entropy": 0.5484141367487609,
      "epoch": 2.542193546903123,
      "grad_norm": 0.1642666459083557,
      "learning_rate": 1.5263593706872757e-05,
      "loss": 0.535,
      "mean_token_accuracy": 0.8318750634789467,
      "num_tokens": 881870994.0,
      "step": 27631
    },
    {
      "entropy": 0.456122612580657,
      "epoch": 2.5422855532745645,
      "grad_norm": 0.14888611435890198,
      "learning_rate": 1.526052688073113e-05,
      "loss": 0.4455,
      "mean_token_accuracy": 0.8594580180943012,
      "num_tokens": 881903707.0,
      "step": 27632
    },
    {
      "entropy": 0.4448514273390174,
      "epoch": 2.5423775596460056,
      "grad_norm": 0.1463066041469574,
      "learning_rate": 1.5257460054589506e-05,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8592074066400528,
      "num_tokens": 881936373.0,
      "step": 27633
    },
    {
      "entropy": 0.5865797456353903,
      "epoch": 2.542469566017447,
      "grad_norm": 0.16663552820682526,
      "learning_rate": 1.5254393228447881e-05,
      "loss": 0.5748,
      "mean_token_accuracy": 0.819670170545578,
      "num_tokens": 881968817.0,
      "step": 27634
    },
    {
      "entropy": 0.5715001868084073,
      "epoch": 2.542561572388888,
      "grad_norm": 0.17100225389003754,
      "learning_rate": 1.5251326402306254e-05,
      "loss": 0.5667,
      "mean_token_accuracy": 0.8257244452834129,
      "num_tokens": 881999498.0,
      "step": 27635
    },
    {
      "entropy": 0.48100972454994917,
      "epoch": 2.542653578760329,
      "grad_norm": 0.1591428965330124,
      "learning_rate": 1.5248259576164629e-05,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8500593453645706,
      "num_tokens": 882031481.0,
      "step": 27636
    },
    {
      "entropy": 0.4632652651052922,
      "epoch": 2.5427455851317706,
      "grad_norm": 0.1554369032382965,
      "learning_rate": 1.5245192750023002e-05,
      "loss": 0.443,
      "mean_token_accuracy": 0.856769721955061,
      "num_tokens": 882063025.0,
      "step": 27637
    },
    {
      "entropy": 0.4711194187402725,
      "epoch": 2.5428375915032118,
      "grad_norm": 0.15287770330905914,
      "learning_rate": 1.5242125923881376e-05,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8524800054728985,
      "num_tokens": 882095605.0,
      "step": 27638
    },
    {
      "entropy": 0.3816979117691517,
      "epoch": 2.542929597874653,
      "grad_norm": 0.14613725244998932,
      "learning_rate": 1.5239059097739749e-05,
      "loss": 0.3606,
      "mean_token_accuracy": 0.882670059800148,
      "num_tokens": 882127768.0,
      "step": 27639
    },
    {
      "entropy": 0.4706883653998375,
      "epoch": 2.543021604246094,
      "grad_norm": 0.15378955006599426,
      "learning_rate": 1.5235992271598124e-05,
      "loss": 0.461,
      "mean_token_accuracy": 0.8514585569500923,
      "num_tokens": 882159463.0,
      "step": 27640
    },
    {
      "entropy": 0.45682069938629866,
      "epoch": 2.543113610617535,
      "grad_norm": 0.14885985851287842,
      "learning_rate": 1.5232925445456497e-05,
      "loss": 0.449,
      "mean_token_accuracy": 0.8553495965898037,
      "num_tokens": 882191257.0,
      "step": 27641
    },
    {
      "entropy": 0.46481182146817446,
      "epoch": 2.5432056169889767,
      "grad_norm": 0.15679949522018433,
      "learning_rate": 1.5229858619314871e-05,
      "loss": 0.4524,
      "mean_token_accuracy": 0.8555116355419159,
      "num_tokens": 882223379.0,
      "step": 27642
    },
    {
      "entropy": 0.4973567221313715,
      "epoch": 2.543297623360418,
      "grad_norm": 0.1645544320344925,
      "learning_rate": 1.5226791793173244e-05,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8421605825424194,
      "num_tokens": 882255427.0,
      "step": 27643
    },
    {
      "entropy": 0.5071465545333922,
      "epoch": 2.543389629731859,
      "grad_norm": 0.15661394596099854,
      "learning_rate": 1.5223724967031619e-05,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8397937454283237,
      "num_tokens": 882287685.0,
      "step": 27644
    },
    {
      "entropy": 0.5200637567322701,
      "epoch": 2.5434816361033,
      "grad_norm": 0.15958978235721588,
      "learning_rate": 1.5220658140889994e-05,
      "loss": 0.5216,
      "mean_token_accuracy": 0.8396167904138565,
      "num_tokens": 882319355.0,
      "step": 27645
    },
    {
      "entropy": 0.47612641751766205,
      "epoch": 2.5435736424747413,
      "grad_norm": 0.16073229908943176,
      "learning_rate": 1.5217591314748367e-05,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8521409928798676,
      "num_tokens": 882350760.0,
      "step": 27646
    },
    {
      "entropy": 0.48938733292743564,
      "epoch": 2.543665648846183,
      "grad_norm": 0.15779033303260803,
      "learning_rate": 1.5214524488606741e-05,
      "loss": 0.483,
      "mean_token_accuracy": 0.8463448844850063,
      "num_tokens": 882382440.0,
      "step": 27647
    },
    {
      "entropy": 0.5351384109817445,
      "epoch": 2.543757655217624,
      "grad_norm": 0.16190457344055176,
      "learning_rate": 1.5211457662465114e-05,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8329937048256397,
      "num_tokens": 882415072.0,
      "step": 27648
    },
    {
      "entropy": 0.5712194899097085,
      "epoch": 2.543849661589065,
      "grad_norm": 0.1640690714120865,
      "learning_rate": 1.5208390836323489e-05,
      "loss": 0.5458,
      "mean_token_accuracy": 0.826498381793499,
      "num_tokens": 882446527.0,
      "step": 27649
    },
    {
      "entropy": 0.4808695064857602,
      "epoch": 2.5439416679605062,
      "grad_norm": 0.15501739084720612,
      "learning_rate": 1.5205324010181862e-05,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8497772216796875,
      "num_tokens": 882479121.0,
      "step": 27650
    },
    {
      "entropy": 0.5875524487346411,
      "epoch": 2.5440336743319474,
      "grad_norm": 0.17261596024036407,
      "learning_rate": 1.5202257184040236e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8195032104849815,
      "num_tokens": 882509729.0,
      "step": 27651
    },
    {
      "entropy": 0.5906483344733715,
      "epoch": 2.544125680703389,
      "grad_norm": 0.1680452972650528,
      "learning_rate": 1.5199190357898613e-05,
      "loss": 0.5716,
      "mean_token_accuracy": 0.822510939091444,
      "num_tokens": 882541964.0,
      "step": 27652
    },
    {
      "entropy": 0.6016996037214994,
      "epoch": 2.54421768707483,
      "grad_norm": 0.1692807972431183,
      "learning_rate": 1.5196123531756987e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8143683932721615,
      "num_tokens": 882574121.0,
      "step": 27653
    },
    {
      "entropy": 0.5400411169975996,
      "epoch": 2.5443096934462712,
      "grad_norm": 0.15754322707653046,
      "learning_rate": 1.519305670561536e-05,
      "loss": 0.5268,
      "mean_token_accuracy": 0.8362237401306629,
      "num_tokens": 882606889.0,
      "step": 27654
    },
    {
      "entropy": 0.5507265077903867,
      "epoch": 2.5444016998177124,
      "grad_norm": 0.1657557338476181,
      "learning_rate": 1.5189989879473735e-05,
      "loss": 0.5377,
      "mean_token_accuracy": 0.8315714225172997,
      "num_tokens": 882639016.0,
      "step": 27655
    },
    {
      "entropy": 0.43328365636989474,
      "epoch": 2.5444937061891535,
      "grad_norm": 0.14698739349842072,
      "learning_rate": 1.5186923053332108e-05,
      "loss": 0.411,
      "mean_token_accuracy": 0.8670594468712807,
      "num_tokens": 882670864.0,
      "step": 27656
    },
    {
      "entropy": 0.4921927673276514,
      "epoch": 2.544585712560595,
      "grad_norm": 0.16347628831863403,
      "learning_rate": 1.5183856227190483e-05,
      "loss": 0.4724,
      "mean_token_accuracy": 0.8477768935263157,
      "num_tokens": 882700837.0,
      "step": 27657
    },
    {
      "entropy": 0.5769752059131861,
      "epoch": 2.544677718932036,
      "grad_norm": 0.16618666052818298,
      "learning_rate": 1.5180789401048856e-05,
      "loss": 0.5666,
      "mean_token_accuracy": 0.8216676488518715,
      "num_tokens": 882733042.0,
      "step": 27658
    },
    {
      "entropy": 0.6747205704450607,
      "epoch": 2.5447697253034773,
      "grad_norm": 0.17546407878398895,
      "learning_rate": 1.517772257490723e-05,
      "loss": 0.6552,
      "mean_token_accuracy": 0.7939896062016487,
      "num_tokens": 882765621.0,
      "step": 27659
    },
    {
      "entropy": 0.5731069166213274,
      "epoch": 2.5448617316749185,
      "grad_norm": 0.16207976639270782,
      "learning_rate": 1.5174655748765603e-05,
      "loss": 0.5596,
      "mean_token_accuracy": 0.8270155116915703,
      "num_tokens": 882797737.0,
      "step": 27660
    },
    {
      "entropy": 0.5248086946085095,
      "epoch": 2.5449537380463596,
      "grad_norm": 0.16124695539474487,
      "learning_rate": 1.5171588922623978e-05,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8399439118802547,
      "num_tokens": 882829607.0,
      "step": 27661
    },
    {
      "entropy": 0.4980604201555252,
      "epoch": 2.545045744417801,
      "grad_norm": 0.1572699099779129,
      "learning_rate": 1.516852209648235e-05,
      "loss": 0.489,
      "mean_token_accuracy": 0.8469434380531311,
      "num_tokens": 882861326.0,
      "step": 27662
    },
    {
      "entropy": 0.49094394501298666,
      "epoch": 2.5451377507892423,
      "grad_norm": 0.15669693052768707,
      "learning_rate": 1.5165455270340725e-05,
      "loss": 0.4843,
      "mean_token_accuracy": 0.844034917652607,
      "num_tokens": 882893420.0,
      "step": 27663
    },
    {
      "entropy": 0.49598605185747147,
      "epoch": 2.5452297571606834,
      "grad_norm": 0.15671327710151672,
      "learning_rate": 1.51623884441991e-05,
      "loss": 0.4835,
      "mean_token_accuracy": 0.8440848737955093,
      "num_tokens": 882925541.0,
      "step": 27664
    },
    {
      "entropy": 0.5782316736876965,
      "epoch": 2.5453217635321246,
      "grad_norm": 0.16821768879890442,
      "learning_rate": 1.5159321618057473e-05,
      "loss": 0.5698,
      "mean_token_accuracy": 0.8204251453280449,
      "num_tokens": 882957434.0,
      "step": 27665
    },
    {
      "entropy": 0.46625447552651167,
      "epoch": 2.5454137699035657,
      "grad_norm": 0.15165011584758759,
      "learning_rate": 1.5156254791915848e-05,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8532943390309811,
      "num_tokens": 882989712.0,
      "step": 27666
    },
    {
      "entropy": 0.5302077196538448,
      "epoch": 2.5455057762750073,
      "grad_norm": 0.1612682342529297,
      "learning_rate": 1.515318796577422e-05,
      "loss": 0.5121,
      "mean_token_accuracy": 0.83967674523592,
      "num_tokens": 883021739.0,
      "step": 27667
    },
    {
      "entropy": 0.5098563252249733,
      "epoch": 2.5455977826464484,
      "grad_norm": 0.16462430357933044,
      "learning_rate": 1.5150121139632595e-05,
      "loss": 0.5067,
      "mean_token_accuracy": 0.8403122685849667,
      "num_tokens": 883053245.0,
      "step": 27668
    },
    {
      "entropy": 0.4558375533670187,
      "epoch": 2.5456897890178896,
      "grad_norm": 0.15979450941085815,
      "learning_rate": 1.5147054313490968e-05,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8535011298954487,
      "num_tokens": 883084493.0,
      "step": 27669
    },
    {
      "entropy": 0.5252083130180836,
      "epoch": 2.5457817953893307,
      "grad_norm": 0.15840044617652893,
      "learning_rate": 1.5143987487349343e-05,
      "loss": 0.5137,
      "mean_token_accuracy": 0.83977385237813,
      "num_tokens": 883115950.0,
      "step": 27670
    },
    {
      "entropy": 0.5275250906124711,
      "epoch": 2.545873801760772,
      "grad_norm": 0.16272282600402832,
      "learning_rate": 1.5140920661207716e-05,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8351394832134247,
      "num_tokens": 883147872.0,
      "step": 27671
    },
    {
      "entropy": 0.6298133935779333,
      "epoch": 2.5459658081322134,
      "grad_norm": 0.1693442016839981,
      "learning_rate": 1.513785383506609e-05,
      "loss": 0.6256,
      "mean_token_accuracy": 0.8069794289767742,
      "num_tokens": 883180175.0,
      "step": 27672
    },
    {
      "entropy": 0.37436226033605635,
      "epoch": 2.5460578145036545,
      "grad_norm": 0.14015918970108032,
      "learning_rate": 1.5134787008924463e-05,
      "loss": 0.3702,
      "mean_token_accuracy": 0.8801637589931488,
      "num_tokens": 883212444.0,
      "step": 27673
    },
    {
      "entropy": 0.4701484120450914,
      "epoch": 2.5461498208750957,
      "grad_norm": 0.15522275865077972,
      "learning_rate": 1.5131720182782838e-05,
      "loss": 0.4648,
      "mean_token_accuracy": 0.8532895408570766,
      "num_tokens": 883244368.0,
      "step": 27674
    },
    {
      "entropy": 0.4761611260473728,
      "epoch": 2.546241827246537,
      "grad_norm": 0.14935174584388733,
      "learning_rate": 1.5128653356641213e-05,
      "loss": 0.4649,
      "mean_token_accuracy": 0.8530998043715954,
      "num_tokens": 883275861.0,
      "step": 27675
    },
    {
      "entropy": 0.4456908777356148,
      "epoch": 2.546333833617978,
      "grad_norm": 0.15045729279518127,
      "learning_rate": 1.5125586530499586e-05,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8575461879372597,
      "num_tokens": 883307334.0,
      "step": 27676
    },
    {
      "entropy": 0.5503247156739235,
      "epoch": 2.5464258399894195,
      "grad_norm": 0.16390793025493622,
      "learning_rate": 1.512251970435796e-05,
      "loss": 0.5423,
      "mean_token_accuracy": 0.8299512043595314,
      "num_tokens": 883339247.0,
      "step": 27677
    },
    {
      "entropy": 0.5142579115927219,
      "epoch": 2.5465178463608606,
      "grad_norm": 0.15751567482948303,
      "learning_rate": 1.5119452878216333e-05,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8437977135181427,
      "num_tokens": 883371518.0,
      "step": 27678
    },
    {
      "entropy": 0.39990393444895744,
      "epoch": 2.546609852732302,
      "grad_norm": 0.14463910460472107,
      "learning_rate": 1.5116386052074708e-05,
      "loss": 0.38,
      "mean_token_accuracy": 0.8759371712803841,
      "num_tokens": 883403805.0,
      "step": 27679
    },
    {
      "entropy": 0.45045374101027846,
      "epoch": 2.546701859103743,
      "grad_norm": 0.14948813617229462,
      "learning_rate": 1.5113319225933081e-05,
      "loss": 0.4452,
      "mean_token_accuracy": 0.8580624014139175,
      "num_tokens": 883436242.0,
      "step": 27680
    },
    {
      "entropy": 0.4950899323448539,
      "epoch": 2.546793865475184,
      "grad_norm": 0.15720021724700928,
      "learning_rate": 1.5110252399791456e-05,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8465541079640388,
      "num_tokens": 883467680.0,
      "step": 27681
    },
    {
      "entropy": 0.5504188863560557,
      "epoch": 2.5468858718466256,
      "grad_norm": 0.1645773947238922,
      "learning_rate": 1.5107185573649832e-05,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8325052224099636,
      "num_tokens": 883499787.0,
      "step": 27682
    },
    {
      "entropy": 0.4403890995308757,
      "epoch": 2.5469778782180668,
      "grad_norm": 0.1476656049489975,
      "learning_rate": 1.5104118747508207e-05,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8657332994043827,
      "num_tokens": 883532003.0,
      "step": 27683
    },
    {
      "entropy": 0.5192102687433362,
      "epoch": 2.547069884589508,
      "grad_norm": 0.1577121764421463,
      "learning_rate": 1.510105192136658e-05,
      "loss": 0.5053,
      "mean_token_accuracy": 0.836200375109911,
      "num_tokens": 883564370.0,
      "step": 27684
    },
    {
      "entropy": 0.5424979861127213,
      "epoch": 2.547161890960949,
      "grad_norm": 0.16129513084888458,
      "learning_rate": 1.5097985095224954e-05,
      "loss": 0.5328,
      "mean_token_accuracy": 0.8315893076360226,
      "num_tokens": 883596715.0,
      "step": 27685
    },
    {
      "entropy": 0.591575350612402,
      "epoch": 2.54725389733239,
      "grad_norm": 0.1719299554824829,
      "learning_rate": 1.5094918269083327e-05,
      "loss": 0.5798,
      "mean_token_accuracy": 0.8183831498026848,
      "num_tokens": 883627797.0,
      "step": 27686
    },
    {
      "entropy": 0.5281448570312932,
      "epoch": 2.5473459037038317,
      "grad_norm": 0.1575048714876175,
      "learning_rate": 1.5091851442941702e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8374450355768204,
      "num_tokens": 883660565.0,
      "step": 27687
    },
    {
      "entropy": 0.6146832313388586,
      "epoch": 2.547437910075273,
      "grad_norm": 0.17131373286247253,
      "learning_rate": 1.5088784616800075e-05,
      "loss": 0.6032,
      "mean_token_accuracy": 0.8088557980954647,
      "num_tokens": 883692088.0,
      "step": 27688
    },
    {
      "entropy": 0.4561497438699007,
      "epoch": 2.547529916446714,
      "grad_norm": 0.14994852244853973,
      "learning_rate": 1.508571779065845e-05,
      "loss": 0.4368,
      "mean_token_accuracy": 0.8623461201786995,
      "num_tokens": 883724455.0,
      "step": 27689
    },
    {
      "entropy": 0.5531552815809846,
      "epoch": 2.547621922818155,
      "grad_norm": 0.16235841810703278,
      "learning_rate": 1.5082650964516822e-05,
      "loss": 0.5388,
      "mean_token_accuracy": 0.833192240446806,
      "num_tokens": 883756668.0,
      "step": 27690
    },
    {
      "entropy": 0.5186484940350056,
      "epoch": 2.5477139291895963,
      "grad_norm": 0.16591286659240723,
      "learning_rate": 1.5079584138375197e-05,
      "loss": 0.5083,
      "mean_token_accuracy": 0.8351085782051086,
      "num_tokens": 883788017.0,
      "step": 27691
    },
    {
      "entropy": 0.5916822236031294,
      "epoch": 2.547805935561038,
      "grad_norm": 0.16518187522888184,
      "learning_rate": 1.507651731223357e-05,
      "loss": 0.5661,
      "mean_token_accuracy": 0.8214198350906372,
      "num_tokens": 883819919.0,
      "step": 27692
    },
    {
      "entropy": 0.5419466681778431,
      "epoch": 2.547897941932479,
      "grad_norm": 0.16373533010482788,
      "learning_rate": 1.5073450486091945e-05,
      "loss": 0.5289,
      "mean_token_accuracy": 0.834194328635931,
      "num_tokens": 883852262.0,
      "step": 27693
    },
    {
      "entropy": 0.4906820834148675,
      "epoch": 2.54798994830392,
      "grad_norm": 0.15690240263938904,
      "learning_rate": 1.507038365995032e-05,
      "loss": 0.477,
      "mean_token_accuracy": 0.8469250872731209,
      "num_tokens": 883883928.0,
      "step": 27694
    },
    {
      "entropy": 0.5429922584444284,
      "epoch": 2.5480819546753612,
      "grad_norm": 0.16006812453269958,
      "learning_rate": 1.5067316833808692e-05,
      "loss": 0.5248,
      "mean_token_accuracy": 0.8331958949565887,
      "num_tokens": 883916046.0,
      "step": 27695
    },
    {
      "entropy": 0.631509416969493,
      "epoch": 2.5481739610468024,
      "grad_norm": 0.17517608404159546,
      "learning_rate": 1.5064250007667067e-05,
      "loss": 0.6227,
      "mean_token_accuracy": 0.8064495548605919,
      "num_tokens": 883947346.0,
      "step": 27696
    },
    {
      "entropy": 0.5645633719395846,
      "epoch": 2.548265967418244,
      "grad_norm": 0.1647799164056778,
      "learning_rate": 1.506118318152544e-05,
      "loss": 0.5505,
      "mean_token_accuracy": 0.8247751630842686,
      "num_tokens": 883979454.0,
      "step": 27697
    },
    {
      "entropy": 0.5427409196272492,
      "epoch": 2.548357973789685,
      "grad_norm": 0.1623416244983673,
      "learning_rate": 1.5058116355383814e-05,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8292931355535984,
      "num_tokens": 884011598.0,
      "step": 27698
    },
    {
      "entropy": 0.4525714414194226,
      "epoch": 2.5484499801611262,
      "grad_norm": 0.15621042251586914,
      "learning_rate": 1.5055049529242187e-05,
      "loss": 0.4491,
      "mean_token_accuracy": 0.8547332771122456,
      "num_tokens": 884043513.0,
      "step": 27699
    },
    {
      "entropy": 0.41818296490237117,
      "epoch": 2.5485419865325674,
      "grad_norm": 0.14266522228717804,
      "learning_rate": 1.5051982703100562e-05,
      "loss": 0.403,
      "mean_token_accuracy": 0.8712203130126,
      "num_tokens": 884075059.0,
      "step": 27700
    },
    {
      "entropy": 0.5264016296714544,
      "epoch": 2.5486339929040085,
      "grad_norm": 0.16427545249462128,
      "learning_rate": 1.5048915876958935e-05,
      "loss": 0.5093,
      "mean_token_accuracy": 0.8354694098234177,
      "num_tokens": 884106461.0,
      "step": 27701
    },
    {
      "entropy": 0.5689514947589487,
      "epoch": 2.54872599927545,
      "grad_norm": 0.1717558056116104,
      "learning_rate": 1.504584905081731e-05,
      "loss": 0.5585,
      "mean_token_accuracy": 0.8228549472987652,
      "num_tokens": 884137588.0,
      "step": 27702
    },
    {
      "entropy": 0.4773866757750511,
      "epoch": 2.548818005646891,
      "grad_norm": 0.1550758183002472,
      "learning_rate": 1.5042782224675683e-05,
      "loss": 0.4668,
      "mean_token_accuracy": 0.8496535569429398,
      "num_tokens": 884169623.0,
      "step": 27703
    },
    {
      "entropy": 0.5115660452283919,
      "epoch": 2.5489100120183323,
      "grad_norm": 0.1610226184129715,
      "learning_rate": 1.5039715398534057e-05,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8397250771522522,
      "num_tokens": 884201698.0,
      "step": 27704
    },
    {
      "entropy": 0.5571389012038708,
      "epoch": 2.5490020183897735,
      "grad_norm": 0.1656779795885086,
      "learning_rate": 1.5036648572392432e-05,
      "loss": 0.5456,
      "mean_token_accuracy": 0.8253315761685371,
      "num_tokens": 884233635.0,
      "step": 27705
    },
    {
      "entropy": 0.5251869540661573,
      "epoch": 2.5490940247612146,
      "grad_norm": 0.15756836533546448,
      "learning_rate": 1.5033581746250805e-05,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8380354829132557,
      "num_tokens": 884265073.0,
      "step": 27706
    },
    {
      "entropy": 0.5591123346239328,
      "epoch": 2.549186031132656,
      "grad_norm": 0.16779546439647675,
      "learning_rate": 1.503051492010918e-05,
      "loss": 0.5536,
      "mean_token_accuracy": 0.8259154222905636,
      "num_tokens": 884297224.0,
      "step": 27707
    },
    {
      "entropy": 0.3905683499760926,
      "epoch": 2.5492780375040973,
      "grad_norm": 0.1450902819633484,
      "learning_rate": 1.5027448093967552e-05,
      "loss": 0.3731,
      "mean_token_accuracy": 0.880497831851244,
      "num_tokens": 884329697.0,
      "step": 27708
    },
    {
      "entropy": 0.5285640666261315,
      "epoch": 2.5493700438755384,
      "grad_norm": 0.15985360741615295,
      "learning_rate": 1.5024381267825927e-05,
      "loss": 0.5213,
      "mean_token_accuracy": 0.8351783119142056,
      "num_tokens": 884361705.0,
      "step": 27709
    },
    {
      "entropy": 0.527408305555582,
      "epoch": 2.5494620502469796,
      "grad_norm": 0.16175666451454163,
      "learning_rate": 1.50213144416843e-05,
      "loss": 0.528,
      "mean_token_accuracy": 0.833770576864481,
      "num_tokens": 884393984.0,
      "step": 27710
    },
    {
      "entropy": 0.43171686585992575,
      "epoch": 2.5495540566184207,
      "grad_norm": 0.14957678318023682,
      "learning_rate": 1.5018247615542675e-05,
      "loss": 0.4067,
      "mean_token_accuracy": 0.8668294847011566,
      "num_tokens": 884425135.0,
      "step": 27711
    },
    {
      "entropy": 0.5000221645459533,
      "epoch": 2.5496460629898623,
      "grad_norm": 0.15814247727394104,
      "learning_rate": 1.5015180789401048e-05,
      "loss": 0.4908,
      "mean_token_accuracy": 0.8432847820222378,
      "num_tokens": 884457052.0,
      "step": 27712
    },
    {
      "entropy": 0.5703133661299944,
      "epoch": 2.5497380693613034,
      "grad_norm": 0.16516362130641937,
      "learning_rate": 1.5012113963259426e-05,
      "loss": 0.5622,
      "mean_token_accuracy": 0.8231674917042255,
      "num_tokens": 884489053.0,
      "step": 27713
    },
    {
      "entropy": 0.5836523398756981,
      "epoch": 2.5498300757327446,
      "grad_norm": 0.16543829441070557,
      "learning_rate": 1.5009047137117799e-05,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8208726309239864,
      "num_tokens": 884521320.0,
      "step": 27714
    },
    {
      "entropy": 0.5715703098103404,
      "epoch": 2.5499220821041857,
      "grad_norm": 0.165090411901474,
      "learning_rate": 1.5005980310976173e-05,
      "loss": 0.5519,
      "mean_token_accuracy": 0.824004303663969,
      "num_tokens": 884552419.0,
      "step": 27715
    },
    {
      "entropy": 0.6489811353385448,
      "epoch": 2.550014088475627,
      "grad_norm": 0.17258404195308685,
      "learning_rate": 1.5002913484834546e-05,
      "loss": 0.6514,
      "mean_token_accuracy": 0.7966431975364685,
      "num_tokens": 884584950.0,
      "step": 27716
    },
    {
      "entropy": 0.5383151625283062,
      "epoch": 2.5501060948470684,
      "grad_norm": 0.1648622304201126,
      "learning_rate": 1.499984665869292e-05,
      "loss": 0.5347,
      "mean_token_accuracy": 0.8276506215333939,
      "num_tokens": 884616352.0,
      "step": 27717
    },
    {
      "entropy": 0.5217035785317421,
      "epoch": 2.5501981012185095,
      "grad_norm": 0.16287560760974884,
      "learning_rate": 1.4996779832551294e-05,
      "loss": 0.5073,
      "mean_token_accuracy": 0.8377256095409393,
      "num_tokens": 884647777.0,
      "step": 27718
    },
    {
      "entropy": 0.5155427530407906,
      "epoch": 2.5502901075899507,
      "grad_norm": 0.15150584280490875,
      "learning_rate": 1.4993713006409668e-05,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8443825840950012,
      "num_tokens": 884679144.0,
      "step": 27719
    },
    {
      "entropy": 0.49809969402849674,
      "epoch": 2.550382113961392,
      "grad_norm": 0.1596212238073349,
      "learning_rate": 1.4990646180268041e-05,
      "loss": 0.4914,
      "mean_token_accuracy": 0.8431305959820747,
      "num_tokens": 884711234.0,
      "step": 27720
    },
    {
      "entropy": 0.5329316803254187,
      "epoch": 2.550474120332833,
      "grad_norm": 0.16413739323616028,
      "learning_rate": 1.4987579354126416e-05,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8306985832750797,
      "num_tokens": 884743472.0,
      "step": 27721
    },
    {
      "entropy": 0.5332930502481759,
      "epoch": 2.5505661267042745,
      "grad_norm": 0.1614294797182083,
      "learning_rate": 1.4984512527984789e-05,
      "loss": 0.5098,
      "mean_token_accuracy": 0.8367998488247395,
      "num_tokens": 884774928.0,
      "step": 27722
    },
    {
      "entropy": 0.6395904310047626,
      "epoch": 2.5506581330757157,
      "grad_norm": 0.1737910658121109,
      "learning_rate": 1.4981445701843164e-05,
      "loss": 0.6248,
      "mean_token_accuracy": 0.804528683423996,
      "num_tokens": 884806889.0,
      "step": 27723
    },
    {
      "entropy": 0.6315637035295367,
      "epoch": 2.550750139447157,
      "grad_norm": 0.1734548658132553,
      "learning_rate": 1.4978378875701538e-05,
      "loss": 0.6094,
      "mean_token_accuracy": 0.8087538070976734,
      "num_tokens": 884838420.0,
      "step": 27724
    },
    {
      "entropy": 0.5086083812639117,
      "epoch": 2.550842145818598,
      "grad_norm": 0.1531190723180771,
      "learning_rate": 1.4975312049559911e-05,
      "loss": 0.4967,
      "mean_token_accuracy": 0.8417605087161064,
      "num_tokens": 884870951.0,
      "step": 27725
    },
    {
      "entropy": 0.5208493256941438,
      "epoch": 2.550934152190039,
      "grad_norm": 0.15686511993408203,
      "learning_rate": 1.4972245223418286e-05,
      "loss": 0.507,
      "mean_token_accuracy": 0.8368440419435501,
      "num_tokens": 884903323.0,
      "step": 27726
    },
    {
      "entropy": 0.5600712988525629,
      "epoch": 2.5510261585614806,
      "grad_norm": 0.1628686934709549,
      "learning_rate": 1.4969178397276659e-05,
      "loss": 0.538,
      "mean_token_accuracy": 0.8296018801629543,
      "num_tokens": 884934498.0,
      "step": 27727
    },
    {
      "entropy": 0.5323799178004265,
      "epoch": 2.5511181649329218,
      "grad_norm": 0.16600818932056427,
      "learning_rate": 1.4966111571135033e-05,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8329383879899979,
      "num_tokens": 884966453.0,
      "step": 27728
    },
    {
      "entropy": 0.5542880687862635,
      "epoch": 2.551210171304363,
      "grad_norm": 0.16105802357196808,
      "learning_rate": 1.4963044744993406e-05,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8317885994911194,
      "num_tokens": 884998010.0,
      "step": 27729
    },
    {
      "entropy": 0.5101466439664364,
      "epoch": 2.551302177675804,
      "grad_norm": 0.15874165296554565,
      "learning_rate": 1.4959977918851781e-05,
      "loss": 0.4939,
      "mean_token_accuracy": 0.8430039547383785,
      "num_tokens": 885030248.0,
      "step": 27730
    },
    {
      "entropy": 0.5649687722325325,
      "epoch": 2.551394184047245,
      "grad_norm": 0.16379885375499725,
      "learning_rate": 1.4956911092710154e-05,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8297393396496773,
      "num_tokens": 885062558.0,
      "step": 27731
    },
    {
      "entropy": 0.49666882539168,
      "epoch": 2.5514861904186867,
      "grad_norm": 0.15951892733573914,
      "learning_rate": 1.4953844266568529e-05,
      "loss": 0.4811,
      "mean_token_accuracy": 0.8440837822854519,
      "num_tokens": 885094011.0,
      "step": 27732
    },
    {
      "entropy": 0.5438062530010939,
      "epoch": 2.551578196790128,
      "grad_norm": 0.16852104663848877,
      "learning_rate": 1.4950777440426902e-05,
      "loss": 0.5398,
      "mean_token_accuracy": 0.8278269544243813,
      "num_tokens": 885126313.0,
      "step": 27733
    },
    {
      "entropy": 0.5579853085801005,
      "epoch": 2.551670203161569,
      "grad_norm": 0.1653009057044983,
      "learning_rate": 1.4947710614285276e-05,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8273465558886528,
      "num_tokens": 885158011.0,
      "step": 27734
    },
    {
      "entropy": 0.4353275840403512,
      "epoch": 2.55176220953301,
      "grad_norm": 0.15073804557323456,
      "learning_rate": 1.4944643788143651e-05,
      "loss": 0.4327,
      "mean_token_accuracy": 0.8640369474887848,
      "num_tokens": 885189971.0,
      "step": 27735
    },
    {
      "entropy": 0.4771358754951507,
      "epoch": 2.5518542159044513,
      "grad_norm": 0.15354929864406586,
      "learning_rate": 1.4941576962002024e-05,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8569183386862278,
      "num_tokens": 885221567.0,
      "step": 27736
    },
    {
      "entropy": 0.5536472499370575,
      "epoch": 2.551946222275893,
      "grad_norm": 0.17027287185192108,
      "learning_rate": 1.4938510135860399e-05,
      "loss": 0.5553,
      "mean_token_accuracy": 0.823649451136589,
      "num_tokens": 885253242.0,
      "step": 27737
    },
    {
      "entropy": 0.45511845126748085,
      "epoch": 2.552038228647334,
      "grad_norm": 0.15681378543376923,
      "learning_rate": 1.4935443309718771e-05,
      "loss": 0.4453,
      "mean_token_accuracy": 0.858614481985569,
      "num_tokens": 885285128.0,
      "step": 27738
    },
    {
      "entropy": 0.5428127218037844,
      "epoch": 2.552130235018775,
      "grad_norm": 0.16012707352638245,
      "learning_rate": 1.4932376483577146e-05,
      "loss": 0.5372,
      "mean_token_accuracy": 0.8290605284273624,
      "num_tokens": 885317158.0,
      "step": 27739
    },
    {
      "entropy": 0.5762438047677279,
      "epoch": 2.5522222413902163,
      "grad_norm": 0.16844969987869263,
      "learning_rate": 1.4929309657435519e-05,
      "loss": 0.5703,
      "mean_token_accuracy": 0.8189755529165268,
      "num_tokens": 885349034.0,
      "step": 27740
    },
    {
      "entropy": 0.5175189767032862,
      "epoch": 2.5523142477616574,
      "grad_norm": 0.16157466173171997,
      "learning_rate": 1.4926242831293894e-05,
      "loss": 0.5059,
      "mean_token_accuracy": 0.8412058502435684,
      "num_tokens": 885380466.0,
      "step": 27741
    },
    {
      "entropy": 0.5354978828690946,
      "epoch": 2.552406254133099,
      "grad_norm": 0.16393831372261047,
      "learning_rate": 1.4923176005152267e-05,
      "loss": 0.5203,
      "mean_token_accuracy": 0.833407212048769,
      "num_tokens": 885411714.0,
      "step": 27742
    },
    {
      "entropy": 0.49221176374703646,
      "epoch": 2.55249826050454,
      "grad_norm": 0.1597551554441452,
      "learning_rate": 1.4920109179010645e-05,
      "loss": 0.484,
      "mean_token_accuracy": 0.8447730205953121,
      "num_tokens": 885444101.0,
      "step": 27743
    },
    {
      "entropy": 0.5830982020124793,
      "epoch": 2.5525902668759812,
      "grad_norm": 0.16880843043327332,
      "learning_rate": 1.4917042352869018e-05,
      "loss": 0.5716,
      "mean_token_accuracy": 0.820639681071043,
      "num_tokens": 885475644.0,
      "step": 27744
    },
    {
      "entropy": 0.5684425716754049,
      "epoch": 2.5526822732474224,
      "grad_norm": 0.16082458198070526,
      "learning_rate": 1.4913975526727392e-05,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8259568624198437,
      "num_tokens": 885507924.0,
      "step": 27745
    },
    {
      "entropy": 0.5679094363003969,
      "epoch": 2.5527742796188635,
      "grad_norm": 0.16368436813354492,
      "learning_rate": 1.4910908700585765e-05,
      "loss": 0.5577,
      "mean_token_accuracy": 0.8234690316021442,
      "num_tokens": 885540526.0,
      "step": 27746
    },
    {
      "entropy": 0.3927176110446453,
      "epoch": 2.552866285990305,
      "grad_norm": 0.13851258158683777,
      "learning_rate": 1.490784187444414e-05,
      "loss": 0.3758,
      "mean_token_accuracy": 0.8781523816287518,
      "num_tokens": 885572739.0,
      "step": 27747
    },
    {
      "entropy": 0.526117155328393,
      "epoch": 2.552958292361746,
      "grad_norm": 0.1678411364555359,
      "learning_rate": 1.4904775048302513e-05,
      "loss": 0.5204,
      "mean_token_accuracy": 0.8297394663095474,
      "num_tokens": 885604769.0,
      "step": 27748
    },
    {
      "entropy": 0.5353841832838953,
      "epoch": 2.5530502987331873,
      "grad_norm": 0.16350361704826355,
      "learning_rate": 1.4901708222160888e-05,
      "loss": 0.5242,
      "mean_token_accuracy": 0.8354743905365467,
      "num_tokens": 885636537.0,
      "step": 27749
    },
    {
      "entropy": 0.528598939999938,
      "epoch": 2.5531423051046285,
      "grad_norm": 0.1606869101524353,
      "learning_rate": 1.489864139601926e-05,
      "loss": 0.5181,
      "mean_token_accuracy": 0.8372217267751694,
      "num_tokens": 885668559.0,
      "step": 27750
    },
    {
      "entropy": 0.5287343803793192,
      "epoch": 2.5532343114760696,
      "grad_norm": 0.16111688315868378,
      "learning_rate": 1.4895574569877635e-05,
      "loss": 0.5114,
      "mean_token_accuracy": 0.8357821218669415,
      "num_tokens": 885700006.0,
      "step": 27751
    },
    {
      "entropy": 0.5696492679417133,
      "epoch": 2.553326317847511,
      "grad_norm": 0.16724130511283875,
      "learning_rate": 1.4892507743736008e-05,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8267586305737495,
      "num_tokens": 885732251.0,
      "step": 27752
    },
    {
      "entropy": 0.47766492620576173,
      "epoch": 2.5534183242189523,
      "grad_norm": 0.1576155573129654,
      "learning_rate": 1.4889440917594383e-05,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8500845730304718,
      "num_tokens": 885764258.0,
      "step": 27753
    },
    {
      "entropy": 0.6552340369671583,
      "epoch": 2.5535103305903935,
      "grad_norm": 0.17102935910224915,
      "learning_rate": 1.4886374091452757e-05,
      "loss": 0.6346,
      "mean_token_accuracy": 0.80001625046134,
      "num_tokens": 885796179.0,
      "step": 27754
    },
    {
      "entropy": 0.5267332335934043,
      "epoch": 2.5536023369618346,
      "grad_norm": 0.1568525731563568,
      "learning_rate": 1.488330726531113e-05,
      "loss": 0.5142,
      "mean_token_accuracy": 0.8362331539392471,
      "num_tokens": 885828678.0,
      "step": 27755
    },
    {
      "entropy": 0.5306254178285599,
      "epoch": 2.5536943433332757,
      "grad_norm": 0.159245565533638,
      "learning_rate": 1.4880240439169505e-05,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8346015587449074,
      "num_tokens": 885861103.0,
      "step": 27756
    },
    {
      "entropy": 0.4968811646103859,
      "epoch": 2.5537863497047173,
      "grad_norm": 0.15916156768798828,
      "learning_rate": 1.4877173613027878e-05,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8442533314228058,
      "num_tokens": 885892992.0,
      "step": 27757
    },
    {
      "entropy": 0.5249291025102139,
      "epoch": 2.5538783560761584,
      "grad_norm": 0.1648005098104477,
      "learning_rate": 1.4874106786886253e-05,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8330894224345684,
      "num_tokens": 885924737.0,
      "step": 27758
    },
    {
      "entropy": 0.5890296436846256,
      "epoch": 2.5539703624475996,
      "grad_norm": 0.16680529713630676,
      "learning_rate": 1.4871039960744626e-05,
      "loss": 0.5777,
      "mean_token_accuracy": 0.8210273757576942,
      "num_tokens": 885957359.0,
      "step": 27759
    },
    {
      "entropy": 0.4557779091410339,
      "epoch": 2.5540623688190407,
      "grad_norm": 0.15289485454559326,
      "learning_rate": 1.4867973134603e-05,
      "loss": 0.4452,
      "mean_token_accuracy": 0.8552682138979435,
      "num_tokens": 885988897.0,
      "step": 27760
    },
    {
      "entropy": 0.5904393931850791,
      "epoch": 2.554154375190482,
      "grad_norm": 0.17368769645690918,
      "learning_rate": 1.4864906308461373e-05,
      "loss": 0.5845,
      "mean_token_accuracy": 0.8165866881608963,
      "num_tokens": 886020401.0,
      "step": 27761
    },
    {
      "entropy": 0.49112062714993954,
      "epoch": 2.5542463815619234,
      "grad_norm": 0.15660755336284637,
      "learning_rate": 1.4861839482319748e-05,
      "loss": 0.4898,
      "mean_token_accuracy": 0.8450849652290344,
      "num_tokens": 886052661.0,
      "step": 27762
    },
    {
      "entropy": 0.4900333760306239,
      "epoch": 2.5543383879333645,
      "grad_norm": 0.16091032326221466,
      "learning_rate": 1.485877265617812e-05,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8469375297427177,
      "num_tokens": 886084173.0,
      "step": 27763
    },
    {
      "entropy": 0.518279604613781,
      "epoch": 2.5544303943048057,
      "grad_norm": 0.16013164818286896,
      "learning_rate": 1.4855705830036495e-05,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8421951457858086,
      "num_tokens": 886116659.0,
      "step": 27764
    },
    {
      "entropy": 0.45587009843438864,
      "epoch": 2.554522400676247,
      "grad_norm": 0.15145806968212128,
      "learning_rate": 1.485263900389487e-05,
      "loss": 0.4464,
      "mean_token_accuracy": 0.8584715984761715,
      "num_tokens": 886149152.0,
      "step": 27765
    },
    {
      "entropy": 0.4309831289574504,
      "epoch": 2.554614407047688,
      "grad_norm": 0.1470658779144287,
      "learning_rate": 1.4849572177753243e-05,
      "loss": 0.4192,
      "mean_token_accuracy": 0.8645019233226776,
      "num_tokens": 886180982.0,
      "step": 27766
    },
    {
      "entropy": 0.5371537450700998,
      "epoch": 2.5547064134191295,
      "grad_norm": 0.16050411760807037,
      "learning_rate": 1.4846505351611618e-05,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8303024247288704,
      "num_tokens": 886212837.0,
      "step": 27767
    },
    {
      "entropy": 0.5060637947171926,
      "epoch": 2.5547984197905707,
      "grad_norm": 0.16177740693092346,
      "learning_rate": 1.484343852546999e-05,
      "loss": 0.4951,
      "mean_token_accuracy": 0.8402470648288727,
      "num_tokens": 886244963.0,
      "step": 27768
    },
    {
      "entropy": 0.5621495940722525,
      "epoch": 2.554890426162012,
      "grad_norm": 0.16222575306892395,
      "learning_rate": 1.4840371699328365e-05,
      "loss": 0.5518,
      "mean_token_accuracy": 0.8256402276456356,
      "num_tokens": 886276895.0,
      "step": 27769
    },
    {
      "entropy": 0.5369127169251442,
      "epoch": 2.554982432533453,
      "grad_norm": 0.16498851776123047,
      "learning_rate": 1.4837304873186738e-05,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8361133709549904,
      "num_tokens": 886308326.0,
      "step": 27770
    },
    {
      "entropy": 0.547296442091465,
      "epoch": 2.555074438904894,
      "grad_norm": 0.16616180539131165,
      "learning_rate": 1.4834238047045113e-05,
      "loss": 0.5294,
      "mean_token_accuracy": 0.8330785259604454,
      "num_tokens": 886339538.0,
      "step": 27771
    },
    {
      "entropy": 0.46630820352584124,
      "epoch": 2.5551664452763356,
      "grad_norm": 0.14981180429458618,
      "learning_rate": 1.4831171220903486e-05,
      "loss": 0.4553,
      "mean_token_accuracy": 0.8542746528983116,
      "num_tokens": 886371894.0,
      "step": 27772
    },
    {
      "entropy": 0.5252022780478001,
      "epoch": 2.5552584516477768,
      "grad_norm": 0.156199648976326,
      "learning_rate": 1.482810439476186e-05,
      "loss": 0.507,
      "mean_token_accuracy": 0.8368638753890991,
      "num_tokens": 886403903.0,
      "step": 27773
    },
    {
      "entropy": 0.4671247978694737,
      "epoch": 2.555350458019218,
      "grad_norm": 0.15423937141895294,
      "learning_rate": 1.4825037568620237e-05,
      "loss": 0.459,
      "mean_token_accuracy": 0.8578252010047436,
      "num_tokens": 886435021.0,
      "step": 27774
    },
    {
      "entropy": 0.5901870913803577,
      "epoch": 2.555442464390659,
      "grad_norm": 0.1700919270515442,
      "learning_rate": 1.4821970742478611e-05,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8190192170441151,
      "num_tokens": 886467190.0,
      "step": 27775
    },
    {
      "entropy": 0.605309035629034,
      "epoch": 2.5555344707621,
      "grad_norm": 0.17214706540107727,
      "learning_rate": 1.4818903916336984e-05,
      "loss": 0.5853,
      "mean_token_accuracy": 0.8155617974698544,
      "num_tokens": 886497917.0,
      "step": 27776
    },
    {
      "entropy": 0.4988372949883342,
      "epoch": 2.5556264771335417,
      "grad_norm": 0.15481148660182953,
      "learning_rate": 1.4815837090195359e-05,
      "loss": 0.4935,
      "mean_token_accuracy": 0.8411728590726852,
      "num_tokens": 886530474.0,
      "step": 27777
    },
    {
      "entropy": 0.4624803503975272,
      "epoch": 2.555718483504983,
      "grad_norm": 0.15218305587768555,
      "learning_rate": 1.4812770264053732e-05,
      "loss": 0.4478,
      "mean_token_accuracy": 0.8555657751858234,
      "num_tokens": 886561819.0,
      "step": 27778
    },
    {
      "entropy": 0.5440258197486401,
      "epoch": 2.555810489876424,
      "grad_norm": 0.1658957600593567,
      "learning_rate": 1.4809703437912107e-05,
      "loss": 0.5275,
      "mean_token_accuracy": 0.8299906067550182,
      "num_tokens": 886593440.0,
      "step": 27779
    },
    {
      "entropy": 0.5813372144475579,
      "epoch": 2.555902496247865,
      "grad_norm": 0.17095547914505005,
      "learning_rate": 1.480663661177048e-05,
      "loss": 0.5691,
      "mean_token_accuracy": 0.8190394267439842,
      "num_tokens": 886624753.0,
      "step": 27780
    },
    {
      "entropy": 0.5823286827653646,
      "epoch": 2.5559945026193063,
      "grad_norm": 0.16710956394672394,
      "learning_rate": 1.4803569785628854e-05,
      "loss": 0.5712,
      "mean_token_accuracy": 0.8201703876256943,
      "num_tokens": 886656023.0,
      "step": 27781
    },
    {
      "entropy": 0.5901878038421273,
      "epoch": 2.556086508990748,
      "grad_norm": 0.16583047807216644,
      "learning_rate": 1.4800502959487227e-05,
      "loss": 0.5743,
      "mean_token_accuracy": 0.8203124217689037,
      "num_tokens": 886687570.0,
      "step": 27782
    },
    {
      "entropy": 0.5283182030543685,
      "epoch": 2.556178515362189,
      "grad_norm": 0.15974834561347961,
      "learning_rate": 1.4797436133345602e-05,
      "loss": 0.5198,
      "mean_token_accuracy": 0.8378865867853165,
      "num_tokens": 886720113.0,
      "step": 27783
    },
    {
      "entropy": 0.5651879943907261,
      "epoch": 2.55627052173363,
      "grad_norm": 0.16526579856872559,
      "learning_rate": 1.4794369307203976e-05,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8277808092534542,
      "num_tokens": 886752166.0,
      "step": 27784
    },
    {
      "entropy": 0.4477510657161474,
      "epoch": 2.5563625281050713,
      "grad_norm": 0.15678711235523224,
      "learning_rate": 1.479130248106235e-05,
      "loss": 0.4345,
      "mean_token_accuracy": 0.8629728741943836,
      "num_tokens": 886783576.0,
      "step": 27785
    },
    {
      "entropy": 0.5726585905067623,
      "epoch": 2.5564545344765124,
      "grad_norm": 0.16683216392993927,
      "learning_rate": 1.4788235654920724e-05,
      "loss": 0.553,
      "mean_token_accuracy": 0.8239772953093052,
      "num_tokens": 886815665.0,
      "step": 27786
    },
    {
      "entropy": 0.5099908832926303,
      "epoch": 2.556546540847954,
      "grad_norm": 0.15408439934253693,
      "learning_rate": 1.4785168828779097e-05,
      "loss": 0.4938,
      "mean_token_accuracy": 0.8418035060167313,
      "num_tokens": 886847972.0,
      "step": 27787
    },
    {
      "entropy": 0.5157418835442513,
      "epoch": 2.556638547219395,
      "grad_norm": 0.16059982776641846,
      "learning_rate": 1.4782102002637472e-05,
      "loss": 0.5004,
      "mean_token_accuracy": 0.8414102271199226,
      "num_tokens": 886879427.0,
      "step": 27788
    },
    {
      "entropy": 0.5860890292096883,
      "epoch": 2.5567305535908362,
      "grad_norm": 0.167585089802742,
      "learning_rate": 1.4779035176495845e-05,
      "loss": 0.57,
      "mean_token_accuracy": 0.8204044699668884,
      "num_tokens": 886910842.0,
      "step": 27789
    },
    {
      "entropy": 0.512324913404882,
      "epoch": 2.5568225599622774,
      "grad_norm": 0.15909262001514435,
      "learning_rate": 1.477596835035422e-05,
      "loss": 0.4904,
      "mean_token_accuracy": 0.843363881111145,
      "num_tokens": 886942383.0,
      "step": 27790
    },
    {
      "entropy": 0.5283442046493292,
      "epoch": 2.5569145663337185,
      "grad_norm": 0.1602851003408432,
      "learning_rate": 1.4772901524212592e-05,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8345499001443386,
      "num_tokens": 886974212.0,
      "step": 27791
    },
    {
      "entropy": 0.3772775400429964,
      "epoch": 2.55700657270516,
      "grad_norm": 0.13854117691516876,
      "learning_rate": 1.4769834698070967e-05,
      "loss": 0.3712,
      "mean_token_accuracy": 0.8810323588550091,
      "num_tokens": 887006839.0,
      "step": 27792
    },
    {
      "entropy": 0.4803765807300806,
      "epoch": 2.557098579076601,
      "grad_norm": 0.15746766328811646,
      "learning_rate": 1.476676787192934e-05,
      "loss": 0.477,
      "mean_token_accuracy": 0.8439763188362122,
      "num_tokens": 887038921.0,
      "step": 27793
    },
    {
      "entropy": 0.4840179421007633,
      "epoch": 2.5571905854480423,
      "grad_norm": 0.15849323570728302,
      "learning_rate": 1.4763701045787714e-05,
      "loss": 0.4828,
      "mean_token_accuracy": 0.8470689579844475,
      "num_tokens": 887070924.0,
      "step": 27794
    },
    {
      "entropy": 0.5456590093672276,
      "epoch": 2.5572825918194835,
      "grad_norm": 0.16252852976322174,
      "learning_rate": 1.4760634219646089e-05,
      "loss": 0.54,
      "mean_token_accuracy": 0.8292318023741245,
      "num_tokens": 887103692.0,
      "step": 27795
    },
    {
      "entropy": 0.5617377441376448,
      "epoch": 2.5573745981909246,
      "grad_norm": 0.16844724118709564,
      "learning_rate": 1.4757567393504462e-05,
      "loss": 0.5605,
      "mean_token_accuracy": 0.823931660503149,
      "num_tokens": 887134893.0,
      "step": 27796
    },
    {
      "entropy": 0.44587725459132344,
      "epoch": 2.557466604562366,
      "grad_norm": 0.14778925478458405,
      "learning_rate": 1.4754500567362837e-05,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8617678955197334,
      "num_tokens": 887167264.0,
      "step": 27797
    },
    {
      "entropy": 0.5428641214966774,
      "epoch": 2.5575586109338073,
      "grad_norm": 0.15973928570747375,
      "learning_rate": 1.475143374122121e-05,
      "loss": 0.5252,
      "mean_token_accuracy": 0.8323867619037628,
      "num_tokens": 887199669.0,
      "step": 27798
    },
    {
      "entropy": 0.4899164019152522,
      "epoch": 2.5576506173052485,
      "grad_norm": 0.15903529524803162,
      "learning_rate": 1.4748366915079584e-05,
      "loss": 0.4732,
      "mean_token_accuracy": 0.846773486584425,
      "num_tokens": 887231325.0,
      "step": 27799
    },
    {
      "entropy": 0.46498246071860194,
      "epoch": 2.5577426236766896,
      "grad_norm": 0.15194833278656006,
      "learning_rate": 1.4745300088937957e-05,
      "loss": 0.4564,
      "mean_token_accuracy": 0.8587567061185837,
      "num_tokens": 887264093.0,
      "step": 27800
    },
    {
      "entropy": 0.5373880341649055,
      "epoch": 2.5578346300481307,
      "grad_norm": 0.15766221284866333,
      "learning_rate": 1.4742233262796332e-05,
      "loss": 0.5225,
      "mean_token_accuracy": 0.8322029113769531,
      "num_tokens": 887295570.0,
      "step": 27801
    },
    {
      "entropy": 0.5575330965220928,
      "epoch": 2.5579266364195723,
      "grad_norm": 0.16105259954929352,
      "learning_rate": 1.4739166436654705e-05,
      "loss": 0.5471,
      "mean_token_accuracy": 0.8266910724341869,
      "num_tokens": 887327071.0,
      "step": 27802
    },
    {
      "entropy": 0.49906399939209223,
      "epoch": 2.5580186427910134,
      "grad_norm": 0.1539108157157898,
      "learning_rate": 1.473609961051308e-05,
      "loss": 0.4892,
      "mean_token_accuracy": 0.845823023468256,
      "num_tokens": 887358741.0,
      "step": 27803
    },
    {
      "entropy": 0.5466690696775913,
      "epoch": 2.5581106491624546,
      "grad_norm": 0.16603243350982666,
      "learning_rate": 1.4733032784371456e-05,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8276195861399174,
      "num_tokens": 887389424.0,
      "step": 27804
    },
    {
      "entropy": 0.43679719883948565,
      "epoch": 2.5582026555338957,
      "grad_norm": 0.15233413875102997,
      "learning_rate": 1.472996595822983e-05,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8649149462580681,
      "num_tokens": 887421170.0,
      "step": 27805
    },
    {
      "entropy": 0.601218743249774,
      "epoch": 2.558294661905337,
      "grad_norm": 0.16703346371650696,
      "learning_rate": 1.4726899132088203e-05,
      "loss": 0.5775,
      "mean_token_accuracy": 0.817918311804533,
      "num_tokens": 887453385.0,
      "step": 27806
    },
    {
      "entropy": 0.5322594717144966,
      "epoch": 2.5583866682767784,
      "grad_norm": 0.16671733558177948,
      "learning_rate": 1.4723832305946578e-05,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8326730541884899,
      "num_tokens": 887484820.0,
      "step": 27807
    },
    {
      "entropy": 0.5309737483039498,
      "epoch": 2.5584786746482195,
      "grad_norm": 0.16037388145923615,
      "learning_rate": 1.4720765479804951e-05,
      "loss": 0.5196,
      "mean_token_accuracy": 0.8378827162086964,
      "num_tokens": 887517274.0,
      "step": 27808
    },
    {
      "entropy": 0.6406680773943663,
      "epoch": 2.5585706810196607,
      "grad_norm": 0.17626477777957916,
      "learning_rate": 1.4717698653663326e-05,
      "loss": 0.6286,
      "mean_token_accuracy": 0.8043615967035294,
      "num_tokens": 887549030.0,
      "step": 27809
    },
    {
      "entropy": 0.428845944814384,
      "epoch": 2.558662687391102,
      "grad_norm": 0.14507938921451569,
      "learning_rate": 1.4714631827521699e-05,
      "loss": 0.41,
      "mean_token_accuracy": 0.8665057048201561,
      "num_tokens": 887581057.0,
      "step": 27810
    },
    {
      "entropy": 0.5006661741062999,
      "epoch": 2.558754693762543,
      "grad_norm": 0.16100528836250305,
      "learning_rate": 1.4711565001380073e-05,
      "loss": 0.487,
      "mean_token_accuracy": 0.8450873605906963,
      "num_tokens": 887613177.0,
      "step": 27811
    },
    {
      "entropy": 0.5714829489588737,
      "epoch": 2.5588467001339845,
      "grad_norm": 0.16638785600662231,
      "learning_rate": 1.4708498175238446e-05,
      "loss": 0.5657,
      "mean_token_accuracy": 0.8210067749023438,
      "num_tokens": 887645696.0,
      "step": 27812
    },
    {
      "entropy": 0.5064878668636084,
      "epoch": 2.5589387065054257,
      "grad_norm": 0.15981778502464294,
      "learning_rate": 1.4705431349096821e-05,
      "loss": 0.4971,
      "mean_token_accuracy": 0.8393071852624416,
      "num_tokens": 887678086.0,
      "step": 27813
    },
    {
      "entropy": 0.5264242519624531,
      "epoch": 2.559030712876867,
      "grad_norm": 0.15909576416015625,
      "learning_rate": 1.4702364522955196e-05,
      "loss": 0.517,
      "mean_token_accuracy": 0.836851567029953,
      "num_tokens": 887710481.0,
      "step": 27814
    },
    {
      "entropy": 0.538302956148982,
      "epoch": 2.559122719248308,
      "grad_norm": 0.16370894014835358,
      "learning_rate": 1.4699297696813569e-05,
      "loss": 0.5168,
      "mean_token_accuracy": 0.8348500169813633,
      "num_tokens": 887741994.0,
      "step": 27815
    },
    {
      "entropy": 0.5216371230781078,
      "epoch": 2.559214725619749,
      "grad_norm": 0.1634582281112671,
      "learning_rate": 1.4696230870671943e-05,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8379722572863102,
      "num_tokens": 887773801.0,
      "step": 27816
    },
    {
      "entropy": 0.5115515114739537,
      "epoch": 2.5593067319911906,
      "grad_norm": 0.16425460577011108,
      "learning_rate": 1.4693164044530316e-05,
      "loss": 0.5021,
      "mean_token_accuracy": 0.8375984579324722,
      "num_tokens": 887805498.0,
      "step": 27817
    },
    {
      "entropy": 0.5274952184408903,
      "epoch": 2.5593987383626318,
      "grad_norm": 0.1602727472782135,
      "learning_rate": 1.469009721838869e-05,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8355798609554768,
      "num_tokens": 887837532.0,
      "step": 27818
    },
    {
      "entropy": 0.5182161012198776,
      "epoch": 2.559490744734073,
      "grad_norm": 0.15567757189273834,
      "learning_rate": 1.4687030392247064e-05,
      "loss": 0.5025,
      "mean_token_accuracy": 0.839700497686863,
      "num_tokens": 887869195.0,
      "step": 27819
    },
    {
      "entropy": 0.48612394789233804,
      "epoch": 2.559582751105514,
      "grad_norm": 0.15481944382190704,
      "learning_rate": 1.4683963566105438e-05,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8494624197483063,
      "num_tokens": 887901684.0,
      "step": 27820
    },
    {
      "entropy": 0.5048131123185158,
      "epoch": 2.559674757476955,
      "grad_norm": 0.15967464447021484,
      "learning_rate": 1.4680896739963811e-05,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8403472453355789,
      "num_tokens": 887934000.0,
      "step": 27821
    },
    {
      "entropy": 0.5459638591855764,
      "epoch": 2.5597667638483967,
      "grad_norm": 0.15926826000213623,
      "learning_rate": 1.4677829913822186e-05,
      "loss": 0.5377,
      "mean_token_accuracy": 0.828897912055254,
      "num_tokens": 887966464.0,
      "step": 27822
    },
    {
      "entropy": 0.5767986536957324,
      "epoch": 2.559858770219838,
      "grad_norm": 0.17316672205924988,
      "learning_rate": 1.4674763087680559e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8184676803648472,
      "num_tokens": 887997112.0,
      "step": 27823
    },
    {
      "entropy": 0.5808033626526594,
      "epoch": 2.559950776591279,
      "grad_norm": 0.17102862894535065,
      "learning_rate": 1.4671696261538934e-05,
      "loss": 0.5729,
      "mean_token_accuracy": 0.8221135474741459,
      "num_tokens": 888028090.0,
      "step": 27824
    },
    {
      "entropy": 0.5417574988678098,
      "epoch": 2.56004278296272,
      "grad_norm": 0.16101601719856262,
      "learning_rate": 1.4668629435397308e-05,
      "loss": 0.5296,
      "mean_token_accuracy": 0.833260428160429,
      "num_tokens": 888060192.0,
      "step": 27825
    },
    {
      "entropy": 0.5546387368813157,
      "epoch": 2.5601347893341613,
      "grad_norm": 0.15997368097305298,
      "learning_rate": 1.4665562609255681e-05,
      "loss": 0.5367,
      "mean_token_accuracy": 0.8328613787889481,
      "num_tokens": 888092503.0,
      "step": 27826
    },
    {
      "entropy": 0.3551357811084017,
      "epoch": 2.560226795705603,
      "grad_norm": 0.1365368366241455,
      "learning_rate": 1.4662495783114056e-05,
      "loss": 0.3504,
      "mean_token_accuracy": 0.8876235969364643,
      "num_tokens": 888125057.0,
      "step": 27827
    },
    {
      "entropy": 0.5122458040714264,
      "epoch": 2.560318802077044,
      "grad_norm": 0.16444221138954163,
      "learning_rate": 1.4659428956972429e-05,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8377237878739834,
      "num_tokens": 888157636.0,
      "step": 27828
    },
    {
      "entropy": 0.6029087835922837,
      "epoch": 2.560410808448485,
      "grad_norm": 0.1636374443769455,
      "learning_rate": 1.4656362130830803e-05,
      "loss": 0.5905,
      "mean_token_accuracy": 0.8145636729896069,
      "num_tokens": 888190244.0,
      "step": 27829
    },
    {
      "entropy": 0.5383347311289981,
      "epoch": 2.5605028148199263,
      "grad_norm": 0.1603499948978424,
      "learning_rate": 1.4653295304689176e-05,
      "loss": 0.513,
      "mean_token_accuracy": 0.8372562862932682,
      "num_tokens": 888222258.0,
      "step": 27830
    },
    {
      "entropy": 0.47993308538571,
      "epoch": 2.5605948211913674,
      "grad_norm": 0.15628361701965332,
      "learning_rate": 1.4650228478547551e-05,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8521546870470047,
      "num_tokens": 888254549.0,
      "step": 27831
    },
    {
      "entropy": 0.582900358363986,
      "epoch": 2.560686827562809,
      "grad_norm": 0.17308922111988068,
      "learning_rate": 1.4647161652405924e-05,
      "loss": 0.5771,
      "mean_token_accuracy": 0.8182068690657616,
      "num_tokens": 888285892.0,
      "step": 27832
    },
    {
      "entropy": 0.5466966489329934,
      "epoch": 2.56077883393425,
      "grad_norm": 0.16441458463668823,
      "learning_rate": 1.4644094826264299e-05,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8298460021615028,
      "num_tokens": 888317543.0,
      "step": 27833
    },
    {
      "entropy": 0.5271537695080042,
      "epoch": 2.5608708403056912,
      "grad_norm": 0.16171178221702576,
      "learning_rate": 1.4641028000122672e-05,
      "loss": 0.5234,
      "mean_token_accuracy": 0.835220854729414,
      "num_tokens": 888349469.0,
      "step": 27834
    },
    {
      "entropy": 0.6165052093565464,
      "epoch": 2.5609628466771324,
      "grad_norm": 0.1666002869606018,
      "learning_rate": 1.463796117398105e-05,
      "loss": 0.6053,
      "mean_token_accuracy": 0.8115911893546581,
      "num_tokens": 888381728.0,
      "step": 27835
    },
    {
      "entropy": 0.4981239391490817,
      "epoch": 2.5610548530485735,
      "grad_norm": 0.15907466411590576,
      "learning_rate": 1.4634894347839423e-05,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8433601297438145,
      "num_tokens": 888412535.0,
      "step": 27836
    },
    {
      "entropy": 0.5324130854569376,
      "epoch": 2.561146859420015,
      "grad_norm": 0.16032002866268158,
      "learning_rate": 1.4631827521697797e-05,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8331669494509697,
      "num_tokens": 888444367.0,
      "step": 27837
    },
    {
      "entropy": 0.5325604192912579,
      "epoch": 2.561238865791456,
      "grad_norm": 0.16300001740455627,
      "learning_rate": 1.462876069555617e-05,
      "loss": 0.5227,
      "mean_token_accuracy": 0.83275056630373,
      "num_tokens": 888476802.0,
      "step": 27838
    },
    {
      "entropy": 0.5162738936487585,
      "epoch": 2.5613308721628973,
      "grad_norm": 0.1602936089038849,
      "learning_rate": 1.4625693869414545e-05,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8406838178634644,
      "num_tokens": 888508901.0,
      "step": 27839
    },
    {
      "entropy": 0.494780297158286,
      "epoch": 2.5614228785343385,
      "grad_norm": 0.15688611567020416,
      "learning_rate": 1.4622627043272918e-05,
      "loss": 0.4811,
      "mean_token_accuracy": 0.8422750048339367,
      "num_tokens": 888539960.0,
      "step": 27840
    },
    {
      "entropy": 0.6469319062307477,
      "epoch": 2.5615148849057796,
      "grad_norm": 0.1704064905643463,
      "learning_rate": 1.4619560217131292e-05,
      "loss": 0.6282,
      "mean_token_accuracy": 0.8037957437336445,
      "num_tokens": 888571950.0,
      "step": 27841
    },
    {
      "entropy": 0.5351004488766193,
      "epoch": 2.561606891277221,
      "grad_norm": 0.16993722319602966,
      "learning_rate": 1.4616493390989665e-05,
      "loss": 0.5267,
      "mean_token_accuracy": 0.8335268460214138,
      "num_tokens": 888603296.0,
      "step": 27842
    },
    {
      "entropy": 0.47536539332941175,
      "epoch": 2.5616988976486623,
      "grad_norm": 0.14958961308002472,
      "learning_rate": 1.461342656484804e-05,
      "loss": 0.4491,
      "mean_token_accuracy": 0.8554249182343483,
      "num_tokens": 888635100.0,
      "step": 27843
    },
    {
      "entropy": 0.5117098689079285,
      "epoch": 2.5617909040201035,
      "grad_norm": 0.15677207708358765,
      "learning_rate": 1.4610359738706415e-05,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8455525562167168,
      "num_tokens": 888667366.0,
      "step": 27844
    },
    {
      "entropy": 0.4725087732076645,
      "epoch": 2.5618829103915446,
      "grad_norm": 0.16034062206745148,
      "learning_rate": 1.4607292912564788e-05,
      "loss": 0.4699,
      "mean_token_accuracy": 0.8521531634032726,
      "num_tokens": 888698870.0,
      "step": 27845
    },
    {
      "entropy": 0.5575875760987401,
      "epoch": 2.5619749167629857,
      "grad_norm": 0.16378040611743927,
      "learning_rate": 1.4604226086423162e-05,
      "loss": 0.5445,
      "mean_token_accuracy": 0.8263400606811047,
      "num_tokens": 888731006.0,
      "step": 27846
    },
    {
      "entropy": 0.4786405535414815,
      "epoch": 2.5620669231344273,
      "grad_norm": 0.15467435121536255,
      "learning_rate": 1.4601159260281535e-05,
      "loss": 0.4593,
      "mean_token_accuracy": 0.8527017347514629,
      "num_tokens": 888762814.0,
      "step": 27847
    },
    {
      "entropy": 0.5489912610501051,
      "epoch": 2.5621589295058684,
      "grad_norm": 0.16342693567276,
      "learning_rate": 1.459809243413991e-05,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8314567543566227,
      "num_tokens": 888794501.0,
      "step": 27848
    },
    {
      "entropy": 0.5573507295921445,
      "epoch": 2.5622509358773096,
      "grad_norm": 0.1655006855726242,
      "learning_rate": 1.4595025607998283e-05,
      "loss": 0.5472,
      "mean_token_accuracy": 0.8278288803994656,
      "num_tokens": 888825745.0,
      "step": 27849
    },
    {
      "entropy": 0.47378135146573186,
      "epoch": 2.5623429422487507,
      "grad_norm": 0.15245293080806732,
      "learning_rate": 1.4591958781856657e-05,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8533378317952156,
      "num_tokens": 888857931.0,
      "step": 27850
    },
    {
      "entropy": 0.5632905717939138,
      "epoch": 2.562434948620192,
      "grad_norm": 0.16033044457435608,
      "learning_rate": 1.458889195571503e-05,
      "loss": 0.5534,
      "mean_token_accuracy": 0.825084775686264,
      "num_tokens": 888890598.0,
      "step": 27851
    },
    {
      "entropy": 0.5382906589657068,
      "epoch": 2.5625269549916334,
      "grad_norm": 0.1653214991092682,
      "learning_rate": 1.4585825129573405e-05,
      "loss": 0.5347,
      "mean_token_accuracy": 0.8311030976474285,
      "num_tokens": 888922669.0,
      "step": 27852
    },
    {
      "entropy": 0.6000905721448362,
      "epoch": 2.5626189613630745,
      "grad_norm": 0.17090415954589844,
      "learning_rate": 1.4582758303431778e-05,
      "loss": 0.5883,
      "mean_token_accuracy": 0.8148649297654629,
      "num_tokens": 888954816.0,
      "step": 27853
    },
    {
      "entropy": 0.48878972604870796,
      "epoch": 2.5627109677345157,
      "grad_norm": 0.15494948625564575,
      "learning_rate": 1.4579691477290153e-05,
      "loss": 0.4789,
      "mean_token_accuracy": 0.845712173730135,
      "num_tokens": 888987076.0,
      "step": 27854
    },
    {
      "entropy": 0.6123590935021639,
      "epoch": 2.562802974105957,
      "grad_norm": 0.1697036772966385,
      "learning_rate": 1.4576624651148527e-05,
      "loss": 0.5946,
      "mean_token_accuracy": 0.813482403755188,
      "num_tokens": 889018377.0,
      "step": 27855
    },
    {
      "entropy": 0.5683414898812771,
      "epoch": 2.562894980477398,
      "grad_norm": 0.167542964220047,
      "learning_rate": 1.45735578250069e-05,
      "loss": 0.5497,
      "mean_token_accuracy": 0.8251914978027344,
      "num_tokens": 889049118.0,
      "step": 27856
    },
    {
      "entropy": 0.5318475123494864,
      "epoch": 2.5629869868488395,
      "grad_norm": 0.1650393009185791,
      "learning_rate": 1.4570490998865275e-05,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8300949558615685,
      "num_tokens": 889080364.0,
      "step": 27857
    },
    {
      "entropy": 0.44798939966131,
      "epoch": 2.5630789932202807,
      "grad_norm": 0.1469188779592514,
      "learning_rate": 1.4567424172723648e-05,
      "loss": 0.431,
      "mean_token_accuracy": 0.8624731115996838,
      "num_tokens": 889112083.0,
      "step": 27858
    },
    {
      "entropy": 0.4814284862950444,
      "epoch": 2.563170999591722,
      "grad_norm": 0.16060402989387512,
      "learning_rate": 1.4564357346582023e-05,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8498095236718655,
      "num_tokens": 889143623.0,
      "step": 27859
    },
    {
      "entropy": 0.4664325970225036,
      "epoch": 2.563263005963163,
      "grad_norm": 0.14877617359161377,
      "learning_rate": 1.4561290520440395e-05,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8603771589696407,
      "num_tokens": 889174857.0,
      "step": 27860
    },
    {
      "entropy": 0.5134729743003845,
      "epoch": 2.563355012334604,
      "grad_norm": 0.15820728242397308,
      "learning_rate": 1.455822369429877e-05,
      "loss": 0.49,
      "mean_token_accuracy": 0.8420812487602234,
      "num_tokens": 889206575.0,
      "step": 27861
    },
    {
      "entropy": 0.6237915866076946,
      "epoch": 2.5634470187060456,
      "grad_norm": 0.1696142554283142,
      "learning_rate": 1.4555156868157143e-05,
      "loss": 0.6247,
      "mean_token_accuracy": 0.8067450858652592,
      "num_tokens": 889237957.0,
      "step": 27862
    },
    {
      "entropy": 0.4692017314955592,
      "epoch": 2.5635390250774868,
      "grad_norm": 0.16030259430408478,
      "learning_rate": 1.4552090042015518e-05,
      "loss": 0.4476,
      "mean_token_accuracy": 0.8550448529422283,
      "num_tokens": 889269683.0,
      "step": 27863
    },
    {
      "entropy": 0.4753136490471661,
      "epoch": 2.563631031448928,
      "grad_norm": 0.15714691579341888,
      "learning_rate": 1.454902321587389e-05,
      "loss": 0.4737,
      "mean_token_accuracy": 0.8508700132369995,
      "num_tokens": 889302249.0,
      "step": 27864
    },
    {
      "entropy": 0.5452870037406683,
      "epoch": 2.563723037820369,
      "grad_norm": 0.16334322094917297,
      "learning_rate": 1.4545956389732269e-05,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8343821465969086,
      "num_tokens": 889334052.0,
      "step": 27865
    },
    {
      "entropy": 0.5938349515199661,
      "epoch": 2.56381504419181,
      "grad_norm": 0.17447532713413239,
      "learning_rate": 1.4542889563590642e-05,
      "loss": 0.5785,
      "mean_token_accuracy": 0.8184165209531784,
      "num_tokens": 889364974.0,
      "step": 27866
    },
    {
      "entropy": 0.5730845178477466,
      "epoch": 2.5639070505632517,
      "grad_norm": 0.16482044756412506,
      "learning_rate": 1.4539822737449016e-05,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8261397331953049,
      "num_tokens": 889397185.0,
      "step": 27867
    },
    {
      "entropy": 0.6023056246340275,
      "epoch": 2.563999056934693,
      "grad_norm": 0.17303699254989624,
      "learning_rate": 1.453675591130739e-05,
      "loss": 0.593,
      "mean_token_accuracy": 0.8110248073935509,
      "num_tokens": 889428826.0,
      "step": 27868
    },
    {
      "entropy": 0.5335569866001606,
      "epoch": 2.564091063306134,
      "grad_norm": 0.16373461484909058,
      "learning_rate": 1.4533689085165764e-05,
      "loss": 0.5275,
      "mean_token_accuracy": 0.8352944888174534,
      "num_tokens": 889460913.0,
      "step": 27869
    },
    {
      "entropy": 0.4197855624370277,
      "epoch": 2.564183069677575,
      "grad_norm": 0.1521814614534378,
      "learning_rate": 1.4530622259024137e-05,
      "loss": 0.411,
      "mean_token_accuracy": 0.8684376999735832,
      "num_tokens": 889492716.0,
      "step": 27870
    },
    {
      "entropy": 0.533538306131959,
      "epoch": 2.5642750760490163,
      "grad_norm": 0.16324461996555328,
      "learning_rate": 1.4527555432882511e-05,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8314023651182652,
      "num_tokens": 889524954.0,
      "step": 27871
    },
    {
      "entropy": 0.6063045430928469,
      "epoch": 2.564367082420458,
      "grad_norm": 0.17043331265449524,
      "learning_rate": 1.4524488606740884e-05,
      "loss": 0.5989,
      "mean_token_accuracy": 0.8107070252299309,
      "num_tokens": 889556751.0,
      "step": 27872
    },
    {
      "entropy": 0.46632667747326195,
      "epoch": 2.564459088791899,
      "grad_norm": 0.1534368246793747,
      "learning_rate": 1.4521421780599259e-05,
      "loss": 0.4489,
      "mean_token_accuracy": 0.856281004846096,
      "num_tokens": 889588891.0,
      "step": 27873
    },
    {
      "entropy": 0.5348928431048989,
      "epoch": 2.56455109516334,
      "grad_norm": 0.16574068367481232,
      "learning_rate": 1.4518354954457634e-05,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8301669284701347,
      "num_tokens": 889620836.0,
      "step": 27874
    },
    {
      "entropy": 0.5206764228641987,
      "epoch": 2.5646431015347813,
      "grad_norm": 0.16734468936920166,
      "learning_rate": 1.4515288128316007e-05,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8417988941073418,
      "num_tokens": 889652457.0,
      "step": 27875
    },
    {
      "entropy": 0.5094137957785279,
      "epoch": 2.5647351079062224,
      "grad_norm": 0.16394419968128204,
      "learning_rate": 1.4512221302174381e-05,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8410102427005768,
      "num_tokens": 889684621.0,
      "step": 27876
    },
    {
      "entropy": 0.6297375410795212,
      "epoch": 2.564827114277664,
      "grad_norm": 0.17136503756046295,
      "learning_rate": 1.4509154476032754e-05,
      "loss": 0.6167,
      "mean_token_accuracy": 0.8095829598605633,
      "num_tokens": 889716718.0,
      "step": 27877
    },
    {
      "entropy": 0.5473395772278309,
      "epoch": 2.564919120649105,
      "grad_norm": 0.16379334032535553,
      "learning_rate": 1.4506087649891129e-05,
      "loss": 0.5355,
      "mean_token_accuracy": 0.8299798220396042,
      "num_tokens": 889749056.0,
      "step": 27878
    },
    {
      "entropy": 0.4561429196037352,
      "epoch": 2.5650111270205462,
      "grad_norm": 0.15654133260250092,
      "learning_rate": 1.4503020823749502e-05,
      "loss": 0.4389,
      "mean_token_accuracy": 0.8588622882962227,
      "num_tokens": 889781027.0,
      "step": 27879
    },
    {
      "entropy": 0.5182159543037415,
      "epoch": 2.5651031333919874,
      "grad_norm": 0.15516060590744019,
      "learning_rate": 1.4499953997607877e-05,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8424523696303368,
      "num_tokens": 889813795.0,
      "step": 27880
    },
    {
      "entropy": 0.49399156332947314,
      "epoch": 2.5651951397634285,
      "grad_norm": 0.15769687294960022,
      "learning_rate": 1.449688717146625e-05,
      "loss": 0.4847,
      "mean_token_accuracy": 0.844147015362978,
      "num_tokens": 889845399.0,
      "step": 27881
    },
    {
      "entropy": 0.44974307995289564,
      "epoch": 2.56528714613487,
      "grad_norm": 0.1489635854959488,
      "learning_rate": 1.4493820345324624e-05,
      "loss": 0.4379,
      "mean_token_accuracy": 0.8596273586153984,
      "num_tokens": 889878122.0,
      "step": 27882
    },
    {
      "entropy": 0.46560955699533224,
      "epoch": 2.565379152506311,
      "grad_norm": 0.15070955455303192,
      "learning_rate": 1.4490753519182997e-05,
      "loss": 0.4658,
      "mean_token_accuracy": 0.8542206771671772,
      "num_tokens": 889910209.0,
      "step": 27883
    },
    {
      "entropy": 0.49511122051626444,
      "epoch": 2.5654711588777523,
      "grad_norm": 0.16124238073825836,
      "learning_rate": 1.4487686693041372e-05,
      "loss": 0.4775,
      "mean_token_accuracy": 0.8507728055119514,
      "num_tokens": 889941661.0,
      "step": 27884
    },
    {
      "entropy": 0.5767561495304108,
      "epoch": 2.5655631652491935,
      "grad_norm": 0.16619133949279785,
      "learning_rate": 1.4484619866899746e-05,
      "loss": 0.572,
      "mean_token_accuracy": 0.8232764266431332,
      "num_tokens": 889973419.0,
      "step": 27885
    },
    {
      "entropy": 0.532209494151175,
      "epoch": 2.5656551716206346,
      "grad_norm": 0.15903155505657196,
      "learning_rate": 1.448155304075812e-05,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8364720046520233,
      "num_tokens": 890005246.0,
      "step": 27886
    },
    {
      "entropy": 0.5512922070920467,
      "epoch": 2.565747177992076,
      "grad_norm": 0.1695273071527481,
      "learning_rate": 1.4478486214616494e-05,
      "loss": 0.5403,
      "mean_token_accuracy": 0.8300509266555309,
      "num_tokens": 890037247.0,
      "step": 27887
    },
    {
      "entropy": 0.5209516631439328,
      "epoch": 2.5658391843635173,
      "grad_norm": 0.16033266484737396,
      "learning_rate": 1.4475419388474867e-05,
      "loss": 0.5026,
      "mean_token_accuracy": 0.8375773504376411,
      "num_tokens": 890068920.0,
      "step": 27888
    },
    {
      "entropy": 0.5310101704671979,
      "epoch": 2.5659311907349585,
      "grad_norm": 0.1596701294183731,
      "learning_rate": 1.4472352562333242e-05,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8396229967474937,
      "num_tokens": 890100649.0,
      "step": 27889
    },
    {
      "entropy": 0.5616439003497362,
      "epoch": 2.5660231971063996,
      "grad_norm": 0.1641140878200531,
      "learning_rate": 1.4469285736191615e-05,
      "loss": 0.5504,
      "mean_token_accuracy": 0.8252043500542641,
      "num_tokens": 890133179.0,
      "step": 27890
    },
    {
      "entropy": 0.6011888366192579,
      "epoch": 2.5661152034778407,
      "grad_norm": 0.17165352404117584,
      "learning_rate": 1.446621891004999e-05,
      "loss": 0.5831,
      "mean_token_accuracy": 0.816658403724432,
      "num_tokens": 890165223.0,
      "step": 27891
    },
    {
      "entropy": 0.5172638213261962,
      "epoch": 2.5662072098492823,
      "grad_norm": 0.1586354374885559,
      "learning_rate": 1.4463152083908362e-05,
      "loss": 0.4912,
      "mean_token_accuracy": 0.8406181558966637,
      "num_tokens": 890196297.0,
      "step": 27892
    },
    {
      "entropy": 0.5345943979918957,
      "epoch": 2.5662992162207234,
      "grad_norm": 0.16026243567466736,
      "learning_rate": 1.4460085257766737e-05,
      "loss": 0.5214,
      "mean_token_accuracy": 0.833476360887289,
      "num_tokens": 890227801.0,
      "step": 27893
    },
    {
      "entropy": 0.521923704771325,
      "epoch": 2.5663912225921646,
      "grad_norm": 0.16030490398406982,
      "learning_rate": 1.445701843162511e-05,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8403667584061623,
      "num_tokens": 890259558.0,
      "step": 27894
    },
    {
      "entropy": 0.504649979993701,
      "epoch": 2.5664832289636057,
      "grad_norm": 0.15870200097560883,
      "learning_rate": 1.4453951605483484e-05,
      "loss": 0.4888,
      "mean_token_accuracy": 0.8442210890352726,
      "num_tokens": 890291310.0,
      "step": 27895
    },
    {
      "entropy": 0.49154115468263626,
      "epoch": 2.566575235335047,
      "grad_norm": 0.15989382565021515,
      "learning_rate": 1.445088477934186e-05,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8469670601189137,
      "num_tokens": 890322711.0,
      "step": 27896
    },
    {
      "entropy": 0.4657503969501704,
      "epoch": 2.5666672417064884,
      "grad_norm": 0.16215190291404724,
      "learning_rate": 1.4447817953200235e-05,
      "loss": 0.4631,
      "mean_token_accuracy": 0.8538453616201878,
      "num_tokens": 890354992.0,
      "step": 27897
    },
    {
      "entropy": 0.5114262253046036,
      "epoch": 2.5667592480779295,
      "grad_norm": 0.1566595733165741,
      "learning_rate": 1.4444751127058608e-05,
      "loss": 0.5091,
      "mean_token_accuracy": 0.839518591761589,
      "num_tokens": 890387141.0,
      "step": 27898
    },
    {
      "entropy": 0.5422809245064855,
      "epoch": 2.5668512544493707,
      "grad_norm": 0.17023839056491852,
      "learning_rate": 1.4441684300916983e-05,
      "loss": 0.5383,
      "mean_token_accuracy": 0.829098779708147,
      "num_tokens": 890418791.0,
      "step": 27899
    },
    {
      "entropy": 0.5237382017076015,
      "epoch": 2.566943260820812,
      "grad_norm": 0.1589289903640747,
      "learning_rate": 1.4438617474775356e-05,
      "loss": 0.5177,
      "mean_token_accuracy": 0.8309347294270992,
      "num_tokens": 890451173.0,
      "step": 27900
    },
    {
      "entropy": 0.47765825781971216,
      "epoch": 2.567035267192253,
      "grad_norm": 0.15448039770126343,
      "learning_rate": 1.443555064863373e-05,
      "loss": 0.4677,
      "mean_token_accuracy": 0.851276807487011,
      "num_tokens": 890483055.0,
      "step": 27901
    },
    {
      "entropy": 0.5215599630028009,
      "epoch": 2.5671272735636945,
      "grad_norm": 0.16228656470775604,
      "learning_rate": 1.4432483822492105e-05,
      "loss": 0.5118,
      "mean_token_accuracy": 0.842516828328371,
      "num_tokens": 890515329.0,
      "step": 27902
    },
    {
      "entropy": 0.4064003434032202,
      "epoch": 2.5672192799351357,
      "grad_norm": 0.1492644101381302,
      "learning_rate": 1.4429416996350478e-05,
      "loss": 0.3927,
      "mean_token_accuracy": 0.874494556337595,
      "num_tokens": 890546705.0,
      "step": 27903
    },
    {
      "entropy": 0.5215918309986591,
      "epoch": 2.567311286306577,
      "grad_norm": 0.1602553278207779,
      "learning_rate": 1.4426350170208853e-05,
      "loss": 0.5108,
      "mean_token_accuracy": 0.8388034850358963,
      "num_tokens": 890578193.0,
      "step": 27904
    },
    {
      "entropy": 0.4414088404737413,
      "epoch": 2.567403292678018,
      "grad_norm": 0.14945168793201447,
      "learning_rate": 1.4423283344067226e-05,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8634303770959377,
      "num_tokens": 890610522.0,
      "step": 27905
    },
    {
      "entropy": 0.4641931760124862,
      "epoch": 2.567495299049459,
      "grad_norm": 0.15173378586769104,
      "learning_rate": 1.44202165179256e-05,
      "loss": 0.4564,
      "mean_token_accuracy": 0.8575015477836132,
      "num_tokens": 890642724.0,
      "step": 27906
    },
    {
      "entropy": 0.5507060494273901,
      "epoch": 2.5675873054209006,
      "grad_norm": 0.16607901453971863,
      "learning_rate": 1.4417149691783973e-05,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8318368121981621,
      "num_tokens": 890674035.0,
      "step": 27907
    },
    {
      "entropy": 0.6044289562851191,
      "epoch": 2.5676793117923418,
      "grad_norm": 0.17256629467010498,
      "learning_rate": 1.4414082865642348e-05,
      "loss": 0.5948,
      "mean_token_accuracy": 0.8129830248653889,
      "num_tokens": 890706376.0,
      "step": 27908
    },
    {
      "entropy": 0.5428751036524773,
      "epoch": 2.567771318163783,
      "grad_norm": 0.15897738933563232,
      "learning_rate": 1.4411016039500721e-05,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8328123651444912,
      "num_tokens": 890738025.0,
      "step": 27909
    },
    {
      "entropy": 0.5205864548915997,
      "epoch": 2.567863324535224,
      "grad_norm": 0.15454499423503876,
      "learning_rate": 1.4407949213359096e-05,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8403965197503567,
      "num_tokens": 890770192.0,
      "step": 27910
    },
    {
      "entropy": 0.4612108822911978,
      "epoch": 2.567955330906665,
      "grad_norm": 0.1529073417186737,
      "learning_rate": 1.4404882387217469e-05,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8587606847286224,
      "num_tokens": 890801672.0,
      "step": 27911
    },
    {
      "entropy": 0.5042638145387173,
      "epoch": 2.5680473372781067,
      "grad_norm": 0.15878555178642273,
      "learning_rate": 1.4401815561075843e-05,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8441482894122601,
      "num_tokens": 890833118.0,
      "step": 27912
    },
    {
      "entropy": 0.540640595369041,
      "epoch": 2.568139343649548,
      "grad_norm": 0.1652693748474121,
      "learning_rate": 1.4398748734934218e-05,
      "loss": 0.5215,
      "mean_token_accuracy": 0.8314888216555119,
      "num_tokens": 890865062.0,
      "step": 27913
    },
    {
      "entropy": 0.5210956952068955,
      "epoch": 2.568231350020989,
      "grad_norm": 0.1580081731081009,
      "learning_rate": 1.4395681908792591e-05,
      "loss": 0.5054,
      "mean_token_accuracy": 0.8407536931335926,
      "num_tokens": 890897001.0,
      "step": 27914
    },
    {
      "entropy": 0.5674803256988525,
      "epoch": 2.56832335639243,
      "grad_norm": 0.1709328144788742,
      "learning_rate": 1.4392615082650965e-05,
      "loss": 0.5693,
      "mean_token_accuracy": 0.8193533942103386,
      "num_tokens": 890929296.0,
      "step": 27915
    },
    {
      "entropy": 0.506020219065249,
      "epoch": 2.5684153627638713,
      "grad_norm": 0.1615322381258011,
      "learning_rate": 1.4389548256509338e-05,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8420915938913822,
      "num_tokens": 890961530.0,
      "step": 27916
    },
    {
      "entropy": 0.4512096121907234,
      "epoch": 2.568507369135313,
      "grad_norm": 0.15390193462371826,
      "learning_rate": 1.4386481430367713e-05,
      "loss": 0.4431,
      "mean_token_accuracy": 0.8565356358885765,
      "num_tokens": 890993393.0,
      "step": 27917
    },
    {
      "entropy": 0.5454994011670351,
      "epoch": 2.568599375506754,
      "grad_norm": 0.16397245228290558,
      "learning_rate": 1.4383414604226086e-05,
      "loss": 0.5331,
      "mean_token_accuracy": 0.8291516676545143,
      "num_tokens": 891025557.0,
      "step": 27918
    },
    {
      "entropy": 0.48944673500955105,
      "epoch": 2.568691381878195,
      "grad_norm": 0.1569342315196991,
      "learning_rate": 1.438034777808446e-05,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8537531718611717,
      "num_tokens": 891056936.0,
      "step": 27919
    },
    {
      "entropy": 0.541633314685896,
      "epoch": 2.5687833882496363,
      "grad_norm": 0.16377580165863037,
      "learning_rate": 1.4377280951942834e-05,
      "loss": 0.5271,
      "mean_token_accuracy": 0.8346587121486664,
      "num_tokens": 891088123.0,
      "step": 27920
    },
    {
      "entropy": 0.48458522465080023,
      "epoch": 2.5688753946210774,
      "grad_norm": 0.1546747386455536,
      "learning_rate": 1.4374214125801208e-05,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8488334827125072,
      "num_tokens": 891119665.0,
      "step": 27921
    },
    {
      "entropy": 0.601488090120256,
      "epoch": 2.568967400992519,
      "grad_norm": 0.17582696676254272,
      "learning_rate": 1.4371147299659581e-05,
      "loss": 0.6183,
      "mean_token_accuracy": 0.8100145496428013,
      "num_tokens": 891151787.0,
      "step": 27922
    },
    {
      "entropy": 0.582222499884665,
      "epoch": 2.56905940736396,
      "grad_norm": 0.16696660220623016,
      "learning_rate": 1.4368080473517956e-05,
      "loss": 0.5745,
      "mean_token_accuracy": 0.8200456015765667,
      "num_tokens": 891183528.0,
      "step": 27923
    },
    {
      "entropy": 0.4184448285959661,
      "epoch": 2.5691514137354012,
      "grad_norm": 0.15064118802547455,
      "learning_rate": 1.436501364737633e-05,
      "loss": 0.4138,
      "mean_token_accuracy": 0.8663708083331585,
      "num_tokens": 891214402.0,
      "step": 27924
    },
    {
      "entropy": 0.4309857194311917,
      "epoch": 2.5692434201068424,
      "grad_norm": 0.1533035784959793,
      "learning_rate": 1.4361946821234704e-05,
      "loss": 0.4223,
      "mean_token_accuracy": 0.8674250245094299,
      "num_tokens": 891245961.0,
      "step": 27925
    },
    {
      "entropy": 0.6238070651888847,
      "epoch": 2.5693354264782835,
      "grad_norm": 0.17144104838371277,
      "learning_rate": 1.435887999509308e-05,
      "loss": 0.6124,
      "mean_token_accuracy": 0.8099132440984249,
      "num_tokens": 891277943.0,
      "step": 27926
    },
    {
      "entropy": 0.5005157166160643,
      "epoch": 2.569427432849725,
      "grad_norm": 0.1505272537469864,
      "learning_rate": 1.4355813168951454e-05,
      "loss": 0.4786,
      "mean_token_accuracy": 0.8468839600682259,
      "num_tokens": 891309748.0,
      "step": 27927
    },
    {
      "entropy": 0.47201241995207965,
      "epoch": 2.569519439221166,
      "grad_norm": 0.1567259579896927,
      "learning_rate": 1.4352746342809827e-05,
      "loss": 0.4571,
      "mean_token_accuracy": 0.8554319441318512,
      "num_tokens": 891341456.0,
      "step": 27928
    },
    {
      "entropy": 0.49358425475656986,
      "epoch": 2.5696114455926073,
      "grad_norm": 0.15991173684597015,
      "learning_rate": 1.4349679516668202e-05,
      "loss": 0.4868,
      "mean_token_accuracy": 0.8440550602972507,
      "num_tokens": 891372425.0,
      "step": 27929
    },
    {
      "entropy": 0.45905011892318726,
      "epoch": 2.5697034519640485,
      "grad_norm": 0.15137462317943573,
      "learning_rate": 1.4346612690526575e-05,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8572192899882793,
      "num_tokens": 891404973.0,
      "step": 27930
    },
    {
      "entropy": 0.46790698275435716,
      "epoch": 2.5697954583354896,
      "grad_norm": 0.1551733762025833,
      "learning_rate": 1.434354586438495e-05,
      "loss": 0.4586,
      "mean_token_accuracy": 0.8525911755859852,
      "num_tokens": 891436655.0,
      "step": 27931
    },
    {
      "entropy": 0.4668181296437979,
      "epoch": 2.569887464706931,
      "grad_norm": 0.15075097978115082,
      "learning_rate": 1.4340479038243324e-05,
      "loss": 0.4544,
      "mean_token_accuracy": 0.8565689884126186,
      "num_tokens": 891468617.0,
      "step": 27932
    },
    {
      "entropy": 0.4341719662770629,
      "epoch": 2.5699794710783723,
      "grad_norm": 0.15051254630088806,
      "learning_rate": 1.4337412212101697e-05,
      "loss": 0.4233,
      "mean_token_accuracy": 0.86410191655159,
      "num_tokens": 891500101.0,
      "step": 27933
    },
    {
      "entropy": 0.5104367267340422,
      "epoch": 2.5700714774498135,
      "grad_norm": 0.16301214694976807,
      "learning_rate": 1.4334345385960072e-05,
      "loss": 0.4979,
      "mean_token_accuracy": 0.8400785177946091,
      "num_tokens": 891531907.0,
      "step": 27934
    },
    {
      "entropy": 0.5294713648036122,
      "epoch": 2.5701634838212546,
      "grad_norm": 0.1629306972026825,
      "learning_rate": 1.4331278559818445e-05,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8357228711247444,
      "num_tokens": 891563213.0,
      "step": 27935
    },
    {
      "entropy": 0.6247416734695435,
      "epoch": 2.5702554901926957,
      "grad_norm": 0.1726834923028946,
      "learning_rate": 1.432821173367682e-05,
      "loss": 0.6172,
      "mean_token_accuracy": 0.8084314316511154,
      "num_tokens": 891594480.0,
      "step": 27936
    },
    {
      "entropy": 0.5382180279120803,
      "epoch": 2.5703474965641373,
      "grad_norm": 0.16155965626239777,
      "learning_rate": 1.4325144907535192e-05,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8326293230056763,
      "num_tokens": 891626359.0,
      "step": 27937
    },
    {
      "entropy": 0.4651665575802326,
      "epoch": 2.5704395029355784,
      "grad_norm": 0.1553574502468109,
      "learning_rate": 1.4322078081393567e-05,
      "loss": 0.451,
      "mean_token_accuracy": 0.8573754988610744,
      "num_tokens": 891658274.0,
      "step": 27938
    },
    {
      "entropy": 0.3787343465955928,
      "epoch": 2.5705315093070196,
      "grad_norm": 0.13580022752285004,
      "learning_rate": 1.431901125525194e-05,
      "loss": 0.3732,
      "mean_token_accuracy": 0.8803141005337238,
      "num_tokens": 891690995.0,
      "step": 27939
    },
    {
      "entropy": 0.5816492736339569,
      "epoch": 2.5706235156784607,
      "grad_norm": 0.16726025938987732,
      "learning_rate": 1.4315944429110315e-05,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8252270594239235,
      "num_tokens": 891722710.0,
      "step": 27940
    },
    {
      "entropy": 0.537945793941617,
      "epoch": 2.570715522049902,
      "grad_norm": 0.16247297823429108,
      "learning_rate": 1.4312877602968688e-05,
      "loss": 0.524,
      "mean_token_accuracy": 0.8325440213084221,
      "num_tokens": 891754670.0,
      "step": 27941
    },
    {
      "entropy": 0.5387726090848446,
      "epoch": 2.5708075284213434,
      "grad_norm": 0.16100311279296875,
      "learning_rate": 1.4309810776827062e-05,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8372866921126842,
      "num_tokens": 891786711.0,
      "step": 27942
    },
    {
      "entropy": 0.5143225025385618,
      "epoch": 2.5708995347927845,
      "grad_norm": 0.15586625039577484,
      "learning_rate": 1.4306743950685437e-05,
      "loss": 0.5038,
      "mean_token_accuracy": 0.8410951383411884,
      "num_tokens": 891819250.0,
      "step": 27943
    },
    {
      "entropy": 0.5333003047853708,
      "epoch": 2.5709915411642257,
      "grad_norm": 0.16351240873336792,
      "learning_rate": 1.430367712454381e-05,
      "loss": 0.508,
      "mean_token_accuracy": 0.8379179984331131,
      "num_tokens": 891850122.0,
      "step": 27944
    },
    {
      "entropy": 0.6330971722491086,
      "epoch": 2.571083547535667,
      "grad_norm": 0.17854207754135132,
      "learning_rate": 1.4300610298402185e-05,
      "loss": 0.6235,
      "mean_token_accuracy": 0.8047675788402557,
      "num_tokens": 891881832.0,
      "step": 27945
    },
    {
      "entropy": 0.5742170177400112,
      "epoch": 2.571175553907108,
      "grad_norm": 0.17309479415416718,
      "learning_rate": 1.4297543472260558e-05,
      "loss": 0.5836,
      "mean_token_accuracy": 0.8182850927114487,
      "num_tokens": 891913552.0,
      "step": 27946
    },
    {
      "entropy": 0.6018597669899464,
      "epoch": 2.5712675602785495,
      "grad_norm": 0.16819734871387482,
      "learning_rate": 1.4294476646118932e-05,
      "loss": 0.597,
      "mean_token_accuracy": 0.8130269832909107,
      "num_tokens": 891945902.0,
      "step": 27947
    },
    {
      "entropy": 0.5878912128973752,
      "epoch": 2.5713595666499907,
      "grad_norm": 0.16594251990318298,
      "learning_rate": 1.4291409819977305e-05,
      "loss": 0.5764,
      "mean_token_accuracy": 0.8172858133912086,
      "num_tokens": 891978295.0,
      "step": 27948
    },
    {
      "entropy": 0.4467625254765153,
      "epoch": 2.571451573021432,
      "grad_norm": 0.15175814926624298,
      "learning_rate": 1.428834299383568e-05,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8598435185849667,
      "num_tokens": 892010907.0,
      "step": 27949
    },
    {
      "entropy": 0.5819784719496965,
      "epoch": 2.571543579392873,
      "grad_norm": 0.1679498702287674,
      "learning_rate": 1.4285276167694053e-05,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8216334991157055,
      "num_tokens": 892041866.0,
      "step": 27950
    },
    {
      "entropy": 0.48603968881070614,
      "epoch": 2.571635585764314,
      "grad_norm": 0.1575077921152115,
      "learning_rate": 1.4282209341552427e-05,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8486887067556381,
      "num_tokens": 892074026.0,
      "step": 27951
    },
    {
      "entropy": 0.4267250762786716,
      "epoch": 2.5717275921357556,
      "grad_norm": 0.14348386228084564,
      "learning_rate": 1.42791425154108e-05,
      "loss": 0.4074,
      "mean_token_accuracy": 0.8695660009980202,
      "num_tokens": 892105667.0,
      "step": 27952
    },
    {
      "entropy": 0.5797330029308796,
      "epoch": 2.5718195985071968,
      "grad_norm": 0.17026041448116302,
      "learning_rate": 1.4276075689269175e-05,
      "loss": 0.5765,
      "mean_token_accuracy": 0.8201380595564842,
      "num_tokens": 892136430.0,
      "step": 27953
    },
    {
      "entropy": 0.44579151505604386,
      "epoch": 2.571911604878638,
      "grad_norm": 0.14512144029140472,
      "learning_rate": 1.427300886312755e-05,
      "loss": 0.4233,
      "mean_token_accuracy": 0.8665755614638329,
      "num_tokens": 892168413.0,
      "step": 27954
    },
    {
      "entropy": 0.4508739758748561,
      "epoch": 2.572003611250079,
      "grad_norm": 0.1470937579870224,
      "learning_rate": 1.4269942036985923e-05,
      "loss": 0.4423,
      "mean_token_accuracy": 0.8604704774916172,
      "num_tokens": 892200830.0,
      "step": 27955
    },
    {
      "entropy": 0.5190306855365634,
      "epoch": 2.57209561762152,
      "grad_norm": 0.15772563219070435,
      "learning_rate": 1.4266875210844297e-05,
      "loss": 0.5,
      "mean_token_accuracy": 0.8398125693202019,
      "num_tokens": 892233060.0,
      "step": 27956
    },
    {
      "entropy": 0.5204372890293598,
      "epoch": 2.5721876239929617,
      "grad_norm": 0.16527044773101807,
      "learning_rate": 1.4263808384702674e-05,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8387764133512974,
      "num_tokens": 892263580.0,
      "step": 27957
    },
    {
      "entropy": 0.5502995015121996,
      "epoch": 2.572279630364403,
      "grad_norm": 0.16883869469165802,
      "learning_rate": 1.4260741558561047e-05,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8335638605058193,
      "num_tokens": 892294064.0,
      "step": 27958
    },
    {
      "entropy": 0.5449499050155282,
      "epoch": 2.572371636735844,
      "grad_norm": 0.16496989130973816,
      "learning_rate": 1.4257674732419421e-05,
      "loss": 0.5423,
      "mean_token_accuracy": 0.8264891691505909,
      "num_tokens": 892325748.0,
      "step": 27959
    },
    {
      "entropy": 0.585951704531908,
      "epoch": 2.572463643107285,
      "grad_norm": 0.16865162551403046,
      "learning_rate": 1.4254607906277794e-05,
      "loss": 0.5718,
      "mean_token_accuracy": 0.8210691027343273,
      "num_tokens": 892358460.0,
      "step": 27960
    },
    {
      "entropy": 0.5394858997315168,
      "epoch": 2.5725556494787263,
      "grad_norm": 0.16654163599014282,
      "learning_rate": 1.4251541080136169e-05,
      "loss": 0.5309,
      "mean_token_accuracy": 0.833041999489069,
      "num_tokens": 892390292.0,
      "step": 27961
    },
    {
      "entropy": 0.5797751042991877,
      "epoch": 2.572647655850168,
      "grad_norm": 0.16374480724334717,
      "learning_rate": 1.4248474253994543e-05,
      "loss": 0.5641,
      "mean_token_accuracy": 0.8234448656439781,
      "num_tokens": 892421399.0,
      "step": 27962
    },
    {
      "entropy": 0.4477890655398369,
      "epoch": 2.572739662221609,
      "grad_norm": 0.15384066104888916,
      "learning_rate": 1.4245407427852916e-05,
      "loss": 0.44,
      "mean_token_accuracy": 0.8585681170225143,
      "num_tokens": 892453504.0,
      "step": 27963
    },
    {
      "entropy": 0.566740179201588,
      "epoch": 2.57283166859305,
      "grad_norm": 0.16396564245224,
      "learning_rate": 1.4242340601711291e-05,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8260596171021461,
      "num_tokens": 892485457.0,
      "step": 27964
    },
    {
      "entropy": 0.5266492031514645,
      "epoch": 2.5729236749644913,
      "grad_norm": 0.15938343107700348,
      "learning_rate": 1.4239273775569664e-05,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8355799652636051,
      "num_tokens": 892517235.0,
      "step": 27965
    },
    {
      "entropy": 0.47260626405477524,
      "epoch": 2.5730156813359324,
      "grad_norm": 0.14840881526470184,
      "learning_rate": 1.4236206949428039e-05,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8506514020264149,
      "num_tokens": 892549934.0,
      "step": 27966
    },
    {
      "entropy": 0.5682886471040547,
      "epoch": 2.573107687707374,
      "grad_norm": 0.17377158999443054,
      "learning_rate": 1.4233140123286412e-05,
      "loss": 0.576,
      "mean_token_accuracy": 0.8214317932724953,
      "num_tokens": 892581091.0,
      "step": 27967
    },
    {
      "entropy": 0.6179476827383041,
      "epoch": 2.573199694078815,
      "grad_norm": 0.17039507627487183,
      "learning_rate": 1.4230073297144786e-05,
      "loss": 0.6064,
      "mean_token_accuracy": 0.8122630566358566,
      "num_tokens": 892613241.0,
      "step": 27968
    },
    {
      "entropy": 0.5347497602924705,
      "epoch": 2.5732917004502562,
      "grad_norm": 0.15969495475292206,
      "learning_rate": 1.422700647100316e-05,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8367887176573277,
      "num_tokens": 892645195.0,
      "step": 27969
    },
    {
      "entropy": 0.572668831795454,
      "epoch": 2.5733837068216974,
      "grad_norm": 0.16911834478378296,
      "learning_rate": 1.4223939644861534e-05,
      "loss": 0.5692,
      "mean_token_accuracy": 0.8216797187924385,
      "num_tokens": 892676485.0,
      "step": 27970
    },
    {
      "entropy": 0.47128820046782494,
      "epoch": 2.5734757131931385,
      "grad_norm": 0.15517233312129974,
      "learning_rate": 1.4220872818719907e-05,
      "loss": 0.4579,
      "mean_token_accuracy": 0.8525437265634537,
      "num_tokens": 892708663.0,
      "step": 27971
    },
    {
      "entropy": 0.5485154001507908,
      "epoch": 2.57356771956458,
      "grad_norm": 0.1662643998861313,
      "learning_rate": 1.4217805992578281e-05,
      "loss": 0.54,
      "mean_token_accuracy": 0.8284154906868935,
      "num_tokens": 892740576.0,
      "step": 27972
    },
    {
      "entropy": 0.4673970816656947,
      "epoch": 2.573659725936021,
      "grad_norm": 0.15898621082305908,
      "learning_rate": 1.4214739166436656e-05,
      "loss": 0.4524,
      "mean_token_accuracy": 0.8545256033539772,
      "num_tokens": 892772163.0,
      "step": 27973
    },
    {
      "entropy": 0.4881863612681627,
      "epoch": 2.5737517323074623,
      "grad_norm": 0.1552416831254959,
      "learning_rate": 1.4211672340295029e-05,
      "loss": 0.4839,
      "mean_token_accuracy": 0.8461059853434563,
      "num_tokens": 892804738.0,
      "step": 27974
    },
    {
      "entropy": 0.4985124636441469,
      "epoch": 2.5738437386789035,
      "grad_norm": 0.15757279098033905,
      "learning_rate": 1.4208605514153404e-05,
      "loss": 0.4839,
      "mean_token_accuracy": 0.8467116020619869,
      "num_tokens": 892836457.0,
      "step": 27975
    },
    {
      "entropy": 0.5355193605646491,
      "epoch": 2.5739357450503446,
      "grad_norm": 0.1630270779132843,
      "learning_rate": 1.4205538688011777e-05,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8332093134522438,
      "num_tokens": 892868325.0,
      "step": 27976
    },
    {
      "entropy": 0.5753793502226472,
      "epoch": 2.574027751421786,
      "grad_norm": 0.16550315916538239,
      "learning_rate": 1.4202471861870151e-05,
      "loss": 0.5646,
      "mean_token_accuracy": 0.8195796050131321,
      "num_tokens": 892900910.0,
      "step": 27977
    },
    {
      "entropy": 0.48852985352277756,
      "epoch": 2.5741197577932273,
      "grad_norm": 0.1520439237356186,
      "learning_rate": 1.4199405035728524e-05,
      "loss": 0.4711,
      "mean_token_accuracy": 0.850315771996975,
      "num_tokens": 892932528.0,
      "step": 27978
    },
    {
      "entropy": 0.5241702296771109,
      "epoch": 2.5742117641646685,
      "grad_norm": 0.16382461786270142,
      "learning_rate": 1.4196338209586899e-05,
      "loss": 0.52,
      "mean_token_accuracy": 0.8350472934544086,
      "num_tokens": 892964242.0,
      "step": 27979
    },
    {
      "entropy": 0.6076832562685013,
      "epoch": 2.5743037705361096,
      "grad_norm": 0.1683720350265503,
      "learning_rate": 1.4193271383445272e-05,
      "loss": 0.6059,
      "mean_token_accuracy": 0.8105281740427017,
      "num_tokens": 892996445.0,
      "step": 27980
    },
    {
      "entropy": 0.3992165403906256,
      "epoch": 2.5743957769075507,
      "grad_norm": 0.1423865407705307,
      "learning_rate": 1.4190204557303646e-05,
      "loss": 0.3845,
      "mean_token_accuracy": 0.8742060773074627,
      "num_tokens": 893028470.0,
      "step": 27981
    },
    {
      "entropy": 0.5328290089964867,
      "epoch": 2.5744877832789923,
      "grad_norm": 0.16029511392116547,
      "learning_rate": 1.418713773116202e-05,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8399805352091789,
      "num_tokens": 893060277.0,
      "step": 27982
    },
    {
      "entropy": 0.6083287103101611,
      "epoch": 2.5745797896504334,
      "grad_norm": 0.16923151910305023,
      "learning_rate": 1.4184070905020394e-05,
      "loss": 0.5962,
      "mean_token_accuracy": 0.8094798624515533,
      "num_tokens": 893092525.0,
      "step": 27983
    },
    {
      "entropy": 0.4960946086794138,
      "epoch": 2.5746717960218746,
      "grad_norm": 0.15495114028453827,
      "learning_rate": 1.4181004078878769e-05,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8498216904699802,
      "num_tokens": 893123611.0,
      "step": 27984
    },
    {
      "entropy": 0.5684484029188752,
      "epoch": 2.5747638023933157,
      "grad_norm": 0.16752584278583527,
      "learning_rate": 1.4177937252737142e-05,
      "loss": 0.551,
      "mean_token_accuracy": 0.8275458067655563,
      "num_tokens": 893154823.0,
      "step": 27985
    },
    {
      "entropy": 0.5321900248527527,
      "epoch": 2.574855808764757,
      "grad_norm": 0.16383430361747742,
      "learning_rate": 1.4174870426595516e-05,
      "loss": 0.5099,
      "mean_token_accuracy": 0.837520994246006,
      "num_tokens": 893186491.0,
      "step": 27986
    },
    {
      "entropy": 0.5863673416897655,
      "epoch": 2.5749478151361984,
      "grad_norm": 0.16512398421764374,
      "learning_rate": 1.4171803600453893e-05,
      "loss": 0.5506,
      "mean_token_accuracy": 0.8236427269876003,
      "num_tokens": 893218147.0,
      "step": 27987
    },
    {
      "entropy": 0.49433601275086403,
      "epoch": 2.5750398215076395,
      "grad_norm": 0.15927410125732422,
      "learning_rate": 1.4168736774312266e-05,
      "loss": 0.4908,
      "mean_token_accuracy": 0.8426626473665237,
      "num_tokens": 893250040.0,
      "step": 27988
    },
    {
      "entropy": 0.6210692934691906,
      "epoch": 2.5751318278790807,
      "grad_norm": 0.17827650904655457,
      "learning_rate": 1.416566994817064e-05,
      "loss": 0.5967,
      "mean_token_accuracy": 0.8138885423541069,
      "num_tokens": 893280888.0,
      "step": 27989
    },
    {
      "entropy": 0.4726563999429345,
      "epoch": 2.575223834250522,
      "grad_norm": 0.15044964849948883,
      "learning_rate": 1.4162603122029013e-05,
      "loss": 0.4652,
      "mean_token_accuracy": 0.852663554251194,
      "num_tokens": 893313385.0,
      "step": 27990
    },
    {
      "entropy": 0.42193164909258485,
      "epoch": 2.575315840621963,
      "grad_norm": 0.14595040678977966,
      "learning_rate": 1.4159536295887388e-05,
      "loss": 0.409,
      "mean_token_accuracy": 0.8694382086396217,
      "num_tokens": 893345026.0,
      "step": 27991
    },
    {
      "entropy": 0.5711971721611917,
      "epoch": 2.5754078469934045,
      "grad_norm": 0.17088863253593445,
      "learning_rate": 1.4156469469745763e-05,
      "loss": 0.5668,
      "mean_token_accuracy": 0.823500283062458,
      "num_tokens": 893376745.0,
      "step": 27992
    },
    {
      "entropy": 0.5254887195769697,
      "epoch": 2.5754998533648457,
      "grad_norm": 0.15623155236244202,
      "learning_rate": 1.4153402643604135e-05,
      "loss": 0.5217,
      "mean_token_accuracy": 0.8368417955935001,
      "num_tokens": 893409465.0,
      "step": 27993
    },
    {
      "entropy": 0.5764715578407049,
      "epoch": 2.575591859736287,
      "grad_norm": 0.17355838418006897,
      "learning_rate": 1.415033581746251e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8203565627336502,
      "num_tokens": 893441075.0,
      "step": 27994
    },
    {
      "entropy": 0.4576921174302697,
      "epoch": 2.575683866107728,
      "grad_norm": 0.15359675884246826,
      "learning_rate": 1.4147268991320883e-05,
      "loss": 0.4466,
      "mean_token_accuracy": 0.8576577752828598,
      "num_tokens": 893473161.0,
      "step": 27995
    },
    {
      "entropy": 0.47995081346016377,
      "epoch": 2.575775872479169,
      "grad_norm": 0.1641051322221756,
      "learning_rate": 1.4144202165179258e-05,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8505962565541267,
      "num_tokens": 893505242.0,
      "step": 27996
    },
    {
      "entropy": 0.5468443230492994,
      "epoch": 2.5758678788506106,
      "grad_norm": 0.15815015137195587,
      "learning_rate": 1.414113533903763e-05,
      "loss": 0.5381,
      "mean_token_accuracy": 0.8299625106155872,
      "num_tokens": 893537028.0,
      "step": 27997
    },
    {
      "entropy": 0.5327145135961473,
      "epoch": 2.5759598852220518,
      "grad_norm": 0.16608621180057526,
      "learning_rate": 1.4138068512896005e-05,
      "loss": 0.5326,
      "mean_token_accuracy": 0.8321197666227818,
      "num_tokens": 893568610.0,
      "step": 27998
    },
    {
      "entropy": 0.5097421179525554,
      "epoch": 2.576051891593493,
      "grad_norm": 0.1641908586025238,
      "learning_rate": 1.4135001686754378e-05,
      "loss": 0.5105,
      "mean_token_accuracy": 0.837648581713438,
      "num_tokens": 893600187.0,
      "step": 27999
    },
    {
      "entropy": 0.5280960937961936,
      "epoch": 2.576143897964934,
      "grad_norm": 0.16367585957050323,
      "learning_rate": 1.4131934860612753e-05,
      "loss": 0.5181,
      "mean_token_accuracy": 0.8380425907671452,
      "num_tokens": 893631731.0,
      "step": 28000
    },
    {
      "entropy": 0.46761382184922695,
      "epoch": 2.576235904336375,
      "grad_norm": 0.15425077080726624,
      "learning_rate": 1.4128868034471126e-05,
      "loss": 0.4582,
      "mean_token_accuracy": 0.8562113270163536,
      "num_tokens": 893664271.0,
      "step": 28001
    },
    {
      "entropy": 0.5099545361008495,
      "epoch": 2.5763279107078167,
      "grad_norm": 0.1539985090494156,
      "learning_rate": 1.41258012083295e-05,
      "loss": 0.4947,
      "mean_token_accuracy": 0.845317430794239,
      "num_tokens": 893696719.0,
      "step": 28002
    },
    {
      "entropy": 0.49378079548478127,
      "epoch": 2.576419917079258,
      "grad_norm": 0.15479028224945068,
      "learning_rate": 1.4122734382187875e-05,
      "loss": 0.4884,
      "mean_token_accuracy": 0.8424283042550087,
      "num_tokens": 893728966.0,
      "step": 28003
    },
    {
      "entropy": 0.4477776982821524,
      "epoch": 2.576511923450699,
      "grad_norm": 0.14910836517810822,
      "learning_rate": 1.4119667556046248e-05,
      "loss": 0.4335,
      "mean_token_accuracy": 0.8643188662827015,
      "num_tokens": 893761036.0,
      "step": 28004
    },
    {
      "entropy": 0.49823045264929533,
      "epoch": 2.57660392982214,
      "grad_norm": 0.15850380063056946,
      "learning_rate": 1.4116600729904623e-05,
      "loss": 0.4785,
      "mean_token_accuracy": 0.8467671163380146,
      "num_tokens": 893793086.0,
      "step": 28005
    },
    {
      "entropy": 0.5128929421771318,
      "epoch": 2.5766959361935813,
      "grad_norm": 0.15778645873069763,
      "learning_rate": 1.4113533903762996e-05,
      "loss": 0.4968,
      "mean_token_accuracy": 0.8429163172841072,
      "num_tokens": 893825191.0,
      "step": 28006
    },
    {
      "entropy": 0.5254715019837022,
      "epoch": 2.576787942565023,
      "grad_norm": 0.15423521399497986,
      "learning_rate": 1.411046707762137e-05,
      "loss": 0.5105,
      "mean_token_accuracy": 0.839458879083395,
      "num_tokens": 893857757.0,
      "step": 28007
    },
    {
      "entropy": 0.4875939078629017,
      "epoch": 2.576879948936464,
      "grad_norm": 0.1548987776041031,
      "learning_rate": 1.4107400251479743e-05,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8485518917441368,
      "num_tokens": 893889610.0,
      "step": 28008
    },
    {
      "entropy": 0.5285850875079632,
      "epoch": 2.576971955307905,
      "grad_norm": 0.1592375934123993,
      "learning_rate": 1.4104333425338118e-05,
      "loss": 0.5179,
      "mean_token_accuracy": 0.8365970589220524,
      "num_tokens": 893921794.0,
      "step": 28009
    },
    {
      "entropy": 0.44693043315783143,
      "epoch": 2.5770639616793463,
      "grad_norm": 0.14636920392513275,
      "learning_rate": 1.4101266599196491e-05,
      "loss": 0.4222,
      "mean_token_accuracy": 0.863738413900137,
      "num_tokens": 893953875.0,
      "step": 28010
    },
    {
      "entropy": 0.6529879234731197,
      "epoch": 2.5771559680507874,
      "grad_norm": 0.1740485578775406,
      "learning_rate": 1.4098199773054866e-05,
      "loss": 0.6311,
      "mean_token_accuracy": 0.8008840940892696,
      "num_tokens": 893985531.0,
      "step": 28011
    },
    {
      "entropy": 0.49650157149881124,
      "epoch": 2.577247974422229,
      "grad_norm": 0.16046440601348877,
      "learning_rate": 1.4095132946913239e-05,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8461982794106007,
      "num_tokens": 894017056.0,
      "step": 28012
    },
    {
      "entropy": 0.4409843790344894,
      "epoch": 2.57733998079367,
      "grad_norm": 0.14743086695671082,
      "learning_rate": 1.4092066120771613e-05,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8634797409176826,
      "num_tokens": 894049449.0,
      "step": 28013
    },
    {
      "entropy": 0.5761608453467488,
      "epoch": 2.5774319871651112,
      "grad_norm": 0.17306651175022125,
      "learning_rate": 1.4088999294629988e-05,
      "loss": 0.5659,
      "mean_token_accuracy": 0.8190472535789013,
      "num_tokens": 894081319.0,
      "step": 28014
    },
    {
      "entropy": 0.4575481144711375,
      "epoch": 2.5775239935365524,
      "grad_norm": 0.15013161301612854,
      "learning_rate": 1.408593246848836e-05,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8567826002836227,
      "num_tokens": 894113950.0,
      "step": 28015
    },
    {
      "entropy": 0.40556412760633975,
      "epoch": 2.5776159999079935,
      "grad_norm": 0.14475876092910767,
      "learning_rate": 1.4082865642346735e-05,
      "loss": 0.4006,
      "mean_token_accuracy": 0.8729747198522091,
      "num_tokens": 894146480.0,
      "step": 28016
    },
    {
      "entropy": 0.5303142722696066,
      "epoch": 2.577708006279435,
      "grad_norm": 0.15840694308280945,
      "learning_rate": 1.4079798816205108e-05,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8344940394163132,
      "num_tokens": 894178947.0,
      "step": 28017
    },
    {
      "entropy": 0.548601046204567,
      "epoch": 2.577800012650876,
      "grad_norm": 0.15955029428005219,
      "learning_rate": 1.4076731990063485e-05,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8294293992221355,
      "num_tokens": 894210736.0,
      "step": 28018
    },
    {
      "entropy": 0.567031092941761,
      "epoch": 2.5778920190223173,
      "grad_norm": 0.17104852199554443,
      "learning_rate": 1.407366516392186e-05,
      "loss": 0.5577,
      "mean_token_accuracy": 0.8228063471615314,
      "num_tokens": 894242489.0,
      "step": 28019
    },
    {
      "entropy": 0.46259811613708735,
      "epoch": 2.5779840253937585,
      "grad_norm": 0.1579248607158661,
      "learning_rate": 1.4070598337780232e-05,
      "loss": 0.4478,
      "mean_token_accuracy": 0.85654691234231,
      "num_tokens": 894273668.0,
      "step": 28020
    },
    {
      "entropy": 0.5618950752541423,
      "epoch": 2.5780760317651996,
      "grad_norm": 0.16269555687904358,
      "learning_rate": 1.4067531511638607e-05,
      "loss": 0.5478,
      "mean_token_accuracy": 0.8284578993916512,
      "num_tokens": 894305637.0,
      "step": 28021
    },
    {
      "entropy": 0.5584892840124667,
      "epoch": 2.578168038136641,
      "grad_norm": 0.1630474179983139,
      "learning_rate": 1.4064464685496982e-05,
      "loss": 0.5398,
      "mean_token_accuracy": 0.8286743722856045,
      "num_tokens": 894337661.0,
      "step": 28022
    },
    {
      "entropy": 0.5518420310690999,
      "epoch": 2.5782600445080823,
      "grad_norm": 0.16632410883903503,
      "learning_rate": 1.4061397859355355e-05,
      "loss": 0.5385,
      "mean_token_accuracy": 0.830918226391077,
      "num_tokens": 894370158.0,
      "step": 28023
    },
    {
      "entropy": 0.4765683338046074,
      "epoch": 2.5783520508795235,
      "grad_norm": 0.15273383259773254,
      "learning_rate": 1.405833103321373e-05,
      "loss": 0.459,
      "mean_token_accuracy": 0.8532802127301693,
      "num_tokens": 894402413.0,
      "step": 28024
    },
    {
      "entropy": 0.5149699486792088,
      "epoch": 2.5784440572509646,
      "grad_norm": 0.16384385526180267,
      "learning_rate": 1.4055264207072102e-05,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8372618556022644,
      "num_tokens": 894434952.0,
      "step": 28025
    },
    {
      "entropy": 0.4480103952810168,
      "epoch": 2.5785360636224057,
      "grad_norm": 0.14666347205638885,
      "learning_rate": 1.4052197380930477e-05,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8605801872909069,
      "num_tokens": 894467179.0,
      "step": 28026
    },
    {
      "entropy": 0.5688767426181585,
      "epoch": 2.5786280699938473,
      "grad_norm": 0.16006797552108765,
      "learning_rate": 1.404913055478885e-05,
      "loss": 0.5544,
      "mean_token_accuracy": 0.8252676874399185,
      "num_tokens": 894499538.0,
      "step": 28027
    },
    {
      "entropy": 0.5012903790920973,
      "epoch": 2.5787200763652884,
      "grad_norm": 0.15825559198856354,
      "learning_rate": 1.4046063728647224e-05,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8452689796686172,
      "num_tokens": 894531119.0,
      "step": 28028
    },
    {
      "entropy": 0.45808917097747326,
      "epoch": 2.5788120827367296,
      "grad_norm": 0.15159830451011658,
      "learning_rate": 1.4042996902505597e-05,
      "loss": 0.454,
      "mean_token_accuracy": 0.8575312532484531,
      "num_tokens": 894563512.0,
      "step": 28029
    },
    {
      "entropy": 0.5253060106188059,
      "epoch": 2.5789040891081707,
      "grad_norm": 0.1662156581878662,
      "learning_rate": 1.4039930076363972e-05,
      "loss": 0.5256,
      "mean_token_accuracy": 0.8311821445822716,
      "num_tokens": 894595498.0,
      "step": 28030
    },
    {
      "entropy": 0.617335207760334,
      "epoch": 2.578996095479612,
      "grad_norm": 0.17269033193588257,
      "learning_rate": 1.4036863250222345e-05,
      "loss": 0.604,
      "mean_token_accuracy": 0.8085249923169613,
      "num_tokens": 894627428.0,
      "step": 28031
    },
    {
      "entropy": 0.5327573111280799,
      "epoch": 2.5790881018510534,
      "grad_norm": 0.162765234708786,
      "learning_rate": 1.403379642408072e-05,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8333285488188267,
      "num_tokens": 894659192.0,
      "step": 28032
    },
    {
      "entropy": 0.4672937449067831,
      "epoch": 2.5791801082224945,
      "grad_norm": 0.15355198085308075,
      "learning_rate": 1.4030729597939094e-05,
      "loss": 0.455,
      "mean_token_accuracy": 0.8541759476065636,
      "num_tokens": 894691108.0,
      "step": 28033
    },
    {
      "entropy": 0.5125269778072834,
      "epoch": 2.5792721145939357,
      "grad_norm": 0.15701064467430115,
      "learning_rate": 1.4027662771797467e-05,
      "loss": 0.5058,
      "mean_token_accuracy": 0.8382247723639011,
      "num_tokens": 894723500.0,
      "step": 28034
    },
    {
      "entropy": 0.530600979924202,
      "epoch": 2.579364120965377,
      "grad_norm": 0.1569787710905075,
      "learning_rate": 1.4024595945655842e-05,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8370745964348316,
      "num_tokens": 894755953.0,
      "step": 28035
    },
    {
      "entropy": 0.5259737521409988,
      "epoch": 2.579456127336818,
      "grad_norm": 0.16097977757453918,
      "learning_rate": 1.4021529119514215e-05,
      "loss": 0.5183,
      "mean_token_accuracy": 0.835309911519289,
      "num_tokens": 894788210.0,
      "step": 28036
    },
    {
      "entropy": 0.5013599419035017,
      "epoch": 2.5795481337082595,
      "grad_norm": 0.15994258224964142,
      "learning_rate": 1.401846229337259e-05,
      "loss": 0.4822,
      "mean_token_accuracy": 0.8475572392344475,
      "num_tokens": 894820192.0,
      "step": 28037
    },
    {
      "entropy": 0.5411869417876005,
      "epoch": 2.5796401400797007,
      "grad_norm": 0.16388799250125885,
      "learning_rate": 1.4015395467230962e-05,
      "loss": 0.5279,
      "mean_token_accuracy": 0.8322090208530426,
      "num_tokens": 894851924.0,
      "step": 28038
    },
    {
      "entropy": 0.44918955489993095,
      "epoch": 2.579732146451142,
      "grad_norm": 0.1528763473033905,
      "learning_rate": 1.4012328641089337e-05,
      "loss": 0.4465,
      "mean_token_accuracy": 0.8597537465393543,
      "num_tokens": 894883471.0,
      "step": 28039
    },
    {
      "entropy": 0.5932127479463816,
      "epoch": 2.579824152822583,
      "grad_norm": 0.16600914299488068,
      "learning_rate": 1.400926181494771e-05,
      "loss": 0.5811,
      "mean_token_accuracy": 0.8202552497386932,
      "num_tokens": 894916239.0,
      "step": 28040
    },
    {
      "entropy": 0.47842860128730536,
      "epoch": 2.579916159194024,
      "grad_norm": 0.1614256650209427,
      "learning_rate": 1.4006194988806085e-05,
      "loss": 0.4714,
      "mean_token_accuracy": 0.8490541577339172,
      "num_tokens": 894948281.0,
      "step": 28041
    },
    {
      "entropy": 0.5092188407434151,
      "epoch": 2.5800081655654656,
      "grad_norm": 0.1572858989238739,
      "learning_rate": 1.4003128162664458e-05,
      "loss": 0.5042,
      "mean_token_accuracy": 0.8431836068630219,
      "num_tokens": 894979877.0,
      "step": 28042
    },
    {
      "entropy": 0.5527383973821998,
      "epoch": 2.5801001719369068,
      "grad_norm": 0.16820071637630463,
      "learning_rate": 1.4000061336522832e-05,
      "loss": 0.5491,
      "mean_token_accuracy": 0.8279079608619213,
      "num_tokens": 895010759.0,
      "step": 28043
    },
    {
      "entropy": 0.4994149599224329,
      "epoch": 2.580192178308348,
      "grad_norm": 0.16527633368968964,
      "learning_rate": 1.3996994510381207e-05,
      "loss": 0.4826,
      "mean_token_accuracy": 0.8458071835339069,
      "num_tokens": 895041655.0,
      "step": 28044
    },
    {
      "entropy": 0.5584120508283377,
      "epoch": 2.580284184679789,
      "grad_norm": 0.16187557578086853,
      "learning_rate": 1.399392768423958e-05,
      "loss": 0.5379,
      "mean_token_accuracy": 0.829451572149992,
      "num_tokens": 895073684.0,
      "step": 28045
    },
    {
      "entropy": 0.49890464171767235,
      "epoch": 2.58037619105123,
      "grad_norm": 0.1551336795091629,
      "learning_rate": 1.3990860858097955e-05,
      "loss": 0.4827,
      "mean_token_accuracy": 0.845428816974163,
      "num_tokens": 895105384.0,
      "step": 28046
    },
    {
      "entropy": 0.5740240644663572,
      "epoch": 2.5804681974226718,
      "grad_norm": 0.1656956672668457,
      "learning_rate": 1.3987794031956327e-05,
      "loss": 0.5694,
      "mean_token_accuracy": 0.8206941671669483,
      "num_tokens": 895137525.0,
      "step": 28047
    },
    {
      "entropy": 0.4401195142418146,
      "epoch": 2.580560203794113,
      "grad_norm": 0.15228302776813507,
      "learning_rate": 1.3984727205814704e-05,
      "loss": 0.434,
      "mean_token_accuracy": 0.8617787212133408,
      "num_tokens": 895169216.0,
      "step": 28048
    },
    {
      "entropy": 0.5691486829891801,
      "epoch": 2.580652210165554,
      "grad_norm": 0.16528752446174622,
      "learning_rate": 1.3981660379673078e-05,
      "loss": 0.5497,
      "mean_token_accuracy": 0.8303701281547546,
      "num_tokens": 895200921.0,
      "step": 28049
    },
    {
      "entropy": 0.44684648187831044,
      "epoch": 2.580744216536995,
      "grad_norm": 0.1512184590101242,
      "learning_rate": 1.3978593553531451e-05,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8594813123345375,
      "num_tokens": 895232731.0,
      "step": 28050
    },
    {
      "entropy": 0.49879721738398075,
      "epoch": 2.5808362229084363,
      "grad_norm": 0.154293954372406,
      "learning_rate": 1.3975526727389826e-05,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8442563377320766,
      "num_tokens": 895264577.0,
      "step": 28051
    },
    {
      "entropy": 0.5690373368561268,
      "epoch": 2.580928229279878,
      "grad_norm": 0.16318269073963165,
      "learning_rate": 1.39724599012482e-05,
      "loss": 0.5514,
      "mean_token_accuracy": 0.8246102631092072,
      "num_tokens": 895296135.0,
      "step": 28052
    },
    {
      "entropy": 0.5370885087177157,
      "epoch": 2.581020235651319,
      "grad_norm": 0.1633731722831726,
      "learning_rate": 1.3969393075106574e-05,
      "loss": 0.5282,
      "mean_token_accuracy": 0.834173571318388,
      "num_tokens": 895327846.0,
      "step": 28053
    },
    {
      "entropy": 0.5426714075729251,
      "epoch": 2.58111224202276,
      "grad_norm": 0.16365137696266174,
      "learning_rate": 1.3966326248964948e-05,
      "loss": 0.5297,
      "mean_token_accuracy": 0.834359809756279,
      "num_tokens": 895360229.0,
      "step": 28054
    },
    {
      "entropy": 0.5628036791458726,
      "epoch": 2.5812042483942013,
      "grad_norm": 0.16550210118293762,
      "learning_rate": 1.3963259422823321e-05,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8286402076482773,
      "num_tokens": 895391749.0,
      "step": 28055
    },
    {
      "entropy": 0.5132559109479189,
      "epoch": 2.5812962547656424,
      "grad_norm": 0.15644866228103638,
      "learning_rate": 1.3960192596681696e-05,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8438626416027546,
      "num_tokens": 895423007.0,
      "step": 28056
    },
    {
      "entropy": 0.5613206876441836,
      "epoch": 2.581388261137084,
      "grad_norm": 0.16741624474525452,
      "learning_rate": 1.3957125770540069e-05,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8287477120757103,
      "num_tokens": 895454983.0,
      "step": 28057
    },
    {
      "entropy": 0.5484776608645916,
      "epoch": 2.581480267508525,
      "grad_norm": 0.16045522689819336,
      "learning_rate": 1.3954058944398444e-05,
      "loss": 0.5398,
      "mean_token_accuracy": 0.8281908743083477,
      "num_tokens": 895487156.0,
      "step": 28058
    },
    {
      "entropy": 0.45663213124498725,
      "epoch": 2.5815722738799662,
      "grad_norm": 0.14967641234397888,
      "learning_rate": 1.3950992118256816e-05,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8627949580550194,
      "num_tokens": 895518476.0,
      "step": 28059
    },
    {
      "entropy": 0.5138166444376111,
      "epoch": 2.5816642802514074,
      "grad_norm": 0.16146288812160492,
      "learning_rate": 1.3947925292115191e-05,
      "loss": 0.5143,
      "mean_token_accuracy": 0.8354360088706017,
      "num_tokens": 895549980.0,
      "step": 28060
    },
    {
      "entropy": 0.6107199843972921,
      "epoch": 2.5817562866228485,
      "grad_norm": 0.17424191534519196,
      "learning_rate": 1.3944858465973564e-05,
      "loss": 0.5997,
      "mean_token_accuracy": 0.8119589425623417,
      "num_tokens": 895581238.0,
      "step": 28061
    },
    {
      "entropy": 0.4666599901393056,
      "epoch": 2.58184829299429,
      "grad_norm": 0.1558225452899933,
      "learning_rate": 1.3941791639831939e-05,
      "loss": 0.4539,
      "mean_token_accuracy": 0.8580175787210464,
      "num_tokens": 895613500.0,
      "step": 28062
    },
    {
      "entropy": 0.5009625111706555,
      "epoch": 2.581940299365731,
      "grad_norm": 0.15377472341060638,
      "learning_rate": 1.3938724813690313e-05,
      "loss": 0.4932,
      "mean_token_accuracy": 0.8433098755776882,
      "num_tokens": 895645660.0,
      "step": 28063
    },
    {
      "entropy": 0.4537616418674588,
      "epoch": 2.5820323057371724,
      "grad_norm": 0.15725396573543549,
      "learning_rate": 1.3935657987548686e-05,
      "loss": 0.4559,
      "mean_token_accuracy": 0.8542776256799698,
      "num_tokens": 895677626.0,
      "step": 28064
    },
    {
      "entropy": 0.5768604654585943,
      "epoch": 2.5821243121086135,
      "grad_norm": 0.16634249687194824,
      "learning_rate": 1.3932591161407061e-05,
      "loss": 0.5577,
      "mean_token_accuracy": 0.8256123140454292,
      "num_tokens": 895708527.0,
      "step": 28065
    },
    {
      "entropy": 0.5477834455668926,
      "epoch": 2.5822163184800546,
      "grad_norm": 0.1684585064649582,
      "learning_rate": 1.3929524335265434e-05,
      "loss": 0.5433,
      "mean_token_accuracy": 0.8296691551804543,
      "num_tokens": 895740048.0,
      "step": 28066
    },
    {
      "entropy": 0.5007168147712946,
      "epoch": 2.582308324851496,
      "grad_norm": 0.15485863387584686,
      "learning_rate": 1.3926457509123809e-05,
      "loss": 0.4949,
      "mean_token_accuracy": 0.8430708199739456,
      "num_tokens": 895772485.0,
      "step": 28067
    },
    {
      "entropy": 0.4835200789384544,
      "epoch": 2.5824003312229373,
      "grad_norm": 0.1594993621110916,
      "learning_rate": 1.3923390682982182e-05,
      "loss": 0.4731,
      "mean_token_accuracy": 0.8510464914143085,
      "num_tokens": 895803593.0,
      "step": 28068
    },
    {
      "entropy": 0.49811689369380474,
      "epoch": 2.5824923375943785,
      "grad_norm": 0.15580376982688904,
      "learning_rate": 1.3920323856840556e-05,
      "loss": 0.49,
      "mean_token_accuracy": 0.8403738141059875,
      "num_tokens": 895835871.0,
      "step": 28069
    },
    {
      "entropy": 0.5030825799331069,
      "epoch": 2.5825843439658196,
      "grad_norm": 0.15520499646663666,
      "learning_rate": 1.3917257030698929e-05,
      "loss": 0.494,
      "mean_token_accuracy": 0.838165108114481,
      "num_tokens": 895868213.0,
      "step": 28070
    },
    {
      "entropy": 0.6250587496906519,
      "epoch": 2.5826763503372607,
      "grad_norm": 0.17217500507831573,
      "learning_rate": 1.3914190204557304e-05,
      "loss": 0.6135,
      "mean_token_accuracy": 0.8040158823132515,
      "num_tokens": 895899689.0,
      "step": 28071
    },
    {
      "entropy": 0.44586161943152547,
      "epoch": 2.5827683567087023,
      "grad_norm": 0.15170100331306458,
      "learning_rate": 1.3911123378415677e-05,
      "loss": 0.4337,
      "mean_token_accuracy": 0.8598935939371586,
      "num_tokens": 895931813.0,
      "step": 28072
    },
    {
      "entropy": 0.552742762491107,
      "epoch": 2.5828603630801434,
      "grad_norm": 0.16675831377506256,
      "learning_rate": 1.3908056552274051e-05,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8291565328836441,
      "num_tokens": 895963845.0,
      "step": 28073
    },
    {
      "entropy": 0.5019473361317068,
      "epoch": 2.5829523694515846,
      "grad_norm": 0.1564164012670517,
      "learning_rate": 1.3904989726132426e-05,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8430112674832344,
      "num_tokens": 895995699.0,
      "step": 28074
    },
    {
      "entropy": 0.5156774874776602,
      "epoch": 2.5830443758230257,
      "grad_norm": 0.1638736128807068,
      "learning_rate": 1.3901922899990799e-05,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8388731926679611,
      "num_tokens": 896027535.0,
      "step": 28075
    },
    {
      "entropy": 0.4858387657441199,
      "epoch": 2.583136382194467,
      "grad_norm": 0.15691940486431122,
      "learning_rate": 1.3898856073849174e-05,
      "loss": 0.4647,
      "mean_token_accuracy": 0.8493615463376045,
      "num_tokens": 896058674.0,
      "step": 28076
    },
    {
      "entropy": 0.5867675887420774,
      "epoch": 2.5832283885659084,
      "grad_norm": 0.16761384904384613,
      "learning_rate": 1.3895789247707547e-05,
      "loss": 0.5672,
      "mean_token_accuracy": 0.8213761299848557,
      "num_tokens": 896090089.0,
      "step": 28077
    },
    {
      "entropy": 0.5303914789110422,
      "epoch": 2.5833203949373496,
      "grad_norm": 0.1650618016719818,
      "learning_rate": 1.3892722421565921e-05,
      "loss": 0.508,
      "mean_token_accuracy": 0.8380478881299496,
      "num_tokens": 896122245.0,
      "step": 28078
    },
    {
      "entropy": 0.5416120456065983,
      "epoch": 2.5834124013087907,
      "grad_norm": 0.16243262588977814,
      "learning_rate": 1.3889655595424298e-05,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8304809369146824,
      "num_tokens": 896154255.0,
      "step": 28079
    },
    {
      "entropy": 0.535273402929306,
      "epoch": 2.583504407680232,
      "grad_norm": 0.16514497995376587,
      "learning_rate": 1.388658876928267e-05,
      "loss": 0.5322,
      "mean_token_accuracy": 0.8342354819178581,
      "num_tokens": 896186237.0,
      "step": 28080
    },
    {
      "entropy": 0.5277675869874656,
      "epoch": 2.583596414051673,
      "grad_norm": 0.1613447368144989,
      "learning_rate": 1.3883521943141045e-05,
      "loss": 0.5072,
      "mean_token_accuracy": 0.8410305120050907,
      "num_tokens": 896217201.0,
      "step": 28081
    },
    {
      "entropy": 0.5497281309217215,
      "epoch": 2.5836884204231145,
      "grad_norm": 0.16616424918174744,
      "learning_rate": 1.388045511699942e-05,
      "loss": 0.5419,
      "mean_token_accuracy": 0.8245006501674652,
      "num_tokens": 896248840.0,
      "step": 28082
    },
    {
      "entropy": 0.5969899748452008,
      "epoch": 2.5837804267945557,
      "grad_norm": 0.16799160838127136,
      "learning_rate": 1.3877388290857793e-05,
      "loss": 0.5944,
      "mean_token_accuracy": 0.8156821355223656,
      "num_tokens": 896280003.0,
      "step": 28083
    },
    {
      "entropy": 0.5058542157057673,
      "epoch": 2.583872433165997,
      "grad_norm": 0.1537942737340927,
      "learning_rate": 1.3874321464716167e-05,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8445666991174221,
      "num_tokens": 896312431.0,
      "step": 28084
    },
    {
      "entropy": 0.5488592023029923,
      "epoch": 2.583964439537438,
      "grad_norm": 0.16077148914337158,
      "learning_rate": 1.387125463857454e-05,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8325539454817772,
      "num_tokens": 896345063.0,
      "step": 28085
    },
    {
      "entropy": 0.464961220510304,
      "epoch": 2.584056445908879,
      "grad_norm": 0.15706725418567657,
      "learning_rate": 1.3868187812432915e-05,
      "loss": 0.4464,
      "mean_token_accuracy": 0.8547363430261612,
      "num_tokens": 896376588.0,
      "step": 28086
    },
    {
      "entropy": 0.5551435314118862,
      "epoch": 2.5841484522803206,
      "grad_norm": 0.16575460135936737,
      "learning_rate": 1.3865120986291288e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8250127770006657,
      "num_tokens": 896408750.0,
      "step": 28087
    },
    {
      "entropy": 0.5103156855329871,
      "epoch": 2.5842404586517618,
      "grad_norm": 0.16284950077533722,
      "learning_rate": 1.3862054160149663e-05,
      "loss": 0.4995,
      "mean_token_accuracy": 0.840576309710741,
      "num_tokens": 896439889.0,
      "step": 28088
    },
    {
      "entropy": 0.44389857165515423,
      "epoch": 2.584332465023203,
      "grad_norm": 0.153767928481102,
      "learning_rate": 1.3858987334008036e-05,
      "loss": 0.4254,
      "mean_token_accuracy": 0.8643326684832573,
      "num_tokens": 896470931.0,
      "step": 28089
    },
    {
      "entropy": 0.5983151625841856,
      "epoch": 2.584424471394644,
      "grad_norm": 0.17362891137599945,
      "learning_rate": 1.385592050786641e-05,
      "loss": 0.5863,
      "mean_token_accuracy": 0.8162340596318245,
      "num_tokens": 896502164.0,
      "step": 28090
    },
    {
      "entropy": 0.6161685101687908,
      "epoch": 2.584516477766085,
      "grad_norm": 0.1797454059123993,
      "learning_rate": 1.3852853681724783e-05,
      "loss": 0.6161,
      "mean_token_accuracy": 0.8070651218295097,
      "num_tokens": 896534138.0,
      "step": 28091
    },
    {
      "entropy": 0.5018267226405442,
      "epoch": 2.5846084841375268,
      "grad_norm": 0.16087229549884796,
      "learning_rate": 1.3849786855583158e-05,
      "loss": 0.4999,
      "mean_token_accuracy": 0.839452650398016,
      "num_tokens": 896566395.0,
      "step": 28092
    },
    {
      "entropy": 0.5283719295402989,
      "epoch": 2.584700490508968,
      "grad_norm": 0.1634930521249771,
      "learning_rate": 1.3846720029441532e-05,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8328247405588627,
      "num_tokens": 896598285.0,
      "step": 28093
    },
    {
      "entropy": 0.5347304202150553,
      "epoch": 2.584792496880409,
      "grad_norm": 0.16230642795562744,
      "learning_rate": 1.3843653203299905e-05,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8382368944585323,
      "num_tokens": 896630008.0,
      "step": 28094
    },
    {
      "entropy": 0.48325309669598937,
      "epoch": 2.58488450325185,
      "grad_norm": 0.16389983892440796,
      "learning_rate": 1.384058637715828e-05,
      "loss": 0.474,
      "mean_token_accuracy": 0.8491056561470032,
      "num_tokens": 896661805.0,
      "step": 28095
    },
    {
      "entropy": 0.5282084187492728,
      "epoch": 2.5849765096232913,
      "grad_norm": 0.16492009162902832,
      "learning_rate": 1.3837519551016653e-05,
      "loss": 0.5102,
      "mean_token_accuracy": 0.8386507146060467,
      "num_tokens": 896693849.0,
      "step": 28096
    },
    {
      "entropy": 0.41410330520011485,
      "epoch": 2.585068515994733,
      "grad_norm": 0.14329518377780914,
      "learning_rate": 1.3834452724875028e-05,
      "loss": 0.3985,
      "mean_token_accuracy": 0.8704049624502659,
      "num_tokens": 896726345.0,
      "step": 28097
    },
    {
      "entropy": 0.5864086654037237,
      "epoch": 2.585160522366174,
      "grad_norm": 0.16220389306545258,
      "learning_rate": 1.38313858987334e-05,
      "loss": 0.5726,
      "mean_token_accuracy": 0.8194712549448013,
      "num_tokens": 896758698.0,
      "step": 28098
    },
    {
      "entropy": 0.4630719097331166,
      "epoch": 2.585252528737615,
      "grad_norm": 0.15829399228096008,
      "learning_rate": 1.3828319072591775e-05,
      "loss": 0.4449,
      "mean_token_accuracy": 0.8604946956038475,
      "num_tokens": 896790262.0,
      "step": 28099
    },
    {
      "entropy": 0.5013225907459855,
      "epoch": 2.5853445351090563,
      "grad_norm": 0.15748651325702667,
      "learning_rate": 1.3825252246450148e-05,
      "loss": 0.48,
      "mean_token_accuracy": 0.8459813185036182,
      "num_tokens": 896821688.0,
      "step": 28100
    },
    {
      "entropy": 0.5188251626677811,
      "epoch": 2.5854365414804974,
      "grad_norm": 0.1605345904827118,
      "learning_rate": 1.3822185420308523e-05,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8405962251126766,
      "num_tokens": 896853386.0,
      "step": 28101
    },
    {
      "entropy": 0.49952845973894,
      "epoch": 2.585528547851939,
      "grad_norm": 0.15931960940361023,
      "learning_rate": 1.3819118594166896e-05,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8438686653971672,
      "num_tokens": 896885172.0,
      "step": 28102
    },
    {
      "entropy": 0.5064111277461052,
      "epoch": 2.58562055422338,
      "grad_norm": 0.15768472850322723,
      "learning_rate": 1.381605176802527e-05,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8408503048121929,
      "num_tokens": 896916640.0,
      "step": 28103
    },
    {
      "entropy": 0.5204603523015976,
      "epoch": 2.5857125605948212,
      "grad_norm": 0.1631646305322647,
      "learning_rate": 1.3812984941883645e-05,
      "loss": 0.5114,
      "mean_token_accuracy": 0.8391935117542744,
      "num_tokens": 896947309.0,
      "step": 28104
    },
    {
      "entropy": 0.46427122689783573,
      "epoch": 2.5858045669662624,
      "grad_norm": 0.15113654732704163,
      "learning_rate": 1.3809918115742018e-05,
      "loss": 0.4539,
      "mean_token_accuracy": 0.8600696101784706,
      "num_tokens": 896980077.0,
      "step": 28105
    },
    {
      "entropy": 0.48281023371964693,
      "epoch": 2.5858965733377035,
      "grad_norm": 0.15038348734378815,
      "learning_rate": 1.3806851289600393e-05,
      "loss": 0.475,
      "mean_token_accuracy": 0.8516939133405685,
      "num_tokens": 897012803.0,
      "step": 28106
    },
    {
      "entropy": 0.5519025269895792,
      "epoch": 2.585988579709145,
      "grad_norm": 0.16055670380592346,
      "learning_rate": 1.3803784463458766e-05,
      "loss": 0.5363,
      "mean_token_accuracy": 0.831156499683857,
      "num_tokens": 897044971.0,
      "step": 28107
    },
    {
      "entropy": 0.5863764397799969,
      "epoch": 2.586080586080586,
      "grad_norm": 0.16670836508274078,
      "learning_rate": 1.380071763731714e-05,
      "loss": 0.5721,
      "mean_token_accuracy": 0.820870291441679,
      "num_tokens": 897076689.0,
      "step": 28108
    },
    {
      "entropy": 0.5244038774399087,
      "epoch": 2.5861725924520274,
      "grad_norm": 0.1596129685640335,
      "learning_rate": 1.3797650811175517e-05,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8401795886456966,
      "num_tokens": 897108195.0,
      "step": 28109
    },
    {
      "entropy": 0.4624980300432071,
      "epoch": 2.5862645988234685,
      "grad_norm": 0.15080475807189941,
      "learning_rate": 1.379458398503389e-05,
      "loss": 0.447,
      "mean_token_accuracy": 0.8591403551399708,
      "num_tokens": 897140241.0,
      "step": 28110
    },
    {
      "entropy": 0.47069514042232186,
      "epoch": 2.5863566051949096,
      "grad_norm": 0.1544683575630188,
      "learning_rate": 1.3791517158892264e-05,
      "loss": 0.446,
      "mean_token_accuracy": 0.8575402423739433,
      "num_tokens": 897172398.0,
      "step": 28111
    },
    {
      "entropy": 0.4854198950342834,
      "epoch": 2.586448611566351,
      "grad_norm": 0.15724247694015503,
      "learning_rate": 1.3788450332750639e-05,
      "loss": 0.4725,
      "mean_token_accuracy": 0.8511341214179993,
      "num_tokens": 897204402.0,
      "step": 28112
    },
    {
      "entropy": 0.5423777238465846,
      "epoch": 2.5865406179377923,
      "grad_norm": 0.16587205231189728,
      "learning_rate": 1.3785383506609012e-05,
      "loss": 0.535,
      "mean_token_accuracy": 0.8304918520152569,
      "num_tokens": 897236310.0,
      "step": 28113
    },
    {
      "entropy": 0.4974405742250383,
      "epoch": 2.5866326243092335,
      "grad_norm": 0.1653268188238144,
      "learning_rate": 1.3782316680467387e-05,
      "loss": 0.4929,
      "mean_token_accuracy": 0.8451525419950485,
      "num_tokens": 897268097.0,
      "step": 28114
    },
    {
      "entropy": 0.5960065769031644,
      "epoch": 2.5867246306806746,
      "grad_norm": 0.16792698204517365,
      "learning_rate": 1.377924985432576e-05,
      "loss": 0.5749,
      "mean_token_accuracy": 0.8247969336807728,
      "num_tokens": 897299720.0,
      "step": 28115
    },
    {
      "entropy": 0.4636851381510496,
      "epoch": 2.5868166370521157,
      "grad_norm": 0.15643127262592316,
      "learning_rate": 1.3776183028184134e-05,
      "loss": 0.4581,
      "mean_token_accuracy": 0.854543287307024,
      "num_tokens": 897332298.0,
      "step": 28116
    },
    {
      "entropy": 0.48482141736894846,
      "epoch": 2.5869086434235573,
      "grad_norm": 0.15834233164787292,
      "learning_rate": 1.3773116202042507e-05,
      "loss": 0.4778,
      "mean_token_accuracy": 0.8460519425570965,
      "num_tokens": 897363875.0,
      "step": 28117
    },
    {
      "entropy": 0.5701260324567556,
      "epoch": 2.5870006497949984,
      "grad_norm": 0.1660914421081543,
      "learning_rate": 1.3770049375900882e-05,
      "loss": 0.5547,
      "mean_token_accuracy": 0.8225078135728836,
      "num_tokens": 897395204.0,
      "step": 28118
    },
    {
      "entropy": 0.5696051390841603,
      "epoch": 2.5870926561664396,
      "grad_norm": 0.16522985696792603,
      "learning_rate": 1.3766982549759255e-05,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8240825720131397,
      "num_tokens": 897426934.0,
      "step": 28119
    },
    {
      "entropy": 0.5934390742331743,
      "epoch": 2.5871846625378807,
      "grad_norm": 0.16817884147167206,
      "learning_rate": 1.376391572361763e-05,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8148620091378689,
      "num_tokens": 897458417.0,
      "step": 28120
    },
    {
      "entropy": 0.4974275156855583,
      "epoch": 2.587276668909322,
      "grad_norm": 0.17376306653022766,
      "learning_rate": 1.3760848897476002e-05,
      "loss": 0.4992,
      "mean_token_accuracy": 0.8439159318804741,
      "num_tokens": 897490985.0,
      "step": 28121
    },
    {
      "entropy": 0.464869195362553,
      "epoch": 2.5873686752807634,
      "grad_norm": 0.1588217169046402,
      "learning_rate": 1.3757782071334377e-05,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8579414188861847,
      "num_tokens": 897522398.0,
      "step": 28122
    },
    {
      "entropy": 0.5667077712714672,
      "epoch": 2.5874606816522046,
      "grad_norm": 0.16997811198234558,
      "learning_rate": 1.3754715245192752e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8262869380414486,
      "num_tokens": 897554567.0,
      "step": 28123
    },
    {
      "entropy": 0.5091244354844093,
      "epoch": 2.5875526880236457,
      "grad_norm": 0.16101105511188507,
      "learning_rate": 1.3751648419051125e-05,
      "loss": 0.4979,
      "mean_token_accuracy": 0.8402592316269875,
      "num_tokens": 897586791.0,
      "step": 28124
    },
    {
      "entropy": 0.4045887663960457,
      "epoch": 2.587644694395087,
      "grad_norm": 0.14107507467269897,
      "learning_rate": 1.37485815929095e-05,
      "loss": 0.3951,
      "mean_token_accuracy": 0.8773943148553371,
      "num_tokens": 897618317.0,
      "step": 28125
    },
    {
      "entropy": 0.5108676645904779,
      "epoch": 2.587736700766528,
      "grad_norm": 0.15767265856266022,
      "learning_rate": 1.3745514766767872e-05,
      "loss": 0.5005,
      "mean_token_accuracy": 0.8423679657280445,
      "num_tokens": 897650410.0,
      "step": 28126
    },
    {
      "entropy": 0.5567603493109345,
      "epoch": 2.5878287071379695,
      "grad_norm": 0.16422688961029053,
      "learning_rate": 1.3742447940626247e-05,
      "loss": 0.5525,
      "mean_token_accuracy": 0.8255249485373497,
      "num_tokens": 897682797.0,
      "step": 28127
    },
    {
      "entropy": 0.4329135335283354,
      "epoch": 2.5879207135094107,
      "grad_norm": 0.14689917862415314,
      "learning_rate": 1.373938111448462e-05,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8637078516185284,
      "num_tokens": 897715346.0,
      "step": 28128
    },
    {
      "entropy": 0.5837183799594641,
      "epoch": 2.588012719880852,
      "grad_norm": 0.17107032239437103,
      "learning_rate": 1.3736314288342994e-05,
      "loss": 0.5821,
      "mean_token_accuracy": 0.8135190904140472,
      "num_tokens": 897746979.0,
      "step": 28129
    },
    {
      "entropy": 0.5025389771908522,
      "epoch": 2.588104726252293,
      "grad_norm": 0.15824845433235168,
      "learning_rate": 1.3733247462201367e-05,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8419886641204357,
      "num_tokens": 897778862.0,
      "step": 28130
    },
    {
      "entropy": 0.5912769790738821,
      "epoch": 2.588196732623734,
      "grad_norm": 0.16942313313484192,
      "learning_rate": 1.3730180636059742e-05,
      "loss": 0.5831,
      "mean_token_accuracy": 0.8139793500304222,
      "num_tokens": 897811078.0,
      "step": 28131
    },
    {
      "entropy": 0.5327993538230658,
      "epoch": 2.5882887389951756,
      "grad_norm": 0.16603243350982666,
      "learning_rate": 1.3727113809918115e-05,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8303708881139755,
      "num_tokens": 897842140.0,
      "step": 28132
    },
    {
      "entropy": 0.4971094075590372,
      "epoch": 2.5883807453666168,
      "grad_norm": 0.1539725959300995,
      "learning_rate": 1.372404698377649e-05,
      "loss": 0.4823,
      "mean_token_accuracy": 0.84947245195508,
      "num_tokens": 897873106.0,
      "step": 28133
    },
    {
      "entropy": 0.5547632835805416,
      "epoch": 2.588472751738058,
      "grad_norm": 0.16235966980457306,
      "learning_rate": 1.3720980157634864e-05,
      "loss": 0.541,
      "mean_token_accuracy": 0.830459713935852,
      "num_tokens": 897905672.0,
      "step": 28134
    },
    {
      "entropy": 0.5603301357477903,
      "epoch": 2.588564758109499,
      "grad_norm": 0.17296938598155975,
      "learning_rate": 1.3717913331493237e-05,
      "loss": 0.5479,
      "mean_token_accuracy": 0.8271014504134655,
      "num_tokens": 897936876.0,
      "step": 28135
    },
    {
      "entropy": 0.41916465293616056,
      "epoch": 2.58865676448094,
      "grad_norm": 0.15447041392326355,
      "learning_rate": 1.3714846505351612e-05,
      "loss": 0.4061,
      "mean_token_accuracy": 0.8687751404941082,
      "num_tokens": 897967767.0,
      "step": 28136
    },
    {
      "entropy": 0.575323473662138,
      "epoch": 2.5887487708523818,
      "grad_norm": 0.16727571189403534,
      "learning_rate": 1.3711779679209985e-05,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8223921284079552,
      "num_tokens": 897999399.0,
      "step": 28137
    },
    {
      "entropy": 0.5494572427123785,
      "epoch": 2.588840777223823,
      "grad_norm": 0.16156508028507233,
      "learning_rate": 1.370871285306836e-05,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8315045349299908,
      "num_tokens": 898031156.0,
      "step": 28138
    },
    {
      "entropy": 0.44597647804766893,
      "epoch": 2.588932783595264,
      "grad_norm": 0.15466099977493286,
      "learning_rate": 1.3705646026926732e-05,
      "loss": 0.4272,
      "mean_token_accuracy": 0.8612591326236725,
      "num_tokens": 898062707.0,
      "step": 28139
    },
    {
      "entropy": 0.5774722527712584,
      "epoch": 2.589024789966705,
      "grad_norm": 0.1713021993637085,
      "learning_rate": 1.3702579200785109e-05,
      "loss": 0.5747,
      "mean_token_accuracy": 0.820287611335516,
      "num_tokens": 898094137.0,
      "step": 28140
    },
    {
      "entropy": 0.512925012037158,
      "epoch": 2.5891167963381463,
      "grad_norm": 0.15929171442985535,
      "learning_rate": 1.3699512374643483e-05,
      "loss": 0.4929,
      "mean_token_accuracy": 0.8410739414393902,
      "num_tokens": 898125743.0,
      "step": 28141
    },
    {
      "entropy": 0.5342249721288681,
      "epoch": 2.589208802709588,
      "grad_norm": 0.15747946500778198,
      "learning_rate": 1.3696445548501858e-05,
      "loss": 0.5197,
      "mean_token_accuracy": 0.837272584438324,
      "num_tokens": 898157283.0,
      "step": 28142
    },
    {
      "entropy": 0.6254636561498046,
      "epoch": 2.589300809081029,
      "grad_norm": 0.17150095105171204,
      "learning_rate": 1.3693378722360231e-05,
      "loss": 0.6137,
      "mean_token_accuracy": 0.8090913705527782,
      "num_tokens": 898189139.0,
      "step": 28143
    },
    {
      "entropy": 0.4516850453801453,
      "epoch": 2.58939281545247,
      "grad_norm": 0.15073958039283752,
      "learning_rate": 1.3690311896218606e-05,
      "loss": 0.4391,
      "mean_token_accuracy": 0.8601688668131828,
      "num_tokens": 898220924.0,
      "step": 28144
    },
    {
      "entropy": 0.5217934674583375,
      "epoch": 2.5894848218239113,
      "grad_norm": 0.15693868696689606,
      "learning_rate": 1.3687245070076979e-05,
      "loss": 0.5068,
      "mean_token_accuracy": 0.8364897966384888,
      "num_tokens": 898252840.0,
      "step": 28145
    },
    {
      "entropy": 0.4867479018867016,
      "epoch": 2.5895768281953524,
      "grad_norm": 0.15914104878902435,
      "learning_rate": 1.3684178243935353e-05,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8493016473948956,
      "num_tokens": 898284441.0,
      "step": 28146
    },
    {
      "entropy": 0.5367108695209026,
      "epoch": 2.589668834566794,
      "grad_norm": 0.16238528490066528,
      "learning_rate": 1.3681111417793726e-05,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8332536220550537,
      "num_tokens": 898316233.0,
      "step": 28147
    },
    {
      "entropy": 0.49243561178445816,
      "epoch": 2.589760840938235,
      "grad_norm": 0.16090337932109833,
      "learning_rate": 1.36780445916521e-05,
      "loss": 0.4747,
      "mean_token_accuracy": 0.8500645980238914,
      "num_tokens": 898347815.0,
      "step": 28148
    },
    {
      "entropy": 0.44714033510535955,
      "epoch": 2.5898528473096762,
      "grad_norm": 0.15217120945453644,
      "learning_rate": 1.3674977765510474e-05,
      "loss": 0.4371,
      "mean_token_accuracy": 0.8610686883330345,
      "num_tokens": 898379465.0,
      "step": 28149
    },
    {
      "entropy": 0.5721219377592206,
      "epoch": 2.5899448536811174,
      "grad_norm": 0.16374215483665466,
      "learning_rate": 1.3671910939368848e-05,
      "loss": 0.564,
      "mean_token_accuracy": 0.8244470059871674,
      "num_tokens": 898411680.0,
      "step": 28150
    },
    {
      "entropy": 0.48351468332111835,
      "epoch": 2.5900368600525585,
      "grad_norm": 0.1577019840478897,
      "learning_rate": 1.3668844113227221e-05,
      "loss": 0.4739,
      "mean_token_accuracy": 0.8491157963871956,
      "num_tokens": 898443178.0,
      "step": 28151
    },
    {
      "entropy": 0.4174171332269907,
      "epoch": 2.590128866424,
      "grad_norm": 0.15485750138759613,
      "learning_rate": 1.3665777287085596e-05,
      "loss": 0.4053,
      "mean_token_accuracy": 0.8692650832235813,
      "num_tokens": 898474128.0,
      "step": 28152
    },
    {
      "entropy": 0.5354734202846885,
      "epoch": 2.590220872795441,
      "grad_norm": 0.1623038351535797,
      "learning_rate": 1.366271046094397e-05,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8334096930921078,
      "num_tokens": 898505604.0,
      "step": 28153
    },
    {
      "entropy": 0.5390862431377172,
      "epoch": 2.5903128791668824,
      "grad_norm": 0.16440820693969727,
      "learning_rate": 1.3659643634802344e-05,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8328700885176659,
      "num_tokens": 898537504.0,
      "step": 28154
    },
    {
      "entropy": 0.4648183537647128,
      "epoch": 2.5904048855383235,
      "grad_norm": 0.15299957990646362,
      "learning_rate": 1.3656576808660718e-05,
      "loss": 0.4497,
      "mean_token_accuracy": 0.8596135564148426,
      "num_tokens": 898569502.0,
      "step": 28155
    },
    {
      "entropy": 0.573387511074543,
      "epoch": 2.5904968919097646,
      "grad_norm": 0.16137684881687164,
      "learning_rate": 1.3653509982519091e-05,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8308342099189758,
      "num_tokens": 898601921.0,
      "step": 28156
    },
    {
      "entropy": 0.5649403017014265,
      "epoch": 2.590588898281206,
      "grad_norm": 0.165973499417305,
      "learning_rate": 1.3650443156377466e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8287822566926479,
      "num_tokens": 898633111.0,
      "step": 28157
    },
    {
      "entropy": 0.4600669788196683,
      "epoch": 2.5906809046526473,
      "grad_norm": 0.15107321739196777,
      "learning_rate": 1.3647376330235839e-05,
      "loss": 0.4318,
      "mean_token_accuracy": 0.8614717200398445,
      "num_tokens": 898664945.0,
      "step": 28158
    },
    {
      "entropy": 0.48311526887118816,
      "epoch": 2.5907729110240885,
      "grad_norm": 0.1498389095067978,
      "learning_rate": 1.3644309504094213e-05,
      "loss": 0.4825,
      "mean_token_accuracy": 0.848312608897686,
      "num_tokens": 898697155.0,
      "step": 28159
    },
    {
      "entropy": 0.48804345913231373,
      "epoch": 2.5908649173955296,
      "grad_norm": 0.16024306416511536,
      "learning_rate": 1.3641242677952586e-05,
      "loss": 0.4797,
      "mean_token_accuracy": 0.8455849140882492,
      "num_tokens": 898728532.0,
      "step": 28160
    },
    {
      "entropy": 0.5466662580147386,
      "epoch": 2.5909569237669707,
      "grad_norm": 0.16228656470775604,
      "learning_rate": 1.3638175851810961e-05,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8273472487926483,
      "num_tokens": 898760791.0,
      "step": 28161
    },
    {
      "entropy": 0.4722052961587906,
      "epoch": 2.5910489301384123,
      "grad_norm": 0.1551559716463089,
      "learning_rate": 1.3635109025669334e-05,
      "loss": 0.4646,
      "mean_token_accuracy": 0.8519289195537567,
      "num_tokens": 898792923.0,
      "step": 28162
    },
    {
      "entropy": 0.5833615465089679,
      "epoch": 2.5911409365098534,
      "grad_norm": 0.1672520488500595,
      "learning_rate": 1.3632042199527709e-05,
      "loss": 0.5809,
      "mean_token_accuracy": 0.8210913278162479,
      "num_tokens": 898824985.0,
      "step": 28163
    },
    {
      "entropy": 0.5385571904480457,
      "epoch": 2.5912329428812946,
      "grad_norm": 0.1642249971628189,
      "learning_rate": 1.3628975373386083e-05,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8338269926607609,
      "num_tokens": 898857579.0,
      "step": 28164
    },
    {
      "entropy": 0.5614807680249214,
      "epoch": 2.5913249492527357,
      "grad_norm": 0.16910722851753235,
      "learning_rate": 1.3625908547244456e-05,
      "loss": 0.5628,
      "mean_token_accuracy": 0.8274957202374935,
      "num_tokens": 898889733.0,
      "step": 28165
    },
    {
      "entropy": 0.49419688433408737,
      "epoch": 2.591416955624177,
      "grad_norm": 0.1620074063539505,
      "learning_rate": 1.3622841721102831e-05,
      "loss": 0.495,
      "mean_token_accuracy": 0.8445340134203434,
      "num_tokens": 898921819.0,
      "step": 28166
    },
    {
      "entropy": 0.46933995094150305,
      "epoch": 2.5915089619956184,
      "grad_norm": 0.15486064553260803,
      "learning_rate": 1.3619774894961204e-05,
      "loss": 0.4616,
      "mean_token_accuracy": 0.8528038784861565,
      "num_tokens": 898953451.0,
      "step": 28167
    },
    {
      "entropy": 0.48801031708717346,
      "epoch": 2.5916009683670596,
      "grad_norm": 0.15753374993801117,
      "learning_rate": 1.3616708068819579e-05,
      "loss": 0.4789,
      "mean_token_accuracy": 0.847422331571579,
      "num_tokens": 898985263.0,
      "step": 28168
    },
    {
      "entropy": 0.5413882434368134,
      "epoch": 2.5916929747385007,
      "grad_norm": 0.16110074520111084,
      "learning_rate": 1.3613641242677951e-05,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8364978469908237,
      "num_tokens": 899016668.0,
      "step": 28169
    },
    {
      "entropy": 0.4558967687189579,
      "epoch": 2.591784981109942,
      "grad_norm": 0.15647286176681519,
      "learning_rate": 1.3610574416536328e-05,
      "loss": 0.4427,
      "mean_token_accuracy": 0.855718769133091,
      "num_tokens": 899048527.0,
      "step": 28170
    },
    {
      "entropy": 0.5605791937559843,
      "epoch": 2.591876987481383,
      "grad_norm": 0.16647058725357056,
      "learning_rate": 1.3607507590394702e-05,
      "loss": 0.5521,
      "mean_token_accuracy": 0.8282223492860794,
      "num_tokens": 899080515.0,
      "step": 28171
    },
    {
      "entropy": 0.5588134815916419,
      "epoch": 2.5919689938528245,
      "grad_norm": 0.16883176565170288,
      "learning_rate": 1.3604440764253077e-05,
      "loss": 0.5494,
      "mean_token_accuracy": 0.8244894966483116,
      "num_tokens": 899112549.0,
      "step": 28172
    },
    {
      "entropy": 0.5100271739065647,
      "epoch": 2.5920610002242657,
      "grad_norm": 0.16155996918678284,
      "learning_rate": 1.360137393811145e-05,
      "loss": 0.5044,
      "mean_token_accuracy": 0.8370888456702232,
      "num_tokens": 899144849.0,
      "step": 28173
    },
    {
      "entropy": 0.603611383587122,
      "epoch": 2.592153006595707,
      "grad_norm": 0.16687989234924316,
      "learning_rate": 1.3598307111969825e-05,
      "loss": 0.5833,
      "mean_token_accuracy": 0.8169094249606133,
      "num_tokens": 899176885.0,
      "step": 28174
    },
    {
      "entropy": 0.48200383968651295,
      "epoch": 2.592245012967148,
      "grad_norm": 0.15411663055419922,
      "learning_rate": 1.3595240285828198e-05,
      "loss": 0.4706,
      "mean_token_accuracy": 0.848433792591095,
      "num_tokens": 899208810.0,
      "step": 28175
    },
    {
      "entropy": 0.5518632438033819,
      "epoch": 2.592337019338589,
      "grad_norm": 0.16824276745319366,
      "learning_rate": 1.3592173459686572e-05,
      "loss": 0.5493,
      "mean_token_accuracy": 0.8276809677481651,
      "num_tokens": 899240082.0,
      "step": 28176
    },
    {
      "entropy": 0.5179026098921895,
      "epoch": 2.5924290257100306,
      "grad_norm": 0.16075196862220764,
      "learning_rate": 1.3589106633544945e-05,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8395738080143929,
      "num_tokens": 899272135.0,
      "step": 28177
    },
    {
      "entropy": 0.4734937287867069,
      "epoch": 2.5925210320814718,
      "grad_norm": 0.15077966451644897,
      "learning_rate": 1.358603980740332e-05,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8590503819286823,
      "num_tokens": 899304003.0,
      "step": 28178
    },
    {
      "entropy": 0.4526001433841884,
      "epoch": 2.592613038452913,
      "grad_norm": 0.14906980097293854,
      "learning_rate": 1.3582972981261693e-05,
      "loss": 0.4367,
      "mean_token_accuracy": 0.8620250672101974,
      "num_tokens": 899336439.0,
      "step": 28179
    },
    {
      "entropy": 0.4167113252915442,
      "epoch": 2.592705044824354,
      "grad_norm": 0.14895063638687134,
      "learning_rate": 1.3579906155120068e-05,
      "loss": 0.4078,
      "mean_token_accuracy": 0.8695353344082832,
      "num_tokens": 899368207.0,
      "step": 28180
    },
    {
      "entropy": 0.5163977155461907,
      "epoch": 2.592797051195795,
      "grad_norm": 0.16566097736358643,
      "learning_rate": 1.357683932897844e-05,
      "loss": 0.508,
      "mean_token_accuracy": 0.8393443450331688,
      "num_tokens": 899399480.0,
      "step": 28181
    },
    {
      "entropy": 0.5245807319879532,
      "epoch": 2.5928890575672368,
      "grad_norm": 0.15991081297397614,
      "learning_rate": 1.3573772502836815e-05,
      "loss": 0.5124,
      "mean_token_accuracy": 0.837839275598526,
      "num_tokens": 899431676.0,
      "step": 28182
    },
    {
      "entropy": 0.5013318620622158,
      "epoch": 2.592981063938678,
      "grad_norm": 0.1551700383424759,
      "learning_rate": 1.357070567669519e-05,
      "loss": 0.4857,
      "mean_token_accuracy": 0.8481391407549381,
      "num_tokens": 899464082.0,
      "step": 28183
    },
    {
      "entropy": 0.6254508709535003,
      "epoch": 2.593073070310119,
      "grad_norm": 0.17042875289916992,
      "learning_rate": 1.3567638850553563e-05,
      "loss": 0.6057,
      "mean_token_accuracy": 0.8060975335538387,
      "num_tokens": 899496193.0,
      "step": 28184
    },
    {
      "entropy": 0.4252132297260687,
      "epoch": 2.59316507668156,
      "grad_norm": 0.14606761932373047,
      "learning_rate": 1.3564572024411937e-05,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8685783334076405,
      "num_tokens": 899528107.0,
      "step": 28185
    },
    {
      "entropy": 0.5891430203337222,
      "epoch": 2.5932570830530013,
      "grad_norm": 0.16900193691253662,
      "learning_rate": 1.356150519827031e-05,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8215434588491917,
      "num_tokens": 899560118.0,
      "step": 28186
    },
    {
      "entropy": 0.49114694725722075,
      "epoch": 2.593349089424443,
      "grad_norm": 0.15849576890468597,
      "learning_rate": 1.3558438372128685e-05,
      "loss": 0.4835,
      "mean_token_accuracy": 0.8433583527803421,
      "num_tokens": 899592270.0,
      "step": 28187
    },
    {
      "entropy": 0.5042822284158319,
      "epoch": 2.593441095795884,
      "grad_norm": 0.16005440056324005,
      "learning_rate": 1.3555371545987058e-05,
      "loss": 0.4905,
      "mean_token_accuracy": 0.845712311565876,
      "num_tokens": 899623907.0,
      "step": 28188
    },
    {
      "entropy": 0.505175887607038,
      "epoch": 2.593533102167325,
      "grad_norm": 0.1555652618408203,
      "learning_rate": 1.3552304719845433e-05,
      "loss": 0.4945,
      "mean_token_accuracy": 0.8479849733412266,
      "num_tokens": 899656202.0,
      "step": 28189
    },
    {
      "entropy": 0.5058012725785375,
      "epoch": 2.5936251085387663,
      "grad_norm": 0.1571321040391922,
      "learning_rate": 1.3549237893703806e-05,
      "loss": 0.497,
      "mean_token_accuracy": 0.8428276367485523,
      "num_tokens": 899688185.0,
      "step": 28190
    },
    {
      "entropy": 0.522138400003314,
      "epoch": 2.5937171149102074,
      "grad_norm": 0.15878257155418396,
      "learning_rate": 1.354617106756218e-05,
      "loss": 0.4972,
      "mean_token_accuracy": 0.8420529924333096,
      "num_tokens": 899720461.0,
      "step": 28191
    },
    {
      "entropy": 0.5119656955357641,
      "epoch": 2.593809121281649,
      "grad_norm": 0.16038256883621216,
      "learning_rate": 1.3543104241420553e-05,
      "loss": 0.5072,
      "mean_token_accuracy": 0.8403251990675926,
      "num_tokens": 899752158.0,
      "step": 28192
    },
    {
      "entropy": 0.42020841408520937,
      "epoch": 2.59390112765309,
      "grad_norm": 0.14307236671447754,
      "learning_rate": 1.3540037415278928e-05,
      "loss": 0.4089,
      "mean_token_accuracy": 0.8706205002963543,
      "num_tokens": 899784866.0,
      "step": 28193
    },
    {
      "entropy": 0.5028666644357145,
      "epoch": 2.5939931340245312,
      "grad_norm": 0.15848112106323242,
      "learning_rate": 1.3536970589137302e-05,
      "loss": 0.4976,
      "mean_token_accuracy": 0.8438576459884644,
      "num_tokens": 899817470.0,
      "step": 28194
    },
    {
      "entropy": 0.48216399690136313,
      "epoch": 2.5940851403959724,
      "grad_norm": 0.15333041548728943,
      "learning_rate": 1.3533903762995675e-05,
      "loss": 0.4736,
      "mean_token_accuracy": 0.8504428267478943,
      "num_tokens": 899849671.0,
      "step": 28195
    },
    {
      "entropy": 0.5440030032768846,
      "epoch": 2.5941771467674135,
      "grad_norm": 0.1580871045589447,
      "learning_rate": 1.353083693685405e-05,
      "loss": 0.529,
      "mean_token_accuracy": 0.833004217594862,
      "num_tokens": 899882141.0,
      "step": 28196
    },
    {
      "entropy": 0.5240934062749147,
      "epoch": 2.594269153138855,
      "grad_norm": 0.16619625687599182,
      "learning_rate": 1.3527770110712423e-05,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8399271108210087,
      "num_tokens": 899914055.0,
      "step": 28197
    },
    {
      "entropy": 0.5195866646245122,
      "epoch": 2.594361159510296,
      "grad_norm": 0.155541330575943,
      "learning_rate": 1.3524703284570798e-05,
      "loss": 0.51,
      "mean_token_accuracy": 0.8403289057314396,
      "num_tokens": 899946375.0,
      "step": 28198
    },
    {
      "entropy": 0.515073130838573,
      "epoch": 2.5944531658817374,
      "grad_norm": 0.15843354165554047,
      "learning_rate": 1.352163645842917e-05,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8381139226257801,
      "num_tokens": 899977818.0,
      "step": 28199
    },
    {
      "entropy": 0.5451608709990978,
      "epoch": 2.5945451722531785,
      "grad_norm": 0.16475257277488708,
      "learning_rate": 1.3518569632287545e-05,
      "loss": 0.5433,
      "mean_token_accuracy": 0.8310026749968529,
      "num_tokens": 900010586.0,
      "step": 28200
    },
    {
      "entropy": 0.5621136538684368,
      "epoch": 2.5946371786246196,
      "grad_norm": 0.17094337940216064,
      "learning_rate": 1.3515502806145922e-05,
      "loss": 0.5458,
      "mean_token_accuracy": 0.8298199512064457,
      "num_tokens": 900041221.0,
      "step": 28201
    },
    {
      "entropy": 0.5056634447537363,
      "epoch": 2.594729184996061,
      "grad_norm": 0.15655094385147095,
      "learning_rate": 1.3512435980004296e-05,
      "loss": 0.4952,
      "mean_token_accuracy": 0.8425271362066269,
      "num_tokens": 900073363.0,
      "step": 28202
    },
    {
      "entropy": 0.5983701115474105,
      "epoch": 2.5948211913675023,
      "grad_norm": 0.1645904779434204,
      "learning_rate": 1.350936915386267e-05,
      "loss": 0.5744,
      "mean_token_accuracy": 0.8158334046602249,
      "num_tokens": 900104860.0,
      "step": 28203
    },
    {
      "entropy": 0.5792715735733509,
      "epoch": 2.5949131977389435,
      "grad_norm": 0.16293327510356903,
      "learning_rate": 1.3506302327721044e-05,
      "loss": 0.5662,
      "mean_token_accuracy": 0.8190562948584557,
      "num_tokens": 900137245.0,
      "step": 28204
    },
    {
      "entropy": 0.4802171002374962,
      "epoch": 2.5950052041103846,
      "grad_norm": 0.15637987852096558,
      "learning_rate": 1.3503235501579417e-05,
      "loss": 0.4793,
      "mean_token_accuracy": 0.8458282947540283,
      "num_tokens": 900169316.0,
      "step": 28205
    },
    {
      "entropy": 0.4413651907816529,
      "epoch": 2.5950972104818257,
      "grad_norm": 0.1529817283153534,
      "learning_rate": 1.3500168675437791e-05,
      "loss": 0.4302,
      "mean_token_accuracy": 0.865674301981926,
      "num_tokens": 900201017.0,
      "step": 28206
    },
    {
      "entropy": 0.5161902965046465,
      "epoch": 2.5951892168532673,
      "grad_norm": 0.16122885048389435,
      "learning_rate": 1.3497101849296164e-05,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8373831138014793,
      "num_tokens": 900233229.0,
      "step": 28207
    },
    {
      "entropy": 0.47774090711027384,
      "epoch": 2.5952812232247084,
      "grad_norm": 0.1560545265674591,
      "learning_rate": 1.3494035023154539e-05,
      "loss": 0.4624,
      "mean_token_accuracy": 0.8524090014398098,
      "num_tokens": 900265661.0,
      "step": 28208
    },
    {
      "entropy": 0.5589319528080523,
      "epoch": 2.5953732295961496,
      "grad_norm": 0.16351665556430817,
      "learning_rate": 1.3490968197012912e-05,
      "loss": 0.5429,
      "mean_token_accuracy": 0.8294409215450287,
      "num_tokens": 900296646.0,
      "step": 28209
    },
    {
      "entropy": 0.5001609358005226,
      "epoch": 2.5954652359675907,
      "grad_norm": 0.15497621893882751,
      "learning_rate": 1.3487901370871287e-05,
      "loss": 0.4856,
      "mean_token_accuracy": 0.8459895364940166,
      "num_tokens": 900328730.0,
      "step": 28210
    },
    {
      "entropy": 0.5521027408540249,
      "epoch": 2.595557242339032,
      "grad_norm": 0.16432277858257294,
      "learning_rate": 1.348483454472966e-05,
      "loss": 0.5485,
      "mean_token_accuracy": 0.8286941386759281,
      "num_tokens": 900360696.0,
      "step": 28211
    },
    {
      "entropy": 0.5150715298950672,
      "epoch": 2.5956492487104734,
      "grad_norm": 0.15542414784431458,
      "learning_rate": 1.3481767718588034e-05,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8401632159948349,
      "num_tokens": 900392322.0,
      "step": 28212
    },
    {
      "entropy": 0.570399955380708,
      "epoch": 2.5957412550819146,
      "grad_norm": 0.16782821714878082,
      "learning_rate": 1.3478700892446409e-05,
      "loss": 0.5627,
      "mean_token_accuracy": 0.8225649297237396,
      "num_tokens": 900424377.0,
      "step": 28213
    },
    {
      "entropy": 0.518876095302403,
      "epoch": 2.5958332614533557,
      "grad_norm": 0.1627325713634491,
      "learning_rate": 1.3475634066304782e-05,
      "loss": 0.5118,
      "mean_token_accuracy": 0.8393551781773567,
      "num_tokens": 900456358.0,
      "step": 28214
    },
    {
      "entropy": 0.50574248470366,
      "epoch": 2.595925267824797,
      "grad_norm": 0.1632971614599228,
      "learning_rate": 1.3472567240163156e-05,
      "loss": 0.5015,
      "mean_token_accuracy": 0.838752020150423,
      "num_tokens": 900487713.0,
      "step": 28215
    },
    {
      "entropy": 0.569315355271101,
      "epoch": 2.596017274196238,
      "grad_norm": 0.1690404862165451,
      "learning_rate": 1.346950041402153e-05,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8167094402015209,
      "num_tokens": 900519447.0,
      "step": 28216
    },
    {
      "entropy": 0.4674865137785673,
      "epoch": 2.5961092805676795,
      "grad_norm": 0.15104569494724274,
      "learning_rate": 1.3466433587879904e-05,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8522581718862057,
      "num_tokens": 900551711.0,
      "step": 28217
    },
    {
      "entropy": 0.5380520345643163,
      "epoch": 2.5962012869391207,
      "grad_norm": 0.16196700930595398,
      "learning_rate": 1.3463366761738277e-05,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8356987610459328,
      "num_tokens": 900583957.0,
      "step": 28218
    },
    {
      "entropy": 0.5046512531116605,
      "epoch": 2.596293293310562,
      "grad_norm": 0.15215645730495453,
      "learning_rate": 1.3460299935596652e-05,
      "loss": 0.4905,
      "mean_token_accuracy": 0.8443394005298615,
      "num_tokens": 900616217.0,
      "step": 28219
    },
    {
      "entropy": 0.517803056165576,
      "epoch": 2.596385299682003,
      "grad_norm": 0.1577947735786438,
      "learning_rate": 1.3457233109455025e-05,
      "loss": 0.5103,
      "mean_token_accuracy": 0.8405432924628258,
      "num_tokens": 900648622.0,
      "step": 28220
    },
    {
      "entropy": 0.5476500056684017,
      "epoch": 2.596477306053444,
      "grad_norm": 0.16586948931217194,
      "learning_rate": 1.34541662833134e-05,
      "loss": 0.5307,
      "mean_token_accuracy": 0.8284727931022644,
      "num_tokens": 900680799.0,
      "step": 28221
    },
    {
      "entropy": 0.5147372949868441,
      "epoch": 2.5965693124248856,
      "grad_norm": 0.15780910849571228,
      "learning_rate": 1.3451099457171772e-05,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8404557108879089,
      "num_tokens": 900712629.0,
      "step": 28222
    },
    {
      "entropy": 0.5480255950242281,
      "epoch": 2.596661318796327,
      "grad_norm": 0.17110292613506317,
      "learning_rate": 1.3448032631030147e-05,
      "loss": 0.5427,
      "mean_token_accuracy": 0.8290177583694458,
      "num_tokens": 900744838.0,
      "step": 28223
    },
    {
      "entropy": 0.569641973823309,
      "epoch": 2.596753325167768,
      "grad_norm": 0.16144204139709473,
      "learning_rate": 1.3444965804888522e-05,
      "loss": 0.5582,
      "mean_token_accuracy": 0.8258572667837143,
      "num_tokens": 900777311.0,
      "step": 28224
    },
    {
      "entropy": 0.5949212089180946,
      "epoch": 2.596845331539209,
      "grad_norm": 0.17667809128761292,
      "learning_rate": 1.3441898978746894e-05,
      "loss": 0.5896,
      "mean_token_accuracy": 0.8115385621786118,
      "num_tokens": 900807622.0,
      "step": 28225
    },
    {
      "entropy": 0.5355099383741617,
      "epoch": 2.59693733791065,
      "grad_norm": 0.159736767411232,
      "learning_rate": 1.3438832152605269e-05,
      "loss": 0.5246,
      "mean_token_accuracy": 0.832745224237442,
      "num_tokens": 900840022.0,
      "step": 28226
    },
    {
      "entropy": 0.6095592640340328,
      "epoch": 2.5970293442820918,
      "grad_norm": 0.17110280692577362,
      "learning_rate": 1.3435765326463642e-05,
      "loss": 0.5935,
      "mean_token_accuracy": 0.8136614859104156,
      "num_tokens": 900871882.0,
      "step": 28227
    },
    {
      "entropy": 0.44097559805959463,
      "epoch": 2.597121350653533,
      "grad_norm": 0.15193869173526764,
      "learning_rate": 1.3432698500322017e-05,
      "loss": 0.4356,
      "mean_token_accuracy": 0.8616142980754375,
      "num_tokens": 900904051.0,
      "step": 28228
    },
    {
      "entropy": 0.5749397296458483,
      "epoch": 2.597213357024974,
      "grad_norm": 0.16740682721138,
      "learning_rate": 1.342963167418039e-05,
      "loss": 0.5597,
      "mean_token_accuracy": 0.8279673606157303,
      "num_tokens": 900935745.0,
      "step": 28229
    },
    {
      "entropy": 0.4730112454853952,
      "epoch": 2.597305363396415,
      "grad_norm": 0.15478718280792236,
      "learning_rate": 1.3426564848038764e-05,
      "loss": 0.458,
      "mean_token_accuracy": 0.8573456928133965,
      "num_tokens": 900967359.0,
      "step": 28230
    },
    {
      "entropy": 0.462701587472111,
      "epoch": 2.5973973697678563,
      "grad_norm": 0.15016965568065643,
      "learning_rate": 1.342349802189714e-05,
      "loss": 0.4411,
      "mean_token_accuracy": 0.8597429394721985,
      "num_tokens": 900998928.0,
      "step": 28231
    },
    {
      "entropy": 0.49225260503590107,
      "epoch": 2.597489376139298,
      "grad_norm": 0.1558152735233307,
      "learning_rate": 1.3420431195755515e-05,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8452869318425655,
      "num_tokens": 901031151.0,
      "step": 28232
    },
    {
      "entropy": 0.653489300981164,
      "epoch": 2.597581382510739,
      "grad_norm": 0.17147962749004364,
      "learning_rate": 1.3417364369613888e-05,
      "loss": 0.6431,
      "mean_token_accuracy": 0.8001320250332355,
      "num_tokens": 901063761.0,
      "step": 28233
    },
    {
      "entropy": 0.4428385307546705,
      "epoch": 2.59767338888218,
      "grad_norm": 0.15672539174556732,
      "learning_rate": 1.3414297543472263e-05,
      "loss": 0.4479,
      "mean_token_accuracy": 0.85982895642519,
      "num_tokens": 901095446.0,
      "step": 28234
    },
    {
      "entropy": 0.5638486817479134,
      "epoch": 2.5977653952536213,
      "grad_norm": 0.17140094935894012,
      "learning_rate": 1.3411230717330636e-05,
      "loss": 0.5554,
      "mean_token_accuracy": 0.8238424845039845,
      "num_tokens": 901127111.0,
      "step": 28235
    },
    {
      "entropy": 0.5179489627480507,
      "epoch": 2.5978574016250624,
      "grad_norm": 0.16132666170597076,
      "learning_rate": 1.340816389118901e-05,
      "loss": 0.5075,
      "mean_token_accuracy": 0.8355398327112198,
      "num_tokens": 901159023.0,
      "step": 28236
    },
    {
      "entropy": 0.5508319400250912,
      "epoch": 2.597949407996504,
      "grad_norm": 0.16155362129211426,
      "learning_rate": 1.3405097065047383e-05,
      "loss": 0.5386,
      "mean_token_accuracy": 0.8279087357223034,
      "num_tokens": 901190959.0,
      "step": 28237
    },
    {
      "entropy": 0.41799213271588087,
      "epoch": 2.598041414367945,
      "grad_norm": 0.14920322597026825,
      "learning_rate": 1.3402030238905758e-05,
      "loss": 0.4072,
      "mean_token_accuracy": 0.8669858239591122,
      "num_tokens": 901223207.0,
      "step": 28238
    },
    {
      "entropy": 0.5421827007085085,
      "epoch": 2.5981334207393862,
      "grad_norm": 0.1634903997182846,
      "learning_rate": 1.3398963412764131e-05,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8298310413956642,
      "num_tokens": 901255259.0,
      "step": 28239
    },
    {
      "entropy": 0.46925269439816475,
      "epoch": 2.5982254271108274,
      "grad_norm": 0.15718407928943634,
      "learning_rate": 1.3395896586622506e-05,
      "loss": 0.4628,
      "mean_token_accuracy": 0.8524570763111115,
      "num_tokens": 901287135.0,
      "step": 28240
    },
    {
      "entropy": 0.4399890740169212,
      "epoch": 2.5983174334822685,
      "grad_norm": 0.1490848809480667,
      "learning_rate": 1.3392829760480879e-05,
      "loss": 0.4286,
      "mean_token_accuracy": 0.8667112775146961,
      "num_tokens": 901318846.0,
      "step": 28241
    },
    {
      "entropy": 0.6308501586318016,
      "epoch": 2.59840943985371,
      "grad_norm": 0.16982804238796234,
      "learning_rate": 1.3389762934339253e-05,
      "loss": 0.612,
      "mean_token_accuracy": 0.8049821332097054,
      "num_tokens": 901350325.0,
      "step": 28242
    },
    {
      "entropy": 0.5401952210813761,
      "epoch": 2.5985014462251512,
      "grad_norm": 0.16271084547042847,
      "learning_rate": 1.3386696108197628e-05,
      "loss": 0.529,
      "mean_token_accuracy": 0.8321045711636543,
      "num_tokens": 901382003.0,
      "step": 28243
    },
    {
      "entropy": 0.4635147023946047,
      "epoch": 2.5985934525965924,
      "grad_norm": 0.15142486989498138,
      "learning_rate": 1.3383629282056001e-05,
      "loss": 0.459,
      "mean_token_accuracy": 0.8541868403553963,
      "num_tokens": 901414589.0,
      "step": 28244
    },
    {
      "entropy": 0.5840107770636678,
      "epoch": 2.5986854589680335,
      "grad_norm": 0.16944913566112518,
      "learning_rate": 1.3380562455914376e-05,
      "loss": 0.5839,
      "mean_token_accuracy": 0.8147887475788593,
      "num_tokens": 901445891.0,
      "step": 28245
    },
    {
      "entropy": 0.5430951435118914,
      "epoch": 2.5987774653394746,
      "grad_norm": 0.1593707799911499,
      "learning_rate": 1.3377495629772749e-05,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8308185040950775,
      "num_tokens": 901477437.0,
      "step": 28246
    },
    {
      "entropy": 0.48058329988271,
      "epoch": 2.598869471710916,
      "grad_norm": 0.15347611904144287,
      "learning_rate": 1.3374428803631123e-05,
      "loss": 0.4658,
      "mean_token_accuracy": 0.8546812906861305,
      "num_tokens": 901509194.0,
      "step": 28247
    },
    {
      "entropy": 0.5709867104887962,
      "epoch": 2.5989614780823573,
      "grad_norm": 0.1666181981563568,
      "learning_rate": 1.3371361977489496e-05,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8210254907608032,
      "num_tokens": 901540633.0,
      "step": 28248
    },
    {
      "entropy": 0.5300663784146309,
      "epoch": 2.5990534844537985,
      "grad_norm": 0.16059331595897675,
      "learning_rate": 1.336829515134787e-05,
      "loss": 0.5164,
      "mean_token_accuracy": 0.8381140977144241,
      "num_tokens": 901572884.0,
      "step": 28249
    },
    {
      "entropy": 0.3744340199045837,
      "epoch": 2.5991454908252396,
      "grad_norm": 0.13875164091587067,
      "learning_rate": 1.3365228325206244e-05,
      "loss": 0.3717,
      "mean_token_accuracy": 0.8823367655277252,
      "num_tokens": 901605430.0,
      "step": 28250
    },
    {
      "entropy": 0.4794623088091612,
      "epoch": 2.5992374971966807,
      "grad_norm": 0.15675584971904755,
      "learning_rate": 1.3362161499064618e-05,
      "loss": 0.4753,
      "mean_token_accuracy": 0.8462494909763336,
      "num_tokens": 901637361.0,
      "step": 28251
    },
    {
      "entropy": 0.42129123234190047,
      "epoch": 2.5993295035681223,
      "grad_norm": 0.1510845571756363,
      "learning_rate": 1.3359094672922991e-05,
      "loss": 0.4125,
      "mean_token_accuracy": 0.8678025230765343,
      "num_tokens": 901669299.0,
      "step": 28252
    },
    {
      "entropy": 0.5704724481329322,
      "epoch": 2.5994215099395634,
      "grad_norm": 0.17035643756389618,
      "learning_rate": 1.3356027846781366e-05,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8246455751359463,
      "num_tokens": 901701471.0,
      "step": 28253
    },
    {
      "entropy": 0.521621155552566,
      "epoch": 2.5995135163110046,
      "grad_norm": 0.1600295901298523,
      "learning_rate": 1.335296102063974e-05,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8403059579432011,
      "num_tokens": 901732891.0,
      "step": 28254
    },
    {
      "entropy": 0.40151338931173086,
      "epoch": 2.5996055226824457,
      "grad_norm": 0.14326894283294678,
      "learning_rate": 1.3349894194498114e-05,
      "loss": 0.3772,
      "mean_token_accuracy": 0.8765738680958748,
      "num_tokens": 901763902.0,
      "step": 28255
    },
    {
      "entropy": 0.5494592888280749,
      "epoch": 2.599697529053887,
      "grad_norm": 0.1664787083864212,
      "learning_rate": 1.3346827368356488e-05,
      "loss": 0.5365,
      "mean_token_accuracy": 0.8310879953205585,
      "num_tokens": 901796046.0,
      "step": 28256
    },
    {
      "entropy": 0.518883396871388,
      "epoch": 2.5997895354253284,
      "grad_norm": 0.1676834225654602,
      "learning_rate": 1.3343760542214861e-05,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8381627462804317,
      "num_tokens": 901828012.0,
      "step": 28257
    },
    {
      "entropy": 0.6034623710438609,
      "epoch": 2.5998815417967696,
      "grad_norm": 0.17131009697914124,
      "learning_rate": 1.3340693716073236e-05,
      "loss": 0.5981,
      "mean_token_accuracy": 0.8165554963052273,
      "num_tokens": 901860467.0,
      "step": 28258
    },
    {
      "entropy": 0.47298349207267165,
      "epoch": 2.5999735481682107,
      "grad_norm": 0.1486651599407196,
      "learning_rate": 1.3337626889931609e-05,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8531575463712215,
      "num_tokens": 901892457.0,
      "step": 28259
    },
    {
      "entropy": 0.5221669524908066,
      "epoch": 2.600065554539652,
      "grad_norm": 0.17288808524608612,
      "learning_rate": 1.3334560063789983e-05,
      "loss": 0.5131,
      "mean_token_accuracy": 0.8364674188196659,
      "num_tokens": 901922603.0,
      "step": 28260
    },
    {
      "entropy": 0.4750172607600689,
      "epoch": 2.600157560911093,
      "grad_norm": 0.14863191545009613,
      "learning_rate": 1.3331493237648356e-05,
      "loss": 0.452,
      "mean_token_accuracy": 0.8551495037972927,
      "num_tokens": 901954748.0,
      "step": 28261
    },
    {
      "entropy": 0.5160471303388476,
      "epoch": 2.6002495672825345,
      "grad_norm": 0.15873827040195465,
      "learning_rate": 1.3328426411506734e-05,
      "loss": 0.4941,
      "mean_token_accuracy": 0.8445414341986179,
      "num_tokens": 901986755.0,
      "step": 28262
    },
    {
      "entropy": 0.43847398925572634,
      "epoch": 2.6003415736539757,
      "grad_norm": 0.15617112815380096,
      "learning_rate": 1.3325359585365107e-05,
      "loss": 0.4278,
      "mean_token_accuracy": 0.8637622408568859,
      "num_tokens": 902018665.0,
      "step": 28263
    },
    {
      "entropy": 0.5217633470892906,
      "epoch": 2.600433580025417,
      "grad_norm": 0.1611657589673996,
      "learning_rate": 1.3322292759223482e-05,
      "loss": 0.5146,
      "mean_token_accuracy": 0.8353440426290035,
      "num_tokens": 902051047.0,
      "step": 28264
    },
    {
      "entropy": 0.4964370774105191,
      "epoch": 2.600525586396858,
      "grad_norm": 0.8510679602622986,
      "learning_rate": 1.3319225933081855e-05,
      "loss": 0.5077,
      "mean_token_accuracy": 0.8462406732141972,
      "num_tokens": 902083525.0,
      "step": 28265
    },
    {
      "entropy": 0.506592015735805,
      "epoch": 2.600617592768299,
      "grad_norm": 0.15731316804885864,
      "learning_rate": 1.331615910694023e-05,
      "loss": 0.4938,
      "mean_token_accuracy": 0.8424609191715717,
      "num_tokens": 902115607.0,
      "step": 28266
    },
    {
      "entropy": 0.5884567648172379,
      "epoch": 2.6007095991397406,
      "grad_norm": 0.16928353905677795,
      "learning_rate": 1.3313092280798603e-05,
      "loss": 0.5833,
      "mean_token_accuracy": 0.8204018510878086,
      "num_tokens": 902147375.0,
      "step": 28267
    },
    {
      "entropy": 0.4821149678900838,
      "epoch": 2.600801605511182,
      "grad_norm": 0.1512814313173294,
      "learning_rate": 1.3310025454656977e-05,
      "loss": 0.4717,
      "mean_token_accuracy": 0.852284375578165,
      "num_tokens": 902179696.0,
      "step": 28268
    },
    {
      "entropy": 0.4868776313960552,
      "epoch": 2.600893611882623,
      "grad_norm": 0.15651291608810425,
      "learning_rate": 1.330695862851535e-05,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8468835428357124,
      "num_tokens": 902212118.0,
      "step": 28269
    },
    {
      "entropy": 0.5535234156996012,
      "epoch": 2.600985618254064,
      "grad_norm": 0.16153469681739807,
      "learning_rate": 1.3303891802373725e-05,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8269290402531624,
      "num_tokens": 902244751.0,
      "step": 28270
    },
    {
      "entropy": 0.5130287623032928,
      "epoch": 2.601077624625505,
      "grad_norm": 0.15925456583499908,
      "learning_rate": 1.3300824976232098e-05,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8397430628538132,
      "num_tokens": 902277388.0,
      "step": 28271
    },
    {
      "entropy": 0.4269328167429194,
      "epoch": 2.6011696309969468,
      "grad_norm": 0.15334734320640564,
      "learning_rate": 1.3297758150090472e-05,
      "loss": 0.4069,
      "mean_token_accuracy": 0.868051465600729,
      "num_tokens": 902308379.0,
      "step": 28272
    },
    {
      "entropy": 0.5427501546218991,
      "epoch": 2.601261637368388,
      "grad_norm": 0.16449619829654694,
      "learning_rate": 1.3294691323948847e-05,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8297178000211716,
      "num_tokens": 902339946.0,
      "step": 28273
    },
    {
      "entropy": 0.5185576044023037,
      "epoch": 2.601353643739829,
      "grad_norm": 0.16293582320213318,
      "learning_rate": 1.329162449780722e-05,
      "loss": 0.5041,
      "mean_token_accuracy": 0.8467773608863354,
      "num_tokens": 902371826.0,
      "step": 28274
    },
    {
      "entropy": 0.5317453574389219,
      "epoch": 2.60144565011127,
      "grad_norm": 0.15658046305179596,
      "learning_rate": 1.3288557671665595e-05,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8361733146011829,
      "num_tokens": 902404530.0,
      "step": 28275
    },
    {
      "entropy": 0.5071073612198234,
      "epoch": 2.6015376564827113,
      "grad_norm": 0.1635550707578659,
      "learning_rate": 1.3285490845523968e-05,
      "loss": 0.4952,
      "mean_token_accuracy": 0.8464798554778099,
      "num_tokens": 902436025.0,
      "step": 28276
    },
    {
      "entropy": 0.5960611011832952,
      "epoch": 2.601629662854153,
      "grad_norm": 0.17304284870624542,
      "learning_rate": 1.3282424019382342e-05,
      "loss": 0.5797,
      "mean_token_accuracy": 0.8178924284875393,
      "num_tokens": 902467476.0,
      "step": 28277
    },
    {
      "entropy": 0.39300623442977667,
      "epoch": 2.601721669225594,
      "grad_norm": 0.14043250679969788,
      "learning_rate": 1.3279357193240715e-05,
      "loss": 0.3874,
      "mean_token_accuracy": 0.876569103449583,
      "num_tokens": 902499836.0,
      "step": 28278
    },
    {
      "entropy": 0.5378907155245543,
      "epoch": 2.601813675597035,
      "grad_norm": 0.1680966168642044,
      "learning_rate": 1.327629036709909e-05,
      "loss": 0.5323,
      "mean_token_accuracy": 0.8316112644970417,
      "num_tokens": 902531703.0,
      "step": 28279
    },
    {
      "entropy": 0.5530280373059213,
      "epoch": 2.6019056819684763,
      "grad_norm": 0.1654997020959854,
      "learning_rate": 1.3273223540957463e-05,
      "loss": 0.5437,
      "mean_token_accuracy": 0.8259734623134136,
      "num_tokens": 902563382.0,
      "step": 28280
    },
    {
      "entropy": 0.48441638331860304,
      "epoch": 2.6019976883399174,
      "grad_norm": 0.15357182919979095,
      "learning_rate": 1.3270156714815837e-05,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8510797843337059,
      "num_tokens": 902595853.0,
      "step": 28281
    },
    {
      "entropy": 0.42462270334362984,
      "epoch": 2.602089694711359,
      "grad_norm": 0.1469718962907791,
      "learning_rate": 1.326708988867421e-05,
      "loss": 0.4066,
      "mean_token_accuracy": 0.8678762689232826,
      "num_tokens": 902627692.0,
      "step": 28282
    },
    {
      "entropy": 0.475219386164099,
      "epoch": 2.6021817010828,
      "grad_norm": 0.15417665243148804,
      "learning_rate": 1.3264023062532585e-05,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8569832034409046,
      "num_tokens": 902659347.0,
      "step": 28283
    },
    {
      "entropy": 0.4839251828379929,
      "epoch": 2.6022737074542412,
      "grad_norm": 0.15723395347595215,
      "learning_rate": 1.326095623639096e-05,
      "loss": 0.4584,
      "mean_token_accuracy": 0.8518504165112972,
      "num_tokens": 902690567.0,
      "step": 28284
    },
    {
      "entropy": 0.5176213551312685,
      "epoch": 2.6023657138256824,
      "grad_norm": 0.16094659268856049,
      "learning_rate": 1.3257889410249333e-05,
      "loss": 0.5072,
      "mean_token_accuracy": 0.8386893682181835,
      "num_tokens": 902722855.0,
      "step": 28285
    },
    {
      "entropy": 0.4889058042317629,
      "epoch": 2.6024577201971235,
      "grad_norm": 0.15278498828411102,
      "learning_rate": 1.3254822584107707e-05,
      "loss": 0.4831,
      "mean_token_accuracy": 0.8503392785787582,
      "num_tokens": 902754797.0,
      "step": 28286
    },
    {
      "entropy": 0.4511449821293354,
      "epoch": 2.602549726568565,
      "grad_norm": 0.15287111699581146,
      "learning_rate": 1.325175575796608e-05,
      "loss": 0.446,
      "mean_token_accuracy": 0.8533364608883858,
      "num_tokens": 902787363.0,
      "step": 28287
    },
    {
      "entropy": 0.5256030266173184,
      "epoch": 2.6026417329400062,
      "grad_norm": 0.15790365636348724,
      "learning_rate": 1.3248688931824455e-05,
      "loss": 0.5148,
      "mean_token_accuracy": 0.8339271657168865,
      "num_tokens": 902819222.0,
      "step": 28288
    },
    {
      "entropy": 0.4206531059462577,
      "epoch": 2.6027337393114474,
      "grad_norm": 0.14611801505088806,
      "learning_rate": 1.3245622105682828e-05,
      "loss": 0.405,
      "mean_token_accuracy": 0.8706445321440697,
      "num_tokens": 902850908.0,
      "step": 28289
    },
    {
      "entropy": 0.499949905090034,
      "epoch": 2.6028257456828885,
      "grad_norm": 0.15330809354782104,
      "learning_rate": 1.3242555279541203e-05,
      "loss": 0.4824,
      "mean_token_accuracy": 0.8457708545029163,
      "num_tokens": 902883070.0,
      "step": 28290
    },
    {
      "entropy": 0.5768110912758857,
      "epoch": 2.6029177520543296,
      "grad_norm": 0.16935767233371735,
      "learning_rate": 1.3239488453399575e-05,
      "loss": 0.5583,
      "mean_token_accuracy": 0.8240936808288097,
      "num_tokens": 902914772.0,
      "step": 28291
    },
    {
      "entropy": 0.45919525530189276,
      "epoch": 2.603009758425771,
      "grad_norm": 0.15374140441417694,
      "learning_rate": 1.3236421627257954e-05,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8558897078037262,
      "num_tokens": 902945807.0,
      "step": 28292
    },
    {
      "entropy": 0.5114441206678748,
      "epoch": 2.6031017647972123,
      "grad_norm": 0.15868070721626282,
      "learning_rate": 1.3233354801116326e-05,
      "loss": 0.4935,
      "mean_token_accuracy": 0.8408354036509991,
      "num_tokens": 902977252.0,
      "step": 28293
    },
    {
      "entropy": 0.6229985933750868,
      "epoch": 2.6031937711686535,
      "grad_norm": 0.16956183314323425,
      "learning_rate": 1.3230287974974701e-05,
      "loss": 0.6121,
      "mean_token_accuracy": 0.8077124655246735,
      "num_tokens": 903008980.0,
      "step": 28294
    },
    {
      "entropy": 0.5544647350907326,
      "epoch": 2.6032857775400946,
      "grad_norm": 0.16108737885951996,
      "learning_rate": 1.3227221148833074e-05,
      "loss": 0.5436,
      "mean_token_accuracy": 0.8310847766697407,
      "num_tokens": 903040670.0,
      "step": 28295
    },
    {
      "entropy": 0.5403539557009935,
      "epoch": 2.6033777839115357,
      "grad_norm": 0.16175585985183716,
      "learning_rate": 1.3224154322691449e-05,
      "loss": 0.5349,
      "mean_token_accuracy": 0.8315732479095459,
      "num_tokens": 903073180.0,
      "step": 28296
    },
    {
      "entropy": 0.5035894960165024,
      "epoch": 2.6034697902829773,
      "grad_norm": 0.15753431618213654,
      "learning_rate": 1.3221087496549822e-05,
      "loss": 0.4945,
      "mean_token_accuracy": 0.8444164879620075,
      "num_tokens": 903104907.0,
      "step": 28297
    },
    {
      "entropy": 0.6202764566987753,
      "epoch": 2.6035617966544184,
      "grad_norm": 0.1707228720188141,
      "learning_rate": 1.3218020670408196e-05,
      "loss": 0.6097,
      "mean_token_accuracy": 0.8066837936639786,
      "num_tokens": 903136705.0,
      "step": 28298
    },
    {
      "entropy": 0.5299674794077873,
      "epoch": 2.6036538030258596,
      "grad_norm": 0.16136720776557922,
      "learning_rate": 1.321495384426657e-05,
      "loss": 0.521,
      "mean_token_accuracy": 0.8349180929362774,
      "num_tokens": 903168307.0,
      "step": 28299
    },
    {
      "entropy": 0.5113131222315133,
      "epoch": 2.6037458093973007,
      "grad_norm": 0.15340976417064667,
      "learning_rate": 1.3211887018124944e-05,
      "loss": 0.493,
      "mean_token_accuracy": 0.8426394872367382,
      "num_tokens": 903199757.0,
      "step": 28300
    },
    {
      "entropy": 0.5137890707701445,
      "epoch": 2.603837815768742,
      "grad_norm": 0.16551907360553741,
      "learning_rate": 1.3208820191983317e-05,
      "loss": 0.5061,
      "mean_token_accuracy": 0.8418612331151962,
      "num_tokens": 903231894.0,
      "step": 28301
    },
    {
      "entropy": 0.4812477696686983,
      "epoch": 2.6039298221401834,
      "grad_norm": 0.15200616419315338,
      "learning_rate": 1.3205753365841692e-05,
      "loss": 0.463,
      "mean_token_accuracy": 0.8467263653874397,
      "num_tokens": 903262672.0,
      "step": 28302
    },
    {
      "entropy": 0.5290214559063315,
      "epoch": 2.6040218285116246,
      "grad_norm": 0.16483215987682343,
      "learning_rate": 1.3202686539700066e-05,
      "loss": 0.5286,
      "mean_token_accuracy": 0.834223035722971,
      "num_tokens": 903294041.0,
      "step": 28303
    },
    {
      "entropy": 0.5562193039804697,
      "epoch": 2.6041138348830657,
      "grad_norm": 0.16365139186382294,
      "learning_rate": 1.3199619713558439e-05,
      "loss": 0.5525,
      "mean_token_accuracy": 0.8252218440175056,
      "num_tokens": 903326274.0,
      "step": 28304
    },
    {
      "entropy": 0.5862515051849186,
      "epoch": 2.604205841254507,
      "grad_norm": 0.1682736873626709,
      "learning_rate": 1.3196552887416814e-05,
      "loss": 0.5651,
      "mean_token_accuracy": 0.8181197419762611,
      "num_tokens": 903358338.0,
      "step": 28305
    },
    {
      "entropy": 0.6012823134660721,
      "epoch": 2.604297847625948,
      "grad_norm": 0.16544298827648163,
      "learning_rate": 1.3193486061275187e-05,
      "loss": 0.5746,
      "mean_token_accuracy": 0.8188791908323765,
      "num_tokens": 903390392.0,
      "step": 28306
    },
    {
      "entropy": 0.5014185199979693,
      "epoch": 2.6043898539973895,
      "grad_norm": 0.15588803589344025,
      "learning_rate": 1.3190419235133561e-05,
      "loss": 0.4821,
      "mean_token_accuracy": 0.8448521085083485,
      "num_tokens": 903422882.0,
      "step": 28307
    },
    {
      "entropy": 0.5936824064701796,
      "epoch": 2.6044818603688307,
      "grad_norm": 0.16672374308109283,
      "learning_rate": 1.3187352408991934e-05,
      "loss": 0.5779,
      "mean_token_accuracy": 0.8195785991847515,
      "num_tokens": 903455037.0,
      "step": 28308
    },
    {
      "entropy": 0.5796952340751886,
      "epoch": 2.604573866740272,
      "grad_norm": 0.15720731019973755,
      "learning_rate": 1.3184285582850309e-05,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8217977024614811,
      "num_tokens": 903487799.0,
      "step": 28309
    },
    {
      "entropy": 0.5514019792899489,
      "epoch": 2.604665873111713,
      "grad_norm": 0.1628049910068512,
      "learning_rate": 1.3181218756708682e-05,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8305836543440819,
      "num_tokens": 903519659.0,
      "step": 28310
    },
    {
      "entropy": 0.5301405536010861,
      "epoch": 2.604757879483154,
      "grad_norm": 0.1608152836561203,
      "learning_rate": 1.3178151930567057e-05,
      "loss": 0.5197,
      "mean_token_accuracy": 0.8354477621614933,
      "num_tokens": 903551475.0,
      "step": 28311
    },
    {
      "entropy": 0.5785166118294001,
      "epoch": 2.6048498858545956,
      "grad_norm": 0.16207890212535858,
      "learning_rate": 1.317508510442543e-05,
      "loss": 0.553,
      "mean_token_accuracy": 0.8272100351750851,
      "num_tokens": 903584044.0,
      "step": 28312
    },
    {
      "entropy": 0.46515848487615585,
      "epoch": 2.604941892226037,
      "grad_norm": 0.15557970106601715,
      "learning_rate": 1.3172018278283804e-05,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8553241267800331,
      "num_tokens": 903616261.0,
      "step": 28313
    },
    {
      "entropy": 0.5638889092952013,
      "epoch": 2.605033898597478,
      "grad_norm": 0.1632772833108902,
      "learning_rate": 1.3168951452142179e-05,
      "loss": 0.5627,
      "mean_token_accuracy": 0.8210031650960445,
      "num_tokens": 903648457.0,
      "step": 28314
    },
    {
      "entropy": 0.46164507418870926,
      "epoch": 2.605125904968919,
      "grad_norm": 0.1451621800661087,
      "learning_rate": 1.3165884626000552e-05,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8578712679445744,
      "num_tokens": 903681225.0,
      "step": 28315
    },
    {
      "entropy": 0.48515452118590474,
      "epoch": 2.60521791134036,
      "grad_norm": 0.15751944482326508,
      "learning_rate": 1.3162817799858926e-05,
      "loss": 0.4705,
      "mean_token_accuracy": 0.8491312861442566,
      "num_tokens": 903713395.0,
      "step": 28316
    },
    {
      "entropy": 0.43777802772819996,
      "epoch": 2.6053099177118018,
      "grad_norm": 0.14396627247333527,
      "learning_rate": 1.31597509737173e-05,
      "loss": 0.4231,
      "mean_token_accuracy": 0.8637995086610317,
      "num_tokens": 903745892.0,
      "step": 28317
    },
    {
      "entropy": 0.5319998059421778,
      "epoch": 2.605401924083243,
      "grad_norm": 0.16491840779781342,
      "learning_rate": 1.3156684147575674e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8348898142576218,
      "num_tokens": 903777256.0,
      "step": 28318
    },
    {
      "entropy": 0.598953009583056,
      "epoch": 2.605493930454684,
      "grad_norm": 0.16713352501392365,
      "learning_rate": 1.3153617321434047e-05,
      "loss": 0.586,
      "mean_token_accuracy": 0.8186911009252071,
      "num_tokens": 903809216.0,
      "step": 28319
    },
    {
      "entropy": 0.6284565310925245,
      "epoch": 2.605585936826125,
      "grad_norm": 0.18103814125061035,
      "learning_rate": 1.3150550495292422e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8031970076262951,
      "num_tokens": 903839945.0,
      "step": 28320
    },
    {
      "entropy": 0.6037060415837914,
      "epoch": 2.6056779431975663,
      "grad_norm": 0.16884230077266693,
      "learning_rate": 1.3147483669150795e-05,
      "loss": 0.593,
      "mean_token_accuracy": 0.8162972629070282,
      "num_tokens": 903872521.0,
      "step": 28321
    },
    {
      "entropy": 0.4517563576810062,
      "epoch": 2.605769949569008,
      "grad_norm": 0.1671735793352127,
      "learning_rate": 1.314441684300917e-05,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8594395071268082,
      "num_tokens": 903903848.0,
      "step": 28322
    },
    {
      "entropy": 0.5462042037397623,
      "epoch": 2.605861955940449,
      "grad_norm": 0.17185059189796448,
      "learning_rate": 1.3141350016867546e-05,
      "loss": 0.5427,
      "mean_token_accuracy": 0.8292316682636738,
      "num_tokens": 903935383.0,
      "step": 28323
    },
    {
      "entropy": 0.5346918134018779,
      "epoch": 2.60595396231189,
      "grad_norm": 0.1606139987707138,
      "learning_rate": 1.313828319072592e-05,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8306994773447514,
      "num_tokens": 903967129.0,
      "step": 28324
    },
    {
      "entropy": 0.43864279706031084,
      "epoch": 2.6060459686833313,
      "grad_norm": 0.15191034972667694,
      "learning_rate": 1.3135216364584293e-05,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8595203906297684,
      "num_tokens": 903999532.0,
      "step": 28325
    },
    {
      "entropy": 0.4920670511201024,
      "epoch": 2.6061379750547724,
      "grad_norm": 0.15922629833221436,
      "learning_rate": 1.3132149538442668e-05,
      "loss": 0.489,
      "mean_token_accuracy": 0.8456680476665497,
      "num_tokens": 904031384.0,
      "step": 28326
    },
    {
      "entropy": 0.5252240691334009,
      "epoch": 2.606229981426214,
      "grad_norm": 0.16592803597450256,
      "learning_rate": 1.312908271230104e-05,
      "loss": 0.5217,
      "mean_token_accuracy": 0.8341379314661026,
      "num_tokens": 904063214.0,
      "step": 28327
    },
    {
      "entropy": 0.6303208852186799,
      "epoch": 2.606321987797655,
      "grad_norm": 0.17247822880744934,
      "learning_rate": 1.3126015886159415e-05,
      "loss": 0.6214,
      "mean_token_accuracy": 0.8078691326081753,
      "num_tokens": 904094443.0,
      "step": 28328
    },
    {
      "entropy": 0.5370352249592543,
      "epoch": 2.6064139941690962,
      "grad_norm": 0.154293954372406,
      "learning_rate": 1.3122949060017788e-05,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8342380970716476,
      "num_tokens": 904126484.0,
      "step": 28329
    },
    {
      "entropy": 0.5660886317491531,
      "epoch": 2.6065060005405374,
      "grad_norm": 0.1669246405363083,
      "learning_rate": 1.3119882233876163e-05,
      "loss": 0.5486,
      "mean_token_accuracy": 0.8243119120597839,
      "num_tokens": 904157958.0,
      "step": 28330
    },
    {
      "entropy": 0.5854050479829311,
      "epoch": 2.6065980069119785,
      "grad_norm": 0.17219731211662292,
      "learning_rate": 1.3116815407734536e-05,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8160492070019245,
      "num_tokens": 904188770.0,
      "step": 28331
    },
    {
      "entropy": 0.49717781599611044,
      "epoch": 2.60669001328342,
      "grad_norm": 0.15709663927555084,
      "learning_rate": 1.311374858159291e-05,
      "loss": 0.4793,
      "mean_token_accuracy": 0.8458669297397137,
      "num_tokens": 904220669.0,
      "step": 28332
    },
    {
      "entropy": 0.6321200234815478,
      "epoch": 2.6067820196548612,
      "grad_norm": 0.17484086751937866,
      "learning_rate": 1.3110681755451285e-05,
      "loss": 0.6273,
      "mean_token_accuracy": 0.8045862391591072,
      "num_tokens": 904252429.0,
      "step": 28333
    },
    {
      "entropy": 0.4589832443743944,
      "epoch": 2.6068740260263024,
      "grad_norm": 0.1568067967891693,
      "learning_rate": 1.3107614929309658e-05,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8614049218595028,
      "num_tokens": 904283242.0,
      "step": 28334
    },
    {
      "entropy": 0.46912096976302564,
      "epoch": 2.6069660323977435,
      "grad_norm": 0.1547614187002182,
      "learning_rate": 1.3104548103168033e-05,
      "loss": 0.4551,
      "mean_token_accuracy": 0.8561983183026314,
      "num_tokens": 904315101.0,
      "step": 28335
    },
    {
      "entropy": 0.5045256251469254,
      "epoch": 2.6070580387691846,
      "grad_norm": 0.15626545250415802,
      "learning_rate": 1.3101481277026406e-05,
      "loss": 0.4889,
      "mean_token_accuracy": 0.844804372638464,
      "num_tokens": 904347494.0,
      "step": 28336
    },
    {
      "entropy": 0.5116377202793956,
      "epoch": 2.607150045140626,
      "grad_norm": 0.1570601612329483,
      "learning_rate": 1.309841445088478e-05,
      "loss": 0.4967,
      "mean_token_accuracy": 0.8409989103674889,
      "num_tokens": 904379467.0,
      "step": 28337
    },
    {
      "entropy": 0.5054028565064073,
      "epoch": 2.6072420515120673,
      "grad_norm": 0.154592826962471,
      "learning_rate": 1.3095347624743153e-05,
      "loss": 0.4977,
      "mean_token_accuracy": 0.8430786617100239,
      "num_tokens": 904411880.0,
      "step": 28338
    },
    {
      "entropy": 0.5233326777815819,
      "epoch": 2.6073340578835085,
      "grad_norm": 0.15749706327915192,
      "learning_rate": 1.3092280798601528e-05,
      "loss": 0.5098,
      "mean_token_accuracy": 0.8398164436221123,
      "num_tokens": 904444645.0,
      "step": 28339
    },
    {
      "entropy": 0.5656864824704826,
      "epoch": 2.6074260642549496,
      "grad_norm": 0.16307906806468964,
      "learning_rate": 1.3089213972459901e-05,
      "loss": 0.5455,
      "mean_token_accuracy": 0.8257585801184177,
      "num_tokens": 904476566.0,
      "step": 28340
    },
    {
      "entropy": 0.6505445586517453,
      "epoch": 2.6075180706263907,
      "grad_norm": 0.17280250787734985,
      "learning_rate": 1.3086147146318276e-05,
      "loss": 0.6398,
      "mean_token_accuracy": 0.797905471175909,
      "num_tokens": 904508685.0,
      "step": 28341
    },
    {
      "entropy": 0.548507702536881,
      "epoch": 2.6076100769978323,
      "grad_norm": 0.1623714715242386,
      "learning_rate": 1.3083080320176649e-05,
      "loss": 0.5398,
      "mean_token_accuracy": 0.8315539285540581,
      "num_tokens": 904541202.0,
      "step": 28342
    },
    {
      "entropy": 0.43674240773543715,
      "epoch": 2.6077020833692734,
      "grad_norm": 0.15180759131908417,
      "learning_rate": 1.3080013494035023e-05,
      "loss": 0.4256,
      "mean_token_accuracy": 0.864042017608881,
      "num_tokens": 904573547.0,
      "step": 28343
    },
    {
      "entropy": 0.5441841860301793,
      "epoch": 2.6077940897407146,
      "grad_norm": 0.16298438608646393,
      "learning_rate": 1.3076946667893398e-05,
      "loss": 0.529,
      "mean_token_accuracy": 0.8314981199800968,
      "num_tokens": 904606126.0,
      "step": 28344
    },
    {
      "entropy": 0.49475818895734847,
      "epoch": 2.6078860961121557,
      "grad_norm": 0.15917466580867767,
      "learning_rate": 1.3073879841751771e-05,
      "loss": 0.4881,
      "mean_token_accuracy": 0.8475888483226299,
      "num_tokens": 904637905.0,
      "step": 28345
    },
    {
      "entropy": 0.5239962125197053,
      "epoch": 2.607978102483597,
      "grad_norm": 0.1600116342306137,
      "learning_rate": 1.3070813015610146e-05,
      "loss": 0.508,
      "mean_token_accuracy": 0.8372312448918819,
      "num_tokens": 904670056.0,
      "step": 28346
    },
    {
      "entropy": 0.5213222680613399,
      "epoch": 2.6080701088550384,
      "grad_norm": 0.16327176988124847,
      "learning_rate": 1.3067746189468518e-05,
      "loss": 0.5041,
      "mean_token_accuracy": 0.839937150478363,
      "num_tokens": 904701166.0,
      "step": 28347
    },
    {
      "entropy": 0.5286225089803338,
      "epoch": 2.6081621152264796,
      "grad_norm": 0.1623900979757309,
      "learning_rate": 1.3064679363326893e-05,
      "loss": 0.5116,
      "mean_token_accuracy": 0.8394653350114822,
      "num_tokens": 904732247.0,
      "step": 28348
    },
    {
      "entropy": 0.4474559426307678,
      "epoch": 2.6082541215979207,
      "grad_norm": 0.15163986384868622,
      "learning_rate": 1.3061612537185266e-05,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8605047129094601,
      "num_tokens": 904763357.0,
      "step": 28349
    },
    {
      "entropy": 0.6489016357809305,
      "epoch": 2.608346127969362,
      "grad_norm": 0.17236848175525665,
      "learning_rate": 1.305854571104364e-05,
      "loss": 0.6324,
      "mean_token_accuracy": 0.802093643695116,
      "num_tokens": 904795468.0,
      "step": 28350
    },
    {
      "entropy": 0.4636388423386961,
      "epoch": 2.608438134340803,
      "grad_norm": 0.1514209806919098,
      "learning_rate": 1.3055478884902014e-05,
      "loss": 0.4552,
      "mean_token_accuracy": 0.8553577065467834,
      "num_tokens": 904827795.0,
      "step": 28351
    },
    {
      "entropy": 0.5101281721144915,
      "epoch": 2.6085301407122445,
      "grad_norm": 0.1591917723417282,
      "learning_rate": 1.3052412058760388e-05,
      "loss": 0.4845,
      "mean_token_accuracy": 0.8426341414451599,
      "num_tokens": 904859154.0,
      "step": 28352
    },
    {
      "entropy": 0.5887274602428079,
      "epoch": 2.6086221470836857,
      "grad_norm": 0.16981323063373566,
      "learning_rate": 1.3049345232618765e-05,
      "loss": 0.5806,
      "mean_token_accuracy": 0.8184574171900749,
      "num_tokens": 904890765.0,
      "step": 28353
    },
    {
      "entropy": 0.43994106352329254,
      "epoch": 2.608714153455127,
      "grad_norm": 0.15104492008686066,
      "learning_rate": 1.304627840647714e-05,
      "loss": 0.4365,
      "mean_token_accuracy": 0.858371376991272,
      "num_tokens": 904922952.0,
      "step": 28354
    },
    {
      "entropy": 0.526735388673842,
      "epoch": 2.608806159826568,
      "grad_norm": 0.16187414526939392,
      "learning_rate": 1.3043211580335512e-05,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8355785608291626,
      "num_tokens": 904954855.0,
      "step": 28355
    },
    {
      "entropy": 0.63600654900074,
      "epoch": 2.608898166198009,
      "grad_norm": 0.16589877009391785,
      "learning_rate": 1.3040144754193887e-05,
      "loss": 0.6297,
      "mean_token_accuracy": 0.8094808422029018,
      "num_tokens": 904986875.0,
      "step": 28356
    },
    {
      "entropy": 0.44948428007774055,
      "epoch": 2.6089901725694506,
      "grad_norm": 0.14481844007968903,
      "learning_rate": 1.303707792805226e-05,
      "loss": 0.4471,
      "mean_token_accuracy": 0.8601429350674152,
      "num_tokens": 905019470.0,
      "step": 28357
    },
    {
      "entropy": 0.46023148437961936,
      "epoch": 2.609082178940892,
      "grad_norm": 0.15091006457805634,
      "learning_rate": 1.3034011101910635e-05,
      "loss": 0.4562,
      "mean_token_accuracy": 0.8532326333224773,
      "num_tokens": 905051738.0,
      "step": 28358
    },
    {
      "entropy": 0.5158172380179167,
      "epoch": 2.609174185312333,
      "grad_norm": 0.1635555922985077,
      "learning_rate": 1.3030944275769007e-05,
      "loss": 0.511,
      "mean_token_accuracy": 0.8366108536720276,
      "num_tokens": 905084219.0,
      "step": 28359
    },
    {
      "entropy": 0.6018341388553381,
      "epoch": 2.609266191683774,
      "grad_norm": 0.16956281661987305,
      "learning_rate": 1.3027877449627382e-05,
      "loss": 0.5949,
      "mean_token_accuracy": 0.8119612783193588,
      "num_tokens": 905116700.0,
      "step": 28360
    },
    {
      "entropy": 0.48768760170787573,
      "epoch": 2.609358198055215,
      "grad_norm": 0.15172818303108215,
      "learning_rate": 1.3024810623485755e-05,
      "loss": 0.4736,
      "mean_token_accuracy": 0.8474522531032562,
      "num_tokens": 905148303.0,
      "step": 28361
    },
    {
      "entropy": 0.6051319167017937,
      "epoch": 2.6094502044266568,
      "grad_norm": 0.16769322752952576,
      "learning_rate": 1.302174379734413e-05,
      "loss": 0.5742,
      "mean_token_accuracy": 0.8185567259788513,
      "num_tokens": 905179075.0,
      "step": 28362
    },
    {
      "entropy": 0.5569892087951303,
      "epoch": 2.609542210798098,
      "grad_norm": 0.16070696711540222,
      "learning_rate": 1.3018676971202504e-05,
      "loss": 0.5473,
      "mean_token_accuracy": 0.8239555433392525,
      "num_tokens": 905210919.0,
      "step": 28363
    },
    {
      "entropy": 0.4555729627609253,
      "epoch": 2.609634217169539,
      "grad_norm": 0.151001438498497,
      "learning_rate": 1.3015610145060877e-05,
      "loss": 0.4477,
      "mean_token_accuracy": 0.8538888283073902,
      "num_tokens": 905243129.0,
      "step": 28364
    },
    {
      "entropy": 0.5702556045725942,
      "epoch": 2.60972622354098,
      "grad_norm": 0.1661531925201416,
      "learning_rate": 1.3012543318919252e-05,
      "loss": 0.542,
      "mean_token_accuracy": 0.8279397450387478,
      "num_tokens": 905274688.0,
      "step": 28365
    },
    {
      "entropy": 0.4867382328957319,
      "epoch": 2.6098182299124213,
      "grad_norm": 0.15781421959400177,
      "learning_rate": 1.3009476492777625e-05,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8477033711969852,
      "num_tokens": 905306564.0,
      "step": 28366
    },
    {
      "entropy": 0.501534512732178,
      "epoch": 2.609910236283863,
      "grad_norm": 0.15479372441768646,
      "learning_rate": 1.3006409666636e-05,
      "loss": 0.4866,
      "mean_token_accuracy": 0.8454982787370682,
      "num_tokens": 905338778.0,
      "step": 28367
    },
    {
      "entropy": 0.49042975064367056,
      "epoch": 2.610002242655304,
      "grad_norm": 0.15344715118408203,
      "learning_rate": 1.3003342840494373e-05,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8496611751616001,
      "num_tokens": 905371126.0,
      "step": 28368
    },
    {
      "entropy": 0.5401394236832857,
      "epoch": 2.610094249026745,
      "grad_norm": 0.16993792355060577,
      "learning_rate": 1.3000276014352747e-05,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8260634355247021,
      "num_tokens": 905402811.0,
      "step": 28369
    },
    {
      "entropy": 0.46417327458038926,
      "epoch": 2.6101862553981863,
      "grad_norm": 0.15596391260623932,
      "learning_rate": 1.299720918821112e-05,
      "loss": 0.4547,
      "mean_token_accuracy": 0.8569624200463295,
      "num_tokens": 905435052.0,
      "step": 28370
    },
    {
      "entropy": 0.4800503448350355,
      "epoch": 2.6102782617696274,
      "grad_norm": 0.15207639336585999,
      "learning_rate": 1.2994142362069495e-05,
      "loss": 0.4575,
      "mean_token_accuracy": 0.8542159460484982,
      "num_tokens": 905466266.0,
      "step": 28371
    },
    {
      "entropy": 0.5743696615099907,
      "epoch": 2.610370268141069,
      "grad_norm": 0.16394104063510895,
      "learning_rate": 1.2991075535927868e-05,
      "loss": 0.5605,
      "mean_token_accuracy": 0.821817371994257,
      "num_tokens": 905498334.0,
      "step": 28372
    },
    {
      "entropy": 0.5432433411478996,
      "epoch": 2.61046227451251,
      "grad_norm": 0.16909398138523102,
      "learning_rate": 1.2988008709786242e-05,
      "loss": 0.5436,
      "mean_token_accuracy": 0.828057125210762,
      "num_tokens": 905530697.0,
      "step": 28373
    },
    {
      "entropy": 0.4852940384298563,
      "epoch": 2.6105542808839513,
      "grad_norm": 0.15646135807037354,
      "learning_rate": 1.2984941883644617e-05,
      "loss": 0.4607,
      "mean_token_accuracy": 0.8523574434220791,
      "num_tokens": 905561855.0,
      "step": 28374
    },
    {
      "entropy": 0.46901047322899103,
      "epoch": 2.6106462872553924,
      "grad_norm": 0.14844255149364471,
      "learning_rate": 1.298187505750299e-05,
      "loss": 0.4619,
      "mean_token_accuracy": 0.854230560362339,
      "num_tokens": 905593982.0,
      "step": 28375
    },
    {
      "entropy": 0.4322726195678115,
      "epoch": 2.6107382936268335,
      "grad_norm": 0.15258347988128662,
      "learning_rate": 1.2978808231361365e-05,
      "loss": 0.4135,
      "mean_token_accuracy": 0.8647962771356106,
      "num_tokens": 905625854.0,
      "step": 28376
    },
    {
      "entropy": 0.39859637524932623,
      "epoch": 2.610830299998275,
      "grad_norm": 0.1435329169034958,
      "learning_rate": 1.2975741405219738e-05,
      "loss": 0.3859,
      "mean_token_accuracy": 0.8740363605320454,
      "num_tokens": 905657992.0,
      "step": 28377
    },
    {
      "entropy": 0.5051388142164797,
      "epoch": 2.6109223063697162,
      "grad_norm": 0.15514451265335083,
      "learning_rate": 1.2972674579078112e-05,
      "loss": 0.4911,
      "mean_token_accuracy": 0.845084223896265,
      "num_tokens": 905689985.0,
      "step": 28378
    },
    {
      "entropy": 0.6106614880263805,
      "epoch": 2.6110143127411574,
      "grad_norm": 0.17078816890716553,
      "learning_rate": 1.2969607752936485e-05,
      "loss": 0.5984,
      "mean_token_accuracy": 0.813265647739172,
      "num_tokens": 905721689.0,
      "step": 28379
    },
    {
      "entropy": 0.4662431748583913,
      "epoch": 2.6111063191125985,
      "grad_norm": 0.15506166219711304,
      "learning_rate": 1.296654092679486e-05,
      "loss": 0.4622,
      "mean_token_accuracy": 0.8535302318632603,
      "num_tokens": 905753337.0,
      "step": 28380
    },
    {
      "entropy": 0.5413099071010947,
      "epoch": 2.6111983254840396,
      "grad_norm": 0.16088257730007172,
      "learning_rate": 1.2963474100653233e-05,
      "loss": 0.5293,
      "mean_token_accuracy": 0.8309621028602123,
      "num_tokens": 905785241.0,
      "step": 28381
    },
    {
      "entropy": 0.43651029164902866,
      "epoch": 2.611290331855481,
      "grad_norm": 0.1536710560321808,
      "learning_rate": 1.2960407274511607e-05,
      "loss": 0.4145,
      "mean_token_accuracy": 0.8654760420322418,
      "num_tokens": 905816391.0,
      "step": 28382
    },
    {
      "entropy": 0.529251717031002,
      "epoch": 2.6113823382269223,
      "grad_norm": 0.1600816398859024,
      "learning_rate": 1.295734044836998e-05,
      "loss": 0.516,
      "mean_token_accuracy": 0.8305042199790478,
      "num_tokens": 905848159.0,
      "step": 28383
    },
    {
      "entropy": 0.4901947192847729,
      "epoch": 2.6114743445983635,
      "grad_norm": 0.16048192977905273,
      "learning_rate": 1.2954273622228358e-05,
      "loss": 0.4792,
      "mean_token_accuracy": 0.8430610299110413,
      "num_tokens": 905880217.0,
      "step": 28384
    },
    {
      "entropy": 0.4864494316279888,
      "epoch": 2.6115663509698046,
      "grad_norm": 0.15533064305782318,
      "learning_rate": 1.2951206796086731e-05,
      "loss": 0.4725,
      "mean_token_accuracy": 0.8485313914716244,
      "num_tokens": 905912702.0,
      "step": 28385
    },
    {
      "entropy": 0.504498478025198,
      "epoch": 2.6116583573412457,
      "grad_norm": 0.16060665249824524,
      "learning_rate": 1.2948139969945106e-05,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8374770320951939,
      "num_tokens": 905945195.0,
      "step": 28386
    },
    {
      "entropy": 0.5254383044084534,
      "epoch": 2.6117503637126873,
      "grad_norm": 0.1577567458152771,
      "learning_rate": 1.2945073143803479e-05,
      "loss": 0.5215,
      "mean_token_accuracy": 0.8364589735865593,
      "num_tokens": 905977133.0,
      "step": 28387
    },
    {
      "entropy": 0.5506015210412443,
      "epoch": 2.6118423700841285,
      "grad_norm": 0.16985046863555908,
      "learning_rate": 1.2942006317661854e-05,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8350461907684803,
      "num_tokens": 906009012.0,
      "step": 28388
    },
    {
      "entropy": 0.4494668892584741,
      "epoch": 2.6119343764555696,
      "grad_norm": 0.14876632392406464,
      "learning_rate": 1.2938939491520227e-05,
      "loss": 0.4338,
      "mean_token_accuracy": 0.8614015243947506,
      "num_tokens": 906040549.0,
      "step": 28389
    },
    {
      "entropy": 0.436582894064486,
      "epoch": 2.6120263828270107,
      "grad_norm": 0.14660252630710602,
      "learning_rate": 1.2935872665378601e-05,
      "loss": 0.431,
      "mean_token_accuracy": 0.8657973445951939,
      "num_tokens": 906072774.0,
      "step": 28390
    },
    {
      "entropy": 0.5749768873210996,
      "epoch": 2.612118389198452,
      "grad_norm": 0.16928085684776306,
      "learning_rate": 1.2932805839236974e-05,
      "loss": 0.5569,
      "mean_token_accuracy": 0.8247369416058064,
      "num_tokens": 906104458.0,
      "step": 28391
    },
    {
      "entropy": 0.5502224704250693,
      "epoch": 2.6122103955698934,
      "grad_norm": 0.16417285799980164,
      "learning_rate": 1.2929739013095349e-05,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8284425772726536,
      "num_tokens": 906136293.0,
      "step": 28392
    },
    {
      "entropy": 0.6164088556542993,
      "epoch": 2.6123024019413346,
      "grad_norm": 0.17010629177093506,
      "learning_rate": 1.2926672186953723e-05,
      "loss": 0.6058,
      "mean_token_accuracy": 0.809988472610712,
      "num_tokens": 906168358.0,
      "step": 28393
    },
    {
      "entropy": 0.5851517580449581,
      "epoch": 2.6123944083127757,
      "grad_norm": 0.16688279807567596,
      "learning_rate": 1.2923605360812096e-05,
      "loss": 0.5644,
      "mean_token_accuracy": 0.8209769800305367,
      "num_tokens": 906199780.0,
      "step": 28394
    },
    {
      "entropy": 0.5865766904316843,
      "epoch": 2.612486414684217,
      "grad_norm": 0.16549886763095856,
      "learning_rate": 1.2920538534670471e-05,
      "loss": 0.5746,
      "mean_token_accuracy": 0.8189774416387081,
      "num_tokens": 906230930.0,
      "step": 28395
    },
    {
      "entropy": 0.4475047034211457,
      "epoch": 2.612578421055658,
      "grad_norm": 0.14966677129268646,
      "learning_rate": 1.2917471708528844e-05,
      "loss": 0.4336,
      "mean_token_accuracy": 0.8587244562804699,
      "num_tokens": 906262580.0,
      "step": 28396
    },
    {
      "entropy": 0.5353839974850416,
      "epoch": 2.6126704274270995,
      "grad_norm": 0.16502976417541504,
      "learning_rate": 1.2914404882387219e-05,
      "loss": 0.5326,
      "mean_token_accuracy": 0.8314050994813442,
      "num_tokens": 906294313.0,
      "step": 28397
    },
    {
      "entropy": 0.5237410767003894,
      "epoch": 2.6127624337985407,
      "grad_norm": 0.15732704102993011,
      "learning_rate": 1.2911338056245592e-05,
      "loss": 0.497,
      "mean_token_accuracy": 0.8385354094207287,
      "num_tokens": 906326821.0,
      "step": 28398
    },
    {
      "entropy": 0.5319394934922457,
      "epoch": 2.612854440169982,
      "grad_norm": 0.1623082011938095,
      "learning_rate": 1.2908271230103966e-05,
      "loss": 0.5253,
      "mean_token_accuracy": 0.8306932859122753,
      "num_tokens": 906358981.0,
      "step": 28399
    },
    {
      "entropy": 0.5501943994313478,
      "epoch": 2.612946446541423,
      "grad_norm": 0.1658920794725418,
      "learning_rate": 1.290520440396234e-05,
      "loss": 0.533,
      "mean_token_accuracy": 0.8310543037950993,
      "num_tokens": 906390365.0,
      "step": 28400
    },
    {
      "entropy": 0.5399463837966323,
      "epoch": 2.613038452912864,
      "grad_norm": 0.16711640357971191,
      "learning_rate": 1.2902137577820714e-05,
      "loss": 0.533,
      "mean_token_accuracy": 0.829064279794693,
      "num_tokens": 906422524.0,
      "step": 28401
    },
    {
      "entropy": 0.4728966448456049,
      "epoch": 2.6131304592843057,
      "grad_norm": 0.1559445559978485,
      "learning_rate": 1.2899070751679087e-05,
      "loss": 0.4637,
      "mean_token_accuracy": 0.8551277816295624,
      "num_tokens": 906454790.0,
      "step": 28402
    },
    {
      "entropy": 0.43484435556456447,
      "epoch": 2.613222465655747,
      "grad_norm": 0.15371328592300415,
      "learning_rate": 1.2896003925537461e-05,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8659226410090923,
      "num_tokens": 906487151.0,
      "step": 28403
    },
    {
      "entropy": 0.4842186700552702,
      "epoch": 2.613314472027188,
      "grad_norm": 0.15425963699817657,
      "learning_rate": 1.2892937099395836e-05,
      "loss": 0.478,
      "mean_token_accuracy": 0.8467161245644093,
      "num_tokens": 906519029.0,
      "step": 28404
    },
    {
      "entropy": 0.5729449857026339,
      "epoch": 2.613406478398629,
      "grad_norm": 0.15991947054862976,
      "learning_rate": 1.2889870273254209e-05,
      "loss": 0.542,
      "mean_token_accuracy": 0.8307716548442841,
      "num_tokens": 906551586.0,
      "step": 28405
    },
    {
      "entropy": 0.49834209866821766,
      "epoch": 2.61349848477007,
      "grad_norm": 0.16002613306045532,
      "learning_rate": 1.2886803447112584e-05,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8442013747990131,
      "num_tokens": 906582783.0,
      "step": 28406
    },
    {
      "entropy": 0.5567482211627066,
      "epoch": 2.6135904911415118,
      "grad_norm": 0.16171160340309143,
      "learning_rate": 1.2883736620970957e-05,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8256620950996876,
      "num_tokens": 906614935.0,
      "step": 28407
    },
    {
      "entropy": 0.47619082941673696,
      "epoch": 2.613682497512953,
      "grad_norm": 0.14954189956188202,
      "learning_rate": 1.2880669794829331e-05,
      "loss": 0.4648,
      "mean_token_accuracy": 0.8513820059597492,
      "num_tokens": 906647472.0,
      "step": 28408
    },
    {
      "entropy": 0.5053963495884091,
      "epoch": 2.613774503884394,
      "grad_norm": 0.15807569026947021,
      "learning_rate": 1.2877602968687704e-05,
      "loss": 0.496,
      "mean_token_accuracy": 0.8442303463816643,
      "num_tokens": 906678842.0,
      "step": 28409
    },
    {
      "entropy": 0.500546651892364,
      "epoch": 2.613866510255835,
      "grad_norm": 0.1568690687417984,
      "learning_rate": 1.2874536142546079e-05,
      "loss": 0.4915,
      "mean_token_accuracy": 0.8433375470340252,
      "num_tokens": 906710844.0,
      "step": 28410
    },
    {
      "entropy": 0.4633509134873748,
      "epoch": 2.6139585166272763,
      "grad_norm": 0.14797885715961456,
      "learning_rate": 1.2871469316404452e-05,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8583218529820442,
      "num_tokens": 906743576.0,
      "step": 28411
    },
    {
      "entropy": 0.44704198942054063,
      "epoch": 2.614050522998718,
      "grad_norm": 0.14615003764629364,
      "learning_rate": 1.2868402490262827e-05,
      "loss": 0.4203,
      "mean_token_accuracy": 0.8623773753643036,
      "num_tokens": 906775133.0,
      "step": 28412
    },
    {
      "entropy": 0.46132844872772694,
      "epoch": 2.614142529370159,
      "grad_norm": 0.1512235701084137,
      "learning_rate": 1.28653356641212e-05,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8579444140195847,
      "num_tokens": 906807311.0,
      "step": 28413
    },
    {
      "entropy": 0.4887592336162925,
      "epoch": 2.6142345357416,
      "grad_norm": 0.15811842679977417,
      "learning_rate": 1.2862268837979578e-05,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8474787101149559,
      "num_tokens": 906838835.0,
      "step": 28414
    },
    {
      "entropy": 0.46245009219273925,
      "epoch": 2.6143265421130413,
      "grad_norm": 0.1515703797340393,
      "learning_rate": 1.285920201183795e-05,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8599442131817341,
      "num_tokens": 906870572.0,
      "step": 28415
    },
    {
      "entropy": 0.5012618713080883,
      "epoch": 2.6144185484844824,
      "grad_norm": 0.16449959576129913,
      "learning_rate": 1.2856135185696325e-05,
      "loss": 0.496,
      "mean_token_accuracy": 0.8398482464253902,
      "num_tokens": 906902703.0,
      "step": 28416
    },
    {
      "entropy": 0.5157023975625634,
      "epoch": 2.614510554855924,
      "grad_norm": 0.15683375298976898,
      "learning_rate": 1.2853068359554698e-05,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8415536917746067,
      "num_tokens": 906935212.0,
      "step": 28417
    },
    {
      "entropy": 0.49589983001351357,
      "epoch": 2.614602561227365,
      "grad_norm": 0.15687665343284607,
      "learning_rate": 1.2850001533413073e-05,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8488186523318291,
      "num_tokens": 906966929.0,
      "step": 28418
    },
    {
      "entropy": 0.48341296426951885,
      "epoch": 2.6146945675988063,
      "grad_norm": 0.15189354121685028,
      "learning_rate": 1.2846934707271446e-05,
      "loss": 0.4718,
      "mean_token_accuracy": 0.8491763211786747,
      "num_tokens": 906998950.0,
      "step": 28419
    },
    {
      "entropy": 0.5115440357476473,
      "epoch": 2.6147865739702474,
      "grad_norm": 0.1626502275466919,
      "learning_rate": 1.284386788112982e-05,
      "loss": 0.4982,
      "mean_token_accuracy": 0.8405604623258114,
      "num_tokens": 907030060.0,
      "step": 28420
    },
    {
      "entropy": 0.49542979476973414,
      "epoch": 2.6148785803416885,
      "grad_norm": 0.16216541826725006,
      "learning_rate": 1.2840801054988193e-05,
      "loss": 0.485,
      "mean_token_accuracy": 0.8488692343235016,
      "num_tokens": 907062342.0,
      "step": 28421
    },
    {
      "entropy": 0.5223367316648364,
      "epoch": 2.61497058671313,
      "grad_norm": 0.15909169614315033,
      "learning_rate": 1.2837734228846568e-05,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8400824405252934,
      "num_tokens": 907093655.0,
      "step": 28422
    },
    {
      "entropy": 0.4810067377984524,
      "epoch": 2.6150625930845712,
      "grad_norm": 0.1540018916130066,
      "learning_rate": 1.2834667402704943e-05,
      "loss": 0.4651,
      "mean_token_accuracy": 0.8500803038477898,
      "num_tokens": 907125301.0,
      "step": 28423
    },
    {
      "entropy": 0.5024853395298123,
      "epoch": 2.6151545994560124,
      "grad_norm": 0.15908902883529663,
      "learning_rate": 1.2831600576563316e-05,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8425003550946712,
      "num_tokens": 907157674.0,
      "step": 28424
    },
    {
      "entropy": 0.49803860671818256,
      "epoch": 2.6152466058274535,
      "grad_norm": 0.15546174347400665,
      "learning_rate": 1.282853375042169e-05,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8431130908429623,
      "num_tokens": 907189309.0,
      "step": 28425
    },
    {
      "entropy": 0.5554602444171906,
      "epoch": 2.6153386121988946,
      "grad_norm": 0.16560494899749756,
      "learning_rate": 1.2825466924280063e-05,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8296909593045712,
      "num_tokens": 907220802.0,
      "step": 28426
    },
    {
      "entropy": 0.4486215626820922,
      "epoch": 2.615430618570336,
      "grad_norm": 0.15424714982509613,
      "learning_rate": 1.2822400098138438e-05,
      "loss": 0.442,
      "mean_token_accuracy": 0.8586292751133442,
      "num_tokens": 907253159.0,
      "step": 28427
    },
    {
      "entropy": 0.5954408594407141,
      "epoch": 2.6155226249417773,
      "grad_norm": 0.17455174028873444,
      "learning_rate": 1.281933327199681e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.812520831823349,
      "num_tokens": 907284712.0,
      "step": 28428
    },
    {
      "entropy": 0.4724067132920027,
      "epoch": 2.6156146313132185,
      "grad_norm": 0.15753591060638428,
      "learning_rate": 1.2816266445855185e-05,
      "loss": 0.4656,
      "mean_token_accuracy": 0.851701233536005,
      "num_tokens": 907316987.0,
      "step": 28429
    },
    {
      "entropy": 0.47951302770525217,
      "epoch": 2.6157066376846596,
      "grad_norm": 0.1570725440979004,
      "learning_rate": 1.2813199619713558e-05,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8493788056075573,
      "num_tokens": 907349004.0,
      "step": 28430
    },
    {
      "entropy": 0.6451071090996265,
      "epoch": 2.6157986440561007,
      "grad_norm": 0.17924527823925018,
      "learning_rate": 1.2810132793571933e-05,
      "loss": 0.6296,
      "mean_token_accuracy": 0.8062796853482723,
      "num_tokens": 907379898.0,
      "step": 28431
    },
    {
      "entropy": 0.5338063463568687,
      "epoch": 2.6158906504275423,
      "grad_norm": 0.16097848117351532,
      "learning_rate": 1.2807065967430306e-05,
      "loss": 0.5347,
      "mean_token_accuracy": 0.83221635222435,
      "num_tokens": 907412007.0,
      "step": 28432
    },
    {
      "entropy": 0.5264148423448205,
      "epoch": 2.6159826567989835,
      "grad_norm": 0.16407915949821472,
      "learning_rate": 1.280399914128868e-05,
      "loss": 0.5213,
      "mean_token_accuracy": 0.8373141400516033,
      "num_tokens": 907443236.0,
      "step": 28433
    },
    {
      "entropy": 0.41364247631281614,
      "epoch": 2.6160746631704246,
      "grad_norm": 0.14656567573547363,
      "learning_rate": 1.2800932315147055e-05,
      "loss": 0.4035,
      "mean_token_accuracy": 0.8709007129073143,
      "num_tokens": 907475420.0,
      "step": 28434
    },
    {
      "entropy": 0.5258886618539691,
      "epoch": 2.6161666695418657,
      "grad_norm": 0.16052483022212982,
      "learning_rate": 1.2797865489005428e-05,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8382388837635517,
      "num_tokens": 907508188.0,
      "step": 28435
    },
    {
      "entropy": 0.4918973036110401,
      "epoch": 2.616258675913307,
      "grad_norm": 0.15632718801498413,
      "learning_rate": 1.2794798662863803e-05,
      "loss": 0.4753,
      "mean_token_accuracy": 0.8500227630138397,
      "num_tokens": 907540203.0,
      "step": 28436
    },
    {
      "entropy": 0.5370103530585766,
      "epoch": 2.6163506822847484,
      "grad_norm": 0.16664482653141022,
      "learning_rate": 1.2791731836722176e-05,
      "loss": 0.5138,
      "mean_token_accuracy": 0.839014895260334,
      "num_tokens": 907571209.0,
      "step": 28437
    },
    {
      "entropy": 0.5033542960882187,
      "epoch": 2.6164426886561896,
      "grad_norm": 0.16290391981601715,
      "learning_rate": 1.278866501058055e-05,
      "loss": 0.4887,
      "mean_token_accuracy": 0.8476892188191414,
      "num_tokens": 907602556.0,
      "step": 28438
    },
    {
      "entropy": 0.48884293297305703,
      "epoch": 2.6165346950276307,
      "grad_norm": 0.15996211767196655,
      "learning_rate": 1.2785598184438923e-05,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8453663438558578,
      "num_tokens": 907634288.0,
      "step": 28439
    },
    {
      "entropy": 0.4135146589251235,
      "epoch": 2.616626701399072,
      "grad_norm": 0.14101813733577728,
      "learning_rate": 1.2782531358297298e-05,
      "loss": 0.4099,
      "mean_token_accuracy": 0.8733113780617714,
      "num_tokens": 907665419.0,
      "step": 28440
    },
    {
      "entropy": 0.4985527840908617,
      "epoch": 2.616718707770513,
      "grad_norm": 0.15886470675468445,
      "learning_rate": 1.2779464532155671e-05,
      "loss": 0.4946,
      "mean_token_accuracy": 0.8455152288079262,
      "num_tokens": 907697563.0,
      "step": 28441
    },
    {
      "entropy": 0.5607749689370394,
      "epoch": 2.6168107141419545,
      "grad_norm": 0.16200730204582214,
      "learning_rate": 1.2776397706014046e-05,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8284430876374245,
      "num_tokens": 907729385.0,
      "step": 28442
    },
    {
      "entropy": 0.5021861912682652,
      "epoch": 2.6169027205133957,
      "grad_norm": 0.15785303711891174,
      "learning_rate": 1.2773330879872419e-05,
      "loss": 0.4863,
      "mean_token_accuracy": 0.8454691357910633,
      "num_tokens": 907760286.0,
      "step": 28443
    },
    {
      "entropy": 0.5360706835053861,
      "epoch": 2.616994726884837,
      "grad_norm": 0.1592334657907486,
      "learning_rate": 1.2770264053730793e-05,
      "loss": 0.5275,
      "mean_token_accuracy": 0.8317881859838963,
      "num_tokens": 907792540.0,
      "step": 28444
    },
    {
      "entropy": 0.5118840970098972,
      "epoch": 2.617086733256278,
      "grad_norm": 0.16009996831417084,
      "learning_rate": 1.276719722758917e-05,
      "loss": 0.5026,
      "mean_token_accuracy": 0.8422753624618053,
      "num_tokens": 907824346.0,
      "step": 28445
    },
    {
      "entropy": 0.6588235311210155,
      "epoch": 2.617178739627719,
      "grad_norm": 0.17193010449409485,
      "learning_rate": 1.2764130401447544e-05,
      "loss": 0.6413,
      "mean_token_accuracy": 0.8020341359078884,
      "num_tokens": 907856745.0,
      "step": 28446
    },
    {
      "entropy": 0.45205259323120117,
      "epoch": 2.6172707459991607,
      "grad_norm": 0.15321390330791473,
      "learning_rate": 1.2761063575305917e-05,
      "loss": 0.4419,
      "mean_token_accuracy": 0.8590777665376663,
      "num_tokens": 907889136.0,
      "step": 28447
    },
    {
      "entropy": 0.42336865002289414,
      "epoch": 2.617362752370602,
      "grad_norm": 0.14363306760787964,
      "learning_rate": 1.2757996749164292e-05,
      "loss": 0.4128,
      "mean_token_accuracy": 0.8683639019727707,
      "num_tokens": 907921172.0,
      "step": 28448
    },
    {
      "entropy": 0.5566326575353742,
      "epoch": 2.617454758742043,
      "grad_norm": 0.16328108310699463,
      "learning_rate": 1.2754929923022665e-05,
      "loss": 0.5386,
      "mean_token_accuracy": 0.82868792116642,
      "num_tokens": 907952799.0,
      "step": 28449
    },
    {
      "entropy": 0.6068343855440617,
      "epoch": 2.617546765113484,
      "grad_norm": 0.16864800453186035,
      "learning_rate": 1.275186309688104e-05,
      "loss": 0.593,
      "mean_token_accuracy": 0.8140186779201031,
      "num_tokens": 907985167.0,
      "step": 28450
    },
    {
      "entropy": 0.5557889677584171,
      "epoch": 2.617638771484925,
      "grad_norm": 0.16484126448631287,
      "learning_rate": 1.2748796270739412e-05,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8275740146636963,
      "num_tokens": 908017102.0,
      "step": 28451
    },
    {
      "entropy": 0.5070758489891887,
      "epoch": 2.6177307778563668,
      "grad_norm": 0.16289915144443512,
      "learning_rate": 1.2745729444597787e-05,
      "loss": 0.5026,
      "mean_token_accuracy": 0.8434283509850502,
      "num_tokens": 908049151.0,
      "step": 28452
    },
    {
      "entropy": 0.5777725242078304,
      "epoch": 2.617822784227808,
      "grad_norm": 0.1661575883626938,
      "learning_rate": 1.2742662618456162e-05,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8190091513097286,
      "num_tokens": 908081368.0,
      "step": 28453
    },
    {
      "entropy": 0.48892054753378034,
      "epoch": 2.617914790599249,
      "grad_norm": 0.15345042943954468,
      "learning_rate": 1.2739595792314535e-05,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8483445309102535,
      "num_tokens": 908114028.0,
      "step": 28454
    },
    {
      "entropy": 0.5136599512770772,
      "epoch": 2.61800679697069,
      "grad_norm": 0.16322869062423706,
      "learning_rate": 1.273652896617291e-05,
      "loss": 0.5012,
      "mean_token_accuracy": 0.8400295116007328,
      "num_tokens": 908146000.0,
      "step": 28455
    },
    {
      "entropy": 0.4791378388181329,
      "epoch": 2.6180988033421313,
      "grad_norm": 0.15779080986976624,
      "learning_rate": 1.2733462140031282e-05,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8508458696305752,
      "num_tokens": 908177736.0,
      "step": 28456
    },
    {
      "entropy": 0.5632055713795125,
      "epoch": 2.618190809713573,
      "grad_norm": 0.1633732169866562,
      "learning_rate": 1.2730395313889657e-05,
      "loss": 0.5552,
      "mean_token_accuracy": 0.8246521055698395,
      "num_tokens": 908209839.0,
      "step": 28457
    },
    {
      "entropy": 0.38095034926664084,
      "epoch": 2.618282816085014,
      "grad_norm": 0.14072102308273315,
      "learning_rate": 1.272732848774803e-05,
      "loss": 0.3692,
      "mean_token_accuracy": 0.8807026296854019,
      "num_tokens": 908242413.0,
      "step": 28458
    },
    {
      "entropy": 0.498297048965469,
      "epoch": 2.618374822456455,
      "grad_norm": 0.15376953780651093,
      "learning_rate": 1.2724261661606404e-05,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8489378243684769,
      "num_tokens": 908274903.0,
      "step": 28459
    },
    {
      "entropy": 0.3838048311881721,
      "epoch": 2.6184668288278963,
      "grad_norm": 0.13801896572113037,
      "learning_rate": 1.2721194835464777e-05,
      "loss": 0.3648,
      "mean_token_accuracy": 0.8794453218579292,
      "num_tokens": 908306599.0,
      "step": 28460
    },
    {
      "entropy": 0.5244917031377554,
      "epoch": 2.6185588351993374,
      "grad_norm": 0.16318616271018982,
      "learning_rate": 1.2718128009323152e-05,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8366780616343021,
      "num_tokens": 908337787.0,
      "step": 28461
    },
    {
      "entropy": 0.5346933221444488,
      "epoch": 2.618650841570779,
      "grad_norm": 0.16216933727264404,
      "learning_rate": 1.2715061183181525e-05,
      "loss": 0.5255,
      "mean_token_accuracy": 0.8331102691590786,
      "num_tokens": 908369789.0,
      "step": 28462
    },
    {
      "entropy": 0.482264406979084,
      "epoch": 2.61874284794222,
      "grad_norm": 0.15435141324996948,
      "learning_rate": 1.27119943570399e-05,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8480692058801651,
      "num_tokens": 908401869.0,
      "step": 28463
    },
    {
      "entropy": 0.47835629142355174,
      "epoch": 2.6188348543136613,
      "grad_norm": 0.1558331847190857,
      "learning_rate": 1.2708927530898274e-05,
      "loss": 0.4699,
      "mean_token_accuracy": 0.848251648247242,
      "num_tokens": 908434019.0,
      "step": 28464
    },
    {
      "entropy": 0.42468306282535195,
      "epoch": 2.6189268606851024,
      "grad_norm": 0.14721408486366272,
      "learning_rate": 1.2705860704756647e-05,
      "loss": 0.4151,
      "mean_token_accuracy": 0.8668523021042347,
      "num_tokens": 908466415.0,
      "step": 28465
    },
    {
      "entropy": 0.49472368881106377,
      "epoch": 2.6190188670565435,
      "grad_norm": 0.1566745638847351,
      "learning_rate": 1.2702793878615022e-05,
      "loss": 0.4866,
      "mean_token_accuracy": 0.8430545069277287,
      "num_tokens": 908498360.0,
      "step": 28466
    },
    {
      "entropy": 0.5221660695970058,
      "epoch": 2.619110873427985,
      "grad_norm": 0.15889960527420044,
      "learning_rate": 1.2699727052473395e-05,
      "loss": 0.5185,
      "mean_token_accuracy": 0.8386930003762245,
      "num_tokens": 908530600.0,
      "step": 28467
    },
    {
      "entropy": 0.5127821303904057,
      "epoch": 2.6192028797994262,
      "grad_norm": 0.15742892026901245,
      "learning_rate": 1.269666022633177e-05,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8411935642361641,
      "num_tokens": 908563311.0,
      "step": 28468
    },
    {
      "entropy": 0.5143163669854403,
      "epoch": 2.6192948861708674,
      "grad_norm": 0.15325585007667542,
      "learning_rate": 1.2693593400190142e-05,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8418286740779877,
      "num_tokens": 908595232.0,
      "step": 28469
    },
    {
      "entropy": 0.5345236957073212,
      "epoch": 2.6193868925423085,
      "grad_norm": 0.16296404600143433,
      "learning_rate": 1.2690526574048517e-05,
      "loss": 0.5241,
      "mean_token_accuracy": 0.834111500531435,
      "num_tokens": 908626737.0,
      "step": 28470
    },
    {
      "entropy": 0.4779608938843012,
      "epoch": 2.6194788989137496,
      "grad_norm": 0.15375496447086334,
      "learning_rate": 1.268745974790689e-05,
      "loss": 0.4664,
      "mean_token_accuracy": 0.8513433448970318,
      "num_tokens": 908659463.0,
      "step": 28471
    },
    {
      "entropy": 0.5440509393811226,
      "epoch": 2.619570905285191,
      "grad_norm": 0.16337433457374573,
      "learning_rate": 1.2684392921765265e-05,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8320069797337055,
      "num_tokens": 908690743.0,
      "step": 28472
    },
    {
      "entropy": 0.5408680494874716,
      "epoch": 2.6196629116566323,
      "grad_norm": 0.16085512936115265,
      "learning_rate": 1.2681326095623638e-05,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8369738571345806,
      "num_tokens": 908722759.0,
      "step": 28473
    },
    {
      "entropy": 0.5228444319218397,
      "epoch": 2.6197549180280735,
      "grad_norm": 0.16351266205310822,
      "learning_rate": 1.2678259269482012e-05,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8354945294559002,
      "num_tokens": 908754798.0,
      "step": 28474
    },
    {
      "entropy": 0.5711441356688738,
      "epoch": 2.6198469243995146,
      "grad_norm": 0.1688336879014969,
      "learning_rate": 1.2675192443340389e-05,
      "loss": 0.5551,
      "mean_token_accuracy": 0.8212306275963783,
      "num_tokens": 908786437.0,
      "step": 28475
    },
    {
      "entropy": 0.524002205580473,
      "epoch": 2.6199389307709557,
      "grad_norm": 0.15644104778766632,
      "learning_rate": 1.2672125617198763e-05,
      "loss": 0.5044,
      "mean_token_accuracy": 0.8397137448191643,
      "num_tokens": 908818304.0,
      "step": 28476
    },
    {
      "entropy": 0.5047356188297272,
      "epoch": 2.6200309371423973,
      "grad_norm": 0.16436022520065308,
      "learning_rate": 1.2669058791057136e-05,
      "loss": 0.4883,
      "mean_token_accuracy": 0.8462070412933826,
      "num_tokens": 908849751.0,
      "step": 28477
    },
    {
      "entropy": 0.5387488435953856,
      "epoch": 2.6201229435138385,
      "grad_norm": 0.1655491292476654,
      "learning_rate": 1.2665991964915511e-05,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8303409032523632,
      "num_tokens": 908881943.0,
      "step": 28478
    },
    {
      "entropy": 0.5604548715054989,
      "epoch": 2.6202149498852796,
      "grad_norm": 0.1673886924982071,
      "learning_rate": 1.2662925138773884e-05,
      "loss": 0.5469,
      "mean_token_accuracy": 0.830522958189249,
      "num_tokens": 908912874.0,
      "step": 28479
    },
    {
      "entropy": 0.5543185919523239,
      "epoch": 2.6203069562567207,
      "grad_norm": 0.16283747553825378,
      "learning_rate": 1.2659858312632259e-05,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8297442831099033,
      "num_tokens": 908945130.0,
      "step": 28480
    },
    {
      "entropy": 0.4837535433471203,
      "epoch": 2.620398962628162,
      "grad_norm": 0.150950625538826,
      "learning_rate": 1.2656791486490631e-05,
      "loss": 0.472,
      "mean_token_accuracy": 0.8494665399193764,
      "num_tokens": 908977794.0,
      "step": 28481
    },
    {
      "entropy": 0.4738158513791859,
      "epoch": 2.6204909689996034,
      "grad_norm": 0.15606895089149475,
      "learning_rate": 1.2653724660349006e-05,
      "loss": 0.4722,
      "mean_token_accuracy": 0.8520720712840557,
      "num_tokens": 909010253.0,
      "step": 28482
    },
    {
      "entropy": 0.49421994574368,
      "epoch": 2.6205829753710446,
      "grad_norm": 0.1562282294034958,
      "learning_rate": 1.265065783420738e-05,
      "loss": 0.486,
      "mean_token_accuracy": 0.8459124341607094,
      "num_tokens": 909042750.0,
      "step": 28483
    },
    {
      "entropy": 0.5929732173681259,
      "epoch": 2.6206749817424857,
      "grad_norm": 0.17257288098335266,
      "learning_rate": 1.2647591008065754e-05,
      "loss": 0.5779,
      "mean_token_accuracy": 0.817989245057106,
      "num_tokens": 909074465.0,
      "step": 28484
    },
    {
      "entropy": 0.5664968844503164,
      "epoch": 2.620766988113927,
      "grad_norm": 0.16820858418941498,
      "learning_rate": 1.2644524181924128e-05,
      "loss": 0.5566,
      "mean_token_accuracy": 0.8242388479411602,
      "num_tokens": 909106422.0,
      "step": 28485
    },
    {
      "entropy": 0.5690034702420235,
      "epoch": 2.620858994485368,
      "grad_norm": 0.1657455712556839,
      "learning_rate": 1.2641457355782501e-05,
      "loss": 0.5662,
      "mean_token_accuracy": 0.8257196843624115,
      "num_tokens": 909138261.0,
      "step": 28486
    },
    {
      "entropy": 0.4818933103233576,
      "epoch": 2.6209510008568095,
      "grad_norm": 0.15625199675559998,
      "learning_rate": 1.2638390529640876e-05,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8512884043157101,
      "num_tokens": 909170663.0,
      "step": 28487
    },
    {
      "entropy": 0.5068693542852998,
      "epoch": 2.6210430072282507,
      "grad_norm": 0.15528902411460876,
      "learning_rate": 1.2635323703499249e-05,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8405085690319538,
      "num_tokens": 909202720.0,
      "step": 28488
    },
    {
      "entropy": 0.4948064526543021,
      "epoch": 2.621135013599692,
      "grad_norm": 0.1528751403093338,
      "learning_rate": 1.2632256877357624e-05,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8479301668703556,
      "num_tokens": 909235473.0,
      "step": 28489
    },
    {
      "entropy": 0.3873220491223037,
      "epoch": 2.621227019971133,
      "grad_norm": 0.14144539833068848,
      "learning_rate": 1.2629190051215997e-05,
      "loss": 0.3775,
      "mean_token_accuracy": 0.8792976886034012,
      "num_tokens": 909267956.0,
      "step": 28490
    },
    {
      "entropy": 0.5360137457028031,
      "epoch": 2.621319026342574,
      "grad_norm": 0.16267043352127075,
      "learning_rate": 1.2626123225074371e-05,
      "loss": 0.5276,
      "mean_token_accuracy": 0.8304232880473137,
      "num_tokens": 909299523.0,
      "step": 28491
    },
    {
      "entropy": 0.5213509043678641,
      "epoch": 2.6214110327140157,
      "grad_norm": 0.15897294878959656,
      "learning_rate": 1.2623056398932744e-05,
      "loss": 0.5042,
      "mean_token_accuracy": 0.8385782688856125,
      "num_tokens": 909331035.0,
      "step": 28492
    },
    {
      "entropy": 0.6062155552208424,
      "epoch": 2.621503039085457,
      "grad_norm": 0.17889709770679474,
      "learning_rate": 1.2619989572791119e-05,
      "loss": 0.6088,
      "mean_token_accuracy": 0.8101099990308285,
      "num_tokens": 909362576.0,
      "step": 28493
    },
    {
      "entropy": 0.4252196728484705,
      "epoch": 2.621595045456898,
      "grad_norm": 0.147322878241539,
      "learning_rate": 1.2616922746649493e-05,
      "loss": 0.42,
      "mean_token_accuracy": 0.8632954619824886,
      "num_tokens": 909394789.0,
      "step": 28494
    },
    {
      "entropy": 0.44485970865935087,
      "epoch": 2.621687051828339,
      "grad_norm": 0.14916880428791046,
      "learning_rate": 1.2613855920507866e-05,
      "loss": 0.4392,
      "mean_token_accuracy": 0.8628986924886703,
      "num_tokens": 909426681.0,
      "step": 28495
    },
    {
      "entropy": 0.5007022339850664,
      "epoch": 2.62177905819978,
      "grad_norm": 0.15526622533798218,
      "learning_rate": 1.2610789094366241e-05,
      "loss": 0.4933,
      "mean_token_accuracy": 0.8436300083994865,
      "num_tokens": 909458679.0,
      "step": 28496
    },
    {
      "entropy": 0.6268287152051926,
      "epoch": 2.6218710645712218,
      "grad_norm": 0.17665192484855652,
      "learning_rate": 1.2607722268224614e-05,
      "loss": 0.6182,
      "mean_token_accuracy": 0.8074675314128399,
      "num_tokens": 909490362.0,
      "step": 28497
    },
    {
      "entropy": 0.41092888195998967,
      "epoch": 2.621963070942663,
      "grad_norm": 0.14891405403614044,
      "learning_rate": 1.2604655442082989e-05,
      "loss": 0.3922,
      "mean_token_accuracy": 0.8730983957648277,
      "num_tokens": 909522170.0,
      "step": 28498
    },
    {
      "entropy": 0.5330140530131757,
      "epoch": 2.622055077314104,
      "grad_norm": 0.15926608443260193,
      "learning_rate": 1.2601588615941362e-05,
      "loss": 0.5202,
      "mean_token_accuracy": 0.8379279710352421,
      "num_tokens": 909554440.0,
      "step": 28499
    },
    {
      "entropy": 0.5314505696296692,
      "epoch": 2.622147083685545,
      "grad_norm": 0.16115252673625946,
      "learning_rate": 1.2598521789799736e-05,
      "loss": 0.5237,
      "mean_token_accuracy": 0.8341157548129559,
      "num_tokens": 909586804.0,
      "step": 28500
    },
    {
      "entropy": 0.5366870099678636,
      "epoch": 2.6222390900569863,
      "grad_norm": 0.1597682535648346,
      "learning_rate": 1.259545496365811e-05,
      "loss": 0.5197,
      "mean_token_accuracy": 0.8375460766255856,
      "num_tokens": 909619169.0,
      "step": 28501
    },
    {
      "entropy": 0.5617544185370207,
      "epoch": 2.622331096428428,
      "grad_norm": 0.16366322338581085,
      "learning_rate": 1.2592388137516484e-05,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8261037692427635,
      "num_tokens": 909650981.0,
      "step": 28502
    },
    {
      "entropy": 0.5216597551479936,
      "epoch": 2.622423102799869,
      "grad_norm": 0.16000941395759583,
      "learning_rate": 1.2589321311374857e-05,
      "loss": 0.5184,
      "mean_token_accuracy": 0.835395734757185,
      "num_tokens": 909682493.0,
      "step": 28503
    },
    {
      "entropy": 0.4814471129793674,
      "epoch": 2.62251510917131,
      "grad_norm": 0.16095900535583496,
      "learning_rate": 1.2586254485233231e-05,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8474032990634441,
      "num_tokens": 909713333.0,
      "step": 28504
    },
    {
      "entropy": 0.5617058463394642,
      "epoch": 2.6226071155427513,
      "grad_norm": 0.16708411276340485,
      "learning_rate": 1.2583187659091606e-05,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8267879113554955,
      "num_tokens": 909745036.0,
      "step": 28505
    },
    {
      "entropy": 0.3730037701316178,
      "epoch": 2.6226991219141924,
      "grad_norm": 0.13812437653541565,
      "learning_rate": 1.2580120832949982e-05,
      "loss": 0.3527,
      "mean_token_accuracy": 0.8897558860480785,
      "num_tokens": 909777338.0,
      "step": 28506
    },
    {
      "entropy": 0.5804052408784628,
      "epoch": 2.622791128285634,
      "grad_norm": 0.17149604856967926,
      "learning_rate": 1.2577054006808355e-05,
      "loss": 0.5736,
      "mean_token_accuracy": 0.8216411471366882,
      "num_tokens": 909809228.0,
      "step": 28507
    },
    {
      "entropy": 0.5940721482038498,
      "epoch": 2.622883134657075,
      "grad_norm": 0.1713334172964096,
      "learning_rate": 1.257398718066673e-05,
      "loss": 0.5848,
      "mean_token_accuracy": 0.8168213032186031,
      "num_tokens": 909841432.0,
      "step": 28508
    },
    {
      "entropy": 0.46102297119796276,
      "epoch": 2.6229751410285163,
      "grad_norm": 0.1497265100479126,
      "learning_rate": 1.2570920354525103e-05,
      "loss": 0.4403,
      "mean_token_accuracy": 0.8599500507116318,
      "num_tokens": 909873495.0,
      "step": 28509
    },
    {
      "entropy": 0.5436926037073135,
      "epoch": 2.6230671473999574,
      "grad_norm": 0.1631515473127365,
      "learning_rate": 1.2567853528383478e-05,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8337978422641754,
      "num_tokens": 909904753.0,
      "step": 28510
    },
    {
      "entropy": 0.5530586948152632,
      "epoch": 2.6231591537713985,
      "grad_norm": 0.16940023005008698,
      "learning_rate": 1.256478670224185e-05,
      "loss": 0.5535,
      "mean_token_accuracy": 0.8281528055667877,
      "num_tokens": 909936537.0,
      "step": 28511
    },
    {
      "entropy": 0.4494082210585475,
      "epoch": 2.62325116014284,
      "grad_norm": 0.1540544331073761,
      "learning_rate": 1.2561719876100225e-05,
      "loss": 0.4327,
      "mean_token_accuracy": 0.8594697788357735,
      "num_tokens": 909967720.0,
      "step": 28512
    },
    {
      "entropy": 0.47317842207849026,
      "epoch": 2.6233431665142812,
      "grad_norm": 0.15196582674980164,
      "learning_rate": 1.25586530499586e-05,
      "loss": 0.459,
      "mean_token_accuracy": 0.8547214232385159,
      "num_tokens": 909999763.0,
      "step": 28513
    },
    {
      "entropy": 0.45542778354138136,
      "epoch": 2.6234351728857224,
      "grad_norm": 0.15048037469387054,
      "learning_rate": 1.2555586223816973e-05,
      "loss": 0.4468,
      "mean_token_accuracy": 0.856327198445797,
      "num_tokens": 910031543.0,
      "step": 28514
    },
    {
      "entropy": 0.5694261705502868,
      "epoch": 2.6235271792571635,
      "grad_norm": 0.1678486317396164,
      "learning_rate": 1.2552519397675347e-05,
      "loss": 0.568,
      "mean_token_accuracy": 0.8234347701072693,
      "num_tokens": 910063513.0,
      "step": 28515
    },
    {
      "entropy": 0.499284902587533,
      "epoch": 2.6236191856286046,
      "grad_norm": 0.15782564878463745,
      "learning_rate": 1.254945257153372e-05,
      "loss": 0.4909,
      "mean_token_accuracy": 0.8450203947722912,
      "num_tokens": 910095970.0,
      "step": 28516
    },
    {
      "entropy": 0.5371801825240254,
      "epoch": 2.623711192000046,
      "grad_norm": 0.1604076623916626,
      "learning_rate": 1.2546385745392095e-05,
      "loss": 0.5158,
      "mean_token_accuracy": 0.8355583027005196,
      "num_tokens": 910127613.0,
      "step": 28517
    },
    {
      "entropy": 0.513969898223877,
      "epoch": 2.6238031983714873,
      "grad_norm": 0.16042852401733398,
      "learning_rate": 1.2543318919250468e-05,
      "loss": 0.5011,
      "mean_token_accuracy": 0.8431829139590263,
      "num_tokens": 910159491.0,
      "step": 28518
    },
    {
      "entropy": 0.55396209936589,
      "epoch": 2.6238952047429285,
      "grad_norm": 0.16199259459972382,
      "learning_rate": 1.2540252093108843e-05,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8288876600563526,
      "num_tokens": 910191339.0,
      "step": 28519
    },
    {
      "entropy": 0.5131684388034046,
      "epoch": 2.6239872111143696,
      "grad_norm": 0.15849533677101135,
      "learning_rate": 1.2537185266967216e-05,
      "loss": 0.5021,
      "mean_token_accuracy": 0.8414049409329891,
      "num_tokens": 910223045.0,
      "step": 28520
    },
    {
      "entropy": 0.5176697261631489,
      "epoch": 2.6240792174858107,
      "grad_norm": 0.15737298130989075,
      "learning_rate": 1.253411844082559e-05,
      "loss": 0.5059,
      "mean_token_accuracy": 0.8413655571639538,
      "num_tokens": 910255436.0,
      "step": 28521
    },
    {
      "entropy": 0.6155740115791559,
      "epoch": 2.6241712238572523,
      "grad_norm": 0.17158713936805725,
      "learning_rate": 1.2531051614683963e-05,
      "loss": 0.616,
      "mean_token_accuracy": 0.8070162795484066,
      "num_tokens": 910287653.0,
      "step": 28522
    },
    {
      "entropy": 0.5141018580179662,
      "epoch": 2.6242632302286935,
      "grad_norm": 0.162644162774086,
      "learning_rate": 1.2527984788542338e-05,
      "loss": 0.5006,
      "mean_token_accuracy": 0.842426173388958,
      "num_tokens": 910319542.0,
      "step": 28523
    },
    {
      "entropy": 0.567445233464241,
      "epoch": 2.6243552366001346,
      "grad_norm": 0.164028599858284,
      "learning_rate": 1.2524917962400713e-05,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8246996849775314,
      "num_tokens": 910351745.0,
      "step": 28524
    },
    {
      "entropy": 0.496801502071321,
      "epoch": 2.6244472429715757,
      "grad_norm": 0.15600475668907166,
      "learning_rate": 1.2521851136259085e-05,
      "loss": 0.4873,
      "mean_token_accuracy": 0.844851978123188,
      "num_tokens": 910383767.0,
      "step": 28525
    },
    {
      "entropy": 0.5084137581288815,
      "epoch": 2.624539249343017,
      "grad_norm": 0.16170063614845276,
      "learning_rate": 1.251878431011746e-05,
      "loss": 0.4982,
      "mean_token_accuracy": 0.8408761322498322,
      "num_tokens": 910414823.0,
      "step": 28526
    },
    {
      "entropy": 0.5111992815509439,
      "epoch": 2.6246312557144584,
      "grad_norm": 0.16149649024009705,
      "learning_rate": 1.2515717483975833e-05,
      "loss": 0.501,
      "mean_token_accuracy": 0.8425846174359322,
      "num_tokens": 910447132.0,
      "step": 28527
    },
    {
      "entropy": 0.5480661736801267,
      "epoch": 2.6247232620858996,
      "grad_norm": 0.1616198569536209,
      "learning_rate": 1.2512650657834208e-05,
      "loss": 0.525,
      "mean_token_accuracy": 0.8291350081562996,
      "num_tokens": 910479517.0,
      "step": 28528
    },
    {
      "entropy": 0.5186334270983934,
      "epoch": 2.6248152684573407,
      "grad_norm": 0.15554706752300262,
      "learning_rate": 1.250958383169258e-05,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8424593545496464,
      "num_tokens": 910511275.0,
      "step": 28529
    },
    {
      "entropy": 0.5764536680653691,
      "epoch": 2.624907274828782,
      "grad_norm": 0.16977976262569427,
      "learning_rate": 1.2506517005550955e-05,
      "loss": 0.5649,
      "mean_token_accuracy": 0.8217010237276554,
      "num_tokens": 910543549.0,
      "step": 28530
    },
    {
      "entropy": 0.5186513913795352,
      "epoch": 2.624999281200223,
      "grad_norm": 0.1605854332447052,
      "learning_rate": 1.2503450179409328e-05,
      "loss": 0.5066,
      "mean_token_accuracy": 0.8414211869239807,
      "num_tokens": 910574898.0,
      "step": 28531
    },
    {
      "entropy": 0.5553284180350602,
      "epoch": 2.6250912875716645,
      "grad_norm": 0.16237714886665344,
      "learning_rate": 1.2500383353267703e-05,
      "loss": 0.5418,
      "mean_token_accuracy": 0.8299385234713554,
      "num_tokens": 910607450.0,
      "step": 28532
    },
    {
      "entropy": 0.4845200404524803,
      "epoch": 2.6251832939431057,
      "grad_norm": 0.15616071224212646,
      "learning_rate": 1.2497316527126078e-05,
      "loss": 0.4808,
      "mean_token_accuracy": 0.8481142073869705,
      "num_tokens": 910639326.0,
      "step": 28533
    },
    {
      "entropy": 0.5600051619112492,
      "epoch": 2.625275300314547,
      "grad_norm": 0.16279670596122742,
      "learning_rate": 1.2494249700984452e-05,
      "loss": 0.5479,
      "mean_token_accuracy": 0.8242406696081161,
      "num_tokens": 910671653.0,
      "step": 28534
    },
    {
      "entropy": 0.4894434022717178,
      "epoch": 2.625367306685988,
      "grad_norm": 0.156472846865654,
      "learning_rate": 1.2491182874842825e-05,
      "loss": 0.4813,
      "mean_token_accuracy": 0.8480213098227978,
      "num_tokens": 910704289.0,
      "step": 28535
    },
    {
      "entropy": 0.537836816161871,
      "epoch": 2.625459313057429,
      "grad_norm": 0.16133864223957062,
      "learning_rate": 1.24881160487012e-05,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8319959416985512,
      "num_tokens": 910736100.0,
      "step": 28536
    },
    {
      "entropy": 0.5233118981122971,
      "epoch": 2.6255513194288707,
      "grad_norm": 0.16240036487579346,
      "learning_rate": 1.2485049222559573e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8362764231860638,
      "num_tokens": 910768429.0,
      "step": 28537
    },
    {
      "entropy": 0.4872579025104642,
      "epoch": 2.625643325800312,
      "grad_norm": 0.15521834790706635,
      "learning_rate": 1.2481982396417947e-05,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8483130484819412,
      "num_tokens": 910800648.0,
      "step": 28538
    },
    {
      "entropy": 0.4909574221819639,
      "epoch": 2.625735332171753,
      "grad_norm": 0.1665601134300232,
      "learning_rate": 1.2478915570276322e-05,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8451548032462597,
      "num_tokens": 910831711.0,
      "step": 28539
    },
    {
      "entropy": 0.45170272258110344,
      "epoch": 2.625827338543194,
      "grad_norm": 0.150875523686409,
      "learning_rate": 1.2475848744134695e-05,
      "loss": 0.4407,
      "mean_token_accuracy": 0.8609897829592228,
      "num_tokens": 910863821.0,
      "step": 28540
    },
    {
      "entropy": 0.5403865985572338,
      "epoch": 2.625919344914635,
      "grad_norm": 0.15916447341442108,
      "learning_rate": 1.247278191799307e-05,
      "loss": 0.5337,
      "mean_token_accuracy": 0.8330178260803223,
      "num_tokens": 910896589.0,
      "step": 28541
    },
    {
      "entropy": 0.5980436131358147,
      "epoch": 2.6260113512860768,
      "grad_norm": 0.16868269443511963,
      "learning_rate": 1.2469715091851443e-05,
      "loss": 0.5871,
      "mean_token_accuracy": 0.8150036633014679,
      "num_tokens": 910929357.0,
      "step": 28542
    },
    {
      "entropy": 0.5129138827323914,
      "epoch": 2.626103357657518,
      "grad_norm": 0.1567656546831131,
      "learning_rate": 1.2466648265709817e-05,
      "loss": 0.499,
      "mean_token_accuracy": 0.8407339565455914,
      "num_tokens": 910960886.0,
      "step": 28543
    },
    {
      "entropy": 0.5706590935587883,
      "epoch": 2.626195364028959,
      "grad_norm": 0.1746145784854889,
      "learning_rate": 1.2463581439568192e-05,
      "loss": 0.547,
      "mean_token_accuracy": 0.8289773017168045,
      "num_tokens": 910991300.0,
      "step": 28544
    },
    {
      "entropy": 0.4650898033287376,
      "epoch": 2.6262873704004,
      "grad_norm": 0.15345853567123413,
      "learning_rate": 1.2460514613426567e-05,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8553797975182533,
      "num_tokens": 911023669.0,
      "step": 28545
    },
    {
      "entropy": 0.4541083090007305,
      "epoch": 2.6263793767718413,
      "grad_norm": 0.14705635607242584,
      "learning_rate": 1.245744778728494e-05,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8600986450910568,
      "num_tokens": 911055451.0,
      "step": 28546
    },
    {
      "entropy": 0.4954109200043604,
      "epoch": 2.626471383143283,
      "grad_norm": 0.1577785164117813,
      "learning_rate": 1.2454380961143314e-05,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8491893336176872,
      "num_tokens": 911087471.0,
      "step": 28547
    },
    {
      "entropy": 0.5003651399165392,
      "epoch": 2.626563389514724,
      "grad_norm": 0.1613709032535553,
      "learning_rate": 1.2451314135001687e-05,
      "loss": 0.4944,
      "mean_token_accuracy": 0.8443612642586231,
      "num_tokens": 911119660.0,
      "step": 28548
    },
    {
      "entropy": 0.5053484085947275,
      "epoch": 2.626655395886165,
      "grad_norm": 0.15981048345565796,
      "learning_rate": 1.2448247308860062e-05,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8455283716320992,
      "num_tokens": 911151447.0,
      "step": 28549
    },
    {
      "entropy": 0.5855758637189865,
      "epoch": 2.6267474022576063,
      "grad_norm": 0.16911378502845764,
      "learning_rate": 1.2445180482718435e-05,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8184811063110828,
      "num_tokens": 911183502.0,
      "step": 28550
    },
    {
      "entropy": 0.4718944523483515,
      "epoch": 2.6268394086290474,
      "grad_norm": 0.148982971906662,
      "learning_rate": 1.244211365657681e-05,
      "loss": 0.4579,
      "mean_token_accuracy": 0.8527277447283268,
      "num_tokens": 911215307.0,
      "step": 28551
    },
    {
      "entropy": 0.5139386653900146,
      "epoch": 2.626931415000489,
      "grad_norm": 0.16033627092838287,
      "learning_rate": 1.2439046830435182e-05,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8422981798648834,
      "num_tokens": 911247303.0,
      "step": 28552
    },
    {
      "entropy": 0.6352408118546009,
      "epoch": 2.62702342137193,
      "grad_norm": 0.17630094289779663,
      "learning_rate": 1.2435980004293557e-05,
      "loss": 0.6265,
      "mean_token_accuracy": 0.8024583086371422,
      "num_tokens": 911279411.0,
      "step": 28553
    },
    {
      "entropy": 0.5294088255614042,
      "epoch": 2.6271154277433713,
      "grad_norm": 0.15857642889022827,
      "learning_rate": 1.2432913178151932e-05,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8348797932267189,
      "num_tokens": 911311985.0,
      "step": 28554
    },
    {
      "entropy": 0.4637564830482006,
      "epoch": 2.6272074341148124,
      "grad_norm": 0.15329892933368683,
      "learning_rate": 1.2429846352010305e-05,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8558578342199326,
      "num_tokens": 911344420.0,
      "step": 28555
    },
    {
      "entropy": 0.6116641452535987,
      "epoch": 2.6272994404862535,
      "grad_norm": 0.17414043843746185,
      "learning_rate": 1.242677952586868e-05,
      "loss": 0.6037,
      "mean_token_accuracy": 0.8109176903963089,
      "num_tokens": 911375303.0,
      "step": 28556
    },
    {
      "entropy": 0.6246916577219963,
      "epoch": 2.627391446857695,
      "grad_norm": 0.1716972142457962,
      "learning_rate": 1.2423712699727052e-05,
      "loss": 0.6069,
      "mean_token_accuracy": 0.8083512149751186,
      "num_tokens": 911407589.0,
      "step": 28557
    },
    {
      "entropy": 0.5644073477014899,
      "epoch": 2.6274834532291362,
      "grad_norm": 0.16585354506969452,
      "learning_rate": 1.2420645873585427e-05,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8239283710718155,
      "num_tokens": 911439429.0,
      "step": 28558
    },
    {
      "entropy": 0.4956664964556694,
      "epoch": 2.6275754596005774,
      "grad_norm": 0.16180679202079773,
      "learning_rate": 1.2417579047443801e-05,
      "loss": 0.4905,
      "mean_token_accuracy": 0.8459570780396461,
      "num_tokens": 911471470.0,
      "step": 28559
    },
    {
      "entropy": 0.43547868449240923,
      "epoch": 2.6276674659720185,
      "grad_norm": 0.15112583339214325,
      "learning_rate": 1.2414512221302176e-05,
      "loss": 0.4214,
      "mean_token_accuracy": 0.8664674200117588,
      "num_tokens": 911503271.0,
      "step": 28560
    },
    {
      "entropy": 0.5239966185763478,
      "epoch": 2.6277594723434596,
      "grad_norm": 0.15615028142929077,
      "learning_rate": 1.2411445395160549e-05,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8412287533283234,
      "num_tokens": 911535497.0,
      "step": 28561
    },
    {
      "entropy": 0.47705564415082335,
      "epoch": 2.627851478714901,
      "grad_norm": 0.15544342994689941,
      "learning_rate": 1.2408378569018924e-05,
      "loss": 0.4758,
      "mean_token_accuracy": 0.8476710170507431,
      "num_tokens": 911567715.0,
      "step": 28562
    },
    {
      "entropy": 0.4100033761933446,
      "epoch": 2.6279434850863423,
      "grad_norm": 0.14335478842258453,
      "learning_rate": 1.2405311742877297e-05,
      "loss": 0.4071,
      "mean_token_accuracy": 0.8694578483700752,
      "num_tokens": 911600168.0,
      "step": 28563
    },
    {
      "entropy": 0.4856917345896363,
      "epoch": 2.6280354914577835,
      "grad_norm": 0.1560276597738266,
      "learning_rate": 1.2402244916735671e-05,
      "loss": 0.4761,
      "mean_token_accuracy": 0.8456128053367138,
      "num_tokens": 911631947.0,
      "step": 28564
    },
    {
      "entropy": 0.47189877461642027,
      "epoch": 2.6281274978292246,
      "grad_norm": 0.15571415424346924,
      "learning_rate": 1.2399178090594044e-05,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8497630022466183,
      "num_tokens": 911664076.0,
      "step": 28565
    },
    {
      "entropy": 0.5563971027731895,
      "epoch": 2.6282195042006657,
      "grad_norm": 0.1650611311197281,
      "learning_rate": 1.2396111264452419e-05,
      "loss": 0.5491,
      "mean_token_accuracy": 0.8276997767388821,
      "num_tokens": 911695103.0,
      "step": 28566
    },
    {
      "entropy": 0.5116386252921075,
      "epoch": 2.6283115105721073,
      "grad_norm": 0.1618286818265915,
      "learning_rate": 1.2393044438310792e-05,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8366227559745312,
      "num_tokens": 911727102.0,
      "step": 28567
    },
    {
      "entropy": 0.4910493600182235,
      "epoch": 2.6284035169435485,
      "grad_norm": 0.1514262557029724,
      "learning_rate": 1.2389977612169167e-05,
      "loss": 0.4737,
      "mean_token_accuracy": 0.8486412614583969,
      "num_tokens": 911759106.0,
      "step": 28568
    },
    {
      "entropy": 0.4091968354769051,
      "epoch": 2.6284955233149896,
      "grad_norm": 0.14538267254829407,
      "learning_rate": 1.2386910786027541e-05,
      "loss": 0.3944,
      "mean_token_accuracy": 0.8718268647789955,
      "num_tokens": 911791312.0,
      "step": 28569
    },
    {
      "entropy": 0.536121972836554,
      "epoch": 2.6285875296864307,
      "grad_norm": 0.16432958841323853,
      "learning_rate": 1.2383843959885914e-05,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8351909294724464,
      "num_tokens": 911823652.0,
      "step": 28570
    },
    {
      "entropy": 0.5049331216141582,
      "epoch": 2.628679536057872,
      "grad_norm": 0.15896442532539368,
      "learning_rate": 1.2380777133744289e-05,
      "loss": 0.4906,
      "mean_token_accuracy": 0.8424909710884094,
      "num_tokens": 911855600.0,
      "step": 28571
    },
    {
      "entropy": 0.4787717889994383,
      "epoch": 2.6287715424293134,
      "grad_norm": 0.1521003246307373,
      "learning_rate": 1.2377710307602662e-05,
      "loss": 0.466,
      "mean_token_accuracy": 0.8520636148750782,
      "num_tokens": 911887406.0,
      "step": 28572
    },
    {
      "entropy": 0.46929718274623156,
      "epoch": 2.6288635488007546,
      "grad_norm": 0.15671516954898834,
      "learning_rate": 1.2374643481461036e-05,
      "loss": 0.4608,
      "mean_token_accuracy": 0.853322971612215,
      "num_tokens": 911919739.0,
      "step": 28573
    },
    {
      "entropy": 0.5110970617970452,
      "epoch": 2.6289555551721957,
      "grad_norm": 0.16353271901607513,
      "learning_rate": 1.237157665531941e-05,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8383074067533016,
      "num_tokens": 911951437.0,
      "step": 28574
    },
    {
      "entropy": 0.4601063234731555,
      "epoch": 2.629047561543637,
      "grad_norm": 0.14891265332698822,
      "learning_rate": 1.2368509829177786e-05,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8587493225932121,
      "num_tokens": 911984135.0,
      "step": 28575
    },
    {
      "entropy": 0.5104622785001993,
      "epoch": 2.629139567915078,
      "grad_norm": 0.15701836347579956,
      "learning_rate": 1.2365443003036159e-05,
      "loss": 0.4928,
      "mean_token_accuracy": 0.8459290750324726,
      "num_tokens": 912015756.0,
      "step": 28576
    },
    {
      "entropy": 0.4538408201187849,
      "epoch": 2.6292315742865195,
      "grad_norm": 0.16003666818141937,
      "learning_rate": 1.2362376176894533e-05,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8580581583082676,
      "num_tokens": 912047021.0,
      "step": 28577
    },
    {
      "entropy": 0.45707631204277277,
      "epoch": 2.6293235806579607,
      "grad_norm": 0.15060614049434662,
      "learning_rate": 1.2359309350752906e-05,
      "loss": 0.4382,
      "mean_token_accuracy": 0.8595397509634495,
      "num_tokens": 912078773.0,
      "step": 28578
    },
    {
      "entropy": 0.4524241331964731,
      "epoch": 2.629415587029402,
      "grad_norm": 0.15686911344528198,
      "learning_rate": 1.2356242524611281e-05,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8569027073681355,
      "num_tokens": 912109482.0,
      "step": 28579
    },
    {
      "entropy": 0.4540207046084106,
      "epoch": 2.629507593400843,
      "grad_norm": 0.14522230625152588,
      "learning_rate": 1.2353175698469654e-05,
      "loss": 0.4434,
      "mean_token_accuracy": 0.8584813103079796,
      "num_tokens": 912141351.0,
      "step": 28580
    },
    {
      "entropy": 0.6053234441205859,
      "epoch": 2.629599599772284,
      "grad_norm": 0.17207637429237366,
      "learning_rate": 1.2350108872328028e-05,
      "loss": 0.6021,
      "mean_token_accuracy": 0.8123660422861576,
      "num_tokens": 912173307.0,
      "step": 28581
    },
    {
      "entropy": 0.5751323889708146,
      "epoch": 2.6296916061437257,
      "grad_norm": 0.16826951503753662,
      "learning_rate": 1.2347042046186401e-05,
      "loss": 0.573,
      "mean_token_accuracy": 0.8216591142117977,
      "num_tokens": 912205083.0,
      "step": 28582
    },
    {
      "entropy": 0.5375095382332802,
      "epoch": 2.629783612515167,
      "grad_norm": 0.15906129777431488,
      "learning_rate": 1.2343975220044776e-05,
      "loss": 0.5242,
      "mean_token_accuracy": 0.8377721160650253,
      "num_tokens": 912237564.0,
      "step": 28583
    },
    {
      "entropy": 0.4391329335048795,
      "epoch": 2.629875618886608,
      "grad_norm": 0.15186254680156708,
      "learning_rate": 1.234090839390315e-05,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8609147556126118,
      "num_tokens": 912270109.0,
      "step": 28584
    },
    {
      "entropy": 0.4653241718187928,
      "epoch": 2.629967625258049,
      "grad_norm": 0.1556038111448288,
      "learning_rate": 1.2337841567761524e-05,
      "loss": 0.4481,
      "mean_token_accuracy": 0.8578742146492004,
      "num_tokens": 912301791.0,
      "step": 28585
    },
    {
      "entropy": 0.4352906006388366,
      "epoch": 2.63005963162949,
      "grad_norm": 0.14776505529880524,
      "learning_rate": 1.2334774741619898e-05,
      "loss": 0.4233,
      "mean_token_accuracy": 0.8656348437070847,
      "num_tokens": 912334213.0,
      "step": 28586
    },
    {
      "entropy": 0.46918365452438593,
      "epoch": 2.6301516380009318,
      "grad_norm": 0.15779194235801697,
      "learning_rate": 1.2331707915478271e-05,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8526128940284252,
      "num_tokens": 912366305.0,
      "step": 28587
    },
    {
      "entropy": 0.4603905864059925,
      "epoch": 2.630243644372373,
      "grad_norm": 0.15165023505687714,
      "learning_rate": 1.2328641089336646e-05,
      "loss": 0.4555,
      "mean_token_accuracy": 0.8555167131125927,
      "num_tokens": 912398763.0,
      "step": 28588
    },
    {
      "entropy": 0.4277596287429333,
      "epoch": 2.630335650743814,
      "grad_norm": 0.14655253291130066,
      "learning_rate": 1.2325574263195019e-05,
      "loss": 0.4245,
      "mean_token_accuracy": 0.862170085310936,
      "num_tokens": 912430958.0,
      "step": 28589
    },
    {
      "entropy": 0.4322954537346959,
      "epoch": 2.630427657115255,
      "grad_norm": 0.15128499269485474,
      "learning_rate": 1.2322507437053395e-05,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8687357343733311,
      "num_tokens": 912461954.0,
      "step": 28590
    },
    {
      "entropy": 0.506429186090827,
      "epoch": 2.6305196634866963,
      "grad_norm": 0.1584501713514328,
      "learning_rate": 1.2319440610911768e-05,
      "loss": 0.4994,
      "mean_token_accuracy": 0.8392626009881496,
      "num_tokens": 912494488.0,
      "step": 28591
    },
    {
      "entropy": 0.5021169641986489,
      "epoch": 2.630611669858138,
      "grad_norm": 0.16065984964370728,
      "learning_rate": 1.2316373784770143e-05,
      "loss": 0.4876,
      "mean_token_accuracy": 0.842839241027832,
      "num_tokens": 912526171.0,
      "step": 28592
    },
    {
      "entropy": 0.47020148672163486,
      "epoch": 2.630703676229579,
      "grad_norm": 0.15363813936710358,
      "learning_rate": 1.2313306958628516e-05,
      "loss": 0.45,
      "mean_token_accuracy": 0.8577128164470196,
      "num_tokens": 912558334.0,
      "step": 28593
    },
    {
      "entropy": 0.5284405751153827,
      "epoch": 2.63079568260102,
      "grad_norm": 0.16013756394386292,
      "learning_rate": 1.231024013248689e-05,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8419649079442024,
      "num_tokens": 912590481.0,
      "step": 28594
    },
    {
      "entropy": 0.5410975497215986,
      "epoch": 2.6308876889724613,
      "grad_norm": 0.1606651097536087,
      "learning_rate": 1.2307173306345263e-05,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8337205611169338,
      "num_tokens": 912622524.0,
      "step": 28595
    },
    {
      "entropy": 0.5391063410788774,
      "epoch": 2.6309796953439024,
      "grad_norm": 0.16173085570335388,
      "learning_rate": 1.2304106480203638e-05,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8355992026627064,
      "num_tokens": 912654422.0,
      "step": 28596
    },
    {
      "entropy": 0.4827126571908593,
      "epoch": 2.631071701715344,
      "grad_norm": 0.1509770005941391,
      "learning_rate": 1.2301039654062011e-05,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8522718250751495,
      "num_tokens": 912686629.0,
      "step": 28597
    },
    {
      "entropy": 0.49157942552119493,
      "epoch": 2.631163708086785,
      "grad_norm": 0.1556216925382614,
      "learning_rate": 1.2297972827920386e-05,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8502452336251736,
      "num_tokens": 912718270.0,
      "step": 28598
    },
    {
      "entropy": 0.508763862773776,
      "epoch": 2.6312557144582263,
      "grad_norm": 0.15805764496326447,
      "learning_rate": 1.229490600177876e-05,
      "loss": 0.4915,
      "mean_token_accuracy": 0.8410438746213913,
      "num_tokens": 912749785.0,
      "step": 28599
    },
    {
      "entropy": 0.5045461116824299,
      "epoch": 2.6313477208296674,
      "grad_norm": 0.15511780977249146,
      "learning_rate": 1.2291839175637133e-05,
      "loss": 0.4896,
      "mean_token_accuracy": 0.8415942080318928,
      "num_tokens": 912781453.0,
      "step": 28600
    },
    {
      "entropy": 0.4521800703369081,
      "epoch": 2.6314397272011085,
      "grad_norm": 0.14504878222942352,
      "learning_rate": 1.2288772349495508e-05,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8612515442073345,
      "num_tokens": 912813644.0,
      "step": 28601
    },
    {
      "entropy": 0.5460655130445957,
      "epoch": 2.63153173357255,
      "grad_norm": 0.16475966572761536,
      "learning_rate": 1.228570552335388e-05,
      "loss": 0.5333,
      "mean_token_accuracy": 0.8328141942620277,
      "num_tokens": 912845158.0,
      "step": 28602
    },
    {
      "entropy": 0.47386425780132413,
      "epoch": 2.6316237399439912,
      "grad_norm": 0.1569187343120575,
      "learning_rate": 1.2282638697212255e-05,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8532017692923546,
      "num_tokens": 912877349.0,
      "step": 28603
    },
    {
      "entropy": 0.5145244905725121,
      "epoch": 2.6317157463154324,
      "grad_norm": 0.16082414984703064,
      "learning_rate": 1.2279571871070628e-05,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8401907719671726,
      "num_tokens": 912909163.0,
      "step": 28604
    },
    {
      "entropy": 0.5160381626337767,
      "epoch": 2.6318077526868735,
      "grad_norm": 0.15804605185985565,
      "learning_rate": 1.2276505044929005e-05,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8372455462813377,
      "num_tokens": 912941329.0,
      "step": 28605
    },
    {
      "entropy": 0.5272350406739861,
      "epoch": 2.6318997590583146,
      "grad_norm": 0.1592501401901245,
      "learning_rate": 1.2273438218787378e-05,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8355843611061573,
      "num_tokens": 912972878.0,
      "step": 28606
    },
    {
      "entropy": 0.4913364630192518,
      "epoch": 2.631991765429756,
      "grad_norm": 0.15935121476650238,
      "learning_rate": 1.2270371392645752e-05,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8486678525805473,
      "num_tokens": 913004833.0,
      "step": 28607
    },
    {
      "entropy": 0.5692519368603826,
      "epoch": 2.6320837718011973,
      "grad_norm": 0.16408900916576385,
      "learning_rate": 1.2267304566504125e-05,
      "loss": 0.5638,
      "mean_token_accuracy": 0.8215127810835838,
      "num_tokens": 913036607.0,
      "step": 28608
    },
    {
      "entropy": 0.5902875307947397,
      "epoch": 2.6321757781726385,
      "grad_norm": 0.1621289700269699,
      "learning_rate": 1.22642377403625e-05,
      "loss": 0.5837,
      "mean_token_accuracy": 0.813748937100172,
      "num_tokens": 913068997.0,
      "step": 28609
    },
    {
      "entropy": 0.5056842090561986,
      "epoch": 2.6322677845440796,
      "grad_norm": 0.16847075521945953,
      "learning_rate": 1.2261170914220873e-05,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8378872722387314,
      "num_tokens": 913100845.0,
      "step": 28610
    },
    {
      "entropy": 0.5966239403933287,
      "epoch": 2.6323597909155207,
      "grad_norm": 0.17228548228740692,
      "learning_rate": 1.2258104088079248e-05,
      "loss": 0.5897,
      "mean_token_accuracy": 0.8145753107964993,
      "num_tokens": 913133245.0,
      "step": 28611
    },
    {
      "entropy": 0.5339623668696731,
      "epoch": 2.6324517972869623,
      "grad_norm": 0.16009674966335297,
      "learning_rate": 1.225503726193762e-05,
      "loss": 0.5177,
      "mean_token_accuracy": 0.8321410827338696,
      "num_tokens": 913165549.0,
      "step": 28612
    },
    {
      "entropy": 0.5511091686785221,
      "epoch": 2.6325438036584035,
      "grad_norm": 0.16465221345424652,
      "learning_rate": 1.2251970435795995e-05,
      "loss": 0.5518,
      "mean_token_accuracy": 0.8275998868048191,
      "num_tokens": 913197179.0,
      "step": 28613
    },
    {
      "entropy": 0.49789247289299965,
      "epoch": 2.6326358100298446,
      "grad_norm": 0.15641701221466064,
      "learning_rate": 1.224890360965437e-05,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8454481400549412,
      "num_tokens": 913227907.0,
      "step": 28614
    },
    {
      "entropy": 0.44562257546931505,
      "epoch": 2.6327278164012857,
      "grad_norm": 0.14820128679275513,
      "learning_rate": 1.2245836783512743e-05,
      "loss": 0.4295,
      "mean_token_accuracy": 0.8596498630940914,
      "num_tokens": 913259874.0,
      "step": 28615
    },
    {
      "entropy": 0.4945051344111562,
      "epoch": 2.632819822772727,
      "grad_norm": 0.1566253900527954,
      "learning_rate": 1.2242769957371117e-05,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8493862152099609,
      "num_tokens": 913292301.0,
      "step": 28616
    },
    {
      "entropy": 0.5335224866867065,
      "epoch": 2.6329118291441684,
      "grad_norm": 0.1600417047739029,
      "learning_rate": 1.223970313122949e-05,
      "loss": 0.527,
      "mean_token_accuracy": 0.8301540240645409,
      "num_tokens": 913324430.0,
      "step": 28617
    },
    {
      "entropy": 0.6241622297093272,
      "epoch": 2.6330038355156096,
      "grad_norm": 0.16417378187179565,
      "learning_rate": 1.2236636305087865e-05,
      "loss": 0.609,
      "mean_token_accuracy": 0.8124510943889618,
      "num_tokens": 913356533.0,
      "step": 28618
    },
    {
      "entropy": 0.5263983175391331,
      "epoch": 2.6330958418870507,
      "grad_norm": 0.15708452463150024,
      "learning_rate": 1.2233569478946238e-05,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8413613699376583,
      "num_tokens": 913388910.0,
      "step": 28619
    },
    {
      "entropy": 0.4708123989403248,
      "epoch": 2.633187848258492,
      "grad_norm": 0.15308718383312225,
      "learning_rate": 1.2230502652804614e-05,
      "loss": 0.4439,
      "mean_token_accuracy": 0.8583148568868637,
      "num_tokens": 913420689.0,
      "step": 28620
    },
    {
      "entropy": 0.37921970430761576,
      "epoch": 2.633279854629933,
      "grad_norm": 0.14159007370471954,
      "learning_rate": 1.2227435826662987e-05,
      "loss": 0.3575,
      "mean_token_accuracy": 0.8825314454734325,
      "num_tokens": 913451892.0,
      "step": 28621
    },
    {
      "entropy": 0.5044188071042299,
      "epoch": 2.6333718610013745,
      "grad_norm": 0.1515602320432663,
      "learning_rate": 1.2224369000521362e-05,
      "loss": 0.48,
      "mean_token_accuracy": 0.8466701023280621,
      "num_tokens": 913484139.0,
      "step": 28622
    },
    {
      "entropy": 0.4714450119063258,
      "epoch": 2.6334638673728157,
      "grad_norm": 0.15663963556289673,
      "learning_rate": 1.2221302174379735e-05,
      "loss": 0.4669,
      "mean_token_accuracy": 0.8511637933552265,
      "num_tokens": 913516218.0,
      "step": 28623
    },
    {
      "entropy": 0.5538614522665739,
      "epoch": 2.633555873744257,
      "grad_norm": 0.16249226033687592,
      "learning_rate": 1.221823534823811e-05,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8297479040920734,
      "num_tokens": 913547912.0,
      "step": 28624
    },
    {
      "entropy": 0.48067272757180035,
      "epoch": 2.633647880115698,
      "grad_norm": 0.1525970697402954,
      "learning_rate": 1.2215168522096482e-05,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8550583273172379,
      "num_tokens": 913580341.0,
      "step": 28625
    },
    {
      "entropy": 0.5278767812997103,
      "epoch": 2.633739886487139,
      "grad_norm": 0.15337324142456055,
      "learning_rate": 1.2212101695954857e-05,
      "loss": 0.5096,
      "mean_token_accuracy": 0.841212585568428,
      "num_tokens": 913612699.0,
      "step": 28626
    },
    {
      "entropy": 0.5017067492008209,
      "epoch": 2.6338318928585807,
      "grad_norm": 0.1567663848400116,
      "learning_rate": 1.2209034869813232e-05,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8456486724317074,
      "num_tokens": 913644794.0,
      "step": 28627
    },
    {
      "entropy": 0.4580888282507658,
      "epoch": 2.633923899230022,
      "grad_norm": 0.15004228055477142,
      "learning_rate": 1.2205968043671605e-05,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8549588583409786,
      "num_tokens": 913676366.0,
      "step": 28628
    },
    {
      "entropy": 0.5716516887769103,
      "epoch": 2.634015905601463,
      "grad_norm": 0.1703519970178604,
      "learning_rate": 1.220290121752998e-05,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8260673321783543,
      "num_tokens": 913707917.0,
      "step": 28629
    },
    {
      "entropy": 0.5148067679256201,
      "epoch": 2.634107911972904,
      "grad_norm": 0.16320723295211792,
      "learning_rate": 1.2199834391388352e-05,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8348728157579899,
      "num_tokens": 913739898.0,
      "step": 28630
    },
    {
      "entropy": 0.47068816632963717,
      "epoch": 2.634199918344345,
      "grad_norm": 0.1567334234714508,
      "learning_rate": 1.2196767565246727e-05,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8509709648787975,
      "num_tokens": 913771874.0,
      "step": 28631
    },
    {
      "entropy": 0.4694254552014172,
      "epoch": 2.6342919247157868,
      "grad_norm": 0.15408442914485931,
      "learning_rate": 1.21937007391051e-05,
      "loss": 0.4596,
      "mean_token_accuracy": 0.8533555120229721,
      "num_tokens": 913803919.0,
      "step": 28632
    },
    {
      "entropy": 0.47191645484417677,
      "epoch": 2.634383931087228,
      "grad_norm": 0.15448342263698578,
      "learning_rate": 1.2190633912963475e-05,
      "loss": 0.4697,
      "mean_token_accuracy": 0.850891437381506,
      "num_tokens": 913835956.0,
      "step": 28633
    },
    {
      "entropy": 0.5111611969769001,
      "epoch": 2.634475937458669,
      "grad_norm": 0.16253215074539185,
      "learning_rate": 1.2187567086821848e-05,
      "loss": 0.5077,
      "mean_token_accuracy": 0.8437350764870644,
      "num_tokens": 913867590.0,
      "step": 28634
    },
    {
      "entropy": 0.4555645710788667,
      "epoch": 2.63456794383011,
      "grad_norm": 0.14639632403850555,
      "learning_rate": 1.2184500260680222e-05,
      "loss": 0.4472,
      "mean_token_accuracy": 0.8596943952143192,
      "num_tokens": 913899737.0,
      "step": 28635
    },
    {
      "entropy": 0.553125720936805,
      "epoch": 2.6346599502015513,
      "grad_norm": 0.16311384737491608,
      "learning_rate": 1.2181433434538597e-05,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8321746215224266,
      "num_tokens": 913932299.0,
      "step": 28636
    },
    {
      "entropy": 0.5119761545211077,
      "epoch": 2.634751956572993,
      "grad_norm": 0.15925249457359314,
      "learning_rate": 1.2178366608396971e-05,
      "loss": 0.5088,
      "mean_token_accuracy": 0.8410978838801384,
      "num_tokens": 913964092.0,
      "step": 28637
    },
    {
      "entropy": 0.5187060004100204,
      "epoch": 2.634843962944434,
      "grad_norm": 0.1561824530363083,
      "learning_rate": 1.2175299782255344e-05,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8420940116047859,
      "num_tokens": 913996130.0,
      "step": 28638
    },
    {
      "entropy": 0.488407573197037,
      "epoch": 2.634935969315875,
      "grad_norm": 0.1541203111410141,
      "learning_rate": 1.2172232956113719e-05,
      "loss": 0.4788,
      "mean_token_accuracy": 0.8481008782982826,
      "num_tokens": 914028898.0,
      "step": 28639
    },
    {
      "entropy": 0.5323407724499702,
      "epoch": 2.6350279756873163,
      "grad_norm": 0.1621667593717575,
      "learning_rate": 1.2169166129972092e-05,
      "loss": 0.527,
      "mean_token_accuracy": 0.8347547575831413,
      "num_tokens": 914060708.0,
      "step": 28640
    },
    {
      "entropy": 0.49367957888171077,
      "epoch": 2.6351199820587574,
      "grad_norm": 0.15868385136127472,
      "learning_rate": 1.2166099303830467e-05,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8440229520201683,
      "num_tokens": 914092580.0,
      "step": 28641
    },
    {
      "entropy": 0.5347346218768507,
      "epoch": 2.635211988430199,
      "grad_norm": 0.16089105606079102,
      "learning_rate": 1.2163032477688841e-05,
      "loss": 0.5252,
      "mean_token_accuracy": 0.8356204256415367,
      "num_tokens": 914124308.0,
      "step": 28642
    },
    {
      "entropy": 0.5284842997789383,
      "epoch": 2.63530399480164,
      "grad_norm": 0.1619178205728531,
      "learning_rate": 1.2159965651547214e-05,
      "loss": 0.5146,
      "mean_token_accuracy": 0.8355135470628738,
      "num_tokens": 914156427.0,
      "step": 28643
    },
    {
      "entropy": 0.4795700889080763,
      "epoch": 2.6353960011730813,
      "grad_norm": 0.1530798226594925,
      "learning_rate": 1.2156898825405589e-05,
      "loss": 0.4703,
      "mean_token_accuracy": 0.850367546081543,
      "num_tokens": 914188165.0,
      "step": 28644
    },
    {
      "entropy": 0.4731675097718835,
      "epoch": 2.6354880075445224,
      "grad_norm": 0.15164361894130707,
      "learning_rate": 1.2153831999263962e-05,
      "loss": 0.4555,
      "mean_token_accuracy": 0.8561969324946404,
      "num_tokens": 914219863.0,
      "step": 28645
    },
    {
      "entropy": 0.5643332004547119,
      "epoch": 2.6355800139159635,
      "grad_norm": 0.1664910763502121,
      "learning_rate": 1.2150765173122337e-05,
      "loss": 0.5605,
      "mean_token_accuracy": 0.8235351406037807,
      "num_tokens": 914251882.0,
      "step": 28646
    },
    {
      "entropy": 0.4297560774721205,
      "epoch": 2.635672020287405,
      "grad_norm": 0.14667996764183044,
      "learning_rate": 1.214769834698071e-05,
      "loss": 0.4189,
      "mean_token_accuracy": 0.8649828992784023,
      "num_tokens": 914283712.0,
      "step": 28647
    },
    {
      "entropy": 0.451479161856696,
      "epoch": 2.6357640266588462,
      "grad_norm": 0.14531712234020233,
      "learning_rate": 1.2144631520839084e-05,
      "loss": 0.4405,
      "mean_token_accuracy": 0.8627183549106121,
      "num_tokens": 914316408.0,
      "step": 28648
    },
    {
      "entropy": 0.382114814594388,
      "epoch": 2.6358560330302874,
      "grad_norm": 0.1451830118894577,
      "learning_rate": 1.2141564694697457e-05,
      "loss": 0.3608,
      "mean_token_accuracy": 0.8807479180395603,
      "num_tokens": 914347698.0,
      "step": 28649
    },
    {
      "entropy": 0.5457235984504223,
      "epoch": 2.6359480394017285,
      "grad_norm": 0.16355405747890472,
      "learning_rate": 1.2138497868555832e-05,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8305564858019352,
      "num_tokens": 914380255.0,
      "step": 28650
    },
    {
      "entropy": 0.49246585415676236,
      "epoch": 2.6360400457731696,
      "grad_norm": 0.154081329703331,
      "learning_rate": 1.2135431042414206e-05,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8461637757718563,
      "num_tokens": 914412008.0,
      "step": 28651
    },
    {
      "entropy": 0.4709891490638256,
      "epoch": 2.636132052144611,
      "grad_norm": 0.15678434073925018,
      "learning_rate": 1.2132364216272581e-05,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8549642637372017,
      "num_tokens": 914443978.0,
      "step": 28652
    },
    {
      "entropy": 0.45426404755562544,
      "epoch": 2.6362240585160523,
      "grad_norm": 0.15250816941261292,
      "learning_rate": 1.2129297390130954e-05,
      "loss": 0.4342,
      "mean_token_accuracy": 0.8591504134237766,
      "num_tokens": 914475215.0,
      "step": 28653
    },
    {
      "entropy": 0.5107673909515142,
      "epoch": 2.6363160648874935,
      "grad_norm": 0.16292515397071838,
      "learning_rate": 1.2126230563989329e-05,
      "loss": 0.4915,
      "mean_token_accuracy": 0.8430817127227783,
      "num_tokens": 914506334.0,
      "step": 28654
    },
    {
      "entropy": 0.5037338696420193,
      "epoch": 2.6364080712589346,
      "grad_norm": 0.16360820829868317,
      "learning_rate": 1.2123163737847702e-05,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8416365161538124,
      "num_tokens": 914537577.0,
      "step": 28655
    },
    {
      "entropy": 0.4954179518390447,
      "epoch": 2.6365000776303757,
      "grad_norm": 0.15586809813976288,
      "learning_rate": 1.2120096911706076e-05,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8475687429308891,
      "num_tokens": 914570014.0,
      "step": 28656
    },
    {
      "entropy": 0.46375011652708054,
      "epoch": 2.6365920840018173,
      "grad_norm": 0.15344689786434174,
      "learning_rate": 1.2117030085564451e-05,
      "loss": 0.4445,
      "mean_token_accuracy": 0.8597763255238533,
      "num_tokens": 914601895.0,
      "step": 28657
    },
    {
      "entropy": 0.46026141569018364,
      "epoch": 2.6366840903732585,
      "grad_norm": 0.14772282540798187,
      "learning_rate": 1.2113963259422824e-05,
      "loss": 0.4394,
      "mean_token_accuracy": 0.8604514934122562,
      "num_tokens": 914634025.0,
      "step": 28658
    },
    {
      "entropy": 0.5655161570757627,
      "epoch": 2.6367760967446996,
      "grad_norm": 0.16885288059711456,
      "learning_rate": 1.2110896433281198e-05,
      "loss": 0.5602,
      "mean_token_accuracy": 0.823468666523695,
      "num_tokens": 914665321.0,
      "step": 28659
    },
    {
      "entropy": 0.4891632830258459,
      "epoch": 2.6368681031161407,
      "grad_norm": 0.155489981174469,
      "learning_rate": 1.2107829607139571e-05,
      "loss": 0.4699,
      "mean_token_accuracy": 0.8485580794513226,
      "num_tokens": 914697123.0,
      "step": 28660
    },
    {
      "entropy": 0.5023351395502687,
      "epoch": 2.636960109487582,
      "grad_norm": 0.15603794157505035,
      "learning_rate": 1.2104762780997946e-05,
      "loss": 0.4894,
      "mean_token_accuracy": 0.842864416539669,
      "num_tokens": 914728875.0,
      "step": 28661
    },
    {
      "entropy": 0.479888089816086,
      "epoch": 2.6370521158590234,
      "grad_norm": 0.15327800810337067,
      "learning_rate": 1.2101695954856319e-05,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8523014597594738,
      "num_tokens": 914761233.0,
      "step": 28662
    },
    {
      "entropy": 0.5032297978177667,
      "epoch": 2.6371441222304646,
      "grad_norm": 0.1567544788122177,
      "learning_rate": 1.2098629128714694e-05,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8399947844445705,
      "num_tokens": 914793367.0,
      "step": 28663
    },
    {
      "entropy": 0.5482601011171937,
      "epoch": 2.6372361286019057,
      "grad_norm": 0.16465702652931213,
      "learning_rate": 1.2095562302573067e-05,
      "loss": 0.5361,
      "mean_token_accuracy": 0.824896514415741,
      "num_tokens": 914824357.0,
      "step": 28664
    },
    {
      "entropy": 0.4967560237273574,
      "epoch": 2.637328134973347,
      "grad_norm": 0.1543126106262207,
      "learning_rate": 1.2092495476431441e-05,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8473295755684376,
      "num_tokens": 914856753.0,
      "step": 28665
    },
    {
      "entropy": 0.5365856606513262,
      "epoch": 2.637420141344788,
      "grad_norm": 0.16769391298294067,
      "learning_rate": 1.2089428650289816e-05,
      "loss": 0.5243,
      "mean_token_accuracy": 0.8334989249706268,
      "num_tokens": 914888837.0,
      "step": 28666
    },
    {
      "entropy": 0.5465413338970393,
      "epoch": 2.6375121477162295,
      "grad_norm": 0.16421394050121307,
      "learning_rate": 1.208636182414819e-05,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8354062512516975,
      "num_tokens": 914920521.0,
      "step": 28667
    },
    {
      "entropy": 0.43226418690755963,
      "epoch": 2.6376041540876707,
      "grad_norm": 0.14735333621501923,
      "learning_rate": 1.2083294998006564e-05,
      "loss": 0.4138,
      "mean_token_accuracy": 0.8653730377554893,
      "num_tokens": 914952695.0,
      "step": 28668
    },
    {
      "entropy": 0.4807998398318887,
      "epoch": 2.637696160459112,
      "grad_norm": 0.156986266374588,
      "learning_rate": 1.2080228171864938e-05,
      "loss": 0.4748,
      "mean_token_accuracy": 0.8507204204797745,
      "num_tokens": 914984564.0,
      "step": 28669
    },
    {
      "entropy": 0.44357173843309283,
      "epoch": 2.637788166830553,
      "grad_norm": 0.14767543971538544,
      "learning_rate": 1.2077161345723311e-05,
      "loss": 0.4339,
      "mean_token_accuracy": 0.8601766005158424,
      "num_tokens": 915016959.0,
      "step": 28670
    },
    {
      "entropy": 0.4816328752785921,
      "epoch": 2.637880173201994,
      "grad_norm": 0.14919748902320862,
      "learning_rate": 1.2074094519581686e-05,
      "loss": 0.4789,
      "mean_token_accuracy": 0.8508860617876053,
      "num_tokens": 915049446.0,
      "step": 28671
    },
    {
      "entropy": 0.527363820001483,
      "epoch": 2.6379721795734357,
      "grad_norm": 0.162318617105484,
      "learning_rate": 1.207102769344006e-05,
      "loss": 0.5168,
      "mean_token_accuracy": 0.8380953781306744,
      "num_tokens": 915081343.0,
      "step": 28672
    },
    {
      "entropy": 0.5223362548276782,
      "epoch": 2.638064185944877,
      "grad_norm": 0.15868447721004486,
      "learning_rate": 1.2067960867298433e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8385086953639984,
      "num_tokens": 915113400.0,
      "step": 28673
    },
    {
      "entropy": 0.5521485069766641,
      "epoch": 2.638156192316318,
      "grad_norm": 0.1599341630935669,
      "learning_rate": 1.2064894041156808e-05,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8264448754489422,
      "num_tokens": 915145395.0,
      "step": 28674
    },
    {
      "entropy": 0.4374299393966794,
      "epoch": 2.638248198687759,
      "grad_norm": 0.14537492394447327,
      "learning_rate": 1.2061827215015181e-05,
      "loss": 0.4277,
      "mean_token_accuracy": 0.864080548286438,
      "num_tokens": 915177228.0,
      "step": 28675
    },
    {
      "entropy": 0.5022753523662686,
      "epoch": 2.6383402050592,
      "grad_norm": 0.1575349122285843,
      "learning_rate": 1.2058760388873556e-05,
      "loss": 0.4895,
      "mean_token_accuracy": 0.8424826078116894,
      "num_tokens": 915209419.0,
      "step": 28676
    },
    {
      "entropy": 0.47795495204627514,
      "epoch": 2.6384322114306418,
      "grad_norm": 0.15496739745140076,
      "learning_rate": 1.2055693562731929e-05,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8528564907610416,
      "num_tokens": 915240943.0,
      "step": 28677
    },
    {
      "entropy": 0.551458794856444,
      "epoch": 2.638524217802083,
      "grad_norm": 0.16731947660446167,
      "learning_rate": 1.2052626736590303e-05,
      "loss": 0.535,
      "mean_token_accuracy": 0.8293592855334282,
      "num_tokens": 915272740.0,
      "step": 28678
    },
    {
      "entropy": 0.5104059539735317,
      "epoch": 2.638616224173524,
      "grad_norm": 0.1610843986272812,
      "learning_rate": 1.2049559910448676e-05,
      "loss": 0.494,
      "mean_token_accuracy": 0.84089295566082,
      "num_tokens": 915304109.0,
      "step": 28679
    },
    {
      "entropy": 0.5430484728422016,
      "epoch": 2.638708230544965,
      "grad_norm": 0.1638769507408142,
      "learning_rate": 1.204649308430705e-05,
      "loss": 0.5381,
      "mean_token_accuracy": 0.8296103067696095,
      "num_tokens": 915336232.0,
      "step": 28680
    },
    {
      "entropy": 0.50894256378524,
      "epoch": 2.6388002369164063,
      "grad_norm": 0.1582222282886505,
      "learning_rate": 1.2043426258165425e-05,
      "loss": 0.4913,
      "mean_token_accuracy": 0.8438608720898628,
      "num_tokens": 915368147.0,
      "step": 28681
    },
    {
      "entropy": 0.5296443430706859,
      "epoch": 2.638892243287848,
      "grad_norm": 0.16124950349330902,
      "learning_rate": 1.20403594320238e-05,
      "loss": 0.526,
      "mean_token_accuracy": 0.8329767771065235,
      "num_tokens": 915399599.0,
      "step": 28682
    },
    {
      "entropy": 0.6747983023524284,
      "epoch": 2.638984249659289,
      "grad_norm": 0.17265664041042328,
      "learning_rate": 1.2037292605882173e-05,
      "loss": 0.6616,
      "mean_token_accuracy": 0.7964690960943699,
      "num_tokens": 915430918.0,
      "step": 28683
    },
    {
      "entropy": 0.4428250160999596,
      "epoch": 2.63907625603073,
      "grad_norm": 0.14719000458717346,
      "learning_rate": 1.2034225779740548e-05,
      "loss": 0.4335,
      "mean_token_accuracy": 0.8595421127974987,
      "num_tokens": 915463161.0,
      "step": 28684
    },
    {
      "entropy": 0.44934767205268145,
      "epoch": 2.6391682624021713,
      "grad_norm": 0.15042106807231903,
      "learning_rate": 1.203115895359892e-05,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8591836057603359,
      "num_tokens": 915495912.0,
      "step": 28685
    },
    {
      "entropy": 0.5410271044820547,
      "epoch": 2.6392602687736124,
      "grad_norm": 0.15928450226783752,
      "learning_rate": 1.2028092127457295e-05,
      "loss": 0.5249,
      "mean_token_accuracy": 0.8342343755066395,
      "num_tokens": 915527708.0,
      "step": 28686
    },
    {
      "entropy": 0.5400179112330079,
      "epoch": 2.639352275145054,
      "grad_norm": 0.1625330150127411,
      "learning_rate": 1.202502530131567e-05,
      "loss": 0.5254,
      "mean_token_accuracy": 0.8311465680599213,
      "num_tokens": 915560018.0,
      "step": 28687
    },
    {
      "entropy": 0.4925200887955725,
      "epoch": 2.639444281516495,
      "grad_norm": 0.15376417338848114,
      "learning_rate": 1.2021958475174043e-05,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8470969498157501,
      "num_tokens": 915592013.0,
      "step": 28688
    },
    {
      "entropy": 0.43722056597471237,
      "epoch": 2.6395362878879363,
      "grad_norm": 0.1488792896270752,
      "learning_rate": 1.2018891649032418e-05,
      "loss": 0.4255,
      "mean_token_accuracy": 0.8638069219887257,
      "num_tokens": 915623909.0,
      "step": 28689
    },
    {
      "entropy": 0.5624607861973345,
      "epoch": 2.6396282942593774,
      "grad_norm": 0.1653810441493988,
      "learning_rate": 1.201582482289079e-05,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8247704394161701,
      "num_tokens": 915655462.0,
      "step": 28690
    },
    {
      "entropy": 0.5387973133474588,
      "epoch": 2.6397203006308185,
      "grad_norm": 0.1650090217590332,
      "learning_rate": 1.2012757996749165e-05,
      "loss": 0.5358,
      "mean_token_accuracy": 0.831901665776968,
      "num_tokens": 915686800.0,
      "step": 28691
    },
    {
      "entropy": 0.5210605440661311,
      "epoch": 2.63981230700226,
      "grad_norm": 0.16561861336231232,
      "learning_rate": 1.2009691170607538e-05,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8404230624437332,
      "num_tokens": 915718807.0,
      "step": 28692
    },
    {
      "entropy": 0.5639687571674585,
      "epoch": 2.6399043133737012,
      "grad_norm": 0.16414682567119598,
      "learning_rate": 1.2006624344465913e-05,
      "loss": 0.548,
      "mean_token_accuracy": 0.8253084495663643,
      "num_tokens": 915750610.0,
      "step": 28693
    },
    {
      "entropy": 0.5216516562504694,
      "epoch": 2.6399963197451424,
      "grad_norm": 0.1623098999261856,
      "learning_rate": 1.2003557518324286e-05,
      "loss": 0.5051,
      "mean_token_accuracy": 0.838804118335247,
      "num_tokens": 915782122.0,
      "step": 28694
    },
    {
      "entropy": 0.6576777137815952,
      "epoch": 2.6400883261165835,
      "grad_norm": 0.17309968173503876,
      "learning_rate": 1.200049069218266e-05,
      "loss": 0.6444,
      "mean_token_accuracy": 0.7974741086363792,
      "num_tokens": 915814716.0,
      "step": 28695
    },
    {
      "entropy": 0.44874109560623765,
      "epoch": 2.6401803324880246,
      "grad_norm": 0.1463397890329361,
      "learning_rate": 1.1997423866041033e-05,
      "loss": 0.4329,
      "mean_token_accuracy": 0.8619155809283257,
      "num_tokens": 915846904.0,
      "step": 28696
    },
    {
      "entropy": 0.6037288028746843,
      "epoch": 2.640272338859466,
      "grad_norm": 0.16955585777759552,
      "learning_rate": 1.199435703989941e-05,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8158805891871452,
      "num_tokens": 915878268.0,
      "step": 28697
    },
    {
      "entropy": 0.450613658875227,
      "epoch": 2.6403643452309074,
      "grad_norm": 0.15663570165634155,
      "learning_rate": 1.1991290213757783e-05,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8566193245351315,
      "num_tokens": 915910072.0,
      "step": 28698
    },
    {
      "entropy": 0.551824807189405,
      "epoch": 2.6404563516023485,
      "grad_norm": 0.16353870928287506,
      "learning_rate": 1.1988223387616157e-05,
      "loss": 0.5494,
      "mean_token_accuracy": 0.8278278149664402,
      "num_tokens": 915942254.0,
      "step": 28699
    },
    {
      "entropy": 0.576007526833564,
      "epoch": 2.6405483579737896,
      "grad_norm": 0.16426634788513184,
      "learning_rate": 1.198515656147453e-05,
      "loss": 0.5625,
      "mean_token_accuracy": 0.820383083075285,
      "num_tokens": 915974688.0,
      "step": 28700
    },
    {
      "entropy": 0.47162360325455666,
      "epoch": 2.6406403643452308,
      "grad_norm": 0.15340620279312134,
      "learning_rate": 1.1982089735332905e-05,
      "loss": 0.4619,
      "mean_token_accuracy": 0.8527082726359367,
      "num_tokens": 916006818.0,
      "step": 28701
    },
    {
      "entropy": 0.5743863042443991,
      "epoch": 2.6407323707166723,
      "grad_norm": 0.1683151125907898,
      "learning_rate": 1.197902290919128e-05,
      "loss": 0.563,
      "mean_token_accuracy": 0.8196663297712803,
      "num_tokens": 916038305.0,
      "step": 28702
    },
    {
      "entropy": 0.4428181266412139,
      "epoch": 2.6408243770881135,
      "grad_norm": 0.15414251387119293,
      "learning_rate": 1.1975956083049652e-05,
      "loss": 0.4362,
      "mean_token_accuracy": 0.860013585537672,
      "num_tokens": 916070867.0,
      "step": 28703
    },
    {
      "entropy": 0.4947105944156647,
      "epoch": 2.6409163834595546,
      "grad_norm": 0.15398941934108734,
      "learning_rate": 1.1972889256908027e-05,
      "loss": 0.4788,
      "mean_token_accuracy": 0.8445515632629395,
      "num_tokens": 916102463.0,
      "step": 28704
    },
    {
      "entropy": 0.5848026536405087,
      "epoch": 2.6410083898309957,
      "grad_norm": 0.1703334003686905,
      "learning_rate": 1.19698224307664e-05,
      "loss": 0.5741,
      "mean_token_accuracy": 0.8206021711230278,
      "num_tokens": 916134576.0,
      "step": 28705
    },
    {
      "entropy": 0.5481984866783023,
      "epoch": 2.641100396202437,
      "grad_norm": 0.1631060093641281,
      "learning_rate": 1.1966755604624775e-05,
      "loss": 0.5341,
      "mean_token_accuracy": 0.8295287303626537,
      "num_tokens": 916166418.0,
      "step": 28706
    },
    {
      "entropy": 0.6077791005373001,
      "epoch": 2.6411924025738784,
      "grad_norm": 0.16836275160312653,
      "learning_rate": 1.1963688778483148e-05,
      "loss": 0.5995,
      "mean_token_accuracy": 0.8107065744698048,
      "num_tokens": 916198977.0,
      "step": 28707
    },
    {
      "entropy": 0.5903102494776249,
      "epoch": 2.6412844089453196,
      "grad_norm": 0.16368727385997772,
      "learning_rate": 1.1960621952341522e-05,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8209252990782261,
      "num_tokens": 916231316.0,
      "step": 28708
    },
    {
      "entropy": 0.5249427612870932,
      "epoch": 2.6413764153167607,
      "grad_norm": 0.16190902888774872,
      "learning_rate": 1.1957555126199895e-05,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8389367572963238,
      "num_tokens": 916263197.0,
      "step": 28709
    },
    {
      "entropy": 0.4519673166796565,
      "epoch": 2.641468421688202,
      "grad_norm": 0.15553665161132812,
      "learning_rate": 1.195448830005827e-05,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8574640490114689,
      "num_tokens": 916295086.0,
      "step": 28710
    },
    {
      "entropy": 0.5193753466010094,
      "epoch": 2.641560428059643,
      "grad_norm": 0.15434221923351288,
      "learning_rate": 1.1951421473916643e-05,
      "loss": 0.5098,
      "mean_token_accuracy": 0.8394358307123184,
      "num_tokens": 916327694.0,
      "step": 28711
    },
    {
      "entropy": 0.43044165801256895,
      "epoch": 2.6416524344310846,
      "grad_norm": 0.14685273170471191,
      "learning_rate": 1.194835464777502e-05,
      "loss": 0.4092,
      "mean_token_accuracy": 0.8655962832272053,
      "num_tokens": 916358900.0,
      "step": 28712
    },
    {
      "entropy": 0.48076781164854765,
      "epoch": 2.6417444408025257,
      "grad_norm": 0.15149952471256256,
      "learning_rate": 1.1945287821633392e-05,
      "loss": 0.4692,
      "mean_token_accuracy": 0.85091907158494,
      "num_tokens": 916391200.0,
      "step": 28713
    },
    {
      "entropy": 0.48505083564668894,
      "epoch": 2.641836447173967,
      "grad_norm": 0.15910124778747559,
      "learning_rate": 1.1942220995491767e-05,
      "loss": 0.4772,
      "mean_token_accuracy": 0.847379595041275,
      "num_tokens": 916423282.0,
      "step": 28714
    },
    {
      "entropy": 0.5461906623095274,
      "epoch": 2.641928453545408,
      "grad_norm": 0.15962575376033783,
      "learning_rate": 1.193915416935014e-05,
      "loss": 0.5333,
      "mean_token_accuracy": 0.8299951106309891,
      "num_tokens": 916456050.0,
      "step": 28715
    },
    {
      "entropy": 0.5445179091766477,
      "epoch": 2.642020459916849,
      "grad_norm": 0.16701623797416687,
      "learning_rate": 1.1936087343208514e-05,
      "loss": 0.5382,
      "mean_token_accuracy": 0.8286100886762142,
      "num_tokens": 916487723.0,
      "step": 28716
    },
    {
      "entropy": 0.5648876866325736,
      "epoch": 2.6421124662882907,
      "grad_norm": 0.16855500638484955,
      "learning_rate": 1.1933020517066889e-05,
      "loss": 0.5587,
      "mean_token_accuracy": 0.8231771662831306,
      "num_tokens": 916520101.0,
      "step": 28717
    },
    {
      "entropy": 0.5309122176840901,
      "epoch": 2.642204472659732,
      "grad_norm": 0.16155970096588135,
      "learning_rate": 1.1929953690925262e-05,
      "loss": 0.518,
      "mean_token_accuracy": 0.8358974456787109,
      "num_tokens": 916551702.0,
      "step": 28718
    },
    {
      "entropy": 0.5218608391005546,
      "epoch": 2.642296479031173,
      "grad_norm": 0.16502362489700317,
      "learning_rate": 1.1926886864783637e-05,
      "loss": 0.5083,
      "mean_token_accuracy": 0.8405493162572384,
      "num_tokens": 916582622.0,
      "step": 28719
    },
    {
      "entropy": 0.5691910209134221,
      "epoch": 2.642388485402614,
      "grad_norm": 0.16565950214862823,
      "learning_rate": 1.192382003864201e-05,
      "loss": 0.5573,
      "mean_token_accuracy": 0.823294498026371,
      "num_tokens": 916614633.0,
      "step": 28720
    },
    {
      "entropy": 0.5187561514321715,
      "epoch": 2.642480491774055,
      "grad_norm": 0.16246090829372406,
      "learning_rate": 1.1920753212500384e-05,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8400212861597538,
      "num_tokens": 916646588.0,
      "step": 28721
    },
    {
      "entropy": 0.5660877898335457,
      "epoch": 2.6425724981454968,
      "grad_norm": 0.1688360869884491,
      "learning_rate": 1.1917686386358757e-05,
      "loss": 0.5502,
      "mean_token_accuracy": 0.8233065493404865,
      "num_tokens": 916677505.0,
      "step": 28722
    },
    {
      "entropy": 0.5032457374036312,
      "epoch": 2.642664504516938,
      "grad_norm": 0.15619178116321564,
      "learning_rate": 1.1914619560217132e-05,
      "loss": 0.4866,
      "mean_token_accuracy": 0.8416066952049732,
      "num_tokens": 916709707.0,
      "step": 28723
    },
    {
      "entropy": 0.5168939866125584,
      "epoch": 2.642756510888379,
      "grad_norm": 0.15955623984336853,
      "learning_rate": 1.1911552734075505e-05,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8388096429407597,
      "num_tokens": 916742463.0,
      "step": 28724
    },
    {
      "entropy": 0.4938740571960807,
      "epoch": 2.64284851725982,
      "grad_norm": 0.1576274186372757,
      "learning_rate": 1.190848590793388e-05,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8467928841710091,
      "num_tokens": 916774424.0,
      "step": 28725
    },
    {
      "entropy": 0.5460786619223654,
      "epoch": 2.6429405236312613,
      "grad_norm": 0.16650350391864777,
      "learning_rate": 1.1905419081792252e-05,
      "loss": 0.5294,
      "mean_token_accuracy": 0.8337629921734333,
      "num_tokens": 916806708.0,
      "step": 28726
    },
    {
      "entropy": 0.450283863581717,
      "epoch": 2.643032530002703,
      "grad_norm": 0.1570683717727661,
      "learning_rate": 1.1902352255650629e-05,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8583840280771255,
      "num_tokens": 916837746.0,
      "step": 28727
    },
    {
      "entropy": 0.5423043705523014,
      "epoch": 2.643124536374144,
      "grad_norm": 0.1591174304485321,
      "learning_rate": 1.1899285429509002e-05,
      "loss": 0.5349,
      "mean_token_accuracy": 0.8327744603157043,
      "num_tokens": 916870064.0,
      "step": 28728
    },
    {
      "entropy": 0.5176257387502119,
      "epoch": 2.643216542745585,
      "grad_norm": 0.16263383626937866,
      "learning_rate": 1.1896218603367376e-05,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8391883820295334,
      "num_tokens": 916901289.0,
      "step": 28729
    },
    {
      "entropy": 0.5711814947426319,
      "epoch": 2.6433085491170263,
      "grad_norm": 0.16446271538734436,
      "learning_rate": 1.189315177722575e-05,
      "loss": 0.5567,
      "mean_token_accuracy": 0.8275195807218552,
      "num_tokens": 916933334.0,
      "step": 28730
    },
    {
      "entropy": 0.4891539169475436,
      "epoch": 2.6434005554884674,
      "grad_norm": 0.14920546114444733,
      "learning_rate": 1.1890084951084124e-05,
      "loss": 0.4719,
      "mean_token_accuracy": 0.8509667478501797,
      "num_tokens": 916965457.0,
      "step": 28731
    },
    {
      "entropy": 0.5676079783588648,
      "epoch": 2.643492561859909,
      "grad_norm": 0.16256743669509888,
      "learning_rate": 1.1887018124942499e-05,
      "loss": 0.5535,
      "mean_token_accuracy": 0.8210791647434235,
      "num_tokens": 916997258.0,
      "step": 28732
    },
    {
      "entropy": 0.5616206261329353,
      "epoch": 2.64358456823135,
      "grad_norm": 0.16387788951396942,
      "learning_rate": 1.1883951298800872e-05,
      "loss": 0.5469,
      "mean_token_accuracy": 0.8278694823384285,
      "num_tokens": 917029436.0,
      "step": 28733
    },
    {
      "entropy": 0.4360199784860015,
      "epoch": 2.6436765746027913,
      "grad_norm": 0.14802837371826172,
      "learning_rate": 1.1880884472659246e-05,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8631043173372746,
      "num_tokens": 917061638.0,
      "step": 28734
    },
    {
      "entropy": 0.5782526088878512,
      "epoch": 2.6437685809742324,
      "grad_norm": 0.1686486452817917,
      "learning_rate": 1.187781764651762e-05,
      "loss": 0.5628,
      "mean_token_accuracy": 0.8201399892568588,
      "num_tokens": 917093514.0,
      "step": 28735
    },
    {
      "entropy": 0.5347378002479672,
      "epoch": 2.6438605873456735,
      "grad_norm": 0.1572728157043457,
      "learning_rate": 1.1874750820375994e-05,
      "loss": 0.5286,
      "mean_token_accuracy": 0.8357962891459465,
      "num_tokens": 917126282.0,
      "step": 28736
    },
    {
      "entropy": 0.5432971119880676,
      "epoch": 2.643952593717115,
      "grad_norm": 0.1614292562007904,
      "learning_rate": 1.1871683994234367e-05,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8323640413582325,
      "num_tokens": 917157712.0,
      "step": 28737
    },
    {
      "entropy": 0.4599590487778187,
      "epoch": 2.6440446000885562,
      "grad_norm": 0.15011891722679138,
      "learning_rate": 1.1868617168092741e-05,
      "loss": 0.4449,
      "mean_token_accuracy": 0.8590729795396328,
      "num_tokens": 917190225.0,
      "step": 28738
    },
    {
      "entropy": 0.49229812901467085,
      "epoch": 2.6441366064599974,
      "grad_norm": 0.16212211549282074,
      "learning_rate": 1.1865550341951114e-05,
      "loss": 0.481,
      "mean_token_accuracy": 0.8481987453997135,
      "num_tokens": 917221935.0,
      "step": 28739
    },
    {
      "entropy": 0.5894629051908851,
      "epoch": 2.6442286128314385,
      "grad_norm": 0.171724334359169,
      "learning_rate": 1.1862483515809489e-05,
      "loss": 0.5879,
      "mean_token_accuracy": 0.8207655400037766,
      "num_tokens": 917253735.0,
      "step": 28740
    },
    {
      "entropy": 0.48553915321826935,
      "epoch": 2.6443206192028796,
      "grad_norm": 0.15779630839824677,
      "learning_rate": 1.1859416689667862e-05,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8496101945638657,
      "num_tokens": 917285070.0,
      "step": 28741
    },
    {
      "entropy": 0.44285490922629833,
      "epoch": 2.644412625574321,
      "grad_norm": 0.15820729732513428,
      "learning_rate": 1.1856349863526238e-05,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8607286773622036,
      "num_tokens": 917316104.0,
      "step": 28742
    },
    {
      "entropy": 0.5720550827682018,
      "epoch": 2.6445046319457624,
      "grad_norm": 0.16918738186359406,
      "learning_rate": 1.1853283037384611e-05,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8226659670472145,
      "num_tokens": 917347740.0,
      "step": 28743
    },
    {
      "entropy": 0.5650491919368505,
      "epoch": 2.6445966383172035,
      "grad_norm": 0.1689033806324005,
      "learning_rate": 1.1850216211242986e-05,
      "loss": 0.557,
      "mean_token_accuracy": 0.8217707723379135,
      "num_tokens": 917380021.0,
      "step": 28744
    },
    {
      "entropy": 0.48541940259747207,
      "epoch": 2.6446886446886446,
      "grad_norm": 0.15377512574195862,
      "learning_rate": 1.1847149385101359e-05,
      "loss": 0.4763,
      "mean_token_accuracy": 0.8485992066562176,
      "num_tokens": 917412357.0,
      "step": 28745
    },
    {
      "entropy": 0.5046316729858518,
      "epoch": 2.6447806510600858,
      "grad_norm": 0.1541416496038437,
      "learning_rate": 1.1844082558959734e-05,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8452470265328884,
      "num_tokens": 917444936.0,
      "step": 28746
    },
    {
      "entropy": 0.5592861492186785,
      "epoch": 2.6448726574315273,
      "grad_norm": 0.1636001169681549,
      "learning_rate": 1.1841015732818108e-05,
      "loss": 0.5517,
      "mean_token_accuracy": 0.8224000558257103,
      "num_tokens": 917477170.0,
      "step": 28747
    },
    {
      "entropy": 0.6077331360429525,
      "epoch": 2.6449646638029685,
      "grad_norm": 0.1659483164548874,
      "learning_rate": 1.1837948906676481e-05,
      "loss": 0.5827,
      "mean_token_accuracy": 0.8124859742820263,
      "num_tokens": 917508987.0,
      "step": 28748
    },
    {
      "entropy": 0.4933638027869165,
      "epoch": 2.6450566701744096,
      "grad_norm": 0.1599252074956894,
      "learning_rate": 1.1834882080534856e-05,
      "loss": 0.4808,
      "mean_token_accuracy": 0.8494591154158115,
      "num_tokens": 917540800.0,
      "step": 28749
    },
    {
      "entropy": 0.45821675239130855,
      "epoch": 2.6451486765458507,
      "grad_norm": 0.1484488844871521,
      "learning_rate": 1.1831815254393229e-05,
      "loss": 0.4367,
      "mean_token_accuracy": 0.8599652610719204,
      "num_tokens": 917572591.0,
      "step": 28750
    },
    {
      "entropy": 0.5651336358860135,
      "epoch": 2.645240682917292,
      "grad_norm": 0.16901636123657227,
      "learning_rate": 1.1828748428251603e-05,
      "loss": 0.5579,
      "mean_token_accuracy": 0.8241643495857716,
      "num_tokens": 917604020.0,
      "step": 28751
    },
    {
      "entropy": 0.43582000490278006,
      "epoch": 2.6453326892887334,
      "grad_norm": 0.14402498304843903,
      "learning_rate": 1.1825681602109976e-05,
      "loss": 0.4002,
      "mean_token_accuracy": 0.8705267235636711,
      "num_tokens": 917635761.0,
      "step": 28752
    },
    {
      "entropy": 0.50933325313963,
      "epoch": 2.6454246956601746,
      "grad_norm": 0.15699726343154907,
      "learning_rate": 1.1822614775968351e-05,
      "loss": 0.4995,
      "mean_token_accuracy": 0.8430102430284023,
      "num_tokens": 917667728.0,
      "step": 28753
    },
    {
      "entropy": 0.4877262907102704,
      "epoch": 2.6455167020316157,
      "grad_norm": 0.15796180069446564,
      "learning_rate": 1.1819547949826724e-05,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8485519029200077,
      "num_tokens": 917699296.0,
      "step": 28754
    },
    {
      "entropy": 0.5044262949377298,
      "epoch": 2.645608708403057,
      "grad_norm": 0.15644749999046326,
      "learning_rate": 1.1816481123685099e-05,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8484356440603733,
      "num_tokens": 917731258.0,
      "step": 28755
    },
    {
      "entropy": 0.5726930049713701,
      "epoch": 2.645700714774498,
      "grad_norm": 0.16766828298568726,
      "learning_rate": 1.1813414297543472e-05,
      "loss": 0.5545,
      "mean_token_accuracy": 0.8242572844028473,
      "num_tokens": 917762878.0,
      "step": 28756
    },
    {
      "entropy": 0.45696032606065273,
      "epoch": 2.6457927211459396,
      "grad_norm": 0.151239812374115,
      "learning_rate": 1.1810347471401846e-05,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8582161776721478,
      "num_tokens": 917794907.0,
      "step": 28757
    },
    {
      "entropy": 0.5076521569862962,
      "epoch": 2.6458847275173807,
      "grad_norm": 0.1636541187763214,
      "learning_rate": 1.180728064526022e-05,
      "loss": 0.5038,
      "mean_token_accuracy": 0.8403318300843239,
      "num_tokens": 917826431.0,
      "step": 28758
    },
    {
      "entropy": 0.5672588739544153,
      "epoch": 2.645976733888822,
      "grad_norm": 0.17682620882987976,
      "learning_rate": 1.1804213819118595e-05,
      "loss": 0.5651,
      "mean_token_accuracy": 0.8215513788163662,
      "num_tokens": 917856558.0,
      "step": 28759
    },
    {
      "entropy": 0.5077988058328629,
      "epoch": 2.646068740260263,
      "grad_norm": 0.15480971336364746,
      "learning_rate": 1.1801146992976968e-05,
      "loss": 0.4925,
      "mean_token_accuracy": 0.841511856764555,
      "num_tokens": 917888952.0,
      "step": 28760
    },
    {
      "entropy": 0.4400537293404341,
      "epoch": 2.646160746631704,
      "grad_norm": 0.14975959062576294,
      "learning_rate": 1.1798080166835343e-05,
      "loss": 0.4245,
      "mean_token_accuracy": 0.8640811778604984,
      "num_tokens": 917921142.0,
      "step": 28761
    },
    {
      "entropy": 0.48222067300230265,
      "epoch": 2.6462527530031457,
      "grad_norm": 0.15726697444915771,
      "learning_rate": 1.1795013340693718e-05,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8538305908441544,
      "num_tokens": 917953492.0,
      "step": 28762
    },
    {
      "entropy": 0.5213451911695302,
      "epoch": 2.646344759374587,
      "grad_norm": 0.15594537556171417,
      "learning_rate": 1.179194651455209e-05,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8387743793427944,
      "num_tokens": 917986059.0,
      "step": 28763
    },
    {
      "entropy": 0.5434514908120036,
      "epoch": 2.646436765746028,
      "grad_norm": 0.1633596569299698,
      "learning_rate": 1.1788879688410465e-05,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8317719399929047,
      "num_tokens": 918018322.0,
      "step": 28764
    },
    {
      "entropy": 0.5603891015052795,
      "epoch": 2.646528772117469,
      "grad_norm": 0.16518478095531464,
      "learning_rate": 1.1785812862268838e-05,
      "loss": 0.5509,
      "mean_token_accuracy": 0.8272021375596523,
      "num_tokens": 918050581.0,
      "step": 28765
    },
    {
      "entropy": 0.5149916561786085,
      "epoch": 2.64662077848891,
      "grad_norm": 0.15552841126918793,
      "learning_rate": 1.1782746036127213e-05,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8422429449856281,
      "num_tokens": 918083000.0,
      "step": 28766
    },
    {
      "entropy": 0.4137429613620043,
      "epoch": 2.6467127848603518,
      "grad_norm": 0.14596933126449585,
      "learning_rate": 1.1779679209985586e-05,
      "loss": 0.402,
      "mean_token_accuracy": 0.8716600984334946,
      "num_tokens": 918114421.0,
      "step": 28767
    },
    {
      "entropy": 0.5230800965800881,
      "epoch": 2.646804791231793,
      "grad_norm": 0.1563464254140854,
      "learning_rate": 1.177661238384396e-05,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8395993672311306,
      "num_tokens": 918146511.0,
      "step": 28768
    },
    {
      "entropy": 0.5640877950936556,
      "epoch": 2.646896797603234,
      "grad_norm": 0.1742185354232788,
      "learning_rate": 1.1773545557702333e-05,
      "loss": 0.5679,
      "mean_token_accuracy": 0.8234096616506577,
      "num_tokens": 918178082.0,
      "step": 28769
    },
    {
      "entropy": 0.5843622091924772,
      "epoch": 2.646988803974675,
      "grad_norm": 0.1697610318660736,
      "learning_rate": 1.1770478731560708e-05,
      "loss": 0.5697,
      "mean_token_accuracy": 0.8212429359555244,
      "num_tokens": 918210289.0,
      "step": 28770
    },
    {
      "entropy": 0.5266028195619583,
      "epoch": 2.6470808103461163,
      "grad_norm": 0.15974664688110352,
      "learning_rate": 1.1767411905419081e-05,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8385157659649849,
      "num_tokens": 918242023.0,
      "step": 28771
    },
    {
      "entropy": 0.5560938399285078,
      "epoch": 2.647172816717558,
      "grad_norm": 0.1614796221256256,
      "learning_rate": 1.1764345079277456e-05,
      "loss": 0.5355,
      "mean_token_accuracy": 0.8326404392719269,
      "num_tokens": 918274412.0,
      "step": 28772
    },
    {
      "entropy": 0.5337304044514894,
      "epoch": 2.647264823088999,
      "grad_norm": 0.1659715324640274,
      "learning_rate": 1.176127825313583e-05,
      "loss": 0.5234,
      "mean_token_accuracy": 0.8335300423204899,
      "num_tokens": 918305760.0,
      "step": 28773
    },
    {
      "entropy": 0.5875033475458622,
      "epoch": 2.64735682946044,
      "grad_norm": 0.17049695551395416,
      "learning_rate": 1.1758211426994205e-05,
      "loss": 0.5738,
      "mean_token_accuracy": 0.8194296173751354,
      "num_tokens": 918336804.0,
      "step": 28774
    },
    {
      "entropy": 0.5057355854660273,
      "epoch": 2.6474488358318813,
      "grad_norm": 0.1593417525291443,
      "learning_rate": 1.1755144600852578e-05,
      "loss": 0.4998,
      "mean_token_accuracy": 0.8445032238960266,
      "num_tokens": 918368488.0,
      "step": 28775
    },
    {
      "entropy": 0.43819689378142357,
      "epoch": 2.6475408422033224,
      "grad_norm": 0.14694347977638245,
      "learning_rate": 1.1752077774710953e-05,
      "loss": 0.4187,
      "mean_token_accuracy": 0.8661615774035454,
      "num_tokens": 918400441.0,
      "step": 28776
    },
    {
      "entropy": 0.461835247464478,
      "epoch": 2.647632848574764,
      "grad_norm": 0.1542579084634781,
      "learning_rate": 1.1749010948569327e-05,
      "loss": 0.4548,
      "mean_token_accuracy": 0.856860276311636,
      "num_tokens": 918432458.0,
      "step": 28777
    },
    {
      "entropy": 0.5682737324386835,
      "epoch": 2.647724854946205,
      "grad_norm": 0.17067085206508636,
      "learning_rate": 1.17459441224277e-05,
      "loss": 0.556,
      "mean_token_accuracy": 0.8235663138329983,
      "num_tokens": 918463360.0,
      "step": 28778
    },
    {
      "entropy": 0.5833244975656271,
      "epoch": 2.6478168613176463,
      "grad_norm": 0.17005330324172974,
      "learning_rate": 1.1742877296286075e-05,
      "loss": 0.5855,
      "mean_token_accuracy": 0.8155643194913864,
      "num_tokens": 918495029.0,
      "step": 28779
    },
    {
      "entropy": 0.5389425428584218,
      "epoch": 2.6479088676890874,
      "grad_norm": 0.15924480557441711,
      "learning_rate": 1.1739810470144448e-05,
      "loss": 0.5335,
      "mean_token_accuracy": 0.831562701612711,
      "num_tokens": 918527044.0,
      "step": 28780
    },
    {
      "entropy": 0.5666505191475153,
      "epoch": 2.6480008740605285,
      "grad_norm": 0.16455380618572235,
      "learning_rate": 1.1736743644002822e-05,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8241432160139084,
      "num_tokens": 918558963.0,
      "step": 28781
    },
    {
      "entropy": 0.41722196247428656,
      "epoch": 2.64809288043197,
      "grad_norm": 0.148321732878685,
      "learning_rate": 1.1733676817861195e-05,
      "loss": 0.4055,
      "mean_token_accuracy": 0.867571733891964,
      "num_tokens": 918590763.0,
      "step": 28782
    },
    {
      "entropy": 0.5554351955652237,
      "epoch": 2.6481848868034112,
      "grad_norm": 0.16216270625591278,
      "learning_rate": 1.173060999171957e-05,
      "loss": 0.551,
      "mean_token_accuracy": 0.8274241015315056,
      "num_tokens": 918622648.0,
      "step": 28783
    },
    {
      "entropy": 0.49259709008038044,
      "epoch": 2.6482768931748524,
      "grad_norm": 0.16039922833442688,
      "learning_rate": 1.1727543165577943e-05,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8430407010018826,
      "num_tokens": 918655205.0,
      "step": 28784
    },
    {
      "entropy": 0.4696511793881655,
      "epoch": 2.6483688995462935,
      "grad_norm": 0.15261246263980865,
      "learning_rate": 1.1724476339436318e-05,
      "loss": 0.4567,
      "mean_token_accuracy": 0.8511829562485218,
      "num_tokens": 918686591.0,
      "step": 28785
    },
    {
      "entropy": 0.47440471430309117,
      "epoch": 2.6484609059177346,
      "grad_norm": 0.15576942265033722,
      "learning_rate": 1.172140951329469e-05,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8534597791731358,
      "num_tokens": 918718123.0,
      "step": 28786
    },
    {
      "entropy": 0.5218381993472576,
      "epoch": 2.648552912289176,
      "grad_norm": 0.1551813781261444,
      "learning_rate": 1.1718342687153065e-05,
      "loss": 0.5126,
      "mean_token_accuracy": 0.8372680135071278,
      "num_tokens": 918750538.0,
      "step": 28787
    },
    {
      "entropy": 0.5330056045204401,
      "epoch": 2.6486449186606174,
      "grad_norm": 0.16664695739746094,
      "learning_rate": 1.171527586101144e-05,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8307797834277153,
      "num_tokens": 918781553.0,
      "step": 28788
    },
    {
      "entropy": 0.6427152808755636,
      "epoch": 2.6487369250320585,
      "grad_norm": 0.17343418300151825,
      "learning_rate": 1.1712209034869815e-05,
      "loss": 0.6273,
      "mean_token_accuracy": 0.8057025782763958,
      "num_tokens": 918813714.0,
      "step": 28789
    },
    {
      "entropy": 0.5866692755371332,
      "epoch": 2.6488289314034996,
      "grad_norm": 0.1644899696111679,
      "learning_rate": 1.1709142208728188e-05,
      "loss": 0.5741,
      "mean_token_accuracy": 0.8181932717561722,
      "num_tokens": 918845720.0,
      "step": 28790
    },
    {
      "entropy": 0.5473570069298148,
      "epoch": 2.6489209377749408,
      "grad_norm": 0.15876436233520508,
      "learning_rate": 1.1706075382586562e-05,
      "loss": 0.5323,
      "mean_token_accuracy": 0.8341119438409805,
      "num_tokens": 918877990.0,
      "step": 28791
    },
    {
      "entropy": 0.5324613144621253,
      "epoch": 2.6490129441463823,
      "grad_norm": 0.16195109486579895,
      "learning_rate": 1.1703008556444937e-05,
      "loss": 0.5261,
      "mean_token_accuracy": 0.8361369110643864,
      "num_tokens": 918910165.0,
      "step": 28792
    },
    {
      "entropy": 0.5300194807350636,
      "epoch": 2.6491049505178235,
      "grad_norm": 0.16201086342334747,
      "learning_rate": 1.169994173030331e-05,
      "loss": 0.5262,
      "mean_token_accuracy": 0.8309770785272121,
      "num_tokens": 918942435.0,
      "step": 28793
    },
    {
      "entropy": 0.5785079002380371,
      "epoch": 2.6491969568892646,
      "grad_norm": 0.1642068475484848,
      "learning_rate": 1.1696874904161684e-05,
      "loss": 0.5575,
      "mean_token_accuracy": 0.8220583759248257,
      "num_tokens": 918974250.0,
      "step": 28794
    },
    {
      "entropy": 0.4624935849569738,
      "epoch": 2.6492889632607057,
      "grad_norm": 0.1514480859041214,
      "learning_rate": 1.1693808078020057e-05,
      "loss": 0.4443,
      "mean_token_accuracy": 0.8580941371619701,
      "num_tokens": 919006255.0,
      "step": 28795
    },
    {
      "entropy": 0.6029708441346884,
      "epoch": 2.649380969632147,
      "grad_norm": 0.16807787120342255,
      "learning_rate": 1.1690741251878432e-05,
      "loss": 0.5766,
      "mean_token_accuracy": 0.8142900727689266,
      "num_tokens": 919037662.0,
      "step": 28796
    },
    {
      "entropy": 0.4673174303025007,
      "epoch": 2.6494729760035884,
      "grad_norm": 0.1571856439113617,
      "learning_rate": 1.1687674425736805e-05,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8541359826922417,
      "num_tokens": 919069870.0,
      "step": 28797
    },
    {
      "entropy": 0.48800642881542444,
      "epoch": 2.6495649823750296,
      "grad_norm": 0.15790994465351105,
      "learning_rate": 1.168460759959518e-05,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8450450412929058,
      "num_tokens": 919102618.0,
      "step": 28798
    },
    {
      "entropy": 0.6250187903642654,
      "epoch": 2.6496569887464707,
      "grad_norm": 0.16792535781860352,
      "learning_rate": 1.1681540773453553e-05,
      "loss": 0.6229,
      "mean_token_accuracy": 0.806537851691246,
      "num_tokens": 919134748.0,
      "step": 28799
    },
    {
      "entropy": 0.5083028310909867,
      "epoch": 2.649748995117912,
      "grad_norm": 0.16409993171691895,
      "learning_rate": 1.1678473947311927e-05,
      "loss": 0.4995,
      "mean_token_accuracy": 0.8409009538590908,
      "num_tokens": 919166083.0,
      "step": 28800
    },
    {
      "entropy": 0.6117040030658245,
      "epoch": 2.649841001489353,
      "grad_norm": 0.1708724945783615,
      "learning_rate": 1.16754071211703e-05,
      "loss": 0.5982,
      "mean_token_accuracy": 0.8129551745951176,
      "num_tokens": 919197757.0,
      "step": 28801
    },
    {
      "entropy": 0.5502398982644081,
      "epoch": 2.6499330078607946,
      "grad_norm": 0.16080158948898315,
      "learning_rate": 1.1672340295028675e-05,
      "loss": 0.532,
      "mean_token_accuracy": 0.8323500491678715,
      "num_tokens": 919230138.0,
      "step": 28802
    },
    {
      "entropy": 0.47903060261160135,
      "epoch": 2.6500250142322357,
      "grad_norm": 0.1572715789079666,
      "learning_rate": 1.166927346888705e-05,
      "loss": 0.4544,
      "mean_token_accuracy": 0.8538713753223419,
      "num_tokens": 919261585.0,
      "step": 28803
    },
    {
      "entropy": 0.48921376012731344,
      "epoch": 2.650117020603677,
      "grad_norm": 0.15673480927944183,
      "learning_rate": 1.1666206642745424e-05,
      "loss": 0.4818,
      "mean_token_accuracy": 0.8477040342986584,
      "num_tokens": 919294012.0,
      "step": 28804
    },
    {
      "entropy": 0.5180733636952937,
      "epoch": 2.650209026975118,
      "grad_norm": 0.15873152017593384,
      "learning_rate": 1.1663139816603797e-05,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8371925167739391,
      "num_tokens": 919325589.0,
      "step": 28805
    },
    {
      "entropy": 0.5249496595934033,
      "epoch": 2.650301033346559,
      "grad_norm": 0.15606962144374847,
      "learning_rate": 1.1660072990462172e-05,
      "loss": 0.5123,
      "mean_token_accuracy": 0.8380061499774456,
      "num_tokens": 919357728.0,
      "step": 28806
    },
    {
      "entropy": 0.6388131342828274,
      "epoch": 2.6503930397180007,
      "grad_norm": 0.17528894543647766,
      "learning_rate": 1.1657006164320546e-05,
      "loss": 0.6252,
      "mean_token_accuracy": 0.803727000951767,
      "num_tokens": 919389901.0,
      "step": 28807
    },
    {
      "entropy": 0.42379310447722673,
      "epoch": 2.650485046089442,
      "grad_norm": 0.14504177868366241,
      "learning_rate": 1.165393933817892e-05,
      "loss": 0.4083,
      "mean_token_accuracy": 0.8720941171050072,
      "num_tokens": 919421930.0,
      "step": 28808
    },
    {
      "entropy": 0.48740063048899174,
      "epoch": 2.650577052460883,
      "grad_norm": 0.15799453854560852,
      "learning_rate": 1.1650872512037294e-05,
      "loss": 0.488,
      "mean_token_accuracy": 0.8480444177985191,
      "num_tokens": 919453817.0,
      "step": 28809
    },
    {
      "entropy": 0.4098177789710462,
      "epoch": 2.650669058832324,
      "grad_norm": 0.1424012929201126,
      "learning_rate": 1.1647805685895667e-05,
      "loss": 0.4001,
      "mean_token_accuracy": 0.8725295104086399,
      "num_tokens": 919485540.0,
      "step": 28810
    },
    {
      "entropy": 0.39254640880972147,
      "epoch": 2.650761065203765,
      "grad_norm": 0.14500655233860016,
      "learning_rate": 1.1644738859754042e-05,
      "loss": 0.378,
      "mean_token_accuracy": 0.8753307349979877,
      "num_tokens": 919517335.0,
      "step": 28811
    },
    {
      "entropy": 0.548517391551286,
      "epoch": 2.6508530715752068,
      "grad_norm": 0.15955474972724915,
      "learning_rate": 1.1641672033612415e-05,
      "loss": 0.5413,
      "mean_token_accuracy": 0.831128466874361,
      "num_tokens": 919549808.0,
      "step": 28812
    },
    {
      "entropy": 0.5489461291581392,
      "epoch": 2.650945077946648,
      "grad_norm": 0.16356602311134338,
      "learning_rate": 1.1638605207470789e-05,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8268807902932167,
      "num_tokens": 919582576.0,
      "step": 28813
    },
    {
      "entropy": 0.5003098761662841,
      "epoch": 2.651037084318089,
      "grad_norm": 0.160952627658844,
      "learning_rate": 1.1635538381329162e-05,
      "loss": 0.4895,
      "mean_token_accuracy": 0.8432111367583275,
      "num_tokens": 919614098.0,
      "step": 28814
    },
    {
      "entropy": 0.4538921982748434,
      "epoch": 2.65112909068953,
      "grad_norm": 0.1483657956123352,
      "learning_rate": 1.1632471555187537e-05,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8583572767674923,
      "num_tokens": 919646352.0,
      "step": 28815
    },
    {
      "entropy": 0.483564087189734,
      "epoch": 2.6512210970609713,
      "grad_norm": 0.15381929278373718,
      "learning_rate": 1.162940472904591e-05,
      "loss": 0.4835,
      "mean_token_accuracy": 0.8496550470590591,
      "num_tokens": 919677923.0,
      "step": 28816
    },
    {
      "entropy": 0.5270320668350905,
      "epoch": 2.651313103432413,
      "grad_norm": 0.15517547726631165,
      "learning_rate": 1.1626337902904284e-05,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8417076282203197,
      "num_tokens": 919710024.0,
      "step": 28817
    },
    {
      "entropy": 0.4876444151159376,
      "epoch": 2.651405109803854,
      "grad_norm": 0.15330296754837036,
      "learning_rate": 1.1623271076762659e-05,
      "loss": 0.4741,
      "mean_token_accuracy": 0.8511433154344559,
      "num_tokens": 919742243.0,
      "step": 28818
    },
    {
      "entropy": 0.522134805098176,
      "epoch": 2.651497116175295,
      "grad_norm": 0.16429181396961212,
      "learning_rate": 1.1620204250621034e-05,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8360271789133549,
      "num_tokens": 919773367.0,
      "step": 28819
    },
    {
      "entropy": 0.5751425940543413,
      "epoch": 2.6515891225467363,
      "grad_norm": 0.17196537554264069,
      "learning_rate": 1.1617137424479407e-05,
      "loss": 0.5593,
      "mean_token_accuracy": 0.821364015340805,
      "num_tokens": 919804148.0,
      "step": 28820
    },
    {
      "entropy": 0.4761507138609886,
      "epoch": 2.6516811289181774,
      "grad_norm": 0.15233781933784485,
      "learning_rate": 1.1614070598337781e-05,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8483154848217964,
      "num_tokens": 919836403.0,
      "step": 28821
    },
    {
      "entropy": 0.4733266527764499,
      "epoch": 2.651773135289619,
      "grad_norm": 0.15399260818958282,
      "learning_rate": 1.1611003772196156e-05,
      "loss": 0.4615,
      "mean_token_accuracy": 0.8517880365252495,
      "num_tokens": 919868179.0,
      "step": 28822
    },
    {
      "entropy": 0.520445916801691,
      "epoch": 2.65186514166106,
      "grad_norm": 0.15567605197429657,
      "learning_rate": 1.1607936946054529e-05,
      "loss": 0.501,
      "mean_token_accuracy": 0.8413637727499008,
      "num_tokens": 919900411.0,
      "step": 28823
    },
    {
      "entropy": 0.4973383608739823,
      "epoch": 2.6519571480325013,
      "grad_norm": 0.16196182370185852,
      "learning_rate": 1.1604870119912903e-05,
      "loss": 0.4781,
      "mean_token_accuracy": 0.8476026505231857,
      "num_tokens": 919932459.0,
      "step": 28824
    },
    {
      "entropy": 0.5009486861526966,
      "epoch": 2.6520491544039424,
      "grad_norm": 0.16490040719509125,
      "learning_rate": 1.1601803293771276e-05,
      "loss": 0.484,
      "mean_token_accuracy": 0.8435639031231403,
      "num_tokens": 919963071.0,
      "step": 28825
    },
    {
      "entropy": 0.4292986420914531,
      "epoch": 2.6521411607753835,
      "grad_norm": 0.1510550081729889,
      "learning_rate": 1.1598736467629651e-05,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8654972203075886,
      "num_tokens": 919995159.0,
      "step": 28826
    },
    {
      "entropy": 0.5045467801392078,
      "epoch": 2.652233167146825,
      "grad_norm": 0.1620698869228363,
      "learning_rate": 1.1595669641488024e-05,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8431127406656742,
      "num_tokens": 920025517.0,
      "step": 28827
    },
    {
      "entropy": 0.40826261416077614,
      "epoch": 2.6523251735182662,
      "grad_norm": 0.14985404908657074,
      "learning_rate": 1.1592602815346399e-05,
      "loss": 0.3979,
      "mean_token_accuracy": 0.8736971691250801,
      "num_tokens": 920058183.0,
      "step": 28828
    },
    {
      "entropy": 0.48144004703499377,
      "epoch": 2.6524171798897074,
      "grad_norm": 0.15618333220481873,
      "learning_rate": 1.1589535989204772e-05,
      "loss": 0.4807,
      "mean_token_accuracy": 0.8494706340134144,
      "num_tokens": 920090133.0,
      "step": 28829
    },
    {
      "entropy": 0.5454852096736431,
      "epoch": 2.6525091862611485,
      "grad_norm": 0.16406214237213135,
      "learning_rate": 1.1586469163063146e-05,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8288193941116333,
      "num_tokens": 920122225.0,
      "step": 28830
    },
    {
      "entropy": 0.5285872649401426,
      "epoch": 2.6526011926325896,
      "grad_norm": 0.16437029838562012,
      "learning_rate": 1.158340233692152e-05,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8407761678099632,
      "num_tokens": 920153442.0,
      "step": 28831
    },
    {
      "entropy": 0.457078842446208,
      "epoch": 2.652693199004031,
      "grad_norm": 0.15769025683403015,
      "learning_rate": 1.1580335510779894e-05,
      "loss": 0.4497,
      "mean_token_accuracy": 0.8558338545262814,
      "num_tokens": 920185075.0,
      "step": 28832
    },
    {
      "entropy": 0.5321084551978856,
      "epoch": 2.6527852053754724,
      "grad_norm": 0.15660378336906433,
      "learning_rate": 1.1577268684638269e-05,
      "loss": 0.52,
      "mean_token_accuracy": 0.8344325944781303,
      "num_tokens": 920217373.0,
      "step": 28833
    },
    {
      "entropy": 0.5211480623111129,
      "epoch": 2.6528772117469135,
      "grad_norm": 0.15766727924346924,
      "learning_rate": 1.1574201858496643e-05,
      "loss": 0.5085,
      "mean_token_accuracy": 0.8375481143593788,
      "num_tokens": 920249460.0,
      "step": 28834
    },
    {
      "entropy": 0.46499805990606546,
      "epoch": 2.6529692181183546,
      "grad_norm": 0.1509428471326828,
      "learning_rate": 1.1571135032355016e-05,
      "loss": 0.4662,
      "mean_token_accuracy": 0.8557081259787083,
      "num_tokens": 920281445.0,
      "step": 28835
    },
    {
      "entropy": 0.5339729944244027,
      "epoch": 2.6530612244897958,
      "grad_norm": 0.16355639696121216,
      "learning_rate": 1.156806820621339e-05,
      "loss": 0.5211,
      "mean_token_accuracy": 0.8354386016726494,
      "num_tokens": 920313210.0,
      "step": 28836
    },
    {
      "entropy": 0.5060730203986168,
      "epoch": 2.6531532308612373,
      "grad_norm": 0.1599336564540863,
      "learning_rate": 1.1565001380071765e-05,
      "loss": 0.4984,
      "mean_token_accuracy": 0.842818059027195,
      "num_tokens": 920345455.0,
      "step": 28837
    },
    {
      "entropy": 0.5193018034333363,
      "epoch": 2.6532452372326785,
      "grad_norm": 0.15960797667503357,
      "learning_rate": 1.1561934553930138e-05,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8408625237643719,
      "num_tokens": 920376268.0,
      "step": 28838
    },
    {
      "entropy": 0.5366523768752813,
      "epoch": 2.6533372436041196,
      "grad_norm": 0.15814200043678284,
      "learning_rate": 1.1558867727788513e-05,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8361287489533424,
      "num_tokens": 920408165.0,
      "step": 28839
    },
    {
      "entropy": 0.4937323946505785,
      "epoch": 2.6534292499755607,
      "grad_norm": 0.1560867875814438,
      "learning_rate": 1.1555800901646886e-05,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8499038815498352,
      "num_tokens": 920440293.0,
      "step": 28840
    },
    {
      "entropy": 0.46452000830322504,
      "epoch": 2.653521256347002,
      "grad_norm": 0.14980503916740417,
      "learning_rate": 1.155273407550526e-05,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8583895340561867,
      "num_tokens": 920472301.0,
      "step": 28841
    },
    {
      "entropy": 0.38741540629416704,
      "epoch": 2.6536132627184434,
      "grad_norm": 0.1445481777191162,
      "learning_rate": 1.1549667249363634e-05,
      "loss": 0.3725,
      "mean_token_accuracy": 0.8811756446957588,
      "num_tokens": 920503712.0,
      "step": 28842
    },
    {
      "entropy": 0.5556134432554245,
      "epoch": 2.6537052690898846,
      "grad_norm": 0.16922315955162048,
      "learning_rate": 1.1546600423222008e-05,
      "loss": 0.543,
      "mean_token_accuracy": 0.826539546251297,
      "num_tokens": 920535377.0,
      "step": 28843
    },
    {
      "entropy": 0.5630683926865458,
      "epoch": 2.6537972754613257,
      "grad_norm": 0.16208823025226593,
      "learning_rate": 1.1543533597080381e-05,
      "loss": 0.5571,
      "mean_token_accuracy": 0.8237583935260773,
      "num_tokens": 920567646.0,
      "step": 28844
    },
    {
      "entropy": 0.45545705361291766,
      "epoch": 2.653889281832767,
      "grad_norm": 0.14963293075561523,
      "learning_rate": 1.1540466770938756e-05,
      "loss": 0.4397,
      "mean_token_accuracy": 0.860712468624115,
      "num_tokens": 920600273.0,
      "step": 28845
    },
    {
      "entropy": 0.49978402629494667,
      "epoch": 2.653981288204208,
      "grad_norm": 0.1617150604724884,
      "learning_rate": 1.1537399944797129e-05,
      "loss": 0.4869,
      "mean_token_accuracy": 0.8457084223628044,
      "num_tokens": 920631121.0,
      "step": 28846
    },
    {
      "entropy": 0.5586301786825061,
      "epoch": 2.6540732945756496,
      "grad_norm": 0.1589001566171646,
      "learning_rate": 1.1534333118655503e-05,
      "loss": 0.5388,
      "mean_token_accuracy": 0.8352234549820423,
      "num_tokens": 920662957.0,
      "step": 28847
    },
    {
      "entropy": 0.541945880278945,
      "epoch": 2.6541653009470907,
      "grad_norm": 0.16378414630889893,
      "learning_rate": 1.1531266292513878e-05,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8277011029422283,
      "num_tokens": 920694585.0,
      "step": 28848
    },
    {
      "entropy": 0.5226876251399517,
      "epoch": 2.654257307318532,
      "grad_norm": 0.16085118055343628,
      "learning_rate": 1.1528199466372253e-05,
      "loss": 0.5083,
      "mean_token_accuracy": 0.8404539227485657,
      "num_tokens": 920725256.0,
      "step": 28849
    },
    {
      "entropy": 0.6017829105257988,
      "epoch": 2.654349313689973,
      "grad_norm": 0.1672353744506836,
      "learning_rate": 1.1525132640230626e-05,
      "loss": 0.5878,
      "mean_token_accuracy": 0.8135151080787182,
      "num_tokens": 920757338.0,
      "step": 28850
    },
    {
      "entropy": 0.4616660689935088,
      "epoch": 2.654441320061414,
      "grad_norm": 0.1490236520767212,
      "learning_rate": 1.1522065814089e-05,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8589731231331825,
      "num_tokens": 920789711.0,
      "step": 28851
    },
    {
      "entropy": 0.5110806245356798,
      "epoch": 2.6545333264328557,
      "grad_norm": 0.14901046454906464,
      "learning_rate": 1.1518998987947375e-05,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8483474105596542,
      "num_tokens": 920822137.0,
      "step": 28852
    },
    {
      "entropy": 0.5296273855492473,
      "epoch": 2.654625332804297,
      "grad_norm": 0.16277827322483063,
      "learning_rate": 1.1515932161805748e-05,
      "loss": 0.5157,
      "mean_token_accuracy": 0.835829459130764,
      "num_tokens": 920853749.0,
      "step": 28853
    },
    {
      "entropy": 0.5225535118952394,
      "epoch": 2.654717339175738,
      "grad_norm": 0.16000565886497498,
      "learning_rate": 1.1512865335664123e-05,
      "loss": 0.5191,
      "mean_token_accuracy": 0.8352160751819611,
      "num_tokens": 920886056.0,
      "step": 28854
    },
    {
      "entropy": 0.5061300825327635,
      "epoch": 2.654809345547179,
      "grad_norm": 0.15158826112747192,
      "learning_rate": 1.1509798509522496e-05,
      "loss": 0.496,
      "mean_token_accuracy": 0.8445197232067585,
      "num_tokens": 920918063.0,
      "step": 28855
    },
    {
      "entropy": 0.44864368974231184,
      "epoch": 2.65490135191862,
      "grad_norm": 0.14974841475486755,
      "learning_rate": 1.150673168338087e-05,
      "loss": 0.435,
      "mean_token_accuracy": 0.8592342399060726,
      "num_tokens": 920950090.0,
      "step": 28856
    },
    {
      "entropy": 0.4751388542354107,
      "epoch": 2.6549933582900618,
      "grad_norm": 0.15677814185619354,
      "learning_rate": 1.1503664857239243e-05,
      "loss": 0.4644,
      "mean_token_accuracy": 0.8536449559032917,
      "num_tokens": 920982278.0,
      "step": 28857
    },
    {
      "entropy": 0.46455737645737827,
      "epoch": 2.655085364661503,
      "grad_norm": 0.14931508898735046,
      "learning_rate": 1.1500598031097618e-05,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8552791886031628,
      "num_tokens": 921013899.0,
      "step": 28858
    },
    {
      "entropy": 0.49210391752421856,
      "epoch": 2.655177371032944,
      "grad_norm": 0.16435809433460236,
      "learning_rate": 1.149753120495599e-05,
      "loss": 0.4829,
      "mean_token_accuracy": 0.8460629545152187,
      "num_tokens": 921045590.0,
      "step": 28859
    },
    {
      "entropy": 0.46547913015820086,
      "epoch": 2.655269377404385,
      "grad_norm": 0.15756583213806152,
      "learning_rate": 1.1494464378814365e-05,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8550718352198601,
      "num_tokens": 921077956.0,
      "step": 28860
    },
    {
      "entropy": 0.4948320398107171,
      "epoch": 2.6553613837758263,
      "grad_norm": 0.1554858535528183,
      "learning_rate": 1.1491397552672738e-05,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8448416739702225,
      "num_tokens": 921109709.0,
      "step": 28861
    },
    {
      "entropy": 0.5655826134607196,
      "epoch": 2.655453390147268,
      "grad_norm": 0.17152836918830872,
      "learning_rate": 1.1488330726531113e-05,
      "loss": 0.5597,
      "mean_token_accuracy": 0.8233070783317089,
      "num_tokens": 921141483.0,
      "step": 28862
    },
    {
      "entropy": 0.4513467224314809,
      "epoch": 2.655545396518709,
      "grad_norm": 0.14878158271312714,
      "learning_rate": 1.1485263900389488e-05,
      "loss": 0.4335,
      "mean_token_accuracy": 0.8566168546676636,
      "num_tokens": 921172679.0,
      "step": 28863
    },
    {
      "entropy": 0.5855150055140257,
      "epoch": 2.65563740289015,
      "grad_norm": 0.16604682803153992,
      "learning_rate": 1.1482197074247862e-05,
      "loss": 0.5721,
      "mean_token_accuracy": 0.8177433498203754,
      "num_tokens": 921204402.0,
      "step": 28864
    },
    {
      "entropy": 0.4818029007874429,
      "epoch": 2.6557294092615913,
      "grad_norm": 0.15875160694122314,
      "learning_rate": 1.1479130248106235e-05,
      "loss": 0.4739,
      "mean_token_accuracy": 0.848325464874506,
      "num_tokens": 921236836.0,
      "step": 28865
    },
    {
      "entropy": 0.4644493702799082,
      "epoch": 2.6558214156330324,
      "grad_norm": 0.14881344139575958,
      "learning_rate": 1.147606342196461e-05,
      "loss": 0.4476,
      "mean_token_accuracy": 0.8580477200448513,
      "num_tokens": 921269095.0,
      "step": 28866
    },
    {
      "entropy": 0.5065218202071264,
      "epoch": 2.655913422004474,
      "grad_norm": 0.15947562456130981,
      "learning_rate": 1.1472996595822985e-05,
      "loss": 0.4895,
      "mean_token_accuracy": 0.8429708480834961,
      "num_tokens": 921299978.0,
      "step": 28867
    },
    {
      "entropy": 0.6082346495240927,
      "epoch": 2.656005428375915,
      "grad_norm": 0.17224891483783722,
      "learning_rate": 1.1469929769681357e-05,
      "loss": 0.6,
      "mean_token_accuracy": 0.8080989085137844,
      "num_tokens": 921331671.0,
      "step": 28868
    },
    {
      "entropy": 0.48745962092652917,
      "epoch": 2.6560974347473563,
      "grad_norm": 0.15119503438472748,
      "learning_rate": 1.1466862943539732e-05,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8457644954323769,
      "num_tokens": 921364131.0,
      "step": 28869
    },
    {
      "entropy": 0.610381655395031,
      "epoch": 2.6561894411187974,
      "grad_norm": 0.17574764788150787,
      "learning_rate": 1.1463796117398105e-05,
      "loss": 0.597,
      "mean_token_accuracy": 0.8120545670390129,
      "num_tokens": 921395366.0,
      "step": 28870
    },
    {
      "entropy": 0.46397441904991865,
      "epoch": 2.6562814474902385,
      "grad_norm": 0.15029343962669373,
      "learning_rate": 1.146072929125648e-05,
      "loss": 0.4506,
      "mean_token_accuracy": 0.8553931452333927,
      "num_tokens": 921427699.0,
      "step": 28871
    },
    {
      "entropy": 0.5122597478330135,
      "epoch": 2.65637345386168,
      "grad_norm": 0.15980160236358643,
      "learning_rate": 1.1457662465114853e-05,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8384366892278194,
      "num_tokens": 921459777.0,
      "step": 28872
    },
    {
      "entropy": 0.5007306970655918,
      "epoch": 2.6564654602331212,
      "grad_norm": 0.1584997922182083,
      "learning_rate": 1.1454595638973227e-05,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8449828438460827,
      "num_tokens": 921491214.0,
      "step": 28873
    },
    {
      "entropy": 0.4075529705733061,
      "epoch": 2.6565574666045624,
      "grad_norm": 0.14543195068836212,
      "learning_rate": 1.14515288128316e-05,
      "loss": 0.3913,
      "mean_token_accuracy": 0.8704181052744389,
      "num_tokens": 921523624.0,
      "step": 28874
    },
    {
      "entropy": 0.5641326643526554,
      "epoch": 2.6566494729760035,
      "grad_norm": 0.16436617076396942,
      "learning_rate": 1.1448461986689975e-05,
      "loss": 0.5436,
      "mean_token_accuracy": 0.8282583057880402,
      "num_tokens": 921555408.0,
      "step": 28875
    },
    {
      "entropy": 0.5708536002784967,
      "epoch": 2.6567414793474446,
      "grad_norm": 0.1640792340040207,
      "learning_rate": 1.1445395160548348e-05,
      "loss": 0.5538,
      "mean_token_accuracy": 0.822257824242115,
      "num_tokens": 921587401.0,
      "step": 28876
    },
    {
      "entropy": 0.5666517075151205,
      "epoch": 2.6568334857188862,
      "grad_norm": 0.16982007026672363,
      "learning_rate": 1.1442328334406723e-05,
      "loss": 0.5521,
      "mean_token_accuracy": 0.8223383538424969,
      "num_tokens": 921618345.0,
      "step": 28877
    },
    {
      "entropy": 0.5700540179386735,
      "epoch": 2.6569254920903274,
      "grad_norm": 0.16704261302947998,
      "learning_rate": 1.1439261508265097e-05,
      "loss": 0.5597,
      "mean_token_accuracy": 0.822387345135212,
      "num_tokens": 921649914.0,
      "step": 28878
    },
    {
      "entropy": 0.5816276972182095,
      "epoch": 2.6570174984617685,
      "grad_norm": 0.16332949697971344,
      "learning_rate": 1.143619468212347e-05,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8216975815594196,
      "num_tokens": 921682580.0,
      "step": 28879
    },
    {
      "entropy": 0.41573788598179817,
      "epoch": 2.6571095048332096,
      "grad_norm": 0.13832968473434448,
      "learning_rate": 1.1433127855981845e-05,
      "loss": 0.4013,
      "mean_token_accuracy": 0.8709391802549362,
      "num_tokens": 921715348.0,
      "step": 28880
    },
    {
      "entropy": 0.48941375222057104,
      "epoch": 2.6572015112046508,
      "grad_norm": 0.16032738983631134,
      "learning_rate": 1.143006102984022e-05,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8448325358331203,
      "num_tokens": 921746974.0,
      "step": 28881
    },
    {
      "entropy": 0.4746137438341975,
      "epoch": 2.6572935175760923,
      "grad_norm": 0.15737198293209076,
      "learning_rate": 1.1426994203698594e-05,
      "loss": 0.4732,
      "mean_token_accuracy": 0.8505882620811462,
      "num_tokens": 921779362.0,
      "step": 28882
    },
    {
      "entropy": 0.41488434188067913,
      "epoch": 2.6573855239475335,
      "grad_norm": 0.149521142244339,
      "learning_rate": 1.1423927377556967e-05,
      "loss": 0.3907,
      "mean_token_accuracy": 0.8720037154853344,
      "num_tokens": 921810408.0,
      "step": 28883
    },
    {
      "entropy": 0.5074068661779165,
      "epoch": 2.6574775303189746,
      "grad_norm": 0.15648092329502106,
      "learning_rate": 1.1420860551415342e-05,
      "loss": 0.4883,
      "mean_token_accuracy": 0.845851369202137,
      "num_tokens": 921842187.0,
      "step": 28884
    },
    {
      "entropy": 0.5140665543731302,
      "epoch": 2.6575695366904157,
      "grad_norm": 0.15848307311534882,
      "learning_rate": 1.1417793725273715e-05,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8425961919128895,
      "num_tokens": 921874343.0,
      "step": 28885
    },
    {
      "entropy": 0.48225456196814775,
      "epoch": 2.657661543061857,
      "grad_norm": 0.15924705564975739,
      "learning_rate": 1.141472689913209e-05,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8480853922665119,
      "num_tokens": 921906281.0,
      "step": 28886
    },
    {
      "entropy": 0.5341044524684548,
      "epoch": 2.6577535494332984,
      "grad_norm": 0.1596551239490509,
      "learning_rate": 1.1411660072990462e-05,
      "loss": 0.5223,
      "mean_token_accuracy": 0.8334156572818756,
      "num_tokens": 921938704.0,
      "step": 28887
    },
    {
      "entropy": 0.579834305215627,
      "epoch": 2.6578455558047396,
      "grad_norm": 0.17013633251190186,
      "learning_rate": 1.1408593246848837e-05,
      "loss": 0.5663,
      "mean_token_accuracy": 0.8207336813211441,
      "num_tokens": 921970898.0,
      "step": 28888
    },
    {
      "entropy": 0.4263748312368989,
      "epoch": 2.6579375621761807,
      "grad_norm": 0.14461533725261688,
      "learning_rate": 1.140552642070721e-05,
      "loss": 0.412,
      "mean_token_accuracy": 0.8686658926308155,
      "num_tokens": 922003533.0,
      "step": 28889
    },
    {
      "entropy": 0.4736303063109517,
      "epoch": 2.658029568547622,
      "grad_norm": 0.15496008098125458,
      "learning_rate": 1.1402459594565584e-05,
      "loss": 0.4568,
      "mean_token_accuracy": 0.8532443232834339,
      "num_tokens": 922035487.0,
      "step": 28890
    },
    {
      "entropy": 0.4361442094668746,
      "epoch": 2.658121574919063,
      "grad_norm": 0.15027831494808197,
      "learning_rate": 1.1399392768423957e-05,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8633706718683243,
      "num_tokens": 922067369.0,
      "step": 28891
    },
    {
      "entropy": 0.5184051217511296,
      "epoch": 2.6582135812905046,
      "grad_norm": 0.15850141644477844,
      "learning_rate": 1.1396325942282332e-05,
      "loss": 0.5075,
      "mean_token_accuracy": 0.8395510278642178,
      "num_tokens": 922099345.0,
      "step": 28892
    },
    {
      "entropy": 0.6079401150345802,
      "epoch": 2.6583055876619457,
      "grad_norm": 0.17237356305122375,
      "learning_rate": 1.1393259116140707e-05,
      "loss": 0.6083,
      "mean_token_accuracy": 0.8101743087172508,
      "num_tokens": 922131261.0,
      "step": 28893
    },
    {
      "entropy": 0.5557158151641488,
      "epoch": 2.658397594033387,
      "grad_norm": 0.160530224442482,
      "learning_rate": 1.139019228999908e-05,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8289968594908714,
      "num_tokens": 922163060.0,
      "step": 28894
    },
    {
      "entropy": 0.45073823537677526,
      "epoch": 2.658489600404828,
      "grad_norm": 0.15137948095798492,
      "learning_rate": 1.1387125463857454e-05,
      "loss": 0.4426,
      "mean_token_accuracy": 0.8568217195570469,
      "num_tokens": 922195483.0,
      "step": 28895
    },
    {
      "entropy": 0.43022868363186717,
      "epoch": 2.658581606776269,
      "grad_norm": 0.14816391468048096,
      "learning_rate": 1.1384058637715829e-05,
      "loss": 0.417,
      "mean_token_accuracy": 0.8633828684687614,
      "num_tokens": 922226366.0,
      "step": 28896
    },
    {
      "entropy": 0.47373247891664505,
      "epoch": 2.6586736131477107,
      "grad_norm": 0.15252232551574707,
      "learning_rate": 1.1380991811574204e-05,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8513346500694752,
      "num_tokens": 922258106.0,
      "step": 28897
    },
    {
      "entropy": 0.5090431307908148,
      "epoch": 2.658765619519152,
      "grad_norm": 0.1535559445619583,
      "learning_rate": 1.1377924985432577e-05,
      "loss": 0.4906,
      "mean_token_accuracy": 0.8464592024683952,
      "num_tokens": 922290201.0,
      "step": 28898
    },
    {
      "entropy": 0.5336899869143963,
      "epoch": 2.658857625890593,
      "grad_norm": 0.16209274530410767,
      "learning_rate": 1.1374858159290951e-05,
      "loss": 0.5286,
      "mean_token_accuracy": 0.8336113542318344,
      "num_tokens": 922321750.0,
      "step": 28899
    },
    {
      "entropy": 0.42937108036130667,
      "epoch": 2.658949632262034,
      "grad_norm": 0.1516285091638565,
      "learning_rate": 1.1371791333149324e-05,
      "loss": 0.4224,
      "mean_token_accuracy": 0.8660693317651749,
      "num_tokens": 922353813.0,
      "step": 28900
    },
    {
      "entropy": 0.4567582723684609,
      "epoch": 2.659041638633475,
      "grad_norm": 0.15385836362838745,
      "learning_rate": 1.1368724507007699e-05,
      "loss": 0.4339,
      "mean_token_accuracy": 0.8593482598662376,
      "num_tokens": 922384895.0,
      "step": 28901
    },
    {
      "entropy": 0.4387959958985448,
      "epoch": 2.659133645004917,
      "grad_norm": 0.14873863756656647,
      "learning_rate": 1.1365657680866072e-05,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8615534342825413,
      "num_tokens": 922416249.0,
      "step": 28902
    },
    {
      "entropy": 0.581670249812305,
      "epoch": 2.659225651376358,
      "grad_norm": 0.168462872505188,
      "learning_rate": 1.1362590854724446e-05,
      "loss": 0.5727,
      "mean_token_accuracy": 0.819544117897749,
      "num_tokens": 922448300.0,
      "step": 28903
    },
    {
      "entropy": 0.5099145423155278,
      "epoch": 2.659317657747799,
      "grad_norm": 0.1533019244670868,
      "learning_rate": 1.135952402858282e-05,
      "loss": 0.4927,
      "mean_token_accuracy": 0.8440049923956394,
      "num_tokens": 922480507.0,
      "step": 28904
    },
    {
      "entropy": 0.4658008674159646,
      "epoch": 2.65940966411924,
      "grad_norm": 0.1546488255262375,
      "learning_rate": 1.1356457202441194e-05,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8538052029907703,
      "num_tokens": 922512021.0,
      "step": 28905
    },
    {
      "entropy": 0.4960828395560384,
      "epoch": 2.6595016704906813,
      "grad_norm": 0.1549452692270279,
      "learning_rate": 1.1353390376299567e-05,
      "loss": 0.4833,
      "mean_token_accuracy": 0.8460543528199196,
      "num_tokens": 922544015.0,
      "step": 28906
    },
    {
      "entropy": 0.45159418880939484,
      "epoch": 2.659593676862123,
      "grad_norm": 0.14671696722507477,
      "learning_rate": 1.1350323550157942e-05,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8609045967459679,
      "num_tokens": 922576777.0,
      "step": 28907
    },
    {
      "entropy": 0.47090710140764713,
      "epoch": 2.659685683233564,
      "grad_norm": 0.15732842683792114,
      "learning_rate": 1.1347256724016316e-05,
      "loss": 0.4552,
      "mean_token_accuracy": 0.8545369990170002,
      "num_tokens": 922608146.0,
      "step": 28908
    },
    {
      "entropy": 0.5424673091620207,
      "epoch": 2.659777689605005,
      "grad_norm": 0.16100727021694183,
      "learning_rate": 1.134418989787469e-05,
      "loss": 0.5372,
      "mean_token_accuracy": 0.8302918262779713,
      "num_tokens": 922640450.0,
      "step": 28909
    },
    {
      "entropy": 0.48831286933273077,
      "epoch": 2.6598696959764463,
      "grad_norm": 0.15724864602088928,
      "learning_rate": 1.1341123071733064e-05,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8466470204293728,
      "num_tokens": 922671810.0,
      "step": 28910
    },
    {
      "entropy": 0.4829055778682232,
      "epoch": 2.6599617023478874,
      "grad_norm": 0.1476743221282959,
      "learning_rate": 1.1338056245591439e-05,
      "loss": 0.4623,
      "mean_token_accuracy": 0.8529555089771748,
      "num_tokens": 922703351.0,
      "step": 28911
    },
    {
      "entropy": 0.48307639081031084,
      "epoch": 2.660053708719329,
      "grad_norm": 0.15661859512329102,
      "learning_rate": 1.1334989419449813e-05,
      "loss": 0.4714,
      "mean_token_accuracy": 0.8516211286187172,
      "num_tokens": 922735345.0,
      "step": 28912
    },
    {
      "entropy": 0.4895117487758398,
      "epoch": 2.66014571509077,
      "grad_norm": 0.16114845871925354,
      "learning_rate": 1.1331922593308186e-05,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8439592756330967,
      "num_tokens": 922767464.0,
      "step": 28913
    },
    {
      "entropy": 0.43370328308083117,
      "epoch": 2.6602377214622113,
      "grad_norm": 0.15138430893421173,
      "learning_rate": 1.132885576716656e-05,
      "loss": 0.4228,
      "mean_token_accuracy": 0.8636542931199074,
      "num_tokens": 922798151.0,
      "step": 28914
    },
    {
      "entropy": 0.5000717239454389,
      "epoch": 2.6603297278336524,
      "grad_norm": 0.15432488918304443,
      "learning_rate": 1.1325788941024934e-05,
      "loss": 0.489,
      "mean_token_accuracy": 0.8409427553415298,
      "num_tokens": 922830642.0,
      "step": 28915
    },
    {
      "entropy": 0.5173124726861715,
      "epoch": 2.6604217342050935,
      "grad_norm": 0.15670529007911682,
      "learning_rate": 1.1322722114883308e-05,
      "loss": 0.5026,
      "mean_token_accuracy": 0.8383710794150829,
      "num_tokens": 922862693.0,
      "step": 28916
    },
    {
      "entropy": 0.5426669083535671,
      "epoch": 2.660513740576535,
      "grad_norm": 0.1609540581703186,
      "learning_rate": 1.1319655288741681e-05,
      "loss": 0.5257,
      "mean_token_accuracy": 0.832837350666523,
      "num_tokens": 922894530.0,
      "step": 28917
    },
    {
      "entropy": 0.47859931737184525,
      "epoch": 2.6606057469479762,
      "grad_norm": 0.15739698708057404,
      "learning_rate": 1.1316588462600056e-05,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8518793247640133,
      "num_tokens": 922926322.0,
      "step": 28918
    },
    {
      "entropy": 0.48703186167404056,
      "epoch": 2.6606977533194174,
      "grad_norm": 0.15671654045581818,
      "learning_rate": 1.1313521636458429e-05,
      "loss": 0.4747,
      "mean_token_accuracy": 0.8481920436024666,
      "num_tokens": 922958804.0,
      "step": 28919
    },
    {
      "entropy": 0.560217110440135,
      "epoch": 2.6607897596908585,
      "grad_norm": 0.1642894595861435,
      "learning_rate": 1.1310454810316804e-05,
      "loss": 0.546,
      "mean_token_accuracy": 0.8296286836266518,
      "num_tokens": 922990981.0,
      "step": 28920
    },
    {
      "entropy": 0.49745198618620634,
      "epoch": 2.6608817660622996,
      "grad_norm": 0.1539844125509262,
      "learning_rate": 1.1307387984175177e-05,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8477859571576118,
      "num_tokens": 923023140.0,
      "step": 28921
    },
    {
      "entropy": 0.488957014400512,
      "epoch": 2.6609737724337412,
      "grad_norm": 0.15478923916816711,
      "learning_rate": 1.1304321158033551e-05,
      "loss": 0.4753,
      "mean_token_accuracy": 0.8487455509603024,
      "num_tokens": 923054850.0,
      "step": 28922
    },
    {
      "entropy": 0.5700964061543345,
      "epoch": 2.6610657788051824,
      "grad_norm": 0.16471901535987854,
      "learning_rate": 1.1301254331891926e-05,
      "loss": 0.5551,
      "mean_token_accuracy": 0.8205985464155674,
      "num_tokens": 923087256.0,
      "step": 28923
    },
    {
      "entropy": 0.47995313070714474,
      "epoch": 2.6611577851766235,
      "grad_norm": 0.15384256839752197,
      "learning_rate": 1.1298187505750299e-05,
      "loss": 0.4701,
      "mean_token_accuracy": 0.8519251644611359,
      "num_tokens": 923119681.0,
      "step": 28924
    },
    {
      "entropy": 0.44236196018755436,
      "epoch": 2.6612497915480646,
      "grad_norm": 0.14812269806861877,
      "learning_rate": 1.1295120679608673e-05,
      "loss": 0.436,
      "mean_token_accuracy": 0.8631572797894478,
      "num_tokens": 923151309.0,
      "step": 28925
    },
    {
      "entropy": 0.5062760766595602,
      "epoch": 2.6613417979195058,
      "grad_norm": 0.15461573004722595,
      "learning_rate": 1.1292053853467048e-05,
      "loss": 0.503,
      "mean_token_accuracy": 0.8404699973762035,
      "num_tokens": 923183217.0,
      "step": 28926
    },
    {
      "entropy": 0.5116226691752672,
      "epoch": 2.6614338042909473,
      "grad_norm": 0.160628542304039,
      "learning_rate": 1.1288987027325423e-05,
      "loss": 0.5035,
      "mean_token_accuracy": 0.8397082686424255,
      "num_tokens": 923214728.0,
      "step": 28927
    },
    {
      "entropy": 0.5115415155887604,
      "epoch": 2.6615258106623885,
      "grad_norm": 0.16249744594097137,
      "learning_rate": 1.1285920201183796e-05,
      "loss": 0.5021,
      "mean_token_accuracy": 0.8400112837553024,
      "num_tokens": 923246426.0,
      "step": 28928
    },
    {
      "entropy": 0.5428371282760054,
      "epoch": 2.6616178170338296,
      "grad_norm": 0.1628090888261795,
      "learning_rate": 1.128285337504217e-05,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8349001184105873,
      "num_tokens": 923277437.0,
      "step": 28929
    },
    {
      "entropy": 0.4685824781190604,
      "epoch": 2.6617098234052707,
      "grad_norm": 0.1525561511516571,
      "learning_rate": 1.1279786548900543e-05,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8568253889679909,
      "num_tokens": 923309084.0,
      "step": 28930
    },
    {
      "entropy": 0.4966426435858011,
      "epoch": 2.661801829776712,
      "grad_norm": 0.15444599092006683,
      "learning_rate": 1.1276719722758918e-05,
      "loss": 0.4831,
      "mean_token_accuracy": 0.8470142707228661,
      "num_tokens": 923340894.0,
      "step": 28931
    },
    {
      "entropy": 0.5290924794971943,
      "epoch": 2.6618938361481534,
      "grad_norm": 0.16040512919425964,
      "learning_rate": 1.1273652896617291e-05,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8340050876140594,
      "num_tokens": 923373406.0,
      "step": 28932
    },
    {
      "entropy": 0.5312575548887253,
      "epoch": 2.6619858425195946,
      "grad_norm": 0.16461235284805298,
      "learning_rate": 1.1270586070475666e-05,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8341732025146484,
      "num_tokens": 923405443.0,
      "step": 28933
    },
    {
      "entropy": 0.46409989334642887,
      "epoch": 2.6620778488910357,
      "grad_norm": 0.15342390537261963,
      "learning_rate": 1.1267519244334038e-05,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8557058349251747,
      "num_tokens": 923437999.0,
      "step": 28934
    },
    {
      "entropy": 0.5511206295341253,
      "epoch": 2.662169855262477,
      "grad_norm": 0.16294333338737488,
      "learning_rate": 1.1264452418192413e-05,
      "loss": 0.536,
      "mean_token_accuracy": 0.8331990540027618,
      "num_tokens": 923469078.0,
      "step": 28935
    },
    {
      "entropy": 0.518022982403636,
      "epoch": 2.662261861633918,
      "grad_norm": 0.16216179728507996,
      "learning_rate": 1.1261385592050786e-05,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8391892798244953,
      "num_tokens": 923501376.0,
      "step": 28936
    },
    {
      "entropy": 0.570892134681344,
      "epoch": 2.6623538680053596,
      "grad_norm": 0.16574759781360626,
      "learning_rate": 1.125831876590916e-05,
      "loss": 0.5568,
      "mean_token_accuracy": 0.821075763553381,
      "num_tokens": 923533125.0,
      "step": 28937
    },
    {
      "entropy": 0.49072714522480965,
      "epoch": 2.6624458743768007,
      "grad_norm": 0.15988412499427795,
      "learning_rate": 1.1255251939767535e-05,
      "loss": 0.4833,
      "mean_token_accuracy": 0.8454981110990047,
      "num_tokens": 923565450.0,
      "step": 28938
    },
    {
      "entropy": 0.5231800805777311,
      "epoch": 2.662537880748242,
      "grad_norm": 0.161393404006958,
      "learning_rate": 1.1252185113625908e-05,
      "loss": 0.5194,
      "mean_token_accuracy": 0.837826069444418,
      "num_tokens": 923597686.0,
      "step": 28939
    },
    {
      "entropy": 0.4748257640749216,
      "epoch": 2.662629887119683,
      "grad_norm": 0.15582625567913055,
      "learning_rate": 1.1249118287484283e-05,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8511350117623806,
      "num_tokens": 923630354.0,
      "step": 28940
    },
    {
      "entropy": 0.546711812261492,
      "epoch": 2.662721893491124,
      "grad_norm": 0.1615271419286728,
      "learning_rate": 1.1246051461342658e-05,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8304548859596252,
      "num_tokens": 923662685.0,
      "step": 28941
    },
    {
      "entropy": 0.5310400873422623,
      "epoch": 2.6628138998625657,
      "grad_norm": 0.1579841822385788,
      "learning_rate": 1.1242984635201032e-05,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8366258218884468,
      "num_tokens": 923694195.0,
      "step": 28942
    },
    {
      "entropy": 0.5774425230920315,
      "epoch": 2.662905906234007,
      "grad_norm": 0.16462105512619019,
      "learning_rate": 1.1239917809059405e-05,
      "loss": 0.5547,
      "mean_token_accuracy": 0.8247920051217079,
      "num_tokens": 923726050.0,
      "step": 28943
    },
    {
      "entropy": 0.42559211514890194,
      "epoch": 2.662997912605448,
      "grad_norm": 0.14829978346824646,
      "learning_rate": 1.123685098291778e-05,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8639443628489971,
      "num_tokens": 923758613.0,
      "step": 28944
    },
    {
      "entropy": 0.5231024813838303,
      "epoch": 2.663089918976889,
      "grad_norm": 0.1589096039533615,
      "learning_rate": 1.1233784156776153e-05,
      "loss": 0.5039,
      "mean_token_accuracy": 0.8426373340189457,
      "num_tokens": 923790333.0,
      "step": 28945
    },
    {
      "entropy": 0.5026270970702171,
      "epoch": 2.66318192534833,
      "grad_norm": 0.156727597117424,
      "learning_rate": 1.1230717330634527e-05,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8454154655337334,
      "num_tokens": 923821996.0,
      "step": 28946
    },
    {
      "entropy": 0.47761406376957893,
      "epoch": 2.663273931719772,
      "grad_norm": 0.15646043419837952,
      "learning_rate": 1.12276505044929e-05,
      "loss": 0.466,
      "mean_token_accuracy": 0.8498580753803253,
      "num_tokens": 923852810.0,
      "step": 28947
    },
    {
      "entropy": 0.5417598355561495,
      "epoch": 2.663365938091213,
      "grad_norm": 0.16505737602710724,
      "learning_rate": 1.1224583678351275e-05,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8321891874074936,
      "num_tokens": 923884367.0,
      "step": 28948
    },
    {
      "entropy": 0.49484779592603445,
      "epoch": 2.663457944462654,
      "grad_norm": 0.15653371810913086,
      "learning_rate": 1.1221516852209648e-05,
      "loss": 0.4812,
      "mean_token_accuracy": 0.8496362753212452,
      "num_tokens": 923916783.0,
      "step": 28949
    },
    {
      "entropy": 0.5768809113651514,
      "epoch": 2.663549950834095,
      "grad_norm": 0.16701634228229523,
      "learning_rate": 1.1218450026068023e-05,
      "loss": 0.569,
      "mean_token_accuracy": 0.8215111494064331,
      "num_tokens": 923948701.0,
      "step": 28950
    },
    {
      "entropy": 0.493265223223716,
      "epoch": 2.6636419572055363,
      "grad_norm": 0.15907634794712067,
      "learning_rate": 1.1215383199926396e-05,
      "loss": 0.4883,
      "mean_token_accuracy": 0.8486412391066551,
      "num_tokens": 923980044.0,
      "step": 28951
    },
    {
      "entropy": 0.5081691481173038,
      "epoch": 2.663733963576978,
      "grad_norm": 0.16195401549339294,
      "learning_rate": 1.121231637378477e-05,
      "loss": 0.4943,
      "mean_token_accuracy": 0.8420471772551537,
      "num_tokens": 924011328.0,
      "step": 28952
    },
    {
      "entropy": 0.46467008988838643,
      "epoch": 2.663825969948419,
      "grad_norm": 0.14548516273498535,
      "learning_rate": 1.1209249547643145e-05,
      "loss": 0.4484,
      "mean_token_accuracy": 0.857046902179718,
      "num_tokens": 924044096.0,
      "step": 28953
    },
    {
      "entropy": 0.4756378170568496,
      "epoch": 2.66391797631986,
      "grad_norm": 0.16125261783599854,
      "learning_rate": 1.1206182721501518e-05,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8479539565742016,
      "num_tokens": 924075525.0,
      "step": 28954
    },
    {
      "entropy": 0.49428918305784464,
      "epoch": 2.6640099826913013,
      "grad_norm": 0.1546645164489746,
      "learning_rate": 1.1203115895359893e-05,
      "loss": 0.477,
      "mean_token_accuracy": 0.8505925536155701,
      "num_tokens": 924106923.0,
      "step": 28955
    },
    {
      "entropy": 0.4964769734069705,
      "epoch": 2.6641019890627424,
      "grad_norm": 0.15028329193592072,
      "learning_rate": 1.1200049069218267e-05,
      "loss": 0.4841,
      "mean_token_accuracy": 0.8454445451498032,
      "num_tokens": 924139691.0,
      "step": 28956
    },
    {
      "entropy": 0.650342158973217,
      "epoch": 2.664193995434184,
      "grad_norm": 0.17911112308502197,
      "learning_rate": 1.1196982243076642e-05,
      "loss": 0.6469,
      "mean_token_accuracy": 0.8007591180503368,
      "num_tokens": 924170969.0,
      "step": 28957
    },
    {
      "entropy": 0.5576014537364244,
      "epoch": 2.664286001805625,
      "grad_norm": 0.16908691823482513,
      "learning_rate": 1.1193915416935015e-05,
      "loss": 0.5536,
      "mean_token_accuracy": 0.8247827850282192,
      "num_tokens": 924202085.0,
      "step": 28958
    },
    {
      "entropy": 0.44817351165693253,
      "epoch": 2.6643780081770663,
      "grad_norm": 0.1482875496149063,
      "learning_rate": 1.119084859079339e-05,
      "loss": 0.4309,
      "mean_token_accuracy": 0.8606200069189072,
      "num_tokens": 924233948.0,
      "step": 28959
    },
    {
      "entropy": 0.438562348484993,
      "epoch": 2.6644700145485074,
      "grad_norm": 0.15333305299282074,
      "learning_rate": 1.1187781764651762e-05,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8633116334676743,
      "num_tokens": 924266013.0,
      "step": 28960
    },
    {
      "entropy": 0.5276313810609281,
      "epoch": 2.6645620209199485,
      "grad_norm": 0.1591552346944809,
      "learning_rate": 1.1184714938510137e-05,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8364941217005253,
      "num_tokens": 924298254.0,
      "step": 28961
    },
    {
      "entropy": 0.5370555780827999,
      "epoch": 2.66465402729139,
      "grad_norm": 0.16106939315795898,
      "learning_rate": 1.118164811236851e-05,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8360434249043465,
      "num_tokens": 924330051.0,
      "step": 28962
    },
    {
      "entropy": 0.41765765147283673,
      "epoch": 2.6647460336628312,
      "grad_norm": 0.14001674950122833,
      "learning_rate": 1.1178581286226885e-05,
      "loss": 0.4058,
      "mean_token_accuracy": 0.8734435774385929,
      "num_tokens": 924362603.0,
      "step": 28963
    },
    {
      "entropy": 0.38977866526693106,
      "epoch": 2.6648380400342724,
      "grad_norm": 0.14650143682956696,
      "learning_rate": 1.1175514460085258e-05,
      "loss": 0.3808,
      "mean_token_accuracy": 0.8768176361918449,
      "num_tokens": 924394952.0,
      "step": 28964
    },
    {
      "entropy": 0.48052484448999166,
      "epoch": 2.6649300464057135,
      "grad_norm": 0.1559106856584549,
      "learning_rate": 1.1172447633943632e-05,
      "loss": 0.4628,
      "mean_token_accuracy": 0.8533286601305008,
      "num_tokens": 924427164.0,
      "step": 28965
    },
    {
      "entropy": 0.5243033645674586,
      "epoch": 2.6650220527771546,
      "grad_norm": 0.16775502264499664,
      "learning_rate": 1.1169380807802005e-05,
      "loss": 0.5055,
      "mean_token_accuracy": 0.8395177535712719,
      "num_tokens": 924458741.0,
      "step": 28966
    },
    {
      "entropy": 0.509163673967123,
      "epoch": 2.6651140591485962,
      "grad_norm": 0.16050565242767334,
      "learning_rate": 1.116631398166038e-05,
      "loss": 0.4989,
      "mean_token_accuracy": 0.8392348699271679,
      "num_tokens": 924490176.0,
      "step": 28967
    },
    {
      "entropy": 0.5119332000613213,
      "epoch": 2.6652060655200374,
      "grad_norm": 0.16130468249320984,
      "learning_rate": 1.1163247155518754e-05,
      "loss": 0.4991,
      "mean_token_accuracy": 0.840829785913229,
      "num_tokens": 924521829.0,
      "step": 28968
    },
    {
      "entropy": 0.5025908355601132,
      "epoch": 2.6652980718914785,
      "grad_norm": 0.15594322979450226,
      "learning_rate": 1.1160180329377127e-05,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8433061204850674,
      "num_tokens": 924553886.0,
      "step": 28969
    },
    {
      "entropy": 0.5218906933441758,
      "epoch": 2.6653900782629196,
      "grad_norm": 0.1625770628452301,
      "learning_rate": 1.1157113503235502e-05,
      "loss": 0.513,
      "mean_token_accuracy": 0.8384376056492329,
      "num_tokens": 924585700.0,
      "step": 28970
    },
    {
      "entropy": 0.599573103711009,
      "epoch": 2.6654820846343608,
      "grad_norm": 0.16805797815322876,
      "learning_rate": 1.1154046677093877e-05,
      "loss": 0.584,
      "mean_token_accuracy": 0.8190628327429295,
      "num_tokens": 924618229.0,
      "step": 28971
    },
    {
      "entropy": 0.5106224594637752,
      "epoch": 2.6655740910058023,
      "grad_norm": 0.1591033637523651,
      "learning_rate": 1.1150979850952251e-05,
      "loss": 0.4969,
      "mean_token_accuracy": 0.8401020877063274,
      "num_tokens": 924650372.0,
      "step": 28972
    },
    {
      "entropy": 0.45761388330720365,
      "epoch": 2.6656660973772435,
      "grad_norm": 0.15524649620056152,
      "learning_rate": 1.1147913024810624e-05,
      "loss": 0.4379,
      "mean_token_accuracy": 0.8608170822262764,
      "num_tokens": 924682073.0,
      "step": 28973
    },
    {
      "entropy": 0.49422946013510227,
      "epoch": 2.6657581037486846,
      "grad_norm": 0.14990702271461487,
      "learning_rate": 1.1144846198668999e-05,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8493291139602661,
      "num_tokens": 924714161.0,
      "step": 28974
    },
    {
      "entropy": 0.5521335415542126,
      "epoch": 2.6658501101201257,
      "grad_norm": 0.16157270967960358,
      "learning_rate": 1.1141779372527372e-05,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8318251296877861,
      "num_tokens": 924745973.0,
      "step": 28975
    },
    {
      "entropy": 0.5220079682767391,
      "epoch": 2.665942116491567,
      "grad_norm": 0.15784096717834473,
      "learning_rate": 1.1138712546385747e-05,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8413025327026844,
      "num_tokens": 924778162.0,
      "step": 28976
    },
    {
      "entropy": 0.5169216124340892,
      "epoch": 2.6660341228630084,
      "grad_norm": 0.15561740100383759,
      "learning_rate": 1.113564572024412e-05,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8420862071216106,
      "num_tokens": 924810475.0,
      "step": 28977
    },
    {
      "entropy": 0.440541296149604,
      "epoch": 2.6661261292344496,
      "grad_norm": 0.14763256907463074,
      "learning_rate": 1.1132578894102494e-05,
      "loss": 0.4295,
      "mean_token_accuracy": 0.8610527142882347,
      "num_tokens": 924842776.0,
      "step": 28978
    },
    {
      "entropy": 0.6521985046565533,
      "epoch": 2.6662181356058907,
      "grad_norm": 0.17497724294662476,
      "learning_rate": 1.1129512067960867e-05,
      "loss": 0.6477,
      "mean_token_accuracy": 0.8013451360166073,
      "num_tokens": 924874352.0,
      "step": 28979
    },
    {
      "entropy": 0.5369883477687836,
      "epoch": 2.666310141977332,
      "grad_norm": 0.16219791769981384,
      "learning_rate": 1.1126445241819242e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8345106616616249,
      "num_tokens": 924905349.0,
      "step": 28980
    },
    {
      "entropy": 0.5402338672429323,
      "epoch": 2.666402148348773,
      "grad_norm": 0.1617148369550705,
      "learning_rate": 1.1123378415677615e-05,
      "loss": 0.5278,
      "mean_token_accuracy": 0.8314130082726479,
      "num_tokens": 924936899.0,
      "step": 28981
    },
    {
      "entropy": 0.6083605196326971,
      "epoch": 2.6664941547202146,
      "grad_norm": 0.16574402153491974,
      "learning_rate": 1.112031158953599e-05,
      "loss": 0.6052,
      "mean_token_accuracy": 0.8122185431420803,
      "num_tokens": 924969078.0,
      "step": 28982
    },
    {
      "entropy": 0.4812767152907327,
      "epoch": 2.6665861610916557,
      "grad_norm": 0.15250396728515625,
      "learning_rate": 1.1117244763394364e-05,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8528083153069019,
      "num_tokens": 925001335.0,
      "step": 28983
    },
    {
      "entropy": 0.5397818330675364,
      "epoch": 2.666678167463097,
      "grad_norm": 0.1622508317232132,
      "learning_rate": 1.1114177937252737e-05,
      "loss": 0.5334,
      "mean_token_accuracy": 0.8318691812455654,
      "num_tokens": 925033496.0,
      "step": 28984
    },
    {
      "entropy": 0.5339321317151189,
      "epoch": 2.666770173834538,
      "grad_norm": 0.1607717126607895,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8337206915020943,
      "num_tokens": 925065556.0,
      "step": 28985
    },
    {
      "entropy": 0.5835020188242197,
      "epoch": 2.666862180205979,
      "grad_norm": 0.16324159502983093,
      "learning_rate": 1.1108044284969486e-05,
      "loss": 0.57,
      "mean_token_accuracy": 0.8209860622882843,
      "num_tokens": 925098295.0,
      "step": 28986
    },
    {
      "entropy": 0.49358377081807703,
      "epoch": 2.6669541865774207,
      "grad_norm": 0.1533805876970291,
      "learning_rate": 1.1104977458827861e-05,
      "loss": 0.4828,
      "mean_token_accuracy": 0.8459164015948772,
      "num_tokens": 925130588.0,
      "step": 28987
    },
    {
      "entropy": 0.5120349284261465,
      "epoch": 2.667046192948862,
      "grad_norm": 0.15795230865478516,
      "learning_rate": 1.1101910632686234e-05,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8420257307589054,
      "num_tokens": 925163205.0,
      "step": 28988
    },
    {
      "entropy": 0.6037245132029057,
      "epoch": 2.667138199320303,
      "grad_norm": 0.16950905323028564,
      "learning_rate": 1.1098843806544609e-05,
      "loss": 0.606,
      "mean_token_accuracy": 0.8110016584396362,
      "num_tokens": 925195382.0,
      "step": 28989
    },
    {
      "entropy": 0.5511310165748,
      "epoch": 2.667230205691744,
      "grad_norm": 0.16071748733520508,
      "learning_rate": 1.1095776980402981e-05,
      "loss": 0.5407,
      "mean_token_accuracy": 0.8323222398757935,
      "num_tokens": 925227685.0,
      "step": 28990
    },
    {
      "entropy": 0.556705204769969,
      "epoch": 2.667322212063185,
      "grad_norm": 0.16261722147464752,
      "learning_rate": 1.1092710154261356e-05,
      "loss": 0.5395,
      "mean_token_accuracy": 0.8338330201804638,
      "num_tokens": 925259691.0,
      "step": 28991
    },
    {
      "entropy": 0.4655759728047997,
      "epoch": 2.667414218434627,
      "grad_norm": 0.15510036051273346,
      "learning_rate": 1.1089643328119729e-05,
      "loss": 0.4562,
      "mean_token_accuracy": 0.8534000813961029,
      "num_tokens": 925291655.0,
      "step": 28992
    },
    {
      "entropy": 0.5545994117856026,
      "epoch": 2.667506224806068,
      "grad_norm": 0.1610172986984253,
      "learning_rate": 1.1086576501978104e-05,
      "loss": 0.548,
      "mean_token_accuracy": 0.8286736272275448,
      "num_tokens": 925324217.0,
      "step": 28993
    },
    {
      "entropy": 0.5525425337255001,
      "epoch": 2.667598231177509,
      "grad_norm": 0.1642543375492096,
      "learning_rate": 1.1083509675836477e-05,
      "loss": 0.5375,
      "mean_token_accuracy": 0.8288926742970943,
      "num_tokens": 925356226.0,
      "step": 28994
    },
    {
      "entropy": 0.5158248571678996,
      "epoch": 2.66769023754895,
      "grad_norm": 0.1586715430021286,
      "learning_rate": 1.1080442849694851e-05,
      "loss": 0.5021,
      "mean_token_accuracy": 0.8415617123246193,
      "num_tokens": 925388448.0,
      "step": 28995
    },
    {
      "entropy": 0.4734045206569135,
      "epoch": 2.6677822439203913,
      "grad_norm": 0.15068365633487701,
      "learning_rate": 1.1077376023553224e-05,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8512603081762791,
      "num_tokens": 925420485.0,
      "step": 28996
    },
    {
      "entropy": 0.534235917031765,
      "epoch": 2.667874250291833,
      "grad_norm": 0.16652213037014008,
      "learning_rate": 1.1074309197411599e-05,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8322240971028805,
      "num_tokens": 925452038.0,
      "step": 28997
    },
    {
      "entropy": 0.5999002158641815,
      "epoch": 2.667966256663274,
      "grad_norm": 0.16932782530784607,
      "learning_rate": 1.1071242371269974e-05,
      "loss": 0.597,
      "mean_token_accuracy": 0.8103773631155491,
      "num_tokens": 925484176.0,
      "step": 28998
    },
    {
      "entropy": 0.5531786326318979,
      "epoch": 2.668058263034715,
      "grad_norm": 0.1621103584766388,
      "learning_rate": 1.1068175545128347e-05,
      "loss": 0.5535,
      "mean_token_accuracy": 0.8269114382565022,
      "num_tokens": 925516118.0,
      "step": 28999
    },
    {
      "entropy": 0.5450896313413978,
      "epoch": 2.6681502694061563,
      "grad_norm": 0.15771810710430145,
      "learning_rate": 1.1065108718986721e-05,
      "loss": 0.5336,
      "mean_token_accuracy": 0.8344024568796158,
      "num_tokens": 925548332.0,
      "step": 29000
    },
    {
      "entropy": 0.42841882072389126,
      "epoch": 2.6682422757775974,
      "grad_norm": 0.14848016202449799,
      "learning_rate": 1.1062041892845094e-05,
      "loss": 0.4204,
      "mean_token_accuracy": 0.8643499873578548,
      "num_tokens": 925580899.0,
      "step": 29001
    },
    {
      "entropy": 0.5997496405616403,
      "epoch": 2.668334282149039,
      "grad_norm": 0.16600824892520905,
      "learning_rate": 1.105897506670347e-05,
      "loss": 0.5791,
      "mean_token_accuracy": 0.8195013925433159,
      "num_tokens": 925613253.0,
      "step": 29002
    },
    {
      "entropy": 0.5914025008678436,
      "epoch": 2.66842628852048,
      "grad_norm": 0.1685250848531723,
      "learning_rate": 1.1055908240561843e-05,
      "loss": 0.5734,
      "mean_token_accuracy": 0.819195207208395,
      "num_tokens": 925645700.0,
      "step": 29003
    },
    {
      "entropy": 0.30611313495319337,
      "epoch": 2.6685182948919213,
      "grad_norm": 0.13365530967712402,
      "learning_rate": 1.1052841414420218e-05,
      "loss": 0.2925,
      "mean_token_accuracy": 0.9050306044518948,
      "num_tokens": 925678015.0,
      "step": 29004
    },
    {
      "entropy": 0.5618971846997738,
      "epoch": 2.6686103012633624,
      "grad_norm": 0.16466428339481354,
      "learning_rate": 1.1049774588278591e-05,
      "loss": 0.5481,
      "mean_token_accuracy": 0.8251046501100063,
      "num_tokens": 925710166.0,
      "step": 29005
    },
    {
      "entropy": 0.43030161084607244,
      "epoch": 2.6687023076348035,
      "grad_norm": 0.153135746717453,
      "learning_rate": 1.1046707762136966e-05,
      "loss": 0.4261,
      "mean_token_accuracy": 0.8647133968770504,
      "num_tokens": 925741635.0,
      "step": 29006
    },
    {
      "entropy": 0.5349757205694914,
      "epoch": 2.668794314006245,
      "grad_norm": 0.1569041907787323,
      "learning_rate": 1.1043640935995339e-05,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8347212187945843,
      "num_tokens": 925773872.0,
      "step": 29007
    },
    {
      "entropy": 0.6193781588226557,
      "epoch": 2.6688863203776862,
      "grad_norm": 0.17353689670562744,
      "learning_rate": 1.1040574109853713e-05,
      "loss": 0.6044,
      "mean_token_accuracy": 0.8108654506504536,
      "num_tokens": 925805553.0,
      "step": 29008
    },
    {
      "entropy": 0.6077210605144501,
      "epoch": 2.6689783267491274,
      "grad_norm": 0.1678159534931183,
      "learning_rate": 1.1037507283712086e-05,
      "loss": 0.5932,
      "mean_token_accuracy": 0.8137641288340092,
      "num_tokens": 925837298.0,
      "step": 29009
    },
    {
      "entropy": 0.6296145925298333,
      "epoch": 2.6690703331205685,
      "grad_norm": 0.17105787992477417,
      "learning_rate": 1.1034440457570461e-05,
      "loss": 0.6249,
      "mean_token_accuracy": 0.8032599575817585,
      "num_tokens": 925869356.0,
      "step": 29010
    },
    {
      "entropy": 0.5117077101022005,
      "epoch": 2.6691623394920097,
      "grad_norm": 0.1557774394750595,
      "learning_rate": 1.1031373631428834e-05,
      "loss": 0.5026,
      "mean_token_accuracy": 0.8448669016361237,
      "num_tokens": 925901459.0,
      "step": 29011
    },
    {
      "entropy": 0.5559321288019419,
      "epoch": 2.6692543458634512,
      "grad_norm": 0.16427622735500336,
      "learning_rate": 1.1028306805287208e-05,
      "loss": 0.5371,
      "mean_token_accuracy": 0.8263683393597603,
      "num_tokens": 925932301.0,
      "step": 29012
    },
    {
      "entropy": 0.44523279275745153,
      "epoch": 2.6693463522348924,
      "grad_norm": 0.15177235007286072,
      "learning_rate": 1.1025239979145583e-05,
      "loss": 0.4339,
      "mean_token_accuracy": 0.858484324067831,
      "num_tokens": 925963901.0,
      "step": 29013
    },
    {
      "entropy": 0.512452770024538,
      "epoch": 2.6694383586063335,
      "grad_norm": 0.15783990919589996,
      "learning_rate": 1.1022173153003956e-05,
      "loss": 0.4938,
      "mean_token_accuracy": 0.8451763913035393,
      "num_tokens": 925995317.0,
      "step": 29014
    },
    {
      "entropy": 0.4814047934487462,
      "epoch": 2.6695303649777746,
      "grad_norm": 0.15452581644058228,
      "learning_rate": 1.101910632686233e-05,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8488652296364307,
      "num_tokens": 926027087.0,
      "step": 29015
    },
    {
      "entropy": 0.6016160231083632,
      "epoch": 2.6696223713492158,
      "grad_norm": 0.16350087523460388,
      "learning_rate": 1.1016039500720704e-05,
      "loss": 0.5823,
      "mean_token_accuracy": 0.8179098591208458,
      "num_tokens": 926059203.0,
      "step": 29016
    },
    {
      "entropy": 0.5663523990660906,
      "epoch": 2.6697143777206573,
      "grad_norm": 0.16379503905773163,
      "learning_rate": 1.101297267457908e-05,
      "loss": 0.5593,
      "mean_token_accuracy": 0.8259646371006966,
      "num_tokens": 926091468.0,
      "step": 29017
    },
    {
      "entropy": 0.500247348099947,
      "epoch": 2.6698063840920985,
      "grad_norm": 0.1549312025308609,
      "learning_rate": 1.1009905848437453e-05,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8451252169907093,
      "num_tokens": 926122950.0,
      "step": 29018
    },
    {
      "entropy": 0.5183204850181937,
      "epoch": 2.6698983904635396,
      "grad_norm": 0.16201159358024597,
      "learning_rate": 1.1006839022295828e-05,
      "loss": 0.5057,
      "mean_token_accuracy": 0.8365412205457687,
      "num_tokens": 926155041.0,
      "step": 29019
    },
    {
      "entropy": 0.6044253166764975,
      "epoch": 2.6699903968349807,
      "grad_norm": 0.16372638940811157,
      "learning_rate": 1.10037721961542e-05,
      "loss": 0.5921,
      "mean_token_accuracy": 0.8163632974028587,
      "num_tokens": 926187365.0,
      "step": 29020
    },
    {
      "entropy": 0.5107974810525775,
      "epoch": 2.670082403206422,
      "grad_norm": 0.15672288835048676,
      "learning_rate": 1.1000705370012575e-05,
      "loss": 0.4994,
      "mean_token_accuracy": 0.8428320549428463,
      "num_tokens": 926219917.0,
      "step": 29021
    },
    {
      "entropy": 0.37289436196442693,
      "epoch": 2.6701744095778635,
      "grad_norm": 0.13316692411899567,
      "learning_rate": 1.0997638543870948e-05,
      "loss": 0.367,
      "mean_token_accuracy": 0.8835142776370049,
      "num_tokens": 926252303.0,
      "step": 29022
    },
    {
      "entropy": 0.569499246776104,
      "epoch": 2.6702664159493046,
      "grad_norm": 0.16618198156356812,
      "learning_rate": 1.0994571717729323e-05,
      "loss": 0.5508,
      "mean_token_accuracy": 0.8238895758986473,
      "num_tokens": 926284342.0,
      "step": 29023
    },
    {
      "entropy": 0.4474105993285775,
      "epoch": 2.6703584223207457,
      "grad_norm": 0.15356937050819397,
      "learning_rate": 1.0991504891587696e-05,
      "loss": 0.4399,
      "mean_token_accuracy": 0.8587850593030453,
      "num_tokens": 926316656.0,
      "step": 29024
    },
    {
      "entropy": 0.49287514202296734,
      "epoch": 2.670450428692187,
      "grad_norm": 0.15802668035030365,
      "learning_rate": 1.098843806544607e-05,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8507375530898571,
      "num_tokens": 926348326.0,
      "step": 29025
    },
    {
      "entropy": 0.5274893287569284,
      "epoch": 2.670542435063628,
      "grad_norm": 0.1630822718143463,
      "learning_rate": 1.0985371239304443e-05,
      "loss": 0.5135,
      "mean_token_accuracy": 0.8351799622178078,
      "num_tokens": 926380365.0,
      "step": 29026
    },
    {
      "entropy": 0.5584040405228734,
      "epoch": 2.6706344414350696,
      "grad_norm": 0.16657142341136932,
      "learning_rate": 1.0982304413162818e-05,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8254241980612278,
      "num_tokens": 926412227.0,
      "step": 29027
    },
    {
      "entropy": 0.4538695691153407,
      "epoch": 2.6707264478065107,
      "grad_norm": 0.15353335440158844,
      "learning_rate": 1.0979237587021193e-05,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8580325096845627,
      "num_tokens": 926444562.0,
      "step": 29028
    },
    {
      "entropy": 0.4798260619863868,
      "epoch": 2.670818454177952,
      "grad_norm": 0.15533462166786194,
      "learning_rate": 1.0976170760879566e-05,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8509709276258945,
      "num_tokens": 926477330.0,
      "step": 29029
    },
    {
      "entropy": 0.4301416538655758,
      "epoch": 2.670910460549393,
      "grad_norm": 0.14955803751945496,
      "learning_rate": 1.097310393473794e-05,
      "loss": 0.4218,
      "mean_token_accuracy": 0.8640177063643932,
      "num_tokens": 926509414.0,
      "step": 29030
    },
    {
      "entropy": 0.6249440498650074,
      "epoch": 2.671002466920834,
      "grad_norm": 0.17913365364074707,
      "learning_rate": 1.0970037108596313e-05,
      "loss": 0.6234,
      "mean_token_accuracy": 0.8069103918969631,
      "num_tokens": 926540749.0,
      "step": 29031
    },
    {
      "entropy": 0.5327433794736862,
      "epoch": 2.6710944732922757,
      "grad_norm": 0.16176027059555054,
      "learning_rate": 1.096697028245469e-05,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8373197987675667,
      "num_tokens": 926572443.0,
      "step": 29032
    },
    {
      "entropy": 0.45493238652125,
      "epoch": 2.671186479663717,
      "grad_norm": 0.15138272941112518,
      "learning_rate": 1.0963903456313063e-05,
      "loss": 0.4521,
      "mean_token_accuracy": 0.8548189401626587,
      "num_tokens": 926604614.0,
      "step": 29033
    },
    {
      "entropy": 0.5542008746415377,
      "epoch": 2.671278486035158,
      "grad_norm": 0.1637515425682068,
      "learning_rate": 1.0960836630171437e-05,
      "loss": 0.5469,
      "mean_token_accuracy": 0.8317673802375793,
      "num_tokens": 926636963.0,
      "step": 29034
    },
    {
      "entropy": 0.39702135510742664,
      "epoch": 2.671370492406599,
      "grad_norm": 0.14265549182891846,
      "learning_rate": 1.095776980402981e-05,
      "loss": 0.3829,
      "mean_token_accuracy": 0.8759907782077789,
      "num_tokens": 926668927.0,
      "step": 29035
    },
    {
      "entropy": 0.5477992929518223,
      "epoch": 2.67146249877804,
      "grad_norm": 0.16643674671649933,
      "learning_rate": 1.0954702977888185e-05,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8283722177147865,
      "num_tokens": 926701059.0,
      "step": 29036
    },
    {
      "entropy": 0.4260215174872428,
      "epoch": 2.671554505149482,
      "grad_norm": 0.1459444910287857,
      "learning_rate": 1.0951636151746558e-05,
      "loss": 0.4171,
      "mean_token_accuracy": 0.8669386059045792,
      "num_tokens": 926733444.0,
      "step": 29037
    },
    {
      "entropy": 0.48572428803890944,
      "epoch": 2.671646511520923,
      "grad_norm": 0.15319594740867615,
      "learning_rate": 1.0948569325604932e-05,
      "loss": 0.485,
      "mean_token_accuracy": 0.8468580581247807,
      "num_tokens": 926765510.0,
      "step": 29038
    },
    {
      "entropy": 0.5966038608457893,
      "epoch": 2.671738517892364,
      "grad_norm": 0.16364480555057526,
      "learning_rate": 1.0945502499463305e-05,
      "loss": 0.5798,
      "mean_token_accuracy": 0.8175028972327709,
      "num_tokens": 926797455.0,
      "step": 29039
    },
    {
      "entropy": 0.4783415552228689,
      "epoch": 2.671830524263805,
      "grad_norm": 0.15561488270759583,
      "learning_rate": 1.094243567332168e-05,
      "loss": 0.4635,
      "mean_token_accuracy": 0.8500254675745964,
      "num_tokens": 926829173.0,
      "step": 29040
    },
    {
      "entropy": 0.5334785301238298,
      "epoch": 2.6719225306352463,
      "grad_norm": 0.16136260330677032,
      "learning_rate": 1.0939368847180053e-05,
      "loss": 0.5344,
      "mean_token_accuracy": 0.8321207873523235,
      "num_tokens": 926860678.0,
      "step": 29041
    },
    {
      "entropy": 0.5602055545896292,
      "epoch": 2.672014537006688,
      "grad_norm": 0.17394712567329407,
      "learning_rate": 1.0936302021038428e-05,
      "loss": 0.5425,
      "mean_token_accuracy": 0.8265578933060169,
      "num_tokens": 926891274.0,
      "step": 29042
    },
    {
      "entropy": 0.41469179652631283,
      "epoch": 2.672106543378129,
      "grad_norm": 0.15083208680152893,
      "learning_rate": 1.0933235194896802e-05,
      "loss": 0.4108,
      "mean_token_accuracy": 0.8712309375405312,
      "num_tokens": 926922953.0,
      "step": 29043
    },
    {
      "entropy": 0.563254477456212,
      "epoch": 2.67219854974957,
      "grad_norm": 0.16624847054481506,
      "learning_rate": 1.0930168368755175e-05,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8277704864740372,
      "num_tokens": 926954380.0,
      "step": 29044
    },
    {
      "entropy": 0.5423695407807827,
      "epoch": 2.6722905561210113,
      "grad_norm": 0.160902738571167,
      "learning_rate": 1.092710154261355e-05,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8309063240885735,
      "num_tokens": 926986716.0,
      "step": 29045
    },
    {
      "entropy": 0.5668747562449425,
      "epoch": 2.6723825624924524,
      "grad_norm": 0.1620187610387802,
      "learning_rate": 1.0924034716471923e-05,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8331626616418362,
      "num_tokens": 927018982.0,
      "step": 29046
    },
    {
      "entropy": 0.5314799649640918,
      "epoch": 2.672474568863894,
      "grad_norm": 0.1597129851579666,
      "learning_rate": 1.0920967890330299e-05,
      "loss": 0.515,
      "mean_token_accuracy": 0.8365638442337513,
      "num_tokens": 927051255.0,
      "step": 29047
    },
    {
      "entropy": 0.5341995162889361,
      "epoch": 2.672566575235335,
      "grad_norm": 0.16147178411483765,
      "learning_rate": 1.0917901064188672e-05,
      "loss": 0.5325,
      "mean_token_accuracy": 0.8315125480294228,
      "num_tokens": 927082480.0,
      "step": 29048
    },
    {
      "entropy": 0.49803900020197034,
      "epoch": 2.6726585816067763,
      "grad_norm": 0.15724505484104156,
      "learning_rate": 1.0914834238047047e-05,
      "loss": 0.4804,
      "mean_token_accuracy": 0.8502714745700359,
      "num_tokens": 927114444.0,
      "step": 29049
    },
    {
      "entropy": 0.5630996113177389,
      "epoch": 2.6727505879782174,
      "grad_norm": 0.1631740927696228,
      "learning_rate": 1.091176741190542e-05,
      "loss": 0.5436,
      "mean_token_accuracy": 0.8274103365838528,
      "num_tokens": 927146158.0,
      "step": 29050
    },
    {
      "entropy": 0.47229135665111244,
      "epoch": 2.6728425943496585,
      "grad_norm": 0.15443281829357147,
      "learning_rate": 1.0908700585763794e-05,
      "loss": 0.4602,
      "mean_token_accuracy": 0.8521202839910984,
      "num_tokens": 927177156.0,
      "step": 29051
    },
    {
      "entropy": 0.47152706515043974,
      "epoch": 2.6729346007211,
      "grad_norm": 0.15628035366535187,
      "learning_rate": 1.0905633759622167e-05,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8539664410054684,
      "num_tokens": 927209387.0,
      "step": 29052
    },
    {
      "entropy": 0.4994216449558735,
      "epoch": 2.6730266070925413,
      "grad_norm": 0.15587659180164337,
      "learning_rate": 1.0902566933480542e-05,
      "loss": 0.4831,
      "mean_token_accuracy": 0.845450758934021,
      "num_tokens": 927240846.0,
      "step": 29053
    },
    {
      "entropy": 0.5427825478836894,
      "epoch": 2.6731186134639824,
      "grad_norm": 0.1632988452911377,
      "learning_rate": 1.0899500107338915e-05,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8316160105168819,
      "num_tokens": 927273099.0,
      "step": 29054
    },
    {
      "entropy": 0.5494410386309028,
      "epoch": 2.6732106198354235,
      "grad_norm": 0.1579010933637619,
      "learning_rate": 1.089643328119729e-05,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8309660628437996,
      "num_tokens": 927305181.0,
      "step": 29055
    },
    {
      "entropy": 0.4748719008639455,
      "epoch": 2.6733026262068647,
      "grad_norm": 0.15362945199012756,
      "learning_rate": 1.0893366455055662e-05,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8525586128234863,
      "num_tokens": 927337949.0,
      "step": 29056
    },
    {
      "entropy": 0.5893669743090868,
      "epoch": 2.6733946325783062,
      "grad_norm": 0.16860277950763702,
      "learning_rate": 1.0890299628914037e-05,
      "loss": 0.5746,
      "mean_token_accuracy": 0.8208395726978779,
      "num_tokens": 927369928.0,
      "step": 29057
    },
    {
      "entropy": 0.5217621009796858,
      "epoch": 2.6734866389497474,
      "grad_norm": 0.16414669156074524,
      "learning_rate": 1.0887232802772412e-05,
      "loss": 0.513,
      "mean_token_accuracy": 0.8366881757974625,
      "num_tokens": 927401984.0,
      "step": 29058
    },
    {
      "entropy": 0.48885321232955903,
      "epoch": 2.6735786453211885,
      "grad_norm": 0.15178456902503967,
      "learning_rate": 1.0884165976630785e-05,
      "loss": 0.4733,
      "mean_token_accuracy": 0.8508575707674026,
      "num_tokens": 927433889.0,
      "step": 29059
    },
    {
      "entropy": 0.621235903352499,
      "epoch": 2.6736706516926296,
      "grad_norm": 0.17128174006938934,
      "learning_rate": 1.088109915048916e-05,
      "loss": 0.6075,
      "mean_token_accuracy": 0.8108186237514019,
      "num_tokens": 927466260.0,
      "step": 29060
    },
    {
      "entropy": 0.50721391197294,
      "epoch": 2.6737626580640708,
      "grad_norm": 0.15768827497959137,
      "learning_rate": 1.0878032324347532e-05,
      "loss": 0.502,
      "mean_token_accuracy": 0.8418320901691914,
      "num_tokens": 927498205.0,
      "step": 29061
    },
    {
      "entropy": 0.5924313496798277,
      "epoch": 2.6738546644355123,
      "grad_norm": 0.1726992428302765,
      "learning_rate": 1.0874965498205907e-05,
      "loss": 0.5751,
      "mean_token_accuracy": 0.8157818801701069,
      "num_tokens": 927529310.0,
      "step": 29062
    },
    {
      "entropy": 0.5859056878834963,
      "epoch": 2.6739466708069535,
      "grad_norm": 0.16258889436721802,
      "learning_rate": 1.0871898672064282e-05,
      "loss": 0.5786,
      "mean_token_accuracy": 0.8159196302294731,
      "num_tokens": 927562078.0,
      "step": 29063
    },
    {
      "entropy": 0.46865563560277224,
      "epoch": 2.6740386771783946,
      "grad_norm": 0.15828044712543488,
      "learning_rate": 1.0868831845922656e-05,
      "loss": 0.4564,
      "mean_token_accuracy": 0.851958766579628,
      "num_tokens": 927594400.0,
      "step": 29064
    },
    {
      "entropy": 0.5302251763641834,
      "epoch": 2.6741306835498357,
      "grad_norm": 0.16796985268592834,
      "learning_rate": 1.086576501978103e-05,
      "loss": 0.5192,
      "mean_token_accuracy": 0.8346922993659973,
      "num_tokens": 927625873.0,
      "step": 29065
    },
    {
      "entropy": 0.48586096335202456,
      "epoch": 2.674222689921277,
      "grad_norm": 0.15882723033428192,
      "learning_rate": 1.0862698193639404e-05,
      "loss": 0.47,
      "mean_token_accuracy": 0.8480932898819447,
      "num_tokens": 927657735.0,
      "step": 29066
    },
    {
      "entropy": 0.5043757744133472,
      "epoch": 2.6743146962927185,
      "grad_norm": 0.15849190950393677,
      "learning_rate": 1.0859631367497777e-05,
      "loss": 0.4921,
      "mean_token_accuracy": 0.8437571413815022,
      "num_tokens": 927689604.0,
      "step": 29067
    },
    {
      "entropy": 0.5230314498767257,
      "epoch": 2.6744067026641596,
      "grad_norm": 0.1593746840953827,
      "learning_rate": 1.0856564541356151e-05,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8406845442950726,
      "num_tokens": 927721135.0,
      "step": 29068
    },
    {
      "entropy": 0.5274232272058725,
      "epoch": 2.6744987090356007,
      "grad_norm": 0.1585666537284851,
      "learning_rate": 1.0853497715214524e-05,
      "loss": 0.5164,
      "mean_token_accuracy": 0.8371040783822536,
      "num_tokens": 927752915.0,
      "step": 29069
    },
    {
      "entropy": 0.47639091219753027,
      "epoch": 2.674590715407042,
      "grad_norm": 0.14903821051120758,
      "learning_rate": 1.0850430889072899e-05,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8517163693904877,
      "num_tokens": 927785598.0,
      "step": 29070
    },
    {
      "entropy": 0.4126891549676657,
      "epoch": 2.674682721778483,
      "grad_norm": 0.14858806133270264,
      "learning_rate": 1.0847364062931272e-05,
      "loss": 0.4027,
      "mean_token_accuracy": 0.8693737499415874,
      "num_tokens": 927817774.0,
      "step": 29071
    },
    {
      "entropy": 0.45480431988835335,
      "epoch": 2.6747747281499246,
      "grad_norm": 0.15132366120815277,
      "learning_rate": 1.0844297236789647e-05,
      "loss": 0.4334,
      "mean_token_accuracy": 0.8579801954329014,
      "num_tokens": 927849848.0,
      "step": 29072
    },
    {
      "entropy": 0.5652332995086908,
      "epoch": 2.6748667345213657,
      "grad_norm": 0.17098771035671234,
      "learning_rate": 1.0841230410648021e-05,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8240426816046238,
      "num_tokens": 927881732.0,
      "step": 29073
    },
    {
      "entropy": 0.5371096171438694,
      "epoch": 2.674958740892807,
      "grad_norm": 0.159988671541214,
      "learning_rate": 1.0838163584506394e-05,
      "loss": 0.526,
      "mean_token_accuracy": 0.8337366320192814,
      "num_tokens": 927913234.0,
      "step": 29074
    },
    {
      "entropy": 0.4687415543012321,
      "epoch": 2.675050747264248,
      "grad_norm": 0.14962151646614075,
      "learning_rate": 1.0835096758364769e-05,
      "loss": 0.4656,
      "mean_token_accuracy": 0.8531821556389332,
      "num_tokens": 927945196.0,
      "step": 29075
    },
    {
      "entropy": 0.5872183423489332,
      "epoch": 2.675142753635689,
      "grad_norm": 0.16805188357830048,
      "learning_rate": 1.0832029932223142e-05,
      "loss": 0.5814,
      "mean_token_accuracy": 0.8183182552456856,
      "num_tokens": 927977560.0,
      "step": 29076
    },
    {
      "entropy": 0.596471106633544,
      "epoch": 2.6752347600071307,
      "grad_norm": 0.1703442931175232,
      "learning_rate": 1.0828963106081517e-05,
      "loss": 0.592,
      "mean_token_accuracy": 0.8125852830708027,
      "num_tokens": 928009525.0,
      "step": 29077
    },
    {
      "entropy": 0.48598845349624753,
      "epoch": 2.675326766378572,
      "grad_norm": 0.15490247309207916,
      "learning_rate": 1.0825896279939891e-05,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8475095555186272,
      "num_tokens": 928041458.0,
      "step": 29078
    },
    {
      "entropy": 0.42429623659700155,
      "epoch": 2.675418772750013,
      "grad_norm": 0.14690105617046356,
      "learning_rate": 1.0822829453798266e-05,
      "loss": 0.419,
      "mean_token_accuracy": 0.8650667928159237,
      "num_tokens": 928073781.0,
      "step": 29079
    },
    {
      "entropy": 0.44496242702007294,
      "epoch": 2.675510779121454,
      "grad_norm": 0.154465913772583,
      "learning_rate": 1.0819762627656639e-05,
      "loss": 0.4292,
      "mean_token_accuracy": 0.8597441203892231,
      "num_tokens": 928105052.0,
      "step": 29080
    },
    {
      "entropy": 0.5285639325156808,
      "epoch": 2.675602785492895,
      "grad_norm": 0.1609007716178894,
      "learning_rate": 1.0816695801515013e-05,
      "loss": 0.5191,
      "mean_token_accuracy": 0.8375102393329144,
      "num_tokens": 928137557.0,
      "step": 29081
    },
    {
      "entropy": 0.4373868564143777,
      "epoch": 2.675694791864337,
      "grad_norm": 0.15063731372356415,
      "learning_rate": 1.0813628975373386e-05,
      "loss": 0.4248,
      "mean_token_accuracy": 0.8606906421482563,
      "num_tokens": 928169460.0,
      "step": 29082
    },
    {
      "entropy": 0.45949739404022694,
      "epoch": 2.675786798235778,
      "grad_norm": 0.15404659509658813,
      "learning_rate": 1.0810562149231761e-05,
      "loss": 0.4524,
      "mean_token_accuracy": 0.8543753251433372,
      "num_tokens": 928201400.0,
      "step": 29083
    },
    {
      "entropy": 0.5622122902423143,
      "epoch": 2.675878804607219,
      "grad_norm": 0.16408319771289825,
      "learning_rate": 1.0807495323090134e-05,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8234252035617828,
      "num_tokens": 928233371.0,
      "step": 29084
    },
    {
      "entropy": 0.5560120809823275,
      "epoch": 2.67597081097866,
      "grad_norm": 0.16728386282920837,
      "learning_rate": 1.0804428496948509e-05,
      "loss": 0.551,
      "mean_token_accuracy": 0.8255364149808884,
      "num_tokens": 928264895.0,
      "step": 29085
    },
    {
      "entropy": 0.5014004707336426,
      "epoch": 2.6760628173501013,
      "grad_norm": 0.1614663451910019,
      "learning_rate": 1.0801361670806882e-05,
      "loss": 0.4879,
      "mean_token_accuracy": 0.8418073058128357,
      "num_tokens": 928296639.0,
      "step": 29086
    },
    {
      "entropy": 0.527387554757297,
      "epoch": 2.676154823721543,
      "grad_norm": 0.15904060006141663,
      "learning_rate": 1.0798294844665256e-05,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8354627452790737,
      "num_tokens": 928329099.0,
      "step": 29087
    },
    {
      "entropy": 0.5219119358807802,
      "epoch": 2.676246830092984,
      "grad_norm": 0.1582222729921341,
      "learning_rate": 1.0795228018523631e-05,
      "loss": 0.5168,
      "mean_token_accuracy": 0.8371809273958206,
      "num_tokens": 928361464.0,
      "step": 29088
    },
    {
      "entropy": 0.5011910413159057,
      "epoch": 2.676338836464425,
      "grad_norm": 0.16050921380519867,
      "learning_rate": 1.0792161192382004e-05,
      "loss": 0.4844,
      "mean_token_accuracy": 0.8480429463088512,
      "num_tokens": 928392724.0,
      "step": 29089
    },
    {
      "entropy": 0.5175329046323895,
      "epoch": 2.6764308428358663,
      "grad_norm": 0.16624395549297333,
      "learning_rate": 1.0789094366240378e-05,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8375846594572067,
      "num_tokens": 928424155.0,
      "step": 29090
    },
    {
      "entropy": 0.512175839394331,
      "epoch": 2.6765228492073074,
      "grad_norm": 0.16082221269607544,
      "learning_rate": 1.0786027540098751e-05,
      "loss": 0.5006,
      "mean_token_accuracy": 0.839100543409586,
      "num_tokens": 928455663.0,
      "step": 29091
    },
    {
      "entropy": 0.5224656986538321,
      "epoch": 2.676614855578749,
      "grad_norm": 0.16018705070018768,
      "learning_rate": 1.0782960713957126e-05,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8393914327025414,
      "num_tokens": 928486422.0,
      "step": 29092
    },
    {
      "entropy": 0.5067157819867134,
      "epoch": 2.67670686195019,
      "grad_norm": 0.16118581593036652,
      "learning_rate": 1.07798938878155e-05,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8393560498952866,
      "num_tokens": 928517931.0,
      "step": 29093
    },
    {
      "entropy": 0.5299272239208221,
      "epoch": 2.6767988683216313,
      "grad_norm": 0.1606595516204834,
      "learning_rate": 1.0776827061673875e-05,
      "loss": 0.5213,
      "mean_token_accuracy": 0.8382042311131954,
      "num_tokens": 928550388.0,
      "step": 29094
    },
    {
      "entropy": 0.646138571202755,
      "epoch": 2.6768908746930724,
      "grad_norm": 0.17283493280410767,
      "learning_rate": 1.0773760235532248e-05,
      "loss": 0.635,
      "mean_token_accuracy": 0.8021733909845352,
      "num_tokens": 928582825.0,
      "step": 29095
    },
    {
      "entropy": 0.5576494336128235,
      "epoch": 2.6769828810645135,
      "grad_norm": 0.16224925220012665,
      "learning_rate": 1.0770693409390623e-05,
      "loss": 0.5425,
      "mean_token_accuracy": 0.8322168737649918,
      "num_tokens": 928615271.0,
      "step": 29096
    },
    {
      "entropy": 0.5102053992450237,
      "epoch": 2.677074887435955,
      "grad_norm": 0.15824055671691895,
      "learning_rate": 1.0767626583248996e-05,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8422297835350037,
      "num_tokens": 928646757.0,
      "step": 29097
    },
    {
      "entropy": 0.38365119183436036,
      "epoch": 2.6771668938073963,
      "grad_norm": 0.1392364799976349,
      "learning_rate": 1.076455975710737e-05,
      "loss": 0.3676,
      "mean_token_accuracy": 0.8829145804047585,
      "num_tokens": 928679067.0,
      "step": 29098
    },
    {
      "entropy": 0.48673626594245434,
      "epoch": 2.6772589001788374,
      "grad_norm": 0.15697930753231049,
      "learning_rate": 1.0761492930965744e-05,
      "loss": 0.4747,
      "mean_token_accuracy": 0.8490366563200951,
      "num_tokens": 928711559.0,
      "step": 29099
    },
    {
      "entropy": 0.544989001005888,
      "epoch": 2.6773509065502785,
      "grad_norm": 0.1608366221189499,
      "learning_rate": 1.0758426104824118e-05,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8326971083879471,
      "num_tokens": 928744112.0,
      "step": 29100
    },
    {
      "entropy": 0.5064699854701757,
      "epoch": 2.6774429129217197,
      "grad_norm": 0.15664240717887878,
      "learning_rate": 1.0755359278682491e-05,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8426880650222301,
      "num_tokens": 928775756.0,
      "step": 29101
    },
    {
      "entropy": 0.5348106152378023,
      "epoch": 2.6775349192931612,
      "grad_norm": 0.15972189605236053,
      "learning_rate": 1.0752292452540866e-05,
      "loss": 0.5131,
      "mean_token_accuracy": 0.8348997160792351,
      "num_tokens": 928807705.0,
      "step": 29102
    },
    {
      "entropy": 0.41262307576835155,
      "epoch": 2.6776269256646024,
      "grad_norm": 0.15042343735694885,
      "learning_rate": 1.074922562639924e-05,
      "loss": 0.4041,
      "mean_token_accuracy": 0.869668610394001,
      "num_tokens": 928838799.0,
      "step": 29103
    },
    {
      "entropy": 0.5619925716891885,
      "epoch": 2.6777189320360435,
      "grad_norm": 0.16356538236141205,
      "learning_rate": 1.0746158800257613e-05,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8237308748066425,
      "num_tokens": 928870097.0,
      "step": 29104
    },
    {
      "entropy": 0.5053810402750969,
      "epoch": 2.6778109384074846,
      "grad_norm": 0.16145078837871552,
      "learning_rate": 1.0743091974115988e-05,
      "loss": 0.495,
      "mean_token_accuracy": 0.8428517319262028,
      "num_tokens": 928902066.0,
      "step": 29105
    },
    {
      "entropy": 0.6307931579649448,
      "epoch": 2.6779029447789258,
      "grad_norm": 0.16719917953014374,
      "learning_rate": 1.0740025147974361e-05,
      "loss": 0.6126,
      "mean_token_accuracy": 0.8083166033029556,
      "num_tokens": 928934012.0,
      "step": 29106
    },
    {
      "entropy": 0.5955058457329869,
      "epoch": 2.6779949511503673,
      "grad_norm": 0.17121584713459015,
      "learning_rate": 1.0736958321832736e-05,
      "loss": 0.5975,
      "mean_token_accuracy": 0.8129660524427891,
      "num_tokens": 928965616.0,
      "step": 29107
    },
    {
      "entropy": 0.4579356829635799,
      "epoch": 2.6780869575218085,
      "grad_norm": 0.1500275731086731,
      "learning_rate": 1.073389149569111e-05,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8584516234695911,
      "num_tokens": 928997452.0,
      "step": 29108
    },
    {
      "entropy": 0.38030233420431614,
      "epoch": 2.6781789638932496,
      "grad_norm": 0.14044588804244995,
      "learning_rate": 1.0730824669549485e-05,
      "loss": 0.3723,
      "mean_token_accuracy": 0.8813891001045704,
      "num_tokens": 929029917.0,
      "step": 29109
    },
    {
      "entropy": 0.46930020255967975,
      "epoch": 2.6782709702646907,
      "grad_norm": 0.1578858196735382,
      "learning_rate": 1.0727757843407858e-05,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8578573986887932,
      "num_tokens": 929062050.0,
      "step": 29110
    },
    {
      "entropy": 0.5661107383202761,
      "epoch": 2.678362976636132,
      "grad_norm": 0.1660747230052948,
      "learning_rate": 1.0724691017266233e-05,
      "loss": 0.5583,
      "mean_token_accuracy": 0.8234492428600788,
      "num_tokens": 929093876.0,
      "step": 29111
    },
    {
      "entropy": 0.46070524491369724,
      "epoch": 2.6784549830075735,
      "grad_norm": 0.1519545018672943,
      "learning_rate": 1.0721624191124605e-05,
      "loss": 0.4478,
      "mean_token_accuracy": 0.8561176769435406,
      "num_tokens": 929126038.0,
      "step": 29112
    },
    {
      "entropy": 0.4608921455219388,
      "epoch": 2.6785469893790146,
      "grad_norm": 0.15386143326759338,
      "learning_rate": 1.071855736498298e-05,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8545432314276695,
      "num_tokens": 929158806.0,
      "step": 29113
    },
    {
      "entropy": 0.5500083081424236,
      "epoch": 2.6786389957504557,
      "grad_norm": 0.16210201382637024,
      "learning_rate": 1.0715490538841353e-05,
      "loss": 0.5425,
      "mean_token_accuracy": 0.8254607208073139,
      "num_tokens": 929190904.0,
      "step": 29114
    },
    {
      "entropy": 0.573618158698082,
      "epoch": 2.678731002121897,
      "grad_norm": 0.168463334441185,
      "learning_rate": 1.0712423712699728e-05,
      "loss": 0.5698,
      "mean_token_accuracy": 0.8217159993946552,
      "num_tokens": 929222338.0,
      "step": 29115
    },
    {
      "entropy": 0.5029428512789309,
      "epoch": 2.678823008493338,
      "grad_norm": 0.15573178231716156,
      "learning_rate": 1.07093568865581e-05,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8464356474578381,
      "num_tokens": 929254866.0,
      "step": 29116
    },
    {
      "entropy": 0.35925727675203234,
      "epoch": 2.6789150148647796,
      "grad_norm": 0.1368696093559265,
      "learning_rate": 1.0706290060416475e-05,
      "loss": 0.3505,
      "mean_token_accuracy": 0.8856182582676411,
      "num_tokens": 929287351.0,
      "step": 29117
    },
    {
      "entropy": 0.5319208633154631,
      "epoch": 2.6790070212362207,
      "grad_norm": 0.16686195135116577,
      "learning_rate": 1.070322323427485e-05,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8324883058667183,
      "num_tokens": 929319648.0,
      "step": 29118
    },
    {
      "entropy": 0.594547163695097,
      "epoch": 2.679099027607662,
      "grad_norm": 0.1634790301322937,
      "learning_rate": 1.0700156408133223e-05,
      "loss": 0.5863,
      "mean_token_accuracy": 0.8168011717498302,
      "num_tokens": 929351806.0,
      "step": 29119
    },
    {
      "entropy": 0.47385852597653866,
      "epoch": 2.679191033979103,
      "grad_norm": 0.1546432077884674,
      "learning_rate": 1.0697089581991598e-05,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8488203808665276,
      "num_tokens": 929383977.0,
      "step": 29120
    },
    {
      "entropy": 0.5531051428988576,
      "epoch": 2.679283040350544,
      "grad_norm": 0.16782991588115692,
      "learning_rate": 1.069402275584997e-05,
      "loss": 0.5514,
      "mean_token_accuracy": 0.824126847088337,
      "num_tokens": 929415999.0,
      "step": 29121
    },
    {
      "entropy": 0.5793291609734297,
      "epoch": 2.6793750467219857,
      "grad_norm": 0.16907919943332672,
      "learning_rate": 1.0690955929708345e-05,
      "loss": 0.5662,
      "mean_token_accuracy": 0.8204370327293873,
      "num_tokens": 929447357.0,
      "step": 29122
    },
    {
      "entropy": 0.5481646647676826,
      "epoch": 2.679467053093427,
      "grad_norm": 0.1667831838130951,
      "learning_rate": 1.0687889103566718e-05,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8270633071660995,
      "num_tokens": 929479443.0,
      "step": 29123
    },
    {
      "entropy": 0.595465637743473,
      "epoch": 2.679559059464868,
      "grad_norm": 0.17061728239059448,
      "learning_rate": 1.0684822277425094e-05,
      "loss": 0.587,
      "mean_token_accuracy": 0.819617249071598,
      "num_tokens": 929510786.0,
      "step": 29124
    },
    {
      "entropy": 0.4347485499456525,
      "epoch": 2.679651065836309,
      "grad_norm": 0.14587590098381042,
      "learning_rate": 1.0681755451283467e-05,
      "loss": 0.4199,
      "mean_token_accuracy": 0.8663195297122002,
      "num_tokens": 929542715.0,
      "step": 29125
    },
    {
      "entropy": 0.5318776601925492,
      "epoch": 2.67974307220775,
      "grad_norm": 0.1604166328907013,
      "learning_rate": 1.0678688625141842e-05,
      "loss": 0.5216,
      "mean_token_accuracy": 0.8357809148728848,
      "num_tokens": 929574968.0,
      "step": 29126
    },
    {
      "entropy": 0.47246203012764454,
      "epoch": 2.679835078579192,
      "grad_norm": 0.1556895673274994,
      "learning_rate": 1.0675621799000215e-05,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8520087599754333,
      "num_tokens": 929607192.0,
      "step": 29127
    },
    {
      "entropy": 0.5288028009235859,
      "epoch": 2.679927084950633,
      "grad_norm": 0.1554131805896759,
      "learning_rate": 1.067255497285859e-05,
      "loss": 0.5057,
      "mean_token_accuracy": 0.8420649766921997,
      "num_tokens": 929639954.0,
      "step": 29128
    },
    {
      "entropy": 0.5561697371304035,
      "epoch": 2.680019091322074,
      "grad_norm": 0.16249243915081024,
      "learning_rate": 1.0669488146716963e-05,
      "loss": 0.5399,
      "mean_token_accuracy": 0.8298728615045547,
      "num_tokens": 929671925.0,
      "step": 29129
    },
    {
      "entropy": 0.4830034370534122,
      "epoch": 2.680111097693515,
      "grad_norm": 0.154496967792511,
      "learning_rate": 1.0666421320575337e-05,
      "loss": 0.4648,
      "mean_token_accuracy": 0.8485722504556179,
      "num_tokens": 929704071.0,
      "step": 29130
    },
    {
      "entropy": 0.5358920525759459,
      "epoch": 2.6802031040649563,
      "grad_norm": 0.15904296934604645,
      "learning_rate": 1.066335449443371e-05,
      "loss": 0.5235,
      "mean_token_accuracy": 0.8352605439722538,
      "num_tokens": 929736499.0,
      "step": 29131
    },
    {
      "entropy": 0.4520348464138806,
      "epoch": 2.680295110436398,
      "grad_norm": 0.14313974976539612,
      "learning_rate": 1.0660287668292085e-05,
      "loss": 0.4378,
      "mean_token_accuracy": 0.8617584928870201,
      "num_tokens": 929768871.0,
      "step": 29132
    },
    {
      "entropy": 0.5348973907530308,
      "epoch": 2.680387116807839,
      "grad_norm": 0.1622764617204666,
      "learning_rate": 1.065722084215046e-05,
      "loss": 0.518,
      "mean_token_accuracy": 0.8343714699149132,
      "num_tokens": 929800900.0,
      "step": 29133
    },
    {
      "entropy": 0.47016852581873536,
      "epoch": 2.68047912317928,
      "grad_norm": 0.15587309002876282,
      "learning_rate": 1.0654154016008832e-05,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8541034385561943,
      "num_tokens": 929832472.0,
      "step": 29134
    },
    {
      "entropy": 0.5553345214575529,
      "epoch": 2.6805711295507213,
      "grad_norm": 0.16228236258029938,
      "learning_rate": 1.0651087189867207e-05,
      "loss": 0.5398,
      "mean_token_accuracy": 0.8282219283282757,
      "num_tokens": 929864424.0,
      "step": 29135
    },
    {
      "entropy": 0.5990844871848822,
      "epoch": 2.6806631359221624,
      "grad_norm": 0.1693442165851593,
      "learning_rate": 1.064802036372558e-05,
      "loss": 0.574,
      "mean_token_accuracy": 0.8180391639471054,
      "num_tokens": 929896517.0,
      "step": 29136
    },
    {
      "entropy": 0.45786064118146896,
      "epoch": 2.680755142293604,
      "grad_norm": 0.14791707694530487,
      "learning_rate": 1.0644953537583955e-05,
      "loss": 0.4365,
      "mean_token_accuracy": 0.8595118224620819,
      "num_tokens": 929928385.0,
      "step": 29137
    },
    {
      "entropy": 0.49356373585760593,
      "epoch": 2.680847148665045,
      "grad_norm": 0.15694810450077057,
      "learning_rate": 1.0641886711442328e-05,
      "loss": 0.4871,
      "mean_token_accuracy": 0.8453834317624569,
      "num_tokens": 929960490.0,
      "step": 29138
    },
    {
      "entropy": 0.5553853365126997,
      "epoch": 2.6809391550364863,
      "grad_norm": 0.16550230979919434,
      "learning_rate": 1.0638819885300704e-05,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8312368057668209,
      "num_tokens": 929992254.0,
      "step": 29139
    },
    {
      "entropy": 0.47371022403240204,
      "epoch": 2.6810311614079274,
      "grad_norm": 0.1482934206724167,
      "learning_rate": 1.0635753059159077e-05,
      "loss": 0.4561,
      "mean_token_accuracy": 0.8553846590220928,
      "num_tokens": 930023179.0,
      "step": 29140
    },
    {
      "entropy": 0.4495510929264128,
      "epoch": 2.6811231677793685,
      "grad_norm": 0.15371932089328766,
      "learning_rate": 1.0632686233017452e-05,
      "loss": 0.4367,
      "mean_token_accuracy": 0.8578386344015598,
      "num_tokens": 930054573.0,
      "step": 29141
    },
    {
      "entropy": 0.5373139139264822,
      "epoch": 2.68121517415081,
      "grad_norm": 0.16416522860527039,
      "learning_rate": 1.0629619406875825e-05,
      "loss": 0.5234,
      "mean_token_accuracy": 0.8332402184605598,
      "num_tokens": 930086329.0,
      "step": 29142
    },
    {
      "entropy": 0.5201195292174816,
      "epoch": 2.6813071805222513,
      "grad_norm": 0.1630067378282547,
      "learning_rate": 1.06265525807342e-05,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8356684297323227,
      "num_tokens": 930118395.0,
      "step": 29143
    },
    {
      "entropy": 0.48565093986690044,
      "epoch": 2.6813991868936924,
      "grad_norm": 0.15940214693546295,
      "learning_rate": 1.0623485754592572e-05,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8473000638186932,
      "num_tokens": 930149757.0,
      "step": 29144
    },
    {
      "entropy": 0.6158801969140768,
      "epoch": 2.6814911932651335,
      "grad_norm": 0.1714402437210083,
      "learning_rate": 1.0620418928450947e-05,
      "loss": 0.6098,
      "mean_token_accuracy": 0.8082390651106834,
      "num_tokens": 930182371.0,
      "step": 29145
    },
    {
      "entropy": 0.4878916386514902,
      "epoch": 2.6815831996365747,
      "grad_norm": 0.1546008139848709,
      "learning_rate": 1.061735210230932e-05,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8509546779096127,
      "num_tokens": 930214207.0,
      "step": 29146
    },
    {
      "entropy": 0.4891133231576532,
      "epoch": 2.6816752060080162,
      "grad_norm": 0.15509751439094543,
      "learning_rate": 1.0614285276167694e-05,
      "loss": 0.4699,
      "mean_token_accuracy": 0.8504340164363384,
      "num_tokens": 930245529.0,
      "step": 29147
    },
    {
      "entropy": 0.5240641534328461,
      "epoch": 2.6817672123794574,
      "grad_norm": 0.16279830038547516,
      "learning_rate": 1.0611218450026069e-05,
      "loss": 0.508,
      "mean_token_accuracy": 0.8422892801463604,
      "num_tokens": 930277807.0,
      "step": 29148
    },
    {
      "entropy": 0.5042914152145386,
      "epoch": 2.6818592187508985,
      "grad_norm": 0.16222591698169708,
      "learning_rate": 1.0608151623884442e-05,
      "loss": 0.4995,
      "mean_token_accuracy": 0.837932612746954,
      "num_tokens": 930309811.0,
      "step": 29149
    },
    {
      "entropy": 0.41874908935278654,
      "epoch": 2.6819512251223396,
      "grad_norm": 0.14813268184661865,
      "learning_rate": 1.0605084797742817e-05,
      "loss": 0.4135,
      "mean_token_accuracy": 0.8675798997282982,
      "num_tokens": 930342518.0,
      "step": 29150
    },
    {
      "entropy": 0.4741004379466176,
      "epoch": 2.6820432314937808,
      "grad_norm": 0.15445643663406372,
      "learning_rate": 1.060201797160119e-05,
      "loss": 0.4677,
      "mean_token_accuracy": 0.8506585955619812,
      "num_tokens": 930373906.0,
      "step": 29151
    },
    {
      "entropy": 0.4328105580061674,
      "epoch": 2.6821352378652223,
      "grad_norm": 0.1471056342124939,
      "learning_rate": 1.0598951145459564e-05,
      "loss": 0.4181,
      "mean_token_accuracy": 0.8645714148879051,
      "num_tokens": 930405673.0,
      "step": 29152
    },
    {
      "entropy": 0.5199329340830445,
      "epoch": 2.6822272442366635,
      "grad_norm": 0.1641828864812851,
      "learning_rate": 1.0595884319317937e-05,
      "loss": 0.52,
      "mean_token_accuracy": 0.8364401198923588,
      "num_tokens": 930438170.0,
      "step": 29153
    },
    {
      "entropy": 0.4421904543414712,
      "epoch": 2.6823192506081046,
      "grad_norm": 0.15187741816043854,
      "learning_rate": 1.0592817493176314e-05,
      "loss": 0.4249,
      "mean_token_accuracy": 0.863849338144064,
      "num_tokens": 930469716.0,
      "step": 29154
    },
    {
      "entropy": 0.5360309602692723,
      "epoch": 2.6824112569795457,
      "grad_norm": 0.16610382497310638,
      "learning_rate": 1.0589750667034687e-05,
      "loss": 0.5278,
      "mean_token_accuracy": 0.8327465951442719,
      "num_tokens": 930501127.0,
      "step": 29155
    },
    {
      "entropy": 0.5386762591078877,
      "epoch": 2.682503263350987,
      "grad_norm": 0.15999069809913635,
      "learning_rate": 1.0586683840893061e-05,
      "loss": 0.5268,
      "mean_token_accuracy": 0.8314958028495312,
      "num_tokens": 930532824.0,
      "step": 29156
    },
    {
      "entropy": 0.593650040216744,
      "epoch": 2.6825952697224285,
      "grad_norm": 0.1655336618423462,
      "learning_rate": 1.0583617014751434e-05,
      "loss": 0.5848,
      "mean_token_accuracy": 0.8164054192602634,
      "num_tokens": 930565420.0,
      "step": 29157
    },
    {
      "entropy": 0.4554415475577116,
      "epoch": 2.6826872760938696,
      "grad_norm": 0.1535850614309311,
      "learning_rate": 1.0580550188609809e-05,
      "loss": 0.4402,
      "mean_token_accuracy": 0.8595408983528614,
      "num_tokens": 930597522.0,
      "step": 29158
    },
    {
      "entropy": 0.4689847733825445,
      "epoch": 2.6827792824653107,
      "grad_norm": 0.15773405134677887,
      "learning_rate": 1.0577483362468182e-05,
      "loss": 0.4593,
      "mean_token_accuracy": 0.8508142456412315,
      "num_tokens": 930628352.0,
      "step": 29159
    },
    {
      "entropy": 0.4597216220572591,
      "epoch": 2.682871288836752,
      "grad_norm": 0.14590951800346375,
      "learning_rate": 1.0574416536326556e-05,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8629725985229015,
      "num_tokens": 930660619.0,
      "step": 29160
    },
    {
      "entropy": 0.5306468876078725,
      "epoch": 2.682963295208193,
      "grad_norm": 0.16176149249076843,
      "learning_rate": 1.057134971018493e-05,
      "loss": 0.5195,
      "mean_token_accuracy": 0.8318662792444229,
      "num_tokens": 930692712.0,
      "step": 29161
    },
    {
      "entropy": 0.532287654466927,
      "epoch": 2.6830553015796346,
      "grad_norm": 0.1568252295255661,
      "learning_rate": 1.0568282884043304e-05,
      "loss": 0.5228,
      "mean_token_accuracy": 0.8337809890508652,
      "num_tokens": 930725469.0,
      "step": 29162
    },
    {
      "entropy": 0.47119265422225,
      "epoch": 2.6831473079510757,
      "grad_norm": 0.1575099676847458,
      "learning_rate": 1.0565216057901679e-05,
      "loss": 0.4491,
      "mean_token_accuracy": 0.853806160390377,
      "num_tokens": 930756450.0,
      "step": 29163
    },
    {
      "entropy": 0.4610559307038784,
      "epoch": 2.683239314322517,
      "grad_norm": 0.15256956219673157,
      "learning_rate": 1.0562149231760052e-05,
      "loss": 0.4487,
      "mean_token_accuracy": 0.854041825979948,
      "num_tokens": 930788520.0,
      "step": 29164
    },
    {
      "entropy": 0.4662434486672282,
      "epoch": 2.683331320693958,
      "grad_norm": 0.1550869345664978,
      "learning_rate": 1.0559082405618426e-05,
      "loss": 0.4527,
      "mean_token_accuracy": 0.8551915287971497,
      "num_tokens": 930819585.0,
      "step": 29165
    },
    {
      "entropy": 0.46269620675593615,
      "epoch": 2.683423327065399,
      "grad_norm": 0.15052516758441925,
      "learning_rate": 1.05560155794768e-05,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8547797910869122,
      "num_tokens": 930851699.0,
      "step": 29166
    },
    {
      "entropy": 0.5121003407984972,
      "epoch": 2.6835153334368407,
      "grad_norm": 0.16183879971504211,
      "learning_rate": 1.0552948753335174e-05,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8393609933555126,
      "num_tokens": 930883661.0,
      "step": 29167
    },
    {
      "entropy": 0.5157331274822354,
      "epoch": 2.683607339808282,
      "grad_norm": 0.1587165892124176,
      "learning_rate": 1.0549881927193547e-05,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8406166359782219,
      "num_tokens": 930915592.0,
      "step": 29168
    },
    {
      "entropy": 0.4556115595623851,
      "epoch": 2.683699346179723,
      "grad_norm": 0.15376439690589905,
      "learning_rate": 1.0546815101051923e-05,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8548634573817253,
      "num_tokens": 930947364.0,
      "step": 29169
    },
    {
      "entropy": 0.5781977456063032,
      "epoch": 2.683791352551164,
      "grad_norm": 0.1667540818452835,
      "learning_rate": 1.0543748274910296e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8180397339165211,
      "num_tokens": 930979384.0,
      "step": 29170
    },
    {
      "entropy": 0.4638390429317951,
      "epoch": 2.683883358922605,
      "grad_norm": 0.15932855010032654,
      "learning_rate": 1.054068144876867e-05,
      "loss": 0.456,
      "mean_token_accuracy": 0.8555225655436516,
      "num_tokens": 931010651.0,
      "step": 29171
    },
    {
      "entropy": 0.4594523645937443,
      "epoch": 2.683975365294047,
      "grad_norm": 0.14927269518375397,
      "learning_rate": 1.0537614622627044e-05,
      "loss": 0.4399,
      "mean_token_accuracy": 0.8600919619202614,
      "num_tokens": 931043038.0,
      "step": 29172
    },
    {
      "entropy": 0.42161701200529933,
      "epoch": 2.684067371665488,
      "grad_norm": 0.15132659673690796,
      "learning_rate": 1.0534547796485418e-05,
      "loss": 0.4149,
      "mean_token_accuracy": 0.8661377541720867,
      "num_tokens": 931075530.0,
      "step": 29173
    },
    {
      "entropy": 0.5341358855366707,
      "epoch": 2.684159378036929,
      "grad_norm": 0.164937824010849,
      "learning_rate": 1.0531480970343791e-05,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8354602456092834,
      "num_tokens": 931107604.0,
      "step": 29174
    },
    {
      "entropy": 0.5312693770974874,
      "epoch": 2.68425138440837,
      "grad_norm": 0.16425643861293793,
      "learning_rate": 1.0528414144202166e-05,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8349555283784866,
      "num_tokens": 931139989.0,
      "step": 29175
    },
    {
      "entropy": 0.48352982057258487,
      "epoch": 2.6843433907798113,
      "grad_norm": 0.15033091604709625,
      "learning_rate": 1.0525347318060539e-05,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8520390428602695,
      "num_tokens": 931172513.0,
      "step": 29176
    },
    {
      "entropy": 0.5099924076348543,
      "epoch": 2.684435397151253,
      "grad_norm": 0.1599452793598175,
      "learning_rate": 1.0522280491918914e-05,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8408362120389938,
      "num_tokens": 931204571.0,
      "step": 29177
    },
    {
      "entropy": 0.5303939636796713,
      "epoch": 2.684527403522694,
      "grad_norm": 0.1587020307779312,
      "learning_rate": 1.0519213665777288e-05,
      "loss": 0.5097,
      "mean_token_accuracy": 0.8401521556079388,
      "num_tokens": 931236823.0,
      "step": 29178
    },
    {
      "entropy": 0.48370025563053787,
      "epoch": 2.684619409894135,
      "grad_norm": 0.15752677619457245,
      "learning_rate": 1.0516146839635661e-05,
      "loss": 0.482,
      "mean_token_accuracy": 0.8464307375252247,
      "num_tokens": 931267352.0,
      "step": 29179
    },
    {
      "entropy": 0.476715630851686,
      "epoch": 2.6847114162655763,
      "grad_norm": 0.15768879652023315,
      "learning_rate": 1.0513080013494036e-05,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8508002534508705,
      "num_tokens": 931299089.0,
      "step": 29180
    },
    {
      "entropy": 0.5810837391763926,
      "epoch": 2.6848034226370174,
      "grad_norm": 0.16606183350086212,
      "learning_rate": 1.0510013187352409e-05,
      "loss": 0.5717,
      "mean_token_accuracy": 0.8185322992503643,
      "num_tokens": 931331191.0,
      "step": 29181
    },
    {
      "entropy": 0.49023588839918375,
      "epoch": 2.684895429008459,
      "grad_norm": 0.16253307461738586,
      "learning_rate": 1.0506946361210783e-05,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8482393175363541,
      "num_tokens": 931362063.0,
      "step": 29182
    },
    {
      "entropy": 0.5249946867115796,
      "epoch": 2.6849874353799,
      "grad_norm": 0.15699920058250427,
      "learning_rate": 1.0503879535069156e-05,
      "loss": 0.5094,
      "mean_token_accuracy": 0.84017763286829,
      "num_tokens": 931394293.0,
      "step": 29183
    },
    {
      "entropy": 0.47583037056028843,
      "epoch": 2.6850794417513413,
      "grad_norm": 0.163823664188385,
      "learning_rate": 1.0500812708927531e-05,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8554549105465412,
      "num_tokens": 931426379.0,
      "step": 29184
    },
    {
      "entropy": 0.45616112649440765,
      "epoch": 2.6851714481227824,
      "grad_norm": 0.1521030068397522,
      "learning_rate": 1.0497745882785906e-05,
      "loss": 0.4467,
      "mean_token_accuracy": 0.8573392070829868,
      "num_tokens": 931458607.0,
      "step": 29185
    },
    {
      "entropy": 0.42048403806984425,
      "epoch": 2.6852634544942235,
      "grad_norm": 0.1516328901052475,
      "learning_rate": 1.049467905664428e-05,
      "loss": 0.4117,
      "mean_token_accuracy": 0.8668138645589352,
      "num_tokens": 931488982.0,
      "step": 29186
    },
    {
      "entropy": 0.5259637981653214,
      "epoch": 2.685355460865665,
      "grad_norm": 0.1587507575750351,
      "learning_rate": 1.0491612230502653e-05,
      "loss": 0.5124,
      "mean_token_accuracy": 0.8379965424537659,
      "num_tokens": 931521052.0,
      "step": 29187
    },
    {
      "entropy": 0.554237593896687,
      "epoch": 2.6854474672371063,
      "grad_norm": 0.16007307171821594,
      "learning_rate": 1.0488545404361028e-05,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8316040970385075,
      "num_tokens": 931553549.0,
      "step": 29188
    },
    {
      "entropy": 0.57468924485147,
      "epoch": 2.6855394736085474,
      "grad_norm": 0.167453333735466,
      "learning_rate": 1.04854785782194e-05,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8221227638423443,
      "num_tokens": 931585516.0,
      "step": 29189
    },
    {
      "entropy": 0.5478267576545477,
      "epoch": 2.6856314799799885,
      "grad_norm": 0.16341184079647064,
      "learning_rate": 1.0482411752077775e-05,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8273701816797256,
      "num_tokens": 931617089.0,
      "step": 29190
    },
    {
      "entropy": 0.3538625380024314,
      "epoch": 2.6857234863514297,
      "grad_norm": 0.12997561693191528,
      "learning_rate": 1.0479344925936148e-05,
      "loss": 0.3355,
      "mean_token_accuracy": 0.8947465755045414,
      "num_tokens": 931649855.0,
      "step": 29191
    },
    {
      "entropy": 0.42961936071515083,
      "epoch": 2.6858154927228712,
      "grad_norm": 0.146204873919487,
      "learning_rate": 1.0476278099794523e-05,
      "loss": 0.4106,
      "mean_token_accuracy": 0.8689808398485184,
      "num_tokens": 931682055.0,
      "step": 29192
    },
    {
      "entropy": 0.5563770895823836,
      "epoch": 2.6859074990943124,
      "grad_norm": 0.1651734858751297,
      "learning_rate": 1.0473211273652898e-05,
      "loss": 0.5421,
      "mean_token_accuracy": 0.8261435516178608,
      "num_tokens": 931713456.0,
      "step": 29193
    },
    {
      "entropy": 0.5897287412080914,
      "epoch": 2.6859995054657535,
      "grad_norm": 0.16699784994125366,
      "learning_rate": 1.047014444751127e-05,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8192525058984756,
      "num_tokens": 931745556.0,
      "step": 29194
    },
    {
      "entropy": 0.5634769015014172,
      "epoch": 2.6860915118371946,
      "grad_norm": 0.16329412162303925,
      "learning_rate": 1.0467077621369645e-05,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8205353319644928,
      "num_tokens": 931777506.0,
      "step": 29195
    },
    {
      "entropy": 0.48265912011265755,
      "epoch": 2.6861835182086358,
      "grad_norm": 0.15263448655605316,
      "learning_rate": 1.0464010795228018e-05,
      "loss": 0.477,
      "mean_token_accuracy": 0.84885049238801,
      "num_tokens": 931809824.0,
      "step": 29196
    },
    {
      "entropy": 0.5616848934441805,
      "epoch": 2.6862755245800773,
      "grad_norm": 0.16630133986473083,
      "learning_rate": 1.0460943969086393e-05,
      "loss": 0.554,
      "mean_token_accuracy": 0.8219681568443775,
      "num_tokens": 931842401.0,
      "step": 29197
    },
    {
      "entropy": 0.519709394313395,
      "epoch": 2.6863675309515185,
      "grad_norm": 0.1570175141096115,
      "learning_rate": 1.0457877142944766e-05,
      "loss": 0.5072,
      "mean_token_accuracy": 0.8369040377438068,
      "num_tokens": 931874329.0,
      "step": 29198
    },
    {
      "entropy": 0.41380641411524266,
      "epoch": 2.6864595373229596,
      "grad_norm": 0.14811186492443085,
      "learning_rate": 1.045481031680314e-05,
      "loss": 0.3994,
      "mean_token_accuracy": 0.8711238577961922,
      "num_tokens": 931904949.0,
      "step": 29199
    },
    {
      "entropy": 0.49824707116931677,
      "epoch": 2.6865515436944007,
      "grad_norm": 0.1575956642627716,
      "learning_rate": 1.0451743490661515e-05,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8424588590860367,
      "num_tokens": 931937285.0,
      "step": 29200
    },
    {
      "entropy": 0.5025965366512537,
      "epoch": 2.686643550065842,
      "grad_norm": 0.15761882066726685,
      "learning_rate": 1.044867666451989e-05,
      "loss": 0.4969,
      "mean_token_accuracy": 0.842497456818819,
      "num_tokens": 931969752.0,
      "step": 29201
    },
    {
      "entropy": 0.45869811810553074,
      "epoch": 2.6867355564372835,
      "grad_norm": 0.1525208204984665,
      "learning_rate": 1.0445609838378263e-05,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8538332842290401,
      "num_tokens": 932002118.0,
      "step": 29202
    },
    {
      "entropy": 0.45185094280168414,
      "epoch": 2.6868275628087246,
      "grad_norm": 0.15199853479862213,
      "learning_rate": 1.0442543012236637e-05,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8594338968396187,
      "num_tokens": 932034236.0,
      "step": 29203
    },
    {
      "entropy": 0.5359283443540335,
      "epoch": 2.6869195691801657,
      "grad_norm": 0.16200707852840424,
      "learning_rate": 1.043947618609501e-05,
      "loss": 0.5315,
      "mean_token_accuracy": 0.833140417933464,
      "num_tokens": 932065886.0,
      "step": 29204
    },
    {
      "entropy": 0.6102082431316376,
      "epoch": 2.687011575551607,
      "grad_norm": 0.1674414873123169,
      "learning_rate": 1.0436409359953385e-05,
      "loss": 0.5981,
      "mean_token_accuracy": 0.813978049904108,
      "num_tokens": 932097901.0,
      "step": 29205
    },
    {
      "entropy": 0.49558766931295395,
      "epoch": 2.687103581923048,
      "grad_norm": 0.15967105329036713,
      "learning_rate": 1.0433342533811758e-05,
      "loss": 0.4875,
      "mean_token_accuracy": 0.8407452628016472,
      "num_tokens": 932129809.0,
      "step": 29206
    },
    {
      "entropy": 0.4618663275614381,
      "epoch": 2.6871955882944896,
      "grad_norm": 0.15839552879333496,
      "learning_rate": 1.0430275707670133e-05,
      "loss": 0.4591,
      "mean_token_accuracy": 0.8548517301678658,
      "num_tokens": 932160589.0,
      "step": 29207
    },
    {
      "entropy": 0.5364617584273219,
      "epoch": 2.6872875946659307,
      "grad_norm": 0.16180817782878876,
      "learning_rate": 1.0427208881528507e-05,
      "loss": 0.5196,
      "mean_token_accuracy": 0.8365568332374096,
      "num_tokens": 932192372.0,
      "step": 29208
    },
    {
      "entropy": 0.5138307837769389,
      "epoch": 2.687379601037372,
      "grad_norm": 0.15898793935775757,
      "learning_rate": 1.042414205538688e-05,
      "loss": 0.5093,
      "mean_token_accuracy": 0.838036298751831,
      "num_tokens": 932223728.0,
      "step": 29209
    },
    {
      "entropy": 0.49181393743492663,
      "epoch": 2.687471607408813,
      "grad_norm": 0.1563669592142105,
      "learning_rate": 1.0421075229245255e-05,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8454081676900387,
      "num_tokens": 932255601.0,
      "step": 29210
    },
    {
      "entropy": 0.5184032276738435,
      "epoch": 2.687563613780254,
      "grad_norm": 0.15870460867881775,
      "learning_rate": 1.0418008403103628e-05,
      "loss": 0.5172,
      "mean_token_accuracy": 0.8364261835813522,
      "num_tokens": 932287834.0,
      "step": 29211
    },
    {
      "entropy": 0.48366621788591146,
      "epoch": 2.6876556201516957,
      "grad_norm": 0.1534072607755661,
      "learning_rate": 1.0414941576962002e-05,
      "loss": 0.473,
      "mean_token_accuracy": 0.8492547124624252,
      "num_tokens": 932320150.0,
      "step": 29212
    },
    {
      "entropy": 0.5285006952472031,
      "epoch": 2.687747626523137,
      "grad_norm": 0.16130995750427246,
      "learning_rate": 1.0411874750820375e-05,
      "loss": 0.5242,
      "mean_token_accuracy": 0.833349134773016,
      "num_tokens": 932352211.0,
      "step": 29213
    },
    {
      "entropy": 0.5567338466644287,
      "epoch": 2.687839632894578,
      "grad_norm": 0.16166143119335175,
      "learning_rate": 1.040880792467875e-05,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8277596272528172,
      "num_tokens": 932383444.0,
      "step": 29214
    },
    {
      "entropy": 0.5008131205104291,
      "epoch": 2.687931639266019,
      "grad_norm": 0.15822719037532806,
      "learning_rate": 1.0405741098537125e-05,
      "loss": 0.4935,
      "mean_token_accuracy": 0.8426543138921261,
      "num_tokens": 932415838.0,
      "step": 29215
    },
    {
      "entropy": 0.47025738283991814,
      "epoch": 2.68802364563746,
      "grad_norm": 0.15469326078891754,
      "learning_rate": 1.04026742723955e-05,
      "loss": 0.458,
      "mean_token_accuracy": 0.854619923979044,
      "num_tokens": 932447561.0,
      "step": 29216
    },
    {
      "entropy": 0.6187727358192205,
      "epoch": 2.688115652008902,
      "grad_norm": 0.1760704666376114,
      "learning_rate": 1.0399607446253872e-05,
      "loss": 0.6131,
      "mean_token_accuracy": 0.8068516664206982,
      "num_tokens": 932479684.0,
      "step": 29217
    },
    {
      "entropy": 0.5422026957385242,
      "epoch": 2.688207658380343,
      "grad_norm": 0.1611599624156952,
      "learning_rate": 1.0396540620112247e-05,
      "loss": 0.5353,
      "mean_token_accuracy": 0.8283149003982544,
      "num_tokens": 932511555.0,
      "step": 29218
    },
    {
      "entropy": 0.5372706586495042,
      "epoch": 2.688299664751784,
      "grad_norm": 0.1655971109867096,
      "learning_rate": 1.039347379397062e-05,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8308033123612404,
      "num_tokens": 932543669.0,
      "step": 29219
    },
    {
      "entropy": 0.534299430437386,
      "epoch": 2.688391671123225,
      "grad_norm": 0.16598068177700043,
      "learning_rate": 1.0390406967828995e-05,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8326968438923359,
      "num_tokens": 932574828.0,
      "step": 29220
    },
    {
      "entropy": 0.5304202325642109,
      "epoch": 2.6884836774946663,
      "grad_norm": 0.15341132879257202,
      "learning_rate": 1.0387340141687368e-05,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8400706201791763,
      "num_tokens": 932607034.0,
      "step": 29221
    },
    {
      "entropy": 0.552578903734684,
      "epoch": 2.688575683866108,
      "grad_norm": 0.16766001284122467,
      "learning_rate": 1.0384273315545742e-05,
      "loss": 0.544,
      "mean_token_accuracy": 0.8306054063141346,
      "num_tokens": 932638523.0,
      "step": 29222
    },
    {
      "entropy": 0.5097727319225669,
      "epoch": 2.688667690237549,
      "grad_norm": 0.157258078455925,
      "learning_rate": 1.0381206489404117e-05,
      "loss": 0.4974,
      "mean_token_accuracy": 0.84267782792449,
      "num_tokens": 932670819.0,
      "step": 29223
    },
    {
      "entropy": 0.4941113879904151,
      "epoch": 2.68875969660899,
      "grad_norm": 0.15117241442203522,
      "learning_rate": 1.037813966326249e-05,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8462105318903923,
      "num_tokens": 932703138.0,
      "step": 29224
    },
    {
      "entropy": 0.5106331431306899,
      "epoch": 2.6888517029804313,
      "grad_norm": 0.15935012698173523,
      "learning_rate": 1.0375072837120864e-05,
      "loss": 0.4959,
      "mean_token_accuracy": 0.8430171571671963,
      "num_tokens": 932735177.0,
      "step": 29225
    },
    {
      "entropy": 0.5450330693274736,
      "epoch": 2.6889437093518724,
      "grad_norm": 0.1594918966293335,
      "learning_rate": 1.0372006010979237e-05,
      "loss": 0.5186,
      "mean_token_accuracy": 0.8367652706801891,
      "num_tokens": 932766859.0,
      "step": 29226
    },
    {
      "entropy": 0.4473277712240815,
      "epoch": 2.689035715723314,
      "grad_norm": 0.1485416740179062,
      "learning_rate": 1.0368939184837612e-05,
      "loss": 0.4303,
      "mean_token_accuracy": 0.8613817468285561,
      "num_tokens": 932798686.0,
      "step": 29227
    },
    {
      "entropy": 0.5004839859902859,
      "epoch": 2.689127722094755,
      "grad_norm": 0.16257615387439728,
      "learning_rate": 1.0365872358695985e-05,
      "loss": 0.5005,
      "mean_token_accuracy": 0.8431133069097996,
      "num_tokens": 932830641.0,
      "step": 29228
    },
    {
      "entropy": 0.4821107944007963,
      "epoch": 2.6892197284661963,
      "grad_norm": 0.14819711446762085,
      "learning_rate": 1.036280553255436e-05,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8519523218274117,
      "num_tokens": 932863018.0,
      "step": 29229
    },
    {
      "entropy": 0.5859891446307302,
      "epoch": 2.6893117348376374,
      "grad_norm": 0.17136679589748383,
      "learning_rate": 1.0359738706412734e-05,
      "loss": 0.5744,
      "mean_token_accuracy": 0.8170598894357681,
      "num_tokens": 932894407.0,
      "step": 29230
    },
    {
      "entropy": 0.5631163055077195,
      "epoch": 2.6894037412090785,
      "grad_norm": 0.17085671424865723,
      "learning_rate": 1.0356671880271109e-05,
      "loss": 0.5612,
      "mean_token_accuracy": 0.8197073973715305,
      "num_tokens": 932926017.0,
      "step": 29231
    },
    {
      "entropy": 0.40431200130842626,
      "epoch": 2.68949574758052,
      "grad_norm": 0.1428239941596985,
      "learning_rate": 1.0353605054129482e-05,
      "loss": 0.3881,
      "mean_token_accuracy": 0.8773490339517593,
      "num_tokens": 932958336.0,
      "step": 29232
    },
    {
      "entropy": 0.4892857843078673,
      "epoch": 2.6895877539519613,
      "grad_norm": 0.15441784262657166,
      "learning_rate": 1.0350538227987857e-05,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8497432433068752,
      "num_tokens": 932990074.0,
      "step": 29233
    },
    {
      "entropy": 0.4864336410537362,
      "epoch": 2.6896797603234024,
      "grad_norm": 0.1514364778995514,
      "learning_rate": 1.034747140184623e-05,
      "loss": 0.479,
      "mean_token_accuracy": 0.8494936600327492,
      "num_tokens": 933022597.0,
      "step": 29234
    },
    {
      "entropy": 0.5343848336488008,
      "epoch": 2.6897717666948435,
      "grad_norm": 0.1568438559770584,
      "learning_rate": 1.0344404575704604e-05,
      "loss": 0.5239,
      "mean_token_accuracy": 0.836336437612772,
      "num_tokens": 933054571.0,
      "step": 29235
    },
    {
      "entropy": 0.5983282793313265,
      "epoch": 2.6898637730662847,
      "grad_norm": 0.17190830409526825,
      "learning_rate": 1.0341337749562977e-05,
      "loss": 0.5869,
      "mean_token_accuracy": 0.8149010241031647,
      "num_tokens": 933086526.0,
      "step": 29236
    },
    {
      "entropy": 0.42800000845454633,
      "epoch": 2.6899557794377262,
      "grad_norm": 0.142505943775177,
      "learning_rate": 1.0338270923421352e-05,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8669461496174335,
      "num_tokens": 933119095.0,
      "step": 29237
    },
    {
      "entropy": 0.43528796965256333,
      "epoch": 2.6900477858091674,
      "grad_norm": 0.143795907497406,
      "learning_rate": 1.0335204097279726e-05,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8699073120951653,
      "num_tokens": 933150951.0,
      "step": 29238
    },
    {
      "entropy": 0.5073248483240604,
      "epoch": 2.6901397921806085,
      "grad_norm": 0.16037467122077942,
      "learning_rate": 1.03321372711381e-05,
      "loss": 0.5013,
      "mean_token_accuracy": 0.8410360030829906,
      "num_tokens": 933181846.0,
      "step": 29239
    },
    {
      "entropy": 0.5555861964821815,
      "epoch": 2.6902317985520496,
      "grad_norm": 0.17449334263801575,
      "learning_rate": 1.0329070444996474e-05,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8314309120178223,
      "num_tokens": 933212769.0,
      "step": 29240
    },
    {
      "entropy": 0.5482701808214188,
      "epoch": 2.6903238049234908,
      "grad_norm": 0.16627824306488037,
      "learning_rate": 1.0326003618854847e-05,
      "loss": 0.5477,
      "mean_token_accuracy": 0.8271259926259518,
      "num_tokens": 933245242.0,
      "step": 29241
    },
    {
      "entropy": 0.510365679860115,
      "epoch": 2.6904158112949323,
      "grad_norm": 0.15781646966934204,
      "learning_rate": 1.0322936792713222e-05,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8407725542783737,
      "num_tokens": 933277066.0,
      "step": 29242
    },
    {
      "entropy": 0.5702436752617359,
      "epoch": 2.6905078176663735,
      "grad_norm": 0.15824002027511597,
      "learning_rate": 1.0319869966571595e-05,
      "loss": 0.5502,
      "mean_token_accuracy": 0.827984482049942,
      "num_tokens": 933309179.0,
      "step": 29243
    },
    {
      "entropy": 0.5268140025436878,
      "epoch": 2.6905998240378146,
      "grad_norm": 0.15891633927822113,
      "learning_rate": 1.031680314042997e-05,
      "loss": 0.5107,
      "mean_token_accuracy": 0.8403922244906425,
      "num_tokens": 933341278.0,
      "step": 29244
    },
    {
      "entropy": 0.49587623961269855,
      "epoch": 2.6906918304092557,
      "grad_norm": 0.15981720387935638,
      "learning_rate": 1.0313736314288342e-05,
      "loss": 0.4932,
      "mean_token_accuracy": 0.8417462781071663,
      "num_tokens": 933373731.0,
      "step": 29245
    },
    {
      "entropy": 0.518040619790554,
      "epoch": 2.690783836780697,
      "grad_norm": 0.16233490407466888,
      "learning_rate": 1.0310669488146718e-05,
      "loss": 0.5006,
      "mean_token_accuracy": 0.8418398387730122,
      "num_tokens": 933405504.0,
      "step": 29246
    },
    {
      "entropy": 0.5555720273405313,
      "epoch": 2.6908758431521385,
      "grad_norm": 0.16458448767662048,
      "learning_rate": 1.0307602662005091e-05,
      "loss": 0.5425,
      "mean_token_accuracy": 0.8283699154853821,
      "num_tokens": 933437475.0,
      "step": 29247
    },
    {
      "entropy": 0.5346228247508407,
      "epoch": 2.6909678495235796,
      "grad_norm": 0.1597287803888321,
      "learning_rate": 1.0304535835863466e-05,
      "loss": 0.5307,
      "mean_token_accuracy": 0.8310592509806156,
      "num_tokens": 933469140.0,
      "step": 29248
    },
    {
      "entropy": 0.5550370384007692,
      "epoch": 2.6910598558950207,
      "grad_norm": 0.16513359546661377,
      "learning_rate": 1.0301469009721839e-05,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8256226666271687,
      "num_tokens": 933501413.0,
      "step": 29249
    },
    {
      "entropy": 0.5071633299812675,
      "epoch": 2.691151862266462,
      "grad_norm": 0.16313818097114563,
      "learning_rate": 1.0298402183580214e-05,
      "loss": 0.5096,
      "mean_token_accuracy": 0.8412250094115734,
      "num_tokens": 933533103.0,
      "step": 29250
    },
    {
      "entropy": 0.60232337936759,
      "epoch": 2.691243868637903,
      "grad_norm": 0.173886239528656,
      "learning_rate": 1.0295335357438587e-05,
      "loss": 0.6033,
      "mean_token_accuracy": 0.8146075718104839,
      "num_tokens": 933565134.0,
      "step": 29251
    },
    {
      "entropy": 0.48569254484027624,
      "epoch": 2.6913358750093446,
      "grad_norm": 0.15934066474437714,
      "learning_rate": 1.0292268531296961e-05,
      "loss": 0.4737,
      "mean_token_accuracy": 0.8508830964565277,
      "num_tokens": 933597187.0,
      "step": 29252
    },
    {
      "entropy": 0.47687705513089895,
      "epoch": 2.6914278813807857,
      "grad_norm": 0.15277011692523956,
      "learning_rate": 1.0289201705155336e-05,
      "loss": 0.4592,
      "mean_token_accuracy": 0.8512817621231079,
      "num_tokens": 933629075.0,
      "step": 29253
    },
    {
      "entropy": 0.37141164403874427,
      "epoch": 2.691519887752227,
      "grad_norm": 0.1368979960680008,
      "learning_rate": 1.0286134879013709e-05,
      "loss": 0.3566,
      "mean_token_accuracy": 0.8841270618140697,
      "num_tokens": 933661404.0,
      "step": 29254
    },
    {
      "entropy": 0.5811316892504692,
      "epoch": 2.691611894123668,
      "grad_norm": 0.17139452695846558,
      "learning_rate": 1.0283068052872084e-05,
      "loss": 0.586,
      "mean_token_accuracy": 0.8131551370024681,
      "num_tokens": 933692767.0,
      "step": 29255
    },
    {
      "entropy": 0.45868054404854774,
      "epoch": 2.691703900495109,
      "grad_norm": 0.14548203349113464,
      "learning_rate": 1.0280001226730456e-05,
      "loss": 0.4408,
      "mean_token_accuracy": 0.8589348271489143,
      "num_tokens": 933724424.0,
      "step": 29256
    },
    {
      "entropy": 0.5850515849888325,
      "epoch": 2.6917959068665507,
      "grad_norm": 0.1612038016319275,
      "learning_rate": 1.0276934400588831e-05,
      "loss": 0.5684,
      "mean_token_accuracy": 0.8213752545416355,
      "num_tokens": 933756475.0,
      "step": 29257
    },
    {
      "entropy": 0.5507652759552002,
      "epoch": 2.691887913237992,
      "grad_norm": 0.1655668169260025,
      "learning_rate": 1.0273867574447204e-05,
      "loss": 0.5486,
      "mean_token_accuracy": 0.8281443789601326,
      "num_tokens": 933788658.0,
      "step": 29258
    },
    {
      "entropy": 0.6450867229141295,
      "epoch": 2.691979919609433,
      "grad_norm": 0.17874585092067719,
      "learning_rate": 1.0270800748305579e-05,
      "loss": 0.6459,
      "mean_token_accuracy": 0.8029948063194752,
      "num_tokens": 933820262.0,
      "step": 29259
    },
    {
      "entropy": 0.5223426083102822,
      "epoch": 2.692071925980874,
      "grad_norm": 0.1629902571439743,
      "learning_rate": 1.0267733922163952e-05,
      "loss": 0.5096,
      "mean_token_accuracy": 0.8392631895840168,
      "num_tokens": 933851834.0,
      "step": 29260
    },
    {
      "entropy": 0.4628622807795182,
      "epoch": 2.692163932352315,
      "grad_norm": 0.14695358276367188,
      "learning_rate": 1.0264667096022328e-05,
      "loss": 0.4478,
      "mean_token_accuracy": 0.858478344976902,
      "num_tokens": 933884456.0,
      "step": 29261
    },
    {
      "entropy": 0.535019769333303,
      "epoch": 2.692255938723757,
      "grad_norm": 0.1577288955450058,
      "learning_rate": 1.0261600269880701e-05,
      "loss": 0.5279,
      "mean_token_accuracy": 0.8318020030856133,
      "num_tokens": 933916924.0,
      "step": 29262
    },
    {
      "entropy": 0.4686298565939069,
      "epoch": 2.692347945095198,
      "grad_norm": 0.15950049459934235,
      "learning_rate": 1.0258533443739076e-05,
      "loss": 0.4663,
      "mean_token_accuracy": 0.8527521416544914,
      "num_tokens": 933947813.0,
      "step": 29263
    },
    {
      "entropy": 0.5709699224680662,
      "epoch": 2.692439951466639,
      "grad_norm": 0.16633421182632446,
      "learning_rate": 1.0255466617597449e-05,
      "loss": 0.5593,
      "mean_token_accuracy": 0.822895485907793,
      "num_tokens": 933979384.0,
      "step": 29264
    },
    {
      "entropy": 0.5360171068459749,
      "epoch": 2.69253195783808,
      "grad_norm": 0.1618087738752365,
      "learning_rate": 1.0252399791455823e-05,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8334452770650387,
      "num_tokens": 934012152.0,
      "step": 29265
    },
    {
      "entropy": 0.5278794299811125,
      "epoch": 2.6926239642095213,
      "grad_norm": 0.16169984638690948,
      "learning_rate": 1.0249332965314196e-05,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8367498405277729,
      "num_tokens": 934044431.0,
      "step": 29266
    },
    {
      "entropy": 0.4804694256745279,
      "epoch": 2.692715970580963,
      "grad_norm": 0.15806612372398376,
      "learning_rate": 1.024626613917257e-05,
      "loss": 0.467,
      "mean_token_accuracy": 0.8492913208901882,
      "num_tokens": 934076542.0,
      "step": 29267
    },
    {
      "entropy": 0.6422549495473504,
      "epoch": 2.692807976952404,
      "grad_norm": 0.17093196511268616,
      "learning_rate": 1.0243199313030945e-05,
      "loss": 0.6397,
      "mean_token_accuracy": 0.8019002825021744,
      "num_tokens": 934109179.0,
      "step": 29268
    },
    {
      "entropy": 0.5431663766503334,
      "epoch": 2.692899983323845,
      "grad_norm": 0.16097547113895416,
      "learning_rate": 1.0240132486889318e-05,
      "loss": 0.525,
      "mean_token_accuracy": 0.8342290557920933,
      "num_tokens": 934140959.0,
      "step": 29269
    },
    {
      "entropy": 0.43664622120559216,
      "epoch": 2.6929919896952863,
      "grad_norm": 0.1497269868850708,
      "learning_rate": 1.0237065660747693e-05,
      "loss": 0.4134,
      "mean_token_accuracy": 0.8659596368670464,
      "num_tokens": 934172587.0,
      "step": 29270
    },
    {
      "entropy": 0.5644866973161697,
      "epoch": 2.6930839960667274,
      "grad_norm": 0.16387256979942322,
      "learning_rate": 1.0233998834606066e-05,
      "loss": 0.537,
      "mean_token_accuracy": 0.8282393552362919,
      "num_tokens": 934204407.0,
      "step": 29271
    },
    {
      "entropy": 0.5997779173776507,
      "epoch": 2.693176002438169,
      "grad_norm": 0.16815908253192902,
      "learning_rate": 1.023093200846444e-05,
      "loss": 0.583,
      "mean_token_accuracy": 0.8165357820689678,
      "num_tokens": 934235771.0,
      "step": 29272
    },
    {
      "entropy": 0.6439721006900072,
      "epoch": 2.69326800880961,
      "grad_norm": 0.1735810935497284,
      "learning_rate": 1.0227865182322814e-05,
      "loss": 0.6232,
      "mean_token_accuracy": 0.8018937930464745,
      "num_tokens": 934267654.0,
      "step": 29273
    },
    {
      "entropy": 0.5736661432310939,
      "epoch": 2.6933600151810513,
      "grad_norm": 0.16475088894367218,
      "learning_rate": 1.0224798356181188e-05,
      "loss": 0.5467,
      "mean_token_accuracy": 0.8233439065515995,
      "num_tokens": 934300080.0,
      "step": 29274
    },
    {
      "entropy": 0.47275217343121767,
      "epoch": 2.6934520215524924,
      "grad_norm": 0.1559646725654602,
      "learning_rate": 1.0221731530039561e-05,
      "loss": 0.468,
      "mean_token_accuracy": 0.8500835597515106,
      "num_tokens": 934332524.0,
      "step": 29275
    },
    {
      "entropy": 0.40725937858223915,
      "epoch": 2.6935440279239335,
      "grad_norm": 0.14856041967868805,
      "learning_rate": 1.0218664703897938e-05,
      "loss": 0.3992,
      "mean_token_accuracy": 0.8724894411861897,
      "num_tokens": 934364518.0,
      "step": 29276
    },
    {
      "entropy": 0.5186889842152596,
      "epoch": 2.693636034295375,
      "grad_norm": 0.162931427359581,
      "learning_rate": 1.021559787775631e-05,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8397401869297028,
      "num_tokens": 934396491.0,
      "step": 29277
    },
    {
      "entropy": 0.5282134972512722,
      "epoch": 2.6937280406668163,
      "grad_norm": 0.16095460951328278,
      "learning_rate": 1.0212531051614685e-05,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8366091139614582,
      "num_tokens": 934428761.0,
      "step": 29278
    },
    {
      "entropy": 0.5203552655875683,
      "epoch": 2.6938200470382574,
      "grad_norm": 0.15762664377689362,
      "learning_rate": 1.0209464225473058e-05,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8325729370117188,
      "num_tokens": 934460680.0,
      "step": 29279
    },
    {
      "entropy": 0.5200632582418621,
      "epoch": 2.6939120534096985,
      "grad_norm": 0.15929864346981049,
      "learning_rate": 1.0206397399331433e-05,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8359840773046017,
      "num_tokens": 934492260.0,
      "step": 29280
    },
    {
      "entropy": 0.474159381701611,
      "epoch": 2.6940040597811397,
      "grad_norm": 0.15485979616641998,
      "learning_rate": 1.0203330573189806e-05,
      "loss": 0.4493,
      "mean_token_accuracy": 0.8531528860330582,
      "num_tokens": 934524158.0,
      "step": 29281
    },
    {
      "entropy": 0.4106456348672509,
      "epoch": 2.6940960661525812,
      "grad_norm": 0.14859503507614136,
      "learning_rate": 1.020026374704818e-05,
      "loss": 0.4011,
      "mean_token_accuracy": 0.8721068315207958,
      "num_tokens": 934555886.0,
      "step": 29282
    },
    {
      "entropy": 0.5009104963392019,
      "epoch": 2.6941880725240224,
      "grad_norm": 0.16386637091636658,
      "learning_rate": 1.0197196920906555e-05,
      "loss": 0.5,
      "mean_token_accuracy": 0.8383962549269199,
      "num_tokens": 934587675.0,
      "step": 29283
    },
    {
      "entropy": 0.49638534570112824,
      "epoch": 2.6942800788954635,
      "grad_norm": 0.15661326050758362,
      "learning_rate": 1.0194130094764928e-05,
      "loss": 0.4946,
      "mean_token_accuracy": 0.8429328538477421,
      "num_tokens": 934620068.0,
      "step": 29284
    },
    {
      "entropy": 0.5027894247323275,
      "epoch": 2.6943720852669046,
      "grad_norm": 0.15924623608589172,
      "learning_rate": 1.0191063268623303e-05,
      "loss": 0.4869,
      "mean_token_accuracy": 0.8428780101239681,
      "num_tokens": 934651480.0,
      "step": 29285
    },
    {
      "entropy": 0.5733790118247271,
      "epoch": 2.6944640916383458,
      "grad_norm": 0.16390696167945862,
      "learning_rate": 1.0187996442481676e-05,
      "loss": 0.556,
      "mean_token_accuracy": 0.8246089443564415,
      "num_tokens": 934682993.0,
      "step": 29286
    },
    {
      "entropy": 0.526812769472599,
      "epoch": 2.6945560980097873,
      "grad_norm": 0.15940111875534058,
      "learning_rate": 1.018492961634005e-05,
      "loss": 0.5278,
      "mean_token_accuracy": 0.836966760456562,
      "num_tokens": 934715154.0,
      "step": 29287
    },
    {
      "entropy": 0.5589410967659205,
      "epoch": 2.6946481043812285,
      "grad_norm": 0.16743184626102448,
      "learning_rate": 1.0181862790198423e-05,
      "loss": 0.5517,
      "mean_token_accuracy": 0.8238426074385643,
      "num_tokens": 934746945.0,
      "step": 29288
    },
    {
      "entropy": 0.496250297408551,
      "epoch": 2.6947401107526696,
      "grad_norm": 0.1585020124912262,
      "learning_rate": 1.0178795964056798e-05,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8408766649663448,
      "num_tokens": 934779263.0,
      "step": 29289
    },
    {
      "entropy": 0.4492460200563073,
      "epoch": 2.6948321171241107,
      "grad_norm": 0.1538180261850357,
      "learning_rate": 1.017572913791517e-05,
      "loss": 0.4523,
      "mean_token_accuracy": 0.858694177120924,
      "num_tokens": 934811285.0,
      "step": 29290
    },
    {
      "entropy": 0.5488769058138132,
      "epoch": 2.694924123495552,
      "grad_norm": 0.16300375759601593,
      "learning_rate": 1.0172662311773547e-05,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8323991373181343,
      "num_tokens": 934843217.0,
      "step": 29291
    },
    {
      "entropy": 0.5101597065804526,
      "epoch": 2.6950161298669935,
      "grad_norm": 0.15797731280326843,
      "learning_rate": 1.016959548563192e-05,
      "loss": 0.5,
      "mean_token_accuracy": 0.8414421044290066,
      "num_tokens": 934875077.0,
      "step": 29292
    },
    {
      "entropy": 0.4859344754368067,
      "epoch": 2.6951081362384346,
      "grad_norm": 0.15103112161159515,
      "learning_rate": 1.0166528659490295e-05,
      "loss": 0.467,
      "mean_token_accuracy": 0.8546303138136864,
      "num_tokens": 934907247.0,
      "step": 29293
    },
    {
      "entropy": 0.6186782345175743,
      "epoch": 2.6952001426098757,
      "grad_norm": 0.16795091331005096,
      "learning_rate": 1.0163461833348668e-05,
      "loss": 0.6004,
      "mean_token_accuracy": 0.8135189078748226,
      "num_tokens": 934939120.0,
      "step": 29294
    },
    {
      "entropy": 0.4623869275674224,
      "epoch": 2.695292148981317,
      "grad_norm": 0.15116751194000244,
      "learning_rate": 1.0160395007207042e-05,
      "loss": 0.4523,
      "mean_token_accuracy": 0.8593169450759888,
      "num_tokens": 934970931.0,
      "step": 29295
    },
    {
      "entropy": 0.5080476179718971,
      "epoch": 2.695384155352758,
      "grad_norm": 0.1542121171951294,
      "learning_rate": 1.0157328181065415e-05,
      "loss": 0.494,
      "mean_token_accuracy": 0.8430944122374058,
      "num_tokens": 935003014.0,
      "step": 29296
    },
    {
      "entropy": 0.5281639085151255,
      "epoch": 2.6954761617241996,
      "grad_norm": 0.16205093264579773,
      "learning_rate": 1.015426135492379e-05,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8344109356403351,
      "num_tokens": 935034969.0,
      "step": 29297
    },
    {
      "entropy": 0.6280229687690735,
      "epoch": 2.6955681680956407,
      "grad_norm": 0.17421941459178925,
      "learning_rate": 1.0151194528782165e-05,
      "loss": 0.6143,
      "mean_token_accuracy": 0.80893250182271,
      "num_tokens": 935066831.0,
      "step": 29298
    },
    {
      "entropy": 0.4604650381952524,
      "epoch": 2.695660174467082,
      "grad_norm": 0.1499304473400116,
      "learning_rate": 1.0148127702640538e-05,
      "loss": 0.4497,
      "mean_token_accuracy": 0.8573344424366951,
      "num_tokens": 935098918.0,
      "step": 29299
    },
    {
      "entropy": 0.5345950499176979,
      "epoch": 2.695752180838523,
      "grad_norm": 0.16396000981330872,
      "learning_rate": 1.0145060876498912e-05,
      "loss": 0.526,
      "mean_token_accuracy": 0.8360309451818466,
      "num_tokens": 935128722.0,
      "step": 29300
    },
    {
      "entropy": 0.4808238008990884,
      "epoch": 2.695844187209964,
      "grad_norm": 0.1537773162126541,
      "learning_rate": 1.0141994050357285e-05,
      "loss": 0.4586,
      "mean_token_accuracy": 0.8518048524856567,
      "num_tokens": 935160925.0,
      "step": 29301
    },
    {
      "entropy": 0.5016135852783918,
      "epoch": 2.6959361935814057,
      "grad_norm": 0.15233628451824188,
      "learning_rate": 1.013892722421566e-05,
      "loss": 0.4844,
      "mean_token_accuracy": 0.8432054631412029,
      "num_tokens": 935193614.0,
      "step": 29302
    },
    {
      "entropy": 0.4921155631309375,
      "epoch": 2.696028199952847,
      "grad_norm": 0.15247498452663422,
      "learning_rate": 1.0135860398074033e-05,
      "loss": 0.4862,
      "mean_token_accuracy": 0.8502289578318596,
      "num_tokens": 935225739.0,
      "step": 29303
    },
    {
      "entropy": 0.5378830004483461,
      "epoch": 2.696120206324288,
      "grad_norm": 0.1639513075351715,
      "learning_rate": 1.0132793571932407e-05,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8323432765901089,
      "num_tokens": 935257128.0,
      "step": 29304
    },
    {
      "entropy": 0.44813902862370014,
      "epoch": 2.696212212695729,
      "grad_norm": 0.14500311017036438,
      "learning_rate": 1.012972674579078e-05,
      "loss": 0.436,
      "mean_token_accuracy": 0.8592525646090508,
      "num_tokens": 935289257.0,
      "step": 29305
    },
    {
      "entropy": 0.4699680022895336,
      "epoch": 2.69630421906717,
      "grad_norm": 0.1515929102897644,
      "learning_rate": 1.0126659919649155e-05,
      "loss": 0.4547,
      "mean_token_accuracy": 0.8548545651137829,
      "num_tokens": 935321545.0,
      "step": 29306
    },
    {
      "entropy": 0.40512541262432933,
      "epoch": 2.696396225438612,
      "grad_norm": 0.14594095945358276,
      "learning_rate": 1.012359309350753e-05,
      "loss": 0.3909,
      "mean_token_accuracy": 0.8768077827990055,
      "num_tokens": 935353625.0,
      "step": 29307
    },
    {
      "entropy": 0.5207513328641653,
      "epoch": 2.696488231810053,
      "grad_norm": 0.16635651886463165,
      "learning_rate": 1.0120526267365904e-05,
      "loss": 0.512,
      "mean_token_accuracy": 0.8344907946884632,
      "num_tokens": 935385128.0,
      "step": 29308
    },
    {
      "entropy": 0.5989111829549074,
      "epoch": 2.696580238181494,
      "grad_norm": 0.17599612474441528,
      "learning_rate": 1.0117459441224277e-05,
      "loss": 0.605,
      "mean_token_accuracy": 0.8136353194713593,
      "num_tokens": 935417238.0,
      "step": 29309
    },
    {
      "entropy": 0.4947807937860489,
      "epoch": 2.696672244552935,
      "grad_norm": 0.1599906086921692,
      "learning_rate": 1.0114392615082652e-05,
      "loss": 0.4746,
      "mean_token_accuracy": 0.847552478313446,
      "num_tokens": 935448375.0,
      "step": 29310
    },
    {
      "entropy": 0.5953299980610609,
      "epoch": 2.6967642509243763,
      "grad_norm": 0.16735689342021942,
      "learning_rate": 1.0111325788941025e-05,
      "loss": 0.5822,
      "mean_token_accuracy": 0.8163962438702583,
      "num_tokens": 935480602.0,
      "step": 29311
    },
    {
      "entropy": 0.5017325319349766,
      "epoch": 2.696856257295818,
      "grad_norm": 0.16154660284519196,
      "learning_rate": 1.01082589627994e-05,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8454618118703365,
      "num_tokens": 935512533.0,
      "step": 29312
    },
    {
      "entropy": 0.639034416526556,
      "epoch": 2.696948263667259,
      "grad_norm": 0.1763482689857483,
      "learning_rate": 1.0105192136657774e-05,
      "loss": 0.6321,
      "mean_token_accuracy": 0.805263988673687,
      "num_tokens": 935544416.0,
      "step": 29313
    },
    {
      "entropy": 0.5961203556507826,
      "epoch": 2.6970402700387,
      "grad_norm": 0.16802315413951874,
      "learning_rate": 1.0102125310516147e-05,
      "loss": 0.5836,
      "mean_token_accuracy": 0.8204915672540665,
      "num_tokens": 935576901.0,
      "step": 29314
    },
    {
      "entropy": 0.4793361518532038,
      "epoch": 2.6971322764101413,
      "grad_norm": 0.15654417872428894,
      "learning_rate": 1.0099058484374522e-05,
      "loss": 0.4685,
      "mean_token_accuracy": 0.850000012665987,
      "num_tokens": 935608740.0,
      "step": 29315
    },
    {
      "entropy": 0.4411387490108609,
      "epoch": 2.6972242827815824,
      "grad_norm": 0.14349143207073212,
      "learning_rate": 1.0095991658232895e-05,
      "loss": 0.4245,
      "mean_token_accuracy": 0.8641449399292469,
      "num_tokens": 935641239.0,
      "step": 29316
    },
    {
      "entropy": 0.5698456950485706,
      "epoch": 2.697316289153024,
      "grad_norm": 0.17418710887432098,
      "learning_rate": 1.009292483209127e-05,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8229559808969498,
      "num_tokens": 935672199.0,
      "step": 29317
    },
    {
      "entropy": 0.48822549264878035,
      "epoch": 2.697408295524465,
      "grad_norm": 0.16012275218963623,
      "learning_rate": 1.0089858005949642e-05,
      "loss": 0.4808,
      "mean_token_accuracy": 0.8508561514317989,
      "num_tokens": 935703673.0,
      "step": 29318
    },
    {
      "entropy": 0.628047950565815,
      "epoch": 2.6975003018959063,
      "grad_norm": 0.1738138496875763,
      "learning_rate": 1.0086791179808017e-05,
      "loss": 0.6198,
      "mean_token_accuracy": 0.8084590137004852,
      "num_tokens": 935735212.0,
      "step": 29319
    },
    {
      "entropy": 0.4194897774141282,
      "epoch": 2.6975923082673474,
      "grad_norm": 0.1497780978679657,
      "learning_rate": 1.008372435366639e-05,
      "loss": 0.4068,
      "mean_token_accuracy": 0.8692225255072117,
      "num_tokens": 935766944.0,
      "step": 29320
    },
    {
      "entropy": 0.5341077925404534,
      "epoch": 2.6976843146387885,
      "grad_norm": 0.15964257717132568,
      "learning_rate": 1.0080657527524765e-05,
      "loss": 0.5277,
      "mean_token_accuracy": 0.8326819725334644,
      "num_tokens": 935799712.0,
      "step": 29321
    },
    {
      "entropy": 0.4829802393214777,
      "epoch": 2.69777632101023,
      "grad_norm": 0.15683172643184662,
      "learning_rate": 1.007759070138314e-05,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8548086769878864,
      "num_tokens": 935831987.0,
      "step": 29322
    },
    {
      "entropy": 0.5455260407179594,
      "epoch": 2.6978683273816713,
      "grad_norm": 0.16412056982517242,
      "learning_rate": 1.0074523875241514e-05,
      "loss": 0.5356,
      "mean_token_accuracy": 0.8306802660226822,
      "num_tokens": 935864383.0,
      "step": 29323
    },
    {
      "entropy": 0.5145674999803305,
      "epoch": 2.6979603337531124,
      "grad_norm": 0.1619911789894104,
      "learning_rate": 1.0071457049099887e-05,
      "loss": 0.5082,
      "mean_token_accuracy": 0.8377228938043118,
      "num_tokens": 935895789.0,
      "step": 29324
    },
    {
      "entropy": 0.4727522018365562,
      "epoch": 2.6980523401245535,
      "grad_norm": 0.15108971297740936,
      "learning_rate": 1.0068390222958261e-05,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8522473685443401,
      "num_tokens": 935927496.0,
      "step": 29325
    },
    {
      "entropy": 0.558165455237031,
      "epoch": 2.6981443464959947,
      "grad_norm": 0.16445815563201904,
      "learning_rate": 1.0065323396816634e-05,
      "loss": 0.5399,
      "mean_token_accuracy": 0.8291667811572552,
      "num_tokens": 935959666.0,
      "step": 29326
    },
    {
      "entropy": 0.4538397123105824,
      "epoch": 2.6982363528674362,
      "grad_norm": 0.14985014498233795,
      "learning_rate": 1.0062256570675009e-05,
      "loss": 0.4504,
      "mean_token_accuracy": 0.858635064214468,
      "num_tokens": 935992053.0,
      "step": 29327
    },
    {
      "entropy": 0.5303164835786447,
      "epoch": 2.6983283592388774,
      "grad_norm": 0.16237477958202362,
      "learning_rate": 1.0059189744533384e-05,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8381507843732834,
      "num_tokens": 936023418.0,
      "step": 29328
    },
    {
      "entropy": 0.5147393317893147,
      "epoch": 2.6984203656103185,
      "grad_norm": 0.15718267858028412,
      "learning_rate": 1.0056122918391757e-05,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8405371159315109,
      "num_tokens": 936055831.0,
      "step": 29329
    },
    {
      "entropy": 0.5365991545841098,
      "epoch": 2.6985123719817596,
      "grad_norm": 0.16277290880680084,
      "learning_rate": 1.0053056092250131e-05,
      "loss": 0.5271,
      "mean_token_accuracy": 0.8347295559942722,
      "num_tokens": 936088297.0,
      "step": 29330
    },
    {
      "entropy": 0.4439056321280077,
      "epoch": 2.6986043783532008,
      "grad_norm": 0.15226441621780396,
      "learning_rate": 1.0049989266108504e-05,
      "loss": 0.4336,
      "mean_token_accuracy": 0.8591758944094181,
      "num_tokens": 936120509.0,
      "step": 29331
    },
    {
      "entropy": 0.5112483464181423,
      "epoch": 2.6986963847246423,
      "grad_norm": 0.16206610202789307,
      "learning_rate": 1.0046922439966879e-05,
      "loss": 0.4954,
      "mean_token_accuracy": 0.8421998657286167,
      "num_tokens": 936153087.0,
      "step": 29332
    },
    {
      "entropy": 0.539549482986331,
      "epoch": 2.6987883910960835,
      "grad_norm": 0.16464123129844666,
      "learning_rate": 1.0043855613825252e-05,
      "loss": 0.5263,
      "mean_token_accuracy": 0.8330859541893005,
      "num_tokens": 936184832.0,
      "step": 29333
    },
    {
      "entropy": 0.541611310094595,
      "epoch": 2.6988803974675246,
      "grad_norm": 0.15987667441368103,
      "learning_rate": 1.0040788787683626e-05,
      "loss": 0.5365,
      "mean_token_accuracy": 0.8294240348041058,
      "num_tokens": 936217038.0,
      "step": 29334
    },
    {
      "entropy": 0.5179210482165217,
      "epoch": 2.6989724038389658,
      "grad_norm": 0.16068647801876068,
      "learning_rate": 1.0037721961542e-05,
      "loss": 0.5011,
      "mean_token_accuracy": 0.8401544727385044,
      "num_tokens": 936248593.0,
      "step": 29335
    },
    {
      "entropy": 0.5126949598779902,
      "epoch": 2.699064410210407,
      "grad_norm": 0.15327002108097076,
      "learning_rate": 1.0034655135400374e-05,
      "loss": 0.499,
      "mean_token_accuracy": 0.840752437710762,
      "num_tokens": 936280711.0,
      "step": 29336
    },
    {
      "entropy": 0.5069447001442313,
      "epoch": 2.6991564165818485,
      "grad_norm": 0.15479129552841187,
      "learning_rate": 1.0031588309258749e-05,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8432827368378639,
      "num_tokens": 936313241.0,
      "step": 29337
    },
    {
      "entropy": 0.48754640691913664,
      "epoch": 2.6992484229532896,
      "grad_norm": 0.15597228705883026,
      "learning_rate": 1.0028521483117123e-05,
      "loss": 0.4812,
      "mean_token_accuracy": 0.8466726765036583,
      "num_tokens": 936345188.0,
      "step": 29338
    },
    {
      "entropy": 0.46410259790718555,
      "epoch": 2.6993404293247307,
      "grad_norm": 0.1543119102716446,
      "learning_rate": 1.0025454656975496e-05,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8586948737502098,
      "num_tokens": 936376354.0,
      "step": 29339
    },
    {
      "entropy": 0.448815296869725,
      "epoch": 2.699432435696172,
      "grad_norm": 0.14958883821964264,
      "learning_rate": 1.0022387830833871e-05,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8585947975516319,
      "num_tokens": 936407677.0,
      "step": 29340
    },
    {
      "entropy": 0.5721731744706631,
      "epoch": 2.699524442067613,
      "grad_norm": 0.1653864085674286,
      "learning_rate": 1.0019321004692244e-05,
      "loss": 0.563,
      "mean_token_accuracy": 0.8194924928247929,
      "num_tokens": 936440306.0,
      "step": 29341
    },
    {
      "entropy": 0.48406486213207245,
      "epoch": 2.6996164484390546,
      "grad_norm": 0.15035246312618256,
      "learning_rate": 1.0016254178550619e-05,
      "loss": 0.476,
      "mean_token_accuracy": 0.8515755534172058,
      "num_tokens": 936473013.0,
      "step": 29342
    },
    {
      "entropy": 0.5872911727055907,
      "epoch": 2.6997084548104957,
      "grad_norm": 0.1623915731906891,
      "learning_rate": 1.0013187352408993e-05,
      "loss": 0.5693,
      "mean_token_accuracy": 0.8229823224246502,
      "num_tokens": 936505690.0,
      "step": 29343
    },
    {
      "entropy": 0.5574037684127688,
      "epoch": 2.699800461181937,
      "grad_norm": 0.1629973202943802,
      "learning_rate": 1.0010120526267366e-05,
      "loss": 0.5421,
      "mean_token_accuracy": 0.8276013061404228,
      "num_tokens": 936537824.0,
      "step": 29344
    },
    {
      "entropy": 0.6027119942009449,
      "epoch": 2.699892467553378,
      "grad_norm": 0.16948752105236053,
      "learning_rate": 1.000705370012574e-05,
      "loss": 0.597,
      "mean_token_accuracy": 0.8108774833381176,
      "num_tokens": 936570137.0,
      "step": 29345
    },
    {
      "entropy": 0.5648125521838665,
      "epoch": 2.699984473924819,
      "grad_norm": 0.1640758365392685,
      "learning_rate": 1.0003986873984114e-05,
      "loss": 0.5619,
      "mean_token_accuracy": 0.8234650753438473,
      "num_tokens": 936602500.0,
      "step": 29346
    },
    {
      "entropy": 0.5782465906813741,
      "epoch": 2.7000764802962607,
      "grad_norm": 0.16841883957386017,
      "learning_rate": 1.0000920047842488e-05,
      "loss": 0.57,
      "mean_token_accuracy": 0.8189262337982655,
      "num_tokens": 936634270.0,
      "step": 29347
    },
    {
      "entropy": 0.5019076853059232,
      "epoch": 2.700168486667702,
      "grad_norm": 0.14939065277576447,
      "learning_rate": 9.997853221700861e-06,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8479167073965073,
      "num_tokens": 936666864.0,
      "step": 29348
    },
    {
      "entropy": 0.5524340188130736,
      "epoch": 2.700260493039143,
      "grad_norm": 0.15751487016677856,
      "learning_rate": 9.994786395559236e-06,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8326072283089161,
      "num_tokens": 936698780.0,
      "step": 29349
    },
    {
      "entropy": 0.49350841250270605,
      "epoch": 2.700352499410584,
      "grad_norm": 0.15732933580875397,
      "learning_rate": 9.991719569417609e-06,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8489820808172226,
      "num_tokens": 936730882.0,
      "step": 29350
    },
    {
      "entropy": 0.582108081318438,
      "epoch": 2.700444505782025,
      "grad_norm": 0.170335590839386,
      "learning_rate": 9.988652743275984e-06,
      "loss": 0.5708,
      "mean_token_accuracy": 0.8227995373308659,
      "num_tokens": 936763195.0,
      "step": 29351
    },
    {
      "entropy": 0.5442016310989857,
      "epoch": 2.700536512153467,
      "grad_norm": 0.16261053085327148,
      "learning_rate": 9.985585917134358e-06,
      "loss": 0.5202,
      "mean_token_accuracy": 0.832390408962965,
      "num_tokens": 936794982.0,
      "step": 29352
    },
    {
      "entropy": 0.5241976389661431,
      "epoch": 2.700628518524908,
      "grad_norm": 0.1617092788219452,
      "learning_rate": 9.982519090992733e-06,
      "loss": 0.5091,
      "mean_token_accuracy": 0.837374821305275,
      "num_tokens": 936827316.0,
      "step": 29353
    },
    {
      "entropy": 0.6319542303681374,
      "epoch": 2.700720524896349,
      "grad_norm": 0.17153164744377136,
      "learning_rate": 9.979452264851106e-06,
      "loss": 0.6218,
      "mean_token_accuracy": 0.804300382733345,
      "num_tokens": 936859096.0,
      "step": 29354
    },
    {
      "entropy": 0.5125544313341379,
      "epoch": 2.70081253126779,
      "grad_norm": 0.16197149455547333,
      "learning_rate": 9.97638543870948e-06,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8370142132043839,
      "num_tokens": 936890835.0,
      "step": 29355
    },
    {
      "entropy": 0.5953981054481119,
      "epoch": 2.7009045376392313,
      "grad_norm": 0.1647098958492279,
      "learning_rate": 9.973318612567855e-06,
      "loss": 0.5804,
      "mean_token_accuracy": 0.8195687346160412,
      "num_tokens": 936923114.0,
      "step": 29356
    },
    {
      "entropy": 0.6408691201359034,
      "epoch": 2.700996544010673,
      "grad_norm": 0.17704324424266815,
      "learning_rate": 9.970251786426228e-06,
      "loss": 0.6318,
      "mean_token_accuracy": 0.8066873624920845,
      "num_tokens": 936955490.0,
      "step": 29357
    },
    {
      "entropy": 0.45240854751318693,
      "epoch": 2.701088550382114,
      "grad_norm": 0.1564779281616211,
      "learning_rate": 9.967184960284603e-06,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8604023233056068,
      "num_tokens": 936987196.0,
      "step": 29358
    },
    {
      "entropy": 0.48583093378692865,
      "epoch": 2.701180556753555,
      "grad_norm": 0.15944457054138184,
      "learning_rate": 9.964118134142976e-06,
      "loss": 0.4841,
      "mean_token_accuracy": 0.8456533700227737,
      "num_tokens": 937019645.0,
      "step": 29359
    },
    {
      "entropy": 0.6104130689054728,
      "epoch": 2.7012725631249963,
      "grad_norm": 0.1719786524772644,
      "learning_rate": 9.96105130800135e-06,
      "loss": 0.6046,
      "mean_token_accuracy": 0.8114256896078587,
      "num_tokens": 937050804.0,
      "step": 29360
    },
    {
      "entropy": 0.4498669309541583,
      "epoch": 2.7013645694964374,
      "grad_norm": 0.14553077518939972,
      "learning_rate": 9.957984481859723e-06,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8582300879061222,
      "num_tokens": 937083241.0,
      "step": 29361
    },
    {
      "entropy": 0.5460790442302823,
      "epoch": 2.701456575867879,
      "grad_norm": 0.1605178266763687,
      "learning_rate": 9.954917655718098e-06,
      "loss": 0.5367,
      "mean_token_accuracy": 0.8331321999430656,
      "num_tokens": 937115850.0,
      "step": 29362
    },
    {
      "entropy": 0.6025282419286668,
      "epoch": 2.70154858223932,
      "grad_norm": 0.1671980321407318,
      "learning_rate": 9.951850829576471e-06,
      "loss": 0.5874,
      "mean_token_accuracy": 0.8180631548166275,
      "num_tokens": 937148502.0,
      "step": 29363
    },
    {
      "entropy": 0.4540129570523277,
      "epoch": 2.7016405886107613,
      "grad_norm": 0.14830252528190613,
      "learning_rate": 9.948784003434846e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8615332432091236,
      "num_tokens": 937180334.0,
      "step": 29364
    },
    {
      "entropy": 0.5438662255182862,
      "epoch": 2.7017325949822024,
      "grad_norm": 0.16113518178462982,
      "learning_rate": 9.945717177293219e-06,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8298593126237392,
      "num_tokens": 937212290.0,
      "step": 29365
    },
    {
      "entropy": 0.5354088172316551,
      "epoch": 2.7018246013536436,
      "grad_norm": 0.1616266518831253,
      "learning_rate": 9.942650351151593e-06,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8349455334246159,
      "num_tokens": 937243759.0,
      "step": 29366
    },
    {
      "entropy": 0.553856423124671,
      "epoch": 2.701916607725085,
      "grad_norm": 0.15925364196300507,
      "learning_rate": 9.939583525009968e-06,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8291758708655834,
      "num_tokens": 937275888.0,
      "step": 29367
    },
    {
      "entropy": 0.5515732131898403,
      "epoch": 2.7020086140965263,
      "grad_norm": 0.16218259930610657,
      "learning_rate": 9.936516698868342e-06,
      "loss": 0.5426,
      "mean_token_accuracy": 0.8274191431701183,
      "num_tokens": 937308238.0,
      "step": 29368
    },
    {
      "entropy": 0.477737913839519,
      "epoch": 2.7021006204679674,
      "grad_norm": 0.15401750802993774,
      "learning_rate": 9.933449872726715e-06,
      "loss": 0.4729,
      "mean_token_accuracy": 0.8534387350082397,
      "num_tokens": 937340326.0,
      "step": 29369
    },
    {
      "entropy": 0.5286146439611912,
      "epoch": 2.7021926268394085,
      "grad_norm": 0.16661345958709717,
      "learning_rate": 9.93038304658509e-06,
      "loss": 0.5135,
      "mean_token_accuracy": 0.8374861292541027,
      "num_tokens": 937372196.0,
      "step": 29370
    },
    {
      "entropy": 0.5245604193769395,
      "epoch": 2.7022846332108497,
      "grad_norm": 0.16082888841629028,
      "learning_rate": 9.927316220443465e-06,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8383408673107624,
      "num_tokens": 937403873.0,
      "step": 29371
    },
    {
      "entropy": 0.5551385600119829,
      "epoch": 2.7023766395822912,
      "grad_norm": 0.16664505004882812,
      "learning_rate": 9.924249394301838e-06,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8315611369907856,
      "num_tokens": 937436086.0,
      "step": 29372
    },
    {
      "entropy": 0.5501063913106918,
      "epoch": 2.7024686459537324,
      "grad_norm": 0.16222886741161346,
      "learning_rate": 9.921182568160212e-06,
      "loss": 0.5438,
      "mean_token_accuracy": 0.8316133320331573,
      "num_tokens": 937468854.0,
      "step": 29373
    },
    {
      "entropy": 0.5024053398519754,
      "epoch": 2.7025606523251735,
      "grad_norm": 0.1567261964082718,
      "learning_rate": 9.918115742018585e-06,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8411301225423813,
      "num_tokens": 937500825.0,
      "step": 29374
    },
    {
      "entropy": 0.5207972470670938,
      "epoch": 2.7026526586966146,
      "grad_norm": 0.15516284108161926,
      "learning_rate": 9.91504891587696e-06,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8370453454554081,
      "num_tokens": 937533265.0,
      "step": 29375
    },
    {
      "entropy": 0.4299504575319588,
      "epoch": 2.7027446650680558,
      "grad_norm": 0.1483698934316635,
      "learning_rate": 9.911982089735333e-06,
      "loss": 0.4197,
      "mean_token_accuracy": 0.8679557293653488,
      "num_tokens": 937565340.0,
      "step": 29376
    },
    {
      "entropy": 0.5052733849734068,
      "epoch": 2.7028366714394974,
      "grad_norm": 0.15659210085868835,
      "learning_rate": 9.908915263593708e-06,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8409812748432159,
      "num_tokens": 937597653.0,
      "step": 29377
    },
    {
      "entropy": 0.4234911169623956,
      "epoch": 2.7029286778109385,
      "grad_norm": 0.15037141740322113,
      "learning_rate": 9.90584843745208e-06,
      "loss": 0.4042,
      "mean_token_accuracy": 0.8686421252787113,
      "num_tokens": 937629732.0,
      "step": 29378
    },
    {
      "entropy": 0.4873985331505537,
      "epoch": 2.7030206841823796,
      "grad_norm": 0.15745662152767181,
      "learning_rate": 9.902781611310455e-06,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8468092083930969,
      "num_tokens": 937661799.0,
      "step": 29379
    },
    {
      "entropy": 0.5100061446428299,
      "epoch": 2.7031126905538208,
      "grad_norm": 0.15508915483951569,
      "learning_rate": 9.899714785168828e-06,
      "loss": 0.492,
      "mean_token_accuracy": 0.8427337519824505,
      "num_tokens": 937694223.0,
      "step": 29380
    },
    {
      "entropy": 0.4590681530535221,
      "epoch": 2.703204696925262,
      "grad_norm": 0.15507547557353973,
      "learning_rate": 9.896647959027203e-06,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8528527729213238,
      "num_tokens": 937725766.0,
      "step": 29381
    },
    {
      "entropy": 0.6861466346308589,
      "epoch": 2.7032967032967035,
      "grad_norm": 0.1772272288799286,
      "learning_rate": 9.893581132885577e-06,
      "loss": 0.6789,
      "mean_token_accuracy": 0.7886240184307098,
      "num_tokens": 937757721.0,
      "step": 29382
    },
    {
      "entropy": 0.5384507086127996,
      "epoch": 2.7033887096681446,
      "grad_norm": 0.16690897941589355,
      "learning_rate": 9.890514306743952e-06,
      "loss": 0.5291,
      "mean_token_accuracy": 0.8327729664742947,
      "num_tokens": 937789924.0,
      "step": 29383
    },
    {
      "entropy": 0.48587938002310693,
      "epoch": 2.7034807160395857,
      "grad_norm": 0.15432921051979065,
      "learning_rate": 9.887447480602325e-06,
      "loss": 0.471,
      "mean_token_accuracy": 0.8491583541035652,
      "num_tokens": 937822168.0,
      "step": 29384
    },
    {
      "entropy": 0.4909462258219719,
      "epoch": 2.703572722411027,
      "grad_norm": 0.15527735650539398,
      "learning_rate": 9.8843806544607e-06,
      "loss": 0.4863,
      "mean_token_accuracy": 0.8478978909552097,
      "num_tokens": 937853705.0,
      "step": 29385
    },
    {
      "entropy": 0.5231052367016673,
      "epoch": 2.703664728782468,
      "grad_norm": 0.1597105860710144,
      "learning_rate": 9.881313828319074e-06,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8361719362437725,
      "num_tokens": 937885931.0,
      "step": 29386
    },
    {
      "entropy": 0.5375697854906321,
      "epoch": 2.7037567351539096,
      "grad_norm": 0.1643153876066208,
      "learning_rate": 9.878247002177447e-06,
      "loss": 0.5307,
      "mean_token_accuracy": 0.8302570693194866,
      "num_tokens": 937918128.0,
      "step": 29387
    },
    {
      "entropy": 0.5565922930836678,
      "epoch": 2.7038487415253507,
      "grad_norm": 0.1636824905872345,
      "learning_rate": 9.875180176035822e-06,
      "loss": 0.5301,
      "mean_token_accuracy": 0.8309844955801964,
      "num_tokens": 937949587.0,
      "step": 29388
    },
    {
      "entropy": 0.49582137912511826,
      "epoch": 2.703940747896792,
      "grad_norm": 0.15249787271022797,
      "learning_rate": 9.872113349894195e-06,
      "loss": 0.4712,
      "mean_token_accuracy": 0.8462495170533657,
      "num_tokens": 937982020.0,
      "step": 29389
    },
    {
      "entropy": 0.5770727219060063,
      "epoch": 2.704032754268233,
      "grad_norm": 0.1633729338645935,
      "learning_rate": 9.86904652375257e-06,
      "loss": 0.5599,
      "mean_token_accuracy": 0.8251688033342361,
      "num_tokens": 938013793.0,
      "step": 29390
    },
    {
      "entropy": 0.5003590872511268,
      "epoch": 2.704124760639674,
      "grad_norm": 0.15249069035053253,
      "learning_rate": 9.865979697610942e-06,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8434446454048157,
      "num_tokens": 938046300.0,
      "step": 29391
    },
    {
      "entropy": 0.6516155544668436,
      "epoch": 2.7042167670111157,
      "grad_norm": 0.17720083892345428,
      "learning_rate": 9.862912871469317e-06,
      "loss": 0.6404,
      "mean_token_accuracy": 0.7989145219326019,
      "num_tokens": 938077542.0,
      "step": 29392
    },
    {
      "entropy": 0.4613434877246618,
      "epoch": 2.704308773382557,
      "grad_norm": 0.1497536599636078,
      "learning_rate": 9.85984604532769e-06,
      "loss": 0.4403,
      "mean_token_accuracy": 0.8579260595142841,
      "num_tokens": 938108792.0,
      "step": 29393
    },
    {
      "entropy": 0.5079781468957663,
      "epoch": 2.704400779753998,
      "grad_norm": 0.16309358179569244,
      "learning_rate": 9.856779219186065e-06,
      "loss": 0.4948,
      "mean_token_accuracy": 0.8436512313783169,
      "num_tokens": 938140311.0,
      "step": 29394
    },
    {
      "entropy": 0.5620687324553728,
      "epoch": 2.704492786125439,
      "grad_norm": 0.16218163073062897,
      "learning_rate": 9.853712393044438e-06,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8268061764538288,
      "num_tokens": 938172385.0,
      "step": 29395
    },
    {
      "entropy": 0.5631493758410215,
      "epoch": 2.70458479249688,
      "grad_norm": 0.16778185963630676,
      "learning_rate": 9.850645566902812e-06,
      "loss": 0.5566,
      "mean_token_accuracy": 0.8261303193867207,
      "num_tokens": 938204306.0,
      "step": 29396
    },
    {
      "entropy": 0.5508441627025604,
      "epoch": 2.704676798868322,
      "grad_norm": 0.16050627827644348,
      "learning_rate": 9.847578740761187e-06,
      "loss": 0.5394,
      "mean_token_accuracy": 0.8313274048268795,
      "num_tokens": 938236533.0,
      "step": 29397
    },
    {
      "entropy": 0.4759468100965023,
      "epoch": 2.704768805239763,
      "grad_norm": 0.15325377881526947,
      "learning_rate": 9.844511914619562e-06,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8558849841356277,
      "num_tokens": 938267797.0,
      "step": 29398
    },
    {
      "entropy": 0.5543584958650172,
      "epoch": 2.704860811611204,
      "grad_norm": 0.16013149917125702,
      "learning_rate": 9.841445088477935e-06,
      "loss": 0.5387,
      "mean_token_accuracy": 0.8285940922796726,
      "num_tokens": 938299916.0,
      "step": 29399
    },
    {
      "entropy": 0.465151762124151,
      "epoch": 2.704952817982645,
      "grad_norm": 0.14346881210803986,
      "learning_rate": 9.83837826233631e-06,
      "loss": 0.459,
      "mean_token_accuracy": 0.8548204302787781,
      "num_tokens": 938331995.0,
      "step": 29400
    },
    {
      "entropy": 0.46698769088834524,
      "epoch": 2.7050448243540863,
      "grad_norm": 0.15566866099834442,
      "learning_rate": 9.835311436194684e-06,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8551886267960072,
      "num_tokens": 938363604.0,
      "step": 29401
    },
    {
      "entropy": 0.43752288969699293,
      "epoch": 2.705136830725528,
      "grad_norm": 0.14855535328388214,
      "learning_rate": 9.832244610053057e-06,
      "loss": 0.4178,
      "mean_token_accuracy": 0.8652308732271194,
      "num_tokens": 938395659.0,
      "step": 29402
    },
    {
      "entropy": 0.4888839013874531,
      "epoch": 2.705228837096969,
      "grad_norm": 0.15580210089683533,
      "learning_rate": 9.829177783911431e-06,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8469050377607346,
      "num_tokens": 938427377.0,
      "step": 29403
    },
    {
      "entropy": 0.35514259489718825,
      "epoch": 2.70532084346841,
      "grad_norm": 0.13633719086647034,
      "learning_rate": 9.826110957769804e-06,
      "loss": 0.3377,
      "mean_token_accuracy": 0.8915996141731739,
      "num_tokens": 938459665.0,
      "step": 29404
    },
    {
      "entropy": 0.47176229301840067,
      "epoch": 2.7054128498398513,
      "grad_norm": 0.14863130450248718,
      "learning_rate": 9.823044131628179e-06,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8543741703033447,
      "num_tokens": 938491665.0,
      "step": 29405
    },
    {
      "entropy": 0.5612962665036321,
      "epoch": 2.7055048562112924,
      "grad_norm": 0.1677524298429489,
      "learning_rate": 9.819977305486552e-06,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8312431983649731,
      "num_tokens": 938523957.0,
      "step": 29406
    },
    {
      "entropy": 0.5797065868973732,
      "epoch": 2.705596862582734,
      "grad_norm": 0.16639162600040436,
      "learning_rate": 9.816910479344927e-06,
      "loss": 0.5668,
      "mean_token_accuracy": 0.818696990609169,
      "num_tokens": 938555800.0,
      "step": 29407
    },
    {
      "entropy": 0.5321558655705303,
      "epoch": 2.705688868954175,
      "grad_norm": 0.16388115286827087,
      "learning_rate": 9.8138436532033e-06,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8324672468006611,
      "num_tokens": 938587436.0,
      "step": 29408
    },
    {
      "entropy": 0.4656363414833322,
      "epoch": 2.7057808753256163,
      "grad_norm": 0.15690837800502777,
      "learning_rate": 9.810776827061674e-06,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8555343486368656,
      "num_tokens": 938618459.0,
      "step": 29409
    },
    {
      "entropy": 0.5708893472328782,
      "epoch": 2.7058728816970574,
      "grad_norm": 0.16330111026763916,
      "learning_rate": 9.807710000920047e-06,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8261488229036331,
      "num_tokens": 938650176.0,
      "step": 29410
    },
    {
      "entropy": 0.4382666046731174,
      "epoch": 2.7059648880684986,
      "grad_norm": 0.14505666494369507,
      "learning_rate": 9.804643174778422e-06,
      "loss": 0.43,
      "mean_token_accuracy": 0.8629513867199421,
      "num_tokens": 938682861.0,
      "step": 29411
    },
    {
      "entropy": 0.5752275623381138,
      "epoch": 2.70605689443994,
      "grad_norm": 0.160879448056221,
      "learning_rate": 9.801576348636796e-06,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8252692557871342,
      "num_tokens": 938714445.0,
      "step": 29412
    },
    {
      "entropy": 0.5175880566239357,
      "epoch": 2.7061489008113813,
      "grad_norm": 0.15693826973438263,
      "learning_rate": 9.798509522495171e-06,
      "loss": 0.4999,
      "mean_token_accuracy": 0.8394310548901558,
      "num_tokens": 938746602.0,
      "step": 29413
    },
    {
      "entropy": 0.4699143124744296,
      "epoch": 2.7062409071828224,
      "grad_norm": 0.15468525886535645,
      "learning_rate": 9.795442696353544e-06,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8524227291345596,
      "num_tokens": 938779234.0,
      "step": 29414
    },
    {
      "entropy": 0.4286231347359717,
      "epoch": 2.7063329135542635,
      "grad_norm": 0.14524734020233154,
      "learning_rate": 9.792375870211919e-06,
      "loss": 0.4097,
      "mean_token_accuracy": 0.8679923228919506,
      "num_tokens": 938811141.0,
      "step": 29415
    },
    {
      "entropy": 0.5766457635909319,
      "epoch": 2.7064249199257047,
      "grad_norm": 0.17183011770248413,
      "learning_rate": 9.789309044070293e-06,
      "loss": 0.5681,
      "mean_token_accuracy": 0.8195791356265545,
      "num_tokens": 938842398.0,
      "step": 29416
    },
    {
      "entropy": 0.5697886997368187,
      "epoch": 2.7065169262971462,
      "grad_norm": 0.1634502410888672,
      "learning_rate": 9.786242217928666e-06,
      "loss": 0.557,
      "mean_token_accuracy": 0.8270718194544315,
      "num_tokens": 938873889.0,
      "step": 29417
    },
    {
      "entropy": 0.4184475615620613,
      "epoch": 2.7066089326685874,
      "grad_norm": 0.14318981766700745,
      "learning_rate": 9.783175391787041e-06,
      "loss": 0.4036,
      "mean_token_accuracy": 0.8711049556732178,
      "num_tokens": 938906226.0,
      "step": 29418
    },
    {
      "entropy": 0.5520606897771358,
      "epoch": 2.7067009390400285,
      "grad_norm": 0.16796471178531647,
      "learning_rate": 9.780108565645414e-06,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8273700997233391,
      "num_tokens": 938938111.0,
      "step": 29419
    },
    {
      "entropy": 0.5111600207164884,
      "epoch": 2.7067929454114696,
      "grad_norm": 0.15550380945205688,
      "learning_rate": 9.777041739503789e-06,
      "loss": 0.493,
      "mean_token_accuracy": 0.8434503376483917,
      "num_tokens": 938970642.0,
      "step": 29420
    },
    {
      "entropy": 0.4784617021214217,
      "epoch": 2.7068849517829108,
      "grad_norm": 0.15212011337280273,
      "learning_rate": 9.773974913362162e-06,
      "loss": 0.465,
      "mean_token_accuracy": 0.854128997772932,
      "num_tokens": 939003054.0,
      "step": 29421
    },
    {
      "entropy": 0.571298761293292,
      "epoch": 2.7069769581543524,
      "grad_norm": 0.16017569601535797,
      "learning_rate": 9.770908087220536e-06,
      "loss": 0.554,
      "mean_token_accuracy": 0.8272219300270081,
      "num_tokens": 939035124.0,
      "step": 29422
    },
    {
      "entropy": 0.41312379331793636,
      "epoch": 2.7070689645257935,
      "grad_norm": 0.1452677994966507,
      "learning_rate": 9.767841261078909e-06,
      "loss": 0.3995,
      "mean_token_accuracy": 0.8761035054922104,
      "num_tokens": 939067433.0,
      "step": 29423
    },
    {
      "entropy": 0.5276785865426064,
      "epoch": 2.7071609708972346,
      "grad_norm": 0.16054177284240723,
      "learning_rate": 9.764774434937284e-06,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8353130184113979,
      "num_tokens": 939099269.0,
      "step": 29424
    },
    {
      "entropy": 0.4896886860951781,
      "epoch": 2.7072529772686758,
      "grad_norm": 0.1540607511997223,
      "learning_rate": 9.761707608795657e-06,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8473938032984734,
      "num_tokens": 939131826.0,
      "step": 29425
    },
    {
      "entropy": 0.557830547913909,
      "epoch": 2.707344983640117,
      "grad_norm": 0.16258573532104492,
      "learning_rate": 9.758640782654031e-06,
      "loss": 0.545,
      "mean_token_accuracy": 0.828764520585537,
      "num_tokens": 939164067.0,
      "step": 29426
    },
    {
      "entropy": 0.5077865980565548,
      "epoch": 2.7074369900115585,
      "grad_norm": 0.15440073609352112,
      "learning_rate": 9.755573956512406e-06,
      "loss": 0.4946,
      "mean_token_accuracy": 0.8444103635847569,
      "num_tokens": 939196700.0,
      "step": 29427
    },
    {
      "entropy": 0.5684558711946011,
      "epoch": 2.7075289963829996,
      "grad_norm": 0.16620001196861267,
      "learning_rate": 9.752507130370779e-06,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8250827975571156,
      "num_tokens": 939227735.0,
      "step": 29428
    },
    {
      "entropy": 0.5109077161177993,
      "epoch": 2.7076210027544407,
      "grad_norm": 0.16117508709430695,
      "learning_rate": 9.749440304229154e-06,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8450235724449158,
      "num_tokens": 939258840.0,
      "step": 29429
    },
    {
      "entropy": 0.5733043737709522,
      "epoch": 2.707713009125882,
      "grad_norm": 0.16416355967521667,
      "learning_rate": 9.746373478087528e-06,
      "loss": 0.562,
      "mean_token_accuracy": 0.824303425848484,
      "num_tokens": 939290636.0,
      "step": 29430
    },
    {
      "entropy": 0.47080767154693604,
      "epoch": 2.707805015497323,
      "grad_norm": 0.15052123367786407,
      "learning_rate": 9.743306651945903e-06,
      "loss": 0.4588,
      "mean_token_accuracy": 0.8532594367861748,
      "num_tokens": 939322966.0,
      "step": 29431
    },
    {
      "entropy": 0.5255593103356659,
      "epoch": 2.7078970218687646,
      "grad_norm": 0.15839508175849915,
      "learning_rate": 9.740239825804276e-06,
      "loss": 0.5018,
      "mean_token_accuracy": 0.8348117135465145,
      "num_tokens": 939353344.0,
      "step": 29432
    },
    {
      "entropy": 0.43154837749898434,
      "epoch": 2.7079890282402057,
      "grad_norm": 0.14521931111812592,
      "learning_rate": 9.73717299966265e-06,
      "loss": 0.4251,
      "mean_token_accuracy": 0.8665536306798458,
      "num_tokens": 939385729.0,
      "step": 29433
    },
    {
      "entropy": 0.4863505130633712,
      "epoch": 2.708081034611647,
      "grad_norm": 0.15736228227615356,
      "learning_rate": 9.734106173521023e-06,
      "loss": 0.4765,
      "mean_token_accuracy": 0.8495798408985138,
      "num_tokens": 939417711.0,
      "step": 29434
    },
    {
      "entropy": 0.5763491718098521,
      "epoch": 2.708173040983088,
      "grad_norm": 0.16466297209262848,
      "learning_rate": 9.731039347379398e-06,
      "loss": 0.5599,
      "mean_token_accuracy": 0.8242186717689037,
      "num_tokens": 939449307.0,
      "step": 29435
    },
    {
      "entropy": 0.607537966221571,
      "epoch": 2.708265047354529,
      "grad_norm": 0.17340034246444702,
      "learning_rate": 9.727972521237771e-06,
      "loss": 0.5962,
      "mean_token_accuracy": 0.8112117946147919,
      "num_tokens": 939481057.0,
      "step": 29436
    },
    {
      "entropy": 0.48483996372669935,
      "epoch": 2.7083570537259707,
      "grad_norm": 0.1575736701488495,
      "learning_rate": 9.724905695096146e-06,
      "loss": 0.4896,
      "mean_token_accuracy": 0.8446231670677662,
      "num_tokens": 939513015.0,
      "step": 29437
    },
    {
      "entropy": 0.46433541364967823,
      "epoch": 2.708449060097412,
      "grad_norm": 0.1545678973197937,
      "learning_rate": 9.721838868954519e-06,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8543913550674915,
      "num_tokens": 939544212.0,
      "step": 29438
    },
    {
      "entropy": 0.5254005650058389,
      "epoch": 2.708541066468853,
      "grad_norm": 0.15690162777900696,
      "learning_rate": 9.718772042812893e-06,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8365977965295315,
      "num_tokens": 939576845.0,
      "step": 29439
    },
    {
      "entropy": 0.3672088396269828,
      "epoch": 2.708633072840294,
      "grad_norm": 0.1367589384317398,
      "learning_rate": 9.715705216671266e-06,
      "loss": 0.357,
      "mean_token_accuracy": 0.8853903152048588,
      "num_tokens": 939609384.0,
      "step": 29440
    },
    {
      "entropy": 0.4853379875421524,
      "epoch": 2.708725079211735,
      "grad_norm": 0.15914711356163025,
      "learning_rate": 9.712638390529641e-06,
      "loss": 0.4845,
      "mean_token_accuracy": 0.8483827374875546,
      "num_tokens": 939641122.0,
      "step": 29441
    },
    {
      "entropy": 0.5711753210052848,
      "epoch": 2.708817085583177,
      "grad_norm": 0.16455940902233124,
      "learning_rate": 9.709571564388016e-06,
      "loss": 0.554,
      "mean_token_accuracy": 0.8265811391174793,
      "num_tokens": 939673545.0,
      "step": 29442
    },
    {
      "entropy": 0.5637918524444103,
      "epoch": 2.708909091954618,
      "grad_norm": 0.16659268736839294,
      "learning_rate": 9.706504738246389e-06,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8259214051067829,
      "num_tokens": 939705579.0,
      "step": 29443
    },
    {
      "entropy": 0.49789175391197205,
      "epoch": 2.709001098326059,
      "grad_norm": 0.15348005294799805,
      "learning_rate": 9.703437912104763e-06,
      "loss": 0.4945,
      "mean_token_accuracy": 0.8447283394634724,
      "num_tokens": 939737891.0,
      "step": 29444
    },
    {
      "entropy": 0.5009822975844145,
      "epoch": 2.7090931046975,
      "grad_norm": 0.15856173634529114,
      "learning_rate": 9.700371085963138e-06,
      "loss": 0.4891,
      "mean_token_accuracy": 0.84312679246068,
      "num_tokens": 939769981.0,
      "step": 29445
    },
    {
      "entropy": 0.5203333215322345,
      "epoch": 2.7091851110689413,
      "grad_norm": 0.16351637244224548,
      "learning_rate": 9.697304259821512e-06,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8383921571075916,
      "num_tokens": 939801247.0,
      "step": 29446
    },
    {
      "entropy": 0.42796873673796654,
      "epoch": 2.709277117440383,
      "grad_norm": 0.1473221629858017,
      "learning_rate": 9.694237433679885e-06,
      "loss": 0.4156,
      "mean_token_accuracy": 0.8650663308799267,
      "num_tokens": 939833373.0,
      "step": 29447
    },
    {
      "entropy": 0.4377625910565257,
      "epoch": 2.709369123811824,
      "grad_norm": 0.14916308224201202,
      "learning_rate": 9.69117060753826e-06,
      "loss": 0.4412,
      "mean_token_accuracy": 0.8606175929307938,
      "num_tokens": 939865867.0,
      "step": 29448
    },
    {
      "entropy": 0.5729145864024758,
      "epoch": 2.709461130183265,
      "grad_norm": 0.15976038575172424,
      "learning_rate": 9.688103781396633e-06,
      "loss": 0.55,
      "mean_token_accuracy": 0.8242787979543209,
      "num_tokens": 939898192.0,
      "step": 29449
    },
    {
      "entropy": 0.5036901757121086,
      "epoch": 2.7095531365547063,
      "grad_norm": 0.1546286791563034,
      "learning_rate": 9.685036955255008e-06,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8392785936594009,
      "num_tokens": 939930507.0,
      "step": 29450
    },
    {
      "entropy": 0.4300869591534138,
      "epoch": 2.7096451429261474,
      "grad_norm": 0.1472901552915573,
      "learning_rate": 9.68197012911338e-06,
      "loss": 0.4248,
      "mean_token_accuracy": 0.8629915490746498,
      "num_tokens": 939962436.0,
      "step": 29451
    },
    {
      "entropy": 0.5139665454626083,
      "epoch": 2.709737149297589,
      "grad_norm": 0.15648040175437927,
      "learning_rate": 9.678903302971755e-06,
      "loss": 0.5075,
      "mean_token_accuracy": 0.8421732597053051,
      "num_tokens": 939994282.0,
      "step": 29452
    },
    {
      "entropy": 0.4963014516979456,
      "epoch": 2.70982915566903,
      "grad_norm": 0.1609032303094864,
      "learning_rate": 9.675836476830128e-06,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8422653861343861,
      "num_tokens": 940025446.0,
      "step": 29453
    },
    {
      "entropy": 0.43595309369266033,
      "epoch": 2.7099211620404713,
      "grad_norm": 0.15129326283931732,
      "learning_rate": 9.672769650688503e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.8643510155379772,
      "num_tokens": 940056689.0,
      "step": 29454
    },
    {
      "entropy": 0.520951533690095,
      "epoch": 2.7100131684119124,
      "grad_norm": 0.15824109315872192,
      "learning_rate": 9.669702824546876e-06,
      "loss": 0.5042,
      "mean_token_accuracy": 0.8383250422775745,
      "num_tokens": 940088351.0,
      "step": 29455
    },
    {
      "entropy": 0.5071330908685923,
      "epoch": 2.7101051747833536,
      "grad_norm": 0.16029204428195953,
      "learning_rate": 9.66663599840525e-06,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8432639539241791,
      "num_tokens": 940119402.0,
      "step": 29456
    },
    {
      "entropy": 0.4565302263945341,
      "epoch": 2.710197181154795,
      "grad_norm": 0.15240882337093353,
      "learning_rate": 9.663569172263625e-06,
      "loss": 0.4401,
      "mean_token_accuracy": 0.8583897650241852,
      "num_tokens": 940151413.0,
      "step": 29457
    },
    {
      "entropy": 0.4756838120520115,
      "epoch": 2.7102891875262363,
      "grad_norm": 0.15586267411708832,
      "learning_rate": 9.660502346121998e-06,
      "loss": 0.457,
      "mean_token_accuracy": 0.8535886593163013,
      "num_tokens": 940182704.0,
      "step": 29458
    },
    {
      "entropy": 0.5403930461034179,
      "epoch": 2.7103811938976774,
      "grad_norm": 0.1680208295583725,
      "learning_rate": 9.657435519980373e-06,
      "loss": 0.523,
      "mean_token_accuracy": 0.8345128931105137,
      "num_tokens": 940213396.0,
      "step": 29459
    },
    {
      "entropy": 0.5304553527384996,
      "epoch": 2.7104732002691185,
      "grad_norm": 0.15435703098773956,
      "learning_rate": 9.654368693838747e-06,
      "loss": 0.5085,
      "mean_token_accuracy": 0.840153306722641,
      "num_tokens": 940245838.0,
      "step": 29460
    },
    {
      "entropy": 0.4818395311012864,
      "epoch": 2.7105652066405597,
      "grad_norm": 0.15371173620224,
      "learning_rate": 9.651301867697122e-06,
      "loss": 0.4602,
      "mean_token_accuracy": 0.8554067239165306,
      "num_tokens": 940277856.0,
      "step": 29461
    },
    {
      "entropy": 0.5673356857150793,
      "epoch": 2.7106572130120012,
      "grad_norm": 0.16269373893737793,
      "learning_rate": 9.648235041555495e-06,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8220394253730774,
      "num_tokens": 940310452.0,
      "step": 29462
    },
    {
      "entropy": 0.579140611924231,
      "epoch": 2.7107492193834424,
      "grad_norm": 0.16885478794574738,
      "learning_rate": 9.64516821541387e-06,
      "loss": 0.5778,
      "mean_token_accuracy": 0.8188778795301914,
      "num_tokens": 940341652.0,
      "step": 29463
    },
    {
      "entropy": 0.453948883805424,
      "epoch": 2.7108412257548835,
      "grad_norm": 0.15291471779346466,
      "learning_rate": 9.642101389272243e-06,
      "loss": 0.4466,
      "mean_token_accuracy": 0.8579147905111313,
      "num_tokens": 940374160.0,
      "step": 29464
    },
    {
      "entropy": 0.45679567754268646,
      "epoch": 2.7109332321263246,
      "grad_norm": 0.15003487467765808,
      "learning_rate": 9.639034563130617e-06,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8556800968945026,
      "num_tokens": 940405525.0,
      "step": 29465
    },
    {
      "entropy": 0.5194982485845685,
      "epoch": 2.7110252384977658,
      "grad_norm": 0.15723580121994019,
      "learning_rate": 9.63596773698899e-06,
      "loss": 0.506,
      "mean_token_accuracy": 0.8367626704275608,
      "num_tokens": 940437134.0,
      "step": 29466
    },
    {
      "entropy": 0.5619834503158927,
      "epoch": 2.7111172448692074,
      "grad_norm": 0.16230976581573486,
      "learning_rate": 9.632900910847365e-06,
      "loss": 0.5434,
      "mean_token_accuracy": 0.82707279920578,
      "num_tokens": 940469200.0,
      "step": 29467
    },
    {
      "entropy": 0.5053897053003311,
      "epoch": 2.7112092512406485,
      "grad_norm": 0.16113726794719696,
      "learning_rate": 9.629834084705738e-06,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8430382646620274,
      "num_tokens": 940500549.0,
      "step": 29468
    },
    {
      "entropy": 0.4651757264509797,
      "epoch": 2.7113012576120896,
      "grad_norm": 0.14883168041706085,
      "learning_rate": 9.626767258564112e-06,
      "loss": 0.462,
      "mean_token_accuracy": 0.8554973527789116,
      "num_tokens": 940533103.0,
      "step": 29469
    },
    {
      "entropy": 0.46917093778029084,
      "epoch": 2.7113932639835308,
      "grad_norm": 0.14386573433876038,
      "learning_rate": 9.623700432422485e-06,
      "loss": 0.4548,
      "mean_token_accuracy": 0.859482616186142,
      "num_tokens": 940565588.0,
      "step": 29470
    },
    {
      "entropy": 0.5867243511602283,
      "epoch": 2.711485270354972,
      "grad_norm": 0.1689629703760147,
      "learning_rate": 9.62063360628086e-06,
      "loss": 0.5788,
      "mean_token_accuracy": 0.8190339505672455,
      "num_tokens": 940598356.0,
      "step": 29471
    },
    {
      "entropy": 0.625205010175705,
      "epoch": 2.7115772767264135,
      "grad_norm": 0.17049282789230347,
      "learning_rate": 9.617566780139235e-06,
      "loss": 0.6145,
      "mean_token_accuracy": 0.8068715482950211,
      "num_tokens": 940630621.0,
      "step": 29472
    },
    {
      "entropy": 0.5346143897622824,
      "epoch": 2.7116692830978546,
      "grad_norm": 0.16651523113250732,
      "learning_rate": 9.614499953997608e-06,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8310666084289551,
      "num_tokens": 940661973.0,
      "step": 29473
    },
    {
      "entropy": 0.48114430834539235,
      "epoch": 2.7117612894692957,
      "grad_norm": 0.16007710993289948,
      "learning_rate": 9.611433127855982e-06,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8505708687007427,
      "num_tokens": 940693775.0,
      "step": 29474
    },
    {
      "entropy": 0.5090610315091908,
      "epoch": 2.711853295840737,
      "grad_norm": 0.15451683104038239,
      "learning_rate": 9.608366301714357e-06,
      "loss": 0.4866,
      "mean_token_accuracy": 0.842536099255085,
      "num_tokens": 940725067.0,
      "step": 29475
    },
    {
      "entropy": 0.44627729849889874,
      "epoch": 2.711945302212178,
      "grad_norm": 0.14641118049621582,
      "learning_rate": 9.605299475572732e-06,
      "loss": 0.4311,
      "mean_token_accuracy": 0.8604666069149971,
      "num_tokens": 940757197.0,
      "step": 29476
    },
    {
      "entropy": 0.5837078290060163,
      "epoch": 2.7120373085836196,
      "grad_norm": 0.16325725615024567,
      "learning_rate": 9.602232649431105e-06,
      "loss": 0.5629,
      "mean_token_accuracy": 0.8251640237867832,
      "num_tokens": 940789963.0,
      "step": 29477
    },
    {
      "entropy": 0.44595308415591717,
      "epoch": 2.7121293149550607,
      "grad_norm": 0.1492750346660614,
      "learning_rate": 9.59916582328948e-06,
      "loss": 0.4378,
      "mean_token_accuracy": 0.8616411089897156,
      "num_tokens": 940821822.0,
      "step": 29478
    },
    {
      "entropy": 0.5667116101831198,
      "epoch": 2.712221321326502,
      "grad_norm": 0.1638147085905075,
      "learning_rate": 9.596098997147852e-06,
      "loss": 0.5622,
      "mean_token_accuracy": 0.8212725855410099,
      "num_tokens": 940854425.0,
      "step": 29479
    },
    {
      "entropy": 0.46278367657214403,
      "epoch": 2.712313327697943,
      "grad_norm": 0.15870465338230133,
      "learning_rate": 9.593032171006227e-06,
      "loss": 0.4481,
      "mean_token_accuracy": 0.8553985953330994,
      "num_tokens": 940886101.0,
      "step": 29480
    },
    {
      "entropy": 0.5029043778777122,
      "epoch": 2.712405334069384,
      "grad_norm": 0.1567719578742981,
      "learning_rate": 9.5899653448646e-06,
      "loss": 0.4976,
      "mean_token_accuracy": 0.843014795333147,
      "num_tokens": 940918713.0,
      "step": 29481
    },
    {
      "entropy": 0.514163332991302,
      "epoch": 2.7124973404408257,
      "grad_norm": 0.16221195459365845,
      "learning_rate": 9.586898518722974e-06,
      "loss": 0.5055,
      "mean_token_accuracy": 0.8397840708494186,
      "num_tokens": 940950812.0,
      "step": 29482
    },
    {
      "entropy": 0.43724628211930394,
      "epoch": 2.712589346812267,
      "grad_norm": 0.1565748155117035,
      "learning_rate": 9.583831692581347e-06,
      "loss": 0.4389,
      "mean_token_accuracy": 0.8557043261826038,
      "num_tokens": 940982340.0,
      "step": 29483
    },
    {
      "entropy": 0.5420735906809568,
      "epoch": 2.712681353183708,
      "grad_norm": 0.16747403144836426,
      "learning_rate": 9.580764866439722e-06,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8331629671156406,
      "num_tokens": 941014155.0,
      "step": 29484
    },
    {
      "entropy": 0.4624431747943163,
      "epoch": 2.712773359555149,
      "grad_norm": 0.15119031071662903,
      "learning_rate": 9.577698040298095e-06,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8551188074052334,
      "num_tokens": 941045616.0,
      "step": 29485
    },
    {
      "entropy": 0.555675673764199,
      "epoch": 2.7128653659265902,
      "grad_norm": 0.15616805851459503,
      "learning_rate": 9.57463121415647e-06,
      "loss": 0.5386,
      "mean_token_accuracy": 0.8319585397839546,
      "num_tokens": 941078189.0,
      "step": 29486
    },
    {
      "entropy": 0.5368008371442556,
      "epoch": 2.712957372298032,
      "grad_norm": 0.15680216252803802,
      "learning_rate": 9.571564388014844e-06,
      "loss": 0.5301,
      "mean_token_accuracy": 0.8331286199390888,
      "num_tokens": 941110941.0,
      "step": 29487
    },
    {
      "entropy": 0.494827467831783,
      "epoch": 2.713049378669473,
      "grad_norm": 0.156403049826622,
      "learning_rate": 9.568497561873217e-06,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8444091565907001,
      "num_tokens": 941142959.0,
      "step": 29488
    },
    {
      "entropy": 0.487176226451993,
      "epoch": 2.713141385040914,
      "grad_norm": 0.15636087954044342,
      "learning_rate": 9.565430735731592e-06,
      "loss": 0.48,
      "mean_token_accuracy": 0.8476056270301342,
      "num_tokens": 941175058.0,
      "step": 29489
    },
    {
      "entropy": 0.5763474958948791,
      "epoch": 2.713233391412355,
      "grad_norm": 0.16623680293560028,
      "learning_rate": 9.562363909589966e-06,
      "loss": 0.5673,
      "mean_token_accuracy": 0.8227771036326885,
      "num_tokens": 941207521.0,
      "step": 29490
    },
    {
      "entropy": 0.5081065036356449,
      "epoch": 2.7133253977837963,
      "grad_norm": 0.15773585438728333,
      "learning_rate": 9.559297083448341e-06,
      "loss": 0.5069,
      "mean_token_accuracy": 0.8436964824795723,
      "num_tokens": 941239735.0,
      "step": 29491
    },
    {
      "entropy": 0.474165519233793,
      "epoch": 2.713417404155238,
      "grad_norm": 0.15304265916347504,
      "learning_rate": 9.556230257306714e-06,
      "loss": 0.4765,
      "mean_token_accuracy": 0.8493592068552971,
      "num_tokens": 941272211.0,
      "step": 29492
    },
    {
      "entropy": 0.41087288537528366,
      "epoch": 2.713509410526679,
      "grad_norm": 0.14176598191261292,
      "learning_rate": 9.553163431165089e-06,
      "loss": 0.3978,
      "mean_token_accuracy": 0.874092485755682,
      "num_tokens": 941304589.0,
      "step": 29493
    },
    {
      "entropy": 0.5401728320866823,
      "epoch": 2.71360141689812,
      "grad_norm": 0.162898451089859,
      "learning_rate": 9.550096605023462e-06,
      "loss": 0.5286,
      "mean_token_accuracy": 0.8334130682051182,
      "num_tokens": 941337243.0,
      "step": 29494
    },
    {
      "entropy": 0.5644473303109407,
      "epoch": 2.7136934232695613,
      "grad_norm": 0.16374875605106354,
      "learning_rate": 9.547029778881836e-06,
      "loss": 0.5505,
      "mean_token_accuracy": 0.8284657038748264,
      "num_tokens": 941368874.0,
      "step": 29495
    },
    {
      "entropy": 0.5110219726338983,
      "epoch": 2.7137854296410024,
      "grad_norm": 0.16040107607841492,
      "learning_rate": 9.54396295274021e-06,
      "loss": 0.4925,
      "mean_token_accuracy": 0.841670598834753,
      "num_tokens": 941400592.0,
      "step": 29496
    },
    {
      "entropy": 0.5601012473925948,
      "epoch": 2.713877436012444,
      "grad_norm": 0.16122746467590332,
      "learning_rate": 9.540896126598584e-06,
      "loss": 0.5346,
      "mean_token_accuracy": 0.830735981464386,
      "num_tokens": 941432722.0,
      "step": 29497
    },
    {
      "entropy": 0.5917423032224178,
      "epoch": 2.713969442383885,
      "grad_norm": 0.16675250232219696,
      "learning_rate": 9.537829300456957e-06,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8176718056201935,
      "num_tokens": 941464510.0,
      "step": 29498
    },
    {
      "entropy": 0.5512966085225344,
      "epoch": 2.7140614487553263,
      "grad_norm": 0.1591653674840927,
      "learning_rate": 9.534762474315332e-06,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8276122882962227,
      "num_tokens": 941496601.0,
      "step": 29499
    },
    {
      "entropy": 0.4531576024601236,
      "epoch": 2.7141534551267674,
      "grad_norm": 0.14627064764499664,
      "learning_rate": 9.531695648173704e-06,
      "loss": 0.4356,
      "mean_token_accuracy": 0.8602516762912273,
      "num_tokens": 941528713.0,
      "step": 29500
    },
    {
      "entropy": 0.544558972120285,
      "epoch": 2.7142454614982086,
      "grad_norm": 0.1682693511247635,
      "learning_rate": 9.528628822032079e-06,
      "loss": 0.5489,
      "mean_token_accuracy": 0.8284273818135262,
      "num_tokens": 941560864.0,
      "step": 29501
    },
    {
      "entropy": 0.49616995453834534,
      "epoch": 2.71433746786965,
      "grad_norm": 0.15120266377925873,
      "learning_rate": 9.525561995890454e-06,
      "loss": 0.4846,
      "mean_token_accuracy": 0.8440384343266487,
      "num_tokens": 941593218.0,
      "step": 29502
    },
    {
      "entropy": 0.49774342868477106,
      "epoch": 2.7144294742410913,
      "grad_norm": 0.15805427730083466,
      "learning_rate": 9.522495169748827e-06,
      "loss": 0.4882,
      "mean_token_accuracy": 0.843717411160469,
      "num_tokens": 941625309.0,
      "step": 29503
    },
    {
      "entropy": 0.5208279844373465,
      "epoch": 2.7145214806125324,
      "grad_norm": 0.16161228716373444,
      "learning_rate": 9.519428343607201e-06,
      "loss": 0.5107,
      "mean_token_accuracy": 0.8412415310740471,
      "num_tokens": 941656664.0,
      "step": 29504
    },
    {
      "entropy": 0.5599145963788033,
      "epoch": 2.7146134869839735,
      "grad_norm": 0.17367525398731232,
      "learning_rate": 9.516361517465576e-06,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8244989663362503,
      "num_tokens": 941687139.0,
      "step": 29505
    },
    {
      "entropy": 0.5424892473965883,
      "epoch": 2.7147054933554147,
      "grad_norm": 0.1627299189567566,
      "learning_rate": 9.51329469132395e-06,
      "loss": 0.5356,
      "mean_token_accuracy": 0.8296914361417294,
      "num_tokens": 941719526.0,
      "step": 29506
    },
    {
      "entropy": 0.48204999417066574,
      "epoch": 2.7147974997268562,
      "grad_norm": 0.15063752233982086,
      "learning_rate": 9.510227865182324e-06,
      "loss": 0.4688,
      "mean_token_accuracy": 0.8485689796507359,
      "num_tokens": 941751341.0,
      "step": 29507
    },
    {
      "entropy": 0.45657989056780934,
      "epoch": 2.7148895060982974,
      "grad_norm": 0.14914539456367493,
      "learning_rate": 9.507161039040698e-06,
      "loss": 0.4412,
      "mean_token_accuracy": 0.8568937964737415,
      "num_tokens": 941783482.0,
      "step": 29508
    },
    {
      "entropy": 0.46702362853102386,
      "epoch": 2.7149815124697385,
      "grad_norm": 0.1557740718126297,
      "learning_rate": 9.504094212899071e-06,
      "loss": 0.4669,
      "mean_token_accuracy": 0.8492706157267094,
      "num_tokens": 941815698.0,
      "step": 29509
    },
    {
      "entropy": 0.44602674152702093,
      "epoch": 2.7150735188411796,
      "grad_norm": 0.14620231091976166,
      "learning_rate": 9.501027386757446e-06,
      "loss": 0.4472,
      "mean_token_accuracy": 0.8600101806223392,
      "num_tokens": 941848205.0,
      "step": 29510
    },
    {
      "entropy": 0.5066583938896656,
      "epoch": 2.715165525212621,
      "grad_norm": 0.16547957062721252,
      "learning_rate": 9.497960560615819e-06,
      "loss": 0.4961,
      "mean_token_accuracy": 0.839994840323925,
      "num_tokens": 941879487.0,
      "step": 29511
    },
    {
      "entropy": 0.6098740380257368,
      "epoch": 2.7152575315840624,
      "grad_norm": 0.1694057285785675,
      "learning_rate": 9.494893734474193e-06,
      "loss": 0.598,
      "mean_token_accuracy": 0.8154836520552635,
      "num_tokens": 941911538.0,
      "step": 29512
    },
    {
      "entropy": 0.5227645514532924,
      "epoch": 2.7153495379555035,
      "grad_norm": 0.16043415665626526,
      "learning_rate": 9.491826908332566e-06,
      "loss": 0.512,
      "mean_token_accuracy": 0.84015978500247,
      "num_tokens": 941943773.0,
      "step": 29513
    },
    {
      "entropy": 0.47369934571906924,
      "epoch": 2.7154415443269446,
      "grad_norm": 0.14708305895328522,
      "learning_rate": 9.488760082190941e-06,
      "loss": 0.4524,
      "mean_token_accuracy": 0.8558394722640514,
      "num_tokens": 941976067.0,
      "step": 29514
    },
    {
      "entropy": 0.5469330567866564,
      "epoch": 2.7155335506983858,
      "grad_norm": 0.15744276344776154,
      "learning_rate": 9.485693256049314e-06,
      "loss": 0.5244,
      "mean_token_accuracy": 0.8328103981912136,
      "num_tokens": 942008805.0,
      "step": 29515
    },
    {
      "entropy": 0.5345748402178288,
      "epoch": 2.715625557069827,
      "grad_norm": 0.15962496399879456,
      "learning_rate": 9.482626429907689e-06,
      "loss": 0.525,
      "mean_token_accuracy": 0.832987803965807,
      "num_tokens": 942040888.0,
      "step": 29516
    },
    {
      "entropy": 0.4384117814479396,
      "epoch": 2.7157175634412685,
      "grad_norm": 0.14387518167495728,
      "learning_rate": 9.479559603766063e-06,
      "loss": 0.4222,
      "mean_token_accuracy": 0.8671610727906227,
      "num_tokens": 942073623.0,
      "step": 29517
    },
    {
      "entropy": 0.5424384325742722,
      "epoch": 2.7158095698127096,
      "grad_norm": 0.1689295619726181,
      "learning_rate": 9.476492777624436e-06,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8277655579149723,
      "num_tokens": 942104737.0,
      "step": 29518
    },
    {
      "entropy": 0.4583223136141896,
      "epoch": 2.7159015761841507,
      "grad_norm": 0.1451146900653839,
      "learning_rate": 9.473425951482811e-06,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8571815304458141,
      "num_tokens": 942137057.0,
      "step": 29519
    },
    {
      "entropy": 0.5224432507529855,
      "epoch": 2.715993582555592,
      "grad_norm": 0.15666259825229645,
      "learning_rate": 9.470359125341186e-06,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8414540514349937,
      "num_tokens": 942169271.0,
      "step": 29520
    },
    {
      "entropy": 0.5231816272716969,
      "epoch": 2.716085588927033,
      "grad_norm": 0.16425912082195282,
      "learning_rate": 9.46729229919956e-06,
      "loss": 0.5169,
      "mean_token_accuracy": 0.8334907814860344,
      "num_tokens": 942201179.0,
      "step": 29521
    },
    {
      "entropy": 0.4906005463562906,
      "epoch": 2.7161775952984746,
      "grad_norm": 0.15223179757595062,
      "learning_rate": 9.464225473057933e-06,
      "loss": 0.4741,
      "mean_token_accuracy": 0.8489337489008904,
      "num_tokens": 942233053.0,
      "step": 29522
    },
    {
      "entropy": 0.4826858504675329,
      "epoch": 2.7162696016699157,
      "grad_norm": 0.1527739018201828,
      "learning_rate": 9.461158646916308e-06,
      "loss": 0.4634,
      "mean_token_accuracy": 0.8548373468220234,
      "num_tokens": 942264858.0,
      "step": 29523
    },
    {
      "entropy": 0.5523564352188259,
      "epoch": 2.716361608041357,
      "grad_norm": 0.16184939444065094,
      "learning_rate": 9.45809182077468e-06,
      "loss": 0.5395,
      "mean_token_accuracy": 0.8357576429843903,
      "num_tokens": 942296917.0,
      "step": 29524
    },
    {
      "entropy": 0.6424685698002577,
      "epoch": 2.716453614412798,
      "grad_norm": 0.17792563140392303,
      "learning_rate": 9.455024994633055e-06,
      "loss": 0.6433,
      "mean_token_accuracy": 0.7972246296703815,
      "num_tokens": 942328589.0,
      "step": 29525
    },
    {
      "entropy": 0.5362963755615056,
      "epoch": 2.716545620784239,
      "grad_norm": 0.15598426759243011,
      "learning_rate": 9.451958168491428e-06,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8332065306603909,
      "num_tokens": 942360861.0,
      "step": 29526
    },
    {
      "entropy": 0.5242857616394758,
      "epoch": 2.7166376271556807,
      "grad_norm": 0.16002561151981354,
      "learning_rate": 9.448891342349803e-06,
      "loss": 0.5026,
      "mean_token_accuracy": 0.8376519940793514,
      "num_tokens": 942392857.0,
      "step": 29527
    },
    {
      "entropy": 0.45163861382752657,
      "epoch": 2.716729633527122,
      "grad_norm": 0.15507030487060547,
      "learning_rate": 9.445824516208176e-06,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8554118536412716,
      "num_tokens": 942425088.0,
      "step": 29528
    },
    {
      "entropy": 0.5544029891025275,
      "epoch": 2.716821639898563,
      "grad_norm": 0.16259752213954926,
      "learning_rate": 9.44275769006655e-06,
      "loss": 0.5253,
      "mean_token_accuracy": 0.8313968293368816,
      "num_tokens": 942456638.0,
      "step": 29529
    },
    {
      "entropy": 0.4911020901054144,
      "epoch": 2.716913646270004,
      "grad_norm": 0.1552354246377945,
      "learning_rate": 9.439690863924924e-06,
      "loss": 0.4866,
      "mean_token_accuracy": 0.8487954027950764,
      "num_tokens": 942488873.0,
      "step": 29530
    },
    {
      "entropy": 0.4854115080088377,
      "epoch": 2.7170056526414452,
      "grad_norm": 0.1588830053806305,
      "learning_rate": 9.436624037783298e-06,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8479112721979618,
      "num_tokens": 942520817.0,
      "step": 29531
    },
    {
      "entropy": 0.5107281575910747,
      "epoch": 2.717097659012887,
      "grad_norm": 0.15638074278831482,
      "learning_rate": 9.433557211641673e-06,
      "loss": 0.5113,
      "mean_token_accuracy": 0.8371774852275848,
      "num_tokens": 942553161.0,
      "step": 29532
    },
    {
      "entropy": 0.4180474542081356,
      "epoch": 2.717189665384328,
      "grad_norm": 0.14483997225761414,
      "learning_rate": 9.430490385500046e-06,
      "loss": 0.4106,
      "mean_token_accuracy": 0.8683444410562515,
      "num_tokens": 942585189.0,
      "step": 29533
    },
    {
      "entropy": 0.46075727697461843,
      "epoch": 2.717281671755769,
      "grad_norm": 0.16033774614334106,
      "learning_rate": 9.42742355935842e-06,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8556770980358124,
      "num_tokens": 942615837.0,
      "step": 29534
    },
    {
      "entropy": 0.48927901219576597,
      "epoch": 2.71737367812721,
      "grad_norm": 0.1564558893442154,
      "learning_rate": 9.424356733216795e-06,
      "loss": 0.4793,
      "mean_token_accuracy": 0.8512209020555019,
      "num_tokens": 942647691.0,
      "step": 29535
    },
    {
      "entropy": 0.5523123014718294,
      "epoch": 2.7174656844986513,
      "grad_norm": 0.16063927114009857,
      "learning_rate": 9.42128990707517e-06,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8307310044765472,
      "num_tokens": 942679318.0,
      "step": 29536
    },
    {
      "entropy": 0.5385976769030094,
      "epoch": 2.717557690870093,
      "grad_norm": 0.16457320749759674,
      "learning_rate": 9.418223080933543e-06,
      "loss": 0.5263,
      "mean_token_accuracy": 0.8288463950157166,
      "num_tokens": 942711369.0,
      "step": 29537
    },
    {
      "entropy": 0.5359412990510464,
      "epoch": 2.717649697241534,
      "grad_norm": 0.16201885044574738,
      "learning_rate": 9.415156254791917e-06,
      "loss": 0.5168,
      "mean_token_accuracy": 0.8345944434404373,
      "num_tokens": 942742965.0,
      "step": 29538
    },
    {
      "entropy": 0.5387691725045443,
      "epoch": 2.717741703612975,
      "grad_norm": 0.16312868893146515,
      "learning_rate": 9.41208942865029e-06,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8304332979023457,
      "num_tokens": 942775455.0,
      "step": 29539
    },
    {
      "entropy": 0.5182561255060136,
      "epoch": 2.7178337099844163,
      "grad_norm": 0.1548660546541214,
      "learning_rate": 9.409022602508665e-06,
      "loss": 0.5012,
      "mean_token_accuracy": 0.8387514464557171,
      "num_tokens": 942807964.0,
      "step": 29540
    },
    {
      "entropy": 0.5114080628845841,
      "epoch": 2.7179257163558574,
      "grad_norm": 0.16080592572689056,
      "learning_rate": 9.405955776367038e-06,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8407025523483753,
      "num_tokens": 942839712.0,
      "step": 29541
    },
    {
      "entropy": 0.5532078746473417,
      "epoch": 2.718017722727299,
      "grad_norm": 0.16597752273082733,
      "learning_rate": 9.402888950225413e-06,
      "loss": 0.5402,
      "mean_token_accuracy": 0.8278957940638065,
      "num_tokens": 942871094.0,
      "step": 29542
    },
    {
      "entropy": 0.5576238073408604,
      "epoch": 2.71810972909874,
      "grad_norm": 0.16924643516540527,
      "learning_rate": 9.399822124083786e-06,
      "loss": 0.5389,
      "mean_token_accuracy": 0.830027524381876,
      "num_tokens": 942902546.0,
      "step": 29543
    },
    {
      "entropy": 0.42738837841898203,
      "epoch": 2.7182017354701813,
      "grad_norm": 0.15169306099414825,
      "learning_rate": 9.39675529794216e-06,
      "loss": 0.4247,
      "mean_token_accuracy": 0.8623219132423401,
      "num_tokens": 942934042.0,
      "step": 29544
    },
    {
      "entropy": 0.5157542433589697,
      "epoch": 2.7182937418416224,
      "grad_norm": 0.15863071382045746,
      "learning_rate": 9.393688471800533e-06,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8408646769821644,
      "num_tokens": 942966810.0,
      "step": 29545
    },
    {
      "entropy": 0.5631628390401602,
      "epoch": 2.7183857482130636,
      "grad_norm": 0.16536839306354523,
      "learning_rate": 9.390621645658908e-06,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8265970535576344,
      "num_tokens": 942997990.0,
      "step": 29546
    },
    {
      "entropy": 0.5431593470275402,
      "epoch": 2.718477754584505,
      "grad_norm": 0.1622903198003769,
      "learning_rate": 9.387554819517282e-06,
      "loss": 0.533,
      "mean_token_accuracy": 0.8326757177710533,
      "num_tokens": 943029443.0,
      "step": 29547
    },
    {
      "entropy": 0.46254440664779395,
      "epoch": 2.7185697609559463,
      "grad_norm": 0.1527068316936493,
      "learning_rate": 9.384487993375655e-06,
      "loss": 0.4558,
      "mean_token_accuracy": 0.857398334890604,
      "num_tokens": 943062069.0,
      "step": 29548
    },
    {
      "entropy": 0.5279805868631229,
      "epoch": 2.7186617673273874,
      "grad_norm": 0.16338324546813965,
      "learning_rate": 9.38142116723403e-06,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8377890177071095,
      "num_tokens": 943093878.0,
      "step": 29549
    },
    {
      "entropy": 0.45413206052035093,
      "epoch": 2.7187537736988285,
      "grad_norm": 0.15075397491455078,
      "learning_rate": 9.378354341092403e-06,
      "loss": 0.4465,
      "mean_token_accuracy": 0.857798270881176,
      "num_tokens": 943126490.0,
      "step": 29550
    },
    {
      "entropy": 0.5252186730504036,
      "epoch": 2.7188457800702697,
      "grad_norm": 0.16004788875579834,
      "learning_rate": 9.37528751495078e-06,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8375021629035473,
      "num_tokens": 943158261.0,
      "step": 29551
    },
    {
      "entropy": 0.5398185262456536,
      "epoch": 2.7189377864417112,
      "grad_norm": 0.16436520218849182,
      "learning_rate": 9.372220688809152e-06,
      "loss": 0.524,
      "mean_token_accuracy": 0.8316056616604328,
      "num_tokens": 943190103.0,
      "step": 29552
    },
    {
      "entropy": 0.5539820622652769,
      "epoch": 2.7190297928131524,
      "grad_norm": 0.16734415292739868,
      "learning_rate": 9.369153862667527e-06,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8313120454549789,
      "num_tokens": 943221264.0,
      "step": 29553
    },
    {
      "entropy": 0.5067363493144512,
      "epoch": 2.7191217991845935,
      "grad_norm": 0.16446061432361603,
      "learning_rate": 9.3660870365259e-06,
      "loss": 0.4959,
      "mean_token_accuracy": 0.8414928838610649,
      "num_tokens": 943252517.0,
      "step": 29554
    },
    {
      "entropy": 0.5138954482972622,
      "epoch": 2.7192138055560346,
      "grad_norm": 0.15598028898239136,
      "learning_rate": 9.363020210384275e-06,
      "loss": 0.4926,
      "mean_token_accuracy": 0.845769289880991,
      "num_tokens": 943284742.0,
      "step": 29555
    },
    {
      "entropy": 0.5439396165311337,
      "epoch": 2.719305811927476,
      "grad_norm": 0.1606169492006302,
      "learning_rate": 9.359953384242647e-06,
      "loss": 0.538,
      "mean_token_accuracy": 0.831633884459734,
      "num_tokens": 943316754.0,
      "step": 29556
    },
    {
      "entropy": 0.5103464722633362,
      "epoch": 2.7193978182989174,
      "grad_norm": 0.1588754802942276,
      "learning_rate": 9.356886558101022e-06,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8359794802963734,
      "num_tokens": 943349522.0,
      "step": 29557
    },
    {
      "entropy": 0.5343229081481695,
      "epoch": 2.7194898246703585,
      "grad_norm": 0.15927842259407043,
      "learning_rate": 9.353819731959395e-06,
      "loss": 0.5255,
      "mean_token_accuracy": 0.8321388885378838,
      "num_tokens": 943381865.0,
      "step": 29558
    },
    {
      "entropy": 0.5423755906522274,
      "epoch": 2.7195818310417996,
      "grad_norm": 0.1671418845653534,
      "learning_rate": 9.35075290581777e-06,
      "loss": 0.5328,
      "mean_token_accuracy": 0.8316516689956188,
      "num_tokens": 943414348.0,
      "step": 29559
    },
    {
      "entropy": 0.5668250471353531,
      "epoch": 2.7196738374132408,
      "grad_norm": 0.17067091166973114,
      "learning_rate": 9.347686079676143e-06,
      "loss": 0.5556,
      "mean_token_accuracy": 0.8275418356060982,
      "num_tokens": 943446668.0,
      "step": 29560
    },
    {
      "entropy": 0.5010566059499979,
      "epoch": 2.719765843784682,
      "grad_norm": 0.1612321138381958,
      "learning_rate": 9.344619253534517e-06,
      "loss": 0.4856,
      "mean_token_accuracy": 0.8445966765284538,
      "num_tokens": 943479013.0,
      "step": 29561
    },
    {
      "entropy": 0.35792711144313216,
      "epoch": 2.7198578501561235,
      "grad_norm": 0.13356424868106842,
      "learning_rate": 9.341552427392892e-06,
      "loss": 0.3472,
      "mean_token_accuracy": 0.8886379711329937,
      "num_tokens": 943510420.0,
      "step": 29562
    },
    {
      "entropy": 0.47766319708898664,
      "epoch": 2.7199498565275646,
      "grad_norm": 0.1505100578069687,
      "learning_rate": 9.338485601251265e-06,
      "loss": 0.4725,
      "mean_token_accuracy": 0.8503442853689194,
      "num_tokens": 943543023.0,
      "step": 29563
    },
    {
      "entropy": 0.54867840744555,
      "epoch": 2.7200418628990057,
      "grad_norm": 0.1679849773645401,
      "learning_rate": 9.33541877510964e-06,
      "loss": 0.5396,
      "mean_token_accuracy": 0.8305001147091389,
      "num_tokens": 943575289.0,
      "step": 29564
    },
    {
      "entropy": 0.546886382624507,
      "epoch": 2.720133869270447,
      "grad_norm": 0.16200384497642517,
      "learning_rate": 9.332351948968013e-06,
      "loss": 0.5339,
      "mean_token_accuracy": 0.8319043032824993,
      "num_tokens": 943606542.0,
      "step": 29565
    },
    {
      "entropy": 0.5371547788381577,
      "epoch": 2.720225875641888,
      "grad_norm": 0.16876019537448883,
      "learning_rate": 9.329285122826389e-06,
      "loss": 0.5171,
      "mean_token_accuracy": 0.8359006568789482,
      "num_tokens": 943637532.0,
      "step": 29566
    },
    {
      "entropy": 0.6683938857167959,
      "epoch": 2.7203178820133296,
      "grad_norm": 0.18114031851291656,
      "learning_rate": 9.326218296684762e-06,
      "loss": 0.6704,
      "mean_token_accuracy": 0.7952069826424122,
      "num_tokens": 943669915.0,
      "step": 29567
    },
    {
      "entropy": 0.5274489168077707,
      "epoch": 2.7204098883847707,
      "grad_norm": 0.15835559368133545,
      "learning_rate": 9.323151470543136e-06,
      "loss": 0.5096,
      "mean_token_accuracy": 0.8375577814877033,
      "num_tokens": 943701581.0,
      "step": 29568
    },
    {
      "entropy": 0.5554957399144769,
      "epoch": 2.720501894756212,
      "grad_norm": 0.1650552898645401,
      "learning_rate": 9.32008464440151e-06,
      "loss": 0.5424,
      "mean_token_accuracy": 0.8292694836854935,
      "num_tokens": 943734239.0,
      "step": 29569
    },
    {
      "entropy": 0.4652759227901697,
      "epoch": 2.720593901127653,
      "grad_norm": 0.1521858125925064,
      "learning_rate": 9.317017818259884e-06,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8579746969044209,
      "num_tokens": 943766805.0,
      "step": 29570
    },
    {
      "entropy": 0.5553964730352163,
      "epoch": 2.720685907499094,
      "grad_norm": 0.16752220690250397,
      "learning_rate": 9.313950992118257e-06,
      "loss": 0.5447,
      "mean_token_accuracy": 0.8255958892405033,
      "num_tokens": 943798456.0,
      "step": 29571
    },
    {
      "entropy": 0.41632277332246304,
      "epoch": 2.7207779138705357,
      "grad_norm": 0.1443548947572708,
      "learning_rate": 9.310884165976632e-06,
      "loss": 0.3925,
      "mean_token_accuracy": 0.8712684214115143,
      "num_tokens": 943829689.0,
      "step": 29572
    },
    {
      "entropy": 0.5260162218473852,
      "epoch": 2.720869920241977,
      "grad_norm": 0.1588772088289261,
      "learning_rate": 9.307817339835005e-06,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8370610512793064,
      "num_tokens": 943862277.0,
      "step": 29573
    },
    {
      "entropy": 0.5115332221612334,
      "epoch": 2.720961926613418,
      "grad_norm": 0.1620284616947174,
      "learning_rate": 9.30475051369338e-06,
      "loss": 0.5111,
      "mean_token_accuracy": 0.8342828340828419,
      "num_tokens": 943894458.0,
      "step": 29574
    },
    {
      "entropy": 0.5179223215673119,
      "epoch": 2.721053932984859,
      "grad_norm": 0.15561611950397491,
      "learning_rate": 9.301683687551752e-06,
      "loss": 0.5055,
      "mean_token_accuracy": 0.8421340323984623,
      "num_tokens": 943926760.0,
      "step": 29575
    },
    {
      "entropy": 0.5905134528875351,
      "epoch": 2.7211459393563002,
      "grad_norm": 0.16326211392879486,
      "learning_rate": 9.298616861410127e-06,
      "loss": 0.5711,
      "mean_token_accuracy": 0.8190884664654732,
      "num_tokens": 943958722.0,
      "step": 29576
    },
    {
      "entropy": 0.5799779482185841,
      "epoch": 2.721237945727742,
      "grad_norm": 0.16340073943138123,
      "learning_rate": 9.295550035268502e-06,
      "loss": 0.5591,
      "mean_token_accuracy": 0.8236204236745834,
      "num_tokens": 943990701.0,
      "step": 29577
    },
    {
      "entropy": 0.4383192863315344,
      "epoch": 2.721329952099183,
      "grad_norm": 0.1499021053314209,
      "learning_rate": 9.292483209126874e-06,
      "loss": 0.4295,
      "mean_token_accuracy": 0.8624138161540031,
      "num_tokens": 944022224.0,
      "step": 29578
    },
    {
      "entropy": 0.6219814382493496,
      "epoch": 2.721421958470624,
      "grad_norm": 0.17073261737823486,
      "learning_rate": 9.289416382985249e-06,
      "loss": 0.6049,
      "mean_token_accuracy": 0.8129013441503048,
      "num_tokens": 944053468.0,
      "step": 29579
    },
    {
      "entropy": 0.5641612745821476,
      "epoch": 2.721513964842065,
      "grad_norm": 0.1620301902294159,
      "learning_rate": 9.286349556843622e-06,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8276461437344551,
      "num_tokens": 944086153.0,
      "step": 29580
    },
    {
      "entropy": 0.5627747625112534,
      "epoch": 2.7216059712135063,
      "grad_norm": 0.16287249326705933,
      "learning_rate": 9.283282730701998e-06,
      "loss": 0.5515,
      "mean_token_accuracy": 0.828267827630043,
      "num_tokens": 944117198.0,
      "step": 29581
    },
    {
      "entropy": 0.478362335357815,
      "epoch": 2.721697977584948,
      "grad_norm": 0.15343336760997772,
      "learning_rate": 9.280215904560371e-06,
      "loss": 0.4681,
      "mean_token_accuracy": 0.851282462477684,
      "num_tokens": 944149404.0,
      "step": 29582
    },
    {
      "entropy": 0.5806732960045338,
      "epoch": 2.721789983956389,
      "grad_norm": 0.16574163734912872,
      "learning_rate": 9.277149078418746e-06,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8210759460926056,
      "num_tokens": 944181673.0,
      "step": 29583
    },
    {
      "entropy": 0.5358434771187603,
      "epoch": 2.72188199032783,
      "grad_norm": 0.16628021001815796,
      "learning_rate": 9.274082252277119e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8356897756457329,
      "num_tokens": 944213539.0,
      "step": 29584
    },
    {
      "entropy": 0.49048814503476024,
      "epoch": 2.7219739966992713,
      "grad_norm": 0.15433037281036377,
      "learning_rate": 9.271015426135494e-06,
      "loss": 0.472,
      "mean_token_accuracy": 0.8496713936328888,
      "num_tokens": 944245838.0,
      "step": 29585
    },
    {
      "entropy": 0.5455464627593756,
      "epoch": 2.7220660030707124,
      "grad_norm": 0.1654725819826126,
      "learning_rate": 9.267948599993867e-06,
      "loss": 0.5377,
      "mean_token_accuracy": 0.8290515691041946,
      "num_tokens": 944276686.0,
      "step": 29586
    },
    {
      "entropy": 0.4717392046004534,
      "epoch": 2.722158009442154,
      "grad_norm": 0.15325377881526947,
      "learning_rate": 9.264881773852241e-06,
      "loss": 0.4539,
      "mean_token_accuracy": 0.8525349870324135,
      "num_tokens": 944308268.0,
      "step": 29587
    },
    {
      "entropy": 0.5656623216345906,
      "epoch": 2.722250015813595,
      "grad_norm": 0.15836845338344574,
      "learning_rate": 9.261814947710614e-06,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8287531547248363,
      "num_tokens": 944340683.0,
      "step": 29588
    },
    {
      "entropy": 0.5388770815916359,
      "epoch": 2.7223420221850363,
      "grad_norm": 0.1647321730852127,
      "learning_rate": 9.258748121568989e-06,
      "loss": 0.5252,
      "mean_token_accuracy": 0.8346909768879414,
      "num_tokens": 944372532.0,
      "step": 29589
    },
    {
      "entropy": 0.5045567732304335,
      "epoch": 2.7224340285564774,
      "grad_norm": 0.16110534965991974,
      "learning_rate": 9.255681295427362e-06,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8453636020421982,
      "num_tokens": 944403528.0,
      "step": 29590
    },
    {
      "entropy": 0.6255577560514212,
      "epoch": 2.7225260349279186,
      "grad_norm": 0.17016375064849854,
      "learning_rate": 9.252614469285736e-06,
      "loss": 0.6129,
      "mean_token_accuracy": 0.8092350289225578,
      "num_tokens": 944435604.0,
      "step": 29591
    },
    {
      "entropy": 0.5915053673088551,
      "epoch": 2.72261804129936,
      "grad_norm": 0.16589447855949402,
      "learning_rate": 9.249547643144111e-06,
      "loss": 0.5856,
      "mean_token_accuracy": 0.8225565254688263,
      "num_tokens": 944467205.0,
      "step": 29592
    },
    {
      "entropy": 0.523325590416789,
      "epoch": 2.7227100476708013,
      "grad_norm": 0.15910154581069946,
      "learning_rate": 9.246480817002484e-06,
      "loss": 0.5154,
      "mean_token_accuracy": 0.834106270223856,
      "num_tokens": 944499065.0,
      "step": 29593
    },
    {
      "entropy": 0.5059903683140874,
      "epoch": 2.7228020540422424,
      "grad_norm": 0.1624099761247635,
      "learning_rate": 9.243413990860859e-06,
      "loss": 0.4904,
      "mean_token_accuracy": 0.8406122997403145,
      "num_tokens": 944530718.0,
      "step": 29594
    },
    {
      "entropy": 0.6068455390632153,
      "epoch": 2.7228940604136835,
      "grad_norm": 0.17373625934123993,
      "learning_rate": 9.240347164719232e-06,
      "loss": 0.603,
      "mean_token_accuracy": 0.8137147463858128,
      "num_tokens": 944562284.0,
      "step": 29595
    },
    {
      "entropy": 0.544063737615943,
      "epoch": 2.7229860667851247,
      "grad_norm": 0.16134260594844818,
      "learning_rate": 9.237280338577608e-06,
      "loss": 0.5419,
      "mean_token_accuracy": 0.8275137208402157,
      "num_tokens": 944594764.0,
      "step": 29596
    },
    {
      "entropy": 0.5224260212853551,
      "epoch": 2.7230780731565662,
      "grad_norm": 0.15844261646270752,
      "learning_rate": 9.234213512435981e-06,
      "loss": 0.5108,
      "mean_token_accuracy": 0.8368390649557114,
      "num_tokens": 944627206.0,
      "step": 29597
    },
    {
      "entropy": 0.491854777559638,
      "epoch": 2.7231700795280074,
      "grad_norm": 0.1501895934343338,
      "learning_rate": 9.231146686294356e-06,
      "loss": 0.476,
      "mean_token_accuracy": 0.8496314324438572,
      "num_tokens": 944658837.0,
      "step": 29598
    },
    {
      "entropy": 0.4588767299428582,
      "epoch": 2.7232620858994485,
      "grad_norm": 0.15485963225364685,
      "learning_rate": 9.228079860152729e-06,
      "loss": 0.4429,
      "mean_token_accuracy": 0.857433658093214,
      "num_tokens": 944690246.0,
      "step": 29599
    },
    {
      "entropy": 0.5965428585186601,
      "epoch": 2.7233540922708896,
      "grad_norm": 0.16626432538032532,
      "learning_rate": 9.225013034011103e-06,
      "loss": 0.5922,
      "mean_token_accuracy": 0.8124168515205383,
      "num_tokens": 944722496.0,
      "step": 29600
    },
    {
      "entropy": 0.5655619772151113,
      "epoch": 2.723446098642331,
      "grad_norm": 0.16734246909618378,
      "learning_rate": 9.221946207869476e-06,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8255213685333729,
      "num_tokens": 944754982.0,
      "step": 29601
    },
    {
      "entropy": 0.4914613279979676,
      "epoch": 2.7235381050137724,
      "grad_norm": 0.1506175696849823,
      "learning_rate": 9.21887938172785e-06,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8513742946088314,
      "num_tokens": 944787349.0,
      "step": 29602
    },
    {
      "entropy": 0.5020689582452178,
      "epoch": 2.7236301113852135,
      "grad_norm": 0.15416616201400757,
      "learning_rate": 9.215812555586224e-06,
      "loss": 0.487,
      "mean_token_accuracy": 0.8413862995803356,
      "num_tokens": 944819557.0,
      "step": 29603
    },
    {
      "entropy": 0.47908373083919287,
      "epoch": 2.7237221177566546,
      "grad_norm": 0.16314207017421722,
      "learning_rate": 9.212745729444598e-06,
      "loss": 0.4582,
      "mean_token_accuracy": 0.8527943715453148,
      "num_tokens": 944850638.0,
      "step": 29604
    },
    {
      "entropy": 0.5113273030146956,
      "epoch": 2.7238141241280958,
      "grad_norm": 0.1660563200712204,
      "learning_rate": 9.209678903302971e-06,
      "loss": 0.5137,
      "mean_token_accuracy": 0.8379736989736557,
      "num_tokens": 944882466.0,
      "step": 29605
    },
    {
      "entropy": 0.5325425281189382,
      "epoch": 2.723906130499537,
      "grad_norm": 0.15420767664909363,
      "learning_rate": 9.206612077161346e-06,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8374937959015369,
      "num_tokens": 944914196.0,
      "step": 29606
    },
    {
      "entropy": 0.4571570996195078,
      "epoch": 2.7239981368709785,
      "grad_norm": 0.1509828418493271,
      "learning_rate": 9.20354525101972e-06,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8560076467692852,
      "num_tokens": 944946570.0,
      "step": 29607
    },
    {
      "entropy": 0.5470130164176226,
      "epoch": 2.7240901432424196,
      "grad_norm": 0.17037256062030792,
      "learning_rate": 9.200478424878094e-06,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8284795470535755,
      "num_tokens": 944978870.0,
      "step": 29608
    },
    {
      "entropy": 0.5072325430810452,
      "epoch": 2.7241821496138607,
      "grad_norm": 0.15575610101222992,
      "learning_rate": 9.197411598736468e-06,
      "loss": 0.4863,
      "mean_token_accuracy": 0.8434260450303555,
      "num_tokens": 945010766.0,
      "step": 29609
    },
    {
      "entropy": 0.5017226468771696,
      "epoch": 2.724274155985302,
      "grad_norm": 0.16015559434890747,
      "learning_rate": 9.194344772594841e-06,
      "loss": 0.4946,
      "mean_token_accuracy": 0.8427359834313393,
      "num_tokens": 945043046.0,
      "step": 29610
    },
    {
      "entropy": 0.4881234308704734,
      "epoch": 2.724366162356743,
      "grad_norm": 0.1513863056898117,
      "learning_rate": 9.191277946453216e-06,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8514093048870564,
      "num_tokens": 945075707.0,
      "step": 29611
    },
    {
      "entropy": 0.5280466098338366,
      "epoch": 2.7244581687281846,
      "grad_norm": 0.15362980961799622,
      "learning_rate": 9.18821112031159e-06,
      "loss": 0.5106,
      "mean_token_accuracy": 0.8383424244821072,
      "num_tokens": 945107495.0,
      "step": 29612
    },
    {
      "entropy": 0.5880910232663155,
      "epoch": 2.7245501750996257,
      "grad_norm": 0.1752110868692398,
      "learning_rate": 9.185144294169965e-06,
      "loss": 0.5764,
      "mean_token_accuracy": 0.8172674663364887,
      "num_tokens": 945138147.0,
      "step": 29613
    },
    {
      "entropy": 0.5487697292119265,
      "epoch": 2.724642181471067,
      "grad_norm": 0.16463971138000488,
      "learning_rate": 9.182077468028338e-06,
      "loss": 0.5333,
      "mean_token_accuracy": 0.8296995311975479,
      "num_tokens": 945169287.0,
      "step": 29614
    },
    {
      "entropy": 0.4627640414983034,
      "epoch": 2.724734187842508,
      "grad_norm": 0.15024037659168243,
      "learning_rate": 9.179010641886713e-06,
      "loss": 0.4444,
      "mean_token_accuracy": 0.8637184426188469,
      "num_tokens": 945201197.0,
      "step": 29615
    },
    {
      "entropy": 0.5321301613003016,
      "epoch": 2.724826194213949,
      "grad_norm": 0.15857335925102234,
      "learning_rate": 9.175943815745086e-06,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8351636305451393,
      "num_tokens": 945233144.0,
      "step": 29616
    },
    {
      "entropy": 0.46153287030756474,
      "epoch": 2.7249182005853907,
      "grad_norm": 0.15090329945087433,
      "learning_rate": 9.17287698960346e-06,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8557507693767548,
      "num_tokens": 945265662.0,
      "step": 29617
    },
    {
      "entropy": 0.6073259972035885,
      "epoch": 2.725010206956832,
      "grad_norm": 0.17131538689136505,
      "learning_rate": 9.169810163461833e-06,
      "loss": 0.5914,
      "mean_token_accuracy": 0.811681654304266,
      "num_tokens": 945298231.0,
      "step": 29618
    },
    {
      "entropy": 0.5692326985299587,
      "epoch": 2.725102213328273,
      "grad_norm": 0.16512183845043182,
      "learning_rate": 9.166743337320208e-06,
      "loss": 0.5608,
      "mean_token_accuracy": 0.8208355270326138,
      "num_tokens": 945330743.0,
      "step": 29619
    },
    {
      "entropy": 0.5473707169294357,
      "epoch": 2.725194219699714,
      "grad_norm": 0.16075965762138367,
      "learning_rate": 9.163676511178581e-06,
      "loss": 0.5305,
      "mean_token_accuracy": 0.831938125193119,
      "num_tokens": 945362362.0,
      "step": 29620
    },
    {
      "entropy": 0.5730983943212777,
      "epoch": 2.7252862260711552,
      "grad_norm": 0.16847656667232513,
      "learning_rate": 9.160609685036956e-06,
      "loss": 0.5705,
      "mean_token_accuracy": 0.8205975592136383,
      "num_tokens": 945394498.0,
      "step": 29621
    },
    {
      "entropy": 0.4838448923546821,
      "epoch": 2.725378232442597,
      "grad_norm": 0.1524457037448883,
      "learning_rate": 9.15754285889533e-06,
      "loss": 0.4586,
      "mean_token_accuracy": 0.8515657186508179,
      "num_tokens": 945425241.0,
      "step": 29622
    },
    {
      "entropy": 0.5278131365776062,
      "epoch": 2.725470238814038,
      "grad_norm": 0.15466158092021942,
      "learning_rate": 9.154476032753703e-06,
      "loss": 0.5244,
      "mean_token_accuracy": 0.8332024626433849,
      "num_tokens": 945457275.0,
      "step": 29623
    },
    {
      "entropy": 0.5265529118478298,
      "epoch": 2.725562245185479,
      "grad_norm": 0.16076140105724335,
      "learning_rate": 9.151409206612078e-06,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8349849097430706,
      "num_tokens": 945488933.0,
      "step": 29624
    },
    {
      "entropy": 0.5538908811286092,
      "epoch": 2.72565425155692,
      "grad_norm": 0.1656368374824524,
      "learning_rate": 9.14834238047045e-06,
      "loss": 0.5493,
      "mean_token_accuracy": 0.8297204710543156,
      "num_tokens": 945521173.0,
      "step": 29625
    },
    {
      "entropy": 0.5572622418403625,
      "epoch": 2.7257462579283613,
      "grad_norm": 0.16058403253555298,
      "learning_rate": 9.145275554328825e-06,
      "loss": 0.5467,
      "mean_token_accuracy": 0.8299688287079334,
      "num_tokens": 945553490.0,
      "step": 29626
    },
    {
      "entropy": 0.4449320264393464,
      "epoch": 2.725838264299803,
      "grad_norm": 0.14714503288269043,
      "learning_rate": 9.1422087281872e-06,
      "loss": 0.435,
      "mean_token_accuracy": 0.8646619133651257,
      "num_tokens": 945585898.0,
      "step": 29627
    },
    {
      "entropy": 0.4982102755457163,
      "epoch": 2.725930270671244,
      "grad_norm": 0.15423038601875305,
      "learning_rate": 9.139141902045575e-06,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8454103618860245,
      "num_tokens": 945617259.0,
      "step": 29628
    },
    {
      "entropy": 0.4518044125288725,
      "epoch": 2.726022277042685,
      "grad_norm": 0.150223970413208,
      "learning_rate": 9.136075075903948e-06,
      "loss": 0.4335,
      "mean_token_accuracy": 0.8599684461951256,
      "num_tokens": 945649886.0,
      "step": 29629
    },
    {
      "entropy": 0.562741881236434,
      "epoch": 2.7261142834141263,
      "grad_norm": 0.16020908951759338,
      "learning_rate": 9.133008249762322e-06,
      "loss": 0.5402,
      "mean_token_accuracy": 0.8319278582930565,
      "num_tokens": 945682343.0,
      "step": 29630
    },
    {
      "entropy": 0.5059346472844481,
      "epoch": 2.7262062897855674,
      "grad_norm": 0.1632010042667389,
      "learning_rate": 9.129941423620695e-06,
      "loss": 0.495,
      "mean_token_accuracy": 0.837911743670702,
      "num_tokens": 945713081.0,
      "step": 29631
    },
    {
      "entropy": 0.48245511669665575,
      "epoch": 2.726298296157009,
      "grad_norm": 0.15377482771873474,
      "learning_rate": 9.12687459747907e-06,
      "loss": 0.4662,
      "mean_token_accuracy": 0.8501903712749481,
      "num_tokens": 945744629.0,
      "step": 29632
    },
    {
      "entropy": 0.6072153551504016,
      "epoch": 2.72639030252845,
      "grad_norm": 0.17108480632305145,
      "learning_rate": 9.123807771337443e-06,
      "loss": 0.6005,
      "mean_token_accuracy": 0.8114023245871067,
      "num_tokens": 945776271.0,
      "step": 29633
    },
    {
      "entropy": 0.4780431610997766,
      "epoch": 2.7264823088998913,
      "grad_norm": 0.1555602103471756,
      "learning_rate": 9.120740945195817e-06,
      "loss": 0.4669,
      "mean_token_accuracy": 0.8521171994507313,
      "num_tokens": 945807731.0,
      "step": 29634
    },
    {
      "entropy": 0.5444823000580072,
      "epoch": 2.7265743152713324,
      "grad_norm": 0.1612011343240738,
      "learning_rate": 9.11767411905419e-06,
      "loss": 0.527,
      "mean_token_accuracy": 0.8308761194348335,
      "num_tokens": 945839498.0,
      "step": 29635
    },
    {
      "entropy": 0.4331234690034762,
      "epoch": 2.7266663216427736,
      "grad_norm": 0.15041349828243256,
      "learning_rate": 9.114607292912565e-06,
      "loss": 0.4197,
      "mean_token_accuracy": 0.8628703095018864,
      "num_tokens": 945871037.0,
      "step": 29636
    },
    {
      "entropy": 0.4784895800985396,
      "epoch": 2.726758328014215,
      "grad_norm": 0.15221376717090607,
      "learning_rate": 9.11154046677094e-06,
      "loss": 0.4635,
      "mean_token_accuracy": 0.8526449277997017,
      "num_tokens": 945902837.0,
      "step": 29637
    },
    {
      "entropy": 0.5712425988167524,
      "epoch": 2.7268503343856563,
      "grad_norm": 0.1692131608724594,
      "learning_rate": 9.108473640629313e-06,
      "loss": 0.5567,
      "mean_token_accuracy": 0.8232521563768387,
      "num_tokens": 945933572.0,
      "step": 29638
    },
    {
      "entropy": 0.5125413937494159,
      "epoch": 2.7269423407570974,
      "grad_norm": 0.15870437026023865,
      "learning_rate": 9.105406814487687e-06,
      "loss": 0.5094,
      "mean_token_accuracy": 0.8373524993658066,
      "num_tokens": 945965234.0,
      "step": 29639
    },
    {
      "entropy": 0.35252045025117695,
      "epoch": 2.7270343471285385,
      "grad_norm": 0.13403166830539703,
      "learning_rate": 9.10233998834606e-06,
      "loss": 0.3363,
      "mean_token_accuracy": 0.8886059559881687,
      "num_tokens": 945997383.0,
      "step": 29640
    },
    {
      "entropy": 0.5206352174282074,
      "epoch": 2.7271263534999797,
      "grad_norm": 0.15982162952423096,
      "learning_rate": 9.099273162204435e-06,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8346998505294323,
      "num_tokens": 946029761.0,
      "step": 29641
    },
    {
      "entropy": 0.5015395674854517,
      "epoch": 2.7272183598714212,
      "grad_norm": 0.15844011306762695,
      "learning_rate": 9.09620633606281e-06,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8426038324832916,
      "num_tokens": 946061123.0,
      "step": 29642
    },
    {
      "entropy": 0.46899318508803844,
      "epoch": 2.7273103662428624,
      "grad_norm": 0.1528700590133667,
      "learning_rate": 9.093139509921184e-06,
      "loss": 0.4531,
      "mean_token_accuracy": 0.856156513094902,
      "num_tokens": 946092597.0,
      "step": 29643
    },
    {
      "entropy": 0.6497541591525078,
      "epoch": 2.7274023726143035,
      "grad_norm": 0.1736035943031311,
      "learning_rate": 9.090072683779557e-06,
      "loss": 0.6428,
      "mean_token_accuracy": 0.7998178713023663,
      "num_tokens": 946124500.0,
      "step": 29644
    },
    {
      "entropy": 0.4862982486374676,
      "epoch": 2.7274943789857446,
      "grad_norm": 0.14830432832241058,
      "learning_rate": 9.087005857637932e-06,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8542132005095482,
      "num_tokens": 946157012.0,
      "step": 29645
    },
    {
      "entropy": 0.4169713295996189,
      "epoch": 2.727586385357186,
      "grad_norm": 0.1404447853565216,
      "learning_rate": 9.083939031496305e-06,
      "loss": 0.407,
      "mean_token_accuracy": 0.8699168264865875,
      "num_tokens": 946189708.0,
      "step": 29646
    },
    {
      "entropy": 0.3984082117676735,
      "epoch": 2.7276783917286274,
      "grad_norm": 0.14666955173015594,
      "learning_rate": 9.08087220535468e-06,
      "loss": 0.391,
      "mean_token_accuracy": 0.8751269020140171,
      "num_tokens": 946221238.0,
      "step": 29647
    },
    {
      "entropy": 0.562229766510427,
      "epoch": 2.7277703981000685,
      "grad_norm": 0.16269834339618683,
      "learning_rate": 9.077805379213052e-06,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8216497376561165,
      "num_tokens": 946253696.0,
      "step": 29648
    },
    {
      "entropy": 0.5213762372732162,
      "epoch": 2.7278624044715096,
      "grad_norm": 0.15899838507175446,
      "learning_rate": 9.074738553071427e-06,
      "loss": 0.5101,
      "mean_token_accuracy": 0.8374737501144409,
      "num_tokens": 946285837.0,
      "step": 29649
    },
    {
      "entropy": 0.5038363942876458,
      "epoch": 2.7279544108429508,
      "grad_norm": 0.15644441545009613,
      "learning_rate": 9.0716717269298e-06,
      "loss": 0.4972,
      "mean_token_accuracy": 0.8448715135455132,
      "num_tokens": 946318208.0,
      "step": 29650
    },
    {
      "entropy": 0.5465747127309442,
      "epoch": 2.728046417214392,
      "grad_norm": 0.15730251371860504,
      "learning_rate": 9.068604900788175e-06,
      "loss": 0.5251,
      "mean_token_accuracy": 0.833989854902029,
      "num_tokens": 946350432.0,
      "step": 29651
    },
    {
      "entropy": 0.4940732913091779,
      "epoch": 2.7281384235858335,
      "grad_norm": 0.15579551458358765,
      "learning_rate": 9.06553807464655e-06,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8458579108119011,
      "num_tokens": 946382463.0,
      "step": 29652
    },
    {
      "entropy": 0.4700454145204276,
      "epoch": 2.7282304299572746,
      "grad_norm": 0.15088273584842682,
      "learning_rate": 9.062471248504922e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.857672818005085,
      "num_tokens": 946414123.0,
      "step": 29653
    },
    {
      "entropy": 0.46988254878669977,
      "epoch": 2.7283224363287157,
      "grad_norm": 0.15839172899723053,
      "learning_rate": 9.059404422363297e-06,
      "loss": 0.4644,
      "mean_token_accuracy": 0.8489965088665485,
      "num_tokens": 946446267.0,
      "step": 29654
    },
    {
      "entropy": 0.5595105644315481,
      "epoch": 2.728414442700157,
      "grad_norm": 0.16320227086544037,
      "learning_rate": 9.05633759622167e-06,
      "loss": 0.5568,
      "mean_token_accuracy": 0.824364498257637,
      "num_tokens": 946478989.0,
      "step": 29655
    },
    {
      "entropy": 0.510800359188579,
      "epoch": 2.728506449071598,
      "grad_norm": 0.15667425096035004,
      "learning_rate": 9.053270770080044e-06,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8424556478857994,
      "num_tokens": 946511122.0,
      "step": 29656
    },
    {
      "entropy": 0.557733491063118,
      "epoch": 2.7285984554430396,
      "grad_norm": 0.16074101626873016,
      "learning_rate": 9.050203943938419e-06,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8256632536649704,
      "num_tokens": 946542930.0,
      "step": 29657
    },
    {
      "entropy": 0.45842165779322386,
      "epoch": 2.7286904618144807,
      "grad_norm": 0.1572568416595459,
      "learning_rate": 9.047137117796794e-06,
      "loss": 0.4568,
      "mean_token_accuracy": 0.8554851375520229,
      "num_tokens": 946574861.0,
      "step": 29658
    },
    {
      "entropy": 0.5045018158853054,
      "epoch": 2.728782468185922,
      "grad_norm": 0.1541786640882492,
      "learning_rate": 9.044070291655167e-06,
      "loss": 0.4994,
      "mean_token_accuracy": 0.8425640724599361,
      "num_tokens": 946607550.0,
      "step": 29659
    },
    {
      "entropy": 0.5821663793176413,
      "epoch": 2.728874474557363,
      "grad_norm": 0.17253249883651733,
      "learning_rate": 9.041003465513541e-06,
      "loss": 0.5712,
      "mean_token_accuracy": 0.8220116570591927,
      "num_tokens": 946637746.0,
      "step": 29660
    },
    {
      "entropy": 0.48069442715495825,
      "epoch": 2.728966480928804,
      "grad_norm": 0.15614914894104004,
      "learning_rate": 9.037936639371914e-06,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8524994738399982,
      "num_tokens": 946670148.0,
      "step": 29661
    },
    {
      "entropy": 0.5232077967375517,
      "epoch": 2.7290584873002457,
      "grad_norm": 0.1619570404291153,
      "learning_rate": 9.034869813230289e-06,
      "loss": 0.5108,
      "mean_token_accuracy": 0.836569469422102,
      "num_tokens": 946701665.0,
      "step": 29662
    },
    {
      "entropy": 0.5841966653242707,
      "epoch": 2.729150493671687,
      "grad_norm": 0.17451192438602448,
      "learning_rate": 9.031802987088662e-06,
      "loss": 0.5757,
      "mean_token_accuracy": 0.8168530464172363,
      "num_tokens": 946733960.0,
      "step": 29663
    },
    {
      "entropy": 0.5330538162961602,
      "epoch": 2.729242500043128,
      "grad_norm": 0.16109608113765717,
      "learning_rate": 9.028736160947037e-06,
      "loss": 0.5259,
      "mean_token_accuracy": 0.835358053445816,
      "num_tokens": 946765523.0,
      "step": 29664
    },
    {
      "entropy": 0.5015799840912223,
      "epoch": 2.729334506414569,
      "grad_norm": 0.16183944046497345,
      "learning_rate": 9.02566933480541e-06,
      "loss": 0.4935,
      "mean_token_accuracy": 0.8453164398670197,
      "num_tokens": 946796982.0,
      "step": 29665
    },
    {
      "entropy": 0.531371034681797,
      "epoch": 2.7294265127860102,
      "grad_norm": 0.16691473126411438,
      "learning_rate": 9.022602508663784e-06,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8338959142565727,
      "num_tokens": 946828652.0,
      "step": 29666
    },
    {
      "entropy": 0.5451765488833189,
      "epoch": 2.729518519157452,
      "grad_norm": 0.16108354926109314,
      "learning_rate": 9.019535682522159e-06,
      "loss": 0.5337,
      "mean_token_accuracy": 0.8278539553284645,
      "num_tokens": 946860336.0,
      "step": 29667
    },
    {
      "entropy": 0.4484567269682884,
      "epoch": 2.729610525528893,
      "grad_norm": 0.1560751050710678,
      "learning_rate": 9.016468856380532e-06,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8569641783833504,
      "num_tokens": 946891521.0,
      "step": 29668
    },
    {
      "entropy": 0.6048677014186978,
      "epoch": 2.729702531900334,
      "grad_norm": 0.16965310275554657,
      "learning_rate": 9.013402030238906e-06,
      "loss": 0.5828,
      "mean_token_accuracy": 0.813883937895298,
      "num_tokens": 946923435.0,
      "step": 29669
    },
    {
      "entropy": 0.4610969563946128,
      "epoch": 2.729794538271775,
      "grad_norm": 0.15348432958126068,
      "learning_rate": 9.01033520409728e-06,
      "loss": 0.4493,
      "mean_token_accuracy": 0.8568824604153633,
      "num_tokens": 946955245.0,
      "step": 29670
    },
    {
      "entropy": 0.5938501637428999,
      "epoch": 2.7298865446432163,
      "grad_norm": 0.1730017364025116,
      "learning_rate": 9.007268377955654e-06,
      "loss": 0.5847,
      "mean_token_accuracy": 0.8161864876747131,
      "num_tokens": 946986355.0,
      "step": 29671
    },
    {
      "entropy": 0.5165241155773401,
      "epoch": 2.729978551014658,
      "grad_norm": 0.16181670129299164,
      "learning_rate": 9.004201551814027e-06,
      "loss": 0.5057,
      "mean_token_accuracy": 0.8407588787376881,
      "num_tokens": 947018350.0,
      "step": 29672
    },
    {
      "entropy": 0.495699810795486,
      "epoch": 2.730070557386099,
      "grad_norm": 0.15789002180099487,
      "learning_rate": 9.001134725672403e-06,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8454557582736015,
      "num_tokens": 947050453.0,
      "step": 29673
    },
    {
      "entropy": 0.5260754683986306,
      "epoch": 2.73016256375754,
      "grad_norm": 0.16036228835582733,
      "learning_rate": 8.998067899530776e-06,
      "loss": 0.518,
      "mean_token_accuracy": 0.8372959047555923,
      "num_tokens": 947082309.0,
      "step": 29674
    },
    {
      "entropy": 0.5182987693697214,
      "epoch": 2.7302545701289813,
      "grad_norm": 0.16141530871391296,
      "learning_rate": 8.995001073389151e-06,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8398929722607136,
      "num_tokens": 947113178.0,
      "step": 29675
    },
    {
      "entropy": 0.4182682844111696,
      "epoch": 2.7303465765004225,
      "grad_norm": 0.1494060605764389,
      "learning_rate": 8.991934247247524e-06,
      "loss": 0.413,
      "mean_token_accuracy": 0.8691479079425335,
      "num_tokens": 947144560.0,
      "step": 29676
    },
    {
      "entropy": 0.48311814945191145,
      "epoch": 2.730438582871864,
      "grad_norm": 0.15918053686618805,
      "learning_rate": 8.988867421105899e-06,
      "loss": 0.4799,
      "mean_token_accuracy": 0.8507554158568382,
      "num_tokens": 947176264.0,
      "step": 29677
    },
    {
      "entropy": 0.5889873504638672,
      "epoch": 2.730530589243305,
      "grad_norm": 0.17376001179218292,
      "learning_rate": 8.985800594964271e-06,
      "loss": 0.6019,
      "mean_token_accuracy": 0.8137708567082882,
      "num_tokens": 947208141.0,
      "step": 29678
    },
    {
      "entropy": 0.4593990119174123,
      "epoch": 2.7306225956147463,
      "grad_norm": 0.15310770273208618,
      "learning_rate": 8.982733768822646e-06,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8534647934138775,
      "num_tokens": 947240117.0,
      "step": 29679
    },
    {
      "entropy": 0.5318596782162786,
      "epoch": 2.7307146019861874,
      "grad_norm": 0.16113431751728058,
      "learning_rate": 8.979666942681019e-06,
      "loss": 0.5141,
      "mean_token_accuracy": 0.8378210179507732,
      "num_tokens": 947272221.0,
      "step": 29680
    },
    {
      "entropy": 0.5062451157718897,
      "epoch": 2.7308066083576286,
      "grad_norm": 0.15780429542064667,
      "learning_rate": 8.976600116539394e-06,
      "loss": 0.4927,
      "mean_token_accuracy": 0.84061698615551,
      "num_tokens": 947304364.0,
      "step": 29681
    },
    {
      "entropy": 0.47126693185418844,
      "epoch": 2.73089861472907,
      "grad_norm": 0.15135739743709564,
      "learning_rate": 8.973533290397768e-06,
      "loss": 0.4565,
      "mean_token_accuracy": 0.853811539709568,
      "num_tokens": 947336197.0,
      "step": 29682
    },
    {
      "entropy": 0.4164144368842244,
      "epoch": 2.7309906211005113,
      "grad_norm": 0.14283674955368042,
      "learning_rate": 8.970466464256141e-06,
      "loss": 0.4026,
      "mean_token_accuracy": 0.8748201839625835,
      "num_tokens": 947368799.0,
      "step": 29683
    },
    {
      "entropy": 0.5494669992476702,
      "epoch": 2.7310826274719524,
      "grad_norm": 0.15570317208766937,
      "learning_rate": 8.967399638114516e-06,
      "loss": 0.5372,
      "mean_token_accuracy": 0.8330225273966789,
      "num_tokens": 947400484.0,
      "step": 29684
    },
    {
      "entropy": 0.5498641487210989,
      "epoch": 2.7311746338433935,
      "grad_norm": 0.16060389578342438,
      "learning_rate": 8.964332811972889e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8328036516904831,
      "num_tokens": 947433111.0,
      "step": 29685
    },
    {
      "entropy": 0.48224139027297497,
      "epoch": 2.7312666402148347,
      "grad_norm": 0.1510043740272522,
      "learning_rate": 8.961265985831264e-06,
      "loss": 0.4634,
      "mean_token_accuracy": 0.8531440570950508,
      "num_tokens": 947465411.0,
      "step": 29686
    },
    {
      "entropy": 0.5464857313781977,
      "epoch": 2.7313586465862763,
      "grad_norm": 0.16498233377933502,
      "learning_rate": 8.958199159689637e-06,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8303550258278847,
      "num_tokens": 947497526.0,
      "step": 29687
    },
    {
      "entropy": 0.5220046187750995,
      "epoch": 2.7314506529577174,
      "grad_norm": 0.16038405895233154,
      "learning_rate": 8.955132333548013e-06,
      "loss": 0.5042,
      "mean_token_accuracy": 0.8392077796161175,
      "num_tokens": 947529408.0,
      "step": 29688
    },
    {
      "entropy": 0.409396936185658,
      "epoch": 2.7315426593291585,
      "grad_norm": 0.14925898611545563,
      "learning_rate": 8.952065507406386e-06,
      "loss": 0.3949,
      "mean_token_accuracy": 0.8723459765315056,
      "num_tokens": 947561748.0,
      "step": 29689
    },
    {
      "entropy": 0.5248338924720883,
      "epoch": 2.7316346657005997,
      "grad_norm": 0.1570858657360077,
      "learning_rate": 8.94899868126476e-06,
      "loss": 0.5124,
      "mean_token_accuracy": 0.839169554412365,
      "num_tokens": 947593682.0,
      "step": 29690
    },
    {
      "entropy": 0.5638986201956868,
      "epoch": 2.731726672072041,
      "grad_norm": 0.16075275838375092,
      "learning_rate": 8.945931855123133e-06,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8248899355530739,
      "num_tokens": 947625926.0,
      "step": 29691
    },
    {
      "entropy": 0.5168521133018658,
      "epoch": 2.7318186784434824,
      "grad_norm": 0.16104188561439514,
      "learning_rate": 8.942865028981508e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.840355034917593,
      "num_tokens": 947657644.0,
      "step": 29692
    },
    {
      "entropy": 0.45424635522067547,
      "epoch": 2.7319106848149235,
      "grad_norm": 0.14881649613380432,
      "learning_rate": 8.939798202839881e-06,
      "loss": 0.4339,
      "mean_token_accuracy": 0.8614268526434898,
      "num_tokens": 947689799.0,
      "step": 29693
    },
    {
      "entropy": 0.49677198752760887,
      "epoch": 2.7320026911863646,
      "grad_norm": 0.1588214933872223,
      "learning_rate": 8.936731376698256e-06,
      "loss": 0.491,
      "mean_token_accuracy": 0.8431148529052734,
      "num_tokens": 947721957.0,
      "step": 29694
    },
    {
      "entropy": 0.42685484304092824,
      "epoch": 2.7320946975578058,
      "grad_norm": 0.14598415791988373,
      "learning_rate": 8.933664550556629e-06,
      "loss": 0.4119,
      "mean_token_accuracy": 0.8723443262279034,
      "num_tokens": 947754565.0,
      "step": 29695
    },
    {
      "entropy": 0.5363648477941751,
      "epoch": 2.732186703929247,
      "grad_norm": 0.16256003081798553,
      "learning_rate": 8.930597724415003e-06,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8335241340100765,
      "num_tokens": 947786647.0,
      "step": 29696
    },
    {
      "entropy": 0.4372754734940827,
      "epoch": 2.7322787103006885,
      "grad_norm": 0.14441999793052673,
      "learning_rate": 8.927530898273378e-06,
      "loss": 0.4227,
      "mean_token_accuracy": 0.8667516298592091,
      "num_tokens": 947818660.0,
      "step": 29697
    },
    {
      "entropy": 0.5185881545767188,
      "epoch": 2.7323707166721296,
      "grad_norm": 0.1609731912612915,
      "learning_rate": 8.924464072131751e-06,
      "loss": 0.4922,
      "mean_token_accuracy": 0.842311929911375,
      "num_tokens": 947850516.0,
      "step": 29698
    },
    {
      "entropy": 0.5564189217984676,
      "epoch": 2.7324627230435707,
      "grad_norm": 0.16546325385570526,
      "learning_rate": 8.921397245990126e-06,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8290635198354721,
      "num_tokens": 947882629.0,
      "step": 29699
    },
    {
      "entropy": 0.5391829870641232,
      "epoch": 2.732554729415012,
      "grad_norm": 0.16195236146450043,
      "learning_rate": 8.918330419848498e-06,
      "loss": 0.5355,
      "mean_token_accuracy": 0.8347656950354576,
      "num_tokens": 947914368.0,
      "step": 29700
    },
    {
      "entropy": 0.459289756603539,
      "epoch": 2.732646735786453,
      "grad_norm": 0.14544479548931122,
      "learning_rate": 8.915263593706873e-06,
      "loss": 0.4453,
      "mean_token_accuracy": 0.858172569423914,
      "num_tokens": 947946055.0,
      "step": 29701
    },
    {
      "entropy": 0.4647158104926348,
      "epoch": 2.7327387421578946,
      "grad_norm": 0.14549556374549866,
      "learning_rate": 8.912196767565246e-06,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8579197824001312,
      "num_tokens": 947977532.0,
      "step": 29702
    },
    {
      "entropy": 0.5393298380076885,
      "epoch": 2.7328307485293357,
      "grad_norm": 0.16053301095962524,
      "learning_rate": 8.909129941423622e-06,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8374634310603142,
      "num_tokens": 948010151.0,
      "step": 29703
    },
    {
      "entropy": 0.5600323937833309,
      "epoch": 2.732922754900777,
      "grad_norm": 0.1649934947490692,
      "learning_rate": 8.906063115281995e-06,
      "loss": 0.5407,
      "mean_token_accuracy": 0.8272761031985283,
      "num_tokens": 948041742.0,
      "step": 29704
    },
    {
      "entropy": 0.530765051022172,
      "epoch": 2.733014761272218,
      "grad_norm": 0.1632613241672516,
      "learning_rate": 8.90299628914037e-06,
      "loss": 0.5248,
      "mean_token_accuracy": 0.8297032080590725,
      "num_tokens": 948073687.0,
      "step": 29705
    },
    {
      "entropy": 0.5167020196095109,
      "epoch": 2.733106767643659,
      "grad_norm": 0.1519050896167755,
      "learning_rate": 8.899929462998743e-06,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8390440493822098,
      "num_tokens": 948105012.0,
      "step": 29706
    },
    {
      "entropy": 0.5207621455192566,
      "epoch": 2.7331987740151007,
      "grad_norm": 0.1598518341779709,
      "learning_rate": 8.896862636857118e-06,
      "loss": 0.5083,
      "mean_token_accuracy": 0.8392688781023026,
      "num_tokens": 948136677.0,
      "step": 29707
    },
    {
      "entropy": 0.5559862535446882,
      "epoch": 2.733290780386542,
      "grad_norm": 0.16873471438884735,
      "learning_rate": 8.89379581071549e-06,
      "loss": 0.5494,
      "mean_token_accuracy": 0.8279499746859074,
      "num_tokens": 948168514.0,
      "step": 29708
    },
    {
      "entropy": 0.5025760885328054,
      "epoch": 2.733382786757983,
      "grad_norm": 0.15659832954406738,
      "learning_rate": 8.890728984573865e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8440040722489357,
      "num_tokens": 948201069.0,
      "step": 29709
    },
    {
      "entropy": 0.5077698063105345,
      "epoch": 2.733474793129424,
      "grad_norm": 0.1586487889289856,
      "learning_rate": 8.887662158432238e-06,
      "loss": 0.4952,
      "mean_token_accuracy": 0.840911254286766,
      "num_tokens": 948233226.0,
      "step": 29710
    },
    {
      "entropy": 0.5772319613024592,
      "epoch": 2.7335667995008652,
      "grad_norm": 0.16661123931407928,
      "learning_rate": 8.884595332290613e-06,
      "loss": 0.5724,
      "mean_token_accuracy": 0.8208019435405731,
      "num_tokens": 948265337.0,
      "step": 29711
    },
    {
      "entropy": 0.45450114738196135,
      "epoch": 2.733658805872307,
      "grad_norm": 0.15197837352752686,
      "learning_rate": 8.881528506148987e-06,
      "loss": 0.4342,
      "mean_token_accuracy": 0.8555973283946514,
      "num_tokens": 948297414.0,
      "step": 29712
    },
    {
      "entropy": 0.46844939794391394,
      "epoch": 2.733750812243748,
      "grad_norm": 0.1582678109407425,
      "learning_rate": 8.87846168000736e-06,
      "loss": 0.4591,
      "mean_token_accuracy": 0.8543141931295395,
      "num_tokens": 948329132.0,
      "step": 29713
    },
    {
      "entropy": 0.3771265564719215,
      "epoch": 2.733842818615189,
      "grad_norm": 0.1443898230791092,
      "learning_rate": 8.875394853865735e-06,
      "loss": 0.3666,
      "mean_token_accuracy": 0.8809162527322769,
      "num_tokens": 948360728.0,
      "step": 29714
    },
    {
      "entropy": 0.6520879659801722,
      "epoch": 2.73393482498663,
      "grad_norm": 0.17005608975887299,
      "learning_rate": 8.872328027724108e-06,
      "loss": 0.6401,
      "mean_token_accuracy": 0.8012840338051319,
      "num_tokens": 948393377.0,
      "step": 29715
    },
    {
      "entropy": 0.4702159631997347,
      "epoch": 2.7340268313580713,
      "grad_norm": 0.15159469842910767,
      "learning_rate": 8.869261201582483e-06,
      "loss": 0.4636,
      "mean_token_accuracy": 0.8519427888095379,
      "num_tokens": 948425208.0,
      "step": 29716
    },
    {
      "entropy": 0.5341153517365456,
      "epoch": 2.734118837729513,
      "grad_norm": 0.16582848131656647,
      "learning_rate": 8.866194375440856e-06,
      "loss": 0.5348,
      "mean_token_accuracy": 0.833034098148346,
      "num_tokens": 948457256.0,
      "step": 29717
    },
    {
      "entropy": 0.48267011693678796,
      "epoch": 2.734210844100954,
      "grad_norm": 0.1558607965707779,
      "learning_rate": 8.863127549299232e-06,
      "loss": 0.4716,
      "mean_token_accuracy": 0.8520090505480766,
      "num_tokens": 948489426.0,
      "step": 29718
    },
    {
      "entropy": 0.5500828567892313,
      "epoch": 2.734302850472395,
      "grad_norm": 0.16568076610565186,
      "learning_rate": 8.860060723157605e-06,
      "loss": 0.5452,
      "mean_token_accuracy": 0.8298346325755119,
      "num_tokens": 948521468.0,
      "step": 29719
    },
    {
      "entropy": 0.49002915155142546,
      "epoch": 2.7343948568438363,
      "grad_norm": 0.15762795507907867,
      "learning_rate": 8.85699389701598e-06,
      "loss": 0.48,
      "mean_token_accuracy": 0.8458360284566879,
      "num_tokens": 948552965.0,
      "step": 29720
    },
    {
      "entropy": 0.40952274575829506,
      "epoch": 2.7344868632152775,
      "grad_norm": 0.14609012007713318,
      "learning_rate": 8.853927070874353e-06,
      "loss": 0.393,
      "mean_token_accuracy": 0.8705171756446362,
      "num_tokens": 948583929.0,
      "step": 29721
    },
    {
      "entropy": 0.5015375185757875,
      "epoch": 2.734578869586719,
      "grad_norm": 0.15915794670581818,
      "learning_rate": 8.850860244732727e-06,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8454880453646183,
      "num_tokens": 948615379.0,
      "step": 29722
    },
    {
      "entropy": 0.5630862824618816,
      "epoch": 2.73467087595816,
      "grad_norm": 0.162466898560524,
      "learning_rate": 8.8477934185911e-06,
      "loss": 0.5508,
      "mean_token_accuracy": 0.8244043663144112,
      "num_tokens": 948646975.0,
      "step": 29723
    },
    {
      "entropy": 0.4876579977571964,
      "epoch": 2.7347628823296013,
      "grad_norm": 0.15589261054992676,
      "learning_rate": 8.844726592449475e-06,
      "loss": 0.472,
      "mean_token_accuracy": 0.8480603136122227,
      "num_tokens": 948678070.0,
      "step": 29724
    },
    {
      "entropy": 0.50002544245217,
      "epoch": 2.7348548887010424,
      "grad_norm": 0.15871009230613708,
      "learning_rate": 8.841659766307848e-06,
      "loss": 0.4941,
      "mean_token_accuracy": 0.8414862640202045,
      "num_tokens": 948710174.0,
      "step": 29725
    },
    {
      "entropy": 0.6030991058796644,
      "epoch": 2.7349468950724836,
      "grad_norm": 0.16913099586963654,
      "learning_rate": 8.838592940166222e-06,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8122539035975933,
      "num_tokens": 948742458.0,
      "step": 29726
    },
    {
      "entropy": 0.5123933907598257,
      "epoch": 2.735038901443925,
      "grad_norm": 0.16436265408992767,
      "learning_rate": 8.835526114024597e-06,
      "loss": 0.5059,
      "mean_token_accuracy": 0.8403882943093777,
      "num_tokens": 948773664.0,
      "step": 29727
    },
    {
      "entropy": 0.45256040059030056,
      "epoch": 2.7351309078153663,
      "grad_norm": 0.15249809622764587,
      "learning_rate": 8.83245928788297e-06,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8579544834792614,
      "num_tokens": 948805838.0,
      "step": 29728
    },
    {
      "entropy": 0.5103991902433336,
      "epoch": 2.7352229141868074,
      "grad_norm": 0.16238893568515778,
      "learning_rate": 8.829392461741345e-06,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8418130576610565,
      "num_tokens": 948837268.0,
      "step": 29729
    },
    {
      "entropy": 0.43892902962397784,
      "epoch": 2.7353149205582485,
      "grad_norm": 0.15477997064590454,
      "learning_rate": 8.826325635599718e-06,
      "loss": 0.4263,
      "mean_token_accuracy": 0.8634524047374725,
      "num_tokens": 948867680.0,
      "step": 29730
    },
    {
      "entropy": 0.4813126539811492,
      "epoch": 2.7354069269296897,
      "grad_norm": 0.14998285472393036,
      "learning_rate": 8.823258809458092e-06,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8549396209418774,
      "num_tokens": 948899795.0,
      "step": 29731
    },
    {
      "entropy": 0.6173803005367517,
      "epoch": 2.7354989333011313,
      "grad_norm": 0.17155764997005463,
      "learning_rate": 8.820191983316465e-06,
      "loss": 0.6076,
      "mean_token_accuracy": 0.8072191029787064,
      "num_tokens": 948931325.0,
      "step": 29732
    },
    {
      "entropy": 0.5580280935391784,
      "epoch": 2.7355909396725724,
      "grad_norm": 0.1617814302444458,
      "learning_rate": 8.81712515717484e-06,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8296086937189102,
      "num_tokens": 948963407.0,
      "step": 29733
    },
    {
      "entropy": 0.5761484047397971,
      "epoch": 2.7356829460440135,
      "grad_norm": 0.17121823132038116,
      "learning_rate": 8.814058331033214e-06,
      "loss": 0.5673,
      "mean_token_accuracy": 0.8189651221036911,
      "num_tokens": 948994867.0,
      "step": 29734
    },
    {
      "entropy": 0.6046598041430116,
      "epoch": 2.7357749524154547,
      "grad_norm": 0.17239098250865936,
      "learning_rate": 8.810991504891589e-06,
      "loss": 0.6005,
      "mean_token_accuracy": 0.8124038018286228,
      "num_tokens": 949026360.0,
      "step": 29735
    },
    {
      "entropy": 0.4984410787001252,
      "epoch": 2.735866958786896,
      "grad_norm": 0.15488840639591217,
      "learning_rate": 8.807924678749962e-06,
      "loss": 0.4841,
      "mean_token_accuracy": 0.8442746102809906,
      "num_tokens": 949058164.0,
      "step": 29736
    },
    {
      "entropy": 0.50642299791798,
      "epoch": 2.7359589651583374,
      "grad_norm": 0.15562666952610016,
      "learning_rate": 8.804857852608337e-06,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8441889360547066,
      "num_tokens": 949090743.0,
      "step": 29737
    },
    {
      "entropy": 0.48437863611616194,
      "epoch": 2.7360509715297785,
      "grad_norm": 0.15529213845729828,
      "learning_rate": 8.80179102646671e-06,
      "loss": 0.4644,
      "mean_token_accuracy": 0.8497821502387524,
      "num_tokens": 949122925.0,
      "step": 29738
    },
    {
      "entropy": 0.46888180542737246,
      "epoch": 2.7361429779012196,
      "grad_norm": 0.15111590921878815,
      "learning_rate": 8.798724200325084e-06,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8546062186360359,
      "num_tokens": 949155357.0,
      "step": 29739
    },
    {
      "entropy": 0.4251591460779309,
      "epoch": 2.7362349842726608,
      "grad_norm": 0.14615067839622498,
      "learning_rate": 8.795657374183457e-06,
      "loss": 0.409,
      "mean_token_accuracy": 0.8654298968613148,
      "num_tokens": 949186736.0,
      "step": 29740
    },
    {
      "entropy": 0.566661411896348,
      "epoch": 2.736326990644102,
      "grad_norm": 0.1622890681028366,
      "learning_rate": 8.792590548041832e-06,
      "loss": 0.5424,
      "mean_token_accuracy": 0.8301825933158398,
      "num_tokens": 949218494.0,
      "step": 29741
    },
    {
      "entropy": 0.5251003657467663,
      "epoch": 2.7364189970155435,
      "grad_norm": 0.15896426141262054,
      "learning_rate": 8.789523721900207e-06,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8410209305584431,
      "num_tokens": 949250630.0,
      "step": 29742
    },
    {
      "entropy": 0.5956835187971592,
      "epoch": 2.7365110033869846,
      "grad_norm": 0.17447932064533234,
      "learning_rate": 8.78645689575858e-06,
      "loss": 0.5851,
      "mean_token_accuracy": 0.8145308457314968,
      "num_tokens": 949282223.0,
      "step": 29743
    },
    {
      "entropy": 0.5379074215888977,
      "epoch": 2.7366030097584257,
      "grad_norm": 0.15924303233623505,
      "learning_rate": 8.783390069616954e-06,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8364526852965355,
      "num_tokens": 949313768.0,
      "step": 29744
    },
    {
      "entropy": 0.5110720186494291,
      "epoch": 2.736695016129867,
      "grad_norm": 0.15512362122535706,
      "learning_rate": 8.780323243475327e-06,
      "loss": 0.4992,
      "mean_token_accuracy": 0.839885950088501,
      "num_tokens": 949346111.0,
      "step": 29745
    },
    {
      "entropy": 0.4914244554238394,
      "epoch": 2.736787022501308,
      "grad_norm": 0.15787605941295624,
      "learning_rate": 8.777256417333702e-06,
      "loss": 0.4868,
      "mean_token_accuracy": 0.8439326249063015,
      "num_tokens": 949378086.0,
      "step": 29746
    },
    {
      "entropy": 0.4943243721500039,
      "epoch": 2.7368790288727496,
      "grad_norm": 0.15682265162467957,
      "learning_rate": 8.774189591192075e-06,
      "loss": 0.487,
      "mean_token_accuracy": 0.8436163365840912,
      "num_tokens": 949409738.0,
      "step": 29747
    },
    {
      "entropy": 0.7021925076842308,
      "epoch": 2.7369710352441907,
      "grad_norm": 0.18535901606082916,
      "learning_rate": 8.77112276505045e-06,
      "loss": 0.6976,
      "mean_token_accuracy": 0.7822442799806595,
      "num_tokens": 949440632.0,
      "step": 29748
    },
    {
      "entropy": 0.5078120147809386,
      "epoch": 2.737063041615632,
      "grad_norm": 0.15920060873031616,
      "learning_rate": 8.768055938908824e-06,
      "loss": 0.4875,
      "mean_token_accuracy": 0.8448071703314781,
      "num_tokens": 949472489.0,
      "step": 29749
    },
    {
      "entropy": 0.5278116384288296,
      "epoch": 2.737155047987073,
      "grad_norm": 0.19658522307872772,
      "learning_rate": 8.764989112767199e-06,
      "loss": 0.5102,
      "mean_token_accuracy": 0.83689871057868,
      "num_tokens": 949503865.0,
      "step": 29750
    },
    {
      "entropy": 0.5002274033613503,
      "epoch": 2.737247054358514,
      "grad_norm": 0.155319944024086,
      "learning_rate": 8.761922286625572e-06,
      "loss": 0.4905,
      "mean_token_accuracy": 0.846098106354475,
      "num_tokens": 949535922.0,
      "step": 29751
    },
    {
      "entropy": 0.6396995382383466,
      "epoch": 2.7373390607299557,
      "grad_norm": 0.17142027616500854,
      "learning_rate": 8.758855460483946e-06,
      "loss": 0.6295,
      "mean_token_accuracy": 0.8047962598502636,
      "num_tokens": 949568141.0,
      "step": 29752
    },
    {
      "entropy": 0.5070598069578409,
      "epoch": 2.737431067101397,
      "grad_norm": 0.15862925350666046,
      "learning_rate": 8.75578863434232e-06,
      "loss": 0.4948,
      "mean_token_accuracy": 0.84397342056036,
      "num_tokens": 949599927.0,
      "step": 29753
    },
    {
      "entropy": 0.41775191109627485,
      "epoch": 2.737523073472838,
      "grad_norm": 0.1473851352930069,
      "learning_rate": 8.752721808200694e-06,
      "loss": 0.4077,
      "mean_token_accuracy": 0.8684256747364998,
      "num_tokens": 949631489.0,
      "step": 29754
    },
    {
      "entropy": 0.508635482750833,
      "epoch": 2.737615079844279,
      "grad_norm": 0.15490971505641937,
      "learning_rate": 8.749654982059067e-06,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8459382094442844,
      "num_tokens": 949663233.0,
      "step": 29755
    },
    {
      "entropy": 0.5193772744387388,
      "epoch": 2.7377070862157202,
      "grad_norm": 0.16357359290122986,
      "learning_rate": 8.746588155917441e-06,
      "loss": 0.5097,
      "mean_token_accuracy": 0.8391096256673336,
      "num_tokens": 949694723.0,
      "step": 29756
    },
    {
      "entropy": 0.5184723809361458,
      "epoch": 2.737799092587162,
      "grad_norm": 0.15812785923480988,
      "learning_rate": 8.743521329775816e-06,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8414072245359421,
      "num_tokens": 949726900.0,
      "step": 29757
    },
    {
      "entropy": 0.5216969978064299,
      "epoch": 2.737891098958603,
      "grad_norm": 0.16108869016170502,
      "learning_rate": 8.740454503634189e-06,
      "loss": 0.511,
      "mean_token_accuracy": 0.83957190066576,
      "num_tokens": 949758280.0,
      "step": 29758
    },
    {
      "entropy": 0.4976958893239498,
      "epoch": 2.737983105330044,
      "grad_norm": 0.1606035828590393,
      "learning_rate": 8.737387677492564e-06,
      "loss": 0.4875,
      "mean_token_accuracy": 0.8458096161484718,
      "num_tokens": 949789830.0,
      "step": 29759
    },
    {
      "entropy": 0.5206500692293048,
      "epoch": 2.738075111701485,
      "grad_norm": 0.16381175816059113,
      "learning_rate": 8.734320851350937e-06,
      "loss": 0.5111,
      "mean_token_accuracy": 0.8389561399817467,
      "num_tokens": 949821472.0,
      "step": 29760
    },
    {
      "entropy": 0.5352764381095767,
      "epoch": 2.7381671180729263,
      "grad_norm": 0.1599886119365692,
      "learning_rate": 8.731254025209311e-06,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8316836804151535,
      "num_tokens": 949853014.0,
      "step": 29761
    },
    {
      "entropy": 0.4357988827396184,
      "epoch": 2.738259124444368,
      "grad_norm": 0.14565686881542206,
      "learning_rate": 8.728187199067684e-06,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8648367524147034,
      "num_tokens": 949884644.0,
      "step": 29762
    },
    {
      "entropy": 0.4722837167792022,
      "epoch": 2.738351130815809,
      "grad_norm": 0.15063776075839996,
      "learning_rate": 8.725120372926059e-06,
      "loss": 0.4621,
      "mean_token_accuracy": 0.857559148222208,
      "num_tokens": 949916760.0,
      "step": 29763
    },
    {
      "entropy": 0.4965417878702283,
      "epoch": 2.73844313718725,
      "grad_norm": 0.15401345491409302,
      "learning_rate": 8.722053546784434e-06,
      "loss": 0.4892,
      "mean_token_accuracy": 0.8449222892522812,
      "num_tokens": 949948712.0,
      "step": 29764
    },
    {
      "entropy": 0.4776016091927886,
      "epoch": 2.7385351435586913,
      "grad_norm": 0.1513943374156952,
      "learning_rate": 8.718986720642808e-06,
      "loss": 0.4624,
      "mean_token_accuracy": 0.8514331877231598,
      "num_tokens": 949980662.0,
      "step": 29765
    },
    {
      "entropy": 0.49673878960311413,
      "epoch": 2.7386271499301325,
      "grad_norm": 0.15532302856445312,
      "learning_rate": 8.715919894501181e-06,
      "loss": 0.4842,
      "mean_token_accuracy": 0.8472764976322651,
      "num_tokens": 950013430.0,
      "step": 29766
    },
    {
      "entropy": 0.3987413505092263,
      "epoch": 2.738719156301574,
      "grad_norm": 0.14307580888271332,
      "learning_rate": 8.712853068359556e-06,
      "loss": 0.3801,
      "mean_token_accuracy": 0.8768569193780422,
      "num_tokens": 950045141.0,
      "step": 29767
    },
    {
      "entropy": 0.44978486048057675,
      "epoch": 2.738811162673015,
      "grad_norm": 0.14943525195121765,
      "learning_rate": 8.709786242217929e-06,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8636938482522964,
      "num_tokens": 950077283.0,
      "step": 29768
    },
    {
      "entropy": 0.5797551330178976,
      "epoch": 2.7389031690444563,
      "grad_norm": 0.1692013442516327,
      "learning_rate": 8.706719416076303e-06,
      "loss": 0.5767,
      "mean_token_accuracy": 0.822377547621727,
      "num_tokens": 950109352.0,
      "step": 29769
    },
    {
      "entropy": 0.5465391222387552,
      "epoch": 2.7389951754158974,
      "grad_norm": 0.16675549745559692,
      "learning_rate": 8.703652589934676e-06,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8327789008617401,
      "num_tokens": 950141649.0,
      "step": 29770
    },
    {
      "entropy": 0.4624445326626301,
      "epoch": 2.7390871817873386,
      "grad_norm": 0.1575317680835724,
      "learning_rate": 8.700585763793051e-06,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8559983596205711,
      "num_tokens": 950173178.0,
      "step": 29771
    },
    {
      "entropy": 0.42594234831631184,
      "epoch": 2.73917918815878,
      "grad_norm": 0.14489507675170898,
      "learning_rate": 8.697518937651426e-06,
      "loss": 0.404,
      "mean_token_accuracy": 0.8678559064865112,
      "num_tokens": 950204999.0,
      "step": 29772
    },
    {
      "entropy": 0.4284608233720064,
      "epoch": 2.7392711945302213,
      "grad_norm": 0.14953382313251495,
      "learning_rate": 8.694452111509799e-06,
      "loss": 0.4125,
      "mean_token_accuracy": 0.8683576360344887,
      "num_tokens": 950237108.0,
      "step": 29773
    },
    {
      "entropy": 0.45553160458803177,
      "epoch": 2.7393632009016624,
      "grad_norm": 0.15483097732067108,
      "learning_rate": 8.691385285368173e-06,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8601999208331108,
      "num_tokens": 950268143.0,
      "step": 29774
    },
    {
      "entropy": 0.6304994728416204,
      "epoch": 2.7394552072731035,
      "grad_norm": 0.16802816092967987,
      "learning_rate": 8.688318459226546e-06,
      "loss": 0.6137,
      "mean_token_accuracy": 0.8093439936637878,
      "num_tokens": 950300401.0,
      "step": 29775
    },
    {
      "entropy": 0.5356822777539492,
      "epoch": 2.7395472136445447,
      "grad_norm": 0.15965233743190765,
      "learning_rate": 8.685251633084921e-06,
      "loss": 0.5285,
      "mean_token_accuracy": 0.8322530016303062,
      "num_tokens": 950332286.0,
      "step": 29776
    },
    {
      "entropy": 0.5074060028418899,
      "epoch": 2.7396392200159863,
      "grad_norm": 0.15723563730716705,
      "learning_rate": 8.682184806943294e-06,
      "loss": 0.4969,
      "mean_token_accuracy": 0.8442843109369278,
      "num_tokens": 950365054.0,
      "step": 29777
    },
    {
      "entropy": 0.5590926837176085,
      "epoch": 2.7397312263874274,
      "grad_norm": 0.16088365018367767,
      "learning_rate": 8.679117980801668e-06,
      "loss": 0.5493,
      "mean_token_accuracy": 0.8275367580354214,
      "num_tokens": 950397798.0,
      "step": 29778
    },
    {
      "entropy": 0.5865186117589474,
      "epoch": 2.7398232327588685,
      "grad_norm": 0.167989119887352,
      "learning_rate": 8.676051154660043e-06,
      "loss": 0.5893,
      "mean_token_accuracy": 0.8168699406087399,
      "num_tokens": 950430145.0,
      "step": 29779
    },
    {
      "entropy": 0.4315749607048929,
      "epoch": 2.7399152391303097,
      "grad_norm": 0.14247071743011475,
      "learning_rate": 8.672984328518418e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8667731992900372,
      "num_tokens": 950462681.0,
      "step": 29780
    },
    {
      "entropy": 0.48795633716508746,
      "epoch": 2.740007245501751,
      "grad_norm": 0.15490563213825226,
      "learning_rate": 8.66991750237679e-06,
      "loss": 0.471,
      "mean_token_accuracy": 0.8488265462219715,
      "num_tokens": 950494893.0,
      "step": 29781
    },
    {
      "entropy": 0.5417912444099784,
      "epoch": 2.7400992518731924,
      "grad_norm": 0.16475032269954681,
      "learning_rate": 8.666850676235165e-06,
      "loss": 0.5213,
      "mean_token_accuracy": 0.8374390564858913,
      "num_tokens": 950525689.0,
      "step": 29782
    },
    {
      "entropy": 0.5469980230554938,
      "epoch": 2.7401912582446335,
      "grad_norm": 0.16867457330226898,
      "learning_rate": 8.663783850093538e-06,
      "loss": 0.5407,
      "mean_token_accuracy": 0.8291767500340939,
      "num_tokens": 950558158.0,
      "step": 29783
    },
    {
      "entropy": 0.5526163908652961,
      "epoch": 2.7402832646160746,
      "grad_norm": 0.16582666337490082,
      "learning_rate": 8.660717023951913e-06,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8275523446500301,
      "num_tokens": 950590475.0,
      "step": 29784
    },
    {
      "entropy": 0.486100060865283,
      "epoch": 2.7403752709875158,
      "grad_norm": 0.151006817817688,
      "learning_rate": 8.657650197810286e-06,
      "loss": 0.4729,
      "mean_token_accuracy": 0.8511312827467918,
      "num_tokens": 950622712.0,
      "step": 29785
    },
    {
      "entropy": 0.49382913392037153,
      "epoch": 2.740467277358957,
      "grad_norm": 0.15331199765205383,
      "learning_rate": 8.65458337166866e-06,
      "loss": 0.4842,
      "mean_token_accuracy": 0.8464754559099674,
      "num_tokens": 950655309.0,
      "step": 29786
    },
    {
      "entropy": 0.5699039734899998,
      "epoch": 2.7405592837303985,
      "grad_norm": 0.1661149561405182,
      "learning_rate": 8.651516545527035e-06,
      "loss": 0.5515,
      "mean_token_accuracy": 0.8240410722792149,
      "num_tokens": 950687297.0,
      "step": 29787
    },
    {
      "entropy": 0.4212853880599141,
      "epoch": 2.7406512901018396,
      "grad_norm": 0.14403675496578217,
      "learning_rate": 8.648449719385408e-06,
      "loss": 0.41,
      "mean_token_accuracy": 0.8680632635951042,
      "num_tokens": 950719380.0,
      "step": 29788
    },
    {
      "entropy": 0.6275374984834343,
      "epoch": 2.7407432964732807,
      "grad_norm": 0.17950782179832458,
      "learning_rate": 8.645382893243783e-06,
      "loss": 0.6088,
      "mean_token_accuracy": 0.8102058172225952,
      "num_tokens": 950751141.0,
      "step": 29789
    },
    {
      "entropy": 0.518241023644805,
      "epoch": 2.740835302844722,
      "grad_norm": 0.15293507277965546,
      "learning_rate": 8.642316067102156e-06,
      "loss": 0.5052,
      "mean_token_accuracy": 0.8390567898750305,
      "num_tokens": 950783505.0,
      "step": 29790
    },
    {
      "entropy": 0.43557193875312805,
      "epoch": 2.740927309216163,
      "grad_norm": 0.14542019367218018,
      "learning_rate": 8.63924924096053e-06,
      "loss": 0.4219,
      "mean_token_accuracy": 0.862947303801775,
      "num_tokens": 950815970.0,
      "step": 29791
    },
    {
      "entropy": 0.5507159605622292,
      "epoch": 2.7410193155876046,
      "grad_norm": 0.1668883115053177,
      "learning_rate": 8.636182414818903e-06,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8261811211705208,
      "num_tokens": 950847958.0,
      "step": 29792
    },
    {
      "entropy": 0.5481104422360659,
      "epoch": 2.7411113219590457,
      "grad_norm": 0.16097183525562286,
      "learning_rate": 8.633115588677278e-06,
      "loss": 0.5312,
      "mean_token_accuracy": 0.8326223790645599,
      "num_tokens": 950879572.0,
      "step": 29793
    },
    {
      "entropy": 0.4134828308597207,
      "epoch": 2.741203328330487,
      "grad_norm": 0.14108514785766602,
      "learning_rate": 8.630048762535651e-06,
      "loss": 0.395,
      "mean_token_accuracy": 0.8760953396558762,
      "num_tokens": 950911876.0,
      "step": 29794
    },
    {
      "entropy": 0.5849418118596077,
      "epoch": 2.741295334701928,
      "grad_norm": 0.17262206971645355,
      "learning_rate": 8.626981936394027e-06,
      "loss": 0.5625,
      "mean_token_accuracy": 0.8245776407420635,
      "num_tokens": 950944005.0,
      "step": 29795
    },
    {
      "entropy": 0.5816346239298582,
      "epoch": 2.741387341073369,
      "grad_norm": 0.16631831228733063,
      "learning_rate": 8.6239151102524e-06,
      "loss": 0.5711,
      "mean_token_accuracy": 0.818681787699461,
      "num_tokens": 950976283.0,
      "step": 29796
    },
    {
      "entropy": 0.5370321981608868,
      "epoch": 2.7414793474448107,
      "grad_norm": 0.16019798815250397,
      "learning_rate": 8.620848284110775e-06,
      "loss": 0.5243,
      "mean_token_accuracy": 0.8365601561963558,
      "num_tokens": 951007921.0,
      "step": 29797
    },
    {
      "entropy": 0.5428119776770473,
      "epoch": 2.741571353816252,
      "grad_norm": 0.16359782218933105,
      "learning_rate": 8.617781457969148e-06,
      "loss": 0.5412,
      "mean_token_accuracy": 0.826892789453268,
      "num_tokens": 951040456.0,
      "step": 29798
    },
    {
      "entropy": 0.4877526720520109,
      "epoch": 2.741663360187693,
      "grad_norm": 0.15292952954769135,
      "learning_rate": 8.614714631827522e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8503101579844952,
      "num_tokens": 951072328.0,
      "step": 29799
    },
    {
      "entropy": 0.5024811350740492,
      "epoch": 2.741755366559134,
      "grad_norm": 0.15402616560459137,
      "learning_rate": 8.611647805685895e-06,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8398952186107635,
      "num_tokens": 951104658.0,
      "step": 29800
    },
    {
      "entropy": 0.5766304852440953,
      "epoch": 2.7418473729305752,
      "grad_norm": 0.1668563336133957,
      "learning_rate": 8.60858097954427e-06,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8232330940663815,
      "num_tokens": 951136246.0,
      "step": 29801
    },
    {
      "entropy": 0.6106442902237177,
      "epoch": 2.741939379302017,
      "grad_norm": 0.1675887256860733,
      "learning_rate": 8.605514153402645e-06,
      "loss": 0.5941,
      "mean_token_accuracy": 0.8125488348305225,
      "num_tokens": 951168700.0,
      "step": 29802
    },
    {
      "entropy": 0.5496670491993427,
      "epoch": 2.742031385673458,
      "grad_norm": 0.160951629281044,
      "learning_rate": 8.602447327261018e-06,
      "loss": 0.5367,
      "mean_token_accuracy": 0.8287177234888077,
      "num_tokens": 951200345.0,
      "step": 29803
    },
    {
      "entropy": 0.49490869883447886,
      "epoch": 2.742123392044899,
      "grad_norm": 0.1562563180923462,
      "learning_rate": 8.599380501119392e-06,
      "loss": 0.4741,
      "mean_token_accuracy": 0.8472774885594845,
      "num_tokens": 951232139.0,
      "step": 29804
    },
    {
      "entropy": 0.5193608738481998,
      "epoch": 2.74221539841634,
      "grad_norm": 0.1601220965385437,
      "learning_rate": 8.596313674977765e-06,
      "loss": 0.5081,
      "mean_token_accuracy": 0.8379615508019924,
      "num_tokens": 951264756.0,
      "step": 29805
    },
    {
      "entropy": 0.475275800563395,
      "epoch": 2.7423074047877813,
      "grad_norm": 0.1550729125738144,
      "learning_rate": 8.59324684883614e-06,
      "loss": 0.4551,
      "mean_token_accuracy": 0.8560411632061005,
      "num_tokens": 951297213.0,
      "step": 29806
    },
    {
      "entropy": 0.5710607413202524,
      "epoch": 2.742399411159223,
      "grad_norm": 0.1660557985305786,
      "learning_rate": 8.590180022694513e-06,
      "loss": 0.5686,
      "mean_token_accuracy": 0.8226867988705635,
      "num_tokens": 951328847.0,
      "step": 29807
    },
    {
      "entropy": 0.5473036961629987,
      "epoch": 2.742491417530664,
      "grad_norm": 0.16090825200080872,
      "learning_rate": 8.587113196552888e-06,
      "loss": 0.5357,
      "mean_token_accuracy": 0.8315405398607254,
      "num_tokens": 951360875.0,
      "step": 29808
    },
    {
      "entropy": 0.5520725265378132,
      "epoch": 2.742583423902105,
      "grad_norm": 0.16410641372203827,
      "learning_rate": 8.58404637041126e-06,
      "loss": 0.5484,
      "mean_token_accuracy": 0.8264982998371124,
      "num_tokens": 951392810.0,
      "step": 29809
    },
    {
      "entropy": 0.3970319477375597,
      "epoch": 2.7426754302735463,
      "grad_norm": 0.14071689546108246,
      "learning_rate": 8.580979544269637e-06,
      "loss": 0.3912,
      "mean_token_accuracy": 0.8735337518155575,
      "num_tokens": 951424359.0,
      "step": 29810
    },
    {
      "entropy": 0.5945173371583223,
      "epoch": 2.7427674366449875,
      "grad_norm": 0.1715121567249298,
      "learning_rate": 8.57791271812801e-06,
      "loss": 0.5828,
      "mean_token_accuracy": 0.8186716102063656,
      "num_tokens": 951456250.0,
      "step": 29811
    },
    {
      "entropy": 0.43948868894949555,
      "epoch": 2.742859443016429,
      "grad_norm": 0.1428239494562149,
      "learning_rate": 8.574845891986384e-06,
      "loss": 0.4225,
      "mean_token_accuracy": 0.8627882041037083,
      "num_tokens": 951488331.0,
      "step": 29812
    },
    {
      "entropy": 0.5358815337531269,
      "epoch": 2.74295144938787,
      "grad_norm": 0.15889889001846313,
      "learning_rate": 8.571779065844757e-06,
      "loss": 0.5283,
      "mean_token_accuracy": 0.8355214931070805,
      "num_tokens": 951521099.0,
      "step": 29813
    },
    {
      "entropy": 0.5337805254384875,
      "epoch": 2.7430434557593113,
      "grad_norm": 0.16564732789993286,
      "learning_rate": 8.568712239703132e-06,
      "loss": 0.5266,
      "mean_token_accuracy": 0.835698164999485,
      "num_tokens": 951552673.0,
      "step": 29814
    },
    {
      "entropy": 0.5938585624098778,
      "epoch": 2.7431354621307524,
      "grad_norm": 0.1641736775636673,
      "learning_rate": 8.565645413561505e-06,
      "loss": 0.5745,
      "mean_token_accuracy": 0.8230671137571335,
      "num_tokens": 951585096.0,
      "step": 29815
    },
    {
      "entropy": 0.5583896404132247,
      "epoch": 2.7432274685021936,
      "grad_norm": 0.1688401699066162,
      "learning_rate": 8.56257858741988e-06,
      "loss": 0.5388,
      "mean_token_accuracy": 0.8246811963617802,
      "num_tokens": 951615461.0,
      "step": 29816
    },
    {
      "entropy": 0.6342075746506453,
      "epoch": 2.743319474873635,
      "grad_norm": 0.17831368744373322,
      "learning_rate": 8.559511761278254e-06,
      "loss": 0.6154,
      "mean_token_accuracy": 0.8060485310852528,
      "num_tokens": 951646748.0,
      "step": 29817
    },
    {
      "entropy": 0.446335406973958,
      "epoch": 2.7434114812450763,
      "grad_norm": 0.15236543118953705,
      "learning_rate": 8.556444935136627e-06,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8597828447818756,
      "num_tokens": 951678361.0,
      "step": 29818
    },
    {
      "entropy": 0.444293767795898,
      "epoch": 2.7435034876165174,
      "grad_norm": 0.14956897497177124,
      "learning_rate": 8.553378108995002e-06,
      "loss": 0.4316,
      "mean_token_accuracy": 0.861313421279192,
      "num_tokens": 951710369.0,
      "step": 29819
    },
    {
      "entropy": 0.49820326268672943,
      "epoch": 2.7435954939879585,
      "grad_norm": 0.14822177588939667,
      "learning_rate": 8.550311282853375e-06,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8511219061911106,
      "num_tokens": 951742762.0,
      "step": 29820
    },
    {
      "entropy": 0.46298938523977995,
      "epoch": 2.7436875003593997,
      "grad_norm": 0.14613881707191467,
      "learning_rate": 8.54724445671175e-06,
      "loss": 0.4536,
      "mean_token_accuracy": 0.8555386923253536,
      "num_tokens": 951774539.0,
      "step": 29821
    },
    {
      "entropy": 0.4809515653178096,
      "epoch": 2.7437795067308413,
      "grad_norm": 0.15036801993846893,
      "learning_rate": 8.544177630570122e-06,
      "loss": 0.4674,
      "mean_token_accuracy": 0.8499081917107105,
      "num_tokens": 951806582.0,
      "step": 29822
    },
    {
      "entropy": 0.4999999813735485,
      "epoch": 2.7438715131022824,
      "grad_norm": 0.1578885167837143,
      "learning_rate": 8.541110804428497e-06,
      "loss": 0.493,
      "mean_token_accuracy": 0.8437510579824448,
      "num_tokens": 951838472.0,
      "step": 29823
    },
    {
      "entropy": 0.4678043080493808,
      "epoch": 2.7439635194737235,
      "grad_norm": 0.15548323094844818,
      "learning_rate": 8.53804397828687e-06,
      "loss": 0.4559,
      "mean_token_accuracy": 0.8525876067578793,
      "num_tokens": 951870169.0,
      "step": 29824
    },
    {
      "entropy": 0.5137793319299817,
      "epoch": 2.7440555258451647,
      "grad_norm": 0.15954971313476562,
      "learning_rate": 8.534977152145246e-06,
      "loss": 0.5038,
      "mean_token_accuracy": 0.841392956674099,
      "num_tokens": 951902742.0,
      "step": 29825
    },
    {
      "entropy": 0.49813865311443806,
      "epoch": 2.744147532216606,
      "grad_norm": 0.15691590309143066,
      "learning_rate": 8.53191032600362e-06,
      "loss": 0.4974,
      "mean_token_accuracy": 0.8423232063651085,
      "num_tokens": 951934686.0,
      "step": 29826
    },
    {
      "entropy": 0.4511820753104985,
      "epoch": 2.7442395385880474,
      "grad_norm": 0.14414982497692108,
      "learning_rate": 8.528843499861994e-06,
      "loss": 0.4354,
      "mean_token_accuracy": 0.8588355928659439,
      "num_tokens": 951966887.0,
      "step": 29827
    },
    {
      "entropy": 0.5535839479416609,
      "epoch": 2.7443315449594885,
      "grad_norm": 0.161504864692688,
      "learning_rate": 8.525776673720367e-06,
      "loss": 0.536,
      "mean_token_accuracy": 0.8253700733184814,
      "num_tokens": 951998520.0,
      "step": 29828
    },
    {
      "entropy": 0.5809453371912241,
      "epoch": 2.7444235513309296,
      "grad_norm": 0.17767831683158875,
      "learning_rate": 8.522709847578742e-06,
      "loss": 0.5758,
      "mean_token_accuracy": 0.8155608363449574,
      "num_tokens": 952029444.0,
      "step": 29829
    },
    {
      "entropy": 0.5802275231108069,
      "epoch": 2.7445155577023708,
      "grad_norm": 0.16458871960639954,
      "learning_rate": 8.519643021437115e-06,
      "loss": 0.5693,
      "mean_token_accuracy": 0.822482768446207,
      "num_tokens": 952061864.0,
      "step": 29830
    },
    {
      "entropy": 0.49002619134262204,
      "epoch": 2.744607564073812,
      "grad_norm": 0.16071631014347076,
      "learning_rate": 8.51657619529549e-06,
      "loss": 0.4844,
      "mean_token_accuracy": 0.8438724614679813,
      "num_tokens": 952093701.0,
      "step": 29831
    },
    {
      "entropy": 0.4831971190869808,
      "epoch": 2.7446995704452535,
      "grad_norm": 0.15958617627620697,
      "learning_rate": 8.513509369153864e-06,
      "loss": 0.4729,
      "mean_token_accuracy": 0.8511350713670254,
      "num_tokens": 952124328.0,
      "step": 29832
    },
    {
      "entropy": 0.5044770250096917,
      "epoch": 2.7447915768166946,
      "grad_norm": 0.1527552306652069,
      "learning_rate": 8.510442543012237e-06,
      "loss": 0.4882,
      "mean_token_accuracy": 0.8472300358116627,
      "num_tokens": 952156448.0,
      "step": 29833
    },
    {
      "entropy": 0.5208719987422228,
      "epoch": 2.7448835831881357,
      "grad_norm": 0.16126419603824615,
      "learning_rate": 8.507375716870611e-06,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8347701691091061,
      "num_tokens": 952188685.0,
      "step": 29834
    },
    {
      "entropy": 0.38837127946317196,
      "epoch": 2.744975589559577,
      "grad_norm": 0.1424439400434494,
      "learning_rate": 8.504308890728984e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8770179562270641,
      "num_tokens": 952220524.0,
      "step": 29835
    },
    {
      "entropy": 0.5246855267323554,
      "epoch": 2.745067595931018,
      "grad_norm": 0.15890417993068695,
      "learning_rate": 8.501242064587359e-06,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8389492705464363,
      "num_tokens": 952252285.0,
      "step": 29836
    },
    {
      "entropy": 0.5551725588738918,
      "epoch": 2.7451596023024596,
      "grad_norm": 0.16146761178970337,
      "learning_rate": 8.498175238445732e-06,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8293983228504658,
      "num_tokens": 952283392.0,
      "step": 29837
    },
    {
      "entropy": 0.509796941652894,
      "epoch": 2.7452516086739007,
      "grad_norm": 0.1564122438430786,
      "learning_rate": 8.495108412304107e-06,
      "loss": 0.4915,
      "mean_token_accuracy": 0.843498982489109,
      "num_tokens": 952315320.0,
      "step": 29838
    },
    {
      "entropy": 0.5118236988782883,
      "epoch": 2.745343615045342,
      "grad_norm": 0.16056139767169952,
      "learning_rate": 8.49204158616248e-06,
      "loss": 0.507,
      "mean_token_accuracy": 0.8356741033494473,
      "num_tokens": 952346717.0,
      "step": 29839
    },
    {
      "entropy": 0.47022003401070833,
      "epoch": 2.745435621416783,
      "grad_norm": 0.15500885248184204,
      "learning_rate": 8.488974760020856e-06,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8534465581178665,
      "num_tokens": 952378296.0,
      "step": 29840
    },
    {
      "entropy": 0.4732345283264294,
      "epoch": 2.745527627788224,
      "grad_norm": 0.15154309570789337,
      "learning_rate": 8.485907933879229e-06,
      "loss": 0.4588,
      "mean_token_accuracy": 0.8535477109253407,
      "num_tokens": 952410355.0,
      "step": 29841
    },
    {
      "entropy": 0.5463935134466738,
      "epoch": 2.7456196341596657,
      "grad_norm": 0.1650286614894867,
      "learning_rate": 8.482841107737604e-06,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8331177644431591,
      "num_tokens": 952441855.0,
      "step": 29842
    },
    {
      "entropy": 0.4246778469532728,
      "epoch": 2.745711640531107,
      "grad_norm": 0.14875276386737823,
      "learning_rate": 8.479774281595976e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8683489635586739,
      "num_tokens": 952473523.0,
      "step": 29843
    },
    {
      "entropy": 0.41965573793277144,
      "epoch": 2.745803646902548,
      "grad_norm": 0.14325596392154694,
      "learning_rate": 8.476707455454351e-06,
      "loss": 0.4108,
      "mean_token_accuracy": 0.8693655990064144,
      "num_tokens": 952505724.0,
      "step": 29844
    },
    {
      "entropy": 0.49061125703155994,
      "epoch": 2.745895653273989,
      "grad_norm": 0.15463563799858093,
      "learning_rate": 8.473640629312724e-06,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8497263826429844,
      "num_tokens": 952537664.0,
      "step": 29845
    },
    {
      "entropy": 0.4693219354376197,
      "epoch": 2.7459876596454302,
      "grad_norm": 0.15412358939647675,
      "learning_rate": 8.470573803171099e-06,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8524628803133965,
      "num_tokens": 952570050.0,
      "step": 29846
    },
    {
      "entropy": 0.5661384742707014,
      "epoch": 2.746079666016872,
      "grad_norm": 0.1623048484325409,
      "learning_rate": 8.467506977029473e-06,
      "loss": 0.5457,
      "mean_token_accuracy": 0.829676903784275,
      "num_tokens": 952602454.0,
      "step": 29847
    },
    {
      "entropy": 0.557768763974309,
      "epoch": 2.746171672388313,
      "grad_norm": 0.1648833304643631,
      "learning_rate": 8.464440150887846e-06,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8280590325593948,
      "num_tokens": 952633936.0,
      "step": 29848
    },
    {
      "entropy": 0.43711232452187687,
      "epoch": 2.746263678759754,
      "grad_norm": 0.14796023070812225,
      "learning_rate": 8.461373324746221e-06,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8611042201519012,
      "num_tokens": 952666509.0,
      "step": 29849
    },
    {
      "entropy": 0.5307588940486312,
      "epoch": 2.746355685131195,
      "grad_norm": 0.15724295377731323,
      "learning_rate": 8.458306498604594e-06,
      "loss": 0.5176,
      "mean_token_accuracy": 0.835557222366333,
      "num_tokens": 952698670.0,
      "step": 29850
    },
    {
      "entropy": 0.4913252405822277,
      "epoch": 2.7464476915026363,
      "grad_norm": 0.15201154351234436,
      "learning_rate": 8.455239672462969e-06,
      "loss": 0.479,
      "mean_token_accuracy": 0.8467991501092911,
      "num_tokens": 952730885.0,
      "step": 29851
    },
    {
      "entropy": 0.5532092275097966,
      "epoch": 2.746539697874078,
      "grad_norm": 0.16366852819919586,
      "learning_rate": 8.452172846321342e-06,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8296989724040031,
      "num_tokens": 952763412.0,
      "step": 29852
    },
    {
      "entropy": 0.5256673917174339,
      "epoch": 2.746631704245519,
      "grad_norm": 0.16003625094890594,
      "learning_rate": 8.449106020179716e-06,
      "loss": 0.503,
      "mean_token_accuracy": 0.8398831151425838,
      "num_tokens": 952795374.0,
      "step": 29853
    },
    {
      "entropy": 0.5356819955632091,
      "epoch": 2.74672371061696,
      "grad_norm": 0.1612854152917862,
      "learning_rate": 8.446039194038089e-06,
      "loss": 0.529,
      "mean_token_accuracy": 0.8317254446446896,
      "num_tokens": 952827518.0,
      "step": 29854
    },
    {
      "entropy": 0.5348044413840398,
      "epoch": 2.7468157169884013,
      "grad_norm": 0.16727179288864136,
      "learning_rate": 8.442972367896464e-06,
      "loss": 0.5251,
      "mean_token_accuracy": 0.835742462426424,
      "num_tokens": 952858734.0,
      "step": 29855
    },
    {
      "entropy": 0.4606066504493356,
      "epoch": 2.7469077233598425,
      "grad_norm": 0.1524716466665268,
      "learning_rate": 8.439905541754838e-06,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8559241630136967,
      "num_tokens": 952891063.0,
      "step": 29856
    },
    {
      "entropy": 0.5436388868838549,
      "epoch": 2.746999729731284,
      "grad_norm": 0.16324421763420105,
      "learning_rate": 8.436838715613213e-06,
      "loss": 0.5357,
      "mean_token_accuracy": 0.8320626877248287,
      "num_tokens": 952922755.0,
      "step": 29857
    },
    {
      "entropy": 0.3760492210276425,
      "epoch": 2.747091736102725,
      "grad_norm": 0.13970047235488892,
      "learning_rate": 8.433771889471586e-06,
      "loss": 0.3632,
      "mean_token_accuracy": 0.8828036785125732,
      "num_tokens": 952955387.0,
      "step": 29858
    },
    {
      "entropy": 0.5742250513285398,
      "epoch": 2.7471837424741663,
      "grad_norm": 0.16877342760562897,
      "learning_rate": 8.43070506332996e-06,
      "loss": 0.5684,
      "mean_token_accuracy": 0.8160293474793434,
      "num_tokens": 952986382.0,
      "step": 29859
    },
    {
      "entropy": 0.5065213311463594,
      "epoch": 2.7472757488456074,
      "grad_norm": 0.15982168912887573,
      "learning_rate": 8.427638237188334e-06,
      "loss": 0.503,
      "mean_token_accuracy": 0.8394764736294746,
      "num_tokens": 953018348.0,
      "step": 29860
    },
    {
      "entropy": 0.6211793422698975,
      "epoch": 2.7473677552170486,
      "grad_norm": 0.17675945162773132,
      "learning_rate": 8.424571411046708e-06,
      "loss": 0.6189,
      "mean_token_accuracy": 0.8072381094098091,
      "num_tokens": 953049912.0,
      "step": 29861
    },
    {
      "entropy": 0.4625325163360685,
      "epoch": 2.74745976158849,
      "grad_norm": 0.1594417542219162,
      "learning_rate": 8.421504584905083e-06,
      "loss": 0.4454,
      "mean_token_accuracy": 0.85829633846879,
      "num_tokens": 953081806.0,
      "step": 29862
    },
    {
      "entropy": 0.5493579925969243,
      "epoch": 2.7475517679599313,
      "grad_norm": 0.16409724950790405,
      "learning_rate": 8.418437758763456e-06,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8294456079602242,
      "num_tokens": 953113565.0,
      "step": 29863
    },
    {
      "entropy": 0.4911146762315184,
      "epoch": 2.7476437743313724,
      "grad_norm": 0.15636686980724335,
      "learning_rate": 8.41537093262183e-06,
      "loss": 0.4862,
      "mean_token_accuracy": 0.8460001423954964,
      "num_tokens": 953145926.0,
      "step": 29864
    },
    {
      "entropy": 0.501208608970046,
      "epoch": 2.7477357807028135,
      "grad_norm": 0.1624017357826233,
      "learning_rate": 8.412304106480203e-06,
      "loss": 0.493,
      "mean_token_accuracy": 0.8427589312195778,
      "num_tokens": 953177227.0,
      "step": 29865
    },
    {
      "entropy": 0.6122909476980567,
      "epoch": 2.7478277870742547,
      "grad_norm": 0.16891533136367798,
      "learning_rate": 8.409237280338578e-06,
      "loss": 0.6013,
      "mean_token_accuracy": 0.8114463090896606,
      "num_tokens": 953209869.0,
      "step": 29866
    },
    {
      "entropy": 0.485756317502819,
      "epoch": 2.7479197934456963,
      "grad_norm": 0.15583524107933044,
      "learning_rate": 8.406170454196951e-06,
      "loss": 0.472,
      "mean_token_accuracy": 0.8521500825881958,
      "num_tokens": 953241732.0,
      "step": 29867
    },
    {
      "entropy": 0.4426771025173366,
      "epoch": 2.7480117998171374,
      "grad_norm": 0.1436615139245987,
      "learning_rate": 8.403103628055326e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8654484190046787,
      "num_tokens": 953273851.0,
      "step": 29868
    },
    {
      "entropy": 0.5083072218112648,
      "epoch": 2.7481038061885785,
      "grad_norm": 0.15721231698989868,
      "learning_rate": 8.400036801913699e-06,
      "loss": 0.4927,
      "mean_token_accuracy": 0.8430715203285217,
      "num_tokens": 953305801.0,
      "step": 29869
    },
    {
      "entropy": 0.4087734576314688,
      "epoch": 2.7481958125600197,
      "grad_norm": 0.14562328159809113,
      "learning_rate": 8.396969975772073e-06,
      "loss": 0.4029,
      "mean_token_accuracy": 0.8708922117948532,
      "num_tokens": 953337820.0,
      "step": 29870
    },
    {
      "entropy": 0.5296937085222453,
      "epoch": 2.748287818931461,
      "grad_norm": 0.1564520299434662,
      "learning_rate": 8.393903149630448e-06,
      "loss": 0.5096,
      "mean_token_accuracy": 0.8341477885842323,
      "num_tokens": 953370020.0,
      "step": 29871
    },
    {
      "entropy": 0.5268379021435976,
      "epoch": 2.7483798253029024,
      "grad_norm": 0.1563519835472107,
      "learning_rate": 8.390836323488823e-06,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8402279950678349,
      "num_tokens": 953402452.0,
      "step": 29872
    },
    {
      "entropy": 0.4785333704203367,
      "epoch": 2.7484718316743435,
      "grad_norm": 0.15639616549015045,
      "learning_rate": 8.387769497347196e-06,
      "loss": 0.459,
      "mean_token_accuracy": 0.8525321446359158,
      "num_tokens": 953434142.0,
      "step": 29873
    },
    {
      "entropy": 0.5469945883378386,
      "epoch": 2.7485638380457846,
      "grad_norm": 0.16496935486793518,
      "learning_rate": 8.38470267120557e-06,
      "loss": 0.5301,
      "mean_token_accuracy": 0.8307308666408062,
      "num_tokens": 953465036.0,
      "step": 29874
    },
    {
      "entropy": 0.4253496043384075,
      "epoch": 2.7486558444172258,
      "grad_norm": 0.1533622443675995,
      "learning_rate": 8.381635845063943e-06,
      "loss": 0.4281,
      "mean_token_accuracy": 0.8634563870728016,
      "num_tokens": 953496184.0,
      "step": 29875
    },
    {
      "entropy": 0.5241974634118378,
      "epoch": 2.748747850788667,
      "grad_norm": 0.15698570013046265,
      "learning_rate": 8.378569018922318e-06,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8379737548530102,
      "num_tokens": 953528043.0,
      "step": 29876
    },
    {
      "entropy": 0.5330012869089842,
      "epoch": 2.7488398571601085,
      "grad_norm": 0.1609855741262436,
      "learning_rate": 8.375502192780692e-06,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8302181139588356,
      "num_tokens": 953560148.0,
      "step": 29877
    },
    {
      "entropy": 0.4823675339575857,
      "epoch": 2.7489318635315496,
      "grad_norm": 0.15915653109550476,
      "learning_rate": 8.372435366639065e-06,
      "loss": 0.469,
      "mean_token_accuracy": 0.8486844152212143,
      "num_tokens": 953592201.0,
      "step": 29878
    },
    {
      "entropy": 0.5526532083749771,
      "epoch": 2.7490238699029907,
      "grad_norm": 0.16542606055736542,
      "learning_rate": 8.36936854049744e-06,
      "loss": 0.5554,
      "mean_token_accuracy": 0.8275087028741837,
      "num_tokens": 953624258.0,
      "step": 29879
    },
    {
      "entropy": 0.5480890208855271,
      "epoch": 2.749115876274432,
      "grad_norm": 0.160817489027977,
      "learning_rate": 8.366301714355813e-06,
      "loss": 0.5326,
      "mean_token_accuracy": 0.8294398859143257,
      "num_tokens": 953655757.0,
      "step": 29880
    },
    {
      "entropy": 0.4936072123236954,
      "epoch": 2.749207882645873,
      "grad_norm": 0.15660275518894196,
      "learning_rate": 8.363234888214188e-06,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8505503349006176,
      "num_tokens": 953687633.0,
      "step": 29881
    },
    {
      "entropy": 0.5101351551711559,
      "epoch": 2.7492998890173146,
      "grad_norm": 0.15316718816757202,
      "learning_rate": 8.36016806207256e-06,
      "loss": 0.4945,
      "mean_token_accuracy": 0.8451264724135399,
      "num_tokens": 953720008.0,
      "step": 29882
    },
    {
      "entropy": 0.5895947907119989,
      "epoch": 2.7493918953887557,
      "grad_norm": 0.1696222871541977,
      "learning_rate": 8.357101235930935e-06,
      "loss": 0.5773,
      "mean_token_accuracy": 0.81939847022295,
      "num_tokens": 953752305.0,
      "step": 29883
    },
    {
      "entropy": 0.45081869151908904,
      "epoch": 2.749483901760197,
      "grad_norm": 0.15467002987861633,
      "learning_rate": 8.354034409789308e-06,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8575029075145721,
      "num_tokens": 953784378.0,
      "step": 29884
    },
    {
      "entropy": 0.5630247052758932,
      "epoch": 2.749575908131638,
      "grad_norm": 0.1688479781150818,
      "learning_rate": 8.350967583647683e-06,
      "loss": 0.5573,
      "mean_token_accuracy": 0.824380949139595,
      "num_tokens": 953816549.0,
      "step": 29885
    },
    {
      "entropy": 0.487217896617949,
      "epoch": 2.749667914503079,
      "grad_norm": 0.15885990858078003,
      "learning_rate": 8.347900757506058e-06,
      "loss": 0.4773,
      "mean_token_accuracy": 0.8504114300012589,
      "num_tokens": 953848560.0,
      "step": 29886
    },
    {
      "entropy": 0.5216638566926122,
      "epoch": 2.7497599208745207,
      "grad_norm": 0.1565675437450409,
      "learning_rate": 8.344833931364432e-06,
      "loss": 0.5046,
      "mean_token_accuracy": 0.841695137321949,
      "num_tokens": 953880602.0,
      "step": 29887
    },
    {
      "entropy": 0.6122578866779804,
      "epoch": 2.749851927245962,
      "grad_norm": 0.1693347990512848,
      "learning_rate": 8.341767105222805e-06,
      "loss": 0.6013,
      "mean_token_accuracy": 0.8111914321780205,
      "num_tokens": 953912862.0,
      "step": 29888
    },
    {
      "entropy": 0.4168141661211848,
      "epoch": 2.749943933617403,
      "grad_norm": 0.14338746666908264,
      "learning_rate": 8.33870027908118e-06,
      "loss": 0.4081,
      "mean_token_accuracy": 0.8715683035552502,
      "num_tokens": 953944867.0,
      "step": 29889
    },
    {
      "entropy": 0.5375016210600734,
      "epoch": 2.750035939988844,
      "grad_norm": 0.1630726158618927,
      "learning_rate": 8.335633452939553e-06,
      "loss": 0.5274,
      "mean_token_accuracy": 0.831487849354744,
      "num_tokens": 953976969.0,
      "step": 29890
    },
    {
      "entropy": 0.5745323738083243,
      "epoch": 2.7501279463602852,
      "grad_norm": 0.16619180142879486,
      "learning_rate": 8.332566626797927e-06,
      "loss": 0.5572,
      "mean_token_accuracy": 0.8231817446649075,
      "num_tokens": 954007983.0,
      "step": 29891
    },
    {
      "entropy": 0.453215173445642,
      "epoch": 2.750219952731727,
      "grad_norm": 0.155622199177742,
      "learning_rate": 8.329499800656302e-06,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8607611805200577,
      "num_tokens": 954039949.0,
      "step": 29892
    },
    {
      "entropy": 0.48515474889427423,
      "epoch": 2.750311959103168,
      "grad_norm": 0.15466073155403137,
      "learning_rate": 8.326432974514675e-06,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8466136865317822,
      "num_tokens": 954072013.0,
      "step": 29893
    },
    {
      "entropy": 0.624629270285368,
      "epoch": 2.750403965474609,
      "grad_norm": 0.17273223400115967,
      "learning_rate": 8.32336614837305e-06,
      "loss": 0.607,
      "mean_token_accuracy": 0.811398945748806,
      "num_tokens": 954104212.0,
      "step": 29894
    },
    {
      "entropy": 0.5277472399175167,
      "epoch": 2.75049597184605,
      "grad_norm": 0.16570086777210236,
      "learning_rate": 8.320299322231423e-06,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8362935818731785,
      "num_tokens": 954135791.0,
      "step": 29895
    },
    {
      "entropy": 0.5253546088933945,
      "epoch": 2.7505879782174913,
      "grad_norm": 0.1592235267162323,
      "learning_rate": 8.317232496089797e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.8366604968905449,
      "num_tokens": 954168138.0,
      "step": 29896
    },
    {
      "entropy": 0.5063671027310193,
      "epoch": 2.750679984588933,
      "grad_norm": 0.15741270780563354,
      "learning_rate": 8.31416566994817e-06,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8458127677440643,
      "num_tokens": 954200158.0,
      "step": 29897
    },
    {
      "entropy": 0.5436564590781927,
      "epoch": 2.750771990960374,
      "grad_norm": 0.15634529292583466,
      "learning_rate": 8.311098843806545e-06,
      "loss": 0.532,
      "mean_token_accuracy": 0.8311962932348251,
      "num_tokens": 954232415.0,
      "step": 29898
    },
    {
      "entropy": 0.42985750269144773,
      "epoch": 2.750863997331815,
      "grad_norm": 0.14584971964359283,
      "learning_rate": 8.308032017664918e-06,
      "loss": 0.42,
      "mean_token_accuracy": 0.8656830005347729,
      "num_tokens": 954264746.0,
      "step": 29899
    },
    {
      "entropy": 0.5699399495497346,
      "epoch": 2.7509560037032563,
      "grad_norm": 0.16208450496196747,
      "learning_rate": 8.304965191523292e-06,
      "loss": 0.5522,
      "mean_token_accuracy": 0.826047308743,
      "num_tokens": 954296598.0,
      "step": 29900
    },
    {
      "entropy": 0.586231246124953,
      "epoch": 2.7510480100746975,
      "grad_norm": 0.16333167254924774,
      "learning_rate": 8.301898365381667e-06,
      "loss": 0.5586,
      "mean_token_accuracy": 0.8217355497181416,
      "num_tokens": 954328513.0,
      "step": 29901
    },
    {
      "entropy": 0.5490612182766199,
      "epoch": 2.751140016446139,
      "grad_norm": 0.1583280861377716,
      "learning_rate": 8.298831539240042e-06,
      "loss": 0.5443,
      "mean_token_accuracy": 0.8297588638961315,
      "num_tokens": 954360917.0,
      "step": 29902
    },
    {
      "entropy": 0.46810030937194824,
      "epoch": 2.75123202281758,
      "grad_norm": 0.15111318230628967,
      "learning_rate": 8.295764713098415e-06,
      "loss": 0.4561,
      "mean_token_accuracy": 0.8541420176625252,
      "num_tokens": 954392618.0,
      "step": 29903
    },
    {
      "entropy": 0.46090863179415464,
      "epoch": 2.7513240291890213,
      "grad_norm": 0.14962004125118256,
      "learning_rate": 8.29269788695679e-06,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8582993261516094,
      "num_tokens": 954424116.0,
      "step": 29904
    },
    {
      "entropy": 0.4546239646151662,
      "epoch": 2.7514160355604624,
      "grad_norm": 0.15293273329734802,
      "learning_rate": 8.289631060815162e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.858315221965313,
      "num_tokens": 954456208.0,
      "step": 29905
    },
    {
      "entropy": 0.49875031039118767,
      "epoch": 2.7515080419319036,
      "grad_norm": 0.15945155918598175,
      "learning_rate": 8.286564234673537e-06,
      "loss": 0.5006,
      "mean_token_accuracy": 0.8407943062484264,
      "num_tokens": 954488724.0,
      "step": 29906
    },
    {
      "entropy": 0.5467359987087548,
      "epoch": 2.751600048303345,
      "grad_norm": 0.1633516252040863,
      "learning_rate": 8.283497408531912e-06,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8338988237082958,
      "num_tokens": 954520490.0,
      "step": 29907
    },
    {
      "entropy": 0.5184823302552104,
      "epoch": 2.7516920546747863,
      "grad_norm": 0.16022615134716034,
      "learning_rate": 8.280430582390285e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.8359988406300545,
      "num_tokens": 954552646.0,
      "step": 29908
    },
    {
      "entropy": 0.39673815108835697,
      "epoch": 2.7517840610462274,
      "grad_norm": 0.15093879401683807,
      "learning_rate": 8.27736375624866e-06,
      "loss": 0.3933,
      "mean_token_accuracy": 0.8704471290111542,
      "num_tokens": 954583884.0,
      "step": 29909
    },
    {
      "entropy": 0.5663810428231955,
      "epoch": 2.7518760674176685,
      "grad_norm": 0.16541293263435364,
      "learning_rate": 8.274296930107032e-06,
      "loss": 0.5602,
      "mean_token_accuracy": 0.8197520300745964,
      "num_tokens": 954615693.0,
      "step": 29910
    },
    {
      "entropy": 0.45247078931424767,
      "epoch": 2.7519680737891097,
      "grad_norm": 0.14884358644485474,
      "learning_rate": 8.271230103965407e-06,
      "loss": 0.4356,
      "mean_token_accuracy": 0.8567561060190201,
      "num_tokens": 954647453.0,
      "step": 29911
    },
    {
      "entropy": 0.5781666450202465,
      "epoch": 2.7520600801605513,
      "grad_norm": 0.16765184700489044,
      "learning_rate": 8.26816327782378e-06,
      "loss": 0.56,
      "mean_token_accuracy": 0.8216214999556541,
      "num_tokens": 954678642.0,
      "step": 29912
    },
    {
      "entropy": 0.4630758622661233,
      "epoch": 2.7521520865319924,
      "grad_norm": 0.15344280004501343,
      "learning_rate": 8.265096451682154e-06,
      "loss": 0.4475,
      "mean_token_accuracy": 0.8528466038405895,
      "num_tokens": 954709720.0,
      "step": 29913
    },
    {
      "entropy": 0.49755996838212013,
      "epoch": 2.7522440929034335,
      "grad_norm": 0.15638577938079834,
      "learning_rate": 8.262029625540527e-06,
      "loss": 0.4801,
      "mean_token_accuracy": 0.845756184309721,
      "num_tokens": 954741938.0,
      "step": 29914
    },
    {
      "entropy": 0.4520229157060385,
      "epoch": 2.7523360992748747,
      "grad_norm": 0.15171663463115692,
      "learning_rate": 8.258962799398902e-06,
      "loss": 0.4355,
      "mean_token_accuracy": 0.8616002202033997,
      "num_tokens": 954773729.0,
      "step": 29915
    },
    {
      "entropy": 0.4313311045989394,
      "epoch": 2.752428105646316,
      "grad_norm": 0.14570854604244232,
      "learning_rate": 8.255895973257275e-06,
      "loss": 0.4119,
      "mean_token_accuracy": 0.8664082735776901,
      "num_tokens": 954805870.0,
      "step": 29916
    },
    {
      "entropy": 0.5683401841670275,
      "epoch": 2.7525201120177574,
      "grad_norm": 0.1683129072189331,
      "learning_rate": 8.252829147115651e-06,
      "loss": 0.5619,
      "mean_token_accuracy": 0.8238731324672699,
      "num_tokens": 954837396.0,
      "step": 29917
    },
    {
      "entropy": 0.5043704101117328,
      "epoch": 2.7526121183891985,
      "grad_norm": 0.15496030449867249,
      "learning_rate": 8.249762320974024e-06,
      "loss": 0.4881,
      "mean_token_accuracy": 0.8442598134279251,
      "num_tokens": 954869764.0,
      "step": 29918
    },
    {
      "entropy": 0.4758687005378306,
      "epoch": 2.7527041247606396,
      "grad_norm": 0.1577085554599762,
      "learning_rate": 8.246695494832399e-06,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8557295240461826,
      "num_tokens": 954901269.0,
      "step": 29919
    },
    {
      "entropy": 0.42573520296718925,
      "epoch": 2.7527961311320808,
      "grad_norm": 0.146132230758667,
      "learning_rate": 8.243628668690772e-06,
      "loss": 0.4094,
      "mean_token_accuracy": 0.8684781827032566,
      "num_tokens": 954933805.0,
      "step": 29920
    },
    {
      "entropy": 0.4694071786943823,
      "epoch": 2.752888137503522,
      "grad_norm": 0.14953818917274475,
      "learning_rate": 8.240561842549146e-06,
      "loss": 0.4521,
      "mean_token_accuracy": 0.8548642657697201,
      "num_tokens": 954966222.0,
      "step": 29921
    },
    {
      "entropy": 0.5065359212458134,
      "epoch": 2.7529801438749635,
      "grad_norm": 0.15289221704006195,
      "learning_rate": 8.237495016407521e-06,
      "loss": 0.4967,
      "mean_token_accuracy": 0.8442559950053692,
      "num_tokens": 954997969.0,
      "step": 29922
    },
    {
      "entropy": 0.5780081851407886,
      "epoch": 2.7530721502464046,
      "grad_norm": 0.16768166422843933,
      "learning_rate": 8.234428190265894e-06,
      "loss": 0.5636,
      "mean_token_accuracy": 0.8250781856477261,
      "num_tokens": 955030578.0,
      "step": 29923
    },
    {
      "entropy": 0.5231542307883501,
      "epoch": 2.7531641566178457,
      "grad_norm": 0.15522919595241547,
      "learning_rate": 8.231361364124269e-06,
      "loss": 0.5097,
      "mean_token_accuracy": 0.840232253074646,
      "num_tokens": 955062229.0,
      "step": 29924
    },
    {
      "entropy": 0.6343635059893131,
      "epoch": 2.753256162989287,
      "grad_norm": 0.17550310492515564,
      "learning_rate": 8.228294537982642e-06,
      "loss": 0.626,
      "mean_token_accuracy": 0.805464930832386,
      "num_tokens": 955093286.0,
      "step": 29925
    },
    {
      "entropy": 0.48859353736042976,
      "epoch": 2.753348169360728,
      "grad_norm": 0.15812505781650543,
      "learning_rate": 8.225227711841016e-06,
      "loss": 0.4683,
      "mean_token_accuracy": 0.8510548770427704,
      "num_tokens": 955124313.0,
      "step": 29926
    },
    {
      "entropy": 0.5145634142681956,
      "epoch": 2.7534401757321696,
      "grad_norm": 0.1549079716205597,
      "learning_rate": 8.22216088569939e-06,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8450781553983688,
      "num_tokens": 955157081.0,
      "step": 29927
    },
    {
      "entropy": 0.4860286461189389,
      "epoch": 2.7535321821036107,
      "grad_norm": 0.15422485768795013,
      "learning_rate": 8.219094059557764e-06,
      "loss": 0.4674,
      "mean_token_accuracy": 0.8548025563359261,
      "num_tokens": 955188459.0,
      "step": 29928
    },
    {
      "entropy": 0.5004647611640394,
      "epoch": 2.753624188475052,
      "grad_norm": 0.1539938896894455,
      "learning_rate": 8.216027233416137e-06,
      "loss": 0.4758,
      "mean_token_accuracy": 0.8469872921705246,
      "num_tokens": 955220733.0,
      "step": 29929
    },
    {
      "entropy": 0.4815553901717067,
      "epoch": 2.753716194846493,
      "grad_norm": 0.15450213849544525,
      "learning_rate": 8.212960407274512e-06,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8499859347939491,
      "num_tokens": 955253052.0,
      "step": 29930
    },
    {
      "entropy": 0.46525106206536293,
      "epoch": 2.753808201217934,
      "grad_norm": 0.1596013903617859,
      "learning_rate": 8.209893581132884e-06,
      "loss": 0.4604,
      "mean_token_accuracy": 0.8521596826612949,
      "num_tokens": 955284646.0,
      "step": 29931
    },
    {
      "entropy": 0.5033528245985508,
      "epoch": 2.7539002075893757,
      "grad_norm": 0.1633491963148117,
      "learning_rate": 8.20682675499126e-06,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8394935056567192,
      "num_tokens": 955316715.0,
      "step": 29932
    },
    {
      "entropy": 0.4193509453907609,
      "epoch": 2.753992213960817,
      "grad_norm": 0.14777927100658417,
      "learning_rate": 8.203759928849634e-06,
      "loss": 0.4121,
      "mean_token_accuracy": 0.8655530512332916,
      "num_tokens": 955349158.0,
      "step": 29933
    },
    {
      "entropy": 0.5262940805405378,
      "epoch": 2.754084220332258,
      "grad_norm": 0.16387522220611572,
      "learning_rate": 8.200693102708008e-06,
      "loss": 0.5186,
      "mean_token_accuracy": 0.8376646153628826,
      "num_tokens": 955380639.0,
      "step": 29934
    },
    {
      "entropy": 0.48397127352654934,
      "epoch": 2.754176226703699,
      "grad_norm": 0.1569935828447342,
      "learning_rate": 8.197626276566381e-06,
      "loss": 0.4836,
      "mean_token_accuracy": 0.8471134528517723,
      "num_tokens": 955412737.0,
      "step": 29935
    },
    {
      "entropy": 0.5262636094121262,
      "epoch": 2.7542682330751402,
      "grad_norm": 0.1579839587211609,
      "learning_rate": 8.194559450424756e-06,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8368801176548004,
      "num_tokens": 955444588.0,
      "step": 29936
    },
    {
      "entropy": 0.47667461074888706,
      "epoch": 2.754360239446582,
      "grad_norm": 0.1525801718235016,
      "learning_rate": 8.19149262428313e-06,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8544154204428196,
      "num_tokens": 955476305.0,
      "step": 29937
    },
    {
      "entropy": 0.4622123222798109,
      "epoch": 2.754452245818023,
      "grad_norm": 0.15595756471157074,
      "learning_rate": 8.188425798141504e-06,
      "loss": 0.4536,
      "mean_token_accuracy": 0.8527952246367931,
      "num_tokens": 955508571.0,
      "step": 29938
    },
    {
      "entropy": 0.4665300268679857,
      "epoch": 2.754544252189464,
      "grad_norm": 0.15452489256858826,
      "learning_rate": 8.185358971999878e-06,
      "loss": 0.4607,
      "mean_token_accuracy": 0.8537804484367371,
      "num_tokens": 955540465.0,
      "step": 29939
    },
    {
      "entropy": 0.457061312161386,
      "epoch": 2.754636258560905,
      "grad_norm": 0.15511681139469147,
      "learning_rate": 8.182292145858251e-06,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8551973886787891,
      "num_tokens": 955571694.0,
      "step": 29940
    },
    {
      "entropy": 0.6086422819644213,
      "epoch": 2.7547282649323463,
      "grad_norm": 0.16367682814598083,
      "learning_rate": 8.179225319716626e-06,
      "loss": 0.5848,
      "mean_token_accuracy": 0.8137224093079567,
      "num_tokens": 955603233.0,
      "step": 29941
    },
    {
      "entropy": 0.4560956433415413,
      "epoch": 2.754820271303788,
      "grad_norm": 0.1490427553653717,
      "learning_rate": 8.176158493574999e-06,
      "loss": 0.4448,
      "mean_token_accuracy": 0.8577105440199375,
      "num_tokens": 955634413.0,
      "step": 29942
    },
    {
      "entropy": 0.5703168753534555,
      "epoch": 2.754912277675229,
      "grad_norm": 0.16225220263004303,
      "learning_rate": 8.173091667433373e-06,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8215284049510956,
      "num_tokens": 955666713.0,
      "step": 29943
    },
    {
      "entropy": 0.4713336252607405,
      "epoch": 2.75500428404667,
      "grad_norm": 0.1546478122472763,
      "learning_rate": 8.170024841291746e-06,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8510037250816822,
      "num_tokens": 955698492.0,
      "step": 29944
    },
    {
      "entropy": 0.5418167682364583,
      "epoch": 2.7550962904181113,
      "grad_norm": 0.16377010941505432,
      "learning_rate": 8.166958015150121e-06,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8324913755059242,
      "num_tokens": 955730435.0,
      "step": 29945
    },
    {
      "entropy": 0.4862560583278537,
      "epoch": 2.7551882967895525,
      "grad_norm": 0.15338708460330963,
      "learning_rate": 8.163891189008494e-06,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8488144613802433,
      "num_tokens": 955762405.0,
      "step": 29946
    },
    {
      "entropy": 0.4867588556371629,
      "epoch": 2.755280303160994,
      "grad_norm": 0.1575908362865448,
      "learning_rate": 8.16082436286687e-06,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8466619849205017,
      "num_tokens": 955794600.0,
      "step": 29947
    },
    {
      "entropy": 0.5512390192598104,
      "epoch": 2.755372309532435,
      "grad_norm": 0.16230526566505432,
      "learning_rate": 8.157757536725243e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8308676034212112,
      "num_tokens": 955826402.0,
      "step": 29948
    },
    {
      "entropy": 0.584863927680999,
      "epoch": 2.7554643159038763,
      "grad_norm": 0.17120496928691864,
      "learning_rate": 8.154690710583618e-06,
      "loss": 0.581,
      "mean_token_accuracy": 0.818032406270504,
      "num_tokens": 955858093.0,
      "step": 29949
    },
    {
      "entropy": 0.5689740860834718,
      "epoch": 2.7555563222753174,
      "grad_norm": 0.16406755149364471,
      "learning_rate": 8.151623884441991e-06,
      "loss": 0.5582,
      "mean_token_accuracy": 0.8273940421640873,
      "num_tokens": 955889896.0,
      "step": 29950
    },
    {
      "entropy": 0.4581419052556157,
      "epoch": 2.7556483286467586,
      "grad_norm": 0.15169581770896912,
      "learning_rate": 8.148557058300366e-06,
      "loss": 0.455,
      "mean_token_accuracy": 0.8549972251057625,
      "num_tokens": 955922234.0,
      "step": 29951
    },
    {
      "entropy": 0.5677355462685227,
      "epoch": 2.7557403350182,
      "grad_norm": 0.1667482554912567,
      "learning_rate": 8.14549023215874e-06,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8231690190732479,
      "num_tokens": 955953750.0,
      "step": 29952
    },
    {
      "entropy": 0.5532951643690467,
      "epoch": 2.7558323413896413,
      "grad_norm": 0.17316986620426178,
      "learning_rate": 8.142423406017113e-06,
      "loss": 0.5486,
      "mean_token_accuracy": 0.8302580304443836,
      "num_tokens": 955986116.0,
      "step": 29953
    },
    {
      "entropy": 0.5199444768950343,
      "epoch": 2.7559243477610824,
      "grad_norm": 0.16208039224147797,
      "learning_rate": 8.139356579875488e-06,
      "loss": 0.5101,
      "mean_token_accuracy": 0.8348654210567474,
      "num_tokens": 956017370.0,
      "step": 29954
    },
    {
      "entropy": 0.5232006749138236,
      "epoch": 2.7560163541325235,
      "grad_norm": 0.1616925597190857,
      "learning_rate": 8.13628975373386e-06,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8370995037257671,
      "num_tokens": 956049532.0,
      "step": 29955
    },
    {
      "entropy": 0.4926131833344698,
      "epoch": 2.7561083605039647,
      "grad_norm": 0.15945963561534882,
      "learning_rate": 8.133222927592235e-06,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8437083847820759,
      "num_tokens": 956081687.0,
      "step": 29956
    },
    {
      "entropy": 0.4818327100947499,
      "epoch": 2.7562003668754063,
      "grad_norm": 0.15753427147865295,
      "learning_rate": 8.130156101450608e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8534070178866386,
      "num_tokens": 956113607.0,
      "step": 29957
    },
    {
      "entropy": 0.538609629496932,
      "epoch": 2.7562923732468474,
      "grad_norm": 0.17000383138656616,
      "learning_rate": 8.127089275308983e-06,
      "loss": 0.5265,
      "mean_token_accuracy": 0.8344250433146954,
      "num_tokens": 956144852.0,
      "step": 29958
    },
    {
      "entropy": 0.580878890119493,
      "epoch": 2.7563843796182885,
      "grad_norm": 0.16475720703601837,
      "learning_rate": 8.124022449167356e-06,
      "loss": 0.5768,
      "mean_token_accuracy": 0.8210177831351757,
      "num_tokens": 956176218.0,
      "step": 29959
    },
    {
      "entropy": 0.557244660332799,
      "epoch": 2.7564763859897297,
      "grad_norm": 0.1659681349992752,
      "learning_rate": 8.12095562302573e-06,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8234551921486855,
      "num_tokens": 956207447.0,
      "step": 29960
    },
    {
      "entropy": 0.5054826210252941,
      "epoch": 2.756568392361171,
      "grad_norm": 0.15759997069835663,
      "learning_rate": 8.117888796884104e-06,
      "loss": 0.5047,
      "mean_token_accuracy": 0.8424173258244991,
      "num_tokens": 956239379.0,
      "step": 29961
    },
    {
      "entropy": 0.5825619995594025,
      "epoch": 2.7566603987326124,
      "grad_norm": 0.16596652567386627,
      "learning_rate": 8.11482197074248e-06,
      "loss": 0.5809,
      "mean_token_accuracy": 0.8195425197482109,
      "num_tokens": 956271602.0,
      "step": 29962
    },
    {
      "entropy": 0.43189507722854614,
      "epoch": 2.7567524051040535,
      "grad_norm": 0.14668886363506317,
      "learning_rate": 8.111755144600853e-06,
      "loss": 0.4248,
      "mean_token_accuracy": 0.8634703606367111,
      "num_tokens": 956302906.0,
      "step": 29963
    },
    {
      "entropy": 0.5032129995524883,
      "epoch": 2.7568444114754946,
      "grad_norm": 0.1605684906244278,
      "learning_rate": 8.108688318459228e-06,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8441488668322563,
      "num_tokens": 956335006.0,
      "step": 29964
    },
    {
      "entropy": 0.5360419261269271,
      "epoch": 2.7569364178469358,
      "grad_norm": 0.15390412509441376,
      "learning_rate": 8.1056214923176e-06,
      "loss": 0.535,
      "mean_token_accuracy": 0.8333363197743893,
      "num_tokens": 956367613.0,
      "step": 29965
    },
    {
      "entropy": 0.49133007042109966,
      "epoch": 2.757028424218377,
      "grad_norm": 0.15632499754428864,
      "learning_rate": 8.102554666175975e-06,
      "loss": 0.4833,
      "mean_token_accuracy": 0.8466712310910225,
      "num_tokens": 956399774.0,
      "step": 29966
    },
    {
      "entropy": 0.48156578885391355,
      "epoch": 2.7571204305898185,
      "grad_norm": 0.1590673178434372,
      "learning_rate": 8.09948784003435e-06,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8454879075288773,
      "num_tokens": 956430672.0,
      "step": 29967
    },
    {
      "entropy": 0.5677047939971089,
      "epoch": 2.7572124369612596,
      "grad_norm": 0.16338308155536652,
      "learning_rate": 8.096421013892723e-06,
      "loss": 0.5468,
      "mean_token_accuracy": 0.8282833956182003,
      "num_tokens": 956463076.0,
      "step": 29968
    },
    {
      "entropy": 0.4989089658483863,
      "epoch": 2.7573044433327007,
      "grad_norm": 0.1558338850736618,
      "learning_rate": 8.093354187751097e-06,
      "loss": 0.4875,
      "mean_token_accuracy": 0.842915765941143,
      "num_tokens": 956495316.0,
      "step": 29969
    },
    {
      "entropy": 0.5394431825261563,
      "epoch": 2.757396449704142,
      "grad_norm": 0.1568731814622879,
      "learning_rate": 8.09028736160947e-06,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8359987437725067,
      "num_tokens": 956527745.0,
      "step": 29970
    },
    {
      "entropy": 0.5145496279001236,
      "epoch": 2.757488456075583,
      "grad_norm": 0.15942978858947754,
      "learning_rate": 8.087220535467845e-06,
      "loss": 0.5116,
      "mean_token_accuracy": 0.8397008292376995,
      "num_tokens": 956559831.0,
      "step": 29971
    },
    {
      "entropy": 0.4574698054930195,
      "epoch": 2.7575804624470246,
      "grad_norm": 0.15112997591495514,
      "learning_rate": 8.084153709326218e-06,
      "loss": 0.4497,
      "mean_token_accuracy": 0.8577198684215546,
      "num_tokens": 956592190.0,
      "step": 29972
    },
    {
      "entropy": 0.6055217105895281,
      "epoch": 2.7576724688184657,
      "grad_norm": 0.1697261780500412,
      "learning_rate": 8.081086883184593e-06,
      "loss": 0.6058,
      "mean_token_accuracy": 0.8109300397336483,
      "num_tokens": 956623711.0,
      "step": 29973
    },
    {
      "entropy": 0.3608243577182293,
      "epoch": 2.757764475189907,
      "grad_norm": 0.14309993386268616,
      "learning_rate": 8.078020057042966e-06,
      "loss": 0.3457,
      "mean_token_accuracy": 0.8879228234291077,
      "num_tokens": 956655428.0,
      "step": 29974
    },
    {
      "entropy": 0.546369576593861,
      "epoch": 2.757856481561348,
      "grad_norm": 0.16606903076171875,
      "learning_rate": 8.07495323090134e-06,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8281718343496323,
      "num_tokens": 956686439.0,
      "step": 29975
    },
    {
      "entropy": 0.5558987748809159,
      "epoch": 2.757948487932789,
      "grad_norm": 0.16123764216899872,
      "learning_rate": 8.071886404759713e-06,
      "loss": 0.5346,
      "mean_token_accuracy": 0.8341857232153416,
      "num_tokens": 956718435.0,
      "step": 29976
    },
    {
      "entropy": 0.5043607288971543,
      "epoch": 2.7580404943042307,
      "grad_norm": 0.15260519087314606,
      "learning_rate": 8.068819578618088e-06,
      "loss": 0.4846,
      "mean_token_accuracy": 0.847457330673933,
      "num_tokens": 956750627.0,
      "step": 29977
    },
    {
      "entropy": 0.5202036374248564,
      "epoch": 2.758132500675672,
      "grad_norm": 0.15871688723564148,
      "learning_rate": 8.065752752476462e-06,
      "loss": 0.5005,
      "mean_token_accuracy": 0.8431636765599251,
      "num_tokens": 956782701.0,
      "step": 29978
    },
    {
      "entropy": 0.4805075703188777,
      "epoch": 2.758224507047113,
      "grad_norm": 0.1570717990398407,
      "learning_rate": 8.062685926334837e-06,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8510277643799782,
      "num_tokens": 956813989.0,
      "step": 29979
    },
    {
      "entropy": 0.6427954100072384,
      "epoch": 2.758316513418554,
      "grad_norm": 0.1758672446012497,
      "learning_rate": 8.05961910019321e-06,
      "loss": 0.6223,
      "mean_token_accuracy": 0.8025556281208992,
      "num_tokens": 956846742.0,
      "step": 29980
    },
    {
      "entropy": 0.5265904229599983,
      "epoch": 2.7584085197899952,
      "grad_norm": 0.15786905586719513,
      "learning_rate": 8.056552274051585e-06,
      "loss": 0.5186,
      "mean_token_accuracy": 0.8361459635198116,
      "num_tokens": 956879014.0,
      "step": 29981
    },
    {
      "entropy": 0.5224381545558572,
      "epoch": 2.758500526161437,
      "grad_norm": 0.15745466947555542,
      "learning_rate": 8.05348544790996e-06,
      "loss": 0.5082,
      "mean_token_accuracy": 0.8368084952235222,
      "num_tokens": 956911472.0,
      "step": 29982
    },
    {
      "entropy": 0.49257446359843016,
      "epoch": 2.758592532532878,
      "grad_norm": 0.15990914404392242,
      "learning_rate": 8.050418621768332e-06,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8467266298830509,
      "num_tokens": 956944227.0,
      "step": 29983
    },
    {
      "entropy": 0.5316358301788568,
      "epoch": 2.758684538904319,
      "grad_norm": 0.1654881089925766,
      "learning_rate": 8.047351795626707e-06,
      "loss": 0.5171,
      "mean_token_accuracy": 0.8355995863676071,
      "num_tokens": 956975295.0,
      "step": 29984
    },
    {
      "entropy": 0.5200402631890029,
      "epoch": 2.75877654527576,
      "grad_norm": 0.15804210305213928,
      "learning_rate": 8.04428496948508e-06,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8363706469535828,
      "num_tokens": 957007116.0,
      "step": 29985
    },
    {
      "entropy": 0.5774671975523233,
      "epoch": 2.7588685516472014,
      "grad_norm": 0.16376914083957672,
      "learning_rate": 8.041218143343455e-06,
      "loss": 0.5699,
      "mean_token_accuracy": 0.8205686956644058,
      "num_tokens": 957039360.0,
      "step": 29986
    },
    {
      "entropy": 0.5137473111972213,
      "epoch": 2.758960558018643,
      "grad_norm": 0.1592244803905487,
      "learning_rate": 8.038151317201827e-06,
      "loss": 0.4949,
      "mean_token_accuracy": 0.8405691012740135,
      "num_tokens": 957070931.0,
      "step": 29987
    },
    {
      "entropy": 0.5160310240462422,
      "epoch": 2.759052564390084,
      "grad_norm": 0.16481052339076996,
      "learning_rate": 8.035084491060202e-06,
      "loss": 0.51,
      "mean_token_accuracy": 0.8380220234394073,
      "num_tokens": 957102966.0,
      "step": 29988
    },
    {
      "entropy": 0.4787656953558326,
      "epoch": 2.759144570761525,
      "grad_norm": 0.15379749238491058,
      "learning_rate": 8.032017664918575e-06,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8497967571020126,
      "num_tokens": 957134852.0,
      "step": 29989
    },
    {
      "entropy": 0.4987211977131665,
      "epoch": 2.7592365771329663,
      "grad_norm": 0.15036554634571075,
      "learning_rate": 8.02895083877695e-06,
      "loss": 0.4844,
      "mean_token_accuracy": 0.8475356474518776,
      "num_tokens": 957166638.0,
      "step": 29990
    },
    {
      "entropy": 0.5079495012760162,
      "epoch": 2.7593285835044075,
      "grad_norm": 0.15850390493869781,
      "learning_rate": 8.025884012635323e-06,
      "loss": 0.4954,
      "mean_token_accuracy": 0.8398894518613815,
      "num_tokens": 957198439.0,
      "step": 29991
    },
    {
      "entropy": 0.554020157083869,
      "epoch": 2.759420589875849,
      "grad_norm": 0.16787739098072052,
      "learning_rate": 8.022817186493697e-06,
      "loss": 0.5242,
      "mean_token_accuracy": 0.8330074846744537,
      "num_tokens": 957229768.0,
      "step": 29992
    },
    {
      "entropy": 0.46518607810139656,
      "epoch": 2.75951259624729,
      "grad_norm": 0.16126543283462524,
      "learning_rate": 8.019750360352072e-06,
      "loss": 0.4373,
      "mean_token_accuracy": 0.8582842350006104,
      "num_tokens": 957258903.0,
      "step": 29993
    },
    {
      "entropy": 0.5858683865517378,
      "epoch": 2.7596046026187313,
      "grad_norm": 0.1705801784992218,
      "learning_rate": 8.016683534210447e-06,
      "loss": 0.5875,
      "mean_token_accuracy": 0.8169808499515057,
      "num_tokens": 957290817.0,
      "step": 29994
    },
    {
      "entropy": 0.5801620036363602,
      "epoch": 2.7596966089901724,
      "grad_norm": 0.17069722712039948,
      "learning_rate": 8.01361670806882e-06,
      "loss": 0.5747,
      "mean_token_accuracy": 0.8167169578373432,
      "num_tokens": 957323043.0,
      "step": 29995
    },
    {
      "entropy": 0.5760022159665823,
      "epoch": 2.7597886153616136,
      "grad_norm": 0.17278152704238892,
      "learning_rate": 8.010549881927194e-06,
      "loss": 0.5745,
      "mean_token_accuracy": 0.8192017041146755,
      "num_tokens": 957354048.0,
      "step": 29996
    },
    {
      "entropy": 0.4517498202621937,
      "epoch": 2.759880621733055,
      "grad_norm": 0.15319111943244934,
      "learning_rate": 8.007483055785569e-06,
      "loss": 0.4425,
      "mean_token_accuracy": 0.8572277426719666,
      "num_tokens": 957385724.0,
      "step": 29997
    },
    {
      "entropy": 0.49849338736385107,
      "epoch": 2.7599726281044963,
      "grad_norm": 0.15751083195209503,
      "learning_rate": 8.004416229643942e-06,
      "loss": 0.4874,
      "mean_token_accuracy": 0.8450451605021954,
      "num_tokens": 957417911.0,
      "step": 29998
    },
    {
      "entropy": 0.38173471996560693,
      "epoch": 2.7600646344759374,
      "grad_norm": 0.14470283687114716,
      "learning_rate": 8.001349403502316e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8804504200816154,
      "num_tokens": 957449546.0,
      "step": 29999
    },
    {
      "entropy": 0.6176809147000313,
      "epoch": 2.7601566408473786,
      "grad_norm": 0.1742590367794037,
      "learning_rate": 7.99828257736069e-06,
      "loss": 0.6143,
      "mean_token_accuracy": 0.8057381026446819,
      "num_tokens": 957481244.0,
      "step": 30000
    },
    {
      "entropy": 0.48474519653245807,
      "epoch": 2.7602486472188197,
      "grad_norm": 0.15339761972427368,
      "learning_rate": 7.995215751219064e-06,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8524862229824066,
      "num_tokens": 957512942.0,
      "step": 30001
    },
    {
      "entropy": 0.4393643712392077,
      "epoch": 2.7603406535902613,
      "grad_norm": 0.14836761355400085,
      "learning_rate": 7.992148925077437e-06,
      "loss": 0.4297,
      "mean_token_accuracy": 0.8633408024907112,
      "num_tokens": 957545249.0,
      "step": 30002
    },
    {
      "entropy": 0.5136038148775697,
      "epoch": 2.7604326599617024,
      "grad_norm": 0.16176770627498627,
      "learning_rate": 7.989082098935812e-06,
      "loss": 0.5073,
      "mean_token_accuracy": 0.8431277945637703,
      "num_tokens": 957576916.0,
      "step": 30003
    },
    {
      "entropy": 0.5843825936317444,
      "epoch": 2.7605246663331435,
      "grad_norm": 0.16813863813877106,
      "learning_rate": 7.986015272794185e-06,
      "loss": 0.5732,
      "mean_token_accuracy": 0.819862850010395,
      "num_tokens": 957609178.0,
      "step": 30004
    },
    {
      "entropy": 0.570758868008852,
      "epoch": 2.7606166727045847,
      "grad_norm": 0.1675911396741867,
      "learning_rate": 7.98294844665256e-06,
      "loss": 0.55,
      "mean_token_accuracy": 0.8253574557602406,
      "num_tokens": 957640941.0,
      "step": 30005
    },
    {
      "entropy": 0.5549767389893532,
      "epoch": 2.760708679076026,
      "grad_norm": 0.16776308417320251,
      "learning_rate": 7.979881620510932e-06,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8290496505796909,
      "num_tokens": 957671887.0,
      "step": 30006
    },
    {
      "entropy": 0.4716509454883635,
      "epoch": 2.7608006854474674,
      "grad_norm": 0.14672057330608368,
      "learning_rate": 7.976814794369307e-06,
      "loss": 0.4636,
      "mean_token_accuracy": 0.8557397276163101,
      "num_tokens": 957704606.0,
      "step": 30007
    },
    {
      "entropy": 0.49375022761523724,
      "epoch": 2.7608926918189085,
      "grad_norm": 0.1542114019393921,
      "learning_rate": 7.973747968227682e-06,
      "loss": 0.4821,
      "mean_token_accuracy": 0.8495172448456287,
      "num_tokens": 957737017.0,
      "step": 30008
    },
    {
      "entropy": 0.5009888093918562,
      "epoch": 2.7609846981903496,
      "grad_norm": 0.15470799803733826,
      "learning_rate": 7.970681142086056e-06,
      "loss": 0.4792,
      "mean_token_accuracy": 0.8454946279525757,
      "num_tokens": 957769032.0,
      "step": 30009
    },
    {
      "entropy": 0.5195128517225385,
      "epoch": 2.7610767045617908,
      "grad_norm": 0.15714654326438904,
      "learning_rate": 7.967614315944429e-06,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8390345461666584,
      "num_tokens": 957800484.0,
      "step": 30010
    },
    {
      "entropy": 0.5079778889194131,
      "epoch": 2.761168710933232,
      "grad_norm": 0.16601811349391937,
      "learning_rate": 7.964547489802804e-06,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8453154340386391,
      "num_tokens": 957832148.0,
      "step": 30011
    },
    {
      "entropy": 0.5134195704013109,
      "epoch": 2.7612607173046735,
      "grad_norm": 0.15933555364608765,
      "learning_rate": 7.961480663661178e-06,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8426645696163177,
      "num_tokens": 957863779.0,
      "step": 30012
    },
    {
      "entropy": 0.5911111757159233,
      "epoch": 2.7613527236761146,
      "grad_norm": 0.1685418337583542,
      "learning_rate": 7.958413837519551e-06,
      "loss": 0.5737,
      "mean_token_accuracy": 0.8183853775262833,
      "num_tokens": 957894818.0,
      "step": 30013
    },
    {
      "entropy": 0.5372423976659775,
      "epoch": 2.7614447300475558,
      "grad_norm": 0.16031751036643982,
      "learning_rate": 7.955347011377926e-06,
      "loss": 0.527,
      "mean_token_accuracy": 0.8353006131947041,
      "num_tokens": 957927470.0,
      "step": 30014
    },
    {
      "entropy": 0.4720791387371719,
      "epoch": 2.761536736418997,
      "grad_norm": 0.14923764765262604,
      "learning_rate": 7.952280185236299e-06,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8507525771856308,
      "num_tokens": 957959947.0,
      "step": 30015
    },
    {
      "entropy": 0.48819294082932174,
      "epoch": 2.761628742790438,
      "grad_norm": 0.15845395624637604,
      "learning_rate": 7.949213359094674e-06,
      "loss": 0.4758,
      "mean_token_accuracy": 0.8473623469471931,
      "num_tokens": 957991681.0,
      "step": 30016
    },
    {
      "entropy": 0.5139734854456037,
      "epoch": 2.7617207491618796,
      "grad_norm": 0.16216206550598145,
      "learning_rate": 7.946146532953047e-06,
      "loss": 0.5007,
      "mean_token_accuracy": 0.840963888913393,
      "num_tokens": 958023377.0,
      "step": 30017
    },
    {
      "entropy": 0.5770497601479292,
      "epoch": 2.7618127555333207,
      "grad_norm": 0.17251715064048767,
      "learning_rate": 7.943079706811421e-06,
      "loss": 0.5675,
      "mean_token_accuracy": 0.8222204856574535,
      "num_tokens": 958054709.0,
      "step": 30018
    },
    {
      "entropy": 0.4508885871618986,
      "epoch": 2.761904761904762,
      "grad_norm": 0.15066035091876984,
      "learning_rate": 7.940012880669794e-06,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8573912717401981,
      "num_tokens": 958086436.0,
      "step": 30019
    },
    {
      "entropy": 0.5124559011310339,
      "epoch": 2.761996768276203,
      "grad_norm": 0.1635720282793045,
      "learning_rate": 7.936946054528169e-06,
      "loss": 0.4949,
      "mean_token_accuracy": 0.841806560754776,
      "num_tokens": 958117570.0,
      "step": 30020
    },
    {
      "entropy": 0.4637379082851112,
      "epoch": 2.762088774647644,
      "grad_norm": 0.1593906730413437,
      "learning_rate": 7.933879228386542e-06,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8496413975954056,
      "num_tokens": 958149803.0,
      "step": 30021
    },
    {
      "entropy": 0.5997186279855669,
      "epoch": 2.7621807810190857,
      "grad_norm": 0.17051395773887634,
      "learning_rate": 7.930812402244916e-06,
      "loss": 0.591,
      "mean_token_accuracy": 0.8163600862026215,
      "num_tokens": 958182196.0,
      "step": 30022
    },
    {
      "entropy": 0.47363984026014805,
      "epoch": 2.762272787390527,
      "grad_norm": 0.1601383239030838,
      "learning_rate": 7.927745576103291e-06,
      "loss": 0.4603,
      "mean_token_accuracy": 0.8557040207087994,
      "num_tokens": 958213126.0,
      "step": 30023
    },
    {
      "entropy": 0.45697830989956856,
      "epoch": 2.762364793761968,
      "grad_norm": 0.14891980588436127,
      "learning_rate": 7.924678749961666e-06,
      "loss": 0.438,
      "mean_token_accuracy": 0.8577915243804455,
      "num_tokens": 958245371.0,
      "step": 30024
    },
    {
      "entropy": 0.441016661003232,
      "epoch": 2.762456800133409,
      "grad_norm": 0.14732810854911804,
      "learning_rate": 7.921611923820039e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8631072528660297,
      "num_tokens": 958277898.0,
      "step": 30025
    },
    {
      "entropy": 0.5536710806190968,
      "epoch": 2.7625488065048502,
      "grad_norm": 0.16535864770412445,
      "learning_rate": 7.918545097678413e-06,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8304311782121658,
      "num_tokens": 958310140.0,
      "step": 30026
    },
    {
      "entropy": 0.4070487869903445,
      "epoch": 2.762640812876292,
      "grad_norm": 0.14777085185050964,
      "learning_rate": 7.915478271536788e-06,
      "loss": 0.3949,
      "mean_token_accuracy": 0.8723949640989304,
      "num_tokens": 958342047.0,
      "step": 30027
    },
    {
      "entropy": 0.5267861019819975,
      "epoch": 2.762732819247733,
      "grad_norm": 0.16180473566055298,
      "learning_rate": 7.912411445395161e-06,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8379891589283943,
      "num_tokens": 958373407.0,
      "step": 30028
    },
    {
      "entropy": 0.40946019859984517,
      "epoch": 2.762824825619174,
      "grad_norm": 0.14804024994373322,
      "learning_rate": 7.909344619253536e-06,
      "loss": 0.4056,
      "mean_token_accuracy": 0.8710776530206203,
      "num_tokens": 958404968.0,
      "step": 30029
    },
    {
      "entropy": 0.4431001003831625,
      "epoch": 2.762916831990615,
      "grad_norm": 0.15147672593593597,
      "learning_rate": 7.906277793111909e-06,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8612251505255699,
      "num_tokens": 958436592.0,
      "step": 30030
    },
    {
      "entropy": 0.5236626900732517,
      "epoch": 2.7630088383620564,
      "grad_norm": 0.16286858916282654,
      "learning_rate": 7.903210966970283e-06,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8353176638484001,
      "num_tokens": 958468466.0,
      "step": 30031
    },
    {
      "entropy": 0.519100128673017,
      "epoch": 2.763100844733498,
      "grad_norm": 0.15544646978378296,
      "learning_rate": 7.900144140828656e-06,
      "loss": 0.499,
      "mean_token_accuracy": 0.8397225588560104,
      "num_tokens": 958500635.0,
      "step": 30032
    },
    {
      "entropy": 0.5416872424539179,
      "epoch": 2.763192851104939,
      "grad_norm": 0.1617060750722885,
      "learning_rate": 7.89707731468703e-06,
      "loss": 0.533,
      "mean_token_accuracy": 0.8311657570302486,
      "num_tokens": 958532372.0,
      "step": 30033
    },
    {
      "entropy": 0.4714395422488451,
      "epoch": 2.76328485747638,
      "grad_norm": 0.15377452969551086,
      "learning_rate": 7.894010488545404e-06,
      "loss": 0.4615,
      "mean_token_accuracy": 0.853739358484745,
      "num_tokens": 958564569.0,
      "step": 30034
    },
    {
      "entropy": 0.4843179890885949,
      "epoch": 2.7633768638478213,
      "grad_norm": 0.1592552661895752,
      "learning_rate": 7.890943662403778e-06,
      "loss": 0.4743,
      "mean_token_accuracy": 0.85071861743927,
      "num_tokens": 958596572.0,
      "step": 30035
    },
    {
      "entropy": 0.5703213345259428,
      "epoch": 2.7634688702192625,
      "grad_norm": 0.16432732343673706,
      "learning_rate": 7.887876836262151e-06,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8241181261837482,
      "num_tokens": 958629033.0,
      "step": 30036
    },
    {
      "entropy": 0.5296684503555298,
      "epoch": 2.763560876590704,
      "grad_norm": 0.15728779137134552,
      "learning_rate": 7.884810010120526e-06,
      "loss": 0.5077,
      "mean_token_accuracy": 0.8361679278314114,
      "num_tokens": 958661139.0,
      "step": 30037
    },
    {
      "entropy": 0.4992525177076459,
      "epoch": 2.763652882962145,
      "grad_norm": 0.1563761979341507,
      "learning_rate": 7.8817431839789e-06,
      "loss": 0.4893,
      "mean_token_accuracy": 0.850221324712038,
      "num_tokens": 958692899.0,
      "step": 30038
    },
    {
      "entropy": 0.6164165968075395,
      "epoch": 2.7637448893335863,
      "grad_norm": 0.1701485961675644,
      "learning_rate": 7.878676357837275e-06,
      "loss": 0.6122,
      "mean_token_accuracy": 0.8068667612969875,
      "num_tokens": 958725282.0,
      "step": 30039
    },
    {
      "entropy": 0.4431082755327225,
      "epoch": 2.7638368957050274,
      "grad_norm": 0.15064550936222076,
      "learning_rate": 7.875609531695648e-06,
      "loss": 0.4256,
      "mean_token_accuracy": 0.8662699945271015,
      "num_tokens": 958757534.0,
      "step": 30040
    },
    {
      "entropy": 0.4146465518279001,
      "epoch": 2.7639289020764686,
      "grad_norm": 0.15072540938854218,
      "learning_rate": 7.872542705554023e-06,
      "loss": 0.4151,
      "mean_token_accuracy": 0.8649309426546097,
      "num_tokens": 958789767.0,
      "step": 30041
    },
    {
      "entropy": 0.5449247335782275,
      "epoch": 2.76402090844791,
      "grad_norm": 0.16225561499595642,
      "learning_rate": 7.869475879412398e-06,
      "loss": 0.5337,
      "mean_token_accuracy": 0.8327939994633198,
      "num_tokens": 958821276.0,
      "step": 30042
    },
    {
      "entropy": 0.4947303617373109,
      "epoch": 2.7641129148193513,
      "grad_norm": 0.15905845165252686,
      "learning_rate": 7.86640905327077e-06,
      "loss": 0.4821,
      "mean_token_accuracy": 0.8472255282104015,
      "num_tokens": 958853036.0,
      "step": 30043
    },
    {
      "entropy": 0.5035746432840824,
      "epoch": 2.7642049211907924,
      "grad_norm": 0.15641196072101593,
      "learning_rate": 7.863342227129145e-06,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8465834818780422,
      "num_tokens": 958884968.0,
      "step": 30044
    },
    {
      "entropy": 0.578776319976896,
      "epoch": 2.7642969275622336,
      "grad_norm": 0.15779681503772736,
      "learning_rate": 7.860275400987518e-06,
      "loss": 0.5636,
      "mean_token_accuracy": 0.823757067322731,
      "num_tokens": 958917119.0,
      "step": 30045
    },
    {
      "entropy": 0.547251008450985,
      "epoch": 2.7643889339336747,
      "grad_norm": 0.1627039909362793,
      "learning_rate": 7.857208574845893e-06,
      "loss": 0.5475,
      "mean_token_accuracy": 0.8266423307359219,
      "num_tokens": 958949367.0,
      "step": 30046
    },
    {
      "entropy": 0.5273563181981444,
      "epoch": 2.7644809403051163,
      "grad_norm": 0.1608188897371292,
      "learning_rate": 7.854141748704266e-06,
      "loss": 0.5216,
      "mean_token_accuracy": 0.8365434668958187,
      "num_tokens": 958981855.0,
      "step": 30047
    },
    {
      "entropy": 0.4781040921807289,
      "epoch": 2.7645729466765574,
      "grad_norm": 0.1503126323223114,
      "learning_rate": 7.85107492256264e-06,
      "loss": 0.4603,
      "mean_token_accuracy": 0.8537978157401085,
      "num_tokens": 959013921.0,
      "step": 30048
    },
    {
      "entropy": 0.558137378655374,
      "epoch": 2.7646649530479985,
      "grad_norm": 0.16375525295734406,
      "learning_rate": 7.848008096421013e-06,
      "loss": 0.5416,
      "mean_token_accuracy": 0.8271457552909851,
      "num_tokens": 959046297.0,
      "step": 30049
    },
    {
      "entropy": 0.44291747873649,
      "epoch": 2.7647569594194397,
      "grad_norm": 0.14678139984607697,
      "learning_rate": 7.844941270279388e-06,
      "loss": 0.4236,
      "mean_token_accuracy": 0.8637195713818073,
      "num_tokens": 959078161.0,
      "step": 30050
    },
    {
      "entropy": 0.5329619823023677,
      "epoch": 2.764848965790881,
      "grad_norm": 0.16190147399902344,
      "learning_rate": 7.841874444137761e-06,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8367320224642754,
      "num_tokens": 959109617.0,
      "step": 30051
    },
    {
      "entropy": 0.5082565452903509,
      "epoch": 2.7649409721623224,
      "grad_norm": 0.16024036705493927,
      "learning_rate": 7.838807617996136e-06,
      "loss": 0.5004,
      "mean_token_accuracy": 0.8388808071613312,
      "num_tokens": 959141467.0,
      "step": 30052
    },
    {
      "entropy": 0.5319426797796041,
      "epoch": 2.7650329785337635,
      "grad_norm": 0.1594778448343277,
      "learning_rate": 7.83574079185451e-06,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8356972225010395,
      "num_tokens": 959173864.0,
      "step": 30053
    },
    {
      "entropy": 0.5697488989681005,
      "epoch": 2.7651249849052046,
      "grad_norm": 0.1646081805229187,
      "learning_rate": 7.832673965712885e-06,
      "loss": 0.5624,
      "mean_token_accuracy": 0.8234316222369671,
      "num_tokens": 959205434.0,
      "step": 30054
    },
    {
      "entropy": 0.5074060568585992,
      "epoch": 2.7652169912766458,
      "grad_norm": 0.15950417518615723,
      "learning_rate": 7.829607139571258e-06,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8427050188183784,
      "num_tokens": 959236709.0,
      "step": 30055
    },
    {
      "entropy": 0.4954108679667115,
      "epoch": 2.765308997648087,
      "grad_norm": 0.1509591042995453,
      "learning_rate": 7.826540313429632e-06,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8495860397815704,
      "num_tokens": 959268989.0,
      "step": 30056
    },
    {
      "entropy": 0.4670657478272915,
      "epoch": 2.7654010040195285,
      "grad_norm": 0.15351584553718567,
      "learning_rate": 7.823473487288007e-06,
      "loss": 0.4646,
      "mean_token_accuracy": 0.8535315170884132,
      "num_tokens": 959301482.0,
      "step": 30057
    },
    {
      "entropy": 0.5240169770549983,
      "epoch": 2.7654930103909696,
      "grad_norm": 0.1628950983285904,
      "learning_rate": 7.82040666114638e-06,
      "loss": 0.5145,
      "mean_token_accuracy": 0.8370962366461754,
      "num_tokens": 959333457.0,
      "step": 30058
    },
    {
      "entropy": 0.45888426434248686,
      "epoch": 2.7655850167624108,
      "grad_norm": 0.14986972510814667,
      "learning_rate": 7.817339835004755e-06,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8576328456401825,
      "num_tokens": 959365209.0,
      "step": 30059
    },
    {
      "entropy": 0.45558553468436,
      "epoch": 2.765677023133852,
      "grad_norm": 0.1507546752691269,
      "learning_rate": 7.814273008863128e-06,
      "loss": 0.4331,
      "mean_token_accuracy": 0.8596733659505844,
      "num_tokens": 959396445.0,
      "step": 30060
    },
    {
      "entropy": 0.4997830595821142,
      "epoch": 2.765769029505293,
      "grad_norm": 0.1545708328485489,
      "learning_rate": 7.811206182721502e-06,
      "loss": 0.4819,
      "mean_token_accuracy": 0.8439701199531555,
      "num_tokens": 959428236.0,
      "step": 30061
    },
    {
      "entropy": 0.45794543880037963,
      "epoch": 2.7658610358767346,
      "grad_norm": 0.15208609402179718,
      "learning_rate": 7.808139356579875e-06,
      "loss": 0.4442,
      "mean_token_accuracy": 0.8608723245561123,
      "num_tokens": 959460323.0,
      "step": 30062
    },
    {
      "entropy": 0.4510920997709036,
      "epoch": 2.7659530422481757,
      "grad_norm": 0.1522589921951294,
      "learning_rate": 7.80507253043825e-06,
      "loss": 0.4313,
      "mean_token_accuracy": 0.857331920415163,
      "num_tokens": 959491791.0,
      "step": 30063
    },
    {
      "entropy": 0.477485170122236,
      "epoch": 2.766045048619617,
      "grad_norm": 0.15296617150306702,
      "learning_rate": 7.802005704296623e-06,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8503394201397896,
      "num_tokens": 959523776.0,
      "step": 30064
    },
    {
      "entropy": 0.5312938038259745,
      "epoch": 2.766137054991058,
      "grad_norm": 0.17254239320755005,
      "learning_rate": 7.798938878154997e-06,
      "loss": 0.5243,
      "mean_token_accuracy": 0.8335315249860287,
      "num_tokens": 959555493.0,
      "step": 30065
    },
    {
      "entropy": 0.44461636152118444,
      "epoch": 2.766229061362499,
      "grad_norm": 0.1537042111158371,
      "learning_rate": 7.79587205201337e-06,
      "loss": 0.4417,
      "mean_token_accuracy": 0.8571821562945843,
      "num_tokens": 959586424.0,
      "step": 30066
    },
    {
      "entropy": 0.47284030821174383,
      "epoch": 2.7663210677339407,
      "grad_norm": 0.15120626986026764,
      "learning_rate": 7.792805225871745e-06,
      "loss": 0.4612,
      "mean_token_accuracy": 0.8517753556370735,
      "num_tokens": 959618622.0,
      "step": 30067
    },
    {
      "entropy": 0.5021800515241921,
      "epoch": 2.766413074105382,
      "grad_norm": 0.1554509699344635,
      "learning_rate": 7.78973839973012e-06,
      "loss": 0.4967,
      "mean_token_accuracy": 0.8417912498116493,
      "num_tokens": 959650219.0,
      "step": 30068
    },
    {
      "entropy": 0.48746518418192863,
      "epoch": 2.766505080476823,
      "grad_norm": 0.15820634365081787,
      "learning_rate": 7.786671573588494e-06,
      "loss": 0.4735,
      "mean_token_accuracy": 0.8483477421104908,
      "num_tokens": 959681659.0,
      "step": 30069
    },
    {
      "entropy": 0.43425661930814385,
      "epoch": 2.766597086848264,
      "grad_norm": 0.14974509179592133,
      "learning_rate": 7.783604747446869e-06,
      "loss": 0.42,
      "mean_token_accuracy": 0.8658670149743557,
      "num_tokens": 959713063.0,
      "step": 30070
    },
    {
      "entropy": 0.5571286845952272,
      "epoch": 2.7666890932197052,
      "grad_norm": 0.16694536805152893,
      "learning_rate": 7.780537921305242e-06,
      "loss": 0.5372,
      "mean_token_accuracy": 0.8303327076137066,
      "num_tokens": 959744977.0,
      "step": 30071
    },
    {
      "entropy": 0.4246750958263874,
      "epoch": 2.766781099591147,
      "grad_norm": 0.14307403564453125,
      "learning_rate": 7.777471095163617e-06,
      "loss": 0.4043,
      "mean_token_accuracy": 0.8715818114578724,
      "num_tokens": 959776910.0,
      "step": 30072
    },
    {
      "entropy": 0.5462467791512609,
      "epoch": 2.766873105962588,
      "grad_norm": 0.16461490094661713,
      "learning_rate": 7.77440426902199e-06,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8364521339535713,
      "num_tokens": 959808026.0,
      "step": 30073
    },
    {
      "entropy": 0.46860216837376356,
      "epoch": 2.766965112334029,
      "grad_norm": 0.1529827117919922,
      "learning_rate": 7.771337442880364e-06,
      "loss": 0.46,
      "mean_token_accuracy": 0.8540273867547512,
      "num_tokens": 959839698.0,
      "step": 30074
    },
    {
      "entropy": 0.4985158839263022,
      "epoch": 2.76705711870547,
      "grad_norm": 0.15241900086402893,
      "learning_rate": 7.768270616738737e-06,
      "loss": 0.4893,
      "mean_token_accuracy": 0.8409603498876095,
      "num_tokens": 959871826.0,
      "step": 30075
    },
    {
      "entropy": 0.5314660102594644,
      "epoch": 2.7671491250769114,
      "grad_norm": 0.16169479489326477,
      "learning_rate": 7.765203790597112e-06,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8362267166376114,
      "num_tokens": 959903863.0,
      "step": 30076
    },
    {
      "entropy": 0.5312954029068351,
      "epoch": 2.767241131448353,
      "grad_norm": 0.15705808997154236,
      "learning_rate": 7.762136964455485e-06,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8358719125390053,
      "num_tokens": 959935745.0,
      "step": 30077
    },
    {
      "entropy": 0.4771018708124757,
      "epoch": 2.767333137819794,
      "grad_norm": 0.1555468887090683,
      "learning_rate": 7.75907013831386e-06,
      "loss": 0.4692,
      "mean_token_accuracy": 0.8483719117939472,
      "num_tokens": 959967614.0,
      "step": 30078
    },
    {
      "entropy": 0.474789930973202,
      "epoch": 2.767425144191235,
      "grad_norm": 0.1465718001127243,
      "learning_rate": 7.756003312172232e-06,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8546957522630692,
      "num_tokens": 960000104.0,
      "step": 30079
    },
    {
      "entropy": 0.4436786463484168,
      "epoch": 2.7675171505626763,
      "grad_norm": 0.15779127180576324,
      "learning_rate": 7.752936486030607e-06,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8598615601658821,
      "num_tokens": 960030647.0,
      "step": 30080
    },
    {
      "entropy": 0.49642858328297734,
      "epoch": 2.7676091569341175,
      "grad_norm": 0.1592485010623932,
      "learning_rate": 7.749869659888982e-06,
      "loss": 0.4859,
      "mean_token_accuracy": 0.8439243882894516,
      "num_tokens": 960061732.0,
      "step": 30081
    },
    {
      "entropy": 0.6079172100871801,
      "epoch": 2.767701163305559,
      "grad_norm": 0.1669425070285797,
      "learning_rate": 7.746802833747355e-06,
      "loss": 0.5952,
      "mean_token_accuracy": 0.8129167295992374,
      "num_tokens": 960093917.0,
      "step": 30082
    },
    {
      "entropy": 0.47738480754196644,
      "epoch": 2.767793169677,
      "grad_norm": 0.1542140543460846,
      "learning_rate": 7.74373600760573e-06,
      "loss": 0.4725,
      "mean_token_accuracy": 0.8509101942181587,
      "num_tokens": 960125577.0,
      "step": 30083
    },
    {
      "entropy": 0.5317522054538131,
      "epoch": 2.7678851760484413,
      "grad_norm": 0.16063496470451355,
      "learning_rate": 7.740669181464104e-06,
      "loss": 0.5134,
      "mean_token_accuracy": 0.834821842610836,
      "num_tokens": 960157983.0,
      "step": 30084
    },
    {
      "entropy": 0.5827318169176579,
      "epoch": 2.7679771824198824,
      "grad_norm": 0.17105095088481903,
      "learning_rate": 7.737602355322479e-06,
      "loss": 0.5694,
      "mean_token_accuracy": 0.8197539187967777,
      "num_tokens": 960189197.0,
      "step": 30085
    },
    {
      "entropy": 0.5049424264580011,
      "epoch": 2.7680691887913236,
      "grad_norm": 0.14799146354198456,
      "learning_rate": 7.734535529180852e-06,
      "loss": 0.4845,
      "mean_token_accuracy": 0.8483883738517761,
      "num_tokens": 960221871.0,
      "step": 30086
    },
    {
      "entropy": 0.5591127034276724,
      "epoch": 2.768161195162765,
      "grad_norm": 0.16333241760730743,
      "learning_rate": 7.731468703039226e-06,
      "loss": 0.5486,
      "mean_token_accuracy": 0.829078059643507,
      "num_tokens": 960254386.0,
      "step": 30087
    },
    {
      "entropy": 0.5397195667028427,
      "epoch": 2.7682532015342063,
      "grad_norm": 0.1579403579235077,
      "learning_rate": 7.728401876897599e-06,
      "loss": 0.5279,
      "mean_token_accuracy": 0.8351899832487106,
      "num_tokens": 960286300.0,
      "step": 30088
    },
    {
      "entropy": 0.5981874521821737,
      "epoch": 2.7683452079056474,
      "grad_norm": 0.174805149435997,
      "learning_rate": 7.725335050755974e-06,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8125674277544022,
      "num_tokens": 960317255.0,
      "step": 30089
    },
    {
      "entropy": 0.5725357127375901,
      "epoch": 2.7684372142770886,
      "grad_norm": 0.16081970930099487,
      "learning_rate": 7.722268224614347e-06,
      "loss": 0.5609,
      "mean_token_accuracy": 0.8234420977532864,
      "num_tokens": 960349474.0,
      "step": 30090
    },
    {
      "entropy": 0.544660521671176,
      "epoch": 2.7685292206485297,
      "grad_norm": 0.1628876030445099,
      "learning_rate": 7.719201398472721e-06,
      "loss": 0.5258,
      "mean_token_accuracy": 0.83529132604599,
      "num_tokens": 960381279.0,
      "step": 30091
    },
    {
      "entropy": 0.46898289024829865,
      "epoch": 2.7686212270199713,
      "grad_norm": 0.15369200706481934,
      "learning_rate": 7.716134572331094e-06,
      "loss": 0.4621,
      "mean_token_accuracy": 0.8535637520253658,
      "num_tokens": 960413284.0,
      "step": 30092
    },
    {
      "entropy": 0.3799782444257289,
      "epoch": 2.7687132333914124,
      "grad_norm": 0.14065004885196686,
      "learning_rate": 7.713067746189469e-06,
      "loss": 0.3706,
      "mean_token_accuracy": 0.8803708218038082,
      "num_tokens": 960445566.0,
      "step": 30093
    },
    {
      "entropy": 0.5373425912111998,
      "epoch": 2.7688052397628535,
      "grad_norm": 0.1614452749490738,
      "learning_rate": 7.710000920047842e-06,
      "loss": 0.5325,
      "mean_token_accuracy": 0.8343115635216236,
      "num_tokens": 960478303.0,
      "step": 30094
    },
    {
      "entropy": 0.618566494435072,
      "epoch": 2.7688972461342947,
      "grad_norm": 0.17293521761894226,
      "learning_rate": 7.706934093906217e-06,
      "loss": 0.6126,
      "mean_token_accuracy": 0.8084633201360703,
      "num_tokens": 960509505.0,
      "step": 30095
    },
    {
      "entropy": 0.45195197430439293,
      "epoch": 2.768989252505736,
      "grad_norm": 0.15112872421741486,
      "learning_rate": 7.703867267764591e-06,
      "loss": 0.4446,
      "mean_token_accuracy": 0.8621690347790718,
      "num_tokens": 960541841.0,
      "step": 30096
    },
    {
      "entropy": 0.4985244544222951,
      "epoch": 2.7690812588771774,
      "grad_norm": 0.16164082288742065,
      "learning_rate": 7.700800441622964e-06,
      "loss": 0.4967,
      "mean_token_accuracy": 0.8454005941748619,
      "num_tokens": 960573410.0,
      "step": 30097
    },
    {
      "entropy": 0.48411201674025506,
      "epoch": 2.7691732652486185,
      "grad_norm": 0.1577129364013672,
      "learning_rate": 7.697733615481339e-06,
      "loss": 0.4739,
      "mean_token_accuracy": 0.8464302718639374,
      "num_tokens": 960605554.0,
      "step": 30098
    },
    {
      "entropy": 0.4497814476490021,
      "epoch": 2.7692652716200596,
      "grad_norm": 0.15145446360111237,
      "learning_rate": 7.694666789339712e-06,
      "loss": 0.4386,
      "mean_token_accuracy": 0.8569777607917786,
      "num_tokens": 960637927.0,
      "step": 30099
    },
    {
      "entropy": 0.4598412432242185,
      "epoch": 2.7693572779915008,
      "grad_norm": 0.14844313263893127,
      "learning_rate": 7.691599963198088e-06,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8608082123100758,
      "num_tokens": 960670395.0,
      "step": 30100
    },
    {
      "entropy": 0.4575711116194725,
      "epoch": 2.769449284362942,
      "grad_norm": 0.1502234786748886,
      "learning_rate": 7.688533137056461e-06,
      "loss": 0.4452,
      "mean_token_accuracy": 0.8557686246931553,
      "num_tokens": 960702945.0,
      "step": 30101
    },
    {
      "entropy": 0.5411399099975824,
      "epoch": 2.7695412907343835,
      "grad_norm": 0.16386887431144714,
      "learning_rate": 7.685466310914836e-06,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8317503593862057,
      "num_tokens": 960734368.0,
      "step": 30102
    },
    {
      "entropy": 0.5905276145786047,
      "epoch": 2.7696332971058246,
      "grad_norm": 0.17647685110569,
      "learning_rate": 7.682399484773209e-06,
      "loss": 0.5734,
      "mean_token_accuracy": 0.8171955868601799,
      "num_tokens": 960764490.0,
      "step": 30103
    },
    {
      "entropy": 0.5593867730349302,
      "epoch": 2.7697253034772658,
      "grad_norm": 0.17242586612701416,
      "learning_rate": 7.679332658631583e-06,
      "loss": 0.548,
      "mean_token_accuracy": 0.825010385364294,
      "num_tokens": 960795733.0,
      "step": 30104
    },
    {
      "entropy": 0.5482597695663571,
      "epoch": 2.769817309848707,
      "grad_norm": 0.1601020097732544,
      "learning_rate": 7.676265832489956e-06,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8295603543519974,
      "num_tokens": 960828500.0,
      "step": 30105
    },
    {
      "entropy": 0.5967402551323175,
      "epoch": 2.769909316220148,
      "grad_norm": 0.16882774233818054,
      "learning_rate": 7.673199006348331e-06,
      "loss": 0.5805,
      "mean_token_accuracy": 0.8186974786221981,
      "num_tokens": 960859633.0,
      "step": 30106
    },
    {
      "entropy": 0.5308121517300606,
      "epoch": 2.7700013225915896,
      "grad_norm": 0.15880228579044342,
      "learning_rate": 7.670132180206704e-06,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8377617448568344,
      "num_tokens": 960891690.0,
      "step": 30107
    },
    {
      "entropy": 0.5584489731118083,
      "epoch": 2.7700933289630307,
      "grad_norm": 0.16274955868721008,
      "learning_rate": 7.667065354065079e-06,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8254114575684071,
      "num_tokens": 960923595.0,
      "step": 30108
    },
    {
      "entropy": 0.5679116537794471,
      "epoch": 2.770185335334472,
      "grad_norm": 0.1640646904706955,
      "learning_rate": 7.663998527923451e-06,
      "loss": 0.5733,
      "mean_token_accuracy": 0.8205646425485611,
      "num_tokens": 960955316.0,
      "step": 30109
    },
    {
      "entropy": 0.5328402773011476,
      "epoch": 2.770277341705913,
      "grad_norm": 0.15464037656784058,
      "learning_rate": 7.660931701781826e-06,
      "loss": 0.5146,
      "mean_token_accuracy": 0.8353264629840851,
      "num_tokens": 960987542.0,
      "step": 30110
    },
    {
      "entropy": 0.5474590891972184,
      "epoch": 2.770369348077354,
      "grad_norm": 0.17020855844020844,
      "learning_rate": 7.6578648756402e-06,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8308403603732586,
      "num_tokens": 961018428.0,
      "step": 30111
    },
    {
      "entropy": 0.5312814535573125,
      "epoch": 2.7704613544487957,
      "grad_norm": 0.1591482311487198,
      "learning_rate": 7.654798049498574e-06,
      "loss": 0.5142,
      "mean_token_accuracy": 0.8349974080920219,
      "num_tokens": 961050103.0,
      "step": 30112
    },
    {
      "entropy": 0.4967648321762681,
      "epoch": 2.770553360820237,
      "grad_norm": 0.15291793644428253,
      "learning_rate": 7.651731223356948e-06,
      "loss": 0.4829,
      "mean_token_accuracy": 0.8478329256176949,
      "num_tokens": 961082639.0,
      "step": 30113
    },
    {
      "entropy": 0.44045969750732183,
      "epoch": 2.770645367191678,
      "grad_norm": 0.15030217170715332,
      "learning_rate": 7.648664397215321e-06,
      "loss": 0.438,
      "mean_token_accuracy": 0.8607475981116295,
      "num_tokens": 961114814.0,
      "step": 30114
    },
    {
      "entropy": 0.4709740597754717,
      "epoch": 2.770737373563119,
      "grad_norm": 0.15674304962158203,
      "learning_rate": 7.645597571073698e-06,
      "loss": 0.464,
      "mean_token_accuracy": 0.855041392147541,
      "num_tokens": 961147095.0,
      "step": 30115
    },
    {
      "entropy": 0.5020008618012071,
      "epoch": 2.7708293799345602,
      "grad_norm": 0.15775451064109802,
      "learning_rate": 7.64253074493207e-06,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8475651778280735,
      "num_tokens": 961179438.0,
      "step": 30116
    },
    {
      "entropy": 0.45951219368726015,
      "epoch": 2.770921386306002,
      "grad_norm": 0.15820825099945068,
      "learning_rate": 7.639463918790445e-06,
      "loss": 0.4521,
      "mean_token_accuracy": 0.8545589745044708,
      "num_tokens": 961211240.0,
      "step": 30117
    },
    {
      "entropy": 0.4877900304272771,
      "epoch": 2.771013392677443,
      "grad_norm": 0.15465344488620758,
      "learning_rate": 7.636397092648818e-06,
      "loss": 0.4719,
      "mean_token_accuracy": 0.8534874022006989,
      "num_tokens": 961243147.0,
      "step": 30118
    },
    {
      "entropy": 0.616034097969532,
      "epoch": 2.771105399048884,
      "grad_norm": 0.17276206612586975,
      "learning_rate": 7.633330266507193e-06,
      "loss": 0.6042,
      "mean_token_accuracy": 0.809585478156805,
      "num_tokens": 961274972.0,
      "step": 30119
    },
    {
      "entropy": 0.5134818106889725,
      "epoch": 2.771197405420325,
      "grad_norm": 0.1640978753566742,
      "learning_rate": 7.630263440365566e-06,
      "loss": 0.506,
      "mean_token_accuracy": 0.8401167690753937,
      "num_tokens": 961306557.0,
      "step": 30120
    },
    {
      "entropy": 0.3888130243867636,
      "epoch": 2.7712894117917664,
      "grad_norm": 0.14123201370239258,
      "learning_rate": 7.6271966142239405e-06,
      "loss": 0.3809,
      "mean_token_accuracy": 0.8772269859910011,
      "num_tokens": 961338854.0,
      "step": 30121
    },
    {
      "entropy": 0.6092704478651285,
      "epoch": 2.771381418163208,
      "grad_norm": 0.17262588441371918,
      "learning_rate": 7.624129788082314e-06,
      "loss": 0.5965,
      "mean_token_accuracy": 0.81248864158988,
      "num_tokens": 961370835.0,
      "step": 30122
    },
    {
      "entropy": 0.4776209418196231,
      "epoch": 2.771473424534649,
      "grad_norm": 0.15470457077026367,
      "learning_rate": 7.621062961940688e-06,
      "loss": 0.4566,
      "mean_token_accuracy": 0.8520879447460175,
      "num_tokens": 961402793.0,
      "step": 30123
    },
    {
      "entropy": 0.5453309919685125,
      "epoch": 2.77156543090609,
      "grad_norm": 0.16032670438289642,
      "learning_rate": 7.617996135799062e-06,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8363240845501423,
      "num_tokens": 961434951.0,
      "step": 30124
    },
    {
      "entropy": 0.5720095941796899,
      "epoch": 2.7716574372775313,
      "grad_norm": 0.1607406884431839,
      "learning_rate": 7.614929309657436e-06,
      "loss": 0.5638,
      "mean_token_accuracy": 0.8213072009384632,
      "num_tokens": 961467474.0,
      "step": 30125
    },
    {
      "entropy": 0.5760832540690899,
      "epoch": 2.7717494436489725,
      "grad_norm": 0.17169930040836334,
      "learning_rate": 7.6118624835158095e-06,
      "loss": 0.5727,
      "mean_token_accuracy": 0.8189746476709843,
      "num_tokens": 961499188.0,
      "step": 30126
    },
    {
      "entropy": 0.5180882681161165,
      "epoch": 2.771841450020414,
      "grad_norm": 0.15851587057113647,
      "learning_rate": 7.608795657374183e-06,
      "loss": 0.5041,
      "mean_token_accuracy": 0.8436696156859398,
      "num_tokens": 961531445.0,
      "step": 30127
    },
    {
      "entropy": 0.4494654922746122,
      "epoch": 2.771933456391855,
      "grad_norm": 0.14811749756336212,
      "learning_rate": 7.605728831232557e-06,
      "loss": 0.4417,
      "mean_token_accuracy": 0.8608969449996948,
      "num_tokens": 961563882.0,
      "step": 30128
    },
    {
      "entropy": 0.5981818032450974,
      "epoch": 2.7720254627632963,
      "grad_norm": 0.1720881164073944,
      "learning_rate": 7.602662005090931e-06,
      "loss": 0.588,
      "mean_token_accuracy": 0.81603067740798,
      "num_tokens": 961595769.0,
      "step": 30129
    },
    {
      "entropy": 0.5621543901506811,
      "epoch": 2.7721174691347374,
      "grad_norm": 0.16507165133953094,
      "learning_rate": 7.599595178949306e-06,
      "loss": 0.5485,
      "mean_token_accuracy": 0.8310214355587959,
      "num_tokens": 961627864.0,
      "step": 30130
    },
    {
      "entropy": 0.4864246482029557,
      "epoch": 2.7722094755061786,
      "grad_norm": 0.1601085066795349,
      "learning_rate": 7.59652835280768e-06,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8477525152266026,
      "num_tokens": 961659768.0,
      "step": 30131
    },
    {
      "entropy": 0.4337195814587176,
      "epoch": 2.77230148187762,
      "grad_norm": 0.14900055527687073,
      "learning_rate": 7.593461526666054e-06,
      "loss": 0.4342,
      "mean_token_accuracy": 0.8624806366860867,
      "num_tokens": 961692386.0,
      "step": 30132
    },
    {
      "entropy": 0.45071278885006905,
      "epoch": 2.7723934882490613,
      "grad_norm": 0.15443088114261627,
      "learning_rate": 7.590394700524428e-06,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8599382489919662,
      "num_tokens": 961724817.0,
      "step": 30133
    },
    {
      "entropy": 0.5164574556984007,
      "epoch": 2.7724854946205024,
      "grad_norm": 0.1600361466407776,
      "learning_rate": 7.587327874382802e-06,
      "loss": 0.4904,
      "mean_token_accuracy": 0.8431217223405838,
      "num_tokens": 961755924.0,
      "step": 30134
    },
    {
      "entropy": 0.5456388090969995,
      "epoch": 2.7725775009919436,
      "grad_norm": 0.16571730375289917,
      "learning_rate": 7.584261048241175e-06,
      "loss": 0.5393,
      "mean_token_accuracy": 0.833349771797657,
      "num_tokens": 961786625.0,
      "step": 30135
    },
    {
      "entropy": 0.5090715894475579,
      "epoch": 2.7726695073633847,
      "grad_norm": 0.16041484475135803,
      "learning_rate": 7.58119422209955e-06,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8414842300117016,
      "num_tokens": 961817883.0,
      "step": 30136
    },
    {
      "entropy": 0.5417686172295362,
      "epoch": 2.7727615137348263,
      "grad_norm": 0.16650491952896118,
      "learning_rate": 7.578127395957924e-06,
      "loss": 0.5438,
      "mean_token_accuracy": 0.8278916142880917,
      "num_tokens": 961850040.0,
      "step": 30137
    },
    {
      "entropy": 0.5481363078579307,
      "epoch": 2.7728535201062674,
      "grad_norm": 0.16506806015968323,
      "learning_rate": 7.575060569816298e-06,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8301356546580791,
      "num_tokens": 961881777.0,
      "step": 30138
    },
    {
      "entropy": 0.4776026784675196,
      "epoch": 2.7729455264777085,
      "grad_norm": 0.15072974562644958,
      "learning_rate": 7.5719937436746714e-06,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8557584099471569,
      "num_tokens": 961914163.0,
      "step": 30139
    },
    {
      "entropy": 0.5529961062129587,
      "epoch": 2.7730375328491497,
      "grad_norm": 0.16241300106048584,
      "learning_rate": 7.568926917533045e-06,
      "loss": 0.5419,
      "mean_token_accuracy": 0.8268226385116577,
      "num_tokens": 961945736.0,
      "step": 30140
    },
    {
      "entropy": 0.4368950469652191,
      "epoch": 2.773129539220591,
      "grad_norm": 0.14588132500648499,
      "learning_rate": 7.565860091391419e-06,
      "loss": 0.4197,
      "mean_token_accuracy": 0.865342278033495,
      "num_tokens": 961978045.0,
      "step": 30141
    },
    {
      "entropy": 0.5126914056017995,
      "epoch": 2.7732215455920324,
      "grad_norm": 0.16245803236961365,
      "learning_rate": 7.562793265249793e-06,
      "loss": 0.5109,
      "mean_token_accuracy": 0.8357092216610909,
      "num_tokens": 962009564.0,
      "step": 30142
    },
    {
      "entropy": 0.4612105544656515,
      "epoch": 2.7733135519634735,
      "grad_norm": 0.15134043991565704,
      "learning_rate": 7.559726439108167e-06,
      "loss": 0.4535,
      "mean_token_accuracy": 0.8537138812243938,
      "num_tokens": 962041766.0,
      "step": 30143
    },
    {
      "entropy": 0.5731869013980031,
      "epoch": 2.7734055583349146,
      "grad_norm": 0.16914604604244232,
      "learning_rate": 7.5566596129665404e-06,
      "loss": 0.5515,
      "mean_token_accuracy": 0.8232237920165062,
      "num_tokens": 962072915.0,
      "step": 30144
    },
    {
      "entropy": 0.44520464539527893,
      "epoch": 2.773497564706356,
      "grad_norm": 0.15309633314609528,
      "learning_rate": 7.553592786824916e-06,
      "loss": 0.4338,
      "mean_token_accuracy": 0.862535510212183,
      "num_tokens": 962104748.0,
      "step": 30145
    },
    {
      "entropy": 0.6381418006494641,
      "epoch": 2.773589571077797,
      "grad_norm": 0.17498357594013214,
      "learning_rate": 7.55052596068329e-06,
      "loss": 0.6225,
      "mean_token_accuracy": 0.8030645065009594,
      "num_tokens": 962136123.0,
      "step": 30146
    },
    {
      "entropy": 0.5370819671079516,
      "epoch": 2.7736815774492385,
      "grad_norm": 0.1641584187746048,
      "learning_rate": 7.5474591345416635e-06,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8312121331691742,
      "num_tokens": 962167883.0,
      "step": 30147
    },
    {
      "entropy": 0.46103311236947775,
      "epoch": 2.7737735838206796,
      "grad_norm": 0.1550750434398651,
      "learning_rate": 7.544392308400037e-06,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8588683940470219,
      "num_tokens": 962200006.0,
      "step": 30148
    },
    {
      "entropy": 0.4717797467019409,
      "epoch": 2.7738655901921208,
      "grad_norm": 0.15585917234420776,
      "learning_rate": 7.541325482258411e-06,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8510122373700142,
      "num_tokens": 962232299.0,
      "step": 30149
    },
    {
      "entropy": 0.5431217034347355,
      "epoch": 2.773957596563562,
      "grad_norm": 0.15874168276786804,
      "learning_rate": 7.538258656116785e-06,
      "loss": 0.5157,
      "mean_token_accuracy": 0.830884825438261,
      "num_tokens": 962263699.0,
      "step": 30150
    },
    {
      "entropy": 0.5018594323191792,
      "epoch": 2.774049602935003,
      "grad_norm": 0.1509477198123932,
      "learning_rate": 7.53519182997516e-06,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8459714725613594,
      "num_tokens": 962296220.0,
      "step": 30151
    },
    {
      "entropy": 0.5741939581930637,
      "epoch": 2.7741416093064446,
      "grad_norm": 0.16467346251010895,
      "learning_rate": 7.532125003833533e-06,
      "loss": 0.5689,
      "mean_token_accuracy": 0.8253219202160835,
      "num_tokens": 962328033.0,
      "step": 30152
    },
    {
      "entropy": 0.4772661894094199,
      "epoch": 2.7742336156778857,
      "grad_norm": 0.14806906878948212,
      "learning_rate": 7.529058177691907e-06,
      "loss": 0.4641,
      "mean_token_accuracy": 0.8530820310115814,
      "num_tokens": 962360522.0,
      "step": 30153
    },
    {
      "entropy": 0.5128633584827185,
      "epoch": 2.774325622049327,
      "grad_norm": 0.15550173819065094,
      "learning_rate": 7.525991351550281e-06,
      "loss": 0.498,
      "mean_token_accuracy": 0.8420321121811867,
      "num_tokens": 962392515.0,
      "step": 30154
    },
    {
      "entropy": 0.5179773396812379,
      "epoch": 2.774417628420768,
      "grad_norm": 0.1571410447359085,
      "learning_rate": 7.522924525408655e-06,
      "loss": 0.5055,
      "mean_token_accuracy": 0.8388318754732609,
      "num_tokens": 962424579.0,
      "step": 30155
    },
    {
      "entropy": 0.5652625244110823,
      "epoch": 2.774509634792209,
      "grad_norm": 0.16821259260177612,
      "learning_rate": 7.519857699267029e-06,
      "loss": 0.555,
      "mean_token_accuracy": 0.8248301856219769,
      "num_tokens": 962456317.0,
      "step": 30156
    },
    {
      "entropy": 0.44014307018369436,
      "epoch": 2.7746016411636507,
      "grad_norm": 0.14918428659439087,
      "learning_rate": 7.516790873125402e-06,
      "loss": 0.4337,
      "mean_token_accuracy": 0.8608517162501812,
      "num_tokens": 962488770.0,
      "step": 30157
    },
    {
      "entropy": 0.5108698960393667,
      "epoch": 2.774693647535092,
      "grad_norm": 0.15893830358982086,
      "learning_rate": 7.513724046983776e-06,
      "loss": 0.5008,
      "mean_token_accuracy": 0.8399659991264343,
      "num_tokens": 962520064.0,
      "step": 30158
    },
    {
      "entropy": 0.409782649949193,
      "epoch": 2.774785653906533,
      "grad_norm": 0.1438545137643814,
      "learning_rate": 7.51065722084215e-06,
      "loss": 0.3926,
      "mean_token_accuracy": 0.873170331120491,
      "num_tokens": 962552042.0,
      "step": 30159
    },
    {
      "entropy": 0.5585016654804349,
      "epoch": 2.774877660277974,
      "grad_norm": 0.16587771475315094,
      "learning_rate": 7.507590394700524e-06,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8288860954344273,
      "num_tokens": 962583879.0,
      "step": 30160
    },
    {
      "entropy": 0.5395308425650001,
      "epoch": 2.7749696666494152,
      "grad_norm": 0.16116905212402344,
      "learning_rate": 7.504523568558899e-06,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8317674659192562,
      "num_tokens": 962615419.0,
      "step": 30161
    },
    {
      "entropy": 0.5080512026324868,
      "epoch": 2.775061673020857,
      "grad_norm": 0.15993373095989227,
      "learning_rate": 7.501456742417273e-06,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8407741971313953,
      "num_tokens": 962646956.0,
      "step": 30162
    },
    {
      "entropy": 0.4959452087059617,
      "epoch": 2.775153679392298,
      "grad_norm": 0.16478511691093445,
      "learning_rate": 7.498389916275647e-06,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8415189757943153,
      "num_tokens": 962678972.0,
      "step": 30163
    },
    {
      "entropy": 0.5165138747543097,
      "epoch": 2.775245685763739,
      "grad_norm": 0.16010916233062744,
      "learning_rate": 7.495323090134021e-06,
      "loss": 0.5098,
      "mean_token_accuracy": 0.8369176350533962,
      "num_tokens": 962711014.0,
      "step": 30164
    },
    {
      "entropy": 0.4998969929292798,
      "epoch": 2.7753376921351802,
      "grad_norm": 0.15684524178504944,
      "learning_rate": 7.4922562639923945e-06,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8411190062761307,
      "num_tokens": 962742216.0,
      "step": 30165
    },
    {
      "entropy": 0.4070316720753908,
      "epoch": 2.7754296985066214,
      "grad_norm": 0.14582602679729462,
      "learning_rate": 7.489189437850769e-06,
      "loss": 0.3999,
      "mean_token_accuracy": 0.8730569258332253,
      "num_tokens": 962773605.0,
      "step": 30166
    },
    {
      "entropy": 0.41355462465435266,
      "epoch": 2.775521704878063,
      "grad_norm": 0.14229656755924225,
      "learning_rate": 7.486122611709143e-06,
      "loss": 0.3911,
      "mean_token_accuracy": 0.8732373453676701,
      "num_tokens": 962805884.0,
      "step": 30167
    },
    {
      "entropy": 0.479584245942533,
      "epoch": 2.775613711249504,
      "grad_norm": 0.1595303863286972,
      "learning_rate": 7.483055785567517e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8469848223030567,
      "num_tokens": 962838174.0,
      "step": 30168
    },
    {
      "entropy": 0.43994728999678046,
      "epoch": 2.775705717620945,
      "grad_norm": 0.14525434374809265,
      "learning_rate": 7.4799889594258905e-06,
      "loss": 0.4229,
      "mean_token_accuracy": 0.8660310991108418,
      "num_tokens": 962870627.0,
      "step": 30169
    },
    {
      "entropy": 0.5471788384020329,
      "epoch": 2.7757977239923863,
      "grad_norm": 0.17115166783332825,
      "learning_rate": 7.476922133284264e-06,
      "loss": 0.5448,
      "mean_token_accuracy": 0.824146531522274,
      "num_tokens": 962901400.0,
      "step": 30170
    },
    {
      "entropy": 0.5363881383091211,
      "epoch": 2.7758897303638275,
      "grad_norm": 0.16021664440631866,
      "learning_rate": 7.473855307142638e-06,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8362605608999729,
      "num_tokens": 962932809.0,
      "step": 30171
    },
    {
      "entropy": 0.5775961745530367,
      "epoch": 2.775981736735269,
      "grad_norm": 0.1741124987602234,
      "learning_rate": 7.470788481001012e-06,
      "loss": 0.5789,
      "mean_token_accuracy": 0.8181728422641754,
      "num_tokens": 962963665.0,
      "step": 30172
    },
    {
      "entropy": 0.5629171495093033,
      "epoch": 2.77607374310671,
      "grad_norm": 0.1630399078130722,
      "learning_rate": 7.467721654859386e-06,
      "loss": 0.564,
      "mean_token_accuracy": 0.8194240219891071,
      "num_tokens": 962995818.0,
      "step": 30173
    },
    {
      "entropy": 0.6260162685066462,
      "epoch": 2.7761657494781513,
      "grad_norm": 0.16996872425079346,
      "learning_rate": 7.4646548287177595e-06,
      "loss": 0.617,
      "mean_token_accuracy": 0.8071018569171429,
      "num_tokens": 963027963.0,
      "step": 30174
    },
    {
      "entropy": 0.5603058841079473,
      "epoch": 2.7762577558495924,
      "grad_norm": 0.1598076969385147,
      "learning_rate": 7.461588002576133e-06,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8285120576620102,
      "num_tokens": 963059709.0,
      "step": 30175
    },
    {
      "entropy": 0.6541766654700041,
      "epoch": 2.7763497622210336,
      "grad_norm": 0.1742902547121048,
      "learning_rate": 7.458521176434509e-06,
      "loss": 0.6295,
      "mean_token_accuracy": 0.8022371456027031,
      "num_tokens": 963090375.0,
      "step": 30176
    },
    {
      "entropy": 0.5632693893276155,
      "epoch": 2.776441768592475,
      "grad_norm": 0.16292855143547058,
      "learning_rate": 7.455454350292883e-06,
      "loss": 0.548,
      "mean_token_accuracy": 0.8264997340738773,
      "num_tokens": 963121731.0,
      "step": 30177
    },
    {
      "entropy": 0.5391836911439896,
      "epoch": 2.7765337749639163,
      "grad_norm": 0.16310860216617584,
      "learning_rate": 7.4523875241512564e-06,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8328126557171345,
      "num_tokens": 963153450.0,
      "step": 30178
    },
    {
      "entropy": 0.48508715629577637,
      "epoch": 2.7766257813353574,
      "grad_norm": 0.1502898782491684,
      "learning_rate": 7.44932069800963e-06,
      "loss": 0.4724,
      "mean_token_accuracy": 0.849189143627882,
      "num_tokens": 963185980.0,
      "step": 30179
    },
    {
      "entropy": 0.4216093705035746,
      "epoch": 2.7767177877067986,
      "grad_norm": 0.1458195447921753,
      "learning_rate": 7.446253871868004e-06,
      "loss": 0.4129,
      "mean_token_accuracy": 0.8683932535350323,
      "num_tokens": 963218016.0,
      "step": 30180
    },
    {
      "entropy": 0.44914725003764033,
      "epoch": 2.7768097940782397,
      "grad_norm": 0.1493915468454361,
      "learning_rate": 7.443187045726379e-06,
      "loss": 0.4296,
      "mean_token_accuracy": 0.8628692626953125,
      "num_tokens": 963249774.0,
      "step": 30181
    },
    {
      "entropy": 0.5806281166151166,
      "epoch": 2.7769018004496813,
      "grad_norm": 0.16578975319862366,
      "learning_rate": 7.4401202195847525e-06,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8185931481420994,
      "num_tokens": 963281623.0,
      "step": 30182
    },
    {
      "entropy": 0.5914466353133321,
      "epoch": 2.7769938068211224,
      "grad_norm": 0.1725785732269287,
      "learning_rate": 7.437053393443126e-06,
      "loss": 0.5884,
      "mean_token_accuracy": 0.8153646178543568,
      "num_tokens": 963314053.0,
      "step": 30183
    },
    {
      "entropy": 0.48391136061400175,
      "epoch": 2.7770858131925635,
      "grad_norm": 0.15843309462070465,
      "learning_rate": 7.4339865673015e-06,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8478245623409748,
      "num_tokens": 963346211.0,
      "step": 30184
    },
    {
      "entropy": 0.47382583329454064,
      "epoch": 2.7771778195640047,
      "grad_norm": 0.15393872559070587,
      "learning_rate": 7.430919741159874e-06,
      "loss": 0.4646,
      "mean_token_accuracy": 0.8509871028363705,
      "num_tokens": 963378763.0,
      "step": 30185
    },
    {
      "entropy": 0.5307508653495461,
      "epoch": 2.777269825935446,
      "grad_norm": 0.15959204733371735,
      "learning_rate": 7.427852915018248e-06,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8351945877075195,
      "num_tokens": 963411387.0,
      "step": 30186
    },
    {
      "entropy": 0.5093901269137859,
      "epoch": 2.7773618323068874,
      "grad_norm": 0.15609899163246155,
      "learning_rate": 7.4247860888766215e-06,
      "loss": 0.4868,
      "mean_token_accuracy": 0.8389490768313408,
      "num_tokens": 963442642.0,
      "step": 30187
    },
    {
      "entropy": 0.5936266761273146,
      "epoch": 2.7774538386783285,
      "grad_norm": 0.16244499385356903,
      "learning_rate": 7.421719262734995e-06,
      "loss": 0.5808,
      "mean_token_accuracy": 0.8163107521831989,
      "num_tokens": 963474750.0,
      "step": 30188
    },
    {
      "entropy": 0.5114194257184863,
      "epoch": 2.7775458450497696,
      "grad_norm": 0.15402820706367493,
      "learning_rate": 7.418652436593369e-06,
      "loss": 0.4995,
      "mean_token_accuracy": 0.8428052514791489,
      "num_tokens": 963507113.0,
      "step": 30189
    },
    {
      "entropy": 0.5158021040260792,
      "epoch": 2.777637851421211,
      "grad_norm": 0.16295546293258667,
      "learning_rate": 7.415585610451743e-06,
      "loss": 0.5096,
      "mean_token_accuracy": 0.8369320333003998,
      "num_tokens": 963538301.0,
      "step": 30190
    },
    {
      "entropy": 0.4419727437198162,
      "epoch": 2.777729857792652,
      "grad_norm": 0.14854229986667633,
      "learning_rate": 7.412518784310118e-06,
      "loss": 0.4314,
      "mean_token_accuracy": 0.8630095459520817,
      "num_tokens": 963570313.0,
      "step": 30191
    },
    {
      "entropy": 0.4240782540291548,
      "epoch": 2.7778218641640935,
      "grad_norm": 0.14779651165008545,
      "learning_rate": 7.409451958168492e-06,
      "loss": 0.4068,
      "mean_token_accuracy": 0.8651255704462528,
      "num_tokens": 963602046.0,
      "step": 30192
    },
    {
      "entropy": 0.4923437489196658,
      "epoch": 2.7779138705355346,
      "grad_norm": 0.15356415510177612,
      "learning_rate": 7.406385132026866e-06,
      "loss": 0.4719,
      "mean_token_accuracy": 0.8518034070730209,
      "num_tokens": 963634139.0,
      "step": 30193
    },
    {
      "entropy": 0.5523361964151263,
      "epoch": 2.7780058769069758,
      "grad_norm": 0.16432617604732513,
      "learning_rate": 7.40331830588524e-06,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8259831294417381,
      "num_tokens": 963665574.0,
      "step": 30194
    },
    {
      "entropy": 0.49316915310919285,
      "epoch": 2.778097883278417,
      "grad_norm": 0.15447719395160675,
      "learning_rate": 7.400251479743614e-06,
      "loss": 0.476,
      "mean_token_accuracy": 0.8481652624905109,
      "num_tokens": 963697546.0,
      "step": 30195
    },
    {
      "entropy": 0.5209617149084806,
      "epoch": 2.778189889649858,
      "grad_norm": 0.16003194451332092,
      "learning_rate": 7.397184653601988e-06,
      "loss": 0.5027,
      "mean_token_accuracy": 0.843558955937624,
      "num_tokens": 963730234.0,
      "step": 30196
    },
    {
      "entropy": 0.4825946483761072,
      "epoch": 2.7782818960212996,
      "grad_norm": 0.15402637422084808,
      "learning_rate": 7.394117827460362e-06,
      "loss": 0.4683,
      "mean_token_accuracy": 0.8502178341150284,
      "num_tokens": 963761953.0,
      "step": 30197
    },
    {
      "entropy": 0.45032392628490925,
      "epoch": 2.7783739023927407,
      "grad_norm": 0.15132655203342438,
      "learning_rate": 7.391051001318736e-06,
      "loss": 0.4419,
      "mean_token_accuracy": 0.8580700010061264,
      "num_tokens": 963794210.0,
      "step": 30198
    },
    {
      "entropy": 0.4964128518477082,
      "epoch": 2.778465908764182,
      "grad_norm": 0.15600095689296722,
      "learning_rate": 7.38798417517711e-06,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8480581976473331,
      "num_tokens": 963826283.0,
      "step": 30199
    },
    {
      "entropy": 0.45562771055847406,
      "epoch": 2.778557915135623,
      "grad_norm": 0.153010755777359,
      "learning_rate": 7.3849173490354834e-06,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8576132990419865,
      "num_tokens": 963858376.0,
      "step": 30200
    },
    {
      "entropy": 0.465802135062404,
      "epoch": 2.778649921507064,
      "grad_norm": 0.14996182918548584,
      "learning_rate": 7.381850522893857e-06,
      "loss": 0.447,
      "mean_token_accuracy": 0.8610555082559586,
      "num_tokens": 963890590.0,
      "step": 30201
    },
    {
      "entropy": 0.5179488565772772,
      "epoch": 2.7787419278785057,
      "grad_norm": 0.152180016040802,
      "learning_rate": 7.378783696752231e-06,
      "loss": 0.4998,
      "mean_token_accuracy": 0.8401413969695568,
      "num_tokens": 963922903.0,
      "step": 30202
    },
    {
      "entropy": 0.5416790703311563,
      "epoch": 2.778833934249947,
      "grad_norm": 0.16073167324066162,
      "learning_rate": 7.375716870610605e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8320708051323891,
      "num_tokens": 963955414.0,
      "step": 30203
    },
    {
      "entropy": 0.49289532098919153,
      "epoch": 2.778925940621388,
      "grad_norm": 0.15420010685920715,
      "learning_rate": 7.372650044468979e-06,
      "loss": 0.4783,
      "mean_token_accuracy": 0.8495307043194771,
      "num_tokens": 963987450.0,
      "step": 30204
    },
    {
      "entropy": 0.4943456156179309,
      "epoch": 2.779017946992829,
      "grad_norm": 0.15993814170360565,
      "learning_rate": 7.3695832183273524e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8460084795951843,
      "num_tokens": 964018527.0,
      "step": 30205
    },
    {
      "entropy": 0.49539165012538433,
      "epoch": 2.7791099533642702,
      "grad_norm": 0.15868693590164185,
      "learning_rate": 7.366516392185728e-06,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8440982587635517,
      "num_tokens": 964050039.0,
      "step": 30206
    },
    {
      "entropy": 0.4475355315953493,
      "epoch": 2.779201959735712,
      "grad_norm": 0.15052904188632965,
      "learning_rate": 7.363449566044102e-06,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8591947630047798,
      "num_tokens": 964082284.0,
      "step": 30207
    },
    {
      "entropy": 0.5013934560120106,
      "epoch": 2.779293966107153,
      "grad_norm": 0.1533956229686737,
      "learning_rate": 7.3603827399024755e-06,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8458231985569,
      "num_tokens": 964114535.0,
      "step": 30208
    },
    {
      "entropy": 0.5395748885348439,
      "epoch": 2.779385972478594,
      "grad_norm": 0.1638399213552475,
      "learning_rate": 7.357315913760849e-06,
      "loss": 0.531,
      "mean_token_accuracy": 0.8327121287584305,
      "num_tokens": 964146928.0,
      "step": 30209
    },
    {
      "entropy": 0.6122716274112463,
      "epoch": 2.7794779788500352,
      "grad_norm": 0.17003817856311798,
      "learning_rate": 7.354249087619223e-06,
      "loss": 0.6063,
      "mean_token_accuracy": 0.8109495714306831,
      "num_tokens": 964178968.0,
      "step": 30210
    },
    {
      "entropy": 0.4864313034340739,
      "epoch": 2.7795699852214764,
      "grad_norm": 0.15327675640583038,
      "learning_rate": 7.351182261477598e-06,
      "loss": 0.4686,
      "mean_token_accuracy": 0.8483983092010021,
      "num_tokens": 964211260.0,
      "step": 30211
    },
    {
      "entropy": 0.44525654474273324,
      "epoch": 2.779661991592918,
      "grad_norm": 0.14571832120418549,
      "learning_rate": 7.348115435335972e-06,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8639616034924984,
      "num_tokens": 964243710.0,
      "step": 30212
    },
    {
      "entropy": 0.5085922749713063,
      "epoch": 2.779753997964359,
      "grad_norm": 0.1567225158214569,
      "learning_rate": 7.345048609194345e-06,
      "loss": 0.4954,
      "mean_token_accuracy": 0.8407198116183281,
      "num_tokens": 964275215.0,
      "step": 30213
    },
    {
      "entropy": 0.5151178520172834,
      "epoch": 2.7798460043358,
      "grad_norm": 0.16316024959087372,
      "learning_rate": 7.341981783052719e-06,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8391397595405579,
      "num_tokens": 964306633.0,
      "step": 30214
    },
    {
      "entropy": 0.45568674383684993,
      "epoch": 2.7799380107072413,
      "grad_norm": 0.14779268205165863,
      "learning_rate": 7.338914956911093e-06,
      "loss": 0.4516,
      "mean_token_accuracy": 0.8593921139836311,
      "num_tokens": 964339088.0,
      "step": 30215
    },
    {
      "entropy": 0.4099721135571599,
      "epoch": 2.7800300170786825,
      "grad_norm": 0.14640162885189056,
      "learning_rate": 7.335848130769467e-06,
      "loss": 0.3975,
      "mean_token_accuracy": 0.8714005649089813,
      "num_tokens": 964370812.0,
      "step": 30216
    },
    {
      "entropy": 0.5262605315074325,
      "epoch": 2.780122023450124,
      "grad_norm": 0.157334566116333,
      "learning_rate": 7.332781304627841e-06,
      "loss": 0.5142,
      "mean_token_accuracy": 0.8375514857470989,
      "num_tokens": 964402664.0,
      "step": 30217
    },
    {
      "entropy": 0.611907402984798,
      "epoch": 2.780214029821565,
      "grad_norm": 0.16840872168540955,
      "learning_rate": 7.329714478486214e-06,
      "loss": 0.5984,
      "mean_token_accuracy": 0.8159163556993008,
      "num_tokens": 964435135.0,
      "step": 30218
    },
    {
      "entropy": 0.5383710730820894,
      "epoch": 2.7803060361930063,
      "grad_norm": 0.16568809747695923,
      "learning_rate": 7.326647652344588e-06,
      "loss": 0.5294,
      "mean_token_accuracy": 0.8303337953984737,
      "num_tokens": 964466144.0,
      "step": 30219
    },
    {
      "entropy": 0.4715210283175111,
      "epoch": 2.7803980425644474,
      "grad_norm": 0.15024176239967346,
      "learning_rate": 7.323580826202962e-06,
      "loss": 0.4615,
      "mean_token_accuracy": 0.8519033268094063,
      "num_tokens": 964498798.0,
      "step": 30220
    },
    {
      "entropy": 0.45935837586876005,
      "epoch": 2.7804900489358886,
      "grad_norm": 0.1501939594745636,
      "learning_rate": 7.320514000061336e-06,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8565562777221203,
      "num_tokens": 964530833.0,
      "step": 30221
    },
    {
      "entropy": 0.5271190698258579,
      "epoch": 2.78058205530733,
      "grad_norm": 0.16582287847995758,
      "learning_rate": 7.317447173919711e-06,
      "loss": 0.5143,
      "mean_token_accuracy": 0.8341496139764786,
      "num_tokens": 964562536.0,
      "step": 30222
    },
    {
      "entropy": 0.5166603324469179,
      "epoch": 2.7806740616787713,
      "grad_norm": 0.1526193469762802,
      "learning_rate": 7.314380347778085e-06,
      "loss": 0.506,
      "mean_token_accuracy": 0.8392908908426762,
      "num_tokens": 964595078.0,
      "step": 30223
    },
    {
      "entropy": 0.5048483200371265,
      "epoch": 2.7807660680502124,
      "grad_norm": 0.15595589578151703,
      "learning_rate": 7.311313521636459e-06,
      "loss": 0.4815,
      "mean_token_accuracy": 0.846055407077074,
      "num_tokens": 964627339.0,
      "step": 30224
    },
    {
      "entropy": 0.6568488907068968,
      "epoch": 2.7808580744216536,
      "grad_norm": 0.1759415715932846,
      "learning_rate": 7.308246695494833e-06,
      "loss": 0.6404,
      "mean_token_accuracy": 0.798169918358326,
      "num_tokens": 964659294.0,
      "step": 30225
    },
    {
      "entropy": 0.4227371287997812,
      "epoch": 2.7809500807930947,
      "grad_norm": 0.15098223090171814,
      "learning_rate": 7.305179869353207e-06,
      "loss": 0.4143,
      "mean_token_accuracy": 0.8671567663550377,
      "num_tokens": 964691279.0,
      "step": 30226
    },
    {
      "entropy": 0.4894923307001591,
      "epoch": 2.7810420871645363,
      "grad_norm": 0.15657517313957214,
      "learning_rate": 7.302113043211581e-06,
      "loss": 0.4735,
      "mean_token_accuracy": 0.8498629406094551,
      "num_tokens": 964723594.0,
      "step": 30227
    },
    {
      "entropy": 0.5681836158037186,
      "epoch": 2.7811340935359774,
      "grad_norm": 0.1658194661140442,
      "learning_rate": 7.299046217069955e-06,
      "loss": 0.5533,
      "mean_token_accuracy": 0.8244877718389034,
      "num_tokens": 964755099.0,
      "step": 30228
    },
    {
      "entropy": 0.5263226144015789,
      "epoch": 2.7812260999074185,
      "grad_norm": 0.1646527796983719,
      "learning_rate": 7.295979390928329e-06,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8329965732991695,
      "num_tokens": 964787501.0,
      "step": 30229
    },
    {
      "entropy": 0.4485214948654175,
      "epoch": 2.7813181062788597,
      "grad_norm": 0.14598335325717926,
      "learning_rate": 7.2929125647867025e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8610577471554279,
      "num_tokens": 964819630.0,
      "step": 30230
    },
    {
      "entropy": 0.49204091913998127,
      "epoch": 2.781410112650301,
      "grad_norm": 0.15743084251880646,
      "learning_rate": 7.289845738645076e-06,
      "loss": 0.478,
      "mean_token_accuracy": 0.8465085476636887,
      "num_tokens": 964851632.0,
      "step": 30231
    },
    {
      "entropy": 0.5761807914823294,
      "epoch": 2.7815021190217424,
      "grad_norm": 0.16958054900169373,
      "learning_rate": 7.28677891250345e-06,
      "loss": 0.5713,
      "mean_token_accuracy": 0.8201095387339592,
      "num_tokens": 964883760.0,
      "step": 30232
    },
    {
      "entropy": 0.4558483371511102,
      "epoch": 2.7815941253931835,
      "grad_norm": 0.1491902470588684,
      "learning_rate": 7.283712086361824e-06,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8607774041593075,
      "num_tokens": 964915477.0,
      "step": 30233
    },
    {
      "entropy": 0.5514753963798285,
      "epoch": 2.7816861317646246,
      "grad_norm": 0.16228321194648743,
      "learning_rate": 7.280645260220198e-06,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8331453278660774,
      "num_tokens": 964947763.0,
      "step": 30234
    },
    {
      "entropy": 0.42571119125932455,
      "epoch": 2.781778138136066,
      "grad_norm": 0.1449476182460785,
      "learning_rate": 7.2775784340785715e-06,
      "loss": 0.4039,
      "mean_token_accuracy": 0.8680808916687965,
      "num_tokens": 964979631.0,
      "step": 30235
    },
    {
      "entropy": 0.43936879374086857,
      "epoch": 2.781870144507507,
      "grad_norm": 0.15097101032733917,
      "learning_rate": 7.274511607936945e-06,
      "loss": 0.4324,
      "mean_token_accuracy": 0.864876601845026,
      "num_tokens": 965010997.0,
      "step": 30236
    },
    {
      "entropy": 0.5466020051389933,
      "epoch": 2.7819621508789485,
      "grad_norm": 0.16216735541820526,
      "learning_rate": 7.271444781795321e-06,
      "loss": 0.5213,
      "mean_token_accuracy": 0.8341222368180752,
      "num_tokens": 965042312.0,
      "step": 30237
    },
    {
      "entropy": 0.500641911290586,
      "epoch": 2.7820541572503896,
      "grad_norm": 0.15934781730175018,
      "learning_rate": 7.268377955653695e-06,
      "loss": 0.486,
      "mean_token_accuracy": 0.8436008878052235,
      "num_tokens": 965073740.0,
      "step": 30238
    },
    {
      "entropy": 0.5412944238632917,
      "epoch": 2.7821461636218308,
      "grad_norm": 0.16222301125526428,
      "learning_rate": 7.2653111295120684e-06,
      "loss": 0.5336,
      "mean_token_accuracy": 0.8300242610275745,
      "num_tokens": 965106122.0,
      "step": 30239
    },
    {
      "entropy": 0.4922706625657156,
      "epoch": 2.782238169993272,
      "grad_norm": 0.15368923544883728,
      "learning_rate": 7.262244303370442e-06,
      "loss": 0.4862,
      "mean_token_accuracy": 0.8446784131228924,
      "num_tokens": 965138671.0,
      "step": 30240
    },
    {
      "entropy": 0.5046289414167404,
      "epoch": 2.782330176364713,
      "grad_norm": 0.15900491178035736,
      "learning_rate": 7.259177477228817e-06,
      "loss": 0.4905,
      "mean_token_accuracy": 0.8407155871391296,
      "num_tokens": 965170496.0,
      "step": 30241
    },
    {
      "entropy": 0.553275722078979,
      "epoch": 2.7824221827361546,
      "grad_norm": 0.17009155452251434,
      "learning_rate": 7.256110651087191e-06,
      "loss": 0.5515,
      "mean_token_accuracy": 0.824972927570343,
      "num_tokens": 965202877.0,
      "step": 30242
    },
    {
      "entropy": 0.5231585232540965,
      "epoch": 2.7825141891075957,
      "grad_norm": 0.1599549800157547,
      "learning_rate": 7.2530438249455645e-06,
      "loss": 0.5123,
      "mean_token_accuracy": 0.8377735503017902,
      "num_tokens": 965234170.0,
      "step": 30243
    },
    {
      "entropy": 0.5024600913748145,
      "epoch": 2.782606195479037,
      "grad_norm": 0.1582275778055191,
      "learning_rate": 7.249976998803938e-06,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8443539217114449,
      "num_tokens": 965265710.0,
      "step": 30244
    },
    {
      "entropy": 0.542324785143137,
      "epoch": 2.782698201850478,
      "grad_norm": 0.16042287647724152,
      "learning_rate": 7.246910172662312e-06,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8336152583360672,
      "num_tokens": 965297359.0,
      "step": 30245
    },
    {
      "entropy": 0.598588889464736,
      "epoch": 2.782790208221919,
      "grad_norm": 0.16633491218090057,
      "learning_rate": 7.243843346520686e-06,
      "loss": 0.5844,
      "mean_token_accuracy": 0.8133337274193764,
      "num_tokens": 965329135.0,
      "step": 30246
    },
    {
      "entropy": 0.5390091594308615,
      "epoch": 2.7828822145933607,
      "grad_norm": 0.16140398383140564,
      "learning_rate": 7.24077652037906e-06,
      "loss": 0.5225,
      "mean_token_accuracy": 0.8330464288592339,
      "num_tokens": 965360972.0,
      "step": 30247
    },
    {
      "entropy": 0.5590539360418916,
      "epoch": 2.782974220964802,
      "grad_norm": 0.15989433228969574,
      "learning_rate": 7.2377096942374335e-06,
      "loss": 0.5378,
      "mean_token_accuracy": 0.828291542828083,
      "num_tokens": 965393236.0,
      "step": 30248
    },
    {
      "entropy": 0.512221448123455,
      "epoch": 2.783066227336243,
      "grad_norm": 0.16235677897930145,
      "learning_rate": 7.234642868095807e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.837503869086504,
      "num_tokens": 965424720.0,
      "step": 30249
    },
    {
      "entropy": 0.5263779917731881,
      "epoch": 2.783158233707684,
      "grad_norm": 0.16317065060138702,
      "learning_rate": 7.231576041954181e-06,
      "loss": 0.5217,
      "mean_token_accuracy": 0.8354438580572605,
      "num_tokens": 965456335.0,
      "step": 30250
    },
    {
      "entropy": 0.4172086585313082,
      "epoch": 2.7832502400791252,
      "grad_norm": 0.14185094833374023,
      "learning_rate": 7.228509215812555e-06,
      "loss": 0.3933,
      "mean_token_accuracy": 0.8704573288559914,
      "num_tokens": 965487929.0,
      "step": 30251
    },
    {
      "entropy": 0.5241026007570326,
      "epoch": 2.783342246450567,
      "grad_norm": 0.15860992670059204,
      "learning_rate": 7.22544238967093e-06,
      "loss": 0.5072,
      "mean_token_accuracy": 0.8384174257516861,
      "num_tokens": 965520295.0,
      "step": 30252
    },
    {
      "entropy": 0.46265684999525547,
      "epoch": 2.783434252822008,
      "grad_norm": 0.15270346403121948,
      "learning_rate": 7.222375563529304e-06,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8519965969026089,
      "num_tokens": 965552967.0,
      "step": 30253
    },
    {
      "entropy": 0.46915148198604584,
      "epoch": 2.783526259193449,
      "grad_norm": 0.15144194662570953,
      "learning_rate": 7.219308737387678e-06,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8569889143109322,
      "num_tokens": 965584935.0,
      "step": 30254
    },
    {
      "entropy": 0.5407870803028345,
      "epoch": 2.7836182655648902,
      "grad_norm": 0.1631370633840561,
      "learning_rate": 7.216241911246053e-06,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8316774964332581,
      "num_tokens": 965616791.0,
      "step": 30255
    },
    {
      "entropy": 0.45625869650393724,
      "epoch": 2.7837102719363314,
      "grad_norm": 0.1513875275850296,
      "learning_rate": 7.213175085104426e-06,
      "loss": 0.4455,
      "mean_token_accuracy": 0.8583700060844421,
      "num_tokens": 965649028.0,
      "step": 30256
    },
    {
      "entropy": 0.49979985412210226,
      "epoch": 2.783802278307773,
      "grad_norm": 0.15705819427967072,
      "learning_rate": 7.2101082589628e-06,
      "loss": 0.4789,
      "mean_token_accuracy": 0.8491889983415604,
      "num_tokens": 965680603.0,
      "step": 30257
    },
    {
      "entropy": 0.5119588486850262,
      "epoch": 2.783894284679214,
      "grad_norm": 0.15476655960083008,
      "learning_rate": 7.207041432821174e-06,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8424711301922798,
      "num_tokens": 965712731.0,
      "step": 30258
    },
    {
      "entropy": 0.5298051137942821,
      "epoch": 2.783986291050655,
      "grad_norm": 0.16469040513038635,
      "learning_rate": 7.203974606679548e-06,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8366053812205791,
      "num_tokens": 965744724.0,
      "step": 30259
    },
    {
      "entropy": 0.503009500913322,
      "epoch": 2.7840782974220963,
      "grad_norm": 0.15950484573841095,
      "learning_rate": 7.200907780537922e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8394712135195732,
      "num_tokens": 965776669.0,
      "step": 30260
    },
    {
      "entropy": 0.5021192748099566,
      "epoch": 2.7841703037935375,
      "grad_norm": 0.16066166758537292,
      "learning_rate": 7.1978409543962954e-06,
      "loss": 0.4998,
      "mean_token_accuracy": 0.8437079638242722,
      "num_tokens": 965809285.0,
      "step": 30261
    },
    {
      "entropy": 0.5317448070272803,
      "epoch": 2.784262310164979,
      "grad_norm": 0.16345150768756866,
      "learning_rate": 7.194774128254669e-06,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8347119502723217,
      "num_tokens": 965840636.0,
      "step": 30262
    },
    {
      "entropy": 0.5141883622854948,
      "epoch": 2.78435431653642,
      "grad_norm": 0.16144680976867676,
      "learning_rate": 7.191707302113043e-06,
      "loss": 0.5097,
      "mean_token_accuracy": 0.8411217890679836,
      "num_tokens": 965872635.0,
      "step": 30263
    },
    {
      "entropy": 0.5559287164360285,
      "epoch": 2.7844463229078613,
      "grad_norm": 0.1665806770324707,
      "learning_rate": 7.188640475971417e-06,
      "loss": 0.5479,
      "mean_token_accuracy": 0.8280952833592892,
      "num_tokens": 965903445.0,
      "step": 30264
    },
    {
      "entropy": 0.49351329635828733,
      "epoch": 2.7845383292793024,
      "grad_norm": 0.154694065451622,
      "learning_rate": 7.185573649829791e-06,
      "loss": 0.4765,
      "mean_token_accuracy": 0.8508049361407757,
      "num_tokens": 965935011.0,
      "step": 30265
    },
    {
      "entropy": 0.5328414086252451,
      "epoch": 2.7846303356507436,
      "grad_norm": 0.16077297925949097,
      "learning_rate": 7.182506823688165e-06,
      "loss": 0.5291,
      "mean_token_accuracy": 0.8348809741437435,
      "num_tokens": 965967484.0,
      "step": 30266
    },
    {
      "entropy": 0.5729996245354414,
      "epoch": 2.784722342022185,
      "grad_norm": 0.16843487322330475,
      "learning_rate": 7.17943999754654e-06,
      "loss": 0.5577,
      "mean_token_accuracy": 0.8239931091666222,
      "num_tokens": 965999650.0,
      "step": 30267
    },
    {
      "entropy": 0.5827785306610167,
      "epoch": 2.7848143483936263,
      "grad_norm": 0.17420215904712677,
      "learning_rate": 7.176373171404914e-06,
      "loss": 0.5764,
      "mean_token_accuracy": 0.8183167465031147,
      "num_tokens": 966030522.0,
      "step": 30268
    },
    {
      "entropy": 0.5178136005997658,
      "epoch": 2.7849063547650674,
      "grad_norm": 0.16450895369052887,
      "learning_rate": 7.1733063452632875e-06,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8355848528444767,
      "num_tokens": 966061204.0,
      "step": 30269
    },
    {
      "entropy": 0.45958715211600065,
      "epoch": 2.7849983611365086,
      "grad_norm": 0.1539660394191742,
      "learning_rate": 7.170239519121662e-06,
      "loss": 0.4352,
      "mean_token_accuracy": 0.8602490611374378,
      "num_tokens": 966092262.0,
      "step": 30270
    },
    {
      "entropy": 0.5310485903173685,
      "epoch": 2.7850903675079497,
      "grad_norm": 0.16152919828891754,
      "learning_rate": 7.167172692980036e-06,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8392962366342545,
      "num_tokens": 966123823.0,
      "step": 30271
    },
    {
      "entropy": 0.5636040782555938,
      "epoch": 2.7851823738793913,
      "grad_norm": 0.16419200599193573,
      "learning_rate": 7.16410586683841e-06,
      "loss": 0.5467,
      "mean_token_accuracy": 0.826422818005085,
      "num_tokens": 966156591.0,
      "step": 30272
    },
    {
      "entropy": 0.5103673273697495,
      "epoch": 2.7852743802508324,
      "grad_norm": 0.1623699814081192,
      "learning_rate": 7.1610390406967836e-06,
      "loss": 0.5077,
      "mean_token_accuracy": 0.8384028375148773,
      "num_tokens": 966188670.0,
      "step": 30273
    },
    {
      "entropy": 0.47381957806646824,
      "epoch": 2.7853663866222735,
      "grad_norm": 0.15250609815120697,
      "learning_rate": 7.157972214555157e-06,
      "loss": 0.4628,
      "mean_token_accuracy": 0.8541064970195293,
      "num_tokens": 966220643.0,
      "step": 30274
    },
    {
      "entropy": 0.5097229750826955,
      "epoch": 2.7854583929937147,
      "grad_norm": 0.15880899131298065,
      "learning_rate": 7.154905388413531e-06,
      "loss": 0.494,
      "mean_token_accuracy": 0.8464817702770233,
      "num_tokens": 966252663.0,
      "step": 30275
    },
    {
      "entropy": 0.4841288619209081,
      "epoch": 2.785550399365156,
      "grad_norm": 0.1561661958694458,
      "learning_rate": 7.151838562271905e-06,
      "loss": 0.4761,
      "mean_token_accuracy": 0.8450148440897465,
      "num_tokens": 966284674.0,
      "step": 30276
    },
    {
      "entropy": 0.6129122413694859,
      "epoch": 2.7856424057365974,
      "grad_norm": 0.17195503413677216,
      "learning_rate": 7.148771736130279e-06,
      "loss": 0.6052,
      "mean_token_accuracy": 0.8100777938961983,
      "num_tokens": 966316102.0,
      "step": 30277
    },
    {
      "entropy": 0.5886732190847397,
      "epoch": 2.7857344121080385,
      "grad_norm": 0.17241120338439941,
      "learning_rate": 7.1457049099886526e-06,
      "loss": 0.5749,
      "mean_token_accuracy": 0.8187889121472836,
      "num_tokens": 966347010.0,
      "step": 30278
    },
    {
      "entropy": 0.533258717507124,
      "epoch": 2.7858264184794796,
      "grad_norm": 0.16373614966869354,
      "learning_rate": 7.142638083847026e-06,
      "loss": 0.5291,
      "mean_token_accuracy": 0.8342349119484425,
      "num_tokens": 966379491.0,
      "step": 30279
    },
    {
      "entropy": 0.5570449670776725,
      "epoch": 2.785918424850921,
      "grad_norm": 0.17085832357406616,
      "learning_rate": 7.1395712577054e-06,
      "loss": 0.5601,
      "mean_token_accuracy": 0.8237059898674488,
      "num_tokens": 966410710.0,
      "step": 30280
    },
    {
      "entropy": 0.5350626558065414,
      "epoch": 2.786010431222362,
      "grad_norm": 0.16022086143493652,
      "learning_rate": 7.136504431563775e-06,
      "loss": 0.5165,
      "mean_token_accuracy": 0.834605511277914,
      "num_tokens": 966443478.0,
      "step": 30281
    },
    {
      "entropy": 0.5329596875235438,
      "epoch": 2.7861024375938035,
      "grad_norm": 0.15945065021514893,
      "learning_rate": 7.133437605422149e-06,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8346828892827034,
      "num_tokens": 966475064.0,
      "step": 30282
    },
    {
      "entropy": 0.38815377675928175,
      "epoch": 2.7861944439652446,
      "grad_norm": 0.1462884545326233,
      "learning_rate": 7.130370779280523e-06,
      "loss": 0.3792,
      "mean_token_accuracy": 0.8773762583732605,
      "num_tokens": 966506930.0,
      "step": 30283
    },
    {
      "entropy": 0.47672743164002895,
      "epoch": 2.7862864503366858,
      "grad_norm": 0.1502138376235962,
      "learning_rate": 7.127303953138897e-06,
      "loss": 0.4669,
      "mean_token_accuracy": 0.8525011315941811,
      "num_tokens": 966539073.0,
      "step": 30284
    },
    {
      "entropy": 0.45923598296940327,
      "epoch": 2.786378456708127,
      "grad_norm": 0.15299546718597412,
      "learning_rate": 7.124237126997272e-06,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8523550294339657,
      "num_tokens": 966571325.0,
      "step": 30285
    },
    {
      "entropy": 0.5298438519239426,
      "epoch": 2.786470463079568,
      "grad_norm": 0.16314175724983215,
      "learning_rate": 7.1211703008556455e-06,
      "loss": 0.5126,
      "mean_token_accuracy": 0.8342049606144428,
      "num_tokens": 966602396.0,
      "step": 30286
    },
    {
      "entropy": 0.42983323708176613,
      "epoch": 2.7865624694510096,
      "grad_norm": 0.14308632910251617,
      "learning_rate": 7.118103474714019e-06,
      "loss": 0.4165,
      "mean_token_accuracy": 0.8663491234183311,
      "num_tokens": 966634621.0,
      "step": 30287
    },
    {
      "entropy": 0.5161064621061087,
      "epoch": 2.7866544758224507,
      "grad_norm": 0.1625148206949234,
      "learning_rate": 7.115036648572393e-06,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8335371911525726,
      "num_tokens": 966666385.0,
      "step": 30288
    },
    {
      "entropy": 0.4067091625183821,
      "epoch": 2.786746482193892,
      "grad_norm": 0.14439593255519867,
      "learning_rate": 7.111969822430767e-06,
      "loss": 0.3957,
      "mean_token_accuracy": 0.8735237270593643,
      "num_tokens": 966697962.0,
      "step": 30289
    },
    {
      "entropy": 0.5686957787256688,
      "epoch": 2.786838488565333,
      "grad_norm": 0.16312240064144135,
      "learning_rate": 7.108902996289141e-06,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8298355117440224,
      "num_tokens": 966729362.0,
      "step": 30290
    },
    {
      "entropy": 0.5196643006056547,
      "epoch": 2.786930494936774,
      "grad_norm": 0.15932367742061615,
      "learning_rate": 7.1058361701475145e-06,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8403480015695095,
      "num_tokens": 966760689.0,
      "step": 30291
    },
    {
      "entropy": 0.48436892591416836,
      "epoch": 2.7870225013082157,
      "grad_norm": 0.15615706145763397,
      "learning_rate": 7.102769344005888e-06,
      "loss": 0.4783,
      "mean_token_accuracy": 0.8461810573935509,
      "num_tokens": 966793337.0,
      "step": 30292
    },
    {
      "entropy": 0.46222922019660473,
      "epoch": 2.787114507679657,
      "grad_norm": 0.15333063900470734,
      "learning_rate": 7.099702517864262e-06,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8555613681674004,
      "num_tokens": 966825167.0,
      "step": 30293
    },
    {
      "entropy": 0.5653870906680822,
      "epoch": 2.787206514051098,
      "grad_norm": 0.1650901436805725,
      "learning_rate": 7.096635691722636e-06,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8228617459535599,
      "num_tokens": 966856646.0,
      "step": 30294
    },
    {
      "entropy": 0.3780062347650528,
      "epoch": 2.787298520422539,
      "grad_norm": 0.14050297439098358,
      "learning_rate": 7.09356886558101e-06,
      "loss": 0.3589,
      "mean_token_accuracy": 0.8825604096055031,
      "num_tokens": 966887898.0,
      "step": 30295
    },
    {
      "entropy": 0.41827560123056173,
      "epoch": 2.7873905267939802,
      "grad_norm": 0.1463109254837036,
      "learning_rate": 7.090502039439384e-06,
      "loss": 0.4068,
      "mean_token_accuracy": 0.8722957819700241,
      "num_tokens": 966919365.0,
      "step": 30296
    },
    {
      "entropy": 0.45035182125866413,
      "epoch": 2.787482533165422,
      "grad_norm": 0.1496853530406952,
      "learning_rate": 7.087435213297758e-06,
      "loss": 0.4468,
      "mean_token_accuracy": 0.8543733768165112,
      "num_tokens": 966951849.0,
      "step": 30297
    },
    {
      "entropy": 0.5756077873520553,
      "epoch": 2.787574539536863,
      "grad_norm": 0.1688118875026703,
      "learning_rate": 7.084368387156133e-06,
      "loss": 0.5709,
      "mean_token_accuracy": 0.8213941566646099,
      "num_tokens": 966984312.0,
      "step": 30298
    },
    {
      "entropy": 0.4935845392756164,
      "epoch": 2.787666545908304,
      "grad_norm": 0.14866232872009277,
      "learning_rate": 7.081301561014507e-06,
      "loss": 0.4741,
      "mean_token_accuracy": 0.8493607901036739,
      "num_tokens": 967016517.0,
      "step": 30299
    },
    {
      "entropy": 0.491499213501811,
      "epoch": 2.7877585522797452,
      "grad_norm": 0.15752016007900238,
      "learning_rate": 7.078234734872881e-06,
      "loss": 0.4805,
      "mean_token_accuracy": 0.844586681574583,
      "num_tokens": 967048961.0,
      "step": 30300
    },
    {
      "entropy": 0.615341316908598,
      "epoch": 2.7878505586511864,
      "grad_norm": 0.17295821011066437,
      "learning_rate": 7.075167908731255e-06,
      "loss": 0.6002,
      "mean_token_accuracy": 0.8111138269305229,
      "num_tokens": 967080349.0,
      "step": 30301
    },
    {
      "entropy": 0.510491332039237,
      "epoch": 2.787942565022628,
      "grad_norm": 0.16156497597694397,
      "learning_rate": 7.072101082589629e-06,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8401804305613041,
      "num_tokens": 967111521.0,
      "step": 30302
    },
    {
      "entropy": 0.5212997838389128,
      "epoch": 2.788034571394069,
      "grad_norm": 0.16217923164367676,
      "learning_rate": 7.069034256448003e-06,
      "loss": 0.507,
      "mean_token_accuracy": 0.8396548219025135,
      "num_tokens": 967142677.0,
      "step": 30303
    },
    {
      "entropy": 0.6143985623493791,
      "epoch": 2.78812657776551,
      "grad_norm": 0.16584938764572144,
      "learning_rate": 7.0659674303063765e-06,
      "loss": 0.6001,
      "mean_token_accuracy": 0.8093461692333221,
      "num_tokens": 967175032.0,
      "step": 30304
    },
    {
      "entropy": 0.4788594446144998,
      "epoch": 2.7882185841369513,
      "grad_norm": 0.1507328599691391,
      "learning_rate": 7.06290060416475e-06,
      "loss": 0.4725,
      "mean_token_accuracy": 0.8497954644262791,
      "num_tokens": 967207467.0,
      "step": 30305
    },
    {
      "entropy": 0.465347183868289,
      "epoch": 2.7883105905083925,
      "grad_norm": 0.14736071228981018,
      "learning_rate": 7.059833778023124e-06,
      "loss": 0.4514,
      "mean_token_accuracy": 0.8547915183007717,
      "num_tokens": 967239408.0,
      "step": 30306
    },
    {
      "entropy": 0.4507964998483658,
      "epoch": 2.788402596879834,
      "grad_norm": 0.1501973420381546,
      "learning_rate": 7.056766951881498e-06,
      "loss": 0.4276,
      "mean_token_accuracy": 0.8620311580598354,
      "num_tokens": 967271242.0,
      "step": 30307
    },
    {
      "entropy": 0.5575670525431633,
      "epoch": 2.788494603251275,
      "grad_norm": 0.16592587530612946,
      "learning_rate": 7.053700125739872e-06,
      "loss": 0.549,
      "mean_token_accuracy": 0.8294631913304329,
      "num_tokens": 967303708.0,
      "step": 30308
    },
    {
      "entropy": 0.4218650236725807,
      "epoch": 2.7885866096227163,
      "grad_norm": 0.15157130360603333,
      "learning_rate": 7.0506332995982455e-06,
      "loss": 0.4202,
      "mean_token_accuracy": 0.8633149415254593,
      "num_tokens": 967335882.0,
      "step": 30309
    },
    {
      "entropy": 0.5076815187931061,
      "epoch": 2.7886786159941575,
      "grad_norm": 0.1557626575231552,
      "learning_rate": 7.047566473456619e-06,
      "loss": 0.5055,
      "mean_token_accuracy": 0.8385738655924797,
      "num_tokens": 967367780.0,
      "step": 30310
    },
    {
      "entropy": 0.5268616951070726,
      "epoch": 2.7887706223655986,
      "grad_norm": 0.1680838167667389,
      "learning_rate": 7.044499647314994e-06,
      "loss": 0.522,
      "mean_token_accuracy": 0.8352295681834221,
      "num_tokens": 967398794.0,
      "step": 30311
    },
    {
      "entropy": 0.5379558000713587,
      "epoch": 2.78886262873704,
      "grad_norm": 0.16563692688941956,
      "learning_rate": 7.041432821173368e-06,
      "loss": 0.5285,
      "mean_token_accuracy": 0.834385734051466,
      "num_tokens": 967430563.0,
      "step": 30312
    },
    {
      "entropy": 0.5237719425931573,
      "epoch": 2.7889546351084813,
      "grad_norm": 0.16403980553150177,
      "learning_rate": 7.038365995031742e-06,
      "loss": 0.5083,
      "mean_token_accuracy": 0.8368545733392239,
      "num_tokens": 967462326.0,
      "step": 30313
    },
    {
      "entropy": 0.5562916975468397,
      "epoch": 2.7890466414799224,
      "grad_norm": 0.16821689903736115,
      "learning_rate": 7.035299168890116e-06,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8273765854537487,
      "num_tokens": 967493566.0,
      "step": 30314
    },
    {
      "entropy": 0.579927358776331,
      "epoch": 2.7891386478513636,
      "grad_norm": 0.16547125577926636,
      "learning_rate": 7.032232342748491e-06,
      "loss": 0.5689,
      "mean_token_accuracy": 0.8174983039498329,
      "num_tokens": 967525086.0,
      "step": 30315
    },
    {
      "entropy": 0.4579476658254862,
      "epoch": 2.7892306542228047,
      "grad_norm": 0.15775203704833984,
      "learning_rate": 7.029165516606865e-06,
      "loss": 0.4403,
      "mean_token_accuracy": 0.8614107482135296,
      "num_tokens": 967556840.0,
      "step": 30316
    },
    {
      "entropy": 0.5079697398468852,
      "epoch": 2.7893226605942463,
      "grad_norm": 0.1660676896572113,
      "learning_rate": 7.026098690465238e-06,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8424001149833202,
      "num_tokens": 967587918.0,
      "step": 30317
    },
    {
      "entropy": 0.49896018020808697,
      "epoch": 2.7894146669656874,
      "grad_norm": 0.16246213018894196,
      "learning_rate": 7.023031864323612e-06,
      "loss": 0.4936,
      "mean_token_accuracy": 0.8421701490879059,
      "num_tokens": 967619872.0,
      "step": 30318
    },
    {
      "entropy": 0.5956637170165777,
      "epoch": 2.7895066733371285,
      "grad_norm": 0.17724211513996124,
      "learning_rate": 7.019965038181986e-06,
      "loss": 0.5744,
      "mean_token_accuracy": 0.8146865107119083,
      "num_tokens": 967650615.0,
      "step": 30319
    },
    {
      "entropy": 0.4555215183645487,
      "epoch": 2.7895986797085697,
      "grad_norm": 0.1572321653366089,
      "learning_rate": 7.01689821204036e-06,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8525288850069046,
      "num_tokens": 967682415.0,
      "step": 30320
    },
    {
      "entropy": 0.5582897309213877,
      "epoch": 2.789690686080011,
      "grad_norm": 0.16098687052726746,
      "learning_rate": 7.013831385898734e-06,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8281438797712326,
      "num_tokens": 967715126.0,
      "step": 30321
    },
    {
      "entropy": 0.40480585931800306,
      "epoch": 2.7897826924514524,
      "grad_norm": 0.14267472922801971,
      "learning_rate": 7.010764559757107e-06,
      "loss": 0.3898,
      "mean_token_accuracy": 0.8733921907842159,
      "num_tokens": 967747195.0,
      "step": 30322
    },
    {
      "entropy": 0.43898750096559525,
      "epoch": 2.7898746988228935,
      "grad_norm": 0.14689897000789642,
      "learning_rate": 7.007697733615481e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8622898831963539,
      "num_tokens": 967779271.0,
      "step": 30323
    },
    {
      "entropy": 0.5133932465687394,
      "epoch": 2.7899667051943347,
      "grad_norm": 0.15855391323566437,
      "learning_rate": 7.004630907473855e-06,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8428266830742359,
      "num_tokens": 967811085.0,
      "step": 30324
    },
    {
      "entropy": 0.5005833636969328,
      "epoch": 2.790058711565776,
      "grad_norm": 0.16321426630020142,
      "learning_rate": 7.001564081332229e-06,
      "loss": 0.4906,
      "mean_token_accuracy": 0.8448703922331333,
      "num_tokens": 967842251.0,
      "step": 30325
    },
    {
      "entropy": 0.5009878054261208,
      "epoch": 2.790150717937217,
      "grad_norm": 0.15157580375671387,
      "learning_rate": 6.9984972551906035e-06,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8460533805191517,
      "num_tokens": 967874372.0,
      "step": 30326
    },
    {
      "entropy": 0.5771335856989026,
      "epoch": 2.7902427243086585,
      "grad_norm": 0.1678728461265564,
      "learning_rate": 6.995430429048977e-06,
      "loss": 0.5737,
      "mean_token_accuracy": 0.8199764154851437,
      "num_tokens": 967906495.0,
      "step": 30327
    },
    {
      "entropy": 0.47447878774255514,
      "epoch": 2.7903347306800996,
      "grad_norm": 0.1504228115081787,
      "learning_rate": 6.992363602907352e-06,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8559262529015541,
      "num_tokens": 967937578.0,
      "step": 30328
    },
    {
      "entropy": 0.45528230955824256,
      "epoch": 2.7904267370515408,
      "grad_norm": 0.1483854204416275,
      "learning_rate": 6.989296776765726e-06,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8608361259102821,
      "num_tokens": 967969933.0,
      "step": 30329
    },
    {
      "entropy": 0.5498452009633183,
      "epoch": 2.790518743422982,
      "grad_norm": 0.16105566918849945,
      "learning_rate": 6.9862299506241e-06,
      "loss": 0.5298,
      "mean_token_accuracy": 0.8320799544453621,
      "num_tokens": 968001172.0,
      "step": 30330
    },
    {
      "entropy": 0.5079976366832852,
      "epoch": 2.790610749794423,
      "grad_norm": 0.16417087614536285,
      "learning_rate": 6.983163124482474e-06,
      "loss": 0.4901,
      "mean_token_accuracy": 0.8446247577667236,
      "num_tokens": 968032778.0,
      "step": 30331
    },
    {
      "entropy": 0.47192604932934046,
      "epoch": 2.7907027561658646,
      "grad_norm": 0.15204721689224243,
      "learning_rate": 6.980096298340848e-06,
      "loss": 0.4588,
      "mean_token_accuracy": 0.8524315506219864,
      "num_tokens": 968065023.0,
      "step": 30332
    },
    {
      "entropy": 0.5982149131596088,
      "epoch": 2.7907947625373057,
      "grad_norm": 0.16502586007118225,
      "learning_rate": 6.977029472199222e-06,
      "loss": 0.5841,
      "mean_token_accuracy": 0.8167938888072968,
      "num_tokens": 968097126.0,
      "step": 30333
    },
    {
      "entropy": 0.5471823262050748,
      "epoch": 2.790886768908747,
      "grad_norm": 0.1624172329902649,
      "learning_rate": 6.9739626460575956e-06,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8325113728642464,
      "num_tokens": 968128308.0,
      "step": 30334
    },
    {
      "entropy": 0.5833583064377308,
      "epoch": 2.790978775280188,
      "grad_norm": 0.1664908230304718,
      "learning_rate": 6.970895819915969e-06,
      "loss": 0.574,
      "mean_token_accuracy": 0.8190548904240131,
      "num_tokens": 968160586.0,
      "step": 30335
    },
    {
      "entropy": 0.507770768366754,
      "epoch": 2.791070781651629,
      "grad_norm": 0.15012748539447784,
      "learning_rate": 6.967828993774343e-06,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8443861715495586,
      "num_tokens": 968192657.0,
      "step": 30336
    },
    {
      "entropy": 0.4388193088816479,
      "epoch": 2.7911627880230707,
      "grad_norm": 0.15123145282268524,
      "learning_rate": 6.964762167632717e-06,
      "loss": 0.4195,
      "mean_token_accuracy": 0.864846833050251,
      "num_tokens": 968224563.0,
      "step": 30337
    },
    {
      "entropy": 0.5338312461972237,
      "epoch": 2.791254794394512,
      "grad_norm": 0.15812712907791138,
      "learning_rate": 6.961695341491091e-06,
      "loss": 0.524,
      "mean_token_accuracy": 0.833313349634409,
      "num_tokens": 968257273.0,
      "step": 30338
    },
    {
      "entropy": 0.5330096771940589,
      "epoch": 2.791346800765953,
      "grad_norm": 0.16310472786426544,
      "learning_rate": 6.9586285153494646e-06,
      "loss": 0.5236,
      "mean_token_accuracy": 0.8350229971110821,
      "num_tokens": 968289145.0,
      "step": 30339
    },
    {
      "entropy": 0.4940840685740113,
      "epoch": 2.791438807137394,
      "grad_norm": 0.1572066992521286,
      "learning_rate": 6.955561689207838e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8459118381142616,
      "num_tokens": 968321242.0,
      "step": 30340
    },
    {
      "entropy": 0.4350016973912716,
      "epoch": 2.7915308135088353,
      "grad_norm": 0.14814776182174683,
      "learning_rate": 6.952494863066213e-06,
      "loss": 0.4249,
      "mean_token_accuracy": 0.8625238016247749,
      "num_tokens": 968352972.0,
      "step": 30341
    },
    {
      "entropy": 0.5113232145085931,
      "epoch": 2.791622819880277,
      "grad_norm": 0.15832068026065826,
      "learning_rate": 6.949428036924587e-06,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8428733050823212,
      "num_tokens": 968384496.0,
      "step": 30342
    },
    {
      "entropy": 0.45792324701324105,
      "epoch": 2.791714826251718,
      "grad_norm": 0.1546132117509842,
      "learning_rate": 6.946361210782961e-06,
      "loss": 0.4476,
      "mean_token_accuracy": 0.855613861232996,
      "num_tokens": 968415557.0,
      "step": 30343
    },
    {
      "entropy": 0.5281066475436091,
      "epoch": 2.791806832623159,
      "grad_norm": 0.16735102236270905,
      "learning_rate": 6.943294384641335e-06,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8390797227621078,
      "num_tokens": 968446192.0,
      "step": 30344
    },
    {
      "entropy": 0.5183103801682591,
      "epoch": 2.7918988389946002,
      "grad_norm": 0.1598062813282013,
      "learning_rate": 6.94022755849971e-06,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8391121551394463,
      "num_tokens": 968477957.0,
      "step": 30345
    },
    {
      "entropy": 0.4327929480932653,
      "epoch": 2.7919908453660414,
      "grad_norm": 0.1478293240070343,
      "learning_rate": 6.937160732358084e-06,
      "loss": 0.4241,
      "mean_token_accuracy": 0.8653839267790318,
      "num_tokens": 968509842.0,
      "step": 30346
    },
    {
      "entropy": 0.45610855519771576,
      "epoch": 2.792082851737483,
      "grad_norm": 0.14879803359508514,
      "learning_rate": 6.9340939062164575e-06,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8598269559442997,
      "num_tokens": 968541573.0,
      "step": 30347
    },
    {
      "entropy": 0.5542249362915754,
      "epoch": 2.792174858108924,
      "grad_norm": 0.16092157363891602,
      "learning_rate": 6.931027080074831e-06,
      "loss": 0.539,
      "mean_token_accuracy": 0.82866171002388,
      "num_tokens": 968573615.0,
      "step": 30348
    },
    {
      "entropy": 0.4690396236255765,
      "epoch": 2.792266864480365,
      "grad_norm": 0.1595858484506607,
      "learning_rate": 6.927960253933205e-06,
      "loss": 0.4523,
      "mean_token_accuracy": 0.8525061793625355,
      "num_tokens": 968605675.0,
      "step": 30349
    },
    {
      "entropy": 0.49374050833284855,
      "epoch": 2.7923588708518063,
      "grad_norm": 0.15264396369457245,
      "learning_rate": 6.924893427791579e-06,
      "loss": 0.4717,
      "mean_token_accuracy": 0.8512186259031296,
      "num_tokens": 968637030.0,
      "step": 30350
    },
    {
      "entropy": 0.5607643527910113,
      "epoch": 2.7924508772232475,
      "grad_norm": 0.16839735209941864,
      "learning_rate": 6.921826601649953e-06,
      "loss": 0.5487,
      "mean_token_accuracy": 0.8259950093924999,
      "num_tokens": 968668603.0,
      "step": 30351
    },
    {
      "entropy": 0.47327105700969696,
      "epoch": 2.792542883594689,
      "grad_norm": 0.14868304133415222,
      "learning_rate": 6.9187597755083265e-06,
      "loss": 0.4644,
      "mean_token_accuracy": 0.8515675701200962,
      "num_tokens": 968701294.0,
      "step": 30352
    },
    {
      "entropy": 0.4781795386224985,
      "epoch": 2.79263488996613,
      "grad_norm": 0.16086208820343018,
      "learning_rate": 6.9156929493667e-06,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8479765094816685,
      "num_tokens": 968732722.0,
      "step": 30353
    },
    {
      "entropy": 0.4901031870394945,
      "epoch": 2.7927268963375713,
      "grad_norm": 0.15874476730823517,
      "learning_rate": 6.912626123225074e-06,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8432901799678802,
      "num_tokens": 968764759.0,
      "step": 30354
    },
    {
      "entropy": 0.5683524836786091,
      "epoch": 2.7928189027090125,
      "grad_norm": 0.1709047108888626,
      "learning_rate": 6.909559297083448e-06,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8261676840484142,
      "num_tokens": 968796357.0,
      "step": 30355
    },
    {
      "entropy": 0.49389295105356723,
      "epoch": 2.7929109090804536,
      "grad_norm": 0.15814827382564545,
      "learning_rate": 6.9064924709418226e-06,
      "loss": 0.486,
      "mean_token_accuracy": 0.8477791175246239,
      "num_tokens": 968828613.0,
      "step": 30356
    },
    {
      "entropy": 0.3787026877980679,
      "epoch": 2.793002915451895,
      "grad_norm": 0.1341909021139145,
      "learning_rate": 6.903425644800196e-06,
      "loss": 0.365,
      "mean_token_accuracy": 0.8836290650069714,
      "num_tokens": 968861232.0,
      "step": 30357
    },
    {
      "entropy": 0.5641222255071625,
      "epoch": 2.7930949218233363,
      "grad_norm": 0.1612086445093155,
      "learning_rate": 6.90035881865857e-06,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8246105499565601,
      "num_tokens": 968893473.0,
      "step": 30358
    },
    {
      "entropy": 0.5795238045975566,
      "epoch": 2.7931869281947774,
      "grad_norm": 0.16418848931789398,
      "learning_rate": 6.897291992516945e-06,
      "loss": 0.5721,
      "mean_token_accuracy": 0.8222954943776131,
      "num_tokens": 968925672.0,
      "step": 30359
    },
    {
      "entropy": 0.4580790160689503,
      "epoch": 2.7932789345662186,
      "grad_norm": 0.1439974308013916,
      "learning_rate": 6.8942251663753195e-06,
      "loss": 0.4386,
      "mean_token_accuracy": 0.8589066453278065,
      "num_tokens": 968957330.0,
      "step": 30360
    },
    {
      "entropy": 0.5169559605419636,
      "epoch": 2.7933709409376597,
      "grad_norm": 0.1575789600610733,
      "learning_rate": 6.891158340233693e-06,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8384739682078362,
      "num_tokens": 968989683.0,
      "step": 30361
    },
    {
      "entropy": 0.5136721034068614,
      "epoch": 2.7934629473091013,
      "grad_norm": 0.16202287375926971,
      "learning_rate": 6.888091514092067e-06,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8392628021538258,
      "num_tokens": 969021370.0,
      "step": 30362
    },
    {
      "entropy": 0.48711987771093845,
      "epoch": 2.7935549536805424,
      "grad_norm": 0.15561635792255402,
      "learning_rate": 6.885024687950441e-06,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8453319892287254,
      "num_tokens": 969053640.0,
      "step": 30363
    },
    {
      "entropy": 0.5174467579927295,
      "epoch": 2.7936469600519835,
      "grad_norm": 0.15631291270256042,
      "learning_rate": 6.881957861808815e-06,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8379488922655582,
      "num_tokens": 969086331.0,
      "step": 30364
    },
    {
      "entropy": 0.46958582662045956,
      "epoch": 2.7937389664234247,
      "grad_norm": 0.1526488959789276,
      "learning_rate": 6.8788910356671885e-06,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8533000312745571,
      "num_tokens": 969117757.0,
      "step": 30365
    },
    {
      "entropy": 0.45339890010654926,
      "epoch": 2.793830972794866,
      "grad_norm": 0.1502118855714798,
      "learning_rate": 6.875824209525562e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8571437671780586,
      "num_tokens": 969150024.0,
      "step": 30366
    },
    {
      "entropy": 0.5368926851078868,
      "epoch": 2.7939229791663074,
      "grad_norm": 0.16401025652885437,
      "learning_rate": 6.872757383383936e-06,
      "loss": 0.5293,
      "mean_token_accuracy": 0.8333758302032948,
      "num_tokens": 969181445.0,
      "step": 30367
    },
    {
      "entropy": 0.5168677056208253,
      "epoch": 2.7940149855377485,
      "grad_norm": 0.1619430035352707,
      "learning_rate": 6.86969055724231e-06,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8423116095364094,
      "num_tokens": 969213958.0,
      "step": 30368
    },
    {
      "entropy": 0.4788007866591215,
      "epoch": 2.7941069919091897,
      "grad_norm": 0.15773333609104156,
      "learning_rate": 6.866623731100684e-06,
      "loss": 0.4727,
      "mean_token_accuracy": 0.847547248005867,
      "num_tokens": 969246206.0,
      "step": 30369
    },
    {
      "entropy": 0.5142628103494644,
      "epoch": 2.794198998280631,
      "grad_norm": 0.1563393771648407,
      "learning_rate": 6.8635569049590575e-06,
      "loss": 0.4999,
      "mean_token_accuracy": 0.8415386453270912,
      "num_tokens": 969278286.0,
      "step": 30370
    },
    {
      "entropy": 0.3807888670125976,
      "epoch": 2.794291004652072,
      "grad_norm": 0.14266978204250336,
      "learning_rate": 6.860490078817432e-06,
      "loss": 0.3569,
      "mean_token_accuracy": 0.880191370844841,
      "num_tokens": 969309164.0,
      "step": 30371
    },
    {
      "entropy": 0.546147920191288,
      "epoch": 2.7943830110235135,
      "grad_norm": 0.16168411076068878,
      "learning_rate": 6.857423252675806e-06,
      "loss": 0.5296,
      "mean_token_accuracy": 0.831158921122551,
      "num_tokens": 969341176.0,
      "step": 30372
    },
    {
      "entropy": 0.5269591575488448,
      "epoch": 2.7944750173949546,
      "grad_norm": 0.15546397864818573,
      "learning_rate": 6.85435642653418e-06,
      "loss": 0.5185,
      "mean_token_accuracy": 0.8411579579114914,
      "num_tokens": 969373241.0,
      "step": 30373
    },
    {
      "entropy": 0.4313344385009259,
      "epoch": 2.7945670237663958,
      "grad_norm": 0.14621761441230774,
      "learning_rate": 6.851289600392554e-06,
      "loss": 0.4146,
      "mean_token_accuracy": 0.8649147190153599,
      "num_tokens": 969405273.0,
      "step": 30374
    },
    {
      "entropy": 0.5878950878977776,
      "epoch": 2.794659030137837,
      "grad_norm": 0.16521503031253815,
      "learning_rate": 6.848222774250929e-06,
      "loss": 0.5773,
      "mean_token_accuracy": 0.8218938522040844,
      "num_tokens": 969437587.0,
      "step": 30375
    },
    {
      "entropy": 0.6236993838101625,
      "epoch": 2.794751036509278,
      "grad_norm": 0.17345942556858063,
      "learning_rate": 6.845155948109303e-06,
      "loss": 0.6011,
      "mean_token_accuracy": 0.8098805882036686,
      "num_tokens": 969469459.0,
      "step": 30376
    },
    {
      "entropy": 0.4789785323664546,
      "epoch": 2.7948430428807196,
      "grad_norm": 0.1585644632577896,
      "learning_rate": 6.842089121967677e-06,
      "loss": 0.4662,
      "mean_token_accuracy": 0.8481402397155762,
      "num_tokens": 969501244.0,
      "step": 30377
    },
    {
      "entropy": 0.47029204946011305,
      "epoch": 2.7949350492521607,
      "grad_norm": 0.15126247704029083,
      "learning_rate": 6.83902229582605e-06,
      "loss": 0.4537,
      "mean_token_accuracy": 0.8569206781685352,
      "num_tokens": 969533898.0,
      "step": 30378
    },
    {
      "entropy": 0.4729101210832596,
      "epoch": 2.795027055623602,
      "grad_norm": 0.15626832842826843,
      "learning_rate": 6.835955469684424e-06,
      "loss": 0.4721,
      "mean_token_accuracy": 0.8467074632644653,
      "num_tokens": 969565903.0,
      "step": 30379
    },
    {
      "entropy": 0.507772452197969,
      "epoch": 2.795119061995043,
      "grad_norm": 0.15970028936862946,
      "learning_rate": 6.832888643542798e-06,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8422327153384686,
      "num_tokens": 969598140.0,
      "step": 30380
    },
    {
      "entropy": 0.5276551449205726,
      "epoch": 2.795211068366484,
      "grad_norm": 0.15350089967250824,
      "learning_rate": 6.829821817401172e-06,
      "loss": 0.5204,
      "mean_token_accuracy": 0.8372143693268299,
      "num_tokens": 969630580.0,
      "step": 30381
    },
    {
      "entropy": 0.5963455941528082,
      "epoch": 2.7953030747379257,
      "grad_norm": 0.166901096701622,
      "learning_rate": 6.826754991259546e-06,
      "loss": 0.5777,
      "mean_token_accuracy": 0.8210935518145561,
      "num_tokens": 969661644.0,
      "step": 30382
    },
    {
      "entropy": 0.5752166183665395,
      "epoch": 2.795395081109367,
      "grad_norm": 0.17256911098957062,
      "learning_rate": 6.823688165117919e-06,
      "loss": 0.564,
      "mean_token_accuracy": 0.8217352479696274,
      "num_tokens": 969692683.0,
      "step": 30383
    },
    {
      "entropy": 0.535552715882659,
      "epoch": 2.795487087480808,
      "grad_norm": 0.1604394018650055,
      "learning_rate": 6.820621338976293e-06,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8342668563127518,
      "num_tokens": 969723982.0,
      "step": 30384
    },
    {
      "entropy": 0.48563607316464186,
      "epoch": 2.795579093852249,
      "grad_norm": 0.15830132365226746,
      "learning_rate": 6.817554512834667e-06,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8498246558010578,
      "num_tokens": 969755771.0,
      "step": 30385
    },
    {
      "entropy": 0.5388203621841967,
      "epoch": 2.7956711002236903,
      "grad_norm": 0.15676848590373993,
      "learning_rate": 6.814487686693042e-06,
      "loss": 0.5216,
      "mean_token_accuracy": 0.836349081248045,
      "num_tokens": 969787527.0,
      "step": 30386
    },
    {
      "entropy": 0.4405926277395338,
      "epoch": 2.795763106595132,
      "grad_norm": 0.14556138217449188,
      "learning_rate": 6.8114208605514155e-06,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8603208884596825,
      "num_tokens": 969819264.0,
      "step": 30387
    },
    {
      "entropy": 0.3650618444662541,
      "epoch": 2.795855112966573,
      "grad_norm": 0.13713742792606354,
      "learning_rate": 6.808354034409789e-06,
      "loss": 0.3567,
      "mean_token_accuracy": 0.885112963616848,
      "num_tokens": 969851766.0,
      "step": 30388
    },
    {
      "entropy": 0.5988733544945717,
      "epoch": 2.795947119338014,
      "grad_norm": 0.16739380359649658,
      "learning_rate": 6.805287208268164e-06,
      "loss": 0.5838,
      "mean_token_accuracy": 0.818601980805397,
      "num_tokens": 969883250.0,
      "step": 30389
    },
    {
      "entropy": 0.4029083363711834,
      "epoch": 2.7960391257094552,
      "grad_norm": 0.1429937183856964,
      "learning_rate": 6.8022203821265386e-06,
      "loss": 0.3941,
      "mean_token_accuracy": 0.8748399280011654,
      "num_tokens": 969915778.0,
      "step": 30390
    },
    {
      "entropy": 0.5245955451391637,
      "epoch": 2.7961311320808964,
      "grad_norm": 0.15919898450374603,
      "learning_rate": 6.799153555984912e-06,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8390928544104099,
      "num_tokens": 969948194.0,
      "step": 30391
    },
    {
      "entropy": 0.46634463872760534,
      "epoch": 2.796223138452338,
      "grad_norm": 0.1545812338590622,
      "learning_rate": 6.796086729843286e-06,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8526017181575298,
      "num_tokens": 969980684.0,
      "step": 30392
    },
    {
      "entropy": 0.44166831113398075,
      "epoch": 2.796315144823779,
      "grad_norm": 0.14417871832847595,
      "learning_rate": 6.79301990370166e-06,
      "loss": 0.4275,
      "mean_token_accuracy": 0.8643584102392197,
      "num_tokens": 970012940.0,
      "step": 30393
    },
    {
      "entropy": 0.5519832530990243,
      "epoch": 2.79640715119522,
      "grad_norm": 0.158479705452919,
      "learning_rate": 6.789953077560034e-06,
      "loss": 0.5333,
      "mean_token_accuracy": 0.8291753977537155,
      "num_tokens": 970045063.0,
      "step": 30394
    },
    {
      "entropy": 0.5590811362490058,
      "epoch": 2.7964991575666613,
      "grad_norm": 0.1630871295928955,
      "learning_rate": 6.7868862514184076e-06,
      "loss": 0.5355,
      "mean_token_accuracy": 0.8329234421253204,
      "num_tokens": 970076360.0,
      "step": 30395
    },
    {
      "entropy": 0.4868656983599067,
      "epoch": 2.7965911639381025,
      "grad_norm": 0.1544070541858673,
      "learning_rate": 6.783819425276781e-06,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8503972217440605,
      "num_tokens": 970108718.0,
      "step": 30396
    },
    {
      "entropy": 0.5175743289291859,
      "epoch": 2.796683170309544,
      "grad_norm": 0.1590314507484436,
      "learning_rate": 6.780752599135155e-06,
      "loss": 0.514,
      "mean_token_accuracy": 0.8364177905023098,
      "num_tokens": 970140920.0,
      "step": 30397
    },
    {
      "entropy": 0.49359590257517993,
      "epoch": 2.796775176680985,
      "grad_norm": 0.15960852801799774,
      "learning_rate": 6.777685772993529e-06,
      "loss": 0.4818,
      "mean_token_accuracy": 0.8463058210909367,
      "num_tokens": 970172679.0,
      "step": 30398
    },
    {
      "entropy": 0.5461814273148775,
      "epoch": 2.7968671830524263,
      "grad_norm": 0.1692432463169098,
      "learning_rate": 6.774618946851903e-06,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8292045406997204,
      "num_tokens": 970204607.0,
      "step": 30399
    },
    {
      "entropy": 0.5546057485044003,
      "epoch": 2.7969591894238675,
      "grad_norm": 0.16649964451789856,
      "learning_rate": 6.7715521207102766e-06,
      "loss": 0.5517,
      "mean_token_accuracy": 0.8262736089527607,
      "num_tokens": 970237052.0,
      "step": 30400
    },
    {
      "entropy": 0.5764221679419279,
      "epoch": 2.7970511957953086,
      "grad_norm": 0.16502591967582703,
      "learning_rate": 6.768485294568651e-06,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8237424157559872,
      "num_tokens": 970269305.0,
      "step": 30401
    },
    {
      "entropy": 0.5026948237791657,
      "epoch": 2.79714320216675,
      "grad_norm": 0.15459416806697845,
      "learning_rate": 6.765418468427025e-06,
      "loss": 0.4885,
      "mean_token_accuracy": 0.8415291085839272,
      "num_tokens": 970301438.0,
      "step": 30402
    },
    {
      "entropy": 0.4559685802087188,
      "epoch": 2.7972352085381913,
      "grad_norm": 0.152982696890831,
      "learning_rate": 6.762351642285399e-06,
      "loss": 0.4456,
      "mean_token_accuracy": 0.8585472218692303,
      "num_tokens": 970332597.0,
      "step": 30403
    },
    {
      "entropy": 0.4674659182783216,
      "epoch": 2.7973272149096324,
      "grad_norm": 0.15376034379005432,
      "learning_rate": 6.759284816143773e-06,
      "loss": 0.4656,
      "mean_token_accuracy": 0.853414312005043,
      "num_tokens": 970364572.0,
      "step": 30404
    },
    {
      "entropy": 0.5465095858089626,
      "epoch": 2.7974192212810736,
      "grad_norm": 0.16200688481330872,
      "learning_rate": 6.756217990002148e-06,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8255648463964462,
      "num_tokens": 970396532.0,
      "step": 30405
    },
    {
      "entropy": 0.4997724243439734,
      "epoch": 2.7975112276525147,
      "grad_norm": 0.15717960894107819,
      "learning_rate": 6.753151163860522e-06,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8499767184257507,
      "num_tokens": 970428397.0,
      "step": 30406
    },
    {
      "entropy": 0.5109283435158432,
      "epoch": 2.7976032340239563,
      "grad_norm": 0.15684561431407928,
      "learning_rate": 6.750084337718896e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.8416594043374062,
      "num_tokens": 970460708.0,
      "step": 30407
    },
    {
      "entropy": 0.3899828674038872,
      "epoch": 2.7976952403953974,
      "grad_norm": 0.14018234610557556,
      "learning_rate": 6.7470175115772695e-06,
      "loss": 0.3763,
      "mean_token_accuracy": 0.8787373378872871,
      "num_tokens": 970492925.0,
      "step": 30408
    },
    {
      "entropy": 0.554346976801753,
      "epoch": 2.7977872467668385,
      "grad_norm": 0.1591276377439499,
      "learning_rate": 6.743950685435643e-06,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8322495706379414,
      "num_tokens": 970524858.0,
      "step": 30409
    },
    {
      "entropy": 0.3822424354730174,
      "epoch": 2.7978792531382797,
      "grad_norm": 0.13349848985671997,
      "learning_rate": 6.740883859294017e-06,
      "loss": 0.3719,
      "mean_token_accuracy": 0.8817722871899605,
      "num_tokens": 970557501.0,
      "step": 30410
    },
    {
      "entropy": 0.5444119875319302,
      "epoch": 2.797971259509721,
      "grad_norm": 0.1620083600282669,
      "learning_rate": 6.737817033152391e-06,
      "loss": 0.5299,
      "mean_token_accuracy": 0.8306681402027607,
      "num_tokens": 970589337.0,
      "step": 30411
    },
    {
      "entropy": 0.5723017565906048,
      "epoch": 2.7980632658811624,
      "grad_norm": 0.169440358877182,
      "learning_rate": 6.734750207010765e-06,
      "loss": 0.5597,
      "mean_token_accuracy": 0.8247070647776127,
      "num_tokens": 970621334.0,
      "step": 30412
    },
    {
      "entropy": 0.4374276986345649,
      "epoch": 2.7981552722526035,
      "grad_norm": 0.1511225700378418,
      "learning_rate": 6.7316833808691385e-06,
      "loss": 0.4283,
      "mean_token_accuracy": 0.862023688852787,
      "num_tokens": 970654102.0,
      "step": 30413
    },
    {
      "entropy": 0.5517580267041922,
      "epoch": 2.7982472786240447,
      "grad_norm": 0.17361736297607422,
      "learning_rate": 6.728616554727512e-06,
      "loss": 0.5463,
      "mean_token_accuracy": 0.8207810446619987,
      "num_tokens": 970685112.0,
      "step": 30414
    },
    {
      "entropy": 0.44111364614218473,
      "epoch": 2.798339284995486,
      "grad_norm": 0.15212002396583557,
      "learning_rate": 6.725549728585886e-06,
      "loss": 0.4329,
      "mean_token_accuracy": 0.8618534952402115,
      "num_tokens": 970716739.0,
      "step": 30415
    },
    {
      "entropy": 0.5458918698132038,
      "epoch": 2.798431291366927,
      "grad_norm": 0.15688586235046387,
      "learning_rate": 6.722482902444261e-06,
      "loss": 0.5254,
      "mean_token_accuracy": 0.8348374702036381,
      "num_tokens": 970748845.0,
      "step": 30416
    },
    {
      "entropy": 0.5723035559058189,
      "epoch": 2.7985232977383685,
      "grad_norm": 0.1651974469423294,
      "learning_rate": 6.7194160763026346e-06,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8288425207138062,
      "num_tokens": 970781111.0,
      "step": 30417
    },
    {
      "entropy": 0.46552191860973835,
      "epoch": 2.7986153041098096,
      "grad_norm": 0.15378262102603912,
      "learning_rate": 6.716349250161008e-06,
      "loss": 0.456,
      "mean_token_accuracy": 0.8521288186311722,
      "num_tokens": 970812556.0,
      "step": 30418
    },
    {
      "entropy": 0.5143972039222717,
      "epoch": 2.7987073104812508,
      "grad_norm": 0.1592797338962555,
      "learning_rate": 6.713282424019382e-06,
      "loss": 0.5115,
      "mean_token_accuracy": 0.839690487831831,
      "num_tokens": 970844461.0,
      "step": 30419
    },
    {
      "entropy": 0.5346896860282868,
      "epoch": 2.798799316852692,
      "grad_norm": 0.1582200676202774,
      "learning_rate": 6.710215597877758e-06,
      "loss": 0.5255,
      "mean_token_accuracy": 0.8339126668870449,
      "num_tokens": 970876739.0,
      "step": 30420
    },
    {
      "entropy": 0.4718807879835367,
      "epoch": 2.798891323224133,
      "grad_norm": 0.15245452523231506,
      "learning_rate": 6.7071487717361315e-06,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8514221981167793,
      "num_tokens": 970909212.0,
      "step": 30421
    },
    {
      "entropy": 0.5915031172335148,
      "epoch": 2.7989833295955746,
      "grad_norm": 0.16949743032455444,
      "learning_rate": 6.704081945594505e-06,
      "loss": 0.573,
      "mean_token_accuracy": 0.816857110708952,
      "num_tokens": 970940565.0,
      "step": 30422
    },
    {
      "entropy": 0.4598831102484837,
      "epoch": 2.7990753359670157,
      "grad_norm": 0.152483731508255,
      "learning_rate": 6.701015119452879e-06,
      "loss": 0.4411,
      "mean_token_accuracy": 0.8600709289312363,
      "num_tokens": 970971675.0,
      "step": 30423
    },
    {
      "entropy": 0.5235786717385054,
      "epoch": 2.799167342338457,
      "grad_norm": 0.15520192682743073,
      "learning_rate": 6.697948293311253e-06,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8384796716272831,
      "num_tokens": 971004339.0,
      "step": 30424
    },
    {
      "entropy": 0.4771978436037898,
      "epoch": 2.799259348709898,
      "grad_norm": 0.15424175560474396,
      "learning_rate": 6.694881467169627e-06,
      "loss": 0.4674,
      "mean_token_accuracy": 0.8505446314811707,
      "num_tokens": 971035743.0,
      "step": 30425
    },
    {
      "entropy": 0.5055127213709056,
      "epoch": 2.799351355081339,
      "grad_norm": 0.1597302258014679,
      "learning_rate": 6.6918146410280005e-06,
      "loss": 0.4923,
      "mean_token_accuracy": 0.8437057323753834,
      "num_tokens": 971067829.0,
      "step": 30426
    },
    {
      "entropy": 0.5376382418908179,
      "epoch": 2.7994433614527807,
      "grad_norm": 0.15918926894664764,
      "learning_rate": 6.688747814886374e-06,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8330714628100395,
      "num_tokens": 971099290.0,
      "step": 30427
    },
    {
      "entropy": 0.5394625980407,
      "epoch": 2.799535367824222,
      "grad_norm": 0.15870702266693115,
      "learning_rate": 6.685680988744748e-06,
      "loss": 0.535,
      "mean_token_accuracy": 0.8315107859671116,
      "num_tokens": 971131335.0,
      "step": 30428
    },
    {
      "entropy": 0.5317857135087252,
      "epoch": 2.799627374195663,
      "grad_norm": 0.15651960670948029,
      "learning_rate": 6.682614162603122e-06,
      "loss": 0.518,
      "mean_token_accuracy": 0.8330998830497265,
      "num_tokens": 971163007.0,
      "step": 30429
    },
    {
      "entropy": 0.5793699477799237,
      "epoch": 2.799719380567104,
      "grad_norm": 0.17161577939987183,
      "learning_rate": 6.679547336461496e-06,
      "loss": 0.5815,
      "mean_token_accuracy": 0.819656565785408,
      "num_tokens": 971194388.0,
      "step": 30430
    },
    {
      "entropy": 0.4323836388066411,
      "epoch": 2.7998113869385453,
      "grad_norm": 0.1453924924135208,
      "learning_rate": 6.67648051031987e-06,
      "loss": 0.4248,
      "mean_token_accuracy": 0.8628164939582348,
      "num_tokens": 971226419.0,
      "step": 30431
    },
    {
      "entropy": 0.4233992053195834,
      "epoch": 2.799903393309987,
      "grad_norm": 0.14413870871067047,
      "learning_rate": 6.673413684178244e-06,
      "loss": 0.3992,
      "mean_token_accuracy": 0.8728900328278542,
      "num_tokens": 971258169.0,
      "step": 30432
    },
    {
      "entropy": 0.5064394064247608,
      "epoch": 2.799995399681428,
      "grad_norm": 0.15747787058353424,
      "learning_rate": 6.670346858036618e-06,
      "loss": 0.4923,
      "mean_token_accuracy": 0.845728013664484,
      "num_tokens": 971290455.0,
      "step": 30433
    },
    {
      "entropy": 0.5409872699528933,
      "epoch": 2.800087406052869,
      "grad_norm": 0.16058823466300964,
      "learning_rate": 6.667280031894992e-06,
      "loss": 0.5211,
      "mean_token_accuracy": 0.8337398394942284,
      "num_tokens": 971322891.0,
      "step": 30434
    },
    {
      "entropy": 0.4452425049385056,
      "epoch": 2.8001794124243102,
      "grad_norm": 0.15119628608226776,
      "learning_rate": 6.664213205753367e-06,
      "loss": 0.426,
      "mean_token_accuracy": 0.8630553372204304,
      "num_tokens": 971354234.0,
      "step": 30435
    },
    {
      "entropy": 0.5000236453488469,
      "epoch": 2.8002714187957514,
      "grad_norm": 0.15931101143360138,
      "learning_rate": 6.661146379611741e-06,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8455277755856514,
      "num_tokens": 971386911.0,
      "step": 30436
    },
    {
      "entropy": 0.44963734271004796,
      "epoch": 2.800363425167193,
      "grad_norm": 0.1500099003314972,
      "learning_rate": 6.658079553470115e-06,
      "loss": 0.4305,
      "mean_token_accuracy": 0.8604603707790375,
      "num_tokens": 971418328.0,
      "step": 30437
    },
    {
      "entropy": 0.5444144096691161,
      "epoch": 2.800455431538634,
      "grad_norm": 0.16181644797325134,
      "learning_rate": 6.655012727328489e-06,
      "loss": 0.5344,
      "mean_token_accuracy": 0.8307526223361492,
      "num_tokens": 971450318.0,
      "step": 30438
    },
    {
      "entropy": 0.4765969570726156,
      "epoch": 2.800547437910075,
      "grad_norm": 0.15627625584602356,
      "learning_rate": 6.651945901186862e-06,
      "loss": 0.4629,
      "mean_token_accuracy": 0.849282618612051,
      "num_tokens": 971482865.0,
      "step": 30439
    },
    {
      "entropy": 0.5478823781013489,
      "epoch": 2.8006394442815163,
      "grad_norm": 0.16560038924217224,
      "learning_rate": 6.648879075045236e-06,
      "loss": 0.5327,
      "mean_token_accuracy": 0.8309085518121719,
      "num_tokens": 971514767.0,
      "step": 30440
    },
    {
      "entropy": 0.5312973856925964,
      "epoch": 2.8007314506529575,
      "grad_norm": 0.16494449973106384,
      "learning_rate": 6.64581224890361e-06,
      "loss": 0.5292,
      "mean_token_accuracy": 0.8340912610292435,
      "num_tokens": 971546133.0,
      "step": 30441
    },
    {
      "entropy": 0.41684141964651644,
      "epoch": 2.800823457024399,
      "grad_norm": 0.1450413465499878,
      "learning_rate": 6.642745422761984e-06,
      "loss": 0.4068,
      "mean_token_accuracy": 0.8685093373060226,
      "num_tokens": 971578121.0,
      "step": 30442
    },
    {
      "entropy": 0.4819603208452463,
      "epoch": 2.80091546339584,
      "grad_norm": 0.15848664939403534,
      "learning_rate": 6.639678596620358e-06,
      "loss": 0.473,
      "mean_token_accuracy": 0.8523603267967701,
      "num_tokens": 971610365.0,
      "step": 30443
    },
    {
      "entropy": 0.5599838870111853,
      "epoch": 2.8010074697672813,
      "grad_norm": 0.16826626658439636,
      "learning_rate": 6.636611770478731e-06,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8261385634541512,
      "num_tokens": 971642233.0,
      "step": 30444
    },
    {
      "entropy": 0.5382668664678931,
      "epoch": 2.8010994761387225,
      "grad_norm": 0.16835641860961914,
      "learning_rate": 6.633544944337105e-06,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8352462910115719,
      "num_tokens": 971673221.0,
      "step": 30445
    },
    {
      "entropy": 0.5232848906889558,
      "epoch": 2.8011914825101636,
      "grad_norm": 0.1617826223373413,
      "learning_rate": 6.63047811819548e-06,
      "loss": 0.5027,
      "mean_token_accuracy": 0.8380308896303177,
      "num_tokens": 971704938.0,
      "step": 30446
    },
    {
      "entropy": 0.5306102419272065,
      "epoch": 2.801283488881605,
      "grad_norm": 0.1628749519586563,
      "learning_rate": 6.627411292053854e-06,
      "loss": 0.523,
      "mean_token_accuracy": 0.835022933781147,
      "num_tokens": 971737195.0,
      "step": 30447
    },
    {
      "entropy": 0.4485223377123475,
      "epoch": 2.8013754952530463,
      "grad_norm": 0.15683552622795105,
      "learning_rate": 6.6243444659122275e-06,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8560801222920418,
      "num_tokens": 971768710.0,
      "step": 30448
    },
    {
      "entropy": 0.4591483920812607,
      "epoch": 2.8014675016244874,
      "grad_norm": 0.16052459180355072,
      "learning_rate": 6.621277639770601e-06,
      "loss": 0.4497,
      "mean_token_accuracy": 0.8565316423773766,
      "num_tokens": 971800498.0,
      "step": 30449
    },
    {
      "entropy": 0.5496203750371933,
      "epoch": 2.8015595079959286,
      "grad_norm": 0.16439813375473022,
      "learning_rate": 6.618210813628977e-06,
      "loss": 0.5381,
      "mean_token_accuracy": 0.8279398940503597,
      "num_tokens": 971832324.0,
      "step": 30450
    },
    {
      "entropy": 0.5049692243337631,
      "epoch": 2.8016515143673697,
      "grad_norm": 0.15890571475028992,
      "learning_rate": 6.6151439874873506e-06,
      "loss": 0.4999,
      "mean_token_accuracy": 0.8432640172541142,
      "num_tokens": 971864922.0,
      "step": 30451
    },
    {
      "entropy": 0.4525249404832721,
      "epoch": 2.8017435207388113,
      "grad_norm": 0.15022626519203186,
      "learning_rate": 6.612077161345724e-06,
      "loss": 0.439,
      "mean_token_accuracy": 0.8595802672207355,
      "num_tokens": 971896217.0,
      "step": 30452
    },
    {
      "entropy": 0.5511234177974984,
      "epoch": 2.8018355271102524,
      "grad_norm": 0.1612405776977539,
      "learning_rate": 6.609010335204098e-06,
      "loss": 0.5323,
      "mean_token_accuracy": 0.8312811590731144,
      "num_tokens": 971927532.0,
      "step": 30453
    },
    {
      "entropy": 0.42735681124031544,
      "epoch": 2.8019275334816935,
      "grad_norm": 0.14764010906219482,
      "learning_rate": 6.605943509062472e-06,
      "loss": 0.4178,
      "mean_token_accuracy": 0.8666453137993813,
      "num_tokens": 971959182.0,
      "step": 30454
    },
    {
      "entropy": 0.4494820907711983,
      "epoch": 2.8020195398531347,
      "grad_norm": 0.15336695313453674,
      "learning_rate": 6.602876682920846e-06,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8563925474882126,
      "num_tokens": 971991147.0,
      "step": 30455
    },
    {
      "entropy": 0.4860591231845319,
      "epoch": 2.802111546224576,
      "grad_norm": 0.15043209493160248,
      "learning_rate": 6.5998098567792196e-06,
      "loss": 0.4677,
      "mean_token_accuracy": 0.8530218601226807,
      "num_tokens": 972023255.0,
      "step": 30456
    },
    {
      "entropy": 0.47580219246447086,
      "epoch": 2.8022035525960174,
      "grad_norm": 0.15609833598136902,
      "learning_rate": 6.596743030637593e-06,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8484720401465893,
      "num_tokens": 972055684.0,
      "step": 30457
    },
    {
      "entropy": 0.5829398930072784,
      "epoch": 2.8022955589674585,
      "grad_norm": 0.17041437327861786,
      "learning_rate": 6.593676204495967e-06,
      "loss": 0.5743,
      "mean_token_accuracy": 0.8180903270840645,
      "num_tokens": 972087997.0,
      "step": 30458
    },
    {
      "entropy": 0.47399924509227276,
      "epoch": 2.8023875653388997,
      "grad_norm": 0.15162284672260284,
      "learning_rate": 6.590609378354341e-06,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8475430905818939,
      "num_tokens": 972120581.0,
      "step": 30459
    },
    {
      "entropy": 0.40237674047239125,
      "epoch": 2.802479571710341,
      "grad_norm": 0.14483578503131866,
      "learning_rate": 6.587542552212715e-06,
      "loss": 0.3971,
      "mean_token_accuracy": 0.8716989904642105,
      "num_tokens": 972152947.0,
      "step": 30460
    },
    {
      "entropy": 0.539516176097095,
      "epoch": 2.802571578081782,
      "grad_norm": 0.16848914325237274,
      "learning_rate": 6.584475726071089e-06,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8327960707247257,
      "num_tokens": 972183727.0,
      "step": 30461
    },
    {
      "entropy": 0.5709201411809772,
      "epoch": 2.8026635844532235,
      "grad_norm": 0.16604192554950714,
      "learning_rate": 6.581408899929463e-06,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8246339075267315,
      "num_tokens": 972214687.0,
      "step": 30462
    },
    {
      "entropy": 0.555259732529521,
      "epoch": 2.8027555908246646,
      "grad_norm": 0.1643468737602234,
      "learning_rate": 6.578342073787837e-06,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8304022029042244,
      "num_tokens": 972246209.0,
      "step": 30463
    },
    {
      "entropy": 0.5013059973716736,
      "epoch": 2.8028475971961058,
      "grad_norm": 0.15843918919563293,
      "learning_rate": 6.575275247646211e-06,
      "loss": 0.4877,
      "mean_token_accuracy": 0.8432217426598072,
      "num_tokens": 972277339.0,
      "step": 30464
    },
    {
      "entropy": 0.46792725287377834,
      "epoch": 2.802939603567547,
      "grad_norm": 0.15347181260585785,
      "learning_rate": 6.572208421504585e-06,
      "loss": 0.4651,
      "mean_token_accuracy": 0.8523520305752754,
      "num_tokens": 972309493.0,
      "step": 30465
    },
    {
      "entropy": 0.44304906390607357,
      "epoch": 2.803031609938988,
      "grad_norm": 0.14832434058189392,
      "learning_rate": 6.56914159536296e-06,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8630924485623837,
      "num_tokens": 972341247.0,
      "step": 30466
    },
    {
      "entropy": 0.44591879891231656,
      "epoch": 2.8031236163104296,
      "grad_norm": 0.15284158289432526,
      "learning_rate": 6.566074769221334e-06,
      "loss": 0.4399,
      "mean_token_accuracy": 0.8588831573724747,
      "num_tokens": 972373242.0,
      "step": 30467
    },
    {
      "entropy": 0.5061817332170904,
      "epoch": 2.8032156226818707,
      "grad_norm": 0.15151607990264893,
      "learning_rate": 6.563007943079708e-06,
      "loss": 0.4905,
      "mean_token_accuracy": 0.8414368964731693,
      "num_tokens": 972405346.0,
      "step": 30468
    },
    {
      "entropy": 0.5180935747921467,
      "epoch": 2.803307629053312,
      "grad_norm": 0.15541726350784302,
      "learning_rate": 6.5599411169380815e-06,
      "loss": 0.511,
      "mean_token_accuracy": 0.8383944407105446,
      "num_tokens": 972437147.0,
      "step": 30469
    },
    {
      "entropy": 0.5622882889583707,
      "epoch": 2.803399635424753,
      "grad_norm": 0.16524575650691986,
      "learning_rate": 6.556874290796455e-06,
      "loss": 0.5486,
      "mean_token_accuracy": 0.8322235606610775,
      "num_tokens": 972469444.0,
      "step": 30470
    },
    {
      "entropy": 0.4654612001031637,
      "epoch": 2.803491641796194,
      "grad_norm": 0.15020887553691864,
      "learning_rate": 6.553807464654829e-06,
      "loss": 0.4533,
      "mean_token_accuracy": 0.8572098426520824,
      "num_tokens": 972501475.0,
      "step": 30471
    },
    {
      "entropy": 0.669417698867619,
      "epoch": 2.8035836481676357,
      "grad_norm": 0.18078412115573883,
      "learning_rate": 6.550740638513203e-06,
      "loss": 0.6506,
      "mean_token_accuracy": 0.7982288971543312,
      "num_tokens": 972531915.0,
      "step": 30472
    },
    {
      "entropy": 0.3834473055321723,
      "epoch": 2.803675654539077,
      "grad_norm": 0.1383625864982605,
      "learning_rate": 6.547673812371577e-06,
      "loss": 0.3751,
      "mean_token_accuracy": 0.8798303194344044,
      "num_tokens": 972564237.0,
      "step": 30473
    },
    {
      "entropy": 0.557874963618815,
      "epoch": 2.803767660910518,
      "grad_norm": 0.16620750725269318,
      "learning_rate": 6.5446069862299505e-06,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8304154388606548,
      "num_tokens": 972596222.0,
      "step": 30474
    },
    {
      "entropy": 0.5108117586933076,
      "epoch": 2.803859667281959,
      "grad_norm": 0.16131913661956787,
      "learning_rate": 6.541540160088324e-06,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8385573774576187,
      "num_tokens": 972627317.0,
      "step": 30475
    },
    {
      "entropy": 0.5306404074653983,
      "epoch": 2.8039516736534003,
      "grad_norm": 0.16037771105766296,
      "learning_rate": 6.538473333946699e-06,
      "loss": 0.5142,
      "mean_token_accuracy": 0.8345720618963242,
      "num_tokens": 972659013.0,
      "step": 30476
    },
    {
      "entropy": 0.6024982165545225,
      "epoch": 2.804043680024842,
      "grad_norm": 0.16544075310230255,
      "learning_rate": 6.535406507805073e-06,
      "loss": 0.592,
      "mean_token_accuracy": 0.8144619539380074,
      "num_tokens": 972691157.0,
      "step": 30477
    },
    {
      "entropy": 0.5179563809651881,
      "epoch": 2.804135686396283,
      "grad_norm": 0.16129447519779205,
      "learning_rate": 6.5323396816634466e-06,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8387492783367634,
      "num_tokens": 972723173.0,
      "step": 30478
    },
    {
      "entropy": 0.5905110910534859,
      "epoch": 2.804227692767724,
      "grad_norm": 0.16741496324539185,
      "learning_rate": 6.52927285552182e-06,
      "loss": 0.5758,
      "mean_token_accuracy": 0.8172625228762627,
      "num_tokens": 972754559.0,
      "step": 30479
    },
    {
      "entropy": 0.43406166858039796,
      "epoch": 2.8043196991391652,
      "grad_norm": 0.14966526627540588,
      "learning_rate": 6.526206029380194e-06,
      "loss": 0.4279,
      "mean_token_accuracy": 0.8617645353078842,
      "num_tokens": 972787146.0,
      "step": 30480
    },
    {
      "entropy": 0.5529153328388929,
      "epoch": 2.8044117055106064,
      "grad_norm": 0.15997135639190674,
      "learning_rate": 6.52313920323857e-06,
      "loss": 0.5327,
      "mean_token_accuracy": 0.8325431384146214,
      "num_tokens": 972819116.0,
      "step": 30481
    },
    {
      "entropy": 0.49908547289669514,
      "epoch": 2.804503711882048,
      "grad_norm": 0.1530493199825287,
      "learning_rate": 6.5200723770969435e-06,
      "loss": 0.4899,
      "mean_token_accuracy": 0.8448295369744301,
      "num_tokens": 972851336.0,
      "step": 30482
    },
    {
      "entropy": 0.4645526260137558,
      "epoch": 2.804595718253489,
      "grad_norm": 0.15149617195129395,
      "learning_rate": 6.517005550955317e-06,
      "loss": 0.4624,
      "mean_token_accuracy": 0.8524939157068729,
      "num_tokens": 972883776.0,
      "step": 30483
    },
    {
      "entropy": 0.45810688857454807,
      "epoch": 2.80468772462493,
      "grad_norm": 0.15644383430480957,
      "learning_rate": 6.513938724813691e-06,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8594631403684616,
      "num_tokens": 972915365.0,
      "step": 30484
    },
    {
      "entropy": 0.44710139429662377,
      "epoch": 2.8047797309963713,
      "grad_norm": 0.14910268783569336,
      "learning_rate": 6.510871898672065e-06,
      "loss": 0.4348,
      "mean_token_accuracy": 0.8624435178935528,
      "num_tokens": 972947334.0,
      "step": 30485
    },
    {
      "entropy": 0.5708191487938166,
      "epoch": 2.8048717373678125,
      "grad_norm": 0.1654990315437317,
      "learning_rate": 6.507805072530439e-06,
      "loss": 0.5579,
      "mean_token_accuracy": 0.8243352361023426,
      "num_tokens": 972979074.0,
      "step": 30486
    },
    {
      "entropy": 0.5076065510511398,
      "epoch": 2.804963743739254,
      "grad_norm": 0.15596745908260345,
      "learning_rate": 6.5047382463888125e-06,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8439218141138554,
      "num_tokens": 973011829.0,
      "step": 30487
    },
    {
      "entropy": 0.5421864164527506,
      "epoch": 2.805055750110695,
      "grad_norm": 0.16121059656143188,
      "learning_rate": 6.501671420247186e-06,
      "loss": 0.5285,
      "mean_token_accuracy": 0.8344056159257889,
      "num_tokens": 973044052.0,
      "step": 30488
    },
    {
      "entropy": 0.5598785893525928,
      "epoch": 2.8051477564821363,
      "grad_norm": 0.16331914067268372,
      "learning_rate": 6.49860459410556e-06,
      "loss": 0.5506,
      "mean_token_accuracy": 0.8291593194007874,
      "num_tokens": 973075813.0,
      "step": 30489
    },
    {
      "entropy": 0.4097729285713285,
      "epoch": 2.8052397628535775,
      "grad_norm": 0.14326457679271698,
      "learning_rate": 6.495537767963934e-06,
      "loss": 0.4014,
      "mean_token_accuracy": 0.8711020052433014,
      "num_tokens": 973107488.0,
      "step": 30490
    },
    {
      "entropy": 0.513747138902545,
      "epoch": 2.8053317692250186,
      "grad_norm": 0.16067026555538177,
      "learning_rate": 6.4924709418223085e-06,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8406629972159863,
      "num_tokens": 973139401.0,
      "step": 30491
    },
    {
      "entropy": 0.5105814291164279,
      "epoch": 2.80542377559646,
      "grad_norm": 0.15965856611728668,
      "learning_rate": 6.489404115680682e-06,
      "loss": 0.4987,
      "mean_token_accuracy": 0.838354978710413,
      "num_tokens": 973171331.0,
      "step": 30492
    },
    {
      "entropy": 0.6165071688592434,
      "epoch": 2.8055157819679013,
      "grad_norm": 0.17489244043827057,
      "learning_rate": 6.486337289539056e-06,
      "loss": 0.6075,
      "mean_token_accuracy": 0.8100163228809834,
      "num_tokens": 973202948.0,
      "step": 30493
    },
    {
      "entropy": 0.5349739408120513,
      "epoch": 2.8056077883393424,
      "grad_norm": 0.15802495181560516,
      "learning_rate": 6.48327046339743e-06,
      "loss": 0.5277,
      "mean_token_accuracy": 0.8343314900994301,
      "num_tokens": 973235445.0,
      "step": 30494
    },
    {
      "entropy": 0.49301846884191036,
      "epoch": 2.8056997947107836,
      "grad_norm": 0.15502195060253143,
      "learning_rate": 6.480203637255804e-06,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8497648313641548,
      "num_tokens": 973267672.0,
      "step": 30495
    },
    {
      "entropy": 0.5762944426387548,
      "epoch": 2.8057918010822247,
      "grad_norm": 0.1601577252149582,
      "learning_rate": 6.477136811114179e-06,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8202486410737038,
      "num_tokens": 973300252.0,
      "step": 30496
    },
    {
      "entropy": 0.5215287493774667,
      "epoch": 2.8058838074536663,
      "grad_norm": 0.1599845588207245,
      "learning_rate": 6.474069984972553e-06,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8412021175026894,
      "num_tokens": 973332492.0,
      "step": 30497
    },
    {
      "entropy": 0.5390173280611634,
      "epoch": 2.8059758138251074,
      "grad_norm": 0.1628226339817047,
      "learning_rate": 6.471003158830927e-06,
      "loss": 0.5256,
      "mean_token_accuracy": 0.8315355516970158,
      "num_tokens": 973364525.0,
      "step": 30498
    },
    {
      "entropy": 0.5316117564216256,
      "epoch": 2.8060678201965485,
      "grad_norm": 0.15606757998466492,
      "learning_rate": 6.467936332689301e-06,
      "loss": 0.5232,
      "mean_token_accuracy": 0.8370052017271519,
      "num_tokens": 973396979.0,
      "step": 30499
    },
    {
      "entropy": 0.40914997179061174,
      "epoch": 2.8061598265679897,
      "grad_norm": 0.14255282282829285,
      "learning_rate": 6.464869506547674e-06,
      "loss": 0.3928,
      "mean_token_accuracy": 0.876232098788023,
      "num_tokens": 973428917.0,
      "step": 30500
    },
    {
      "entropy": 0.4161196493078023,
      "epoch": 2.806251832939431,
      "grad_norm": 0.14334042370319366,
      "learning_rate": 6.461802680406048e-06,
      "loss": 0.411,
      "mean_token_accuracy": 0.8679492771625519,
      "num_tokens": 973460610.0,
      "step": 30501
    },
    {
      "entropy": 0.4899792140349746,
      "epoch": 2.8063438393108724,
      "grad_norm": 0.1526852250099182,
      "learning_rate": 6.458735854264422e-06,
      "loss": 0.4862,
      "mean_token_accuracy": 0.8487161360681057,
      "num_tokens": 973492638.0,
      "step": 30502
    },
    {
      "entropy": 0.473813614808023,
      "epoch": 2.8064358456823135,
      "grad_norm": 0.15518473088741302,
      "learning_rate": 6.455669028122796e-06,
      "loss": 0.454,
      "mean_token_accuracy": 0.8521016277372837,
      "num_tokens": 973524632.0,
      "step": 30503
    },
    {
      "entropy": 0.5865078419446945,
      "epoch": 2.8065278520537547,
      "grad_norm": 0.17410552501678467,
      "learning_rate": 6.45260220198117e-06,
      "loss": 0.5831,
      "mean_token_accuracy": 0.8173665031790733,
      "num_tokens": 973556830.0,
      "step": 30504
    },
    {
      "entropy": 0.5460788663476706,
      "epoch": 2.806619858425196,
      "grad_norm": 0.1607493907213211,
      "learning_rate": 6.449535375839543e-06,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8309501595795155,
      "num_tokens": 973588603.0,
      "step": 30505
    },
    {
      "entropy": 0.38505503279156983,
      "epoch": 2.806711864796637,
      "grad_norm": 0.1338462382555008,
      "learning_rate": 6.446468549697918e-06,
      "loss": 0.373,
      "mean_token_accuracy": 0.8784095980226994,
      "num_tokens": 973621304.0,
      "step": 30506
    },
    {
      "entropy": 0.5286556538194418,
      "epoch": 2.8068038711680785,
      "grad_norm": 0.16406309604644775,
      "learning_rate": 6.443401723556292e-06,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8337962739169598,
      "num_tokens": 973653125.0,
      "step": 30507
    },
    {
      "entropy": 0.4332569227553904,
      "epoch": 2.8068958775395196,
      "grad_norm": 0.1484663486480713,
      "learning_rate": 6.440334897414666e-06,
      "loss": 0.4206,
      "mean_token_accuracy": 0.8650980219244957,
      "num_tokens": 973685302.0,
      "step": 30508
    },
    {
      "entropy": 0.5521515840664506,
      "epoch": 2.8069878839109608,
      "grad_norm": 0.164997398853302,
      "learning_rate": 6.4372680712730395e-06,
      "loss": 0.5415,
      "mean_token_accuracy": 0.8298868201673031,
      "num_tokens": 973716786.0,
      "step": 30509
    },
    {
      "entropy": 0.5733310673385859,
      "epoch": 2.807079890282402,
      "grad_norm": 0.1633637398481369,
      "learning_rate": 6.434201245131413e-06,
      "loss": 0.5605,
      "mean_token_accuracy": 0.8224911838769913,
      "num_tokens": 973748738.0,
      "step": 30510
    },
    {
      "entropy": 0.47113856533542275,
      "epoch": 2.807171896653843,
      "grad_norm": 0.15312598645687103,
      "learning_rate": 6.431134418989789e-06,
      "loss": 0.4431,
      "mean_token_accuracy": 0.8558329045772552,
      "num_tokens": 973779225.0,
      "step": 30511
    },
    {
      "entropy": 0.4913472547195852,
      "epoch": 2.8072639030252846,
      "grad_norm": 0.1496400386095047,
      "learning_rate": 6.4280675928481626e-06,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8463104739785194,
      "num_tokens": 973811452.0,
      "step": 30512
    },
    {
      "entropy": 0.6547150444239378,
      "epoch": 2.8073559093967257,
      "grad_norm": 0.1780930906534195,
      "learning_rate": 6.425000766706536e-06,
      "loss": 0.6432,
      "mean_token_accuracy": 0.8005599789321423,
      "num_tokens": 973843222.0,
      "step": 30513
    },
    {
      "entropy": 0.4745524504687637,
      "epoch": 2.807447915768167,
      "grad_norm": 0.1487613320350647,
      "learning_rate": 6.42193394056491e-06,
      "loss": 0.4575,
      "mean_token_accuracy": 0.8511054217815399,
      "num_tokens": 973875245.0,
      "step": 30514
    },
    {
      "entropy": 0.5789120635017753,
      "epoch": 2.807539922139608,
      "grad_norm": 0.1718229204416275,
      "learning_rate": 6.418867114423284e-06,
      "loss": 0.5635,
      "mean_token_accuracy": 0.8185384273529053,
      "num_tokens": 973906789.0,
      "step": 30515
    },
    {
      "entropy": 0.4542483971454203,
      "epoch": 2.807631928511049,
      "grad_norm": 0.15284189581871033,
      "learning_rate": 6.415800288281658e-06,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8576742708683014,
      "num_tokens": 973939135.0,
      "step": 30516
    },
    {
      "entropy": 0.42598086409270763,
      "epoch": 2.8077239348824907,
      "grad_norm": 0.14874184131622314,
      "learning_rate": 6.4127334621400316e-06,
      "loss": 0.418,
      "mean_token_accuracy": 0.8649560622870922,
      "num_tokens": 973970912.0,
      "step": 30517
    },
    {
      "entropy": 0.5373815149068832,
      "epoch": 2.807815941253932,
      "grad_norm": 0.16300587356090546,
      "learning_rate": 6.409666635998405e-06,
      "loss": 0.527,
      "mean_token_accuracy": 0.8289806097745895,
      "num_tokens": 974002517.0,
      "step": 30518
    },
    {
      "entropy": 0.36156840925104916,
      "epoch": 2.807907947625373,
      "grad_norm": 0.13701185584068298,
      "learning_rate": 6.406599809856779e-06,
      "loss": 0.3478,
      "mean_token_accuracy": 0.884867824614048,
      "num_tokens": 974034352.0,
      "step": 30519
    },
    {
      "entropy": 0.4349507298320532,
      "epoch": 2.807999953996814,
      "grad_norm": 0.1527503877878189,
      "learning_rate": 6.403532983715153e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8618253357708454,
      "num_tokens": 974065731.0,
      "step": 30520
    },
    {
      "entropy": 0.5650129290297627,
      "epoch": 2.8080919603682553,
      "grad_norm": 0.17180147767066956,
      "learning_rate": 6.400466157573528e-06,
      "loss": 0.5504,
      "mean_token_accuracy": 0.82822410389781,
      "num_tokens": 974097758.0,
      "step": 30521
    },
    {
      "entropy": 0.44568252097815275,
      "epoch": 2.808183966739697,
      "grad_norm": 0.15803413093090057,
      "learning_rate": 6.397399331431901e-06,
      "loss": 0.4471,
      "mean_token_accuracy": 0.857530515640974,
      "num_tokens": 974130391.0,
      "step": 30522
    },
    {
      "entropy": 0.5297908931970596,
      "epoch": 2.808275973111138,
      "grad_norm": 0.16486890614032745,
      "learning_rate": 6.394332505290275e-06,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8361472450196743,
      "num_tokens": 974161756.0,
      "step": 30523
    },
    {
      "entropy": 0.5321033101063222,
      "epoch": 2.808367979482579,
      "grad_norm": 0.158965066075325,
      "learning_rate": 6.391265679148649e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.8358894251286983,
      "num_tokens": 974193151.0,
      "step": 30524
    },
    {
      "entropy": 0.4531360436230898,
      "epoch": 2.8084599858540202,
      "grad_norm": 0.1558816283941269,
      "learning_rate": 6.388198853007023e-06,
      "loss": 0.4452,
      "mean_token_accuracy": 0.8540102988481522,
      "num_tokens": 974224678.0,
      "step": 30525
    },
    {
      "entropy": 0.5738856215029955,
      "epoch": 2.8085519922254614,
      "grad_norm": 0.16288766264915466,
      "learning_rate": 6.385132026865397e-06,
      "loss": 0.5609,
      "mean_token_accuracy": 0.8214993961155415,
      "num_tokens": 974256898.0,
      "step": 30526
    },
    {
      "entropy": 0.46217628428712487,
      "epoch": 2.808643998596903,
      "grad_norm": 0.15170352160930634,
      "learning_rate": 6.382065200723772e-06,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8608363941311836,
      "num_tokens": 974288989.0,
      "step": 30527
    },
    {
      "entropy": 0.4893840057775378,
      "epoch": 2.808736004968344,
      "grad_norm": 0.15427544713020325,
      "learning_rate": 6.378998374582146e-06,
      "loss": 0.4781,
      "mean_token_accuracy": 0.8475771397352219,
      "num_tokens": 974320898.0,
      "step": 30528
    },
    {
      "entropy": 0.5080059226602316,
      "epoch": 2.808828011339785,
      "grad_norm": 0.16039161384105682,
      "learning_rate": 6.37593154844052e-06,
      "loss": 0.497,
      "mean_token_accuracy": 0.8431824035942554,
      "num_tokens": 974352020.0,
      "step": 30529
    },
    {
      "entropy": 0.5967588741332293,
      "epoch": 2.8089200177112263,
      "grad_norm": 0.1749284714460373,
      "learning_rate": 6.3728647222988935e-06,
      "loss": 0.5799,
      "mean_token_accuracy": 0.8170932345092297,
      "num_tokens": 974383814.0,
      "step": 30530
    },
    {
      "entropy": 0.5324973645620048,
      "epoch": 2.8090120240826675,
      "grad_norm": 0.15994790196418762,
      "learning_rate": 6.369797896157267e-06,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8364770151674747,
      "num_tokens": 974415679.0,
      "step": 30531
    },
    {
      "entropy": 0.6299627907574177,
      "epoch": 2.809104030454109,
      "grad_norm": 0.17419545352458954,
      "learning_rate": 6.366731070015641e-06,
      "loss": 0.6236,
      "mean_token_accuracy": 0.8079745918512344,
      "num_tokens": 974447526.0,
      "step": 30532
    },
    {
      "entropy": 0.5057573450030759,
      "epoch": 2.80919603682555,
      "grad_norm": 0.15170718729496002,
      "learning_rate": 6.363664243874015e-06,
      "loss": 0.4869,
      "mean_token_accuracy": 0.8442255295813084,
      "num_tokens": 974479742.0,
      "step": 30533
    },
    {
      "entropy": 0.5682681584730744,
      "epoch": 2.8092880431969913,
      "grad_norm": 0.16579660773277283,
      "learning_rate": 6.360597417732389e-06,
      "loss": 0.5568,
      "mean_token_accuracy": 0.8246041275560856,
      "num_tokens": 974512081.0,
      "step": 30534
    },
    {
      "entropy": 0.5191947878338397,
      "epoch": 2.8093800495684325,
      "grad_norm": 0.15461841225624084,
      "learning_rate": 6.3575305915907625e-06,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8396650366485119,
      "num_tokens": 974544556.0,
      "step": 30535
    },
    {
      "entropy": 0.47446395037695765,
      "epoch": 2.8094720559398736,
      "grad_norm": 0.15822021663188934,
      "learning_rate": 6.354463765449137e-06,
      "loss": 0.4606,
      "mean_token_accuracy": 0.8541994765400887,
      "num_tokens": 974575383.0,
      "step": 30536
    },
    {
      "entropy": 0.3824410680681467,
      "epoch": 2.809564062311315,
      "grad_norm": 0.13534924387931824,
      "learning_rate": 6.351396939307511e-06,
      "loss": 0.3648,
      "mean_token_accuracy": 0.8840357251465321,
      "num_tokens": 974607194.0,
      "step": 30537
    },
    {
      "entropy": 0.6272602025419474,
      "epoch": 2.8096560686827563,
      "grad_norm": 0.17090007662773132,
      "learning_rate": 6.348330113165885e-06,
      "loss": 0.6237,
      "mean_token_accuracy": 0.8067593760788441,
      "num_tokens": 974639525.0,
      "step": 30538
    },
    {
      "entropy": 0.41993342339992523,
      "epoch": 2.8097480750541974,
      "grad_norm": 0.1471415013074875,
      "learning_rate": 6.3452632870242586e-06,
      "loss": 0.4111,
      "mean_token_accuracy": 0.867431029677391,
      "num_tokens": 974671800.0,
      "step": 30539
    },
    {
      "entropy": 0.5096746571362019,
      "epoch": 2.8098400814256386,
      "grad_norm": 0.16002404689788818,
      "learning_rate": 6.342196460882632e-06,
      "loss": 0.4963,
      "mean_token_accuracy": 0.8392217345535755,
      "num_tokens": 974703468.0,
      "step": 30540
    },
    {
      "entropy": 0.4092093533836305,
      "epoch": 2.8099320877970797,
      "grad_norm": 0.14510340988636017,
      "learning_rate": 6.339129634741006e-06,
      "loss": 0.3936,
      "mean_token_accuracy": 0.8721749931573868,
      "num_tokens": 974735321.0,
      "step": 30541
    },
    {
      "entropy": 0.4826081623323262,
      "epoch": 2.8100240941685213,
      "grad_norm": 0.15498802065849304,
      "learning_rate": 6.336062808599382e-06,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8475674688816071,
      "num_tokens": 974767745.0,
      "step": 30542
    },
    {
      "entropy": 0.47660670429468155,
      "epoch": 2.8101161005399624,
      "grad_norm": 0.15716411173343658,
      "learning_rate": 6.3329959824577555e-06,
      "loss": 0.4752,
      "mean_token_accuracy": 0.8513933680951595,
      "num_tokens": 974799541.0,
      "step": 30543
    },
    {
      "entropy": 0.5275083808228374,
      "epoch": 2.8102081069114035,
      "grad_norm": 0.16032074391841888,
      "learning_rate": 6.329929156316129e-06,
      "loss": 0.5229,
      "mean_token_accuracy": 0.8340963162481785,
      "num_tokens": 974831523.0,
      "step": 30544
    },
    {
      "entropy": 0.5268382877111435,
      "epoch": 2.8103001132828447,
      "grad_norm": 0.1618504375219345,
      "learning_rate": 6.326862330174503e-06,
      "loss": 0.525,
      "mean_token_accuracy": 0.8365030400454998,
      "num_tokens": 974863795.0,
      "step": 30545
    },
    {
      "entropy": 0.48919318709522486,
      "epoch": 2.810392119654286,
      "grad_norm": 0.15361575782299042,
      "learning_rate": 6.323795504032877e-06,
      "loss": 0.4766,
      "mean_token_accuracy": 0.84586276486516,
      "num_tokens": 974896108.0,
      "step": 30546
    },
    {
      "entropy": 0.45609473600052297,
      "epoch": 2.8104841260257274,
      "grad_norm": 0.15118886530399323,
      "learning_rate": 6.320728677891251e-06,
      "loss": 0.4381,
      "mean_token_accuracy": 0.8567127250134945,
      "num_tokens": 974927626.0,
      "step": 30547
    },
    {
      "entropy": 0.5528857354074717,
      "epoch": 2.8105761323971685,
      "grad_norm": 0.1628599315881729,
      "learning_rate": 6.3176618517496245e-06,
      "loss": 0.5416,
      "mean_token_accuracy": 0.8248345851898193,
      "num_tokens": 974959650.0,
      "step": 30548
    },
    {
      "entropy": 0.5465880837291479,
      "epoch": 2.8106681387686097,
      "grad_norm": 0.16052532196044922,
      "learning_rate": 6.314595025607998e-06,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8306493796408176,
      "num_tokens": 974991876.0,
      "step": 30549
    },
    {
      "entropy": 0.5224515339359641,
      "epoch": 2.810760145140051,
      "grad_norm": 0.1626156121492386,
      "learning_rate": 6.311528199466372e-06,
      "loss": 0.5059,
      "mean_token_accuracy": 0.8403061889111996,
      "num_tokens": 975023224.0,
      "step": 30550
    },
    {
      "entropy": 0.520873511210084,
      "epoch": 2.810852151511492,
      "grad_norm": 0.16045694053173065,
      "learning_rate": 6.308461373324747e-06,
      "loss": 0.4969,
      "mean_token_accuracy": 0.839028388261795,
      "num_tokens": 975055035.0,
      "step": 30551
    },
    {
      "entropy": 0.5976796858012676,
      "epoch": 2.8109441578829335,
      "grad_norm": 0.17149606347084045,
      "learning_rate": 6.3053945471831205e-06,
      "loss": 0.5834,
      "mean_token_accuracy": 0.8170138038694859,
      "num_tokens": 975086750.0,
      "step": 30552
    },
    {
      "entropy": 0.577871510758996,
      "epoch": 2.8110361642543746,
      "grad_norm": 0.1686127930879593,
      "learning_rate": 6.302327721041494e-06,
      "loss": 0.5625,
      "mean_token_accuracy": 0.8229619599878788,
      "num_tokens": 975117630.0,
      "step": 30553
    },
    {
      "entropy": 0.59995355643332,
      "epoch": 2.8111281706258158,
      "grad_norm": 0.17028404772281647,
      "learning_rate": 6.299260894899868e-06,
      "loss": 0.5875,
      "mean_token_accuracy": 0.8160918541252613,
      "num_tokens": 975149764.0,
      "step": 30554
    },
    {
      "entropy": 0.49974226858466864,
      "epoch": 2.811220176997257,
      "grad_norm": 0.15034693479537964,
      "learning_rate": 6.296194068758242e-06,
      "loss": 0.4807,
      "mean_token_accuracy": 0.8449908867478371,
      "num_tokens": 975181908.0,
      "step": 30555
    },
    {
      "entropy": 0.4952424429357052,
      "epoch": 2.811312183368698,
      "grad_norm": 0.15825584530830383,
      "learning_rate": 6.293127242616616e-06,
      "loss": 0.478,
      "mean_token_accuracy": 0.8482599146664143,
      "num_tokens": 975213523.0,
      "step": 30556
    },
    {
      "entropy": 0.5044479952193797,
      "epoch": 2.8114041897401396,
      "grad_norm": 0.158309668302536,
      "learning_rate": 6.290060416474991e-06,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8419233746826649,
      "num_tokens": 975245222.0,
      "step": 30557
    },
    {
      "entropy": 0.3638612092472613,
      "epoch": 2.8114961961115807,
      "grad_norm": 0.1405401974916458,
      "learning_rate": 6.286993590333365e-06,
      "loss": 0.3575,
      "mean_token_accuracy": 0.8830577731132507,
      "num_tokens": 975276892.0,
      "step": 30558
    },
    {
      "entropy": 0.421908188611269,
      "epoch": 2.811588202483022,
      "grad_norm": 0.146645188331604,
      "learning_rate": 6.283926764191739e-06,
      "loss": 0.4107,
      "mean_token_accuracy": 0.8704155050218105,
      "num_tokens": 975308969.0,
      "step": 30559
    },
    {
      "entropy": 0.45622160751372576,
      "epoch": 2.811680208854463,
      "grad_norm": 0.14992298185825348,
      "learning_rate": 6.280859938050113e-06,
      "loss": 0.44,
      "mean_token_accuracy": 0.8607391715049744,
      "num_tokens": 975341475.0,
      "step": 30560
    },
    {
      "entropy": 0.5847187796607614,
      "epoch": 2.811772215225904,
      "grad_norm": 0.16999424993991852,
      "learning_rate": 6.277793111908486e-06,
      "loss": 0.5711,
      "mean_token_accuracy": 0.8204625174403191,
      "num_tokens": 975373824.0,
      "step": 30561
    },
    {
      "entropy": 0.5357662634924054,
      "epoch": 2.8118642215973457,
      "grad_norm": 0.16248150169849396,
      "learning_rate": 6.27472628576686e-06,
      "loss": 0.5312,
      "mean_token_accuracy": 0.830376498401165,
      "num_tokens": 975405652.0,
      "step": 30562
    },
    {
      "entropy": 0.5876124799251556,
      "epoch": 2.811956227968787,
      "grad_norm": 0.1673559993505478,
      "learning_rate": 6.271659459625234e-06,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8236293271183968,
      "num_tokens": 975436842.0,
      "step": 30563
    },
    {
      "entropy": 0.6007164511829615,
      "epoch": 2.812048234340228,
      "grad_norm": 0.16999749839305878,
      "learning_rate": 6.268592633483608e-06,
      "loss": 0.5846,
      "mean_token_accuracy": 0.8170763999223709,
      "num_tokens": 975468933.0,
      "step": 30564
    },
    {
      "entropy": 0.5476798582822084,
      "epoch": 2.812140240711669,
      "grad_norm": 0.16513966023921967,
      "learning_rate": 6.265525807341982e-06,
      "loss": 0.5326,
      "mean_token_accuracy": 0.8318161331117153,
      "num_tokens": 975499664.0,
      "step": 30565
    },
    {
      "entropy": 0.5733172260224819,
      "epoch": 2.8122322470831103,
      "grad_norm": 0.1642134040594101,
      "learning_rate": 6.262458981200356e-06,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8256812989711761,
      "num_tokens": 975531870.0,
      "step": 30566
    },
    {
      "entropy": 0.5404245764948428,
      "epoch": 2.812324253454552,
      "grad_norm": 0.16019585728645325,
      "learning_rate": 6.25939215505873e-06,
      "loss": 0.5265,
      "mean_token_accuracy": 0.8359227105975151,
      "num_tokens": 975563596.0,
      "step": 30567
    },
    {
      "entropy": 0.5448697226820514,
      "epoch": 2.812416259825993,
      "grad_norm": 0.16039006412029266,
      "learning_rate": 6.256325328917104e-06,
      "loss": 0.5408,
      "mean_token_accuracy": 0.8341628052294254,
      "num_tokens": 975596286.0,
      "step": 30568
    },
    {
      "entropy": 0.529661450535059,
      "epoch": 2.812508266197434,
      "grad_norm": 0.15533694624900818,
      "learning_rate": 6.253258502775478e-06,
      "loss": 0.5262,
      "mean_token_accuracy": 0.836797684431076,
      "num_tokens": 975628689.0,
      "step": 30569
    },
    {
      "entropy": 0.5740058501251042,
      "epoch": 2.8126002725688752,
      "grad_norm": 0.16604425013065338,
      "learning_rate": 6.2501916766338515e-06,
      "loss": 0.5622,
      "mean_token_accuracy": 0.8226158916950226,
      "num_tokens": 975659945.0,
      "step": 30570
    },
    {
      "entropy": 0.48708884743973613,
      "epoch": 2.8126922789403164,
      "grad_norm": 0.153049498796463,
      "learning_rate": 6.247124850492226e-06,
      "loss": 0.467,
      "mean_token_accuracy": 0.8526181168854237,
      "num_tokens": 975692165.0,
      "step": 30571
    },
    {
      "entropy": 0.5373837625375018,
      "epoch": 2.812784285311758,
      "grad_norm": 0.16174614429473877,
      "learning_rate": 6.2440580243506e-06,
      "loss": 0.5237,
      "mean_token_accuracy": 0.8345158323645592,
      "num_tokens": 975724312.0,
      "step": 30572
    },
    {
      "entropy": 0.5237835487350821,
      "epoch": 2.812876291683199,
      "grad_norm": 0.16366064548492432,
      "learning_rate": 6.240991198208974e-06,
      "loss": 0.5168,
      "mean_token_accuracy": 0.8389166481792927,
      "num_tokens": 975756403.0,
      "step": 30573
    },
    {
      "entropy": 0.5129001159220934,
      "epoch": 2.81296829805464,
      "grad_norm": 0.1582709103822708,
      "learning_rate": 6.2379243720673475e-06,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8433110304176807,
      "num_tokens": 975787523.0,
      "step": 30574
    },
    {
      "entropy": 0.4954087296500802,
      "epoch": 2.8130603044260813,
      "grad_norm": 0.15678392350673676,
      "learning_rate": 6.234857545925721e-06,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8423198014497757,
      "num_tokens": 975820017.0,
      "step": 30575
    },
    {
      "entropy": 0.5541054457426071,
      "epoch": 2.8131523107975225,
      "grad_norm": 0.16090060770511627,
      "learning_rate": 6.231790719784096e-06,
      "loss": 0.5404,
      "mean_token_accuracy": 0.8289679735898972,
      "num_tokens": 975852098.0,
      "step": 30576
    },
    {
      "entropy": 0.5941604059189558,
      "epoch": 2.813244317168964,
      "grad_norm": 0.17034459114074707,
      "learning_rate": 6.22872389364247e-06,
      "loss": 0.5856,
      "mean_token_accuracy": 0.8147741593420506,
      "num_tokens": 975883897.0,
      "step": 30577
    },
    {
      "entropy": 0.4598654934670776,
      "epoch": 2.813336323540405,
      "grad_norm": 0.1435524821281433,
      "learning_rate": 6.2256570675008436e-06,
      "loss": 0.4408,
      "mean_token_accuracy": 0.8618306778371334,
      "num_tokens": 975915340.0,
      "step": 30578
    },
    {
      "entropy": 0.5497800500597805,
      "epoch": 2.8134283299118463,
      "grad_norm": 0.16436651349067688,
      "learning_rate": 6.222590241359217e-06,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8306715004146099,
      "num_tokens": 975946970.0,
      "step": 30579
    },
    {
      "entropy": 0.4540921268053353,
      "epoch": 2.8135203362832875,
      "grad_norm": 0.15405505895614624,
      "learning_rate": 6.219523415217591e-06,
      "loss": 0.4547,
      "mean_token_accuracy": 0.8544642440974712,
      "num_tokens": 975978999.0,
      "step": 30580
    },
    {
      "entropy": 0.5394385904073715,
      "epoch": 2.8136123426547286,
      "grad_norm": 0.16567854583263397,
      "learning_rate": 6.216456589075966e-06,
      "loss": 0.5294,
      "mean_token_accuracy": 0.830023005604744,
      "num_tokens": 976010182.0,
      "step": 30581
    },
    {
      "entropy": 0.5284359073266387,
      "epoch": 2.81370434902617,
      "grad_norm": 0.15754467248916626,
      "learning_rate": 6.21338976293434e-06,
      "loss": 0.5189,
      "mean_token_accuracy": 0.8359335511922836,
      "num_tokens": 976042709.0,
      "step": 30582
    },
    {
      "entropy": 0.531605800613761,
      "epoch": 2.8137963553976113,
      "grad_norm": 0.1636608988046646,
      "learning_rate": 6.210322936792713e-06,
      "loss": 0.5216,
      "mean_token_accuracy": 0.8337466828525066,
      "num_tokens": 976074641.0,
      "step": 30583
    },
    {
      "entropy": 0.6225151065737009,
      "epoch": 2.8138883617690524,
      "grad_norm": 0.17730122804641724,
      "learning_rate": 6.207256110651088e-06,
      "loss": 0.622,
      "mean_token_accuracy": 0.8037230633199215,
      "num_tokens": 976105987.0,
      "step": 30584
    },
    {
      "entropy": 0.5222407476976514,
      "epoch": 2.8139803681404936,
      "grad_norm": 0.15795321762561798,
      "learning_rate": 6.204189284509462e-06,
      "loss": 0.5144,
      "mean_token_accuracy": 0.8375210836529732,
      "num_tokens": 976137806.0,
      "step": 30585
    },
    {
      "entropy": 0.48523419722914696,
      "epoch": 2.8140723745119347,
      "grad_norm": 0.15166863799095154,
      "learning_rate": 6.201122458367836e-06,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8505929037928581,
      "num_tokens": 976169985.0,
      "step": 30586
    },
    {
      "entropy": 0.4768767524510622,
      "epoch": 2.8141643808833763,
      "grad_norm": 0.15969082713127136,
      "learning_rate": 6.1980556322262095e-06,
      "loss": 0.4641,
      "mean_token_accuracy": 0.852387759834528,
      "num_tokens": 976201683.0,
      "step": 30587
    },
    {
      "entropy": 0.3732161740772426,
      "epoch": 2.8142563872548174,
      "grad_norm": 0.14166145026683807,
      "learning_rate": 6.194988806084583e-06,
      "loss": 0.3627,
      "mean_token_accuracy": 0.8836814425885677,
      "num_tokens": 976233271.0,
      "step": 30588
    },
    {
      "entropy": 0.5905681643635035,
      "epoch": 2.8143483936262585,
      "grad_norm": 0.16886158287525177,
      "learning_rate": 6.191921979942957e-06,
      "loss": 0.572,
      "mean_token_accuracy": 0.8231450691819191,
      "num_tokens": 976263953.0,
      "step": 30589
    },
    {
      "entropy": 0.5280918944627047,
      "epoch": 2.8144403999976997,
      "grad_norm": 0.1537124514579773,
      "learning_rate": 6.188855153801331e-06,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8367123007774353,
      "num_tokens": 976296196.0,
      "step": 30590
    },
    {
      "entropy": 0.35394806833937764,
      "epoch": 2.814532406369141,
      "grad_norm": 0.13188716769218445,
      "learning_rate": 6.185788327659705e-06,
      "loss": 0.3388,
      "mean_token_accuracy": 0.8904109112918377,
      "num_tokens": 976328545.0,
      "step": 30591
    },
    {
      "entropy": 0.4816479869186878,
      "epoch": 2.8146244127405824,
      "grad_norm": 0.15249046683311462,
      "learning_rate": 6.182721501518079e-06,
      "loss": 0.476,
      "mean_token_accuracy": 0.8484517000615597,
      "num_tokens": 976360802.0,
      "step": 30592
    },
    {
      "entropy": 0.46905397810041904,
      "epoch": 2.8147164191120235,
      "grad_norm": 0.14950576424598694,
      "learning_rate": 6.179654675376453e-06,
      "loss": 0.4428,
      "mean_token_accuracy": 0.8605567999184132,
      "num_tokens": 976392206.0,
      "step": 30593
    },
    {
      "entropy": 0.5477710049599409,
      "epoch": 2.8148084254834647,
      "grad_norm": 0.16508282721042633,
      "learning_rate": 6.176587849234827e-06,
      "loss": 0.5201,
      "mean_token_accuracy": 0.832053005695343,
      "num_tokens": 976423216.0,
      "step": 30594
    },
    {
      "entropy": 0.47884791996330023,
      "epoch": 2.814900431854906,
      "grad_norm": 0.15202106535434723,
      "learning_rate": 6.173521023093201e-06,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8520701043307781,
      "num_tokens": 976455984.0,
      "step": 30595
    },
    {
      "entropy": 0.5651941522955894,
      "epoch": 2.814992438226347,
      "grad_norm": 0.16482514142990112,
      "learning_rate": 6.170454196951575e-06,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8261648938059807,
      "num_tokens": 976488377.0,
      "step": 30596
    },
    {
      "entropy": 0.5399523265659809,
      "epoch": 2.8150844445977885,
      "grad_norm": 0.163997083902359,
      "learning_rate": 6.167387370809949e-06,
      "loss": 0.5202,
      "mean_token_accuracy": 0.83495057746768,
      "num_tokens": 976519877.0,
      "step": 30597
    },
    {
      "entropy": 0.5108399530872703,
      "epoch": 2.8151764509692296,
      "grad_norm": 0.16301363706588745,
      "learning_rate": 6.164320544668323e-06,
      "loss": 0.5038,
      "mean_token_accuracy": 0.8391782604157925,
      "num_tokens": 976551430.0,
      "step": 30598
    },
    {
      "entropy": 0.48760390223469585,
      "epoch": 2.8152684573406708,
      "grad_norm": 0.15239116549491882,
      "learning_rate": 6.161253718526698e-06,
      "loss": 0.4808,
      "mean_token_accuracy": 0.8485977202653885,
      "num_tokens": 976583659.0,
      "step": 30599
    },
    {
      "entropy": 0.47892950754612684,
      "epoch": 2.815360463712112,
      "grad_norm": 0.15406838059425354,
      "learning_rate": 6.158186892385071e-06,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8516927361488342,
      "num_tokens": 976615564.0,
      "step": 30600
    },
    {
      "entropy": 0.5372488854918629,
      "epoch": 2.815452470083553,
      "grad_norm": 0.16344454884529114,
      "learning_rate": 6.155120066243445e-06,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8362475819885731,
      "num_tokens": 976646151.0,
      "step": 30601
    },
    {
      "entropy": 0.4620950189419091,
      "epoch": 2.8155444764549946,
      "grad_norm": 0.14144816994667053,
      "learning_rate": 6.152053240101819e-06,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8562042266130447,
      "num_tokens": 976678398.0,
      "step": 30602
    },
    {
      "entropy": 0.4115473818965256,
      "epoch": 2.8156364828264357,
      "grad_norm": 0.1469084471464157,
      "learning_rate": 6.148986413960193e-06,
      "loss": 0.4048,
      "mean_token_accuracy": 0.8674778640270233,
      "num_tokens": 976710482.0,
      "step": 30603
    },
    {
      "entropy": 0.5700972899794579,
      "epoch": 2.815728489197877,
      "grad_norm": 0.16296955943107605,
      "learning_rate": 6.145919587818567e-06,
      "loss": 0.5642,
      "mean_token_accuracy": 0.824481438845396,
      "num_tokens": 976742954.0,
      "step": 30604
    },
    {
      "entropy": 0.4053597906604409,
      "epoch": 2.815820495569318,
      "grad_norm": 0.14520293474197388,
      "learning_rate": 6.14285276167694e-06,
      "loss": 0.3951,
      "mean_token_accuracy": 0.8740951381623745,
      "num_tokens": 976775190.0,
      "step": 30605
    },
    {
      "entropy": 0.49051460553891957,
      "epoch": 2.815912501940759,
      "grad_norm": 0.15284942090511322,
      "learning_rate": 6.139785935535314e-06,
      "loss": 0.483,
      "mean_token_accuracy": 0.8497637175023556,
      "num_tokens": 976807194.0,
      "step": 30606
    },
    {
      "entropy": 0.43303670873865485,
      "epoch": 2.8160045083122007,
      "grad_norm": 0.1499023735523224,
      "learning_rate": 6.136719109393689e-06,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8633654564619064,
      "num_tokens": 976839783.0,
      "step": 30607
    },
    {
      "entropy": 0.5602650297805667,
      "epoch": 2.816096514683642,
      "grad_norm": 0.16833338141441345,
      "learning_rate": 6.133652283252063e-06,
      "loss": 0.5517,
      "mean_token_accuracy": 0.8242268338799477,
      "num_tokens": 976871950.0,
      "step": 30608
    },
    {
      "entropy": 0.4998796796426177,
      "epoch": 2.816188521055083,
      "grad_norm": 0.15484003722667694,
      "learning_rate": 6.1305854571104365e-06,
      "loss": 0.4952,
      "mean_token_accuracy": 0.841492772102356,
      "num_tokens": 976904236.0,
      "step": 30609
    },
    {
      "entropy": 0.5057190358638763,
      "epoch": 2.816280527426524,
      "grad_norm": 0.1583196520805359,
      "learning_rate": 6.12751863096881e-06,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8444086760282516,
      "num_tokens": 976936398.0,
      "step": 30610
    },
    {
      "entropy": 0.47754121106117964,
      "epoch": 2.8163725337979653,
      "grad_norm": 0.15268497169017792,
      "learning_rate": 6.124451804827185e-06,
      "loss": 0.463,
      "mean_token_accuracy": 0.8491850644350052,
      "num_tokens": 976968383.0,
      "step": 30611
    },
    {
      "entropy": 0.4804488494992256,
      "epoch": 2.816464540169407,
      "grad_norm": 0.1499428153038025,
      "learning_rate": 6.121384978685559e-06,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8498909436166286,
      "num_tokens": 976999939.0,
      "step": 30612
    },
    {
      "entropy": 0.528547041118145,
      "epoch": 2.816556546540848,
      "grad_norm": 0.16160105168819427,
      "learning_rate": 6.1183181525439325e-06,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8347940519452095,
      "num_tokens": 977032163.0,
      "step": 30613
    },
    {
      "entropy": 0.5593496770597994,
      "epoch": 2.816648552912289,
      "grad_norm": 0.1627102643251419,
      "learning_rate": 6.115251326402307e-06,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8308653645217419,
      "num_tokens": 977064044.0,
      "step": 30614
    },
    {
      "entropy": 0.4850474721752107,
      "epoch": 2.8167405592837302,
      "grad_norm": 0.1571303755044937,
      "learning_rate": 6.112184500260681e-06,
      "loss": 0.4624,
      "mean_token_accuracy": 0.8520283810794353,
      "num_tokens": 977095271.0,
      "step": 30615
    },
    {
      "entropy": 0.5488944035023451,
      "epoch": 2.8168325656551714,
      "grad_norm": 0.16232644021511078,
      "learning_rate": 6.109117674119055e-06,
      "loss": 0.5353,
      "mean_token_accuracy": 0.8305374719202518,
      "num_tokens": 977126575.0,
      "step": 30616
    },
    {
      "entropy": 0.4830475226044655,
      "epoch": 2.816924572026613,
      "grad_norm": 0.14927938580513,
      "learning_rate": 6.1060508479774286e-06,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8494472242891788,
      "num_tokens": 977158561.0,
      "step": 30617
    },
    {
      "entropy": 0.4679707339964807,
      "epoch": 2.817016578398054,
      "grad_norm": 0.15289753675460815,
      "learning_rate": 6.102984021835802e-06,
      "loss": 0.4596,
      "mean_token_accuracy": 0.8529548645019531,
      "num_tokens": 977190403.0,
      "step": 30618
    },
    {
      "entropy": 0.49682356184348464,
      "epoch": 2.817108584769495,
      "grad_norm": 0.15474095940589905,
      "learning_rate": 6.099917195694176e-06,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8478069715201855,
      "num_tokens": 977221785.0,
      "step": 30619
    },
    {
      "entropy": 0.5313549304846674,
      "epoch": 2.8172005911409363,
      "grad_norm": 0.15935954451560974,
      "learning_rate": 6.09685036955255e-06,
      "loss": 0.526,
      "mean_token_accuracy": 0.8359923958778381,
      "num_tokens": 977254316.0,
      "step": 30620
    },
    {
      "entropy": 0.4055721468757838,
      "epoch": 2.8172925975123775,
      "grad_norm": 0.14211955666542053,
      "learning_rate": 6.093783543410924e-06,
      "loss": 0.3939,
      "mean_token_accuracy": 0.8727536983788013,
      "num_tokens": 977286108.0,
      "step": 30621
    },
    {
      "entropy": 0.5418731113895774,
      "epoch": 2.817384603883819,
      "grad_norm": 0.1685665100812912,
      "learning_rate": 6.090716717269298e-06,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8309857584536076,
      "num_tokens": 977318339.0,
      "step": 30622
    },
    {
      "entropy": 0.5756905619055033,
      "epoch": 2.81747661025526,
      "grad_norm": 0.16749845445156097,
      "learning_rate": 6.087649891127672e-06,
      "loss": 0.5693,
      "mean_token_accuracy": 0.8198658637702465,
      "num_tokens": 977350049.0,
      "step": 30623
    },
    {
      "entropy": 0.484671363607049,
      "epoch": 2.8175686166267013,
      "grad_norm": 0.15778625011444092,
      "learning_rate": 6.084583064986046e-06,
      "loss": 0.4738,
      "mean_token_accuracy": 0.847363792359829,
      "num_tokens": 977381751.0,
      "step": 30624
    },
    {
      "entropy": 0.5980933830142021,
      "epoch": 2.8176606229981425,
      "grad_norm": 0.1707300990819931,
      "learning_rate": 6.081516238844421e-06,
      "loss": 0.5873,
      "mean_token_accuracy": 0.81221654266119,
      "num_tokens": 977413197.0,
      "step": 30625
    },
    {
      "entropy": 0.45952247735112906,
      "epoch": 2.8177526293695836,
      "grad_norm": 0.1519535481929779,
      "learning_rate": 6.0784494127027945e-06,
      "loss": 0.444,
      "mean_token_accuracy": 0.8573103845119476,
      "num_tokens": 977445591.0,
      "step": 30626
    },
    {
      "entropy": 0.4743762742727995,
      "epoch": 2.817844635741025,
      "grad_norm": 0.15159961581230164,
      "learning_rate": 6.075382586561168e-06,
      "loss": 0.453,
      "mean_token_accuracy": 0.8539119213819504,
      "num_tokens": 977477389.0,
      "step": 30627
    },
    {
      "entropy": 0.5430637616664171,
      "epoch": 2.8179366421124663,
      "grad_norm": 0.16340947151184082,
      "learning_rate": 6.072315760419542e-06,
      "loss": 0.5265,
      "mean_token_accuracy": 0.8314450532197952,
      "num_tokens": 977509380.0,
      "step": 30628
    },
    {
      "entropy": 0.5582068469375372,
      "epoch": 2.8180286484839074,
      "grad_norm": 0.166301891207695,
      "learning_rate": 6.069248934277916e-06,
      "loss": 0.5548,
      "mean_token_accuracy": 0.8236098885536194,
      "num_tokens": 977541761.0,
      "step": 30629
    },
    {
      "entropy": 0.5260424492880702,
      "epoch": 2.8181206548553486,
      "grad_norm": 0.16500163078308105,
      "learning_rate": 6.0661821081362905e-06,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8361476473510265,
      "num_tokens": 977573343.0,
      "step": 30630
    },
    {
      "entropy": 0.4140526270493865,
      "epoch": 2.8182126612267897,
      "grad_norm": 0.14589501917362213,
      "learning_rate": 6.063115281994664e-06,
      "loss": 0.4065,
      "mean_token_accuracy": 0.8705046884715557,
      "num_tokens": 977605643.0,
      "step": 30631
    },
    {
      "entropy": 0.5593183445744216,
      "epoch": 2.8183046675982313,
      "grad_norm": 0.1625441461801529,
      "learning_rate": 6.060048455853038e-06,
      "loss": 0.5468,
      "mean_token_accuracy": 0.8266198039054871,
      "num_tokens": 977637444.0,
      "step": 30632
    },
    {
      "entropy": 0.45993457781150937,
      "epoch": 2.8183966739696724,
      "grad_norm": 0.14925260841846466,
      "learning_rate": 6.056981629711412e-06,
      "loss": 0.4558,
      "mean_token_accuracy": 0.854380551725626,
      "num_tokens": 977669597.0,
      "step": 30633
    },
    {
      "entropy": 0.4902703892439604,
      "epoch": 2.8184886803411136,
      "grad_norm": 0.15527135133743286,
      "learning_rate": 6.053914803569786e-06,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8498076163232327,
      "num_tokens": 977702185.0,
      "step": 30634
    },
    {
      "entropy": 0.5540941786020994,
      "epoch": 2.8185806867125547,
      "grad_norm": 0.16514402627944946,
      "learning_rate": 6.0508479774281595e-06,
      "loss": 0.5421,
      "mean_token_accuracy": 0.82966398447752,
      "num_tokens": 977734522.0,
      "step": 30635
    },
    {
      "entropy": 0.39599848352372646,
      "epoch": 2.818672693083996,
      "grad_norm": 0.14575470983982086,
      "learning_rate": 6.047781151286533e-06,
      "loss": 0.384,
      "mean_token_accuracy": 0.877319198101759,
      "num_tokens": 977766129.0,
      "step": 30636
    },
    {
      "entropy": 0.5458386926911771,
      "epoch": 2.8187646994554374,
      "grad_norm": 0.16284386813640594,
      "learning_rate": 6.044714325144908e-06,
      "loss": 0.5352,
      "mean_token_accuracy": 0.8312662653625011,
      "num_tokens": 977798037.0,
      "step": 30637
    },
    {
      "entropy": 0.45873565319925547,
      "epoch": 2.8188567058268785,
      "grad_norm": 0.15293405950069427,
      "learning_rate": 6.041647499003282e-06,
      "loss": 0.4527,
      "mean_token_accuracy": 0.8547204807400703,
      "num_tokens": 977829693.0,
      "step": 30638
    },
    {
      "entropy": 0.5577714201062918,
      "epoch": 2.8189487121983197,
      "grad_norm": 0.16721096634864807,
      "learning_rate": 6.0385806728616556e-06,
      "loss": 0.5485,
      "mean_token_accuracy": 0.8234479129314423,
      "num_tokens": 977861141.0,
      "step": 30639
    },
    {
      "entropy": 0.4087636498734355,
      "epoch": 2.819040718569761,
      "grad_norm": 0.1476532369852066,
      "learning_rate": 6.03551384672003e-06,
      "loss": 0.4037,
      "mean_token_accuracy": 0.8695716038346291,
      "num_tokens": 977893241.0,
      "step": 30640
    },
    {
      "entropy": 0.49195212777704,
      "epoch": 2.819132724941202,
      "grad_norm": 0.15762200951576233,
      "learning_rate": 6.032447020578404e-06,
      "loss": 0.4905,
      "mean_token_accuracy": 0.8480879478156567,
      "num_tokens": 977924926.0,
      "step": 30641
    },
    {
      "entropy": 0.559745498932898,
      "epoch": 2.8192247313126435,
      "grad_norm": 0.17023922502994537,
      "learning_rate": 6.029380194436778e-06,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8248878605663776,
      "num_tokens": 977956302.0,
      "step": 30642
    },
    {
      "entropy": 0.5832675248384476,
      "epoch": 2.8193167376840846,
      "grad_norm": 0.16549210250377655,
      "learning_rate": 6.026313368295152e-06,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8206571564078331,
      "num_tokens": 977988596.0,
      "step": 30643
    },
    {
      "entropy": 0.48908940656110644,
      "epoch": 2.8194087440555258,
      "grad_norm": 0.1568450778722763,
      "learning_rate": 6.023246542153525e-06,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8484173342585564,
      "num_tokens": 978020456.0,
      "step": 30644
    },
    {
      "entropy": 0.5093068955466151,
      "epoch": 2.819500750426967,
      "grad_norm": 0.16369087994098663,
      "learning_rate": 6.0201797160119e-06,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8423797972500324,
      "num_tokens": 978051607.0,
      "step": 30645
    },
    {
      "entropy": 0.46174316853284836,
      "epoch": 2.819592756798408,
      "grad_norm": 0.15093007683753967,
      "learning_rate": 6.017112889870274e-06,
      "loss": 0.4305,
      "mean_token_accuracy": 0.8596860691905022,
      "num_tokens": 978082978.0,
      "step": 30646
    },
    {
      "entropy": 0.5529978442937136,
      "epoch": 2.8196847631698496,
      "grad_norm": 0.16279812157154083,
      "learning_rate": 6.014046063728648e-06,
      "loss": 0.5488,
      "mean_token_accuracy": 0.8297162689268589,
      "num_tokens": 978115124.0,
      "step": 30647
    },
    {
      "entropy": 0.5570217184722424,
      "epoch": 2.8197767695412908,
      "grad_norm": 0.16482499241828918,
      "learning_rate": 6.0109792375870215e-06,
      "loss": 0.5341,
      "mean_token_accuracy": 0.8315592706203461,
      "num_tokens": 978146734.0,
      "step": 30648
    },
    {
      "entropy": 0.458372121443972,
      "epoch": 2.819868775912732,
      "grad_norm": 0.14857804775238037,
      "learning_rate": 6.007912411445395e-06,
      "loss": 0.4434,
      "mean_token_accuracy": 0.8577955849468708,
      "num_tokens": 978179051.0,
      "step": 30649
    },
    {
      "entropy": 0.5656350441277027,
      "epoch": 2.819960782284173,
      "grad_norm": 0.16013361513614655,
      "learning_rate": 6.004845585303769e-06,
      "loss": 0.5528,
      "mean_token_accuracy": 0.826419934630394,
      "num_tokens": 978211459.0,
      "step": 30650
    },
    {
      "entropy": 0.4806794570758939,
      "epoch": 2.820052788655614,
      "grad_norm": 0.15092401206493378,
      "learning_rate": 6.001778759162143e-06,
      "loss": 0.4584,
      "mean_token_accuracy": 0.852159459143877,
      "num_tokens": 978242864.0,
      "step": 30651
    },
    {
      "entropy": 0.5691798310726881,
      "epoch": 2.8201447950270557,
      "grad_norm": 0.1737612783908844,
      "learning_rate": 5.998711933020517e-06,
      "loss": 0.5608,
      "mean_token_accuracy": 0.8215178735554218,
      "num_tokens": 978274719.0,
      "step": 30652
    },
    {
      "entropy": 0.5629078503698111,
      "epoch": 2.820236801398497,
      "grad_norm": 0.16663892567157745,
      "learning_rate": 5.995645106878891e-06,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8264601044356823,
      "num_tokens": 978306452.0,
      "step": 30653
    },
    {
      "entropy": 0.4681504787877202,
      "epoch": 2.820328807769938,
      "grad_norm": 0.1508292406797409,
      "learning_rate": 5.992578280737265e-06,
      "loss": 0.4602,
      "mean_token_accuracy": 0.8528414778411388,
      "num_tokens": 978339184.0,
      "step": 30654
    },
    {
      "entropy": 0.44807169679552317,
      "epoch": 2.820420814141379,
      "grad_norm": 0.1519341617822647,
      "learning_rate": 5.98951145459564e-06,
      "loss": 0.4236,
      "mean_token_accuracy": 0.8615587018430233,
      "num_tokens": 978370662.0,
      "step": 30655
    },
    {
      "entropy": 0.4535581097006798,
      "epoch": 2.8205128205128203,
      "grad_norm": 0.15608444809913635,
      "learning_rate": 5.9864446284540136e-06,
      "loss": 0.4462,
      "mean_token_accuracy": 0.8545732647180557,
      "num_tokens": 978402763.0,
      "step": 30656
    },
    {
      "entropy": 0.4843961736187339,
      "epoch": 2.820604826884262,
      "grad_norm": 0.1524607539176941,
      "learning_rate": 5.983377802312387e-06,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8516100198030472,
      "num_tokens": 978434862.0,
      "step": 30657
    },
    {
      "entropy": 0.4768097852356732,
      "epoch": 2.820696833255703,
      "grad_norm": 0.15593256056308746,
      "learning_rate": 5.980310976170761e-06,
      "loss": 0.4668,
      "mean_token_accuracy": 0.8507202230393887,
      "num_tokens": 978466913.0,
      "step": 30658
    },
    {
      "entropy": 0.5530218966305256,
      "epoch": 2.820788839627144,
      "grad_norm": 0.1661440134048462,
      "learning_rate": 5.977244150029135e-06,
      "loss": 0.539,
      "mean_token_accuracy": 0.8295857571065426,
      "num_tokens": 978497993.0,
      "step": 30659
    },
    {
      "entropy": 0.49552164878696203,
      "epoch": 2.8208808459985852,
      "grad_norm": 0.16452491283416748,
      "learning_rate": 5.97417732388751e-06,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8442538417875767,
      "num_tokens": 978529387.0,
      "step": 30660
    },
    {
      "entropy": 0.49210032168775797,
      "epoch": 2.8209728523700264,
      "grad_norm": 0.15056224167346954,
      "learning_rate": 5.971110497745883e-06,
      "loss": 0.4805,
      "mean_token_accuracy": 0.8469334878027439,
      "num_tokens": 978561714.0,
      "step": 30661
    },
    {
      "entropy": 0.45718653686344624,
      "epoch": 2.821064858741468,
      "grad_norm": 0.15081390738487244,
      "learning_rate": 5.968043671604257e-06,
      "loss": 0.4466,
      "mean_token_accuracy": 0.8584586009383202,
      "num_tokens": 978593653.0,
      "step": 30662
    },
    {
      "entropy": 0.48523078951984644,
      "epoch": 2.821156865112909,
      "grad_norm": 0.15714320540428162,
      "learning_rate": 5.964976845462631e-06,
      "loss": 0.4759,
      "mean_token_accuracy": 0.846836507320404,
      "num_tokens": 978626158.0,
      "step": 30663
    },
    {
      "entropy": 0.4891779227182269,
      "epoch": 2.82124887148435,
      "grad_norm": 0.151772141456604,
      "learning_rate": 5.961910019321005e-06,
      "loss": 0.475,
      "mean_token_accuracy": 0.849338211119175,
      "num_tokens": 978658467.0,
      "step": 30664
    },
    {
      "entropy": 0.45531059242784977,
      "epoch": 2.8213408778557914,
      "grad_norm": 0.1504693478345871,
      "learning_rate": 5.958843193179379e-06,
      "loss": 0.4516,
      "mean_token_accuracy": 0.8567344881594181,
      "num_tokens": 978690543.0,
      "step": 30665
    },
    {
      "entropy": 0.49697323562577367,
      "epoch": 2.8214328842272325,
      "grad_norm": 0.15830782055854797,
      "learning_rate": 5.955776367037752e-06,
      "loss": 0.4904,
      "mean_token_accuracy": 0.8482373580336571,
      "num_tokens": 978722401.0,
      "step": 30666
    },
    {
      "entropy": 0.5956569365225732,
      "epoch": 2.821524890598674,
      "grad_norm": 0.16733293235301971,
      "learning_rate": 5.952709540896126e-06,
      "loss": 0.5688,
      "mean_token_accuracy": 0.8145471140742302,
      "num_tokens": 978754092.0,
      "step": 30667
    },
    {
      "entropy": 0.4350906666368246,
      "epoch": 2.821616896970115,
      "grad_norm": 0.14930814504623413,
      "learning_rate": 5.949642714754501e-06,
      "loss": 0.4346,
      "mean_token_accuracy": 0.8616169281303883,
      "num_tokens": 978786380.0,
      "step": 30668
    },
    {
      "entropy": 0.4654869828373194,
      "epoch": 2.8217089033415563,
      "grad_norm": 0.15123142302036285,
      "learning_rate": 5.946575888612875e-06,
      "loss": 0.4497,
      "mean_token_accuracy": 0.8546097837388515,
      "num_tokens": 978817452.0,
      "step": 30669
    },
    {
      "entropy": 0.5336575880646706,
      "epoch": 2.8218009097129975,
      "grad_norm": 0.1579148918390274,
      "learning_rate": 5.943509062471249e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.836739644408226,
      "num_tokens": 978849709.0,
      "step": 30670
    },
    {
      "entropy": 0.5384118296205997,
      "epoch": 2.8218929160844386,
      "grad_norm": 0.1677444875240326,
      "learning_rate": 5.940442236329623e-06,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8312722817063332,
      "num_tokens": 978881345.0,
      "step": 30671
    },
    {
      "entropy": 0.5322920456528664,
      "epoch": 2.82198492245588,
      "grad_norm": 0.16749367117881775,
      "learning_rate": 5.937375410187997e-06,
      "loss": 0.5176,
      "mean_token_accuracy": 0.8350643254816532,
      "num_tokens": 978912251.0,
      "step": 30672
    },
    {
      "entropy": 0.4112345587927848,
      "epoch": 2.8220769288273213,
      "grad_norm": 0.14138060808181763,
      "learning_rate": 5.934308584046371e-06,
      "loss": 0.3908,
      "mean_token_accuracy": 0.8753745630383492,
      "num_tokens": 978944298.0,
      "step": 30673
    },
    {
      "entropy": 0.5393845830112696,
      "epoch": 2.8221689351987624,
      "grad_norm": 0.1641545593738556,
      "learning_rate": 5.9312417579047445e-06,
      "loss": 0.5121,
      "mean_token_accuracy": 0.8356085531413555,
      "num_tokens": 978975461.0,
      "step": 30674
    },
    {
      "entropy": 0.45645564375445247,
      "epoch": 2.8222609415702036,
      "grad_norm": 0.15081220865249634,
      "learning_rate": 5.928174931763119e-06,
      "loss": 0.4458,
      "mean_token_accuracy": 0.8568639829754829,
      "num_tokens": 979007923.0,
      "step": 30675
    },
    {
      "entropy": 0.531689414056018,
      "epoch": 2.8223529479416447,
      "grad_norm": 0.16138485074043274,
      "learning_rate": 5.925108105621493e-06,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8343422673642635,
      "num_tokens": 979040234.0,
      "step": 30676
    },
    {
      "entropy": 0.5289607830345631,
      "epoch": 2.8224449543130863,
      "grad_norm": 0.1556061953306198,
      "learning_rate": 5.922041279479867e-06,
      "loss": 0.5185,
      "mean_token_accuracy": 0.8316984623670578,
      "num_tokens": 979072363.0,
      "step": 30677
    },
    {
      "entropy": 0.5056796036660671,
      "epoch": 2.8225369606845274,
      "grad_norm": 0.15761838853359222,
      "learning_rate": 5.9189744533382406e-06,
      "loss": 0.492,
      "mean_token_accuracy": 0.8428798913955688,
      "num_tokens": 979104010.0,
      "step": 30678
    },
    {
      "entropy": 0.6055741814197972,
      "epoch": 2.8226289670559686,
      "grad_norm": 0.17062623798847198,
      "learning_rate": 5.915907627196614e-06,
      "loss": 0.5912,
      "mean_token_accuracy": 0.8162107095122337,
      "num_tokens": 979135707.0,
      "step": 30679
    },
    {
      "entropy": 0.5713395345956087,
      "epoch": 2.8227209734274097,
      "grad_norm": 0.1690729260444641,
      "learning_rate": 5.912840801054988e-06,
      "loss": 0.5748,
      "mean_token_accuracy": 0.8176518157124519,
      "num_tokens": 979167279.0,
      "step": 30680
    },
    {
      "entropy": 0.48549505416303873,
      "epoch": 2.822812979798851,
      "grad_norm": 0.1551162749528885,
      "learning_rate": 5.909773974913362e-06,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8485466502606869,
      "num_tokens": 979199179.0,
      "step": 30681
    },
    {
      "entropy": 0.48792925383895636,
      "epoch": 2.8229049861702924,
      "grad_norm": 0.15174415707588196,
      "learning_rate": 5.906707148771736e-06,
      "loss": 0.4679,
      "mean_token_accuracy": 0.8481421507894993,
      "num_tokens": 979230719.0,
      "step": 30682
    },
    {
      "entropy": 0.4484820021316409,
      "epoch": 2.8229969925417335,
      "grad_norm": 0.14732269942760468,
      "learning_rate": 5.90364032263011e-06,
      "loss": 0.4347,
      "mean_token_accuracy": 0.8600813709199429,
      "num_tokens": 979262722.0,
      "step": 30683
    },
    {
      "entropy": 0.5712294429540634,
      "epoch": 2.8230889989131747,
      "grad_norm": 0.16825266182422638,
      "learning_rate": 5.900573496488484e-06,
      "loss": 0.5684,
      "mean_token_accuracy": 0.8211793601512909,
      "num_tokens": 979295443.0,
      "step": 30684
    },
    {
      "entropy": 0.5309921456500888,
      "epoch": 2.823181005284616,
      "grad_norm": 0.16232801973819733,
      "learning_rate": 5.897506670346859e-06,
      "loss": 0.5179,
      "mean_token_accuracy": 0.8348929733037949,
      "num_tokens": 979327822.0,
      "step": 30685
    },
    {
      "entropy": 0.5215324927121401,
      "epoch": 2.823273011656057,
      "grad_norm": 0.16012616455554962,
      "learning_rate": 5.894439844205233e-06,
      "loss": 0.5116,
      "mean_token_accuracy": 0.8391219116747379,
      "num_tokens": 979359854.0,
      "step": 30686
    },
    {
      "entropy": 0.4457060443237424,
      "epoch": 2.8233650180274985,
      "grad_norm": 0.14728757739067078,
      "learning_rate": 5.8913730180636065e-06,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8615280538797379,
      "num_tokens": 979392549.0,
      "step": 30687
    },
    {
      "entropy": 0.4765072828158736,
      "epoch": 2.8234570243989396,
      "grad_norm": 0.15808729827404022,
      "learning_rate": 5.88830619192198e-06,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8497406505048275,
      "num_tokens": 979424161.0,
      "step": 30688
    },
    {
      "entropy": 0.5553435208275914,
      "epoch": 2.8235490307703808,
      "grad_norm": 0.16468647122383118,
      "learning_rate": 5.885239365780354e-06,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8318768814206123,
      "num_tokens": 979455348.0,
      "step": 30689
    },
    {
      "entropy": 0.5105995927006006,
      "epoch": 2.823641037141822,
      "grad_norm": 0.1550377607345581,
      "learning_rate": 5.882172539638728e-06,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8385362289845943,
      "num_tokens": 979487877.0,
      "step": 30690
    },
    {
      "entropy": 0.45496634766459465,
      "epoch": 2.823733043513263,
      "grad_norm": 0.14992964267730713,
      "learning_rate": 5.8791057134971025e-06,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8586935251951218,
      "num_tokens": 979520175.0,
      "step": 30691
    },
    {
      "entropy": 0.46663113962858915,
      "epoch": 2.8238250498847046,
      "grad_norm": 0.1533048152923584,
      "learning_rate": 5.876038887355476e-06,
      "loss": 0.459,
      "mean_token_accuracy": 0.8565846011042595,
      "num_tokens": 979551611.0,
      "step": 30692
    },
    {
      "entropy": 0.3910221103578806,
      "epoch": 2.8239170562561458,
      "grad_norm": 0.1443222016096115,
      "learning_rate": 5.87297206121385e-06,
      "loss": 0.3657,
      "mean_token_accuracy": 0.8778267987072468,
      "num_tokens": 979582566.0,
      "step": 30693
    },
    {
      "entropy": 0.5123604238033295,
      "epoch": 2.824009062627587,
      "grad_norm": 0.16390658915042877,
      "learning_rate": 5.869905235072224e-06,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8394579701125622,
      "num_tokens": 979614496.0,
      "step": 30694
    },
    {
      "entropy": 0.49132039677351713,
      "epoch": 2.824101068999028,
      "grad_norm": 0.15964899957180023,
      "learning_rate": 5.866838408930598e-06,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8464353196322918,
      "num_tokens": 979646079.0,
      "step": 30695
    },
    {
      "entropy": 0.4842971384059638,
      "epoch": 2.824193075370469,
      "grad_norm": 0.15781472623348236,
      "learning_rate": 5.8637715827889715e-06,
      "loss": 0.48,
      "mean_token_accuracy": 0.8479972220957279,
      "num_tokens": 979677858.0,
      "step": 30696
    },
    {
      "entropy": 0.47871224395930767,
      "epoch": 2.8242850817419107,
      "grad_norm": 0.15414239466190338,
      "learning_rate": 5.860704756647345e-06,
      "loss": 0.4634,
      "mean_token_accuracy": 0.8513352684676647,
      "num_tokens": 979710382.0,
      "step": 30697
    },
    {
      "entropy": 0.40267108799889684,
      "epoch": 2.824377088113352,
      "grad_norm": 0.14055156707763672,
      "learning_rate": 5.85763793050572e-06,
      "loss": 0.3873,
      "mean_token_accuracy": 0.8752901628613472,
      "num_tokens": 979742314.0,
      "step": 30698
    },
    {
      "entropy": 0.46483452804386616,
      "epoch": 2.824469094484793,
      "grad_norm": 0.1485041081905365,
      "learning_rate": 5.854571104364094e-06,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8561561703681946,
      "num_tokens": 979774268.0,
      "step": 30699
    },
    {
      "entropy": 0.5460274163633585,
      "epoch": 2.824561100856234,
      "grad_norm": 0.16521137952804565,
      "learning_rate": 5.851504278222468e-06,
      "loss": 0.537,
      "mean_token_accuracy": 0.8278980478644371,
      "num_tokens": 979806555.0,
      "step": 30700
    },
    {
      "entropy": 0.4754151524975896,
      "epoch": 2.8246531072276753,
      "grad_norm": 0.15530316531658173,
      "learning_rate": 5.848437452080842e-06,
      "loss": 0.4712,
      "mean_token_accuracy": 0.8513538874685764,
      "num_tokens": 979838564.0,
      "step": 30701
    },
    {
      "entropy": 0.5357860047370195,
      "epoch": 2.824745113599117,
      "grad_norm": 0.15577299892902374,
      "learning_rate": 5.845370625939216e-06,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8366206586360931,
      "num_tokens": 979871332.0,
      "step": 30702
    },
    {
      "entropy": 0.5206956584006548,
      "epoch": 2.824837119970558,
      "grad_norm": 0.16263645887374878,
      "learning_rate": 5.84230379979759e-06,
      "loss": 0.5206,
      "mean_token_accuracy": 0.8364713713526726,
      "num_tokens": 979903755.0,
      "step": 30703
    },
    {
      "entropy": 0.510059641674161,
      "epoch": 2.824929126341999,
      "grad_norm": 0.16078579425811768,
      "learning_rate": 5.839236973655964e-06,
      "loss": 0.5146,
      "mean_token_accuracy": 0.8377950228750706,
      "num_tokens": 979936168.0,
      "step": 30704
    },
    {
      "entropy": 0.45015039574354887,
      "epoch": 2.8250211327134402,
      "grad_norm": 0.155058354139328,
      "learning_rate": 5.836170147514337e-06,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8579364977777004,
      "num_tokens": 979968398.0,
      "step": 30705
    },
    {
      "entropy": 0.45773681811988354,
      "epoch": 2.8251131390848814,
      "grad_norm": 0.14887858927249908,
      "learning_rate": 5.833103321372712e-06,
      "loss": 0.4501,
      "mean_token_accuracy": 0.85500717908144,
      "num_tokens": 980000624.0,
      "step": 30706
    },
    {
      "entropy": 0.523702540434897,
      "epoch": 2.825205145456323,
      "grad_norm": 0.158290833234787,
      "learning_rate": 5.830036495231086e-06,
      "loss": 0.5105,
      "mean_token_accuracy": 0.8394437283277512,
      "num_tokens": 980032665.0,
      "step": 30707
    },
    {
      "entropy": 0.5575965833850205,
      "epoch": 2.825297151827764,
      "grad_norm": 0.16496066749095917,
      "learning_rate": 5.82696966908946e-06,
      "loss": 0.5397,
      "mean_token_accuracy": 0.8280420750379562,
      "num_tokens": 980064392.0,
      "step": 30708
    },
    {
      "entropy": 0.5103539302945137,
      "epoch": 2.825389158199205,
      "grad_norm": 0.15585334599018097,
      "learning_rate": 5.8239028429478335e-06,
      "loss": 0.4943,
      "mean_token_accuracy": 0.8448812067508698,
      "num_tokens": 980096662.0,
      "step": 30709
    },
    {
      "entropy": 0.4790516151115298,
      "epoch": 2.8254811645706464,
      "grad_norm": 0.1572667360305786,
      "learning_rate": 5.820836016806207e-06,
      "loss": 0.4696,
      "mean_token_accuracy": 0.848761536180973,
      "num_tokens": 980128356.0,
      "step": 30710
    },
    {
      "entropy": 0.5093629863113165,
      "epoch": 2.8255731709420875,
      "grad_norm": 0.1586901843547821,
      "learning_rate": 5.817769190664581e-06,
      "loss": 0.4943,
      "mean_token_accuracy": 0.8446452654898167,
      "num_tokens": 980160068.0,
      "step": 30711
    },
    {
      "entropy": 0.4468823978677392,
      "epoch": 2.825665177313529,
      "grad_norm": 0.15353244543075562,
      "learning_rate": 5.814702364522955e-06,
      "loss": 0.4335,
      "mean_token_accuracy": 0.860167745500803,
      "num_tokens": 980191569.0,
      "step": 30712
    },
    {
      "entropy": 0.49982450250536203,
      "epoch": 2.82575718368497,
      "grad_norm": 0.15122289955615997,
      "learning_rate": 5.8116355383813295e-06,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8459341712296009,
      "num_tokens": 980224112.0,
      "step": 30713
    },
    {
      "entropy": 0.4808986969292164,
      "epoch": 2.8258491900564113,
      "grad_norm": 0.1572885811328888,
      "learning_rate": 5.808568712239703e-06,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8493192754685879,
      "num_tokens": 980255724.0,
      "step": 30714
    },
    {
      "entropy": 0.6167458230629563,
      "epoch": 2.8259411964278525,
      "grad_norm": 0.17475128173828125,
      "learning_rate": 5.805501886098078e-06,
      "loss": 0.6012,
      "mean_token_accuracy": 0.8123305551707745,
      "num_tokens": 980287463.0,
      "step": 30715
    },
    {
      "entropy": 0.5052719851955771,
      "epoch": 2.8260332027992936,
      "grad_norm": 0.15749770402908325,
      "learning_rate": 5.802435059956452e-06,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8425124622881413,
      "num_tokens": 980319375.0,
      "step": 30716
    },
    {
      "entropy": 0.5469092973507941,
      "epoch": 2.826125209170735,
      "grad_norm": 0.15767431259155273,
      "learning_rate": 5.7993682338148256e-06,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8292575404047966,
      "num_tokens": 980351549.0,
      "step": 30717
    },
    {
      "entropy": 0.6190937533974648,
      "epoch": 2.8262172155421763,
      "grad_norm": 0.17760969698429108,
      "learning_rate": 5.796301407673199e-06,
      "loss": 0.6122,
      "mean_token_accuracy": 0.8080600127577782,
      "num_tokens": 980383164.0,
      "step": 30718
    },
    {
      "entropy": 0.5352273555472493,
      "epoch": 2.8263092219136174,
      "grad_norm": 0.1586073487997055,
      "learning_rate": 5.793234581531573e-06,
      "loss": 0.5203,
      "mean_token_accuracy": 0.8365787453949451,
      "num_tokens": 980415225.0,
      "step": 30719
    },
    {
      "entropy": 0.5370484218001366,
      "epoch": 2.8264012282850586,
      "grad_norm": 0.16470053791999817,
      "learning_rate": 5.790167755389947e-06,
      "loss": 0.5336,
      "mean_token_accuracy": 0.8288317993283272,
      "num_tokens": 980446624.0,
      "step": 30720
    },
    {
      "entropy": 0.5479905591346323,
      "epoch": 2.8264932346564997,
      "grad_norm": 0.1580156832933426,
      "learning_rate": 5.787100929248322e-06,
      "loss": 0.5341,
      "mean_token_accuracy": 0.8325903676450253,
      "num_tokens": 980479392.0,
      "step": 30721
    },
    {
      "entropy": 0.4404736696742475,
      "epoch": 2.8265852410279413,
      "grad_norm": 0.15573161840438843,
      "learning_rate": 5.784034103106695e-06,
      "loss": 0.4298,
      "mean_token_accuracy": 0.8627054654061794,
      "num_tokens": 980510336.0,
      "step": 30722
    },
    {
      "entropy": 0.44932823348790407,
      "epoch": 2.8266772473993824,
      "grad_norm": 0.1459510624408722,
      "learning_rate": 5.780967276965069e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8612214289605618,
      "num_tokens": 980542422.0,
      "step": 30723
    },
    {
      "entropy": 0.4759682109579444,
      "epoch": 2.8267692537708236,
      "grad_norm": 0.15419597923755646,
      "learning_rate": 5.777900450823443e-06,
      "loss": 0.4636,
      "mean_token_accuracy": 0.854646809399128,
      "num_tokens": 980574750.0,
      "step": 30724
    },
    {
      "entropy": 0.49297099094837904,
      "epoch": 2.8268612601422647,
      "grad_norm": 0.16193923354148865,
      "learning_rate": 5.774833624681817e-06,
      "loss": 0.4804,
      "mean_token_accuracy": 0.8438408747315407,
      "num_tokens": 980606235.0,
      "step": 30725
    },
    {
      "entropy": 0.4897456313483417,
      "epoch": 2.826953266513706,
      "grad_norm": 0.16127702593803406,
      "learning_rate": 5.771766798540191e-06,
      "loss": 0.4837,
      "mean_token_accuracy": 0.8459902592003345,
      "num_tokens": 980638160.0,
      "step": 30726
    },
    {
      "entropy": 0.4363760482519865,
      "epoch": 2.8270452728851474,
      "grad_norm": 0.14531256258487701,
      "learning_rate": 5.768699972398564e-06,
      "loss": 0.4254,
      "mean_token_accuracy": 0.8637722693383694,
      "num_tokens": 980670151.0,
      "step": 30727
    },
    {
      "entropy": 0.4884539758786559,
      "epoch": 2.8271372792565885,
      "grad_norm": 0.156505286693573,
      "learning_rate": 5.765633146256939e-06,
      "loss": 0.4786,
      "mean_token_accuracy": 0.846405565738678,
      "num_tokens": 980702358.0,
      "step": 30728
    },
    {
      "entropy": 0.4709008689969778,
      "epoch": 2.8272292856280297,
      "grad_norm": 0.15103039145469666,
      "learning_rate": 5.762566320115313e-06,
      "loss": 0.4551,
      "mean_token_accuracy": 0.8566405475139618,
      "num_tokens": 980734491.0,
      "step": 30729
    },
    {
      "entropy": 0.5427050925791264,
      "epoch": 2.827321291999471,
      "grad_norm": 0.1649104803800583,
      "learning_rate": 5.7594994939736875e-06,
      "loss": 0.5261,
      "mean_token_accuracy": 0.8339856080710888,
      "num_tokens": 980765008.0,
      "step": 30730
    },
    {
      "entropy": 0.43452922254800797,
      "epoch": 2.827413298370912,
      "grad_norm": 0.14713752269744873,
      "learning_rate": 5.756432667832061e-06,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8626778982579708,
      "num_tokens": 980796492.0,
      "step": 30731
    },
    {
      "entropy": 0.5952228512614965,
      "epoch": 2.8275053047423535,
      "grad_norm": 0.17106515169143677,
      "learning_rate": 5.753365841690435e-06,
      "loss": 0.5828,
      "mean_token_accuracy": 0.8185066394507885,
      "num_tokens": 980827575.0,
      "step": 30732
    },
    {
      "entropy": 0.4049360987264663,
      "epoch": 2.8275973111137946,
      "grad_norm": 0.1453365832567215,
      "learning_rate": 5.750299015548809e-06,
      "loss": 0.3899,
      "mean_token_accuracy": 0.8727223202586174,
      "num_tokens": 980859536.0,
      "step": 30733
    },
    {
      "entropy": 0.4245196804404259,
      "epoch": 2.8276893174852358,
      "grad_norm": 0.1474565863609314,
      "learning_rate": 5.747232189407183e-06,
      "loss": 0.4071,
      "mean_token_accuracy": 0.8671754561364651,
      "num_tokens": 980891007.0,
      "step": 30734
    },
    {
      "entropy": 0.5188368950039148,
      "epoch": 2.827781323856677,
      "grad_norm": 0.1607431024312973,
      "learning_rate": 5.7441653632655565e-06,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8392727598547935,
      "num_tokens": 980922549.0,
      "step": 30735
    },
    {
      "entropy": 0.45021918904967606,
      "epoch": 2.827873330228118,
      "grad_norm": 0.1432957798242569,
      "learning_rate": 5.741098537123931e-06,
      "loss": 0.4338,
      "mean_token_accuracy": 0.8629451803863049,
      "num_tokens": 980954876.0,
      "step": 30736
    },
    {
      "entropy": 0.4375721737742424,
      "epoch": 2.8279653365995596,
      "grad_norm": 0.15015611052513123,
      "learning_rate": 5.738031710982305e-06,
      "loss": 0.4233,
      "mean_token_accuracy": 0.862211998552084,
      "num_tokens": 980986301.0,
      "step": 30737
    },
    {
      "entropy": 0.4759975429624319,
      "epoch": 2.8280573429710008,
      "grad_norm": 0.1553383767604828,
      "learning_rate": 5.734964884840679e-06,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8461357913911343,
      "num_tokens": 981018712.0,
      "step": 30738
    },
    {
      "entropy": 0.5220368313603103,
      "epoch": 2.828149349342442,
      "grad_norm": 0.16791073977947235,
      "learning_rate": 5.7318980586990526e-06,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8349144719541073,
      "num_tokens": 981050760.0,
      "step": 30739
    },
    {
      "entropy": 0.4760425901040435,
      "epoch": 2.828241355713883,
      "grad_norm": 0.15850703418254852,
      "learning_rate": 5.728831232557426e-06,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8493364974856377,
      "num_tokens": 981083289.0,
      "step": 30740
    },
    {
      "entropy": 0.4491696336772293,
      "epoch": 2.828333362085324,
      "grad_norm": 0.14644743502140045,
      "learning_rate": 5.7257644064158e-06,
      "loss": 0.4273,
      "mean_token_accuracy": 0.8628015965223312,
      "num_tokens": 981115793.0,
      "step": 30741
    },
    {
      "entropy": 0.6213086172938347,
      "epoch": 2.8284253684567657,
      "grad_norm": 0.16849030554294586,
      "learning_rate": 5.722697580274174e-06,
      "loss": 0.6113,
      "mean_token_accuracy": 0.8081787303090096,
      "num_tokens": 981148423.0,
      "step": 30742
    },
    {
      "entropy": 0.5609804578125477,
      "epoch": 2.828517374828207,
      "grad_norm": 0.16737361252307892,
      "learning_rate": 5.719630754132549e-06,
      "loss": 0.5475,
      "mean_token_accuracy": 0.8244437538087368,
      "num_tokens": 981180005.0,
      "step": 30743
    },
    {
      "entropy": 0.5187018159776926,
      "epoch": 2.828609381199648,
      "grad_norm": 0.15968452394008636,
      "learning_rate": 5.716563927990922e-06,
      "loss": 0.5121,
      "mean_token_accuracy": 0.8402453511953354,
      "num_tokens": 981211838.0,
      "step": 30744
    },
    {
      "entropy": 0.586634092964232,
      "epoch": 2.828701387571089,
      "grad_norm": 0.16855624318122864,
      "learning_rate": 5.713497101849297e-06,
      "loss": 0.5741,
      "mean_token_accuracy": 0.81878811866045,
      "num_tokens": 981243754.0,
      "step": 30745
    },
    {
      "entropy": 0.43090315256267786,
      "epoch": 2.8287933939425303,
      "grad_norm": 0.1496388465166092,
      "learning_rate": 5.710430275707671e-06,
      "loss": 0.4303,
      "mean_token_accuracy": 0.8641488552093506,
      "num_tokens": 981275772.0,
      "step": 30746
    },
    {
      "entropy": 0.6209806464612484,
      "epoch": 2.828885400313972,
      "grad_norm": 0.1730778068304062,
      "learning_rate": 5.707363449566045e-06,
      "loss": 0.6024,
      "mean_token_accuracy": 0.8111903332173824,
      "num_tokens": 981308122.0,
      "step": 30747
    },
    {
      "entropy": 0.4652902130037546,
      "epoch": 2.828977406685413,
      "grad_norm": 0.15498612821102142,
      "learning_rate": 5.7042966234244184e-06,
      "loss": 0.451,
      "mean_token_accuracy": 0.8565200828015804,
      "num_tokens": 981340019.0,
      "step": 30748
    },
    {
      "entropy": 0.42890664935112,
      "epoch": 2.829069413056854,
      "grad_norm": 0.15043799579143524,
      "learning_rate": 5.701229797282792e-06,
      "loss": 0.4225,
      "mean_token_accuracy": 0.8667032867670059,
      "num_tokens": 981371788.0,
      "step": 30749
    },
    {
      "entropy": 0.5188657604157925,
      "epoch": 2.8291614194282952,
      "grad_norm": 0.15568311512470245,
      "learning_rate": 5.698162971141166e-06,
      "loss": 0.5044,
      "mean_token_accuracy": 0.8392857648432255,
      "num_tokens": 981403497.0,
      "step": 30750
    },
    {
      "entropy": 0.5107304895063862,
      "epoch": 2.8292534257997364,
      "grad_norm": 0.1557496339082718,
      "learning_rate": 5.69509614499954e-06,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8449386209249496,
      "num_tokens": 981435991.0,
      "step": 30751
    },
    {
      "entropy": 0.524192813783884,
      "epoch": 2.829345432171178,
      "grad_norm": 0.15437975525856018,
      "learning_rate": 5.6920293188579145e-06,
      "loss": 0.5035,
      "mean_token_accuracy": 0.8398679830133915,
      "num_tokens": 981467448.0,
      "step": 30752
    },
    {
      "entropy": 0.44923115335404873,
      "epoch": 2.829437438542619,
      "grad_norm": 0.14449234306812286,
      "learning_rate": 5.688962492716288e-06,
      "loss": 0.4313,
      "mean_token_accuracy": 0.8649622239172459,
      "num_tokens": 981499919.0,
      "step": 30753
    },
    {
      "entropy": 0.5067380275577307,
      "epoch": 2.82952944491406,
      "grad_norm": 0.16109438240528107,
      "learning_rate": 5.685895666574662e-06,
      "loss": 0.496,
      "mean_token_accuracy": 0.8415318913757801,
      "num_tokens": 981531832.0,
      "step": 30754
    },
    {
      "entropy": 0.5012020859867334,
      "epoch": 2.8296214512855014,
      "grad_norm": 0.15479475259780884,
      "learning_rate": 5.682828840433036e-06,
      "loss": 0.4912,
      "mean_token_accuracy": 0.8451342694461346,
      "num_tokens": 981563829.0,
      "step": 30755
    },
    {
      "entropy": 0.5161574184894562,
      "epoch": 2.8297134576569425,
      "grad_norm": 0.16147741675376892,
      "learning_rate": 5.67976201429141e-06,
      "loss": 0.514,
      "mean_token_accuracy": 0.8377743512392044,
      "num_tokens": 981595930.0,
      "step": 30756
    },
    {
      "entropy": 0.5944633092731237,
      "epoch": 2.829805464028384,
      "grad_norm": 0.1678607016801834,
      "learning_rate": 5.6766951881497835e-06,
      "loss": 0.5938,
      "mean_token_accuracy": 0.8154339790344238,
      "num_tokens": 981628114.0,
      "step": 30757
    },
    {
      "entropy": 0.5006955247372389,
      "epoch": 2.829897470399825,
      "grad_norm": 0.15470664203166962,
      "learning_rate": 5.673628362008158e-06,
      "loss": 0.491,
      "mean_token_accuracy": 0.844619557261467,
      "num_tokens": 981660876.0,
      "step": 30758
    },
    {
      "entropy": 0.48524318542331457,
      "epoch": 2.8299894767712663,
      "grad_norm": 0.15498578548431396,
      "learning_rate": 5.670561535866532e-06,
      "loss": 0.4715,
      "mean_token_accuracy": 0.8480599410831928,
      "num_tokens": 981692596.0,
      "step": 30759
    },
    {
      "entropy": 0.4613969293422997,
      "epoch": 2.830081483142708,
      "grad_norm": 0.1473885476589203,
      "learning_rate": 5.667494709724907e-06,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8591958619654179,
      "num_tokens": 981725136.0,
      "step": 30760
    },
    {
      "entropy": 0.5209539271891117,
      "epoch": 2.8301734895141486,
      "grad_norm": 0.15895788371562958,
      "learning_rate": 5.66442788358328e-06,
      "loss": 0.5114,
      "mean_token_accuracy": 0.839583832770586,
      "num_tokens": 981757245.0,
      "step": 30761
    },
    {
      "entropy": 0.49122844776138663,
      "epoch": 2.83026549588559,
      "grad_norm": 0.15494033694267273,
      "learning_rate": 5.661361057441654e-06,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8475398011505604,
      "num_tokens": 981789940.0,
      "step": 30762
    },
    {
      "entropy": 0.608969546854496,
      "epoch": 2.8303575022570313,
      "grad_norm": 0.17324140667915344,
      "learning_rate": 5.658294231300028e-06,
      "loss": 0.6086,
      "mean_token_accuracy": 0.8110148310661316,
      "num_tokens": 981821839.0,
      "step": 30763
    },
    {
      "entropy": 0.44588272273540497,
      "epoch": 2.8304495086284724,
      "grad_norm": 0.14703841507434845,
      "learning_rate": 5.655227405158402e-06,
      "loss": 0.4309,
      "mean_token_accuracy": 0.8622820265591145,
      "num_tokens": 981853763.0,
      "step": 30764
    },
    {
      "entropy": 0.5390018466860056,
      "epoch": 2.830541514999914,
      "grad_norm": 0.1574896275997162,
      "learning_rate": 5.652160579016776e-06,
      "loss": 0.518,
      "mean_token_accuracy": 0.8372841849923134,
      "num_tokens": 981886191.0,
      "step": 30765
    },
    {
      "entropy": 0.5335904574021697,
      "epoch": 2.8306335213713547,
      "grad_norm": 0.16236627101898193,
      "learning_rate": 5.649093752875149e-06,
      "loss": 0.5206,
      "mean_token_accuracy": 0.8329044654965401,
      "num_tokens": 981918004.0,
      "step": 30766
    },
    {
      "entropy": 0.4184075314551592,
      "epoch": 2.8307255277427963,
      "grad_norm": 0.14028549194335938,
      "learning_rate": 5.646026926733524e-06,
      "loss": 0.4029,
      "mean_token_accuracy": 0.8719499818980694,
      "num_tokens": 981950183.0,
      "step": 30767
    },
    {
      "entropy": 0.5258149644359946,
      "epoch": 2.8308175341142374,
      "grad_norm": 0.15707919001579285,
      "learning_rate": 5.642960100591898e-06,
      "loss": 0.5048,
      "mean_token_accuracy": 0.8410048969089985,
      "num_tokens": 981982293.0,
      "step": 30768
    },
    {
      "entropy": 0.6177536565810442,
      "epoch": 2.8309095404856786,
      "grad_norm": 0.17158116400241852,
      "learning_rate": 5.639893274450272e-06,
      "loss": 0.611,
      "mean_token_accuracy": 0.8087483271956444,
      "num_tokens": 982014171.0,
      "step": 30769
    },
    {
      "entropy": 0.4836545465514064,
      "epoch": 2.83100154685712,
      "grad_norm": 0.15162163972854614,
      "learning_rate": 5.6368264483086454e-06,
      "loss": 0.4706,
      "mean_token_accuracy": 0.8528161570429802,
      "num_tokens": 982045789.0,
      "step": 30770
    },
    {
      "entropy": 0.5584819592768326,
      "epoch": 2.831093553228561,
      "grad_norm": 0.1608712375164032,
      "learning_rate": 5.633759622167019e-06,
      "loss": 0.5357,
      "mean_token_accuracy": 0.8294152952730656,
      "num_tokens": 982077469.0,
      "step": 30771
    },
    {
      "entropy": 0.49154803063720465,
      "epoch": 2.8311855596000024,
      "grad_norm": 0.15500931441783905,
      "learning_rate": 5.630692796025393e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.84512098133564,
      "num_tokens": 982109392.0,
      "step": 30772
    },
    {
      "entropy": 0.4140441018389538,
      "epoch": 2.8312775659714435,
      "grad_norm": 0.14062243700027466,
      "learning_rate": 5.627625969883768e-06,
      "loss": 0.3939,
      "mean_token_accuracy": 0.8749253042042255,
      "num_tokens": 982141276.0,
      "step": 30773
    },
    {
      "entropy": 0.4587291283532977,
      "epoch": 2.8313695723428847,
      "grad_norm": 0.15371759235858917,
      "learning_rate": 5.6245591437421415e-06,
      "loss": 0.4525,
      "mean_token_accuracy": 0.852049745619297,
      "num_tokens": 982172876.0,
      "step": 30774
    },
    {
      "entropy": 0.4937840709462762,
      "epoch": 2.8314615787143262,
      "grad_norm": 0.153734028339386,
      "learning_rate": 5.621492317600516e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8437619470059872,
      "num_tokens": 982204810.0,
      "step": 30775
    },
    {
      "entropy": 0.5473852527793497,
      "epoch": 2.831553585085767,
      "grad_norm": 0.16592560708522797,
      "learning_rate": 5.61842549145889e-06,
      "loss": 0.529,
      "mean_token_accuracy": 0.8318765349686146,
      "num_tokens": 982236525.0,
      "step": 30776
    },
    {
      "entropy": 0.48574246698990464,
      "epoch": 2.8316455914572085,
      "grad_norm": 0.15510044991970062,
      "learning_rate": 5.615358665317264e-06,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8495673947036266,
      "num_tokens": 982268097.0,
      "step": 30777
    },
    {
      "entropy": 0.5177850658074021,
      "epoch": 2.8317375978286496,
      "grad_norm": 0.16456636786460876,
      "learning_rate": 5.6122918391756375e-06,
      "loss": 0.5035,
      "mean_token_accuracy": 0.8403764627873898,
      "num_tokens": 982299263.0,
      "step": 30778
    },
    {
      "entropy": 0.5911246575415134,
      "epoch": 2.8318296042000908,
      "grad_norm": 0.1706560105085373,
      "learning_rate": 5.609225013034011e-06,
      "loss": 0.5908,
      "mean_token_accuracy": 0.8173394091427326,
      "num_tokens": 982330927.0,
      "step": 30779
    },
    {
      "entropy": 0.518230794928968,
      "epoch": 2.8319216105715324,
      "grad_norm": 0.15378785133361816,
      "learning_rate": 5.606158186892385e-06,
      "loss": 0.5098,
      "mean_token_accuracy": 0.8395098634064198,
      "num_tokens": 982363070.0,
      "step": 30780
    },
    {
      "entropy": 0.42597562773153186,
      "epoch": 2.832013616942973,
      "grad_norm": 0.1502479612827301,
      "learning_rate": 5.603091360750759e-06,
      "loss": 0.4164,
      "mean_token_accuracy": 0.8639299273490906,
      "num_tokens": 982394313.0,
      "step": 30781
    },
    {
      "entropy": 0.5301134269684553,
      "epoch": 2.8321056233144146,
      "grad_norm": 0.16273942589759827,
      "learning_rate": 5.600024534609134e-06,
      "loss": 0.5292,
      "mean_token_accuracy": 0.8342523835599422,
      "num_tokens": 982425662.0,
      "step": 30782
    },
    {
      "entropy": 0.5227723009884357,
      "epoch": 2.8321976296858558,
      "grad_norm": 0.1636069118976593,
      "learning_rate": 5.596957708467507e-06,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8381098732352257,
      "num_tokens": 982457496.0,
      "step": 30783
    },
    {
      "entropy": 0.5127747878432274,
      "epoch": 2.832289636057297,
      "grad_norm": 0.15973149240016937,
      "learning_rate": 5.593890882325881e-06,
      "loss": 0.497,
      "mean_token_accuracy": 0.8423040993511677,
      "num_tokens": 982488416.0,
      "step": 30784
    },
    {
      "entropy": 0.4870662880130112,
      "epoch": 2.8323816424287385,
      "grad_norm": 0.15169215202331543,
      "learning_rate": 5.590824056184255e-06,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8509495630860329,
      "num_tokens": 982520594.0,
      "step": 30785
    },
    {
      "entropy": 0.4763760860078037,
      "epoch": 2.832473648800179,
      "grad_norm": 0.15279942750930786,
      "learning_rate": 5.587757230042629e-06,
      "loss": 0.4658,
      "mean_token_accuracy": 0.8526883125305176,
      "num_tokens": 982552689.0,
      "step": 30786
    },
    {
      "entropy": 0.5011884350096807,
      "epoch": 2.8325656551716207,
      "grad_norm": 0.1542358696460724,
      "learning_rate": 5.584690403901003e-06,
      "loss": 0.4893,
      "mean_token_accuracy": 0.842780590057373,
      "num_tokens": 982584257.0,
      "step": 30787
    },
    {
      "entropy": 0.4408895878586918,
      "epoch": 2.832657661543062,
      "grad_norm": 0.15300388634204865,
      "learning_rate": 5.581623577759377e-06,
      "loss": 0.4333,
      "mean_token_accuracy": 0.862041600048542,
      "num_tokens": 982615330.0,
      "step": 30788
    },
    {
      "entropy": 0.536591898300685,
      "epoch": 2.832749667914503,
      "grad_norm": 0.16071471571922302,
      "learning_rate": 5.578556751617751e-06,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8355703055858612,
      "num_tokens": 982647339.0,
      "step": 30789
    },
    {
      "entropy": 0.509639548137784,
      "epoch": 2.8328416742859446,
      "grad_norm": 0.16082000732421875,
      "learning_rate": 5.575489925476126e-06,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8435748256742954,
      "num_tokens": 982678813.0,
      "step": 30790
    },
    {
      "entropy": 0.5434460481628776,
      "epoch": 2.8329336806573853,
      "grad_norm": 0.1637696474790573,
      "learning_rate": 5.5724230993344995e-06,
      "loss": 0.5312,
      "mean_token_accuracy": 0.8329576253890991,
      "num_tokens": 982710511.0,
      "step": 30791
    },
    {
      "entropy": 0.5874759647995234,
      "epoch": 2.833025687028827,
      "grad_norm": 0.17405186593532562,
      "learning_rate": 5.569356273192873e-06,
      "loss": 0.5871,
      "mean_token_accuracy": 0.8143781535327435,
      "num_tokens": 982742550.0,
      "step": 30792
    },
    {
      "entropy": 0.5521581941284239,
      "epoch": 2.833117693400268,
      "grad_norm": 0.16535548865795135,
      "learning_rate": 5.566289447051247e-06,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8317292332649231,
      "num_tokens": 982774032.0,
      "step": 30793
    },
    {
      "entropy": 0.42204171791672707,
      "epoch": 2.833209699771709,
      "grad_norm": 0.14864133298397064,
      "learning_rate": 5.563222620909621e-06,
      "loss": 0.3973,
      "mean_token_accuracy": 0.8727068826556206,
      "num_tokens": 982804849.0,
      "step": 30794
    },
    {
      "entropy": 0.45346603682264686,
      "epoch": 2.8333017061431507,
      "grad_norm": 0.1511583775281906,
      "learning_rate": 5.560155794767995e-06,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8585865013301373,
      "num_tokens": 982836626.0,
      "step": 30795
    },
    {
      "entropy": 0.4778182804584503,
      "epoch": 2.8333937125145914,
      "grad_norm": 0.15419700741767883,
      "learning_rate": 5.5570889686263685e-06,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8499455563724041,
      "num_tokens": 982868487.0,
      "step": 30796
    },
    {
      "entropy": 0.5770160835236311,
      "epoch": 2.833485718886033,
      "grad_norm": 0.16805121302604675,
      "learning_rate": 5.554022142484743e-06,
      "loss": 0.5578,
      "mean_token_accuracy": 0.8229421488940716,
      "num_tokens": 982900361.0,
      "step": 30797
    },
    {
      "entropy": 0.4690220768097788,
      "epoch": 2.833577725257474,
      "grad_norm": 0.15200263261795044,
      "learning_rate": 5.550955316343117e-06,
      "loss": 0.4526,
      "mean_token_accuracy": 0.8551665842533112,
      "num_tokens": 982932274.0,
      "step": 30798
    },
    {
      "entropy": 0.586215203627944,
      "epoch": 2.8336697316289152,
      "grad_norm": 0.16454531252384186,
      "learning_rate": 5.547888490201491e-06,
      "loss": 0.5652,
      "mean_token_accuracy": 0.8193839080631733,
      "num_tokens": 982964182.0,
      "step": 30799
    },
    {
      "entropy": 0.5674469098448753,
      "epoch": 2.833761738000357,
      "grad_norm": 0.16725873947143555,
      "learning_rate": 5.5448216640598645e-06,
      "loss": 0.5469,
      "mean_token_accuracy": 0.8226227574050426,
      "num_tokens": 982995695.0,
      "step": 30800
    },
    {
      "entropy": 0.5352579923346639,
      "epoch": 2.8338537443717975,
      "grad_norm": 0.16175317764282227,
      "learning_rate": 5.541754837918238e-06,
      "loss": 0.5171,
      "mean_token_accuracy": 0.8371655084192753,
      "num_tokens": 983026829.0,
      "step": 30801
    },
    {
      "entropy": 0.4755205390974879,
      "epoch": 2.833945750743239,
      "grad_norm": 0.15058408677577972,
      "learning_rate": 5.538688011776612e-06,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8557251617312431,
      "num_tokens": 983058837.0,
      "step": 30802
    },
    {
      "entropy": 0.5963310617953539,
      "epoch": 2.83403775711468,
      "grad_norm": 0.16825750470161438,
      "learning_rate": 5.535621185634987e-06,
      "loss": 0.591,
      "mean_token_accuracy": 0.8169018737971783,
      "num_tokens": 983091254.0,
      "step": 30803
    },
    {
      "entropy": 0.47828649263828993,
      "epoch": 2.8341297634861213,
      "grad_norm": 0.15391774475574493,
      "learning_rate": 5.532554359493361e-06,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8499655500054359,
      "num_tokens": 983123783.0,
      "step": 30804
    },
    {
      "entropy": 0.527013260871172,
      "epoch": 2.834221769857563,
      "grad_norm": 0.1651645302772522,
      "learning_rate": 5.529487533351735e-06,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8345692865550518,
      "num_tokens": 983155469.0,
      "step": 30805
    },
    {
      "entropy": 0.5635727867484093,
      "epoch": 2.8343137762290036,
      "grad_norm": 0.16516827046871185,
      "learning_rate": 5.526420707210109e-06,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8252821639180183,
      "num_tokens": 983187468.0,
      "step": 30806
    },
    {
      "entropy": 0.49112529028207064,
      "epoch": 2.834405782600445,
      "grad_norm": 0.15268383920192719,
      "learning_rate": 5.523353881068483e-06,
      "loss": 0.4683,
      "mean_token_accuracy": 0.8483290150761604,
      "num_tokens": 983219522.0,
      "step": 30807
    },
    {
      "entropy": 0.520997291430831,
      "epoch": 2.8344977889718863,
      "grad_norm": 0.1592126190662384,
      "learning_rate": 5.520287054926857e-06,
      "loss": 0.5106,
      "mean_token_accuracy": 0.8360397182404995,
      "num_tokens": 983251722.0,
      "step": 30808
    },
    {
      "entropy": 0.6253173900768161,
      "epoch": 2.8345897953433274,
      "grad_norm": 0.17194989323616028,
      "learning_rate": 5.5172202287852304e-06,
      "loss": 0.6101,
      "mean_token_accuracy": 0.8083648346364498,
      "num_tokens": 983284249.0,
      "step": 30809
    },
    {
      "entropy": 0.44241382472682744,
      "epoch": 2.834681801714769,
      "grad_norm": 0.1517944484949112,
      "learning_rate": 5.514153402643604e-06,
      "loss": 0.426,
      "mean_token_accuracy": 0.8608533143997192,
      "num_tokens": 983315779.0,
      "step": 30810
    },
    {
      "entropy": 0.49941227841190994,
      "epoch": 2.8347738080862097,
      "grad_norm": 0.1526833027601242,
      "learning_rate": 5.511086576501978e-06,
      "loss": 0.4813,
      "mean_token_accuracy": 0.8490854650735855,
      "num_tokens": 983348277.0,
      "step": 30811
    },
    {
      "entropy": 0.5026094671338797,
      "epoch": 2.8348658144576513,
      "grad_norm": 0.15999311208724976,
      "learning_rate": 5.508019750360352e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8446546159684658,
      "num_tokens": 983379736.0,
      "step": 30812
    },
    {
      "entropy": 0.4463031133636832,
      "epoch": 2.8349578208290924,
      "grad_norm": 0.150692880153656,
      "learning_rate": 5.5049529242187265e-06,
      "loss": 0.4335,
      "mean_token_accuracy": 0.8610712923109531,
      "num_tokens": 983410912.0,
      "step": 30813
    },
    {
      "entropy": 0.5710107535123825,
      "epoch": 2.8350498272005336,
      "grad_norm": 0.16533899307250977,
      "learning_rate": 5.5018860980771e-06,
      "loss": 0.5541,
      "mean_token_accuracy": 0.8267124071717262,
      "num_tokens": 983442798.0,
      "step": 30814
    },
    {
      "entropy": 0.4289022843586281,
      "epoch": 2.835141833571975,
      "grad_norm": 0.1757938414812088,
      "learning_rate": 5.498819271935474e-06,
      "loss": 0.4216,
      "mean_token_accuracy": 0.865375205874443,
      "num_tokens": 983474611.0,
      "step": 30815
    },
    {
      "entropy": 0.5481075886636972,
      "epoch": 2.835233839943416,
      "grad_norm": 0.16558592021465302,
      "learning_rate": 5.495752445793848e-06,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8269090056419373,
      "num_tokens": 983506404.0,
      "step": 30816
    },
    {
      "entropy": 0.4304634225554764,
      "epoch": 2.8353258463148574,
      "grad_norm": 0.15343423187732697,
      "learning_rate": 5.492685619652222e-06,
      "loss": 0.4256,
      "mean_token_accuracy": 0.8654625490307808,
      "num_tokens": 983538381.0,
      "step": 30817
    },
    {
      "entropy": 0.5104298209771514,
      "epoch": 2.8354178526862985,
      "grad_norm": 0.1613747924566269,
      "learning_rate": 5.489618793510596e-06,
      "loss": 0.5046,
      "mean_token_accuracy": 0.8360678032040596,
      "num_tokens": 983570139.0,
      "step": 30818
    },
    {
      "entropy": 0.5624992307275534,
      "epoch": 2.8355098590577397,
      "grad_norm": 0.1656021922826767,
      "learning_rate": 5.48655196736897e-06,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8245978988707066,
      "num_tokens": 983601981.0,
      "step": 30819
    },
    {
      "entropy": 0.4794329348951578,
      "epoch": 2.8356018654291812,
      "grad_norm": 0.15124276280403137,
      "learning_rate": 5.483485141227345e-06,
      "loss": 0.4717,
      "mean_token_accuracy": 0.8483744151890278,
      "num_tokens": 983634388.0,
      "step": 30820
    },
    {
      "entropy": 0.5022896807640791,
      "epoch": 2.835693871800622,
      "grad_norm": 0.15884463489055634,
      "learning_rate": 5.480418315085719e-06,
      "loss": 0.4976,
      "mean_token_accuracy": 0.8404826149344444,
      "num_tokens": 983666303.0,
      "step": 30821
    },
    {
      "entropy": 0.5057252375409007,
      "epoch": 2.8357858781720635,
      "grad_norm": 0.15863195061683655,
      "learning_rate": 5.477351488944092e-06,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8388830795884132,
      "num_tokens": 983698720.0,
      "step": 30822
    },
    {
      "entropy": 0.557211431208998,
      "epoch": 2.8358778845435046,
      "grad_norm": 0.15912842750549316,
      "learning_rate": 5.474284662802466e-06,
      "loss": 0.5437,
      "mean_token_accuracy": 0.8294980525970459,
      "num_tokens": 983731360.0,
      "step": 30823
    },
    {
      "entropy": 0.6040778309106827,
      "epoch": 2.835969890914946,
      "grad_norm": 0.17241698503494263,
      "learning_rate": 5.47121783666084e-06,
      "loss": 0.59,
      "mean_token_accuracy": 0.8125056810677052,
      "num_tokens": 983763236.0,
      "step": 30824
    },
    {
      "entropy": 0.3882106754463166,
      "epoch": 2.8360618972863874,
      "grad_norm": 0.13855735957622528,
      "learning_rate": 5.468151010519214e-06,
      "loss": 0.3801,
      "mean_token_accuracy": 0.8772694654762745,
      "num_tokens": 983795804.0,
      "step": 30825
    },
    {
      "entropy": 0.4634431889280677,
      "epoch": 2.836153903657828,
      "grad_norm": 0.15315233170986176,
      "learning_rate": 5.465084184377588e-06,
      "loss": 0.4584,
      "mean_token_accuracy": 0.853467658162117,
      "num_tokens": 983827596.0,
      "step": 30826
    },
    {
      "entropy": 0.42409518733620644,
      "epoch": 2.8362459100292696,
      "grad_norm": 0.1391432136297226,
      "learning_rate": 5.462017358235961e-06,
      "loss": 0.4128,
      "mean_token_accuracy": 0.8681178018450737,
      "num_tokens": 983860070.0,
      "step": 30827
    },
    {
      "entropy": 0.5633469466120005,
      "epoch": 2.8363379164007108,
      "grad_norm": 0.15990519523620605,
      "learning_rate": 5.458950532094336e-06,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8252794705331326,
      "num_tokens": 983891915.0,
      "step": 30828
    },
    {
      "entropy": 0.4976943600922823,
      "epoch": 2.836429922772152,
      "grad_norm": 0.15326306223869324,
      "learning_rate": 5.45588370595271e-06,
      "loss": 0.4834,
      "mean_token_accuracy": 0.84910948574543,
      "num_tokens": 983923919.0,
      "step": 30829
    },
    {
      "entropy": 0.46072495402768254,
      "epoch": 2.8365219291435935,
      "grad_norm": 0.15576298534870148,
      "learning_rate": 5.452816879811084e-06,
      "loss": 0.4517,
      "mean_token_accuracy": 0.8585781343281269,
      "num_tokens": 983955871.0,
      "step": 30830
    },
    {
      "entropy": 0.49939229153096676,
      "epoch": 2.836613935515034,
      "grad_norm": 0.15964554250240326,
      "learning_rate": 5.4497500536694574e-06,
      "loss": 0.4927,
      "mean_token_accuracy": 0.8454192914068699,
      "num_tokens": 983986992.0,
      "step": 30831
    },
    {
      "entropy": 0.3800981077365577,
      "epoch": 2.8367059418864757,
      "grad_norm": 0.137970432639122,
      "learning_rate": 5.446683227527831e-06,
      "loss": 0.3718,
      "mean_token_accuracy": 0.8810350149869919,
      "num_tokens": 984019614.0,
      "step": 30832
    },
    {
      "entropy": 0.534972257912159,
      "epoch": 2.836797948257917,
      "grad_norm": 0.1629253774881363,
      "learning_rate": 5.443616401386206e-06,
      "loss": 0.5143,
      "mean_token_accuracy": 0.8350354135036469,
      "num_tokens": 984051238.0,
      "step": 30833
    },
    {
      "entropy": 0.5459138974547386,
      "epoch": 2.836889954629358,
      "grad_norm": 0.17034128308296204,
      "learning_rate": 5.44054957524458e-06,
      "loss": 0.5363,
      "mean_token_accuracy": 0.8293285295367241,
      "num_tokens": 984082547.0,
      "step": 30834
    },
    {
      "entropy": 0.5584123432636261,
      "epoch": 2.8369819610007996,
      "grad_norm": 0.16808563470840454,
      "learning_rate": 5.4374827491029535e-06,
      "loss": 0.5558,
      "mean_token_accuracy": 0.824787862598896,
      "num_tokens": 984114018.0,
      "step": 30835
    },
    {
      "entropy": 0.4689779947511852,
      "epoch": 2.8370739673722403,
      "grad_norm": 0.1503787785768509,
      "learning_rate": 5.434415922961328e-06,
      "loss": 0.4575,
      "mean_token_accuracy": 0.8545525185763836,
      "num_tokens": 984145967.0,
      "step": 30836
    },
    {
      "entropy": 0.48147964011877775,
      "epoch": 2.837165973743682,
      "grad_norm": 0.15791285037994385,
      "learning_rate": 5.431349096819702e-06,
      "loss": 0.4778,
      "mean_token_accuracy": 0.8497544303536415,
      "num_tokens": 984177439.0,
      "step": 30837
    },
    {
      "entropy": 0.48829318629577756,
      "epoch": 2.837257980115123,
      "grad_norm": 0.15887834131717682,
      "learning_rate": 5.428282270678076e-06,
      "loss": 0.4871,
      "mean_token_accuracy": 0.8466715067625046,
      "num_tokens": 984209427.0,
      "step": 30838
    },
    {
      "entropy": 0.49491854361258447,
      "epoch": 2.837349986486564,
      "grad_norm": 0.1535016894340515,
      "learning_rate": 5.4252154445364495e-06,
      "loss": 0.4706,
      "mean_token_accuracy": 0.8489282205700874,
      "num_tokens": 984240960.0,
      "step": 30839
    },
    {
      "entropy": 0.5371818579733372,
      "epoch": 2.8374419928580057,
      "grad_norm": 0.1647314429283142,
      "learning_rate": 5.422148618394823e-06,
      "loss": 0.5176,
      "mean_token_accuracy": 0.8363812267780304,
      "num_tokens": 984272531.0,
      "step": 30840
    },
    {
      "entropy": 0.5293594840914011,
      "epoch": 2.8375339992294464,
      "grad_norm": 0.16063489019870758,
      "learning_rate": 5.419081792253197e-06,
      "loss": 0.5164,
      "mean_token_accuracy": 0.8418919444084167,
      "num_tokens": 984304957.0,
      "step": 30841
    },
    {
      "entropy": 0.4951758135575801,
      "epoch": 2.837626005600888,
      "grad_norm": 0.1538560837507248,
      "learning_rate": 5.416014966111571e-06,
      "loss": 0.479,
      "mean_token_accuracy": 0.847920186817646,
      "num_tokens": 984337206.0,
      "step": 30842
    },
    {
      "entropy": 0.5270031600957736,
      "epoch": 2.837718011972329,
      "grad_norm": 0.158014714717865,
      "learning_rate": 5.412948139969946e-06,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8370269723236561,
      "num_tokens": 984368767.0,
      "step": 30843
    },
    {
      "entropy": 0.4940642435103655,
      "epoch": 2.8378100183437702,
      "grad_norm": 0.16319096088409424,
      "learning_rate": 5.409881313828319e-06,
      "loss": 0.4906,
      "mean_token_accuracy": 0.8397546410560608,
      "num_tokens": 984400375.0,
      "step": 30844
    },
    {
      "entropy": 0.46044142078608274,
      "epoch": 2.837902024715212,
      "grad_norm": 0.1514016091823578,
      "learning_rate": 5.406814487686693e-06,
      "loss": 0.439,
      "mean_token_accuracy": 0.8595194406807423,
      "num_tokens": 984431930.0,
      "step": 30845
    },
    {
      "entropy": 0.5612105466425419,
      "epoch": 2.8379940310866525,
      "grad_norm": 0.16407941281795502,
      "learning_rate": 5.403747661545067e-06,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8267320692539215,
      "num_tokens": 984464224.0,
      "step": 30846
    },
    {
      "entropy": 0.5810034102760255,
      "epoch": 2.838086037458094,
      "grad_norm": 0.167433962225914,
      "learning_rate": 5.400680835403441e-06,
      "loss": 0.5575,
      "mean_token_accuracy": 0.8251894041895866,
      "num_tokens": 984496207.0,
      "step": 30847
    },
    {
      "entropy": 0.42753190733492374,
      "epoch": 2.838178043829535,
      "grad_norm": 0.14843961596488953,
      "learning_rate": 5.3976140092618154e-06,
      "loss": 0.4127,
      "mean_token_accuracy": 0.8655072003602982,
      "num_tokens": 984527511.0,
      "step": 30848
    },
    {
      "entropy": 0.5212423019111156,
      "epoch": 2.8382700502009763,
      "grad_norm": 0.16146595776081085,
      "learning_rate": 5.394547183120189e-06,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8393901698291302,
      "num_tokens": 984558851.0,
      "step": 30849
    },
    {
      "entropy": 0.5399132360471413,
      "epoch": 2.838362056572418,
      "grad_norm": 0.1587369441986084,
      "learning_rate": 5.391480356978563e-06,
      "loss": 0.5284,
      "mean_token_accuracy": 0.8352274857461452,
      "num_tokens": 984591305.0,
      "step": 30850
    },
    {
      "entropy": 0.5359340030699968,
      "epoch": 2.8384540629438586,
      "grad_norm": 0.16005000472068787,
      "learning_rate": 5.388413530836938e-06,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8351939804852009,
      "num_tokens": 984623528.0,
      "step": 30851
    },
    {
      "entropy": 0.527529438957572,
      "epoch": 2.8385460693153,
      "grad_norm": 0.16297169029712677,
      "learning_rate": 5.3853467046953115e-06,
      "loss": 0.5143,
      "mean_token_accuracy": 0.8373116739094257,
      "num_tokens": 984655293.0,
      "step": 30852
    },
    {
      "entropy": 0.5705566597171128,
      "epoch": 2.8386380756867413,
      "grad_norm": 0.15724313259124756,
      "learning_rate": 5.382279878553685e-06,
      "loss": 0.5606,
      "mean_token_accuracy": 0.8246529437601566,
      "num_tokens": 984687739.0,
      "step": 30853
    },
    {
      "entropy": 0.4970882300985977,
      "epoch": 2.8387300820581824,
      "grad_norm": 0.1491517722606659,
      "learning_rate": 5.379213052412059e-06,
      "loss": 0.4781,
      "mean_token_accuracy": 0.8475197814404964,
      "num_tokens": 984720248.0,
      "step": 30854
    },
    {
      "entropy": 0.4941301168873906,
      "epoch": 2.838822088429624,
      "grad_norm": 0.1604161411523819,
      "learning_rate": 5.376146226270433e-06,
      "loss": 0.4924,
      "mean_token_accuracy": 0.844475619494915,
      "num_tokens": 984752440.0,
      "step": 30855
    },
    {
      "entropy": 0.5075097922235727,
      "epoch": 2.8389140948010647,
      "grad_norm": 0.164576455950737,
      "learning_rate": 5.373079400128807e-06,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8386057503521442,
      "num_tokens": 984784519.0,
      "step": 30856
    },
    {
      "entropy": 0.5314411390572786,
      "epoch": 2.8390061011725063,
      "grad_norm": 0.16011282801628113,
      "learning_rate": 5.3700125739871805e-06,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8336268737912178,
      "num_tokens": 984816629.0,
      "step": 30857
    },
    {
      "entropy": 0.45366244832985103,
      "epoch": 2.8390981075439474,
      "grad_norm": 0.14830008149147034,
      "learning_rate": 5.366945747845555e-06,
      "loss": 0.4356,
      "mean_token_accuracy": 0.8624216578900814,
      "num_tokens": 984849103.0,
      "step": 30858
    },
    {
      "entropy": 0.4796434943564236,
      "epoch": 2.8391901139153886,
      "grad_norm": 0.1494174301624298,
      "learning_rate": 5.363878921703929e-06,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8535391800105572,
      "num_tokens": 984881123.0,
      "step": 30859
    },
    {
      "entropy": 0.5433413553982973,
      "epoch": 2.83928212028683,
      "grad_norm": 0.16520258784294128,
      "learning_rate": 5.360812095562303e-06,
      "loss": 0.533,
      "mean_token_accuracy": 0.8298856355249882,
      "num_tokens": 984912743.0,
      "step": 30860
    },
    {
      "entropy": 0.4592350609600544,
      "epoch": 2.839374126658271,
      "grad_norm": 0.15304315090179443,
      "learning_rate": 5.3577452694206765e-06,
      "loss": 0.4489,
      "mean_token_accuracy": 0.856724925339222,
      "num_tokens": 984944623.0,
      "step": 30861
    },
    {
      "entropy": 0.6013162657618523,
      "epoch": 2.8394661330297124,
      "grad_norm": 0.17073333263397217,
      "learning_rate": 5.35467844327905e-06,
      "loss": 0.5803,
      "mean_token_accuracy": 0.8167915046215057,
      "num_tokens": 984976336.0,
      "step": 30862
    },
    {
      "entropy": 0.5308603877201676,
      "epoch": 2.8395581394011535,
      "grad_norm": 0.1597377061843872,
      "learning_rate": 5.351611617137425e-06,
      "loss": 0.5198,
      "mean_token_accuracy": 0.8363183736801147,
      "num_tokens": 985008017.0,
      "step": 30863
    },
    {
      "entropy": 0.47947958670556545,
      "epoch": 2.8396501457725947,
      "grad_norm": 0.15353946387767792,
      "learning_rate": 5.348544790995799e-06,
      "loss": 0.475,
      "mean_token_accuracy": 0.8483670055866241,
      "num_tokens": 985040095.0,
      "step": 30864
    },
    {
      "entropy": 0.4408190846443176,
      "epoch": 2.8397421521440362,
      "grad_norm": 0.15143945813179016,
      "learning_rate": 5.345477964854173e-06,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8591712638735771,
      "num_tokens": 985072477.0,
      "step": 30865
    },
    {
      "entropy": 0.5290170609951019,
      "epoch": 2.839834158515477,
      "grad_norm": 0.1588984578847885,
      "learning_rate": 5.342411138712547e-06,
      "loss": 0.5291,
      "mean_token_accuracy": 0.833920031785965,
      "num_tokens": 985104703.0,
      "step": 30866
    },
    {
      "entropy": 0.43374165962450206,
      "epoch": 2.8399261648869185,
      "grad_norm": 0.14418205618858337,
      "learning_rate": 5.339344312570921e-06,
      "loss": 0.4155,
      "mean_token_accuracy": 0.8628778979182243,
      "num_tokens": 985136570.0,
      "step": 30867
    },
    {
      "entropy": 0.4934403607621789,
      "epoch": 2.8400181712583596,
      "grad_norm": 0.16149728000164032,
      "learning_rate": 5.336277486429295e-06,
      "loss": 0.478,
      "mean_token_accuracy": 0.8476103954017162,
      "num_tokens": 985167692.0,
      "step": 30868
    },
    {
      "entropy": 0.5222466886043549,
      "epoch": 2.840110177629801,
      "grad_norm": 0.1635347306728363,
      "learning_rate": 5.333210660287669e-06,
      "loss": 0.5206,
      "mean_token_accuracy": 0.8337588049471378,
      "num_tokens": 985199251.0,
      "step": 30869
    },
    {
      "entropy": 0.5570830339565873,
      "epoch": 2.8402021840012424,
      "grad_norm": 0.1663794070482254,
      "learning_rate": 5.3301438341460424e-06,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8280361369252205,
      "num_tokens": 985230979.0,
      "step": 30870
    },
    {
      "entropy": 0.41675170324742794,
      "epoch": 2.840294190372683,
      "grad_norm": 0.1422896534204483,
      "learning_rate": 5.327077008004416e-06,
      "loss": 0.4034,
      "mean_token_accuracy": 0.8700130991637707,
      "num_tokens": 985263076.0,
      "step": 30871
    },
    {
      "entropy": 0.4790400620549917,
      "epoch": 2.8403861967441246,
      "grad_norm": 0.15478971600532532,
      "learning_rate": 5.32401018186279e-06,
      "loss": 0.4752,
      "mean_token_accuracy": 0.8490935042500496,
      "num_tokens": 985295819.0,
      "step": 30872
    },
    {
      "entropy": 0.4652027147822082,
      "epoch": 2.8404782031155658,
      "grad_norm": 0.14859072864055634,
      "learning_rate": 5.320943355721164e-06,
      "loss": 0.4475,
      "mean_token_accuracy": 0.8566161766648293,
      "num_tokens": 985327841.0,
      "step": 30873
    },
    {
      "entropy": 0.4955671103671193,
      "epoch": 2.840570209487007,
      "grad_norm": 0.1579618602991104,
      "learning_rate": 5.3178765295795385e-06,
      "loss": 0.4874,
      "mean_token_accuracy": 0.8436746075749397,
      "num_tokens": 985360215.0,
      "step": 30874
    },
    {
      "entropy": 0.3789389943704009,
      "epoch": 2.8406622158584485,
      "grad_norm": 0.1437903195619583,
      "learning_rate": 5.314809703437912e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.8805196173489094,
      "num_tokens": 985392552.0,
      "step": 30875
    },
    {
      "entropy": 0.4424340189434588,
      "epoch": 2.840754222229889,
      "grad_norm": 0.14925776422023773,
      "learning_rate": 5.311742877296286e-06,
      "loss": 0.4318,
      "mean_token_accuracy": 0.862743467092514,
      "num_tokens": 985425062.0,
      "step": 30876
    },
    {
      "entropy": 0.5175318804103881,
      "epoch": 2.8408462286013307,
      "grad_norm": 0.15535707771778107,
      "learning_rate": 5.30867605115466e-06,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8397960364818573,
      "num_tokens": 985457830.0,
      "step": 30877
    },
    {
      "entropy": 0.49720079731196165,
      "epoch": 2.840938234972772,
      "grad_norm": 0.1578425168991089,
      "learning_rate": 5.3056092250130345e-06,
      "loss": 0.4887,
      "mean_token_accuracy": 0.8409937135875225,
      "num_tokens": 985489507.0,
      "step": 30878
    },
    {
      "entropy": 0.47129555616993457,
      "epoch": 2.841030241344213,
      "grad_norm": 0.15585801005363464,
      "learning_rate": 5.302542398871408e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8519791662693024,
      "num_tokens": 985521878.0,
      "step": 30879
    },
    {
      "entropy": 0.5054392227903008,
      "epoch": 2.8411222477156546,
      "grad_norm": 0.15958043932914734,
      "learning_rate": 5.299475572729782e-06,
      "loss": 0.4976,
      "mean_token_accuracy": 0.8411745801568031,
      "num_tokens": 985554006.0,
      "step": 30880
    },
    {
      "entropy": 0.46018961491063237,
      "epoch": 2.8412142540870953,
      "grad_norm": 0.15228869020938873,
      "learning_rate": 5.296408746588157e-06,
      "loss": 0.4494,
      "mean_token_accuracy": 0.8533428981900215,
      "num_tokens": 985586636.0,
      "step": 30881
    },
    {
      "entropy": 0.4683815920725465,
      "epoch": 2.841306260458537,
      "grad_norm": 0.15300822257995605,
      "learning_rate": 5.293341920446531e-06,
      "loss": 0.4571,
      "mean_token_accuracy": 0.8526016995310783,
      "num_tokens": 985618466.0,
      "step": 30882
    },
    {
      "entropy": 0.475822564214468,
      "epoch": 2.841398266829978,
      "grad_norm": 0.15564963221549988,
      "learning_rate": 5.290275094304904e-06,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8523133657872677,
      "num_tokens": 985649599.0,
      "step": 30883
    },
    {
      "entropy": 0.4281503828242421,
      "epoch": 2.841490273201419,
      "grad_norm": 0.1476971060037613,
      "learning_rate": 5.287208268163278e-06,
      "loss": 0.41,
      "mean_token_accuracy": 0.8666361384093761,
      "num_tokens": 985681685.0,
      "step": 30884
    },
    {
      "entropy": 0.5234494532924145,
      "epoch": 2.8415822795728607,
      "grad_norm": 0.16155987977981567,
      "learning_rate": 5.284141442021652e-06,
      "loss": 0.5058,
      "mean_token_accuracy": 0.83685302734375,
      "num_tokens": 985713616.0,
      "step": 30885
    },
    {
      "entropy": 0.5307904845103621,
      "epoch": 2.8416742859443014,
      "grad_norm": 0.1602986752986908,
      "learning_rate": 5.281074615880026e-06,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8379770070314407,
      "num_tokens": 985745375.0,
      "step": 30886
    },
    {
      "entropy": 0.5739568546414375,
      "epoch": 2.841766292315743,
      "grad_norm": 0.16715571284294128,
      "learning_rate": 5.2780077897384e-06,
      "loss": 0.5594,
      "mean_token_accuracy": 0.8215081468224525,
      "num_tokens": 985777385.0,
      "step": 30887
    },
    {
      "entropy": 0.5530880724545568,
      "epoch": 2.841858298687184,
      "grad_norm": 0.16279549896717072,
      "learning_rate": 5.274940963596773e-06,
      "loss": 0.5417,
      "mean_token_accuracy": 0.8283645361661911,
      "num_tokens": 985808973.0,
      "step": 30888
    },
    {
      "entropy": 0.4878236148506403,
      "epoch": 2.8419503050586252,
      "grad_norm": 0.15519340336322784,
      "learning_rate": 5.271874137455148e-06,
      "loss": 0.4858,
      "mean_token_accuracy": 0.847225695848465,
      "num_tokens": 985841385.0,
      "step": 30889
    },
    {
      "entropy": 0.4705556258559227,
      "epoch": 2.842042311430067,
      "grad_norm": 0.15141384303569794,
      "learning_rate": 5.268807311313522e-06,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8534190505743027,
      "num_tokens": 985873994.0,
      "step": 30890
    },
    {
      "entropy": 0.49640288669615984,
      "epoch": 2.8421343178015075,
      "grad_norm": 0.15788842737674713,
      "learning_rate": 5.265740485171896e-06,
      "loss": 0.4917,
      "mean_token_accuracy": 0.843115258961916,
      "num_tokens": 985906428.0,
      "step": 30891
    },
    {
      "entropy": 0.5168431680649519,
      "epoch": 2.842226324172949,
      "grad_norm": 0.16373948752880096,
      "learning_rate": 5.2626736590302694e-06,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8369100615382195,
      "num_tokens": 985937461.0,
      "step": 30892
    },
    {
      "entropy": 0.48473438527435064,
      "epoch": 2.84231833054439,
      "grad_norm": 0.1585601568222046,
      "learning_rate": 5.259606832888644e-06,
      "loss": 0.4771,
      "mean_token_accuracy": 0.8513865992426872,
      "num_tokens": 985969735.0,
      "step": 30893
    },
    {
      "entropy": 0.4906923243543133,
      "epoch": 2.8424103369158313,
      "grad_norm": 0.15813089907169342,
      "learning_rate": 5.256540006747018e-06,
      "loss": 0.482,
      "mean_token_accuracy": 0.8468774147331715,
      "num_tokens": 986001238.0,
      "step": 30894
    },
    {
      "entropy": 0.5434903912246227,
      "epoch": 2.842502343287273,
      "grad_norm": 0.16837534308433533,
      "learning_rate": 5.253473180605392e-06,
      "loss": 0.5208,
      "mean_token_accuracy": 0.8322231359779835,
      "num_tokens": 986032111.0,
      "step": 30895
    },
    {
      "entropy": 0.49585324991494417,
      "epoch": 2.8425943496587136,
      "grad_norm": 0.15829573571681976,
      "learning_rate": 5.2504063544637655e-06,
      "loss": 0.4862,
      "mean_token_accuracy": 0.8474205620586872,
      "num_tokens": 986064866.0,
      "step": 30896
    },
    {
      "entropy": 0.4420115730026737,
      "epoch": 2.842686356030155,
      "grad_norm": 0.14652545750141144,
      "learning_rate": 5.24733952832214e-06,
      "loss": 0.431,
      "mean_token_accuracy": 0.8621942959725857,
      "num_tokens": 986096833.0,
      "step": 30897
    },
    {
      "entropy": 0.5189852137118578,
      "epoch": 2.8427783624015963,
      "grad_norm": 0.1617887169122696,
      "learning_rate": 5.244272702180514e-06,
      "loss": 0.5131,
      "mean_token_accuracy": 0.836378138512373,
      "num_tokens": 986129523.0,
      "step": 30898
    },
    {
      "entropy": 0.3598820809274912,
      "epoch": 2.8428703687730374,
      "grad_norm": 0.13089272379875183,
      "learning_rate": 5.241205876038888e-06,
      "loss": 0.3476,
      "mean_token_accuracy": 0.8878457099199295,
      "num_tokens": 986161940.0,
      "step": 30899
    },
    {
      "entropy": 0.5555359926074743,
      "epoch": 2.842962375144479,
      "grad_norm": 0.16475778818130493,
      "learning_rate": 5.2381390498972615e-06,
      "loss": 0.5414,
      "mean_token_accuracy": 0.8293236494064331,
      "num_tokens": 986194508.0,
      "step": 30900
    },
    {
      "entropy": 0.5294822948053479,
      "epoch": 2.8430543815159197,
      "grad_norm": 0.1601799577474594,
      "learning_rate": 5.235072223755635e-06,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8348379842936993,
      "num_tokens": 986226420.0,
      "step": 30901
    },
    {
      "entropy": 0.5299427639693022,
      "epoch": 2.8431463878873613,
      "grad_norm": 0.1584051549434662,
      "learning_rate": 5.232005397614009e-06,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8383761830627918,
      "num_tokens": 986258011.0,
      "step": 30902
    },
    {
      "entropy": 0.4870434086769819,
      "epoch": 2.8432383942588024,
      "grad_norm": 0.16269761323928833,
      "learning_rate": 5.228938571472383e-06,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8482878804206848,
      "num_tokens": 986289692.0,
      "step": 30903
    },
    {
      "entropy": 0.49538907315582037,
      "epoch": 2.8433304006302436,
      "grad_norm": 0.1512833684682846,
      "learning_rate": 5.225871745330758e-06,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8473032638430595,
      "num_tokens": 986321732.0,
      "step": 30904
    },
    {
      "entropy": 0.5336834050249308,
      "epoch": 2.843422407001685,
      "grad_norm": 0.16373132169246674,
      "learning_rate": 5.222804919189131e-06,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8323710970580578,
      "num_tokens": 986352878.0,
      "step": 30905
    },
    {
      "entropy": 0.5076505020260811,
      "epoch": 2.843514413373126,
      "grad_norm": 0.15562652051448822,
      "learning_rate": 5.219738093047505e-06,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8458886742591858,
      "num_tokens": 986384613.0,
      "step": 30906
    },
    {
      "entropy": 0.5148686291649938,
      "epoch": 2.8436064197445674,
      "grad_norm": 0.16533437371253967,
      "learning_rate": 5.216671266905879e-06,
      "loss": 0.5088,
      "mean_token_accuracy": 0.8387455157935619,
      "num_tokens": 986416581.0,
      "step": 30907
    },
    {
      "entropy": 0.5498736230656505,
      "epoch": 2.8436984261160085,
      "grad_norm": 0.16610053181648254,
      "learning_rate": 5.213604440764254e-06,
      "loss": 0.5451,
      "mean_token_accuracy": 0.828874833881855,
      "num_tokens": 986448338.0,
      "step": 30908
    },
    {
      "entropy": 0.5550685534253716,
      "epoch": 2.8437904324874497,
      "grad_norm": 0.16254404187202454,
      "learning_rate": 5.2105376146226274e-06,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8274310231208801,
      "num_tokens": 986480732.0,
      "step": 30909
    },
    {
      "entropy": 0.4777596453204751,
      "epoch": 2.8438824388588912,
      "grad_norm": 0.152607724070549,
      "learning_rate": 5.207470788481001e-06,
      "loss": 0.4683,
      "mean_token_accuracy": 0.8494230695068836,
      "num_tokens": 986512912.0,
      "step": 30910
    },
    {
      "entropy": 0.5303374119102955,
      "epoch": 2.843974445230332,
      "grad_norm": 0.15608102083206177,
      "learning_rate": 5.204403962339375e-06,
      "loss": 0.5228,
      "mean_token_accuracy": 0.8348559774458408,
      "num_tokens": 986545092.0,
      "step": 30911
    },
    {
      "entropy": 0.418519401922822,
      "epoch": 2.8440664516017735,
      "grad_norm": 0.14442569017410278,
      "learning_rate": 5.20133713619775e-06,
      "loss": 0.406,
      "mean_token_accuracy": 0.8695583194494247,
      "num_tokens": 986577108.0,
      "step": 30912
    },
    {
      "entropy": 0.5478366203606129,
      "epoch": 2.8441584579732146,
      "grad_norm": 0.16322208940982819,
      "learning_rate": 5.1982703100561235e-06,
      "loss": 0.5327,
      "mean_token_accuracy": 0.8307689689099789,
      "num_tokens": 986608352.0,
      "step": 30913
    },
    {
      "entropy": 0.4998806118965149,
      "epoch": 2.844250464344656,
      "grad_norm": 0.1552649289369583,
      "learning_rate": 5.195203483914497e-06,
      "loss": 0.4845,
      "mean_token_accuracy": 0.842898715287447,
      "num_tokens": 986640069.0,
      "step": 30914
    },
    {
      "entropy": 0.5605117594823241,
      "epoch": 2.8443424707160974,
      "grad_norm": 0.1633223444223404,
      "learning_rate": 5.192136657772871e-06,
      "loss": 0.5455,
      "mean_token_accuracy": 0.8252279572188854,
      "num_tokens": 986672421.0,
      "step": 30915
    },
    {
      "entropy": 0.3671462256461382,
      "epoch": 2.844434477087538,
      "grad_norm": 0.14037632942199707,
      "learning_rate": 5.189069831631245e-06,
      "loss": 0.3574,
      "mean_token_accuracy": 0.8840219713747501,
      "num_tokens": 986704556.0,
      "step": 30916
    },
    {
      "entropy": 0.45863037556409836,
      "epoch": 2.8445264834589796,
      "grad_norm": 0.1446625292301178,
      "learning_rate": 5.186003005489619e-06,
      "loss": 0.4368,
      "mean_token_accuracy": 0.8565450720489025,
      "num_tokens": 986736083.0,
      "step": 30917
    },
    {
      "entropy": 0.4322343006497249,
      "epoch": 2.8446184898304208,
      "grad_norm": 0.14237858355045319,
      "learning_rate": 5.1829361793479925e-06,
      "loss": 0.4094,
      "mean_token_accuracy": 0.8683502189815044,
      "num_tokens": 986768636.0,
      "step": 30918
    },
    {
      "entropy": 0.49322031904011965,
      "epoch": 2.844710496201862,
      "grad_norm": 0.15391692519187927,
      "learning_rate": 5.179869353206367e-06,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8538631796836853,
      "num_tokens": 986800129.0,
      "step": 30919
    },
    {
      "entropy": 0.6043797917664051,
      "epoch": 2.8448025025733035,
      "grad_norm": 0.17508727312088013,
      "learning_rate": 5.176802527064741e-06,
      "loss": 0.5975,
      "mean_token_accuracy": 0.8155517242848873,
      "num_tokens": 986831516.0,
      "step": 30920
    },
    {
      "entropy": 0.5689660022035241,
      "epoch": 2.844894508944744,
      "grad_norm": 0.15918119251728058,
      "learning_rate": 5.173735700923115e-06,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8272027857601643,
      "num_tokens": 986863867.0,
      "step": 30921
    },
    {
      "entropy": 0.49544854555279016,
      "epoch": 2.8449865153161857,
      "grad_norm": 0.1521468311548233,
      "learning_rate": 5.1706688747814885e-06,
      "loss": 0.483,
      "mean_token_accuracy": 0.8484451174736023,
      "num_tokens": 986896489.0,
      "step": 30922
    },
    {
      "entropy": 0.43608041014522314,
      "epoch": 2.845078521687627,
      "grad_norm": 0.14520937204360962,
      "learning_rate": 5.167602048639863e-06,
      "loss": 0.4274,
      "mean_token_accuracy": 0.8659699819982052,
      "num_tokens": 986928775.0,
      "step": 30923
    },
    {
      "entropy": 0.5859024822711945,
      "epoch": 2.845170528059068,
      "grad_norm": 0.1691850870847702,
      "learning_rate": 5.164535222498237e-06,
      "loss": 0.5759,
      "mean_token_accuracy": 0.8203875757753849,
      "num_tokens": 986961177.0,
      "step": 30924
    },
    {
      "entropy": 0.5790256522595882,
      "epoch": 2.8452625344305096,
      "grad_norm": 0.1717074066400528,
      "learning_rate": 5.161468396356611e-06,
      "loss": 0.57,
      "mean_token_accuracy": 0.825512308627367,
      "num_tokens": 986992468.0,
      "step": 30925
    },
    {
      "entropy": 0.5047995245549828,
      "epoch": 2.8453545408019503,
      "grad_norm": 0.15309076011180878,
      "learning_rate": 5.158401570214985e-06,
      "loss": 0.4831,
      "mean_token_accuracy": 0.8499044515192509,
      "num_tokens": 987023992.0,
      "step": 30926
    },
    {
      "entropy": 0.5357163043227047,
      "epoch": 2.845446547173392,
      "grad_norm": 0.16288264095783234,
      "learning_rate": 5.155334744073359e-06,
      "loss": 0.5249,
      "mean_token_accuracy": 0.8346902057528496,
      "num_tokens": 987056108.0,
      "step": 30927
    },
    {
      "entropy": 0.4718465209007263,
      "epoch": 2.845538553544833,
      "grad_norm": 0.15146972239017487,
      "learning_rate": 5.152267917931733e-06,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8510568290948868,
      "num_tokens": 987087937.0,
      "step": 30928
    },
    {
      "entropy": 0.5852206218987703,
      "epoch": 2.845630559916274,
      "grad_norm": 0.1632966846227646,
      "learning_rate": 5.149201091790107e-06,
      "loss": 0.5765,
      "mean_token_accuracy": 0.8181763477623463,
      "num_tokens": 987119482.0,
      "step": 30929
    },
    {
      "entropy": 0.4988195765763521,
      "epoch": 2.8457225662877157,
      "grad_norm": 0.15552230179309845,
      "learning_rate": 5.146134265648481e-06,
      "loss": 0.4875,
      "mean_token_accuracy": 0.8453903719782829,
      "num_tokens": 987151896.0,
      "step": 30930
    },
    {
      "entropy": 0.5835787113755941,
      "epoch": 2.8458145726591564,
      "grad_norm": 0.1694699078798294,
      "learning_rate": 5.1430674395068544e-06,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8210051394999027,
      "num_tokens": 987182791.0,
      "step": 30931
    },
    {
      "entropy": 0.6017418252304196,
      "epoch": 2.845906579030598,
      "grad_norm": 0.16753119230270386,
      "learning_rate": 5.140000613365228e-06,
      "loss": 0.5903,
      "mean_token_accuracy": 0.8125535808503628,
      "num_tokens": 987214994.0,
      "step": 30932
    },
    {
      "entropy": 0.5366645725443959,
      "epoch": 2.845998585402039,
      "grad_norm": 0.1604587435722351,
      "learning_rate": 5.136933787223602e-06,
      "loss": 0.5168,
      "mean_token_accuracy": 0.8338412158191204,
      "num_tokens": 987247184.0,
      "step": 30933
    },
    {
      "entropy": 0.5451708417385817,
      "epoch": 2.8460905917734802,
      "grad_norm": 0.15881657600402832,
      "learning_rate": 5.133866961081976e-06,
      "loss": 0.5347,
      "mean_token_accuracy": 0.829517800360918,
      "num_tokens": 987278827.0,
      "step": 30934
    },
    {
      "entropy": 0.5620360802859068,
      "epoch": 2.846182598144922,
      "grad_norm": 0.16636639833450317,
      "learning_rate": 5.1308001349403505e-06,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8254901245236397,
      "num_tokens": 987310035.0,
      "step": 30935
    },
    {
      "entropy": 0.5448927078396082,
      "epoch": 2.8462746045163625,
      "grad_norm": 0.16362300515174866,
      "learning_rate": 5.127733308798724e-06,
      "loss": 0.5433,
      "mean_token_accuracy": 0.8271367363631725,
      "num_tokens": 987342059.0,
      "step": 30936
    },
    {
      "entropy": 0.5151621089316905,
      "epoch": 2.846366610887804,
      "grad_norm": 0.15852448344230652,
      "learning_rate": 5.124666482657098e-06,
      "loss": 0.5053,
      "mean_token_accuracy": 0.8430433496832848,
      "num_tokens": 987373748.0,
      "step": 30937
    },
    {
      "entropy": 0.5332212708890438,
      "epoch": 2.846458617259245,
      "grad_norm": 0.16177698969841003,
      "learning_rate": 5.121599656515473e-06,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8350173644721508,
      "num_tokens": 987405437.0,
      "step": 30938
    },
    {
      "entropy": 0.5447258832864463,
      "epoch": 2.8465506236306863,
      "grad_norm": 0.15943299233913422,
      "learning_rate": 5.1185328303738465e-06,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8329392112791538,
      "num_tokens": 987437512.0,
      "step": 30939
    },
    {
      "entropy": 0.47161909751594067,
      "epoch": 2.846642630002128,
      "grad_norm": 0.146324023604393,
      "learning_rate": 5.11546600423222e-06,
      "loss": 0.4593,
      "mean_token_accuracy": 0.8541536256670952,
      "num_tokens": 987469559.0,
      "step": 30940
    },
    {
      "entropy": 0.5182045237161219,
      "epoch": 2.8467346363735686,
      "grad_norm": 0.16193273663520813,
      "learning_rate": 5.112399178090594e-06,
      "loss": 0.5113,
      "mean_token_accuracy": 0.8382688499987125,
      "num_tokens": 987501241.0,
      "step": 30941
    },
    {
      "entropy": 0.5425730175338686,
      "epoch": 2.84682664274501,
      "grad_norm": 0.16503208875656128,
      "learning_rate": 5.109332351948969e-06,
      "loss": 0.5327,
      "mean_token_accuracy": 0.8332066163420677,
      "num_tokens": 987532666.0,
      "step": 30942
    },
    {
      "entropy": 0.4345288500189781,
      "epoch": 2.8469186491164513,
      "grad_norm": 0.15165652334690094,
      "learning_rate": 5.106265525807343e-06,
      "loss": 0.4325,
      "mean_token_accuracy": 0.8638653680682182,
      "num_tokens": 987564174.0,
      "step": 30943
    },
    {
      "entropy": 0.4642467638477683,
      "epoch": 2.8470106554878924,
      "grad_norm": 0.1521826535463333,
      "learning_rate": 5.103198699665716e-06,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8550541624426842,
      "num_tokens": 987595953.0,
      "step": 30944
    },
    {
      "entropy": 0.44902859162539244,
      "epoch": 2.847102661859334,
      "grad_norm": 0.14838410913944244,
      "learning_rate": 5.10013187352409e-06,
      "loss": 0.4252,
      "mean_token_accuracy": 0.858080729842186,
      "num_tokens": 987628017.0,
      "step": 30945
    },
    {
      "entropy": 0.4555679988116026,
      "epoch": 2.8471946682307747,
      "grad_norm": 0.1471869796514511,
      "learning_rate": 5.097065047382464e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8645266890525818,
      "num_tokens": 987659796.0,
      "step": 30946
    },
    {
      "entropy": 0.47185801167506725,
      "epoch": 2.8472866746022163,
      "grad_norm": 0.15087087452411652,
      "learning_rate": 5.093998221240838e-06,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8550442755222321,
      "num_tokens": 987691228.0,
      "step": 30947
    },
    {
      "entropy": 0.5119786942377687,
      "epoch": 2.8473786809736574,
      "grad_norm": 0.15457601845264435,
      "learning_rate": 5.090931395099212e-06,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8430260121822357,
      "num_tokens": 987723184.0,
      "step": 30948
    },
    {
      "entropy": 0.555682064499706,
      "epoch": 2.8474706873450986,
      "grad_norm": 0.16228033602237701,
      "learning_rate": 5.087864568957585e-06,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8324959874153137,
      "num_tokens": 987755352.0,
      "step": 30949
    },
    {
      "entropy": 0.4850449487566948,
      "epoch": 2.84756269371654,
      "grad_norm": 0.15572834014892578,
      "learning_rate": 5.08479774281596e-06,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8506049290299416,
      "num_tokens": 987787563.0,
      "step": 30950
    },
    {
      "entropy": 0.49568894726689905,
      "epoch": 2.847654700087981,
      "grad_norm": 0.1580287665128708,
      "learning_rate": 5.081730916674334e-06,
      "loss": 0.4866,
      "mean_token_accuracy": 0.8468748964369297,
      "num_tokens": 987819122.0,
      "step": 30951
    },
    {
      "entropy": 0.4447654103860259,
      "epoch": 2.8477467064594224,
      "grad_norm": 0.14885559678077698,
      "learning_rate": 5.078664090532708e-06,
      "loss": 0.4356,
      "mean_token_accuracy": 0.8592921048402786,
      "num_tokens": 987851684.0,
      "step": 30952
    },
    {
      "entropy": 0.49678147956728935,
      "epoch": 2.8478387128308635,
      "grad_norm": 0.1541690230369568,
      "learning_rate": 5.075597264391082e-06,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8468656986951828,
      "num_tokens": 987884289.0,
      "step": 30953
    },
    {
      "entropy": 0.590606851503253,
      "epoch": 2.8479307192023047,
      "grad_norm": 0.1719757616519928,
      "learning_rate": 5.072530438249456e-06,
      "loss": 0.5811,
      "mean_token_accuracy": 0.818294957280159,
      "num_tokens": 987916240.0,
      "step": 30954
    },
    {
      "entropy": 0.5339004555717111,
      "epoch": 2.8480227255737462,
      "grad_norm": 0.1610279530286789,
      "learning_rate": 5.06946361210783e-06,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8295700736343861,
      "num_tokens": 987947907.0,
      "step": 30955
    },
    {
      "entropy": 0.40451743081212044,
      "epoch": 2.848114731945187,
      "grad_norm": 0.13980720937252045,
      "learning_rate": 5.066396785966204e-06,
      "loss": 0.396,
      "mean_token_accuracy": 0.8732596524059772,
      "num_tokens": 987980675.0,
      "step": 30956
    },
    {
      "entropy": 0.46836667507886887,
      "epoch": 2.8482067383166285,
      "grad_norm": 0.15199804306030273,
      "learning_rate": 5.0633299598245775e-06,
      "loss": 0.4526,
      "mean_token_accuracy": 0.8558648712933064,
      "num_tokens": 988012120.0,
      "step": 30957
    },
    {
      "entropy": 0.5636201724410057,
      "epoch": 2.8482987446880697,
      "grad_norm": 0.16473819315433502,
      "learning_rate": 5.060263133682952e-06,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8230922780930996,
      "num_tokens": 988043646.0,
      "step": 30958
    },
    {
      "entropy": 0.5025576129555702,
      "epoch": 2.848390751059511,
      "grad_norm": 0.15208904445171356,
      "learning_rate": 5.057196307541326e-06,
      "loss": 0.4875,
      "mean_token_accuracy": 0.8409582041203976,
      "num_tokens": 988075628.0,
      "step": 30959
    },
    {
      "entropy": 0.5817444082349539,
      "epoch": 2.8484827574309524,
      "grad_norm": 0.16693459451198578,
      "learning_rate": 5.0541294813997e-06,
      "loss": 0.5679,
      "mean_token_accuracy": 0.8183943592011929,
      "num_tokens": 988107135.0,
      "step": 30960
    },
    {
      "entropy": 0.5069543402642012,
      "epoch": 2.848574763802393,
      "grad_norm": 0.1566770374774933,
      "learning_rate": 5.0510626552580735e-06,
      "loss": 0.4837,
      "mean_token_accuracy": 0.8444923497736454,
      "num_tokens": 988138876.0,
      "step": 30961
    },
    {
      "entropy": 0.5277909748256207,
      "epoch": 2.8486667701738346,
      "grad_norm": 0.15918530523777008,
      "learning_rate": 5.047995829116447e-06,
      "loss": 0.5061,
      "mean_token_accuracy": 0.8349845036864281,
      "num_tokens": 988170867.0,
      "step": 30962
    },
    {
      "entropy": 0.4899155432358384,
      "epoch": 2.8487587765452758,
      "grad_norm": 0.15858207643032074,
      "learning_rate": 5.044929002974821e-06,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8491893857717514,
      "num_tokens": 988202572.0,
      "step": 30963
    },
    {
      "entropy": 0.49879045132547617,
      "epoch": 2.848850782916717,
      "grad_norm": 0.16158771514892578,
      "learning_rate": 5.041862176833195e-06,
      "loss": 0.4939,
      "mean_token_accuracy": 0.845827229321003,
      "num_tokens": 988234632.0,
      "step": 30964
    },
    {
      "entropy": 0.48824515054002404,
      "epoch": 2.8489427892881585,
      "grad_norm": 0.16657011210918427,
      "learning_rate": 5.03879535069157e-06,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8425556682050228,
      "num_tokens": 988265719.0,
      "step": 30965
    },
    {
      "entropy": 0.5493351863697171,
      "epoch": 2.849034795659599,
      "grad_norm": 0.15933960676193237,
      "learning_rate": 5.035728524549943e-06,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8306088484823704,
      "num_tokens": 988298122.0,
      "step": 30966
    },
    {
      "entropy": 0.48818809166550636,
      "epoch": 2.8491268020310407,
      "grad_norm": 0.1564672589302063,
      "learning_rate": 5.032661698408317e-06,
      "loss": 0.4915,
      "mean_token_accuracy": 0.8440530970692635,
      "num_tokens": 988330430.0,
      "step": 30967
    },
    {
      "entropy": 0.42426119931042194,
      "epoch": 2.849218808402482,
      "grad_norm": 0.14728973805904388,
      "learning_rate": 5.029594872266692e-06,
      "loss": 0.4142,
      "mean_token_accuracy": 0.8686725236475468,
      "num_tokens": 988362607.0,
      "step": 30968
    },
    {
      "entropy": 0.6076685329899192,
      "epoch": 2.849310814773923,
      "grad_norm": 0.16505587100982666,
      "learning_rate": 5.026528046125066e-06,
      "loss": 0.5937,
      "mean_token_accuracy": 0.8123616315424442,
      "num_tokens": 988393799.0,
      "step": 30969
    },
    {
      "entropy": 0.5727687645703554,
      "epoch": 2.8494028211453646,
      "grad_norm": 0.16834169626235962,
      "learning_rate": 5.0234612199834394e-06,
      "loss": 0.5694,
      "mean_token_accuracy": 0.8192884735763073,
      "num_tokens": 988425609.0,
      "step": 30970
    },
    {
      "entropy": 0.4965915381908417,
      "epoch": 2.8494948275168053,
      "grad_norm": 0.1580386608839035,
      "learning_rate": 5.020394393841813e-06,
      "loss": 0.4892,
      "mean_token_accuracy": 0.8463782332837582,
      "num_tokens": 988456860.0,
      "step": 30971
    },
    {
      "entropy": 0.4282669806852937,
      "epoch": 2.849586833888247,
      "grad_norm": 0.14606338739395142,
      "learning_rate": 5.017327567700187e-06,
      "loss": 0.4223,
      "mean_token_accuracy": 0.8611077778041363,
      "num_tokens": 988489426.0,
      "step": 30972
    },
    {
      "entropy": 0.41844259668141603,
      "epoch": 2.849678840259688,
      "grad_norm": 0.15240681171417236,
      "learning_rate": 5.014260741558562e-06,
      "loss": 0.4156,
      "mean_token_accuracy": 0.8686834275722504,
      "num_tokens": 988520977.0,
      "step": 30973
    },
    {
      "entropy": 0.44983058515936136,
      "epoch": 2.849770846631129,
      "grad_norm": 0.1518130600452423,
      "learning_rate": 5.0111939154169355e-06,
      "loss": 0.4425,
      "mean_token_accuracy": 0.8563351817429066,
      "num_tokens": 988552509.0,
      "step": 30974
    },
    {
      "entropy": 0.5058556734584272,
      "epoch": 2.8498628530025707,
      "grad_norm": 0.15145958960056305,
      "learning_rate": 5.008127089275309e-06,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8445240743458271,
      "num_tokens": 988584772.0,
      "step": 30975
    },
    {
      "entropy": 0.4838848179206252,
      "epoch": 2.8499548593740114,
      "grad_norm": 0.15281254053115845,
      "learning_rate": 5.005060263133683e-06,
      "loss": 0.4662,
      "mean_token_accuracy": 0.8526947759091854,
      "num_tokens": 988616796.0,
      "step": 30976
    },
    {
      "entropy": 0.5094943647272885,
      "epoch": 2.850046865745453,
      "grad_norm": 0.15584596991539001,
      "learning_rate": 5.001993436992057e-06,
      "loss": 0.505,
      "mean_token_accuracy": 0.8395688496530056,
      "num_tokens": 988648724.0,
      "step": 30977
    },
    {
      "entropy": 0.516405462520197,
      "epoch": 2.850138872116894,
      "grad_norm": 0.15906773507595062,
      "learning_rate": 4.998926610850431e-06,
      "loss": 0.5149,
      "mean_token_accuracy": 0.8400152772665024,
      "num_tokens": 988680021.0,
      "step": 30978
    },
    {
      "entropy": 0.44866085052490234,
      "epoch": 2.8502308784883352,
      "grad_norm": 0.15027713775634766,
      "learning_rate": 4.9958597847088045e-06,
      "loss": 0.424,
      "mean_token_accuracy": 0.8598613999783993,
      "num_tokens": 988711407.0,
      "step": 30979
    },
    {
      "entropy": 0.4459956674836576,
      "epoch": 2.850322884859777,
      "grad_norm": 0.14503946900367737,
      "learning_rate": 4.992792958567179e-06,
      "loss": 0.438,
      "mean_token_accuracy": 0.8609614074230194,
      "num_tokens": 988744009.0,
      "step": 30980
    },
    {
      "entropy": 0.5184634160250425,
      "epoch": 2.8504148912312175,
      "grad_norm": 0.1629553735256195,
      "learning_rate": 4.989726132425553e-06,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8374007977545261,
      "num_tokens": 988776109.0,
      "step": 30981
    },
    {
      "entropy": 0.4347132823895663,
      "epoch": 2.850506897602659,
      "grad_norm": 0.1492612361907959,
      "learning_rate": 4.986659306283928e-06,
      "loss": 0.4275,
      "mean_token_accuracy": 0.8636230900883675,
      "num_tokens": 988808820.0,
      "step": 30982
    },
    {
      "entropy": 0.4928192272782326,
      "epoch": 2.8505989039741,
      "grad_norm": 0.15638679265975952,
      "learning_rate": 4.983592480142301e-06,
      "loss": 0.4753,
      "mean_token_accuracy": 0.8473424054682255,
      "num_tokens": 988840248.0,
      "step": 30983
    },
    {
      "entropy": 0.564812446013093,
      "epoch": 2.8506909103455413,
      "grad_norm": 0.1661076694726944,
      "learning_rate": 4.980525654000675e-06,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8257232010364532,
      "num_tokens": 988872248.0,
      "step": 30984
    },
    {
      "entropy": 0.542037351988256,
      "epoch": 2.850782916716983,
      "grad_norm": 0.16199475526809692,
      "learning_rate": 4.977458827859049e-06,
      "loss": 0.533,
      "mean_token_accuracy": 0.8304618746042252,
      "num_tokens": 988903989.0,
      "step": 30985
    },
    {
      "entropy": 0.7007123343646526,
      "epoch": 2.8508749230884236,
      "grad_norm": 0.18664203584194183,
      "learning_rate": 4.974392001717423e-06,
      "loss": 0.7075,
      "mean_token_accuracy": 0.7781974896788597,
      "num_tokens": 988935964.0,
      "step": 30986
    },
    {
      "entropy": 0.5133791503030807,
      "epoch": 2.850966929459865,
      "grad_norm": 0.15529251098632812,
      "learning_rate": 4.971325175575797e-06,
      "loss": 0.4818,
      "mean_token_accuracy": 0.8447114452719688,
      "num_tokens": 988966961.0,
      "step": 30987
    },
    {
      "entropy": 0.48227667063474655,
      "epoch": 2.8510589358313063,
      "grad_norm": 0.15338031947612762,
      "learning_rate": 4.968258349434171e-06,
      "loss": 0.4664,
      "mean_token_accuracy": 0.849270187318325,
      "num_tokens": 988999270.0,
      "step": 30988
    },
    {
      "entropy": 0.44166705664247274,
      "epoch": 2.8511509422027475,
      "grad_norm": 0.1461368352174759,
      "learning_rate": 4.965191523292545e-06,
      "loss": 0.4257,
      "mean_token_accuracy": 0.8642513677477837,
      "num_tokens": 989031324.0,
      "step": 30989
    },
    {
      "entropy": 0.5405570198781788,
      "epoch": 2.851242948574189,
      "grad_norm": 0.16022464632987976,
      "learning_rate": 4.962124697150919e-06,
      "loss": 0.5319,
      "mean_token_accuracy": 0.8302718214690685,
      "num_tokens": 989063957.0,
      "step": 30990
    },
    {
      "entropy": 0.5208493387326598,
      "epoch": 2.8513349549456297,
      "grad_norm": 0.15868328511714935,
      "learning_rate": 4.959057871009293e-06,
      "loss": 0.5149,
      "mean_token_accuracy": 0.8352275602519512,
      "num_tokens": 989096151.0,
      "step": 30991
    },
    {
      "entropy": 0.4704625648446381,
      "epoch": 2.8514269613170713,
      "grad_norm": 0.15079152584075928,
      "learning_rate": 4.9559910448676664e-06,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8557249791920185,
      "num_tokens": 989128419.0,
      "step": 30992
    },
    {
      "entropy": 0.48116118647158146,
      "epoch": 2.8515189676885124,
      "grad_norm": 0.1550997644662857,
      "learning_rate": 4.95292421872604e-06,
      "loss": 0.4704,
      "mean_token_accuracy": 0.8506804816424847,
      "num_tokens": 989160611.0,
      "step": 30993
    },
    {
      "entropy": 0.5284871580079198,
      "epoch": 2.8516109740599536,
      "grad_norm": 0.16408339142799377,
      "learning_rate": 4.949857392584414e-06,
      "loss": 0.5256,
      "mean_token_accuracy": 0.8342362232506275,
      "num_tokens": 989192188.0,
      "step": 30994
    },
    {
      "entropy": 0.5326850857818499,
      "epoch": 2.851702980431395,
      "grad_norm": 0.15678586065769196,
      "learning_rate": 4.946790566442789e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8343392834067345,
      "num_tokens": 989224744.0,
      "step": 30995
    },
    {
      "entropy": 0.5608168123289943,
      "epoch": 2.851794986802836,
      "grad_norm": 0.16908927261829376,
      "learning_rate": 4.9437237403011625e-06,
      "loss": 0.5578,
      "mean_token_accuracy": 0.8277950137853622,
      "num_tokens": 989256373.0,
      "step": 30996
    },
    {
      "entropy": 0.4758100116159767,
      "epoch": 2.8518869931742774,
      "grad_norm": 0.15511730313301086,
      "learning_rate": 4.940656914159537e-06,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8566987477242947,
      "num_tokens": 989288219.0,
      "step": 30997
    },
    {
      "entropy": 0.5371631495654583,
      "epoch": 2.8519789995457185,
      "grad_norm": 0.16640642285346985,
      "learning_rate": 4.937590088017911e-06,
      "loss": 0.5261,
      "mean_token_accuracy": 0.8319735303521156,
      "num_tokens": 989320623.0,
      "step": 30998
    },
    {
      "entropy": 0.5837738886475563,
      "epoch": 2.8520710059171597,
      "grad_norm": 0.1705716848373413,
      "learning_rate": 4.934523261876285e-06,
      "loss": 0.5787,
      "mean_token_accuracy": 0.8154936023056507,
      "num_tokens": 989351927.0,
      "step": 30999
    },
    {
      "entropy": 0.5337894465774298,
      "epoch": 2.8521630122886013,
      "grad_norm": 0.1649247258901596,
      "learning_rate": 4.9314564357346585e-06,
      "loss": 0.5265,
      "mean_token_accuracy": 0.8330414034426212,
      "num_tokens": 989383573.0,
      "step": 31000
    },
    {
      "entropy": 0.47054633870720863,
      "epoch": 2.852255018660042,
      "grad_norm": 0.1527155339717865,
      "learning_rate": 4.928389609593032e-06,
      "loss": 0.4471,
      "mean_token_accuracy": 0.8545568659901619,
      "num_tokens": 989415890.0,
      "step": 31001
    },
    {
      "entropy": 0.4349274127744138,
      "epoch": 2.8523470250314835,
      "grad_norm": 0.1423732489347458,
      "learning_rate": 4.925322783451406e-06,
      "loss": 0.429,
      "mean_token_accuracy": 0.8642898686230183,
      "num_tokens": 989448473.0,
      "step": 31002
    },
    {
      "entropy": 0.49536374385934323,
      "epoch": 2.8524390314029247,
      "grad_norm": 0.1519639641046524,
      "learning_rate": 4.922255957309781e-06,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8498359359800816,
      "num_tokens": 989480949.0,
      "step": 31003
    },
    {
      "entropy": 0.5244959879200906,
      "epoch": 2.852531037774366,
      "grad_norm": 0.15993647277355194,
      "learning_rate": 4.919189131168155e-06,
      "loss": 0.5047,
      "mean_token_accuracy": 0.8383530899882317,
      "num_tokens": 989512822.0,
      "step": 31004
    },
    {
      "entropy": 0.550473777577281,
      "epoch": 2.8526230441458074,
      "grad_norm": 0.16677889227867126,
      "learning_rate": 4.916122305026528e-06,
      "loss": 0.5449,
      "mean_token_accuracy": 0.8306699357926846,
      "num_tokens": 989545152.0,
      "step": 31005
    },
    {
      "entropy": 0.4639761494472623,
      "epoch": 2.852715050517248,
      "grad_norm": 0.15079861879348755,
      "learning_rate": 4.913055478884902e-06,
      "loss": 0.4497,
      "mean_token_accuracy": 0.8565508350729942,
      "num_tokens": 989576719.0,
      "step": 31006
    },
    {
      "entropy": 0.47761829756200314,
      "epoch": 2.8528070568886896,
      "grad_norm": 0.15250733494758606,
      "learning_rate": 4.909988652743276e-06,
      "loss": 0.4545,
      "mean_token_accuracy": 0.855520274490118,
      "num_tokens": 989609487.0,
      "step": 31007
    },
    {
      "entropy": 0.5068490550620481,
      "epoch": 2.8528990632601308,
      "grad_norm": 0.15620796382427216,
      "learning_rate": 4.90692182660165e-06,
      "loss": 0.5012,
      "mean_token_accuracy": 0.8407004624605179,
      "num_tokens": 989642253.0,
      "step": 31008
    },
    {
      "entropy": 0.5676681715995073,
      "epoch": 2.852991069631572,
      "grad_norm": 0.16637800633907318,
      "learning_rate": 4.903855000460024e-06,
      "loss": 0.5497,
      "mean_token_accuracy": 0.8289338052272797,
      "num_tokens": 989673700.0,
      "step": 31009
    },
    {
      "entropy": 0.5209064157679677,
      "epoch": 2.8530830760030135,
      "grad_norm": 0.16022342443466187,
      "learning_rate": 4.900788174318398e-06,
      "loss": 0.515,
      "mean_token_accuracy": 0.8407990112900734,
      "num_tokens": 989705584.0,
      "step": 31010
    },
    {
      "entropy": 0.5709184929728508,
      "epoch": 2.853175082374454,
      "grad_norm": 0.1630183458328247,
      "learning_rate": 4.897721348176772e-06,
      "loss": 0.5493,
      "mean_token_accuracy": 0.826214287430048,
      "num_tokens": 989737779.0,
      "step": 31011
    },
    {
      "entropy": 0.48557083681225777,
      "epoch": 2.8532670887458957,
      "grad_norm": 0.1536952704191208,
      "learning_rate": 4.894654522035147e-06,
      "loss": 0.4739,
      "mean_token_accuracy": 0.8514500185847282,
      "num_tokens": 989770120.0,
      "step": 31012
    },
    {
      "entropy": 0.48349877819418907,
      "epoch": 2.853359095117337,
      "grad_norm": 0.1499452441930771,
      "learning_rate": 4.8915876958935205e-06,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8506691604852676,
      "num_tokens": 989802243.0,
      "step": 31013
    },
    {
      "entropy": 0.5177935091778636,
      "epoch": 2.853451101488778,
      "grad_norm": 0.1593460738658905,
      "learning_rate": 4.888520869751894e-06,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8400457799434662,
      "num_tokens": 989834395.0,
      "step": 31014
    },
    {
      "entropy": 0.46877338574267924,
      "epoch": 2.8535431078602196,
      "grad_norm": 0.15125198662281036,
      "learning_rate": 4.885454043610268e-06,
      "loss": 0.4584,
      "mean_token_accuracy": 0.8545791171491146,
      "num_tokens": 989866403.0,
      "step": 31015
    },
    {
      "entropy": 0.5276855169795454,
      "epoch": 2.8536351142316603,
      "grad_norm": 0.16113774478435516,
      "learning_rate": 4.882387217468642e-06,
      "loss": 0.5196,
      "mean_token_accuracy": 0.8347262106835842,
      "num_tokens": 989898406.0,
      "step": 31016
    },
    {
      "entropy": 0.5382525231689215,
      "epoch": 2.853727120603102,
      "grad_norm": 0.16379743814468384,
      "learning_rate": 4.879320391327016e-06,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8322954326868057,
      "num_tokens": 989929245.0,
      "step": 31017
    },
    {
      "entropy": 0.5652036406099796,
      "epoch": 2.853819126974543,
      "grad_norm": 0.16694508492946625,
      "learning_rate": 4.8762535651853895e-06,
      "loss": 0.5534,
      "mean_token_accuracy": 0.8266158439218998,
      "num_tokens": 989961475.0,
      "step": 31018
    },
    {
      "entropy": 0.529891368933022,
      "epoch": 2.853911133345984,
      "grad_norm": 0.15801899135112762,
      "learning_rate": 4.873186739043764e-06,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8413974568247795,
      "num_tokens": 989993377.0,
      "step": 31019
    },
    {
      "entropy": 0.43529437901452184,
      "epoch": 2.8540031397174257,
      "grad_norm": 0.1470368355512619,
      "learning_rate": 4.870119912902138e-06,
      "loss": 0.4313,
      "mean_token_accuracy": 0.8646189533174038,
      "num_tokens": 990026145.0,
      "step": 31020
    },
    {
      "entropy": 0.5089855331461877,
      "epoch": 2.8540951460888664,
      "grad_norm": 0.1588105857372284,
      "learning_rate": 4.867053086760512e-06,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8424705676734447,
      "num_tokens": 990058372.0,
      "step": 31021
    },
    {
      "entropy": 0.6070628929883242,
      "epoch": 2.854187152460308,
      "grad_norm": 0.17573325335979462,
      "learning_rate": 4.8639862606188855e-06,
      "loss": 0.6024,
      "mean_token_accuracy": 0.811151310801506,
      "num_tokens": 990089884.0,
      "step": 31022
    },
    {
      "entropy": 0.5363995172083378,
      "epoch": 2.854279158831749,
      "grad_norm": 0.16120144724845886,
      "learning_rate": 4.860919434477259e-06,
      "loss": 0.5256,
      "mean_token_accuracy": 0.831540934741497,
      "num_tokens": 990121482.0,
      "step": 31023
    },
    {
      "entropy": 0.5268956637009978,
      "epoch": 2.8543711652031902,
      "grad_norm": 0.162922203540802,
      "learning_rate": 4.857852608335633e-06,
      "loss": 0.5203,
      "mean_token_accuracy": 0.838768444955349,
      "num_tokens": 990152898.0,
      "step": 31024
    },
    {
      "entropy": 0.44974452443420887,
      "epoch": 2.854463171574632,
      "grad_norm": 0.15155944228172302,
      "learning_rate": 4.854785782194008e-06,
      "loss": 0.4464,
      "mean_token_accuracy": 0.8577988594770432,
      "num_tokens": 990184574.0,
      "step": 31025
    },
    {
      "entropy": 0.45087254513055086,
      "epoch": 2.8545551779460725,
      "grad_norm": 0.15316691994667053,
      "learning_rate": 4.851718956052382e-06,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8587225005030632,
      "num_tokens": 990215894.0,
      "step": 31026
    },
    {
      "entropy": 0.42683859425596893,
      "epoch": 2.854647184317514,
      "grad_norm": 0.14729750156402588,
      "learning_rate": 4.848652129910756e-06,
      "loss": 0.4192,
      "mean_token_accuracy": 0.8670236356556416,
      "num_tokens": 990248167.0,
      "step": 31027
    },
    {
      "entropy": 0.41029771510511637,
      "epoch": 2.854739190688955,
      "grad_norm": 0.14160679280757904,
      "learning_rate": 4.84558530376913e-06,
      "loss": 0.388,
      "mean_token_accuracy": 0.8735663741827011,
      "num_tokens": 990280373.0,
      "step": 31028
    },
    {
      "entropy": 0.5127438995987177,
      "epoch": 2.8548311970603963,
      "grad_norm": 0.15565705299377441,
      "learning_rate": 4.842518477627504e-06,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8390130624175072,
      "num_tokens": 990312595.0,
      "step": 31029
    },
    {
      "entropy": 0.532985508441925,
      "epoch": 2.854923203431838,
      "grad_norm": 0.15982523560523987,
      "learning_rate": 4.839451651485878e-06,
      "loss": 0.5299,
      "mean_token_accuracy": 0.8316979371011257,
      "num_tokens": 990344797.0,
      "step": 31030
    },
    {
      "entropy": 0.46515205316245556,
      "epoch": 2.8550152098032786,
      "grad_norm": 0.1586884707212448,
      "learning_rate": 4.8363848253442514e-06,
      "loss": 0.4537,
      "mean_token_accuracy": 0.8521144203841686,
      "num_tokens": 990376880.0,
      "step": 31031
    },
    {
      "entropy": 0.4768691072240472,
      "epoch": 2.85510721617472,
      "grad_norm": 0.15294261276721954,
      "learning_rate": 4.833317999202625e-06,
      "loss": 0.4619,
      "mean_token_accuracy": 0.8527505435049534,
      "num_tokens": 990409034.0,
      "step": 31032
    },
    {
      "entropy": 0.5586417652666569,
      "epoch": 2.8551992225461613,
      "grad_norm": 0.16204988956451416,
      "learning_rate": 4.830251173060999e-06,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8289080485701561,
      "num_tokens": 990440561.0,
      "step": 31033
    },
    {
      "entropy": 0.4531955267302692,
      "epoch": 2.8552912289176025,
      "grad_norm": 0.15317480266094208,
      "learning_rate": 4.827184346919374e-06,
      "loss": 0.4352,
      "mean_token_accuracy": 0.8622684441506863,
      "num_tokens": 990471482.0,
      "step": 31034
    },
    {
      "entropy": 0.41689923498779535,
      "epoch": 2.855383235289044,
      "grad_norm": 0.1493026465177536,
      "learning_rate": 4.8241175207777475e-06,
      "loss": 0.4005,
      "mean_token_accuracy": 0.8671316765248775,
      "num_tokens": 990502795.0,
      "step": 31035
    },
    {
      "entropy": 0.39969280431978405,
      "epoch": 2.8554752416604847,
      "grad_norm": 0.13844136893749237,
      "learning_rate": 4.821050694636121e-06,
      "loss": 0.3841,
      "mean_token_accuracy": 0.8759171925485134,
      "num_tokens": 990535312.0,
      "step": 31036
    },
    {
      "entropy": 0.5316884499043226,
      "epoch": 2.8555672480319263,
      "grad_norm": 0.16577835381031036,
      "learning_rate": 4.817983868494495e-06,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8345972932875156,
      "num_tokens": 990566791.0,
      "step": 31037
    },
    {
      "entropy": 0.4995403364300728,
      "epoch": 2.8556592544033674,
      "grad_norm": 0.15834350883960724,
      "learning_rate": 4.814917042352869e-06,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8478107750415802,
      "num_tokens": 990598415.0,
      "step": 31038
    },
    {
      "entropy": 0.539911475032568,
      "epoch": 2.8557512607748086,
      "grad_norm": 0.17101843655109406,
      "learning_rate": 4.811850216211243e-06,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8289022408425808,
      "num_tokens": 990628964.0,
      "step": 31039
    },
    {
      "entropy": 0.5456389095634222,
      "epoch": 2.85584326714625,
      "grad_norm": 0.16339664161205292,
      "learning_rate": 4.808783390069617e-06,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8318040743470192,
      "num_tokens": 990660778.0,
      "step": 31040
    },
    {
      "entropy": 0.5112208109349012,
      "epoch": 2.855935273517691,
      "grad_norm": 0.1541031152009964,
      "learning_rate": 4.805716563927991e-06,
      "loss": 0.4994,
      "mean_token_accuracy": 0.8402377590537071,
      "num_tokens": 990693099.0,
      "step": 31041
    },
    {
      "entropy": 0.559214262291789,
      "epoch": 2.8560272798891324,
      "grad_norm": 0.16210679709911346,
      "learning_rate": 4.802649737786366e-06,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8268850184977055,
      "num_tokens": 990725008.0,
      "step": 31042
    },
    {
      "entropy": 0.5670617558062077,
      "epoch": 2.8561192862605735,
      "grad_norm": 0.16493651270866394,
      "learning_rate": 4.79958291164474e-06,
      "loss": 0.5587,
      "mean_token_accuracy": 0.8237275816500187,
      "num_tokens": 990757017.0,
      "step": 31043
    },
    {
      "entropy": 0.4385327701456845,
      "epoch": 2.8562112926320147,
      "grad_norm": 0.14834432303905487,
      "learning_rate": 4.796516085503113e-06,
      "loss": 0.4206,
      "mean_token_accuracy": 0.8611421510577202,
      "num_tokens": 990789228.0,
      "step": 31044
    },
    {
      "entropy": 0.4921995420008898,
      "epoch": 2.8563032990034563,
      "grad_norm": 0.15579791367053986,
      "learning_rate": 4.793449259361487e-06,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8433990739285946,
      "num_tokens": 990820911.0,
      "step": 31045
    },
    {
      "entropy": 0.5400874204933643,
      "epoch": 2.856395305374897,
      "grad_norm": 0.16348996758460999,
      "learning_rate": 4.790382433219861e-06,
      "loss": 0.5264,
      "mean_token_accuracy": 0.83069122210145,
      "num_tokens": 990852864.0,
      "step": 31046
    },
    {
      "entropy": 0.5013998160138726,
      "epoch": 2.8564873117463385,
      "grad_norm": 0.15576520562171936,
      "learning_rate": 4.787315607078235e-06,
      "loss": 0.4946,
      "mean_token_accuracy": 0.8436354026198387,
      "num_tokens": 990885404.0,
      "step": 31047
    },
    {
      "entropy": 0.5068009868264198,
      "epoch": 2.8565793181177797,
      "grad_norm": 0.1578468382358551,
      "learning_rate": 4.784248780936609e-06,
      "loss": 0.496,
      "mean_token_accuracy": 0.8452158756554127,
      "num_tokens": 990917670.0,
      "step": 31048
    },
    {
      "entropy": 0.5806452482938766,
      "epoch": 2.856671324489221,
      "grad_norm": 0.17427457869052887,
      "learning_rate": 4.781181954794983e-06,
      "loss": 0.5744,
      "mean_token_accuracy": 0.8220440968871117,
      "num_tokens": 990949488.0,
      "step": 31049
    },
    {
      "entropy": 0.4546896982938051,
      "epoch": 2.8567633308606624,
      "grad_norm": 0.15475232899188995,
      "learning_rate": 4.778115128653357e-06,
      "loss": 0.4363,
      "mean_token_accuracy": 0.858076985925436,
      "num_tokens": 990981487.0,
      "step": 31050
    },
    {
      "entropy": 0.5254027608316392,
      "epoch": 2.856855337232103,
      "grad_norm": 0.1589546650648117,
      "learning_rate": 4.775048302511731e-06,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8371860012412071,
      "num_tokens": 991013227.0,
      "step": 31051
    },
    {
      "entropy": 0.5268791913986206,
      "epoch": 2.8569473436035446,
      "grad_norm": 0.16174374520778656,
      "learning_rate": 4.771981476370105e-06,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8310419246554375,
      "num_tokens": 991044697.0,
      "step": 31052
    },
    {
      "entropy": 0.4813780691474676,
      "epoch": 2.8570393499749858,
      "grad_norm": 0.15142759680747986,
      "learning_rate": 4.7689146502284784e-06,
      "loss": 0.4688,
      "mean_token_accuracy": 0.8499045707285404,
      "num_tokens": 991077021.0,
      "step": 31053
    },
    {
      "entropy": 0.5461976602673531,
      "epoch": 2.857131356346427,
      "grad_norm": 0.1635885387659073,
      "learning_rate": 4.765847824086852e-06,
      "loss": 0.5229,
      "mean_token_accuracy": 0.8356252126395702,
      "num_tokens": 991108209.0,
      "step": 31054
    },
    {
      "entropy": 0.4847530536353588,
      "epoch": 2.8572233627178685,
      "grad_norm": 0.15683631598949432,
      "learning_rate": 4.762780997945227e-06,
      "loss": 0.4732,
      "mean_token_accuracy": 0.8481242917478085,
      "num_tokens": 991140643.0,
      "step": 31055
    },
    {
      "entropy": 0.511363735422492,
      "epoch": 2.857315369089309,
      "grad_norm": 0.15451449155807495,
      "learning_rate": 4.759714171803601e-06,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8420250900089741,
      "num_tokens": 991172413.0,
      "step": 31056
    },
    {
      "entropy": 0.5301727056503296,
      "epoch": 2.8574073754607507,
      "grad_norm": 0.16422878205776215,
      "learning_rate": 4.756647345661975e-06,
      "loss": 0.5242,
      "mean_token_accuracy": 0.8341356478631496,
      "num_tokens": 991203574.0,
      "step": 31057
    },
    {
      "entropy": 0.5401285728439689,
      "epoch": 2.857499381832192,
      "grad_norm": 0.16268035769462585,
      "learning_rate": 4.753580519520349e-06,
      "loss": 0.5302,
      "mean_token_accuracy": 0.8344483859837055,
      "num_tokens": 991235849.0,
      "step": 31058
    },
    {
      "entropy": 0.4797625527717173,
      "epoch": 2.857591388203633,
      "grad_norm": 0.1543111354112625,
      "learning_rate": 4.750513693378723e-06,
      "loss": 0.4648,
      "mean_token_accuracy": 0.8515814542770386,
      "num_tokens": 991267678.0,
      "step": 31059
    },
    {
      "entropy": 0.49122488335706294,
      "epoch": 2.8576833945750746,
      "grad_norm": 0.16026058793067932,
      "learning_rate": 4.747446867237097e-06,
      "loss": 0.4822,
      "mean_token_accuracy": 0.8479299694299698,
      "num_tokens": 991298893.0,
      "step": 31060
    },
    {
      "entropy": 0.46189178340137005,
      "epoch": 2.8577754009465153,
      "grad_norm": 0.1461651474237442,
      "learning_rate": 4.7443800410954705e-06,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8580897264182568,
      "num_tokens": 991330954.0,
      "step": 31061
    },
    {
      "entropy": 0.527945261914283,
      "epoch": 2.857867407317957,
      "grad_norm": 0.1572139859199524,
      "learning_rate": 4.741313214953844e-06,
      "loss": 0.5125,
      "mean_token_accuracy": 0.8394239470362663,
      "num_tokens": 991362884.0,
      "step": 31062
    },
    {
      "entropy": 0.6012300318107009,
      "epoch": 2.857959413689398,
      "grad_norm": 0.17082789540290833,
      "learning_rate": 4.738246388812218e-06,
      "loss": 0.594,
      "mean_token_accuracy": 0.8129290267825127,
      "num_tokens": 991394091.0,
      "step": 31063
    },
    {
      "entropy": 0.4055753326974809,
      "epoch": 2.858051420060839,
      "grad_norm": 0.14148376882076263,
      "learning_rate": 4.735179562670593e-06,
      "loss": 0.3879,
      "mean_token_accuracy": 0.8746793009340763,
      "num_tokens": 991425423.0,
      "step": 31064
    },
    {
      "entropy": 0.5383847178891301,
      "epoch": 2.8581434264322807,
      "grad_norm": 0.16192257404327393,
      "learning_rate": 4.732112736528967e-06,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8354639410972595,
      "num_tokens": 991456202.0,
      "step": 31065
    },
    {
      "entropy": 0.5089679416269064,
      "epoch": 2.8582354328037214,
      "grad_norm": 0.1575816571712494,
      "learning_rate": 4.72904591038734e-06,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8414457477629185,
      "num_tokens": 991488709.0,
      "step": 31066
    },
    {
      "entropy": 0.5890535376966,
      "epoch": 2.858327439175163,
      "grad_norm": 0.16368858516216278,
      "learning_rate": 4.725979084245714e-06,
      "loss": 0.5803,
      "mean_token_accuracy": 0.818593941628933,
      "num_tokens": 991521212.0,
      "step": 31067
    },
    {
      "entropy": 0.5104729700833559,
      "epoch": 2.858419445546604,
      "grad_norm": 0.15910600125789642,
      "learning_rate": 4.722912258104088e-06,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8388135954737663,
      "num_tokens": 991553530.0,
      "step": 31068
    },
    {
      "entropy": 0.4688290925696492,
      "epoch": 2.8585114519180452,
      "grad_norm": 0.1553010195493698,
      "learning_rate": 4.719845431962462e-06,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8510657399892807,
      "num_tokens": 991585379.0,
      "step": 31069
    },
    {
      "entropy": 0.5829480839893222,
      "epoch": 2.858603458289487,
      "grad_norm": 0.16891026496887207,
      "learning_rate": 4.7167786058208364e-06,
      "loss": 0.5726,
      "mean_token_accuracy": 0.821544948965311,
      "num_tokens": 991617337.0,
      "step": 31070
    },
    {
      "entropy": 0.5818604566156864,
      "epoch": 2.8586954646609275,
      "grad_norm": 0.1699824184179306,
      "learning_rate": 4.71371177967921e-06,
      "loss": 0.5745,
      "mean_token_accuracy": 0.8150949478149414,
      "num_tokens": 991648513.0,
      "step": 31071
    },
    {
      "entropy": 0.55419209622778,
      "epoch": 2.858787471032369,
      "grad_norm": 0.16863074898719788,
      "learning_rate": 4.710644953537585e-06,
      "loss": 0.5488,
      "mean_token_accuracy": 0.8261091485619545,
      "num_tokens": 991680157.0,
      "step": 31072
    },
    {
      "entropy": 0.48226687405258417,
      "epoch": 2.85887947740381,
      "grad_norm": 0.15463723242282867,
      "learning_rate": 4.707578127395959e-06,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8517268858850002,
      "num_tokens": 991711819.0,
      "step": 31073
    },
    {
      "entropy": 0.4477140768431127,
      "epoch": 2.8589714837752513,
      "grad_norm": 0.1507241278886795,
      "learning_rate": 4.7045113012543325e-06,
      "loss": 0.4395,
      "mean_token_accuracy": 0.8612537086009979,
      "num_tokens": 991743750.0,
      "step": 31074
    },
    {
      "entropy": 0.577746294438839,
      "epoch": 2.859063490146693,
      "grad_norm": 0.16514775156974792,
      "learning_rate": 4.701444475112706e-06,
      "loss": 0.5598,
      "mean_token_accuracy": 0.8224420882761478,
      "num_tokens": 991775804.0,
      "step": 31075
    },
    {
      "entropy": 0.39578627794981003,
      "epoch": 2.8591554965181336,
      "grad_norm": 0.14046919345855713,
      "learning_rate": 4.69837764897108e-06,
      "loss": 0.3817,
      "mean_token_accuracy": 0.8788251988589764,
      "num_tokens": 991808317.0,
      "step": 31076
    },
    {
      "entropy": 0.4970679492689669,
      "epoch": 2.859247502889575,
      "grad_norm": 0.14795717597007751,
      "learning_rate": 4.695310822829454e-06,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8515340201556683,
      "num_tokens": 991840445.0,
      "step": 31077
    },
    {
      "entropy": 0.4163001247216016,
      "epoch": 2.8593395092610163,
      "grad_norm": 0.1454896181821823,
      "learning_rate": 4.692243996687828e-06,
      "loss": 0.4079,
      "mean_token_accuracy": 0.8700380958616734,
      "num_tokens": 991872421.0,
      "step": 31078
    },
    {
      "entropy": 0.4247060150373727,
      "epoch": 2.8594315156324575,
      "grad_norm": 0.148522287607193,
      "learning_rate": 4.6891771705462015e-06,
      "loss": 0.4144,
      "mean_token_accuracy": 0.8662765510380268,
      "num_tokens": 991904554.0,
      "step": 31079
    },
    {
      "entropy": 0.4947613161057234,
      "epoch": 2.859523522003899,
      "grad_norm": 0.154581218957901,
      "learning_rate": 4.686110344404576e-06,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8493479155004025,
      "num_tokens": 991936181.0,
      "step": 31080
    },
    {
      "entropy": 0.37764533795416355,
      "epoch": 2.8596155283753397,
      "grad_norm": 0.13984721899032593,
      "learning_rate": 4.68304351826295e-06,
      "loss": 0.3663,
      "mean_token_accuracy": 0.880608294159174,
      "num_tokens": 991968060.0,
      "step": 31081
    },
    {
      "entropy": 0.5015263753011823,
      "epoch": 2.8597075347467813,
      "grad_norm": 0.157615527510643,
      "learning_rate": 4.679976692121324e-06,
      "loss": 0.4948,
      "mean_token_accuracy": 0.8437376581132412,
      "num_tokens": 992000089.0,
      "step": 31082
    },
    {
      "entropy": 0.6554372403770685,
      "epoch": 2.8597995411182224,
      "grad_norm": 0.17785866558551788,
      "learning_rate": 4.6769098659796975e-06,
      "loss": 0.6423,
      "mean_token_accuracy": 0.7967857718467712,
      "num_tokens": 992031946.0,
      "step": 31083
    },
    {
      "entropy": 0.4786383459577337,
      "epoch": 2.8598915474896636,
      "grad_norm": 0.1474701464176178,
      "learning_rate": 4.673843039838071e-06,
      "loss": 0.4588,
      "mean_token_accuracy": 0.851359162479639,
      "num_tokens": 992064366.0,
      "step": 31084
    },
    {
      "entropy": 0.44758815318346024,
      "epoch": 2.859983553861105,
      "grad_norm": 0.14628925919532776,
      "learning_rate": 4.670776213696446e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8613411784172058,
      "num_tokens": 992096566.0,
      "step": 31085
    },
    {
      "entropy": 0.479907903354615,
      "epoch": 2.860075560232546,
      "grad_norm": 0.15366660058498383,
      "learning_rate": 4.66770938755482e-06,
      "loss": 0.4662,
      "mean_token_accuracy": 0.8539710268378258,
      "num_tokens": 992128503.0,
      "step": 31086
    },
    {
      "entropy": 0.5056688347831368,
      "epoch": 2.8601675666039874,
      "grad_norm": 0.1561715453863144,
      "learning_rate": 4.6646425614131944e-06,
      "loss": 0.5055,
      "mean_token_accuracy": 0.8381306678056717,
      "num_tokens": 992160426.0,
      "step": 31087
    },
    {
      "entropy": 0.4940162915736437,
      "epoch": 2.8602595729754285,
      "grad_norm": 0.1572146862745285,
      "learning_rate": 4.661575735271568e-06,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8457158729434013,
      "num_tokens": 992192590.0,
      "step": 31088
    },
    {
      "entropy": 0.5595698347315192,
      "epoch": 2.8603515793468697,
      "grad_norm": 0.16504010558128357,
      "learning_rate": 4.658508909129942e-06,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8253695592284203,
      "num_tokens": 992224203.0,
      "step": 31089
    },
    {
      "entropy": 0.5342556359246373,
      "epoch": 2.8604435857183113,
      "grad_norm": 0.15780609846115112,
      "learning_rate": 4.655442082988316e-06,
      "loss": 0.5207,
      "mean_token_accuracy": 0.837006215006113,
      "num_tokens": 992256534.0,
      "step": 31090
    },
    {
      "entropy": 0.5168205564841628,
      "epoch": 2.860535592089752,
      "grad_norm": 0.1552872359752655,
      "learning_rate": 4.65237525684669e-06,
      "loss": 0.5081,
      "mean_token_accuracy": 0.8377760760486126,
      "num_tokens": 992288946.0,
      "step": 31091
    },
    {
      "entropy": 0.5286414548754692,
      "epoch": 2.8606275984611935,
      "grad_norm": 0.1570819616317749,
      "learning_rate": 4.6493084307050634e-06,
      "loss": 0.5139,
      "mean_token_accuracy": 0.8386033177375793,
      "num_tokens": 992321300.0,
      "step": 31092
    },
    {
      "entropy": 0.46595456916838884,
      "epoch": 2.8607196048326347,
      "grad_norm": 0.15313021838665009,
      "learning_rate": 4.646241604563437e-06,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8555036336183548,
      "num_tokens": 992353534.0,
      "step": 31093
    },
    {
      "entropy": 0.4778445940464735,
      "epoch": 2.860811611204076,
      "grad_norm": 0.15471641719341278,
      "learning_rate": 4.643174778421811e-06,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8490543626248837,
      "num_tokens": 992385431.0,
      "step": 31094
    },
    {
      "entropy": 0.5399540718644857,
      "epoch": 2.8609036175755174,
      "grad_norm": 0.16406121850013733,
      "learning_rate": 4.640107952280186e-06,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8323011882603168,
      "num_tokens": 992417624.0,
      "step": 31095
    },
    {
      "entropy": 0.5136526860296726,
      "epoch": 2.860995623946958,
      "grad_norm": 0.15497708320617676,
      "learning_rate": 4.6370411261385595e-06,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8455179333686829,
      "num_tokens": 992450319.0,
      "step": 31096
    },
    {
      "entropy": 0.5173973073251545,
      "epoch": 2.8610876303183996,
      "grad_norm": 0.15912386775016785,
      "learning_rate": 4.633974299996933e-06,
      "loss": 0.5022,
      "mean_token_accuracy": 0.838664073497057,
      "num_tokens": 992482150.0,
      "step": 31097
    },
    {
      "entropy": 0.4514890182763338,
      "epoch": 2.8611796366898408,
      "grad_norm": 0.14864304661750793,
      "learning_rate": 4.630907473855307e-06,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8562837950885296,
      "num_tokens": 992514067.0,
      "step": 31098
    },
    {
      "entropy": 0.5334462523460388,
      "epoch": 2.861271643061282,
      "grad_norm": 0.16807545721530914,
      "learning_rate": 4.627840647713681e-06,
      "loss": 0.5375,
      "mean_token_accuracy": 0.830092404037714,
      "num_tokens": 992545099.0,
      "step": 31099
    },
    {
      "entropy": 0.578710999339819,
      "epoch": 2.8613636494327235,
      "grad_norm": 0.16759027540683746,
      "learning_rate": 4.6247738215720555e-06,
      "loss": 0.5742,
      "mean_token_accuracy": 0.8195499517023563,
      "num_tokens": 992576925.0,
      "step": 31100
    },
    {
      "entropy": 0.4867170648649335,
      "epoch": 2.861455655804164,
      "grad_norm": 0.15763068199157715,
      "learning_rate": 4.621706995430429e-06,
      "loss": 0.476,
      "mean_token_accuracy": 0.8452002815902233,
      "num_tokens": 992609693.0,
      "step": 31101
    },
    {
      "entropy": 0.4876565276645124,
      "epoch": 2.8615476621756057,
      "grad_norm": 0.1544945389032364,
      "learning_rate": 4.618640169288804e-06,
      "loss": 0.4752,
      "mean_token_accuracy": 0.847437396645546,
      "num_tokens": 992642051.0,
      "step": 31102
    },
    {
      "entropy": 0.49015711434185505,
      "epoch": 2.861639668547047,
      "grad_norm": 0.15285071730613708,
      "learning_rate": 4.615573343147178e-06,
      "loss": 0.4824,
      "mean_token_accuracy": 0.8486115597188473,
      "num_tokens": 992674526.0,
      "step": 31103
    },
    {
      "entropy": 0.46890577115118504,
      "epoch": 2.861731674918488,
      "grad_norm": 0.15765032172203064,
      "learning_rate": 4.612506517005552e-06,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8535914793610573,
      "num_tokens": 992705662.0,
      "step": 31104
    },
    {
      "entropy": 0.4732665382325649,
      "epoch": 2.8618236812899296,
      "grad_norm": 0.16020740568637848,
      "learning_rate": 4.609439690863925e-06,
      "loss": 0.4729,
      "mean_token_accuracy": 0.8488015234470367,
      "num_tokens": 992737201.0,
      "step": 31105
    },
    {
      "entropy": 0.5188466385006905,
      "epoch": 2.8619156876613703,
      "grad_norm": 0.16173464059829712,
      "learning_rate": 4.606372864722299e-06,
      "loss": 0.5077,
      "mean_token_accuracy": 0.8360721357166767,
      "num_tokens": 992768992.0,
      "step": 31106
    },
    {
      "entropy": 0.518133794888854,
      "epoch": 2.862007694032812,
      "grad_norm": 0.15939296782016754,
      "learning_rate": 4.603306038580673e-06,
      "loss": 0.5163,
      "mean_token_accuracy": 0.835816103965044,
      "num_tokens": 992800859.0,
      "step": 31107
    },
    {
      "entropy": 0.42818605387583375,
      "epoch": 2.862099700404253,
      "grad_norm": 0.14099228382110596,
      "learning_rate": 4.600239212439047e-06,
      "loss": 0.4086,
      "mean_token_accuracy": 0.8696761354804039,
      "num_tokens": 992833063.0,
      "step": 31108
    },
    {
      "entropy": 0.5008003041148186,
      "epoch": 2.862191706775694,
      "grad_norm": 0.16431300342082977,
      "learning_rate": 4.597172386297421e-06,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8391229510307312,
      "num_tokens": 992864127.0,
      "step": 31109
    },
    {
      "entropy": 0.4711577710695565,
      "epoch": 2.8622837131471357,
      "grad_norm": 0.15447933971881866,
      "learning_rate": 4.594105560155795e-06,
      "loss": 0.453,
      "mean_token_accuracy": 0.8542061634361744,
      "num_tokens": 992896468.0,
      "step": 31110
    },
    {
      "entropy": 0.6042974080191925,
      "epoch": 2.8623757195185764,
      "grad_norm": 0.17175105214118958,
      "learning_rate": 4.591038734014169e-06,
      "loss": 0.5854,
      "mean_token_accuracy": 0.8174034915864468,
      "num_tokens": 992927174.0,
      "step": 31111
    },
    {
      "entropy": 0.5080666914582253,
      "epoch": 2.862467725890018,
      "grad_norm": 0.15924930572509766,
      "learning_rate": 4.587971907872543e-06,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8456603959202766,
      "num_tokens": 992958794.0,
      "step": 31112
    },
    {
      "entropy": 0.4161553392186761,
      "epoch": 2.862559732261459,
      "grad_norm": 0.14455954730510712,
      "learning_rate": 4.584905081730917e-06,
      "loss": 0.4071,
      "mean_token_accuracy": 0.8701988309621811,
      "num_tokens": 992991218.0,
      "step": 31113
    },
    {
      "entropy": 0.5811319500207901,
      "epoch": 2.8626517386329002,
      "grad_norm": 0.16908983886241913,
      "learning_rate": 4.5818382555892904e-06,
      "loss": 0.563,
      "mean_token_accuracy": 0.8198605068027973,
      "num_tokens": 993022738.0,
      "step": 31114
    },
    {
      "entropy": 0.53909044386819,
      "epoch": 2.862743745004342,
      "grad_norm": 0.15696609020233154,
      "learning_rate": 4.578771429447665e-06,
      "loss": 0.5284,
      "mean_token_accuracy": 0.8317051120102406,
      "num_tokens": 993054941.0,
      "step": 31115
    },
    {
      "entropy": 0.5518376813270152,
      "epoch": 2.8628357513757825,
      "grad_norm": 0.16156868636608124,
      "learning_rate": 4.575704603306039e-06,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8265420533716679,
      "num_tokens": 993087224.0,
      "step": 31116
    },
    {
      "entropy": 0.5016796831041574,
      "epoch": 2.862927757747224,
      "grad_norm": 0.1540202796459198,
      "learning_rate": 4.572637777164413e-06,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8461192958056927,
      "num_tokens": 993119011.0,
      "step": 31117
    },
    {
      "entropy": 0.5113821607083082,
      "epoch": 2.863019764118665,
      "grad_norm": 0.1583355963230133,
      "learning_rate": 4.569570951022787e-06,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8418862260878086,
      "num_tokens": 993150888.0,
      "step": 31118
    },
    {
      "entropy": 0.46798832388594747,
      "epoch": 2.8631117704901063,
      "grad_norm": 0.14965605735778809,
      "learning_rate": 4.566504124881161e-06,
      "loss": 0.457,
      "mean_token_accuracy": 0.8539325781166553,
      "num_tokens": 993183656.0,
      "step": 31119
    },
    {
      "entropy": 0.4314656755886972,
      "epoch": 2.863203776861548,
      "grad_norm": 0.1493433564901352,
      "learning_rate": 4.563437298739535e-06,
      "loss": 0.429,
      "mean_token_accuracy": 0.8610068745911121,
      "num_tokens": 993216095.0,
      "step": 31120
    },
    {
      "entropy": 0.4943870007991791,
      "epoch": 2.8632957832329886,
      "grad_norm": 0.1554824411869049,
      "learning_rate": 4.560370472597909e-06,
      "loss": 0.476,
      "mean_token_accuracy": 0.8500502593815327,
      "num_tokens": 993247777.0,
      "step": 31121
    },
    {
      "entropy": 0.5167360045015812,
      "epoch": 2.86338778960443,
      "grad_norm": 0.15812593698501587,
      "learning_rate": 4.5573036464562825e-06,
      "loss": 0.5102,
      "mean_token_accuracy": 0.8362532742321491,
      "num_tokens": 993279980.0,
      "step": 31122
    },
    {
      "entropy": 0.5195717420428991,
      "epoch": 2.8634797959758713,
      "grad_norm": 0.15769334137439728,
      "learning_rate": 4.554236820314656e-06,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8387519903481007,
      "num_tokens": 993311769.0,
      "step": 31123
    },
    {
      "entropy": 0.47329605533741415,
      "epoch": 2.8635718023473125,
      "grad_norm": 0.1518731266260147,
      "learning_rate": 4.55116999417303e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8523741774260998,
      "num_tokens": 993343841.0,
      "step": 31124
    },
    {
      "entropy": 0.5027115782722831,
      "epoch": 2.863663808718754,
      "grad_norm": 0.15586495399475098,
      "learning_rate": 4.548103168031405e-06,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8463596925139427,
      "num_tokens": 993375987.0,
      "step": 31125
    },
    {
      "entropy": 0.522064640186727,
      "epoch": 2.8637558150901947,
      "grad_norm": 0.16448117792606354,
      "learning_rate": 4.545036341889779e-06,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8369849286973476,
      "num_tokens": 993407598.0,
      "step": 31126
    },
    {
      "entropy": 0.5601519662886858,
      "epoch": 2.8638478214616363,
      "grad_norm": 0.16386888921260834,
      "learning_rate": 4.541969515748152e-06,
      "loss": 0.5499,
      "mean_token_accuracy": 0.829575277864933,
      "num_tokens": 993440115.0,
      "step": 31127
    },
    {
      "entropy": 0.527539744740352,
      "epoch": 2.8639398278330774,
      "grad_norm": 0.16112814843654633,
      "learning_rate": 4.538902689606526e-06,
      "loss": 0.5189,
      "mean_token_accuracy": 0.8361978530883789,
      "num_tokens": 993471930.0,
      "step": 31128
    },
    {
      "entropy": 0.4574029939249158,
      "epoch": 2.8640318342045186,
      "grad_norm": 0.14385585486888885,
      "learning_rate": 4.5358358634649e-06,
      "loss": 0.4383,
      "mean_token_accuracy": 0.8613579720258713,
      "num_tokens": 993503719.0,
      "step": 31129
    },
    {
      "entropy": 0.36533514596521854,
      "epoch": 2.86412384057596,
      "grad_norm": 0.14026188850402832,
      "learning_rate": 4.532769037323275e-06,
      "loss": 0.3657,
      "mean_token_accuracy": 0.8835528157651424,
      "num_tokens": 993535917.0,
      "step": 31130
    },
    {
      "entropy": 0.4742334047332406,
      "epoch": 2.864215846947401,
      "grad_norm": 0.1567487120628357,
      "learning_rate": 4.5297022111816484e-06,
      "loss": 0.4683,
      "mean_token_accuracy": 0.8514313064515591,
      "num_tokens": 993567550.0,
      "step": 31131
    },
    {
      "entropy": 0.5481370938941836,
      "epoch": 2.8643078533188424,
      "grad_norm": 0.16437716782093048,
      "learning_rate": 4.526635385040022e-06,
      "loss": 0.5425,
      "mean_token_accuracy": 0.8300848342478275,
      "num_tokens": 993599143.0,
      "step": 31132
    },
    {
      "entropy": 0.5107259079813957,
      "epoch": 2.8643998596902835,
      "grad_norm": 0.160638228058815,
      "learning_rate": 4.523568558898397e-06,
      "loss": 0.4959,
      "mean_token_accuracy": 0.8417050838470459,
      "num_tokens": 993630587.0,
      "step": 31133
    },
    {
      "entropy": 0.45521863643079996,
      "epoch": 2.8644918660617247,
      "grad_norm": 0.1571221500635147,
      "learning_rate": 4.520501732756771e-06,
      "loss": 0.4446,
      "mean_token_accuracy": 0.8572830483317375,
      "num_tokens": 993661961.0,
      "step": 31134
    },
    {
      "entropy": 0.4654705449938774,
      "epoch": 2.8645838724331663,
      "grad_norm": 0.14726361632347107,
      "learning_rate": 4.5174349066151445e-06,
      "loss": 0.4424,
      "mean_token_accuracy": 0.8602954223752022,
      "num_tokens": 993693932.0,
      "step": 31135
    },
    {
      "entropy": 0.5180297889746726,
      "epoch": 2.864675878804607,
      "grad_norm": 0.15924593806266785,
      "learning_rate": 4.514368080473518e-06,
      "loss": 0.5102,
      "mean_token_accuracy": 0.8414258025586605,
      "num_tokens": 993726035.0,
      "step": 31136
    },
    {
      "entropy": 0.43400405254215,
      "epoch": 2.8647678851760485,
      "grad_norm": 0.1457241028547287,
      "learning_rate": 4.511301254331892e-06,
      "loss": 0.4206,
      "mean_token_accuracy": 0.8638955913484097,
      "num_tokens": 993757846.0,
      "step": 31137
    },
    {
      "entropy": 0.5512167329434305,
      "epoch": 2.8648598915474897,
      "grad_norm": 0.16050654649734497,
      "learning_rate": 4.508234428190266e-06,
      "loss": 0.5358,
      "mean_token_accuracy": 0.8343472927808762,
      "num_tokens": 993790431.0,
      "step": 31138
    },
    {
      "entropy": 0.5514167062938213,
      "epoch": 2.864951897918931,
      "grad_norm": 0.1644747257232666,
      "learning_rate": 4.50516760204864e-06,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8261870332062244,
      "num_tokens": 993822891.0,
      "step": 31139
    },
    {
      "entropy": 0.5609780326485634,
      "epoch": 2.8650439042903724,
      "grad_norm": 0.1693408340215683,
      "learning_rate": 4.5021007759070135e-06,
      "loss": 0.5472,
      "mean_token_accuracy": 0.824471477419138,
      "num_tokens": 993854872.0,
      "step": 31140
    },
    {
      "entropy": 0.44001428748015314,
      "epoch": 2.865135910661813,
      "grad_norm": 0.1517821103334427,
      "learning_rate": 4.499033949765388e-06,
      "loss": 0.4264,
      "mean_token_accuracy": 0.864250160753727,
      "num_tokens": 993886162.0,
      "step": 31141
    },
    {
      "entropy": 0.37735517090186477,
      "epoch": 2.8652279170332546,
      "grad_norm": 0.13976401090621948,
      "learning_rate": 4.495967123623762e-06,
      "loss": 0.3655,
      "mean_token_accuracy": 0.8838309645652771,
      "num_tokens": 993918159.0,
      "step": 31142
    },
    {
      "entropy": 0.43241212412249297,
      "epoch": 2.8653199234046958,
      "grad_norm": 0.1511191427707672,
      "learning_rate": 4.492900297482136e-06,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8703251741826534,
      "num_tokens": 993950545.0,
      "step": 31143
    },
    {
      "entropy": 0.540729841333814,
      "epoch": 2.865411929776137,
      "grad_norm": 0.15910036861896515,
      "learning_rate": 4.4898334713405095e-06,
      "loss": 0.532,
      "mean_token_accuracy": 0.8308380097150803,
      "num_tokens": 993982648.0,
      "step": 31144
    },
    {
      "entropy": 0.3794414948206395,
      "epoch": 2.8655039361475785,
      "grad_norm": 0.14573197066783905,
      "learning_rate": 4.486766645198884e-06,
      "loss": 0.373,
      "mean_token_accuracy": 0.8784600049257278,
      "num_tokens": 994014705.0,
      "step": 31145
    },
    {
      "entropy": 0.4002938694320619,
      "epoch": 2.865595942519019,
      "grad_norm": 0.14101608097553253,
      "learning_rate": 4.483699819057258e-06,
      "loss": 0.3889,
      "mean_token_accuracy": 0.8715782053768635,
      "num_tokens": 994046939.0,
      "step": 31146
    },
    {
      "entropy": 0.5030062645673752,
      "epoch": 2.8656879488904607,
      "grad_norm": 0.15754054486751556,
      "learning_rate": 4.480632992915632e-06,
      "loss": 0.4944,
      "mean_token_accuracy": 0.8429109454154968,
      "num_tokens": 994078188.0,
      "step": 31147
    },
    {
      "entropy": 0.450618636328727,
      "epoch": 2.865779955261902,
      "grad_norm": 0.14844457805156708,
      "learning_rate": 4.4775661667740064e-06,
      "loss": 0.4353,
      "mean_token_accuracy": 0.8599964007735252,
      "num_tokens": 994110221.0,
      "step": 31148
    },
    {
      "entropy": 0.5733611546456814,
      "epoch": 2.865871961633343,
      "grad_norm": 0.1643732637166977,
      "learning_rate": 4.47449934063238e-06,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8271585702896118,
      "num_tokens": 994142462.0,
      "step": 31149
    },
    {
      "entropy": 0.501433159224689,
      "epoch": 2.8659639680047846,
      "grad_norm": 0.15726013481616974,
      "learning_rate": 4.471432514490754e-06,
      "loss": 0.489,
      "mean_token_accuracy": 0.8437099829316139,
      "num_tokens": 994174056.0,
      "step": 31150
    },
    {
      "entropy": 0.46389450691640377,
      "epoch": 2.8660559743762253,
      "grad_norm": 0.14622513949871063,
      "learning_rate": 4.468365688349128e-06,
      "loss": 0.4511,
      "mean_token_accuracy": 0.8568523600697517,
      "num_tokens": 994206764.0,
      "step": 31151
    },
    {
      "entropy": 0.4379460201598704,
      "epoch": 2.866147980747667,
      "grad_norm": 0.14367727935314178,
      "learning_rate": 4.465298862207502e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8637102730572224,
      "num_tokens": 994238903.0,
      "step": 31152
    },
    {
      "entropy": 0.4665264915674925,
      "epoch": 2.866239987119108,
      "grad_norm": 0.1532783806324005,
      "learning_rate": 4.4622320360658754e-06,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8527999669313431,
      "num_tokens": 994271224.0,
      "step": 31153
    },
    {
      "entropy": 0.5307346172630787,
      "epoch": 2.866331993490549,
      "grad_norm": 0.15931133925914764,
      "learning_rate": 4.459165209924249e-06,
      "loss": 0.5107,
      "mean_token_accuracy": 0.8375858552753925,
      "num_tokens": 994303418.0,
      "step": 31154
    },
    {
      "entropy": 0.48048109002411366,
      "epoch": 2.8664239998619907,
      "grad_norm": 0.15326614677906036,
      "learning_rate": 4.456098383782623e-06,
      "loss": 0.475,
      "mean_token_accuracy": 0.8488607481122017,
      "num_tokens": 994335267.0,
      "step": 31155
    },
    {
      "entropy": 0.5773802220355719,
      "epoch": 2.8665160062334314,
      "grad_norm": 0.16777516901493073,
      "learning_rate": 4.453031557640998e-06,
      "loss": 0.5702,
      "mean_token_accuracy": 0.8209003508090973,
      "num_tokens": 994366776.0,
      "step": 31156
    },
    {
      "entropy": 0.5791630540043116,
      "epoch": 2.866608012604873,
      "grad_norm": 0.16563016176223755,
      "learning_rate": 4.4499647314993715e-06,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8174149245023727,
      "num_tokens": 994398653.0,
      "step": 31157
    },
    {
      "entropy": 0.5244216402061284,
      "epoch": 2.866700018976314,
      "grad_norm": 0.16178365051746368,
      "learning_rate": 4.446897905357745e-06,
      "loss": 0.5068,
      "mean_token_accuracy": 0.8354716412723064,
      "num_tokens": 994429796.0,
      "step": 31158
    },
    {
      "entropy": 0.5638384632766247,
      "epoch": 2.8667920253477552,
      "grad_norm": 0.1631016731262207,
      "learning_rate": 4.443831079216119e-06,
      "loss": 0.5471,
      "mean_token_accuracy": 0.8272509276866913,
      "num_tokens": 994461905.0,
      "step": 31159
    },
    {
      "entropy": 0.4813322573900223,
      "epoch": 2.866884031719197,
      "grad_norm": 0.1603105366230011,
      "learning_rate": 4.440764253074494e-06,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8495865911245346,
      "num_tokens": 994494071.0,
      "step": 31160
    },
    {
      "entropy": 0.4825950332451612,
      "epoch": 2.8669760380906375,
      "grad_norm": 0.15436050295829773,
      "learning_rate": 4.4376974269328675e-06,
      "loss": 0.4717,
      "mean_token_accuracy": 0.8527932092547417,
      "num_tokens": 994525955.0,
      "step": 31161
    },
    {
      "entropy": 0.5472350250929594,
      "epoch": 2.867068044462079,
      "grad_norm": 0.16501256823539734,
      "learning_rate": 4.434630600791241e-06,
      "loss": 0.5379,
      "mean_token_accuracy": 0.827666699886322,
      "num_tokens": 994558001.0,
      "step": 31162
    },
    {
      "entropy": 0.5730064231902361,
      "epoch": 2.86716005083352,
      "grad_norm": 0.1604122519493103,
      "learning_rate": 4.431563774649616e-06,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8250283002853394,
      "num_tokens": 994589950.0,
      "step": 31163
    },
    {
      "entropy": 0.4792739152908325,
      "epoch": 2.8672520572049613,
      "grad_norm": 0.15368078649044037,
      "learning_rate": 4.42849694850799e-06,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8541169352829456,
      "num_tokens": 994622159.0,
      "step": 31164
    },
    {
      "entropy": 0.4364418974146247,
      "epoch": 2.867344063576403,
      "grad_norm": 0.1481766402721405,
      "learning_rate": 4.4254301223663636e-06,
      "loss": 0.4134,
      "mean_token_accuracy": 0.864356990903616,
      "num_tokens": 994653390.0,
      "step": 31165
    },
    {
      "entropy": 0.5446920283138752,
      "epoch": 2.8674360699478436,
      "grad_norm": 0.16236884891986847,
      "learning_rate": 4.422363296224737e-06,
      "loss": 0.5348,
      "mean_token_accuracy": 0.8290444500744343,
      "num_tokens": 994685661.0,
      "step": 31166
    },
    {
      "entropy": 0.5005400106310844,
      "epoch": 2.867528076319285,
      "grad_norm": 0.15641766786575317,
      "learning_rate": 4.419296470083111e-06,
      "loss": 0.4922,
      "mean_token_accuracy": 0.843978438526392,
      "num_tokens": 994717707.0,
      "step": 31167
    },
    {
      "entropy": 0.4922049278393388,
      "epoch": 2.8676200826907263,
      "grad_norm": 0.1563546359539032,
      "learning_rate": 4.416229643941485e-06,
      "loss": 0.4723,
      "mean_token_accuracy": 0.849853090941906,
      "num_tokens": 994749893.0,
      "step": 31168
    },
    {
      "entropy": 0.49468292854726315,
      "epoch": 2.8677120890621675,
      "grad_norm": 0.1550363004207611,
      "learning_rate": 4.413162817799859e-06,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8507157452404499,
      "num_tokens": 994781442.0,
      "step": 31169
    },
    {
      "entropy": 0.4466555565595627,
      "epoch": 2.867804095433609,
      "grad_norm": 0.15237437188625336,
      "learning_rate": 4.410095991658233e-06,
      "loss": 0.4328,
      "mean_token_accuracy": 0.859869722276926,
      "num_tokens": 994812958.0,
      "step": 31170
    },
    {
      "entropy": 0.5645270240493119,
      "epoch": 2.8678961018050497,
      "grad_norm": 0.16049183905124664,
      "learning_rate": 4.407029165516607e-06,
      "loss": 0.5479,
      "mean_token_accuracy": 0.8274504467844963,
      "num_tokens": 994845222.0,
      "step": 31171
    },
    {
      "entropy": 0.46273063588887453,
      "epoch": 2.8679881081764913,
      "grad_norm": 0.1511777639389038,
      "learning_rate": 4.403962339374981e-06,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8588530272245407,
      "num_tokens": 994877613.0,
      "step": 31172
    },
    {
      "entropy": 0.4332657316699624,
      "epoch": 2.8680801145479324,
      "grad_norm": 0.14450110495090485,
      "learning_rate": 4.400895513233355e-06,
      "loss": 0.4155,
      "mean_token_accuracy": 0.8659929595887661,
      "num_tokens": 994909330.0,
      "step": 31173
    },
    {
      "entropy": 0.3774621128104627,
      "epoch": 2.8681721209193736,
      "grad_norm": 0.13700631260871887,
      "learning_rate": 4.397828687091729e-06,
      "loss": 0.3672,
      "mean_token_accuracy": 0.8801240846514702,
      "num_tokens": 994941215.0,
      "step": 31174
    },
    {
      "entropy": 0.49674852658063173,
      "epoch": 2.868264127290815,
      "grad_norm": 0.15518184006214142,
      "learning_rate": 4.394761860950103e-06,
      "loss": 0.4913,
      "mean_token_accuracy": 0.841999888420105,
      "num_tokens": 994973475.0,
      "step": 31175
    },
    {
      "entropy": 0.6033667726442218,
      "epoch": 2.868356133662256,
      "grad_norm": 0.1724766194820404,
      "learning_rate": 4.391695034808477e-06,
      "loss": 0.5903,
      "mean_token_accuracy": 0.8144565187394619,
      "num_tokens": 995005240.0,
      "step": 31176
    },
    {
      "entropy": 0.5294983700150624,
      "epoch": 2.8684481400336974,
      "grad_norm": 0.15795862674713135,
      "learning_rate": 4.388628208666851e-06,
      "loss": 0.517,
      "mean_token_accuracy": 0.8394545949995518,
      "num_tokens": 995037639.0,
      "step": 31177
    },
    {
      "entropy": 0.438827978214249,
      "epoch": 2.8685401464051385,
      "grad_norm": 0.14420434832572937,
      "learning_rate": 4.385561382525225e-06,
      "loss": 0.4262,
      "mean_token_accuracy": 0.8635871820151806,
      "num_tokens": 995070114.0,
      "step": 31178
    },
    {
      "entropy": 0.5736258905380964,
      "epoch": 2.8686321527765797,
      "grad_norm": 0.16471712291240692,
      "learning_rate": 4.382494556383599e-06,
      "loss": 0.5547,
      "mean_token_accuracy": 0.8254453167319298,
      "num_tokens": 995102280.0,
      "step": 31179
    },
    {
      "entropy": 0.5775042856112123,
      "epoch": 2.8687241591480213,
      "grad_norm": 0.15810614824295044,
      "learning_rate": 4.379427730241973e-06,
      "loss": 0.5541,
      "mean_token_accuracy": 0.822974544018507,
      "num_tokens": 995134125.0,
      "step": 31180
    },
    {
      "entropy": 0.538669393863529,
      "epoch": 2.868816165519462,
      "grad_norm": 0.164259135723114,
      "learning_rate": 4.376360904100347e-06,
      "loss": 0.5316,
      "mean_token_accuracy": 0.8351107649505138,
      "num_tokens": 995165442.0,
      "step": 31181
    },
    {
      "entropy": 0.602238648571074,
      "epoch": 2.8689081718909035,
      "grad_norm": 0.17015811800956726,
      "learning_rate": 4.373294077958721e-06,
      "loss": 0.5991,
      "mean_token_accuracy": 0.8162700310349464,
      "num_tokens": 995197266.0,
      "step": 31182
    },
    {
      "entropy": 0.4945082003250718,
      "epoch": 2.8690001782623447,
      "grad_norm": 0.1572244018316269,
      "learning_rate": 4.3702272518170945e-06,
      "loss": 0.4887,
      "mean_token_accuracy": 0.8445051237940788,
      "num_tokens": 995229067.0,
      "step": 31183
    },
    {
      "entropy": 0.45686858240514994,
      "epoch": 2.869092184633786,
      "grad_norm": 0.15147337317466736,
      "learning_rate": 4.367160425675468e-06,
      "loss": 0.4559,
      "mean_token_accuracy": 0.8547775708138943,
      "num_tokens": 995261517.0,
      "step": 31184
    },
    {
      "entropy": 0.523355164565146,
      "epoch": 2.8691841910052274,
      "grad_norm": 0.1593364179134369,
      "learning_rate": 4.364093599533842e-06,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8389433361589909,
      "num_tokens": 995293418.0,
      "step": 31185
    },
    {
      "entropy": 0.46084903134033084,
      "epoch": 2.869276197376668,
      "grad_norm": 0.14749407768249512,
      "learning_rate": 4.361026773392217e-06,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8561536781489849,
      "num_tokens": 995325667.0,
      "step": 31186
    },
    {
      "entropy": 0.5364325419068336,
      "epoch": 2.8693682037481096,
      "grad_norm": 0.16770818829536438,
      "learning_rate": 4.3579599472505906e-06,
      "loss": 0.5222,
      "mean_token_accuracy": 0.8316379189491272,
      "num_tokens": 995357402.0,
      "step": 31187
    },
    {
      "entropy": 0.4593327927868813,
      "epoch": 2.8694602101195508,
      "grad_norm": 0.14878985285758972,
      "learning_rate": 4.354893121108964e-06,
      "loss": 0.4478,
      "mean_token_accuracy": 0.856392253190279,
      "num_tokens": 995389426.0,
      "step": 31188
    },
    {
      "entropy": 0.5622357139363885,
      "epoch": 2.869552216490992,
      "grad_norm": 0.16093938052654266,
      "learning_rate": 4.351826294967338e-06,
      "loss": 0.5519,
      "mean_token_accuracy": 0.8292459100484848,
      "num_tokens": 995421673.0,
      "step": 31189
    },
    {
      "entropy": 0.5136206168681383,
      "epoch": 2.8696442228624335,
      "grad_norm": 0.15643104910850525,
      "learning_rate": 4.348759468825713e-06,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8409014604985714,
      "num_tokens": 995453532.0,
      "step": 31190
    },
    {
      "entropy": 0.4840678363107145,
      "epoch": 2.869736229233874,
      "grad_norm": 0.15183672308921814,
      "learning_rate": 4.345692642684087e-06,
      "loss": 0.4731,
      "mean_token_accuracy": 0.849112655967474,
      "num_tokens": 995485562.0,
      "step": 31191
    },
    {
      "entropy": 0.5525637315586209,
      "epoch": 2.8698282356053157,
      "grad_norm": 0.1678321361541748,
      "learning_rate": 4.3426258165424604e-06,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8281722739338875,
      "num_tokens": 995516478.0,
      "step": 31192
    },
    {
      "entropy": 0.5738853290677071,
      "epoch": 2.869920241976757,
      "grad_norm": 0.1682407259941101,
      "learning_rate": 4.339558990400834e-06,
      "loss": 0.5646,
      "mean_token_accuracy": 0.8226166591048241,
      "num_tokens": 995548660.0,
      "step": 31193
    },
    {
      "entropy": 0.5057942103594542,
      "epoch": 2.870012248348198,
      "grad_norm": 0.1583353877067566,
      "learning_rate": 4.336492164259209e-06,
      "loss": 0.4943,
      "mean_token_accuracy": 0.8384686186909676,
      "num_tokens": 995580717.0,
      "step": 31194
    },
    {
      "entropy": 0.4865074912086129,
      "epoch": 2.8701042547196396,
      "grad_norm": 0.1570555716753006,
      "learning_rate": 4.333425338117583e-06,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8499159775674343,
      "num_tokens": 995612708.0,
      "step": 31195
    },
    {
      "entropy": 0.5578481141710654,
      "epoch": 2.8701962610910803,
      "grad_norm": 0.1603919416666031,
      "learning_rate": 4.3303585119759565e-06,
      "loss": 0.5427,
      "mean_token_accuracy": 0.82754822447896,
      "num_tokens": 995644743.0,
      "step": 31196
    },
    {
      "entropy": 0.5155012449249625,
      "epoch": 2.870288267462522,
      "grad_norm": 0.16049902141094208,
      "learning_rate": 4.32729168583433e-06,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8417127393186092,
      "num_tokens": 995676776.0,
      "step": 31197
    },
    {
      "entropy": 0.4790263641625643,
      "epoch": 2.870380273833963,
      "grad_norm": 0.15070278942584991,
      "learning_rate": 4.324224859692704e-06,
      "loss": 0.4586,
      "mean_token_accuracy": 0.853542648255825,
      "num_tokens": 995708696.0,
      "step": 31198
    },
    {
      "entropy": 0.413134076166898,
      "epoch": 2.870472280205404,
      "grad_norm": 0.14436504244804382,
      "learning_rate": 4.321158033551078e-06,
      "loss": 0.4019,
      "mean_token_accuracy": 0.8676903620362282,
      "num_tokens": 995740531.0,
      "step": 31199
    },
    {
      "entropy": 0.39088229509070516,
      "epoch": 2.8705642865768457,
      "grad_norm": 0.14023731648921967,
      "learning_rate": 4.318091207409452e-06,
      "loss": 0.3806,
      "mean_token_accuracy": 0.87609026953578,
      "num_tokens": 995772920.0,
      "step": 31200
    },
    {
      "entropy": 0.49430989602115005,
      "epoch": 2.8706562929482864,
      "grad_norm": 0.15551888942718506,
      "learning_rate": 4.3150243812678255e-06,
      "loss": 0.476,
      "mean_token_accuracy": 0.8490261808037758,
      "num_tokens": 995804663.0,
      "step": 31201
    },
    {
      "entropy": 0.52553017786704,
      "epoch": 2.870748299319728,
      "grad_norm": 0.15745334327220917,
      "learning_rate": 4.3119575551262e-06,
      "loss": 0.5113,
      "mean_token_accuracy": 0.8358984179794788,
      "num_tokens": 995836627.0,
      "step": 31202
    },
    {
      "entropy": 0.44715092750266194,
      "epoch": 2.870840305691169,
      "grad_norm": 0.14927688241004944,
      "learning_rate": 4.308890728984574e-06,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8637816272675991,
      "num_tokens": 995868071.0,
      "step": 31203
    },
    {
      "entropy": 0.45800512097775936,
      "epoch": 2.8709323120626102,
      "grad_norm": 0.1475718468427658,
      "learning_rate": 4.305823902842948e-06,
      "loss": 0.4377,
      "mean_token_accuracy": 0.857784878462553,
      "num_tokens": 995899551.0,
      "step": 31204
    },
    {
      "entropy": 0.5379366278648376,
      "epoch": 2.871024318434052,
      "grad_norm": 0.1620662808418274,
      "learning_rate": 4.302757076701322e-06,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8328554220497608,
      "num_tokens": 995931572.0,
      "step": 31205
    },
    {
      "entropy": 0.4652764042839408,
      "epoch": 2.8711163248054925,
      "grad_norm": 0.15103226900100708,
      "learning_rate": 4.299690250559696e-06,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8524115905165672,
      "num_tokens": 995963612.0,
      "step": 31206
    },
    {
      "entropy": 0.5590747017413378,
      "epoch": 2.871208331176934,
      "grad_norm": 0.16561615467071533,
      "learning_rate": 4.29662342441807e-06,
      "loss": 0.546,
      "mean_token_accuracy": 0.8259571045637131,
      "num_tokens": 995995419.0,
      "step": 31207
    },
    {
      "entropy": 0.49856501538306475,
      "epoch": 2.871300337548375,
      "grad_norm": 0.1621222347021103,
      "learning_rate": 4.293556598276444e-06,
      "loss": 0.4879,
      "mean_token_accuracy": 0.8480130322277546,
      "num_tokens": 996026946.0,
      "step": 31208
    },
    {
      "entropy": 0.5491717983968556,
      "epoch": 2.8713923439198163,
      "grad_norm": 0.1648215800523758,
      "learning_rate": 4.290489772134818e-06,
      "loss": 0.5312,
      "mean_token_accuracy": 0.8291566446423531,
      "num_tokens": 996057895.0,
      "step": 31209
    },
    {
      "entropy": 0.5520942201837897,
      "epoch": 2.871484350291258,
      "grad_norm": 0.16221830248832703,
      "learning_rate": 4.287422945993192e-06,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8292748108506203,
      "num_tokens": 996090112.0,
      "step": 31210
    },
    {
      "entropy": 0.6238908190280199,
      "epoch": 2.8715763566626986,
      "grad_norm": 0.16874873638153076,
      "learning_rate": 4.284356119851566e-06,
      "loss": 0.6177,
      "mean_token_accuracy": 0.8025118373334408,
      "num_tokens": 996121778.0,
      "step": 31211
    },
    {
      "entropy": 0.5588373923674226,
      "epoch": 2.87166836303414,
      "grad_norm": 0.16055719554424286,
      "learning_rate": 4.28128929370994e-06,
      "loss": 0.5473,
      "mean_token_accuracy": 0.828247245401144,
      "num_tokens": 996153927.0,
      "step": 31212
    },
    {
      "entropy": 0.5264856908470392,
      "epoch": 2.8717603694055813,
      "grad_norm": 0.15518039464950562,
      "learning_rate": 4.278222467568314e-06,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8416153118014336,
      "num_tokens": 996186408.0,
      "step": 31213
    },
    {
      "entropy": 0.5317849311977625,
      "epoch": 2.8718523757770225,
      "grad_norm": 0.16391511261463165,
      "learning_rate": 4.2751556414266874e-06,
      "loss": 0.5242,
      "mean_token_accuracy": 0.8332283087074757,
      "num_tokens": 996218047.0,
      "step": 31214
    },
    {
      "entropy": 0.5392787461169064,
      "epoch": 2.871944382148464,
      "grad_norm": 0.16367653012275696,
      "learning_rate": 4.272088815285061e-06,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8310276195406914,
      "num_tokens": 996248828.0,
      "step": 31215
    },
    {
      "entropy": 0.43933171639218926,
      "epoch": 2.8720363885199047,
      "grad_norm": 0.15139394998550415,
      "learning_rate": 4.269021989143435e-06,
      "loss": 0.4239,
      "mean_token_accuracy": 0.8609170690178871,
      "num_tokens": 996280489.0,
      "step": 31216
    },
    {
      "entropy": 0.5655048210173845,
      "epoch": 2.8721283948913463,
      "grad_norm": 0.16033893823623657,
      "learning_rate": 4.26595516300181e-06,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8285175822675228,
      "num_tokens": 996312612.0,
      "step": 31217
    },
    {
      "entropy": 0.4265790390782058,
      "epoch": 2.8722204012627874,
      "grad_norm": 0.1493423581123352,
      "learning_rate": 4.2628883368601835e-06,
      "loss": 0.4206,
      "mean_token_accuracy": 0.8661077208817005,
      "num_tokens": 996344750.0,
      "step": 31218
    },
    {
      "entropy": 0.4763551829382777,
      "epoch": 2.8723124076342286,
      "grad_norm": 0.1558840274810791,
      "learning_rate": 4.259821510718557e-06,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8481355607509613,
      "num_tokens": 996376183.0,
      "step": 31219
    },
    {
      "entropy": 0.4970233915373683,
      "epoch": 2.87240441400567,
      "grad_norm": 0.15635336935520172,
      "learning_rate": 4.256754684576932e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8435080423951149,
      "num_tokens": 996408576.0,
      "step": 31220
    },
    {
      "entropy": 0.4794151643291116,
      "epoch": 2.8724964203771113,
      "grad_norm": 0.15485717356204987,
      "learning_rate": 4.253687858435306e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8531177155673504,
      "num_tokens": 996441021.0,
      "step": 31221
    },
    {
      "entropy": 0.45831229351460934,
      "epoch": 2.8725884267485524,
      "grad_norm": 0.14997999370098114,
      "learning_rate": 4.2506210322936795e-06,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8541617952287197,
      "num_tokens": 996473249.0,
      "step": 31222
    },
    {
      "entropy": 0.6029296461492777,
      "epoch": 2.8726804331199935,
      "grad_norm": 0.1659042090177536,
      "learning_rate": 4.247554206152053e-06,
      "loss": 0.5927,
      "mean_token_accuracy": 0.8156997710466385,
      "num_tokens": 996505697.0,
      "step": 31223
    },
    {
      "entropy": 0.5786794554442167,
      "epoch": 2.8727724394914347,
      "grad_norm": 0.16864532232284546,
      "learning_rate": 4.244487380010428e-06,
      "loss": 0.5655,
      "mean_token_accuracy": 0.8222114406526089,
      "num_tokens": 996537664.0,
      "step": 31224
    },
    {
      "entropy": 0.5361781022511423,
      "epoch": 2.8728644458628763,
      "grad_norm": 0.15700846910476685,
      "learning_rate": 4.241420553868802e-06,
      "loss": 0.5254,
      "mean_token_accuracy": 0.8306274004280567,
      "num_tokens": 996569915.0,
      "step": 31225
    },
    {
      "entropy": 0.4807998249307275,
      "epoch": 2.8729564522343174,
      "grad_norm": 0.15181167423725128,
      "learning_rate": 4.2383537277271756e-06,
      "loss": 0.4699,
      "mean_token_accuracy": 0.8478300496935844,
      "num_tokens": 996601892.0,
      "step": 31226
    },
    {
      "entropy": 0.4857446225360036,
      "epoch": 2.8730484586057585,
      "grad_norm": 0.15585225820541382,
      "learning_rate": 4.235286901585549e-06,
      "loss": 0.4812,
      "mean_token_accuracy": 0.8457190692424774,
      "num_tokens": 996633974.0,
      "step": 31227
    },
    {
      "entropy": 0.5513904355466366,
      "epoch": 2.8731404649771997,
      "grad_norm": 0.16239911317825317,
      "learning_rate": 4.232220075443923e-06,
      "loss": 0.5556,
      "mean_token_accuracy": 0.8291348218917847,
      "num_tokens": 996666479.0,
      "step": 31228
    },
    {
      "entropy": 0.594289917498827,
      "epoch": 2.873232471348641,
      "grad_norm": 0.17117489874362946,
      "learning_rate": 4.229153249302297e-06,
      "loss": 0.5982,
      "mean_token_accuracy": 0.8122495785355568,
      "num_tokens": 996698991.0,
      "step": 31229
    },
    {
      "entropy": 0.43455158127471805,
      "epoch": 2.8733244777200824,
      "grad_norm": 0.14761847257614136,
      "learning_rate": 4.226086423160671e-06,
      "loss": 0.4233,
      "mean_token_accuracy": 0.8634544536471367,
      "num_tokens": 996731180.0,
      "step": 31230
    },
    {
      "entropy": 0.4865298082586378,
      "epoch": 2.8734164840915235,
      "grad_norm": 0.15601183474063873,
      "learning_rate": 4.2230195970190446e-06,
      "loss": 0.4792,
      "mean_token_accuracy": 0.8450096361339092,
      "num_tokens": 996763258.0,
      "step": 31231
    },
    {
      "entropy": 0.3780389765743166,
      "epoch": 2.8735084904629646,
      "grad_norm": 0.13284042477607727,
      "learning_rate": 4.219952770877419e-06,
      "loss": 0.3704,
      "mean_token_accuracy": 0.8820306845009327,
      "num_tokens": 996795878.0,
      "step": 31232
    },
    {
      "entropy": 0.4830935993231833,
      "epoch": 2.8736004968344058,
      "grad_norm": 0.1589847356081009,
      "learning_rate": 4.216885944735793e-06,
      "loss": 0.4805,
      "mean_token_accuracy": 0.8485394828021526,
      "num_tokens": 996828050.0,
      "step": 31233
    },
    {
      "entropy": 0.5342163443565369,
      "epoch": 2.873692503205847,
      "grad_norm": 0.1621285229921341,
      "learning_rate": 4.213819118594167e-06,
      "loss": 0.5256,
      "mean_token_accuracy": 0.8334249034523964,
      "num_tokens": 996859820.0,
      "step": 31234
    },
    {
      "entropy": 0.45839804131537676,
      "epoch": 2.8737845095772885,
      "grad_norm": 0.15086272358894348,
      "learning_rate": 4.2107522924525415e-06,
      "loss": 0.4402,
      "mean_token_accuracy": 0.8574959337711334,
      "num_tokens": 996891738.0,
      "step": 31235
    },
    {
      "entropy": 0.492339376360178,
      "epoch": 2.8738765159487296,
      "grad_norm": 0.15063709020614624,
      "learning_rate": 4.207685466310915e-06,
      "loss": 0.48,
      "mean_token_accuracy": 0.8476428911089897,
      "num_tokens": 996924506.0,
      "step": 31236
    },
    {
      "entropy": 0.5705893728882074,
      "epoch": 2.8739685223201707,
      "grad_norm": 0.16704565286636353,
      "learning_rate": 4.204618640169289e-06,
      "loss": 0.5667,
      "mean_token_accuracy": 0.819702573120594,
      "num_tokens": 996956025.0,
      "step": 31237
    },
    {
      "entropy": 0.42346640583127737,
      "epoch": 2.874060528691612,
      "grad_norm": 0.14510485529899597,
      "learning_rate": 4.201551814027663e-06,
      "loss": 0.408,
      "mean_token_accuracy": 0.867692906409502,
      "num_tokens": 996988227.0,
      "step": 31238
    },
    {
      "entropy": 0.614421758800745,
      "epoch": 2.874152535063053,
      "grad_norm": 0.16994427144527435,
      "learning_rate": 4.198484987886037e-06,
      "loss": 0.6025,
      "mean_token_accuracy": 0.8106539286673069,
      "num_tokens": 997020437.0,
      "step": 31239
    },
    {
      "entropy": 0.5144102121703327,
      "epoch": 2.8742445414344946,
      "grad_norm": 0.15147651731967926,
      "learning_rate": 4.195418161744411e-06,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8406936638057232,
      "num_tokens": 997052267.0,
      "step": 31240
    },
    {
      "entropy": 0.521929457783699,
      "epoch": 2.8743365478059357,
      "grad_norm": 0.16289810836315155,
      "learning_rate": 4.192351335602785e-06,
      "loss": 0.5048,
      "mean_token_accuracy": 0.8410388678312302,
      "num_tokens": 997082941.0,
      "step": 31241
    },
    {
      "entropy": 0.5551186744123697,
      "epoch": 2.874428554177377,
      "grad_norm": 0.15773098170757294,
      "learning_rate": 4.189284509461159e-06,
      "loss": 0.5418,
      "mean_token_accuracy": 0.8300276771187782,
      "num_tokens": 997114897.0,
      "step": 31242
    },
    {
      "entropy": 0.6316133057698607,
      "epoch": 2.874520560548818,
      "grad_norm": 0.1747281700372696,
      "learning_rate": 4.186217683319533e-06,
      "loss": 0.6107,
      "mean_token_accuracy": 0.8068725913763046,
      "num_tokens": 997146321.0,
      "step": 31243
    },
    {
      "entropy": 0.523737951181829,
      "epoch": 2.874612566920259,
      "grad_norm": 0.1602228879928589,
      "learning_rate": 4.1831508571779065e-06,
      "loss": 0.5027,
      "mean_token_accuracy": 0.8405417539179325,
      "num_tokens": 997177729.0,
      "step": 31244
    },
    {
      "entropy": 0.5394886825233698,
      "epoch": 2.8747045732917007,
      "grad_norm": 0.16572622954845428,
      "learning_rate": 4.18008403103628e-06,
      "loss": 0.5248,
      "mean_token_accuracy": 0.831692535430193,
      "num_tokens": 997208984.0,
      "step": 31245
    },
    {
      "entropy": 0.4060301352292299,
      "epoch": 2.874796579663142,
      "grad_norm": 0.14368648827075958,
      "learning_rate": 4.177017204894654e-06,
      "loss": 0.3993,
      "mean_token_accuracy": 0.8690629042685032,
      "num_tokens": 997240649.0,
      "step": 31246
    },
    {
      "entropy": 0.5048107001930475,
      "epoch": 2.874888586034583,
      "grad_norm": 0.15899351239204407,
      "learning_rate": 4.173950378753029e-06,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8425822965800762,
      "num_tokens": 997272892.0,
      "step": 31247
    },
    {
      "entropy": 0.4615498948842287,
      "epoch": 2.874980592406024,
      "grad_norm": 0.15122875571250916,
      "learning_rate": 4.1708835526114026e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8548302948474884,
      "num_tokens": 997305272.0,
      "step": 31248
    },
    {
      "entropy": 0.5335235139355063,
      "epoch": 2.8750725987774652,
      "grad_norm": 0.15999968349933624,
      "learning_rate": 4.167816726469776e-06,
      "loss": 0.5291,
      "mean_token_accuracy": 0.8336076363921165,
      "num_tokens": 997337462.0,
      "step": 31249
    },
    {
      "entropy": 0.5110153055284172,
      "epoch": 2.875164605148907,
      "grad_norm": 0.15331892669200897,
      "learning_rate": 4.164749900328151e-06,
      "loss": 0.5054,
      "mean_token_accuracy": 0.8400247246026993,
      "num_tokens": 997369633.0,
      "step": 31250
    },
    {
      "entropy": 0.5344690689817071,
      "epoch": 2.875256611520348,
      "grad_norm": 0.15900681912899017,
      "learning_rate": 4.161683074186525e-06,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8355192691087723,
      "num_tokens": 997401871.0,
      "step": 31251
    },
    {
      "entropy": 0.5533622878137976,
      "epoch": 2.875348617891789,
      "grad_norm": 0.16592776775360107,
      "learning_rate": 4.158616248044899e-06,
      "loss": 0.543,
      "mean_token_accuracy": 0.8315554298460484,
      "num_tokens": 997433562.0,
      "step": 31252
    },
    {
      "entropy": 0.49935809848830104,
      "epoch": 2.87544062426323,
      "grad_norm": 0.15741579234600067,
      "learning_rate": 4.155549421903272e-06,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8469501435756683,
      "num_tokens": 997465598.0,
      "step": 31253
    },
    {
      "entropy": 0.5242416225373745,
      "epoch": 2.8755326306346713,
      "grad_norm": 0.1573609560728073,
      "learning_rate": 4.152482595761646e-06,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8376751691102982,
      "num_tokens": 997497696.0,
      "step": 31254
    },
    {
      "entropy": 0.5249722963199019,
      "epoch": 2.875624637006113,
      "grad_norm": 0.15829946100711823,
      "learning_rate": 4.149415769620021e-06,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8352988213300705,
      "num_tokens": 997530023.0,
      "step": 31255
    },
    {
      "entropy": 0.4910608073696494,
      "epoch": 2.875716643377554,
      "grad_norm": 0.1577412486076355,
      "learning_rate": 4.146348943478395e-06,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8453929089009762,
      "num_tokens": 997562710.0,
      "step": 31256
    },
    {
      "entropy": 0.5786822235677391,
      "epoch": 2.875808649748995,
      "grad_norm": 0.1627744734287262,
      "learning_rate": 4.1432821173367685e-06,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8235183916985989,
      "num_tokens": 997594825.0,
      "step": 31257
    },
    {
      "entropy": 0.48405909165740013,
      "epoch": 2.8759006561204363,
      "grad_norm": 0.15072619915008545,
      "learning_rate": 4.140215291195142e-06,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8466893769800663,
      "num_tokens": 997626689.0,
      "step": 31258
    },
    {
      "entropy": 0.46814881917089224,
      "epoch": 2.8759926624918775,
      "grad_norm": 0.15089735388755798,
      "learning_rate": 4.137148465053516e-06,
      "loss": 0.4584,
      "mean_token_accuracy": 0.8526739776134491,
      "num_tokens": 997659050.0,
      "step": 31259
    },
    {
      "entropy": 0.545580554753542,
      "epoch": 2.876084668863319,
      "grad_norm": 0.16800086200237274,
      "learning_rate": 4.13408163891189e-06,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8328114710748196,
      "num_tokens": 997688686.0,
      "step": 31260
    },
    {
      "entropy": 0.5433481093496084,
      "epoch": 2.87617667523476,
      "grad_norm": 0.1642759144306183,
      "learning_rate": 4.131014812770264e-06,
      "loss": 0.5345,
      "mean_token_accuracy": 0.8278446905314922,
      "num_tokens": 997720316.0,
      "step": 31261
    },
    {
      "entropy": 0.4900012437719852,
      "epoch": 2.8762686816062013,
      "grad_norm": 0.15345865488052368,
      "learning_rate": 4.1279479866286375e-06,
      "loss": 0.4823,
      "mean_token_accuracy": 0.8492505773901939,
      "num_tokens": 997753037.0,
      "step": 31262
    },
    {
      "entropy": 0.5303038069978356,
      "epoch": 2.8763606879776424,
      "grad_norm": 0.1529984027147293,
      "learning_rate": 4.124881160487012e-06,
      "loss": 0.5129,
      "mean_token_accuracy": 0.836335513740778,
      "num_tokens": 997785549.0,
      "step": 31263
    },
    {
      "entropy": 0.532194547355175,
      "epoch": 2.8764526943490836,
      "grad_norm": 0.1619383543729782,
      "learning_rate": 4.121814334345386e-06,
      "loss": 0.5262,
      "mean_token_accuracy": 0.8371518403291702,
      "num_tokens": 997817568.0,
      "step": 31264
    },
    {
      "entropy": 0.5595115553587675,
      "epoch": 2.876544700720525,
      "grad_norm": 0.16682428121566772,
      "learning_rate": 4.1187475082037606e-06,
      "loss": 0.547,
      "mean_token_accuracy": 0.825828094035387,
      "num_tokens": 997848823.0,
      "step": 31265
    },
    {
      "entropy": 0.5054666826035827,
      "epoch": 2.8766367070919663,
      "grad_norm": 0.1596326231956482,
      "learning_rate": 4.115680682062134e-06,
      "loss": 0.4892,
      "mean_token_accuracy": 0.8446760252118111,
      "num_tokens": 997880056.0,
      "step": 31266
    },
    {
      "entropy": 0.5175486076623201,
      "epoch": 2.8767287134634074,
      "grad_norm": 0.1595136672258377,
      "learning_rate": 4.112613855920508e-06,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8411890715360641,
      "num_tokens": 997911531.0,
      "step": 31267
    },
    {
      "entropy": 0.4841271201148629,
      "epoch": 2.8768207198348485,
      "grad_norm": 0.14960187673568726,
      "learning_rate": 4.109547029778882e-06,
      "loss": 0.4636,
      "mean_token_accuracy": 0.8519152849912643,
      "num_tokens": 997942973.0,
      "step": 31268
    },
    {
      "entropy": 0.622483154758811,
      "epoch": 2.8769127262062897,
      "grad_norm": 0.17019395530223846,
      "learning_rate": 4.106480203637256e-06,
      "loss": 0.614,
      "mean_token_accuracy": 0.8099816590547562,
      "num_tokens": 997975120.0,
      "step": 31269
    },
    {
      "entropy": 0.5041049437131733,
      "epoch": 2.8770047325777313,
      "grad_norm": 0.1555529534816742,
      "learning_rate": 4.10341337749563e-06,
      "loss": 0.49,
      "mean_token_accuracy": 0.8449587300419807,
      "num_tokens": 998007447.0,
      "step": 31270
    },
    {
      "entropy": 0.5493228696286678,
      "epoch": 2.8770967389491724,
      "grad_norm": 0.16898302733898163,
      "learning_rate": 4.100346551354004e-06,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8244623057544231,
      "num_tokens": 998039631.0,
      "step": 31271
    },
    {
      "entropy": 0.44119550473988056,
      "epoch": 2.8771887453206135,
      "grad_norm": 0.15223196148872375,
      "learning_rate": 4.097279725212378e-06,
      "loss": 0.4317,
      "mean_token_accuracy": 0.8603708446025848,
      "num_tokens": 998071152.0,
      "step": 31272
    },
    {
      "entropy": 0.5013704965822399,
      "epoch": 2.8772807516920547,
      "grad_norm": 0.15424032509326935,
      "learning_rate": 4.094212899070752e-06,
      "loss": 0.4875,
      "mean_token_accuracy": 0.8437557145953178,
      "num_tokens": 998103227.0,
      "step": 31273
    },
    {
      "entropy": 0.5156728029251099,
      "epoch": 2.877372758063496,
      "grad_norm": 0.15906314551830292,
      "learning_rate": 4.091146072929126e-06,
      "loss": 0.5059,
      "mean_token_accuracy": 0.8392297364771366,
      "num_tokens": 998134561.0,
      "step": 31274
    },
    {
      "entropy": 0.4793338980525732,
      "epoch": 2.8774647644349374,
      "grad_norm": 0.15442131459712982,
      "learning_rate": 4.088079246787499e-06,
      "loss": 0.4732,
      "mean_token_accuracy": 0.8495701476931572,
      "num_tokens": 998166986.0,
      "step": 31275
    },
    {
      "entropy": 0.4848207300528884,
      "epoch": 2.8775567708063785,
      "grad_norm": 0.15594668686389923,
      "learning_rate": 4.085012420645873e-06,
      "loss": 0.4747,
      "mean_token_accuracy": 0.8464813865721226,
      "num_tokens": 998199307.0,
      "step": 31276
    },
    {
      "entropy": 0.5108547806739807,
      "epoch": 2.8776487771778196,
      "grad_norm": 0.1633753776550293,
      "learning_rate": 4.081945594504247e-06,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8395392522215843,
      "num_tokens": 998230344.0,
      "step": 31277
    },
    {
      "entropy": 0.5361057259142399,
      "epoch": 2.8777407835492608,
      "grad_norm": 0.15970903635025024,
      "learning_rate": 4.078878768362622e-06,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8326511159539223,
      "num_tokens": 998262280.0,
      "step": 31278
    },
    {
      "entropy": 0.49110369582194835,
      "epoch": 2.877832789920702,
      "grad_norm": 0.1559518426656723,
      "learning_rate": 4.0758119422209955e-06,
      "loss": 0.4868,
      "mean_token_accuracy": 0.8460261411964893,
      "num_tokens": 998294620.0,
      "step": 31279
    },
    {
      "entropy": 0.5236918982118368,
      "epoch": 2.8779247962921435,
      "grad_norm": 0.16085156798362732,
      "learning_rate": 4.07274511607937e-06,
      "loss": 0.5029,
      "mean_token_accuracy": 0.8396500535309315,
      "num_tokens": 998326142.0,
      "step": 31280
    },
    {
      "entropy": 0.5397739774780348,
      "epoch": 2.8780168026635846,
      "grad_norm": 0.17064109444618225,
      "learning_rate": 4.069678289937744e-06,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8240397460758686,
      "num_tokens": 998357001.0,
      "step": 31281
    },
    {
      "entropy": 0.4082274080719799,
      "epoch": 2.8781088090350258,
      "grad_norm": 0.14522093534469604,
      "learning_rate": 4.066611463796118e-06,
      "loss": 0.3878,
      "mean_token_accuracy": 0.8740737922489643,
      "num_tokens": 998388616.0,
      "step": 31282
    },
    {
      "entropy": 0.4451181956101209,
      "epoch": 2.878200815406467,
      "grad_norm": 0.14908266067504883,
      "learning_rate": 4.0635446376544915e-06,
      "loss": 0.4262,
      "mean_token_accuracy": 0.8633779212832451,
      "num_tokens": 998419858.0,
      "step": 31283
    },
    {
      "entropy": 0.5758809766266495,
      "epoch": 2.878292821777908,
      "grad_norm": 0.16633547842502594,
      "learning_rate": 4.060477811512865e-06,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8231260403990746,
      "num_tokens": 998452465.0,
      "step": 31284
    },
    {
      "entropy": 0.5311901327222586,
      "epoch": 2.8783848281493496,
      "grad_norm": 0.16250966489315033,
      "learning_rate": 4.05741098537124e-06,
      "loss": 0.5135,
      "mean_token_accuracy": 0.8371080793440342,
      "num_tokens": 998484450.0,
      "step": 31285
    },
    {
      "entropy": 0.4894248126074672,
      "epoch": 2.8784768345207907,
      "grad_norm": 0.15893715620040894,
      "learning_rate": 4.054344159229614e-06,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8495708480477333,
      "num_tokens": 998515953.0,
      "step": 31286
    },
    {
      "entropy": 0.508858148008585,
      "epoch": 2.878568840892232,
      "grad_norm": 0.17616601288318634,
      "learning_rate": 4.0512773330879876e-06,
      "loss": 0.4837,
      "mean_token_accuracy": 0.8445065282285213,
      "num_tokens": 998547715.0,
      "step": 31287
    },
    {
      "entropy": 0.5803709756582975,
      "epoch": 2.878660847263673,
      "grad_norm": 0.16744501888751984,
      "learning_rate": 4.048210506946361e-06,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8238446712493896,
      "num_tokens": 998579640.0,
      "step": 31288
    },
    {
      "entropy": 0.5555455880239606,
      "epoch": 2.878752853635114,
      "grad_norm": 0.16587890684604645,
      "learning_rate": 4.045143680804735e-06,
      "loss": 0.5423,
      "mean_token_accuracy": 0.8310403451323509,
      "num_tokens": 998611098.0,
      "step": 31289
    },
    {
      "entropy": 0.4406985864043236,
      "epoch": 2.8788448600065557,
      "grad_norm": 0.15021689236164093,
      "learning_rate": 4.042076854663109e-06,
      "loss": 0.437,
      "mean_token_accuracy": 0.8579268045723438,
      "num_tokens": 998642633.0,
      "step": 31290
    },
    {
      "entropy": 0.5502172168344259,
      "epoch": 2.878936866377997,
      "grad_norm": 0.16164661943912506,
      "learning_rate": 4.039010028521483e-06,
      "loss": 0.5375,
      "mean_token_accuracy": 0.8305261731147766,
      "num_tokens": 998674457.0,
      "step": 31291
    },
    {
      "entropy": 0.5258241882547736,
      "epoch": 2.879028872749438,
      "grad_norm": 0.15715518593788147,
      "learning_rate": 4.0359432023798566e-06,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8353083357214928,
      "num_tokens": 998706919.0,
      "step": 31292
    },
    {
      "entropy": 0.5435229651629925,
      "epoch": 2.879120879120879,
      "grad_norm": 0.1662316918373108,
      "learning_rate": 4.032876376238231e-06,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8336584493517876,
      "num_tokens": 998738590.0,
      "step": 31293
    },
    {
      "entropy": 0.4696411266922951,
      "epoch": 2.8792128854923202,
      "grad_norm": 0.1491681933403015,
      "learning_rate": 4.029809550096605e-06,
      "loss": 0.4606,
      "mean_token_accuracy": 0.8496014550328255,
      "num_tokens": 998770591.0,
      "step": 31294
    },
    {
      "entropy": 0.604468896985054,
      "epoch": 2.879304891863762,
      "grad_norm": 0.16786432266235352,
      "learning_rate": 4.02674272395498e-06,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8188901618123055,
      "num_tokens": 998802652.0,
      "step": 31295
    },
    {
      "entropy": 0.5545415785163641,
      "epoch": 2.879396898235203,
      "grad_norm": 0.16648411750793457,
      "learning_rate": 4.0236758978133535e-06,
      "loss": 0.5522,
      "mean_token_accuracy": 0.825324285775423,
      "num_tokens": 998834014.0,
      "step": 31296
    },
    {
      "entropy": 0.6097477450966835,
      "epoch": 2.879488904606644,
      "grad_norm": 0.17000553011894226,
      "learning_rate": 4.020609071671727e-06,
      "loss": 0.5937,
      "mean_token_accuracy": 0.8139801695942879,
      "num_tokens": 998866099.0,
      "step": 31297
    },
    {
      "entropy": 0.5811591614037752,
      "epoch": 2.879580910978085,
      "grad_norm": 0.1662900447845459,
      "learning_rate": 4.017542245530101e-06,
      "loss": 0.5609,
      "mean_token_accuracy": 0.8229537457227707,
      "num_tokens": 998897243.0,
      "step": 31298
    },
    {
      "entropy": 0.5647624358534813,
      "epoch": 2.8796729173495264,
      "grad_norm": 0.16905628144741058,
      "learning_rate": 4.014475419388475e-06,
      "loss": 0.5419,
      "mean_token_accuracy": 0.826876986771822,
      "num_tokens": 998928649.0,
      "step": 31299
    },
    {
      "entropy": 0.4643087845761329,
      "epoch": 2.879764923720968,
      "grad_norm": 0.15087509155273438,
      "learning_rate": 4.011408593246849e-06,
      "loss": 0.4544,
      "mean_token_accuracy": 0.8560409694910049,
      "num_tokens": 998960390.0,
      "step": 31300
    },
    {
      "entropy": 0.44931311602704227,
      "epoch": 2.879856930092409,
      "grad_norm": 0.15146410465240479,
      "learning_rate": 4.008341767105223e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8609739057719707,
      "num_tokens": 998992865.0,
      "step": 31301
    },
    {
      "entropy": 0.5031686236616224,
      "epoch": 2.87994893646385,
      "grad_norm": 0.15608350932598114,
      "learning_rate": 4.005274940963597e-06,
      "loss": 0.495,
      "mean_token_accuracy": 0.8457356914877892,
      "num_tokens": 999025031.0,
      "step": 31302
    },
    {
      "entropy": 0.5622586570680141,
      "epoch": 2.8800409428352913,
      "grad_norm": 0.17077511548995972,
      "learning_rate": 4.002208114821971e-06,
      "loss": 0.5659,
      "mean_token_accuracy": 0.8260948359966278,
      "num_tokens": 999057391.0,
      "step": 31303
    },
    {
      "entropy": 0.593403865583241,
      "epoch": 2.8801329492067325,
      "grad_norm": 0.17045412957668304,
      "learning_rate": 3.999141288680345e-06,
      "loss": 0.5871,
      "mean_token_accuracy": 0.8152286410331726,
      "num_tokens": 999089019.0,
      "step": 31304
    },
    {
      "entropy": 0.5495115965604782,
      "epoch": 2.880224955578174,
      "grad_norm": 0.16491472721099854,
      "learning_rate": 3.9960744625387185e-06,
      "loss": 0.5249,
      "mean_token_accuracy": 0.8317015320062637,
      "num_tokens": 999119695.0,
      "step": 31305
    },
    {
      "entropy": 0.5110496655106544,
      "epoch": 2.880316961949615,
      "grad_norm": 0.1539221704006195,
      "learning_rate": 3.993007636397092e-06,
      "loss": 0.5007,
      "mean_token_accuracy": 0.84166194871068,
      "num_tokens": 999152330.0,
      "step": 31306
    },
    {
      "entropy": 0.5230075367726386,
      "epoch": 2.8804089683210563,
      "grad_norm": 0.15523354709148407,
      "learning_rate": 3.989940810255466e-06,
      "loss": 0.5001,
      "mean_token_accuracy": 0.8382917307317257,
      "num_tokens": 999183847.0,
      "step": 31307
    },
    {
      "entropy": 0.4677544068545103,
      "epoch": 2.8805009746924974,
      "grad_norm": 0.1548316478729248,
      "learning_rate": 3.986873984113841e-06,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8529916852712631,
      "num_tokens": 999215277.0,
      "step": 31308
    },
    {
      "entropy": 0.4481594869866967,
      "epoch": 2.8805929810639386,
      "grad_norm": 0.14918948709964752,
      "learning_rate": 3.9838071579722146e-06,
      "loss": 0.4373,
      "mean_token_accuracy": 0.8586076013743877,
      "num_tokens": 999247419.0,
      "step": 31309
    },
    {
      "entropy": 0.5124827465042472,
      "epoch": 2.88068498743538,
      "grad_norm": 0.1658295840024948,
      "learning_rate": 3.980740331830589e-06,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8435980528593063,
      "num_tokens": 999278156.0,
      "step": 31310
    },
    {
      "entropy": 0.4436284340918064,
      "epoch": 2.8807769938068213,
      "grad_norm": 0.14569304883480072,
      "learning_rate": 3.977673505688963e-06,
      "loss": 0.4335,
      "mean_token_accuracy": 0.8613918088376522,
      "num_tokens": 999310415.0,
      "step": 31311
    },
    {
      "entropy": 0.45013231597840786,
      "epoch": 2.8808690001782624,
      "grad_norm": 0.14795148372650146,
      "learning_rate": 3.974606679547337e-06,
      "loss": 0.4257,
      "mean_token_accuracy": 0.8614009618759155,
      "num_tokens": 999342456.0,
      "step": 31312
    },
    {
      "entropy": 0.4766194950789213,
      "epoch": 2.8809610065497036,
      "grad_norm": 0.15573373436927795,
      "learning_rate": 3.971539853405711e-06,
      "loss": 0.4539,
      "mean_token_accuracy": 0.8555333130061626,
      "num_tokens": 999373784.0,
      "step": 31313
    },
    {
      "entropy": 0.4738125381991267,
      "epoch": 2.8810530129211447,
      "grad_norm": 0.15539449453353882,
      "learning_rate": 3.968473027264084e-06,
      "loss": 0.4606,
      "mean_token_accuracy": 0.8538845479488373,
      "num_tokens": 999405798.0,
      "step": 31314
    },
    {
      "entropy": 0.4475711612030864,
      "epoch": 2.8811450192925863,
      "grad_norm": 0.14801909029483795,
      "learning_rate": 3.965406201122458e-06,
      "loss": 0.4304,
      "mean_token_accuracy": 0.8617581874132156,
      "num_tokens": 999437991.0,
      "step": 31315
    },
    {
      "entropy": 0.4664712288649753,
      "epoch": 2.8812370256640274,
      "grad_norm": 0.1557716578245163,
      "learning_rate": 3.962339374980833e-06,
      "loss": 0.4567,
      "mean_token_accuracy": 0.8547178879380226,
      "num_tokens": 999469598.0,
      "step": 31316
    },
    {
      "entropy": 0.544954190030694,
      "epoch": 2.8813290320354685,
      "grad_norm": 0.16157203912734985,
      "learning_rate": 3.959272548839207e-06,
      "loss": 0.5388,
      "mean_token_accuracy": 0.8309501595795155,
      "num_tokens": 999501360.0,
      "step": 31317
    },
    {
      "entropy": 0.5481997635215521,
      "epoch": 2.8814210384069097,
      "grad_norm": 0.16234320402145386,
      "learning_rate": 3.9562057226975805e-06,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8268891796469688,
      "num_tokens": 999533312.0,
      "step": 31318
    },
    {
      "entropy": 0.5121154161170125,
      "epoch": 2.881513044778351,
      "grad_norm": 0.15037237107753754,
      "learning_rate": 3.953138896555954e-06,
      "loss": 0.4977,
      "mean_token_accuracy": 0.8433976769447327,
      "num_tokens": 999565804.0,
      "step": 31319
    },
    {
      "entropy": 0.5800366662442684,
      "epoch": 2.8816050511497924,
      "grad_norm": 0.1651807725429535,
      "learning_rate": 3.950072070414328e-06,
      "loss": 0.5578,
      "mean_token_accuracy": 0.8218883015215397,
      "num_tokens": 999597968.0,
      "step": 31320
    },
    {
      "entropy": 0.49190973676741123,
      "epoch": 2.8816970575212335,
      "grad_norm": 0.15593954920768738,
      "learning_rate": 3.947005244272702e-06,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8460346348583698,
      "num_tokens": 999630528.0,
      "step": 31321
    },
    {
      "entropy": 0.5653857411816716,
      "epoch": 2.8817890638926746,
      "grad_norm": 0.16716259717941284,
      "learning_rate": 3.943938418131076e-06,
      "loss": 0.5452,
      "mean_token_accuracy": 0.8266244679689407,
      "num_tokens": 999662048.0,
      "step": 31322
    },
    {
      "entropy": 0.5570958770113066,
      "epoch": 2.8818810702641158,
      "grad_norm": 0.16258499026298523,
      "learning_rate": 3.94087159198945e-06,
      "loss": 0.5456,
      "mean_token_accuracy": 0.8289714269340038,
      "num_tokens": 999693789.0,
      "step": 31323
    },
    {
      "entropy": 0.53520926553756,
      "epoch": 2.881973076635557,
      "grad_norm": 0.1612854152917862,
      "learning_rate": 3.937804765847824e-06,
      "loss": 0.5331,
      "mean_token_accuracy": 0.8303773477673531,
      "num_tokens": 999725874.0,
      "step": 31324
    },
    {
      "entropy": 0.41958001954481006,
      "epoch": 2.8820650830069985,
      "grad_norm": 0.15103179216384888,
      "learning_rate": 3.934737939706199e-06,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8642576038837433,
      "num_tokens": 999757720.0,
      "step": 31325
    },
    {
      "entropy": 0.3282533483579755,
      "epoch": 2.8821570893784396,
      "grad_norm": 0.12896113097667694,
      "learning_rate": 3.9316711135645726e-06,
      "loss": 0.3171,
      "mean_token_accuracy": 0.8975245393812656,
      "num_tokens": 999790075.0,
      "step": 31326
    },
    {
      "entropy": 0.4618805437348783,
      "epoch": 2.8822490957498808,
      "grad_norm": 0.1543899029493332,
      "learning_rate": 3.928604287422946e-06,
      "loss": 0.4426,
      "mean_token_accuracy": 0.8573475480079651,
      "num_tokens": 999821761.0,
      "step": 31327
    },
    {
      "entropy": 0.48334518168121576,
      "epoch": 2.882341102121322,
      "grad_norm": 0.15031851828098297,
      "learning_rate": 3.92553746128132e-06,
      "loss": 0.458,
      "mean_token_accuracy": 0.8510084263980389,
      "num_tokens": 999853816.0,
      "step": 31328
    },
    {
      "entropy": 0.5076165767386556,
      "epoch": 2.882433108492763,
      "grad_norm": 0.15211182832717896,
      "learning_rate": 3.922470635139694e-06,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8421941362321377,
      "num_tokens": 999885981.0,
      "step": 31329
    },
    {
      "entropy": 0.46056328620761633,
      "epoch": 2.8825251148642046,
      "grad_norm": 0.15504871308803558,
      "learning_rate": 3.919403808998068e-06,
      "loss": 0.4537,
      "mean_token_accuracy": 0.8555531539022923,
      "num_tokens": 999918291.0,
      "step": 31330
    },
    {
      "entropy": 0.5924987886101007,
      "epoch": 2.8826171212356457,
      "grad_norm": 0.16703715920448303,
      "learning_rate": 3.916336982856442e-06,
      "loss": 0.5718,
      "mean_token_accuracy": 0.819462563842535,
      "num_tokens": 999949743.0,
      "step": 31331
    },
    {
      "entropy": 0.5743249375373125,
      "epoch": 2.882709127607087,
      "grad_norm": 0.1661352962255478,
      "learning_rate": 3.913270156714816e-06,
      "loss": 0.5641,
      "mean_token_accuracy": 0.8220742046833038,
      "num_tokens": 999981227.0,
      "step": 31332
    },
    {
      "entropy": 0.5035208500921726,
      "epoch": 2.882801133978528,
      "grad_norm": 0.1529919058084488,
      "learning_rate": 3.91020333057319e-06,
      "loss": 0.4884,
      "mean_token_accuracy": 0.8453472293913364,
      "num_tokens": 1000013251.0,
      "step": 31333
    },
    {
      "entropy": 0.42084903991781175,
      "epoch": 2.882893140349969,
      "grad_norm": 0.1475006639957428,
      "learning_rate": 3.907136504431564e-06,
      "loss": 0.4042,
      "mean_token_accuracy": 0.8713021986186504,
      "num_tokens": 1000045441.0,
      "step": 31334
    },
    {
      "entropy": 0.5573714142665267,
      "epoch": 2.8829851467214107,
      "grad_norm": 0.16773557662963867,
      "learning_rate": 3.904069678289938e-06,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8296543546020985,
      "num_tokens": 1000076379.0,
      "step": 31335
    },
    {
      "entropy": 0.4662314420565963,
      "epoch": 2.883077153092852,
      "grad_norm": 0.15170705318450928,
      "learning_rate": 3.901002852148311e-06,
      "loss": 0.4515,
      "mean_token_accuracy": 0.8557563535869122,
      "num_tokens": 1000107872.0,
      "step": 31336
    },
    {
      "entropy": 0.4783631367608905,
      "epoch": 2.883169159464293,
      "grad_norm": 0.14803795516490936,
      "learning_rate": 3.897936026006685e-06,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8552236594259739,
      "num_tokens": 1000139544.0,
      "step": 31337
    },
    {
      "entropy": 0.49596782051958144,
      "epoch": 2.883261165835734,
      "grad_norm": 0.1585795283317566,
      "learning_rate": 3.89486919986506e-06,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8471024073660374,
      "num_tokens": 1000171168.0,
      "step": 31338
    },
    {
      "entropy": 0.48750976473093033,
      "epoch": 2.8833531722071752,
      "grad_norm": 0.16297228634357452,
      "learning_rate": 3.8918023737234345e-06,
      "loss": 0.4898,
      "mean_token_accuracy": 0.8465131968259811,
      "num_tokens": 1000203049.0,
      "step": 31339
    },
    {
      "entropy": 0.4553116918541491,
      "epoch": 2.883445178578617,
      "grad_norm": 0.1473173350095749,
      "learning_rate": 3.888735547581808e-06,
      "loss": 0.4457,
      "mean_token_accuracy": 0.8573811165988445,
      "num_tokens": 1000235465.0,
      "step": 31340
    },
    {
      "entropy": 0.5577451396966353,
      "epoch": 2.883537184950058,
      "grad_norm": 0.16841186583042145,
      "learning_rate": 3.885668721440182e-06,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8279734887182713,
      "num_tokens": 1000266833.0,
      "step": 31341
    },
    {
      "entropy": 0.4646059329388663,
      "epoch": 2.883629191321499,
      "grad_norm": 0.1496892273426056,
      "learning_rate": 3.882601895298556e-06,
      "loss": 0.4539,
      "mean_token_accuracy": 0.8533219881355762,
      "num_tokens": 1000299011.0,
      "step": 31342
    },
    {
      "entropy": 0.4529569330625236,
      "epoch": 2.88372119769294,
      "grad_norm": 0.15390510857105255,
      "learning_rate": 3.87953506915693e-06,
      "loss": 0.445,
      "mean_token_accuracy": 0.8580760769546032,
      "num_tokens": 1000331297.0,
      "step": 31343
    },
    {
      "entropy": 0.5762043062131852,
      "epoch": 2.8838132040643814,
      "grad_norm": 0.1669618934392929,
      "learning_rate": 3.8764682430153035e-06,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8216305561363697,
      "num_tokens": 1000363167.0,
      "step": 31344
    },
    {
      "entropy": 0.5664607379585505,
      "epoch": 2.883905210435823,
      "grad_norm": 0.16429747641086578,
      "learning_rate": 3.873401416873677e-06,
      "loss": 0.5595,
      "mean_token_accuracy": 0.823222778737545,
      "num_tokens": 1000394953.0,
      "step": 31345
    },
    {
      "entropy": 0.4658992630429566,
      "epoch": 2.883997216807264,
      "grad_norm": 0.15182535350322723,
      "learning_rate": 3.870334590732052e-06,
      "loss": 0.4546,
      "mean_token_accuracy": 0.8552075885236263,
      "num_tokens": 1000426819.0,
      "step": 31346
    },
    {
      "entropy": 0.5108497987966985,
      "epoch": 2.884089223178705,
      "grad_norm": 0.1574421375989914,
      "learning_rate": 3.867267764590426e-06,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8432298712432384,
      "num_tokens": 1000458269.0,
      "step": 31347
    },
    {
      "entropy": 0.5094836074858904,
      "epoch": 2.8841812295501463,
      "grad_norm": 0.16614863276481628,
      "learning_rate": 3.8642009384487996e-06,
      "loss": 0.5016,
      "mean_token_accuracy": 0.8414914645254612,
      "num_tokens": 1000489863.0,
      "step": 31348
    },
    {
      "entropy": 0.5228829374536872,
      "epoch": 2.8842732359215875,
      "grad_norm": 0.15995804965496063,
      "learning_rate": 3.861134112307173e-06,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8373846486210823,
      "num_tokens": 1000521197.0,
      "step": 31349
    },
    {
      "entropy": 0.5211272407323122,
      "epoch": 2.884365242293029,
      "grad_norm": 0.1603499948978424,
      "learning_rate": 3.858067286165547e-06,
      "loss": 0.5164,
      "mean_token_accuracy": 0.8354693613946438,
      "num_tokens": 1000553773.0,
      "step": 31350
    },
    {
      "entropy": 0.5542767755687237,
      "epoch": 2.88445724866447,
      "grad_norm": 0.1696052998304367,
      "learning_rate": 3.855000460023921e-06,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8246998600661755,
      "num_tokens": 1000584721.0,
      "step": 31351
    },
    {
      "entropy": 0.512807235121727,
      "epoch": 2.8845492550359113,
      "grad_norm": 0.15926523506641388,
      "learning_rate": 3.851933633882296e-06,
      "loss": 0.504,
      "mean_token_accuracy": 0.8391728177666664,
      "num_tokens": 1000616797.0,
      "step": 31352
    },
    {
      "entropy": 0.5093614235520363,
      "epoch": 2.8846412614073524,
      "grad_norm": 0.16057150065898895,
      "learning_rate": 3.848866807740669e-06,
      "loss": 0.4857,
      "mean_token_accuracy": 0.8470631390810013,
      "num_tokens": 1000648424.0,
      "step": 31353
    },
    {
      "entropy": 0.464734866283834,
      "epoch": 2.8847332677787936,
      "grad_norm": 0.15297754108905792,
      "learning_rate": 3.845799981599044e-06,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8583525456488132,
      "num_tokens": 1000680399.0,
      "step": 31354
    },
    {
      "entropy": 0.46068653278052807,
      "epoch": 2.884825274150235,
      "grad_norm": 0.15263864398002625,
      "learning_rate": 3.842733155457418e-06,
      "loss": 0.4446,
      "mean_token_accuracy": 0.8562167808413506,
      "num_tokens": 1000712477.0,
      "step": 31355
    },
    {
      "entropy": 0.4661556165665388,
      "epoch": 2.8849172805216763,
      "grad_norm": 0.1605611890554428,
      "learning_rate": 3.839666329315792e-06,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8492453619837761,
      "num_tokens": 1000743830.0,
      "step": 31356
    },
    {
      "entropy": 0.5090312091633677,
      "epoch": 2.8850092868931174,
      "grad_norm": 0.15661554038524628,
      "learning_rate": 3.8365995031741655e-06,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8369695730507374,
      "num_tokens": 1000776216.0,
      "step": 31357
    },
    {
      "entropy": 0.3857209477573633,
      "epoch": 2.8851012932645586,
      "grad_norm": 0.13816706836223602,
      "learning_rate": 3.833532677032539e-06,
      "loss": 0.3632,
      "mean_token_accuracy": 0.8802574388682842,
      "num_tokens": 1000807934.0,
      "step": 31358
    },
    {
      "entropy": 0.5808370588347316,
      "epoch": 2.8851932996359997,
      "grad_norm": 0.16279934346675873,
      "learning_rate": 3.830465850890913e-06,
      "loss": 0.5701,
      "mean_token_accuracy": 0.8214392922818661,
      "num_tokens": 1000840016.0,
      "step": 31359
    },
    {
      "entropy": 0.5106098770629615,
      "epoch": 2.8852853060074413,
      "grad_norm": 0.16140003502368927,
      "learning_rate": 3.827399024749287e-06,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8398513346910477,
      "num_tokens": 1000871696.0,
      "step": 31360
    },
    {
      "entropy": 0.46563585940748453,
      "epoch": 2.8853773123788824,
      "grad_norm": 0.1532893031835556,
      "learning_rate": 3.824332198607661e-06,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8528696484863758,
      "num_tokens": 1000904363.0,
      "step": 31361
    },
    {
      "entropy": 0.4816677989438176,
      "epoch": 2.8854693187503235,
      "grad_norm": 0.15487657487392426,
      "learning_rate": 3.821265372466035e-06,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8489088229835033,
      "num_tokens": 1000936351.0,
      "step": 31362
    },
    {
      "entropy": 0.5624307952821255,
      "epoch": 2.8855613251217647,
      "grad_norm": 0.1681782603263855,
      "learning_rate": 3.818198546324409e-06,
      "loss": 0.5512,
      "mean_token_accuracy": 0.8256294839084148,
      "num_tokens": 1000967619.0,
      "step": 31363
    },
    {
      "entropy": 0.4686747780069709,
      "epoch": 2.885653331493206,
      "grad_norm": 0.1544167846441269,
      "learning_rate": 3.815131720182783e-06,
      "loss": 0.461,
      "mean_token_accuracy": 0.850154560059309,
      "num_tokens": 1000999308.0,
      "step": 31364
    },
    {
      "entropy": 0.5132130272686481,
      "epoch": 2.8857453378646474,
      "grad_norm": 0.1666790246963501,
      "learning_rate": 3.812064894041157e-06,
      "loss": 0.5092,
      "mean_token_accuracy": 0.836430586874485,
      "num_tokens": 1001030560.0,
      "step": 31365
    },
    {
      "entropy": 0.49101167684420943,
      "epoch": 2.8858373442360885,
      "grad_norm": 0.15393921732902527,
      "learning_rate": 3.808998067899531e-06,
      "loss": 0.4764,
      "mean_token_accuracy": 0.8484195284545422,
      "num_tokens": 1001062629.0,
      "step": 31366
    },
    {
      "entropy": 0.5380766540765762,
      "epoch": 2.8859293506075296,
      "grad_norm": 0.15993227064609528,
      "learning_rate": 3.8059312417579047e-06,
      "loss": 0.5223,
      "mean_token_accuracy": 0.8329933993518353,
      "num_tokens": 1001094710.0,
      "step": 31367
    },
    {
      "entropy": 0.48656788654625416,
      "epoch": 2.8860213569789708,
      "grad_norm": 0.1501987725496292,
      "learning_rate": 3.8028644156162785e-06,
      "loss": 0.4717,
      "mean_token_accuracy": 0.8506922647356987,
      "num_tokens": 1001127111.0,
      "step": 31368
    },
    {
      "entropy": 0.6098427753895521,
      "epoch": 2.886113363350412,
      "grad_norm": 0.16997799277305603,
      "learning_rate": 3.799797589474653e-06,
      "loss": 0.5871,
      "mean_token_accuracy": 0.8132799379527569,
      "num_tokens": 1001159024.0,
      "step": 31369
    },
    {
      "entropy": 0.5503489375114441,
      "epoch": 2.8862053697218535,
      "grad_norm": 0.16186712682247162,
      "learning_rate": 3.796730763333027e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8261216878890991,
      "num_tokens": 1001190619.0,
      "step": 31370
    },
    {
      "entropy": 0.503169001895003,
      "epoch": 2.8862973760932946,
      "grad_norm": 0.15461675822734833,
      "learning_rate": 3.793663937191401e-06,
      "loss": 0.4869,
      "mean_token_accuracy": 0.8446392677724361,
      "num_tokens": 1001221526.0,
      "step": 31371
    },
    {
      "entropy": 0.5247924281284213,
      "epoch": 2.8863893824647358,
      "grad_norm": 0.16046680510044098,
      "learning_rate": 3.790597111049775e-06,
      "loss": 0.5149,
      "mean_token_accuracy": 0.8352019116282463,
      "num_tokens": 1001253686.0,
      "step": 31372
    },
    {
      "entropy": 0.5046870270743966,
      "epoch": 2.886481388836177,
      "grad_norm": 0.15839765965938568,
      "learning_rate": 3.787530284908149e-06,
      "loss": 0.4956,
      "mean_token_accuracy": 0.843292623758316,
      "num_tokens": 1001285598.0,
      "step": 31373
    },
    {
      "entropy": 0.5196921182796359,
      "epoch": 2.886573395207618,
      "grad_norm": 0.15511275827884674,
      "learning_rate": 3.7844634587665226e-06,
      "loss": 0.5105,
      "mean_token_accuracy": 0.8378648199141026,
      "num_tokens": 1001317792.0,
      "step": 31374
    },
    {
      "entropy": 0.5631193569861352,
      "epoch": 2.8866654015790596,
      "grad_norm": 0.1625712811946869,
      "learning_rate": 3.7813966326248964e-06,
      "loss": 0.5553,
      "mean_token_accuracy": 0.8249815180897713,
      "num_tokens": 1001349615.0,
      "step": 31375
    },
    {
      "entropy": 0.5516986083239317,
      "epoch": 2.8867574079505007,
      "grad_norm": 0.15986981987953186,
      "learning_rate": 3.7783298064832702e-06,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8297471404075623,
      "num_tokens": 1001382328.0,
      "step": 31376
    },
    {
      "entropy": 0.511344657978043,
      "epoch": 2.886849414321942,
      "grad_norm": 0.1553650200366974,
      "learning_rate": 3.775262980341645e-06,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8409059606492519,
      "num_tokens": 1001414459.0,
      "step": 31377
    },
    {
      "entropy": 0.4381791641935706,
      "epoch": 2.886941420693383,
      "grad_norm": 0.15026284754276276,
      "learning_rate": 3.7721961542000187e-06,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8625035807490349,
      "num_tokens": 1001446067.0,
      "step": 31378
    },
    {
      "entropy": 0.5384059250354767,
      "epoch": 2.887033427064824,
      "grad_norm": 0.1585063338279724,
      "learning_rate": 3.7691293280583925e-06,
      "loss": 0.5275,
      "mean_token_accuracy": 0.836004976183176,
      "num_tokens": 1001477794.0,
      "step": 31379
    },
    {
      "entropy": 0.5951441891957074,
      "epoch": 2.8871254334362657,
      "grad_norm": 0.16913282871246338,
      "learning_rate": 3.7660625019167667e-06,
      "loss": 0.5829,
      "mean_token_accuracy": 0.8182841800153255,
      "num_tokens": 1001509883.0,
      "step": 31380
    },
    {
      "entropy": 0.4776777299121022,
      "epoch": 2.887217439807707,
      "grad_norm": 0.1571924239397049,
      "learning_rate": 3.7629956757751405e-06,
      "loss": 0.467,
      "mean_token_accuracy": 0.8536336459219456,
      "num_tokens": 1001541305.0,
      "step": 31381
    },
    {
      "entropy": 0.5162696205079556,
      "epoch": 2.887309446179148,
      "grad_norm": 0.16026538610458374,
      "learning_rate": 3.7599288496335143e-06,
      "loss": 0.5103,
      "mean_token_accuracy": 0.8378934748470783,
      "num_tokens": 1001573178.0,
      "step": 31382
    },
    {
      "entropy": 0.5158943072892725,
      "epoch": 2.887401452550589,
      "grad_norm": 0.1601237654685974,
      "learning_rate": 3.756862023491888e-06,
      "loss": 0.5039,
      "mean_token_accuracy": 0.8383112996816635,
      "num_tokens": 1001605035.0,
      "step": 31383
    },
    {
      "entropy": 0.4422268131747842,
      "epoch": 2.8874934589220302,
      "grad_norm": 0.15306456387043,
      "learning_rate": 3.753795197350262e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8643792159855366,
      "num_tokens": 1001637163.0,
      "step": 31384
    },
    {
      "entropy": 0.43683998100459576,
      "epoch": 2.887585465293472,
      "grad_norm": 0.14901979267597198,
      "learning_rate": 3.7507283712086365e-06,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8611868321895599,
      "num_tokens": 1001669111.0,
      "step": 31385
    },
    {
      "entropy": 0.46248890832066536,
      "epoch": 2.887677471664913,
      "grad_norm": 0.147746741771698,
      "learning_rate": 3.7476615450670103e-06,
      "loss": 0.4537,
      "mean_token_accuracy": 0.856044489890337,
      "num_tokens": 1001701493.0,
      "step": 31386
    },
    {
      "entropy": 0.5017292750999331,
      "epoch": 2.887769478036354,
      "grad_norm": 0.15307417511940002,
      "learning_rate": 3.7445947189253846e-06,
      "loss": 0.4905,
      "mean_token_accuracy": 0.8460048250854015,
      "num_tokens": 1001734050.0,
      "step": 31387
    },
    {
      "entropy": 0.4977693387772888,
      "epoch": 2.887861484407795,
      "grad_norm": 0.1597558557987213,
      "learning_rate": 3.7415278927837584e-06,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8431467711925507,
      "num_tokens": 1001765569.0,
      "step": 31388
    },
    {
      "entropy": 0.5096885403618217,
      "epoch": 2.8879534907792364,
      "grad_norm": 0.16646552085876465,
      "learning_rate": 3.738461066642132e-06,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8379825763404369,
      "num_tokens": 1001797104.0,
      "step": 31389
    },
    {
      "entropy": 0.3985362323001027,
      "epoch": 2.888045497150678,
      "grad_norm": 0.14728254079818726,
      "learning_rate": 3.735394240500506e-06,
      "loss": 0.393,
      "mean_token_accuracy": 0.874590776860714,
      "num_tokens": 1001829369.0,
      "step": 31390
    },
    {
      "entropy": 0.5517448550090194,
      "epoch": 2.888137503522119,
      "grad_norm": 0.1642114371061325,
      "learning_rate": 3.7323274143588798e-06,
      "loss": 0.5446,
      "mean_token_accuracy": 0.8282720297574997,
      "num_tokens": 1001860855.0,
      "step": 31391
    },
    {
      "entropy": 0.48244852921925485,
      "epoch": 2.88822950989356,
      "grad_norm": 0.14956510066986084,
      "learning_rate": 3.7292605882172544e-06,
      "loss": 0.4717,
      "mean_token_accuracy": 0.8516433872282505,
      "num_tokens": 1001893533.0,
      "step": 31392
    },
    {
      "entropy": 0.505201761610806,
      "epoch": 2.8883215162650013,
      "grad_norm": 0.15419992804527283,
      "learning_rate": 3.7261937620756282e-06,
      "loss": 0.4983,
      "mean_token_accuracy": 0.841070294380188,
      "num_tokens": 1001926263.0,
      "step": 31393
    },
    {
      "entropy": 0.5100685218349099,
      "epoch": 2.8884135226364425,
      "grad_norm": 0.15466824173927307,
      "learning_rate": 3.723126935934002e-06,
      "loss": 0.4948,
      "mean_token_accuracy": 0.8405955471098423,
      "num_tokens": 1001958271.0,
      "step": 31394
    },
    {
      "entropy": 0.5117922476492822,
      "epoch": 2.888505529007884,
      "grad_norm": 0.16041138768196106,
      "learning_rate": 3.7200601097923762e-06,
      "loss": 0.4939,
      "mean_token_accuracy": 0.8398463949561119,
      "num_tokens": 1001989788.0,
      "step": 31395
    },
    {
      "entropy": 0.4336009216494858,
      "epoch": 2.888597535379325,
      "grad_norm": 0.14785951375961304,
      "learning_rate": 3.71699328365075e-06,
      "loss": 0.4182,
      "mean_token_accuracy": 0.8662091419100761,
      "num_tokens": 1002021715.0,
      "step": 31396
    },
    {
      "entropy": 0.5075069894082844,
      "epoch": 2.8886895417507663,
      "grad_norm": 0.15695008635520935,
      "learning_rate": 3.713926457509124e-06,
      "loss": 0.4944,
      "mean_token_accuracy": 0.8432704955339432,
      "num_tokens": 1002053135.0,
      "step": 31397
    },
    {
      "entropy": 0.48345300182700157,
      "epoch": 2.8887815481222074,
      "grad_norm": 0.1501331478357315,
      "learning_rate": 3.7108596313674976e-06,
      "loss": 0.4674,
      "mean_token_accuracy": 0.8495795242488384,
      "num_tokens": 1002085563.0,
      "step": 31398
    },
    {
      "entropy": 0.48772477079182863,
      "epoch": 2.8888735544936486,
      "grad_norm": 0.1581398844718933,
      "learning_rate": 3.7077928052258714e-06,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8479868769645691,
      "num_tokens": 1002117662.0,
      "step": 31399
    },
    {
      "entropy": 0.5710143679752946,
      "epoch": 2.88896556086509,
      "grad_norm": 0.16097256541252136,
      "learning_rate": 3.704725979084246e-06,
      "loss": 0.562,
      "mean_token_accuracy": 0.8219433277845383,
      "num_tokens": 1002150119.0,
      "step": 31400
    },
    {
      "entropy": 0.522172839846462,
      "epoch": 2.8890575672365313,
      "grad_norm": 0.16085419058799744,
      "learning_rate": 3.70165915294262e-06,
      "loss": 0.5094,
      "mean_token_accuracy": 0.8389863967895508,
      "num_tokens": 1002182056.0,
      "step": 31401
    },
    {
      "entropy": 0.5290586017072201,
      "epoch": 2.8891495736079724,
      "grad_norm": 0.16602900624275208,
      "learning_rate": 3.698592326800994e-06,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8388316221535206,
      "num_tokens": 1002213166.0,
      "step": 31402
    },
    {
      "entropy": 0.48118123901076615,
      "epoch": 2.8892415799794136,
      "grad_norm": 0.15226192772388458,
      "learning_rate": 3.695525500659368e-06,
      "loss": 0.4701,
      "mean_token_accuracy": 0.8514592722058296,
      "num_tokens": 1002244505.0,
      "step": 31403
    },
    {
      "entropy": 0.6509510334581137,
      "epoch": 2.8893335863508547,
      "grad_norm": 0.1733172982931137,
      "learning_rate": 3.6924586745177417e-06,
      "loss": 0.6416,
      "mean_token_accuracy": 0.8004279509186745,
      "num_tokens": 1002276175.0,
      "step": 31404
    },
    {
      "entropy": 0.5067733414471149,
      "epoch": 2.8894255927222963,
      "grad_norm": 0.15357652306556702,
      "learning_rate": 3.6893918483761155e-06,
      "loss": 0.4963,
      "mean_token_accuracy": 0.844310101121664,
      "num_tokens": 1002308231.0,
      "step": 31405
    },
    {
      "entropy": 0.4892267594113946,
      "epoch": 2.8895175990937374,
      "grad_norm": 0.15652510523796082,
      "learning_rate": 3.6863250222344893e-06,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8473416976630688,
      "num_tokens": 1002340506.0,
      "step": 31406
    },
    {
      "entropy": 0.44451572734396905,
      "epoch": 2.8896096054651785,
      "grad_norm": 0.14501510560512543,
      "learning_rate": 3.683258196092864e-06,
      "loss": 0.4319,
      "mean_token_accuracy": 0.8633502461016178,
      "num_tokens": 1002372769.0,
      "step": 31407
    },
    {
      "entropy": 0.47519912384450436,
      "epoch": 2.8897016118366197,
      "grad_norm": 0.1565016508102417,
      "learning_rate": 3.6801913699512378e-06,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8497588597238064,
      "num_tokens": 1002405264.0,
      "step": 31408
    },
    {
      "entropy": 0.5011252053081989,
      "epoch": 2.889793618208061,
      "grad_norm": 0.15708979964256287,
      "learning_rate": 3.6771245438096116e-06,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8453945927321911,
      "num_tokens": 1002437368.0,
      "step": 31409
    },
    {
      "entropy": 0.513928635045886,
      "epoch": 2.8898856245795024,
      "grad_norm": 0.15674932301044464,
      "learning_rate": 3.674057717667986e-06,
      "loss": 0.5121,
      "mean_token_accuracy": 0.8388689495623112,
      "num_tokens": 1002469541.0,
      "step": 31410
    },
    {
      "entropy": 0.45488819014281034,
      "epoch": 2.8899776309509435,
      "grad_norm": 0.1531049758195877,
      "learning_rate": 3.6709908915263596e-06,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8529062084853649,
      "num_tokens": 1002501926.0,
      "step": 31411
    },
    {
      "entropy": 0.537932613864541,
      "epoch": 2.8900696373223846,
      "grad_norm": 0.16157276928424835,
      "learning_rate": 3.6679240653847334e-06,
      "loss": 0.5268,
      "mean_token_accuracy": 0.8350026346743107,
      "num_tokens": 1002534112.0,
      "step": 31412
    },
    {
      "entropy": 0.5625253198668361,
      "epoch": 2.8901616436938258,
      "grad_norm": 0.16584038734436035,
      "learning_rate": 3.664857239243107e-06,
      "loss": 0.554,
      "mean_token_accuracy": 0.824642613530159,
      "num_tokens": 1002565650.0,
      "step": 31413
    },
    {
      "entropy": 0.4968835115432739,
      "epoch": 2.890253650065267,
      "grad_norm": 0.15069250762462616,
      "learning_rate": 3.661790413101481e-06,
      "loss": 0.4716,
      "mean_token_accuracy": 0.8468423262238503,
      "num_tokens": 1002596980.0,
      "step": 31414
    },
    {
      "entropy": 0.5340477274730802,
      "epoch": 2.8903456564367085,
      "grad_norm": 0.15777333080768585,
      "learning_rate": 3.6587235869598556e-06,
      "loss": 0.533,
      "mean_token_accuracy": 0.8347911536693573,
      "num_tokens": 1002629335.0,
      "step": 31415
    },
    {
      "entropy": 0.49387452472001314,
      "epoch": 2.8904376628081496,
      "grad_norm": 0.1500483751296997,
      "learning_rate": 3.6556567608182294e-06,
      "loss": 0.4805,
      "mean_token_accuracy": 0.8465676121413708,
      "num_tokens": 1002661618.0,
      "step": 31416
    },
    {
      "entropy": 0.4829049437539652,
      "epoch": 2.8905296691795908,
      "grad_norm": 0.1503726840019226,
      "learning_rate": 3.6525899346766037e-06,
      "loss": 0.4643,
      "mean_token_accuracy": 0.852664839476347,
      "num_tokens": 1002694175.0,
      "step": 31417
    },
    {
      "entropy": 0.5031869634985924,
      "epoch": 2.890621675551032,
      "grad_norm": 0.15836825966835022,
      "learning_rate": 3.6495231085349775e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8438434787094593,
      "num_tokens": 1002725569.0,
      "step": 31418
    },
    {
      "entropy": 0.4622834464535117,
      "epoch": 2.890713681922473,
      "grad_norm": 0.1568465232849121,
      "learning_rate": 3.6464562823933513e-06,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8542176559567451,
      "num_tokens": 1002756901.0,
      "step": 31419
    },
    {
      "entropy": 0.6260617077350616,
      "epoch": 2.8908056882939146,
      "grad_norm": 0.17840272188186646,
      "learning_rate": 3.643389456251725e-06,
      "loss": 0.6219,
      "mean_token_accuracy": 0.8080848418176174,
      "num_tokens": 1002788612.0,
      "step": 31420
    },
    {
      "entropy": 0.4950318248011172,
      "epoch": 2.8908976946653557,
      "grad_norm": 0.1537701040506363,
      "learning_rate": 3.640322630110099e-06,
      "loss": 0.472,
      "mean_token_accuracy": 0.8509997949004173,
      "num_tokens": 1002821038.0,
      "step": 31421
    },
    {
      "entropy": 0.44409867841750383,
      "epoch": 2.890989701036797,
      "grad_norm": 0.14568454027175903,
      "learning_rate": 3.6372558039684727e-06,
      "loss": 0.4201,
      "mean_token_accuracy": 0.8657036647200584,
      "num_tokens": 1002852403.0,
      "step": 31422
    },
    {
      "entropy": 0.42540982784703374,
      "epoch": 2.891081707408238,
      "grad_norm": 0.14956136047840118,
      "learning_rate": 3.6341889778268473e-06,
      "loss": 0.4032,
      "mean_token_accuracy": 0.8699908629059792,
      "num_tokens": 1002883644.0,
      "step": 31423
    },
    {
      "entropy": 0.5404672054573894,
      "epoch": 2.891173713779679,
      "grad_norm": 0.15890230238437653,
      "learning_rate": 3.631122151685221e-06,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8326514363288879,
      "num_tokens": 1002916412.0,
      "step": 31424
    },
    {
      "entropy": 0.48919372796081007,
      "epoch": 2.8912657201511207,
      "grad_norm": 0.1582607924938202,
      "learning_rate": 3.6280553255435953e-06,
      "loss": 0.4719,
      "mean_token_accuracy": 0.8503203615546227,
      "num_tokens": 1002948017.0,
      "step": 31425
    },
    {
      "entropy": 0.5582905365154147,
      "epoch": 2.891357726522562,
      "grad_norm": 0.16295886039733887,
      "learning_rate": 3.624988499401969e-06,
      "loss": 0.5483,
      "mean_token_accuracy": 0.8269627094268799,
      "num_tokens": 1002980244.0,
      "step": 31426
    },
    {
      "entropy": 0.48498890828341246,
      "epoch": 2.891449732894003,
      "grad_norm": 0.15405738353729248,
      "learning_rate": 3.621921673260343e-06,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8493396900594234,
      "num_tokens": 1003012236.0,
      "step": 31427
    },
    {
      "entropy": 0.5003170417621732,
      "epoch": 2.891541739265444,
      "grad_norm": 0.157309427857399,
      "learning_rate": 3.6188548471187167e-06,
      "loss": 0.491,
      "mean_token_accuracy": 0.8404478318989277,
      "num_tokens": 1003044693.0,
      "step": 31428
    },
    {
      "entropy": 0.5694604888558388,
      "epoch": 2.8916337456368852,
      "grad_norm": 0.16275560855865479,
      "learning_rate": 3.6157880209770905e-06,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8278425745666027,
      "num_tokens": 1003077032.0,
      "step": 31429
    },
    {
      "entropy": 0.5891969371587038,
      "epoch": 2.891725752008327,
      "grad_norm": 0.16790063679218292,
      "learning_rate": 3.612721194835465e-06,
      "loss": 0.5765,
      "mean_token_accuracy": 0.8145990185439587,
      "num_tokens": 1003107949.0,
      "step": 31430
    },
    {
      "entropy": 0.40249595255590975,
      "epoch": 2.891817758379768,
      "grad_norm": 0.13478687405586243,
      "learning_rate": 3.609654368693839e-06,
      "loss": 0.3828,
      "mean_token_accuracy": 0.8783187419176102,
      "num_tokens": 1003140358.0,
      "step": 31431
    },
    {
      "entropy": 0.42680397361982614,
      "epoch": 2.891909764751209,
      "grad_norm": 0.14363539218902588,
      "learning_rate": 3.606587542552213e-06,
      "loss": 0.4114,
      "mean_token_accuracy": 0.8674963675439358,
      "num_tokens": 1003172364.0,
      "step": 31432
    },
    {
      "entropy": 0.4301823005080223,
      "epoch": 2.8920017711226502,
      "grad_norm": 0.14176124334335327,
      "learning_rate": 3.603520716410587e-06,
      "loss": 0.4126,
      "mean_token_accuracy": 0.8653039187192917,
      "num_tokens": 1003203716.0,
      "step": 31433
    },
    {
      "entropy": 0.4788942816667259,
      "epoch": 2.8920937774940914,
      "grad_norm": 0.1608441323041916,
      "learning_rate": 3.600453890268961e-06,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8531585037708282,
      "num_tokens": 1003234080.0,
      "step": 31434
    },
    {
      "entropy": 0.49124930147081614,
      "epoch": 2.892185783865533,
      "grad_norm": 0.15743164718151093,
      "learning_rate": 3.5973870641273346e-06,
      "loss": 0.4846,
      "mean_token_accuracy": 0.8474209271371365,
      "num_tokens": 1003266031.0,
      "step": 31435
    },
    {
      "entropy": 0.40799681190401316,
      "epoch": 2.892277790236974,
      "grad_norm": 0.13928256928920746,
      "learning_rate": 3.5943202379857084e-06,
      "loss": 0.3911,
      "mean_token_accuracy": 0.8729040399193764,
      "num_tokens": 1003298333.0,
      "step": 31436
    },
    {
      "entropy": 0.5381042174994946,
      "epoch": 2.892369796608415,
      "grad_norm": 0.16535069048404694,
      "learning_rate": 3.5912534118440826e-06,
      "loss": 0.5489,
      "mean_token_accuracy": 0.8254694603383541,
      "num_tokens": 1003329982.0,
      "step": 31437
    },
    {
      "entropy": 0.5597025286406279,
      "epoch": 2.8924618029798563,
      "grad_norm": 0.16233958303928375,
      "learning_rate": 3.588186585702457e-06,
      "loss": 0.5481,
      "mean_token_accuracy": 0.8313385471701622,
      "num_tokens": 1003362750.0,
      "step": 31438
    },
    {
      "entropy": 0.570269487798214,
      "epoch": 2.8925538093512975,
      "grad_norm": 0.16756050288677216,
      "learning_rate": 3.585119759560831e-06,
      "loss": 0.5606,
      "mean_token_accuracy": 0.8215921334922314,
      "num_tokens": 1003394070.0,
      "step": 31439
    },
    {
      "entropy": 0.5149770537391305,
      "epoch": 2.892645815722739,
      "grad_norm": 0.15986886620521545,
      "learning_rate": 3.582052933419205e-06,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8402603901922703,
      "num_tokens": 1003425763.0,
      "step": 31440
    },
    {
      "entropy": 0.544053903548047,
      "epoch": 2.89273782209418,
      "grad_norm": 0.16100646555423737,
      "learning_rate": 3.5789861072775787e-06,
      "loss": 0.5267,
      "mean_token_accuracy": 0.8326406069099903,
      "num_tokens": 1003457674.0,
      "step": 31441
    },
    {
      "entropy": 0.49620294920168817,
      "epoch": 2.8928298284656213,
      "grad_norm": 0.15950819849967957,
      "learning_rate": 3.5759192811359525e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8441822677850723,
      "num_tokens": 1003489315.0,
      "step": 31442
    },
    {
      "entropy": 0.5168964308686554,
      "epoch": 2.8929218348370624,
      "grad_norm": 0.16527850925922394,
      "learning_rate": 3.5728524549943263e-06,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8391883112490177,
      "num_tokens": 1003520806.0,
      "step": 31443
    },
    {
      "entropy": 0.43447909387759864,
      "epoch": 2.8930138412085036,
      "grad_norm": 0.14191457629203796,
      "learning_rate": 3.5697856288527e-06,
      "loss": 0.4228,
      "mean_token_accuracy": 0.8636290989816189,
      "num_tokens": 1003553117.0,
      "step": 31444
    },
    {
      "entropy": 0.5820436631329358,
      "epoch": 2.893105847579945,
      "grad_norm": 0.173892080783844,
      "learning_rate": 3.5667188027110743e-06,
      "loss": 0.5824,
      "mean_token_accuracy": 0.8206462524831295,
      "num_tokens": 1003584887.0,
      "step": 31445
    },
    {
      "entropy": 0.6369138238951564,
      "epoch": 2.8931978539513863,
      "grad_norm": 0.17237398028373718,
      "learning_rate": 3.5636519765694485e-06,
      "loss": 0.6248,
      "mean_token_accuracy": 0.8048393055796623,
      "num_tokens": 1003616244.0,
      "step": 31446
    },
    {
      "entropy": 0.5213831551373005,
      "epoch": 2.8932898603228274,
      "grad_norm": 0.16239586472511292,
      "learning_rate": 3.5605851504278228e-06,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8356528058648109,
      "num_tokens": 1003647987.0,
      "step": 31447
    },
    {
      "entropy": 0.5931692346930504,
      "epoch": 2.8933818666942686,
      "grad_norm": 0.17045553028583527,
      "learning_rate": 3.5575183242861966e-06,
      "loss": 0.5908,
      "mean_token_accuracy": 0.81211381778121,
      "num_tokens": 1003679745.0,
      "step": 31448
    },
    {
      "entropy": 0.5652593169361353,
      "epoch": 2.8934738730657097,
      "grad_norm": 0.16631245613098145,
      "learning_rate": 3.5544514981445704e-06,
      "loss": 0.5538,
      "mean_token_accuracy": 0.8225578665733337,
      "num_tokens": 1003711169.0,
      "step": 31449
    },
    {
      "entropy": 0.4714051093906164,
      "epoch": 2.8935658794371513,
      "grad_norm": 0.15750011801719666,
      "learning_rate": 3.551384672002944e-06,
      "loss": 0.4541,
      "mean_token_accuracy": 0.8540040068328381,
      "num_tokens": 1003743043.0,
      "step": 31450
    },
    {
      "entropy": 0.47640044428408146,
      "epoch": 2.8936578858085924,
      "grad_norm": 0.15523454546928406,
      "learning_rate": 3.548317845861318e-06,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8522394746541977,
      "num_tokens": 1003775300.0,
      "step": 31451
    },
    {
      "entropy": 0.49469313584268093,
      "epoch": 2.8937498921800335,
      "grad_norm": 0.159311905503273,
      "learning_rate": 3.545251019719692e-06,
      "loss": 0.4945,
      "mean_token_accuracy": 0.8452989123761654,
      "num_tokens": 1003807168.0,
      "step": 31452
    },
    {
      "entropy": 0.48751228069886565,
      "epoch": 2.8938418985514747,
      "grad_norm": 0.15745943784713745,
      "learning_rate": 3.5421841935780664e-06,
      "loss": 0.476,
      "mean_token_accuracy": 0.8491556458175182,
      "num_tokens": 1003839101.0,
      "step": 31453
    },
    {
      "entropy": 0.5262865088880062,
      "epoch": 2.893933904922916,
      "grad_norm": 0.1604430079460144,
      "learning_rate": 3.5391173674364406e-06,
      "loss": 0.5096,
      "mean_token_accuracy": 0.8422942869365215,
      "num_tokens": 1003870859.0,
      "step": 31454
    },
    {
      "entropy": 0.5074722487479448,
      "epoch": 2.8940259112943574,
      "grad_norm": 0.1624491810798645,
      "learning_rate": 3.5360505412948144e-06,
      "loss": 0.4993,
      "mean_token_accuracy": 0.843735683709383,
      "num_tokens": 1003901889.0,
      "step": 31455
    },
    {
      "entropy": 0.5866063805297017,
      "epoch": 2.8941179176657985,
      "grad_norm": 0.1728864163160324,
      "learning_rate": 3.5329837151531882e-06,
      "loss": 0.5772,
      "mean_token_accuracy": 0.82065225020051,
      "num_tokens": 1003932943.0,
      "step": 31456
    },
    {
      "entropy": 0.4986608596518636,
      "epoch": 2.8942099240372396,
      "grad_norm": 0.15117351710796356,
      "learning_rate": 3.529916889011562e-06,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8490350916981697,
      "num_tokens": 1003964996.0,
      "step": 31457
    },
    {
      "entropy": 0.574238421395421,
      "epoch": 2.894301930408681,
      "grad_norm": 0.16488461196422577,
      "learning_rate": 3.526850062869936e-06,
      "loss": 0.5629,
      "mean_token_accuracy": 0.8217437081038952,
      "num_tokens": 1003997143.0,
      "step": 31458
    },
    {
      "entropy": 0.5145159258972853,
      "epoch": 2.894393936780122,
      "grad_norm": 0.15862447023391724,
      "learning_rate": 3.5237832367283096e-06,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8428985141217709,
      "num_tokens": 1004029006.0,
      "step": 31459
    },
    {
      "entropy": 0.5200365888886154,
      "epoch": 2.8944859431515635,
      "grad_norm": 0.16047672927379608,
      "learning_rate": 3.520716410586684e-06,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8404472693800926,
      "num_tokens": 1004060629.0,
      "step": 31460
    },
    {
      "entropy": 0.49377044290304184,
      "epoch": 2.8945779495230046,
      "grad_norm": 0.14921662211418152,
      "learning_rate": 3.517649584445058e-06,
      "loss": 0.474,
      "mean_token_accuracy": 0.8487715721130371,
      "num_tokens": 1004092425.0,
      "step": 31461
    },
    {
      "entropy": 0.4701189836487174,
      "epoch": 2.8946699558944458,
      "grad_norm": 0.1598646640777588,
      "learning_rate": 3.5145827583034323e-06,
      "loss": 0.4635,
      "mean_token_accuracy": 0.850188884884119,
      "num_tokens": 1004123667.0,
      "step": 31462
    },
    {
      "entropy": 0.5139106521382928,
      "epoch": 2.894761962265887,
      "grad_norm": 0.15997561812400818,
      "learning_rate": 3.511515932161806e-06,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8415982276201248,
      "num_tokens": 1004155441.0,
      "step": 31463
    },
    {
      "entropy": 0.48181752336677164,
      "epoch": 2.894853968637328,
      "grad_norm": 0.15753374993801117,
      "learning_rate": 3.50844910602018e-06,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8491030670702457,
      "num_tokens": 1004185737.0,
      "step": 31464
    },
    {
      "entropy": 0.5684602558612823,
      "epoch": 2.8949459750087696,
      "grad_norm": 0.16624915599822998,
      "learning_rate": 3.5053822798785537e-06,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8248395547270775,
      "num_tokens": 1004217290.0,
      "step": 31465
    },
    {
      "entropy": 0.4672912092646584,
      "epoch": 2.8950379813802107,
      "grad_norm": 0.15614759922027588,
      "learning_rate": 3.5023154537369275e-06,
      "loss": 0.4596,
      "mean_token_accuracy": 0.850574005395174,
      "num_tokens": 1004250058.0,
      "step": 31466
    },
    {
      "entropy": 0.5873409165069461,
      "epoch": 2.895129987751652,
      "grad_norm": 0.16640928387641907,
      "learning_rate": 3.4992486275953017e-06,
      "loss": 0.5843,
      "mean_token_accuracy": 0.8205162882804871,
      "num_tokens": 1004282121.0,
      "step": 31467
    },
    {
      "entropy": 0.5506027890369296,
      "epoch": 2.895221994123093,
      "grad_norm": 0.16523128747940063,
      "learning_rate": 3.496181801453676e-06,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8287897035479546,
      "num_tokens": 1004313368.0,
      "step": 31468
    },
    {
      "entropy": 0.5385595336556435,
      "epoch": 2.895314000494534,
      "grad_norm": 0.1638544201850891,
      "learning_rate": 3.49311497531205e-06,
      "loss": 0.5277,
      "mean_token_accuracy": 0.831238642334938,
      "num_tokens": 1004345243.0,
      "step": 31469
    },
    {
      "entropy": 0.4957367032766342,
      "epoch": 2.8954060068659757,
      "grad_norm": 0.15717288851737976,
      "learning_rate": 3.490048149170424e-06,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8457636646926403,
      "num_tokens": 1004377683.0,
      "step": 31470
    },
    {
      "entropy": 0.5205638613551855,
      "epoch": 2.895498013237417,
      "grad_norm": 0.16606296598911285,
      "learning_rate": 3.4869813230287978e-06,
      "loss": 0.5111,
      "mean_token_accuracy": 0.8370200358331203,
      "num_tokens": 1004409528.0,
      "step": 31471
    },
    {
      "entropy": 0.6240067146718502,
      "epoch": 2.895590019608858,
      "grad_norm": 0.17197352647781372,
      "learning_rate": 3.4839144968871716e-06,
      "loss": 0.615,
      "mean_token_accuracy": 0.8089226521551609,
      "num_tokens": 1004441231.0,
      "step": 31472
    },
    {
      "entropy": 0.5934379026293755,
      "epoch": 2.895682025980299,
      "grad_norm": 0.1703081876039505,
      "learning_rate": 3.4808476707455454e-06,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8170631639659405,
      "num_tokens": 1004473127.0,
      "step": 31473
    },
    {
      "entropy": 0.4973849318921566,
      "epoch": 2.8957740323517402,
      "grad_norm": 0.15945293009281158,
      "learning_rate": 3.477780844603919e-06,
      "loss": 0.488,
      "mean_token_accuracy": 0.8433755598962307,
      "num_tokens": 1004504472.0,
      "step": 31474
    },
    {
      "entropy": 0.45920765958726406,
      "epoch": 2.895866038723182,
      "grad_norm": 0.1554589718580246,
      "learning_rate": 3.4747140184622934e-06,
      "loss": 0.4546,
      "mean_token_accuracy": 0.8529760800302029,
      "num_tokens": 1004536454.0,
      "step": 31475
    },
    {
      "entropy": 0.6398505847901106,
      "epoch": 2.895958045094623,
      "grad_norm": 0.1746576577425003,
      "learning_rate": 3.4716471923206676e-06,
      "loss": 0.6334,
      "mean_token_accuracy": 0.8001558147370815,
      "num_tokens": 1004569216.0,
      "step": 31476
    },
    {
      "entropy": 0.5014181775040925,
      "epoch": 2.896050051466064,
      "grad_norm": 0.1670459359884262,
      "learning_rate": 3.468580366179042e-06,
      "loss": 0.4914,
      "mean_token_accuracy": 0.841837901622057,
      "num_tokens": 1004600990.0,
      "step": 31477
    },
    {
      "entropy": 0.5014363462105393,
      "epoch": 2.8961420578375052,
      "grad_norm": 0.1614161878824234,
      "learning_rate": 3.4655135400374157e-06,
      "loss": 0.4927,
      "mean_token_accuracy": 0.8437250927090645,
      "num_tokens": 1004632278.0,
      "step": 31478
    },
    {
      "entropy": 0.47857082821428776,
      "epoch": 2.8962340642089464,
      "grad_norm": 0.15312601625919342,
      "learning_rate": 3.4624467138957895e-06,
      "loss": 0.4628,
      "mean_token_accuracy": 0.8550199531018734,
      "num_tokens": 1004663944.0,
      "step": 31479
    },
    {
      "entropy": 0.4588276371359825,
      "epoch": 2.896326070580388,
      "grad_norm": 0.1522202491760254,
      "learning_rate": 3.4593798877541633e-06,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8568122871220112,
      "num_tokens": 1004695529.0,
      "step": 31480
    },
    {
      "entropy": 0.5075106052681804,
      "epoch": 2.896418076951829,
      "grad_norm": 0.1620834618806839,
      "learning_rate": 3.456313061612537e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8449946716427803,
      "num_tokens": 1004726474.0,
      "step": 31481
    },
    {
      "entropy": 0.4788057536352426,
      "epoch": 2.89651008332327,
      "grad_norm": 0.15635384619235992,
      "learning_rate": 3.4532462354709113e-06,
      "loss": 0.4643,
      "mean_token_accuracy": 0.8506386466324329,
      "num_tokens": 1004757854.0,
      "step": 31482
    },
    {
      "entropy": 0.5670123863965273,
      "epoch": 2.8966020896947113,
      "grad_norm": 0.16267086565494537,
      "learning_rate": 3.450179409329285e-06,
      "loss": 0.5512,
      "mean_token_accuracy": 0.8249162845313549,
      "num_tokens": 1004789766.0,
      "step": 31483
    },
    {
      "entropy": 0.555198878981173,
      "epoch": 2.8966940960661525,
      "grad_norm": 0.16142535209655762,
      "learning_rate": 3.4471125831876597e-06,
      "loss": 0.5278,
      "mean_token_accuracy": 0.831569392234087,
      "num_tokens": 1004821506.0,
      "step": 31484
    },
    {
      "entropy": 0.5400498835369945,
      "epoch": 2.896786102437594,
      "grad_norm": 0.15953858196735382,
      "learning_rate": 3.4440457570460335e-06,
      "loss": 0.5172,
      "mean_token_accuracy": 0.8372411206364632,
      "num_tokens": 1004854048.0,
      "step": 31485
    },
    {
      "entropy": 0.45697123266290873,
      "epoch": 2.896878108809035,
      "grad_norm": 0.15463407337665558,
      "learning_rate": 3.4409789309044073e-06,
      "loss": 0.4384,
      "mean_token_accuracy": 0.8589780367910862,
      "num_tokens": 1004885133.0,
      "step": 31486
    },
    {
      "entropy": 0.6256903917528689,
      "epoch": 2.8969701151804763,
      "grad_norm": 0.17397968471050262,
      "learning_rate": 3.437912104762781e-06,
      "loss": 0.6046,
      "mean_token_accuracy": 0.8086534328758717,
      "num_tokens": 1004916325.0,
      "step": 31487
    },
    {
      "entropy": 0.5263466984033585,
      "epoch": 2.8970621215519174,
      "grad_norm": 0.15930941700935364,
      "learning_rate": 3.434845278621155e-06,
      "loss": 0.5168,
      "mean_token_accuracy": 0.8362919762730598,
      "num_tokens": 1004948547.0,
      "step": 31488
    },
    {
      "entropy": 0.5120844226330519,
      "epoch": 2.8971541279233586,
      "grad_norm": 0.15620924532413483,
      "learning_rate": 3.4317784524795287e-06,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8432424440979958,
      "num_tokens": 1004980789.0,
      "step": 31489
    },
    {
      "entropy": 0.5290216105058789,
      "epoch": 2.8972461342948,
      "grad_norm": 0.1621103286743164,
      "learning_rate": 3.428711626337903e-06,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8338179811835289,
      "num_tokens": 1005012563.0,
      "step": 31490
    },
    {
      "entropy": 0.5951716173440218,
      "epoch": 2.8973381406662413,
      "grad_norm": 0.1697954535484314,
      "learning_rate": 3.425644800196277e-06,
      "loss": 0.5831,
      "mean_token_accuracy": 0.8190478757023811,
      "num_tokens": 1005043593.0,
      "step": 31491
    },
    {
      "entropy": 0.5352549459785223,
      "epoch": 2.8974301470376824,
      "grad_norm": 0.16027900576591492,
      "learning_rate": 3.4225779740546514e-06,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8334017768502235,
      "num_tokens": 1005075379.0,
      "step": 31492
    },
    {
      "entropy": 0.503842013888061,
      "epoch": 2.8975221534091236,
      "grad_norm": 0.15924477577209473,
      "learning_rate": 3.419511147913025e-06,
      "loss": 0.4914,
      "mean_token_accuracy": 0.8411090523004532,
      "num_tokens": 1005106441.0,
      "step": 31493
    },
    {
      "entropy": 0.4432165178004652,
      "epoch": 2.8976141597805647,
      "grad_norm": 0.1521245390176773,
      "learning_rate": 3.416444321771399e-06,
      "loss": 0.4282,
      "mean_token_accuracy": 0.864767499268055,
      "num_tokens": 1005138745.0,
      "step": 31494
    },
    {
      "entropy": 0.5419009979814291,
      "epoch": 2.8977061661520063,
      "grad_norm": 0.159159854054451,
      "learning_rate": 3.413377495629773e-06,
      "loss": 0.5315,
      "mean_token_accuracy": 0.8339160345494747,
      "num_tokens": 1005171120.0,
      "step": 31495
    },
    {
      "entropy": 0.5486129922792315,
      "epoch": 2.8977981725234474,
      "grad_norm": 0.1584559977054596,
      "learning_rate": 3.4103106694881466e-06,
      "loss": 0.546,
      "mean_token_accuracy": 0.8292376212775707,
      "num_tokens": 1005203759.0,
      "step": 31496
    },
    {
      "entropy": 0.4189792168326676,
      "epoch": 2.8978901788948885,
      "grad_norm": 0.14391793310642242,
      "learning_rate": 3.407243843346521e-06,
      "loss": 0.411,
      "mean_token_accuracy": 0.8677356392145157,
      "num_tokens": 1005235821.0,
      "step": 31497
    },
    {
      "entropy": 0.5458352100104094,
      "epoch": 2.8979821852663297,
      "grad_norm": 0.16536052525043488,
      "learning_rate": 3.4041770172048946e-06,
      "loss": 0.524,
      "mean_token_accuracy": 0.835627842694521,
      "num_tokens": 1005267539.0,
      "step": 31498
    },
    {
      "entropy": 0.5822840565815568,
      "epoch": 2.898074191637771,
      "grad_norm": 0.16359061002731323,
      "learning_rate": 3.4011101910632693e-06,
      "loss": 0.568,
      "mean_token_accuracy": 0.8223526254296303,
      "num_tokens": 1005299410.0,
      "step": 31499
    },
    {
      "entropy": 0.5821058712899685,
      "epoch": 2.8981661980092124,
      "grad_norm": 0.16895368695259094,
      "learning_rate": 3.398043364921643e-06,
      "loss": 0.5697,
      "mean_token_accuracy": 0.8189950846135616,
      "num_tokens": 1005331214.0,
      "step": 31500
    },
    {
      "entropy": 0.5653657764196396,
      "epoch": 2.8982582043806535,
      "grad_norm": 0.16370351612567902,
      "learning_rate": 3.394976538780017e-06,
      "loss": 0.56,
      "mean_token_accuracy": 0.8274785093963146,
      "num_tokens": 1005363444.0,
      "step": 31501
    },
    {
      "entropy": 0.527248153463006,
      "epoch": 2.8983502107520946,
      "grad_norm": 0.1592574566602707,
      "learning_rate": 3.3919097126383907e-06,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8364933468401432,
      "num_tokens": 1005395474.0,
      "step": 31502
    },
    {
      "entropy": 0.4925485122948885,
      "epoch": 2.898442217123536,
      "grad_norm": 0.1532878428697586,
      "learning_rate": 3.3888428864967645e-06,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8461514785885811,
      "num_tokens": 1005427584.0,
      "step": 31503
    },
    {
      "entropy": 0.49550397938583046,
      "epoch": 2.898534223494977,
      "grad_norm": 0.1560148447751999,
      "learning_rate": 3.3857760603551383e-06,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8447991274297237,
      "num_tokens": 1005459730.0,
      "step": 31504
    },
    {
      "entropy": 0.4551802477799356,
      "epoch": 2.8986262298664185,
      "grad_norm": 0.15141625702381134,
      "learning_rate": 3.3827092342135125e-06,
      "loss": 0.4432,
      "mean_token_accuracy": 0.8581014014780521,
      "num_tokens": 1005491991.0,
      "step": 31505
    },
    {
      "entropy": 0.5203266041353345,
      "epoch": 2.8987182362378596,
      "grad_norm": 0.1576836109161377,
      "learning_rate": 3.3796424080718863e-06,
      "loss": 0.5116,
      "mean_token_accuracy": 0.8378792516887188,
      "num_tokens": 1005524185.0,
      "step": 31506
    },
    {
      "entropy": 0.38026899471879005,
      "epoch": 2.8988102426093008,
      "grad_norm": 0.1426028311252594,
      "learning_rate": 3.376575581930261e-06,
      "loss": 0.3703,
      "mean_token_accuracy": 0.8783901296555996,
      "num_tokens": 1005555763.0,
      "step": 31507
    },
    {
      "entropy": 0.5132184284739196,
      "epoch": 2.898902248980742,
      "grad_norm": 0.1583632081747055,
      "learning_rate": 3.3735087557886348e-06,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8382533974945545,
      "num_tokens": 1005587695.0,
      "step": 31508
    },
    {
      "entropy": 0.5810971856117249,
      "epoch": 2.898994255352183,
      "grad_norm": 0.16650788486003876,
      "learning_rate": 3.3704419296470086e-06,
      "loss": 0.5651,
      "mean_token_accuracy": 0.8169837631285191,
      "num_tokens": 1005619132.0,
      "step": 31509
    },
    {
      "entropy": 0.4931301986798644,
      "epoch": 2.8990862617236246,
      "grad_norm": 0.1533353328704834,
      "learning_rate": 3.3673751035053824e-06,
      "loss": 0.4769,
      "mean_token_accuracy": 0.848039548844099,
      "num_tokens": 1005651240.0,
      "step": 31510
    },
    {
      "entropy": 0.5519805829972029,
      "epoch": 2.8991782680950657,
      "grad_norm": 0.1597803682088852,
      "learning_rate": 3.364308277363756e-06,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8317958489060402,
      "num_tokens": 1005683563.0,
      "step": 31511
    },
    {
      "entropy": 0.5850732959806919,
      "epoch": 2.899270274466507,
      "grad_norm": 0.16327480971813202,
      "learning_rate": 3.3612414512221304e-06,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8161827623844147,
      "num_tokens": 1005715325.0,
      "step": 31512
    },
    {
      "entropy": 0.4079348719678819,
      "epoch": 2.899362280837948,
      "grad_norm": 0.1405681073665619,
      "learning_rate": 3.358174625080504e-06,
      "loss": 0.3935,
      "mean_token_accuracy": 0.8711353093385696,
      "num_tokens": 1005747459.0,
      "step": 31513
    },
    {
      "entropy": 0.5239689331501722,
      "epoch": 2.899454287209389,
      "grad_norm": 0.15909230709075928,
      "learning_rate": 3.355107798938879e-06,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8395080827176571,
      "num_tokens": 1005779129.0,
      "step": 31514
    },
    {
      "entropy": 0.5109254075214267,
      "epoch": 2.8995462935808307,
      "grad_norm": 0.15587447583675385,
      "learning_rate": 3.3520409727972526e-06,
      "loss": 0.4963,
      "mean_token_accuracy": 0.8442014344036579,
      "num_tokens": 1005811020.0,
      "step": 31515
    },
    {
      "entropy": 0.5857046023011208,
      "epoch": 2.899638299952272,
      "grad_norm": 0.16929425299167633,
      "learning_rate": 3.3489741466556264e-06,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8211199752986431,
      "num_tokens": 1005842819.0,
      "step": 31516
    },
    {
      "entropy": 0.4423145179171115,
      "epoch": 2.899730306323713,
      "grad_norm": 0.1510288417339325,
      "learning_rate": 3.3459073205140002e-06,
      "loss": 0.422,
      "mean_token_accuracy": 0.8650008887052536,
      "num_tokens": 1005874629.0,
      "step": 31517
    },
    {
      "entropy": 0.5482199238613248,
      "epoch": 2.899822312695154,
      "grad_norm": 0.16713447868824005,
      "learning_rate": 3.342840494372374e-06,
      "loss": 0.5493,
      "mean_token_accuracy": 0.8302322551608086,
      "num_tokens": 1005906761.0,
      "step": 31518
    },
    {
      "entropy": 0.47212606738321483,
      "epoch": 2.8999143190665952,
      "grad_norm": 0.15082073211669922,
      "learning_rate": 3.339773668230748e-06,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8486284986138344,
      "num_tokens": 1005939249.0,
      "step": 31519
    },
    {
      "entropy": 0.4616201496683061,
      "epoch": 2.900006325438037,
      "grad_norm": 0.15026533603668213,
      "learning_rate": 3.336706842089122e-06,
      "loss": 0.4407,
      "mean_token_accuracy": 0.8573712147772312,
      "num_tokens": 1005970817.0,
      "step": 31520
    },
    {
      "entropy": 0.4645165540277958,
      "epoch": 2.900098331809478,
      "grad_norm": 0.1521545946598053,
      "learning_rate": 3.333640015947496e-06,
      "loss": 0.4477,
      "mean_token_accuracy": 0.8537127003073692,
      "num_tokens": 1006002532.0,
      "step": 31521
    },
    {
      "entropy": 0.4636175250634551,
      "epoch": 2.900190338180919,
      "grad_norm": 0.15337876975536346,
      "learning_rate": 3.3305731898058705e-06,
      "loss": 0.4439,
      "mean_token_accuracy": 0.8583484999835491,
      "num_tokens": 1006034109.0,
      "step": 31522
    },
    {
      "entropy": 0.5437806975096464,
      "epoch": 2.9002823445523602,
      "grad_norm": 0.15615496039390564,
      "learning_rate": 3.3275063636642443e-06,
      "loss": 0.5321,
      "mean_token_accuracy": 0.831374604254961,
      "num_tokens": 1006066746.0,
      "step": 31523
    },
    {
      "entropy": 0.5185043624369428,
      "epoch": 2.9003743509238014,
      "grad_norm": 0.15772078931331635,
      "learning_rate": 3.324439537522618e-06,
      "loss": 0.5054,
      "mean_token_accuracy": 0.8407393284142017,
      "num_tokens": 1006099009.0,
      "step": 31524
    },
    {
      "entropy": 0.4412778550758958,
      "epoch": 2.900466357295243,
      "grad_norm": 0.15362468361854553,
      "learning_rate": 3.321372711380992e-06,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8602082170546055,
      "num_tokens": 1006130748.0,
      "step": 31525
    },
    {
      "entropy": 0.4118336362298578,
      "epoch": 2.900558363666684,
      "grad_norm": 0.1428271234035492,
      "learning_rate": 3.3183058852393657e-06,
      "loss": 0.4012,
      "mean_token_accuracy": 0.871694665402174,
      "num_tokens": 1006162378.0,
      "step": 31526
    },
    {
      "entropy": 0.5337100885808468,
      "epoch": 2.900650370038125,
      "grad_norm": 0.15661467611789703,
      "learning_rate": 3.31523905909774e-06,
      "loss": 0.5139,
      "mean_token_accuracy": 0.8401233442127705,
      "num_tokens": 1006193974.0,
      "step": 31527
    },
    {
      "entropy": 0.4894585858564824,
      "epoch": 2.9007423764095663,
      "grad_norm": 0.1558380275964737,
      "learning_rate": 3.3121722329561137e-06,
      "loss": 0.4879,
      "mean_token_accuracy": 0.8465963788330555,
      "num_tokens": 1006226437.0,
      "step": 31528
    },
    {
      "entropy": 0.5211427100002766,
      "epoch": 2.9008343827810075,
      "grad_norm": 0.15813888609409332,
      "learning_rate": 3.3091054068144884e-06,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8419184125959873,
      "num_tokens": 1006258469.0,
      "step": 31529
    },
    {
      "entropy": 0.5192622151225805,
      "epoch": 2.900926389152449,
      "grad_norm": 0.16047906875610352,
      "learning_rate": 3.306038580672862e-06,
      "loss": 0.5057,
      "mean_token_accuracy": 0.8367398418486118,
      "num_tokens": 1006290458.0,
      "step": 31530
    },
    {
      "entropy": 0.46523651760071516,
      "epoch": 2.90101839552389,
      "grad_norm": 0.14959746599197388,
      "learning_rate": 3.302971754531236e-06,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8548380173742771,
      "num_tokens": 1006322782.0,
      "step": 31531
    },
    {
      "entropy": 0.44858345040120184,
      "epoch": 2.9011104018953313,
      "grad_norm": 0.15546540915966034,
      "learning_rate": 3.2999049283896098e-06,
      "loss": 0.4347,
      "mean_token_accuracy": 0.8588090427219868,
      "num_tokens": 1006354618.0,
      "step": 31532
    },
    {
      "entropy": 0.5626152222976089,
      "epoch": 2.9012024082667724,
      "grad_norm": 0.16569732129573822,
      "learning_rate": 3.2968381022479836e-06,
      "loss": 0.5546,
      "mean_token_accuracy": 0.8255561701953411,
      "num_tokens": 1006387011.0,
      "step": 31533
    },
    {
      "entropy": 0.5801016166806221,
      "epoch": 2.9012944146382136,
      "grad_norm": 0.16973631083965302,
      "learning_rate": 3.2937712761063574e-06,
      "loss": 0.5804,
      "mean_token_accuracy": 0.820213820785284,
      "num_tokens": 1006418036.0,
      "step": 31534
    },
    {
      "entropy": 0.5505087310448289,
      "epoch": 2.901386421009655,
      "grad_norm": 0.16412346065044403,
      "learning_rate": 3.2907044499647316e-06,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8279671594500542,
      "num_tokens": 1006449827.0,
      "step": 31535
    },
    {
      "entropy": 0.5310875079594553,
      "epoch": 2.9014784273810963,
      "grad_norm": 0.15910454094409943,
      "learning_rate": 3.2876376238231054e-06,
      "loss": 0.5223,
      "mean_token_accuracy": 0.8332143202424049,
      "num_tokens": 1006481702.0,
      "step": 31536
    },
    {
      "entropy": 0.5936904605478048,
      "epoch": 2.9015704337525374,
      "grad_norm": 0.16947390139102936,
      "learning_rate": 3.28457079768148e-06,
      "loss": 0.5844,
      "mean_token_accuracy": 0.8160713464021683,
      "num_tokens": 1006513545.0,
      "step": 31537
    },
    {
      "entropy": 0.528499536216259,
      "epoch": 2.9016624401239786,
      "grad_norm": 0.15993405878543854,
      "learning_rate": 3.281503971539854e-06,
      "loss": 0.5142,
      "mean_token_accuracy": 0.8366597443819046,
      "num_tokens": 1006545757.0,
      "step": 31538
    },
    {
      "entropy": 0.5646276418119669,
      "epoch": 2.9017544464954197,
      "grad_norm": 0.15994174778461456,
      "learning_rate": 3.2784371453982277e-06,
      "loss": 0.5512,
      "mean_token_accuracy": 0.82488664239645,
      "num_tokens": 1006577568.0,
      "step": 31539
    },
    {
      "entropy": 0.49315325543284416,
      "epoch": 2.9018464528668613,
      "grad_norm": 0.15184767544269562,
      "learning_rate": 3.2753703192566015e-06,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8473564386367798,
      "num_tokens": 1006609353.0,
      "step": 31540
    },
    {
      "entropy": 0.49224151112139225,
      "epoch": 2.9019384592383024,
      "grad_norm": 0.16077107191085815,
      "learning_rate": 3.2723034931149753e-06,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8456478528678417,
      "num_tokens": 1006640494.0,
      "step": 31541
    },
    {
      "entropy": 0.5203180620446801,
      "epoch": 2.9020304656097435,
      "grad_norm": 0.16169868409633636,
      "learning_rate": 3.2692366669733495e-06,
      "loss": 0.5106,
      "mean_token_accuracy": 0.8374363072216511,
      "num_tokens": 1006672577.0,
      "step": 31542
    },
    {
      "entropy": 0.4805606543086469,
      "epoch": 2.9021224719811847,
      "grad_norm": 0.15165621042251587,
      "learning_rate": 3.2661698408317233e-06,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8482798710465431,
      "num_tokens": 1006704794.0,
      "step": 31543
    },
    {
      "entropy": 0.4363651215098798,
      "epoch": 2.902214478352626,
      "grad_norm": 0.14829368889331818,
      "learning_rate": 3.263103014690097e-06,
      "loss": 0.4236,
      "mean_token_accuracy": 0.8663235455751419,
      "num_tokens": 1006737156.0,
      "step": 31544
    },
    {
      "entropy": 0.44634343427605927,
      "epoch": 2.9023064847240674,
      "grad_norm": 0.1487882137298584,
      "learning_rate": 3.2600361885484717e-06,
      "loss": 0.4368,
      "mean_token_accuracy": 0.8630665056407452,
      "num_tokens": 1006769564.0,
      "step": 31545
    },
    {
      "entropy": 0.3983557624742389,
      "epoch": 2.9023984910955085,
      "grad_norm": 0.14082013070583344,
      "learning_rate": 3.2569693624068455e-06,
      "loss": 0.3898,
      "mean_token_accuracy": 0.874240018427372,
      "num_tokens": 1006802190.0,
      "step": 31546
    },
    {
      "entropy": 0.5517287468537688,
      "epoch": 2.9024904974669496,
      "grad_norm": 0.167547807097435,
      "learning_rate": 3.2539025362652193e-06,
      "loss": 0.542,
      "mean_token_accuracy": 0.826974093914032,
      "num_tokens": 1006834788.0,
      "step": 31547
    },
    {
      "entropy": 0.48695842223241925,
      "epoch": 2.902582503838391,
      "grad_norm": 0.15513822436332703,
      "learning_rate": 3.250835710123593e-06,
      "loss": 0.4788,
      "mean_token_accuracy": 0.850761853158474,
      "num_tokens": 1006866406.0,
      "step": 31548
    },
    {
      "entropy": 0.4774362873286009,
      "epoch": 2.902674510209832,
      "grad_norm": 0.1580434888601303,
      "learning_rate": 3.247768883981967e-06,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8486513793468475,
      "num_tokens": 1006898755.0,
      "step": 31549
    },
    {
      "entropy": 0.5447227703407407,
      "epoch": 2.9027665165812735,
      "grad_norm": 0.16247224807739258,
      "learning_rate": 3.244702057840341e-06,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8319791480898857,
      "num_tokens": 1006930676.0,
      "step": 31550
    },
    {
      "entropy": 0.5153209650889039,
      "epoch": 2.9028585229527146,
      "grad_norm": 0.16143395006656647,
      "learning_rate": 3.241635231698715e-06,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8381151780486107,
      "num_tokens": 1006962402.0,
      "step": 31551
    },
    {
      "entropy": 0.5157618727535009,
      "epoch": 2.9029505293241558,
      "grad_norm": 0.1645643264055252,
      "learning_rate": 3.2385684055570896e-06,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8378011956810951,
      "num_tokens": 1006994289.0,
      "step": 31552
    },
    {
      "entropy": 0.5140593273099512,
      "epoch": 2.903042535695597,
      "grad_norm": 0.15574456751346588,
      "learning_rate": 3.2355015794154634e-06,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8430406861007214,
      "num_tokens": 1007025705.0,
      "step": 31553
    },
    {
      "entropy": 0.44442208111286163,
      "epoch": 2.903134542067038,
      "grad_norm": 0.15221352875232697,
      "learning_rate": 3.232434753273837e-06,
      "loss": 0.4339,
      "mean_token_accuracy": 0.8613984175026417,
      "num_tokens": 1007057996.0,
      "step": 31554
    },
    {
      "entropy": 0.5976265127537772,
      "epoch": 2.9032265484384796,
      "grad_norm": 0.16906780004501343,
      "learning_rate": 3.229367927132211e-06,
      "loss": 0.5859,
      "mean_token_accuracy": 0.8168173618614674,
      "num_tokens": 1007090113.0,
      "step": 31555
    },
    {
      "entropy": 0.42008482583332807,
      "epoch": 2.9033185548099207,
      "grad_norm": 0.15009652078151703,
      "learning_rate": 3.226301100990585e-06,
      "loss": 0.4087,
      "mean_token_accuracy": 0.8696645349264145,
      "num_tokens": 1007122075.0,
      "step": 31556
    },
    {
      "entropy": 0.4837853229837492,
      "epoch": 2.903410561181362,
      "grad_norm": 0.1521686613559723,
      "learning_rate": 3.223234274848959e-06,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8499681241810322,
      "num_tokens": 1007154085.0,
      "step": 31557
    },
    {
      "entropy": 0.5702483151108027,
      "epoch": 2.903502567552803,
      "grad_norm": 0.1626317948102951,
      "learning_rate": 3.220167448707333e-06,
      "loss": 0.5547,
      "mean_token_accuracy": 0.824523564428091,
      "num_tokens": 1007186492.0,
      "step": 31558
    },
    {
      "entropy": 0.5154353734105825,
      "epoch": 2.903594573924244,
      "grad_norm": 0.15318149328231812,
      "learning_rate": 3.2171006225657066e-06,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8394853994250298,
      "num_tokens": 1007218819.0,
      "step": 31559
    },
    {
      "entropy": 0.5383674055337906,
      "epoch": 2.9036865802956857,
      "grad_norm": 0.16433897614479065,
      "learning_rate": 3.2140337964240813e-06,
      "loss": 0.5302,
      "mean_token_accuracy": 0.8298264853656292,
      "num_tokens": 1007250460.0,
      "step": 31560
    },
    {
      "entropy": 0.5043973792344332,
      "epoch": 2.903778586667127,
      "grad_norm": 0.16186104714870453,
      "learning_rate": 3.210966970282455e-06,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8456127643585205,
      "num_tokens": 1007281344.0,
      "step": 31561
    },
    {
      "entropy": 0.5687700640410185,
      "epoch": 2.903870593038568,
      "grad_norm": 0.16750314831733704,
      "learning_rate": 3.207900144140829e-06,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8209189362823963,
      "num_tokens": 1007312178.0,
      "step": 31562
    },
    {
      "entropy": 0.4772051442414522,
      "epoch": 2.903962599410009,
      "grad_norm": 0.15277297794818878,
      "learning_rate": 3.2048333179992027e-06,
      "loss": 0.4686,
      "mean_token_accuracy": 0.852705430239439,
      "num_tokens": 1007344411.0,
      "step": 31563
    },
    {
      "entropy": 0.5629870644770563,
      "epoch": 2.9040546057814502,
      "grad_norm": 0.15945376455783844,
      "learning_rate": 3.2017664918575765e-06,
      "loss": 0.5502,
      "mean_token_accuracy": 0.8277216777205467,
      "num_tokens": 1007376551.0,
      "step": 31564
    },
    {
      "entropy": 0.5501813343726099,
      "epoch": 2.904146612152892,
      "grad_norm": 0.16777487099170685,
      "learning_rate": 3.1986996657159507e-06,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8289535008370876,
      "num_tokens": 1007408468.0,
      "step": 31565
    },
    {
      "entropy": 0.42640900146216154,
      "epoch": 2.904238618524333,
      "grad_norm": 0.14660978317260742,
      "learning_rate": 3.1956328395743245e-06,
      "loss": 0.4059,
      "mean_token_accuracy": 0.8707652613520622,
      "num_tokens": 1007440400.0,
      "step": 31566
    },
    {
      "entropy": 0.426222356967628,
      "epoch": 2.904330624895774,
      "grad_norm": 0.14701193571090698,
      "learning_rate": 3.1925660134326983e-06,
      "loss": 0.4144,
      "mean_token_accuracy": 0.8651508241891861,
      "num_tokens": 1007472256.0,
      "step": 31567
    },
    {
      "entropy": 0.5582355819642544,
      "epoch": 2.9044226312672152,
      "grad_norm": 0.16503295302391052,
      "learning_rate": 3.189499187291073e-06,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8249890953302383,
      "num_tokens": 1007504310.0,
      "step": 31568
    },
    {
      "entropy": 0.485361160710454,
      "epoch": 2.9045146376386564,
      "grad_norm": 0.15477268397808075,
      "learning_rate": 3.1864323611494468e-06,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8486686050891876,
      "num_tokens": 1007536309.0,
      "step": 31569
    },
    {
      "entropy": 0.5523416865617037,
      "epoch": 2.904606644010098,
      "grad_norm": 0.16727662086486816,
      "learning_rate": 3.1833655350078206e-06,
      "loss": 0.5472,
      "mean_token_accuracy": 0.8275586292147636,
      "num_tokens": 1007568573.0,
      "step": 31570
    },
    {
      "entropy": 0.5313183572143316,
      "epoch": 2.904698650381539,
      "grad_norm": 0.15975676476955414,
      "learning_rate": 3.1802987088661944e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.8332741968333721,
      "num_tokens": 1007600118.0,
      "step": 31571
    },
    {
      "entropy": 0.5841391105204821,
      "epoch": 2.90479065675298,
      "grad_norm": 0.16703222692012787,
      "learning_rate": 3.1772318827245686e-06,
      "loss": 0.5634,
      "mean_token_accuracy": 0.8201775662600994,
      "num_tokens": 1007632088.0,
      "step": 31572
    },
    {
      "entropy": 0.6021374445408583,
      "epoch": 2.9048826631244213,
      "grad_norm": 0.16688482463359833,
      "learning_rate": 3.1741650565829424e-06,
      "loss": 0.5897,
      "mean_token_accuracy": 0.8147394917905331,
      "num_tokens": 1007664319.0,
      "step": 31573
    },
    {
      "entropy": 0.5454305512830615,
      "epoch": 2.9049746694958625,
      "grad_norm": 0.16014398634433746,
      "learning_rate": 3.171098230441316e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8307584263384342,
      "num_tokens": 1007697087.0,
      "step": 31574
    },
    {
      "entropy": 0.4904369255527854,
      "epoch": 2.905066675867304,
      "grad_norm": 0.15288996696472168,
      "learning_rate": 3.168031404299691e-06,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8490511178970337,
      "num_tokens": 1007729533.0,
      "step": 31575
    },
    {
      "entropy": 0.5232648253440857,
      "epoch": 2.905158682238745,
      "grad_norm": 0.15664124488830566,
      "learning_rate": 3.1649645781580646e-06,
      "loss": 0.517,
      "mean_token_accuracy": 0.8378329128026962,
      "num_tokens": 1007761759.0,
      "step": 31576
    },
    {
      "entropy": 0.4304900458082557,
      "epoch": 2.9052506886101863,
      "grad_norm": 0.143419548869133,
      "learning_rate": 3.1618977520164384e-06,
      "loss": 0.4272,
      "mean_token_accuracy": 0.8625184558331966,
      "num_tokens": 1007794374.0,
      "step": 31577
    },
    {
      "entropy": 0.5753717888146639,
      "epoch": 2.9053426949816274,
      "grad_norm": 0.16931338608264923,
      "learning_rate": 3.1588309258748122e-06,
      "loss": 0.5635,
      "mean_token_accuracy": 0.8203911669552326,
      "num_tokens": 1007826484.0,
      "step": 31578
    },
    {
      "entropy": 0.4707006346434355,
      "epoch": 2.9054347013530686,
      "grad_norm": 0.15341539680957794,
      "learning_rate": 3.155764099733186e-06,
      "loss": 0.4599,
      "mean_token_accuracy": 0.8533500507473946,
      "num_tokens": 1007858436.0,
      "step": 31579
    },
    {
      "entropy": 0.48521264689043164,
      "epoch": 2.90552670772451,
      "grad_norm": 0.15396103262901306,
      "learning_rate": 3.1526972735915603e-06,
      "loss": 0.4725,
      "mean_token_accuracy": 0.8531202338635921,
      "num_tokens": 1007889677.0,
      "step": 31580
    },
    {
      "entropy": 0.5183248305693269,
      "epoch": 2.9056187140959513,
      "grad_norm": 0.1544097512960434,
      "learning_rate": 3.149630447449934e-06,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8386948518455029,
      "num_tokens": 1007922206.0,
      "step": 31581
    },
    {
      "entropy": 0.5512849506922066,
      "epoch": 2.9057107204673924,
      "grad_norm": 0.16731055080890656,
      "learning_rate": 3.146563621308308e-06,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8321472108364105,
      "num_tokens": 1007953375.0,
      "step": 31582
    },
    {
      "entropy": 0.48357674677390605,
      "epoch": 2.9058027268388336,
      "grad_norm": 0.15476591885089874,
      "learning_rate": 3.1434967951666825e-06,
      "loss": 0.462,
      "mean_token_accuracy": 0.854445319622755,
      "num_tokens": 1007984286.0,
      "step": 31583
    },
    {
      "entropy": 0.44135486718732864,
      "epoch": 2.9058947332102747,
      "grad_norm": 0.14640596508979797,
      "learning_rate": 3.1404299690250563e-06,
      "loss": 0.4345,
      "mean_token_accuracy": 0.8573675341904163,
      "num_tokens": 1008016445.0,
      "step": 31584
    },
    {
      "entropy": 0.44826501328498125,
      "epoch": 2.9059867395817163,
      "grad_norm": 0.14845646917819977,
      "learning_rate": 3.13736314288343e-06,
      "loss": 0.4288,
      "mean_token_accuracy": 0.8602013476192951,
      "num_tokens": 1008048478.0,
      "step": 31585
    },
    {
      "entropy": 0.5641799729783088,
      "epoch": 2.9060787459531574,
      "grad_norm": 0.1607801616191864,
      "learning_rate": 3.134296316741804e-06,
      "loss": 0.5546,
      "mean_token_accuracy": 0.826539658010006,
      "num_tokens": 1008080284.0,
      "step": 31586
    },
    {
      "entropy": 0.38239237666130066,
      "epoch": 2.9061707523245985,
      "grad_norm": 0.14188101887702942,
      "learning_rate": 3.131229490600178e-06,
      "loss": 0.3774,
      "mean_token_accuracy": 0.8783796317875385,
      "num_tokens": 1008112622.0,
      "step": 31587
    },
    {
      "entropy": 0.45367622282356024,
      "epoch": 2.9062627586960397,
      "grad_norm": 0.15234455466270447,
      "learning_rate": 3.128162664458552e-06,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8561719246208668,
      "num_tokens": 1008144393.0,
      "step": 31588
    },
    {
      "entropy": 0.5354903251864016,
      "epoch": 2.906354765067481,
      "grad_norm": 0.16315287351608276,
      "learning_rate": 3.1250958383169257e-06,
      "loss": 0.5329,
      "mean_token_accuracy": 0.8304736651480198,
      "num_tokens": 1008176406.0,
      "step": 31589
    },
    {
      "entropy": 0.523020414635539,
      "epoch": 2.9064467714389224,
      "grad_norm": 0.14952205121517181,
      "learning_rate": 3.1220290121753e-06,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8380216136574745,
      "num_tokens": 1008208994.0,
      "step": 31590
    },
    {
      "entropy": 0.4147814222378656,
      "epoch": 2.9065387778103635,
      "grad_norm": 0.14092965424060822,
      "learning_rate": 3.1189621860336738e-06,
      "loss": 0.4012,
      "mean_token_accuracy": 0.8715152516961098,
      "num_tokens": 1008240996.0,
      "step": 31591
    },
    {
      "entropy": 0.5402420293539762,
      "epoch": 2.9066307841818046,
      "grad_norm": 0.1651100218296051,
      "learning_rate": 3.115895359892048e-06,
      "loss": 0.5275,
      "mean_token_accuracy": 0.8329867795109749,
      "num_tokens": 1008272053.0,
      "step": 31592
    },
    {
      "entropy": 0.467072163708508,
      "epoch": 2.906722790553246,
      "grad_norm": 0.15561813116073608,
      "learning_rate": 3.1128285337504218e-06,
      "loss": 0.4606,
      "mean_token_accuracy": 0.8503327332437038,
      "num_tokens": 1008303540.0,
      "step": 31593
    },
    {
      "entropy": 0.5088983926689252,
      "epoch": 2.906814796924687,
      "grad_norm": 0.1508224904537201,
      "learning_rate": 3.1097617076087956e-06,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8431078344583511,
      "num_tokens": 1008335539.0,
      "step": 31594
    },
    {
      "entropy": 0.5376829076558352,
      "epoch": 2.9069068032961285,
      "grad_norm": 0.1598944216966629,
      "learning_rate": 3.10669488146717e-06,
      "loss": 0.5393,
      "mean_token_accuracy": 0.8318348787724972,
      "num_tokens": 1008367649.0,
      "step": 31595
    },
    {
      "entropy": 0.5486671943217516,
      "epoch": 2.9069988096675696,
      "grad_norm": 0.1632557213306427,
      "learning_rate": 3.103628055325544e-06,
      "loss": 0.5438,
      "mean_token_accuracy": 0.8290096558630466,
      "num_tokens": 1008400096.0,
      "step": 31596
    },
    {
      "entropy": 0.5329328905791044,
      "epoch": 2.9070908160390108,
      "grad_norm": 0.1665772944688797,
      "learning_rate": 3.100561229183918e-06,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8359791226685047,
      "num_tokens": 1008431966.0,
      "step": 31597
    },
    {
      "entropy": 0.5627332236617804,
      "epoch": 2.907182822410452,
      "grad_norm": 0.16651403903961182,
      "learning_rate": 3.0974944030422916e-06,
      "loss": 0.5449,
      "mean_token_accuracy": 0.8278806582093239,
      "num_tokens": 1008463366.0,
      "step": 31598
    },
    {
      "entropy": 0.4963144166395068,
      "epoch": 2.907274828781893,
      "grad_norm": 0.15680305659770966,
      "learning_rate": 3.0944275769006654e-06,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8424457423388958,
      "num_tokens": 1008495824.0,
      "step": 31599
    },
    {
      "entropy": 0.49684586841613054,
      "epoch": 2.9073668351533346,
      "grad_norm": 0.15986284613609314,
      "learning_rate": 3.0913607507590397e-06,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8449944332242012,
      "num_tokens": 1008526688.0,
      "step": 31600
    },
    {
      "entropy": 0.4465630556223914,
      "epoch": 2.9074588415247757,
      "grad_norm": 0.15352092683315277,
      "learning_rate": 3.0882939246174135e-06,
      "loss": 0.4371,
      "mean_token_accuracy": 0.859958179295063,
      "num_tokens": 1008558371.0,
      "step": 31601
    },
    {
      "entropy": 0.5683343876153231,
      "epoch": 2.907550847896217,
      "grad_norm": 0.16791731119155884,
      "learning_rate": 3.0852270984757877e-06,
      "loss": 0.5479,
      "mean_token_accuracy": 0.8253293819725513,
      "num_tokens": 1008590056.0,
      "step": 31602
    },
    {
      "entropy": 0.47182382084429264,
      "epoch": 2.907642854267658,
      "grad_norm": 0.1548403948545456,
      "learning_rate": 3.0821602723341615e-06,
      "loss": 0.4582,
      "mean_token_accuracy": 0.8579439520835876,
      "num_tokens": 1008621538.0,
      "step": 31603
    },
    {
      "entropy": 0.5265463758260012,
      "epoch": 2.907734860639099,
      "grad_norm": 0.16188135743141174,
      "learning_rate": 3.0790934461925357e-06,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8370391577482224,
      "num_tokens": 1008653089.0,
      "step": 31604
    },
    {
      "entropy": 0.50462656468153,
      "epoch": 2.9078268670105407,
      "grad_norm": 0.1613793820142746,
      "learning_rate": 3.0760266200509095e-06,
      "loss": 0.4977,
      "mean_token_accuracy": 0.8408206887543201,
      "num_tokens": 1008685235.0,
      "step": 31605
    },
    {
      "entropy": 0.5778474592370912,
      "epoch": 2.907918873381982,
      "grad_norm": 0.1656084656715393,
      "learning_rate": 3.0729597939092833e-06,
      "loss": 0.5714,
      "mean_token_accuracy": 0.8184494152665138,
      "num_tokens": 1008716820.0,
      "step": 31606
    },
    {
      "entropy": 0.5489211669191718,
      "epoch": 2.908010879753423,
      "grad_norm": 0.15710194408893585,
      "learning_rate": 3.069892967767657e-06,
      "loss": 0.5348,
      "mean_token_accuracy": 0.8354113735258579,
      "num_tokens": 1008748529.0,
      "step": 31607
    },
    {
      "entropy": 0.5889763366430998,
      "epoch": 2.908102886124864,
      "grad_norm": 0.16712027788162231,
      "learning_rate": 3.0668261416260313e-06,
      "loss": 0.576,
      "mean_token_accuracy": 0.8192365430295467,
      "num_tokens": 1008780638.0,
      "step": 31608
    },
    {
      "entropy": 0.5734177827835083,
      "epoch": 2.9081948924963053,
      "grad_norm": 0.16682463884353638,
      "learning_rate": 3.063759315484405e-06,
      "loss": 0.5614,
      "mean_token_accuracy": 0.8195853754878044,
      "num_tokens": 1008812375.0,
      "step": 31609
    },
    {
      "entropy": 0.4877708191052079,
      "epoch": 2.908286898867747,
      "grad_norm": 0.15251833200454712,
      "learning_rate": 3.0606924893427794e-06,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8501756452023983,
      "num_tokens": 1008844696.0,
      "step": 31610
    },
    {
      "entropy": 0.5730983931571245,
      "epoch": 2.908378905239188,
      "grad_norm": 0.16584867238998413,
      "learning_rate": 3.0576256632011536e-06,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8238728120923042,
      "num_tokens": 1008876030.0,
      "step": 31611
    },
    {
      "entropy": 0.5097004637354985,
      "epoch": 2.908470911610629,
      "grad_norm": 0.1519850343465805,
      "learning_rate": 3.0545588370595274e-06,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8448539115488529,
      "num_tokens": 1008908652.0,
      "step": 31612
    },
    {
      "entropy": 0.5208407770842314,
      "epoch": 2.9085629179820702,
      "grad_norm": 0.15480978786945343,
      "learning_rate": 3.051492010917901e-06,
      "loss": 0.5068,
      "mean_token_accuracy": 0.8413561061024666,
      "num_tokens": 1008940765.0,
      "step": 31613
    },
    {
      "entropy": 0.4764237739145756,
      "epoch": 2.9086549243535114,
      "grad_norm": 0.15402160584926605,
      "learning_rate": 3.048425184776275e-06,
      "loss": 0.4638,
      "mean_token_accuracy": 0.851925864815712,
      "num_tokens": 1008971962.0,
      "step": 31614
    },
    {
      "entropy": 0.474448140244931,
      "epoch": 2.908746930724953,
      "grad_norm": 0.15485703945159912,
      "learning_rate": 3.045358358634649e-06,
      "loss": 0.4722,
      "mean_token_accuracy": 0.8528354056179523,
      "num_tokens": 1009003957.0,
      "step": 31615
    },
    {
      "entropy": 0.5162083068862557,
      "epoch": 2.908838937096394,
      "grad_norm": 0.15733757615089417,
      "learning_rate": 3.042291532493023e-06,
      "loss": 0.5048,
      "mean_token_accuracy": 0.8399391658604145,
      "num_tokens": 1009036581.0,
      "step": 31616
    },
    {
      "entropy": 0.6146967275999486,
      "epoch": 2.908930943467835,
      "grad_norm": 0.174052432179451,
      "learning_rate": 3.0392247063513972e-06,
      "loss": 0.6107,
      "mean_token_accuracy": 0.8150640837848186,
      "num_tokens": 1009067504.0,
      "step": 31617
    },
    {
      "entropy": 0.5351252045948058,
      "epoch": 2.9090229498392763,
      "grad_norm": 0.1624765694141388,
      "learning_rate": 3.036157880209771e-06,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8374026976525784,
      "num_tokens": 1009099049.0,
      "step": 31618
    },
    {
      "entropy": 0.5386321740224957,
      "epoch": 2.9091149562107175,
      "grad_norm": 0.16007831692695618,
      "learning_rate": 3.0330910540681453e-06,
      "loss": 0.5181,
      "mean_token_accuracy": 0.8370063751935959,
      "num_tokens": 1009130916.0,
      "step": 31619
    },
    {
      "entropy": 0.5071009285748005,
      "epoch": 2.909206962582159,
      "grad_norm": 0.1553984433412552,
      "learning_rate": 3.030024227926519e-06,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8436948098242283,
      "num_tokens": 1009163342.0,
      "step": 31620
    },
    {
      "entropy": 0.5736031970009208,
      "epoch": 2.9092989689536,
      "grad_norm": 0.16804218292236328,
      "learning_rate": 3.026957401784893e-06,
      "loss": 0.5484,
      "mean_token_accuracy": 0.8233486972749233,
      "num_tokens": 1009195056.0,
      "step": 31621
    },
    {
      "entropy": 0.49400392500683665,
      "epoch": 2.9093909753250413,
      "grad_norm": 0.15106551349163055,
      "learning_rate": 3.0238905756432667e-06,
      "loss": 0.4827,
      "mean_token_accuracy": 0.8447491750121117,
      "num_tokens": 1009226796.0,
      "step": 31622
    },
    {
      "entropy": 0.5841475054621696,
      "epoch": 2.9094829816964825,
      "grad_norm": 0.1668873280286789,
      "learning_rate": 3.020823749501641e-06,
      "loss": 0.5775,
      "mean_token_accuracy": 0.8199221938848495,
      "num_tokens": 1009259029.0,
      "step": 31623
    },
    {
      "entropy": 0.45760053861886263,
      "epoch": 2.9095749880679236,
      "grad_norm": 0.15325871109962463,
      "learning_rate": 3.017756923360015e-06,
      "loss": 0.4452,
      "mean_token_accuracy": 0.8581708446145058,
      "num_tokens": 1009291372.0,
      "step": 31624
    },
    {
      "entropy": 0.5139081878587604,
      "epoch": 2.909666994439365,
      "grad_norm": 0.1532834768295288,
      "learning_rate": 3.014690097218389e-06,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8392605371773243,
      "num_tokens": 1009323857.0,
      "step": 31625
    },
    {
      "entropy": 0.4610051040071994,
      "epoch": 2.9097590008108063,
      "grad_norm": 0.14767156541347504,
      "learning_rate": 3.0116232710767627e-06,
      "loss": 0.4436,
      "mean_token_accuracy": 0.8603173233568668,
      "num_tokens": 1009355377.0,
      "step": 31626
    },
    {
      "entropy": 0.4651310946792364,
      "epoch": 2.9098510071822474,
      "grad_norm": 0.14925280213356018,
      "learning_rate": 3.008556444935137e-06,
      "loss": 0.4434,
      "mean_token_accuracy": 0.8560331836342812,
      "num_tokens": 1009387501.0,
      "step": 31627
    },
    {
      "entropy": 0.5012597348541021,
      "epoch": 2.9099430135536886,
      "grad_norm": 0.15255172550678253,
      "learning_rate": 3.0054896187935107e-06,
      "loss": 0.4808,
      "mean_token_accuracy": 0.8470074981451035,
      "num_tokens": 1009419178.0,
      "step": 31628
    },
    {
      "entropy": 0.5060306899249554,
      "epoch": 2.9100350199251297,
      "grad_norm": 0.15836448967456818,
      "learning_rate": 3.0024227926518845e-06,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8448220826685429,
      "num_tokens": 1009449372.0,
      "step": 31629
    },
    {
      "entropy": 0.5651253085816279,
      "epoch": 2.9101270262965713,
      "grad_norm": 0.17296932637691498,
      "learning_rate": 2.9993559665102583e-06,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8253622390329838,
      "num_tokens": 1009480711.0,
      "step": 31630
    },
    {
      "entropy": 0.4658669624477625,
      "epoch": 2.9102190326680124,
      "grad_norm": 0.14268329739570618,
      "learning_rate": 2.9962891403686326e-06,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8566499724984169,
      "num_tokens": 1009513479.0,
      "step": 31631
    },
    {
      "entropy": 0.5145214814692736,
      "epoch": 2.9103110390394535,
      "grad_norm": 0.15145088732242584,
      "learning_rate": 2.9932223142270068e-06,
      "loss": 0.4994,
      "mean_token_accuracy": 0.8436353839933872,
      "num_tokens": 1009545699.0,
      "step": 31632
    },
    {
      "entropy": 0.597349239513278,
      "epoch": 2.9104030454108947,
      "grad_norm": 0.16990788280963898,
      "learning_rate": 2.9901554880853806e-06,
      "loss": 0.5872,
      "mean_token_accuracy": 0.8133483491837978,
      "num_tokens": 1009577332.0,
      "step": 31633
    },
    {
      "entropy": 0.42105412809178233,
      "epoch": 2.910495051782336,
      "grad_norm": 0.14672738313674927,
      "learning_rate": 2.987088661943755e-06,
      "loss": 0.4202,
      "mean_token_accuracy": 0.8680449463427067,
      "num_tokens": 1009609189.0,
      "step": 31634
    },
    {
      "entropy": 0.44890528451651335,
      "epoch": 2.9105870581537774,
      "grad_norm": 0.15114480257034302,
      "learning_rate": 2.9840218358021286e-06,
      "loss": 0.4385,
      "mean_token_accuracy": 0.8588906191289425,
      "num_tokens": 1009640403.0,
      "step": 31635
    },
    {
      "entropy": 0.54056769865565,
      "epoch": 2.9106790645252185,
      "grad_norm": 0.1671488732099533,
      "learning_rate": 2.9809550096605024e-06,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8316262289881706,
      "num_tokens": 1009671175.0,
      "step": 31636
    },
    {
      "entropy": 0.48831605073064566,
      "epoch": 2.9107710708966597,
      "grad_norm": 0.15273576974868774,
      "learning_rate": 2.977888183518876e-06,
      "loss": 0.4731,
      "mean_token_accuracy": 0.8499517627060413,
      "num_tokens": 1009703404.0,
      "step": 31637
    },
    {
      "entropy": 0.5255767293274403,
      "epoch": 2.910863077268101,
      "grad_norm": 0.16143545508384705,
      "learning_rate": 2.9748213573772504e-06,
      "loss": 0.5207,
      "mean_token_accuracy": 0.8391364812850952,
      "num_tokens": 1009735496.0,
      "step": 31638
    },
    {
      "entropy": 0.4433043079916388,
      "epoch": 2.910955083639542,
      "grad_norm": 0.14718285202980042,
      "learning_rate": 2.9717545312356247e-06,
      "loss": 0.428,
      "mean_token_accuracy": 0.8628178425133228,
      "num_tokens": 1009767302.0,
      "step": 31639
    },
    {
      "entropy": 0.48239945131354034,
      "epoch": 2.9110470900109835,
      "grad_norm": 0.15507185459136963,
      "learning_rate": 2.9686877050939985e-06,
      "loss": 0.4646,
      "mean_token_accuracy": 0.8497467152774334,
      "num_tokens": 1009797807.0,
      "step": 31640
    },
    {
      "entropy": 0.5009360015392303,
      "epoch": 2.9111390963824246,
      "grad_norm": 0.1556379348039627,
      "learning_rate": 2.9656208789523723e-06,
      "loss": 0.4908,
      "mean_token_accuracy": 0.8442758470773697,
      "num_tokens": 1009829309.0,
      "step": 31641
    },
    {
      "entropy": 0.5162770971655846,
      "epoch": 2.9112311027538658,
      "grad_norm": 0.1584232896566391,
      "learning_rate": 2.9625540528107465e-06,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8397867940366268,
      "num_tokens": 1009861638.0,
      "step": 31642
    },
    {
      "entropy": 0.4839450251311064,
      "epoch": 2.911323109125307,
      "grad_norm": 0.15905074775218964,
      "learning_rate": 2.9594872266691203e-06,
      "loss": 0.473,
      "mean_token_accuracy": 0.8480511084198952,
      "num_tokens": 1009894121.0,
      "step": 31643
    },
    {
      "entropy": 0.5286220572888851,
      "epoch": 2.911415115496748,
      "grad_norm": 0.15655481815338135,
      "learning_rate": 2.956420400527494e-06,
      "loss": 0.5124,
      "mean_token_accuracy": 0.8372854292392731,
      "num_tokens": 1009925597.0,
      "step": 31644
    },
    {
      "entropy": 0.5160848796367645,
      "epoch": 2.9115071218681896,
      "grad_norm": 0.1584775447845459,
      "learning_rate": 2.953353574385868e-06,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8367670737206936,
      "num_tokens": 1009958064.0,
      "step": 31645
    },
    {
      "entropy": 0.6534235496073961,
      "epoch": 2.9115991282396307,
      "grad_norm": 0.17550985515117645,
      "learning_rate": 2.950286748244242e-06,
      "loss": 0.64,
      "mean_token_accuracy": 0.7986322306096554,
      "num_tokens": 1009990485.0,
      "step": 31646
    },
    {
      "entropy": 0.5132122915238142,
      "epoch": 2.911691134611072,
      "grad_norm": 0.1604534387588501,
      "learning_rate": 2.9472199221026163e-06,
      "loss": 0.513,
      "mean_token_accuracy": 0.8361475095152855,
      "num_tokens": 1010022753.0,
      "step": 31647
    },
    {
      "entropy": 0.5409465436823666,
      "epoch": 2.911783140982513,
      "grad_norm": 0.15687209367752075,
      "learning_rate": 2.94415309596099e-06,
      "loss": 0.526,
      "mean_token_accuracy": 0.8322770409286022,
      "num_tokens": 1010055360.0,
      "step": 31648
    },
    {
      "entropy": 0.5523829339072108,
      "epoch": 2.911875147353954,
      "grad_norm": 0.16190841794013977,
      "learning_rate": 2.941086269819364e-06,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8304870240390301,
      "num_tokens": 1010087520.0,
      "step": 31649
    },
    {
      "entropy": 0.41511599090881646,
      "epoch": 2.9119671537253957,
      "grad_norm": 0.14611946046352386,
      "learning_rate": 2.938019443677738e-06,
      "loss": 0.4076,
      "mean_token_accuracy": 0.8696324005723,
      "num_tokens": 1010119107.0,
      "step": 31650
    },
    {
      "entropy": 0.49565735552459955,
      "epoch": 2.912059160096837,
      "grad_norm": 0.1546175181865692,
      "learning_rate": 2.934952617536112e-06,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8512743040919304,
      "num_tokens": 1010151012.0,
      "step": 31651
    },
    {
      "entropy": 0.5509389787912369,
      "epoch": 2.912151166468278,
      "grad_norm": 0.16450297832489014,
      "learning_rate": 2.9318857913944858e-06,
      "loss": 0.5312,
      "mean_token_accuracy": 0.832123164087534,
      "num_tokens": 1010182932.0,
      "step": 31652
    },
    {
      "entropy": 0.5375024452805519,
      "epoch": 2.912243172839719,
      "grad_norm": 0.16287247836589813,
      "learning_rate": 2.92881896525286e-06,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8361030966043472,
      "num_tokens": 1010214255.0,
      "step": 31653
    },
    {
      "entropy": 0.4236013904446736,
      "epoch": 2.9123351792111603,
      "grad_norm": 0.15103203058242798,
      "learning_rate": 2.925752139111234e-06,
      "loss": 0.425,
      "mean_token_accuracy": 0.8668359071016312,
      "num_tokens": 1010246360.0,
      "step": 31654
    },
    {
      "entropy": 0.45108175836503506,
      "epoch": 2.912427185582602,
      "grad_norm": 0.14598585665225983,
      "learning_rate": 2.922685312969608e-06,
      "loss": 0.436,
      "mean_token_accuracy": 0.8573505394160748,
      "num_tokens": 1010278690.0,
      "step": 31655
    },
    {
      "entropy": 0.4998539686203003,
      "epoch": 2.912519191954043,
      "grad_norm": 0.1612376570701599,
      "learning_rate": 2.919618486827982e-06,
      "loss": 0.4863,
      "mean_token_accuracy": 0.8435728549957275,
      "num_tokens": 1010310936.0,
      "step": 31656
    },
    {
      "entropy": 0.5131943188607693,
      "epoch": 2.912611198325484,
      "grad_norm": 0.15934942662715912,
      "learning_rate": 2.916551660686356e-06,
      "loss": 0.498,
      "mean_token_accuracy": 0.8429062999784946,
      "num_tokens": 1010343256.0,
      "step": 31657
    },
    {
      "entropy": 0.5321753183379769,
      "epoch": 2.9127032046969252,
      "grad_norm": 0.16112186014652252,
      "learning_rate": 2.91348483454473e-06,
      "loss": 0.5192,
      "mean_token_accuracy": 0.8335392735898495,
      "num_tokens": 1010375069.0,
      "step": 31658
    },
    {
      "entropy": 0.4927747920155525,
      "epoch": 2.9127952110683664,
      "grad_norm": 0.15587098896503448,
      "learning_rate": 2.9104180084031036e-06,
      "loss": 0.4793,
      "mean_token_accuracy": 0.8482653498649597,
      "num_tokens": 1010406957.0,
      "step": 31659
    },
    {
      "entropy": 0.5061252880841494,
      "epoch": 2.912887217439808,
      "grad_norm": 0.15851615369319916,
      "learning_rate": 2.9073511822614774e-06,
      "loss": 0.4971,
      "mean_token_accuracy": 0.8420426174998283,
      "num_tokens": 1010438593.0,
      "step": 31660
    },
    {
      "entropy": 0.4822808550670743,
      "epoch": 2.912979223811249,
      "grad_norm": 0.14951299130916595,
      "learning_rate": 2.9042843561198517e-06,
      "loss": 0.4641,
      "mean_token_accuracy": 0.8513177409768105,
      "num_tokens": 1010470791.0,
      "step": 31661
    },
    {
      "entropy": 0.5065961740911007,
      "epoch": 2.91307123018269,
      "grad_norm": 0.1535574495792389,
      "learning_rate": 2.901217529978226e-06,
      "loss": 0.483,
      "mean_token_accuracy": 0.84660205245018,
      "num_tokens": 1010503047.0,
      "step": 31662
    },
    {
      "entropy": 0.52178855240345,
      "epoch": 2.9131632365541313,
      "grad_norm": 0.16166672110557556,
      "learning_rate": 2.8981507038365997e-06,
      "loss": 0.5141,
      "mean_token_accuracy": 0.8348442874848843,
      "num_tokens": 1010534532.0,
      "step": 31663
    },
    {
      "entropy": 0.5033799186348915,
      "epoch": 2.9132552429255725,
      "grad_norm": 0.15941636264324188,
      "learning_rate": 2.8950838776949735e-06,
      "loss": 0.5016,
      "mean_token_accuracy": 0.8387733176350594,
      "num_tokens": 1010567204.0,
      "step": 31664
    },
    {
      "entropy": 0.6181393302977085,
      "epoch": 2.913347249297014,
      "grad_norm": 0.1771569401025772,
      "learning_rate": 2.8920170515533477e-06,
      "loss": 0.6033,
      "mean_token_accuracy": 0.8109242916107178,
      "num_tokens": 1010598945.0,
      "step": 31665
    },
    {
      "entropy": 0.41257854271680117,
      "epoch": 2.913439255668455,
      "grad_norm": 0.14270257949829102,
      "learning_rate": 2.8889502254117215e-06,
      "loss": 0.3988,
      "mean_token_accuracy": 0.8723224177956581,
      "num_tokens": 1010630978.0,
      "step": 31666
    },
    {
      "entropy": 0.5051873670890927,
      "epoch": 2.9135312620398963,
      "grad_norm": 0.1536223292350769,
      "learning_rate": 2.8858833992700953e-06,
      "loss": 0.4938,
      "mean_token_accuracy": 0.8453224152326584,
      "num_tokens": 1010663746.0,
      "step": 31667
    },
    {
      "entropy": 0.5252764374017715,
      "epoch": 2.9136232684113375,
      "grad_norm": 0.15574558079242706,
      "learning_rate": 2.8828165731284695e-06,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8394999839365482,
      "num_tokens": 1010696338.0,
      "step": 31668
    },
    {
      "entropy": 0.4795083049684763,
      "epoch": 2.9137152747827786,
      "grad_norm": 0.15101025998592377,
      "learning_rate": 2.8797497469868437e-06,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8502002954483032,
      "num_tokens": 1010728525.0,
      "step": 31669
    },
    {
      "entropy": 0.4987202435731888,
      "epoch": 2.91380728115422,
      "grad_norm": 0.15657947957515717,
      "learning_rate": 2.8766829208452175e-06,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8452239595353603,
      "num_tokens": 1010760052.0,
      "step": 31670
    },
    {
      "entropy": 0.5420360490679741,
      "epoch": 2.9138992875256613,
      "grad_norm": 0.16103225946426392,
      "learning_rate": 2.8736160947035914e-06,
      "loss": 0.5287,
      "mean_token_accuracy": 0.830456081777811,
      "num_tokens": 1010792753.0,
      "step": 31671
    },
    {
      "entropy": 0.5478196805343032,
      "epoch": 2.9139912938971024,
      "grad_norm": 0.16446156799793243,
      "learning_rate": 2.8705492685619656e-06,
      "loss": 0.5268,
      "mean_token_accuracy": 0.8366538919508457,
      "num_tokens": 1010824423.0,
      "step": 31672
    },
    {
      "entropy": 0.5007362198084593,
      "epoch": 2.9140833002685436,
      "grad_norm": 0.16428814828395844,
      "learning_rate": 2.8674824424203394e-06,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8386535234749317,
      "num_tokens": 1010855863.0,
      "step": 31673
    },
    {
      "entropy": 0.4766214694827795,
      "epoch": 2.9141753066399847,
      "grad_norm": 0.1498776227235794,
      "learning_rate": 2.864415616278713e-06,
      "loss": 0.466,
      "mean_token_accuracy": 0.8534175343811512,
      "num_tokens": 1010888142.0,
      "step": 31674
    },
    {
      "entropy": 0.4307875777594745,
      "epoch": 2.9142673130114263,
      "grad_norm": 0.14174772799015045,
      "learning_rate": 2.861348790137087e-06,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8691376373171806,
      "num_tokens": 1010920523.0,
      "step": 31675
    },
    {
      "entropy": 0.5977502670139074,
      "epoch": 2.9143593193828674,
      "grad_norm": 0.16539889574050903,
      "learning_rate": 2.858281963995461e-06,
      "loss": 0.576,
      "mean_token_accuracy": 0.8184392563998699,
      "num_tokens": 1010952670.0,
      "step": 31676
    },
    {
      "entropy": 0.4964477811008692,
      "epoch": 2.9144513257543085,
      "grad_norm": 0.15655851364135742,
      "learning_rate": 2.8552151378538354e-06,
      "loss": 0.486,
      "mean_token_accuracy": 0.8424384295940399,
      "num_tokens": 1010984902.0,
      "step": 31677
    },
    {
      "entropy": 0.5024928720667958,
      "epoch": 2.9145433321257497,
      "grad_norm": 0.161249577999115,
      "learning_rate": 2.8521483117122092e-06,
      "loss": 0.497,
      "mean_token_accuracy": 0.8391375690698624,
      "num_tokens": 1011016421.0,
      "step": 31678
    },
    {
      "entropy": 0.48950728750787675,
      "epoch": 2.914635338497191,
      "grad_norm": 0.14951542019844055,
      "learning_rate": 2.849081485570583e-06,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8512056581676006,
      "num_tokens": 1011048763.0,
      "step": 31679
    },
    {
      "entropy": 0.6742918267846107,
      "epoch": 2.9147273448686324,
      "grad_norm": 0.1776774376630783,
      "learning_rate": 2.8460146594289572e-06,
      "loss": 0.6599,
      "mean_token_accuracy": 0.7956428043544292,
      "num_tokens": 1011081286.0,
      "step": 31680
    },
    {
      "entropy": 0.4860797682777047,
      "epoch": 2.9148193512400735,
      "grad_norm": 0.15313833951950073,
      "learning_rate": 2.842947833287331e-06,
      "loss": 0.4818,
      "mean_token_accuracy": 0.8454750776290894,
      "num_tokens": 1011114054.0,
      "step": 31681
    },
    {
      "entropy": 0.43394132889807224,
      "epoch": 2.9149113576115147,
      "grad_norm": 0.14794811606407166,
      "learning_rate": 2.839881007145705e-06,
      "loss": 0.4193,
      "mean_token_accuracy": 0.8624856323003769,
      "num_tokens": 1011145834.0,
      "step": 31682
    },
    {
      "entropy": 0.5178061127662659,
      "epoch": 2.915003363982956,
      "grad_norm": 0.16045522689819336,
      "learning_rate": 2.836814181004079e-06,
      "loss": 0.5101,
      "mean_token_accuracy": 0.8405504040420055,
      "num_tokens": 1011177790.0,
      "step": 31683
    },
    {
      "entropy": 0.565799793228507,
      "epoch": 2.915095370354397,
      "grad_norm": 0.1662212610244751,
      "learning_rate": 2.8337473548624533e-06,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8223086856305599,
      "num_tokens": 1011209842.0,
      "step": 31684
    },
    {
      "entropy": 0.4792796600377187,
      "epoch": 2.9151873767258385,
      "grad_norm": 0.15776340663433075,
      "learning_rate": 2.830680528720827e-06,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8477111756801605,
      "num_tokens": 1011241947.0,
      "step": 31685
    },
    {
      "entropy": 0.49656126741319895,
      "epoch": 2.9152793830972796,
      "grad_norm": 0.1588110774755478,
      "learning_rate": 2.827613702579201e-06,
      "loss": 0.4877,
      "mean_token_accuracy": 0.8425460010766983,
      "num_tokens": 1011273465.0,
      "step": 31686
    },
    {
      "entropy": 0.5098253623582423,
      "epoch": 2.9153713894687208,
      "grad_norm": 0.16245092451572418,
      "learning_rate": 2.8245468764375747e-06,
      "loss": 0.4882,
      "mean_token_accuracy": 0.842391312122345,
      "num_tokens": 1011304409.0,
      "step": 31687
    },
    {
      "entropy": 0.49966050893999636,
      "epoch": 2.915463395840162,
      "grad_norm": 0.15349076688289642,
      "learning_rate": 2.821480050295949e-06,
      "loss": 0.4789,
      "mean_token_accuracy": 0.8466973677277565,
      "num_tokens": 1011336336.0,
      "step": 31688
    },
    {
      "entropy": 0.4444881835952401,
      "epoch": 2.915555402211603,
      "grad_norm": 0.15310357511043549,
      "learning_rate": 2.8184132241543227e-06,
      "loss": 0.432,
      "mean_token_accuracy": 0.8578585162758827,
      "num_tokens": 1011368492.0,
      "step": 31689
    },
    {
      "entropy": 0.4998810840770602,
      "epoch": 2.9156474085830446,
      "grad_norm": 0.153120756149292,
      "learning_rate": 2.8153463980126965e-06,
      "loss": 0.486,
      "mean_token_accuracy": 0.8432980589568615,
      "num_tokens": 1011401203.0,
      "step": 31690
    },
    {
      "entropy": 0.5195723855867982,
      "epoch": 2.9157394149544857,
      "grad_norm": 0.16672664880752563,
      "learning_rate": 2.8122795718710707e-06,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8379898928105831,
      "num_tokens": 1011432849.0,
      "step": 31691
    },
    {
      "entropy": 0.4694164600223303,
      "epoch": 2.915831421325927,
      "grad_norm": 0.1520937830209732,
      "learning_rate": 2.809212745729445e-06,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8534672595560551,
      "num_tokens": 1011465228.0,
      "step": 31692
    },
    {
      "entropy": 0.4728469140827656,
      "epoch": 2.915923427697368,
      "grad_norm": 0.14774729311466217,
      "learning_rate": 2.8061459195878188e-06,
      "loss": 0.459,
      "mean_token_accuracy": 0.8531510643661022,
      "num_tokens": 1011497536.0,
      "step": 31693
    },
    {
      "entropy": 0.4884731210768223,
      "epoch": 2.916015434068809,
      "grad_norm": 0.15448133647441864,
      "learning_rate": 2.8030790934461926e-06,
      "loss": 0.4789,
      "mean_token_accuracy": 0.8463282585144043,
      "num_tokens": 1011529536.0,
      "step": 31694
    },
    {
      "entropy": 0.5279954876750708,
      "epoch": 2.9161074404402507,
      "grad_norm": 0.15788273513317108,
      "learning_rate": 2.800012267304567e-06,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8332233093678951,
      "num_tokens": 1011561428.0,
      "step": 31695
    },
    {
      "entropy": 0.5783958751708269,
      "epoch": 2.916199446811692,
      "grad_norm": 0.1663161963224411,
      "learning_rate": 2.7969454411629406e-06,
      "loss": 0.5704,
      "mean_token_accuracy": 0.8201868906617165,
      "num_tokens": 1011593783.0,
      "step": 31696
    },
    {
      "entropy": 0.5686865085735917,
      "epoch": 2.916291453183133,
      "grad_norm": 0.1749512106180191,
      "learning_rate": 2.7938786150213144e-06,
      "loss": 0.5664,
      "mean_token_accuracy": 0.8209868296980858,
      "num_tokens": 1011624135.0,
      "step": 31697
    },
    {
      "entropy": 0.5910235308110714,
      "epoch": 2.916383459554574,
      "grad_norm": 0.1704583615064621,
      "learning_rate": 2.7908117888796886e-06,
      "loss": 0.5788,
      "mean_token_accuracy": 0.8196577839553356,
      "num_tokens": 1011655850.0,
      "step": 31698
    },
    {
      "entropy": 0.5581608153879642,
      "epoch": 2.9164754659260153,
      "grad_norm": 0.16311725974082947,
      "learning_rate": 2.787744962738063e-06,
      "loss": 0.5475,
      "mean_token_accuracy": 0.8263854719698429,
      "num_tokens": 1011687832.0,
      "step": 31699
    },
    {
      "entropy": 0.5480462908744812,
      "epoch": 2.916567472297457,
      "grad_norm": 0.16277042031288147,
      "learning_rate": 2.7846781365964366e-06,
      "loss": 0.5346,
      "mean_token_accuracy": 0.829914677888155,
      "num_tokens": 1011719673.0,
      "step": 31700
    },
    {
      "entropy": 0.47436994686722755,
      "epoch": 2.916659478668898,
      "grad_norm": 0.15067608654499054,
      "learning_rate": 2.7816113104548104e-06,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8542046882212162,
      "num_tokens": 1011752171.0,
      "step": 31701
    },
    {
      "entropy": 0.5510345678776503,
      "epoch": 2.916751485040339,
      "grad_norm": 0.16556870937347412,
      "learning_rate": 2.7785444843131842e-06,
      "loss": 0.554,
      "mean_token_accuracy": 0.8255844116210938,
      "num_tokens": 1011784349.0,
      "step": 31702
    },
    {
      "entropy": 0.5588245241669938,
      "epoch": 2.9168434914117802,
      "grad_norm": 0.16403868794441223,
      "learning_rate": 2.7754776581715585e-06,
      "loss": 0.5506,
      "mean_token_accuracy": 0.8235465101897717,
      "num_tokens": 1011815859.0,
      "step": 31703
    },
    {
      "entropy": 0.5596060156822205,
      "epoch": 2.9169354977832214,
      "grad_norm": 0.16525335609912872,
      "learning_rate": 2.7724108320299323e-06,
      "loss": 0.5511,
      "mean_token_accuracy": 0.826134692877531,
      "num_tokens": 1011847705.0,
      "step": 31704
    },
    {
      "entropy": 0.4598851501941681,
      "epoch": 2.917027504154663,
      "grad_norm": 0.15391482412815094,
      "learning_rate": 2.769344005888306e-06,
      "loss": 0.444,
      "mean_token_accuracy": 0.8560588248074055,
      "num_tokens": 1011879174.0,
      "step": 31705
    },
    {
      "entropy": 0.5082038203254342,
      "epoch": 2.917119510526104,
      "grad_norm": 0.16376687586307526,
      "learning_rate": 2.7662771797466803e-06,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8411647006869316,
      "num_tokens": 1011910749.0,
      "step": 31706
    },
    {
      "entropy": 0.4546252158470452,
      "epoch": 2.917211516897545,
      "grad_norm": 0.14843203127384186,
      "learning_rate": 2.7632103536050545e-06,
      "loss": 0.4408,
      "mean_token_accuracy": 0.8593180403113365,
      "num_tokens": 1011942708.0,
      "step": 31707
    },
    {
      "entropy": 0.5021251305006444,
      "epoch": 2.9173035232689863,
      "grad_norm": 0.15559092164039612,
      "learning_rate": 2.7601435274634283e-06,
      "loss": 0.4948,
      "mean_token_accuracy": 0.841538991779089,
      "num_tokens": 1011974680.0,
      "step": 31708
    },
    {
      "entropy": 0.5452813692390919,
      "epoch": 2.9173955296404275,
      "grad_norm": 0.16383862495422363,
      "learning_rate": 2.757076701321802e-06,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8311031237244606,
      "num_tokens": 1012006598.0,
      "step": 31709
    },
    {
      "entropy": 0.4255920685827732,
      "epoch": 2.917487536011869,
      "grad_norm": 0.15354737639427185,
      "learning_rate": 2.754009875180176e-06,
      "loss": 0.4166,
      "mean_token_accuracy": 0.8657374195754528,
      "num_tokens": 1012037816.0,
      "step": 31710
    },
    {
      "entropy": 0.4680136349052191,
      "epoch": 2.91757954238331,
      "grad_norm": 0.1565343588590622,
      "learning_rate": 2.75094304903855e-06,
      "loss": 0.4535,
      "mean_token_accuracy": 0.8556649498641491,
      "num_tokens": 1012069540.0,
      "step": 31711
    },
    {
      "entropy": 0.5196498169098049,
      "epoch": 2.9176715487547513,
      "grad_norm": 0.1571260243654251,
      "learning_rate": 2.747876222896924e-06,
      "loss": 0.4929,
      "mean_token_accuracy": 0.8431192450225353,
      "num_tokens": 1012101180.0,
      "step": 31712
    },
    {
      "entropy": 0.6111965454183519,
      "epoch": 2.9177635551261925,
      "grad_norm": 0.16694658994674683,
      "learning_rate": 2.744809396755298e-06,
      "loss": 0.5949,
      "mean_token_accuracy": 0.8137258514761925,
      "num_tokens": 1012133396.0,
      "step": 31713
    },
    {
      "entropy": 0.5639216550625861,
      "epoch": 2.9178555614976336,
      "grad_norm": 0.16423092782497406,
      "learning_rate": 2.7417425706136724e-06,
      "loss": 0.5495,
      "mean_token_accuracy": 0.82720036059618,
      "num_tokens": 1012165121.0,
      "step": 31714
    },
    {
      "entropy": 0.5340166375972331,
      "epoch": 2.917947567869075,
      "grad_norm": 0.15981276333332062,
      "learning_rate": 2.738675744472046e-06,
      "loss": 0.5301,
      "mean_token_accuracy": 0.8331610560417175,
      "num_tokens": 1012197444.0,
      "step": 31715
    },
    {
      "entropy": 0.5216633211821318,
      "epoch": 2.9180395742405163,
      "grad_norm": 0.1639036238193512,
      "learning_rate": 2.73560891833042e-06,
      "loss": 0.5099,
      "mean_token_accuracy": 0.8344588875770569,
      "num_tokens": 1012228779.0,
      "step": 31716
    },
    {
      "entropy": 0.5645678099244833,
      "epoch": 2.9181315806119574,
      "grad_norm": 0.16350406408309937,
      "learning_rate": 2.732542092188794e-06,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8233195096254349,
      "num_tokens": 1012261098.0,
      "step": 31717
    },
    {
      "entropy": 0.43952575605362654,
      "epoch": 2.9182235869833986,
      "grad_norm": 0.14854522049427032,
      "learning_rate": 2.729475266047168e-06,
      "loss": 0.4264,
      "mean_token_accuracy": 0.8633929155766964,
      "num_tokens": 1012293373.0,
      "step": 31718
    },
    {
      "entropy": 0.517444709315896,
      "epoch": 2.9183155933548397,
      "grad_norm": 0.1599203497171402,
      "learning_rate": 2.726408439905542e-06,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8397021777927876,
      "num_tokens": 1012324478.0,
      "step": 31719
    },
    {
      "entropy": 0.5666746674105525,
      "epoch": 2.9184075997262813,
      "grad_norm": 0.16495554149150848,
      "learning_rate": 2.7233416137639156e-06,
      "loss": 0.5611,
      "mean_token_accuracy": 0.823512826114893,
      "num_tokens": 1012356114.0,
      "step": 31720
    },
    {
      "entropy": 0.5228293491527438,
      "epoch": 2.9184996060977224,
      "grad_norm": 0.16306181252002716,
      "learning_rate": 2.72027478762229e-06,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8342958875000477,
      "num_tokens": 1012387693.0,
      "step": 31721
    },
    {
      "entropy": 0.5537099950015545,
      "epoch": 2.9185916124691635,
      "grad_norm": 0.16126035153865814,
      "learning_rate": 2.717207961480664e-06,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8287349790334702,
      "num_tokens": 1012420058.0,
      "step": 31722
    },
    {
      "entropy": 0.5478099884930998,
      "epoch": 2.9186836188406047,
      "grad_norm": 0.16461244225502014,
      "learning_rate": 2.714141135339038e-06,
      "loss": 0.5353,
      "mean_token_accuracy": 0.8318675570189953,
      "num_tokens": 1012452358.0,
      "step": 31723
    },
    {
      "entropy": 0.48316869465634227,
      "epoch": 2.918775625212046,
      "grad_norm": 0.15521939098834991,
      "learning_rate": 2.7110743091974117e-06,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8527202531695366,
      "num_tokens": 1012484165.0,
      "step": 31724
    },
    {
      "entropy": 0.46900575049221516,
      "epoch": 2.9188676315834874,
      "grad_norm": 0.15264862775802612,
      "learning_rate": 2.7080074830557855e-06,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8548115082085133,
      "num_tokens": 1012516273.0,
      "step": 31725
    },
    {
      "entropy": 0.5851594191044569,
      "epoch": 2.9189596379549285,
      "grad_norm": 0.16812758147716522,
      "learning_rate": 2.7049406569141597e-06,
      "loss": 0.5679,
      "mean_token_accuracy": 0.8210597150027752,
      "num_tokens": 1012547127.0,
      "step": 31726
    },
    {
      "entropy": 0.5330981398001313,
      "epoch": 2.9190516443263697,
      "grad_norm": 0.160051167011261,
      "learning_rate": 2.7018738307725335e-06,
      "loss": 0.518,
      "mean_token_accuracy": 0.8380129151046276,
      "num_tokens": 1012578984.0,
      "step": 31727
    },
    {
      "entropy": 0.4813415170647204,
      "epoch": 2.919143650697811,
      "grad_norm": 0.14841102063655853,
      "learning_rate": 2.6988070046309077e-06,
      "loss": 0.4764,
      "mean_token_accuracy": 0.8462304584681988,
      "num_tokens": 1012611341.0,
      "step": 31728
    },
    {
      "entropy": 0.4679763885214925,
      "epoch": 2.919235657069252,
      "grad_norm": 0.1500910222530365,
      "learning_rate": 2.6957401784892815e-06,
      "loss": 0.4556,
      "mean_token_accuracy": 0.8509758487343788,
      "num_tokens": 1012643866.0,
      "step": 31729
    },
    {
      "entropy": 0.5588549552485347,
      "epoch": 2.9193276634406935,
      "grad_norm": 0.16787482798099518,
      "learning_rate": 2.6926733523476557e-06,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8302885480225086,
      "num_tokens": 1012675391.0,
      "step": 31730
    },
    {
      "entropy": 0.4668615674600005,
      "epoch": 2.9194196698121346,
      "grad_norm": 0.14733274281024933,
      "learning_rate": 2.6896065262060295e-06,
      "loss": 0.4473,
      "mean_token_accuracy": 0.858049251139164,
      "num_tokens": 1012707827.0,
      "step": 31731
    },
    {
      "entropy": 0.5227158665657043,
      "epoch": 2.9195116761835758,
      "grad_norm": 0.1593148410320282,
      "learning_rate": 2.6865397000644033e-06,
      "loss": 0.5106,
      "mean_token_accuracy": 0.8368420042097569,
      "num_tokens": 1012739632.0,
      "step": 31732
    },
    {
      "entropy": 0.5090091931633651,
      "epoch": 2.919603682555017,
      "grad_norm": 0.1593145877122879,
      "learning_rate": 2.6834728739227776e-06,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8413753993809223,
      "num_tokens": 1012771210.0,
      "step": 31733
    },
    {
      "entropy": 0.5637792246416211,
      "epoch": 2.919695688926458,
      "grad_norm": 0.16283030807971954,
      "learning_rate": 2.6804060477811514e-06,
      "loss": 0.5503,
      "mean_token_accuracy": 0.825458500534296,
      "num_tokens": 1012802132.0,
      "step": 31734
    },
    {
      "entropy": 0.43047125078737736,
      "epoch": 2.9197876952978996,
      "grad_norm": 0.14735275506973267,
      "learning_rate": 2.677339221639525e-06,
      "loss": 0.4172,
      "mean_token_accuracy": 0.8686269000172615,
      "num_tokens": 1012834072.0,
      "step": 31735
    },
    {
      "entropy": 0.47237207437865436,
      "epoch": 2.9198797016693407,
      "grad_norm": 0.1524677723646164,
      "learning_rate": 2.6742723954978994e-06,
      "loss": 0.457,
      "mean_token_accuracy": 0.8523063696920872,
      "num_tokens": 1012866407.0,
      "step": 31736
    },
    {
      "entropy": 0.5321150254458189,
      "epoch": 2.919971708040782,
      "grad_norm": 0.1592411994934082,
      "learning_rate": 2.6712055693562736e-06,
      "loss": 0.5148,
      "mean_token_accuracy": 0.8357622511684895,
      "num_tokens": 1012898877.0,
      "step": 31737
    },
    {
      "entropy": 0.495721897110343,
      "epoch": 2.920063714412223,
      "grad_norm": 0.15538723766803741,
      "learning_rate": 2.6681387432146474e-06,
      "loss": 0.483,
      "mean_token_accuracy": 0.8448802568018436,
      "num_tokens": 1012931045.0,
      "step": 31738
    },
    {
      "entropy": 0.5231772544793785,
      "epoch": 2.920155720783664,
      "grad_norm": 0.16151833534240723,
      "learning_rate": 2.6650719170730212e-06,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8328370228409767,
      "num_tokens": 1012962839.0,
      "step": 31739
    },
    {
      "entropy": 0.447361066006124,
      "epoch": 2.9202477271551057,
      "grad_norm": 0.14922600984573364,
      "learning_rate": 2.662005090931395e-06,
      "loss": 0.4374,
      "mean_token_accuracy": 0.8579003065824509,
      "num_tokens": 1012995213.0,
      "step": 31740
    },
    {
      "entropy": 0.5960277887061238,
      "epoch": 2.920339733526547,
      "grad_norm": 0.16524837911128998,
      "learning_rate": 2.6589382647897692e-06,
      "loss": 0.5834,
      "mean_token_accuracy": 0.8176283687353134,
      "num_tokens": 1013027543.0,
      "step": 31741
    },
    {
      "entropy": 0.4791737413033843,
      "epoch": 2.920431739897988,
      "grad_norm": 0.15334735810756683,
      "learning_rate": 2.655871438648143e-06,
      "loss": 0.4665,
      "mean_token_accuracy": 0.8485603220760822,
      "num_tokens": 1013059359.0,
      "step": 31742
    },
    {
      "entropy": 0.4899951908737421,
      "epoch": 2.920523746269429,
      "grad_norm": 0.15800726413726807,
      "learning_rate": 2.6528046125065173e-06,
      "loss": 0.4882,
      "mean_token_accuracy": 0.8491741381585598,
      "num_tokens": 1013090997.0,
      "step": 31743
    },
    {
      "entropy": 0.543082382529974,
      "epoch": 2.9206157526408703,
      "grad_norm": 0.16081850230693817,
      "learning_rate": 2.649737786364891e-06,
      "loss": 0.5401,
      "mean_token_accuracy": 0.8263990841805935,
      "num_tokens": 1013123736.0,
      "step": 31744
    },
    {
      "entropy": 0.5319846458733082,
      "epoch": 2.920707759012312,
      "grad_norm": 0.16519789397716522,
      "learning_rate": 2.6466709602232653e-06,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8353897966444492,
      "num_tokens": 1013154733.0,
      "step": 31745
    },
    {
      "entropy": 0.5868540713563561,
      "epoch": 2.920799765383753,
      "grad_norm": 0.17321357131004333,
      "learning_rate": 2.643604134081639e-06,
      "loss": 0.5816,
      "mean_token_accuracy": 0.8198974616825581,
      "num_tokens": 1013186579.0,
      "step": 31746
    },
    {
      "entropy": 0.5349886901676655,
      "epoch": 2.920891771755194,
      "grad_norm": 0.1631455272436142,
      "learning_rate": 2.640537307940013e-06,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8333188220858574,
      "num_tokens": 1013219003.0,
      "step": 31747
    },
    {
      "entropy": 0.5630631241947412,
      "epoch": 2.9209837781266352,
      "grad_norm": 0.16576986014842987,
      "learning_rate": 2.6374704817983867e-06,
      "loss": 0.5382,
      "mean_token_accuracy": 0.8292986042797565,
      "num_tokens": 1013250740.0,
      "step": 31748
    },
    {
      "entropy": 0.460366805549711,
      "epoch": 2.9210757844980764,
      "grad_norm": 0.15208236873149872,
      "learning_rate": 2.634403655656761e-06,
      "loss": 0.4419,
      "mean_token_accuracy": 0.8564093001186848,
      "num_tokens": 1013282329.0,
      "step": 31749
    },
    {
      "entropy": 0.6067542713135481,
      "epoch": 2.921167790869518,
      "grad_norm": 0.17222054302692413,
      "learning_rate": 2.6313368295151347e-06,
      "loss": 0.5911,
      "mean_token_accuracy": 0.8121697790920734,
      "num_tokens": 1013313249.0,
      "step": 31750
    },
    {
      "entropy": 0.43379749567247927,
      "epoch": 2.921259797240959,
      "grad_norm": 0.14759962260723114,
      "learning_rate": 2.628270003373509e-06,
      "loss": 0.4171,
      "mean_token_accuracy": 0.8641929924488068,
      "num_tokens": 1013344753.0,
      "step": 31751
    },
    {
      "entropy": 0.5435611456632614,
      "epoch": 2.9213518036124,
      "grad_norm": 0.16250555217266083,
      "learning_rate": 2.6252031772318827e-06,
      "loss": 0.535,
      "mean_token_accuracy": 0.8276898339390755,
      "num_tokens": 1013376932.0,
      "step": 31752
    },
    {
      "entropy": 0.49098570458590984,
      "epoch": 2.9214438099838413,
      "grad_norm": 0.15270814299583435,
      "learning_rate": 2.622136351090257e-06,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8461875580251217,
      "num_tokens": 1013408873.0,
      "step": 31753
    },
    {
      "entropy": 0.46455681463703513,
      "epoch": 2.9215358163552825,
      "grad_norm": 0.15764877200126648,
      "learning_rate": 2.6190695249486308e-06,
      "loss": 0.4501,
      "mean_token_accuracy": 0.8556985445320606,
      "num_tokens": 1013439388.0,
      "step": 31754
    },
    {
      "entropy": 0.5712600275874138,
      "epoch": 2.921627822726724,
      "grad_norm": 0.1652008444070816,
      "learning_rate": 2.6160026988070046e-06,
      "loss": 0.5611,
      "mean_token_accuracy": 0.8227217458188534,
      "num_tokens": 1013471217.0,
      "step": 31755
    },
    {
      "entropy": 0.45485780760645866,
      "epoch": 2.921719829098165,
      "grad_norm": 0.150771826505661,
      "learning_rate": 2.612935872665379e-06,
      "loss": 0.4501,
      "mean_token_accuracy": 0.8554886877536774,
      "num_tokens": 1013503604.0,
      "step": 31756
    },
    {
      "entropy": 0.5274342028424144,
      "epoch": 2.9218118354696063,
      "grad_norm": 0.16336211562156677,
      "learning_rate": 2.6098690465237526e-06,
      "loss": 0.5096,
      "mean_token_accuracy": 0.8381139822304249,
      "num_tokens": 1013535575.0,
      "step": 31757
    },
    {
      "entropy": 0.49456009827554226,
      "epoch": 2.9219038418410475,
      "grad_norm": 0.15537448227405548,
      "learning_rate": 2.606802220382127e-06,
      "loss": 0.4835,
      "mean_token_accuracy": 0.8459806255996227,
      "num_tokens": 1013567995.0,
      "step": 31758
    },
    {
      "entropy": 0.5206192089244723,
      "epoch": 2.9219958482124886,
      "grad_norm": 0.16146180033683777,
      "learning_rate": 2.6037353942405006e-06,
      "loss": 0.5099,
      "mean_token_accuracy": 0.8402730971574783,
      "num_tokens": 1013600215.0,
      "step": 31759
    },
    {
      "entropy": 0.4393008681945503,
      "epoch": 2.92208785458393,
      "grad_norm": 0.14802145957946777,
      "learning_rate": 2.600668568098875e-06,
      "loss": 0.427,
      "mean_token_accuracy": 0.8644580878317356,
      "num_tokens": 1013632459.0,
      "step": 31760
    },
    {
      "entropy": 0.46326209767721593,
      "epoch": 2.9221798609553713,
      "grad_norm": 0.14769063889980316,
      "learning_rate": 2.5976017419572486e-06,
      "loss": 0.4526,
      "mean_token_accuracy": 0.856969378888607,
      "num_tokens": 1013664458.0,
      "step": 31761
    },
    {
      "entropy": 0.5483143255114555,
      "epoch": 2.9222718673268124,
      "grad_norm": 0.17152808606624603,
      "learning_rate": 2.5945349158156224e-06,
      "loss": 0.5481,
      "mean_token_accuracy": 0.8327157609164715,
      "num_tokens": 1013695495.0,
      "step": 31762
    },
    {
      "entropy": 0.5452362280339003,
      "epoch": 2.9223638736982536,
      "grad_norm": 0.16819968819618225,
      "learning_rate": 2.5914680896739962e-06,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8307125680148602,
      "num_tokens": 1013727404.0,
      "step": 31763
    },
    {
      "entropy": 0.472548789344728,
      "epoch": 2.9224558800696947,
      "grad_norm": 0.1591678410768509,
      "learning_rate": 2.5884012635323705e-06,
      "loss": 0.4623,
      "mean_token_accuracy": 0.8502104803919792,
      "num_tokens": 1013759215.0,
      "step": 31764
    },
    {
      "entropy": 0.5980387711897492,
      "epoch": 2.9225478864411363,
      "grad_norm": 0.16690655052661896,
      "learning_rate": 2.5853344373907443e-06,
      "loss": 0.587,
      "mean_token_accuracy": 0.8169295750558376,
      "num_tokens": 1013791646.0,
      "step": 31765
    },
    {
      "entropy": 0.48281127121299505,
      "epoch": 2.9226398928125774,
      "grad_norm": 0.1530444473028183,
      "learning_rate": 2.5822676112491185e-06,
      "loss": 0.4735,
      "mean_token_accuracy": 0.8492398522794247,
      "num_tokens": 1013823802.0,
      "step": 31766
    },
    {
      "entropy": 0.5759454239159822,
      "epoch": 2.9227318991840185,
      "grad_norm": 0.1642029881477356,
      "learning_rate": 2.5792007851074923e-06,
      "loss": 0.5712,
      "mean_token_accuracy": 0.8224954307079315,
      "num_tokens": 1013855914.0,
      "step": 31767
    },
    {
      "entropy": 0.4334556922549382,
      "epoch": 2.9228239055554597,
      "grad_norm": 0.14723919332027435,
      "learning_rate": 2.5761339589658665e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.8644561991095543,
      "num_tokens": 1013888172.0,
      "step": 31768
    },
    {
      "entropy": 0.6005426216870546,
      "epoch": 2.922915911926901,
      "grad_norm": 0.17307747900485992,
      "learning_rate": 2.5730671328242403e-06,
      "loss": 0.6048,
      "mean_token_accuracy": 0.8136522956192493,
      "num_tokens": 1013919466.0,
      "step": 31769
    },
    {
      "entropy": 0.6403626021929085,
      "epoch": 2.9230079182983424,
      "grad_norm": 0.17645008862018585,
      "learning_rate": 2.570000306682614e-06,
      "loss": 0.6218,
      "mean_token_accuracy": 0.8029905259609222,
      "num_tokens": 1013951173.0,
      "step": 31770
    },
    {
      "entropy": 0.5787704344838858,
      "epoch": 2.9230999246697835,
      "grad_norm": 0.1703077107667923,
      "learning_rate": 2.566933480540988e-06,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8222888894379139,
      "num_tokens": 1013982491.0,
      "step": 31771
    },
    {
      "entropy": 0.5150749702006578,
      "epoch": 2.9231919310412247,
      "grad_norm": 0.15702736377716064,
      "learning_rate": 2.563866654399362e-06,
      "loss": 0.503,
      "mean_token_accuracy": 0.8420481458306313,
      "num_tokens": 1014014532.0,
      "step": 31772
    },
    {
      "entropy": 0.526671452447772,
      "epoch": 2.923283937412666,
      "grad_norm": 0.16156889498233795,
      "learning_rate": 2.5607998282577364e-06,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8332667015492916,
      "num_tokens": 1014046139.0,
      "step": 31773
    },
    {
      "entropy": 0.5016065519303083,
      "epoch": 2.923375943784107,
      "grad_norm": 0.16201089322566986,
      "learning_rate": 2.55773300211611e-06,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8429101593792439,
      "num_tokens": 1014077553.0,
      "step": 31774
    },
    {
      "entropy": 0.5188759043812752,
      "epoch": 2.9234679501555485,
      "grad_norm": 0.16235889494419098,
      "learning_rate": 2.5546661759744844e-06,
      "loss": 0.5179,
      "mean_token_accuracy": 0.838822990655899,
      "num_tokens": 1014109215.0,
      "step": 31775
    },
    {
      "entropy": 0.5142664834856987,
      "epoch": 2.9235599565269896,
      "grad_norm": 0.16145849227905273,
      "learning_rate": 2.551599349832858e-06,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8422900699079037,
      "num_tokens": 1014141705.0,
      "step": 31776
    },
    {
      "entropy": 0.46680191811174154,
      "epoch": 2.9236519628984308,
      "grad_norm": 0.15135565400123596,
      "learning_rate": 2.548532523691232e-06,
      "loss": 0.4495,
      "mean_token_accuracy": 0.8556445091962814,
      "num_tokens": 1014173508.0,
      "step": 31777
    },
    {
      "entropy": 0.617265859618783,
      "epoch": 2.923743969269872,
      "grad_norm": 0.17530739307403564,
      "learning_rate": 2.545465697549606e-06,
      "loss": 0.6089,
      "mean_token_accuracy": 0.810190986841917,
      "num_tokens": 1014204926.0,
      "step": 31778
    },
    {
      "entropy": 0.4179935157299042,
      "epoch": 2.923835975641313,
      "grad_norm": 0.1424587219953537,
      "learning_rate": 2.54239887140798e-06,
      "loss": 0.4033,
      "mean_token_accuracy": 0.8702496737241745,
      "num_tokens": 1014236758.0,
      "step": 31779
    },
    {
      "entropy": 0.5649140737950802,
      "epoch": 2.9239279820127546,
      "grad_norm": 0.16688962280750275,
      "learning_rate": 2.539332045266354e-06,
      "loss": 0.5495,
      "mean_token_accuracy": 0.8246295191347599,
      "num_tokens": 1014267864.0,
      "step": 31780
    },
    {
      "entropy": 0.6308311996981502,
      "epoch": 2.9240199883841957,
      "grad_norm": 0.1783447563648224,
      "learning_rate": 2.536265219124728e-06,
      "loss": 0.6245,
      "mean_token_accuracy": 0.8049199432134628,
      "num_tokens": 1014299580.0,
      "step": 31781
    },
    {
      "entropy": 0.4563632905483246,
      "epoch": 2.924111994755637,
      "grad_norm": 0.15261642634868622,
      "learning_rate": 2.533198392983102e-06,
      "loss": 0.4476,
      "mean_token_accuracy": 0.8585711009800434,
      "num_tokens": 1014330681.0,
      "step": 31782
    },
    {
      "entropy": 0.537350345402956,
      "epoch": 2.924204001127078,
      "grad_norm": 0.16106557846069336,
      "learning_rate": 2.530131566841476e-06,
      "loss": 0.5247,
      "mean_token_accuracy": 0.8357470221817493,
      "num_tokens": 1014362396.0,
      "step": 31783
    },
    {
      "entropy": 0.5359706999734044,
      "epoch": 2.924296007498519,
      "grad_norm": 0.1584431231021881,
      "learning_rate": 2.52706474069985e-06,
      "loss": 0.526,
      "mean_token_accuracy": 0.8323953002691269,
      "num_tokens": 1014394724.0,
      "step": 31784
    },
    {
      "entropy": 0.4857017621397972,
      "epoch": 2.9243880138699607,
      "grad_norm": 0.15865299105644226,
      "learning_rate": 2.5239979145582237e-06,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8488707691431046,
      "num_tokens": 1014425812.0,
      "step": 31785
    },
    {
      "entropy": 0.5502953995019197,
      "epoch": 2.924480020241402,
      "grad_norm": 0.16354504227638245,
      "learning_rate": 2.5209310884165975e-06,
      "loss": 0.5344,
      "mean_token_accuracy": 0.829794391989708,
      "num_tokens": 1014457374.0,
      "step": 31786
    },
    {
      "entropy": 0.47881444077938795,
      "epoch": 2.924572026612843,
      "grad_norm": 0.15266768634319305,
      "learning_rate": 2.5178642622749717e-06,
      "loss": 0.4564,
      "mean_token_accuracy": 0.8548004105687141,
      "num_tokens": 1014489630.0,
      "step": 31787
    },
    {
      "entropy": 0.5685544870793819,
      "epoch": 2.924664032984284,
      "grad_norm": 0.16803540289402008,
      "learning_rate": 2.514797436133346e-06,
      "loss": 0.5508,
      "mean_token_accuracy": 0.826168067753315,
      "num_tokens": 1014520984.0,
      "step": 31788
    },
    {
      "entropy": 0.539248077198863,
      "epoch": 2.9247560393557253,
      "grad_norm": 0.1693422645330429,
      "learning_rate": 2.5117306099917197e-06,
      "loss": 0.5337,
      "mean_token_accuracy": 0.8310528621077538,
      "num_tokens": 1014552773.0,
      "step": 31789
    },
    {
      "entropy": 0.4506468018516898,
      "epoch": 2.924848045727167,
      "grad_norm": 0.15235792100429535,
      "learning_rate": 2.5086637838500935e-06,
      "loss": 0.442,
      "mean_token_accuracy": 0.8562762700021267,
      "num_tokens": 1014584647.0,
      "step": 31790
    },
    {
      "entropy": 0.4318255139514804,
      "epoch": 2.924940052098608,
      "grad_norm": 0.14643001556396484,
      "learning_rate": 2.5055969577084677e-06,
      "loss": 0.405,
      "mean_token_accuracy": 0.8677957430481911,
      "num_tokens": 1014615496.0,
      "step": 31791
    },
    {
      "entropy": 0.4590385966002941,
      "epoch": 2.925032058470049,
      "grad_norm": 0.1489037275314331,
      "learning_rate": 2.5025301315668415e-06,
      "loss": 0.4569,
      "mean_token_accuracy": 0.853086918592453,
      "num_tokens": 1014647728.0,
      "step": 31792
    },
    {
      "entropy": 0.5555484155192971,
      "epoch": 2.9251240648414902,
      "grad_norm": 0.16631750762462616,
      "learning_rate": 2.4994633054252153e-06,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8330865316092968,
      "num_tokens": 1014678288.0,
      "step": 31793
    },
    {
      "entropy": 0.4374724659137428,
      "epoch": 2.9252160712129314,
      "grad_norm": 0.14767484366893768,
      "learning_rate": 2.4963964792835896e-06,
      "loss": 0.4267,
      "mean_token_accuracy": 0.8637932576239109,
      "num_tokens": 1014711030.0,
      "step": 31794
    },
    {
      "entropy": 0.4648166475817561,
      "epoch": 2.925308077584373,
      "grad_norm": 0.14528243243694305,
      "learning_rate": 2.493329653141964e-06,
      "loss": 0.4575,
      "mean_token_accuracy": 0.8571274094283581,
      "num_tokens": 1014743370.0,
      "step": 31795
    },
    {
      "entropy": 0.4918287182226777,
      "epoch": 2.925400083955814,
      "grad_norm": 0.15668189525604248,
      "learning_rate": 2.4902628270003376e-06,
      "loss": 0.4836,
      "mean_token_accuracy": 0.8477023169398308,
      "num_tokens": 1014775343.0,
      "step": 31796
    },
    {
      "entropy": 0.5489394282922149,
      "epoch": 2.925492090327255,
      "grad_norm": 0.16053560376167297,
      "learning_rate": 2.4871960008587114e-06,
      "loss": 0.5426,
      "mean_token_accuracy": 0.8299192190170288,
      "num_tokens": 1014807336.0,
      "step": 31797
    },
    {
      "entropy": 0.4948321571573615,
      "epoch": 2.9255840966986963,
      "grad_norm": 0.15627391636371613,
      "learning_rate": 2.4841291747170856e-06,
      "loss": 0.4891,
      "mean_token_accuracy": 0.843318585306406,
      "num_tokens": 1014839706.0,
      "step": 31798
    },
    {
      "entropy": 0.47013200260698795,
      "epoch": 2.9256761030701375,
      "grad_norm": 0.1515626758337021,
      "learning_rate": 2.4810623485754594e-06,
      "loss": 0.4634,
      "mean_token_accuracy": 0.8526234850287437,
      "num_tokens": 1014872028.0,
      "step": 31799
    },
    {
      "entropy": 0.5503885867074132,
      "epoch": 2.925768109441579,
      "grad_norm": 0.16112937033176422,
      "learning_rate": 2.4779955224338332e-06,
      "loss": 0.5427,
      "mean_token_accuracy": 0.8269736617803574,
      "num_tokens": 1014903847.0,
      "step": 31800
    },
    {
      "entropy": 0.49100455455482006,
      "epoch": 2.92586011581302,
      "grad_norm": 0.15130679309368134,
      "learning_rate": 2.474928696292207e-06,
      "loss": 0.4807,
      "mean_token_accuracy": 0.848766952753067,
      "num_tokens": 1014936510.0,
      "step": 31801
    },
    {
      "entropy": 0.5589046903187409,
      "epoch": 2.9259521221844613,
      "grad_norm": 0.1722332239151001,
      "learning_rate": 2.4718618701505812e-06,
      "loss": 0.5503,
      "mean_token_accuracy": 0.8290222659707069,
      "num_tokens": 1014968261.0,
      "step": 31802
    },
    {
      "entropy": 0.5001875625457615,
      "epoch": 2.9260441285559025,
      "grad_norm": 0.15268993377685547,
      "learning_rate": 2.4687950440089555e-06,
      "loss": 0.4811,
      "mean_token_accuracy": 0.8459999822080135,
      "num_tokens": 1014999727.0,
      "step": 31803
    },
    {
      "entropy": 0.43762048962526023,
      "epoch": 2.9261361349273436,
      "grad_norm": 0.1456039994955063,
      "learning_rate": 2.4657282178673293e-06,
      "loss": 0.4181,
      "mean_token_accuracy": 0.864673875272274,
      "num_tokens": 1015031949.0,
      "step": 31804
    },
    {
      "entropy": 0.5127415955066681,
      "epoch": 2.926228141298785,
      "grad_norm": 0.16095797717571259,
      "learning_rate": 2.462661391725703e-06,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8417551852762699,
      "num_tokens": 1015063980.0,
      "step": 31805
    },
    {
      "entropy": 0.5326036754995584,
      "epoch": 2.9263201476702263,
      "grad_norm": 0.16137591004371643,
      "learning_rate": 2.4595945655840773e-06,
      "loss": 0.5244,
      "mean_token_accuracy": 0.8328909575939178,
      "num_tokens": 1015096253.0,
      "step": 31806
    },
    {
      "entropy": 0.49617819767445326,
      "epoch": 2.9264121540416674,
      "grad_norm": 0.16500011086463928,
      "learning_rate": 2.456527739442451e-06,
      "loss": 0.4946,
      "mean_token_accuracy": 0.8406014963984489,
      "num_tokens": 1015128295.0,
      "step": 31807
    },
    {
      "entropy": 0.5766541510820389,
      "epoch": 2.9265041604131086,
      "grad_norm": 0.17179319262504578,
      "learning_rate": 2.453460913300825e-06,
      "loss": 0.5695,
      "mean_token_accuracy": 0.8218878097832203,
      "num_tokens": 1015159861.0,
      "step": 31808
    },
    {
      "entropy": 0.5356638766825199,
      "epoch": 2.9265961667845497,
      "grad_norm": 0.15949773788452148,
      "learning_rate": 2.450394087159199e-06,
      "loss": 0.5339,
      "mean_token_accuracy": 0.8332209438085556,
      "num_tokens": 1015191411.0,
      "step": 31809
    },
    {
      "entropy": 0.5301405349746346,
      "epoch": 2.9266881731559913,
      "grad_norm": 0.1594104915857315,
      "learning_rate": 2.4473272610175733e-06,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8357100710272789,
      "num_tokens": 1015223144.0,
      "step": 31810
    },
    {
      "entropy": 0.5070986058562994,
      "epoch": 2.9267801795274324,
      "grad_norm": 0.15827055275440216,
      "learning_rate": 2.444260434875947e-06,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8407880179584026,
      "num_tokens": 1015255340.0,
      "step": 31811
    },
    {
      "entropy": 0.46296371042262763,
      "epoch": 2.9268721858988735,
      "grad_norm": 0.15943875908851624,
      "learning_rate": 2.441193608734321e-06,
      "loss": 0.4605,
      "mean_token_accuracy": 0.8571669831871986,
      "num_tokens": 1015286464.0,
      "step": 31812
    },
    {
      "entropy": 0.571125216782093,
      "epoch": 2.9269641922703147,
      "grad_norm": 0.1653498411178589,
      "learning_rate": 2.4381267825926947e-06,
      "loss": 0.5587,
      "mean_token_accuracy": 0.8260752633213997,
      "num_tokens": 1015318219.0,
      "step": 31813
    },
    {
      "entropy": 0.4638118236325681,
      "epoch": 2.927056198641756,
      "grad_norm": 0.15019658207893372,
      "learning_rate": 2.435059956451069e-06,
      "loss": 0.4551,
      "mean_token_accuracy": 0.8536990135908127,
      "num_tokens": 1015349920.0,
      "step": 31814
    },
    {
      "entropy": 0.5400100033730268,
      "epoch": 2.9271482050131974,
      "grad_norm": 0.16151654720306396,
      "learning_rate": 2.4319931303094428e-06,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8301758542656898,
      "num_tokens": 1015382218.0,
      "step": 31815
    },
    {
      "entropy": 0.4884088281542063,
      "epoch": 2.9272402113846385,
      "grad_norm": 0.1525789052248001,
      "learning_rate": 2.4289263041678166e-06,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8469706699252129,
      "num_tokens": 1015414761.0,
      "step": 31816
    },
    {
      "entropy": 0.39252142142504454,
      "epoch": 2.9273322177560797,
      "grad_norm": 0.149458110332489,
      "learning_rate": 2.425859478026191e-06,
      "loss": 0.3802,
      "mean_token_accuracy": 0.8733635991811752,
      "num_tokens": 1015446447.0,
      "step": 31817
    },
    {
      "entropy": 0.5429373476654291,
      "epoch": 2.927424224127521,
      "grad_norm": 0.15866073966026306,
      "learning_rate": 2.422792651884565e-06,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8338942416012287,
      "num_tokens": 1015478491.0,
      "step": 31818
    },
    {
      "entropy": 0.42760237911716104,
      "epoch": 2.927516230498962,
      "grad_norm": 0.1449303776025772,
      "learning_rate": 2.419725825742939e-06,
      "loss": 0.4166,
      "mean_token_accuracy": 0.8710307739675045,
      "num_tokens": 1015511259.0,
      "step": 31819
    },
    {
      "entropy": 0.5001507932320237,
      "epoch": 2.9276082368704035,
      "grad_norm": 0.15767452120780945,
      "learning_rate": 2.4166589996013126e-06,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8442264460027218,
      "num_tokens": 1015543091.0,
      "step": 31820
    },
    {
      "entropy": 0.4908609064295888,
      "epoch": 2.9277002432418446,
      "grad_norm": 0.14977192878723145,
      "learning_rate": 2.413592173459687e-06,
      "loss": 0.4793,
      "mean_token_accuracy": 0.8474039360880852,
      "num_tokens": 1015575693.0,
      "step": 31821
    },
    {
      "entropy": 0.4660340053960681,
      "epoch": 2.9277922496132858,
      "grad_norm": 0.1547018140554428,
      "learning_rate": 2.4105253473180606e-06,
      "loss": 0.437,
      "mean_token_accuracy": 0.8573995046317577,
      "num_tokens": 1015606609.0,
      "step": 31822
    },
    {
      "entropy": 0.5500672478228807,
      "epoch": 2.927884255984727,
      "grad_norm": 0.1719299703836441,
      "learning_rate": 2.4074585211764344e-06,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8295006118714809,
      "num_tokens": 1015637266.0,
      "step": 31823
    },
    {
      "entropy": 0.48989981785416603,
      "epoch": 2.927976262356168,
      "grad_norm": 0.15497882664203644,
      "learning_rate": 2.4043916950348087e-06,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8470989428460598,
      "num_tokens": 1015669567.0,
      "step": 31824
    },
    {
      "entropy": 0.48269051779061556,
      "epoch": 2.9280682687276096,
      "grad_norm": 0.15256471931934357,
      "learning_rate": 2.401324868893183e-06,
      "loss": 0.47,
      "mean_token_accuracy": 0.8491369038820267,
      "num_tokens": 1015700913.0,
      "step": 31825
    },
    {
      "entropy": 0.4689708719961345,
      "epoch": 2.9281602750990507,
      "grad_norm": 0.14997228980064392,
      "learning_rate": 2.3982580427515567e-06,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8554032072424889,
      "num_tokens": 1015733261.0,
      "step": 31826
    },
    {
      "entropy": 0.42702114023268223,
      "epoch": 2.928252281470492,
      "grad_norm": 0.14362786710262299,
      "learning_rate": 2.3951912166099305e-06,
      "loss": 0.4178,
      "mean_token_accuracy": 0.8678942136466503,
      "num_tokens": 1015765806.0,
      "step": 31827
    },
    {
      "entropy": 0.5394585598260164,
      "epoch": 2.928344287841933,
      "grad_norm": 0.16322395205497742,
      "learning_rate": 2.3921243904683043e-06,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8355248235166073,
      "num_tokens": 1015797894.0,
      "step": 31828
    },
    {
      "entropy": 0.5181923108175397,
      "epoch": 2.928436294213374,
      "grad_norm": 0.15478505194187164,
      "learning_rate": 2.3890575643266785e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.8383127637207508,
      "num_tokens": 1015830057.0,
      "step": 31829
    },
    {
      "entropy": 0.44628699868917465,
      "epoch": 2.9285283005848157,
      "grad_norm": 0.14917904138565063,
      "learning_rate": 2.3859907381850523e-06,
      "loss": 0.4369,
      "mean_token_accuracy": 0.8625404313206673,
      "num_tokens": 1015862300.0,
      "step": 31830
    },
    {
      "entropy": 0.5413677834440023,
      "epoch": 2.928620306956257,
      "grad_norm": 0.16025298833847046,
      "learning_rate": 2.382923912043426e-06,
      "loss": 0.5281,
      "mean_token_accuracy": 0.831074807792902,
      "num_tokens": 1015894406.0,
      "step": 31831
    },
    {
      "entropy": 0.4865019442513585,
      "epoch": 2.928712313327698,
      "grad_norm": 0.1532217264175415,
      "learning_rate": 2.3798570859018003e-06,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8482651822268963,
      "num_tokens": 1015926421.0,
      "step": 31832
    },
    {
      "entropy": 0.4352242548484355,
      "epoch": 2.928804319699139,
      "grad_norm": 0.14681288599967957,
      "learning_rate": 2.3767902597601746e-06,
      "loss": 0.4234,
      "mean_token_accuracy": 0.864451490342617,
      "num_tokens": 1015958117.0,
      "step": 31833
    },
    {
      "entropy": 0.48680308647453785,
      "epoch": 2.9288963260705803,
      "grad_norm": 0.15219087898731232,
      "learning_rate": 2.3737234336185484e-06,
      "loss": 0.4835,
      "mean_token_accuracy": 0.847474280744791,
      "num_tokens": 1015990377.0,
      "step": 31834
    },
    {
      "entropy": 0.5819701449945569,
      "epoch": 2.928988332442022,
      "grad_norm": 0.1646842211484909,
      "learning_rate": 2.370656607476922e-06,
      "loss": 0.5628,
      "mean_token_accuracy": 0.821190133690834,
      "num_tokens": 1016022447.0,
      "step": 31835
    },
    {
      "entropy": 0.5504291709512472,
      "epoch": 2.929080338813463,
      "grad_norm": 0.16576990485191345,
      "learning_rate": 2.3675897813352964e-06,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8304933980107307,
      "num_tokens": 1016053160.0,
      "step": 31836
    },
    {
      "entropy": 0.5270520886406302,
      "epoch": 2.929172345184904,
      "grad_norm": 0.1611815094947815,
      "learning_rate": 2.36452295519367e-06,
      "loss": 0.5194,
      "mean_token_accuracy": 0.8346703983843327,
      "num_tokens": 1016085789.0,
      "step": 31837
    },
    {
      "entropy": 0.46977974101901054,
      "epoch": 2.9292643515563452,
      "grad_norm": 0.15456072986125946,
      "learning_rate": 2.361456129052044e-06,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8560431450605392,
      "num_tokens": 1016117516.0,
      "step": 31838
    },
    {
      "entropy": 0.46722599025815725,
      "epoch": 2.9293563579277864,
      "grad_norm": 0.15276995301246643,
      "learning_rate": 2.3583893029104182e-06,
      "loss": 0.457,
      "mean_token_accuracy": 0.8547835759818554,
      "num_tokens": 1016149603.0,
      "step": 31839
    },
    {
      "entropy": 0.43160336883738637,
      "epoch": 2.929448364299228,
      "grad_norm": 0.1421881765127182,
      "learning_rate": 2.3553224767687924e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8671656735241413,
      "num_tokens": 1016182173.0,
      "step": 31840
    },
    {
      "entropy": 0.4081073310226202,
      "epoch": 2.929540370670669,
      "grad_norm": 0.14026784896850586,
      "learning_rate": 2.3522556506271662e-06,
      "loss": 0.3933,
      "mean_token_accuracy": 0.871900774538517,
      "num_tokens": 1016214527.0,
      "step": 31841
    },
    {
      "entropy": 0.5853007724508643,
      "epoch": 2.92963237704211,
      "grad_norm": 0.17149214446544647,
      "learning_rate": 2.34918882448554e-06,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8194554410874844,
      "num_tokens": 1016246405.0,
      "step": 31842
    },
    {
      "entropy": 0.46794142527505755,
      "epoch": 2.9297243834135513,
      "grad_norm": 0.15217973291873932,
      "learning_rate": 2.346121998343914e-06,
      "loss": 0.4542,
      "mean_token_accuracy": 0.8568478785455227,
      "num_tokens": 1016278580.0,
      "step": 31843
    },
    {
      "entropy": 0.5532304411754012,
      "epoch": 2.9298163897849925,
      "grad_norm": 0.1631949096918106,
      "learning_rate": 2.343055172202288e-06,
      "loss": 0.5455,
      "mean_token_accuracy": 0.8295558094978333,
      "num_tokens": 1016310655.0,
      "step": 31844
    },
    {
      "entropy": 0.5575850736349821,
      "epoch": 2.929908396156434,
      "grad_norm": 0.16690362989902496,
      "learning_rate": 2.339988346060662e-06,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8255917355418205,
      "num_tokens": 1016342294.0,
      "step": 31845
    },
    {
      "entropy": 0.4923320454545319,
      "epoch": 2.930000402527875,
      "grad_norm": 0.15967340767383575,
      "learning_rate": 2.3369215199190357e-06,
      "loss": 0.4859,
      "mean_token_accuracy": 0.8460588492453098,
      "num_tokens": 1016373688.0,
      "step": 31846
    },
    {
      "entropy": 0.5929568894207478,
      "epoch": 2.9300924088993163,
      "grad_norm": 0.1681520938873291,
      "learning_rate": 2.33385469377741e-06,
      "loss": 0.5741,
      "mean_token_accuracy": 0.8208691105246544,
      "num_tokens": 1016405893.0,
      "step": 31847
    },
    {
      "entropy": 0.43029348459094763,
      "epoch": 2.9301844152707575,
      "grad_norm": 0.1487070769071579,
      "learning_rate": 2.330787867635784e-06,
      "loss": 0.4219,
      "mean_token_accuracy": 0.8672554604709148,
      "num_tokens": 1016438536.0,
      "step": 31848
    },
    {
      "entropy": 0.5267316764220595,
      "epoch": 2.9302764216421986,
      "grad_norm": 0.16183118522167206,
      "learning_rate": 2.327721041494158e-06,
      "loss": 0.5195,
      "mean_token_accuracy": 0.8369435705244541,
      "num_tokens": 1016470656.0,
      "step": 31849
    },
    {
      "entropy": 0.4299102472141385,
      "epoch": 2.93036842801364,
      "grad_norm": 0.14371545612812042,
      "learning_rate": 2.3246542153525317e-06,
      "loss": 0.4152,
      "mean_token_accuracy": 0.8650827035307884,
      "num_tokens": 1016503218.0,
      "step": 31850
    },
    {
      "entropy": 0.5304584968835115,
      "epoch": 2.9304604343850813,
      "grad_norm": 0.15858617424964905,
      "learning_rate": 2.3215873892109055e-06,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8360789008438587,
      "num_tokens": 1016535254.0,
      "step": 31851
    },
    {
      "entropy": 0.5005966820754111,
      "epoch": 2.9305524407565224,
      "grad_norm": 0.15687420964241028,
      "learning_rate": 2.3185205630692797e-06,
      "loss": 0.491,
      "mean_token_accuracy": 0.8464709334075451,
      "num_tokens": 1016566805.0,
      "step": 31852
    },
    {
      "entropy": 0.40490155667066574,
      "epoch": 2.9306444471279636,
      "grad_norm": 0.14237353205680847,
      "learning_rate": 2.3154537369276535e-06,
      "loss": 0.3941,
      "mean_token_accuracy": 0.8747315816581249,
      "num_tokens": 1016599260.0,
      "step": 31853
    },
    {
      "entropy": 0.47504247003234923,
      "epoch": 2.9307364534994047,
      "grad_norm": 0.14704562723636627,
      "learning_rate": 2.3123869107860278e-06,
      "loss": 0.4642,
      "mean_token_accuracy": 0.8563927449285984,
      "num_tokens": 1016631679.0,
      "step": 31854
    },
    {
      "entropy": 0.46811259374953806,
      "epoch": 2.9308284598708463,
      "grad_norm": 0.1544298380613327,
      "learning_rate": 2.309320084644402e-06,
      "loss": 0.46,
      "mean_token_accuracy": 0.8514516614377499,
      "num_tokens": 1016663167.0,
      "step": 31855
    },
    {
      "entropy": 0.45336119877174497,
      "epoch": 2.9309204662422874,
      "grad_norm": 0.15037164092063904,
      "learning_rate": 2.306253258502776e-06,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8586073368787766,
      "num_tokens": 1016695248.0,
      "step": 31856
    },
    {
      "entropy": 0.4624192325863987,
      "epoch": 2.9310124726137285,
      "grad_norm": 0.1511041671037674,
      "learning_rate": 2.3031864323611496e-06,
      "loss": 0.4403,
      "mean_token_accuracy": 0.8563654087483883,
      "num_tokens": 1016726608.0,
      "step": 31857
    },
    {
      "entropy": 0.4411197677254677,
      "epoch": 2.9311044789851697,
      "grad_norm": 0.14751146733760834,
      "learning_rate": 2.3001196062195234e-06,
      "loss": 0.4279,
      "mean_token_accuracy": 0.8632020354270935,
      "num_tokens": 1016758397.0,
      "step": 31858
    },
    {
      "entropy": 0.4658185434527695,
      "epoch": 2.931196485356611,
      "grad_norm": 0.14946973323822021,
      "learning_rate": 2.2970527800778976e-06,
      "loss": 0.458,
      "mean_token_accuracy": 0.8566535264253616,
      "num_tokens": 1016790899.0,
      "step": 31859
    },
    {
      "entropy": 0.4889997225254774,
      "epoch": 2.9312884917280524,
      "grad_norm": 0.15636591613292694,
      "learning_rate": 2.2939859539362714e-06,
      "loss": 0.4752,
      "mean_token_accuracy": 0.8498530499637127,
      "num_tokens": 1016822578.0,
      "step": 31860
    },
    {
      "entropy": 0.533542257733643,
      "epoch": 2.9313804980994935,
      "grad_norm": 0.16283245384693146,
      "learning_rate": 2.2909191277946452e-06,
      "loss": 0.5169,
      "mean_token_accuracy": 0.8373825699090958,
      "num_tokens": 1016853589.0,
      "step": 31861
    },
    {
      "entropy": 0.5873375153169036,
      "epoch": 2.9314725044709347,
      "grad_norm": 0.16492138803005219,
      "learning_rate": 2.2878523016530194e-06,
      "loss": 0.5761,
      "mean_token_accuracy": 0.8214927613735199,
      "num_tokens": 1016885567.0,
      "step": 31862
    },
    {
      "entropy": 0.4200665820389986,
      "epoch": 2.931564510842376,
      "grad_norm": 0.1489538997411728,
      "learning_rate": 2.2847854755113937e-06,
      "loss": 0.4086,
      "mean_token_accuracy": 0.869022898375988,
      "num_tokens": 1016917858.0,
      "step": 31863
    },
    {
      "entropy": 0.5721034584566951,
      "epoch": 2.931656517213817,
      "grad_norm": 0.16578391194343567,
      "learning_rate": 2.2817186493697675e-06,
      "loss": 0.5663,
      "mean_token_accuracy": 0.8239295668900013,
      "num_tokens": 1016949917.0,
      "step": 31864
    },
    {
      "entropy": 0.5113514279946685,
      "epoch": 2.9317485235852585,
      "grad_norm": 0.1570456176996231,
      "learning_rate": 2.2786518232281413e-06,
      "loss": 0.5081,
      "mean_token_accuracy": 0.8411736451089382,
      "num_tokens": 1016982320.0,
      "step": 31865
    },
    {
      "entropy": 0.4037983138114214,
      "epoch": 2.9318405299566996,
      "grad_norm": 0.14378871023654938,
      "learning_rate": 2.275584997086515e-06,
      "loss": 0.3845,
      "mean_token_accuracy": 0.8742812052369118,
      "num_tokens": 1017014320.0,
      "step": 31866
    },
    {
      "entropy": 0.4946711426600814,
      "epoch": 2.9319325363281408,
      "grad_norm": 0.15392398834228516,
      "learning_rate": 2.2725181709448893e-06,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8431714773178101,
      "num_tokens": 1017046775.0,
      "step": 31867
    },
    {
      "entropy": 0.5698843076825142,
      "epoch": 2.932024542699582,
      "grad_norm": 0.16984274983406067,
      "learning_rate": 2.269451344803263e-06,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8229168355464935,
      "num_tokens": 1017078184.0,
      "step": 31868
    },
    {
      "entropy": 0.41875117737799883,
      "epoch": 2.932116549071023,
      "grad_norm": 0.14541591703891754,
      "learning_rate": 2.2663845186616373e-06,
      "loss": 0.4133,
      "mean_token_accuracy": 0.867579497396946,
      "num_tokens": 1017110380.0,
      "step": 31869
    },
    {
      "entropy": 0.4859205358661711,
      "epoch": 2.9322085554424646,
      "grad_norm": 0.15406835079193115,
      "learning_rate": 2.263317692520011e-06,
      "loss": 0.472,
      "mean_token_accuracy": 0.8505454994738102,
      "num_tokens": 1017142136.0,
      "step": 31870
    },
    {
      "entropy": 0.43754747277125716,
      "epoch": 2.9323005618139057,
      "grad_norm": 0.15280751883983612,
      "learning_rate": 2.2602508663783853e-06,
      "loss": 0.4223,
      "mean_token_accuracy": 0.8616149798035622,
      "num_tokens": 1017174263.0,
      "step": 31871
    },
    {
      "entropy": 0.43586302921175957,
      "epoch": 2.932392568185347,
      "grad_norm": 0.14607171714305878,
      "learning_rate": 2.257184040236759e-06,
      "loss": 0.4288,
      "mean_token_accuracy": 0.8644662909209728,
      "num_tokens": 1017207031.0,
      "step": 31872
    },
    {
      "entropy": 0.5491888727992773,
      "epoch": 2.932484574556788,
      "grad_norm": 0.16220644116401672,
      "learning_rate": 2.254117214095133e-06,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8285768665373325,
      "num_tokens": 1017238753.0,
      "step": 31873
    },
    {
      "entropy": 0.39852727856487036,
      "epoch": 2.932576580928229,
      "grad_norm": 0.1423807144165039,
      "learning_rate": 2.2510503879535067e-06,
      "loss": 0.3867,
      "mean_token_accuracy": 0.8707529343664646,
      "num_tokens": 1017271083.0,
      "step": 31874
    },
    {
      "entropy": 0.4635793917113915,
      "epoch": 2.9326685872996707,
      "grad_norm": 0.147701233625412,
      "learning_rate": 2.247983561811881e-06,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8577213808894157,
      "num_tokens": 1017303023.0,
      "step": 31875
    },
    {
      "entropy": 0.49756394140422344,
      "epoch": 2.932760593671112,
      "grad_norm": 0.15468250215053558,
      "learning_rate": 2.2449167356702548e-06,
      "loss": 0.4875,
      "mean_token_accuracy": 0.8485403805971146,
      "num_tokens": 1017335214.0,
      "step": 31876
    },
    {
      "entropy": 0.4109184518456459,
      "epoch": 2.932852600042553,
      "grad_norm": 0.14459457993507385,
      "learning_rate": 2.241849909528629e-06,
      "loss": 0.3961,
      "mean_token_accuracy": 0.8724530301988125,
      "num_tokens": 1017365963.0,
      "step": 31877
    },
    {
      "entropy": 0.4461100082844496,
      "epoch": 2.932944606413994,
      "grad_norm": 0.14786817133426666,
      "learning_rate": 2.2387830833870032e-06,
      "loss": 0.428,
      "mean_token_accuracy": 0.8647199459373951,
      "num_tokens": 1017398132.0,
      "step": 31878
    },
    {
      "entropy": 0.38667009002529085,
      "epoch": 2.9330366127854353,
      "grad_norm": 0.14079909026622772,
      "learning_rate": 2.235716257245377e-06,
      "loss": 0.3783,
      "mean_token_accuracy": 0.8779804594814777,
      "num_tokens": 1017430451.0,
      "step": 31879
    },
    {
      "entropy": 0.4658886604011059,
      "epoch": 2.933128619156877,
      "grad_norm": 0.14897577464580536,
      "learning_rate": 2.232649431103751e-06,
      "loss": 0.4549,
      "mean_token_accuracy": 0.8562872409820557,
      "num_tokens": 1017462310.0,
      "step": 31880
    },
    {
      "entropy": 0.5524609591811895,
      "epoch": 2.933220625528318,
      "grad_norm": 0.16645626723766327,
      "learning_rate": 2.2295826049621246e-06,
      "loss": 0.541,
      "mean_token_accuracy": 0.8286476880311966,
      "num_tokens": 1017493552.0,
      "step": 31881
    },
    {
      "entropy": 0.5730965109542012,
      "epoch": 2.933312631899759,
      "grad_norm": 0.16319330036640167,
      "learning_rate": 2.226515778820499e-06,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8210746422410011,
      "num_tokens": 1017525762.0,
      "step": 31882
    },
    {
      "entropy": 0.4656250597909093,
      "epoch": 2.9334046382712002,
      "grad_norm": 0.15613210201263428,
      "learning_rate": 2.2234489526788726e-06,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8527276329696178,
      "num_tokens": 1017558363.0,
      "step": 31883
    },
    {
      "entropy": 0.4432260962203145,
      "epoch": 2.9334966446426414,
      "grad_norm": 0.14654192328453064,
      "learning_rate": 2.220382126537247e-06,
      "loss": 0.4383,
      "mean_token_accuracy": 0.86286611109972,
      "num_tokens": 1017590742.0,
      "step": 31884
    },
    {
      "entropy": 0.5554750873707235,
      "epoch": 2.933588651014083,
      "grad_norm": 0.16290423274040222,
      "learning_rate": 2.2173153003956207e-06,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8270453847944736,
      "num_tokens": 1017622650.0,
      "step": 31885
    },
    {
      "entropy": 0.5064868237823248,
      "epoch": 2.933680657385524,
      "grad_norm": 0.16115662455558777,
      "learning_rate": 2.214248474253995e-06,
      "loss": 0.5012,
      "mean_token_accuracy": 0.839052002876997,
      "num_tokens": 1017654531.0,
      "step": 31886
    },
    {
      "entropy": 0.4720356110483408,
      "epoch": 2.933772663756965,
      "grad_norm": 0.1598621904850006,
      "learning_rate": 2.2111816481123687e-06,
      "loss": 0.4682,
      "mean_token_accuracy": 0.8536981604993343,
      "num_tokens": 1017686147.0,
      "step": 31887
    },
    {
      "entropy": 0.5446380572393537,
      "epoch": 2.9338646701284063,
      "grad_norm": 0.1617157906293869,
      "learning_rate": 2.2081148219707425e-06,
      "loss": 0.5241,
      "mean_token_accuracy": 0.8324738256633282,
      "num_tokens": 1017717704.0,
      "step": 31888
    },
    {
      "entropy": 0.46494124084711075,
      "epoch": 2.9339566764998475,
      "grad_norm": 0.14464975893497467,
      "learning_rate": 2.2050479958291163e-06,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8603070601820946,
      "num_tokens": 1017749996.0,
      "step": 31889
    },
    {
      "entropy": 0.47443937323987484,
      "epoch": 2.934048682871289,
      "grad_norm": 0.1562570333480835,
      "learning_rate": 2.2019811696874905e-06,
      "loss": 0.463,
      "mean_token_accuracy": 0.8572038114070892,
      "num_tokens": 1017782171.0,
      "step": 31890
    },
    {
      "entropy": 0.5013790428638458,
      "epoch": 2.93414068924273,
      "grad_norm": 0.15703104436397552,
      "learning_rate": 2.1989143435458643e-06,
      "loss": 0.4905,
      "mean_token_accuracy": 0.8445847369730473,
      "num_tokens": 1017813889.0,
      "step": 31891
    },
    {
      "entropy": 0.48756731837056577,
      "epoch": 2.9342326956141713,
      "grad_norm": 0.15677174925804138,
      "learning_rate": 2.1958475174042385e-06,
      "loss": 0.4687,
      "mean_token_accuracy": 0.849238283932209,
      "num_tokens": 1017845091.0,
      "step": 31892
    },
    {
      "entropy": 0.472154650837183,
      "epoch": 2.9343247019856125,
      "grad_norm": 0.1529429405927658,
      "learning_rate": 2.1927806912626123e-06,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8535237237811089,
      "num_tokens": 1017877005.0,
      "step": 31893
    },
    {
      "entropy": 0.40906788129359484,
      "epoch": 2.9344167083570536,
      "grad_norm": 0.14156168699264526,
      "learning_rate": 2.1897138651209866e-06,
      "loss": 0.3954,
      "mean_token_accuracy": 0.8749310821294785,
      "num_tokens": 1017909264.0,
      "step": 31894
    },
    {
      "entropy": 0.6190978744998574,
      "epoch": 2.934508714728495,
      "grad_norm": 0.17588181793689728,
      "learning_rate": 2.1866470389793604e-06,
      "loss": 0.6063,
      "mean_token_accuracy": 0.80882802978158,
      "num_tokens": 1017941230.0,
      "step": 31895
    },
    {
      "entropy": 0.4447037121281028,
      "epoch": 2.9346007210999363,
      "grad_norm": 0.1497122198343277,
      "learning_rate": 2.183580212837734e-06,
      "loss": 0.4315,
      "mean_token_accuracy": 0.8604005686938763,
      "num_tokens": 1017973123.0,
      "step": 31896
    },
    {
      "entropy": 0.4718636814504862,
      "epoch": 2.9346927274713774,
      "grad_norm": 0.1539813131093979,
      "learning_rate": 2.1805133866961084e-06,
      "loss": 0.4609,
      "mean_token_accuracy": 0.8514469675719738,
      "num_tokens": 1018004008.0,
      "step": 31897
    },
    {
      "entropy": 0.5197385204955935,
      "epoch": 2.9347847338428186,
      "grad_norm": 0.16423337161540985,
      "learning_rate": 2.177446560554482e-06,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8390932157635689,
      "num_tokens": 1018035464.0,
      "step": 31898
    },
    {
      "entropy": 0.47312190849334,
      "epoch": 2.9348767402142597,
      "grad_norm": 0.15302179753780365,
      "learning_rate": 2.1743797344128564e-06,
      "loss": 0.4649,
      "mean_token_accuracy": 0.8504030033946037,
      "num_tokens": 1018068052.0,
      "step": 31899
    },
    {
      "entropy": 0.6004688208922744,
      "epoch": 2.9349687465857013,
      "grad_norm": 0.16812525689601898,
      "learning_rate": 2.1713129082712302e-06,
      "loss": 0.5883,
      "mean_token_accuracy": 0.8129074089229107,
      "num_tokens": 1018100529.0,
      "step": 31900
    },
    {
      "entropy": 0.5673424479318783,
      "epoch": 2.9350607529571424,
      "grad_norm": 0.1645883172750473,
      "learning_rate": 2.1682460821296044e-06,
      "loss": 0.5538,
      "mean_token_accuracy": 0.8259999975562096,
      "num_tokens": 1018132562.0,
      "step": 31901
    },
    {
      "entropy": 0.5119147077202797,
      "epoch": 2.9351527593285835,
      "grad_norm": 0.15730255842208862,
      "learning_rate": 2.1651792559879782e-06,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8409453257918358,
      "num_tokens": 1018164973.0,
      "step": 31902
    },
    {
      "entropy": 0.4768095528706908,
      "epoch": 2.9352447657000247,
      "grad_norm": 0.15072426199913025,
      "learning_rate": 2.162112429846352e-06,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8506843112409115,
      "num_tokens": 1018197018.0,
      "step": 31903
    },
    {
      "entropy": 0.5233382740989327,
      "epoch": 2.935336772071466,
      "grad_norm": 0.15660379827022552,
      "learning_rate": 2.159045603704726e-06,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8381690122187138,
      "num_tokens": 1018229499.0,
      "step": 31904
    },
    {
      "entropy": 0.538415415212512,
      "epoch": 2.9354287784429074,
      "grad_norm": 0.1662788689136505,
      "learning_rate": 2.1559787775631e-06,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8341865167021751,
      "num_tokens": 1018261659.0,
      "step": 31905
    },
    {
      "entropy": 0.4987659747712314,
      "epoch": 2.9355207848143485,
      "grad_norm": 0.16008035838603973,
      "learning_rate": 2.152911951421474e-06,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8429819494485855,
      "num_tokens": 1018294251.0,
      "step": 31906
    },
    {
      "entropy": 0.5199867910705507,
      "epoch": 2.9356127911857897,
      "grad_norm": 0.16211873292922974,
      "learning_rate": 2.149845125279848e-06,
      "loss": 0.5106,
      "mean_token_accuracy": 0.8378139808773994,
      "num_tokens": 1018325691.0,
      "step": 31907
    },
    {
      "entropy": 0.6080751121044159,
      "epoch": 2.935704797557231,
      "grad_norm": 0.17172957956790924,
      "learning_rate": 2.146778299138222e-06,
      "loss": 0.6003,
      "mean_token_accuracy": 0.811909768730402,
      "num_tokens": 1018357845.0,
      "step": 31908
    },
    {
      "entropy": 0.5615025609731674,
      "epoch": 2.935796803928672,
      "grad_norm": 0.1625656634569168,
      "learning_rate": 2.143711472996596e-06,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8275767415761948,
      "num_tokens": 1018389877.0,
      "step": 31909
    },
    {
      "entropy": 0.534761302638799,
      "epoch": 2.9358888103001135,
      "grad_norm": 0.16178590059280396,
      "learning_rate": 2.14064464685497e-06,
      "loss": 0.518,
      "mean_token_accuracy": 0.8350806683301926,
      "num_tokens": 1018422113.0,
      "step": 31910
    },
    {
      "entropy": 0.4741458436474204,
      "epoch": 2.9359808166715546,
      "grad_norm": 0.14809750020503998,
      "learning_rate": 2.1375778207133437e-06,
      "loss": 0.4539,
      "mean_token_accuracy": 0.8539791889488697,
      "num_tokens": 1018454698.0,
      "step": 31911
    },
    {
      "entropy": 0.4986515175551176,
      "epoch": 2.9360728230429958,
      "grad_norm": 0.16071145236492157,
      "learning_rate": 2.1345109945717175e-06,
      "loss": 0.4981,
      "mean_token_accuracy": 0.8394823558628559,
      "num_tokens": 1018487292.0,
      "step": 31912
    },
    {
      "entropy": 0.5362431388348341,
      "epoch": 2.936164829414437,
      "grad_norm": 0.16190539300441742,
      "learning_rate": 2.1314441684300917e-06,
      "loss": 0.5307,
      "mean_token_accuracy": 0.8347228579223156,
      "num_tokens": 1018519121.0,
      "step": 31913
    },
    {
      "entropy": 0.5180257251486182,
      "epoch": 2.936256835785878,
      "grad_norm": 0.16082654893398285,
      "learning_rate": 2.128377342288466e-06,
      "loss": 0.5059,
      "mean_token_accuracy": 0.8385635763406754,
      "num_tokens": 1018550807.0,
      "step": 31914
    },
    {
      "entropy": 0.5561826550401747,
      "epoch": 2.9363488421573196,
      "grad_norm": 0.16613531112670898,
      "learning_rate": 2.1253105161468398e-06,
      "loss": 0.5414,
      "mean_token_accuracy": 0.8299732506275177,
      "num_tokens": 1018581663.0,
      "step": 31915
    },
    {
      "entropy": 0.48910319060087204,
      "epoch": 2.9364408485287607,
      "grad_norm": 0.15152607858181,
      "learning_rate": 2.122243690005214e-06,
      "loss": 0.4788,
      "mean_token_accuracy": 0.8492613211274147,
      "num_tokens": 1018614283.0,
      "step": 31916
    },
    {
      "entropy": 0.48720199055969715,
      "epoch": 2.936532854900202,
      "grad_norm": 0.15599492192268372,
      "learning_rate": 2.1191768638635878e-06,
      "loss": 0.4828,
      "mean_token_accuracy": 0.8451715260744095,
      "num_tokens": 1018646362.0,
      "step": 31917
    },
    {
      "entropy": 0.46120519982650876,
      "epoch": 2.936624861271643,
      "grad_norm": 0.15213419497013092,
      "learning_rate": 2.1161100377219616e-06,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8596899323165417,
      "num_tokens": 1018678792.0,
      "step": 31918
    },
    {
      "entropy": 0.4917578902095556,
      "epoch": 2.936716867643084,
      "grad_norm": 0.15303394198417664,
      "learning_rate": 2.1130432115803354e-06,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8489046171307564,
      "num_tokens": 1018710488.0,
      "step": 31919
    },
    {
      "entropy": 0.4953863089904189,
      "epoch": 2.9368088740145257,
      "grad_norm": 0.15746837854385376,
      "learning_rate": 2.1099763854387096e-06,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8457692973315716,
      "num_tokens": 1018742446.0,
      "step": 31920
    },
    {
      "entropy": 0.4861604515463114,
      "epoch": 2.936900880385967,
      "grad_norm": 0.15808406472206116,
      "learning_rate": 2.1069095592970834e-06,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8482159227132797,
      "num_tokens": 1018774116.0,
      "step": 31921
    },
    {
      "entropy": 0.45176790095865726,
      "epoch": 2.936992886757408,
      "grad_norm": 0.15432070195674896,
      "learning_rate": 2.1038427331554576e-06,
      "loss": 0.4544,
      "mean_token_accuracy": 0.8580611199140549,
      "num_tokens": 1018805532.0,
      "step": 31922
    },
    {
      "entropy": 0.5721796322613955,
      "epoch": 2.937084893128849,
      "grad_norm": 0.16700537502765656,
      "learning_rate": 2.1007759070138314e-06,
      "loss": 0.558,
      "mean_token_accuracy": 0.8254141248762608,
      "num_tokens": 1018837566.0,
      "step": 31923
    },
    {
      "entropy": 0.4811301724985242,
      "epoch": 2.9371768995002903,
      "grad_norm": 0.15677954256534576,
      "learning_rate": 2.0977090808722057e-06,
      "loss": 0.4682,
      "mean_token_accuracy": 0.8495360352098942,
      "num_tokens": 1018868893.0,
      "step": 31924
    },
    {
      "entropy": 0.4324828233802691,
      "epoch": 2.937268905871732,
      "grad_norm": 0.14781829714775085,
      "learning_rate": 2.0946422547305795e-06,
      "loss": 0.4202,
      "mean_token_accuracy": 0.8672522492706776,
      "num_tokens": 1018901295.0,
      "step": 31925
    },
    {
      "entropy": 0.4908102182671428,
      "epoch": 2.937360912243173,
      "grad_norm": 0.154385045170784,
      "learning_rate": 2.0915754285889533e-06,
      "loss": 0.4851,
      "mean_token_accuracy": 0.8428973779082298,
      "num_tokens": 1018933102.0,
      "step": 31926
    },
    {
      "entropy": 0.508835673565045,
      "epoch": 2.937452918614614,
      "grad_norm": 0.1602274477481842,
      "learning_rate": 2.088508602447327e-06,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8447184190154076,
      "num_tokens": 1018964839.0,
      "step": 31927
    },
    {
      "entropy": 0.4033956313505769,
      "epoch": 2.9375449249860552,
      "grad_norm": 0.1468179076910019,
      "learning_rate": 2.0854417763057013e-06,
      "loss": 0.3971,
      "mean_token_accuracy": 0.87021454423666,
      "num_tokens": 1018997071.0,
      "step": 31928
    },
    {
      "entropy": 0.423188557382673,
      "epoch": 2.9376369313574964,
      "grad_norm": 0.14218716323375702,
      "learning_rate": 2.0823749501640755e-06,
      "loss": 0.4144,
      "mean_token_accuracy": 0.8681578151881695,
      "num_tokens": 1019029441.0,
      "step": 31929
    },
    {
      "entropy": 0.3939781042281538,
      "epoch": 2.937728937728938,
      "grad_norm": 0.13439029455184937,
      "learning_rate": 2.0793081240224493e-06,
      "loss": 0.3789,
      "mean_token_accuracy": 0.8791588023304939,
      "num_tokens": 1019061591.0,
      "step": 31930
    },
    {
      "entropy": 0.509125585667789,
      "epoch": 2.937820944100379,
      "grad_norm": 0.160905659198761,
      "learning_rate": 2.076241297880823e-06,
      "loss": 0.504,
      "mean_token_accuracy": 0.8411611206829548,
      "num_tokens": 1019093680.0,
      "step": 31931
    },
    {
      "entropy": 0.5803996045142412,
      "epoch": 2.93791295047182,
      "grad_norm": 0.16711175441741943,
      "learning_rate": 2.0731744717391973e-06,
      "loss": 0.5772,
      "mean_token_accuracy": 0.8174074962735176,
      "num_tokens": 1019125338.0,
      "step": 31932
    },
    {
      "entropy": 0.5352561729960144,
      "epoch": 2.9380049568432614,
      "grad_norm": 0.16871005296707153,
      "learning_rate": 2.070107645597571e-06,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8312118761241436,
      "num_tokens": 1019157448.0,
      "step": 31933
    },
    {
      "entropy": 0.5473991613835096,
      "epoch": 2.9380969632147025,
      "grad_norm": 0.16102084517478943,
      "learning_rate": 2.067040819455945e-06,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8287815302610397,
      "num_tokens": 1019189539.0,
      "step": 31934
    },
    {
      "entropy": 0.4857138483785093,
      "epoch": 2.938188969586144,
      "grad_norm": 0.15841029584407806,
      "learning_rate": 2.0639739933143187e-06,
      "loss": 0.4741,
      "mean_token_accuracy": 0.8510875627398491,
      "num_tokens": 1019221583.0,
      "step": 31935
    },
    {
      "entropy": 0.4938088618218899,
      "epoch": 2.938280975957585,
      "grad_norm": 0.15819412469863892,
      "learning_rate": 2.060907167172693e-06,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8437407910823822,
      "num_tokens": 1019253303.0,
      "step": 31936
    },
    {
      "entropy": 0.5426892507821321,
      "epoch": 2.9383729823290263,
      "grad_norm": 0.1624724268913269,
      "learning_rate": 2.057840341031067e-06,
      "loss": 0.5324,
      "mean_token_accuracy": 0.8297702074050903,
      "num_tokens": 1019285225.0,
      "step": 31937
    },
    {
      "entropy": 0.5545469224452972,
      "epoch": 2.9384649887004675,
      "grad_norm": 0.16594456136226654,
      "learning_rate": 2.054773514889441e-06,
      "loss": 0.5468,
      "mean_token_accuracy": 0.8253882974386215,
      "num_tokens": 1019316620.0,
      "step": 31938
    },
    {
      "entropy": 0.518985079601407,
      "epoch": 2.9385569950719086,
      "grad_norm": 0.1555393934249878,
      "learning_rate": 2.051706688747815e-06,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8398794308304787,
      "num_tokens": 1019348761.0,
      "step": 31939
    },
    {
      "entropy": 0.4573060004040599,
      "epoch": 2.93864900144335,
      "grad_norm": 0.15427973866462708,
      "learning_rate": 2.048639862606189e-06,
      "loss": 0.4482,
      "mean_token_accuracy": 0.8571955785155296,
      "num_tokens": 1019380729.0,
      "step": 31940
    },
    {
      "entropy": 0.45194597356021404,
      "epoch": 2.9387410078147913,
      "grad_norm": 0.14913706481456757,
      "learning_rate": 2.045573036464563e-06,
      "loss": 0.4421,
      "mean_token_accuracy": 0.8615854680538177,
      "num_tokens": 1019412898.0,
      "step": 31941
    },
    {
      "entropy": 0.4467101311311126,
      "epoch": 2.9388330141862324,
      "grad_norm": 0.14669372141361237,
      "learning_rate": 2.0425062103229366e-06,
      "loss": 0.4413,
      "mean_token_accuracy": 0.8592363558709621,
      "num_tokens": 1019444311.0,
      "step": 31942
    },
    {
      "entropy": 0.4392906394787133,
      "epoch": 2.9389250205576736,
      "grad_norm": 0.15395939350128174,
      "learning_rate": 2.039439384181311e-06,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8646626695990562,
      "num_tokens": 1019476168.0,
      "step": 31943
    },
    {
      "entropy": 0.37580857425928116,
      "epoch": 2.9390170269291147,
      "grad_norm": 0.13512523472309113,
      "learning_rate": 2.036372558039685e-06,
      "loss": 0.3622,
      "mean_token_accuracy": 0.8868941441178322,
      "num_tokens": 1019507832.0,
      "step": 31944
    },
    {
      "entropy": 0.4393664738163352,
      "epoch": 2.9391090333005563,
      "grad_norm": 0.14297644793987274,
      "learning_rate": 2.033305731898059e-06,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8616402931511402,
      "num_tokens": 1019540005.0,
      "step": 31945
    },
    {
      "entropy": 0.555123059079051,
      "epoch": 2.9392010396719974,
      "grad_norm": 0.15868264436721802,
      "learning_rate": 2.0302389057564327e-06,
      "loss": 0.5453,
      "mean_token_accuracy": 0.8288728147745132,
      "num_tokens": 1019572037.0,
      "step": 31946
    },
    {
      "entropy": 0.6390635315328836,
      "epoch": 2.9392930460434386,
      "grad_norm": 0.17135673761367798,
      "learning_rate": 2.027172079614807e-06,
      "loss": 0.6185,
      "mean_token_accuracy": 0.8048166930675507,
      "num_tokens": 1019603562.0,
      "step": 31947
    },
    {
      "entropy": 0.47182036750018597,
      "epoch": 2.9393850524148797,
      "grad_norm": 0.15016598999500275,
      "learning_rate": 2.0241052534731807e-06,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8520496226847172,
      "num_tokens": 1019636006.0,
      "step": 31948
    },
    {
      "entropy": 0.5663012843579054,
      "epoch": 2.939477058786321,
      "grad_norm": 0.16705691814422607,
      "learning_rate": 2.0210384273315545e-06,
      "loss": 0.553,
      "mean_token_accuracy": 0.8249349817633629,
      "num_tokens": 1019668753.0,
      "step": 31949
    },
    {
      "entropy": 0.5315374284982681,
      "epoch": 2.9395690651577624,
      "grad_norm": 0.1622197926044464,
      "learning_rate": 2.0179716011899283e-06,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8364822193980217,
      "num_tokens": 1019701153.0,
      "step": 31950
    },
    {
      "entropy": 0.5174348065629601,
      "epoch": 2.9396610715292035,
      "grad_norm": 0.1576225608587265,
      "learning_rate": 2.0149047750483025e-06,
      "loss": 0.5192,
      "mean_token_accuracy": 0.8375776000320911,
      "num_tokens": 1019733195.0,
      "step": 31951
    },
    {
      "entropy": 0.5265309382230043,
      "epoch": 2.9397530779006447,
      "grad_norm": 0.16323190927505493,
      "learning_rate": 2.0118379489066767e-06,
      "loss": 0.5075,
      "mean_token_accuracy": 0.8403994180262089,
      "num_tokens": 1019764925.0,
      "step": 31952
    },
    {
      "entropy": 0.5687125702388585,
      "epoch": 2.939845084272086,
      "grad_norm": 0.16301199793815613,
      "learning_rate": 2.0087711227650505e-06,
      "loss": 0.5611,
      "mean_token_accuracy": 0.8221835866570473,
      "num_tokens": 1019797419.0,
      "step": 31953
    },
    {
      "entropy": 0.485572159755975,
      "epoch": 2.939937090643527,
      "grad_norm": 0.15235209465026855,
      "learning_rate": 2.0057042966234243e-06,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8502679690718651,
      "num_tokens": 1019829351.0,
      "step": 31954
    },
    {
      "entropy": 0.5704564843326807,
      "epoch": 2.9400290970149685,
      "grad_norm": 0.15847253799438477,
      "learning_rate": 2.0026374704817986e-06,
      "loss": 0.5619,
      "mean_token_accuracy": 0.8246534019708633,
      "num_tokens": 1019861337.0,
      "step": 31955
    },
    {
      "entropy": 0.47834829427301884,
      "epoch": 2.9401211033864096,
      "grad_norm": 0.1490108221769333,
      "learning_rate": 1.9995706443401724e-06,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8494736477732658,
      "num_tokens": 1019893639.0,
      "step": 31956
    },
    {
      "entropy": 0.5128578385338187,
      "epoch": 2.9402131097578508,
      "grad_norm": 0.15852339565753937,
      "learning_rate": 1.996503818198546e-06,
      "loss": 0.5055,
      "mean_token_accuracy": 0.8419775888323784,
      "num_tokens": 1019925676.0,
      "step": 31957
    },
    {
      "entropy": 0.5101419743150473,
      "epoch": 2.940305116129292,
      "grad_norm": 0.16361133754253387,
      "learning_rate": 1.9934369920569204e-06,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8430198654532433,
      "num_tokens": 1019957634.0,
      "step": 31958
    },
    {
      "entropy": 0.5437237769365311,
      "epoch": 2.940397122500733,
      "grad_norm": 0.16524960100650787,
      "learning_rate": 1.9903701659152946e-06,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8364427946507931,
      "num_tokens": 1019989284.0,
      "step": 31959
    },
    {
      "entropy": 0.46125171054154634,
      "epoch": 2.9404891288721746,
      "grad_norm": 0.1491917073726654,
      "learning_rate": 1.9873033397736684e-06,
      "loss": 0.4623,
      "mean_token_accuracy": 0.8572357073426247,
      "num_tokens": 1020021741.0,
      "step": 31960
    },
    {
      "entropy": 0.4808612987399101,
      "epoch": 2.9405811352436158,
      "grad_norm": 0.15878628194332123,
      "learning_rate": 1.984236513632042e-06,
      "loss": 0.4747,
      "mean_token_accuracy": 0.8457922637462616,
      "num_tokens": 1020053859.0,
      "step": 31961
    },
    {
      "entropy": 0.5046213453169912,
      "epoch": 2.940673141615057,
      "grad_norm": 0.15946100652217865,
      "learning_rate": 1.9811696874904164e-06,
      "loss": 0.4822,
      "mean_token_accuracy": 0.8450268507003784,
      "num_tokens": 1020084468.0,
      "step": 31962
    },
    {
      "entropy": 0.5428722156211734,
      "epoch": 2.940765147986498,
      "grad_norm": 0.16222497820854187,
      "learning_rate": 1.9781028613487902e-06,
      "loss": 0.5249,
      "mean_token_accuracy": 0.8349394388496876,
      "num_tokens": 1020116809.0,
      "step": 31963
    },
    {
      "entropy": 0.5355948433279991,
      "epoch": 2.940857154357939,
      "grad_norm": 0.17156736552715302,
      "learning_rate": 1.975036035207164e-06,
      "loss": 0.5334,
      "mean_token_accuracy": 0.8311324566602707,
      "num_tokens": 1020147909.0,
      "step": 31964
    },
    {
      "entropy": 0.4891096008941531,
      "epoch": 2.9409491607293807,
      "grad_norm": 0.158516526222229,
      "learning_rate": 1.971969209065538e-06,
      "loss": 0.4909,
      "mean_token_accuracy": 0.8482609875500202,
      "num_tokens": 1020180131.0,
      "step": 31965
    },
    {
      "entropy": 0.5630268603563309,
      "epoch": 2.941041167100822,
      "grad_norm": 0.16060471534729004,
      "learning_rate": 1.968902382923912e-06,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8284397199749947,
      "num_tokens": 1020211488.0,
      "step": 31966
    },
    {
      "entropy": 0.5075676329433918,
      "epoch": 2.941133173472263,
      "grad_norm": 0.15860339999198914,
      "learning_rate": 1.9658355567822863e-06,
      "loss": 0.5144,
      "mean_token_accuracy": 0.8355743959546089,
      "num_tokens": 1020243736.0,
      "step": 31967
    },
    {
      "entropy": 0.5553919449448586,
      "epoch": 2.941225179843704,
      "grad_norm": 0.16270217299461365,
      "learning_rate": 1.96276873064066e-06,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8306823149323463,
      "num_tokens": 1020275869.0,
      "step": 31968
    },
    {
      "entropy": 0.6557868253439665,
      "epoch": 2.9413171862151453,
      "grad_norm": 0.17865192890167236,
      "learning_rate": 1.959701904499034e-06,
      "loss": 0.6478,
      "mean_token_accuracy": 0.7975081689655781,
      "num_tokens": 1020307267.0,
      "step": 31969
    },
    {
      "entropy": 0.5009624073281884,
      "epoch": 2.941409192586587,
      "grad_norm": 0.1575886458158493,
      "learning_rate": 1.956635078357408e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8460453525185585,
      "num_tokens": 1020338986.0,
      "step": 31970
    },
    {
      "entropy": 0.39392140367999673,
      "epoch": 2.941501198958028,
      "grad_norm": 0.1421598494052887,
      "learning_rate": 1.953568252215782e-06,
      "loss": 0.3811,
      "mean_token_accuracy": 0.8762484900653362,
      "num_tokens": 1020370772.0,
      "step": 31971
    },
    {
      "entropy": 0.590209168381989,
      "epoch": 2.941593205329469,
      "grad_norm": 0.16156364977359772,
      "learning_rate": 1.9505014260741557e-06,
      "loss": 0.5756,
      "mean_token_accuracy": 0.8216020986437798,
      "num_tokens": 1020403183.0,
      "step": 31972
    },
    {
      "entropy": 0.5084224883466959,
      "epoch": 2.9416852117009102,
      "grad_norm": 0.1538233608007431,
      "learning_rate": 1.94743459993253e-06,
      "loss": 0.5037,
      "mean_token_accuracy": 0.8386480025947094,
      "num_tokens": 1020435388.0,
      "step": 31973
    },
    {
      "entropy": 0.5619555814191699,
      "epoch": 2.9417772180723514,
      "grad_norm": 0.1634993702173233,
      "learning_rate": 1.944367773790904e-06,
      "loss": 0.5483,
      "mean_token_accuracy": 0.8277643732726574,
      "num_tokens": 1020467783.0,
      "step": 31974
    },
    {
      "entropy": 0.4370388473616913,
      "epoch": 2.941869224443793,
      "grad_norm": 0.14769479632377625,
      "learning_rate": 1.941300947649278e-06,
      "loss": 0.4221,
      "mean_token_accuracy": 0.8678931072354317,
      "num_tokens": 1020499967.0,
      "step": 31975
    },
    {
      "entropy": 0.4761411380022764,
      "epoch": 2.941961230815234,
      "grad_norm": 0.15274930000305176,
      "learning_rate": 1.9382341215076518e-06,
      "loss": 0.4585,
      "mean_token_accuracy": 0.8546558544039726,
      "num_tokens": 1020531482.0,
      "step": 31976
    },
    {
      "entropy": 0.45077611692249775,
      "epoch": 2.942053237186675,
      "grad_norm": 0.15243691205978394,
      "learning_rate": 1.935167295366026e-06,
      "loss": 0.4407,
      "mean_token_accuracy": 0.8592940531671047,
      "num_tokens": 1020562929.0,
      "step": 31977
    },
    {
      "entropy": 0.5961405634880066,
      "epoch": 2.9421452435581164,
      "grad_norm": 0.16715380549430847,
      "learning_rate": 1.9321004692243998e-06,
      "loss": 0.5832,
      "mean_token_accuracy": 0.8175336718559265,
      "num_tokens": 1020595498.0,
      "step": 31978
    },
    {
      "entropy": 0.492584609426558,
      "epoch": 2.9422372499295575,
      "grad_norm": 0.1549905389547348,
      "learning_rate": 1.9290336430827736e-06,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8467886745929718,
      "num_tokens": 1020627293.0,
      "step": 31979
    },
    {
      "entropy": 0.5295019363984466,
      "epoch": 2.942329256300999,
      "grad_norm": 0.162834033370018,
      "learning_rate": 1.925966816941148e-06,
      "loss": 0.5118,
      "mean_token_accuracy": 0.8392284773290157,
      "num_tokens": 1020658708.0,
      "step": 31980
    },
    {
      "entropy": 0.5036382814869285,
      "epoch": 2.94242126267244,
      "grad_norm": 0.15531177818775177,
      "learning_rate": 1.922899990799522e-06,
      "loss": 0.4951,
      "mean_token_accuracy": 0.8409178107976913,
      "num_tokens": 1020691001.0,
      "step": 31981
    },
    {
      "entropy": 0.4928086744621396,
      "epoch": 2.9425132690438813,
      "grad_norm": 0.1493474543094635,
      "learning_rate": 1.919833164657896e-06,
      "loss": 0.4651,
      "mean_token_accuracy": 0.8516578935086727,
      "num_tokens": 1020723272.0,
      "step": 31982
    },
    {
      "entropy": 0.5098895560950041,
      "epoch": 2.9426052754153225,
      "grad_norm": 0.1623993068933487,
      "learning_rate": 1.9167663385162696e-06,
      "loss": 0.4981,
      "mean_token_accuracy": 0.8463074490427971,
      "num_tokens": 1020754168.0,
      "step": 31983
    },
    {
      "entropy": 0.4811521675437689,
      "epoch": 2.9426972817867636,
      "grad_norm": 0.15236471593379974,
      "learning_rate": 1.9136995123746434e-06,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8505706414580345,
      "num_tokens": 1020786191.0,
      "step": 31984
    },
    {
      "entropy": 0.5725087560713291,
      "epoch": 2.942789288158205,
      "grad_norm": 0.16299721598625183,
      "learning_rate": 1.9106326862330177e-06,
      "loss": 0.5502,
      "mean_token_accuracy": 0.8266764096915722,
      "num_tokens": 1020817895.0,
      "step": 31985
    },
    {
      "entropy": 0.48213307186961174,
      "epoch": 2.9428812945296463,
      "grad_norm": 0.15550991892814636,
      "learning_rate": 1.9075658600913915e-06,
      "loss": 0.4676,
      "mean_token_accuracy": 0.8489363379776478,
      "num_tokens": 1020850078.0,
      "step": 31986
    },
    {
      "entropy": 0.5216829327400774,
      "epoch": 2.9429733009010874,
      "grad_norm": 0.16087330877780914,
      "learning_rate": 1.9044990339497655e-06,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8366078808903694,
      "num_tokens": 1020882233.0,
      "step": 31987
    },
    {
      "entropy": 0.520468776114285,
      "epoch": 2.9430653072725286,
      "grad_norm": 0.15558743476867676,
      "learning_rate": 1.9014322078081393e-06,
      "loss": 0.5037,
      "mean_token_accuracy": 0.8416167348623276,
      "num_tokens": 1020913961.0,
      "step": 31988
    },
    {
      "entropy": 0.46504425816237926,
      "epoch": 2.9431573136439697,
      "grad_norm": 0.1508774310350418,
      "learning_rate": 1.8983653816665135e-06,
      "loss": 0.457,
      "mean_token_accuracy": 0.8562387563288212,
      "num_tokens": 1020945840.0,
      "step": 31989
    },
    {
      "entropy": 0.4513892778195441,
      "epoch": 2.9432493200154113,
      "grad_norm": 0.15510691702365875,
      "learning_rate": 1.8952985555248875e-06,
      "loss": 0.4476,
      "mean_token_accuracy": 0.859651118516922,
      "num_tokens": 1020978199.0,
      "step": 31990
    },
    {
      "entropy": 0.5149519834667444,
      "epoch": 2.9433413263868524,
      "grad_norm": 0.15727512538433075,
      "learning_rate": 1.8922317293832613e-06,
      "loss": 0.5118,
      "mean_token_accuracy": 0.838647373020649,
      "num_tokens": 1021010305.0,
      "step": 31991
    },
    {
      "entropy": 0.6678859069943428,
      "epoch": 2.9434333327582936,
      "grad_norm": 0.18249724805355072,
      "learning_rate": 1.8891649032416351e-06,
      "loss": 0.6632,
      "mean_token_accuracy": 0.7918914370238781,
      "num_tokens": 1021041583.0,
      "step": 31992
    },
    {
      "entropy": 0.5164060877868906,
      "epoch": 2.9435253391297347,
      "grad_norm": 0.15865927934646606,
      "learning_rate": 1.8860980771000093e-06,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8421055041253567,
      "num_tokens": 1021072776.0,
      "step": 31993
    },
    {
      "entropy": 0.49164728727191687,
      "epoch": 2.943617345501176,
      "grad_norm": 0.15799832344055176,
      "learning_rate": 1.8830312509583833e-06,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8454901874065399,
      "num_tokens": 1021105422.0,
      "step": 31994
    },
    {
      "entropy": 0.520497826859355,
      "epoch": 2.9437093518726174,
      "grad_norm": 0.15764658153057098,
      "learning_rate": 1.8799644248167571e-06,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8395322486758232,
      "num_tokens": 1021136978.0,
      "step": 31995
    },
    {
      "entropy": 0.5600766744464636,
      "epoch": 2.9438013582440585,
      "grad_norm": 0.16577602922916412,
      "learning_rate": 1.876897598675131e-06,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8277389593422413,
      "num_tokens": 1021168660.0,
      "step": 31996
    },
    {
      "entropy": 0.5510672628879547,
      "epoch": 2.9438933646154997,
      "grad_norm": 0.16110859811306,
      "learning_rate": 1.8738307725335052e-06,
      "loss": 0.5352,
      "mean_token_accuracy": 0.8302560374140739,
      "num_tokens": 1021200000.0,
      "step": 31997
    },
    {
      "entropy": 0.44621425005607307,
      "epoch": 2.943985370986941,
      "grad_norm": 0.15109704434871674,
      "learning_rate": 1.8707639463918792e-06,
      "loss": 0.4287,
      "mean_token_accuracy": 0.8611067533493042,
      "num_tokens": 1021231353.0,
      "step": 31998
    },
    {
      "entropy": 0.49877991038374603,
      "epoch": 2.944077377358382,
      "grad_norm": 0.15681718289852142,
      "learning_rate": 1.867697120250253e-06,
      "loss": 0.482,
      "mean_token_accuracy": 0.8475156687200069,
      "num_tokens": 1021262914.0,
      "step": 31999
    },
    {
      "entropy": 0.471724275033921,
      "epoch": 2.9441693837298235,
      "grad_norm": 0.15491081774234772,
      "learning_rate": 1.8646302941086272e-06,
      "loss": 0.4694,
      "mean_token_accuracy": 0.8504035025835037,
      "num_tokens": 1021295206.0,
      "step": 32000
    },
    {
      "entropy": 0.47871334513183683,
      "epoch": 2.9442613901012646,
      "grad_norm": 0.15486633777618408,
      "learning_rate": 1.861563467967001e-06,
      "loss": 0.465,
      "mean_token_accuracy": 0.8489585779607296,
      "num_tokens": 1021326771.0,
      "step": 32001
    },
    {
      "entropy": 0.47682512551546097,
      "epoch": 2.9443533964727058,
      "grad_norm": 0.1561036854982376,
      "learning_rate": 1.858496641825375e-06,
      "loss": 0.4715,
      "mean_token_accuracy": 0.8515845462679863,
      "num_tokens": 1021358741.0,
      "step": 32002
    },
    {
      "entropy": 0.531024269759655,
      "epoch": 2.944445402844147,
      "grad_norm": 0.15916797518730164,
      "learning_rate": 1.8554298156837488e-06,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8359980694949627,
      "num_tokens": 1021390793.0,
      "step": 32003
    },
    {
      "entropy": 0.555805429816246,
      "epoch": 2.944537409215588,
      "grad_norm": 0.16437286138534546,
      "learning_rate": 1.852362989542123e-06,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8320870921015739,
      "num_tokens": 1021422565.0,
      "step": 32004
    },
    {
      "entropy": 0.5230866353958845,
      "epoch": 2.9446294155870296,
      "grad_norm": 0.16016480326652527,
      "learning_rate": 1.849296163400497e-06,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8357190638780594,
      "num_tokens": 1021454530.0,
      "step": 32005
    },
    {
      "entropy": 0.48805062659084797,
      "epoch": 2.9447214219584708,
      "grad_norm": 0.15109466016292572,
      "learning_rate": 1.8462293372588709e-06,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8486891128122807,
      "num_tokens": 1021486902.0,
      "step": 32006
    },
    {
      "entropy": 0.4402818668168038,
      "epoch": 2.944813428329912,
      "grad_norm": 0.14382225275039673,
      "learning_rate": 1.8431625111172447e-06,
      "loss": 0.4271,
      "mean_token_accuracy": 0.8639349341392517,
      "num_tokens": 1021518124.0,
      "step": 32007
    },
    {
      "entropy": 0.5210155099630356,
      "epoch": 2.944905434701353,
      "grad_norm": 0.1614706665277481,
      "learning_rate": 1.8400956849756189e-06,
      "loss": 0.5066,
      "mean_token_accuracy": 0.8380300812423229,
      "num_tokens": 1021550874.0,
      "step": 32008
    },
    {
      "entropy": 0.5141850588843226,
      "epoch": 2.944997441072794,
      "grad_norm": 0.1572600156068802,
      "learning_rate": 1.837028858833993e-06,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8398799784481525,
      "num_tokens": 1021583221.0,
      "step": 32009
    },
    {
      "entropy": 0.5633502686396241,
      "epoch": 2.9450894474442357,
      "grad_norm": 0.15806856751441956,
      "learning_rate": 1.8339620326923667e-06,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8274615332484245,
      "num_tokens": 1021614735.0,
      "step": 32010
    },
    {
      "entropy": 0.5825268477201462,
      "epoch": 2.945181453815677,
      "grad_norm": 0.1656578928232193,
      "learning_rate": 1.8308952065507405e-06,
      "loss": 0.5745,
      "mean_token_accuracy": 0.8162197098135948,
      "num_tokens": 1021646534.0,
      "step": 32011
    },
    {
      "entropy": 0.5827208627015352,
      "epoch": 2.945273460187118,
      "grad_norm": 0.16733285784721375,
      "learning_rate": 1.8278283804091147e-06,
      "loss": 0.5811,
      "mean_token_accuracy": 0.818036887794733,
      "num_tokens": 1021679145.0,
      "step": 32012
    },
    {
      "entropy": 0.5251020831055939,
      "epoch": 2.945365466558559,
      "grad_norm": 0.1566496193408966,
      "learning_rate": 1.8247615542674887e-06,
      "loss": 0.5029,
      "mean_token_accuracy": 0.8387703932821751,
      "num_tokens": 1021710551.0,
      "step": 32013
    },
    {
      "entropy": 0.4720872766338289,
      "epoch": 2.9454574729300003,
      "grad_norm": 0.14991794526576996,
      "learning_rate": 1.8216947281258625e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8523674942553043,
      "num_tokens": 1021743170.0,
      "step": 32014
    },
    {
      "entropy": 0.5709024844691157,
      "epoch": 2.945549479301442,
      "grad_norm": 0.16305890679359436,
      "learning_rate": 1.8186279019842363e-06,
      "loss": 0.5589,
      "mean_token_accuracy": 0.8253582194447517,
      "num_tokens": 1021775494.0,
      "step": 32015
    },
    {
      "entropy": 0.4091324951732531,
      "epoch": 2.945641485672883,
      "grad_norm": 0.1446717381477356,
      "learning_rate": 1.8155610758426106e-06,
      "loss": 0.3979,
      "mean_token_accuracy": 0.8702147603034973,
      "num_tokens": 1021807959.0,
      "step": 32016
    },
    {
      "entropy": 0.5715493466705084,
      "epoch": 2.945733492044324,
      "grad_norm": 0.16328151524066925,
      "learning_rate": 1.8124942497009846e-06,
      "loss": 0.5588,
      "mean_token_accuracy": 0.8211340233683586,
      "num_tokens": 1021840295.0,
      "step": 32017
    },
    {
      "entropy": 0.4816752327606082,
      "epoch": 2.9458254984157652,
      "grad_norm": 0.15810498595237732,
      "learning_rate": 1.8094274235593584e-06,
      "loss": 0.4618,
      "mean_token_accuracy": 0.8486754819750786,
      "num_tokens": 1021870928.0,
      "step": 32018
    },
    {
      "entropy": 0.41182148549705744,
      "epoch": 2.9459175047872064,
      "grad_norm": 0.14631237089633942,
      "learning_rate": 1.8063605974177326e-06,
      "loss": 0.4064,
      "mean_token_accuracy": 0.8727832362055779,
      "num_tokens": 1021902700.0,
      "step": 32019
    },
    {
      "entropy": 0.46645348239690065,
      "epoch": 2.946009511158648,
      "grad_norm": 0.15099821984767914,
      "learning_rate": 1.8032937712761066e-06,
      "loss": 0.4571,
      "mean_token_accuracy": 0.8538354374468327,
      "num_tokens": 1021934465.0,
      "step": 32020
    },
    {
      "entropy": 0.5275453981012106,
      "epoch": 2.946101517530089,
      "grad_norm": 0.1620856523513794,
      "learning_rate": 1.8002269451344804e-06,
      "loss": 0.5037,
      "mean_token_accuracy": 0.8402669727802277,
      "num_tokens": 1021965186.0,
      "step": 32021
    },
    {
      "entropy": 0.4796474170871079,
      "epoch": 2.94619352390153,
      "grad_norm": 0.15633587539196014,
      "learning_rate": 1.7971601189928542e-06,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8509690538048744,
      "num_tokens": 1021997225.0,
      "step": 32022
    },
    {
      "entropy": 0.5303644026862457,
      "epoch": 2.9462855302729714,
      "grad_norm": 0.1595885455608368,
      "learning_rate": 1.7940932928512284e-06,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8353297486901283,
      "num_tokens": 1022028743.0,
      "step": 32023
    },
    {
      "entropy": 0.47556094476021826,
      "epoch": 2.9463775366444125,
      "grad_norm": 0.1545715183019638,
      "learning_rate": 1.7910264667096024e-06,
      "loss": 0.4717,
      "mean_token_accuracy": 0.8512580506503582,
      "num_tokens": 1022061064.0,
      "step": 32024
    },
    {
      "entropy": 0.4590216185897589,
      "epoch": 2.946469543015854,
      "grad_norm": 0.15045039355754852,
      "learning_rate": 1.7879596405679762e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8550000563263893,
      "num_tokens": 1022092965.0,
      "step": 32025
    },
    {
      "entropy": 0.4432495441287756,
      "epoch": 2.946561549387295,
      "grad_norm": 0.1514224410057068,
      "learning_rate": 1.78489281442635e-06,
      "loss": 0.4298,
      "mean_token_accuracy": 0.863517876714468,
      "num_tokens": 1022124833.0,
      "step": 32026
    },
    {
      "entropy": 0.49868031160440296,
      "epoch": 2.9466535557587363,
      "grad_norm": 0.14945445954799652,
      "learning_rate": 1.7818259882847243e-06,
      "loss": 0.4758,
      "mean_token_accuracy": 0.8470295071601868,
      "num_tokens": 1022157165.0,
      "step": 32027
    },
    {
      "entropy": 0.5536439437419176,
      "epoch": 2.9467455621301775,
      "grad_norm": 0.16675163805484772,
      "learning_rate": 1.7787591621430983e-06,
      "loss": 0.5471,
      "mean_token_accuracy": 0.8291610926389694,
      "num_tokens": 1022188681.0,
      "step": 32028
    },
    {
      "entropy": 0.588651554659009,
      "epoch": 2.9468375685016186,
      "grad_norm": 0.1682382971048355,
      "learning_rate": 1.775692336001472e-06,
      "loss": 0.5701,
      "mean_token_accuracy": 0.8184392787516117,
      "num_tokens": 1022219899.0,
      "step": 32029
    },
    {
      "entropy": 0.4674591748043895,
      "epoch": 2.94692957487306,
      "grad_norm": 0.16140124201774597,
      "learning_rate": 1.772625509859846e-06,
      "loss": 0.4564,
      "mean_token_accuracy": 0.8553835712373257,
      "num_tokens": 1022251233.0,
      "step": 32030
    },
    {
      "entropy": 0.3981453189626336,
      "epoch": 2.9470215812445013,
      "grad_norm": 0.14344538748264313,
      "learning_rate": 1.7695586837182203e-06,
      "loss": 0.3825,
      "mean_token_accuracy": 0.873452939093113,
      "num_tokens": 1022282489.0,
      "step": 32031
    },
    {
      "entropy": 0.4871666565304622,
      "epoch": 2.9471135876159424,
      "grad_norm": 0.15472137928009033,
      "learning_rate": 1.7664918575765941e-06,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8514176942408085,
      "num_tokens": 1022314723.0,
      "step": 32032
    },
    {
      "entropy": 0.6001225216314197,
      "epoch": 2.9472055939873836,
      "grad_norm": 0.17099878191947937,
      "learning_rate": 1.763425031434968e-06,
      "loss": 0.5921,
      "mean_token_accuracy": 0.8140510469675064,
      "num_tokens": 1022346479.0,
      "step": 32033
    },
    {
      "entropy": 0.49765645805746317,
      "epoch": 2.9472976003588247,
      "grad_norm": 0.15389849245548248,
      "learning_rate": 1.760358205293342e-06,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8469386100769043,
      "num_tokens": 1022378822.0,
      "step": 32034
    },
    {
      "entropy": 0.5383893437683582,
      "epoch": 2.9473896067302663,
      "grad_norm": 0.1614699512720108,
      "learning_rate": 1.7572913791517162e-06,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8325440436601639,
      "num_tokens": 1022410670.0,
      "step": 32035
    },
    {
      "entropy": 0.5185810755938292,
      "epoch": 2.9474816131017074,
      "grad_norm": 0.15802454948425293,
      "learning_rate": 1.75422455301009e-06,
      "loss": 0.506,
      "mean_token_accuracy": 0.8381531424820423,
      "num_tokens": 1022442617.0,
      "step": 32036
    },
    {
      "entropy": 0.4612330002710223,
      "epoch": 2.9475736194731486,
      "grad_norm": 0.15022945404052734,
      "learning_rate": 1.7511577268684638e-06,
      "loss": 0.4491,
      "mean_token_accuracy": 0.854473914951086,
      "num_tokens": 1022474785.0,
      "step": 32037
    },
    {
      "entropy": 0.4516699503874406,
      "epoch": 2.9476656258445897,
      "grad_norm": 0.15046828985214233,
      "learning_rate": 1.748090900726838e-06,
      "loss": 0.4334,
      "mean_token_accuracy": 0.8622746057808399,
      "num_tokens": 1022506272.0,
      "step": 32038
    },
    {
      "entropy": 0.5059669008478522,
      "epoch": 2.947757632216031,
      "grad_norm": 0.15916675329208374,
      "learning_rate": 1.745024074585212e-06,
      "loss": 0.4915,
      "mean_token_accuracy": 0.8431802950799465,
      "num_tokens": 1022537192.0,
      "step": 32039
    },
    {
      "entropy": 0.5717412373051047,
      "epoch": 2.9478496385874724,
      "grad_norm": 0.16423852741718292,
      "learning_rate": 1.7419572484435858e-06,
      "loss": 0.5594,
      "mean_token_accuracy": 0.8223466835916042,
      "num_tokens": 1022568820.0,
      "step": 32040
    },
    {
      "entropy": 0.5304757783887908,
      "epoch": 2.9479416449589135,
      "grad_norm": 0.15705609321594238,
      "learning_rate": 1.7388904223019596e-06,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8345508649945259,
      "num_tokens": 1022601024.0,
      "step": 32041
    },
    {
      "entropy": 0.4774935278110206,
      "epoch": 2.9480336513303547,
      "grad_norm": 0.1533011794090271,
      "learning_rate": 1.7358235961603338e-06,
      "loss": 0.4626,
      "mean_token_accuracy": 0.8516473397612572,
      "num_tokens": 1022633385.0,
      "step": 32042
    },
    {
      "entropy": 0.5113905053585768,
      "epoch": 2.948125657701796,
      "grad_norm": 0.15856024622917175,
      "learning_rate": 1.7327567700187078e-06,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8422679416835308,
      "num_tokens": 1022665240.0,
      "step": 32043
    },
    {
      "entropy": 0.4200936835259199,
      "epoch": 2.948217664073237,
      "grad_norm": 0.14325745403766632,
      "learning_rate": 1.7296899438770816e-06,
      "loss": 0.4108,
      "mean_token_accuracy": 0.8672145158052444,
      "num_tokens": 1022697781.0,
      "step": 32044
    },
    {
      "entropy": 0.6066513620316982,
      "epoch": 2.9483096704446785,
      "grad_norm": 0.16882896423339844,
      "learning_rate": 1.7266231177354556e-06,
      "loss": 0.587,
      "mean_token_accuracy": 0.8111228197813034,
      "num_tokens": 1022729668.0,
      "step": 32045
    },
    {
      "entropy": 0.47235419414937496,
      "epoch": 2.9484016768161196,
      "grad_norm": 0.1562238186597824,
      "learning_rate": 1.7235562915938299e-06,
      "loss": 0.4663,
      "mean_token_accuracy": 0.8528483137488365,
      "num_tokens": 1022761391.0,
      "step": 32046
    },
    {
      "entropy": 0.5160524137318134,
      "epoch": 2.9484936831875608,
      "grad_norm": 0.1563263237476349,
      "learning_rate": 1.7204894654522037e-06,
      "loss": 0.501,
      "mean_token_accuracy": 0.8425595834851265,
      "num_tokens": 1022793679.0,
      "step": 32047
    },
    {
      "entropy": 0.5142077663913369,
      "epoch": 2.948585689559002,
      "grad_norm": 0.16063007712364197,
      "learning_rate": 1.7174226393105775e-06,
      "loss": 0.5,
      "mean_token_accuracy": 0.8394503332674503,
      "num_tokens": 1022826025.0,
      "step": 32048
    },
    {
      "entropy": 0.49256439693272114,
      "epoch": 2.948677695930443,
      "grad_norm": 0.16108223795890808,
      "learning_rate": 1.7143558131689515e-06,
      "loss": 0.4881,
      "mean_token_accuracy": 0.8416983895003796,
      "num_tokens": 1022857810.0,
      "step": 32049
    },
    {
      "entropy": 0.4533622907474637,
      "epoch": 2.9487697023018846,
      "grad_norm": 0.1495075821876526,
      "learning_rate": 1.7112889870273257e-06,
      "loss": 0.4442,
      "mean_token_accuracy": 0.8605195358395576,
      "num_tokens": 1022889725.0,
      "step": 32050
    },
    {
      "entropy": 0.517442811280489,
      "epoch": 2.9488617086733258,
      "grad_norm": 0.16024482250213623,
      "learning_rate": 1.7082221608856995e-06,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8437106572091579,
      "num_tokens": 1022920944.0,
      "step": 32051
    },
    {
      "entropy": 0.5585487530333921,
      "epoch": 2.948953715044767,
      "grad_norm": 0.16070464253425598,
      "learning_rate": 1.7051553347440733e-06,
      "loss": 0.5515,
      "mean_token_accuracy": 0.8249349594116211,
      "num_tokens": 1022953240.0,
      "step": 32052
    },
    {
      "entropy": 0.48702805768698454,
      "epoch": 2.949045721416208,
      "grad_norm": 0.1525048166513443,
      "learning_rate": 1.7020885086024473e-06,
      "loss": 0.4802,
      "mean_token_accuracy": 0.8481351062655449,
      "num_tokens": 1022985619.0,
      "step": 32053
    },
    {
      "entropy": 0.5730780400335789,
      "epoch": 2.949137727787649,
      "grad_norm": 0.16286219656467438,
      "learning_rate": 1.6990216824608215e-06,
      "loss": 0.5569,
      "mean_token_accuracy": 0.8267212994396687,
      "num_tokens": 1023017643.0,
      "step": 32054
    },
    {
      "entropy": 0.5428705539088696,
      "epoch": 2.9492297341590907,
      "grad_norm": 0.15559892356395721,
      "learning_rate": 1.6959548563191953e-06,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8352856636047363,
      "num_tokens": 1023049330.0,
      "step": 32055
    },
    {
      "entropy": 0.5044092936441302,
      "epoch": 2.949321740530532,
      "grad_norm": 0.15781475603580475,
      "learning_rate": 1.6928880301775691e-06,
      "loss": 0.4802,
      "mean_token_accuracy": 0.846137274056673,
      "num_tokens": 1023080669.0,
      "step": 32056
    },
    {
      "entropy": 0.55393385887146,
      "epoch": 2.949413746901973,
      "grad_norm": 0.16678392887115479,
      "learning_rate": 1.6898212040359432e-06,
      "loss": 0.5371,
      "mean_token_accuracy": 0.8269033096730709,
      "num_tokens": 1023112505.0,
      "step": 32057
    },
    {
      "entropy": 0.6327042561024427,
      "epoch": 2.949505753273414,
      "grad_norm": 0.17736920714378357,
      "learning_rate": 1.6867543778943174e-06,
      "loss": 0.6183,
      "mean_token_accuracy": 0.8035491928458214,
      "num_tokens": 1023144481.0,
      "step": 32058
    },
    {
      "entropy": 0.4713002562057227,
      "epoch": 2.9495977596448553,
      "grad_norm": 0.14944423735141754,
      "learning_rate": 1.6836875517526912e-06,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8533292748034,
      "num_tokens": 1023176790.0,
      "step": 32059
    },
    {
      "entropy": 0.5729109533131123,
      "epoch": 2.949689766016297,
      "grad_norm": 0.16674169898033142,
      "learning_rate": 1.6806207256110652e-06,
      "loss": 0.5602,
      "mean_token_accuracy": 0.8246070966124535,
      "num_tokens": 1023207710.0,
      "step": 32060
    },
    {
      "entropy": 0.5587168168276548,
      "epoch": 2.949781772387738,
      "grad_norm": 0.165485680103302,
      "learning_rate": 1.6775538994694394e-06,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8275048136711121,
      "num_tokens": 1023239324.0,
      "step": 32061
    },
    {
      "entropy": 0.4867154147941619,
      "epoch": 2.949873778759179,
      "grad_norm": 0.1481008380651474,
      "learning_rate": 1.6744870733278132e-06,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8509980849921703,
      "num_tokens": 1023271730.0,
      "step": 32062
    },
    {
      "entropy": 0.5063167167827487,
      "epoch": 2.9499657851306202,
      "grad_norm": 0.1655537635087967,
      "learning_rate": 1.671420247186187e-06,
      "loss": 0.4862,
      "mean_token_accuracy": 0.8431991934776306,
      "num_tokens": 1023302222.0,
      "step": 32063
    },
    {
      "entropy": 0.5466684214770794,
      "epoch": 2.9500577915020614,
      "grad_norm": 0.16203589737415314,
      "learning_rate": 1.668353421044561e-06,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8308196291327477,
      "num_tokens": 1023334025.0,
      "step": 32064
    },
    {
      "entropy": 0.4555301498621702,
      "epoch": 2.950149797873503,
      "grad_norm": 0.15245801210403442,
      "learning_rate": 1.6652865949029353e-06,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8598806038498878,
      "num_tokens": 1023366293.0,
      "step": 32065
    },
    {
      "entropy": 0.5543874781578779,
      "epoch": 2.950241804244944,
      "grad_norm": 0.16234268248081207,
      "learning_rate": 1.662219768761309e-06,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8288470059633255,
      "num_tokens": 1023398255.0,
      "step": 32066
    },
    {
      "entropy": 0.4060304887825623,
      "epoch": 2.950333810616385,
      "grad_norm": 0.14277814328670502,
      "learning_rate": 1.6591529426196829e-06,
      "loss": 0.3961,
      "mean_token_accuracy": 0.874453216791153,
      "num_tokens": 1023430352.0,
      "step": 32067
    },
    {
      "entropy": 0.4283428159542382,
      "epoch": 2.9504258169878264,
      "grad_norm": 0.14313577115535736,
      "learning_rate": 1.6560861164780569e-06,
      "loss": 0.4162,
      "mean_token_accuracy": 0.8661682307720184,
      "num_tokens": 1023462735.0,
      "step": 32068
    },
    {
      "entropy": 0.6277387626469135,
      "epoch": 2.9505178233592675,
      "grad_norm": 0.16729682683944702,
      "learning_rate": 1.653019290336431e-06,
      "loss": 0.6185,
      "mean_token_accuracy": 0.806989811360836,
      "num_tokens": 1023494943.0,
      "step": 32069
    },
    {
      "entropy": 0.5365962591022253,
      "epoch": 2.950609829730709,
      "grad_norm": 0.15940140187740326,
      "learning_rate": 1.6499524641948049e-06,
      "loss": 0.5317,
      "mean_token_accuracy": 0.8320334441959858,
      "num_tokens": 1023527648.0,
      "step": 32070
    },
    {
      "entropy": 0.5623654434457421,
      "epoch": 2.95070183610215,
      "grad_norm": 0.16232500970363617,
      "learning_rate": 1.6468856380531787e-06,
      "loss": 0.544,
      "mean_token_accuracy": 0.8257686831057072,
      "num_tokens": 1023560195.0,
      "step": 32071
    },
    {
      "entropy": 0.49811084661632776,
      "epoch": 2.9507938424735913,
      "grad_norm": 0.15430380403995514,
      "learning_rate": 1.6438188119115527e-06,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8491408787667751,
      "num_tokens": 1023591837.0,
      "step": 32072
    },
    {
      "entropy": 0.46247350983321667,
      "epoch": 2.9508858488450325,
      "grad_norm": 0.15126203000545502,
      "learning_rate": 1.640751985769927e-06,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8544009476900101,
      "num_tokens": 1023624343.0,
      "step": 32073
    },
    {
      "entropy": 0.50666319206357,
      "epoch": 2.9509778552164736,
      "grad_norm": 0.1611812561750412,
      "learning_rate": 1.6376851596283007e-06,
      "loss": 0.5057,
      "mean_token_accuracy": 0.8406178541481495,
      "num_tokens": 1023655981.0,
      "step": 32074
    },
    {
      "entropy": 0.4416622156277299,
      "epoch": 2.951069861587915,
      "grad_norm": 0.15430979430675507,
      "learning_rate": 1.6346183334866747e-06,
      "loss": 0.432,
      "mean_token_accuracy": 0.8593244440853596,
      "num_tokens": 1023687797.0,
      "step": 32075
    },
    {
      "entropy": 0.4395762253552675,
      "epoch": 2.9511618679593563,
      "grad_norm": 0.14907792210578918,
      "learning_rate": 1.6315515073450485e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8666926883161068,
      "num_tokens": 1023719899.0,
      "step": 32076
    },
    {
      "entropy": 0.5043083671480417,
      "epoch": 2.9512538743307974,
      "grad_norm": 0.1570325642824173,
      "learning_rate": 1.6284846812034228e-06,
      "loss": 0.4979,
      "mean_token_accuracy": 0.8421953395009041,
      "num_tokens": 1023751864.0,
      "step": 32077
    },
    {
      "entropy": 0.5467188712209463,
      "epoch": 2.9513458807022386,
      "grad_norm": 0.1710360199213028,
      "learning_rate": 1.6254178550617966e-06,
      "loss": 0.5409,
      "mean_token_accuracy": 0.82618722692132,
      "num_tokens": 1023782466.0,
      "step": 32078
    },
    {
      "entropy": 0.411650444380939,
      "epoch": 2.9514378870736797,
      "grad_norm": 0.14429031312465668,
      "learning_rate": 1.6223510289201706e-06,
      "loss": 0.3994,
      "mean_token_accuracy": 0.8703246675431728,
      "num_tokens": 1023814982.0,
      "step": 32079
    },
    {
      "entropy": 0.5142369475215673,
      "epoch": 2.9515298934451213,
      "grad_norm": 0.15606611967086792,
      "learning_rate": 1.6192842027785448e-06,
      "loss": 0.5048,
      "mean_token_accuracy": 0.8437004238367081,
      "num_tokens": 1023847509.0,
      "step": 32080
    },
    {
      "entropy": 0.475602216552943,
      "epoch": 2.9516218998165624,
      "grad_norm": 0.15950143337249756,
      "learning_rate": 1.6162173766369186e-06,
      "loss": 0.4586,
      "mean_token_accuracy": 0.8483615107834339,
      "num_tokens": 1023878426.0,
      "step": 32081
    },
    {
      "entropy": 0.5056816097348928,
      "epoch": 2.9517139061880036,
      "grad_norm": 0.1620495766401291,
      "learning_rate": 1.6131505504952924e-06,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8419463373720646,
      "num_tokens": 1023909878.0,
      "step": 32082
    },
    {
      "entropy": 0.549053393304348,
      "epoch": 2.9518059125594447,
      "grad_norm": 0.1627330780029297,
      "learning_rate": 1.6100837243536664e-06,
      "loss": 0.5355,
      "mean_token_accuracy": 0.8285077624022961,
      "num_tokens": 1023941796.0,
      "step": 32083
    },
    {
      "entropy": 0.5110219726338983,
      "epoch": 2.951897918930886,
      "grad_norm": 0.16155202686786652,
      "learning_rate": 1.6070168982120406e-06,
      "loss": 0.4967,
      "mean_token_accuracy": 0.8417469039559364,
      "num_tokens": 1023973450.0,
      "step": 32084
    },
    {
      "entropy": 0.47363340063020587,
      "epoch": 2.9519899253023274,
      "grad_norm": 0.15047572553157806,
      "learning_rate": 1.6039500720704144e-06,
      "loss": 0.4625,
      "mean_token_accuracy": 0.8521812669932842,
      "num_tokens": 1024006054.0,
      "step": 32085
    },
    {
      "entropy": 0.5091374581679702,
      "epoch": 2.9520819316737685,
      "grad_norm": 0.15549054741859436,
      "learning_rate": 1.6008832459287882e-06,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8398714624345303,
      "num_tokens": 1024038740.0,
      "step": 32086
    },
    {
      "entropy": 0.51521797478199,
      "epoch": 2.9521739380452097,
      "grad_norm": 0.15543623268604279,
      "learning_rate": 1.5978164197871623e-06,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8380301147699356,
      "num_tokens": 1024070391.0,
      "step": 32087
    },
    {
      "entropy": 0.42569743073545396,
      "epoch": 2.952265944416651,
      "grad_norm": 0.14212332665920258,
      "learning_rate": 1.5947495936455365e-06,
      "loss": 0.4076,
      "mean_token_accuracy": 0.8679503574967384,
      "num_tokens": 1024102722.0,
      "step": 32088
    },
    {
      "entropy": 0.45321388076990843,
      "epoch": 2.952357950788092,
      "grad_norm": 0.15107361972332,
      "learning_rate": 1.5916827675039103e-06,
      "loss": 0.4436,
      "mean_token_accuracy": 0.8549899831414223,
      "num_tokens": 1024135394.0,
      "step": 32089
    },
    {
      "entropy": 0.5920851714909077,
      "epoch": 2.9524499571595335,
      "grad_norm": 0.16591697931289673,
      "learning_rate": 1.5886159413622843e-06,
      "loss": 0.5773,
      "mean_token_accuracy": 0.8171552792191505,
      "num_tokens": 1024167413.0,
      "step": 32090
    },
    {
      "entropy": 0.5401100683957338,
      "epoch": 2.9525419635309746,
      "grad_norm": 0.16052784025669098,
      "learning_rate": 1.585549115220658e-06,
      "loss": 0.5293,
      "mean_token_accuracy": 0.8358100913465023,
      "num_tokens": 1024199253.0,
      "step": 32091
    },
    {
      "entropy": 0.46808263566344976,
      "epoch": 2.952633969902416,
      "grad_norm": 0.15246528387069702,
      "learning_rate": 1.5824822890790323e-06,
      "loss": 0.4553,
      "mean_token_accuracy": 0.8543905504047871,
      "num_tokens": 1024231273.0,
      "step": 32092
    },
    {
      "entropy": 0.46675979532301426,
      "epoch": 2.952725976273857,
      "grad_norm": 0.15517331659793854,
      "learning_rate": 1.5794154629374061e-06,
      "loss": 0.4575,
      "mean_token_accuracy": 0.8543650060892105,
      "num_tokens": 1024263539.0,
      "step": 32093
    },
    {
      "entropy": 0.5527291148900986,
      "epoch": 2.952817982645298,
      "grad_norm": 0.1664735972881317,
      "learning_rate": 1.5763486367957801e-06,
      "loss": 0.5415,
      "mean_token_accuracy": 0.8291473910212517,
      "num_tokens": 1024295738.0,
      "step": 32094
    },
    {
      "entropy": 0.6268958374857903,
      "epoch": 2.9529099890167396,
      "grad_norm": 0.17405569553375244,
      "learning_rate": 1.573281810654154e-06,
      "loss": 0.6087,
      "mean_token_accuracy": 0.8101629503071308,
      "num_tokens": 1024327151.0,
      "step": 32095
    },
    {
      "entropy": 0.5489833950996399,
      "epoch": 2.9530019953881808,
      "grad_norm": 0.1620088517665863,
      "learning_rate": 1.5702149845125282e-06,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8278182148933411,
      "num_tokens": 1024358648.0,
      "step": 32096
    },
    {
      "entropy": 0.4869064884260297,
      "epoch": 2.953094001759622,
      "grad_norm": 0.15508903563022614,
      "learning_rate": 1.567148158370902e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8492361716926098,
      "num_tokens": 1024389879.0,
      "step": 32097
    },
    {
      "entropy": 0.4064714387059212,
      "epoch": 2.953186008131063,
      "grad_norm": 0.1470608115196228,
      "learning_rate": 1.564081332229276e-06,
      "loss": 0.3963,
      "mean_token_accuracy": 0.8713431730866432,
      "num_tokens": 1024421944.0,
      "step": 32098
    },
    {
      "entropy": 0.4560503140091896,
      "epoch": 2.953278014502504,
      "grad_norm": 0.14830683171749115,
      "learning_rate": 1.56101450608765e-06,
      "loss": 0.4467,
      "mean_token_accuracy": 0.8549391441047192,
      "num_tokens": 1024453978.0,
      "step": 32099
    },
    {
      "entropy": 0.4978223480284214,
      "epoch": 2.9533700208739457,
      "grad_norm": 0.19851292669773102,
      "learning_rate": 1.557947679946024e-06,
      "loss": 0.496,
      "mean_token_accuracy": 0.8390278294682503,
      "num_tokens": 1024485380.0,
      "step": 32100
    },
    {
      "entropy": 0.38617671001702547,
      "epoch": 2.953462027245387,
      "grad_norm": 0.13890519738197327,
      "learning_rate": 1.5548808538043978e-06,
      "loss": 0.3746,
      "mean_token_accuracy": 0.8811933882534504,
      "num_tokens": 1024517724.0,
      "step": 32101
    },
    {
      "entropy": 0.5296175442636013,
      "epoch": 2.953554033616828,
      "grad_norm": 0.16449956595897675,
      "learning_rate": 1.551814027662772e-06,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8318645618855953,
      "num_tokens": 1024549132.0,
      "step": 32102
    },
    {
      "entropy": 0.5238952077925205,
      "epoch": 2.953646039988269,
      "grad_norm": 0.15745387971401215,
      "learning_rate": 1.5487472015211458e-06,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8366127163171768,
      "num_tokens": 1024581440.0,
      "step": 32103
    },
    {
      "entropy": 0.5547643639147282,
      "epoch": 2.9537380463597103,
      "grad_norm": 0.16425205767154694,
      "learning_rate": 1.5456803753795198e-06,
      "loss": 0.5569,
      "mean_token_accuracy": 0.8248942270874977,
      "num_tokens": 1024613239.0,
      "step": 32104
    },
    {
      "entropy": 0.4981997311115265,
      "epoch": 2.953830052731152,
      "grad_norm": 0.16166137158870697,
      "learning_rate": 1.5426135492378938e-06,
      "loss": 0.4929,
      "mean_token_accuracy": 0.8433126918971539,
      "num_tokens": 1024645345.0,
      "step": 32105
    },
    {
      "entropy": 0.5229432266205549,
      "epoch": 2.953922059102593,
      "grad_norm": 0.16673579812049866,
      "learning_rate": 1.5395467230962679e-06,
      "loss": 0.509,
      "mean_token_accuracy": 0.8352012783288956,
      "num_tokens": 1024676462.0,
      "step": 32106
    },
    {
      "entropy": 0.5540642724372447,
      "epoch": 2.954014065474034,
      "grad_norm": 0.16453394293785095,
      "learning_rate": 1.5364798969546417e-06,
      "loss": 0.539,
      "mean_token_accuracy": 0.8298035971820354,
      "num_tokens": 1024707480.0,
      "step": 32107
    },
    {
      "entropy": 0.4991475101560354,
      "epoch": 2.9541060718454752,
      "grad_norm": 0.15959177911281586,
      "learning_rate": 1.5334130708130157e-06,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8409345857799053,
      "num_tokens": 1024739795.0,
      "step": 32108
    },
    {
      "entropy": 0.41893932421226054,
      "epoch": 2.9541980782169164,
      "grad_norm": 0.1526029407978058,
      "learning_rate": 1.5303462446713897e-06,
      "loss": 0.4077,
      "mean_token_accuracy": 0.8669273555278778,
      "num_tokens": 1024771288.0,
      "step": 32109
    },
    {
      "entropy": 0.31831643206533045,
      "epoch": 2.954290084588358,
      "grad_norm": 0.13009651005268097,
      "learning_rate": 1.5272794185297637e-06,
      "loss": 0.3114,
      "mean_token_accuracy": 0.9010227583348751,
      "num_tokens": 1024803029.0,
      "step": 32110
    },
    {
      "entropy": 0.49491268023848534,
      "epoch": 2.954382090959799,
      "grad_norm": 0.15674211084842682,
      "learning_rate": 1.5242125923881375e-06,
      "loss": 0.4871,
      "mean_token_accuracy": 0.8462946861982346,
      "num_tokens": 1024834607.0,
      "step": 32111
    },
    {
      "entropy": 0.4987502368167043,
      "epoch": 2.9544740973312402,
      "grad_norm": 0.16448472440242767,
      "learning_rate": 1.5211457662465115e-06,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8410846665501595,
      "num_tokens": 1024866723.0,
      "step": 32112
    },
    {
      "entropy": 0.5324021484702826,
      "epoch": 2.9545661037026814,
      "grad_norm": 0.16374778747558594,
      "learning_rate": 1.5180789401048855e-06,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8359346762299538,
      "num_tokens": 1024898907.0,
      "step": 32113
    },
    {
      "entropy": 0.4280213990714401,
      "epoch": 2.9546581100741225,
      "grad_norm": 0.14140717685222626,
      "learning_rate": 1.5150121139632595e-06,
      "loss": 0.4144,
      "mean_token_accuracy": 0.8658622428774834,
      "num_tokens": 1024931641.0,
      "step": 32114
    },
    {
      "entropy": 0.4846067032776773,
      "epoch": 2.954750116445564,
      "grad_norm": 0.15348568558692932,
      "learning_rate": 1.5119452878216333e-06,
      "loss": 0.472,
      "mean_token_accuracy": 0.8525438793003559,
      "num_tokens": 1024963725.0,
      "step": 32115
    },
    {
      "entropy": 0.4926098622381687,
      "epoch": 2.954842122817005,
      "grad_norm": 0.16045933961868286,
      "learning_rate": 1.5088784616800076e-06,
      "loss": 0.4784,
      "mean_token_accuracy": 0.8464617766439915,
      "num_tokens": 1024995556.0,
      "step": 32116
    },
    {
      "entropy": 0.49587690585758537,
      "epoch": 2.9549341291884463,
      "grad_norm": 0.1569909006357193,
      "learning_rate": 1.5058116355383814e-06,
      "loss": 0.4828,
      "mean_token_accuracy": 0.8449804075062275,
      "num_tokens": 1025027804.0,
      "step": 32117
    },
    {
      "entropy": 0.486534146592021,
      "epoch": 2.9550261355598875,
      "grad_norm": 0.15565821528434753,
      "learning_rate": 1.5027448093967554e-06,
      "loss": 0.4771,
      "mean_token_accuracy": 0.8471653833985329,
      "num_tokens": 1025060153.0,
      "step": 32118
    },
    {
      "entropy": 0.507850406691432,
      "epoch": 2.9551181419313286,
      "grad_norm": 0.15904587507247925,
      "learning_rate": 1.4996779832551292e-06,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8408239558339119,
      "num_tokens": 1025092593.0,
      "step": 32119
    },
    {
      "entropy": 0.5162360183894634,
      "epoch": 2.95521014830277,
      "grad_norm": 0.1561402827501297,
      "learning_rate": 1.4966111571135034e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.8378500267863274,
      "num_tokens": 1025124061.0,
      "step": 32120
    },
    {
      "entropy": 0.49067461397498846,
      "epoch": 2.9553021546742113,
      "grad_norm": 0.15528561174869537,
      "learning_rate": 1.4935443309718774e-06,
      "loss": 0.4813,
      "mean_token_accuracy": 0.8463025167584419,
      "num_tokens": 1025156048.0,
      "step": 32121
    },
    {
      "entropy": 0.5012251646257937,
      "epoch": 2.9553941610456524,
      "grad_norm": 0.15054619312286377,
      "learning_rate": 1.4904775048302512e-06,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8467920757830143,
      "num_tokens": 1025187984.0,
      "step": 32122
    },
    {
      "entropy": 0.5082824714481831,
      "epoch": 2.9554861674170936,
      "grad_norm": 0.15580356121063232,
      "learning_rate": 1.4874106786886252e-06,
      "loss": 0.4893,
      "mean_token_accuracy": 0.8446563184261322,
      "num_tokens": 1025220630.0,
      "step": 32123
    },
    {
      "entropy": 0.529580669477582,
      "epoch": 2.9555781737885347,
      "grad_norm": 0.1584487408399582,
      "learning_rate": 1.4843438525469992e-06,
      "loss": 0.525,
      "mean_token_accuracy": 0.8312206044793129,
      "num_tokens": 1025253282.0,
      "step": 32124
    },
    {
      "entropy": 0.46369996992871165,
      "epoch": 2.9556701801599763,
      "grad_norm": 0.1482163667678833,
      "learning_rate": 1.4812770264053732e-06,
      "loss": 0.4274,
      "mean_token_accuracy": 0.8585988469421864,
      "num_tokens": 1025284459.0,
      "step": 32125
    },
    {
      "entropy": 0.5839184671640396,
      "epoch": 2.9557621865314174,
      "grad_norm": 0.17080338299274445,
      "learning_rate": 1.478210200263747e-06,
      "loss": 0.5753,
      "mean_token_accuracy": 0.8179742060601711,
      "num_tokens": 1025315842.0,
      "step": 32126
    },
    {
      "entropy": 0.5567651093006134,
      "epoch": 2.9558541929028586,
      "grad_norm": 0.16160748898983002,
      "learning_rate": 1.475143374122121e-06,
      "loss": 0.5414,
      "mean_token_accuracy": 0.8266993910074234,
      "num_tokens": 1025347347.0,
      "step": 32127
    },
    {
      "entropy": 0.4437043946236372,
      "epoch": 2.9559461992742997,
      "grad_norm": 0.15042132139205933,
      "learning_rate": 1.472076547980495e-06,
      "loss": 0.4302,
      "mean_token_accuracy": 0.863942239433527,
      "num_tokens": 1025379538.0,
      "step": 32128
    },
    {
      "entropy": 0.47782915737479925,
      "epoch": 2.956038205645741,
      "grad_norm": 0.14957338571548462,
      "learning_rate": 1.469009721838869e-06,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8497806899249554,
      "num_tokens": 1025411850.0,
      "step": 32129
    },
    {
      "entropy": 0.5127208782359958,
      "epoch": 2.9561302120171824,
      "grad_norm": 0.16234825551509857,
      "learning_rate": 1.4659428956972429e-06,
      "loss": 0.5003,
      "mean_token_accuracy": 0.840380672365427,
      "num_tokens": 1025444040.0,
      "step": 32130
    },
    {
      "entropy": 0.48038568440824747,
      "epoch": 2.9562222183886235,
      "grad_norm": 0.15291796624660492,
      "learning_rate": 1.462876069555617e-06,
      "loss": 0.4616,
      "mean_token_accuracy": 0.8494956009089947,
      "num_tokens": 1025476310.0,
      "step": 32131
    },
    {
      "entropy": 0.5410595064749941,
      "epoch": 2.9563142247600647,
      "grad_norm": 0.1622391790151596,
      "learning_rate": 1.459809243413991e-06,
      "loss": 0.5325,
      "mean_token_accuracy": 0.833068460226059,
      "num_tokens": 1025508825.0,
      "step": 32132
    },
    {
      "entropy": 0.4125160709954798,
      "epoch": 2.956406231131506,
      "grad_norm": 0.14752492308616638,
      "learning_rate": 1.456742417272365e-06,
      "loss": 0.399,
      "mean_token_accuracy": 0.8700212202966213,
      "num_tokens": 1025540209.0,
      "step": 32133
    },
    {
      "entropy": 0.5455748019739985,
      "epoch": 2.956498237502947,
      "grad_norm": 0.16009493172168732,
      "learning_rate": 1.4536755911307387e-06,
      "loss": 0.5375,
      "mean_token_accuracy": 0.832282941788435,
      "num_tokens": 1025572919.0,
      "step": 32134
    },
    {
      "entropy": 0.552881333976984,
      "epoch": 2.9565902438743885,
      "grad_norm": 0.16032804548740387,
      "learning_rate": 1.450608764989113e-06,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8295330256223679,
      "num_tokens": 1025604631.0,
      "step": 32135
    },
    {
      "entropy": 0.5388886230066419,
      "epoch": 2.9566822502458296,
      "grad_norm": 0.16372396051883698,
      "learning_rate": 1.4475419388474867e-06,
      "loss": 0.5254,
      "mean_token_accuracy": 0.8338641971349716,
      "num_tokens": 1025635658.0,
      "step": 32136
    },
    {
      "entropy": 0.4884359007701278,
      "epoch": 2.956774256617271,
      "grad_norm": 0.15563179552555084,
      "learning_rate": 1.4444751127058608e-06,
      "loss": 0.4833,
      "mean_token_accuracy": 0.8475779481232166,
      "num_tokens": 1025667474.0,
      "step": 32137
    },
    {
      "entropy": 0.4584669219329953,
      "epoch": 2.956866262988712,
      "grad_norm": 0.14840856194496155,
      "learning_rate": 1.4414082865642348e-06,
      "loss": 0.4378,
      "mean_token_accuracy": 0.8594172075390816,
      "num_tokens": 1025699306.0,
      "step": 32138
    },
    {
      "entropy": 0.498941233381629,
      "epoch": 2.956958269360153,
      "grad_norm": 0.15657860040664673,
      "learning_rate": 1.4383414604226088e-06,
      "loss": 0.4931,
      "mean_token_accuracy": 0.843257799744606,
      "num_tokens": 1025731383.0,
      "step": 32139
    },
    {
      "entropy": 0.46949294907972217,
      "epoch": 2.9570502757315946,
      "grad_norm": 0.1514056921005249,
      "learning_rate": 1.4352746342809828e-06,
      "loss": 0.4558,
      "mean_token_accuracy": 0.855690211057663,
      "num_tokens": 1025763639.0,
      "step": 32140
    },
    {
      "entropy": 0.5381053145974874,
      "epoch": 2.9571422821030358,
      "grad_norm": 0.15675802528858185,
      "learning_rate": 1.4322078081393566e-06,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8333076201379299,
      "num_tokens": 1025795543.0,
      "step": 32141
    },
    {
      "entropy": 0.49452594155445695,
      "epoch": 2.957234288474477,
      "grad_norm": 0.15388061106204987,
      "learning_rate": 1.4291409819977306e-06,
      "loss": 0.4866,
      "mean_token_accuracy": 0.8447450511157513,
      "num_tokens": 1025827679.0,
      "step": 32142
    },
    {
      "entropy": 0.5159891527146101,
      "epoch": 2.957326294845918,
      "grad_norm": 0.15855970978736877,
      "learning_rate": 1.4260741558561046e-06,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8419123664498329,
      "num_tokens": 1025859031.0,
      "step": 32143
    },
    {
      "entropy": 0.45709094731137156,
      "epoch": 2.957418301217359,
      "grad_norm": 0.1489274650812149,
      "learning_rate": 1.4230073297144786e-06,
      "loss": 0.4443,
      "mean_token_accuracy": 0.8564713001251221,
      "num_tokens": 1025891088.0,
      "step": 32144
    },
    {
      "entropy": 0.3917133458890021,
      "epoch": 2.9575103075888007,
      "grad_norm": 0.1433103382587433,
      "learning_rate": 1.4199405035728524e-06,
      "loss": 0.3814,
      "mean_token_accuracy": 0.8775855824351311,
      "num_tokens": 1025923210.0,
      "step": 32145
    },
    {
      "entropy": 0.47424285113811493,
      "epoch": 2.957602313960242,
      "grad_norm": 0.15345770120620728,
      "learning_rate": 1.4168736774312266e-06,
      "loss": 0.4612,
      "mean_token_accuracy": 0.8496404588222504,
      "num_tokens": 1025955528.0,
      "step": 32146
    },
    {
      "entropy": 0.5740520413964987,
      "epoch": 2.957694320331683,
      "grad_norm": 0.16449284553527832,
      "learning_rate": 1.4138068512896004e-06,
      "loss": 0.561,
      "mean_token_accuracy": 0.8213963061571121,
      "num_tokens": 1025987578.0,
      "step": 32147
    },
    {
      "entropy": 0.6640539206564426,
      "epoch": 2.957786326703124,
      "grad_norm": 0.17401687800884247,
      "learning_rate": 1.4107400251479745e-06,
      "loss": 0.6683,
      "mean_token_accuracy": 0.7948333211243153,
      "num_tokens": 1026019496.0,
      "step": 32148
    },
    {
      "entropy": 0.5027279881760478,
      "epoch": 2.9578783330745653,
      "grad_norm": 0.15286917984485626,
      "learning_rate": 1.4076731990063483e-06,
      "loss": 0.4889,
      "mean_token_accuracy": 0.845367219299078,
      "num_tokens": 1026051307.0,
      "step": 32149
    },
    {
      "entropy": 0.4867476299405098,
      "epoch": 2.957970339446007,
      "grad_norm": 0.15362298488616943,
      "learning_rate": 1.4046063728647225e-06,
      "loss": 0.4794,
      "mean_token_accuracy": 0.8464484103024006,
      "num_tokens": 1026083611.0,
      "step": 32150
    },
    {
      "entropy": 0.5108291778014973,
      "epoch": 2.958062345817448,
      "grad_norm": 0.15658453106880188,
      "learning_rate": 1.4015395467230963e-06,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8464674465358257,
      "num_tokens": 1026115447.0,
      "step": 32151
    },
    {
      "entropy": 0.46302385069429874,
      "epoch": 2.958154352188889,
      "grad_norm": 0.15076103806495667,
      "learning_rate": 1.3984727205814703e-06,
      "loss": 0.4562,
      "mean_token_accuracy": 0.8539990186691284,
      "num_tokens": 1026147147.0,
      "step": 32152
    },
    {
      "entropy": 0.5379674723371863,
      "epoch": 2.9582463585603302,
      "grad_norm": 0.16108600795269012,
      "learning_rate": 1.3954058944398443e-06,
      "loss": 0.5197,
      "mean_token_accuracy": 0.832439485937357,
      "num_tokens": 1026179393.0,
      "step": 32153
    },
    {
      "entropy": 0.43599938321858644,
      "epoch": 2.9583383649317714,
      "grad_norm": 0.1502433717250824,
      "learning_rate": 1.3923390682982183e-06,
      "loss": 0.4246,
      "mean_token_accuracy": 0.8616433888673782,
      "num_tokens": 1026210807.0,
      "step": 32154
    },
    {
      "entropy": 0.5898248217999935,
      "epoch": 2.958430371303213,
      "grad_norm": 0.16890326142311096,
      "learning_rate": 1.3892722421565921e-06,
      "loss": 0.5649,
      "mean_token_accuracy": 0.822101179510355,
      "num_tokens": 1026241096.0,
      "step": 32155
    },
    {
      "entropy": 0.5304880636977032,
      "epoch": 2.958522377674654,
      "grad_norm": 0.16140949726104736,
      "learning_rate": 1.3862054160149661e-06,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8374636620283127,
      "num_tokens": 1026272198.0,
      "step": 32156
    },
    {
      "entropy": 0.5033831093460321,
      "epoch": 2.9586143840460952,
      "grad_norm": 0.15634848177433014,
      "learning_rate": 1.3831385898733401e-06,
      "loss": 0.4905,
      "mean_token_accuracy": 0.8439279869198799,
      "num_tokens": 1026303812.0,
      "step": 32157
    },
    {
      "entropy": 0.4949455726891756,
      "epoch": 2.9587063904175364,
      "grad_norm": 0.1596129685640335,
      "learning_rate": 1.3800717637317142e-06,
      "loss": 0.4816,
      "mean_token_accuracy": 0.845242328941822,
      "num_tokens": 1026335848.0,
      "step": 32158
    },
    {
      "entropy": 0.49176677456125617,
      "epoch": 2.9587983967889775,
      "grad_norm": 0.15386638045310974,
      "learning_rate": 1.377004937590088e-06,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8487289696931839,
      "num_tokens": 1026368305.0,
      "step": 32159
    },
    {
      "entropy": 0.3728747391141951,
      "epoch": 2.958890403160419,
      "grad_norm": 0.13231050968170166,
      "learning_rate": 1.373938111448462e-06,
      "loss": 0.3478,
      "mean_token_accuracy": 0.8851602040231228,
      "num_tokens": 1026399463.0,
      "step": 32160
    },
    {
      "entropy": 0.5105081875808537,
      "epoch": 2.95898240953186,
      "grad_norm": 0.1584860384464264,
      "learning_rate": 1.3708712853068362e-06,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8407845683395863,
      "num_tokens": 1026431534.0,
      "step": 32161
    },
    {
      "entropy": 0.4713140120729804,
      "epoch": 2.9590744159033013,
      "grad_norm": 0.15292374789714813,
      "learning_rate": 1.36780445916521e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8540832027792931,
      "num_tokens": 1026463727.0,
      "step": 32162
    },
    {
      "entropy": 0.4621247057802975,
      "epoch": 2.9591664222747425,
      "grad_norm": 0.14858265221118927,
      "learning_rate": 1.364737633023584e-06,
      "loss": 0.4478,
      "mean_token_accuracy": 0.8549796305596828,
      "num_tokens": 1026496098.0,
      "step": 32163
    },
    {
      "entropy": 0.480269139748998,
      "epoch": 2.9592584286461836,
      "grad_norm": 0.1557343304157257,
      "learning_rate": 1.3616708068819578e-06,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8509322479367256,
      "num_tokens": 1026528139.0,
      "step": 32164
    },
    {
      "entropy": 0.4869471685960889,
      "epoch": 2.959350435017625,
      "grad_norm": 0.1563590168952942,
      "learning_rate": 1.358603980740332e-06,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8500192016363144,
      "num_tokens": 1026559279.0,
      "step": 32165
    },
    {
      "entropy": 0.5209658220410347,
      "epoch": 2.9594424413890663,
      "grad_norm": 0.16062255203723907,
      "learning_rate": 1.3555371545987058e-06,
      "loss": 0.506,
      "mean_token_accuracy": 0.8388379067182541,
      "num_tokens": 1026590741.0,
      "step": 32166
    },
    {
      "entropy": 0.46857895981520414,
      "epoch": 2.9595344477605074,
      "grad_norm": 0.15343818068504333,
      "learning_rate": 1.3524703284570798e-06,
      "loss": 0.4623,
      "mean_token_accuracy": 0.8512794263660908,
      "num_tokens": 1026622359.0,
      "step": 32167
    },
    {
      "entropy": 0.5222775117726997,
      "epoch": 2.9596264541319486,
      "grad_norm": 0.15388107299804688,
      "learning_rate": 1.3494035023154539e-06,
      "loss": 0.508,
      "mean_token_accuracy": 0.8371779248118401,
      "num_tokens": 1026654316.0,
      "step": 32168
    },
    {
      "entropy": 0.4162462083622813,
      "epoch": 2.9597184605033897,
      "grad_norm": 0.14207181334495544,
      "learning_rate": 1.3463366761738279e-06,
      "loss": 0.4073,
      "mean_token_accuracy": 0.8675755336880684,
      "num_tokens": 1026686862.0,
      "step": 32169
    },
    {
      "entropy": 0.48497254867106676,
      "epoch": 2.9598104668748313,
      "grad_norm": 0.15527844429016113,
      "learning_rate": 1.3432698500322017e-06,
      "loss": 0.4806,
      "mean_token_accuracy": 0.8479638434946537,
      "num_tokens": 1026718497.0,
      "step": 32170
    },
    {
      "entropy": 0.41607017535716295,
      "epoch": 2.9599024732462724,
      "grad_norm": 0.14560146629810333,
      "learning_rate": 1.3402030238905757e-06,
      "loss": 0.3959,
      "mean_token_accuracy": 0.8693778961896896,
      "num_tokens": 1026749711.0,
      "step": 32171
    },
    {
      "entropy": 0.5368530713021755,
      "epoch": 2.9599944796177136,
      "grad_norm": 0.1667083203792572,
      "learning_rate": 1.3371361977489497e-06,
      "loss": 0.5319,
      "mean_token_accuracy": 0.8336309790611267,
      "num_tokens": 1026780704.0,
      "step": 32172
    },
    {
      "entropy": 0.4556618304923177,
      "epoch": 2.9600864859891547,
      "grad_norm": 0.15745742619037628,
      "learning_rate": 1.3340693716073237e-06,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8554634377360344,
      "num_tokens": 1026811781.0,
      "step": 32173
    },
    {
      "entropy": 0.5101319309324026,
      "epoch": 2.960178492360596,
      "grad_norm": 0.15975606441497803,
      "learning_rate": 1.3310025454656975e-06,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8418052494525909,
      "num_tokens": 1026844055.0,
      "step": 32174
    },
    {
      "entropy": 0.4389064854476601,
      "epoch": 2.9602704987320374,
      "grad_norm": 0.15010394155979156,
      "learning_rate": 1.3279357193240715e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.861704271286726,
      "num_tokens": 1026875898.0,
      "step": 32175
    },
    {
      "entropy": 0.46137698646634817,
      "epoch": 2.9603625051034785,
      "grad_norm": 0.1495361030101776,
      "learning_rate": 1.3248688931824455e-06,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8552439101040363,
      "num_tokens": 1026908155.0,
      "step": 32176
    },
    {
      "entropy": 0.4547022597398609,
      "epoch": 2.9604545114749197,
      "grad_norm": 0.14605996012687683,
      "learning_rate": 1.3218020670408195e-06,
      "loss": 0.4458,
      "mean_token_accuracy": 0.8576490841805935,
      "num_tokens": 1026940422.0,
      "step": 32177
    },
    {
      "entropy": 0.5036822138354182,
      "epoch": 2.960546517846361,
      "grad_norm": 0.15844276547431946,
      "learning_rate": 1.3187352408991933e-06,
      "loss": 0.4979,
      "mean_token_accuracy": 0.8403379842638969,
      "num_tokens": 1026972484.0,
      "step": 32178
    },
    {
      "entropy": 0.5441761258989573,
      "epoch": 2.960638524217802,
      "grad_norm": 0.16015556454658508,
      "learning_rate": 1.3156684147575674e-06,
      "loss": 0.5331,
      "mean_token_accuracy": 0.8332786224782467,
      "num_tokens": 1027004915.0,
      "step": 32179
    },
    {
      "entropy": 0.4693728066049516,
      "epoch": 2.9607305305892435,
      "grad_norm": 0.14986476302146912,
      "learning_rate": 1.3126015886159414e-06,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8535972833633423,
      "num_tokens": 1027037542.0,
      "step": 32180
    },
    {
      "entropy": 0.5053011057898402,
      "epoch": 2.9608225369606846,
      "grad_norm": 0.15692894160747528,
      "learning_rate": 1.3095347624743154e-06,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8477112203836441,
      "num_tokens": 1027069465.0,
      "step": 32181
    },
    {
      "entropy": 0.5286132507026196,
      "epoch": 2.960914543332126,
      "grad_norm": 0.1553308218717575,
      "learning_rate": 1.3064679363326894e-06,
      "loss": 0.5127,
      "mean_token_accuracy": 0.8364853858947754,
      "num_tokens": 1027101598.0,
      "step": 32182
    },
    {
      "entropy": 0.5233627147972584,
      "epoch": 2.961006549703567,
      "grad_norm": 0.16082240641117096,
      "learning_rate": 1.3034011101910634e-06,
      "loss": 0.5052,
      "mean_token_accuracy": 0.8381575681269169,
      "num_tokens": 1027133367.0,
      "step": 32183
    },
    {
      "entropy": 0.5760040739551187,
      "epoch": 2.961098556075008,
      "grad_norm": 0.16582943499088287,
      "learning_rate": 1.3003342840494374e-06,
      "loss": 0.5641,
      "mean_token_accuracy": 0.8285480961203575,
      "num_tokens": 1027165760.0,
      "step": 32184
    },
    {
      "entropy": 0.4899447616189718,
      "epoch": 2.9611905624464496,
      "grad_norm": 0.15206165611743927,
      "learning_rate": 1.2972674579078112e-06,
      "loss": 0.4822,
      "mean_token_accuracy": 0.8482298962771893,
      "num_tokens": 1027198390.0,
      "step": 32185
    },
    {
      "entropy": 0.6119703082367778,
      "epoch": 2.9612825688178908,
      "grad_norm": 0.16799677908420563,
      "learning_rate": 1.2942006317661852e-06,
      "loss": 0.5982,
      "mean_token_accuracy": 0.8105991445481777,
      "num_tokens": 1027229890.0,
      "step": 32186
    },
    {
      "entropy": 0.5092736128717661,
      "epoch": 2.961374575189332,
      "grad_norm": 0.16003268957138062,
      "learning_rate": 1.2911338056245592e-06,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8414034731686115,
      "num_tokens": 1027261816.0,
      "step": 32187
    },
    {
      "entropy": 0.4424059501616284,
      "epoch": 2.961466581560773,
      "grad_norm": 0.14868855476379395,
      "learning_rate": 1.2880669794829333e-06,
      "loss": 0.4297,
      "mean_token_accuracy": 0.8636302761733532,
      "num_tokens": 1027294398.0,
      "step": 32188
    },
    {
      "entropy": 0.5371302170678973,
      "epoch": 2.961558587932214,
      "grad_norm": 0.16063261032104492,
      "learning_rate": 1.285000153341307e-06,
      "loss": 0.5239,
      "mean_token_accuracy": 0.833907674998045,
      "num_tokens": 1027326420.0,
      "step": 32189
    },
    {
      "entropy": 0.5680709425359964,
      "epoch": 2.9616505943036557,
      "grad_norm": 0.16657346487045288,
      "learning_rate": 1.281933327199681e-06,
      "loss": 0.5585,
      "mean_token_accuracy": 0.8248233683407307,
      "num_tokens": 1027358536.0,
      "step": 32190
    },
    {
      "entropy": 0.45749491825699806,
      "epoch": 2.961742600675097,
      "grad_norm": 0.15120932459831238,
      "learning_rate": 1.278866501058055e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8540220260620117,
      "num_tokens": 1027390431.0,
      "step": 32191
    },
    {
      "entropy": 0.5421678200364113,
      "epoch": 2.961834607046538,
      "grad_norm": 0.16130249202251434,
      "learning_rate": 1.275799674916429e-06,
      "loss": 0.5283,
      "mean_token_accuracy": 0.8294514678418636,
      "num_tokens": 1027422133.0,
      "step": 32192
    },
    {
      "entropy": 0.4712884873151779,
      "epoch": 2.961926613417979,
      "grad_norm": 0.15288913249969482,
      "learning_rate": 1.272732848774803e-06,
      "loss": 0.4561,
      "mean_token_accuracy": 0.8542916178703308,
      "num_tokens": 1027453975.0,
      "step": 32193
    },
    {
      "entropy": 0.4341368577443063,
      "epoch": 2.9620186197894203,
      "grad_norm": 0.14743490517139435,
      "learning_rate": 1.269666022633177e-06,
      "loss": 0.422,
      "mean_token_accuracy": 0.8630747497081757,
      "num_tokens": 1027485919.0,
      "step": 32194
    },
    {
      "entropy": 0.43911546072922647,
      "epoch": 2.962110626160862,
      "grad_norm": 0.14401715993881226,
      "learning_rate": 1.266599196491551e-06,
      "loss": 0.4308,
      "mean_token_accuracy": 0.8632588535547256,
      "num_tokens": 1027518497.0,
      "step": 32195
    },
    {
      "entropy": 0.5145508456043899,
      "epoch": 2.962202632532303,
      "grad_norm": 0.15365834534168243,
      "learning_rate": 1.263532370349925e-06,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8411691002547741,
      "num_tokens": 1027550962.0,
      "step": 32196
    },
    {
      "entropy": 0.5349120412720367,
      "epoch": 2.962294638903744,
      "grad_norm": 0.1607525795698166,
      "learning_rate": 1.2604655442082987e-06,
      "loss": 0.5255,
      "mean_token_accuracy": 0.8339854292571545,
      "num_tokens": 1027582798.0,
      "step": 32197
    },
    {
      "entropy": 0.5410489998757839,
      "epoch": 2.9623866452751852,
      "grad_norm": 0.158787801861763,
      "learning_rate": 1.257398718066673e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8311082683503628,
      "num_tokens": 1027615371.0,
      "step": 32198
    },
    {
      "entropy": 0.6508749919012189,
      "epoch": 2.9624786516466264,
      "grad_norm": 0.17725370824337006,
      "learning_rate": 1.2543318919250468e-06,
      "loss": 0.6482,
      "mean_token_accuracy": 0.7959261834621429,
      "num_tokens": 1027647116.0,
      "step": 32199
    },
    {
      "entropy": 0.4036416760645807,
      "epoch": 2.962570658018068,
      "grad_norm": 0.14039799571037292,
      "learning_rate": 1.2512650657834208e-06,
      "loss": 0.3879,
      "mean_token_accuracy": 0.8743336349725723,
      "num_tokens": 1027679525.0,
      "step": 32200
    },
    {
      "entropy": 0.4751293668523431,
      "epoch": 2.962662664389509,
      "grad_norm": 0.1512291431427002,
      "learning_rate": 1.2481982396417948e-06,
      "loss": 0.4658,
      "mean_token_accuracy": 0.8523337431252003,
      "num_tokens": 1027711424.0,
      "step": 32201
    },
    {
      "entropy": 0.5074322801083326,
      "epoch": 2.9627546707609502,
      "grad_norm": 0.158411905169487,
      "learning_rate": 1.2451314135001688e-06,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8390339724719524,
      "num_tokens": 1027742732.0,
      "step": 32202
    },
    {
      "entropy": 0.5934736803174019,
      "epoch": 2.9628466771323914,
      "grad_norm": 0.16786357760429382,
      "learning_rate": 1.2420645873585428e-06,
      "loss": 0.5835,
      "mean_token_accuracy": 0.8169018961489201,
      "num_tokens": 1027774126.0,
      "step": 32203
    },
    {
      "entropy": 0.45876084454357624,
      "epoch": 2.9629386835038325,
      "grad_norm": 0.14989948272705078,
      "learning_rate": 1.2389977612169166e-06,
      "loss": 0.4542,
      "mean_token_accuracy": 0.8539936430752277,
      "num_tokens": 1027806894.0,
      "step": 32204
    },
    {
      "entropy": 0.5713170422241092,
      "epoch": 2.963030689875274,
      "grad_norm": 0.1652839034795761,
      "learning_rate": 1.2359309350752906e-06,
      "loss": 0.5569,
      "mean_token_accuracy": 0.8253287896513939,
      "num_tokens": 1027838546.0,
      "step": 32205
    },
    {
      "entropy": 0.4859502734616399,
      "epoch": 2.963122696246715,
      "grad_norm": 0.15280002355575562,
      "learning_rate": 1.2328641089336646e-06,
      "loss": 0.4676,
      "mean_token_accuracy": 0.8497059233486652,
      "num_tokens": 1027870139.0,
      "step": 32206
    },
    {
      "entropy": 0.496772401034832,
      "epoch": 2.9632147026181563,
      "grad_norm": 0.15090107917785645,
      "learning_rate": 1.2297972827920386e-06,
      "loss": 0.4812,
      "mean_token_accuracy": 0.8481771275401115,
      "num_tokens": 1027901551.0,
      "step": 32207
    },
    {
      "entropy": 0.47774574742652476,
      "epoch": 2.9633067089895975,
      "grad_norm": 0.15030384063720703,
      "learning_rate": 1.2267304566504124e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8517349064350128,
      "num_tokens": 1027933330.0,
      "step": 32208
    },
    {
      "entropy": 0.4785393534693867,
      "epoch": 2.9633987153610386,
      "grad_norm": 0.16054844856262207,
      "learning_rate": 1.2236636305087867e-06,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8485161289572716,
      "num_tokens": 1027964504.0,
      "step": 32209
    },
    {
      "entropy": 0.5811847420409322,
      "epoch": 2.96349072173248,
      "grad_norm": 0.16586431860923767,
      "learning_rate": 1.2205968043671605e-06,
      "loss": 0.5701,
      "mean_token_accuracy": 0.8214950002729893,
      "num_tokens": 1027995979.0,
      "step": 32210
    },
    {
      "entropy": 0.4645373895764351,
      "epoch": 2.9635827281039213,
      "grad_norm": 0.14885200560092926,
      "learning_rate": 1.2175299782255345e-06,
      "loss": 0.4579,
      "mean_token_accuracy": 0.856586329638958,
      "num_tokens": 1028028103.0,
      "step": 32211
    },
    {
      "entropy": 0.44452530494891107,
      "epoch": 2.9636747344753624,
      "grad_norm": 0.1482878178358078,
      "learning_rate": 1.2144631520839083e-06,
      "loss": 0.432,
      "mean_token_accuracy": 0.8584084287285805,
      "num_tokens": 1028059933.0,
      "step": 32212
    },
    {
      "entropy": 0.5291555039584637,
      "epoch": 2.9637667408468036,
      "grad_norm": 0.1606413722038269,
      "learning_rate": 1.2113963259422825e-06,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8364062197506428,
      "num_tokens": 1028092127.0,
      "step": 32213
    },
    {
      "entropy": 0.5480206599459052,
      "epoch": 2.9638587472182447,
      "grad_norm": 0.16281239688396454,
      "learning_rate": 1.2083294998006563e-06,
      "loss": 0.5285,
      "mean_token_accuracy": 0.8335162065923214,
      "num_tokens": 1028123437.0,
      "step": 32214
    },
    {
      "entropy": 0.5166218504309654,
      "epoch": 2.9639507535896863,
      "grad_norm": 0.1600278913974762,
      "learning_rate": 1.2052626736590303e-06,
      "loss": 0.5094,
      "mean_token_accuracy": 0.8360125310719013,
      "num_tokens": 1028155005.0,
      "step": 32215
    },
    {
      "entropy": 0.4945978485047817,
      "epoch": 2.9640427599611274,
      "grad_norm": 0.1527179628610611,
      "learning_rate": 1.2021958475174043e-06,
      "loss": 0.4823,
      "mean_token_accuracy": 0.8499787822365761,
      "num_tokens": 1028187193.0,
      "step": 32216
    },
    {
      "entropy": 0.5621816357597709,
      "epoch": 2.9641347663325686,
      "grad_norm": 0.1692175418138504,
      "learning_rate": 1.1991290213757783e-06,
      "loss": 0.5475,
      "mean_token_accuracy": 0.8268318697810173,
      "num_tokens": 1028218383.0,
      "step": 32217
    },
    {
      "entropy": 0.49210862442851067,
      "epoch": 2.9642267727040097,
      "grad_norm": 0.15707719326019287,
      "learning_rate": 1.1960621952341521e-06,
      "loss": 0.4788,
      "mean_token_accuracy": 0.8475061468780041,
      "num_tokens": 1028250702.0,
      "step": 32218
    },
    {
      "entropy": 0.39643796207383275,
      "epoch": 2.964318779075451,
      "grad_norm": 0.14288893342018127,
      "learning_rate": 1.1929953690925262e-06,
      "loss": 0.3771,
      "mean_token_accuracy": 0.8763169795274734,
      "num_tokens": 1028282897.0,
      "step": 32219
    },
    {
      "entropy": 0.5067685376852751,
      "epoch": 2.9644107854468924,
      "grad_norm": 0.16005070507526398,
      "learning_rate": 1.1899285429509002e-06,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8415132462978363,
      "num_tokens": 1028314686.0,
      "step": 32220
    },
    {
      "entropy": 0.44491367042064667,
      "epoch": 2.9645027918183335,
      "grad_norm": 0.15109634399414062,
      "learning_rate": 1.1868617168092742e-06,
      "loss": 0.43,
      "mean_token_accuracy": 0.863400649279356,
      "num_tokens": 1028346005.0,
      "step": 32221
    },
    {
      "entropy": 0.6148242987692356,
      "epoch": 2.9645947981897747,
      "grad_norm": 0.17442777752876282,
      "learning_rate": 1.1837948906676482e-06,
      "loss": 0.6126,
      "mean_token_accuracy": 0.8091343231499195,
      "num_tokens": 1028378269.0,
      "step": 32222
    },
    {
      "entropy": 0.4815577024128288,
      "epoch": 2.964686804561216,
      "grad_norm": 0.15044893324375153,
      "learning_rate": 1.180728064526022e-06,
      "loss": 0.4585,
      "mean_token_accuracy": 0.8527756370604038,
      "num_tokens": 1028410290.0,
      "step": 32223
    },
    {
      "entropy": 0.5370516823604703,
      "epoch": 2.964778810932657,
      "grad_norm": 0.1617552638053894,
      "learning_rate": 1.1776612383843962e-06,
      "loss": 0.5345,
      "mean_token_accuracy": 0.8327954560518265,
      "num_tokens": 1028442570.0,
      "step": 32224
    },
    {
      "entropy": 0.5937168318778276,
      "epoch": 2.9648708173040985,
      "grad_norm": 0.1657361388206482,
      "learning_rate": 1.17459441224277e-06,
      "loss": 0.5853,
      "mean_token_accuracy": 0.8159993216395378,
      "num_tokens": 1028474454.0,
      "step": 32225
    },
    {
      "entropy": 0.6148591972887516,
      "epoch": 2.9649628236755396,
      "grad_norm": 0.1753617525100708,
      "learning_rate": 1.171527586101144e-06,
      "loss": 0.6039,
      "mean_token_accuracy": 0.8116931729018688,
      "num_tokens": 1028505278.0,
      "step": 32226
    },
    {
      "entropy": 0.6037893537431955,
      "epoch": 2.965054830046981,
      "grad_norm": 0.16961492598056793,
      "learning_rate": 1.1684607599595178e-06,
      "loss": 0.595,
      "mean_token_accuracy": 0.8125064224004745,
      "num_tokens": 1028537288.0,
      "step": 32227
    },
    {
      "entropy": 0.5022375863045454,
      "epoch": 2.965146836418422,
      "grad_norm": 0.1629834622144699,
      "learning_rate": 1.165393933817892e-06,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8382101953029633,
      "num_tokens": 1028568621.0,
      "step": 32228
    },
    {
      "entropy": 0.46327031310647726,
      "epoch": 2.965238842789863,
      "grad_norm": 0.15633659064769745,
      "learning_rate": 1.1623271076762659e-06,
      "loss": 0.458,
      "mean_token_accuracy": 0.8550660163164139,
      "num_tokens": 1028600399.0,
      "step": 32229
    },
    {
      "entropy": 0.4248107271268964,
      "epoch": 2.9653308491613046,
      "grad_norm": 0.14194607734680176,
      "learning_rate": 1.1592602815346399e-06,
      "loss": 0.4071,
      "mean_token_accuracy": 0.8690138533711433,
      "num_tokens": 1028632831.0,
      "step": 32230
    },
    {
      "entropy": 0.45262608560733497,
      "epoch": 2.9654228555327458,
      "grad_norm": 0.14771734178066254,
      "learning_rate": 1.1561934553930139e-06,
      "loss": 0.447,
      "mean_token_accuracy": 0.8591305799782276,
      "num_tokens": 1028665267.0,
      "step": 32231
    },
    {
      "entropy": 0.5536364782601595,
      "epoch": 2.965514861904187,
      "grad_norm": 0.1677790880203247,
      "learning_rate": 1.153126629251388e-06,
      "loss": 0.5453,
      "mean_token_accuracy": 0.8274675607681274,
      "num_tokens": 1028697089.0,
      "step": 32232
    },
    {
      "entropy": 0.5402936162427068,
      "epoch": 2.965606868275628,
      "grad_norm": 0.16764578223228455,
      "learning_rate": 1.1500598031097617e-06,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8338696509599686,
      "num_tokens": 1028728726.0,
      "step": 32233
    },
    {
      "entropy": 0.5449135070666671,
      "epoch": 2.965698874647069,
      "grad_norm": 0.1608424335718155,
      "learning_rate": 1.1469929769681357e-06,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8317479901015759,
      "num_tokens": 1028760212.0,
      "step": 32234
    },
    {
      "entropy": 0.42924045177642256,
      "epoch": 2.9657908810185107,
      "grad_norm": 0.14362475275993347,
      "learning_rate": 1.1439261508265097e-06,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8666749857366085,
      "num_tokens": 1028792767.0,
      "step": 32235
    },
    {
      "entropy": 0.5043237628415227,
      "epoch": 2.965882887389952,
      "grad_norm": 0.16164515912532806,
      "learning_rate": 1.1408593246848837e-06,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8404127918183804,
      "num_tokens": 1028824438.0,
      "step": 32236
    },
    {
      "entropy": 0.5222768876701593,
      "epoch": 2.965974893761393,
      "grad_norm": 0.15482006967067719,
      "learning_rate": 1.1377924985432575e-06,
      "loss": 0.5103,
      "mean_token_accuracy": 0.8380273766815662,
      "num_tokens": 1028856352.0,
      "step": 32237
    },
    {
      "entropy": 0.5299500059336424,
      "epoch": 2.966066900132834,
      "grad_norm": 0.1622035950422287,
      "learning_rate": 1.1347256724016315e-06,
      "loss": 0.5156,
      "mean_token_accuracy": 0.8354271650314331,
      "num_tokens": 1028888400.0,
      "step": 32238
    },
    {
      "entropy": 0.5285862870514393,
      "epoch": 2.9661589065042753,
      "grad_norm": 0.16351303458213806,
      "learning_rate": 1.1316588462600056e-06,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8372374549508095,
      "num_tokens": 1028920434.0,
      "step": 32239
    },
    {
      "entropy": 0.5247831866145134,
      "epoch": 2.966250912875717,
      "grad_norm": 0.16410815715789795,
      "learning_rate": 1.1285920201183796e-06,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8351546861231327,
      "num_tokens": 1028952269.0,
      "step": 32240
    },
    {
      "entropy": 0.5091661503538489,
      "epoch": 2.966342919247158,
      "grad_norm": 0.1565173715353012,
      "learning_rate": 1.1255251939767534e-06,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8427984192967415,
      "num_tokens": 1028983974.0,
      "step": 32241
    },
    {
      "entropy": 0.5696617811918259,
      "epoch": 2.966434925618599,
      "grad_norm": 0.16693848371505737,
      "learning_rate": 1.1224583678351274e-06,
      "loss": 0.5663,
      "mean_token_accuracy": 0.8202609680593014,
      "num_tokens": 1029016711.0,
      "step": 32242
    },
    {
      "entropy": 0.4965949389152229,
      "epoch": 2.9665269319900403,
      "grad_norm": 0.15437646210193634,
      "learning_rate": 1.1193915416935016e-06,
      "loss": 0.483,
      "mean_token_accuracy": 0.8470289707183838,
      "num_tokens": 1029048504.0,
      "step": 32243
    },
    {
      "entropy": 0.43484185775741935,
      "epoch": 2.9666189383614814,
      "grad_norm": 0.14570844173431396,
      "learning_rate": 1.1163247155518754e-06,
      "loss": 0.423,
      "mean_token_accuracy": 0.8674041219055653,
      "num_tokens": 1029080279.0,
      "step": 32244
    },
    {
      "entropy": 0.4722154795890674,
      "epoch": 2.966710944732923,
      "grad_norm": 0.14788436889648438,
      "learning_rate": 1.1132578894102494e-06,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8555599227547646,
      "num_tokens": 1029112796.0,
      "step": 32245
    },
    {
      "entropy": 0.47404505126178265,
      "epoch": 2.966802951104364,
      "grad_norm": 0.15292181074619293,
      "learning_rate": 1.1101910632686234e-06,
      "loss": 0.4646,
      "mean_token_accuracy": 0.8521328307688236,
      "num_tokens": 1029145027.0,
      "step": 32246
    },
    {
      "entropy": 0.4610542682930827,
      "epoch": 2.9668949574758052,
      "grad_norm": 0.14932456612586975,
      "learning_rate": 1.1071242371269974e-06,
      "loss": 0.4521,
      "mean_token_accuracy": 0.8563302271068096,
      "num_tokens": 1029177273.0,
      "step": 32247
    },
    {
      "entropy": 0.5482483403757215,
      "epoch": 2.9669869638472464,
      "grad_norm": 0.16578851640224457,
      "learning_rate": 1.1040574109853712e-06,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8278865329921246,
      "num_tokens": 1029209069.0,
      "step": 32248
    },
    {
      "entropy": 0.4314740113914013,
      "epoch": 2.9670789702186875,
      "grad_norm": 0.14661312103271484,
      "learning_rate": 1.1009905848437453e-06,
      "loss": 0.413,
      "mean_token_accuracy": 0.8662921339273453,
      "num_tokens": 1029241056.0,
      "step": 32249
    },
    {
      "entropy": 0.5531725380569696,
      "epoch": 2.967170976590129,
      "grad_norm": 0.1635354459285736,
      "learning_rate": 1.0979237587021193e-06,
      "loss": 0.5445,
      "mean_token_accuracy": 0.8258672170341015,
      "num_tokens": 1029273310.0,
      "step": 32250
    },
    {
      "entropy": 0.5185016300529242,
      "epoch": 2.96726298296157,
      "grad_norm": 0.16595251858234406,
      "learning_rate": 1.0948569325604933e-06,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8353414796292782,
      "num_tokens": 1029305324.0,
      "step": 32251
    },
    {
      "entropy": 0.35556650650687516,
      "epoch": 2.9673549893330113,
      "grad_norm": 0.1330936849117279,
      "learning_rate": 1.091790106418867e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.8898858763277531,
      "num_tokens": 1029337777.0,
      "step": 32252
    },
    {
      "entropy": 0.4980971257900819,
      "epoch": 2.9674469957044525,
      "grad_norm": 0.15334518253803253,
      "learning_rate": 1.088723280277241e-06,
      "loss": 0.4859,
      "mean_token_accuracy": 0.8444637395441532,
      "num_tokens": 1029370185.0,
      "step": 32253
    },
    {
      "entropy": 0.5761081799864769,
      "epoch": 2.9675390020758936,
      "grad_norm": 0.16821280121803284,
      "learning_rate": 1.0856564541356151e-06,
      "loss": 0.5772,
      "mean_token_accuracy": 0.8167408853769302,
      "num_tokens": 1029402001.0,
      "step": 32254
    },
    {
      "entropy": 0.5062299938872457,
      "epoch": 2.967631008447335,
      "grad_norm": 0.15301500260829926,
      "learning_rate": 1.0825896279939891e-06,
      "loss": 0.4948,
      "mean_token_accuracy": 0.8407672643661499,
      "num_tokens": 1029433981.0,
      "step": 32255
    },
    {
      "entropy": 0.5006883158348501,
      "epoch": 2.9677230148187763,
      "grad_norm": 0.15740962326526642,
      "learning_rate": 1.079522801852363e-06,
      "loss": 0.4898,
      "mean_token_accuracy": 0.8423534072935581,
      "num_tokens": 1029465969.0,
      "step": 32256
    },
    {
      "entropy": 0.42812182079069316,
      "epoch": 2.9678150211902175,
      "grad_norm": 0.14088882505893707,
      "learning_rate": 1.076455975710737e-06,
      "loss": 0.4193,
      "mean_token_accuracy": 0.8684586212038994,
      "num_tokens": 1029498253.0,
      "step": 32257
    },
    {
      "entropy": 0.46068246476352215,
      "epoch": 2.9679070275616586,
      "grad_norm": 0.15039314329624176,
      "learning_rate": 1.073389149569111e-06,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8572991564869881,
      "num_tokens": 1029530498.0,
      "step": 32258
    },
    {
      "entropy": 0.43031419976614416,
      "epoch": 2.9679990339330997,
      "grad_norm": 0.14304468035697937,
      "learning_rate": 1.070322323427485e-06,
      "loss": 0.41,
      "mean_token_accuracy": 0.8665422536432743,
      "num_tokens": 1029562001.0,
      "step": 32259
    },
    {
      "entropy": 0.4980227751657367,
      "epoch": 2.9680910403045413,
      "grad_norm": 0.16005510091781616,
      "learning_rate": 1.0672554972858588e-06,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8424009867012501,
      "num_tokens": 1029594010.0,
      "step": 32260
    },
    {
      "entropy": 0.5509489849209785,
      "epoch": 2.9681830466759824,
      "grad_norm": 0.16820791363716125,
      "learning_rate": 1.064188671144233e-06,
      "loss": 0.5381,
      "mean_token_accuracy": 0.8310459852218628,
      "num_tokens": 1029625278.0,
      "step": 32261
    },
    {
      "entropy": 0.5007706279866397,
      "epoch": 2.9682750530474236,
      "grad_norm": 0.1547941267490387,
      "learning_rate": 1.061121845002607e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8424835726618767,
      "num_tokens": 1029657939.0,
      "step": 32262
    },
    {
      "entropy": 0.51571992225945,
      "epoch": 2.9683670594188647,
      "grad_norm": 0.16049888730049133,
      "learning_rate": 1.0580550188609808e-06,
      "loss": 0.4968,
      "mean_token_accuracy": 0.8385275267064571,
      "num_tokens": 1029689849.0,
      "step": 32263
    },
    {
      "entropy": 0.5489074420183897,
      "epoch": 2.968459065790306,
      "grad_norm": 0.16203747689723969,
      "learning_rate": 1.0549881927193548e-06,
      "loss": 0.5414,
      "mean_token_accuracy": 0.8281932771205902,
      "num_tokens": 1029721661.0,
      "step": 32264
    },
    {
      "entropy": 0.5232167541980743,
      "epoch": 2.9685510721617474,
      "grad_norm": 0.15995238721370697,
      "learning_rate": 1.0519213665777288e-06,
      "loss": 0.5061,
      "mean_token_accuracy": 0.8389117680490017,
      "num_tokens": 1029753364.0,
      "step": 32265
    },
    {
      "entropy": 0.5950754433870316,
      "epoch": 2.9686430785331885,
      "grad_norm": 0.16580048203468323,
      "learning_rate": 1.0488545404361028e-06,
      "loss": 0.583,
      "mean_token_accuracy": 0.8161311969161034,
      "num_tokens": 1029784758.0,
      "step": 32266
    },
    {
      "entropy": 0.4085292899981141,
      "epoch": 2.9687350849046297,
      "grad_norm": 0.14219669997692108,
      "learning_rate": 1.0457877142944766e-06,
      "loss": 0.3936,
      "mean_token_accuracy": 0.8714838549494743,
      "num_tokens": 1029817106.0,
      "step": 32267
    },
    {
      "entropy": 0.45228573121130466,
      "epoch": 2.968827091276071,
      "grad_norm": 0.14983730018138885,
      "learning_rate": 1.0427208881528506e-06,
      "loss": 0.4337,
      "mean_token_accuracy": 0.8598489090800285,
      "num_tokens": 1029848848.0,
      "step": 32268
    },
    {
      "entropy": 0.4294070064788684,
      "epoch": 2.968919097647512,
      "grad_norm": 0.14000093936920166,
      "learning_rate": 1.0396540620112247e-06,
      "loss": 0.4136,
      "mean_token_accuracy": 0.8661414086818695,
      "num_tokens": 1029880410.0,
      "step": 32269
    },
    {
      "entropy": 0.5045245578512549,
      "epoch": 2.9690111040189535,
      "grad_norm": 0.1550406515598297,
      "learning_rate": 1.0365872358695987e-06,
      "loss": 0.4888,
      "mean_token_accuracy": 0.8443292267620564,
      "num_tokens": 1029912261.0,
      "step": 32270
    },
    {
      "entropy": 0.4676247730385512,
      "epoch": 2.9691031103903947,
      "grad_norm": 0.15253852307796478,
      "learning_rate": 1.0335204097279725e-06,
      "loss": 0.4516,
      "mean_token_accuracy": 0.8540171273052692,
      "num_tokens": 1029943806.0,
      "step": 32271
    },
    {
      "entropy": 0.40541810216382146,
      "epoch": 2.969195116761836,
      "grad_norm": 0.14437323808670044,
      "learning_rate": 1.0304535835863465e-06,
      "loss": 0.3938,
      "mean_token_accuracy": 0.8775572627782822,
      "num_tokens": 1029975276.0,
      "step": 32272
    },
    {
      "entropy": 0.5264608999714255,
      "epoch": 2.969287123133277,
      "grad_norm": 0.16128945350646973,
      "learning_rate": 1.0273867574447205e-06,
      "loss": 0.5042,
      "mean_token_accuracy": 0.839641347527504,
      "num_tokens": 1030007027.0,
      "step": 32273
    },
    {
      "entropy": 0.5501187220215797,
      "epoch": 2.969379129504718,
      "grad_norm": 0.1580071598291397,
      "learning_rate": 1.0243199313030945e-06,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8288349136710167,
      "num_tokens": 1030039391.0,
      "step": 32274
    },
    {
      "entropy": 0.45394227746874094,
      "epoch": 2.9694711358761596,
      "grad_norm": 0.15236036479473114,
      "learning_rate": 1.0212531051614683e-06,
      "loss": 0.442,
      "mean_token_accuracy": 0.8600396066904068,
      "num_tokens": 1030071104.0,
      "step": 32275
    },
    {
      "entropy": 0.5335117280483246,
      "epoch": 2.9695631422476008,
      "grad_norm": 0.15689371526241302,
      "learning_rate": 1.0181862790198425e-06,
      "loss": 0.5193,
      "mean_token_accuracy": 0.833508487790823,
      "num_tokens": 1030103615.0,
      "step": 32276
    },
    {
      "entropy": 0.5819128947332501,
      "epoch": 2.969655148619042,
      "grad_norm": 0.16442255675792694,
      "learning_rate": 1.0151194528782163e-06,
      "loss": 0.5691,
      "mean_token_accuracy": 0.8198184594511986,
      "num_tokens": 1030135485.0,
      "step": 32277
    },
    {
      "entropy": 0.4877621568739414,
      "epoch": 2.969747154990483,
      "grad_norm": 0.1575632244348526,
      "learning_rate": 1.0120526267365903e-06,
      "loss": 0.478,
      "mean_token_accuracy": 0.8475013300776482,
      "num_tokens": 1030167207.0,
      "step": 32278
    },
    {
      "entropy": 0.5135930581018329,
      "epoch": 2.969839161361924,
      "grad_norm": 0.15113763511180878,
      "learning_rate": 1.0089858005949641e-06,
      "loss": 0.503,
      "mean_token_accuracy": 0.8409087434411049,
      "num_tokens": 1030199780.0,
      "step": 32279
    },
    {
      "entropy": 0.5483936425298452,
      "epoch": 2.9699311677333657,
      "grad_norm": 0.168921560049057,
      "learning_rate": 1.0059189744533384e-06,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8322047963738441,
      "num_tokens": 1030230687.0,
      "step": 32280
    },
    {
      "entropy": 0.41649895580485463,
      "epoch": 2.970023174104807,
      "grad_norm": 0.14903345704078674,
      "learning_rate": 1.0028521483117122e-06,
      "loss": 0.4063,
      "mean_token_accuracy": 0.8654147461056709,
      "num_tokens": 1030262073.0,
      "step": 32281
    },
    {
      "entropy": 0.5496848258189857,
      "epoch": 2.970115180476248,
      "grad_norm": 0.15801699459552765,
      "learning_rate": 9.997853221700862e-07,
      "loss": 0.5382,
      "mean_token_accuracy": 0.829232532531023,
      "num_tokens": 1030293880.0,
      "step": 32282
    },
    {
      "entropy": 0.4454535283148289,
      "epoch": 2.970207186847689,
      "grad_norm": 0.14812766015529633,
      "learning_rate": 9.967184960284602e-07,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8632466569542885,
      "num_tokens": 1030326194.0,
      "step": 32283
    },
    {
      "entropy": 0.5558443069458008,
      "epoch": 2.9702991932191303,
      "grad_norm": 0.16517774760723114,
      "learning_rate": 9.936516698868342e-07,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8288338780403137,
      "num_tokens": 1030358037.0,
      "step": 32284
    },
    {
      "entropy": 0.46605911757797003,
      "epoch": 2.970391199590572,
      "grad_norm": 0.14746522903442383,
      "learning_rate": 9.905848437452082e-07,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8572420068085194,
      "num_tokens": 1030390474.0,
      "step": 32285
    },
    {
      "entropy": 0.5149324899539351,
      "epoch": 2.970483205962013,
      "grad_norm": 0.1622145175933838,
      "learning_rate": 9.87518017603582e-07,
      "loss": 0.5125,
      "mean_token_accuracy": 0.8392997533082962,
      "num_tokens": 1030421922.0,
      "step": 32286
    },
    {
      "entropy": 0.4715610947459936,
      "epoch": 2.970575212333454,
      "grad_norm": 0.15142831206321716,
      "learning_rate": 9.84451191461956e-07,
      "loss": 0.4598,
      "mean_token_accuracy": 0.8522660322487354,
      "num_tokens": 1030453803.0,
      "step": 32287
    },
    {
      "entropy": 0.49365594796836376,
      "epoch": 2.9706672187048953,
      "grad_norm": 0.15785880386829376,
      "learning_rate": 9.8138436532033e-07,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8443404771387577,
      "num_tokens": 1030485725.0,
      "step": 32288
    },
    {
      "entropy": 0.4799515618942678,
      "epoch": 2.9707592250763364,
      "grad_norm": 0.15208406746387482,
      "learning_rate": 9.78317539178704e-07,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8500353246927261,
      "num_tokens": 1030518291.0,
      "step": 32289
    },
    {
      "entropy": 0.4619027730077505,
      "epoch": 2.970851231447778,
      "grad_norm": 0.1607932597398758,
      "learning_rate": 9.752507130370779e-07,
      "loss": 0.4544,
      "mean_token_accuracy": 0.8538635484874249,
      "num_tokens": 1030548653.0,
      "step": 32290
    },
    {
      "entropy": 0.5363004859536886,
      "epoch": 2.970943237819219,
      "grad_norm": 0.16112969815731049,
      "learning_rate": 9.72183886895452e-07,
      "loss": 0.5215,
      "mean_token_accuracy": 0.8366954512894154,
      "num_tokens": 1030580825.0,
      "step": 32291
    },
    {
      "entropy": 0.47714396100491285,
      "epoch": 2.9710352441906602,
      "grad_norm": 0.15193746984004974,
      "learning_rate": 9.691170607538259e-07,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8505403064191341,
      "num_tokens": 1030612776.0,
      "step": 32292
    },
    {
      "entropy": 0.5251020456198603,
      "epoch": 2.9711272505621014,
      "grad_norm": 0.16026192903518677,
      "learning_rate": 9.660502346121999e-07,
      "loss": 0.518,
      "mean_token_accuracy": 0.8400356136262417,
      "num_tokens": 1030644381.0,
      "step": 32293
    },
    {
      "entropy": 0.42331167683005333,
      "epoch": 2.9712192569335425,
      "grad_norm": 0.1502801477909088,
      "learning_rate": 9.62983408470574e-07,
      "loss": 0.4176,
      "mean_token_accuracy": 0.8643608652055264,
      "num_tokens": 1030676296.0,
      "step": 32294
    },
    {
      "entropy": 0.44222155038733035,
      "epoch": 2.971311263304984,
      "grad_norm": 0.1472693830728531,
      "learning_rate": 9.59916582328948e-07,
      "loss": 0.4363,
      "mean_token_accuracy": 0.8599495962262154,
      "num_tokens": 1030708844.0,
      "step": 32295
    },
    {
      "entropy": 0.5177637487649918,
      "epoch": 2.971403269676425,
      "grad_norm": 0.15887722373008728,
      "learning_rate": 9.568497561873217e-07,
      "loss": 0.5102,
      "mean_token_accuracy": 0.8389346525073051,
      "num_tokens": 1030741197.0,
      "step": 32296
    },
    {
      "entropy": 0.4232736211270094,
      "epoch": 2.9714952760478663,
      "grad_norm": 0.14480677247047424,
      "learning_rate": 9.537829300456957e-07,
      "loss": 0.4074,
      "mean_token_accuracy": 0.8695213012397289,
      "num_tokens": 1030773573.0,
      "step": 32297
    },
    {
      "entropy": 0.5612757280468941,
      "epoch": 2.9715872824193075,
      "grad_norm": 0.16232788562774658,
      "learning_rate": 9.507161039040696e-07,
      "loss": 0.5587,
      "mean_token_accuracy": 0.8253078386187553,
      "num_tokens": 1030805827.0,
      "step": 32298
    },
    {
      "entropy": 0.5419466532766819,
      "epoch": 2.9716792887907486,
      "grad_norm": 0.16176539659500122,
      "learning_rate": 9.476492777624438e-07,
      "loss": 0.5346,
      "mean_token_accuracy": 0.8329378627240658,
      "num_tokens": 1030838063.0,
      "step": 32299
    },
    {
      "entropy": 0.500976949930191,
      "epoch": 2.97177129516219,
      "grad_norm": 0.15940485894680023,
      "learning_rate": 9.445824516208176e-07,
      "loss": 0.4954,
      "mean_token_accuracy": 0.8424787223339081,
      "num_tokens": 1030869453.0,
      "step": 32300
    },
    {
      "entropy": 0.4606150649487972,
      "epoch": 2.9718633015336313,
      "grad_norm": 0.15324339270591736,
      "learning_rate": 9.415156254791917e-07,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8541411273181438,
      "num_tokens": 1030901751.0,
      "step": 32301
    },
    {
      "entropy": 0.46865101251751184,
      "epoch": 2.9719553079050725,
      "grad_norm": 0.15266941487789154,
      "learning_rate": 9.384487993375655e-07,
      "loss": 0.4589,
      "mean_token_accuracy": 0.8527810238301754,
      "num_tokens": 1030933871.0,
      "step": 32302
    },
    {
      "entropy": 0.6276939585804939,
      "epoch": 2.9720473142765136,
      "grad_norm": 0.1736716479063034,
      "learning_rate": 9.353819731959396e-07,
      "loss": 0.6158,
      "mean_token_accuracy": 0.8075175508856773,
      "num_tokens": 1030965214.0,
      "step": 32303
    },
    {
      "entropy": 0.5850006602704525,
      "epoch": 2.9721393206479547,
      "grad_norm": 0.170444518327713,
      "learning_rate": 9.323151470543136e-07,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8197363317012787,
      "num_tokens": 1030997134.0,
      "step": 32304
    },
    {
      "entropy": 0.4010355811333284,
      "epoch": 2.9722313270193963,
      "grad_norm": 0.1364968866109848,
      "learning_rate": 9.292483209126875e-07,
      "loss": 0.3849,
      "mean_token_accuracy": 0.8745917715132236,
      "num_tokens": 1031029565.0,
      "step": 32305
    },
    {
      "entropy": 0.5124551327899098,
      "epoch": 2.9723233333908374,
      "grad_norm": 0.15803663432598114,
      "learning_rate": 9.261814947710615e-07,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8374841026961803,
      "num_tokens": 1031061512.0,
      "step": 32306
    },
    {
      "entropy": 0.5323857488110662,
      "epoch": 2.9724153397622786,
      "grad_norm": 0.1536620408296585,
      "learning_rate": 9.231146686294354e-07,
      "loss": 0.5196,
      "mean_token_accuracy": 0.8364644683897495,
      "num_tokens": 1031093536.0,
      "step": 32307
    },
    {
      "entropy": 0.47805634408723563,
      "epoch": 2.9725073461337197,
      "grad_norm": 0.15168625116348267,
      "learning_rate": 9.200478424878094e-07,
      "loss": 0.4643,
      "mean_token_accuracy": 0.8508875519037247,
      "num_tokens": 1031124627.0,
      "step": 32308
    },
    {
      "entropy": 0.45635552424937487,
      "epoch": 2.972599352505161,
      "grad_norm": 0.15315008163452148,
      "learning_rate": 9.169810163461833e-07,
      "loss": 0.4436,
      "mean_token_accuracy": 0.855119314044714,
      "num_tokens": 1031156753.0,
      "step": 32309
    },
    {
      "entropy": 0.5963024543598294,
      "epoch": 2.9726913588766024,
      "grad_norm": 0.16816383600234985,
      "learning_rate": 9.139141902045574e-07,
      "loss": 0.5859,
      "mean_token_accuracy": 0.8183597587049007,
      "num_tokens": 1031188687.0,
      "step": 32310
    },
    {
      "entropy": 0.45074520632624626,
      "epoch": 2.9727833652480435,
      "grad_norm": 0.15000170469284058,
      "learning_rate": 9.108473640629313e-07,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8584470599889755,
      "num_tokens": 1031221187.0,
      "step": 32311
    },
    {
      "entropy": 0.5712340641766787,
      "epoch": 2.9728753716194847,
      "grad_norm": 0.1732204407453537,
      "learning_rate": 9.077805379213053e-07,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8247973844408989,
      "num_tokens": 1031251796.0,
      "step": 32312
    },
    {
      "entropy": 0.5764369145035744,
      "epoch": 2.972967377990926,
      "grad_norm": 0.16801726818084717,
      "learning_rate": 9.047137117796792e-07,
      "loss": 0.5666,
      "mean_token_accuracy": 0.8197418823838234,
      "num_tokens": 1031283645.0,
      "step": 32313
    },
    {
      "entropy": 0.45279229059815407,
      "epoch": 2.973059384362367,
      "grad_norm": 0.1452091783285141,
      "learning_rate": 9.016468856380533e-07,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8607050627470016,
      "num_tokens": 1031315692.0,
      "step": 32314
    },
    {
      "entropy": 0.499296513851732,
      "epoch": 2.9731513907338085,
      "grad_norm": 0.1547030657529831,
      "learning_rate": 8.985800594964271e-07,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8444201275706291,
      "num_tokens": 1031347044.0,
      "step": 32315
    },
    {
      "entropy": 0.6162654850631952,
      "epoch": 2.9732433971052497,
      "grad_norm": 0.17113548517227173,
      "learning_rate": 8.955132333548012e-07,
      "loss": 0.5959,
      "mean_token_accuracy": 0.8156956471502781,
      "num_tokens": 1031378694.0,
      "step": 32316
    },
    {
      "entropy": 0.4789689276367426,
      "epoch": 2.973335403476691,
      "grad_norm": 0.15136702358722687,
      "learning_rate": 8.92446407213175e-07,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8521506637334824,
      "num_tokens": 1031410587.0,
      "step": 32317
    },
    {
      "entropy": 0.518044475466013,
      "epoch": 2.973427409848132,
      "grad_norm": 0.15921765565872192,
      "learning_rate": 8.893795810715491e-07,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8422594927251339,
      "num_tokens": 1031442558.0,
      "step": 32318
    },
    {
      "entropy": 0.5500902403146029,
      "epoch": 2.973519416219573,
      "grad_norm": 0.17061297595500946,
      "learning_rate": 8.86312754929923e-07,
      "loss": 0.5407,
      "mean_token_accuracy": 0.8276944644749165,
      "num_tokens": 1031474377.0,
      "step": 32319
    },
    {
      "entropy": 0.4846153296530247,
      "epoch": 2.9736114225910146,
      "grad_norm": 0.15910638868808746,
      "learning_rate": 8.832459287882971e-07,
      "loss": 0.4831,
      "mean_token_accuracy": 0.8490417823195457,
      "num_tokens": 1031505818.0,
      "step": 32320
    },
    {
      "entropy": 0.49742671847343445,
      "epoch": 2.9737034289624558,
      "grad_norm": 0.1544404923915863,
      "learning_rate": 8.80179102646671e-07,
      "loss": 0.4818,
      "mean_token_accuracy": 0.8438292555510998,
      "num_tokens": 1031538054.0,
      "step": 32321
    },
    {
      "entropy": 0.5028064800426364,
      "epoch": 2.973795435333897,
      "grad_norm": 0.15509334206581116,
      "learning_rate": 8.77112276505045e-07,
      "loss": 0.4829,
      "mean_token_accuracy": 0.8460366651415825,
      "num_tokens": 1031570583.0,
      "step": 32322
    },
    {
      "entropy": 0.49886561650782824,
      "epoch": 2.973887441705338,
      "grad_norm": 0.1553271859884262,
      "learning_rate": 8.74045450363419e-07,
      "loss": 0.4908,
      "mean_token_accuracy": 0.8419875018298626,
      "num_tokens": 1031602973.0,
      "step": 32323
    },
    {
      "entropy": 0.4858549414202571,
      "epoch": 2.973979448076779,
      "grad_norm": 0.15522648394107819,
      "learning_rate": 8.709786242217929e-07,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8480829112231731,
      "num_tokens": 1031634405.0,
      "step": 32324
    },
    {
      "entropy": 0.44577604345977306,
      "epoch": 2.9740714544482207,
      "grad_norm": 0.15098726749420166,
      "learning_rate": 8.679117980801669e-07,
      "loss": 0.4438,
      "mean_token_accuracy": 0.8601540885865688,
      "num_tokens": 1031666155.0,
      "step": 32325
    },
    {
      "entropy": 0.4369589372072369,
      "epoch": 2.974163460819662,
      "grad_norm": 0.1435539275407791,
      "learning_rate": 8.648449719385408e-07,
      "loss": 0.4261,
      "mean_token_accuracy": 0.8662040121853352,
      "num_tokens": 1031698423.0,
      "step": 32326
    },
    {
      "entropy": 0.5370141549501568,
      "epoch": 2.974255467191103,
      "grad_norm": 0.1610584706068039,
      "learning_rate": 8.617781457969149e-07,
      "loss": 0.5191,
      "mean_token_accuracy": 0.8351154699921608,
      "num_tokens": 1031728717.0,
      "step": 32327
    },
    {
      "entropy": 0.5342743480578065,
      "epoch": 2.974347473562544,
      "grad_norm": 0.15936657786369324,
      "learning_rate": 8.587113196552887e-07,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8394401147961617,
      "num_tokens": 1031761078.0,
      "step": 32328
    },
    {
      "entropy": 0.5164936259388924,
      "epoch": 2.9744394799339853,
      "grad_norm": 0.15595905482769012,
      "learning_rate": 8.556444935136629e-07,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8379749990999699,
      "num_tokens": 1031793399.0,
      "step": 32329
    },
    {
      "entropy": 0.5761733530089259,
      "epoch": 2.974531486305427,
      "grad_norm": 0.16754569113254547,
      "learning_rate": 8.525776673720367e-07,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8243238478899002,
      "num_tokens": 1031824858.0,
      "step": 32330
    },
    {
      "entropy": 0.41128361877053976,
      "epoch": 2.974623492676868,
      "grad_norm": 0.14305199682712555,
      "learning_rate": 8.495108412304108e-07,
      "loss": 0.407,
      "mean_token_accuracy": 0.8700186088681221,
      "num_tokens": 1031856985.0,
      "step": 32331
    },
    {
      "entropy": 0.48043999820947647,
      "epoch": 2.974715499048309,
      "grad_norm": 0.15493130683898926,
      "learning_rate": 8.464440150887846e-07,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8499801158905029,
      "num_tokens": 1031888817.0,
      "step": 32332
    },
    {
      "entropy": 0.4739791937172413,
      "epoch": 2.9748075054197503,
      "grad_norm": 0.1523011475801468,
      "learning_rate": 8.433771889471587e-07,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8515085652470589,
      "num_tokens": 1031920891.0,
      "step": 32333
    },
    {
      "entropy": 0.5341730555519462,
      "epoch": 2.9748995117911914,
      "grad_norm": 0.15774112939834595,
      "learning_rate": 8.403103628055326e-07,
      "loss": 0.5242,
      "mean_token_accuracy": 0.8354517817497253,
      "num_tokens": 1031953598.0,
      "step": 32334
    },
    {
      "entropy": 0.47701889742165804,
      "epoch": 2.974991518162633,
      "grad_norm": 0.15166401863098145,
      "learning_rate": 8.372435366639066e-07,
      "loss": 0.4643,
      "mean_token_accuracy": 0.8515617214143276,
      "num_tokens": 1031986007.0,
      "step": 32335
    },
    {
      "entropy": 0.49997374787926674,
      "epoch": 2.975083524534074,
      "grad_norm": 0.15843361616134644,
      "learning_rate": 8.341767105222805e-07,
      "loss": 0.4863,
      "mean_token_accuracy": 0.8472138307988644,
      "num_tokens": 1032017323.0,
      "step": 32336
    },
    {
      "entropy": 0.42298223660327494,
      "epoch": 2.9751755309055152,
      "grad_norm": 0.1464833766222,
      "learning_rate": 8.311098843806545e-07,
      "loss": 0.41,
      "mean_token_accuracy": 0.8684764243662357,
      "num_tokens": 1032048784.0,
      "step": 32337
    },
    {
      "entropy": 0.47787413001060486,
      "epoch": 2.9752675372769564,
      "grad_norm": 0.1536768525838852,
      "learning_rate": 8.280430582390284e-07,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8477063588798046,
      "num_tokens": 1032080169.0,
      "step": 32338
    },
    {
      "entropy": 0.4532371307723224,
      "epoch": 2.9753595436483975,
      "grad_norm": 0.14752760529518127,
      "learning_rate": 8.249762320974024e-07,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8600115329027176,
      "num_tokens": 1032111714.0,
      "step": 32339
    },
    {
      "entropy": 0.48838580353185534,
      "epoch": 2.975451550019839,
      "grad_norm": 0.15293598175048828,
      "learning_rate": 8.219094059557764e-07,
      "loss": 0.4782,
      "mean_token_accuracy": 0.848510030657053,
      "num_tokens": 1032144201.0,
      "step": 32340
    },
    {
      "entropy": 0.47638132236897945,
      "epoch": 2.97554355639128,
      "grad_norm": 0.15861710906028748,
      "learning_rate": 8.188425798141504e-07,
      "loss": 0.4606,
      "mean_token_accuracy": 0.851166345179081,
      "num_tokens": 1032175276.0,
      "step": 32341
    },
    {
      "entropy": 0.539448247407563,
      "epoch": 2.9756355627627213,
      "grad_norm": 0.15797892212867737,
      "learning_rate": 8.157757536725243e-07,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8361501544713974,
      "num_tokens": 1032206568.0,
      "step": 32342
    },
    {
      "entropy": 0.6622688695788383,
      "epoch": 2.9757275691341625,
      "grad_norm": 0.17230388522148132,
      "learning_rate": 8.127089275308983e-07,
      "loss": 0.6567,
      "mean_token_accuracy": 0.7946532927453518,
      "num_tokens": 1032239333.0,
      "step": 32343
    },
    {
      "entropy": 0.5866507821483538,
      "epoch": 2.9758195755056036,
      "grad_norm": 0.16112500429153442,
      "learning_rate": 8.096421013892724e-07,
      "loss": 0.565,
      "mean_token_accuracy": 0.8281792514026165,
      "num_tokens": 1032270928.0,
      "step": 32344
    },
    {
      "entropy": 0.4154736869968474,
      "epoch": 2.975911581877045,
      "grad_norm": 0.14074759185314178,
      "learning_rate": 8.065752752476462e-07,
      "loss": 0.4102,
      "mean_token_accuracy": 0.8703285306692123,
      "num_tokens": 1032303696.0,
      "step": 32345
    },
    {
      "entropy": 0.553498150780797,
      "epoch": 2.9760035882484863,
      "grad_norm": 0.1710578203201294,
      "learning_rate": 8.035084491060203e-07,
      "loss": 0.5381,
      "mean_token_accuracy": 0.8306807242333889,
      "num_tokens": 1032335694.0,
      "step": 32346
    },
    {
      "entropy": 0.4920405038865283,
      "epoch": 2.9760955946199275,
      "grad_norm": 0.15915876626968384,
      "learning_rate": 8.004416229643941e-07,
      "loss": 0.482,
      "mean_token_accuracy": 0.8481951393187046,
      "num_tokens": 1032367348.0,
      "step": 32347
    },
    {
      "entropy": 0.4771037846803665,
      "epoch": 2.9761876009913686,
      "grad_norm": 0.15849730372428894,
      "learning_rate": 7.973747968227682e-07,
      "loss": 0.4688,
      "mean_token_accuracy": 0.8508323766291142,
      "num_tokens": 1032398348.0,
      "step": 32348
    },
    {
      "entropy": 0.5231748120859265,
      "epoch": 2.9762796073628097,
      "grad_norm": 0.15807032585144043,
      "learning_rate": 7.943079706811421e-07,
      "loss": 0.5093,
      "mean_token_accuracy": 0.8395749814808369,
      "num_tokens": 1032430517.0,
      "step": 32349
    },
    {
      "entropy": 0.41840339358896017,
      "epoch": 2.9763716137342513,
      "grad_norm": 0.14719446003437042,
      "learning_rate": 7.912411445395162e-07,
      "loss": 0.4046,
      "mean_token_accuracy": 0.8665513582527637,
      "num_tokens": 1032462244.0,
      "step": 32350
    },
    {
      "entropy": 0.5190475266426802,
      "epoch": 2.9764636201056924,
      "grad_norm": 0.15744948387145996,
      "learning_rate": 7.881743183978901e-07,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8371450081467628,
      "num_tokens": 1032493484.0,
      "step": 32351
    },
    {
      "entropy": 0.596458725980483,
      "epoch": 2.9765556264771336,
      "grad_norm": 0.1668752282857895,
      "learning_rate": 7.851074922562641e-07,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8150375038385391,
      "num_tokens": 1032525195.0,
      "step": 32352
    },
    {
      "entropy": 0.6054442403838038,
      "epoch": 2.9766476328485747,
      "grad_norm": 0.16586807370185852,
      "learning_rate": 7.82040666114638e-07,
      "loss": 0.5933,
      "mean_token_accuracy": 0.8153080530464649,
      "num_tokens": 1032557532.0,
      "step": 32353
    },
    {
      "entropy": 0.5443276851437986,
      "epoch": 2.976739639220016,
      "grad_norm": 0.16577494144439697,
      "learning_rate": 7.78973839973012e-07,
      "loss": 0.5237,
      "mean_token_accuracy": 0.831793300807476,
      "num_tokens": 1032588158.0,
      "step": 32354
    },
    {
      "entropy": 0.5097162239253521,
      "epoch": 2.9768316455914574,
      "grad_norm": 0.15649476647377014,
      "learning_rate": 7.75907013831386e-07,
      "loss": 0.4901,
      "mean_token_accuracy": 0.8453604020178318,
      "num_tokens": 1032620505.0,
      "step": 32355
    },
    {
      "entropy": 0.44206675328314304,
      "epoch": 2.9769236519628985,
      "grad_norm": 0.14804020524024963,
      "learning_rate": 7.728401876897599e-07,
      "loss": 0.4248,
      "mean_token_accuracy": 0.8610699325799942,
      "num_tokens": 1032652556.0,
      "step": 32356
    },
    {
      "entropy": 0.45140325417742133,
      "epoch": 2.9770156583343397,
      "grad_norm": 0.14476290345191956,
      "learning_rate": 7.697733615481339e-07,
      "loss": 0.4444,
      "mean_token_accuracy": 0.8572501391172409,
      "num_tokens": 1032685196.0,
      "step": 32357
    },
    {
      "entropy": 0.4684425871819258,
      "epoch": 2.977107664705781,
      "grad_norm": 0.15304717421531677,
      "learning_rate": 7.667065354065078e-07,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8512157499790192,
      "num_tokens": 1032717214.0,
      "step": 32358
    },
    {
      "entropy": 0.46279097348451614,
      "epoch": 2.977199671077222,
      "grad_norm": 0.15390850603580475,
      "learning_rate": 7.636397092648818e-07,
      "loss": 0.4584,
      "mean_token_accuracy": 0.8539050109684467,
      "num_tokens": 1032748877.0,
      "step": 32359
    },
    {
      "entropy": 0.38386787939816713,
      "epoch": 2.9772916774486635,
      "grad_norm": 0.13441719114780426,
      "learning_rate": 7.605728831232558e-07,
      "loss": 0.3638,
      "mean_token_accuracy": 0.8778478875756264,
      "num_tokens": 1032781137.0,
      "step": 32360
    },
    {
      "entropy": 0.5539260469377041,
      "epoch": 2.9773836838201047,
      "grad_norm": 0.1609802097082138,
      "learning_rate": 7.575060569816298e-07,
      "loss": 0.5508,
      "mean_token_accuracy": 0.8298937827348709,
      "num_tokens": 1032812914.0,
      "step": 32361
    },
    {
      "entropy": 0.5146436078939587,
      "epoch": 2.977475690191546,
      "grad_norm": 0.15235276520252228,
      "learning_rate": 7.544392308400038e-07,
      "loss": 0.5054,
      "mean_token_accuracy": 0.841616790741682,
      "num_tokens": 1032845263.0,
      "step": 32362
    },
    {
      "entropy": 0.5955823343247175,
      "epoch": 2.977567696562987,
      "grad_norm": 0.17130300402641296,
      "learning_rate": 7.513724046983777e-07,
      "loss": 0.5894,
      "mean_token_accuracy": 0.8142514899373055,
      "num_tokens": 1032876536.0,
      "step": 32363
    },
    {
      "entropy": 0.44159476459026337,
      "epoch": 2.977659702934428,
      "grad_norm": 0.15183141827583313,
      "learning_rate": 7.483055785567517e-07,
      "loss": 0.4353,
      "mean_token_accuracy": 0.8587217628955841,
      "num_tokens": 1032908170.0,
      "step": 32364
    },
    {
      "entropy": 0.46571092400699854,
      "epoch": 2.9777517093058696,
      "grad_norm": 0.1484919637441635,
      "learning_rate": 7.452387524151256e-07,
      "loss": 0.4549,
      "mean_token_accuracy": 0.8555879257619381,
      "num_tokens": 1032940730.0,
      "step": 32365
    },
    {
      "entropy": 0.5793533176183701,
      "epoch": 2.9778437156773108,
      "grad_norm": 0.15906330943107605,
      "learning_rate": 7.421719262734996e-07,
      "loss": 0.554,
      "mean_token_accuracy": 0.826703418046236,
      "num_tokens": 1032973366.0,
      "step": 32366
    },
    {
      "entropy": 0.4179733913624659,
      "epoch": 2.977935722048752,
      "grad_norm": 0.14022445678710938,
      "learning_rate": 7.391051001318735e-07,
      "loss": 0.4089,
      "mean_token_accuracy": 0.8700457476079464,
      "num_tokens": 1033006086.0,
      "step": 32367
    },
    {
      "entropy": 0.5481031723320484,
      "epoch": 2.978027728420193,
      "grad_norm": 0.1660323292016983,
      "learning_rate": 7.360382739902475e-07,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8310101479291916,
      "num_tokens": 1033037071.0,
      "step": 32368
    },
    {
      "entropy": 0.3996511506848037,
      "epoch": 2.978119734791634,
      "grad_norm": 0.13404303789138794,
      "learning_rate": 7.329714478486214e-07,
      "loss": 0.3886,
      "mean_token_accuracy": 0.8794882744550705,
      "num_tokens": 1033069839.0,
      "step": 32369
    },
    {
      "entropy": 0.5792920328676701,
      "epoch": 2.9782117411630757,
      "grad_norm": 0.1672493815422058,
      "learning_rate": 7.299046217069955e-07,
      "loss": 0.5622,
      "mean_token_accuracy": 0.8210929334163666,
      "num_tokens": 1033100722.0,
      "step": 32370
    },
    {
      "entropy": 0.532009051181376,
      "epoch": 2.978303747534517,
      "grad_norm": 0.15936079621315002,
      "learning_rate": 7.268377955653694e-07,
      "loss": 0.5169,
      "mean_token_accuracy": 0.8384346254169941,
      "num_tokens": 1033131818.0,
      "step": 32371
    },
    {
      "entropy": 0.4865764924325049,
      "epoch": 2.978395753905958,
      "grad_norm": 0.15249191224575043,
      "learning_rate": 7.237709694237434e-07,
      "loss": 0.4682,
      "mean_token_accuracy": 0.848790168762207,
      "num_tokens": 1033163980.0,
      "step": 32372
    },
    {
      "entropy": 0.39257694291882217,
      "epoch": 2.978487760277399,
      "grad_norm": 0.1412283480167389,
      "learning_rate": 7.207041432821174e-07,
      "loss": 0.3798,
      "mean_token_accuracy": 0.8746127262711525,
      "num_tokens": 1033196250.0,
      "step": 32373
    },
    {
      "entropy": 0.5199130759574473,
      "epoch": 2.9785797666488403,
      "grad_norm": 0.1583741158246994,
      "learning_rate": 7.176373171404914e-07,
      "loss": 0.5063,
      "mean_token_accuracy": 0.8409878127276897,
      "num_tokens": 1033227415.0,
      "step": 32374
    },
    {
      "entropy": 0.39308267552405596,
      "epoch": 2.978671773020282,
      "grad_norm": 0.14271192252635956,
      "learning_rate": 7.145704909988653e-07,
      "loss": 0.3783,
      "mean_token_accuracy": 0.8743228949606419,
      "num_tokens": 1033258535.0,
      "step": 32375
    },
    {
      "entropy": 0.6001141089946032,
      "epoch": 2.978763779391723,
      "grad_norm": 0.1686936467885971,
      "learning_rate": 7.115036648572393e-07,
      "loss": 0.5779,
      "mean_token_accuracy": 0.8151560761034489,
      "num_tokens": 1033290337.0,
      "step": 32376
    },
    {
      "entropy": 0.49579619243741035,
      "epoch": 2.978855785763164,
      "grad_norm": 0.15801894664764404,
      "learning_rate": 7.084368387156133e-07,
      "loss": 0.481,
      "mean_token_accuracy": 0.8465945944190025,
      "num_tokens": 1033322072.0,
      "step": 32377
    },
    {
      "entropy": 0.4890263043344021,
      "epoch": 2.9789477921346053,
      "grad_norm": 0.15454551577568054,
      "learning_rate": 7.053700125739872e-07,
      "loss": 0.4866,
      "mean_token_accuracy": 0.8471232503652573,
      "num_tokens": 1033353938.0,
      "step": 32378
    },
    {
      "entropy": 0.5717684382107109,
      "epoch": 2.9790397985060464,
      "grad_norm": 0.16571691632270813,
      "learning_rate": 7.023031864323612e-07,
      "loss": 0.5684,
      "mean_token_accuracy": 0.8217038549482822,
      "num_tokens": 1033385885.0,
      "step": 32379
    },
    {
      "entropy": 0.5796127263456583,
      "epoch": 2.979131804877488,
      "grad_norm": 0.16700194776058197,
      "learning_rate": 6.992363602907351e-07,
      "loss": 0.5651,
      "mean_token_accuracy": 0.8254427202045918,
      "num_tokens": 1033417817.0,
      "step": 32380
    },
    {
      "entropy": 0.49641203135252,
      "epoch": 2.979223811248929,
      "grad_norm": 0.14950281381607056,
      "learning_rate": 6.961695341491092e-07,
      "loss": 0.4775,
      "mean_token_accuracy": 0.8464965559542179,
      "num_tokens": 1033449837.0,
      "step": 32381
    },
    {
      "entropy": 0.46788534335792065,
      "epoch": 2.9793158176203702,
      "grad_norm": 0.15077327191829681,
      "learning_rate": 6.931027080074831e-07,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8541533164680004,
      "num_tokens": 1033481789.0,
      "step": 32382
    },
    {
      "entropy": 0.6074517611414194,
      "epoch": 2.9794078239918114,
      "grad_norm": 0.1712934672832489,
      "learning_rate": 6.900358818658571e-07,
      "loss": 0.5924,
      "mean_token_accuracy": 0.8129815980792046,
      "num_tokens": 1033513977.0,
      "step": 32383
    },
    {
      "entropy": 0.492969511076808,
      "epoch": 2.9794998303632525,
      "grad_norm": 0.1518772393465042,
      "learning_rate": 6.86969055724231e-07,
      "loss": 0.4823,
      "mean_token_accuracy": 0.848633885383606,
      "num_tokens": 1033544915.0,
      "step": 32384
    },
    {
      "entropy": 0.5108314715325832,
      "epoch": 2.979591836734694,
      "grad_norm": 0.15716740489006042,
      "learning_rate": 6.83902229582605e-07,
      "loss": 0.498,
      "mean_token_accuracy": 0.8413144275546074,
      "num_tokens": 1033577544.0,
      "step": 32385
    },
    {
      "entropy": 0.5090573001652956,
      "epoch": 2.979683843106135,
      "grad_norm": 0.156625896692276,
      "learning_rate": 6.808354034409789e-07,
      "loss": 0.4962,
      "mean_token_accuracy": 0.8407343812286854,
      "num_tokens": 1033609447.0,
      "step": 32386
    },
    {
      "entropy": 0.5834165681153536,
      "epoch": 2.9797758494775763,
      "grad_norm": 0.16583675146102905,
      "learning_rate": 6.777685772993529e-07,
      "loss": 0.5774,
      "mean_token_accuracy": 0.82062678784132,
      "num_tokens": 1033641822.0,
      "step": 32387
    },
    {
      "entropy": 0.5313225481659174,
      "epoch": 2.9798678558490175,
      "grad_norm": 0.16413050889968872,
      "learning_rate": 6.747017511577269e-07,
      "loss": 0.5176,
      "mean_token_accuracy": 0.8374691531062126,
      "num_tokens": 1033672817.0,
      "step": 32388
    },
    {
      "entropy": 0.5170269263908267,
      "epoch": 2.9799598622204586,
      "grad_norm": 0.15849743783473969,
      "learning_rate": 6.716349250161008e-07,
      "loss": 0.5018,
      "mean_token_accuracy": 0.8384084701538086,
      "num_tokens": 1033705084.0,
      "step": 32389
    },
    {
      "entropy": 0.4999324092641473,
      "epoch": 2.9800518685919,
      "grad_norm": 0.15820729732513428,
      "learning_rate": 6.685680988744748e-07,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8379489034414291,
      "num_tokens": 1033736897.0,
      "step": 32390
    },
    {
      "entropy": 0.5188646268798038,
      "epoch": 2.9801438749633413,
      "grad_norm": 0.15695111453533173,
      "learning_rate": 6.655012727328488e-07,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8427647650241852,
      "num_tokens": 1033769230.0,
      "step": 32391
    },
    {
      "entropy": 0.5153922635363415,
      "epoch": 2.9802358813347825,
      "grad_norm": 0.16262952983379364,
      "learning_rate": 6.624344465912228e-07,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8365371525287628,
      "num_tokens": 1033801307.0,
      "step": 32392
    },
    {
      "entropy": 0.5108836246654391,
      "epoch": 2.9803278877062236,
      "grad_norm": 0.15898142755031586,
      "learning_rate": 6.593676204495967e-07,
      "loss": 0.5068,
      "mean_token_accuracy": 0.8373308144509792,
      "num_tokens": 1033832879.0,
      "step": 32393
    },
    {
      "entropy": 0.5345505927689373,
      "epoch": 2.9804198940776647,
      "grad_norm": 0.1584562361240387,
      "learning_rate": 6.563007943079707e-07,
      "loss": 0.5252,
      "mean_token_accuracy": 0.8317182622849941,
      "num_tokens": 1033865039.0,
      "step": 32394
    },
    {
      "entropy": 0.47770991176366806,
      "epoch": 2.9805119004491063,
      "grad_norm": 0.151068776845932,
      "learning_rate": 6.532339681663447e-07,
      "loss": 0.4676,
      "mean_token_accuracy": 0.8498101457953453,
      "num_tokens": 1033896436.0,
      "step": 32395
    },
    {
      "entropy": 0.5272120493464172,
      "epoch": 2.9806039068205474,
      "grad_norm": 0.16224311292171478,
      "learning_rate": 6.501671420247187e-07,
      "loss": 0.5196,
      "mean_token_accuracy": 0.8370568603277206,
      "num_tokens": 1033927989.0,
      "step": 32396
    },
    {
      "entropy": 0.5333994266111404,
      "epoch": 2.9806959131919886,
      "grad_norm": 0.1563442498445511,
      "learning_rate": 6.471003158830926e-07,
      "loss": 0.5154,
      "mean_token_accuracy": 0.8350666724145412,
      "num_tokens": 1033960028.0,
      "step": 32397
    },
    {
      "entropy": 0.5128248082473874,
      "epoch": 2.9807879195634297,
      "grad_norm": 0.15992765128612518,
      "learning_rate": 6.440334897414666e-07,
      "loss": 0.4946,
      "mean_token_accuracy": 0.842837642878294,
      "num_tokens": 1033992136.0,
      "step": 32398
    },
    {
      "entropy": 0.5311938747763634,
      "epoch": 2.980879925934871,
      "grad_norm": 0.15865273773670197,
      "learning_rate": 6.409666635998405e-07,
      "loss": 0.511,
      "mean_token_accuracy": 0.8365944251418114,
      "num_tokens": 1034024107.0,
      "step": 32399
    },
    {
      "entropy": 0.49067461118102074,
      "epoch": 2.9809719323063124,
      "grad_norm": 0.15559089183807373,
      "learning_rate": 6.378998374582145e-07,
      "loss": 0.4775,
      "mean_token_accuracy": 0.8462906554341316,
      "num_tokens": 1034056005.0,
      "step": 32400
    },
    {
      "entropy": 0.600898802280426,
      "epoch": 2.9810639386777535,
      "grad_norm": 0.172745019197464,
      "learning_rate": 6.348330113165885e-07,
      "loss": 0.5984,
      "mean_token_accuracy": 0.8117482028901577,
      "num_tokens": 1034088447.0,
      "step": 32401
    },
    {
      "entropy": 0.5015659015625715,
      "epoch": 2.9811559450491947,
      "grad_norm": 0.1533370167016983,
      "learning_rate": 6.317661851749625e-07,
      "loss": 0.4874,
      "mean_token_accuracy": 0.8453106135129929,
      "num_tokens": 1034120670.0,
      "step": 32402
    },
    {
      "entropy": 0.5055651115253568,
      "epoch": 2.981247951420636,
      "grad_norm": 0.15512211620807648,
      "learning_rate": 6.286993590333365e-07,
      "loss": 0.4915,
      "mean_token_accuracy": 0.842899739742279,
      "num_tokens": 1034152995.0,
      "step": 32403
    },
    {
      "entropy": 0.5480817835777998,
      "epoch": 2.981339957792077,
      "grad_norm": 0.16276682913303375,
      "learning_rate": 6.256325328917104e-07,
      "loss": 0.5283,
      "mean_token_accuracy": 0.8326586149632931,
      "num_tokens": 1034183971.0,
      "step": 32404
    },
    {
      "entropy": 0.4380393009632826,
      "epoch": 2.9814319641635185,
      "grad_norm": 0.14590735733509064,
      "learning_rate": 6.225657067500844e-07,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8641374632716179,
      "num_tokens": 1034215734.0,
      "step": 32405
    },
    {
      "entropy": 0.5036380048841238,
      "epoch": 2.9815239705349597,
      "grad_norm": 0.15923874080181122,
      "learning_rate": 6.194988806084583e-07,
      "loss": 0.4961,
      "mean_token_accuracy": 0.8413756377995014,
      "num_tokens": 1034247205.0,
      "step": 32406
    },
    {
      "entropy": 0.5122233163565397,
      "epoch": 2.981615976906401,
      "grad_norm": 0.1618775725364685,
      "learning_rate": 6.164320544668323e-07,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8423952460289001,
      "num_tokens": 1034278415.0,
      "step": 32407
    },
    {
      "entropy": 0.47216995619237423,
      "epoch": 2.981707983277842,
      "grad_norm": 0.1492568850517273,
      "learning_rate": 6.133652283252062e-07,
      "loss": 0.4619,
      "mean_token_accuracy": 0.8522838316857815,
      "num_tokens": 1034311183.0,
      "step": 32408
    },
    {
      "entropy": 0.48691192362457514,
      "epoch": 2.981799989649283,
      "grad_norm": 0.15752644836902618,
      "learning_rate": 6.102984021835802e-07,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8498133607208729,
      "num_tokens": 1034343263.0,
      "step": 32409
    },
    {
      "entropy": 0.4736970430240035,
      "epoch": 2.9818919960207246,
      "grad_norm": 0.15401691198349,
      "learning_rate": 6.072315760419541e-07,
      "loss": 0.4662,
      "mean_token_accuracy": 0.853071641176939,
      "num_tokens": 1034375508.0,
      "step": 32410
    },
    {
      "entropy": 0.5312348743900657,
      "epoch": 2.9819840023921658,
      "grad_norm": 0.1606704592704773,
      "learning_rate": 6.041647499003282e-07,
      "loss": 0.5186,
      "mean_token_accuracy": 0.8328997828066349,
      "num_tokens": 1034407500.0,
      "step": 32411
    },
    {
      "entropy": 0.5021482273004949,
      "epoch": 2.982076008763607,
      "grad_norm": 0.16126272082328796,
      "learning_rate": 6.010979237587022e-07,
      "loss": 0.4932,
      "mean_token_accuracy": 0.84629962220788,
      "num_tokens": 1034439093.0,
      "step": 32412
    },
    {
      "entropy": 0.5436648358590901,
      "epoch": 2.982168015135048,
      "grad_norm": 0.1628803163766861,
      "learning_rate": 5.980310976170761e-07,
      "loss": 0.5236,
      "mean_token_accuracy": 0.8311222419142723,
      "num_tokens": 1034470757.0,
      "step": 32413
    },
    {
      "entropy": 0.46428613690659404,
      "epoch": 2.982260021506489,
      "grad_norm": 0.1507827639579773,
      "learning_rate": 5.949642714754501e-07,
      "loss": 0.4465,
      "mean_token_accuracy": 0.856250386685133,
      "num_tokens": 1034503144.0,
      "step": 32414
    },
    {
      "entropy": 0.5087641859427094,
      "epoch": 2.9823520278779307,
      "grad_norm": 0.1608603447675705,
      "learning_rate": 5.918974453338241e-07,
      "loss": 0.4995,
      "mean_token_accuracy": 0.8397073373198509,
      "num_tokens": 1034534962.0,
      "step": 32415
    },
    {
      "entropy": 0.45786841213703156,
      "epoch": 2.982444034249372,
      "grad_norm": 0.1543562263250351,
      "learning_rate": 5.888306191921981e-07,
      "loss": 0.4511,
      "mean_token_accuracy": 0.8567524179816246,
      "num_tokens": 1034566579.0,
      "step": 32416
    },
    {
      "entropy": 0.5557301663793623,
      "epoch": 2.982536040620813,
      "grad_norm": 0.15800222754478455,
      "learning_rate": 5.85763793050572e-07,
      "loss": 0.5363,
      "mean_token_accuracy": 0.8290520198643208,
      "num_tokens": 1034598242.0,
      "step": 32417
    },
    {
      "entropy": 0.5351392191369087,
      "epoch": 2.982628046992254,
      "grad_norm": 0.16277658939361572,
      "learning_rate": 5.82696966908946e-07,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8311499878764153,
      "num_tokens": 1034630692.0,
      "step": 32418
    },
    {
      "entropy": 0.5112783666700125,
      "epoch": 2.9827200533636953,
      "grad_norm": 0.15182967483997345,
      "learning_rate": 5.796301407673199e-07,
      "loss": 0.495,
      "mean_token_accuracy": 0.843654926866293,
      "num_tokens": 1034662508.0,
      "step": 32419
    },
    {
      "entropy": 0.5833793114870787,
      "epoch": 2.982812059735137,
      "grad_norm": 0.1681280881166458,
      "learning_rate": 5.76563314625694e-07,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8179415315389633,
      "num_tokens": 1034694513.0,
      "step": 32420
    },
    {
      "entropy": 0.47089724033139646,
      "epoch": 2.982904066106578,
      "grad_norm": 0.15425701439380646,
      "learning_rate": 5.734964884840679e-07,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8531941697001457,
      "num_tokens": 1034726573.0,
      "step": 32421
    },
    {
      "entropy": 0.43106127879582345,
      "epoch": 2.982996072478019,
      "grad_norm": 0.14276200532913208,
      "learning_rate": 5.704296623424419e-07,
      "loss": 0.4154,
      "mean_token_accuracy": 0.8693778999149799,
      "num_tokens": 1034758581.0,
      "step": 32422
    },
    {
      "entropy": 0.4958057962357998,
      "epoch": 2.9830880788494603,
      "grad_norm": 0.1542145162820816,
      "learning_rate": 5.673628362008158e-07,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8453873321413994,
      "num_tokens": 1034790381.0,
      "step": 32423
    },
    {
      "entropy": 0.48020518384873867,
      "epoch": 2.9831800852209014,
      "grad_norm": 0.1485714614391327,
      "learning_rate": 5.642960100591898e-07,
      "loss": 0.4725,
      "mean_token_accuracy": 0.8514685109257698,
      "num_tokens": 1034822763.0,
      "step": 32424
    },
    {
      "entropy": 0.5214766198769212,
      "epoch": 2.983272091592343,
      "grad_norm": 0.1548001915216446,
      "learning_rate": 5.612291839175637e-07,
      "loss": 0.4923,
      "mean_token_accuracy": 0.8436683192849159,
      "num_tokens": 1034853649.0,
      "step": 32425
    },
    {
      "entropy": 0.4134130470920354,
      "epoch": 2.983364097963784,
      "grad_norm": 0.13942936062812805,
      "learning_rate": 5.581623577759377e-07,
      "loss": 0.391,
      "mean_token_accuracy": 0.8705797269940376,
      "num_tokens": 1034885289.0,
      "step": 32426
    },
    {
      "entropy": 0.532732049934566,
      "epoch": 2.9834561043352252,
      "grad_norm": 0.16073979437351227,
      "learning_rate": 5.550955316343117e-07,
      "loss": 0.5191,
      "mean_token_accuracy": 0.8363544940948486,
      "num_tokens": 1034917458.0,
      "step": 32427
    },
    {
      "entropy": 0.49507930176332593,
      "epoch": 2.9835481107066664,
      "grad_norm": 0.14904631674289703,
      "learning_rate": 5.520287054926856e-07,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8501357398927212,
      "num_tokens": 1034949579.0,
      "step": 32428
    },
    {
      "entropy": 0.42733982810750604,
      "epoch": 2.9836401170781075,
      "grad_norm": 0.14411622285842896,
      "learning_rate": 5.489618793510596e-07,
      "loss": 0.4096,
      "mean_token_accuracy": 0.8714241199195385,
      "num_tokens": 1034981606.0,
      "step": 32429
    },
    {
      "entropy": 0.4476691447198391,
      "epoch": 2.983732123449549,
      "grad_norm": 0.1464494913816452,
      "learning_rate": 5.458950532094335e-07,
      "loss": 0.4348,
      "mean_token_accuracy": 0.8627040199935436,
      "num_tokens": 1035013326.0,
      "step": 32430
    },
    {
      "entropy": 0.48783194413408637,
      "epoch": 2.98382412982099,
      "grad_norm": 0.15200676023960114,
      "learning_rate": 5.428282270678076e-07,
      "loss": 0.474,
      "mean_token_accuracy": 0.8491176925599575,
      "num_tokens": 1035045652.0,
      "step": 32431
    },
    {
      "entropy": 0.49444226175546646,
      "epoch": 2.9839161361924313,
      "grad_norm": 0.1558907926082611,
      "learning_rate": 5.397614009261815e-07,
      "loss": 0.4804,
      "mean_token_accuracy": 0.8438982702791691,
      "num_tokens": 1035077934.0,
      "step": 32432
    },
    {
      "entropy": 0.5037533710710704,
      "epoch": 2.9840081425638725,
      "grad_norm": 0.16043145954608917,
      "learning_rate": 5.366945747845555e-07,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8389142528176308,
      "num_tokens": 1035109526.0,
      "step": 32433
    },
    {
      "entropy": 0.5275922371074557,
      "epoch": 2.9841001489353136,
      "grad_norm": 0.16426421701908112,
      "learning_rate": 5.336277486429294e-07,
      "loss": 0.529,
      "mean_token_accuracy": 0.8329727984964848,
      "num_tokens": 1035140399.0,
      "step": 32434
    },
    {
      "entropy": 0.48451465368270874,
      "epoch": 2.984192155306755,
      "grad_norm": 0.15208663046360016,
      "learning_rate": 5.305609225013035e-07,
      "loss": 0.4676,
      "mean_token_accuracy": 0.8503731414675713,
      "num_tokens": 1035172707.0,
      "step": 32435
    },
    {
      "entropy": 0.5183189865201712,
      "epoch": 2.9842841616781963,
      "grad_norm": 0.15731850266456604,
      "learning_rate": 5.274940963596774e-07,
      "loss": 0.5073,
      "mean_token_accuracy": 0.8438038043677807,
      "num_tokens": 1035204898.0,
      "step": 32436
    },
    {
      "entropy": 0.5209946222603321,
      "epoch": 2.9843761680496375,
      "grad_norm": 0.16736666858196259,
      "learning_rate": 5.244272702180514e-07,
      "loss": 0.5097,
      "mean_token_accuracy": 0.8339290283620358,
      "num_tokens": 1035236427.0,
      "step": 32437
    },
    {
      "entropy": 0.5579792829230428,
      "epoch": 2.9844681744210786,
      "grad_norm": 0.16063867509365082,
      "learning_rate": 5.213604440764253e-07,
      "loss": 0.5403,
      "mean_token_accuracy": 0.8277308791875839,
      "num_tokens": 1035268424.0,
      "step": 32438
    },
    {
      "entropy": 0.48381583066657186,
      "epoch": 2.9845601807925197,
      "grad_norm": 0.15525005757808685,
      "learning_rate": 5.182936179347993e-07,
      "loss": 0.467,
      "mean_token_accuracy": 0.8498439602553844,
      "num_tokens": 1035300281.0,
      "step": 32439
    },
    {
      "entropy": 0.5619140658527613,
      "epoch": 2.9846521871639613,
      "grad_norm": 0.1635187268257141,
      "learning_rate": 5.152267917931732e-07,
      "loss": 0.5561,
      "mean_token_accuracy": 0.8281073607504368,
      "num_tokens": 1035332677.0,
      "step": 32440
    },
    {
      "entropy": 0.4576850577723235,
      "epoch": 2.9847441935354024,
      "grad_norm": 0.15399593114852905,
      "learning_rate": 5.121599656515473e-07,
      "loss": 0.4517,
      "mean_token_accuracy": 0.8561508245766163,
      "num_tokens": 1035363909.0,
      "step": 32441
    },
    {
      "entropy": 0.4891635235399008,
      "epoch": 2.9848361999068436,
      "grad_norm": 0.15664450824260712,
      "learning_rate": 5.090931395099213e-07,
      "loss": 0.481,
      "mean_token_accuracy": 0.8482675403356552,
      "num_tokens": 1035396418.0,
      "step": 32442
    },
    {
      "entropy": 0.518792262300849,
      "epoch": 2.9849282062782847,
      "grad_norm": 0.16215862333774567,
      "learning_rate": 5.060263133682952e-07,
      "loss": 0.5108,
      "mean_token_accuracy": 0.835899256169796,
      "num_tokens": 1035428123.0,
      "step": 32443
    },
    {
      "entropy": 0.5640861731953919,
      "epoch": 2.985020212649726,
      "grad_norm": 0.16569873690605164,
      "learning_rate": 5.029594872266692e-07,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8276932090520859,
      "num_tokens": 1035459535.0,
      "step": 32444
    },
    {
      "entropy": 0.5352602172642946,
      "epoch": 2.9851122190211674,
      "grad_norm": 0.15956901013851166,
      "learning_rate": 4.998926610850431e-07,
      "loss": 0.5184,
      "mean_token_accuracy": 0.8345181979238987,
      "num_tokens": 1035491888.0,
      "step": 32445
    },
    {
      "entropy": 0.5276532471179962,
      "epoch": 2.9852042253926085,
      "grad_norm": 0.15406645834445953,
      "learning_rate": 4.968258349434171e-07,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8392791859805584,
      "num_tokens": 1035523656.0,
      "step": 32446
    },
    {
      "entropy": 0.4452056121081114,
      "epoch": 2.9852962317640497,
      "grad_norm": 0.1429823487997055,
      "learning_rate": 4.93759008801791e-07,
      "loss": 0.4356,
      "mean_token_accuracy": 0.8612748980522156,
      "num_tokens": 1035556395.0,
      "step": 32447
    },
    {
      "entropy": 0.5504662431776524,
      "epoch": 2.985388238135491,
      "grad_norm": 0.16591575741767883,
      "learning_rate": 4.90692182660165e-07,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8289615213871002,
      "num_tokens": 1035588800.0,
      "step": 32448
    },
    {
      "entropy": 0.5442882813513279,
      "epoch": 2.985480244506932,
      "grad_norm": 0.16162005066871643,
      "learning_rate": 4.876253565185389e-07,
      "loss": 0.5262,
      "mean_token_accuracy": 0.831499245017767,
      "num_tokens": 1035621314.0,
      "step": 32449
    },
    {
      "entropy": 0.5040675206109881,
      "epoch": 2.9855722508783735,
      "grad_norm": 0.16049279272556305,
      "learning_rate": 4.845585303769129e-07,
      "loss": 0.4915,
      "mean_token_accuracy": 0.8463453687727451,
      "num_tokens": 1035652940.0,
      "step": 32450
    },
    {
      "entropy": 0.48174608778208494,
      "epoch": 2.9856642572498147,
      "grad_norm": 0.15753601491451263,
      "learning_rate": 4.81491704235287e-07,
      "loss": 0.47,
      "mean_token_accuracy": 0.849089901894331,
      "num_tokens": 1035684216.0,
      "step": 32451
    },
    {
      "entropy": 0.5451541263610125,
      "epoch": 2.985756263621256,
      "grad_norm": 0.16259737312793732,
      "learning_rate": 4.784248780936609e-07,
      "loss": 0.5325,
      "mean_token_accuracy": 0.8334095291793346,
      "num_tokens": 1035716768.0,
      "step": 32452
    },
    {
      "entropy": 0.4907906934386119,
      "epoch": 2.985848269992697,
      "grad_norm": 0.14767985045909882,
      "learning_rate": 4.753580519520348e-07,
      "loss": 0.4721,
      "mean_token_accuracy": 0.8511421009898186,
      "num_tokens": 1035749107.0,
      "step": 32453
    },
    {
      "entropy": 0.508453257381916,
      "epoch": 2.985940276364138,
      "grad_norm": 0.15876604616641998,
      "learning_rate": 4.722912258104088e-07,
      "loss": 0.5081,
      "mean_token_accuracy": 0.836811039596796,
      "num_tokens": 1035781083.0,
      "step": 32454
    },
    {
      "entropy": 0.5439638067036867,
      "epoch": 2.9860322827355796,
      "grad_norm": 0.1715569943189621,
      "learning_rate": 4.6922439966878274e-07,
      "loss": 0.5271,
      "mean_token_accuracy": 0.8363178707659245,
      "num_tokens": 1035811862.0,
      "step": 32455
    },
    {
      "entropy": 0.5053128153085709,
      "epoch": 2.9861242891070208,
      "grad_norm": 0.16296438872814178,
      "learning_rate": 4.661575735271568e-07,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8411299772560596,
      "num_tokens": 1035843266.0,
      "step": 32456
    },
    {
      "entropy": 0.46998933143913746,
      "epoch": 2.986216295478462,
      "grad_norm": 0.1516619622707367,
      "learning_rate": 4.6309074738553076e-07,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8562774658203125,
      "num_tokens": 1035875356.0,
      "step": 32457
    },
    {
      "entropy": 0.43248677952215075,
      "epoch": 2.986308301849903,
      "grad_norm": 0.1472802460193634,
      "learning_rate": 4.600239212439047e-07,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8615469373762608,
      "num_tokens": 1035907915.0,
      "step": 32458
    },
    {
      "entropy": 0.4088680420536548,
      "epoch": 2.986400308221344,
      "grad_norm": 0.1489328294992447,
      "learning_rate": 4.569570951022787e-07,
      "loss": 0.3995,
      "mean_token_accuracy": 0.871172595769167,
      "num_tokens": 1035940052.0,
      "step": 32459
    },
    {
      "entropy": 0.45934921968728304,
      "epoch": 2.9864923145927857,
      "grad_norm": 0.15381614863872528,
      "learning_rate": 4.5389026896065264e-07,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8537452667951584,
      "num_tokens": 1035972257.0,
      "step": 32460
    },
    {
      "entropy": 0.4884538119658828,
      "epoch": 2.986584320964227,
      "grad_norm": 0.14908434450626373,
      "learning_rate": 4.5082344281902665e-07,
      "loss": 0.479,
      "mean_token_accuracy": 0.8489637076854706,
      "num_tokens": 1036004846.0,
      "step": 32461
    },
    {
      "entropy": 0.4258498589042574,
      "epoch": 2.986676327335668,
      "grad_norm": 0.1438973993062973,
      "learning_rate": 4.477566166774006e-07,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8689714558422565,
      "num_tokens": 1036036631.0,
      "step": 32462
    },
    {
      "entropy": 0.41494948975741863,
      "epoch": 2.986768333707109,
      "grad_norm": 0.14181436598300934,
      "learning_rate": 4.4468979053577457e-07,
      "loss": 0.4008,
      "mean_token_accuracy": 0.8724239878356457,
      "num_tokens": 1036069186.0,
      "step": 32463
    },
    {
      "entropy": 0.5779502540826797,
      "epoch": 2.9868603400785503,
      "grad_norm": 0.1675933599472046,
      "learning_rate": 4.4162296439414853e-07,
      "loss": 0.5732,
      "mean_token_accuracy": 0.8168110139667988,
      "num_tokens": 1036101081.0,
      "step": 32464
    },
    {
      "entropy": 0.5840766374021769,
      "epoch": 2.986952346449992,
      "grad_norm": 0.16748476028442383,
      "learning_rate": 4.385561382525225e-07,
      "loss": 0.5716,
      "mean_token_accuracy": 0.8178421184420586,
      "num_tokens": 1036133502.0,
      "step": 32465
    },
    {
      "entropy": 0.3054589161183685,
      "epoch": 2.987044352821433,
      "grad_norm": 0.12545467913150787,
      "learning_rate": 4.3548931211089645e-07,
      "loss": 0.2812,
      "mean_token_accuracy": 0.9048275947570801,
      "num_tokens": 1036164567.0,
      "step": 32466
    },
    {
      "entropy": 0.45418372098356485,
      "epoch": 2.987136359192874,
      "grad_norm": 0.15009424090385437,
      "learning_rate": 4.324224859692704e-07,
      "loss": 0.4389,
      "mean_token_accuracy": 0.8586706891655922,
      "num_tokens": 1036196328.0,
      "step": 32467
    },
    {
      "entropy": 0.5617285231128335,
      "epoch": 2.9872283655643153,
      "grad_norm": 0.1659560203552246,
      "learning_rate": 4.2935565982764437e-07,
      "loss": 0.5407,
      "mean_token_accuracy": 0.8313969410955906,
      "num_tokens": 1036228637.0,
      "step": 32468
    },
    {
      "entropy": 0.4577325638383627,
      "epoch": 2.9873203719357564,
      "grad_norm": 0.14699050784111023,
      "learning_rate": 4.262888336860183e-07,
      "loss": 0.4443,
      "mean_token_accuracy": 0.8584665507078171,
      "num_tokens": 1036261148.0,
      "step": 32469
    },
    {
      "entropy": 0.4761097077280283,
      "epoch": 2.987412378307198,
      "grad_norm": 0.1524362862110138,
      "learning_rate": 4.232220075443923e-07,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8513114601373672,
      "num_tokens": 1036292977.0,
      "step": 32470
    },
    {
      "entropy": 0.4344371429178864,
      "epoch": 2.987504384678639,
      "grad_norm": 0.14225421845912933,
      "learning_rate": 4.201551814027663e-07,
      "loss": 0.4255,
      "mean_token_accuracy": 0.8671648874878883,
      "num_tokens": 1036325397.0,
      "step": 32471
    },
    {
      "entropy": 0.5290812011808157,
      "epoch": 2.9875963910500802,
      "grad_norm": 0.16516035795211792,
      "learning_rate": 4.1708835526114026e-07,
      "loss": 0.5211,
      "mean_token_accuracy": 0.8339022472500801,
      "num_tokens": 1036356574.0,
      "step": 32472
    },
    {
      "entropy": 0.529133060015738,
      "epoch": 2.9876883974215214,
      "grad_norm": 0.15680481493473053,
      "learning_rate": 4.140215291195142e-07,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8331286534667015,
      "num_tokens": 1036388564.0,
      "step": 32473
    },
    {
      "entropy": 0.6212648991495371,
      "epoch": 2.9877804037929625,
      "grad_norm": 0.1682511270046234,
      "learning_rate": 4.109547029778882e-07,
      "loss": 0.6104,
      "mean_token_accuracy": 0.8095633573830128,
      "num_tokens": 1036420610.0,
      "step": 32474
    },
    {
      "entropy": 0.447754240129143,
      "epoch": 2.987872410164404,
      "grad_norm": 0.1482512205839157,
      "learning_rate": 4.0788787683626214e-07,
      "loss": 0.4401,
      "mean_token_accuracy": 0.8587283939123154,
      "num_tokens": 1036452989.0,
      "step": 32475
    },
    {
      "entropy": 0.46582782082259655,
      "epoch": 2.987964416535845,
      "grad_norm": 0.14510758221149445,
      "learning_rate": 4.048210506946362e-07,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8555235005915165,
      "num_tokens": 1036485299.0,
      "step": 32476
    },
    {
      "entropy": 0.5384462028741837,
      "epoch": 2.9880564229072863,
      "grad_norm": 0.1588117927312851,
      "learning_rate": 4.0175422455301016e-07,
      "loss": 0.5217,
      "mean_token_accuracy": 0.8358105197548866,
      "num_tokens": 1036517073.0,
      "step": 32477
    },
    {
      "entropy": 0.49430576991289854,
      "epoch": 2.9881484292787275,
      "grad_norm": 0.15566784143447876,
      "learning_rate": 3.986873984113841e-07,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8454501442611217,
      "num_tokens": 1036549589.0,
      "step": 32478
    },
    {
      "entropy": 0.41917638247832656,
      "epoch": 2.9882404356501686,
      "grad_norm": 0.14956212043762207,
      "learning_rate": 3.956205722697581e-07,
      "loss": 0.4103,
      "mean_token_accuracy": 0.8696881234645844,
      "num_tokens": 1036581013.0,
      "step": 32479
    },
    {
      "entropy": 0.5485188630409539,
      "epoch": 2.98833244202161,
      "grad_norm": 0.1594991832971573,
      "learning_rate": 3.9255374612813204e-07,
      "loss": 0.5363,
      "mean_token_accuracy": 0.8339180424809456,
      "num_tokens": 1036613325.0,
      "step": 32480
    },
    {
      "entropy": 0.45450872578658164,
      "epoch": 2.9884244483930513,
      "grad_norm": 0.14976361393928528,
      "learning_rate": 3.89486919986506e-07,
      "loss": 0.4435,
      "mean_token_accuracy": 0.857601772993803,
      "num_tokens": 1036645773.0,
      "step": 32481
    },
    {
      "entropy": 0.5655024461448193,
      "epoch": 2.9885164547644925,
      "grad_norm": 0.16473664343357086,
      "learning_rate": 3.8642009384487996e-07,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8247174136340618,
      "num_tokens": 1036678159.0,
      "step": 32482
    },
    {
      "entropy": 0.4739213725551963,
      "epoch": 2.9886084611359336,
      "grad_norm": 0.15590094029903412,
      "learning_rate": 3.833532677032539e-07,
      "loss": 0.4618,
      "mean_token_accuracy": 0.8537748456001282,
      "num_tokens": 1036709610.0,
      "step": 32483
    },
    {
      "entropy": 0.5685183797031641,
      "epoch": 2.9887004675073747,
      "grad_norm": 0.16482920944690704,
      "learning_rate": 3.802864415616279e-07,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8289441429078579,
      "num_tokens": 1036741467.0,
      "step": 32484
    },
    {
      "entropy": 0.5019770488142967,
      "epoch": 2.9887924738788163,
      "grad_norm": 0.16034694015979767,
      "learning_rate": 3.772196154200019e-07,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8445442989468575,
      "num_tokens": 1036773624.0,
      "step": 32485
    },
    {
      "entropy": 0.5086040599271655,
      "epoch": 2.9888844802502574,
      "grad_norm": 0.16040977835655212,
      "learning_rate": 3.7415278927837585e-07,
      "loss": 0.4923,
      "mean_token_accuracy": 0.8416010737419128,
      "num_tokens": 1036805592.0,
      "step": 32486
    },
    {
      "entropy": 0.46376107539981604,
      "epoch": 2.9889764866216986,
      "grad_norm": 0.15286992490291595,
      "learning_rate": 3.710859631367498e-07,
      "loss": 0.4533,
      "mean_token_accuracy": 0.857163067907095,
      "num_tokens": 1036837926.0,
      "step": 32487
    },
    {
      "entropy": 0.4949878789484501,
      "epoch": 2.9890684929931397,
      "grad_norm": 0.1534898728132248,
      "learning_rate": 3.6801913699512377e-07,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8476954065263271,
      "num_tokens": 1036869778.0,
      "step": 32488
    },
    {
      "entropy": 0.38793920748867095,
      "epoch": 2.989160499364581,
      "grad_norm": 0.13799548149108887,
      "learning_rate": 3.649523108534977e-07,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8804032988846302,
      "num_tokens": 1036902247.0,
      "step": 32489
    },
    {
      "entropy": 0.5523587029892951,
      "epoch": 2.9892525057360224,
      "grad_norm": 0.1640305519104004,
      "learning_rate": 3.618854847118717e-07,
      "loss": 0.5371,
      "mean_token_accuracy": 0.8294119201600552,
      "num_tokens": 1036934047.0,
      "step": 32490
    },
    {
      "entropy": 0.56776072550565,
      "epoch": 2.9893445121074635,
      "grad_norm": 0.1679437756538391,
      "learning_rate": 3.588186585702457e-07,
      "loss": 0.5558,
      "mean_token_accuracy": 0.8265392407774925,
      "num_tokens": 1036966435.0,
      "step": 32491
    },
    {
      "entropy": 0.5500362385064363,
      "epoch": 2.9894365184789047,
      "grad_norm": 0.16117210686206818,
      "learning_rate": 3.5575183242861966e-07,
      "loss": 0.5396,
      "mean_token_accuracy": 0.8269031532108784,
      "num_tokens": 1036998511.0,
      "step": 32492
    },
    {
      "entropy": 0.6030830880627036,
      "epoch": 2.989528524850346,
      "grad_norm": 0.16925545036792755,
      "learning_rate": 3.526850062869936e-07,
      "loss": 0.5945,
      "mean_token_accuracy": 0.8118065185844898,
      "num_tokens": 1037030811.0,
      "step": 32493
    },
    {
      "entropy": 0.5552235515788198,
      "epoch": 2.989620531221787,
      "grad_norm": 0.1612447053194046,
      "learning_rate": 3.496181801453676e-07,
      "loss": 0.5467,
      "mean_token_accuracy": 0.8284021094441414,
      "num_tokens": 1037062373.0,
      "step": 32494
    },
    {
      "entropy": 0.5030144779011607,
      "epoch": 2.9897125375932285,
      "grad_norm": 0.1611350029706955,
      "learning_rate": 3.4655135400374153e-07,
      "loss": 0.4967,
      "mean_token_accuracy": 0.8448388613760471,
      "num_tokens": 1037094589.0,
      "step": 32495
    },
    {
      "entropy": 0.5601963207591325,
      "epoch": 2.9898045439646697,
      "grad_norm": 0.1629462093114853,
      "learning_rate": 3.434845278621155e-07,
      "loss": 0.5435,
      "mean_token_accuracy": 0.83073665574193,
      "num_tokens": 1037126387.0,
      "step": 32496
    },
    {
      "entropy": 0.619656153023243,
      "epoch": 2.989896550336111,
      "grad_norm": 0.16777384281158447,
      "learning_rate": 3.4041770172048945e-07,
      "loss": 0.6115,
      "mean_token_accuracy": 0.8144576326012611,
      "num_tokens": 1037157541.0,
      "step": 32497
    },
    {
      "entropy": 0.6053085448220372,
      "epoch": 2.989988556707552,
      "grad_norm": 0.1675145924091339,
      "learning_rate": 3.3735087557886347e-07,
      "loss": 0.5908,
      "mean_token_accuracy": 0.8134604878723621,
      "num_tokens": 1037189513.0,
      "step": 32498
    },
    {
      "entropy": 0.4810337023809552,
      "epoch": 2.990080563078993,
      "grad_norm": 0.1516447365283966,
      "learning_rate": 3.342840494372374e-07,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8501156643033028,
      "num_tokens": 1037222000.0,
      "step": 32499
    },
    {
      "entropy": 0.4512067511677742,
      "epoch": 2.9901725694504346,
      "grad_norm": 0.15003550052642822,
      "learning_rate": 3.312172232956114e-07,
      "loss": 0.425,
      "mean_token_accuracy": 0.8621104657649994,
      "num_tokens": 1037253244.0,
      "step": 32500
    },
    {
      "entropy": 0.5768688470125198,
      "epoch": 2.9902645758218758,
      "grad_norm": 0.1649358868598938,
      "learning_rate": 3.2815039715398534e-07,
      "loss": 0.5628,
      "mean_token_accuracy": 0.8213451616466045,
      "num_tokens": 1037285255.0,
      "step": 32501
    },
    {
      "entropy": 0.4624253618530929,
      "epoch": 2.990356582193317,
      "grad_norm": 0.15265107154846191,
      "learning_rate": 3.2508357101235936e-07,
      "loss": 0.4559,
      "mean_token_accuracy": 0.8561736047267914,
      "num_tokens": 1037317441.0,
      "step": 32502
    },
    {
      "entropy": 0.4578947069821879,
      "epoch": 2.990448588564758,
      "grad_norm": 0.14729756116867065,
      "learning_rate": 3.220167448707333e-07,
      "loss": 0.4446,
      "mean_token_accuracy": 0.8589144013822079,
      "num_tokens": 1037349883.0,
      "step": 32503
    },
    {
      "entropy": 0.43758129701018333,
      "epoch": 2.990540594936199,
      "grad_norm": 0.14759521186351776,
      "learning_rate": 3.189499187291073e-07,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8664525113999844,
      "num_tokens": 1037381508.0,
      "step": 32504
    },
    {
      "entropy": 0.5367978224530816,
      "epoch": 2.9906326013076407,
      "grad_norm": 0.16238008439540863,
      "learning_rate": 3.1588309258748123e-07,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8353521674871445,
      "num_tokens": 1037413828.0,
      "step": 32505
    },
    {
      "entropy": 0.4409590009599924,
      "epoch": 2.990724607679082,
      "grad_norm": 0.14758096635341644,
      "learning_rate": 3.128162664458552e-07,
      "loss": 0.4247,
      "mean_token_accuracy": 0.8654706366360188,
      "num_tokens": 1037445914.0,
      "step": 32506
    },
    {
      "entropy": 0.4018377379979938,
      "epoch": 2.990816614050523,
      "grad_norm": 0.14356312155723572,
      "learning_rate": 3.0974944030422915e-07,
      "loss": 0.3864,
      "mean_token_accuracy": 0.876817874610424,
      "num_tokens": 1037477634.0,
      "step": 32507
    },
    {
      "entropy": 0.5643242616206408,
      "epoch": 2.990908620421964,
      "grad_norm": 0.16376373171806335,
      "learning_rate": 3.066826141626031e-07,
      "loss": 0.5556,
      "mean_token_accuracy": 0.8259813115000725,
      "num_tokens": 1037509874.0,
      "step": 32508
    },
    {
      "entropy": 0.4620575502049178,
      "epoch": 2.9910006267934053,
      "grad_norm": 0.15229657292366028,
      "learning_rate": 3.0361578802097707e-07,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8572931326925755,
      "num_tokens": 1037541413.0,
      "step": 32509
    },
    {
      "entropy": 0.5012321984395385,
      "epoch": 2.991092633164847,
      "grad_norm": 0.16120034456253052,
      "learning_rate": 3.005489618793511e-07,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8459660559892654,
      "num_tokens": 1037572975.0,
      "step": 32510
    },
    {
      "entropy": 0.4926568267401308,
      "epoch": 2.991184639536288,
      "grad_norm": 0.15445098280906677,
      "learning_rate": 2.9748213573772504e-07,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8472104854881763,
      "num_tokens": 1037605405.0,
      "step": 32511
    },
    {
      "entropy": 0.6311904875328764,
      "epoch": 2.991276645907729,
      "grad_norm": 0.171850323677063,
      "learning_rate": 2.9441530959609905e-07,
      "loss": 0.6332,
      "mean_token_accuracy": 0.8006501011550426,
      "num_tokens": 1037637647.0,
      "step": 32512
    },
    {
      "entropy": 0.5601234700297937,
      "epoch": 2.9913686522791703,
      "grad_norm": 0.1633949726819992,
      "learning_rate": 2.91348483454473e-07,
      "loss": 0.541,
      "mean_token_accuracy": 0.8293094523251057,
      "num_tokens": 1037668907.0,
      "step": 32513
    },
    {
      "entropy": 0.5011802716180682,
      "epoch": 2.9914606586506114,
      "grad_norm": 0.15292884409427643,
      "learning_rate": 2.88281657312847e-07,
      "loss": 0.4823,
      "mean_token_accuracy": 0.8433573320508003,
      "num_tokens": 1037701050.0,
      "step": 32514
    },
    {
      "entropy": 0.5212602252140641,
      "epoch": 2.991552665022053,
      "grad_norm": 0.1619303673505783,
      "learning_rate": 2.8521483117122093e-07,
      "loss": 0.516,
      "mean_token_accuracy": 0.8377443440258503,
      "num_tokens": 1037732840.0,
      "step": 32515
    },
    {
      "entropy": 0.4585135008674115,
      "epoch": 2.991644671393494,
      "grad_norm": 0.15441235899925232,
      "learning_rate": 2.821480050295949e-07,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8553984314203262,
      "num_tokens": 1037763465.0,
      "step": 32516
    },
    {
      "entropy": 0.5172540536150336,
      "epoch": 2.9917366777649352,
      "grad_norm": 0.15850549936294556,
      "learning_rate": 2.7908117888796885e-07,
      "loss": 0.5038,
      "mean_token_accuracy": 0.8399342969059944,
      "num_tokens": 1037795388.0,
      "step": 32517
    },
    {
      "entropy": 0.4056778568774462,
      "epoch": 2.9918286841363764,
      "grad_norm": 0.13648779690265656,
      "learning_rate": 2.760143527463428e-07,
      "loss": 0.3835,
      "mean_token_accuracy": 0.8746361806988716,
      "num_tokens": 1037827265.0,
      "step": 32518
    },
    {
      "entropy": 0.458367726765573,
      "epoch": 2.9919206905078175,
      "grad_norm": 0.15138812363147736,
      "learning_rate": 2.7294752660471677e-07,
      "loss": 0.4449,
      "mean_token_accuracy": 0.8584590181708336,
      "num_tokens": 1037859886.0,
      "step": 32519
    },
    {
      "entropy": 0.564876414835453,
      "epoch": 2.992012696879259,
      "grad_norm": 0.16065166890621185,
      "learning_rate": 2.6988070046309073e-07,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8245764970779419,
      "num_tokens": 1037891588.0,
      "step": 32520
    },
    {
      "entropy": 0.4600023290840909,
      "epoch": 2.9921047032507,
      "grad_norm": 0.15312615036964417,
      "learning_rate": 2.668138743214647e-07,
      "loss": 0.4462,
      "mean_token_accuracy": 0.8585522919893265,
      "num_tokens": 1037923188.0,
      "step": 32521
    },
    {
      "entropy": 0.5307360012084246,
      "epoch": 2.9921967096221413,
      "grad_norm": 0.1624392867088318,
      "learning_rate": 2.637470481798387e-07,
      "loss": 0.5225,
      "mean_token_accuracy": 0.8326991610229015,
      "num_tokens": 1037955437.0,
      "step": 32522
    },
    {
      "entropy": 0.5113668564008549,
      "epoch": 2.9922887159935825,
      "grad_norm": 0.15242670476436615,
      "learning_rate": 2.6068022203821266e-07,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8414390310645103,
      "num_tokens": 1037987715.0,
      "step": 32523
    },
    {
      "entropy": 0.4935246976092458,
      "epoch": 2.9923807223650236,
      "grad_norm": 0.1616232544183731,
      "learning_rate": 2.576133958965866e-07,
      "loss": 0.4859,
      "mean_token_accuracy": 0.8451517857611179,
      "num_tokens": 1038019084.0,
      "step": 32524
    },
    {
      "entropy": 0.43808496720157564,
      "epoch": 2.992472728736465,
      "grad_norm": 0.14740405976772308,
      "learning_rate": 2.5454656975496063e-07,
      "loss": 0.4324,
      "mean_token_accuracy": 0.861184936016798,
      "num_tokens": 1038051366.0,
      "step": 32525
    },
    {
      "entropy": 0.4897879003547132,
      "epoch": 2.9925647351079063,
      "grad_norm": 0.1518833041191101,
      "learning_rate": 2.514797436133346e-07,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8491738215088844,
      "num_tokens": 1038083699.0,
      "step": 32526
    },
    {
      "entropy": 0.5617822166532278,
      "epoch": 2.9926567414793475,
      "grad_norm": 0.16409268975257874,
      "learning_rate": 2.4841291747170855e-07,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8266457244753838,
      "num_tokens": 1038115350.0,
      "step": 32527
    },
    {
      "entropy": 0.4918698398396373,
      "epoch": 2.9927487478507886,
      "grad_norm": 0.15508981049060822,
      "learning_rate": 2.453460913300825e-07,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8472187221050262,
      "num_tokens": 1038147146.0,
      "step": 32528
    },
    {
      "entropy": 0.383085492067039,
      "epoch": 2.9928407542222297,
      "grad_norm": 0.13845953345298767,
      "learning_rate": 2.4227926518845647e-07,
      "loss": 0.3676,
      "mean_token_accuracy": 0.8817590847611427,
      "num_tokens": 1038179299.0,
      "step": 32529
    },
    {
      "entropy": 0.42966079618781805,
      "epoch": 2.9929327605936713,
      "grad_norm": 0.15027639269828796,
      "learning_rate": 2.3921243904683043e-07,
      "loss": 0.423,
      "mean_token_accuracy": 0.8634446486830711,
      "num_tokens": 1038211532.0,
      "step": 32530
    },
    {
      "entropy": 0.45595214422792196,
      "epoch": 2.9930247669651124,
      "grad_norm": 0.15411965548992157,
      "learning_rate": 2.361456129052044e-07,
      "loss": 0.4413,
      "mean_token_accuracy": 0.8571213334798813,
      "num_tokens": 1038243025.0,
      "step": 32531
    },
    {
      "entropy": 0.4032047837972641,
      "epoch": 2.9931167733365536,
      "grad_norm": 0.14654335379600525,
      "learning_rate": 2.330787867635784e-07,
      "loss": 0.3895,
      "mean_token_accuracy": 0.873693935573101,
      "num_tokens": 1038273863.0,
      "step": 32532
    },
    {
      "entropy": 0.47856413573026657,
      "epoch": 2.9932087797079947,
      "grad_norm": 0.15280979871749878,
      "learning_rate": 2.3001196062195236e-07,
      "loss": 0.4706,
      "mean_token_accuracy": 0.8503470942378044,
      "num_tokens": 1038306022.0,
      "step": 32533
    },
    {
      "entropy": 0.5457407347857952,
      "epoch": 2.993300786079436,
      "grad_norm": 0.1578868329524994,
      "learning_rate": 2.2694513448032632e-07,
      "loss": 0.5407,
      "mean_token_accuracy": 0.8305842652916908,
      "num_tokens": 1038338364.0,
      "step": 32534
    },
    {
      "entropy": 0.5068913102149963,
      "epoch": 2.9933927924508774,
      "grad_norm": 0.1527256965637207,
      "learning_rate": 2.238783083387003e-07,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8445498235523701,
      "num_tokens": 1038370557.0,
      "step": 32535
    },
    {
      "entropy": 0.4411937119439244,
      "epoch": 2.9934847988223185,
      "grad_norm": 0.15043559670448303,
      "learning_rate": 2.2081148219707426e-07,
      "loss": 0.4262,
      "mean_token_accuracy": 0.8611917346715927,
      "num_tokens": 1038402344.0,
      "step": 32536
    },
    {
      "entropy": 0.5175738674588501,
      "epoch": 2.9935768051937597,
      "grad_norm": 0.16315248608589172,
      "learning_rate": 2.1774465605544822e-07,
      "loss": 0.4933,
      "mean_token_accuracy": 0.8400431722402573,
      "num_tokens": 1038433659.0,
      "step": 32537
    },
    {
      "entropy": 0.5388886630535126,
      "epoch": 2.993668811565201,
      "grad_norm": 0.16030704975128174,
      "learning_rate": 2.1467782991382218e-07,
      "loss": 0.5181,
      "mean_token_accuracy": 0.8367697857320309,
      "num_tokens": 1038466064.0,
      "step": 32538
    },
    {
      "entropy": 0.45815816475078464,
      "epoch": 2.993760817936642,
      "grad_norm": 0.14799827337265015,
      "learning_rate": 2.1161100377219614e-07,
      "loss": 0.4421,
      "mean_token_accuracy": 0.8589058183133602,
      "num_tokens": 1038498336.0,
      "step": 32539
    },
    {
      "entropy": 0.5313301337882876,
      "epoch": 2.9938528243080835,
      "grad_norm": 0.15383802354335785,
      "learning_rate": 2.0854417763057013e-07,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8368460200726986,
      "num_tokens": 1038530828.0,
      "step": 32540
    },
    {
      "entropy": 0.5360145694576204,
      "epoch": 2.9939448306795247,
      "grad_norm": 0.15896502137184143,
      "learning_rate": 2.054773514889441e-07,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8343473002314568,
      "num_tokens": 1038562769.0,
      "step": 32541
    },
    {
      "entropy": 0.4604075066745281,
      "epoch": 2.994036837050966,
      "grad_norm": 0.1521497666835785,
      "learning_rate": 2.024105253473181e-07,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8567976802587509,
      "num_tokens": 1038594720.0,
      "step": 32542
    },
    {
      "entropy": 0.544972438365221,
      "epoch": 2.994128843422407,
      "grad_norm": 0.16226300597190857,
      "learning_rate": 1.9934369920569206e-07,
      "loss": 0.5359,
      "mean_token_accuracy": 0.829968772828579,
      "num_tokens": 1038626766.0,
      "step": 32543
    },
    {
      "entropy": 0.5589906955137849,
      "epoch": 2.994220849793848,
      "grad_norm": 0.16654044389724731,
      "learning_rate": 1.9627687306406602e-07,
      "loss": 0.5411,
      "mean_token_accuracy": 0.827381145209074,
      "num_tokens": 1038658146.0,
      "step": 32544
    },
    {
      "entropy": 0.5228387527167797,
      "epoch": 2.9943128561652896,
      "grad_norm": 0.15996070206165314,
      "learning_rate": 1.9321004692243998e-07,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8375005461275578,
      "num_tokens": 1038690594.0,
      "step": 32545
    },
    {
      "entropy": 0.4629976721480489,
      "epoch": 2.9944048625367308,
      "grad_norm": 0.15012411773204803,
      "learning_rate": 1.9014322078081394e-07,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8537956774234772,
      "num_tokens": 1038723262.0,
      "step": 32546
    },
    {
      "entropy": 0.5279366634786129,
      "epoch": 2.994496868908172,
      "grad_norm": 0.15618379414081573,
      "learning_rate": 1.8707639463918792e-07,
      "loss": 0.5109,
      "mean_token_accuracy": 0.8339962363243103,
      "num_tokens": 1038755259.0,
      "step": 32547
    },
    {
      "entropy": 0.5671944040805101,
      "epoch": 2.994588875279613,
      "grad_norm": 0.16662000119686127,
      "learning_rate": 1.8400956849756188e-07,
      "loss": 0.5547,
      "mean_token_accuracy": 0.8262070268392563,
      "num_tokens": 1038787254.0,
      "step": 32548
    },
    {
      "entropy": 0.5727398842573166,
      "epoch": 2.994680881651054,
      "grad_norm": 0.1598084270954132,
      "learning_rate": 1.8094274235593584e-07,
      "loss": 0.5602,
      "mean_token_accuracy": 0.8208267539739609,
      "num_tokens": 1038819741.0,
      "step": 32549
    },
    {
      "entropy": 0.5299432007595897,
      "epoch": 2.9947728880224957,
      "grad_norm": 0.1576143354177475,
      "learning_rate": 1.7787591621430983e-07,
      "loss": 0.5263,
      "mean_token_accuracy": 0.833133514970541,
      "num_tokens": 1038851516.0,
      "step": 32550
    },
    {
      "entropy": 0.502737844362855,
      "epoch": 2.994864894393937,
      "grad_norm": 0.15251940488815308,
      "learning_rate": 1.748090900726838e-07,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8436849191784859,
      "num_tokens": 1038883793.0,
      "step": 32551
    },
    {
      "entropy": 0.6512576807290316,
      "epoch": 2.994956900765378,
      "grad_norm": 0.17518053948879242,
      "learning_rate": 1.7174226393105775e-07,
      "loss": 0.641,
      "mean_token_accuracy": 0.8011804558336735,
      "num_tokens": 1038915568.0,
      "step": 32552
    },
    {
      "entropy": 0.44325100909918547,
      "epoch": 2.995048907136819,
      "grad_norm": 0.14696884155273438,
      "learning_rate": 1.6867543778943173e-07,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8605886623263359,
      "num_tokens": 1038948336.0,
      "step": 32553
    },
    {
      "entropy": 0.5668432172387838,
      "epoch": 2.9951409135082603,
      "grad_norm": 0.16381089389324188,
      "learning_rate": 1.656086116478057e-07,
      "loss": 0.5575,
      "mean_token_accuracy": 0.8242855444550514,
      "num_tokens": 1038981104.0,
      "step": 32554
    },
    {
      "entropy": 0.5217581456527114,
      "epoch": 2.995232919879702,
      "grad_norm": 0.16281262040138245,
      "learning_rate": 1.6254178550617968e-07,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8419105187058449,
      "num_tokens": 1039012438.0,
      "step": 32555
    },
    {
      "entropy": 0.4782115323469043,
      "epoch": 2.995324926251143,
      "grad_norm": 0.15541908144950867,
      "learning_rate": 1.5947495936455364e-07,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8501060754060745,
      "num_tokens": 1039044326.0,
      "step": 32556
    },
    {
      "entropy": 0.5458064679987729,
      "epoch": 2.995416932622584,
      "grad_norm": 0.16566140949726105,
      "learning_rate": 1.564081332229276e-07,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8336336761713028,
      "num_tokens": 1039075722.0,
      "step": 32557
    },
    {
      "entropy": 0.5797093529254198,
      "epoch": 2.9955089389940253,
      "grad_norm": 0.17091773450374603,
      "learning_rate": 1.5334130708130156e-07,
      "loss": 0.5666,
      "mean_token_accuracy": 0.8207139372825623,
      "num_tokens": 1039107689.0,
      "step": 32558
    },
    {
      "entropy": 0.46307712816633284,
      "epoch": 2.9956009453654664,
      "grad_norm": 0.15370526909828186,
      "learning_rate": 1.5027448093967554e-07,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8573250211775303,
      "num_tokens": 1039139990.0,
      "step": 32559
    },
    {
      "entropy": 0.574143236503005,
      "epoch": 2.995692951736908,
      "grad_norm": 0.166622132062912,
      "learning_rate": 1.4720765479804953e-07,
      "loss": 0.5601,
      "mean_token_accuracy": 0.8256207220256329,
      "num_tokens": 1039172253.0,
      "step": 32560
    },
    {
      "entropy": 0.5849631559103727,
      "epoch": 2.995784958108349,
      "grad_norm": 0.16948184370994568,
      "learning_rate": 1.441408286564235e-07,
      "loss": 0.5716,
      "mean_token_accuracy": 0.8230915926396847,
      "num_tokens": 1039204262.0,
      "step": 32561
    },
    {
      "entropy": 0.5466214530169964,
      "epoch": 2.9958769644797902,
      "grad_norm": 0.16961687803268433,
      "learning_rate": 1.4107400251479745e-07,
      "loss": 0.5367,
      "mean_token_accuracy": 0.8291136063635349,
      "num_tokens": 1039235680.0,
      "step": 32562
    },
    {
      "entropy": 0.49312388175167143,
      "epoch": 2.9959689708512314,
      "grad_norm": 0.1580779254436493,
      "learning_rate": 1.380071763731714e-07,
      "loss": 0.4844,
      "mean_token_accuracy": 0.8460257761180401,
      "num_tokens": 1039267301.0,
      "step": 32563
    },
    {
      "entropy": 0.4128842381760478,
      "epoch": 2.9960609772226725,
      "grad_norm": 0.14764712750911713,
      "learning_rate": 1.3494035023154536e-07,
      "loss": 0.4181,
      "mean_token_accuracy": 0.8680311627686024,
      "num_tokens": 1039299135.0,
      "step": 32564
    },
    {
      "entropy": 0.5731050679460168,
      "epoch": 2.996152983594114,
      "grad_norm": 0.1644684225320816,
      "learning_rate": 1.3187352408991935e-07,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8198948614299297,
      "num_tokens": 1039331572.0,
      "step": 32565
    },
    {
      "entropy": 0.49408612214028835,
      "epoch": 2.996244989965555,
      "grad_norm": 0.15806028246879578,
      "learning_rate": 1.288066979482933e-07,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8437324464321136,
      "num_tokens": 1039363052.0,
      "step": 32566
    },
    {
      "entropy": 0.47649630066007376,
      "epoch": 2.9963369963369964,
      "grad_norm": 0.15427705645561218,
      "learning_rate": 1.257398718066673e-07,
      "loss": 0.462,
      "mean_token_accuracy": 0.8526766113936901,
      "num_tokens": 1039393829.0,
      "step": 32567
    },
    {
      "entropy": 0.4876908613368869,
      "epoch": 2.9964290027084375,
      "grad_norm": 0.15040777623653412,
      "learning_rate": 1.2267304566504126e-07,
      "loss": 0.4844,
      "mean_token_accuracy": 0.8466574586927891,
      "num_tokens": 1039426293.0,
      "step": 32568
    },
    {
      "entropy": 0.4633932542055845,
      "epoch": 2.9965210090798786,
      "grad_norm": 0.1544238179922104,
      "learning_rate": 1.1960621952341521e-07,
      "loss": 0.4481,
      "mean_token_accuracy": 0.8561103828251362,
      "num_tokens": 1039458426.0,
      "step": 32569
    },
    {
      "entropy": 0.4465142614208162,
      "epoch": 2.99661301545132,
      "grad_norm": 0.1468755155801773,
      "learning_rate": 1.165393933817892e-07,
      "loss": 0.4402,
      "mean_token_accuracy": 0.8586387187242508,
      "num_tokens": 1039490155.0,
      "step": 32570
    },
    {
      "entropy": 0.5133038926869631,
      "epoch": 2.9967050218227613,
      "grad_norm": 0.15211813151836395,
      "learning_rate": 1.1347256724016316e-07,
      "loss": 0.5105,
      "mean_token_accuracy": 0.8422166705131531,
      "num_tokens": 1039522619.0,
      "step": 32571
    },
    {
      "entropy": 0.6132980640977621,
      "epoch": 2.9967970281942025,
      "grad_norm": 0.176359623670578,
      "learning_rate": 1.1040574109853713e-07,
      "loss": 0.6059,
      "mean_token_accuracy": 0.8085598945617676,
      "num_tokens": 1039553224.0,
      "step": 32572
    },
    {
      "entropy": 0.5359079893678427,
      "epoch": 2.9968890345656436,
      "grad_norm": 0.15525488555431366,
      "learning_rate": 1.0733891495691109e-07,
      "loss": 0.5271,
      "mean_token_accuracy": 0.8354080840945244,
      "num_tokens": 1039585701.0,
      "step": 32573
    },
    {
      "entropy": 0.5553355626761913,
      "epoch": 2.9969810409370847,
      "grad_norm": 0.1667434275150299,
      "learning_rate": 1.0427208881528506e-07,
      "loss": 0.544,
      "mean_token_accuracy": 0.8247694410383701,
      "num_tokens": 1039617740.0,
      "step": 32574
    },
    {
      "entropy": 0.4849728960543871,
      "epoch": 2.9970730473085263,
      "grad_norm": 0.15033572912216187,
      "learning_rate": 1.0120526267365905e-07,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8561784885823727,
      "num_tokens": 1039649786.0,
      "step": 32575
    },
    {
      "entropy": 0.5201933146454394,
      "epoch": 2.9971650536799674,
      "grad_norm": 0.15671683847904205,
      "learning_rate": 9.813843653203301e-08,
      "loss": 0.5081,
      "mean_token_accuracy": 0.8401416651904583,
      "num_tokens": 1039681812.0,
      "step": 32576
    },
    {
      "entropy": 0.5700886305421591,
      "epoch": 2.9972570600514086,
      "grad_norm": 0.16642138361930847,
      "learning_rate": 9.507161039040697e-08,
      "loss": 0.5608,
      "mean_token_accuracy": 0.8212116472423077,
      "num_tokens": 1039713553.0,
      "step": 32577
    },
    {
      "entropy": 0.4662871566833928,
      "epoch": 2.9973490664228497,
      "grad_norm": 0.15317252278327942,
      "learning_rate": 9.200478424878094e-08,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8534101285040379,
      "num_tokens": 1039745660.0,
      "step": 32578
    },
    {
      "entropy": 0.5439997753128409,
      "epoch": 2.997441072794291,
      "grad_norm": 0.16222256422042847,
      "learning_rate": 8.893795810715491e-08,
      "loss": 0.5386,
      "mean_token_accuracy": 0.8300264440476894,
      "num_tokens": 1039777590.0,
      "step": 32579
    },
    {
      "entropy": 0.4498083684593439,
      "epoch": 2.9975330791657324,
      "grad_norm": 0.15002478659152985,
      "learning_rate": 8.587113196552887e-08,
      "loss": 0.4417,
      "mean_token_accuracy": 0.8609254211187363,
      "num_tokens": 1039809540.0,
      "step": 32580
    },
    {
      "entropy": 0.45930738863535225,
      "epoch": 2.9976250855371736,
      "grad_norm": 0.14743363857269287,
      "learning_rate": 8.280430582390285e-08,
      "loss": 0.4331,
      "mean_token_accuracy": 0.8604478724300861,
      "num_tokens": 1039841038.0,
      "step": 32581
    },
    {
      "entropy": 0.5693882189225405,
      "epoch": 2.9977170919086147,
      "grad_norm": 0.15961895883083344,
      "learning_rate": 7.973747968227682e-08,
      "loss": 0.562,
      "mean_token_accuracy": 0.8266643323004246,
      "num_tokens": 1039873009.0,
      "step": 32582
    },
    {
      "entropy": 0.4707766892388463,
      "epoch": 2.997809098280056,
      "grad_norm": 0.152934730052948,
      "learning_rate": 7.667065354065078e-08,
      "loss": 0.4599,
      "mean_token_accuracy": 0.8518671542406082,
      "num_tokens": 1039905039.0,
      "step": 32583
    },
    {
      "entropy": 0.49732233211398125,
      "epoch": 2.997901104651497,
      "grad_norm": 0.15353447198867798,
      "learning_rate": 7.360382739902476e-08,
      "loss": 0.4757,
      "mean_token_accuracy": 0.846423365175724,
      "num_tokens": 1039936617.0,
      "step": 32584
    },
    {
      "entropy": 0.4879288421943784,
      "epoch": 2.9979931110229385,
      "grad_norm": 0.1495877355337143,
      "learning_rate": 7.053700125739872e-08,
      "loss": 0.463,
      "mean_token_accuracy": 0.8504310362040997,
      "num_tokens": 1039968390.0,
      "step": 32585
    },
    {
      "entropy": 0.43780668172985315,
      "epoch": 2.9980851173943797,
      "grad_norm": 0.1481904238462448,
      "learning_rate": 6.747017511577268e-08,
      "loss": 0.4267,
      "mean_token_accuracy": 0.8651520647108555,
      "num_tokens": 1040001087.0,
      "step": 32586
    },
    {
      "entropy": 0.4674735330045223,
      "epoch": 2.998177123765821,
      "grad_norm": 0.1535358875989914,
      "learning_rate": 6.440334897414666e-08,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8541503511369228,
      "num_tokens": 1040032957.0,
      "step": 32587
    },
    {
      "entropy": 0.3516848888248205,
      "epoch": 2.998269130137262,
      "grad_norm": 0.13588255643844604,
      "learning_rate": 6.133652283252063e-08,
      "loss": 0.3422,
      "mean_token_accuracy": 0.8860380575060844,
      "num_tokens": 1040065282.0,
      "step": 32588
    },
    {
      "entropy": 0.4517837893217802,
      "epoch": 2.998361136508703,
      "grad_norm": 0.1485452502965927,
      "learning_rate": 5.82696966908946e-08,
      "loss": 0.4365,
      "mean_token_accuracy": 0.8619481585919857,
      "num_tokens": 1040096804.0,
      "step": 32589
    },
    {
      "entropy": 0.5070592244155705,
      "epoch": 2.9984531428801446,
      "grad_norm": 0.15603210031986237,
      "learning_rate": 5.5202870549268566e-08,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8425775989890099,
      "num_tokens": 1040129200.0,
      "step": 32590
    },
    {
      "entropy": 0.4931426299735904,
      "epoch": 2.9985451492515858,
      "grad_norm": 0.15894587337970734,
      "learning_rate": 5.213604440764253e-08,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8466703593730927,
      "num_tokens": 1040161118.0,
      "step": 32591
    },
    {
      "entropy": 0.45312025398015976,
      "epoch": 2.998637155623027,
      "grad_norm": 0.15387864410877228,
      "learning_rate": 4.9069218266016505e-08,
      "loss": 0.4469,
      "mean_token_accuracy": 0.856114849448204,
      "num_tokens": 1040192214.0,
      "step": 32592
    },
    {
      "entropy": 0.5182721205055714,
      "epoch": 2.998729161994468,
      "grad_norm": 0.15639516711235046,
      "learning_rate": 4.600239212439047e-08,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8423645906150341,
      "num_tokens": 1040223510.0,
      "step": 32593
    },
    {
      "entropy": 0.5402618023799732,
      "epoch": 2.998821168365909,
      "grad_norm": 0.1601385623216629,
      "learning_rate": 4.293556598276444e-08,
      "loss": 0.5329,
      "mean_token_accuracy": 0.8353582993149757,
      "num_tokens": 1040255651.0,
      "step": 32594
    },
    {
      "entropy": 0.5512095429003239,
      "epoch": 2.9989131747373508,
      "grad_norm": 0.16194838285446167,
      "learning_rate": 3.986873984113841e-08,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8279604278504848,
      "num_tokens": 1040287967.0,
      "step": 32595
    },
    {
      "entropy": 0.510215693153441,
      "epoch": 2.999005181108792,
      "grad_norm": 0.1584617644548416,
      "learning_rate": 3.680191369951238e-08,
      "loss": 0.5085,
      "mean_token_accuracy": 0.8395956940948963,
      "num_tokens": 1040320488.0,
      "step": 32596
    },
    {
      "entropy": 0.4503476994577795,
      "epoch": 2.999097187480233,
      "grad_norm": 0.15084485709667206,
      "learning_rate": 3.373508755788634e-08,
      "loss": 0.4424,
      "mean_token_accuracy": 0.8585429899394512,
      "num_tokens": 1040353256.0,
      "step": 32597
    },
    {
      "entropy": 0.42705095652490854,
      "epoch": 2.999189193851674,
      "grad_norm": 0.15003405511379242,
      "learning_rate": 3.0668261416260314e-08,
      "loss": 0.4222,
      "mean_token_accuracy": 0.8653552532196045,
      "num_tokens": 1040385111.0,
      "step": 32598
    },
    {
      "entropy": 0.5067622396163642,
      "epoch": 2.9992812002231153,
      "grad_norm": 0.15833155810832977,
      "learning_rate": 2.7601435274634283e-08,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8449217192828655,
      "num_tokens": 1040416590.0,
      "step": 32599
    },
    {
      "entropy": 0.4428671661298722,
      "epoch": 2.999373206594557,
      "grad_norm": 0.1485234647989273,
      "learning_rate": 2.4534609133008252e-08,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8644893020391464,
      "num_tokens": 1040448262.0,
      "step": 32600
    },
    {
      "entropy": 0.44809275586158037,
      "epoch": 2.999465212965998,
      "grad_norm": 0.150173157453537,
      "learning_rate": 2.146778299138222e-08,
      "loss": 0.4399,
      "mean_token_accuracy": 0.8566832840442657,
      "num_tokens": 1040479852.0,
      "step": 32601
    },
    {
      "entropy": 0.5706205610185862,
      "epoch": 2.999557219337439,
      "grad_norm": 0.16554294526576996,
      "learning_rate": 1.840095684975619e-08,
      "loss": 0.558,
      "mean_token_accuracy": 0.8221446499228477,
      "num_tokens": 1040511624.0,
      "step": 32602
    },
    {
      "entropy": 0.5464887912385166,
      "epoch": 2.9996492257088803,
      "grad_norm": 0.1625102162361145,
      "learning_rate": 1.5334130708130157e-08,
      "loss": 0.5331,
      "mean_token_accuracy": 0.8294104002416134,
      "num_tokens": 1040543191.0,
      "step": 32603
    },
    {
      "entropy": 0.48829369246959686,
      "epoch": 2.9997412320803214,
      "grad_norm": 0.15551809966564178,
      "learning_rate": 1.2267304566504126e-08,
      "loss": 0.479,
      "mean_token_accuracy": 0.8443508818745613,
      "num_tokens": 1040575295.0,
      "step": 32604
    },
    {
      "entropy": 0.5489434562623501,
      "epoch": 2.999833238451763,
      "grad_norm": 0.1569431573152542,
      "learning_rate": 9.200478424878095e-09,
      "loss": 0.5319,
      "mean_token_accuracy": 0.8297187462449074,
      "num_tokens": 1040607283.0,
      "step": 32605
    },
    {
      "entropy": 0.5856823613867164,
      "epoch": 2.999925244823204,
      "grad_norm": 0.17174363136291504,
      "learning_rate": 6.133652283252063e-09,
      "loss": 0.5775,
      "mean_token_accuracy": 0.8155203238129616,
      "num_tokens": 1040637950.0,
      "step": 32606
    },
    {
      "entropy": 0.5558843291722811,
      "epoch": 3.0,
      "grad_norm": 0.1852726936340332,
      "learning_rate": 3.0668261416260315e-09,
      "loss": 0.5358,
      "mean_token_accuracy": 0.8309442034134498,
      "num_tokens": 1040663286.0,
      "step": 32607
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 32607,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.03191513407551e+19,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}